University of Pardubice

UNIVERSITY OF PARDUBICE
Faculty of Chemical Technology
Department of Theory and Technology of Explosives
NEW TRENDS IN RESEARCH

OF ENERGETIC MATERIALS
PROCEEDINGS OF THE VIII. SEMINAR
CD version
Pardubice, Czech Republic

April 19 - 21, 2005
UNIVERSITY OF PARDUBICE
Faculty of Chemical Technology
Department of Theory & Technology of Explosives
CZ-532 10 Pardubice
PROCEEDINGS
of the eight Seminar
NEW TRENDS IN RESEARCH OF ENERGETIC MATERIALS
CD version
held at the University of Pardubice,

Pardubice, the Czech Republic
April 19 21, 2005
intended as a meeting of students, postgraduate students, university

teachers and young research and development workers concerned
from the whole world
NOTICE
This publication has not been submitted to language corrections
and contributions have not been reviewed.
Formatting of the contributions has been checked by editor.
The only distributor of the present publication is the
Department of Theory & Technology of Explosives, University of Pardubice,
CZ-532 10 Pardubice,
where the publication can be ordered or gained by exchange of similar publications.
Contributions of the Proceedings will be quoted in the Chemical Abstracts.
March 14th, 2005

Close
Editor:
Ji Vgenknecht
Edition 1st, Limited, 80 pcs Original Copies, Unprotected CD-R
Published by the University of Pardubice
Document designed by J. Vgenknecht at DTTX by Adobe Acrobat 5.0.5.0 CZ
University of Pardubice, 2005

ISBN
(for printed version)
part I. and II.
ISBN 80-7194-758-X
Seminar is supported by:

Defence Science & Technology Lab. (DSTL), Sevenoaks
(an Agency of the Ministry of Defence of U.K)
United States Army International Technology Center (Atlantic) - ERO, London
Austin Detonator, Inc., Vsetn,
Indet Safety Systems, Inc., Vsetn,
OZM Research, Ltd,., Hrochv Tnec
CZ Team, Ltd,., Praha
Chairman of the Seminar:

Prof. Svatopluk Zeman, D.Sc.
Scientific Committee:
Chairman of the Committee:
Dr. Adam Cumming
(DSTL, Sevenoaks, U.K.)
Members of the Committee:

Prof. Ang How-Ghee
Dr. Anthony J. Bellamy
Dr. Stanislaw Cudzilo
Prof. Mikhail A. Ilyushin
Prof. Thomas Klapoetke
Prof. Michel Lefevre
Prof. Frantiek Ludvk
Prof. Andrzej Maranda
Assoc. Prof. Yuanjie SHU
Dr. Muhamed Sueska
Dr. Allen Tulius
Prof. Waldemar A. Trzciski
Assoc. Prof. Pavel Vvra
Prof. Heming XIAO
Dr. Fred Volk
Prof. Ian G. Wallace
Dr. Woodward Waesche
(National Univ. of Singapore)

(Cranfield Univ, UK)
(Military Univ. Technol., Warsaw, Poland)
(St. Petersburg State Inst. of Technol., Russia)
(Ludwig-Maximilians-Universitt Mnchen)
(Royal Military Academy, Belgium)
(Military Academy Brno, Czech Rep.)
(Inst. Chem. Materials, CAEP, Mian Yang, PRC)
(Brodarski Inst., Zagreb, Croatia)
(Applied Res. Assoc., Inc. USA)
(Univ. of Pardubice, Czech Rep.)
(Nanjing Univ. of Sci. & Technol., PRC)
(ICT Pfinztal, Germany)
(Cranfield Univ, UK)
(Office of Naval Res. Int. Field Office, USA)
Organizing Committee
Chairman of the Committee:
Dr. Ji Vgenknecht
(Univ. Pardubice)
Members of the Committee:

Assoc. Prof. Ladislav Lehk
Dr. Jan Jakubko
Dr. Marcela Jungov
Dr. Pavel Valenta
Dr. Iva Ulbrichov
(Explosia, Ltd., Pardubice)

(Indet Safety Systems, Vsetn)
(DTTX, Univ. of Pardubice)
(Austin Detonator, Vsetn)
(Dean Office, University of Pardubice)
CONTENT
PROCEEDING - CD VERSION
PREFACE
16
SPECIAL PRESENTATIONS
AUSTIN DETONATOR Ltd., VSETN
18
CZ TEAM Ltd. UXO CLEARANCE IN CZECH REPUBLIC, PRAHA
19
EXPLOSIA a.s., PARDUBICE SEMTN
21
OZM RESEARCH, Ltd. , HROCHUV TYNEC
23
CEJEM
27
INVITED LECTURES
Ronald W. Armstrong
Center for Energetic Concepts Development, University of Maryland, College Park, MD 20742, U.S.A.
DISLOCATION - ASSISTED INITIATION OF ENERGETIC MATERIALS
32
Heming Xiao*, Zhaoxu Chen**, and Xiaojuan Xu*

* Department of Chemistry, Nanjing University of Science and Technology, Nanjing 210094, P.R.China
** Institute of Theoretical and Computational Chemistry, Nanjing University, Nanjing 210093, P.R.China
A QUANTUM CHEMICAL STUDY ON THERMOLYSIS INITIATION MECHANISMS

AND IMPACT SENSITIVITY OF ENERGETIC MATERIALS
48
Hans J. Pasman*, Richard Bouma**, Nikolaos Zarzalis***, and Max Weiss***

* Delft Uni of Technology, TNW, MSP, Kramers Lab. Prins Bernardlaan 6, 2628 BW, Delft, NL
** TNO Defense, Security and Safety,, PO Box 45, 2280 AA Rijswijk, The Netherlands
*** University Karlsruhe, Engler-Bunte-Institute, 76131 Karlsruhe, Germany
SOME NOTES ON THE FUNDAMENTALS OF E.M. INITIATION
Towards a UNIVERSAL SENSITIVITY CHARACTERISTIC?
63
Thomas D. (Tommy) Sewell

Theoretical Division, Explosives and Organic Materials Group,
Los Alamos National Laboratory, Los Alamos, New Mexico 87545, USA
ATOMISTIC STUDIES OF FUNDAMENTAL PROPERTIES AND PROCESSES IN ENERGETIC

MATERIALS: RELEVANCE TO MESOSCALE INITIATION PHENOMENA
74
Allen J. Tulis
Applied Research Associates, Inc., Chicago, IL, U.S.A.
HIGH-TEMPERATURE PROPELLANT TORCH-SYSTEM FOR NON-DETONATIVE

NEUTRALIZATION OF MINES; SOME PHYSICAL AND CHEMICAL ASPECTS
91
F. Volk
Fraunhofer Institut fr Chemische Technologie, ICT, Postfach 1240, D-76318 Pfinztal-Berghausen, Germany
HIGHLIGHTS OF STABILITY RESEARCH IN THE FIRST DOZEN

OF JAN HANSSON SYMPOSIA 1967 2001
100
Janusz Wrzesiski, Jadwiga Popawska-Jach, and Andrzej Koaczkowski

Wroclaw University of Technology, Wybrzeze Wyspianskiego 27, 50-372 Wroclaw, PL
LIABILITY OF AMMONIUM NITRATE EMULSIONS TO SPONTANEOUS DECOMPOSITION
119
LECTURES
Eugeniya A. Bakhmatova* , Tatyana V. Petukhova**, Vyacheslav L. Korolev*, Tatyana S. Pivina*,
and Victor P. Ivshin**
* Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Leninsky Prospect 47, Moscow 119991, RUSSIA
** Mari State University, Lenin square 1, Yoshkar-Ola 424000, Mari El Republic, RUSSIA
COMPUTER SIMULATION OF NITRO-1,2,4-TRIAZOLES DECOMPOSITION REACTIONS
130
Anthony J. Bellamy*, Alessandro E. Contini*, Peter Golding**, and Stephen J. Trussell**

* Cranfield University, Royal Military College of Science, Shrivenham, Swindon SN6 8LA, UK
** AWE Aldermaston, Reading RG7 4PR, UK
BOMB CALORIMETRIC STUDY

OF A SERIES OF ENERGETIC LINEAR POLYPHOSPHAZENES
138
Jan Bdek, Stanisaw Cudzio, and Sylwia Pietrzyk

Institute of Chemistry, Military University of Technology, 00-908 Warsaw, POLAND
DETERMINATION OF POST-EXPLOSION RESIDUES

OF AMMONIUM NITRATE IN ENVIRONMENTAL SAMPLES
146
Manfred A. Bohn, Manuela Drich, Jasmin Aniol, Heike Pontius, and Peter Gerber
Fraunhofer-Institut fr Chemische Technologie (ICT), Postfach 1240, D-76318 Pfinztal-Berghausen, Germany
REACTIVITY BETWEEN -CL20-GAP AND -HMX-GAP INVESTIGATED BY MASS LOSS,

ADIABATIC SELF HEATING AND DYNAMIC MECHANICAL ANALYSIS
151
A.G.Anshits*, N.N.Anshits*, A.A.Deribas**, S.M.Karakhanov**, N.S.Kasatkina***, A.V.Plastinin**,

A.Yu. Reshetnyak****, V.V.Silvestrov**
* Institute of Chemistry and Chemical Technology SB RAS, Krasnoyarsk
** Lavrentyev Institute of Hydrodynamics SB RAS Novosibirsk
*** Boreskov Institute of Catalysis SB RAS Novosibirsk
**** Institute of Theoretical and Applied Mechanics SB RAS Novosibirsk
DETONATION VELOCITY OF EMULSION EXPLOSIVES WITH CENOSPHERES
179
Stefan Ek*, Carina Eldster*, Patrick Goede*, Erik Holmgren*, Rolf Tryman*, Nikolaj Latypov*,
Yang Guo Ying Raymond**, and Lee Yiew Wang**
* FOI, Swedish Defence Research Agency, Department of Energetic Materials, S-147 25 Tumba, Sweden
** DSO National Laboratories, 20 Science Park, Singapore 118230
SYNTHESIS AND CHARACTERISATION

OF 2,2-DINITRO-1,3-PROPANEDIOL-BASED PLASTICISERS
184
G. Fischer*, G. Holl**, T. M. Klaptke*, P. Mayer*, J. J. Weigand*

* Contribution from the Chair of Inorganic Chemistry, Ludwig-Maximilian University of Munich,
D-81377 Munich, Germany
** Bundeswehr Research Institute for Materials, Fuels and Lubricants, Swisttal-Heimerzheim, Groes Cent,
D-53913 Swisttal, Germany
PROPERTIES OF A HIGHLY FRICTION SENSITIVE DERIVATIVE OF 1,5-DIAMINO-1HTETRAZOLE (DAT): 1,5-DIAMINO-4-METHYLTETRAZOLIUM DINITRAMIDE
194
Alexander Gromov*, Ulrich Frter-Barth**, and Ulrich Teipel**

* Tomsk Polytechnic University, 30, Lenin Ave., 634050, Tomsk, Russia
** Fraunhofer Institute for Chemical Technology, P.O. Box 1240, D-76318 Pfinztal (Berghausen), Germany
STUDY OF NON-ISOTHERMAL NITRIDATION OF ALUMINUM NANOPOWDERS PASSIVATED

BY NON-OXIDE LAYERS
204
Zbigniew Chyek, Stanisaw Cudzio, Jan Bdek, and Sylwia Pietrzyk
Institute of Chemistry, Military University of Technology 00-908 Warsaw, POLAND
OPTIMIZATION OF 1,1-DIAMINO-2,2-DINITROETHENE SYNTHESIS
211
Mikhail A. Ilyushin, and Igor V. Tselinskii

Saint-Petersburg State Institute of Technology, Saint-Petersburg, Moskovsky pr. 26
THE INFLUENCE OF THE STRUCTURE OF THE SALTS OF AZOLES

UPON THE PROCESSES OF THEIR THERMAL AND LASER INITIATION
217
Guy Jacob*, Stphane Bnazet*, Rolf Tryman**, Patrick Goede**, and Henric stmark**
* SME Research Centre, 9 Rue Lavoisier, 91710 Vert le Petit, France
** FOI, Swedish Defence Research Agency, Department of Energetic Materials, S-147 25 Tumba, Sweden.
CALCULATIONS OF DENSITIES AND HEATS OF FORMATIONS OF ENERGETIC MOLECULES

FOR THE USE IN THERMOCHEMICAL CODES
226
L. Jeunieau*, M.H. Lefebvre*, M.C. Pirlot**, and P. Guillaume***
* Laboratory for Energetic Materials, Ecole Royale Militaire, Avenue de la Renaissance 30, 1000 Bruxelles (Belgium)
** Department of Weapon Systems and Ballistics, Ecole Royale Militaire,Avenue de la Renaissance 30,
1000 Bruxelles (Belgium)
*** PB Clermont SA, Rue de Clermont 176, 4480 Engis, (Belgium)
CHARACTERIZATION OF DETERRED PROPELLANTS BY CLOSED VESSEL TESTS:

IMPORTANCE OF THE IGNITION METHOD
236
Xiao Jijun, Ma Xiufang, Zhu Wei, Huang Yucheng, and Xiao Heming
Chemical Engineering Institute, Nanjing University of Science and Technology, Nanjing, 210094, P. R. China
A MOLeCULAR DYNAMICS SIMULATION STUDY

OF ELASTIC PROPERTIES OF HMX-BASED AND TATB-BASED PBXS
249
Lu Chunxu
Nanjing University of Science and Technology,Nanjing210094,China
THE APPLICATION OF A SURFACE ACTIVE THEORY TO ENERGETIC MATERIALS RESEARCH ON EXPANSION AMMONIUM NITRATE EXPLOSIVE
260
Alexander N. Lukin
Physics-Chemical Mechanics Department, Institute of Applied Mechanics,
Ural Branch of the Russian Academy of Sciences, Bldg. 34, T.Baramzinoy Str., 426067, Izhevsk, Udmurtia Republic,
Russian Federation
CONCEPT OF THE SPATIAL-PERIODIC MICRO-STRUCTURES EXCITATION AT THE

EVAPORATED ENERGETIC MATERIALS TRANSIENT COMBUSTION AS A WAY FOR
IMPROVEMENT OF THE SOLID PROPULSION TECHNOLOGIES
271
Sreekumar Pisharath and Ang How Ghee

Energetic Materials Research Centre, School of Materials Science &Engineering, Nanyang Technological University,
50, Nanyang Avenue, Singapore 639 798
DECOMPOSITION KINETICS OF GAP BINDER

IN THE PRESENCE OF AN ENERGETIC COMPONENT
300
Miroslav Pospil* and Pavel Vvra**

* Charles University Prague, Faculty of Mathematics and Physics,
Department of Chemical Physics and Optics, Ke Karlovu 3, 12116 Prague 2, Czech Republic
** University of Pardubice, Faculty of Chemical Technology, Department of Theory and Technology of Explosives,
Studentsk 95, 53210 Pardubice, Czech Republic
DECOMPOSITION OF ENERGETIC MATERIALS

INVESTIGATED BY MOLECULAR SIMULATIONS: HNIW, TNA
306
G. Santhosh, S. Venkatachalam, K. Krishnan, K.N. Ninan

Propellants and Special Chemicals Group, Vikram Sarabhai Space Centre, Trivandrum 695 022, INDIA.
THE PHASE STABILIZATION OF AMMONIUM NITRATE

BY POTASSIUM DINITRAMIDE A DIFFERENTIAL SCANNING CALORIMETRIC STUDY
316
Jonas arlauskas*, Aura Nemeikait-nien**, ilvinas Anuseviius*, Henrikas Nivinskas*,

Lina Miseviien*, Valentina Vilutien***, and Narimantas nas*
* Institute of Biochemistry, Mokslinink 12, LT-08669 Vilnius, Lithuania
** Institute of Immunology, Vilnius University, Molt Pl. 29, LT-08409 Vilnius, Lithuania
*** The General J. emaitis Military Academy of Lithuania, ilo 5a, LT-10322 Vilnius, Lithuania
MAMMALIAN CELL CYTOTOXICITY OF NITROAROMATIC EXPLOSIVES

AND THEIR DEGRADATION PRODUCTS: THE ROLE OF OXIDATIVE STRESS
324
SUN Bin
Dept. of Safety Defence, Zhejiang Police Institue, Hangzhou 310018 , China
APPLICATION OF THE METHOD OF GRAY INTERRELATION ANALYZING

TO ASSESS HAZARD OF GAS EXPLOSION ACCIDENTS
334
Ulrich Teipel, and Ulrich Frter-Barth

Fraunhofer Institut fr Chemische Technologie (ICT), Joseph-von-Fraunhofer-Strae 7, 76327 Pfinztal, Germany
DESIGN OF PARTICULATE ENERGETIC MATERIALS
340
Titica Vasile, Cristian Barbu, and Doru Safta

Military Technical Academy, 81-83 George Cosbuc Avenue, Bucharest, Romania
REGARDING THE DETERMINATION OF EXPONENT FROM BURNING RATE LAW OF POWDER,

USING THE CLOSED BOMB DATA
353
Richard Wild
PBX-Center Maasberg, Diehl BGT Defence GmbH & Co KG, Karl-Diehl-Strae 1, D-66620 Nonnweiler, Germany
MAIN CHARGE INITIATION OF INSENSITIVE MUNITIONS BYUNPLANNED STIMULI.

FAST COOK-OFF RESISTANCE OF PRESSED PBX
361
Stephan Wilker, Gabriele Pantel, Gerhard Holl, and Uldis Ticmanis

WIWEB ASt Heimerzheim, Groes Cent, 53913 Swisttal (DE)
FROM BINARY MIXTURES TO COMPLETE AMMUNITION COMPATIBILITY

A WHOLE CARTRIDGE IN ONE TAM AMPOULE
370
Svatopluk Zeman
Department of Theory & Technology of Explosives, University of Pardubice CZ-532 10 Pardubice, Czech Republic
THE STUDY OF CHEMICAL MICRO-MECHANISM OF NITRAMINES

INITIATION WITH UTILIZATION OF 15N NMR CHEMICAL SHIFTS
387
CONTRIBUTIONS PUBLISHED ONLY AS ABSTRACT

D. Clment, J. Diener, and D. Kovalev
TU Muenchen - Physik Department E16, James-Franck-Strasse, 85747 Garching b. Muenchen
HIGHLY EXPLOSIVE NANOSILICON-BASED COMPOSITE MATERIALS
399
Maciej Duda*, and Wincenty Skupiski**

* Department of Highenergetic Materials, Faculty of Chemistry,
Warsaw University of Technology, Noakowskiego 3, 00-664 Warsaw, Poland
** Industrial Chemistry Research Institute, Rydygiera 8, 01-793 Warsaw, Poland
THERMAL STABILITY OF THE IMPURITIES AND THE POLYMORPHS OF HNIW
400
Ernst-Christian Koch, and Karl P. Rudolf

Diehl BGT Defence GmbH & Co. KG, Fischbachstrasse 16, D-90552 Rthenbach a d Pegnitz, Germany
DETONATIVE IGNITION OF METAL/FLUOROCARBON PYROLANTS
401
L.A.Merzhievsky*, P.I.Zubkov*, K.A.Ten*, E.R.Pruuel*, L.A.Lukyanchikov*, V.M.Titov*,

B.P.Tolochko**, M.G.Fedotov**, M.R.Sharafutdinov**, M.A.Sheromov**, I.L.Jogin**, V.V.Julanov***,
L.I.Shehtman***, and V.M.Aulchenko***
* Lavrentyev Institute of Hydrodynamics SB RAS, Novosibirsk, Russia
** Budker Institute of Nuclear Physics SB RAS, Novosibirsk, Russia
** Institute of Solid Chemistry and Mechanochemistry SB RAS, Novosibirsk, Russia
INVESTIGATION OF DEFLAGRATION TO DETONATION TRANSITION

IN POROUS HIGH EXPLOSIVE BY SYNCHROTRON RADIATION APPROACH
402
Karl Rudolf
Diehl BGT Defence GmbH & Co. KG, Fischbachstrae 16, 90552 Rthenbach, Germany
INSENSITIVE INITIATION CHAINS FOR INSENSITIVE EXPLOSIVE FILLS
403
Rudolf S. Stepanov, Ludmila A. Kruglyakova, and Alexander M. Astachov

Siberian State Technological University, prosp. Mira, 82, Krasnoyarsk 660049, Russia
STRUCTURE-KINETIC LAWS
OF THERMAL DECOMPOSITION OF SIX-MEMBERED CYCLIC N-NITRAMINES
404
Alexander Tsoglin* **, and Chagit Denekamp*

* Department of Chemistry and Institute of Catalysis Science and Technology, Technion Israel Institute of Technology,
Haifa 32000, Israel
** Rafael Ltd., Department M4, P.O.B. 2250, Haifa 31021, Israel
MASS SPECTROMETRIC ANALYSIS OF NOVEL NITRAMINE EXPLOSIVES

- HNIW AND DTIW
405
Alexander Tsoglin, and Levi Gotlieb

Rafael Ltd., Department M4, P.O.B. 2250, Haifa 31021, Israel
VACUUM STABILITY TEST (VST) CALIBRATION AND COMPARISON OF MERCURY

AND PRESSURE TRANSDUCER METHODS
406
I. G. Wallace
Cranfield University, Shrivenham Campus, Swindon SN6 8LA, UK
DEVELOPING AND MAINTAINING SKILLS IN THE EXPLOSIVES SECTOR
407
POSTERS
Vladislav Adamk*, and Zenon Wilk**
* Department of Theory and Technology of Explosives (DTTX), University of Pardubice, 532 10 Pardubice,
Czech Republic
** Institute of Organic Industry Branch in Krupski Myn, Zawadzkiego 1 str., 42-693 Krupski Myn, Poland
NUMERICAL SIMULATIONS OF OIL-WELL PERFORATOR

USING 3D LAGRANGIAN LS-DYNA CODE
408
E.I. Aleshkina and G.D. Kozak

Mendeleev University of Chemical Technology, 125047, Miusskaja sq. 9, Moscow, Russia.
PARAMETERS OF HEAT EXPLOSION AND DETONATION OF HEXANITROMANNITE

AND NITROTHREEAZOLONE
415
S.V. Arinina, and G.D. Kozak

Mendeleev University of Chemical Technology 125047, 125047, Miusskaja sq. 9, Moscow, Russia
COMPARISION OF EXPLOSIVE PARAMETERS OF PEROXIDES AND OF NITROCOMPOUNDS 422

Alexander M. Astachov*, Alexander D. Vasiliev**, Maxim S. Molokeev**, Andrew A. Nefedov*,
Ludmila A. Kruglyakova*, Vitaliy A. Revenko*, and Eduard S. Buka*
* Siberian State Technological University, Prosp. Mira 82, 660049 Krasnoyarsk, Russia
** Institute of Physics RAS (Sib. branch), Akademgorodok, 660036 Krasnoyarsk, Russia
COMPARISION OF EXPLOSIVE PARAMETERS OF PEROXIDES

AND OF NITROCOMPOUNDS 2-NITRIMINO-5-NITROHEXAHYDRO-1,3,5-TRIAZINE:
STRUCTURE AND PROPERTIES
430
A.A. Astratv, D.V. Dashko, and A.I.Stepanov

Special Design and Construction Bureau SDCB Technolog of the Saint-Petersburg State Institute of Technology
(technical university), 190013, Russia, Saint-Petersburg, Moskovskiy pr.26
REACTIVITY OF 2-(DINITROMETHYLENE)- 4,6-DIHYDROXY-5,5-DINITROPYRIMIDINE

IN THE PROCESSES OF NUCLEOPHILIC SUBSTITUTION
444
Anthony J. Bellamy
Cranfield University, Royal Military College of Science, Shrivenham, Swindon SN6 8LA, UK
STUDIES ON THE HYDRODENITRATION BY TIN (II) CHLORIDE

OF POLYNITROHEXAAZAISOWURTZITANES
451
Anthony J. Bellamy, Luigi Cassioli and Alessandro E. Contini

SYNTHESIS AND PROPERTIES OF SALTS OF 3,5-DIAMINOPICRIC ACID
459
Anthony J. Bellamy*, Alistair MacCuish* and Peter Golding**

*Cranfield University, Royal Military College of Science, Shrivenham, Swindon SN6 8LA, UK
**AWE Aldermaston, Reading RG7 4PR, UK
THE USE OF THE TRIFLUOROACETYL GROUP TO PROTECT NH AND OH GROUPS

DURING NITROLYSIS REACTIONS
465
Jan Bdek, Sylwia Pietrzyk, Stanisaw Cudzio, and Zbigniew Chyek

TLC ANALYSIS OF DADNE AND SOME INTERMEDIATE PRODUCTS OF ITS SYNTHESIS
472
Michael A. Bohn*, Anton Hammerl**, Kate Harris***, and Thomas M. Klaptke**

* Fraunhofer Institut Chemische Technologie, Pfinztal, Germany
** Chair of Inorganic Chemistry, LMU Munich, Munich, Germany
*** Department of Chemistry, University of Edinburgh, Edinburgh, U.K.
INTERACTIONS BETWEEN THE NITRAMINES RDX, HMX AND CL20

WITH THE ENERGETIC BINDER GAP
10
477
Daniel Buczkowski*, Adam Presz**, and Bogdan Zygmunt***

* Institute of Industrial Organic Chemistry, Annopol St. 6, 03-236, Warszawa, PL
** Institute of High Pressure Physics of PAS, Sokoowska St. 29/37, Warszawa, PL
*** Military Academy of Technology, Kaliskiego St. 2, 00-908 Warszawa, PL
THE INFLUENCE OF PHYSICAL STRUCTURE OF AMMONIUM NITRATE PRILLS

ON DETONATION PROPERTIES OF ANFO-EXPLOSIVES
485
A. Dbkowski*, A. Sapiski**, A. Teodorczyk*, and W. Witkowski***

* Warsaw University of Technology, ITC, Nowowiejska 21/25, 00-665 Warszawa, Poland
** Sapinski Engineering Consulting. Lidzbarska 52, 03-085 Warszawa
*** Institute of Industrial Organic Chemistry, Annopol 6, 03-236 Warszawa, Poland
DETONATION ARRESTER PERFORMANCE FOR HYDROGEN-AIR MIXTURES
495
Moritz v. Denffer*, Gerhard Heeb**, Thomas M. Klaptke*, Gernot Kramer*, Gunnar Spie*,
and Jan M. Welch*
* Chair of Inorganic Chemistry, University of Munich, Butenandtstr. 5-13 (D), D-81377 Munich, Germany
** Bundeswehr Research Institute for Materials, Fuels and Lubricants, Swisttal-Heimerzheim; Groes Cent,
D-53913 Swisttal (Germany)
IMPROVED SYNTHESIS AND X-RAY STRUCTURE OF 5-AMINOTETRAZOLIUM NITRATE
500
Mario Dobrilovi, Zvonimir Ester, and Branimir Jankovi

Faculty of Mining, Geology and Petroleum Engineering, University of Zagreb, Croatia
MEASUREMENT IN BLAST HOLE STEM

AND INFLUENCE OF STEMMING MATERIAL ON BLASTING QUALITY
510
Mario Dobrilovi, Zvonimir Ester, and Trpimir Kujundi

Faculty of Mining, Geology and Petroleum Engineering, University of Zagreb, Croatia
MEASUREMENT OF SHOCK WAVE FORCE IN SHOCK TUBE WITH INDIRECT METHODS
520
I.V. Egorova, and B.N.Kondrikov

Mendeleev University of Chemical Technology, 125047, Miusskaja sq. 9, Moscow, Russia
CHARACTERISTICS OF EXPLOSION-SAFETY OF AIR-METAL DUST SUSPENSIONS
528
Michael Gbel, Thomas M. Klaptke, and Peter Mayer

Department of Chemistry and Biochemistry, Ludwig-Maximilians University of Munich, Butenandtstr. 5 - 13 (Haus D),
CRYSTAL STRUCTURES OF THE SILVER AND POTASSIUM SALTS OF NITROFORM
537
Hu Bingcheng, L Chunxu, Liu Zuliang

School of Chemical Engineering, Nanjing University of Science & Technology, Nanjing 210094, P.R. China
STUDY ON SAFETY OF EXPANDED AMMONIUM NITRATE EXPLOSIVE
539
L. Huskov*, J. rmkov*, K. Ventura*, Z. Aktein**, and M. tancl**

*Department of Analytical Chemistry, Faculty of Chemical Technology, University of Pardubice, nm. s. legi 565, 532
10 Pardubice, CZ
** Research Institute of Industrial Chemistry, Explosia a.s., 532 17 Pardubice-Semtn, CZ
DETERMINATION OF METAL OXIDES AND THEIR CONCENTRATION

IN THE IDENTIFICATION PARTICLES
545
Martina Chovancova, Peter Ocko, Alzbeta Pechova, and Milos Lazar

VTSU (Military technical and Testing Institute) Zahorie, 905 24 Senica, Slovak Republic
RESEARCH OF ARTIFICIAL AGEING SUITABLE CONDITIONS

DURING THE INVESTIGATION OF PROPELLANTS LIFETIME
549
Zdenk Jalov*, Pavel Mareek**, Kamil Dudek**, Ondej Fohl*, Nikolaj V. Latypov***,
Stefan Ek***, and Martin Johansson***
* University of Pardubice, Department of Theory and Technology of Explosives, Studentsk 95, CZ-532 10 Pardubice,
Czech Republic
** Explosia, a.s., CZ-532 17 Pardubice-Semtn, Czech Republic
*** FOI, Swedish Defence Research Agency, Department of Energetic Materials, S-147 25 Tumba, Sweden
IMPROVED SYNTHESIS OF 2-(DINITROMETHYLENE)-4,5-IMIDAZOLIDINEDIONE
11
566
Vra Jeov, Ale Eisner, Jana Chalnkov, and Karel Ventura

Department of Analytical Chemistry, University of Pardubice, 532 10 Pardubice, Czech Republic
ANALYSIS OF NITROAROMATIC COMPOUNDS IN SOIL ENVIRONMENTS
571
T. M. Klaptke*, J. J. Weigand*, and G. Holl**

** Bundeswehr Research Institute for Materials, Fuels and Lubricants, Swisttal-Heimerzheim, Groes Cent,
D-53913 Swisttal, Germany
PROPERTIES OF A HIGHLY FRICTION SENSITIVE DERIVATIVE

OF 1,5-DIAMINO-1H-TETRAZOLE (DAT):
1,5-DIAMINO-4-METHYLTETRAZOLIUM DINITRAMIDE
575
Milan Klusek, and Marcela Jungov

Department of Theory and Technology of Explosives (DTTX), University of Pardubice, 532 10 Pardubice, Czech Republic
DEVELOPMENT OF TESTING METHOD FOR DETERMINATION INITIATION STRENGTH

OF DETONATORS BY MEASUREMENTIN IMPULSE CHAMBER
585
.V. Kostitsyn, B.G. Loboiko, V.P. Filin, .V. Vershinin, .. Nikulin, .B. Smirnov,
and S.N. Lyubyatinsky
Zababakhin Russian Federal Nuclear Centre - VNIITF, P.O. Box 245, Snezhinsk, Chelyabinsk region, 456770 Russia
PROPAGATION OF DETONATION IN CYLINDRICAL LOW-SENSITIVE HE SAMPLES

*
**
595
***
N.V. Kozak , G.D. Kozak , and Zhou Lin
*Academy of labor and Social Relation, Lobachevskogo st. 90, Moscow, 119454.
**Mendeleev University of Chemical Technology Miusskaya Sq.9, Moscow, 125190.
**Beijing Institute of Technology, P O Box 327, Beijing 100081, China.
THE SPIN-PULSATING REGIME OF DETONATION IN SOLID AND LIQUID EXPLOSIVES
600
Miloslav Krupka
Department of Theory and Technology of Explosives, University of Pardubice, Czech Republic
INTRODUCTION TO THERMAL STABILITY STUDIES OF NITROBENZENE

AND NITROPHENOL DERIVATIVES IN SULFURIC ACID
607
Richard Kuracina
COMPARISON OF SMOKELESS NITROCELLULOSE POWDER PRODUCTION ACCIDENTS

IN EXPLOSIA WITH THE WORLDS ACCIDENTS DATABASE
613
Karel Leinweber, and Petr Pchouek

EXPLOSIA a.s., 532 17 Pardubice - Semtn
PROTOTYPA-ZM, s.r.o., Hudcova 533 / 78c, 612 00 Brno
EFFECTIVE AMMUNITION 20x102
623
Katarzyna Lipiska*, Marek Lipiski*, and Andrzej Maranda **

* Institute of Industrial Organic Chemistry, 6 Annopol St., 03-236 Warszawa, PL
** Military University of Technology, 2 Kaliskiego St., 00-908 Warszawa, PL
SOME DETONATION CHARACTERISTICS

OF AMMONIUM NITRATE-ALUMINUM-DOUBLE BASE PROPELLANT MIXTURES
632
Liu Dabin Tang, and Shuangling Ni Ouqi

Nanjing University of Science& Technology, Nanjing, China 210094
INACTIVE MODIFICATION OF AMMONIUM NITRATE AND THE DETERMINATION

OF ITS ACTIVITY AS OXIDIZER IN EXPLOSIVES
639
Ji Majzlk, and Ji Strnad

Department of Theory & Technology of Explosives, University of Pardubice, CZ 532 10 Pardubice, Czech Republic
EFFICIENCY OF APPARATUS FOR TESTING SENSITIVITY OF EM

TO ELECTROSTATIC DISCHARGE
12
643
Andrzej Maranda*, Adam Kubecki*, Jerzy Nowaczewski*, Andrzej Orzechowski**,

and Dorota Powaa**
* Faculty of Mechatronics, Military University of Technology, 00-908 Warsaw, POLAND
* Institute of Industrial Organic Chemistry, 03-236 Warszawa, POLAND
DEVELOPMENT OF INSENSITIVE HIGH EXPLOSIVES NITROGUANIDINE
652
Sanja Matei Muani*, Muhamed Sueska*, Maa Raji Linari*, and Sanko Bakija**
* Brodarski Institute, Av. V. Holjevca 20, 10 020 Zagreb, Croatia
** Ministry of Defence, Bauerova 33, 10000 Zagreb, Croatia
APPLICATION OF TIME-TEMPERATURE SUPERPOSITION PRINCIPLES TO PREDICTION

OF MECHANICAL PROPERTIES OF DOUBLE BASED ROCKET PROPELLANTS
663
Robert Maty
Department of Theory and Technology of Explosives, University of Pardubice, 532 10 Pardubice, Czech Republic
INFLUENCE OF OIL ON SENSITIVITY AND THERMAL STABILITY

OF TRIACETONE TRIPEROXIDE AND HEXAMETHYLENETRIPEROXIDE DIAMINE
674
A. Mouloud
Laboratoire des Systmes Pyrotechniques, UER de Chimie Applique,
B.P 17, EMP, Bordj-El-Bahri, 16111 Algiers, Algeria
EFFECT OF ALUMINIUM POWDER ON THE THERMAL AND ENERGETIC PROPERTIES

OF COMPOSITE ROCKET PROPELLANT BASED EPOXY RESIN
680
Andrzej Orzechowski*, Dorota Powaa *, Andrzej Maranda**, and Wojciech Pawowski***

* Institute of Industrial Organic Chemistry, 6 Annopol St, 03-236 Warsaw, PL
** Military University of Technology, 2 Kaliskiego St, 00-908 Warsaw 49, PL
*** Warsaw University of Technology, 3 Noakowskiego St, 00-665 Warsaw, PL
CRYSTALLIZATION OF HEXANITROSTILBENE
687
Ji Pachm, and Jakub eleovsk

Department of Theory and Technology of Explosives, University of Pardubice, CZ - 530 09 Pardubice
CHARACTERIZATION OF COMPOSITE SOLID ROCKET PROPELLANT USING DMA
695
Jzef Paszula*, Andrzej Maranda*, Andrzej Papliski*, Barbara Gobek**, and Johann Kasperski **
* Military University of Technology, 00-908 Warszawa, ul. Kaliskiego 2, Poland
** Blastexpol, 59-145 Duninw, Poland
AN ANALYSIS OF BLAST WAVES PARAMETERS AND UNDERWATER EXPLOSION TEST

OF EMULSION EXPLOSIVES AND DYNAMITES
708
P. Paunovic*, O. Popovski**, M. Tasev R. Smileski**, and S. Hadzi Jordanov*

* Faculty of Technology and Metallurgy, University Sts. Cyril and Methodius, Skopje, R. Macedonia
** Military Academy "Mihailo Apostolski", Skopje, R. Macedonia
SOPHISTICATED ELECTROCATALYSTS FOR ECONOMICAL PRODUCTION

OF HYDROGEN; PART A: STRUCTURAL CHARACTERIZATION
719
Dmitriy V. Pleshakov
Mendeleyev University, Miusskaya pl. 9, Moscow, 125047 Russia
ESTIMATION OF THE CONDENSATION PROBABILITY OF SUPERSATURATED

NITROGLYCERINE VAPOR DURING MANUFACTURING OF ENERGETIC MATERIALS
730
O. Popovski**, P. Paunovic, R. Smileski* and S. Hadzi Jordanov*

*Faculty of Technology and Metallurgy, University Sts. Cyril and Methodius, Skopje, R. Macedonia
SOPHISTICATED ELECTROCATALYSTS FOR ECONOMICAL PRODUCTION

OF HYDROGEN; PART B: ELECTROCHEMICAL CHARACTERIZATION
739
Pavel Prchal, Jan Zigmund, and Jindich Veverka

Research Institute of Industrial Chemistry, Explosia a.s., 532 17 Pardubice-Semtn, Czech Republic
COMPARISON OF MEASURED AND CALCULATED PARAMETERS OF PROPELLANTS
13
745
Maa Raji Linari, Muhamed Sueska, and Sanja Matei Muani

Brodarski institut Marine Research & Special Technologies, Av. V. Holjevca 20, 10020 Zagreb, Croatia
CHANGES OF SOME THERMOPHYSICAL PROPERTIES OF NITROCELLULOSE PROPELLANTS

DURING ACCELERATED AGING
750
Jakub eleovsk, and Ji Pachm
* Department of Theory and Technology of Explosives, University of Pardubice, CZ-532 10 Pardubice, Czech Republic
ON THE ESTIMATION OF COOKOFF TEMPERATURE USING FEM
757
Galina A. Shraiber, and Vlada M. Raikova

Mendeleev University of Chemical Technology, Miusskaya sq. 9, Moscow A-47, Russia
HAZARD OF EXOTHERMAL REACTIONS IN PRODUCTION OF NITROCOMPOUNDS
764
Yuanjie Shu, Chaoyang Zhang, Xinfeng Wang, Xiaodong Zhao

Institute of Chemical Materials CAEP, 621900, P.O.Box 919-301,Mianyang, Sichuan, P. R. China
THEORETICAL STUDIES ON THE PROPERTIES OF SOME ENERGETIC MATERIALS
771
Martin imek*, Vt Kuttelwascher*, and Petr Stojan**

* University of Defense, Faculty of Military Technologies, Kounicova 65, CZ-602 00, Brno, Czech Republic
** Explosia a.s., Research Institute for Industrial Chemistry, CZ-532 17, Pardubice - Semtn, Czech Republic
INFLUENCE OF VELOCITY OF GAS FLUX GOING TO EROSIVE BURNING PROCESS INSIDE

OF COMBUSTION CHAMBER
781
J. Skldal*, V. Jeov**, A. Eisner**, M. Kouba*, and K. Ventura**

* Research Institute of Industrial Chemistry, Explosia a.s.,532 17 Pardubice - Semtn, Czech Republic
** Department of Analytical Chemistry, University of Pardubice, 532 10 Pardubice, Czech Republic
DETERMINATION OF NITRATE ESTERS CONTENT IN WATER SAMPLES
789
M. korpkov
Research Institute of Industrial Chemistry, Explosia a.s. 532 17, Pardubice Semtn, Czech Republic
IDENTIFICATION AND MUTUAL DIFFERENTIATION BETWEEN NITROESTERS

AND NITRAMINES BY THE COLOUR REACTION WITH THYMOL
795
M. korpkov
Research Institute of Industrial Chemistry, Explosia a.s. 532 17, Pardubice Semtn, Czech Republic
UV-VIS SPECTRA OF SOME ENERGETIC MATERIALS MEASURED

WITH THE OPTICAL FIBER PROBE
802
Renata psov
Explosia a. s., Research Institute for Industrial Chemistry, Safety Engineering Laboratory, 532 17 Pardubice-Semtn, CZ
DIFFERENTIAL SCANNING CALORIMETERS
808

Siberian State Technological University 660049, Krasnoyarsk, prosp. Mira, 82, Russia
STRUCTURE-KINETIC LAWS OF THERMAL DECOMPOSITION

OF SIX-MEMBERED CYCLIC N-NITRAMINES
816
N.P. Taibinov, B.G. Loboiko, V.P. Filin, .V. Kostitsin, V.V. Shaposhnikov, S.N. Lubyatinsky,
.B. Smirnov, .B. Syrtsov, .V. Vershinin, and .. Nikulin
Zababakhin Russian Federal Nuclear Centre - VNIITF, P.O. Box 245, Snezhinsk, Chelyabinsk region, 456770 Russia
RECORDING OF PARTICLE VELOCITY PROFILES IN SHOCK

AND DETONATION WAVES BY ELECTROMAGNETIC TECHNIQUE
825
Waldemar A. Trzciski, Marcin Kutkiewicz, and Leszek Szymaczyk

Military University of Technology, Kaliskiego 2, 00 908 Warsaw, Poland
THE USE OF THE GAP TEST TO INVESTIGATE THE SHOCK TO DETONATION TRANSITION
IN LOW-SENSITIVITY EXPLOSIVES; PART I EXPERIMENTAL APPROACH
832
Waldemar A. Trzciski
THE USE OF THE GAP TEST TO INVESTIGATE THE SHOCK TO DETONATION TRANSITION
IN LOW-SENSITIVITY EXPLOSIVES; PART II NUMERICAL SIMULATION
14
839
Yukihiro Tsunezumi*, Miyako Akiyoshi**, Hiroshi Miya***, and Hidetsugu Nakamura*

* Department of Applied Science for Integrated System Engineering, Kyusyu Institute of Technology, Sensui-machi,
Tobata-ku, Kitakyusyu-shi 804-8550, Japan
** National Institute of Advanced Industrial Science and Technology (AIST), Higashi1-1-1, Tsukuba, Ibaraki,
305-8565, Japan
*** Asahi Kasei Chemicals Corporation, Engineering and Development Department , Oita Plant, 2620, Oaza-Sato, Oita
870-0392, Japan
AGING BEHAVIOR OF PROPELLANTS
846
J. Vgenknecht, V. Adamk, and Z. Aktein

**Research Institute of Industrial Chemistry, Explosia a.s. 532 17, Pardubice Semtn, Czech Republic
CONTRIBUTION TO STUDIES OF DETONATION OF A CURVED CYLINDRICAL CHARGE
853
R. Varga*, S. Zeman*, J. Vgenknecht*, and M. Kolovri**

** Institute of Forensic Science of the Slovak Police Corps, Sklabinsk 1, SK-812 72 Bratislava, Slovak Republic
APPLICATION OF LC-MSD AND GC-ECD TECHNIQUES IN THE ANALYSIS

OF POST EXPLOSION RESIDUES OF EXPLOSIVES ON THE BASIS
OF 1,3,5-TRINITROSO-1,3,5-TRIAZINANE
864
Xue-Hai Ju and He-Ming Xiao

Department of Chemistry, Nanjing University of Science and Technology, Nanjing 210094, P. R. China
INTERMOLECULAR INTERACTIONS OF ENERGETIC COMPOUNDS IN GASEOUS

AND SOLID STATES
875
Ye Zhiwen, Lu Chunxu, and Liu Zuliang

Nanjing University of Science & Technology, Nanjing, 210094
SYNTHESIS AND APPLICATION OF GEMINI SURFACTANT IN EMULSION EXPLOSIVE
889
Adam Zakrzewski, and Zenon Wilk

Institute of Organic Industry Branch in Krupski Myn, Zawadzkiego 1 Str, 42-693 Krupski Myn, Poland
EFFECTS OF THE REACTION OF THE LINER CONTAINING THE MAGNESIUM

AND ALUMINUM WITH WATER AND WATER SOLUTION OF AMMONIUM NITRATE
896
Svatopluk Zeman and Rbert Varga

Department of Theory and Technology of Explosives, University of Pardubice, CZ-532 10 Pardubice, Czech Republic
STUDY OF THERMAL AND DETONATION REACTIVITIES OF THE MIXTURES

CONTAINING 1,3,5-TRINITROSO-1,3,5-TRIAZINANE (TMTA)
901
Svatopluk Zeman*, Yuanjie Shu**, and Xinfeng Wang**.

* Department of Theory and Technology of Explosives, University of Pardubice CZ-532 10 Pardubice, Czech Republic
** Institute of Chemical Materials CAEP, P. O. Box 919-301, Mianyang,
Sichuan 621900, P. R. China
STUDY ON MECHANISMS OF INITIATION OF TWO POLYNITRO ARENES

Stjepan ganec*, Ester Zvonimir** and Mario Dobrilovi**
910
* Minervo Ltd, Ljubljana, Slovenia

** Faculty of Mining, Geology and Petroleum Engineering, University of Zagreb, Croatia
COMPARISON OF CONTINOUS AND DISCONTINOUS METHOD OF DETONATION VELOCITY

MEASUREMENT IN MINING DRILLS (VOD)
917
Igor Zori, and Dalibor Kuhinek
University of Zagreb, Faculty of Mining, Geology and Petroleum Engineering, Pierottijeva 6, 10000 Zagreb, CROATIA
NON STANDARD METHODES IN TESTING FIRING CURRENT OF ELECTRIC

DETONATORSPREFACE
925
Additionally published poster, which has not been filed into the Proceeding of the 7. Seminar
Zvonimir Ester, Mario Dobrilovi, Darko Vrkljan
University Of Zagreb, Faculty of Mining, Geology and Petroleum Engineering, Pierottijeva 6, 10 000 Zagreb, Croatia
COMPARISON OF VOD MEASUREMENT RESULTS WITH CONTINUOUS AND DISCONTINUOUS

METHOD
932
15
PREFACE
This year, 85 years will have elapsed since the start of education in the field of
science and technology of explosives in Bohemia. In 1920 this education began at the
present Pragues Institute of Chemical Technology in Prague under the guidance of Prof. Dr.
Cyril Krauz. In the same year Explosia Ltd., a factory producing explosives, was established
at Pardubice. These two events significantly affected the development in the field of
energetic materials in Czechoslovakia, a young state formed in 1918 after break-up of
Austrian-Hungarian monarchy. The said development along with the high level of Czech
machine industry caused Czechoslovakia to become an important producer and exporter of
arms, ammunition and explosives in the years before the World War II.
Teaching the field of science and technology of explosives was interrupted by World
War II. Czech universities were closed in the period of 1939-1945. It was renewed to a
limited extent in 1946 again at the present Pragues Institute of Chemical Technology in the
framework of activities of the existing Department of Organic Technology. In 1952, a
subsection Technology of Special Production was established at this Department under the
leadership of Dr. Josef Seifert, who exploited his contacts with Explosia Comp., and in 1953
transferred the teaching activities to Institute of Chemical Technology in Pardubice (the
present Faculty of Chemical Technology, University of Pardubice), where he established
Department of Technology of Special Production. In the period of 1951-1959 the teaching in
the area of chemistry and technology of explosives was also realized at Military Technical
Academy in Brno. Several professors of this Academy were external co-workers of the
Department in Pardubice, and this cooperation has been continuing ever since.
After World War II, Czechoslovakia did not attain its former position in production
and export of arms, ammunition and explosives. In spite of that, the export of abovementioned commodities remained highly important for its economy. However, this
importance was not reflected by the extent of support of scientific and pedagogical activities
of the Department of Technology of Special Production from the side of the Institutes
management. In 1963 a two-storied building was found (the present Technological hall of
the Faculty of Chemical Technology) originally dedicated for scientific and pedagogical
activities in the field of explosives. The department however has obtained there only two
thirds of the second floor and further problems were yet to come. The independency of the
department has ended in 1965, and until its renewal in 1986 its fragments existed as parts of
two other Departments of the Institute.
The renewed Department started its activities in September 1986 under the name of
Department of Theory and Technology of Explosives (DTTX). In 1995, DTTX went through
a severe personal crisis. This crisis resulted in the fact that the present scientific and
pedagogical staff does not include workers in the age category of 40-60 years. The
Department also struggles with a shortage of necessary rooms; this problem has been
existing since its establishing in 1953 up to now since the interest in studies at the
Department continuously increases.
The workplaces of Department were and still are unique of their kind on the territory
of former Czechoslovakia. Even despite the unfavorable situation mentioned, they were able
to provide education to 320 MS graduates, more than 350 graduates of four-term license
courses (technological and blasting), and 58 PhD graduates. In academic year 2004/2005 the
total number of students studying in the said programs is 70, out of whom 15 are Slovak
citizens. The ongoing restructuring of Czech universities will also affect DTTX in
unpredictable way. A change of name to Department of Energetic Materials is anticipated.
16
The former Departments graduates also include Hungarian citizens, Afghanis,

Yugoslavs and Egyptians. A systematic cooperation with foreign partners began only after
1994. The international seminars New Trends in Research of Energetic Materials belong
among the most important activities of the said type. They started from cooperation with
Polish colleagues from Military University of Technology and Institute of Industrial Organic
Chemistry, both the partners from Warsaw. The original stimulus for starting the seminars
was an attempt to teach young research workers to present their results in front of scientific
audience. These Seminars gradually changed into international meetings of young
specialists, university teachers and senior top-experts working in the field of energetic
materials. They became a part of education of students at DTTX. With regard to economic
situation of the middle European countries, Balkan and former Soviet Union, no conference
fee was ever asked from the participants. This would never be possible without the generous
support of many sponsoring institution and individuals. In the last four years a very
significant financial support has been regularly provided by the European Office of
Aerospace Research & Development of the USAF in London, by Austin Detonator Comp.
(Vsetn) and by Indet Safety Systems Comp. (a member of Nippon Kayaku group in Vsetn).
This year the list of sponsors has been extended by Explosia Comp. Pardubice, the firm CZ
Team, United States Army International Technology Center (Atlantic) - European Research
Office in London and a substantial support provided by Defence Science & Technology Lab.
(DSTL), Sevenoaks (an Agency of the Ministry of Defence of U.K). The efficient help in
ensuring smooth and successful course of the meeting obtained from all these institutions is
gratefully acknowledged. We greatly appreciate that thanks to this support all specifics of the
previous seminars can be maintained. It would be quite unfair not to mention personal efforts
of our foreign friends helping to find sponsoring organizations, publicize the Seminar abroad
and personally participate in organizing activities during the Seminar. In this context, we are
especially obliged to Dr. Woodward Waesche from ONRIFO (USA), and to Dr. Adam
Cumming from DSTL (UK).
Finally, I wish to thank the members of the Scientific Committee, the authors of all the
seminar papers and, last but not least, you, the participants in this seminar, for its success and
its influence on the continued success and growth of all future meetings at our University of
young people and university teachers working in the field of all kinds of energetic materials.
Allow me to use this opportunity for inviting you in the name of my co-workers and
mine: we are looking forward to meet you at the ninth seminar on April 25th - 27th, 2006 in
Pardubice.
Pardubice, March 10th, 2005
Svatopluk Z e m a n
17
Austin Detonator s.r.o.

Jasenice 712
755 01 Vsetn CZ
tel. 571404156, 571404120, fax. 571431926
E-mail: austin@austin.cz, http:// www.austin.cz
The company is a manufacturer and supplier of electric detonators, non-electric

detonators and blasting machines. Austin Detonator s.r.o. product range includes:
a) Complete range of industry electric and non-electric detonators for use in
standard and specific conditions, such as:
explosive environment (underground coal mines)
high water pressure
high temperatures
foreign sources of electric energy
large variability of delays
b) blasting machines PR-12 and DBR-12 for initiation of electric initiation system
blasting machines JR-1 and MR-1 for initiation of non-electric initiation system
Indetshock/Shockstar
c) Through its subsidiaries in Romania, Austria, Poland and
Hungary, the company provides service, which includes:
Complex rock-on-the-ground service - supply of know -how and
blasting material
Small and large scale blast w ork outside the scope of mining
industry
Sales and distribution of commercial explosives, detonators and
accessories for blasting of Austin Pow der International company
Consultancy and expertise for use of explosives
18
CZ team Ltd.
UXO and Mine Clearance
Equipment * Detecting * Expert Services
CZ team Ltd. UXO clearance in Czech republic

Dr. Jiri Chladek
CZ team Ltd. is a high-specialized company focused on UXO and mine
clearance and related services and equipment. CZ team Ltd. was established
in spring 2004 by integration of companies Chladek (est. 1992) and Zazvorka
(est.1994).
UXO and aircraft bombs can be finding in former battlefields, military training
camps or bases. These procedures are very important especially for building
industry, because old ammunition is still very dangerous. We are able to find
and dispose off not only UXO and mines, but also aircraft bombs or other deep
buried ferrous object. We are able to deal with UXO also underwater.
CZ team Ltd., Sulova 1247, 156 00 Prague 5 Zbraslav, Czech Republic

Tel./Fax: ++ 420 257 922 907, e-mail: info@chladek.cz , cell. ++420 602 343 958
19
CZ team Ltd.
UXO and Mine Clearance
Equipment * Detecting * Expert Services
Services:
9 UXO and mine clearance
9 Detection of aircraft bombs and other deep buried metal objects by
magnetometry
9 Underwater UXO clearance
9 Detection of metals objects
9 EOD training
9 Repair and maitenance our products
Goods:
9
9
9
9
9
Metal detectors for field

Metal detectors for underwater operation
Metal detectors for Security
Explosive detector
Mine Resistant Boots
CZ team Ltd., Sulova 1247, 156 00 Prague 5 Zbraslav, Czech Republic

Tel./Fax: ++ 420 257 922 907, e-mail: info@chladek.cz , cell. ++420 602 343 958
20
Explosia a.s.
532 17 Pardubice Semtn
Explosia a.s. is production and trading company operating in the
field of production of explosives, ammunition and providing the
services, connected with application of energetic materials for
commercial as well as military use.
The company with original name Czechoslovak Joint-Stock
Factory for Explosive Materials was established in 1920, and this
year celebrates the 85th anniversary. Name Explosia, used first
unofficially only, was adopted in 1934. During its existence
Explosia has gone through a number of organizational changes. An
important step in modern history of Explosia was the rise of jointstock company Explosia in 1998 and its consequent property splitoff from Aliachem a.s. At present time, 100% Explosia a.s. shares
are owned by the state.
Explosia a.s. holds an important position in the field of production of explosives and
propellants not only in the Czech Republic, but it is also an important exporter of
propellants, commercial explosives, ammunition and special pyrotechnic products.
A part of Explosia a.s. is the Research Institute of Industrial Chemistry (VPCH)
providing research and development in the field of explosives and ammunition not only for
Explosia a.s. but also for other partners in the
Czech Republic as well as in abroad.
Apart from R&D, VPCH offers services in the
field of analytics of explosives, testing of
explosives and propellants and short-run
production of new energetic materials, special
explosives and pyrotechnic components for
aircraft rescue systems (the figure shows rocket
motor which is a part of pilot seat in L 159
aircraft).
Explosia a.s. produces a wide assortment of qualified
commercial explosives for open pits blasting, explosives
designed for underground blasting, special plastic
explosives for military use, smokeless powders for both
commercial and military use and a number of other
products, out of which it is possible to mention e.g.
modular charges for weapons of calibre 155 mm,
subcalibre ammunition of calibre 125 mm, pyrotechnic
components for aircraft rescue systems, plastic
explosives etc.
21
Explosia a.s. has the quality system certified according to SN EN ISO 9001:2000 and, for
the supplies for the Army of the Czech Republic, the Certificate of Conformity of Quality
System with AQAP 110 requirements.
Further information concerning Explosia a.s., assortment of produced explosives, smokeless
powders and other products you can find on our web site: www:explosia.cz.
22
Energetic Materials
Testing Instruments Technologies Expert Services
TESTING INSTRUMENTS FOR ENERGETIC MATERIALS

Instruments for EM Testing Programs
OZM Research manufactures or supplies testing instruments for energetic materials
completely covering needs of the following testing programs:
EC Conformity Attestation for Placing of Commercial Explosives on EU Market (93/15/EC)

NATO Qualification of Explosive Materials for Military Use (STANAG 4170)
Surveillance of Military Explosive Materials
Quality Control of Manufactured or Processed Explosive Ingredients (STANAG)
Quality Control in Ammunition Manufacture
Transport Classification Tests for Explosive Materials (UN, ADR, RID, IMDG, IATA DGR)
Testing of Explosive Properties of Chemicals and Wastes
Forensic Examination of Explosives
Research and Development of New Energetic Materials
Instruments for EM Properties
Our testing instruments cover wide range of properties of energetic materials:
Thermal Analysis, Stability and Compatibility

Chemical Stability and Compatibility Tests
Mechanical and Thermomechanical Analysis
Sensitiveness Tests
Explosiveness Tests
Performance Tests
High-Speed Measurements
Ballistic Tests
Flaw Detection
Physical Properties
Chemical Analysis
Environmental Tests
Type-Specific and Customer-Defined Tests
Complete Solution of Your Needs in EM Testing
OZM Research is a knowledge-based company formed by experienced explosives
scientists and engineers with advanced academic background in this field. The company is
capable to offer COMPLETE SOLUTIONS of customers needs in testing of energetic materials
NOT JUST DELIVERIES of the instruments. The company is prepared to provide turn-key
installations with detailed testing procedures, methodologies for practical interpretation of
results, consultancy for organizing customer's testing programs, qualified training of lab
personnel and continuous expert support, all according to the customer's specific requirements.
Do not hesitate to test us!
OZM Research, Nadrazni 266, 538 62 Hrochuv Tynec, Czech Republic

Tel: +420 608 742 777, Fax: +420 469 692 882, E-mail: ozm@ozm.cz, http://www.ozm.cz
23
Energetic Materials
DIFFERENTIAL THERMAL ANALYSER FOR EXPLOSIVES

DTA 551 Ex is a testing instrument designed specifically for differential thermal
analysis (DTA) of explosive materials which explosion might cause damage to standard
commercial analysers. Its main features are represented by robust design resistant against
explosion of up to several hundred milligrams of explosives (allowing analysis of truly
representative samples), high sensitivity due to direct contact of a thermocouple with a
sample, high precision and accuracy, large variety of accessories, user-friendly software
for data acquisition, analysis and archiving, many testing modes, very low costs of investment
and operation. These features brought popularity of DTA 551 Ex at dozens of explosives
laboratories of many countries, some of them operating the instrument for more than 8 years.
DTA as a method is applied for evaluation of
thermal stability of explosive materials, their
purity
(melting
point,
solidification
point),
compatibility, thermal decomposition parameters.
It is used in quality control of manufactured
explosives, characterization and qualification of
new
explosives,
surveillance
of
in-service
explosives, R&D and many other testing programs.
Based on proven design of DTA 551 Ex we
manufacture
many
other
customerdefined
instruments, e.g. large-scale DTA for sample
weights of several grams with simultaneous
temperaturepressure measurement using high
pressure cells.
VACUUM STABILITY TESTER FOR EXPLOSIVES

Vacuum stability tester STABIL is used for measuring chemical stability of explosive
materials based on volume of gas evolved from heating the samples in evacuated test tubes in a
heating block maintained at a constant temperature (30 - 180 oC) for a specified period of time.
STABIL is a fully instrumental tester equipped with sensitive electronic pressure transducers,
communication with PC for direct control, continuous data acquisition, analysis and archiving and
allowing to test multiple samples (up to 20) from a single measuring unit.
The apparatus fully complies with requirements of STANAG 4556, but can also be used
with minor modifications for other customer-defined tests, such as long-term heating at lower
temperatures or measurements under initial atmospheric pressure or in different atmospheres.
Vacuum stability test is frequently used for
determination of compatibility of energetic materials
with contact materials as defined in STANAG 4147 and
for quality tests of energetic ingredients (multiple
STANAGs). The test is able to discover with high
sensitivity, precision and reproducibility chemical
instability of energetic materials caused by presence of
destabilizing impurities, incompatibility with surrounding
materials or ageing. Vacuum stability test finds its wide
application
in
qualification,
surveillance,
manufacture, quality control and R&D of a whole
range of energetic materials.
Please contact us for more information and technical data on DTA 551 Ex, STABIL and the others!

24
Energetic Materials
SMALL-SCALE ELECTRIC SPARK TESTER

Electrostatic discharge is one of
the most frequent and the least
characterized causes of accidental
explosions of energetic materials. To
have reliable data on electric spark
sensitiveness of energetic materials is
thus critical imperative in their R&D,
manufacture, processing, loading or
demilitarization.
Small-scale
electric
spark
tester ESZ KTTV 2001, as one of very
few available instruments of this type,
allows to precisely measure both
total spark energy discharged into
the sample and a fraction of this energy really absorbed by the sample initiating its explosion.
This feature allows to determine the true minimum energy sufficient for accidental
initiation of the sample which allows to stay on the safe side when applying the results for any
real-world scenario. The apparatus is designed according to specification of NATO AOP-7 method
CZ 201.03.001.
Tester ESZ KTTV 2001 can load the sample with very wide range
of spark energies from 1.10-5 to 1,6.101 J allowing to test all
categories of energetics ranging from extremely sensitive primary
explosives to insensitive high explosives. Input energy and the fraction of
the energy consumed by the sample are measured by high voltage
sensors connected with oscilloscope and analyzed by a special
software in the operators PC. Discharge duration can be varied allowing
to test the samples in modes incorporating different profiles of the
spark (shock wave or thermal initiation mechanisms by the spark).
LARGE-SCALE ELECTRIC SPARK TESTER

Large scale electric spark tester ESZ KTTV LS is used for large scale electrostatic
discharge test as defined in STANAG 4490. This test involves loading of an explosive sample
weighing several hundred grams by a series of 30 discharges with constant spark energy
of 15.6 J. The test simulates possible
real-world scenario of accidental ignition
of large charges of energetic materials
(especially propellants) during their
manufacture, processing, loading or
demilitarization.
From safety reasons, this test has to be
carried out in detonation chambers,
bunkers or in detonation pits (with
necessary protection of the instrument)
and the possible interactions (extent of
reaction) of the electrostatic discharge
with the sample detected remotely by
CCD cameras.
Please contact us for more information and technical data on electric spark testers and the others!

25
Energetic Materials
INDUSTRIAL DETONATION CHAMBERS

Industrial detonation chambers are automated machinery designed to withstand
repeated detonations as part of a manufacturing process, with many applications:
Research, development, evaluation and testing of explosives, ordnance and confined explosions
Quality control in manufacture of explosives and munitions
Environmentally friendly demilitarization of munitions
Safe destruction of explosive wastes or improvised explosive devices
Explosive forming of metals hardening, welding, cutting, pressing
Manufacture of artificial diamonds
Industrial detonation chambers are manufactured in
partnership between OZM Research and Design and Technology
Branch of Lavrentyev Institute of Hydrodynamics, a leading
Russian scientific and technological institute for high pressure
hydrodynamics. These detonation chambers have been produced
in Russia for more than 40 years and installed in many countries.
During these years of service, they have proven to have long
service life (100,000s detonations) and reliable, safe and
simple operation with low investment and operating costs.
For scientific applications, the chambers can contain windows covered by steel or
plexiglas discs for safe optical or electrical measurements of detonation processes. For testing or
destruction of fragmentation ammunition, replaceable armor shields are installed. The
chambers are equipped with modern hydraulic system and control panel providing for full
control over automatic remote operations of opening and closing the chamber, as well as electric
firing of the explosive charge. The modernized detonation chambers comply with all relevant EU
regulations and, where required (for demilitarization technologies), they can be equipped with
type-specific off-gas treatment units.
Chamber
type
KV-0.2
KV-2
RADUGA
KV-5
KVG-8
KVG-16
Maximum
charge weight
0.2 kg TNT
2 kg TNT
2.6 kg TNT
5 kg TNT
8 kg TNT
16 kg TNT
Overall dimensions
lxwxh
1930 x 1190 x 1980
2840 x 1800 x 2740
5300 x 2800 x 2600
4640 x 2800 x 4300
16360 x 2200 x 2460
27210 x 2200 x 2460
mm
mm
mm
mm
mm
mm
Total
weight
2 tons
10.5 tons
20.5 tons
35 tons
48 tons
76 tons
The table above summarizes basic parameters of standard types of detonation chambers.
The manufacturing team is however experienced in designing detonation chambers fulfilling very
specific needs and requirements of a customer. Please contact us for further details!

26
Central European Journal of Energetic Materials

CEJEM - a new forum for the exchange of science
and technology in energetic materials
CEJEM provides a platform for those scientists

interested in publishing and reading research papers
in the field of energetic materials propellants,
explosives and pyrotechnics.
The aim of the journal is to focus on the latest research
results in energetic materials in general.
Topics:
ignition, combustion and detonation phenomenon,

formulation, synthesis and processing,
analysis and thermal decomposition,
toxicological, environmental and safety aspects of energetic
materials production, application, utilization and demilitarization,
molecular orbital calculations,
detonation properties and ballistics,
biotechnology and hazards testing.
CEJEM presents original research and interesting reviews.

Contributions are from experts in chemistry, physics and engineering.
All submissions are independently refereed by editorial board
members and by external referees chosen on international basis.
Indexed by CHEMICAL ABSTRACTS

Andrzej Maranda - chairman (amaranda@wat.edu.pl)
Svatopluk Zeman - vice chairman (svatopluk.zeman@upce.cz)
EDITORIAL BOARD
PROMOTION!
30
Two issues at the price of one!

Ang How-Ghee, Nanyang Technological University, Singapore
Alexander Astachov, Siberian State Technological University, Russia
Karol Balog, Slovak University of Technology Bratislava, Slovak Republic
Anthony Bellamy, Cranfield University, Royal Military College of Science, Swindon
Andrzej K. Bi, Warsaw University of Technology, Poland
Jaroslav Buchar, Mendel University of Agricult. & Foresty, Brno, Czech Rep.
Stanisaw Cudzio, Military University of Technology, Warsaw, Poland
Bogdan Florczak, Institute of Industrial Organic Chemistry, Warsaw, Poland
Zdenek Friedl, University of Technology, Brno, Czech Republic
Sergey Frolov, Semenov Institute of Chemical Physics Russian Academy of Sciences, Moscow, Russia
Michael A. Ilyushin, Saint-Petersburg State Institute of Technology, Russia
Thomas M. Klapoetke, University of Munich, Mnchen, Germany
Andrzej Ksiczak, Warsaw University of Technology, Poland
Nikolaj Latypov, Swedish Defence Agency (FOI), Tumba, Sweden
Michel Lefebvre, Royal Military Academy, Brusseles, Belgium
Only during the Conference!

Carl-Otto Leiber, Rheinbach, Germany
Frantiek Ludvk, Military Academy Brno, Czech Rep.
Andrzej Maranda, Military University of Technology, Warsaw, Poland
Ryszard Morawa, University of Mining and Metallurgy, Cracow, Poland
Peter Politzer, University of New Orleans, USA
Yuanjie Shu, Institute of Chemical Materials, CAEP, Sichuan, China
Valerii P. Sinditskii, Mendeleev University. of Chemical Technology, Moscow, Russia
Wincenty Skupiski, Institute of Industrial Chemistry, Warsaw, Poland
Muhamed Sueska, Brodarski Instytut, Zagreb, Croatia
Ulrich Teipel, University of Applied Sciences Nrnberg and Fraunhofer ICT, Germany
Fred Volk, Fraunhofer Inst. fr Chemische Technologie, Pfinztal, Germany
R. H. Woodward Waesche, Office of Naval Res. Int. Field Office, USA
Waldemar Witkowski, Institute of Industrial Organic Chemistry, Warsaw, Poland
Svatopluk Zeman, University of Pardubice, Czech Rep.
Annual subscription price 2005: in Poland 400 PLN, outside 120 (exclusive of VAT, incl. postage and handling charges).
Price may be subject to be changed. The orders should be submitted to publisher by post (address above) or e-mail: karas@ipo.waw.pl
More information can be found in the enclosed leaflet including details on subscription and Web www.ipo.waw.pl
Publisher
INSTITUTE OF INDUSTRIAL ORGANIC CHEMISTRY

Promotion and Information Department, Tel: 48 22 8111231, Fax: 48 22 8110799, Web www.ipo.waw.pl
03-236 Warsaw, 6 Annopol Str., Poland
Managing Editor: Wanda Kara (e-mail: karas@ipo.waw.pl)
Quarterly
ISSN 1733-7178
Publisher
INSTITUTE OF INDUSTRIAL ORGANIC CHEMISTRY
Editorial Office
Promotion and Information Department
03-236 Warsaw, 6 Annopol Str., Poland
Tel: 48 22 8111231
Fax: 48 22 8110799
www.ipo.waw.pl
Andrzej Maranda chairman

amaranda@wat.edu.pl
Svatopluk Zeman vice chairman

svatopluk.zeman@upce.cz
EDITORIAL BOARD
Ang How-Ghee, Nanyang Technological University, Singapore
Alexander Astachov, Siberian State Technological University, Russia
Karol Balog, Slovak University of Technology Bratislava, Slovak Republic
Anthony Bellamy, Cranfield University, Royal Military College of Science, Swindon
Andrzej K. Bi, Warsaw University of Technology, Poland
Jaroslav Buchar, Mendel University of Agricult. & Foresty, Brno, Czech Republic
Stanisaw Cudzio, Military University of Technology, Warsaw, Poland
Bogdan Florczak, Institute of Industrial Organic Chemistry, Warsaw, Poland
Zdenek Friedl, University of Technology, Brno, Czech Republic
Sergey Frolov, Semenov Institute of Chemical Physics Russian Academy of Sciences, Moscow, Russia
Michael A. Ilyushin, Saint-Petersburg State Institute of Technology, Russia
Thomas M. Klapoetke, University of Munich, Mnchen, Germany
Andrzej Ksiczak, Warsaw University of Technology, Poland
Nikolaj Latypov, Swedish Defence Agency (FOI), Tumba, Sweden
Michel Lefebvre, Royal Military Academy, Brusseles, Belgium
Carl-Otto Leiber, Rheinbach, Germany
Frantiek Ludvk, Military Academy Brno, Czech Republic
Andrzej Maranda, Military University of Technology, Warsaw, Poland
Ryszard Morawa, University of Mining and Metallurgy, Cracow, Poland
Peter Politzer, University of New Orleans, USA
Yuanjie Shu, Institute of Chemical Materials, CAEP, Sichuan, China
Valerii P. Sinditskii, Mendeleev University. of Chemical Technology, Moscow, Russia
Wincenty Skupiski, Institute of Industrial Chemistry, Warsaw, Poland
Muhamed Sueska, Brodarski Instytut, Zagreb, Croatia
Ulrich Teipel, University of Applied Sciences Nrnberg and Fraunhofer ICT, Germany
Fred Volk, Fraunhofer Inst. fr Chemische Technologie, Pfinztal, Germany
R. H. Woodward Waesche, Office of Naval Res. Int. Field Office, USA
Waldemar Witkowski, Institute of Industrial Organic Chemistry, Warsaw, Poland
Svatopluk Zeman, University of Pardubice, Czech Republic
Managing Editor: Wanda Kara (e-mail: karas@ipo.waw.pl)

CEJEM - a new forum for the exchange of science
and technology in energetic materials
CEJEM provides a platform for those scientists interested in publishing
and reading research papers in the field of energetic materials propellants,
explosives and pyrotechnics. The aim of the journal is to focus on the latest
research results in energetic materials in general.
Topics:
ignition, combustion and detonation phenomenon,
formulation, synthesis and processing,
analysis and thermal decomposition,
toxicological, environmental and safety aspects of energetic materials
production, application, utilization and demilitarization,
molecular orbital calculations,
detonation properties and ballistics,
biotechnology and hazards testing.
CEJEM presents original research and interesting reviews. Contributions
are from experts in chemistry, physics and engineering. All submissions are
independently refereed by editorial board members and by external
referees chosen on international basis.
Indexed by CHEMICAL ABSTRACTS

Annual subscription price 2005:
in Poland 400 PLN, outside 120 (exclusive of VAT, incl. postage and handling charges).
Price may be subject to be changed. The orders should be submitted to the publisher
by post (address above) or e-mail: karas@ipo.waw.pl
For more information visit our Web at www.ipo.waw.pl

Volume 1, Number 1, 2004
Svatopluk ZEMAN and Zdenk FRIEDL
Relationship between Electronic Charges at Nitrogen Atoms
of Nitro Groups and Onsets of Thermal Decomposition of Polynitro Arenes.......................................................3
Muhamed SUESKA and Sanja MATEI MUANI
Numerical Modeling of Self-Ignition of Energetic Materials ...............................................................................23
Zhang CHAOYANG, Shu YUANJIE, Zhao XIAODONG and Dong HAISHAN
DFT Study on Non-reversible Expansion of TATB Crystal.................................................................................43
Zbigniew A. WALENTA, Andrzej TEODORCZYK, Andrzej DBKOWSKI and Waldemar WITKOWSKI
Direct Monte-Carlo Simulation of a Detonation Wave in a Narrow Channel, Containing Flammable Gas........51
Dorota POWAA, Andrzej ORZECHOWSKI, Andrzej MARANDA,
Jerzy NOWACZEWSKI and Adam LOREK
Researches on New Crystalline Form of Some Insensitive High Explosives in PBX ........................................63
Bogdan CZAJKA, Stanisaw TABAT, Sawomir STYCZYSKI
Leszek WACHOWSKI, Micha ZIELISKI and Mariusz PIETROWSKI
Physicochemical Properties of Iron Powders Used as Reducing Agents
in Materials of High Calorific Value ....................................................................................................................75
Volume 2, Number 1, 2005

Valery P. SINDITSKII
Reason for Heat Instability of Combustion of Energetic Materials
with Condensed Phase Leading Reaction ......................................................................................................... 3
Mikhail A. ILYUSHIN, Igor V. TSELINSKY, Igor A. UGRYUMOV,
Valerii Yu. DOLMATOV and Irina V. SHUGALEI
Study of Submicron Structured Energetic Coordination Metal Complexes for Laser Initiation Systems .......... 21
Zbigniew STRADOMSKI, Henryk DYJA and Andrzej MARANDA
Tribological Properties of High Manganese Austenit Strengthened by Explosion ............................................ 35
Xinfeng WANG, Yuanjie SHU and Chaoyang ZHANG
Investigation on NO2 Catalysis Mechanism in Dimethylnitramine Decomposition Using DFT Method ............ 47
Bogdan CZAJKA and Leszek WACHOWSKI
Some Thermochemical Properties of High Calorific Mixture of Fe-KClO4 ........................................................ 55
Katarzyna LIPISKA, Marek LIPISKI and Andrzej MARANDA
Demilitarized Double Base Propellants as Ingredients of Commercial Explosives .......................................... 69
PROMOTION!
30
Two issues at the price of one! Only during the Conference!
DISLOCATION - ASSISTED INITIATION OF ENERGETIC MATERIALS

Ronald W. Armstrong
Center for Energetic Concepts Development,
University of Maryland, College Park, MD 20742, U.S.A.
Abstract
The role of dislocations in assisting initiation of (explosive) chemical decomposition of
energetic materials has connection with the known influences for crystals and polycrystals
of dislocations facilitating permanent deformations and phase transformations. X-ray
topographic observation of relatively few dislocations in solution-grown crystals relates to
the influence of large Burgers (displacement) vectors that are characteristic of molecular
crystal bonding. Both model evaluations of the load dependence of cracking at hardness
indentations and the derived hardness stress-strain behaviors show that dislocation
movement is difficult whether in the indentation strain fields or at the tips of indentationinduced cracks. Thus, energetic crystals are elastically compliant, plastically hard, and
relatively brittle [1]. Nevertheless, cracking is shown to be facilitated by the shear stress
driven, normally limited, dislocation flow that, on molecular dynamics and dislocation pileup model bases, is shown to be especially prone to producing localized hot spot heating for
explosive initiations. Such model consideration is in agreement with greater drop-weight
heights being required to initiate smaller crystals. The crystal size effect carries over to
more difficult combustion occurring for compaction of smaller crystals. The total results
relate to dual advantages of greater strength and reduced mechanical sensitivity accruing
for the development of nanocrystal formulations. In consequence, also, several levels of
dislocation-assisted modeling are described for initiation mechanisms under shock wave
loading conditions.
Keywords:
dislocations, initiation, energetic crystals, x-ray topography, hardness,

cracking, drop-weight impact, combustion, shock wave loading, crystal
growth, plasticity, RDX, PETN, HMX, hot spots, sensitivity, crystal lattice
modeling
1.
DISLOCATION ORIGINS IN ENERGETIC CRYSTALS
The crystallographically imperfect structures of conventional, solution grown, energetic

crystals, for example, of RDX (cyclotrimethylenetrinitramine), show a segmented growth sector
structure in agreement with dislocation-based crystal growth predictions [2], as shown in Fig. 1ad. In 1a, first, a cross-sectional sketch is shown of the overall growth sector and boundary
structure that is normally developed during progressive growth of a polygonally-shaped crystal
and, in 1b, is shown separately the threading dislocation bundle structure emanating from a
crystal seed, S. The dislocations produce crystal growth more easily than any other mechanism.
As indicated in 1b, the dislocation bundles may emanate from inclusion particles. Profuse
nucleation of such dislocation bundles was shown to occur at interrupted sapphire crystal
growth surfaces via x-ray diffraction topographic imaging of crystal sections [3]. Here, a tabular
RDX crystal sectioned parallel to its planar (001) top/bottom surface [4] is shown in Fig. 1c. The
32
crystallographic labels in the Figure were identified from angular measurements on a standard
(001) stereographic projection for the RDX orthorhombic lattice [5]. The growth sector
boundaries provide sinks for solute pushed laterally during the facet-sensitive extension of the
adjacent growth sector regions along their normal, ni , directions in accordance with locallyoperative temperature and solute gradients that can be different at adjacent ni levels.
c)
Fig 1.
d)
[2]
a) From Klapper , showing growth sectors and boundaries building

onto a seed; and,
b) Growth-promoting dislocation lines emanating from seed and
inclusions.
c) An (001) sectioned RDX crystal [4], after chemical polishing, with
identified traces [5] of the original polygonal crystal growth surfaces
at earlier stages of growth.
d) A {210} type Lang transmission x-ray diffraction topograph of an
RDX crystal slice showing dislocation bundles emanating from a
central crystal seed [6].
33
The black lines in Fig. 1d, relating to the schematic dislocation lines in Fig. 1b, are revealed
because of the dislocation strain fields locally enhancing in their vicinity the diffracted x-ray
intensity [7], much the same as thinner dislocation lines are observed in metals through
transmission electron microscopy. In Fig. 1d, small metal spheres, shown as white circles, had
been put into etch pit depressions on the right-hand {111} type crystal facet to confirm that any
dislocation lines revealed in the topographic images would run into the pits, as shown.
Thus, energetic crystals, as all others, contain an intrinsic population of dislocations that
play an intimate role in the crystal formation process. The special nature of the dislocations in
molecularly bonded energetic crystals, however, relates first to the dislocation self-energy, Es
expressed as:
Es = [G b2/4] ln [R/ro] + Ec
(1)
where G is the shear modulus; b is the dislocation Burgers (displacement) vector; is a

factor depending on whether the dislocation is of screw ( = 1), edge ( = [1-]), or mixed type;
is Poissons ratio; R is an outer cut-off radius for negligible displacement compared to lattice
vibrations; ro is an inner core cut-off radius determined by the limit of linear elasticity; and Ec is
the remaining dislocation core energy [8]. For the molecularly-bonded energetic crystals
compared to others, the large value of b outweighs the lower value of G, and so the dislocation
self-energies are relatively large [9]. The consequence is that energetic crystals generally contain
few dislocations --- but as will be seen, the dislocation influence on strength properties and
plastically-generated hot spots is large [1].
2.
MICROINDENTATION PROPERTIES
The known brittleness of energetic crystals promoted early measurements being made of
their cracking behaviors by the method of indentation fracture mechanics (IFM) [10]. Fig. 2 is a
comparison on a log-log basis of measured diamond pyramid indentation diagonal lengths and
tip-to-tip crack lengths (along the diagonals) produced at different load values [11 -14].
34
Fig 2.
A comparison between MgO and RDX of loading force versus indentation

diagonal and tip-to-tip (diagonal) crack lengths for assessment of cracking
behaviors.
In Fig. 2, obtainment of a constant Vickers diamond pyramid hardness would be indicated

by a line of slope 2.0 as indicated by the line with that slope drawn to the left-side triangle. For
the MgO result, cracking occurred at the second highest applied load of 25 g. and produced the
indicated trend of a reduced hardness continuing to occur at increasing load values. In Fig. 2,
RDX is shown to have a lower hardness than MgO because of the downward shifted parallel
line of slope 2.0 for RDX and also is shown to be more brittle than MgO because of cracking
occurring at the lowest applied load of 15g. The important characterization of cracking for the
two materials in Fig. 2, however, relates to the prediction of a slope of 1.5 for the load
dependence on crack size predicted on an IFM basis [10, 15]. For the MgO measurements, a
reasonable fit to the IFM prediction is seen to apply at the larger applied loads and, moreover,
the occurrence of cracking has apparently influenced the plastic hardness value in the same
manner. Alternatively, the RDX crack size measurements appear to show a slope value greater
that 1.5, perhaps approaching 2.0, as indicated for the hardness measurements themselves.
Furthermore, an IFM analysis by Elban [16] of the Hagan and Chaudhri [12] cracking
35
measurements gave the important result that relatively little plastic flow was produced at the
crack tips.
Further assessment of the relative hardness and cracking properties of RDX is able to be
made on the basis of the relation between the hardness pressures and the effective hardness
strains associated with those pressures. Fig. 3 shows such a comparison.
Fig 3.
Elastic/plastic/hardness stress-strain comparison among different materials
In Fig. 3, that is based on a hardness stress-strain description given by Hammond and

Armstrong [17], the hardness stress is defined on the basis of a ball indentation test as the applied
load divided by the projected circular area of contact with the material surface. The hardness
strain is specified as the contact diameter, d, divided by the ball diameter, D. Actual Vickers
hardness numbers (VHN) are plotted in the Figure along a vertical range of values at an
effective d/D value of 0.375. A number of conventional Meyer hardness measurements for
RDX, made with a 1.59 mm ball diameter, are plotted as open-square points in the Figure [18]. A
36
linear dependence of hardness stress, on (d/D) is predicted for elastic loading according to the
Hertzian relationship
H = (4/3) [{(1-m2)/Em} + {(1-b2)/Eb}]-1 (d/D)
(2)
where m , Em and b , Eb are the Poissons ratios and Youngs modulus values for the tested
material and ball, respectively. The straight dashed lines shown at the left-side of the Figure
were computed for the listed materials while the heavy line result for NaCl, even extending
laterally to a comparable (d/D) value to the VHN measurement, was determined from points
taken along a continuous indentation test [19]. At the upper terminal points of the lines are given
C values that represent the IFM cracking stresses to be achieved on elastic loading in
accordance with the relationship [20]
C = [4Em/{D(1 s2)(K12 + K22)}]1/2 (d/D)-1/2
(3)
where is the crack surface energy and (K12 + K22) is a numerical factor equal to 2.5 x 10-5
. As indicated in equation (3) and along the loading line for MgO in Fig. 2, a lower value of C
is obtained at larger D values. The topmost 0.124 mm D value indicated on the MgO line is an
equivalent ball diameter corresponding to the VHN indentation size.
Among the various material identifying labels in Fig. 3 are those, in addition to RDX, for
PETN (pentaerythritol tetranitrate), AP (ammonium perchlorate), and HMX (cyclotetramethylenetetranitramine. Whereas ionic-bonded AP has approximately the same VHN hardness
as NaCl, the hardnesses of PETN, RDX and HMX are increasingly greater in that order. The
result is interesting to compare with the material elastic stiffnesses that are included in the
bracketed effective modulus factor in equation (2). Thus, as expected the Youngs modulus, Em ,
for PETN and RDX are indicated to be significantly lower in Fig. 3 than Em for NaCl, compared
to the reversed result for the VHN measurements and the ball test points for RDX. The
comparison between elastic moduli and measured hardnesses provides a basis for the assertion
that energetic crystals are relatively hard. Also, in all cases, the measured hardnesses are lower
than the theoretical stresses required for cracking in the absence of plastic flow. The
experimental hardness points for RDX show that cracking occurs after significant plastic flow
has occurred. Thus, whereas the presence of dislocations allows plastic flow to occur at lower
stresses than would result in their absence, the dislocation motion also provides cracking at
lower stresses than would result only from elastic deformation!
Modeled dislocation displacements in RDX have shown, in general, that shearing
displacements and, hence, dislocation movement are difficult because of the intertwined nature
of the oddly shaped molecules otherwise packed into relatively dense crystal structures [21, 22].
Fig. 4 shows an example for the projection of slip (in and out of the Figure) in the [100]
direction on the {021} type slip plane in the orthorhombic RDX lattice. An important aspect of
such modeling of shear displacements, as shown in Fig. 4, is to show that, beyond mutual
blockages between adjacent molecules, their forced shearing in dislocation motion can bring the
respective outcropping intermolecular appendages into critical reaction coordinate separations.
For the case projected in Fig. 4 near to the cell coordinate position (0,0.5,1), the hydrogenoxygen interactions could promote nitroso compound formations as have been detected both in
partially initiated material recovered from drop-weight impact tests [23] and in combustion
37
residues [24]. Such forced shearing of molecules to achieve chemical decomposition obviates the
more difficult requirement of imposing greater hydrostatic pressures to achieve the same atomic
separations. Gilman [25] has emphasized the importance of elastic shear strains even, compared
to hydrostatic ones, in facilitating chemical reactions.
Fig 4.
3.
An a-axis [100] projection, of the eight RDX molecule positions in each

of two adjacent unit cells, shown in the (100) plane containing the horizontal
c-axis [001] direction (to the left) and the b-axis [010] direction vertical, also
with labeled hydrogen, oxygen and nitrogen atom position obstructions near to
the crystal {021}type slip plane.
DROP-WEIGHT IMPACT SENSITIVITY
Early work on drop-weight impact testing of energetic crystal powders led to the model of
converting the plastic work to hot spot heating for material initiation [26]. In further research
measuring the strain responses in drop weight impact tests performed on a wide variety of
materials, Heavens and Field [27] found that explosive decompositions were associated with
discontinuous drops in load occurring during the material loading. Such load-drop behavior is
well-known in the materials community and is associated with dislocation pile-up behavior [28],
as shown schematically in Fig. 5, although attention had not been previously directed to the
thermal energy release attending pile-up breakthroughs. After essentially isothermal build-up of
internal energy in the pile-up, as indicated in Fig. 5a, a critical condition for pile-up release is
achieved, as depicted in Fig. 5b whereupon the dislocations are driven past the collapsed
obstacle by an internal stress magnified by the number of dislocations in the pile-up. The critical
breakthrough condition of Fig. 5b involves the number of dislocations, n2 , that is linearly
38
proportional to the slip length, , and 2 , so that for c* being determined by a constant
localized stress requirement, say, for cracking, then
Fig 5.
Dislocation pile-up avalanche model for quick dissipation of stored energy [28].
2 = 0 + [ c (G c* b)] -1/2
(4)
39
where 0 is the shear stress required for intrinsic dislocation movement, c is a numerical
constant, and the other symbols have been defined in equation (1). Such reciprocal square root
Fig 6.
Initiation drop-weight height versus -1/2, extended to nano-scale crystal sizes.
of slip length dependence for the applied shear stress is known as a Hall-Petch (H-P)
dependence after these researchers independently measured such dependencies for the plastic
yield stress and cleavage fracture stresses of steel materials that, not incidentally, normally
exhibit sudden load drop behavior during their initial yielding behaviors. The rather broad
application of such pile-up theory for explaining the influence of grain size on the mechanical
properties of materials has been recently reviewed [29]. The model consideration of Fig. 5,
involving the grain size dependence described in equation (4), was employed with addition of a
thermally-activated model for the intrinsic dislocation movement, to give a log-log dependence
for the drop-weight sensitivity for 50% probability of initiation, H50 , and -1/2 , with taken as
the crystal particle size [1, 30]. Fig. 6 presents those results and added ones [31] for other RDX and
CL-12 crystals. The dashed lines in the Figure enclose a region of crystal sizes larger than 0.4
m, and the extension of the Figure has been made to indicate, by extrapolation, the regime of
mechanical insensitivity that might be achievable with nanometric-sized crystals. Some
supporting evidence for improved crystal insensitivities at nanocrystal sizes has been presented
40
[32]
. Also, such indication in Figs. 5 and 6 of importance of the dislocation loading rate, it being
greater for greater drop-weight heights, relates to Fig. 3 where the plastic and cracking hardness
stress results for RDX show that there is not a very large range in stress levels available for
investigation of the dynamic strain rate dependence of dislocation behavior. And, furthermore,
the number of dislocations required in a pile-up to produce cracking in RDX is not very great
either, hence, relating to the intrinsic brittleness of RDX, as assessed earlier by Armstrong and
Elban [9] in terms of the dislocation self-energy and the surface energy for cracking, including its
experimental determination on an IFM basis [16].
4.
COMBUSTION
The cracking behavior of RDX and related energetic crystals is of interest for thermal stress
influences during combustion of conventional energetic material formulations that, generally,
incorporate polymer binding materials and a small amount of surfactants. The crystals
themselves burn over a melt layer formed on their surfaces. Fig. 7 shows evidence of such result
at a burn spot produced with a laser beam shone onto the surface of an isolated RDX crystal [33].
The relatively low energy beam was directed onto the crystal surface at a low glancing angle
The outer elliptical extent of the melt spot is made recognizable by a network of fine cracks
separated vertically by ~5 m. Their occurrence was explained [18] in terms of the thermal
stresses more easily producing cracking under the melt layer because then only the liquid-solid
interfacial energy had to be supplied for cracking as compared with the vapor-solid value
outside of the melt spot. Such consideration was employed on a conventional fracture mechanics
basis, also, to inquire whether such cracking could be responsible for a pressure-induced
instability of the energetic material burn rates through the crack magnification of induced burn
rate pressures [34].
Fig 7.
A laser-beam-induced melt spot produced on an RDX crystal surface with

thermally induced cracking having occurred underneath the crystal melted
layer [33].
41
Fig 8.
HMX particle size dependence of combustion initiated during modeled pistondriven compaction, after computed results presented by Gonthier [35].
Another aspect of particle size influence, this time, on the combustion behavior of energetic
materials is indicated in the modified H-P type dependence; see equation (4) in Section 3; now
shown in Fig. 8 derived from computations presented by Gonthier for the piston-driven
compaction of a granular bed [35]. The computational model was based on experimental results
reported by Jacobs, Sandusky and Elban [36] for particle diameters of size, 2R. The demarcation
between extinction versus sustained combustion was associated with the boundary between
local deformation and full particle-to-particle yielding of the compressed bed. In this case, the
thermally activated model of dislocation motion mentioned with respect to the particle size
dependence in Fig. 6, where = 2R, generally gives a logarithmic dependence on imposed
strain rate of the thermal stress and, so, the expectation of a linear H-P type dependence in Fig.
8, following equation (4) if transmission of plastic flow through the particle bed is required to
sustain combustion.
5.
SHOCK EFFECTS
The limit of imposed deformation rate on dislocation behavior in crystals and polycrystals is
achieved with the propagation of a shock wave. A recent symposium has been concerned with
understanding such influence, mostly, on structural metals and alloys [37]. Pioneering results had
been reported by Smith [38]. The Hugoniot elastic limit (HEL) stress for shock-induced
42
Fig 9.
Model for nanometer-scale dislocation generation at a shock front [43-45].
yielding of Armco iron material has been shown to follow an H-P dependence [39, 40]. There is
equal concern about understanding shock wave influences on the properties of energetic
crystals, including their initiations [5, 41]. Concern has been expressed for the shock propagation
front moving too fast for thermally activated chemical reactions to occur [42].
A broadly applicable model for achieving a residual state of one-dimensional strain at the
shock front has been described [43], as shown in Fig. 9, and elaborated [44, 45]. A fine scale
dislocation structure is generated at the shock front for two reasons: (1) the significant shear
strain magnitudes developed at the shock front are not able to be relieved by the displacement,
at some distance away, of the intrinsic dislocation population; and, (2) the resident dislocation
population can only travel, at fastest, at a slower elastic wave speed. The shock front itself is
envisioned to have a width on the scale of nanometers. As indicated in Fig. 9, dislocation loops
form on conjugate shear planes and then intersect to react, presumably, after nanoscale
movement to produce an immobile network of interstitial-type prismatic dislocation loops with
Burgers vectors parallel to the direction of shock wave propagation. The dislocation network is
available for interaction with the resident dislocation population traveling behind the shock
front. The interaction has been proposed to provide analogous behavior to that established in
neutron irradiation experiments, for example, leading to dislocation channeling because of
follow-on further reactions between the shock-generated dislocation network and the original
resident dislocations [1, 5]. Shock hardening in copper in post-shock deformation experiments is
well-established and attributed to the presence of a shock-induced defect structure [46].
43
Because of the small material volume involved at the shock front, and short time for wave
passage, the subject lends itself to description of individual atomic or molecular force
interactions via the method of molecular dynamics (MD) [47]. Of particular concern has been the
influence of such fast compression on defects such as voids. An important result obtained for
collapse of a void in a simple atomic lattice into a dislocation defect structure [48] has been the
observation that such permanent relaxation converted the higher potential energy of atoms
adjacent to the void into thermal hot spot energy, thus, giving evidence that such permanent
deformation process always generates significant thermal energy above the normal atomic
lattice modes. Recent model computations have been presented for large-scale MD simulation
of the orientation dependence of shock deformations in a face-centered-cubic lattice [49]. Such
experiments possibly relate to pioneering measurements made on the orientation dependence of
run-to-detonation measurements made on body-centered-tetragonal PETN crystals by Dick
and attributed by him to the orientation dependence of shearing actions in the perfect lattice [50],
akin to the model considerations given in references [21] and [22].
Acknowledgements
The present article has drawn heavily on reference [1] with benefit from association with
my co-author there, Wayne Elban. Again, appreciation is expressed both to Byron Allmon for
help given in compiling Figures; and to the following agencies who have supported the overall
research effort: the Office of Naval Research, at Arlington, VA; the Naval Surface Weapons
Center, at the White Oak Laboratory, Silver Spring, MD; the Naval Surface Warfare Center, at
Indian Head, MD; the Air Force Research Laboratory, AFRL/MNME, at Eglin Air Force Base,
FL; and the Center for Energetic Concepts Development (CECD), University of Maryland, at
College Park.
44
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
R.W. Armstrong and W.L. Elban: Dislocations in Energetic Crystals, (Dislocations in Solids,
F.R.N. Nabarro and J.P. Hirth, Eds.), Elsevier B.V., Amsterdam, The Netherlands, Vol. 12,
Chap. 69, p. 403-446, 2004.
H. Klapper: Defects in Non-metal Crystals (Characterization of Crystal Growth Defects by X-ray
Methods, B.K. Tanner and H.K. Bowen, Eds.), Plenum Press, N.Y., p. 133, 1980.
E.N. Farabaugh: Ph.D. Thesis, University of Maryland, 1977; see R.W. Armstrong,
Characterization of Materials by X-ray Diffraction Topography, (Crystal Properties and
Preparation), Trans Tech Publ., Switzerland, Vol.16, p. 1, 1988.
A.C. van der Steen and W. Duvalois: What do Explosive Particles Look Like (ONR/TNO
Workshop on Desensitization of Explosives and Propellants, A.C. Van der Steen, Ed.), TNO
Prins Maurits Laboratory, Rijswijk, The Netherlands, Vol. 3, p. 1, 1991.
R.W. Armstrong: Dislocation Mechanisms for Shock-induced Hot Spots, J. de Phys. IV-Coll. 5,
C4-89, 1995.
I.T. McDermott and P.P. Phakey: A Method of Correlating Dislocations and Etch Pits:
Application to Cyclotrimethylene Trinatramine, J. Appl. Phys. 4, 479, 1971; An X-ray
Topographic Study of Defect Structure, Phys. Stat. Sol. (a) 8, 505, 1971.
A.R. Lang: X-ray Diffraction Topography, (Diffraction and Imaging Techniques in Materials
Science, S. Amelinckx, R. Gevers and J. van Landuyt, Eds.), North-Holland Publ., Amsterdam,
Vol. II, p. 678, 1978.
J.P. Hirth and J. Lothe: Theory of Dislocations, McGraw-Hill Book Co., NY, Part 2, p. 201,
1968.
R.W. Armstrong and W.L. Elban: Cracking at Hardness Micro-indentations in RDX Explosive
and MgO Single Crystals, Mater. Sci. Eng. A111, 35, 1989.
F.C. Frank and B.R. Lawn: On the Theory of Hertzian Fracture, Proc. Roy. Soc. Lond. A299,
291, 1967.
R.W. Armstrong and A.C. Raghuram: Anisotropy of Micro-hardness in Crystals, (The Science of
Hardness Testing and Its Research Applications, J.H. Westbrook and H. Conrad, Eds.) ASM,
Metals Park, OH, p. 174, 1973.
J.T. Hagan and M.M. Chaudhri: Fracture Surface Energies of High Explosives PETN and RDX,
J. Mater. Sci. 12, 1055, 1977.
K.-C. Yoo, R.G. Rosemeier, W.L. Elban and R.W. Armstrong: X-ray Topography Evidence for
Energy Dissipation at Indentation Cracks in MgO Crystals, J. Mater. Sci. Lett. 3, 560, 1984.
W.L. Elban, R.G. Rosemeier and R.W. Armstrong: Summary Report: Microstructural Origins of
Hot Spots in RDX Explosive and Several Reference Inert Materials, Naval Surface Weapons
Center, Silver Spring, MD, Report NSWC MP 84-358, 1984.
B.R. Lawn and E.R. Fuller: Equilibrium Penny-like Cracks in Indentation Fracture, J. Mater.
Sci. 10, 2016, 1975.
W.L. Elban: Surface Energies of High Explosives PETN and RDX from Contact Angle
Measurements, J. Mater. Sci. 14, 1008, 1979.
B.L. Hammond and R.W. Armstrong: Recovered Elastic and Plastic Strains at Residual Microindentations in an MgO Crystal, Philos. Mag. (Lett.) 57, 41, 1988.
W.L. Elban, R.W. Armstrong and T.P. Russell: Plasticity/Interfacial Energy Influences
onCombustion-driven Cracking of RDX Energetic Crystals, Philos. Mag. A78, 907, 1998.
45
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
R.W. Armstrong and W.H. Robinson: Combined Elastic and Plastic Deformation Behavior from
a Continuous Indentation Hardness Test, New Zealand J. Sci. 17, 429, 1974.
B.R. Lawn: Hertzian Fracture in Single Crystals with the Diamond Structure, J. Appl. Phys. 39,
4828, 1968.
R.W. Armstrong and W.L Elban: ONR Workshop on Energetic Material Initiation
Fundamentals, Los Alamos Nat. Lab., Chem. Publ. Inform. Agency, CPIA Publ. 475, p. 171,
1987.
J.J. Dick, R.N. Mulford, W.J. Spencer, D.R. Pettit, E. Garcia and D.C. Shaw: Shock Response of
Pentaerythritol Tetranitrate, J. Appl. Phys. 70, 3572, 1991.
J.C. Hoffsommer, D.J. Glover and W.L. Elban: Quantitative Evidence for Nitroso Compound
Formation in Drop-Weight Impacted RDX Crystals, J. Energetic Mater. 3, 149, 1985.
R. Behrens Jr. and S. Bulushu: Thermal Decomposition of Energetic Materials. 3. Temporal
Behaviors of the Rates of Formation of the Gaseous Pyrolysis Products from Condensed Phase
Decomposition of 1,3,5-Trinitrahexahydro-s-triazine(RDX), J. Phys. Chem. 96, 8877, 1992.
J.J. Gilman: Direct Evidence of Chemical Reactions Induced by Shear Strains, (Synthesis,
Characterization, and Properties of Energetic/Reactive Nanomaterials, R.W. Armstrong, N.N.
Thadhani, W.H. Wilson, J.J. Gilman and R.L. Simpson, Eds.) Materials Research Society,
Warrendale, PA, Proc. Vol. 800, p. 287- 297, 2004.
F.P. Bowden and A.D. Yoffe: Initiation and Growth of Explosion in Liquids and Solids,
Cambridge University Press, London, 1952; Fast Reactions in Solids, Butterworth Scientific
Publ., London, 1958.
S.N. Heavens and J.E. Field: The Ignition of a Thin Layer of Explosive by Impact, Proc. Roy.
Soc. Lond., A338, 77, 1974.
R.W. Armstrong, C.S. Coffey and W.L. Elban: Adiabatic Heating at a Dislocation Pile-up
Avalanche, Acta Metall. 30, 2111-2118, 1982.
R.W. Armstrong: Plasticity: Grain Size Effects II, (Encyclopedia of Materials: Science and
Technology Updates, K.H.J. Bischow, R.W. Cahn, M.C. Flemings, E.J. Kramer, S. Mahajan
and P. Veyssiere, Eds), Elsevier Sci. Ltd., Oxford, UK, in print, 2005.
R.W. Armstrong, C.S. Coffey, V.F. DeVost and W.L. Elban: Crystal Size Dependence for
Impact Initiation of Cyclotrimethylenetrinitramine, J. Appl. Phys., 68, 979, 1990.
A.T. Nielson: Impact Sensitivity Versus Particle Size for RDX and Octanitrobenzidene (CL-12),
(Working Group Meeting on Sensitivity of Explosives, Center for Energy Technology and
Research), New Mexico Institute of Technology, Soccoro, p. 256, 1987.
R.W. Armstrong, K. Kline, M.P. Kramer and W.H. Wilson: The Power of Energetic
Nanomaterials (Twenty-Ninth International Pyrotechnics Seminar Proceedings, F.J. Schelling,
Ed.), IPSUSA, Inc., p. 239 (Abstract for unpublished presentation), 2002.
R.W. Armstrong, A.L. Ramaswamy and J.E. Field: Thermomechanical Influences on the
Combustion of RDX Crystals, (ONR/SNPE/ONERA Workshop on Combustion Mechanisms,
R.W. Armstrong, Ed.) ONR, London, UK, p. 168, 1991.
R.W. Armstrong, C.F. Clark and W.L. Elban: Influence of Micro-cracking on Pressuredependent Energetic Crystal Combustion, (Combustion of Energetic Materials, K.K. Kuo and L.
DeLuca, Eds.) Begell House, Inc., NY, p. 354, 2003.
K.A. Gonthier: Modeling and Analysis of Reactive Compaction for Granular Energetic Solids,
Technical Report AFRL-MN-EG-TR-2001-7091, Eglin AFB, August, 2001.
S.J. Jacobs, H.W. Sandusky and W.L. Elban: Quasi-static Compaction of Porous Propellant
Beds. I. Modeling Ball Powder Experiments with Deformed Spheres in a Regular Lattice,
Powder Technology, 89. 209-217, 1996.
46
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
G.T. Gray III, M.A. Meyers, N.N. Thadhani, and K.S. Vecchio: Symposium on Dynamic
Deformation, Metall. Mater. Trans. 35A, pp. 2542-2745, 2004.
C.S. Smith: Metallographic Studies of Metals after Explosive Shock, Trans. TMS-AIME, 212,
574, 1958.
W. Arnold: Dynamische Werkstoffverhalten von Armco-Eisen bei Stosswellenbelastung,
Fortschritt-Beriche VDI-Verlag Gmbh, Dusseldorf Germany, 1992.
R.W. Armstrong, W. Arnold, and F.J. Zerilli: Grain Size Dependence of Shock-induced
Twinning Stresses for Armco Iron (International Workshop on New Models and Hydrocodes for
Shock Wave Processes in Condensed Matter), Edinburgh, UK, Chemical Physics (Russian), in
print, 2002.
Approches Microscopique et Macroscopique des Detonations, S. Odiot (Ed.) J. de Physique,
Coll C4, Suppl. 9, 48, pp. 1-433, 1987.
J.J. Gilman: Shear-induced Metallization, Philos. Mag. B, 67 207, 1993; Chemical Reactions at
Detonation Fronts in Solids, Philos. Mag. B, 71, 1057, 1995.
R.W. Armstrong, H.W. Sandusky and R.S. Miller: Indentation Hardness Testing, Defect
Structure, and Shock Model for RDX Explosive Crystals, (ONR Workshop on Dynamic
Deformation, Fracture and Transient Combustion), Chemical Propulsion Information Agency,
CPIA Publ. 474, p. 77, 1987.
F.A. Bandak, R.W. Armstrong and A.S. Douglas: Dislocation Structure for One-dimensional
Strain in a Shocked Crystal, Phys. Rev. B, 46, 3228, 1992.
F.A. Bandak, D.H. Tsai, R.W. Armstrong and A.S. Douglas: Formation of Nanodislocation
Dipoles in Shock-compressed Crystals, Phys. Rev. B, 47, 11681, 1993.
D.H. Lassila, T. Shien, B.Y. Cao and M.A. Meyers: Effect of Low-temperature Shock
Compression on the Microstructure and Strength of Copper, Metall. Mater. Trans. 35A, 27292739, 2004.
D.H. Tsai: Structural Defects and Hot Spot Formation in a Crystalline Solid under Rapid
Compression. I. Vacancy Clusters and Slip bands, J. Chem. Phys. 95, 7497, 1991.
D.H. Tsai and R.W. Armstrong: Defect-enhanced Structural Relaxation Mechanism for the
Evolution of Hot Spots in Rapidly Compressed Crystals, J. Phys. Chem. 98, 10997.1994.
T.C. Germann, D. Tanguy, B.L. Holian, T.S. Lomdahl, M. Mareschal and R. Ravelo: Dislocation
Structure Behind a Shock front in FCC Perfect Crystals: Atomistic Simulation Results, Metall.
Mater. Trans., 35A, 2609-2615, 2004.
J.J. Dick: Effect of Crystal Orientation on Shock Initiation Sensitivity of Pentaerythritol
Tetranitrate Explosive, Appl. Phys. Lett. 60, 2494, 1992.
47
A QUANTUM CHEMICAL STUDY

ON THERMOLYSIS INITIATION MECHANISMS
Heming Xiao*, Zhaoxu Chen**, and Xiaojuan Xu*
* Department of Chemistry, Nanjing University of Science and Technology,
Nanjing 210094, P.R.China
** Institute of Theoretical and Computational Chemistry, Nanjing University,
Nanjing 210093, P.R.China
Abstract:
Thermolysis initiation mechanisms of polynitro compounds, tetrazole derivatives and
their metallic salts, and cage high energy density compounds have been investigated
using quantum chemical approaches. Our calculations show that the trigger bonds whose
breaking initiate a decomposition or an explosion are C-NO2 or N-NO2 bonds for nitro
derivatives of benzene and aminobenzenes, CL-20 and polynitroadamantanes. Explosion
of nitro derivatives of phenol and toluene are most likely triggered by the isomerization
reactions involving the H-shift. Due to larger strain energy, the trigger bond is found to be
the C-C bond in the framework of polynitrocubanes. Regarding tetrazoles and their
metallic salts, opening of the tetrazole ring, i.e., scission of the N-N bond, followed by
formation of N2 molecules, initiate explosive reactions. We found for energetic materials
having similar molecular structures and following similar thermal decomposition
mechanisms, the bond orders of the trigger bond and the activation energy to break the
bond are directly related to the impact sensitivity. We thus proposed two criteria used to
evaluate the relative ordering of impact sensitivity of energetic materials with similar
structures: the smaller the bond order, the more sensitive an energetic material, which is
called the principle of the smallest bond order (PSBO). And the higher the activation
energy, the less sensitive a material. We demonstrated that in most cases the principle of
the smallest bond order (PSBO) is equivalent to the activation energy criterion. The
former is more convenient and easier to obtain while the latter can be applied more
universally.
Kyewords: thermolysis initiation mechanisms, N-N, C-NO2 or N-NO2 bonds,
nitro derivatives of benzene, nitro derivatives, of aminobenzenes, CL-20
and polynitroadamantanes
48
1.
INTRODUCTION
Thermal decomposition directly related to the stability and deflagration of energetic

materials. It is an important factor affecting the safety of production, usage and storage of
energetic materials. Therefore, study of thermolysis mechanism is of great significance, and is
a necessity in the field of explosive chemistry and technology. In the course of studying
thermolysis mechanisms, we, in particular, paid great attention to the trigger step or initiation
step not only because trigger step is usually rate-limiting step determined by molecular
structure (The subsequent reaction steps are automatically accelerated, and thus not so
important in the sense of determining the ordering of sensitivity), but also because the
sensitivity of explosives is directly associated with thermolysis initiation step.
Basing on the basic theory that thermolysis takes ahead of an explosion and the idea about
the thermal entity of initiation [1-4], people always put thermolysis, initiation and sensitivity
together to study.
During the last two decades, we have systematically investigated the mechanisms of
thermal decomposition for various series of energetic materials including nitrocompounds,
nitroamines, nitroesters, azides, tetrazloe derivatives and their metallic salts, metallic salts of
NTO, specially organic cage high energy density compounds. In these studies, we adopted
quantum chemical methods, combined with some experimental approaches. We found that the
relative order of impact sensitivity of energetic materials can be predicted by examining the
bond order of the trigger bond or activation energy of the trigger bond-breaking [5-12]. Most of
the achievements are summarized in the monographs [13-15], respectively. Here we present and
discuss some selected typical compounds only.
2.
POLY-NITRO AROMATIC COMPOUNDS
Much attention has been paid to polynitro aromatic compounds. TATB

(1,3,5-triamino-2,4,6-trinitrobenzene) and TNT (2,4,6-trinitrotoluene) et al. are among the
most effective and widely-used explosives and monopropellants.
49
2.1
Nitro derivatives of benzene and aminobenzenes [7-8]
Table 1. The AM1 calculated Wiberg bond order of the weakest C-NO2 bonds in the
nitro derivatives of benzene and aminobenzenes (BC-N), the activation
energies (Ea) of pyrolysis initiation reactions and the experimental impact
sensitivity of them (h50%)
Compounds
BC-N
Ea (kJ/mol)
h50% (m)
nitro derivatives of benzene

nitrobenzene
m-dinitrobenzene
p-dinitrobenzene
s-trinitrobenzene
o-dinitrobenzene
1,2,3,4-tetranitrobenzene
hexanitrobenzene
0.857
140.069
75.83
0.852
131.429
3.75
0.848
129.379
0.844
124.226
0.60
0.839
118.996
0.812
96.678
0.797
84.185
0.05
nitro derivatives of aminobenzenes

TATB(1,3,5-triamino-2,4,6-trinitrobenzene)
0.934 157.21
o-nitroaniline
0.889 146.21
DATB(1,3-diamino-2,4,6-trinitrobenzene)
0.882 140.40
2,6-dinitroaniline
0.880 135.35
TNA(2,4,6-trinitroaniline)
0.871 130.64
2,3,4,6- tetranitroaniline
0.806
95.94
pentanitroaniline
0.801
88.63
>>3,20
3.20
1.77
0.41
0.15
UHF-SCF-AM1 method [16] has been applied to study the pyrolysis initiation reactions for
seven nitro derivatives of benzene and seven nitro derivatives of aminobenzenes. From Table
1 one can see that for each series of compounds there is a good linear relationship between the
bond order (BC-N)of the trigger bond C-NO2 and the activation energy (Ea) of the homolysis
reaction of the bond. Comparing the experimental impact sensitivity values(H50%) (BC-N) and
Ea, one finds that there is parallel relationship between them. That is, for homologous
compounds, the smaller the bond order of the trigger bond, (there is less electron density
around this bond, thus the bond is relatively weak.) the lower the activation energy to break the
bond. Therefore thermal decomposition proceeds more easily. Accordingly it is easier for this
compound to be initiated to explosion or equivalently its sensitivity is larger. The finding that
the smallest bond order corresponds to the largest sensitivity is called the principle of the
smallest bond order (PSBO) used for estimating the sequence of impact sensitivity. In the case
of nitro derivatives of benzene and aminobenzenes, PSBO leads to the same conclusion as that
derived from the activation energy (Ea) criterion.
Table 2 lists Mulliken bond order (MC-N) calculated at the density functional theory
(DFT)-B3LYP [17,18]/6-31G* level and the experimental impact sensitivity values. Data at the
first -principles level (Table 2) further corroborate [18] the validity of PSBO previously
derived based on semi-empirical molecular orbital calculations.
50
Table 2. Table 2 The Mulliken bond order of the weakest C-NO2 bonds (MC-N)
from DFT-B3LYP/6-31G* calculations and the experimental impact
sensitivity (h50% and E)
compounds
m-dinitrobenzene
s-trinitrobenzene
1,2,3,5-tetranitrobenzene
pentanitrobenzene
hexanitrobenzene
TATB
DATB
TNA
2,3,4,6- tetranitroaniline
pentanitroaniline
2.2
MC-N
h50% (m)
0.1611
0.1511
71
0.1346
33
0.1211
7
0.1323
15
0.1788
>>320
0.1640
320
0.1567
177
0.1350
041
0.1169
15
E/Nm
39
74
Nitro derivatives of phenols and methylbenzenes [9,10]
Table 3 and Table 4 list the results of the nitro derivatives of Phenols and methylbenzenes
calculated by UAM1 method.
For nitro derivatives of phenols, the activation energy for breaking C-NO2 bond Eah is
larger than that for the H-shift isomerization reaction Eai . Hence it is reasonable to assume
that O-H bond is the trigger bond in nitro derivatives of phenols. In fact, the experimental
impact sensitivity results vary in parallel with the bond order (BH-O) and Eai.
Our calculated Eah for cleaving the C-NO2 bond in nitro derivatives of methylbenzenes is
larger than the one for the H-shift isomerization reaction (Eai), indicating that the trigger bond
is the -C-H bond. There is basically a parallel relationship between Eai and the impact
sensitivity values.
Table 3. Table 3 The experimental impact sensitivity (h50%) of six nitro derivatives
of Phenols, the activation energies of two kinds of reactions (Eah and Eai)
and the Wiberg bond order in the reactant molecules (BC-N and BH-O)
No. of compounda
(1)
(2)
(3)
(4)
(5)
(6)
h50%
>3.2
2.96
0.87
0.43
0.27
h
b
Ea (kJ/mol)
136.859 146.289 143.369 133.691 132.495 142.812
i
Ea (kJ/mol)
94.889 97.613 81.350 85.475 72.094 67.354
BC-Nc
0.859
0.879
0.885
0.852
0.858
0.886
BH-O
0.886
0.885
0.878
0.877
0.869
0.866
a
(1),(2),(6) indicate 2,4-dinitrophenol, 4,6-dinitroresorcinol, 2,4-dinitroresorcinol, 2,4,6-tri-nitrophenol,

2,4,6-trinitroresorcinol and 2,4,6-trinitro-1,3,5-trihydroxyl benzenes, respectively
Eah and Eai correspond to the activation energies of the homolysis and isomerization reactions, respectively
BC-N and BH-O stand for the Wiberg bond orders of the weakest C-NO2 and H-O bonds in the reactant molecules,
respectively
51
Table 4. Table 4 The experimental impact sensitivity (impact energy) of seven nitro
derivatives of methylbenzenes, the activation energies of two kinds
of reactions (Eah and Eai)
impact energy
Eah
Eai
(10-4kg/m2)
(kJ/mol)
(kJ/mol)
2,4-dinitromethylbenzene
18.9
134.076
79.194
4,6-dinitro-1, 3- dinitromethylbenzene
14.6
129.691
74.760
2,4-dinitro-1, 3- dinitromethylbenzene
120.068
68.572
2,4-dinitro-1, 3,5- trinitromethylbenzene
13.8
121.650
68.382
2,4,6- trinitromethylbenzene
11.4
114.504
64.446
2,4,6-trintro-1,3- dinitromethylbenzene
5.7
111.516
61.768
s- trinitromethylbenzene
5.9
112.813
69.626
Compounds
3.
TETRAZOLE DERIVATIVES AND THEIR METALLIC SALTS [11-12]
Tetrazole is a kind of five-membered heterocyclic compound containing four nitrogen

atoms. Tetrazole derivatives and their metal salts (complexes) have been investigated as
detonators to replace lead azide. A series of investigations on tetrazole derivatives show that
5-substituted (or C-substituted) tetrazoles are more stable than the 1- and 2-substituted (or
N-substituted) isomers. The impact sensitivity of 5- substituted tetrazole derivatives follows
the following order [20,21]: R: CH3 < NH2 < H < NHNO2 < Cl < NO2 < N3 < N+2 (R are
substitutents; see Fig. 1). The sensitivity of some metal salts of tetrazole derivatives has also
been measured [22] (see Table 6). Various efforts have been made to explain the experimentally
observed sensitivity order. Unfortunately none of them can explain the observed order of
sensitivity satisfactorily.
Fig 1 illustrates the thermal decomposition of C-substituted tetrazole derivatives and their
metal
salts. A systematic investigation on a series of tetrazole derivatives shows that the
rate-controlling step of thermal decomposition is the first step, i.e., the opening of the tetrazole
ring. Thus it is reasonable to predict the sensitivity order for tetrazole derivatives just by
comparing the activation energy of this step. We employed both DFT-B3LYP and
semi-empirical PM3 [23] methods.
C5
N4
N3
N1
N2
Reactant
H(M)
C5
N1
N4
N3
H(M)
N2
Transition State
52
C5
N4
N3
N1
N2
Product
H(M)
Fig. 1. Illustration of the thermolysis initiation step of 5-substituted tetrazoles and their
metallic salts(R=substituent, M=metal atom for tetrazole metal salts)
3.1
5-Substituted tetrazole derivatives
Table 5. Activation energies obtained from B3LYP/6-31Gand PM3 calculations,

respectively (unit: kJ/mol,).
R
N+2
OH
N3
NO2
Cl
NHNO2
NH2
H
CH3
No.
1
2
3
4
5
6
7
8
9
Ea DFT
89.02
97.46
103.01
108.21
103.24
100.08
99.53
110.14
114.32
Ea(PM3)
99.11
124.45
132.16
132.76
133.66
134.24
134.55
138.05
140.85
It is notable in Table 5 that all activation energies from B3LYP calculation are smaller
than the PM3s results. The average difference between two sets is 27.19 kJ/mol. A close look
at the data in Table 5 shows that PM3 produces the similar trend for activation energy as
B3LYP. For example, both methods predict that 5- N+2-tetrazole has the smallest reaction
barriers (89.02, 99.11 kJ/mol, respectively), and 5-methyltetrazole the largest ones (114.32,
140.85 kJ/mol, respectively). The lower the barrier, the more easily the thermal decomposition
proceeds. Thus the compound will be more sensitive, provided that the reaction barrier of this
step is the decisive factor for detonation of 5-substituted tetrazoles. According to the obtained
activation energies from B3LYP calculation, we might conclude that 5-N+2-tetrazole is the
most sensitive whereas 5-methyltetrazole the most stable, which is in line with experiments.
Based on the calculated activation energies, we can obtain the following theoretical sensitivity
order for the eight derivatives (1, 3, 4, 5, 6, 7, 8, and 9): CH3 < H < NO2 < Cl < NHNO2 < N3
< NH2 < N+2.
Comparing the experimental sensitivity ordering (CH3 < NH2 < H < NHNO2 < Cl < NO2
< N3 <N+2), one can see that there are some discrepancies between the two sets. In our opinion
there are two reasons for it. First, the discrepancies might be due to the different
decomposition pathway undergone by 5-nitranminotetrazole, 5-nitrotetrazole, and
5-aminotetrazole. The self-consistent reaction field (SCRF) calculations show that unlike
most C-substituted tetrazoles, 5-NO2- and 5-NHNO2-tetrazole derivatives exist mainly in the
2H-tetrazole (not 1H-isomer) forms in condensed phase. Therefore in the cases of 5-nitro- and
5-nitramino tetrazoles, explosion is initiated, most probably by the decomposition of
5-substituted 2H-tetrazole derivatives. For 5-aminotetrazole, there is indeed experimental
evidence indicating that the pyrolysis mechanism of 5-aminotetrazole is different from Fig 1.
The second reason for the discrepancies is the smaller activation gap (see especially the
reaction barrier from compound 3 to compound 7), which is beyond the accuracy of our
theoretical calculation. Analysis of activation energy from PM3 calculation shows that the
53
activation energies Ea (PM3) are in the following increasing order: N+2 > N3 > NO2 > Cl >
NHNO2 > NH2 > H > CH3, in accordance with the experimental sensitivity order except for
1H-tetrazole and 5-aminotetrazole.
3.2
Metallic salts of 5-substituted tetrazole derivatives
Table 6. Activation energies obtained from PM3 calculations(unit: kJ/mol) and

relative sensitivity(s)
Compounds
Mercuric
Mercurous
Mercuric
Lead
Thallous
Potassium
NitroTzc
NitroTz
nitroaminoTz
nitroaminoTz
AzoTz
NitroTz
(I)
(II)
(III)
(IV)
(V)
(VI)
Ea
130.76
156.68
183.43
217.01
232.94
233.36
15
50
70
100
150
200
Table 6 contains the activation energies and the experimental impact sensitivity. It is
notable from Table 6 that the smaller the activation energy, the smaller S is thus the more
sensitive the compound is. Fig. 2 illustrates the relationship between the calculated activation
energies E and the sensitivity S. From Table 6 we can see that the calculated activation energy
for potassium nitrotetrazole VI is a little smaller, which results in that the last point in Fig. 2
deviates from the line a little. This might be due to the use of the potassium-like sparkle model
as the parameter for potassium in the program. Regressing the line for the first five points, we
obtained:
Ea=s122.6+0.779S
The correlation coefficient is 0.973. The higher correlation coefficient indicates that the
proposed substituent model for treating the metal salts of tetrazole derivatives is successful.
Ea[Activaiton Energy (kJ/mol)]
240
220
200
180
160
140
120
0
50
100
150
200
S (S e n s itiv ity )
Fig. 2. Activation energy of the N (1)-N (2) bond breaking to the experimental impact
sensitivity of some tetrazole metal salts
54
4.
ORGANIC CAGE HIGH ENERGY DENSITY COMPOUNDS
High energy density materials (HEDMs) have been receiving heated attention because of
superior explosive performances over conventional energetic materials. Due to strain energy
and compact structures, organic cage compounds are investigated as an important category of
HEDMs. Typical examples of these compounds are hexanitrohexaazaisowurtzitane (Cl-20)
and octanitrocubane (ONC) [24]. Polynitrocubanes and polynitroadamantanes are also
investigated as promising candidates of HEDMs [25].
4.1
-CL-20 (HNIW) [15]
Fig3 gives the molecular structure of -CL-20. Table 7 and Table 8 list bond orders and
activation energies (Ea) of homolysis different kinds of the weakest bond C-C, C-N and N-N
from UPM3 calculations, respectively.
O 15
O 14
O 29
N 13
N2
C7
O 23
C 10
N9
N 22
C1
N 28
N 12
O 30
N 11
N 25
O 24
O 27
O 26
C8
C6
C3
N4
N5
O 21
N 19
N 16
O 20
O 17
O 18
Fig. 3. Molecular structure of -CL-20
Table 7. The Mulliken bond order of -CL-20 from UPM3 calculations

C-C
Bond order
C1-C3
0.275
C7-C10
0.279
C6-C8
0.288
C-N
Bond order
C1-N2
0.217
C1-N9
0.223
C7-N2
0.216
C10-N9
0.231
C8-N5
0.229
C8-N12
0.225
C6-N4
0.233
C6-N11
0.230
55
N-N
N2-N13
N12-N28
N4-N16
N5-N19
N9-N22
N11-N25
Bond order
0.172
0.177
0.210
0.217
0.186
0.194
Table 8.
The heat of formation and the activation energies(Ea) in different themolysis

reaction approaches of CL-20(unit: kJmol1)
Homolysis
Bond
C1-C3
C7-N2
N12-N28
N2-N13
heat of formation
R
TS
1007.5
1139.9
1007.5
1196.8
1007.5
1107.7
1007.5
1103.7
Ea
132.4
189.3
100.2
96.2
From Table 7,we can see that the bond whose bond order is the smallest in the same class,
i.e., the weakest bond is C1-C3, C7-N2, N2-N13 and N12-N28, respectively.
Its shown from Table 8 and Fig4 that the activation energies for the homolysis reaction of
N-NO2 bond, 96.2 and 100.2 kJ mol-1, are notably smaller than those for breaking the C-C and
C-N bonds in the cage framework, 132.4 and 189.3 kJ mol-1, indicating that the initiation step
for -CL-20 is homolysis of the N-NO2 bond on the side chain, and the framework made of
cyclohexane and cyclopentane is rather stable.
C1-C3
C7-N 2
N2-N 13
10
680
heat of f or mat i on( KJ / mol )
660
640
620
600
580
560
540
520
500
480
460
440
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
0
3.2
r eac t i on c oodi nat e
Fig. 4. Potential energy curves of C-C, C-N and N-N in the homolysis reaction
of -CL-20 from PM3 calculation
4.2
Polynitrocubanes [26]
Fig 5 shows the thermolysis mechanism of cubane. The diradical intermediate P1 is

produced through the homolysis of the skeletal C-C via the transition state TS1. Further
breaking of another C-C bond (via TS2) leads to the product. When 1,4-ditrinocubane and
octanitrocubane are talked about, we just need replace -H with NO2 radical and this figure is
also applicable.
56
.
R
TS1
TS2
P1
P2
Fig. 5. Thermolysis mechanism of cubane

In order to elucidate the pyrolysis mechanism of polynitrocubanes,two possible initial
steps in the pyrolysis route are considered: (1) breaking the first C-C bond of the cube to form
a transition state TS1 and diradical intermediate P1 and subsequently breaking the second
C-C bond in P1 to form TS2 and product P2. (2) breaking a C-NO2 bond to form a TS3 and the
product P3 containing two radicals.
Table 9. MINDO/3 calculated the activation energies(kJmol1)*
1,4-dinitrocubane
ONC
*
Rupture of C-C bond

Ea1
Ea2
133.17
2.03
86.65
9.99
Rupture of C-NO2 bond

Ea3
149.94
102.66
Ea1 and Ea2 are the activation energies via the pyrolysis routes of R-TS1-P1 and P1-TS2-P2
respectively, Ea3 are the activation energies via pyrolysis route for the rupture of C-NO2 bond
From Table 9, we can see that Ea3 for breaking the C-NO2 bonds in both ONC and
1,4-dinitrocubane molecules are larger than the corresponding Ea1 for breaking the first C-C
bonds in their cubic structures. This result indicates that the initial step in the pyrolysis of
polynitrocubanes is the rupture of the C-C bond on cubane cage skeleton, which is different
from the pyrolysis initiation step via breaking the bond between the ring and NO2 for other
caged energetic compounds such as Cl-20 and polynitroadamantanes, and for many noncaged
energetic compound such as TATB, RDX and HMX. Obviously, this difference is because the
strain energies on cubane cage skeletons are much more than those on the ring in other caged
and non-caged molecules.
From Table 9 we can also find that for the two compounds, Ea2 for breaking the second
C-C bonds are much smaller than Ea1 for breaking the first C-C bond in their cubes. This
means that compared with the first C-C bond, the rupture of the second C-C bond is very easy.
So, in the pyrolysis initiation reaction of polynitrocubane, the rate-controlling step, is to form
a diradical by the single C-C bond breaking in the cube. The second C-C bond breaking is
easily followed to form a nitrocyclooctatetraene for ONC. The corresponding energy level
diagram calculated by MINDO/3 method in the pyrolysis of ONC is shown in Fig 6.
57
heat of formation (kJ/mol)
350
300
250
200
150
100
50
0.1
0.15
0.2
0.25
0.3
0.35
rc-c (nm)
Fig. 6. Thermolysis potential energy curve of octanitrocubane from MINDO/3

calculation
4.3
Polynitroamadantane
In order to elucidate the pyrolysis mechanism of PNAs, two possible initial steps in the
pyrolysis route were considered: (1) breaking the C-NO2 bond. (2) breaking the C-C bond on
the skeleton. It should be pointed out that, in this work, we selected the weakest C-C or C-NO2
bond based on the Mulliken population analyses as initially breaking bond. Table 10 gave the
results of the homolysis of the C-C and C-NO2 bonds using UHF-PM3 method.
C(1)
C(8)
C(2)
C(9)
C(7)
C(3)
C(5)
C(10)
C(6)
C(4)
Fig. 7. The structure and atomic numbering of adamantine

From Table 10, we can see that, Ea for breaking C-NO2 bonds are much smaller than the
corresponding activation energies for breaking C-C bonds of PNAs. These results indicate that
the initial step in the pyrolysis of PNAs is the rupture of C-NO2 bond, not C-C bond on cage
skeleton. As a whole, with the number (n) of nitro groups increasing, Ea for the homolysis of
C-NO2 bonds decrease. In addition, we found that Ea are also relative to the electronic
structure parameters, such as bond orders of C-N (BC-N) of PANs. From Fig. 8, we can see that
there is the good relationships between Ea and BC-N with correlation coefficients 0.9747,
which proves the reliability of PSBO again. And the thermal stability and impact sensitivity
for PANs can be identified.
58
Table 10. The HOFs of the Reactants, Transition States and Activation Energies( Ea)
for the Homolysis of C-C and C-NO2 Bond of PNAs
Heat
Compoundsa
of
formation (kJ.mol-1)
Reactants
TS
C-C
Ea(kJmol1)
C-NO2
C-C
C-NO2
1-
477.60
663.77
621.66
186.17
144.06
2-
485.00
668.65
630.12
183.65
145.12
1,2-
498.92
661.15
630.01
162.23
131.09
1,3
483.23
674.17
621.27
191.04
138.04
1,4-
487.09
674.11
628.24
187.02
141.15
2,4-
498.44
683.07
638.79
184.63
140.35
2,6-
493.05
677.41
635.48
184.36
142.43
2,2-
535.73
707.09
635.98
171.36
100.25
1,3,5-
499.07
695.03
631.41
195.96
132.34
2,4,6-
515.96
1,4,4-
545.08
719.10
641.72
174.02
96.64
1,2,3-
533.28
694.21
650.30
160.93
117.02
1,3,8-
510.85
636.63
125.78
1,3,6-
500.88
635.56
134.68
1,3,5,7-
523.96
2,4,6,8-
555.99
2,2,4,4-
649.74
1,4,4,7-
564.62
659.76
95.14
2,2,6,6-
606.21
703.64
97.43
651.23
729.43
651.24
135.27
205.47
682.18
826.75
731.40
127.28
126.19
177.01
81.66
1,3,4,5,7-
592.61
692.52
99.91
2,4,6,8,10-
598.29
721.31
123.02
2,2,4,4,6,6-
778.29
848.69
70.40
1,3,4,5,6,7-
657.83
750.84
93.01
2,4,6,8,9,10-
642.75
761.76
119.01
1,2,3,4,5,6,7-
743.90
837.25
93.35
1,3,4,4,5,7,8-
745.42
819.33
73.91
2,2,4,4,6,6,8,8-
979.53
1042.72
63.19
1,2,3,4,5,6,7,8-
820.39
913.29
92.90
1,2,3,4,5,6,7,8,9-
932.45
1026.02
93.57.
1,2,3,4,5,6,7,8,9,10-
1046.12
1121.99
75.87
2,2,4,4,6,6,8,8,10,10-
1343.26
1369.28
26.02
1,2,3,4,4,5,6,7,8,9,10-
1198.78
1223. 55
24.77
1- and 1,2- denote 1-nitroadamantane and 1,2-dinitroadamantane, respectively,

the others are similar.
59
Based on quantum-chemical calculation of heats of formation, molecular volumes and

theoretical densities, using Kormlet-Jacobs equation [27], the detonation velocity (D) and
detonation pressure (P) of PANs also obtained. From Table 11, we can find that only when the
number of nitro groups of PNA is equal to or more than eight, can it be possible for PANs to
used as a HEDM [15].
160
Activaton energy Ea(kJ/mol)
R=0.9747
140
120
100
80
60
40
20
0.16
0.18
0.20
0.22
0.24
0.26
0.28
BC-N
Fig. 8. Linear relationship between Ea and C-N bond orders (BC-N)
Table 11. The Average Volumes (V), theoretical Densities(),

Detonation Velocities(D) and Pressures (P) of PNAsa
V
1-
135.49
1.34
4.32
6.77
2-
135.50
1.34
4.36
6.92
1,2-
151.45
1.49
5.99
14.12
Compdsb
180.78
1.75
7.35
23.58
2,2,6,6-
178.97
1.77
7.32
23.89
1,3,4,5,7-
200.97
1.80
7.75
26.67
Compdsb
2,2,4,4-
1,3
151.65
1.49
5.66
12.61
2,4,6,8,10-
198.30
1.82
7.84
27.44
1,4-
152.80
1.48
5.65
12.49
1,3,4,5,6,7-
213.55
1.90
8.76
35.15
2,4-
147.38
1.53
5.81
13.52
1,3,4,4,5,7-
210.95
1.92
8.45
32.93
2,6-
151.60
1.49
5.70
12.77
2,4,6 8,9 10-
214.25
1.89
8.33
31.68
2,2-
150.20
1.50
5.79
13. 25
1,2,3,4,5,6,7-
231.85
1.95
8.77
35.74
1,3,5-
165.30
1.64
6.49
17.66
1,3,4,4,5,7,8-
229.64
1.96
8.81
36.21
2,4,6-
167.41
1.62
6.47
17.41
1,2,3,4,5,6,7,8-
244.66
2.03
9.24
40.59
1,4,4-
165.38
1.64
6.47
17.53
2,2,4,4,6,6,8,8-
250.23
2.00
9.18
39.82
2,4,6,8-
183.57
1.72
7.18
22.72
1,2,3,4,5,6,7,8,9-
267.42
2.02
9.39
41.86
1,3,5,7-
184.52
3
1.71
-1
7.11
-3
21.74
-1
Units: V:cm mol , :gcm , D: kms , P: GPa.
1- and 1,2- denote 1-nitroadamantane and 1,2-dinitroadamantane respectively, the others are similar.
60
5.
CONCLUSIONS
We have studied molecular geometries, electronic structures, thermolysis initiation

mechanisms and theoretical criterion of impact sensitivity for series of energetic materials by
quantum chemical methods. Our systematic work leads to following conclusions:
1. The thermolysis initiation step of energetic compounds can be determined by
comparing the bond orders of neighboring atoms, or more safely, via calculating
and comparing the activation energies of bond breaking reactions.
2. For energetic materials with the similar molecular structures and thermolysis
mechanisms, there is normally a parallel relationship between the bond order of
the trigger bond, the activation energy to break the bond and the experimental
impact sensitivity. Using the bond orders of trigger bonds and the activation
energies for cleavage of the trigger bonds, the sequence of impact sensitivity of
these materials can be estimated or predicted.
3. Tetrazole derivatives and their metallic salts possess similar mechanisms of
thermal decomposition. The trigger step is found to be the ring-breaking, leading
to formation of azides. The relative magnitude of the experimental impact
sensitivity can be correctly determined only with the activation energy of the
ring-breaking.
Acknowledgements
We gratefully thank the National Natural Science Foundation of China (Grant No.
20173028 and 10176012) for the support of this work.
61
REFERENCES
[1]
BOWDEN F P, YOFFE A D: Initiation and Crowth of Explosives in Liquids and Solids. Cambridge
University Press, Cambridge, 1952
[2]
CHAUDHRI M M: Stab-initation of explosions. Nature (London), , 263: 121, 1976
[3]
SVATOPLUK ZEMAN:Thermochimica Acta,384, , 137-154, 2002
[4]
SVATOPLUK ZEMAN. Chapter2 in a monograph by Peter A. Politzer abd Jone S.Murray(Eds):

Energetic Materials, Part2, Elsevies B.V.Amsterdam,2003,pp,25-52
[5]
XIAO, H. M.; WANG, Z. Y.; YAO, J. M.: Acta Chimica Sinica.1985, 43, 14
[6]
XIAO, H. M.; LI, Y. F. Science in China (Series B). 1995, 38, 538
[7]
FAN, J. F.; GU, Z. M.; XIAO, H. M.: J. Mol.Struct. (Theochem,),1996, 365, 246
[8]
XIAO, H. M.; FAN, J. F.; GU, Z. M.; DONG, H. S. Chem.Phys. 1998, 226, 15
[9]
FAN, J. F.; GU, Z. M.; XIAO, H. M.;DONG, H. S. J.Phys. Org.Chem. 1998, 11, 360.
[10] GU ZIMING, FAN JIANFEN, XIAO HEMING, DONG HAISHAN, Chemical Research in Chinese
Universities. 2000, 16(1) : 21-30
[11] CHEN ZHAOXU, XIAO HEMING, YANG SHULIN. Chemical Physics 1999, 250: 243-248
[12] CHEN ZHAOXU, XIAO HEMIN: International Journal of Quantum Chemistry 2000, 79: 350-357
[13] XIAO, H. M.: Molecular Orbital Theory for Nitro Compounds. Beijing, national defence industry
press. 1993
[14] XIAO, H. M.,CHEN ZHAOXU: Modern Theory of Tetrazole Chemistry, Science Press, Beijing,
2004
[15] XIAO,H.M.: Structures and Properties for Energetic Compounds, Beijing, National Defence
Industry Press, 2004
[16] DEWAR, M. J. S.; ZOEBISCH, E. G.; HEALY, E. F.; STEWART, J. J. P.: J. Am. Chem. Soc., 107,
3902, 1985
[17] LEE, C .; YANG ,W.; PARR, R. G.: Phys. Rav.B., 37, 785, 1988
[18] BECKE, A. D.: J. Chem. Phys., 97, 9173, 1992
[19] SONG HUAJU,YU BAIHENG,XIAO HEMING: Chinese J.Chem.Phys.16, 337, 2003
[20] TAYLOR,G.W.;JENKINS,J.M: Third Symposium on Chemical Problems Connected with the
Stability of Explosives, Ystad, May 28-30, pp.43-46, 1973
[21] BATES,L.R.; JENKINS,J.M.: Search for New Detonants, Proceedings of the International
Coference on Reach in Primary Explosives,Vol.2,No.14, March 1975
[22] BATES,L.R.: Proc Symp Pyrotec, 13, V17-V22., 1986
[23] STEWART, J. J.: P. J. Comput. Chem., 10, 209, 1989
[24] EATON P.E.,GILARD R.L.,ZHANG M.X.:B Adv.Mater.,12,1143, 2000
[25] SCHULMAN, J.M.; DISH, R.L.: J. Am. Chem. Soc., 106, 1202, 1984
[26] ZHANG, J.; XIAO, H. M.: J. Chem. Phys., 116, 10674, 2002
[27] KAMLET, M. J.; JACOBS, S. J.: J.Chem. Phys., 48, 23, 1968
62
SOME NOTES ON THE FUNDAMENTALS OF E.M. INITIATION

Towards a UNIVERSAL SENSITIVITY CHARACTERISTIC?
Hans J. Pasman*, Richard Bouma**, Nikolaos Zarzalis***, and Max Weiss***

* Delft Uni of Technology, TNW, MSP, Kramers Lab. Prins Bernardlaan 6, 2628 BW, Delft, NL
** TNO Defense, Security and Safety,, PO Box 45, 2280 AA Rijswijk, The Netherlands
*** University Karlsruhe, Engler-Bunte-Institute, 76131 Karlsruhe, Germany
Abstract:
The stimulus needed for initiation of an energetic material is an important quantity to
determine for reliable functioning and safe handling. When considering the gamma of
impact, friction, spark and heating tests in use one runs into the problem of how to
compare results and how to make a prediction for practical situations that an initiation
certainly will happen or definitely can be excluded. The way forward is not easy. The
initiation process, which in order to be successful, should be followed by sustained
propagation, is not simply accessible. This paper tries to produce a contribution by
considering the fundamental aspects and to show as an example some simulation results
of a gas phase initiation model. Some suggestions are made for making further
progress.
Keywords: Initiation, chemistry simulation, gas phase ignition
1.
INTRODUCTION
1.1
Why looking at initiation
There are two obvious reasons for determining the sensitivity to initiation of an energetic
material:
Reliable functioning of an E.M. loaded device
Avoidance of any unwanted accidental explosive reaction of the material
The relative ability of a material to show an unwanted reaction after being subjected to a
stimulus is in English often referred to as sensitiveness. To quantify sensitivity we need
experimental methods to increase in an incremental and measurable way the loading of an
E.M. until explosion always occurs, or alternatively until the first time an explosive reaction
starts, and we need criteria to decide when the desired level of reliable functioning is
reached or the chance of an unwanted, hazardous reaction can be excluded. The stimulus can
be heat by flame, spark, mechanical impact or shock. Over the years many methods have
been developed and a huge amount of work has been done to measure sensitivity of
explosives, propellants, pyrotechnic materials and their sensitiveness. The latter is also
relevant for safe transport and handling of all kinds of industrial products, which can exhibit
explosive behaviour. Historically one takes a hammer and hits a small sample of the material
on an anvil. The higher and stronger the swing before it cracks, the less sensitive the
material. In the so called mallet test the guy swinging the hammer had to do it in a
standardised way! The contemporary methods are described in detail in the U.N.
63
Recommendations for the Transport of Dangerous Goods [12]. It is not the intention to review
the methods here or to go in any detail, but to discuss some common shortcomings, which
hamper progress.
When one synthesises a new energetic material one of the first things to test is the
impact and friction sensitivity. The equipment to use is the BAM friction machine and drop
hammer apparatus. Both produce an outcome in Joules. What can one do with the figures
obtained. In absolute terms very little. The data does not give any information on what
situations in practice of handling the material are safe or unsafe. It is not possible to translate
the initiation energy in information of e.g. from what height a case with the material can be
dropped to avoid any initiation. After having done the tests successfully one would only
know something on a relative basis. Is it more or less sensitive than TNT or PETN. A
problem may arise if the material under examination is more sensitive than a reference
material in one type of test and less sensitive in another. This picture is further blurred by the
fact that the test outcomes are distributed around a mean. The conditions in a test can never
be controlled in such way that all variability can be excluded. Since the test procedures
always vary the strength of the stimulus in incremental steps, quite sophisticated statistics
have to be applied to obtain results which can be reproduced. Numerous round-robins have
been organised between laboratories to compare test outcomes and to eliminate variability in
assessment. Of course all experts known with the field realise this state of affairs. A new
promising method for correlating sensitivity test results is pattern recognition by using
neural networks [7]. However, unfortunately funding agencies embark reluctantly on
programmes on a longer term to have more advanced methods developed. Instead one tries
to harmonise the methods and the criteria to avoid the situation that one national authority
admits a certain material while another blocks its transportation.
1.2
Explosiveness
When considering explosive reactions generated by reactive chemicals, it is instructive

to consider first the three different types of explosion that can occur in gases, as well as in
liquids and solids:
Thermal explosion
o reaction throughout the entire mass simultaneously
o balance of rate of heat production vs. loss: critical temperature, mass
o induction period (from nanoseconds upwards)
Deflagration
o reaction in a separate zone; propagation mechanism by heat transfer
o burning velocity ranging from some cm/hr to hundreds m/s
o total energy release rate increases with burning surface area
o higher pressure and temperature generally increase burning rate, hence
confinement and heat insulation are important
Detonation
o reaction in a separate zone, propagation mechanism is shock compression
o detonation velocity is supersonic, range: 1 -- 9 km/s
o below a critical diameter no propagation
o confinement partly compensates losses.
Since Bowden and Yoffe in the forties of the past century, the thermal nature of
initiation is clear. Hot spot initiation by stimuli of different nature was demonstrated in many
ways. Thermal explosion is therefore key. It is further helpful to make a strict distinction
between initiation and propagation. Sensitivity has in the first place to do with initiation, but
only successful initiation will result in propagation. So, in fact we hit here the weak points of
64
many methods: It is not clear what type of reaction is tried to initiate and it is not established
properly whether propagation occurs after initiation. In a way propagation is a continued
initiation process. To study initiation it is therefore necessary to consider the propagation
processes. Beside the above, practice is complicated by the transitions in propagation
mechanism. Many initial and local thermal explosions result in deflagration of the remaining
of the material and in larger mass or adequate confinement deflagrations can accelerate and
run-up to detonation.
1.3
Chemistry and physics of explosive processes
An aspect that is often neglected is the chemistry of decomposition and energy

production, mainly because it is too complex. The explosive potential of a reactive chemical
can of course be calculated still relatively simple with chemical thermodynamic and energy
release codes based on Gibbs free energy minimization, e.g. [1, 2], but the real problem is the
kinetics of the reactions. Until recently chemistry was purely empirical, and the zero-order
Arrhenius law was the model par excellence to predict the overall rate of reaction. For
nuances in initiation it was possible in some instances to introduce autocatalytic effects. This
is however changing. Gas phase energetic processes can now be modeled in detail and
hydrocarbon oxidation models contain thousand or more reactions between hundreds of
species. For condensed materials the state of affairs is perhaps much less advanced, but more
sensitive and precise experimental methods, supported by ab initio calculations and
molecular dynamics will improve the situation. In solid explosives the detailed chemistry
and energy transfer are being studied within the reaction zone of the detonation front [14].
As regards the physics of heat and mass transfer for developing explosive processes the
situation is certainly not simpler. For gases the physics of flow interacts with the reactions.
Turbulence generated may intensify mixing and therefore enhance rates but the stresses can
increase to such an extent that reaction quenches. In solids and liquids during the reaction
usually gases are formed which depending on the structure of the material being continuous
or grained, build up pressure and start to flow. This can spread ignition rapidly. In detonation
shock waves dominate the scene and the acoustic impedance of the material matters.
Analytic models of thermal explosion on the basis of the heat balance in differential
form and simple, zero-order Arrhenius kinetics are abundant. Two extremes can be
distinguished: the Semenov - no temperature gradient, ideal convection model and the
Frank-Kamenetzkii - pure conduction model; for an overview see [5]. Apart from the fact that
these extremes often do not cover practical situations, usually the situation is mixed, the
mathematical solution makes use of approximations which can only be realized with the
zero-order exponential law. Numerical models of e.g. the Finite Element type are more
flexible but the lack of kinetic data limits their usefulness. Also diffusion of e.g. gaseous
decomposition products, or of oxygen and water from outside, influences the reaction rates.
Since near the critical condition the temperature-time history is rather flat, it is sometimes, in
particular with larger masses, quite difficult to decide whether a certain starting condition in
the end will lead to an explosion or not. Induction periods of several weeks are possible.
The energy release rate in deflagration is determined by burning velocity and burning
area. Models for the burning rate are scarce and mostly limited to propellant type of
energetic materials. Geometry, confinement, physics of flow and structure of the material
can have a very strong influence on the fraction of the energy generated that is used to
initiate a new layer of material. If a point initiation source is applied the strong initial growth
of the burning surface requires relatively much energy to sustain the front and may lead to
slow down or quench of reaction. On the other hand in linear charges in all substances in
(partly) confined situations a strong acceleration of the energy release rate can occur. In
65
gases this is due to the self generated turbulence as a result of the expansion of the hot gases
behind the flame front and in liquids and solids by the increase in burning rate by pressure
build-up and the increase in burning area by a surface instability of the reaction front in
liquids and the presence of a porous grain structure in solids. In all cases this can lead to
pressure waves and reactions ahead of the flame in pre-compressed material and sudden
initiation of detonation.
In many respects modeling and testing of detonation is conceptually simple, although
the process completes in microseconds. For the models the Chapman-Jouguet overall set of
equations is common knowledge. Further, the criterion for its presence is a sustained,
stationary detonation wave, which can be traced either by a fragmentation pattern of
confining tube, a continuous pressure sensitive probe, high speed cinematography, optical
diagnostic technique as laser imaging or interferometry etc. Also in case of detonation
geometry is important. Because of the energy diversion a point detonation is much more
difficult to initiate than a detonation initiated over the full cross-sectional area of a
cylindrical charge [6]. This shows up in particular when trying to initiate a low energetic
material as ammonium nitrate or a so called non-ideal explosive. Attempting to initiate over
a full cross- section of a (confined) cylindrical charge AN ammonium nitrate can still
detonate when the initiating booster only contains 10% penthrite, PETN, but a small
spherical 100% PETN booster in the centre fails to do it. Of course, the process of coupling
of a shock wave to the reactions it triggers and the associated energy release rate profile is
still one of the great challenges to science to solve.
Overlooking these various processes one can conclude that in initiation some form of
mechanical energy (impact, shock or friction), electrical energy (spark, plasma), laser light,
or just thermal heating is deposited in a certain volume in a certain time, which triggers
reaction. The distribution of the energy will in general not be uniform (hot spots) and by the
nature of the material (slide planes, shear, crystal defects, gas bubbles) it may also work out
heterogeneously. For propagation to succeed the reaction shall to a sufficient extent selfaccelerate. In sustained propagation of an explosive process there is an energy flow forward
initiating fresh material. In case of point initiation this energy flow is diverging. So beside a
sufficient amount of initial energy input to raise temperature, there will be minimum volume
required with a sufficiently large radius and surface area. The energy input shall also be
within a sufficiently short time, so that heat losses do not frustrate and quench the process.
This brings us to the utopian concept of a 'universal sensitivity characteristic', in which a
stimulus is defined in terms of energy, time and volume of deposition. In Figure 1 schematic representations are given of the initiation energy and initiation energy density (energy
deposited per unit of volume) to initiate and propagate an explosive reaction as a function of
the radius of volume in which the energy is deposited. In small volumes the energy required
is relatively high in particular at long deposition times because during chemical induction
heat leaks away. When the volume gets large despite the losses the centre will be only
slightly affected and the chemical induction time becomes determining. Hence the energy
density decreases till a certain minimum level. In Figure 2 the initiation energy density as a
function of time (in log scale) is schematically shown. It can again be seen how an optimum
arises between energy influx, chemical response rate and energy losses to the surroundings.
For a given test set-up the minimum ignition energy (MIE) apparently depends on the
duration of the stimulus and the volume: Not too short, not too long; not too small, not too
large.
Propagation will depend on a critical power density or energy flux to ignite the next
layer. Front curvature and diameter (lateral losses) are important. In gases preferential
diffusion of fuel or oxidizer species play a role. This will not be the subject of this paper.
66
td,3 >> td,2 >>td,1
td,3
td,2
td,1
Log Initiation Energy Density
Log Initiation Energy
Deposition times, td
Log Deposition radius, rd
Log Deposition radius, rd
Initiation energy (left) and energy density (per unit of volume) (right), schematic, as
a function of the deposition volume radius at three different deposition duration
times. At small volume heat loss prevails. The energy density tends to a constant
minimum value, although for longer deposition times at higher level.
Log Initiation Energy Density
Fig 1.
td,3
td,2
td,1
Deposition radius, rd
rd,3 >> rd,2 >> rd,1
rd,1
rd,2
rd,3
Log Deposition time, td

Fig 2.
Initiation energy density, schematic, as a function of the deposition time at three

different deposition radii. At longer times the effect of deposition volume increases.
At ultra short times (nano-seconds) the chemical relaxation times are going to play
a role and energy density goes up again.
67
2.
SENSITIVITY TESTING: A GAS PHASE MODEL CASE
The gas phase is relatively accessible for testing. The most convenient way of initiation
is by electric spark. Within certain limits energy, deposition time and to a lesser extent
deposition volume can be controlled. Although many experiments have already been
reported, special equipment is now installed by one of the authors (Nikolaos Zarzalis) to
vary the ignition parameters as freely as possible and to study ignition also at higher
pressures1.
Since gas phase hydrocarbon oxidation kinetics are well modeled in detail one can also
simulate on the basis of a simplified version of the Navier-Stokes equations, written in
Lagrange co-ordinates. The reactions are represented by over 300 equations with more than
50 species [4]. In the ignition process the acceleration of reaction in gases under the influence
of energy input is by radical branching processes (net increase in number of radicals) and
self-heating. Such scheme cannot be described by a simple overall Arrhenius equation, since
the oxidation mechanism changes drastically with temperature. This makes e.g. that at a
certain temperature range induction time even becomes longer with further temperature
increase. The solution is by a finite difference scheme. Below results are given of
calculations made with the program INSFLA [4, 17], which fully support the schematic of
Figure 1. For details of the work, see Weiss [13]. In Figure 3 the calculated ignition energies2
and densities for stoichiometric propane-air mixtures and different deposition times are
plotted against deposition radius.
For small deposition radii the ignition energy is largely dependent on the deposition time
and seems to tend to a constant value with decreasing energy deposition radius. The energy
Diffusion
Fig 3.
Induction controlled
Ignition energies (left) and ignition energy densities (right) of propane-air mixtures
versus energy deposition radius at different deposition times.
Financial support of this work by the European Commission within the Fifth Framework Programme on
Energy, Environment and Sustainable Development, contract EVG1-CT-2002-00072, Project SAFEKINEX is
gratefully acknowledged.
2
This is the minimum energy required for successful ignition in a certain condition. The term Minimum
Ignition Energy or MIE is reserved for the real overall minimum considering both volume and deposition time.
68
level of this constant part decreases with deposition time but this effect is slowing down in
the microsecond time order of magnitude. The minimum ignition energy seems to depend
therefore particularly on a certain minimum deposition time.
With longer deposition times the amount of energy for a successful ignition increases.
This effect is stronger the smaller the deposition radius. This ignition region is called the
diffusion controlled region. In this region the heat loss, which is mainly due to thermal
diffusion, is the determining process of the ignition.
For radii larger than about 1 mm the ignition energy density reaches a nearly constant
minimal value for all deposition radii and times, the minimum energy density. The ignition
energy shows an asymptotic and deposition time independent behavior. It increases
proportional to the energy deposition volume (3rd power of the deposition radius). This
region is called the induction controlled region. In this region the ignition process is mainly
controlled by the formation of a radical pool within the induction time.
The simulation results allow too an examination of the temperature distribution of the
hot spot. In the left diagram of Figure 4 temperature profiles are shown at three different
points in time for a simulation with an input energy of 0.1 mJ below the ignition energy. The
deposition time, td has been set to a value of 600 s and the deposition radius, rd to 4 mm.
After the deposition time near the boundary the temperature starts to decrease and
3000
3000
2500
0.6 m
0.6 ms
5 ms
8 ms
4.5 ms
5 ms
2500
2000
T [K ]
T [K ]
2000
1500
attd
1000
rd
500
6
10
r[
mm]
Fig 4.
attd
1000
rd
500
1500
10
r[
mm]
Induction controlled ignition (td=600 s, rd=4 mm) of stoichiometric propane-air at

atmospheric conditions: temperature profiles at three different points in time;
unsuccessful ignition at Ei=144.3 mJ (left), successful ignition at Ei=144.4 mJ
(right)
the temperature profile flattens outwards. In the right diagram of Figure 4 the energy value is
chosen only 0.1 mJ higher to 144.4 mJ. In the latter case the kernel temperature increases
after the deposition time and the flame front starts to propagate. For large deposition radii the
basic criterion for a successful ignition is that the temperature reaches a minimum ignition
value after energy deposition. This is comparable with the auto ignition temperature, but the
values in case of an induced ignition are slightly higher.
For large deposition radii (induction controlled region) the ratio of surface of the
deposition volume to deposition volume is smaller than for small radii (diffusion controlled
region). As the heat of reaction is proportional to the volume and the heat flux is
proportional to the area, heat loss by diffusion in comparison to heat production by the
exothermic chemical reaction is much smaller than in the diffusion controlled region. In the
69
induction controlled region the thermal diffusion is almost negligible, compared to the
diffusion controlled region, so that the ignition energy approximately depends only on the
deposition radius and is independent on the deposition time. Thus the energy density is
nearly constant. As one can see in Figure 4 the energy density tends for a radius larger than
1 mm toward a constant minimal value and is not depending on the deposition time.
The criterion for initiation is successful propagation. Figure 5 shows three temperature
profiles at different points in time in case of ignition leading to successful and unsuccessful
propagation. Looking at these graphs one can see that the temperature after energy
deposition is much higher in comparison to the temperature in the induction controlled
region, even if the energy amount is not high enough for a successful ignition with adjacent
flame front propagation. Further one can notice a local rise of the temperature profile starting
in the centre of the energy deposition volume. This is a sign of already starting exothermic
chemical reaction, meaning that the exothermic reaction can start in the diffusion controlled
region even if the energy input is lower than the ignition energy. Because of the small energy
deposition radius the heat diffusion is high and the temperature is then decreasing rapidly in
case of an energy value smaller than the ignition energy. If the ignition energy is deposited,
the temperature is first decreasing. After passing a critical radius the temperature is
increasing and reaches the burned gas temperature.
Fig 5.
Diffusion controlled ignition (td=600 s, rd=0.2 mm) of stoichiometric propane-air

at atmospheric conditions: temperature profiles at three different points in time;
unsuccessful ignition at Ei=542.1 J (left), successful ignition at Ei=542.2 J (right)
3.
THE PRACTICE OF E.M. TESTING
When testing it should be clear that the applied test is relevant for the situation being
examined for safety, and the test effort should be efficient in view of the utility and
reliability of the result obtained. This is still utopia. Because of the bad control given by the
test set-up and the nature of the sample a large dispersion in result usually occurs.
The criterion for a substance tested with e.g. the Julius Peters fall hammer machine to be
safe is no explosion (or rather no 'bang': Is any 'bang' really an explosion?) at 2 Nm or 2 J
70
drop energy, i.e. 1 kg hammer at 20 cm drop height [12]. Under conditions of more efficient
energy transfer the same substance may be initiated by 10 - 100 times less energy. Basic
studies have been performed amongst others by Field and coworkers [11], however more
knowledge is needed to find out how much energy in what time is deposited in what volume.
Further, the criterion should be met as only one explosion in 6 tests. This seems risky since
for safely applying go/no-go statistics, the establishment of a 50% probability point (up- and
down method assuming normal distribution of outcomes, or the Robins-Munroe method not
requiring an a priori distribution, e.g. [3]) requires many tens of tests. For both reliability and
safety investigation one should in fact go further and try to establish e.g. a 99% or a 1%
probability level of an explosion event, given a certain stimulus. This would however, take
several hundreds of tests.
An improvement to traditional impact testing can be found in [15], describing the Energy
to Ignition test and the Ballistic Impact Chamber test. With a photo sensor the moment of
ignition in a solid explosive is determined, and from the deceleration of the drop weight
during the impact, one can calculate the loss in kinetic energy of the drop weight being a
measure of the energy transferred to the explosive. This measure of energy is a better
measure compared to the previously mentioned potential energy of the drop weight.
Correlating the thermal decomposition characteristics with impact and friction
sensitivity has been attempted in view of the hot spot mechanism. This has certainly been a
success. One of the possible ways is the Thermal Step Test [10]. It consists of a stainless steel
capillary tube, suddenly heated by a charged capacitor discharge with the tube as part of the
circuit, acting simultaneously as a heat sensing element (resistance thermometer). Consistent
measurements have been possible of induction time of a millisecond to a hundred seconds at
a given jump-wise-attained initial temperature level. Plotting the logarithm of induction time
versus reciprocal absolute temperature produces straight lines down to the millisecond
regime. Below this value however, anomalies occur starting at different time values for
different substances, probably due to changes in heat transfer mechanism. Above this
apparent energy of activation values are well in agreement with values otherwise
determined. Higher sensitivity can be directly correlated with shorter induction times at a
given temperature.
There are further new possibilities. Laser initiation is one of these, see e.g. [8, 9]. A laser
can deliver in principle a well controlled amount of energy into a material with respect to
power, area of energy deposition, pulse duration and total energy. It is less controllable in
terms of volume in which the energy is delivered and short pulses are associated with shock.
Systematic experimentation determining the energy density for ignition as a function of
pulse exposure time deposition volume is scarce.
Shock initiation of solid explosives by flyer impact does provide valuable data as well.
The impact flyer induces a rectangular shock wave in the explosive sample with the induced
pressure related to the velocity of the impact flyer and the pulse duration to its thickness.
Typically shock initiation data are given as minimum impact velocity and/or energy fluence
(energy input per unit of area) as function of flyer impact thickness. Data on the initiation of
pressed samples of TATB and HNS-IV, with a density of 80% of the Theoretical Maximum
Density, are given in [16] and represented here as well as energy density versus pulse
duration, see Figure 6. From the graph of energy fluence vs. flyer thickness one can conclude
that from energetic considerations (minimizing the kinetic energy of the impacting flyer) a
flyer thickness near 125 m is optimal for shock initiation of the TATB sample, whereas a
flyer thickness of even less than 50 m is preferred for shock initiation of the HNS-IV. Such
a consideration may be important in the optimization of a detonation train. However, looking
at the energy density vs. duration of the incoming shock one can clearly see the minimum
71
energy density value, which is specific to the energetic material. For the TATB sample the
minimum of 1.8 J/mm3 holds for pulse durations down to about 0.1 s. For HNS-IV the
induction controlled region extends down to pulse durations of 0.06 s, and a corresponding
minimum energy density of 0.85 J/mm3.
0.7
10
0.5
0.4
e / J/mm3
E/A / J/mm2
0.6
0.3
0.2
TATB C
HNS-IV
0.1
0
0
100
200
300
0.1
0.01
flyer thickness / micrometer
0.1
pulse duration / microseconds
Fig 6.
Minimum energy fluence vs. flyer thickness (left) and minimum energy density vs.
pulse duration (right) of pressed TATB and HNS-IV explosive samples.
4.
CONCLUSIONS
During the last ten years considerable progress has been made in understanding
mechanisms and in modelling. However, many test methods for determining
sensitivity are still archaic and require too much effort given the quality and
applicability of the answers.
Rationalization and harmonization of tests and criteria is one way to go. Even more
important is to unify efforts under expert leadership and create cooperative
programmes to develop improved methods and to try to determine fundamentally
based quantities enabling extrapolation and risk estimation in practical situations.
72
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
FRURIP, D.J., FREEDMAN, E AND HERTEL, G.R.: A New Release of the ASTM CHETAH
Program for Hazard Evaluation: Versions for mainframe and Personal Computer,
Plant/Operations Progress, Vol. 8, No.2, 100 - 104, April 1989
Frurip, D.J., 1992, Using the ASTM CHETAH Program in Chemical Process hazard
Evaluation , Plant/Operations Progress, Vol. 11, No.4 224-228, Oct.
LEEUW, M.W., BAL, E.A. AND PRINSE, W.C.: Parametric Analysis of the Reliability of Igniter
Systems, Rev. Scientific Instr., Vol.63 (6), 1-10; Leeuw, M.W., 1990, Non-destructive
Reliability Analysis of a Bridgewire Fuze Head, Prop. Expl. & Pyrotechnics, Vol.15, 92-96.
1992
Maas, U.: Mathematische Modellierung instationrer Verbrennungsprozesse unter
Verwendung detaillierter Reaktionsmechanismen, Dissertation, Ruprecht-Karls-Universitt
Heidelberg, 1988
MERZHANOV, A.G. AND ABRAMOV, V.G.: Propellants and Explosives, 6, 130. 1981
MIYAKE, A, VAN DER STEEN, A.C. AND KODDE H.H.: Detonation Velocity and Pressure of the
Non-Ideal Explosive Ammonium Nitrate, Proc.Ninth Symp. (Intl) on Detonation, 560-565,
Aug 28-Sept 1, Portland Oregon, OCNR 113291-7, Arlington, VA.22217-5000, 1989
NEFATI, H.: Prdiction de la sensibilit au choc des substances explosives ou non : approches
statistique et neuronale , Thse pour l'obtention du grade de Docteur de l'Universit Paris VI,
Chimie Applique et Gnie des Procds Industriels, le 18 juillet, V-153 p.+ annexes, 1994
STMARK, H.: Laser as a Tool in Sensitivity Testing of Explosives, The Eighth Symposium
(Intl) on Detonation, July 15-19, 46-53, 1985
Roman, N.B.M., 1991, Laser Ignition of Explosives and its Application in a Laser Diode Based
Ignition System, Proc. 16th Intl Pyrotechnics Seminar, 822-836, 24-28 June, ISBN 91-3600551-4.
SCHRADER, M.A., LEEUW, M.W. AND VAN DER STEEN, A.C.: The Thermal Step Test: A Key to
High Temperature Behavior of Explosives, Proc. 9th Intl Pyrotechnic Seminar, Colo Springs
USA, 881-891, 6-10 August, IITRI, Chicago Ill, 1984
SWALLOWE, G.M., AND FIELD, J.E.: The ignition of a thin layer of explosive by impact; the
effect of polymer particles, Proc. R. Soc. London A 379, 389-408, 1982
U.N.: Recommendations on the Transport of Dangerous Goods, Tests and Criteria (Orange
Book), 2nd ed. U.N. ST/SG/AC 10/11/Rev.1, New York, U.N. Publ. Sales No. E90 VIII.1,
ISBN 92-1-139033-8, ISSN 1014-7160, 1990
WEISS, M.: Model, software development for calculation of Markstein numbers and minimum
ignition energies (MIE), Deliverable No.20 Project SAFEKINEX, September 2004
TARVER, C.M., FRIED, L.E., RUGGIERO, A.J., CALEF, D.F.: Energy transfer in solid
explosives, Proc. of the Tenth International Detonation Symposium, 3-8, July 12-16, 1993,
Boston.
COFFEY, C.S., DE VOST, V.F.: Impact testing of explosives and propellants, Propellants,
Explosives, Pyrotechnics 20, 105-115, 1995
BOUMA, R.H.B., JORDAN, T., CARDELL, A., VAN ESVELD, R.J., PRINSE, W.C., VAN DER
STEEN, A.C.: Flyer impact characterisation of booster explosives for IM applications,
Insensitive Munitions & Energetic Materials Technology Symposium, 16-19 November 1998,
San Diego.
MAAS, U., WARNATZ, J.: Detailed numeric modelling of H2-O2 ignition by hot spots, Progress
in Astronautics and Aeronautics 131, 1991
73
ATOMISTIC STUDIES OF FUNDAMENTAL PROPERTIES

AND PROCESSES IN ENERGETIC MATERIALS:
RELEVANCE TO MESOSCALE INITIATION PHENOMENA
Thomas D. (Tommy) Sewell
Theoretical Division, Explosives and Organic Materials Group,
Los Alamos National Laboratory, Los Alamos, New Mexico 87545, USA
Abstract:
Genuine, physics-based understanding of initiation phenomena in plastic-bonded
explosives (PBXs) requires knowledge of the physics and chemistry at mesoscopic scales
that are far larger than can be simulated directly using atomistic detail, yet far smaller
than is directly resolvable in practical engineering scale continuum simulations.
Initiation is determined by localization phenomena that arise due to the heterogeneous
character of most explosive formulations. Indeed, the average temperature behind a
weak shock is not a useful measure for understanding initiation phenomena; rather, it is
the tails of the distributions in temperature, stress, and strain rates, localized to small,
spatially distributed volumes in the material (hot spots), that dictate the outcome of a
given loading event.
Important factors for predicting hot spot formation and subsequent extinction or
growth/coalescence include particle size, concentration, morphology, and void content;
physical and chemical interactions between grains and binder; thermophysical and
mechanical properties of the constituents and interfaces between them; and, of course,
the inherent chemical stability of the explosive component(s) in the formulation.
We are in the process of computing many of the thermophysical and mechanical
properties required for a complete specification of constituent models for use in
mesoscale simulations, wherein grains and binder in representative volumes of a PBX
are spatially resolved and then studied within a continuum hydrodynamic framework. In
addition to calculating specific properties of interest, we have recently undertaken a
series of large-scale molecular dynamics simulations of energetic crystals to understand
dissipation phenomena in dynamically loaded single- or poly-crystalline samples; for
instance, plastic deformation and stress/energy localization mechanisms, phase
transitions, and so on. Recent and ongoing work in these areas will be discussed, along
with their specific relevance to emerging mesoscale simulation capabilities.
Keywords:
1.
atomistic simulation, quantum chemistry,

thermal and mechanical properties, localization, initiation
INTRODUCTION
In order to understand and predict initiation in plastic-bonded explosives (PBXs), for

example behind a weak shock, there are at least three key classes of knowledge that we must
have in order to claim a credible predictive capability:
the mechanisms and resulting spatio-temporal distributions of energy localization,

typically measured in a reference frame that moves with the shock front [1];
the thermal and mechanical parameters required for specification of the equation of state,
which enable predictions of local temperature and stress [2]; and
74
the correct chemical reactions rates, which should ideally be expressed as functions of
temperature and pressure, rather than pressure.
While the need for the second and third points above is presumably obvious, the need to
characterize the distributions of temperature and stress state behind the shock, and the
underlying unit mechanisms by which these non-uniform conditions arise [3,4], may be
somewhat less well recognized. For a weak shock, the bulk temperature due to simple
hydrodynamic heating is not high enough to lead to prompt chemical reaction [5]. Rather, it is
only in regions of space corresponding to the extreme tails of the temperature and pressure
distributions the hot spots that fast chemistry will occur and then either spread or quench
depending on the spatial density and size of the hot spots. Given the complexity of a real
PBX, it necessary to capture this information within a statistical framework. (It is important
to emphasize the focus in this paper on ignition phenomena as opposed to strong shock
initiation, for which the role of hot spots may be substantially less than has generally been
assumed in the past [6].)
Experimental determination of many of the processes and properties mentioned above is
difficult for high explosives, due in part to the complicated molecular/crystal structures
involved and also to the extremely fast reaction rates that arise for many thermodynamic
conditions of interest. Indeed, for many common explosives, even the simplest parameters
are known only at a few temperatures and pressures, and often with relatively large
uncertainties. Moreover, the error bars typically increase the further one departs from
ambient conditions [7]. As a concrete example that applies to weak shock ignition, absence of
knowledge of the temperature (and pressure?) dependence of the specific heat can lead to
errors of hundreds of degrees in predicted temperature such that, if real rather than
calibrated chemical rates are to be used, induction times for reaction can be off by orders of
magnitude [8]. The most common way around this problem is to take as fixed a few key
parameters in the equation of state, and then adjust others to match, for example, the
principal shock hugoniot locus. These knobs become problematic for more complicated
scenarios, however; as, for example, in the case of multi-shock stimuli that take one far from
the principal hugoniot. However, assessment of experiments even for such simple quantities
as the bulk modulus and its initial pressure derivative can be problematic [9].
Over roughly the last decade we have employed classical molecular dynamics and
Monte Carlo methods to predict some of the thermophysical and mechanical properties that
are needed as input to realistic mesoscale equation of state and dissipation models for
plastic-bonded explosives (PBXs). The main materials we have studied by these methods
include TATB [10]; RDX [11,12]; -, -, -, and liquid HMX [13-18]; Estane, a poly(ester
urethane) copolymer [19]; and bis(2,2-dinitropropyl)formal/acetal (BDNPF/A) [20]. The threedimensionally periodic simulation cells in our more recent work typically contain ~102
molecules, with simulation times of 2-10 nanoseconds (ns). While much of this work has
focused on verification and validation (V&V [21]), the overarching goal of these medium
scale atomistic studies is to develop internally consistent model thermodynamic and elastic
mechanical descriptions for the materials, suitable for use in, or as a guide to the formulation
of, mesoscale simulations in which the grains and binder in a PBX are resolved within a
continuum hydrodynamic or material-point method (MPM) framework. Emphasis in our
work is placed on non-reactive properties and thermodynamic states relevant to cookoff and
shock initiation phenomena. Prioritization among those properties amenable to atomistic
simulation is made based on ongoing interactions with mesoscale modelers [8,22],
predominantly at the Los Alamos and Sandia National Laboratories. With the recent
completion of a force field for BDNPF/A [20], we should, in principle, be able to study all
relevant intermolecular interactions in PBX-9501 [23].
75
In addition to classical molecular dynamics and Monte Carlo simulations, we have

employed condensed phase electronic-structure methods to study the zero Kelvin structures
and properties of crystalline HMX [24] and PETN [25]. Highly parallel, O(N) algorithms
implemented in the MondoSCF quantum chemistry code [26] used in Ref. [25] enable routine
fully optimized, all-electron, density functional electronic-structure predictions of zero
temperature molecular crystal structures as a function of material density that is, the cold
curve. In this work we are able to use high-level functionals and basis sets (e.g., PBE/631G**), and primary cells containing hundreds of atoms. The PETN study described in Ref.
[25] is the first published application of MondoSCF to high explosives.
In Section 2 we outline our view on good roles for atomistic simulations in connection
with high explosives (and most other materials, for that matter), tempered by a brief
philosophical digression on the skepticism with which atomistic results should be
interpreted. We also include a few results from our older work, which serve as reference
points for what can realistically be extracted from atomistic simulation. Vignettes of work
presently underway are given in Section 3; as such, the results shown there should be
regarded as snapshots in time rather than final product. We conclude in Section 4 with
some thoughts about probable future roles for classical and electronic structure-based
atomistic simulations of energetic materials.
2.
THE IMPORTANCE OF ATOMISTIC COMPUTATIONS

IN EXPLOSIVE SCIENCE
2.1
Useful roles for atomistic studies
A large number of atomistic studies of condensed phase high explosives have been
published, with a rate that has increased noticeably in recent years [27]. Based on our own
experience, we have come to view judicious application of atomistic simulation approaches
as a credible tool for prediction of high explosive thermophysical and elastic mechanical
properties. In particular, we think that, given a validated potential-energy surface, molecular
simulations can serve as a complement to experimental data by extending regions in
thermodynamic space for which information is available [12]. Another useful role is to
provide information about properties in thermodynamic regimes that are difficult, if not
impossible, to probe experimentally for some high explosives (e.g., liquid HMX [16,17]).
Although unavailable experimentally, some of these quantities are key to the
parameterization of improved mesoscale models [2,8]. In situations for which independent
experimental determinations of some property are at odds with one another, molecular
simulation approaches are potentially useful as a tool for interpreting the differences based
on, for example, differing methods of data collection and/or analysis. Such situations are
more common than one might think, even for simple properties such as the isotherm and
bulk modulus [9,15,25,28,29] and coefficients of thermal expansion [14,30-33], let alone more
complicated objects such as the elastic tensor [12,15,34]. Finally, molecular-scale calculations
can provide insights into fundamental processes and mechanisms at a level unlikely to ever
be matched experimentally.
2.2
Factors to consider when assessing atomistic simulation results
In spite of the fact that we are gaining increased confidence in atomistic simulations as a
viable tool for use in explosive science, it is vitally important to maintain a healthy
skepticism when assessing the results of a molecular dynamics simulation. From a
philosophical point of view, Caveat Emptor (Let the Buyer [End User] Beware.) Among the
76
questions that should be asked by anyone contemplating the use of atomistic simulation
results as input to their own research are:
Is the force field used validated against a variety of experimental data?
Are the computational methods employed sound for the problem addressed?
Are the time and length scales accessible in the simulations compatible with the
phenomena of interest?
Are the results likely to be highly corrupted by pesky quantum effects not accounted for
in a classical simulation?
Was care taken to ensure convergence of the results, and are statistical measures of
precision provided?
With a little imagination, one can draw meaningful analogies between each of the items
above and corresponding ones relevant to larger spatial scales such as continuum
hydrodynamic simulations (constitutive models, mesh refinement, etc.).
2.3
Representative reults from older work
Table 1 contains calculated and measured elastic tensors for RDX at room temperature
and atmospheric pressure. The calculated results are based on early isothermal-isobaric
Monte Carlo realizations that used small primary cells and rigid molecules [12]. (The rigidmolecule approach has been used extensively by others in studies of high explosives, most
notably in a series of papers by Thompson and co-workers [35].) The two sets of experimental
results are based on resonant ultrasound spectroscopy (RUS) [36,37]. In the present
comparison we make no attempt to account for the difference between isothermal and
isentropic elastic tensors, as they are expected to differ by only a few percent. In spite of the
relatively severe approximations made in this early study, the agreement between the
simulated results and experiment for individual elements of the elastic tensor is surprisingly
good, and derived values for the isotropic moduli are in even better agreement. (The level of
agreement seen here for the experimental and calculated elastic tensor is somewhat unusual,
since the elastic tensor for polyatomic molecules is highly sensitive to the curvature of the
force field near the potential energy minimum [34]; in those instances for which experimental
data are available, the elastic tensor is among the most incisive measures of the accuracy of
the force field near equilibrium.)
Table 1. Table 1. Calculated and measured atmospheric pressure elastic tensors for
RDX.
Source
Simulationa (2000)
Hausshlb (2001)
Schwarz,
Dick,
Hooksc (2005)
C11
27.5
19.58
and 21.3
C22
24.6
24.98
25.6
C33
18.0
17.89
19
C44
8.4
4.06
4.27
C55
5.3
5.15
5.38
C66
8.0
6.90
7.27
C12
6.5
8.19
8.67
C13
5.9
5.90
6.4
C23
6.4
5.80
5.72
a. Ref. [12], isothermal-isobaric, rigid molecule Monte Carlo, T=304 K.

b. Ref. [36], resonant ultrasound spectroscopy.
c. Ref. [37], resonant ultrasound spectroscopy.
A much more sophisticated, flexible-molecule force field has been used in a series of
studies of liquid and crystalline HMX [13]. Among the liquid state properties of HMX for
which predictions have been reported are atmospheric pressure self-diffusion and shear
77
viscosity coefficients [16] and thermal conductivity [17], each extending from 550 K to 800 K.
These temperatures bracket the nominal HMX melting and boiling points, respectively, at
atmospheric pressure. Although there are no experimental data for direct validation, the
predictions of the liquid thermal conductivity are generally consistent with data for the solid
at elevated temperature (and which most likely consists of a mixture of - and -HMX, with
some amount of degradation products), and correspond fairly closely to the (constant) liquid
state value for thermal conductivity used in some of the combustion simulations of
Beckstead and co-workers [38].
Linear fit
Experiments
MD simulations
Log(viscosity) in cP
-1
1.5
2.0
2.5
3.0
3.5
-1
1000/T (K )
Fig 1.
Calculated and measured shear viscosity of BDNPF/A, plotted as

log(viscosity) -vs- 1/T. Line is a fit through calulcated results.
Very recently [20], we completed the development of a force field for bis(2,2dinitropropyl)formal/acetal, which is the plasticizer in the binder for PBX-9501. A
comparison of the measured and calculated shear viscosities for that eutectic mixture is
depicted in Fig. 1. It was necessary to use temperature extrapolation to compare to
experiment, since the simulation times required to obtain converged values of the shear
viscosity near room temperature would have been prohibitively long. However, the
measured and calculated results appear to be consistent. Additional force field validation
calculations for BDNPF/A and other model compounds can be found in Ref. [20].
3.
SELECTED HIGHLIGHTS FROM ONGOING RESEARCH

3.1
Further studies of the HMX liquid equation of state
Continued studies of the HMX liquid equation of state are underway by Jaramillo and
Sewell, to yield liquid-state properties as functions of pressure and temperature, and to map
out where in thermodynamic space the onset of glassy behavior occurs (based on the ~10 ns
time scale of our simulations). The force field and computational protocol are essentially
identical to those used in our recent study of the elastic properties of crystalline HMX [15],
except that in the liquid studies we constrain the primary simulation cell to cubic geometries.
The calculations span the intervals 0 GPa P 10 GPa and 550 K T 850 K, and were
designed to bracket the existence interval for the liquid at atmospheric pressure. (Obviously,
78
the melting and boiling points will increase with increasing pressure.) In Fig. 2 we show the
calculated bulk modulus as a function of temperature, with separate curves for several
pressures between zero and ten GPa. The bulk modulus at a given thermodynamic state was
obtained from an analysis of the volume fluctuations sampled in our isothermal-isobaric
simulations via the relation B=<V>T/V2, where <V> and V2 are the average volume and
its variance, determined from the simulation.
Fig 2.
Calculated bulk modulus of liquid and glassy HMX as a function of

temperature and pressure.
Based on calculations of molecular self-diffusion coefficients, time scales for molecular

rotation, and radial distribution functions (none of which are shown), we conclude that, for
the relatively short (but nevertheless relevant for mesoscale simulation) time scale of ~10 ns,
the onset of glassy-like behavior in liquid HMX occurs at fairly low temperatures, even for
modest pressures. The bottom-line results obtained to date are shown in Fig. 3. The figure is
something of an amalgam of the results for the various measures listed above, but suggests
that, for most pressures above about one GPa, HMX behaves more like an amorphous,
glassy material than as a liquid, for temperatures below about 900 K. This may be of
practical importance in discriminating among candidate dissipative mechanisms in HMX at
the moderately high pressures and temperatures relevant to weak-shock initiation [5]. It is
important to understand that Fig. 3 does not reflect a rigorous prediction based on free
energy calculations to obtain the coexistence curves by matching chemical potentials, or
even from two-phase simulations such as have been performed recently for nitromethane by
Thompson and co-workers [39], in which crystalline and liquid volumes contained within a
single periodic simulation cell were gradually heated until melting of the crystalline phase
into the liquid was observed.
79
Fig 3.
3.2
Predicted phase diagram for the onset of glass-like behavior of HMX at

elevated pressures and temperatures, based on a variety of structural and
dynamical measures.
Electronic structure studies of hydrostatic compression

of -HMX
As a follow-on to our preceding electronic-structure studies of the PETN zero Kelvin

isotherm (cold curve) [25], we have undertaken analogous calculations for -HMX. The
interest in re-visiting -HMX isotherm, in spite of the existence of at least three each
experimental [28,29,40] and computational [11,15,35,41] studies (including one each at non-ambient
temperatures [40,41]) is that, while all of the published computed isotherms are in better
agreement with the 1978 isotherm of Olinger et al. [28], Menikoff and Sewell [8] concluded
that the 1999 Yoo and Cynn isotherm [29] is more consistent with the preponderance of data
for shocked -HMX. In particular, the extrapolated Olinger isotherm appears to be too
stiff/steep (in the P-V and Us-Up planes, respectively) compared to shock hugoniot data,
which is an aphysical result.
As in our PETN study [25], we use a three-dimensionally periodic, all-electron density
functional treatment, at the PBE/6-31G** level; and fully optimize the cell contents and the
symmetry class-restricted lattice parameters (in this case lattice lengths a, b, and c; and
monoclinic angle ) at a given volume compression V/V0. Most of the calculations were
performed using 32 or 64 processors. Two different computational protocols were used, both
of which yield hydrostatic conditions through a series of sequential line searches along the
monoclinic lattice parameters, with geometry optimization of the cell contents at each step.
There are two differences between the protocols. In one case, which we refer to as uniform
optimization, we initially assume uniform relative compression along a, b, and c to obtain a
starting guess for the first geometry optimization [i.e., (a1,b1,c1)= (a0, b0, c0),where is
slightly less than unity]; in subsequent steps, we simply enforce xk+1/xk,r = xk,r/xk-1,r (where x
80
= a, c, are varied independently; b provides closure at the desired volume V=abcsin;

subscript r denotes a fully relaxed structure from a preceding sequence of line searches; and
no reference to experiment is required). The second approach, which we refer to as
experimentally-based, differs in that we initially follow the experimental linear
compressions along a, b, c, and , using a 5th-degree polynomial fit to Olingers data [28]; and
enforce volume closure via the monoclinic angle instead of the lattice length b.
In Fig. 4 we show the PBE/6-31G** energy difference E versus hydrostatic volumetric
compression ratio V/V0 (left panel) and the associated P-V/V0 relationship (right panel). We
also include both published experimental P-V data sets for comparison. From the relative
energies, we conclude that the results are not highly sensitive to primary cell size. (This is
only true for large values of V/V0, however; thus, all subsequent calculations will only be
shown for the 2x1x2 case.) The right panel is perhaps more interesting in that, for the 2x1x2
uniform case shown, PBE/6-31G** predicts better agreement with Yoo and Cynn than
with Olinger et al., in contrast to all preceding computational studies of which we are aware.
In the absence of the present results, one might credibly argue that reconciliation among
the various experimental and theoretical results could be rationalized in terms of crystal
purity and differences between the shock hugoniot of single crystals and the initial porosity
in pressed powders.
0.22
45
0.2
40
2x1x2 "uniform"
1x1x1 "uniform"
1x1x1 "expt-based"
0.18
30
0.14
Pressure (GPa)
E (Hartree/Unit Cell)
MondoSCF 2x1x2 "uniform"

Yoo -- Cynn
Olinger et al.
35
0.16
0.12
0.1
0.08
25
20
0.06
15
0.04
10
0.02
5
0
0.7
0.75
Fig 4.
0.8
0.85
V / V0
0.9
0.95
0.6
0.7
0.8
V / V0
0.9
Calculated relative energies (left panel) and pressures (right panel) versus
compression ratio V/V0 for -HMX, at the PBE/6-31G** level. The full
curves in both panels are based on a primary cell containing eight
molecules, with full optimization of cell contents at each volume.
In Fig. 5 we compare the results for calculated linear compressions of the monoclinic
lattice parameters to the results of Olinger et al. [28,42] based on the uniform and
experimentally-guided protocols described above. Note that the MondoSCF PBE/6-31G**
predictions for variation in lattice length a and monoclinic angle are predicted fairly
accurately using either protocol. By contrast, calculations based on the uniform approach
underestimate the compressibility along the b axis while overestimating it along the c axis.
The calculated linear compressibilities based on the uniform approach are essentially
isotropic (in the scaled sense depicted in Fig. 5), whereas experiment indicates significant
anisotropy.
Our motivation for designing the experimentally-guided protocol was concern that
there might exist geometrically nearby energy valleys, due for instance to subtle
81
conformational changes in molecular geometry, that would depend sensitively on our initial
guess for the crystal structure away from zero pressure. In such an instance, an initially
shallow barrier between valleys would likely become steeper and deeper with increasing
compression, effectively steering our predictions away from the experimental result (in
much the same way that a hiker might veer far off course by choosing the wrong drainage at
the foot of a mountain range). However, the results in Fig. 5 indicate that, with repeated
iterations of the line search algorithm, cycling over the monoclinic lattice parameters, the
PBE/6-31G** prediction using the experimental protocol tends towards, and appears to be
converging to, the uniform result. (The arrows in Fig. 5 point from the result of the first
iteration towards that of the final one.) This suggests that the chosen model chemistry and/or
basis set is insufficient for -HMX crystal. It also underscores the need to perform repeated
optimization cycles in this kind of calculation since, at least for the specified lattice
optimization protocol, a single iteration can lead to incorrect conclusions.
1.02
Linear compressibility (x/x0)
0.98
0.96
Crosses (x) denote

experimentally-guided
MondoSCF results
a (MondoSCF)
b (MondoSCF)
c (MondoSCF)
(MondoSCF)
a (Olinger et al.)
b (Olinger et al.)
c (Olinger et al.)
(Olinger et al.)
0.94
0.92
0.9
0.88
0.72
0.76
0.8
0.84
0.88
0.92
0.96
V / V0
Fig 5.
Comparisons of the linear compressions in -HMX predicted using

PBE/6-31G**, based on the uniform and experimentally-guided
protocols. The results suggest that PBE/6-31G** is an insufficient model
chemistry/basis set combination for -HMX crystal.
Very recently, the ability to automatically perform optimization of lattice degrees of

freedom in conjunction with optimization of molecular geometry has been incorporated into
MondoSCF, along with the ability to employ hybrid density functionals such as PBE0 that
generally do a better job at predicting molecular crystal structures for materials in which
dispersion interactions are significant. While still at the validation stage, we expect that these
two advances in code capability will improve our ability to calculate cold curves for
energetic materials.
82
3.3
Preliminary medium- and large-scale molecular dynamics

simulations of shocked HMX
We have begun to consider the feasibility of million or even multi-million molecule

simulations of plasticity (and dissipation more generally) in polyatomic molecular and
polymeric systems, under a variety of dynamic loading scenarios. We strongly suspect that
the dominant dissipative mechanisms culminating in plastic deformation in complicated
polyatomic molecular materials will differ significantly from the relatively tidy models that
have evolved for metals since the first suggestion of dislocations in the 1930s. With specific
connection to energetic materials, we view this work as a logical extension to some of the
studies performed by Armstrong and co-workers [3,4].
Figure 6 depicts 2x2x1 projections of the crystal structures of aluminum and PETN,
shown on the same scale. Beyond the obvious fact that the molecular materials of interest to
us contain many intramolecular degrees of freedom (as opposed to none for simple metals),
we expect that specific topological features (rings, floppy arms, pendent nitro groups, etc.)
will play important roles in determining the origins and time scales for plastic relaxation in
these kinds of materials.
PETN
Al
Fig 6. Comparison of Aluminum and PETN
crystal structures.
In Fig. 7 we show initial and final snapshots from a simulation of shocked -HMX,
calculated using the LAMMPS computer code developed at Sandia National Laboratory [43].
Our choice to consider -HMX is based on its orthorhombic structure, which is a present
limitation with the LAMMPS code. Microcanonical (NVE) conditions were used, with
periodic boundary conditions applied in the transverse directions. A large region of vacuum
was included at the right-hand end of the simulation cell. The total system size in this
medium-scale simulation was 4096 -HMX molecules (512 unit cells, 114,688 atoms;
initial physical dimensions: 12 nm x 41.4 nm x 2.4 nm), 256 of which formed a rigid,
immobile slab (left edge; see Fig. 7). The remaining molecules were assigned an initial
particle velocity of 3 km/sec directed towards the immobile slab. The orientation is such that
vertical and horizontal axes in the figure correspond to the a and b crystallographic
directions, respectively; hence, the shock is directed along b, and propagates from left to
right. The bottom snapshot corresponds to a time of approximately 4.4 ps, for which roughly
six molecular layers are unshocked (right-hand edge). Close inspection reveals an initial
zone of relatively ordered uniaxial compression (starting about seven layers in from the
right-hand edge), behind which significant disordering occurs. The preliminary results
shown here do not indicate obvious plastic deformation of the compressed solid along slip
planes, but may hint at melting within the material.
83
Fig 7.
Snapshots from a simulation of shocked -HMX containing 4096 fully

flexible molecules. The shock is directed along the b crystallographic axis,
with a particle velocity of 3.0 km/sec.
Results for a shock analogous to the one just described, but directed along the c
crystallographic axis are shown in Figs. 8 and 9. Figure 8 contains the initial and final
snapshots from the simulation, and is intended for comparison with Fig. 7 to underscore the
anisotropic nature of the molecular crystal. Figure 9 contains plots of the density (left panel)
and components of temperature parallel and transverse to the direction of shock
propagation (right panel).
Fig 8.
As in Fig. 7, but for a shock directed along the c crystal axis.
84
Fig 9.
Density (left panel) and temperature (right panel) for the shock shown in
Fig. 8. Densities at several different instants are superimposed, referenced
to the shock front; temperatures were obtained by averaging over several
instants, yielding three traces.
We have also performed much larger simulations, containing 65,536 fully flexible HMX
molecules colliding with a fixed slab of 1,024 molecules. We think that these much larger
simulations (physical dimensions 3.0 nm x 156.4 nm x 38.7 nm) are required in order to
eliminate (or at least mitigate) boundary condition effects in the directions transverse to
shock propagation. An example is shown in Fig. 10, for the case of a shock directed along
the b-crystallographic axis with a particle velocity of 2.0 km/sec. In this case, complex
structure behind the shock front is evident, and signs of plastic relaxation are present (notice
for example the rough upper and lower edges of the simulation cell). Even in this case, we
are still considering thin slab geometries in which the constraint of a system whose
dimension into the plane of the page is only 3.0 nm possibly affects the observed result.
Nevertheless, these calculations are useful as a scoping tool prior to performing much
larger simulations corresponding to a simulation cell that is nearly square in the directions
normal to shock propagation.
85
Fig 10. Cutout from a shock in system containing 65,536 flexible -HMX
molecules. The shock is propagating from left to right along the
b crystallographic axis with a particle velocity of 2.0 km/sec.
4.
CONCLUDING REMARKS
Atomistic simulation methods are increasingly useful for the prediction of constituent
properties of high explosives that are difficult to obtain experimentally. Up until the present
time, most of our studies of real condensed-phase explosives have been based on classical
molecular dynamics or Monte Carlo and focused on predictions of equilibrium properties of
HMX requested by mesoscale modelers. Among the high-priority properties yet to be
calculated are the temperature and pressure dependent specific heats of the various phases
[44]
, the melting curve Tm=Tm(P) [45], and the (rate dependent) shear modulus of the
liquid/glassy states as functions of thermodynamic state. In addition to calculations of the
properties just listed for individual materials, the recent completion of the force field for
BDNPF/A [20] should enable detailed study of essentially all non-reactive intermolecular
interactions among the constituents of PBX-9501.
As the new results shown here indicate, emerging condensed phase electronic-structure
methods enable cold-curve calculations using sophisticated model chemistries and basis sets,
for periodic systems containing hundreds of atoms. However, these results also demonstrate
the importance of careful thought and a skeptical eye in designing the computational
protocol. Even with the very recent incorporation of automated lattice optimization into
MondoSCF, it may be necessary to incorporate at least some level of finite-temperature
annealing to achieve high confidence in the calculated zero Kelvin properties. It is
anticipated that practical calculations of vibrational spectra and chemical energetics (and
dynamics, for carefully chosen situations) will soon be feasible, for primary cells
approaching 1000 atoms. These calculations should become even more reliable, and
interesting, once spin-unrestricted states are incorporated into MondoSCF.
86
With reasonable allocations of time on large Linux clusters, million-molecule

simulations of shocks and other dynamic loading scenarios in inert high explosives are
essentially within reach, using realistic particle velocities to study dissipation and plastic
deformation in these large polyatomic systems. This line of research should be of use in the
development of improved formulations of plasticity in organics. One of the major hurdles
that must be overcome, however, is the question of how to identify, characterize, and
quantify plasticity and plastic deformation mechanisms for this class of materials. While it is
unclear exactly what these mechanisms will be, it is almost certain that they will be
significantly different from those in metals. Moreover, with the ongoing validation and
application of reactive force fields to high explosives (e.g., ReaxFF [46], developed at
CalTech), it should be possible in the not-too-distant future to consider similar calculations
for reactive shocks or cookoff in systems containing tens of thousands of molecules.
Finally, given the difficulty in obtaining reliable experimental data for force field
validation for states far from equilibrium, it may be profitable to merge the capabilities of
cutting-edge tools like MondoSCF and ReaxFF (and related force field formulations under
development by others), by using high-level, benchmark electronic-structure calculations of
energies and forces for condensed-phase geometries sampled in ReaxFF dynamics
simulations, in order to improve the accuracy of the force field in the regions of
configuration space of most interest. (This is essentially an extension to the condensed phase
of the philosophy used in the gas phase to parameterize the present version of ReaxFF,
which appears already to yield fairly credible results [47].)
Thus, it is our opinion that, with careful thought and planning, combined with a healthy
dose of self-criticism and skepticism, atomistic simulation methods constitute a practically
useful means of obtaining equilibrium properties of molecular systems in regimes that might
not otherwise be obtainable. They will begin to play even larger roles than at present [47,48] in
understanding non-equilibrium processes and mechanisms in energetic materials, for both
non-reactive and reactive systems. In summary, we think that, in the not-too-distant future,
atomistic simulation methods should be capable of providing significant insights into each of
the three distinct classes of knowledge required to understand initiation, as outlined in the
opening paragraph of this contribution.
Acknowledgment
I wish to acknowledge my numerous collaborators in the efforts described above: Dan
Hooks, Ralph Menikoff, Chak Ayyagari, Dmitry Bedrov, Oleg Borodin, Hemali Davande,
Grant Smith, Chee Kwan Gan, Matt Challacombe, Eugenio Jaramillo, and Alejandro
Strachan. Although I used the first person plural (we, our) in this document, the editorial
opinions expressed here are mine and do not necessarily coincide with those of all
collaborators under all circumstances. [LA-UR 05-1826]
87
REFERENCES
[1]
M.R. BAER, Modeling heterogeneous energetic materials at the mesoscale, Thermochimica

Acta 384, 351 (2002).
[2]
T.D. SEWELL and R. MENIKOFF, Complete equation of state for -HMX and implications
for initiation, in Proceedings of the 2003 APS Topical Conference on Shock Compression of
Condensed Matter, M.D. Furnish, Y.M. Gupta, and J.W. Forbes Eds. (AIP, Melville, NY,
2004) p. 157.
R.W. ARMSTRONG, H.L. AMMON, W.L. ELBAN, and D.H. TSAI, Investigation of hot
spot characteristics in energetic crystals, Thermochimica Acta 384, 303 (2002).
R.W. ARMSTRONG and W.L. ELBAN, Dislocations in Energetic Crystals, in Dislocations
in Solids, F.R.N. Nabarro and J.P. Hirth Eds. (North-Holland, New York, 2004) Ch. 68.
R. MENIKOFF, Pore collapse and hot spots in HMX, in Proceedings of the 2003 APS
Topical Conference on Shock Compression of Condensed Matter, M.D. Furnish, Y.M. Gupta,
and J.W. Forbes, Eds. (AIP, Melville, NY 2004) p. 393.
R. MENIKOFF, Detonation Wave Profile in PBX-9501 (LA-UR 05-1633) to appear in the
Proceedings of the 2005 APS Topical Meeting on Shock Compression of Condensed Matter;
presently available online at: http://t14web.lanl.gov/Staff/rsm/Preprints/CJprofile.pdf .
D.M. HANSON-PARR and T.P. PARR, Thermal properties measurements of solid rocket
propellant oxidizers and binder materials as a function of temperature, Journal of Energetic
Materials 17, 1 (1999).
R. MENIKOFF and T.D. SEWELL, Constituent properties of HMX needed for mesoscale
simulations, Combustion Theory and Modeling 6, 103 (2002).
R. MENIKOFF and T.D. SEWELL, Fitting forms for isothermal data, High Pressure
Research 21, 121 (2001).
T.D. SEWELL, Monte Carlo simulations of crystalline TATB, in Decomposition,
Combustion, and Detonation Chemistry of Energetic Materials, Materials Research Society
Symposium Proceedings, T.B. Brill, T.P. Russell, W.C. Tao, and R.B. Wardle Eds., (MRS,
Pittsburgh, 1996) Volume 418, p. 67.
T.D. SEWELL, Monte Carlo calculations of the hydrostatic compression of hexahydro-1,3,5trinitro-1,3,5-triazine and -octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine, Journal of
Applied Physics 83, 4142 (1998).
T.D. SEWELL and C.M. BENNETT, Monte Carlo calculations of the elastic moduli and
pressure-volume-temperature equation of state for hexahydro-1-3-5-trinitro-1,3,5-triazine,
Journal of Applied Physics 88, 88 (2000).
The HMX simulations resulting from the collaboration between T.D. Sewell at Los Alamos and
the G.D. Smith group at the University of Utah are based on a flexible molecule, quantum
chemistry-based force field: G.D. SMITH and R.K. BHARADWAJ, Quantum chemistry
based force field for simulations of HMX, Journal of Physical Chemistry B 103, 3570 (1999).
D. BEDROV, C. AYYAGARI, G.D. SMITH, T.D. SEWELL, R. MENIKOFF, and J.M.
ZAUG, Molecular dynamics simulations of HMX crystal polymorphs using a flexible
molecule force field, Journal of Computer-Aided Materials Design 8, 77 (2002).
T.D. SEWELL, R. MENIKOFF, D. BEDROV, and G.D. SMITH, A molecular dynamics
simulation study of elastic properties of HMX, Journal of Chemical Physics 119, 7417 (2003).
D. BEDROV, G.D. SMITH, and T.D. SEWELL, Temperature-dependent shear viscosity
coefficient of octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX): A molecular dynamics
simulation study, Journal of Chemical Physics 112, 7203 (2000).
D. BEDROV, G.D. SMITH, and T.D. SEWELL, Thermal conductivity of liquid octahydro1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX) from molecular dynamics simulations, Chemical
Physics Letters 324, 64 (2000).
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
88
[18] D. BEDROV, G.D. SMITH, and T.D. SEWELL, Thermodynamics and Mechanical Properties
of HMX from Atomistic Simulations, in Energetic Materials Part 1. Decomposition, Crystal,
and Molecular Properties, P. Politzer and J.S. Murray Eds. (Elsevier, Boston, 2003), Ch. 10, p.
279.
[19] G.D. SMITH, D. BEDROV, O. BYUTNER, O. BORODIN, C. AYYAGARI, and T.D.
SEWELL, A quantum chemistry-based potential for a poly(ester urethane), Journal of
Physical Chemistry A. 107, 7552 (2003).
[20] H. DAVANDE, O. BORODIN, G.D. SMITH, and T.D. SEWELL, Quantum chemistry-based
force field for simulations of energetic dinitro compounds, Journal of Energetic Materials
(submitted).
[21] Verification: are we solving the equations right? Validation: are we solving the right equations?
[22] M.R. BAER and T.D. SEWELL, Linking molecular dynamics to mesoscale simulation, in
Molecular Dynamics Simulations of Detonation Phenomena, B.L. Holian, Ed. (ITRI, Laurel,
MD, 2003) Ch. 5, p. 79.
[23] PBX-9501 is 95% by weight HMX, 2.5% Estane, and 2.5% BDNPF/A. A small amount of
stabilizer is also present in the formulation, but we have no plans to develop a force field for
this trace component.
[24] J.P. LEWIS, T.D. SEWELL, R.B. EVANS, and G.A. VOTH, Electronic structure calculation
of the structures and energies of the three pure polymorphic forms of crystalline HMX,
Journal of Physical Chemistry B 104, 1009 (2000).
[25] C.K. GAN, T.D. SEWELL, and M. CHALLACOMBE, All electron density-functional studies
of hydrostatic compression of pentaerythritol tetranitrate (PETN), Physical Review B 69,
035116 (2004).
[26] M. CHALLACOMBE, E. SCHWEGLER, C.J. TYMCZAK, C.K. GAN, K. NEMETH, A.M.N.
NIKLASSON, H. NYMEYER, and G. HENKLEMAN, MONDOSCF v1.07. A program suite
for massively parallel, linear scaling SCF theory and ab initio molecular dynamics, 2001,
http://www.t12.lanl.gov/home/mchalla, Los Alamos National Laboratory (LA-CC 01-2),
copyright University of California.
[27] The 2003 study Molecular Dynamics Simulations of Detonation Phenomena, chaired by B.L.
Holian and funded by the International Technology Research Institute (ITRI), contains what is
probably the most complete and up-to-date (through mid-2003) compendium of theoretical
studies of energetic materials that has been compiled. Contributing authors include B.L.
HOLIAN, M.R. BAER, D. BRENNER, D.D. DLOTT, A. REDONDO, B.M. RICE, T.D.
SEWELL, and C.A. WIGHT.
[28] B. OLINGER, B. ROOF, and H. CADY, The linear and volume compression of -HMX and
RDX , in Symposium International Sur le Comportement Des Milieux Denses Sous Hautes
Pressions (CEA, Paris, 1978) p. 3.
[29] C.-S. YOO and H. CYNN, Equation of state, phase transition, decomposition of -HMX
(octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine) at high pressures, Journal of Chemical
Physics 111, 10229 (1999).
[30] M. HERRMANN, W. ENGEL, and N. EISENREICH, Propellants, Explosives, and
Pyrotechnics 17, 190 (1992).
[31] M. HERRMANN, W. ENGEL, and N. EISENREICH, Thermal analysis of the phases of
HMX using x-ray diffraction, Zeitschrift fur Kristallographie 204, 121 (1993).
[32] C.K. SAW, Kinetics of HMX and phase transitions: effects of grain size at elevated
temperature, to appear in 12th International Detonation Symposium; currently at
http://www.sainc.com/onr/detsymp/technicalProgram.htm.
[33] See Ref. [18], Fig. 9 (p. 305).
[34] G.M. DAY, S.L. PRICE, and M. LESLIE, Elastic constant calculations for molecular
crystals, Crystal Growth and Design 1, 13 (2001). Note especially entries in the tables where
multiple experimental data are presented for the same substance.
89
[35] D.C. SORESCU, B.M. RICE, and D.L. THOMPSON, Theoretical studies of the hydrostatic
compression of RDX, HMX, HNIW, and PETN crystals, Journal of Physical Chemistry B
103, 6783 (1999); and numerous references therein.
[36] S. HAUSSHL, Elastic and thermoelastic properties of organic crystals Zeitschrift fur
Kristallographie 216, 339 (2001).
[37] DAN HOOKS (LANL), private communication to Tommy Sewell (March 2005).
[38] MERRILL BECKSTEAD, private communication to Tommy Sewell.
[39] P.M. AGRAWAL, B.M. RICE, and D.L. THOMPSON, Molecular dynamics study of the
melting of nitromethane, Journal of Chemical Physics 119, 9617 (2003).
[40] J.C. GUMP and S.M. PEIRIS, Isothermal equations of state of -octahydro-1,3,5,7-tetranitro1,3,5,7-tetrazocine, Journal of Applied Physics 97, 53513 (2005).
[41] T.D. SEWELL (in preparation).
[42] For the present purpose, the differences between the linear compressions of Olinger et al. and
Yoo & Cynn are probably negligible.
[43] S.J. PLIMPTON, Fast Parallel Algorithms for Short-Range Molecular Dynamics, Journal of
Computational Physics 117, 1 (1995); S.J. PLIMPTON, R. POLLOCK, M. STEVENS,
Particle-Mesh Ewald and rRESPA for Parallel Molecular Dynamics Simulations, in
Proceedings of the Eighth SIAM Conference on Parallel Processing for Scientific Computing,
Minneapolis, MN, March 1997.
[44] W.A. GODDARD III and co-workers have done some preliminary calculations of the specific
heat Cv of -HMX for temperatures up to 3000 K. These can presently be found in annual
reports
of
the
CalTech
ASC/ASAP
Level
One
Alliance:
http://csdrm.caltech.edu/publications/annualreps.html; see also Ref. [8], p. 108.
[45] See Fig. 1 of Ref. [8] and associated discussion.
[46] A. STRACHAN, A.C.T. VAN DUIN, D. CHAKRABORTY, S. DASGUPTA, and W.A.
GODDARD III, Shock waves in high-energy materials: the initial chemical events in
nitramine RDX, Physical Review Letters 91, 098301 (2003).
[47] On the basis of many calculations performed recently by Alejandro Strachan (LANL) and Adri
van Duin (CalTech).
[48] M.R. MANAA, L.E. FRIED, C.F. MELIUS, M. ELSTNER, and T. FRAUENHEIM,
Decomposition of HMX at extreme conditions: A molecular dynamics simulation, Journal of
Physical Chemistry A 106, 9024 (2002).
90
HIGH-TEMPERATURE PROPELLANT TORCH-SYSTEM

FOR NON-DETONATIVE NEUTRALIZATION OF MINES;
SOME PHYSICAL AND CHEMICAL ASPECTS
Allen J. Tulis
Applied Research Associates, Inc., Chicago, IL, U.S.A.
Abstract:
A novel, proprietary rocket-type torch, using very-high-temperature pyrotechnic and
propellant compositions initially developed at IIT Research Institute (IITRI) under U.S.
Government contract, has been extensively further developed and demonstrated in recent
years at Applied Research Associates, Inc. (ARA) to be effective in penetrating all explosive
mines and neutralizing all mine explosive types tested without causing the explosive to
detonate. This ARA proprietary PTS torch is based mainly on the use of aluminum and
potassium perchlorate to optimize its temperatue and performance; the computed chamber
temperatures are in excess of 4000 K. Total torch burning times are generally in excess of
30 sec, although penetrations of thick steel are within a few sec, mandatory for the
penetration of steel casings since long penetration times cause excessive heating of the steel
which can prevent penetration as well as allow the buildup of slag and other solid
condensables from the torch products that can interfere with the subsequent capability of
igniting the explosive. The long burn time of the torch after penetration is required to assure
adequate ignition of the explosive, particularly TNT. A similar but more robust and heavier
torch developed in Russia at the Semenov Institute of Chemical Physics (SICP) of the
Russian Academy of Sciences penetrates 25mm steel within seconds (Ref 1). This earlier
effort emphasized the penetration of thick steel, wood, and thermoplastics of surface
exposed mines, the so-called hardened mines. This work has now been extended to mines
that are covered with water and/or soil, to the extent encountered in the Global
Humanitarian Demining Program. Work involving mines under water and/or soil is
currently underway by the U.S, Government; some work conducted under IR&D in the U.S.
at. ARA as well as at SICP is described here, with emphasis on the physical and chemical
aspects of neutralizing the explosive within the mine after penetration of the casing is
achieved. In particular, because of its inadequate oxidizer balance, TNT is very difficult to
initiate and achieve sustained autocatalytic decomposition when attempted under conditions
wherein air (oxygen) is inadequate or absent; e.g., underwater and underground. When
burned in open air, TNT decomposes/combusts but still provides voluminous black smoke
due to inadequate combustion of carbon.
Keywords:
1.
torchion, high-temperature compositions, penetrating mines,

neutralizing explosive
INTRODUCTION
Humanitarian Demining that addresses the global proliferation of landmines continues to be

a major effort of the U.S. Government; especially effort to develop expedient, safe, simple, and
economic methodology. This paper describes a highly-effective extremely-high temperature
thermal-initiation that provides for the complete explosive destruction of mine explosives that of
91
itself precludes detonation by the concomitant penetration of the mine casing and autocatalytic
neutralization of the main explosive charge within the mine, irrespective of type of casing or
explosive. It is based on rocket technology, using very-high-temperature pyrotechnic/propellant
compositions that are ignited in a throttled-down form from a cylindrical reaction chamber
through a nozzle to provide a flame jet that readily penetrates even heavy steel casings. The PTS
torch is illustrated in Figure 1.
Fig 1.
Propellant Torch System (PTS) in Early (1998) Field Application
Figure 1 illustrates this PTS torch positioned to neutralize a surrogate wood case, cast TNT
mine. Of course, mines in place worldwide are fuzed and initiation of the detonator and/or
booster cannot be absolutely precluded using burn-type non-detonative neutralization
techniques. However, in numerous tests against fuzed AT/AP mines, the fuzes did detonate after
many minutes as the burning progressed and/or were nearly completely neutralized, and
generally did not cause detonation of any unburned explosive, which continued to burn to
completion. The burned out plastic and steel casings from these mines were intact after
neutralization was completed.
The PTS is safe, simple, reliable, highly effective, environmentally friendly and with a very
low-cost potential. Scientists at SICP in Russia, in collaboration with ARA, helped in the
development of this PTS technology, particularly with their modified Propellant (thermite)
Torch, for the tremendously rapid penetration of thick steel.. Figure 2 illustrates the first
prototype of this torch, which was designed to penetrate 40 mm steel. Figure 3 illustrates the
torch penetrating 22 mm steel and Figure 4 shows the penetrated plate.
92
Fig 2.
Russian Torch for Penetrating Thick Steel
In the initial development of the U.S. Government PTS, longer than desired burn times were
required to initiate the autocatalytic decomposition of TNT that was not adequately exposed to
oxygen of the air, such as when buried, confined in a heavy casing (very large steel, wood,
plastic casings with a very small penetration hole), or underwater. In the 1970s, the hypergolic
initiation of TNT and TNT-based explosives were successfully developed using reactive amines
and metal alkyls, and to some extent the interhalogens (although these smetimes sometimes
caused instantaneous detonation) (Ref 2). A similar approach was taken in the development of
the PTS for the autocatalytic decomposition of TNT. Further effort is under study to
demonstrate such capability with the use of zinc in the PTS (proprietary development).
Fig 3.
Russian Torch Penetrating

by the Russian PT
Fig 4.
93
Penetrated 20-mm Steel Plate
The basis of this appears to be as follows: (1) for penetration of mine casings, the coarse
zinc (which does not react to any significant extent within the torch flame) melts and vaporizes
to provide an abrasive/kinetic and highly effective heat transfer mechanism on the target, and
(2) for the autocatalytic initiation of TNT, the molten/vaporized zinc enters the penetrated mine
casings and induces a heterogeneous hypergolic reaction with the hot, molten/vaporizing TNT.
The PTS composition developed for the U.S. Government consisted of aluminum (Al),
potassium perchlorate (KP), Teflon (C2F4 polymer) and zinc (Zn). The computed products are
mainly Al2O3(l) and its gaseous decomposition products AlO, O, Al and Al2O and KCl(g),
AlOF2(g), CO(g) and Zn(g). Although the PTS technology and the specific PTS composition
developed for the U.S. Governments demonstration tests are the result of prior government
contracts, most of the concepts presented in this paper are based on effort are ARA proprietary.
2.
TECHNICAL DISCUSSION
Because of TNTs low melting point of about 78 C, it is very difficulty to burn unless
oxygen from ambient air is sufficiently available. TNT burns initially in the vapor phase. The
TNT melts and, as further heat is applied, vaporizes and creates a boiling froth. The vapors
above this froth will ignite and combust/decompose. However, if the TNT is confined in a
closed space under excessive heat and high temperature, it will not ignite but will decompose to
gaseous species and pyrolize; unless an adequate heterogeneous decomposition/dissociation of
the solid TNT can be achieved, as is believed to be the case under critical conditions involving
hypergolic reactions with the solid and/or molten TNT. This has been demonstrated and results
in a very white-hot jet-type flame and the disappearance of the typical black smoke; i.e., a nonsmoke product. This mechanism has not been specifically identified at this time, although the
conditions are believed to be empirically known. The following is a description, or a hypothesis
of this mechanism.
If TNT is confined in a metal, wood or plastic casing, especially if buried, simply applying a
flame to the TNT through a penetration hole is generally not sufficient to achieve TNT
combustion. Non-detonative neutralization becomes very difficult, requiring extensive case
rupture to expose the TNT to air or continuing to maintain the ignition source for a very long
time, such that eventually the TNT will shift to the heterogeneous autocatalytic decomposition
mode (which has been identified experimentally on numerous occasions) or will simply be
consumed by such forced combustion/decomposition/pyrolysis. Hence, there is good reason to
anticipate that small amounts of high temperature melted/vaporized metals (such as zinc,
magnesium, copper or iron) blasted into hot, molten and decomposing TNT (as achieved by the
PTS) will induce this transition to autocatalytic decomposition. The mechanism is the same as
with the metal alkyls; i.e., the highly reactive (hot!) metal will react with the nitro groups
directly upon contact and induce the highly exothermic autocatalytic decomposition. Once
induced, this reaction will not quench. Efforts to quench it by deep burial or confinement will
cause very rapid pressure buildup and expulsion from its confinement or possibly transition to
detonation. This autocatalytic decomposition mechanism also proceeds without the typical
evolution of voluminous quantities of black, carbonaceous smoke observed in the simple air
combustion of TNT. However, only zinc will remain essentially as an atomic element in the
extremely high temperatures of the PTS flame. The inclusion of large enough particles of highly
reactive zinc in the PTS will not alter the stoichiometry of the PTS reaction to any great extent,
as the large particles of zinc will simply melt and possibly vaporize before being carried into the
94
explosive as very hot and reactive molten/vaporized zinc. Other metals, such as magnesium,
copper, and iron oxide could also be effective in this manner, but readily form stable oxides
whereas zinc does not. Zinc vaporizes to a gas at the PTS temperatures. In the case of
magnesium, although considerable magnesium gas is produced at these temperatures, MgCl and
MgO are produced, as well as liquid MgAl2O4 along with lesser amounts of other magnesium
species. Magnesium readily reacts with O2 to form MgO, which has a stable boiling point of
nearly 3900 K. With the inclusion of iron oxide in place of zinc; the iron oxide is reduced to
gaseous Fe and provides additional oxygen. This mechanism is believed to be especially
effective for steel penetration.
Hence, the major significance of zinc additive is that it
exists as gaseous zinc in the PTS jet and impinges upon the mine casing, giving up its latent heat
of condensation to the casing, and after the mine casing is penetrated, enters into the mine to
cause hypergolic chemical reaction with the hot, melting/vaporizing TNT
Figurers 5 and 6 are included here to demonstrate the influence of zinc on the neutralization
of TNT in similar experiments. This is highly suggestive, but to date not sufficiently conclusive
proof of the zincs adequate hypergolic performance.
Fig 5.
Surrogate Wooden Mine Test

(without zinc)
Fig 6.
Surrogate Wooden Mine Test

(with zinc)
In the combustion of the PTS composition, the KP dissociates and provides the oxidizer
(O2) to react with the aluminum powder. This extremely exothermic reaction decomposes the
Teflon and releases carbon and F2, which then reacts with the Al and O2 to form AlOF2 and
KF. The Cl from the KP reacts to form chlorides with K. The Zn remains as elemental Zn,
which does not oxidize but melts/vaporizes. All of the chemical oxidative reactions are
exothermic, whereas all physical phase changes solid to liquid to gas are endothermic. Zn
particles (dependent on particle size, induction time, heat/mass transport) will result as gaseous,
molten, and solid phases before being expelled out of the PTS nozzle. At these high
temperatures the Al2O3 is a liquid. However, as the reaction temperature exceeds about 3500 K,
molten Al2O3 begins to dissociate into gaseous species. The KP is in a gaseous state. Both KCl
gas and Al2O3 liquid/gas provide tremendous latent heats to the mine casings as the PTS jet
impinges upon them. Gaseous Zn also contributes to this, although its specific chemical/kinetic
action for penetration, although positively demonstrated, has not been completely established;
its major inclusion is for its potential catalytic/hypergolic reaction with TNT and TNT-based
explosives.
95
3.
NEUTRALIZATION OF BURIED MINES AND MINES UNDERWATER
Recent effort has been extended to the neutralization of mines that are covered with an
overburden of soil and other debris, as well as those covered by shallow water. The following
early results were conducted under Internal Research and Development at ARA and SICP.
Because of the extremely high temperatures generated by the PTS, most materials will vaporize
and/or dissociate as well as being physically removed from the powerful jets that are created.
Both aspects are being used to develop the optimum approach to both of these applications. The
following figures illustrate results of early experiments; additional effort is anticipated.
Fig 7. The experimental setup for GPPTS

testing; soil overburden (sand, gravel,
etc.) was inserted over the buried mine
and compacted to within 1-in. standoff
of the PTS torch) and a 1/8-in. thick
steel plate was placed on top of the
surrogate mine (when used).
Fig 8. Test setup prior to placing the steel

plate on the mine and burying it with
2.5-in. of Overburden (soil, sand,
aggregate).
Fig 9. Post-test results in of this test; all of the Fig 10. The hole from this test; (burned out
Comp B explosive was burned out.
mine and torch setup removed). Note
Note that the casing showed no signs of
the large hole blasted out by the PTS
burnout, being insulated by the soil
torch
surrounding it; although the top of the
mine was completely burned out by the
Comp B explosive flames.
96
Fig 11. A photo of the 5/8-in. Diam. clean hole

in the steel plate and the partially
burned out PTS torch, also due to the
Comp B explosive flames.
Fig 12. A test using a thermite/propellant

formulation PTS (longer burn time) on
a buried TNT surrogate PVC mine
Fig 13. Results from the test depicted in Figure Fig 14. . Pyrolized TNT in an underwater
12 using TNT; the mine was penetrated
experiment; only a small hole was
and the TNT ignited but quenched after
penetrated and the TNT was confined
burning partially. (Note: No zinc used
in steel without access to air. The TNT
in this PTS torche.)
was completely pyrolized inside the
steel surrogate mine, although initially
the TNT burned as observed by flames
above the water surface.
97
Fig 16. Large surrogate 5.7 Kg cast TNT

mine also tested underwater, with the
penetrated ~3 mm (1/8-in.) steel top
plate removed and pyrolized TNT
wherever the UWPTS torch jet-flame
contacted the explosive; about 650 g
TNT pyrolized during the burn time of
the UWPTS torch which burned
completely through the mine
Fig 15. Recovered UWPTS pyrolized TNT

residue from a galvanized steel
surrogate AP mine with >200 g cast
TNT.
4.
CONCLUSIONS
As was anticipated, the complete autocatalytic decomposition of buried and/or underwater

TNT mines, and sometimes even Comp B mines, is difficult. A potential advantage may be
useful by inducing hypergolic heterogeneous reaction of these explosives, as appears to be
feasible and has been demonstrated on some experiments with these torches. Another approach
is to achieve the complete pyrolysis of the mine explosives with more powerful and extended
burn times of the torches. A major variable is the size and casing of the mines. There are also
several other approaches involving the chemical kinetics of the explosives which are expected to
be investigated.
The PTS provides for a safe, simple, fast, effective, reliable and low-cost means to
neutralize surface-exposed hardened mines, and is expected to be developed to also be effective
against buried and underwater mines (demonstrated effectively against explosives other than
TNT). It needs to be noted that these PTS torches do not contain any explosives, toxic materials,
and are environmentally acceptable
98
Acknowlegments
The author wishes to express his deep appreciation to his colleagues at Applied Research
Associates, Inc., as well as former colleagues at IIT Research Institute. Support from Profs.
Yurii V. Frolov and Alla Pivkina of the SICP in Russia, Dr. David R. Dillehay, president of
Technical Consultants, Inc., and Mr. C. James Dahn, president of Safety Consulting Engineers,
Inc., is especially recognized. Finally, Dr. Divyakant L. Patel of CECOM NVESD, the U.S.
Government COTR for most of the past government programs, is especially appreciated.
REFERENCES
[1]
[2]
A.J. TULIS, D.D. DILLEHAY, Y.V. FROLOV, D.L. PATEL, AND J. DILLON: Chemical and Physical
Aspects of a Pyrotechnic Torch in Penetrating and Non-Detonatively Neutralizing all UXO, Proc.
28th Int. Pyrotechnics Seminar, Adelaide, Australia, Nov., 2001.
A.J. TULIS ET. AL.,: Hypergolic fluid Jet Destruction of Landmines, Ninth Explosives and
Pyrotechnics Symposium, Philadelphia, PA, Sept. 1976.
99
HIGHLIGHTS OF STABILITY RESEARCH IN THE FIRST DOZEN

OF JAN HANSSON SYMPOSIA 1967 2001
F. Volk
Fraunhofer Institut fr Chemische Technologie, ICT
Postfach 1240, D-76318 Pfinztal-Berghausen, Germany
Abstract
This paper is dealing with a review of reports on chemical stability of propellants
presented in the first Twelve Jan Hansson Symposia (S1 S12) on Chemical Problems
Connected with the Stability of Explosives 1967 2001.
Keywords:
1.
review, chemical stability, propellants, Jan Hansson Symposia
INTRODUCTION
The aim of this presentation is to review the first dozen of stability symposia in Sweden
1967-2001. This review is not intended to be comprehensive, but reflects also the interest of
the author, who was involved for many years with the problems of testing the chemical
stability of explosives, gun propellants and double base rocket propellants.
Moreover, he was witness of the rapid development of the analytical chemistry,
especially in the field of the chromatographic techniques such as gas chromatography (GC),
high performance liquid chromatography (HPLC), gel permeation chromatography (GPC),
microcalorimetry and the chemiluminescence techniques. He attended all these symposia
from 1973 to 1995 and had many fruitful discussions with the late Jan Hansson, the initiator
of these conferences.
We would like to thank Jan Hansson very much for the invitation of many famous
scientists from different countries, also from East Europe, who most of us could meet only in
Sweden, before the fall of the wall in 1989. In this way we could win a good insight into
their scientific work at that time.
As a chemist, Jan Hansson was interested in chemical reactions connected with stability.
Therefore the emphasis of the symposia is on chemical problems with regard to stability.
This is best shown with the schematic reaction
k
1 B + C + S
A
HR,1
This means, we have educts, products, and reaction enthalpy. We are interested in
reaction mechanisms, reaction rates, reaction conversion, group effects, bond energy,
activation energies, pre-exponential factors and last but not least in measurement methods
which are able to characterise and describe the chemical reactions.
100
2.
DECOMPOSITION OF NITRATE ESTERS
The thermal decomposition of nitrate esters has been studied by several investigators.
J. Tranchant (S6 / 1) and B. Lurie (S9 / 119) gave a detailed description of the different
reaction steps.
The thermolysis of nitrate esters begins with the breaking of the nitrated groups and the
formation of a free radical and NO2.
R O
NO2
RONO2 + R
. + .NO
R O
.
O
RCHO
RONO2 + NO2
RCOOH
+
COOH
COOH
NO + N2O + CO + CO2 + N2 + H2O

RONO2 + H2O
ROH + HNO3
NO + Air
NO2 + N2O4
The initial decomposition phase is immediately followed by secondary reactions, either

recombination of the R O radical with the remaining nitrate ester or oxidation of the
neighbouring chains by nitric dioxide. These reactions are characterized by abstraction of the
hydrogen atom, disproportionation and destructive decomposition with the fission of a C C
bond:
RCH2O
+ RCH2ONO2
. + NO
.
.
RCH O + RCH O
RCH2O
RCH2OH + RCHO + NO2
RCHO + HNO2
RCH2OH + RCHO
RCH2OH + 2 NO2
RCH2ONO + HNO3
RCH2O
R + HCHO
The activation energy of the primary decomposition process, which is calculated from
the temperature dependence of the initial reaction rate, is approximately 160 kJ/mole. In an
overview, about 36 different Arrhenius parameters of nitrocellulose, nitroglycerine, ethylene
diglycol dinitrate (DEGN), triethylene glycol dinitrate (TEGDN) and other nitrate esters are
tabulated from different investigations and authors.
101
Also Caire-Maurisier and Tranchant (S4 / 249) found in connection with the
investigation of the decomposition mechanism of nitroglycerine products such as
formaldehyde, glyoxal, formic acid and oxalic acid.
Finally we can explain the low stability of nitrate esters by the relatively low bond
strength of the O N bond which needs only 37 kcal for its rupture (P. Gray: Transactions
Faraday Society 51 (1955), 1367).
But beside the bond rupture (thermolysis), there are some other possibilities of the
degradation of nitrate esters during the investigation of the ageing behaviour such as ester
hydrolysis or saponification of the nitric acid ester group and chain splitting by hydrolysis of
the nitrocellulose chain. As we see in the following picture , there are significantly different
activation energies, see Bohn (S11 / 89), Lurie (S9 / 119) and (S12 / 163), and Bellerby and
Elsby (S11 / 175).
-16
ln(kA [1/s])
Arrhenius plot of the three

main decomposition types of NC
temperature range 30C to 100C
90C
-20
65C
-24
30C
ester thermolysis, 163 kJ/mol (38.9 kcal/mol)

-28
ester hydrolysis,
chain hydrolysis,
-32
0,0026
100 kJ/mol (23.9 kcal/mol)

56 kJ/mol (13.4 kcal/mol), catalysed
1/T [1/K]
0,0028
0,0030
0,0032
0,0034
Consequence of the different activation energies of the three main decomposition

reactions of NC on the experimental determination of the ageing data of NC-based
propellants. The intersection points of the three Arrhenius lines lie in the temperature range
55C to 70C.. But dependent on the decomposition product spectrum (water and HNO3
content) they can be shifted to higher or lower temperatures because of changes in the preexponential factor.
In connection with the stability tests we should mention that many investigations have
also been conducted with the use of differential thermoanalytic methods such as TG, DTA
and DSC. For those publications in the field of explosives, P.G. Laye has prepared a detailed
review in S11 / 241.
The formation of NO2 and of traces of acids and aldehydes catalyzes and accelerates the
decomposition process, so that after a distinct time the so-called autocatalytic decomposition
process is starting (Brook S4 / 55). In order to prevent this process or to guarantee a longer
life time of a nitrate ester containing propellants, stabilizers are added in the propellant
formulation. The function of the stabilizer is to react rapidly with the nitrogen oxides so that
they are removed from the system before they can catalyze the decomposition reaction. So
propellants can be safely stored for many years.
For quality assurance, tests are required to assess the stability of propellant batches
during the manufacturing. These tests should be sufficiently rapid to enable the propellant to
be tested within a short time of propellant manufacture. In this case, the propellants are aged
102
at a relatively high temperature to complete the test in sufficiently short time. The
disadvantage of this procedure is that decomposition reactions at higher temperatures are
usually not the same as those at normal storage temperatures. Therefore for the accurate
assessment of storage life, a temperature as close as possible to ambient temperature must be
used, so that the extrapolation to several years remains valid.
3.
STABILITY TESTS
The most important stability tests are:
a)
measuring the gas evolution
b)
measuring the weight loss
c)
measuring the heat evolution
d)
measuring the stabilizer consumption
e)
following the variation of the physical-chemical properties of the polymer,

such as the change of the mean molecular weight
f)
measuring the ballistic stability
The most common tests are listed in the following table (Brook S4 / 55):
Stability tests
Test
Temperature
Time of test
Abel Heat
80C
10 30 min
Remarks
KJ-starch paper
Methylviolet
120C
134.5C
30 40 min
NOX turns methylviolet paper pink
Silvered vessel
80C
> 600 h
80C surveillance
80C
150 h
Up to self heating in a Dewar vessel

Decomposition measurement by Abel
Heat Test before and after ageing
65.5C or 65C
surveillance
65.5C
65C
Very long
2h
Up to the appearing of brown fumes

(NO2)
Stability of Nitrocellulose
Nitrogen oxide evolved is measured
quantitatively
Stability of Nitrocellulose
NOX decomposition is measured
as N2 quantitatively
Bergmann-Junk
132C
Will test
135C
Dutch
110C
105C
72 h
Weight loss: 8 h open test tubes,

64 h closed test tubes
90C
weight loss test
90C
Several days
Up to the formation of brown fumes

(NO2)
103
Although the 65.5C or 65C surveillance test is the dominating test especially for gun
propellants, this test is not always reliable, because sometimes the emission of red fumes
does not appear at all. So during the long storage of samples at least 6 months at 65C a
spontaneous ignition of the propellant can occur. The following times up to the beginning of
the formation of red fumes at 65C serve as the standard for assessment in Sweden, as Jan
Hansson described in S2 / 1.
Assessment according to the 65.5C surveillance test

Time in months
Assessment
>6
No restriction
46
Use up first
24
Destroy
<2
Destroy immediately
So we understand that modern methods are required, methods that should tell which
changes have occurred in the powder but also methods that should be easy and cheap to be
used in practice. Because the stabilizer is changing in the same way as nitrocellulose
decomposes during the ageing process, we have to look for a method to analyze the stabilizer
consumption.
In S1 / 143 and S1/ 179, Jacobsson and Cronwall described in 1967 a method, which can
overcome the problems mentioned before. By using a quantitative thin layer
chromatographic method (TLC), the transformation of the stabilizers diphenylamine (DPA)
and ethylcentralite (EC) could be determined as a function of the ageing at 65C. This
stabilizer consumption test is a valuable complement to the test at 65 C. It was found that a
propellant develops no red fumes after a storage of > 6 months if the content of
diphenylamine (DPA) is higher than 0.2 % and less than 0.4 % of N-Nitroso-DPA. The same
holds for propellants with ethylcentralite (EC) if the consumption of EC is less than 50 % of
the original stabilizer content. The authors had not found any emission of red fumes from the
powder during the test of 65C without having previously measured that the stabilizing
agents had been consumed.
4.
STABILIZER REACTIONS AND WEIGHT LOSS TESTS
In order to understand better the stabilizer decrease in propellants, A. Alm (S1 / 162)
studied the reactions between nitrogen oxides and diphenylamine (DPA). In 1889, DPA was
introduced by Nobel as a stabilizer for nitrocellulose and is still used in both nitrocellulose
and double base propellants in Sweden. Alm found no reaction between DPA and nitrogen
oxides (NO), there were no changes in either colour or weight. On the other side, if DPA was
exposed to a constant flow of NO together with oxygen, an increase in weight was measured.
The reaction products have been separated by two dimensional thin layer chromatography.
The following main substances could be identified:
DPA, NNODPA, 4-NO-DPA, 2-NO2-DPA, 4-NO2-DPA, and 2,4-(NO2)2-DPA.
104
This means that DPA is able to effectively eliminate nitrogen dioxide during the ageing
of gun propellants.
In addition it was found that also the low-substituted derivatives of DPA react readily
with NO2, such as N-NO-DPA, 2-NO2-DPA and 4-NO2-DPA. The dinitro derivatives
reacted very slowly.
The next step to identify the reaction products of DPA and EC during the ageing of
propellants was explored by several investigators: Ripper (S1 / 188), Alm (S2 / 34), Blay (S3
/ 61), Ammann (S4 / 9 and S5 / 127), Volk (S4 / 29), Sopranetti (S4 / 73), Lebert (S4 / 99),
Lindblom (S5 / 107). A new procedure had been developed to determine the content of
stabilizers and to analyze their reaction products by high performance liquid chromatography
(HPLC). Double base propellants with 2-Nitro-DPA as a stabilizer were artificially aged at
60C, 70C and 80C and the various nitro derivatives are plotted against the time so that we
can see the concentration change of the reaction products up to the formation of
pentanitrodiphenylamine and picric acid. About 30 well resolved components can be seen
with HPLC under isocratic conditions. The powders tested had DPA or 2nitrodiphenylamine or ethylcentralite as stabilizers.
In S3 / 86 and S3 / 98, Tranchant, Maget and Lucotte recommended a method for the
determination of the life time by storing the propellants at different temperatures and storage
times and measuring the stabilizer decrease down to a distinct limit.
In S3 / 112, the influence of different burning modifiers on the chemical stability of
double base propellants has been investigated.
In a study conducted in the ICT (S4 / 29), it was shown that it is possible to get a good
insight into the ageing behavior of propellants by analyzing the stabilizer reaction products.
The ball propellant K503 contains 1% of DPA and is surface coated with 5% ethylcentralite.
The content of nitroglycerine is 9%. If we compare the time to the begin of the autocatalytic
decomposition (NO2 - formation) of K503 with the consumption of diphenylamine (DPA),
we get a time factor of more than twenty if we take into account the same kinetic conditions
for both reactions. The following table may explain this behavior:
Temperature (C)
Autocatalysis by
weight loss (days)
DPA consumption
(days)
Factor
90
32
1.3
24.61
85
62.4
80
126.8
5.5
75
263.0
70
516.6 (interpol.)
25.9
65
1061.8
60
2307 (extrapol.)
100
23.07
50
11306 (extrapol.)
420
26.92
E (kJ/mole)
142.2
140.9
Lg (B [sec-1])
13.8
15.8
105
23.05
20.66
For example: at 80C the weight loss autocatalysis starts after 126 days and the DPA
content has been consumed after 5.5 days. In this case, the time factor is about 23. With
regard to the activation energies of both different reactions, we see a quite good agreement:
142.2 kJ/mole compared with 140.9 kJ/mole. A similar behavior is to be seen also for the
other temperatures between 50C and 90C. This means that both reactions exhibit the same
decomposition mechanism.
The same holds for the single base propellant MRP2a containing also DPA as a
stabilizer: In this case we compare the begin of autocatalysis (NO2 formation) of the
weight loss investigation of MRP2a with the 50% stabilizer decrease as well as with the
100% stabilizer consumption. The three straight lines, which are more or less parallel,
indicate that it is allowed to establish a temperature correlation between weight loss and
diphenylamine consumption.
We investigated also the double base propellant MRP6a, containing DPA and found a
very good correlation of the 50% stabilizer decrease between 95C and 50C. In this case we
are also sure that we have no change of the reaction mechanism down to 50C.
106
In S4 / 29 we find also some plots with the consumption of ethylcentralite, the

autocatalysis and the formation of N-Nitroso-2-Nitro-Ethylaniline.
In another study, we investigated propellants containing the same single base and double
base composition with different stabilizers:
a) without stabilizer, but with dibutylphthalate instead of the stabilizer.
b) with diphenylamine
c) with ethylcentralite
d) with acardite II
The storage was conducted in glass tubes with glass stoppers loosely put on the tubes, at
different temperatures between 90C and 65C up to the formation of brown fumes
(autocatalytic decomposition by weight analysis). It was found that in all cases, for single
base as well as for double base propellants, Akardite II was the best stabilizer. The second
best was diphenylamine followed by ethylcentralite. In all cases, the activation energies
exhibited values between about 150 and 130 kJ/mole. Only the propellants without
stabilizers have shown much lower energies: 103 kJ/mole for the single base propellant and
117 kJ/mole for the double base propellant, see the following two tables:
107
Single base propellant: Time until autocatalysis in days

Temperature
MRP1a
MRP2a
MRP3a
MRP4a
(C)
Without stabilizer
With DPA
With EC
With acardite II
90
10
66
54
123
85
17
153
115
261
80
28
296
209
483
75
46
575
361
931
70
78
1089
686
1790
65
133
1185
E(kJ/mole)
103.1
151.5
130.5
138. 2
Lg B
6.9
13.0
10.1
10.8
Double base propellant

50%
decrease
of DPA
Time until autocatalysis in days

Temperature
MRP5a
MRP6a
MRP7a
MRP8a
MRP6a
(C)
Without
With DPA
With EC
With
With DPA
stabilizer
acardite II
90
31
33
76
0.33
85
70
65
166
0.58
80
10
139
130
328
1.54
75
20
277
232
693
2.45
70
32
494
430
1359
4.54
65
52
1117
729
2970
8.72
60
22.03
E(kJ/mole)
117.5
139.3
136.4
146.2
129.5
Lg B
9.5
11.5
11.2
12.2
13.9
108
The stabilizer decrease of these propellants described containing DPA and EC was
measured by thin layer chromatography and is depicted in S4 / 29. In this connection we
have to mention that acardite II produces the same reaction products as diphenylamine.
In S4 / 99, a detailed description of the HPLC analysis of diphenylamine and
ethylcentralite together with their reaction products is given by M. Lebert, M. Stephan and
B. Zeller. In addition, the paper of Caire-Maurisier, Tranchant and Kucera (S4 / 265) should
be mentioned which is dealing with the mechanism of the thermal decomposition of
nitroglycerine.
L. Druet et al. studied the reaction products of ethylcentralite of a mortar propellant
using two different chromatographic techniques (S7 / 233). In S7 / 285, S8 / 201, S10 / 305
and S12 / 267 , the blasting oil migration in propellant grains, which is important for the
interior ballistics behavior of propellants, has been investigated by Sopranetti, Brnnimann,
Stalder and Vogelsanger.
The decrease of acardite II of two different double base propellants in connection with
the NATO criteria of the 65.5C surveillance test 4117 is reported in S8 / 149.
Curtis and Kempson (S8 / 179) have also been involved in surveillance and sentencing
procedures of propellants using HPLC as the most versatile method of stabilizer analysis.
5.
GAS EVOLUTION
In order to understand better the reaction behavior of propellants containing energetic

nitrate esters, we should know more about the gases evolved during a thermal decomposition
process.
Blay (S3 / 61) gave a good overview entitled with Analysis of gases evolved from
simple propellant type composition during storage at 32C and 50C.
We know that the function of propellant stabilizers is to react with the oxides of nitrogen
or acids which are responsible for the autocatalysis.
Using the modified Will test, nitroglycerine, diethyleneglycol dinitrate, nitrocellulose
and different propellants have been heated to temperatures between 60C and 95 C in order
to measure the amount of NO2 formed during the thermal decomposition. The data for
nitroglycerine (NG) and nitrocellulose (NC) give Arrhenius type plots with activation
energies of 178 kJ/mole for NG and 162 kJ/mole for NC.
The gas evolution of propellants was determined by heating the sample under vacuum at
50 C. The evolved gases were analyzed by gas chromatography (GC). The principal gases
were nitric oxide (NO) and carbon dioxide (CO2), carbon monoxide (CO), nitrogen (N2) and
nitrous oxide (N2O).
In the Fraunhofer Institute (S5 / 479) double base propellants in the shape of cubes (5 *
5 * 5 cm) with a mass of each about 200 g were stored in glass vessels of 1,5 l, which have
been closed with metal caps of aluminium, containing a small needle valve, so that the air
after closing the vessel could be replaced by an inert gas (argon). In order to compare the
influence of air on the ageing procedure, the same propellants were stored in air. As an
example, the results of an ageing at 80 C will be presented. The storage has been conducted
in specially constructed ovens which were heated by hot water of 80 C. For the analysis, the
vessel was connected with a mass spectrometer or a gas chromatograph for analyzing of H2,
O2, N2, NO, N2O, CO and CO2. The following results of an ageing test at 80 C after 14 and
109
28 days as well in argon as in air were found. The double base propellant was stabilized with
2-Nitro-DPA.
Double base propellant D714

Gaseous reaction products in vol-%
Storage in argon (14 days)
Storage in air (14 days)
H2
0.3
0.1
O2
1.9
N2
26.1
80.4
NO
16.5
N2O
8.2
1.2
CO
0.7
0.5
CO2
48.2
15.9
Storage in argon (28 days)
Storage in air (28 days)
H2
0.4
0.2
O2
N2
25.0
59.1
NO
12.0
3.3
N2O
8.4
5.2
CO
0.8
0.8
CO2
53.2
31.6
Trinitro-DPA
at the outside
of the propellant
No
Yes
We see gaseous reaction products consisting of H2, N2, NO, N2O, CO and CO2 in
addition to water (not analyzed) in the case of 14 days ageing in argon atmosphere. The
result of a longer study of 28 days was not very much different from the 14 days ageing
process, only that the concentration of CO2 was a little higher. The amount of N2O of about
8 vol-% is quite high. On the other hand, we see a dramatically change of oxygen during
ageing in air. At the begin of the ageing process we started with 21 vol-% O2 and 78 vol-%
N2. After 14 days, most of the O2 content has been consumed by the reaction
NO + O2 NO2
The reason that we did not analyze NO2 was that this product reacts immediately after
production with the propellant stabilizer:
NO2 + stabilizer (2-Nitro-DPA) nitro derivatives of this stabilizer
110
Because of these reactions described, after a storage time of 14 days most of the oxygen
was consumed by reactions, which are exothermal: see the high content of CO2.
In addition we should mention that the ageing of propellants in air is more severe than in
argon atmosphere. By analyzing the surface of the propellant cubes after opening the glass
vessel (see figure 1) we found trinitrodiphenylamine in the propellant aged in air, but only
dinitrodiphenylamine as the highest nitrated stabilizer reaction product in the propellant aged
in argon atmosphere.
Using this information, we are possibly able to explain the heat

calorimetric measurements during the first time of the investigation of
The initial heat flow is rather high because of the reaction of air with
gases of the propellant, especially with NO. If all oxygen is consumed,
attain a constant level.
out-put of microa fresh propellant.

the decomposition
the heat flow may
A very good confirmation of our results that decomposition of nitrate esters is influenced
by oxygen during ageing in air, is published in S11 / 205 and in Propellants, Explosives,
Pyrotechnics (PEP) 14 (1989), 89-92, where Kimura studied the effect of autoxidation on the
ageing of explosives and gun propellants by chemiluminescence techniques. Other papers by
Wilker and Petrzilek et al. mention the same influence of oxygen: S11 / 383 and S12 / 221.
111
Another interesting reaction between diphenylamine, nitric oxides and nitrocellulose is

reported in a recent publication by T. Lindblom in PEP 27 (2002), 197, where he described a
green reaction product, which is not extractable.
6.
HEAT EVOLUTION
During the storage of energetic materials such as nitrate esters, thermal decomposition
occurs by exothermal reactions, which can lead to a disastrous self ignition caused by self
heating of propellant, especially stored in bulk.
With the use of so called heat generation meters, heat flow calorimeters or microcalorimeters we are able to forecast the dangerous behavior of propellants and explosives.
Jan Hansson studied some materials that have self-ignited in S7 / 87. In this connection we
should thank him very much for his successful effort in the development of new and better
instruments for measuring the heat evolution of explosives.
Van Geel, TNO, The Netherlands, presented one of the first isothermal heat-generation
meter in 1967 (S1 / 105). He proposed a method for the assessment of the chemical stability
of propellants by measuring the heat generation for 6 days at 85C. This should correspond
to 10 years of storage at 30C (ICT-Conference in 1971). Later he developed a method for
the estimation of safe diameters for the storage of gun propellants (S4 / 299).
An interesting introduction into the field of microcalorimetry was presented by Ingemar
Wads, Lund University, in S5 / 93. He discussed the different methods of heat evolution
measurements.
Max Frey, who developed also a heat flow calorimeter using a dewar vessel which was
integrated into a metal block oven, determined the safe life of different gun propellants. In
addition, he used his apparatus to conduct compatibility experiments and to calculate the
critical diameter for self ignition of TNT (S5 / 345). He also measured the heat evolution of
double base and triple base propellants. In the low temperature range (40 C) he found a
break of the Arrhenius plot. This means that the decomposition mechanism can change at
low temperatures (S8 / 15).
The influence of humidity and other parameters on the heat evolution has been
investigated by M. Rat (S5 / 361). It was found that the humidity increases also the
consumption of the stabilizer diphenylamine, see also S5 / 381. Dreyfus and Leveque
modelled the self-ignition of different ammunition calibres of propellants using Arrhenius
parameters determined by heat flow calorimetry (S7 / 99).
A comparative study of the ageing behavior of double base propellants has been
performed by Volk (S8 / 91) using a TAM apparatus made at that time by company LKB,
Sweden, today Thermometric AB, Sweden. He compared the stabilizer decrease of the
double base propellant RLC 470/33 with the stabilizer 2-Nitro-DPA with the heat evolution
during a storage at 80C. The following results were obtained.
112
Propellant RLC 470/33 with 1.56% 2-Nitro-DPA at 80C.

Days
0
Stabilizer
(%)
1.56
Heat gerneration rate

(W/g)
29.4
0.73
29.5
16
0.25
33.6
24
0.05
40.8
32
0.01
40.1
We see that the decrease of the stabilizer content from 1.56% 2-Nitro-DPA down to
0.01% did only lead to a small change in the heat generation rate from 29.4 to 40.1 W/g.
This means that the analysis of the stabilizer in this case is a better method for controlling
the ageing behavior than microcalorimetry.
On the other hand, Lindblom, Paulssen and Svensson have shown how useful
microcalorimetry can be in connection with the determination of the oxidation behavior of
N-NO-Diphenylamine (S9 / 157). This substance is in so far very much of interest, because
it is the first reaction product of diphenylamine during ageing, by reaction with NO2. But NNO-DPA is not a good stabilizer by itself; therefore it needs to be converted to 2-NO2 -DPA
and 4-NO2 -DPA, before the stabilizing effect can start. In order to get more information, the
reaction behavior of N-NO-DPA together with air, oxygen and nitrogen was studied in a
TAM microcalorimeter. With N-NO-DPA and air in the ampoule at 70C, a steady decrease
in the heat flow curve from 250 W/g to zero after 2 days was obtained, whereas in nitrogen
atmosphere a curve close to zero was found. In another try with N-NO-DPA mixed with
nitrocellulose, a heat flow curve, starting with > 100 W/g decreased but not until 8 days at
70C it was below 5 W/g. After this run the N-NO-DPA content decreased very much
according to HPLC analyses. The energy release was quite high: more than 290 kJ/mole O2.
From these results it seems reasonable to expect that one more mole N-NO-DPA will
react with one mole O2 to give NO2 - DPA:
N-NO-DPA + O2 2-NO2 -DPA (or 4-NO2 -DPA)
After the test in the microcalorimeter, the samples have been tested by HPLC and FTIR: 2-nitro-DPA and 4-nitro-DPA were the main products found after the reaction.
By comparing the enthalpy of formation of N-NO-DPA with that of the reaction product
2-NO2-DPA (see ICT-Database of Thermochemical Data, 7th update 2004), a high reaction
enthalpy is found:
Enthalpy of formation
N-NO-DPA
197.06 kJ/mole
2-NO2 DPA
53.14 kJ/mole
Reaction enthalpy:
243.9 kJ/mole
113
In this way it is easy to explain the high value of reaction enthalpy.
7.
MOLECULAR WEIGHT DECREASE
In his paper on Problems connected with the stability of propellants, B.V. Adsersen
(S1 / 206) recommended in 1967 the measurement of the viscosity of solutions of
nitrocellulose. He knew that it is a well known fact that when Nitrocellulose (NC) degrades,
not only NOX and other degradation products are released, but also the viscosity of the
solution of NC decreases. It was found that the decrease in viscosity is greater by storage of
the propellant in an atmosphere containing oxygen than by storage in an oxygen-free
atmosphere such as CO2 or N2. The reason for this behavior is that NO, which is always in
the decomposition products, will be oxidized in air to NO2 and this product reacts
immediately with nitrocellulose and the decomposition products.
In S6 / 47 and S7 / 197, gel permeation chromatography was used for the
characterization of the molecular weight degradation of nitrocellulose during ageing of
propellants between 60C and 90C.
In S3/125, the degradation of polyvinylnitrate (PVN) was measured by gel permeation
chromatography (GPC) and osmometry in order to obtain the weight average (MW) and the
number average (MN) molecular weight. Ageing tests which have been conducted at 90C
for several times have shown how much the molecular mass has been decreased. In addition,
the depletion of the stabilizer 2-nitro-DPA was evaluated by Poulin et al..
In S7/ 247, HPLC, GPC and microcalorimetry have been applied by Lindblom et al. to
characterize a single base propellant stored at 50C, 65C and 80C. Also ballistic
measurements were conducted. The aim of the investigation was to find out if it is possible
to replace the old test procedures, conducted at higher temperatures, by modern analytical
methods, conducted at a lower temperature range. In order to get rid of the initial relaxation
and equilibration processes, which we have always to take into account with
microcalorimetric measurements, closed glass ampoules have been used.
From these measurements it can be seen that all methods used give a rather high
acceleration factor of about 4.0 / 10C. It is also interesting that the reaction order (n) is not
1 as expected, but lies between 0.6 and 0.35. n varies linearly with the temperature:
n = 0 at ~ 120C
n = 1 at ~ 0C.
In an extensive paper, M.A. Bohn (S11 / 89) discussed the use of kinetic equations to
evaluate the ageing behavior of energetic materials. Several models are discussed to describe
the stabilizer reactions, the chain splitting and their consideration in modelling. The effective
stabilizer concentration can be used to predict an extended safe service life time. Mechanical
strength properties of a propellant were discussed and a model established for molecular
mass degradation of nitrocellulose. A close relationship between stabilizer decrease and
molecular mass degradation was found.
114
8.
DISCUSSION
Although we dont understand completely the mechanism of the thermal decomposition

of nitrocellulose, we have learnt a lot during our ageing studies in the last 37 years:
We know that oxygen plays an important role in the decomposition of propellants based
on nitrocellulose. It has three major impacts:
It oxidizes the nitrocellulose,
it oxidizes the stabilizer and some reaction products, and
it reacts with nitric oxide (NO) forming the highly reactive product NO2, which
reacts very fast with the stabilizer as well as with nitrocellulose and its
decomposition products.
In this connection, oxygen is also of influence on the storage behavior of explosives, e.g.
for weight loss measurement and heat flux measurement in the microcalorimetry. So we
know that the effect of preconditioning during heat flux measurements is very much
dependent on the oxygen content of the ampoules used for propellant heating in a
microcalorimeter.
Moreover, we have seen that there is a correlation between autocatalysis by weight loss
measurements, stabilizer consumption and molecular weight degradation; and when using
the 50% stabilizer depletion (measured by HPLC), it is possible to forecast the ageing
behavior of propellants in a much shorter time than with the use of conventional weight loss
tests. The same holds for the molecular mass decrease measured by gel permeation
chromatography (GPC). By measuring the chain break down of the nitrocellulose polymer,
we are able to get a better insight into the mechanical properties of propellant grains, which
is very important in the case of pressure build up during the ignition process.
With regard to the best stability test, we should consider the following facts:
-
If we have energetic materials without stabilizers, as it was the case with the film
material consisting of celluloid (S7 / 87), the best stability testing method is
microcalorimetry.
For nitrate ester propellants containing stabilizers, both methods are
recommended, measuring the stabilizer depletion by HPLC or heat flow
measurements using a microcalorimeter. The last method should be improved by
excluding oxygen from the ampoule, either by using an inert gas or by the
increase of the loading density (S11 / 383).
Finally, we have to mention the chemiluminescence techniques in connection
with the study of the effect of autoxidation on the ageing of explosives and nitrate
esters, and, in addition, the study of NOX evolution during the ageing of
propellants (S5 / 221).
115
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
S1 / 105 J.L.C. van Geel: National Defence Research Organization TNO, The Netherlands
The isothermal heat generation meter, Stockholm, 1967
S1 / 143 . Jacobsson: Research Institute of National Defence, Stockholm, Sweden Thin
layer chromatography of diphenylamine and centralite stabilized powders,
Stockholm, 1967
S1 / 162 A. Alm: Research Institute of National Defence, Stockholm, Sweden Studies on
reactions between nitrogen oxides and diphenylamine compounds, Stockholm, 1967
S1 / 179 D. Cronwall and B. Hildeby: Royal Swedish Naval Administration, Stockholm The
use of thin layer chromatography in connection with protective technical supervision
of powders, Stockholm, 1967
S1 / 188 E. Ripper: CTI Heimerzheim, Germany Dnnschichtchromatographische
Untersuchung von Pulvern, Stockholm, 1967
S1 / 206 B.V. Adsersen: Royal Danish Navy, Copenhagen, Denmark Problems with the
stability of propellants, Stockholm, 1967
S2 / 1
J. Hansson: Research Institute of National Defence, Stockholm, Sweden On
surveillance of propellants Tyringe, 1970
S2 / 34 A. Alm: Research Institute of National Defence, Stockholm, Sweden Mass spectra
of stabilizers, Tyringe, 1970
S3 / 61 N.J. Blay: Explosive Research and Development Establishment, Waltham Abbey,
Essex, England Some relationships between the stability of nitrate esters and the
stability of propellants, Ystad, 1973
S3 / 86 J. Tranchant: SNPE, St. Mdard-en-Jalles, France Stabilit chimique evaluation de
la dure de vie des poudres propulsives,Ystad, 1973
S3 / 98 J. Mayet and J.P. Lucotte: SNPE, Centre de Recherches du Bouchet, Vert-le-Petit,
France Etude cinetique de la decomposition des poudres pour armes, Ystad, 1973
S3 / 112 F. Volk: Fraunhofer Institut fr chemische Technologie, Pfinztal, Germany Der
Einflu von Abbrandmodifikatoren auf die chemische Stabilitt von DoublebaseFesttreibstoffen,Ystad, 1973
S3 / 125 G. Poulin, C. Michaud, S. Poulard: Commissariat lEnergie Atomique (CEA),
Sevran, France Etudes par chromatographie sur gel et sur couche mince de la
dgradation du nitrate de polyvinyle lors de vieillissement acceleres, Ystad, 1973
S4 / 9
R. Ammann: Gruppe fr Rstungsdienste, Thun, Schweiz Is there a critical
composition of reaction products of DPA stabilized nitrocellulose propellant during
ageing,Mlle, 1976
S4 / 29
F. Volk: Fraunhofer Institut fr chemische Technologie, Pfinztal, Germany
Dnnschichtchromatographische
Ermittlung
der
Lebensdauer
von
Treibladungspulvern, Mlle, 1976
S4 / 55 Mlle, 1976 A.J.W. Brook: MQAD, Woolwich, London, England Quality assurance
aspects of stability
S4 / 73 A. Sopranetti and H.U. Reich: Swiss Federal Propellant Plant, Wimmis, Switzerland
Simultaneous analysis of stabilizers and their decomposition products by
gaschromatography application of new methods for the judging of propellants by
the 65,5 C surveillance test,Mlle, 1976
S4 / 99 M. Lebert, M. Stephan, B. Zeller: SNPE, Centre de Recherches du Bouchet, Vert-lePetit, France Dosage de la diphenylamine, de la centralite et de leur derivs dans les
poudres et propergols par HPLC, Mlle, 1976
S4 / 249 M. Caire-Maurisier and J. Tranchant: SNPE, St. Mdard-en-Jalles, France
Mechanisme de la decomposition thermique de la nitroglycerine, Mlle, 1976
116
[20] S4 / 265 V. Kucera: Research Institute of Industrial Chemistry, Pardubice, CSSR Einflu der
Zersetzung von Nitroglycerin bei hheren Temperaturen auf den Verlauf der
Stabilittsprfungen von rauchlosen Pulvern, Mlle, 1976
[21] S4 / 299 J.L.C. van Geel, J. Verhoeff: Technological Laboratory TNO, Rijswijk, The
Netherlands Heat generation measurements for the stability control of nitrate ester
propellants, Mlle, 1976
[22] S5 / 93 J. Wads: University Lund Microcalorimetry,Bstad, 1979
[23] S5 / 107 T. Lindblom: AB Bofors, Bofors, Sweden Determination of stabilizer and its
nitroderivatives in double base powders, Bstad, 1979
[24] S5 / 127 R. Ammann, H. Hilty, H.R. Pfeiffer, W. Rauber: Armament Technology and
Procurement Group, Thun, Switzerland Methods to evaluate the chemical stability
of propellants using thin-layer and liquid chromatography, Bstad, 1979
[25] S5 / 221 J.G. Wallace, R.J. Powell, B.J. Downes: Propellants, Explosives and Rocket Motor
Establishment, Waltham Abbey, Essex, England A chemiluminescence study of
oxides of nitrogen evolved from propellants, Bstad, 1979
[26] S5 / 345 M. Frey: BICT, Swisttal-Heimerzheim, Germany Stabilittsuntersuchungen an
Treibladungspulvern und Sprengstoffen mittels Wrmeflusskalorimetrie,Bstad,
1979
[27] S5 / 361 M. Rat: SNPE, Centre de Recherches du Bouchet, Vert-le-Petit, France Application
de la microcalorimetrie isotherme letude de la stabilit des poudres pour armes,
Bstad, 1979
[28] S5 / 381 M. Dreyfus, M. Leveque: E.T.P.S./ D.T.A.T., France Puissances calorifiques
degages par les poudres homogenes: Resultat de mesures effectuees sur divers lots
de propergols neufs et anciens par microcalorimetrie isotherme, Bstad, 1979
[29] S5 / 479 F. Volk: Fraunhofer Institut fr chemische Technologie, Pfinztal, Germany
Alterungsverhalten von Doublebase-Festtreibstoffen und Treibladungspulvern,
Bstad, 1979
[30] S6 / 1
J. Tranchant: SNPE, Centre de Recherches du Bouchet, Vert-le-Petit, FranceInternal
mechanism of the chemical evolution of nitrocellulose propellants: Hypotheses and
consequences,Kunglv, 1982
[31] S6 / 47 F. Volk, G. Wunsch: Fraunhofer Institut fr chemische Technologie, Pfinztal,
Germany Anwendung der Gel-Permeations-Chromatographie zur Charakterisierung
des Nitrocelluloseabbaus in Treibladungspulvern, Kunglv, 1982
[32] S6 / 239 R.N. Palmaeus: Defence Material Administration, Stockholm, Sweden Acceptance
and surveillance inspection of propellants in Sweden,Kunglv, 1982
[33] S7 / 87 J. Hansson: Swedish National Defence Research Institute, Sweden
Microcalorimetric measurement on substances having self-ignited, Smygehamn,
1985
[34] S7 / 99 M. Dreyfus, M. Leveque: Etablissement Technique de Bourges, France
Modelisation du phnomne dauto-inflammation des propergols et examples
dapplication, Smygehamn, 1985
[35] S7 / 197 F. Volk, K.M. Bucerius, G. Wunsch: Fraunhofer Institut fr chemische Technologie,
Pfinztal, Germany Ermittlung von Einflugren auf die Genauigkeit von GPCErgebnissen, Smygehamn, 1985
[36] S7 / 233 L. Druet, M. Asselin, A. Bolvari: Defence Research Etablissement Valcartier,
Canada Stability testing of gun and mortar propellants A series of aged
propellants stabilized by ethylcentralite was investigated with two chromatographic
techniques, Smygehamn, 1985
[37] S7 / 247 T. Lindblom, P. Lagerkvist, L.-G. Svensson: AB Bofors, Bofors, Sweden
Comparison and evaluation of modern analytical methods used for stability testing
of a single base propellant, Smygehamn, 1985
117
[38] S7 / 285 E. Brnnimann, A. Sopranetti, Ch. Stalder Eidgenssische Pulverfabrik Wimmis,

Schweiz A universal test procedure to predict the shelf life of propellants,
Smygehamn, 1985
[39] S8 / 15 M. Frey: Rheinbach, Germany Mikrokalorimetrische Untersuchungen an
Nitratestertreibstoffen, Strmstad, 1988
[40] S8 / 91 F. Volk: Fraunhofer Institut fr chemische Technologie, Pfinztal, Germany Ageing
behavior of solid rocket propellants by measuring the heat evolution and the
stabilizer decrease, Strmstad, 1988
[41] S8 / 149 A.M. Fraser: DQA / TS, Bishopton, Scotland Assessment of the stability of
propellants stabilized with acardite II, Strmstad, 1988
[42] S8 / 201 E. Brnnimann, Ch. Stalder, A. Sopranetti Swiss Federal Propellant Plant, Wimmis,
Switzerland A new safety concept for development, processing and surveillance
of propellants, Strmstad, 1988
[43] S9 / 119 B.A. Lurie, B.S. Svetlov, A.N. Chernyshov: Mendeleev Institute of Chemical
Technology, Moscow, Russia Primary process of the nitrate esters thermal
decomposition, Margretetorp, 1992
[44] S9 / 157 T. Lindblom, L.-E. Paulsson, L.-G. Svensson: FFV Materialteknik, Karlskoga AB,
Sweden Oxidation of N-NO-Diphenylamine, Margretetorp, 1992
[45] S10 / 305 B. Vogelsanger, B. Ossola, E. Brnnimann: Swiss Munition Enterprise, Thun,
Switzerland The diffusion of deterrents and blasting oils into propellants, observed
by FTIR-Microspectroscopy,Margretetorp, 1995
[46] S11 / 89 Hemmeslvs Herrgrd, M.A. Bohn: Fraunhofer Institut fr chemische Technologie,
Pfinztal, Germany The use of kinetic equations to evaluate the ageing behavior
of energetic materials possible problems, Bstad, 1998
[47] S11 / 175J.M. Bellerby and S.J. Elsby: Department of Environmental Ordnance System,
Cranfield University, Shrivenham, U.K. Direct monitoring of nitrocellulose
degradation using visible spectroscopy, Hemmeslvs Herrgrd, Bstad, 1998
[48] S11 / 205Jun-ichi Kimura: Technical Research and Development Institute, Japan Defence
Agency, Tokyo, Japan Application of chemiluminescence to mechanistic and
kinetic studies of explosives - nitrocellulose and diethyleneglycol dinitrate (DEGN),
Hemmeslvs Herrgrd, Bstad, 1998
[49] S11 / 241P.G. Laye: School of Chemistry, University of Leeds, Leeds, U.K. Review of recent
thermal studies of pyrotechnics, Hemmeslvs Herrgrd, Bstad, 1998
[50] S11 / 383S. Wilker, U. Ticmanis, G. Pantel (WIWEB, Swisttal, Germany): P. Guillaume (B 4480 Engis, Belgium) Detailed investigation of sensitivity and reproducibility of
heat flow calorimetry, Hemmeslvs Herrgrd, Bstad, 1998
[51] S12 / 163 B. Lurie, B. Svetlov: Mendeleev University of Chemical Technology of Russia,
Moscow Chemistry and kinetics of nitrocellulose thermal decomposition,
Karlsborgs Fstning, Karlsborg, 2001
[52] S12 / 221J. Petrzilek, S. Wilker, J. Sklaldal, U. Ticmanis, G. Pantel, L. Stottmeister: WIWEB,
Swisttal, Germany Stability analyses of spherical propellants in dependence of their
stabilizer and nitroglycerine content, Karlsborgs Fstning, Karlsborg, 2001
[53] S12 / 267 B. Vogelsanger, R. Sopranetti, B. Ossola and K. Ryf: Nitrochemie Wimmis AG,
Wimmis, CH Compatibility and service life predictions of propulsion / ammunition
systems,Karlsborgs Fstning, Karlsborg, 2001
118
LIABILITY OF AMMONIUM NITRATE EMULSIONS

TO SPONTANEOUS DECOMPOSITION
Janusz Wrzesiski, Jadwiga Popawska-Jach, and Andrzej Koaczkowski
Wroclaw University of Technology, Wybrzeze Wyspianskiego 27, 50-372 Wroclaw, PL
Abstract:
Thermal properties of some ammonium nitrate emulsions were investigated by means of
a heat flow calorimeter SETARAM C80D. Kinetic parameters of ammonium nitrate
emulsions like activation energy, frequency factor were determined. Based on these
data, applying Frank-Kamenetskii's model, the critical radius and temperature for a
sphere were estimated.
Keywords:
ammonium nitrate, safety hazards, accidents
1. INTRODUCTION
In spite of the great progress in investigations on circumstances leading to uncontrolled
exothermal decomposition of ammonium nitrate, still we encounter disaster of immense
extent like that one in Toulouse (Grande Paroisse, 21 September 2001, France) or some
years earlier in Port Neal (Terra Industries, 13 December 1994, Iowa, USA). The analysis of
accidents involving ammonium nitrate shows that all of them with only few exceptions were
caused by its spontaneous exothermal decomposition. Under conditions enabling heat
accumulation, an uncontrolled reaction may take place and lead to temperature increase, fire,
explosion or even to detonation. This phenomenon is not limited only to the ammonium
nitrate in solid phase or melt but happens also in its aqueous solutions and emulsions.
The aim of the paper is to discuss some aspects of ammonium nitrate decomposition in
aqueous solutions in connection with two accidents and increasing production of ammonium
nitrate emulsions.
2.
HEAT BALANCE OF AMMONIUM NITRATE DECOMPOSITION

IN AQUEOUS SOLUTIONS
The heat balance of ammonium nitrate decomposition means that the whole heat
generated in the system is dissipated into the surroundings. The heat balance of unconfined
ammonium nitrate decomposition is hold up to a particular extent even at limited heat
dissipation into the surroundings. The balance is influenced generally by the evaporation of
water generated in the process and partly also by endothermic chemical processes.
Such balance takes place also in the process of ammonium nitrate decomposition in
aqueous solutions. The boiling temperature and ammonium nitrate concentration in the
steady state of decomposition were calculated for different reaction based on thermodynamic
data under following conditions:
119
a) The decomposition of ammonium nitrate may proceed according to the set of six
equations:
NH4NO3 = NH3 + HNO3
Ho298= +145 kJ
(1)
2NH4NO3 = 2N2 + O2 + 4H2O
H o298= 413 kJ
(2)
NH4NO3 = N2O + 2H2O
H o298= 124 kJ
(3)
2NH4NO3 = N2 + 2NO + 4H2O
H o298= 232 kJ
(4)
4NH4NO3 = 3N2 + 2NO2 + 8H2O
H o298= 758 kJ
(5)
5NH4NO3 = 4N2 + 2HNO3 + 9H2O
H o298= 1094 kJ
(6)
The reactions 2 to 6 are the only source of heat energy generated in the system under
consideration. The system is supplied with a boiling aqueous solution of ammonium
nitrate; the products of ammonium nitrate decomposition are carried off.
b) The heat generated in the ammonium nitrate decomposition is balanced entirely by
the evaporation heat (Qe) connected with the evaporation of water from the
ammonium nitrate aqueous solution. To simplify the calculation, Qe is assumed to be
constant in the temperature range under consideration and is equal to 2255 kJ/mole.
c) The next simplification is the assumption that heat losses (Ql), due to the convection,
transmission and radiation, compared to the heat generated in the very fast
proceeding ammonium nitrate decomposition are negligible and does not influence
essential the results of calculation.
Based on the results of calculation presented in table 1. the conclusion can be drawn that
solution containing more ammonium nitrate then about 60% by mass are liable to violent
decomposition. In case of more concentrated solutions, the amount of water that must
evaporate to compensate the heat generated during the chemical reaction is much higher than
it is available in the solution. Consequently, with the progress of reaction the concentration
of ammonium nitrate in the solution as well as its boiling temperature increases in the last
stage decomposes ammonium nitrate melt. The consecutive decomposition of ammonium
nitrate melt may lead to fire and explosion, especially in the presence of impurities, which
catalyse the decomposition of ammonium nitrate.
Under common conditions, generally encountered while producing and handling
ammonium nitrate bearing materials, the participation of water evaporation and endothermic
reactions can be negligible; e.g. in closed vessels. Then, the heat balance is determined by
the relation of heat dissipation into surroundings to the heat generation.
2.1.
Heat generated in the ammonium nitrate decomposition
Ammonium nitrate is capable of spontaneous exothermal decomposition from the

thermodynamics point of view. The composition of the products can vary vastly in quality
and quantity. The products usually include water (H2O), oxygen (O2), nitrogen (N2), nitrous
oxide (N2O), nitrogen dioxide (NO2), and nitrogen oxide (NO). The mass ratios of the
components in the product and the heat generated depend on the conditions of the
decomposition. Temperature, pressure, and, above all, catalysts have a great effect. The
reaction mechanisms are not fully understood, however the decomposition is usually
described by means of equations 1 to 6 summarizing the substrates and products.
In order to understand self-heating in ammonium nitrate, it is useful to understand the
influence of the primary factors stimulating the decomposition of ammonium nitrate in
particular, the effects of acid concentration and chlorides as well as reducing agents.
120
Table 1.
Results of calculation.
Boiling
temperature
NH4NO3
concentration
Standard
reaction heat (Qr) calculated for
% b. mass
kJ/kg aqueous solution of ammonium nitrate
(2)
(3)
(4)
(5)
(6)
Dissolving heat
(Qs)
Evaporation
heat
(Qe)
Qr + Qe - Qs calculated for
Energy for
heating
solution to
boiling **
(2)
(3)
(4)
(5)
(6)
10
11
12
13
14
15
101.0
10
-154.49
-52.84
-41.41
-133.55
-161.89
25.6
2029.66
1849.67
1951.22
1962.65
1850.71
1842.17
78.4 (25)
102.5
20
-308.88
-105.58
-82.70
-267.00
-323.93
46.4
1804.14
1448.86
1652.16
1675.04
1490.74
1433.81
80.8 (25
104.2
30
-463.61
-158.63
-124.35
-400.80
-485.89
64.4
1578.62
1050.61
1355.59
1389.87
1113.42
1028.33
82.5 (25)
106.6
40
-618.50
-211.90
-166.15
-534.80
-648.25
79.7
1353.11
664.64
1061.24
1106.99
738.34
624.89
83.8 (25)
109.3
50
-773.75
-265.50
-208.31
-669.13
-811.06
92.3
1127.59
261.54
759.54
826.98
366.16
224.23
84.3 (25)
113.1
60
-929.55
-319.73
-251.03
-804.08
-947.48
102.4
902.07
-129.88
479.94
548.64
-4.41
-147.81
84.6 (25)
119.3
70
-1086.49 -375.11
-294.79
-940.19
-1139.16 110.7
676.55
-520.64
190.74
271.06
-374.34
-573.31
51.0 (30)
128.9
80
-1274.90 -462.00
-370.10
-1107.80 -1335.60 73.1
451.04
-896.96
-84.06
7.84
-729.86
-957.65
35.1 (60)
144.1
90
-1459.58 -545.18
-441.56
-1271.81 -1528.76 75.1
225.52
-1309.16 -394.75
-291.14
-1121.39 -1378.34 20.0 (100)
* - for constant evaporation heat equal to 2255.19 kJ/kg H2O

** - initial temperature in brackets.
121
2.1.1. Nitric acid and chlorides

The stimulating effect of nitric acid and chlorides on ammonium nitrate decomposition
(i.e., increasing the thermal power) has been revealed in many publications. In particular, a
distinct reaction rate increase has been observed in the presence of chlorides with
concentrations as low as 10 ppm.
Generally, following conclusions concerning aqueous solutions of ammonium nitrate
(about 80% by mass) can be drawn:
Chlorides stimulate the decomposition of ammonium nitrate after an induction period

connected with the increase of acidity of solution.
The induction period is inversely depended to the temperature, the acidity of the solution
(the nitric acid concentration), and the chloride concentration.
The induction period for the chloride-catalyzed decomposition of ammonium nitrate

should be distinguished from the induction time of self-heating (the time after the conditions
for self-heating have been established to the time the temperature of the solution rises
spontaneously). The induction period plays a very important role in the decomposition of
ammonium nitrate.
2.1.2. Spontaneous acidification of ammonium nitrate melt and solution
The ammoniating of ammonium nitrate melt or solution may inhibit the dissociation of
ammonium nitrate. Nevertheless, a very slow, but highly exothermic, reaction takes place
simultaneously, even at relatively low temperatures (e.g., about 100 C, or at least below the
melting point of pure ammonium nitrate):
Ho298= -1094 kJ
5NH4NO3 = 4N2 + 2HNO3 + 9H2O
(6)
As a consequence of this reaction, acidification of stored ammonium nitrate melts or

solutions takes place. The acidification proceeds up to an equilibrium (or steady state) known
to be dependent on temperature, ammonium nitrate concentration and the presence of other
substances. Therefore, alkaline or neutral melts and solutions may become acidic under
normal conditions of storage if their pH is not maintained and the solution is not ammoniated.
Furthermore, if the ammonia added is not sufficient to ensure alkaline conditions in the liquid
phase, then reaction (6) proceeds continuously and generates heat. In spite of wide
represented opinion that the escape of ammonia is the main cause of acidification,
acidification takes place also in closed vessels. This phenomenon plays a very important role
in activating catalytic properties of chlorides.
2.1.3. Influence of reducing agents
Materials containing ammonium nitrate and purposely added reducing agents are
available as explosives. The increasing share of ammonium nitrate emulsions in production
and consumption of explosives enhances the safety hazards, especially the self-heating risk.
With the aim to estimate the safety hazard connected with ammonium nitrate emulsions,
investigation on their liability to self-heating were performed.
2.2.
Experimental
The influence of chloride ions and nitric acid on an ammonium nitrate solution
decomposition has been investigated by means of a SETARAM (France) C80 D heat flow
calorimeter. The measurements (scanning and isothermal) were performed in the temperature
range from ambient to 300 C in closed pressure vessels (pressure resistant up to 350 bar)
122
made of Hastelloy. The samples were held in an open glass tube placed in the vessel. The
pressure-sensitive transducer closed the top of the vessel.
Pure ammonium nitrate aqueous solutions (about 83 % by mass NH4NO3) and containing
admixture were applied for the experiments. A special series of ammonium nitrate emulsions
(matrixes) was investigated. Similar determinations were performed on samples of
ammonium nitrate emulsions manufactured in full-scale installation: m1, m2, m3, m4, m5, and
m6. The composition and structure of the ammonium nitrate emulsions showed great
diversity, for example: in ammonium nitrate content (about 80 do 98 % by mass), in the kind
of emulsifier or buffer applied. The mentioned samples were also modified by adding
ammonium chloride and investigated in the presence of materials widely used for the
construction of installation in the chemical industry titanium, aluminum, and stainless steel
304.
Ammonium nitrate aqueous solution
The results of investigation are shown in Figures 1. Calorimetric measurements have
shown that higher concentrations of nitric acid in ammonium nitrate solution increase the
decomposition rate of ammonium nitrate. From Figure 1 it follows that in the presence of
nitric acid, the heat flux generated during decomposition is greater than with pure ammonium
nitrate solution. The effect of nitric acid is enhanced in the presence of 304 stainless steel.
The results of these experiments have confirmed that the decomposition of ammonium
nitrate in acidic aqueous solution is strongly stimulated by chloride ions, considerably more
than by nitric acid alone. In the presence of 304 stainless steel, the effect of chlorides and
nitric acid on the decomposition rate of ammonium nitrate is the strongest.
500
450
TAR 207 AN
TAR 208 AN+NA
TAR 254 AN+NA+Cl
TAR 256 AN+NA+Cl+SS
TAR 260 AN+NA+SS
400
Heat flow [ W/kg NH4NO3]
350
300
250
200
150
100
50
0
-50
-100
120 130 140 150 160 170 180 190 200 210 220 230 240 250 260 270 280 290 300
Temperature [oC]
Fig 1.
The influence of additives on the heat flux generated in ammonium nitrate

decomposition versus temperature (heating rate 0.25 C/min)
123
Ammonium nitrate emulsions

Our investigation on the thermal properties of non-activated ammonium nitrate emulsions
shows that their behavior is similar to that one of ammonium nitrate aqueous solution. Selfheating constitute the main hazard in handling ammonium nitrate emulsions, because
generally they consists of high concentrated ammonium nitrate aqueous solution as oxidizer
and organic compounds as fuel that equalizes the oxygen balance.
Typical profiles of heat flux versus temperature resulting from the scanning
measurements are presented in figures 2 and 3 (heating rate 0.25 oC/min, the mass of the
samples about 250 mg). The pressure axis responds to the changes of pressure related to the
value of pressure at 30oC calibrated equal to 0 Pa. The heat flux axis responds to the heat flux
calculated for 1 kg of ammonium nitrate emulsion. In some analysis the heat flow overrun the
limit - it is evidenced by a pick with the top cut off (heat flow greater than about 750 W/kg
emulsion).
T e m p e ra tu re , C
233
23 5
2 37
23 9
24 1
2 43
5 10
51 2
51 4
5 16
24 5
2 47
2 49
25 1
2 53
255
25 7
2 59
51 8
5 20
5 22
52 4
5 26
528
53 0
5 32
8 00
7 50
m1
m2
m3
m4
m5
m6
7 00
6 50
6 00
5 50
5 00
Heat flux, W
4 50
4 00
3 50
3 00
2 50
2 00
1 50
1 00
50
0
-50
506
50 8
T e m p e ra tu re , K
Fig 2.
Results of calorimetric measurements of ammonium nitrate emulsions,

heat flux versus temperature
124
T e m p e ra tu re , C
233
235
237
239
241
243
245
247
249
251
253
255
257
259
510
512
514
516
518
520
522
524
526
528
530
532
19
18
m1
m2
m3
m4
m5
m6
17
16
15
14
13
Pressure, MPa
12
11
10
9
8
7
6
5
4
3
2
1
0
506
508
T e m p e ra tu re , K
Fig 3.
Results of calorimetric measurements of ammonium nitrate emulsions,

pressure versus temperature
Figures 4 to 7 displays the results of isothermal measurements. They show the changes of
heat flux and pressure respectively. The investigation proved the diversity of the thermal
behaviour of the emulsions under the conditions of experiments. There are noticeable
differences in the temperature of the initiation of decomposition, from about 506 K to above
532 K (Fig.2 and 3).
The results of scanning measurements were confirmed by the isothermal determinations.
As expected, the heat flux reached noticeable value in the isothermal measurements at much
lower temperatures compared with the temperature of the initiation of the decomposition
during scanning measurements. On figures 4 and 5 are presented the results of isothermal
measurements of sample m3. From these figures it follows that the sample is liable to intrinsic
decomposition at about 488 K after a induction period of about 9 hours. There are also
differences in the induction period in the presence of additives like: chlorides or metals
usually applied in the construction of chemical installations (figures 6 and 7). Under the
conditions of measurements sample m4 does not underwent a violent decomposition even
after about 90 000 s and in the presence of aluminium. However, under the same conditions
other additives caused violent decomposition.
125
400
473 K
488 K
498 K
350
300
Heat flux, W
250
200
150
100
50
0
-50
0
10000
20000
30000
40000
50000
60000
Time, s
Fig 4.
Results of calorimetric measurements of ammonium nitrate emulsion m3 ,

heat flux versus time
3.2
3.0
473 K
488 K
498 K
2.8
2.6
2.4
Pressure, MPa
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0
10000
20000
30000
40000
50000
60000
Time, s
Fig 5.
Results of calorimetric measurements of ammonium nitrate emulsion m3 ,

pressure versus time
126
900
850
m4
+Al
+C l
+SS
+Ti
+C l+Al
+C l+SS
+C l+Ti
800
750
700
650
Heat flux, W
600
550
500
450
400
350
300
250
200
150
100
50
0
-50
25000
30000
35000
40000
45000
50000
55000
60000
65000
70000
75000
80000
85000
90000
Tim e, s
Fig 6.
Results of calorimetric measurements of ammonium nitrate emulsion m4

in the presence of chlorides, metallic titanium, aluminium,
and stainless steel 304 (heat flux versus time)
13
m4
+SS
+Cl
+Cl+SS
12
11
10
+Al
+Ti
+Cl+Al
+Cl+Ti
Pressure, MPa
8
7
6
5
4
3
2
1
0
-1
25000 30000 35000 40000 45000 50000 55000 60000 65000 70000 75000 80000 85000 90000
Time, s
Fig 7.
Results of calorimetric measurements of ammonium nitrate emulsion m4

in the presence of chlorides, metallic titanium, aluminium,
and stainless steel 304 (pressure versus time)
127
2.3.
Critical parameters
From the loss prevention point of view it is very important to estimate the conditions for
self-heating of ammonium nitrate bearing materials. Based on Frank-Kamenetzkii's model an
attempt was made to determine the critical parameters of temperature related to the radius of a
sphere.
RTo2
E
EQA exp
RTo
r=
where:
r radius of the sphere; R - universal gas constant; To wall temperature;

- Frank-Kamenetskii critical parameter; - heat transfer coefficient;
E activation energy; Q heat of the reaction; density;
A frequency factor.
The frequency factor and acti-vation energy were calculated from
the calorimetric data.
The results of calculation are presented in figure 8. They should be considered only a
rough estimation resulting from the assumption made by Frank Kamenetskii as well as the
difficulties with the precise determination of the thermal effect. The actual value of the critical
parameters are expected to be lower.
5.5
m1
m5
m6
5.0
4.5
Critical radius r, m
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
320
340
360
380
400
420
440
460
480
500
520
Temperature T, K
Fig 8.
Comparison of the critical radius (sphere) versus critical temperature

for three ammonium nitrate samples
128
3. CONCLUSION
The main cause of accidents involving ammonium nitrate is its liability to spontaneous
exothermal decomposition. Limited mass and heat exchange with the surroundings are the
main factors enabling heat accumulating and in consequence run-away reaction, deflagrating
or even detonating.
The aim of investigation was to recognize the specific feathers of decomposition of
ammonium nitrate emulsion, intrinsic for effective loss prevention.
By means of the heat flow calorimeter SETARAM C80D such kinetic parameters like
activation energy, frequency factor were obtained. Based on these data, applying
the Frank-Kamenetskii's model the critical parameters for a sphere were estimated.
The values of critical radius and critical temperature are influenced by all the assumption
made by Frank Kamenetskii and the difficulties with the precise determining of the thermal
effect. Nevertheless, they remain in reasonable relation to the results of investigation into
causes of two accidents.
REFERENCES
(up-to-date!)
[1]
[2]
X.-R. LI and H. KOSEKI: Study on the contamination of chlorides in ammonium nitrate, Trans
ICemE, Process Safety and Environment Protection, No. 83(B1), p.31-37, 2005
ZHANHUI SUN, JINHUA SUN, HUI DING AND CHUANSHENG JIANG: Thermal Hazard
Evaluation of Ammonium Nitrate Mixtures, Proc. of 11th Int. Symp. Loss Prevention and Safety
Promotion in the Process Industries, Praha , p.2364-2371, 2004
129
COMPUTER SIMULATION
OF NITRO-1,2,4-TRIAZOLES DECOMPOSITION REACTIONS
Eugeniya A. Bakhmatova* , Tatyana V. Petukhova**, Vyacheslav L. Korolev*,
Tatyana S. Pivina*, and Victor P. Ivshin**
* Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences,

Leninsky Prospect 47, Moscow 119991, RUSSIA
** Mari State University,
Lenin square 1, Yoshkar-Ola 424000, Mari El Republic, RUSSIA
Abstract:
The combination of nitro-groups with polynitrogenous heterocycles and, in particular,
with 1,2,4-triazoles is often used for the synthesis of power-consuming compounds. For
these materials creation, an advance estimation of the basic characteristics is crucial to
select the most prospective structures for subsequent synthesis. One of the most
important characteristics of power-consuming compounds is the thermal stability
connected with the processes of thermal decomposition. Nevertheless, nowadays a
theoretical description of multi-channel process of decomposition represents a
complicated problem and has been mainly limited by the primary decomposition act
with simplified schemes of the process.
On the basis of Recombination Reaction Networks, a stringent approach to the
generation of a set of hypotheses for the mechanism of thermolysis has been elaborated.
In this work, our methodology has been used for computer simulation of decomposition
reactions of N-nitro and C-nitro-1,2,4-triazoles. Subsequent estimation of the thermochemical preference of these or other decomposition pathways has been estimated using
DFT B3LYP 6-31G* method.
The schemes obtained have been collated with the available experimental data.
Keywords:
1.
computer simulation, thermal decomposition,

N-nitro and C-nitro 1,2,4-triazoles
INTRODUCTION
The combination of polynitrogenous heterocycles and, in particular, with 1,2,4-triazoles

is used for the synthesis of power-consuming compounds for different applied purposes.
During these materials creation, an advance estimation of the basic characteristics is crucial
to select the most prospective structures for synthesis. One of the most important
characteristics of compounds is the thermal stability connected with the processes of thermal
decomposition. Nevertheless, nowadays a theoretical description of multi-stepped
thermolysis processes represents a complicated problem. That is why in a series of works
methodology of organic compounds thermolysis channels modeling has been proposed. In
the basis of this methodology has been representation of decomposition processes with
recombination reaction nets (RRN). In present work this approach has been used for N-nitro1,2,4-triazole (1) and C-nitro-1,2,4-triazole (2) thermolysis channels modeling.
130
N
N
NO2
NH
NO2
1
2.
N
2
METHODS OF INVESTIGATIONS
To generate pathways of thermolysis we created a base of experimental data of nitro1,2,4-triazoles decomposition reactions. On the basis of this data a generator of hypoteses for
the mechanisms of thermolysis of these compounds has been elaborated. To evaluate the
thermochemical preference of these or other pathways of decomposition, which has been got
by computer-based simulation, quantum-chemical calculations of the first steps activation
energies of decomposition were carried out. All calculations were accomplished by method
of density functional (DFT) using hybrid potential B3LYP /8/ and standard valence-split
basic 6-31G* /9/. All calculations were executed in the context of GAUSSIAN 98 /10/ in
Computing Centre of N.D.Zelinsky Institute of Organic Chemistry of Russian Academy of
Sciences.
3.
DISCUSSION OF THE RESULTS
Generation of 1-nitro-1,2,4-triazole (1) channels thermolysis allowed us to choose four

probable pathways of its thermal decomposition which has been caused by radical cleavage
of nitro-group (1), dissociation of N-N bond in heterocycle (2), 1,5 sigmatropical shift of
nitro-group with formation of 5-nitro-1,2,4-triazole (2c) (3) and nitro-nitrite rearrangement.
N
N
NO2
1
2
N
N
.
. ONO
N
N
N
N
ONO
O2N
NH
2c
NO2
Fig 1.
Computer simulation scheme of the first step

of 1-nitro-1,2,4-triazole (1) thermolysis.
On this and other figures final products corresponding to the experimental data are
detailed in a bold type.
For determination of the thermochemial preference of this or other thermolysis pathways
of compound 1 activation energies of the first step of decomposition have been calculated.
It has been found that for radical cleavage of nitro-group (Fig.1, channel 1) 39.4
kkal/mol is expended. Dissociation energy of exocycle N-N bond (channel 2) is 62.2
kkal/mol, activation energy of 1,5-sigmatropical shift of nitro-group (channel 3) is 31.8
131
kkal/mol. We havent managed to localize the transition state of corresponding nitro-nitrite

rearrangement.
According to the data /11/ for N-nitro derivatives this rearrangement is not favorable
thermochemically in comparison with the cleavage of nitro-group. The results of fulfilled
calculations show that 1.5-sigmatropic shift of nitro-group is 7.6 kkal/mol and more
favorable than its homolytic cleavage. That is why 1.5-sigmatropic shift will predominate in
thermolysis of 1.
According to /12/ C-nitro compound was prepared in not less than 54% yield by
heating of 1 in benzonitrile during 36 hours at 1200C. However radical cleavage of NO2group should be considered as a competitive process, since among the products of
thermolysis of 1 unsubstituted 1,2,4-triazole was not found.
Compound 2c formed by thermolysis of 1 (channel 3) is regioisomer of the latter.
Besides several tautomeric forms of C-nitro-1,2,4-triazole (2) are probable.
NO2
N
NH
NO2
NH
2
N
N
N
N
H
+N
-O
O
N+
O
O-
NO2
NH
H
+N
-O
O
2d-1 (E,E)
N
N
N+
O
OH
H
2d-2 (E,Z)
2e-1 (Z,E)
2e-2 (Z,Z)
According to our calculations tautomer 2c is the most thermodynamically stable one in

the gas phase (Table 1).
Table 1. Energy characteristics of tautomers 2a-e.
Compounds
Total Energy,
E, a.e.
2a
2b
2c
2d-1
2d-2
2e-1
2e-2
- 446.736095
- 446.727251
- 446.737404
-446.698206
-446.691123
-446.697340
-446.691393
Energy of zeropoint oscillations

correction,
ZPE, a.e.
0.062405
0.061738
0.062507
0.061071
0.060557
0.060977
0.060553
132
E + ZPE, a.e.
Tautomerization
energy 2,
kcal/mol
- 446.673689
- 446.665514
- 446.674897
-446.637135
-446.630566
-446.636363
-446.630484
0.8
5.9
22.9
27.8
24.2
27.9
The energy of tautomerization was calculated by formula [(E + ZPE) (E +

ZPE)]627.5 kcal/mol /13/, where ZPE is a correction for the energy of zero-point
oscillations. At the same time for N-unsubstituted 3(5)-nitro-1,2,4-triazoles in solutions Hatom is localized at N-atom distant to nitro-group. These data obtained from NMR /14/ and
dipole moment analysis are in accordance to the tautomer 2a. The existence of compound 2
in the form of this or another tautomer, probably, will depend on reaction conditions.
To get a clear description of C-nitro-1,2,4-triazole (2) thermolysis mechanisms we have
generated some channels of thermolysis for all probable tautomers (2a-e). The results of
computer modeling for 2c and 2e are outlined in Fig. 2 and Fig. 3. It turned out the
decomposition channels of tautomers 2a-c and 2d-e are similar. Thus, four probable ways of
decomposition for tautomers 2a-c are observed. They are homolytic cleavage of nitro-group
(channel 1), dissociation of N-N bond in heterocycle (channel 2), exocyclic C-N bond
dissociation (channel 3), and nitro-nitrite rearrangement (channel 4).
N
N
NO 2
2
N
NH
N.
. ONO
NH .
NO2
Fig 2.
N
N
NH
.
NH
ONO
NO 2
Scheme of modelling for thermolysis first step

of 1-H-5-nitro-1,2,4-triazole (2c).
Probable thermolysis pathways of thermodynamically most stable tautomers 2d-e are

homolytic cleavage of hydroxyl radical (channel 1), destruction of 1,2,4-triazole cycle with
throwing nitrogen (channel 2), endocyclic C-N bond dissociation (channel 3), and cleavage
of oxygen in aci-nitro-unit (channel 4).
N
N
N
N
HO
2
OH
+M
N
N
H2O
N
HO
N
N
N
O
Fig 3.
N2
HO
Scheme of modelling for thermolysis first step

of 3-aci-nitro-1,2,4-triazole (2e).
133
N
N
N
HO
O
+M
H2O
Data of activation energy calculations for the reactions proceeding on the first step of
2a-e compounds decomposition are outlined in Table 2.
Table 1. Activation energy of the reactions proceeding for the first step
of 2a-e tautomers decomposition.
Number of
the channel
1
2
3
4
2
67.2
71.3
105.6
65.2
2
68.2
66.7
90.4
63.5
2
67.4
67.8
93.3
60.1
, kcal/mol
2d-1
51.5
59.7
95.4
78.0
2d-2
48.2
55.4
91.1
75.0
2e-1
50.2
59.2
94.9
75.9
2e-2
46.4
55.4
91.1
73.7
As it follows from obtained results, tautomer 2e, for which homolytic N-O bond
cleavage with removal of hydroxyl radical is the most favorable process on the first step is
termochemically least stable from all thermolysis conditions.
Noteworthy, experimentally found activation energy of the first step of 3-nitro-1,2,4triazole thermolysis is 38.28 kcal/mol /16/ in solid phase at 180-200. At the same time,
activation energy for 1-ethyl-3-nitro-1,2,4-triazole incapable of formation of aci-nitro-form
due to the proton of NH group migration, is 65.06 kcal/mol /17/.
A clear description of probable thermolysis mechanisms is outlined in Fig.4, which
shows that formed 3-nitroso-1,2,4-triazole and 1,2,4-triazolone can be considered as
intermediates and as possible final products of thermolysis. According to the data /16/ massspectra of decomposition products contain molecular ions peaks of these compounds.
Probable formations of 1,2,4-triazalone is discussed in /18/. N2, N2O, NO, CO2, HCN,
HCNO, H2O, cyanamide, cyanuric acid and melamine are also the final products of
thermolysis. Besides the first six compounds from the series were experimentally registered
before.
For tautomers 2a-c thermochemically most favorable thermolysis pathway is nitro-nitrite
rearrangement which is theoretically considered as one of possible decomposition channels
of 2 in /18/. However, it is 13.7-18.8 kcal/mol less preferable then hydroxyl homolysis of
2e-2 (Fig.3, channel 1). Furthermore, this way of 2 decomposition excludes formation of Cnitroso-1,2,4-triazole, which was experimentally registered in products of thermolysis.
134
N
N
N
N
HO
ON
N
N
H2O
N
O
HN
CN
+M
HCN N C O O
NH
NH
O
N
+M
NH C O
HCN
O
HN
O NH
HN
N
. .
+M
H2N
N
NH
O
. .
Fig 4.
NH
NH
N
N
N
H2O
+M
HN
N
CN
H2N
HCN
H2N
N
+M
+R
H2N
NH2
OH
OH
N
H2O
NO
N H2O
+M
HNO
+NO HCN
HN
O NH
NH
+HONO
+M
CHO
+R
N2 OH
+ONO
CO CO2 + HNO
. .
H2O
+M
HCN
N O CH N N OH
O HCN
N2O+ OH
+M
CN
HCN
N2
CN HCN
CN
+M
HCN
NH C O
N C NO
NH
C N O
NH
N
O
HCN
+R
H2N
N2
+M
NO
NO
HO
N
N NO
NO
NH
HO
OH
+M
N
NH
HN
H2O
NH
+M
HO
N
O
N
+OH
NH C O
C
N
OH
HO
NH
NH
+M
N
N
N
O
HCN C N O
+M
+M
N2
N2
+M
H2O
3
2
OH
OH
+M
H2O
+NO2 -NO
CO2
NH2
. .
. .
Computer modeling scheme

of 3-aci-nitro-1,2,4-triazole (2e) decomposition.
From Table 2 and Fig.2 activation energies of homolytic NO2-group cleavage (channel
1), endocyclic N-N bond dissociation (channel 2) and nitro-nitrite rearrangement (channel
4) are close. So these three channels of thermolysis are equiprobable. Triazole cycle
destruction on the first step of thermolysis (channel 2) excludes existence of experimentally
registered 1,2,4-triazole derivatives among decomposition products, though the rest of
above-listed compounds are formed in the reaction.
For dissociation of heteroaromatic C-N bond in 2a-c by channel 3 (Fig.2) 22.2-40.4
kcal/mol more is expended then for process by channels 1,2,4. Consequently, possibility of
destruction by this pathway is not very high. Besides, as in the case of dissociation by
channel 2, the triazole cycle decomposition on the initial step excludes formation of
experimentally registered 1,2,4-triazole derivatives.
135
4.
CONCLUSION
Computer simulation schemes of triazoles thermolysis, which have been predicted, allow
to describe the whole variety of formally possible reactions during decomposition processes.
They also permit to mark out thermochemically most preferable ones. Obtained data can be
used for triazoles thermal stability prediction and also for the further thermochemical and
kinetic analysis of these compounds thermolysis process.
Acknowledgment:
The authors are grateful to Dr. A. Porollo for help in this work carrying-out.
136
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
M. S. PEVZNER: Derivatives of 1,2,4-triazole high-energy compounds. Russian chemical

journal, 41, No 2, p.73-83, 1997
H.H. LICHT, S. BRAUN, M. SCHAFER, B. WANDERS, and H. RITTER: Nitrotriazole:
Chemische Struktur und explosive Eigenschaften. Proceedings of the 29th International Annual
Conference of ICT (Energetic Materials), Karlsruhe, Germany, p. 47.1-15, 1998
A. M. ASTACHOV, V. A. REVENKO, and E.S. BUKA: Comparative characteristics of two
isomeric explosives: 4-nitro-5-nitrimino-1H-1,2,4-triazole and 3-nitro-5-nitrimino-1,4H-1,2,4triazole. Proceedings of the 7th Seminar "New Trends in Research of Energetic Materials",
Pardubice, the Czech Republic, p. 424-431, 2004
A.A. POROLLO, D.E. LUSHNIKOV, T.S. PIVINA, and V.P. IVSHIN: J. Mol. Structure
(THEOCHEM), vol. 391, p. 117-124, 1997
A.A. POROLLO, T.V. PETUKHOVA, V.P. IVSHIN, T.S. PIVINA, and D.E. LUSHNIKOV:
Azetidine and its mono-, di- and tri-nitro substituted derivatives: computer modeling of
decomposition reactions, Proceedings of the 30-th International Annual Conference of ICT,
Karlsruhe, Germany, p. 15 (1-3), 1999
T.S. PIVINA, A.A. POROLLO, T.V. PETUKHOVA, and V.P. IVSHIN: Basic scheme for
computer simulation of decomposition reactions for energetic compounds. Proceedings of the
5th International Symposium on Special Topics in Chemical Propulsion (5-ISICP), Italy, p.
343-344, 2000
T.V. PETUKHOVA, A.A. POROLLO, V.L. KOROLEV, V.P. IVSHIN, Y.N. SURIKOVA,
and T.S. PIVINA: Computer Modeling of Decomposition Reactions for Varies Representatives
of Nitroamines. Proceedings of the 33rd International Annual Conference of ICT, Karlsruhe,
Germany, p. 60 (1-10), 2002
W. KOCH, and M. C. HOLTHAUSEN: A Chemistss Guide to Density Functional Theory,
Wiley-VCH, Weinheim, p. 300, 2001
T. KLARK: Computed chemistry (in Russian), Mir, Moscow, 383 p., 1990
Program Complex Gaussian-98 of the Computer Chemical Investigations Center Russian
Academy of Sciences.
G.B. MANELIS, G.M. NAZIN, Yu.I. RUBTSOV, and V.A. STRUNIN: Thermal
decomposition and combustion of explosives and powders. Nauka, Moscow, 223 p., 1996
M.S. PEVZNER, T.N. KULIBABINA, N.A. POVAROVA and L.V. KILINA: Heterocycle
nitrocompounds. XXV, Nitration of-5-amino-1,2,4-triazole and 5-acetamido-1,2,4-triazole by
acetylnitrate and nitronium salt. Khim. Geterotsikl. Soedin. (Chem Heterocycle Compounds).
No. 8, p. 1132-1135, 1979
P. POLITZER, M. E. GRICE, and J.M. SEMINARIO: Density Functional Analysis of a
Decomposition of 4-Nitro-1,2,3-Triazole Trough the Evolution of N2. J. Quantum Chem. Vol.
61, No 3, p.389-392, 1997
H. H. LICHT, H. RITTER, H. R. BIRCHER, and P. BIGLER: Tautomerism in Nitrotriazoles:
Structure Investigation by Combined 1H, 13C and 15N NMR Spectroscopy, Magn. Reson. Chem,
No 36, p. 343, 1998
M.S. PEVZNER, E.YA. FEDOROVA, I.N. SHOKHOR, and L.I. BAGAL: Heterocycle
nitrocompounds. 9. Dipole moments of 3(5)-nitro-1,2,4-triazoles, Khim. Geterotsikl. Soedin.
(Chem. Heterocycle Compounds), No 2, p. 275. 1971
R.S. STEPANOV, L.A. KRUQLYAKOVA, and M.A. STEPANOVA: Journal of general
Chemistry, Vol. 67(2), p. 324. 1997
R.S. STEPANOV and L.A. KRUGLYAKOVA: Kinetics and catalysis, 37(3), p. 339 1996
J.C. OXLEY, J.L. SMITH, Z.L. ZHOU, and R.L. MCKENNEY: Thermal Decomposition
Studies on NTO and NTO/TNT, J. Phys. Chem., No 99, p. 10383 10391, 1995
137
BOMB CALORIMETRIC STUDY

OF A SERIES OF ENERGETIC LINEAR POLYPHOSPHAZENES
Anthony J. Bellamy*,1 Alessandro E. Contini*,
Peter Golding**, and Stephen J. Trussell**
Abstract:
Energetic linear polyphosphazenes are of potential interest as novel binders for
energetic formulations. However, the presence of phosphorus in these materials renders
their combustion chemistry significantly different to that of conventional binders. This
paper explores the combustion chemistry of a series of energetic polyphosphazenes. The
heat of combustion of each member of the series has been measured by bomb
calorimetry, and the water-soluble combustion products have been identified and
quantified using NMR Spectroscopy and Ion Chromatography. Since some of the
combustion products are hydrolytically unstable, it was necessary to stabilise the initial
combustion product mixtures by using a buffer solution instead of pure water in the
bomb, and then to determine the composition of the stabilized product mixtures in order
to obtain meaningful values for the heats of combustion and thence heats of formation.
The thermochemical measurements themselves were made with pure water in the bomb.
The composition and structures of the various polyphosphazenes have been correlated
with their heats of combustion and should enable polyphosphazenes with specific
thermochemical properties to be identified. The combustion of the polyphosphazenes
under a nitrogen atmosphere has also been studied.
Keywords:
1.
energetic polyphosphazenes, heat of combustion, heat of formation.
INTRODUCTION
Energetic polyphosphazenes have been suggested as promising, low hazard binders with
enhanced energy-density [1]. The oxygen combustion chemistry of five such energetic, linear
polyphosphazenes, the random unit structures of which are shown in Figure 1, has been
studied. The values of the measured internal energy of combustion (Uc) were employed to
calculate the standard enthalpies of combustion (Hc) and thence the enthalpies of
formation (Hf) of the polymers, which are necessary for the semi-empirical estimation of
the enthalpies of detonation.
Corresponding author; fax (44) 01793-783878, e-mail: A.J.Bellamy@cranfield.ac.uk
138
2.
RESULTS AND DISCUSSION

2.1
Preparation and structure elucidation
Polymers I, II, III, IV and V (Figure 1) were prepared by reacting an appropriate

alkoxide with linear poly[bis(2,2,2-trifluoroethoxy)phosphazene] to generate a random,
mixed substituent polymer, and then nitrating the initial substitution product [1]. The degrees
of trifluoroethoxy group replacement in the final polymers was estimated by 1H-NMR
spectroscopy and CHN elemental analysis, and were found to range between 31% and 76%
depending on the size of the nucleophile reacted with the homopolymer (Table 1), and the
reaction time. [100% energetic substituent (100% ES) represents complete replacement of
both trifluoroethoxy groups in the substitution reaction.] The empirical formulas of each
energetic polymer were derived from the values of the percent of energetic substituent.
Approximate oxygen balance values for the highest substituted specimens of each
polymer were estimated assuming ideal combustion behaviour viz. assuming only N2, CO2,
H2O, HF and H3PO4 are formed (Table 2). All four polymers are oxygen-deficient, with
polymer II exhibiting the least negative oxygen balance. Most of the polymers are, however,
less oxygen-deficient than the carbon-based energetic polymers polyNIMMO, polyGLYN
and GAP, as shown in Table 2.
2.2
Combustion chemistry studies
Samples (300 mg) of each energetic polymer were combusted in oxygen at 30 atm
pressure in a Gallenkamp Autobomb 305 static, adiabatic, oxygen bomb calorimeter, using
a Parr 1108-Cl halogen-resistant bomb. The calorimeter was calibrated with benzoic acid
thermochemical standard (Parr Inc.). A measured volume of either pure water
(thermochemical measurements) or pH 7 buffer (imidazole-imidazolium oxalate 0.6M; used
to stabilise the hydrolytically unstable products for product analysis) was added to the bomb
before firing. The combustion products, other than CO2, N2 and H2O, arising from all
samples were identified by 19F NMR spectroscopy (Figure 2) and Ion exchange
Chromatography (Figure 3) as nitric acid, hydrogen fluoride, orthophosphoric acid,
monofluoro- and difluoro-phosphoric acids. There is no evidence for the formation of
polycondensed phosphorus acid species [2], their fluorinated analogues [3], and carbon-based
fluorinated gases [4,5] (e.g. CF4, C2F6); hexafluorophosphoric acid is only formed when
energetic polyphosphazenes with low ES values, and the precursor linear poly[bis(2,2,2trifluoroethoxy)phosphazene], are burned. The hydrolytic instability of monofluoro- and
difluoro-phosphoric acid in unbuffered solution, which has been investigated by 19F and 31P
NMR spectroscopy by other workers [6,7,8], was confirmed by 19F NMR spectroscopy. The
use of 19F NMR spectroscopy is preferred over 31P NMR spectroscopy for the analysis of
these systems due to the simpler spectra in the case of the former [3,9].
The chemical analysis of the combustion products indicates that the overall combustion
process of the polymers in excess oxygen follows the stoichiometry shown in Equation 1,
which assumes that no polycondensed phosphorus acids and their fluorinated analogues are
formed.
139
2a + b c e 0.5z + 2.5y + w + 2.5f

combustion
Ca H b Oc Nd Fe Pf +
O2 (g) + nH2O(l)
2
b (e z 2w)
d y
(1)
aCO2(g) +
H 2 O(l) +
N 2(g) + (e z 2w)HF(aq) +
2
+ (f z w )H3PO4 (aq) + zH2 PO3F(aq) + wHPO2 F2 (aq) + yHNO3(aq)
Table 3 summarises the internal energy of combustion (Uc) of different batches of the
five energetic polymers, with different percent energetic substituent. Thermochemical
corrections were subsequently made to account for the amounts of orthophosphoric acid [10],
nitric acid [11], monofluoro- and difluoro-phosphoric acids [12], [and hexafluorophosphoric
acid when present], formed in the bomb, while the Washburn corrections [13] to standard
states were neglected.
As expected, the measured values of the internal energy of combustion were found to be
dependant upon the percent energetic substituent in the polymer. Polymer V showed the
highest calorific value, despite having a lower percent energetic substituent than any of the
other polymers. By converting the experimentally derived Uc values into Hc values [14]
and direct subtraction of the latter quantities from the sum of the standard enthalpies of
formation of the combustion products [12,15,16] it was possible to calculate the values of the
standard enthalpy of formation (Hf (298)) of each energetic polymer at different %ES values
(Table 4). Unfortunately however the percentage error in the Hf values is high (<5%).
This is because the absolute error in the Hf values is essentially the same as that in the
Hc values, there being relatively small errors in the Hf values of the combustion
products. The ultimate goal of the project will involve graphical correlation of Hf and
hence Hdet with %ES for each polymer, and should enable energetic polyphosphazenes
with specific thermochemical properties to be identified.
A further step towards estimating the Hdet of the energetic linear polyphosphazenes
was made by initiating combustion of polymer II (%ES 64) in an atmosphere of nitrogen
(30 atm). In these experiments the products were not greatly different from when II was
combusted in an oxygen atmosphere, except that HPF6 and CF3CH2OH (the latter is
normally observed when the polyphosphazenes are pyrolysed) were also present. The
accountancy for F and for P in the combustion products was >80%. The residue amounted to
4.0-6.4 wt%. These results suggest that in the absence of external oxygen, the preferred
processes are conversion of F and P to HF, P2O5 (H3PO4) and fluorinated analogues. Any
surplus oxygen is then expected to be distributed between the oxidation of C and the
remaining H in the normal manner. The Uc value was 3510 200 J g-1, compared to 8960
260 J g-1 for combustion of the same batch in oxygen.
British Crown Copyright 2005/MOD
Published with the permission of the Controller of Her Britannic Majestys Stationery
Office.
140
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
P.GOLDING and S.J.TRUSSELL: Energetic polyphosphazenes a new category of binders

for energetic formulations, Insensitive Munitions & Energetic Materials Technology
Symposium, San Francisco, 15-17 November 2004.
D.R.KIRKLIN and E.S.DOMALSKI: Enthalpies of combustion of triphenylphosphine and
triphenylphosphine oxide, J. Chem. Thermodynamics, 20(6), 743-754, 1988.
D.P.AMES, S.OHASHI, C.F.CALLIS and VAN J.R.WAZER:, Principles of phosphorus
chemistry. IV. The system of fluorophosphoric acids, J.Amer.Chem.Soc., 81, 6350-6357, 1959.
W.D.GOOD., D.W.SCOTT and G.WADDINGTON: Combustion calorimetryof organic
fluorine compounds by a rotating-bomb method, J.Phys.Chem., 60, 1080-1089, 1959
J.D.COX, H.A.GUNDRY and A.J.HEAD: Thermodynamic properties of fluorine compounds.
I. Heats of combustion of p-fluorobenzoic acid, pentafluorobenzoic acid, hexafluorobenzene,
and decafluorocyclohexane, Trans.Faraday Soc., 60, 653-665, 1964.
V.N.PLAKHOTNIK,
N.N.SHAMAKOVA,
V.B.TULCHINSKII,
E.G.LLIN
and
M.E.IGNATOV: Formation equilibrium of monofluorophosphoric acid, Zhurnal
Neorgranicheskoi Khimii, 30(11), 2773-2777, 1985.
V.V.SHELUCHENKO, S. S. DUBOV, G. I. DROZD and S. Z. IVIN: N.M.R. of phosphorus-31
and fluorine-19 of some tricoordinated phosphorus compounds containing phosphorus-fluorine
bonds, Zhurnal Strukturnoi Khimii , 9(5), 909-910, 1968.
P.VAST, A.SEMMOUD, A.ADDOU and G.PALAVIT: New method for preparation of
difluorophosphoric acid, J.Fluorine Chem., 27(3), 319-325, 1985.
J.F.NIXON, and R.SCHMUTZLER: Phosphorus-31 nuclear magnetic resonance studies of
phosphorus-fluorine compounds, Spectrochim.Acta, 20(12), 1835-1842, 1964.
A.F.BEDFORD and C.T.MORTIMER: Heats of formation and bond energies. II. Triethyl
phosphate, triphenylphosphine, and triphenylphosphine oxide, J. Chem. Soc. (Abstracts), 20,
743-754, 1988.
M.V.ROUX, T.A.TORRES and J.Z.DAVALOS: 1,2,4-Triazole as a reference material for
combustion calorimetry of N-containing compounds, J.Chem. Thermodynamics, 33, 949-957,
2000.
J.W.LARSON and B. SU: Thermodynamics of formation of aqueous monofluoro-, difluoroand amidofluoro-phosphoric acids, J.Chem.& Eng.Data, 39(1), 36-38, 1994.
E.M.WASHBURN, Standard states for bomb calorimetry, Bur.Standards J. Research, 10, 525528, 1933.
E.DIAZ, P.BRUSSEAU, G.AMPLEMAN and R.E.PRUDHOMME: Heats of combustion and
formation of new energetic thermoplastic elastomers based on GAP, PolyNIMMO and
PolyGLYN, Propellants, Explosives, Pyrotechnics, 28(3), 101-106, 2003.
S.G.LIAS, J.E.BARTMESS, J.F LIEBMAN, J.L.HOLMES, R.D.LEVIN and
W.G.MALLARD: Gas-phase ion and neutral thermochemistry, J.Physical & Chemical
Reference Data, Supplement, 1988, 17(1).
I.CONTINEANU and D.I. MARCHIDAN: The enthalpies of combustion and formation of
some isomers of aminobutanoic acid, Revue Romaine de Chimie, 39(12), 1391-1395, 1994.
141
OCH2CF3
[P
N ]x
OCH2CF3
OR
[P
OR
N ]y
OCH2CF3
R = CH2CH2ONO2
[P
N ]z
OR
CH2CH(ONO2)CH2ONO2
II
CH2CH2CH(ONO2)CH2ONO2
III
CH2CH2CH2CH(ONO2)CH2ONO2
IV
CH2CH2CH2CH2CH(ONO2)CH2ONO2
Fig 1.
Random structures of five energetic, linear polyphosphazenes
142
F-
2-
PO3F
PO2F2-
Fig 2.
19
F
internal
standard
Typical 19F NMR spectrum (acetone-d6 internal probe) of the undiluted,

buffered bomb solution.
FPO2F2NO3-
Fig 3.
PO43-
PO3F2-
Typical ion chromatogram (eluent: aqueous Na2CO3/ NaHCO3, 18/17 mM,

pH 10.3; retention time in minutes) of the diluted, buffered bomb solution.
143
Table 1. Percent energetic substituent as estimated by 1H NMR spectroscopy

and CHN elemental analysis.
Resultant energetic polyphosphazene
Energetic
Energetic
substituent
substituent
(%
Polymer
(%
by CHN
by
elemental
1
H NMR)
analysis)
[NP(OCH2CF3)0.48(OCH2CH2ONO2)1.52]
I (single
batch)
76
74
[NP(OCH2CF3)1.38(OCH2CH(ONO2)CH2ONO2)0.62]
II
batch 3
31
II
batch 1
65
II
batch 2
70
69
[NP(OCH2CF3)0.82(OCH2CH2CH(ONO2)CH2ONO2)1.18]
III
batch 1
59
[NP(OCH2CF3)0.68(OCH2CH2CH(ONO2)CH2ONO2)1.32]
III
batch 2
61
62
[NP(OCH2CF3)0.66(OCH2CH2CH2CH(ONO2)CH2ONO2)1.34]
IV
batch 1
67
67
[NP(OCH2CF3)1.00(OCH2CH2CH2CH2CH(ONO2)CH2ONO2)1.00]
V
batch 1
50
[NP(OCH2CF3)0.98(OCH2CH2CH2CH2CH
(ONO2)CH2ONO2)1.02]
V
batch 2
51
54
Table 2. Values of the oxygen balance (%) of five energetic polyphosphazene-based

polymers and of three carbon-based polymers.
Energetic Polymer
(% energetic substituent
Unit Empirical Formula
Oxygen Balance
by 1H NMR)
I (76%)
C4.00H7.04O6.56N2.52F1.44P1.00
-42.5
II (70%)
C5.40H8.20O10.40N5.20F1.80P1.00
-25.9
III (61%)
C6.42H10.06O9.27N3.42F2.36P1.00
-44.2
IV (67%)
C8.01H13.36O10.03N3.68F1.99P1.00
-58.1
V (51%)
C8.06H13.14O8.10N3.03F2.95P1.00
-67.7
PolyGLYN
C3.00H5.00O4.00N1.00
-60.5
PolyNIMMO
C5.00H9.00O4.00N1.00
-114.3
GAP
C3.00H5.00O1.00N3.00
-121.2
144
Table 3. Measured values of the internal energy of combustion (Uc) for polymers
I-V with different percent energetic substituent.
Energetic Polymer
%ES*
Uc (J g-1)
76
-10520 180 (1.7%)
31**
-8880 140 (1.6%)
65
-8640 105 (1.2%)
70
-9220 160 (1.7%)
59
-10890 160 (1.5%)
61
-11250 100 (0.9%)
67
-13060 210 (1.6%)
50
-13910 520 (3.7%)
51
-14450 180 (1.2%)
II
III
IV
V
* %ES = percent energetic substituent (as measured by 1H NMR spectroscopy)

** Some hexafluorophosphoric acid is formed on combustion
Table 4. Standard enthalpy of combustion (Hc) and standard enthalpy

of formation (Hf) of energetic polymers I-V.
Energetic
Polymer
%ES*
Hc (J g-1)
Hc (kJ mol-1)
Hf (J g-1)
Hf (kJ mol-1)
76
-10520 180
-2670 46
-5719 180
-1451 46
31**
-8880 140
-2612 41
-6974 140
-2052 41
70
-9190 160
-3290 57
-4496 160
-1609 57
III
61
-11250 100
-4043 36
-4512 100
-1621 36
IV
67
-13050 210
-5092 82
-4115 210
-1606 82
51
-14460 180
-5338 66
-4106 180
-1516 66
PolyGLYN
-2710***
-323***
PolyNIMMO
-2290***
-337***
GAP
-1150***
-114***
II
* %ES = percent energetic substituent (as measured by 1H NMR spectroscopy)

** Some hexafluorophosphoric acid is formed on combustion
*** Reference 14.
145
TLC ANALYSIS OF DADNE

AND SOME INTERMEDIATE PRODUCTS OF ITS SYNTHESIS
Jan Bdek, Sylwia Pietrzyk, Stanisaw Cudzio, and Zbigniew Chyek
Abstract:
In the paper, results of research on application of thin layer chromatography (TLC) for
determination of 1,1-diamino-2,2-dinitoethene (DADNE) and its precursors produced in
the synthetic path starting from 2-methylpyrimidine-4,6(3H,5H)-dione. Analytical
parameters of the substances and methodology of their quantitative analysis were
determined, and the results obtained were used for controlling the DADNE synthesis
process.
Keywords:
1.
1,1-Diamino-2,2-dinitoethene, analysis, TLC
INTRODUCTION
1,1-Diamino-2,2-dinitroethene is a novel explosive with low sensitivity and high

performance. Its detonation parameters are close to RDX whereas its sensitivity to different
stimuli is lower than that of TNT [1-3]. This is a result of the structure of the compound. The
molecular packing of DADNE consists of layers with strong intermolecular hydrogen bonds
within the layers and the bonds stabilize the molecule [4-5]. The unique combination of the
valuable properties arouses a lot of interest in the compound synthesis [6-9], analysis [10-13] and
performance [14-15].
O
C 2 H5 O
H2 N
O2 N
O
OC2H5
NH HCl
CH3ONa
CH3OH
O
HNO3
HN
H2SO4
CH3
HN
H2 N
NO2
(2)
NH2
C
O2N
(1)
H2 O
NH
O2 N
CH3
Fig 1.
NO2
NO2
+
CH2(NO2)2 + 2 CO2
DADNE synthesis from 2-methylpyrimidine-4,6(3H,5H)-dione (1) [9]
At our laboratory, DADNE is prepared by nitration of from 2-methylpyrimidine4,6(3H,5H)-dione (1) in concentrated HNO3/H2SO4 acids at room temperature (Fig. 1) [16, 17].
This route was first described by Astratev at al. [9], and according to them the final products
of nitration is 2-dinitromethylene-5,5-dinitropyrimidine-4,6(1H,3H)-dione (2) which is
directly hydrolyzed in water to DADNE and dinitromethane. Direct DADNE precursor (2)
can be isolated by filtration of the reaction mixture and consecutive washing out of the
deposit with trifluoroacetic anhydride and trifluoroacetic acid [9].
146
HN
98 % HNO3
HN
CH3
H2O
CH3
(1)
Fig 2.
NO2
(3)
Synthesis of 2-methyl-5-nitropyrimidine-4,6(3H,5H)-dione [9]
Additionally, we prepared 2-methyl-5-nitropyrimidine-4,6(3H,5H)-dione (3) (also

described by Astratev at al. [9]) from from 2-methylpyrimidine-4,6(3H,5H)-dione, using
98% HNO3 as nitrating agent (Fig. 2). Compound (3) is supposed to be a possible
intermediate product on the synthetic path shown in Fig. 1 [10].
2.
EXPERIMENTAL
The main aim of the present work was to find a methodology of quantitative analysis of
mixtures of the mentioned above compounds (2-methylpyrimidine-4,6(3H,5H)-dione (1), 2dinitromethylene-5,5-dinitropyrimidine-4,6(1H,3H)-dione (2), 2-methyl-5-nitropyrimidine4,6(3H,5H)-dione (3) and DADNE), that could be used to control DADNE synthesis. All the
analyses reported here were performed using instrumental TLC. We tried to find the
chromatographic system enabling separation of the compounds from their mixture as well as
quantitative dependences of the surface area of densitometric peaks on the amount of
analytes in the chromatographic bands.
2.1
Apparatus and materials
The equipment, chemicals and reagents used were: applicator Linomat IV (CAMAG);
horizontal chromatographic chamber DS; densitometer CS-9000 (SHIMADZU);
chromatoplates with silica gel 60 F254 HPLC (MERCK, Cat. No. 1.05548) or silica gel with
chemically bonded octadecyl (MERCK, Cat. No. 1.05559); organic solvents (J. T. Baker,
Merck); standards of the analysed substances (Military University of Technology).
2.2
Selection of chromatographic system
The first stage of investigations included finding a chromatographic system, which made
the separation possible in order to identify components and estimate their analytical
parameters. All experiments were conducted using acetone solutions containing 0.3 g of an
analyte in one l of acetone, and 10-l samples of the solutions were applied on the start line
of chromatoplates (Linomat IV, band width 4 mm). Chromatograms were developed to the
length of 59 cm in the horizontal chamber DS. The correctness of separation was assessed
by observation of fluorescence quenching.
Unfortunately, in this way we did not obtain good results. The analytes are very polar
compounds so that they are difficult to retain in conventional reversed phase conditions (on
silica gel with chemically bounded octadecyl). To make matter worse such compounds are
practically insoluble in many organic solvents, and this makes it difficult to choose a good
mobile phase.
The best separation (Fig. 3) was gained when two-steps elution in normal phase system
was applied. The conditions of separation were as follows:
147
the first step: (2 cm length) methanoldichloromethane (3:2; v/v),

the second step: (3 cm length) carbon tetrachlorideacetonitrile (3:2; v/v).
Fig 3. Chromatogram of the analytes:

(1) 2-methyl-pyrimidine-4,6(3H,5H)dione,
(2) 2-dinitromethylene-5,5dinitropyrimidine-4,6(1H,3H)-dione,
(3) 2-methyl-5-nitropyrimidine4,6(3H,5H)-dione,
(4) DADNE,
unmarked peak at zf ~ 20 dinitromethane,
A absorbance; zf distance of elution
2.3
Quantification
In order to increase the reliability of analyses, UV absorption spectra of separated

mixture components were measured in situ, getting information about a wave length max
which corresponds to the absorption maximum. Next, variable volumes of analyte solutions
(150 l) with concentrations from 0.1 to 1.0 g/l were applied on the start line of
chromatoplates, and the chromatograms were developed in the proposed phase system. The
amounts of the analyte were determined by scanning of a chromatogram (zig-zag technique),
at the characteristic value of max,. The average results from five measurements enabled us to
specify the equation of the standardization curve and evaluate the limit of detection and the
maximum range of linearity of detector indications (Tab. 1, Fig. 4).
Table 1. Analytical parameters of the anlytes tested
RF
max
[nm]
Calibration curve
2-Methylpyrimidine4,6(3H,5H)-dione
0.28
235
A = 16615m - 710
0.997
2.8
0.5
2-Methyl-5nitropyrimidine4,6(3H,5H)-dione
0.78
318
A = 4843m + 10202
0.960
25.0
2.0
2-Dinitromethylene5,5-dinitropyrimidine- 0.58
4,6(1H,3H)-dione
320
A= 13649m - 2276
0.978
2.5
0.5
Analyte
A = f(m)*
Correlation Maximum Detection

coefficient linearity [g] limit [g]
DADNE
0.92 328
A = 81042m - 1726
0.994
0.6
0.1
*A surface area of the densitometric peak, m mass of the analyte in the chromatographic band
[g]
148
140000
50000
120000
40000
100000
30000
80000
20000
60000
10000
40000
20000
0
-10000
10
a
40000
35000
30000
25000
20000
15000
10000
5000
0
-5000 0
3.
30
b
50000
40000
30000
20000
10000
0
-10000
c
Fig 4.
20
0,2
0,4
0,6
Calibration curves obtained in the densitometric measurements: a) 2methylpyrimidine-4,6(3H,5H)-dione, b) 2-methyl-5-nitropyrimidine4,6(3H,5H)-dione, c) 2-dinitromethylene-5,5-dinitropyrimidine4,6(1H,3H)-dione, and d) DADNE (Y-axis: absorbance, X-axis: the mass
of an analyte in the chromatographic band [g])
CONCLUSION
The performed experiments have shown that the method proposed is suitable for purity
assessment of DADNE. All analytes can be detected and determined side by side in a single
run. The detection limit is at the level of 0.11 g, but on the other side it is possible to
apply big samples even a few hundred of l in volume. Thanks to this the detection levels
are comparable to that of typical for HPLC. An advantage of TLC is the possibility of
simultaneous analysis of several samples that reduces the total costs of analysis.
Acknowledgment:
This research was supported by the State Committee of Scientific Research through
Military University of Technology, Grant 0 T00C 01528.
149
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
B. JANZON, H. BERGMAN, C. ELDSATER, C. LAMNEVIK, H. OSTMARK: FOX-7 a novel, high

performance, low-vulnerability high explosive for warhead applications, 20th Internationa
Symposium on Ballistcs, Orlando, FL, USA, 23-27.09.2003.
Y. N. MATYUSHIN, G. T. AFANASEV, V. P. LEBEDEV, M. N. MAHOV, V. I. PEPEKIN, TATB
and FOX-7: thermochemistry, performance, detonability, sensitivity, 34th International Annual
Conference of ICT, Karlsruhe, Germany, 24-27.06.2003.
H. OSTMARK, A. LANGLET, H. BERGMAN, N. WINGBORG, U. WELLMAR, U. BEMM, FOX-7
a new explosive with low sensitivity and high performance, 11th Internationa Symposium on
Detonation, Snowmass, CO, USA, 31.08-04.09.1998.
U. BEMM, H. OSTMARK, 1,1-Diamino-2,2-dinitroethylene a novel energetic material with
infinite layers in two dimensions, Acta Cryst., C54, pp. 1997-1998, 1998.
A. GINDULYTE, L. MASSA, L. HUANG, J. KARLE: Proposed mechanism of 1,1-diamino-2,2dinitroethylene decomposition: A functional theory study, J. Phys. Chem. A 103, pp. 1104511051, 1999.
N. LATYPOV, J. BERGMAN, A. LANGLET, U. WELLMAR, U. BEMM: Synthesis and reactios of
1,1-diamino-2,2-dinitroethylene, Tetrahedron 54, pp. 11525-11536, 1998.
N. LATYPOV, A. LANGLET, U. WELLMAR, Chemical compound suitable for use as an
explosive, intermediate and method for preparing the compound, US Patent 6 312 538 B1,
Nov. 6, 2001.
H. OSTMARK, H. BERGMAN, U. BEMM, P. GOEDE, E. HOLMGREN, M. JOHANSSON, A.
LANGLET, N. LATYPOV, A. PETTERSSON, M-L PETTERSSON, N. WINGBORG, C. VORDE, H.
STENMARK, L. KARLSSON, M. HIHKIO, 2,2-Dinitro-ethene-1,1-diamine (FOX-7) properties,
analysis and scale-up, 32nd International Annual Conference of ICT, Karlsruhe, Germany,
2001.
A.A. ASTRATIEV, A. DASHKO, D. MERSHIN, A. STEPANOV, A. URAZGILDIEV: Russian Journal
of Organic Chemistry 37, pp. 729-733, 2001.
E. HOLMGREN, H. CARLSSON, P. GOEDE, N. LATYPOV, C. CRESCENZI: Characterization of
FOX-7, its precursors and possible byproducts, 34th International Annual Conference of ICT,
Karlsruhe, Germany, 24-27.06.2003, also A. J. BELLAMY, N.V. LATYPOV, P. GOEDE, Studies
on the nitration of new potential precursors for FOX-7, New Trends in Research of Energetic
Materials, Pardubice, Czech Republic, 116-122, 2004.
U. TICMANIS, M. KAISER, G. PANTEL, I. FUHR, U. TEIPEL, Kinetics and chemistry of thermal
decomposition of FOX-7, 35th International Annual Conference of ICT, Karlsruhe, Germany,
2004.
P. B. KEMPA, M. HERRMANN, F. J. MOLINA METZGER, V. THOME, A. KJELLSTROM, N.
LATYPOV, Phase transitions of FOX-7 studied by X-ray diffraction and thermal analysis, 35th
International Annual Conference of ICT, Karlsruhe, Germany, 2004.
N. V. GARMASHEVA, I. V. CHEMAGINA, V. P. FILIN, M. B. KAZAKOVA, G. B. LOBOIKO,
Investigation of diaminodinitroethylene, New Trends in Research of Energetic Materials,
Pardubice, Czech Republic, 116-122, 2004.
R. WILD, U. TEIPEL, Characterization and explosive properties of FOX-7, 35th International
Annual Conference of ICT, Karlsruhe, Germany, 2004.
U. TEIPEL, I. FUHR, K. HARTLIEB, A. KJELLSTROM, C. ELDSATER, Crystallization of 1,1diamino-2,2-dinitroethylene (FOX-7), 35th International Annual Conference of ICT, Karlsruhe,
Germany, 2004.
S. CUDZIO, W. KICISKI, Properties and preparation of DADNE (in Polish), Biul. WAT 53,
(2-3), pp. 183-200, 2004.
S. CUDZIO, Z. CHYEK, J. BDEK, S. PIETRZYK, Optimization of DADNE synthesis, This
proceedings, 2005.
150
REACTIVITY BETWEEN -CL20-GAP AND -HMX-GAP

INVESTIGATED BY MASS LOSS, ADIABATIC SELF HEATING
AND DYNAMIC MECHANICAL ANALYSIS
Manfred A. Bohn, Manuela Drich, Jasmin Aniol, Heike Pontius, and Peter Gerber
Fraunhofer-Institut fr Chemische Technologie (ICT)
Postfach 1240, D-76318 Pfinztal-Berghausen, Germany
Abstract
The crystalline energetic component -CL20 (-HNIW, hexanitro-hexaazaisowurtzitane, crystallized in -phase) has the potential to be used as ingredient in high
performance formulations, for example in high burning rate rocket propellants. A closer
inspection of the reactivity of -CL20 revealed a marked difference between -CL20GAP and -HMX-GAP. Experimental data of the reactivity between the 1:1 by mass
mixtures -CL20-GAP diol, -HMX-GAP diol and of 45:55 by mass formulations CL20/GAP-N100 and -HMX/GAP-N100 (GAP-N100 means GAP diol cured with
DesmodurTM N100) are presented. The data have been obtained by the following
measurement methods:
mass loss
gas generation
adiabatic self heat rate measured by ARCTM
dynamic mechanical analysis (DMA)
Gas generation determined with a standard vacuum stability test apparatus was used to
sustain the evaluation. With all four measurement methods the difference in reactivity
between -CL20-GAP and -HMX-GAP is found. Further to this a difference in
reactivity is observed between the mere mixture of -CL20-GAP diol and the 45:55
formulation -CL20/GAP-N100. So-named reactivity functions are obtained from the
raw data of mass loss, which describe the reactive part between the components. The
investigation of the mechanical properties of the 45:55 formulations -CL20/GAP-N100
and -HMX/GAP-N100 by dynamic mechanical analysis shows the marked difference in
binder ageing between them.
Keywords:
1.
CL20 (HNIW), HMX, GAP, GAP-N100, compatibility, reactivity

functions, ageing, mass loss, gas generation, adiabatic self heating,
dynamic mechanical analysis, shear modulus, glass transition
temperature shift
INTRODUCTION
The high energy material -CL20 (-HNIW) together with energetic binders as GAPN100 gives a realistic base for new high energy fast burning rocket propellants [1,2]. Besides
the successful life time assessment of formulations containing -CL20 [3,4] bonded by
GAP-N100, the experimental results indicate a difference in the compatibility behaviour of
-HMX-GAP and -CL20-GAP. This is surprising because both -CL20 and -HMX are
nitramines and are stable substances. The thermal stability of -CL20 as substance alone is
good as shown in several investigations [5,6]. Two types of component pairs have been under
151
investigation: (i) 1:1 mixtures by mass of -CL20-GAP diol and -HMX-GAP diol, (ii)
kneaded and cured 45:55 by mass formulations of -CL20/GAP-N100 and -HMX/GAPN100. The reactivity differences have been determined by mass loss ML, gas generation GG
and adiabatic self heating h. The measurement methods probe the materials differently: ML
and GG by the split off reaction gases and h by the rate of the net sum of the reaction heats.
The found result will be assured if both probings reveal the same behaviour. In [7] results
obtained with heat generation are presented and discussed and an explanation is given for the
found differences in reactivity based on molecular modelling.
2.
SUBSTANCES AND MEASUREMENT METHODS
The chemical formulas of the substances used are shown in Fig. 1. The diol GAP (GAP
has in average not two OH end groups per GAP chain, it has between 1.3 and 1.85, lot
depending) reacts with oligoisocyanate by addition reaction to a three dimensional
polyurethane elastomer. The 1:1 mixtures by mass of -CL20-GAP diol and -HMX-GAP
diol were mixed intensively by hand. The GAP diol was from SNPE (this corresponding part
of SNPE company is named now SME), France, the -CL20 and -HMX lots have been:
-CL20 from Thiokol, USA, with mean particle size of 3.2 m
-HMX from Dyno, Norway, with mean particle size of 5 m
NO2
H
O2NN
O2NN
CH2
O2 N N
N NO2
H
O2NN
H2C
NNO2
NNO2
nitramine
group
H2C
N NO2
N
CH2
HMX
NO2
CL20 or HNIW
O
HO
CH
CH2
CH2 CH2
CH2
N3
CH2 CH
OH
N3
N C N C N
H
H
CH2
CH2
NCO
n
DesmodurTM N100
GAP diol
Fig 1.
CH2
CH2
n
OCN
The chemical formulas of the substances used. The oligoisocyanate

DesmodurTM N100 and GAP diol are shown with idealized formulas.
N100 contains only about 2.5 active isocyanate groups per molecule.
The unit multiplier n is about 10 with standard GAP diol.
152
NCO
The formulations -CL20/GAP-N100 (CGN) and -HMX/GAP-N100 (HGN) have been

made in the technical plant of ICT in a kneader [8]. The curing agent DesmodurTM N100 from
Bayer Company, Germany, was added at the end of the main kneading. As curing catalyst
dibutyltin dilaurate was added in a very small amount. The filling of plate like forms was
done under vacuum. The curing conditions have been 60C over 1 day under confinement in
the plate and additionally 60C over 3 days in the mould without confinement. The same
GAP diol from SNPE (now SME), France was used as with the mixtures. The following CL20 and -HMX lots were taken:
-CL20 from SNPE (now SME), France, with mean particle size of 8.4 m
-HMX from Dyno, Norway, with mean particle size of 10 m
The ratio of the crystalline energetic materials to the binder GAP-N100 (GN) was 45:55
(0.818 : 1) by mass. The relative reduction of the energetics part was done to facilitate the
casting of the formulation into the plate mould. The molar ratios of the formulations are with
the molar masses m of the three components mGN = 1871 g/mol, mCL20 = 438.2 g/mol and
mHMX = 296.2 g/mol:
CGN: 0.777 : 0.223 = 3.48 : 1
HGN: 0.838 : 0.162 = 5.17 : 1
The measurement methods used have been:
mass loss by storing the samples in loosely stoppered glass vials placed in
aluminium block oven with PID controlled temperature and remote weighing at
intervals adapted to the substance decomposition rate; two samples in parallel,
sample amounts 2g each.
gas generation determined with the standard vacuum stability test apparature at
100C over 40h with 2.5g sample amount, two samples in parallel.
adiabatic self heating determined with an ARCTM, made by former company
Columbia Scientific Industries, now TIAX LLC, 15F/320 Acorn Park, Cambridge,
MA 02140-2390, USA; sample amounts about 200mg, heat step 4C, sensitivity
limit for detection of exotherme 0.02C/min, initial atmosphere vacuum.
dynamic mechanical analyser (DMA) ARESTM (Advanced Rheometric Expansion
System) from Rheometric Scientific Inc, Piscataway, NJ, USA. Determination of
complex shear modulus G*() = G()+iG() by torsion DMA with forced
sinusoidal deformation (shear) (t) = 0sin(t) of a rectangular sample as function
of temperature and deformation frequency f, which gives a deformation rate d(t)/dt
= 0cos(t). The symbol means angular frequency =2f, f is frequency in Hz,
0 is deformation or shear amplitude. Temperature scan mode, 1C/min with 4 min
equilibration time with frequency scan at each temperature. Sample dimensions
have been: 12mm broad, 4 to 5 mm thick and about 50mm long. Deformation with
frequency: 0.1 Hz, 1 Hz and 10 Hz. The shear amplitudes have been adjusted to
deform the samples in the range of linear deformation behaviour.
153
3.
METHOD TO DETERMINE REACTIVITY
A chemical reaction between two components is recognizable by its effects, namely

formation of reaction products, reaction heat and consumption of reaction educts, Eq.(1).
A + B gases + liquids + solids + reaction heat
(1)
Ideally compatible components will not create any reaction between themselves when
brought into contact, but this is not often the case. To the already existing decomposition
reactions of the components alone additional decomposition ways generally for both
components are created. Educt A forms PA, educt B forms PB and A+B form P and each of
them open additional reactions with A and B each. An additional chemical conversion is
created with regard to the chemical conversions of the components alone. Several methods
can be used to determine the additional chemical conversion:
gas generation
mass loss
heat generation
analysis of educts and products as function of time
For assessing compatibility or reactivity the best is to use a procedure based on the
excess chemical conversion. The excess reaction part or the excess conversion is best
obtained by the difference of measured quantities P of so-named extensive character, which
are in amount proportional to the excess conversion. That means the values of the single
components are subtracted from the value of the mixture. This is analogous to the
thermodynamic description of mixtures with excess quantities. The here used reactivity
functions RP(t,T) are defined as the difference of measured quantities P, Eq.(2), which are
scaled in substance amount and taken as absolute or reference value normalized quantities. It
can be formed also with mixture ratios based on molar amount, volume, surface or number
of reactive groups. With 1:1 mixtures by mass one has MGi / MG = 0.5. Eq.(3) shows the fully
normalized reactivity expression RPr(t,T), which is a conversion quantity and correctly
RPr(t,T) is named excess conversion function.
RP ( t , T) =
M
P ( t , T) M G,2
PG ( t , T)

G,1 1
MG
M G M 1
MG
P ( t , T)
2
M2
PG( t , T)
P1( t , T)
M G,1 M 1 M G,2
MG
RPr ( t , T) =
MG
M G P1,ref
PG,ref
154
P2 ( t , T)
M2
P2,ref
(2)
(3)
Meaning of symbols in Eq.(2) and Eq.(3):

RP
RPr
MG
MG1
MG2
PG
P1
P2
PG,ref
P1,ref
P2,ref
reactivity function or excess reaction function, determined with

measurement quantity P
normalized reactivity function or excess conversion function, determined
with measurement quantity P and the corresponding reference values Pi,ref
mass of the mixture at t=0
mass of component 1 in mixture at t=0
mass of component 2 in mixture at t=0
measurement quantity P of mixture
measurement quantity P of component 1, with sample mass M1 at t=0
measurement quantity P of component 2, with sample mass M2 at t=0
reference value of measurement quantity P of mixture, amount normalized
reference value of measurement quantity P of component 1, amount
normalized
reference value of measurement quantity P of component 2, amount
normalized
To investigate not well known substances, one must use several measurement methods
to determine compatibility in order to get a high enough reliability for the assessment, if one
uses so-named summarizing measurement methods as mass loss, heat generation and gas
generation. In the mixture of components we have ongoing decomposition reactions of the
components and between the components with changing reaction intensity and possible
changes in decomposition mechanisms with temperature. Therefore single measurement
values often are not enough to assess compatibility, especially with unknown substances. A
fairly reliable way is to use two methods which probe the samples in a different way, means
for example with gas generation and heat generation. The following three measurement
quantities P can be used preferably isothermally:
mass loss ML (ML(t,T) = 1M(t,T)/M(0))
gas generation GG
heat generation (HG) Q.
Mass loss is complementary to gas generation, but because of the normalization
advantage the mass loss is a very convenient measurement quantity. With mass loss, one
has immediately the formulation as excess conversion function or normalized reactivity
function, Eq.(4) and Eq.(5). With gas generation GG and heat generation Q mostly one uses
the excess reaction functions, Eq.(6) and Eq.(7). In these equations GGi(t,T) and Qi(t,T) are
already initial mass normalized. In [7] and [9] more can be found to the definition of
conversion type reactivity functions.
RMr ( t , T) =
M G (0) M G ( t , T) M G,1(0) M 1(0) M 1( t , T) M G,2 (0) M 2 (0) M 2 ( t , T)

(4)

M G (0 )
M G (0)
M 1 (0 )
M 2 (0 )
M G (0 )
RML ( t , T) = RMr ( t , T) = ML G ( t , T)
M G,1(0)
M G (0 )
155
ML 1( t , T)
M G,2 (0)
M G (0 )
ML 2 ( t , T)
(5)
RGG( t , T) = GGG ( t , T)
RQ ( t , T) = Q G ( t , T)
4.
M G,1(0)
M G (0 )
M G,1(0)
M G (0 )
GG1( t , T)
Q1( t , T)
M G,2 (0)
M G (0 )
M G,2 (0)
M G (0 )
GG2 ( t , T)
Q 2 ( t , T)
(6)
(7)
ASSESSMENT CRITERIA FOR REACTIVITIES RP
For the assessment of compatibility or reactivity but also for stability, one must fix three
conditions with single point measurements:
test temperature
test time
limit value of the excess conversion at test temperature and test time
for stability: limit value for conversion at test temperature and test time
To get the data in an effective time period suitable temperature-time data for the tests
must be chosen, which represent a sufficient use time at ambient temperatures. The
compilation in Table 1 shows such data extrapolated from some temperature-time data used
in stability and compatibility tests. Eq.(8) was used to obtain the corresponding times tE at
25C from test time and test temperature.
t E[a] = t T [d] F( TT TE ) / TF / 365d
tE
tT
F
TT
TE
TF
(8)
time in years at temperature TE, here 25C

test time in days at test temperature TT
acceleration or deceleration factor per 10C change, here 3
test temperature in C
environmental temperature in C
temperature interval for actual value of factor F, here always 10C
Table 1. Extrapolated times in years at 25C from some temperature-time data

used in stability and compatibility tests.
Temperature, time
Extrapolated time in years at

25C with factor 3 per 10C
temperature change
100C, 40h
17.3
90C, 18d
62.3
90C,
3d
10.4
80C, 10d
11.5
75C, 15d
9.99
8.9
65C, 40d
To assess mixtures of not well known components some general validity for the limit
values must be found. Based on the upper limit value from the reactivity test done with gas
156
generation at 100C, 40h, namely 1.0 ml/g (5 ml per 2.5g+2.5g) (lower limit value is 0.6
ml/g), the limit values for the other measurement methods are deduced. With this limit value
of excess gas generation 0.25% excess conversion is assigned using the reference value of
400 ml/g for the mixture in test, means it is able to produce in average 400 ml gas per gram
of mass at total decomposition. This is a low limit value in order to be on the safe side.
Because of possible negatively directed interactions between the components the limits for
RGG are defined as -1.0 ml/g RGG +1.0 ml/g. The same holds for all other limit values.
With the above gas evolution limit at 100C and 40h (=1.667d) we obtain the ML reactivity
criterion at 90C, 18d to be 0.9% excess conversion: 0.25% x 18d / (3 x 1.667d) =
0.9%. If we go now to 80C and maintain the same degree of excess conversion we could
extent the time by a factor of 3. But if measurements not longer than 10 days at 80C are
needed, the limit of 0.9% is divided by a factor of 3x1.8 = 5.4 and the limit excess
conversion of 0.17% at 80C, 10d is obtained for RML. Table 2 lists some of the possible
limit values. More on the systematic to obtain such data can be found in [9].
Table 2. Limit values for two reactivity quantities RP. The allowed range for RP
is from minus to plus: low limit RP high limit. The master limit values
are given in the first line for every method. The deduced limit values
are obtained with factor 3 per 10C temperature change.
RP
limits for assessment of

excess conversion
reference value PG,ref
0.9 % at 90C, 18d

ML
GG
RML [%]
RGG [ml/g]
0.9 % at 100C, 6d
0.25 % at 100C, 40h
0.17 % at 80C, 10dh
1 ml/g at 100C, 40h
3.6 ml/g at 90C, 18d
3.6 ml/g at 100C, 6d
Initial mass MG(0).

Base is 0.25% at 100C, 40h.
Base is 0.25% of 400 ml/g at
100C, 40h. This is the average
final mixture gas generation value
GGG(te,T).
To find the right reference value with gas generation can be a problem. The question is
what is the total gas generated when all the component mixture has reacted to the state where
no further gas is generated, and this has to be achieved at the chosen test temperature.
Additionally the ongoing reactions should be the same along the complete reaction
coordinate or conversion. Experimentally this is almost never achievable at typical lower to
medium high test temperatures, therefore only assumptions are possible, which clearly
depend on the nature of the components. For completion, the limit values LP to determine the
stability of a component or formulation are listed in Table 3 for the two measurement
methods ML and GG. Again in the first line the master limit value is given and below this
some deduced limit values, whereby the factor 3 per 10C temperature change was taken.
The master limit value for mass loss was taken from the testing of NC-based gun and rocket
propellants with the ML stability criterion at 90C, 18d to be 3% of conversion, means
measured mass loss.
157
Table 3. Limit values LP for the two measurement methods ML and GG to determine
the stability of a component or formulation. The master limit values are
given in the first line for every method. The deduced limit values are
obtained with factor 3 per 10C temperature change.
LP
limits for stability

assessment
reference value Pref
3% at 90C, 18d
ML
GG
5.
LML [%]
LGG [ml/g]
0.83% at 100C, 40h

0.56% at 80C, 10d
3% at 85C, 31.2d
2 ml/g at 100C, 40h
7.2 ml/g at 90C, 18d
1.33 ml/g at 80C, 10d
Initial
mass
Base is 3% at 90C, 18d.
M(0).
Base is 0.25% of 800 ml/g at

100C, 40h. This is the average
final substance gas generation
value GG(te,T).
RESULTS
5.1
Mass loss measurements
The Fig. 2 shows the mass loss curves ML(t,T) at 100C of the 1:1 mixtures by mass and
of their components, as well as the reactivity function RML(t,T) of the mixtures. Clearly CL20-GAP diol shows higher reactivity than -HMX-GAP diol. But according to the criteria
of stability, 100C, 6d, 3% and reactivity, 100C, 6d, 0.9%, the mixture -CL20-GAP
diol is assessed as stable and -CL20 as compatible with GAP diol. In Fig. 3 the reactivity
functions RML(t,T) of the 1:1 mixture -CL20-GAP diol at 90C, 100C, 110C and 120C
can be seen. The mass loss curves of -CL20/GAP-N100 (CGN) at 70C, 80C and 90C
and that of -HMX/GAP-N100 (HGN) at 100C, 110C and 120C can be seen in Fig. 4.
The choice of temperatures has been such to have comparable conversions between CGN
and HGN, however for pairs of temperatures different by 30C for the two formulations.
This indicates the great reactivity difference of these two substances. The reactivity rates
differ by a factor of about 40 to 50. The stability criterion at 90C, 18d, 3%, is exceeded by
CGN and this is the same at all other temperatures. Fig. 5 shows the mass loss curves of
CGN and HGN and of the components and the reactivity curves of CGN and HGN at 90C.
HMX is not shown, because no mass loss other than some moisture loss could be detected.
The information from this figure is that the reactivity of CGN is much higher than that of
HGN and CGN must be stated as instable and -CL20 incompatible with GAP-N100.
158
1,8
e-CL20-GAP diol
b-HMX-GAP diol
Reactivity function RML at 100C
ML [%]
RML [%]
1,6
1,4
(GAP/CL20) -GAP -CL20

(GAP/HMX) -GAP -HMX
GAP/HMX
GAP/CL20
R ML CL20-GAP
GAP
CL20
ML= 0.4% at 6d
HMX
zero line
1,2
1
0,8
0,6
0,4
RML HMX-GAP
0,2
0
time [d]
-0,2
-2
10
12
14
16
18
Mass loss ML at 100C of -CL20, -HMX, GAP diol, 1:1-mixtures

-CL20-GAP diol and -HMX-GAP diol and reactivity functions RML
of the mixtures. The figure shows smoothed curves.
Fig 2.
3.5
RML [%]
reactivity functions R ML(t,T)

e-CL20-GAP diol
1:1 mixture by mass
3
120C
2.5
110C
90C
100C
110C
1.5
120C
100C
90C
0.5
time [d]
0
0
Fig 3.
10
15
20
25
30
35
40
Reactivity functions RML(t,T) of the 1:1 mixture by mass -CL20-GAP diol

determined from mass loss measurements.
159
mass loss comparison of

e-CL20 / GAP-N100 (45:55)
b-HMX / GAP-N100 (45:55)
reactivity difference corresponds
to more than 30C
ML [%]
4
CGN-80C
CGN-90C
CGN, 90C
HGN, 120C
CGN-70C
CGN, 80C
HGN-120C
HGN-110C
HGN, 110C
HGN-100C
2
CGN, 70C
1
HGN, 100C
time [d]
0
0
10
Fig 4.
ML [%]
RML [%]
20
30
40
50
60
70
80
Mass loss curves ML(t,T) at several temperatures of -CL20/GAP-N100

(CGN) and -HMX/GAP-N100 (HGN). A strong reactivity difference
between CGN and HGN can be recognized. To reach with HGN the same
mass loss rate as with CGN the temperature of HGN must be raised
by somewhat more than 30C.
reactivity function RML of

e-CL20/GAP-N100 or CGN (45:55)
b-HMX/GAP-N100 or HGN (45:55)
90C
CGN : R ML = 5.1%
at 18d
HGN: RML = 0.14% at 18d

R ML limit at 90C and 18d: 0.9%
ML-CL20-GN
ML CGN
R-CGN
ML-GN
ML-CL20
RML CGN
R-HGN
ML-HMX-GN
ML HGN
ML GN
ML CL20
RML HGN
time [d]
0
0
Fig 5.
12
15
18
21
24
27
Mass loss ML at 90C of -CL20/GAP-N100 (CGN), -HMX/GAP-N100

(HGN), GAP-N100 (GN) and -CL20. ML of HMX is not shown because
there was no mass loss. The reactivity function RML of CGN exceeds that
of HGN very much.
160
30
The mass loss curves of GAP-N100 can be seen in Fig. 6. GAP decomposes without
autocatalytic acceleration, the curve at 130C seems to be caused by a first order
decomposition. In detail the decomposition is a combination of first order and consecutive
reactions in splitting of molecular nitrogen and backbone decomposition products. At the
beginning the main decomposition reaction is loss of nitrogen from the azide group.
35
ML [%]
GAP-N100 (GN)
90C
30
100C
110C
25
130C
120C
130C
20
15
120C
10
90C
100C
110C
5
0
time [d]
-5
-10
10
Fig 6.
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
Mass loss as function of time and temperature of cured GAP, GAP-N100.

The ML reaches a plateau value at about 32%.
55
ML [%]
80C
b-HMX / GAP-N100 (HGN)
90C
45
100C
110C
35
120C
120C
130C
GAP-N100, 130C
130C
GN-130C
25
15
110C
5
-5
-10
100C
time [d]
0
Fig 7.
10
20
30
40
50
60
70
80
90
100
110
120
Mass loss of 45:55 formulation -HMX/GAP-N100 as function of time and

temperature. For comparison the ML curve of GAP-N100 (GN) at 130C is
shown also.
161
5
ML [%]
80C
b-HMX / GAP-N100 (HGN)
90C
GN, 130C
100C
120C
110C
GN, 120C
110C
130C
120C
130C
GN-130C
GN, 110C
GN-120C
GN-110C
2
100C
1
90C
time [d]
80C
0
-10
10
Fig 8.
20
30
40
50
60
70
80
90
100
110
120
Mass loss of 45:55 formulation -HMX/GAP-N100 as function of time and

temperature up to ML = 5%. For comparison the ML curves
of GAP-N100 (GN) are shown at 110C, 120C and 130C.
40
ML [%]
e-CL20 / GAP-N100 (CGN)
35
70C
80C
90C
100C
110C
120C
130C
GN-130C
130C
30
110C
25
90C
100C
120C
20
GAP-N100, 130C
15
80C
10
5
70C
0
time [d]
-5
-10
Fig 9.
10
20
30
40
50
60
70
80
90
100
110
Mass loss of 45:55 formulation -CL20/GAP-N100 as function of time

and temperature. For comparison the ML curve of GAP-N100 at 130C
is shown.
162
120
The 45:55 formulation -HMX/GAP-N100 (HGN) shows an increasing decomposition

rate as found with autocatalytic decomposition, especially at the higher measurement
temperatures. A plateau value of 42% to 44% of ML will be reached. HGN decomposes
much faster than GAP-N100, which can be seen in Fig. 7 and Fig. 8 in comparison with the
curves for GAP-N100. Fig. 9 shows the mass loss curves of the 45:55 formulation CL20/GAP-N100 (CGN) at temperatures between 70C and 130C. Compared to HGN the
decomposition rate is much higher as already stated with Fig. 4. Also CGN shows
autocatalytic type increase of reaction rate and a plateau value is indicated at about 36% to
37% of ML. The molar ratio of -HMX to -CL20 in the formulations is about 1.48. The
mass loss part of GAP-N100 may be assumed to be 17.6%. HMX contributes then 25.4%
and CL20 18.4% to the total mass loss. The number of molecules accessible at the surface of
the explosive particles is taken as proportional to the surface area assuming a similar
occupied size in the surface by both molecules. The surface ratio CL20 to HMX particles is
about 0.71, which is about the same as their mass loss contribution ratio.
5.2
Gas generation
Table 4 summarizes the data obtained from gas generation GG at 100C and 40h. It lists
on the left side the results for the 1:1 mixtures and their components and on the right side the
results of the formulations. The reactivity RGG between -CL20 and GAP diol is about 4
times that of -HMX and GAP diol. All substances and mixtures are assessed stable and the
two mixture reactivity values RGG fulfil the limit criterion, 100C, 40h, -1 ml/g RGG +1
ml/g. The criterion for stability is 2 ml/g. The formulation -CL20/GAP-N100 must be
assessed as unstable and according to RGG -CL20 is incompatible in GAP-N100.
Table 4. Gas generations GG ( 2ml/g) and reactivities RGG (-1 ml/g RGG 1
ml/g) at 100C over 40h of the 1:1-mixtures and their components as well
as of the two formulations -CL20/GAP-N100 and -HMX/GAP-N100
and their components, determined with the standard vacuum stability test.
Volumes normalized to 0C and 1 atm.
1:1 mixtures by mass and components 45:55 formulations by mass and components
GG [ml/g]
assessment
GG [ml/g]
assessment
GAP diol
0.25
stable
GAP-N100
0.29
stable
-CL20,
3.2m
0.06
stable
-CL20, 25 m
0.07
stable
-HMX, 5
m
0.02
stable
-HMX, 10 m
0.08
stable
-CL20GAP
diol
0.42
stable
-CL20/GAP-N100
3.03
instable
-HMX
GAP diol
0.20
stable
-HMX/GAPN100
0.43
stable
RGG [ml/g]
RGG [ml/g]
-CL20GAP
diol
+ 0.265
compatible
-CL20/GAP-N100
+ 2.85
incompatible
-HMX
GAP diol
+ 0.065
compatible
-HMX/GAPN100
+ 0.25
compatible
163
5.3
Adiabatic self heating
The results of adiabatic self heating are presented in Fig. 10 for the mixtures and in Fig.
11 for the formulations, with their components each. They have been determined with an
ARCTM, for details of the method see [10,11]. The adiabatic self heat rate h is shown as
function of the inverse absolute temperature, centigrade are assigned for convenience. The
temperature is the one which the sample has reached by its self heating. The adiabatic self
heat rate curves start from the onset temperature detected by the apparatus and end at the
transition from by the apparatus controlled self heating to deflagration. Only the curves of
GAP diol and GAP-N100 are fully measured curves until the substance was consumed. The
mixture -HMX-GAP diol has an onset temperature coinciding with GAP diol alone, means
the onset of self heating of the mixture is controlled by GAP diol. Some reactivity in the
mixture is detected, because the slope of the curve of -HMX-GAP diol is greater than the
one of GAP diol. With the mixture -CL20-GAP diol reactivity is clearly detected by the
greater slope compared to GAP diol and by the onset temperature of 155C of the mixture
compared to GAP diol with 167C and -CL20 with 172C. The shift of 12C in onset
temperature between GAP diol and -CL20-GAP diol can be assigned as compatible based
on the results from mass loss, heat generation [7] and gas generation. The onset temperature
of -CL20/GAP-N100, Fig. 11, with 123C is largely shifted to lower temperatures, namely
by 49C, compared to the onset of GAP-N100 with 172C. From this the guess is
incompatibility. This guess is confirmed by the following results of mass loss and of heat
generation given in [7] and by the data of Table 4, right side. The gas generation of CL20/GAP-N100 exceeds with 3.03 ml/g at 100C, 40h also the limit of stability, maximum
2ml/g.
Fig 10. Adiabatic self heating of the 1:1 mixtures and their components,
determined with an ARCTM. The curve of -CL20-GAP diol is clearly
shifted to lower temperatures compared to GAP alone, indicating a higher
reactivity than in -HMX-GAP diol. The adiabatic self heating
of the -HMX-GAP mixture starts at about the same temperature
as the one of GAP diol alone.
164
Fig 11. Adiabatic self heating of the formulations and their components,
determined with an ARCTM. The great shift to lower onset temperature
of -CL20/GAP-N100 is indicative for a high reactivity between
-CL20 and GAP-N100.
Compared to -HMX-GAP diol the formulation -HMX/GAP-N100 shows some higher
reactivity too, see Table 4. With adiabatic self heating the higher reactivity of HGN
compared to the 1:1 mixture is revealed also, because the onset temperature of HGN is
shifted below the one of GAP-N100. The -CL20/GAP-N100 curve shows a convex
curvature in direction of the abscissa, means an accelerated increase of self heat rate. This is
the same effect seen with the isothermal curves in Fig. 2 to Fig. 5 and the cause is assumed
to be autocatalytic acceleration. Also -HMX/GAP-N100 shows this curvature, not so
significant but recognizable and also the -CL20 and -HMX alone in the upper part of their
self heat rates. Non autocatalytic behaviour is shown by GAP-N100 in Fig. 11 and by GAP
diol in Fig. 10. They show a maximum because of substance consumption.
5.4
DMA measurements
Dynamic mechanical analysis in torsion mode was performed with aged sample of GAPN100 and the two formulations. The data of moderate and strongly aged samples are
presented here. The measurements were made at three different deformation rates
distinguished by the frequency of the forced sinusoidal deformation. Fig. 12 shows an
example with -CL20/GAP-N100 aged at 80C over 2 days. The storage modulus G and the
loss factor tan()= G/G can be seen as function of temperature and deformation frequency.
The higher the frequency the higher the deformation rate and the response is like a material
with increasing strength, recognizable by the shift of the glass transition to higher
temperatures. In this work the glass transition temperature is taken as the maximum of loss
factor tan() = f(T). Fig. 13 shows the storage modulus G and the loss factor tan() of GAP-
165
N100 as function of temperature at deformation with 1 Hz at different ageing states. The

ageing shifts the glass transition to higher temperatures. This is achieved by additional crosslinking between neighbouring GAP chain. This explanation was discussed already in [12]. But
the quite strong ageing has already first effects on the strength in the rubbery region. The
value of G seems to be lower for the sample aged over 50 days in comparison to the one
aged over 2 days. This means a beginning chain cleavage of the binder in competition to the
strength increasing cross-linking. The behaviour of formulation -HMX/GAP-N100 is
shown in Fig. 14. The ageing temperature was the same as with GAP-N100, but the strong
ageing extends to even 80 days. There is no indication of chain splitting in the binder. The
two curves seem to coincide in the rubbery part. However the glass transition temperature is
shifted to relatively high temperatures, higher than to be expected with GAP-N100 alone.
This is indicative for an increased cross-linking in GAP-N100. The effect of ageing on the
formulation -CL20/GAP-N100 is presented in Fig. 15. The ageing temperature was only
80C. From Fig. 8 one knows that CGN ages at the same extend as HGN but at a
temperature difference of about 30C. The pure binder ageing is therefore not as strong as at
110C. However, the most striking result is the clearly stated decrease in binder strength as
shown by significant lowering of the storage modulus in the rubbery region compared to the
weakly aged and unaged samples.
1,E+10
2,5
torsion DMA
e-CL20 / GAP-N100
80C, 2d
G' [Pa]
1 Hz
1,E+09
G'-0.1Hz
1,E+08
G'-1Hz
G'-10Hz
tan(d) [-]
2
10 Hz
0.1 Hz
1,5
tan(d)-0.1Hz
1,E+07
tan(d)-1Hz
tan(d)-10Hz
1,E+06
1 Hz
10 Hz
1,E+05
0,5
T [C]
1,E+04
-60
-50
-40
-30
-20
-10
10
Fig 12. Storage shear modulus G and loss factor tan() of 45:55 formulation
-CL20/GAP-N100 as function of temperature and deformation aged
at 80C over 2 days.
Fig. 18, Fig. 19 and Fig. 20 show the storage shear modulus, the loss shear modulus and
the loss factor for all samples together to allow immediate comparison. A further result is
revealed with Fig. 20, in that the high temperature tailing of the loss factor shows a clear
separation between weakly and strongly aged samples. The coincidence of the tan() tails of
the two formulations states again the same ageing extent, but induced by a temperature
difference of 30C. The lowering of the height of tan() in going from unfilled GAP-N100 to
filled ones, especially to see with weakly aged samples, can be explained by the restriction in
polymer chain mobility, in part caused by binder-polymer interaction and in part by filling
166
up of free volume by the filler material. The restricted possibility for reorientation for the
polymer chains in going from the rubbery state to the glassy state reduces the extent of
energy consumed for these processes and therefore the height in tan(). A further aspect is
that the filler raises the elastic behaviour of the sample and this reduces the phase shift
between source of deformation and measurement of deformation.
1,E+10
2,5
G' [Pa]
tan(d) [-]
GAP-N100
deformation with 1 Hz
1,E+09
2
1,E+08
2d, 110C
1,E+07
50d, 110C
1,5
G'-2d, 110C
2d,
G'-50d, 110C
1,E+06
tan(d)-2d, 110C
tan(d)-50d, 110C
1,E+05
50d, 110C
0,5
1,E+04
T [C]
1,E+03
-60
-50
-40
-30
-20
-10
10
Fig 13. Storage shear modulus G and loss factor tan() of GAP-N100 as function
of temperature and ageing at deformation with 1 Hz.
1,E+10
2,5
G' [Pa]
b-HMX/GAP-N100
tan(d) [-]
G'-2d, 110C
1,E+09
G'-80d, 110C
tan(d)-2d, 110C
1,E+08
tan(d)-80d, 110C
2d, 110C
1,5
1,E+07
1
1,E+06
80d,
110C
0,5
1,E+05
T [C]
1,E+04
0
-60
-50
-40
-30
-20
-10
10
20
-HMX/GAP-N100 as function of temperature and ageing at deformation
with 1Hz.
167
2,5
1,E+10
G' [Pa]
e-CL20/GAP-N100
tan(d) [-]
G'-unaged
G'-2d, 80C
1,E+09
G'-60d, 80C
60d, 80C
tan(d)-unaged
2d, 80C
1,E+08
tan(d)-2d, 80C
1,5
tan(d)-60d, 80C
2d, 80C
1,E+07
2d, 80C
60d, 80C
1,E+06
60d, 80C
2d, 80C
T [C]
1,E+05
-60
-50
-40
-30
-20
0,5
-10
10
0
20
-CL20/GAP-N100 as function of temperature and ageing at deformation
with 1 Hz.
1,E+10
G' [Pa]
torsion DMA
storage shear modulus G'=f(T)
1,E+09
G'-GN-2d-110C
G'-GN-50d-110C
G'-HGN-2d-110C
G'-HGN-80d-110C
G'-CGN-2d-80C
G'-CGN-60d-80C
HGN, 80d, 110C

GN, 2d, 110C
1,E+08
CGN, 60d, 80C
GN, 50d,
1,E+07
CGN, 2d, 80C

HGN, 2d, 110C
1,E+06
CGN, 2d, 80C
GN, 50d, 110C
T [C]
1,E+05
-60
-50
-40
-30
-20
-10
10
20
Fig 16. Storage shear modulus G as function of temperature at deformation with

1 Hz of 45:55 formulations CGN, HGN and GAP-N100 (GN) aged
at indicated conditions.
168
1.E+09
G''-GN-2d-110C
G'' [Pa]
G''-GN-50d-110C
GN, 50d, 110C
G''-HGN-2d-110C
HGN, 80d, 110C
G''-HGN-80d-110C
1.E+08
G''-CGN-2d-80C
G''-CGN-60d-80C
GN, 2d, 110C
CGN, 60d, 80C
1.E+07
HGN, 2d, 110C

CGN, 2d, 80C
1.E+06
torsion DMA
loss shear modulus G''=f(T)
T [C]
1.E+05
-60
-50
-40
-30
-20
-10
10
20
Fig 17. Loss shear modulus G as function of temperature at deformation

with 1 Hz of 45:55 formulations CGN, HGN and GAP-N100 (GN) aged
at indicated conditions.
2.5
tan(d) [-]
torsion DMA
loss factor tan(d)=f(T)
GN, 2d, 110C
tan(d)-GN-2d-110C
tan(d)-GN-50d-110C
tan(d)-HGN-2d-110C
tan(d)-HGN-80d-110C
GN, 50d, 110C
1.5
HGN, 2d, 110C
tan(d)-CGN-60d-80C
CGN, 60d, 80C
CGN, 2d, 80C
tan(d)-CGN-2d-80C
HGN, 80d, 110C

0.5
HGN, 80d, 110C
T [C]
0
-60
-50
-40
-30
-20
-10
10
Fig 18. Loss factor tan() as function of temperature at deformation with 1 Hz

of 45:55 formulations CGN, HGN and GAP-N100 (GN) aged at indicated
conditions.
169
20
6.
DISCUSSION
The shape of the mass loss curves and of the excess reaction functions RML in Fig. 3
looks like autocatalytic decomposition behaviour, means here an autocatalytic excess
decomposition is revealed. The model ML: first order + autocatalytic [6] has been applied to
describe these data up to 4% RML. In Eq.(9) the principal autocatalytic reaction scheme is
shown. Eq.(10) describes the decomposition rate of component A, which is formulated for
masses in Eq.(11) with the corresponding symbols given in Eq.(12) using MAr(t,T) =
MA(t,T)/MA(0). The integration of Eq.(12) with the initial condition MAr(0) = 1 results in
Eq.(13). B stands for an autocatalytically effective decomposition product of A, C
symbolizes the decomposition gases and S all solid products. F considers already present B.
More about this modelling see [6].
k
1 B + C + Sintrinsic decomposition
A
2 2 B + C + S autocatalytic decomposition
A + B
(9)
dA ( t , T)
= k 1( T) A ( t , T) k 2 ( T) A ( t , T) B( t , T)
dt
T
(10)
dM Ar ( t , T)
2
= k 1ML ( T) M Ar ( t , T) k ML
( T) M Ar ( t , T) (F + 1 M Ar ( t , T)) (11)
dt
T
k 1ML ( T) = k 1( T)
M Ar ( t , T) =
and
2
k ML
( T) = k 2 ( T) A ( 0 )
and
F=
M B (0) m A
M A (0) m B
2
k 1ML ( T) + (F + 1) k ML
( T)
((
) )
2
2
2
k ML
( T) + k 1ML ( T) + F k ML
( T) exp k 1ML ( T) + (F + 1) k ML
( T) t
(12)
(13)
The general relation between the measurement quantity M(t,T) and the looked for
quantity MA(t,T) is given in Eq.(14). The symbol mi means molar mass of substance i. With
Eq.(15) and Eq.(16) the relations between mass loss ML and measured and looked for
quantities are given. O means an offset not caused by the considered decomposition.
M r ( t , T) =
m m B m S M A (0 )
M ( t , T)
= 1 A
(1 M Ar ( t , T))
M (0 )
mA
M (0 )
ML ( t , T) = O + (1 M r ( t , T)) 100%
ML ( t , T) = O +
(14)
(15)
mC
(1 M Ar ( t , T)) 100%
mA
with
m C = m A mB m S
(16)
The results of the evaluation of the excess reaction in the 1:1 mixture -CL20-GAP diol
are shown in Table 5 and in Fig. 21 as Arrhenius plots. The value 143 kJ/mol was found for
the activation energy of the autocatalytic part, which dominates the first order part at higher
conversions and temperatures. Fig. 22 shows the modelling of the mass loss data of the 1:1
170
mixture -CL20-GAP diol itself. The data are described very well, which can be seen from
Table 6 and from Fig. 23, which presents the Arrhenius plot of the two reaction rate
constants. The values of the two activation energies are comparable to the ones of the excess
reaction regarding the standard deviation limits of the values. This means, the decomposition
of -CL20 in contact with GAP dominates the reactivity and the decomposition of the single
substances has almost no effect on the overall decomposition of the mixture. The modelling
of the ML data of the formulation -CL20/GAP-N100 can be seen in Fig. 24. Also here the
description of the data is very good, see Table 7 and Fig. 25 for the Arrhenius plot. The
values of the activation energies differ from those of the 1:1 mixture. The first order reaction
has a higher value, from about 90 kJ/mol to 120 kJ/mol and the autocatalytic part shows a
decrease from 140 kJ/mol to 122 kJ/mol. In Table 8 all the data are compiled for
comparison. It also includes the Arrhenius parameters of the -CL20 alone. From this
comparison one states a marked reduction in the values of the decomposition of -CL20 in
contact with GAP. In part this can explain the higher burning rates of -CL20-GAP
formulations in comparison to -HMX-GAP formulations. In Table 9 the Arrhenius
parameters of the thermal decomposition of GAP-N100 are given determined from mass loss
in the temperature range 90C to 130C. Fig. 26 shows the Arrhenius plot. The data from the
linear region of mass loss increase have been taken only, means a mass loss range up to 1%
and 5%, depending on the temperature. With 150 kJ/mol GAP-N100 has a higher activation
energy than all the -CL20-GAP combinations. Therefore the acceleration of the thermal
decomposition is caused by the contact between -CL20 with GAP.
Table 5. Excess reaction rate constants of the excess thermal decomposition
of the 1:1 mixture by mass -CL20-GAP diol calculated with the reaction
kinetic model 'ML: first-order + autocatalytic' applied formally to the
isothermal reactivity function RML(t,T) up to RML = 4% with mC/mA =
1 and fixed offset value O.
k 1RML (first order)
2
k RML
(autocat.)
90
[1/d]
2.314 E-4 0.02 E-4
[1/d]
1.097 E-2 0.02 E-2
100
5.951 E-4 0.09 E-4
110
temperature
[C]
offset O
correl.
[%]
coefficient
0
0.99980
5.108 E-2 0.10 E-2
0.99924
1.025 E-3 0.03 E-3
1.363 E-1 0.05 E-1
0.99846
120
1.066 E-3 0.03 E-3
1.292 E-1 0.04 E-1
0.99839
EaiRML [kJ/mol]
86.8 5
7.556 E+8
8.878 0.7
0.9934
143.4 8.3
4.975 E+18
18.697 1.15
0.9933
ZiRML
ZiRML
[1/d]
lg(
[1/d])
correl.coeff.
171
1
Arrhenius plot
description of reactivity RML(t,T) up to 4%
with model 'ML: first order + autocatalytic'
Arrhenius parameters for the excess reaction
ln(kX [1/d])
0
autocatalytic part
-1
120C
-2
-3
-4
-5
Ea2 = 143 kJ/mol

Z2 = 4.975 E+18 1/d
110C
Ea1 = 87 kJ/mol
Z1 = 7.556 E+8 1/d
2
R = 0.9934
2
R = 0.9933
100C
90C
first order part
-6
-7
-8
-9
0,0025
1/T [1/K]
0,00255
0,0026
0,00265
0,0027
0,00275
0,0028
Fig 19. Evaluation of the reactivity functions RML(t,T) of the 1:1-mixtures

-CL20-GAP diol by the model ML: first order + autocatalytic.
The Arrhenius parameters of the formal assignment to first order
and utocatalytic reactivity between the components are given.
Fig 20. Isothermal mass loss data at several temperatures of the 1:1 mixture
by mass -CL20-GAP diol and their kinetic modelling with model ML:
first order + autocatalytic.
172
Table 6. Reaction rate constants of the thermal decomposition of the 1:1 mixture
by mass -CL20-GAP diol calculated with the reaction kinetic model 'ML:
first-order + autocatalytic' applied to the isothermal mass loss data
with mC/mA = 0.8 and fixed offset O.
k 1ML (first order)
2
k ML
(autocat.)
90
[1/d]
3.024 E-4 1.1 E-4
[1/d]
1.102 E-2 0.60 E-2
100
9.746 E-4 3.1 E-4
110
120
temperature
[C]
offset O
correl.
[%]
coefficient
0.157
0.99987
3.699 E-2 1.70 E-2
0.157
0.99864
1.647 E-3 6.5 E-4
1.096 E-1 0.43 E-1
0.157
0.99899
3.256 E-3 1.6 E-3
3.763 E-1 1.40 E-1
0.157
0.99848
91.1 10
4.457 E+9
9.649 1.4
0.976
138.6 3.5
9.226 E+17
17.965 0.50
0.9987
EaiML
ZiML
ZiML
[kJ/mol]
[1/d]
lg(
[1/d])
correl.coeff.
0
ln(kX[1/d])
Arrhenius plot
decomposition of 1:1 mixture e-CL20-GAP diol
description with model 'ML: first order + autocatalytic'
-1
120C
-2
110C
-3
-4
Ea 1 = 91 kJ/mol
Z1 = 4.457 E+9 1/d
-5
R 2 = 0.976
Ea 2 = 139 kJ/mol
Z2 = 9.226 E+17 1/d
100C
autocatalytic
part
R 2 = 0.9987
90C
first order part
-6
-7
-8
1/T [1/K]
-9
0,0025
0,00255
0,0026
0,00265
0,0027
0,00275
0,0028
Fig 21. Evaluation of the decomposition of the 1:1 mixture -CL20-GAP diol
by the model ML: first order + autocatalytic. The Arrhenius parameters
of the first order and autocatalytic reaction rate constants are given
.
173
Fig 22. Isothermal mass loss data at several temperatures of the 45:55 formulation
by mass -CL20/GAP-N100 and their kinetic modelling with model ML:
first order + autocatalytic.
Table 7. Reaction rate constants of the thermal decomposition of the 45:55 by mass
formulation -CL20/GAP-N100 calculated with the reaction kinetic model
'ML: first-order + autocatalytic' applied to the isothermal mass loss data
with mC/mA = 0.36 and fixed offset O.
k 1ML (first order)
2
k ML
(autocat.)
70
[1/d]
5.297 E-4 0.38 E-4
[1/d]
5.237 E-3 0.51 E-3
80
1.483 E-3 0.25 E-3
90
temperature
[C]
offset O
[%]
R2
0.15
0.99973
1.892 E-2 0.38 E-2
0.15
0.99912
4.720 E-3 0.82 E-3
6.262 E-2 1.25 E-2
0.15
0.99974
100
1.466 E-2 0.35 E-2
1.828 E-1 0.50 E-1
0.15
0.99977
110
4.117 E-2 1.02 E-2
4.433 E-1 1.40 E-1
0.15
0.99948
EaiML [kJ/mol]
120.2 2.6
9.505 E+14
14.978 0.38
0.9986
122.0 2.7
2.089 E+16
16.320 0.39
0.9985
ZiML
ZiML
[1/d]
lg(
[1/d])
correl.coeff.
174
0
Arrhenius plot
decomposition of 45:55 formulation e-CL20/GAP-N100
description with model 'ML: first order + autocatalytic'
ln(kX [1/d])
-1
110C
-2
autocatalytic part
100C
-3
90C
Ea 2 = 122 kJ/mol
Z2 = 2.09 E+16 1/d
-4
80C
-5
-6
Ea 1 = 120 kJ/mol
Z1 = 9.51 E+14 1/d
R 2 = 0.9986
R 2 = 0.9985
70C
first order part
-7
1/T [1/K]
-8
0,00255
0,0026
0,00265
0,0027
0,00275
0,0028
0,00285
0,0029
0,00295
0,003
Fig 23. Evaluation of the decomposition of the 45:55 formulation -CL20/GAPN100 by the model ML: first order + autocatalytic. The Arrhenius
parameters of the first order and autocatalytic reaction rate constants
are given.
175
Table 8. Arrhenius parameters of -CL20 alone, for the excess reaction of the 1:1
mixture -CL20-GAP diol and of the mixture itself and of the formulation
-CL20/GAP-N100. All data used for evaluation are isothermal mass loss
measurements.
[6]
-CL20 alone
therm. decomp.
160C to 185C
first ord.
autocat.
-CL20-GAP diol
therm. excess
reaction
90C to 120C
-CL20-GAP diol
therm. decomp.
90C to 120C
-CL20/GAPN100
therm. decomp.
70C to 110C
first ord. autocat. first ord. autocat. first ord. autocat
Eai [kJ/mol] 185.24.0 183.30.4
87 5
143 8
91 10
139 4
120 3
122 3
15.980.5 17.78.05
3.942
13.760
4.71
13.03
14.978
16.320
0.9934
0.9933
0.976
0.9987
0.9986
0.9985
lg(Zi [1/s])
2
0.99896
0.999989
0
ln(k[%/d])
GAP-N100
Arrhenius plot
description of mass loss data with reaction of zero order
ML range between 1% and 5%
-1
130C
-2
120C
-3
110C
-4
100C
Ea = 150.0 kJ/mol
lg(Z [%/d]) = 18.923
KK = 0.99917
-5
90C
-6
1/T [1/K]
-7
0.00245
0.0025
0.00255
0.0026
0.00265
0.0027
0.00275
0.0028
Fig 24. Evaluation of the thermal decomposition of GAP-N100 by the model ML:
zero order.
Table 9. Arrhenius parameters of GAP-N100 determined from the mass loss data.
For the modelling a reaction of zero order was used to describe the linear
part of the data in the beginning.
GAP-N100
thermal decomposition
90C to 130C
Ea [kJ/mol]
150.0 2.5
lg(Z [1/s])
11.986
lg(Z [%/d])
18.923 0.34
KK = R
0.99917
176
7.
SUMMARY AND CONCLUSION
The reactivity of the nitramine -CL20 with GAP diol and GAP cured with
oligoisocyanate N100 is higher than the one of the nitramine -HMX in these environments.
This was confirmed by measurements of mass loss, gas generation and adiabatic self heat
rate. They probe the samples in two different ways: by split off gas products and by reaction
heat. The results of both probings are the same. Mass loss is complementary to gas
generation but it has a normalization advantage compared to gas generation. The reactivity is
formulated as excess reaction or as excess conversion as function of time and temperature.
Additionally it was found that the reactivity of -CL20 with GAP-N100 is much increased
over that with GAP diol. Dynamic mechanical analysis in torsion mode of aged samples of
cured binder GAP-N100, 45:55 formulation -CL20/GAP-N100 (CGN) and 45:55
formulation -HMX/GAP-N100 (HGN) confirm this reactivity difference also. However
with this method a further result was found. By ageing the glass transition temperature shifts
to higher temperatures for all three sample types, but the shift is highest with HGN. With
CGN the binder is weakened at stages of stronger ageing indicated by the lowering of the
storage modulus G compared to the values of G of the unaged sample in the rubbery
region. With HGN the conclusion is that the cross-link density in the GAP-N100 is more
increased than for unfilled GAP-N100. The activation energies obtained by evaluation of
mass loss measurements of -CL20-GAP diol and -CL20/GAP-N100 are significantly
smaller compared to the one of thermal decomposition of -CL20 alone. The higher
reactivity of -CL20 with GAP compared with that of -HMX can be seen as one cause for
the higher burning rates of -CL20 based formulations. The results indicate that an
adjustment of the component mixture must be made to handle -CL20 in the presence of the
binder GAP-N100.
177
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
S. Eisele, K. Menke: About the Burning Behaviour of Smoke Reduced Composite Propellants
based on AP / CL20 / GAP, Proceed. 32nd Internat. Annual Conference of ICT, pages 149-1 to
149-18, July 3-6, 2001, Karlsruhe, Germany. Fraunhofer-Institut fr Chemische Technologie
(ICT), D-76318 Pfinztal-Berghausen, Germany.
Klaus Menke, Siegfried Eisele, Manfred A. Bohn, Peter Gerber: Minimum Smoke Propellants
with High Burning Rates and Thermodynamic Performance, RTA/RTO-AVT 089 Specialist's
Meeting on 'Advances in Rocket Propellant Performance, Life and Disposal for Improved
System Performance and Reduced Costs', Aalborg, Denmark, September 23 - 26, 2002.
Proceedings on CD, RTO-MP-091 (AVT-089), published October 2003.
M.A. Bohn: Stability and Ageing Assessment of Rocket Propellant Formulation Batches with
high Burn Rates containing CL20, Proceed. of the 3rd International Heat Flow Calorimetry
Symposium for Energetic Materials, April 8-11, 2002, French Lick, Indiana, USA, pages R1
to R15. Edited by James A. Wilson, US Navy Surface Warfare Center, Crane, Indiana, USA,
2002.
M.A. Bohn: Thermal Ageing of rocket propellant formulations containing -HNIW (-CL20)
investigated by heat generation rate and mass loss, Thermochim. Acta 401, 27-41 (2003).
S. Lbbecke, M.A. Bohn, A. Pfeil, H.H. Krause: Thermal Behavior and Stability of HNIW
(CL20), Paper 145 in Proceed. of the 29th Internat. Annual Conference of ICT 1998, pages
145-1 to 145-15, June 30-July 3, 1998, Karlsruhe, Germany. Fraunhofer-Institut fr Chemische
Technologie (ICT), D-76318 Pfinztal, Germany.
M.A. Bohn: Kinetic Description of Mass Loss Data for the Assessment of Stability,
Compatibility and Aging of Energetic Components and Formulations Exemplified with -CL20,
Propellants Explosives Pyrotechnics 27, 125-135 (2002).
M.A. Bohn, M. Drich, J. Aniol, H. Pontius, P.B. Kempa, V. Thome: Reactivity Between CL20 and GAP in Comparison to -HMX and GAP, Paper 4 in Proceed. of the 35th Internat.
Annual Conference of ICT 2004, pages 4-1 to 4-30, June 29-July 2, 2004, Karlsruhe, Germany.
Fraunhofer-Institut fr Chemische Technologie (ICT), D-76318 Pfinztal-Berghausen,
Germany. The support of my colleagues Peter Gerber and Dr. Klaus Menke in manufacturing
the formulations is much acknowledged.
[8] M.A. Bohn: Compatibility of Energetic Material Aspects of Measurement and Assessment
13th Jan-Hansson-Symposium on Chemical Problems Connected with the Stability of
Explosives, June 6 to 10, 2004, Bckaskog Slott, Kristianstad, Sweden. Proceeddings to be
published in 2005.
[9] M.A. Bohn: Determination of the kinetic data of the thermal decomposition of energetic
plasticizers and binders by adiabatic self heating.Proceed. of the 2nd International Heat Flow
Calorimetry Symposium of Energetic Materials (named at time Workshop on the
Microcalorimetry of Energetic Materials), May 17-19, 1999, Leeds (Otley), UK, pages F1 to
F25. Edited by T.T. Griffiths, DERA (now QinetiQ), Fort Halstead, Sevenoaks, UK.
[10] M.A. Bohn: Determination of the kinetic data of the thermal decomposition of energetic
plasticizers and binders by adiabatic self heating, Thermochim. Acta 337, 121-139 (1999).
[11] M.A. Bohn, P. Elsner: Aging of the Binders GAP-N100 and HTPB-IPDI Investigated by
Torsion-DMA, Propellants Explosives Pyrotechnics 24, 199-205 (1999)
[7]
178
DETONATION VELOCITY OF EMULSION EXPLOSIVES

WITH CENOSPHERES
A.G. Anshits*, N.N. Anshits*, A.A. Deribas**, S.M. Karakhanov**,
N.S. Kasatkina***, A.V. Plastinin**, A.Yu. Reshetnyak****, V.V. Silvestrov**
* Institute of Chemistry and Chemical Technology SB RAS, Krasnoyarsk
** Lavrentyev Institute of Hydrodynamics SB RAS Novosibirsk
*** Boreskov Institute of Catalysis SB RAS Novosibirsk
**** Institute of Theoretical and Applied Mechanics SB RAS Novosibirsk
Abstract:
The velocity of detonation of emulsion explosive with microspheres from coal ash
(cenospheres) was measured. The diameter of cenospheres was changed from 50 to 250
microns. The comparison of the dependence of detonation velocity vs. density and
diameter of charge with different sensitizers was carried out. The maximal detonation
velocity for cenospheres 70 -100 micron fraction was determined to be 5.5 5.6 km/s.
This value of detonation velocity is equal to the maximal detonation velocity determined
for the charges of emulsion with glass microballoons produced by 3M Company [1].
The critical diameter for the charge with cenospheres was determined as 35 40 mm to
be 1.5 2 times more than that of for the charge with 3M microballoons.
Keywords:
1.
emulsion explosives, coal ash
EMULSION
We used the emulsion of typical composition: industrial oil 6.9% volume, water 15.2%,
emulsifier 1%, ammonium nitrate
balance. We used the standard
procedure for obtaining of emulsion [1].
The structure of emulsion is shown in
Fig.1. It is possible to see that the
droplets sizes of the water solution of
ammonium nitrate are in the interval 5 50 microns. The microspheres of
sensitizers were add to hot emulsion (85
90 degree centigrades), the
concentration of sensitizer was changed
from 7% to 32% volume (from 1% to
18% weight). The density of emulsion
without sensitixer by itself was 1.35
g/cc. The densities of charges varied
Fig 1. Emulsion.
from 1.14 to 1.30 g/cc in dependence of
the volume of sensitizers.
179
2.
CENOSPHERES
The fractional composition of ash concentrate from power station No. 5, Novosibirsk,
with bulk density 0.38- 0.4 g/cc is shown in Fig.2. It is possible to see (curve1), that the
maximum of distribution of particles is situated in the interval of cenospheres diameters 100
-150 microns, about 85% of particles are of 100 300 microns in sizes. There are about 5 %
(weight) of particles with the sizes less than 100 microns and less than 10% (weight)
particles are of the sizes more than 300 microns. The floatation of ash in water showed that
about 0.5% (weight) of particles has the density more than 1 g/cc. The main part of the
particles of ash cenospheres has the form close to spherical one.
Fig 2. Fraction distribution for ash concentrate (1) and for glass microballoons
3M Company (2).
The curve of fractional distribution for glass microballoons produced by 3M Company
as received is shown in Fig.2 for comparison (curve 2). The bulk density of them is 0.12
g/cc, the maximum is near to 35 70 microns. The thickness of shells for 3M microballoons
is about 0.3 0.6 microns and the same parameter for cenospheres is in the interval 2.5 7.6
microns.
3.
DETONATION VELOCITY
Cylindrical charges of emulsion explosive 55 mm in diameter in plastic tubes 1 mm wall

thickness were used in our experiments. The length of it was mainly 350 mm. The booster
was used for the ignition of detonation (Comp B, d = 55 mm, L = d/2). The detonation
velocity was measured at 120 mm from the end of charge by using of four contact wires. The
accuracy of the measuring of velocity was about 1%. The charges of the diameters less than
55 mm were used at the determination of critical diameters of detonation. It was made in
plastic tubes with the wall thickness 0.5 1 mm or in paper tubes. The length of it was 10
15 of diameters.
180
The dependence of detonation velocity vs. charge density is shown in Fig.3 It is possible
to see that the maximal detonation velocity for charges with cenospheres 70 100 microns
in size (3) is equal to that of for charges with microballoons produced by 3M Company (2).
The value of this maximum is 5.5 5.6 km/s. The density of charges in the area of this
maximum is about 1.18 1.20 g/cc. The failure of detonation was observed at the density of
charges about 1.28 1.29 g/cc. Calculations showed that near the maximum of detonation
velocity the weight of cenospheres must be 4 6 times more than that of for glass
microballoons produced by 3M Company. The quantity of cenospheres provided the
maximal detonation velocity must be about 1.7 times more in comparison with glass
microballoons. Therefore increasing of the wall thickness of microspheres demand to
increase the quantity of hot spots which are necessary to support the detonation in
emulsion explosive. The experiments showed that increasing of the wall thickness onto 10%
decreased the value of detonation velocity on about 0.5 km/sec.
Fig 3. The dependence of detonation
velocity vs. charge density.
1 Lee et al. [1]
charge 52.5 mm,
glass microballoons 3M.
2 6: our data for charge 55 mm:
2 glass microballoons 3M,
3 cenospheres 70100 m,
6 ash concentrate without fracturing.
The data on detonation velocity vs. density for our charge with 3M microballoons and
the data for approximately the same charge (d=52.5 mm) with the same microballoons
described in [1] are presented here. The maxima of the detonation velocities in both cases at
the densities 1.14 1.26 are close to each other despite of the difference in size of AN
droplets in emulsion. But the critical density which determines the failure of detonation
depends probably from the sizes of emulsion droplets (critical density 1.26 in [1] and that of
1.24 in our experiments).
We have studied the influence of the sizes of microspheres on the detonation velocity.
These curves are shown in Fig.4. The maximum of detonation velocity for cenospheres was
determined for particles 70 100 microns. The relationship the wall thicknesses vs.
diameters for these cenospheres is about 0.04. The experiments with cenospheres with the
relationship wall thickness vs. diameter to be equal 0.067 (bulk density 0.46 g/cc) showed
that detonation is failed and so the thick-walled cenospheres cannot to generate the hot
spots to be sufficient to support the detonation (arrow in Fig.4).
181
Fig 4. The dependence of detonation

velocity vs. the sizes
of microspheres.
Our data for charge 55 mm:
1 cenospheres;
2 glass microballoons 3M.
Data for glass microballoons
3M:
3 30 mm [2],
4 23.6 mm [3].
4.
THE CRITICAL DIAMETER OF CHARGES
The data on the influence of the diameter of charge on the detonation velocity is shown
in Fig.5. The critical diameter of our charges with optimal quantity of glass microballoons
3M was determined to be in the interval 20 22 mm (19 mm no detonation, 24 mm
detonation). In paper [1] the critical diameter at the same condition approximately has been
determined as 14 mm. We assume that the reason of this difference is the size of droplets of
ammonium nitrate solution (1-5 microns vs. 10 -15 microns). The value of critical diameters
for cenospheres 70 -100 microns was determined as 35 40 mm. The critical diameter for
the cenospheres as received was determined as 40 -45 mm.
The detonation velocity for the charges to be close to critical diameter was determined to
be 3.7 4.1 km/s. The same value of detonation velocity was determine in the charges close
to the critical value of density. It is possible to assume that the reason of the detonation
failure is the same in both cases.
5.
CONCLUSION
Our investigations showed that the cenospheres from coal ash can be used as sensitizer
for emulsion explosives. The maximal detonation velocity can be obtained by using the
fraction of cenospheres 70 -100 m with bulk density to be 0.36 g/cc and charge density
1.18 1.20 g/cc. The value of maximal velocity is 5.5 5.6 km/s at the charge diameter 55
mm. The volume concentration of cenosphere must be 17 20%, the weight concentrationmust be 8 10%. The critical diameter for emulsion of studied composition is about 35 -40
mm.
182
Fig 5.
The dependence of detonation velocity vs. charge diameter.
Sensitizer
Legend
glass microballoons 3M
1 our data, 2 [1], 3 [4]
gas bubbles
4 poremit P-1 [5]
cenospheres (m):
5 70100
6 80180
7 70250
ash concentrate
REFERENCES
[1]
[2]
[3]
[4]
[5]
LEE J., SANDSTROM F. W., CRAIG B. G., PERSSON P.-A.: Detonation and shock initiation
properties of emulsion explosives // Proc. 9th Int. Symposium on Detonation, 1989. P. 263-271.
LEE J., PERSSON P.-A. Detonation behavior of emulsion explosive // Propellants, Explosives,
Pyrotechnics. 1990. V. 15. P. 208-216.
CHAUDHRI M.M., ALMGREN L.-A., PERSSON A.: Detonation behavior of a water-in-oil type
emulsion explosive containing glass microballoons of selected sizes // Proc. 10th Int.
Symposium on Detonation, 1993..
SHVEDOV K.K., LAVROV V.V.: Parametry detonaciji emulsionnyh VV. Problemy vzryvnogo
dela. Sbornik statey I dokladov. Izd.MGTU 2002 no.1 s.90 -99 (in Russian)
SOSNIN V.A., KOLGANOV E.V.: Issledovanive .processa detonacii v emulsionnyh
promyshlennyh vzryvchatyh veshestvah. Trudy Meshdunarodnoi konferencii Kharitonovskije
tematicheskije naychnye Chtenija 17 -21 March 2003 Sarov Vniief 2003 c.288 -297 (In
Russian )
183
SYNTHESIS AND CHARACTERISATION

OF 2,2-DINITRO-1,3-PROPANEDIOL-BASED PLASTICISERS
Stefan Ek*, Carina Eldster*, Patrick Goede*, Erik Holmgren*, Rolf Tryman*,
Nikolaj Latypov*, Yang Guo Ying Raymond**, and Lee Yiew Wang**
* FOI, Swedish Defence Research Agency, Department of Energetic Materials,

S-147 25 Tumba, Sweden
** DSO National Laboratories, 20 Science Park, Singapore 118230
Abstract:
In this paper, the synthesis of two energetic plasticisers (2,2-dinitro-1,3-bis(2-azido
acetoxy) propane and 2,2-dinitro-1,3-bis(formyloxy)propane) and the attempted
syntheses of even further derivatives are described. The prepared compounds were
characterised and evaluated as plasticisers. Their glass transition temperatures are
acceptable to excellent, but both of them are thermally unstable.
Keywords:
1.
energetic material, energetic plasticiser, synthesis
BACKGROUND
One of the key steps in the improvement of explosive charges and propellants is the
development of new energetic ingredients. The aim of such additives is to enhance the
performance as well as the mechanical properties of propellants and plastic bonded
explosives (PBX) in comparison with the current binders and/or plasticisers. The reason for
our interest in energetic plasticisers is the dissatisfactory mechanical properties of
commercially available, energetic binders. The polar groups in the molecular structure of
these compounds render them increasingly viscous and elevate their glass transition
temperatures. The rise in glass transition temperature downgrades the low temperature
properties, which are especially important for missile propellants. The higher viscosity can
result in processibility problems.
The inert binders currently in use have excellent mechanical properties. However, they
contain little energy and requires large amount of fillers to have a sufficient performance.
The disadvantages of available energetic polymers are mentioned above. Hitherto, the
known energetic plasticisers have inconveniences such as low thermal stability, low energy
content, high migratory ability and sometimes, they dissolve the filler. The low thermal
stability can be remedied with stabilisers, but finding a molecule that is stable as such is a
very interesting area of research.
Recently, attention has been directed towards gem-dinitro-based plasticisers [1]. Great
effort has been put into industrial-scale production of BDNPF/A [2,3] and to develop similar
mixtures containing formals of other gem-dinitro alcohols [4,5]. The research of the last
decade has been focused mostly on azido-plasticisers [1]. These considerations focused our
efforts on structures with energetically derivatised gem-dinitro compounds. Our earlier
preparation of the interesting but thermally labile energetic plasticiser 2,2-dinitro-1,3-bis-
184
nitrooxy-propane [6] (NPN). The properties of this compound prompted us to prepare other
2,2-dinitro-1,3-propanediol-based compounds.
2.
CALCULATIONS
Our first criterion in the selection of new molecules to synthesis is the result from the
calculations of their heat of formation, density and performance. The heat of formation in the
gaseous state is quantum mechanically calculated by the use of Gaussian (G98) [7]. The heat
of formation in the solid state is then calculated with a code developed by professor Tore
Brinck at the Royal Institute of Technology in Stockholm. This code is built on models
developed by Rice [8] and Politzer [9] MOLPAK [10] was used to estimate the density. Cheetah
2.0 is used to estimate the performance of the new molecules. The results were promising for
the selected substances, as shown in Table 1. Our previously published [6] plasticiser NPN
and butyl-NENA are in table for comparison.
Table 1. Calculated properties of the energetic plasticisers
Structure
H0f
(kcal/
mol)
(g/cm3
)
DCJ
(km/s)
PCJ
(GPa)
ISP
(s)
OB
(%)
-75
1.90
8.7
31.7
248
+12.5
-183
1.66
6.9
18.4
200
-36.0
-68
1.46
6.6
15.0
201
-48.2
111
1.62
8.3
27.8
265
-29.6
-371
1.73
6.1
16.0
211
-22.3
-71
1.42
6.4
13.8
205
-104.2
NO2
O2NO
ONO2
NO2
O
H
O
O
O 2N
O
NO2
O
N3
H
O
O
O 2N
N3
O
NO2
NO2
N3
N3
NO2
O
F3C
O
O
O2N
O2N
CF3
NO2
ONO2
To calculate the enthalpy of formation of fluorine-containing compounds, the calculation

code had to be slightly modified, since the work of Rice [8] and Politzer [9] does not include
any other elements than carbon, hydrogen, nitrogen and oxygen.
185
3.
EXPERIMENTS
3.1
Synthesis
3.1.1. 2,2-Dinitropropanediol and dinitromethane salts

The parent compound of the above-mentioned energetic plasticiser is 2,2-dinitro -1,3propanediol (1). This is synthesised by a so-called Henry reaction between an alkali metal
salt of dinitromethane and formaldehyde[11], cf. Scheme 1.
O 2N
NO2
O
M +
O 2N
H2O/AcOH
HO
OH
NO2
1
Scheme 1. Synthesis of 2,2-dinitropropanediol

This compound (1) can also be synthesised by the Henry reaction between nitromethane
and formaldehyde, followed by oxidation of the product with sodium persulphate [12]. Our
reason to choose the synthesis shown above, Scheme 1, was our experience in the area of
gem-dinitro chemistry. There were two main sources of dinitromethane salts: nitration of
barbituric acids [13] (2) and hydrolysis of 1,1-diamino-2,2-dinitroethylene [14] (FOX-7, 3), cf.
Scheme 2. The experimental section [14] of the paper describing the hydrolysis of FOX-7
claims that the hydrolysis is complete in three hours. Further studies have shown that twelve
to sixteen hours are required for the reaction to reach completion.
O
HN
O
HNO3/H2SO4
O
HN
2
HN
NO2
HN
NO2
O
MOH
H2 N
NO2
H2 N
NO2
MOH
O 2N
M
O 2N
Scheme 2. Synthesis of dinitromethane salts

3.1.2. 2,2-Dinitro-1,3-bis(formyloxy)propane
2,2-Dinitro-1,3-bis(formyloxy)propane (4) was synthesised according to a literature
procedure [15], cf. Scheme 3. 2,2-Dinitro-1,3-propanediol (1) is reacted with formic acetic
anhydride in a basic, organic solution. One inconvenience of this synthesis is that the
anhydride must be freshly prepared.
186
NO2
OH +
HO
Py, DCM
NO2
O
O2N
O
4
NO2
Scheme 3. Synthesis of 2,2-dinitro-1,3-bis(formyloxy)propane

3.1.3. 2,2-Dinitro-1,3-bis(2-azidoacetoxy)propane
The hitherto unpublished 2,2-dinitro-1,3-bis(2-azidoacetoxy)propane (6) was synthesised
according to the procedure described in Scheme 4. 2,2-dinitro1,3-propanediol (1) is treated
with chloroacetyl chloride to yield chloro-acetic acid 2,2-dinitro-1,3-bis(2-chloro acetoxy)
propane (5). Chloroacetic anhydride can also be used. Substitution of the chlorines with
sodium azide produces the desired product (6).
NO2
HO
OH
Cl
Cl
NO2
1
Py, DCM
Cl
O
O 2N
O
NaN3
N3
O
Cl
O
5
NO2
O
O
O2N
N3
O
6
NO2
Scheme 4. Synthesis of 2,2-dinitro-1,3-bis(2-azidoacetoxy)propane

3.1.4. Attempted synthesis of 2,2-dinitro-1,3-bis(trifluoroacetoxy)propane
When 2,2-dinitro-1,3-propanediol (1) was treated with triflic anhydride, under the same
reaction conditions as with formic acetic anhydride, no reaction occurred. No product (7)
was obtained, despite various attempts.
NO2
HO
NO2
OH +
F3C
O
O
Py, DCM
F3C
CF3
O
O 2N
O
NO2
CF3
Scheme 5. Attempted synthesis of 2,2-dinitro-1,3-bis(trifluoroacetoxy)propane

3.1.5. Attempted synthesis of 2,2-dinitro-1,3-diazidopropane
The synthesis of 2,2-dinitro-1,3-diazidopropane (10) was attempted in two different
ways, cf. Scheme 6. In the first 2,2-dinitro-1,3-propanediol (1) is transformed into 2,2dinitro-1,3-dichloropropane (8), by the use of thionyl chloride and pyridine. The second way
is the nitration of the same parent compound (1) in mixed acids into 2,2-dinitro-1,3bis(nitrooxy)propane [6] (10). Experience in our group prompted us to try nitrate esters as
leaving groups. The two strategies share the same problem. Both intermediates are
decomposed, presumably by proton abstraction by the azide ion, the nitrate ester faster (9)
than the chloride (8).
187
NO2
Cl
SOCl2/Py
NaN3
NO2
8
NO2
HO
Cl
NO2
OH
NO2
1
HNO3/
H2SO4
N3
NO2
O2NO
ONO2
NaN3
N3
NO2
10
NO2
9
Scheme 6. Attempted syntheses of 2,2-dinitro-1,3-diazopropane

It later became known to us that 2,2-dinitro-1,3-diazidopropane (10) has been
synthesised [16]. It is very sensitive to impact and friction and has been tested as a primary
explosive [17, 18], a fact that dissuaded us from further attempts of preparation.
3.2
Characterisation
Our group has selected two criteria to evaluate potential plasticisers. The first is the glass
transition temperature of the pure compound, which was measured using a Mettler DSC 30,
differential scanning calorimeter, equipped with a ceramic sensor. Every test as well as the
calibration was performed in a nitrogen atmosphere (50 ml/min) at a heating rate of
10 C/min. The samples were put in 40 l aluminium cups with a pierced lid. The sample
weight was between 10 and 20 mg. Calibration was made with indium (Mettler-Toledo,
Sweden) (Tm= 156.6 C) and anhydrous n-octane (Sigma-Aldrich, Sweden) (Tm= -57 C).
All tests were performed in duplicate.
The second is the thermal stability as measured by heat flow calorimetry. STANAG 4582
describes the NATO requirements on the stability of explosives and nitrocellulose-based
propellants [19]. This procedure specifies that a sample should not at any time have a higher
heat-flow than 63.1 W/g, being kept at 75 C for 19 days, or higher than 114 W/g at 80 C
for 11 days to be considered as stable. Approximately 0.5 grams of each sample was used.
The measurements were done isothermally in a Thermal Activity Monitor, TAM 2277, heat
flow (micro) calorimeter. The samples were put in 3 ml glass ampoules and hermetically
sealed with aluminium caps. If a new product is stable, according to these conditions, further
evaluation of its compatibility with different fillers and binders are performed.
A first test of the thermal stability is a so-called block test, which is a measure of precaution
to avoid damaged microcalorimeter cells. The sample is stored in an ordinary, metal heating
block for 11 days at 80 C. If no change in colour and/or gas evolution is observed, a new
sample will be put into the heat flow calorimeter for a more thorough measurement.
3.2.6. 2,2-Dinitro-1,3-bis(formyloxy)propane
This compound (4, Scheme 3) is an excellent plasticiser in terms of its glass transition
temperature, which is 77 C, cf. Figure 1. A sample of this product turned black without
any larger gas evolution in a block test, vide supra. Therefore, no further heat flow
calorimetric evaluation of this molecule was performed.
188
Fig 1.
Glass transition temperature measurement

of 2,2-dinitro-1,3-bis(formyloxy)propane
3.2.7. 2,2-Dinitro-1,3-bis(2-azidoacetoxy)propane
This compound (6, Scheme 4) has an acceptable glass transition temperature of 49 C, cf.
Figure 2. Though, it has low thermal stability, as shown in Figure 3. The heat-flow peaks at
approximately 3300 W/g, which is more than fifty times higher than the acceptance limit of
STANAG 4582 [19] (63 W/g).
Fig 2.
Glass transition temperature measurement

of 2,2-dinitro-1,3-bis(2-azidoacetoxy)propane
189
Fig 3.
4.
Heat flow produced by 2,2-dinitro-1,3-bis(2-azidoacetoxy)propane at 75C,

measured by calorimetry
DISCUSSION
4.3
Calculation vs. observed values
There are discrepancies between the calculated and the observed properties of the
plasticisers, as shown in Table 2. A major problem when performing calculations on
potential plasticisers is that they generally are liquids. Our codes are adapted to estimate the
properties of gases and solids.
The heat of formation is first calculated in the gas phase. The heat of sublimation is then
calculated to enable the conversion from H0f(g) to H0f(s). To obtain the heat of formation
of a liquid, the heat of evaporation is needed, instead of the heat of sublimation. This can
explain the discrepancies in this area. The enthalpy of formation has not been measured for
any of the new products. Therefore, butyl-NENA is added for comparison.
A similar problem is encountered when calculating the density. MOLPAK [10] assumes
that the substance is crystalline to be able to calculate the best packing of the molecules.
190
Table 2.
The calculated and observed heats of formation and densities

of some plasticisers
H0f
(kcal/
mol)
Structure
(g/cm3)
Calc.
Obs.
Calc.
Obs.
-75
n.m.
1.90
1.661
-183
n.m.
1.66
1.47
-68
n.m.
1.46
1.50
-71
-46 [20]
1.42
1.22
NO2
O2NO
ONO2
NO2
O
H
O
O
O 2N
O
NO2
O
N3
O
O
O 2N
O2N
N3
NO2
ONO2
4.4
Synthesis
It is noteworthy that 2,2-dinitro propanediol (1) is stable to pyridine in dichloromethane,

whereas it decomposes into 2,2-dinitroethanol (11) and formaldehyde in basic, aqueous
solutions [11], as shown below.
NO2
HO
OH
NO2
1
NO2
OH-/H2O
O2N
O
OH +
11
Scheme 7. Decomposition of 2,2-dinitro-1,3-propanediol (1) in basic,

aqueous solution
The reaction of the diol with chloroacetyl chloride, formic acetic anhydride and chloro
acetic anhydride in the presence of pyridine was successful. In the case of trifluoroacetic
anhydride no product was obtained. Our hypothesis is that the equilibrium in basic, organic
solution is unfavourable for the product.
The problems encountered in the attempted syntheses of 2,2-dinitro-1,3-diazidopropane
can be explained by the instability of the intermediates in even slightly basic conditions. 2,2Dinitro-1,3-dichloropropane can undergo a retro-Henry reaction, by loosing formaldehyde
and hydrochloric acid. 2,2-Dinitro-1,3-bis(nitrooxy)propane is very sensitive to bases. Its
protons are supposed to be easily abstracted, despite the lack of pKa measurements. This
191
assumption is based on the observation of oligomerisation in basic media. The formation of

oligomers was observed by mass spectrometry (not shown in this paper), but their structures
were not determined.
Analogous results from these two compounds and NPN indicate that 2,2-dinitro-1,3diol-based compounds have low glass transition temperatures, but they seem to be thermally
unstable. The close proximity of the electron withdrawing groups in these structures might
be a reason for their low stabilities. These results made us discard 2,2-dinitro-1,3propanediol as a core for energetic plasticisers. The effects of extended carbon chains on the
glass transition temperature and thermal stability are under current investigation.
Acknowledgements
We gratefully acknowledge the financial support from DSTA in Singapore.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
A. PROVATAS: Energetic Polymers and Plasticisers for Explosive Formulations. A Review of

Recent Advances, 2000, DSTO Aeronautical and Maritime Research Laboratory
H. G. ADOLPH, J. M. GOLDWASSER, W. M. KOPPES: Synthesis of Polyformals from Nitro- and
Fluorodiols. Substituent and Chainlength Effects., J. Polym. Sc. 1987, 25, 805-822
R. B. WARDLE, R. S. HAMILTON, M. GESLIN, V. MANCINI, D. MERRILL: An Environmentally
Favorable Continous Process for the Synthesis of BDNPF/A, 30th International Annual
Conference of ICT, June 1999, Karlsruhe June 1999
J. R. CHO, J. S. KIM, K. D. LEE, Y. G. CHEUN, B. S. PARK: Patent US2001-808379, 2001,
J. R. CHO, J. S. KIM, K. D. LEE, B. S. PARK: An Improved Mixed Formal Energetic Plasticizer,
1999 Insensitive Munitions and Energetic Materials Symposium, Tampa 1999, p.404-412
N. WINGBORG, C. ELDSTER: 2,2-Dinitro-1,3-Bis-Nitrooxy-Propane (NPN): A New Energetic
Plasticizer, Propellants, Explosives, Pyrotechnics 2002, 27, 314-319
M. J. FRISCH, H. B. S. G. W. TRUCKS, G. E. SCUSERIA, M. A. ROBB, V. G. Z. J. R.
CHEESEMAN, J. J. A. MONTGOMERY, J. C. B. R. E. STRATMANN, S. DAPPRICH, J. M. MILLAM,
A. D. DANIELS, M. C. S. K. N. KUDIN, O. FARKAS, J. TOMASI, V. BARONE, M. COSSI, R.
CAMMI, C. P. B. MENNUCCI, C. ADAMO, S. CLIFFORD, J. OCHTERSKI, G. A. PETERSSON, Q. C.
P. Y. AYALA, K. MOROKUMA, D. K. MALICK, A. D. RABUCK, K. RAGHAVACHARI, J. C. J. B.
FORESMAN, J. V. ORTIZ, B. B. STEFANOV, G. LIU, A. LIASHENKO, I. K. P. PISKORZ, R.
GOMPERTS, R. L. MARTIN, D. J. FOX, T. KEITH, C. Y. P. M. A. AL-LAHAM, A.
NANAYAKKARA, C. GONZALEZ, M. CHALLACOMBE, B. J. P. M. W. GILL, W. CHEN, M. W.
WONG, J. L. ANDRES, C. GONZALEZ, E. S. R. M. HEAD-GORDON, J. A. POPLE: Gaussian.
,Gaussian Inc., 1998
B. M. RICE, S. V. PAI, J. HARE: Predicting Heats of Formation of Energetic Materials Using
Quantum Mechanical Calculations, Combustion and Flame 1999, 118, 445-458
192
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
P. POLITZER, J. S. MURRAY, M. E. GRICE, M. DESALVO, E. MILLER: Calculation of heats of

sublimation and solid phase heats of formation, Molecular Physics 1997, 91, 923-928
J. R. HOLDEN, Z. DU, H. L. AMMON: Prediction of Possible Crystal Structures for C-, H-, N-,
O-, and F-Containing Organic Compounds, Journal of Computational Chemistry 1993, 14,
422-437
H. FEUER, G. B. BACHMAN, J. P. KISPERSKY: A New Preparation of Potassium
Dinitromethane and its Conversion to 2,2-Dinitro-1,3-propanediol, J. Am. Chem. Soc. 1951,
73, 1360
L. C. GARVER, V. GRAKAUSKAS, K. BAUM: Catalyzed Oxidative Nitration of Nitronate Salts,
Journal of Organic Chemistry 1985, 50, 1699-1702
A. LANGLET, N. V. LATYPOV, U. WELLMAR, U. BEMM, P. GOEDE, J. BERGMAN, I. ROMERO:
Nitration of 2-Substituted Pyrimidine-4,6-diones, Structure and Reactivity of 5,5-gemDinitropyrimidine-4,6-diones, J. Org. Chem. 2002, 67, 7833-7838
C. SANDBERG, P. GOEDE, N. V. LATYPOV, A. BELLAMY: Acid-Base Characteristics of FOX-7
and its Monohydrazo Analogue, New Trends In Research of Energetic Materials, Pardubice
2002
T. K. HIGHSMITH, D. W. DOLL, L. F. CANNIZZO: Energetic Plasticizer, and Explosive and
Propellant Composition Containing Same, Patent US 6,425,966 2002
P. R. DAVE, R. DUDDU, K. YANG, R. DAMAVARAPU, N. GELBER, R. SURAPANENI,
R. GILARDI: Preparatio of 'cage molecule' based polyazido core units for dendrimer synthesis,
Tet. Lett. 2004, 45, 2159-2162
S. P. VELARDE: Thiokol, Personal communications, 2003
J. HIRLINGER: Investigating Alternative 'Green' Primary Explosives, 39th Annual Guns &
Ammunition Conference, Baltimore, MD, USA 2004
M. A. f. S. (MAS), STANAG 4582 (First Draft) - Explosives, Nitrocellulose based propellants,
stability test procedure and requirements using heat flow calorimetry, in: (2002)
R. L. SIMMONS: Thermochemistry of NENA Plasticizers, 25th Annual Conference of ICT,
Karlsruhe 1994
193

G. Fischer*, G. Holl**, T. M. Klaptke*, P. Mayer*, J. J. Weigand*
* Contribution from the Chair of Inorganic Chemistry,
Ludwig-Maximilian University of Munich, D-81377 Munich, Germany
** Bundeswehr Research Institute for Materials, Fuels and Lubricants,
Swisttal-Heimerzheim, Groes Cent, D-53913 Swisttal, Germany
Abstract
The reaction of 1,5-diamino-1H-tetrazole (1) with iodomethane followed by the
methathesis of the iodide (2) with silver dinitramide yields the highly friction sensitive
1,5-diamino-4-methyltetrazolium dinitramide (3). 3 crystallize in the orthorhombic
space groups P212121 and shows an intriguing interaction of one nitro group of the
dinitramide anion with the tetrazolium cation. Preliminary sensitivity testing of the
crystalline compound 3 indicates rather low impact sensitivity with a value of 7 J in
contrast to the relatively high friction sensitivity of 24 N. The thermal decomposition of
3 was investigated by thermogravimetric analysis (TGA) and differential scanning
calorimetry (DSC). Mass spectrometry and IR spectroscopy were used to identify the
gaseous products.
1.
INTRODUCTION
Since the generation of molecular nitrogen as an end-product of a propulsion or

explosion is highly desired in order to avoid environmental pollution, health risks as well as
the reduction of detectable plume signatures, compounds containing a backbone of directly
linked nitrogen atoms (nitrogen catenation) are of great interest [1]. The ability of the
dinitramide (DN) anion to form stable oxygen-rich salts with high densities with a variety of
cations [2] makes DN a promising candidate in the development of new tetrazole based
oxidizers. Such salts would be expected to possess both a high nitrogen and a high oxygen
content. The combination of the oxygen-rich anion (DN) with a nitrogen-rich cation forms a
class of energetic materials whose energy is derived from their very high positive heat of
formation (directly attributed to the large number of inherently energetic NN, NO and N
C bonds) as well as the combustion of the carbon backbone.
Here we wish to report (i) the synthesis of the highly friction sensitive 1,5-diamino-4methyl-1H-tetrazolium dinitramide (3), (ii) the X-ray determination with respect to an
intriguing interaction of one nitro group with the tetrazolium cation and (iii) a study on the
thermal decomposition behavior of 3 [3].
194
2.
SYNTHESIS
Scheme 1.
The two step synthesis of 3 starts from 1,5-diamino-1H-tetrazole (DAT, 1) [4a] via
quaternization of 1 with MeI in acetonitrile (2, 86%). The metathetical reaction of 2 with a
freshly prepared solution of AgDN [5] yields 3 as light yellow crystals suitable for X-ray
structure determination (85%) (Scheme 1).
3.
CRYSTAL STRUCTURE
3 crystallizes in the orthorhombic space group P212121 with four formula units in the
unit cell as racemic twin. The structure contains one crystallographically independent cation
and one anion. As the methyl group hardly influences the molecular parameter of the
tetrazole moiety (MeDAT) we abstain from a detailed discussion. (cf. X-ray data of 1,5diamino-1H-tetrazolium cation see [4b]). The overall geometry of the dinitramide ion is
similar to that observed for other comparable dinitramide salts, like guanidinium and
biguanidinium salts [2]. The two NN bond lengths are asymmetric (N7N8 1.350(3) ; N8
N9 1.391(3) ; cf. a typical N=N double bond (1.245 ); NN single bond (1.454 )) [6]
and a NNN angles of 115.4(2) was observed. The nitro groups are twisted out of the central
NNN plane (O3N9N7O1 -24.5(2)). Hence, the local symmetry of the anion is C1. As
known also form others [2] the metrical parameters of the DN ion are easily and strongly
influenced by the environment.
Fig 1.
View of the strand formed by alternating cations (MeDAT) and anions (DN)
displayed along the a-axis in 3. Short contacts between the cations and anions
are shown by dotted (, N5H5bO2) and dashed (- - - , O(dinitramide) and
tetrazole moiety) lines, respectively. (ORTEP Plot, thermal ellipsoid represents
50% probability).
195
In Figure 1 the view along the a-axis is displayed showing the zigzag strand composed
of alternating MeDAT and DN ions. Each strand is further connected by strong hydrogen
bonds (N5H5aO1i and N5H5aO2i [symmetry code: (i) -1+x, y, z]) forming a threedimensional network. In turn, these strands are formed by two different type of interactions:
(1) a strong hydrogen bond between N5H5bO2ii (symmetry code: (ii) -0.5+x, 0.5-y, 1-z)
and (2) a closed-shell interaction between the O atoms of one nitro group (O3-N9-O4) with
the tetrazole moiety.
The interionic distances for the O4 and O3 contacts with the tetrazole ring are in the
range of 3.020(3) to 3.278(3) . For closed-shell interactions there is a somewhat arbitrary
dividing line between what is or is not an interaction when only the sum of the van der
Waals radii are applied as criterion. Since the MeDAT is a non-spheric ion which is also
able to form hydrogen bonds it is hard to compare the anion-cation interactions with those in
alkali DN salts which are observed in the range of 2.810(2) (K+[N3O4]-) to 3.534(6)
(Cs+[N3O4]-) [2b].
4.
DECOMPOSITION EXPERIMENTS
3 belongs to the new class of energetic ionic liquids related to the low melting salts of
1,2,4- and 1,2,3-triazole systems, introduced by Drake et al [7]. The thermal stability was
initially evaluated using differential scanning calorimetry (DCS) and thermogravimetric
analysis (TGA). For initial safety testing the impact and friction sensitivity was tested
according to BAM methods with the BAM Fallhammer and BAM friction tester [8]. The
energy of activation for the decomposition step was estimated by the method of Ozawa and
Kissinger [9] by following the differential heating rate method of the American Society for
Testing and Materials (ASTM) according to the ASTM protocol E 698 99 [10].
In order to analyse the gases from the stepwise decomposition of 3, a specially equipped
IR-cell was loaded with the compounds (~ 2 mg) and evacuated. The sample holder of the IR
cell was heated at a rate of 4C/min (CARBOLITE 900C Tube Furnace type MTF 9/15)
and the reaction products were allowed to expand continuously into the gas cell. During this
heating, the IR-spectra were recorded continuously as a function of the heating rate using a
Perkin-Elmer Spektrum One FT-IR instrument. To record the mass spectra, a sample (~
1mg) of the 3 was heated at a heating rate of 4C/min (CARBOLITE 900C Tube Furnace
type MTF 9/15) in a one side closed glass tube (length: 500 mm; diameter: 5 mm)
connected to the reservoir of the mass spectrometer (Jeol MStation JMS 700). In this case,
the spectra were also recorded as a function of the heating rate. Residual decomposition
products (e.g. sublimed material and residue after the decomposition) were identified by
means of mass spectrometry (EI and DEI mode) and NMR (1H, 13C) techniques.
Preliminary sensitivity testing of the crystalline compound 3 indicates rather low impact
sensitivity with a value of 7 J in contrast to the relatively high friction sensitivity of 24 N.
The detonation pressures of 3 was calculated to be P = 33.6 GPa (comparable to RDX [11], P
= 34.4 GPa) and a detonation velocities of D = 8827 m s-1 (comparable to RDX [12], D =
8750 m s-1) was estimated [13]. The oxygen balance of 3 shows an reasonable value of -25.3%
and heat of formation was determined to be positive with a value of 92.1 kcal mol-1.
DSC and TGA were used to evaluate the relative thermal stabilities of 3 and are
measured under comparable condition (Al-containers with a hole (1m) on the top for gas
release for DSC measurements). Figure 2 shows the obtained DSC and TGA thermograph of
3.
196
Fig 2.
DSC and TGA thermographs of 3 ( = 10C/min).
3 shows an relatively low melting point (85C (onset)) and start to decompose in the
range of 150 230C. The enthalpy of melting, m H , were found to be 118 J g-1. The
DSC curves of the investigated 3 is divided into three parts, the first one corresponds to a
phase transition region which appears as one small endothermic reaction (Tg = 82 C), the
second part shortly after, is related to the melting region of 3 as a endothermic peak. The
third exothermic peak corresponds to the decomposition reaction. The purity of 3 was
determined to be higher than 98% by the evaluation of the corresponding melting peaks with
the vant Hoff equation according ASTM protocol E 928 96 [14]. The activation energy was
estimated to be 32.89 kcal mol-1 (Ozawa) and 32.91 kcal mol-1 (Kissinger), respectively, i.e.,
in close agreement.
3 does not decompose residual free in the temperature range of 50 400C, as the mass
loss was determined by TGA to be ~ 90%. The reason for this is found in the formation of
condensation products with higher molecular mass, e.g. melam, melem or melom [15]. These
products decompose at higher temperatures (> 500C) to volatile substances like cyanamide,
hydrogen cyanide and ammonia [16]. According to the mass loss derived from the TGA
experiment, the dinitramide salt 3 decomposes in one single step.
197
Fig 3.
Infrared spectroscopic evolved gas analysis of 3.
3 shows a relatively stable melt up to ~ 150C where short after, also in accordance with
the DSC experiment, decomposition occurs and a broad exothermic peak in the DSC and a
sudden rise of gaseous products in the IR spectra indicates a spontaneous decomposition in
one step. The thermal decomposition of 3 under reduced pressure in the temperature range
from 40 to 220C is depicted in Figure 3. Typical products observed during the
decomposition of 3 are N2O (6) as the main product, MeN3 (14), MeONO2 (9), 1,3,5-triazine
(13), HCN, NH3 and H2O (Figure 4, Table 1).
Fig 4.
Gas phase IR spetrum of the decomposition products of 3.
198
Table 1.
Vibrational frequencies (cm-1) of the experimentally observed molecules

in the gas phase.
Species
Frequencies
ref
MeN3
2940 (m), 2818 (w), 2539 (vw), 2198 (s), 2104 (vs), 1466 (vw), 1450 (vw),
1284 (s), 920 (w), 676 (w)
3336 (m), 1626 (s), 968 (vs), 933 (vs)
3311 (s), 2097 (w), 712 (vs)
2959 (m), 2917 (m), 1678 (vs), 1661 (vs), 1442 (m), 1430 (m), 1296 (s),
1287 (s), 1278 (s), 1017 (s), 862 (s), 855 (s), 853 (s), 844 (s) 758 (m), 658
(m)
3891 (w), 3480 (m), 2809 (w), 2591 (m), 2488 (m), 2457 (vs), 2217 (vs),
1890 (w),
1302 (vs), 1275 (vs), 1183 (m), 1155 (m), 694 (w), 588 (w)
3657 (s), 1595 (s)
[17]
NH3
HCN
MeONO2
N2O
H2O
[18]
[19]
[20]
[21,22]
[18]
The thermal decomposition is initiated by a proton transfer to subsequently produce

dinitraminic acid HN3O4 (5) in the melt. According to [23], 5 decomposes under formation of
N2O (6) and HNO3 (7) which produce 8 from the recombination of 4 and HNO3 (7). This
reaction process is shown in Scheme 2 (I). No evidence, whether from the explosion
experiments nor the decomposition experiments, was found for the formation of DAT (1)
(Scheme 2, II) and methyldinitramide (18).
Interestingly, during the decomposition of 3, the formation of NO2 was not observed
which excludes the favoured mechanism of decomposition of HN3O4 (5) to NO2 (15) and
HNNO2 (16) according [24] at higher temperatures (Scheme 2, V). The main decomposition
process of in situ formed 8 involves a methyl group transfer to 1 and methyl nitrate (9)
(Scheme 2, III) and not a proton transfer to 4 and HNO3 (2) (Scheme 2, IV). The main
process observed is the formation of 6 which was found to be the major product (Figure 4) in
the IR spectra of the explosion products as well as in the spectra of the decomposition
experiments and could also be identified in the mass spectrometry experiments.
Interestingly, under this reaction condition, the decomposition pathway discussed in [15],
which involves the imino form of DAT (1), undergoes not the decomposition with the
evolution of HN3, but rather eliminates nitrogen from DAT (1) (Scheme 2, III).
199
Scheme 2.
The resulting unstable nitrene (10) decomposes to HCN, NH3 and N2 according Scheme
2 and in a minor process to cyanamide which is less stable and dimerizes to dicyandiamide
that reacts through an intermediate to melamine (11) [25].
200
The hydrogen cyanide undergoes in situ cyclotrimerization to the 1,3,5-triazine (12),

which is known to proceed under harsh condition [26] and was also identified in the IR as
well as mass spectrum (Figure 4). Only small amounts of MeN3 indicates the partly
decomposition of 4 according Scheme 2 (IV) resulting from a proton transfer. Nitric acid
and MeN3 were found as components in the mass spectra of the residual explosion products
in small quantities.
Fig 5.
13
C-NMR spectra of decomposition products of 3 recorded in [d6]-DMSO.
1,2,4-Triazole (13), 1-amino-4-methyl-5-imino-4,5-dihydro-1H-tetrazole (4), 1,5diamino-1H-tetrazole (1), 1,3,5-triazine (12) and melamine (11) could be identified by
means of 13C NMR spectroscopy (Figure 5) and were found in the residue of the
decomposition experiments. Ammonium nitrate (17), resulting from the recombination of
HNO3 and NH3, was identified as one component of the obtained sublimate (Scheme 2, VI)
and explains the low concentration of ammonia. The thermal decomposition of 9 above
200C leads to the formation of CH2O, CH3OH, H2O, CO, NO and NO2 [27] (Scheme 2,
VII). The thermal decomposition of AN during explosion processes leads to further
formation of N2O (6) and H2O (Scheme 2, VIII).
201
5.
CONCLUSION
3 shows an distinctive decomposition pathways which leads to the formation of a series

of products, which are for the major gaseous decomposition products are still highly
endothermic. According the UN guideline ST/SG/AC.10/11 [8a], 3 is not safe for transport
but prilling of 3 might increase the stability toward friction. The predicted detonation
performance of 3 is higher than those found for RDX and the oxygen balance is in a
reasonable limit. Therefore, 3 might find application in liquid monopropellants, similar to
HAN [28] and ADN [29] monopropellants.
Acknowledgements
Financial support of this work by the University of Munich (LMU) and the Fonds der
Chemischen Industrie is gratefully acknowledged (JJW thanks for a FCI scholarship, DO
171/46). The authors are also indebted to and like to thank Mr. Gunnar Spie for the drop
hammer and friction tests. We are also indebted to and thank Mr. Christoph Rohbogner for
his very valuable help concerning the preparation of the used compounds during his
research project.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
(a) Dixon, D. A.; Feller, D.; Christe, K. O.; Wilson, W. W.; Vij, A.; Vij, V.; Brooke, H. D. B.;
Olson, R. M.; Gordon, M. S. J. Am. Chem. Soc. 2004, 126(3), 834-843; (b) Vij, A.; Pavlovich,
J. G.; Wilson, W. W.; Vij, V.; Christe, K. O. Angew. Chem., Int. Ed. Engl. 2002, 41(16), 30513054; (c) Huynh, M.-H. V.; Hiskey, M. A.; Hartline, E. L.; Montoya, D. P.; Gilardi, R. Angew.
Chem., Int. Ed. Engl. 2004, 43(37), 4924-4928.
(a) Bottaro, J. C.; Penwell, P. E.; Schmitt, R. J. J. Am. Chem. Soc. 1997, 119(40), 9405-9410;
(b) Gilardi, R.; Flippen-Anderson, J.; Georg, C.; Butcher, R. J. J. Am. Chem. Soc. 1997,
119(40), 9411-9416; (c) Sitzmann, M. E.; Gilardi, R.; Butcher, R. J.; Koppes, W. M.; Stern, A.
G.; Trasher, J. S.; Trivedi, N. J.; Yang, Z.-Y. Inorg. Chem. 2000, 39(4), 843-850; (d) Martin,
A.; Pinkerton, A. A.; Gilardi, R. D.; Bottaro, J. C. Acta Crystallogr. 1997, B53, 504-512.
(a) Klaptke, T. M.; Mayer, P.; Schulz, A.; Weigand, J. J. J. Am. Chem. Soc. 2005, in press; (b)
Glvez-Ruiz, J. C.; Holl, G.; Karaghiosoff, K.; Klaptke, T. M.; Lhnwitz, K.; Mayer, P.;
Nth, H.; Polborn, K.; Rohbogner, C. J.; Suter, M.; Weigand, J. J. 2005, manuscript submitted;
(c) Fischer, G.; Holl, G.; Klaptke, T. M.; Weigand, J. J. 2005, manuscript submitted.
(a) Gaponik, P. N.; Karavai, V. P. Khim. Geterotsikl. Soedin. 1984, 1388-1391; (b) Matulis, V.
E.; Lyakhov, A. S.; Gaponik, P. N.; Voitekhovich, S. V.; Ivashkevich, O. A. J. Mol. Struc.
2003, 649, 309-312.
How-Ghee, A.; Fraenk, W.; Karaghiosoff, K.; Klaptke, T. M.; Mayer, P.; Nth, H.; Sprott, J.;
Warchhold, M. Z. Anorg. Allg. Chem. 2002, 628, 2894.
N-N values and N=N values from: International Tables for X-ray Crystallography; Kluwer
Academic Publisher: Dordrecht, 1992; Volume C.
202
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
(a) Drake, G.; Hawkins, T.; Brand, A.; Hall, L.; Mckay, M. Propellants, Explosives,
Pyrotechnics. 2003, 28(4), 174-180; (b) Xue, H.; Arritt, S. W.; Twamely, B.; Shreeve, J. M.
Inorg. Chem. 2004, 43(25), 7972-7977.
(a) Test methods according to the UN Recommendations on the Transport of Dangerous
Goods, Manual of Tests and Criteria, fourth revised edition, United Nations Publication, New
York and Geneva, 2003, ISBN 92-1-139087-7, Sales No. E.03.VIII.2.; 13.4.2 Test 3(a)(ii)
BAM Fallhammer; (b) Reichel & Partner GmbH, http://www.reichel-partner.de/.
(a) Ozawa, T., Bull. Chem. Soc. Jpn. 1965, 38, 1881; (b) Kissinger, H. E. Anal. Chem. 1957,
29, 1702.
Standard Test Methods for Arrhenius Kinetic Constants for Thermally Unstable Materials.
ASTM designation E698-99, 1999.
Mader, C. L. Detonation Properties of Condensed Explosives Computed Using the BeckerKistiakowsky-Wilson Equation of State, Los Alamos Scientific Laboratory 1963, Rept. LA2900.
Khler, J.; Mayer, R. Explosivstoffe, 7. Aufl., Wiley-VCH, 1991, Weinheim.
(a) Kamlet, M. J.; Jacobs, S. J. J. Chem. Phys. 1968, 48, 23; (b) Kamlet, M. J.; Ablard, J. E. J.
Chem. Phys. 1968, 48, 36; c) Kamlet, M. J.; Dickinson, C. J. Chem. Phys. 1968, 48, 43.
Standard Test Methods for Mol Percent Impurity by DSC. ASTM designation E928-96, 2003.
Stoner, C. E.; Brill, T. B. Combust. Flame 1991, 83, 302.
Costa, L.; Camino, G. J. Therm. Anal. 1988, 34, 423.
Milligan, D. E.; Jacox, M. E. J. Chem. Phys. 1975, 56, 333.
Shimanouchi, T. Tables of Molecular Vibrational Frequencies Consolidated Volume I,
National Bureau of Standards (1972) 1.
Shimanouchi, T. Tables of Molecular Vibrational Frequencies Consolidated Volume II, J.
Phys. Chem. Ref. Data, 6(3) (1972) 993.
Brand, C. D.; Cawthon, T. M. J. Am. Chem. Soc. 1955, 77, 319.
Nakamoto, N. Infrared and Raman Spectra of Inorganic and Coodination Compounds, 4th ed.;
Wiley & Sons: New York, Chichester, Brisbane, Toronto, Singapore, 1986.
Mecke, R.; Langenbucher, F. Infrared Spectra, Heyden & Son, 1965, London, Serial No. 106.
Mishra, I. B.; Russell, T. B. Thermochim. Acta 2002, 384, 47.
Brill, T. B.; Brush, P. J.; Patil, P. J. Combust. Flame 1993, 92, 178.
Gmelin, L. Gmelins Handbuch der anorgnischen Chemie, 8. Aufl. Verlag Chemie, Weinheim,
1971, Syst.Nr. 14, Teil D1, 347.
Low, M. J. D.; Ramamurthy, P. J. Res. Inst. Catal. 1968, 16(2), 535.
Jain, S. R.; Rao, M. V.; Verneker, V. R. Propellants, Explosives, Pyrotechnics 1979, 3, 761.
Meinhardt, D.; Brewster, G.; Christofferson, S.; Wucherer, E. J. AIAA 98-4006, 34th
AIAA/ASME/SAE/ASEE Joint Propulsion Conference, 13 15 July, Cleveland, OH, USA,
1998.
Anflo, K.; Grnland, T. A.; Wingborg, N. AIAA 2000-3162, 36th AIAA/ ASME/SAE/ASEE
Joint Propulsion Conference, 16 19 July, Huntsville, AL, USA, 2000.
203
STUDY OF NON-ISOTHERMAL NITRIDATION

OF ALUMINUM NANOPOWDERS PASSIVATED
BY NON-OXIDE LAYERS
Alexander Gromov*, Ulrich Frter-Barth**, and Ulrich Teipel**
* Tomsk Polytechnic University, 30, Lenin Ave., 634050, Tomsk, Russia
** Fraunhofer Institute for Chemical Technology, P.O. Box 1240,
D-76318 Pfinztal (Berghausen), Germany
Abstract:
Results of DTA-TG investigation and chemical analysis of electro-exploded aluminum
nanopowders coated and/or passivated with the reactive reagents: nitrocellulose (NC),
oleic (C17H33COOH) and stearic (C17H35COOH) acids, amorphous boron and air (for a
comparison) are discussed. Surface protection of aluminum nanopowders by coatings of
different origin results in significant advantages in the energetic properties of the
powders. Aluminum nanopowders with a protecting surface show increased stability to
oxidation during storage period.
Keywords: aluminum nanopowder, surface passivation, energetic properties, DTA-TG
1.
INTRODUCTION
Nowadays a number of new synthesized energetic materials have been reported and
generated much discussion. One of the most interesting newly developed material is
aluminum nanopowder (ANP) used as a propellant booster [1]. This material can be produced
by several methods, but the highest productivity of them (~150 g/hour for ANPs) is the
electrical explosion of wires (EEW). This method is characterized by a high rate of energy
input into the material. Electro-exploded ANPs after stabilization and air-passivation contain
80-85 wt. % of metallic aluminium and are protected from further oxidation by an Al2O3
film. But, 20-25 wt. % of oxide ballast in ANPs leads to the loss of a number of
advantages for their use in propellants in comparison to micron-sized powders which have a
content of metallic aluminium up to 99.5 wt. %. Thus, studies how to increase the metal
content of EEW-ANPs should be concentrated in two directions: search for appropriate
passivating and coating reagents for nanopowders to reduce the oxide formation during
production and simultaneously to increase the stability of ANPs towards oxidation during
storage and mixing with propellant binders (HTPB, GAP etc).
This specific motivation attracts powders which are coated by reactive reagents directly
during ANP production (explosion in Ar) or after particles formation and cooling before
contact with air [2], because the surface of Al particles immediately oxidizes when contacted
with air. All after-oxidation-treatment cannot remove such oxide layers from the particle
surface. Hence, we cannot store more metal in the particle.
The results of DTA-TG investigation in nitrogen of electro-exploded ANPs coated
and/or passivated with the non-oxide reagents: nitrocellulose (NC), oleic (C17H33COOH) and
stearic (C17H35COOH) acids, amorphous boron and air (for the comparison) are given in this
204
work. Nitrogen as a media for DTA-TG study has been selected for 2 reasons: it is one of the
components of a flame for modern fast-burning propellants and nitrogen-containing
compounds (ADN, HMX, TNT etc.) come into contact with ANPs during propellant
fabrication and storage.
2.
EXPERIMENTAL RESULTS AND DISCUSSION

2.1
Aluminum nanopowders preparation and passivation
Parameters for powder production are shown in table 1. Production of the ANP samples
was carried out by using an EEW machine of the 4th generation UDP-4G which was
constructed for the mass production of metallic nanopowders at the High Voltage Research
Institute, Tomsk Polytechnic University, Russia. The characteristics of the UDP-4G machine
are comprehensively described elsewhere [3].
Table 1. Parameters of production for electro-exploded aluminum nanopowders
Nanopowder Voltage by
EEW,
Aluminum
Energy
input,
kV
a.u.
26
1.4
Capacit Inducti Resist Wire Wire

Ar
ance,
vity,
ance, diame length press
ure,
Ohm ter,
mm
F
H
mm
MPa
2.17
0.6
0.07
0.35
70
0.15
Non-passivated ANPs immediately self-ignite if exposed to air, making passivation

essential. Hence, after the EEW machine stopped, the collector with prepared powder was
placed into a separate hermetic passivation chamber immediately after powder production.
Four samples, studied within this work and their specific surface area, determined by BET
method, as well as metal aluminum content (CAl), measured by volumetric analysis, and
volume mean particle diameter (av) are shown in table 2.
Table 2. Properties of aluminum nanopowders
Sample
code
Gas
Passivation condition
media in
explosive
chamber
Ar
Air
BET, av,
m2/g nm
CAl,
wt. %
1. Al (Air)
Initial
wire
compos
ition
Al
18,6
553
85
2. Al (B)
Al (B)
Ar
Air
12,0
610
84
3. Al (St Ac)
Al
Ar
Stearic acid in ethanol
12,1
255
74
4. Al (Ol Ac)
Al
Ar
Oleic acid in ethanol
14,3
393
45
5. Al (NC)
Al
Ar
Nitrocellulose in ethanol 12.6
68
205
Sample 1 ALEX analogue was produced in the High Voltage Research Institute by
the UDP-4G machine. Sample 2 was obtained from the composite Al-B wire. Airpassivation was carried out at room temperature in the medium of argon gas (pressure of
1.1 atm) having an air content of about 0.1 vol. %. These conditions of air-passivation for
metal nanopowders were discussed in a previous paper [2]. ANP samples 1-2, table 2, were
completely passivated by an Ar+0.1 vol. % air mixture for 72 hours. The end of the
passivation period was determined as the moment when the nanopowder stopped reacting
with air, i.e. when the pressure of the gas mixture in the passivation chamber stopped
decreasing. Samples 3-5 were passivated by organic substances in solvents:
-
0.1 wt. % stearic acid, C18H36O2 solution in ethanol (C2H6O);

0.1 wt. % oleic acid, C18H34O2 solution in ethanol (C2H6O);
0.1 wt. % nitrocellulose solution in ethanol (C2H6O).
The solution for passivation was added to the fresh powder immediately after production and
the powder solution was mechanically stirred for ~2 hours. The temperature was kept lower
than 30 oC in order to avoid self-heating and partial self-sintering of the powder. All powders
were stored in open-to-atmosphere-boxes for 2 months after passivation to simulate
conditions, close to industrial.
Reduction in the specific surface area (BET) of samples 2-5, passivated by non-oxide
layers, can be caused by the residue of solvent on the particles compared to sample 1
passivated by air. This is confirmed by a reduction of the content of Al metal (CAl) in
samples 2-5: the reduction is maximal for ANP passivated by oleic acid down to 45 wt. %.
Hence, coating of particles by organic reagents leads to considerable reduction of the
specific metal content in the powder.
2.2
Aluminum nanopowders characterization
DTA-DSC-TG (SHIMADZU 60, Japan) was used for the study of the reactivity of
ANPs in N2. TEM-EDX study provided by the University of Karlsruhe (apparatus Philips
(FEI) CM 200 FEG, electron source field emitter (Shottky-type), 200 kV). SEM-EDX
investigations were performed with the JSM-6500 F (JEOL, Japan) apparatus, which
includes a field emission gun. XRD analyses were carried out using a MAX-B
diffractometer (Rigaku, Japan) with CuK radiation (Ni-monochromated). The Zetasizer
3000 by Malvern Instruments Ltd., UK, was used for the determination of the particle size.
The results of the TEM-EDX and XRD study of ANPs are presented in table 3. All
powders contain more than 10 wt. % of oxygen (as AlxOy) on the particle surface. Traces of
aluminum carbide were found by XRD for samples 3 and 4 (C was not determined by EDX).
206
Table 3. Elemental and phase composition of aluminum nanopowders
Sample code
Wt. content of elements, % (EDX) Phase composition

(XRD)
O
Al
1. Al (Air)
15.5
84.5
2. Al (B)
10.9
89.1
3. Al (St Ac)
15.1
84.9
4. Al (Ol Ac)
18.3
81.7
5. Al (NC)
Al
Al, traces of Al4C3
n/a
According to TEM observation, the concentration of Al inside the particles increases by

application of organic coatings (fig. 1), while the specific metal content CAl is lower for
powders passivated by organic substances (table 2). It should be noted that the term "specific
metal content" characterizes the content of metal in the powder, but not in the particle. The
content of metal in ANP, passivated by oleic acid, can be higher than for ANP, passivated by
air (fig. 1a). According to TEM data, particles, passivated by air, are covered with an oxide
film (thickness of 4-5 nm), while particles, passivated by oleic acid (fig. 1b) do not have an
oxide layer. Analyzing fig. 1a, we can observe the beginning of oxide film crystallization,
i.e. the critical thickness for amorphous oxide film is 7-8 nm, after that crystallization begins.
SEM results confirm that organic-passivated particles hold on their surface a lot of residual
solvents (fig. 2): particles are not separated as in the case of dry powders.
10 nm
10 nm
a
Fig 1.
TEM images of ANPs passivated by air (a, sample 1) and oleic acid
(b, sample 4).
207
Fig 2.
2.3
SEM images of ANPs passivated by air (a, sample 1) and stearic acid
(b, sample 3).
Aluminum nanopowders non-isothermal nitridation
Results of the indirect estimation of the content of metallic aluminum in ANPs

according to the heat of melting ( melt Al, table 4) show that the highest metal content
possesses the powder, whose particles are covered by boron (see also CAl, table 1). The
minimum temperature of the beginning of nitridation had ANP (NC) (sample 5, table 4),
which, probably, is caused by the preliminary activation of its surface during NC
decomposition (Tdec=195oC, fig. 3). The most stable to nitridation (highest Tnitrid) is sample
ANP (Air), which, probably, is explained by the large strength of fresh amorphous oxide
films on the particles. The exothermic effect of nitridation ( nitrid Al) is maximal for the
sample ANP (Air). Sample ANP (NC) also had the high heat of nitridation (3940 J/g), which
can be caused by both the simultaneous nitridation of aluminum with decomposition by
residual NC and by nitridation of aluminum due to the preliminary activation of the surface.
The degree of transformation of Al to AlN (), was maximum for ANP passivated by oleic
acid (sample 4, table 4). It should be noted that the mass of coating on the particles in the
case of NC, stearic and oleic acids is high, which reduces the specific metal content in the
powders.
208
Fig 3.
DTA curves for ANPs under non-isothermal heating in nitrogen

(m = 4.4 mg, vheat= 10 K/min, etalon -Al2O3). Numbers of samples relate
to table 1.
Table 4. Reactivity parameters of aluminum nanopowders (m=4.4 mg)

by non-isothermal heating in nitrogen
1.
nitrid,
Sample code melt Al
o
(at 660),
J/g
Al (Air)
-280
798
nitrid Al, Weight of coating ***

(gases), %
J/g
(5001000oC),
%
4270
0,4
76.0
2.
Al (B)
-287
749
534
3,3
72.9
3.
Al (StAc)
-109
690
584
11,0
37.9
4.
Al (Ol Ac)
-84
700
230
11,6
86.0**
5.
Al (NC)
-89
553
3940
24,0*
79.3
* Desorption of coating accompanies exo-peak on DTA curve (NC burning for sample 5)
** Up to 900
*** Degree of conversion (AlAlN) assuming metal content in the powders (table 1)
209
3.
CONCLUSION
Experimental studies of ANPs passivated by organic and inorganic compositions have

been carried out. The applied coatings provide the reaction of metallic particles with nitrogen
under lower temperatures. In the case of NC and oleic acid, the degree of transformation of
Al to AlN () is higher than for air-passivated ANP. Oleic and stearic acid passivation of
ANPs results in carbonization of the particles surface (traces of Al4C3 fixed by XRD). The
best characteristics under reaction with N2 show ANPs with NC and boron coatings. Hence,
their study in aluminized propellants is most promising in future experiments. Passivation of
ANPs by non-oxide coatings results in an increase of the metal content inside the particles
and, hence, in an increasing of the powder burning enthalpy.
Acknowledgement
This work has been financially supported by INTAS grant YSF 03-55-671.
REFERENCES
[1]
[2]
[3]
H. H. KRAUSE: New Energetic Materials,(Energetic Materials, U. Teipel), Wiley, p.1-25, 2004

Y. S. KWON, A.A. GROMOV, A.P. ILYIN AND G.H. RIM: Passivation Process for Superfine
Aluminum Powders Obtained by Electrical Explosion of Wires, Applied Surface Science, 211,
p. 57-67, 2003
Y. S. KWON, Y.H. JUNG, N.A. YAVOROVSKY, A.P. ILYIN AND J.S. KIM: Ultrafine Metal
Powders by Wires Electric Explosion Method, Scripta mater., 44, p. 2247-2251, 2001
210
OPTIMIZATION
OF 1,1-DIAMINO-2,2-DINITROETHENE SYNTHESIS
Zbigniew Chyek, Stanisaw Cudzio, Jan Bdek, and Sylwia Pietrzyk
Institute of Chemistry, Military University of Technology 00-908 Warsaw, POLAND
Abstract:
Known syntheses of 2-methylpyrimidine-4,6(3H,5H)-dione and 1-diamino-2,2dinitoethene (DADNE), have been studied in order to improve their yield. Parameters
affecting yield of both the compound were recognized and optimized. A new method of
DADNE stabilization and purification was developed. The overall yield of DADNE
synthesis starting from acetamidine hydrochloride was shown to be around 65%.
Key words:
1.
2-Methylpyrimidine-4,6(3H,5H)-dione, 1,1-Diamino-2,2-dinitoethene,
synthesis
INTRODUCTION
1,1-Diamino-2,2-dinitroethene (DADNE) combines comparatively high performance

and low sensitivity to mechanical and thermal stimuli, and thanks to this it attracts a lot of
interest in the explosive community [111]. The compound is prepared by nitrating
heterocyclic (5- or 6-ring) containing the structural element of acetamidine [1215]. The
reaction path (Fig.1) starting from 2-methylpyrimidine-4,6(3H,5H)-dione (1) is considered to
be the most promising one among so far disclosed methods of DADNE synthesis [16]. The
substrate (1) can be prepared from acetamidine hydrochloride and diethyl malonate [17].
O
HNO3
HN
H2SO4
CH3
(1)
Fig 1.
NO2
HN
N
CH3
(2)
O2N
O
NO2
O
HNO3
H2SO4 HN
H2O
NH
C
O2N
H2N
NO2
NH2
+
C
O2N
NO2
CH2(NO2)2 + 2 CO2
(3)
DADNE
(2')
DADNE synthesis from 2-methylpyrimidine-4,6(3H,5H)-dione (1),

possible intermediates (2) (2) and by-products (3) [13, 15]
However the destructive nitration of the known DADNE precursors consumes a lot of
nitrating agent (concentrated nitric acid) and in some cases produce unwanted by-products,
e.g. dinitromethane (3). Recently reported search for new heterocyclic substrates devoid of
the drawbacks was not successful [18]. Therefore an optimization of the known synthetic
methods may provide a provisional solution to the problem, which is quite important as the
present routes to prepare DADNE makes it a very expensive explosive ca. 3000
EURO/kg [19].
211
Even cursory survey of the literature suggests that the authors still recommend quite
different synthesis conditions. For example Astratiev at. al. [15] in 2001 reported nitration of
(1) with HNO3/H2O4 mixture containing ca. 13 equivalents of HNO3, whereas Bellamy at
al. [18] (2004) and Latypov at al. [13] stated that the optimum yield of DADNE was obtained
when around 5 moles of nitric acid per 1 mole of (1) was used. There are also discrepancies
in the amount of the sulfuric acid used (3.8 mole [13, 18] or 0.5 mole [15] of H2SO4 per 1 mole
of HNO3). The time-temperature regime of both the syntheses differs significantly, as well.
In the present work, results of a parametric study on synthesis of 2-methylpyrimidine4,6(3H,5H)-dione (1) and its nitration with nitrating acid are presented. We checked how the
reagent concentrations, temperature and duration of the reactions influence the yield of the
main products when their synthesis is carried out on a laboratory scale batches of around
100 g.
2.
EXPERIMENTAL
Preparation 2-methylpyrimidine-4,6(3H,5H)-dione. The condensation of acetamidine

hydrochloride with diethyl malonate to form 2-methylpyrimidine-4,6(3H,5H)-dione (1) was
initially performed under identical conditions to those reported by L. P Ferris at al. [17],
Fig. 2.
O
C2H5O
OC2H5
H2 N
NH HCl
CH3OH
CH3
CH3ONa
HN
N
CH3
(1)
Fig 2.
Synthesis of 2-methylpyrimidine-4,6(3H,5H)-dione
They obtained (1) with 43% yield using one-eighth mole of acetamidinie hydrochloride
and one-twentieth mole of a malonic ester and carrying out the reaction in a solution of 3.3 g
sodium in 150 ml of absolute alcohol. After standing for three days, the solution was
neutralized with concentrated hydrochloric acid and enough water was added to dissolve the
co-precipitated sodium chloride [17].
The method was modified as follows: sodium methylate (135 g, 2.5 mole) was dissolved
in 1300 ml of methanol and 94.5 g (1 mole) of acetamidine hydrochloride was added to the
solution, at room temperature. At the same temperature, while stirring, a solution of 160 g (1
mole) of diethyl malonate in 200 ml of methanol was added. The reaction mixture was
stirred and heated at 50 oC, under reflux condenser, for 2 hours and then left in the reactor
for 24 hours at ambient temperature. Precipitation of 2-methylpyrimidine-4,6(3H,5H)-dione
was caused by addition of concentrated hydrochloric acid, adjusting pH to around 6. Sodium
chloride was removed by addition of 1 dm3 of water. The undissolved main product was
filtered off, washed out with water and dried. This method gives (1) in high yield (up to
97 %), Fig. 3.
212
100
97 %
90
Yield of (1) [%]
80
81 %
70
68%
60
50
54 %
40
30
20
10
0
Fig 3.
7
8
9
Concentration of CH3ONa [% wg.]
10
11
Yield of 2-methylpyrimidine-4,6(3H,5H)-dione versus concentration of

sodium methylate
Parametric study on the synthesis has shown that there are two main factors affecting the
yield of 2-methylpyrimidine-4,6(3H,5H)-dione concentration of sodium methoxide in
methanol (used as the reaction medium) and duration of the reaction. The latter can be
reduced practically to a few hours if the reaction mixture is heated at 50 oC. Almost
theoretical yield of (1) can be gained when CH3ONa concentration equals to around 10 %.
Preparation of DADNE. The nitration of 2-methylpyrimidine-4,6(3H,5H)-dione was
initially carried out according to the complete instruction given in patent [13]. The molar ratio
HNO3/(1) equaled to 4.75/1, because at least 4 equivalents of HNO3 is needed to form the
tetranitro intermediate (2), Fig. 1. Around 18 moles of H2SO4 was used to dissolve 1 mole
of 2-methylpyrimidine-4,6(3H,5H)-dione. The only change of the method disclosed by
Latypov at al. [13] consisted in an elongated duration of the hydrolysis of (2). After pouring
the reaction mixture into water, we kept the resulting solution for 24 hours at ambient
temperature to precipitate DADNE. This caused the yield of DADNE to increase to ca. 70%
(Latypov at al. reported 35% yield).
Taking into account that sulfuric acid causes nitric acid to dissociate, producing actual
nitrating agent (cation NO2+), the minimum required amount of H2SO4 can be calculated
from the following reaction:
HNO3 + 2H2SO4 = NO2+ + 2HSO4- + H3O+
Only 2 moles of H2SO4 is needed for complete conversion of HNO3 into nitronium ion,
so that it seems possible to reduce almost by a half the amount of the reactant recommended
in Ref. [13, 18] (ca. 4 moles of H2SO4 per 1 mole of HNO3). Envisaging that such a big change
in reaction mixture volume may cause problems with stirring and heat abstraction, we
performed several experiments systematically decreasing the content of sulfuric acid in the
nitrating acid. Other parameters of the synthesis were kept constant. Results of the
experiments are shown in Fig. 4.
213
80
Yield of DADNE [%]
70
60
50
40
30
20
10
0
1,8
Fig 4.
2,0
2,2
2,4 2,6 2,8 3,0 3,2 3,4

H2SO4/HNO3 molar ratio
3,6
3,8
4,0
Dependence of DADNE yield on composition of nitrating acid molar

ratio H2SO4/HNO3
An optimum molar ratio of sulfuric to nitric acids was found to be around 3, but even
reducing this parameter to 2, still it is possible to obtain DADNE in quite high yield of
around 70%. However, under this conditions, stirring and cooling systems must be very
effective, as the reaction mixture is thick, practically from the very beginning of the nitration
process.
Summing up the following method gave DADNE in the highest yield. 2-Methylpyrimidine-4,6(3H,5H)-dione (126 g, 1 mole) was dissolved in concentrated sulfuric acid
(600 ml, 1.84 g/cm3) at around 20 oC during stirring. To the resulting yellow solution, nitric
acid (200 ml, 1.52 g/cm3) was added (at the same temperature) over a 60-min period. The
mixture was briskly stirred at 15 oC for 3 hours, and then poured into 4.5 dm3 of icy water.
The yellow precipitate was filtered off, washed with water and dried. 110 g of raw DADNE
was recovered (74% yield).
Stabilization and purification of DADNE. In order to purify DADNE, 110 g of the
crude product was poured into 3 dm3 of 0,1% aqueous solution of NaHCO3 and the mixture
was boiled under reflux condenser for 8 hours. Next, the mixture was chilled, filtered, and
the solid material was dissolved in 6.5 dm3 of boiling water. After cooling down, DADNE
crystals were filtered and dried at 60 oC. During the first stage of purification the sample
mass was reduced by ca. 8%. After crystallization from water the mass losses were around
3%. The pure product obtained (ca. 98 g) gave correct analyses (elemental, TG/DTA, XRD,
1
H and 13C NMR, TLC) [20].
214
3.
CONCLUSION
All the performed experiments have shown that using a 2-liter reactor cooled with water
jacket and equipped in a paddle mixer it is possible to produce DADNE from 2-methylpyrimidine-4,6(3H,5H)-dione with quite high yield of ca. 75%. Taking into account almost
theoretical yield of the condensation stage, the total costs of synthesis of 1 kg of pure
DADNE from commercial chemicals (delivered by Merck and Aldrich) was calculated to be
ca. 2000 EURO (including wage costs and overheads).
Acknowledgment:
REFERENCES
[1]
B. JANZON, H. BERGMAN, C. ELDSATER, C. LAMNEVIK, H. OSTMARK, FOX-7 a novel, high

performance, low-vulnerability high explosive for warhead applications, 20th Internationa
Symposium on Ballistcs, Orlando, FL, USA, 23-27.09.2003.
[2] Y. N. MATYUSHIN, G. T. AFANASEV, V. P. LEBEDEV, M. N. MAHOV, V. I. PEPEKIN, TATB
and FOX-7: thermochemistry, performance, detonability, sensitivity, 34th International Annual
Conference of ICT, Karlsruhe, Germany, 24-27.06.2003.
[3] H. OSTMARK, A. LANGLET, H. BERGMAN, N. WINGBORG, U. WELLMAR, U. BEMM, FOX-7
a new explosive with low sensitivity and high performance, 11th Internationa Symposium on
Detonation, Snowmass, CO, USA, 31.08-04.09.1998.
[4] U. BEMM, H. OSTMARK, 1,1-Diamino-2,2-dinitroethylene a novel energetic material with
infinite layers in two dimensions, Acta Cryst., C54, pp. 1997-1998, 1998.
[5] S. CUDZIO, W. KICISKI, Properties and preparation of DADNE (in Polish), Biul. WAT 53,
(2-3), pp. 183-200, 2004.
[6] A. GINDULYTE, L. MASSA, L. HUANG, J. KARLE, Proposed mechanism of 1,1-diamino-2,2dinitroethylene decomposition: A functional theory study, J. Phys. Chem. A 103, pp. 1104511051, 1999.
[7] U. TICMANIS, M. KAISER, G. PANTEL, I. FUHR, U. TEIPEL, Kinetics and chemistry of thermal
decomposition of FOX-7, 35th International Annual Conference of ICT, Karlsruhe, Germany,
2004.
[8] P. B. KEMPA, M. HERRMANN, F. J. MOLINA METZGER, V. THOME, A. KJELLSTROM, N.
LATYPOV, Phase transitions of FOX-7 studied by X-ray diffraction and thermal analysis, 35th
International Annual Conference of ICT, Karlsruhe, Germany, 2004.
[9] N. V. GARMASHEVA, I. V. CHEMAGINA, V. P. FILIN, M. B. KAZAKOVA, G. B. LOBOIKO,
Investigation of diaminodinitroethylene, New Trends in Research of Energetic Materials,
Pardubice, Czech Republic, 116-122, 2004.
[10] R. WILD, U. TEIPEL, Characterization and explosive properties of FOX-7, 35th International
Annual Conference of ICT, Karlsruhe, Germany, 2004.
[11] U. TEIPEL, I. FUHR, K. HARTLIEB, A. KJELLSTROM, C. ELDSATER, Crystallization of 1,1diamino-2,2-dinitroethylene (FOX-7), 35th International Annual Conference of ICT, Karlsruhe,
Germany, 2004.
215
[12] N. LATYPOV, J. BERGMAN, A. LANGLET, U. WELLMAR, U. BEMM, Synthesis and reactios of

1,1-diamino-2,2-dinitroethylene, Tetrahedron 54, pp. 11525-11536, 1998.
[13] N. LATYPOV, A. LANGLET, U. WELLMAR, Chemical compound suitable for use as an
explosive, intermediate and method for preparing the compound, US Patent 6 312 538 B1,
Nov. 6, 2001.
[14] H. OSTMARK, H. BERGMAN, U. BEMM, P. GOEDE, E. HOLMGREN, M. JOHANSSON, A.
LANGLET, N. LATYPOV, A. PETTERSSON, M-L PETTERSSON, N. WINGBORG, C. VORDE, H.
STENMARK, L. KARLSSON, M. HIHKIO, 2,2-Dinitro-ethene-1,1-diamine (FOX-7) properties,
analysis and scale-up, 32nd International Annual Conference of ICT, Karlsruhe, Germany,
2001.
[15] A.A. ASTRATIEV, A. DASHKO, D. MERSHIN, A. STEPANOV, A. URAZGILDIEV, Russian Journal
of Organic Chemistry 37, pp. 729-733, 2001.
[16] E. HOLMGREN, H. CARLSSON, P. GOEDE, N. LATYPOV, C. CRESCENZI, Characterization of
FOX-7, its precursors and possible byproducts, 34th International Annual Conference of ICT,
Karlsruhe, Germany, 24-27.06.2003.
[17] L. P. FERRIS, A. R. RONZIO, A series of 2-methyl-5-alkyl-4,6-dihydroksypyrimidines, J. Amer.
Chem. Soc. 62, pp. 606-607, 1940.
[18] A. J. BELLAMY, N.V. LATYPOV, P. GOEDE, Studies on the nitration of new potential
precursors for FOX-7, New Trends in Research of Energetic Materials, Pardubice, Czech
Republic, 116-122, 2004.
[19] U. TEIPEL (Ed.), Energetic Materials, Particle Processing and Characterization, p. 21, WileyVCH, Weinheim, 2005.
[20] S.CUDZIO, Z. CHYEK, R. DIDUSZKO, Crystallization and characterization of 1,1-diamino2,2-dintroethene (DADNE), will be published in Centr. Europ. J Energ. Mat., 2005.
216
THE INFLUENCE OF THE STRUCTURE OF THE SALTS

OF AZOLES UPON THE PROCESSES OF THEIR THERMAL
AND LASER INITIATION
Mikhail A. Ilyushin, and Igor V. Tselinskii
Saint-Petersburg State Institute of Technology, Saint-Petersburg, Moskovsky pr. 26
Abstract:
It is experimentally shown that for metal salts of azoles there is no universal factor
determining the processes of deflagration under thermal and laser initiation.
For the series of azoles having a common initial stage of thermal degradation, the step
of deflagration-to-detonation transition (DDT) for the same metal cation depends on
Hf0 value of the salt.
At the same time within the range of silver salts of isomeric N-nitroaminotetrazoles Hf0
values of the compounds, their structure, reactivity of the products of the initial
decomposition and the composition of gaseous products of burning all influence their
initiating ability under thermal initiation.
But the ionization potential of complex perchlorates of d-metals with 3(5)-hydrazino-4amino-1,2,4-triazoles as ligands determines the intensity of initial steps of
decomposition under laser initiation which, in its tern controls the threshold of ignition.
Hence the knowledge of the nature of initial decomposition stages of azole salts, taking
into account the mechanism of energetic effects, is necessary for prediction of their
behavior under initiation.
Keywords: deflagration-to-detonation transition (DDT), salts of azidoazoles, silver
salts of N-nitroaminotetrazoles, complex perchlorates of d-metals with
3(5)-hydrazino-4-amino-1,2,4-triazole
1.
INTRODUCTION
Primary explosives are used to initiate detonation (explosion) processes in the charges of
high explosives (HE). By initiation ability of the primary explosive charges there is meant
the ability after their initiation by a simple pulse (mechanical impact, mechanical friction,
beam of fire, laser beam, etc.) to initiate the detonation processes in the charges of HE.[1].
Detonation in charges of HE is commonly initiated by the pressure and energy of explosion
products of primary explosives, which are more sensitive to external stimuli and readily
detonate under the action of a thermal or mechanical pulse. Combustion in primary
explosives quickly transits into detonation within a distance not exceeding several
millimeters from the ignition place. The shorter the region within which combustion
transforms into detonation, the higher the efficiency of a primary explosive
217
2.
INITIATION OF DETONATION PROCESSES IN EXPLOSIVES
Modern theory of the initiation of explosives under the action of a pulsed stimulus is
based on the concept of "hot spots - the centers of intense local heating, that are responsible
for the fast reactions of thermal decomposition including explosion. Local heating centers
are formed due to the structural inhomogeneity of the material (dot defects, congestions of
vacancies, dislocations, interfaces, etc.), chemical impurities, microinclusions occurring
owing to the non-stoichiometric composition of the initial products, addition of various
impurities (presence of various additives) and so on.
The defects can also arise in the field of an intense light wave. The main hypothesis of
warming up the optical heterogeneities resulting in the ignition of explosives has been
adopted for laser initiation due to the short duration and therefore the high power of laser
impulse. Carbon inclusions contained in organic high explosives, or metal microinclusions
present in azides of heavy metals can play the role of initial centers.
The mean diameter of hot spots in the case of mechanical impact or mechanical friction
is of the order ~10-3- ~10-5 cm [2]. The smallest sizes of the experimentally found hot spots
under the action of laser monopulse (Q-switch) ( = 1064 nm, q = 8 ns) depend upon the
nature of an explosive. In a compound with covalent bonds (RDX, hexogen) it equals to ~20
nm. For ionic compounds (ammonium perchlorate, NH4CIO4) the size of the centers is larger
and equals to about 200 - 300 nm and even attains ~1000 nm.[3].
The decomposition of an explosive can be accelerated if the temperature at the reaction
site is such that the heat generation by a chemical reaction exceeds the heat dissipation into
the environment (the so-called "flash point"). After a certain timelag, tig, from the start of
heating the specimen ignites at the decomposition site
tig = B exp(E/RT0)
where B is an empirical constant, E is the activation energy of thermal decomposition, R
is the universal gas constant, and tig is the flash point of the explosive. It is necessary that the
specimen should remain at this temperature during the induction period of thermal
decomposition. The physical meaning of the pre-exponential factor B is the minimum
possible ignition delay. The B value (10-8 10-13 s) depends on the sample properties and
experimental conditions.
The pressure of explosion products and the detonation velocity of an explosive are
related by
P=
0 * D2
n+1
where P is the detonation pressure, 0 is the density of the explosive, D is the detonation
velocity, and n is the polytropic exponent (for most explosives n ~ 3). To initiate an
explosion in HE, a shock wave should have the pulse energy more or equal to that given by
the equation below,
J critical Pcritical d
0
Thus, the initiating ability of a shock wave depends on the front pressure (Pcritical) and
the time of compression (). The value of Jcritical for RDX is near 46 J/cm2 which
218
corresponds to P2 = 460 [4].The initiating power of primary explosives can be estimated

from the value of their minimal blasting charge necessary to initiate HE placed in the
standard cartridge shell detonator (for example, no. 8) [4]. The initiating ability of a primary
explosive mainly depends on the distance from deflagration-to-detonation transition (DDT).
The chemical structure of a primary explosive and physical properties of charges both
influence its initiating ability.
During the first step of the decomposition of explosives the weak bonds in energetic
molecules are broken. This is the fragmentation step. It is an endothermic step common for
all explosives.
Secondary processes (for example, Red-Ox reactions, recombination of free radicals
etc.) are exothermic and different in nature. They produce gaseous reaction products at high
temperatures and pressures [5].
Among primary explosives metal azides are investigated in most detail. Under the action
of an energy pulse they decompose to produce the metal and molecular nitrogen by the
mechanism involving electron transfer. A simplified scheme of the decomposition of
inorganic azides is as follows
_
N3
N3 + e
2N3
2+
M + 2e
3N2 + Q
M
where Q is the heat of decomposition [4,6]. The first step of decomposition of inorganic
azides is the M-N bond breaking. The energy of breaking the weak bond decreases with an
increasing metal ionization potential (I). This energy nearly equals the energy of electron
excitation to the zone of conductivity [6]. Thus, the reaction N3- - e N30 becomes more
real when the power of cations field estimated from the sum value of their potentials of
ionization divided on their ion radius (I/r) increases and the crystal lattice becomes less
ionic but more covalent.
Consequently, the cations nature I main factor determining the explosive properties of
inorganic azides.
The increase in cations enthalpy of formation, which primangly depends on potential of
ionization (I), leads to elevation of the enthalpy of formation of the azide.
Therefore, the values of minimal blasting charges of the salts (W) correlate with the
enthalpy of formation of the inorganic azide (Hfo) or its heat of explosion [7]:
lnW = d + bln Hfo;
where b, d are coefficients.
The simplified scheme for the decomposition of inorganic azides given above is not
fully correct and needs in further refining. For example, it was found that the explosion
products of Pb(N3)2 contained not only N2, but such species as N3, N4; N is in concentration
equal the concentration of N2 [8].
That is why the dependence of the velocity of explosion processes on DDT distance for
inorganic azides may be the result of on incomplete release of potential energy during
decomposition of energetic molecules.
219
3.
INITIATION OF SALTS OF AZIDOSUBSTITUTED NHAND OH-ACIDS
The first step in the decomposition of organic azides and salts of azidosubstituted acids
is the degradation of azidogroups with the elimination of N2 molecule [4].
N3
N
Reaction products
N2 +
R
It can be stated that both the initial reaction and secondary processes will become more
probable with the gain in the heat of thermal decomposition or/and enthalpy of formation of
salts of azidosubstituted acids. Consequently, the value of the energy of degradation must
affect the DDT step and minimal blasting charge of the salts of azidosubstituted acids.
We synthesized and investigated Cu (II), Ag (I), Pb (II) and Hg (II) salts for the
following azidosubstituted acids: 1,2-diazidoglyoxim (1); bis(1-H-3-azido-1,2,4-triazole-5il) (2): 1-H-3-azido-5-nitro-1,2,4-triazole (3); 1-H-3-azido-5-amino-1,2,4-triazole (4); 1-H3,5-diazido-1,2,4-triazole (5); 3-azido4-nitramino-1,2,5-oxadiazole (6) [5,9]
N3
N3
N3
N
NOH
HON
N3
(4)
N
N3
O2N
(3)
N3
N3
(2)
N
N
N3
(1)
H2 N
NHNO 2
N3
(5)
(6)
It was found that the salts of azidosubstituted acids with the same cation have a distinct
tendency of reducing the minimal blasting charge with increasing the enthalpy of formation
[10]
. This tendency is the similar to that for inorganic azides [7].
Consequently, the potential energy releases only partially on the DDT step.
4.
INITIATION PROPERTIES OF AG-SALTS

OF (N-NITRAMINO)TETRAZOLES
It is known that some heavy and transition metal salts of 5-(N-nitramino)tetrazole have a
short DDT period and thus possess initiating properties, i.e., can cause the detonation of a
HE, e.g., RDX [9,11,12].
In order to study the influence of the position of the N-nitramine group on the explosive
properties of salts of (N-nitramino)tetrazoles, we prepared silver salts of l-(N-nitramino)(7), 2-(N-nitramino)- (8), 5-(N-nitramino)- (9) and l-methyl-5-(N-nitramino)tetrazole (10)
[13]
.
220
N --NHNO2
HN
N
(7)
HN
O2NHN-
N
N
N--NHNO2
(8)
NH
O2NHN -
(9)
N-CH3
N
(10)
The initiating power of the salts was estimated from their minimal blasting charge in
RDX by the procedure described in [10]. The tests showed that the minimal blasting charges
of silver salts are dependent on the position of the N-nitramino group and increase in the
sequence:
Ag-8: Ag-7: Ag2-9: Ag-10 1:5:60:120.
So it can be concluded that the salts of 1- and 2-(N-nitramino)tetrazoles have a DDT
period shorter than the salts of 5-(N-nitramino)tetrazoles. The average accuracy of the
determination of minimal charges does not exceed 20%. Salt Ag-8 is a more powerful
initiating explosive than lead azide.
As it may be seen for such initiators as inorganic azides and salts of azidoazoles the
DDT period tends to decrease with the increasing enthalpy of formation of the salt.
According to approximate calculations, the enthalpies of formation (Hfo) of 1- and 2-(Nnitramino)tetrazoles are by ca. 30 kJ. mol-1 higher than those for 5-(N-nitramino)tetrazoles.
However, such negligible increase in Hfo cannot be responsible for the considerable rise in
the initiating power on going from the salts of C-substituted (N-nitramino)tetrazoles (9) and
(10) to the salts of N-substituted (N-nitramino)tetrazoles (7) and (8). Most probably, the
increase in the initiating power suggests significant differences in the mechanism of
decomposition of the salts of (N-nitramino)tetrazoles at the stage of DDT.
In order to evaluate possible differences in the contents of the primary decomposition
products of (7-10), which can be indicative of the plausible routes of their decomposition, we
examined their mass spectra, since, as it was experimentally established for tetrazoles, the
initial steps of thermal decomposition were similar to the fragmentation pattern under the
electron impact [14]. As model compounds we used volatile ammonium salts of (Nnitramino)tetrazoles. The results are presented in Table 1.
Ammonium salts of (7-10) give no molecular ion peaks because of the dissociation to
give ammonia and (N-nitramino)tetrazoles. The spectra of (7-9) contain the molecular ion
peaks of (N-nitramino)etrazoles (M+ 129). Their intensities relative to the total ion current
are 0.6, 1.0, and 0.4%, respectively. Low intensities of the molecular ion peaks of (Nnitramino)tetrazoles may be due to the fact that the oxygen atom of the nitro group is a
strong electron acceptor and so destabilizes the positive charge. Furthermore, the nitro group
is a stable fragment and is readily eliminated under the electron impact. In the spectrum of
salt (10) the molecular ion peak (M+144) is strong: its intensity amounts to 26.1% of the total
ion current. The relative stability of the molecular ion peak of (10) may be due to the
presence of the electron-donating methyl group. The main peaks of (N-nitramino)tetrazoles
(8) and (9) are those with m/z 28. Their intensity amounts to 35.6 and 34.1%, respectively,
of the total ion current. These peaks can be assigned to three possible compounds: molecular
nitrogen, CH2N+, or carbon monoxide. The main peak of (N-nitramino)tetrazole (7) is that
with m/z 30, which corresponds to nitrogen monoxide; its intensity amounts to 18.2% of the
total ion current.
221
Table 1. Mass spectra of ammonium (N-nitramino)tetrazoles

Salt
Fragmentation peaks (% of total ion current)

H
+
NH4
NNO2
(1,2), 55(1,7), 46(12,0), 44(14,2), 42(6,9), 41
129(0,6), 113(0,4), 97(0,7), 83(0,1), 69(1,3), 60

(6,5,), 36(5,5), 30(18,2), 28(17,5), 18(13,9)
M 147
+
NH 4
129(1,0), 97(0,9), 83(1,4), 69(2,3), 60(2,0), 55
(2,9), 46(7,2), 44(8,1), 42(7,4), 41(4,7), 30
N -NNO 2
(16,3), 28(35,5), 18(10,2)
M 147
129(0,4), 97(0,7), 83(0,7), 69(1,2), 60(1,2), 55
2-
+
NH 4
O2 NN
N
(1,6), 46(6,5), 44(4,0), 42(7,0), 41(2,6), 36
(5,7), 30(14,0), 28(34,1), 18(20,3)
M 164
+
NH4
N CH3
O2 NN
N
M 161
145(1,7), 144(26,1), 143(0,1), 129(0,2), 128

(0,2), 114(3,6), 98(29,6), 83(0,3), 82(0,1), 75
(0,2), 68(0,1), 60(0,3), 56(6,3), 55(5,9), 54(0,6),
46(3,9), 44(0,4), 42(0,7), 40(0,7), 32(1,5), 30
(4,7), 26(0,2), 18(2,2)
The peak with m/z 28 is also intense (17.6% of the total ion current). The main peak of
(N-nitramino)tetrazole (10) is that with m/z 98, corresponding to the compound of the
general formula C2H4N5. Its intensity amounts to 29.6% of the total ion current.
-N -CH3 |
HNN
The main fragmentation pathway of the molecular ion peaks of (N-nitramino)tetrazoles

is the loss of nitrogen dioxide. Substantially less intense is the peak of the protonated
tetrazole, m/z 69 (for (N-nitramino)tetrazoles (7-9), resulting from the elimination of the NNO2 group. The major fragmentation pathway of the tetrazole ring in the salts of (7-9) is the
loss of molecular nitrogen, but some portion of the product decomposes with the formation
of the azide radical ion with m/z 42. In (N-nitramino)tetrazole (10) the tetrazole ring
decomposes mainly with the loss of molecular nitrogen. The intermediate products of its
fragmentation after the elimination of the nitro group are relatively stable to oxygen, and
their further fragmentation is practically not accompanied by oxidation, which is evidenced
222
by the presence of free oxygen and only minor amounts of carbon dioxide (m/z 44) in the
decomposition products. It mat be concluded that, (N-nitramino)tetrazole (10) is hardly
suitable for preparing salts with a high initiating power. Salts of (10) should have a
prolonged DDT period because of the low rate of oxidation of the initial decomposition
products and, therefore, reduced heat of decomposition at the stage of acceleration of
explosion. The properties of the silver salt of (10) have been studied experimentally, and
fully support the above conclusion.
Fragmentation of the diammonium salt of (N-nitramino)tetrazole (9) gives fragment ions
with m/z 28 with the intensity three times higher as compared to (10). Examination of the
high-resolution mass spectra of the fragment ions with m/z 28 showed that of the three
possible compounds having this molecular mass molecular nitrogen is the main component.
Furthermore, in the mass spectrum of (N-nitramino)tetrazole (9) the intensity of the carbon
dioxide peak is ten times higher as compared to (10). It means that the fragmentation of this
molecular ion is accompanied by more intense oxidation of carbon. These facts suggest that
decomposition and oxidation at the stage of DDT are probably more close to complete for
the salts of (9) than for those of (10). If this is the case, then salts of (N-nitramino)tetrazole
(9) will have greater heats of decomposition at the stage of acceleration of explosion, higher
pressures of the deflagration products, and higher initiating power. The properties of the
disilver salt of (9) studied experimentally, fully confirm this suggestion.
The fragment ions of the ammonium salts of (7) and (8) are oxidized still more readily.
The relative intensity of the carbon dioxide peak in their mass spectra is from two to three
times higher compared to (9) and from 20 to 30 times higher compared to (10). It should be
noted also that the relative intensity of the molecular nitrogen peak in the mass spectrum of
(7) is by a factor of ca. 2 lower compared to (8) and (9). Therefore, out of the studied
compounds the salts of (N-nitramino)tetrazoles (7) and (8) will have the highest heats of
decomposition at the initial stages and the shortest DDT periods, the latter being probably
somewhat longer for (7) because of a lesser nitrogen content in the initial decomposition
products. The experimental data on the initiating power of the silver salts of isomeric (Nnitramino)tetrazoles (7) and (8) are in full agreement with this hypothesis.
Thus, we showed experimentally that the length of the DDT period and thus the
initiating power of silver salts of isomeric (N-nitramino)tetrazoles (7-10) is controlled not
only by their Hfo values, but also by the structure and reactivity of their primary
decomposition products and by the composition of the gas phase in the course of
deflagration.
5.
LASER INITIATION OF METAL COMPLEXES

WITH 3(5)-HYDRAZINO-4-AMINO-1,2,4-TRIAZOLE AS LIGAND
In search for new photosensitive energetic coordination compounds and for the study of
the central ion effect of the on the sensitivity of metal complexes to laser mono pulse, we
prepared and investigated complex perchlorates of copper (II) (11), cobalt (II) (12), nickel
(II) (13) and cadmium (II) (14) with 3(5)-hydrazino-4-amino-1,2,4-triazole (HATr) as ligand
[15]
. The choice of the ligand is based on the high sensitivity of the perchlorate bis
(3(5)hydrazino-4-amino-1,2,4-triazole)copper (II) to laser mono pulse [16].
The prepared complex salts were tested for sensitivity to laser mono pulse (table. 2).
223
Table 2. Laser initiation of the synthesized complexes

Compound
Critical energy
Ecritical., J
The sum of two

ionization potentials of
the metal cation, I1+I2,
Ev
28.02
[Cu (HATr) 2] (CIO4)2
1.110-5
[Cd (HATr) 2] (CIO4)2
5.0310-4
25.90
[Ni (HATr) 2] (CIO4)2
5.7510-4
25.78
[Co (HATr) 2] (CIO4)2
1.3610-3
24.92
In this case Ecritical. is the minimum energy sufficient for the initiation of the complex
resulting in detonation.
On the basis of the results of testing it is possible to conclude that compounds (11-14)
may be regarded as energetic light-sensitive primary explosives. Their sensitivity to laser
mono pulse varies according to the oxidizing ability of the central metal ion, expressed as the
sum of two ionization potentials I1+I2 (Table. 2). The more this value, the higher the
sensitivity of a compound towards laser radiation:
Cu> Cd> Ni> Co.
It was established that at the initial stages of the process there occurs predominantly the
inner-sphere ligand oxidation effected by the metal cation. The results of our study allow to
assume that the following mechanism of oxidation takes place on the initial stages:
2 CIO4 + Me 2+ CIO4 + Me + + CIO4
CIO4 + Me + CIO4 + Me
The two-stage interaction of the metal cation and the perchlorate anion results in the
formation of a highly-active perchlorate radical participating at the second stage of ligand
oxidation.
Consequently, it is experimentally shown that for the metal salts of azoles there is no
universal factor governing the processes of deflagration under thermal and laser initiation.
Acknowledgments
The work was financially supported by the Ministry of Education of the Russian
Federation in the framework of the Intercollegiate Research and Development Program
Scientific Research of High School in the Field of Chemistry and Chemical Products.
224
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
Physics of explosion/ Ed M.N. Stanyukovich. Moskva.: Nayka. 1975 (in Russian).

Bowden F.P., Yoffe A.D. Fast reactions in solids. London. Butterworths.1958.
Sharma J., Coffey C.S., Ramaswamy A.L., Armstrong R.W. Atomic force microscopy of hot
spot reaction sites in impacted RDX and laser heated AP // Materials Research Society
Symposium Proc. (Detonation, Combustion and Detonation Chemistry of Energetic Materials.
Ed. T.B.Brill, T.P.Russell, W.C.Tao, R.B.Wardle). 1996. 418. 257-264.
Energetic Materials V.2. Technology of the Inorganic Azides / Ed.Fair H.D., Walker R.F. N.V. - London, Plenum press, 1977.
Danilov Yu.N., Ilyushin M.A., Tselinsky I.V. Industrial explosives. Part I. Primary explosives.
Text-book. Saint-Petersburg: Synthesis press. 2004. (in Russian)
Energetic Materials V.1.Physics and Chemistry of the Inorganic Azides / Ed Fair H.D., Walker
R.F. - N.Y. - London: Plenum press, 1977
Zeman S., Dimun M.M., Trachlic K.S., Kobatova V. The relationship between kinetics data of
low temperature thermolysis and the heats of explosion of inorganic azides//Thermochim.Acta
1984. v.80. p.137 - 141.
Trunks H., Shiff N. Gasdinamic investigations of lead azide/lead styphnate processes in
vacuum by multichannel mass spectrometry // Progress in Astronautics and Aeronautics. 1981 .
v.75 . . 242 - 252.
Ilyushin M.A., Tselinskii I.V. Primary explosives.//Ross. Khim. Zh. 1997, v.41, N 4, p. 3 - 13
(in Russian)
Avanesov D.S. Laboratory Manual on Physicochemical Tests of explosives. Moscow. GIOP.
1959.(in Russian)
Bates J.R. The potential of tetrazoles in initiating explosives systems //Proc. Symp. Explos.
Pyrotechn, 13-th. 1986. Paper 12 . P. III/1-III/13
Australian Patent no. 284707, 1968, Ref. Zh. Khim..1970. 4N842P
Ilyushin M.A., Terpigorev A.N., Tselinskii I.V. Effect of Isomeric Structure of (NNitrammo)tetrazoles on the Properties of Their Salts// Russian Journal of General Chemistry
1999. V.69. No.10. p.1654 1657.
ShurukhinYu. V., Klyuev N.A. and Grandberg I.I. Analogies in reactions of thermodestruction
and
mass-spectrometric
fragmentation
of
tetrazoles//
Khim.Geterotsikl.Soedin.
1985.no.6.p.723-741. (in Russian)
Ugryumov I.A., Ilyushin M.A., Tselinsky I.V., Kozlov A.S. Synthesis and properties of lightsensitive of perchlorate complexes of d-metals with 3(5)-hydrazino-4-amino-1,2,4-triazole as
ligand.// Russian Journal of Applied Chemistry 2003. vol.76. no.3. p.454-456 (in Russian)
Ilyushin M.A., Petrova N.A., Tselinskii I.V. The correlation between thermal decomposition
and laser ignition parameters for complexes of HAT.// Hunneng cailiao=Energetic
materials.1993. Vol.1. no 4. P.41-43.
225
CALCULATIONS OF DENSITIES AND HEATS OF FORMATIONS

OF ENERGETIC MOLECULES
FOR THE USE IN THERMOCHEMICAL CODES
Guy Jacob*, Stphane Bnazet*, Rolf Tryman**, Patrick Goede**,
and Henric stmark**
* SME Research Centre, 9 Rue Lavoisier, 91710 Vert le Petit, France
** FOI, Swedish Defence Research Agency, Department of Energetic Materials,
S-147 25 Tumba, Sweden.
Abstract:
Performance of energetic molecules is evaluated by thermo chemical codes which need
the input of density, heat of formation and formula. Density and heat of formation for
target compounds are values that need to be determined. In this paper two different
methods to calculate these values have been applied to a common set of known energetic
molecules. These methods are briefly described and the results from these calculations
are compared.
Keywords: QM-calculations, semi empirical-calculations, heat of formation, density
1.
INTRODUCTION
In a joint research program between France and Sweden the possibility of developing
energetic materials for the future is being explored. The two main driving forces for the
development of energetic materials are lower sensitivity and increased performance. Until
now, new energetic materials have either been less sensitive with retained performance
compared to current materials or more powerful but also more sensitive (Figure 1). A longterm objective is of course to merge these two qualities and produce energetic molecules that
lead to both more powerful and less sensitive materials than the current ones. Compounds
with the performance of CL-20 and the sensitivity of TATB would be desirable long-term
goals. The most obvious military applications of new molecules are in the explosive and
propellant fields, leading to higher performance of the systems where these compounds will
be incorporated (detonation energy, specific impulse, burn rate, operational temperature and
range) and better safety (insensitive munitions requirements).
226
Current
RDX, HMX
TNT
Fig 1.
Sensitivity
FOX-7, NTO, TATB
Performance
CL-20
Lower Sensitivity
Increased Performance
Development of new explosives
To focus the laboratory efforts on synthetic targets with desired performance it is

necessary to have computational tools that help to evaluate the potential of the target
molecules. Computational methods allow quite accurate prediction of density and heat of
formation, which makes it possible to evaluate the performance of large numbers of
candidate molecules before their preparation is attempted. It is more difficult to predict the
sensitivity of new energetic materials, which leads to a more qualitative analysis of the
proposed target molecules. This input will help the difficult decisions of deciding which
synthetic target molecules to pursue.
Performance of energetic molecules is evaluated by thermo chemical codes which need
the input of density, heat of formation and formula. Density and heat of formation for target
compounds are values that need to be determined. In this paper two different methods to
calculate these values have been applied to a common set of known energetic molecules.
These methods are briefly described and the results from these calculations are compared.
1.1
CRB-method description
Density is estimated by computing the molecular volume with a molecular mechanics

code, using a traditional force field enlarged to functional groups present in energetic
molecules. Fast evaluation of the heat of formation is performed with semi-empirical PM3
code. It allows calculations on a large number of molecules with an accuracy that is
sufficient to direct the synthesis work on the most promising target molecules.
1.2
FOI-method description
Quantum mechanical calculations (Gaussian 98) were used to theoretically optimise the
geometry and calculate the total energy of the common set of molecules. The level of theory
used was B3LYP with the basis set 6-31G(d). For theoretical estimates of the density, the
program MOLPAK (MOLecular PAcKing) from the University of Maryland [1] was used. To
theoretically estimate the heat of formation (Hf(Solid)) a method, developed by Betsy Rice
et al, U.S. Army Research Lab, Aberdeen, was used [2,3].
227
2.
CALCULATION METHODS
2.3
CRB Calculation methods
Traditional methods to evaluate physico-chemical properties like density and heat of

formation use increments for each atom or functional groups that are present in the molecule;
a review of these methods for density estimation has been presented [4]. Despite their
advantages in term of the speed and easiness, they lack generality and fail if new groups are
imagined. For this reason, it has been searched for methods based on the computation of
geometry with molecular or quantum mechanics.
2.3.1. Density
Molecular volume is a value that can be evaluated after the optimization of the geometry
by every kind of computation. From that, density is easily deduced. By using molecular
mechanics, it appeared that calculated density is linearly correlated to experimental
measurements[5]. The slope of the line has been firstly determined for non energetic
molecules. A good correlation was obtained by partitioning products between solids and
liquids. For energetic compounds, the same sort of correlations was also observed, based on
142 molecules. This procedure [6] involves MM2 force field with added parameters specific
for energetic groups. Calculation of the volume of the molecule is performed by enclosing
the molecule in a box and dividing it into small cubes that are then scanned to detect the
presence of an atom in it.
2.3.2. Heat of formation
Gas phase heat of formation is readily obtained from semi-empirical quantum mechanics
using the PM3 hamiltonian whose parameterization largely included data from heat of
formation and then affords more reliable values than older semi-empirical methods. For
computational uses, Hyperchem 6.0 was used. First, the structure is optimized by molecular
mechanic with an MM+ force field. Then energy minima of conformers are searched by
molecular dynamics and then energy is minimised using molecular mechanics. The most
stable conformer is retained from this step. Next, geometry is optimized with PM3 method.
Minimisation algorithms are first steepest descent for speed and then, near minima of
energy, conjugate gradient (Polak-Ribire or Fletcher-Reeves for larger molecule).
Minimum of the energy on the potential surface is checked by vibration analysis. If a large
number of molecules are to be screened, a small interface of Hyperchem with Tcl has been
written. It allows reading each structure from standard chemical drawing program,
converting them to Hyperchem file, optimizing the structures and compute properties as
described before in a batch process. Several papers have presented estimations of heat of
formation with PM3 but were limited in each case to one family or sort of energetic
molecules [7-10].
2.4
FOI Calculation methods
Quantum mechanical calculations were used to theoretically optimise the geometry of a

number of interesting molecules, to estimate the density () and to estimate the heat of
formation (Hf(Solid)). Calculations for a common test set of 19 reference molecules are
presented.
Most of the quantum mechanical calculations were done on a 8 processor PC-cluster.
The H(Sublimation) and the density calculation with MOLPAK were done on a Compaq
Alpha Workstation.
228
2.4.1. Geometry
The geometry needs to be calculated as a first step in determining other properties of a
molecule. The geometry is optimized by varying the geometry until a stationary point on the
potential energy surface (PES) is found. This can be done at different levels of theory
depending on the needed accuracy. Higher accuracy costs more in terms of computer time.
Therefore, it is necessary to weigh the cost to the desired accuracy when choosing method.
At FOI, a hybrid method between Density Functional Theory (DFT) and Hartree-Fock
approximation (HF) called B3LYP is normally used for comparative purposes and as a
starting point. The basis set normally used is 6-31G(d). After the geometry has been
determined, other properties, such as heat of formation, stability, spectral properties (IR,
Raman, NMR) density and state of aggregation, can be calculated at the determined
geometry. This method does not give the best possible values for the heat of formation for
these methods but it is good enough for comparative purposes and to estimate the potential
of a substance.
Different computer codes are available for this type of calculations. Commonly used soft
wares are various versions of Gaussian, at FOI normally 98. To generate good input data to
Gaussian 98 the molecules were optimized in Chem-3D by the semi-empirical program
MOPAC on a PC. The quantum mechanical calculations in Gaussian were done on an 8
processor LINUX-cluster. The calculation time, depending on the size of the molecule, was
10 to 24 hours. The result is the optimized geometry and total energy of the molecule. If any
of the calculated vibrational frequencies are negative it means that the stationary point is not
a minimum on the PES but a saddle point, i.e. a minimum with respect to some directions on
the surface and a maximum in others. If all the frequencies are positive, the calculated
minimum is a true minimum. If there are very low frequency vibrations present, this could
indicate a shallow minimum on the PES, i.e. limited stability. All the reference molecules
vibrational frequencies were positive.
2.4.2. Heat of formation
A number of methods have been developed for the calculation of gas phase heats of
formation. These methods range from advanced ab initio schemes to simple bond additivity
models. Ab initio schemes such as the G2 and G3 models of Pople and coworkers [11] and the
CBS models of Petersson and coworkers [12] are able to predict heat of formations to an
accuracy of around 2 kcal/mol. However, these are very computationally demanding and can
only be used for molecules with up to between seven and eight non hydrogen atoms. A less
demanding approach is the use of density functional theory with B3LYP parametrization for
heat of formation calculations. The method uses quantum mechanical calculations, where
Hf(g) at 0K is determined as the difference between the total energy and the sum of the
atom equivalents. The atom equivalents are semi-empirically calculated from quantum
mechanical calculations and correlated with experimental values. The approach is based on
the scheme of Habibollahzadeh et. al. [13]. To calculate the solid state heat of formation it is
necessary to subtract the sublimation enthalpy from the gas phase heat of formation
(Equation 1).
Hf(s) = Hf(g) - Hf(subl)
(1)
The sublimation enthalpy can be estimated from quantum mechanical calculations where
the molecular surface area and the electrostatic potential on the surface are used to generate
two statistically based quantities which describe the polarity of the molecule. These
quantities together with the surface area are correlated by three empirical parameters to
predict the sublimation enthalpy. This approach was developed by Politzer and Murray [14].
At FOI, a program, hs95fo made by Tore Brinck at the Royal Institute of Technology in
229
Stockholm, is used to calculate the sublimation energyHf(s). These calculations are made
on a Compaq work station. The heat of formation estimations described in this work were
essentially implemented for energetic materials use by Rice et al.[2]. When evaluated by Rice
et al the root mean square(rms) deviation of the solid heats of formation from 75
experimental values corresponding to 44 molecules was 9.0 kcal/mol with a maximum
deviation of 35.4 kcal/mol [2].
The estimation of the heat of formation done on our selected basis set is however more
ambiguous. For 8 molecules the difference between calculated and measured Hf(Solid) was
less than 6 kcal/mol. For the others the differences are between 10 and 30 kcal/mol. The
average divergence of the calculated Hf(Solid) from the experimental values [15] was 13.6
kcal/mol, with a standard deviation of 11.4 kcal/mol (based on 17 molecules).
2.4.3. Density
For theoretical estimates of the density, the program MOLPAK (MOLecular PAcKing)
from the University of Maryland [1] was used. MOLPAK uses the calculated electrostatic
potential surfaces from Gaussian98 (B3LYP/6-31G(d)) as the size of a rigid-body
representing the volume of the molecule. This molecule is used to build packing arrangement
of possible crystal structures in various space groups. MOLPAK uses group theory to pack
the molecules in 29 different space groups and calculates the minimum energy for each
space group. The space group which gave the lowest energy was considered the most
probable packing and the corresponding density was estimated as the theoretical density. The
MOLPAK calculations were done on a Compaq work station with a process time of 15-35
hours per molecule.
A comparison between calculated (calc) and experimental (exp) determined density
shows that on average the calculated density is 0.98 of the experimental density, with a
deviation of 4.2% based on 19 molecules.
We can use following formula to estimate the experimental density:
exp=1.02calc
3.
(2)
COMMON TEST SET
In order to compare the different codes used by FOI and CRB, a common list of test
molecules was established and calculations of physical chemical properties were performed
by the two parties. The common test set includes nitramines, aliphatics, aromatics and
different heterocycles. This highly divers set of energetic molecules was made up of 19
reference molecules.
230
3.1
Common test set of molecules
Table 1. Common test set of molecules

Name (formula)
Structure
NO2
Keto-RDX
(C3H4N6O7)
O2N
NO2
FOX-7
(C2H4N4O4)
H2N
NO2
H2N
NO2
NO2
N
RDX
(C3H6N6O6)
O2N
CL-20
(C6H6N12O12)
O2N
O2N N
N
NO2
NO2
N NO2
N
O2N
NO2
NO2
TEX
(C6H6N4O8)
NO2
N
O2N
NTO
(C2H2N4O3)
NH
HN
O
O2N
TNGU (Sorguyl)
(C4H2N8O10)
NO2
N
O
N
O2N
NO2
O2N
NO2
TNDA
(C4H6N6O8)
O2N
231
NO2
NH2
LAX-112
(C2H4N6O2)
N+
N+
NH2
NH2
TATB
(C6H6N6O6)
O2N
NO2
H2N
NH2
NO2
NH2
HN
NGU
(CH4N4O2)
NH
O2N
O2N
HMX
(C4H8N8O8)
NO2
O2N
NO2
O2N
DINGU
(C4H4N6O6)
H
N
N
H
NO2
NO2
O N
Dinitrobifurazane
(C4N6O6)
N
N O
NO2
O2N
N-DNAT
(C4H2N10O4)
N
N
N
NO2
O2N
HHTDD
(C6H4N12O14)
NO2
O
N
O2N
O2N
DATNTAN
(C5H10N14O8)
N
N
N3
NO2
NO2
N
NO2
NO2
N
NO2
232
N3
NO2
3.2
Comparison of common test set of molecules
Table 2. Calculated and Experimental data for common test set of molecules
Calculated CRB Calculated FOI
Literature data
Density
Hf (s)
Hf (g) Density Hf (g) Hf (s) Density
15,
(g/cm3)
3
3 kcal/mol
kcal/mol (g/cm ) kcal/mol kcal/mol (g/cm ) 16
15,16
Name
Keto-RDX
FOX-7
RDX
CL-20
TEX
NTO
TNGU (sorguyl)
TNDA
LAX-112
TATB
NGU
HMX
DINGU
Dinitrobifurazane
N-DNAT
HHTDD
DATNTAN
4.68
-11.18
40.33
107
-113.33
-3
-12.96
26.08
93.06
-9.93
14.84
63.64
-13.01
154.45
190.17
49.87
204.88
1.868
1.824
1.756
2.051
1.927
1.851
2.056
1.818
1.794
1.891
1.725
1.788
1.872
1.991
1.813
2.089
1.726
22.51
-4.4
45.04
142.53
-76.07
2.26
24.01
24.4
88.82
7.22
18.28
62.16
-26.2
113.74
193.88
101.48
192.85
- 5.2
-30.6
20.69
96.92
-126.92
-16.04
-9.44
-4.09
65.09
-20.41
-7.85
27.27
-55.38
93.32
159.14
56.48
142.4
1.905
1.900
1.772
1.865
1.935
1.955
1.953
1.794
2.052
1.963
1.798
1.747
1.901
1.856
1.753
2.014
1.632
- 10
-32
16.8
98
-129, -106
-25.8
-8, 13
-4.3, 12.7
39.2
-33.8
-20.7
20.1
-82, -42
101
94.4
36.7
163.54
2,1
SORGUYL
LAX112
Density SNPE
Density FOI
DNBIF
calculated density
CL20
1,9
CL20
1,8
HMX
1,7
1,6
1,6
1,7
1,8
1,9
observed density
Fig 2.
Plot of calculated versus observed density
233
2,1
1.932
1.885
1.816
2.04
1.99
1.93
2.01
1.82
1.86
1.937
1.77
1.902
1.98
1.85
1.831
2.07
1.67
estimated Enthalpy of formation
250
200
150
DHf gas SNPE
100
DHf gas FOI
50
0
-150
-50
50
150
250
-50
DHf condensed FOI
-100
-150
measured Enthalpy of formation
Fig 3.
4.
Plot of calculated versus observed heat of formation
CONCLUSIONS
The FOI method for density calculations underestimates compounds with large
differences between polymorphs like CL-20 and HMX. It also over estimates LAX-112, as
this is the only reference compound of this type there is a need for more calculations on
similar compounds. The SNPE density method works very well for compounds which are
well parameterised. As furazans are not fully parameterised there is a bad correlation with
Dinitrobifurazane. The FOI method is however easier to apply on totally new classes of
compounds. When comparing the whole test set of molecules the density predictions are
good.
On average the heats of formation in the gas phase as calculated both by the FOI and
SNPE methods are quite good. Heats of formation in the condensed phase as calculated by
the FOI method are also quite good. It is important to mention that experimental values for
the heat of formation may vary quite a lot, depending on the source of the information.
These methods for the determination of heat of formation and density where applied to a
set of potential target molecules and their theoretical performance was estimated. Further
factors influenced the decisions of which molecules were selected for actual synthesis work.
These were the potential cost of producing the compounds, the delay of availability based
upon synthetic difficulties and their potential vulnerability.
Acknowledgments
Financial support from DGA and the Swedish Armed Forces is greatly acknowledged
234
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
J. R. HOLDEN, Z. DU, H. L. AMMON: Prediction of Possible Crystal Structures for C-, H-, N-,
O-, and F-Containing Organic Compounds, Journal of Computational Chemistry, 14, 422-437,
1993.
B. M. RICE, S. V. PAI, J. HARE: Predicting Heats of Formation of Energetic Materials Using
Quantum Mechanical Calculations, Combustion and Flame, 118, 445-458, 1999.
M. J. FRISCH, H. B. S. G. W. TRUCKS, G. E. SCUSERIA, M. A. ROBB, , V. G. Z. J. R.
CHEESEMAN, J. A. MONTGOMERY, JR., , J. C. B. R. E. STRATMANN, S. DAPPRICH, J. M.
MILLAM, A. D. DANIELS,, M. C. S. K. N. KUDIN, O. FARKAS, J. TOMASI, V. BARONE, M.
COSSI, R. CAMMI,, C. P. B. MENNUCCI, C. ADAMO, S. CLIFFORD, J. OCHTERSKI, G. A.
PETERSSON,, Q. C. P. Y. AYALA, K. MOROKUMA, D. K. MALICK, A. D. RABUCK, K.
RAGHAVACHARI, , J. C. J. B. FORESMAN, J. V. ORTIZ, B. B. STEFANOV, G. LIU, A.
LIASHENKO,, I. K. P. PISKORZ, R. GOMPERTS, R. L. MARTIN, D. J. FOX, T. KEITH,, C. Y. P. M.
A. AL-LAHAM, A. NANAYAKKARA, C. GONZALEZ, M. CHALLACOMBE,, B. J. P. M. W. GILL,
W. CHEN, M. W. WONG, J. L. ANDRES, C. GONZALEZ,, E. S. R. M. HEAD-GORDON, AND J. A.
POPLE: Gaussian 98, Revision A.3. Gaussian, Inc., 1998.
G. PIACENZA, G. JACOB, R. GALLO: The Prediction of Density, in: M. Charton, B. I. Charton
(Eds.), Advances in quantitative structure-property p 67-108, 2002.
G. PIACENZA, G. JACOB, H. GRAINDORGE, B. BLAIVE, R. GALLO: ICT Annual Conference,
Karlsruhe, p.123-121 - 123-114, 1997.
B. BLAIVE, G. LEGSAI, J. BOILEAU: Europyro, 1995.
Y. AKUTSU, M. TAMURA: Journal of Energetic Materials, 11, 39-46, 1993.
Y. AKUTSU, R. CHE, M. TAMURA: Journal of Energetic Materials, , 11, 195-204, 1993.
Y. AKUTSU, M. TAMURA: Journal of Energetic Materials, 11, 205-208, 1993.
J. L. DE PAZ, J. CILLER: Propellants, Explosives and Pyrotechnics, 18, 1993.
L. A. CURTISS, P. C. REDFERN, K. RAGHAVACHIRI, J. A. POPLE: Gaussian-3X (G3X) Theory:
Use of Improved Geometries Zero-Point Energies and Hartree-Fock Basis Sets., Journal of
Chemical Physics, 114, 108, 2001.
J. A. MONTGOMERY, M. J. FRISCH, J. W. OCHTERSKI, G. A. PETERSSON: A Complete Basis Set
Model Chemistry. VII Use of the Minimum Population Localization Method., Journal of
Chemical Physics, 112, 6352, 2000.
D. HABIBOLLAHZADEH, M. E. GRICE, M. C. CONCHA, J. S. MURRAY, P. POLITZER: Nonlocal
Density Functional Calculation of Gas-Phase Heats of Formation, Journal of Computational
Chemistry, 16, 654-658, 1995.
P. POLITZER, J. S. MURRAY, M. E. GRICE, M. DESALVO, E. MILLER: Calculation of heats of
sublimation and solid phase heats of formation, Molecular Physics, 91, 923-928, 1997.
ICT Database of Thermochemical Version 2004
http://webbook.nist.gov/
235
CHARACTERIZATION OF DETERRED PROPELLANTS

BY CLOSED VESSEL TESTS: IMPORTANCE
OF THE IGNITION METHOD
L. Jeunieau*, M.H. Lefebvre*, M.C. Pirlot**, and P. Guillaume***
* Laboratory for Energetic Materials, Ecole Royale Militaire,
Avenue de la Renaissance 30, 1000 Bruxelles (Belgium)
** Department of Weapon Systems and Ballistics, Ecole Royale Militaire,
Avenue de la Renaissance 30, 1000 Bruxelles (Belgium)
*** PB Clermont SA, Rue de Clermont 176, 4480 Engis, (Belgium)
Abstract:
Different gaseous ignition systems have been used for the characterisation of spherical
deterred propellants in closed vessel tests. It has been observed that, with an
appropriate ignition system, a good correlation is obtained between closed vessel tests,
deterrent concentration profiles and ballistic firing.
Keywords:
1.
deterred propellant, closed vessel tests, gaseous ignition,

IR microscopy, ballistic firing
INTRODUCTION
The ballistic performance of guns can be improved by using of less degressive

propellants. For this purpose, deterred propellants are used especially in small arms. In
deterred propellants, the concentration of deterrent in the outer layer is larger than in the core
of the propellant particles. The deterrent gradient results in a larger pressure peak during the
combustion process, compensating for the decreasing burning surface by an increasing
burning rate.
The characteristics of the deterrent concentration gradient have an effect on the
propellant combustion behavior. These concentration profiles can be obtained by infrared
microscopy analysis but this procedure is highly time consuming. Standard closed vessel
tests with black powder as ignition mixture do not permit to observe significant differences
between the combustion of the two chemical compositions of the propellant.
In this work, different gaseous ignition systems have been investigated. These ignition
systems have been tested on propellants with different deterrent concentrations. The
combustion rates have been calculated and the combustion features have been correlated
with the deterrent concentration profiles obtained by infrared microscopy. Furthermore,
ballistic firings have been performed, and their results have been compared with both the IR
analysis and the results of closed vessel test.
236
2.
EXPERIMENTAL
2.1
Propellant used
Several formulations of deterred spherical propellants have been used in this work.
Their characteristics are reported in Table 1. The used deterrent is the dibutylphthalate
(DBP). Some of the propellants have all the deterrent in the outer layer of the particles,
some have also dibuthylphthalate in the core of the particle.
The used propellants can be sorted in two types as a function of their size distribution,
one with an average diameter of about 0.7 mm and one with an average diameter of about
1 mm.
Table 1. Characteristics of the investigated propellants
Average % DBP**
in the ball powder*
Diameter (mm)
Propellant A
Propellant B
Propellant C
Propellant D
Propellant E
Propellant F
Propellant G
Propellant H
Propellant I
Propellant J
*
**
2.2
0.7
0.7
0.7
0.7
0.7
0.7
1.0
1.0
1.0
1.0
7.4
1.3
7.4
12
16.9
18.5
7.0
7.4
8.5
7.9
Local % DBP**
in the core
2
1.8
determined experimentally by HPLC

dibutylphthalate
Closed vessel experiments
Closed vessel experiments were carried out in a vessel of 140 cm3 using a piezoelectric
pressure transducer (Kistler 6201B4) for recording the pressure. The output voltage of the
pressure gauge was transferred to a data acquisition system (Nicolet Multipro, resolution 12
bit, sampling frequency 250 KHz). The ignition system consists of two electrodes, which
are connected with a nickeline hot wire. The vessel is equipped with a valve to introduce the
gaseous ignition mixture. The partial pressures of these mixtures are measured with a piezoelectric transducer (Kistler 4070). When black powder is used as ignition method, 1g of
powder is used. The combustion rate is calculated according to Stanag 4115[1].
The compositions of the various ignition mixtures are given in Table 2. The gaseous
ignition systems are characterized by their partial pressures (for the actual composition, the
initial atmospheric pressure must be added). In this table, the maximum pressure of the
ignition mixtures alone in the closed vessel, the oxygen balance and the energy of the
ignition mixtures are mentioned. They are calculated using Eq.1.
E = Cv*T*migniter
(1)
237
where Cv, T , migniter are the specific heat at constant volume, the difference between
the adiabatic explosion temperature and the ambient temperature and the mass of ignition
mixture respectively.
Others thermodynamical characteristics of these ignition systems are reported by
Jeunieau et al [2].
Table 2. Characteristics of the different investigated ignition mixtures
Composition
Mixture I
Mixture II
Mixture III
Mixture IV
Mixture V
Mixture VI
2.3
Pmax
(MPa)
0.1 MPa CH4- 0.2 MPa O2
5.4
0.1 MPa CH4 0.14 MPa O2
4.7
0.05 MPa CH4- 0.075 MPa O2
2.5
0.1 MPa CH4 0.1 MPa O2
3.6
0.05 MPa CH4 0.05 MPa O2
1.9
Black powder (loading density 2.6
0.007g/cm3)
Energy
(J)
3252
2841
1632
2308
1400
2186
Oxygen
balance (%)
+5.2
-13.7
-1.9
-32.6
-1.9
-9.7
Infrared microscopy
The deterrent concentration profile is measured by InfraRed (IR) microscopy: the

propellant grain is placed into an adhesive, cut by a microtome into small slices (7 m) and
analyzed by IR microscopy. A Bruker Hyperion infrared microscope mounted on a Vector
33 Fourier-transform spectrometer is used in this study. A medium-band MCT (HgCdTe)
detector in the microscope gives high sensitivity in the 4000-600 cm-1 range. A 15X
cassegrain mirror objective is used to obtain the infrared spectra. The IR spectrometer is
operated at a resolution of 4 cm-1 and 32 scans are acquired for each measurement position.
The measuring window has an aperture of 10 m x 50 m, the larger side of the aparture is
placed perpendicularly to the measured diameter. The step for each data point is 3 m. The
quantitative DBP concentration is obtained using the ratio of two IR bands, one typical of
the DBP (1720 cm-1) and one typical of the nitrocellulose (1160 cm-1).
2.4
Ballistic firing
The experimental work is carried out in a .50 inch weapon instrumented with a piezoelectric transducers (Kistler 6215) coupled to a charge amplifiers (Kistler 5011A) in order to
record the pressure. The pressure measurement is located in the cartridge case. The output
charges of the pressure gauge is amplified and filtered to get an output voltage, which is
transferred to a data acquisition system (resolution 12 bit, sampling rate 1 MHz). 13 g of the
propellant are poured in a standard 12.7 x 99 mm cartridge. The corresponding loading
density is 0.75 g/cm3.
3.
RESULTS
3.1
Ignition mixtures
To investigate the role played by the ignition mixture, propellant A (see table 1) has
been used at a loading density of 0.20 g/cm3. Figure 1 shows for some propellants the
pressure as function of time and their corresponding derivatives. It can be noticed that
important combustion instabilities are present for the ignition mixture I. As these
238
instabilities are present in the calculated combustion rate, this ignition mixture has to be
disregarded. The pressure derivative permits to distinguish clearly two phases in the
combustion process when mixture II or III are used as ignition system. In fact, two different
positive slopes can be observed in the pressure derivative vs. time.
For the sake of clarity, these slopes have been emphasized by the two drawn lines. In other
cases, the variation of the derivative vs. time is more regular, the pressure increase is
smoother and the combustion process can no more be divided into two parts. One
hypothesis to explain the lack of discontinuity is the inhomogeneous ignition of the
propellants by the black powder. When ignition is inhomogeneous, the deterred layers of all
the particles do not burn simultaneously. When one particle has its deterred layer burnt and
therefore its combustion rate increased the other particles may have their deterred layer not
completely burnt. Thus the observed combustion rate is more or less a smooth average of
the overall combustion rate.
300
(I)
160
(III)
140
dP/dt (MPa/ms)
Pressure (Mpa)
250
200
(VI)
150
100
120
80
60
40
20
0
-2
0
2
4
time (ms)
-4 -3 -2 -1 0 1 2
time (ms)
(1)
Fig 1.
(VI)
100
50
-4
(II and III)
(2)
(1) Pressure time histories and (2) derivative of pressure vs. time of propellant
A for different ignition systems and for a loading density of 0.20 g/cm3.
Numbers are refering to the formulations as reported in table 2.
Figure 2 shows the calculated combustion rate of propellant A when different ignition
mixtures are used. The largest combustion rates of the second part of the combustion, i.e.
the combustion rate of the inner part of the propellant grain, are obtained when mixtures II
and III are used. For these mixtures two parts in the combustion rate vs. pressure can also be
observed. When ignition mixture II is used, instabilities are present in the curve; this can be
related to the high energy of this ignition mixture (see table 2). For the others gaseous
ignition systems, a similar combustion rate to the one obtained with black powder are
obtained.
For the observation of a discontinuity, two factors seem important: the total ignition
energy and the oxygen balance (Table 2). For mixture IV, there is a great lack of oxygen
and no discontinuity is observable despite the large energy value. For mixture V, the oxygen
balance is close to zero but the energy is low. For mixture II, the oxygen balance is
negative, but as the energy is high, a discontinuity is observable in the combustion rate. For
mixture VI, the type of ignition mixture is different and the pressure history has to be taken
into account. In fact, the pressure increase of the ignition mixture is much slower than for
239
the gaseous ignition mixture (see ref [2]). This could explain an inhomogeneous ignition
and therefore the absence of a discontinuity in the combustion rate curve.
Combustion rate (cm/s) .
25
20
(III)
(I)
(II)
(IV)
15
10
(V)
(VI)
5
0
0
50
100
150
200
Pressure (MPa)
Fig 2.
Combustion rate of propellant A obtained with different ignition mixtures

I to VI as defined in table 2. The loading density is 0.20 g/cm3.
From these results, the ignition mixture III and II have been selected and will be tested
on different propellants. The mixture VI will also be used, as it is used traditionally in
closed vessel tests.
At this stage of the discussion, the fact that the observed discontinuity in the derivative
curve corresponds to the burning of the deterred layer has to be confirmed. This will be
done by using different propellants with different deterrent concentrations and different
concentration profiles. If the discontinuity corresponds to the burning of the deterred layer,
its location will vary with the deterrent concentration profile.
3.2
Correlation between closed vessel tests and infrared microscopy
The propellants have been divided in two series, one with an average diameter of 0.7
mm and one with an average diameter of 1.0 mm.
3.2.1 Propellant with an average diameter of 0.7 mm
Figure 3 shows the deterrent concentration profiles of the four propellant formulations
with an average diameter close to 0.7 mm.
240
%DBP
20
18
16
14
12
10
8
6
4
2
0
Propellant C
Propellant D
Propellant E
Propellant F
Fig 3.
50
100
150
200
250
Deterrent concentration profiles of the propellant with an average diameter of

0.7 mm. The vertical lines corresponds to the values of the deterrent
penetration depth calculated from the closed vessel tests (see further in the text).
Figure 4 shows the combustion rate of propellants C, D, E and F obtained with the
ignition mixtures II, III and VI. Fig. 5 shows the corresponding derivative curves. It can be
observed that if mixture VI is used, no discontinuity is observed in the derivative curve and
the variation of the combustion rate with pressure evolves regularly.
When mixture III is used, an important sharp discontinuity is observed for propellant C
in the combustion rate curve and in the derivative curve. A less important discontinuity is
observable for propellant D which its consistent with its deterrent concentration profile.
If mixture II is used, the discontinuity is observable for propellant D, E and F. As
propellant E and F have a higher deterrent concentration and a smaller percentage of
propellant volume without deterrent, a more energetic mixture (see table 2) is needed for the
observation of the two propellant chemical compositions.
14
14
(D)
10
8
(E)
6
4
2
(F)
0
0
50
100
Pressure (MPa)
(1)
Fig 4.
(C)
Combustion rate (cm/s)
12
Combustion rate ( cm/s)
14
12
10
(D)
(E)
6
4
2
(F)
0
0
50
100
Pressure (MPa)
(2)
12
(D)
10
(C)
(E)
6
4
2
(F)
0
0
50
100
Pressure (MPa)
(3)
Combustion rate of Propellant C, D, E and F (gray line). The labels in brackets

refer to the type of propellant. The ignition mixture is (1) igniter II, (2) igniter
III and (3) igniter VI. The loading density is 0.15 g/cm3.
241
90
90
80
80
(D)
40
30
20
70
60
50
(D)
40
(E)
30
(F)
-2
3
time (ms)
(1)
(C)
60
50
40
(E)
(D)
30
20
20
10
dP/dt (MPa/ms)
(E)
50
-7
80
70
60
dP/dt (MPa/ms)
dP/dt (MPa/ms)
70
90
(C)
10
(F)
0
-7
10
(F)
-2
3
time (ms)
-7
-2
3
time (ms)
(2)
(3)
Fig 5. Derivative of the pressure vs. time of Propellant C, D, E and F (gray line).
The labels in brackets refer to the type of propellant. The ignition mixture is
(1) igniter II, (2) igniter III and (3) igniter VI. The loading density is 0.15 g/cm3.
The brunt mass fraction corresponding to the discontinuity in the derivative time curve
can be calculated using Eq. (2)[3].
1 1

z break =
Pmax Pmin 1
P
1
1 Pmin
+ max +
Pbreak
Pbreak Pbreak
1 1
(2)
Pmin
Pmax Pmin 1
1
Pbreak + Pbreak Pmax + Pmin
where , , Pmax, Pmin, Pbreak, are the loading density, the propellant density, the
experimental maximal pressure, the pressure due to the ignition system, the pressure
corresponding to the discontinuity in the derivative curve and the covolume respectively. In
this equation is the ratio between the theoretical maximal pressure of the propellant, which
has the chemical composition of the deterred propellant, and the theoretical maximal
pressure, which has the actual composition of the propellant. More information about this
equation can be found in [3].
As the burnt mass fraction is defined by Eq.(3) where V is the propellant volume and V0
is its initial value, the burnt thickness can be easily calculated.
z = 1
V
V0
(3)
The obtained results are given in table 3. These values are in good agreement with the
deterrent penetration depths measured by infrared microscopy, excepting for propellant E
and F (values for propellant F should be greater than the one of propellant E). But the
difference between these two values is lower than the experimental uncertainty (at least
10m). Theses values have been added to the chart of the deterrent concentration profile
242
(Fig.3). It is observed that the calculated deterrent penetration depth are generally not
situated at the end of the deterred layer (Fig.3) but before the large drop of the deterrent
concentration. In fact, the discontinuity in the derivative curve occurs between the
combustion of a more or less constant deterred layer and the combustion of a less deterred
layer with an abrupt variation of the deterrent concentration.
Table 3. Values of the deterrent penetration depths calculated from the closed
vessel experiments. Experiments with two ignition mixtures are reported.
Deterrent penetration depth (m)

Igniter II
Igniter III
Propellant C
70
Propellant D
124
127
Propellant E
154
Propellant F
146
3.2.2 Propellant with an average diameter of 1.0 mm

A study similar to the one described in the previous section has been performed on the
propellants with the larger average diameter. Figure 6 shows the deterrent concentration
profiles of the propellant with an average diameter of 1.0 mm.
%DBP
16
14
Propellant G
12
Propellant H
Propellant I
10
Propellant J
8
6
4
2
0
0
Fig 6.
50
100
150
200
m
250
300
350
400
Deterrent concentration profiles of Propellant G, Propellant H, Propellant I

and Propellant J. The vertical lines correspond to the deterrent penetration
depths calculated from the closed vessel tests.
Figure 7 shows the corresponding combustion rate obtained with ignition mixtures III
and VI and Fig.8 shows the corresponding derivative curves. When ignition mixture VI is
used, a difference is observed only between propellant G, H and propellant I, J. If ignition
mixture III is used, a difference is observed between all the propellants. For this ignition
mixture, a discontinuity is observable for all the propellants in the derivative curves and the
order of appearance of this one corresponds to the difference of deterrent penetration depth.
243
20
18
16
14
12
10
8
6
4
2
0
(G)
(H)
(J)
(I)
0
50
100
Pressure (MPa)
20
15
(G),(H)
10
150
(I),(J)
0
0
50
100
Pressure (MPa)
(1)
Fig 7.
150
(2)
Combution rate of Propellant G, H, I and J. The labels in bracket refer to the

type of propellant. The igniton mixture for (1) is mixture III and for (2) is
mixture VI. The loading density is 0.20 g/cm3. The curves of propellant H and
J are in gray.
.
120
(J)
80
60
40
(I)
20
0
-5
0 (ms)
time
(I),(J)
-5
(1)
Fig 8.
(G),(H)
dP/dt (MPa/ms)
(G)
100
dP/dt (MPa/ms)
120
100
80
60
40
20
0
(H)
0
time (ms)
(2)
Derivative of the pressure vs. time of Propellant G, H, I and J. The ignition

mixture for (1) is mixture II and for (2) is mixture VI. The loading density is
0.20 g/cm3. The curves of Propellant H and J are in gray.
The calculated deterrent penetration depths from Eq.(1) and (2) are listed in table 4 and
these values are located on the deterrent concentration profiles in Fig.6, accounting for a
good agreement with the actual deterrent profiles.
244
Table 4. Values of the deterrent penetration depth calculated

from the closed vessel tests. Mixture III is used as ignition mixture.
Deterrent penetration depth (m)

Propellant G
94
Propellant H
113
Propellant I
119
Propellant J
112
3.2.3 Correlation between the two series of propellant

In Fig. 9, the obtained combustion rate at 30 MPa has been plotted as a function of the
average DBP concentration (obtained by HPLC) and of the percentage of DBP in surface for
ignition mixtures III and VI. The combustion rate at 30 MPa corresponds to the burning of
the outer surface of the propellant. It can be observed that only a roughly linear variation of
the combustion rate as a function of the average concentration can be observed when ignition
mixture VI is used. If this combustion rate is plotted as a function of the surface deterrent
concentration, a linear variation can be observed as well for mixture III as for mixture VI. In
this chart, the linear trendline is calculated using only the propellant with an average
diameter of 0.7 mm. The values corresponding to the propellant with an average diameter of
1.0 mm put themselves nicely on these trendlines. The slope of this trendline is smaller for
mixture VI than for mixture III. This could be an explanation of the better differentiation of
deterred propellant by mixture III than by mixture VI. As the variation of the combustion
rate with the deterrent concentration is greater for mixture III than for mixture VI, two
propellants with a small variation of the DBP concentration are only distinguishable with
mixture III. The presence of the discontinuity is also explainable by this regression. When
the propellant deterred part is burnt, the combustion of the center part begins. As the
difference between the combustion rates is high for mixture III, this is expressed by a
discontinuity in the derivative curve.
diameter
0.7mm,
igniter III
3
2.5
diameter
1.0mm,
igniter III
2
1.5
diameter
0.7mm,
igniter VI
1
0.5
0
0
10
%DBP (HPLC)
20
diameter
1.0mm,
igniter VI
3.5
Combustion rate @30 MPa
(cm/s)
Combustion rate @30 MPa

(cm/s)
3.5
3
2.5
diameter
1.0mm,
igniter III
2
1.5
diameter
0.7mm,
igniter VI
1
0.5
0
0
(1)
Fig 9.
diameter
0.7mm,
igniter III
10
%DBP @0m
20
diameter
1.0mm,
igniter VI
(2)
Variation of the combustion rate at 30MPa as a function of the percentage of (1)

average DBP concentration and (2) DBP in surface for the different propellants
(from B to J) and for igniting mixture III and VI. The loading density is
0.20 g/cm3.
245
3.3
Correlation between ballistic firing, closed vessel tests and infrared

microscopy
First, it must be mentioned that only propellant C is manufactured for the used weapon
system. For the others propellants, the combustion is incomplete, residues have been
observed during the firing.
3.3.1 Propellants with an average diameter of 0.7 mm
Figure 10 shows the pressure records in the cartridge for the different propellants with an
average diameter of 0.7 mm. This has to be correlated with figures 3 and 4 showing their
respective deterrent concentration profiles and their corresponding combustion rates. The
peak pressure of propellant C shows an asymmetric shape. At the beginning of the
combustion, the deterred part of the propellant burnt, the burning rate is low but the
propellant surface is high, when the deterred part of the propellant is burnt the combustion
rate is higher but the propellant surface is smaller and the gas production remains more or
less constant. This explains the particular shape of the peak pressure. This peak corresponds
to a sequence of two phenomena, the burning of the deterred part of the propellant and the
combustion of the inner core of the propellant.
This asymmetric shape of the peak is no more present for the other propellants, due to
incomplete combustion. It is probable that only the deterred part of the propellant is burnt.
The slopes of the pressure buildup correspond to the deterrent concentration profiles of
the different propellants and their respective combustion rates. When the deterrent
concentration in surface is higher, the pressure increase is slower. By consequence, the
maximum pressure is lower. During the ballistic firing, a difference is observed between
propellant E and F. This difference corresponds to the beginning of the combustion (burning
of the propellant outer layer), as the combustion is incomplete. This discrepancy is not
noticeable in the closed vessel tests if igniter VI is used, in this case a difference is only
present at the end of the combustion (combustion rate at the high pressure values).
350
Pressure (MPa)
300
(C)
(D)
250
200
(E)
150
100
(F)
50
0
1000
2000
3000
4000
5000
6000
7000
time (s)
Fig 10. Pressure time curve of Propellant C, D, E and F measured in the cartridge
during the ballistic firing (pretrigger from muzzle flash at 10 000s).
The labels in brackets refers to the type of propellant.
246
3.3.2 Propellants with an average diameter of 1.0 mm

Figure 11 shows the pressure time curves measured in the cartridge during ballistic
firing for the propellant with an average diameter of 1.0 mm.
Pressure (MPa)
300
(I)
250
(H)
200
(G)
150
100
(J)
50
0
3000
4000
5000
time (s)
6000
7000
Fig 11. Pressure time curve of Propellant G, H, I and J measured in the cartridge during
the ballistic firing (the curves are not correlated in time). The labels in brackets
refer to the type of propellant.
A difference is observable between all the propellants. This difference is not observable
in the closed vessel tests when mixture VI is used. The curve of propellant G shows a
discontinuity in the pressure decrease. This is due to the presence of the two chemical
compositions. For the other propellants this discontinuity is no more present, due to
incomplete combustion and corresponding to their greater deterrent penetration depth. In
fact, as the combustion is incomplete, only the outer layer of the deterrent burnt during the
ballistic firing.
4.
CONCLUSIONS
Different gaseous ignition mixtures have been tested on a deterred propellant. It has
been shown that the selected ignition mixtures permit to observe a difference between the
two chemical compositions during the closed vessel tests, namely mixtures II and III.
Ignition mixtures II and III have been used on different deterred propellants with
mixture VI as benchmark. It has been observed that a greater difference is observed between
the deterred propellants if a gaseous ignition system is used. A discontinuity is present in the
pressure derivative curve when an appropriate gaseous ignition system is used. From this
discontinuity a deterrent penetration depth has been calculated. These calculated values
correspond to the propellant thickness with has a high deterrent concentration. The
combustion rate corresponding to the combustion of the deterred part of the propellant is
linear as a function of the percentage of DBP in surface. The slope of this linear variation is
greater for a gaseous ignition than if mixture VI is used as ignition system.
247
To explain the presence of a discontinuity in the closed vessel tests when a gaseous
ignition system is used, two hypotheses have been proposed:
-
The gaseous ignition is more homogeneous than the ignition by black powder.
The variation of the combustion rate with the deterrent concentration is higher when
mixture II or III is used and this permits to differentiate the two chemical
compositions of the propellant.
These two hypotheses are not incompatible, if the ignition is not homogeneous with black
powder, the combustion rate will be an average of the combustion rate corresponding to the
different deterrent concentrations. This will decrease artificially the calculated combustion
rate. To differentiate these two hypotheses, propellants with a constant DBP concentration
will be manufactured. With these propellants ignition will have no influence on the
combustion rate and the effect of the ignition method will be investigated.
Ballistic firing of different deterred propellants have been realized. It has been observed
difference between the pressure records of different propellants which exhibit none
difference in the closed vessel tests when mixture VI is used.
This work permit to observe that there is a good correlation between infrared
microscopy, ballistic firing and closed vessel tests if an appropriate ignition system for the
closed vessel tests is used. It shown that closed vessel tests can be used as a rapid method of
characterization of deterred propellants. In fact, infrared microscopy is highly time
consuming and for ballistic firing appropriate laboratory conditions are needed.
REFERENCES
[1]
[2]
[3]
STANAG 4115 (1997): Definition and determination of ballistic properties of gun propellants,
North Atlantic Council.
L. JEUNIEAU, M.H. LEFEBVRE, A. PAPY, M.C. PIRLOT, P. GUILLAUME, C. REYNAUD: Closed
vessel test: influence of the ignition method on the combustion rate, 33rd International Annual
Conference of ICT, June 25, 2002, Karlsruhe, Germany.
L. JEUNIEAU, M.H. LEFEBVRE, A. PAPY, M.C. PIRLOT, P. GUILLAUME, C. REYNAUD:
Spherical deterred propellant: influence of the concentration gradient on the burning rate
calculation, 34th International Annual Conference of ICT, June 24-27, 2003, Karlsruhe,
Germany.
248
A MOLECULAR DYNAMICS SIMULATION STUDY

OF ELASTIC PROPERTIES OF HMX-BASED AND TATB-BASED PBXS
Xiao Jijun, Ma Xiufang, Zhu Wei, Huang Yucheng, and Xiao Heming
Chemical Engineering Institute, Nanjing University of Science and Technology,
Nanjing, 210094, P. R. China
Abstract:
Three different models, i.e., inserting, covering and cutting models, for PBXs were
proposed for different researching aspects. Used for choosing polymeric binders, the
inserting and covering models are mainly applied to find the relations of temperatures
and concentrations respectively with elastic properties of the PBXs. The cutting model is
especially used to describe the highly anisotropic behavior of TATB. These models were
simulated with molecular dynamics using COMPASS force field. It is found that the
mechanical properties of -HMX and TATB can be effectively improved by blending
fluorine containing polymers in small amount. The moduli of the PBXs decrease with
increasing temperature and concentration of binders. Different crystal surfaces interacting
with the same polymer binder have different modulus-decreasing effects due to the highly
anisotropic behavior of TATB. The modulus-decreasing effect ranking order is (010)
(100) > (001).
Keywords:
1.
molecular dynamics, simulation, HMX, TATB, PBXs, elastic properties
INTRODUCTION
Plastic-bonded explosives (PBXs) are highly filled (ca. 90% w/w filler) composite materials
comprising grains of a high-explosive (HE) material held together by a polymeric binder and
widely used in many defense and economy scopes because PBXs have notable merits such as
good safety, high strength, ease of process and so on [1-3]. In general, a solid energetic compound
is brittle, so the key problem in PBX formulation is in choosing a polymeric binder which can
reduce brittleness of the compound. It is a costly work to determine a PBX formulation, the
process of which lacks safety [4,5]. Therefore, it is needed to better understand the physical,
chemical, and mechanical behaviors of the constituents of PBX formulations, as well as the
interactions between them, from fundamental theoretical principles. There are many reports on
the behaviors mentioned above of explosives or polymers by quantum chemistry method and
molecular dynamics (MD) methods [6-10], but few reports [11] have been found on simulation
computation for PBXs, especially for its most important properties mechanical properties by
MD method. In this work, not only the elastic properties of explosive compounds but also those
of PBXs were simulated by MD method using COMPASS [12] force field.
Author to whom correspondence should be addressed. E-mail: xiao@mail.njust.edu.cn

This research was supported by the National Natural Science Foundation of China (Grant No.10176012 and
20173028).
249
The solid energetic compound octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX) is

widely used with best comprehensive properties. The HMX crystal structure of -phase has
been determined by neutron diffraction [13]. -HMX is the thermodynamically stable form under
ambient conditions and is the energetic constituent in several high-performance PBX
formulations [2,14]. Thus -HMX-based PBXs were taken as one of researching objects in this
paper.
The solid energetic compound 1,3,5-triamino-2,4,6-trinitrobenzene (TATB) has generated
considerable interest since its first synthesis [15] because of its extraordinary stability under
thermal, impact, or shock-initiation conditions [16]. TATB is a triclinic crystal with two
molecules in the unit cell [17]. The crystal structure is graphitelike, containing hydrogen-bonded
sheets of molecules along the plane of the a and b crystallographic axes. TATB crystal has the
highly anisotropic properties. Therefore TATB-based PBXs were chosen as the other
researching object.
As to polymeric binders, 4 types of fluorine containing polymers were selected. They were
Poly(vinylidene difluoride) (PVDF), Polychlorotrifluoroethylene (PCTFE), F2311 and F2314.
F2311 and F2314 are copolymers polymerized from vinylidene difluoride and
chlorotrifluoroethylene with the ratios of 1:1, 1:4, respectively.
2.
MODELLING AND SIMULATION
The initial HMX and TATB structures in Choi et al.[13] and Cady et al.[17] were used in the
condensed phase simulations in this paper. The periodic MD simulation cells containing 48 and
72 molecules were used for -HMX and TATB, corresponding to 24 (423) and 36 (334)
unit cells, respectively. The MD simulation data were generated under constraints of constant
particle number, pressure, and temperature (NPT) dynamics for -HMX. TATB system was run
respectively in NVT and NPT statistical ensembles. After an equilibrium run, configurations of
the production of run trajectory were analyzed by static force method to obtain the elastic
coefficient, and on the base of static force analysis, using a least-squares fit, isotropic elastic
properties (tensile, bulk, and shear moduli, as well as Poissons ratios) of the system were
computed. All computations were carried out using commercial software programs from
Accelrys Inc. [18].
As to the polymer chains, the end groups were saturated by H or F according to their types.
Each chain was run under NVT (constant volume, temperature, and particle number) condition
using COMPASS force field for 2.5 ns to get an equilibrium configuration. During the
following PBXs model construction, all polymer chains were in equilibrium states.
A.
INSERTING MODEL
A PBX composite model was constructed by inserting a equilibrium polymer chain in a

periodic MD simulation cell and then deleting all HMX or TATB molecules which contain any
atom with distance to any atom in the polymer chain less than or equal to 1. Then the PBX
model was allowed to evolve dynamically in NPT ensemble. As an example, the PBX model
HMX/F2311 is shown in Fig.1. The advantage of this model is making it convenient to study
the temperature effect on elastic properties of PBXs.
250
Fig 1.
B.
HMX/F2311 PBX periodic MD simulation cell.
COVERING MODEL
According to the initial TATB structure [17], the super cell (444 unit cells) was made;
keeping a ball consisted of 71 molecules in the middle of the super cell, deleting the rest of
molecules, then making polymer chains close to the ball surface along x, y, z axis directions,
after the super cell was compressed to the calculation density, a periodic MD simulation cell
was obtained, i.e. a PBX model covering model was finished. Then NVT MD simulations of
the model were performed. When using this model in MD simulation, different polymer
concentrations in PBXs can be easily obtained by adjusting the number of polymer chains,
and/or the length of polymer chains. The following Fig. 3 and Fig. 4 show the covering models
of TATB/ PCTFE before and after MD, respectively.
Fig 2.
Covering model before being compressed. FIG. 3 Covering model after MD.
251
C.
CUTTING MODEL
Because of the complex nature of the extremely anisotropic behavior of TATB, the cutting
model was considered. Also according to the initial TATB structure [17], (001), (010) and (100)
crystal surfaces were taken in three different periodic MD simulation cells with each containing
90 molecules, as shown in Fig.4. Putting polymer chains on the surfaces, compressing these
cells to the calculation densities, the periodic cells for MD simulation were finished, as shown in
Fig. 5. Then these cells were run in NVT statistical ensemble.
(001)
Fig 3.
(100)
Different crystal surface of TATB.
(001)
Fig 4.
(010)
(010)
(100)
Different MD simulation cells for different TATB crystal surface.
252
3.

3.1
Elastic properties of crystal HMX and TATB and their PBXs

calculated by the inserting model
This model makes it easy to calculate the elastic properties with the temperature changing.
The elastic coefficients, (tensile, bulk, shear) moduli, and Poissons ratios of crystal HMX and
HMX-based PBXs starting at a temperature of 245K in steps of 50K to a maximum temperature
of 345K were obtained by analyzing the configurations of the production of run trajectory in
NPT ensemble. These data are tabulated in Tables 1 and 2. The elastic properties of crystal
TATB and TATB-based PBXs at the same starting temperature and in the same steps mentioned
above to a maximum temperature 545K were obtained by the same way, and the data are given
in Table 3. For the purpose of comparison, the NVT MD simulation of crystal TATB was also
run at 295K temperature and the atomistic simulation was used to calculate the elastic properties
for TATB.
Table 1. Elastic coefficients and isotropic moduli, Poissons ratios for HMX
and HMX/F2311 PBX at different temperatures.
245K
295K
345K
C11
13.2 (8.9)
12.8 (7.2)
10.8 (7.1)
C22
12.9 (7.1)
10.9 (5.8)
9.7 (5.9)
C33
12.9 (8.3)
11.4 (7.4)
10.4 (6.8)
C44
6.4 (3.3)
5.9 (2.9)
5.5 (2.8)
C55
4.9 (2.7)
4.6 (2.5)
4.2 (2.2)
C66
5.4 (2.9)
4.8 (2.9))
4.3 (2.5)
C12
3.9 (1.4)
3.4 (1.9))
3.0 (1.3)
C13
5.5 (2.8)
4.9 (2.7)
3.8 (2.1)
C23
6.8 (3.1)
5.2 (2.8)
4.1 (2.1)
C15
-0.8 (0.4)
-0.4 (0.0)
-0.1 (0.3)
C25
-3.0 (-0.8)
-2.7 (-0.9))
-2.3 (-0.7)
C35
-0.7 (0.6)
-0.4 (0.2))
-0.4 (0.6)
C46
-2.0 (-0.3)
-1.8 (-0.5)
-1.5 (-0.3)
Tensile modulus
9.9 (7.0)
9.2 (5.5)
8.4 (5.8)
Poissons ratio
0.3 (0.2)
0.3 (0.3)
0.3 (0.2)
Bulk modulus
7.9 (4.3)
6.9 (3.9)
5.9 (3.4)
Shear modulus
3.8 (2.9)
3.6 (2.2)
3.3 (2.4)
Lam coefficient
5.4 (2.4)
4.7 (2.3)
3.7 (1.8)
Lam coefficient
3.8 (2.9)
3.6( 2.4)
3.3 (2.4)
Data in parentheses are corresponding results for HMX/F2311. Elastic coefficient,

modulus and Lam coefficient units: GPa.
253
Table 2. Elastic coefficients and isotropic moduli, Poissons ratios for HMX
and various PBXs at 295K
-HMX
-HMX/PVDF
-HMX/ PCTFE
-HMX/F2311
-HMX/F2314
C11
12.8
7.6
6.5
7.2
6.9
C22
10.9
5.3
5.4
5.8
6.2
C33
11.4
5.1
6.9
7.4
4.7
C44
5.9
3.0
2.1
2.9
3.2
C55
4.6
2.0
2.2
2.5
1.7
C66
4.8
1.9
2.6
2.9
2.1
C12
3.4
1.0
1.5
1.9
1.0
C13
4.9
1.8
2.4
2.7
1.3
C23
5.2
1.8
2.7
2.8
1.8
C15
-0.4
0.0
0.1
0.0
0.4
C25
-2.7
-0.4
-0.8
-0.9
-0.5
C35
-0.4
0.4
0.1
0.2
0.6
C46
-1.8
-0.3
-0.4
-0.5
-0.3
Tensile modulus
9.2
5.4
5.2
5.5
5.4
Poissons ratio
0.3
0.2
0.3
0.3
0.2
Bulk modulus
6.9
3.0
3.5
3.9
2.9
Shear modulus
3.6
2.2
2.1
2.2
2.3
Elastic coefficient, modulus and Lam coefficients units: GPa.
From the data in Tables 1 and 2, we can find that the moduli of HMX and HMX-based
PBXs decreased with the temperatures increasing, and the moduli of HMX also decreased when
it was blended with polymer binders. The same trends were seen in Table 3 for TATB and
TATB/F2311.
254
Table 3. Table3 Elastic coefficients and isotropic moduli, Poissons ratios for TATB
and TATB/F2311 PBX at different temperatures
245K
C11
C22
C33
C44
295K
345K
395K
445K
495K
545k
38.2 (29.6) 35.0 [43.7] (26.8) 31.6 (25.3) 28.3 (18.2) 26.4 (17.1) 20.5 (14.2) 15.0 (14.1)
34.8 (26.1) 33.2 [38.2] (24.2) 31.1 (22.6) 26.1 (15.3) 24.0 (15.3) 20.0 (14.5) 16.5 (13.5)
12.3
11.9 [16.5] (7.6) 10.7 (6.4) 10.1 (5.7) 9.6 (4.9) 6.5 (3.8) 4.5 (2.3)
(8.910
13.1 (10.3) 12.6 [14.9] (9.1) 11.8 (9.0) 10.4 (7.2) 9.9 (6.4) 7.6 (6.1) 6.4 (5.8)
C55
1.4 (1.0)
1.4 [0.4] (0.9)
1.2 (0.7)
1.3 (0.8)
1.3 (0.7)
0.5 (0.5)
0.8 (0.5)
C66
1.3 (1.0)
1.2 [1.2] (0.9)
1.0 (1.0)
1.2 (0.5)
1.2 (0.8)
0.6 (0.6)
0.6 (0.4)
C12
7.6 (5.1)
7.2 [10.0] (4.6)
6.7 (4.9)
5.4 (3.4)
5.0 (3.5)
4.0 (3.6)
3.6 (1.9)
C13
1.9 (2.3)
2.0 [2.7] (1.2)
1.8 (1.2)
1.8 (3.0)
1.8 (2.2)
4.3 (2.1)
3.5 (0.9)
C23
1.9 (2.1)
2.2 [4.6] (1.5)
1.6 (1.6)
1.9 (2.4)
1.8 (2.0)
4.4 (2.0)
2.9 (0.5)
C15
0.6 (0.4)
0.6 [0.3] (0.3)
0.5 (0.4)
0.8 (-0.2)
0.6 (0.1)
0.2 (0.1) -0.1 (0.5)
C25
0.7 (0.5)
0.6 [0.7] (0.4)
0.6 (0.6)
0.6 (-0.1)
0.5 (0.0)
0.3 (0.0)
C35
0.2 (0.1)
0.1 [-0.5] (0.0)
0.1 (-0.2)
0.0 (0.0)
0.0 (0.2) -0.1 (0.2) 0.0 (0.3)
C46
0.0 (0.0)
-0.0 [-0.5] (0.0)
-0.1 (0.0) 0.0 (-0.2)
0.0 (0.0) 0.1 (-0.3) 0.2 (0.4)
0.3 (0.7)
Tensile modulus 27.5 (20.6) 25.7 [31.1] (19.0) 23.6 (17.4) 20.8 (12.0) 19.3 (11.8) 13.9 (10.1) 10.6 (9.7)
Poissons ratio
0.1 (0.1)
0.1 (0.1)
0.1 (0.2)
0.1 (0.2)
0.2 (0.2)
0.2 (0.1)
Bulk modulus
12.0 (9.5) 11.4 [14.8] (8.2) 10.4 (7.7)
9.2 (6.3)
8.6 (5.9)
8.0 (5.0)
6.2 (4.1)
Shear modulus
12.3 (9.0) 11.4 [13.5] (8.6) 10.5 (7.8)
9.2 (5.1)
8.6 (5.1)
5.7 (4.3)
4.3 (4.4)
3.4 (2.6)
3.0 (2.9)
2.9 (2.5)
4.2 (2.2)
3.3 (1.1)
Lam coefficient 12.3 (9.0) 11.4 [13.5] (8.6) 10.5 (7.8)
9.2 (5.1)
8.6 (5.1)
5.7 (4.3)
4.3 (4.4)
Lam coefficient 3.8 (3.4)
0.1 [0.2] (0.1)
3.8 [5.8] (2.4)
Elastic coefficient, modulus and Lam coefficients units: GPa. Data in parentheses for
TATB/F2311.
Data in brackets for TATB in NVT ensemble.
This is because when the temperature becomes more elevated, the internal free volume is
expanded more and the molecules get more movement energies. All of these changes make the
materials deform more easily as they are subjected to external force. This means that the moduli
decrease. The polymer chain of binders is flexible due to the rotation of bonds, and this also
makes deformation easy.
The data for elastic properties of HMX obtained in this work is smaller than the literature
value9. But this does not affect the obtained results in this paper on the trend of elastic properties
with temperature changing.
255
3.2
Elastic properties of crystal TATB and TATB-based PBXs calculated

by the covering model
The polymer binder concentrations of PBXs can be easily modified by the covering model.
The data in Table 4 show that the more the concentration of PCTFE increases, the more the
moduli of the PBX decrease. This character may be attributed to the cooperation of polymer
binder chains.
Table 4. Elastic coefficients and isotropic moduli, Poissons ratios for TATB/PCTFE
PBXs with different PCTFE concentrations at 298K
6.2% PCTFE
16.5% PCTFE
19.0% PCTFE
28.3% PCTFE
C11
13.4
10.0
9.5
9.1
C22
12.7
12.1
10.0
8.9
C33
12.0
9.3
8.6
7.4
C44
3.6
3.1
3.0
2.3
C55
2.1
1.9
1.4
1.6
C66
2.5
1.9
1.6
1.6
C12
5.3
5.2
4.4
4.2
C13
5.0
5.5
4.1
4.5
C23
6.7
5.1
4.8
4.8
C15
-0.9
-0.5
-0.2
-0.3
C25
0.2
0.2
-0.4
-0.5
C35
-0.4
0.1
-0.2
-0.5
C46
-0.4
-0.4
0.1
-0.2
Tensile modulus
8.9
7.0
6.5
5.4
Poissons ratio
0.3
0.3
0.3
0.3
Bulk modulus
8.2
7.0
6.1
5.8
Shear modulus
3.4
2.6
2.5
2.0
Elastic coefficients, moduli units: GPa. Concentration: PCTFE (W)[TATB+PCTFE (W)].
This model can also be used to investigate the temperature impact on the moduli of PBXs.
As an example of TATB-based PBXs, the model of 16.5% PCTFE was allowed to dynamically
evolve at the temperature of 298K, 348K, and 398K, respectively. From the data in Table 5, the
moduli changing trends are the same as the mentioned in Section A.
256
Table 5. Elastic coefficients, and isotropic moduli, Poissons ratios for the TATB-based
PBX with 16.5% PCTFE at different temperatures
C11
C22
C33
C44
C55
C66
C12
C13
C23
C15
C25
C35
C46
Tensile modulus
Poissons ratio
Bulk modulus
Shear modulus
3.3
298K
348K
398K
10.0
12.1
9.3
3.1
1.9
1.9
5.2
5.5
5.1
-0.5
0.2
0.1
-0.4
7.0
0.3
7.0
2.6
10.3
11.1
8.1
2.2
1.8
1.7
4.2
5.4
5.0
-0.4
0.1
-0.1
-0.3
6.6
0.3
6.5
2.5
9.0
10.3
8.4
1.5
1.5
2.0
4.9
5.2
4.8
-0.3
0.3
0.4
-0.5
5.8
0.4
6.4
2.3
Elastic coefficients, moduli units: GPa.
Elastic properties of crystal TATB and TATB-based PBXs calculated

by the cutting model
Because of the highly anisotropic behavior of TATB, the modulus-decreasing effect can be
investigated by the cutting model. The data in Table 6 show that not only different polymer
binders have different modulus-decreasing impact but also for the same polymer binder,
different crystal surfaces have the different impacts. The interaction of different polymer
molecules with TATB is different and the polymer bond rotation barriers are not the same for
different molecular structures. This causes the change of the moduli-decreasing effect. TATB
crystal structure is graphitelike, containing hydrogen-bonded sheets of molecules along the
(001) surface, thus the interaction among TATB molecules parallel to (001) surface is the
strongest and polymer binders play an less important role on the surface. This leads to the
modulus-decreasing effect ranking (010) (100) > (001).
257
Table 6.
Elastic coefficients, and isotropic moduli, Poissons ratios for TATB

and TATB-based PBXs with various polymer binder on the different
crystalline surface
Tensile module
Poissons Ratio
Bulk module
Shear module
31.5
0.15
15.2
13.7
TATB
PBX 010
PVDF
F2311
F2314
PCTFE
9.5
9.7
9.4
9.2
0.27
0.25
0.26
0.22
7.0
6.5
6.6
5.5
3.7
3.9
3.7
3.8
PBX 100
PVDF
F2311
F2314
PCTFE
12.8
10.9
10.0
10.0
0.19
0.17
0.22
0.21
7.0
5.6
6.0
5.8
5.4
4.6
4.1
4.1
PBX 001
PVDF
F2311
F2314
PCTFE
18.8
17.7
14.1
14.4
0.12
0.13
0.17
0.16
8.2
8.1
7.0
7.1
8.4
7.8
6.0
6.2
4.
Elastic coefficients, moduli units: GPa.
CONCLUSIONS
Through the atomic MD simulation using the three models discussed above, the following
conclusions were obtained:
1. Blending with a polymer binder can reduce the elastic moduli of HMX and TATB
crystal.
2. With the temperature elevating and the polymer binder concentration increasing, the
moduli of HMX, TATB, HMX-based PBXs, and TATB-based PBXs decrease.
3. The modulus-decreasing effect rank ordering of TATB crystal is: (010) (100)
>(001).
258
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
T. R. GIBBS, A. POPOLATO, EDS.: LASL Explosive Property Data, University of California Press,
Berkeley, 1980.
H. S. DONG, F. F. ZHOU: High Energy Explosives and Correlative Physical Properties, Science
Press, Beijing, 1984.
G. X. SUN: Polymer Blended Explosives, Defense Industry Press, Beijing, 1984.
D. M. HOFFMAN, L. E. CALY: Polym. Engin. Sci. 26, 1489 (1986).
Y. B. SUN, J. M. HUI, X. M. CAO: Military Use Blended Explosives, Weapon Industry Press,
Beijing, 1995.
J. S. LI, H. M. XIAO, H. S. DONG: Exploding and Shock, 20(3), 221 (2000).
H. M. XIAO, J. S. LI, H. S. DONG: J. Phys. Org. Chem. 14, 644 (2001).
J. P. LEWIS, T. D. SEWELL, R. B. EVANS, G. A. VOTH: J Phys. Chem., B104, 1009 (2000).
T. D. SEWELL, R. MENIKOFF, D. BEDROV, G. D. SMITH: J Chem. Phys., 119(14) 7417 (2003).
X. Z. YANG: Molecular Simulation and Polymer materials, Science Press, Beijing, 2002.
J. J. XIAO, G. Y. FAN, G. F. JI, H. M. XIAO: Chinese Science in Bulletin,50(1),21 (2005).
H. SUN: J. Chem. Phys., B102, 7338 (1998).
C. S. CHOI: H. P. Boutin, Acta Crystallogr. B26, 1235 (1970).
T. R. GIBBS, A. POPOLATO, EDS.: LASL Explosive Property Data, University of California Press,
Berkeley, 1980.
C. L. JACKSON, J. F. WING: Am. Chem. J. 10, 283 (1888).
B. M. DOBRATZ: Report No. LA-13014-H, 1995.
H. H. CADY, A. C. LARSON: Acta Crystallogr. 18, 485 (1965).
ACCELRYS INC., San Diego, CA.
259
THE APPLICATION OF A SURFACE ACTIVE THEORY

TO ENERGETIC MATERIALS RESEARCH ON EXPANSION AMMONIUM NITRATE EXPLOSIVES
Lu Chunxu
Nanjing University of Science and Technology,Nanjing210094,China
Abstract:
The expansion of Ammonium Nitrate(AN)is a novel technology. The guideline of this
innovation is the self-sensitization of AN. It is a breakthrough to classic methods. The
approach to self-sensitization is the expansion of AN. Its essence is a surfactant
technology applied to powdery industrial explosive, and it is a physical and chemical
process under coercive crystallization. In this paper, the mechanism and technical
characteristics of expansion of AN are discussed and its unique advantage is shown. The
expansion technology of AN is mainly applied to rock expanded AN explosive and its
comparison with other industry explosives is also given. The expansion technology is also
used in manufacturing of permitted explosive and other industrial explosives.
Keywords: expanded ammonium nitrate; rock explosive; micro structure; mechanism
1.
DESIGN AND TECHNICAL PROCESS OF EXPANSION

OF AMMONIUM NITRATE
1.1
Background
Although emulsion explosive along with ANFO and heavy ANFO has made some
remarkable progress and possesses increasing market share, the powdery AN explosives
containing TNT or NG sensitizer still occupy a remarkable proportion. But these explosives
have some shortcomings as following: (1)Big occupancy of TNT leads to serious toxicity,
which would successively do bad to the bodies of workers and pollute environment; (2)The
products have high cost; (3)The production safety can not be ensured; (4)The products has bad
physical performance because of serious hygroscopicity.
With the elapse of more than 100 years, many researchers have being engaged in this task
and made some progress, e.g.., the addition of additive and surfactants for coating and
hydrophobes; the modification of charging and packing; the substitution of TNT by RDX,
urea nitrate and metal powder. These methods are do have some efficiency in a way, but they
are unilateral and uneconomical. So the essential problem is still present. In a word, these
problems are imperative to be resolved but have not been resolved.
260
1.2
Basic point of design
After technically analyzing of powdery industrial explosive and trying many methods to
resolve these problems, we propose that only the self-sensitization of AN, whose theoretic
foundation is that the detonation of AN explosive is a kind of non-ideal one, could innovate its
performance and get completely rid of TNT from powdery industrial AN explosive. Under the
action of outside energy, some pointed parts of explosive would form hot point with quite a
little energy, which would make explosive decompose and emit energy rapidly, which would
successively cause the explosive around it to react till all the system to detonate. The so-called
hot point could be formed through adiabatic compression of micro bubble or the acutely
attrition and collision between pointed parts. So the theory mentioned above inspire us that
introducing micro bubble into AN particles and making particles malformation and
coarseness may realize self-sensitization [1-4].
1.3
Creation of self-sensitization theory
The detonation mechanism via the formation of hot point is a conformable view and is
considered to be essentially important in the detonation theory of explosive. It can not only
satisfactorily explain the reason why it can explode under little exoteric energy, but also guide
people to adjust the product process to a safe way and design the sensitivity of liquid explosive
and emulsion explosive.
We applied special technique to imbed micro bubble into AN particles. When the
detonating energy is put into action, micro bubble would be compressed adiabaticly to form
hot point with extremely high temperature and pressure. At the same time, distorted
particles would attrite and collide violently together. All above would lead to the formation of
hot point. It is the action of interior self-sensitization of crystal that completely changes the
physical performance and detonation sensitivity of AN explosive.
1.4
Expansion is the core of self-sensitization of AN
In order to make micro bubble and malformation in AN particles, there is only one
way to compel AN to rapidly crystallize from excessive boiling state. On the basis of deep
research on surfactant technology, we synthesized a new composite surfactant, which can
decrease greatly the surface tension of AN solution, after a great many of experiments. Under
the condition of reduced pressure and high temperature AN crystal would form a incredible
number of pores, holes and interspaces. Contemporarily, there would be many active
surface to form outside. Therefore this kind of AN is called expanded AN. The results show
that the design of self-sensitization is proper and efficient.
The key technology of expanded AN free of TNT is the expansion of AN, and the core of
expansion is surfactant technology, which is a physical and chemical process of compulsive
froth and crystallization under the action of surfactant.
The modification of AN has being done through traditional ways, e.g., addition of other
materials, emulsion after dispersing, etc. But the author breakout it and find a new way and
finally make it.
261
1.5
Design of self-sensitization theory of expanded ammonium nitrate
The application of micro-pore, acting as hot spot, can efficiently promote the detonation
performance and shock sensitivity of emulsion explosive, which contributes to find new ways
to sensitize powder explosive.
It is extremely important to introduce micro-pore in powder explosive. But obviously,
introducing micro-pore in sawdust and oil is very difficult and whose amount can be
neglected. Then the only practical way is to introduce it in the crystal of ammonium nitrate.
This is the basis to design self-sensitization of expanded ammonium nitrate.
The detonation initiated by exterior energy of AN-based explosive is a kind of non-ideal
detonation. Under the action of exterior energy, some rough parts in explosive would form hot
spot with a great deal of energy, which can lead to rapid decomposition of explosive in these
hot spot and then proceed to initiate reaction of other explosive till the whole system to
explode. Hot spot can be formed by adiabatic compression on micro-pore or violent friction or
collision between edges of explosive particles. What showed above inspire us that it is
reasonable to introduce micro-pore in AN and roughen its crystal. This kind of design makes it
possible to form hot spot with high temperature and pressure to initiate explosive under the
strong action of exterior factors. This is so-called self-sensitization design.
Through analyzing production technologies and methods of powder industrial explosive,
a great number of results can be obtained, which indicate that self-sensitization of AN is one of
ways to realize producing powder explosive free from TNT.
This kind of interior self-sensitization of AN can take place sensitizer TNT that sensitize
explosive, which give rise to tremendous change of detonation performance and shock
sensitivity of expanded AN and explosive derived from it.
1.6
Distribution and regularity of micro-pore of expanded AN
Expanding process makes a great of pore, crack, burrs and irregular edges and corners on
crystal surface. And these factors would lead to the formation of hot spot under the action of
detonation shock. Besides concerning surface characteristics of micro-pore, its distribution
and regularity also have been studies [5-9].
We have used OMNISORP (TM) specific surface area and total volume apparatus made
by American Coulter Corp. to study distribution and regularity of micro-pore of expanded AN
under different adsorption pressure and obtained its Kelvin radius, thickness of adsorption
layer, pore radius, volume increment of desorption, surface area of pore and percentage of
total surface area. The results listed listed in Table 1.
262
Table 1. Distribution and regularity of micro-pore in expanded ammonium nitrate

Pressure
Volume
Kelvin Thickness
Pore
Volume
Surface Percentage
ratio of
of
of
radius
radius
increment area of
of total
adsorption Rk/10-10m adsorption Rp/10-10m desorption
pore
of
surface
p/p0
Vd/mlg-1 desorption Sdp/mg-1
layer
area
-10
-1
d/10 m
Vad/mlg
Szdp%
0.9376
147.97
15.024
162.99
0.2402
0.00045
0.05436
2.1171
0.91913
105.89
13.834
119.72
0.2033
0.00083
0.1482
7.8888
0.8899
81.684
12.853
94.537
0.2119
0.00125
0.2545
17.798
0.8664
66.473
12.054
78.527
0.1941
0.00165
0.391
33.023
0.8418
55.354
11.34
66.693
0.1665
0.00199
0.584
55.765
0.8184
47.581
10.752
58.333
0.152
0.00230
0.8819
90.111
0.794
41.325
10.210
51.536
0.1484
0.00262
1.3421
100.0
0.7697
36.427
9.734
41.161
0.148
0.00294
2.0238
100.0
0.7454
32.449
9.3049
41.754
0.1338
0.00323
2.9808
100.0
0.7214
29.185
8.9194
38.105
0.1255
0.00350
4.3421
100.0
0.6975
26.465
8.5712
35.036
0.1093
0.00374
6.1974
100.0
0.6729
24.064
8.2401
32.304
0.1053
0.00396
8.693
100.0
0.6493
22.071
7.9461
30.017
0.0961
0.00417
11.900
100.0
0.6246
20.257
7.6609
27.918
0.0982
0.00438
15.837
100.0
0.6011
18.73
7.4046
26.136
0.1022
0.00461
20.446
100.0
0.5768
17.322
7.1586
24.430
0.0940
0.00482
25.278
100.0
0.5518
16.034
6.9195
22.954
0.0868
0.00500
29.711
100.0
0.05277
14.914
6.7006
21.614
0.0795
0.00517
33.057
100.0
0.5032
13.880
6.489
20.369
0.0590
0.00527
34.006
100.0
0.4794
12.966
6.2928
19.259
0.0551
0.00536
31.897
100.0
0.4546
12.093
6.097
18.19
0.0610
0.00547
25.278
100.0
0.4311
11.328
5.9179
17.246
0.0386
0.00550
14.743
100.0
0.4065
10.59
5.7378
16.327
0.0415
0.00550
0.2149
100.0
0.3811
9.8821
5.7378
16.327
0.0415
0.00560
0.2149
100.0
0.357
9.2548
5.7378
16.327
0.0415
0.00564
0.2149
100.0
The relation between pore radius and volume of expanded AN is also determined and the
results are listed in Table 2.
263
Table 2. The relation between pore radius and pore volume

of expanded ammonium nitrate
pore radius/10-10m
Pore volume /mlg-1
Percentage over total

volume/%
200
0.00000
0.00
200~100
0.00087
15.41
100~50
0.00178
31.52
50~40
0.00061
10.80
40~30
0.00094
16.65
30~20
0.00109
19.30
20~10
0.00037
6.55
The total volume of pore with a radius R bigger than 1010-10m is 0.00568mlg-1.
According to the standard for pore radius distribution from IUPAC, the pores are
classified respectively as following:
R50010-10m: big pore;
R=2010-10m~50010-10m:mesopore;
R=710-10m~2010-10m: supermicro-pore;
R710-10m: imperceptible-pore
From this standard, 95% of pore in expanded AN belongs to mesopores. So it is defined as
mesopore exactly, whereas we call it micro-pore customarily.
According to self-sensitization theory of expanded AN, fine pore (supremicro-pore and
imperceptible micro-pore) is invalid. Because it is inevitable to mix expanded AN, sawdust
and oil together when manufacturing expanded ammonium nitrate explosive so that any pore
with a radius smaller than 2010-10m is sure to be filled by sawdust and oil. On the other hand,
big pore should also be avoided because nimiety big pores would decrease pore ratio so as to
influence the amount of hot spot, which is disadvantage to self-sensitization. In a word,
mesopore is the best satisfactory for self-sensitization. Because it not only can not be filled by
sawdust and oil but also can make sawdust and oil disperse evenly on the surface of pore, so as
to oxide and combustible substance can be blended tightly. In addition, it can maintain
adequate pore, namely adequate hot spot, so as to make sure it can reach the best detonation
sensitivity and explosion performance, which is the substantial basis of self-sensitization to
succeed and be applied practically.[10-17,19,20]
264
2.
EXPANSION MECHANISM OF AMMONIUM NITRATE [6,7]
After the procedure conditions and the influence of surfactant to expansion effect being
studied, it is presumed that the expansion mechanism of AN could be described as following:
under certain temperature and vacuum, the water in saturated solution of AN would evaporate
and form a huge number of stable bubbles with the action of surfactant. As a result, if the
temperature and pressure get a particular value these bubbles would breakdown and have its
water evaporate rapidly, then the crystal of AN would form a great number of pores, holes
and interspaces and pointed surface, which would cause particles to join together in a
extremely irregular way. The expanded AN produced form this process is light, porous and
loose.
Research results show that evaporation under reduced pressure itself has somewhat effect
on the expansion of AN. At the same time, the application of surfactant would also reinforce it.
Experimental results indicate that whether the expansion effect is stand or fall is in line with
the frothing capability of surfactant.
Research results also show that composite surfactants have satisfactory frothing capability
to saturated solution of AN. Among the composite surfactants, the composite amphoteric
surfactant has the best surface activity and frothing capability. Composite surfactant would
have its components intensify their activity each other. The composite amphoteric surfactant
has the best interaction parameter. The effect of surfactant used in expansion modification of
AN mainly act as vesicant, coating agent and crystal modifier.
3.
TECHNICAL CHARACTERISTIC
OF EXPANDED AMMONIUM NITRATE [6,7]
The expanded AN treated with surfactant have a porous and structure by SEM, which is
shown as Fig.1
(a) Common AN(900 times enlarged)
(b) Expanded AN(850 times enlarged)
Fig. 1. Micro-structure of ammonium Nitrate(AN)
265
Comparing with common AN and grainy AN and porous grainy AN, expanded AN has
some unique characteristics:
(1)
Unbelievable specific surface area. Its specific surface area is 3328.54cm2g-1 by

drop-with-resistance method. In contrast, common AN has only a specific surface
area of 758.76 cm2g-1 under the same condition. So it increases almost 4.4 times.
(2)
A great many of micro bubbles in its particles. It is obviously from Fig.1 that
there are numerous holes and cracks inside while many burrs and edges outside,
which ensure that the expanded AN can be sensitized by itself. This is the physical
foundation to get rid of TNT.
(3)
Decreased hygroscopicity. Its hygroscopic increment at 20 and 93 relative

humidity is shown in Table 3.
Table 3. Hygroscopic increment of AN(%)

Hygroscopic time/h
Pure AN
Expanded AN
#
2 rock ammonite explosive
Rock expanded AN explosive
2
0.5832
0.4258
0.5569
0.3958
4
1.1084
0.7697
1.1100
0.6432
6
1.5792
1.1212
1.5456
0.9397
8
2.0660
1.4829
2.00014
1.3025
10
2.5370
1.8746
2.4811
1.5904
It is obvious from Table 3 that the average hygroscopic rate of expanded AN is only half
of that of common AN. Expanded AN and common AN are put into a surrounding of 80 for
3h respectively then cooled to room temperature. Repeating this process for 5 times and
measure its ultimate pressure endured and the data are shown in Table 4.
Table 4. The pressure endured by charge through cycling test

Loading
Common AN
Expanded AN
Pressure endured/MPa
8.06
0.81
8.11
0.81
7.99
0.80
8.02
0.83
Average pressure
endured/MPa
8.06
0.813
Analyzing Table 4, we can found that the ultimate pressure endured of expanded AN is
1/81/9 over that of common AN. So we can conclude that expanded AN has good
anti-caking capability.
(4)
Brittleness. So it is quite easy to treat it to fine powder. The expanded AN

produced freshly would become brittle along with the drop of temperature. After
being ground softly for 20 min by wheel-grinding machine, 90% particles of it
would have its diameter less than 250m, which could satisfy the fine need of AN.
266
4.
THE APPLICATION OF EXPANDED AN

4.1
Rock expanded AN explosive [9]
LC-2 surfactant(0.15% over the weight of AN) is solved in the solution of common AN
then undergoes evaporation and crystallization to get expanded AN. In succession, 92%
expanded AN, 4% wood powder and 4% composite oil are added into wheel-grinding machine
all together to mix for 20min. as a result, rock expanded AN explosive is got.
The prescription and performance parameters of rock expanded AN together with 2# rock
ammonite explosive and 2# rock ammonite-oil explosive are shown in Table 5 and Table 6.
Table 5. The prescription of three powdery industrial explosives

Explosive
w(AN)/%
w(TNT)/%
w(oil)/%
11.01.0
w(wood
powder)/%
4.00.5
w(addictive)/
%
2# rock ammonite
explosive
2# rock ammonite-oil
explosive
Rock expanded AN
Explosive
85.01.5
87.51.5
7.00.7
4.00.5
1.50.3
0.1000.005
92.02.0
4.00.5
4.00.5
0.1200.005
Table 6. The detonation performance of three powdery industrial explosives

Item
Loading
density/gcm-3
Gap distance/cm
Brisance/mm
Strength/ml
Detonation
velocity/ms-1
w(water)/%
Storage
lifetime/a
2# rock ammonite
explosive
0.951.10
2# rock ammonite-oil
explosive
0.951.05
Rock expanded AN
Explosive
0.851.00
4
12
320
3200
4
12
320
3300
5
14.516.0
330380
32003500
0.30
0.5
0.30
0.5
0.30
0.5
The data shown in Table 6 indicate that the detonation data of rock expanded AN
explosive reach, even exceed, those of 2# rock ammonite explosive and 2# rock ammonite
explosive.
Rock expanded AN explosive has another advantage of very distinctive econocimal
benefit. Because of being free of TNT, the cost of raw materials have been cut down for
300500 yuan(China)/t so the benefit of product have being increased greatly. In addition,
267
some investments of the equipment and workshop for wheel-grinding procedure for TNT are
saved. So does the cost of disposal of waste water as well as staff and power consume, etc.
At the same time, rock expanded AN explosive has great social benefit because TNT is
excluded, so the harm to workers bodies and environment is decreased.
4.1
Permitted expanded AN explosive [3]
Permitted expanded AN explosive is among the serial products of expanded AN

explosives with a prescript of 79% AN, 3% oil, 15% common salt and 3% wood powder.
Up to now, this kind of explosive has been undergone test-production in some factories.
This technology is much accounted by Coal Industry Bureau of China and has been ranked as
key generalizing item of the ninth-five-year plan in coal-mining industry.
4.2
Seismic expanded AN explosive [6]
Research results indicate that expanded AN could replace common AN to be used in

seismic explosive, which may decrease its cost greatly and ensure production safety and
alleviate pollution but would not reduce its performance.
A product line with a output of 4000t over a capability of 5000t expanded AN has been
established in Huber Xiangsha Chemical Factory. The product has excellent performance and
high stability.
4.3
Low detonation velocity of expanded AN explosive [18]
Low detonation
velocity explosive is widely used in explosive welding (or
compounding), smooth explosion, and lately it has been applied increasingly in petroleum
industry.
The low detonation velocity explosive with universal composition (Expanded AN
72~83%, wood powder 2.5~3.5%, composite oil 2.5~3.5%, pearlite 5~9%, sodium chloride
5~8%, stabilizer 2~4%) has satisfactory performance: detonation velocity is 2152ms-1,
sensitivity to initial detonation is 4cm and charge density is 0.62gcm-3.
Fruit tree blasting agent also belongs to low detonation velocity explosive contained
nutrient which can promote growing of fruit tree.
The compositions and characteristics of fruit tree blasting agent are: expanded AN
80.0~85.0%, composite oil 2.0~4.0%, nutrient 12.0~14.0%, moisture content 0.30%,
detonation velocity(D)2600 ms-1, brisance(HLe) 10mm, strength(VL) 280ml, diameter of
the soil loosened by blasting1.1m, nitrogen content15.0%, other nutrient content10.0%,
sensitivety to impact0%, sensitivity to friction0% and noxious gas content after blasting
100Lkg-1.
Expansion technology of AN has being widely popularized in China because of the low
cost, high safety and less pollution.
268
CONCLUSION [19,20]
5.
(1)
According to the design idea of self-sensitization of AN, we could apply

surfactant technology. Then under the action of surfactant, expanded AN is
produced with many advantages: possession of tremendous micro-pores, huge
specific surface area and excellent hygroscopicity and caking performance.
Expansion technology, original and innovative in the world, breaks a new way for
research on and application of powdery industrial explosive.
Although the rock expanded AN explosive is free of TNT, it has satisfactory
detonating and physical characteristics and reaches advanced level in the would.
(2)
Rock expanded AN explosive is characteristic of abundant raw material, low cost

and simple product process, especially remarkable economical benefit derived
from low cost and social benefit because of getting rid of TNT.
(3)
The production technology of expanded AN has been adopted by more than 80

factories. Accordingly, powdery industrial explosive prepared from it has a total
output of around 1200 kt in 213 plants all over the China. In a word, expansion
technology of AN is prospective in popularization and application.
REFERENCE
[1]
DING YUN: The influence of surfactant on modification of ammonium nitrate, [D], Nanjing
China: East-china Institute of Technology, 1990
[2]
L&u& CHUNXU: Theory and technology of surfactant, [M], Nanjing China: Jiangsu Science press,
1991
[3]
CHEN TIANYUN: Research on surface characteristic of ammonium nitrate, [D], Nangjing:

East-China Institute of Technology, 1992
[4]
L&u& CHUNXU: Surface chemical research on expanded,
AN [J], Journal of Nangjing
University, 1995(31):286
[5]
LIU ZULIANG,HUI JUNMING, L&u& CHUNXU: Light ammonium nitrate and HF-type ammonium
nitrate explosive, [J], Explosive materials,1991, (5) :5
[6]
LU MING: Study on expanded ammonium nitrate explosive, [D], Nanjing China: Nanjing
University Of Science & Technology,1999
[7]
YE ZHIWEN: Study on expanding mechanism of ammonium nitrate, [D], Nanjing China: Nanjing
university of science & technology, 1995
[8]
CHEN TIANYUN, L&u& CHUNXU, YE ZHIWEN: Study on performance of modified ammonium

nitrate, [J], Energetic Materials,1996, (4) :169
[9]
L&u& CHUNXU, LIU ZULIANG, HUI JUNMING, WANG YILING: Study on rock Expanded, AN [J],
Explosive Materials,1997,(1):5
269
[10] L&u& CHUNXU, LIU ZULIANG, HUI JUNMING: Form and develop of the self-sensitization of
expanded AN Light ammonium nitrate and HF-type ammonium nitrate explosive, [J], Chinese
Journal of Explosives and Propellants, 2000, (4) :1
[11] L&u& CHUNXU, LIU ZULIANG: Basic and experiment of the theory of self-sensitization of
Expanded AN[J], Explosive Materials, 2000, (4) :1
[12] L&u& CHUNXU: Study on the theory of self-sensitization of Expanded AN[J], China Enginering
Science, 2000 (11) :73
[13] L&u& CHUNXU, LU MING, CHEN TIANYUN YE ZHIWEN: Study on tiny bubble of the theory of
self-sensitization of Expanded AN[J], Acta Armamentarii, 2001 (4) :485
[14] L&u& CHUNXU: Study on the crystal characteristics of Expanded, AN[J], Acta Armamentarii,
2002 (3) :316
[15] YE ZHIWEN, LIU ZULIANG, L&u& CHUNXU: Study on an improved, AN[J], Acta Armamentarii,
2002(3): 412
[16] CHEN TIANYUN: Study on the self-sensitization of Expanded AN and its application, [D],
Nanjing China, East-china Institute of Technology,2000
[17] ZHOU XINLI: Study on the self-sensitization of Expanded AN and its application, [D], Nanjing
China:East-china Institute of Technology, 2003
[18] XU BINCHEN, L&u& CHUNXU: Research on Compositions of Low detonation velocity expanded
AN explosive, Journal of Propellants and explosives, 2000(3):1
[19] L&u& CHUNXU, ETC.: Expanded AN Explosives, Beijing, Ordnance industrial press, 2001
[20] L&u& CHUNXU, ETC.: Theory of industrial explosive, Beijing, Ordnance industrial press, 2003
270
CONCEPT OF THE SPATIAL-PERIODIC MICRO-STRUCTURES

EXCITATION AT THE EVAPORATED ENERGETIC MATERIALS
TRANSIENT COMBUSTION AS A WAY FOR IMPROVEMENT
OF THE SOLID PROPULSION TECHNOLOGIES
Alexander N. Lukin
Physics-Chemical Mechanics Department, Institute of Applied Mechanics,
Ural Branch of the Russian Academy of Sciences, Bldg. 34, T.Baramzinoy Str.,
426067, Izhevsk, Udmurtia Republic, Russian Federation
Abstract:
In connection with development of solid propulsion systems (SPS) of a new generation
with high energy and mass characteristics, the problem of prevention of development of
combustion instability and anomalies of burning of the energetic materials (EM) again
began to have extreme importance. For successful solution of problems, connected with
suppression of the SPS combustion instability are necessary have detailed
understanding about essence of mechanisms of the EM unstable burning on the new
qualitative level. The experimental data independently obtained by various authors and
for various kinds of evaporated EM shows the on the burning surface are formed the
existential cellular structures. This phenomenon was a subject of many investigations
already more than four decades, however within the framework of the combustion
classical theory the mechanism of its excitation and development does not find an
adequate explanation. Detailed analysis and mutual comparison of investigations
executed in this area and also detailed analysis of the experimental data, obtained in
adjacent areas of science and engineering has allowed suggesting the new mechanism
of occurrence of this phenomenon. This new concept connected, mainly, with excitation
of the synergetic dissipative spatial-periodic micro-structures (SPMS) in the thin liquidviscous layer and on the EM burning surface and determining the burning wave spatial
instability. Obviously, it is possible to speak about the fundamental law determining the
processes of the EM burning: the phenomenon of the SPMS formation has universal
nature. On the EM burning surface occurs transition from the isotropic medium to the
medium with existential structure. The process of the SPMS formation provides
influence on the process of agglomeration and on the spectrum of agglomerates on the
burning surface of heterogeneous EM. Within the scope of suggested concept, the
physical-chemical effects taking place at erosive burning of the EM, receive a new
explanation. On the basis of suggested concept the new technologies for suppression of
combustion instability are developed. Application of suggested technologies in practical
systems will allow control the development of physical- chemical processes at a new
qualitative level.
Keywords:
evaporated energetic materials, burning surface, unstable combustion

mechanisms, cellular-pulsating burning phenomena.
271
1.
INTRODUCTION
In connection with development of solid propulsion systems (SPS) of a new generation

with high energy and mass characteristics for various aero-space applications (Fig 1 Fig 3),
the problem of prevention of development of combustion instability and anomalies of
burning of the energetic materials (EM) again began to have extreme importance. For
successful solution of problems, connected with suppression of the SPS combustion
instability are necessary to have detailed understanding about essence of mechanisms of the
EM unstable burning on the new qualitative level.
Fig 1. Sectional view of the space solid propulsion systems
Fig 2. Sectional view of high-loading-density sustainer solid

propulsion system design of the
intercontinental ballistic rocket
upper stage
Fig 3. Sectional view of the space solid

propulsion system
The EM that melt and evaporate in the combustion wave are widely used in various
aero-space applications and SPS. The combustion mechanism for such materials is
characterized by melting of the material at a temperature below the burning surface
temperature and simultaneously occurring processes of thermal decomposition and
vaporization in the surface layer.
272
In the past a lot of effort has been put into the development of models for transient
burning of solid propellants. The experimental data independently obtained by various
authors and for various kinds of evaporated EM shows the same on the burning surface are
formed the cellular structures. This phenomenon was a subject of many investigations
already more than four decades, however within the framework of the combustion classical
theory the mechanism of its excitation and development does not find an adequate
explanation.
The principles of the EM combustion anomalies theory were established by outstanding
Russian scientist Ya.B.Zel'dovich. [1, 2] However, this theory does not allow explain the
reason of the cellular structures formation on the burning surface of evaporated EM.
For the first time the conception of formation of the cellular structures on the burning
surface of evaporated EM the conception of cellular-pulsating burning (CPB) has been
suggested by K.I.Synaiyev in 1968. [3] And since that time exists the scientific problem of
adequate understanding of the mechanism of development of this phenomenon.
The extensive experimental researches executed by V.N. Marshakov (Semenov Institute
of Chemical Physics of the Russian Academy of Sciences, Moscow, Russia) have shown the
complex and ambiguous nature of the given phenomenon. He has executed a large number
of experiments for understanding of this phenomenon. Investigations of V.N.Marshakov
have shown, that for majority of gasified EM is typical the CPB phenomena significantly
influencing on the combustion processes stability. [4 8] These phenomena were observed in
all propellants investigated (in double-base, composite solid propellants, nitrocellulose and
same individual explosives) under various experimental conditions (at pressures up to about
80 MPa), by means of different experimental techniques (visual and cinema observation,
measurements on extinguished propellant surface, analysis of electric conductivity and
radiation fluctuations, and temperature measurements with thermo-couples). The CPB
phenomenon is observed not only near the critical conditions of burning, but also at the
increased pressures and initial temperatures. The experimental data have shown that this
phenomenon develops irrespective of properties and structure of the specific EM. At
realization of this burning mode the sizes of burning cells are not connected with sizes of
components and the sizes and structure of tested EM. For example, for composite solid
propellants, the dispersiveness of the components (for example, the sizes of crystals of the
ammonium perchlorate) does not influence on the CPB parameters. Structural heterogeneity
of the EM does not interrupt development of this phenomenon.
The EM CPB phenomenon has been a subject of many investigations already more than
four decades, however within the framework of the combustion classical theory the
mechanism of its excitation and development does not find an adequate explanation.
Existence of scientific problem connected with studying of the cellular structures on the
burning surface and the CPB phenomenon also is underlined in the recent papers of
Arkhipov V.A., Bondarchuk S.S. and Vorozhtsov A.B. (Tomsk State University and
Institute of Problems of Chemical & Energetic Technologies, SB of the RAS, Russia) [9] ,
Korobeinichev O.P., Paletsky A.A., Tereschenko A.G., and Volkov E.N. (Institute of
Chemical Kinetics and Combustion, Siberian Branch of the RAS, Novosibirsk, Russia) [10
12]
and Yarushin S.G. (Perm State Technical University, Perm, Russia and Developmental
Bureau "TEMP", Perm, Russia) [13]. The extensive experimental investigations in this area
have shown the complex and ambiguous nature of the given phenomenon.
273
2.
SPATIAL-PERIODIC CELLULAR MICRO-STRUCTURES

ON THE BURNING SURFACE: SOME EXPERIMENTAL DATA
Micro-video-recording of the EM samples burning surface has allowed observe behavior

of the spatial-periodic cellular micro-structures (SPMS) on the burning surface. [5, 7, 14]
The images of video film of burning of the NB powder at the atmospheric pressure are
shown on the Fig. 4. Diameter of the sample was equal to 0.012 m., the initial temperature of
the sample is equal to 293 K and the time interval between the images is 80 milliseconds.
Fig 4. Sequential images (the time interval between the images is 80 milliseconds)
of the video film of burning of the NB powder at the atmospheric pressure. [7, 14]
The images of video film of burning of the colloxylin sample at the atmospheric pressure
are shown on Fig. 5. [7, 14] Diameter of the sample was equal to 0.010 m., the initial
temperature of the sample is equal to 343 K and the time interval between the images is 40
milliseconds.
Fig 5. Sequential images (the time interval between the images is 40 milliseconds) of the
video film of burning of the colloxylin sample at the atmospheric pressure. [7, 14]
The images of video film of burning of the sample of model nitroglycerin gunpowder
A at the atmospheric pressure is shown on Fig. 6. [7, 14] Diameter of the sample was equal
to 0.012 m. The view of the burning surface from the bottom end face of the sample, "on the
translucence". On the figures it is possible to see the cellular micro-structures that develops
on the burning surface. The burning cells appear and disappear on the burning surface.
274
Fig 6. Sequential images (the time interval between the images is 80 milliseconds) of the
video film of burning of the sample of the model nitroglycerin gunpowder A. [7, 14]
The typical shapes of extinguished burning surfaces of the pyroxylin samples is shown
on Fig. 7. [5] Diameter of the samples was equal to 0.01 m and the density is equal to 1500
[5]
At the left side - the imprint (trace) of the burning front on the plasticine (modeling
kg / m 3 .
clay) at the atmospheric pressure, upon the air. At the right side the extinguished burning
surfaces of the pyroxylin sample. The experiment was executed in the nitrogen atmosphere
at pressure equal to 0.51 MPa. The burning surface on the extinguished samples (Fig. 7) is
not quite planar, and represents a set of superficial (shallow) craters (cells). The crosssectional sizes of these craters do not depend on diameter of the sample and at (10 15)
times more than characteristic thickness h h of the EM heated up layer at burning.
Fig 7. Typical shapes of the burning surfaces of the pyroxylin samples. [5]
On the Fig. 4 - Fig. 7 it is possible to see that the boundaries of craters (cells) forms
similarity of a grid. The sizes of the burning cells decrease with the pressure increase and do
not depend on the specimen diameter. However the ratio of the cells sizes to the thickness of
the heated up layer is kept by a constant. Also, the sizes of cells have correlation with the
size of critical diameter of burning which exceeds the cross-sectional size of the cells in 1.52 times.
The experimental data have shown that CPB phenomenon develops irrespective of
properties and structure of the specific EM. At realization of this burning mode the sizes of
burning cells are not connected with sizes of components and the sizes and structure of
tested EM.
275
3.
TEMPERATURE FLUCTUATION ON THE BURNING SURFACE
Temperature distributions in the burning wave for nitroglycerin powder A were

measured with using of two thermocouples located in the same cross section on the distance
of (4-5) mm from each other. [7, 14] Figure 8 and Figure 9 shows an examples of such
temperature variations at the burning wave propagation through the nitroglycerin powder
A sample at the atmospheric pressure in the nitrogen atmosphere. Similar temperature
distributions were observed and for the N powder. From the figures it is possible to see the
thermal instability of the burning wave front - the phenomenon of temperature fluctuation
on the burning surface.
In the papers [10 12] the mechanism of combustion of the composite solid-rocket pseudopropellant based on ammonium dinitramide ADN and polycaprolactone (PCL) at 0.1 MPa
has been studied. Initially, fuel (PCL) and oxidizer (ADN) are distributed uniformly in
propellant mixture. During combustion of the ADN/PCL(1250) at 0.1 MPa, redistribution of
fuel on the burning surface, which is caused by formation of carbon-containing drops, takes
place. It results in variation of oxidizer/fuel ratio in the gas phase near the burning surface.
Separate flame jets moving over the strand burning surface during the combustion were
observed. So, the combustion of ADN/PCL(1250) at 0.1 MPa has a torch character with
formation of separate seats (cells) of burning on the burning surface. During the combustion
Fig 8. Temperature variations at the burning wave propagation. [7, 14]
276
Fig 9.
Temperature variations at the burning wave propagation. [7, 14] Figure shows
the signals from two thermocouples in the same cross-section of the sample.
of the ADN/HTPB (97/3) at 0.10.6 MPa, separate seats (cells) of burning were also
observed. The probe (or thermocouple) during the combustion was located either in the
luminous zone (torch) or in dark zone (between torches or far from them). Results of two
experiments on measurement of the temperature profile in the ADN/PCL(1250) flame at 0.1
MPa, which confirm the conclusion regarding torch combustion of this propellant, are
presented in Fig. 10. [10 12] Curve 1 in Fig. 10 corresponds to the case when the
thermocouple moved from the torch to the burning surface. Abrupt fallings of temperature
on curve 1 are connected with changing of torch location with respect to thermocouple.
Curve 2 corresponds to the case when the thermocouple moved to the burning surface from
the dark zone. As will be shown below, such temperature distribution is not random or
chaotic and is not connected with non-perfect technique or errors of measurements. The
spread in the T S values is connected with a real existing set of the T S values appropriate to
various burning rates in pulsations.
Fig 10. Temperature profiles in fame of propellant ADN/PCL(1250) at 0.1 MPa. [10 12]
277
4.
MICRO-OSCILLATIONS OF ELECTRIC CONDUCTIVITY

IN THE COMBUSTION ZONES
One of methods for investigation of the mechanisms of solid propellants burning is

measurement of electric conductivity of the reaction layer and burning surface. The electric
conductivity value characterizes concentration of the burning cells on the solid propellant
surface. Measurement of electric conductivity of ammonium perchlorate (AP) reaction layer
and burning surface, at low pressure, shows intensive low-frequency pulsations. [15] At
transition to high pressures, the amplitude of these pulsations gradually decreases. And at
high pressures is registered change of the conducting layer structure. [8] In the paper [16] is
shown, that the amplitude of electric conductivity micro-oscillations permits determination
of amount of extinguished parts on the burning surface. After beginning of the pressure
decrease, the burning surface becomes non continuous, and the burning wave has a non onedimensional structure. For an example, diagrams of electric conductivity of the surface layer
of pure AP at pulsating burning are shown on Fig. 11 ( P = 3 4 MPa; 1 a sample having
profile (2.7 x 12) mm, 2 (12 x 12) mm. ). [16]
Fig 11. Diagrams of electric conductivity of the surface layer of pure AP at pulsating
burning. [16]
Also, investigations of electric conductivity micro-oscillations in the combustion zones
of the solid propellants was done in the Scientific-Research Institute of Polymeric Materials
(Perm, Russia). The method of measurement of micro-oscillations of electric conductivity in
the EM combustion zones was used for detailed investigation of the non-simultaneous
burning out of the EM components. The method are presented in papers. [17, 18] In particular,
measurement of the electric conductivity of the EM combustion zones allows observe
pulsation of the burning cells concentration in the combustion zone and deviations of
unstable oscillatory process directly in the combustion zones. The experiments show direct
dependence of unstable combustion of EM from the regular micro-oscillations in the EM
combustion zones.
278
5.
CELLULAR-PULSATING BURNING PHENOMENA:

CHAOS OR SELF-ORDERING ?
Analysis of micro-pulsations on the EM burning surface are presented in papers of

N.M.Pivkin. [17, 18] As the reason for appearance of micro-pulsations on the burning surface,
the non-simultaneous burning out of the EM components was considered. The idea of nonsimultaneous burning out of the EM components was suggested at the end of 19th century by
D.I.Mendeleyev. [19] Authors have shown that each specific solid propellant has its intrinsic
(own) frequencies of micro-oscillations in the combustion zones. [17, 18] Thus, the new
universal integrated parameter ("intrinsic frequencies of micro-oscillations on the burning
surface of propellant") was entered. This parameter allows characterize the most probable
range of the acoustic oscillations, radiating from the burning surface of each specific
propellant, from the point of view of the phenomenon of the non-simultaneous burning out
of the components.
The results, obtained in papers [17, 18] shows that in case of coincidence of the "propellant
intrinsic frequencies" with the acoustic frequencies of the charge channel cavity (combustion
chamber cavity), are observed the oscillations of the combustion products (CP) pressure and
the combustion instability. However, these experimental results and the theory [17, 18] have
not allowed explain the reason of the spatial-periodic cellular micro-structures excitation on
the burning surface and all other mechanisms of the CPB phenomena. At the same time, the
"propellant intrinsic frequencies of micro-oscillations" are display existence of some
universal mechanism.
Present-day theoretical concepts of CPB phenomenon are based, mainly, on the ideas
generated under influence of researches of processes of self-propagating high-temperature
synthesis (SHS) on the models of spinning combustion processes developed for description
of dynamics of the SHS front structure. [20] Spinning waves at the SHS represent a spiralhelix motion of a localized combustion center, which occurs along the lateral surface of a
cylindrical sample. While the combustion center moves along a spiral path, its dimensions
do not vary.
In particular, theoretical explanation of the CPB phenomenon suggested by V.N.
Marshakov is based on this concept. V.N.Marshakov has suggested theoretical model of the
burning wave thermal instability. [7, 14, 21] In accordance with this model, on the burning
surface there is a mobile system (grid) of burning transverse waves, which is the waves,
spreading on the heated-up layer, in the transverse direction to the burning propagation.
Between these transverse waves burning stops. In accordance with theoretical scheme of
V.N.Marshakov, the transverse burning waves are propagating one after another. Also, in
accordance with this model the burning cells are considered as a wave cells and is supposed
that the picture of this phenomenon corresponds to the theoretical model of transverse spin
burning waves at the SHS.
These theoretical model is very similar to theoretical model of spiral propagation of the
burning wave, which for the first time was suggested by V.N.Fomenko, in 1990 (The
Federal Centre of Dual-Use Technologies SOYUZ, Dzerzhinsky, Moscow Region,
Russia). V.N.Fomenko has used this model for description of the end-burning charges
rhythmic extinctions phenomenon.
However, use of theoretical concepts from the area of SHS does not allow explain the
mechanism of formation of the burning cells on the burning surface of the EM and solid
rocket propellants. Firs of all there is no full analogy between the burning processes of the
279
solid rocket propellants and the SHS-systems. At the SHS the condensed reaction products
continue produce essential influence on further development of the burning process. The
solid-flame combustion [20, 22] is an self-wave chemical process in the system of solid-phase
reagents. This process leads to formation of the intermediate and finished products in the
solid-phase. The solid-flame combustion is only solid-phase process, not connected with
formation of liquid and gaseous phases. For example, in the reaction zone there is no liquidviscous layer (LVL). After the burning wave propagation, in the material develops the
volumetric physics-chemical processes that not connected with burning. In these conditions,
considerable influence on the process development is provided by significant heterogeneity
of the EM structure. For example, for the SHS-systems, the transverse burning waves may
exist at various ratios between propagation velocities of the transverse and longitudinal
burning waves. However, at the gun-powders and solid propellants burning, the transverse
burning waves can be considered only if their propagation velocity considerably exceeds the
longitudinal burning wave propagation velocity. The transverse burning waves exist, but
only for SHS-systems.
For analysis of system of the burning cells on the burning surface, G.V.MelikGaiykazov (Semenov Institute of Chemical Physics of the Russian Academy of Sciences,
Moscow, Russia) has suggested to use the theory of clusters. [16, 23] In particular, was
suggested consider the system of burning cells as infinite cluster that disintegrates at the time
of the cells extinction. However, the mechanism of formation of the burning cells on the
burning surface was not explained.
On the other hand, behavior of the complex micro-structures, developing on the EM
burning surface is unanalyzable by traditional methods. V.N.Marshakovs theoretical model
of the burning wave thermal instability, considers only separate components and connections
between them. As noted by B.V. Novozhilov (Semenov Institute of Chemical Physics of the
Russian Academy of Sciences, Moscow, Russia) [24], the solid propellant burning surface
represents oscillatory system with infinite number of freedom degrees. However, such
composite systems cannot be understood, analyzing their parts separately.
A lot of efforts has been put into the development of models for transient burning of
solid propellants.
In B.V.Novozhilov's recent paper [25] was suggested synergetic approach for
consideration of the problem of combustion instability - the anomalies of the EM burning
and the phenomenon of non-acoustic instability (the "chuffing" phenomenon) in the
combustion chambers. In particular, was suggested consider the processes on the solid
propellant burning surface as a chaotic process. However the CPB phenomenon was not
considered.
The CPB phenomenon is similar to the phenomenon of hydrodynamic instability at
transition of laminar flow into the turbulent flow, connected with self-organizing of dynamic
dissipative spatial-periodic vortex micro-structures. In this case some part of energy of the
system will be transformed in the organized eddy movement. Until recently such
hydrodynamic instability was identified as transition to chaos. However last investigations
have shown that in this case, vice versa, takes place transition to the order, i.e. to formation
of the self-ordering vortex structures.
280
6.
FORMATION OF THE SPATIAL-PERIODIC MICRO-STRUCTURES

AT THE PHASE TRANSITIONS UNIVERSAL PROPERTY OF THE
EXISTENTIAL CONTINUUM
The universal phenomenon of the SPMS formation is very wide-spread in the natural
objects.
The first, who has paid attention on existence of the common universal law of formation
of the SPMS in the natural objects and the processes of self-organizing of the dissipative
structures in the nature, was well-known German astronomer and mathematician Johannes
Kepler. In 1611 he published a short treatise On the Six-Cornered Snowflake, which is
perhaps the first scientific reference to snow crystals. [26] In his treatise Kepler ponders the
question of why snow crystals always exhibit six-fold symmetry. Kepler was astute in
recognizing that the genesis of crystalline symmetry was an interesting scientific question,
and he also realized that he did not have the means to answer it. In particular, Kepler
suggested that all forms and symmetries in the nature are subordinated to the "Forming
Force" of the Earth and develop under influence of this force (Fig. 12, A). Such processes of
forming are not connected with the properties of specific substances.
Fig 12. Examples of formation of the SPMS in the natural objects

The next step in understanding of processes of formation of the SPMS was made
through three centuries by French fluid mechanist Bnard. In 1900 he investigated a fluid,
with a free surface, heated from below in a dish, and noticed a rather regular cellular pattern
of hexagonal convection cells (Fig. 12, B). Benard experimented on only very thin layers (a
millimeter or less) that had a free surface and observed hexagonal cells when the convection
developed. The convection cells appearing at marginal instability are called Bnard cells
(also called Rayleigh-Bnard convection cells) in honor of French fluid mechanist
H.B.Benard who performed pioneering experiments on convective instability in his doctoral
thesis in 1900. [27] Namely after Benard experiments, any convection cells often names as
Benard cells. The Benard cells is a classical example of synergetic phenomena. The
mechanism of formation of such cells is actuated by buoyancy (lifting) force F L .
Stimulated by these experiments, Rayleigh in 1916 explained this in terms of a buoyancy
driven instability, by doing a linear expansion of the equations in the fluid velocities. It was
later understood that in buoyancy driven convection the expected pattern would be a stripe
pattern of convection rolls, rather than the cellular pattern observed by Benard. (This is a
non-linear phenomena, that was not accessible by the methods Rayleigh used). The
convection observed by Benard is now understood to be driven by temperature dependent
surface tension forces rather than by buoyancy. Nevertheless, the stripe or roll state formed
281
in buoyancy driven convection is today referred to as Rayleigh-Benard convection. Surface

tension induced convection is known as Maringoni convection.
In 1949 B.P.Zhukov (Scientific-Research Institute 125 (Scientific Research
Chemical-Technological Institute), Dzerjinskiy, Moscow Region) has shown, that at the EM
burning on the boundary between of solid and gas phases there is a LVL. [28] These data were
received by experimental way. Namely from that time has appeared opportunity for analyze
the EM combustion instability and the abnormal physics-chemical processes at the EM
burning at a new qualitative level.
At the last decades were obtained the experimental data on excitation of convective
flows and the convection cells at heating of the solids from above, from the side of free
surface. [29 34] Such convective, cellular movement is observed at smelting of solids by laser
radiation and especially in experiments on alloying of the steels and on preparation of the
semi-conductor alloys. It is necessary underline, that between processes of formation of
SPMS at the EM burning and at smelting of the solids under influence of the laser radiation
energy there is an analogy. In both cases on the surface of materials there is the LVL that is
heating up from above. At the same time, at the influence of the laser radiation energy on the
propellant surface, the LVL can be absent.
However, such cellular movements can not be excited by the non-electric mechanisms.
As shows the experiments, under influence of heating can arise the electric charge and the
electric field. The experimental data have shown that the SPMS develops on the surface of
solids under influence of the laser radiation energy. [29 34] In particular, the experiments
have shown that the SPMS do not arise at the temperatures lower than melting point. At once
above the melting point arise the non-stationary SPMS which disappear through tens of
nano-seconds. At increase of the intensity of influence, is observed the SPMS formation,
which are kept after the ending of the impulse. Also, the experiments have shown, that at the
SPMS formation on the surface of materials is observed the strong increase of the optical
radiation absorption by the surface of these materials. As show the experimental data, the
thermo-electric mechanism is excited in the melted layers, which thickness less than h = (1.0
- 0.1) mm. These are the most thin layers that observed by the experimental way. Taking into
account above mentioned experiments, it is necessary underline, that characteristic thickness
of the solid propellant heated up layer at burning (and especially LVL) much less than this
critical thickness of the melted layer.
Moreover, available experimental data [35] indicate on abnormally-fast redistribution of
impurity in the liquid phase at superficial alloying of metals which cannot be explained
within the framework of usual convection and diffusions, and, apparently, testifies on
occurrence of the hydro-dynamical flows in the laser fusion.
Both the experiment, and the theory confirm, that the SPMS formation is rather
universal phenomenon. Such micro-structures arise at the diversified modes of the materials
processing by laser radiation.
7.
PHENOMENON OF SPATIAL-PERIODIC MICRO-STRUCTURES

EXCITATION
Detailed analysis and mutual comparison of investigations executed in this area

(K.I.Synayev, B.V.Novozhilov, Merrill W.Beckstead, Kenneth K.Kuo, Herman
F.R.Schoyer, Vigor Yang, Fred E.C.Culick, Luigi T.DeLuca, B.P.Zhukov, A.D.Margolin,
P.F.Pohil, V.N.Marshakov, A.G.Istratov, N.M.Pivkin, G.V.Melik-Gaykazov, I.G.Assovskii,
V.Ye.Zarko, V.N.Fomenko, O.Ya.Romanov, V.A.Babuk, S.G.Yiarushin, S.A.Rashkovskii,
282
O.P.Korobeynichev), and also detailed analysis of the experimental data, obtained in

adjacent areas of science and engineering has allowed suggesting the new mechanism of
occurrence of this phenomenon. This new concept connected, mainly, with excitation of the
synergetic dissipative spatial-periodic micro-structures (SPMS) in the thin LVL and on the
EM burning surface and determining the burning wave spatial instability. Obviously, it is
possible to speak about the fundamental law determining the processes of the EM burning:
the phenomenon of the SPMS formation has universal nature. As the system becomes
complicated (during heating up), it gets such phenomenological features that are difficult for
finding out, at studying of more simple subsystems. On the EM burning surface occurs
transition from the isotropic medium to the medium with existential structure.
In accordance with suggested theoretical model, at heating from above in the thin LVL
occurs interaction of the hydrodynamic, electric and thermal subsystems of the disordered
system - the thermo-electric convection excitation. The thermo-electric mechanism can
induce instability of the LVL and excitation of cellular movement and formation of the
synergetic SPMS. Excitation is possible at any direction of heating, including heating of the
LVL from above. Besides the velocity (convection) cells, in the LVL arise the electromagnetic field cellular structures.
Thus, in comparison with B.V.Novozhilovs paper, suggested concept asserts the
opposite opinion. On the solid propellant burning surface there is no chaotic combustion, but
on the contrary, takes place the order the process of self-organizing of dissipative
hydrodynamic micro-structures. This phenomenon is similar to the phenomenon of
hydrodynamic instability at transition of laminar flow into the turbulent flow, connected with
formation of dynamic dissipative spatial-periodic vortex micro-structures.
The reality of existence of thermo-electric convection excitation phenomenon is
confirmed by large volume of experimental works executed in adjacent areas of science and
engineering and now cannot cause any doubts.
In 1976 I.V.Ioffe and E.D.Eidelman of Russia [36] have shown that thermo-electric
mechanism of excitation is characterized by the dimensionless Eidelman number Ed:
Ed = ( 2 A 2 h 2 ) ( a )
(1)
The dimensionless parameter Ed shows condition of excitation, that is the ratio of

destabilizing effect of the volumetric thermo-electric (electrostatic) force arising at heating
to stabilizing effect of dissipative viscous force. The physical sense of this number is similar
to sense of the Rayleigh (Ra) [37] and Maringoni (M) [38] numbers:
Ra = ( g A h 4 ) ( a )
(2)
M = ( A h2 ) ( a)
(3)
As is possible to see from written down formulas (2) and (3), numbers Ra and M are
proportional to A in the first degree and therefore depends from direction of heating.
Excitation at heating from above or from free surface, with using of mechanisms
determining these numbers, is impossible. But Ed number does not depend from direction of
heating.
283
From the qualitative point of view, it can be understood so, that arising of the
temperature fluctuation on value of T1 , in the non-uniformly heated up LVL, induces the
thermo-electric field fluctuation TE 1 = T 1 and also fluctuation of the volumetric
charge, connected with this field ( e n 1 = T 1 - small fluctuation of the charge
density, connected with electric field and determined by dielectric permittivity of the LVL
). Influence of stationary, external thermo-electric field ( A ), caused by nonuniformity of heating, on the fluctuation electric charge creates the volumetric force
( e n 1 A = 2 A T 1 ). This volumetric thermo-electric force can induce instability
and movement in the LVL in case this force exceeds the dissipation forces
( a T 1 / A ). Besides the velocity cells, in the LVL arise the electric field cellular
structures. By other words, the electric field cellular structures are the original structures or
the centers of self-organizing at formation of the burning cells in the EM LVL. They are
enough steady structures.
Fig. 13 shows thin LVL in the EM reaction zone. Distribution of the forces interacting in
the LVL is shown. This is the buoyancy (lifting) force FL (the difference between
Archimedean force and the force of gravity), the surface tension force (thermo-capillary
force) F ST , and the thermo-electric (electrostatic) force FTE . For realization of specific
mechanism, the excitation force should exceeded the dissipation force F D in many times.
From the figure it is possible to see that at heating from above, the instability can be excited
Fig 13. Thin liquid-viscous layer in the EM reaction zone, having thermo-electric
properties at heating from above
284
only by thermo-electric force. During interaction of hydrodynamic, electric and thermal

subsystems in the open thermodynamic system - in the EM LVL arise the self-organizing
and appearance of macroscopic, existential structures. The burning surface is spontaneously
divided on the hexagonal hydrodynamic Benard cells, and we have result of this instability
as a CPB mode. Each of Benard cells have an own electric field. The cells will have
convective oscillations with intrinsic frequency. Thermo-electricity excites also the electric
field structures. In accordance with investigations executed in papers [39, 40] the boundaries of
the electric field structures coincide with boundaries of convection cells (the structures of
velocity). And on the LVL surface under influence of thermo-electric field is excited the
electric charge. Self-organizing (forming and evolution) of such SPMS occurs near with the
point of phase transition of the EM heterogeneous structure.
At the stationary burning mode the cells arise in the very thin superficial LVL and their
lifetime will be so small, that usually it is very difficult to register them. Obviously, in the
real conditions this mechanism will take place simultaneously with other physics-chemical
processes.
The process of thermo-electric convection excitation in the LVL can be described as
fllows. Certainly, this is one of possible variants of mathematical description of thermoelectric convection processes in the LVL. Taking into account usually used approaching of
the electric convection, it is possible to write down the following equation system (the
equation system are linearized on the fluctuation parameters, on the small perturbations,
index "1"):
The equations of movement (Navier-Stokes equations), which includes buoyancy
(lifting) force ( T 1 g ), Coulombs force (the electric force inducing thermo-electric
convection) ( e n 1 E 0 / 0 = e n 1 T 0 / 0 ) and dissipation by viscous forces:

e n 1 T0
P1
W =
T1 g +
0
0
(4)
where: P 1 T 1 g - the difference between Archimedean force and the force of gravity
0
which takes place because of the LVL thermal expansion, usually inducing Rayleigh
convection. [37]
In case LVL rotates around z axis with angular velocity :

t
P1
e n 1 T0
W + 2 W Z =
T1 g +
0
0
(5)
where 2 W Z - additional Coriolis force.

In accordance with paper [40] rotation suppresses thermo-electric convection and size of
the cell in cross-sectional direction decreases a little. In this case, a condition of excitation:
2
3
2
4
2 Te 2 8 h 3 2

Ed rot
=
4
2
4
285
(6)
The thermal conductivity equation where is neglected by Joule heat effect, but taken into
account the heat release distribution due to chemical reactions in the LVL and heat flux
issuing from the gaseous phase (flame) to the surface of the propellant:
q0

i
t a T1 = W T 0 + C ( 1 i ) K i ( T C )
T1
t
a T1 + ( V ) T 0 = 0 ,
(7)
(8)
The continuity equation of incompressible liquid:

div W = 0 ,
(9)
The electrostatics equation (Poissons equation):

div E 1 =
e n 1
(10)
The continuity equation of electric field:

div ( E 1 T 1 ) = 0 ,
(11)
rat E 1 = 0 .
(12)
In accordance with experimental data, the ratio of the longitudinal and cross-sectional
sizes of the elementary burning cell are the most stable characteristic of the thermo-electric
mechanism of the cellular movement excitation.
Description of heat transfer inside the solid propellant burning cell with number i
corresponds to the parabolic partial differential equation of thermal conductivity to which is
added a term related to the coordinate system that moves with the regressing surface:
T (Ci, j )
t
=u
( i, j )
C
T (Ci, j )
z
q
2 T (Ci, j )
C
+ 0 ( 1 ( i, j) ) K ( i, j ) ( T C( i, j ) )
+
2
C C C
CC
z
(13)
Equation of kinetics:
( i, j)
t
= u (Ci,j)
( i, j)
z
+ ( 1 ( i, j) ) K (i,j) ( T Ci,j) )
Factor of the chemical reaction rate inside the burning cell:
286
(14)
K ( i, j ) ( T
( i, j )
C
)=
Ea
k 0 exp
R T ( i,j )
0
C
0 , at T C( i,j )
, at T C( i,j ) > TR
TR
(15)
This implies that the reactive zone is very small in comparison with heated propellant
thickness. The high value of the activation energy justifies this assumption.
Boundary conditions:
z = 0:
- C
T (Ci, j )
z
( i, j )
( i, j )
( T 0 , h(hi, j ) , u (Ci, j ) , P CP )
= m p q 0 + Q 0 + IGN ; T S = f
(16)
The last condition takes into account also temperature variations at the burning wave
propagation.
( i, j )
Z : c T C = 0 ,
t = 0:
T = T0 ,
(17)
(i, j) = 0
= 0.
(18)
At formation of the cellular structure, body forces play a role (see Eqs. 4 and 5). This
implies that the cellular structure would be different depending on the attitude of the
propellant. I.e. the gravity vector can be directed towards the burning surface, is parallel to
the burning surface or can be directed away from the burning surface. However, the effect of
gravitation is taken into account here, as the additional factor, because this effect is not main
or determining in formation of these cellular structures. The body forces can be present and
can be absent. Existence of the body forces at heating from above stabilizes the LVL. Also
under (strong) acceleration the formation of the cellular surface and hence the frequencies of
oscillation (combustion instability) could be different and maybe just match or match not
with the acoustic frequencies of the combustion chamber. Such high accelerations would be
experienced in the flight of a rocket but could be reproduced in a centrifuge. V.N.Marshakov
carried out experimental researches only in stationary conditions. Unfortunately, the
experiments connected with formation of the cellular structure on the burning surface under
strong acceleration were not carried out. It is a theme for future investigations.
The thermo-electric instability, also, are the reason of appearance of the burning wave
thermal instability phenomenon. [4, 5, 7, 14] The unstable (vibrating) combustion arises at
achievement of amplitude of micro-oscillations in the burning zone of some threshold value,
that dependent on geometrical characteristics of acoustic cavity and from the propellant
properties. Suggested theoretical model allows give an explanation of the experimental
results obtained by V.N.Marshakov. The frequency of micro-oscillations on the EM burning
surface is determined by the parameters of self-organizing synergetic micro-structures on its
burning surface. Obviously that for each specific EM the considered phenomenon of selforganizing synergetic microstructures on the burning surface will have own peculiarities and
will appear in a various degree.
At the same time, thermo-electric field fluctuation phenomenon, which leads to
occurrence of a volumetric charge in the LVL of the burning EM and to excitation of the
thermo-electric convection, is very similar to the mechanism of occurrence of the electric
polarization phenomenon at the SHS-systems burning. In the papers [41, 42], on example of
wide class of the SHS-systems, is shown, that, simultaneously with propagation of the SHS
287
wave spontaneously arise the wave of electric polarization of the reactionary system. The
electric polarization may be of three types: formation of mainly positive or negative potential
of the burning wave or alternation of potentials of both marks. Excitation of the electromotive force at the SHS, authors of papers [41, 42] basically connect with the phenomenon of
thermal diffusion of the mobile ions inside the condensed phases along the temperature
gradient of the burning wave.
8.
SPATIAL-PERIODIC MICRO-STRUCTURES IN THE LIQUIDVISCOUS LAYER AND THE UNIVERSAL FREQUENCY CODE
OF THE ENERGETIC MATERIAL
Taking into account suggested mechanism of thermo-electric convection excitation in

the EM LVL, is possible introduce the conditional concept of the Universal Frequency
Code of the EM. This concept is connected with formation of the electric field structures
and the SPMS in the EM LVL. By other words, the Universal Frequency Code of specific
EM includes a set of the parameters determining the governing laws of development of the
thermo-electric convection and the SPMS in the LVL of this EM. Each specific EM already
has a built-in Universal Frequency Code that carries out the adaptation of the system to
external influence. If there is an external source of the entropy (for example, the heating),
then the standard condition of the medium are the SPMS. Within the scope of suggested
theoretical model, the integrated parameter of the propellant intrinsic frequencies of microoscillations [17, 18], which characterizes features of burning of each specific solid propellant,
can receive a new explanation. The propellant intrinsic frequencies of micro-oscillations is
a result of formation of individual SPMS on the burning surface of specific EM. Such SPMS
are the source of oscillatory process on the burning surface. At the EM burning, SPMS
radiate oscillating flows of the CP which interact with spatial-periodic vortex microstructures in the turbulent CP flow and with the acoustic cavities of the combustion chamber
and the solid propellant charge channel. It is possible to say, that the elementary burning cell
is the specific carrier of the EM Universal Frequency Code.
In accordance with experimental data, the ratio of the longitudinal and cross-sectional
sizes of the elementary burning cell are the most stable characteristic of thermo-electric
mechanism of the cellular movement excitation. Change of the boundary conditions
practically does not influence this parameter. At change (reduction) of the thickness of the
LVL occurs change of the basic mechanism of excitation of the cellular movement leading to
instantaneous change of the sizes of arising cells. Replacing of the mechanism of excitation
occurs as the phase transition of the second type.
At execution of thermocouple measurements was detected the temperature distributions
in the burning wave the spread of the surface temperature values TS . This spread is not
random, that is connected with non-perfect technique or errors of measurements. The spread
in the TS values is connected with a real existing set of the TS values appropriate to various
burning rates in pulsations. The spread of the TS value express a temperature range of
existence of the LVL. Variation of the mode of convective movement in the LVL, formation
and destruction of the cellular micro-structures influence on distribution of the temperature
field in the LVL. The convective flows and cellular movement in the LVL also determine
oscillations of T S on the burning surface. On the other hand, oscillatory character of change
of values of T S and R (t) is connected with periodic change of thickness of the LVL.
288
9.
SPATIAL-PERIODIC MICRO-STRUCTURES AND PROCESSES OF

EROSIVE BURNING AND AGGLOMERATION ON THE BURNING
SURFACE
Within thye scope of suggested concept, physical-chemical effects taking place at

erosive burning of the EM, receive a new explanation.
First of all, interconnection between gas-dynamics processes and process of the EM
burning is carried out through the phenomenon of the EM erosive burning. The basic and
universal gas-dynamics parameter determining the coefficient of erosion is tension of
friction on the burning surface.
At description of mechanism of the EM erosive burning usually are considered physicalchemical (and convective) processes in the gas phase (in the boundary layer) above the
burning surface. At the same time, processes of convection in the LVL of burning EM are
not considered. At intensive flowing around of the burning surface by the CP, the thickness
of the LVL will decrease. In these conditions, the process of thermo-electric convection and
the cellular movement will be activated. The intensive cellular movement will provide
increase of the heat flow into the EM, to the cold boundary of the LVL. At the same time,
increase in the heat supply from above to the LVL will make active the process of thermoelectric convection. In the result, will be observed increase of the EM burning rate. In
particular, the threshold (critical) value of velocity of the CP flow, below which there is no
erosive burning, obtains new interpretation connected with intensification of the thermoelectric convection and with amplification of cellular movement in the LVL. This threshold
values are determined by critical thickness of the LVL. After reduction of thickness of the
LVL below critical value, the phenomenon of thermo-electric convection becomes more
active and increases heat supply in the depth of EM.
The process of the SPMS formation provides influence on the process of agglomeration
and on the spectrum of agglomerates on the burning surface of heterogeneous EM.
For increase of energetic properties of solid rocket propellants, usually, in their
composition include pulverous metals, for example, the aluminum having high heatproducing capability of reaction of oxidizing. Unfortunately, the metal particles having an
initial size of 5-20 microns unite in the burning wave and create agglomerates with a size of
100-500 microns. This phenomenon induces a number of undesirable effects. In particular,
full combustion of agglomerates requires significant time, and they have not time to burn
away in the combustion chamber. Development of any methods providing reduction of the
sizes of agglomerates, leaving from the EM burning surface, finally, allows increase the
ballistic efficiency of the solid propulsion systems.
At burning of the heterogeneous EM, each of the cells of burning can be considered,
also, as the center of agglomeration. The process of the SPMS formation provides influence
on the process of agglomeration and on the spectrum of agglomerates on the burning surface
of heterogeneous EM. The structures of electric field and the structures of convection in the
LVL provides essential influence on formation of the agglomerates on the EM burning
surfaces. Accordingly, the technology of control of the electric field structures in the EM
LVL is a key to control of the agglomeration processes on the EM burning surface.
Changing the sizes, orientation and energy of structures of the electric field in the LVL it is
possible to control the sizes of agglomerates on the EM burning surface. Taking into account
intrinsic frequencies of each specific solid propellant is possible control the electric field
structures and provide essential reduction of agglomeration.
289
10.
CONTROLLING BY PROCESSES OF SELF-ORGANIZING IN THE

ENERGETIC MATERIAL LIQUID-VISCOUS LAYER
Controlling of self-organizing processes in the LVL and on the EM burning surface

opens possibilities for efficient control by processes of agglomeration and erosive burning
and also stabilize burning process.
One of perspective methods for control of formation of the SPMS on the burning surface
is inclusion of nano- and micro-particles of semi-conductor materials in the EM structure. In
this case, conditions of formation of the electric field cellular structures in the LVL will be
considerably changed. Change of parameters of thermo-electric convection will lead to
changes of process of self-organizing of the cellular micro-structures on the EM burning
surface. In the result, the spectrum of frequencies of micro-oscillations on the EM burning
surface will be corrected or displaced.
As an example of practical realization of this technology, it is possible to consider the
experimental data obtained in the Scientific-Research Institute of Polymeric Materials
(Perm, Russia). For suppression of combustion instability in the specific solid propulsion
system, the various kinds of stabilizers of burning were tested [43 45] : CaCO3, MgO, ZrO2,
ZrC, TiO2, TiC, NiN, TiO, TiN, SiO2, SiC and TiSi2. These experiments were executed for
case of coincidence of the "propellant intrinsic frequencies" and the acoustic frequencies of
the charge channel cavity. Practically in all fire tests, oscillations of the CP pressure and
combustion instability were observed. However, experiments have shown, that at use of the
titanium disilicide (TiSi2 ) the most effective suppression of the combustion instability in the
SPS is provided. In accordance with the patents [44, 45] for stabilization of combustion process
of the charge it was suggested insert the titanium disilicide into the structure of the
propellant in quantity from 1 up to 5 % from the charge mass. In this case, the ratio of length
of cylindrical section of the charge channel to its diameter should be in the limits from 6 up
to 28.
As explanation of reasons of effective suppression of combustion instability it was
supposed that on the propellant burning surface occurs reaction of decomposition of the
titanium disilicide. [43, 44] In particular, it was supposed that as a result of this reaction, two
other stabilizers of burning will be formed [43]:
Ti Si 2 + 3 O 2 Ti O 2 + Si O 2
(19)
In this thesis [43] also it was noted that at use of the titanium disilicide, non-simultaneous
burn-out of the propellant components will change. In result, the spectrum of frequencies of
micro-fluctuations, radiating from the burning surface should be changed. However, taking
into account physical and chemical properties of the titanium disilicide (Table 1) it is very
difficult to agree with such explanations. First of all, melting point of the titanium disilicide
(2033 K) is much higher, than temperature in the propellant LVL and on the burning surface.
And chemical reaction, mentioned above, can not occur on the burning surface. In the result,
the real mechanism of suppression of combustion instability at use of the titanium disilicide
was not understood. And in the thesis [43] the incorrect explanation of this phenomenon was
given. In accordance with suggested universal concept of the SPMS excitation at the EM
combustion, these experimental data obtain a new explanation. Titanium disilicide (TiSi2) is
a material used in the semiconductor industry. It has the lowest resistivity of all refractory
290
Table 1. Physics-chemical properties of stabilizers of burning.

Chemical Name
Chemical Formula
Density, kg m 3
Melting Point, K
Titanium Disilicide
(size of powder: 3-13
micron)
TiSi2
4390
2033
Titanium Dioxide
TiO2
4260
2103 - 2123
Silicon
Dioxide
SiO2
2200
1986
metal silicides and good thermal stability. As it was already indicated, the temperature of
melting of titanium disilicide essentially exceeds temperature of melting of propellant. At the
process of burning of propellant, in the propellant LVL and on the burning surface,
mentioned micro-particles of titanium disilicide will remain in the solid phase. Accordingly,
presence of such micro-particles in the melt of solid propellant (LVL) will change dielectric
properties of the melt (dielectric permittivity of the LVL ). In this case the conditions of
development of thermo-electric convection in the LVL will be considerably changed.
Change of parameters of thermo-electric convection will lead to changes of the process of
self-organizing of the cellular micro-structures on the propellant burning surface. In the
result, the spectrum of frequencies of micro-oscillations on the propellant burning surface
will be corrected or displaced. And will be provided mismatching of the spectrum of
frequencies of micro-oscillations on the propellant burning surface and the acoustic
frequencies of the charge channel cavity.
Moreover, high-melting micro-particles of titanium disilicide in the LVL are the centers
of agglomeration. Increase of agglomeration on the burning surfaces and increase of quantity
of agglomerates in the CP will lead to suppression of oscillatory processes directly on the
burning surface and in the CP flowing from the burning surface. As LVL contains a large
amount of agglomerates, the coefficient of thermal conductivity of this LVL at the burning
surface is higher than that of the propellant without additive. As a result, additional heat
release appears near the burning surface. This heat release accelerates reactions in both CP
and gas phase near the burning surface, and as a consequence the burning rate increases.
Also, for control of formation of the SPMS in the LVL and on the burning surface, in the
structure of solid propellant can be included high-melting nano- and micro-particles of
ferromagnetic materials. Controlled change of the physical-chemical properties of the LVL
will lead to change of basic parameters of the process of self-organizing of the SPMS on the
burning surface and to change of the spectrum of frequencies of the micro-fluctuations on
the burning surface. In this case also will be changed the mechanism of development of the
CPB. Controlled increase of viscosity of the LVL will allow to change or to decelerate the
process of formation of the burning cells, to reduce diffusion processes in the LVL and to
reduce the frequency of micro-pulsations on the burning surfaces of the EM.
Also, there is a possible variant when provided preliminary electrostatic or ionization
charging of nano- and micro-particles of semi-conductor materials. Such charging of the
nano- and micro-particles can be carried out in the field of electric corona discharge, due to
flight of this particles through the electric corona discharge. Degree of charging of nano- and
micro-particles increases with growth of current and voltage of the electric corona discharge.
291
11.
SOLID PROPELLANT CHARGES OF NEW GENERATION

WITH ACTIVE SELF-ADAPTING MICRO-STRUCTURES
On the basis of suggested new concept suitable design of the solid propellant charge is
developed. The design of this charge has active, self-adapting SPMS and provides stable
mode of burning of the EM.
The basic idea used in this solid propellant charge design application of special
supporting elements, manufactured from active high-porous penetrable cellular EM with
self-adaptive spatial-periodic micro-structure.
Such charge can be manufactured as the grain containing a set of sections. Each of
sections can be manufactured as a spatial hollow gas-permeable skeleton (latticework) from
active, high-porous permeable cellular EM with self-adaptive spatial-periodic microstructure (Fig. 14). As active high-porous EM can be used both gas-free SHS mixtures
(SHS-systems), and hydrogen-generating SHS mixtures (SHS-systems). As exothermic
hydrogen-generating mixture can be used mechanical mixture of powders of the solid fuel carbide - or nitride-forming metals of IV - V groups of D.I.Mendelyeev's Periodic System
and solid oxidizer - hydrogen-containing compounds of carbon or nitrogen. Spatial cavities
of the supporting skeleton are filled by some energetic material or solid rocket propellant. In
particular, such skeleton can be formed from active elements with a spatial-periodic microstructure (Fig. 14) which are connected with each other in the certain sequence. Specified
Fig 14. Variants of supporting cellular structures in the solid propellant charge design
(A cross-section of volumetric cell of the high-porous cellular material;
B active element (solid flame) with a spatial-periodic micro-structure).
spatial skeleton also execute function of accompanying of burning. For the charges having
such design, there will not be restrictions for modes of steady burning of the EM determined
in accordance with the theory of abnormal burning, suggested by Ya. B. Zel'dovich. [2]
Such design of the charge allows to suppress pressure peaks and various kinds of
abnormal and unstable burning in the combustion chamber. Self-adaptive spatial-periodic
micro-structures will stabilize oscillatory processes on the charge burning surface. Suggested
design of the charge allows to increase energy characteristics of propulsion system due to
mechanism of heat transfer from the high-temperature zone of burning to the propellant
burning surface and in the LVL. In this case, propellant burning rate will be increased. At
flowing of CP through the spatial-periodic micro-structure of supporting material there will
be afterburning of them. Moreover such charge will be less sensitive to external loadings.
292
Changing the sizes and the shape of cells and also heat conductivity of supporting
material it is possible to program a mode of burning of the energetic material in the
combustion chamber. Also programming a spectrum of frequencies of oscillations, radiated
from the charge burning surface can be organized.
Use of suggested design of the charge allows to control and to reduce the process of
agglomeration in the combustion chamber.
Such design of the charge opens a unique opportunity for realization of new
technologies of deep-control of thrust of the SPS. Additionally, suggested design of the
charge allow provide required completeness of combustion of components of solid
propellant, reduce amount of harmful CP and provide cleaner combustion.
12.
CONCLUSION
The mechanisms of the EM burning comply of the general universal concept: if there is
an external source of the entropy (for example, the heating), then the standard condition of
the medium are the SPMS. Actually, mechanism of the SPMS excitation at the EM
combustion is one of the fundamental laws determining the burning processes of the EM. On
the solid propellant burning surface there is no chaotic combustion, but on the contrary, takes
place the order the process of self-organizing of dissipative hydrodynamic microstructures. This phenomenon is similar to the phenomenon of hydrodynamic instability at
transition of laminar flow into the turbulent flow, connected with formation of dynamic
dissipative spatial-periodic vortex micro-structures.
Suggested theoretical scheme changes our representations about the EM burning
processes and opens new opportunities both for controlling of the EM burning processes and
for effective prevention of the unstable combustion and burning abnormal modes in the
practical systems. In particular, controlling of the self-organizing processes on the EM
burning surface opens possibilities for change a range of acoustic frequencies, radiating from
the burning surface. Technology of control of the electric field structures in the EM LVL is a
key to control of the agglomeration processes on the EM burning surface. Suggested new
concept is baseline for creation of the methodology of control by the physics-chemical
properties of the LVL and by processes of self-organizing on the EM burning surface.
Application of suggested concept and technologies in practical systems will allow
control the development of physical-chemical processes at a new qualitative level.
Suggested concept can find application at development of some practical systems:
-
Nano-technologies for safe and ecology-friendly utilization (recycling) of the largesized solid propulsion systems;
Solid propellant charges of a new generation with active, self-adapting microstructures on the burning surface;
Technologies for control of processes of erosive burning and agglomeration at the
solid propellants burning in the combustion chamber;
Systems for suppression of unstable and abnormal modes of burning of the solid
propellants in the solid propulsion systems combustion chambers;
Systems for support of combustion processes in the solid propulsion systems
combustion chambers;
Solid propellant gas generators with programmed frequency of the CP pressure
oscillations for increase of production of the old oil wells;
Active devices for control of hydro-meteorological processes;
293
New technologies of deep-control of thrust of the solid propulsion systems that can
be used, for example, as a propulsion systems of the intercontinental ballistic rocket
upper stages (of the multiple independently targetable reentry vehicle (MIRVed)
warheads).
Acknowledgments:
Author would like express deepest gratitude for the experimental data provided by Prof.
Vladimir N. Marshakov (Chief of the Group of Non-Stationary Combustion of the
N.N.Semenov Institute of Chemical Physics of the Russian Academy of Sciences, Moscow,
Russia), by Prof. Nickolay M. Pivkin (Chief of the Department No. 012 of the ScientificResearch Institute of Polymeric Materials, Perm, Russia) and by Prof. Oleg P.
Korobeinichev (Head of Laboratory of Kinetics of Combustion Processes, Institute of
Chemical Kinetics and Combustion, Siberian Branch of the Russian Academy of Sciences,
Novosibirsk, Russia).
Nomenclature
A
a
C
C0
e
Ed
E
Ea
E1
E TE
E TE 1
FD
FL
F ST
F TE
g
h
hC
h M, E
- / T 0 / = ( T h T c ) / h - Equilibrium gradient of temperature (variation of

temperature T per unit of thickness), degree of heat, heating up;
- Thermal diffusivity of the liquid-viscous layer;
- Specific heat capacity of the condensed phase;
- Density of luminous flux of laser radiation;
- Electric charge of carrier;
- ( 2 A 2 h 2 ) ( a ) - Dimensionless Eidelman (thermo-electric)
number, the ratio of the destabilizing effect of thermo-electric (electrostatic)
force to the stabilizing effect of viscous force [36, 39];
- Electric field intensity;
activation energy of the condensed phase reactions;
- ( E E 0 ) - Fluctuation of the electric (thermo-electric) field intensity;
- T 0 - Thermo-electric field of equilibrium state;
- T 1 - Thermo-electric field fluctuation;
- a - Dissipation, viscous force;
- g A h 4 - Buoyancy (lifting) force (the difference between
Archimedean force and the force of gravity), (Rayleigh, 1916);
- A h 2 - Surface tension force (thermo-capillary force), (G.K.A. Pearson,
1958), (the convection cells arising under influence of this mechanism is
known as Maringoni cells);
- 2 A 2 h 2 - Thermo-electric (electrostatic) force, Coulombs force,
(I.V.Ioffe and E.D.Eidelman, 1976);
- Acceleration due to gravity;
- Characteristic size of fluctuation movement (the thickness of the liquidviscous layer, cross-sectional size of the cell);
- Critical thickness of the liquid-viscous layer;
- h M, E = h C - Critical thickness of the liquid-viscous layer, at which the action
294
h Ra,E
hh
of the thermo- capillary (surface tension) mechanism is equivalent to action of

the thermo-electric mechanism;
- Critical thickness of the liquid-viscous layer, at which the action of the
buoyancy (lifting) force is equivalent to action of the thermo-electric force;
- a / u c - Characteristic thickness of the EM heated up layer at burning;
- Characteristic thickness of the EM heated up layer at burning, obtained from

the experiments;
- Width of the thermal channel, that burnt out by the transverse burning wave;
- Pre-exponential factor of chemical reaction rate;
- Factor of the chemical reaction rate inside the burning cell;
- (10 15) h h - Characteristic size of the burning cell, longitudinal size of
LS
the burning cell;

- Distance from the burning surface;
exp
h
H
k0
K ( i, j ) ( T C( i, j ) )
LB
mp
M*
n1
n1
P
P0
P1
q0
Q0
Q T1
R
R (t)
Ra
Ra *
t
T
T0
T1
- Characteristic size of the Benard cell in the liquid-viscous layer (for ideal
conditions);
- Propellant mass burning rate;
- FST / F D = ( A h 2 ) ( a ) - Maringoni number, the ratio of the
destabilizing effect of the surface
tension (thermo-capillary) force to the
stabilizing effect of viscous force;
- Critical value of Maringoni number;
- Number of influencing impulses in the series of laser radiation;
- T 1 - Concentration of the electric charge (e) carriers;
- ( n n 0 ) Fluctuation of concentration of the electric charge (e) carriers;
- Pressure;
- Pressure equilibrium value;
- ( P P0 ) - Pressure fluctuation (deviation of the pressure from the
equilibrium value P0 );
- Heat release distribution due to chemical reactions in the solid phase and
liquid-viscous layer of the cell;
- Heat flux issuing from the gaseous phase (flame) to the surface of the
burning cell;
- a T1 h 2 - The heat flow arising because of temperature fluctuation;
- Universal gas constant;
- Electric conductivity of the solid propellant burning surface;
- ( g A h 4 ) ( a ) - Rayleigh number, the ratio of the destabilizing
effect of buoyancy force to the stabilizing effect of viscous force (criterion of
convection excitation);
- Critical value of Rayleigh number;
- Time of the process;
- Temperature;
- Temperature equilibrium value;
- ( T T 0 ) Temperature fluctuation (deviation of the temperature from the
equilibrium
TC
value
T0
);
- Temperature of the cold boundary surface of the liquid-viscous layer;
295
Te
Th
TS
T S*
Ts
u*
u (Ci,j)
U
W
w
x
z
4 2 h 4
- Dimensionless Taylors number;
- Temperature of the hot boundary surface of the liquid-viscous layer (on the
interface of liquid-viscous layer and gasification zone);
- Temperature on the burning surface;
- Temperature on the gunpowder surface in the intervals between the
transverse burning waves;
- Pulsation of the temperature on the energetic material burning surface;
- Propagation velocity of the transverse burning wave;
- Linear burning rate of the energetic material sample in the cell with number
i;
- Linear burning rate of the energetic material sample;
- Flow velocity of the liquid-viscous layer;
- w = 4 h 2 L 2 - Parameter describing the symmetry of arising cell in the
longitudinal and cross directions (ratio of sizes of the convection cell);
- Distance between the transverse burning waves;
- Space coordinate, perpendicular to the liquid-viscous layer;
Greek Symbols
( i, j )
IGN
0
1
- ( 1 ) T - Volumetric coefficient of the thermal expansion of the

liquid;
- Degree of chemical conversion in the cell;
- Nabla operator;
- Coefficient of the thermo-electro-motive force. Coefficient shows efficiency of
transformation of the heating in the electric field;
- Dielectric constant, factor of dielectric permittivity of the liquid-viscous layer;
- Wave length of the laser radiation;
- conductivity;
- Dynamic viscosity coefficient;
- - Kinematic viscosity coefficient of the liquid-viscous layer;
- Ignition heat flux;
- Density of the liquid-viscous layer, density of the condensed phase;
- Density equilibrium value;
- ( 0 ) Density fluctuation ( deviation of the density from the equilibrium
value 0 );
- Temperature coefficient of the superficial tension, coefficient that shows
the decrease of the surface tension with the temperature increase;
- Duration of impulse of the laser radiation;
- Dimensionless time;
- Wave period;
- Angular velocity of the liquid-viscous layer rotation (relatively of space

coordinate z, directed perpendicularly to the liquid-viscous layer);
296
Subscripts and Superscripts

c
( i, j )
g
s
0
1
*
- condensed phase;
- coordinates of the burning cell on the burning surface perpendicular z axis;
- gas phase;
- surface;
- parameters of equilibrium state;
- parameters in fluctuation conditions;
- critical value of parameter.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
YA. B. ZEL'DOVICH: To the Theory of Combustion of the Gun-powders and Explosives, J. of

Experimental and Theoretical Physics, vol. 12, No. 11-12, p. 498-510, 1942, (in Russian)
YA. B. ZEL'DOVICH: About Stability of Burning of Gun-powder in the Semi-Closed Cavity, J.
of Appl. Mech. and Tech. Phys., No. 1 , p. 67-76, 1963, (in Russian)
K. I. SYNAIYEV: The Cellular-Pulsating Burning of the Nitrocellulose Solid Propellants, Book
of Abstracts, The 1st All-Union Symposium on Combustion and Explosion, Publishing House
Science, Moscow, p. 59-63, 1968, (in Russian)
A. G. ISTRATOV and V. N. MARSHAKOV: The Burning Wave Instability, The Concept of
Development of Combustion and Explosion as a Field of Scientific-Technical Progress,
Publishing House Territoriya, Chernogolovka, p. 37-38, 2001, (in Russian)
A. V. ANANIYEV, A. G. ISTRATOV, Z. V. KIRSANOVA, V. N. MARSHAKOV, and G. V. MELIKGAIYKAZOV: Instability at the Steady-State Burning of the Gun-powders and Explosives, J.
Chemical Physics, vol. 20, No. 12, p. 47-52, 2001, (in Russian)
V. N. MARSHAKOV: Parameters of the Cellular-Pulsating Burning of the Nitroglycerin
Gunpowder, J. Chemical Physics, vol. 6, No. 4, p. 530-537, 1987, (in Russian)
V. N. MARSHAKOV, A. G. ISTRATOV and V. M. PUCHKOV: Non One-Dimensional Front of
Burning of the Propellant Compositions on the Basis of Nitrocellulose and Nitroglycerin, J.
Physics of Combustion and Explosion, vol. 39, No. 4, p. 100-106, 2003, (in Russian)
G. V. MELIK-GAIYKAZOV, V. N. MARSHAKOV: Investigation of the Electric Conductivity of
the Ammonium Perchlorate at Burning, J. Physics of Combustion and Explosion, vol. 23, No.
4, p. 15-20, 1987, (in Russian)
V. A. ARKHIPOV, S. S. BONDARCHUK and A. B. VOROZHTSOV: Experimental Study of Solid
Propellant Nonsteady Burning Rate under Rapid Depressurization, Novel Energetic Materials
and Applications, 9-IWCP Book of Abstracts, The Ninth International Workshop on
Combustion and Propulsion, Lerici, La Spezia, Italy, September 14-18, 2003, Politecnico di
Milano, SP Lab, Milan, MI, Italy, p. 168, 2003
O. P. KOROBEINICHEV, A. A. PALETSKY, A. G. TERESCHENKO and E. N. VOLKOV:
Combustion of Ammonium Dinitramide/Polycaprolactone Propellants, Proceedings of the
Combustion Institute, vol.29, p.2955-2961, 2002
O. P. KOROBEINICHEV, E. N. VOLKOV, A. A. PALETSKY, T. A. BOLSHOVA and A. G.
TERESCHENKO: Environmentally Friendly ADN-based Solid Rocket Propellant with Good
Ballistic Characteristics, Rocket Propulsion: Present and Future, edited by L.T. DeLuca,
Politecnico di Milano, 2003, Pozzuoli, Naples, Italy, p. (28-1)- (28-16), 2002
O. P. KOROBEINICHEV, A. A. PALETSKY, A. G. TERESCHENKO and E. N. VOLKOV: Study of
Combustion Characteristics of Ammonium Dinitramide/Polycaprolactone Propellants, Journal
of Propulsion and Power, vol.19, No. 2, p. 203-212, 2003
S. G. YARUSHIN: Investigation of Basic Schemes, Properties and Characteristics of
Controllable Solid Propulsion Systems on the Propellants with Specific Properties, A pre-print,
Perm: Institute of Continuous Mechanics UB of the RAS, 80 p., 1997, (in Russian)
297
[14] V. N. MARSHAKOV and A. G. ISTRATOV: Wave Structure of Solid Propellants Burning Front,
Proceedings of International Conference on Combustion and Detonation: Zel'dovich Memorial
II, (August 30 September 3, 2004), Moscow, Russia, p. W2-2.1-W2-2.11, 2004
[15] YU. S. IVASHENKO, A. L. SADYRIN, and V. L. PAVLENKO: Investigation of the Electric
Conductivity of the Ammonium Perchlorate Burning Surface, J. Physics of Combustion and
Explosion, vol. 22, No. 4, p. 30-33, 1986, (in Russian)
[16] G. V. MELIK-GAIYKAZOV: Experimental Investigations of the Gunpowder Non-Stationary
Burning Modes in the Semi-Closed Cavity, Ph.D. Thesis (Phys.-Math. Sci., Spec. 01.04.17
Chem. Phys., incl. Physics of Combustion and Explosion), N.N.Semionov Inst. of Chem. Phys.
RAS, Moscow, p. 147, 1998, (in Russian)
[17] N. M. PIVKIN and N. M. PELYKH: The High-Frequency Instability of Combustion in Solid
Rocket Motor, J. of Propulsion and Power, vol. 11, No. 4, p. 651-656, 1995
[18] N. M. PIVKIN: Non-Simultaneous Burnout of Energetic Material Components, Theory and
Experiment, Propellants, Explosives, Rockets, and Guns (Proceedings of the Second
International High Energy Materials Conference and Exhibit), (December 8-10, 1998, IIT
Madras, Chennai, India), Ed. S.Krishnan, S.K. Athithan., Allied Publishers Limited, India, p.
106-110, 1998
[19] D. I. MENDELEYEV: The Collected Works, Publishing House of the USSR Academy of
Sciences, vol. 9, Moscow, p. 43-45, 1949, (in Russian)
[20] T. P. IVLEVA and A. G. MERZHANOV: Three-Dimensional Spinning Waves in the Case of GasFree Combustion, Doklady Physics, vol. 45, No. 4, p. 136141, 2000, (Translated from
Doklady Akademii Nauk, vol. 371, No. 6, p. 753758, 2000)
[21] A. G. ISTRATOV and V. N. MARSHAKOV: Critical Diameter of Burning of Gun Powders and
Solid Explosives, Proceedings of the XIII Symposium on Combustion and Explosion
(Chernogolovka, February 7-11, 2005), Publishing House ISMAN, Chernogolovka, p. (46-1)
(46-11), 2005, (in Russian)
[22] A. G. MERZHANOV: Solid-Flame Combustion, Publishing House of the ISMAN,
Chernogolovka, p. 224, 2000, (in Russian)
[23] G. V. MELIK-GAIYKAZOV: Critical Phenomena at the Condensed Substances Combustion, J.
Physics of Combustion and Explosion, vol. 29, No. 1, p. 3-9, 1993, (in Russian)
[24] B. V. NOVOZHILOV: Non-Stationary Combustion of the Solid Rocket Propellants, Publishing
House "Nauka", Moscow, p. 176, 1973, (in Russian)
[25] B. V. NOVOZHILOV: Propellant Combustion From Steady State up to Chaos, Novel Energetic
Materials and Applications, 9-IWCP Book of Abstracts, The Ninth International Workshop on
Combustion and Propulsion, Lerici, La Spezia, Italy, September 14-18, 2003, Politecnico di
Milano, SP Lab, Milan, MI, Italy, p. 126-127, 2003
[26] J. KEPLER: The Six-Cornered Snowflake, 1611, translated by L.L.,Whyte, (Oxford Univ. Press),
1966
[27] H. BENARD: Les Tourbillons Cellulaires Dans Une Nappe Liquide Transportant de la Chaleur
Par Convection en Regime Permanent, Revue Generale des Sciences Pures et Appliquees, p.
11:12611271, 13091328, 1900
[28] B. P. ZHUKOV: Research and Development of the New Types of Ballistite Rocket Gun-powders,
Dr. of Sci. Thesis (Tech.), Scientific-Research Institute 125 (Scientific Research ChemicalTechnological Institute), Moscow, p. 224, 1951, (in Russian)
[29] Laser Beam Effects on the Materials, YE. P. VELIKHOV, Eds., Moscow, Publishing House
Science, p. 367, 1989, (in Russian)
[30] J. F. YOUNG, J. S. PRESTON, H. M. DRIEL VAN and J. E. SIPE: Laser Induced Periodic Surface
Structures, J. Phys. Rev. B., vol. B27, No. 2, p. 1141-1172, 1983
298
[31] YU. N. ORLOV: Dynamics of Formation of the Periodic Structures at Interaction of the Laser
Radiation with the Surface of Solids, Authors Abstract of the Ph.D. Thesis (Phys.-Math. Sci.),
L.D. Landau Inst. of Theor. Physics, USSR Acad. of Sci., Chernogolovka, p. 12, 1987, (in
Russian)
[32] I. URSU, I. N. MIHAILESCU, A. M. PROKHOROV ET. AL.: On the Role of Periodical Surface
Structures Induced by Powerful Laser Irradiation on Metallic Surfaces in the Energy Coupling
Process, J. Physica, vol. 132C, No. 3, p. 396-402, 1985
[33] Z. GUO SHENG, P. M. FAUCHET and A. E. SIEGMAN: Growth of Spontaneous Periodic
Structures on Solids During Laser Illumination, J. Phys. Rev., B., vol. B26, No. 10, p. 53665381, 1982
[34] G. M. GANDELMAN, and P. S. KONDRATENKO: To the Theory of Formation of Periodic
Structures Under Influence of the Laser Radiation on the Surface of Metals and SemiConductors, J. of Experimental and Theoretical Physics, vol. 88, No. 4, p. 1470-1480, 1985, (in
Russian)
[35] YE. B. LEVCHENKO, and A. L. CHERNIYAKOV: Instability of Superficial Waves in the NonUniformly Heated up Liquid, Journal of Experimental and Theoretical Physics, vol. 81, issue 1
(7), p. 202-209, 1981, (in Russian)
[36] I. V. IOFFE, N. V. KALININ and E. D. EIDELMAN: Possibility of Pre-Rayleigh Convection in the
Liquid Semi-Conductors, Letters in Journal of Technical Physics, vol. 2, issue 9, p. 395-396,
1976, (in Russian)
[37] R. S. CHANDRASEKHAR: Hydrodynamic and Hydromagnetic Stability, Clarendon Press,
Oxford, 654 p., 1961, (reprinted by Dover Publications, Inc., 1981)
[38] J. R. A. PEARSON: On Convection Cells Induced by Surface Tension, J. Fluid Mech., v. 4, p.
489-500, 1958
[39] E. D. EIDELMAN: Convection Under Influence of Thermo-Electric Field in the Liquid SemiConductors, Journal of Experimental and Theoretical Physics, vol. 103, issue 5, p. 1633-1643,
1993, (in Russian)
[40] E. D. EIDELMAN: Amplitudes of Convective Values at Sub-Rayleigh Convection in the Liquid
Semi-Conductors, Journal of Technical Physics, vol. 63, issue 10, p. 192-195, 1993, (in
Russian)
[41] YU. G. MOROZOV, M. V. KUZNETSOV and A. G. MERZHANOV: Electric Fields in the
Processes of Self-Propagating High-Temperature Synthesis, Int. J. of Self-Propag. High-Temp.
Synth., vol. 6, No. 1, p. 1-13, 1997
[42] YU. M. MAKSIMOV, A. I. KIRDIYASHKIN, V. K. SMOLIYAKOV and V. I. ITIN: The Features of
SHS Reactions in the Electric and Magnetic Fields, Self-Propagating High-Temperature
Synthesis: the Theory and Practice, Publishing House "Territoriya", Chernogolovka, p. 215235, 2001, (in Russian)
[43] N. M. PELYKH: Non-Stationary Burning of the Solid Propellant Charges and Use of This Effect
in the National Economy, Dr. of Sci. Thesis (Tech., Spec. 01.02.05 Mech. of the Liquid, Gas
and Plasma and 01.04.17 Chem. Phys., incl. Physics of Combustion and Explosion),
Scientific-Research Institute of Polymeric Materials, Perm, 244 p., 2003, (in Russian)
[44] N. M. PELYKH, N. G. IBRAGIMOV, A. P. TALALAIYEV and L. L. HIMENKO: Method of
Stabilization of Burning of Solid Propellant, Pat. 2 194 689 of Russia, C 06 D 5/06,
Application 2000121109/02, Priority from August 04, 2000, Registered at August 04, 2000,
The Inventions Bulletin (Russia) No. 31, December 20, 2002, (in Russian)
[45] G. N. AMARANTOV ET AL: The Charge of the Non-Metallized Solid Propellant, Pat. 2 202
096 of Russia, F 42 B 1/04, E 21 B 43/26, Application 2001105439/02, Priority from
February 26, 2001, Registered at February 26, 2001, The Inventions Bulletin (Russia) No. 10,
April 04, 2003, (in Russian)
299
DECOMPOSITION KINETICS OF GAP BINDER

IN THE PRESENCE OF AN ENERGETIC COMPONENT
Sreekumar Pisharath and Ang How Ghee*1
Energetic Materials Research Centre, School of Materials Science &Engineering,
Nanyang Technological University, 50, Nanyang Avenue, Singapore 639 798
Abstract:
Glycidyl Azide Polymers (GAP) are promising candidates as energetic binders for
future solid composite propellants. They produce minimum smoke, cause reduced
pollution, and have low sensitivity. In our hands, we have undertaken a study of the
decomposition kinetics of energetic binder in the presence of the energetic oxidizer, to
assess the stability of the composite formulation.
We now report the decomposition kinetics of GAP binder in the presence of two high
performance and environment friendly oxidizers like Ammonium Dinitramide (ADN)
and 4,10-dinitro-2,6,8,12-4,10-diazatetracyclo-[5.5.0.05,9 03,11]dodecane (TEX). Thermal
gravimetric analyses (TGA) and differential scanning calorimetry (DSC) were used to
investigate the decomposition characteristics and heat of decomposition of TEX/GAP
and ADN/GAP gum formulations. Addition of GAP has improved the thermal stabilities
of both AND and TEX. GAP has been found to be more vulnerable to thermal
decomposition in the presence of TEX than in the presence of ADN.
Keywords:
1.
glycidyl azide polymer, thermal stability, thermal decomposition,

activation energy
INTRODUCTION
Nitramine based energetic materials such as RDX and HMX are widely used as
propellant ingredients for increased specific impulse at low flame temperatures. However,
nitramines have low burning rates coupled with a relatively high and undesirable burning
rate exponent. Therefore research efforts were directed to employ nitramine explosives as
fillers in oxidizer/binder matrix propellants.
Glycidyl Azide Polymers (GAP) are promising candidates as energetic binders for
rocket propellant applications. Owing to its highly exothermic decomposition, high burning
rates are usually observed during the combustion of GAP. Also, its fuel rich composition
results in low flame temperature. Composite propellants of nitramines and energetic azido
polymers are characterized by low flame temperature, high detonation velocity, low
vulnerability and high thermal stability [1].
In this paper, we report the study of thermal decomposition characteristics of two
nitramine/GAP pseudo propellants using thermo gravimetric analysis (TGA) and differential
scanning calorimetry (DSC). The nitramines used for this study are Ammonium Dinitramide
(ADN) [2] and 4,10-dinitro-2,6,8,12-4,10-diazatetracyclo-[5.5.0.05,9 03,11]dodecane
(TEX) [3]. ADN and TEX are two high performance environment friendly oxidizers. ADN
is a potential halogen free replacement to ammonium perchlorate (AP) in solid rocket
1
Corresponding author: E mail: hgang@ntu.edu.sg
300
propellants. TEX is envisioned as a prospective oxidizer component in nitramine based

composite propellant formulations.
It is important to understand the thermal decomposition processes that occur in the
condensed phase of composite propellants because; these processes control the burn rate of
the propellant. Moreover, an understanding of the underlying molecular processes behind the
thermal decomposition behavior of propellants will provide insight into their relative
stability. This information will in turn provide guidance into the development of more stable
composite propellants.
2.
EXPERIMENTAL
GAP, ADN and TEX were synthesized in our laboratory by adopting standard
procedures. All the synthesized compounds were characterized using GPC, NMR, and IR
analysis. Requisite amounts of the compounds were hand mixed in a vacuum chamber at
room temperature and the samples were degassed for 1hr prior to testing.
Thermo gravimetric analyses were carried out using a Shimadzu Hires TG-50 equipped
with thermal analysis software for data analysis.
Differential scanning calorimetry measurements were carried out in TA DSC 2010
equipment containing thermal analysis software for data interpretation. In representative
runs, 2-3 mg of samples in sealed hermetic pans were ramped from room temperature to
3200C at varying rates of 20C/min, 50C/min, 100C/min and 150C/min in a steady flow of
nitrogen (50ml/min).
3.

3.1
Thermogravimetric Analysis
TGA curves of ADN, GAP and ADN+ GAP compounds are presented in figure 1.
100
ADN+GAP
GAP
ADN
Weight (%)
80
0
250.3 C(40.9%)
60
390.8 C(60%)
40
0
197.1 C(76.1%)
20
559 C(81%)
0
732.5 C(95%)
247.8 C(86.2%)
202 C(95.0%)
200
400
600
800
Temperature( C)
Fig 1.
TGA curves of GAP,ADN and ADN +GAP formulations
GAP decomposes in multiple stages. The predominant weight loss of 37% between
197 C and 2500C is due to the decomposition of azide groups, which is the common
decomposition mechanism associated with the azido polymers [4]. After 2500 C, GAP exhibits
a gradual weight loss until 7320C due to the decomposition of the polymer back bone.
0
301
ADN shows a single step decomposition behavior with 91% weight loss between 1350C
and 2020C. The major decomposition mechanism associated with ADN is its breakdown to
various oxides of nitrogen (NO2, N2O and NO), ammonia, ammonium nitrate and nitrogen
[5]
. ADN+GAP show multi step decomposition with a major weight loss of 75% between
1400C and 1970C owing to ADN decomposition, followed by a 10% weight loss until
247.80C due to GAP decomposition. TGA curve of ADN+GAP lies above that of ADN
indicating that thermal stability of ADN has improved with the mixing of GAP.
Figure 2 shows the TGA curves of TEX, GAP and TEX+GAP formulations.
100
GAP
TEX
GAP+TEX
Weight(%)
80
223.51 C
60
268.95 C(43.9%)
40
0
272.12 C(71.24%)
20
0
254.5 C(80%)
0
100
200
300
400
500
600
700
800
Temperature( C)
Fig 2.
TGA curves of GAP,TEX and TEX+GAP formulations
TGA curve of TEX shows a single step decomposition behavior of 80% weight loss
between 1800C and 2540C associated with the scission of N-NO2 bonds in the molecule, the
primary decomposition mechanism in nitramines [6]. TEX+ GAP compound also shows a
single step decomposition of 80%weight loss between 1800C and 2540C, followed by a
gradual weight loss associated with GAP component in the formulation. It is noteworthy that
TEX and TEX+GAP compounds exhibit a higher thermal stability than GAP until 2230C.
After 2500C, thermal stability of TEX+GAP is higher than that of TEX alone. Clearly
addition of GAP has improved the thermal stability of TEX.
Comparing figures 1&2, in ADN + GAP and TEX+GAP pseudo propellant formulations
the major mass loss step primarily relates to that of decomposition of ADN and TEX
followed by the degradation of GAP binder. Thermal stabilities of ADN + GAP & TEX +
GAP formulations have increased with respect to ADN and TEX alone. ADN + GAP
propellant formulation shows a gentle slope of variation of mass loss with temperature when
compared to that of ADN. On the other hand, for TEX+GAP and TEX, the slopes are the
identical.
302
Activation energy of thermal decomposition was calculated by employing CoatsRedfern method [7]. Activation energy of thermal decomposition of ADN increases from
150KJ/mole to 170KJ/mole in ADN+GAP formulation. For TEX, activation energy
increases from 120KJ/mole to 128KJ/mole in TEX + GAP formulation. The increase in the
activation energy of ADN+GAP formulation is higher than that for TEX+GAP.
Consequently, it could be concluded that, GAP stabilizes the thermal decomposition of ADN
more than that of TEX.
3.2
Differential Scanning Calorimetry
Differential Scanning Calorimetry (DSC) is a more versatile technique to understand the

thermal decomposition process of propellant formulations. Figure 3 presents the DSC curves
for GAP, ADN and ADN + GAP formulations.
16
177.4 C
Heater Power (mW)
14
12
0
166.32 C
10
GAP+ADN
8
6
86.22 C
ADN
0
231.6 C
76.4 C
GAP
2
0
50
100
150
200
250
300
350
Temperature ( C)
Fig 3.
DSC curves of GAP,ADN, and ADN+GAP formulations
GAP exhibits an exothermic decomposition peak at 231.60C with a heat of

decomposition of 1294 J/g. ADN displays an endothermic melting peak at 76.40C, followed
by an exothermic decomposition of the melt at 166.320C. The heat of decomposition of ADN
is 2131J/g. ADN + GAP propellant formulation, also shows an analogous DSC curve as that
of ADN. The exothermic decomposition temperature has shifted to 177.40C. The heat of
decomposition is 2042 J/g. Exothermic peak due to decomposition of GAP in the
composition could not be observed.
DSC curves for GAP, TEX and TEX+ GAP formulations are presented in figure 4. TEX
displays a narrow decomposition peak at 297.150C with a heat of decomposition of 730J/g.
In contrast to the ADN + GAP formulation, TEX + GAP exhibits two exothermic
decomposition peaks at 2280C and 297.150C. The heats of decomposition are 600J/g and
380J/g respectively. This double decomposition behavior could not be assessed clearly from
the TGA experiments. It is noteworthy that the decomposition temperature of GAP in
GAP+TEX formulation shifts to lower values.
303
Comparing figures 3&4, decomposition temperature of ADN shifts to higher values in

the presence of GAP. On the other hand, in the presence of GAP, the decomposition
temperature of TEX remains the same. Shift in the decomposition temperature of the solid
propellant demonstrates that GAP is capable of enhancing the thermal stability of ADN. This
result is consistent with our TGA results.
18
297.15
228.01 C
Heater Power(mW)
16
14
GAP+TEX
12
297.15
TEX
10
231.68 C
8
6
GAP
4
2
0
50
100
150
200
250
300
Temperature( C)
Fig 4.
DSC curves of GAP,TEX, and TEX+GAP formulations
Kissinger equation [8] (equation 1) was employed to evaluate the activation energy of
thermal decomposition of GAP and the GAP component in GAP+TEX formulation.
E
=
+K
2
RTd
T
d
- ln
(1)
Where is the heating rate, Td is the temperature of decomposition, E is the energy of

activation for thermal decomposition, R is gas constant and K is a constant. The Kissinger
plots are presented in figure 5.
12.0
GAP
GAP in TEX+GAP
11.5
-ln(/Td )
11.0
10.5
10.0
9.5
1.86
1.88
1.90
1.92
1.94
1.96
1.98
2.00
-1
1000/Td(K )
Fig 5.
Kissinger plots for GAP and GAP component in TEX+GAP formulation
304
Activation energies are calculated from the slopes of the Kissinger plots. The activation
energy for pure GAP was computed as 155.6 KJ/mole and that for the GAP component in
GAP+TEX was computed as 126.6 KJ/mole. It could be concluded that, GAP is more
vulnerable to thermal degradation in the presence of TEX.
4.
CONCLUSIONS
Thermogravimetry analysis (TGA) and Differential Scanning Calorimetry were used to

study the thermal decomposition characteristics of GAP and its two propellant formulations
with two high performance oxidizers; ADN & TEX. TGA results indicated that inclusion of
GAP has improved the thermal stability of its pseudo propellant formulations with TEX and
ADN. Clearly, mixing of nitramine oxidizers with energetic polymers has resulted in the
formation of more thermally stable propellant formulations. The thermal stabilizing effect of
GAP is more pronounced in ADN when compared to that of TEX. DSC results provided a
comprehensive assessment of the thermal degradation process. It was shown that, GAP is
more vulnerable to thermal degradation in the presence of TEX.
Acknowledgements
We thank Dr.D.Bahulayan and Mr. Santhosh Gopalakrishnan for providing the TEX and
ADN samples. Mr.Erwin Wouterson Merijn, School of materials science and engineering,
NTU helped us with the DSC measurements.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
Y.J.LEE, C.J.TANG, and T.A.LITZINGER: Thermal Decomposition of RDX/BAMO PseudoPropellants, Combustion & Flame, 117: P.795-809, 1999.
M.J.ROSSI, J.C.BOTTARO, and D.F. McMILLEN: Thermal Decomposition of the New
Energetic Material ADN in Relation to Nitramide and Ammonium Nitrate, International
Journal of Chemical Kinetics, 25(7) : P. 549-70, 1993.
T.M.KLAPOTKE, and H.-G.ANG: Estimation of Crystalline Density of Nitramine High
Energy Materials, Propellants,Explosives,Pyrotechnics, 26(5) : P.221-4, 2001.
B. GAUR, B. LOCHAB, V. CHOUDHARY and I.K.VARMA: Azido Polymers: Energetic
Binders for Solid Rocket Propellants, Journal of Macromolecular Science, Polymer Reviews,
C43(4): P. 505-45, 2003.
T.B. BRILL: Surface Chemistry of Energetic Materials at High Temperature, Materials
Research Society Symposium Proceedings, Pittsburgh, No: 296, P.269-280, 1992.
S. ZEMAN: Analysis and Prediction of Arrhenius Parameters of Low Temperature
Thermolysis of Nitramines by means of the 15N NMR Spectroscopy. Thermochimica Acta, 333:
P. 121-129, 1999.
A.W. COATS and J.P. REDFERN: Kinetic Parameters from Thermogravimetric Data. Nature,
201(4914), P68-69.1963.
H.E. KISSINGER: Reaction Kinetics in Differential Thermal Analysis. Analytical Chemistry,
29(11), P.1702-1706, 1957.
305
DECOMPOSITION OF ENERGETIC MATERIALS

INVESTIGATED BY MOLECULAR SIMULATIONS: HNIW, TNA
Miroslav Pospil* and Pavel Vvra**
* Charles University Prague, Faculty of Mathematics and Physics,
Department of Chemical Physics and Optics, Ke Karlovu 3, 12116 Prague 2, Czech Republic
** University of Pardubice, Faculty of Chemical Technology,
Department of Theory and Technology of Explosives,
Studentsk 95, 53210 Pardubice, Czech Republic
Abstract:
Classical molecular dynamics simulations in Cerius2 modelling environment were used
to describe the decomposition process of crystal structure 2,4,6,8,10,12-hexanitro2,4,6,8,10,12-hexaazatetracyclo [5.5.0.05,9.03,11] dodecane known as HNIW and 1amino-2,4,6-trinitrobenzene, known as TNA. The dynamics simulations were carried out
in 3D periodic systems of above mentioned crystal structures under high initial
temperatures related to impact sensitivity. Dynamics trajectories were saved for a series
of initial temperatures in the range 300 K - 1000 K for HNIW and in the range 1000 K
5000 K for TNA. The detailed analysis of the dynamics trajectories allows us to describe
the mechanism and kinetics of decomposition processes under selected high
temperatures. Higher values of temperature decrease the calculation time
corresponding to the rupture of the first N-NO2 bond and also dramatically increase the
kinetic energy of the system.
Keywords:
1.
energetic materials, HNIW, TNA, molecular decomposition, classical

molecular dynamics simulations
INTRODUCTION
The presented work is a continuation of our previous two studies focused to explain the
molecular crystal decomposition processes in energetic materials [1-2]. The first one describes
the decomposition of cyclotrimethylene-trinitramine, known as RDX [3], cyclotetramethylene-tetranitramine, known as HMX [4], 1,1-diamino-2,2-dinitroethene,
known as DADNE [5] and nitroquanidine, known as NQ [6] under high temperature. The
second one describes the decomposition of RDX under high pressure. The classical
molecular dynamics simulations in Cerius2 modelling environment [7] were used for the
description of molecular crystals decomposition. The aim of this work is to enlarge the set of
energetic materials investigated by molecular dynamics simulations, Monte Carlo and other
computational methods [8-14]. In this case it is 2,4,6,8,10,12-hexanitro-2,4,6,8,10,12hexaazatetracyclo [5.5.0.05,9.03,11] dodecane, known as HNIW [15], see figure 1 and 1-amino2,4,6-trinitrobenzene, known as TNA [16], see figure 2. The main aim of the presented work
was step by step describe the mechanism of explosion in crystal structure and find some
reliable parameters characterizing the decomposition process of these energetic materials
with using molecular dynamics simulations.
306
2.
STRATEGY OF MOLECULAR DYNAMICS SIMULATIONS
The decomposition process under high temperature was simulated by classical

molecular dynamics. Molecular simulations are based on the empirical force field which
describes basic atoms parameters such as valence, atom mass, charge, van der Waals radius,
hybridization, etc. The force field cvff_950 [17] has been used. This force field allows
interrupting bonds in crystal structures under high pressure or under high temperature due to
included Morse potential describing bond energy. The harmonic potential included in the
most of force field is not suitable for calculations of decomposition. Charges were calculated
by Charge equilibration method (Qeq) [18]. During calculation, the decomposition process of
crystal structures is carried out to trajectory file for following investigation and analysis. The
dynamics trajectories include information about the model structure, temperature and
energies in dependence on the time. Moreover the trajectories allow us to visualise the time
dependence of the crystal structure decomposition process observable from the first bond
interrupting to releasing of nitro groups from structure. Data from trajectories allows us to
estimate the reliable parameters characterising the decomposition process of crystal
structures.
The molecular dynamics was run under these conditions: The Impulse type dynamics
was used. That means the system obtain the initial impulse at the beginning of calculation,
corresponding approximately to double of the origin selected temperature to assign the initial
velocities in agreement with Maxwell Boltzmann distribution. Impulse dynamic gives to
the selected atoms (all atoms in crystal structure in this work) an initial impulse in direction
of x-axis (this is valid for calculations in the present work; the axis can be generally
changed). Dynamics were performed in microcanonical ensemble (NVE), the number of
atoms, volume and energy are constant. Although the temperature is not constant in NVE
ensemble of impulse dynamic, Cerius2 allows us to fix the temperature within specified
tolerances by periodic rescaling of the velocities. The dynamics time step was 1 fs for TNA
crystal and 0.1 fs for HNIW crystal because of higher impact sensitivity in case of HNIW.
The maximum length of calculation in ps was chosen about 100 ps. If the crystal structure
does not decompose, the higher temperature was used because of saving the computer time.
That means the structure can decompose after 100 ps but this situation was not tested in this
work, see [1, 2]. The initial temperature was partially derived from the impact sensitivity of
the investigated crystal structures. In all calculations the crystallographic data were used as a
starting geometry for initial models. No minimization calculations and methods were used
during simulations. Bonding tolerance was 1.150 and molecular dynamics were calculated
with ideal periodical crystal structure without any defects.
3.
RESULTS AND DISCUSSIONS
The decomposition of TNA crystal structure is quite different from decomposition of

HNIW crystal structure. Some similarities can be seen among decomposition of TNA and
DADNE, NQ, published in our previous study, but in case DADNE and NQ the individual
molecules were decomposed [1]. Also decomposition process of HNIW crystal structure
shows some analogue with individual molecule decomposition of RDX and HMX [1] but in
both cases the differences were observed.
The impact sensitivity of TNA crystal structure is 170 cm [19] and on the base of this the
initial temperature was selected. The initial temperature for TNA decomposition was
estimated for the following values: 5000 K, 1000K, 3000 K, 2000 K, 2500 K and 2800 K.
The trajectories for 1000 K and 2000 K are not presented because of beginning time of
307
decomposition over 100 ps of calculation time. For development of decomposition process

in time with respect to initial temperature see figure 3, curves from the left: 5000 K, 3000 K,
2800 K and 2500 K. It is also important to mention, that the real initial temperature values
are double in the beginning of molecular dynamics calculations. The decomposition of TNA
generally starts with rupture of the N-H bond. Then the ring from six carbons is break up.
Usually two C-C bonds are decomposed. These dissociations are followed by rupture of the
first C-NO2 bond in the crystal structure. In this time the total decomposition processes are
started in the whole volume of crystal cell and after several steps the molecular dynamics is
interrupted due to high degree of dissociation of calculated structure. The decomposition
process is characterized with fluctuation of kinetic energy, as we can see in the figures 4-7
and also in the work [1].
The impact sensitivity of HNIW crystal structure is 23 cm [19]. The initial temperature
for HNIW decomposition was estimated for these values: 1000K, 500 K, 400 K and 300 K.
The HNIW crystal structure was decomposed for all selected temperatures. The overview of
decomposition in time with respect to initial temperature see figure 8, curves from the left:
1000 K, 500 K, 400 K and 300 K. The decomposition process in this case starts with rupture
of the C-N bonds. Decomposition of C-N bonds destroys the closed chain structure of HNIW
and new non-cyclic mainly linear or branched chain is created. Later this chain is dissociated
to two or more fragments which are predominantly compounded from CHNNO2 or
CH(NNO2)2. (Some far analogue can be seen with decomposition of RDX and HMX crystal
structures [1]. These cyclic molecules are also decomposed mainly to linear chains or directly
to fragments compounded from CH2NNO2). After fragmentation of the HNIW molecule the
N-NO2 bond is ruptured in whichever fragment. NO2 group is released and the total
decomposition process of the HNIW crystal structure is followed. See figures 9-12 for the
description of dissociation processes of HNIW crystal structure under different initial values.
4.
CONCLUSIONS
Molecular dynamics simulations in the present work were used for decomposition of the
perfect 3D crystals without any structural defects. Analyses of dynamics trajectories allow us
in details describe the mechanism of the crystal structure decomposition for TNA and
HNIW. Moreover we obtain information about parameter values characterizing the explosive
materials such as kinetic energy, temperature, pressure [2], etc. These parameters are usually
very difficult compared with experimental results perhaps due to difference among real and
ideal crystal structures. But it is possible in some cases. Anyway, results from molecular
dynamics simulations offered us new way for characterization of explosive materials without
any human risks.
Acknowledgement:
Experiments were performed thanks to financial support from the Ministry of Education,
Youth and Sports of the Czech Republic (Project MSM0021627501).
308
Fig 1.
Crystal structure of HNIW.
Fig 2.
Crystal structure of TNA.
309
5000
Kinetic energy [kcal]
4000
3000
2000
1000
0
0
20
40
60
80
100
time [ps]
Fig 3.
Decomposition of TNA crystal under different initial temperatures.
10000
Decomposition of TNA, T = 2500 K
8000
C-NO2
6000
C-C
4000
C-NH2
2000
0
95
96
97
98
99
100
time [ps]
Fig 4.
Rupture of bonds in TNA crystal under initial temperature 2500 K.
310
101
5000
4000
C-NO2
3000
C-C
2000
N-H
1000
0
45.0
45.2
45.4
45.6
45.8
46.0
46.2
46.4
time [ps]
Fig 5.
7000

C-NO2
6000
5000
4000
2x C-C
3000
N-H
2000
1000
0
31.0
31.5
32.0
32.5
33.0
time [ps]
Fig 6.
311
33.5
30000
25000
20000
15000
C-NO2
2x C-C
10000
N-H
5000
0
0.0
0.1
0.2
0.3
0.4
0.5
time [ps]
Fig 7.
40000
35000
30000
25000
20000
15000
10000
5000
0
-1
time [ps]
Fig 8.
Decomposition of HNIW crystal under different initial temperatures.
312
10
10000
Decomposition of HNIW, T = 300 K
8000
fragments
6000
N-NO2
4000
2x C-N
2000
0
8.5
8.6
8.7
8.8
8.9
9.0
time [ps]
Fig 9.
Rupture of bonds in HNIW crystal under initial temperature 300 K.
10000
8000
6000
N-NO2
4000
fragments
C-N
2000
0
2.5
2.6
2.7
2.8
2.9
time [ps]
Fig 10. Rupture of bonds in HNIW crystal under initial temperature 400 K.
313
10000
8000
6000
N-NO2
4000
C-N
fragments
2000
0
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
time [ps]
10000
8000
6000
4000
fragments
N-NO2
C-N
2000
0
0.00
0.05
0.10
0.15
0.20
time [ps]
314
0.25
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
M. Pospil, P. apkov, P. Vvra and S. Zeman: Characterization of Explosive Materials

Using Molecular Dynamics Simulations, Proc. of the 5th Seminar New Trends in Research of
Energetic Materials, University of Pardubice, Czech Republic, 5, p. 262-270, 2002.
M. Pospil, P. apkov, P. Vvra and S. Zeman: Classical Molecular Dynamics Simulations of
RDX Decomposition under High Pressure, Proceedings of the 6th Seminar New Trends in
Research of Energetic Materials, University of Pardubice, Czech Republic, 6, p. 275-284,
2003.
C. S. Choi and E. Prince: The Crystal Structure of Cyclotrimethylene-Trinitramine, Acta Cryst.,
B 28, p. 2857-2862, 1972.
C. S. Choi and H. P. Boutin: A Study of the Crystal Structure of -Cyclotetramethylene
Tetranitramine by Neutron Diffraction, Acta Cryst., B 26, p. 1235-1240, 1970.
U. Bemm and H. stmark: 1,1-Diamino-2,2-dinitroethylene: a Novel Energetic Material with
Infinite Layers in Two Dimensions, Acta Crystallographica C, 54, p. 1997-1999, 1998.
A.J. Bracuti: Crystal structure refinement of nitroguanidine, Journal of Chemical
Crystallography, 29, p. 671-676, 1999.
Cerius2 documentation, June 2000, San Diego: Molecular Simulations Inc, 2000.
P. Politzer, J. S. Murray, J. M. Seminario, P. Lane, M. E. Grice and M. C. Concha:
Computational Characterization of Energetic Materials, J. Mol. Struct. (Theochem), 573 p. 110, 2001.
M. M. Kuklja: Thermal Decomposition of Solid Cyclotrimethylene Trinitramine, J. Phys.
Chem. B, 105, p. 10159-10162, 2001.
D. V. Shalashilin and D. L. Thompson: Monte Carlo Variational Transition-State Theory Study
of the Unimolecular Dissociations of RDX, J. Phys. Chem. A, 101, p. 961-966, 1997.
Y. Kohno, K. Ueda and A. Imamura: Molecular Dynamics Simulations of Initial
Decomposition Process on the Unique N-N Bond in Nitramines in the Crystalline State, J.
Phys. Chem., 100, p. 4701-4712, 1996.
H. Dorsett: Computational Studies of FOX-7, a New Insensitive Explosive, DSTO Aeronautical
and Maritime Research Laboratory, Salisbury, South Australia, p. 1-19, 2000.
M.R. Manaa, L.E. Fried, C.F. Melius, M. Elstner and T. Frauenheim: Decomposition of HMX
at Extreme Conditions: A Molecular Dynamics Simulation, J. Phys. Chem. A., 106, p. 90249029, 2002
D.C. Sorescu, B.M. Rice, D.L. Thompson: Intermolecular Potential for the Hexahydro-1,3,5trinitro-1,3,5-s-triazine Crystal (RDX): A Crystal Packing, Monte Carlo, and Molecular
Dynamics Study, J. Phys. Chem. B, 101, p. 798-808, 1997.
Y. Ou, H. Jia, Y. Xu, B. Chen, G. Fan, L. Liu, F. Zheng, Z. Pan and C. Wang: Synthesis and
crystal structure of -hexanitrohexaazaisowurtzitane, Science in China B, 42, p. 217-224,
1999.
J.R. Holden, C. Dickinson and C.M. Bock: The Crystal Structure of 2,4,6-Trinitroaniline,
Journal of Phys. Chem., 76, p. 3597-3602, 1972.
A. T. Hagler, E. Huler and S. Lifson: Energy Functions for Peptides and Proteins. I. Derivation
of a Consistent Force Field Including the Hydrogen Bond from the Amide Crystals, J. Am.
Chem. Soc, 96, p. 5319-5327, 1974.
A. K. Rapp and W. A. Goddard III: Charge Equilibration for Molecular Dynamics
Simulations, J. Phys. Chem., 95, p. 3358-3363, 1991.
M. Pospil and P. Vvra: Study of electron density of molecules, intermolecular forces and
impact sensitivity of explosives, Proceedings of the 7th Seminar New Trends in Research of
Energetic Materials, University of Pardubice, Czech Republic, 7, p. 600-605, 2004.
315
THE PHASE STABILIZATION OF AMMONIUM NITRATE

BY POTASSIUM DINITRAMIDE
A DIFFERENTIAL SCANNING CALORIMETRIC STUDY
G. Santhosh,1 S. Venkatachalam,2 K. Krishnan, and K.N. Ninan
Propellants and Special Chemicals Group
Vikram Sarabhai Space Centre, Trivandrum 695 022, INDIA.
Abstract:
Ammonium nitrate (AN) has long been considered as a desirable oxidizer for solid
rocket propellants and explosives because of its low cost, low sensitivity, low smoke
signature and the absence of halogen atoms. The stumbling block for the application of
AN as rocket propellant oxidizer is its dimensional instability caused by polymorphic
transitions near the propellant processing and storage temperatures and its extreme
hygroscopicity leading to unpredictable ballistic performance and catastrophic rocket
motor failure. The phase transitions leading to volume change can be overcome by the
use of phase stabilizers. Unfortunately most of the phase stabilizers used to prepare
phase stabilized ammonium nitrate (PSAN) are non-energetic and the overall energy
content of the formulation is reduced to a great extent. Therefore it is necessary to have
PSAN with improved energy content. The present paper discusses the differential
scanning calorimetry (DSC) study of potassium dinitramide (KDN)-AN compositions.
Concentration levels of KDN in the range of 0.2 to 5% (by weight) were studied. The
results obtained show that at concentration levels above 3%, KDN has better phase
stabilization effect on AN.
Keywords:
1.
Ammonium nitrate, potassium dinitramide (KDN), phase stabilization,

differential scanning calorimetry, thermal decomposition.
INTRODUCTION
AN finds extensive use in industrial explosives, gas generators and solid rocket
propellants. Much attention has been paid on the development and use of environmentalfriendly solid rocket propellant composition containing AN. Some important reasons are its
low cost, chemical stability, low sensitivity to friction and impact as well as its low polluting
characteristics. Its use as a key oxidizer in propellants has been limited due to its extreme
hygroscopicity, phase transition characteristics, low energy content and least reactivity. A
review describing the history, manufacture, safety, physical and chemical properties and
other applications of AN has been appeared [1].
Many efforts have been made in the recent years to modify the phase transition
behaviour of AN with the addition of KF, KNO3, oxides of Zn, Cu, Ni or metal complexes.
Different processes for the phase stabilization of AN using various phase stabilizers have
been reported [2-13]. In these systems, the IV-III phase transition (discussed later) depends on
a variety of parameters like moisture content, mode of crystallization, grain size, purity and
experimental technique adopted [14]. PSAN has been considered as an oxidizer in various
316
propellant compositions in combination with different binders [15-17]. The thermal

decomposition of PSAN has received much attention in the recent years [18-20].
KDN (KN(NO2)2) has attracted a wide interest as a promising new class of energetic
oxidizer and is being explored as an energetic phase stabilizer [21-22] . Ammonium nitrate
phase stabilized with KDN exhibits less hygroscopicity, improved energy in propellant and
pyrotechnic formulations, increased burning rate and no combustion instability.
In the present investigation the effect of KDN on the phase stabilization of AN has been
studied by differential scanning calorimetry (DSC). The results indicate that the KDN-AN
formulation is a promising candidate in which the KDN has good phase-stabilizing effect on
AN.
2.
EXPERIMENTAL
2.1
Materials
Ammonium nitrate (s.d Fine chem., Mumbai) was recrystallized from methanol and
dried under vacuum in a desiccator. KOH (AR Grade, s.d Fine chem., Mumbai) and
methanol (AR Grade, SRL Mumbai) were used as received. Potassium dinitramide (KDN)
was prepared by double decomposition reaction of ammonium dinitramide (ADN) with
potassium hydroxide (KOH).
2.2
Methods
The PSAN samples based on KDN were prepared by co-crystallization from methanol.
The amount of KDN in AN was varied from 0.2 % to 5% (by weight). The KDN-AN
mixtures were dissolved in methanol with slight warming and evaporated under vacuum.
The obtained solid was further dried under vacuum in a desiccator.
The DSC curves were recorded by a Mettler TA 3000 thermal analysis system in
conjunction with DSC-20 standard cell. All the experiments were carried out at 5C/min.
Constant sample mass of 50.3mg was used in all the experiments. The samples were placed
in an encapsulated aluminium pan with a pierced lid and the atmosphere was static air.
3.
AN undergoes different crystal transformations viz., , , and in the temperature

range of 18C to 125C before its melting at 169C [23]. AN used in this study shows phase
transformations at 53.8C, 90.2C, 126.1C and a melting transition at 166.2C followed by
its endothermic decomposition with a peak maximum at 242.4C. Figure 1 shows the DSC
trace of AN at a heating rate of 5C/min.
317
20
10
Heat Flow (mW)
0
o
53.8 C
90.2 C
-10
o
126.1 C
-20
-30
166.2 C
-40
242.4 C
50
100
150
200
250
300
350
Temperature ( C)
Fig 1. DSC trace of ammonium nitrate (AN) at a heating rate of 5oC/min

The crystal transformation of AN has been the subject of examination by numerous
investigators. Under ordinary pressure AN exhibits five stable polymorphic transformations.
The transition temperatures of the five phases are given below .
tetragonal crystal
(5)
-18oC
orthorhombic crystal
(4)
32.1oC
orthorhombic crystal
(3)
84.2oC
liquid
169.6oC
cubic crystal
(1)
125.2oC
tetragonal crystal
(2)
The transition temperatures, crystallographic data and stability ranges of the various
phases of AN are given in Table 1.
318
Table 1. Crystallographic details of ammonium nitrate phases.

Phase
Crystal
system
IV
III
Orthorhombic Orthorhombic Orthorhombic
II
Tetragonal
Cubic
Space group
Pccn
Pmmn
Pnma
P4/mbm
Pm3m
Cell
parameter
a=7.943
b=7.972
c=9.832
a=5.745
b=5.438
c=4.942
a=7.677
b=5.820
c=7.140
a=5.7193
c=49326
a=4.366
Measured at
-100C
22C
24C
82C
150C
Stability
-Moist
<-18C
-18 - 32C
32-84C
84-125C
<-18C
-18 - 55C
--
55-125C
>125C
1.76
1.72
1.68
1.60
1.55
-Dry
Density
(g/cm3)
>125C
3.1 Phase transition and modifications

A search in the literature shows wide variations in the reported phase transition
temperature. While the fusion temperature and transition temperature at 125C remain
unambiguous in most of the schemes, the transition IV-II is debated widely [1]. The reason
for the dissimilarity in the literature on the transition temperature is ascribed to its
dependence on various factors. Some of these include moisture, mode of crystallization,
thermal history of the sample, sample weight, heating mode, purity of samples, grain size,
experimental technique etc., Though a solid-solid transition takes place during IV-II, the
involvement of moisture as a catalyst is needed for this transition. Thus only moist AN will
exhibit the transition at 32C in the DSC study. The transitions do not take place precisely in
the same temperature range during the heating and cooling cycle. The II to III transition
occurs anywhere between 48 and 84C and III to II transition occurs between 84 and 89C. It
is also observed that IV to III transition takes place between 37.5 and 55.5C and III to IV
transition occurs between 0 and 35C. Thus the phase transition of AN is too complex and
inconsistent and depends on moisture content of the sample and whether the samples are
heated from lower to higher temperatures or vice versa.
Realizing the technical significance of the IV-III transition, there has been many
attempts to stabilize this phase. The stabilization of IV-III phase is necessary to prevent
volume changes occurring during processing and storage of AN based propellants. In the
present study, phase stabilization of AN has been studied using KDN as a phase stabilizer.
Introduction of small amounts of KDN can have phase stabilization effect on AN without
319
energy penalty. KDN-AN system also shows improved ballistic performance-increased

thrust, reduced variability, and better predictability, compared with AN propellants without
KDN.
3.2
Thermal studies of AN phase stabilized with KDN
A process for the preparation of phase-stabilized ammonium nitrate by addition of KDN

is disclosed [24]. This process involves making of a saturated solution of AN and KDN in
water followed by precipitation using a miscible solvent.
The method used in our study to prepare PSAN samples containing 0.2, 0.5, 1, 3 and 5%
KDN is described in 2.2, and their thermal properties were examined by DSC.
The DSC curve of AN (Figure 1) shows transitions at 53.8C, 90.2C, 126.1C and
166.2C. The DSC curve for AN does not show the transition at 32C because this transition
is shifted to 53.8C due to the absence of moisture in the AN used in the present study. The
melting endotherm was observed at 166.2C followed by endothermic decomposition
between 170 and 260C with a peak maximum at 242C.
The DSC curve of KDN is shown in Figure 2. It shows an endotherm at 130.2C
corresponding to the melting of KDN, followed by exothermic decomposition in the range of
170 to 200C. The DSC curve also shows an endotherm at 322C corresponding to the
melting of the in-situ formed potassium nitrate (KN) in the decomposition of KDN.
50
171.3 C
40
Heat Flow (mW)
30
20
10
0
o
322.0 C
-10
130.2 C
50
100
150
200
250
300
Temperature ( C)
Fig 2. DSC trace of KDN at a heating rate of 5oC/min
320
350
The DSC curves of PSAN with different proportion of KDN (0.2 to 5%) are shown in
Figure 3.
5% KDN
3% KDN
1% KDN
exo >
0.5% KDN
0.2% KDN
50
100
150
200
250
300
350
400
Temperature ( C)
Fig 3. DSC overlay of PSAN samples with different amounts of KDN.

The observed transition temperatures (from DSC) in AN phase stabilized with KDN are
given in Table 2.
Table 2. Observed characteristic transition temperature of PSAN.
Sample
Temperature (C)
AN
53.8
90.2
126.1
166.2
0.2% KDN
54.0
91.0
130.0
169.0
0.5% KDN
45.0
91.2
126.6
166.7
1% KDN
41.2
92.5
128.0
168.7
3% KDN
--
94.0
128.9
163.2
5% KDN
--
95.8
126.7
163.3
321
From Table 2, it can be seen that the PSAN samples with 3 and 5% KDN did not show
any transition, while other samples do show the transition. For the PSAN samples with
0.5 and 1% KDN, the transition is lowered to 45 to 40C. The other phase transitions viz.,
, and are least affected by the incorporation of KDN in the AN lattice.
4.
CONCLUSION
The efficiency of KDN in the phase stabilization of AN was studied by DSC technique.
A concentration level of 3% (by weight) of KDN is found to be sufficient for the suppression
of the phase transition at 52.8C. It is also observed that the transition at 88.1C is shifted to
95C. The phase transition at 126.5C and the melting transition at 169C show no shift upto
a concentration of 3% KDN. At higher concentration of KDN, an endothermic transition at
325C is observed which is attributed to the in-situ formation of KNO3 during the
decomposition of KDN. The results indicate that the KDN-AN formulation is a promising
candidate in which the KDN has good phase-stabilizing effect on AN.
Acknowledgment:
One of the authors GS thanks the Indian Space Research Organization (ISRO) for award
of a research fellowship.
REFERENCES
[1]
C. OOMMEN, S.R. JAIN: Ammonium nitrate: a promising rocket propellant oxidizer, Journal
of Hazardous Materials, A67, 253, 1999.
[2] MISHRA INDU B: Potassium fluoride stabilized ammonium nitrate, United states patent
4552736, 1985
[3] [OBERTH ADOLF E: Phase-stabilization of ammonium nitrate by zinc diammine complexes,
United states patent 5071630, 1991
[4] ENGEL WALTER, EISENREICH NORBERT, THIEL KLAUS-DIETER: Process for
producing phase-stabilized ammonium nitrate, United states patent 5063036, 1991
[5] MEHROTRA ASHOK K, MARKOVICH I LEE: Potassium fluoride stabilized ammonium
nitrate and method of producing potassium fluoride stabilized ammonium nitrate, United states
patent 5098683, 1992
[6] HIGHSMITH THOMAS K, HINSHAW CAROL J, WARDLE ROBERT B: Phase-stabilized
ammonium nitrate and method of making same, United states patent 5292387, 1994
[7] KIM, JUN-HYUNG: Preparation of phase stabilized ammonium nitrate (PSAN) by a salting out
process, Journal of Chemical Engineering of Japan, 30(2), 336-338,1997
[8] KIM JUN-HYUN, HON MYUN-PYO, IMU YO JIN: Production of phase stabilized
ammonium nitrate by salting out method, Japanese patent JP 11100210, 1999
[9] MENDENHALL IVAN V, SMITH ROBERT G, TAYLOR ROBERT D: Phase stabilized ammonium
nitrate and method of making the same, United states patent 6372191, 2002
[10] BLOMQUIST, HAROLD R: Process for preparing free-flowing particulate phase stabilized
ammonium nitrate, United states patent application 20020096235, 2002
[11] WILLIAM P. SAMPSON, PETER J.ASTRAUSKAS: Process for preparing phase stabilized
ammonium nitrate, United states patent 6641622, 2003
322
[12] ENGEL WALTER, HEINISCH HERBERT: Process for producing phase-stabilized

ammonium nitrate, United states patent 6508995, 2003
[13] HAMILTON, BRIAN K: Phase-stabilized ammonium nitrate, United states patent application
20040149363, 2004
[14] L. FILIPESCU, D. FATU, T. COSEAC, MOCIOI MIHAELA, E. SEGAL: On the chemical and
thermal stabilization of ammonium nitrate(IV), Thermochimica Acta, 97, 229-41,1986
[15] MENKE, KLAUS, BOEHNLEIN-MAUSS JUTTA: Properties of AN and PSAN/GAP-propellants.
International Annual Conference of ICT, 27th(Energetic Materials), 21.1-21.15, 1996
[16] MENKE KLAUS, BOEHNLEIN-MAUSS JUTTA, SCHUBERT HILTMAR: Characteristic properties of
AN/GAP-propellants, Propellants, Explosives, Pyrotechnics, 21(3), 139-145,1996
[17] MENKE KLAUS, BOEHNLEIN-MAUSS JUTT, SCHMID HELMUT, BUCERIUS KLAUS
M, ; ENGEL WALTHER: Solid propellant based on phase-stabilized ammonium nitrate,
European Patent EP 705808,1996
[18] MATHEW SURESH, K. KRISHNAN, K.N. NINAN: A DSC study on the effect of RDX and HMX on
the thermal DECOMPOSITION of phase stabilized ammonium nitrate, Propellants, Explosives,
Pyrotechnics, 23(3), 150-154, 1998
[19] P.N. SIMOES, L.M. PEDROSO, A.A. PORTUGAL, J.L. CAMPOS: Study of the thermal
decomposition of PHASE-stabilized ammonium nitrate and determination of kinetic parameters,
Thermochimica Acta, 319(1-2), 55-65, 1998
[20] MATHEW Suresh, K. KRISHNAN, K.N. NINAN: Effect of energetic materials on thermal
decomposition of phase-stabilized ammonium nitrate - an eco-friendly oxidizer, Defence Science
Journal, 49(1), 65-69, 1999
[21] MING LEI, ZHI-ZHONG ZHANG, YANG-HUI KONG, ZI-RU LIU, CHUN-HUA ZHU, YING-HUI
SHAO AND PEI ZHANG: The thermal behavior of potassium dinitramide. Part 1. Thermal
stability, Thermochimica Acta, 335(1-2) , 105-112, 1999
[22] Ming LEI, ZI-RU LIU, YANG-HUI KONG, CUI-MEI YIN, BO-ZHOU WANG, YUAN WANG AND
PEI ZHANG: The thermal behavior of potassium dinitramide Part 2. Mechanism of thermal
decomposition, Thermochimica Acta 335(1-2), 113-120, 1999
[23] T. HATEKEYAMA, ZHENHAI LIU: Handbook of Thermal Analysis, John Wiley & Sons, 1998.
[24] KIM JUN Hyung, NO MAN KYUN, SEO TAE SEOK: Purification of phase-stabilized ammonium
nitrate by addition of potassium dinitramide, Korean Patent KR 2000015463, 2000, CAN:
136:120703
323
MAMMALIAN CELL CYTOTOXICITY OF NITROAROMATIC

EXPLOSIVES AND THEIR DEGRADATION PRODUCTS:
THE ROLE OF OXIDATIVE STRESS
Jonas arlauskas*, Aura Nemeikait-nien**, ilvinas Anuseviius*,
Henrikas Nivinskas*, Lina Miseviien*, Valentina Vilutien***,
and Narimantas nas*
* Institute of Biochemistry, Mokslinink 12, LT-08669 Vilnius, Lithuania
** Institute of Immunology, Vilnius University, Molt Pl. 29, LT-08409 Vilnius, Lithuania
*** The General J. emaitis Military Academy of Lithuania, ilo 5a, LT-10322 Vilnius, Lithuania
Abstract:
Toxic effects of nitroaromatic explosives on mammalian species are attributed to their
single-electron reduction by flavoenzymes (redox cycling of free radicals and oxidative
stress), and/or to their two-electron reduction (formation of alkylating nitroso- or
hydroxylamino products). In this work, we summarize the data on the enzymatic

reactivity of explosives (2,4,6-trinitrotoluene (TNT) and its amino- and
hydroxylamino- reduction products, tetryl, pentryl, tetranitrocarbazole,
dinitrobenzofuroxan, nitrotriazoles, tetranitrobenzimidazolone) and model
nitroaromatic compounds, and their cytotoxicity in bovine leukemia virustransformed lamb kidney fibroblasts (line FLK). Cytotoxicity of compounds
increased with an increase in their single-electron reduction potential (E17) or
their reactivity in enzymatic single-electron reduction reactions catalyzed by
NADPH:cytochrome P-450 reductase or ferredoxin:NADP+ reductase, and did
not follow their reactivity in DT-diaphorase-catalyzed two-electron reduction.
Further, the cytotoxicity was reduced by antioxidants, whereas dicumarol, an
inhibitor for DT-diaphorase, showed minor and equivocal effects. It shows that
the cytotoxicity of explosives in this cell line is caused mainly by the oxidative
stress. The cytotoxicity of amino- and hydroxylamino-DNTs was lower than that
of TNT, but higher than one may expect from their redox cycling activity. Thus,
the additional modes of their cytotoxicity may be involved. Our data imply that
the evaluation of redox properties of new explosives and high energy compounds
in enzymatic systems may be a useful approach to characterize or predict their
toxic properties.
Keywords:
1.
nitroaromatic explosives; free radicals; oxidative stress; cytotoxicity;
INTRODUCTION
Explosives and their degradation products are toxic and mutagenic to humans and other
mammalian species. Long lasting contacts with 2,4,6-trinitrotoluene (TNT) cause
methemoglobinemia, cataract, the reproductive toxicity, and increased incidence of urinary
tract, kidney, and liver tumours.[1] In spite of the numerous clinical and ecotoxicological
reports, the biochemical mechanisms of the (cyto)toxicity of nitroaromatic explosives,
324
especially their enzymatic aspects, have been poorly studied. Like for other nitroaromatics
(ArNO2), the cytotoxicity of TNT may be attributed to the single-electron enzymatic
reduction to radicals accompanied by the formation of the reactive oxygen species (oxidative
stress type of cytotoxicity):
ArNO2 + e- ArNO2-..
(1)
ArNO2-. + O2 ArNO2 + O2-.,
(2)
2O2-. + 2H+ H2O2 + O2,
(3)
O2-. + Fe3+ O2 + Fe2+,
(4)
Fe2+ + H2O2 Fe3+ + OH- + OH.,
(5)
and their two-electron reduction to nitroso and hydroxylamino metabolites causing the
cytotoxicity by their covalent binding to proteins and DNA: [2.3]
ArNO2 + 2e- + 2H+ ArNO + H2O,
(6)
ArNO + 2e- + 2H+ ArNHOH
(7)
Both types of reactions are catalyzed by flavoenzymes, e.g. NADPH:cytochrome P-450

reductase (P-450R, EC 1.6.2.4) performs a single-electron reduction, whereas
NADPH:quinone oxidoreductase (DT-diaphorase, NQO1, EC 1.6.99.2) performs a twoelectron reduction. Besides, hydroxylamines may undergo transition metal-catalyzed redox
cycling: [4]
ArNHOH + Cu2+ ArNHO-. + Cu+ + H+,
(8)
ArNHO-. + NADH ArNHOH + NAD. ,
(9)
NAD. + O2 NAD+ + O2-.
(10)
One of the possible additional mechanisms of toxicity of nitroaromatics is the inhibition

and/or inactivation of enzymes responsible for the cell antioxidant defence or cell growth
regulation, e.g., flavosulfoselenoenzyme NADPH:thioredoxin reductase (TrxR, EC 1.8.1.9),
or flavosulfoenzyme NADPH:glutathione reductase (GR, EC 1.6.4.2). [5]
This paper summarizes the data on the enzymatic reactivity and the mammalian cell
cytotoxicity of classical explosives like TNT and its hydroxylamino- and amino metabolites,
tetryl, and 2,4,6-trinitrophenyl-N-nitraminoethylnitrate (pentryl), and explosives of novel
generation, 1,3,6,8-tetranitrocarbazole (TNC), 4,6-dinitrobenzofuroxan (DNBF), 5-nitro1,2,4-triazol-3-one (NTO), 5-nitro-1,2,4-triazol-3-amine (ANTA), and 4,5,6,7tetranitrobenzimidazol-2-one (TNBO) (Fig. 1).
2.
MATERIALS AND METHODS
Explosives and their metabolites (Fig. 2) were synthesized according to the described
methods. [6-9] All the the compounds were characterized by melting points, 1H NMR, UV
and IR spectroscopy. All the other chemicals were obtained from Sigma, and used as
received.
Pig liver NADPH:cytochrome P-450 reductase and rat liver DT-diaphorase were
prepared as described. [6,7] Recombinant Anabaena ferredoxin:NADP+ reductase (FNR, EC
1.18.1.2) was a generous gift from Professor Carlos Gomez-Moreno (Zaragoza University,
325
Spain), recombinant rat NADPH:thioredoxin reductase (TrxR) was a generous gift from Dr.
Elias Arner (Karolinska Institutet, Stockholm, Sweden). The enzyme concentrations were
determined using 460 = 22 mM-1 cm-1 (P-450R), 459 = 9.4 mM-1 cm-1 (FNR), and 460 = 11
mM-1 cm-1. (NQO1, TrxR).
Fig 1.
Structural formulae of examined nitroaromatic explosives and their

degradation products.
The kinetic measurements were carried out spectrophotometrically using a Hitachi-557

spectrophotometer in 0.1 M K-phosphate buffer (pH 7.0) containing 1 mM EDTA at 25 oC.
The rates of enzymatic oxidation of NADPH by nitroaromatic compounds were determined
according to 340 = 6.2 mM-1cm-1, using 100 M NADPH (P-450R, NQO1), 200 M
NADPH (FNR), or 50 M NADPH (TrxR) as an electron donor. The catalytic constant (kcat)
and the bimolecular rate constant (kcat/Km) of the reduction of aromatic nitrocompounds
correspond to the reciprocal intercepts and slopes of plots [E]/v vs. 1/[ArNO2], where [E] is
326
the enzyme concentration and [ArNO2] is the concentration of nitrocompound. kcat is the
number of NADPH molecules oxidized by the single active center of an enzyme per second.
The activity of TrxR was determined using 2 mM 5,5-dithiobis-(2-nitrobenzoic acid)
(DTNB) as an electron acceptor, according to 412 = 13.6 mM-1cm-1. The reduction of
cytochrome c added in separate experiments, was monitored using 550 = 20 mM-1 cm-1.
The rates of oxygen consumption were monitored using a Clark electrode.
The culture of bovine leukemia virus-transformed lamb kidney fibroblasts (line FLK)
was grown and maintained in Eagles medium supplemented with 10 % fetal bovine serum
at 37 oC as described previously.[6-9] In the cytotoxicity experiments, cells (3.0104/ml) were
grown in the presence of various amounts of aromatic nitrocompounds for 24 h, and counted
using a hematocytometer with viability determined by the exclusion of Trypan blue. Before
the count, the cells were trypsinized.
The regression analysis was performed using Statistica software (version 4.3; Statsoft,
1993).
3.

3.1
Reactivity of nitroaromatic explosives towards single and two-electron transferring flavoenzymes
Under aerobic conditions, the cytotoxicity of a number of nitroaromatic compounds in

several cell lines increased with an increase in their single-electron reduction potential (E17)
with a relationship log cL50/E17 ~ -10 V-1, where cL50 was the compound concentration
for 50% cell survival. [6,10] Nitroaromatics induced lipid peroxidation, and their toxic effects
were decreased by antioxidants and Fe-ion chelator, desferrioxamine. Taken together with
these data, the observed log cL50 vs. E17 relationships point to the dominant role of the
oxidative stress-type cytotoxicity caused by free radical redox cycling (Eq. (1-5)). Indeed,
single-electron transferring flavoenzyme P-450R is not specific towards particular structures
of nitroaromatics, its reactivity follows the relationship log (kcat/Km)/E17 ~ 10 V-1, where
kcat/Km is the reaction bimolecular rate constant.[11,12] The analogous relationships are
characteristic of other single-electron transferring flavoenzymes, e.g., ferredoxin:NADP+
reductase (FNR, EC 1.18.1.2).[11,13] For this reason, it is worthwhile examining the reactivity
of nitroaromatic explosives towards P-450R and FNR (Table 1). The enzyme catalytic
constants (kcat) are not presented, because they showed little variation. The reactions were
accompanied by the oxidation of significant excess of NADPH over nitrocompound, and
stoichiometric to NADPH consumption of oxygen. In FNR-catalyzed reactions, the
reduction of added cytochrome c took place at a doubled rate of NADPH oxidation. The
reactions were partly inhibited by superoxide dismutase, thus showing an involvement of
both ArNO2-. (Eq. (1)) and superoxide (Eq. (2)) in cytochrome c reduction. The reduction
and redox cycling of tetryl by P-450R and FNR is accompanied by N-denitration and
formation of nitrite:
Tetryl
O2-.
e-
[Tetryl]-.
O2
e-, H+
N-Methylpicramide
(11)
NO2-
The same reaction mechanism is characteristic of pentryl.[6] Although the E17 values for
most nitroaromatic explosives except TNT and its amino metabolites are unavailable, one
327
may note that the reactivity of polynitroaromatic compounds tetryl, pentryl, TNC, DNBF,
and TNBO is similar to or even higher than that of TNT (Table 1). This is in line with a
general trend of an increase in the electron-accepting potency with an increase in the number
of nitrogroups.
Table 1. Single-electron reduction potentials of nitroaromatic explosives (E17),
and their reduction rate constants (kcat/Km) by NADPH:cytochrome
P-450 reductase (P-450R), ferredoxin:NADP+ reductase (FNR),
and DT-diaphorase (NQO1) at pH 7.0 and 25 oC.[6-9] aThe kcat values (s-1)
are given in parentheses.
No.
Compound
E17 (V)[14]
kcat/Km (M-1s-1)
P-450R
NQO1a
FNR
1.
Pentryl
1.3 x 107
5.5 x 105
5.7 x 105 (75)
2.
Tetryl
2.3 x 107
1.1 x 105
2.6 x 105 (73)
3.
TNC
2.6 x 107
7.3 x 104
5.0 x 106 (102)
4.
TNBO
8.1 x 105
3.3 x 105
2.6 x 105 (73)
5.
DNBF
3.9 x 105
3.8 x 104
4.5 x 103 (> 0.5)
6.
TNT
1.7 x 106
1.1 x 104
6.5 x 102 (1.0)
7.
2-NHOH-DNT
4.5 x 104
3.1 x 103
6.5 x 103 (0.35)
8.
4-NHOH-DNT
1.3 x 104
2.1 x 102
4.1 x 103 (0.52)
9.
2-NH2-DNT
-0.417
6.1 x 103
6.2 x 102
1.3 x 102 (> 0.04)
10.
4-NH2-DNT
-0.449
2.5 x 103
3.4 x 102
1.0 x 102 (> 0.04)
11.
2,4-(NH2)2-NT
-0.502
2.0 x 103
2.1 x 102
50
12.
NTO
9.0 x 102
5.5 x 101
30
13.
ANTA
1.7 x 103
2.6 x 102
30
-0.253
In contrast, the relationship between the activity of nitroaromatics towards two-(four)electron transferring mammalian DT-diaphorase (NQO1) and their reduction potential is
almost absent (Table 1). The substrate specificity of nitroreductase reaction of NQO1 is
currently poorly understood.[15] This enzyme reduces nitroaromatic compounds to
corresponding hydroxylamines. However, NQO1-catalyzed reduction of tetryl and pentryl
proceeds in a single-electron way (Eq. (11)).[6]
3.2
Nitroaromatic explosives as inhibitors and subversive substrates

for glutathione reductase and thioredoxin reductase
Glutathione reductase (GR) in the active center contains both FAD and a catalytic
disulfide, the latter reducing oxidized glutathione (GSSG) to antioxidant GSH. Mammalian
thioredoxin reductase (TrxR) contains FAD, disulfide, and an additional redox center, a
terminal selenosulfide group, which transfers electrons to the 12 kD disulfide redox protein
thioredoxin (Trx). Trx performs antioxidant and cytokine functions. The inhibition of these
328
enzymes may confer an additional pathway of cytotoxicity of nitroaromatics. Besides,

nitroaromatics may act as subversive substrates for these enzymes, i.e., generate free
radicals (Eq. (1)). Erythrocyte GR and rat TrxR perform a single-electron reduction of tetryl
(Eq. (11)) with kcat of 5 s-1 and 1.8 s-1, and kcat/Km = 2.0 x 103 M-1s-1, and 1.4 x 104 M-1s-1,
respectively, at pH 7.0.[16] TNT is reduced 10 times slower. However, it is unclear whether
these slow reactions may contribute to the oxidative stress-type cytotoxicity. On the other
hand, tetryl and pentryl react with GSH directly (k = 0.6 M-1s-1) forming trinitrophenylSG.[17] DNBF also reacts with GSH directly (k = 0.02 M-1s-1) forming the Meisenheimertype adduct.[9] Tetryl, TNT, and trinitrophenyl-SG inhibit GR with the inhibition constants
(Ki) of 4.5-14 M.[16] TNT is almost inactive as TrxR inhibitor (Ki = 400 M). On the other
hand, tetryl inhibits rat TrxR with Ki = 12 M, and acts as the irreversible inactivator of the
reduced enzyme with the maximal rate, 3.8 min-1 and the bimolecular rate constant, 730
M-1s-1. This reaction most probably proceeds via the covalent modification of reduced
selenocysteine or cysteine. Analogously, DNBF inhibits TrxR with Ki = 5 M, and
inactivates it with the maximal rate, 0.75 min-1, and
the bimolecular rate constant, 600
M-1s-1. The latter phenomena may be considered as potentially important in cytotoxicity.
3.3
Cytotoxicity of nitroaromatic explosives in FLK cells
Mammalian cell culture cytotoxicity data of TNT, its metabolites, and other
nitroaromatic explosives, expressed as concentrations causing 50 % cell death (cL50), are
summarized in Table 2. The most thorough cytotoxicity studies were performed with FLK
cells (Table 2).[6-9] The cytotoxicity of TNT, tetryl, DNBF, NTO, ANTA, TNBO, NH2DNTs, and NHOH-DNTs in this cell line was reduced by the antioxidant N,N-diphenyl-pphenylene diamine (DPPD), and desferrioxamine, an iron ion chelator which prevents the
Fenton reaction (Eq. (4-5)). Typically, at compound concentrations causing 40-60 % cell
death, DPPD or desferrioxamine increased cell viability to 70-85 %. In contrast, the
alkylating agent 1,3-bis-(2-chloroethyl)-1-nitrosourea (BCNU) which inactivates glutathione
reductase and depletes intracellular GSH, potentiated the cellular toxicity, reducing cell
viablity to 15-30 %. Cytotoxicity was accompanied by lipid peroxidation, suggesting the
involvement of oxidative stress, resulting from the enzymatic redox cycling of radicals of
explosives.[6-9] Enzymes responsible for free radical formation in FLK cells have not yet
been identified. Dicumarol, an inhibitor of NQO1, partly reduced the cytotoxicity of TNT,
TNBO, and DNBF, but, surprisingly, potentiated the cytotoxicity of tetryl and TNC. Both
protective and potentiating effects of dicumarol may not be considered as significant, since it
affected the cL50 values of the above compounds by approximately 1.5 times. It shows that
the cytotoxic consequences of reaction of NQO1 with all explosives in FLK cells are
equivocal, evidently depending on the further metabolic fate of reaction products. For a
number of explosives and model nitroaromatics, log cL50 in FLK cells decreased with an
increase in their E17 or the geometric mean of their reactivity in FNR- and P-450R-catalyzed
reactions, (log kcat/Km (FNR) + log kcat/Km (P-450R))/2 (Fig. 2).[6-9] This type of dependence
shows that the main cytotoxicity mechanisms of nitroaromatics in this cell line is the
oxidative stress, initiated by flavoenzyme-catalyzed single-electron reduction.
Table 2. The concentrations of nitroaromatic explosives and their metabolites for
the 50 % cell survival (cL50) in mammalian cell culture cytotoxicity
studies of FLK cells (24 h exposure),[6-9] Chinese hamster ovary K1(24 h)
and rat hepatoma H4IIE cells (24 h),[18] Chinese hamster lung V79 (24 h)
and human lymphoblast TK-6 cells (48 h),[19] and the geometrical mean
of their reactivity in single-electron enzymatic reduction,
(log kcat/Km (FNR) + log kcat/Km (P-450R))/2 .
329
No
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
Compound
Pentryl
Tetryl
TNC
TNBO
TNT
DNBF
2-NHOH-DNT
4-NHOH-DNT
2-NH2-DNT
4-NH2-DNT
2,4-(NH2)2-NT
NTO
ANTA
(log kcat/Km
(FNR) +
log kcat/Km
(P-450R))/2
6.426
6.200
6.140
5.850
5.136
5.080
4.072
3.218
3.289
2.965
2.812
2.350
2.822
FLK
5.01.0
2.20.3
8.02.0
305.0
255.0
708.0
407.0
11210
44035
31620
35040
>3500
3000400
KI
cL50 (M)
V79
106
18.8
>1270
>1270
>1500
H4IIE
TK-6
197
17.6
22
222
>328
>600
28.2
91.3
335
>1500
168
248
>600
Because the amino metabolites of TNT are less cytotoxic than TNT in several cell lines
(Table 2), it is possible that the flavoenzyme-catalyzed redox cycling is their universal
cytotoxicity mechanism. One may note that FLK cell cytotoxicity of NHOH-DNTs, NH2DNTs and (NH2)2-NT is somewhat higher than expected from their redox cycling activities
(Fig. 2). Besides, in other cell lines, NHOH-DNTs are as toxic as TNT, or even more toxic
(Table 2). It points to the possibility of the parallel involvement of other cytotoxicity
mechanisms, e.g., redox cycling of hydroxylamines (Eq. (8-10)) and/or their binding to
macromolecules. It is possible that the enhanced cytotoxicity of the amino metabolites of
TNT is caused by the formation of toxic hydroxylamines from amines due to Nhydroxylation catalyzed by cytochromes P-450.[20] The latter reaction may take place in FLK
cells, because we found that the cytotoxicity of NH2-DNTs was partly reduced by the
inhibitors of cytochromes P-450, -naphthoflavone and izoniazide. Another important point
is that the ability of compounds to inactivate TrxR and directly react with GSH apparently
did not enhance their cytotoxicity in FLK cells, i.e., the cytotoxicity of tetryl, pentryl, and
DNBF is in line with their redox cycling properties (Fig. 2).
330
Fig 2.
Relationship between the cytotoxicity of nitroaromatic explosives and

model nitroaromatics in FLK cells (cL50), and their reactivity towards
single-electron transferring flavoenzymes, FNR and P-450R.
Numbers of explosives (1-13) correspond to those in Tables 1,2. Other
nitroaromatic compounds: p-dinitrobenzene (14), o-dinitrobenzene (15),
p-nitrobenzaldehyde (16), m-dinitrobenzene (17), p-nitroacetophenone
(18), p-nitrobenzoic acid (19), p-nitrobenzyl alcohol (20), and nitrobenzene
(21).
331
4.
CONCLUSIONS
This work summarizes the data on the mammalian cytotoxicity of nitroaromatic

explosives and their metabolites obtained by us and several other groups. Our studies show
that the main cytotoxicity mechanism of explosives in FLK cells is their redox cycling with
concomitant oxidative stress. The identification of enzymes responsible for this mode of
cytotoxicity is currently under way. The mechanisms of cytotoxicity of explosives in other
cell lines are not disclosed so far. One may suppose that the relative role of the oxidative
stress may depend on the conditions of experiment, e.g., the time of incubation, and, more
importantly, on the amount of specific enzymes, e.g., P-450R, NQO1, cytochromes P-450,
and, possibly, glutathione-S-transferase, in particular cell lines. Nevertheless, the oxidative
stress will remain an important cytotoxicity mechanism of nitroaromatic explosives. The
prooxidant cytotoxicity of nitroaromatics correlates with their single-electron reduction
potential values, which may be obtained in pulse-radiolysis experiments. Our data imply
that the evaluation of redox properties of new explosives and high energy compounds in
enzymatic systems may be an alternative and simple approach in order to characterize or
predict their toxic properties.
Acknowledgement
This work was supported in part by Lithuanian Science and Studies Foundation and by
the NATO Collaborative Linkage Grant LST.CLG.978882.
REFERENCES
[1]
[2]
[3]
[4]
S. LETZEL, T. GOEN, M. BADER, J. ANGERER and T. KRAUS: Exposure to Nitroaromatic

Explosives and Health Effects During Disposal of Military Waste, Occup. Environ. Med., 60,
483-488, 2003, and references therein.
L. KONG, Q. JIANG and Q. QU: Formation of Superoxide Radical and Hydrogen Peroxide
Enhanced by Trinitrotoluene in Rat Liver, Brain, Kidney, and Testicle in Vitro and Monkey
Liver in Vivo, Biomed. Environ. Sci., 2, 72-77, 1989
K.H. LEUNG, M. YAO, R. STEARNS and S-H.L. CHIU: Mechanism of Bioactivation and
Covalent Binding of 2,4,6-Trinitrotoluene, Chem.-Biol. Interact., 97, 37-51, 1995
S. HOMMA-TAKEDA, Y. HIRAKU, Y. OHKUMA, S. OIKAWA, M. MURATA, K.
OGAWA, T. IWAMURO, S. LI, G-F. SUN, Y. KUMAGAI, N. SHIMOJO and S.
KAWANISHI: 2,4,6-Trinitrotoluene-Induced Reproductive Toxicity via Oxidative DNA
Damage by its Metabolite, Free Rad. Res., 36, 555-566, 2002
332
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
J. NORDBERG and E.S.J. ARNER: Reactive Oxygen Species, Antioxidants, and the
Mammalian Thioredoxin System, Free Rad. Biol. & Med., 31, 1287-1311, 2001
N. NAS, A. NEMEIKAIT-NIEN, E. SERGEDIEN, H. NIVINSKAS, .
ANUSEVIIUS and J. ARLAUSKAS: Quantitative Structure-Activity Relationships in
Enzymatic Single-Electron Reduction of Nitroaromatic Explosives: Implications for their
Cytotoxicity, Biochim. Biophys. Acta, 1528, 31-38, 2001
J. ARLAUSKAS, A. NEMEIKAIT-NIEN, . ANUSEVIIUS, L. MISEVIIEN, M.
MARTINEZ-JULVEZ, M. MEDINA, C. GOMEZ-MORENO and N. NAS: FlavoenzymeCatalyzed Redox Cycling of Hydroxylamino- and Amino Metabolites of 2,4,6-Trinitrotoluene:
Implications for their Cytotoxicity, Arch. Biochem. Biophys., 425, 184-192, 2004
J. ARLAUSKAS, A. NEMEIKAIT-NIEN, . ANUSEVIIUS, L. MISEVIIEN, A.
MAROZIEN, A. MARKEVIIUS and N. NAS: Enzymatic Redox Properties of Novel
Nitrotetrazole Explosives: Implications for their Toxicity, Z. Naturforsch., 59c, 399-404, 2004
A. NEMEIKAIT-NIEN, J. ARLAUSKAS, L. MISEVIIEN, . ANUSEVIIUS, A.
MAROZIEN and N. NAS: Enzymatic Redox Reactions of the Explosive 4,6Dinitrobenzofuroxan (DNBF): Implications for its Toxic Action, Acta Biochim. Polon., 51,
1081-1086, 2004
P.J. OBRIEN, W.C. WONG, J. SILVA and S. KHAN: Toxicity of Nitrobenzene Compounds
Towards Isolated Hepatocytes: Dependence on Reduction Potential, Xenobiotica, 20, 945-955,
1990
M.V. ORNA and R.P. MASON: Correlation of Kinetic Parameters of Nitroreductase Enzymes
with Redox properties of Nitroaromatic Compounds, J. Biol. Chem., 264, 12379-12384, 1989
N. NAS, . ANUSEVIIUS, D. BIRONAIT, G.I. BACHMANOVA, A.I. ARCHAKOV
and K. OLLINGER: The Electron Transfer Reactions of NADPH:Cytochrome P-450 Reductase
with Nonphysiological Oxidants, Arch. Biochem. Biophys., 315, 400-406, 1994
. ANUSEVIIUS, M. MARTINEZ-JULVEZ, C. GENZOR, H. NIVINSKAS, C. GOMEZMORENO and N. NAS: Electron Transfer Reactions of Anabaena PCC 7119
Ferredoxin:NADP+ Reductase with Nonphysiological Oxidants, Biochim. Biophys. Acta, 1320,
247-255, 1997
R.G. RIEFLER and B.F. SMETTS: Enzymatic Reduction of 2,4,6-Trinitrotoluene and Related
Nitroarenes: Kinetics Linked to One-Electron Redox Potentials, Environ. Sci. Technol., 34,
3900-3906, 2000
R.J. KNOX and S. CHEN: Quinone Reductase-Mediated Nitro-Reduction: Clinical
Applications, Methods Enzymol., 382B, 194-221, 2004
V. MIKINIEN, . ANUSEVIIUS, A. MAROZIEN, R. KLIUKIEN, H. NIVINSKAS, J.
ARLAUSKAS, N. NAS and K. BECKER: Tetryl as Inhibitor and Subversive Substrate
for Human Erythrocyte Glutathione Reductase, (Flavins and Flavoproteins 1999, S. Ghisla, P.
Kroneck, P. Macheroux and H.Sund, Eds.), Rudolf Weber, Berlin, p. 703-707, 1999
A. MAROZIEN, R. KLIUKIEN, J. ARLAUSKAS and N. NAS: Methemoglobin
Formation in Human Erythrocytes by Nitroaromatic Explosives, Z. Naturforsch., 56c, 11571163, 2001
B. LACHANCE, P.Y. ROBIDOUX, J. HAWARI, G. AMPLEMAN, S. THIBOUTOT and G.I.
SUNAHARA: Cytotoxic and Genotoxic Effects of Energetic Compounds on Bacterial and
Mammalian Cells in Vitro, Mutat. Res., 444, 25-39, 1999
M.E. HONEYCUTT, A.S. JARVIS and V.A. McFARLAND: Cytotoxicity and Mutagenicity of
2,4,6-Trinitrotoluene and Its Metabolites, Ecotoxicol. Environ. Saf., 35, 282-287, 1996
F.P. GUENGERICH: N-Hydroxylarylamines, Drug Metab. Rev., 34, 607-623, 2002
333
APPLICATION OF THE METHOD OF GRAY INTERRELATION

ANALYZING TO ASSESS HAZARD OF GAS EXPLOSION ACCIDENTS
SUN Bin
Dept. of Safety Defence, Zhejiang Police Institue, Hangzhou 310018, China
Abstract:
Based on safety engineering science principle and local investigation, with an aim at
situation of preventing gas, gas inherent hazard is put forward, and the gas gush quantity
and gas pressure are inherent. The change of gas gush quantity in a period is nonlinear,
gas inherent hazard is nonlinear and dynamic, the relation between gas inherent hazard
and gas lure hazard is analyzed by nonlinear gray interrelation method. Furthermore, risk
of gas hazard in roadway is analyzed with the gas gush quantity, the wind quantity and the
gas concentration, it is significance on spot.
Key words:
1.
gas inherent hazard, accident ris, gray interrelation
INTRODUCTION
At present, gas explosion accidents are one of disasteres happening frequently in coal mine,
causing effect badly. The reason for accident exists in many aspects, but essential reasons are
short of scientific identifaction and systematic elimination. Safety Production in Coal Mines,
Gas as God.. Thus the study on the rational method for risk rank and sequence of hazard for gas
explosion accidents is of realistic significance to improve the level of risk management and
safety benefits in mine. In risk assessment of hazard of gas explosion accidents, by the method
of gray interrelation analyses, forecasting the extent of risk for hazard of gas explosion
accidents in heading, the risk for hazard is nolinear, whose change course takes on gray, so gray
theroy is feasible for analysing hazard. In order to effectively prevent gas explosion accidents
from happening based on data on spot, select 215 tailentry, 412 tailentry etc. 4 pieces of
heading in Cheng Jia Shang Mine. According to index system of hazard risk assessment, the
paper applies the method of gray interrelation analyzing to assess hazard of gas explosion
accidents in heading.
334
2.
NONLINEAR CHARACTERANALYSE FOR GAS INHERENT HAZARD

This paper takes gas inherent hazard as studying object, analysing characteristic of gas
inherent hazard. In high gas mine, a quantity of gas is naturally released from coal seam. Gas
inherent hazard, namely eigenvalue describing gas characteristic itself, can use gas gush quantity,
gas concentration, gas content and gas pressure etc.expresssed Change of gas gush quantity in an
observed period takes on nonlineaar, substantially, and gas inherent hazard has the characteristic
of nonlinear and dynamical. Accumulation and even exceeding limit of gas, which can arouse
explosion and suffocation accident at last, which considerablely rest with human factor, such as
macthed wind quantity etc.. Gas educing hazard, for instance, wind quantity and wind velocity.
Actually, wind quantity and wind velocity have an impact on gas concentration. The extent of
gas educing hazard impacting on gas inherent hazard and the relation between gas educing
hazard and gas inherent hazard need be urgently confirmed. At present, owing to the limited
technology, changing trend of gas gush quantity and sudden increase of gas gush quantity etc.
have not been scentificly known, further more, gas inherent hazard is involved in many factors,
so hazard of gas explosion accidents is very complex and varied. In fact, gas inherent hazard
has connection with ventilation system, viz. part fan wind quantity, wind velocity, geography
condition in heading, personnel and management. The change trend of gas inherent hazard
factor shows nonlinear, for example, change of gas gush quantity to be nonlinear in an observed
cycle. At the present time, it is difficult to predict scientificly risk extent of gas inherent hazard.
Direct reason of gas accindent leading to severe economy losses is that gas concentration exceeds
limit. In fact, Wind quantity is gas educing hazard. Gas concentration has connection with wind
quantity. With the method of gray interrelation, the author forecasts the extent of risk for hazard
of gas explosion accidents, and analyses the relation between gas inherent hazard and gas
educing hazard, investigating fine compounding concern among gas gush quantity, wind quantity
and gas concentration. By forecasting risk extent of gas inherent hazard, we can enhance the
ability to prevent gas explosion accidents, to the greatest extent, reduce accident losses resulting
from gas hazard. In the past, how to match wind quantity and adjust gas concentration almost
rely on experience. As a matter of fact, gas inherent hazard and gas educing hazard make up of
typical gray system, so forecasting gas gush quantity can adopt gray methmatics. This paper applies the
Method of Gray Interrelation to study interrelation degree among gush quantity, wind quantity
and gas concentration, analysing risk extent of hazard of gas explosion accident, in order to prevent
effectively accident from happening.The reasons for gas explosion accidents are quite complicated.
In fact, in heading of mining area, interaction of natural condition factor gas condition factor
and ventilation condition factor etc. can influence frequency and extent of gas explosion
accidents, moreover natural condition change may affect operator, as well as manipulation
defect and machine fault, which are potential reasons. Further more interior connection is also
335
fuzzy and uncertain, no single physics prototype and mathematics model, and mutual relation,
extent and data collection connform to gray system characteristic. Therefore, hazard of gas
explosion accidents is also gray system. The theroy of gray system get over Probability Theory
imprecise, thus evaluating risk extent of hazard of gas explosion accidents in heading with the
method of gray interrelation, which can effectively predict risk of hazard. But quantitative
analysis of hazard still need more perfection.
3.
THE ANALYSES FOR GRAY INTERRELATION OF HAZARD DATA

3.1
The method of calculating gray interrelation modulus

and gray interrelation degree
In general, data disposal can adopt initial disposal, equal disposal and section disposal etc.
Provided that accident and influencing factor are the function of time, changing along with time.
Take X 0 as the first sequence, X 1 X 2 X 3 X n representing influencing factor of hazard, then
defining relating modulus between X 0 and X i as follow :
0i ( k ) =
0i ( k ) =
min + max
0i (k ) + max
(1)
min min X 0 (k ) X i (k ) + max max X 0 (k ) X i (k )

i
X 0 (k ) X i (k ) + max max X 0 (k ) X i (k )
i
(2)
in formula (1) and (2) : 0i (k ) as k st relating modulus, as differentiating modulus,

0 1, commonly chosing =0.5--1 in this paper =1, max , min as max and min of
absolute difference in each time. Clearly, 0i (k ) is the function about time. Take average of
0i (k ) , If we can educe relating degree which reflects related extent in whole course
between X 0 and X i , namely

r0i =
1
n
(3)
0i ( k )
k =1
compare r0i with r0 j (ij), if r0i > r0 j , indicating i st factor to accident more extent than
j st
factor.
3.2
Calculating method about slope interrelation degree
In above sequence, often appearring zero, we can adopot calculating method of slope
interrelation degree or noninitial disdpopsal, but calculating interrelation degree, in former
literature, often adopt initial disdpopsal or slope interrelation degree or applying gray
interrelation analyzing. On earth which means calculated result reasable or not, only by
checking relating order to judge.
336
Calculating principle about slope interrelation degree

(k ) =
1
x(t )
y (t )
1+
x(t ) y(t )
t {1,2,L, N }
(4)
relating degree between two sequeences

r=
4.
1
N 1
N 1
(t )
t =1
t={1,2,,N-1}
(5)
APPLIED CASE OF ANALYZING HAZARD IN HEADING

Select 3 pieces of heading and standard heading, comparing hazard risk extent. Compare
standard heading with 3 pieces of heading about wind quantity gas concentration and gush
quantity in 3 calsses on one day, analyzing risk extent of hazard in heading. Supposing
A 0 K
as standard series, A 1 K A 2 K A 3 K as compared series. based on
formula (1)(3), analyzing Tab.1 data, results after calculation as Tab.2Tab.5.

Table 1. Assessed normal data series and refered data series
1st class
index
A 0
A 1
A 2
A 3
K
K
K
K
K
2nd class
wind
gas
gush
quantity concentration
quantity
m3/min
1
740
490
600
590
%
2
0.75
0.62
0.6
0.8
m3/min
3
3.03
2.96
3.6
4.72
3rd class
wind
gas
gush
quantity
m3/min
4
740
510
678
610
%
5
0.75
0.58
0.7
0.76
m3/min
6
3.03
2.96
4.05
4.64
wind
gas
gush
quantity
m3/min
7
740
510
606
594
%
8
0.75
0.62
0.72
0.76
m3/min
9
3.03
3.16
4.36
4.51
Yield
on day
10
3448.6
3671
3385
4529
Table 2. Data after initialized disposal

1
1.2231
0.8099
0.9917
0.9725
2
1.0830
0.8953
0.8664
1.552
3
1.7447
1.0496
1.2766
1.6738
4
1.2141
0.8368
0.9483
1.0008
5
1.0753
0.8315
1.0036
1.0896
6
0.5256
0.8065
1.1035
1.2643
7
1.2082
0.8327
0.9894
0.9698
8
1.0526
0.8702
1.0105
1.0667
9
0.8048
0.8393
1.1580
1.1979
10
0.9176
0.9767
0.9006
1.2050
7
0.3755
0.2188
0.2384
8
0.1824
0.0421
0.141
9
0.0345
0.3532
0.3931
10
0.0591
0.017
0.2874
Table 3. Grade margin data after calculation

1
0.4132
0.2314
0.2060
2
0.1877
0.2166
0.0722
3
0.6951
0.4681
0.0709
4
0.3773
0.2658
0.2133
5
0.2438
0.0717
0.0143
6
0.0191
0.2779
0.4387
337
Table 4.
0i (1)
0.0639
0.7655
0.7499
Interrelation modulus results after calculation
0i (2)
0.8034
0.7779
0.9243
Table 5.
0i (3)
0.6613
0.5101
0.9258
0i (4)
0.6613
0.7381
0.7807
0i (5)
0.7554
0.9249
0.9249
0i (6)
0.9930
0.7289
0.6255
0i (7)
0.8482
0.7760
0.7760
0i (8)
0.8082
0.9602
0.8482
0i (9)
0.9720
0.9709
0.9709
0i (10)
0.9403
0.9959
0.7218
Interrelation degree results after calculation
R 0 (1)
0.7748
R 0 (2)
R 0 (3)
0.82498
0.79874
From Tab. 5 Interrelation degree results after calculation, ranking sequence, risk extenet of
3 pieces of heading as follow :
R
(2) R 0 (3) R 0 (1)
Interrelation degree more, showing more close to prescriptive standard, therefore safety
situation better. So 2 # heading gas hazard risk extenet least. 1# heading gas hazard risk extenet
most.
5.
CONCLUSIONS
Gas gush quantity and gas pressure take on characteristic of gas inherent hazard, and Gas
gush quantity in a cycle, its change is nonlionar, wind quantity is gas inducing hazard, the
relation between gas inducing hazard and gas inherent hazard can be analysed with the method
of gray interrelation that adopts gas gush quantity, wind quantity and gas concentration etc..
This research indicates that, based on data on spot, owing to scientificly using The Method of
Gray Interrelation to analyse risk extent of 4 pieces of heading in Cheng Jia Shang Mine, result
of risk assessment offers foundation and consult to identification and control of hazard of gas
explosion accidents.
Throuth the reaserch on application of the method of gray interrelation to assess hazard of
gas explosion accidents, this paper discusses how to assess risk extent of hazard, depending on
gas gush quantity, wind quantity and gas concentration etc., and applies the Method of Gray
Interrelation to caculate interrelation degree. According to interrelation degree we know
difference between each index and standard index, to predict risk extent of hazaed accordingly,
scientificly and effectively to enhance the level of forecast, prevention and control of hazard of
gas explosion accidents. So applying the method of gray interrelation, may much more
factually manifest risk extent of hazard of gas explosion accidents in heading, prone to judge.
338
The Method of Gray Interrelation is an important method of identifying and removing

factors about hazards of gas explosion accidents, however how to assess impersonally and
actually hazard of gas explosion accidents, in time affirm currently authentic level of risk over
system of dynamics and complexity, namely hazard of gas explosion accidents in headings, so
as to achieve advancing veracity and reliability of assessment over hazard, still more research.
REFERENCES
[1]
WU ZONG-ZHI, GAO JING-DONG : Identification and Control of The Major Hazard [M].
Metallurgy Industy Presss 2001. 15-25
[2]
SHI SHI-LIANG,WANG HAI-QIAO.: Safety Assessment of Nonlinear Dynamics in Mine [M].

CoalMetallurgy Industy Presss 2001
[3]
LIU XIN-TANG,WU XIAO-YAN. Modern Times System Model and Emulation Tecohnology [M].
Northwest Politechnolic University Press 2001 8. 82-109
[4]
TIAN SHUI-CHENG: Study on Identification and Control of The Third Class of Hazard. [Doctor
thesis]. Beijing Institue of Technology, 2001 7. 36-39
[5]
CHENG WUI- MING. Gas Powder State safety assessing System and Application. [Doctor
thesis].China Mining University, 1997 7.
[6]
SUNBIN: Study on Risk Assessment of Gas Accidents in Coal Mine Based on Theroy of Hazard.[
Master thesis]. Xian University of Science and Technology, 2003 5. 43-52
[7]
SUN BIN TIAN SHUI-CHENG CHANG XIN-TAN.: Study on Risk Assessment and Risk
Management for Accident [J] China Mining, 2003 Vol.12 No.1 71-73
[8]
R.L.BRUCE, A.MINTY, C.A.J.GREGORY: Holistic Risk Management [J].

Assessment, Melchers & Stewart 1995
339
Intergrated Risk
DESIGN OF PARTICULATE ENERGETIC MATERIALS

Ulrich Teipel, 1and Ulrich Frter-Barth
Fraunhofer Institut fr Chemische Technologie (ICT)
Joseph-von-Fraunhofer-Strae 7, 76327 Pfinztal, Germany
Abstract:
The crystal quality and the internal microstructure of crystals have a great influence on
the sensitivity of energetic materials. Besides, the particle size and the particle size
distribution are of great importance to the processing technology of energetic materials.
Particle properties can especially be influenced by applying different crystallization
techniques, such as cooling crystallization, membrane crystallization, emulsion
crystallization and others.
The goal of the investigations was to determine the interrelationship between the
properties of the gained crystals and the process parameters. Special attention was
directed to the qualitative and quantitative examination of crystal defects and their
dependence on the experimental conditions. Besides, the morphology and structure of
crystals were calculated by molecular modelling. The effect of crystal defects on the
sensitivity of the material was tested on different collectives of particles having varying
amount of crystal defects.
Keywords:
1.
crystal quality, microstructure, morphology, sensitivity, crystallization.
INTRODUCTION
Particulate energetic materials and systems produced with such particulates, such as
solid propellants, propellant powders and explosives, have assumed an ever more important
role in industry. Important material properties of these materials, including burn rate
behavior and sensitivity, are decisively affected by particulate properties such as particle
morphology, size and size distribution.
Particles are generally manufactured using well-known solids formation techniques such
as crystallization, precipitation, size-reduction, spray drying or deagglomeration [1].
Although a large amount of information exists on these different processing techniques,
there are still gaps in our understanding of how to completely control the particle formation
process or reliably predict processing phenomena and the properties of the particulate
products.
A number of subjects related to particle technology and the design of particulate
products require further investigation, including the characterization and experimental
determination of various particle properties, the simulation and modeling of process
operations involving dispersed products and, in particular, the relationship between product
properties and the properties of dispersed systems and particles.
Ulrich.Teipel@ict.fhg.de
340
2.
PRODUCT DESIGN OF PARTICLES
The goal of product design is to develop end products with a defined set of properties
that can be economically commercialized in an environmentally acceptable manner [2, 3]. The
design of particulate products and dispersed systems is complex because obtaining the
desired end product properties requires achieving an optimal combination of the chemical
and physical properties of each component of the system. An essential problem in product
design is to selectively and reproducibly achieve the desired material characteristics that
meet the demands required of the particulate product or dispersed system under
development. Such characteristics may include:
insensitivity
defect free particles
absence of gas or fluid inclusions
high density
high purity
surface and adhesion properties
wettability, particle-binder interactions
high packing density
high performance
rheological properties
storage stability
low propensity to form fine dusts
filterability
low tendency to agglomerate
Many of these product material properties are currently optimized empirically based on
process and product requirements. The product properties are strongly dependent on the
physical properties of the incorporated powder and on the dispersity characteristics of the
system. The functional relationship between the product properties and the dispersity
characteristics for chemically identical products can be described via the property function:
i = f ( j )
(1)
The dependence of the product properties i on the dispersity charactericstics j is

expressed via this equation.
341
Product material
properties
Property
function
Application oriented
product properties
e.g., size,
structure,
morphology
Process
function
Dispersity
properties
Processes,
methods
processes
Fig 1.
Product design in particulate technology.
Important dispersity properties for particulate materials include:
particle size
particle size distribution
morphology
polymorphism
crystallinity
inner structure, porosity
particle density, bulk density
One of the main challenges in the field of particle technology is attaining a better
understanding of the relationship between the product properties i and the dispersity
properties for different types of products and product applications.
By considering the process function, particles with defined dispersity properties can be
manufactured using a variety of processing technologies, including methods such as
crystallization or spray drying.
The dispersity properties of particulate products represent, through the property and
process functions, a connection between the feedstock, the chosen processing method and
the required application oriented product properties of the end product (see Fig. 1). Besides
appropriate chemical design of the feedstock, when designing particulate products one or
more suitable processing methods with appropriate process parameters must be selected and
the required dispersity properties must be set to create an end product with the required
material properties.
342
3.
CRYSTAL DEFECTS AND FLAWS
In addition to hot spots in the binder [4,5] and interactions between particles and binder,
another property that significantly influences the sensitivity of energetic systems is the
crystal quality of the particulates. When describing the product quality of particulate
materials, one distinguishes between flaws within the crystal [6-10] and the crystals surface
properties. Inner flaws that can significantly impair the crystal quality include point defects
(Schottky or Frenkel defects), screw and edge dislocations, two dimensional defects (e.g.,
twin formation) and three dimensional defects (e.g., fluid or gas inclusions).
Defects are essentially a deviation in the perfect periodicity of a crystal. In an ideal
crystal all the atoms assume a configuration corresponding to the global energy minimum.
However, a real crystal does not exhibit a global energy minimum, but instead adopts
configurations corresponding to numerous local energy minima. Because crystal defects
have a significant effect on the physical properties of a crystal (e.g., the internal stress
distribution), they also have a significant influence on the sensitivity of energetic materials.
Figure 2 shows an example defect in the FOX 7 crystal.
Fig 2.
Defect in a FOX 7 crystal.
Because crystal defects are related to the crystals formation history, optimizing the
crystallization method offers the possibility to minimize crystal defects. Figures 3 and 4
show FOX 7 crystal surfaces; surface defects are clearly visible in the crystal depicted in
Fig. 3. In contrast, Fig. 4 shows crystal growth layers of FOX 7 produced using optimized
crystallization conditions.
343
Fig 3.
Fig 4.
Surfaces of FOX 7 crystals.
Crystal growth layers (FOX 7).
At high suspension concentrations and local regions of high supersaturation, nucleation

may occur in liquid bridges. After crystallization, this causes the transition to solid bridges
(see Fig. 5) and the agglomeration of particles.
344
Fig 5.
Solids bridges.
Subjecting the agglomerated crystals to mechanical energy (e.g., through processing)

leads to their deagglomeration and can also produce crystals with sharp edges (as also occurs
in particle grinding), a condition which leads to a clear increase in the materials sensitivity.
4.
PRODUCT DESIGN VIA CRYSTALLIZATION
The starting point for many particle formation processes is a fluid phase sufficiently
oversaturated with an active component. Transport processes, chemical reactions and other
processes cause nucleation, which then leads to the formation of clusters. Typically,
nucleation is categorized either as primary (both homogeneous and heterogeneous) or
secondary. During suspension crystallization, nucleation is followed by crystal growth of the
dispersed phase in the oversaturated solution. Besides the decisive particle growth stage,
critical steps in the product design process include: homogenization/stabilization of the
dispersed system, or suppressing aggregation of the dispersed phase, depending on the
specific application requirements of the end product.
achieving
oversaturation
nucleation
crystal growth
aggregation
homogenization
stabilization
Fig 6.
Product design via crystallization.
345
4.1
Particle design via molecular modelling
Besides particle size and particle size distribution, other important dispersity properties
include morphology, polymorphism and crystal structure. The morphology and structure of
crystals are largely determined by the type of interactions that occur between the crystal
surface and the surrounding fluid [11]. In the last few years, computation programs have been
developed and become commercially available that allow the morphology of various
crystalline materials to be calculated. The inner crystalline structure and distance and
symmetry positions of the elementary cells are known exactly for many energetic materials.
There are various theories for describing the relationship between the inner crystalline
structure and outer crystal form, such as the Bravais-Friedel-Donnay-Harker model [12],
which allows the crystal structure to be pre-calculated based on pure geometry, and the
Hartman-Perdok method [13], which, besides geometrical parameters, includes the effect of
energetic interactions. Another method of modelling the crystal structure is the Attachment
Energy method [14, 15, 16]. The surface deposition energy EAn is the energy released when a
layer deposits onto the crystal surface.
E An = E Kri E Sch
(2)
Here, EKri is the lattice energy and ESch is the energy of a layer of infinite thickness. The
growth rate of the crystal surface is directly proportional to the surface deposition energy
EAn, so that surfaces with low deposition energy grow more slowly and therefore have the
greatest influence on the crystalline morphology.
The environment in which crystallization takes place (i.e,. the solvent) has a decisive
influence on the crystal habit. Use of different solvents leads to formation of various
morphologies of a material. This influence on the crystal growth process is still not fully
understood or amenable to modelling. A well-known method from the literature describes
the incorporation of solvent into the crystal lattice structure (Build-up method; see Fig. 8).
One newly developed method, instead of simulating the substitution of molecules at the
crystal lattice points, describes the accumulation of solvent on the crystalline surface and
calculates the crystal habit on the basis of interactions between the crystal surface and
solvent (see Fig. 7).
E = E Krist + E Lsungsm. E Krist + Lsungsm.
(3)
This method allows one to more realistically simulate the growth of the crystal habit and
its dependence on the solvent. Figure 8 shows an example of the solvent nMPs influence on
FOX 7 crystals.
346
FOX7
Fig 7.
nMP
nMP-molecules
on the FOX 7 (020) surface
Accumulation method.
Build-up method
Fig 8.
4.2
Accumulation method
Solvent influences; Fox7/nMP.
Crystallization of 1,1 diamino-2,2-dinitroethylene (FOX 7)
Cooling crystallization was used to recrystallize FOX 7. Example results using nMP as
the solvent are presented here. It is clear from Fig. 9 that due to its various surfaces, FOX 7
crystallizes at a range of different crystallization rates and surface defects form on those
crystal surfaces in which the growth process is hindered.
Fig 9.
347
FOX 7 crystal.
All of the following experiments were conducted using a cooling gradient dT/dt of
10 K/h. Pure nMP and an nMP/water mixture were employed as solvents. It was found that
using pure solvent and a certain portion of water as antisolvent yielded crystals ranging in
size x50.3 from 40 to 50 m. Using solvent mixtures, particles with sizes ranging from 100
m x50.3 400 m can be produced via cooling crystallization.
x50.3 = 55 m
x50.3 = 306 m
Fig 10. FOX 7 crystals.
4.3
Crystallization of insensitive HMX
Using a modified cooling crystallization process, insensitive HMX (ICT-i-HMX) was

produced at the Fraunhofer Institute-ICT. Figure 11 shows the original HMX and ICT-iHMX.
Fig 11. HMX
ICT-i-HMX
348
Using this special crystallization process, HMX with a significantly higher particle
density was produced. The higher density is directly attributable to the reduced number of
inner crystal defects in the molecules. Model PBX test slabs were produced using this HMX
and subjected to gap tests. The ICT-i-HMX exhibits an insensitivity nearly a factor of two
better than conventional HMX based on the shock wave loading test (Fig. 12).
14
recrystallized HMX
Original-HMX
12
gap / mm
10
8
6
1.885
1.89
1.895
1.9
1.905
density /g/cm3
Fig 12. Gap test results, HMX and ICT-i-HMX.
4.4
Product improvement using ultrasound to initiate nucleation
The use of ultrasound to initiate precipitation leads to a significant improvement in the

quality of the end product and a narrower particle size distribution [17, 18]. Primary
homogeneous nucleation is a process that occurs under thermodynamically unstable
conditions. For many materials, a solid phase forms only under conditions of
supersaturation. The point at which nucleation begins depends on the concentration of
foreign material, the type of secondary material and the amount of mechanical energy
applied to the system. When ultrasound is used during the cooling crystallization process, the
onset temperature for nucleation is nearly independent of the temperature gradient and
significantly higher than in the absence of ultrasound. In addition, the particle size
distribution is significantly narrower when nucleation is initiated via ultrasound. The particle
size distribution can be described in terms of the distribution coefficient i:
i =
x84,3 x16,3
(4)
2 x50,3
where xi,3 is the particle size for a volume sum distribution of i%.
4.5
Particle manufacture via the membrane crystallization process
Membrane crystallization is a new crystallization process that offers an interesting

alternative to cooling or steam crystallization, especially for energetic materials [19, 20, 21].
This method is especially interesting for use with thermolabile products, because the process
can be carried out at low temperatures. In addition, the energy balance for membrane
349
crystallization is considerably more favorable compared to cooling or steam processes. Like

other processes, membrane crystallization occurs in an oversaturated solution, in which the
solvent is transported through a semi-permeable membrane, thereby becoming more
concentrated at a nearly constant process temperature. The membrane must be chosen so that
it provides a selective barrier to the active ingredient while allowing the solvent to penetrate
the membrane (see Fig. 13). Membrane processes are typically pressure driven methods,
such as reverse osmosis or micro or nano-filtration [22]. After transport of the solvent through
the membrane to form a highly concentrated fluid phase, nucleation begins and is followed
by the desired crystal growth stage.
Membranmodul
Kristallsuspension
Feed
Permeat
Fig 13. Membrane crystallization.
5.
PRODUCT DESIGN OF SPHERICAL AMMONIUM DINITRAMIDE

(ADN)
Emulsion crystallization is particularly suitable for producing spherical particles of

materials that melt at temperatures accessible with conventional processing technologies.
The emulsion crystallization process consists of two linked process steps. The first step is
production of an emulsion consisting of molten material dispersed in an inert continuous
phase. The droplet size of an emulsion can be controlled by choosing the appropriate
emulsification method and equipment. Various dispersion and size reduction processes exist,
using equipment such as rotor-stator mixers (gear rim dispersers and colloid mills) or static
mixers.
The second step in emulsion crystallization, which is decisive to the process, is the
transition of the fluid dispersed phase into solid particles. For materials with particularly
sluggish nucleation behavior, in addition to application of a sharp temperature gradient,
some combination of mechanical energy input, particle-particle and/or particle-wall
interactions or addition of seed crystals is necessary to initiate the nucleation process. The
possibility of producing ammonium dinitramide (ADN) via the emulsion crystallization
process is demonstrated in Fig. 14.
350
a. Synthetic product
b. Emulsion crystallization product
Fig 14. Crystallization of ammonium dinitramide.

Figure 14a shows the product after synthesis and 14b depicts the spherical ammonium
dinitramide after the emulsion crystallization process [23]. Afterwards, the continuous and
dispersed phase of the suspension are separated and the product is dried.
A process newly developed at the Fraunhofer ICT offers the possibility of designing
stabilized ADN particles. During crystallization of molten ADN, stabilator particles are
incorporated into the ADN. Figure 15 shows the weight loss as a function of time for ADN
immediately after synthesis and for conventional ADN particles and ADN particles
stabilized via the ICT process. It is clearly evident that ADN particles produced using the
new process exhibit significantly less weight loss in these experiments.
Fig 15. Weight loss function of ADN.
351
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
U. TEIPEL: Production of Particles of Explosives, Propellants Explosives and Pyrotechnics, 24,

p.134 139, 1999
U. TEIPEL: Produktgestaltung in der Partikeltechnologie, Fraunhofer-IRB-Verlag, Stuttgart,
2002
E. Favre, L. Marchal-Heusler, M. Kind: Chemical Product Engineering: Research and
Educational Challenges, Trans IChemE, 80, p. 65 74, 2002
R. W. Armstrong, H. L. Ammon, W. L. Elban, D. H. Tsai: Investigation of hot spot
characteristics in energetic crystals, Thermochimica Acta, 384, p. 303 313, 2002
R. W. ARMSTRONG, C. S. COFFEY, V. F. DE VOST, W. L. ELBAN: Crystal size dependence for
impact initiation of cyclotrimethylenetrinitramine explosive, Journal of Applied Phys., 68,
p. 979 983, 1990
K. A. GROSS: Liquid Inclusions in RDX Crystals, Journal of Crystal Growth, 6, p. 210 212,
1970
L. BORNE, J.-C. PATEDOYE: Quantitative Characterization of Internal Defects in RDX
Crystals, Propellants Explosives and Pyrotechnics, 24, p. 255 259, 1999
R. JANSSEN- VAN ROSMALEN, P. BENNEMA: The role of hydrodynamics and supersaturation in
the formation of liquid inclusions in KDP, Journal of Crystal Growth, 42, p. 224 227, 1977
A. E. D. M.VAN DER HEIJDEN, W. DUVALOIS, C. J. M. VAN DER WULP: Micro-inclusions in
HMX Crystals, Proc. 30th Int. Annual Conference of ICT, Karlsruhe, p. 41/1 41/10, 1999
H. H. CADY: Growth and Defects of Explosives Crystals, Proc. Material Research Society
Symposium, 296, Pittsburgh, p. 243 254, 1993
I. FUHR, H. KRBER, U. TEIPEL: Molecular Modelling: Produktgestaltung in der
Kristallisation (Produktgestaltung in der Partikeltechnologie, U. Teipel), Fraunhofer IRBVerlag, Stuttgart, p. 275 285, 2002
J. D. H. DONNAY, D. HARKER: A New Law for Crystal Morphology Extending the Law of
Brovais, Am. Mineral., 22, p. 446 467, 1937
P. HARTMAN, W. G. PERDOK: On the relation between structure and morphology of crystals,
Acta Cryst., 8, p. 49 -52, 1955
P. HARTMAN, P. BENNEMA: The Attachment Energy as a Habit Controlling Factor, 1.
Theoretical Considerations, J. Grystal Growth, 49, p. 145 156, 1980
Z. BERKOVITCH-YELLIN: Toward an ab initio derivation of crystal morphology, J. Am. Chem.
Soc., 107, p. 8239 8253, 1985
R. DOCHERTY, G. CLYDESDALE, K. J. ROBERTS, P. BENNEMA: Application of Bravais-FriedelDonnay-Harker, Attachment-Energy and Ising Models to Prediction and Understanding the
Morphology of Molecular Crystals, J. Phys. D: Appl. Phys., 24, p. 88 99,1991
U. TEIPEL, I. MIKONSAARI, H. KRAUSE, J. ULRICH: Influence of ultrasound on potassium alum
crystallization, Proc. 4th World Congress on Particle Technology, Sydney, 2002
I. MIKONSAARI, U. TEIPEL, J. ULRICH: Kristallisation unter Anwendung von Ultraschall,
(Produktgestaltung in der Partikeltechnologie, U. Teipel), Fraunhofer IRB-Verlag, Stuttgart
p. 255 261, 2002
U. TEIPEL, U. FRTER-BARTH: Verfahren zur Kristallisation aus Lsungen, EP 1 140 313 B1
U. TEIPEL, U. FRTER-BARTH: Method for producing crystals from solutions, US 6.616.847
B1
E. CURCIO, A. CRISCUOLI, E. ORIOLI: Membrane Crystallizers, Ind. Eng. Chem. Res., 40,
p. 2679 2684, 2001
R. RAUTENBACH: Membranverfahren: Grundlagen der Modul- und Anlagenauslegung,
Springer Verlag, Berlin, 1996
U. TEIPEL, T. HEINTZ, H. KRAUSE: Crystallization of Spherical Ammonium Dinitramide (ADN)
Particles, Propellants Explosives and Pyrotechnics, 25, p. 81 85, 2000
352
REGARDING THE DETERMINATION OF EXPONENT

FROM BURNING RATE LAW OF POWDER,
USING THE CLOSED BOMB DATA
Titica Vasile, Cristian Barbu, and Doru Safta
Military Technical Academy, 81-83 George Cosbuc Avenue, Bucharest, Romania
Abstract
In this paper it is presented a mathematical model of firing phenomenon from a ballistic
system. In this model is used an expression for burning rate law of powder, in which the
exponent has a variable value. This exponent is calculated using the curve of gases
pressure versus time obtained at the burning of powder in closed bomb. On the base of
the mathematical model was elaborated an interior ballistic soft, which allows to study
the variation of the powder gases pressure and the projectile velocity versus its
displacement inside of barrel and versus time. For an extant ballistic system, the
theoretical results, obtained with the aid of this soft and the experimental data are
compared.
Keywords:
closed bomb, burning rate law, firing phenomenon, ballistic system
1. INTRODUCTION
In the proposed mathematical model, the fundamental problem of interior ballistics it is
solved on the base of physical law for powders burning, which reflects better the reality of
firing phenomenon in comparison with geometrical law [1, 2, 3, 4].
Fig. 1. Pressure in closed bomb versus time.

In this paper, the index from the burning rate law of powder is variable. It is
determinated as a function of the pressure using the pressure in closed bomb P() , which are
presented in Fig. 1.
The elaborated mathematical model is consisted of the differential and algebraical
equations. By solving the differential and algebraical equations system, in numerical way
353
with the aid of the computer, it is obtained the variation of gases pressure and projectile
velocity versus its displacement within the barrel and versus time since the beginning of
powder burning, which defines the content of fundamental problem of interior ballistics.
The signification of the used parameters in this paper is that established in specialty
literature [1, 2, 3, 4].
2. MATHEMATICAL MODEL
The elaborated mathematical model is based on the main differential and algebraical
equations of interior ballistics [1, 2, 3] , thus: the fundamental equation of interior ballistics; the
equation of projectile translation motion; the burning rate law; the rate of gases forming.
The differential equation of displacement is obtained with the aid of the fundamental
equation of interior ballistics and of the equation of projectile translation motion, in
following form:
dl
l +l
q
f
vdv
(1)
2
+ v
2
vlim
where:
f
l = l 0 (a b ) ; a = 1 ; b = 1 ; = 1 1 / + 1 .
For the obtaining of velocity differential equation, it is admitted that in closed bomb and
in gun has been burned the same thickness of powder grain and it is took into account the
equation of projectile translation motion, thus:
dv
dI
= s P
q p
(2)
The differential equation for time has got the following form:
dt = dl .
(3)
These equations have been adequately transformed in order to allow the utilizing of
experimental data, obtained during the powder burning in closed bomb, at the solving of
fundamental problem of interior ballistics.
So, it is adopted as an independent variable , defined as ratio between impulse value of
powder gases pressure at certain moment I and impulse value of powder gases pressure at
the fragmentation moment I s , in the case of multiperforated powders. In the case of simple
shape powders, impulse I s is changed with impulse I k , i.e. the impulse at the end of
powder burning.
In order to increase the level of generality of these equations are utilized following
dimensionless variables:
= v ; = l ; = t ,
v lim
where:
l0
t lim
354
2 f
; t lim
q
vlim =
q v lim
.
s P
Utilizing the dimensionless variables, the equation (1) becomes

d = 2 E d ,
d
d
(4)
where
E=
a b +
2
Taking into account that dI = I s d and v = vlim , the equation (2) gets the form
d
=
d
B P

2 p
(5)
The differential equation for time (3), after the utilizing of the dimensionless variables
and the changing of d d with its expression from equation (4), becomes
B P

2 p
l0 E
d
=2 1
d
t lim v lim
(6)
After the introducing of the dimensionless variables in the equation of projectile

translation motion, this gets the form
p=
f
.
E
(7)
It is eliminated the pressure from equations (4), for displacement, from (5), for velocity,
from (6), for time and, finally, it is obtained:
d
=
d
2 B E EP
f
d
=
d
B EP

2 f
(8a)
(8b)
I
d
= 1 s EP ,
d t lim P f
(8c)
p=
f
E
(8d)
In order to determine the relation for index from burning rate law, is particularised the
logarithmic expression of burning rate law for two charge densities obtaining one system of
two equations with two unknown variable. From this equations system it can be obtained the
relation for index in the form
355
log u 2
log u1
log P2
log P1
(9)
If it is considered that has been burned the same thickness of powder grain for two
charge densities, than the relation (9) can be written thus:
= 1
dI
log 2
dI1
P
log 2
P1
d1
d 2
P
log 2
P1
log
(10)
The differential and algebraical equations system (8) can be used in first period and as
well as second period of the firing phenomenon.
The differential and algebraical equations system of projectile motion within grooved
barrel (8) represents the new mathematical model which allows to study the variation of
gases pressure and projectile velocity versus its displacement and versus time in conditions
closer of the reality of the firing phenomenon.
3. INTEGRATION OF EQUATIONS SYSTEM

The differential and algebraical equations system (8) was integrated using the fourth
order Runge-Kutta method, with the aid of following initial conditions:
= 0 0 ; = 0 = 0 ; = 0 = 0 ; = 0 0 ,
in which:
0
I
0 = 0 = 0
s
Is
Pd
Pd
; 0 = 0 = 2.303
t lim
t lim
I k (1. )
p
log( 0 ) .
f
pa
The first period has got two phases, in the case of multiperforated powder grains. In the
first phase, from the moment of the inflammation of powder until the moment of breaking of
powder grains, when s and 1 , it is integrated equations system (8).
In the second phase, from the moment of powder grains breaking until the moment of
the end of powder charge burning, when s 1 and = 1 , the system (8) becomes:
d
=
d
2 B E ,
(11a)
d
=
d
B
,
2
(11b)
d
d
p=
1 Is E
tlim f
(11c)
f
.
E
(11d)
356
In the equations of this system is not included the pressure P from closed bomb.
However, the solving of this system is done with the aid of the curve P( ), because the
values for from E relation are calculated on the base of experimental data.
In the second period, from the moment of the end of powder charge burning until the
moment when projectile leaves the barrel, = 1 and = 1 , the expression for E becomes
E = a b +2 .
1
The initial conditions for the integration of differential and algebraical equations system
(11) are just values of ballistic magnitudes at the moment of end of powder burning.
If the solving is effectuated for the powder that burns without breaking and in the
conditions of the burning rate law u = u1P , in which = 1 , than it is utilized the system
(11), taking into consideration the particularities of the second phase of the powders that
burn with the breaking of grains.
4. RESULTS AND CONCLUSIONS

In the Fig.2 is presented the variation of index in function of burned powder fraction
during of powder burning. It is observed that this variation tends to one.
niu
1,2
1
0,8
niu
0,6
0,4
0,2
0
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
Fig. 2. Variation of index versus burned powder fraction
In Fig.3 and Fig.4 it is presented the variation of gases pressure* and projectile velocity*
versus its displacement* and time*, obtained with the aid of the interior ballistics soft that
was elaborated on the base of proposed mathematical model, for an extant ballistic system.
357
The magnitudes with index are relative variables that were obtained as the ratio between
the value of current variable and its maximum value.
1,2
1
0,8
V*
0,6
P*
0,4
0,2
0
0
0,2
0,4
0,6
0,8
1,2
Fig. 3. The variation of pressure and velocity versus time
From the analysis of the data, one of them are presented in the diagrams from Fig.3 and
Fig.4, can be concluded that this way of the fundamental problem solving of interior
ballistics it is closer of firing phenomenon intimacy with guns that use multiperforated
powders [5, 6, 7, 8, 9].
It is necessary to remark that the finish of the powder burning takes place after the
projectile leaves the barrel of gun.
1,2
1
0,8
V*
0,6
P*
0,4
0,2
0
0
0,2
0,4
0,6
0,8
1,2
Fig. 4. The variation of pressure and velocity versus displacement
358
The magnitudes with index from graphics are obtained as a ratio between current values
of magnitudes and their maximum values.
For an extant ballistic system, the experimental and theoretical data are presented in
Table 1.
Table 1. Experimental and theoretical data
Type of data
Parameters
Maximum pressure
[MPa]
Muzzle velocity
[m/s]
Experimental data
Theoretical data
240
256,6
840
847,1
From the comparing of the experimental data with the theoretical values obtained in the
case of the proposed mathematical model, had resulted that the difference between the
muzzle velocities values is approximate 0,5% and between the maximum pressures values is
approximate 6%.
The advantages of this manner of the fundamental problem solving comparative with the
manner based on geometrical law of the powder burning consist of followings:
-
it is utilized whole curve of the pressure P( ) that is obtained at the combustion of

the powder in closed bomb and that takes into account all particularities of powder
burning;
it is used the variable values for index from the burning rate law, in the case of
the powders that burn with the breaking of the grains, which allow to present all
aspects of the powder burning.
359
REFERENCES
[1] E.V. CIURBANOV: Vnutrenniaia Ballistika, Izdatelstvo VAOLKA, Leningrad, 1975.
[2] M.E. SEREBRIAKOV: Internal ballistics of guns and solid propellant rocket engines, vol. 1 and 2,
Military Academy Publishing House, Bucharest, 1970.
[3] T. VASILE: Internal ballistics of guns, vol. 1, Military Academy Publishing House, Bucharest,
1993.
[4] T. VASILE: Internal ballistics of guns, vol. 2, Military Academy Publishing House, Bucharest,
1996.
[5] T. VASILE: Contributions to the fundamental problem solving of interior ballistics,
PROCEEDINGS of the II-nd International Conference Artillery Barrel Systems, Ammunition,
Means of Artillery Reconnaissance and Fire Control, October, 27-29, 1998, Kiev, pg.440-445.
[6] T. VASILE and I. NISTOR: Mathematical modeling of projectile motion in grooved barrel, The
XXIIIth National Conference of Solids Mechanics, University Bulletin, Petrol-Gaze Ploiesti,
vol. 1, May, 28-29, 1999, Ploiesti, pg.61-66.
[7] T. VASILE: Studies and researches regarding the fundamental problem solving of interior
ballistics for weapon with grooved barrel, PROCEEDINGS of III-rd International Armament
Conference on Scientific Aspect of Armament Technology, October, 11-13, 2000, Waplewo,
pg.157-165.
[8] T. VASILE and C. BARBU: The Fundamental Problem Solving of Interior Ballistics Using the
Experimental Data from Manometrical Bomb, PROCEEDINGS of the XXXIVth International
Scientific Symposium of the Defense Research Agency, May 29-30, 2003, Bucharest, pg. 55-60.
[9] T. VASILE, C. BARBU and D. SAFTA: The Fundamental Problem Solving of Interior
Ballistics Using the Physical Law of Powder Burning, Academic and Applied Research in
Military Science Journal, Vol. 3, ISSN 1588-8789, 2004, Budapest, pg. 407-414.
360
MAIN CHARGE INITIATION OF INSENSITIVE MUNITIONS

BYUNPLANNED STIMULI.
FAST COOK-OFF RESISTANCE OF PRESSED PBX
Richard Wild
PBX-Center Maasberg, Diehl BGT Defence GmbH & Co KG,
Karl-Diehl-Strae 1, D-66620 Nonnweiler, Germany
Abstract:
Energy transfer by shock, impact and heat are the basis for the initiation of warheads
on demand as well as by unplanned stimuli. Heat is involved in every kind of HE
initiation and fire is a threat for a warhead which is always present, from the production
over transportation and storage up to its use.
So the thermal behaviour of an explosive is important for the Initiation of a PBX charge
and the Cook-off safety of ammunitions.
Moulding powder properties, heat transfer calculations and small scale tests help in the
design of Fast Cook-off resistant press-filled warheads.
Keywords:
1.
PBX, Insensitive Munitions, Cook-off Tests, Initiation, Press Filling.
INTRODUCTION
Accidents with ammunition are often caused by a simple fire.
Fig 1.
Fire on the aircraft carrier USS Forrestal 1967
The fire spreads to the ammunition and the whole story ends in a mass detonation of the
warheads.
361
Examples from the last 40 years like

aviation fires on ships (USS Oriskany, 1966; USS Nimitz 1981),
fires on trucks (Iraq Camp Doha,1991; Sinkiang 2000),
forest fires (East-Russia ammunition store, 1997; Ukraine storage, 2003),
show that heat resistance was and still is a challenge for IM development.
On the other hand heat is involved in every kind of HE initiation.
So the thermal behaviour of an explosive is important for the Initiation of a PBX charge
and the Cook-off safety of ammunitions.
2.
IM INITIATION
Low sensitive PBX in Insensitive Munitions use special multiple stage boosters for their
on demand initiation.
Fuze
Booster
Main Charge
Fig 2.
Multiple stage booster for the PBX initiation
Electric, mechanic, shock and heat energy are used to initiate the main charge as directed
via the initiation train.
Unfortunately also unplanned stimuli, delivering energy in form of shock, impact or heat
initiate the main charge in an undirected way.
362
These threats and the corresponding tests are defined and described in IM standards:
Heat Energy
Fast and Slow Heating, with the involved reaction mechanisms thermal
explosion and deflagration to detonation transition;
Impact Energy
Bullet and Fragment Impact, with the involved reaction mechanisms
deflagration to detonation transition and delayed shock to detonation
transition;
Shock Energy
Sympathetic Reaction and Shaped Charge Jet Impact, with the involved
reaction mechanisms shock to detonation transition.
3.
HEAT INITIATION MECHANISM

The decomposition of an explosive depends on its temperature.
The higher the temperature of the explosive, the faster is the decomposition rate and the
greater is the heat and the amount of gas, evolved by the HE decomposition.
As long as the heating temperature will not rise and/or the evolved decomposition heat
could be carried off by heat transfer, further reaction will only occur, when the explosive is
not long-term stable at this higher temperature.
However when the produced heat by the decomposition reaction is more than the heat
that could be transported out of the explosive, or if the heating temperature from its
environment is still increasing, temperature and pressure will increase until the HE explodes.
Heat transfer equilibrium:
Heat rate evolved = Heat transfer rate
Heat rate evolved = f (, H, Ea )
Heat transfer rate = f (, T, , cp)
with
: Thermal Conductivity
cp: Heat Capacity
H: Heat of Reaction
T: Temperature
: density
Ea: Activation Energy
363
4.
HEAT INITIATION OF CONFINED MAIN CHARGES IN A FIRE

The flame temperature of a fuel fire is between 900C and 1100C.
If no special insulation is used, the heat of the open fire is transferred through the metal
case of a shell direct to the explosive charge inside, depending only on the wall thickness.
Within a few minutes the surface of the explosive charge reaches the decomposition
temperature of the explosive. The explosive ignites, starts to deflagrate and, confined by a
thick-walled shell, the gas pressure of the reaction products builds up.
Pressure waves compact the explosive and accelerate the process to a detonation.
Fig. 4. shows what happens if no additional measures are taken to restrain this reaction.
A usual in trade PBX was pressed in a 155mm steel cylinder confined by 20mm wall
thickness. The commercial quality PBX detonates after 3.5 minutes, (Type I reaction):
Fig 3.
Test Setup Detonation of a RDX based PBX in a steel tube

with 20mm wall thickness
The effect of a thinner confinement shows Fig.5. A cook-off test with the same PBX in a
50mm steel tube, with 5mm wall thickness, electrically heated (ca 90 K/min). The thinner
shell opened before the deflagration transferred to a detonation. The result is an explosion, a
Type III reaction:
Fig 4.
Test Setup Explosion Reaction (Photos by WIWEB

of a RDX based PBX in a steel tube with 5mm wall thickness
364
Fig.4 and Fig.5 demonstrate the necessity, but also the possibility for mitigating a too
high pressure build up, by shell/explosive combinations, that will break before a detonation
is reached.
5.
PARAMETER MODIFICATION FOR COOK-OFF IMPROVEMENTS
For a given explosive/binder composition, the scope for thermal improvements and a
more convenient sensitiveness of the IHE is limited.
Nevertheless, potential for improvements are to be found in the modification of both, the
PBX and the press process.
5.1
PBX Modification
Points of departure for PBX modifications are first of all: the HE crystals, the coating
process and mechanical properties.
The thermal characteristics of a given PBX can be optimized to a certain degree by:
-
the crystal size, surface and shape, e.g. finer grain sizes,
the crystallization process of the crystals, e.g. higher purity of the HE crystals,
the bonding of the binder on the crystal surface, e.g. with pre-coated crystals,
the conductivity and elasticity of the binder by using special additives.
One method to assess the effects of these modifications on the thermal behavior of the
granules are DSC measurements:
Fig 5.
DSC of a RDX based moulding powder (Photo by WIWEB)
Comparisons of DSC curves of the explosive decomposition and the material loss help,
in a very early stage of a HE optimization, to predict the effectiveness of the measures.
365
5.2
Press Process Modification
PBX press filling has grown out of a simple fill in squeeze eject process over a
complicated and lengthy high temperature fill evacuate - press eject - cool down
process into a more sophisticated compacting of granules, to support IM compatibility,
performance and production efficiency.
Since some years a few explosive manufacturers offer PBX granules which allow
densities of the pressed PBX close to 100% of the theoretical material density. Among other
advantages this reduces the porosity and improves the heat transfer in the explosive charge
by influencing the thermal conductivity and the heat capacity cp of the PBX charges.
For these new kinds of explosives, low compaction pressures, reduced wall friction and
a smaller compacting expansion of the PBX are additional steps to the cook-off
improvement by the manufacturing process of the main charge.
The effectiveness of the measures can be derived for example from confined tube tests,
which show the explosiveness response of PBX compacting variations.
Mainly three versions of confined tube tests are in use:
-
fuel fire,
internal ignition and
electrical heating
of steel tubes filled with PBX pellets and closed by steel covers.
These tests give orders of precedence for explosiveness and for application of a PBX to
assess the IM suitability for thermal threats under confinement.
Fuel fire test with steel tubes, 6mm wall thickness

Fig 6.
Confined burning tube tests of a RDX based PBXmod
5.3
366
5.4
Shell Modification
The main design techniques for thick-walled shells to withstand cook-off are:
-
protection against thermal stimuli by insulation and
prevention or mitigation of an increasing pressure build up by a venting device.
Insulation of the inner surface of the shell body is used in two respects:
-
6.
to get a well defined heat transfer through the shell to the explosive as well as a
shock absorber or reducer.
OPTIMIZATION OF THE COOK-OFF RESISTANCE
Making use of these tools in a step by step optimization, the cook-off behavior of an
RDX based PBX was improved.
In small scale tests and close to original tests, visible progress by the parameter
optimization can be proved, see Fig. 7 and Fig.8.
The test results of the PBXmod filled in the steel tubes with 5mm wall thickness is now
burning, a Type V reaction:
Fig 7. Test Setup Burning Reaction (Photos by WIWEB)

of a RDX based PBXmod in a steel tube with 5mm wall
thickness
The same test as in section 4, but now with a PBXmod directly pressed in an 155mm
steel cylinder confined by 20mm wall thickness is shown in Fig.8.
The PBXmod in the press filled steel cylinder reacts after 3.3 minutes.
367
The result is a Type V reaction:
Fig 8.
7.
Test Setup Burning Reaction (Photos by WIWEB) of a RDX based

PBXmod in a steel tube with 5mm wall thickness
IM BEHAVIOR OF ORIGINAL WARHEADS
This basis of examinations and tests allows the application of similar PBXmod in
original warheads to get Cook-off resistance.
One of the more challenging examples is a press filled 155mm HE artillery round, tested
with its transport pallet in a wood fire.
A picture of the round, standing in the pallet on a burning stack of wood is on the left
side of Fig. 9.
On the right side of the same figure comprises an overview of the test remains.
The result of this Fast Cook-off test was a Type V reaction.
-
using thermal insulation together with a mitigation technique for the shell body.
Round and Pallet in the fire
Test remains
Fig 9. Original Fast Cook-off test, burning reaction TypeV
368
8.
SUMMARY
Even in existing explosive/binder compositions it is possible to improve the thermal and

mechanical characteristics with regard to Cook-off compatibility.
For a RDX based PBX example, the Cook-off response of the explosive in a strongly
confined shell could be reduced from a detonation reaction to a Type V reaction.
These improvements are attainable by:
-
investigating the initiation mechanisms of explosives under unconfined and

confined conditions,
modifications of the parameters influencing the thermal behavior of the PBX,
optimizing the compacting process of a press filled shell in a way that improves
the cook-off resistance,
using thermal insulation together with a mitigation technique for the shell body.
369
FROM BINARY MIXTURES

TO COMPLETE AMMUNITION COMPATIBILITY
A WHOLE CARTRIDGE IN ONE TAM AMPOULE
Stephan Wilker, Gabriele Pantel, Gerhard Holl, and Uldis Ticmanis
WIWEB ASt Heimerzheim, Groes Cent, 53913 Swisttal (DE)
Abstract:
The determination of compatibility between explosives and contact materials in
ammunition is an important test. It is mandatory while a final (type) qualification is
performed. STANAG 4147 allows different methods to determine the compatibility
which are able to demonstrate whether a reaction between explosives and other
ammunition components occur or not. All tests require binary mixtures of explosive and
contact material in a 50:50 ratio. This is said to be a worst case scenario, meaning that
if no reaction is observed in a 50:50 mixture of minced materials then there will be no
reaction in the real environment, where the mixture ratio is often like 95:5 and the
materials consist of big pieces with a small surface to volume ratio.
But binary mixtures do not consider the reality at all. One may think about reactions
with three (or more) partners. For example, we have found out that in a 9 mm
ammunition a reaction is going on between the solvent in the propellant with one
component of the primer mixture which is catalyzed by the presence of copper. If only
binary mixtures were checked then this reaction would have remained undetected.
Therefore we have decided to measure complete ammunition in the microcalorimeter
(TAM) to be sure to have followed all reactions that are going on in the cartridge.
Unfortunately most ammunition articles are too big to be introduced into a 4 or 20 ml
TAM ampoule. Therefore we had to simulate a complete cartridge by adding all
components in a realistic weight or volume ratio in a TAM ampoule.
The comparison of the microcalorimetry signal of these demonstrators with the
microcalorimetry signal of real ammunition shows whether this is a suitable way to go
ahead or whether other parameters than mass, surface or volume ratio determine the
microcalorimetry signal of the ammunition. We have tested two small calibre
ammunition articles (a 9 mm cartridge and a 5.56 mm cartridge) by variation of their
ingredients (propellant, metal, lacquer, primer). It could be shown that besides the
overall loading density the ratio between propellant and primer is the main parameter
to shift the microcalorimetry signal of this demonstrator as close as possible into the
pattern of the microcalorimetry signal of the complete cartridge.
Keywords: compatibility, ammunition stability, double base propellants, small calibre
ammunition, microcalorimetry
1.
INTRODUCTION
Ammunition stability is a product of the stability of the ingredients and the compatibility
of the explosives (propellant and primers) with each other and with the surrounding inert
contact material (like metal, lacquers, paints etc.). In course of new developments and
qualification exercises the compatibility between the explosives and contact materials is
examined. Generally we follow the regulations of STANAG 4147 [1] which demands a 50:50
mixture ratio with minced materials to ensure a big contact area. If there were a reaction
between the explosive and the contact material it would be easily detected in this case.
370
Usually these conditions are far from reality. This has several reasons: The first reason is
that in ammunition the explosive/contact material ratio usually is far away from 50:50. The
second reason is the fact that the materials in ammunition usually are not minced. Therefore
the conditions in STANAG 4147 tests are much more severe than in reality (this is good in
terms of having a conservative approach). But what is missing in STANAG 4147 tests is the
fact that only binary mixtures are tested. Any incompatibility that requires more than two
reaction partners will remain undetected in the compatibility test. One example with an
incompatibility between a residual solvent in a propellant with a component of the primer
mixture under the catalytic influence of copper has been reported recently [2]. Also special
physical conditions (pressure, moisture, adsorption ability.) in the ammunition can not be
reproduced in the STANAG 4147 test.
Therefore the best alternative to overcome this disadvantage is to measure the compatibility of all components in an ammunition by putting the complete ammunition into the
microcalorimeter ampoule. Unfortunately most ammunition items are much too big to fit
into a 3 ml or 20 ml ampoule. Some TAM users help themselves by construction of
calorimeters with measuring cylinders big enough to even hold rocket motors [3], but this is
not a practical way to proceed. We decided to overcome this problem by mixing different
ammunition ingredients in a 3 ml ampoule. To study the influence of different ingredients
we varied parameters like mixing ratio, sample cutting and loading density. We did this at
the example of two ammunition articles which we could completely measure in the TAM.
2.
PROPELLANT AND AMMUNITION
Two types of propellants (both of them double base ball propellants) are tested. One of
them is the K5810 (used in 5.5645 mm cartridges), the other (K6210) is used in 919 mm
cartridges. The propellants used for the experiments were directly taken from the cartridges.
Three types of ammunition are tested. Two of them are 5.5645 mm cartridges, one
produced in Germany containing a lead free primer (Sintox), the other one produced in the
UK with a lead containing primer. The third is a 919 mm ammunition, produced in
Germany containing a lead free primer (Sintox).
The contact materials examined are the brass shells of the ammunition, the Sintox
primer, the projectile and the bitumen lacquer used in the ammunition.
3.
EXPERIMENTS
Microcalorimetric measurements were conducted with a Thermal Activity Monitor

(TAM III, Thermometric AB, Sverige). The measurements were performed in 3 ml glass
ampoules and in 4 ml steel ampoules. The 5.56 mm cartridge was measured in a 20 ml glass
ampoule (part of the projectile had to be cut to fit into the length of the ampoule). Usually
the ampoules are completely filled and sealed. Glass balls of diameter < 2 mm were added to
some experiments to reduce the free volume.
Tables 1 and 2 show the experiments that were performed during this study. To avoid
lengthy descriptions of the experimental setup during the discussion of the results the
experiment numbers, outlined in tables 1 and 2, are used throughout this text. The
arrangement of different experiments is sketched out in the Annex. All results are expressed
as W/g propellant, even if propellant and primer(s) are present in the ampoule. The same
approach is performed with the integrated values for compatibility evaluation (J/g
propellant).
371
Table 1. HFC Experiments with propellant K5810 and components of 5.56 mm

ammunition. All experiments were conducted at 89C
Exp. No Ampoule
type
Propellant
Brass Primers Lacquer Projectile Glass balls
Free volume
cm
1,353
no
no
n.d.
1,160
no
no
n.d.
no
no
no
n.d.
596
glass
no
no
no
597
glass
1,165
no
no
d)
1065
glass
no
no
0,022
1626
glass
3,495
no
no
no
no
no
1,0
1640
glass
0,875
no
no
no
no
no
2,7
1916
glass 20 ml
Cartridge 5.56 mm * 45 (UK)
n.d.
1924
glass 20 ml
Cartridge 5.56 mm * 45 (DE)
n.d.
2411
glass
3,414
no
no
no
no
no
1,1
2412
glass
3,432
no
no
no
no
no
1,1
2417
glass
3,000
3,000b)
no
traces
no
no
1,0
2418
glass
3,037
3,024b)
no
traces
no
no
0,9
2419
glass
0,801
0,807c)
no
traces
no
no
2,6
1,735
c)
no
traces
no
no
2,0
a)
no
traces
no
1,181
1,8
2420
glass
0,824
2433
steel
2,045
3,609
2457
steel
0,879
5,600a)
0,231
traces
3,999
0,635
2,0
2458
steel
1,496
6,301a)
0,228
traces
4,012
0,460
1,6
1,189
b)
0,228
traces
4,006
0,606
1,8
c)
0,233
traces
no
no
1,4
2459
steel
5,651
2460
glass
2,600
0,696
2490
glass
3,402
no
0,231
no
no
no
1,0
2491
glass
3,454
no
0,232
no
no
no
1,0
2492
glass
3,061
no
1,622
no
no
no
0,9
2493
glass
3,165
no
1,615
no
no
no
0,9
b)
0,231
traces
4,068
no
0,9
2494
glass
2,717
1,367
2495
glass
2,763
1,276b)
0,231
traces
4,292
no
0,8
2516
glass
2,494
1,275b)
1,618
traces
3,998
no
0,7
2517
glass
2,445
1,296b)
1,619
traces
3,996
no
0,7
Note to tables 1 and 2:

n.d. = not determined
a) bigger pieces (usually one shell divided into three parts)
b) small powder (grain size < 0,5 mm)
c) voluminous long thin wires
d) referring to mass of explosives inside one primer
e) traces: small amounts (< 0,01 g) present along some parts of the brass surface
372
Table 2. HFC Experiments with propellant K6210 and components of 9 mm

ammunition. All experiments were conducted at 89C
Exp. No
Ampoule Propellant
type
g
Brass
Primers
Lacquer
Projectile Glass balls

g
Free
volume
cm
772
steel
Cartridge 9 mm * 19
n.d.
773
steel
Cartridge 9 mm * 19
n.d.
1075
glass
2,704
no
no
0,271
no
no
n.d.
1184
glass
2,700
no
0,126
no
no
no
1,5
1959
glass
no
no
no
1,410
no
no
n.d.
1961
glass
2,400
no
no
0,239
no
no
n.d.
2409
glass
2,843
no
no
no
no
no
1,4
2410
glass
2,854
no
no
no
no
no
1,4
2413
glass
0,747
0,747c)
no
no
no
no
2,6
0,762
c)
no
traces
no
no
2,6
b)
no
traces
no
no
1,5
2414
glass
0,761
2415
glass
2,395
1,720
2416
glass
2,526
1,743b)
no
traces
no
no
1,4
2432
steel
0,768
7,493a)
no
traces
no
1,364
2,0
2435
steel
0,802
7,511a)
no
traces
no
1,446
1,9
2,516
3,484
a)
no
traces
no
no
2,0
a)
0,214
traces
7,988
1,814
2,0
2439
steel
2455
steel
0,462
2,570
2456
steel
0,465
2,607a)
0,237
traces
7,972
1,541
2,1
2457
steel
0,567
4,249b)
0,231
traces
3,999
0,636
2,4
2461
glass
2,057
0,978c)
0,213
traces
no
no
1,7
2486
glass
2,831
no
0,206
no
no
no
1,4
2487
glass
2,876
no
0,206
no
no
no
1,4
2488
glass
2,696
no
1,444
no
no
no
1,2
2489
glass
2,685
no
2496
2497
steel
steel
2,216
2,245
1,463
no
no
no
1,2
2,585
b)
0,211
traces
7,979
no
1,7
2,511
b)
0,210
traces
7,971
no
1,7
373
4.
RESULTS
4.1
Microcalorimetric measurements of the single components
The propellants, the primers and the lacquer were measured alone. Fig. 1 and 2 (in
chapter 4.2) show the microcalorimetric curves of the propellants and the lacquer. The
primer curve is depicted in figure 3. The brass of these cartridges were not examined, but we
know from other comparable experiments that their heat production rate is zero [4].
4.2
Microcalorimetric measurements of binary mixtures
This chapter shows the compatibility of the two propellants with bitumen lacquer, with
brass and with primers. Fig. 1 depicts the mixture between K5810 and bitumen, in fig. 2 the
compatibility between K6210 and bitumen is shown. Please note the different mixing ratio
(K6210 with a much higher nitroglycerin content requires a much higher propellant/bitumen
ratio for a comparable microcalorimetric signal).
P, W/g
T = 89C
600
400
Mixture
200
Propellant K5810
Lacquer
0
1.5
Fig 1.
3.0
4.5 Time, day
Microcalorimetry of propellant K5810, bitumen lacquer and their mixture

(mixing ratio 50:50).
The Sintox primer was also measured alone (see fig. 3). The heat generation rate is
reduced to the amount of explosive ingredients and not to the weight of the whole primer.
Six primers were mixed with propellant K6210 (see fig. 3), whereas a mixture of one and of
seven primers with the same propellant can be found in fig. 5.
374
P, W/g
T = 89C
800
600
Mixture
400
Propellant K6210
200
Bitumen
0
30
Fig 2.
60
Time, hour
Microcalorimetry of propellant K6210, bitumen lacquer and their mixture

(mixing ratio 90:10).
P, W/g
T = 89C
600
DM1465
K6210 + 6 DM1465
400
K6210
200
0
0
Fig 3.
Time, day
Microcalorimetry of the Sintox primer (DM1465 curve) and of K6210
375
350
2411
2412
300
2490
2491
2492
2493
250
200
150
100
50
0
0
t [days] at 89C
Fig 4.
Microcalorimetry of propellant K5810 (2411/2412), K5810 with one

Sintox primer (2490/91), and with seven Sintox primers (2492/93).
Approx. free volumes: 2411/12 = 1,1 cm, 2490-93 = 1,0 cm.
450
400
2409
2410
2486
2487
2488
2489
350
300
P [W/g]
P [W/g]
400
250
200
150
100
50
0
0
t [days] at 89C
Fig 5.
Microcalorimetry of propellant K6210 (2409/2410), K6210 with one

Sintox primer (2486/87), and with seven Sintox primers (2488/89).
Approx. free volumes: 2409/10 = 1,4 cm, 2486-89 = 1,2-1,4 cm.
376
P [W/g]
400
350
2411
2412
300
2419
2420
2417
2418
250
200
150
100
50
0
0
t [days] at 89C
Fig 6.
Microcalorimetry of propellant K5810 (2411/2412), K5810 with brass

powder (2417/18), and with brass wire (2419/20). Approx. free volumes:
2411/12 = 1,1 cm, 2417/18 = 1,0 cm, 2419/20 = 2,3 cm.
450
2409
2415
2414
400
350
2410
2413
2439
P [W/g]
300
250
200
150
100
50
0
0
t [days] at 89C
Fig 7.
Microcalorimetry of propellant K6210 (2409/2410), K6210 with brass

powder (2415), brass wire (2413/14) and with two big brass pieces (2439).
Approx. free volumes: 2409/10 = 1,4 cm, 2413/14 = 2,6 cm, 2415 = 1,5
cm, 2439 = 2,0 cm.
377
4.3
Microcalorimetric measurements of ingredients of 5.56 mm

ammunition
The following figures show different experiments with propellant K5810, with
ingredients of the 5.56 mm ammunition and of the complete cartridge. Please refer to table 1
for individual amounts of materials and to the figure in the Annex for the experimental setup.
400
350
P [W/g]
300
250
2411
2412
2460
2458
2459
1916
1924
200
150
100
50
0
0
t [days] at 89C
P [W/g]
Fig 8.
Microcalorimetry of propellant K 5810 (2411/12), of ingredient mixture

with one primer (2458/60), of a rebuilt ammunition (2459), of UK
ammunition (1916) and of DE ammunition (1924).
400
2411
2412
350
2494
2495
300
2516
2517
250
1916
1924
200
150
100
50
0
0
t [days] at 89C
Fig 9.
Microcalorimetry of propellant K 5810 (2411/12), of ingredient mixture

with one primer (2494/95), with seven primers (2516/17), of UK
ammunition (1916) and of DE ammunition (1924). Approx. free volumes:
2411/12 = 1,1 cm, 2494/95 = 0,9 cm, 2516/17 = 0,7 cm.
378
4.4
Microcalorimetric measurements of ingredients of 9 mm

ammunition
The following figures show different experiments with propellant K6210, with
ingredients of the 9 mm ammunition and of the complete cartridge. Please refer to table 2 for
individual amounts of materials and to the figure in the Annex for the experimental setup.
450
400
P [W/g]
350
300
250
200
150
100
2409
2410
2455
50
2456
2461
773
2457
0
0
t [days] at 89C
Fig 10. Microcalorimetry of propellant K 6210 (2409/10), of different mixtures

with all components (2455-2461) and of 9 mm ammunition (773). Approx.
free volumes: 2409/10 = 1,4 cm, 2455-57 = 2,4 cm, 2461 = 1,7 cm.
450
2409
400
2410
2496
2497
773
P [W/g]
350
300
250
200
150
100
50
0
0
t [days] at 89C
Fig 11. Microcalorimetry of propellant K 6210 (2409/10), of ingredient mixture

with one primer (2496/97) and of 9 mm ammunition (773). The free
volume of all samples is comparable (1,4 - 1,7 cm).
379
5.
DISCUSSION
5.1
Microcalorimetry of single components
The propellants show the usual behaviour of double base DPA stabilized propellants.
The propellant K6210 with a higher nitroglycerin content has a higher heat flow level and an
earlier second maximum. The lacquer depicts a first maximum, due to oxidation reactions.
The primer also has a first maximum, and further on the heat generation rate is nearly
constant at a level which is comparable to that of propellant K6210.
5.2
Microcalorimetry of binary mixtures
The mixtures of propellant with lacquer show an incompatibility. This incompatibility

increases with lacquer content [5] and with the nitroglycerin content of the propellant.
The mixture of propellant with one primer show no significant change of the heat flow
pattern. If seven primers are added to the propellant the heat generation rate increases during
the whole measurement. The times of first and second maximum do not change.
The mixture of propellant and metal does not show any incompatibility as could be
expected from recent studies [4]. Nevertheless there is a slight catalytic reaction that leads to
an earlier second maximum (compare curves 2417/18 with 2411/12 in fig. 6 and curves
2415/2439 with 2409/10 in fig. 7). In the case of the 9 mm ammunition the first maximum
also is increased. If brass wires instead of brass powder are used the microcalorimetric signal
is changed significantly. Instead of a long first minimum a sharp peak is observed. This peak
is very similar to the additional peak that is always observed when propellant is measured
under a low loading density [6]. Fig. 12 compares the curve shape of these two experiments.
400
350
2411
2419
2420
1640
P [W/g]
300
250
200
150
100
50
0
0
t [days] at 89C
Fig 12. Microcalorimetry of propellant K 5810 with loading density 1.0 (2411), of
propellant K5810 with loading density 0.25 (1640) and of mixtures of
propellant K5810 with brass wires (2419/2420). Approx. free volumes:
2411 = 1,1 cm, 1640 = 2,7 cm, 2419/20 = 2,3 cm.
380
Using voluminous brass wires leads to a much lower loading density of the whole
ampoule which easily explains the behaviour of these mixtures.
5.3
Microcalorimetry of ingredients of 5.56 mm ammunition
Figure 8 shows different experiments that were conducted with a combination of

ingredients of the 5.56 mm ammunition. The experiments 2458 and 2460 have a comparable
pattern of the heat generation rate than the propellant but with a longer and higher first
maximum (oxidation reactions) and an earlier second maximum (complete stabilizer depletion). The curves 2459 and 2475 show an additional peak during the first day which is
comparable to the peak depicted in fig. 12 with low loading density experiments. This underlines the importance of the propellant/oxygen ratio for the whole decomposition process.
In the following experiments the loading density was increased by adding as much of
propellant as possible into the TAM ampoule. By doing this the weight ratio between propellant and contact material is much higher than in the ammunition itself. But it helps
extremely to shift the heat generation rate curve of the mixture towards the shape of the heat
generation rate curve of the complete ammunition (see fig. 9). The addition of one primer to
the propellant/brass mixture leads to an earlier second maximum. If further primers are
added the level of heat generation is increased without a further shift of the second maximum. Thus the heat flow curves of the demonstrator curves with seven primers are quite
close to the heat generation rate curve of the complete cartridge. The biggest difference
between these curves is the absence of early oxidation reactions (first maximum) in the
ammunition. This can easily be explained by the absence of freshly prepared metal surfaces
and by the absence of fresh oxygen in the cartridge all oxidation reactions between
oxygen and the propellant in the cartridge have already been finished.
5.4
Microcalorimetry of ingredients of 9 mm ammunition
The heat flow pattern of the ingredients of the 9 mm ammunition are very comparable to
those recorded for the 5.56 mm ammunition (see figs. 10 and 11). Figure 10 depicts several
experiments with a too low loading density, especially experiments 2455 and 2461. Experiment 2455 was conducted with amounts of propellant that are identical with the ammunition
(which is obviously too low in the TAM ampoule), experiment 2461 was conducted with
metal wire to study the surface area influence (with the effect of a much too big empty space
in the ampoule).
If these errors are overcome by adding as much as possible propellant into the TAM
ampoule (curves 2496/2497 in fig. 11) the heat flow curve becomes quite similar to the curve
of the complete ammunition. Small differences (duration of first minimum, level of second
maximum) still remain and as far as we can see now, it will be impossible to reach a completely identical heat flow curve by just mixing components together.
5.5
Compatibility and stability evaluation
The compatibility of the ingredients is calculated analogously to STANAG 4147, test 2

of the new draft edition (QR procedure, see also [7]). To overcome the difficulty of
summarizing the weight of all (inclusive inert) ingredients the evaluation was performed by
only regarding the weight of the propellant. The stability is evaluated by maximum heat
generation rates described in STANAG 4582 [8].
Table 3 summarizes the heat generation of all experiments. To be in line with STANAG
4147 and STANAG 4582 the value after 3.83 days ( 10 years at 25C) was taken.
381
Table 3. energy release of microcalorimetric experiments (all values taken after

3.83 days) (left part: K5810/5.56 mm; right part: K6210/9 mm). Bold
figures: QR exceeds 30 J/g
Exp. No
Energy
release
Energy
Stability
difference evaluation
4,8 c)
597
89,5
c)
1065
92,7
a)
1626
Energy
release
Energy
Stability
difference evaluation
d)
772
82,5
13,2
46,0
773
83,9
14,6
145,7
22,4
J/g prop. J/g prop.
596
Exp. No
J/g prop. J/g prop.
b)c)
d)
1075
215
42,9
-0,6
1184
93,3 c)
1640
52,3
8,8
1959
1916
67,5
24,1
1961
171,5
102,2
1924
82,6
39,2
(S)
2409
69,1
2411
43,5
2410
69,5
2412
43,4
2413
65,9
-3,4
(S)
2417
46,8
3,3
2414
66,6
-2,7
(S)
2418
45,9
2,5
2415
68,2
-1,1
2419
53,3
9,8
2416
63,4
-5,9
2420
53,8
10,4
2432
80,0
10,7
2433
49,3
5,8
2435
85,2
15,9
2458
58,5
15,1
2439
69,2
-0,1
2459
53,0
9,6
2455
96,8
27,5
(S)
2460
46,1
2,7
2456
91,0
21,7
(S)
2490
45,3
1,8
2457
82,5
13,2
(S)
2491
44,5
1,1
2461
63,7
-5,6
2492
57,5
14,1
2486
65,7
-3,6
2493
59,7
16,3
2487
69,2
-0,1
2494
50,2
6,8
2488
91,2
21,9
2495
48,5
5,1
2489
91,5
22,2
2516
67,6
24,2
2496
69,0
-0,3
2517
66,2
22,8
2497
65,7
-3,6
2,1 c)
Explanation to table 3:
a) per gramme primer mixture
b) estimated value; experiment has ended before 3.83 days
c) per gramme lacquer or per gramme mixture
d) S = stable, (S) = 2nd maximum > 314 W/g, N = not stable
382
The compatibility between propellant and primer is good, if only one primer is used.
With six or seven primers the interreaction heat rises to 20-25 J/g which is still below the
limit value of 30 J/g.
As known beforehand, the propellants are not compatible with bitumen lacquer. The
incompatibility is bigger in the case of K6210. If the mixing ratio is shifted dramatically
(like 99:1) the microcalorimetric curve of the mixture becomes comparable to that of the
propellant alone [5].
The compatibility between propellant and brass is good, independent of the shape of the
metal material and the shape of the curve (remember the big influence of loading density).
The compatibility between propellant and brass + primer is acceptable (QR values
between 10 and 25 J/g). The interreaction heat is independent of loading density and design
of the experiment. Generally propellant K6210 shows slightly higher interreaction values
than K5810.
The compatibility of the whole ammunition is given for the 919 mm ammunition and
for the 5.5645 mm ammunition of UK production. But the QR value of the 5.56 mm
ammunition of DE production is higher than 30 J/g. This is mainly an arithmetic effect of the
lower propellant mass (in comparison with experiments No. 2516/2517) in the cartridge.
The stability of most of the samples is good, because their maximum heat generation
rate is below the limit value of the STANAG 4582 (314 W/g). Some samples (5.56 mm
ammunition of DE production and the samples of K6210/9 mm with a low LD) exceed the
limit value for a short time in their 2nd maximum and are thus marked with a (S). The
mixture of K6210 + seven primers exceeds the limit value in the end, combined with a high
gassing tendency. Therefore the entries 1184, 2488 and 2489 are marked with a N in the
stability column.
6.
CONCLUSION
The determination of compatibility of an explosive with contact materials should not

exclusively be conducted by simply testing binary mixtures. An interreaction between three
or more components is very likely to occur, even if one of the partners only by catalysing a
reaction. The best way to proceed is to measure the whole ammunition item by microcalorimetry. If this is not possible (e.g. because the item does not fit into the microcalorimetry
measuring channel) then a mixture of different ingredients may be a suitable way to go.
But as our experiments show an excellent agreement of the microcalorimetry curves
of the mixtures with the complete cartridge is not feasible. The reason for this is the general
problem of miniaturisation. Shall the weight ratio of the ingredients remain constant? Shall
the surface area ratio remain constant? And how big is the influence of the atmosphere inside
the ammunition? What about physical effects like adsorption ability of some surfaces?
What we can learn from these experiments is that the microcalorimetric signal is very
sensitively depending on the loading density and the oxygen/propellant ratio. Therefore if
any compatibility studies are to be conducted it is more important to choose the correct
loading density and closing procedures than to vary the mixing ratio. If the materials are
compatible it also does not seem to matter whether the materials are ground or compact. The
surface/volume ratio begins to be important when reactions between the explosive and the
contact material are observed.
383
As long as we can test complete ammunition we can compare the microcalorimetric

signal of individual mixtures against the microcalorimetric signal of the ammunition. If this
is not the case we can only suggest which mixture signal comes close to the complete
ammunition signal. But at least, any multi-component mixture signal seems to be better than
the signal of simple binary mixtures as long as the loading density and the closing of the
ampoule corresponds with the ammunition.
REFERENCES
[1]
[2]
[3]
STANAG 4147, Explosives Chemical Compatibility of Ammunition Components with

Explosives and Propellants (Non-Nuclear Applications), Edition 2 (1999).
ULDIS TICMANIS, PIERRE GUILLAUME, ANDR FANTIN, GABRIELE PANTEL, JEAN-PIERRE
MARCHANDISE and STEPHAN WILKER: Assessment of the Ballistic Service Lifetime of a
Small Calibre Cartridge by Heat Flow Calorimetry, 3rd International symposium on Heat Flow
Calorimetry and its applications for Energetic materials (2002).
DAN ELLISON, ANTON CHIN, NSWC Crane IN, personal communication (2001).
[4] STEPHAN WILKER and GABRIELE PANTEL, Vertrglichkeit von TLP mit Metallen, WIWEB
[5]
[6]
[7]
[8]
internal report (2004).

STEPHAN WILKER, GERHARD HOLL, ULDIS TICMANIS, ANNE SOMMER, GABRIELE PANTEL,
Chemische Vertrglichkeit von TLP mit Kontaktmaterialien, Int. Annu. Conf. ICT 28, 17
(1997).
STEPHAN WILKER, PIERRE GUILLAUME, MICHEL H. LEFEBVRE, SAMMY CHEVALIER,
LAURENCE JEUNIEAU, GABRIELE PANTEL, ULDIS TICMANIS and LUTZ STOTTMEISTER:
Stability analyses of rolled ball propellants Part 1: Microcalorimetric studies and stabilizer
depletion, Int. Annu. Conf. ICT 34, 83 (2003).
STEPHAN WILKER, PIERRE GUILLAUME, International Round Robin Test on Compatibility,
Int. Annu. Conf. ICT 33, 110 (2002); STEPHAN WILKER, STANAG 4147 Concepts for a 3rd
Edition, 4. Forum Explosivstoffe WIWEB 2003, 119-131.
STANAG 4582, Explosives, Nitrocellulose Based Propellants, Stability Test Procedures and
Requirements Using Heat Flow Calorimetry, 1st Edition, March 2003.
384
Figure A. Experimental setups (experiments with some ingredients)
385
Figure B. Experimental setups (experiments with all ingredients)
386
THE STUDY OF CHEMICAL MICRO-MECHANISM OF NITRAMINES

INITIATION WITH UTILIZATION OF 15N NMR CHEMICAL SHIFTS
Svatopluk Zeman
Department of Theory & Technology of Explosives,
University of Pardubice CZ-532 10 Pardubice, Czech Republic
Abstract:
A brief survey is presented of the authors results obtained from studies of the chemical
micro-mechanisms of nitramines initiation from the point of view of organic chemistry. The
relationships have been presented and discussed between the characteristics of impact and
electric spark sensitivities, detonation and thermal decomposition, on the one hand, and 15N
NMR chemical shifts of nitrogen atoms of nitramino groups, on the other. In the case of the
impact sensitivity, the said relationships involve the 15N shifts of the amino nitrogen atoms
carrying the nitro group primarily split off from the molecule. In the case of the initiation by
shock, heat and electric spark, the 15N shifts of nitrogen atoms in the primarily split off nitro
groups themselves are involved. It has been stated that the chemical micro-mechanisms of
primary fission processes of molecules of nitramines in the initiation by mechanical stimuli
(inclusive the detonation course) and electric spark should be the same as in the case of
their low-temperature thermal decomposition. Also mentioned is relevance of the modified
EvansPolanyiSemenov relationship. On the basis of the findings presented it also has
been stated that the detonation transformation itself of the nitramines should be preceded by
an induction period.
Keywords:
1.
detonation, electric spark sensitivity, impact sensitivity, initiation,

mechanism, nitramines, 15N NMR chemical shifts, thermal decomposition
INTRODUCTION
At present it is a generally accepted idea that nitro groups represent the primary cause of
initiation reactivity of polynitro compounds. A number of papers, inclusive of some striking
pieces of experimental evidence, deal with relationships between the chemism of homolytic
fragmentation of these compounds in the initiation processes and the chemism of lowtemperature thermal decomposition (see Introduction in Ref. 1 and references herein). However,
the approaches to study of micro-mechanism of initiation of these compounds are dominated by
opinions of physicists (e.g. Refs 2-5). If the problems of initiation chemism are mentioned in
literature at all, then this is done very reluctantly and mostly without any broader chemical
contexts. Until now some authors [3,5] have been referring to the study on the significance and
relevancy of molecular-structural dependences of initiation reactivity [6] based on non-real ideas
about the primary radical reactions of nitro groups within the detonation transformations of
several polynitro arenes. With regard to the reasons mentioned, this present paper offers a
review of results of studies of initiation mechanisms obtained on the bases of approaches of
physical organic chemistry to the problem given over the last seven years. These results are
demonstrated on nitramines, which in their molecular structure are relatively simple polynitro
387
compounds, and the mechanism of primary homolysis of their molecules is well understood [7.
11]
2.
PHILOSOPHY OF APPROACH
Recently the relationships were found between outputs of non-isothermal differential

thermal analysis and characteristics of detonation of polynitro arenes [12,13] and also nitramines,
nitrosamines and nitric esters [13,14]. A more detailed analysis of these results for polynitro arenes
shows that their classification in the sense of the relationships found is given primarily by steric
conditions and electron configuration in the ground state of the reaction centre of the molecule
[12,13]
. In this context, by the reaction centre is meant the grouping of atoms and/or functional
groups in the molecule whose primary chemical changes initiate decomposition of this
molecule. At the same time, the said method represents one of the basic principles of approach
of organic chemistry to dealing with reactivity problems in general.
The electron configuration and steric conditions within the reaction centre of the molecule
can be represented by NMR chemical shifts of the key atoms of the centre. It is known that the
application of 15N NMR chemical shifts to the study of chemical micro-mechanism of initiation
of nitramines by heat [1,15], impact [1,16], shock [1,17] or electric spark [1,18,19] has given highly
valuable results. It can be argued against this approach that the NMR studies carried out in
solutions neglect important crystal-lattice effects that are vital in the determination of explosive
properties [20]. We pointed out this possibility as early as in Ref. 21 using the particular case of
2,4,6,8,10,12-hexanitro-2,4,6,8,10,12-hexaaza-isowurtzitane (HNIW). From several papers [1,1517]
and the following text it will be seen that this fact has no fundamental significance for studies
of chemical micro-mechanism of initiation of energetic materials. Also a study of direct relation
between electronic charges at nitrogen atoms of primarily split off nitro group and thermal
reactivity of nitramines [22] is a logical supplementation of this type of investigation.
3.
DISCUSSION
We were engaged in study of impact sensitivity, defined as the so-called first reaction (i.
e. the first chemical change of the sample is taken as positive result), in the past [16,23]. An
analysis of the relationship between the corresponding drop energies, Edr, and 15N NMR
chemical shifts of nitrogen atoms of nitramino groups led to formulation of the relationship [16]
documented in Fig. 1. The correlation involves the chemical shifts, A, of nitrogen aza-atoms
carrying the nitro groups that are leaving first during the initiation (i. e. the most reactive nitro
groups). The aza (amino) nitrogen chemical shifts, A, in the nitramino groups are expected to
be influenced by the nitrogen hybridisation, the size and conformation of molecule, and the
extent to which the nitrogen lone pair is involved in -bonding with NO2 group. As the
conformation and size of molecule plays dominant role in the intermolecular interactions in the
corresponding crystal, the interactions should have a significant influence on the impact
sensitivity. This hypothesis is verified by Fig. 2, which represents a relationship between the Edr
values and heats of fusion, Hm,tr, of nitramines [16]. The heat of fusion is defined here as a sum
of all heats of the polymorphous transitions and of the heat of melting. As the heat represents the
work needed for formation of defects in crystal lattice, the relationships found seem to stand in
accordance with the ideas about the decisive role of plastic deformations of crystal [24-26] and of
388
dislocations in energetic crystals [27,28] which they played in the initiation of energetic materials
by impact or shock (also the crack temperature can be mentioned in this connection [29,30]).
25
Cyclic nitramines
Drop energy of he first reaction Edr /J
CH 3
20
y = -0.7936x - 150.81
2
R = 0.9919
O2N
N
CH 3
DMNA
Linear nitramines
N
O2N
y = 0.6738x + 141.11
2
R = 0.9932
NO 2
CO CH 3
CPX
15
NO 2
N
H3C
DADN
NO 2
NO 2
TETROGEN
N
N
NO 2
NO 2
RDX
N
N
N
O2N
-220
-215
-210
-205
DECAGEN
NO 2
HNIW(2-)
O2N
O2N
NO 2
NO 2
N
N
NO 2
NO 2
NO 2
N
CH2 CO OCH 3
N
CH2
CH2
CH2
NO 2
NO 2
CH3
CH3
NO 2
NO 2
TEX
N
O2N
O2N
N
NO 2
O 2N
-200
N
O2N
NO 2
NO 2
NO 2
H3C
O2 N
NO 2
H3 C
TNADEC(4,7-)
O2N
0
-225
OHMX(4,6-)
NO 2
O2 N
HMX
NO 2
NO 2
NO 2
CH3O CO CH 2
AcAn(4,6-)
HOMO(1,5-)
NO 2
NO 2
O2 N
DPT
CH3
N
NO 2
N
O 2N
DNDC
O 2N
ORDX(4-)
CH3 CO
10
NO 2
NO 2
NO 2
NO 2
N
NO 2
NO 2
-195
-190
-185
-180
15
N NMR Chemical shift A / ppm
Fig 1.
Relationship between drop energy, Edr, of the first reaction and 15N NMR
chemical shifts, , of aza nitrogen atoms carrying the most reactive nitro
groups in the nitramine molecule (in bracket given are the respective positions
in molecule) taken from Ref. 16, complemented by the data of TETROGEN,
HNIW and TEX [1].
Paper [17] deals with the relationships between square of detonation velocities, D2, or heats
of explosion, Qreal, on the one hand, and values of the 15N NMR chemical shifts of the nitrogen
atoms in nitramino groups, on the other. These relationships for the D2 values are represented in
Fig. 3. In contrast to the foregoing case, in this case such chemical shifts correlate whose values
correspond to nitrogen atoms of nitro groups primarily split off during detonation [17].
Similarly, Fig. 4 represents the relation of the electric spark sensitivity to electronic
structure and steric conditions in the reaction centre of nitramine molecules: here the electric
spark energy, EES, correlates with the 15N NMR chemical shifts, N, of nitrogen atoms of the
most reactive nitro groups [1,19]. The relationships in this diagram are similar to those between
the square of detonation velocity, D2, of nitramines and their N values in Fig. 3. The mechanism
of primary fragmentation should be identical in the two cases, even if there is no shock
component present in the electric discharge. Absence of this component was justified on
samples of 1,3,5-trinitro-1,3,5-triazinane (RDX) that were contaminated with crushed glass (i. e.
by hot spots): the crushed glass markedly reduced electric spark sensitivity of the resulting
mixture [31]. The said effect can be interpreted [31] on the basis of separation of RDX grains from
each other by the introduced glass particles (hot spots develop here at the thinnest part of the
individual solid under discharge, i. e. at inter-grain points [32]). An analogous effect of the
introduced hot spots in lead azide was found by Stengach [33].
389
Logaritm of drop enery of the first reaction ln Edr / J
3.5
Cyclic nitramines
y = -0.0749x + 4.3909
R 2 = 0.9838
CH3
O 2N
N
CH3
DMNA
O2N
H3C
NO 2
CPX
O2N
CH3
NO 2
NO 2
ORDX
NO 2
N
2.5
N
O 2N
NO 2
TETROGEN
DNDC
CH2
CH2
N
CH2
HOMO
N
N
NO 2
H3C
N
O 2N
NO 2
NO 2
N
CH3
TNADEC
NO 2
NO 2
N
NO 2
NO 2
CH3
OHMX
DPT
NO 2
NO 2
O 2N
H3 C
NO 2
NO 2
NO 2
NO 2
N
O 2N
HMX
N
N
NO 2
O 2N
O 2N
RDX
NO 2
O2 N
N
N
O2 N
NO 2
O 2N
TEX
NO 2
NO 2
HNIW
NO 2
N
1.5
y = -0.012x + 2.962
R 2 = 0.9598
O 2N
DECAGEN
Linear nitramines
NO 2
NO 2
N
N
O 2N
NO 2
O 2N
NO 2
1
5
15
25
35
45
Heat of fusion H m,tr
Fig 2.
Square of detonation velocity D2 /km2s-2
110
55
/ kJ.mol
65
75
-1
Relationship between drop energy, Edr, of the first reaction and heat of
fusion, Hm,tr, of nitramines; taken from Ref. 16, completed by data of TEX
[1]
.
y = -3.404x - 44.662
2
R = 0.9677
NO 2
O 2N
N
N
N
HNIW
90
O 2N
O2 N
N
O 2N
NO 2
MDN
N
NO 2
HMX
CH 3
O 2N
NO 2
N
CH2
NO 2
NO 2
NO 2
N
H 3C
NO 2
N
CH2
CH2
CH3
NO 2
N
O2 N
RDX
OHMX
NO 2
O 2N
NO 2
DECAGEN
OHMX
TEX
HOMO
ORDX
N
CH 3
70
NO 2
O 2N
HMX
NO 2
O 2N
NO 2
CH 3
NO 2
CH3
CH 2
DIGEN
TNADEC
H3C
50
NO 2
NO 2
NO 2
NO 2
N
O 2N
NO 2
CPX
O 2N
H3C
N
NO 2
NO 2
CH3
CH3
CH2
H2C
NO 2
OCPX
DNDC
NO 2
CH2 CO OCH 3
CH 3O CO CH 2
NO 2
NO 2
O 2N
NO 2
AcAn
NO 2
N
N
NO 2
CH 2
NO 2
TETROGEN
O 2N
NO 2
N
O2 N
y = -1.529x + 29.475
2
R = 0.992
O2 N
N
NO 2
NO 2
DPT
NO 2
N
N
N
N
NO 2
DMNA
CH 3
30
O 2N
DADN
CO CH 3
CH 3
y = -22.946x - 714.68
2
R = 0.986
NO 2
N
N
N
N
NO 2
CH3 CO
10
-45
-40
-35
15
Fig 3.
-30
-25
N NMR chemical shift N /ppm
Relationship between square of detonation velocity, D2, and 15N NMR

chemical shifts, N, of nitro-nitrogen atoms of the most reactive nitro groups
(in brackets given are the respective positions in molecule) repainted from
Ref. [17].
390
-20
20
NO 2
18
DPT
NO 2
DNDC
16
CH3 CO O
NO 2
NO 2
NO 2
NO 2
AcAn
14
Spark energy EES / J
O CO CH 3
NO 2
NO 2
H3C
12
y = 8.0187x + 281.68
2
R = 0.9896
OCPX
O 2N
O 2N
NO 2
CH3
CO
N
CO
NO 2
H3C
DMNO
N
CH2
N
N
O2N
O 2N
NO 2
N
N
TETRYL
y = 0.4859x + 19.278
2
R = 0.9576
NO 2
HMX
TETROGEN
NO 2
O 2N
O2N
O2N
DECAGEN
NO 2
N
TNAZ
NO 2
NO 2
NO 2
N
N
NO 2
DIGEN
N
OHMX(4,6-)
O2N
O2N
NO 2
NO 2
CH3
HNIW(2-)
N
CH2
CH2
H2C
CH3
N
NO 2
ORDX(4-)
NO 2
NO 2
NO 2
N
O2N
NO 2
NO 2
O 2N
O 2N
NO 2
y = 1.1439x + 45.214
2
R = 0.9794
CH3
H3C
H3C
N
N
NO 2
CPX
10
8
H3C
NO 2
NO 2
O2N
NO 2
N
NO 2
RDX
O 2N
NO 2
O2N
NO 2
0
-44
-39
-34
-29
-24
-19
15
N NMR Chemical shift N / ppm
Fig 4.
Relationship between spark energy, EES (i. e. sensitivity to electric spark), and
N NMR chemical shifts, N, of nitro-nitrogen atoms of the most reactive
nitro groups (in brackets given are the respective positions in molecule) in
nitramines - taken from Refs. [1] and [19].
15
Activation energy of thermolysis Ea / kJ mol
-1
260
NO 2
240
O 2N
N
N
N
HMX
220
Liquid state
200
CH3
O 2N
CO
N
H3C
HMX
NO 2
NO 2
RDX
RDX
N
O2 N
N
NO 2
CO CH 3
O 2N
RDX
O 2N
DMNA
RDX
CH 3
HNIW
Sterically crowded
O 2N
HNIW
DMNA
CPX
NO 2
N
N
N
DADN
NO 2
CH3 CO
Solid state
N
N
N
HNIW
NO 2
DPT
NO 2
CH 3
DNDC
DMNO
180
160
CO
HMX
N
O 2N
NO 2
DPT
TETROGEN
TETRYL
H3C
140
NO 2
NO 2
O 2N
NO 2
TNAZ
120
100
-50
-45
-40
15
Fig 5.
-35
-30
-25
-20
-15
-10
N NMR Chemical shift d N of nitro-nitrogen atoms / ppm
Relationship between activation energy, Ear, of thermal decomposition under

conditions of Russian manometric method and 15N NMR chemical shifts,
, of nitro-nitrogen of the most reactive nitro groups in the nitramine
molecule (see Refs. [1] and [15]).
391
The 15N NMR chemical shifts were used in analysis and prediction of the Arrhenius
parameters of low-temperature decomposition of nitramines [15]. Figure 5 presents the
relationship between the activation energies, Ea, of this decomposition and the aforesaid
chemical shifts, N, of nitrogen atoms in the most reactive nitro groups [15]. No analogous
relationship was found for the chemical shifts, A, of aza-nitrogen atoms in nitramino groups.
The found dependences of the type shown in Fig. 5 are discussed in detail elsewhere [15].
From the point of view of electron structure and close neighbourhood, the individual
nitramino groups in polynitroamines are usually not equivalent to each other. Therefore, also
their abilities of participation in primary initiation processes are different. This fact is
documented in Figs 1-4, and also by the electronic charges at nitrogen atoms of the nitramines
calculated on the basis of the Mulliken population analysis of electron densities, qN, obtained by
ab initio DFT B3LYP/6-31G** method [22]. The relationships found between these charges at
nitrogen atoms of primarily leaving nitro groups and thermal reactivity of a set of nitramines are
documented in Fig. 6. The thermal reactivity is expressed here as the slope (activation energy)
Ea.R1 in the Kissinger relationship [34], which is used for evaluation of results of non-isothermal
differential thermal analysis. The logical relationships of Fig. 6 are discussed in Ref. [22], and the
most reactive nitro groups in molecules are pointed out again.
80000
y = 2E+06x - 1E+06
2
R = 0.9996
70000
TEX
60000
Kissinger Slope Ea.R-1 / K
O 2N
NO 2
y = -1E+06x + 830370
2
R =1
50000
NO 2
DMEDNA
H3C
N
CH2
CH2
O 2N
40000
O 2N
O 2N
NO 2
O 2N
NO 2
-HNIW(2)
RDX
O2 N
30000
O2N
ORDX(4)
DNDC
NO 2
O 2N
CH 3
N
CH 2
TNAZ(1)
N
CH 2
NO 2
NO 2
HOMO(1)
NO 2
y = 483403x - 290540
2
R = 0.9999
0
0.62
0.63
NO 2
CH 3
NO 2
N
O 2N
NO 2
10000
HMX
NO 2
N
NO 2
-HNIW(2)
CPX
y = 307657x - 186974
2
R = 0.9995
0.64
O 2N
O2N
O2N
N
N
N
N
NO 2
NO 2
N
N
NO 2
N
O 2N
NO 2
N
O2N
0.65
NO 2
O 2N
NO 2
H2N
20000
N
N
CH3
0.66
N
NO 2
0.67
0.68
0.69
Charge q / e
Fig 6.
Relationship between the Ea/R values and Mulliken B3LYP/6-31G** charges,

qN, at the nitrogen atoms of the primarily split off nitro groups from nitramino
groupings in molecule (in bracket given are the respective positions in
molecule) taken from Ref. [22]
392
From what has been given so far it follows that the initiation by impact should be dominated
by the key role of the aza-atoms carrying the primarily leaving nitro groups [1]. The dominant
factor in the initiation by shock, electric spark, and in low-temperature thermolysis should be the
electron structure and close neighbourhood of the primarily leaving nitro group [1]. For nitramino
molecules, whose nitrogen atoms are not isochronous, the said fact is documented in Fig. 7.
OHMX
H3C
N
N
O2 N
heat
N
O2 N
CH3
O2 N
NO2
heat
O2N N
O 2N N
N
heat
shock
electric spark
impact
NO2
N
O2 N
impact
shock
impact
O2 N
Fig 7.
electric spark
HOMO
shock
electric spark
N NO2
N NO2
N
O2 N
NO2
HNIW
Summarization of the findings about the initiation reactivities can be presented

as follows:
dominating reactivity of inner nitramino groups of OHMX molecule in impact and

shock [1,16,17] was confirmed by molecular dynamics simulation [35];
an electronic configuration and steric effect of the nitrogen carriers of the primarily
split off nitro group should play the key role in the initiation by impact [1,16], whereas
the state of nitrogen atom of this most reactive nitro group should be dominating in
the remaining kinds of initiations [1,17,22];
the most reactive nitramino group in HOMO molecule is that one at 1 position of its
molecule [1,16,17,22] and in HNIW molecule at position 2 [1,16,17](there is the longest N
N bond in molecule of its -polymorph [36]);
In the case of 2,4,6,8-tetranitro-2,4,6,8-tetraazanonane (OHMX) molecule, the primary

reactivity of inner nitramino groups (positions 4, 6) was confirmed by Kohno et al. [35] using
the molecular-dynamic simulation. This paper tends to be quoted [3,5] without mentioning one of
its important merits, i.e. the selectivity in the initiation reactivity of nitramino groups [35]: in
connection with action of impact or shock on a nitramine crystal, the intramolecular vibration
energy is transferred into nitramino groups. There is a difference between linear and cyclic
nitramines in transfer of the energy. According to Kohno et al. [35], the linear molecule OHMX
shows only low probability of transfer of surplus energy from the inner to outer nitramino
groups.
The longest NN bond in -HNIW is that of nitramino group at 2 position of its molecule
(143.6 nm, the single bond NN length being 142.5 nm [37]). Therefore, this bond will be the
first to undergo homolysis in initiation processes. This statement is supported by the findings
presented in Figs 1-4, 6 and 7.
[36]
393
Comparison of equations given in Figs 3-6 leads to the relationship

E = a D2 + b
(1)
and application of the definition relationship between detonation velocities and explosion heats
Q [38] in the form
Q = D2 {2 (2 + 1)}1
(2)
where is the polytropy coefficient, transforms Eq. 1 into the following form
E=Q+
(3)
which is a modified EvansPolanyiSemenov (E-P-S) equation [39] for energetic materials. The
original E-P-S describes a relationship between activation energies E of most substitution
reactions of free radicals and corresponding heats of reaction H of the narrow sets of substance
structures [40]. The equation documents that the strength of bond being split is a decisive factor in
the given reaction. A motive stimulus for study of modified E-P-S relationship was the
homolytic character of primary fission in an initiation of both the detonation and lowtemperature thermal decompositions of energetic materials [39] including explosive mixtures (see
e.g. Ref. [41]). Figure 8 was taken from Ref. [39] and documents this equation for nitramines,
whose activation energies, Ea, for cyclic derivatives correspond to their thermal decomposition
in the solid state.
225
HMX
NO 2
y = 11.131x + 151.46
2
R = 0.8803
Activation energy of thermal decompn. Ea /kJ mol-1
O2N
NO 2
TNAD
HMX
RDX
NO 2
N
N
TNAD
O 2N
HMX
DNDC
N
N
N
O 2N
CO CH 3
NO 2
N
N
CH 3 CO
185
DMNA(s)
N
N
DPT
NO 2
N
N
N
NO 2
NO 2
NO 2
NO 2
N
NO 2
NO 2
N
O2 N
NO 2
NO 2
DPT
NO 2
N
O 2N
NO 2
NO 2
DECAGEN
NO 2
NO 2
NO 2
N
O 2N
HOMO
205
RDX
TEX
O 2N
N
NO 2
CH3
O 2N
N
CH3
DADN
NO 2
TNP
O 2N
NO 2
NO 2
N
NVA
H3C
y = 2.7866x + 165.87
2
R = 0.9349
COOH
OCPX
DMEDNA
NO 2
NO 2
N
H3C
NO 2
O 2N
OHMX
ORDX
H3C
N
NO 2
NO 2
NO 2
CH3
N
CH2
NO 2
NO 2
NO 2
N
H 3C
CH2
N
CH2
CH3
CH3
H3C
N
CH3
O2 N
165
3
3.5
4.5
5.5
6.5
Heat of explosion Q real /MJ kg-1
Fig 8.
Modified E-P-S equation for relationship between activation energies, Ea,

of low-temperature thermolysis of nitramines (the cyclic ones in the solid state)
and real heat of explosion, Qreal (calculated according to the Pepekins semiempirical method [44]) taken from Ref. [39].
394
We also made molecular-dynamic simulation of thermal decomposition of some individual

energetic materials, including RDX, at extremely high temperatures [42]. It turned out that the
primary fragmentation mechanism at these conditions is entirely different from the lowtemperature variant. In the case of the RDX unimolecular decomposition, it can be mentioned
that elimination of NO2 group by homolysis of one NN bond is observed at all the reaction
conditions whereas triazinane ring fission (depolymerization to 1-nitro-1-azaethylene) occurs
predominantly in the gas phase thermal decomposition of this nitramine, i.e. at higher
temperatures [43]. The correlation of activation energies of low-temperature thermal
decomposition (i.e. in the range up to 600 K) of nitramines in the sense of E-P-S equation thus
means that the primary fission processes in this decomposition should be identical with those in
the detonation transformation of these polynitro compounds [1]. That also means that the
detonation transformation itself of the given substance should be preceded by an induction
period. Dremin postulated the necessity of such induction period also from the point of view of
physics of explosion [2].
4.
CONCLUSION
The primary fission processes of nitramine molecules in the low-temperature thermal

decomposition should be identical with those in the impact, electric spark and shock initiations.
The relationships presented here focus attention on those atoms in molecules of polynitro
compounds whose electron configurations and steric conditions play a key role in the primary
fission processes of these molecules during initiation by mechanic stimuli, electric spark, or
heat. That means that centres of initiation reactivity in molecules of nitramines can be
advantageously determined by means of 15N NMR spectroscopy. From the point of view of
physics of explosion the centres can be taken as hot spots.
Electron configuration and steric effects on aza-atoms, which are carriers of the most
reactive nitro groups in molecule, should play a key role in the initiation by impact.
Characteristics of detonation, electric spark sensitivity, and thermal decomposition, on the other
hand, correspond to electron configuration of the nitrogen atoms of the most reactive nitro
group in nitramine molecules.
The relationships presented here, namely the modified EvansPolanyiSemenov equation,
lead to a premise that the primary fragmentation of the studied polynitro compounds in their
detonation transformation proceeds at milder conditions than those present at the front of
detonation wave or in its reaction zone [1]. That means that the detonation transformation itself
of the given substances should be preceded by an induction period [1,2].
Acknowledgement
The paper is a part of authors lecture at the Nanyang Technological University in Singapore on
Jan. 19th; in this connection author expresses his gratitude to Prof. Ang How-Ghee for arranging the
lecture and for his valuable suggestions to its topic.
The work was created as a part of the project of the Ministry of Education, Youth and Sports of the
Czech Republic No. MSM 0021627501.
395
REFERENCES:
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
ZEMAN S.: A study of chemical micromechanism of the organic polynitro compounds initiation, in:
POLITZER P. & MURRAY J. S. (Eds.), Energetic materials, Part 2, , Detonation, combustion,
Elsevier B. V., 2003, pp. 25-52.
DREMIN A. N.:, Toward detonation theory, Springer, New York, 1999.
POLITZER P., MURRAY J. S. (Eds.): Energetic materials, Part 2, Detonation, combustion, Elsevier
B. V., Amsterdam, 2003.
LEIBER C.-O.: Assesment of safety and risk with a microscopic model of detonation, Elsevier B. V.,
Amsterdam, 2003.
DLOTT D.: Fast molecular processes in energetic materials, in POLITZER P. & MURRAY J. S.
(Eds.), Energetic materials, Part 2, Detonation, combustion, Elsevier B. V., Amsterdam, 2003, p.
125.
BRILL T. B., JAMES K. J.: Thermal decomposition of energetic materials. 61. Perfidy in the amino2,4,6-trinitrobenzene series of explosives, J. Phys. Chem. 97 (1993) 8752.
F. I. DUBOVITSKII F. I., KORSOONSKII B. L.: Kinetika termicheskogo razlozheniya Nnitrosoedinenii (Kinetics of thermal decomposition of N-nitrocompounds), Usp. Khim. 50 (1981)
1828.
NAZIN G. M., MANELIS G. B.: Thermal decomposition of aliphatic nitrocompounds Russ. Chem.
Revs., 63 (1994) 313.
STEPANOV V. P., FEDOTOV A. A., PAVLOV A. N., MANELIS G. B. in: NOVOZHILOV B. N. (Ed.):
Khimicheskaya fizika protsessov goreniya i vzryva (Chemical physics of the combustion and
explosion), Proc. 9th All-union Symp. Combust. Explos, Acad. Scienses USSR, Chernogolovka,
1989, p. 100.
COOK M. D.: Thermal decomposition of RDX: A critical review, J. Energ. Mater., 5 (1987) 257.
MANELIS G. B., NAZIN G. M., RUBTSOVYU. I., STRUNIN V. A.: Termicheskoe razlozhenie i
goreniye vzryvchatykh veschestv i porokhov (Thermal decomposition and combustion of explosives
and powders). Izdat. Nauka, Moscow, 1996.
ZEMAN S., The relationship between differential thermal analysis data and the detonation
characteristics of polynitroaromatic compounds, Thermochim. Acta 41 (1980) 199.
ZEMAN S.: A new aspects of relations between differential thermal analysis data and the detonation
characteristics of polynitro compounds, in: Proc. 29th Int. Annual Conf. ICT, Karlsruhe, 1998, pp.
ZEMAN S., FEDK J., DIMUN M.: The relationship between differential thermal analysis data and
the detonation characteristics of thermodynamically unstable aliphatic series compounds, Zbornik
Radova (Tech. Fac. Bor) 18 (1982) 119-135.
ZEMAN S.: Analysis and prediction of the Arrhenius parameters of low-temperature thermolysis of
nitramines by means of the 15N NMR spectroscopy, Thermochim. Acta 333 (1999) 121.
ZEMAN S.: New aspects of the impact reactivity of nitramines, Propellants, Explos., Pyrotech. 25
(2000) 66.
ZEMAN S.: Relationship between detonation characteristics and 15N NMR chemical shifts of
nitramines, J. Energet. Mater. 17 (1999) 305.
ZEMAN S., ZEMAN V., KAMENSK Z.: Relationships between the electric spark sensitivity and the
NMR chemical shifts of some organic polynitro compounds, in: Proc. 28th Int. Annual Conf. ICT,
Karlsruhe, 1997, pp. 66/1.
ZEMAN V.: Sensitivity of secondary explosives to electric spark, Ph.D. Thesis, Univ. Pardubice,
2003.
396
[20] CLAWSON J. S., STROHMEIER M., STUEBER D., ORENDT A. M., BARICH D. H., ASAY B. A., HISKEY
M. A., PUGMIRE R. J., GRANT D. M: 15N Chemical shift tensors of HMX, J. Phys. Chem., A 106
(2002) 6352.
[21] ZEMAN S., JALOV J.: Heats of fusion of polynitro derivatives of polyazaisowurtzitane,
Thermochim. Acta, 345 (2000) 31.
[22] ZEMAN S., FRIEDL Z.: Relationship between Electronic Charges at Nitrogen Atoms of Nitro Groups
and Thermal Reactivity of Nitramines. J. Thermal Anal. Calorim. 77 (2004) 217.
[23] ZEMAN S.: New aspects of impact reactivity of polynitro compounds. Part IV. Allocation of
polynitro compounds on the basis of their impact sensitivities. Propellants, Explos., Pyrotech. 28
(2003) 308.
[24] COFFEY C. S., SHARMA J.: Initiation of Crystalline Explosives due to Energy Dissipated During
Plastic Flow, in: Proc. 11th Int. Detonation Symposium, Snowmass Village, Colorado, August
1998, p. 751.
[25] COFFEY C. S., SHARMA J.: Plastic deformation, energy dissipation and initiation of crystalline
explosives, Phys. Rev. B: Condens. Mater. Phys., 60 (1999) 9365.
[26] MCNESBY K. L., COFFEY C. S.: Spectroscopic determination of impact sensitivities of explosives,
J. Phys. Chem., B 101 (1997) 3097.
[27] ARMSTRONG R. W., ELBAN W. L.: Dislocations in energetic crystals, in: NABARRO F. R. N., HIRTH
J. P. (Eds.), Dislocation in solids, Elsevier B. V., Amsterdam, 2004, p. 405.
[28] ARMSTRONG R. W.: Dislocations assisted initiation of energetic materials, in: VGENKNECHT J.
(Ed.), Proc. of the 8th Seminar New Trends in Research of Energetic Materials, Univ. of
Pardubice, April 2005
[29] FOX P. G., SORIA-RUIZ J.: Fracture-induced thermal decomposition in brittle crystalline solids,
Proc. Roy. Soc. (London) A 317 (1970) 79.
[30] HOLMES W., FRANCIS R. S., FAYER M. D.: Crack propagation induced heating in crystalline
energetic materials, J. Phys. Chem. 110 (1999) 3576.
[31] KO J., ZEMAN S., MAJZLK J., STRNAD J.: Notices to determination of the electric spark
sensitivity of energetic materials, in: VGENKNECHT J. (Ed.), Proc. of the 5th Seminar New Trends
in Research of Energtic Materials, Univ. of Pardubice, 2002, p. 110.
[32] AUZANNEAU M., ROUX M.: Electric spark and ESD sensitivity of reactive solids. Part. II: Energy
transfer mechanism and comprehensive study on E50, Propellants, Explos., Pyrotech. 20 (1995) 99.
[33] STENGACH V. V.: Electric spark sensitivity of lead(II) azide, Fyz. Goreniya i Vzryva 6 (1970) 113..
[34] KISSINGER H. E.: Reaction kinetics in differential thermal analysis, Anal. Chem., 29 (1957) 1702.
[35] KOHNO Y., UEDA K., IMAMURA A.: Molecular dynamics simulation of initial decomposition
process on the unique N-N bond in nitramines in the crystalline state. J. Phys. Chem. 100 (1996)
4701.
[36] OU Y., CHEN B., JIA H., XU Y., PAN Z.: Structural identification of hexanitrohexaazaisowurtzitane,
HanNeng CaiLiao 3 (1995) 1.
[37] MURRAY J. S., POLITZER P.: Computational studies of energetic nitramines, in: BULUSU S.N. (Ed.),
Chemistry and Physics of Energetic Materials, NATO ASI Series, Vol. 309, Kluwer, Dordrecht,
1990, p. 175.
[38] PRICE D.: Dependence of damage effects upon detonation parameters of organic high explosives.
Chem. Revs. 59 (1959) 801.
[39] ZEMAN S.: Modified Evans-Polanyi-Semenov relationship in the study of chemical
micromechanism governing detonation initiation of individual energetic materials, Thermochim.
Acta 384 (2002) 137.
397
[40] SEMENOV N. N.: O nekotorykh problemakh khimicheskoy kinetiki i reaktsionnoy sposobnosti

(Some Problems of Chemical Kinetics and of Reaction Capability). USSR Acad. Sci., Moscow,
1958, pp. 41 and 101, see also German version: SEMJONOW N. N.: Einige Probleme der
chemischen Kinetik und Reaktionsfhigkeit. Akademie-Verlag, Berlin, 1961, pp. 27, 84.
[41] ZEMAN S., KOHLEK P., MARANDA M.: Study of chemical micromechanism governing detonation
initiation of condensed explosive mixtures by means of differential thermal analysis. Thermochim.
Acta 398 (2003) 185.
[42] POSPIL M., APKOV P., VVRA P., ZEMAN S.: Characterisation of Explosive Materials Using
Molecular Dynamics Simulations, in POLITZER P. & MURRAY J. S. (Ed.), Energetic Materials, Part
I, Elsevier B. V., Amsterdam, 2003, p. 49.
[43] SHALASHILIN D. V., THOMPSON D. L.: Monte Carlo variational transition-state theory study of the
unimolecular dissociation of RDX, J. Phys. Chem., A 101 (1997) 961.
[44] PEPEKIN V. I., MAKHOV M. N., LEBEDEV YU. A.: Teploty vzryvchatogo razlozheniya
individualnykh VV (Heats of explosive decomposition of the individual explosives), Dokl. Akad.
Nauk SSSR 232 (1977) 853.
398
HIGHLY EXPLOSIVE
NANOSILICON-BASED COMPOSITE MATERIALS
D. Clment, J. Diener, and D. Kovalev
TU Muenchen - Physik Department E16,
James-Franck-Strasse, 85747 Garching b. Muenchen
Abstract:
We present a new composite energetic material based on nano-silicon. Two
configurations of possible explosive systems can be realized. First, porous silicon (PSi)
itself as explosive material in combination with common oxidizers and second, as
enhancing additive to commonly used high explosives. Commonly, PSi has been produced
from bulk Si wafers by electrochemical etching in an HF/ethanol mixture. This method is
completely compatible with the standard silicon technology and full bulk silicon wafers
can be processed to achieve the required explosive configuration. After filling the pores
with oxidizer a 2D-explosive system can be realized. It reveals a variety of new possible
applications in different industrial fields, e.g. as a novel, very fast airbag igniter.
Unfortunately, this technique does not allow the production of large amounts of PSi and
its costs are extremely high. Therefore, we developed a new concept of large scale
production of PSi powder. It is based on "stain-etching" of commercially available
polycrystalline bulk Si powder. This new technique expands further the variety of new
possible applications of Si nanocrystals in energetic materials. In our presentation we
will explain the preparation techniques and discuss the morphology and the physical
properties of the resulting porous materials. Different parameters of the reaction were
determined for several combinations of PSi and oxidizers by laboratory experiments.
Among them are the time of chemical reactions, the reaction temperature, and the
propagation speed of the shockwave. The energy yield was found to be as high as 9.2 kJ/g
for a stoichiometric ratio of fuel (PSi) and oxidizer.Finally, we will also present the
results of first experiments with PSi and fine polycrystalline Si powder as an enhancing
additive to common high explosives. We will outline the advantages of the composite
energetic materials containing nanosilicon, with respect to commonly used aluminum
powder.
Keywords: nano-silicon, porous silicon- PSi, PSi and oxidizer, explosives
399
THERMAL STABILITY OF THE IMPURITIES

AND THE POLYMORPHS OF HNIW
Maciej Duda*, and Wincenty Skupiski**
* Department of Highenergetic Materials, Faculty of Chemistry,
Warsaw University of Technology, Noakowskiego 3, 00-664 Warsaw, Poland
** Industrial Chemistry Research Institute, Rydygiera 8, 01-793 Warsaw, Poland
Abstract:
2,4,6,8,10,12-heksanitro-2,4,6,8,10,12-heksaazatetracyclo
[5.5.0.05,903,11]dodecane
(HNIW) is highly dense and one of most powerful explosive materials. Four polymorphs
(, , and ) have different molecular conformation, mechanical sensibility and
thermal stability.
Phase transitions in the -, -, -, and -polymorphs of HNIW have been studied as a
function of temperature. High temperature equilibrium solvation studies coupled with
Fourier transform infrared spectroscopy (FTIR) for the identification of polymorphic
conversion indicated a thermodynamic stability order of > > -hydrate > , with the
epsilon polymorph the most thermodynamically stable phase of HNIW. These results are
agreement with data from Differential Scanning Calorimetry (DSC).
The thermal stability is also related with purity of energetic materials. 4,10-dinitro2,6,8,12-tetraoxa-4,10-diazatetracyclo[5.5.0.05,903,11] dodecane (TEX) and 2-acetyl4,6,8,10,12-pentanitro-2,4,6,8,10,12-heksaazatetracyclo[5.5.0.05,903,11]dodecane
are
two main separated and identified impurities. Both these impurities are more
thermodynamically stable than HNIW. Differential Scanning Calorimetry (DSC) studies
showed impure HNIW to be more thermodynamically stable then pure 2,4,6,8,10,12heksanitro-2,4,6,8,10,12-heksaazatetracyclo [5.5.0.05,903,11]dodecane.
Keywords:
HNIW, TEX, polymorphs, impurities, stability, DSC
400
DETONATIVE IGNITION OF METAL/FLUOROCARBON PYROLANTS

Ernst-Christian Koch, and Karl P. Rudolf
Diehl BGT Defence GmbH & Co. KG, Fischbachstrasse 16,
D-90552 Rthenbach a d Pegnitz, Germany
Abstract:
Fuel rich metal/fluorocarbon pyrolants play an important role as either infrared decoy flare
materials or propellant igniters [1]. Recently the viability of metal/fluorocarbon pyrolants as
reactive fragments for enhanced blast effects has been studied in the US [2]. In addition the
behaviour of Magnesium/Teflon/Viton payloads in common IR decoy flares is now
investigated in several countries with respect to sensitivity towards bullet impact and close
detonation of warheads. So far there have been only reports about mechanical shock
ignition of pyrotechnics and propellants with accelerated thin films [3]. We report in this
paper about the detonative ignition of a series of consolidated metal fluorocarbon pyrolants
in close contact to high explosives. Upon subjection to a shock wave metal fluorocarbon
pyrolants ignite. The paper discusses the ignition mechanism and propagation aspects of the
reaction. A series of metal fluorocarbon mixtures based on either magnesium, aluminium
and titanium as well as polytetrafluoroethylene and graphite fluoride are investigated.
Keywords:
magnesium, graphite fluoride, polytetrafluoroethylene, pyrolant,

detonation, ignition, initiation, shock wave
REFERENCES
[1]
[2]
[3]
E.-C. KOCH: Metal-Fluorocarbon Pyrolants. III. Development and Application of Magnesium/

Teflon/Viton(MTV), Propellants, Explos. Pyrotech. 27 2002, 262.
NATIONAL RESEARCH COUNCIL: Advanced Energetic Materials, The national Academy, Press,
Washington DC, 2004.
A. P. HARDT, R. H. MARTINSON: Initiation of Pyrotechnic Mixtures by Shock, 8th Symposium on
Explosives and Pyrotechnics, February 5 7, 1974, Los Angeles, CA, 1974, p. 53
401
INVESTIGATION
OF DEFLAGRATION TO DETONATION TRANSITION
IN POROUS HIGH EXPLOSIVE
BY SYNCHROTRON RADIATION APPROACH
L.A.Merzhievsky*, P.I.Zubkov*, K.A.Ten*, E.R.Pruuel*, L.A.Lukyanchikov*,
V.M.Titov*, B.P.Tolochko**, M.G.Fedotov**, M.R.Sharafutdinov**,
M.A.Sheromov**, I.L.Jogin**, V.V.Julanov***, L.I.Shehtman***, and
V.M.Aulchenko***
* Lavrentyev Institute of Hydrodynamics SB RAS, Novosibirsk, Russia
** Budker Institute of Nuclear Physics SB RAS, Novosibirsk, Russia

** Institute of Solid Chemistry and Mechanochemistry SB RAS, Novosibirsk, Russia
Abstract:
We present facility of new experimental approach for physics explosion. Contemporary
electron accelerator can produce roentgen radiation (synchrotron radiation) with
features that are very attractive for experimental physics at all and for explosion physics
especially. In present paper we discuss one method for detecting density dynamics by
synchrotron radiation.
The main idea of method is next. We can radiograph investigated charge of explosive by
synchrotron radiation beam and detect weakened radiation intensity. Use this
information and appropriate calibration we can restore amount of radiograph matter.
At present we can use radiation beam with knife shape: thickness 0.1 mm and width
15 mm. Gas line detector, which measure weakened radiation intensity, has 256 space
strips with 0.1 mm interval and memory for 32 frames, with time between frames
0.5 mks. As a result of one experiment we have slit roentgen cinema of explosion
process. Such high space and time resolution are very attractive for explosion detecting.
As a one result of method we present some new experimental data for deflagration to
detonation transition in charges of porous hay explosive (petn, rdx), with density near 1
g/cm3. In experiment we measure weakening of transmitted beam through bit core of
investigated charge. We can obtain compression dynamic
dl
versus time and space
duration along charge axes.

As a result in one experiment we obtain all density dynamic for deflagration to
detonation process, from initiation moment to stationary detonation wave.
Keywords:
experimental approach, physics explosion, synchrotron radiation,

deflagration to detonation transition
402
INSENSITIVE INITIATION CHAINS

FOR INSENSITIVE EXPLOSIVE FILLS
Karl Rudolf
Diehl BGT Defence GmbH & Co. KG, Fischbachstrae 16, 90552 Rthenbach, Germany
Abstract:
Based on the work on insensitive high explosive mixtures presented on the G.Int. Seminar and other high explosive materials we are developing insensitive
initiation chains for existing sensitive fuze systems and new designs to fully meet
the existing MIL Std 2105 B and the follow on STANAG 4439. During
presentation technical related issues deducted from the high explosive main
fillings required by the given threat requirements consequences on the
components of initiation chains will be discussed and some of the existing
technical solutions presented.
Keywords:
insensitive high explosive mixtures, MIL Std 2105,

STANAG 4439
403

Siberian State Technological University, prosp. Mira, 82, Krasnoyarsk 660049, Russia
Abstract:
The thermal decomposition of six-membered cyclic N-nitramines, contained in ring one,
two and three nitramine fragments, and also oxy-, gem-dinitro, azido and tetrazole
function, is investigated, using manometer method combined with chromatography,
FTIR-spectroscopy and mass-spectrometry. Influence of functional groups on the rate
and activation parameters of N-nitramines decomposition limited stage is established.
Keywords:
thermal decomposition, N-nitramines, limited stage,

limited stage of decomposition
404
MASS SPECTROMETRIC ANALYSIS OF NOVEL NITRAMINE

EXPLOSIVES - HNIW AND DTIW
Alexander Tsoglin*,**, and Chagit Denekamp*
* Department of Chemistry and Institute of Catalysis Science and Technology
Technion Israel Institute of Technology, Haifa 32000, Israel
** Rafael Ltd., Department M4, P.O.B. 2250, Haifa 31021, Israel
Abstract
The current paper describes the development of mass spectrometric methods for the
analysis of two novel nitramine explosives - Hexanitrohexaazaisowurzitane (HNIW) and
4,10-Dinitro-2,6,8,12-Tetraoxa-4,10-Diazaisowurzitane (DTIW). The methods include
electrospray (ESI) and atmospheric pressure chemical ionization (APCI) techniques for
liquid chromatography mass spectrometry (LC/MS), chemical ionization for direct
introduction (DCI) and gas chromatography mass spectrometry (CI-GC/MS). It is found
that HNIW (438 Da) is detectable using both positive and negative modes of DCI and in the
negative mode, using ESI in combination with LC/MS. Several anions were found to
complex with HNIW, e.g. CF3CO2-, Cl-, Br-, I-, NO3- and NO2-. On the other hand, DTIW
could only be detected using positive DCI and CI-GC/MS, where an MH+ ion (m/z 263) was
formed. An interesting loss of an oxygen atom was observed ([MH-17]+, m/z 246) in the
iBuCI mass spectra of DTIW, which could be explained by the loss of iBuOH from the
[M+57]+ attachment ion. The fragmentation pathways of the two nitramines were further
studied by MS2 experiments. It was found that the main fragmentation pathway of the MH+
ion of DTIW involves the loss of nitrous acid (HNO2). The fragmentations of anion adducts
of HNIW were studied using both CID (using an FTICR and an Ion Trap) and IRMPD (with
an FTICR). Several anion adducts gave rise to [438 Da + anion] ions, which fragmented
back to the neutral HNIW and the added anions. However, Cl-, Br-, NO3- and NO2 afforded
a series of fragments that resulted from the isowurzitane structure. It was found that under
more energetic conditions (e.g. IRMPD and high pressure CID) more informative CID
spectra are produced. It is also apparent that the attached anion affects the course of
fragmentation. Thus Ion trap CID MS2 experiment of the Br- adduct produced the poorest
spectrum with one fragment ion at m/z 371 (for 79Br), attributed to the loss of C2H2[NNO2]2,
while the corresponding [M+Cl]- and [M+NO2]- ions gave rise to HNO2 losses together
with the consecutive loss of C2H2[NNO2]2. Hence the CID spectrum of the [M+Br]- ion was
found to be the best suited for analytical purposes. In order to understand some of the
HNIW fragmentation pathways, theoretical calculations are being used.
Keywords:
HNIW, DTIW, ESI-MS, IRMPD, Ion Trap, CI-GCMS, adduct ions
405
VACUUM STABILITY TEST (VST) CALIBRATION

AND COMPARISON OF MERCURY
AND PRESSURE TRANSDUCER METHODS
Alexander Tsoglin, and Levi Gotlieb
Rafael Ltd., Department M4, P.O.B. 2250, Haifa 31021, Israel
Abstract
VST is the most fundamental method used to assess thermal stability and compatibility of
energetic materials. The volume of gas evolved upon heating under vacuum is measured
and according to the amount of gas a material is declared stable, or a mixture compatible.
This paper compares mercury manometer VST and pressure transducer VST.
The traditional method of measuring the volume of evolved gas is using a mercury
manometer (VSTHg). The desire to eliminate the extensive use of mercury motivated the
development of an electronic version using pressure transducers (VSTPT). Both VSTHg and
VSTPT have been accepted as NATO standards (STANAG 4556 and STANAG 4147).
Comparison between the two methods is essential before abandoning the traditional
mercury base method.
For this purpose a unique calibration method was developed. It is based on a reaction
which evolves a predictable amount of gas in the temperature range of the test. A mixture of
a solid acid which melts in the 90C-120C range and only then reacts with a metal
carbonate releasing a predicable amount of carbon dioxide was looked for. Based on
melting points, glutaric acid was chosen. When melted, the diacid reacts quantitatively with
a simple carbonate, giving carbon dioxide according to the following reaction:
M2CO3 (s) + 2RCOOH (s) ? 2RCOOM (s) + H2O (l) + CO2 (g)
Where M = K or Na.
The volume of CO2 evolved is stochiometrically proportional to the salt weight and only
carbon dioxide remains as a gas at room temperature, hence the amount of gas could be
calculated theoretically. The results obtained from VSTHg and VSTPT were almost identical
and are in excellent agreement with theoretical values. Hence the calibration method is
quantitative, reproducible and may be applied on both VSTHg and VSTPT methods.
In addition, test results from VSTHg and VSTPT of a series of 18 mixtures (energetic and nonenergetic) were compared. A standard deviation of 7.3% 5.8% between VSTHg and VSTPT
was obtained. Therefore the two methods are interchangeable and the VSTPT method was
adopted for stability evaluation.
Keywords:
VST, pressure transducer, mercury, metal carbonate, glutaric acid
406
DEVELOPING AND MAINTAINING SKILLS

IN THE EXPLOSIVES SECTOR
I. G. Wallace
Cranfield University, Shrivenham Campus, Swindon SN6 8LA, UK
Abstract:
The manufacture and use of explosives, propellants and pyrotechnics underpins a
significant part of the European Union economic and industrial activity. An
understanding of explosives science and technology and the competence to harness it is
central to maintaining European Explosives capability, national security, and in
sustaining a competitive European industry. There is a perception and some evidence
that in Europe, competence level in this key technological area is being eroded. In
several member nations a high proportion of the most experienced and knowledgeable
personnel are retiring or nearing retirement. Urgent efforts are therefore underway in
some partner nations to replenish this expertise. The UK together with Sweden,
Norway, Finland and Italy are taking part in a Leonardo Da Vinci programme to
develop a comprehensive framework which describes and categories all of the
competences of workers engaged in the manufacture or use of explosives. This paper
will outline i) the procedures used to develop the competences for UK workers and ii)
the development of National Occupational Standards for the UK and Europe.
Keywords:
Explosives; Skills; Competence; Training and Education
407
NUMERICAL SIMULATIONS OF OIL-WELL PERFORATOR

USING 3D LAGRANGIAN LS-DYNA CODE
Vladislav Adamk*, and Zenon Wilk**
* Department of Theory and Technology of Explosives (DTTX),
University of Pardubice, 532 10 Pardubice, Czech Republic
** Institute of Organic Industry Branch in Krupski Myn,
Zawadzkiego 1 str., 42-693 Krupski Myn, Poland
Abstract:
Fast dynamics of an oil-well perforator was numerically simulated by using 3D
Lagrangian LS-DYNA code covering explosive detonation, liner collapse and jet
formation. In the analysis, the influence of liner mesh density on jet parameters was
studied. It was demonstrated that for a perforator analysed, the liner had to be divided
into around ten elements across the liner thickness. The analysis showed that 3D explicit
LS-DYNA is still an efficient and useful numerical tool for prediction of shaped charge
jets formation, in spite of current effort that is focused on ALE numerical techniques in
this field.
Keywords:
1.
numerical simulation LS-DYNA code, oil-well perforator,

jet formation
INTRODUCTION
Generally, oil-well perforator performance is analysed in two consecutive phases: the

first one covers explosive detonation, liner collapse and jet formation, the second is devoted
to a simulation of shaped charge jet penetration into various targets materials, usually steel
and concrete. This paper deals with the first phase simulation regarding an IOI (Institute of
Organic Industry, Poland) oil-well perforator by using 3D Lagrangian code LS-DYNA.
In present, a prevailing opinion exists that the Lagrangian codes can only simulate the
initial stages of jet formation, mainly because of the element distortion problems, see for
example Jin [1] and Berg [2]. Therefore, the problem solution, currently very often adapted, is
to use multi-material ALE numerical methods, see references mentioned above.
In spite of this, we try to demonstrate that 3D Lagrangian code LS-DYNA can be an
efficient and useful numerical tool for prediction of shaped charge jets formation, i.e. it it can
be useful for the numerical simulation of the first dynamic perforator phase, at the least. In
more, we feel that by this approach the conservation laws involved are very well respected
on contrary to ALE methods where transport terms could be difficult to evaluate in time.
As for results of our analysis, we observed that the perforator liner should had to be
divided into approximately ten elements across the liner thickness in order to obtain real jet
parameters.
408
2.
NUMERICAL PERFORATOR MODEL
The oil-well perforator analysed is composed of a liner, explosive and casing, see Fig.1
where the perforator cross-section is presented. The conical shaped liner is made of copper.
At the top, it has thickness 0.54 mm and at the bottom its thickness is around 1.2 mm. The
outer casing is made of ordinary steel, the explosive is RDX or HMX, both with
fluoropolymers.
Three different finite element (FE) models were generated (due to symmetry, only one
perforator quarter was modelled for each case): case A with five elements across the liner
thickness, model B with eight elements across the liner thickness, model C with ten elements
across the liner thickness. The total number of Lagrangian elements was: model A - 44772,
model B - 154270, model C - 217824. For the model C, a detail of the mesh in vicinity of the
liner top is presented for illustration in Fig.2, note a regular mesh on contact explosive-liner
surface.
Fig 1.
Perforator cross-section
409
Fig 2.
Mesh in vicinity of liner top
The crucial problem for the analysis presented here was a correct description of the
contacts involved: explosive-casing and explosive-liner. After many trial computations it
was decided to model these contacts through LS-DYNA contact type: tied explosive nodes
to casing and liner surface. This contact can be considered too restrictive but the process of
explosive detonation is very fast and following gas products expansion is confined by both
the liner and casing, i.e. without significant explosive detonation products turbulence.
Therefore, we believe that this simulation is quite correct, at least for this specific perforator
case.
3.
MATERIAL MODELS
For the explosives, JWL equation of state was used in form:
p = A 1 R1 0 e
R1
+ B 1 R2 0 e
R2
+ E 0
(1)
where p is pressure, is density, E is internal energy, A, B, R1, R2, are material

parameters.
For the casing and liner, Johnson-Cook constitutive equation was used in form:
= (A + B.e n ) (1,0 + C ln( * )) (1,0 T *m )

T * = (T Tr ) (Tmelt Tr )
410
(2)
where is yield stress, is plastic strain, * is strain rate, T is current temperature, Tr is

room temperature, Tmelt is melting temperature, A, B, C, n, m are material parameters.
The parameters for materials involved are summarized in Tab.1 and 2. The parameters
were recommended by Trzciski[3] and Buchar[4].
Table 1. Explosive parameters
Explosive based on RDX or HMX with fluoropolymers binder ( 4% weight) [3]
Explosive 0, kg/m3
RDXfl
1743.0
HMXfl
1814.0
E0, GPa
10.0
10.5
A, GPa
908.471
1214.843
B, GPa
19.10836
22.18767
R1
4.92
5.16
R2
1.41
1.41
0.31
0.33
Table 2. Solid material parameters

Material
Copper
Casing
4.
0
kg/m3
8960.0
7860.0
A
MPa
85.0
350.0
B
MPa
293.0
275.0
C
0.026
0.022
n
0.31
0.36
m
1.05
1.00
Tr
K
300
300
Tmelt
K
1083
1536
RESULTS
The results presented are only for based on HMX explosive. The jet formation process
for model C is illustrated in Figs.3-6. Note that the jet formed has a conical shape like in the
work of Berg [2].
In Fig.7, there are average jet tip velocity histories for all FE models involved. It can be
seen that model A offers too high jet tip average velocity without reaching a stationary final
velocity while models B and C give quasi identical average velocities with a stationary
average velocity of about 8500 m/s. The average velocity for each model was computed as
average value for adjacent 5 external nodes (lying initially on the external top liner surface
near the liner axis).
411
Fig 3. Liner contours at 0,000 ms
412
Fig 7.
5.
Jet tip velocity for different liner meshing
CONCLUSIONS
The crucial problem for the analysis presented here was a correct description of the
contacts involved: explosive-casing and explosive-liner. It was verified that these contacts
could be modelled using tied explosive nodes to casing and liner surface.
As for analysis results, it was observed that the perforator liner should had to be divided
into approximately ten elements across the liner thickness in order to obtain real jet
parameters.
The analysis demonstrated that 3D explicit LS-DYNA is still an efficient and useful
numerical tool for prediction of shaped charge jets formation, in spite of current effort that is
mainly focused on ALE numerical techniques in this field.
Aknowledgment
The work was performed on a non-commercial cooperation base between the University
of Pardubice and Institute of Organic Industry - Branch in Krupski Myn, Poland. The
authors would like to thank prof. J. Buchar for his help in the field of dynamic solid
materials behavior and prof. W. Trzciski for his contribution regarding JWL explosive
parameters determination.
413
REFERENCES
[1]
[2]
[3]
[4]
QIANKUN JIN, GARY DING, YIANJUN, CUI BINGGI: 3D numerical simulations of penetration of
oil-well perforator into concrete targets, 7th International LS-DYNA Users Conference,
Dearborn, Michigan USA, May 19-21, 2001.
V.S.BERG, D.S.PREECE: Shaped charge induced concrete damage predictions using RHT
constitutive modelling, 30th Annual Conference on Explosives and Blasting Techniques, New
Orleans, Louisiana USA, February 1-4, 2004.
W.A.TRZCISKI: Detonation characteristics of phlegmatised and putty-type explosives with
fluorocarbon binders, Military University of Technology, Poland, Biuletyn WAT Vol. L, Nr 7,
2001.
J.BUCHAR: Private communication, Mendel University of Brno, Czech Republic, January 2005.
414
PARAMETERS OF HEAT EXPLOSION AND DETONATION

OF HEXANITROMANNITE AND NITROTHREEAZOLONE
E.I. Aleshkina, and G.D. Kozak
Mendeleev University of Chemical Technology
125047, Miusskaja sq. 9, Moscow, Russia.
Abstract
The main goal of investigation was to compare the explosion parameters of essentially
different explosives: hexanitromannite (HNM) and nitrothreeazolone (NTO). HNM is
a nitroester, and NTO is insensitive explosive. Explosion hazard of liquid nitoesters
supposes to be very high, failure detonation diameter of them is very low (df ~ 2 mm for
nitroglycerine and nitroglycole), and character oh heat explosion of them is very
intensive. HNM is solid substance, and its detonability was investigated for charges
in cast state and for a solutions of HNM in liquid nitroglycole. The dependence of df of
nitroglycole vs. initial temperature was used for estimation df of HNM in molten state.
The estimation gave a value of critical diameter df = 0.7 mm at t = 120 0C. At d = 1.3
mm in long glass capillary low velocity detonation was registered in cast charge of
HNM, at diameter d = 2.2 mm in such charge high velocity detonation propagated.
Detonation velocity of pressed NTO was measured in steel tubes at density =
1.48 g/sm3, D = 6.34 6.38 km/s. The dependencies of heat explosion delay vs. initial
temperature was measured for both substances. Their kinetic parameters were
calculated using these dependencies.
Keyword:
1.
detonation, heat explosion, detonation velocity, critical diameter
INTRODUCTION.
Substances under investigation are known to belong to sufficient different classes of

high explosives. The first of them is nitroester Hexanitromannite, the second one is
insensitive explosive Nitrothreeazolone. The properties of them that were obtained by means
of computing according to method [1] are collected in Table 1. In spite of that HNM was
synthesized long time ago, almost simultaneously with nitroglycerine, some explosion
properties of it were not be estimated. It detonability (critical or failure diameter of
detonation df), delay time of heat explosion () etc. are unknown. Evidently it was connected
with the fact of unsung of HNM in practice. Heat explosion of NTO was not investigated
too. As to detonability of NTO, its critical (failure) diameter of detonation is known to be
rather large df = 16 mm at density = 1.8 g/sm3 [2]. Reliable estimation of HNM explosion
properties and its arrangement in the raw of high explosives according to explosion hazard
called these data. Slow thermal decomposition of HNM was thoroughly investigated, the
data were analyzed in [3], and parameters of NTO decomposing was published in work [4].
415
Table 1. Calculated parameters of detonation

Explosive *, g/cm3 Qv, kJ/kg PCJ, GPa
NTO
HNM
*
2.
D, m/s
TD, K
1.911
4109
33.12
8493
2796
3.16
1.48
4000
17.02
6948
3218
3.20
1.604
5895
21.63
7498
4535
3.17
1.68
5902
24.34
7796
4469
3.19
A, %
-24.6
7.1
The greatest value is theoretical maximum density, the lowest one is experimental density
of charge
EXPERIMENTAL.
HNM was synthesized in laboratory by means of nitration of mannite with mixture of

nitric and sulfuric acids at temperature t < 20 0C. Ratio of the components was: mannite:
HNO3 : H2SO4 = 1: 5: 10. HNM was washed by a cold and hot (t 50 0C) water, and it was
recrystallized from ethanol, intercystalline acid was neutralized by means of sodium
carbonate adding. Melting point of HNM was Tfus = 110 0C. HNM.
The same sample of NTO, that was investigated in [4], was used in our runs.
Induction period of heat explosion versus initial temperature was investigated in inertial
aluminum thermostat in the range of temperature from 100 to 200 0C. Glass test-tubes were
put into hole in aluminum block and heated to desired temperature. Small mass of substance
(0.1 g HNM or 0.3 g NTO) was introduced into test-tube, delay time of heat explosion was
measured by means of seconds counter.
df of the analogues of HNM (NGL and nitroglycerin) was well know [5], and we tried to
investigate df of HNM solutions in NGL.
Cast charges of HNM was obtained in air thermostat at filling up the glass cylinders
tubes having widening in upper part of them with melted HNM. Before this operation glass
tubes and glass vessel with HNM were placed into thermostat at t = 120 0C and sustained in
it during t=15 20 min for HNM fusing. A canal of tubes previously was closed not tightly
with glass needle, and when liquid was filled up widened part of tube, canal was closed with
light pressing on tube. Crystallization proceeded in opened thermostat during a period t = 5
10 min. Widened part of a tube after crystallization was filled up phlegmatized powdered
RDX. Density of cast charge (d = 5 mm) of HNM was estimated by means of pycnometer, it
was = 1.595 g/sm3.
df of cast HNM was estimated by means of method go no go with charges of various
diameters in glass tubes that were arranged on steel or brass witness-plates. Detonation
velocity was measured by means of streak-camera registration.
416
3.
EXPERIMENTAL RESULTS.
Dependencies of delay time of thermal explosion versus reciprocal temperature of HNM

and NTO are presented in Fig. 1 and Fig. 2. There was not HNM heat explosion at
temperature t < 165 0C at m = 0.1 g. Heat explosion at higher temperature t = 166 167 0C
was not intensive. Their intensity increased at higher temperature, but glass test tube that
contained HNM was not destroyed in the most of runs. Heat explosion of NTO (m = 0.3 g)
occurred at temperature higher t > 244 0C, and delay time of it decreased with temperature
growth. There was no flame and sound effect at heat explosion of NTO.
-2,8
lg(/T ), (s/ )
-3,0
-3,2
-3,4
-3,6
2
lg(/T ) = -20,4 + 7,6(1/)

= 145,6 kJ/mol
-3,8
-4,0
2,15
-2,5
-1
10 /,
2,20
2,25
Fig 1.
The delay time dependence

vs. reciprocal temperature
for HNM.
Fig 2.
The delay time dependence

vs. reciprocal temperature
for NTO.
2,30
lg(t/T ), (s/K )
-3,0
-3,5
-4,0
2
lg(t/T ) = -23,5 + 10,5(1/)

= 201,2 kJ/mol
-4,5
-5,0
1,75
-1
10 /T, K
1,80
1,85
1,90
1,95
2,00
417
12
df, mm
10
8
6
4
Fig 3.
Dependence of df of NGL
vs. HNM content: dark
points are detonation,
open points are failure
of detonation, crosses are
damping of detonation.
Fig 4.
Dependence of df
of mixture NGL/HNM:
90/10 vs. DNT content:
notation are in Fig. 3.
2
CHNM, %
10 11
5 d , mm
f
4
3
2
1
CDNT, %
0
10
15
20
25
30
The limit of solubility of HNM in NGL was found to be 10 %. As one can see in Fig.3,
df of NGL at introducing of HNM did not changed in this interval of solution composition.
The limit of solubility of dinitrotoluene in mixture NGL/HNM: 90/10 was found to be
CDNT = 35 %. One can see the minimum in dependence of df versus DNT content (Fig. 4).
Typical streak camera record of detonation in system DNT/NGL/HNM is preprinted in
Fig. 5.
418
Fig 5. Typical streak camera record of detonation

in system DNT/NGL/HNM.
Results of measuring of HNM df in cast state are presented in Table 2. High velocity
detonation propagated in HNM charges at the diameter d 2.2 mm. Low velocity detonation
was registered in charge at d = 1.3 mm (Fig. 6). The process was stable at length L = 115
mm, and then it damped, the end part of tube (l = 20 mm) was found in the explosion
chamber after experiment. The length of detonated part of charge coincided with extent of
trace in streak camera record.
Table 2. Cast charges of HNM. Experimental results of df and D measurement
D, km/s
d, mm
Length
of charge, mm
D, km/s
Result of run
Calc.
3.0
97
HVD*
2.5
160
HVD
In glass tube
7.39
0.11
7.29
0.21
6.89
0.61
2.4
70
HVD
2.4
110
HVD
2.3
90
HVD
2.2
90
HVD *
1.3
135
LVD
7,50
1.67
HVD is high velocity detonation, LVD is low velocity detonation
Fig 6. Streak camera record of low velocity detonation: d = 1.3 mm, length
of detonation 130mm, D = 1.67 km/s.
419
Detonation velocity of NTO was measured in still tubes (diameter d = 5 mm, wall
thickness = 11 mm). NTO at density =1.48 g/cm3 detonated at powerful initiation with
velocity 6.34 and 6.38 km/s. Detonation velocity that was calculated by means of method [1]
at this density was 6.95 km/s.
4.
DISCUSSION.
The equations that were obtained at mathematical treatment of dependencies log (i/T2)
= f(1/T) at Fig. 1 2 are:
For HNM: lg i/T2 = -20.4 + 7.6/T (correlation factor R = 0.98);
For NTO: lg i/T2 = -23.5 + 10.5/T (correlation factor R = 0.98).
Activation energies are calculated from inclinations of straight lines:
EHNM = 145.6 kJ/mol;
ENTO = 201.2 kJ/mol
These values are found to be in good agreement with the data that was obtained at
treatment of slow thermal decomposition experiment results. Calculation of preexponational
factors with using data in Fig. 1 Fig. 2 connected with some assumptions, and that was
why the coinciding of factors k0 obtained by both of the experimental method was some
worse.
The influence of initial temperature (T0) on df of NGL was investigated in [6]. The
relationships were linear in 1/T0 lg df coordinates. The critical diameter decreased at initial
temperature increasing:
lg df = -1.5 + 521/T0
(df in mm, N0 in K)
(3)
Equation (3) was used to estimate of the detonability of the hexonitromannite at a

temperature above its melting point (m.p. = 110 0C). As shown in Fig 3., there was no
variation of df for NGL when HNM was dissolved in it. The critical diameter stayed constant
at df = 2 mm up to 10% HNM which is the limit of solubility. It was postulated that, if HNM
were a liquid at room temperature, it would have the same df as NGL df=2 mm. Calculation
of dfHNM at t=120 0C gave the value df=0.67 mm.
Calculated df at tf = 110 0C was df = 0.7 mm. Thus df of liquid HNM was found to be in
~ 2 time lesser that of a solid one. Such difference was observed earlier for NG in work [7],
and for NGL in work [6].
When DNT was adding to mixture NGL/HNM: 90/10 (oxygen balance is A = + 0.6 %)
the mixture of oxidizer with fuel formed. Oxygen balance of system investigated A=0% was
at content of DNT CDNT = 0.59 %. Minimum of df at Fig. situated approximately at the same
DNT content, as it usual observed for oxidizer-fuel systems.
420
5.
CONCLUSION.
The data justify the opinion about high explosion hazard of HNM that are congruent
with one of nitroglycerine. Cast HNM charge in weak confinement by the diameter only
d=1.3 mm detonated in low velocity regime of detonation. Df of HNM estimated at it
melting point df = 0.7 mm. Heat explosion of HNM at m = 0.1 was rather intensive, and it
proceeded at m = 1 g with explosion and crashing glass vessel that contained the HNM. Heat
explosion of NTO proceeded without flame. Measured detonation velocity of NTO was
closed to calculated one.
Acknowledgment
Student: M.V. Efanov, took part in carrying out some of the experiments.
We are grateful to Russian Foundation of Basic Research (RFBR) for partial financial
support. (Grant 01-03-32713).
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
SUMIN A.I., GAMEZO V.N., KONDRIKOV B.N., RAIKOVA V.M.: Shock and detonation
general kinetics and thermodynamics in reactive systems computer package, Trans. of 11th
Detonation Symposium (Int.), Bookcomp, Ampersand, 2000, 30-35.
ZHILIN V.F., ZBARSKII V.L.: Methods of production of high explosives. 2Nitrothreeazolone. Khimicheskaya Tekhnologiya, N 5, 2001, p. 6 12.
ALESHKINA E.I.: Kinetic investigation of thermal decomposition, heat explosion, and
detonation of nitroesters. Cand. Sci. (Ph. D.) Thesis (Rus.). Mendeleev Univ. of Chem.
Technol. Moscow, 2004.
KONDRIKOV B.N., SMIRNOV S.P., KANIKOVA YA.A., DOHERTY R.M.: Chemical
kinetics of the macroscopic stages of NTO thermal decomposition. Energetic Materials
Structure and Properties,35th International Annual Conference of ICT, Karlsruhe, FRG, June 29
July 2, 2004, pp. 89-1 - 89-14.
KOZAK G.D., KONDRIKOV B.N., STARSHINOV A.V.: Critical conditions of low and high
velocity regimes in liquid nitrocompounds. Trans. of 11th Int. Detonation Symposium (Int),
Bookcomp, Ampersand, 2000, 81-86.
KONDRATEV V.V.: Elaboration of liquid explosive mixtures on the base of nitroesters.
Cand. Sci. (Ph. D.) Thesis (Rus.). Mendeleev Univ. of Chem. Technol. Moscow, 1983.
APIN A.YA., VELINA N.F.: Failure diameters of detonation crystals of high explosives.
Trans. of Second burning and explosion Symp. (USSR), Chernogolovka, 1969, p. 317-318.
421
COMPARISION OF EXPLOSIVE PARAMETERS

OF PEROXIDES AND OF NITROCOMPOUNDS
S.V. Arinina, and G.D. Kozak
Mendeleev University of Chemical Technology 125047, Miusskaja sq. 9, Moscow, Russia
Abstract:
The ultimate objectives of investigation were benzoyl peroxide and cumene
hydroperoxide. We carried out the experiments with above named peroxides and with
typical explosive (TNT, PETN and NTO) for comparison applying method of Differential
Scanning Calorimeter. The method does not permit to measure the heat of
decomposition of the reactions which are accompanied with gas formation. Comparison
of dependencies of heat flux vs.temperature of benzoyl peroxide decomposition and of
TNT, NTO and even of PENT visually demonstrates that decomposition of peroxide
begins at fusion, and nitrocompounds decompose at rather higher temperature than its
melting point. High explosion hazard of benzoyl peroxide is explained by low
temperature of decomposition, high intensity of heat explosion and high ability and high
burning velocity.
Keywords:
1.
explosion, hazard, peroxide, burning, detonation
NOMENCLATURE
BP Benzoyl peroxide
CH Cumene hydroperoxide
TNT 2,4,6-Trinitrotoluene
PETN Pentaerythritol tetranitrate
NTO Nitrotriazolon
DSC Differential scanning calorimetry
CSB US Chemical Safety and Hazard Investigation Board
2.
INTRODUCTION
Peroxide compounds are widely used in chemical industry, and in polymer industry as
curing agents of polyester and other resins.
Peroxides and hydroperoxides contain peroxide group -O-O-, which is very unstable and
therefore this compounds have explosive properties. Although they are not applying as
explosives they consider to be explosive substances nevertheless, and as a matter of fact are
non-traditional explosives. This conclusion follows from works [1-7]. Theirs explosion can
occur as a result of action of shock wave, heating, friction or pollution. The processing and
especially drying of benzoyl peroxide (BP) in significant quantities is rather dangerous
operation, and it is testified by numerous incidents that are described in literature [6,8].
422
In the Internet site of US Chemical Safety and Hazard Investigation Board (CSB) one
can find the final reports of huge incidents, which were accompanied by fires and
explosions. These reports include photos of destruction, data of researches of the reasons of
incidents and recommendation to avoid similar ones. One of such reports concerns to
explosion of benzoyl peroxide occurred in 2003 in USA at it drying.
The method of differential scanning calorimetry (DSC) was applied in this work to study
exo- and endothermic effects while heating of BP, cumene hydroperoxide (CH) in
comparison with traditional explosives TNT and PETN and insensitive explosive NTO. The
analysis of the references dealing with fusion enthalpy calorimetric data measurements [9,10]
was carried our to justify the working capacity of the metod.
3.
EXPERIMENTAL
The DSC method [11] is based on simultaneous heating (or cooling) of sample
investigating in comparison with standard one. The given speed was regulated by computer
of arrangement, measurement of the compensating heat flux was supported temperature of a
sample within the limits of the given program. The DSC experimental curves are
dependencies of a heat flux (millijoules per second) versus temperature (in isothermal runs versus time). They permit to register the temperature and the heat of fusion, the temperature
of very decomposition starting with great accuracy. Unfortunately the heat of decomposition
could not be measured with sufficient accuracy because of heat dissipation at gas formation.
Mass of a sample was 1-1.5 mg. In present work all experiment were carried out by means of
Mettler Toledo DSC 30 calorimeter. The technique of experiments is presented in work
[11]
.
TNT, PETN, NTO were chosen as traditional explosive BP and CH as nontraditional
ones. TNT was chosen as there were sure calorimetric data heat of fusion (Hfus) in literature
[10]
. There were not sure data of PETN melting heat and that was measured, Hfus = 171 J/g.
Fusion enthalpy of TNT, obtained in our experiment, was coincided precisely with literature
DSC data [9] - Hfus = 103 J/g. This coinciding gave an opportunity to judge about
reliability of our measurements. According to earlier sources [10] this value was 92.3 J/g but
it was rather close to enthalpy obtained by means of DSC method.
4.
RESULTS AND DISCUSSION.
The estimation of explosive parameters of BP, and CH was carried out in the previous
works [1-8]. Both of the substances were shown to be weak explosives that were capable to
burn without access of air, to heat explosion and to detonate with a low (1-2 km/s) velocity.
The sample of BP was heated in calorimeter with speed 10 0C/min. The results are
reprinted in Fig.1. Melting of BP occurred at t = 108 C after the linear stage of heating. The
substance began to decompose at t = 109 C sharply at once after melting, and it was
impossible to separate sharp peak of decomposition and endothermic effect of fusion and to
calculate enthalpies of every process.
Fig.2 illustrates the dependence of a heat flux of decomposition of CH versus
temperature. Heating rates were 10 and 20 0C/min. The substance began decompose slowly
at t = 105 C after the linear stage and maximum temperature was t = 158 C.
423
35
128f
m=3.11 mg
dH/dt,mW
30
25
109f
m=3.76 mg
20
15
10
5
0
T, C
-5
50
30
100
150
200
250
300
Fig 1. Heat flux vs temperature

at BP decomposing.
Heating rate 10 0C/min
350
dH/dt,mW
20
302f
m=1.35 mg
258f
10
m=1.08 mg
-10
0
T, C
80
100
120
140
160
180
Fig 2. Heat flux vs temperature

at CH decomposing.
Heating rate 10 C/min.
Cumene hydroperoxide is a liquid, containing not less than 89 % of neat hydroperoxide,

showed two maximums in calorimetric curves during decomposition. The decomposition
reactions of CH are known from literature:
65 + 3- - 3 + 309,6 /
(1)
653 + 3 + 230 /
(2)
65(3)2
424
According to the first reaction, phenol and acetone are produced in industry [12]. Both of
the reactions are exothermic and give volatile acetone and methanol as products of reactions.
Probably, the form of calorimetric curve with two peaks is a result of the exothermic effect
of decomposition and endothermic effects of evaporation of volatile substances that occur
simultaneously. The heat produced at reaction 1 and 2 proceeding without evaporation of
methanol and acetone should be in 1.5 - 2 times greater then registered one. The values of
heat evaluation, registered in experiment, thermal effects of reactions 1 and 2, and
calculated heat of explosion of cumene hydroperoxide according method [5] are presented in
Table1
As for BP there was an endothermic peak at t = 108 C (Fig.1) that corresponded to
fusion. Decompositions began immediately after fusion. A fusion enthalpy are estimated
with great precise by means of DSC method, but in the case of BP it was practically
impossible to do because of overlapping of decomposition effect on fusion peak. The heat of
reaction, registered by calorimeter was approximately 0.56-0.57 from thermodynamically
calculated explosion heat.
Decomposition reaction of BP is known from literature:
C6H5(O)C-O-O-C(O)C6H5 C6H5(O)C C6H5 + CO2 +100.6 J/mol
(3)
C6H5(O)C-O-O-C(O)C6H5 C6H5-C6H5 + 2CO2 + 167.6 J/mol
(4)
Both of them are all exothermic ones.

Table 1. DSC registered decomposition heat (HDSC), and calculated explosion heat
of benzoyl peroxide and cumene hydroperoxide.
Substance
DSC,
J/g
g/mol
Qv,
/Qv
J/g
1383
BP
1380
0,57
242,2
2446
0,56
1004
CH
982
0,30
152,2
3141
0,31
Registered decomposition heat of CH was lower in 3 time than thermodynamically

calculated explosion heat.
In spite of impossibility of precession heat decomposition estimation it was useful to
compare calorimetric data that obtained for crystal BP with traditional explosives. The
comparison gives visual demonstration of peroxide hazard.
This comparison presented in figures 3-5. The accurate endothermic fusion effects and
exothermic decomposition peaks presented in all curves. BP heightened sensibility to
heating in compare even with nitroester (PETN) is seen clear: fusion of BP at 108 C
immediately leads to heat explosion whereas the difference between melting temperature and
initial decomposition temperature of PETN is approximately 60 degrees. This difference is
425
rather greater in case of TNT. As for the insensitive explosive - NTO, the melting (t~180 0C)
also foregoes to decomposition but the last one starts at rather higher temperature t = 280 C.
High explosion hazard of BP probably connects with its high ability to burning [7,13,14],
with high sensitivity to heating (Fig. 3-5) and high intensity of heat explosion [8]. BP burns
faster than nitrocompounds [13], and it able to burn in even in vacuum [14] in contrary to
nitrocompounds
30
dH/dt, mW
20
302f
m=1.35 mg
305f
m=1.54mg
10
0
O
T, C
-10
50
Fig 3.
100
150
200
250
300
350
Heat flux comparison of BP (302f) and TNT (305f).
426
400
30
dH/dt, mW
25
20
302f
m=1.35 mg
15
303f
m=1.41mg
10
5
0
-5
-10
-15
T, C
-20
50
Fig 4.
100
150
200
250
300
350
Heat flux comparison of BP (302f) and PETN (303f)
160
dH/dt, mW
140
120
100
218f
m=2.210 mg
80
60
302f
m=1.35 mg
40
20
0
-20
50
Fig 5.
T, C
100
150
200
250
300
Heat flux comparison of BP (302f) and NTO (218f).
427
350
5.
CONCLUSION
The properties of BP and CH were studied by means of differential scanning calorimetry

(DSC). The calorimetric data were obtained and compared with characteristic DSC curves
of typical explosives: TNT, PETN and insensitive explosive NTO.
It was shown, that peroxides have lower decomposition temperature resulting to heat
explosion in comparison with typical explosives. The decomposition of BP begins
immediately after its melting i.e. at T = 108-109 C. High explosion hazard of BP connects
with its high ability to burning, with high sensitivity to heating in contrary to
nitrocompounds, and with high intensity of heat explosion.
Acknowledgment
Students: I.A. Burtseva, A.V. Muratova, and I.G. Chumakova took part in carrying out
some of the experiments and calculation.
428
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
KOZAK G.D., AKININ N.I., RAIKOVA V.M., ARININA S.V.: Cumene hydroperoxide explosion
hazard characteristics study,. Chemical industry. 2002, 9, p.48-54.
ARININA S.V., KOZAK G.D., AKININ N.I.: Cumene hydroperoxide explosion hazard study,
Works of XVII International young scientist conference, 2002 "Progress in chemistry and
chemical technology". Mendeleev Institute of Chem. Technology, ., 2002.
AKININ N.I., KOZAK G.D., RAIKOVA V.M., ARININA S.V.: The Parameters estimating an
exhlosion hazard of organic peroxides, Trans. of XI International conference of organic and
organoelement peroxides chemistry . IChF RAN, 2003, p. 220-221.
Kozak G.D., Akinin N.I., Raikova V.M., Arinina S.V.: Explosion hazard of some organic
peroxides, Proc. of The 6th seminar New trends in research of energetic materials, Univ.
Pardubice, CSR, 2003, pp.173-181.
ALESHKINA E.I., MURATOVA A.V., KOZAK G.D: Study of organic peroxides explosion hazard
parameters, Trans. of XVII International young scientist conference, 2002 "Progress in
chemistry and chemical technology"., Mendeleev Institute of Chem. Technology ., 2003.
AKININ N.I., ARININA S.V., KOZAK G.D., PONOMAREV I.N.: Parameters, necessary to define
explosion hazard of cumene hydroperoxide, benzoyl peroxide and cyclohexanone peroxide,.
Chem. Prom, 2003, 11, p.50-56.
ARININA S.V., BURTSEVA I.A., MURATOVA A.V., AKININ N.I., KOZAK G.D.: Study of
parameters, necessary for explosion hazard estimation, Trans. of XVIII International young
scientist conference, 2003 "Progress in chemistry and chemical technology". Mendeleev
Institute of Chem. Technology, ., 2003.
NIKOLAY.I. AKININ, SVETLANA.V. ARININA, GEORGY.D.KOZAK, IGOR.N. PONOMAREV: The
explosion parameters of benzoyl and cyclohexanone peroxides, The 7th seminar new trends in
research of energetic materials, univ. Pardubice, csr, 2004, pp.409-418.
INTERNET: Data Base of National Institute of Science and Technology, (Issue: Acree W.E. Jr.
Therochim. Acta, 1991, 37).
E.YU. ORLOVA.: Chemistry and technology of high explosives, Chemistry, 1973. P. 687.
V.A. BERSHTAIN., V.M. EGOROV: Differential Scanning Calorimetry in polymer physical
chemistry, Leningrad Chemistry 1990.
B.D. KRUZHALOV, B.I. GOLOVANENKO: Sovmestnoe poluchenie fenola i atsetona (Rus.),
M.:Goskhimizdat, 1963, 87.
ANTONOVSKII V.L.: Organicheskie perekisnye initsiatory (Rus.), M.: Khimiya, 1972. P. 230.
A.E. FOGELZANG, V.YA. ADZHEMYAN, ET AL: Investigation of burning of benxoyl
peroxide, Voprosy Teorii Cond. Vzryvchatykh System (Rus.). M.: Mendeleev Institute of
Chemical Technology. 1980. V. 112, 67.
429
2-NITRIMINO-5-NITROHEXAHYDRO-1,3,5-TRIAZINE:
STRUCTURE AND PROPERTIES
Alexander M. Astachov*, Alexander D. Vasiliev**, Maxim S. Molokeev**,
Andrew A. Nefedov*, Ludmila A. Kruglyakova*, Vitaliy A. Revenko*,
and Eduard S. Buka*
* Siberian State Technological University,
Prosp. Mira 82, 660049 Krasnoyarsk, Russia
** Institute of Physics RAS (Sib. branch),
Akademgorodok, 660036 Krasnoyarsk, Russia
Abstract:
The 2-nitrimino-5-nitrohexahydro-1,3,5-triazine (NNHT) structure was solved by
methods of X-ray analysis. Kinetics of a thermal decomposition was investigated under
isothermal conditions by manometric method in solid phase and solution. The
decomposition products were investigated by mass-spectrometric method. The impact
sensitivity of NNHT was investigated with the help of fallhammer test. The sensitivity to
explosion by heat was determined as the temperature of flash and time to explosion
delay. Energetic and detonation parameters of NNHT were estimated using
thermodynamic method with BKW equation of state and by simple correlation methods.
Keywords: high explosives, nitramines, nitrimines, X-ray structure,
thermal decomposition, impact sensitivity
1.
INTRODUCTION
2-Nitrimino-5-nitrohexahydro-1,3,5-triazine (NNHT) is considered in different countries

as a perspective energetic compound during last fifteen years [1].
NNO2
NH
HN
NNHT
N
NO2
First report about the compound in an accessible literature had been done by Chinese
scientists in 1989 [2]. In the USA patent [3] there is cited a method of the compound receiving
and considered some of its physicochemical properties. Authors of works [3, 4] give to NNHT
properties of an insensitive energetic material which has essentially lesser impact sensitivity
than RDX. In another USA patent [5] there are proposed insensitive propellants for an
artillery or for a tank system on a base of NNHT. According to the patent, propellants on a
base of NNHT, in comparison with regular American propellants M30A1, M43 and JA2,
have an enhanced safety, allow raising an impetus without increasing of flame temperature
and have extremely low pressure exponents and high burning rate coefficient.
At the same time in the work of Australian scientists [6] there was shown that an impact
sensitivity of NNHT on application of another method to define it (Rotter Tests vs. Bureau
of Mines Tests [3, 4]), is close to RDX or even exceed the latter.
430
In the works [7, 8] there were proposed methods of NNHT synthesis which differ from
considered in [2, 3] one. In the recent work [9] there described a synthesis and properties of
onium salts of NNHT with nitric and perchloric acids. Besides, NNHT is the mother
compound for synthesis of other energetic compounds [2, 10].
The aims of the present work are an investigation of NNHT structure, a research of
kinetics and thermal decomposition mechanism, impact sensitivity (with Russian apparatus)
and an estimation of the compound energetic characteristics.
2.
SYNTHESIS
NNHT was synthesized from the nitroguanidine (NQ) as described previously Cliff et al. [8]:
NNO2
NNO2
H2N
NH2
CH2O + t-BuNH2
NH
HN
N
NNO2
HNO3/NH4Cl
NH
HN
87%
89%
NO2
The compound was refined by recrystallization from water. For all that, there arise
transparent needle-like crystals good for X-ray analysis.
3.
X-RAY ANALYSIS, UV AND FTIR SPECTROSCOPY
The NNHT structure was investigated by the single-crystal X-ray technique on an

automated KUMA-diffraction KM-4 four-circle diffractometer (T=293(2)K, variable scan
speed, CuK radiation, incident beam graphite monochromator). The structure was resolved
by direct methods and refined with full-matrix least-squares refinement in anisotropic
approximation (isotropic for hydrogen atoms) using SHELX-97 software [11]. Hydrogen
atoms were located in a difference Fourier synthesis.
Molecular and crystal structures are shown in Figures 1 and 2. Bond lengths and angles
are represented in Table 1. The single crystal belongs to monoclinic system, space group
P21/c, with parameters: a = 9.3907(4) , b = 8.5862(8) , c = 9.0100(5) , = 88.851(4),
Z = 4, V = 726.334(1) 3, dx-ray = 1.738 g/cm3. The hexacycle of a NNHT molecule has halfboat (half-chair) conformation; the structure of nitrimine fragment of a molecule corresponds
to the structure of analogous fragments of other nitroguanidine derivatives [12]; as in latter,
the nitroguanile fragment is plane. Because of delocalization of -electron density bond
lengths CN, NN and NO have intermediate values between the values characteristic for
corresponding single and double bonds. Like other nitroguanidines, the planar conformation
of nitrimine fragment is stabilized by intramolecular N3H1O1 hydrogen bond (Figure 1).
The peculiarity of nitramine fragment consists in considerable pyramidality of N5
nitrogen atom (the sum of bond angles is 337.6). An absence of conjugation between
unshared electron pair of amine nitrogen atom N5 and delocalized -electron density of a
nitro-group, connected with this atom, is the reason of comparatively long N5N6 bond
[1.409(2)]. This peculiarity of the molecular structure facilitates breaking of the bond under
the thermal decomposition.
Length of N5N6 bond is slightly less in NNHT than the length of the least strong N
NO2 bond in RDX [1.398(3), and the sum of bond angles at the amine nitrogen 347.9 [13].
431
O4
N6
H6
C3
H2
O3
N4
H5
N5
N2
C2
C1
N1
N3
H3
H4
H1
O2
O1
Fig 1.
The molecule of NNHT shown with 50% probability displacement

ellipsoids for non-H atoms; dashed line indicate intramolecular
hydrogen bond.
b
a
Fig 2.
NNHT molecular packing; dashed lines indicate hydrogen bonds.
432
Table 1. Bond lengths () and angles () for NNHT

N1N2
1.343(2)
N4C3
1.448(2)
N1O1
1.242(2)
N2C1
1.357(2)
C2N5
1.452(2)
N1O2
1.230(2)
C1N3
1.324(2)
C3N5
1.456(2)
N6O3
1.212(2)
C1N4
1.323(2)
N5N6
1.409(2)
N6O4
1.215(2)
N3C2
1.454(2)
N1N2C1
119.5(1)
N3C2N5
110.0(1)
O2N1N2
114.5(1)
N2C1N3
127.8(1)
N4C3N5
109.7(1)
O1N1O2
121.7(2)
N2C1N4
112.8(1)
C2N5C3
111.3(2)
O3N6N5
116.9(2)
N3C1N4
119.4(1)
C2N5N6
113.1(1)
O4N6N5
117.9(2)
C1N3C2
122.7(1)
C3N5N6
113.2(1)
O3N6O4
125.1(2)
C1N4C3
122.0(1)
O1N1N2
123.8(1)
The 2N3, C2N5, C3N4 and C3N5 bond lengths (1.4481.456 ) correspond to
single bonds and indicate an absence of conjugation between nitrimine and nitramine groups.
Molecular formula of NNHT, taking into consideration the geometric structure and electron
density distribution, is shown in Figure 3.
O
N
H
N
Fig 3.
N O
H O
The delocalization of -electron density in the NNHT molecule.
Molecules of NNHT in a crystal are connected with each other in pairs by two hydrogen
bonds (Figure 2). Table 2 contains geometrical parameters of intra- and intermolecular Hbonds.
Table 2. Hydrogen-bonding geometry (,) for NNHT
DH...A
DH
H...A
D...A
DH...A
N3H1...O1
0.84(2)
1.98(2)
2.590(2)
129(2)
N4H2...N2
0.83(2)
2.10(2)
2.929(2)
172(2)
More detailed crystallographic structure data of NNHT will be considered in other

publication.
433
The UV absorption spectra of NNHT in water solution were recorded on the Shimadzu
UV-1601 spectrophotometer. The UV spectrum of NNHT show an absorption at
wavelengths max1 = 206.0 nm and max2 = 270.5 nm with logarithm of extinction
coefficients lg 1 = 3.835 and lg 2 = 4.152 (M1cm1 for ), accordingly. The UV spectrum
of NNHT is typical for nitroguanidine derivatives and usually have intensive maximum at
wavelengths 260290 nm. This maximum is associated with n * transition in a nitrimine
fragment, for example, in a nitroguanidine max2 = 265 nm and lg 2 = 4.162 (Figure 4).
A
1.4
NQ
1.2
N N HT
1
0.8
0.6
Fig 4.
UV spectrums of NNHT
and NQ in water solution
0.2
with concentration 104 M
0
190
210
230
250
270
290
310
330
350
370
390 n m
(thickness of the UV cell
is 1cm).
The infrared spectra (4000650 cm1) were obtained on the Nicolet IMPACT-400D
FTIR spectrometer with the IR microscope SpectraTech INSPECT IR+. IR spectra were
recorded at room temperature and 4 cm1 resolution. Figure 5 shows the IR spectrum of
NNHT.
0.4
1600
0.36
0.34
0.32
0.30
0.28
0.26
0.24
1124
1296
1244
0.18
1388
1570
0.20
969
0.04
908
1033
828
0.06
2970
2902
3214
0.08
3121
0.10
3045
3327
0.12
761
0.14
715
0.16
1486
1441
Absorbance
0.22
0.02
3500
3000
2500
2000
1500
1000
Wavenumbers (cm-1)
Fig 5.
The FTIR spectrum of NNHT.
Table 3 summarizes the frequencies and proposed assignments. Tentative frequencies

assignments have been done by analogy with the assignments for nitroguanidine [14, 15].
434
Table 3.
Observed IR frequencies of NNHT and their tentative assignments
Wavenumber, cm1
Wavenumber, cm1
Assignment
Assignment
656 vw
1441 vw
(CH2)
715 m
1486 m
as(NO2) nitrimine
1570 s
as(NO2) nitramine
761 m
828 w
N O 2 , N -N and
ring deform ation
1600 vs
908 m
2902 vw
969 m
2970 w
1033 m
(NHCNH)
3045 m
1124 s
(CNH)
3121 m
1244 s
1296 s
as ( N C N )
(NH) and
as , s(CH2)
3214 m
s ( N C N )
s(NO2) nitramine
3327 m
1388 s
s(NO2) nitrimine
vs = very strong; s = strong; m = medium; w = weak; vw = very weak
4.
THERMAL DECOMPOSITION AND EXPLOSION TEMPERATURE

TEST
Kinetics of NNHT thermal decomposition has been studied under isothermal conditions
by manometric method using glass compensation zero-manometers Bourdon type (Soviet
manometric method) [16]. Thermal decomposition has been studied in solid phase and in
diluted solutions of phenyl benzoate (PhB) (low polar aprotic inert solvent).
Gaseous products of thermal decomposition were analyzed in a mass-spectrometer MI
12-01-V (USSR) with special system for sample input [17] under the following conditions:
pressure inside the ion source 1104 Pa; accelerating voltage of EI source 3 kV; electron
energy of EI source 35 eV; cathode current 0.6 A; multiplier voltage 2 kV; the amount of
NNHT sample used is about 3 mg. The mass-spectrometer resolution allows distinguishing
between molecules and molecular fragments with close values of m/e (N2O and CO2, N2 and
CO etc.), but does not allow to say about comparative quantitative contents of water in
decomposition products.
The isothermal kinetic curves of thermal decomposition in a solid phase have an S-shape
appearance (Figure 4a). Ratios m/V and S/V (ratio of sample weight to volume of vessel and
ratio of surface of vessel to volume of vessel, correspondingly) affect decomposition rate.
So, at the temperature 175 when the ratio m/V increases from 104 to 102 g/cm3, a rate
constant of a thermal decomposition on the initial part of a kinetic curve becomes 3.3 times
less (from 11.8104 to 3.6104 s1). Packing of vacant volume of reaction vessel with glass
capillaries for the purpose of ratio S/V increasing raises the rate of a thermal decomposition
at 175 in five times.
Thermal decomposition in solution is satisfactorily described by a kinetic equation of the
first order to the extent of transformation of 5055% (Figure 4b). The change in a solution
435
concentration does not affect decomposition rate (studied at 0.55%). In a case with S-shape
appearance kinetic curves, rate constants of thermal decomposition reactions were defined
via initial rates of decomposition with assumption of first-order reaction.
190oC
V, cm 3/g
a)
185oC
360
V, cm3/g
b)
180oC
240
180oC
300
175oC
170oC
180
240
160oC
120
180
150oC
170oC
120
60
140oC
60
0
0
0
80
160
240
320
400
480
t, min
40
80
120
160
200 t, min
Fig 6. Kinetic curves of thermal decomposition of NNHT at various temperatures in a solid

phase at loading density m/V = 1103 g/cm3 (a) and in 1% PhB solutions (b).
The rate constants of thermal decomposition and activation parameters are shown in
Table 4, and Arrhenius temperature dependences are shown in Figure 6.
Table 4. Kinetic parameters of thermal decomposition of NNHT
Constant rate k, s1
T, C
Solid phase
(m/V = 1103 g/cm3)
1% PhB solution
150
5.67105
160
1.42104
170
0.68105
2.28104
180
1.92105
3.97104
185
3.23105
190
6.88105
Ea, kJ/mol
193.3
100.9
lgA
17.59
8.25
0.9950
0.9908
436
ln k
-8
-9
-1 0
-1 1
-1 2
2.15
-1
2.35 1000 / T, K
2.25
Fig 7. Arrhenius plot for the rate

constants of thermal
decomposition of NNHT
in a solid phase (1) and in
1% PhB solution (2).
During the thermal decomposition of NNHT there arise NO2 (m/e = 46), N2O (m/e =
44), NH2CN (m/e = 42) and HCN (m/e = 27, 26). Ratio of maxima intensities NO2/N2O =
1:1.62.2 at temperatures 175180C. The maximum with m/e = 30 corresponds to NO. The
latter is generated under the action of electron impact from N2O and NO2, but not as a result
of secondary oxidation-reduction reaction passing with the assistance of NO2, because of in
primary products of thermal decomposition there was not discovered CO and CO2 peaks,
whish usually accompany to oxidation processes [18]. Because of considerable contents of
NO2 in NNHT thermal decomposition products and low strength of NNO2 bond in a
nitramine molecular fragment (Table 1) one can suppose a possibility of primary breaking
just this bond. The rate of NNHT decomposition in a solution 2030 times as much then in a
solid phase (Table 3, Figure 5), that indicate considerable inhibitory effect of a crystal lattice
on the thermal decomposition process [16]. Therefore activation parameters which were found
for thermal decomposition of NNHT in a solid state not allow saying about limiting stage of
the thermal decomposition. S-shaped character of kinetic curves and an affect of ratios m/V
and S/V on a rate of the thermal decomposition make the found activation parameters even a
greater extent formal. On the other hand the activation parameters of the NNHT thermal
decomposition of in solution (Table 4) are not typical for radical gomolysis reaction of N
NO2 bond, and a contents of N2O in products of thermal decomposition is higher than NO2.
Therefore it is not inconceivable that a trigger-linkage of NNHT decomposition is the
nitrimine fragment of a molecule with a primary stage of a proton transfer, as was shown
recently for range variety of other nitrimines [19, 20]. Equality of pro et contra arguments does
not allow to conclude unambiguously about the first stage of a thermal decomposition.
Probably, the thermal decomposition of NNHT proceeds at once two-way:
O2N N
NH2
C=O + N2O
N
NH
2
O2N N
NH
NNO2
NH
NO2 +
N
NH
NNO2
H2O
O2N N
N
C
N
N
O2N N
N
N
NO2
N
m/e = 96
In a mass-spectrum of thermal decomposition products appear peaks which responsible

for a decomposition or a fragmentation of particles both after the primary detachment of NO2
and after the initial decomposition of a nitrimine group with opening of a triazine cycle:
C3H4N4 (m/e = 96), C2H2N2O (m/e = 70), NH2CH2NCH2 (m/e = 58), CH2NCO (m/e = 56).
437
Sensitivity to explosion by heat was determined as the temperature of flash and time to
explosion delay. The known method is used [2123]. The sample of explosive (10 mg) is
placed in a 8 detonator shell, and the lower end of this shell is plunged into a bath of
molten Woods metal at a measured temperature (T). The time required to cause explosion or
flashing is noted with a stop watch. Each experiment was repeated ten times and average
values are fixed. Then the similar operation is repeated with another temperature. The
specified method allows most simply to find activation parameters of thermal decomposition
of explosive compounds. However, the accuracy of definition of activation parameters using
such way is, usually, low.
The temperature time to explosion curve for NNHT is shown in Figure 8a. To find
activation parameters from relationship of temperature time to explosion, data processing
in coordinates ln = f(1/T), shown in Figure 8b, were applied.
b)
3.5
40
3.0
30
ln[(s)]
Time to thermal explosion, s
4.0
a)
50
20
10
Fig 8.
2.0
1.5
1.0
1.90
0
200
2.5
210
220
230
240
Temper at ur e, oC
250
1.96
2.02
2.08
1000/ T, K -1
Relationship between the temperature and time to explosion for NNHT (a) and
plot of ln vs. 1/T (b)
This relationship can be expressed by the following equation:

ln = 27.47 + 14930.8/T, R = 0.9937
From this equation the values of energy of activation (Ea) and pre-exponential factor (A)
are found. The calculation is based on the following expression:
= {cpRT 2/(Qexpl. Ea A)}exp(Ea/RT),
where cp heat capacity (we used approximate values for heat capacity organic
explosives cp = 1.25 J/(gK)); R ideal gas constant, 8.314 J/(molK); Qexpl. heat
decomposition reaction (J/g), identical with heat of explosion (calculated in the chapter 6).
Calculated activation parameters of thermal explosion of NNHT are represented below:
Texpl.( =5 c) = 240 C
T = 206248 C
Ea = 124.1 kJ/mol
lg A = 9.509.57
average lg A = 9.53
438
The obtained values of activation parameters of NNHT and calculated out of them rate
constants visibly differ from obtained in manometric experiments parameters (Table 4). This
fact may be concerned both large error of activation parameters values from explosion
temperature test and/or greatly discriminate experimental conditions.
5.
IMPACT SENSITIVITY
Huang and Rindone [3, 4] investigated an impact sensitivity of NNHT using Bureau of
Mines fallhammer apparatus. They have determined that the 50% explosion probability is
observed under dropping of 2 kg weight from 80 cm height, whereas for RDX this value is
equal to 3242 cm (in the patent [5] 89 cm vs. 22 cm ). On the basis of the fact they
concluded that NNHT is an insensitive high explosive.
However, Australian scientists [6], using Rotter apparatus tests, have shown that NNHT
in other test conditions may possess higher impact sensitivity. A 5 kg drop-weight was used
to determine the Figure of Insensitiveness (F of I), average explosiveness (X*) and Hazard
Index (HI) (Table 5).
Table 5.
Rotter tests results for NNHT, NQ and RDX [6]

Rotter Apparatus Tests
NNHT
NQ
RDX
Figure of Insensitiveness F of I
55
100105
80
Average explosiveness X*
39.4
3.8
51
Hazard Index HI, m1
64.2
3.8
88.5
From article [6] it follows that NNHT is easiest to initiate than RDX (the less index F of
I, the most sensitive). However, X* and HI indexes indicate on somewhat lesser disposition
of a reaction to the propagation after initiating NNHT than RDX. In any case, a sensitivity of
NNHT was much higher then of nitroguanidine (NQ).
In this article sensitivity to impact of NNHT was investigated with K-44-II Russian
drop-weight impact machine [22]. Experimental conditions: 5 kg weight drop hammer; the
sample of explosive is placed in the plunger assembly 1 [21], consisting of two steel rollers
with diameter 10 mm and height 12 mm [21]; samples weight is 30 mg. The results of tests are
collected in Table 6, the curves of sensitivity are shown in Figure 9. Each point was obtained
by results of 25 trials. For comparison the data for PETN are given. Samples of RDX with
weight 50 mg are used as a standard for K-44-II apparatus (f = 488% for H = 25 cm).
Table 6.
Impact sensitivity of NNHT
Load level (drop height) H0, cm
11
14
20
25
30
Explosion probability f, %
32
56
80
84
92
From the data obtained it follows that the sensitivity of NNHT approximately
corresponds to PETN, i.e. it is impossible to refer the compound to an insensitive high
explosive.
439
f,%
100
NNHT
PETN
80
60
RDX 50 mg
40
20
Fig 9.
12
16
20
24
28
32
36
40
44 H, cm
Explosion probability (f) for NNHT vs drop height of load (H); for comparison the
data for PETN are given.
Thus, a modification of test conditions and used apparatus has strong affect on a result
and on our estimation of a compound sensitivity. It is necessary to consider more correct
those conditions under which maximal sensitivity is exhibited, since just they allow estimate
adequately an explosive application safety.
The best method of the explosive impact sensitivity determination is a method of critical
pressure of initiation [24], however it needs in use of complicated apparatus engineering. To
estimate critical parameters of initiation by impact [24] we have tried to use a numerical
method [25] and activation parameters of thermal decomposition obtained by us (chapter 4 in
this paper). Calculated value of Pcr. for NNHT less then 600 MPa, whereas for RDX Pcr. =
700 MPa [24, 25]. In calculations we used manometric data for solid phase. If use data for
NNHT in a solution or activation parameters from explosion temperature test then we could
not get reasonable values of Pcr..
6.
ENERGETIC AND DETONATION PARAMETERS
Energetic and detonation parameters of NNHT were estimated using thermodynamic

method with BKW equation of state [26, 27] at various sets of parameters (BKW-RDX, -R, RR, -C) and, also, by simple correlation methods [27, 28]. The calculation was performed at
maximum density of NNHT. The condensed carbon formed by detonation was taken into
consideration in diamond (BKW-RDX, -R, -C) and ultra fine diamond (BKW-RR) phases.
The experiment and thermodynamic calculation show that diamond phase is preferable in
Chapman-Jouguet point for high density explosives [29]. Table 7 shows averaged values
obtained by the above mentioned methods. Experimental values of detonation parameters for
RDX [30] are shown in Table 7 for comparison.
440
Table 7. Some energetic and detonation parameters of NNHT and RDX.

NNO2
Explosives
HN
O2NN
NH
NNO2
N
NO2
N
NNHT NO2
RDX
Brutto-formula
C3H6N6O4
C3H6N6O6
Molecular weight
190.12
222.12
Oxygen balance, %
42.1
21.6
Hf 0, kJ/mol
+68.2 [3, 4]
+61.5
0, g/cm3
1.738
1.70
1.80
1.70 g/cm3
Qexpl., MJ/kg
5.1752.2%
5.40
V0, m3/kg
0.7670.85%
0.7621.70 g/cm3
D, km/s
8.282.3%
8.39
8.77
PCJ, GPa
28.45.5%
31.4
36.1
TCJ, K
2252 a) 3543 b)
3740
3520
3.229.3%
2.83
2.83
Notes: a) BKW-RDX; b) BKW-C.
It is clear from Table 7 data that NNHT distinctly yields to RDX by its detonation
parameters.
7.
CONCLUSION
NNHT is a hybrid of NQ and RDX by its molecular structure. It expected that the fact
will lead to the combination of positive properties peculiar to both compounds low
sensitivity of NQ associated with a developed net of intermolecular H-bonds, and high
power of RDX [6]. However that did not happened. X-ray analysis showed that NNHT does
not possess three-dimensional net of H-bonds and the strength of NNO2 bond in a nitramine
group even less than in RDX.
On set of NNHT properties one may conclude that this compound has not any
preferences in comparison with RDX a well known regular explosive. NNHT possesses
worse oxygen balance and lesser density at slightly higher heat of formation. This inevitably
leads to lesser, in comparison with RDX, values of detonation parameters. Declared in [3, 4]
low impact sensitivity of NNHT did not confirmed in our investigation and also earlier in
work [6]. NNHT not only is not insensitive high explosive but even conversely is a very
sensitive high explosive, more dangerous than RDX. Thus, it is unlikely that NNHT will
have prospect of industrial application as an energetic compound.
441
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
P. F. PAGORIA, G. S. LEE, A. R. MITCHELL, AND R. D. SCHMIDT: A review of energetic

materials synthesis. Thermochimica Acta. Vol. 384. No. 1, p. 187204, 2002
Y. YONGZHONG, S. ZHUANG, D. BAORU, AND C. FUBO: Synthesis of polynitrocompounds from
nitroguanidine. Propellants, Explosives, Pyrotechnics. Vol. 14. No. 4, p. 150152, 1989
D. S. HUANG, AND R. R. RINDONE: High energy insensitive cyclic nitramines. US Patent
4,937,340, 1990
D. S. HUANG, AND R. R. RINDONE: NNHT: A new low cost insensitive cyclic nitramine. Proc.
of the Joint Intern. Symp. On Compatibility of Plastics and Other Materials with Explosives,
Propellants, Pyrotechnics and Processing of Explosives, Propellants and Ingredients, San
Diego, CA, USA, p. 6268, 1991
B. STRAUSS, AND S. M. MOY: Insensitive gun propellant. US Patent 5,325,782, 1994
I. J. DAGLEY, AND M. KONY: Properties and impact sensitiveness of cyclic nitramine
explosives containing nitroguanidine groups. J. Energetic Mat. Vol. 13. No. 1-2, p.3556, 1995
A. S. ERMAKOV, S. A. SERKOV, V. A. TARTAKOVSKII, T. S. NOVIKOVA, AND L. I.
KHMELNITSKII: Use of potassium sulfamate for synthesis of heterocyclic nitramines. Khim.
Geterotsikl. Soedin. (Chem. Heterocycle Compounds). No. 8, p. 11291132, 1994 (In Russian)
M. D. CLIFF, I. J. DAGLEY, R. P. PARKER, AND G. WALKER: Synthesis of 2-nitroimino-5nitrohexahydro-1,3,5-triazine via chloride-assisted nitrolysis of a tertiary amine. Propellants,
Explosives, Pyrotechnics. Vol. 23. No. 4, p. 179181, 1998
H. S. JADHAV, M. B. TALAWAR, D. D. DHAVALE, S. N. ASTHANA, AND V. N.
KRISHNAMURTHY: Synthesis and characterization of nitroguanidine based nitrate and
perchlorate salts of 5-nitro-2-nitroiminohexahydro-1,3,5-triazine. Proc. 7th Int. Seminar New
Trends in Research of Energetic Materials. Part I, Pardubice, Czech Republic, p. 150157,
2004
D. S. HUANG: Preparation of 2,4,6-trinitro-2,4,6-triaza-cyclohexanone. US Patent 5,391,736,
1995
G. M. SHELDRICK: SHELXLS-97 and SHELXL-97 Release 97-2. University of Gttingen,
Germany, 1995
Cambridge Structural Database, April 2004 release (about database: F. N. ALLEN: The
Cambridge Structural Database: A quarter of a million crystal structures and rising. Acta
Cryst. B58. No. 2, p. 380388, 2002)
C. S. CHOI, AND E. PRINCE: The crystal structure of cyclotrimethylene-trinitramine. Acta
Cryst. B28. No. 2, p. 28572862, 1972
N. S. MOROZOVA, E. L. METELKINA, T. A. NOVIKOVA, V. A. SCHLYAPOCHNIKOV, AND O. I.
SERGIENKO: Interpretation of IR spectra of nitroguandine. Russ. J. Org. Chem. Vol. 19. No. 6,
p. 12281232, 1983 (In Russian)
Y. OYUMI, A. L. RHEINGOLD, AND T. B. BRILL: Thermal decomposition of energetic materials.
XXIV. A comparison of the crystal structures, IR spectra, thermolysis and impact sensitivities of
nitroguanidine and trinitroethylnitroguanidine. Propellants, Explosives, Pyrotechnics. Vol. 12.
No. 1, p. 4652, 1987
G. B. MANELIS, G. M. NAZIN, YU. I. RUBTSOV, AND V. A. STRUNIN: Thermal Decomposition
and Combustion of Explosives and Powders, (Edited by F.I. Dubovitsky), Nauka, Moscow,
1996 (In Russian)
V. M. GRANKIN, AND P. P. SEMYANNIKOV: The ion source and the high-temperature source of
molecular beam for mass-spectrometer MI 12-01-V. Instruments and Experimental Tachniques.
Vol. 34. No. 4, p. 129131, 1991 (In Russian)
442
[18] YU. SHU, B. L. KORSOUNSKII, AND G. M. NAZIN: Mechanism of thermal decomposition of

secondary nitramines. Uspekhi Khimii (Russ. Chem. Rev.). Vol. 28. No. 3, p. 320335, 2004
(In Russian)
[19] A. M. ASTACHOV, A. A. NEFEDOV, L. A. KRUGLYAKOVA, A. D. VASILIEV, M. S. MOLOKEEV,
S. V. TRUBIN, P. P. SEMYANNIKOV, AND R. S. STEPANOV: Structure thermal stability
relationship for derivatives of nitroguanidine. Proc. International Scientific and Techn. Annual
Conf. Modern Problems of Technical Chemistry, Kazan, Russia, p. 2735, 2004 (In Russian)
[20] A. A. NEFEDOV: Thermal decomposition derivatives of nitroguanidine, Ph.D thesis, Siberian
State Technological University, Krasnoyarsk, 2004 (In Russian)
[21] K. K. ANDREEV, AND A. F. BELYAEV: Theory of explosives, Oborongiz, Moscow, 1960 (In
Russian)
[22] S. G. ANDREEV, A. V. BABKIN, F. A. BAUM ET. AL.: Physics of explosion, Vol. 1 (Edited by L.
P. Orlenko third edition), Fizmatlit, Moscow, 2002 (In Russian)
[23] A. I. GOLBINDER: Laboratory works on a course in the theory of explosives, Rosvuzizdat,
Moscow, 1963 (In Russian)
[24] G. T. AFANASIEV, AND V. K. BOBOLEV: Initiation of solid explosives by impact, Nauka,
Moscow, 1968 (In Russian)
[25] A. V. DUBOVIK: Calculation of impact sensitivity parameters of solid explosives. Dokl. Acad.
Nauk USSR (Doklady Chemistry USSR). Vol. 286. No. 2, p. 377380, 1986 (In Russian)
[26] C. L. MADER: Numerical modeling of explosives and propellants (second edition), CRC Press,
Boca Raton, 1998
[27] V. I. PEPEKIN, AND S. A. GUBIN: Methods of calculation of detonation parameters of
explosives. Khim. Fizika (Chem. Phys. Report). Vol. 22. No. 9, p. 7297, 2003 (In Russian)
[28] A. M. ASTACHOV, V. I. VLASENKO, AND E. S. BUKA: Estimation of energetic characteristic
some nitrimines. Proc. All-Russia Scientific and Techn. Annual Conf. Modern Problems of
Technical Chemistry, Vol. 1., Kazan, Russia, p. 249253, 2002 (In Russian)
[29] V. V. ODINTSOV, S. A. GUBIN, V. I. PEPEKIN, AND L. N. AKIMOVA: The form and size of
diamond crystals definition behind a detonation wave in the condensed explosives. Khim.
Fizika (Chem. Phys. Report). Vol. 10. No. 5, p. 687695, 1991 (In Russian)
[30] M. F. GOGULYA, AND M. A. BRAZHNIKOV: Temperature of detonation products of the
condensed explosive. 1.The solid explosives. Khim. Fizika (Chem. Phys. Report). Vol. 13.
No.1, p. 5263, 1994 (In Russian)
443
REACTIVITY OF 2-(DINITROMETHYLENE)4,6-DIHYDROXY-5,5-DINITROPYRIMIDINE
IN THE PROCESSES OF NUCLEOPHILIC SUBSTITUTION
A.A. Astratv, D.V. Dashko, and A.I.Stepanov
Special Design and Construction Bureau SDCB Technolog
of the Saint-Petersburg State Institute of Technology (technical university)
190013, Russia, Saint-Petersburg, Moskovskiy pr.26
Abstract:
The main routes of transformations of 2-(dinitromethylene)-4,6-dihydroxy-5,5dinitropyrimidine in reactions with primary aliphatic amines, ammonia, hydrazine and
its derivatives in aqueous solution are considered and the mechanism of the processes is
suggested. The formation of derivatives of 2-dinitromethyl-5-N-alkyl substituted 1,3,5triazines and 3-dinitromethyl-1,2,4-triazolone is demonstrated.
Keywords:
1.
synthesis, 2-(dinitromethylene)-4,6-dihydroxy-5,5-dinitropyrimidine
INTRODUCTION
Synthesis of gem-dinitromethylene compounds both in the aliphatic and heterocyclic

series opens up attractive synthetic perspectives for the preparation of high explosives
including insensitive ones (IHE). Recently special interest was focused on 1,1-diamino-2,2dinitroethylene (DADNE, FOX-7) as an outstanding representative of IHE [1-8]. Substitution
of the nitroamino group in nitroguanidine and its analogues for the dinitromethylene moiety
ensures, all other factors being equal, a considerable improvement in the oxygen balance
which eventually favorably affects the energetic characteristics of IHE.
H2N
NH2
H2N
NO2
NO2
NO2
=0.50
=0.66
DADNE
NH2
Nitroguanidine
444
A combination of the dinitromethylene fragment with the , diaminomethylene group

implies the existence of amino-imino tautomerizm which exerts a favorable influence on the
molecular characteristics(density, chemical and thermal stability). Such situation is observed
for DADNE which can be considered as a dinitromethylene analogue of nitroguanidine.
Both compounds have double C=C (C=N) bonds of similar nature which is reflected by the
reactions of DADNE with nucleophiles proceeding with the substitution of the amino
moieties similarly to the reactions of nitroguanidine.
Traditionally, the introduction of a gem-dinitoromethyl(ene) fragment into an organic
molecule is carried out by means of a destructive acidic nitration of compounds with active
methylene groups. The main shortcoming of such an approach is connected with the
difficulty of quenching the reaction on the stage of the dinitromethyl moiety and
impracticability of its application for compounds containing groups unstable under the
conditions of acidic nitration. In this connection the indirect routes of forming such
compounds are of an obvious interest. One of such approaches was demonstrated by Baum
et al in their study of reactions of 1,1-diiodo-2,2-dinitroethylene with various amines [10].
Along these lines special attention should be paid to employing 2-(dinitromethylene)-4,6dihydroxy-5,5-dinitropyrimidine*) (1) as a source of the dinitomethylene moiety. Compound
(1) was first synthesized by Astratev et al in SDCB Technolog via direct nitration
of 2-methyl-4,6-dihydroxypyrimidine with mixed acid [3].
NO2
CH3
N
OH
NH
O
NO2
HNO3
HN
H2SO4
NH
O
NO2
H2O
NO2
NH2
NO2
NH2
NO2
1
Fig 1.
Scheme 1: Synthesis of 2-(dinitromethylene)-4,6-dihydroxy-5,5dinitropyrimidine and DADNE carried out in the SDCBTechnolog.
It was shown earlier that compound (1) was the key intermediate responsible for the
synthesis of DADNE, one of the most promising IHEs. However, the other aspects of the
synthetic potential of (1) as a versatile building block have not yet been identified.
*)
This compound apparently exists in the diketo

2-(dinitromethylene)-4,6-dioxo-1,3-dihydro-5,5-dinitropyrimidine
445
tautomeric
form,
2.
A priori in the structure of intermediate (1) there may be picked out three weak, from
the point of view of the vulnerability toward nucleophilic attack, chemical bonds (A, B, C).
Decomposition of intermediate (1) under the action of nucleophilic agents was shown by
NO2
NO2
A
HN
NH
B
O
NO2
NO2
us earlier to proceed in stages and to be determined to a large extent by the acid-base

properties of the medium. Judging from the structure of (1) and the results obtained earlier, it
is reasonable to assume that theweakest bond in (1) is bond "C" [1]. To check this
assumption, we studied the interaction of (1) with ammonia and aliphatic amines and found
that treatment of (1) with primary aliphatic amines and ammonia in aqueous solution
furnished the corresponding derivatives of 2-dinitromethyl-5-N-alkyl substituted 1,3,5triazines. The sym-triazine structure of the products was supported by the results of
hydrolysis of 2-dinitromethylene-4,6-dioxo-1,3,5-trihydro-1,3,5-triazine (2) in the diluted
hydrochloric acid which afforded cyanuric acid (3) identified by its IR spectrum. The
following scheme of chemical transformations of compound (1) under the action of ammonia
may be suggested on the basis of the above discussion.
NO2
NO2
NO2
NH
HN
O
NO2
NO2
NH3
H2O
NO2
O
O
N
H
C
O
NO2
NO2
NH2
NO2
NO2
O
N
H
NO2
NO2
NH
- CH-(NO2)2
N
H
NH2
HN
O
NH
N
H
Fig 2.
HN
Scheme 2: Chemical transformations of 2-(dinitromethylene)-4,6dihydroxy-5,5-dinitropyrimidine in the reaction with ammonia.
Since cyclic compounds are produced in this case, it is reasonable to assume that the
transformation of the dinitroacetyl moiety proceeds in stages and begins with the splitting
out of dinitromethane molecule and formation of an isocyanate.
446
- CH2(NO2)2
HN
CH NO2
NO2
H2O
R1-NH2
R
O
R
HN
HN
OH
R1
N
H
Fig 3.
NH2
Scheme 3: Chemical transformations of the dinitroacetyl moiety under

nucleophilic attack on 2-(dinitromethylene)-4,6-dihydroxy-5,5dinitropyrimidine.
Meanwhile, on treatment of (1) with sodium methoxide for 48 hours a dinitromethyl

analogue of nitroguanylurea, 1-amino-1-carbamino-2,2-dinitroethylene (5), was isolated in
25% yield. The formation of the carbamino moiety apparently takes place in the course of
fragmentation of intermediates in the presence of a strong base, the evolving ammonia being
entrained into ammonolysis of the methylcarbamate group.
NO2
NO2
NO2
NH
HN
CH3ONa
O
HN
N
H
C
O
NO2
NO2
O
NO2
NO2
NO2
1
- CH-(NO2)2
H2O
-CO2
NO2
NO2
NO2
NO2
O
O
N
H
H2N
OMe
NH3
H2N
N
H
OMe
NH2
Fig 4.
Scheme 4: Chemical transformations of 2-(dinitromethylene)-4,6dihydroxy-5,5-dinitropyrimidine in the reaction with sodium methoxide.
Since no cyclic structures are formed in this case, the ammonolysis takes place after the
complete fragmentation of the dinitroacetyl moiety.
447
Rather interesting opportunities are offered by the action of hydrazine on compound (1).
The corresponding products may be both of linear and cyclic nature depending on the order
of mixing the reactants and reaction conditions. So, when compound (1) was metered into a
20 % aqueous solution of hydrazine hydrate at 10-15 oC, the hydrazinium salt of 3dinitromethyl-1,2,4-triazolone (6) was produced. When compound (1) was first added to
water and a solution of hydrazine was subsequently poured into the reaction mass, 1-amino1-hydrazino-2,2-dinitroethylene (7) was formed.
H
N N
NO2
C
NO2
NH2NH2
NO2
NO2
HN
NH
O
NO2
O
NO2
1
N
H
NH3+NH2
1. H2O
2. NH2NH2
NO2
NO2
H
N NH2
NH2
Fig 5.
Scheme 5: Chemical transformations of 2-(dinitromethylene)4,6-dihydroxy-5,5-dinitropyrimidine in the reaction with hydrazine.
In this respect the behavior of 1,1-dimethylhydrazine is demonstrative because it

apparently attacks compound (11), the product of fission of 2-(dinitromethylene)-4,6dihydroxy-5,5-dinitropyrimidine across the C bond, eventually forming the intermediate
amidrazone (8). Since the two alkyl groups deactivated the other nitrogen atom of the
hydrazine moiety, the formation of cyclic structures in the process of destruction of 5,5dinitroacetyl fragment did not occur, rather the addition of the second molecule of 1,1dimethylhydrazine took place with the formation of N,N-dimethylsemicarbazide derivative
(9) which was identified on the basis of its NMR1H and mass-spectrometric data. If the
opposite were true, the formation of a corresponding 1-N-dialkyl sybstituted 1,3,5triazine(10) would take place by analogy with the mechanism of the interaction with
primary amines.
448
CH3
H2N N
CH3
NH
HN
O
NO2
NO2
NO2
NO2
NO2
NO2
O
CH3
N
H
HN
C
O
NO2
NO2
N N
H
CH3
CH3
H2N N
CH3
H3C
NO2
NO2
CH3
H3N+
N
CH3
H3C
NO2
NO2
HN
NH
H3C
Fig 6.
N
N
10
NH
C
N
H
8
NO2
NO2
H3N+
CH3
N
CH3
-CO2
-CH2(NO2)2
11
CH3
N
NO2
NO2
CH3
N
NH
N
H
H
N N
CH3
CH3
9
CH3
Scheme 6: Chemical transformations of 2-(dinitromethylene)-4,6dihydroxy-5,5-dinitropyrimidine in the reaction with 1,1dimethylhydrazine.
Hence a certain difference in the reaction route between N-alkyl substituted primary
aliphatic amines, ammonia, hydrazine and its derivatives is observed which is caused by the
process of hydrazinolysis of the amine moiety in 1,1-dinitro-2,2-diaminoethylene. Such
difference eventually expands the possibilities of chemical transformations of 2(dinitromethylene)-4,6-dihydroxy-5,5-dinitropyrimidine.
3.
CONCLUSION
In conclusion it should be mentioned that the interaction of compound (1) with ammonia
gives results differing from those received on the ammonolysis of 5,5-dinitrobarbituric acid
in which case, according to the work by Latypov et al [9], a linear product, viz., the
symmetrical triurea, is formed. Besides, evident distinctions are observed in the results of
ammonolysis of 2-(dinitromethylene)-4,6-dihydroxy-5,5-dinitropyrimidine and another
source of DADNE, 2-dinitromethylene-5,5-dinitroimidazolidin-4-one. [4]
Acknowledgements
The authors would like to thank the International Science and Technology Center for
giving financial support to this project.
449
REFERENCES.
[1]
[2]
USSR Inventors Certificate 318468; 1989.

DOROKHOV V.A., GORDEYEV M.F., BOGDANOV V.S., LAIKHER A.L., KISLYI V.P., SEMENOV
V.V.: A novel synthesis of aminals of nitroketenes, Bull. of Russian Acad. Sci, Ser. Khim,
1990. 11 p. 2660
[3] ASTRAT`EV,A; DASHKO,D.; MERSHIN,A; STEPANOV,A. URAZGIL`DEEV, N.: Russian Journal
of Organic Chemistry 2001,37,729-733
[4] LATYPOV,N.V.; BERGMAN, J.; LANGLET, A.;WELLMAR,U.;BEMM,U.: Tetrahedron 1998, 54,
11525-11536
[5] BERGMAN, H.; STMARK, H; PETTERSSON, A; PETTERSSON, M.-L.; HIHKIO,M.: Some Initial
Properties and Thermal Stability of FOX-7, Tampa, 1999, pp346-351
[6] STMARK, H; BERGMAN, H.; BEMM,U.; GOEDE, P.; HOLMGREN,E.; JOHANSSON, M.;
LANGLET, A.; LATYPOV,N.V.; PETTERSSON, A.; PETTERSSON, M.-L.; WINGBORG,N.;
VORDE,C.; STENMARK, H.; KARLSSON,L.; HIHKIO,M.: 2,2-Dinitro-ethene-1,1-diamine(FOX-7)
- Properties, Analysis and Scale Up: Karlsruhe, 2001.
[7] KARLSSON,S.; STMARK, H; ELDSATER, C.; CARLSSON,T.; BERGMAN, H.; WALLIN, S.;
PETTERSSON, A.: Detonation and Properties of FOX-7 and formulations containing FOX-7;
San Diego,2002, pp 5.
[8] LOCHERT, I.: FOX-7- New Insensitive Explosive, DSTO, 2001
[9] LANGLET, A.; LATYPOV,N.V.; WELLMAR,U.; BEMM,U; GOEDE , A; BERGMAN, J.;
ROMERO,I.: Nitration of 2-substituted pyrimidine-4,6-diones , structure and reactivity of 5,5gem-dinitropyrimidine 4,6-diones.; J.Org.Chem.2002, 67, 7833-7838
[10] BAUM K.; BIGELOW,S.S.; NGUYEN,N.V.; ARCHIBALD,T.G.; GILARDI,R.; FLIPPEN-ANDERSON,
J.L. AND GEORGE, C.: Synthesis and Reactions of 1,1-Diiododinitroethylene; J.Org.Chem.1992,
57, 235-241
450
STUDIES ON THE HYDRODENITRATION BY TIN (II) CHLORIDE

OF POLYNITROHEXAAZAISOWURTZITANES
Anthony J. Bellamy1
Abstract:
The hydrodenitration of 2,4,6,8,10,12-hexanitro- (HNIW, CL-20), 2,4,6,8,12-pentanitroand 2,6,8,12-tetranitro-hexazaisowurtzitane using SnCl2 has been studied. Removal of
N-nitro groups from HNIW to give the two isomeric mono-amines occurs with almost
statistical distribution of products, but hydrodenitration of 2,4,6,8,12pentanitrohexaazaisowurtzitane more strongly favours reaction at the nitramine groups
in the 5-membered rings. Surprisingly, reduction of the di-amine 2,6,8,12tetranitrohexaazaisowurtzitane did not give the expected tri-amine 2,6,8trinitrohexaazaisowurtzitane but the ring-cleavage product 1,1,2,2-tetranitraminoethane
instead.
Keywords:
1.
Hexanitrohexazaisowurtzitane, pentanitrohexaazaisowurtzitane,
tetranitrohexaazaisowurtzitane, tetranitraminoethane,
hydrodenitration.
INTRODUCTION
Nitramines may be chemically reduced to the corresponding amines using a variety of

reagents [1]. These reagents have been used mostly for the deprotection of N-nitroargenine
derivatives, but, as we have reported earlier [2], they may also be used for the removal of
nitro groups from hexanitrohexaazaisowurtzitane (I, HNIW, CL-20). The reagents that we
had used previously were (i) cyclohexa-1,4-diene/Pd-C in EtOH, MeOH or CH3CN [3], (ii)
formic acid/Pd black in MeOH [4], (iii) hydrazine/Pd black in CH3CN [5] and (iv) SnCl2.2H2O
in H2O-HCOOH, MeOH-HCOOH or MeOH alone [6]. All of these reagents with HNIW gave
broadly similar results, but reagents (i) and (iv) appeared to be the most selective, and
SnCl2.2H2O in MeOH was used for preparative experiments.
Our earlier work had led us to conclude that HNIW is reduced preferentially at position
2 (6, 8 or 12) to give the 2,4,6,8,10-pentanitro derivative (II) as the main primary product.
This mono-amine is then reduced further to give di-amines, two of which were isolated and
characterised as 2,6,8,10- (III) and 2,4,6,8-tetranitro (IV) derivatives. A study of the course
of the reduction by HPLC analysis of samples removed at intervals indicated that there was a
second primary product, assumed to be the 2,4,6,8,12-pentanitro derivative (V), but that this
was formed at a slower rate (~0.2X). There were also other secondary products, assumed to
be di-amines with both amine groups in the 5-membered rings.
Our recent synthesis [7] of authentic samples of 2,4,6,8,12-pentanitrohexaazaisowurtzitane (V) and 2,6,8,12-tetranitrohexaazaisowurtzitane (VI) has allowed us to check
these conclusions, and has indicated that some modification of the above picture is
1
451
necessary. We have found that unlike the previously isolated and characterised amines (IIIV), both V and VI are unstable under the HPLC conditions used in the previous study, and
on flash column chromatography on silica gel. Consequently, our estimates of the amounts
of these potential HNIW reduction products were in error. We have now shown by careful
1
H NMR analysis of the reduction product mixtures from HNIW, and by separate reduction
of the mono-amine V, that reduction at positions 2 (6, 8 or 12) and 4 (or 10) is close to being
statistically distributed.
2.

2.1
Reduction of 2,4,6,8,10,12-hexanitrohexaazaisowurtzitane (I)

and 2,4,6,8,12-pentanitrohexaazaisowurtzitane (V).
Hexanitrohexaazaisowurtzitane (I) was reduced with a limited amount of SnCl2 in order

to minimise the amount of secondary reduction products. The resultant mixture of products
was then analysed by 1H NMR spectroscopy. Since the 1H spectra of both potential primary
reduction products (II and V) were available [refs. 2 and 7b respectively; see Figure 2], it
was possible to identify all of the peaks exhibited by these products. This indicated that the
extent of reduction was ~25% (~75% I remained) and that the ratio of the primary products
II/V was ~2.4. Our earlier report [2] therefore overestimated the relative reactivities of the
two different nitramine groups in I. The revised figure is much closer to a statistical
distribution.
Similar reduction of the pentanitro derivative V gave only ~8% reduction, indicating that
its reactivity is only about 1/3 that of HNIW. The di-amine product distribution was III
(4.8%), IV (2.7%) and VI (0.5%). The relative reactivities of the nitramine groups in the
5-membered rings and the remaining nitramine group in the 6-membered ring [(III + IV)/VI]
was ~15, much higher than a statistical distribution (4), whilst reduction at positions 2 and 6
was favoured over positions 8 and 12 (III/IV).
2.2
Reduction of 2,6,8,12-tetranitrohexaazaisowurtzitane (VI)
It was anticipated that reduction of 2,6,8,12-tetranitrohexaazaisowurtzitane (VI) might

give the tri-amine 2,6,8-trinitrohexaazaisowurtzitane. However, due to the low solubility of
VI in MeOH, it was not possible to perform a selective, low conversion reaction. The
reaction was allowed to proceed until all of the substrate had gone into solution (24h). The
product mixture isolated from such a reaction contained 1,1,2,2-tetranitraminoethane (VII)
as the major product, identified by comparison of its spectroscopic properties with those of
an authentic sample. The latter was prepared from tetranitroglycoluril (VIII) [8] by hydrolysis
with alkali [9].
One rationalisation of how VII might be formed from VI is shown in Figure 3. An
alternative could involve the addition of one electron to each of the two nitro groups
involved, followed by a similar cleavage. No product derived from the central 6-membered
piperazine ring was detected.
The formation of VII from both VI and VIII demonstrates that HNIW, with its cage
structure, and tetranitroglycoluril (TNGU or SORGUYL, VIII), with its 2,4,6,8-tetraazabicyclo[3,3,0]octane structure, are structurally related (see Figure 4), each incorporating a
1,1,2,2-tetranitraminoethane unit.
452
3.
EXPERIMENTAL
Many of the amines obtained appear to have a strong affinity for EtOAc; final traces of
the latter are difficult to remove. NMR spectra were recorded in acetone-d6.
3.1
Reduction of 2,4,6,8,10,12-hexanitrohexaazaisowurtzitane (I)
HNIW (100mg, 0.23mmol) was dissolved in MeOH (5.0ml) and a solution of

SnCl2.2H2O in MeOH (0.50ml, 0.22mmol; 1.00g/10ml) was added dropwise. After 15 min
the solution was concentrated and the residue was taken up in EtOAc (15ml). The solution
was washed with 3M HCl (2x2.5ml) and brine (3x2.5ml), and then evaporated. Residue
(which contained EtOAc by NMR) 133mg. TLC (hexane/EtOAc 3/2): Rf 0.64 (HNIW), 0.44
(mono-amines II and V; both present by NMR), 0.32 (di-amine III), 0.17 (di-amine IV),
0.00.
3.2
Reduction of 2,4,6,8,12-pentanitrohexazaisowurtzitane (V)
As for HNIW. Residue (which contained EtOAc by NMR) 162mg. TLC: as for the
reduction mixture from HNIW except that HNIW was absent.
3.3
Reduction of 2,6,8,12-tetranitrohexaazaisowurtzitane (VI)
VI (250mg, 0.72mmol) was slurried in MeOH (15ml) and two portions (3.15ml each,
total 2.79mmol) of a solution of SnCl2.2H2O in MeOH (1.00g/10ml) were added at 0h and
4h. After 24h the mixture had become almost clear. The solvent was evaporated and the
residue was dissolved in EtOAc (50ml). The solution was washed with brine/water (1/1,
25ml) and brine (25ml). Evaporation of the EtOAc gave a solid residue (378mg), which
contained some EtOAc (by NMR). TLC: no starting material remained, the only spot was at
the origin. NMR (acetone-d6): 1H 6.58 (s, 2H, CH, VII), 9.00 (s, 0.52H, unknown no 13C
side-bands were visible), 11.92ppm (br s, 3.25H, NH, VII), plus peaks for EtOAc (1.2eq);
13
C 65.01ppm (CH, VII), plus peaks for EtOAc. The spectra in DMSO-d6 and D2O were
unstable (see behaviour of 1,1,2,2-tetranitraminoethane below). FTIR (KBr) spectrum
exhibited the characteristic bands of 1,1,2,2-tetranitraminoethane.
A sample of the reduction products (77mg) was dissolved in water (4.0ml; solution pH
1) and 1M NaOH was added dropwise until the pH was 13-14 (total volume added ~0.95ml).
Evaporation and drying gave a solid residue (107mg). NMR (D2O + DSS reference): 1H
5.90ppm (s, CH); 13C 77.32ppm (CH).
3.4
Synthesis of 1,1,2,2-tetranitraminoethane (VII)
Tetranitroglycoluril (VIII) was prepared by nitration of glycoluril using N2O5/HNO3

according to a literature method [8]. A sample (100mg) was then treated with 1M NaOH
(2.50ml, 8eq). The solid dissolved within minutes, with mild evolution of heat. The final pH
was ~12.
EtOAc (5ml) and 1M HCl (1.3ml) were added to half of the solution. The final pH was
~1. The organic phase was separated, washed with brine (2x2.5ml) and concentrated to give
1,1,2,2-tetranitraminoethane as a white crystalline solid. DSC (10K/min) onset of exotherm
108.1C. NMR (acetone-d6): 1H 6.58 (s, 2H, CH), 11.90ppm (br s, 3.66H, NH); 13C
65.03ppm (CH). The NMR spectra in DMSO-d6 initially exhibited peaks at 6.08 (s) and 13.2
(br s) in the 1H spectrum and 63.11 in the 13C spectrum. However, within a few hours this
had changed, with the appearance of several new peaks in the 1H spectrum and a new peak at
162.9 (CH) in the 13C spectrum. The latter was the major peak after 20h. Gas was also
453
evolved. In D2O, the initial 1H spectrum exhibited a peak at 6.39 (s), but within the time
required to acquire the 13C spectrum (1.25h) there were two peaks (66.52 and 93.21). After
20h only the 93.21 peak remained, with a 1H peak at 4.80 (peak at 6.39 absent). There was
considerable gas evolution. Thus 1,1,2,2-tetranitraminoethane is unstable in both DMSO-d6
and D2O; N2O was identified (GC-MS) as the gas evolved in both solvents.
The other half of the hydrolysis solution was evaporated and dried to give the tetrasodium salt of 1,1,2,2-tetranitraminoethane (+ Na2CO3) as a solid. NMR (D2O with DSS as
reference): 1H 5.90ppm (s, CH), 13C 77.28ppm (CH). Unlike the parent 1,1,2,2tetranitraminoethane (see above) the spectrum in D2O did not change with time.
Treatment of 1,1,2,2-tetranitraminoethane with SnCl2 (as for VI above) gave back
unchanged material.
Acknowledgment:
I wish to thank Dr A. MacCuish for helpful discussions, Dr P. Gill for GC-MS analyses
and Mr A. Agha for HPLC analyses.
454
REFERENCES
[1] T.W.GREENE and P.G.M.WUTS, Protective groups in organic synthesis, 2nd edition, Wiley
1991, p374.
[2] A.J.BELLAMY, Polynitrohexaazaisowurtzitane derivatives related to hexanitrohexaazaisowurtzitane (HNIW, WN6), ICT Conference on Energetic Materials, Karlsruhe, Germany, 1091 to 109-14, 2000.
[3] A.M.FELIX, E.P.HEIMER, T.J.LAMBROS, C.TZOUGRAKI and J.MEIENHOFER, Rapid
removal of protecting groups from peptides by catalytic transfer hydrogenation with 1,4cyclohexadiene, J.Org.Chem., 43, 4194-4196, 1978.
[4] B.ELAMIN, G.M.ANANTHARAMAIAH, G.P.ROYER and G.E.MEANS, Removal of benzyltype protecting groups from peptides by catalytic transfer hydrogenation with formic acid,
J.Org.Chem., 44, 3442-3444, 1979.
[5] R.D.CHAPMAN, R.A.OBRIEN and P.A.KONDRACKI, N-Denitration of octahydro-1,3,5,7tetranitro-1,3,5,7-tetrazocine (HMX) by hydrazine catalyzed by palladium, J.Energetic
Materials, 16, 147-171, 1998.
[6] T.HAYAKAWA, Y.FUJIWARA and J.NOGUCHI, A new method of reducing nitroargininepeptide into arginine-peptide, with reference to the synthesis of poly(L-arginine hydrochloride,
Bull.Chem.Soc.Japan, 40, 1205-1208, 1967.
[7] A.J.BELLAMY, A.MACCUISH and P.GOLDING, (a) The use of the trifluoroacetyl protecting
group in the synthesis of mono- (4) and di- (4,10) amines in the polynitrohexaazaisowurtzitane
series, Propellants, Explosives, Pyrotechnics, 28(3), 157-158, 2003; (b) Synthesis and reactions
of 2,4,6,8,12-pentanitrohexaaza-isowurtzitane (WN5H) and 2,6,8,12-tetranitrohexaazaisowurtzitane (WN4H2), ICT Conference on Energetic Materials, Karlsruhe, Germany, 133-1 to
133-6, 2004.
[8] J.BOILEAU, J-M.L.EMEURY and J-P.KEHREN, Tetranitroglycoluril and method of
preparation thereof, US 4487938 (1984).
[9] D.WAN, Reactions of 1,1,2,2-tetranitroaminoethane and properties of its tetra sodium salt,
Proceedings of the 17th International Pyrotechnics Seminar, Vol 1, 231-234, 1991.
455
NO 2
N
NO 2
NO2
NO 2
NO 2
NO 2
NO 2
10
8
6
12
2
NO 2
NO 2
NO2
NO 2
NO 2
NO2
NO 2
NO2
(II)
(I)
(III)
NO 2
NO 2
NO2
NO2
NO 2
NO 2
NO 2
NO2
NO 2
NO 2
NO2
NO2
NO 2
(V)
(IV)
(VI)
O
O2NHN
NHNO 2
O 2N N
N NO2
O2NHN
NHNO 2
O 2N N
N NO2
(VII)
Fig 1.
Formulae
456
(VIII)
Fig 2.
H NMR spectra (acetone-d6) of isomeric pentanitro- and tetranitrohexaazaisowurtzitanes
457
NO2
NO2
2e
(i)
NO2
NO2
NO2
NO2
(ii)
O
N
O
Repeat of (i) and (ii)

H
O2NHN
NHNO2
+
O2NHN
NHNO2
N
H
Fig 3.
Rationalisation of the formation of VII from VI
NO2
NO2
NO2
NO2
(VI)
Fig 4.
SnCl2
MeOH
O2NHN
NHNO2
O2NHN
NHNO2
(i) NaOH
(ii) HCl
O2N N
N NO2
O2N N
N NO2
O
(VII)
Formation of VII from both VI and VIII
458
(VIII)
SYNTHESIS AND PROPERTIES

OF SALTS OF 3,5-DIAMINOPICRIC ACID
Anthony J. Bellamy,1 Luigi Cassioli and Alessandro E. Contini
Abstract:
Seven salts of 3,5-diaminopicric acid, ammonium 3,5-diaminopicrate (ADAP),
hydrazinium 3,5-diaminopicrate (HDAP), guanidinium 3,5-diaminopicrate (GDAP),
aminoguanidinium
3,5-diaminopicrate
(AGDAP),
diaminoguanidinium
3,5diaminopicrate (DAGDAP), triaminoguanidinium 3,5-diaminopicrate (TAGDAP) and
guanylurea 3,5-diaminopicrate (GUDAP) have been synthesized [from 3,5diaminopicric acid (DAPA) and the corresponding base (or its salt)] and characterized
by 1H, 13C, and 15N NMR spectroscopy, FTIR, DSC, TG, SEM and CHN elemental
analysis. Their heats of combustion were measured by bomb calorimetry and their heats
of formation calculated. Their impact sensitivities were also measured. Some of these
salts (GDAP, AGDAP and GUDAP) are extremely insensitive to impact (F of I > 130),
whilst others (HDAP and TAGDAP) might be useful as initiators (F of I 30).
Keywords:
1.
salts of 3,5-diaminopicric acid, heat of combustion, impact sensitivity.
INTRODUCTION
During the development of a new synthetic route to TATB [1], we observed that
amination of the immediate precursor of TATB, 1,3,5-trimethoxy-2,4,6-trinitrobenzene, at
elevated temperature gave the ammonium salt of 3,5-diamino-2,4,6-trinitrophenol
(ammonium 3,5-diaminopicrate, ADAP, I) as the major product [2]. The same salt could also
be made, in higher yield, by selective hydrolysis of one methoxy group in the precursor to
give the sodium salt of 3,5-dimethoxy-2,4,6-trinitrophenol, followed by amination [2]. ADAP
was subsequently found to show potential as an insensitive explosive. We have now
prepared a range of related salts of 3,5-diaminopicric acid and examined their physical,
thermochemical and explosive properties.
2.
EXPERIMENTAL
NMR spectra were recorded in DMSO-d6. The heats of combustion were determined by
burning 170-250mg of the salt, compressed into a disc, in an oxygen atmosphere (30 bar)
using a Gallenkamp Autobomb 305 static adiabatic oxygen bomb calorimeter, fitted with a
Parr 1108-Cl halogen-resistant bomb. The calorimeter was calibrated with benzoic acid
thermochemical standard (Parr Inc.). The results were corrected for the formation of 3.5-5.7
molar % nitric acid.
1 Corresponding author; fax (44) 01793-783878, e-mail: A.J.Bellamy@cranfield.ac.uk
459
3.
The 3,5-diaminopicrate salts of guanidine (GDAP), aminoguanidine (AGDAP),

diaminoguanidine (DAGDAP), triaminoguanidine (TAGDAP) and guanylurea (GUDAP)
were prepared by reacting a hot aqueous suspension of 3,5-diaminopicric acid (DAPA) with
one equivalent of NaOH and then adding an excess of an aqueous solution of the
hydrochloride or sulphate of the base. For the hydrazinium salt (HDAP), treatment with
NaOH was omitted and an excess of aqueous hydrazine was added. The salts which
separated were filtered off and recrystallised from water.
+
Salt
ADAP
NH4
HDAP
NH2NH3
GDAP
(NH2)3C
(NH2)2(NH2NH)C
M
O
O2N
NO2
2
H2N
NO2
NH2
AGDAP
DAGDAP (NH2)(NH2NH)2C
TAGDAP (NH2NH)3C
GUDAP
(NH2CONH)(NH2)2C
All of the salts were characterized by 1H (Table 1), 13C (Table 2), and 15N (Table 3)
NMR spectroscopy, FTIR, DSC (Table 4), TG, SEM and CHN elemental analysis. ADAP
was included for comparison. The explosive properties of the salts [Figure of Insensitiveness
(F of I; limited Rotter impact test)], detonation velocity and detonation pressure calculated
by the Rothstein and Petersen method [3]] are also reported in Table 4. The heats of
combustion were measured and, after correction, used to calculate the enthalpies of
formation of the salts. The results are presented in Table 5. GUDAP failed to ignite in the
bomb calorimeter on several occasions (see Figure 1).
Fig 1.
Burn marks on a pellet of GUDAP after attempted ignition

in the bomb calorimeter.
460
4.
CONCLUSIONS
Some of these salts (GDAP, AGDAP and GUDAP) are extremely insensitive to impact
(F of I > 130) and possess high decomposition temperatures. In addition GUDAP was
particularly difficult to ignite in the bomb calorimeter. Other salts (HDAP and TAGDAP)
are quite sensitive to impact (F of I 30) and might be useful as initiators.
REFERENCES
[1]
[2]
[3]
[4]
[5]
A.J. BELLAMY, P. GOLDING and S.J. WARD, A new synthetic route to 1,3,5-triamino-2,4,6trinitrobenzene (TATB), Insensitive Munitions and Energetic Materials Technology
Symposium, Bordeaux, France, 626-634, October 2001; A new synthetic route to 1,3,5triamino-2,4,6-trinitrobenzene (TATB), Propellants, Explosives, Pyrotechnics, 27, 49-58, 2002;
Synthesis of TATB and DATB from trihydroxybenzene and dihydroxybenzene by nitration,
alkylation and amination, GB 2355713 A1 (2001).
A.J. BELLAMY, P. GOLDING and S.J. WARD: Synthesis of ammonium diaminopicrate
(ADAP), a new secondary explosive, Propellants, Explosives, Pyrotechnics, 27, 59-61, 2002;
Synthesis of ammonium diaminopicrate from trihydroxybenzene as candidate novel insensitive
explosive, GB 2355714 A1 (2001); Amination of 1,3,5-trialkoxy-2,4,6-trinitrobenzene;
formation of TATB vs. ADAP (ammonium 3,5-diaminopicrate), ICT Conference on Energetic
Materials, Karlsruhe, Germany, 1-1 to 1-10, June 2002.
L.R. ROTHSTEIN and R. PETERSEN: Predicting high explosive detonation velocities from
their composition and structure, Propellants and Explosives, 4, 56-60, 1979; 4, 86, 1979.
R. MEYER: Explosives, 20, Verlag Chemie, 1977.
J. AKHAVAN: The Chemistry of Explosives, 41, The Royal Society of Chemistry, 1998.
461
Table 1.
H NMR Chemical Shifts
Salt
Chemical shift
(ppm from TMS)
9.1 (s)
7.1 (s)
9.1 (s)
7.0 (broad s)
9.1 (s)
6.9 (s)
9.1 (s)
8.6 (s)
7.0 (broad d)
4.7 (s)
9.1 (s)
8.5 (s)
7.1 (s)
4.6 (s)
9.1 (s)
8.6 (s)
4.5 (s)
9.6 (s)
9.1 (s)
8.1 (s)
7.2 (broad s)
Proton
aromatic NH2
NH4
aromatic NH2
NH2NH3
aromatic NH2
C(NH2)3
aromatic NH2
NH2NH
C(NH2)2
NH2NH
aromatic NH2
NH2NH
CNH2
NH2NH
aromatic NH2
NH2NH
NH2NH
CONH
aromatic NH2
C(NH2)2
CONH2
ADAP
HDAP
GDAP
AGDAP
DAGDAP
TAGDAP
GUDAP
Table 2.
13
Integral
4.00
3.95
4.00
2.98
4.00
6.11
4.00
1.03
4.00
2.01
4.00
1.95
1.94
3.98
4.00
3.04
6.08
1.02
4.00
3.95
1.97
C NMR Chemical Shifts

13
Salt
C-1
DAPA*
C Chemical Shifts (ppm from TMS)

C-2
C-3
C-4
cation carbons
153.1
115.3
146.0
117.8
156.7
113.0
146.5
117.0
163.4
109.4
147.2
120.5
163.2
163.3
163.3
163.3
163.3
163.3
109.5
109.4
109.4
109.4
109.4
109.4
147.1
147.2
147.1
147.1
147.1
147.2
120.4
120.4
120.5
120.5
120.5
120.5
157.8
158.7
158.7
158.9
155.3
154.3
(calculated)
DAPA*
(observed)
ADAP
HDAP
GDAP
AGDAP
DAGDAP
TAGDAP
GUDAP
* DAPA = 3,5-diaminopicric acid
462
Table 3.
15
N NMR Chemical Shifts

15
N Chemical Shifts (ppm from NH3)

NO2
aromatic NH2
cation nitrogens
368.8 365.8
87.1
22.1
368.6 365.7
87.7
49.2
368.8 365.8
87.5
75.5
-
Salt
ADAP
HDAP
GDAP
A
G
D 368.6
365.8
88.4
97.2
55.4
365.6
365.4
364.5
87.8
87.0
88.0
68.8
92.2
110.2
53.5
51.7
89.0
83.9
A
P
DAGDAP
TAGDAP
GUDAP
b
c
d
e
f
368.7
368.8
368.7
DSC 10K/min
Figure of Insensitiveness (Rotter Impact Test), RDX = 80
By Rothstein and Petersens method [3]
Onset of melting 228C
Onset of melting 257C
Ammonium picrate
Table 4. Thermal decomposition and explosive properties

Salt
ADAP
HDAP
GDAP
AGDAP
DAGDAP
TAGDAP
GUDAP
Explosive
Df
TATB
Decomposition
temperature
by DSCa (C)
267d
189
331
231
202
197
271e
F of Ib
(approx.)
90
30
>130
>130
100
30
>>130
Detonation
velocity (ms-1)
(calculated)c
7764
7919
7737
7874
7998
8113
7578
280 [4]
105
7360
350 [5]
110
7940
463
Detonation
pressure (kbar)
(calculated)c
268
283
266
279
290
301
251
234
277
Table 5.
Thermochemical data
Salt
Uc (J g-1)
Hc (J g-1)
Hf (J g-1)
ADAP
HDAP
GDAP
AGDAP
DAGDAP
TAGDAP
GUDAP
Explosive Da
TATB
-11730 190
-11930 30
-12190 210
-12120 70
-12470 50
-13060 120
-11670 320
-
-11690 190
-11890 30
-12151 210
-12081 70
-12431 50
-13021 120
-11631 320
-
-1042 190
-677 30
-1040 210
-944 70
-443 50
286 120
-1480 320
-1581[4]
-598 [5]
Ammonium picrate
464
THE USE OF THE TRIFLUOROACETYL GROUP TO PROTECT

NH AND OH GROUPS DURING NITROLYSIS REACTIONS
Anthony J. Bellamy*,1 Alistair MacCuish* and Peter Golding**
Abstract:
N and O trifluoroacetylation has been used to protect secondary amine groups and
hydroxyl groups respectively during nitrolysis reactions. The trifluoroacetyl group may
be readily removed from the energetic products by solvolysis under mild conditions.
Examples include the synthesis of 2,4,6,8,12-pentanitrohexaazaisowurtzitane, 2,6,8,12tetranitrohexaazaisowurtzitane, 2,2-bis(nitratomethyl)propan-1-ol and N-nitroazetidin3-ol.
Keywords: N-trifluoroacetylation, O-trifluoroacetylation, protection, nitrolysis
1.
INTRODUCTION
Until relatively recently, energetic compounds for practical applications have had simple
structures and have been prepared by simple one- or two-step processes. More recent
developments in energetic materials have seen the synthesis of increasingly complex
molecules via multi-step synthetic procedures. In order to facilitate the synthesis of some
novel energetic molecules, we have protected hydroxyl and secondary amino groups in nonenergetic precursor compounds by trifluoroacetylation to give the O- or N-trifluoroacetyl
derivative respectively. Trifluoroacetate esters and trifluoroacetamides are known to be
resistant to nitrolysis conditions [1]. Following the nitrolysis of other functionalities in these
compounds e.g. N-acetyl, N-H, oxetane, the protecting group may then be removed under
mild conditions to regenerate the hydroxyl or secondary amine group, which can then be
used for further elaboration of the molecule.
Related work [2,3] has previously been carried out using acetyl as a protecting group for
hydroxyl groups. However, the use of trifluoroacetyl offers several advantages.
Trifluoroacetic anhydride (TFAA) is highly reactive and the protection reaction can
normally be carried out under mild conditions in the absence of catalysts. The high volatility
of TFAA and the by-product of the reaction, trifluoroacetic acid, also allows excess
reagent/by-product to be removed by evaporation. Furthermore, trifluoroacetate esters and
trifluoroactamides are more readily cleaved than the corresponding acetate esters and
acetamides; in most cases deprotection of the nitrolysed compounds can be achieved by
solvolysis in alcohols at room temperature. Such mild deprotection conditions could be
important in cases where compounds are acid-, base- or heat-sensitive, as is often the case
with polynitro compounds and in some recently developed, novel materials which contain a
range of functionalities e.g. amino, azido, nitrourea and N-oxide groups. To avoid premature
cleavage of the protecting group, it is important that the nitrolysis step is carried out under
essentially anhydrous conditions.
1
465
2.

2.1
Protection of secondary amine groups during nitrolysis

of N-acetyl and/or nitration of secondary amine groups;
synthesis of polynitro-2,4,6,8,10,12-hexaazaisowurtzitanes [4]
The synthesis of 2,6,8,12-tetranitro-2,4,6,8,10,12-hexaazaisowurtzitane (V) used

2,6,8,12-tetraacetyl-2,4,6,8,10,12-hexazaisowurtzitane (I) as starting material. The protected
compound, 2,6,8,12-tetraacetyl-4,10-bis(trifluoroacetyl)-2,4,6,8,10,12-hexaazaisowurtzitane
(II) was synthesised by reaction of (I) with excess TFAA; II was then nitrolysed in
anhydrous mixed acid to give 2,6,8,12-tetranitro-4,10-bis(trifluoroacetyl)-2,4,6,8,10,12hexaazaisowurtzitane (III). Solvolysis in EtOH alone gave 2,6,8,12-tetranitro-4(trifluoroacetyl)-2,4,6,8,10,12-hexaazaisowurtzitane (IV) exclusively. Full deprotection to
give 2,6,8,12-tetranitro-2,4,6,8,10,12-hexaazaisowurtzitane (V) was achieved with an
ethanolic solution of sodium acetate.
The synthesis of 2,6,8,10,12-pentanitro-2,4,6,8,10,12-hexaazaisowurtzitane (VIII) was
carried out in a similar manner. We found that trifluoroacetylation of I with a solution of
TFAA in trifluoroacetic acid (TFA) gave 2,6,8,12-tetraacetyl-4-trifluoroacetyl-2,4,6,8,10,12hexaazaisowurtzitane (VI) exclusively, even when a large molar excess of TFAA was used.
A similar mono-acetylation of I using acetic anhydride in acetic acid solution has been
previously reported [5]. VI was nitrated in anhydrous mixed acid to yield 2,6,8,10,12pentanitro-4-(trifluoroacetyl)-2,4,6,8,10,12-hexaazaisowurtzitane (VII). VII was also
accessible by nitration of IV; both methods demonstrate protection of a secondary amine
group during nitration of another secondary amine group in the same molecule. (VII) was
readily deprotected with EtOH to give VIII.
2.2
Protection of hydroxyl groups during the nitrolysis of an oxetane;

synthesis of 2-methyl-2-(hydroxymethyl)-1,3-dinitratopropane
(XII)
A similar strategy was applied to the protection of an hydroxyl group during the
nitrolysis of an oxetane. It has been previously shown that in the nitrolysis of 3-methyl-3(hydroxymethyl)oxetane (IX), nitration of the hydroxyl group occurs much more rapidly
than nitrolysis of the oxetane ring, to the extent that it is possible to synthesise pure
3-methyl-3-(nitratomethyl)oxetane by careful control of the reaction conditions. However,
extended reaction times and excess reagent give 2-methyl-(2-nitratomethyl)1,3-trinitratopropane (meritol trinitrate) through nitrolytic ring opening of the oxetane [6]. We
found that it is possible to protect the hydroxyl group of IX during the nitrolysis of the
oxetane, allowing the synthesis of 2-methyl-2-(hydroxymethyl)-1,3-dinitratopropane (XII).
The trifluoroacetate ester (X) was easily prepared by trifluoroacetylation of IX with
TFAA in dichloromethane. This ester is hydrolytically stable, and any remaining traces of
trifluoroacetic acid could be removed by washing with water, without any apparent
hydrolysis of the ester taking place. The nitrolysis of X was carried out using a solution of
dinitrogen pentoxide in dichloromethane; this gave 2-methyl-2-(trifluoroacetoxymethyl)-1,3dinitratopropane (XI) contaminated with only trace amounts of the over-nitration product.
Deprotection to give XII was achieved by solvolysis in methanol at 40oC.
466
2.3
Protection of hydroxyl groups during the nitrolysis of an N-acetyl

group; synthesis of N-nitroazetidin-3-ol (XVI)
The strategy was also applied to the protection of an hydroxyl group during nitrolysis of
an N-acetyl group. N-Acetylazetidin-3-ol (XIII) was trifluoroacetylated with TFAA to give
the trifluoroacetate (XIV). Nitrolysis of XIV with dinitrogen pentoxide in dichloromethane
gave N-nitroazetidin-3-yl trifluoracetate (XV) which was deprotected by solvolysis in EtOH
at room temperature to give XVI. XVI had previously [3] been synthesised in a similar
manner from the acetate ester, but in that case much harsher conditions (acid-catalysed
hydrolysis under reflux) were required for deprotection.
3.
EXPERIMENTAL
3.1
Synthesis of polynitro-2,4,6,8,10,12-hexaazaisowurtzitanes V and

VIII and synthesis of 2-methyl-2-(hydroxymethyl)-1,3dinitratopropane (XII)
Detailed procedures for the syntheses of these compounds have already been made
available in a patent application [7].
3.2
Synthesis of N-nitroazetidin-3-ol (XVI)
3.2.1 N-Acetylazetidin-3-yl trifluoroacetate (XIV)

N-Acetylazetidin-3-ol (1.0g) was dissolved in dichloromethane (DCM, 3.5ml) and
cooled in an ice-bath before the dropwise addition of trifluoroacetic anhydride (TFAA,
5.0ml). Following the addition, the reaction was removed from the ice-bath and stirred at
room temperature for 1h. The solvent and excess of TFAA were removed on the rotary
evaporator to leave a yellowish oil. This was dissolved in DCM and washed with water. The
organic phase was dried over MgSO4, and then evaporated to leave a yellow oil (2.75g,
150% crude yield). On the basis of the 19F NMR spectrum, the crude product contained
approximately 55 mol% (40 wt%) trifluoroacetic acid, giving an overall yield of Nacetylazetidin-3-yl trifluoroacetate of around 90%. No further purification was attempted
before the nitrolysis step. NMR (CDCl3) 1H 2.0 (s, 3.0H), 4.1-4.3 (m, 2.1H), 4.4-4.7 (m,
2.4H) and 5.5ppm (m, 1.0H); 19F NMR -75.3ppm (s).
3.2.2 N-Nitroazetidin-3-yl trifluoracetate (XV)
Crude N-acetylazetidin-3-yl trifluoroacetate [2.0g crude material, containing approx.
1.2g (5.7mmol) of substrate] was dissolved in DCM (4.0ml) and cooled in an ice-bath during
the addition of a solution of N2O5 in DCM (8.5ml of approximately 140mg ml-1
concentration, 11mmol). After the addition was complete, the reaction mixture was removed
from the ice-bath and stirred at room temperature for 1h. The solution was washed with
saturated NaHCO3 (2x30ml), then brine (30ml), and dried over MgSO4. The solution was
evaporated to leave a viscous, yellow oil (0.91mg, 74%). NMR (CDCl3) 1H 4.4-4.5 (m,
2.0H), 4.7-4.8 (m, 2.0H), 5.3-5.4ppm (m, 1.0H); 13C NMR 62.6 (s), 63.6 (s), 114.0 (q,
J=283Hz) and 156.7 (q, J=44Hz); 19F NMR -75.2ppm (s).
467
3.2.3 N-Nitroazetidin-3-ol (XVI)

N-Nitroazetidin-3-yl trifluoroacetate (65mg) was dissolved in EtOH and stirred at room
temperature for 24h. Evaporation of the solvent gave a pale yellow oil (38mg, 108%) which
did not solidify on standing to give a waxy solid as previously reported [3]. NMR (CDCl3) 1H
4.2-4.3 (m, 2.0H) and 4.5-4.6ppm (m, 3.0H); 13C 58.4 and 66.34ppm. The DEPT 135,
1
H-1H correlation and 1H-13C correlation spectra were all consistent with the expected
structure.
Acknowledgements
We thank AWE (Aldermaston) for funding this research and ATK Thiokol for supplying
compound I.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
B.S. FEDOROV AND V.V. ARAKCHEEVA: Synthesis of trifluoroacetates of N-(2-hydroxyethyl)substituted trifluoroacetaminosuccindiamide and succinimide, Izv. Akad. Nauk, Ser. Khim., 5,
1321-1322, 1996.
[2] M.E. COLCLOUGH, J. HAMID, A. PELTER AND G.S. KANG: Preparation of Glycerol
Dinitrates, UK Patent Aplication GB 2 352 239A (2001).
P.R. DAVE, Acylative dealkylation of N-tert-butyl-3-substituted azetidines, J. Org. Chem., 61,
5453-5455, 1996.
A.J. BELLAMY, PETER GOLDING AND A. MACCUISH: The use of the trifluoroacetyl protecting
group in the synthesis of mono- (4) and di-amines (4,10) in the polynitroisowurtzitane series,
Propellants, Explosives, Pyrotechnics, 28, 157-158, 2003.
K-H. CHUNG, H-S. KIL, I-Y. CHOI, C-K. CHU, AND I-M. LEE: New precursors for
hexanitrohexaazaisowurtzitane (HNIW, CL-20), J. Heterocyclic Chem., 37, 1647-1649, 2000.
P. GOLDING, R.W. MILLAR, N.C. PAUL AND D.H. RICHARDS, Preparation of di- and
polynitrates by ring-opening nitration of oxetanes by dinitrogen pentoxide (N2O5), Tetrahedron,
49, 7051-7062, 1993.
A.J. BELLAMY, PETER GOLDING AND A. MACCUISH: Trifluoroacetyl as a protective group for
amino and hydroxyl groups in the synthesis of CL-20 derivatives as novel explosives, PCT Int.
Appl. WO 2004/76384 A2 (2004).
468
Ac
Ac
N
H
N
COCF3
Ac
Ac
N
HNO3/oleum
TFAA
N
N
H
Ac
N
Ac
Ac
N
O 2N
Ac
(II)
COCF3
N
COCF3
(I)
O 2N
NO2
N
EtOH
N
NO2
COCF3
COCF3
O 2N
N
H
O 2N
(III)
NO2
N
N
NO2
(IV)
NaOAc/EtOH
O 2N
NO2
N
H
N
N
H
O 2N
N
NO2
(V)
Fig 1.
Synthesis of 2,6,8,12-tetranitro-2,4,6,8,10,12-hexaazaisowurtzitane (V)
469
Ac
Ac
N
H
N
Ac
N
TFAA/TFA
N
H
Ac
Ac
N
Ac
Ac
N
O2N
HNO3/oleum
N
Ac
(VI)
NO2
N
COCF3
(I)
O2N
H
N
NO2
N
EtOH
N
NO2
COCF3
NO2
O2N
N
H
O2N
NO2
N
N
NO2
(VIII)
(VII)
HNO3/oleum
O2N
NO2
N
N
O2N
H
N
N
COCF3
N
NO2
(IV)
Fig 2.
Synthesis of 2,6,8,10,12-pentanitro-2,4,6,8,10,12-hexaazaisowurtzitane
(VIII)
470
COCF3
OH
N2O5/DCM
TFAA/DCM
(IX)
(X)
COCF3
O
OH
MeOH
ONO2
ONO2
ONO2
(XI)
Fig 3.
ONO2
(XII)
Synthesis of 2-methyl-2-(hydroxymethyl)-1,3-dinitratopropane (XII)
COCF3
OH
TFAA
N
COCF3
N2O5/DCM
N
OH
EtOH
N
Ac
Ac
NO2
NO2
(XIII)
(XIV)
(XV)
(XVI)
Fig 4.
Synthesis of N-nitroazetidin-3-ol (XVI)
471
DETERMINATION OF POST-EXPLOSION RESIDUES

OF AMMONIUM NITRATE IN ENVIRONMENTAL SAMPLES
Jan Bdek, Stanisaw Cudzio, and Sylwia Pietrzyk
Abstract:
The paper presents results of determination of ammonium nitrate, which remains in
detonation products of industrial explosives. Charges of the explosives containing
ammonium nitrate were detonated inside samples of concrete, sand and wood. After
detonation, we tried to detect ammonium nitrate in the matrixes using instrumental thin
layer chromatography (TLC). In this way a possibility of identification and
determination of ammonium nitrate on the spot of criminal events was confirmed.
Keywords:
1.
analysis, TLC, explosives, ammonium nitrate
INTRODUCTION
Analysis of post-explosion residues of explosives in environmental samples, taken on

the spot of criminal or terrorist events, is an important element of forensic investigation.
Results of such analyses enable determination of the kind of explosives used, their
composition, the method of initiation, and consequently, these make easier the search for the
culprit [1, 2].
Continuing works on this topic [3, 4], we undertook an effort to check the possibility of
application of instrumental thin layer chromatography (TLC) for analysis of ammonium
nitrate (AN) residues in samples taken from the direct neighborhood of explosion. AN is
widely used, not only as component of mining explosives but also as a fertilizer. Therefore it
is commonly available and frequently applied in criminal or terrorist activities [5].
Experimental part of this work included:
searching for a visualization reaction and chromatographic system enabling
isolation of AN from the matrixes (especially from substances that may cause
interference),
determination of the quantitative relations i.e. the dependence of the surface area
of densitometric peak on the mass of the analyte in the chromatographic band,
evaluation of the recovery of AN during extraction from the matrixes,
concentration and purification,
measurement of quantity of AN remaining in detonation products.
2.
EXPERIMENTAL
The equipment, chemicals and reagents used were: applicator Linomat IV (Camag);
vertical chromatographic chamber (Camag); densitometer CS-9000 (Shimadzu); a set for
solid phase extraction (J.T. Baker), chromatoplates HPTLC (Merck), solvents of reagents
(Merck) and mobile phases made of organic solvents (J.T. Baker, Merck).
472
Visualization of the analyte. Ammonium nitrate is a colorless substance which does not
absorb in UV region and does not cause the fluorescence quenching. Attempts of AN
visualization with diphenylamine this is a well known reagent for nitrates detection [6]
turned out ineffective. After addition of a few droplets of diphenylamine and concentrated
sulfuric acid to the aqueous AN, the solution changes to dark-blue, but the color is unstable
and gets brown very quickly. This reaction could not be carried out directly on the
chromatoplate, because of too high acidity of the medium, which destroy the adsorbent.
Good analytical effects were obtained using modified Griess-Ilosvays method [6]. It
includes reduction of nitrate ions to nitrite ions, reaction with an aromatic amine to form
diazonium cation and consecutive coupling with another aromatic amine to form a colorful
compound, Fig.1.
+
NH 2 + NO 2 + 2 H + Cl
HSO 3
N2Cl
HSO3
HSO3
+
NH2
Fig 1.
N 2 Cl + 2 H 2 O
HSO 3
N N
+ HCl
NH2
Exemplary reactions of Griress procedure of visualization of nitrates
Ammonium nitrate is soluble in water what facilitates extraction, and the Griess reaction
can be carried out in the aqueous extracts. Of course, in the case of trace analyses, the
concentration of the reaction products is too low to cause noticeable changes of the solution
color. So that compound created must be concentrated and extracted with an organic solvent,
because of the technique of sample application.
The procedure of analyses was as follows: aqueous solution of sodium hydroxide
(0.5 M), aqueous solution of copper sulfate and hydrazine sulfate (1 g of N2H5HSO4 and
0,0164 g of CuSO4 in 1 l of water), solution of sulfanilic acid (1.2 g of the acid and 55.5 ml
of concentrated HCl in 1 l of water), solution of -naphthylamine (0.9 g of the compound
and 2.4 ml of concentrated HCl in 1 l of water) were prepared. Solutions of NaOH, CuSO4
and N2H5HSO4 were added to a aqueous solution of AN, adjusting pH to 9.6. The obtained
solution was heated at 70 oC for 15 min. (reduction stage). Next it was cooled down to 0 oC
and then acetone was added to remove an excess of hydrazine. The nitrite formed at the
reduction stage was used as diazotation agent of sulfanilic acid (added in the form of the
above mentioned solution), and the resulting diazonium cation was finally coupled with naphthylamine, Fig. 1.
Different extraction techniques were used to concentrate the diazo compound and
introduced it into an organic phase. The best results were gained using solid phase extraction
(SPE). The compound is retained at the top of column filled with C18 Polar Plus packing.
After extraction, the column was washed out with hexane, dichloromethane and water (1 ml
in volume each), and the purified compound was eluted with methanol (two-time, 0.5 ml).
The obtained eluents were applied on the start line of the chromatoplates. A good
separation was achieved in an ascending chamber on silica gel with chemically bonded
473
octadecyl using methanol-water (4:1 v/v) as a mobile phase. In this system, the distance
traveled Rf by red-violet diazo compound equals to 0.84.
Quantification. The quantitative analyses were carried out using densitometric
technique. To this end a standard of the diazo compound was prepared and dissolved in
methanol (concentration of 1 ng/l). Different volumes (1200 l) of the solution was
applied on the start line of chromatoplates. After developing in the above described system,
the chromatograms were dried out and put to densitometric measurements. The
chromatograms were scanned (zig-zag technique) using light with the wavelength
corresponding to the absorbance maximum of the compound, max = 510 nm. Average
results of the measurements (each was repeated five times) were used to determine the
dependence of absorbance A (arbitrary units) on the mass of the compound in the
chromatographic band, m, Fig. 2. Analytical parameters of AN, obtained in this way, are
presented in Table 1.
Fig 2.
Dependence of absorbance on the mass of AN in the chromatographic

band
Table 1. Analytical parameters of ammonium nitrate

RF
max
[nm]
Calibration curve
A = f(m)*
Correlation
coefficient
Detection limit
[ng]
0.84
510
0.998
3
A = 4505m + 260
*A absorbance a.u., m mass of the analyte in the chromatographic band [ng]
Evaluation of AN recoveries. Samples of concrete, sand wood (sawdust) of 20 g in
mass were powdered and washed out with water to extract possible traces of nitrates. Next
the samples were contaminated with aqueous solution of AN. The degree of contamination
of each matrix was 1.5, 5.0 and 12 g/g.
The contaminated samples were extracted with water (twice, 25 ml of water each) in an
ultrasonic bath and reactions of reduction, diazotation and coupling were carried out in the
474
obtained extracts. The concentration and determination of the content of the resulting diazo
compound were performed according to the above described procedure. The results of the
experiments are shown in Tab. 2.
Table 2. Recovery of AN from different matrixes
Matrix
Recovery* [%]
Standard deviation [%]
Concrete
55
2.1
Sand
78
1.4
Wood
69
* an average value from 5 experimental results
1.8
Measurement of post-explosion residues. Heaving found a scheme for the AN

identification in environmental samples, an attempt was made to determine how much of the
compound reacts in the detonation wave of typical mining explosives ammonals and
ammonits. Experimental explosions were carried out in a steel tube with inside diameter of
120 mm, wall thickness of 30 mm and length of 800 mm. The tube was placed on a steel
plate of 10 mm thick which was used as the tube lock. The charges of 25 g in mass
surrounded by the matrix material (Fig. 3, 600 g of concrete blocks or sand, and 80 g of
wood) was positioned axially inside the tube at a distance of 1/3 tube length measured from
the bottom.
Fig 3.
Photos of fused ammonit charges immersed in matrixes:

1 concrete, 2 sawdust, 3 sand
After detonation the internal tube walls were swept with a bore brush, the tube was taken
off the base and some matrix fragments that remained on it (treated like representative
samples) were analyzed. In each case, i.e. regardless of the matrix type and the explosive,
experiments and consecutive analyses were repeated five times in order to get the possibility
to assess the methods repeatability and measuring errors. Results of the analyses are
presented in Tab. 3. The amount of AN remaining in the matrixes after detonation is given as
mass fraction of unreacted AN. This parameter was calculated as a ratio of AN mass found
in a matrix to the mass of AN in the explosive charges.
475
Table 3. Residues of AN in matrixes after detonation of a 25-g charge of ammonit

Matrix
Mass fraction of unreacted AN*
Concrete
Sand
Wood
* an average value from 5 experimental results
3.
0.0011
0.0025
0.0029
CONCLUSIONS
Searching for the visualization reaction and chromatographic system enabling isolation
of AN from environmental samples was most difficult part of the work. The procedure is
time consuming, but is characterized by comparatively high selectivity, sensitivity and
adequate precision. The proposed modification of Griess-Ilosvays reaction is especially
worth mentioning as it enables determination of nitrates in dilute solutions with
concentrations lower by two orders than it has been possible so far.
The obtained results indicate that the recoveries of AN are relatively low. However, the
information collected in Tab. 2 concerns not only the efficiency of extraction, in the classical
meaning of the parameter, but also the yield of consecutive steps of the visualization
reactions. The most important fact is that AN concentration does not affect its recovery what
implies that the procedure is correct.
From an analysis of the results presented in Tab. 3, it follows that the amount of
unreacted AN (and passing into matrixes) ranges from 0.1 to 0.3%. These figures are quite
high from the analytic point of view, i.e. they are big enough for determination with
satisfactory precision. This means that TLC can be applied for detection of AN postexplosion residues in samples taken from sites, where ammonium nitrate explosives were
detonated. The developed procedure provides reliable information on the type of the
explosive used what may have key meaning for a successful investigation.
Acknowledgment:
Military University of Technology, Grant PBS 700.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
J. YINON, S. ZITRIN: Modern Methods and Applications in Analysis of Explosives, John

Wiley&Sons, Chichester, 1993.
T. BARAN, Problemy kryminalistyki, 123 p. 562, 1976.
J. BDEK, A. PAPLISKI, A. ROSTKOWSKI AND S. NEFFE: Application of TLC for
determination of high explosive residues in water and soil samples, Chem. Anal. 43 p. 711,
1998.
J. BDEK, S. CUDZIO, S. PIETRASIAK: Determination of residues of explosives in
enviromental samples, Proc. of the 7th Seminar: New Trends in Research of Energetic
Materials, Pardubice, pp. 90-94, April 20-22, 2004.
R. HAAS: Review of suspected warfare-related environmental damage in the Federal Republic
of Germany, Part 2, Explosives Dictionary, Umweil Bundes AMT, Berlin, 1996.
H. JORK, W. FUNK, W. FISCHER, H. WIMMER: Thin-Layer Chromatography, VCH, Weinhem,
1994.
476
INTERACTIONS BETWEEN THE NITRAMINES RDX, HMX

AND CL20 WITH THE ENERGETIC BINDER GAP
Michael A. Bohn*, Anton Hammerl**, Kate Harris***, and Thomas M. Klaptke**
* Fraunhofer Institut Chemische Technologie, Pfinztal, Germany
** Chair of Inorganic Chemistry, LMU Munich, Munich, Germany
*** Department of Chemistry, University of Edinburgh, Edinburgh, U.K.
Abstract:
The nitramines RDX, HMX and Cl-20 are commercially used as explosives.
We investigated their decomposition pathways computationally and modelled their
decomposition in the presence of the energetic binder GAP.
Keywords:
1.
CL20, RDX, HMX, GAP, decomposition, computational chemistry
INTRODUCTION
The cyclic nitramines RDX and HMX as well as the cage nitramine CL20 are explosives
which are either in commercial use or under investigation as potential military explosives.
For the formulation of energetic materials GAP (glycidyl azide polymer) is often used as an
energetic binder. The purpose of this study is the theoretical determination of the influence
of GAP on the decomposition mechanism of the nitramines and its influence on the stability
of the mixtures.
2.
CALCULATIONS
The most stable isomers of the three molecules are
-HMX, -CL20 and the diaxial conformation of RMX. For our calculations we used
the conformation of CL20 since this the most promising polymorph.
-HMX
Fig 1.
diaxial RDX
Ground states of -HMX, diaxial RDX and -CL20.
477
-CL20
The decomposition of the nitramines is very complicated. While the breaking

of a N-NO2 bond is considered to be the most favored initial decomposition step, several
other decomposition mechanism have been investigated. Kinetic data suggest that during the
actual decompostion of CL20 the breaking of C-H bonds is the rate determining step and
thermodynamically this represents the most favored decomposition path.
NO2(-)
O
O
N
N
N
NO2
NO2
(b)
(h)
O
N
NO2
(c)
NO2
O2N
N
N
NO2
NO2
O2N
HMX
O2N
O2N
(d)
N
N
(g)
NO2
NO2
H2C
NO2
(a)
NO2
O2N
NO2
NO2
O2N
O2N
N
O2N
N
N(+)
(e)
(f)
NO2
O2N
NO2
N
N
N
NO2
O2N
NO2
O2N
N
N
N
N
N
N
H
O2N
HC
N
NO2
NO2
NO2
(a) Transfer of an oxygen from an NO2 group to an adjacent CH2 group.

(b) Heterolytic fission of a C-N bond. (c) Elimination of H2CNNO2.
(d) Concerted depolymerisation to four H2CNNO2 molecules.
(e) Homolytic fission of an N-NO2 bond. (f) Homolytic fission of a C-N bond.
(g) Elimation of HONO (nitric acid). (h) Elimation of HNO2 (nitric acid).
Fig 2.
Proposed pathways for the first step in the decomposition of HMX.
We calculated the transition states for the decomposition pathways shown above for the
nitramines at the B3LYP/6-31G(d) level of theory. The obtained values for the activation
energies agree well with literature data (s. Table).
We also determined the lowest energy conformations of mixtures of the nitramines with
GAP-Diol.
478
Fig 3.
Alignment I of the mixtures of the nitramines HMX, RDX and CL20

with GAP.
In order to evaluate the influence of different GAP conformations we also used a second
conformation in our calculations:
Fig 4.
Alignment II of the mixtures of the nitramines HMX, RDX and CL20

with GAP.
The transition states obtained for the decomposition of the nitramines were investigated
in the mixtures of the nitramines with GAP-Diol on the B3LYP/3-21G level of theory. A
comparison of the values obtained for the decomposition of the nitramines on with
the 6-31G(d) and the 3-21G basis sets shows that the smaller basis set yields lower activation
energies.
479
Table 1.
Ea
[kJ mol1]
Calculated activation energies for the decomposition pathways given

in Fig 2.
Ref.
(a)
HMX
+ GAP
RDX
RDX
+ GAP
(d)
231.4
CL20
+ GAP
2.1
(f)
(g)
186.6
[2]
166.2
177.5
[3]
193.3
[4] (i)
(ii)
172.4
201.3
174.9
234.3
300.4
169.5
251.0
159.0
(h)
159.0
179.5
B3LYP/631G(d)
192.5
229.0
178.5
292.8
182.0
B3LYP/3-21G
182.5
162.8
299.0
149.6
B3LYP/3-21G /
I
215.6
161.2
173.4
B3LYP/3-21G /
II
169.3
166.4
[1]
220.5
[5]
228.9
143.5
[6]
248.5
163.2
164.0
B3LYP/631G(d)
219.9
262.4
248.6
173.2
164.1
B3LYP/3-21G
213.8
244.7
152.1
133.6
B3LYP/3-21G /
I
209.6
261.9
174.0
152.7
B3LYP/3-21G /
II
154.7
115.7
[7]
CL20
(e)
166.5
[1]
HMX
(b)
177
B3LYP/631G(d)
267.9
246.7
B3LYP/3-21G
244.4
153.1
222.1
B3LYP/3-21G /
I
241.7
151.7
B3LYP/3-21G /
II
225.3
145.4
224.4
N-NO2 bond fission
The N-NO2 bond fission is considered to have no significant activation energy. We

calculated the reaction paths for the singlet and triplet.
480
1200
-HMX/GAP
-HMX/GAP
-HMX/GAP II
-HMX/GAP II
relative energy / kJ mol
-1
1000
Singulet
Triplet
Singulet
Triplet
800
600
400
200
0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
d (N-N) /
2200
RDX/GAP Singulet
RDX/GAP Triplet
RDX/GAP II Singulet
RDX/GAP II Triplet
2000
-1
1800
1600
1400
1200
1000
800
600
400
200
0
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
d(N-N) /
1200
-CL20/GAP Singulet
-CL20/GAP Triplet
-CL20/GAP II Singulet
-CL20/GAP II Triplet
-1
1000
800
600
400
200
0
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
d(N-N) /
Fig 5.
Reaction profile for the N-NO2 bond fission of mixtures of -HMX (a),
RDX (b) und -CL20 (c) with GAP-diol at the B3LYP / 3-21 G level of
theory.
481
In agreement with experimental evidence we find that the decomposition of CL20 and to
a lesser extent that of HMX is favored in the presence of GAP. The conformation of GAPDiol plays an important role in this decomposition reaction with energy differences up to 60
kJ mol1.
2.2
HONO Elimination
While the axial elimination of HONO is preferred by RDX, axial and equatorial
elimantion are similar for HMX. HONO elimination is not favored for CL20 due to the
formation of a bridgehead double bond. Transition states for the elimination of the other
tautomer of nitric acid, HNO2, were not found. The optimization always leads to the
transition states for the elimination of HONO. The addition of GAP-Diol favors the
elimination of HONO only in one conformation for RDX, in all other cases the activation
barrier is increased by the addition of GAP-Diol.
Fig 6.
Transition states for the elimination of HONO from HMX, RDX and CL20
and mixtures of these nitramines with GAP-diol. The bond distances of the
transition are given in .
482
2.3
Oxygen Transfer
The activation barriers for the oxygen transfer reaction for HMX and CL20 have about
the same magnitude as the barriers for the HONO elimination. For RDX the activation
barrier is higher. The addition of GAP-Diol lowers the activation barriers for RDX and CL20
and increases the activation barrier for HMX.
Fig 7.
Transition states for the transfer of an oxygen atom from the NO2 group
of HMX, RDX and CL20 to an adjacent CH2 group. The bond distances
of the transition states of the nitramines at the B3LYP/3-21G and
B3LYP/6-31G(d) level of theory and the bond distances of the transition
states of the mixtures of the nitramines with GAP-diol at the B3LYP/3-21G
level of theory are given in .
483
3.
CONCLUSION
The calculated activation barriers for the decomposition of the nitramines RDX, HMX
and CL20 change in the mixtures with GAP-Diol, a monomer unit of the energetic binder
GAP.
The conformation of GAP-Diol is very important for the decomposition, different
conformers accord for differences of up to 60 kJ mol1 in activation energy. Thus our
calculated results show only an upper limit for the activation barrier, since other
conformations, which were not calculated for cost reasons, may lower the activation barrier
even more.
Nonetheless our results show that the decomposition barriers of the nitramines are
lowered by the addition of GAP-Diol and that the decomposition mechanism changes upon
addition of GAP-Diol, which agrees well with the observed decomposition products.[8]
REFERENCES
All calculations were performed with the Gaussian03 package.
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
D. CHAKRABORTY, R. P. MULLER, S. DASGUPTA UND W. A. GODDARD III: J. Phys. Chem. A,

2001, 105, 1302-1314.
S. ZHANG, H. N. NGUYEN UND T. N. TRUONG, J. PHYS. CHEM. A, 2003, 107, 2981-2989.
R. SHAW UND F. E. WALKER, J. PHYS. CHEM., 1977, 81, 2572-2576.
J. P. LEWIS, K. R. GLAESEMANN, K. VANOPDORP UND G. A. VOTH, J. PHYS. CHEM. A, 2000,
104, 11384-11389.
C. J. WU UND L. E. FRIED, J. PHYS. CHEM., 1997, 101, 8675-8679.
D. CHAKRABORTY, R. P. MULLER, S. DASGUPTA UND W. A. GODDARD III, J. PHYS. CHEM. A,
2000, 104, 2261-2272.
J. C. OXLEY, A. B. KOOH, R. SZEKERES UND W. ZHENG, J. PHYS. CHEM., 1994, 98, 70047008.
M.A. BOHN, THERMOCHIM. ACTA, 2003, 401, 27-41.
484
THE INFLUENCE OF PHYSICAL STRUCTURE

OF AMMONIUM NITRATE PRILLS ON DETONATION PROPERTIES
OF ANFO-EXPLOSIVES
Daniel Buczkowski*, Adam Presz**, and Bogdan Zygmunt***
* Institute of Industrial Organic Chemistry, Annopol St. 6, 03-236, Warszawa, PL

** Institute of High Pressure Physics of PAS, Sokoowska St. 29/37, Warszawa, PL
*** Military Academy of Technology, Kaliskiego St. 2, 00-908 Warszawa, PL
Abstract:
A structure of ammonium nitrate prills about different porosity has been examined by
using scanning electron microscope. Porous prills have been obtained from agricultural
product by thermal treatment. Detonation velocity and cylinder expansion test have
been run for ANFO made from above-mentioned prills. Disclosed by microscopic
examinations prills structures have been related with detonation properties of ANFO
manufactured from different kind of prills.
Keywords: porous ammonium nitrate, ANFO, detonation properties, physical structure
1.
INTRODUCTION
Almost all used explosives have heterogeneous physical structure. Physical form of
industrial explosives may be solid, liquid, plastic, half-liquid or gelatinous. Explosives may
have homogenic or multiphases form. Results of experiments point out that physical
structure of explosives has a significant influence on detonation ability of explosives, like
minimum stimuli needs for initiation, critical diameter and detonation velocity. This is
especially important for non-ideal detonation. Very large influence on afore-mentioned
properties affect such elements of physical structure as: granulometric composition, porosity
(related with density), existence of impurities (especially about different densities) and, in a
case of mixtures, rate of contact between ingredients [1-4]. Taking into account explosives
about low sensitivity like ammonites, ANFO or slurry explosives physical structure has
crucial influence on ability to detonation [5,6]. In the case of explosives about high sensitivity,
like strong brisant explosives, affect of physical structure on some explosive properties is
also significant [7]. Changing physical structure of explosives, by using suitable technological
ways, make possible manufacturing product about various properties, fitted to customers
needs.
Above half of explosives consumes by worlds mining are ANFO. From physical point
of view this kind of explosive is a two component, multiphases mixture consists of inorganic
oxidizer and organic liquid fuel. Explosive properties of such type material are in a
significant degree determined by physical structure of ammonium nitrate (AN) prills. In
papers [8,9] authors tried to obtain porous AN about different porosity, from one batch of
agricultural AN. Next the influence of prills porosity on detonation velocity of ANFO was
examined. Increasing of AN prills porosity affected on significant increasing of detonation
velocity of ANFO (manufactured from afore-named prills), although density of the charges
485
decreased significantly. ANFO prepared from AN about oil absorption less than 2,5 cm3/100
g didnt detonate during experiments. The least oil absorption of AN prills, that ANFO made
of such prills detonated, were about 2,5-3,0 cm3/100 g and measured detonation velocities
were about 1,6-1,7 km/s. Accordingly to rising oil absorption of prills, rose also detonation
velocity of ANFO. Maximum values of detonation velocities were reached for oil absorption
about 12-15 cm3/100 g and were about 2,7 km/s. If oil absorption of prills still increased,
detonation velocity of ANFO decreased.
Prills of AN used as fertilizer shouldnt have porous structure; prills ought to be compact
and mechanically resistant. Porous AN is used only for producing explosives and there are a
few industrial methods of manufacturing porous prills. One of the earliest used methods was
applied in Poland thermal treatment. This method utilizes the phenomenon of changing
density of AN during polymorphic transitions [10,11,12]. Nowadays porous AN is mostly
manufactured immediately during the process of tower granulation. During this process a
deposition of a few percent water is evaporated or gas producing substances are added to
molten AN. Because that kind of AN is produced only for mining, in the next part of this
paper such AN will be called mining AN.
The aim of this work was disclosure microscopic changes which run in single prill of
AN according to rising its porosity. Observed changes in physical structure of prills were
connected with detonation properties of ANFO (detonation velocity and Guerneys energy
determined by cylinder expansion test) manufactured from examined prills.
2.
A METHOD OF MAKING POROUS AMMONIUM NITRATE
The most suitable for making porous prills by thermal treatment is AN with no inorganic
additives. But such type of AN is produced very rarely. Almost all AN contain inorganic
additives which improve their quality and make a process of manufacturing porous AN more
difficult. Relatively a small influence on this process has ammonium sulfate. Another
additives like calcium or magnesium nitrate cause, that it is very hard to produce AN about
high porosity. Used in this work AN contained about 1 % of ammonium sulfate.
The process of making porous prills was based on a polymorphic transition of AN,
which runs at + 32oC.
32C
3
IV (1,72 g/cm )
32C
3
III (1,66 g/cm )
IV (1,72 g/cm3)
During the transition IV III IV density of AN crystals significantly changes, which

causes creating porous structure and cracking of prills. Porous product has bigger oil
absorption and ability to detonation measured by tube test [11].
A technology relied on heating AN above 32C and cooling; an additive which increased
porosity was also used [12]. Due to different number of thermal cycles heating/cooling and
amount of mentioned additive, AN about broad scope of physical characteristics were
produced. Repeating the thermal cycles five times and dosing optimal amount of the
additive, AN about oil absorption on a level of 20 cm3/100 g were obtained. Further
increasing of porosity wasnt useful because prills have small mechanical resistance and
crushed during next operations. Examinations of physical properties of produced AN were
done; next we made ANFO and determined their detonation properties.
486
3.
CHARACTERISTICS OF AMMONIUM NITRATE PRILLS

AND ANFO-EXPLOSIVES
Absorption of fuel oil by AN prills was measured by sticking prills in wall of

Erlenmeyer flask. Maximal amount of fuel oil added to 100 g AN which didnt caused
sticking prills in the wall was named as oil absorption and admitted as a measure of porosity.
Absolute uncertainty of this method is assessed about 0,5 cm3/100 g.
Measurements of physical properties of AN prills showed, that increasing porosity of
prills (measured by oil absorption) were accompanied by decreasing of products density and
mechanical resistance of prills.
Examinations of prills physical structure were done by using samples about different
porosity manufactured from fertilizer product. Samples about oil absorption 0 (non-porous),
8 and 15 cm3/100 g were chosen.
Detonation properties of ANFO (94,5% AN, 5,5% fuel oil) were measured by shooting
charge of such explosive. ANFO were placed in a copper tube about internal diameter 25
mm and walls thickness 2,5 mm. A charge consists of 14 g RDX was used as a booster.
During the shot detonation velocity was measured and a photo of expanded tube was done.
Using the photo so called Guerneys energy (the sum energies of tube and detonation
products) was next determined.
Physical properties of examined AN samples and detonation characteristics of ANFO
manufactured from these samples are presented in table 1. This table contains also physical
properties of mining AN and milled AN and so detonation characteristics of ANFO.
Table 1. Physical properties of various kinds of ammonium nitrate and detonation
properties of ANFO.
Sample
A
B
C
D
E
Kind of
AN
Oil
absorption
[cm3/100g]
agricultural
0
agricultural
8
porous
agricultural
15
porous
Mining
11
Milled
-
Prills
diameter
[mm]
2,0-2,5
2,2-2,8
Density of Detonation Guerneys

ANFO
velocity
energy
3
[g/cm ]
[km/s]
[MJ/kg]
1,03
didnt detonate
0,77
1,79
0,65
2,3-2,9
0,67
2,14
0,86
1,0-1,5
< 0,5
0,78
0,84
2,50
3,44
1,47
As may be seen from the table samples A, B and C, manufactured from the same raw
product, characterise increased oil absorption as a result of thermal treatment. From the same
reason diameter of these prills grow slightly. Density of ANFO manufactured from that
kinds of AN successively falls down and detonation characteristics improve.
Mining AN has oil absorption between samples B and C, but diameter of prills is
significantly smaller. ANFO produced from that AN has relatively high density and better
detonation properties than explosive manufactured from samples A, B and C. ANFO
produced from milled AN characterizes the best detonation properties which are
significantly better than the others.
487
4.
MICROSCOPIC EXAMINATIONS
OF AMMONIUM NITRATE PRILLS
An object of microscopic examinations were single prills or grains of AN. Before tests
prills were mechanically fractured, so the inner structure, including contraction cavity, might
be observed.
Photos were done by scanning electron microscope (LEO 1530), with Gemini column
and field emision gun (Shottk'y cahtode). This is high resolution microscope which works
with low energy of electrons bundle. This property makes possible obtaining large
magnifying power (high resolutions) without significant heating examining objects. This
feature is very important in the case of materials such AN, which takes a course of
polymorphic transition in temperature as low as 32C. Used magnifyings were 75, 500 and
2000 times. The least from chosen magnifyings enable showing in photo practically all
fractured prill. Magnifying 500 times made possible to identify single crystals, which put
together to a crystallite creates prill. And finally magnifying 2000 times enable determining
the nature of voids (crevices and pores) inside the prill. Photos are enclosed at the end of the
paper.
The subsequent photos (number 1-3) show series of four photos, made by electron
microscope. Magnifying power in these series grows accordingly 75, 500 and 2000 times.
Each series consists of four photos marked A-D, which corresponds samples presented in
table 1.
The fourth series of photos (Fig. 4) shows fine grains (sample E) of AN manufactured by
mechanical milling non-porous prilled AN (sample A); used magnifyings are 75, 200, 500
and 25000 times. ANFO made from such AN characterizes the highest detonation
parameters.
5.
RESULTS OF EXAMINATIONS
In the first series of photos (A-D), made by magifying 75 times, differences in prills
structure, caused by different technologies of manufacturing, are very strong presented. Prill
of agricultural AN (photo 1A) characterizes compact, solid structure with occurred single
fractures induced by stress created during solidifying liquid droplet of AN. The inside of
contraction cavity and especially outer surface of prill are relatively smooth. This makes that
prills surface seen without any equipment is shiny. Porous prills made from fertilizer
product by the way described in chapter 2 are completely different (photos 1B and 1C).
Polymorphic transition induces creating a space net of fractures, existed as well on external
surface of prills as on the surface of contraction cavity inside prills. Differences between
photos of prills about oil absorption 8 and 15 cm3/100 g arent so big. Disclosed in the
photos the net of capillary fractures, existing on outside surface of prills is doubtlees the
reason of ability of absorbing and permanent retaining liquid hydrocarbons. These properties
affect on detonation ability and other explosive properties such the mixture.
Physical structure of mining AN is signifacantly different (photo 1D). That prill is a
polycrystal consists of many single, slightly elongated crystals about length of a few ten
micrometers, which are tightly placed and partly fused. The whole prill is penetrated by a
net of jointed pores and crevices which enables the ability of absorbing liquids. The
regularity of prills structure and similar dimensions of single crystals may testify about
programmed crystallization of AN prills with using a substance creating crystallization seeds
during prilling process. Polycrystalline structure gives this kind of AN better detonation
properties.
488
In the photos series 2 (A-D), made by magnifying 500 times, it is possible to seen details
of crystals surface. The surface of crystal of agricultural AN (photo 2A) is smooth,
boundaries among crystals are clear and sharply marked. Intercrystalline space dont create a
net of connected pores, which confirm the evidence about a lack of ability to absorb organic
liquids. Result of polymorphic transitions run in fertilizer prills is fracturing of bigger
crystals inside prill and creating many smaller. These phenomena cause that boundaries
among crystals arent so sharp as previous (photo 2B) and uplifting crystals surfaces,
formerly smooth; this may be especially good seen on photo 2C. Increasing the volume of
pores and accordingly development of surface, which in turn will be the place of contact
between AN and oil (oxidizer/fuel), acts for rising detonation properties of such the mixture.
Photo 2D shows the regular, fine crystalline structure of mining AN prill. Such structure
determines good resistance properties of prills and high detonation characteristics of ANFO
manufactured from such AN.
Images presented in photos series 3 (A-D) confirm above-mentioned conclusions. This
large magnifying (2000 times) enables quite well visibility of high development of pores
surface, especially for prills about big oil absorption. Additionally dimensions of pores
inside prills may be estimated. In the case of prills thermally treated diameter of pores is
above 10 m, but for mining AN it is a few m.
Photos series 4 (A-D) show irregular grains of AN obtained by mechanical milling nonporous prilled AN. Dimensions of grains are from about 15 to a few hundred m, but the
share of smaller grains is very significant. Because grains arent porous, oil added to AN is
placed on their surface. Very high development of the surface makes this ANFO more
energetic even than ANFO manufactured from prilled AN about high porosity. The surfaces
of milled grains are rough and additionally make the absorption of oil easier, so in turn
induces improving the run of oxidizing reaction of organic fuel in the zone of chemical
reaction during detonation process.
6.
CONCLUSIONS
Realized experiments showed, that it is possible to produce prilled AN about the same
dimensions but significantly different physical structure. Microscopic photos of crystalline
structure of prills inner of agricultural AN showed clear differences induced by thermal
treatment and run of polymorphic transition accompanied by density changes. Created as a
result of density changes inner stresses induce increasing of uneveness inside the volume of
prill fracturing and diminishing dimensions of single crystals, uplifting previously smooth
crystals surfaces, rising volumes of pores and crevices determining oil absorption.
The crystalline structure of AN prills, which are polycrystals, implies detonation
properties manufactured next ANFO. If we compare crystalline structure prills of
agricultural AN after thermal treatment and prills of mining AN and successively grains of
mechanically milled AN it is possible to correlate detonation properties manufactured from
them ANFO with elements of prills physical structure (porosity oil absorption,
prills/grains dimensions, dimensions of crystals creating prills). The first rank deciding
about detonation ability of ANFO is doubtless porosity of AN prills (grains) determining the
level of oil absorption, increasing sensitivity to initiation and significantly improving
energetic characteristics of explosive. As it was determined in formerly papers ANFO
manufactured from prilled AN about minimum level of oil absorption 6-8 cm3/100 g has
satisfactory energetic properties. Optimum range of oil absorption is 10-12 cm3/100 g. An
additional useful requirement about relatively high density of ANFO is needed. These
requirements fulfil ANFO manufactured from mining AN. Second factor in the rank, also
489
important element of physical structure influenced on detonation characteristics of ANFO is

granulometric composition of prills/grains. This was confirmed by detonation parameters;
ANFO manufactured from milled, fine, nonporous AN has the highest detonation
characteristics. But it is worth to add, that milled AN fulfils the first requirement - high oil
absorption as a result of large development of surface by small crystals.
Dimensions of crystals forming prills, seem to influence in not so big degree on
detonation parameters of ANFO. However taking into account their topology and space
array, their define the level of porosity as well as geometry of pores. These factors have
direct influence on detonation properties of ANFO. More important is affect of crystals
dimensions on prills mechanical resistance important, utilizable parameter of AN and
ANFO.
REFERENCES
[1]
[2]
[3]
[4]
F. A. BAUM I IN. Fizika vzryva, Nauka, Moskva 1975.

D. SMOLESKI. Teoria materiaw wybuchowych, MON, Warszawa 1954.
E. WODARCZYK. Wstp do mechaniki wybuchu, PWN, Warszawa 1994.
A. MARANDA, J. NOWACZEWSKI, J. STATUCH, M. SYCZEWSKI, B. ZYGMUNT. Chemia
stosowana, WAT, Warszawa 1985.
[5] B. ZYGMUNT: Wpyw struktury fizycznej na inicjowanie detonacji niejednorodnych materiaw wybuchowych. Organika - Prace naukowe IPO, 1999, s.7-19.
[6] B. ZYGMUNT: Rola fizycznej struktury niejednorodnych materiaw wybuchowych w procesach inicjowania i propagacji fal detonacyjnych. Organika - Prace Naukowe IPO, Warszawa
1998.
[7] . :
BB . , 16, 4, 1980, s. 89-93.
[8] D. BUCZKOWSKI, B. ZYGMUNT: Wpyw struktury fizycznej granul saletry amonowej na
waciwoci wybuchowe saletrolu (ANFO). Biul. WAT 53 ( ) 2004, s.
[9] D. BUCZKOWSKI, B. ZYGMUNT: Influence of ammonium nitrate prills porosity and dimesions
on detonation velocity of ANFO explosives. V Int. Seminar New trends in research of
energetic materials. Pardubice, 21-23.04.2003.
[10] D. BUCZKOWSKI, W. PGOWSKI, B. ZYGMUNT: Evaluation of the influence of modificating
substances on resistance to detonation of fertilizer grade ammonium nitrate. Proceedings of the
III Seminar - New Trends in Research of Energetic Materials. Pardubice, Czech Republic,
April 12-13, 2000, p. 29-39.
[11] D. BUCZKOWSKI, W. TRZCISKI, B. ZYGMUNT: Badanie waciwoci energetycznych saletrolu
z wykorzystaniem testu cylindrycznego. V Midzynarodowa Konf. Uzbrojeniowa Naukowe
aspekty techniki uzbrojenia. Waplewo, 9-11.10.2004, s.74-82.
[12] Polish Patent nr 95331 (1974).
490
Fig 1.
Photos series 1 magnifying 75 times.

491
Fig 2.
Photo series 2 magnifying 500 times.

492
Fig 3.
Photo series 3 magnifying 2000 times (except photo 3C magnifying 1000 times).
493
Fig 4.
Photo series 4 (sample E) magnifyings 75, 500, 2000 and 25000 times.
494
DETONATION ARRESTER PERFORMANCE

FOR HYDROGEN-AIR MIXTURES
A. Dbkowski*, A. Sapiski**, A. Teodorczyk*, and W. Witkowski***
* Warsaw University of Technology, ITC, Nowowiejska 21/25, 00-665 Warszawa, Poland
** Sapinski Engineering Consulting. Lidzbarska 52, 03-085 Warszawa
*** Institute of Industrial Organic Chemistry, Annopol 6, 03-236 Warszawa, Poland
Abstract:
Experimental data on the feasibility of a specific detonation arrester are provided in this
study. This new type of arrester consists in a series of segments filled with shot. The data
are aimed at demonstrating the pertinent use of this type of arrester for the safety in the
chemical and oil industry. The experimental results show that this device is appropriate
for quenching of flame and detonation in hydrogen-air mixtures.
Keywords:
1.
arrester, flame, detonation, hydrogen-air mixtures
INTRODUCTION
In a number of industrial processes hazardous gases build up in the working

environment. Examples include transport of combustible gases in the petrochemical industry,
the operation of process plant machinery and vent pipes of storage tanks containing
flammable liquids. For safety reasons it is common to install flame and detonation arresters
at some points of installation capable to withstand flame or detonation impacts without
combustion propagating downstream.
Detonation arresters are designed with a shock absorber in front of the flame arresting
element (Halstrick, 1995) [1] or with so called momentum attenuator (Westech, 1989) [2].
Detonations occurring in piping have velocities of about 2000 m/s and in closed process
vessels, pipes and equipment can generate pressures from 20 to 100 times the initial
pressure. Most detonation arresters have crimped metal ribbon elements, although expanded
metal cartridges, perforated plates, ceramic balls and metal shot are also used (Grossel,
2002) [3]. Arrester elements for detonation arresters are usually longer than for deflagration
arresters. The role of internal arrester element (matrix) is to quench the flame and cool the
products of combustion.
Taking into consideration complexity of detonation wave it is surprising that
thermodynamic predictions of detonation properties are very accurate. Despite the success of
CJ theory in predicting the parameters of the established detonation, such as velocity and
pressure, the theory is unable to give any guidance on whether detonation can be initiated
readily in a mixture for a given geometry and other initial conditions (e.g. pressure and
temperature) in a pipe or vessel of given dimensions. The cause of this difficulty is the
complex interaction between the reaction chemistry and the gas flow dynamics, as
characterized by the transverse wave structure.
495
Initiation of a steady-state detonation requires that the self-sustaining multidimensional

structure must develop. In a related way, if transverse structure can be eliminated, than the
detonation fails.
Despite some attempts made to provide a theoretical basis for the design of flame and
detonation arresters (Edwards, 1991) [4] and the significant effort expended on flame arrester
development, there is no quantitative theoretical basis for their design and operation.
Qualitatively, they must operate by reducing the momentum associated with the gas
motion in a detonation, while reducing the temperature of the products of combustion to the
point where reaction in the unburned mixture can no longer be self-sustaining, and the flame
is quenched.
Any design guidelines that do exist are usually based on empirical observations rather
than fundamental theory.
2.
EXPERIMENTAL
This experimental study was devoted to the evaluation of the quenching capabilities of
specific arrester with quenching element filled with metal shot. Figure 1 shows schematic of
the arrester with specially designed segmented housing (Fig.2). Figure 3 shows the view of
arrester attached to detonation tube.
Fig 1.
Schematic of detonation arrester with metal shot
496
Fig 2.
The view of arrester housing containing metal shot
Fig 3.
The view of arrester attached to detonation tube
The size of the steel balls was 1 mm in diameter. The housing for metal shot was 155
mm in diameter and constructed in sections. First section is 60 mm long and other 35 mm
long. It is possible to form housing of different length up to 200 mm, according to demand.
In this study only first section was tested. The balls are packed tightly together within the
housing to prevent movement. One advantage of this arrester is ease of assembly and
disassembly for cleaning purposes. Another advantage is that it can be made sufficiently
robust to withstand detonations. Praxair has released information that their design for nickel
shot contained in a thick-walled housing successfully stops acetylene detonation at initial
pressures from 1 to 25 bar (Praxair, 2005) [5].
Experiments were carried out using detonation tube (Fig.4) of 50 mm in diameter in
which detonation wave was initiated by spark plug in stoichiometric hydrogen-air mixture at
P = 1 bar and room temperature. Then, detonation propagates through arrester and comes out
to receiver section. Behavior of combustion wave in the receiver section was studied using
PCB pressure transducer and photodiode.
497
Fig 4.
Schematic of the experimental set-up;

1 spark ignitor, 2 acetylene-oxygen mixture, 3 valve,
4 vacuum pump, 5 hydrogen-air mixture, 6 valve,
7 vacuum pump, 8 arrester,
9 - PCB pressure transducer (P0, P1, PP, P4, P3),
10 - photodiode (P2, D), 11- dump tank
3.
RESULTS
Preliminary tests were performed with the use of one section of 60 mm long inner
arrester housing filled with metal shot of 1 mm in diameter. Figure 5 shows pressure and
photodiode profiles registered during experiment. Colours of the profiles are the same as the
respective lines in Fig.4. The first black line shows pressure profile recorded by transducer
P1 before arrester. The profile is typical for steady detonation wave. The second red line
shows pressure profile recorded by transducer PP located behind arrester. The profile shows
weak shock wave which passed through arrester. The third brown line shows light emission
profile recorded by photodiode D located behind arrester and 0.1 m in front of transducer PP.
This profile shows that there is re-ignition of hydrogen-air mixture behind arrester but it is
much delayed with respect to shock wave passage. The last lines (green and blue) show
pressure profiles recorded by transducers P4 and P3 at the end of the tube. These profiles
show the existence of weak shock wave.
Figure 6 presents average wave velocity versus distance showing that arrester attenuated
detonation wave to a shock wave of about 650 m/s in velocity and the flame.
Further studies are planned to analyse longer sections of metal shot filling as well as
influence of different size of shot on the effectiveness of arrester.
498
Fig 5.
Recorded pressure and photodiode profiles for detonation passage through

arrester
2000
1800
1600
1400
m/s
1200
1000
800
600
400
200
0
0
Fig 6.
Average velocity of the wave along the detonation tube
REFERENCES
[1]
[2]
[3]
[4]
GROSSEL S.S.: Deflagration and Detonation Flame Arresters, AIChE, 2002

EDWARDS J.C.: Thermal Models of a Flame Arrester, Bureau of Mines, RI 9378, 1991
WESTECH INDUSTRIAL LTD.: Flame Arrester Seminar Notes, Calgary Canada Halstrick V.
(1995), Technical Report Part 1, Protego Fundamentals, Braunschweiger Flammenfilter
GmbH, Braunschweig, Germany, 1989
www.praxair.com, 2005
499
IMPROVED SYNTHESIS AND X-RAY STRUCTURE

OF 5-AMINOTETRAZOLIUM NITRATE
Moritz v. Denffer*, Gerhard Heeb**, Thomas M. Klaptke*, Gernot Kramer*,
Gunnar Spie*, and Jan M. Welch*
* Chair of Inorganic Chemistry, University of Munich,
Butenandtstr. 5-13 (D), D-81377 Munich, Germany
Swisttal-Heimerzheim; Groes Cent, D-53913 Swisttal (Germany)

Abstract:
5-Aminotetrazolium nitrate was synthesized in high yield and characterized using
Raman and multinuclear NMR spectroscopy (1H, 13C, 15N). The molecular structure of
5-aminotetrazolium nitrate in the crystalline state was determined by X-ray
crystallography: monoclinic, P 21/c, a = 10.5493(8), b = 3.4556(4), c = 14.606(1) ,
= 90.548(9), V = 532.44(8) 3, Z = 4, = 1.847 g cm-1, R1 = 0.034, wR2 (all data) =
0.090. The thermal stability of 5-aminotetrazolium nitrate was determined using
differential scanning calorimetry, the compound decomposes at 169C. The enthalpy of
combustion (Hcomb.) of 5-aminotetrazolium nitrate ([CH4N5]+[NO3]-) was determined
experimentally using oxygen bomb calorimetry: Hcomb.( [CH4N5]+[NO3]-) = -6020
200 kJ kg-1. The standard enthalpy of formation (Hf) of [CH4N5]+[NO3]- was
obtained on the basis of quantum chemical computations at the electron-correlated ab
initio MP2 (second order Mller-Plesset perturbation theory) level of theory using a
correlation consistent double-zeta basis set (cc-pV-DZ): Hf([CH4N5]+[NO3]- (s)) =
+ 87 kJ mol-1 = + 586 kJ kg-1. The detonation velocity (D) and the detonation pressure
(P) of 5-aminotetrazolium nitrate was calculated using the empirical equations by
Kamlet and Jacobs: D([CH4N5]+[NO3]-) = 8.90 mm s-1 and P([CH4N5]+[NO3]-) = 35.7
GPa.
Keywords:
1.
aminotetrazole, 5-aminotetrazolium nitrate, explosives
INTRODUCTION
The synthesis of energetic, non-nuclear materials for possible military application has
been a long term goal in our research group [1-3]. We recently prepared and determined the
molecular structures of diaminotetrazolium nitrate (HDAT-NO3) (Table 1) [4,5]. Although 5aminotetrazole monohydrate has been known for a long time and its structure and acid/base
behavior were investigated extensively [6-11], the chemistry and structure of its nitrate salt,
namely 5-aminotetrazolium nitrate, has not been the subject of intensive investigations.
Here we report on an improved synthesis of 5-aminotetrazolium nitrate and for the first time
on its spectroscopic characterization and the determination of the molecular structure using
single crystal X-ray crystallography.
500
2.
EXPERIMENTAL
2.1
Synthesis
5-Aminotetrazolium nitrate: To 2 mL boiling water were added 1.03g (10.0 mmol) 5aminotetrazole (Aldrich). Following complete solvation of 5-aminotetrazole, 2.0 mL
concentrated nitric acid (65%) were added over one minute and heating was immediately
thereafter removed. Within a few minutes feathery white crystals were observed. After 5
minutes cooling time, the reaction vessel was transferred to an icewater bath resulting in the
crystalization of an increased amount of feathery white of 5-aminotetrazolium nitrate. This
material was washed 3 times with diethylether, recrystalized once from boiling water,
washed 3 additional times with diethylether and then dried under vacuum. Yield: 1.24g
(84%).
2.2
Analytical data
CH4N6O3 (148.08), calcd., found: C, 8.0, 8.3; H, 2.7, 3.2; N, 57.3, 57.2 %.
m.p. 173C (decomp.), Buechi B-540 melting point apparatus, uncorrected.
DSC: s. Fig. 3.
MS (FAB+) m/z, %: 86 (100) CH4N5+.
1
H NMR (DMSO-d6, TMS): 11.5 ppm.
13
C NMR (DMSO-d6, TMS): 152.3 ppm.
15
N NMR (DMSO-d6, MeNO2): -10.0 (NO3-), -24.5 (N), -165.2 (N-H), -329.1 (NH2).
Raman (100 mW, 90, 1064 nm):
1489 (2), 1465 (1), 1400 (1), 1340 (2), 1091 (1), 1057 (3), 1042 (10), 745 (6),
730 (1), 400 (2), 152 (2).
2.3
X-Ray structure determination
The X-ray diffraction study for 5-aminotetrazolium nitrate, [CH4N5]+[NO3]-, was carried
out at 200 K using a Nonius Kappa CCD instrument, graphite monochromated Mo-K
radiation ( = 0.071073 ), SHELX97 program, direct methods, least squares refinement
(F2) [34].
2.4
Bomb Calorimetry
For all calorimetric measurements a Parr 1356 bomb calorimeter (static jacket) equipped
with a Parr 207A oxygen bomb for the combustion of highly energetic materials was
used [30]. The samples (ca. 100 mg) were loaded in (energetically) calibrated Parr gelatin
capsules (0.9 mL). A Parr 1755 printer was furnished with the 1356 calorimeter to produce a
permanent record of all activity within the calorimeter. The experimentally determined
values, five averaged measurements each, are:
Hcomb.(CH4N6O3)
2.5
Sensitivity Data
Drophammer test [31, 32]:

Friction test
2.6
= 6020 200 cal g-1.
[32]
> 30 Nm
(BAM method, 6 tests, 0 explosions @ 30 Nm)
> 360 N
(BAM method, 6 tests, 0 explosions @ 360 N)
Ab initio Calculations
All computations were carried out using the program package Gaussian-98 [13]. The
stability of the HF (Hartree-Fock) wave function was checked using the stable keyword
option. All final calculations were done at the electron correlated MP2 (Mller-Plesset) level
501
of theory using Dunnings correlation consistent (cc), polarized (p) valence triple-zeta (VTZ)
basis set (MP2/cc-pVTZ) [12,13].
3.

3.1
X-ray Structure
5-aminotetrazolium nitrate was prepared in high yield and purity from the reaction of 5aminotetrazole and nitric acid (eq. 1, see experimental).
CH3N5
HNO3
[CH4N5]+[NO3]-
(1)
After recrystallization from water, crystals of 5-aminotetrazolium nitrate suitable for a

single crystal X-ray structure determination were obtained. The molecular structure of 5aminotetrazolium nitrate in the solid state consists of discrete 5-aminotetrazolium cations
and nitrate anions (Fig. 1, Tab. 1). The CH4N5+ cations and NO3- anions are linked by
reasonably strong hydrogen bonds: O1H10-N2 (O1-N2 2.85 ), O1H9-N3 (O1-N3 2.85
), O2H7-N4 (O2-N4 3.2 ), O3H8-N4 (O3-N4 3.2 ). The reasonably strong
hydrogen bonds may also account for the relatively high thermal stability of 5aminotetrazolium nitrate (see below).
Table 1. Crystallographic data for 5-aminotetrazolium nitrate.
5-aminotetrazolium nitrate
Formula
CH4N6O3
Formula weight [g mol-1]
148.08
Crystal system
monoclinic
Space group
P21/c
a [pm]
10.5493(8)
b [pm]
3.4556(4)
c [pm]
14.606(1)
[]
90
[]
90.548(9)
[]
90
VUC [nm3] a
532.44(8)
Za
[g cm-3]
1.847
R1 [F > 2(F)]
0.0339
wR2 (all data)
0.0901
VUC: volume of unit cell; Z: number of molecules in unit cell.
502
Fig 1.
Molecular structure of 5-aminotetrazolium nitrate in the crystalline state

(ORTEP plot, 50% probability).
Table 2. Selected structural parameters for 5-aminotetrazolium nitrate.a

d/
</
C1-N4
1.308(2)
N4-C1-N2
127.8(1)
C1-N2
1.334(2)
N4-C1-N3
127.8(1)
C1-N3
1.342(2)
N2-C1-N3
104.4(1)
N2-N5
1.366(2)
C1-N2-N5
109.9(1)
N5-N6
1.268(2)
N6-N5-N2
108.0(1)
N6-N3
1.363(2)
N5-N6-N3
108.2(1)
N-O2
1.238(1)
C1-N3-N6
109.7(1)
N-O3
1.240(1)
O2-N-O3
122.3(1)
N-O1
1.277(1)
O2-N-O1
119.1(1)
O3-N-O1
118.6(1)
Within the standard deviation, all torsion angles are 180 or 0, respectively.
503
3.2
Spectroscopy
Due to fast proton exchange in DMSO solution, 5-aminotetrazolium nitrate shows only
one resonance in the 1H NMR spectrum at 11.5 ppm. The 13C NMR spectrum shows a
single resonance at 152.3 ppm in the expected region for sp2 hybridized carbon. In order to
avoid quadrupolar broadening in the 14N NMR spectrum, a 15N NMR spectrum was recorded
(natural abundance) in DMSO solution and is shown in Fig. 2. Four well resolved 15N
resonances were observed and were be assigned as follows: -10.0 ppm (NO3- anion), -24.5
ppm (N5, N6, cf. Fig. 1), -165.2 ppm (N2, N3, cf. Fig. 1), -329.1 ppm (N4, cf. Fig. 1).
-24.5
-329.1
-10.0
-165.2
100.0
Fig 2.
15
-100.0
-200.0
(ppm)
-300.0
-400.0
N NMR spectrum of 5-aminotetrazolium nitrate.
In addition to the peaks of the 5-aminotetrazolium cation the Raman spectrum of

[CH4N5]+[NO3]- shows and is dominated by the very strong 1 vibration of the NO3- anion
which appears at 1042 cm-1. In addition, the vibrations at 1400 cm-1 (3) and 730 cm-1 (4)
can be assigned to the nitrate anion (see experimental).
3.3
Differential Scanning Calorimetry (DSC)
The DSC plot of 5-aminotetrazolium nitrate is shown in Figure 3. With a heating rate of
2C min-1 exothermic decomposition is detectable above 167C with H = -3085 J g-1. The
melting point of 5-aminotetrazolium nitrate was independently determined to be 173C
(decomposition, see experimental).
504
Heat flow endo down (mW)

169C
-20
-16
-12
-8
-4
0
150
130
Fig 3.
3.4
190
170
Temperature (C)
210
DSC plot of 5-aminotetrazolium nitrate, heating rate 2C/min.
Thermodynamics
In order to determine the enthalpy of formation of 5-aminotetrazolium nitrate a BornHaber energy cycle [12] as shown in Scheme 1 was used to calculate the enthalpy of reaction
eq. (2).
[CH4N5]+[NO3]- (s)
HNO3 (g)
[CH4N5]+[NO3]- (s)
CH3N5 (g)
(2)
HNO3 (g) + CH3N5 (g)
HL
CH4N5+ (g) + NO3- (g)

PA(CH3N5)
PA(NO3 )
CH3N5 (g) + H+ (g)
Fig 4.
Born-Haber energy cycle for reaction (2);proton affinity

of CH3N5 (5-aminotetrazole) = - 893 kJ mol-1 (MP2) [13], proton affinity
of NO3- = -1358 kJ mol-1 [14] (all data with zero point energy correction and
corrections for the work term, and translational and rotational corrections)
[15]
.
The lattice energy (UL) for 5-aminotetrazolium nitrate was calculated from the molecular
volume (VM = VUC / Z = 0.133 nm3, see Tab. 1) using Jenkins equation (eq. 3) [16] where
(kJ mol-1 nm) and (kJ mol-1) are coefficients of best fit, z+ and z- are the respective charges
505
on the cations and anions, is the number of ions per molecule, and VM (nm3) is the
molecular (formula unit) volume of the salt [13].
UL = |z+| |z-| {( / VM 0.33) + }
(3)
For 1:1 salts: = 117.3 kJ mol-1 nm, = 51.9 kJ mol-1.

UL(5-aminotetrazolium nitrate) = + 563 kJ mol-1.
The corresponding lattice enthalpy, HL for 5-aminotetrazolium nitrate was derived
from UL using eq. (4) [17], where nM and nX depend on the nature of the ions, Mq+ and Xq- of
the salt MpXq, and are equal to 3 for monoatomic, 5 for linear polyatomic ions, and 6 for
non-linear polyatomic ions.
HL = UL + {p[(nM/2) -2)] + q[(nX/2) - 2]} RT
(4)
HL(5-aminotetrazolium nitrate) = + 568 kJ mol-1.

With the calculated enthalpy of reaction (2) of H(2) = + 103 kJ mol-1 (cf. Scheme 1)
and the experimentally known enthalpies of formation of HNO3 (g) (Hf, HNO3 = -134 kJ
mol-1) [18,19] and 5-aminotetrazole (g) (Hf, CH3N5 = +323.8 kJ mol-1) [19-21], it was now
possible to calculate the standard enthalpy of formation of 5-aminotetrazolium nitrate:
Hf([CH4N5]+[NO3]- (s)) = + 87 kJ mol-1 = + 586 kJ kg-1.
Using the so obtained value for the enthalpy of formation of 5-aminotetrazolium nitrate
and the literature values for the enthalpy of formation of H2O (l) (Hf, H2O (l) = -285 kJ
mol-1) [18,22] and CO2 (Hf, CO2 (g) = -394 kJ mol-1) [18,19,22] it was now possible to calculate
the enthalpy of the (oxygen) combustion of [CH4N5]+[NO3]- according to eq. (5).
[CH4N5]+[NO3]- (s) + 1/2 O2 (g) CO2 (g) + 2 H2O (l) + 3 N2 (g)
(5)
H(5) = -1051 kJ mol-1 = -7093 kJ kg-1.

This value is in reasonable agreement with the experimentally determined value (this
work, see experimental) of Hcomb.( [CH4N5]+[NO3]-) = 6020 200 kJ kg-1.
5-Aminotetrazolium nitrate, [CH4N5]+[NO3]-, is a compound with only a slightly
negative oxygen balance [23] of 10.8%.
In order to assess more quantitatively the expected detonation properties of 5aminotetrazolium nitrate, [CH4N5]+[NO3]-, we calculated the expected detonation pressure
(P) and detonation velocity (D) using the semi-empirical equations suggested by Kamlet and
Jacobs (eq. 6 and 7) [24-28].
P [108 Pa] = K 2
(6)
with: K = 15.58; density in g cm-3

= 0.49 N M0.5 Q0.5 with : N = moles of gas per g of explosive;
M = g of gas per mol of gas (average
molar mass of formed gases);
Q = estimated or guessed heat of
detonation (cf. eq. 8) (in kJ kg-1)
D [mm s-1] = A 0.5 (1 + B )
(7)
with: A = 1.01; B = 1.30
506
To estimate the enthalpy of detonation, the Kistiakowsky-Wilson rules [29] were applied
in order to predict the detonation products in equation (8).
[CH4N5]+[NO3]- (s)
CO (g)
2 H2O (g)
3 N2 (g)
(8)
With the experimentally known enthalpy of formation of H2O (g) (Hf, H2O = -242 kJ
mol-1) [12,13] and CO (g) (Hf, CO = -111 kJ mol-1) [11-13] it was possible to calculate the
enthalpy of detonation Q for 5-aminotetrazolium nitrate:
HR(8) = Q([CH4N5]+[NO3]-) = - 682 kJ mol-1 = - 4603 kJ kg-1.
For 5-aminotetrazolium nitrate we calculated a detonation pressure of P = 35.7 GPa and
a detonation velocity of 8.90 mm s-1 (8898 ms-1), which compares nicely with RDX (8750
ms-1) and HMX (9100 ms-1) at their highest densities [23].
4.
CONCLUSIONS
From this combined experimental and theoretical study the following conclusions can be
drawn:
(i)
(ii)
(iii)
(iv)
5-Aminotetrazolium nitrate, [CH4N5]+[NO3]-, is a powerful and promising

explosive with a good oxygen balance (-10.8%) and low sensitivity.
5-Aminotetrazolium nitrate, [CH4N5]+[NO3]-, can easily be prepared in a one step
synthesis and is expected to have sufficient thermal stability for application.
The crystal structure of 5-aminotetrazolium nitrate, [CH4N5]+[NO3]-, was
determined by single crystal X-ray diffraction.
5-Aminotetrazolium nitrate, [CH4N5]+[NO3]-, shows estimated detonation
velocities and detonation pressures similar to high explosives such as PETN,
RDX or HMX.
Acknowledgments:
Financial support of this work by the University of Munich (LMU), the Fonds der
Chemischen Industrie and the German Federal Office of Defense Technology and
Procurement (BWB, WIWEB) is gratefully acknowledged. The authors are also indebted to
and like to thank PD Dr. Konstantin Karaghiosoff for recording the NMR spectra and Dipl.Chem. Jan Weigand for the DSC measurements. We also like to thank Prof. Dr. Peter
Klfers for generous allocation of X-ray spectrometer time.
507
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
W.-H. WALTHER, T. M. KLAPTKE, AND G. HOLL: Physical Chemical Characterization and

Explosive Properties of Hydrazine Azide, in: Proceedings 29th International Annual Conference
of ICT, June 30 July 3, 1998, Energetic Materials Production, Processing and
Characterization, Fraunhofer Institut fr Chemische Technologie: Karlsruhe, Germany, 1998,
pages 134-1 to 134-1.
T. M. KLAPTKE, B. KRUMM, G. HOLL, AND M. KAISER: Synthesis, Characterization and
Quantum Chemical Computations of Hexanitrohexaaza-Isowurtzitane, in: Proceedings 30th
International Annual Conference of ICT, June 30 July 3, 1999, Energetic Materials Modeling of Phenomena, Experimental Characterization, Environmental Engineering,
Fraunhofer Institut fr Chemische Technologie: Karlsruhe, Germany, 1999.
T. HABEREDER, A. HAMMERL, G. HOLL, T. M. KLAPTKE, J. KNIZEK, P. MAYER, AND
H. NTH: Methylsubstituted Hyrdazinium Azide Derivatives, in: Proceedings 31st International
Annual Conference of ICT, June 27 30, 2000, Energetic Materials
Analysis,
Diagnostics and Testing, Fraunhofer Institut fr Chemische Technologie: Karlsruhe, Germany,
2000,
pages 150-1 to 150-7.
T. M. KLAPTKE: New highly energetic materials based on hydrazine and tetrazole
compounds, ACS Spring 2003 Award Symposium, New Orleans, March 23 27, 2003.
J. GEITH, T. M. KLAPTKE, J. WEIGAND, G. HOLL: Calculation of the Detonation Velocities
and Detonation Pressures of Dinitrobiuret (DNB) and Diaminotetrazole Nitrate (HDAT-NO3),
Propellants Explosives and Pyrotechnics 2004, 29, 3 - 8.
W. S. EWAN, M. W. RIGG: The heats of combustion of compounds containing the tetrazole
ring, J. Chem. Soc. 1951, 4725-27.
K. BRITTS, I. L. KARLE: The crystal and molecular structure of 5-aminotetrazole monohydrate,
Acta Cryst. 1967, 22, 308-312.
D. D. BRAY: Refinement of the crystal structure of 5-aminotetrazole monohydrate, Acta Cryst.
1979, 35, 3089-3091.
J. H. BYRDEN: The structure of the hydrazine salt of 5-aminotetrazole, Acta Cryst. 1958, 11,
31-37.
V. SINDITSKY, A. E. FOGELZANG: Combustion of 5-aminotetrazole salts, Proc, Intl. Pyrotech.
Seminar 1995, 21, 762-773.
O. A. LUKYANOV, V. A. TARTAKOVSKY: Synthesis and characterization of dinitraminic acid
and its salts, Prog. Astronaut. Aeronaut. 2000, 185, 207-220.
(a) T. M. KLAPTKE, I. C. TORNIEPORTH-OETTING: in Nichtmetallchemie, VCH, Weinheim,
1994, chapter 2.1.2, pp 96 103;
(b) T. M. KLAPTKE, A. SCHULZ, R. D. HARCOURT: Quantum Chemical Methods in MainGroup Chemistry; Wiley, Chichester, Chapter 5, 1998.
Gaussian 98, Revision A.3, M. J. FRISCH, G. W. TRUCKS, H. B. SCHLEGEL, G. E. SCUSERIA,
M. A. ROBB, J. R. CHEESEMAN, V. G. ZAKRZEWSKI, J. A. MONTGOMERY, JR., R. E.
STRATMANN, J. C. BURANT, S. DAPPRICH, J. M. MILLAM, A. D. DANIELS, K. N. KUDIN, M. C.
STRAIN, O. FARKAS, J. TOMASI, V. BARONE, M. COSSI, R. CAMMI, B. MENNUCCI, C.
POMELLI, C. ADAMO, S. CLIFFORD, J. OCHTERSKI, G. A. PETERSSON, P. Y. AYALA, Q. CUI,
K. MOROKUMA, D. K. MALICK, A. D. RABUCK, K. RAGHAVACHARI, J. B. FORESMAN, J.
CIOSLOWSKI, J. V. ORTIZ, B. B. STEFANOV, G. LIU, A. LIASHENKO, P. PISKORZ, I.
KOMAROMI, R. GOMPERTS, R. L. MARTIN, D. J. FOX, T. KEITH, M. A. AL-LAHAM, C. Y.
PENG, A. NANAYAKKARA, C. GONZALEZ, M. CHALLACOMBE, P. M. W. GILL, B. JOHNSON,
W. CHEN, M. W. WONG, J. L. ANDRES, C. GONZALEZ, M. HEAD-GORDON, E. S. REPLOGLE,
AND J. A. POPLE, Gaussian, Inc., Pittsburgh PA (1998); and corresponding references given in
theGaussian 98 users reference handbook.
508
[14] J. A. DAVIDSON, F. C. FEHSENFELD, C. J. HOWARD: The heats of formation of NO3- and NO3association complexes with HNO3 and HBr, Int. J. Chem. Kinet., 1977, 9, 17.
[15] T. M. KLAPTKE, A. SCHULZ, R. D. HARCOURT: Quantum Chemical Methods in Main Group
Chemistry, Wiley, Chichester, 1998.
[16] H. D. B. JENKINS, H. K. ROOBOTTOM, J. PASSMORE, L. GLASSER: Relationships among Ionic
Lattice Energies, Molecular (Formula Unit) Volumes, and Thermochemical Radii, Inorg.
Chem., 1999, 38, 3609 3620.
[17] H. D. B. Jenkins, D. Tudela, L. Glasser: Lattice Potential Energy Estimation for Complex Ionic
Salts from Density Measurements, Inorg. Chem., 2002, 41, 2364 2367.
[18] M. W. CHASE, JR.: NIST-JANAF Themochemical Tables, 4th edn., J. Phys. Chem. Ref. Data,
Monograph 9, 1998, 1-1951.
[19] NIST Chemistry WebBook, NIST Standard Reference Database Number 69 - March, 2003
Release, www version: http://webbook.nist.gov/chemistry/.
[20] A. A. KOZYRO, V. V. SIMIRSKII, A. P. KRASULIN, V. M. SEVRUK, G. YA KABO, M. L.
FRENKEL, P. N. GAPONIK, YU V. GRIGOREV: Thermodynamic properties of tetrazole
derivatives in different aggregation states, Russ. J. Phys. Chem. (Engl. Transl.), 1990, 64, 348350.
[21] Hsub.( CH3N5) = 116.0 kJ mol-1, Hf( CH3N5, s) = +207.8 kJ mol-1 [19,20]
[22] J. D. COX, D. D. WAGMAN, V. A. MEDVEDEV: CODATA Key Values for Thermodynamics,
Hemisphere Publishing Corp., New York, 1984, 1.
[23] J. KHLER AND R. MEYER: Explosivstoffe, 9th ed.; VCH: Weinheim, 1998.
[24] M. J. KAMLET AND S. J. JACOBS: Chemistry of detonations. I. Simple method for calculating
detonation properties of carbon-hydrogen-nitrogen-oxygen explosives, J. Chem. Phys., 1968,
48, 23 - 35
[25] M. J. KAMLET AND J. E. ABLARD: Chemistry of detonations. II. Buffered equilibrium, J.
Chem. Phys., 1968, 48, 36 - 42
[26] M. J. KAMLET AND C. DICKINSON: Chemistry of detonations. III. Evaluation of the simplified
calculational method for Chapman-Jouguet detonation pressures on the basis of avilable
experimental information, J. Chem. Phys., 1968, 48, 43 - 50.
[27] L. T. EREMENKO AND D. A. NESTERENKO: Energetics of the decomposition of
polynitrocubanes (analytical prediction), Chem. Phys. Reports, 1997, 16, 1675.
[28] A. M. ASTAKHOV, R. S. STEPANOV, AND A. Y. BABUSHKIN: Detonation parameters of
octanitrocubane, Combust. Explos. Shock Waves, (Engl. Transl.), 1998, 34, 85.
[29] A. A. KOZYRO, M. L. FRENKEL, A. P. KRASULIN, V. V. SIMIRSKII, AND G. YA. KABO:
Thermodynamic properties of biuret in different states of aggregation, Russ. J. Phys. Chem.
(Engl. Transl.), 1988, 62, 897-899.
[30] http://www.parrinst.com/.
[31] T. M. KLAPTKE, C. M. RIENCKER: Drophammer Test Investigations on Some Inorganic and
Organic Azides, Propellants Explosives and Pyrotechnics 2001, 26, 43 - 47.
[32] REICHEL & PARTNER GMBH, http://www.reichel-partner.de/.
[33] T. M. KLAPTKE, G. SPIE, unpublished.
[34] G. M. SHELDRICK: SHELXL97, Program for Crystal Structure Refinement, Universitt
Gttingen, 1997.
509
MEASUREMENT IN BLAST HOLE STEM AND INFLUENCE

OF STEMMING MATERIAL ON BLASTING QUALITY
Mario Dobrilovi, Zvonimir Ester, and Branimir Jankovi
Faculty of Mining, Geology and Petroleum Engineering,
University of Zagreb, Croatia
Abstract:
Paper presents results of blast hole stemm materials, that were conducted to assert best
stemm materials for surface blasting in quarry of technical / construction stone.
Blasting has been performed with equal explosives and test-minefield parameters
(length of the stem, volume of the explosive charge, mine-drill depth, and angle) on
various sites / quarry. Results are guidelines in materials to be chosen in surface
blasting of quarry works, adding to quality of mining works and reduction of costs.
Keywords:
1.
blasting, civil-usage explosives, shock-wave, stem materials miniranje
INTRODUCTION
Blasting is a technological process used in mining for acquiring of mineral rawmaterials, and for profiling of underground structures and tunnels. Blasting of explosive
charge in blasting hole creates significant amount of energy released, shock-wave of great
force in explosive and surrounding rock-formation and gases of high temperature and
pressure. Blasting is considered successful when gas-energy produced by blasting is held in
blast-hole long enough to create web of cracks in rock-formation and discharge of materials.
Also, amount of energy needs to be sufficient to surpass strength of rock-formation and
assures controlled movement of materials. If blasting parameters meet requirement, majority
of energy is used to crush and discharge materials, while inadequate parameters result in
surplus oscillation of ground and air (shock-wave in ground and air) and greater discharge of
material. Blast energy reduction is a key factor of control for size and nature of oscillation
produced. Test-blasts are required to successfully determine amount of explosive charge,
stem size and drill parameters (diameter of the hole, placing of the holes and size of
discharge). For understanding of the stem and its role in the process, underground or surface,
knowledge of the processes in hole during blasting is required. Parameters involved are:
amount and type, blasting-technical characteristics of explosives, detonation method, and
geometry of the blast-field and influence of blasting to rock-formation. For optimal usage of
explosive energy and reduction of unnecessary discharge, blast-holes need to be stemmed
with adequate inert materials, from explosive charge to mouth of the hole.
Underground blasting, where distance between holes is lesser, where rock-formation
levels and is under influence of previous blasting shock-waves, stem holds explosive charge
in holes. Experience, both lab and field, proves that optimal stem increases efficiency of
blasting up to even 300%. Stem usage in underground construction and tunnelling by
blasting is considered hindering factor in speed of process and is mostly avoid in praxis, but
testing has proven that 300% increase in efficiency of usage of energy, so that delays in
explosives charge production are compensated by savings in explosives and possible
reduction in blasting cycles. Stems are characterised by; type of blasting (surface,
underground, special blasting), shape (patronized, powder), material (discharged material,
gravel, crushed rock of different fractions, dirt, clay, water, water-based gels, air, mineral
510
clay, polyurethane foam, etc), method of stemming (hand or machine). For surface blasting,
stems used most are discharged material, grovel, crushed rock of different classes.
2.
TEST OF VARIOUS GRANULATED STEMMS

UNDER EQUAL BLASTING CONDITIONS
2.1
Stem Material
Optimal stem type test were conducted in two technical / construction rock quarrys
(Ivanec and Oura) in test-fields of 5 blast-holes. Classes used for determination of
optimal fraction were:
-
Discharged blast-holes material, classes of: 4; 4/8; 8/16; 16/32.
Class 16/32 material is shown on figure 1. In table 1 are densities of used classes.
Fig 1.
16/32 Class Material
Table 1. Density of classs used in stemming

Class (mm)
Sample volume (dm3)
Sample Mass (kg)
Density (kg/ dm3)
2.2
Discharged
Material
1
1.766
1.766
-4
4/8
8/16
16/32
1
1.716
1.716
1
1.609
1.609
1
1.430
1.430
1
1.383
1.383
Blasting-technical qualities of explosive used in test-blasting
Ivanec quarry test-blasts were conducted with emulsion explosive EMUNIT,

manufacturer Elmech Razvoj, for both primary and secondary charges.
An Oura quarry blast was with ELMULEXAL explosive of Elmech Razvoj, and
for secondary charge powder ANFO explosive was used.
Emunit is emulsion explosive of secondary heavy-ANFO explosive type and is
sensitized with glass micro spheres and ammonium nitrate granules to produce hot-spot. Its
purpose is in civil usage in surface and underground blasting of methane and coal-dust free
conditions and in wet blasting due to its water-resistance. It is white, pasty material, packed
in PVC patroness of 70 mm diameter / 2.600 g weight.
511
ANFO explosives are mixture of nitrates and fuel. Basic material for its production is
ammonium nitrate with addition of oils, especially mineral oils, with or without aluminium
powder to increase energy and efficiency of explosive. ANFO explosive is used in
combination with emulsion, plastic and other stronger explosives.
Elmulexal is emulsion explosive of secondary type. Aluminium powder is on of its
component for increase of energy and efficiency, and is used for blasting of hardest rock, for
special construction and underwater works. Comes in patroness of 60 mm diameter / 1.714 g
weight. Characteristics (declared) of explosives are given in Tab. 2.
Table 2. , Characteristics (declared) of explosives used
Explosive
Velocity of detonation (m/s)
Density (g/cm3)
Blast Temperature(C)
toplina eksplozije (kJ/kg)
Gas Volume (l/kg)
Oxygen Bilance (%)
2.3
ANFO
2700
0,882
2190
3600
1050
1,5
Emunit
4862
1.201
2050
3650
920
-1.5
Elmulexal
5200
1.15
2057
3120
910
0.7
Test I Ivanec quarry
Velocity of detonation was measured in 5 blast-holes. Stem material changed, other

parameters equal. Blasting Parameters:
Drill Angle
Drill Diameter
Drill Depth
Stem Depth
Distance between Holes
Burden
70,
86 mm,
13,5 m,
3,8 m,
3 m,
2,7 m.
Initiation system non-electric detonator Crodet with 500 ms retarder. Firing retarder
for each hole 25 ms. Drill depth, stem length and number of cartridge (charging mass)
were parameters monitored. Each hole was stemmed with different material.
Geometry of charging is given in table 3. On figure 2 is schematic view of blasting field.
Figure 3 presents charge of blast hole.
Table 3.
Hole
No.
1
2
3
4
5
Drill parameters and explosive charge in Ivanec quarry
Hole Depth
(m)
13,5
13,5
13,5
13,6
13,2
Stem Length (m)
Stem Class (mm)
No.of cartridge
3,8
3,8
3,4
3,8
3,6
4
4/8
8/16
16/32
Discharged Material
26
25
26
27
24
512
Fig 2.
1.
2.
3.
4.
5.
Blasting Field Schematic
Stem
EMUNIT explosive charge
70/2600
Non-electric Detonator
VOD Measurement Cable
Shock Tube
Fig 3.
2.4
Ivanec quary, Blast Hole Schematics
Oura Quarry Test II
Detonation speed was measured in 5 blast-holes. Stem material was different, while
other parameters stayed same. Elmulexal 60/1714 g explosive was used as primer
cartridge for ANFO explosive making rest of explosive charge.
Drill Parameters:
Angle
Diameter
Depth
Stem length
Distance between holes
Burden
70,
90 mm,
24 m,
4,5 m,
5 m,
5 m.
513
Non-electric detonator Crodet with 500 ms retarder was initialising element.

Detonation retardation was 42 ms. Blast-hole length, stem length and ANFO explosives
volume were parameters monitored. Holes were stemmed with different material.
Charging data are given in table 4. On figure 2 is schematic view of blasting field.
Figure 5 presents charge of blast hole.
Table 4.
Holel size and explosive charge volume in Oura quarry
Hole
No.
Hole
Depth
(m)
Blast hole explosive

charge
1
2
3
4
24
24
24
24
24kgA+UP+95kgA
24kgA+UP+90kgA
24kgA+UP+95kgA
24kgA+UP+100kgA
Measured
Hole
Length
(m)
21
21
21
21
24
24kgA+UP+100kgA
21
Stem
Length
(m)
4,5
5
4,5
4,5
4,5
Stem Class
(mm)
-4
4/8
8/16
16/32
Discharged
material
1p one cartridge
25kgA 25 kg of ANFO explosive
UP primer cartridge
2.5
Velocity of detonation measurement in blasting holes
VOD measurement is crucial parameter, based on which we interpret retardation

(decrease in speed) of shock-wave in specific stem. Stem with highest retardation has best
energy retention and discharged material prevention. Detonation speed is measured with
VODMate (VOD velocity of detonation), Instantel Inc. Adequate measurement cable,
equipment and connector were used with VODMate instrument.
VODMate is measuring speed continuously, giving real-time detonation speed data for
explosive used and helping to optimise blast effect. Computer software calculates data
received and interprets achieved detonation speed and gives boundaries between explosives
in hole. Instrument also monitors shock-wave speed in stem. Fig. 5, Measurement Schematic
in Blast-hole.
514
1.
2.
3.
4.
5.
6.
ANFO
neelektrini detonator
udarna patrona
ELMULEXAL- a
ep
kabel za mjerenje brzine
detonacije
udarna cjevica neelektrinog
sistema iniciranja
Fig 4.
Oura quary, Blast Hole Schematics
Measurement Cable
Coaxial
Cable
BNC
Connecto
VODMate
Attachment Cable
Fig 5.
VODMate Measurement Schematic
VODMate operates on measurement of change in electric resistance in measurement

cable, where constant power supply is secured. During detonation, explosives detonation
front advances and shortens (melts) measurement cable. Based on known specific electric
resistance per m of cable, instrument measures total electric resistance in time frame and
calculate cable length. Detonation speed is calculation of measured values.
Upon selection of charges in test-holes, measuring equipment is selected. Primer cartrige
is equipped with coaxial measurement cable. Primer cartridge is lowered to the bottom of
hole, and is charged with pre-determined amount of explosives. Uncut measurement cable is
lowered in all test-holes. Primer cartridge is placed on the bottom of bore hole.
515
Fig 6.
2.6
Primer cartridge
Rezultati mjerenja brzina detonacija u minskim buotinama
Table 5 gives velocity of detonation for specific holes. Table 6 presents measured
Shock-wave velocity in stems.
Computing, with Blast Ware III software, of VOD diagram reveals data open for
interpretation, and can be computed upon detonation speed in blasting-hole, hole depth,
explosive charge length or stem length in hole. Correct ascertain of start and finish of
detonation front influence, shock-wave area in hole and in stem, and middle value of
velocity of detonation in specific interval.
Small change in angle of axis or dependence of measurement cable length per time,
speed value can change significantly. Following general axis angle, differences in measured
intervals are in 100 m/s range, or app. 2,5% error in detonation speed. Detonation speed
diagrams are analysed and median value determined according to axis angle.
Figure 7 presents VOD measurement characteristic diagrame and figure 8, Segment of
measurement for hole 1, Test I, Stem material class-4
Table 5. VOD of explosive used
Detonation Speed in Holes (m/s)
Explosive type
Declared Detonation
Speed (m/s)
EMUNIT
4725
4815
4739
4567
4719
4862
ANFO
3642
3590
3660
3705
3528
2700
ELMULEXAL
Used as primer cartridge
516
5200
Table 6. Velocity of shock wave in stems

Stem Type (mm)
1
2
3
4
5
-4
4/8
8/16
16/32
Discharged Material
Velocity (m/s)
Measurement I
665
966
1663
476
1331
Measurement II
433
634
1218
393
923
Lenght (m)
Hole No.
Time (ms)
VOD measurement diagram
Lenght (m)
Fig 7.
Time (ms)
Fig 8.
Hole no. 1, Time I, Stem class -4
517
2.7
Blasting Video Analysis
Tape analysed is of FUJI S500 digital camera, 320x240 resolution, and 33 frames7sec.
In VirtualDub 1.4.8 software, drill positions are determined inside blast-field with stem
discharge height visible for specific holes. Hole no. 3, stemmed with granulation 8/16, has
highest discharge, confirming detonation speed interpretation that this particular granulation
presents lowest stemming quality. Next highest discharge / lowest quality are 16/32. This
granulation was asserted as best for stemming using previous interpretation of detonation
speed results. Location of holes in blasting-field are given on figure 9.
Fig 9.
Holes location
Analysis of blasting in Oura quarry shows that no stem discharge occurred. Reason for it is
use of non-electric initiation system that doesnt destroy stem, and stem over-dimensioning. Stem
material (fraction) is relevant with border-line stem length, as seen in Ivanec quarry. Further
inspection of blasting field revealed that part of material surrounding stem was also not discharged.
Figure 10, hole stemmed with 16/32 fraction material left in hole after blasting.
Fig 10.
Hole stemmed with 16/32 fraction
518
3.
CONCLUSION
Video-tape interpretation of Ivanec quarry blasting revealed that 8/16 and 16/32
materials used for stemming were discharged from hole while stayed intact. Oura quarry
test produced no discharge in both tests.
Detonation speed measurements in hole gives conclusion that best suited material for
stemming in technical / construction quarries is 16/32 fraction, which produced lowest
shock-wave speed inside the stem. Usage of this fraction allows reduction of stem length
without loss of explosive energy. This results in increase in explosive charge and
geometrical parameters of mining, or lesser no. of holes for same amount of discharged
material.
8/16 proved to be least favourite stem material, with shock-wave speed of over 1.200
m/s. Material most frequently used is discharged material.
Shock-wave speed is dependant also upon explosives type. In Ivanec quarry, Emunit
explosive was used, producing greater detonation speed than ANFO explosive used in
Oura quarry. Greater detonation speed in explosive charge results in greater shock-wave
speed in stem. So Oura quarry measurements revealed for separate fractions lesser shockwave speed in stem.
Stem type depends on needs or conditions in which blasting is performed and of
fractioned material disponibility and desired blasting effects. 16/32 fraction is recommended
in all cases, dependant only to supply economics.
REFERENCES
[1]
[2]
[3]
[4]
[5]
ESTER Z. (2001.): Miniranje I, eksplozivne tvari, svojstva i metode ispitivanja,

Rudarsko-geoloko-naftni fakultet, Interna skripta, Zagreb, 145
E. I. DU PONT DE NEMOURS & COMPANY (INC.) (1952): Blasters' Handbook, sesquicentennial Edition,
Wilmineton 98, Delaware, 477
INSTANTEL INC. (1998-1999.): VODMate Operator Manual, Canada
KRSNIK J. (1989.): Miniranje, Rudarsko-geoloko-naftni fakultet, Zagreb, 180
ESTER Z. (1996): Utjecaj zaloma na rezultate miniranja podzemnih prostorija, Disertacija,
Rudarsko-geoloko- naftni fakultet, Zagreb, 83,8
519
MEASUREMENT OF SHOCK WAVE FORCE IN SHOCK TUBE

WITH INDIRECT METHODS
Mario Dobrilovi, Zvonimir Ester, and Trpimir Kujundi
Faculty of Mining, Geology and Petroleum Engineering,
Abstract:
Explosives Lab of Mining Geological & Oil Faculty in Zagreb University is performing testing
with aim of design of detonator that would unify advantages of non-electric system and precision
in regulating interval in electric system. Sum of energy released by wave force in shock tube is a
condition for work of this new detonator, and measurement of wave force is first step in
determining of sum of energy. Sum of energy is measured indirectly, on two principles:
movement sensors and strain gauges sensors.
Keywords:
1.
civil explosive, initiation, retardation, shock tube, impact energy,

shock wave force, blasting
INTRODUCTION
Non-electric system of firing by civil blasting has a number of advantages toward other
initiation methods. Those are: resistance to static electricity and currents, simple handling,
connections overview, resistance to humidity, greater mechanical durability and volume of the
tube, no limits to number of blasting drills (degrees of firing) in mining field, doesn't destroy
blasting drill cork, no air shock wave is produced, multiple time-differentiated detonations,
preserved structure of emulsion and ANFO explosives. One of the disadvantages is significant
difference in delay time for greater retardations (5000 6000 m/s); used more frequently in
construction of tunnels, underground objects and ravines. For contour blasting, time
synchronization is paramount for shape of contour blast. Retardation time is determined by
retarding-element of the detonator, pyrotechnic mix with precise time consumption, which is
undeterminable for total time greater than 5000 m/s. This element requires new retarding
element that can solve this problem, and was found in new system (detonator) that combines
advantages of non-electric (simplicity, speed, and cost) and electric regulation of retardation.
Phase 1 is measurement of shock wave force in shock tube.
2.
MEASUREMENT OF SHOCK WAVE FORCE IN SHOCK TUBE

2.1
Shock wave velocity measurement
Shock tube used for testing was of: inner diameter 1,2 5(%), and outer diameter 3 +/5(%), octogen-HMX filling 15-17 (mg/m). Shock tube was initiated using Spark blasting
machine jr-1 in all tests.
Shock wave measurement in shock-tube was conducted to determine influence of
velocity to force of the shock wave. Electro-optical method was used for measurements.
Velocity of shock wave is measured with electronic stop-timer and direct measurement of
shock wave travel time in shock-tube. Device is receiving impulses through optical fibre
cables. In measuring, device calculates velocity on basis of time and distance between
probes. Measuring device is electronic stop-timer Explomet-FO-2000 of following
characteristics: max measurable velocity 10.000 m/s, time interval from 0,1 do 10 s,
accuracy +/- 0,1 s. Velocity was measured on meter-segments on 1 to 4 meters of the tube to
determine velocity per length, and its influence on shock wave force. Measured velocities
were in range from 1642 to 1754 m/s. Table 1. present measurements results. Measurement
equipment is shown on figure 1.
Table 1. Velocity of detonation measurements results
Measurement
No.
Fig 1.
2.2
Type of
explosive
Initiating
System
Probe
distance (m)
Time t(s)
Velocity
(m/s)
L 1 =1
577,1
1732
Shock
SPARK
L 2 =1
570,1
1754
tube
BLASTING
L 3 =1
575,5
1737
MACHINE
L 4 =1
581,3
1720
L 1 =1
608,8
1642
Shock
SPARK
L 2 =1
583
1715
tube
BLASTING
L 3 =1
595
1680
MACHINE
L 4 =1
595,5
1679
L 1 =1
571,8
1748
Shock
SPARK
L 2 =1
575,7
1737
tube
BLASTING
L 3 =1
582
1718
MACHINE
L 4 =1
579,8
1724
Velocity of detonation, measurement equipment
Shock-wave force Measurement with LVDT sensors
LVDT sensors work on shift principles. Shift in tin plate membrane during firing of
shock tube is measured. Prior to measurement, measuring apparatus is inflated on
dynamometer on hydraulic press. Equipment and instruments used are: hydraulic press,
dynamometer, LVDT sensor, measuring frame, computer registrator (Spider 8). Measuring
frame consists of: body, tin-plate membrane, transition piston, basis for tube. Tube is fixed
into basis, sitting tightly to transition piston attached to tin-plate membrane. During inflation,
determined force is applied through additional piston to transition piston that elastic deforms
tin-plate membrane upon shock wave measurement. Membrane shifts are measured on
LVDT probes, transferred to frugal curve. Measuring frame is displayed on Fig. 2 and
measuring apparatus on Fig. 3.
Fig 2.
Measuring frame
2.2.1. Measurement Description

Measuring is indirect, by measuring of deformation of circular tin-plate membrane
using LVDT probe, size of deformation is proportional to force deforming.
Measuring Phases are:
a) Calibration of LVDT probes
b) Measurement of shock-wave in shock tube
LVDT probe measures shifts with electro induction method. Calibration of probes with
static force produces correlation of shift (elastic deformation of tin-plate membrane) and
force used. Curve with highest correlation factor R was extracted from specific Calibration
series and was used in further results handling. Calibration was performed till shifts, after
force applied, were repeated, or till Calibration and permanent plastic deformation of tinplate membrane was achieved. Measuring app. schematic is given on fig. 3. Frugal curve and
LVDT diagram are on fig. 4. Max force used for calibration was 50 (N). This value was
maximal toward strength upon static force, greater plastic deformations occurrences and
maximal shift of LVDT probes.
Equipment as for Calibration of LVDT probes is used for shock-wave measurements in
shock-tube, with conditions unchanged from Calibration. This secures accuracy of
measurement. For Calibration, force was applied in very slow pace, almost statically, while
shock-wave force travels at +1.700 m/s.
Shock-tube is fired at segment, where for calibration additional piston was placed.
Additional piston was attached to transition piston of 4,1 mm in diameter applied to tinplate-membrane. Measurement was performed with tubes different in length, to correlate
amount of charge to shock-wave force. Shock-wave reaches the piston, transferring impulse
to membrane and shifting probe. LVDT sensor data are computed in registrar. Correlations
to calibration curve, values of shock-wave force are calculated.
Measured shock-wave force values are given in table 2.
Dynamometer
Additional piston
Transition piston
Tin-plate
LVDT sensor
Hydraulic press
Fig 3.
Measuring apparatus
Table 2. Measured shock-wave force values

Measure no.
LVDT
[mV/V]
9,96048
8,80944
10,26144
12,24432
10,47216
1
2
3
4
5
Force
[N]
190,685
156,790
216,0931
308,5739
225,291
60
2
F (N)
5) y= 1,6878x + 8,6544x - 50,434

2
R = 0,9999
40
10
4) y= 1,73x + 8,0421x - 50,429

2
R = 0,9999
1. Badarenje
2. Badarenje
3. Badarenje
20
0
0
-20
4
6
8
2
3) y= 1,6833x + 7,8516x - 50,413
2
R = 0,9999
2
2) y = 1,7187x + 7,3842x - 50,432
4. Badarenje
5. Badarenje
Poly. (5. Badarenje)
R = 0,9999
-40
Fig 4.
5
0
0
10
15
-5
-10
-15
-20
1) y = 1,5619x + 7,8754x - 50,434

2
R =1
-60
LVDT (mV/V)
15
LVDT (mV/V)
Calibration curve and LVDT diagram
Time (s)
20
25
30
Negative values on LVDT diagram are caused with inertness of probes. Transition shock
impulse travel with speed of shock wave (~1700 m/s), and probes are constructed for static
loading. Because of that measuring system inertness, probe is confused and that value is
negative. That negative value has no practical meaning.
2.3
Shock-wave force measurement with strain gages
Equipment and apparatus used in this measurement are same used in LVDT
measurements. While LVDT measurement observed shift in membrane, this measurement
monitors change in electric resistance. Measurement is indirect, of the deformation of
circular tin-plate membrane using strain gage, where deformation size is proportional to
force that caused it.
Measurement phases are: Calibration of strain gage, measurement of shock-wave in
shock-tube. Measurement frame with sensor is given on figure 5.
Fig 5.
Measurement Frame with Sensor
Measurement equipment is placed same a sin LVDT process. Tin-plate membrane is 1

mm thick, with sensor applied to it 24 h prior to test.
Calibration of strain gage is performed same as Calibration of LVDT probes. Calibration
curve and strain gage diagram is given on figure 6. Middle values of measured force for
different length of shock tube are given in table 3.
Table 3. Middle values of measured force Strain gage
Tube length
[m]
0,5 m
1.0 m
2.0 m
4.0 m
1.0 m, fixed
Force
[N]
71,32
53,63
64,43
208.87
208,29
0,025
120
F (N)
y = 3138,3x - 27,001
100
R = 0,9509
drugo umjeravanje
80
y = 2792,1x - 22,404
2
R = 0,9376
60
y = 2910,5x - 22,417
20
0,015
Linear (prvo umjeravanje)
Series1
0,01
R = 0,9142
Linear (drugo umjeravanje)
0
0
0,01
0,02
0,03
0,04
0,05
-20
y = 3054,4x - 28,828
2
R = 0,9477
(mV/V)
-60
0,005
Linear (tree umjeravanje)

Linear (etvrto
umjeravanje)
0
0
-0,005
Fig 6.
3.
tree umjeravanje
0,02
etvrto umjeravanje
40
-40
(mV/V)
prvo umjeravanje
10
Time (s)
Calibration curve and strain gage diagram
CONCLUSION
Shock-wave force in non-electric shock-tube in this test is determined by: tube length
and velocity, outer diameter of the tube, tube fixation
Tube length is directly influencing speed of shock-wave inside the tube. Greater the
length, greater the speed, and proportionally greater the force. After speed stabilization,
velocity depends of amount of explosive charge used in different segments of the tube.
Increasing of velocity during the first meter of length is 5% and that result with force
increasing upon 20%.
Diagram of correlation of speed of shock-wave and force is constructed in dependence
to max median values. Max value on diagram (100%) is represented by max median force
value and max median speed, and all other values are in percentage to max values. X-axis is
series of measurement. Diagram is given on figure 7. Influence of velocity to force is
evident. Changes in speed are relatively small and range between 96 and 100 %, while
changes in force vary higher at 73 to 100 %. Incremental changes in speed result in
significant change in force inside the tube.
Tube fixation. Tubes of different length and various diameter and thickness were used in
this test. Declared diameter of the tube is 3 mm +/- 5%, resulting in range of 2,85 to 3,15
mm outer diameter. Tube is fixed into 3 mm frame. Dependant to tube diameter, tube is
fixed with greater/lesser friction which is relevant to loss of energy during firing.
105
100
%v (m/s)
%F (N)
95
velocity
90
force
85
Srednje brzine
Srednje sile
80
75
70
65
series
60
1
Fig 7.
Velocity-force diagram
Fixation and tight sealing of tube in frame multiplicities shock wave force. Results of
compared measurements are given in table 4 and diagram is on figure 8.
F (N)
300
250
200
bz
freepridravanja
150
fixed
sa
pridravanjem
100
50
series
0
1
Fig 8.
Shock wavwe force, free and fixed shock tube
Table 4.
Series no.
1
2
3
4
5
Force (free)
[N]
57,75
49,76
50,06
62,66
47,92
Force (fixed)
[N]
248,15
221,74
241,39
113,028
217,132
REFERENCES
[1]
[2]
ESTER Z.: Miniranje I, eksplozivne tvari, svojstva i metode

Rudarsko-geoloko-naftni fakultet, Interna skripta, Zagreb, 145 (2001)
Krsnik J.: Miniranje, Rudarsko-geoloko-naftni fakultet, Zagreb, 180 . (1989)
ispitivanja,
CHARACTERISTICS OF EXPLOSION-SAFETY
OF AIR-METAL DUST SUSPENSIONS
I.V. Egorova, and B.N.Kondrikov
Mendeleev University of Chemical Technology,
125047, Miusskaja sq. 9, Moscow, Russia
Abstract:
Ignition and burning of particulate metals was a subject of many investigations during
for about half-of-a-century. Within the framework of the problem of this paper the
modern state of the theory of ignition of metal particles in a hot oxidizing gas is
described. The theory of ignition was developed on the base of the overall heat
explosion theory of Semenov - Frank-Kamenetsky. Theory of burning of a particle was
elaborated based, first, on the burning theory of a droplet of a liquid fuel and then it
was modified with taking into account natural complications accompanying a metal
particle oxidation. Experimental results that satisfy to the theories can be easily found
among the numerous experimental data till now produced and it is the obvious success
of the theory. The main goal of this paper is summarizing of literature experimental
results in the light of modern scientific ideas.
Keyword:
1.
ignition, burning, particle, droplet, metal
INTRODUCTION
A problem of initiation and propagation of deflagration and detonation in air-dust and

air-liquid droplets clouds presents the great interest in many respects, especially, of course,
from the point of view of providing reliability and safety of technological processes and
avoiding of man-caused catastrophes. The problem attracted attention of specialists in the
field of burning and detonation still in the 50th due to intensive usage of the particulate
metals in formulations of solid rocket propellants.
The new direction of explosive systems development has given birth to many of theories
and experimental investigations that converted, in fact, this rather badly developed field of
science into one of the most advanced directions of theory of explosives, as well as, of
course, the theory of reliability and safety in chemical and other branches of modern
technology.
2.
IGNITION AND BURNING OF A SINGLE PARTICLE

OF A METAL FUEL
The first experimental investigations considering burning of a metal particle in a hot,

usually stagnant, gas appeared on the end of the 50th beginning of the 60th [2-14]. There
were well known and widely quoted works. Results of these works were used in many
subsequent investigations as a base for probation of different theoretical constructions. These
investigations were prolonged in the Soviet Union on the late 60th [15-17]. The state of the
528
problem in the West up to 1963 was considered in the work [2], the overview of those and
consequent investigations, up to 1972, was fulfilled in the book [1].
The theory of ignition was developed on the base of the overall heat explosion theory of
Semenov - Frank-Kamenetsky. It describes quite well the critical ignition conditions of the
particles of magnesium but in the case of aluminum essential deviation from the theory
postulates is observed consisting in presence of minimum on the curves critical temperature
of gas versus particle diameter which is not predicted by the theory.
Fig 1. Explosion ranges of aluminum

particles
in air (1)
and
water stream (2) [18].
The presence of a minimum on the curves describing dependence of the critical

temperature of gas on the aluminum particles dimension was explained by the complex
character of chemical reaction on the surface covered by the oxide film. The surface was
assumed to be covered by the crystalline -Al2O3 just partially, and the area of the surface
being covered by -Al2O3 increases during oxidation. The small particles are covered by the
-Al2O3 earlier than the oxidation reaction would evolve enough heat to ignite the particle
whereas the bigger particles come to ignition before the surface will be covered by the
crystalline oxide and the temperature of gas resulting in ignition can be lower than that in the
case of the smaller particles.
It was stated that dependence of ignition temperature on diameter of Al particles in water
steam, and probably also in air, has a minimum, and the ignition temperature for particles of
7-27 m dia varies from 700 to 1300 0C (1000-1600 K) being much lower that 2300K, the
fusion point of Al2O3 [18].
In the work [19] the particles of Al covered by the amorphous and crystalline Al2O3 were
investigated. The crystalline oxide film was formed from the amorphous one by means of
heating of the particles in atmosphere of Argon, at 870 K. The experimental curves, at autoignition of the particles in the hot water steam, are presented in Fig. 2. Not only the ignition
temperature is much less than 2300 K even for the particles of up to 70 m diameter but the
529
course of the curves is not monotonic, and points of minimum are observed, at a temperature
as small as 1050-1270 K.
Fig 2.
Critical temperature vs particle size in water steam: 1-particles

covered by -oxide; 2-particles covered by the amorphous oxide [19].
Experimental data are taken from [20].
Many Russian scientists studied ignition of magnesium particles

the particles of magnesium was investigated in [21].
[21-25]
. Auto-ignition of
The temperatures of a gas critical for ignition of magnesium particles are compared for
different works in Fig. 3. We see sharp reduction of critical temperature of gas in the field of
small particles at increase of particle diameter independence of this temperature from
particle size in further.
Fig 3.
Critical temperature of gas versus magnesium particles diameter

from the works 1 from [17], 2 from [26], 3 from [26], 4 from [21].
Ignition temperature, ignition delay and the reduced time of burning for particles
of magnesium are much smaller than those for the aluminum particles.
530
3.
BURNING OF ALUMINUM AND MAGNESIUM PARTICLES
Theory of burning of a particle was elaborated based, first, on the burning theory of a
droplet of a liquid fuel and then modified with taking into account natural complications
accompanying a metal particle oxidation. For the particles of aluminum the burning time is
proportional to the particle diameter in the extent of about 1.5 and decreases with the relative
content of oxidizer. In this case the magnesium particles reveal an abnormal behavior,
reduced time of burning, /d2, of the small particles increases when diameter grows, for very
big ones it is independent on diameter whereas the particles of intermediate dimensions burn
with decreasing of the /d2 number, at increasing of the diameter. Presumably, the process of
destruction of the intermediate droplets of the liquid magnesium in the course of burning
takes place.
Three types of functions of the were obtained (Figs. 4-6):
Fig 4. Burning time of Al particle vs ak:

1-particles 70 m dia;
2-particles 140 m dia from [16]
Fig 5. Burning time of Al particle vs

pressure in the bomb.
1-ak=37.5
2-ak=71.5 from [16]
Fig 6. Burning time of Al particles at

different values of ak:
1-ak=37.5;
2- ak=57.5:
3- ak=71.5 from [16]
= 0.67 d1.5/ak0.9
where
is the burning time in ms,

d is diameter of the particle in m, and
ak is the mole fraction of oxidizing gases (water, carbon dioxide) in percents.
531
The burning constant has the form:

kb = d2/ = 1.5 d0.5ak0.9,
i.e. it slightly grows with diameter and essentially decreases with mole percent of the
oxygen containing gases in the gaseous mixture.
For magnesium particles the dependence is more complicated. Burning of magnesium
particles is widely presented in [21,26-39].
Fig 7.
Burning time of magnesium in air versus initial particle size using radiation
of Mg (or MgO) vapor (1), and using the full time of luminescence of
particles (2) from [40].
The line (1) in Fig. 7 has a form:

= 0.042d1.36 ms (d in m)
what gives
/d2 = 0.042 d-0.64, ms/m2
Joint construction of three dependences from [15,41,40], is presented in Figs. 8 and 9.
532
Fig 8.
Reduced burning time versus diameter of particles:1 from [15], 2 from

, line 3 from [40] is not taken into account at drawing of the general
curve.
[41]
One may conclude that, at first, burning time is proportional to cube of the diameter
(Fig. 8) and then, at higher diameters, is proportional to the square of it. Obviously, the small
particles liquefy, collapse, i.e. as though blow up at burning. But
the data in the intermediate diapason of diameters show that a region exists where /d2 falls
down at increasing of d and consequently the burning time is proportional to diameter in the
extent lower than 2. It means that there are some sorts of magnesium intensity of destruction
of that increases with particle size erasing.
Fig 9.
Reduced burning time

versus diameter of
particles:1 from [15],
2 from [41], line 3 from [40] is

taken into account at drawing
of the general dependence.
533
4.
CONCLUSIONS
Within the framework of the problem of this paper the modern state of the theory of
ignition of metal particles in a hot oxidizing gas is described.
Theory of burning of a particle was elaborated based, first, on the burning theory of a
droplet of a liquid fuel and then modified with taking into account natural complications
accompanying a metal particle oxidation. For the particles of aluminum the burning time is
proportional to the particle diameter in the extent of about 1.5 and decreases with the relative
content of oxidizer. In this case the magnesium particles reveal an abnormal behavior,
reduced time of burning, /d2, of the small particles increases when diameter grows, for very
big ones it is independent on diameter whereas the particles of intermediate dimensions burn
with decreasing of the /d2 number, at increasing of the diameter. Presumably, the process of
destruction of the intermediate droplets of the liquid magnesium in the course of burning
takes place.
Experimental results that satisfy to the theories can be easily found among the numerous
experimental data till now produced and it is the obvious success of the theory. There are
many contradictions between the data of different investigators, however, and obvious
complications that do not taken into account at these theories construction.
Acknowledgment
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
POKHIL, P. F., BELYAEV, A. F., FROLOV, Yu. V., LOGACHEV, V. S., KOROTKOV,
A. I.: Burning of powdered metals in active media, Nauka, Moscow, 1972, 294 pp.
MARKSTEIN, G. H.: AIAA, J., 1963, Vol. I, 3.
COFFIN, K.: 5th Symposium on Combustion, N.Y., 1953
CASSEL, H. M., LIEBMAN, I.: Combustion and Flame, 1959, Vol. 3, 4, 1963, Vol. 7, 1.
CASSEL, H. M., LIEBMAN, I.: Combustion and Flame, 1962, Vol. 6, 3.
FRIDMAN, R., MAEK, A.: Combustion and Flame, 1962, Vol. 6, 9.
FRIDMAN, R., MAEK, A.: 9th Symposium (Int.) on Combustion, New York-London,
Academic Press, 1963, pp. 703.
MAEK, A.: 11th Symposium (Int.) on Combustion. The Combustion Institute. Pittsburgh,
1967, pp. 203.
DAVIS, A.: COMBUSTION AND FLAME, 1963, VOL. I, 4.
BARTLETT, R. W., ONG, J. N. at al.: Combustion and Flame, 1963, Vol. 7, 3.
GLASSMAN, I.: Metal combustion processes, A.R.S. Preprint, 938, 1959.
BRZUSTOWSKI, T. A.: Canad. Aer. And Space J., 1963, Vol. 9, 5.
BRZUSTOWSKI, T. A. and GLASSMAN, I.: Spectroscopic investigation on metal
combustion, Heterog. Comb., N.Y.-Ld., 1964.
BRZUSTOWSKI, T. A. and GLASSMAN, I.: Vapor-phase diffusion flames in the combustion
of magnesium aluminum, Heterog. Comb., N.Y.-Ld., 1964.
534
[15] KLYACHKO, L. A.: Burning of unmoved particle of easily boiled metal, Combustion,
Explosion and Shock Waves, 1969, Vol. 5, 3, pp. 404-413.
[16] BELYAEV, A. F., KOROTKOV, A. I., FROLOV, YU. V.: Ignition and combustion of single
particle of small-grained aluminum, Combustion, Explosion and Shock Waves, 1968, Vol. 4,
3, pp. 323-330.
[17] GUREVICH, M. A., STEPANOV, A. M.: Critical conditions for the metallic particle ignition,
Combustion, Explosion and Shock Waves, 1968, Vol. 4, 1, pp. 189-196.
[18] GUREVICH, M. A., LAPKINA, K. I., OZEROV, E. S.: Maximum conditions of aluminum
particle ignition, Combustion, Explosion and Shock Waves, 1970, Vol. 6, 2, pp. 172-176.
[19] GUREVICH, M. A., OZEROV, E. S., YURINOV, A. A.: On the influence of oxide film on the
characteristics of aluminum ignition, Combustion, Explosion and Shock Waves, 1978, Vol. 14,
4, pp. 50-55.
[20] GUREVICH, M. A., LAPKINA, K. I., OZEROV, E. S.: Ignition limits of aluminum particles,
Combustion, Explosion and Shock Waves, 1970, Vol.6, 172.
[21] EZHOVSKY, G. K., MOCHALOVA, A. S., OZEROV, E. S., YURINOV, A. A.: Ignition and
combustion of magnesium particle, Combustion and explosion, Materials of III All-Union
Symposium on combustion and explosion, ., Science, 1972, pp. 234-241.
[22] VALOV, A. E., GUSACHENKO, E. I., SHEVTSOV, V. I.: Influence of pressure of oxidizing
media and oxygen concentration on ignition of single magnesium particle, Combustion,
Explosion and Shock Waves, 1991, Vol. 27, 4, pp. 3-7.
[23] GOSTEEV, YU. A., FEDOROV, A. V.: Ignition of magnesium particle (Divided Model),
Combustion, Explosion and Shock Waves, 1996, Vol. 32, 4, pp.5-12.
[24] RUMANOV, E. N., KHAYKIN, B. I.: Critical conditions of spontaneous ignition of whole
particles, Acad. of Science of USSR, M, 1969, 9 pp.
[25] FLORKO, A. V., ZOLOTKO, A. N., KAMINSKAYA, N. V., SHEVCHUK, V. G.: Spectral
investigations of magnesium particles ignition, Combustion, Explosion and Shock Waves,
1982, Vol. 18, 1, pp. 17-22.
[26] EZHOVSKY, G. K., OZEROV, E. S.: Ignition of magnesium powder, Combustion, Explosion
and Shock Waves, 1977, Vol. 13, 6, pp. 845-852.
[27] OZEROV, E. S., YURINOV, A. A.: Combustion of aluminum-magnesium alloy particles in the
water vapor, Combustion, Explosion and Shock Waves, 1977, Vol. 13, 6, pp. 913-916.
[28] DEREVYAGINA, M.E., STESIK, L.N., FEDORIN, E.A.: Study on ignition and combustion of
aluminum and zinc in air, Combustion, Explosion and Shock Waves, 1977, Vol. 13, 6, pp.
852-857.
[29] DEREVYAGA, M. E., STESIC, L. N., FEDORIN, E. A.: Combustion regimes of magnesium,
[30] DEREVYAGA, M. E., STESIK, L. N., FEDORIN, E. A.: Experimental study of critical
condition in an unmovable layer of nonporous catalyst, Chemical physics, 1978, Vol. 14, 6,
pp. 44-49.
[31] SHAFIROVICH, E. YA., GOLDSHLEGER, U. I.: Ignition and combustion of magnesium
particles in carbon dioxide, Combustion, Explosion and Shock Waves, 1990, Vol. 26, 1, pp.
3-11.
[32] GOLDSHLEGER, U.I., SHAFIROVOCH, E.YA.: Regimes of magnesium burning in carbon
oxides. 1. Burning in CO2, Combustion, Explosion and Shock Waves, 1999, Vol. 35, 6,
pp.42-50.
[33] GRACHUKHO, V. P., OZEROV, E. S., YURINOV, A. A.: Magnesium particle burning in
water steam, Combustion, Explosion and Shock Waves, 1971, Vol. 7, 2, pp. 232-236.
[34] ALTMAN, I. S., VOVCHUK, YU. I., The thermal regime of magnesium particle vapor-phase
burning, Combustion, Explosion and Shock Waves, 2000, Vol. 36, 2, pp.7476.
535
[35] OZEROV, E.S., SKVORTSOV, I.I.: Burning of magnesium particles conglomeration,

Combustion, Explosion and Shock Waves, 1971, Vol. 7, 2 pp. 227-232.
[36] GUREVICH, M. A., IGNATOV, G. V., OZEROV, E. S., SOTNICHENKO, B. I.: Statistic
method of study of metal particles burning, Combustion, Explosion and Shock Waves, 1969,
Vol. 5, 3, pp. 441-444.
[37] FLORKO, A. V., KOZITSKY S. V., ZOLOTKO, A. N., GOLOVKO, V. V.: On transfer
mechanism of condensed combustion products to the surface of burning magnesium particle,
[38] KLYACHKO, L. A.: Burning of easily boiled metal particle moving relatively gaseous
oxidizer, Combustion, Explosion and Shock Waves, 1971, Vol. 7, 2, pp. 236-241.
[39] GUREVICH, M. A., OZEROV, E. S., RYBINA, L. S.: Calculation of a speed of steam phase
diffusive burning of a metal particle, Combustion, Explosion and Shock Waves, 1974, Vol. 10,
3, pp. 363-372.
[40] VALOV, A. E., KUSTOV, YA. A., SHEVTSOV, V. A.: Spectroscopic study of single
magnesium particles burning in air and carbon dioxide, Combustion, Explosion and Shock
Waves, 1994, Vol. 30, 4, pp.29-35.
[41] FLORKO, A. V., KOZITSKY, S. V., PISARENKO, A. N., MATSKO, A. M.: Investigation of
magnesium single particles under reduced pressure, Combustion, Explosion and Shock Waves,
1993, Vol. 29, 3, pp. 115-120.
536
CRYSTAL STRUCTURES
OF THE SILVER AND POTASSIUM SALTS OF NITROFORM
Michael Gbel, Thomas M. Klaptke, and Peter Mayer
Department of Chemistry and Biochemistry, Ludwig-Maximilians University of Munich,
Butenandtstr. 5 - 13 (Haus D), D-81377 Munich, Germany
Abstract:
Silver and potassium nitroformate were synthesized and the crystal structures were
obtained.
Keywords:
Salts of trinitromethane, nitroform, crystal structures
Silver and potassim nitroformate were synthesized according to equations (1) and (2).[1,2]
(1)
NO2
O2N
NO2
KOH, Glycerol, H2O

O2N
NO2
C
NO2
NO2
(2)
NO2
O2N
NO2
Ag2O
H
O2N
CH3CN
Ag
C
NO2
NO2
537
Potassium nitroformate was recrystallized from nitromethane yielding yellow octahedral

single crystals. The molecular structure in the crystalline state was determined by X-ray
spectroscopy:
tetragonal, I -4 2 d (no. 122), a = 11.3594(5), c = 8.7883(5) , V = 1134.0(1) 3 ,
Z = 8, = 2,215 g/cm3 , R1 = 0,0273, wR2 = 0.0496.
Silver nitroformate was recrystallized from water yielding orange needle-like single
crystals of the monohydrate of silver nitroformate.
The molecular structure in the crystalline state was determined by X-ray spectroscopy:
monoclinic, P 1 21/n 1 (no. 14), a = 7.5172(2), b = 7.8572(2), c = 20.7387(5) ,
= 91.4() , V = 1224.55(6) 3, Z = 8, = 2.993 g/cm3, R1 = 0.0315, wR2 = 0.0709.
REFERENCES
[1]
[2]
A. K. MACBETH, W. B. ORR, J. Chem. Soc., p.542, 1932

G. S. HAMMOND, W. D. EMMONS, Tetrahedron, 19 Suppl. 1, p.177ff, 1932
538
STUDY ON SAFETY
OF EXPANDED AMMONIUM NITRATE EXPLOSIVE
Hu Bingcheng, L Chunxu, Liu Zuliang
School of Chemical Engineering, Nanjing University of Science & Technology,

Nanjing 210094, P.R. China
Abstract:
The safety properties of expanded ammonium nitrate(AN) explosive and its main
composition were tested, such as the mechanical sensitivity, the sensitivity to auget, the
sensitivity to shock wave and the sensitivity to spark of static electricity. The results show
that the safety of expanded AN and expanded AN explosive is similar to that of common
AN and its TNT-explosive.
Keywords: safety, expanded AN, common AN, explosive, sensitivity
As a new type of no-TNT explosive, expanded AN explosive has become popular with
more and more factories in China, for its perfect physical and explosive properties. The
thoroughly study on the safety properties of expanded AN and its explosive is of great
importance to the research and manufacture of expanded AN explosive.
1.
SAFETY EXPERIMENTS
1.1
Mechanical sensitivity
The ability for explosives to explode by mechanical action is called mechanical sensitivity
of explosive. As the main stuff of expanded AN explosive, expanded AN should be low in
mechanical sensitivity, in order to suit the industrial manufacturing of explosives. The
mechanical sensitivity of expanded AN and common AN was tested simultaneously in the
experiment.
1.1.1. Impact sensitivity
The impact sensitivity of expanded AN was tested by Karstel dropping-hammer
instrument. The experiment was performed in groups, and each group had 25 charges. Two
groups of samples were tested under every condition The results show in table 1.
539
Table 1. The impact sensitivity of expanded AN

Weight of the
Dropping height of
Weight of the
Percent of the
hammer/
the hammer/mm
charge/mg
explosion100
25
10
250
50
Expended AN
25
10
250
50
Common AN
100
10
250
50
Expanded AN
100
10
250
50
Common AN
100
10
500
50
Expanded AN
100
10
500
50
Sample
Temperature/
Commom AN
The energy of the dropping-hammer to impact on the explosive can be calculated through
the following equation:
E = 1/2 mv2 = mgh
in the equation, m - the weight of the dropping-hammer;
h - the height of the dropping hammer.
If h = 250 mm, the energy of the dropping-hammer E1 = 109.80.25 = 24.5 J.
If h = 500 mm, the energy of the dropping-hammer E2 = 109.80.5 = 49.0 J.
1.1.2. Friction sensitivity
The friction sensitivity of expanded AN was tested by Kezlof friction-pendulum. The
experimental conditions are as follows: the length of the pendulum is 760 mm, the gauge
pressure is 539.2 MPa, the swinging angle is 96. The experiment was performed in groups,
and each group had 25 charges. Two groups of samples were tested under every condition The
results show in table 2.
Table 2. The friction sensitivity of expanded AN
Weight of the
Swinging
Weight of the
Percent of the
hammer/
angle/
charge/mg
explosion100
25
1.5
96
30
Expended AN
25
1.5
96
30
Common AN
100
1.5
96
30
Expanded AN
100
1.5
96
30
Sample
Temperature/
Commom AN
The energy of the pendulum hammer to hit the staff can be calculated through the
following equation:
540
E = 1/2 mv2 = mgh = mgL (1cos)

in the equation, m - the weight of the pendulum hammer;
L - the length of the pendulum; - the swinging angle.
E = 1.59.80.76 (1 - cos 96) 12.34 J
1.2
Sensitivity to auget
The sensitivity of expanded AN to auget was tested under various densities, in order to
seek the density range of expanded AN with secure sensitivity to auget. The experiments were
carried out through the following method: the sample which was put in a drum (80 mm, H
200 mm) was detonated by an 8# electric detonator. The results are shown in the table 3.
Table 3. The sensitivity of expanded AN to auget

Pulverization state of the
sample
Density/gcm-3
Temperature of the
charge/
Result of detonating
Not pulverized
0.25
40~45
Not detonated
Not pulverized
0.30
40~45
Not detonated
Not pulverized
0.35
33.5
Not detonated
Not pulverized
0.37
33.5
Detonated
Not pulverized
0.40
33.5
Detonated
Not pulverized
0.45
29.0
Detonated
Not pulverized
0.50
29.0
Detonated
Not pulverized
0.55
40~45
Detonated
Not pulverized
0.60
33.5
Not detonated
Not pulverized
0.65
40.0
Not detonated
Pulverized by the ball mill
0.45
22
Detonated
0.50
22
Detonated
0.55
22
Detonated
0.60
22
Detonated
1.3
Sensitivity to shock wave
Expanded AN has very large surface area, which will efficiently increase the explosive
reaction surface and promote the transmission of detonation wave. Therefore, the sensitivity
of Expanded AN explosive to shock wave was tested. The experiment method is displayed in
the following scheme:
Driving charge
Passive charge
L
L - the distance for 50% of the passive charge to detonate
541
The air between the driving and passive charge acts as the medium, which will form the
shock wave when the driving charge is detonated. The shock wave will decay after moving
along some distance and enter the passive charge. Whether the passive charge can be
detonated or not depends on the intensity of the shock wave. The shorter is the distance
between the driving and passive charge, the stronger is the shock wave which acts on the
passive charge. So this distance reflects the sensitivity of the sample to shock wave. The Tetryl
charge ( = 42 mm, h = 22 mm, m = 50 g) was used here as the fixed driving charge to test the
distance (L) of the passive charge ( = 35 mm, h = 150 mm, m = 130 g). The results show that
the value (L) of the expanded AN explosive and 2# rock AN-TNT explosive are 21.0 cm and
17.0 cm, respectively.
1.4
Sensitivity to spark of static electricity
Just as common AN explosive, expanded AN explosive is also pulverous in the industrial

process, especially in the air-flow manufacturing technology quantities of dust are produced.
They accumulate to form the dust cloud with static electricity. Once the accumulated static
electricity discharge, it will give off electric sparks. whether these electric sparks can light the
dust or not hasnt been made clear. Although the energy to light the dust can be measured out
through spark discharging method, it is still difficult to obtain the accurate results with
different experimental conditions and methods. Therefore, we designed and processed a series
of equipments to measure the sensitivity of explosive to the spark of static electricity. With
them the sensitivities of expanded AN explosive and its compositions to the spark of static
electricity were tested.
The experimental principle is as follows: The air from the blast fan is heated by a heat
plate to form the heated air flow, of which the velocity and temperature can be altered by
adjusting the air inlet of the blast fan and the resistance of the heat plate. Through a spiral
feeder the sample is added to the heated-air-flow pipe. After the velocity and temperature of
the sample reach a stable value respectively, an electric spark is given off from a certain place.
The energy of the electric spark can be controlled and exactly measured.
The velocity and temperature of the sample in the experiment are 13.2 ms-1 and 70 C
respectively, which are basically close to those of the heated air flow in the air-flow
technology of the AN-TNT explosive. The feeding velocity of the spiral feeder is 5 gs-1. The
results are displayed in the table 4.
542
Table 4. Stability of expanded AN explosive and its compositions to static electricity

Discharge voltage/V
800
1200
1600
2000
2400
2800
3200
3600
4000
Energy of spark/J
0.7
1.4
2.3
3.6
5.0
6.9
9.0
11.6
14.7
Common AN
Expanded AN
Oil phase
Wood powder
Sample
Expanded AN +
Wood powder
Expanded AN +
Oil phase
Expanded AN
explosive
Note: 0 indicates that the sample is not lighted; 1 indicates that the sample is lighted.
2.
DISCUSSION
The above-mentioned results come to a conclusion that the sensitivity of expanded AN

and its new no-TNT explosive without increases to a certain extent, especially in the
sensitivity to auget and shock wave, and their mechanical and static electricity sensitivity are
basically close to those of common AN and its TNT explosive. This conclusion indicates that
it is advantageous and practical to produce expanded AN explosive from expanded AN. This
new type of explosive is easy to be detonated and safe to be processed and used as well. In
addition, following conclusions can also be drawn from the experiment results:
1. Expanded AN is as the same dull as common AN to mechanical action. The results in
table 1 show that expanded AN has a bit more sensitivity to impact sensitivity than
common AN only under the strict conditions, such as the temperature 100 C and the
dropping height 500 mm.
2. Expanded AN may has auget sensitivity when its density exceeds 0.3gcm-3, and loss
its auget sensitivity when its density increases to a certain value. For the reason above,
it is easy to artificially control the density of expanded AN, abate its auget sensitivity
and ensure the safety in the industrial process.
3. Compared with 2# rock AN-TNT explosive, the sensitivity of expanded AN explosive
to shock wave is enhanced, which improves the explosibility of the AN-TNT explosive
system without affecting the safety in the industrial process.
4. The results in table 4 show that the static electricity sensitivity of expanded AN and its
no-TNT explosive varies little within the normal range. Only the conditions with the
discharge voltage less than 4000 V and the spark energy less than 14.7 J were
considered in the experiment, though the circumstance that the static electricity voltage
is more than 4000 V and a spark exits simultaneously will appears unlikely. So the
air-flow technology can also be used in the manufacture of expanded AN explosive
from expanded AN.
543
REFERENCE
[1]
LU HUA, WAN SHANHONG: Ammonium Nitrate Explosives. Publishing House of National

Defence Industry, Beijing.1972
[2]
LU CHUNXU, LIU ZULIANG, NI OUQI: Industrial explosives, Publishing house of Weapon

Industry, Beijing,1994
[3]
T.URBANSKI: The Chemistry and Technology of Explosives, 1965, (2):332
544
DETERMINATION OF METAL OXIDES AND THEIR

CONCENTRATION IN THE IDENTIFICATION PARTICLES
L. Huskov*, J. rmkov*, K. Ventura*, Z. Aktein**, and M. tancl**
*Department of Analytical Chemistry, Faculty of Chemical Technology,
University of Pardubice, nm. s. legi 565, 532 10 Pardubice, CZ
** Research Institute of Industrial Chemistry, Explosia a.s., 532 17 Pardubice-Semtn, CZ
Abstract:
Identification Particles used for the purpose of the post blast identification of explosives,
have a coding system based on the combination of metal oxides and their different
concentration. These materials are consisted of the polymeric matrix, iron powder
(ferromagnetic properties), UV light active dyestuff and various metal oxides in different
ratio. A suitable analytical method must be applied for an accurate characterization of
these metals components in the particles for finding needed information, it means to
determine the production place, year of production and eventually production batch of
misusing explosives. The flame atomic absorption spectrometry (F-AAS) performs rapid
and sufficiently sensitive method suitable for this purchase. In this work, the method for
the determination of ratio of oxides in a few types of identification particles using F-AAS
is described. The microwave digestion was applied for the samples containing
polybutadiene matrix, leaching with nitrohydrochloric acid was used if the samples have
contained epoxide matrix. Using the described procedure, the determination was free of
interferences so that a simple calibration curve method could be applied. The detection
limits (LODs) for individual elements (lead, copper, magnesium and zinc) were 0.53
mg.l-1, 0.11 mg.l-1, 0.01 mg.l-1 and 0.11 mg.l-1, respectively.
Keywords:
1.
identification particles, F-AAS, lead, copper, magnesium, zinc
INTRODUCTION
The marking of explosives is an important part of the defense against misusing of plastic
explosives. The marking can be divided into a pre-blast tagging and a post-blast one. The
objective of pre-blast marking is to enable an easy detection of an explosive by means of the
addition of the substances with higher vapour pressure. The post-blast marking is done in
order to make easy the disclosure of the explosive used after the explosion. So called
identification particles (IP) are used for this purpose. For instance, Microtaggant
Identification Particles of Microtrace Inc. (USA) developed by the firm 3M International are
used for this purpose at present. Hardened polymeric matrix, so called reactoplast based on
alkyl-melamin resin, is a carrier of identification particles. The particles should have such
properties to be able to survive the extreme conditions of the explosion relatively
undamaged. IP are basically mixtures of a polymeric carrier and solid powdered filling
agents, which ensures the function of a body. A particle is in the form of a homogenous unit
or it consists of more layers [1]. Up-to-date research in our country are focused on types of IP
based on the different concentration of metal oxides, which could be able to search out by
means of a magnet and subsequently to determine metals using a suitable analytical method.
545
In this work the most useful AAS method for this purchase was developed and applied
to the determination of lead, copper, magnesium and zinc in several types of identification
mixtures.
2.
EXPERIMENTAL
2.1
Equipment
All measurements were carried out using Zeiss AAS 3 atomic absorption spectrometer
(Carl Zeiss, Jena, Germany) with deuterium arc correction in single beam arrangement. The
hollow cathode lamps for copper and magnesium were obtained from Photron Pty.
(Australia), the Narva lamps were used for lead and zinc. In all cases the lamp current was
5 mA. The wavelength was set at 285.2 nm for Mg, 213.9 nm for Zn, 217.0 nm for Pb and
324.8 nm for Cu. Spectral slit-width was 0.2 nm for Mg and Zn and 0.3 nm for Pb and Cu.
An air-acetylene fuel-lean flame of 100 nm length and with an optimum of 8 mm height of
light beam was employed in all cases.
Microwave digestion was performed using BM 1S/II and Uniclever II microwave
systems (Poland). The maximum power which could be reached in one vessel is 100 W for
BM 1S/II and 600W for Uniclever II.
The water used in experiments was filtered using DEMIWA 5-ROI (WATEK, Czech
Republic) pure water system to 0.05 S cm-1.
2.2
Reagents and solutions
All solutions used in the study were of analytical grade or higher quality. All solutions
were prepared using deionised water. Standard solutions of Pb, Mg, Cu and Zn (1 g l-1) were
obtained from Analytica Co. Ltd (Prague, Czech Republic). Nitric and hydrochloric (36%)
acids (65%) were Suprapur (Analytica).
2.3
Procedure
Sample preparation:
(i)
Sample containing epoxide matrix: 0.1-0.2 g (0.0001) was placed to the

volumetric flask and 5 ml of nitrohydrochlorid acid (HNO3+HCl,1+3) was added
and refilled with deionised water to final volume of 50 ml.
(ii)
Sample containing polybutadiene matrix: 0.1-0.2 g of sample was weighed out to

the polytetrafluorethylene microwave vessels. Than 3 ml of 65% HNO3 and 2 ml
of HCl (36%) was added. The microwave digestion programme consisted of two
following steps: (1) 10 min at 60% power, (2) 5 min at 80% power, pressure =
(17-20 atm). Than the mineralized sample was placed to the volumetric flask and
volume refilled to 50 ml.
Calibration:
Calibration was prepared in the following ranges: 0.1-0.5 mg l-1 for Mg, 0.5-2.5 mg l-1
for Zn, 2.0-15.0 for Pb and 1.0-4.0 mg l-1 for Cu. All calibrations were linear in the
investigated range. The correlation coefficients were in all cases higher than 0.999.
546
3.
As it was already mentioned in the introduction section, the aim of this study was to
develop the rapid and accurate procedure for the determination of selected metals in the
Identification Particles. The accurate sample preparation represents the non-separable part of
the analytical procedure. In our case the sample preparation represented the most
problematic part of the analytical procedure. With respect to disposition of the samples
finally the two ways of sample preparations were chosen. The microwave digestion was
applied for the samples containing polybutadiene matrix and leaching with nitrohydrochloric
acid was used if the samples have contained epoxide matrix. Because of high metals content,
the samples could be further diluted (at least 25 fold), which resulted in a decrease of matrix
interference effects, so that a simple calibration curve could be applied. Using the described
procedure the detection limits calculated [2] were: 0.53 mg l-1, 0.11 mg l-1, 0.01 mg l-1, 0.11
mg l-1 for Pb, Cu, Mg and Zinc respectively. The method developed was applied to the
determination of all metals in three samples with expected values. The results obtained were
in a relatively good agreement with given values (Table 1). It is evident from the table that in
several cases values for zinc are lower when compared with expected values. This fact was
probably caused owing to high dilution of sample before the analysis (up to 100 fold).
Table 1. Determination of Pb,Cu,Mg and Zinc in the Identification Particles
expressed as appropriate oxides by AAS method given in %
Component
IP1
IP2
IP3
Given
Found
Given
Found
Given
Found
PbO
10
9.6 0.8
12
12.3 0.1
CuO
10
8.8 0.3
2.6 0.5
MgO
10
9.0 0.1
ZnO
10
7.5 0.1
10
9.8 0.2
6.6 0.1
4.
CONCLUSIONS
It is possible to see from the presented results that atomic absorption spectrometry is
a suitable method for the purpose of the determination of lead, copper, magnesium and zinc
in the identification particles. The proposed method performed the simple, rapid and reliable
control using instrumentation which is easy accessible in many laboratories. Moreover it can
be supposed that described method can be easily applicable for the determination of many
further metals.
Acknowledgement:
Youth and Sport of the Czech Republic (Project MSMT 0021627502).
547
REFERENCES
[1]
[2]
M. VEEA, L. PROKPEK: Selected polymers as the carriers for the identification particles,
Annual Report, Institute of polymeric materials, University of Pardubice, 2003
M. MELOUN, J. MILITK: Chemometrics for analytical chemistry, Ellis Horwood, England,
1994
548
RESEARCH OF ARTIFICIAL AGEING SUITABLE CONDITIONS

DURING THE INVESTIGATION OF PROPELLANTS LIFETIME
Martina Chovancova, Peter Ocko, Alzbeta Pechova, and Milos Lazar
VTSU (Military technical and Testing Institute) Zahorie, 905 24 Senica, Slovak Republic
Abstract:
This paper describes the investigation of the effect of different artificial ageing
conditions on chemical stability of propellants. The chemical life of two samples of
propellant was calculated by the depletion of stabilizer. The ageing conditions were
choosing according to first study draft of AOP-48 edition 2.
Keywords:
1.
artificial ageing, propellants, chemical life, stabilizer depletion
INTRODUCTION
Nitrocellulose based propellants show a slow, but constant decomposition of nitrate ester
groups under the formation of nitrogen oxides and nitric acids. These components catalyse
the further decomposition of the propellant and may finally lead to an autocatalytic
decomposition. It can be followed with the temperature increase and ignition of propellant.
To prevent autocatalysis or at least reduce the risk of it, stabilisers are added to the
propellants. These substances have the ability to response with nitrogen oxides by the
formation of nitrosated and nitrated compounds. Typical stabilisers for propellants are
diphenylamine DPA, N-nitrosodiphenylamine NODPA and the urea derivates ethylcentralite
CI and methylcentralite CII. The concentrations of stabilisers in the propellant give us
important information about the actual grade of propellant decomposition provided that we
know aborigine concentration of stabilisers. The investigation of stabiliser contents is
notable process of the stability assessment and plays significant role in the prediction of
propellants lifetime. The artificial ageing is suitable and effective form for the investigation
of propellants lifetime. It is process of accelerated ageing, which simulates spontaneous
ageing in ammunition system. It is usually realized by heating at the higher temperatures
which accelerate decomposition processes in the propellant. Its goal is to send the propellant
to such state in which it will be to occur with most probability after number of years of
spontaneous ageing. It is possible by means of artificial ageing to predict the changes of
stability, sensitivity, mechanical and function properties of propellants, thus their lifetime.
From previous publications in this area is known that the process of DPA chemical
reactions and its derivates in the propellant is very effected by ageing conditions it is
mainly the possibility of propellant to contact with air and gases escaping from propellant.
Petrlek and Skldal [2] investigated the artificial ageing condition effect on DPA and CI
reactions and found out that DPA losses are higher in case if the air is enterable during the
ageing. From works [4,5] is known that these losses are dependable from amount of oxygen
around the sample and they have maximal value during the initial phase of artificial ageing.
It is known from summary of work [2] that propellant samples aged with entrance of air offer
markedly higher evolve rates of gaseous products after induction period then samples
549
artificial aged in inert atmosphere. Authors of work [3] suppose increased rate of gase evolve
by means VST relates with increased ratio of skeleton urea products fission at higher initial
CI concentrations. Wilker and col. [6,7] investigated stability of different propellants during
ageing and stabiliser depletion. He accented the meaning of ageing conditions. In closed and
completely filled systems (HFC, VST) NO2 is rapidly reduced to NO, it lead to a faster
oxidation of the nitrocellulose. DPA is easily reacting with NO to NODPA which stabilises
the propellant as well but is no more reacting with NO. In open systems (WL tests) NO and
NO2 present and NODPA is then nitrated and converted into di- and trinitro-DPA derivates.
Wilker noticed that oxygen has three major effects on the samples: the nitrocellulose itself is
oxidized, the stabilisers can be oxidized and it backs NO to NO2, which is a strong oxidizing
agent. So they recommended to do ageing studies in complete ammunition articles or, at
least, under storage conditions as closely as possible related to the conditions in ammunition.
2.
EXPERIMENTAL
2.1
Artificial ageing
The chemical and thermal stability of propellants had been tested by VST and DTA
during their artificial ageing. On the base kinetic depletion of stabilizer were calculated
chemical lifetimes of investigated propellants according to the first study draft of AOP-48
edition 2. This publication recommended different types of artificial ageing in viewpoint of a
temperature, duration of ageing and type of propellant packing.
There were two propellants tested. The first one was single base propellant S6 stabilised
with DPA and CI and the second one was three base propellant D3 stabilised with CI. The
propellants were aged under three different temperatures, ageing period and type of
packaging. Circumstances of artificial ageing are in table 1.
Table 1. Circumstances of artificial ageing
Temperature Ageing
(C)
(days)
50
60
65
period Corresponding Type of packaging

natural ageing
(year)
60, 120, 180,

2, 4, 6, 10
301
24, 48, 72,
2, 4, 6, 10
123
13, 26, 39, 65 2, 4, 6, 10
Glass tubes with stoppers (NATO

tube)
Sealed polymer-coated aluminium
bags
Original cartridge only for S6
sample
Ageing periods were chosen according to AOP-48 and STANAG 4582. Propellant S6 is
used in small calibre cartridges and therefore it was aged in original cartridge. Propellant S6
is used in large calibre cartridges and therefore it wasnt aged in original cartridge. The
loading density of S6 was the same in glass tube as in original cartridge.
550
2.2
Tests
Chemical stability test was tested by VST according to STANAG 4556 and TP-SHVV108-(08)-00-ZH by the devices Stabil 16-Ex. The test conditions were: 100C for 40 hours
and 5 g sample. DTA were performed according to TP-SHVV-1080-(05)-00-ZH nonisothermal method by the devices DTA 550-Ex. The test conditions were: 50 mg sample and
rate of heating 5C/minute.
Stabiliser depletion and calculation of propellant lifetime were assessment according to
the first study draft of AOP-48 edition 2 and TP-SHVV-1080-(12)-00-ZH by HPLC (High
Pressure Liquid Chromatography).
2.3
Stabilizer depletion investigation by HPLC
The amount of effective stabilizer was determined by High Performance Liquid

Chromatography using the method with the internal standard. The samples of propellants
were extracted in the ultrasonic bate minimal 4 hours in acetonitrile, the nitrocellulose was
precipitated by 2% aqueous calcium chloride solution and eliminated using the centrifuging
and consequently the sample supernatant liquid was filtrated through GF50 filter.
HPLC measurement conditions:
Column - MAG 1, 4,6x250mm, C 18
Temperature - 35C
Eluent composition -
acetonitrile/water (55/45 volume) for propellant S6

acetonitrile/water (80/20 volume) for propellant D3
Eluent flow - 1ml/min

Injection volume - 20 l
UV detector, wave length - 205 nm
If the propellant is stabilized with mixture of DPA and CI (prop.S6), the decrease in
DPA content during the standard ageing period has been adopted as the stability criterion of
the propellant, because there are not significant changes in CI content. The content of
effective stabilizer (ES) was calculated in % according to the relation:
% ES = % DPA + (0,85 x % NODPA)
2.4
a)
(1)
Calculation of propellant chemical life

For each test temperature the plot time t of ageing versus the natural log of the
ratio of the original stabilizer a to the concentration of stabilizer after ageing (ax) was created
t = f ln a / (a-x)
(2)
b)
The rate constant k of stabilizer depletion at each of the test temperature was
determined by determining the slope of the line t = f ln a / (a-x), the slope is
equal to the reciprocal of the rate constant 1/k
c)
The plot the natural log of the rate constant k versus the reciprocal of the
respective ageing temperatures T (in Kelvins) was created
ln k = f1/T
(3)
551
d)
By multiplying of the slope of the line ln k = f 1/T by the Gas Constant the
Activation Energy Ea for the stabilizer consumption process was calculated
e)
The value of the rate constant k25 for the storage temperature at 25C was
determined by the extrapolation of the relation ln k = f 1/T
f)
f) The chemical life of propellant as the time t25 which is needed for the
decrease of the
stabilizer content to 50% of initial value at the storage
temperature at 25C was calculated
according to the equation
t25 = 1/k25 . ln a/a-x,
(4)
where x = 50% of initial value of the stabilizer content.
3.
RESULTS A DISCUSSION
3.1
Results of VST
The results of VST are in table 2. The table includes volume of gas evolved on heating
the propellant in ml for all condition of artificial ageing.
Table 2. Results of VST
AT
(C)
AP
(days)
65
13
26
39
65
24
48
72
123
60
120
180
301
60
50
Glass
tube
3,311
3,821
4,958
8,637
2,915
3,224
5,713
8,619
3,074
3,996
9,359
13,559
Propellant S6
Sealed bag
Cartridge
2,528
2,864
3,514
4,542
2,757
2,831
3,372
4,048
2,474
2,746
4,434
6,280
3,176
3,365
4,756
5,531
3,140
3,416
4,302
4,446
2,825
2,862
4,644
7,280
Propellant D3
Glass tube Sealed bag
14,787
14,665
13,383
12,672
12,185
13,127
13,316
13,512
12,002
14,1465
13,457
13,862
13,627
13,275
15,604
15,631
12,980
12,678
12,886
14,352
11,691
12,854
12,558
12,767
Note:
AT Ageing Temperature
AP Ageing Period
Volume of gas evolved of original propellants:
S6 - 2,044 ml
D3 -12,282 ml
From the results of VST include that chemical stability of single base propellant S6 had
been decreasing, especially in the case artificial ageing in glass tube. There was the largest
increase the volume of gas evolved. The volume of gas evolved is increased six times for
ageing temperature 50C. The volume of gas evolved is increased only softly in the case
propellant S6 and ageing condition: artificial ageing in sealed bag, higher temperatures of
ageing and shorter ageing period. The largest volume of gas was evolved after the longer
ageing period and the lowest temperature of ageing. These values are smaller than values in
552
the case ageing in glass tube. On the base results of VST and from conclusions of
publication [2,3] followed that the samples aged in glass tube were more influenced by oxygen
in tube however the loading density was same as in original cartridge.
The changes volume of gas evolved werent significant and werent observed any
decrease of chemical stability during ageing in the case of propellant D3. There was no
significant increase volume of gas evolved in the cases artificial ageing in the higher
temperature, sealed bag and lowest temperature, glass tube. The samples which were ageing
in glass tube werent influenced by oxygen as the glasses were completely filled.
3.2
Result of DTA
From the results of DTA followed that the thermal stability wasnt changed for all
ageing temperature, ageing period and type of packaging.
3.3
Results of the stabilizer depletion investigation and the calculation

of chemical life
The results of the measurements and the calculations are included in tables 3 and 4
Table 3. Results of the stabilizer depletion investigation and the calculation of
chemical life for the propellant S6, a = 1,0659 %
AT
65
60
50
AP
13
26
39
65
24
48
72
123
60
120
180
301
NATO tube
a-x
1,0563
0,8608
0,8293
0,7218
1,0552
0,9605
0,8090
0,5728
0,9153
0,9013
0,7722
0,5527
ln a/a-x
0,0091
0,2137
0,2510
0,3898
0,0101
0,1041
0,2758
0,6210
0,1523
0,1677
0,3223
0,6568
-4
ln k
-4,8954
-5,0536
-6,0565
Sealed bag
a-x
1,0173
0,9874
0,9182
0,8797
1,0645
0,9580
0,9226
0,8376
1,0422
0,9634
0,9129
0,7611
ln a/a-x
0,0467
0,0765
0,1492
0,1920
0,0013
0,1067
0,1444
0,2410
0,0023
0,1011
0,1515
0,3368
-4
Orig. cartridge
ln k
-5,7813
-6,0310
-6,6418
k25
t25
2,473 . 10
2,539 . 10
2802,9 days = 7,68 years 2730 days = 7,48 years
Ea
73 445,2 J.mol-1
52 592,1 J.mol-1
553
a-x
1,0451
1,0257
0,9349
0,7601
0,9960
0,9080
0,8638
0,7319
1,0338
0,9663
0,8392
0,6481
ln a/a-x
0,0197
0,0384
0,1311
0,3381
0,0678
0,1603
0,2102
0,3759
0,0306
0,0981
0,2391
0,4975
-4
ln k
-5,0018
-5,7886
-6,2048
2,264 . 10
3061,6 days = 8,39
years
67 445,6 J.mol-1
Table 4. Results of the stabilizer depletion investigation and the calculation of

chemical life for the propellant D3, a = 1,4529 %
AT
65
60
50
AP
13
26
39
65
24
48
72
123
60
120
180
301
NATO tube
a-x
1,3537
1,3509
1,2724
1,1741
1,3786
1,3041
1,2346
1,1490
1,3700
1,2912
1,1697
1,0373
ln a/a-x
0,0707
0,0728
0,1327
0,2131
0,0525
0,1080
0,1628
0,2347
0,0588
0,1180
0,2168
0,3369
Sealed bag
ln k
-5,7769
-6,2892
-6,7345
k25
t25
2,030 . 10-4
3414,5 days = 9,35 years
Ea
55 273,47 J.mol-1
a-x
1,4409
1,3676
1,3021
1,2497
1,3230
1,3097
1,2766
1,1632
1,3457
1,2479
1,2050
1,0823
ln a/a-x
0,0083
0,0605
0,1096
0,1507
0,0937
0,1038
0,1294
0,2224
0,0766
0,1521
0,1871
0,2945
ln k
-5,8591
-6,5520
-7,0295
1,574 . 10-4
4403,73 days = 12,07 years
57 450,98 J.mol-1
Commentary to tables 3 and 4

a initial content of the effective stabilizer (%)
x stabilizer depletion (%)
a x content of effective stabilizer during ageing (%)
k rate constant of stabilizer depletion at ageing temperature
k25 rate constant of stabilizer depletion at storage temperature 25C
t25 time needed for the decrease of the stabilizer content to 50% of initial value at the storage
temperature at 25C
- propellant chemical life
Ea - Activation Energy for the stabilizer consumption process
AT Ageing Temperature
AP Ageing Period
By the kinetic investigation of stabilizer depletion during the artificial ageing the
chemical life of propellants was determined. The chemical life for propellant S6 aged in
original cartridge was determined 8,4 years, and for aged in NATO tube and sealed bag the
value was determined near lower. This result was expected and supports recommendation to
make the artificial ageing of propellants in the original ammunition application if it is
practically and possible. The oxygen which was presented in the tube and foil accelerated
measurably the stabilizer decomposition. From result we can recommend the sealed foil bag
as a cover for the artificial ageing of propellants which are stabilized with CI if it is not
possible to use the original cartridge.
554
4.
SUMARY
In term of the VST results the sealed bag is preferable as a cover for artificial ageing of
propellants than tube. It is possible to use all three tested cover, but in case of NATO tube or
foil these covers have to be properly sealed, so as the propellant sample could be aged
without the entrance of air. The tube has to be completely filled by sample. It is very
important to do ageing studies in complete ammunition articles or, at least, under storage
conditions as closely as possible related to the conditions in ammunition. In term of ageing
temperatures the range 50 65C is relatively acceptable, whereas plots the rate constants
versus ageing temperature gave good linear relations.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
J. SKLDAL: Sledovn rozkladu bezdmnch prach pomoc HPLC, Proceedings of the 2-nd
Seminar New Trends in Research of Energetic Materials, University of Pardubice, p.258,
Apr. 1999
J. PETRLEK, J. SKLDAL: Vliv podmnek urychlenho strnut na reakce difenylaminu
v bezdmnm prachu, Proceedings of the 3-th Seminar New Trends in Research of Energetic
Materials, University of Pardubice, p. 148, Apr.2000
J. PETRLEK, J. SKLDAL: Vliv podmnek urychlenho strnut na reakce centralitu I
v bezdmnm prachu, Proceedings of the 3-th Seminar New Trends in Research of Energetic
Materials, University of Pardubice, p. 187, Apr.2000
P. GUILLAUME, M. RAT, S. WILKER, G. PANTEL: Int. Annu. Conf. ICT 1998, p. 133
S. WILKER, P. GUILLAUME, M. RAT. Second Workshop on Thermal Analysis and Stability of
Popellants, WIWEB, SRN 1999, p. 19
S. WILKER, J. PTERLEK, J. SKLDAL, U. TICMANIS, G. PANTEL: Stability Analyses of Double
Base Propellants in Dependence of their DPA and NGL Content, Proceedings of the 4-th
Seminar New Trends in Research of Energetic Materials, University of Pardubice, p. 369,
Apr.2001
S. WILKER, G. PANTEL, L. STOTTMEISTER: Stability Ananlyses of Porous Propellants,
Proceedings of the 6-th Seminar New Trends in Research of Energetic Materials, University
of Pardubice, p. 506, Apr.2003
M. CHOVANCOV, P. OKO, M. LAZAR, A. PECHOV: Vyuitie kvapalinovej chromatografie
pre skky vbunn, Zbornk prspevkov 4. pyrotechnickej konferencie s medzinrodnou
asou, Zemplnska rava 2003, p. 83
STANAG 4620
AOP-48
STANAG 4556
STANAG 4582
TP-SHVV-1080-(05)-00-ZH
TP-SHVV-1080-(08)-00-ZH
TP-SHVV-1080-(12)-00-ZH
555
70
y = 322,76x - 3,7329
R2 = 0,9486
ageing time (date)
60
50
40
30
20
10
0,05
0,07
0,09
0,11
0,13
0,15
0,17
0,19
0,21
0,23
ln(a/(a-x))
Fig 1.
Stabilizer depletion ln(a/(a-x)) for propellant D3, ageing temperature

65C, glass tube
130
120
110
y = 538,73x - 8,4028
R2 = 0,9817
ageing time (date)
100
90
80
70
60
50
40
30
20
0,025
0,075
0,125
0,175
0,225
ln(a/(a-x))
Fig 2.

60C, glass tube
556
ageing time (date)
315
295
275
255
235
215
195
175
155
135
115
95
75
55
0,05
y = 840,96x + 11,671
2
R = 0,9901
0,1
0,15
0,2
0,25
0,3
0,35
ln(a/(a-x))
Fig 3.

50C, glass tube
75
65
ageing time (date)
55
y = 350,41x + 6,9245
R2 = 0,9442
45
35
25
15
5
0
0,02
0,04
0,06
0,08
0,1
0,12
0,14
0,16
ln(a/(a-x))
Fig 4.

65C, sealed bag
557
125
110
y = 700,67x - 29,463
R2 = 0,9443
ageing time (date)
95
80
65
50
35
20
0,08
0,1
0,12
0,14
0,16
0,18
0,2
0,22
0,24
ln(a/(a-x))
Fig 5.

60C, sealed bag
350
300
y = 1129,5x - 35,319
R2 = 0,9881
ageing time (date)
250
200
150
100
50
0
0,05
0,1
0,15
0,2
0,25
0,3
ln (a/(a-x))
Fig 6.

50C, sealed bag
558
T-1(K)
0,00295
-5,5
0,003
0,00305
0,0031
y = -6648x + 13,799
R2 = 0,9416
lnk
-6
-6,5
-7
Fig 7.
Rate constant for propellant D3, glass tube
T-1(K)
0,00295
-5,8
0,00305
y = -6909,9x + 14,423
R2 = 0,8477
-6,1
lnk
0,003
-6,4
-6,7
-7
Fig 8.
Rate constant for propellant D3, sealed bag
559
0,0031
70
60
y = 133,68x + 6,8896
R2 = 0,8975
ageing time (date)
50
40
30
20
10
0
0
0,05
0,1
0,15
0,2
0,25
0,3
0,35
0,4
0,45
ln(a/(a-x))
Fig 9.
Stabilizer depletion ln(a/(a-x)) for propellant S6, ageing temperature

65C, glass tube
125
110
y = 156,58x + 27,175
R2 = 0,9911
ageing time (date)
95
80
65
50
35
20
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
ln(a/(a-x))
Fig 10. Stabilizer depletion ln(a/(a-x)) for propellant S6, ageing temperature
60C, glass tube
560
ageing time (date)
300
y = 426,9x + 26,603
R2 = 0,9446
250
200
150
100
50
0,1
0,2
0,3
0,4
0,5
0,6
0,7
ln(a/(a-x))
50C, glass tube
70
y = 361,28x - 4,0004
R2 = 0,9971
ageing time (date)
60
50
40
30
20
10
0,04
0,06
0,08
0,1
0,12
0,14
0,16
0,18
0,2
ln(a/(a-x))
65C, sealed bag
561
115
ageing time (date)
100
y = 416,13x + 15,422
R2 = 0,9497
85
70
55
40
25
10
0
0,05
0,1
0,15
0,2
0,25
ln(a/(a-x))
60C, sealed bag
300
ageing time (date)
250
y = 766,46x + 48,003
R2 = 0,9894
200
150
100
50
0
0,05
0,1
0,15
0,2
0,25
0,3
0,35
0,4
ln (a/(a-x))
50C, sealed bag
562
70
y = 148,68x + 16,151
R2 = 0,9544
ageing time (date)
60
50
40
30
20
10
0
0,04
0,08
0,12
0,16
0,2
0,24
0,28
0,32
ln(a/(a-x))
65C, original cartridge
125
ageing time (date)
110
y = 326,57x + 0,2772
R2 = 0,9938
95
80
65
50
35
20
0,05
0,1
0,15
0,2
0,25
0,3
0,35
0,4
ln(a/(a-x))
563
290
y = 495,14x + 58,138
R2 = 0,9881
ageing time (date)
240
190
140
90
40
0,025
0,075
0,125
0,175
0,225
0,275
0,325
0,375
0,425
0,475
ln(a/(a-x))
T-1(K)
0,00295
0,003
0,00305
-4,6
y = -8833,6x + 21,328
R2 = 0,9628
lnk
-5,1
-5,6
-6,1
Fig 18. Rate constant for propellant D3, glass tube
564
0,0031
T-1(K)
0,00295
-5,8
0,003
0,00305
0,0031
y = -5657,1x + 10,888
R2 = 0,9799
lnk
-6,1
-6,4
-6,7
Fig 19. Rate constant for propellant S6, sealed bag
T-1(K)
0,00295
0,003
0,00305
-5,05
-5,25
y = -8112,6x + 18,821
R2 = 0,8627
lnk
-5,45
-5,65
-5,85
-6,05
-6,25
Fig 20. Rate constant for propellant D3, original cartridge
565
0,0031
IMPROVED SYNTHESIS
OF 2-(DINITROMETHYLENE)-4,5-IMIDAZOLIDINEDIONE
Zdenk Jalov*, Pavel Mareek**, Kamil Dudek**, Ondej Fohl*,
Nikolaj V. Latypov***, Stefan Ek***, and Martin Johansson***
* University of Pardubice, Department of Theory and Technology of Explosives, Studentsk 95,
CZ-532 10 Pardubice, Czech Republic
** Explosia, a.s., CZ-532 17 Pardubice-Semtn, Czech Republic
*** FOI, Swedish Defence Research Agency, Department of Energetic Materials,
S-147 25 Tumba, Sweden
Abstract:
Acetamidine sulphate and acetamidine hydrogensulphate were prepared
from acetamidine hydrochloride by two different methods. Both acetamidine sulphate
and hydrogensulphate were used for condensation reaction with diethyloxalate to
produce 2-methoxy-2-methyl-4,5-imidazolidinedione, which was subsequently nitrated
to 2-(dinitromethylene)-4,5-imidazolidinedione. The use of the mentioned acetamidines
is convenient, since Soxhlet extraction can be removed from the reaction process and
the yield is higher than in the case of acetamidine hydrochloride.
Keywords:
1.
synthesis, acetamidine sulphate, acetamidine hydrogensulphate,

2-methoxy-2-methyl-4,5-imidazolidinedione, 2-(dinitromethylene)4,5-imidazolidinedione
INTRODUCTION
The method for synthesis of 2-methoxy-2-methyl-4,5-imidazolidinedione starting

from acetamidine hydrochloride and diethyloxalate was published in 1998, by Latypov
et al. [1]. The reaction medium was sodium methoxide and methanol. The main condensation
product was 2-methoxy-2-methyl-4,5-imidazolidinedione (1), which is soluble in methanol.
The condensation product contained some amount of sodium chloride, which is partially
soluble in methanol. Therefore, it was necessary to use Soxhlet extraction in order to isolate
pure condensation product without inorganic impurities.
Abbreviations in the reaction schemes: Me methyl, Et ethyl, R - methyl or ethyl
H
NH2 Cl
NH2
EtO
MeONa
N
H
MeOH
OEt
NaCl
2-Methoxy-2-methyl-4,5-imidazolidinedione (1) was nitrated to 2-(dinitromethylene)4,5-imidazolidinedione (2) with mixed acids and the nitrated product was subsequently
hydrolysed to 2,2-dinitroethylene-1,1-diamine (FOX-7) [1].
566
H
O
N
N
H
HNO3
H2SO4
N
N
H
NO2
NH3
H2N
NO2
NO2
H2O
H2N
NO2
FOX-7
When acetamidine sulphate (3) or hydrogensulphate (4) are used for the condensation
reaction with diethyloxalate, sodium sulphate forms, instead of sodium chloride in the case
of acetamidine hydrochloride. Sodium sulphate is insoluble in methanol and precipitates
completely from the reaction mixture. Thus, the condensation product is isolated
without inorganic salt and Soxhlet extraction is no longer necessary.
H
EtO
NH2
1/2
NH2
MeONa
SO4
MeOH
OEt
N
N
H
1
Na2SO4
H
NH2 HSO4
NH2
EtO
MeONa
MeOH
OEt
2.
N
N
H
1
Na2SO4

2.1
Synthesis of acetamidine sulphates
Acetamidine hydrochloride is commonly prepared by the Pinner method [2,3], where

acetonitrile and alcohol react in the presence of hydrogen chloride forming imidoester hydrochloride (5), which is then treated with ammonia to produce acetamidine hydrochloride.
C N + R OH
HCl
NH2 Cl
NH3
NH2 Cl
NH2
OR
+ NH4Cl
The analogous reaction of acetonitrile, alcohol and sulphuric acid (instead of hydrogen
chloride) to form imidoester sulphate or hydrogensulphate did not work properly in that
manner [4]. A solid product was obtained, which is under current investigation.
Thus, acetamidine sulphate and hydrogensulphate were prepared from acetamidine
hydrochloride by two different methods.
Acetamidine sulphate was prepared in two steps. The first one is the reaction of
acetamidine hydrochloride with equimolar ratio of sodium ethoxide in ethanol to precipitate
567
sodium chloride. The free base, acetamidine, is then treated with half an equivalent of
sulphuric acid to form acetamidine sulphate (3) as a precipitate.
NH2 Cl
NH2
NH
+
EtONa
NH
2
NH2
H2SO4
NaCl
NH2
NH2
NH2
SO4
2
Acetamidine hydrogensulphate was prepared in a similar manner; equimolar ratios

of acetamidine hydrochloride and sulphuric acid were mixed. This induces hydrogen
chloride evolution and acetamidine hydrogensulphate (4) formation.
NH2 Cl
NH2
+ H2SO4
NH2 HSO4
NH2
HCl (g)
2.2
Synthesis of 2-methoxy-2-methyl-4,5-imidazolidinedione
from the sulphates
Both products were condensed with diethyl oxalate in the same manner as described
in the original paper[1], vide supra. In this case, the yield of the final product, 2-methoxy2-methyl-4,5-imidazolidinedione, was increased from 64 % in the original method up to
92-94 %. Moreover, this method does not require the tedious Soxhlet extraction to remove
inorganic impurities. The product contains 10-15 % of 2-methyl-4,5-imidazolidinedione,
which is nitrated in the same way as the desired 2-methoxy-2-methyl-4,5imidazolidinedione, yielding the same nitrated product.
2.3
Nitration of 2-methoxy-2-methyl-4,5-imidazolidinedione
The product thus formed was nitrated directly without further purification, giving the
desired product in high purity and with a reproducible yield as described earlier (62-64 %)[1].
3.
EXPERIMENTAL SECTION
3.1
Acetamidine sulphate (3)
Sodium (7.2 g, 0.313 mol) was dissolved in 230 ml of absolute ethanol. The solution
was added under stirring to solution of acetamidine hydrochloride (30 g, 0.317 mol)
in 150 ml of absolute ethanol. The precipitated sodium chloride was removed by filtration.
Sulphuric acid (100%, 15.6 g, 0.159 mol) was gradually added to the solution of acetamidine
in ethanol. The temperature was kept below 25oC. The precipitate was filtrated and washed
with ethanol to give 26.8 g (79%) of acetamidine sulphate; m.p. 206-210 C. The product
was crystallised from methanol, m.p. 211-212 C. 1H NMR (D2O) 2.23 (s, CH3), 8.25 bs
(NH) ppm. 13C NMR (D2O) 15.9, 166.3 ppm. Elementary analysis calc. for C4H14N4O4S:
C, 22.42 %; H, 6.58 %; N, 26.16 %; S, 14.96 %. Found: C, 23.16 %; H, 6.27 %; N, 25.98 %;
S, 15.26 %.
568
3.2
Acetamidine hydrogensulphate (4)
Sulphuric acid (100%, 9.8 g, 0.1 mol) was added to acetamidine hydrochloride (9.45 g,
0.1 mol) and the solution was stirred for 30 minutes. Hydrogen chloride evolved. After the
stipulated time, the mixture was concentrated in vacuo, at 35 oC for 75 minutes. The mixture
was cooled and acetamidine hydrogensulphate precipitated. Yield 13.7 g (88%), m.p.
56-58 C. 1H NMR (D2O) 1.97 (s, CH3), 7.90 bs (NH) ppm. 13C NMR (D2O) 16.0, 166.4
ppm.
Elementary analysis calc. for C2H8N2O4S: C, 15.38 %; H, 5.13 %; N, 17.95 %; S,
20.51 %. Found: C, 15.65 %; H, 5.80 %; N, 17.13 %; S, 21.68 %.
3.3
2-Methoxy-2-methyl-4,5-imidazolidinedione (1)
(from acetamidine hydrogensulphate)
Acetamidine hydrogensulphate (8.1 g, 0.052 mol) was dissolved in 120 ml of methanol.

Sodium methoxide (39 g of 30% solution of MeONa in methanol, 0.217 mol) was added and
sodium sulphate precipitated. Diethyloxalate (7.6 g, 0.052 mol) in 100 ml of methanol was
added to the solution of acetamidine with stirring under temperature up to 10 oC for 3 hours.
Than the reaction mixture was stirred for 1 hour at 15oC. The pH of the mixture was adjusted
to 5,5 with sulphuric acid. Sodium sulphate was removed by filtration and methanolic
solution was evaporated in vacuum rotary evaporator. The yield of 2-methoxy-2-methyl-4,5imidazolidinedione was 7 g (93%), m.p. 153-157 C (lit. [1], 158 C). According to 1H NMR
spectra, the product is 2-methoxy-2-methyl-4,5-imidazolidinedione, with a variable amount
of 2-methyl-4,5-imidazoledione. 1H NMR (DMSO-d6) 1.56 s (CH3), 2.97 s (O-CH3), 9.98
bs (NH) ppm, 13C NMR (DMSO-d6) 27.2, 48.2, 92,0, 160.2.
3.4
2-Methoxy-2-methyl-4,5-imidazolidinedione (1)
(from acetamidine sulphate)
Acetamidine sulphate (5.4 g; 0.0252 mol) was suspended in 200 ml of methanol and the
mixture was heated to dissolve acetamidine sulphate. Sodium methoxide (30 g of 30%
solution in methanol, 0.167 mol) was added and the mixture was immediately cooled
to ambient temperature. Diethyloxalate (7.6 g, 0.052 mol) in 100 ml of methanol was added
to the solution of acetamidine with stirring over three hours. The temperature was kept
below 10 oC. The reaction mixture was stirred for 1 hour at 15 oC. The mixture was acidified
with sulphuric acid to pH 5.5. Sodium sulphate was filtered off and the methanolic solution
was concentrated in vacuo to give 6.7 g (92%) of 2-methoxy-2-methyl-4,5-imidazolidinedione, m.p. 153-160 C (lit. [1], 158 C). According to 1H NMR spectra, the product is
2-methoxy-2-methyl-4,5-imidazolidinedione, with a variable amount of 2-methyl-4,5-imidazolidinedione. 1H NMR (DMSO-d6) 1.56 s (CH3), 2.97 s (O-CH3) 9.98 bs (NH) ppm,
13
C NMR (DMSO-d6) 27.2, 48.2, 92,0, 160.2.
3.5
Nitration of the dione from 3.3
The dione (4.3 g, 0.030 mol) was dissolved in sulphuric acid (d 1.83, 18 ml) and nitric
acid (d 1.5, 3.3 ml, 0.015 mol) was added dropwise during 5 minutes. The temperature was
kept below 10 C by external cooling. After 10 minutes 2-(dinitromethylene)-4,5imidazolidinedione precipitated, which was removed by filtration, washed with
trifluoroacetic acid and dried under vacuum to give 3.8 g (63%) of 2-(dinitromethylene)-4,5imidazolidinedione (2). 1H NMR (DMSO-d6) 11.01 ppm, 13C NMR (DMSO-d6) 131.1,
153.5, 159.8 ppm.
569
3.6
Nitration of the dione from 3.4
The dione (4.3 g, 0.030 mol) was dissolved in sulphuric acid (d 1.83, 18 ml) and nitric
acid (d 1.5, 3.3 ml, 0.015 mol) was added dropwise during 5 minutes, the temperature was
kept below 10 C by external cooling. After 10 minutes a precipitate of 2(dinitromethylene)-4,5-imidazolidinedione was formed, which was removed by filtration,
washed with trifluoroacetic acid and dried under vacuum to give 3.9 g (65 %) of 2(dinitromethylene)-4,5-imidazolidinedione (2). 1H NMR (DMSO-d6) 11.01 ppm, 13C
NMR (DMSO-d6) 131.1, 153.5, 159.8 ppm.
4.
CONCLUSIONS
New salts of acetamidine were prepapared from acetamidine hydrochloride.
Acetamidine sulphate is a stable compound with melting point of 211-212 C.
Acetamidine hydrogensulphate is very hygroscopic at normal conditions.
When acetamidine sulphate or hydrogensulphate are used for condensation

reaction with diethyloxalate, 2-methoxy-2-methyl-4,5-imidazolidinedione was
prepared in high yield and purity.
All inorganic impurities are removed by filtration, thus avoiding the tedious use
of Soxhlet extraction
The use of acetamidine sulphate or hydrogen sulphate increases the overall yield
of FOX-7 synthesis starting from acetamidine.
Acknowledgements
The authors would like to thank to Tom Weidlich (VOS, a.s., Czech republic) for
NMR measurements.
REFERENCES
[1]
[2]
[3]
[4]
N.V. LATYPOV, J. BERGMAN, A. LANGLET, U. WELLMAR and U. BEMM: Tetrahedron 54,

11525-11536, 1998
A. PINNER: Die Imidoaether and ihre Derivative, Berlin, 1892
A.W. DOX: Org. Syntheses, Coll. vol. 1, 5-7, 1932
Z. JALOV and J. OTTIS: unpublished results, University of Pardubice, 2003
570
ANALYSIS OF NITROAROMATIC COMPOUNDS

IN SOIL ENVIRONMENTS
Vra Jeov, Ale Eisner, Jana Chalnkov, and Karel Ventura
Department of Analytical Chemistry, University of Pardubice, 532 10 Pardubice, Czech Republic
Abstract:
Polynitro organic explosives are typical labile environmental pollutants that can
biotransform with soil indigenous micro-organisms, photodegrade by sunlight and
migrate through subsurface to groundwater contamination. To be able to determine
nitroaromatic compounds and their (bio)transformation products in soil, a
comprehensive analytical methodology of sample preparation, separation and detection
is required. At first were taken different soil samples contaminated with nitroaromatic
compounds and then they were analysed. Modern extraction techniques were tested for
the determination pressurised fluid extraction (PFE), supercritical fluid extraction
(SFE) and ultrasonic extraction. Gas chromatography with mass detection (GC-MS)
was used for the separation and detection.
Keywords:
1.
Explosives, Pressurised fluid extraction, Ultrasonic extraction, GC/MS
INTRODUCTION
Explosives are labile and in the environment can be transformed by sunlight, soil
microflora and other indigenous plant species and that can migrate through subsurface soil to
cause groundwater contamination.
The contamination of soil and water samples with different explosives, such as 2,4,6trinitrotoluene (TNT) and their degradation products are widespread and often caused by
military activities. TNT is one of the most commonly used highly energetic compounds that
have toxic effects on a number of species including humans. The biodegradation and
photolytical by-products of TNT introduce highly reactive compounds often greater polarity
and water solubility, which may constitute an even environmental concern than TNT itself.
The contaminated soils by TNT can contain different products of TNT
(bio)transformation. The formation of these compounds follows the reduction of TNT to its
associated hydroxylamino- and amino-nitrotoluene intermediates. Soil bacteria in either
aerobic or anaerobic conditions can form the amino-nitrotoluene intermediates and they
present the most commonly detected products of degradation of TNT.
Different extraction techniques are applied for the determination of TNT and their
degradation products from various soil samples. Widely used is the extraction in
ultrasonicate bath. In many studies were employed the plant tissues samples grown in
contaminated soils [1]. The supercritical fluid extraction (SFE) was applied for the tissues
samples and it were compared dynamic and static SFE. In the same study the contaminated
soil samples by TNT were extracted in the ultrasonicate bath with acetonitrile and analysed
by HPLC-UV. TNT and two of its monoamino derivates were found (2-amno-dinitrotoluene
571
and 4-amno-dinitrotoluene). The concentration of these compounds is depended on the depth

sampling and is rapidly decreased with increased depth.
Gas chromatography with an electron capture detector (GC-ECD) in the combination
with ultrasonic extraction was applied for the determination of nitroaromatic and nitramine
explosives in soil [2]. The soil was extracted with acetonitrile for 18 hours in a sonic bath.
Each extract was filtered prior the injection into the GC. TNT was detected in all samples
ranging from 1.3 to 273 g kg -1.
This study describes the use of two extraction techniques - extraction in ultrasonicate
bath and pressurised fluid extraction (PFE) for the determination of TNT and its degradation
products in soil samples. The obtained extracted were separated and identified by GC/MS.
Pressurised fluid extraction (PFE) is relatively new extraction techniques such as
supercritical fluid extraction [3]. In principle PFE is an extraction process taking place in the
solid / liquid phase system and being performed at a increased temperature (50-200C),
increased pressure (5-20 MPa) and within a relatively short time interval (max. 20 min.)
In order to perform PFE, the same solvents as for Soxhlet extraction can be used. More
over, it is possible to choose even such solvents that are less effective when used in classical
techniques. This is due to specially enhanced extraction kinetics of the extraction process in
ASE. Very advantageous is that extraction work with ASE can be made under conditions of
liquid phase. This allows one to use even mixed solvents, as there is no risk of distilling one
component separately.
2.
EXPERIMENTAL
It were took away eight samples (about 50g) of the soil contaminated by TNT. Each
sample was extracted by ultrasonic extraction.
The sample of the soil (about 10g) was extracted to 30ml of acetonitrile in ultrasonicate
bath for 15 minutes. The obtained extracts were filtered and most of the solvent was
evaporated to volume 2 ml.
The obtained extracts were analysed by using of the gas chromatograph GC 17A
coupled with mass spectrometry detector QP 5050A (EI, NCI, both Shimadzu) and GC/MS
solution data system (Shimadzu). The helium (grade 5.0, Linde) was used as carrier gas.
Separations were performed on a 30m x 25m i.d. capillary column coated with a 0,25 m
film of polymethylsiloxane (DB-5 MS). The electronic ionization in SCAN mode and the
split injection 1:150 were used. The volume of dosed extract was 1 l of the evaporated
sample. The column oven was isothermally maintained at 50C for 3 minutes and than
temperature increase for 20C/min to 220C that was kept for 3 minutes. The temperature of
injector was 220C and the temperature of interface was 230C. The identification of
compounds was based on the comparison their mass spectrum with the spectrum in the
library (NIST 62 and NIST 12, Shimadzu).
572
3.
Eight soil samples contaminated by TNT were extracted to the acetonitrile in the
ultrasonicate bath. The filtered and evaporated extracts were analysed with gas
chromatography with mass detector. TNT was detected in all soil samples in the different
amount (Fig 1).
40000000
Intensity
30000000
20000000
10000000
0
1
Soil samples
Fig 1.
The comparison of soil samples extracted in the sonication bath
Different degradation products were identified with mass detector. A mass spectrum in
the peak was compared with library or with the measured standard. Some of the peaks were
identified, but any peaks were not identified with the library. DNTs and amino-DNTs were
found in the most samples and dibuthylphalate was identified in some of samples.
The chromatogram of 1st soil sample is shown on the figure 2. This soil sample has
included the highest amount of TNT and following degradations products: 2,6dinitrotoluene, 2,4-dinitrotoluene, 3,4-dinitrotoluene, 6-amino-2,4-dinitrotoluene, 4-amino2,6-dinitrotoluene, methyl-dinitrofenol.
TIC
300e3
TNT
250e3
200e3
2,6-DNT
150e3
2,4-DNT
3,4-DNT
amino-DNT
amino-DNT
methyl-DNP
100e3
50e3
3
Fig 2.
The chromatogram of 1st soil sample
573
10
11
2,4,6-trinitrotoluene was quantified in soil samples number 1, 4, 5 and 8. TNT was

determined in the concentration of 0,8 to 0,075 % in these four samples.
4.
CONCLUSION
This study describes the determination of degradation products of 2,4,6-trinitroloulene in

soil samples. The combination ultrasonic extraction with gas chromatography-mass
spectroscopy was applied for the determination and identification of TNT. Some degradation
products of TNT were determined, e. g. 2,6-DNT, 3,4-DNT, 2,4-DNT and amino-DNT. TNT
was found in soil samples in the concentration between 0,8 % and 0,075 %.
Acknowledgement:
Youth and Sports of the Czech Republic (Project MSM0021627502).
REFERENCES
[1]
[2]
[3]
A. HALASZ, C. GROOM, E. ZHOU, L. PAQUET, C. BEAULIEU, S. DESCHAMPS,

A. CORRIVEAU, S. THIBOUTOT, G. AMPLEMAN, C. DUBOIS, J. HAWARI: Detection
of explosives and their degradation products in soil environments, Journal of
Chromatography A, 963, 411-418, 2002.
M. E. WALSH: Determination of nitroaromatic, nitrateamine and nitrate ester explosives in
soil by gas chromatography and an electron capture detector, Talanta 54, 427-438, 2001.
J. R. DEAN: Extraction Methods for Environmental Analysis, John Wiley & Sons,
Chichester 1998.
574

T. M. Klaptke*, J. J. Weigand*, and G. Holl**
Swisttal-Heimerzheim, Groes Cent, D-53913 Swisttal, Germany
Abstract:
The reaction of 1,5-diamino-1H-tetrazole (1) with iodomethane followed by the
methathesis of the iodide (2) with silver dinitramide yields the highly friction sensitive
1,5-diamino-4-methyl-1H-tetrazolium dinitramide (3). 3 crystallize in the orthorhombic
space groups P212121 and shows an intriguing interaction of one nitro group of the
dinitramide anion with the tetrazolium cation. Preliminary sensitivity testing of the
crystalline compound 3 indicates rather low impact sensitivity with a value of 7 J in
contrast to the relatively high friction sensitivity of 24 N. The thermal decomposition of
3 was investigated by thermogravimetric analysis (TGA) and differential scanning
calorimetry (DSC). Mass spectrometry and IR spectroscopy were used to identify the
gaseous decomposition products.
Keywords:
1.
tetrazoles, DAT, dinitramide, thermal decomposition, TGA, DSC
INTRODUCTION
Since the generation of molecular nitrogen as an end-product of a propulsion or

explosion is highly desired in order to avoid environmental pollution, health risks as well as
the reduction of detectable plume signatures, compounds containing a backbone of directly
linked nitrogen atoms (nitrogen catenation) are of great interest [1]. The ability of the
dinitramide (DN) anion to form stable oxygen-rich salts with high densities with a variety of
cations [2] makes DN a promising candidate in the development of new tetrazole based
oxidizers. Such salts would be expected to possess both a high nitrogen and a high oxygen
content. The combination of the oxygen-rich anion (DN) with a nitrogen-rich cation forms a
class of energetic materials whose energy is derived from their very high positive heat of
formation (directly attributed to the large number of inherently energetic NN, NO and N
C bonds) as well as the combustion of the carbon backbone.
Here we wish to report (i) the synthesis of the highly friction sensitive 1,5-diamino-4methyl-1H-tetrazolium dinitramide (3), (ii) the X-ray determination with respect to an
intriguing interaction of one nitro group with the tetrazolium cation and (iii) a study on the
thermal decomposition behavior of 3 [3].
575
2.
SYNTHESIS
Scheme 1.
The two step synthesis of 3 starts from 1,5-diamino-1H-tetrazole (DAT, 1) [4a] via
quaternization of 1 with MeI in acetonitrile (2, 86%). The metathetical reaction of 2 with a
freshly prepared solution of AgDN [5] yields 3 as light yellow crystals suitable for X-ray
structure determination (85%) (Scheme 1).
3.
CRYSTAL STRUCTURE
Fig 1.
View of the strand formed by alternating cations (MeDAT) and anions (DN)
displayed along the a-axis in 3. Short contacts between the cations and anions
are shown by dotted (, N5H5bO2) and dashed (- - - , O(dinitramide) and
tetrazole moiety) lines, respectively. (ORTEP Plot, thermal ellipsoid represents
50% probability).
3 crystallizes in the orthorhombic space group P212121 with four formula units in the
unit cell as racemic twin. The structure contains one crystallographically independent cation
and one anion. As the methyl group hardly influences the molecular parameter of the
tetrazole moiety (MeDAT) we abstain from a detailed discussion. (cf. X-ray data of 1,5diamino-1H-tetrazolium cation see [4b]). The overall geometry of the dinitramide ion is
similar to that observed for other comparable dinitramide salts, like guanidinium and
biguanidinium salts [2]. The two NN bond lengths are asymmetric (N7N8 1.350(3) ; N8
N9 1.391(3) ; cf. a typical N=N double bond (1.245 ); NN single bond (1.454 )) [6]
576
and a NNN angles of 115.4(2) was observed. The nitro groups are twisted out of the central
NNN plane (O3N9N7O1 -24.5(2)). Hence, the local symmetry of the anion is C1. As
known also form others [2] the metrical parameters of the DN ion are easily and strongly
influenced by the environment.
In Fig. 1 the view along the a-axis is displayed showing the zigzag strand composed of
alternating MeDAT and DN ions. Each strand is further connected by strong hydrogen bonds
(N5H5aO1i and N5H5aO2i [symmetry code: (i) -1+x, y, z]) forming a threedimensional network. In turn, these strands are formed by two different type of interactions:
(1) a strong hydrogen bond between N5H5bO2ii (symmetry code: (ii) -0.5+x, 0.5-y, 1-z)
and (2) a closed-shell interaction between the O atoms of one nitro group (O3-N9-O4) with
the tetrazole moiety.
The interionic distances for the O4 and O3 contacts with the tetrazole ring are in the
range of 3.020(3) to 3.278(3) . For closed-shell interactions there is a somewhat arbitrary
dividing line between what is or is not an interaction when only the sum of the van der
Waals radii are applied as criterion. Since the MeDAT is a non-spheric ion which is also
able to form hydrogen bonds it is hard to compare the anion-cation interactions with those in
alkali DN salts which are observed in the range of 2.810(2) (K+[N3O4]-) to 3.534(6)
(Cs+[N3O4]-) [2b].
4.
DECOMPOSITION EXPERIMENTS
3 belongs to the new class of energetic ionic liquids related to the low melting salts of
1,2,4- and 1,2,3-triazole systems, introduced by Drake et al [7]. The thermal stability was
initially evaluated using differential scanning calorimetry (DCS) and thermogravimetric
analysis (TGA). For initial safety testing the impact and friction sensitivity was tested
according to BAM methods with the BAM Fallhammer and BAM friction tester [8]. The
energy of activation for the decomposition step was estimated by the method of Ozawa and
Kissinger [9] by following the differential heating rate method of the American Society for
Testing and Materials (ASTM) according to the ASTM protocol E 698 99 [10].
In order to analyse the gases from the stepwise decomposition of 3, a specially equipped
IR-cell was loaded with the compounds (~ 2 mg) and evacuated. The sample holder of the IR
cell was heated at a rate of 4C/min (CARBOLITE 900C Tube Furnace type MTF 9/15)
and the reaction products were allowed to expand continuously into the gas cell. During this
heating, the IR-spectra were recorded continuously as a function of the heating rate using a
Perkin-Elmer Spektrum One FT-IR instrument. To record the mass spectra, a sample (~
1mg) of the 3 was heated at a heating rate of 4C/min (CARBOLITE 900C Tube Furnace
type MTF 9/15) in a one side closed glass tube (length: 500 mm; diameter: 5 mm)
connected to the reservoir of the mass spectrometer (Jeol MStation JMS 700). In this case,
the spectra were also recorded as a function of the heating rate. Residual decomposition
products (e.g. sublimed material and residue after the decomposition) were identified by
means of mass spectrometry (EI and DEI mode) and NMR (1H, 13C) techniques.
Preliminary sensitivity testing of the crystalline compound 3 indicates rather low impact
sensitivity with a value of 7 J in contrast to the relatively high friction sensitivity of 24 N.
The detonation pressures of 3 was calculated to be P = 33.6 GPa (comparable to RDX [11], P
= 34.4 GPa) and a detonation velocities of D = 8827 m s-1 (comparable to RDX [12], D =
8750 m s-1) was estimated [13]. The oxygen balance of 3 shows an reasonable value of -25.3%
and heat of formation was determined to be positive with a value of 92.1 kcal mol-1.
577
DSC and TGA were used to evaluate the relative thermal stabilities of 3 and are
measured under comparable condition (Al-containers with a hole (1m) on the top for gas
release for DSC measurements). Fig. 2 shows the obtained DSC and TGA thermograph of 3.
Fig 2.
DSC and TGA thermographs of 3 ( = 10C/min).
3 shows an relatively low melting point (85C (onset)) and start to decompose in the
range of 150 230C. The enthalpy of melting, m H , were found to be 118 J g-1. The
DSC curves of the investigated 3 is divided into three parts, the first one corresponds to a
phase transition region which appears as one small endothermic reaction (Tg = 82 C), the
second part shortly after, is related to the melting region of 3 as a endothermic peak. The
third exothermic peak corresponds to the decomposition reaction. The purity of 3 was
determined to be higher than 98% by the evaluation of the corresponding melting peaks with
the vant Hoff equation according ASTM protocol E 928 96 [14]. The activation energy was
estimated to be 32.89 kcal mol-1 (Ozawa) and 32.91 kcal mol-1 (Kissinger), respectively, i.e.,
in close agreement.
3 does not decompose residual free in the temperature range of 50 400C, as the mass
loss was determined by TGA to be ~ 90%. The reason for this is found in the formation of
condensation products with higher molecular mass, e.g. melam, melem or melom [15]. These
products decompose at higher temperatures (> 500C) to volatile substances like cyanamide,
hydrogen cyanide and ammonia [16]. According to the mass loss derived from the TGA
experiment, the dinitramide salt 3 decomposes in one single step.
578
Fig 3.
Infrared spectroscopic evolved gas analysis of 3.
3 shows a relatively stable melt up to ~ 150C where short after, also in accordance with
the DSC experiment, decomposition occurs and a broad exothermic peak in the DSC and a
sudden rise of gaseous products in the IR spectra indicates a spontaneous decomposition in
one step. The thermal decomposition of 3 under reduced pressure in the temperature range
from 40 to 220C is depicted in Fig. 3. Typical products observed during the decomposition
of 3 are N2O (6) as the main product, MeN3 (14), MeONO2 (9), 1,3,5-triazine (13), HCN,
NH3 and H2O (Fig. 4, Table 1).
579
Fig 4.
Gas phase IR spetrum of the decomposition products of 3.
The thermal decomposition is initiated by a proton transfer to subsequently produce

dinitraminic acid HN3O4 (5) in the melt. According to [23], 5 decomposes under formation of
N2O (6) and HNO3 (7) which produce 8 from the recombination of 4 and HNO3 (7). This
reaction process is shown in Scheme 2 (I). No evidence, whether from the explosion
experiments nor the decomposition experiments, was found for the formation of DAT (1)
(Scheme 2, II) and methyldinitramide (18).
Table 1. Vibrational frequencies (cm-1) of the experimentally observed molecules
in the gas phase.
Species
MeN3
NH3
HCN
MeONO2
N2O
H2O
Frequencies
ref
2940 (m), 2818 (w), 2539 (vw), 2198 (s), 2104 (vs), 1466 (vw), 1450 (vw), [17]
1284 (s), 920 (w), 676 (w)
[18]
3336 (m), 1626 (s), 968 (vs), 933 (vs)
[19]
3311 (s), 2097 (w), 712 (vs)
2959 (m), 2917 (m), 1678 (vs), 1661 (vs), 1442 (m), 1430 (m), 1296 (s), [20]
1287 (s), 1278 (s), 1017 (s), 862 (s), 855 (s), 853 (s), 844 (s) 758 (m), 658
(m)
3891 (w), 3480 (m), 2809 (w), 2591 (m), 2488 (m), 2457 (vs), 2217 (vs), [21,22]
1890 (w),1302 (vs), 1275 (vs), 1183 (m), 1155 (m), 694 (w), 588 (w)
[18]
3657 (s), 1595 (s)
Interestingly, during the decomposition of 3, the formation of NO2 was not observed
which excludes the favoured mechanism of decomposition of HN3O4 (5) to NO2 (15) and
HNNO2 (16) according [24] at higher temperatures (Scheme 2, V). The main decomposition
process of in situ formed 8 involves a methyl group transfer to 1 and methyl nitrate (9)
(Scheme 2, III) and not a proton transfer to 4 and HNO3 (2) (Scheme 2, IV). The main
process observed is the formation of 6 which was found to be the major product (Fig. 4) in
580
the IR spectra of the explosion products as well as in the spectra of the decomposition
experiments and could also be identified in the mass spectrometry experiments.
Interestingly, under this reaction condition, the decomposition pathway discussed in [15],
which involves the imino form of DAT (1), undergoes not the decomposition with the
evolution of HN3, but rather eliminates nitrogen from DAT (1) (Scheme 2, III).
Scheme 2.
581
The resulting unstable nitrene (10) decomposes to HCN, NH3 and N2 according Scheme
2 and in a minor process to cyanamide which is less stable and dimerizes to dicyandiamide
that reacts through an intermediate to melamine (11) [25]. The hydrogen cyanide undergoes in
situ cyclotrimerization to the 1,3,5-triazine (12), which is known to proceed under harsh
condition [26] and was also identified in the IR as well as mass spectrum (Fig. 4). Only small
amounts of MeN3 indicates the partly decomposition of 4 according Scheme 2 (IV) resulting
from a proton transfer. Nitric acid and MeN3 were found as components in the mass spectra
of the residual explosion products in small quantities.
Fig 5.
13
C-NMR spectra of decomposition products of 3 recorded in [d6]-DMSO.
1,2,4-Triazole (13), 1-amino-4-methyl-5-imino-4,5-dihydro-1H-tetrazole (4), 1,5diamino-1H-tetrazole (1), 1,3,5-triazine (12) and melamine (11) could be identified by
means of 13C NMR spectroscopy (Fig. 5) and were found in the residue of the
decomposition experiments. Ammonium nitrate (17), resulting from the recombination of
HNO3 and NH3, was identified as one component of the obtained sublimate (Scheme 2, VI)
and explains the low concentration of ammonia. The thermal decomposition of 9 above
200C leads to the formation of CH2O, CH3OH, H2O, CO, NO and NO2 [27] (Scheme 2,
VII). The thermal decomposition of AN during explosion processes leads to further
formation of N2O (6) and H2O (Scheme 2, VIII).
582
5.
CONCLUSION
3 shows an distinctive decomposition pathways which leads to the formation of a series

of products, which are for the major gaseous decomposition products are still highly
endothermic. According the UN guideline ST/SG/AC.10/11 [8a], 3 is not safe for transport
but prilling of 3 might increase the stability toward friction. The predicted detonation
performance of 3 is higher than those found for RDX and the oxygen balance is in a
reasonable limit. Therefore, 3 might find application in liquid monopropellants, similar to
HAN [28] and ADN [29] monopropellants.
Acknowledgements
Financial support of this work by the University of Munich (LMU) and the Fonds der
Chemischen Industrie is gratefully acknowledged (JJW thanks for a FCI scholarship, DO
171/46). The authors are also indebted to and like to thank Mr. Gunnar Spie for the drop
hammer and friction tests. We are also indebted to and thank Mr. Christoph Rohbogner for
his very valuable help concerning the preparation of the used compounds during his
research project.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
(A) DIXON, D. A.; FELLER, D.; CHRISTE, K. O.; WILSON, W. W.; VIJ, A.; VIJ, V.; BROOKE, H.
D. B.; OLSON, R. M.; GORDON, M. S.: J. Am. Chem. Soc. 2004, 126(3), 834-843; (b) VIJ, A.;
PAVLOVICH, J. G.; WILSON, W. W.; VIJ, V.; CHRISTE, K. O.: Angew. Chem., Int. Ed. Engl.
2002, 41(16), 3051-3054; (C) HUYNH, M.-H. V.; HISKEY, M. A.; HARTLINE, E. L.; MONTOYA,
D. P.; GILARDI, R.: Angew. Chem., Int. Ed. Engl. 2004, 43(37), 4924-4928.
(A) BOTTARO, J. C.; PENWELL, P. E.; SCHMITT, R. J.: J. Am. Chem. Soc. 1997, 119(40), 94059410; (b) GILARDI, R.; FLIPPEN-ANDERSON, J.; GEORG, C.; BUTCHER, R. J.: J. Am.
Chem. Soc. 1997, 119(40), 9411-9416; (c) SITZMANN, M. E.; GILARDI, R.; BUTCHER, R.
J.; KOPPES, W. M.; STERN, A. G.; TRASHER, J. S.; TRIVEDI, N. J.; YANG, Z.-Y.: Inorg.
Chem. 2000, 39(4), 843-850; (d) MARTIN, A.; PINKERTON, A. A.; GILARDI, R. D.;
BOTTARO, J. C.: Acta Crystallogr. 1997, B53, 504-512.
(a) KLAPTKE, T. M.; MAYER, P.; SCHULZ, A.; WEIGAND, J. J.: J. Am. Chem. Soc. 2005,
127, 2032-2033; (b) GLVEZ-RUIZ, J. C.; HOLL, G.; KARAGHIOSOFF, K.; KLAPTKE,
T. M.; LHNWITZ, K.; MAYER, P.; NTH, H.; POLBORN, K.; ROHBOGNER, C. J.;
SUTER, M.; WEIGAND, J. J.: 2005, Inorg. Chem. 2005, in press; (c) FISCHER, G.; HOLL,
G.; KLAPTKE, T. M.; WEIGAND, J. J.: Thermochim. Acta 2005, manuscript submitted.
(a) GAPONIK, P. N.; KARAVAI, V. P.: Khim. Geterotsikl. Soedin. 1984, 1388-1391; (b)
MATULIS, V. E.; LYAKHOV, A. S.; GAPONIK, P. N.; VOITEKHOVICH, S. V.;
IVASHKEVICH, O. A.: J. Mol. Struc. 2003, 649, 309-312.
HOW-GHEE, A.; FRAENK, W.; KARAGHIOSOFF, K.; KLAPTKE, T. M.; MAYER, P.;
NTH, H.; SPROTT, J.; WARCHHOLD, M.: Z. Anorg. Allg. Chem. 2002, 628, 2894.
N-N values and N=N values from: International Tables for X-ray Crystallography; Kluwer
Academic Publisher: Dordrecht, 1992; Volume C.
583
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
(a) DRAKE, G.; HAWKINS, T.; BRAND, A.; HALL, L.; MCKAY, M.: Propellants,
Explosives, Pyrotechnics. 2003, 28(4), 174-180; (b) XUE, H.; ARRITT, S. W.; TWAMELY,
B.; SHREEVE, J. M.: Inorg. Chem. 2004, 43(25), 7972-7977.
(a) Test methods according to the UN Recommendations on the Transport of Dangerous
Goods, Manual of Tests and Criteria, fourth revised edition, United Nations Publication, New
York and Geneva, 2003, ISBN 92-1-139087-7, Sales No. E.03.VIII.2.; 13.4.2 Test 3(a)(ii)
BAM Fallhammer; (b) REICHEL & PARTNER GmbH, http://www.reichel-partner.de/.
(a) OZAWA, T.: Bull. Chem. Soc. Jpn. 1965, 38, 1881; (b) KISSINGER, H. E. Anal. Chem.
1957, 29, 1702.
Standard Test Methods for Arrhenius Kinetic Constants for Thermally Unstable Materials.
ASTM designation E698-99, 1999.
MADER, C. L.: Detonation Properties of Condensed Explosives Computed Using the BeckerKistiakowsky-Wilson Equation of State, Los Alamos Scientific Laboratory 1963, Rept. LA2900.
KHLER, J.; MAYER, R.: Explosivstoffe, 7. Aufl., Wiley-VCH, 1991, Weinheim.
(a) KAMLET, M. J.; JACOBS, S. J.: J. Chem. Phys. 1968, 48, 23; b) KAMLET, M. J.;
ABLARD, J. E.: J. Chem. Phys. 1968, 48, 36; c) KAMLET, M. J.; DICKISON, C.: J. Chem.
Phys. 1968, 48, 43.
Standard Test Methods for Mol Percent Impurity by DSC. ASTM designation E928-96, 2003.
STONER, C. E.; BRILL, T. B.: Combust. Flame 1991, 83, 302.
COSTA, L.; CAMINO, G.: J. Therm. Anal. 1988, 34, 423.
MILLIGAN, D. E.; JACOX, M. E. : J. Chem. Phys. 1975, 56, 333.
SHIMANOUCHI, T.: Tables of Molecular Vibrational Frequencies Consolidated Volume I,
National Bureau of Standards (1972) 1.
SHIMANOUCHI, T.: Tables of Molecular Vibrational Frequencies Consolidated Volume II, J.
Phys. Chem. Ref. Data, 6(3) (1972) 993.
BRAND, C. D.; CAWTHON, T. M.: J. Am. Chem. Soc. 1955, 77, 319.
NAKAMOTO, N.: Infrared and Raman Spectra of Inorganic and Coodination Compounds, 4th
ed.; Wiley & Sons: New York, Chichester, Brisbane, Toronto, Singapore, 1986.
MECKE, R.; LANGENBUCHER, F.: Infrared Spectra, Heyden & Son, 1965, London, Serial
No. 106.
MISHRA, I. B.; RUSSELL, T. B.: Thermochim. Acta 2002, 384, 47.
BRILL, T. B.; BRUSH, P. J.; PATIL, P. J.: Combust. Flame 1993, 92, 178.
GMELIN, L.: Gmelins Handbuch der anorgnischen Chemie, 8. Aufl. Verlag Chemie,
Weinheim, 1971, Syst.Nr. 14, Teil D1, 347.
LOW, M. J. D.; RAMAMURTHY, P.: J. Res. Inst. Catal. 1968, 16(2), 535.
JAIN, S. R.; RAO, M. V.; VERNEKER, V. R.: Propellants, Explosives, Pyrotechnics 1979, 3,
761.
MEINHARDT, D.; BREWESTER, G.; CHRISTOFFERSON, S.; WUCHERER, E. J.: AIAA
98-4006, 34th AIAA/ASME/SAE/ASEE Joint Propulsion Conference, 13 15 July, Cleveland,
OH, USA, 1998.
ANFLO, K.; GRNLAND, T. A.; WINGBORG, N.: AIAA 2000-3162, 36th AIAA/
ASME/SAE/ASEE Joint Propulsion Conference, 16 19 July, Huntsville, AL, USA, 2000.
584
DEVELOPMENT OF TESTING METHOD FOR DETERMINATION

INITIATION STRENGTH OF DETONATORS BY MEASUREMENT
IN IMPULSE CHAMBER
Milan Klusek, and Marcela Jungov
Department of Theory and Technology of Explosives (DTTX),
Abstract:
A new testing method for determination of initiation strength of detonators is described.
Results of some experimental investigations of measurement of explosion of detonators
in impulse chamber are presented. The principle, mathematical apparatus and
instrumental equipment of this new non-direct testing method are described. The method
seems to be simple, prompt, with little demands for staff and measuring installation,
economically practicable and safe, and could be used as a standard testing method for
commercial detonators.
Keywords:
1.
detonator, initiation strength, impulse chamber
INTRODUCTION
1.1
Initiation Strength of Detonators (ISD)
In both experimental and industrial area it is often necessary to compare various types,
sorts and dimensions of detonators. A parameter called initiation strength is appropriate for
such comparison.
The initiation strength of detonators (ISD) is defined as the ability of detonator to initiate
stable detonation of explosive material of definite sensitivity.
1.2
Factors Affecting ISD
The initiation strength depends in particular on the construction proper of a given

detonator, this construction being determined by [1]:
- mass, material and wall thickness of casing
- mass, material and wall thickness of blasting fuse
- primary charge of detonator
- secondary charge of detonator
- interface area between the charges
- orientation of detonator with respect to the witness object
585
1.3
Method of Determination of ISD
Various methods have been developed for determination of ISD: they differ in the
complexity of evaluation of necessary parameters, reproducibility of results and
instrumentation demands. Important parameters also involve the costs and the needs of
industrial practice.
1.3.1. Indirect Methods
The indirect methods [2] find out the effect of detonation of the detonator upon the
witness object or substrate, i.e. the ISD is proportional to the mechanical work done and/or
the degree of destruction of the witness object. These methods are simple and with low
demands for instrumentation; however, they only reflect some of manifestations of the tested
detonator.
1.4
Direct Methods
In the direct methods [2] the investigated detonator is brought to detonation by a standard
charge, and the effect of their combined detonation is evaluated. These methods are usually
simple, but highly time-consuming and costly. The information content of results for
evaluation of ISD is usually insufficient, because not only the primary effect of detonator but
also the non-negligible co-action of the standard charge is evaluated.
1.4.2. New Trends in Determination of ISD
Also into practical use are being implemented new methods [3] of evaluation and
determination of ISD by measuring the parameters of air shock waves or those of water
shock waves (aquarium test) or of the shock waves spreading within a defined solid medium
(Hopkinson rod). These methods first of all make use of the gradually improving sensor
techniques. They also include the numerical simulations.
2.
SYSTEM OF IMPULSE CHAMBER
The staff of DTTX have been dealing with evaluation of ISD for rather a long time.
These activities have resulted in designing a methodology for evaluation of ISD by means of
the Air Gap Test [4] (in cooperation with Austin Detonator Comp., Vsetn), and also the
methodology described in the present article. The method is new and is being experimentally
tested at present.
2.5
Construction of Impulse Chamber (IC)
The arrangement proper of the IC system is relatively simple. It consists of a concrete

foundation block with a stable tower attached to it by means of fixation bolts securing its
vertical position. Attached to the tower are a vertical measuring scale and a horizontal
working table (ca 1.2 m above the foundation) with a layer of cushioning material. The table
serves for placing the set of impulsion chamber itself: it consists of supporting steel plate
with a series of arrestment bolts attached at the bottom side, and the body of impulse
chamber (Fig. 1).
586
Fig 1.
Scheme of tower
The body of chamber (Fig. 2) is of cylindrical shape (the height to width ratio 26.3/12.6
cm), the cylinder having widened circular fixation flanges at the top and at the bottom. The
bottom flange is bolted to the supporting plate. The space between both metal parts is filled
with rubber sealing. The transition between the flange and steel support is protected from the
inner side of chamber by a steel ring of 2 mm thickness. The upper part of IC is equipped
with a bolted-on and sealed cover with circular manipulation hole enabling insertion and
arrestment of the detonator. The said manipulation hole is a part of circular centric recess
the space for placing the lid (projectile).
Fig 2.
Scheme of impulse chamber

1 body, 2 foundation supporting plate,
3 cover with recess for projectile,
4 projectile, 5 rubber sealing, 6 steel sheet sealing,
7 connecting bolts
587
In the upper third of the impulse chamber body there is a hole for electric wires to the
detonator (or detonation tube). The whole set of IC (Fig. 3) is placed on the table by means
of the above-mentioned arrestment bolts of supporting plate, which keep it in horizontal
position. A series of disk projectiles were made for the impulsion chamber; they are of
various masses and can be connected/combined, as the case may be, to produce projectiles of
desired mass. The surface area affected by the detonator impulse is always constant.
Fig 3.
2.6
View of impulse chamber set
Principle of Method and Evaluation
The method is based on the principle of reading the height to which the lid of impulse
chamber is thrown by the detonation of tested detonator, and determination of impulse.
The mathematical evaluation starts from the energy balance (Eq. (1)): the detonation of
the detonator tested imparts kinetic energy to the lid mass, which is equal to the work needed
to overcome its potential energy. The impulse is a measure of ISD.
1 2
mv = EK mgh = EP
2
(1)
where: m - mass, v - velocity, h - height, g - gravitational acceleration,

EK, EP - kinetic, potential energy
Comparison of EK and EP can be used for an orientation calculation of the maximum

height to which the projectile (lid) will be thrown (Eq. (2)).
h=
v2
2g
(2)
In order to express the specific impulse, we start from the equation (Eq. (3)) taking into
account the shape of charge and the shape of the surface affected by the impulse. The
velocity is expressed by the equation of free fall (Eq. (4)).
589
I=
mv
A
(3)
v = 2 gh
I = 2 gh
(4)
m
A
(5)
where: I - impulse, A - area

From the point of view of ballistics, the following simplifying presumptions were
introduced: the movement of lid is only affected by its weight, and the resistance
characteristics of medium (air) are considered constant.
3.
EXPERIMENTAL
3.7
Detonators Tested
Two sets of detonators were used in the tests:

1) Special test detonators
Whole Case Detonators (D-WC-)

Bottomless Detonators (D-BL-)
Caseless Detonators (D-CL-)
The special test detonators were produced for evaluation of selected factors affecting the
ISDs.
2) Standard industrial detonators
3.8
Preparation of Samples
The special whole-casing and bottomless detonators and secondary particles for
detonators without casing were pressed by means of a mechanical lever press in the press
laboratory of DTTX. The semi-products with primary sections for detonators without casing
(Fig. 4) and the initiation particles for whole-casing and bottomless detonators were supplied
by Austin Detonator Comp., Vsetn.
590
Fig 4.
4.
Caseless detonator

The aim of the experiments carried out particularly included verification of:
physical practicality of the measuring apparatus
sensitivity to some factors generally influencing the initiation strength.
The following were included among the factors studied:
effect of press pressure (140 kg/ks, 220 kg/ks)
effect of orientation of detonator (axial, lateral)
effect of casing (material of casing, D-BL+D-CL)
other effects.
With each of the types of detonators and pressures we carried out 5 measurements with
various masses of projectile. The table presents the average values of height and calculated
impulse.
The projectile mass given was experimentally determined as the optimum value for the
measurements of the whole set of the detonators tested.
591
Table 1. Results of measurements
(mass of projectile m = 1546 g)

Type denotation of detonator
D-WC-Cu, 20.3g Np2, 140 kg/ks

D-WC-Cu, 20.3g Np2, 220 kg/ks
D-BL-Cu, 20.3g Np2, 140 kg/ks
D-BL-Cu, 20.3g Np2, 220 kg/ks
D-CL-Cu, 20.3g Np2, 140 kg/ks
D-CL-Cu, 20.3g Np2, 220 kg/ks
D-WC-Al, 20.3g Np2, 140 kg/ks
D-WC-Al, 20.3g Np2, 220 kg/ks
D-BL-Al, 20.3g Np2, 140 kg/ks
D-BL-Al, 20.3g Np2, 220 kg/ks
D-CL-Al, 20.3g Np2, 140 kg/ks
D-CL-Al, 20.3g Np2, 220 kg/ks
std. D-0-NME-Cu, 20.3g Np, 180 kg/ks
std. D-0-NME-Al, 20.3g Np, 180 kg/ks
el. D-0-NME-Al, (Austin Detonator)
ignition R-Al-type II (Sellier & Bellot)
standard ignition detonator No. 3
standard ignition detonator No. 8
hAXIS
IAXIS
hSIDE
ISIDE
[m]
[Pa.s]
[m]
[Pa.s]
2.8
3.0
3.0
3.5
3.2
3.2
5.3
5.9
7.0
7.5
3.6
3.6
2.4
5.5
5.1
6.3
0.4
4.1
2824.9
2884.2
2884.3
3138.1
3000.2
2981.7
3861.1
4074.4
4061.5
4593.7
3164.6
3182.4
2598.2
3946.7
3787.6
4203.2
1060.9
3412.5
3.6
2.8
2.5
3.5
3.1
2.9
5.5
5.7
6.0
5.5
3.2
3.5
2.4
5.1
4.7
5.5
0.4
4.5
3156.0
2792.1
2672.4
3138.1
2976.8
2875.9
3918.3
4000.0
4384.1
3933.8
2990.7
3128.8
2597.9
3794.2
3620.5
3946.0
1004.0
3541.5
The results presented in Table 1 allow the following conclusions to be drawn:
the differences between the impulse values achieved with various pressures of
pressing (140 kg/ks and 220 kg/ks) secondary charge are not markedly reflected in
ISD
impulses of detonators with aluminium casing are higher than those of the copper
equivalents
the impulse values of laterally ( ISIDE) and axially ( IAXIS) oriented detonators do
not necessarily have to differ, but in contrast they can be diametrally different (Figs
5,6)
no additional evaluation means and apparatus are necessary visual readings are
sufficient
592
5000
AXIS
4500
SIDE
Impulse [Pa.s]
4000
3500
3000
2500
2000
1500
1000
500
0
D-WC-Cu
Fig 5.
D--BLCu
D-CL-Cu
D-WC-Al
D-BL-Al
D-CL-Al
Impulse values for selected detonators (pressure 140 kg/ks)
5000
AXIS
SIDE
4500
4000
Impulse [Pa.s]
3500
3000
2500
2000
1500
1000
500
0
D-WC-Cu
Fig 6.
D-BL-Cu
D-CL-Cu
D-WC-Al
D-BL-Al
D-CL-Al
Impulse values for selected detonators (pressure 220 kg/ks)
593
5.
CONCLUSION
On the basis of the results obtained it was possible to suggest further possible
development of this methodology, particularly refinement of several details of the technique:
- optimization of shape of projectile
- elimination of erosive effect of testing (a more robust construction of chamber)
- enlargement of manipulation hole in the cover
- adding of efficient protective/damping means to the projectile or the chamber proper
It is suggested to continue the work by measurements that can document:
- dependence of impulse on mass and surface area of projectile
- dependence of impulse on working space of chamber
The methodology tested appears to be simple, fast, undemanding for operators and
measuring apparatus, economically accessible and safe. Its drawbacks can lie in the
relatively large weight of the chamber and overall dimensions of the measuring system.
The IC system provides data of good reproducibility and reliably reflects the substantial
selected factors. The methodology is now in a phase of verification and research, and it
offers possibilities of further development and improvement.
REFERENCES
[1]
[2]
[3]
[4]
V. SVACHOUEK: Initiation strength of detonators, PhD Thesis (in Czech), VCHT Pardubice,
1993
J.STRNAD: Initiation properties of the most frequently used high explosives and development of
new methodology for their measurement, PhD Thesis (in Czech), VCHT Pardubice, 1972
M.SUCESKA: M.: Test Methods for Explosives, Springer-Verlag, 1995
M.JUNGOV, J.STRNAD: J.: Initiation Strength of Detonators Air Gap Test Method, 5th
seminar New Trends in Research of Energetic Materials, University of Pardubice, 2002
594
PROPAGATION OF DETONATION
IN CYLINDRICAL LOW-SENSITIVE HE SAMPLES
.V. Kostitsyn, B.G. Loboiko, V.P. Filin, .V. Vershinin, .. Nikulin,
.B. Smirnov, and S.N. Lyubyatinsky
Zababakhin Russian Federal Nuclear Centre - VNIITF,
P.O. Box 245, Snezhinsk, Chelyabinsk region, 456770 Russia
Abstract:
The electrocontact and photochronographic procedures were used to study detonation
propagation in cylindrical samples of different diameter made of insensitive HE for the
cases of single-point and plane-wave initiation. The diameter of samples ranged from 15
to 120mm. The stationary detonation rate vs diameter of the cylindrical sample was
constructed. The relation between the curvature of the detonation front and its rate is
identified.
Keyword:
initiation, detonation, TATB, IHE
Significant value of the cutoff diameter, extended area of detonation wave reaction,
relatively slow kinetics of reaction of low-sensitive HE frequently results in essential
difference between their real detonation rate differs and ideal value. Difference in some
cases is as high as 30 % from ideal detonation rate [1].
The paper covers the results of studies of detonation wave front propagation in cylinder
samples of different diameter made of insensitive TATB-based HE (IHE) for the cases of
single-point and plane-wave initiation. Detonation rate was recorded by electric contact
technique. Time profiles of detonation wave arrival to the end surface of the examined
samples, which were recomputed to the form of detonation front taking into account
recorded values of detonation rate, were recorded using photographic recorder with linear
slot scanning.
We found in our study that when detonation wave passes a certain distance along the
cylinder charge made of IHE its rate D reaches certain value, typical for this diameter of
samples (steady detonation rate Dst), thus deformation of detonation front occurs: detonation
front takes a certain shape, also characteristic for this diameter of samples.
As an illustration of the above Figures 1, 2 plot dependencies D/Dst and S=1/RS versus
the length of detonation wave path (x) in samples with diameter 60 mm (here S is curvature
of detonation front in the zone of sample axis, RS is radius of curvature).
We can see from results given in Figures 1 and 2, that the nature of detonation buildup
in IHE differs essentially for different types of initiation. Process of detonation buildup
consists of two phases, i.e. the phase of rapid buildup D and phase of slow buildup and
monotonous approach to level Dst. The phase of rapid growth D is completed after ~20-30
mm in case of divergent detonation wave, thus D reaches value of ~0.97Dst. Further
detonation velocity slowly increases and reaches the level Dst only at a distance ~200-250
mm.
595
Detonation velocity reaches level Dst already in ~20 mm in case of plane-wave initiation
of IHE. Moreover, detonation rate exceeds level Dst in ~60...70 mm, reaches its maximum
value (~1.1Dst) and only then slowly decreases to the level Dst (at a distance ~150...170 mm).
Curvature of detonation front is stabilized approximately at the same distances that shows
that front shape becomes stable.
Figures 3, 4 illustrate the dependencies Dst/Dlim (Dlim is limiting detonation rate for the
infinite diameter of the samples) and RS versus diameter of cylinder samples d.
Curvature (S=1/RS) of detonation front was determined in the zone of charge axis in
cylinder samples made of IHE for different conditions of propagation of detonation wave,
namely:
At stationary propagation of detonation wave in samples of different diameter;
At propagation of divergent detonation wave in samples with diameter 60 mm in

the domain of slow growth of detonation rate;
For cases of plane-wave initiation and propagation of the detonation wave in the
domain where Dst value is exceeded.
Fig 1.
Dependence D/Dst(x)
596
Fig 2.
Dependence S(x)
Fig 3.
Dependence Dst/Dlim(d). d=15120 mm
597
Fig 4.
Dependence RS(d). d=15120 mm
Figure 5 illustrates D/Dlim(S) dependences for each type of detonation propagation.
D/Dlim=(1+155S)/(1+164S)
Fig 5.
Dependence D/Dlim(S)
598
One can see from the Figure 5 that the obtained results is a single dependence D/Dlim(S)
of IHE detonation rate on detonation front curvature. Similar dependences were recorded for
another non-ideal HE [2, 3].
Obtained results testify to the great role of detonation front curvature in buildup and
propagation processes and allow the following conclusion: after some buildup site, value of
which depends on specific conditions of initiation and propagation, detonation wave velocity
in IHE is virtually completely determined by detonation front curvature.
Existence of virtually unequivocal interrelation between detonation wave velocity and
curvature of its front allows explanation of experimentally recorded nature of dependence
D/Dst(x) (see Figure 1).
Such nature of detonation buildup depends on the fact that stationary propagating
detonation front is set by transpiration of two processes, influencing detonation rate, that is:
process of detonation front structure formation, determined by kinetics of reaction of
explosive transformation, and process of change of shape (curvature) of detonation front
under the influence of relaxation wave on the chemical reaction site. Thus process of change
of detonation front shape under impact of relaxation wave goes much longer than the process
of detonation front structure formation in samples of rather big diameter. Hence detonation
rate increases more rapidly up to the value corresponding to the curvature of detonation front
within initial site of detonation buildup. Further, in case of divergent detonation wave,
detonation rate increases evidently more slowly up to level Dst in accordance with decrease
of curvature of detonation front alongside with the growth of distance from the place of
initiation, and at plane-wave initiation detonation rate decreases up to level Dst according to
detonation front curvature increase under the impact of relaxation wave.
REFERENCES
[1]
[2]
[3]
D.L. KENNEDY, G.B. DRIVE: Multi-valued normal shock velocity versus curvature
relationships for highly non-ideal explosives, 11th Symposium (Int.) on Detonation (1998).
A.W. CAMPBELL, R. ENGELKE: The diameter effect in high-density heterogeneous explosives.
Los Alamos Scientific Laboratory, 6th Symposium (Int.) on Detonation (1976).
P.C. SOUERS, R. GARZA: Kinetic information from detonation front curvature, 11th Symposium
(Int.) on Detonation (1998).
599
THE SPIN-PULSATING REGIME OF DETONATION

IN SOLID AND LIQUID EXPLOSIVES
N.V. Kozak*, G.D. Kozak**, and Zhou Lin***
*Academy of labor and Social Relation, Lobachevskogo st. 90, Moscow, 119454.
**Mendeleev University of Chemical Technology Miusskaya Sq.9, Moscow, 125190.
**Beijing Institute of Technology, P O Box 327, Beijing 100081, China.
Abstract:
The work concerns to investigation of quite distinct inhomogenity in the

detonation front in cast and liquid high explosives.. The experimental
methodology of simultaneous registration of luminosity at detonation of cast
charges on a base of TNT near charge surface and deformation of witnessplate
are described. Experiments with ballistite propellant NB were carried out, and
spin pulsation at detonation were observed too. Solutions of DNT in nitroglycole
(NGL) were investigated with respect to possibility of spin detonation
propagation. Results of investigation justify the conclusion of previous work that
the necessary properties for low frequency spin detonation propagation are:
detonation velocity of system must be D = 7.3-7.4 km/s and heat of explosion
QV = 5.4 MJ/kg.
Keyword:
1.
detonation, sensitizer, detonation velocity, explosion heat,

streak camera.
INTRODUCTION
Quite distinct inhomogenity in the detonation front was found earlier at detonation of cast
charges on a base of TNT and DNT with sensitizers: RDX, PETN [1,2], DINA, TETRYL [3-5]
and at detonation of liquid charges: Nitromethane (NM), Diethyleneglycoledinitrate (DGDN),
solutions of DNT in melted DINA [6,7] The results were summarized last year in work [4].
Investigation in this direction was prolonged. In this report we would like to present the
experimental data about mechanism of spin detonation and pictures sustained the compliance
of alternating of luminosity inhomogenity and step of dents on witnessplate, that spin
pulsations stamp. The experimental methodology of simultaneous registration of luminosity
near charge surface and deformation of witnessplate are described. These runs were carried
out with cast TNT/RDX charges. Experiments with ballistite propellant NB were carried out
and spin pulsation were observed too. Solutions of DNT in nitroglycole (NGL) were
investigated with respect to possibility of spin detonation propagation. All of the results are
compared and discussed on a base of thermodynamically calculated detonation parameters.
600
2.
EXPERIMENTAL
Commercial grade TNT and RDX were uses for cast charges preparation [1,2]. The surface
wave fine structure was investigated in runs conducted as follows: cast charges of mixtures of
TNT with RDX (content 15 or 12.5%) were placed coaxially into a thick-walled brass tubes
that had diameter on 15-20 mm greater than the charge diameter. A tube witness that was
made from copper plate was placed close to be a bushing wall of thick-walled brass tube. The
wave process propagated over the charge surface marked tracks on tube witness. The
luminosity in clearance between a copper tube and a cylinder charge surface was recorded by
means of streak camera (SFR device) through a mirror disposed under the bottom of a charge.
After experiment the copper tube was withdrawn from the thick-walled tube, straightened,
and photographed. One can see in Fig. 1 the experimental arrangement, photo of straightened
copper tube and streak camera record, all of them were made it the same scale.
Fig 1.
The streak camera record (A), photo of straightened copper tube (B),
and photo of the experimental arrangement (C). Section I is visible (front)
side of charge, section II is back side of one.
The explanation one can find in the text.
NGL was synthesized and purified in the Laboratory. DNT was a commercial grade
product, the freezing point was Tf = 60C. The limit of solubility of DNT in NGL at room
temperature (T=19-20C) was C*DNT=30%. At content of DNT CDNT =35-40% the mixtures
were heterogeneous. The density () of solutions was measured by means of picnometer [8].
Measurement of critical (failure) diameter (df) were carried out by means method go-nogo in paper tubes. The tracing paper was reeled up on steel needles or cylinder bobbins of
various diameters and glued by means of rubber glue. Specific weight of tracing paper (on
unit of square) was g= 0.0042 g/cm2. Diameter of steel needles was measured by means of
micrometer. If the needle after drying glue was extracted from the tube with strength the tube
was selected for carrying out the experiment. There was only one layer of tracing paper in
tubes by a diameter d=3-10 mm, and two layers at d<2.0-2.5 mm. There were a paper conic
funnels (angle was ~25-300) in top parts of tubes by a diameter d<6 mm to make possible a
normal initiation of detonation in the charges. The pellets (d~5 mm) of pressed phlegmatized
HMX were used as boosters.
The cylinder part of paper tube was fixed to the surface of brass witness-plate, that in turn
was fastened to heavy steel plate. The tracing paper tubes were transparent and it allowed
making sure that the gas bubbles were removed from the liquid.
601
The Russian streak camera GFR-3 was used to register the luminescence of detonation
front. To give the streak camera records from a site of the bottom of charge the tube by a
diameter d=10 mm was glued to thin glass plate and the mirror was fastened under glass
bottom at the angle 450 to axis of charge. Thus the direction of propagating process was
perpendicular to axis of rotating mirror of GFR-3, as in runs with cast charges (Fig. 1)
Deformation of the surface of witness-plate after the experiments allowed distinguishing
in runs detonation from detonation damping and from detonation failure.
3.
The stamp of spin detonation on straightened cooper tube witness plate, inside those the
cast charge of TNT/RDX 85/15 has been placed, was published in [1,4]. The streak camera
record is shown in Fig. 1 A. One can see in the upper part of it the track of luminosity (2 in
Fig 1A) of upper part of cylindrical charge, that run out from a brass tube (3 in Fig. 1C).
Luminosity of detonation of lower part of the charge correspondingly marked as 1 in Fig. 1A.
Luminosity of middle part could not be seen because it was placed in brass tube, and
imaginary track of detonation was drawn as dash line in Fig. 1A that connected lines 1 and 2.
The streak camera record and the photo of straightened copper tube in Fig. 1 are printed in the
same scale. It permits with help dash line in Fig. 1A to connect each point of luminosity in the
clearance between surface charge (letters with touches) with deformation of copper tube
surface (letters in Fig. 1B). One can see streak camera record making through the mirror (6 in
Fig. 1C) in lower part of Fig. 1A. Dark zones in the streak camera record coincide with pitted
surface of copper tube, i.e. damping of detonation took place in these parts of charge. As it
was assumed in [1,2] then spin detonation wave occurred and propagated in supper compressed
regime. The slopping line on the surface of witness plate is the stamp of spin wave, splashes
of luminosity in clearance that could be seen in upper part of Fig. 1A, coincide with points of
this line. The zone of propagating of detonation in normal regime follows after that, smooth
deformation in copper tube and luminosity without pulsating could be seen in Fig. 1B and
Fig. 1A, correspondingly. Then consistency of these phenomena repeats.
The results of measuring of df of DNT / NGL solutions are shown in Fig.2. Curve df
versus content of DNT divides the area of dark points (detonation in the experiments) and
open ones (detonation failure). NGL detonated in paper tubes by a diameter d=2.1 mm. There
was detonation damping in tubes by a diameter d=2.0 mm, lengths of detonated parts of
charges were lD=10-13 mm, i.e ID=(5-6) d. The detonation damping was observed in tube by a
diameter d=1.9 mm, but lD was less, 1D=4 d. Detonation did not propagate at all in tubes by a
diameter d=l.5 mm. Addition of DNT up to CDNT=15% did not change df of solution, it
remained constant df =2 mm. Arising of CDNT more than 15% led to growth of df.
Fig 2.
602
Crinical detonation
diameter of NGL/DNT
solutions vs. DNT
content. Dark points
signify detonation, open
ones - failure of
detonation.
The oscillations were registered in streak camera records as alternating of dark and white
borders near the lateral surface of charge (Fig.3). Simultaneously on the surface of witnessplate the oscillations of detonation front were registered as small ferrous arranging across the
direction of detonation front propagation. The distance between white border on the records
(the step of oscillation) corresponded to average distance between the furrows on the plate
The step of oscillation of detonation front did not depend on concentration of DNT in mixture
and was approximately h= I mm.
Fig 3.
Streak camera record at detonation of

the charge (d=10 mm) of solution
NGL/DNT 95/5 (A) and static photo
of a charge (B).
1 charge in paper confinement,
2 - glass bottom of it,
3 - end-wall, that is visible through
the mirror placed under it at angle
45.
The failure diameter of detonation of NGL in glass tubes [9] was found to be df=1.9 mm.
In paper tubes practically the same value was measured, df =2 mm, but paper in contrary to
glass confinement did not block the registration of detonation front oscillation neither in
records nor on witness-plates. The parameters of detonation of NGL-DNT solutions at
experimental density calculated by means of SD computer code [10,11] are given in Table 1.
At detonation velocity of NGL D=7 km/s the frequency of oscillation of detonation front
f=D/h ~ 7 MHz. It is to 3-3.5 times higher than spin pulsating at detonation of cast charges of
TNT-RDX and TNT-PETN was [1,2,4,5]. All explosives which detonate in low frequency
regime were shown to have the detonation velocity D=7.3-7.4 km/s and practically the same
value of the explosion heat Qv=5.4 MJ/kg [4,5]. Double base propellant (NB) have D and Qv
that coincide with these values (Table 1), and fluctuation of luminosity [2] were registered in
streak camera record (Fig. 4).
The NGL-DNT solutions, and itself NGL have Qv values sufficiently higher than 5.4
MJ/kg (Table 1), and although the velocity of detonation is D=7.3-7.4 km/s at CDNT=10-20%
the low frequency regime of spin detonation does not realize
603
Table 1. The failure detonation diameter, step and frequency of oscillation,

and calculated parameters of detonation
Explosive system
Density,
g/cm3
Experimental parameters
df, mm
h, mm
f, MHz
Propellant NB
1.56
~9*
2.5
NGL
1.48
2.0
~7
Calculated
parameters
Qv,
D,
km/s
MJ/kg
7.40
5 39
6.69
7.41
NGL/DNT
90/10
1.4'7
-7
NGL/DNT
80/20
1.45
2.5
~7
7.33
6.06
NGL/DNT
70/30
1.44
4.5
~7
7.16
5.87
6.55
at t=600C
Fig 4.
604
Fluctuation
of luminosity at
detonation
of propellant NB.
Diameter d=9.4
mm. This run was
carried out at
temperature t=600C.
D=7.32 km/s.
4.
CONCLUSION.
The experiments that made clear mechanism of spin detonation of cast charges of
mixtures TNT with RDX were carried out at simultaneous registration of fluctuation of
detonation front luminosity and tracks of spin detonation at witness-plate. The mechanism of
spin detonation consists of consecutive propagation of damping wave along the lateral surface
of a charge, reinitiating of detonation and propagation of super compressed detonation wave
in the thin subsurface layer of explosive. Last process propagates along a spiral trajectory and
is accompanied by bright luminosity and higher deforming action then usual detonation
process. All of the stages of spin pulsating detonation were fixed by means of suggested
method.
Failure detonation diameter of solutions DNT in NGL was measured vs. DNT
concentration in paper tubes. Oscillation of high frequency at detonation was registered by
means of streak camera and of copper witness-plate. Frequency of oscillation at detonation of
solution on a base of NGL (f=7 MHz) is rather more than at spin detonation of cast charges
on a base of TNT, propellant NB and other systems that are capable to low frequency spin
detonation (f=2-3 MHz).
Results of investigation justify the conclusion of work [4] that the necessary properties for
low frequency spin detonation propagation are: detonation velocity of system is D=7.3-7.4
km/s and heat of explosion QV=5.4 MJ/kg. If one of them did not coincide with these values
low frequency oscillations at detonation of a system are not observed.
Acknowledgement
605
REFERENCES
[1]
[2]
[3]
[4]
[5]
KOZAK G.D., KONDRIKOV B.N., OBLOMSKY V.B.: Spin detonation in solid substances,
Comb., Expl. & Shock Waves, 1989, v.25, 4, 88-93.
OBLOMSKY V.B.: Non-ideal detonation of homogeneous and porous explosives. Cand. Sci.
(Ph. D.) Thesis. Mendeleev University of Chemical Technology, Moscow, 1988.
ZHOU LIN.: Surface phenomena at detonation of condensed explosives. Cand. Sci. (Ph. D.)
Thesis. Mendeleev University of Chemical Technology, Moscow, 2000.
KOZAK G.D., LIN Z:. The condensed explosive systems capable of propagating detonation in
the spin-pulsating regime. New trends in research of energetic materials, proceedings of the 7th
seminar Univ. Pardubice, Czech. Republic, April 20-22, 2004, p. 516-523.
KOZAK G. D., ZHOU LIN.: Spin pulsating detonation of cast charges. Trans. of the
conf. (Int). The theory and practice of energetic materials. V. 5, Science press. BeijingNew York. 2003, pp.404-407.
[6]
KOZAK G.D., KONDRIKOV B.N., OBLOMSKY V.B.: Spin wave and attenuation of liquid
explosive detonation, Combustion, Explosion, and Shock Waves, vol.28, No.2, pp. 195-199,
March-April, 1992.
[7] KOZAK G.D., KONDRIKOV B.N., ELISEENKOV M.YU. ZHUK YU.S: Influence of viscosity
on detonation stability of dietileneglycoldinitrate. Combustion, Explosion, and Shock Waves,
1996, v. 32, 1, 111-114
[8] KOZAK G.D., RAIKOVA V.M., RASSHCHUPKINA N.V.: Critical diameter and fluctuation
in Detonation of Solution of Dinitrotoluene in Nitroglycol. Combustion, Explosion and Shock
Waves v.35, 3, 1999, p. 303-308.
[9] KOZAK G.D., KONDRIKOV B.N., STARSHINOV A.V.: Critical conditions of low and high
velocity regimes in liquid nitrocompounds. Trans. of 11th Int. Detonation Symposium (Int),
Bookcomp, Ampersand, 2000, 81-86.
[10] KONDRIKOV B.N., SUMIN A.I.: Equation of state of gases at a high pressure, Comb., Expl.
& Shock Waves, 1987, v.23, 1, 114-123.
[11] SUMIN A.I., GAMEZO V.N., KONDRIKOV B.N., RAIKOVA V.M.: Shock and detonation
general kinetics and thermodynamics in reactive systems computer package, Trans. of 11th
Detonation Symposium (Int.), Bookcomp, Ampersand, 2000, 30-35.
606
INTRODUCTION TO THERMAL STABILITY STUDIES

OF NITROBENZENE AND NITROPHENOL DERIVATIVES
IN SULFURIC ACID
Miloslav Krupka
Abstract:
Differential thermal analysis was used for evaluation of thermal stability of
nitroaromatic byproducts from industrial manufacture of nitrobenzene. Influence of
70% sulfuric acid and traces of Fe2O3 on thermal stability of these byproducts was
investigated for initial characterisation of risks of runaway reactions in waste liquids. It
was found that 2-nitrophenol forms thermally unstable mixture with sulfuric acid and
can thus act as a initiator of runaway reactions. As the only one from tested substances,
it showed exothermal decomposition at temperatures below 100 oC. Presence of Fe2O3
in reaction mixtures at the applied experimental conditions does not influence onsets of
exothermal decomposition reactions but changes their mechanisms.
Keywords: nitrobenzene, 1,3-dinitrobenzene, 2-nitrophenol, 4-nitrophenol,
2,4-dinitrophenol, 2,6-dinitrophenol, 2,4,6-trinitrophenol, sulfuric acid,
DTA, thermal stability, thermal decomposition
1.
INTRODUCTION
In industrial manufacture of nitroaromatic compounds, significant amounts of

byproducts are produced, which typically remain in waste liquid after nitration. During
manufacture of nitrobenzene (NB), which is one of the most widely used nitroaromatics, the
waste liquid is formed by 70% sulfuric acid containing 10 20 % of a mixture of
nitrobenzene, dinitrobenzene and nitrophenols. In continuous manufacturing process, the
waste liquid intended for recycling is kept at elevated temperatures for a period as long as
several weeks, which can lead to increase in concentration of dissolved nitroaromatics.
Mixtures of nitroaromatics with sulfuric acid are reported to be less stable than the
nitroaromatics alone[1]. Thermal initiation of their diluted solutions in waste liquid is
improbable at normal conditions as it would require a strong initiating impulse. Conditions
at industrial manufacture can however form coincidence of factors which can lead to release
of thermal energy sufficient for initiation of runaway thermal decomposition of the waste
liquid. Local overheating can start exothermal reactions which take place at temperatures
higher than those normally used during the manufacture.
In nitrobenzene manufacture, the following risk factors are involved (among others):
Nitrophenols, the main byproducts of NB manufacture[2], form very sensitive

explosive salts in contact with metals.
In sulfuric acid, exothermal decomposition reactions can proceed in several

steps. Primary decomposition reactions can start at relatively low temperatures.
607
Side reactions can be catalysed by metal ions released to the reaction mixture by
corrosion of manufacturing equipment or contained in raw materials.
Due to complexity of a manufacturing equipment, reactive compounds can

accumulate in hard-to-reach places.
Side exothermal reactions are not expected outside the reactor in normal manufacturing
conditions. Unexpected accidental situations may however happen during repairs or
shutdowns. For example, long-term circulation of waste liquids at elevated temperatures can
initiate exothermal reactions with long induction periods. Dilution heat can be evolved from
mixing of less and more concentrated solutions of acids etc.
This paper describes introductory study of a thermal behavior of selected nitroaromates
with sulfuric acid and Fe2O3. Its results will be used as a starting point for more detailed
studies characterising conditions which can lead to accidental situations at manufacture of
nitrobenzene.
2.
EXPERIMENTAL
The following nitroaromatic compounds were studied: nitrobenzene (NB), 1,3dinitrobenzene (1,3-DNB), 2-nitrophenol (2-NP), 4-nitrophenol (4-NP), 2,4-dinitrophenol
(2,4-DNP), 2,6-dinitrophenol (2,6-DNP) and 2,4,6-trinitrophenol (2,4,6-TNP) in Analytical
Reagent purity, alone and in mixtures with 70% H2SO4 (A.R.) and H2SO4 + Fe2O3 (A.R.)
Differential thermal analyser DTA 551 Ex (manufactured by OZM Research) was used
with the following experimental conditions:
Heating rate: 5C .min-1
Sample weights:
a) 30 mg of a nitroaromatic compound
b) 30 mg of a nitroaromatic compound + 70 mg of H2SO4
c) 30 mg of a nitroaromatic compound + 70 mg of H2SO4 + 1 mg of Fe2O3
Temperature range:
30 550 C
Analyses were carried out in open test tubes and a thermocouple protected by a glass
sheath was inserted directly into the sample. Data were evaluated using MEAVY 2.0.0.4
software of DTA 551 Ex analyser. Onsets of the first exothermal effect were evaluated at the
thermograms.
3.
RESULTS AND DISCUSSIONS
For evaluation of the results it is necessary to take into account DTA principles, which
do not allow to distinguish different thermal processes if occurring simultaneously. For
example, exothermal decomposition of a substance can be accompanied by endothermic
evaporation of either the substance itself or its reaction products. If we however analyse
reactions occurring below the boiling point of the mixture, the results can be considered
representative. For more detailed understanding of the reaction it is however necessary to run
more analyses in conditions preventing evaporation of the gas phase from the reaction cell.
608
3.1
Pure nitroaromatics
Figure 1 shows DTA thermograms of pure nitroaromatics aligned according to intensity

of their exothermal decomposition. It is evident that 2,4,6-TNP, 2,6-DNP, 2,4-DNP a 4-NP
decompose exothermally at temperatures above 225 oC. Thermograms of remaining
compounds (2-NP, NB, 1,3-DNB) show only endothermic phenomena (melting,
evaporation).
Temperature Diference [C]
Comparison of DTA Thermograms of Pure Nitroaromatics

16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
-1
-2
-3
-4
-5
-6
-7
-8
-9
-10
-11
-12
-13
-14
2,4,6-TNP
2,6-DNP
2,4-DNP
4-NP
1,3-DNB
2-NP
NB
50
100
150
200
250
Temperature [C]
300
350
400
450
Meavy
Comparison of DTA thermograms of pure nitroaromatics
3.2
Mixtures of nitroaromatics with sulfuric acid (Fig. 2)
Nitrobenzene in sulfuric acid did not show any exothermal effects at the applied
experimental conditions.
1,3-dinitrobenzene in H2SO4 did not show any significant deviation from decomposition
pattern of a pure substance until 180 oC. Above this temperature, endothermic evaporation of
reaction products start and it is overwhelmed by intense exothermal decomposition above
235 oC.
Thermogram of 2,4,6-trinitrophenol in sulfuric acid is an example of exothermal
decomposition masked by endothermic evaporation effect. The mixture apparently shows
higher thermal stability than pure substance. Evaporation of a liquid phase starts at about 170
o
C and it causes cooling of the reaction mixture accelerating to exothermal decomposition
reactions at higher temperatures.
Thermograms of 2,4- and 2,6-dinitrophenol in H2SO4 are very similar. Melting points
are almost identical with pure substances and then exothermal decompositions start at about
180oC.
609
4-nitrophenol is known to form an adduct with sulfuric acid having a melting point of 90
C[3]. Melting of this adduct can be seen on the thermogram of Fig. 2. Unlike pure 4-NP
which is thermally stable until 240 oC, its solution in H2SO4 starts to decompose
exothermally at temperature of 170 oC.
o
Comparison of Thermograms of Nitroaromatics in 70% Sulfuric Acid

40
35
30
25
2-NP
Temperature diference [C]
20
15
4-NP
10
2,6-DNP
2,4-DNP
0
2,4,6-TNP
-5
1,3-DNB
-10
-15
NB
-20
70% Sulfuric Acid
-25
50
100
150
200
250
300
Temperature [C]
350
400
450
500
Meavy
Comparison of thermograms of nitroaromatics in 70% sulfuric acid

As evident from a thermogram detail on Fig. 4, 2-nitrophenol is very unstable in mixture
with sulfuric acid. As the only one of tested substances, it shows exothermal decomposition
at temperatures below 100 oC. The decomposition starts at a temperature of about 60 oC and
accelerates at 140 oC. The decomposition is partially compensated by an evaporation process
(Fig. 2).
3.3
Mixtures of nitroaromatics, H2SO4 and Fe2O3 (Fig. 3)
If we compare thermograms of nitroaromatics in sulfuric acid with addition of Fe2O3

(Fig. 3) with corresponding thermograms without this substance (Fig. 2), it is clear than
Fe2O3 causes more vigorous exothermal decomposition of the mixtures. The only exception
is a mixture based on NB, which thermogram does not show any significant exothermal
effect. Addition of Fe2O3 to 2-NP in sulfuric acid clearly accelerates its decomposition
reactions as it can be seen from Fig. 4.
Fig. 5 compares thermograms of mixtures of 1,3-DNB with Fe2O3 and different amount
of sulfuric acid. It is clear than decrease in H2SO4 amount leads to unmasking of
exothermal reactions taking place in parallel with endothermic evaporation.
610
Comparison of Thermograms of Nitroaromaticsin 70% Sulfuric Acid and Fe2O3

65
60
55
50
2-NP
45
Temperature diference[C]
40
4-NP
35
30
2,6-DNP
25
2,4-DNP
20
1,3-DNB
15
2.4.6-TNP
10
NB
5
0
-5
50
100
150
200
250
Temperature[C]
300
350
400
Meavy
Comparison of thermograms of nitroaromatics in 70% H2SO4 with 1 mg of Fe2O3 added
DTA Detail of 2-NP Mixtures
DTA Thermograms of 1,3-DNB Mixtures

18
17
16
15
14
13
12
11
10
2NP + H2SO4
8
7
6
1,3-DNB + H2SO4 + Fe2O3 (30+70+1)
5
4
3
2
1,3-DNB + H2SO4 + Fe2O3 (30+90+1)
1
0
-1
-2
2NP + H2SO4 + Fe2O3
1,3-DNB + H2SO4 (30+70)
-3
-4
-5
1,3-DNB
-6
-7
-8
-9
60
70
80
90
100
110
120
130
Temperature [C]
140
150
160
170
180
Thermogram detail
of 2-nitrophenol in 70% H2SO4
with and without Fe2O3
60
80
100
120
140
160
180
200
220
Temperature [C]
240
260
280
300
320
340
360
380
Comparison of thermograms
1,3-dinitrobenzene in 70% H2SO4
(30:70 mg and 30:90 mg) with
Fe2O3, 1,3-DNB in 70% H2SO4 and
pure 1,3-DNB
Table 1 compares experimentally found onsets temperatures of exothermal

decompositions with tabulated data on physical properties of pure nitroaromates.
611
Table 1. The influence of Sulfuric acid and Fe2O3 to thermal stability

of nitroaromatic compounds
Substance
Tabulated data
of pure
substances
Onset of decomposition peak [C]

at heating rate of 5C.min-1
exothermal or endothermal Pure
+ H2SO4
+ H2SO4 + Fe2O3
189
172
210,9b[3,5]
303b[3],
302b[5],291b[6]
2-nitrophenol
214-216b[3,4],
152
76
b[6]
217,2
4-nitrophenol
279c[4,6]
252
183
1812,4-dinitrophenol
> 70d[3]
256
190
1872,6-dinitrophenol
245
191
1872,4,6-trinitrophenol
122,5d[3]
180225
184a[6]
> 300
a) Decomposition, b) Boiling, c) Boiling with decomposition, d) Sublimation
Nitrobenzene
1,3-dinitrobenzene
4.
CONCLUSION
Investigated nitrobenzenes and nitrophenols are thermally stable in pure state and their
exothermal decomposition starts at temperatures above 200 oC under conditions of the test.
Addition of 70% sulfuric acid shifts the decomposition to lower temperatures in all cases.
Presence of Fe2O3 in reaction mixtures does not significantly influence onsets of exothermal
decomposition reactions but changes their mechanisms. The least stable system of those
tested is 2-nitrophenol in sulfuric acid which decomposition starts at about 60 oC and
steadily grows to a thermal explosion. This interaction can be an initiator of runaway
reactions in a waste liquid from nitrobenzene industrial manufacture and thus needs further
investigation.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
T. GREWER: Thermal Hazards of Chemical Reactions, Elsevier, Amsterdam, p. 246,

1994
A.A. GUENKEL, T.W. MALONEY: Recent Advances in the Technology of
Mononitrobenzene Manufacture, in L. F. Albright, at al, Nitration: Recent Laboratory
and Industrial Developments, (Acs Symposium Series, 623) American Chemical
Society Meeting , Anaheim, 1995
T. URBANSKI: Chemistry and Technology of Explosives, SNTL Praha, 1958
S.M. KAYE: Encyclopedia of Explosives and Related Items, Volume 8, US Army
Research and Development Command, Dover, New Jersey, 1978
B.T. FEDOROFF, O.E. SHEFFIELD: Encyclopedia of Explosives and Related Items,
Volume 2, Picatiny Arsenal, Dover, New Jersey, 1962
M. VECERA at al: Chemick tabulky organickch slouenin, SNTL Praha, 1975
612
COMPARISON OF SMOKELESS NITROCELLULOSE POWDER

PRODUCTION ACCIDENTS IN EXPLOSIA WITH THE WORLDS
ACCIDENTS DATABASE
Richard Kuracina
Abstract:
Smokeless nitrocellulose powder production technology is one of the most dangerous
technologies in the chemical industry, because all chemicals used in high amount
(without water) can be considered as explosives. Smokeless powder production in
Explosia has begun in year 1924. Several causes from this year to present, defined as
accidents, occurred. Therefore this contribution is focused on statistical evaluation of
accidents in Explosia. Typical notes of these accidents are compared with worlds
accidents database for the same time period.
Keywords:
1.
accident, smokeless nitrocellulose powder
INTRODUCTION
Smokeless powder production database of accidents began in the year 1925 with the
accident in object B12 and was continuing until the year 2002 with another 30 accidents
taking place.
1.1
Basic operations used in processing of nitrocellulose into smokeless

powder
Various technologies are used in smokeless powder production. Smokeless powder with
required properties is obtained by combining of these technologies. Here are the main
technologies used in process:
Transport technology which is linked with all the operations, metal or plastic transport
vessels are used.
Gelatination nitrocellulose conversion to gelatin state using ether and ethanol mixture,
primary process in Explosia
Pressing gelatin pressing to required form using hydraulic press / pressing of
nitrocellulose to create slide powders, extruding through matrices to create bar powders
Soaking technology of dissolving used during nitrocellulose processing, nitrocellulose
material contains crystals of well soluble salts, e.g. KNO3, NaCl
Surface finishing technology used in modification of powders surface conditions,
generally it means creating a graphite foil covering the surface of powder particle or creating
an organic foil which is modifying ballistic properties.
Cribbling homogenization of produced batch of smokeless powder according to the
differences in size and the measure of powder particles deformation.
613
Drying it is a partial drying of fugitive solvents contained in powder particles to

improve its mechanical properties before surface finishing or at the final stage as a complete
drying of remaining fugitive solvents and also drying out the water in porous powders
production.
Warehousing, homogenization of batches final phase in smokeless powder processing
technology
1.2
Chemicals used in smokeless powder production
1.2.1. Nitrocellulose
Nitrocellulose is imported to Explosia from Synthesia plant as compressed cakes. To
ensure the safety, these cakes must contain minimum 20 mass % of alcohol to eliminate its
explosive conditions.
Nitrocellulose as a mixture of chemicals shows very low stability; therefore the use of
stabilisators (centralites, diphenylamin ...) is necessary. In dry state it is extremely sensitive
to friction, strike, heat and fire. Minimum initiation energy of nitrocellulose is 0,60 mJ.
1.2.2. Diethylether
Diethylether is used as main gelatining agent in nitrocellulose processing. It is unstable
exposed to light because it creates peroxides, therefore must be stored in dark tanks or urea
must be used as a stabilisator.
Concerning the safety, diethylether is the most hazardous chemical used in smokeless
powder production. Temperature of combustion is 45C, temperature of spontaneous
combustion is 160C. Ether vapours in combination with air create an explosive mixture in
concentration range of (1,9 36,0) vol. %. Minimum initiation energy of ether vapours is
0,20 mJ.
1.2.3. Ethylalcohol
Ethanol is used in technology of powder processing mainly as wetting and gelatining
agent together with diethylether. Temperature of ethanol combustion is 16,6C and
temperature of spontaneous combustion is 363C. Ethanol vapours combined with air create
an explosive mixture in concentration dimension of (3,3 19,0) vol. %.
1.2.4. Stabilisators
Because nitrocellulose is nitrated natural material, there are some chemical changes
taking place, generally named as unstability of this material. Therefore it is necessary to use
stabilisators, which generally fix nitro-oxides (NOX) to avoid the changes of powder
particles properties.
Urea derivates (centralite I, centralite II) and diphenylamin are most often used as
stabilisators. In pioneer-days of smokeless powder production were used CaCO3 and MgO as
stabilisators. Because in smokeless powder these agents exist only in small amount, they do
not affect the final conditions in the matter of safety.
1.2.5. Graphite
Graphite is used for surface finishing of smokeless powder particles as a reduction of
static electricity induction in the particles. Its addition eliminates conglomeration of powder
particles.
614
1.2.6. Flegmatisators
Camphor, dibutylphtalate and dinitrotoluenes are used as flegmatisators for smokeless
powder particles. They are used for modification of ballistic properties of powders; therefore
these agents do not affect safety qualities of powders.
1.2.7. Inorganic additions
Inorganic additions are added into the powders as well soluble salts for porous powders
(KNO3, NaCl are added), optionally agents modifying ballistic properties are added (for
example K2SO4 as flame damper).
1.2.8. Water
This compound creates the major environment for smokeless powder production. In the
matter of safety, water does not create any potential hazards.
1.3
Basic equipment
1.3.9. Mixers
Equipment used for nitrocellulose gelatination using ether-alcohol mixture. This
equipment contains spinning components, which can create friction.
1.3.10. Hydraulic press
Equipment used for extruding or pressing of nitrocellulose gelatin. This equipment can
endanger the process by overheating, friction, high pressure and adiabatic compression.
1.3.11. Cutting machines
Equipment used for cutting of nitrocellulose gelatin and creating of powder particles.
Using of this equipment can cause seizing or overheating.
1.3.12. Transport vessels
Transport vessels are used for transportation of intermediate products and final products
in the process of nitrocellulose processing. The problems could occur during transportation
such as spilling of content followed by initiation.
1.3.13. Drying chamber
Equipment used for drying out the fugitive agents (water, ether-alcohol mixture) from
powder particles. Overheating or initiation of mechanical friction can occur using this
technology.
1.3.14. Graphiting drum
Graphiting drums with wooden balls are used for surface finishing. Strike is the most
probable ignition source of powder in the barrel.
1.3.15. Sifting machine
Equipment used for shape homogenization of produced powder batch. In the safety
matter, this technology can endanger the material by too much friction.
615
2.
COMPARISON OF EXPLOSIA ACCIDENTS DATABASE WITH

THE WORLDS ACCIDENTS DATABASE
2.4
Worlds accidents database
The worlds accident database was obtained from two sources[1,2]. This database covers
time interval between 1900 and 2003, practically the whole last century. Number of records
is almost 100, but the big disadvantage of this database is in too short accident description
using only one or two rows. After all, in many cases (more than 60 %) initiation and
technology involved in accident can be determined. Numbers of accidents in specified
technology and numbers of the most probable ignition sources is written in the Table 1.
Table 1. Numbers of accidents in specified technology and numbers of the most
probable ignition sources, Worlds accidents database
technology
2.5
most probable initiation
gelatination
22
friction
14
pressing
spark, static electricity
transport
strike
drying
14
overheating, heat
warehousing
10
fire
surface finishing
strange element in equipment
service
equipment failure
sorting
Explosia accidents database
Explosia accidents database[3], contain records of accidents in smokeless powder

production since 1924 till 2000. Factors of each particular accident were described in this
database: processing technology in which the accident occurred, most probable cause of
accident and chemical which initiation caused the accident. Finally, database was divided
according to comparison of consequences and damages (Table 2.).
616
Table 2. Numbers of accidents in specified technology, numbers of the most

probable ignition sources, numbers of damages and numbers of initiated
chemicals, Explosia accidents database
initiated chemical
damages
nitrocellulose
44
minimum damage, 1
diethylether
16
middle damage, 2
12
aceton
equipment damage, 3
total destruction, 4
technology
2.6
most probable initiation
gelatination
14
heat
pressing, cutting
friction
surface finishing
spark
11
warehousing
strike
sorting
drying
service
Timeline of Explosia accidents
Fig 1. shows, that until the beginning of 2nd world war the rate of accidents in Explosia
was high. It was the beginning of production of smokeless powders, so the safe methods for
nitrocellulose processing and work with solvents were only in development (e.g. exhaustion
of vapours from mixers was introduced in 1928). During 2nd world war and after the war, the
rate of accidents was approximately 3 accidents per 12 years.
617
number of accidents
2000
1992
1984
1976
1968
1960
1952
1944
1936
1928
year
Fig 1.
2.7
Timeline of Explosia accident database
Evaluation of Explosia accidents database
2.7.1. Evaluation of Explosia accidents database based on damages

The result of Table 3. evaluation shows that according to expectations the greatest
number of accidents produced middle damages (mainly damage of building plaster, roof,
minimum damage of equipment e.g. burned tubes). This result was connected with good
working morality and work in accordance with regulations. Possible consequences were
eliminated by these factors. Section minimum damage refers to accidents, in which
minimum amount of materials was involved. On the other hand big damages (damage of
equipment) were caused by accidents involving big amount of materials. Big damages were
also caused by malfunction of extinguishing system, or by storing of large amount of
materials very close to the building. The section of total destruction shows the smallest
percentage of accidents. In this case we talk about accidents which resulted into total
decomposition of equipment and building caused by violation of regulations or sabotage.
According to weighted average of consequences - see equation (1) we can assume that in
case of accident, the damage will probably attain the level of middle damage.
618
Table 3. Evaluation of Explosia accidents database based on damages
x=
i =1
damages
minimum damage
19
middle damage
46
damage of equipment
23
total destruction
12
xi ni 59
=
= 2, 27
ni
26
(1)
Weighted average of damages evaluation

2.7.2. Evaluation of Explosia accidents database based on initiated chemical
From Table 4. it can be seen that diethylether is the most often initiated chemical. It is
because diethylether is the most dangerous chemical used in nitrocellulose processing.
Nitrocellulose as chemical present in the largest amount is involved in 39 % of all events.
Most often it is combustion of nitrocellulose after friction or strike. In most cases
nitrocellulose contained small amount of water (accidents in the process of gelatination or
drying, eventually warehousing and sorting).
Table 4. Evaluation of Explosia accidents database based on initiated chemical
chemical
2.8
diethylether
57
nitrocellulose
39
acetone
Comparison of Explosia accidents database with worlds accidents

database
2.8.1. Comparison of technologies used in processing of nitrocellulose into

smokeless powder
According to Table 5. gelatination seems to be the most dangerous technology used in
nitrocellulose processing. Gelatination is involved almost in 50 % of all accidents in
Explosia, a.s. Most of these accidents had happened before year 1970. They were old
accidents which became a base for the processing technology improvement. After year 1970
accidents in the process of gellatination still occasionally happens until today. Therefore
running of this technology requires constant attention.
Pressing and cutting of nitrocellulose gelatin is the next dangerous technology used in
Explosia's smokeless powder production. Adiabatic compression of gas bubbles in gelatin is
the biggest problem. This is the cause of almost all accidents. Other processing technologies
619
are comparable with the world in the view of the number of accidents. Accidents in some
technologies in Explosia are less frequent (drying, warehousing) as in identical technologies
used in the world.
Table 5. Comparison of smokeless powder production technologies
technology
Explosia (%)
world (%)
gelatination
47
36
pressing, cutting
20
11
surface finishing
10
warehousing
16
sorting
drying
10
23
service
transport
2.8.2. Comparison of the most probable ignition source in Explosia and in the
world
In Table 6. it can be seen that the most probable ignition source of Explosia's accidents
is spark. Probability of spark ignition (static electricity, mechanical spark ...) is more that
twice as in the world. This can be explained by the fact that control-operators do not know
about the danger of ignition by the spark. Recently accidents with this cause do not happen
anymore because of using appliances and uniforms without the capacity to generate static
electricity.
Heat can cause the ignition with the same probability. Ignition by heat happens mainly
in pressing technology, where gas bubbles cannot be properly removed from gelatin and
following adiabatic compression produces enough heat for ignition. This ignition source can
be eliminated removing of bubbles from gelatin by vacuum.
Friction as an initiation cause of accidents compared to the world is much more
eliminated, but still it is a cause of one third of all accidents. Last ignition source strike, is
comparable with the worlds statistics.
Comparing to the world, in Explosia there was no record of initiation caused by direct
fire, strange element in equipment or equipment failure. This reflects good training and
morality of workers.
620
Table 6. Comparison of the most probable ignitions sources

initiation
3.
Explosia (%)
world (%)
friction
38
43
spark, static electricity
35
15
strike
12
overheating, heat
28
15
direct fire
strange element in equipment
failure of equipment
CONCLUSION
By simple comparison we came to a conclusion that in global, the technology of

smokeless powder in Explosia is comparable with technology used in the world. But some
operations are more dangerous compared to the world, mainly gelatination. In present times
accidents in gelatination process are declining. Last series of accidents in gelatination
occurred in period of 1970 1971. New materials, new technologies of smokeless powder
production and using of conductive uniforms and equipment materials decreased the chances
of accident.
Lately, pressing is the most dangerous technology with adiabatic compression as the
initiation cause. This fact is probably connected to a new technology used for pressing. This
technology uses higher pressing pressures.
Morality of control-operators could be the next possible accident cause, but because of
low accident rate (3 accidents in the same technology with the same course) this accident
cause is not very crucial.
In most of the accidents the most crucial processing section is the section with solvents
(diethylether, ethanol) present in the powder. Therefore it is recommended to focus on the
following sections: exhaustion, elimination of static electricity induction, elimination of
conditions for adiabatic compression of bubbles.
Ignition sources spark and heat appear more often in the world, but spark is not
that common anymore thanks to use of conductive materials. Causes friction and strike
are less common compared to the world. Morality of control-operators plays a big role. In
Explosia friction and strikes during the process are generally more eliminated than static
electricity induction or adiabatic compression during pressing.
High morality of control-operators was confirmed after comparison of initiation causes.
There was not a record of direct fire, strange element in equipment and failure of equipment
in Explosia plant.
Concerning the damages, there is an important fact showing: the amount of accidents in
level of total destruction was the lowest of all the levels of accidents damages, so these
accidents are results of not very probable events. On the other hand the largest numbers of
all accidents could be find in the level of middle damage accidents. This corresponds to
621
properties of materials used in processing. In case of accident these materials will cause
small damage only in certain conditions: minimum amount of combustible materials exposed
to fire, good morality of control-operators and trying to eliminate possible negative
consequences in the primary state.
REFERENCES
[1]
[2]
[3]
Safex International, International exchange of experience on industry-connected accidents of

the explosives, EMS-Dottikon AG, Switzerland
G.S.BIASUTTI: Historie netst v prmyslu vbuin, Semtn, 1978
Smokeless nitrocellulose powder production accident accounts 1925 - 2003, Explosia accident
database
622
EFFECTIVE AMMUNITION 20x102

Karel Leinweber*, and Petr Pchouek**
* EXPLOSIA a.s., 532 17 Pardubice - Semtn
** PROTOTYPA-ZM, s.r.o., Hudcova 533 / 78c, 612 00 Brno
Abstract:
Within the framework of arming the Czech Army with weapon system PL20, the
requirement has been raised for development of effective ammunition for firing on
armored targets. This ammunition should serve at both weapon systems used at combat
fighters, and for carriage weapon systems for land fire.
Keywords:
1.
penetrator, ammunition, projectile
INTRODUCTION
Possibilities of solution:
This problem could be solved using several construction approaches.
Variant 1 : Classical full-calibre armor piercing projectile
Armor is penetrated with full-calibre projectile body (manufactured usually of quality

steel), secondary incendiary effect is produced by means of incendiary filler that is
positioned between projectile body and ballistic cap which serves to improve aerodynamic
shape. The front part of projectile body is flat-tipped to reduce the risk of projectile
reflection on greater angles of impact.
Advantage of this projectile consists in a low price and relatively high accuracy of firing,
a disadvantage is lower penetrating performance, especially on short distances.
Fig 1.
Full-calibre projectile
623
Variant 2 : Full-calibre projectile with armor piercing core

Armor is penetrated with sub-calibre core (manufactured of quality steel, or, better, of
tungsten alloy or depleted uranium), inserted in the jacket. Secondary incendiary effect is
produced by means of incendiary filler that is positioned between projectile body and
ballistic cap which serves to improve aerodynamic shape. To achieve a high initial velocity
it is necessary that the whole projectile is as light as possible. Therefore the case is
manufactured of aluminium alloy or of plastic.
Advantage of this construction lies in higher penetrating performance on shorter
distances and relatively high accuracy of firing. Disadvantage is a rapid decrease of
penetrating performance with distance.
Fig 2.
Full-calibre projectile with armor piercing core
Variant 3 : Spin stabillized projectile with subcalibre core and discarding sabots
Armor is penetrated with subcalibre core (manufactured of tungsten alloy or of depleted
uranium). During its movement in the barrel, the projectile including the core is accelerated
as a unit. Divided base and projectile jacket are discarded on the muzzle of the barrel from
the core which goes on spin-stabilized flying separately. In comparison with Variant 2, the
cross section load of the flying part is higher, ballistic coefficient is lower (projectile
velocity decreases more slowly) and in this way the projectile keeps its penetrating
performance also in greater distance. Weight of the whole projectile is comparable with
Variant 2 and therefore the projectile has also a considerable initial velocity.
Advantage of this projectile consists in relatively high penetration performance,
decreasing more slowly with distance (when compared with Variant 2), and short time of
flying to the target. Disadvantage lies in a high price and lower accuracy of firing caused by
uneven separating of discarding sabots.
624
Fig 3.
Armor piercing spin stabilized

discarding sabot projectile
Variant 4 : Armor piercing fin stabilized discarding sabot projectile

Armor is penetrated with subcalibre core (manufactured of quality steel, or, better, of
tungsten alloy or of depleted uranium). During its movement in the barrel, the core is
accelerated by means of discarding sabots which are separated from the core on the muzzle
of the barrel (due to air gust) and the core then goes on flying separately. Due to great crosssection load of the core, low ballistic coefficient is achieved which secures low decrease of
velocity. To achieve a high initial velocity it is necessary that the whole projectile is as light
as possible and therefore the case is manufactured of aluminium alloy or of plastic. As the
core is (when compared with Variants 2 and 3) considerably slender, it must be finstabilized, as in such a case the stabilizing revolutions, imparted to the projectile in the barrel
are detrimental. For this reason the projectiles of this type are fitted with so called floating
driving band which avoids transmission of rotation on the projectile. In case of calibre
20mm, however, so called labyrinth packing is often used, driving the projectile in bore
lands only. Volume of the gas outrunning the projectile in the gaps between bore grooves
and discarding sabots is reduced just by means of labyrinth packing.
Advantage of this projectile is a high penetration performance, decreasing only slowly
with distance, and a short time of flying to the target. Disadvantage is a high price and lower
accuracy of firing caused by uneven discarding of driving segments and also the fact, that the
part of the core with stabilizers protrudes into the space designed for powder charge, the
quantity of which in the case then has to be reduced.
625
Fig 4.
Armor piercing fin stabilized

discarding sabot projectile
Note: Basic disadvantage of armor piercing discarding sabot projectiles (Variants 3 and
4) is limited (sometimes complete) inability of being fired from the weapon fitted with
muzzle brake, and a risk also exists here of a damage to the own firing aircraft due to
discarded sabots.
2.
EXPERIMENTAL
2.1
Solution
The projectile according to Variant 2, it means armor piercing projectile with tungsten
armor piercing core was chosen for construction solution. This variant offers the possibillity
to gain experience with penetrating performance of tungsten cores at high impact velocities
which will be achieved by decreasing the projectile weight (compared to standard projectile
of weight 101g).
Fig 5.
Design of projectile 20x102
626
After completion of the first two pieces of the designed armor piercing projectiles (see
Fig.5), the following weights were ascertained :
- projectile body with driving band 36g
- cap 10g
- core D=7 and L=65 ms=35g
Average weight of projectile assembly is 86g.
At designed penetrators, L/D ratio = 9.3 seems to be the limit for use, when demands on
mechanical properties of penetrators material are judged. At penetrators of lower slenderness
below L/D < 8 the demands concerning toughness are usually lower. On the basis of this
fact we chose the material with 92.5 % tungsten (by weight), used under name TK of density
17.67 g/cm3. After cold forming and ageing this material has strength limit Rm= 1,550
1,600 Mpa at ductility A5 = 5%.
2.2
Checking of internal ballistic values
For development and verification of production technology for propellant charges,

dummy weight equivalents of 20x102TP projectiles were used for ballistic testing. These
projectiles were modified by cutting material from the projectile cap, so that their weight
corresponded with preliminary calculation of weight of the newly designed projectile
20x102 VOSA. With this way modified projectile the existing as well as newly developed
propellant charges were tested. Results of individual tests are tabelled in Table No.1
Table 1. Results of tests with weight equivalent
Sample
2700
D100-SM-1/04
D100-SM-1/04
D103-SE-1/03
D103-SE-1/03
D103-SE-1/03
D103-SE-2/03
Where:
Note:
Dummy project.
Dummy project.
Dummy project.
Dummy project.
Dummy project.
Dummy project.
Dummy project.
Assembly
20x102VOSA
20x102VOSA
20x102VOSA
20x102VOSA
20x102VOSA
20x102VOSA
20x102VOSA
35.000
35.000
39.000
35.000
38.000
39.000
35.000
T
21
21
21
21
21
21
21
N
1
2
4
1
1
5
1
Ppi
5429
3757
4291
3664
4080
4234
2865
SD D N Pcr SD D N
SD
360 509 2 3530 88 125

259 581 4 3694 195 419 4 1102.9 20.56 43.7
1 1073.7
169 406 5 3540 94 206 4 1084.5 8.82 20.1
weight of powder (g)
Tep
temperature of powder
Ppi
pressure measured using piezo method (bar)
Pcr
pressure measured using crusher method (kg/cm2)
V10
projectile velocity in 10 m
SD
standard deviation
difference between maximum and minimum value
number of measurements
(oC)
(m/s)
627
V10
2.3
Checking of projectile mechanical strength at fire
For initial loading of the projectile, lower weight of propellant was used and passage of
the projectile through paper screens in the distances 10 and 50 m was observed. Shot holes in
both screens were symmetrical and without other perforations. The test was repeated also
with increased propellant weight with the same result. Firing test of mechanical strength of
the projectile (2pcs) on fire was carried out with satisfying result.
High measured pressure (which exceeds the pressure at standard projectiles at +50C
see Table No.2) subjected the projectile construction to a considerable strain.
Table 2. The pressures achieved at strength test of the projectiles
Ppi
4012
4666
2.4
Pcr
3401
3905
Checking of ballistic assumptions of VOSA projectiles
Firing tests were carried out according to methodology KO-01, which is a set of methods
for testing of projectiles 20x102 and their components.
For verification of ballistic properties 16 projectiles 20x102 VOSA were manufactured,
of which 7 projectiles with penetrators 1/04 and 9 projectiles with penetrators 2/04. The
results obtained at testing are tabelled in Tables No. 3 and 4.
Table 3. Results of firings with penetrator 1/04
N
Ppi
Pcr
Bar
kg/cm
Pcr
V10
Bar
m/s
1
2
3
4
5
6
7
4309
4412
4362
4417
4401
4320
4290
3786
3942
3829
3914
3914
3843
3757
3713
3866
3755
3838
3838
3769
3684
1131.2
1140.8
1136.4
1145.5
1143.6
1146.0
1139.7
SD
MAX
MIN
7
4359
52.7
4417
4290
127
7
3855
70.4
3942
3757
185
7
3780
69
3866
3684
181
7
1140.5
5.31
1146.0
1131.2
14.8
628
Armor S3
Armor S5
Table 4. Results of firings with penetrator 2/04

N
1
2
3
4
5
6
7
8
9
Ppi
Bar
4238
4234
4261
4208
4295
4290
4289
4298
4316
Pcr
Kg/cm2
3786
3674
3647
3743
3786
3800
3772
3772
3857
Pcr
Bar
3713
3603
3577
3671
3713
3727
3699
3699
3782
V10
m/s
1141.3
1136.6
1136.4
1135.1
1141.3
1140.0
1129.4
1136.6
1141.6
SD
MAX
MIN
9
4270
36.2
4316
4208
108
9
3760
64.3
3857
3647
210
9
3687
63
3782
3577
206
9
1137.6
3.97
1141.6
1129.4
12.2
Armor S1
Armor S4
During the test, stability of the projectile on trajectory was observed by means of paper
screens positioned in distances 50m and 100m. Ovality of some shot holes was ascertained
and greater scattering of hits, than that prescribed for standard ammunition of the same
calibre.
A part of this test was evaluation of final effect of the projectile on armor. The armor
used was of thickness 50 mm and according to measured values of hardness 325 HV30, its
strength limit is 1100 MPa at minimum. The armor was positioned in the distance 50m from
the muzzle of the weapon.
Fig. 6 presents a view of the armor and details of hits marked S1,S3, S4 a S5. All hits
were cut out/excised by means of water beam under formation of cylinders of diameter 80
mm that were milled so, that approximately half of penetrator material was removed. Fig.7
shows hits S1 and S5, in S1 no remainders of penetrator were present. There was a reminder
of penetrator in S5, as seen in Fig. 7, but it fell out during milling the cut. Fig.7 shows also
hits S3 and S4 with remainders of penetrators. The cuts of hits with penetrator in the armor
were heated to 300C. Oxidation of steel surface improved visibility of penetrator
remainders.
629
Fig 6.
Overall view of the armor with individual hits
Fig 7.
Individual cuts of penetrator penetration
630
Hits S1 and S5 were very oblique and penetrator material reflected, or directly turned the
direction of its movement (S5). Hits S3 and S4 were also oblique and it is obvious from the
manner of entering the armor, that the penetrator did not turned its direction in the armor, but
already on contact of the projectile with the surface the penetrator was diverted from vertical
incidence. Real depth of penetration into the armor was thus greater than the achieved
penetration vertical to the surface of the armor, amounting at S4 - 26 mm and at S3 30
mm. Certain destruction of penetrator material on the front of penetration is usual, because
of consumption of penetrator material. In this case, however, a damage to the penetrator also
occured as a result of bending stress acting on slender penetrator.
3.
During firing tests, correctness of initial calculations was proved and the results obtained
correspond with initial requirements for projectile 20x102 VOSA.
From the results achieved, a low stability of the projectile on the trajectory is evident and
especially on incidence, due to insufficient compactness of projectile body and penetrator.
Further, from the metallographic cut insufficient perpendicularity of penetrator impact on
target plate was ascertained and its deflection from vertical. Assumed penetration depth on
ideal impact and penetration of penetrator through the armor is approximately 40mm.
For next stages of development the attention is paid to the possibility of increasing the
impact energy of the penetrator on the target by means of methods of reducing projectile
weight vs armor piercing core weight ratio and through appropriate choice of propellant
charge. Furthermore, it is necessary to deal with stability of projectile on trajectory and with
vertical impact of the projectile on target plate with vertical penetration of penetrator through
target plate.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
SN 39 51 05: Testing ammunition, Praha 1993

TP VD-239 57 III.: Procedure ballistic examinations,VV ZVS Brno, Brno 1977
P. PCHOUEK: Stela 20KE pro nboj 20x102Zprva o nvrhu a jeho oven, Prototypa ZM,
Brno 2004
V. HAMOUZ: Zprva k vrob penetrtor 7x65, UJP Praha, Zbraslav 2004
BOHUMIL PLHAL: Vnitn balistika een zvltnch pklad, VA Brno 2000.
JAN KUSK, JOSEF ROUSEK, JI MIKULK: Metodick postup KO-01 soubor metodik zkouek
nboj 20x102 a jejich dl, ZVI Vsetn , Vsetn 2003
631
SOME DETONATION CHARACTERISTICS OF AMMONIUM

NITRATE-ALUMINUM-DOUBLE BASE PROPELLANT MIXTURES
Katarzyna Lipiska*, Marek Lipiski*, and Andrzej Maranda**
* Institute of Industrial Organic Chemistry, 6 Annopol St., 03-236 Warszawa, PL
** Military University of Technology, 2 Kaliskiego St., 00-908 Warszawa, PL
Abstract:
The influence of double base propellants (DBP) on detonation characteristics of
ammonium nitrate-aluminium mixtures was assessed. Investigated explosive mixtures
contained 20 and 40% of the propellants. We determined velocity of detonation,
comparative strength and intensity of blast waves (peak blast overpressures) generated
by the explosion of the mixtures. Positive phase impulses were also calculated. Some
parameters of the compositions increased along with the propellant content.
Keywords:
1.
detonation velocity , blast wave, ballistic mortar, peak overpressure,

positive phase impulse
INTRODUCTION
The performance potential of an explosive, especially non-homogeneous explosive

mixtures, cannot be described by a single parameter. It is determined by detonation velocity
of the explosive, the amount of gas liberated per unit weight and the energy evolved in the
detonation process. The detonation velocity and density of the explosive are specific
physical parameters and they both require no conventions. Classic tests for comparative
strength determination are lead block and ballistic mortar tests. In both cases the explosive is
enclosed in a confined space so that the measured parameter is the work of explosive
products in a borehole. The main disadvantage of both methods is that the quantity of the
sample is rather small. Practical methods for determining the strength of explosives that are
of significant heterogeneity, as ammonium nitrate-aluminium-double base propellant
mixtures, require larger samples. Working capacity of ammonals with demilitarised double
base propellant content was determined by blast wave method [1]. In the case of investigated
explosives the main advantage of the applied method is relatively big mass of the sample.
Conditions of experiments enable more complete reactions among components and
influences repeatability of measurements.
Our research is to support the development of methods of double base propellant
disposal [2, 3, 5]. Double base propellants contain nitrocellulose of various nitration levels,
nitroglycerine and inert plasticizers to modify either the flame temperature or the physical
properties of the propellant. The idea to use the excess military propellants in commercial
explosives seems to be attractive and economically justified [4].
Commercial explosives should have an oxygen balance close to zero in order to
minimize the amount of toxic gases, particularly carbon monoxide and nitrous gases, which
are evolved in the fumes. All investigated explosive mixtures were zero balanced.
632
2.
EXPERIMENTS
2.1
Materials
2.1.1. Double base propellants

We used two different kinds of double base propellants, identified as propellant I and
propellant II. They were both in the form of squares of 22 mm, having a thickness of
0,25 mm. Compositions of the propellants differed primarily in nitroglycerine content.
Propellant I contained 40 % of nitroglycerine. The nitroglycerine content in propellant II was
only 26 %, but dinitrotoluene (9 %) was present. Oxygen balance of propellant I was
-26,77 % and its detonation velocity in steel pipe (ID 36 mm, OD 42 mm) was 4290 m/s at
746 kg/m3 density. Oxygen balance of propellant II was lower, -47,17 %. Its detonation
velocity, measured in the same conditions, was 3380 m/s at 836 kg/m3 density.
2.1.2. Aluminium
Two types of metallic aluminium powder were applied. Their refinement, determined
by IPS-U System Infrared Particle Sizer, is shown (Fig. 1, 2). Most of particles of flaked
aluminium were of size between 10 and 35 m. The dimensions of aluminium powder
particles were much bigger mainly between 50 and 250 m. Due to differences in particle
sizes aluminium powders differed in development of their surfaces.
Fig 1.
Refinement of flaked aluminium
Fig 2.
Refinement of aluminium powder
633
2.1.3. Explosive mixtures

We examined the influence of double base propellants on detonation parameters of
ammonium nitrate-aluminium mixtures. This type of explosives is known as ammonals.
Ammonals were prepared by mixing crystalline ammonium nitrate (particle size below
1mm) with fine metallic aluminium and then with double base propellants in given ratio. All
explosive mixtures were zero balanced. Their compositions are given in table 1. Type of
propellant is marked by propellant content.
Table 1. Compositions of ammonium nitrate-aluminium-propellant mixtures
Explosive Component [%]
amonium flaked
nitrate
aluminum
1
70,2
9,8
2
58,8
1,2
3
70,2
4
58,8
5
74
6
6
74
-
2.2
aluminum
powder
9,8
1,2
6
propellant
20 (I)
40 (I)
20 (I)
40 (I)
20 (II)
20 (II)
Methods
We measured the intensity of blast waves generated by the explosion of ammonal/double

base propellant mixtures. The mixtures were loaded into paper tubes having an inside
diameter of 45 mm and an outer diameter of 50 mm. The weight of every explosive was
350 g and density was 1000 kg/m3. The explosives were initiated using boosters HT-14. Two
PCB quartz pressure transducers were located: first one 2 m and the second one 3 m away
from the centre of charges. The charges were hung 1,5 m above the ground. The gauges
registered overpressure in the transient shock waves, sliding on the working surface of the
device.
Simultaneously, detonation velocity tests were conducted. Fibre optic cable sensors were
placed inside the explosive. Detonation velocity was measured by means of EXPLOMETFO-2000.
Moreover we carried out ballistic mortal tests. We employed a standard procedure.
3.

3.3
Peak blast overpressure
Peak blast overpressures obtained for the explosives were collected in table 2.
Comparative tests were made for TNT. Peak blast overpressure of TNT explosion products
was 133,55 kPa for 2 m distance from the charge and 53,97 kPa for 3 m. Mass of TNT was
also 350 g, but charge density was lower, only 840 kg/m3. That was connected with the fact
that we used TNT in the form of pellets.
A typical blast overpressure signal, obtained for explosive 2, was presented in Figure 3.
634
Fig 3.
Blast overpressure signal for explosive 2
Table 2. Peak blast overpressures for investigated mixtures

Explosive Peak overpressure
for 2 m [kPa]
1
135,18
Peak overpressure
for 3 m [kPa]
123,78
48,93
135,18
58,93
93,65
42,62
5
6
119,27
106,15
39,74
37,07
53,37
The peak pressures generated by 2 and 4 were smaller than those for mixtures 1 and 4.
Peak blast overpressures for mixtures 1 and 4 resembled peak overpressures obtained for
TNT. Peak overpressures measured for mixtures 5 and 6 containing 20% of propellant II
differed because of aluminium type higher results were obtained for flaked aluminium.
3.4
Blast wave impulse
The positive phase impulses were calculated by integrating the blast overpressure signals
until a time when the pressure falls below atmospheric pressure. They were shown in table 3.
That impulses calculated for TNT charges were, respectively, 59,38 Pa s for 2 m and 36,13
Pa s for 3 m.
Both values obtained at the first and the second transducer (2 m and 3 m) were the
highest for explosives 1 and 3. Both in the case of overpressures and wave impulses the
lowest values were obtained for mixtures containing 1,2 % of Al and 40 % of propellant I
and mixtures containing 6 % of aluminium of 20 % of propellant II. Presumably this results
from the fact that so small amount of aluminium sensitizes the mixture only in the slight
degree and the experiments were conducted in the conditions close to critical.
635
Table 3. Positive phase impulses for investigated mixtures
3.5
Explosive Positive phase impulse

for 2 m [Pa s]
1
59,66
Positive phase impulse

for 3 m [Pa s]
53,40
32,72
61,99
38,19
44,55
26,53
5
6
48,80
42,88
25,67
23,39
38,06
Ballistic mortar
Practical test for comparative strength determination is ballistic mortar test. In this case
the explosive is enclosed in a confined space so that the measured parameter is the work of
explosive products in a borehole. The main disadvantage of the method is that the quantity of
the sample is small. The ballistic mortar test values were shown in table 4.
Table 4. Ballistic mortar test values
Explosive Ballistic Mortar Test
[%]
1
98
2
77
3
98
4
84
5
80
6
73
Obtained values were definitely higher for explosives with lower content of double
base propellant. The values obtained for mixtures with 20% of propellant II were lower than
mixtures with 20% of propellant I. Ballistic mortar test values for explosives 1 and 3, which
differed in aluminium powder form, are similar. This can show that the reaction between
aluminium and oxygen forming during exothermic ammonium nitrate decomposition is of
great importance in the conditions of measurements (small mass of the sample).
3.6
Detonation velocity
Detonation velocity and blast wave measurements were conducted at the same time.
Detonation velocity of propellant I in paper tube was 4050 m/s at density 825 kg/m3.
Propellant II did not detonate in the condition of the experiment. Obtained values were
shown in table 5.
All charges were very lightly confined and their diameter was rather small (only
45 mm). In the case of investigated mixtures the achievement and propagation of detonation
is very complex. The large fraction of the total energy of the mixtures derives from reactions
between components.
636
Table 5. Detonation velocities in paper tubes (density 1000 kg/m3)

Explosive Detonation velocity
[m/s]
3200
1
3650
2
3100
3
3170
4
2950
5
1400
6
A main factor, which determines detonation velocity, is the amount of energy generating
during chemical reactions of the propellant in the detonation wave. There were no
differences in densities and it made comparison among the explosive mixtures easier. It can
be noticed that detonation velocities are higher in the case of explosives 2 and 4. They
contained 40 % of propellant. At this propellant content refinement of aluminium powder did
not affect measured parameter. However, in the case of explosive 5 and 6, containing 20 %
of propellant II, there is a noticeable distinction between detonation velocity of explosive 5
and 6. That difference meant that for explosive 6 the diameter of the charge was near critical
diameter.
4.
SUMMARY
Investigated mixtures contained two explosives: non-ideal (ammonal) and ideal

(propellant). Depending on experimental conditions (mass of the sample, strength and mass
of envelope) predominant chemical reaction, which decides on values of measured
detonation parameters, is co-reacting between ammonium nitrate decomposition products
and aluminum or exothermic decomposition of propellant. That is why we could not get any
correlation between measured detonation velocities and performances of the explosives.
Thus precise explanation of reaction mechanism proceeding in non-ideal explosives
containing high energetic individual explosive requires more experiments on other systems.
The results are also of practical meaning. Double base rocket propellants can be used as
ingredients of ammonals, increasing their detonation parameters. The incorporation of
double base propellants into explosives for civil uses offers the most controlled, safe and
relatively environmentally friendly method of disposing military propellants. Moreover, the
use of these propellants as ingredients of commercial blasting agents would be an
economical means of their disposing and in the process, producing industrial explosives
exhibiting a favourable cost for performance.
637
REFERENCES
[1]
[2]
[3]
[4]
[5]
K. BUCHALIK , W. WITKOWSKI, R. TRBISKI: Determination of TNT equivalent using

the blast wave measurements, Proc. of the IV Seminar, University of Pardubice, p. 34-39, 2001
A. N. AFANASENKOV, W. W. GALKIN: Ispolzovanie utiliziruemykh vzryvchatykh
materialov dlya povysheniya ffektivnosti destviya vzryva, Fizika Goreniya i Vzryva, 2 (37),
p. 131-134, 2001
E. F. OKHRIMENKO, N. G. IBRAGIMOV, E. KH. AFIATULLOV, L. S. LUKIN, YU. M.
YUKOV, I. P. IVANOVA: Use of Powders in National Economy, Conversion Concepts for
Commercial Applications and Disposal Technologies of Energetic Systems, Kluwer Academic
Publishers, Dordrecht, p. 213-219, 1997
O. MACHACEK, G. R. ECK: Waste propellants and smokeless powders as ingredients in
commercial explosives, Proc. of 23th International Annual Conference of ICT, Karlsruhe,
p. 11-1 11-14, 1992
K. LIPISKA, M. LIPISKI, A. MARANDA: Influence of double base propellants on
detonation velocity of mining blasting agents, Polish Journal of Applied Chemistry XLVII 3,
p. 221-225, 2003
638
INACTIVE MODIFICATION OF AMMONIUM NITRATE

AND THE DETERMINATION OF ITS ACTIVITY
AS OXIDIZER IN EXPLOSIVES
Liu Dabin Tang, and Shuangling Ni Ouqi
Nanjing University of Science& Technology, Nanjing, China 210094
Abstract:
In China, in order to guard against terrifying activities by the explosive made from
ammonium nitrate, Chinese government prescribes that agriculture ammonium nitrate is
not allowed to put in market before it is modified so well that explosive can not be made
from it through simple mixing. The method of modification is changing ammonium nitrate
to compound fertilizer by adding calcium carbonate, urea, ammonium sulfate, etc.
A testing method similar to UN cap sensitivity test was established primarily to assess the
modification. Based on the content of ammonium nitrate in modified compound, diesel oil
and wooden powder are mixed into it homogeneously by the proportion of ammonium
nitrate/diesel oil/wooden powder=92%/2%/6%. The mixture is restricted in an iron
tube( 80mm160mm1mm) then initiated by 8th instantaneous electric detonator.
A witness and steel base plate with thickness of 1mm as verifying plate is put underside
the tube to estimate whether the detonation is complete.
Key words: ammonium nitrate, explosive, modification, test
1.
INTRODUCTION
Ammonium nitrate is both an oxidant and a kind of farm fertilizer, and at the same time
also a basic material of explosive mixture. It becomes explosive after mixed with combustible
agent like diesel oil, wooden powder, etc. So it may be a tool with which terrorists go in for
destruction activities. With the happening of a series world-shocked explosive events, such as
the explosive event at Yulin, Shanxi province, China and the tragedy one at Shi Jiazhuang,
China etc, harmfulness of ammonium nitrate arouses high attention all over the world.
In order to safeguard public security, Chinese government put ammonium nitrate into two
categories: industrial one and farming one. The former is strictly controlled as dangerous
goods by police system, while the latter can be saled as common goods on market only if it is
modified not to become explosive through simple mix and not to be restored with a simple
method.
To modify ammonium nitrate, many institutes and factories have done a lot of work and
finished lots of formula to eliminate or reduce the explosion effect of mixture of ammonium
nitrate with combustible agent. The technical way is mainly changing ammonium nitrate into
compound fertilizer. That is adding CaCO3, urea, NH4H2PO4, (NH4)2SO4 to ammonium
nitrate, and improving its crystal density technically. So far, there have been dozens of formula
on experiment.
639
2.
EXPERIMENTAL
At present, we need a testing method to assess if the modified ammonium nitrate as the
oxidant in explosive is insensitive enough. Nowadays a testing method is primarily built up by
which we can assess the effect of modification of ammonium nitrate. The method is described
as follows:
Sample
drying and
sifting
determine the content
of nitrate in the sample
add combustible agent (ammonium nitrate /diesel oil
/wooden powder=92% / 2% / 6%)
Determining the cap sensitivity
of the sample
in the specified conditions.
The Recommendations on the Transport of Dangerous Goods of United Nations-Manual
of Tests and Criteria Test 5 (a) recommends a method and a apparatus by which cap sensitivity
test can be done. In our method, we cite most of the apparatus, materials and the criteria from
UN cap sensitivity test, only change the cardboard tube to 80mm160mm1mm steel tube.
The apparatus is as follows:
c
f
g
Fig 1.
Test apparatus.
c - standard detonator d - test substance
e -steel tube f - witness plate of steel g - steel ring
640
Fig 2.
The witness plate and the sample tube after not complete explosion.
Fig 3.
The remains of the witness plate after

complete explosion
3.
RESULTS
The test must be done in specified condition. Some factor influence test result
fundamentally: granularity of sample, sample density, displaying time of explosive mixture
and standard detonator. The details are as follows:
(1)
Modified ammonium nitrate must be grinded into powder with the
granularity as follows: all going through 20-mesh sieve and then 50%
going through 60-mesh sieve.
(2)
Density of sample must be 0.85~0.89g/cm3. The tests show that explosive

mixture within this density range has the strongest cap sensitivity.
(3)
The content of the ammonium nitrate in the dried sample is determined.

Then modified ammonium nitrate, diesel oil, wooden powder are mixed
for 40mins into explosive mixture in the proportion of ammonium nitrate
/diesel oil /wooden powder=92% / 2% / 6%, then the mixture is sealed and
displayed for 48h. Finally, the mixture is tested. The aim of displaying it is
to make diesel oil fully diffuse into ammonium nitrate. The aim of sealing
it is to prevent volatilization of diesel oil.
(4)
8# instantaneous copper-shell electric detonator appointed as the standard

detonator is used for initiation. The standard detonator is inserted centrally
into the top of the sample. The detonator is then fired from a safe position
and the witness examined. The test is conducted three times unless
detonation of the sample occurs.
641
4.
RESULTS
If the sample is not detonated fully in the above conditions, we think it is not so easy to be
used for explosive by terrorists. So in some degree, modified ammonium nitrate can guard
against terrifying attack in some extent.
But in the theory, modified ammonium nitrate with combustible agent added in it is in
nature an explosive mixture. It can detonate under the condition of strong enough initiation
and restriction.
The method above of modification of ammonium nitrate is that ammonium nitrate is
changed into compound fertilizer. In the research of formula for the modification, some other
factors should be considered except its insensitivity, such as its cost of production, the
additives effect on soil and environment, the additives accommodation to production line of
ammonium nitrate, etc. So selection of the additive must be careful.
642
EFFICIENCY OF APPARATUS FOR TESTING SENSITIVITY

OF EM TO ELECTROSTATIC DISCHARGE
Ji Majzlk, and Ji Strnad
Department of Theory & Technology of Explosives, University of Pardubice,
CZ 532 10 Pardubice, Czech Republic
Abstract:
Expert literature describes various apparatuses for testing sensitivity of energetic
materials (EM) to effects of electrostatic discharge. Besides, also various
methodological and interpretation procedures have been extensively or in secondary
issues introduced, which hinders comparisons of sensitivity of comparable EM. For
instance, Kirshenbaum [1] gathered available data of initiation energies of discharge for
Lead Azide RD1333, which vary in the interval from 4.1010 to 2.102 Joule. This
contribution presents besides the analysis of efficiency of the test circuit working in
Short (oscillation) regime also several rules obeyed in the methodology of testing of a
wide varied of EM by means of ESZ apparatus at DTTX. The paper has been written
with the aim to contribute to more extensive standardisation of sensitivity tests of
various EM by means of condenser pulse devices.
Keywords:
1.
electrostatic sensitivity, explosives, pyrotechnics
INTRODUCTION
A considerable number of electrostatic test devices work in the Fixed Gap regime,
whose advantage is inter alia its electro-mechanic simplicity, versatility and easy operation.
A similar design is used also in the Small Scale test device of the type of ESZ at DTTX; its
simplified scheme is presented in Fig. 1. Charging of the working battery of condensers from
a HV source as well as the discharge of the circuit, whose part is the spark gap with tested
sample, are carried out by means of a mechanic switch. Conceptually, the whole apparatus
fulfils the requirements of operation in a widely variable regime.
Standard testing of electrostatic sensitivity of secondary explosives requires a fast
increase of liberated output, (dP/dt 107 W/s), which is why the values of loss resistances
of circuit as well as its parasitic inductance are limited to the attainable minimum.
The discharge into the test circuit is short and oscillating, and it is usually presumed that
the Joulean energy of the electric discharge consumed in the medium of the tested sample,
ESA, is equal to the energy of charge at the battery of working condensers, E0, according to
Eq. (1).
ESA = E0 = 0.5 C*U2
(1)
643
Standard sensitivity tests of readily inflammable loose materials that react with
atmospheric oxygen, pyrotechnic materials and primary explosives are usually carried out at
1 k < RP < 1 M. In this variant it is possible to measure the time course of electrically
active output consumed in the medium of tested sample by means of a fairly trouble-free
voltampere method see e.g. [1], [2], [3], [4], [5].
CH1
HV SOURCE
3 10 kV
CH2
HV PROBE
100 M /3 pF
MANUAL
SWITCH
u(t)
40 M
ANODE
CC + CM
EM
SAMPLE
RP
0 10 k
C
i(t)
CATHODE
COND. BANK
0,03 360 nF
CVR
Fig. 1. Function scheme of Small Scale test device ESZ KTTV. The apparatus is
briefly described in Refs [2], [3].
2.
OSCILLATION REGIME
2.1
Efficiency of the system
The tests of secondary explosives and substances whose activation needs an

instantaneous evolution of a high power necessitate minimization of both the inductance L
and all the parasitic Joulean losses of discharging circuit.
This means that the aim is to reach, as far as possible, a loss-free function of the
discharge circuit and thus approach the fulfilment of the requirement defined in Eq. (1).
However, any real electro-mechanic apparatus working in the area of high energies in the
frequency region of 10 Mc/s (see Fig. 1) has to take into account certain energy losses.
Therefore, in order to make the test results reliable, it is necessary to identify the losses of
the circuit and, at the same time, evaluate the energy efficiency of the given system.
Nevertheless, there are only sporadic results of evaluation of this parameter published in
literature. Ref. [1] gives the efficiency of discharge circuit in the interval of 80 - 90 %, the
active component of resistance of the circuit being 0.15 . Similarly, Ref. [6] gives the
644
efficiency of ca 88 % for C0 = 1 nF, U0 = 5 kV, R = 0.15 . The problem thus delimited is

also treated in a contribution to Proceedings [7].
The following text concentrates on methodology of determination of efficiency of the
apparatus working in Short (oscillation) regime.
In the Short regime the time dependence of the current through circuit and of the voltage
at the anode of spark gap resemble the diagrams presented in Fig. 2. However, with some
apparatuses it is difficult to measure the high current with sufficient accuracy in this regime.
The reasons are the swirling currents in steel parts of the sandwich structure, parasitic
inductance, strong electromagnetic fields, and ohmic resistance of the set cathode CWR
earth return to zero point.
The circuit current according to Eq.(5)
i1r
U1
USH1
u;i
U 1
U 2
to
T
Time
The voltage drop measured by CWR
The voltage signal measured by HV probe
Fig. 2. Time courses of the current in the circuit and the voltage at spark gap
of the apparatus working in the regime of oscillation discharge
The energetic efficiency of discharging circuit can approximately be expressed by
relationship (2):
R = (R S A /R) * 100
[%]
(2)
where the sample resistance in the first quarter-period of discharge RSA is given by the
ratio of the voltage drop at sample and the flowing current; with regard to notation of the
individual quantities depicted in Fig. 2:
R SA = (U 1 U SH1 )/i 1r
(3)
R represents, according to Eq. (4), a sum of real losses of the discharged circuit:
R = (R SW + R HF + R DC + R SA +R SH ) = 2Lf * ln(i 1 /i 2 )
(4)
where (i 1 /i 2 ) = (U S H 1 /U S H 2 ) and R SW , R HF , R DC ,, R SA , RSH mean the

resistance of switch, the high-frequency and DC losses, resistance of the channel in the
sample, and the resistance of CVR.
645
Efficiency of ESZ
(%
100
75
50
Efficiency after Eq. (2)
25
Efficiency after Eq. (6)
0
1
10
100
1000
10000
100000
Electrostatic Charge Energy Eo (mJ)

Fig. 3. Efficiency R and J of ESZ KTTV (PETN D, grain size 0.125 mm).
The current in the oscillating circuit is defined by Eq. (5):
i r = ( I 0 * e R t / 2 L ) sin t
(5)
where I0 = U0 / L .
The value of the first maximum of the corrected time course of oscillating current at the
time t = t0 + T/4 is introduced into Eq. (3).
The second, more promising way of evaluation of the testing apparatus is based on
comparison of the Joule integral of the measured current-time course in the circuit corrected
with respect to Eq. (5) with the energy of the charged capacitor according to Eq. (1); the
resulting product J being defined by Eq. (6):
J =
[ ( R i dt ) / E ] * 100
SA
[%; ; A; s; J]
646
(6)
2.2
Efficiency of ESZ KTTV
The time courses of current and voltage and, subsequently, the efficiency of the
apparatus ESZ KTTV [2], [3] were evaluated by the above-described procedure using PETN
D, RDX and TEX as the tested explosives whose crystals have the dimensions below 0.125
mm, 0.1 mm and 0.3 mm, respectively.
The voltage levels were chosen in the interval of U0 = 5 7 10 kV. The methodology
of both evaluation procedures can now be judged also on the basis of resulting data of the
analysis of R and J graphically, as presented in Fig 4 for illustration. In reality, the
efficiency of circuit J is somewhat higher than that calculated from Eq. (6), because with
decreasing amplitude of current beats the resistance of inter-electrode medium is evidently
increased during the subsequent oscillations.
A simple analysis of data J = f(EO) leads to a linear equatio, Eq. (7):
J = A* EO + B
[ %; mJ]
(7)
Table 1. Constants A; B of Eq. 7; valuable for energy E0 region

of 5 mJ EO 1000 mJ
Substance tested
Energy E50
[mJ]
[%/mJ]
[%]
PETN D, grain size 0.123 mm
51
0.02079
69.46
70.52
RDX, grain size 0.100 mm
86
0.02240
65.61
67.63
TEX, grain size < 0.3 mm
193
0.01183
70.11
72.39
0.00375
62.58
Crushed glass, particles < 0.3 mm
Efficien. at E50
J
[%]
For instance, a correction of energy value E50, which was determined along with the
efficiency J for the loose explosive PETN D, provides the corrected value of E50* = 35.7
mJ.
A way to attaining a high efficiency of the series resonance circuit (Fig. 1) consists in
lowering of the ratio of the losses represented by the parameters (RSW + RHF + RDC + RSH ) to
RSA to obtain an acceptable minimum. For better understanding of the reasons of formation
of the individual Joulean losses in the circuit, Fig. 5 represents the dependences of
resistances R ans RSA on the capacity of battery of source condensers, i.e. by through Eq.
(5) also on the current flowing through the circuit
From the diagram in Fig. 4 and data given in Ref. [4] it is obvious that the efficiency of
apparatus, J, in the area of low energies (E0 < 30 mL; C0 1 nF; resonance frequency of
circuit f > 4 Mc/s) varies at a level of 60 -70 %, which to a certain extent is due to skin
effect, losses in dielectric of the condensers used, and field losses.
647
On the other hand, the mechanism of heat energy exchange increases the resistance of
channel in the sample, which obviously contributes to keeping the efficiency of apparatus at
an acceptable level.
100
Resist.
R;
RSA
U0 = 7 kV;
[]
10
RSA
U0 = 5 kV;
R
RSA
U0 = 10 kV;
R
RSA
0,1
0,1
10
Capacity
C0
100
1000
[n F]
Fig. 4. Total loss resistance of discharge circuit (Fig. 1) R and resistance of channel
in the medium of tested sample PETN D RSA at time t = t0 + T/4 depending
on capacity C0
With increasing test voltage U0 at constant capacity of the circuit, the time and trajectory
of premature contact of electrodes in mechanical switch are prolonged, which results in an
increase in RSW and, hence, also decrease in efficiency of the discharge.
648
This finding follows from data of Table 2, which gives the mean values of J depending
on the test voltage.
Table 2. Table 2. Effect of voltage U0 on efficiency of discharge circuit.

Energy 5 mJ < E0 < 2 J
Voltage
U0 [kV]
Apparatus Efficiency
PETN D
RDX
J
TEX
[%]
GLASS
81
74
74
70
78
72
74
65
10
76
70
77
66
With C0 increasing above ca 20 nF at constant U0, the amplitude and time of passage of
current beats through the tested sample are increased. The volume of channel formed in the
medium of sample is increased, its mean resistance is decreased and, simultaneously, the
efficiency of the system is decreased too see the structure of Eqs (3) and (4).
In the region of E0 > 1 J, the effect of active component of resistance of CVR makes
itself felt. For most secondary explosives tested by means of ESZ KTTV this region
represents a region of supercritical energies. At U0 = 10 kV; C0 = 256 nF, the resistances
attain the values of R = 0.368 , RSA = 0.187 . The difference between these two values
(R RSA) represents nearly the DC resistance of the short-circuited circuit, which can be
measured e.g. by means of a DC ohmmeter. In this regime, the first maximum of the current
passing through the circuit attains the value of ca 5000 A. For the efficiency of the circuit it
is favourable that the resistance of conducting trajectory between the electrodes of
mechanical switch is negligible as compared with other resistances of the circuit.
After exceeding the critical energy of the sample, the primary decomposition processes
are preferably realized in the contact of particles EM surface with the plasmatic discharge
channel. Hence the pressure effects of generated gases can cause a diffusion of plasma into
the free volume of sample. As a consequence of growing thermal energy exchange, both the
ohmic resistance RSA and the efficiency of discharge are increased see Eq. (2). This effect
cannot operate in inert materials, which can be the reason of the considerably lower
efficiency of discharge into crushed glass see Table 2.
2.3
Working procedure
If RP = 0, and the capacity of testing condenser C0 is (as in Fig. 1) comparable with the
capacity of the connecting cable and other capacities of the instrument (CC + CM ) , then we
have to respect the principle of conservation of charge, where the charge
Q = C0*U0 = C *U1, and
C = CO + CC + CM
(8)
649
The value of energy of charge after the mechanical switch has switched over is then
given by Eq. (9):
E1 = 0,5*CO2U02/C
(9)
When applying the Bruceton procedure, we must in advance determine the field of
activation energies E = Emax Emin and then calculate a series of equidistant energy steps,
Ei, with overall span of E = 10 Ei. The respective voltage levels, which cover the energy
field E , are then defined by Eq. (10):
Ui = E i ( C ) / 0 . 5 C O 2
(10)
Thereupon follows the usual test and evaluation procedure. The calculated value of
energy E50 can then be corrected, if necessary, by means of the coefficient expressing the
efficiency of the discharge circuit see Table 1.
3.
CONCLUSION
The paper suggests a method for evaluation of energy efficiency of a testing RLC circuit
working in the Fixed Gap oscillation regime. The method was used to evaluate a number of
functional parameters of ESZ KTTV.
It appears that the per cent ratio between Joulean energy consumed immediately in the
medium of the EM sample tested and the charge energy calculated from Eq. (1) depends on
the adjusted working conditions of the discharge circuit and on the type of the tested primary
explosive. A complicated interplay of various factors affecting the course and results of
individual tests results in a certain scattering of the measured recordings of i(t) and u(t). As a
result, the values of J vary in an interval from 55 to 95 %. In the vicinity of the testing
conditions where PETN D, RDX a TEX are activated the efficiency of apparatus J reaches
ca 70 %.
Significant factors lowering the efficiency of apparatus include the HF losses of circuit
and the resistance of plasmatic channel formed between approaching electrodes of
mechanical switch. In constructing a test discharge instrument working in Fixed Gap
Oscillating regime there always will be antagonistic requirements: on the one hand the
highest possible efficiency, and on the other hand versatility, a wide range of test conditions,
safety and simple operation.
650
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
KIRSHENBAUM, M. S.: Electrostatic Sensitivity. In: Energetic Materials, H. D. Fair and R. F.

Walker (Eds.), Vol. 2, Chapter 5, Plenum Press, New York, NY, 1977.
STRNAD, J., MAJZLK, J.: Determination of Electrostatic Spark Sensitivity of Energetic
th
Materials, 4 . Seminar DTTE, Pardubice, 2001, pp 303 - 307
STRNAD J., MAJZLK J.: DTTE Documentation of ESZ KTTV, 2001
KO J., ZEMAN S., MAJZLK J., STRNAD J.: Notice to Determination of the Electric Spark
th
Sensitivity of Energetic Materials. Proc. of 5 Seminar: New Trends in Research of Energetic
Materials, University of Pardubice, CR, 2002, pp 110 - 118
JARMAN D., PRINSE W., BOUMA R.: Electrostatic Discharge Initiation of CL-20: Effect of
th
Discharge Time and Spark Energy. Proc. of 34 Conference: Energetic Materials, Reactions of
Propellants, Explosives and Pyrotechnics, Karlsruhe, FRG, 2003, P 71.
PRIEDE, T.: Initiation of Explosion in Gases, Ph. D. Thesis, University of London, 1958
MAJZLK, J., STRNAD, J.: Efficiency of Apparatus for Testing Electrostatic Discharge
th
Sensitivity of Energetic Materials. Proc. of 35 Conference: Energetic Materials, Structure and
Properties, Karlsruhe, FRG, 2004, P 113
651
DEVELOPMENT
OF INSENSITIVE HIGH EXPLOSIVES NITROGUANIDINE
Andrzej Maranda*, Adam Kubecki*, Jerzy Nowaczewski*, Andrzej Orzechowski**,
and Dorota Powaa**
* Faculty of Mechatronics, Military University of Technology, 00-908 Warsaw, POLAND
* Institute of Industrial Organic Chemistry, 03-236 Warszawa, POLAND
Abstract:
In the paper an insensitive high explosive nitroguanidine (NQ) is investigated.
Obtaining technology of nitroguanidine is described and investigation of its explosive
properties is carried out. Methods of NQ crystallization from various solvents are
described. Spherical forms of NQ crystals were obtained, grain morphology dissolution,
bulk density and tap bulk density were determined Differential thermal patterns are
determined and spectrum in infra-red radiation is investigated. Density of obtained NQ,
sensitivity to friction and impact were inspected. Mixtures of crystalline forms of NQ
with high energy explosives (HE) such as hexogene (RDX) and oktogene (HMX) were
prepared. The detonation velocity for selected mixtures was determined.
Key words:
1.
nitroguanidine, insensitive high explosives, detonation velocity
INTRODUCTION
At several conferences and in the latest years literature the growing interest in high
energy (HE) mixtures of low sensitivity to external stimuli is observed [14]. The attention is
also paid to develop of new low-sensitive HE such as TATB, TACOT, PYX, FOX-7 and
quite intensive efforts in investigation of their characteristic features are made for. This
interest arises from the care about safety of people who produce HE, deal with elaboration of
ammunition, are busy at transport and by applications.
To the most popular methods of decreasing the sensitivity of explosives belong methods
of covering of the crystals of high explosives like hexogene (RDX), octogene (HMX) and
pentaerythritol tetranitrate (PETN) with the wax or polymers (i.e. phlegmatizing agents) or
by adding to these HEs components of low sensitivity. But phlegmatizators have
considerably smaller density than HE and have no explosive properties. Therefore the
detonation parameters of obtained in such a way explosive mixtures decrease, in comparison
to the base HE. Among the low-sensitive homogeneous explosives the NQ and
nitrotriazolone (NTO) are the cheapest and the easiest to obtain. However, an important
hindrance in use of these explosives on a larger scale, in mixtures with other HE, is their
improper crystalline form they are obtained in a form of long needles. Only by gaining the
spherical form of crystals that leads to much higher density than in previously obtained
specimens have caused rise of interest in their application. The use of NQ or NTO instead of
wax or phlegmatizing polymers has this advantage that mass density of these explosives is
similar to density of base HE and by adding them, even in large quantities, no decrease of
detonative properties of resultant mixtures is observed.
652
2.
SYNTHESIS OF NITROGUANIDINE (NQ)
NQ was obtained in dehydration reaction of guanidine nitrate (QN) by use of

concentrated sulfur acid in accordance with the recipe taken from Organic Syntheses [5].
Synthesis proceeded in following way: to cooled sulfur acid, during mixing by hand, small
portions of QN were added, so that the temperature does not exceed 20 C. After adding all
the prescribed amount of QN the content of reactor was periodically mixed for over three
hours and the mixture was left at room temperature till the next day (about 20 hours). Then
the mixture was poured out to about of 2 dm3 water with ice. The precipitated sediment of
NQ was then filtered, bathed with cold water and crystallized from as little amount of water
as possible (1.5 1.7 dm3), which was heated boil. A water solution of NQ was cooled
slowly and left till next day. From 200 g of QN of about of 140149 g of NQ was obtained
in form of long needles (about of 85 % of efficiency).
2.
METHODS OF CRYSTALLIZATION OF NQ
To crystallize of NQ various solvents were used: glycerin (G), ethylene glycol (EG),
diethylene glycol (DEG), formamide (F), dimethyloformamide (DMF), dimethylosulfoxide
(DMSO), N-methylopyrrolide (NMP) and mixtures of some of them [68]. The purpose of
these searches was to find of suitable solvent and to match conditions of crystallization by
which an spherical form of product of high density will be obtained. In most cases the
crystallization proceeded in following way: into a reactor the solvent and prescribed portion
NQ were put in, then the mixture was heated until the crystals melted. Next the solution was
cooled in water/ice bath by intensive mixing. In consecutive variants the speed of cooling
and rotational velocity of mixer was changed. In several cases this operation was modified in
such a way: during cooling the nodular form of NQ was added, scheduled to be the
crystallization nuclei of or some cases, extra component was introduced, that not dissolve
NQ. After the crystallization ended of, the obtained sediment was filtered under decreased
pressure, rinsed with acetone or ethanol and dried at a room temperature.
Not in every case the satisfactory crystalline form of product was obtained. Negative
effects of crystallizations of NQ, i.e. lack of crystals have been revealed in the cases when
as the solvents, water, glycerin, ethylene glycol and formamide were used (Fig. 1).
a
Fig 1.
Nitroguanidine crystals obtained after crystallization from:

a H2O , b - formamide
653
In the Table 1 parameters of crystallization procedures are set up. The cases in which the
final product had the form of needle again, were not included.
Table 1. Parameters of NQ crystallization
Solvents
Features
GDE
DMF.
2
NMP.
2
NMP.
3
Concentration of NQ [%]
4.7
25
14
18
28
25
37.5
42
Maximum temperature [ C]
100
85
100
100
85
85
85
85
Cooling temperature [ C]
20
20
Speed of cooling [C/min.]
Speed of mixing [r.p.m]
250
320
250
250
250
250
275
250
Time of cooling [hours]
1.5
2.5
0.25
0.25
Melting temperature [ C]
232.8
237.9
231.4
233.8
237.9
236.8
231.1
232.7
Bulk density [g/cm3]
0.25
0.47
0.64
0.68
0.79
0.79
0.54
0.70
Tap bulk density [g/cm3]
0.36
0.57
0.72
0.77
0.87
0.86
0.59
0.74
66
54
67
64
61
72
64
43
Crystallization
[%]
effectiveness
DMF/GDE DMF/GDE
2
3
DMSO DMSO
.2
.3
In the Fig 2, photos of spherical crystals of NQ obtained by various methods of

crystallization are shown. The pictures were acquired by scanning microscope of JEOL,
model JSM 5400, 25 kV.
654
Fig 2.
Nitroguanidine crystals obtained after crystallisation from various solvents:

1 GDE, 2 DMF.2, 3 DMF/GDE.2, 4 DMF/GDE.3, 5 NMP.2,
6 NMP.3, 7 DMSO.2, 8 DMSO.2
The efficiency of crystallization (Fig. 3) was determined from the relation of initial mass
of NQ to the quantity of product obtained after crystallization. Bulk density was calculated
by use a measuring glass of capacity of 25 cm3, as the amount of mass that can be poured
into it to its volume. And tap bulk density was measured in an similar manner but after 300
cycles of shakes that were made on Stampfvolumeter STAV 2003.
655
[%]
80
70
60
50
66
40
67
54
72
61
64
64
43
30
20
DMSO.3
DMSO.2
NMP.3
0,29
0,74
0,7
0,59
0,54
0,3
0,79
0,4
0,47
0,36
0,5
0,68
0,6
0,64
0,57
0,7
0,86
0,87
0,72
0,8
0,2
0,77
0,9
[g/cm3]
NMP.2
DMF/GDE.3
DMF.2
Effectiveness of NQ crystallization from various solvents
0,79
Fig 3.
GDE
DMF/GDE.2
10
bulk density
Fig 4.
3.
DMSO.3
DMSO.2
NMP.3
NMP.2
DMF/GDE.
DMF/GDE.
DMF.2
GDE
0,1
tap density
Bulk and tap density of NQ samples obtained from different solvents
GRAIN SIZE DISTRIBUTION OF NQ CRYSTALS
An recognition of grain size distribution of the explosives brings an indispensable

information by press moulding or by developing of explosives. In order to obtain high
detonation parameters it is advantageous to use of various fractions of polycrystals.
656
Grain size distribution of NQ crystals were obtained by various crystallization

procedures were determined by use IPS U (Infrared Particle Size) analyzer version 8.12.
The principle of analyzer consists in measuring of changes of infrared rays stream that
diffuse on moving NQ particles. The changes of the rays stream are registered and the
dispersed spectrum is depicted in a suitable graph. (Fig. 5)
[%] 18
16
14
12
10
8
6
4
2
0
[%] 12
10
8
6
4
2
0
0
100
200
300
400
[ m]
100
GDE
4
0
100
200
300
300
400
[ m]
DMF2
[%]
200
400
500
600
700 800
[ m]
18
[%]
16
14
12
10
8
6
4
2
0
0
100
DMF/GDE.2
200
300
400
[ m]
DMF/GDE.3
[%] 12
10
[%] 10
2
0
0
0
100
200
300
400
100
200
300
400
[ m]
[ m]
NMP.2
NMP.3
657
10
[%]
8
10
[%]
8
0
0
100
200
300
[ m]
400
100
200
300
400
DMSO.2
Fig 5.
4.
500
600 700
[m]
DMSO.3
Grain size of NQ crystals spectrum as obtained by infrared radiation scatter .
THERMAL ANALYSIS OF NQ
To control the temperature susceptibility an thermal differential analysis of obtained NQ

crystals was performed. The measurements were made by Labsys TG_DTA/DSC apparatus
made by SETARAM . TG/DSC suspension of 1600oC was used. Samples of mass of about
10 mg (exact masses of each samples are specified on thermograms) were heated up in
platinum crucibles of capacity of 100 l. The crucible of reference was empty on view of
small mass of investigated samples. The tests were performed at following parameters:
Range of temperature:
20500oC,
The speed of heating:
5oC/min,
Furnace atmosphere: Nitrogen, intensity of flux/flow 50 ml/min.
Thermograms for two kinds of NQ were made: crystallized from water and from N,Ndimethyloformamide.
50
45
NMP
o
243,2 C
35
30
TG, %
Signal DSC [a.u.]
40
25
20
H2O
15
232,7 oC
10
5
0
0
50
100
150
200
250
300
350
400
5.
H2O
NMP
50
100
150
200
250
300
350
400
Temperature [ C]
Fig 6.
100
90
80
70
60
50
40
30
20
10
0
-10
Temperature [ C]
Thermal decomposition of NQ;
FRACTION SENSITIVITY OF NQ MIXTURES WITH HE
Determination of sensitivity to fraction of explosives containing NQ was made on

Peters apparatus which is generally used in this type of investigations [9]. For comparison
also measurements to selected crystalline MW were made: PETN, RDX, HMX. Series of
tests were performed in which the critical force with which the plunger acts on the sample of
explosive was investigated. In the investigations the smallest magnitude of pressure force
658
determined, by which in 6 measurements at least one positive result occur and in contrary,
the largest force of pressure by which in any of 6 samples there was no explosive
transformation, were determined. The results of investigations are illustrated in Fig. 7, on
which sensitivity of RDX was assumed as the level of reference. The samples to
investigations were prepared by mixing the components at the same comminution degree (of
100200 m).
200
180
160
[%]
140
120
200
100
80
100
60
57
40
51
20
42
34
0
PETN
Fig 7.
6.
RDX, HMX PETN/NQ30/70
RDX/NQ70/30
RDX/NQ50/50
RDX/NQ30/70
Sensitivity to fraction of selected explosives.
THE DETONATION VELOCITY OF NQ

AND OF HE MIXTURES WITH NQ
NQ was molded directly into steel tubes of 36.5/44.5 mm in diameter to various

densities in range 1.0 1.45 g/cm3. Two openings were made in tube , through which short
circuit gauges were pulled up (Fig. 8). A charge was boosted by RDX/Al=70/30 detonator of
mass of 25 g. The detonation evolution over 3 segments of tube was registered by use a
microseconds timer type T-1520. The results of measurements are set up in Table 2.
1
4
Fig 8.
The NQ charge prepared to detonation test:

1 detonator, 2 NQ charge, 3 steel tube, 4 - gauges
659
Table 2.
Detonative parameters of nitroguanidine

NITROGUANIDINE
Mass density [g/cm3]
1,0
1,2
1,45
Velocity of detonation [m/s]
4660
5900
2640
Mixtures of high energetic explosives (RDX and HMX) with NQ were prepared in order
to ensure their high density and to be safe by moulding. Not phlegmatized explosives were
used. In order the mixtures did not electrify, 0,5 % of graphite was added. The following
mixtures were set up for investigations:
Hx/NQ/Cgr (29.5/70/0.5), Hx/NQ/Cgr (49.5/50/0.5),
HMX/NQ/Cgr (29.5/70/0.5), HMX/NQ/Cgr (49.5/50/0.5)
Specimens of 30 or 50 g were weighed out and moulded in matrix of 40 mm in diameter
on a hydraulic press under pressure of 200 MPa. Assuming the mass and volume of cube
were, the density of charges was calculated. Gauges were fasted to prepared samples by use
the nitrocellulose varnish and whole charge was put in tube of 45/50 mm in diameter. The
free space was poured with water with small sand. Detonator RDX/wax = 95/5 was used to
initiate the detonation. The results of measurements of charges density and velocity of
detonation are presented in table 3 and on fig. 9. Also measured detonation velocity of TNT,
RDX i HMX were added. The table contains also the calculations of theoretical maximum
density and evaluated theoretical velocity of detonation.
9000
Velocity of detonation [m/s]
8700
8500
8500
8000
8070
8300
7990
8000
7500
7000
6900
6500
6000
TNT
HMX/NQ(29,5/70)
HMXphleg.
Fig 9.
RDX/NQ(29,5/70)
HMX/NQ(49,5/50)
RDX/NQ(49,5/50)
RDXphleg.
Detonation velocity of HE and of selected mixtures of HE with NQ.
660
Table 3. Measured and calculated magnitudes of density and velocity

of detonation examined mixtures.
Composition of
mixture
Theoretical
Velocity of
0
Velocity
of
Density maximum
detonation D0 100%
100%
detonation Dteor for Dteor
density max max
0
[g/cm3]
D0 [m/s] 0 [m/s]
[g/cm3]
[%]
[%]
(TIGER)
(TIGER)
Hx/NQ/Cgr
1,62
1,78
91
8000
8091
98,9
1,62
1,79
90
8070
8139
99,1
1,63
1,81
90
8100
8160
99,3
1,72
1,84
93
8500
8505
99,9
29,5/70/0,5
Hx/NQ/Cgr
49,5/50/0,5
HMX/NQ/Cgr
29,5/70/0,5
HMX/NQ/Cgr
49,5/50/0,5
7.
FINAL CONCLUSIONS
As it follows from the carried out investigations on crystallization of NQ, spherical form
of crystals can be obtained from many solvents. It should be noted that the density of these
crystals was considerably high and the efficiency of crystallization were satisfactory. Also
the grain size distribution of obtained product was suitable. On the basis of the analysis of
results of measurements (fig. 24) we can state, that the highest densities and efficiency
were gained when N-methylopyrrolide was used as a solvent. Analysis of distribution of NQ
grain size NQ points out that not in every case most of crystals are obtained in size of 100300 m. The smallest crystals were obtained from dieethylene glycol and their mass density
was also the smallest one. From thermograms which were gained during thermal NQ
analysis appears that NQ which is crystallized from N-methylopyrrolide is clearer than one
obtained from water it gives high and sharp peak on the DSC graph and smaller rest on the
TG curve. The investigations on sensitivity to fraction and measurements of velocity of
detonation of explosives RDX and HMX which contain 70% and 50% NQ confirm the
possibility to obtain an considerable decrease of sensitivity to fraction. The measured
velocity of detonation of there mixtures differs from theoretical velocity calculated by use of
the numerical code Tiger merely of about of 1%. Obtained mouldering densities of these
mixtures are of about of 9093 % of theoretical maximum density.
Generally it is to state, that NQ can be used to effective decrease of HE sensitivity and
brings no important influence on detonation parameters these mixtures.
Acknowledgment:
661
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
LAMY P., et al.:(Technical Group, ASNR LTTP on IHE), Air Senior National Representative
Long Term Technology Project on Insensitive High Explosives, Proc. of 27-th International
Annual Conference of ICT, Karlsruhe, Germany, 25-28.06.1996, 1/1 1/13.
FOUCHE F.C., VAN SCHALWYK G.C.: TNT Based Insensitive Munitions, ibid, 69/169/12.
AGRAWAL P.: Recent Trends in High-Energy Materials,Prog. Energy Combust. Sci., 1998
BECUWE A.,DELCLOS A.: Low-Sensitivity Explosive Compounds for Low Vulnerability
Warheads, Propellants, Explosives, Pyrotechnic, 18, 1-10, 1993.
TENNEY L. D., CLARKE H. T., PHILLIPS R.: Nitroguanidine,Organic Syntheses, CV, 1,
399
PRITCHARD E. J., WRIGHT G. F.: Production of nitroguanidine with high bulk-density,
Canadian Journal of Reaserch, vol. 25 (July 1947), sec.F, 257 263.
ENGEL W., HEINISCH H.: Process for preparation of compact nitroguanidine, pat. USA 4
544 769 (1985).
THIEL K., HEINISH H.: Process for the production of spferulitic partcles, pat. USA 5 696 407
(1997).
POLISH STANDARD PN-C-86019:1994 (in Polish): Materiay wybuchowe, Oznaczanie
wraliwoci na tarcie.
662
APPLICATION OF TIME-TEMPERATURE SUPERPOSITION

PRINCIPLES TO PREDICTION OF MECHANICAL PROPERTIES
OF DOUBLE BASED ROCKET PROPELLANTS
Sanja Matei Muani*, Muhamed Sueska*, Maa Raji Linari*,
and Sanko Bakija**
* Brodarski Institute, Av. V. Holjevca 20, 10 020 Zagreb, Croatia
** Ministry of Defence, Bauerova 33, 10000 Zagreb, Croatia
Abstract:
Because of their viscoelastic nature, double based rocket propellants may undergo the
degradation of their mechanical properties, which can significantly influence their stability.
Because of safety reasons, it is necessary to establish an appropriate method for the
prediction of the propellants behaviour over a long-term period of time under defined
temperature conditions.
Applying accelerated temperature measurements and time-temperature superposition
(TTS) treatment of the data, limited lab testing (creep or stress relaxation) is sufficient to
project long-term properties under different conditions. Dynamic mechanical analysis,
which measures the modulus and damping properties of material, is one of the most
convenient thermal analysis techniques that can be applied for time-temperature
superposition approach.
The goal of this work was to investigate the applicability of the TTS principle to the
prediction of the mechanical properties of double based rocket propellants during the
ageing. The master curve was generated on the basis of the creep measurements in the
temperature region from -71 to +20C, by a dynamic mechanical analyser. The generated
master curve was mathematically described by Williams-Lander-Ferry (WLF) and
Arrhenious models. Material constants in the WLF equation were determined to be
C1=44.43 and C2=308.4, and the viscoelastic activation energy from the Arrhenious
equation was determined to be Ea= 152.7 kJ/mol. The results have shown that the TTS
principle is applicable to the prediction of the propellants behaviour during a long-term
period of time at defined temperature conditions.
Keywords:
1.
time-temperature superposition, double based rocket propellants,

dynamic mechanical analysis, creep, shift factor, master curve, WLF
model, WLF constants, Arrhenious model, viscoelastic activation energy
INTRODUCTION
Double based rocket propellants are due to their viscoelastic nature undergo to degradation
of their mechanical properties. This can significantly influence their stability and safety at use.
The stability of explosives and propellants is one of the major problems of the explosive
community. Therefore, it is necessary to find out an appropriate method for prediction of the
663
propellants behavior over a long-term period under defined temperature conditions. The timetemperature superposition (TTS) is one of the possible approaches that can be applied to treat
this problem.
Viscoelastic materials during deformation exhibit behavior which is time, or frequency and
temperature, dependent. This means that, if a viscoelastic material is subjected to a constant
load, the elastic modulus exhibited by the material will decrease over a period of time (Fig 1).
This occurs because the viscoelastic material under a load undergoes molecular rearrangement
in attempt to minimize localized stresses. Hence, modulus measurements performed over a short
time span result in higher values than longer term measurements. This time dependent behaviour
would seem to imply that in order to evaluate accurately material properties, one need to test the
material under the actual time and temperature conditions.
log E, Pa
T1<T2<T3
T1
T2
T3
log t, s
Fig 1.
Creep modulus as a function of time and temperature
However, using the time-temperature superposition principle, that type of prolonged

measurements is not necessary. Applying accelerated temperature measurements and theoretical
time-temperature superposition treatment of the data, limited laboratory testing (creep or stress
relaxation) is sufficient to project long-term properties under different conditions [1].
Dynamic mechanical analysis, that measures the modulus and damping properties of a
material, can be used for time-temperature superposition approach.
1.1
Time-temperature superposition principle
According to the TTS principle, the time function of elastic modulus (E) at one temperature
T0 is similar in shape to the same function at neighboring temperatures (Fig 1). The curve of E
versus logarithmic response of time at one temperature can be horizontally shifted along the
reduced time axis until overlapped the curve at neighbouring temperature. The shift distance
along the logarithmic time axis is called the time-temperature shift factor (aT). It is given by
equation (1):
aT =
tT
t ref
(1)
where tT is the time needed to give a certain response at a temperature T, and tref is the time
needed to give the same response at the reference temperature Tref. [1, 2, 3].
664
Mathematically the idea of the TTS may be expressed as:
E (TT , t T ) = E (Tref , t T / T ) ,
(2)
where TT is the temperature T, and E is the elastic modulus of material that is equal to the
flexural creep modulus (1/S):
E = 1/ S
(3)
where S is creep compliance (S = /).
log (1/S), Pa
In the TTS, every curve at a temperature other than the reference temperature can be shifted
along a proper distance, in order to overlap the curve at the neighboring temperature (Fig 2).
The value of the shift distance is dependent on the reference temperature and the material
properties of the polymer. For every reference temperature chosen, a fully overlapped curve
could be formed, and it is called the master curve.
T1
log T
T2
log t1
Fig 2.
log t2
log t, s
Graphical presentation of TTS principle and determination of aT
The degree of horizontal shifting (i.e. time) can be mathematically described with respect to
temperature. Two models are commonly used in practice. The first is the well-known WilliamsLander-Ferry (WLF) equation [1, 4, 5]:
log T =
C1 (T Tref
C 2 + (T Tref )
(4)
where Tref is a reference temperature (typically Tref is taken as the glass transition
temperature, Tg) according the master curve is generated by shifting the dynamic mechanical
test data at other temperatures, and C1 i C2 are material constants [6].
The WLF equation is based on the assumption that above the glass transition temperature,
the fractional free volume increases linearly with respect to temperature [1,7]. The WLF equation
is typically used to describe the time/temperature behaviour of polymers in the glass transition
region [4,5]. The typical temperature range in which the WLF works well is between Tg and
Tg+100C. At the temperatures well below Tg, the WLF equation deviates significantly from the
test data [6].
665
For the temperature range below the glass transition temperature, the Arrhenious model is
generally accepted as suitable to describe the relationship between the shift factors (aT) of
master curve and the reference temperature: [1, 2]
ln T =
E a (1 / T 1 / Tref )
R
(5)
where Ea is the viscoelastic activation energy (the activation energy associated with the
relaxation transition), R is the universal gas constant (R = 8.314 kJ/molC), T is the
measurement temperature. The Arrhenious equation is typically used to describe the viscoelastic
events associated with and relaxation transitions, or for the glass transition associated with
semi-crystalline polymers. Frequently, it is used to obtain the activation energy associated with
the glass transition event.
2.
EXPERIMENTAL
The experiments were carried out on the double based rocket propellant of the following
chemical composition: 55.1% of nitrocellulose, 34.0% nitroglycerine, 3.1% diethilftalate, 2,8%
centralite, 5.0% of other additives.
3 mm
The samples for dynamic mechanical measurements, having rectangular bar shape (50 x 10
x 2.5 mm) (Fig 3), were cut from propellant grain.
10 mm
50 mm
Fig 3.
Form and dimensions of test specimen for DMA measurement
Dynamic mechanical measurements were carried out using TA Instruments, DMA Model
983.
For the determination of the temperature range for the application of the time-temperature
superposition, the measurement was conducting using the fixed frequency mode at the following
testing conditions:
temperature range: -120C to +80C
heating rate: 2C/min
frequency of an oscillation load: 1 Hz
amplitude of deformation: 0.2 mm
length to thickness ratio (L/T): ~10
666
The testing conditions for the creep experiment necessary for the time-temperature
superposition to be done were:
temperature range: -70C to +80C
temperature increment: 10C
displace and recovery time: 20 min
length to thickness ratio (L/T): ~10
3.

3.1
Determination of temperature range of time temperature

superposition principles application
For the determination of the optimum temperature range for the creep experiment, the
scouting experiment was carried out on the rocket propellant in the fixed frequency mode. The
DMA thermogram of the studied double based rocket propellant is shown in Fig 4.
Fig 4.
DMA thermogram of the studied double based rocket propellant
From the theoretical point of view, the lower temperature limit (T0) of time-temperature
superposition principle application is temperature at which the activation of the relaxation
processes starts (value of loss modulus starts to grow). In effect, that temperature can be
determined from tangents intersection (or onset) at the peak of the loss modulus (E) curve (Fig
4). As the upper temperature limit (Tg+50C), a temperature that is about 50C higher than the
temperature of glass transition (Fig 4) is usually taken.
This principle is applicable to the materials that have common relaxation spectrum and
relatively tight peak of loss modulus (tight region of relaxation transition). Compared with some
other materials, in the case of the double based rocket propellants this interval is much higher.
It follows from DMA thermogram of the double based rocket propellant (Fig 4), that the
onset temperature at the curve of loss modulus (T0) is approximately -88C, the temperature of
667
glass transitions (Tg) is -43C, and the temperature Tg +50C is +7C. It is obvious from these
results that in accordance with the before described principle for the determination of the
temperature region of the time-temperature superposition principle application, for double
basted rocket propellants the temperature region of application is between -88 and +7C.
3.2
Creep measurements and generation of the master curve
Although from the theoretical point of view the temperature region of the creep experiment
should be between the -88 and +7C, because of some experimental limitations (capacity of
liquid nitrogen cooling accessory was 50 L) the creep experiment was carried out starting from 71C. In order to see the creep behaviour of material at a temperature that is higher than the
temperature in the region of the glass transition (viscoelastic and viscous region), the creep
experiment was carried out in the region from -71 to +80C. The creep data for the studied
double based rocket propellant is shown in Fig 5.
aT
Fig 5.
Creep data for the studied double based rocket propellant in temperature range
from -71 to +80C
It is obvious from Fig 5 that the values of the flexure creep modulus decrease much faster in
the region of propellant softening (temperatures from +20 to +60C), which indicates that the
relaxation processes in that region are much faster than the relaxation processes in the region of
the glass transition.
From the creep experiment data obtained in the temperature region from -71 to 80C, in
accordance with the previously described time-temperature superposition principle, the
temperature -61C was chosen as a reference (Tref). After that, other isothermally obtained
curves of the flexure creep modulus were shifted horizontally along the time axis to overlap
with the reference curve (Fig 5).
The generated master curve and corresponding shift factor data in the temperature range
from -71 to +80C are shown in Figs 6 and 7.
668
Fig 6.
Master curve generated from creep data in the temperature range

from -71 to 80C
Fig 7.
Shift factor vs. temperature in the temperature range from -71 to 80C
The master curve generated from the flexural creep modulus curves, obtained in the
temperature region from -71 to +80C (Fig 6) shows that the value of the flexural creep modulus
in the glass transition region (-71 to 20C) slowly decreases with time. The decrease of the
flexural creep modulus is much faster in the region of the propellant softening temperatures
(+20 to +80C). This is in accordance with the theory that relaxation processes in the viscous
stage of matter are much faster than relaxation processes in the region of the glass transition,
which is also obvious from the dependency of the shift factor on temperature (Fig 7).
669
log E=9,68 Pa
log E=9,42 Pa
t=0,001s
Fig 8.
t=100s
Master curve generated from creep data in the temperature range

from -71 to +20C
From the theoretical point of view previously described (i.e. linearity requirements), and
from the experimental results given earlier, it is obvious that the temperature range for the
application of the time-temperature superposition principle to the double based rocket
propellants is from -71C to +20C.
Accordingly, from the creep data obtained in the temperature region from -71 to +20C, a
new master curve was generated (Fig 8). As a reference curve, in accordance with a theory, the
temperature nearest to the temperature of the glass transition (Tg = -43C), Tref = -40,9C was
chosen. The master curve and the corresponding horizontal shift factor are shown in Fig 8 and
Table 1.
Table 1. The values of the horizontal shift factors in the temperature region from -71
to +20 oC
3.3
Temperature, oC
log(aT), s
-70.9
-60.9
-50.9
-40.9
-30.9
-20.9
-10.9
-0.9
9.1
19.1
4.1346
3.3550
1.,6705
0 (reference)
-1.6241
-3.0905
-4.2552
-5.0302
-6.3387
-7.0905
Dependence of shift factor on temperature
Dependence of the shift factor on temperature can be mathematically described by two

models: by WLF model and by Arrhenius model.
670
By nonlinear regression analysis of the data shown in Table 1, in accordance with the WLF
and Arrhenious equations, the WLF constants (C1 and C2), and the value of the activation
energy (Ea) associated with the relaxation transition were calculated. The obtained results are
shown in Figs. 9 and 10 and Table 2.
Fig 9. Dependence of shift factor on temperature (solid line obtained applying

the WLF model)
Fig 10. Dependence of shift factor on temperature (solid line obtained applying
the Arrhenious model)
671
Table 2. The WLF constants (C1 and C2), and viscoelastic activation energy (Ea)
calculated from creep data (experiments)
Temperature region
of measurement
-71 to +20 oC
Arrhenious
model
WLF model
Reference
temperature, oC
Constant C1
Constant C2
Ea, kJ/mol
-40.9
44.43
308.4
152.7
It is obvious from Figs 9 and 10 that both models can satisfactorily describe the shift factortemperature relationship. But, it should be noted that the obtained results -WLF constants (C1
and C2) and activation energy (Ea) are very sensitive to the values of the shift factor. From the
practical aspect that means that during the generation of the master curve very small differences
in the values of the shift factors can significantly influence the final results: WLF constants and
activation energy.
The practical importance of the master curve is explained in Fig 8. Namely, the master
curve shows the dependence of relaxation modulus at temperature 40,9C on time of stress
duration. It is obvious that the time axis is, thanks to the principle of time-temperature
superposition, expanded from very short time of the loaded stress (20 min), to the time that
considerably exceeds the experiment time limit. For example, the value of the modulus after 100
s (log t = 2) is 9.42 Pa, and value of the modulus after 0.001 s (log t = -3) is 9.68 Pa. The value
of the modulus at very short time of the loaded stress is important for prediction of propellant
behaviour under dynamic stress conditions, for example the impact of the ignition charge
pressure.
Expanding the temperature of the creep experiment to room temperatures, it is possible to
determine the values of the relaxation modulus after long term loading stress which is very
important for the prediction of the propellant behaviour under long term loading stress
conditions, for example, during the storage of the propellants.
672
4.
CONCLUSIONS
On the basis of the experimental results the following can be concluded:
-
The temperature range, obtained on the basis of the fixed frequency and creep mode
measurements, for the application of the time temperature superposition principle in
the case of a double based rocket propellant is between -88 to +20C.
The obtained master curve (Tref= -40.9C) can be mathematically described very well
by the WLF and the Arrhenious model.
The constants in the WLF equation are determined to be: C1= 44.43 and C2= 308.4.
The activation energy (Ea) associated with the relaxation transition, calculated from
the Arrhenious model is 152.7 kJ/mol.
Generally, it can be concluded that the DMA technique and the time-temperature
superposition principles, are applicable to the prediction of the double based rocket
propellants mechanical properties during a long-term period. However, great care
should be taken during the data treatment procedure, since the WLF constants (C1 and
C2) and activation energy (Ea) are very sensitive to the values of the shift factor.
REFERENCES
[1] J.D.FERRY: Viscoelastic Properties of Polymers, Wiley, New York, 1980.
[2] R. LI: Time temperature superposition method for glass transition temperature of plastic materials,
Materials Science and Engineering A 278 (2000) 36-45
[3] W. BROSTOW: Macromol., Chem. Symp. 41 (1991) 119
[4] M.L. WILLIAMS, R.F. LANDEL, J.D.FERRY: J.Am. Chem.Soc., 77, 3701 (1955)
[5] J.J. ALCONIS, W.J. MACKNIGHT, S. MICHEL: Introduction to Polym. Viscoelasticity, John Wiley &
Sons, New York, 1972.
[6] B. FAN, D. KAZMER: Effect of Low Temperature Shift Factor Modeling on Predicted Part Quality,
ANTEC 2003, 938-941
[7] A.K.DOOLITTLE, D.B.DOOLITLE, J.APPL.PHYS., 31, 1164 (1959)
673
INFLUENCE OF OIL ON SENSITIVITY

AND THERMAL STABILITY OF TRIACETONE TRIPEROXIDE
AND HEXAMETHYLENETRIPEROXIDE DIAMINE
Robert Maty
Abstract:
The desensitization of triacetone triperoxide (TATP) and hexamethylenetriperoxide
diamine (HMTD) was investigated. WD-40 oil was used as a possible candidate. The
sensitivity to friction and impact were studied for pure TATP and HMTD as well as
mixtures with varying ratio of WD-40 oil. Thermal stability of such mixtures was
determined using DTA.
Keywords:
1.
triacetone triperoxide; TATP; hexamethylenetriperoxide diamine;

HMTD; sensitivity; thermal stability.
INTRODUCTION
An increasing number of criminal incidents was carried using triacetone triperoxide

(TATP, 3,3,6,6,9,9-hexamethyl- [1,2,4,5,7,8]hexaoxonane) and hexamethylene-triperoxide
diamine (HMTD, 3,4,8,9,12,13-hexaoxa-1,6-diaza-bicyklo [4,4,4] tetradecane) recently [1,2].
These organic peroxides have also been reported to be used as improvised explosives by
terrorist organizations [3,4] and also are very popular among amateur chemists [1,5]. Their
popularity is mainly caused due to an easy preparation well described on the internet.
Information including properties, synthesis in detail, hands-on experience of other users,
photographs of applications and many others can be easily found on extremists webs [5,6,7].
Discussion conferences than explain practical aspects of synthesis and applications. The
synthesis is very simple and can by carry out without deeper knowledge of chemistry. There
is no need for special equipment and the necessary starting substances can be obtained in
most of the drugstores.
Preparation of TATP is based on reaction acetone with hydrogen peroxide [8,9]. This
reaction is catalyzed by acid. Hydrogen peroxide is accessible in drugstores as a bleaching
agent (in concentration 30%); acetone is used as paint thinner. Many acids are sold in
drugstores that can be used as catalysts in preparation of TATP, for example hydrochloric,
sulphuric or nitric acids. The diacetone diperoxide is frequently observed as a by-product of
the synthesis of TATP when the sulphuric acid is used as a catalyst [10].
674
H3C
H3C
C H3
O O
C
O
O
O
H3C
CH2 O O CH2
C H3
N CH2 O O CH2
CH2 O O CH2
O
C H3
TATP
(triacetone triperoxide)
Fig 1.
HMTD
(hexamethylenetriperoxide diamine)
Chemical formulas of TATP and HMTD
Preparation of HMTD is based on reaction of hexamethylenetetramine with hydrogen

peroxide; reaction is catalyzed by acid [11]. Hexamethylenetetramine is accessible as fuel for
tourists cooker in Czech Republic or as chemical. Most popular acid for catalysis of this
reaction is citric acid. Citric acid is used as a relish.
TATP and HMTD are powerful explosives what leaded to consideration about their
applications [9,11]. On the other hand TATP and HMTD are unstable substances. TATP is
very volatile substance (weight loss 6.5% during 24 hours, 68.6% during 14 days at room
temperature) [9]. Therefore they have never been used in any applications even though they
were studied and patented as primary explosives in detonators [12] or in mixture with other
explosives [13].
TATP and HMTD are very sensitive primary explosives. Many accidents happened
during handling [14,15]. Liquidation these findings by pyrotechnic service are dangerous due to
their high sensibility. Several ways of destruction were published. TATP could be
decomposed by burning of it in toluene or xylene solution [10]. Saturated solution
(~ 28 wt. %) couldnt be initiated by detonator and is chemically stable. Thermal
decomposition of its solution is rather slow (half-time decomposition of benzene solution at
100C in the presence of H2SO4 is reported to be 44 min) [10]. Several agents were tested for
chemical destruction of TATP and HMTD [10,16]. TATP decomposes by conc. hydroiodic
acid within 1-3 min. (for fine crystals) and higher crystals decompose within 2 hours at room
temperature. HMTD decomposes by conc. hydrobromic acid within a few seconds [16].
All of these ways require handling with TATP and HMTD that could be dangerous. The
possibility was therefore studied for reduction high sensibility of these explosives by using
desensitization agent that could be used by pyrotechnic service to decrease sensibility during
handling.
The desensitization agent must fulfil the following conditions. At first it has to decrease
sensitivity of explosive to mechanical stimuli. The agent has to penetrate well through layer
of powdered organic peroxide and must not react with them. The WD-40 oil was chosen for
desensitization of dry triacetone triperoxide and hexamethylenetriperoxide diamine. The
advantage of this oil is its good accessibility and low price. The oil WD-40 has low viscosity
and well penetrates through powdered organic peroxide. It has good adhesion to TATP and
HMTD crystals. Because oil WD-40 is provided as a spray it can be apply by spritz. It
doesnt react neither TATP non HMTD.
675
2.
EXPERIMENTS AND RESULTS

Samples preparation
Triacetone triperoxide was prepared according to Fedoroff [9] method. Hydrochloric acid
was used as a catalyst because used acid has a big influence on TATP properties (especially
thermal stability). HMTD was prepared according to Urbaski [17] method. TATP and
HMTD desensitization was carried out by WD-40 oil addition.
WD-40 oil was added to powdered TATP and HMTD in 15, 30 and 45 wt. %. Further
they were prepared samples where TATP and HMTD were completely under oil surface.
These samples are called excess of oil further. All samples were well homogenized.
Sensitivity to friction
The friction sensitivity was performed using BAM friction apparatus. The friction
sensitivity data obtained according to up and down method were used for assessment of
force causing ignition in 50% of the trails. The test was performed with TATP, HMTD, their
oil mixtures and with mercury fulminate (MF) as a standard of primary explosives. The
results of this test summarized in table 1 and a comparison with mercury fulminate is plotted
in fig. 3.
Table 1. Friction sensitivity of TATP, HMTD and oil mixtures.
content of oil in peroxide

(wt. %)
F50
(mN)
TATP
HMTD
15
12
15
30
45
excess of oil
60
50
49
75
54
87
53
50
As can be seen sensitivity of mercury fulminate (51 mN) is comparable to the oil
mixtures of peroxides. A significant desensitization was observed even with small amounts
of oil.
676
Sensitivity to friction
(mN)
80
60
MF
40
TATP
20
HMTD
0
0
15
30
45
excess
of oil
Content of oil (wt. %)
Fig 2.
Sensitivity of TATP and HMTD and their oil mixtures to friction

compared with mercury fulminate (up and down method).
Sensitivity to impact
The sensitivity to impact was tested using the BAM drop hammer apparatus. The test
was performed until 6 negative results occurred. The results are compared with values from
the literature in table 2.
Table 2. Results from impact sensitivity test. The experimental values obtained
during this investigation are compared with Mayer values [18].
content of oil in peroxide
(wt. %)
published value for pure
org. peroxide by Meyer [18]
0
15
30
45
impact energy
(J)
TATP
HMTD
0.3
0.6
0.75
0.75
0.75
0.75
1.5
3.0
1.0
1.0
Thermal stability
Thermal stability and compatibility of TATP and HMTD with oil WD-40 were
investigated by DTA. The amount used was 25 mg for TATP and TATP/oil compositions
and 10 mg for HMTD and HMTD/oil compositions. The lower amount of HMTD and its oil
compositions was chosen since the decomposition is more violent compared to TATP. The
heating rate was 5C/min. DTA thermogram of TATP and its oil compositions is presented
on figure 3 and DTA thermogram of HMTD and its oil compositions is presented on
figure 4.
677
T AT P
Difference of temperature [C]
-1
T AT P + 15% W D 40
-2
-3
T AT P + 30% W D 40
-4
-5
T AT P + 45% W D 40
-6
-7
T AT P + "e xce ss" W D 40

-8
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
200
210
Temperature [C]
Meavy
Fig 3.
DTA thermogram of TATP and TATP/oil compositions.
Difference of temperature [C]
H MT D
0
H MT D + 15% W D 40
-1
H MT D + 30% W D 40
-2
H MT D + 45% W D 40
-3
H MT D + "e xce ss" W D 40

-4
60
65
70
75
80
85
90
95
100
105
110
115
120
125
130
135
140
145
150
Teperature [C]
Meavy
Fig 4.
DTA thermogram of TATP and TATP/oil compositions.
678
3.
CONCLUSIONS:
Effect of WD-40 oil in desensitization TATP and HMTD was investigated. Thermal
stability of these compositions was measured by DTA. Several organic peroxide/oil
compositions were tested with different content of oil. Addition WD-40 oil to powdered
TATP and HMTD noticeably decreases their sensitivity to friction. Approx. 3-4 times lower
sensitivity compared to pure TATP and HMTD is achieved which makes the mixtures about
as sensitive as mercury fulminate. Decrease of sensitivity to friction doesnt much depend on
amount of oil in TATP/oil and HMTD/oil compositions.
Presence of oil in powdered TATP and HMTD doesnt have considerable influence on
their sensitivity to impact and is approx. same as pure organic peroxides. Addition of WD-40
oil to TATP and HMTD doesnt have negative influence on the thermal stability of TATP
and HMTD. The onset of decomposition is shifted to higher temperatures with addition of
oil.
Based on these results WD-40 oil can be considered suitable for desensitization of TATP
and HMTD. Since friction appears to be more probable source of initiation WD-40 oil can be
recommended.
REFERENCES:
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
J. ARAPATKA: Analza udlost za rok 2000; Bulletin Ochrann sluby Policie esk
republiky; MV R; Praha; 2001.
R. SCHULTE-LADBECK, P. KOLLA,U. KARST: Anal. Chem.; 75; 713; 2003.
The Independent; October 8; p. 7; London; 1996.
R. T. COOPER: Los Angeles Times; December 29; p. A12; 2001.
R. MATY: Grant of Ministry of Internal Affairs Czech Republic, no. RN 20012003003, not
approved for public releace; 2002.
http://www.roguesci.org/megalomania/explosives.htm [02.03.2005].
http://pxd.zde.cz/ [02.03.2005].
R. WOLFFENSTEIN: Chem. Ber.; 28; 2265; 1895.
T. FEDOROFF: Encyclopedia of explosives and related items, Volume 1, A 41, Picatinny
Arsenal, New Jersey, USA; 1960.
A. J. BELLAMY: J. Forensic Sci., 44(3); 1999; 603.
T. FEDOROFF: Encyclopedia of explosives and related items, Volume 7, H 83, Picatinny
Arsenal, New Jersey USA 1975.
G. E. MAVRODI: Brit. pat. 620,498 (1949); Chem. Abstr.; 43; 6418; 1949.
GVELOT, GAUPILAT: Fr. pat. 893,941 (1944); Chem. Abstr.; 47; 8374; 1953.
I. PETEREK: personal communication.
A. NOPONEN: Chem. Eng. News; 55; 5; 1977.
R. MATY: Chemical destruction of triacetone triperoxide and hexamethylenetriperoxidediamine, Proceeding of VI. seminar New trends in research of energetic materials,
Pardubice; 2004.
T. URBASKI: Chemie a technologie vbuin; Volume 3; 166; SNTL; Praha; 1959.
R. MEYER: Explosives, 2 ed., Verlag Chemie, Weinheim; 1981.
679
EFFECT OF ALUMINIUM POWDER ON THE THERMAL

AND ENERGETIC PROPERTIES OF COMPOSITE
ROCKET PROPELLANT BASED EPOXY RESIN
A. Mouloud
Laboratoire des Systmes Pyrotechniques, UER de Chimie Applique,
B.P 17, EMP, Bordj-El-Bahri, 16111 Algiers, Algeria
Abstract:
The objective of this work is to study the effect of aluminium rate on the thermal, kinetic
and energetic properties of catalysed composite propellant based DGEBA epoxy resin
binder. The first part of this study is consisting of the formulation of the propellant at
different rates of the aluminium powder (from 0 to 20 % in weight), and by
incorporation copper chromite as burning catalyst. The second part consists to
characterize thermally, kinetically and energetically the propellant samples by using
DSC (auto-inflammation temperature), TGA (activation energy) and adiabatic
calorimeter (calorific potential) of the elaborated propellant.
Keywords: aluminised composite propellant, aluminium, DGEBA epoxy resin,
thermokinetic properties
1.
INTRODUCTION
The aluminium powder is used as combustible charge in solid propellants, because of its
high density, its combustion energy when burning, its low price and relative security
(properties wanted particularly in missiles and boosters).
As the most of aluminized propellants have a great ratio of fuel, it seems that the
addition of an other concentration of aluminium will be excessive, particularly when the
burning product (Alumina) is in liquid state. However, the combustion take oxidizer vapors
(H2O and CO2) to H2 and CO which are the best propulsive gases. This is achieved with a
great burning temperature which consquetely give high specific impulse of about 10 % and a
gain of propellant density of 15 %. Also, the addition of aluminium powder, even in small
quantities cancels the combustion instability problems, which are encountered in the
development programs of propellers [1].
Aluminium is known for its power to increase the burning rate of solid propellants. It is
also well known for its power to make an increase in the combustion temperature and
specific impulse. Globally, ballistic characteristics increase sensitively [2].
The present work consist on the incorporation of aluminium powder finely divided (at
different concentrations) acting as combustible in the elaboration of composite propellant
with ammonium perchlorate as oxidizer, epoxy resin as binder and copper chromite as
burning catalyst.
The oxidation basic product of aluminium is aluminium oxide (Al2O3), a product of
condensed phase. The resulted particles generate undesirable secondary effect such as
erosion of nozzle. However, they product an extremely desirable effect, which consist to
effective reducing of combustion instabilities [3].
680
2.
EXPERIMENTAL
In order to proceed to the elaboration of highly performed propellants, it is necessary to

adjust some parameters such as: the rate between the ingredients, particle size of the oxidizer
and aluminium powder, the composition of binder.
The chemicals used in this study were catalysed composite propellants consisted of an
oxidizer charge: ammonium perchlorate, an epoxy-binder with specific additives, and copper
chromite (Cu2Cr2O5) as combustion catalyst at a rate of 85/14/1 by weight respectively [4,5,6].
The experimental plan of this study is structured in two parts, in order to investigate
thermally and energetically, the effect of aluminium powder on the formulation of
propellant:
-
The first part, consists to the elaboration of the different composite propellant systems
with ammonium perchlorate as oxidizer, epoxy resin as binder, copper chromite as
catalyst, and by adding aluminium powder at different rates (between 0 and 20 % in
weight of the total mass of the propellant with a step of 2.5 %). The rate of copper
chromite is constant for all the preparation, while the rate of ammonium perchlorate
varies between 85 % to 65 % (wholly, nine systems of propellants were prepared).
The second part consist to the characterisation of propellants samples:

-
by DSC in order to determine thermal parameters, with a sample mass around 1.2
mg heated from 50 C to 400 C at a constant heating rate of 5 C/mn
by TGA to study kinetically the decomposition of propellants, with sample masses
of 10 mg and scanning until 400 C at a heating rate of 5 C/mn, a high purity
nitrogen purge gas is used (flow rate of 20 cm3/mn)
using adiabatic calorimeter to determine the calorimetric potential of specimens,
using sample mass of 0.6 g (in powder form)
The materials used are listed in Table 1.

Table 1. Materials used
Designation
Description
Supplier
Oxidizer
Ammonium perchlorate
Fluka
Epoxy Binder
DGEBA Diglycidyl ether of bisphenol A: GY-260; MW = 381
Shering France
Company
Fuel
Aluminium powder (< 45 m)
Merck
Binder additive
Bentonite (nanocharge) Mx(Al4-xMgx)Si8O20(OH)4
Maghnia Algeria
Curing agent
Imidazoline polyamine (Eurdur 460)
Shering France
Company
Combustion
catalyst
Copper chromite (Cu2Cr2O5)
Merck
Solvant
Tetrahydrofuran (THF = C4H8O)
Fluka
Moderator
Aerosil 200 (Silicone dioxyde)
Merck
681
3.
RESULTS
3.1.
DSC results
Exothermic
From DSC plots of the Fig. 1, the main parameters are determined by the 1022 Series
Software of Perkin Elmer [7], and they are decomposition temperature (auto-inflammation) of
propellant (Td) and maximum exothermic decomposition peak temperature (Tm).
0 % Al
2.5 % Al
5 % Al
12.5 % Al
210
240
270
300
330
360
390
Temperature (C)
Fig 1.
DSC plots of propellant Systems with 0 %, 2.5 %, 5 % and 12.5 % of Al
DSC results i.e. decomposition temperatures (Td) of propellants for all the samples are
shown in figure 2.
360
T (C)
350
340
330
320
310
300
0
10
15
Al content (%)
Fig 2.
Decomposition Temperature of Propellant Systems by DSC
682
20
3.2.
TGA results
From TGA Plots, the decomposition temperature (Td) versus aluminium rate is
determined. By an other way, the activation energy Ea of the different systems is calculated
using Freemann and Caroll method.
TGA analyses gives the results reported below in figure 3 (Td) and figure 4 (Ea).
400
T (C)
380
360
340
0
10
15
20
Al content (%)
Fig 3.
Decomposition Temperature of Propellant Systems by TGA
Ea (KCal/mol)
25
20
15
10
0
10
15
Al content (%)
Fig 4.
Effect of Aluminium Content on Activation Energy
683
20
3.1
Calorimetric Potential results
Figure 5 gives the main results of the measured calorimetric potentials H0 for each
system of propellants.
9000
8500
H0 (J/g)
8000
7500
7000
6500
6000
0
10
15
20
Al content (%)
Fig 5.
4.
Effect of Aluminium Content on Calorimetric Potential
DISCUSSION
From obtained DSC plots, it can be observed two thermal transitions characterized by
the yielding of two peaks on all thermograms as following:
-
an endothermic peak an endothermic peak appeared at around 240 250 C and

has been assigned to the allotropic transformation of crystal structure of
ammonium perchlorate from orthorhombic to cubic
and a maximum exothermic peak appearing at different temperatures above
300C, resulting from the high temperature decomposition reaction of the
propellant.
It can be emerged a negative displacement of the onset as well as an increase of the

maximum decomposition peak temperature (temperature at which the reaction will proceed
to completion), when the aluminium rate increases until the value of 12.5 %, see figure 1.
Above 12.5 % in aluminium, this displacement becomes positive (higher obtained
temperatures).
The addition of aluminium powder allow to facilitate the decomposition of propellant.
The DSC results show a decrease of 50 C in decomposition peak temperature of aluminised
propellant, in comparison with the non-aluminised propellant (fig. 2). This can be explained
by the fact that the aluminium powder acts as combustion accelerator, allowing the increase
684
of thermal transfers by increasing the contact surface when burning the epoxy resin and
ammonium perchlorate and the rest of ingredients.
When aluminium rate is up to 12.5 %, the decomposition temperature of propellant will
increase; this may be explained by the fact that the combustion reaction becomes less fast,
because the oxygen amount is not sufficient to make a total combustion; in an other hand
because this is due to the decrease of ammonium perchlorate ratio, with higher rate of
aluminium powder.
Otherwise, this can be attributed to the inhibitor effect of alumina (Al2O3) generated by
the process combustion from a critical concentration of 12.5 % (weight %). This effect has
been anyway set experimentally, where at the end of DSC experiment, it is deposing in the
measurement cell pan, an amount of residual white solid attributed to alumina.
Consequently, in the case of high energetically rocket propellants as the case of
aluminised propellants based epoxy resin as binder, it is necessary to adjust the ratio between
aluminium powder and ammonium perchlorate to optimize thermal properties (the lower
decomposition temperature) and energetic characteristics (a maximum generated energy).
Finally, from the obtained DSC results, it emerges that the optimal weight rate of
aluminium powder to be used in propellant is 12.5 %, corresponded to the lower autoinflammation temperature (the faster chemical reaction decomposition).
Concerning TGA results, it has been found that in the one hand the decomposition
temperature of propellant follow the same profile like shown with DSC results (fig. 3);
i.e. Td decrease until a minimum value corresponded to a content of aluminium powder of
12.5 %. Thus TGA results have confirmed the results of DSC measurements.
In the other hand, values of activation energy revealed that Ea decreases with the add of
aluminium powder reaching a minimum at a critical concentration of 12.5 % (fig. 4.). Then
Ea increases by adding more aluminium content. The reaction of chemical decomposition
becomes faster at an optimum of 12.5 % in aluminium powder, which is in agreement of
DSC results.
The calorimetric potential shows an increase while aluminium content increases (fig. 5).
This is attributed to aluminium effect, which increase sensitively the burning temperature of
propellant [2,8,9]. It has been found that when aluminium content is up to 12.5 %, a deposit of
alumina is observed on calorimetric cell. A maximum of generated energy is reached at a
concentration of 12.5 % in aluminium powder (H0 = 8280.63 J/g).
This result can be explained by the role of burning product (essentially alumina), when
the content of aluminium is up to a concentration of 12.5 %. Alumina play a role of inhibitor
at this stage and then the chemical reaction of decomposition of propellant becomes so less
energetic and more slowly (as shown in DSC and TGA analyses). In catalyse, alumina is
well known for its contaminated character. In the case of aluminised propellants, during their
combustion, the reactional sites will be contaminated, which effect can generate combustion
with low vivacity (low combustion energy and burning rate), alumina will saturate these
reactional sites by agglomeration.
685
5.
CONCLUSION
The incorporation of aluminium powder in the composite propellant allow to increase

sensitively thermal, kinetic and energetic properties of rocket propellant system based
DGEBA epoxy resin as binder, ammonium perchlorate as oxidizer and copper chromite as
burning catalyst.
On the whole of this work, the three methods of characterisation have shown the same
result: the optimal weight rate of aluminium powder given the best thermal, energetic and
kinetic properties for this propellant formulation is around 12.5 %.
The thermal study carried out by DSC has shown that the add of aluminium powder
decrease the decomposition temperature of propellant until a temperature corresponded to a
rate around 12.5 % in weight of the total mass of propellant.
The TGA analysis has in one hand confirmed the results obtained by DSC (by adding
aluminium, the degradation temperature of propellant determined TGA follow the same
evolution that the one obtained by DSC measure). In second hand, by using Freemann and
Carroll method, it has been possible to determine the activation energy of the different
propellant system. A critical concentration of aluminium powder around 12.5 % has been
found, corresponded to the minimal activation energy.
The energetic study by adiabatic calorimeter has shown an increase of calorimetric
potential with the add of aluminium powder. The maximum value was reached around a rate
per cent of 12.5 in weight of aluminium powder.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
A. DOKHAN, E. W. PRICE, R. K. SIGMAN, and J. M. SEITZMAN: The Effect of Aluminium

Particles Size on the Burning Rate and Residual Oxide in Aluminized Propellants, School of
Aerospace Engineering Georgia Institute of Technology, Atlanta, GA, 2001
A. DAVENAS et COLL.: Technologie des propergols solides, Masson Editor, Paris, 1989
V. YANG, T. B. BRILL and W. Z. REN: Solid Propellant Chemistry Combustion and Motor
Interior Ballistics, Progress in Astronautics and Aeronautics, Zarchan Edition, No. 185, 2000
H. ATTOU and A. REDJEM: Effet de la Bentonite sur les Proprits d'un Systme poxy DGEBA/
Polyamine Utilis dans un Propergol Composite, EMP, Algiers, 2003
M. B. MAOUCHE and A. REDJEM: Les Catalyseurs de Combustion et leur Influence sur les
Proprits Energtiques et Cintiques des Propergols Composites, EMP, Algiers, 1995
A. ABDI and K. KHALDOUN: Influence des Catalyseurs de Combustion sur les Proprits
Thermocintiques des Propergols Composites, EMP, Algiers, 1996
T. L. Mc NAUGHTON and C. T. MORTIMER: Differential Scanning Calorimetry, Perkin Elmer,
USA, 1975
J. QUINCHON, J. TRANCHANT and E. COHEN-NIR : Les Poudres, Propergols et Explosifs,
Tome4 : Les Propergols, Technique et Documentation - Lavoisier, Paris, 1991
G. P. SUTTON and O. BIRLARZ: Rocket Propulsion Elements, Wiley-Interscience, 7th Edition
N.Y, USA, 2001
686
CRYSTALLIZATION OF HEXANITROSTILBENE
Andrzej Orzechowski*, Dorota Powaa *,
Andrzej Maranda**, and Wojciech Pawowski***
* Institute of Industrial Organic Chemistry, 6 Annopol St, 03-236 Warsaw, PL
** Military University of Technology, 2 Kaliskiego St, 00-908 Warsaw 49, PL
*** Warsaw University of Technology, 3 Noakowskiego St, 00-665 Warsaw, PL
Abstract:
In this paper, the method of synthesis and crystallizations of hexanitrostilbene are
presented. The aim of the work was to obtain cubic crystals of HNS. Several solvents for
crystallization of HNS were checked. The 100% fuming nitric acid as a solvent to
crystallization was the best. The influence of various parameters on crystallization
process was investigated. The particle size distribution and the tap density in the
crystals obtained are presented.
Keywords:
1.
crystallization, hexanitrostilbene, HNS
INTRODUCTION
The development of space programmes, the drilling of deep oil wells etc. has resulted in
the need for heat resistant explosives. This is a new line of world-wide researches with the
aim of producing explosives or explosive compositions that are safe, reliable and stable at
high temperatures.
Hexanitrostilbene is interesting explosive from the point of view of thermal stability.
The interest in this material has been growing in recent years. This explosive was employed
for seismic experiments on the Moon and was used in achieving stage separation in space
rockets. Moreover, the additive of hexanitrostilbene to melt-cast TNT improves internal
structure of obtained mixture.
In the literature, various synthesis methods of hexanitrostilbene are referred. The
researches on HNS preparing are described in [1-6] and chosen methods of HNS
crystallization are shown in [7,8].
In our laboratory, we conduct researches on crystallization of hexanitrostilbene in order
to obtain a suitable form of this explosive for application in Plastic Bonded Explosives.
2.
EXPERIMENTAL
2.1
Properties of hexanitrostilbene
Hexanitrostilbene (HNS) is light yellow crystalline substance. The melting point of

hexanitrostilbene is 318 C, and the density is 1,74 g/cm3. This material is a heat resistance
explosive with thermal stability over 300 C. The detonation parameters of HNS are similar
to those of trinitrotoluene [9]. The impact sensitivity of this explosive is 5 Nm and the friction
sensitivity is 240 N [10]. Heat of detonation of HNS determined by calorimetric method for
24 g of charge and density 1,65 g/cm3 is 4600 kJ/kg [11].
687
Fig 1.
2.2
Crystals of HNS without crystallization
Synthesis of hexanitrostilbene
The hexanitrostilbene was prepared by reacting TNT with sodium hypochlorite in

tetrahydrofurane. The synthesis coursed in diphase system with catalyst of interfacial
transfer (N,N,N-tert-butylbenzylamineamonium chloride).
TNT was dissolved in tetrahdrofurane and cooled to 0 C. This solution was then added
rapidly with stirring to 5% aqueous sodium hypochlorite with addition of catalyst, previously
cooled 0 C. The temperature increased sharply and was held at boiling point of THF for 30
minutes. Next, the crystals of product were filtered and washed with cold water.
2.3
Crystallization of hexanitrostilbene
In order to obtain the applicable form of hexanitrostilbene we carried out several

experiments of crystallization. As a solvent we used N,N- dimethylformamide,
dimethylsulfoxide, N-methylpyrrolidone and fuming nitric acid. During the tests different
parameters were changed, for example: solvents, mass ratio of compound, stirring speed and
cooling rate. Among the solvents, which were chosen to crystallization only, fuming HNO3
allowed to obtain favourable results of experiments.
Crystallization of HNS from fuming nitric acid
The crystallization of hexanitrostilbene was made in glass laboratory equipment. The
mixture consisted of HNS and fuming nitric acid in different weight ratios. The mixture was
stirred and gradually heated until the crystals were dissolved. Then the solution was
gradually cooled with various cooling rate. Next, the crystals of hexanitrostilbene were
filtered and washed with water and acetone. The shape and size of obtained crystals were
depended on parameters of crystallization.
Example 1
20 g of fine HNS was added to a flask with 900 ml fuming nitric acid. The mixture was
gradually heated with mechanical stirring to 60 C. Then the stirring was stopped and the
mixture was cooled to ambient temperature. The product was filtered and washed with water
and acetone. The HNS crystallized in the form of long plates with bulk density 0,27 g/cm3
and tap bulk density 0,49 g/cm3. The crystals are presented in Fig. 2.
688
Fig 2.
Crystals of HNS obtained by the method described in example 1
Example 2
20 g of fine HNS was added to a flask with 300 ml fuming nitric acid. The mixture was
gradually heated with mechanical stirring (700 rpm) to 80 C. Then the mixture was
gradually cooled to 5 C with cooling rate 15 C/min. The product was filtered and washed
with water and acetone. The crystals were similar to these, which were obtained in
example 1. The bulk density of HNS was 0,24 g/cm3 and tap bulk density 0,45 g/cm3 The
product is shown in Fig. 3.
Fig 3.
Example 3
This test was carried out in the same way like previous (example 2). The difference
between these two examples was cooling rate. In this case it was about 5 C/min. The
obtained product about bulk density 0,44 g/cm3 and tap bulk density 0,64 g/cm3 is presented
in Fig. 4. This experiment allowed to obtain crystals of hexanitrostilbene about higher bulk
density than above.
689
Fig 4.
Example 4
In order to obtain suitable form of HNS crystals, a few crystallization tests with low
cooling rate (below 0,5 C/min) were carried out. The temperature 20 C was attained after
5 h of cooling. These experiments differed in stirring speed of the mixture in comparison
with the previous examples. The crystallization parameters and obtained bulk density are
presented intable 1. After these tests coarse hexanitrostilbene about high bulk density
(~1 g/cm3) was obtained. The crystals of HNS and grain size characteristics of its are shown
in Fig. 5 Fig.12.
Table 1. Parameters of HNS crystallization with cooling rate under 0,5 C
1/25
Speed of
stirring
[rpm]
700
Bulk
density
[g/cm3]
0,86
Tap bulk
density
[g/cm3]
1,00
1/25
1200
0,91
1,12
1/25
1400
0,93
1,08
1/35
1400
0,97
1,13
1/25
1600
0,68
1,00
Sample
Volume ratio
HNS/HNO3
Fig 5.
Crystals of HNS after crystallization from HNO3 with stirring 700 rpm
690
8
7
Volume [%]
6
5
4
3
2
1
0
5,7
71
137
204
270
336
403
469
535
602
668
734
801
Particle diameter [m]
Fig 6.
Particle size distribution of HNS after crystallization from HNO3

with stirring 700 rpm
Fig 7.
10
9
Volume [%]
8
7
6
5
4
3
2
1
0
5,7
71
137 204
270
336
403 469
535
602
668 721
787
Fig 8.
Particle size distribution of HNS after crystallization from HNO3

691
Fig 9.
8
7
Volume [%]
6
5
4
3
2
1
0
5,7
71
137
204
270
336
403
469
535
602
668
734
801
Fig 10. Particle size distribution of HNS after crystallization from HNO3
Fig 11. Crystals of HNS after crystallization from HNO3 with stirring 1600 rpm
692
10
9
8
Volume [%]
7
6
5
4
3
2
1
0
5
71
137
204
270
336
403
469
535
Fig 12. Particle size distribution of HNS after crystallization from HNO3
3.
SUMMARY
The modification of hexanitrostilbene synthesis was carried out. The application of

diphase method allowed enlargement of process scale. The following parameters were
changed:
the solvent tetrahydrofurane as solvent about high boiling temperature was
used,
the quantity of catalyst was decreased,
reacting substances were dozed in the portions.
The process of hexanitrostilbene crystallization according to example 4 allowed to
obtain coarse product of high bulk density (above 1 g/cm3). The highest bulk density of the
HNS was 0,97 g/cm3 and tap bulk density was 1,13 g/cm3.
Our experiments showed that the size and form of crystals could be regulated by speed
and time of stirring.
The yield of the crystallization process ranged from 60 to 65%. The melting point of the
product after crystallization was 322 323 C.
693
REFERENCES
[1]
E.E. GILBERT: The Preparation of Hexanitrostilbene from Hexanitrobibenzyl, Propellants and

Explosives, 1980, 5, 15.
[2] R. KUBOSZEK, K. SO, M. MKOSZA, W. ZAACHOWSKI, R. PRZYBYLIK: Nowa
metoda otrzymywania heksanitrostilbenu, Patent polski 101104, 1978.
[3] R. KUBOSZEK, K. SO, M. MKOSZA, W. ZAACHOWSKI, R. PRZYBYLIK: Nowa
metoda otrzymywania heksanitrodibenzylu, Patent polski 101422, 1979.
[4] P. GOLDING, G.F. HAYES: Studies on the Synthesis of 2,2,4,4,6,6 Hexanitrostilbene,
Propellants and Explosives, 4, 115-120, 1979.
[5] G.P. SOLLOT, M. WARMAN, E.E. GILBERT: Base-Catalyzed Dehydrogenation of 2,2, 4,4,
6,6Hexanitrobibenzyl by Quinones, J. Org. Chem., 44, 3328-3330, 1979.
[6] P. GOLDING, G.F. HAYES: A Parametric Study of the Synthesis of 2,2,4,4,6,6
Hexanitrostilbene from Trinitrotoluene and Sodium Hypochlorite, Propellants, Explosives,
Pyrotechnics 8, 35-39, 1983.
[7] Z. JALOVY, P. MARECEK, K. DUDEK: Crystallization of Hexanitrostilbene, Proc. of VI
Seminar, New Trends in Research of Energetic Materials, Pardubice 2003.
[8] S.M. HARRIS, S.E. KLASSEN: Hexanitrostilbene (HNS) Development for Modern Slapper
Detonators, Proc. of 32nd International Annual Conference of ICT, Karlsruhe 2001.
[9] A. MARANDA, S. CUDZIO, J. NOWACZEWSKI, A. PAPLISKI: Podstawy chemii
materiaw wybuchowych, WAT Warszawa, 1997.
[10] J. KHLER, R. MEYER: Explosives, 4 rev. and extended ed., Wiley VCH Verlag GmbH,
Weincheim, 1993.
[11] D.L. ORNELLAS: Calorimetric determinations of the heat and products of detonation for
explosives, Rep. UCRL-52821, Lawrence Livermore Laboratory, University of California,
USA, 5.04.1982.
694
CHARACTERIZATION OF COMPOSITE SOLID ROCKET

PROPELLANT USING DMA
Ji Pachm, and Jakub eleovsk
University of Pardubice, CZ - 530 09 Pardubice
Abstract:
Dynamic mechanical analysis has been used to characterize mechanical properties
of selected composite solid rocket propellant. Different experimental approaches
to obtain data for determination of activation energies Ea associated with glass
transition are presented. The experimental results show that the Ea values differ
depending on the method used for evaluation. Some limitations of the time temperature
superposition principle are discussed.
Keywords:
1.
DMA, solid propellant, activation energy, glass transition,

time-temperature superposition, WLF, mechanical properties
INTRODUCTION
The composite rocket propellants are highly filled composite materials. The typical
composition of composite solid rocket propellant is based on crystalline solid oxidizer bound
together with polymeric matrix. Many other additives like metal particles, nitramine crystals,
ballistic modifiers etc. may be included in the final mixture depending on the desired use of
the propellant. Deterioration of propellants properties may be negligible at room
temperature but quite indispensable in case of thermal cycling due to a daily or annual
temperature oscillations (stress buildup, cracking), higher storage temperatures (creep) or
even at very low storage temperatures (physical aging). Chemical changes taking place in
propellant may result in additional cross-linking, degradation of polymer backbone, surface
hardening, heat evolution, swelling etc. These chemical changes may in the long term affect
mechanical properties. The ballistic behavior may be affected when cracking or significant
deformation of propellant grain occurs.
In standard tensile or compression tests samples are usually tested to failure either under
constant deformation rate (STANAG 4506, 4443). Creep or relaxation behavior may be
determined usually using the same apparatus but holding the deformation (strain) or force
(stress) constant after some initial step loading (STANAG 4507). These approaches require a
large amount of samples and a significant amount of time. Parameters describing fracture
behavior as well as a rheological behavior in non-linear region (high deformation and/or
high deformation rates) are accessible in this way. Structural analysis including
determination of changes in transition regions can however hardly be determined. A
sensitive determination of structural changes is difficult even when relatively small
temperature steps are taken. Interpolations are possible; however values such as modulus
may change rapidly around a glass transition temperature in the orders of magnitude making
the interpolations inaccurate.
695
Dynamic mechanical analysis on the other hand tests material in the entire temperature
interval and since small deformations within the linear region are applied only a few pieces
of samples are required. This method is an excellent choice for identification of structural
changes occurring in the material including determination of apparent activation energies as
well as temperature and frequency dependencies of these transitions. Transition regions are
closely related with molecular mobility of polymeric matrix of solid propellant and changes
in chemical composition due to an aging may therefore studied using DMA. The failure
properties are however not attainable with this technique.
A careful application of theory of time temperature equivalence (superposition TTS)
principle can lead to a prediction of long term behavior from short term tests. This principle
in its simplest form implies that viscoelastic behavior at one temperature can be related to
that at another temperature by change in time scale only.
Shift factors determined from DMA measurements may than be used for prediction of
failure parameters from ambient temperature measurements. This approach however
requires, that all relaxation processes in material are influenced equally with the change of
temperature (the relaxation spectrum changes in the same manner for all relaxation times)
and the mechanism of relaxation does not change (new relaxation processes do not appear
and old do not disappear with change of temperature!). An application of the superposition
principle in the temperature range containing changes in relaxation mechanisms is quite
questionable but has been reported in literature [1], [2].
The applicability of TTS principle has been evaluated by [3] for polymer blends. It has
been pointed out that not all materials that have been expected to obey TTS really obey it. In
case of rocket propellants the situation is further complicated by addition of modifiers and
presence of crystalline phase. The general applicability of TTS should therefore not be
assumed.
A simple way to check whether superposition can or can not be used is to plot loss factor
vs. log of modulus so called Wicket plot [4]. The plot must be a unique curve and valid data
must lie on or close to this curve. Data points that deviate significantly may be considered
suspect. A modification of this approach has also been applied plotting loss angle vs.
absolute value of the complex modulus creating so called van Gurp-Palmen plot [3].
For the interval of temperatures where TTS holds activation energies of transition (based
on barrier theories) can be calculated using Arrhenius equation [5] in. equation 1. The
activation energy of transition must be constant for the superposition to make sense. This is
often assumed without proof and activation energies are calculated [6]. It has been however
reported by [7] that assumption of constant activation energy should not be generally
expected.
In Arrhenius equation f stands for frequency (Hz), Ea for activation energy (J.mol-1), R
for universal gas constant (8,314 J.mol-1.K-1) and T for temperature (K).
f = f0e
Ea
RT
(1)
According to Ward [8] equation 1 bares a name Arrhenius because he first used it to show
the influence of temperature on the velocity of chemical reactions. The same equation is
according to Malkin [9] called Arrhenius-Eyring. This relation can be rewritten for two
measured frequencies in the form of:
696
log aT =
log f1 E a 1 1
log e
=
log f 2
R T2 T1
(2)
where log aT is called the shift factor. This shift factor can also be obtained using Williams
Landel Ferry (WLF) equation [10] in the form:
log aT =
C1 (T T0 )
C 2 + (T T0 )
(3)
where T0 is a chosen reference temperature (usually in range of Tg - Tg + 50K) and C1 and C2

are constants. The applicability of this equation is usually reported to be in temperature
range from Tg to Tg + 100K. Other shift equations also exist but are not as widely used [4].
2.
EXPERIMENTAL
2.1.
Sample preparation
The composite propellant described herein has been drawn out from a rocket motor in
the course of periodical surveillance testing. The samples were prepared in the form of
rectangular bars 40 x 10 x 5 (2) mm by die cutting of plates milled from the propellant grain
and vacuum-sealed in an aluminum-plastic foil. One set of samples was used for chemical
analysis, one for initial testing of mechanical properties and the rest used for accelerated
aging. Sample described herein as aged was stored for 113 days at 70C. The results of the
chemical analysis and the influence of accelerated aging on chemical composition and
sensitivity have been described elsewhere [11].
2.2.
Dynamic mechanical analysis
Measurements of mechanical properties were carried out using dynamic mechanical

analyzer DMA 04 DX produced by RMI. Liquid nitrogen evaporation unit was used for sub
ambient temperature measurements down to 100C. Samples (rectangular blocks) were
tested in a single cantilever mode. Measurements were carried with constant amplitudes of
deformation set to 0.030 mm (sample thickness 2 mm, width 10 mm and free length 27
mm) or 0.050 mm (sample thickness 5 mm, width 10 mm and free length 27). The
frequencies ranged from 0,01Hz to 20 Hz. Simple sine wave form was used at a linear
heating rate as well as at a stepwise temperature function.
2.3.
Determination of activation energy of transition using Arrhenius

relationship
The apparent activation energy of relaxation transition can be determined in a number of

different ways using DMA. All methods require determination of dynamic data with relation
to temperature and frequency. The usual approach is to hold either frequency or temperature
constant and measure dynamic properties while varying the other one. It is also possible to
use multiple wave instead of a sine wave and than decompose the resulting signal with the
aim of FFT. Result of any such approach can be expressed as a 3D plot. An example of such
data matrix is presented for E of unaged propellant in fig. 1.
697
Fig 1.
Plot of E vs. temperature and frequency for unaged propellant
The first method of determination of activation energy of transition (according to [6], [8])
consists of a measurement of dynamic properties at various heating rates and constant
frequencies. In this case Ea can be determined according to Arrhenius equation (eq. 2) from
the plot of inverse of the temperature 1/T vs. ln (loge) of frequency (or log10 than 2.303
correction must be applied). Since powerful computer algorithms for non-linear least square
fitting are readily available there is a possibility to determine the activation energy directly
from dependence of peak temperature vs. frequency. A GNU OCTAVE routine leasqr
utilizing Marguardt Levenberg method was used to fit the data presented herein. An easy
to identify point on the temperature dependence of dynamic property (peak of E, peak of tg
, onset of E or E*) is chosen and appropriate temperature is evaluated (a constant
activation energy through the process is assumed, however without proof).
Second possibility is to measure of dynamic properties at constant temperatures
(isothermal steps) while varying frequencies. A dynamic property (most often E, E or E) is
determined at discrete frequencies. Data are plotted as a dynamic property (commonly E or
E) vs. log f and reference temperature is chosen. Data corresponding to the temperatures
below and above this reference temperature are shifted horizontally along the log f axis. The
magnitude of the shifts necessary to form unique curve are then plotted against temperature
and evaluated using Arrhenius or WLF equation. Reference temperature as well as
temperature interval should be chosen with the TTS limitations in mind.
698
3.

3.1.
Isochronal and isothermal measurements
Temperatures scan at 1Hz was run first to determine the temperature range of interest.
As can be seen from fig. 2 and 3 there is practically only one transition taking place at a
temperature range from 5 to + 20C (depending on the method of evaluation) at 1Hz. It was
therefore decided to run measurements with temperature program from 60C to + 100C.
Fig 2.
Temperature dependence of E* and tg for aged and unaged propellant
Fig 3.
Temperature dependence of E* and tg for aged and unaged propellant
699
The influence of heating rate on the position of Tg was studied using up to 12 heating
rates. Linear heating rates from -7 to 7C.min-1 were used for frequencies 0.1, 1 and 5 Hz. It
has been expected that thermal gradient would build up in the sample and that the
temperature of the sample would lag behind the surrounding temperature in case of heating
and surpass during cooling. The thermal gradient was expected to be smaller for slower
heating/cooling rates.
The results presented in fig. 4 show, that extrapolation of peak temperature is possible
giving reasonable fit as shown in tab. 1.
5Hz
50
1Hz
0,1HZ
E peak temperature [ C ]
40
isotherms
30
20
10
0
-10
-20
-10
-5
10
-1
dT/dt [ C*min ]
Fig 4.
Glass transition temperatures of unaged propellant determined from E peak

at different heating and cooling rates. The diamonds represent peaks of E
determined from isothermal steps.
Data obtained at different heating rates were interpolated to yield intersection with yaxis. This intersection corresponds to extremely slow heating rate isotherm and is reported
in tab. 1 as a peak temperature of E at dT/dt = 0. These values are compared to the values
obtained from isothermal measurements by Bezier interpolation. Since frequency of 5 Hz
has not been used for isothermal measurements, interpolated value is used.
Table 1. Dependence of E peak on heating rate compared with isothermal results.
frequency [Hz]
5
1
0.1
R2
dT/dt = 0 C.min-1
isotherm
[C]
[C]
0.9978
17.1
13.8 interpolated
0.9986
12.6
9.4
0.9985
4.8
4.9
700
Next stepwise temperature function was used. In the first step material was cooled to
60C with heating rate of 3 C.min-1 and held 5 min at this temperature. Then steps of 10C
with heating rate 3C.min-1 were applied. Duration of isothermal ramp was set to measure at
least 4 points for each frequency (usually 40 minutes). As an example of such measurement
results a raw data of E for the unaged propellant are presented in fig. 5.
Fig 5.
Plot of E vs. frequency for isothermal steps from -60C to 100C
It can be seen from the temperature preview fig. 2 and 3, that not all points can be
superimposed. There is a glass transition practically splitting the temperature interval into
two regions. The Wicket plot (fig.6) also shows that some points at lower modulus values
corresponding to high temperatures are suspicious.
Fig 6.
Wicket plot of unaged and aged propellant
701
In spite of the above mentioned facts superposition was applied to the entire data set to
show, how shift factors and calculated activation energies differ from the ones determined
for above Tg region. The reference temperature was chosen to be 20 C. The resulting
composite master curves are presented in fig. 7. They may be used for a brief comparison of
aged and unaged propellant.
Fig 7.
3.2.
Plot of master curves obtained from stepwise measurements by time

temperature superposition for unaged and aged propellant.
Ea from isothermal measurements
The mastercurves presented on fig. 7 were obtained by application of time temperature

superposition with the shift factors shown in fig. 8. The values of shift factor log aT differed
slightly when determined from E and E. This can be considered as an indication of
problematic applicability of TTS. In spite of this TTS has been applied in the usual way. It
can be seen, that there is a change of pattern at 10 or 20C associated with the glass
transition region as identified by standard temperature scan.
The region above Tg is enlarged in fig. 9 with fit only for displayed values. Results
together with the results of fitting the entire range are summarized in tab. 2. It is evident that
Ea incorrectly calculated from the entire temperature range are only slightly different from
the values calculated for temperatures above Tg. The quality of the fit in terms of R2 is
approximately the same and therefore should not be used as a criterion of goodness of
superposition. WLF parameters differ to a much bigger extent.
702
Fig 8.
Plot of shift factor vs. temperature determined from shift of E the entire
temperature range of measurement. The data are fitted with Arrhenius and WLF
equations.
Fig 9.
Plot of shift factor vs. temperature determined from shift of E

for the temperatures above Tg . The data are fitted with Arrhenius and WLF
equations.
703
Table 2. Comparison of Ea and WLF parameters calculated from different

temperature ranges
range -60 +100 C
Ea [kJ mol-1] R2
E
240.7
0.9913
233.7
0.9852
242.6
0.9805
273.1
0.9887
E
E
3.3.
unaged propellant
range +20 +100C
range -60 +100 C
Ea [kJ mol-1] R2
WLF C1/C2 R2
50.7
245.8
0.9782
0.9930
339.3
98.1
276.7
0.9719
0.9934
644.0
aged propellant
90.8
277.0
0.9418
0.9844
596.5
86.3
305.9
0.9489
0.9906
505.3
range +20 +100C

WLF
R2
19.7
0.9935
98.7
23.5
0.9863
108.9
16.1
55.5
20.9
75.2
0.9891
0.9804
Ea from fixed frequency measurements (isochrones)
Two approaches have been applied to demonstrate determination of activation energies

using peak E values of the unaged propellant from isochronal measurements. The results
are summarized in tab. 3.
In he first approach fixed frequency measurements at varying heating/cooling rates as
presented before in fig. 4 were evaluated. Activation energy was calculated for each heating
rate, including interpolation to dT/dt = 0C.min-1. The cooling trials are presented only for
demonstration purposes. They were primarily used to confirm the extrapolation to
dT/dt = 0C.min-1 from heating rates.
5 C/min
3C/min
1.5
1C/min
0C/min - int.
isotherm
ln f [Hz]
0.5
-1C/min
-3C/min
-0.5
-5C/min
-1
-1.5
-2
-2.5
-3
0.003 0.003 0.003 0.003 0.004 0.004 0.004 0.004 0.004 0.004
1 / TE peak temperature [C-1]
Fig 10. Dependence of logarithm of frequency (ln f) vs. inverse of E peak temperature
(1/T) at heating rates 1, 3, 5 C/min. Peak temperature at heating rate
0C/min-int. is an intersection with y axis. Diamonds represent data from
isothermal measurements.
704
Second approach is based on a creation of the isochrones from isothermal data by means
of interpolation. This is obviously more precise if small temperature steps are applied. The
values of E peak maximums shown in fig. 11 are used in this article to demonstrate the
approach. These isochronal measurements were obtained from isothermal stepwise data
presented in fig. 5 and therefore do not represent simple temperature run as could be
expected. This approach eliminates the need for isochronal run at very low heating rate that
would be necessary because of the 0.01Hz frequency. Results are compared with the
previous ones in fig. 10 (labeled isotherm).
Table 3. Activation energies determined from isochronal measurements
heating rate [C.min-1]
-5
-3
-1
1
3
5
0 int.
isotherm
Ea [kJ.mol-1] 166.5 164.2 220.5 228.1 265.3 235.0 205.4 278.4
R2 [ ]
0.9733 0.9935 0.9839 0.9998 0.9990 0.9959 0.9974 0.9697
As can be seen the activation energies for unaged propellant determined from the peak
values differ significantly depending on the heating rate and it was therefore decided to omit
the measurements for aged propellant. A comparison of activation energies determined by
presented methods is summarized in fig. 12.
Fig 11. Peaks of E of unaged propellant at frequencies from 0.01 Hz to 20 Hz.

Curves were obtained from isothermal measurements by Bezier interpolation.
705
linear heating rate
300
interp.
WLF
250
-1
Ea [kJ.mol ]
Bezier interp.
200
150
-6
-4
-2
-1
heating rate [C.min ]
Fig 12. Activation energies associated with glass transition of unaged propellant
(inter. is interpolation of isochronal peaks of E to dT/dt = 0, Bezier interp.
is an interpolation of isothermal data to form isochronal peaks of E, WLF
is calculated from TTS above Tg.
4.
CONCLUSION
A characterization of one selected solid rocket propellant has been investigated by

means of dynamic mechanical analysis. Three possible methods - isochronal measurements
at fixed frequency, isothermal measurements at constant temperature and manipulation of
isothermal data to obtain isochronal relations - of determination of activation energy
associated with glass transition have been presented.
The activation energies determined from isothermal measurements by incorrect
application of TTS were compared with correctly evaluated ones. Activation energies and
WLF constants have been calculated for aged and unaged propellant using temperature
interval from -60 to + 100 as well as the above Tg interval from +20 to +100C for
superposition. The aged propellant showed slightly higher activation energies and also a
change in WLF parameters.
Isochronal dependencies of loss modulus E measured directly and determined from
isothermal measurements were used to calculate activation energies associated with glass
transition. The resulting values differed from 200 to 280 kJ.mol-1 (when excluding data from
cooling trials).
Acknowledgements
The cooperation with VTUVM Slavicin during preparation and aging of the sample is
gratefully acknowledged.
706
REFERENCES
[1]
[2]
[3]
[4]
HERDER G. WETERINGS F.P., KLERK W.P.C.: J. therm. Anal. Calorim., 72, 921 (2003).
FUENTE J.L., RODRIGUEZ O.: J. Appl. Polymer. Sci. 87, 2397 (2003).
GURP M., PALMEN J.:Rheol. Bull., 67, 5(1998).
DAVID I. G. JONES: Viscoelastic vibration damping, John Willey and Sons, Chchester, Great
Britain (2001).
[5] Z. ARRHENIUS: J. Phys. Chem., 4, 226, (1889).
[6] LI G., SULLIVAN P.L., THRING R.W.: J. Therm. Anal. Calor., 60, 377 (2000).
[7] VYAZOVKIN S., SBIRRAZZOULI N., DRANCA I.: Macromol. Rapid Commun., 25, 1708 (2004).
[8] WARD I.M., HADLEZ D.W.: An Introduction to the Mechanical Properties of Polymers, John
Willey and Sons, Chichester, Great Britain (1993).
[9] MALKIN A.Y.: Rheology Fundamentals, ChemTec publishing, Toronto, Canada (1994).
[10] WILLIAMS M.L., LANDEL R.F., FERRY J.D.:J. Am. Chem Soc., 77, 3701 (1955).
[11] PACHMAN J., HANUS M. SELESOVSKY J.: Some aspects of service life evaluation of composite
rocket propellants, In Proc: V. Seminar New Trends in Research of Energetic Materials,
Pardubice, Czech Republic 2002.
707
AN ANALYSIS OF BLAST WAVES PARAMETERS

AND UNDERWATER EXPLOSION TEST
OF EMULSION EXPLOSIVES AND DYNAMITES
Jzef Paszula*, Andrzej Maranda*, Andrzej Papliski*,
Barbara Gobek**, and Johann Kasperski **
* Military University of Technology, 00-908 Warszawa, ul. Kaliskiego 2, Poland
** Blastexpol, 59-145 Duninw, Poland
Abstract:
In the paper results of experimental investigations as well as numerical analyses of blast
waves parameters and underwater explosion test results of water-in-oil emulsion
explosives and dynamites is presented. In the experiments the overpressure magnitude
in transient shock wave generated by explosion of investigated explosives in air and
in water as well as time of bubble collapse was measured. Also numerical evaluation
of blast field in the space surrounding the explosive charge is performed. Comparison
of the blast wave intensity generated by various kinds of explosives with underwater
explosion test results is performed.
Keywords:
1.
emulsion explosives, blast waves, explosive performance,

underwater explosion test
INTRODUCTION
Many explosives is used in blasting industry are accomplished as multi-ingredient

mixtures in which the detonation process occurs in a nonideal regime. Differences
in physical and chemical characteristics of the components of the mixture results
in differences in their reaction susceptibility and overall kinetic mechanisms. In such
explosive compositions the chemical reactions in Chapman Jouguet point, as usually, aren't
still completed. Same additional reactions continue to proceed in the first stage of expansion
of detonation products. As the result, additional energy is liberated after the detonation zone.
Therefore, in such explosives the value of detonation velocity which is immediately
connected to the detonation energy i.e. energy evolved before the C-J point, cannot be
considered as the adequate measure of the energetic performance of the explosives
[Ornellas, 1982] [1].
Also other classic method like lead block test or ballistic mortar test used for estimation
of work capacity do not assure the correct illustration of real energetic potential of nonideal
explosives, mainly due to limited mass of samples used in this test [Persson and others,
1994] [2].
In the paper, which is continuation of earlier works [Paszula and others, Pardubice 2003,
Paszula and others, Beijing 2003, Paszula and others, Pardubice 2004] [3-5], the method of
underwater explosion test is compared with overpressure amplitude estimation in the blast
wave as the method of appraising of the work performance of an investigated explosives.
708
2.
EXPERIMENTAL
2.1
Explosives tested
Emulsion explosives at various mass densities as well as dynamites were used in

experiments. Base detonative parameters of the considered explosives are set up in Table 1.
Energetic parameters of investigated explosives were evaluated along with the Polish
Norm BN-80-6091-42.
Table 1. Composition and energetic parameters of tested explosives.
Explosive
Explosive Parameters
Emulsion explosive
Dynamite
ALAN 3
82GP
10G5H
Nitrates
88,50
87,60
83,40
62,0
Nitroesters
25,0
Aluminium
2,00
5,00
4,00
Water
4,00
3,20
7,10
Others
5,50
4,20
5,50
13,0
Mass density, [kg/m3]
1120
1130
1110
1140
Work capacity, [cm3]
310
360
Heat of explosion, [kJ/kg]
3640,60
4307,29
Explosion temperature, [K]
2668
3006
Explosion pressure, [MPa]
953,65
1021,13
Ideal work of explosion, [kJ/kg]
3046,38
3581,13
Composition
[cg/g]
ALAN 1
2.2
Blast overpressure measurements
The overpressure characteristics in the blast wave generated by investigated explosive were
measured by pressure gauges of PCB Piezotronics, Inc, Serie 137A. The gauges register
overpressure in the transient shock wave, sliding on the working surface of the device.
It works as an electromechanical transducer that converts the pressure (ratio of the force
acting on the working surface of the active element to magnitude of this surface) into the
proportional voltage signal. In the Table 2 technical characteristics of the pressure sensors
used in overpressure measurements are presented. The pressure gauges as well as the
explosive charge were so disposed to avoid of influence of reflected waves upon the
registered pressure records.
The sketch of the experimental set is presented in Fig. 1. Electric igniters initiated
emulsion explosive charges while electric igniters and aluminum hexogene detonators,
boosted dynamite charges.
Charges were executed from commercial ones with diameter about 72 mm. The masses
of charges carried out 350 5 g. in this way, prepared charges characterised the height to
diameter relation close to one.
709
Table 2. Technical parameters of used PCB pressure gauge.

Characteristic
Units
PCB 137A21
PCB 137A22
Measurement range
[MPa]
34,50
34,50E-03
Pressure resolution
[kPa]
0,69
0,69E-03
Sensitivity
[mV/kPa]
0,145
1,45
Resonance frequency
kHz
500
500
Temperature sensitivity coefficient.
[%/C]
0,054
0,054
quartz
quartz
Active element
Fig 1.
2.3
Experimental set up: 1 explosive charge; 2 pressure gauge; 3 PCB

signal conditioner; 4 digital oscilloscope; r distance from explosive
charge to pressure gauges
Underwater explosion test
During an underwater explosion appreciable part of explosion energy is following

through surrounding water as shock waves. The gas bubble expands due to high internal
pressure and temperature generated during explosion. Because of the water inertia, the
bubble expands so long, that inner pressure drops below hydrostatic pressure of surrounding
water. And then bubble reaches the maximum volume. Difference between hydrostatic and
bubble pressure is reason that water starts moving into the centre of the bubble. And the
bubble collapses.
We can record of prime shock waves and time of bubble first collapse. These records
allow us to calculate shock wave and bubble energy. We used well-known [5] equation to
calculate:
- The shock energy:
710
Es = ks
6 , 7
p 2 (t )dt
where: - characteristic shock wave time; ks coefficient.

- The bubble energy:
Eb = k bTb3
where: Tb - characteristic time; kb coefficient.
Time-overpressure distribution in the blast wave generated by investigated explosive
charges was measured by pressure gauges of PCB Piezotronics, Inc, Serie 138A. The gauges
register overpressure in the transient shock wave, sliding on the working surface of the
device. It works as an electromechanical transducer that converts the pressure (ratio of the
force acting on the working surface of the active element to magnitude of this surface) into
the proportional voltage signal. The sketch of the experimental set is presented in Fig 2 and
view of test arrangement on Fig 3.
Fig 2.
Experimental set up: 1 explosive charge; 2 pressure gauge; 3 detonator, H - depth of experiments, R - distance between explosive charge
and pressure sensors
Fig 3.
The view of test arrangement placed in water reservoir hung

under buoys
711
3.
RESULTS OF EXPERIMENTS AND DISCUSSION

3.1
Blast waves
In the Fig 4 an exemplary registration of overpressure course for Alan3 explosive is

presented. The 350 g charge of mass density of 1130 kg/m3 was tested. The registration is
made at the distance of 2 and 3 m. Several measurements of overpressure of shock wave
were executed for every charge the studied explosives. Upon the registered overpressure
history the maximal pressure value is estimated. The results of overpressure measurement as
well as the calculations of the specific shock wave impulse together with standard deviations
were put in Table 3.
0,07
0,06
Overpressure, MPa
0,05
0,04
0,03
0,02
0,01
0
-2
10
-0,01
-0,02
Time, ms
Fig 4.
Overpressure vs time diagram for Alan 3 emulsion explosive

mexp = 0,35 kg; exp = 1,15 g/cm3
Table 3. The results of measurements and the calculations of overpressure

and the specific impulse of shock wave for studied explosive materials
Explosive
D,
m/s
Density,
kg/m3
Alan 1
4550
Alan 3
Overpressure, kPa
Impulse, kPa*ms
2m
3m
2m
3m
1120
74,30 0,46
30,46 0,79
38,80 3,10
21,16 0,35
4280
1130
66,84 3,22
28,56 0,01
32,17 2,04
19,35 0,29
82GP
4600
1110
67,48 1,18
28,75 0,22
36,74 1,98
19,61 0,30
10G5H
6100
1480
63,95 2,45
34,58 0,75
38,22 2,14
24,22 0,65
As an important factor influencing upon the detonation parameters aluminium particles

reaction degree inside the detonation zone can be indicated. In first stage aluminum particles
absorb energy and act as an inert additive. The process of aluminum activation occurs on
behalf of energy liberated in other chemical reactions. Absorption of energy evolved in
parallel chemical reactions by aluminum alters the kinetic of this process. Therefore,
aluminium lowers detonation parameters, in particular the detonation velocity and as the
result lower the overpressure of blast waves. In the Fig 5 and Fig 6 the maximum
overpressure and specific impulse are presented.
712
2m
80
3m
70
Overpressure, KPa
60
67,48
66,84
74,3
63,95
50
40
34,58
30
28,56
30,46
28,75
20
10
0
Alan1
Fig 5.
Alan3
82GP
10G5H
Maximum overpressure of blast waves for investigated explosives

2m
40
3m
35
Impulse, kPa*ms
30
38,8
38,22
36,74
32,17
25
20
21,16
24,22
19,61
19,35
15
10
5
0
Alan1
Fig 6.
Alan3
82GP
10G5H
Calculated blast waves impulse for investigated explosives
How we can notice in Fig 3, the growth of content of aluminium powder addition causes
the lowering the value of shock wave overpressure (Alan 1 - Alan 3). A quick expansion
causing cooling of explosion products stops the reaction of aluminium with oxygen. So the
measurements of blast waves pressure do not show potential work performance of explosives
with addition of aluminum powder. The maximum overpressure for emulsion explosive
Alan 1 is near value for Dynamite despite its higher detonation parameters.
713
3.2
Underwater explosion test
The example of primary shock wave overpressure recorded as well as time of first
bubble collapse for Alan 3 emulsions explosives were shown in the Fig 7 and Fig 8.
In Table 4 the results of calculations of energy of shock wave and the gas bubble for
investigated explosives were shown. The loss factor was establish accordingly to
[Bjarnholt G., 1976] [6] for measured detonation velocity.
12
10
Overpressure, MPa
0
-0,2
0,2
0,4
0,6
0,8
-2
Time, ms
Fig 7.
Overpressure in time for shock wave in water - Alan 3 emulsion explosive

1,6
1,4
1,2
1
Volts
0,8
0,6
0,4
0,2
0
-0,05
0,05
0,1
0,15
0,2
0,25
0,3
-0,2
Seconds
Fig 8.
The record of first buble pulse for Alan 3 emulsion explosive
714
Table 4. The results of calculations of energy of shock wave and the gas bubble
for investigated explosives
Explosive
Density, ESW,
kg/m3 MJ/kg
Eb,
MJ/kg
A0,
MJ/kg
Alan1
1120
0,76
1,94
1,58
3,13
Alan3
1140
0,87
1,82
1,55
3,17
82GP
1109
0,75
1,63
1,616
2,84
10G5H
1480
0,895
1,765
1,9
3,47
The main output of energy produced in reactions of aluminium combustion occurs after
the as the C-J point. The liberated energy influences the process of expansion of detonation
products. Due to relatively large energy efficiency of aluminium oxidation the energy
liberated in aluminium combustion can bring a considerable contribution to the work
capacity of expanding gases. Therefore, value of useful explosion work significantly
increase for explosives with aluminium addition (Fig 9).
3,50
3,00
3,14
3,23
3,47
2,99
A0, MJ/kg
2,50
2,00
1,50
1,00
0,50
0,00
Alan1
Fig 9.
4.
Alan3
82GP
10G5H
Useful explosion work for investigated explosives
NUMERICAL ANALYSIS OF THE BLAST FIELD
A numerical evaluation of blast field parameters was also performed to compare the
measurement results with theoretical predictions. Blast wave propagation was evaluated
by the program package elaborated for numerical analysis of shock and detonation wave
propagation in reactive gaseous suspensions [Papliski 1999] [7]. Three-step numerical
algorithm is employed. A second-order Godunov type scheme is implemented, spatial
interpolation is used as a higher-order extension to Godunovs method. Air properties were
described with the real-gas effects including equation of state [Papliski 1993] [8]. Explosion
product were described with the JWL equation of state.
715
Assuming exact description of the detonation products and air properties an attempt was
made to estimate the explosion energy upon the registered blast wave parameters. In Fig. 10
a pressure distribution for the blast waves reaching the distance of 2m and 3m are presented.
The charge of 450 gram was considered. Two JWL parameterizations were taken from
Hornberg [Hornberg 1986] [9], as adjusted for Aquanal (Slurry +Al) and ANFO (6 % of oil
content).
The explosion energy (Qv) was so chosen to match the measured pressure profiles.
In the case of ALAN1 an registered overpressure peak at the distance of 2 m was
of p = 74,30 kPa (Table 3). In evaluations, an agreement was attained for the explosion
energy of 3,47 MJ/kg (in the case of Aquanal JWL parameterization). However, as can be
seen from the Fig. 10, with growing distance an slower decrease of evaluated pressure peak
was observed, in comparison with registered values. An measured peak overpressure for
ALAN1 at the distance of 3 m was of 30,46 kPa (Table 3). In evaluations the values
of p = 35,87 kPa (when ANFO JWL is applied ) and p = 36,55 kPa for Aquanal JWL
were obtained.
p [kPa]
80,00
60,00
40,00
20,00
0,00
-20,00
500
1000
1500
2000
2500
3000
r [mm]
Fig 10. Numerically evaluated blast field for an explosive charge in air;
mexp = 450 g
Aquanal JWL parameters
ANFO JWL parameters
In general, the energy assumed in evaluations is slightly lower than the quantity
of available energy, as obtained by multipling the mass charge by the explosion energy
(from Table 1). This can be explained by not full reaction of the explosive material. By an
explosion in air, in particular when we consider explosives detonating in an non-ideal
regime, not full energy is liberated, due to border effects. As a result not full explosion
energy is deposited into the shock wave. Then, to attain the same values of overpressure
lower explosion energy should be assumed in numerical evaluations. The other differences
observed in the registered and evaluated blast field are in intensity of blast wave decreasing.
These can be explained by assumed and real properties of detonation products. The JWL
equations of state are estimated from tests of metal tubes driving. Therefore at larger
716
distances when detonation products expand at relatively low mass densities the gas
properties may be not exactly re-mapped, by the assumed equation of state. This can explain
that once attained agreement between evaluated and registered magnitude of the pressure
wave e.g. at the distance of 2m is not further preserved (i.e. at 3 m), in an exact manner.
As is seen from the results obtained for ALAN type explosives a general agreement
between the value of ideal explosion energy and the numerically evaluated blast wave
intensity can be observed. However, further theoretical and experimental investigations are
necessary to correlate the energy liberated in explosion with the energy deposited to the
loading wave generated in material continuum surrounding the explosive charge. These
problems will be the subject of further investigations of authors.
5.
CONCLUSIONS
In the paper analysis of blast waves parameters as well as underwater explosion test
results of mining explosives are performed. Emulsion explosives of water-in-oil type and
dynamites were investigated. Values of maximal overpressure magnitude and specific blast
wave impulse generated by explosion of investigated explosive and shock waves and bubble
energy are considered as characteristics of energetical performance of explosive. Both
experimental measurements and numerical evaluation of blast wave field is performed.
Explosive characteristics of heterogeneous explosive mixtures, like emulsion explosives
are considered. An attention is paid to the influence of additional energy output, which can
proceed after the detonation zone. In this case, the detonation velocity cannot be the measure
of the global explosion energy as it is strictly coupled with the energy liberated before the
C-J point, through the detonative adiabate. The additional energy that is produced during
expansion of detonation products will enhance intensity of loading wave generated in
material surrounding the explosive charge. Therefore, by measuring the blast wave intensity,
a more integral characteristic of energy output occurring in explosion can be obtained.
In the paper, experimental measurements as well as numerical evaluation of blast field
produced by explosive charges is performed. As can be seen from obtained results, in the
case of aluminized explosives when the detonation velocity is lowered due to inertial effects
of aluminum particles in the first stage of the detonation process, the value of blast
overpressure can give a more realistic estimation of the energy performance.
The proposed method can be used as the optional characteristic of the explosive
performance. However, the more comprehensive methodic should be elaborated, to
introduce this approach as the criterion for estimation energetic effectiveness of explosives.
717
REFERENCES
[1] ORNELLAS D. L.: Calorimetric determinations of the heat and products of detonation for
explosives: October 1961 to April 1982. Report UCRL 52821, Lawrence Livermore National
Laboratory.
[2] PERSSON P-A., HOLMBERG R., Lee J.: Rock Blasting and explosives engineering, Ch. 4,
Boca Raton, Florida, 1994.
[3] PASZULA J., MARANDA A., PAPLISKI A., GOBEK B., KASPERSKI J., A comparative
analysis of blast waves parameters generated by explosion of emulsion explosives and
dynamites, Proc. VI Seminar New Trends in Research of Energetic Materials, Pardubice, April
22 - 24, 2003, pp. 266-273.
[4] PASZULA J., MARANDA A., PAPLISKI A., GOBEK B., KASPERSKI J., Investigation
of Blast Performance of Emulsion Explosives, Proc. 2003 International Autumn Seminar on
Propellants, Explosives and Pyrotechnics, Guilin, China, October 15-18, 2003, Theory and
Practice of Energetic Materials (vol. V), Science Press, Beijing/New York, 2003, pp. 442-448.
[5] PASZULA J., MARANDA A., GOBEK B., KASPERSKI J., An underwater test for LWC
emulsion explosives, 7th International Seminar - New trends in research of energetic materials,
Pardubice, the Czech Republic, April 20 - 22, 2004, pp. 589-598.
[6] BJARNHOLT G., HOLMBERG R., Explosive expansion works in underwater detonations,
Proceedings Sixth Symposium (International) on Detonation, August 24 27, 1976, Coronado,
California.
[7] PAPLISKI A., Evaluation of Nonstationary Detonation Initiation in Combustible Gaseous
Suspensions, 16-th International Symposium on Combustion Processes, Kazimierz, 1 5.08.1999, pp. 63 66.
[8] PAPLISKI A., Evaluation of chemical composition and state parameters in complex chemical
systems (in Polish), Bull. Milit. Univ. Techn., 42, Nr 2, 1993, pp. 41- 54.
[9] HORNBERG H., Determination of fume state parameters from expansion measurements of
metal tubes, Prop. Expl. Pyrot., N0 1, 1986, pp. 23 31.
718
SOPHISTICATED ELECTROCATALYSTS
FOR ECONOMICAL PRODUCTION OF HYDROGEN
PART A: STRUCTURAL CHARACTERIZATION
P. Paunovic*, O. Popovski**, M. Tasev R. Smileski**, and S. Hadzi Jordanov*
* Faculty of Technology and Metallurgy, University Sts. Cyril and Methodius,
Skopje, R. Macedonia
Abstract:
An original procedure was applied in order to prepare composite electrocatalysts for
hydrogen evolution reaction (HER), containing Ni, Co and CoNi as a hyper delectronic phase and TiO2 as a hypo-phase deposited on conductive carbon substrate
Vulcan XC-72. The catalysts support (TiO2 and carbon) was thermally treated on 250
and 480oC in the atmosphere of H2 and N2. XRD and SEM methods were employed in
order to elucidate the influence of temperature on the structural characteristics of
catalysts.
XRD analysis has shown that all catalysts contain hyper d- metallic phase of crystalline
structure, the size of grains being 1020 nm, except the Co-phase that showed to be
amorphous (grain size < 2 nm). TiO2 calcinated on 480oC has shown typical anatase
structure, while no crystalline state was observed for that calcinated on 250oC.
For all types of catalysts, an aggregation of grains in clusters of a size of typically 100
nm was observed by SEM analysis. Also, a presence of empty space, that improves
materials porosity and specific surface area respectively, has been shown. Further
improvements of the non-platinum hypo-hyper d- catalysts are in course.
Keywords:
1.
hypo-hyper d- catalyst, hydrogen evolution reaction (HER), Co, CoNi,

Ni, anatase
INTRODUCTION
The electrochemical technologies for hydrogen production (water electrolysis, chloralkaly industry) and fuel cells become more and more significant in the field of alternative
sources of energy. The use of electrochemically generated hydrogen, either in fuel cells or
for heat generation, is characterized by advantages of both high content of energy and
avoidance of environmental pollution.
Electrode materials for hydrogen evolution have to satisfy some criterions from different
aspects. By the technical point of view, the electrode material has to be stable (no corrosion,
no passivation etc.) and to catalyses the hydrogen evolution reaction (high current density at
lower overpotential). Economical reasons favor cheaper material, while environmental
concern is in favor of usage of nonpolluting materials. So, the choice of electrode material is
very difficult due to the conflict of technical and economical issues. Platinum was found to
be electrode on top of which hydrogen evolves with minimum overpotential [1]. But Pt and
other precious metals dont satisfy economical reason, because they are expensive, while
719
cheaper substituents using non-precious metals as Ni, Co etc. suffer of corrosion, passivation
and similar problems.
There are two basic approaches how to find more active, stable and cheaper electrode
material: i) developing of real surface area of the electrode (using of porous electrodes) and
ii) developing of multicomponent catalysts.
Based on Brewers theory of bonding in metals and intermetallic phases [2], Jak{i} and
his group successfully alloyed non-precious dual combination of Mo and Co and produced
materials with electrocatalytic activity even superior to that of individual metals of platinum
group. Jak{i} has extended Brewer valence-bond theory giving an explanation of alloying
effect on the rate of hydrogen evolution on electrodes based on transition metals [3].
According to his consideration, the mutual combination of the metals on the right side of
transition series, which are good individual electrocatalysts (hyper d- electronic phase) and
the metals of the left side, which are poor electrocatalysts (hypo d- electronic phase), shows
pronounced synergetic effect. As a result, composite hypo-hyper d- electrocatalysts become
subject of scientific and technical interest, so a new era of electrocatalysis was opened.
Hypo and hyper d- electronic phases can be in elemental or in valence state, so that
many combinations are possible, for example, metal-metal, metal-metal sulfide, metal-metal
oxide etc [4].
The subject of study in this paper is preparation and structural characterization of
nanostructured composite catalysts with hyper d- metallic phase (Ni, Co, CoNi) and hypo doxide phase (TiO2).
2.
EXPERIMENTAL
A modified sol-gel procedure [5] was applied to prepare composite hypo-hyper delectrocatalysts for hydrogen evolution. Firstly, the hypo d- electronic oxide phase was
deposited on carbon substrate. Ti-isopropoxide (Aldrich, 97%) was added into dispersed
Vulcan XC-72 (Cabot Corp. Boston Mass.) in anhydrous ethanol (Merck, p.a.). To provide
hydrolization of Ti-isopropoxide to Ti(OH)4, small amount of 1M HNO3 (Merck, p.a.) in the
ratio 10:1 was added. This mixture was evaporated at 60oC with intensive stirring until fine
nano-structured powder of catalyst support was obtained. Further, the metallic hyper d-phase
was grafted on the catalyst support. An individual or mixed Me-2,4-pentaedionate (Alfa
Aesar, Johnson Matthey, GmbH) was dissolved in absolute ethanol or acetone. This solution
was added into dispersed catalyst support in anhydrous ethanol. Evaporation was carried out
under the same conditions as above. The obtained catalyst powder consists of Ti(OH)4 with
respective amount of 18% TiO2, 10% Me or composite metallic phase and the rest is carbon.
To decompose Ti(OH)4 to TiO2 and the rest amount of oraganometallics, the powder was
calcinated at 250oC in the atmosphere of H2 and N2. Three different types of catalysts were
prepared, e.g.: 18% TiO2 + 10% Ni + Carbon, 18% TiO2 + 10% CoNi + Carbon and 18%
TiO2 + 10% Co + Carbon.
To obtain anatase structure of TiO2, the catalyst support was thermally treated at 480oC
and after that dispersed into ethanol in order to graft the hyper d-metallic phase. Finally the
catalyst was calcinated at 250oC. The composition of these catalysts was the same as the
previous ones. The only difference was the structure of TiO2.
For visual observation of the samples structure, Scanning Electronic Microscope (SEM)
was employed (JEOL, model JEM 200 CX).
720
X-ray diffraction method was used to determine the structural characteristics of the
obtained catalysts. XRD measurements were carried out on XRD diffractometer Philips APD
15, CuK radiation.
The BET-method was used to determine and compare the specific surface area (ssa) of
the catalysts. The measurements were performed by AREA-METER STROHLEIN
instrumentation.
3.
Microphotographies from the SEM observation are shown in the Fig 1. As one can be
seen, the catalysts particles deposited on Vulcan are of a spherical shape. There is a
grouping of the particles in clusters sized of 100-200 nm, whereat good adherence between
particles is achieved. Presence of a numbers of holes between the aggregates is evident, thus
contributing to the higher specific surface area, and consequently to a higher electrocatalytic
activity.
a)
b)
Fig 1.
SEM photografies of
a) 10%Ni + 18%TiO2 + Vulcan
XC-72,
b) 10%Co + 18%TiO2 + Vulcan
XC-72
c) 10%CoNi + 18%TiO2 +
Vulcan XC-72 catalyst
(Magnification x10000)
c)
Further information about the structure of the catalysts was obtained by XRD analysis.
XRD spectra for the catalysts thermally treated on 250oC are shown in the Fig. 2. The first
XRD pattern (Fig. 2a) corresponds to pure Ni of crystalline f.c.c structure with cell
parameter a=3,524 and particles size of 10-20 nm. XRD- spectra of Co based catalyst
shows very wide range of peaks, corresponding to particles of size less than 2 nm (Fig. 2b).
Co is almost amorphous and phase identification is impossible. For the third catalyst (Fig.
2b) the value of cell parameter a = 3,5318 is between the values of pure Ni (a = 3,524 )
721
and pure Co (a = 3,544 ). This implies a presence of a solid-state solution of Co and Ni.
Dominant peaks correspond to Ni cristalline cell, what means that Co atoms are incorporated
into Ni crystals. In this case, Co impurity atoms substitute Ni host atoms in local crystalline
environment of Ni trough structural phase transition of Co from amorphous to crystalline
solid-state solution. The crystal size of hyper-metallic d component is the same as that of Ni
based catalyst (10-20 nm). The crystalline state of the hypo-oxide phase was not detected,
what means that TiO2 is amorphous. This nano-strucured state of the catalysts contribute to
increase the surface area and hence, development of the electrocatalytic characteristics.
Ni
111
6000
18% TiO 2 + 10% Ni + Carbon
5000
Intensity, cps
4000
Ni
Ni
1000
a)
220
Al-holder
2000
Al-holder
Carbon
200
3000
0
0
20
40
60
80
100
2, degrees
111
Al-holder
Al-holder
1000
220
Al-holder
1500
200
Carbon
Intensity, cps
2000
Al-holder
18% TiO2 + 10% CoNi + Carbon
2500
500
b)
0
0
20
40
60
80
100
Carbon
2, degrees
900
18%TiO2 + 10%Co + Vulcan
800
700
600
Carbon
500
400
300
c)
200
100
0
20
40
60
80
100
2, grad
Fig 2.
XRD spectra of catalysts thermally treated on 250oC a) 10%Ni + 18%TiO2 +

Vulcan XC-72, b) 10%Co + 18%TiO2 + Vulcan XC-72 and c) 10%CoNi +
18%TiO2 + Vulcan XC-72 catalyst
722
XRD spectra for the catalysts thermally treated at 480oC are shown in the Fig. 3. All
metallic phases show the same crystal characteristic as in previous series, i.e. crystalline Ni
(Fig.3a), amorphous Co (Fig.3b) and solid-state solution of Co in Ni crystalline lattice
(Fig.3c). At 480oC, TiO2 transforms from amorphous to crystalline form as anatase.
3500
Holder
18% TiO2 + 10% Ni + Vulcan XC-72

A - TiO2 (anatase)
3000
Intensisty, cps
2500
Ni
111
2000
A
1500
Ni
200
AA
1000
a)
Ni
200
500
0
0
20
40
60
80
100
2, degrees
3500
18% TiO2 + 10% Co + Vulcan XC-72

A - TiO2 (anatase)
Holder
3000
Intensity, cps
2500
2000
1500
1000
500
b)
0
0
20
40
60
80
100
2, degrees
Holder
18% TiO2 + 10% CoNi + Vulcan XC-72

A - TiO2 (anatase)
3000
Intensity, cps
2500
2000
Ni(Co)
111
1500
Ni(Co)
A200A
A
1000
500
Ni(Co)
200 A
A
c)
20
40
60
80
100
2, degrees
Fig 3.
XRD spectra of catalysts thermally treated on 480oC a) 10%Ni + 18%TiO2 +

Vulcan XC-72, b) 10%Co + 18%TiO2 + Vulcan XC-72 and c) 10%CoNi +
18%TiO2 + Vulcan XC-72 catalyst
723
BET analysis has shown that the Ni based catalyst deposited on Vulcan with 18%TiO2
has the highest specific surface area of 108 m2/g, while the CoNi based catalyst has 96 m2/g
and the Co based has 70 m2/g. These high-developed surface areas of the nano-structured
catalysts are in favor of achieving high electrocatalytic activities.
4.
CONCLUSION
As main conclusions of the investigations we can mention that:
1) The applied simplified sol-gel procedure provides synthesis of composite
hypo oxide-hyper metal nano-structured d-electrocatalysts (below 2 nm for
Co based catalyst and 10-20 nm for Ni and CoNi based catalysts),
2) Metallic phases have the same structure for both temperatures crystalline
Ni, amorphous Co and solid state solution of CoNi. TiO2 prepared on 250oC
is amorphous, while at 480oC it transforms to crystalline anatase,
3) Structural characteristic of these catalysts are basis for improved
electrocatalytic activities for both hydrogen evolution in electrolytic
processes and hydrogen oxidation in fuel cells.
Acknowledgement
This paper has been supported by and investigations carried out within the EU Project
PROMETHEAS PL ICA2-2001-10037.
The authors are grateful and indebted to Professor Milan Jak{i}, University of Belgrade,
for his bountiful and extensive help in the literature supply and valuable advices for both
theoretical consideration and practical aspects of catalysts preparation.
Great thank to the staff of Institute of Electrochemistry and Energy Systems,
Bulgarian Academy of Science, Sofia for the fruitful collaboration and the hospitality in
their labs during structural analysis and electrode preparation. Special thank to
Academician Evegeny Budevski, Dr. Evelina Slavcheva and Dr. Tamara Petkova.
REFERENCES
[1]
[2]
[3]
[4]
[5]
H. KITA: Electrocatalysis by d and sp Metals, (Electrochemistry: The Past Thirty and the Next
Thirty Years, H. BLOOM and F. GOOTMAN, Eds.), p. 117-137, Plenum Press, New York
(1977)
L. BREWER: Thermodynamic Stability and Bond Character in Relation to Electronic
Structure and Crystal Structure (Electronic Structure and Alloy Chemistry of Transition
Elemenets, P.A. BECK eds.), p. 221-235, Intercsience, New York (1963), Dover New York
(1965),
M. M. JAKSIC: Advances in electrocatalysis for Hydroge Evolution in the Light of the BrewerEngel Valence-Bond Theory, Int. J. Hydr. Energy, 12, pp. 727-752 (1987)
S. HADZI JORDANOV, P. PAUNOVIC, O. POPOVSKI, A. T. DIMITROV
and D. SLAVKOV: Electrocatalysts in the Last 30 Years - From Precious Metals to Cheaper
but Sophisticated Complex Systems, Bull. Chem. Technol. Macedonia, 23, No2, p. 101-112
(2004)
C. J. BRINKER, G. W. SCHERER: Sol-Gel Science, The Physic and Chemistry of Sol-Gel
Processing, Academic Press, New York, 1990
724
ESTIMATION OF THE CONDENSATION PROBABILITY

OF SUPERSATURATED NITROGLYCERINE VAPOR
DURING MANUFACTURING OF ENERGETIC MATERIALS
Dmitriy V. Pleshakov
Mendeleyev University, Miusskaya pl. 9, Moscow, 125047 Russia
Abstract:
Thermodynamic conditions for the condensation of supersaturated nitroglycerine vapor
on the cold surfaces of apparatus during energetic materials manufacturing are
analyzed. The tendencies in condensation center generation upon cooling of onecomponent (nitroglycerine) and two-component (nitroglycerine + coplasticizer) vapor
are considered. The use of a two-component plasticizer reduces the probability of liquid
plasticizer drops appearing on the apparatus surface.
Keywords:
1.
nitroglycerine, condensation of vapor, manufacturing of energy materials
INTRODUCTION
Nitroglycerine (NG) is used in manufacturing of various energetic materials (explosives,

gun powders, propellants). NG is an explosive with high shock and friction sensitivities.
Therefore, NG segregation from energetic materials is undesirable. NG drops can appear
given that an energetic material has a poor thermodynamic stability [1-3] or given that
nitroether vapor condenses on the cold surfaces of apparatus [4]. Here, we will consider how
one can reduce the probability of supesaturated NG vapor condensing during manufacturing
of energetic materials. A two-component plasticizer (NG + coplasticizer) was used. We will
consider tendencies in condensation center generation for a two-component plasticizer vapor
and compare the probability of condensation center appearance for one-component and twocomponent supersaturated vapors.
2.
RESULTS AND DISCUTION
We will consider two hypothetical energetic materials. In one of them, the binder
contains a one-component plasticizer A; in the other, a two-component plasticizer A + B.
Under a thermodynamic equilibrium at temperature Th , the partial vapor pressure of the
composition containing one-component plasticizer A is described by
p Ah = p A0 (Th )a Ah
(1)
In the case of two-component plasticizer A + B,

p Ah = p 0A (Th )a Ah ,
(2)
pBh = pB0 (Th )aBh ,
(3)
730
P = p Ah + pBh ,
x Ah =
(4)
p Ah
,
P
(5)
Here p A0 (Th ) and pB0 (Th ) are saturated vapor pressures of plasticizers A and B at
temperature Th ; p Ah and pBh are the partial pressures of plasticizers A and B at temperature
Th ; a Ah and aBh are the activities of plasticizers A and B in the binder; P is two-component
vapor pressure; and x Ah is the A mole fraction in two-component vapor.
If the apparatus surface temperature Tl is lower than the vapor temperature Th , the vapor
is cooled and becomes supersaturated. It can, therefore, condense. Condensation center
formation is a fluctuation process [5]. If the condensation center resulting from density
fluctuations in the supersaturated vapor has a radius smaller than the critical condensation
radius, this condensation center spontaneously dissociates as a result of vaporization. If the
fluctuation radius is larger than the critical condensation radius, this condensation center
spontaneously grows up. For one-component vapor,
G =
4Rc3
GA + 4Rc2 A ,
3VA
(6)
p A0 (Tl )
GA = RTl ln
,
p Ah
(7)
2 AVA
,
GA
(8)
4
GcrA = ( RcrA ) 2 A ,
3
(9)
RcrA =
Here, G is the Gibbs energy of formation for a condensation center with radius Rc ; VA
and A are the molar volume and surface tension energy of plasticizer A; p 0A (Tl ) is the
saturated vapor pressure of plasticizer A at temperature Tl ; GA is the Gibbs energy change
in the condensation of one mole of plasticizer A supersaturated vapor; RcrA is the critical
condensation center radius for plasticizer A; GcrA is the Gibbs energy change corresponding
to the formation of a plasticizer A condensation center with the critical radius; and R is the
universal gas constant.
For the two-component plasticizer, a fluctuation in the supersaturated vapor density can
be accompanied by a fluctuation in composition; that is, the condensate can differ in its
composition from the vapor. In this case,
G =
4 Rc3
(GM + GF ) + 4 Rc3 M ,
3VM
l
pB0 (Tl )aBl
p A0 (Tl )a lA
l
GM = RTl x A ln
+ (1 x A ) ln
h
h
(
1
)
P
x
Px
A
A
731
(10)
(11)
l x Ah
(1 x lA )
l
( x lA < x Ah ),
GF = RTl x A ln l + (1 x A ) ln
h
xA
(1 x A )
(12)
l x lA
(1 x Ah )
l
( x lA > x Ah ),
GF = RTl x A ln h + (1 x A ) ln
l
xA
(1 x A )
(13)
2 M VM
,
(GM + GF )
(14)
4
GcrM = ( RcrM ) 2 M ,
3
(15)
RcrM =
where VM and M are the molar volume and surface tension energy for the twocomponent condensate; p A0 (Tl ) and pB0 (Tl ) are the saturated vapor pressures of plasticizers
A and B at temperature Tl ; x lA is the plasticizer A mole fraction in the two-component
condensate; a lA and aBl are the activities of components A and B in the two-component
condensate; GM is the Gibbs energy change in the condensation of one mole of the twocomponent supersaturated vapor; GF is the Gibbs energy change in compositional
fluctuation for one mole of the two-component supersaturated vapor; RcrM is the critical
radius of the two-component vapor condensation center; and GcrM is the Gibbs energy
change corresponding to the formation of a two-component vapor condensation center with
the critical radius.
From Eqs. (11) - (13), if the temperature and composition of the supersaturated vapor do
not change, GM and GF are functions of one variable x lA . Note that the smaller the
GM + GF sum (i.e., since GM + GF 0, the larger the absolute value of this sum),
the smaller the RcrM and GcrM , and therefore, the higher the formation probability of a
condensation center of a given radius. The most probable composition of the condensate can
be determined by solving
(GM + GF )
= 0,
x lA
(16)
under the condition that

2 (GM + GF )
x lA
> 0,
(17)
Expressions (16) and (17) correspond to the GM + GF minimum. With allowance for
expressions (11) - (13), the solution to Eq. (16) reads as
xlA =
x Ah
e
pB0 (Tl )
( x lA > x Ah ),
0
p A (Tl )
(18)
732
xlA = 1
(1 x Ah )
e
p A0 (Tl )
(if x lA < x Ah )
0
pB (Tl )
(19)
Note that Eq. (18) has a meaning when

pB0 (Tl )
>e
p A0 (Tl )
(20)
Eq. (19) has a meaning when

p A0 (Tl )
>e
pB0 (Tl )
(21)
If for pB0 (Tl ) > p A0 (Tl ) ,

pB0 (Tl )
e,
p A0 (Tl )
(22)
and if for p 0A (Tl ) > pB0 (Tl ) ,

p A0 (Tl )
e,
pB0 (Tl )
(23)
there are no compositional fluctuations in two-component supersaturated vapor. The

condensate composition corresponds to the vapor composition ( xlA = x Ah ), and GF = 0.
As known [5],
GcrA
LA ~ exp
,
RT
(24)
GcrM
LM ~ exp
,
RT
(25)
where LA and LM are condensation center formation probabilities for one- and twocomponent vapor, respectively. Thus, at a fixed temperature, GcrA and GcrM control LA and
LM . Therefore, as a comparison criterion for the condensation center formation probabilities
for one- and two-component supersaturated vapors, we can use
GcrA
GcrM
(26)
The results of the calculations are displayed in Figs. 1 - 4 and Tables 1 - 4. In the
calculations, we assumed that the major plasticizer A is nitroglycerine and that component B
is a hypothetical plasticizer that is mixed with NG to give ideal solutions. The
physicochemical properties of the coplasticizer (the enthalpy of vaporization, surface tension
energy, molar volume), except for the saturated vapor pressure, are close to the properties of
NG. We set that the saturated vapor pressures of NG and component B are related through
733
Fig 1.
The plots vs. z . z < 1. T = Th Tl = 5 (1), 10 (2) and 15 C (3).

Tl = 25 C. A = 0.68, B =0.12. = 0.1.
Fig 2.
The plots vs. z . z > 1. T = Th Tl = 5 (1), 10 (2) and 15 C (3).

Tl = 25 C. A = 0.68, B =0.12. = 0.1.
734
Fig 3.
The plots vs. z . z < 1. T = Th Tl = 10 C. Tl = 25 C. = 0.1.

A = 0.68, B =0.12 (1); A = 0.52, B =0.28 (2), A = 0.4, B =0.4 (3).
Fig 4.
The plots vs. z . z > 1. T = Th Tl = 10 C. Tl = 25 C. = 0.1.

A = 0.76, B =0.04 (1); A = 0.68, B =0.12 (2) , A = 0.6, B =0.2 (3).
735
Table 1.
Dependence 0 on T = Th Tl . Tl = 25 C. A = 0.68, B =0.12.

= 0.1.
T
0.472
10
0.708
15
0.807
Table 2. Dependence z0 on T = Th Tl . Tl = 25 C. A = 0.68, B =0.12.

= 0.1.
z0
8.2 0.2
10
25.5 0.5
15
198 2
Table 3. Dependence 0 on B . T = Th Tl = 10 C. Tl = 25 C. = 0.1.
0.12
0.708
0.28
0.335
0.40
0.104
Table 4. Dependence z0 on B . T = Th Tl = 10 C. Tl = 25 C. = 0.1.
z0
0.04
34 1
0.12
25.5 0.5
0.20
47 1
0.28
2335 10
736
z=
pB0
0
pNG
(27)
The NG saturated vapor pressure versus temperature is expressed through [1,4]

0
= 29.62
ln pNG
9518
T
(28)
With allowance for Eqs. (27) and (28), the coplasticizer saturated vapor pressure
depends on temperature as follows:
ln pB0 = 29.62
9518
+ ln z
T
(29)
In the calculations z was varied from 0 to 2345.

To calculate the activity coefficients of plasticizer A and two-component plasticizer
A + B, we used the lattice solutions model [6-8]. For one-component plasticizer A,
ln a Ah = ln A + (1 A ) + (1 A ) 2 ,
(30)
A + P = 1,
for two-component plasticizer A + B,
ln a Ah = ln A + (1 A B ) + (1 A )(1 A B ) ,
(31)
ln aBh = ln B + (1 A B ) + (1 B )(1 A B ) ,
(32)
A + B + P = 1
Here, A and B are the volume fractions of the plasticizers in binders; P is the
volume fractions of the polymer in binders; and is the plasticizer - polymer intermolecular
interaction parameter in the binder of the energetic material. The plasticizer activities in the
two-component condensate were set equal to their mole fractions.
In Figs 1 and 2 and in Tables 1 and 2, we present the results obtained for a fixed
composition of the two-component plasticizer and various temperature differences T
between the "hot" energy material and the "cold" apparatus surface ( T = Th Tl ). The
= f (z ) plots for z < 1 are shown in Fig. 1. When z 0, decreases. Table 1 displays
the 0 calculated in accordance with the constraint
0 = lim
(3
z)
1
424
(33)
z 0
For all T , 0 > 0 and 0 increases with increasing T .

When z > 1 (Fig. 2), = f ( z ) is a decreasing function. In this case, however,
decreases to zero. Denote the abscissas at which = 0 as z0 . The z0 corresponding to
various T are listed in Table 2. One can see that z0 increases as T increases. Note that,
when = 0 , GcrM = and LM = 0 (see Eq. (25)); that is, the condensation center formation
737
probability for two-component supersaturated vapor is zero, and condensate drops do not
appear.
In Figs. 3 and 4 and in Tables 3 and 4, we present the results of the calculations for a
fixed temperature difference T and a variable composition of the two-component
plasticizer. When z < 1, an increase in the coplasticizer B concentration decreases the
condensation center formation probability for two-component supersaturated vapor (Fig. 3,
Table 3). When z > 1, the condensation probability is an extreme function of B (Fig. 4,
Table 4). The optimal component B concentration, to which a minimal z0 corresponds, is
B = 0.08 - 0.12. This concentration range corresponds to mixed plasticizers containing 10 15% component B.
3.
CONCLUSIONS
We have analyzed thermodynamic conditions for condensation of NG supersaturated

vapor during energetic materials manufacturing. We have shown that the use of a twocomponent plasticizer (NG + coplasticizer) reduces the probability of liquid nitroether drops
appearing on the apparatus surfaces. These results allow us to conclude that compounds
more volatile than NG are the best coplasticizers. With these compounds, one can fully avoid
NG condensation on the apparatus surfaces and decrease hazard.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
YU. M. LOTMENTSEV AND D. V. PLESHAKOV: Phase State of Nitrocellulose Plasticized with

Trinitroglycerin, Propellant, Explosives, Pyrotechnics, vol. 22, p. 203-206, 1997
D. V. PLESHAKOV, YU. M. LOTMENTSEV, SHAO ZIQIANG, N. N. KONDAKOVA AND A. V.
LUKASHEV: Solvation Effects and Thermodynamics of Plasticization of Cellulose Nitrate,
Polymer Science, A, vol. 41, no. 3, p. 364-370, 1999
D. V. PLESHAKOV, YU. M. LOTMENTSEV, N. N. KONDAKOVA, T. A. BESTUZHEVA,
E. A. BUTENKO, I. N. KOLESNIKOV AND E. V. ZHURAVLEV: Thermodynamic Stability
of Polyurethane Binders, Proc. of International Autumn Seminar on Propellants, Explosives
and Pyrotechnics, Guilin, China, p. 156-160, 2003
D. V. PLESHAKOV AND YU. M. LOTMENTSEV: Prediction of Thermodynamic Conditions for
Nitroester Vapor Condensation on the Surfaces of Process Apparatus during the Production of
Energetic Materials, Proc. of the 7th Seminar New Trends in Research of Energetic
Materials, Pardubice, Czech Republic, p. 591-599, 2004
YU.G. FROLOV: Kurs koloidnoy himii (Colloid Chemistry), Moscow: Khimiya, 1982
P.J. FLORY: Principles of Polymer Chemistry, Ithaca: Cornell Univ. Press, 1953
N.A. SMIRNOVA: Molekulyarnyie teorii rastvorov (Molecular Theories of Solutions),
Leningrad: Khimiya, 1987
YU.M. LOTMENTSEV AND D.V. PLESHAKOV: Thermodynamics of Phase Separation in
Plasticized Polymer Networks, Polymer Science, A, vol. 35, no. 7, p. 824-829, 1993
738
SOPHISTICATED ELECTROCATALYSTS
FOR ECONOMICAL PRODUCTION OF HYDROGEN
PART B: ELECTROCHEMICAL CHARACTERIZATION
O. Popovski**, P. Paunovic, R. Smileski* and S. Hadzi Jordanov*
* Faculty of Technology and Metallurgy,
University Sts. Cyril and Methodius, Skopje, R. Macedonia
Abstract
Electrochemical characterization of composite hypo-hyper d- electrocatalysts for
hydrogen evolution reaction (HER) was performed using cyclic voltammetry and steadystate galvanostatic method.
Cyclic voltammograms are without any pronounced peaks corresponding to some
surface processes, typically for porous electrodes.
The influence of the temperature on electrocatalytic activity was studied. The catalysts
with support treated on 480oC are more active than those on 250oC. The overpotential
difference ( 250 C 480 C ) for hydrogen evolution at 60 mAcm-2 is 15 mV for Co based
electrocatalyst and even 60 mV for Ni based. Comparing the electrocatalytic activity of
the catalysts with different hyper d-metallic phase, one can conclude that the activity
decreases in the sequence: Co > CoNi > Ni. Evidence is produced that the investigated
non-precious catalysts show significant synergetic effect. These results are valuable per
se and also as indicators for further improvement of non-platinum hypo-hyper dcatalysts.
o
Key words:
1.
hypo-hyper d- catalyst, hydrogen evolution reaction (HER), Co, CoNi,

Ni and porous electrodes
INTRODUCTION
The search for new, sophisticated and less expensive materials for electrocatalysis with
pronounced synergetic characteristics, (i.e. high current density at overpotential near 0) is
one of the most actual field of contemporary electrochemistry. Significant synergetic effect
can be achieved by combining the metals of the opposite sides of transition series, i.e. metals
with empty or partialy filled d- orbitals, thus being bad catalysts and metals with internally
paired d- electrons, thus being good catalysts for hydrogen evolution reaction (HER) [1].
These materials are named hypo-hyper d- electrocatalysts, some of them even exceed the
catalytic activity of individual precious metals.
There is an analogy between the hypo-hyper d- electronic compound stability and
electrocatalytic activity, in which Laves phases of high symmetry and extra stability have the
highest activity for hydrogen evolution. This is confirmed for non-precious hypo-hyper dcombinations as MoFe3, WFe3 [1], LaCo5, LaNi5 [2] etc.
Depending on the applied preparation procedure, the crystallinity of the catalysts can
vary from amorphous up to high crystalline. Hypo-hyper d-components can be in metallic
(zero valence state) or in higher valence state as oxide, sulfide, phosphide etc. Depending on
739
the state of the hypo and hyper d- phases, these catalysts can be divided into 5 groups, as
shown in table 1 [3].
Table 1. Grouping of hypo-hyper d- intermetallic electrocatalysts
Type of hypo-hyper d- electrocatalyst
Examples
Intermetallic catalysts
Mo-Co, MoPt3, MoNi3, MoFe3,

LaCo5, Ti1-xNix (Tiney Ni), Ni-CoMo, Ni-Zr, Co-Zr, etc.
II
Sulfides, Phosphides
Ni-W-S, Ni-W-P, Fe-Ni-P etc.
III
Catalysts based on Raney Ni
Ni-Al-Ti, Ni-Al-Mo, Ni-Al-Cr,

Ni-Zn-Ti, Ni-Zn-Mo etc.
IV
Mixed Oxides
TiO2-RuO2, TiO2-IrO2, RuxTi1-xO2,

RuO2-IrO2 etc.
Hypo Oxide Hyper metal
Pt-TiO2, Pt-Ru-WO3, Ni-V2O5 etc.
The aim of this work was to obtain and electrochemically characterize non-platinum
nanostructured hypo-hyper d- electrocatalysts. The hypo component is oxide phase TiO2,
thermally treated on 250 and 480oC, thus resulting in transformation of the TiO2 from
amorphous into anatase structure. The hyper d-phase is in metallic state Co, Ni or CoNi.
2.
EXPERIMENTAL
Catalysts prepared by the procedure described elsewhere [4], where further transformed
into porous gas-diffusion electrodes. Pressing at 300 kgcm-2 on 300oC [5] was employed, so
that a two layer electrode consisting of acetylenes carbon black + PTFE as a substrate and a
top layer of catalyst + PTFE mixture (See Fig. 1).
Electrochemical cell used for testing is shown in Fig. 2. The counter electrode was a
spiral shaped platinum wire, the reference Hg/HgO, while the electrolyte was a solution of
3,5 M KOH (p.a., Merck) in deionizated ultrapurified water. During the electrochemical
testing, the solution in the cathodic area was purged with hydrogen.
The measurements were performed by AMEL electrochemical line (Function Generator
AMEL 568, Potentiostate/Galvanostate 2053 and software package SOFTASSIST 2.0).
740
Fig 1.
Cross section of three phase gasdiffusion porous electrode
Fig 2.
3.
Electrochemical
cell used for the
electrochemical
characterization
In Fig. 3, cyclic voltammograms of the investigated electrodes are shown. Generally, the
voltammograms of the catalysts containing TiO2 as anatase (thermally treated at 480oC),
show more intensive surface processes as well as electrode reactions. Hydrogen evolution
(A) occurs at lower potential on the catalyst thermally treated at 480oC for all investigated
systems. Also the oxygen evolution reaction is more intensive and occurs at lower potentials
for the same systems. This is due to achievement of stronger metal-support d-interaction,
when TiO2 is in crystalline form [6,7]. Ni system shows less intensive peaks for the surface
processes as Ni to Ni2+ transition (B) and corresponding reduction peak (D) (Fig. 3a). The
systems in Fig. 3b and c, show pronounced peaks of surface processes due to presence of Co.
The peak F represents transition of elemental Co to Co2+, peak B transition of Co2+ to Co3+
and oxygen evolution, while peak C oxygen evolution reaction. D and E are corresponding
peaks of oxygen reduction and metal oxide reduction.
741
a)
0.03
480 C
0.02
250 C
B
i, Acm
-2
0.01
0.00
D
-0.01
-0.02
-1.4
-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
E, V vs. Hg/HgO
b)
0.05
480 C
0.04
250 C
0.03
0.02
0.01
-2
i,
Ac
m
0.00
-0.01
-0.02
-0.03
-0.04
-0.05
-0.06
-1.4
-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
E, V vs. Hg/HgO
c)
0.05
o
480 C
0.04
250 C
0.03
i, Acm
-2
0.02
B
F
0.01
0.00
-0.01
-0.02
-0.03
-0.04
-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
E, V vs. Hg/HgO
Fig 3.
Cyclic voltamograms for the catalysts containing
The electrocatalytic activity of the investigated electrodes is shown in the Fig. 4. The
order of activity of the different hyper d-metallic systems decreases in the sequence: Co >
CoNi > Ni with corresponding values of overpotential at current density of 60 mAcm-2: -275
mV for Co based catalyst, -290 mV for CoNi based and -485 mV for Ni based. These values
are related for the catalysts containing non-crystalline TiO2 (thermally treated at 250oC). For
the second series of catalysts containing TiO2 as anatase, the order of activity is the same as
above with corresponding values of overpotential at 60 mAcm-2: -260 mV for Co based
742
catalyst, -290 mV for CoNi based and 425 mV for Ni based. Comparing the activity of the
catalysts with the same hyper d-metallic phase with different form of TiO2, it is obvious that
the catalysts containing anatase are more active than those containing non-crystalline TiO2,
except for CoNi based system where the activities are almost the same. For instance, for Co
based catalysts this difference at 60 mAcm-2 is 15 mV, for CoNi based 0 mV and even 60
mV for Ni based. As mentioned above, the stronger metal-support interaction is achieved
between crystalline TiO2 and the hyper d-metallic phase. Otherwise, hyper d-metallic
component gives the prevailing catalytic activity, while hypo d-electronic phase contributes
to the overall synergetic effect of such composite electrocatalyst by so called Strong Metal
Support Interaction (SMSI) [6,7]. The hypo-oxide phase is not only a support, but also active
and significant part of the catalyst, which provides the second, bifuntional role. Thus, TiO2 is
a component that improves the catalytic activity in this case.
Vulcan XC-72 + 18%TiO2 + 10%Me
0
-100
, mV
-200
o
Co(480 oC)
Co(250 C)o
CoNi (480 oC)
CoNi (250 C)
-300
-400
Ni (480 C)
-500
Ni (250 C)
-600
0
20
40
i, Acm
Fig 4.
4.
60
80
-2
Polarization curves of the investigated catalysts for hydrogen evolution in

alkaline solution
CONCLUSION
As main conclusion of the electrochemical investigations one can mention that:
1) Applying the modified sol-gel procedure we produced catalysts that exhibit pronounced
synergetic effect for the hydrogen evolution reaction. Among the 6 prepared and tested
hypo oxide-hyper metal d-catalysts, the most active were those with Co and CoNi
phases.
2) This effect is due to both high developed surface area as result of obtained
nanostructured components of the catalysts and the strong metal-support interaction
between hypo and hyper d- electronic phases.
3) Thermal treatment of the metal support (hypo oxide d-phase) at 480oC where anatase
crystalline TiO2 was obtained, contribute to stronger metal-support interaction as result
the higher electrocatalytic activity was achieved.
743
Acknowledgement
This paper has been supported by and investigations carried out within the EU Project
PROMETHEAS PL ICA2-2001-10037.
The authors are grateful and indebted to Professor Milan Jak{i}, University of
Belgrade, for his bountiful and extensive help in the literature supply and valuable advices
for both theoretical consideration and practical aspects of catalysts preparation.
Great thank to the staff of Institute of Electrochemistry and Energy Systems,
Bulgarian Academy of Science, Sofia for the fruitful collaboration and the hospitality in
their labs during structural analysis and electrode preparation.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
M. M. JAKSIC: Advances in Electrocatalysis for Hydrogen Evolution in the Light of the

Brewer-Engel Valence-Bond Theory, Int. J. Hydr. Energy, 12, p. 727-752 (1987)
T. KITAMURA, C. IWAKURA and H. TAMURA: Embrittlement of LaNi5- Type Alloy
Electrode During the Cathodic Evolution of Hydrogen, Electrochimica Acta, 27, p. 1723-1727
(1982)
S. HADZI JORDANOV, P. PAUNOVIC, O. POPOVSKI, A. DIMITROV and D. SLAVKOV:
Electrocatalysts in the last 30 years - from precious metals to cheaper but sophisticated
complex systems, Bull. Chem. Technol. Macedonia, 23, No2, p. 101-112, (2004)
P. PAUNOVIC, O. POPOVSKI, V. NIKOLOVA, P. ILIEV, M. TASEV, S. HADZI
JORDANOV: Ni-Me/TiO2 electrocatalysts for hydrogen evolution reaction (HER), XVIII-th
Congress on the Chemists and Technologists of Macedonia, Extended abstracts, [ECH-1],
Ohrid, september, (2004)
BG PATENT Appl. No 38581
S.G. NEOPHYTIDES, S. ZAFEIRATOS, G.D. PAPAKONSTANTINOU, J.M. JAKSIC, F.E.
PALOUKIS, M.M. JAKSIC: Extended Brewer hypohyper-d-interionic bonding theoryI.
Theoretical considerations and examples for its experimental confirmation, Int. J. Hydr.
Energy, in press.
S.G. NEOPHYTIDES, S. ZAFEIRATOS, G.D. PAPAKONSTANTINOU, J.M. JAKSIC, F.E.
PALOUKIS, M.M. JAKSIC: Extended Brewer hypohyper-d-interionic bonding theoryII.
Strong metal-support interaction grafting of composite electrocatalysts, Int. J. Hydr. Energy, in
press.
744
COMPARISON OF MEASURED
AND CALCULATED PARAMETERS OF PROPELLANTS
Pavel Prchal, Jan Zigmund, and Jindich Veverka
Research Institute of Industrial Chemistry, Explosia a.s.,
532 17 Pardubice-Semtn, Czech Republic
Abstract:
This paper deals with comparison of experimentally observed and calculated values of
heat of explosion of several propellant compositions. For calculations TCHAR

program was used.
Keywords:
1.
propellant, heat of explosion, calculation program
INTRODUCTION
There are several calculation programs for estimation of characteristics of propellants.

In our company the software called TCHAR is used.
Out of propellant characteristics we chose heat of explosion under constant volume as
a well comparable parameter in contrast to the other parameters, because of their problematic
formulation without errors, heat losses and other influences, which often depend on the
concrete measure equipment used.
Selected propellant formulations prepared in our institute were used to evaluate the
conformity of their calculated and measured heats of explosion.
2.
EXPERIMENTAL
2.1
Calculation program
TCHAR [1] program exists in several DOS versions (by Jan Zigmund - latest version
3.40) and in one version for Windows (by Ivo Bauer from the previous - version TCHAR
Expert 0.10.0.56). All versions enable selection of many components of the projected
propellant and their content in per cent by weight. The database of substances is
well-arranged and it is simple to add new components with knowledge of number of atoms
C, H, O, N, Al and Cl in molecule and enthalpy of formation of the newly added substance.
The program can calculate parameters of composition with assumption of constant volume,
constant pressure and for the case of burning with air with assumption of constant pressure.
The parameters calculated are: Average summary formula, oxygen balance, oxygen
coefficient, amount and composition of expected burning products, temperature of
explosion, heat of explosion, force constant, specific heats ratio, total quantity, specific
volume and molar weight of burning products, velocity of sound in burning products, rate of
burning and many other parameters. Older version has some other options, which are not
incorporated into the newer version - for example the option called triangle - the user can
choose 3 components in the formulation and then, in triangle display, to change their mutual
ratio under continuous displaying of main characteristics.
Inorganic components were substituted with an inert (nitrogen) with zero value of
enthalpy of formation.
745
2.2
Measurement of heat of explosion
The measurement of heat of explosion [2] is performed as explosion of sample in

calorimeter bomb with volume 300 cm3 placed in water bath. The changes of temperature of
water bath after explosion are recalculated to heat of explosion at 18-21 C and vacuum in
J.g-1.
2.3
The propellants used
Samples of single-base 7-perforated powders SA [3] were prepared according to

American M14 powder (88% nitrocellulose with 13.3% N, 1.5% diphenylamine, 2%
dibutylphthalate, 7% 2,4-dinitrotoluene) with 0-2.5% polyethylenglycol - see Table 1.
Table 1. SA samples
Sample
Volatile
components
PEG
K2SO4
QE,exp.
-1
QE,calc.
-1
Error
[%]
[%]
[J.g ]
[J.g ]
[%]
[%]
SA-160
1.21
0.93
3,433
3,437
0.12
SA-161
0.77
0.37
1.83
3,407
3,413
0.18
SA-162
0.87
2.54
1.78
3,467
3,211
-7.38
SA-074/1
0.68
0.77
1.49
3,379
3,391
0.36
SA-074/2
0.63
0.65
1.53
3,354
3,401
1.40
SA-074/3
0.66
0.46
1.52
3,358
3,419
1.82
SA-075/1
0.76
0.51
1.74
3,411
3,403
-0.23
SA-075/2
0.69
0.36
1.72
3,382
3,419
1.09
SA-075/3
0.60
0.43
1.72
3,380
3,417
1.09
Samples of double-base 7-perforated powders S105 were prepared (approximately 80%

nitrocellulose with 13.2% N, 11.5-14.5% nitroglycerine, 3% Centralite I, 0.85%
diphenylamine) - see Table 2.
Table 2. S105 samples
Sample
NG
[%]
Volatile
components
QE,exp.
QE,calc.
Error
[J.g-1]
[J.g-1]
[%]
[%]
S105-VU-26/04
11.42
1.25
4,005
3,992
-0.32
S105-VU-27/04
13.00
1.09
4,067
4,057
-0.25
S105-VU-28/04
14.15
1.10
4,118
4,108
-0.24
S105-VU-29/04
14.53
1.01
4,078
4,068
-0.25
746
Samples of double-base ball powders D100 were prepared in defined composition in

accordance with specifications given [4] (80% nitrocellulose with min. 13.00% N, 8-11%
nitroglycerine, 5-10% dibutylphthalate, min. 0.7% stabilizers (Centralite I, Acardite II,
diphenylamine), max. 0.5% graphite, max. 0.3% Na2SO4, max. 1.5% KNO3, max. 1% ethyl
acetate, max. 1.5% SnO2, 0.4-1.2% water) - see Table 3.
Table 3. D100 samples
Sample
Inorganic
components
Volatile
components
[%]
[%]
D100-SE-2/01
1.47
D100-SE-1/04
QE,exp.
-1
QE,calc.
-1
Error
[J.g ]
[J.g ]
[%]
0.83
3,502
3,352
-4.28
1.25
0.91
3,586
3,533
-1.48
D100-SM-1/04
0.77
0.93
3,548
3,541
-0.20
D100-SM-4/04
1.12
0.91
3,520
3,545
0.71
D100-VU-1/05
0.29
0.59
3,602
3,506
-2.67
Samples of triple-base powders PP were prepared as modifications of our powder

D350 [5] (1.5% Centralite I, 3% 2,4-dinitrotoluene, 0.5% graphite, 44% nitrocellulose with
13.1%N, 27% nitroguanidine, 25.5% nitroglycerine) - see Table 4.
Table 4. PP samples
Sample
Description
Volatile
components
QE,exp.
QE,calc.
Error
[J.g-1]
[J.g-1]
[%]
[%]
PP-0039
original D350
0.57
4,249
4,236
-0.31
PP-0040
+ linear nitramines s.08
0.69
4,284
4,270
-0.33
PP-0041
0.77
4,333
4,276
-1.32
PP-0042
0.60
4,286
4,326
0.93
PP-0043
+ polymer s.19
0.78
4,211
4,164
-1.12
PP-0044
+ Acardite II
0.72
4,227
4,235
0.19
PP-0045
+ hexogen
1.94
4,026
4,162
3.38
Graphical comparison of measured and calculated heats of explosion is presented in

Fig. 1
747
4,300
SA
S105
D100
PP
4,100
QE,calc.
3,900
3,700
3,500
3,300
3,100
3,100
Fig 1.
3.
3,300
3,500
3,700 3,900
QE,exp.
4,100
4,300
Comparison of measured and calculated heats of explosion
CONCLUSIONS
TCHAR easily provides estimation of values of heat of explosion with acceptable

accuracy. The error did not generally exceed the range from -7.38 to +3.38 % and for 22
samples out of 25 did not exceed the range from -2.67 to +1.82 % - all related to the
measured values of heat of explosion. Small number of higher differences can also be caused
by preparation or measurement errors or sample structure inhomogeneities. For more
complex study with statistics it is necessary to have much more measured values of heats of
explosion of particular mixtures with exact analysis of their complete composition.
Interesting, but problematic, is also comparison of many other parameters, which can be
calculated with this program. In any case, in scientific research TCHAR program is a good
tool for estimation of properties, when new propellants are designed, and for comparison of
similar samples with little changes in their composition.
748
REFERENCES
[1]
[2]
[3]
[4]
[5]
J. PETRLEK and J. ZIGMUND:
Modelling of Thermodynamical Properties of Propellants, Proc. of

27 Int. Ann. Conf. of ICT, Karlsruhe, Federal Republic of Germany, p.93, 1996
Procedure 701.1, Explosia Company, Pardubice, Czech Republic
V. PU: Vroba jednoslokovch prach bez tkavch elatintor, Proc. of 3rd Sem. ,,New
Trends in Research of Energetic Materials,, , Pardubice, Czech Republic, p.159-164, 2000
Technical Documentation of Product TDV 556/04, Explosia Company, Pardubice, Czech
Republic, 2004
Technical Documentation of Product TDV 538/04, Explosia Company, Pardubice, Czech
Republic, 2004
th
749
CHANGES OF SOME THERMOPHYSICAL PROPERTIES

OF NITROCELLULOSE PROPELLANTS
DURING ACCELERATED AGING
Maa Raji Linari, Muhamed Sueska, and Sanja Matei Muani
Brodarski institut Marine Research & Special Technologies,
Av. V. Holjevca 20, 10020 Zagreb, Croatia
Abstract:
The ageing of nitrocellulose propellants causes changes of a number of relevant
propellants properties (decrease of nitrocellulose molar mass, decrease of stabiliser
content, change of thermophysical properties, etc.). The quality of propellants at a given
moment of time, i.e. their stability, may be determined by measuring these properties.
The changes of some thermophysical properties of nitrocellulose propellant during
accelerated aging have been studied in this paper applying differential scanning
calorimetry (DSC).
The series of nitrocellulose propellant sample were aged in temperature range between
80 up to 100 C. It was found out that some DSC parameters of aged nitrocellulose
propellants differ significantly (temperature of maximum and width at half height of
DSC peak increase, height and reaction enthalpy decrease). This suggests that the
information derived from DSC measurements can help in nitrocellulose propellant
stability evaluation.
Keywords:
1.
nitrocellulose propellant, DSC, thermophysical properties
INTRODUCTION
Thermal decomposition rates and mechanisms of nitrocellulose are of obvious

importance to efforts designed to enhance in manufacturing, storage and handling of
propellants. Because of that, its degradation mechanism has been the subject of numerous
investigations [1,2,3,4].
The safety and functional features of nitrocellulose change during the aging. The
chemical aging of nitrocellulose starts with the homolytic breaking of the weak O-NO2 bond
of the aliphatic nitrate esters thus forming nitrogen dioxide and corresponding alkoxyl
radical. These reactive free radicals immediately undergo consecutive reactions with nearby
nitrate esters molecules. Another main decomposition pathway is the neutral to acid
hydrolysis of the nitrate esters. The reaction is catalysed by moisture and residual acids
(which were not fully removed after nitrate ester synthesis), or by water and acids formed
during decomposition process[5]. The radicals and acids formed by reactions strongly selfaccelerated decomposition of the nitrate esters. This accelerating effect is called
autocatalysis. Whereas the primary homolytic reaction cannot be suppressed, the consecutive
reaction can be slowed down by bonding or elimination of acids, nitric oxides and water
from the system. This fact is used for the stabilisation of propellants based on nitric esters
(nitrocellulose) by aromatic amines (diphenylamine). It prevents the further decomposition
of nitrocellulose by chemically bonding with the autocatalytically effective nitric oxides and
nitric acid reaction product of the nitrocellulose decomposition.
750
A number of investigators have used different methods with a view to determine the
details of propellant chemistry at a molecular level, that the macroscopic behaviour can be
better understood. To assess the decomposition rate, thermal stability or self life, the use of
different thermal analysis is necessary.
In this work we have studies changes of some DSC parameters i.e. the initial
temperature, temperature of maximum and final temperature, height, width at half height of
peak and reaction enthalpy of nitrocellulose propellant under elevated temperature
conditions (80 100 C).
2.
EXPERIMENTAL
Experiments were carried out using nitrocellulose propellant samples. The samples of
nitrocellulose were cut to small grains having mass less then 5 mg. Then, the samples
weighing 10 g were put in glass vessels having volume of 100 cm3, and subjected to five
different elevated temperatures: 80, 85, 90, 95 and 100 C. The samples were taken and
analysed periodically. In addition, two samples are aged by simultaneous DTA-TGA (about
2 mg in mass, in open sample pans, at 140 and 150 C).
The thermal measurements were performed using a differential scanning calorimetry
(TA Instruments, DSC Model 2910) and simultaneous DTA-TGA (TA Instruments, DTATGA Model 2960). The samples were tested in small aluminium sample pans. The DSC
curves were obtained by using the samples of about 0,5 mg in mass, in closed sample pans,
with heating rate of 3 C/min and nitrogen (purity above 99,996 %) flowing at the rate of 50
ml/min. The DSC was calibrated with indium. Each of sample was measured three times.
3.
The characteristic DSC and TGA curves of nitrocellulose propellant are presented in
Figure 1.
Fig 1.
Typical DSC and TGA curves of nitrocellulose propellant
The self life covers the period of time during which the explosive can safety be stored without representing
any hazard to its environment. The safe life is limited by the extent of chemical aging reactions.
751
In temperature range from 25 C up to 150 C small weight loss is detected (0,69 %),
which may correspond to the removal of moisture and residual solvents. The TGA curve is
dominated by one-step decomposition, weight loss about 75 %. In same temperature range
(between 150 and 250 C) the DSC curve has one exothermic maximum at 194 C, which
corresponds to thermal decomposition process of nitrocellulose. Remaining 25 % mass of
sample is solid carbon which oxidation took place slowly and was followed by elimination
of carbon oxides.
Attempts have been made to identify and characterise the changes in the shape of
nitrocellulose propellants during the accelerated aging. Because of that, a series of DSC
measurements of aged nitrocellulose propellant was carried out. As an example the DSC
curves of samples aged at 100 C and non-aged sample of nitrocellulose propellant are
shown in Figure 2.
Fig 2.
DSC curves of accelerated aging samples at 100 C and non-aged sample

of nitrocellulose propellant
From curves given in Figure 2 it seems that the changes of DSC parameters are not to
pronounced during the aging. However, the shape of DSC peak changes slightly, which
results in change of the peak area with the aginig, and consequently in change of the reaction
enthalpy (heat of decomposition).
In order to quantify the changes observed the following DSC parameters were
determinated and separately analysed:
-
the initial temperature,

the temperature of maximum,
the final temperature,
height of peak,
width at half height of peak,
reaction enthalpy.
As mentioned earlier the accelerated aging testes were carried out in temperature range
80 100 C. Maximum weight loss reached in this way was about 2 % (because of safety
reasons). In order to broad range of weight loss, i.e. to see trends outside this region,
additional aging test were carried out using TGA, at 140 and 150 C, and sample weighing 2
mg. Weight loss of all samples during aging vs. coordinated of reduced time at 100 C are
shown in the Figure 3.
752
One can note that all data in Figure 3 are converted in reduced time coordinate. The
weight loss as a function of time is converted into the coordinate of reduced time, according
to the following equation:
T T
t = t100 10 100
10
(1)
where 10 is acceleration factor, in this work was taken to be 3.
Fig 3.
Sample mass loss as a function of reduced time (at 100 C)

at different temperature
From the curve given in Figure 3, one can conclude that weight loss up to 2 %
corresponds to the initial stage of decomposition during which low molecular decomposition
products of nitrocellulose were liberated, with no autocatalysis. This period is important
because very intensive reactions of stabiliser depletion take place during this early
decomposition stage. Autocatalytic decomposition reactions begin after that period, i.e. after
depletion of all original stabiliser and its stabilising nitroderivates.
The results of experimentally obtained DSC parameters are presented in Figures 4 - 9.
Fig 4.
Initial onset temperature change with aging time (reduced at 100 C)

at different aging temperature
753
Fig 5.
DSc peak maximum temperature change with aging time

(reduced at 100 C) at different aging temperature
Fig 6.
Final endset temperature change with aging time (reduced at 100 C)

Fig 7.
Height of DSC peak change with aging time (reduced at 100 C)

754
Fig 8.
Width of DSC peak at half height change with aging time

(reduced at 100 C) at different aging temperature
Fig 9.
Reaction enthalpy change with aging time (reduced at 100 C)

The results presented in Figures 4 to 9 are summarised and given in Table 1.

It follows from Table 1 that the DSC parameters for nitrocellulose propellant during
accelerated aging change in different ways. For example, temperature of maximum and
widths at half height increase with aging, height and reaction enthalpy decrease with aging,
while initial temperature and final temperature of DSC peak doesnt show significant
changes. It seems that height of DSC peak and reaction enthalpy are the most sensitive DSC
parameters to aging of nitrocellulose propellant.
755
Table 1.
Changes of DSC parameters in temperature ranges between

80 and 100 C for accelerated aging nitrocellulose propellant
80 150 / C
deviation / %
Ti
no change
Tmax
increase
low
Tf
no change
0,4
decrease
strong
increase
medium
decrease
strong
DSC parameter
Also, one can note from Figure 2 and Table 1 that the DSC peak becoming asymmetrical
during aging. Namely, the shape of DSC peak between initial temperature and peak
maximum increases with aging, indications that increase of decomposition rate with the
aging.
4.
CONCLUSION
The aging process in nitrocellulose changes the important propellant properties such as
stabiliser content, Arrehenius kinetic parameters, heat capacity, etc. Thus, ageing limits the
safe services lifetime and increases the possibility of self-ignition[6].
In the presented work nitrocellulose propellant samples were aged at 80, 85, 90, 95 and
100 C, during different time. It was found out that the maximum temperature and widths of
DSC peak increase with the accelerated aging time and temperature, while a final
temperature and reaction enthalpy decrease.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
A. BERGENS, B. NYGARD: Decomposition of Diphenylamine in Nitrocellulose Based

Propellants Proceedings of the Ninth Symposium on Chemical Problems Connected with
Stability of Explosives, Margretetorp, Sweden, 1992
M. A. BOHN, F. VOLK: Propell. Explos. Pyrotech. 17 (1992) 171-178
A. MISZCZAK, J. BLADEK: Chemical Stability Investigation of Propellants Under
Conditions of Accelerated and Natural Aged Proceedings of the Ninth Symposium on
Chemical Problems Connected with Stability of Explosives, Margretetorp, Sweden, 1992
M. RAJI, M. SUESKA: Influence of aging on nitrocellulose thermal properties and hazard
potential figure-of-merit High Temperatures-High Pressures 32 171-178, 2000
B. VOGELSANGER: Chemical stability, Compatibility and Self Life of Explosives,Chimia 58
401-408, 2004
M. RAJI LINARI: Thermal decomposition and ignition possibility of nitrocellulose
stabilised by diphenylamine Doctoral dissertation, 2005
756
ON THE ESTIMATION
OF COOKOFF TEMPERATURE USING FEM
Jakub eleovsk, and Ji Pachm
Department of Theory and Technology of Explosives, University of Pardubice,
Abstract
The simulation of temperature distribution during cookoff is presented in this
paper. Cylinder charge of RDX in steel confinement was heated by heating rate 3
K.hr-1. The ignition was predicted at the temperature 160C and occurred in the
central part of the charge. The simulation was carried out using the ANSYS
Multiphysics finite element code.
Keywords:
1.
FEM, cookoff, simulation, heat transfer
INTRODUCTION
The cookoff tests are used for evaluating of a hazard connected with heating of energetic
materials (especially EIDS Extremely Insensitive Detonating Substances) or ammunition
(LOVA LOw Vulnerability Ammunition). In case of energetic materials the explosive is
placed into a steel confinement and heated. In case of LOVA the whole ammunition object is
tested. The temperature of decomposition and the violence of decomposition reaction are
evaluated.
Because of the cost of experimental procedures, the numerical simulation is often used in
the field of characterization of various loading effect on the material. One of the most
commonly employed numerical techniques is the Finite Element Method FEM. The mean
principle of this method is to divide the body into small elements and to calculate the effect
of applied loads only in discrete points nodes. This creates a large system of equations,
which can then be solved numerically.
Several research groups published results of their effort in the field of cookoff computer
modeling to investigate energetic materials response to thermal heating, in a confined
geometry. Sandia National Laboratory (USA) is working on a multidimensional fully coupled
thermal/chemical/mechanical
code
to
calculate
the
response
of
reactive
materials [1-5]. They also try to develop model for thermally degraded energetic materials [6].
The LLNL (USA) is working on a single computer code specialized only for the cookoff
response [7-11] of energetic materials. The cookoff was also simulated in TNO (Netherlands)
[12]
, or in Australia [13].
The cookoff response simulation can be essentially divided into the two parts. The first
part (pre-ignition) is the calculation of temperature distribution during heating. The second
part (post-ignition) is the description of burning, deflagration, eventually DDT (Deflagration
to Detonation Transition) and the reaction violence evaluation.
The first part of cookoff simulation, carried in the ANSYS Multiphysics finite element
code, is presented in tis paper.
757
2.
THEORETICAL DESCRIPTION
The calculation of the temperature distribution during cookoff can be described as the
transient heat conduction with the heat source. The transient character of the problem is given
by consecutive heating of the confinement and also by the heat generation caused by
decomposition of the explosive. The problem is nonlinear because of exponential dependence
of the heat generation on the temperature.
The equation for transient heat conduction with heat source is given:
C
T T T
T
+ q& ;
=
+
+
t
x y z
(1)
where C is specific heat, is density, T is temperature, t is time, is thermal conductivity, x,

y, z are cartesian coordinates and q& is a heat source.
The source of heat (heat generated by decomposing explosive) can be in the simplest
form expressed by Arrhenius equation (for unit volume):
q& = H r A exp( E / RT ) ;
(2)
where H r is reaction heat, A is preexponential factor, E is activation energy and R is

universal gas constant.
To solve the equation (1) one has to define the boundary conditions. The first one is the
temperature distribution in the time t = 0, the second one is the condition describing the
heating of the confinement from the outer side.
3.
SIMULATION
The simulation of cookoff temperature estimation in the ANSYS Multiphysics code was
carried out on the geometrically simple problem. The cylinder charge of RDX (diameter 30
mm, length 150 mm) was confined by steel sleeve with the thickness of 3 mm (see figure 1).
The initial temperature of the whole set was 400 K (127C) and the confinement was heated
from the outer side (heating rate 3 K.hod-1).
Fig 1.
Schema of modeled problem.
758
3.1.
Material properties
Material properties of RDX can be found in the LLNL Explosive Handbook [14], the
properties of steel are from chemical engineering tables [15]. All the material constants are
given in table 1.
Table 1.
Material properties.
Property
H r
A
E
3.2.
Steel
7840
465
49.8
RDX
1720
1226
0.106
2.092 106
2.02 1018
197.1 103
Units
kg.m-3
J.kg-1.K-1
W.m-1.K-1
J.kg-1
s-1
J.mol-1
Geometric model
The modeled problem is axisymmetric, therefore the simulation was carried out in two
dimensions only. The symmetry of central plane was also used. The meshed geometric model
can be seen in the figure 2. The steel confinement is colored light whereas the explosive is
dark. The element PLANE77 (planar element with 8 nodes) was used.
3.3.
Loads
Two types of loads were used in the simulation. The first one is represented by the
boundary condition (heating with the rate of 3 K.hr-1). The second one is the heat generated
by the decomposition of the explosive. This heat was expressed from the equation (2) as the
function of temperature. The heat generation was applied on all inner nodes of the explosive.
The loads were applied in two load steps. The first one longed for 10 hours and the
temperature of 430 K was reached on the outer side of the confinement. This load step was
divided into ten substeps (by 3 600 s). The heat generation above the temperature of 430 K is
bigger, therefore the time substep had to be reduced in the second load step. The second
loadstep longed for 7 200 s with the substeps of 50 s.
4.
The calculated temperature distribution can be seen in figures 3 - 5. The first picture (3)
corresponds to the time of 28 800 s (480 min) from the beginning of heating. The temperature
of the confinement is 424 K (151C). The heating of the explosive by decomposition can be
seen, despite the maximal difference of temperature is only 1C.
The temperature difference grows with the subsequent heating of the confinement. This
can be seen in the figure 4, which corresponds to the time 39 900 s (665 min). The temperature
of the confinement is 433 K (160C), whereas the temperature in the central part of the charge
is by 25C higher. This is the last time substep before the ignition.
The ignition of the explosive can be seen in figure 5. The temperature in the center of the
charge rises rapidly and the temperature difference between the center and the confinement
exceeds 100C. This figure corresponds to the time of 39 950 s (666 min) and the temperature
of the confinement is 433 K (160C).
759
Fig 2.
Meshed geometric model (light confinement, dark explosive).
Fig 3.
Temperature (in K) distribution at time t = 28 800 s (480 min).
760
Fig 4.
Fig 5.
761
Fig 6.
The time behavior of temperature.
The time behavior of the temperature is presented in the last figure 6. The temperature in
the center of the charge is plotted by solid curve; the temperature of the confinement is
plotted by dashed line. The rise of central temperature can be seen clearly as the ignition
arises.
The calculated ignition temperature is in accordance with the experimental results,
published for another RDX based high explosives. The PBXN-109 heated at the similar
condition (the same initial temperature and heating rate) was ignited at the temperature of
162.5C [16] and Comp. B (again the same initial temperature and heating rate) run away at the
temperature of 166C [17].
The temperature distribution was modeled successfully and the ignition point was
localized as expected for slow heating rate. Therefore one can say, that the modeling of the
first part of cookoff experiment is coped. The future work will be concerned on the
simulation of real experimental geometries and also on the simulation of the second part of
cookoff experiment to describe the violence of the reaction.
5.
CONCLUSIONS
The simulation of thermal distribution by heating of RDX at the conditions similar to

slow cookoff (steel confinement, heating rate 3C.hr-1) was carried out in the ANSYS
Multiphysics code. The explosive is ignited from the central part of the charge at the time of
666 min (corresponding to the temperature of confinement 160C) what is in accordance with
experimental results for another RDX based explosives.
REFERENCES
[1]
M.R. BAER, M.E. KIPP, R.G. SCHMITT, AND M.L. HOBBS: Towards assessing the violence of
reaction during cookoff of confined energetic materials. In JANNAF Combustion
Subcommittee and Propulsion Systems Hazards Subcommittee Joint Meeting, Monterey,
California, USA, November 4-8, 1996.
762
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
M.L. HOBBS AND M.R. BAER: Multidimensional fully-coupled thermal/chemical/mechanical

response of reactive materials. In JANNAF Propulsion Systems Hazards Subcommittee
Meeting, Huntsville, Alabama, USA, October 23-27 1995.
M.R. BAER, M.L. HOBBS, R.J. GROSS, D.K. GARTLING, AND R.E. HOGAN: Modeling
thermal/chemical/mechanical response of reactive materials. In 26th International Annual
Conference ICT, Karlsruhe, Germany, 1995.
M.R. BAER, M.L. HOBBS, R.J. GROSS, AND G.R. SCHMITT: Cookoff of energetic materials. In
11th International Detonation Symposium, Snowmass, Coloredo, USA, August 31 - September
4 1998.
W.W. ERIKSON, R.G. SCHMITT, A.I. ATWOOD, AND P.D. CURRAN: Coupled thermal-chemicalmechanical modeling of validation cookoff experiments. In JANNAF 37th Combustion and 19th
Propulsion Systems Hazards Subcommittee Meeting, Monterey, California, USA, November
2000.
M.L. HOBBS, M.R. BAER, AND R.J. GROSS: A constitutive mechanical model for energetic
materials. In 20th International Pyrotechnic Seminar, Colorado Springs, Colorado, USA, 1994.
A.L. NICHOLS, R. COUCH, R.C. MCCALLEN, I. OTERO, AND R. SHARP: Modelling thermally
driven energetic response of high explosives. In 11th International Detonation Symposium,
Snowmass, Colorado, USA, August 31 - September 4 1998.
M.A. MCCLELLAND, T.D. TRAN, B.J. CUNNINGHAM, R.K. WEESE, AND J.L. MAIENSCHEIN:
Cookoff response of PBXN-109: Material characterization and ALE3D thermal predictions. In
50th Joint Army-Navy NASA Air Force (JANNAF) Propulsion Meeting, Salt Lake City, Utah,
USA, July 11-13 2001.
A.L. NICHOLS, A. ANDERSON, R. NEELY, AND B. WALLIN: A model for high explosive cookoff.
In 12th Internatioanl Detonation Symposium, San Diego, California, USA, August 11-16 2002.
M.A. MCCLELLAND, J.L. MAIENSTEIN, A.L. NICHOLS, J.F. WARDELL, A.I. ATWOOD, AND P.O.
CURRAN: ALE3D model predications and materials characterization for the cookoff response of
PBXN-109. In Joint Army Navy NASA Air Force 38th Combustions Subcommittee, 26th
Airbreathing Propulsion Subcommittee, 20th Propulsion Systems Hazards Subcommittee and
2nd Modelind and Simulation Subcommittee Joint Meeting, Destin, Florida, USA, April 8-12
2002.
J.J. YOH AND M.A. MCCLELLAND: Simulating the thermal response of high explosives on time
scales of days to microseconds. In 13th APS SCCM Conference, Portland, Oregon, USA, 2003.
G. SCHOLTES, J. MAKKUS, AND N. VAN DER MEER: Measuring and modeling the cook-off
process. In 28th International Annual ICT Conference, Karlsruhe, Germany, 1997.
S.-Y. HO: Modeling of cookoff reaction violence of confined energetic materials. In Solid
Propellant Chemistry, Combustion, and Motor Interior Ballistics, AIAA 185, p. 591606.
Reston, Virginia, USA, 2000.
B.M. DOBRATZ AND P.C. CRAWFORD: LLNL Explosives Handbook. University of California,
Livermore, California, USA, 1 edition, 1985.
J. CAKL, A. KOV, AND J. HEMER: Pruka k chemickoinenrskm vpotm I. University
of Pardubice, Pardubice, Czech Republic, 3 edition, 2001.
H.W. SANDUSKY AND G.P. CHAMBRES: Validation experiments for modeling slow cook-off. In
JANNAF 38th Combustion and 20th Propulsion Systems Hazards Subcommittee Meeting,
Destin, Florida, USA, April 8-12 2002.
J.L. MAIENSCHEIN AND J.F. WARDELL: Thermal explosion violence of HMX-based and RDXbased explosives effects of confinement, and solid phase using the scaled thermal explosion
experiment. In Joint Army Navy NASA Air Force 38th Combustions Subcommittee, 26th
Airbreathing Propulsion Subcommittee, 20th Propulsion Systems Hazards Subcommittee and
2nd Modelind and Simulation Subcommittee Joint Meeting, Destin, Florida, USA, April 8-12
2002.
763
HAZARD OF EXOTHERMAL REACTIONS

IN PRODUCTION OF NITROCOMPOUNDS
Galina A. Shraiber, and Vlada M. Raikova
Mendeleev University of Chemical Technology, Miusskaya sq. 9, Moscow A-47, Russia
Abstract
Hazard of runaway of nitration processes in nitrocompounds production was
investigated. Temperature rise in DNT nitration mixes under near adiabatic conditions
was measured, and derivative dT/dt was computed. Heat evolution rate caused by
nitration (primary reaction) and oxidation (secondary reaction) were examined by
mathematical model. Limiting values of T and dT/dt for safe process were determined.
Keywords:
1.
hazard, safety, nitration, oxidation, heat evolution,

exothermal reactions
INTRODUCTION
Production of nitrocompoumds is connected with the danger of the accidental explosion.

Runaway of synthesis process (nitration) causes to frequently accident. Often not primary
reaction is causing trouble, but secondary reactions (oxidation or decomposition) starting at
higher than usual temperature and having a higher energy of activation than intended
reaction. Runaway starts with self-heating after decomposition of reaction mix and can be
occurred as thermal explosion, that causes to destruction of reactor or sometimes to fire and
explosion.
From a hazard point of view study of exothermal self-accelerating processes should be
consist of experimental testing, mathematical modeling of runaway reactions and
determination of practical safety criterion parameters.
Measurement of the heat generation rate (or the temperature rise) in different heat
exchange conditions is a goal of experimental study. The evaluation of thermal
characteristics of chemical reactions is different mater. Many empirical calorimetric
techniques have been used for these testing. The choice of suitable test method depends on
given objective. Investigation can be performed either by examining samples taken during
the course of reaction, or by simulating the complete operating procedure in calorimeter
device. In the first case, differential thermal analysis (DTA), differential scanning
calorimeter (DSK), adiabatic rate calorimeter (ARC) and others are used. In the second case
lager instruments with different technical constructions and methods of operation are used [1,
2]
.
Special test method called thermokinetic for investigation of exothermal selfaccelerating reactions was developed [3]. The reaction rate and heat generation were
measured by this method simultaneously. Thermokinetic method was used for hazard study
of nitration processes too. Most considerable results were obtained for TNT production. The
explosions during the nitration of DNT to TNT were shown to be connected with the violent
exothermal reactions and detonation of reaction mixtures [4-6].
764
The experimental results on thermal stability of DNT nitration mixes obtained by

thermokinetic method and examined framework of kinetic theory of exothermal reactions
under nonisothermal conditions are presented in this work.
2.
EXPERIMENTAL
Materials. Sulfuric acid (conc.100%) was prepared by mixing 93-95% sulfuric acid
(chemical grade) with 112-114% oleum (pure per analyses). Nitric acid (conc. 99%) was
purified by two-stage distillation, final distillate contained less than 0.05% nitrogen
oxides.2,4-dinitrotolune (DNT) and 2.4.6-trinitrotolune (TNT) were synthesized in
laboratory and purified by recrystallization from nitric acid (conc. 63%).
Instrumental. Experiments were carried out in the glass spherical reactor with jacket and
stirrer. Reactor volume was 250 cm3, diameter was 85 mm, and jacket diameter was 102
mm. The mass of reaction mix was 110 120 g. Heat removal conditions were varied by
using of various heat carrier (water, air) and changing of intensity of mixing. Heat transfer
coefficient (t) was determined by special experiments with neat sulfuric acid. t value for
water heat carrier changed from 200 to 470 W/(m2K) depending on stirring rate, t value
for air heat carrier was 10 W/(m2K) independently of stirring rate.
3.
RESULTS
Reaction mixes consisted of DNT, TNT, nitric (NA) and sulfuric (SA) acids. Total
concentration of DNT and TNT was 30% in mix. The solution of TNT and DNT in sulfuric
acid was poured into reactor and heated to 95 97 0C. Then water was removal from the
jacket, and nitric acid was introduced in the mix. Temperature of the mix during reaction
was measured by differential thermocouple and recorded by potentiometer. Thermograms
T(t) were treated, and derivatives dT/dt were computed.
Typical temperature curves T(t) for reaction mixes contained 5.5% NA are represented
in Fig. 1. Temperature rises rather, and its maximal value is greater when the concentration
of DNT is increased from 5 to 22%. At the beginning of process temperature rises from 95
to 105 110 0C, because NA dissolves in sulfuric acid and then temperature rise is reduced.
Further temperature curve turns up, temperature rises very fast (18 and 22% DNT).
200
22 % DNT
18 % DNT
T, C
160
9 % DNT
120
5 % DNT
80
Fig 1.
50
t, s
100
Temperature curves for reaction mixes contained 5.5 % NA.
765
Two different kinds of temperature curves are presented in Fig. 2. Lower curve
corresponds to slow temperature rise, and decreasing of derivative dT/dt (Fig.3). Higher
curve (Fig. 2) corresponds to fast temperature rise, and dependence of dT/dt vs. time has
minimum and maximum (Fig. 3).
160
22 % DNT
T, C
140
9 % DNT
120
100
80
Fig 2.
50
150
200
Temperature curves for reaction mixes contained 3.5 % NA.
22 % DNT
0.6
dT/dt, C/s
100
t ,s
0.4
0.2
9 % DNT
0.0
Fig 3.
50
100
t, s
150
200
The dependencies of derivative dT/dt vs. time calculated by temperature

curves (Fig.2).
766
In some tests DNT was introduced in mix after NA. Thus reaction of DNT with NA was
isolated from dissolving of NA in sulfuric acid. The temperature curves are similar with
each other and have maximal values about 2000C (Fig.4). The dependences of dT/dt vs.
time have identical maximum (Fig.5).
160
T, C
200
DNT
120
80
50
100
150
200
t,s
Fig 4.
Temperature curves for reaction mixes contained 7 % NA and 22 % DNT.

(1) NA was introduced in the mix DNT-TNT-SA; (2) NA was introduced
in the mix TNT-SA, and DNT was introduced after 82 s.
1
2
dT/dt, C/s
8
6
4
2
0
Fig 5.
50
t, s
100
150
The dependencies of derivative dT/dt vs. time computed by temperature

curves (Fig. 4).
767
4.
DISCUSSION
Temperature rise in nonisothermal conditions as a result of exothermal reactions is

described by equation:
cp
dT
=
dt
W Q
i =1
t F
m
(T To ) ,
(1)
where is the density of solution, cp is the specific heat of solution, Wi is the reaction
rate, Qi is reaction heat, t is heat transfer coefficient, F is the surface area of heat exchange,
m is mass of reaction mix and T0 is temperature of heat carrier.
Value of t in experiments was 10 W/(m2K), so heat loss was very small, and it could
be suppose that process was nearly adiabatic.
Two exothermal reactions proceed in mix, primer reaction (desired) is nitration, and
secondary reaction (unwanted) is destructive oxidation of DNT. A rate of both reactions can
be described by second order kinetic equations:
Wn = C N C D Bn e En / RT ,
(2)
Wox = C N C D Box e Eox / RT ,
(3)
where CN and CD are concentrations of NA and DNT in the solution (mol/l), En and Eox
are energies of activation, Bn and Box - pre-exponent factors.
The reaction heat values calculated from tabulated data and the values of kinetic
parameters obtained by treatment of experimental curves T(t) [5] are collected in Table 1.
Comparison of calculated kinetic constants Kn and Kox vs. 1/T are presented in Fig. 6. The
lines are crossed at value T = 160 0C, and the rates of nitration and oxidation are equal with
each other Wn= Wox. Equal values of heat generation rates (WiQi) are observed at 130 0C.
Table 1. Reaction heats and kinetic parameters of nitration and oxidation of DNT
Reaction heat ,
Kinetic parameters
Reaction
kJ/mol
E, kJ/mol
B, l/(mol s)
Nitration
100
84
1110 8
Oxidation
201-213
117.2
1110 12
768
10
Kn,Kox, l/(mol s)
1
0.1
0.01
Kn
Kox
1E-3
103/T, K-1
1E-4
2.0
Fig 6.
2.2
2.4
2.6
2.8
3.0
The rate constants of nitration (Kn) and oxidation (Kox) vs. 1/T.
Maximum adiabatic temperature rise Tad [2] that can be generated by reaction is
calculated by expression:
Tad =
Qi
cp
For solutions contained 5% NA the value Tad is 53 K for nitration, and 106 K for
oxidation. The temperature Tox at which oxidation becomes noticeable far lower than
maximum possible temperature for nitration (150 0C). In the case a considerable hazard may
exist. Experimental results are shown that limiting value of temperature in reaction mix is
about 120 0C. Runaway can occur in nitration, if temperature is higher than 120 0C.
Estimation of limiting temperature is necessary but is not enough for decision of safety
synthesis process. Second important parameter is rate of temperature rise, in other words, the
derivative dT/dt. At the beginning of process reaction rate rises. If dT/dt value tends to zero,
that the temperature tends to constant, and steady process is established. If dT/dt value no
tends to zero, that the temperature and reaction rate will rise rapidly, and reaction will be
runaway.
5.
CONCLUSIONS
Experimental dependencies of temperature vs. time for DNT nitration mixes were
obtained by thermokinetic method, and dT/dt values were computed. Mathematical model
described heat generation caused exothermal nitration and oxidation reactions is constructed.
It is shown that DNT nitration should be runaway, if the temperature of reaction mix is
higher 120 0C and dT/dt rises fast.
769
Acknowledgements
We thank Russian Foundation of Basic Research (RFBR) for partial financial support of
this work (Grant 04-03-32713).
REFERENCES
[1] PASMAN H.J. and ZEVENBERGEN J.: New safety thinking applicable to em synthesis and
manufacture, Proc. of the 7th seminar New trends in research of energetic materials, Univ.
Pardubice, CSR, 2004, p. 35-50.
[2] GREVER T., KLUSACEK H., LFFLER U.: Determination and assessment of the
characteristic values for the thermal safety of chemical processes, J. Loss Prev. Process
Ind.,1989,v. 2, p. 215-223.
[3] KONDRIKOV B.N., RAIKOVA B.M., VEGERA Ia.S.: Thermokinetic of nitration process of
DNT, Journal Applied Chemistry, 1991, 4, p. 877-883.
[4] RAIKOVA V.M., KONDRIKOV B.N.: Investigation of technological safety of nitration of
dinitrotolune by mathematical modeling, Chemical industry, 1997, 10, . 54-59.
[5] RAIKOVA V.M., KONDRIKOV B.N.: Mathematical modeling of the most hazardous situations
at TNT production, Proc. of 27th International Annual Conference of ICT, Karlsruhe, FRG, 1996,
p.105-1-14.
[6] KONDRIKOV B.N.: Some problems in the field of reliability and safety of new energetic
materials, Novel Energetic Materials and Applications. 9th Int. Workshop on Combustion and
Propulsion, Grafiche CSS, Bergano, Italy, 2004, p.3-19.
770
THEORETICAL STUDIES ON THE PROPERTIES

OF SOME ENERGETIC MATERIALS
Yuanjie Shu,*Chaoyang Zhang, Xinfeng Wang, and Xiaodong Zhao
Institute of Chemical Materials CAEP,
621900, P.O.Box 919-301,Mianyang, Sichuan, P. R. China
Abstract:
The theoretical results of structure-sensitivity relationship of several energetic materials
were introduced. It is demonstrated that the nitro group charges can be used to assess
impact sensitivity like bond length, oxygen balance and molecular electrostatic potential.
When nitro group Mulliken charges less than 0.23e, the compound may have H500.4m. In
addition, the thermal properties and the detonation performance of single and multi
furazans and furoxans compounds were introduced. The results suggested that those
compounds have high energy and relatively high sensitivity. They are the important
compounds in the viewpoint of high energy density materials designing.
Keywords:
1.
property, energetic material, theoretical calculation, sensitivity,

nitroaromatic compound, furazan, furoxan, nitro group Mulliken charge.
INTRODUCTION
The application fields of energetic materials depend on the properties such as denotation
performance, thermal stability, sensitivities etc.. However these properties were mainly
determined by the molecular structures of the energetic materials (EM). The molecular
geometric structures can be obtained by quantum chemistry calculation, while the properties
may be measured and estimated by experiments. There are many studies on the relationships
between the properties and molecular structures, but this is still an unclear subject until today.
In order to clarify the relationship further as well as the design of HEDM, we carried a series
of studies on the relationship of structure-property and the property prediction of EM.
2.
THE RELATIONSHIP BETWEEN THE STRUCTURES

As to nitro groups in nitro compounds, they are very strong electron-acceptors, that is to
say, they have big potentials to attract electrons. The same as the potentials of H atoms in
above hydrides, these potentials can also be represented by charges of nitro group, but not
only of one atom differently. Similarly, we can draw a primary and tentative conclusion
similar to the former: the more negative charges the nitro groups have, the more their
potentials decrease, the more stable the nitro compounds will be. For nitro covalent
*
syjfree@sohu.com or syjfree@sina.com
771
compounds, C-NO2, N-NO2, O-NO2 bonds (R-NO2) are the weakest in molecule and their
ruptures are the initial steps in the decomposition or detonation. Moreover, the impact
sensitivities are used to show the stabilities of nitro compounds. In order to answer the
questions: do exist some relationships between the nitro group charges and the stabilities (for
example, impact sensitivities) of compounds, and what are the relationships? Mulliken nitro
group charges and R-NO2 bond lengths of optimized nitro compounds are calculated by the
Acceryls code Dmol3, General Gradient Approximation method (GGA) and Beck LYP
hybrid functional and DNP basis set. The nitro group charges (QNO2) are calculated by
equation (1), where QN, QO1 and QO2 are charges of N and O atoms on nitro group
respectively.
(1)
2.1
Nitroaromatic molecules
Structure-sensitivity relationships of nitroaromatic compounds have been studied most

by many scientists. And the work of Kamlet, Politzer and Owen must be the most famous
and accepted widely. Here, the nitroaromatic molecules in Murray and Politzers [1] and Rice
and Owens [2] work are taken as our investigation objects.
No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Fig 1.
X, Y, Z (position)
OH,OH,OH (A)
NH2,OH,OH (B)
CHO,H,H (A)
NH2,NO2,H (Y)
OH,OH,H (C)
NH2Cl,H,H (A)
NH2OH,H,H (A)
C2H4OH,H,H (A)
OCH3,Cl,Cl (A)
Cl, H,H (A)
COCH3,H,H (A)
OH,H,H (A)
COOCH3,H,H (A)
H,H,H (A)
COOH,H,H (A)
OH, NH2, NH2 (A)
OH, NH2,H (C)
CN, NH2,H (C)
CH3,H,H (A)
NH2,H,H (C)
OC2H5,H,H (A)
OCH3,H,H (A)
OCH3,OCH3,H (A)
NH2, NH2,H (C)
O CH3, NH2,H (C)
NH2, NH2, NH2 (A)
The relationship among Vmid, max, -QNO2 and LB,max and H50
of nitroaromatics
In order to show the tendency or relationship more clearly among Vmid, max, QNO2 and
772
LB,max and H50 , we converted the data in ref.[1] into Fig. 1. In Fig. 1: the more -QNO2, the
less Vmid, max, the less LB,max, the more H50. In a certain extent, the bond strength can be
represented by the bond length approximately. Specially, the -QNO2 and Vmid, max are
discussed here. As the impact sensitivities of the nitroaromatic compounds decrease, there is
an increasing trend for -QNO2, but a decreasing one for Vmid, max. And there are the same
locations of inflexions on the curves on the bottom of Fig.1: M4, M5, M16, M17 and M20,
which are hydroxy or amino nitroaromatics. For hydroxy nitroaromatics, one of the
explanations of these inflexions may be due to the formation of unstable nitronic acid
tautomers, i.e., a minor part of nitro-phenols (NP) can change into nitronic acids (NA) as
equation (2), which have higher total energies and higher sensitivities. We calculated two
isomers of PA and TNAP respectively (drawn on Fig. 2) and found that it was the true case PA: -921.2358675 hartree (NP) and -921.1879466 hartree (NA), TNAP: -976.5969929
hartree (NP) and -976.5654405 hartree (NA). So, it is no surprise for these exceptions
existing: -QNO2 is calculated according to NP, but H50 must be of NA. The other is that the
orders of C-OH bond are less than those of C-NO2 bond, that is, it is not C-NO2 bond but COH bond the weakest in the entire molecule (described in Table 1). For amino nitroaromatics,
their low -QNO2 but high H50 must be owing to their resonant structures, which can diminish
sensitivities.
Note: Position and QNO2 are the location and charges respectively of nitro group whose
negative charges are the least, H50 and Vmid, max are from Ref. [1] and LB,max is the
maximum C-NO2 bond length.
(2)
Table 1. Bond orders of C-NO2 bond (BOC-N) and C-OH bond (BOC-O)
in hydroxy nitroaromatics at CNDO/2 level [3]
Compounds
H50, m
BO
C-N
>3.20
0.6859
>3.20
0.6889
2.96
0.6895
0.87
0.6859
0.43
0.6887
0.27
0.6992
BO
C-O
0.5316
0.5314
0.5307
0.5305
0.5300
0.5295
It can be concluded from the above discussion that there is a relationship between nitro
group Mulliken charges and impact sensitivities of nitroaromatics: the nitroaromatic
compound whose -QNO2 is low has low H50 when the C-NO2 bond is the weakest and its
rupture is the primary step in decomposition. It has the same role in assessing impact
sensitivity as the molecular electrostatic potentials introduced by Murray and Politzer.
773
Correlations between impact sensitivity and chemical composition reported earlier by

Kamlet and Adoplph [4] are probably the most widely applied in the area of energetic
materials research. Figure 2 indicates the Kamlets suggestion: the less oxygen balance, the
lower sensitivity. But special cases such as TNAP and TNT must be due to other
decomposition mechanisms. Figure 3 also shows there are similar tendencies on BDE and QNO2 curves. But it seems that correlation of nitro group charges and impact sensitivity is
better that of bond dissociation energy and impact sensitivity for there are only two obvious
inflexions on the curve: PA and TNAP. This can be explained as hydroxy nitroaromatics
discussed above.
Fig 2.
The relationships between Vmid, max, -QNO2 ,OBand LB,max and H50
of nitroaromatics
So, it shows that investigation on nitroaromatics can relate the impact sensitivity to the
nitro group charges: the more -QNO2 , the bigger H50. That is to say, -QNO2 can be used as an
alternate method to estimate impact sensitivities for nitroaromatics.
2.2
Other EM molecules
Fig 3.
Nitro charges and impact sensitivity of other nitro-compounds in Ref. [5, 6]
774
Two main nitric ester explosives PETN and NG, two main nitramine explosives RDX
and HMX, and other explosives are shown in Fig. 3. From the figure, it can be found that
two nitric esters have positive nitro group charges and this must be due to nitro groups
joining with oxygen atom, which has more ability to attract electrons. Apparently, nitro
groups potential to attract electrons can not decrease. Accordingly, the nitric esters are
unstable and have smaller H50. Explosive molecules such as CL-20, RDX, HMX, 2# and
TNAZ have less than 0.12e -QNO2 and higher impact sensitivities (H500.32m), but other
explosives such as FOX-7, LLM-105, NTO and NQ have more -QNO2 (>0.26) and higher
H50 (>0.72m). So, for these explosive molecules, that the molecules have high H50 if they
have low QNO2 can also be showed as a rule, even though there is not a corresponding
relationship between QNO2 and H50 (for example, PETN has the least H50 but not the largest
QNO2, and FOX-7 has the least QNO2 but not the largest H50). Also, it can be used -QNO2 to
predict that some compounds such as NTO and TEX have high impact sensitivities, but TNA
and DAMN may be not too sensitive for it has not too few -QNO2. Additionally, we can
roughly conclude there is an increasing sensitivity order for nitro compounds in which the CNO2 bond, N-NO2 bond and O-NO2 bond are the weakest from the viewpoint of nitro group
charges.
Obviously, it has much application on estimating impact sensitivity by nitro charges: for
almost nitro molecules in which the C-NO2 bond or N-NO2 bond or O-NO2 bond is the
weakest. It is important to emphasize that the rule cant run when the primary decomposition
step is not those bonds break. These exceptions are including above hydroxy nitroaromatics
and following special cases.
Fig 4.
Special cases for the rule
According to the rule of nitro group charges, there is an increasing order for H50 on Fig.
4: b<d<a<c. It is clearly that order does not accord with experiment. Reference [7,8]
suggests that loss of nitrogen (N1-N2) may conceivably be a possibility in the decomposition
of 1-picryl isomers. But for compounds c and d, it can use the rule. However, according to
the rule of OB compounds a and c, b and d should have the same sensitivity, but they do not.
So, considering chemical composition is not enough, structure is probably the main factor
affecting the sensitivity. For compound e, it is very sensitive because N1-O2 bond is very
weak and it is the initial decomposition step, even though it has -0.239e QNO2.
775
3.
THEORETICAL CALCULATION ON NITRO FURAZAN

(FUROXAN) SERIES: THERMAL PROPERTIES
AND DETONATION PERFORMANCE
Energetic furazan compounds have being investigated for more than 20 years by
scientists from Russia and other countries. It is regarded that furazan group is an effective
unit in high energetic density materials (HEDM) composed of C, H, O and N atoms [9].
Scientists have synthesized many furazans and furoxans [10-15]. As a matter of fact, the
density and detonation velocity can be increased by 0.06-0.08g/cm3 and about 300m/s
respectively when a nitro group is displaced by a furazan group of HEDM [16]. Furazan and
furoxan are very important compounds we must take into account in HEDM designing. In
this designing, the standard heats of formation (HOF) and densities of designed molecules
should be considered, which is highly advantageous to predict the partial detonation
properties theoretically. T.S.Pivina et al [17] adopted comparative characteristic of energy
content calculating methods to calculate the HOF of furazan series, but only partial results
are satisfied due to the limitations of methods. Sheremeteeev A. B et als work gives much
help to researchers who engage in furazans synthesizing [18], but it does not offer theoretical
values of detonation properties. The most advanced calculation methods are employed and
results of bigger molecules are obtained in this paper, and the promising compounds are
offered for syntheses.
3.1
Single nitro furazans and furoxans
Comparing with the furazan molecules of M1~M12, the furoxan molecules of M13~M24
have more than one oxygen atom correspondingly. It must be the main reason that furoxans
have less values of HOF than furazans correspondingly. Owing to low atom density of
hydrogen, the furazans or furoxans will have low densities if they have substituted groups
containing hydrogen atom (s). For example, M2, M3, M6, M8, M14, M15, M18 and M20
compounds having low densities must be due to their containing amino group(s) or methyl
or methylene, which has hydrogen atom(s). Figure 5 shows that the weakest bonds in
furazan molecules are not on but outside the rings and joined with the energetic groups and
the weakest bonds in furoxan molecules are on the rings, that is to say, the furazan rings are
stable, but furoxan rings are unstable. Table 2 shows that the orders of these bonds are all
less than 1, and the furoxans have less bond orders than furazans correspondingly. So it can
be concluded that nitro furazans are more stable than nitro furoxans, and the trigger points
and initial mechanisms of these two sort energetic compounds decomposition are different:
the former must be the rupture of C-NO2 or C-NO2, the later must be the break of N-O on the
ring. Table 2 also shows the most nitro furazans and nitro furoxans are very energetic but
sensitive for being substituted by nitro group and other energetic groups such as -N3, NHNO2, -ONO2, -C(NO2)3 and -COOCH2C(NO2)3, namely, M1, M7, M9, M10, M11, M12,
M13, M16, M19, M21, M22 and M24 have higher calculated heats of detonation, detonation
velocities, detonation pressures and relative specific impulse. But M1 and M9 are
recommended for a comprehensive account including synthesis, molecular stability and
detonation ability.
776
Fig 5.
Single nitro furazans and furoxans (the arrow showes the weakest bond,
whose bond order (BD, listed in Table 1) is the smallest in the entire
molecule)
Table 2. Calculated results of nitro-furazans and nitro furoxans
3.2
Et
hartree
HOF
kJ/mol
Density
-H
g/cm
kJ/g
km/s
GPa
160.04
-671.2296198
266.69
2.02
6.584
9.44
100.08
-372.8363098
125.07
1.77
3.666
7.57
CNHO
2
130.06
-522.0536832
144.85
1.82
2.813
C N HO
131.05
-541.9264713
36.09
1.93
CNO
140.06
-558.9238589
393.70
CNHO
129.07
-505.9903162
CNO
156.06
CNHO
C N HO
10
CNO
11
12
No
Formula
CNO
CNHO
RIs
BD
42.29
1.023
0.921
25.17
0.768
0.801
7.66
26.23
0.725
0.932
6.124
8.71
35.05
0.990
0.923
1.75
5.178
8.52
31.71
0.970
0.925
119.95
1.63
3.498
7.05
20.73
0.709
0.928
-630.3143471
472.64
1.85
6.258
9.05
36.91
1.032
0.923
170.09
-669.6185013
447.86
1.74
4.583
8.71
32.97
0.921
0.922
175.06
-726.6009700
265.82
2.01
6.014
9.35
41.32
0.989
0.922
294.23
2.09
5.585
9.33
42.08
0.951
0.918
0.917
264.07
-1119.6787651
CNO
2
176.05
-746.4732312
155.97
2.04
5.356
9.05
39.09
0.932
CNHO
2
322.10
-1347.6049467
-38.23
1.96
6.859
9.32
40.50
1.023
0.918
13
CNO
176.05
-746.4426487
196.66
2.04
5.588
9.15
39.93
0.951
0.842
14
CNHO
116.08
-448.0351254
12.25
1.80
4.272
7.75
26.66
0.806
0.862
15
CNHO
-597.2535481
103.31
1.84
5.814
8.51
32.53
0.970
0.855
16
C N HO
146.06
147.05
-617.1248965
0.42
2.04
6.177
9.42
42.35
0.983
0.854
17
CNO
156.06
-634.1311067
333.59
1.80
6.076
8.97
35.72
1.017
0.855
18
CNHO
145.07
-581.1958854
64.21
1.64
4.848
8.05
27.18
0.913
0.860
19
CNO
172.06
-705.5147730
431.98
1.88
7.085
9.07
37.48
1.067
0.859
20
CNHO
186.09
-744.8236682
390.57
1.73
5.180
8.57
31.81
0.962
0.861
21
C N HO
-801.8037483
216.62
2.03
5.886
9.37
41.77
0.977
0.860
22
CNO
191.06
-1194.8813074
243.51
2.11
5.085
9.17
40.86
0.912
0.859
23
CNO
24
CNHO
11
2
2
10
280.07
192.04
-821.6825499
89.35
1.99
4.563
8.56
34.48
0.867
0.841
338.02
-1422.8109618
-95.61
2.00
6.253
9.25
40.39
0.982
0.853
12
Poly-furazans (furoxans) on a chain
According to the mode of linking furazan rings, these molecules are discussed in eight
groups. (I) The first includes M25~M29, which formed by linking furazan rings directly. It
can be found from Table 3 that -HD, VD, PD and RIs change in a descendant order from
M25 to M29. As a calculated result, the detonation abilities of this sort of compound
decrease as furazan rings increase. There are seemly not regular total energy changes from
M25 to M29, so how the molecular stabilities change can not be confirmed. Moreover, this
sort of compound all has high density according to the calculated results. (II) M30~M34
777
belong to the second group, the furazan rings are linked by azoic group alternately. This kind
of compound has lower density than the first. At the same time, we also can not confirm the
molecular stabilities for the total energies change irregularly from M30 to M34. (III) The
third is M35~M37, the nitro furazans are linked by furoxan(s). As a distinctive characteristic,
these compounds have very excellent detonation abilities: very high -HD, VD, PD, RIs and
high crystal densities. (IV) M38~M41 belong to the fourth group, the furazan rings are
linked by azoxy group alternately. Comparing with the second group, this group has better
detonation properties. (V) The fifth one includes M42~M44, which linked by the nitro
Table 3. Calculated results of poly-furazans on a chain
Note: Et is energy change when a structural repeat unit increases. For M26, Et =Et (M26)-Et (M25)
oxygen atom. There is a decreasing trend of detonation ability when the chain increases for
this sort of molecule. (VI) The sixth is linked by sulfur atom, and includes M45 and M46. It
can be found that the detonation abilities of these compounds are common. (VII) This sort
molecule is linked by carbonyl group(s). (VIII) These molecules are much bigger and linked
by some different groups.
778
At the same time, there are some tendencies for these molecules according to the
calculated results in Table 3. At first, effects on the properties by the linking groups can be
concluded as some descendant orders: (1) for density: I>VI,VII,III,V>IV,II; (2) for heat of
denotation: III>VI>II>I>VII>VI; (3) for detonation velocity: III>I,VI,V,II,VII>VI; (4) for
detonation pressure: III,I,VI>V,VII>II>VI; (5) for relative specific impulse:
III>VI>II,V>I>VII>VI. Taking it all round, the furoxan group (III) is the most helpful to
enhance detonation properties, and the sulfur atom (VI) is the least. Secondly, it seems that
increasing the length of the chain is useless to increase explosive performances of all sorts of
molecules according to the table. Thirdly, it cant be confirmed the variation rule of
molecular stability for the irregular total energies change of all sorts of molecules when they
augment. At last, the detonation properties must be affected by the linking order and quantity
of groups. For M49 and M51, M50 and M52, there is no obvious evidence to show that the
azoxy group is superior to the azoic group for detonation.
3.3
Poly-furazans (furoxans) on a circle
Fig 6.
Poly-furazans (furoxans)on a circle
From the calculated results of ring-shaped Poly-furazans (furoxans) molecules in Table 4,

we can draw some conclusion as following. At first, it is useless to enhance detonation
ability by increasing the circle. Secondly, the strain energies of the rings take big effect on
the heat of detonation. Moreover, the azoxy group is more efficient than the azoic group to
increase crystal density. Additionally, much negative oxygen balances must be the main
reason for some compounds have not good denotation parameters.
Table 4. Calculated results of poly-furazans on a circle
No
Formula
MW
53
54
55
56
57
58
59
60
61
62
C4N4O2
C4N8O2
C4N8O4
C6N6O3
C6N12O3
C6N12O6
C8N8O4
C8N16O4
C8N16O8
C10N10O5
136.07
192.10
224.09
204.10
288.14
336.14
272.14
284.19
448.19
340.17
Et
hartree
-521.6971295
-740.7234593
-891.1547197
-782.7350627
-1111.1002303
-1336.7408860
-1043.6433564
-1481.4890473
-1782.3056642
-1304.5336289
HOF
kJ/mol
696.87
956.61
844.77
561.91
1398.55
1247.12
760.27
1810.72
1704.19
1005.29
779
Density
3
g/cm
1.89
1.90
2.04
1.96
1.91
1.98
1.95
1.88
1.96
1.94
-HD
kJ/g
6.749
6.134
5.745
4.380
6.007
5.685
4.439
5.867
5.778
4.582
VD
km/s
8.73
8.82
9.76
8.04
8.80
9.52
6.96
8.65
9.49
8.07
PD
Gpa
34.79
35.61
22.52
30.14
34.79
30.20
22.52
30.14
35.61
30.20
RIs
1.025
1.002
1.019
0.835
0.992
1.014
0.728
0.981
1.022
0.853
4.
CONCLUSION
1)
Nitro group charges can be regarded as a structural parameter to assess impact

sensitivity (the compound will has high H50 when it has much -QNO2).
Considering the molecular structure has much availability in almost all nitro
compounds when the C-NO2 bond or N-NO2 bond or O-NO2 bond is the weakest
in the entire molecule. According to the data in this paper, the compound may
have H500.4m when its nitro group has less negative charges than about 0.23e.
2)
Furazans and furoxans are commonly dense, energetic but sensitive, that is to say,
their densities, heats of detonation, detonation velocities and pressures, specific
impulse and sensitivities are high. So they should be one of the important subjects
for HEDM designing and synthesizing.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
MURRAY J., POLITZER P: Chemisty and Physics of Energetic Materials, 157~173, Kluwer
Academic Publishers. Printed in the Netherlands, 1990.
B.M. RICE ET AL.: Journal of molecular structure (Theochem), (583),69-72 2002
HEMING X.: Molecular Orbital Theory of Nitro-compound, Peking: Publishing House of
defense industry(in Chinese), 1994
KAMLET M J, ADOLPH H G.: Prop., Explos., Pyrotech., 4(2):30 1979
DOBRATZ B. M., CRAWFORD P.C. LLNL Explosives Handbook: Properties of Chemical
Explosives and Explosive Simulants (UCRL-52997-Chg.2),1985:187~188 (translated into
Chinese)
PHILLP F.P., GREGORY S.L., ALEXANDER R.M., ET AL.: Thermochimica Acta,
2002,384:187~204
NEUMAN P.N.: J.Heterocycl.Chem.1989, 8, 51
STORN, C.B., RYAN, R.R., RITCHIE, J.P., ET AL.: J.Phys.Chem. 1989, 93, 1000
A. B. SHEREMETEEEV.: Chemistry of furazans fused to five-membered rings. J. Heterocyclic
hem., 1995, 32, 371~384
[10] A.K.ZELENIN, M.L.TRUDLL.: J. Heterocycl. Chem. 1997, 34, 1057

[11] A.GUNASEKARAN, T.JAYACHANDRAN, J.H.BOYER, M.L.TRUDELL.: J. Heterocycl. Chem. 1995,
32, 1405
[12] R.D.SCHMIDT, G.S.LEE, P.F.PAGORIA, ET AL.: J. Heterocycl. Chem. 2001, 38, 1227
[13] D.CHAVEZ, L.HILL, M.HISKEY, ET AL.: J. Energetic Mater. 2000, 18, 219
[14] A.K.ZELENIN, M.L.TRUDLL, R.D.GILARDI. J. Heterocycl. Chem. 1998, 35, 151
[15] MAKHOVA N N, KULIKOV A S, BLINNIKOV A N ET AL.: 30th international annual conference of
ICT, June29 ~ July 2, 1999, 58
[16] OU Y.X.: Synthesizing Chemistry of Modern Explosives (in Chinese). Beijing. National
Defense Industry Publishing house.1998, 192
[17] TATYANA S. P., DMITRII V. S ET AL.: Propellants, Explosives, Pyrotechnics. 1995, 20,5~10
[18] ALEKSEI B. SHEREMETEV, VALENTINA O. KULAGINA ET AL.: Propellants, Explosives,
Pyrotechnics.1998, 23,142~149
780
INFLUENCE OF VELOCITY OF GAS FLUX

GOING TO EROSIVE BURNING PROCESS
INSIDE OF COMBUSTION CHAMBER
Martin imek*, Vt Kuttelwascher*, and Petr Stojan**
* University of Defense, Faculty of Military Technologies, Kounicova 65,
CZ-602 00, Brno, Czech Republic
** Explosia a.s., Research Institute for Industrial Chemistry,
CZ-532 17, Pardubice - Semtn, Czech Republic
Abstract:
This paper includes the results of Testing Solid Propellant Rocket Motor with erosive
burning part of burning process when gas generator is used for the purpose mentioned.
Students of doctors study programs of University of Defense in Brno and members of
Explosia, a.s. provided the realization of these tests.
Keywords:
1.
erosive burning, solid propellant, gas generator.
INTRODUCTION
The Solid Propellant Rocket Motor (SPRM) with high level of specific impulse needs to
realize the case of erosive burning of solid propellant. There is the problem with free crosssection of Combusting Chamber (CC). High velocity of gas stream brings very important
influence to the surface of propellant and internal surface of chamber, mainly when the
burning process is starting.
The flow of burnt elements passes through the free cross-section of CC. At the
beginning of burning process the products of burning flow through the channel with the
lowest cross-section has the maximum velocity. Then his value goes to lower according to
the burning rate. The erosion kind of burning is expired in the first part of mass propellant
burning.
Theoretical solution of testing Solid Propellant Rocket Motor with erosive burning part
of burning process was described early [5]. This paper includes the results of testing solid
propellant rocket motor with erosive burning part of burning process when gas generator is
used for the purpose mentioned. There are compared values of theoretical solution and
testing program.
2.
THEORY
Values of gas pressure inside of the CC, burning temperature and velocity of gas flow
are important parameters to express the function of SPRM. These values determine the rate
of burning. The erosive burning rate of rocket motors can be expressed as follows [1,4]
ue = u0 f (Tph ) psk K e ,
(1)
781
where u0 is unit burning rate, function f(Tph) includes sensitivity of SP with respect to the
temperature of SP surface, psk is pressure of gases inside of CC, is exponent of burning and
Ke > 1 is the parameter of erosive burning.
Parameter Ke of erosive burning SP can be expressed in equation [1,2]
K e = 1 + KW ( w wmez ) ,
(2)
where KW is expressions for coefficients of SP sensitivity towards the erosive burning and
wmez is a critical velocity of erosive burning rate (wmez ~ (100 - 150) m.s-1) [1,4].
Usual expression of internal ballistics solution can be mentioned in following manner :
1) Solution of long and thick SP grain.
2) Solution of SP charge using gas generator (GG) in order to obtain gas stream velocity
to be able to increase same conditions as first. There is chance to change the gas velocity
through the CC free cross-section. [4]
The second solution was chosen[5,6]. Testing rocket motor was made in the shape as
shown Fig. 4, having four main construction parts, i.e.:
Gas generator (2)
Regulator of gas flux to the main CC (3)
Main combusting camber (1)
Main CC nozzle (4)
Fig 1.
Testing rocket motor (1- main combusting camber, 2- gas generator,

3- regulator of gas flux to the main CC, 4- nozzle of main CC), 5- solid
propellant of gas generator, 6- tested solid propellant grain
782
Experimental measurement on the testing rocket motors involved measures of pressure

in main combustion chamber (see Fig. 2). Parameter Ke of erosive burning SP could be
obtained from follow equation
u
e
Ke = e = e ,
un un te
(3)
where un is theoretical non-erosive burning rate of used solid propellant, ue is erosive

burning rate of tested grain, te is duration of erosive burning which is obtained from
measurements of p = p(t) and ee is initial thickness of solid propellant grain to the time te.
p (MPa)
Pressure in main
combusting chamber
Ignite pressure
te
Fig 2.
th
t (s)
Presure p vs. time t in main combusting chamber and definition of time te

and time th.
The initial burning thickness ee in the time te (see Fig. 2) is obtained as

ee = e0
ie
,
ic i z
(4)
where e0 is overall burning thickness of propellant grain, ic is total pressure impulse

(presure impulse is defined as the integration of pressure p in main CC over the certain
duration t), ie is pressure impulse to the erosion burning time te and ic is pressure impulse
from igniter.
Then from experimental data is estimated suitable analytical relation between parameter
Ke of erosive burning SP and speed of erosive burning rate wpl
( ).
K e = K w pl
(5)
The equation (6) could be used for determination of critical speed level of erosive
burning rate wmez .
3.
EXPERIMENTS
783
UD in Brno had made the analysis of burning process and basic computing solution of
testing rocket motor. Exploitation and preparation of testing area were made in cooperation
between UD in Brno and Research Institute for Industrial Chemistry (Explosia a.s.,
PardubiceSemtn), which one is allowed to offer the testing area. Experimental
measurement on the testing rocket motors involved measures of pressure in gas generator
and measure of pressure in main combustion chamber.
For the experiments were used double based propellants designated TPH-ROP. It was
also used as propellant in gas generator. Theoretical (non-erosive) burning rate of tested
propellant is
u = 2,46942.105. p 0,401
(6)
Other thermo chemical characteristics of tested propellant are: specific weight

ph = 1650 kg.m-3; flame temperature Th = 2342 K; specific heat ratio (adiabatic exponent)
= 1,2471; molecular weight of burning products Mm = 23,95066 kg.kmol-1; parameter of
heat losses = 0,9.
Dimensions of tested propellant grain (TPG) and main CC are: external diameter of TPG
D2 = 0,0497 m; internal diameter of TPG D3 = 0,0135 m (e.g. overall burning thickness
e0 = 0,00905 m); length of TPG Lphz = 0,140 m; internal diameter of CC D1 = 0,050 m.
The gas velocity of around testing SP grain was controlled by number of SP grains in
gas generator and by critical diameter of nozzle of main combustion chamber [5]. Typical
presure p vs. time t diagrams for different gas velocity around testing SP grain in pressure
interval (14,8 15,8) MPa , resp. (8,7 9) MPa, are shown in Fig. 3, resp. Fig. 4. Note that
for wpl = 160 m.s-1 in pressure interval (14,8 15,8) MPa, resp. for wpl = 200 m.s-1 in interval
(8,7 9) MPa brings non-erosive burning (pressure-time profile have not erosive pike).
18
wpl = 280 ms
16
-1
14
12
p 10
(MPa)
wpl = 240 ms
-1
8
6
wpl = 200 ms
-1
4
wpl = 160 ms
2
0
Fig 3.
0.2
0.3
-1
0.4
0.5
t (s)
0.6
0.7
0.8
Presure p vs. time t in main combusting chamber including four different

gas velocity wpl around testing SP grain in pressures iterval (14,8
784
15,8) MPa. Note: for wpl = 160 m.s-1 with non-erosive burning (profile
without erosive pike).
10
wpl = 280 ms
-1
8
7
6
p
(MPa) 5
wpl = 240 ms
4
wpl = 200 ms
-1
-1
2
1
0
Fig 4.
4.
0.2
0.3
0.4
0.5
0.6
t (s)
0.7
0.8
0.9
Presure p vs. time t in main combusting chamber for different gas velocity
of around testing SP grain in pressures iterval (8,7 9) MPa. Note: for wpl
= 160 m.s-1 arrive non-erosive burning (profile without erosive pike).
RESULTS
Were calculated (see eqs.(3), (4)) of parameters of erosive burning Ke (given from
experiments) for diverse several gas velocities wpl. There are shown values in table 1 for
pressure interval (14, 8-15, 8) MPa, and. table 2 for pressure interval (8, 7-9) MPa.
Table 1. Measured and calculated values for interval of pressure (14,8 15,8)MPa.
wpl (ms-1)
un (ms-1)
te (s)
ue (ms-1)
Ke
160
1
200
240
280
0,01825 0,01805 0,01799
0,20043 0,19004 0,15981
0,01954 0,02091 0,02174
1,07026 1,15831 1,21838
785
Table 2. Measured and calculated values for interval of pressure (8,7 - 9)MPa.
wpl (ms-1)
un (ms-1)
te (s)
ue (ms-1)
Ke
200
1
240
0,01435
0,41874
0,01629
1,13555
280
0,01460
0,38190
0,01661
1,13716
Experimental relation between parameter of erosive burning Ke and and speed of erosive
burning rate wpl is:
a) for pressure interval (14,8-15,8) MPa:
Critical speed level of erosive burning rate wM = 162,56 ms-1, parameter of erosive
burning Ke = 1 + kw (w 162,56) , where kw = 1,618.10-8 w2 1,224.10-5 w + 0,00403.
b) for pressure interval (8,7-9) MPa:
Critical speed level of erosive burning rate wM = 201,33 ms-1, parameter of erosive
burning Ke = 1 + kw (w 201,33) , where kw = 5,523.10-9 w2 5,248.10-6 w + 0,00334.
From theoretical calculations
respect to gas velocities wpl :
[5,6]
are derived parameters of erosive burning Ke with
a) For pressure interval (14,8-15,8) MPa:
K e = 1 + k w (w 159,14 ) , where k w = 3,1468 106 w + 7,9267 104

b) For pressure interval (8,7-9) MPa are:
K e = 1 + k w (w 233,99) , where k w = 7,9525 106 w 6,3388 104 .

Note, that final relations are fashioned on calculated critical speed level of erosive
burning rate wM = 159,14 ms-1, resp. wM = 233,99 ms-1.
786
Both theoretical and experimental relation of parameter of erosive burning Ke vs. gas
velocity wpl is shown in fig. 5, resp. fig. 6. From comparison in fig. 5 for pressure interval
(14,8-15,8) MPa could be find in good accordance between theoretical and experimental
relation. Experiments show dependence between of values of working pressure and velocity
of gas stream, that burning process reaches erosive burning goes up with increasing of higher
values of these physical magnitudes. Intervals of gas velocity for working pressure (8,7
9) MPa we see at the (fig.6) seems to be changed of start point of erosive burning process. It
offers the reason for studying of this the problem with a new direction of searching.
1.25
1.2
Experimental
experimentln
1.15
Ke
1.1
1.05
Theoretical
teoretick
1
0.95
140
Fig 5.
160
180
200
220
240
-1
wpl (ms )
260
280
300
Theoretical and experimental relation of parameter of erosive burning Ke

vs. gas velocity wpl for pressure interval (14, 8 15, 8) MPa.
1.25
1.2
1.15
Ke
experimentln
Experimental
teoretick
Theoretical
1.1
1.05
0.95
140
Fig 6.
160
180
200
220
240
-1
wpl (ms )
260
280
300
Tehoretical and experimental relation of parameter of erosive burning Ke

vs. gas velocity wpl for pressure interval (8,7 9) MPa.
787
5.
CONCLUSION
Performed experiments prove, that theoretical part of this work was given correctly.
Chosen solid propellant seems to be able to burn according to estimated kind of burning
process. Solution of erosive burning process will be developed in future with respect to
working pressure interval.
This cooperation between University of Defense in Brno and Research Institute for
Industrial Chemistry (Explosia a.s., PardubiceSemtn) is presupposed to continue also in
future. Both partners going to cooperate in field of these practical tests of special testing
rocket motors for others students of doctor study programs too.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
F. LUDVK AND P. KONEN: Vnitn balistika raketovch motor na tuhou pohonnou hmotu,
Uebnice VA v Brn U-1153/1, Brno, 1999 (in Czech).
A. A. IKOV: Gazodinamika porochovych raketnych dvigatlej, Mainostrojenie, Moscow,
1963.
F. LUDVK: Zkuebn zazen pro sledovn vlivu rychlosti plynovho proudu na rychlost
hoen tuh pohonn hmoty, Sbornk VA v Brn, Brno, 1985 (in Czech).
A. H. KARAM ALLA: The influence of erosive burning of solid propellant on the internal
ballistics of solid propellant rocket motors, MA Brno, 1985.
M. IMEK, V. KUTTELWASCHER AND P. STOJAN: Influence of velocity of gas flux to the
burning surface, Proc. 7th Seminar New Trends in Research of Energetic Materials, Univ.
Pardubice, April 2004, p. 268.
M. IMEK: een erozivnho hoen tuh pohonn hmoty, Ph.D. Thesis, UO Brno, 2004
(in Czech).
788
DETERMINATION OF NITRATE ESTERS CONTENT

IN WATER SAMPLES
J. Skldal*, V. Jeov**, A. Eisner**, M. Kouba*, and K. Ventura**
* Research Institute of Industrial Chemistry, Explosia a.s.,
532 17 Pardubice - Semtn, Czech Republic
** Department of Analytical Chemistry, University of Pardubice,
532 10 Pardubice, Czech Republic
Abstract:
This paper describes a method for the analysis a mixture of nitroglycerin (NG)
and ethylene glycol dinitrate (EGDN) in water samples. Nitroglycerin has used
in explosives in many dynamites and multi-base propellants. Nitrate esters low
contents were determined with two extraction techniques solid phase
extraction (SPE) and solid phase microextraction (SPME). Several types and
adsorbents of the SPE columns were tested. SPME parameters including
adsorption and desorption time, coating type, rate of stirring and salt addition
were optimized. After extraction the analytes were separated and quantified
using high-pressure liquid chromatography with ultraviolet detection
(HPLC/UV).
Keywords:
1.
nitrate esters, SPME, SPE, HPLC-UV
INTRODUCTION
Nitrate esters such as nitroglycerin (NG; glycerol nitrate) and ethylene glycol dinitrate
(EGDN; nitroglycol) are typical compounds contained different explosives material. At first
nitroglycerin is a widely used in industrial explosive as dynamites and it is an ingredient in
multi-base propellants. Currently NG in low concentrations is also a drug for the therapy of
cardiovascular diseases but in contrast to the property it is toxic and it is good absorbed
through the skin. NG is colourless viscous liquid explosive and is sensitive against the
manipulation. NG is instability in water and its solubility in water is depended on the
temperature (at 20 C is dissolved 0,15 g of NG in 100 g of water). EGDN is a transparent,
colourless, liquid explosive. EGDN is less sensitive than NG and it is mainly used in
mixtures with NG for low-freezing dynamites. EGDN is more volatile and soluble in water
than NG (at 20 C is dissolved 0,50 g of NG in 100 g of water) [1]. These compounds are
present in water owing to manipulation of army munitions. Contamination by compounds is
often associated with explosives manufacturing and with loading, assembling and packing of
explosives into munitions.
The preconcentration technique is not required for the determination of nitrate esters in
high concentration. The samples of water are diluted in mobile phase and are applied direct
into HPLC system.
For the determination of low concentration of explosive in water is important to use
some preconcentration technique. Typical applied methods for the analyses of water samples
789
are liquid-liquid extraction (LLE), solid phase extraction (SPE) and solid phase
microextraction (SPME) [2] . Principle of LLE is based on two immiscible liquid phases.
In practice, one phase is usually aqueous and the other one is an organic solvent. There are
some parameters, which influence liquid-liquid extraction (miscibility, density, and
solubility). The difficulties of LLE are large volumes of organic solvents, formation of
emulsions and extraction time [3].
Solid phase extraction (SPE) is a simply physical extraction process between a liquid
and a solid phase [2]. The adsorbent is placed in a plastic SPE column by porous frits and
the commercial pre-packed columns containing stationary phases such as silica gel. SPE is
performed in the following steps: column conditioning, sample addition, column wash and
sample elution.
Solid-phase microextraction (SPME) is rapid, precise, simple, solventless method [2].
It is based on extraction using a thin polymeric-coated fused-silica fiber, fitted in a special
syringe-type holder for protection and sampling. When the fiber coated with stationary phase
is exposed to an aqueous solution or to the headspace above it, organic compounds are
extracted from their matrix to the stationary phase until equilibrium is achieved. After
the sorption the fiber is transferred to the desorption part of the GC or LC system. SPME
was applied to the determination of a large variety of volatile and semi-volatile analytes in
several types of environmental matrices.
The analysis of nitrate esters by gas chromatography is problematic because of the
thermal instability and sorption in the analytical system. That is why these thermally labile
compounds are usually determined by liquid chromatography [4,5].
The comparison of liquid versus gas chromatography for the analysis of mixture
explosive compounds is described with different SPME fibres [5]. It was shown that SPME is
able to detect these compounds in lower detection limits with shorter analysis times and
minimal sample handling.
Widely recommended technique is liquid chromatography with mass spectroscopy [4].
The comparison of electrospray ionization and atmospheric pressure chemical ionization is
demonstrated high specificity and sensitivity in the identification of nitrate esters.
EXPERIMENTAL
2.
2.1
Chemicals and reagents
Nitrate ester standards were dissolved in methanol. All solvents were obtained from
Riedel-de Han, Sigma Aldrich (Seelze, Germany). The redistilled water used for sample
preparation was prepared on a Demiwa 5ROI station Watek (Lede nad Szavou, Czech
Republic) and Ultra Clear UV system SG Water USA (Nashua, USA).
2.2
SPME/HPLC procedure
The standard solution mixed with EGDN at concentration 20 g/ml and NG at

concentration 20 g/ml in redistilled water was prepared each day. Four different SPME
fibres were applied for their ability to extract NG and EGDN: a 50 m carbowax/templated
resin (CW/TPR); a 75 m carboxene/polydimethylsiloxane (CAR/PDMS); a 65 m
carbowax/divinylbenzene (CW/DVB) and a 65 m polydimethylsiloxane/divinylbenzene
(PDMS/DVB). SPME fibres were conditioned in mobile phase for 30 min before the
sorption and cleaned in methanol for 5 min between sorptions. SPME fibres and SPME fibre
holder assembly were obtained from Supelco (Bellefonte, PA). Aliquot of 4 ml of standard
790
solution was extracted from 8 ml vials. Type of fibre, time and temperature of sorption, time
of desorption, stirring rate, concentration of NaCl and direct versus headspace sorption were
optimized. After sample extraction the SPME fibre was insert into the desorption chamber.
HPLC analyses were made with a chromatography system (Ecom s.r.o., Praque, Czech
Republic) with UV-detector. Separation was performed with a capillary column LiChrospher
100 RP-18e, MERCK (Darmstadt, Germany). It was used a Security Guard C18
Phenomenex (Torrance, CA, USA) too. For SPME/HPLC coupling was used an interface
desorption reservoir Supelco (Bellefonte, PA). An isocratic mobile phase acetonitrile : water
(60:40 v/v) was used at a flow of 0,5 ml/min and UV wavelength was set at 210 nm.
2.3
SPE procedure
SPE procedure was tested on different types of columns - STRATA X 6 ml, STRATA
C18 3 ml (Phenomenex), DISCOVERY SUPELCO-18LT and DISCOVERY SUPELCO-18
(Supelco) both 6 ml. The packing materials were chosen specially because they are reversed
phase, which are used for analyses of explosives.
The condition of SPE columns was made with 10 ml of methanol and then with
10 ml of redistilled water. It was extracted 500 ml redistilled water standard with EGDN at
concentration 1 g/ml and NG at concentration 1 g/ml. This volume was applied to all
kinds of columns and the column with the best efficiency was searched. The extraction time
was between 45 and 100 minutes. The extraction time depended on the used SPE column.
The elution of analytes was made with 10 ml of methanol. The extracted samples were
analysed with using HPLC/UV system.
2.4
HPLC/UV method
A Hewlett-Packard series HP 1050 system with diode array detection was used to
separate the extracted compounds by SPE and integrate the resolved peaks. For the
determination of NG and EGDN was used high performance liquid chromatography coupled
to UV-spectrometry. Separation was made on SEPARON SGX, C18 Tessek (Praque, Czech
Republic) column. An isocratic mobile phase of methanol : water (65:35 v/v) was used at a
flow of 0,6 ml/min and UV-wavelength was set at 210 nm, see Fig 1(a,b).
2450
2450
1950
1450
1450
1950
950
950
450
450
-50
180
230
280
330
-50
180
380
280
Wavelenght [nm]
Wavelenght [nm]
a) NG
Fig 1.
230
b) EGDN
UV-spectrums of NG and EGDN
791
330
380
3.

3.1
Optimization of SPME/HPLC-UV analysis
3.1.1. Comparison of SPME fibres

SPME fibres were exposed into the water solution with concentration 20 g/ml of each
nitrate ester for 15 min at 50 C. Direct sampling and magnetic stirring were used. It was
chosen static desorption for 10 min and the desorption chamber was filled with mobile phase
or methanol. The extraction efficiencies of fibres for EGDN and NG are shown in Fig 2.
The best extraction effects had PDMS/DVB fibre and the second best is CW/DVB. For the
next analysis was chosen PDMS/DVB fibre and the additions of mobile phase into the
desorption cell.
8000
Area [mV/s]
6000
EGDN
4000
NG
2000
0
CW /TPR
PDMS/DVB
Type of fibre
Fig 2.
The comparison of SPME fibers
3.1.2. Sorption process

The influence of extraction time was measured by immersing the PDMS/DVB fibre in
the sample for 1 to 70 min. In all cases, 4 ml of the standard solution, magnetic stirring,
sorption at 50 C and desorption 10 min were used. Adsorption profile was determined as a
function of time (Fig 3a). The extraction time from 30 min was demonstrated to be enough
to obtain a quantitative extraction in the case of the mixed samples. This time is a reasonable
compromise between good peak area and acceptable extraction time.
3.1.3. Desorption process
The optimization was done under the same conditions as the effect of adsorption time
(4 ml of the standard solution, magnetic stirring and sorption at 50 C for 15 min).
The optimized desorption time was determined at 5 min and no carry-over effect is obtained
at this time (Fig 3b).
792
EGDN
EGDN
12000
5000
NG
NG
4500
8000
Area [mV.s]
Area [mV.s]
10000
6000
4000
4000
3500
2000
0
0
10
20
30
40
50
60
3000
70
a) sorption time
Fig 3.
3.2
10
15
20
Desorption time [min]
Sorption time [min]
b) desorption time
The effect of sorption and desorption time
SPE/HPLC-UV analysis
First of all the best SPE column was found for analyses water samples. The best column
had to have short extraction time and high efficiency of extraction.
The best column was STRATA X (6 ml, Phenomenex) and the analytes were eluted with
10 ml of methanol. The characteristics of SPE columns are shown in Table 1.
Table 1. The comparison of different SPE columns used by extraction samples with
concentration 1 g/ml NG and 1 g/ml EGDN sample volume 500 ml
Sorbent Column Extraction Extraction
quantity volume
time
efficiency
[mg]
[ml]
[min]
NG [%]
Extraction
efficiency
EGDN [%]
Column
Sorbent
STRATA X
polymeric*
200
45
95,34
97,22
STRATA C18
C18
100
100
9,02
2,92
C18
1000
60
29,73
6,18
C18
1000
60
5,37
1,39
DISCOVERY
SUPELCO-18
DISCOVERY
SUPELCO-18 LT
polymeric* - patent pending
793
4.
CONCLUSION
The various types of SPME fibres were applied for the determination of nitrate esters.
PDMS/DVB fibre was the highest extraction efficiency for both nitrate esters and it was used
for the following optimization of conditions. The best conditions for analyses of low
concentration nitrate esters from water samples were found sorption for 30 min at 50C and
desorption time 5 min.
Four different SPE columns were tested in this study. The extracts were analysed with
HPLC/UV. The SPE column STRATA X with polymeric sorbent (6 ml, Phenomenex) had
the highest efficiency. The SPE columns with sorbent C18 are not suitable for the
preconcentration of nitrate esters in the comparison with STRATA X.
Acknowledgement:
Youth and Sports of the Czech Republic (Project MSM 0021627502) and from FRV
1932/2005.
REFERENCES
[1]
[2]
[3]
[4]
[5]
J. YINON, S. ZITRIN: Modern Methods and Applications in Analysis of Explosives, John Wiley
& Sons, Chichester 1993
J. R. DEAN: Extraction Methods for Environmental Analysis, John Wiley & Sons, Chichester
1998
S. MITRA: Sample Preparation Techniques in Analytical Chemistry, John Wiley & Sons,
Hoboken, 2003
X. ZHAO, J. YINON: Identification of Nitrate Ester Explosives by Liquid ChromatographyElectrospray Ionization and Atmospheric Pressure Chemical Ionization Mass Spectrometry,
J. of Chromatogr. A 977, 59-68, 2002
K. G. FURTON, J. R. ALMIRALL, M. BI, J. WANG, L. WU: Application of Solid-Phase
Microextraction to the Recovery of Explosives and Ignitable Liquid Residues from Forensic
Specimens, J. of Chromatogr. A, 885, 419-432, 2000
794
IDENTIFICATION AND MUTUAL DIFFERENTIATION

BETWEEN NITROESTERS AND NITRAMINES
BY THE COLOUR REACTION WITH THYMOL
M. korpkov
Research Institute of Industrial Chemistry, Explosia a.s.
532 17, Pardubice Semtn, Czech Republic
Abstract:
In the present paper there is discussed the colour reaction of nitramines and nitroesters
with thymol what can be used for their mutual differentiation. The reaction is
supplemental to these employed in commercial kits for detection of explosives. It was
examined in different modifications as a spot test, a test-tube reaction and also on
filters.
Keywords:
1.
colour reactions, detection of explosives, nitroesters, nitramines, thymol
INTRODUCTION
Nowdays when in the media there are still more often seen the news about the terroristic
attacks or their threats, the prompt and reliable analysis of materials suspicious to be
explosives again comes in on the forefront of interest. It especially refers to methods
allowing the analysis to be carried out in field conditions. Further there is a need to remark
such techniques are not only for above mentioned situations but it is possible to utilize them
for the orientation analysis in the field with the ecological damage e.g. former ammunition
magazines, military premises etc.
On the market there are available the commercial detection kits working on the
principles of colour reactions that detect the materials of groups such as polynitroaromatics
(TNT, DNT, picric acid, tetryl), nitroesters (PETN, NC, NG) and nitramines (RDX, HMX).
Some of the kits can detect also inorganic nitrates, chlorates (eventually perchlorates) and
bromates. But none according to performed literature search is able to discriminate if it is
nitramine or nitroester. From resultant developed colour there is only evident that the tested
material is generally due to the group of nitroesters and nitamines.
2.
THEORY
Because of the concerning the paper of colour reactions of nitroesters and nitramines
there are not mentioned the detection principles for polynitroaromatics, inorganic nitrates
and chlorates (eventually perchlorates and bromates). This part is focused on reactions found
in literature.
795
2.1
Nitramines
For their proof there are available some colour reactions such as:
2.2
reaction with aromatic amines in the presence of acetic acid,

reaction with ferrous sulfate in concentrated sulphuric acid,
Liebermann reaction as for nitrosocompounds a few milligrammes of tested
material is heated on boiling bath with the grain of phenol and 1 ml of
concentrated sulphuric acid, in the positive case there appears the red colour,
reaction with diphenylamine in concentrated sulphuric acid similar to nitrates
but they can be differentiate by no giving such reaction in the presence of
phosphoric acid,
after reduction to respective amines, they are then identified by consequential
reactions [1,3].
Nitroesters
For their proof there are available these colour reactions:
2.3
reaction with ferrous sulfate in concentrated sulphuric acid ( reaction is also given
by nitramines),
reaction with phenoldisulfonic acid,
reaction with diphenylamine in concentrated sulphuric acid (also given by
nitramines),
after reduction of nitrate ion to nitrite ion they (nitroesters) react in acidic
medium with sulfanilic acid and it leads to the formation of respective diazonium
salt what can be coupled with e.g. N-1-naphtylethylenediamine [1,3].
Additional found reactions
According to literature search [1,2,3,4,5] there can be said the commercial detection kits
employ for identification of nitramines and nitroesters the Griess reaction (the first step
is diazotation and the next one is coupling with the suitable coupling reagent) after the
previous dissolving of a sample and releasing of the nitrite ion by the reagent containing the
hydroxide (e.g. KOH, NaOH) and organic solvents such as dimethylsulfoxide, methanol,
ethanol, isopropyl alcohol and the like. Reaction with diphenylamine in sulphuric acid is
convenient as well but it is not possible to differentiate between nitroesters and nitramines
without using the TLC technique.
[6,7,8]
There are additional possibilities:
reaction of nitamines and nitroesters with a solution of the alkaloid brucine in

concentrated sulphuric acid gives orange-to-red colour (placed on the same
principles as reaction with diphenylamine), but has relatively low specification
inorganic nitrates react similarly,
reaction with quinalizarin solution (1,2,5,8-tetrahydroanthraquinone) in
concentrated sulphuric acid gives the colour what varies from violet to yellow,
both nitramines and nitroesters react,
796
decomposition by concentrated sulphuric acid what causes the formation of

formaldehyde and its following proof by chromotropic acid undergo cyclic
nitramines (RDX, HMX) and the colour is violet-pink,
reaction with thymol in concentrated sulphuric acid differentiate between
nitroesters (green colour) and nitramines (red-brown colour), reaction can give
the false response in the case of presence of sugars and aldehydes according to
[9]
there is possible to increase its selectivity by moderate heating of mixture of
sample and reagent and then to add a small amount of ethanol. Sugars and
aldehydes gives the brown colour but RDX gives a rich blue solution and HMX
gives a pale blue-green colour.
2.4
Thymol
According to U.S.P. [10] thymol is a phenol (isopropyl-m-cresol) and is isomeric with

carvacrol (isopropyl-o-cresol).
The thymol of commerce being derived from thyme oil and may be obtained by
fractionally distilling the pure oil of thyme, cymene and thymene pass over, and afterward
the thymol, which may be purified by crystallization from alcohol or the thyme oil may be
agitated with the solution of caustic soda, the hydrocarbons forming the tipper layer are
removed, and the lower solution containing the thymolate of sodium is neutralized with
hydrochloric acid, when thymol will rise to the surface, and may be purified by
crystallization from its alcoholic solution.
Fig 1.
Chemical structure of thymol
Thymol forms large, colourless, translucent crystals of the hexagonal system, having an
aromatic, thyme-like odour, and a pungent, aromatic taste with a very slight caustic effect
upon the lips. When liquefied by fusion it is lighter than water. When triturated with about
equal quantities of camphor, menthol or chloral, it liquefies. Soluble in about 1200 parts of
water at 15C and in less than its own weight of alcohol, ether or chloroform; also readily
soluble in carbon disulphide, glacial acetic acid and in fixed or volatile oils. Its alcoholic
solution is optically inactive. Being a phenol, thymol readily dissolves in caustic alkalies and
is regenerated by acids. Thymol also dissolves in 120 parts of glycerin.
797
Table 1. Published properties of thymol

Compound
isopropyl-m-cresol
Formula index
C10H14O
Chemical abstract (CAS)
89-83-8
Molecular weight [g. mol-1]
149,66 [10]; 150,22 [11]
Density [g.cm-3]
1.069 (15C) [10]; 0,965 (25C) [11]
Melting point [C]
50-51C [10]; 48-51C [11]
Boiling point [C]
230C [10]; 232C [11]
If a very small crystal of thymol be dissolved in 1 ml of glacial acetic acid and then 6
drops of suphuric acid and 1 drop of nitric acid be added, the liquid will assume a deep
bluish-green colour.
As a specific reaction for RDX it was published by Amas and Yallop [9]. It is based on
the production of the red colour formed by this compound in the presence of thymol and
nitrogen-free sulphuric acid. An objection to this test is, however, that a red colour is
produced by other compounds, particular sugars and aldehydes; a deep pink colour may be
produced unless the sulphuric scid is of the quality marketed as nitrogen free. They found
a method of surmounting this difficulty. After the procedure RDX produced a rich blue
solution. Under the conditions of the procedure sugars and aldehydes give a brown colour
and HMX gives a pale blue-green colour.
RDX is the only compound commonly found in explosive formulations that gives a blue
colour in this test. Authors tested a series of twenty-four nitramines, the only others that gave
a blue colour were 1,9-dinitroxypentamethylene-2,4,6,8-tetramine, 1-acetoxymethyl-3,5dinitro-1,3,5-triazacyclohexane, dinitrosopentamethylenetetramine and 1,3,5-trinitro-7nitroso-1,3,5,7-tetraazacyclooctane. Many sugars and aldehydes were tested and although all
gave the initial red colouration none gave a blue colour under the conditions of this test.
3.
EXPERIMENTAL
As optimal the reaction with thymol was chosen as the only one markedly differentiate
between nitramines and nitroesters. It was proposed to try out it as a spot test, a test-tube
reaction and also on filters for RDX, HMX and PETN.
After executing of the spot test to a few crystals of RDX, HMX and PETN standards
on a porcelain plate there was added approximately the same amount of powdered thymol
(recrystallized in advance) and a few drops of concentrated nitrogen-free sulphuric acid. The
product of RDX turned to red-brown, HMX to yellow-orange and this of PETN to yellowgreen. The reaction took about 1 minute maximum.
798
2,5
absorbance
1,5
0,5
0
350
400
450
500
550
600
650
700
750
800
wave lenght /nm/

Fig 2.
Vis spectra of RDX colour product in a test-tube
In the same way there was proceeded the reaction on filters as the best seemed those
made of glass microfibers what allow as skimmings as a good sinking with the same
results.
For the test-tube reaction there was added a larger quantity of concentrated sulphuric
acid and the resulting colour products were in agreement with those above-mentioned. The
PETN colour product was less green then in a spot test. Then the products were
spectrofotometrically measured on AvaSpec-2048 (see the in Fig 2. and Fig 3.). There was
also experimented to dilute the resulting colour products in ethanol or distilled water but
with no success. The solutions were observed to discolour.
2
1,8
1,6
absorbance
1,4
1,2
1
0,8
0,6
0,4
0,2
0
350
400
450
500
550
600
650
700
wave lenght /nm/

Fig 3.
Vis spectra of HMX colour product in a test-tube
799
750
800
1,4
1,2
absorbance
1
0,8
0,6
0,4
0,2
0
-0,2
350
400
450
500
550
600
650
700
750
800
wave lenght /nm/

Fig 4.
4.
Vis spectra of PETN colour product in a test-tube
CONCLUSION
The executed tests shows that samples of standards react with powdered thymol after
decomposition by concentrated nitrogen-free sulphuric acid in laboratory temperature
PETN (nitroester) gives bright yellow-green colour, RDX gives red-brown and HMX
yellow-orange colour.
For decomposition it is necessary to use concentrated sulphuric acid and therefore
common filter papers cannot be employed. So there were tried teflon filters with no false
reaction for sugars but their micropores do not allow the sinking of sulphuric acid and of
course the skimmings. Next the glass microfiber filters were tested. They seemed to be very
good liquid-absorbing, have sufficiently porous surface for skimming and do not contain
remains of cellulose so the material of a filter does not give the false reaction.
Colour products of test-tube reactions were measured by spectrophotometer AvaSpec2048. Resulting spectra show differencies.
Acknowledgment:
This research was supported by the Ministry of Industry and Trade, Project FT-TA/048.
800
REFERENCES
[1]
[2]
M. VEEA and J. GASPARI: Dkaz a identifikace organickch ltek, SNTL, Praha, 1973
B. T. FEDOROFF and O. E. SHEFFIELD: Encyclopedia of Explosives and Related Items, Volume
3, Picatinny Arsenal , Dover, New Jersey, p. 405-420, 1962
[3] J. YINON and S. ZITRIN: The Analysis of Explosives, Pergamon Press, Oxford, 1981
[4] J. YINON and S. ZITRIN: Modern Methods and Applications in Analysis of Explosives, John
Wiley & Sons, Chichester, 1993
[5] M. HANUS: Zklady analytick chemie vbunin, skripta KTTV, Univerzita Pardubice,
Pardubice, 1999
[6] B. GLATTSTEIN: US Patent 4788039, 1988
[7] Y. MARGALIT: US Patent 5296380, 1994
[8] Y. MARGALIT: US Patent 5480612, 1996
[9] S. A. H. AMAS and H. J. YALLOP: A Test for Cyclotrimethylenetrinitramine, Analyst, No. 94,
p.828, 1969
[10] http://www.ibiblio.org/herbmed/eclectic/kings/thymol.html
[11] https://www.sigmaaldrich.com/cgibin/hsrun/Suite7/Suite/Suite.hjx;start=Suite.HsViewHierarchy.run?Detail=Product&ProductNu
mber=SIGMA-T0501&VersionSequence=1
801
UV-VIS SPECTRA OF SOME ENERGETIC MATERIALS MEASURED

WITH THE OPTICAL FIBER PROBE
M. korpkov
Research Institute of Industrial Chemistry, Explosia a.s.
532 17, Pardubice Semtn, Czech Republic
Abstract:
The present paper is focused on possibilities of identification of explosives or their
decomposition products using the UV-Vis spectrometry. The spectral analysis is
employed there in the region of 200-800 nm. The spectra were measured using the
optical fibre probe what is especially set for powders and thick fluids.
Keywords:
1.
UV-VIS spectra, energetic materials, optical fibre probe
INTRODUCTION
Recently there is still put the accent on development of instrumental methods for
identification of explosives because of their need for environmental or security purposes.
The accent is given to instruments from the view of their size, necessity of additional
devices, power consumption and also qualification of an operator. The preparation of a
sample to analysis plays an important role as well.
All these requirements lead definitely to such instrumental techniques to be time-saving,
with no need of special qualified staffs and their utilization in field conditions is desirable.
2.
THEORY
Spectroscopic analytical techniques [1,2,3,4] are mainly based on measurement of

absorption of electromagnetic radiation. In visible (Vis) or ultraviolet (UV) spectroscopy the
absorption process is a result from electronic transitions in atoms or molecules. On
absorbing the incident radiation, the molecule rises from its ground electronic state to a
higher energy electronic excited state. The energy absorbed E is proportional to the
frequency of the electromagnetic radiation. The energy rises with the shortening of . The
UV-Vis radiation is therefore energetically more effective then e.g. the infrared radiation.
The larger quantity of excitation energy causes the excitation of the electron system that
means the valence electrons. Usually the wavelenght is measured, the unit is the nanometer
(1 nm = 10-9 m = 10 ).
The UV-Vis radiation takes a short part of the region of electromagnetic wave motion in
the range from 10 to 1000 nm. The region between 10 and 200 nm is called the far or
vacuum UV region (absorption of oxygen is removed by evacuation of apparatus there), the
commonly used near UV region is in interval of 200 to 400 nm, the range of 400 to 760 nm
is known as visible (Vis), between 760 and 1000 nm there is the region what is not enough
802
significant for the infrared spectroscopy but some of analytical important compounds absorb
there.
The amount of radiation absorbed is defined by the absorbance A what is the logarithmic
rate of I0 and I the intensities of the incident and transmitted (unabsorbed) radiation,
respectively. According to the Beer-Lambert law the absorbance depends on the
concentration c of absorbing species and the radiation path length l (usually in cm). Also
there is employed the molar extinction coefficient .
In most cases UV-Vis spectra are taken in dilute solutions using quartz (for UV) or glass
(only for Vis) cuvettes. Typical solvents are ethanol, water, acetonitrile or hexane but the
choice of a solvent depends on the desired measuring range. Table 1. shows the transmission
limits of UV radiation for some of them.
Table 1. UV radiation transmission limits of some solvents
solvent
transmission limit of UV radiation
water
200 nm (re-distilled up to 185 nm)
aliphatic hydrocarbons, cyclohexane
195 nm
ether, ethanol, methanol
200 nm
acetonitrile, dioxane
220 nm
glycerol
230 nm
dichlormethane, chloroform
240 nm
benzene, toluene, tetrahydrofuran
275 nm
pyridine
305 nm
acetone
330 nm
carbon disulfide
340 nm
Measurements must be carried out under the same conditions, especially the temperature
is important because the increasing temperature makes the absorption coefficients increasing
too. The coefficient also depends on the type of a solvent.
The electron absorption spectrum mainly shows the changes in external orbitals of
electron covers but contains the changes of vibrational and rotational energies as well their
lines create the smooth structure of a spectrum.
Organic molecules contain bonding orbitals and (usually occupied by electrons
which form the chemical bonds), non-bonding orbitals n (occupied by electrons which do not
participate in chemical bonds) and anti-bonding orbitals * and * (usually unoccupied by
electrons). Their relative energies are in the following order: < < n < * < * . In UV-Vis
absorption processes transitions of electrons from bonding or non-bonding orbitals to antibonding orbitals take place. The lowest energy transition is n * which is responsible for
the weak absorption band of the carbonyl group of acetone at max = 280 nm where the nonbonding electrons of the oxygen atom undergo this transition. A * transition is typical
803
to a double bond. An isolated double bond absorbs strongly at about 190 nm, while
conjugation of them lowers the energy of the transition, resulting in a higher max .
The specific groups in the molecule which absorb in the UV-Vis region are called
chromophores or colour carriers. The identification of a chromophore is often difficult due to
the many factors which affect the spectrum. A structural change in the molecule or a change
of a solvent could shift max either towards a longer wavelength (red shift or bathochromic
effect) or towards a shorter wavelength (blue shift or hypsochromic effect).
The UV-Vis spectra contain broad peaks because the electron transitions are
accompanied by vibrational and rotational changes in the molecules. For identification
purposes such spectra have much less fingerprinting value than IR spectra. UV-Vis
spectroscopy is useful for quantitative analysis better than for identification.
3.
EXPERIMENTAL
For the purpose of instrumental methods research concerning the explosive detection
there was employed the spectrometer AvaSpec-2048 with the wavelength range of 200-1000
nm. The instrument utilizes the fiber optics that is very flexible and allows the usage of as a
probe as a cuvette (or other similar additional devices for the data collection). The
spectrometer has very sensitive CCD detector with 2048 pixels, also contain the additional
collimating lens for the sensitivity enhancement. As a light source there is the combined
deuterium (78 W, 0,75 A) and halogen (5 W, 0,5 A) lamp what supplies a continuous
spectrum with high efficiency and highest stability in the UV, visible and near-infrared range
from 215-2000 nm. The disadvantage of such device is a worse value of the signal/noise
rate.
Fig 1.
A typical setup for absorption measurement
The probe is designed for measurements of powders and thick fluids the reflected light
is measured there. The cuvette (made of quartz) is for solutions and the absorbed radiation is
measured there.
804
The UV-Vis spectra of standards of exlosives were taken using the optical fibre probe.
The standards were of powders or small crystals.
Fig 2.
The detailed scheme of the reflection probe
At first the reference spectrum of the white reference tile was taken as well as the
background spectrum. Next the statically fixed probe was inserted into a sample in a bottle
so the material enclosed the measuring part of the probe in 1-2 mm approximatelly. Both the
deuterium and halogen lamp were employed, every spectrum was taken in average of 500
scans. The resulting spectra of TNT, RDX and PETN are shown in Fig 3., Fig 4. and Fig 5.
1,4
1,2
absorbance
1
0,8
0,6
0,4
0,2
0
200
300
400
500
wavelength /nm/
Fig 3.
UV-Vis spectra of TNT
805
600
700
800
1,4
1,2
1,0
absorbance
0,8
0,6
0,4
0,2
0,0
-0,2
200
300
400
500
600
700
800
600
700
800
wavelength /nm/
Fig 4.
UV-Vis spectra of RDX
1,0
absorbance
0,8
0,6
0,4
0,2
0,0
200
300
400
500
wavelength /nm/
Fig 5.
4.
UV-Vis spectra of PETN
CONCLUSION
In the literature search there were found two interesting papers concerning the utilization
of UV-Vis spectroscopy for characterization of raw energetic materials. The first paper [6]
deals with the spectral behavior of energetic materials after they were melted. The second
one [7] discussed the energetic materials spectra in the solid state. The solid state absorption
spectrum in the range 190 to 550 nm has been obtained for four common high explosives:
HNS, RDX, HMX and PETN. Studies were performed with high explosives crystals
embedded in potassium chloride disks and compared to spectra in acetonitrile solution. It
was observed that the solid state spectra generally have red shifted absorption maxima and
806
extended red edge absorption compared with the solution phase spectra. These effects are
attributed to intermolecular stabilization of the energetic molecules in the solid phase.
Energetic materials of organic origin especially with aromatic structure absorb very
strongly in the UV range. So the signal reaching the detector is thus too small. The error of
measurement is influenced by the adverse signal/noise rate of such type of a spectrometer
what was used for measuring. The low signal reaching the detector increases the error too.
The average of scans was thus put on 500 to obtain better results.
The broad peak of TNT up to 420 nm approximately is influenced by the yellow colour
of the standard. The differencies among the spectra of these three widely used explosives can
be observed.
Acknowledgment:
This research was supported by the Ministry of Industry and Trade, Project FT-TA/048.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
J. YINON and S. ZITRIN: The Analysis of Explosives, Pergamon Press, Oxford, 1981
Z. HOLZBECHER and J. CHUREK: Analytick chemie, SNTL, Praha, 1987
M. HANUS: Zklady analytick chemie vbunin, skripta KTTV, Univerzita Pardubice,
Pardubice, 1999
M. NEPRA and M. TITZ: Zklady teorie elektronovch spekter, SNTL, Praha, 1983
J. YINON and S. ZITRIN: Modern Methods and Applications in Analysis of Explosives, John
Wiley & Sons, Chichester, 1993
M. L. PANTOYA and B. D. SHAW: Molten Salt Destruction of Energetic Materials: Emission
and Absorption Measurements, Journal of Energetic Materials, No. 20, p.001-037, 2002
K. J. SMIT: Ultraviolet and Visible Absorption Spectroscopy of Some Energetic Molecules in
the Solid State, Journal of Energetic Materials, No. 9/2, p.081-103, 1991
807
DIFFERENTIAL SCANNING CALORIMETERS

Renata psov
Explosia a. s., Research Institute for Industrial Chemistry,
Safety Engineering Laboratory, 532 17 Pardubice-Semtn,CZ
Abstract:
In this paper the principles are outlined of differential scanning calorimetry. Main
characteristics that can be determined by means of this method are listed. Examples of
results are presented of thermal stabilities of some explosives.
Keywords:
1.
thermal stability, differential scanning calorimetry, explosives
INTRODUCTION
Calorimetry is the measurement of the heat absorbed or generated in a substance under

study as it undergoes a change from a well-defined initial state to a well-defined final state.
The change of state in question may result from chemical reaction, dissolution or dilution, or
a physical change such as vaporization, sublimation, melting or the processes reverse to
these. It may also result from the changes in temperature and other thermodynamic variables
such as pressure, magnetic field and electric field. Among these the heat capacity
calorimetry by the adiabatic method differs from the others in principle and in practice; in
this type of calorimetry the temperature increment rather than the heat is measured against a
fixed amount of dispensed energy, the result of the measurement being presented as the heat
capacity. The dynamic methods include differential scanning calorimetry (DSC), modulated
DSC, AC calorimetry, heat capacity spectroscopy and relaxation calorimetry. Two timeconstants are involved in these methods. One characterizes the equilibration in the sample
and the other the heat conduction between the sample and its surroundings. The basic
requirement in the use of these methods is that the former time constant is much shorter than
the latter[1].
2.
THEORY
2.1
Differential Scanning Calorimetry
Differential scanning calorimetry is a modern differential thermal analysis in which the

sample compartment of the apparatus is a differential calorimeter, allowing an exact measure
of the heat of transition independent of the specific heat, thermal conductivity and other
variables of the sample. DSC is the technique most widely applied to study any physical or
chemical change that is accompanied by a change in energy. The DSC measurement takes
place in an insulated furnace with two separate compartments. On the laboratory DSC
instrument, a positive deflection on the plot corresponds to an exothermic process and a
negative deflection corresponds to an endothermic one. The use of DSC for the
characterization of the thermal stability of energetic materials has a big advantage and brings
a lot of information[2]. If a sample and reference are heated at the same rate, the heating
powers required are proportional to their respective heat capacities. The sample heat capacity
808
is determined from the heating power and baseline and sensitivity calibrations. Heats of
transitions are determined by integration of the peak in the power vs. time curve
corresponding to the transition[1].
2.2
Definition of Terms
For instance a transition or reaction occurring on the sample yields a peak on the DSC
curve. In general, an exothermic change occurring on the sample is represented by a convex
peak, whereas a concave peak represents an endothermic reaction. The initial peak
temperature, Ti, is the temperature (at time ti) where the measured curve first deviates from
the baseline; the final peak temperature, Tf, is the temperature (at time tf) where the curve
reaches the final baseline; the interpolated baseline is a horizontal straight line between Ti
and Tf (assuming that cp 0); the extrapolated peak onset temperature, Te, is the
temperature (at time te) where the extrapolated ascending peak slope intersects the
interpolated baseline[1].
2.3
Utilization DSC
2.3.1. Glass Transition

The glass transition is a second order thermodynamic process, since there is no change
in enthalpy unlike melting or crystallization, but only a change in the heat capacity i.e. dH/dt
shows a step change and d2H/dT2 goes to infinity.
2.3.2. Crystallization
Crystallization is a first-order transition that appears as a relatively sharp peak in DSC.
Crystallization refers to the formation of solid crystals from a homogeneous solution.
The mechanical properties of polymers are markedly dependent on the degree of
crystallinity and it is important to determine the temperature range over which the
crystallinity changes and affects material properties.
2.3.3. Purity
DSC purity determination is based on the fact that eutectic impurities lower the melting
point of a eutectic system. This effect is described by the Vant Hoff equation.
This method depends only on the physical-chemical behavior of the compound, and no
reference standard is necessary. The pure melting point of the sample does not even need to
be known. The method is accurate for samples over 98% purity, but it does not measure
impurities, which are soluble in the solid phase or insoluble in the melt. This technique is
particularly valuable in the pharmaceutical and basic chemicals industry[4].
2.3.4. Heat capacity
The specific heat capacity of a solid or liquid is defined as the heat required to raise unit
mass of substance by one degree of temperature. When a sample material is subjected to a
linear temperature increase, the rate of heat flow into the sample is proportional to its
instantaneous specific heat. By regarding this rate of heat flow as a function of temperature
and comparing it with a standard material under the same conditions, we can obtain the
specific heat, Cp, as a function of temperature.
809
2.3.5. Reaction kinetic

Reaction kinetics can be studied to determine Arrhenius parameters, such as Activation
Energy (E), the pre-exponential term (Z), rate constant (k) and reaction order (n).
Experiments can be carried out isothermally at temperatures of interest[3].
3.
EXPERIMENTS
The main advantage of DSC method is the possibility to work with minimum quantity of
substance, even tenths of mg. This is very important when explosives and substances very
sensitive and easy to decompose are to be measured. Qualitative and quantitative
measurements were carried out for TNT, Np and RDX. For these substances melting points
and characteristics of high exothermic thermal decomposition were measured. Important
features of thermal decomposition include decomposition temperature or peak temperature,
the width of thermal reaction, or the onset and endset temperatures of the peak and the
enthalpy of thermal decomposition, or area under the peak.
Measurements were performed with about 1 mg sample at various heating rates.
Aluminium crucible was used for measurement. Conversion was performed from the
measured curves and activation energy was calculated.
Fig. 1 shows a series of dynamic DSC curves of TNT measured at different heating
rates. It can be seen that the peak maximum of about 280 C at 2 C/min shifts to about 332
C at 20 C/min (see Tab. 1). In figure 2, conversion curves have been calculated from DSC
curves. Model free kinetics (MFK software) is then used to calculate a value of activation
energy - conversion dependence from the curves.
mW
20 C/min
50
15 C/min
40
10 C/min
5 C/min
30
2 C/min
20
10
0
-10
-20
-30
-40
40
60
80
Fig 1.
100
120
140
160
180
200
220
240
260
280
300
320
DSC curves of the TNT measured at different heating rates.

810
340
360
Table 1. TNT.
Heating rate
2
5
10
15
20
Tm [C]
79.8
79.8
79.6
79.4
80.1
Ti [C]
260.5
271.3
287.4
294.7
297.5
Peak [C]
279.7
296.0
318.3
326.1
332.3
H [J.g-1]
245.44
224.31
314.92
272.59
181.48
kJm ol^-1
2 C/min
Activation Energy
5 C/m in
100
10 C/min
120
15 C/m in
20 C/m in
80
118
116
60
114
40
112
110
20
108
0
106
260
Fig 2.
280
300
320
340
20
40
60
80
Conversion curves calculated from DSC curves (Fig. 1) In the right

coordinate system, the activation energy calculated using MFK SW is
plotted as a function of conversion.
Fig. 3 shows a series of dynamic DSC curves of Np measured at different heating rates.
It can be seen that the peak maximum of about 191 C at 2 C/min shifts to about 215 C at
20 C/min (see Tab. 2). In figure 4, conversion curves have been calculated from DSC
curves. Model free kinetics is then used to calculate a value of activation energy - conversion
dependence from the curves.
811
mW
20
10
20 C/min
15 C/min
-10
10 C/min
5 C/min
2 C/min
-20
-30
120
130
Fig 3.
140
150
160
170
180
190
200
210
220
230
240
DSC curves of Np measured at different heating rates
kJm ol^-1
Activation Energy
205
100
20 C/min
200
15 C/min
10 C/min
80
195
5 C/min
190
60
185
180
40
175
170
20
165
160
160
Fig 4.
180
200
220
20
40
60
80
Conversion curves calculated from DSC curves of Np. In the right

coordinate system, the activation energy calculated using MFK is plotted
as a function of conversion.
812
Table 2. Np.
Heating rate
2
5
10
15
20
Tm [C]
140.8
141.0
140.9
141.1
141.1
Ti [C]
145.6
149.1
153.4
154.9
158.2
H [J.g-1]
910.56
868.97
847.58
830.97
785.39
Peak [C]
191.0
200.2
206.8
212.0
215.3
Fig. 5 shows a series of dynamic DSC curves of RDX measured at different heating
rates. It can be seen that the peak maximum of about 223 C at 2 C/min shifts to about 251
C at 20 C/min (see Tab. 3). In figure 6, conversion curves have been calculated from DSC
curves. Model free kinetics is then used to calculate a value of activation energy - conversion
dependence from the curves.
It can be seen from the results, that melting point is the same for different heating rates.
Melting point is independent of the heating rate. Initial peak temperature increases with
heating rate (see Tab. 1, 2, 3).
mW
30
20 C/min
15 C/min
20
10 C/min
5 C/min
2 C/min
10
-10
-20
180
185
Fig 5.
190
195
200
205
210
215
220
225
230
235
240
245
250
DSC curves of RDX measured at different heating rates.
813
255
260
265
Table 3. RDX
Tm [C]
204.4
204.4
204.7
204.4
204.5
Heating rate
2
5
10
15
20
Ti [C]
205.7
206.5
208.0
209.1
210.6
k Jm ol^-1
2 0 C /m in
5 50
1 5 C /m in
90
H [J.g-1]
1,887.12
1,804.76
1,883.51
2,032.04
2,014.38
Peak [C]
223.1
233.9
242.4
247.5
251.0
A c tivatio n E n e rg y
10 C /m in
5 00
5 C /m in
80
2 C /m in
4 50
70
4 00
60
50
3 50
40
3 00
30
2 50
20
2 00
10
1 50
0
1 80
2 00
Fig 6.
4.
22 0
24 0
260
20
40
60
80
Conversion curves determined from DSC curves (Fig. 5) In the right

coordinate system, the activation energy calculated using MFK SW is
plotted as a function of conversion.
CONCLUSION
Today DSC is the most appropriate method to determine thermal characteristics of an

explosive. This technology allows a lot of information to be obtained in short time
concerning the thermal stability of an explosive and not only the ignition temperature. The
purpose of this paper was not to present the results of determinations of thermal stabilities of
compounds (these are presented as examples only), but experimental method serving to
obtain basis for study of stability.
814
REFERENCES
[1]
[2]
[3]
[4]
MICHIO SORAI: Comprehensive Handbook of Calorimetry and Thermal Analysis, John Wiley
and Sons, Ltd West Sussex , 2004
PATRICK FOLLY: Thermal Stability of Explosives, Chimia 58, Schweizerische Chemische
Gessellschaft, No. 6, p.394-396, 2004
JRGEN SCHAWE: Kinetics studies of complex reactions, Part 1: model free kinetics, User
Come 2/2003.
E.L. CHARSLEY AND S.B. WARRINGTON: Thermal analysis Techniques and Applications, The
Royal Society of Chemistry, Leeds, 1992
815

Siberian State Technological University
660049, Krasnoyarsk, prosp. Mira, 82, Russia
Abstract:
The thermal decomposition of six-membered cyclic N-nitramines, contained in ring one,
two and three nitramine fragments, and also oxy-, gem-dinitro, azido and tetrazole
function, is investigated, using manometer method combined with chromatography,
FTIR-spectroscopy and mass-spectrometry. Influence of functional groups on the rate
and activation parameters of N-nitramines decomposition limited stage is established.
Keywords:
1.
thermal decomposition, structure, conformation, kinetics,

decomposition rate
INTRODUCTION
Nowadays amond the six-membered N-nitramines thermal decomposition of 1,3,5trinitro-1,3,5-triazacyclohexane (RDX) is investigated in most detail. The basic knowledge
on this question are generalized in the reviews [1-5]. Possible mechanisms of RDX thermal
decomposition are analysed here and the conclusion is made, that the primary act of reaction
is homolytic break of N-NO2 bond. The same mechanism is offered for RDX analogues
containing one or two nitramines functions [5]. However the thermal decomposition kinetics
of six-membered N-nitramines, which molecules contain other functional groups, except
nitramines, till now were not considered in the literature practically. In this connection the
thermal decomposition of the compounds (I-VIII) is investigated in this work with the
purpose of revealing influence of functional groups and of molecules conformation on the
rate and activation parameters of a limiting stage.
NO2
N
NO2
N
O
O2N
N NO2
N
N
O2N
NO2
VI
N
NO2
IV
III
II
NO2
O2N
N NO2
N
O
CH2N3
I
NO2
N
O2N
N
N
N
NO2
VII
816
NO2
NO2
N
N
N
N
NO2
V
O2N
O
N
NO2
NO2
VIII
IX
NO2
2.
EXPERIMENTAL
The compounds I-IX are synthesized and purified by known methods, were
chromatography pure and contained 99,6-99,8 % of the basic substance. Research of thermal
decomposition kinetics was carried out in isothermal conditions by a manometric method
with using Bourdon's manometer. Residual air pressure in a reactionary vessel was 10-3-10-2
mm of mercury. Thermal decomposition of compounds I-IX was studied in melt and
solutions of inert solvents, basically, in dibutyl phtalate (DBPh) and dinitrobenzene (DNB).
The substance concentration in a solution from 1,2 up to 10 % (mass)., practically, did not
influence on decomposition rate. Diluted 2 % solutions were used. Thermal decomposition
of compound V was investigated also in a solid phase. The rate constants calculated from
initial rates. The rate constants definition error did not exceed 10 %. An r.m.s. deviation for
activation energy was no more than 6,5 kJ/mole, and for lgA - 0,6 log. units. Decomposition
gaseous products were analyzed using chromatograph (column with sorbent Polysorb-1,
temperature 18-20) and FTIR-spectrometer UR-20, and also mass-spectrometer -1312
(chromatomass, column with sqalane) and I-12-01 (pressure in the chamber 10-5-10-7 mm
of mercury, ionizating voltage 70 eV, accelerating voltage 5 kV).
3.
The researches have shown, that the thermal decomposition of compounds I-IX has
much identical in the kinetic attitude. First of all, it concerns the law of decomposition. The
reaction proceeds with acceleration and characterizes by S-shaped kinetic curves. Thus in a
solution kinetic curves in comparison with melt phase have weak S-shapeness. The brightly
expressed S-shapeness of kinetic curves is observed in a solid state (compound V).
Irrespective of a phase state of nitramines III-V, VIII (melt, solid phase) the reaction does
not complicated by heterogeneous and chain processes, what about one may judge on the
change of S/V (ratio of a surface of a reactionary vessel to its volume). Besides m/V (ratio of
substance mass to reactionary volume) exerts influence upon initial decomposition rate of
nitramines in melt. The more m/V, the more induction period and the less decomposition rate
are. Such influence is connected to the simultaneous contribution of gaseous reaction in
common rate of process [6]. Therefore the temperature influence upon decomposition rate of
nitramines we studied under constant m/V. However for decomposition of nitramine V in a
solid phase the increase of m/V in 20 times, on the contrary, results in reduction of an
induction period and, practically does not influence upon the initial decomposition rate. One
may explain this by proceeding of reaction solely in the condensed phase. The change of
crystals sizes (0,01-0,2 mm) of nitramine V does not influence upon the initial rate and
induction period. Role of topochemistry in acceleration of decomposition is not displayed,
probably. The change of decomposition rate at acceleration stage is, obviously, connected
with autocatalytic action of decomposition products on initial substance.
Kinetic parameters for nitramines are presented in Table 1.
The data of Table 1 allow analyzing the influence of structure of compounds I-XI upon
the rate and kinetic parameters of thermal decomposition. So, compound I has the least
decomposition rate, compound VII the most one. Distinctions in decomposition rates
consist ~ 7 orders. The same picture is observed in activation parameters. The activation
energy in a liquid phase (melt, solution) varies from 101 up to 198 kJ/mole. In our opinion,
the values a = 177-198 kJ/mole are overestimated, probably, because of the disproportional
contribution of gaseous reaction in common rate of process at different temperatures.
817
Preexponential factor changes within the limits of 8,25-16,25. We can note also the
increased decomposition rate of nitramines in melt in comparison with a solution.
As it was mentioned above already, a limiting thermal decomposition stage of compound
VIII (RDX) is homolytic break of N-NO2 bond. This process characterizes in the following
parameters: Ea = 166-176 kJ/mole and lgA = 14,30-15,95 (Table 1).
Table 1. Kinetic parameters of thermal decomposition of cyclic N-nitramines
Formula of
compound
O2 N N
II
III
O2N N
O
O
O
CH2N3
O2N N N NO2
O
IV
O2N N
V
O2 N N
VI
N NO2
N
N
N NO2
O2N NO2
O2N N N NO2
VII
VIII
NO2
N N
N N
N N
NO2
NO2
N
O2N N N NO2
O
IX
O2N N N NO2
N
NO2
X
XI
XII
O2N N N NO2
O2N N
NH
NNO2
NH
O2N NO2
O O
a,
kJ/mol
168,7
174,6
173,9
163,9
162,3
169,2
171,2
13,95
14,46
14,40
14,38
14,29
14,71
14,93
k200C,
s-1
2,110-5
1,510-5
1,610-5
1,910-4
2,310-4
1,110-4
1,110-4
solution in DNB[7] 225-245

melt[7]
216-250
[8]
solution in TNB*
230-245
melt
180-220
solution in DBPh
150-185
solution in DNB
150-180
solid
150-180
solution in DBPh 145-170
solution in DNB
150-175
melt
165-185
solution in NB*
50-70
solution in DBS*
50-75
155,3
198,4
191,7
177,1
148,7
153,2
172,9
157,9
160,4
159,1
123,1
128,8
12,00
17,30
15,70
14,86
13,38
13,91
15,64
14,47
14,78
14,71
16,25
16,87
7,010-6
2,510-5
3,010-6
2,010-5
9,110-4
9,910-4
3,510-4
1,110-3
1,210-3
1,410-3
4,55102
4,45102
solution in DNB[6]
solution in DBPh
melt
solution in DBS*
solution in DNB
160-200
170-210
205-224
160-180
160-180
166,3
167,5
176,3
159,5
162,9
14,30
14,62
15,95
14,53
14,74
8,710-5
1,310-4
3,010-4
8,310-4
5,710-4
solution in EtOH[9]
melt[9]
solution in PhB*[10]
200-300
240-260
140-180
167,4
175,4
100,9
14,60
15,40
8,25
1,310-4
1,110-4
1,310-3
solution in DBPh[11] 195-215

melt[11]
180-200
174,6
176,8
14,75
15,09
3,010-5
3,710-5
Decomposition
conditions
gas phase
solution in DBPh
solution in DNB
solution in DBPh
solution in DNB
solution in DBPh
melt
T,
200-230
185-220
185-225
175-210
180-220
170-210
180-220
lgA
* TNB trinitrobenzene, NB - nitrobenzene, DBS dibuthyl sebacinate, PhB phenyl benzoate
If take into consideration, that the decomposition rate in melt is depending on m/V, so
true parameters for a limiting stage it is necessary to consider the data received in a solution:
Ea = 166,3-171,6 kJmole-1, lgA = 14,30-15,46. It is visible from Table 1, that the majority of
818
the investigated compounds in a solution have the similar kinetic parameters within the
limits of an experience error.
The analysis of thermal decomposition products of compound II in melt has shown, that
while decomposition degree increases the intensity of an absorption peak of azide group
(2140 cm-1) in FTIR-spectra decreases. At the same time intensity of peaks of N-NO2 group
(1555 cm-1, 1295 cm-1), practically, did not change. At transformation degree 1-3 % by
mass-spectroscopy method only molecular nitrogen and 5-nitro-5-aza-1,3-dioxa-2methyleneiminecyclohexane were found out. Nitrogen dioxide has appeared more late (at
transformation degree 7-12 %), when the composition of products became rather
complicated. This fact testifies, that the decomposition of compound II begins from
homolysis of N-N bond in azidogroup.
Change of reactionary centre may be expected also under decomposition of compound
VI which has potentially reactive dinitromethyl group besides secondary nitramine group. It
is visible in Table 1, that dinitromethyl group in a cycle (compound XII) is close to
nitramine one in stablility (compound I). Transition from compound X to compound VIII
changes decomposition rate insignificantly. At the same time for compound VI
decomposition rate increases for the order, that may be tied with steric influence of
dinitromethyl group (Es = 3,22) on on of nitramine groups [12]. According to X-ray data
nitramine groups in compound VI are unequvalent, the length of less stable N-NO2bond is
1.408 , but the sum of valence angles at amine nitrogen is 349.7 [12]. The similar steric
influense was observed if replace one methylgroup in dimethyl-N-nitroamine on 2,2dinitropropyl group [13]. Thus, one can certify, that decomposition of compound VI, as well
as in RDX, is limiting by homolysis of N-NO2 bond.
The qualitative composition of gaseous decomposition products, found by FTIR- and
mass-spectrometry method for compounds I, III, VIII and IX, is, practically, identical and
contains N2, N2O, NO, CH2O, H2O, CO, CO2 and HCN (traces). Quantitative change of
basic nitrogen-containing gaseous products N2, N2O, NO, and also CO and CO2 for
compound III in melt are submitted in Table 2.
Table 2. Composition of basic gaseous decomposition products of nitramines III
and VIII at different temperatures (mole per mole of substance, rel.102)
Nitramine
III
VIII
Temperature, Decomposition
degree, %
180
5
190
5
190
20
190
30
210
98
210
98
NO
N2O
0,06
0,27
1,61
6,14
1,78
2,66
0,07
0,28
2,95
9,52
3,73
6,31
N2
0,12
0,54
3,88
9,03
10,00
13,2
CO
CO2
0,08
0,28
2,49
8,06
5,55
4,21
0,07
0,16
1,87
3,68
7,81
4,98
It follows from the analysis of the data of Table 2, that at temperature 190C the increase
of transformation degree of compound III leads to decrease of ratio NO/N2O from 0,97 up to
0,64, N2/N2O from 1,93 up to 0,95, but NO/N2 is increased from 0,5 up to 0,68 and
CO/CO2 from 1,8 up to 2,2.
When the temperature falls up to 180C the ratio NO/N2O decreases insignificantly from
0,97 up to 0,86 and N2/N2O from 1,93 up to 1,8. But the ratio /CO2 sharply decreases
819
from 1,8 up to 1,1, that testifies about decrease of the rate of oxidation-reduction processes
which are proceeding with NO2 participation. If compare the ratio /CO2 for compounds
III and VIII, one may see, that it is hardly more for compound VIII (0,85), than for
compound III (0,71).
It was not possible to find out NO2 in the gaseous decomposition products of compound
III, despite of using of a high-sensitivity photoelectrocolorimeter method. However, it is
possible, that NO2 formed in the primary act of decomposition enters into oxidationreduction reaction more fast, than it is possible to find out it [5].
It is visible from Table 1, that the compounds III, VIII and X have in a solution,
practically, the same decomposition rate, despite of, apparently, their different chemical
structure. The decomposition rate of six-membered cyclic nitramines, in the main, depends
upon conformation of nitramine group. According to neutron-diffraction data [14] in crystal
structure of RDX nitramine groups are unequivalent: two of them have pyramidal atoms of
amine nitrogen and third has plane one. Such nonequivalence is caused by influence of
crystal structure and conducts to marked difference in the length and, consequently, strength
of NNO2 bonds. The lengths of the less stable NNO2 bonds consists 1,398 . The sum of
valent angles at amine nitrogen atom is 347.9. Over the electron-diffraction data [15] in RDX
molecule in gas phase all nitramine groups are equivalent and their geometry is similar to
plane (the sum of valent angles at amine nitrogen atom is 356.3). Length of NNO2 bonds
increases up to 1.413 . In solution the intermediate situation among strong influence of
intramolecular interaction in crystal state and its absent in gas phase is realized. If to judge
on value of the dipole moments of compounds VIII and III, which are equal to 5,78 and 5,64,
accordingly [16], these nitramines in a solution must have equvalent nitramine groups. As to
conformation of amine nitrogen atom, it has, probably, more pyramidal structure than in gas
phase. This conclusion follows from comparing of thermal decomposition rate of RDX,
nitramines III, X and compound IV.
Among investigated nitramines compound IV is most stable. Its thermal decomposition
rate is 1-2 orders less than for its isomer compound X. In crystal state compound IV has chair
conformation and two pyramidal nitramine groups (sum of valent angles at amine nitrogen
atom is 347). However, the length of NN bond 1.37 [17] in compound IV is the less than
in RDX. Its possible that the length of NN bond in compound IV in solution is the less than
at other considered here nitramines. It explains its more high thermal stability in compare with
other nitramines.
Unfortunately, there are no data on structure of compound I. However, value of its
decomposition rate allow to assume, that the length of NN bond in compound I is the less
than in RDX.
Compound IX contains flattened conjugated dinitroureido fragment. Its conformation is
half-chair (half-boat) [18]. The length of all NN bonds (the most value is 1.434 ) are more,
than in RDX. All this explains less thermal stability of compound IX in comparison with
RDX. The same time in work [19] composition of decomposition products of compound IX
(mainly CH2O, N2O and CO2 with little amount of HCN, CO, NO and NO2) allowed to
establish different from other cyclic nitramines decomposition mechanism. It initiates by
attack of one of the NO2 oxygen atoms on the keto carbon atom, forming four-centre C-N-NO complex, followed by initial elimination of CO2 and rapid transformation of remaining
fragment into CH2O and N2O.
The decomposition rate of compound V in a solid phase and solution, as a whole, is
more, than at RDX and is close to those for compound IX (Table 1). The reason of it,
probably, consists in its conformation. In DPT crystal conformation chair - chair is realized
820
[19]
. The angles between planes of nitrogroups and CNC of a cycle are, accordingly, 34 and
26, as at RDX. The angles CNC in six-membered cycles are equal to 113-118, i.e. are
close to those in RDX, but it is less, than in a molecule HMX. The structure of compound V
is more rigid, than cyclooctane structure, i.e. conversion of cycles is complicated [20].
Conformation of a type an chair - boat and boat - boat energetically are considerably less
favourable because of instability.
At the same time analysis gaseous products of DPT at different temperatures (Table 3)
show, that the basic component is N2O, and NO is absent.
Table 3. Composition of noncondensing gases at thermal decomposition of DPT
in solid phase (mol. %)
Temperature,
C
160
160
160
160
160
150
150
170
Decomposition
degree, %
10
20
50
75
100
10
20
100
N2
0,98
0,69
0,64
0,66
0,64
0,68
0,46
0,95
Composition of decomposition products

N2O
CO
CO2
96,19
0,90
1,93
96,47
0,83
2,01
95,57
1,09
2,69
93,94
1,29
4,11
91,87
2,39
5,10
96,32
0,83
2,18
96,55
0,76
2,23
88,22
2,87
7,94
The ratio N2/N2O is approximately for two order less, than at compounds III and VIII.
Taking into account above-stated, its possible to expect, that thermal decomposition of DPT
proceeds not homolytically with break of N-NO2 bond, but on the mechanism including
inter- or intramolecular carry of nitrogroups oxygen on the carbon atom of methylene
group. This is promoted by presence of many short intermolecular contacts [20] in DPT
crystal structure. In particular, distances between oxygen atoms and carbon atom of
methylene group are equal 2,62-2,69 , that is less than sum of Van der Vaals radiuses [21]. It
allows to assume the intermolecular attack of oxygen atom on carbon atom of methylene
group [22] with the subsequent break of C-N bond with formation of CH2O and N2O. The
CH2O presence in decomposition products of DPT is proved by mass-spectrometer and
chemical-analytical (dimedon) methods. The reality of this mechanism is supporting by the
absence of NO in decomposition products and presence of small amounts CO and CO2
(Table 3), which are inevitable in moderate amounts in oxidation-reduction reactions with
NO2 participation.
Besides, as the molecule DPT contains two tertiary amine function, the display of acidbasic catalysis [23] already at initial stages of thermal decomposition is possible.
The experiences on decomposition of DPT in a solution of DBPh and DNB have shown
rate increase with addition of 0,3-0,6 % water (from a solvent mass). Thus not only growth
of rate occurs, but also change of the law of decomposition from autocatalytical to first order
reaction.
So, one can consume, that the DPT decomposition begins from initial homolysis of NNO2 bond. Formed NO2 vigorously oxidizes intermediate decomposition products and initial
DPT, generating water. The occurrence of water includes acid-basic catalysis It is
impossible to exclude also autocatalysis from the very beginning of decomposition, because
the crystals DPT contain a moisture seized with crystallization[24].
821
Thermal decomposition of compound XI proceeds with the same rate, as compound VI

(Table 1). According to X-ray data the length of N-N bond in secondary nitraminogroup of
this compound is 1.409 , and sum of valent angles at amine nitrogen atom is 337.6.
However at the same time with primary homolysis of N-NO2 bond, for compound X the
decomposition upon nitrimine group [10] is possible. In this favour low activation parameters
and considerable amount of N2O in decomposition products testify.
The largest decomposition rate compound VII. In compare with other nitramines, whose
structure data are known, one of N-NO2 bond of this compound has a record value of length
(1.494 ) [25] (nitramine group which pyramidal amine nitrogen atom is directly tied with
one of nitrogen atom of tetrazole cycle). As NO2 is primary found out decomposition
product of compound VII in solid state, authors of work [25] suggest that homolysis of NNO2 bond limits whole decomposition process. This mechanism of decomposition can be
offered and for our case in a solution.
4.
CONCLUSION
General conformities of influence of different functional groups upon mechanism of

limiting stage of decomposition and its activation parameters in six-membered N-nitramines.
Combination of secondary nitramine group with oxo-, gem-dinitro- and tetrazole groups does
not change the limiting stage homoltsis of N-NO2 bond, but increases decomposition rate
for 2-7-orders. Activation energy and preexponential factor in the Arrhenius equation vary
within the limits of 166,3-171,6 kJ/mole, lgA = 14,30-15,46. In a case of tetrazole Ea =
123,1-128,8 kJ/mole, lgA = 16,25-16,87. The combination of nitraminogroup with azido-,
carbonyl and nitroquanyl groups can change limiting stage.
822
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
F. I. DUBOVITSKII, B. L. KORSOUNSKII: Thermal decomposition kinetics of N-nitro

compounds, Usp. Khim., Vol. 50, No. 10, p. 1828-1871, 1981 (In Russian)
T. L. BOGGS: The thermal behavior of cyclotrimethylenetrinitramine (RDX) and
cyclotetramethylenetetranitramine (HMX), Fundamental of solid-propellant combustion,
K.Kuo and Summerfield (Eds). New York, AIAA Inc., p. 121-175, (Progress Astronaut and
Aeronaut; Vol. 90), 1984
Y. OYUMI, T. B. BRILL: Thermal decomposition of energetic materials. 3. A high-rate, in situ,
FTIR, study of the thermolysis of RDX and HMX with pressure and heating rate as variables,
Combust. And Flame, Vol.62, p. 213-224, 1985
N. E. ERMOLIN, V. E. ZARKO: Mechanism and kinetics of thermal decomposition of the cyclic
nitramines, Fiz. Goreniya i Vzryva, Vol. 33, No. 3, p.10-31, 1997 (In Russian)
YU SHU, B. L. KORSOUNSKII, G. M. NAZIN: Mechanism of thermal decomposition of secondary
nitramines, Usp. Khim., Vol. 73, No. 3, p.320-335, 2004 (In Russian)
YU. YA. MAKSIMOV: Thermal decomposition of hexogen and oktogen, (Theory of
explosives,Proc. D. I. Mendeleev MchTI, Editted by Andreev K. K., Vysshaya Shkola),
Moscow, Is. 53, p. 73-84, 1967 (In Russian)
G. V. SITONONA, B. L. KORSOUNSKII, N. F. PYATAKOV, V. G. SHVAIKO, I. SH.
ABDRAKHMANOV, F. I. DUBOVITSKII: Thermal decomposition kinetics
of N,Ndinitropiperazine and 1,3-dinitro-1,3-diazacyclopenthane, Izvestiya AN SSSR, Ser. Chem.,
No. 2, p. 311-314, 1979 (In Russian)
B. A. LURIE, V. N. IVAKHOV: Thermal decomposition of 1,4-dinitro-1,4-diazacyclohexane,
Proc. D. I. Mendeleev MchTI, Moscow , Vol. 104, p. 12-17, 1979 (In Russian)
J. C. OXLEY, A. B. KOOH, R. SZEKERES, W. ZHENG: Mechanisms of nitramine thermolysis, J.
Phys. Chem. , Vol. 98. p. 7004-7008, 1994.
A. A. NEFEDOV: Thermal decomposition derivatives of nitroguanidines, Diss., Siberian State
Technological University, 2004.
R. S. STEPANOV, L. A. KRUGLYAKOVA, A. M. ASTACHOV, O. A. GOLUBTSOVA: Kinetics and
mechanism of thermal decomposition of 2-substituted 5,5-dinitro-1,3-dioxanes, Russ. J.
General Chem., Vol. 74, Is. 10, p. 1702-1705, 2004 (In Russian)
Y. OYUMI, T. B. BRILL, A. L. RHEINGOLD, T. M. HALLER: Crystal structure and molecular
dynamics of the energetic nitramine 1,3,3,5-tetranitrohexahydropyrimidine and a comparison
with 1,3,3,5,7,7-hexanitro-1,5-diazacyclooctane and 1,3,3-trinitroazetidine J. Phys. Chem.,
Vol. 89, N 20, p. 43174324, 1985
R. S. STEPANOV, L. A. KRUGLYAKOVA, A. M. ASTACHOV: Thermal decomposition of gemdinitroethylnitroamines, Russ. J. General Chem., Vol. 74, Is. 10, p. 1669-1673, 2004 (In
Russian)
C. S. CHOI, E. PRINCE: The crystal structure of cyclotrimethylene-trinitramine, Acta
Crystallogr., Vol. B 28, p. 2857-2862, 1972
I. F. SHISHKOV, L. V. VILKOV, M. KOLONITS, B. ROZSONDAI: The molecular geometries of
some cyclic nitramines in the gas phase, Struct. Chem., Vol. 2., No. 1, p. 5764, 1991
J. F. WRIGHT: Methods of introduction of nitraminogroup, its properties and reactions,
(Chemistry of nitro- and nitrosogroup, Editted by G. Foyer, Mir), Moscow, Part 1, p. 516,
1972 (In Russian)
R. GILARDI: Privat communication, Cambridge Structural Database, April 2004 release (about
database: F. N. ALLEN: The Cambridge Structural Database: A quarter of a million crystal
structures and rising. Acta Cryst. B58. No. 2, p. 380388, 2002)
823
[18] J. L. ANDERSON, C. GEORGE: Structures of 1,3,5-trinitro-2-oxo-1,3,5-triazayclohexane and

1,4-dinitro-2,5-dioxo-1,4-diazacyclohexane, Acta Crystallogr., Vol. C 46, No. 4, p. 706-708,
1990
[19] S. BULUSU, R. BEHRENS: A Review of thermal decomposition pathways in RDX, HMX and
other closely rlated cyclic nitramine, Defence Science Journal, Vjl. 46, No. 5, p. 347-360, 1996
[20] C. S. CHOI, S. BULUSU: The crystal structure of dinitropentamethylenetetramine, Acta
Crystallogr., Vol. B 30, No. 6, p. 1576-1580, 1974
[21] YU. V. ZEFIROV, P. M. ZORKII: New using of Van der Vaals radiuses in chemistry, Usp. Khim.,
Vol. 14, No. 5, p. 446-461, 1995 (In Russian)
[22] A. I. B. ROBERTSON: The thermal decomposition of explosives. Part II.
Cyclotrimethylenetrinitramine and cyclotetramethylenetetranitramin, Trans. Faraday Soc.,
Vol.45, p. 85-93, 1949
[23] A. P. COONEY, M. R. CRAMPTON, J. K. SERANAGE, R. COLDING: Kinetic studies of the pH
dependence of the decomposition of 3,7-dinitro-1,3,5,7-tetraazabicyclo[3,3,1]nonane (DPT)
and related compounds, J. Chem. Soc. Perkin Trans., No. 2, p.77-81, 1989
[24] V. L. ZBARSKII, YU. A. MAKSIMOV, E. YU. ORLOVA: About dependence of purification
method upon the thermal stability of oktogen, (Theory of explosives, Vysshaya Shkola),
Moscow, Is. 53, p. 84-92, 1967 (In Russian)
[25] A. GAO, A. I. RHEINGOLD, T. B. BRILL: Thermal decomposition of energetic materials. 47. A
trigger linkage study of high-nitrogen content nitraminotetrazoles and nitramino-1,2,4triazoles. Propellants, Explos. Pyrotechn., Vol.16, p.97-104, 1991
824
RECORDING OF PARTICLE VELOCITY PROFILES IN SHOCK

AND DETONATION WAVES BY ELECTROMAGNETIC TECHNIQUE
N.P. Taibinov, B.G. Loboiko, V.P. Filin, .V. Kostitsin, V.V. Shaposhnikov,
S.N. Lubyatinsky, .B. Smirnov, .B. Syrtsov, .V. Vershinin, and .. Nikulin
Zababakhin Russian Federal Nuclear Centre - VNIITF,
P.O. Box 245, Snezhinsk, Chelyabinsk region, 456770 Russia
Abstract:
The electrocontact and photochronographic procedures were used to study detonation
propagation in cylindrical samples of different diameter made of insensitive HE for the
cases of single-point and plane-wave initiation. The diameter of samples ranged from 15
to 120mm. The stationary detonation rate vs diameter of the cylindrical sample was
constructed. The relation between the curvature of the detonation front and its rate is
identified
Keyword:
electromagnetic gauges, shock wave, TATB, IHE,
First electromagnetic technique of recording of particle velocity in shock waves was

applied in Russia (.K. Zavoisky, 1948).
Operation of electromagnetic gauges (EMG) of particle velocity is based on a simple
physical principle (Figure 1). When the conductor moves in the magnetic field normal to
induction vector B in it, according to the law of electromagnetic induction, the
electromotive force (EMF) =Uh is induced, where is EMF, U is the conductor
velocity, B is induction of the magnetic field; h is the length of the conductor crossing the
lines of the magnetic field.
h
r r
U u
Shock wave
R
Fig 1.
Principle of EMG operation
825
Thus, conductor movement velocity as function of time U (t) can be experimentally

found. Then after shock wave (SW) passage, if the conductor is made of rather thin foil, its
velocity will be close to particle velocity of substance behind SW front, repeating its
variation in time, i.. U(t)u(t).
Currently this technique is widely applied in multi-channel version in studies of
detonation initiation and parameters of phase transitions as well as of chemical
transformations in shock waves [1-3]. Channeling allows recording of evolution of particle
velocity profile during the whole process of initiation in one experiment. Such data is a
valuable information concerning the nature of detonation growth, necessary for simulation of
these processes.
As a rule, application of electromagnetic technique is related to application of
electromagnet with dc source of several hundreds of amperes, in order to make a magnetic
field of ~700-1200 GS with a required degree of homogeneity. Such electromagnet is a
stationary installation having rather strict limitation of HE charge mass, critical to
environmental conditions. Feed current stabilization and neutralization of temperature drift
at heating of electromagnet windings are required for it.
Improvement of electromagnetic procedure where magnetic field is generated using
Helmholtz coils is performed in RFNC - VNIITF.
Helmholtz coils is a system of two equal coils, using them one can get homogeneous
magnetic field in the inter-coil space, if the following conditions are met: the distance
between coils should be equal to their average radius, and the dimensions (thickness and
width) of windings are many times less than coils diameter. Value of magnetic-field intensity
of such system of coils is unambiguously determined by their current.
Helmholtz coils are rather simple device allowing feed from pulsed current source. As
the coils are to be destroyed during experiment, requirements to coils protection are reduced
and there are no practical problems with increase of dimensions of examined HE samples
when needed (that is especially important for studies of insensitive HE). Moreover,
parameters of Helmholtz coils are easily adjusted following the requirements of specific
experiment.
Some results of recording particle velocity profiles with application of electromagnetic
technique on the basis of Helmholtz coils are given in the paper.
The scheme of the experimental device is given in Figure 2.
826
Generator of a plane
shock wave
Glass shield
Helmholtz coils
Electromagnetic
gauge
Examined sample
Fig 2.
Experimental device
Pulsed current feed to Helmholtz coils was applied. Magnetic field with induction of
~1000 G during ~300 s was generated in the working area. Inhomogeneity of the magnetic
field did not exceed 0.5 % during this time.
Diameter of the examined samples was 60 mm. The examined samples were loaded by
the plane shock wave.
Helmholtz coils were preserved from premature destruction by the air shock wave by
introduction of glass shields between them and HE charge.
The unit of electromagnetic gauges (EMG) is located between two parts of the
examined sample at the angle of 30 to the plane of shock wave front. Multi-channel EMG
with various number of active components of particle velocity gauges were applied. The
gauge for measuring of wave velocity in the form of meander with a constant step was
introduced into the design of some EMG.
One of the versions of EMG units is given in Figure 3.
827
Fig 3.
EMG unit view
Profiles of particle velocity U (t) in teflon, recorded using four-element EMG are
given in Figure 4. For comparison, similar results obtained using the same gauge at
electromagnetic installation with static magnetic field with induction of ~500 G are given in
Figure 5.
Fig 4.
Profiles U(t) in teflon. Helmholtz coils
828
Fig 5.
Profiles U(t) in teflon. Static magnetic field
Figures 6 and 7 show profiles U (t) of the shock wave in teflon and stationary
detonation wave in insensitive HE based on TATB (IHE) in comparison with the similar
profiles obtained by recalculation of data, obtained using manganin gauge (MG) and ratio
P=0DU, where P is pressure, 0 is initial density, U is particle velocity.
Strictly speaking, this recalculation is valid only for rather small area in the
neighborhood of shock front. More exact analysis using differential equations in Lagrangian
coordinates is required for correct coupling of pressure profiles and particle velocity.
However, as it has been illustrated in reference [4], application of this recalculation results in
the differences, which are not exceeding experimental error.
Fig 6.
U(t) profiles in teflon
829
Fig 7.
U(t) profiles in IHE
Experimental results illustrate that profiles of particle velocity obtained using

electromagnetic technique with Helmholtz coils are identical to the profiles obtained in
experiment with a static magnetic field (Figures 4, 5), and also to the profiles obtained using
MS (Figures 6, 7).
Application of wave velocity gauges together with EMG in some experiments allowed
recording shock front movement diagram simultaneously with profiles of particle velocity
(x-t) in one experiment. Typical signal from wave velocity gauge is given in Figure 8.
Fig 8.
Signal from wave velocity gauge
830
Figure 9 gives the results of research of initiation of detonation in IHE by a shock wave
of near-critical intensity (approximately half of Chapman-Jouguet pressure).
Fig 9.
U(t) profiles in IHE. Initiation of detonation
It is obvious from figure 9, that slow build-up of particle velocity in the initial site of
detonation growth is observed at this shock-wave impact in IHE.
Thus, application of Helmholtz coils with pulsed generation of magnetic field allows
studies of evolution of profiles of particle velocity using rather simple experimental devices,
without special requirements in terms of protection of electromagnet and environmental
conditions, practically without limitation of the dimensions of the examined HE samples.
REFERENCES
[1]
[2]
[3]
[4]
P.. YRTEV, L.M. ERIKSON, B. HEIS, M.. PARKER: FGV, #5, p. 113, 1986.
J.E. VORTMAN, G. ANDREWS, J. WACKERLE: Reaction rate from electromagnetic gauge data.
8th Symposium (Int.) on Detonation (1985).
S.A. SHEFFIELD, R.L. GUSTAVSEN, L.G. HILL, R.R. ALCON: Electromagnetic gauge
measurements of shock initiating PBX9501 and PBX9502 explosives, 11th Symposium (Int.)
on Detonation (1998).
V.S. TROFIMOV, G.P. TROFIMOVA: About opportunity of decomposition of cast TNT in shock
jump, FGV, V.16, #2, p.p. 92-99, 1980.
831
THE USE OF THE GAP TEST TO INVESTIGATE

THE SHOCK TO DETONATION TRANSITION
IN LOW-SENSITIVITY EXPLOSIVES
PART I EXPERIMENTAL APPROACH
Waldemar A. Trzciski, Marcin Kutkiewicz, and Leszek Szymaczyk
Abstract:
In the present work an attempt was undertaken to apply the gap test configuration for
investigation of the process of buildup of detonation in low-sensitivity explosives. The
set-up was proposed in which sensitivity on shock wave and shock wave velocity can be
measured simultaneously. The detonation initiation was investigated in the compositions
containing 3-nitro-1,2,4-triazol-5-on (NTO).
Keywords:
1.
insensitive munitions, initiation of detonation
INTRODUCTION
A wedge test is the main method of experimental investigation of the shock wave
initiation sensitivity of explosives and determination of a run-to-detonation distance, i.e, the
distance inside the explosive at which the shock wave is transformed into a full detonation
wave[1-3]. However, the run-to-detonation distance can be also determined in the gap test
configuration[4-5]. Mainly, the gap test enables the determination of the minimum shock wave
pressure that can cause complete detonation of the tested explosives[6-7].
New TNT/RDX/NTO and HMX/NTO formulations were successfully manufactured and
characterized in Ref. [8-9]. Detonation performance and energetic characteristics of these
mixtures were determined. The shock sensitivity of tested formulations was also estimated in
the gap test.
In this work, an attempt is undertaken to use the gap test configuration for investigation
of the process of buildup of detonation in NTO-based explosives.
2.
COMPOSITIONS
To investigate the process of shock-to-detonation transition in low-sensitivity

explosives, four formulations were chosen, in which 3-nitro-1,2,4-triazol-5-on (NTO) was
included. The following explosive compositions were tested:
NTO/RDX/TNT 30/40/30 (denoted by C30)
NTO/RDX/TNT 50/20/30 (C50)
NTO/HMX/Viton 30/66.5/3.5 (CH30)
NTO/HMX/Viton 50/47.5/2.5 (CH50)
The density and detonation velocity of the mixtures investigated are given in Table 1[8-9].
832
Table 1. Density (0) and detonation velocity(D) of the mixtures tested
C30
C50
CH30
0
[kg/m3]
1717
1738
1810
D
[m/s]
7850
7730
8580
CH50
1825
8450
Explosive
3.
EXPERIMENTAL ARRANGEMENT
The gap test configuration is shown in Fig. 1. A booster made of phlegmatized RDX of
the density of 1.64 g/cm3 served as a shock wave generator. The booster high was 50 mm, its
diameter was also 50 mm. The explosive tested was placed inside a cooper tube with internal
diameter of 25 mm and wall thickness of 2.5 mm. Holes were made in the explosive charge
and copper tube. The short-circuit sensors were placed inside the holes at distances of 0, 20,
40 and 60 mm from the end of an inert attenuator. They enable us to measure time intervals
corresponding to the passage of the shock wave. The inert attenuator of 71 mm in diameter
was mounted from synthetic polyurethane resin with ceramic flour of the density of
1600 kg/m3. The length of resin attenuator was chosen in a manner making possible the
observation of complete detonation and failure of explosion process. The complete
detonation of explosive charge was indicated when a clean hole was cut in the steel witness
plate.
1
2
3
4
5
6
7
8
9
Fig 1.
Scheme of gap test set-up: 1 fuse, 2 holder, 3 booster (donor), 4

attenuator, 5,6 sensors, 7 copper tube, 8 tested explosive, 9 steel plate
833
4.
RESULTS
A time interval needed by the shock wave to travel a distance between two short-circuit
sensors through the explosive was measured by an electronic counter. Thus, the shock wave
velocity is determined as a ratio between the distance traveled and the corresponding time
interval. The results obtained in the gap test configuration for composition K30 are shown in
Figs. 2-3 as the time-dependence of the distance traveled or the dependence of the mean
velocity of shock wave on the distance. The value of mean velocity is shown as the point
corresponding to the centre of the course.
70
Distance [mm]
60
50
40
30
48 mm
49 mm
51 mm
54 mm
20
10
0
0
20
40
60
80
Time [s]
Fig 2. Gap test time-space diagram of shock wave acceleration for composition K30
10000
48 mm
49 mm
51 mm
54 mm
Mean velocity [m/s]
8000
6000
4000
2000
10
20
30
40
50
60
Distance [mm]
Fig 3. Mean shock wave velocity as a function of the distance traveled for K30
The shock-to-detonation transition in the composition K30 sample is observed only for
the gap length of 48 mm. This process does not occur for the resin blocks of 49, 51 or
54 mm. For these lengths, a copper tube was partially fragmented and there was not a hole in
the steel witness plate (Fig. 4).
834
Fig 4.
Photo of copper tubes recovered from gap tests with 49, 51 and 54 mm resin
length for composition K30
The shock-to-detonation behaviour of other compositions is illustrated in Figs. 5-10.

70
Distance [mm]
60
50
40
30
45 mm
46 mm
48 mm
20
10
0
0
20
40
60
80
Time [s]
Fig 5.
Gap test time-space diagram of shock wave acceleration for composition K50
Mean velocity [m/s]
10000
8000
6000
45 mm
46 mm
48 mm
4000
2000
0
0
10
20
30
40
50
60
Distance [mm]
Fig 6.
Mean shock wave velocity as a function of the distance traveled for K50
835
70
Distance [mm]
60
52 mm
57 mm
62 mm
50
40
30
20
10
0
0
20
40
60
80
100
Time [s]
Fig 7.
Gap test time-space diagram of shock wave acceleration for composition KH30
10000
52 mm
57 mm
62 mm
Mean velocity [m/s]
8000
6000
4000
2000
0
0
10
20
30
40
50
60
Distance [mm]
Fig 8.
Mean shock wave velocity as a function of the distance traveled for KH30
70
Distance [mm]
60
50
40
30
50 mm
55 mm
57 mm
20
10
0
0
20
40
60
80
100
Time [s]
Fig 9.
Gap test time-space diagram of shock wave acceleration for composition KH30
836
Mean velocity [m/s]
10000
50 mm
55 mm
57 mm
8000
6000
4000
2000
0
0
10
20
30
40
50
60
Distance [mm]
Fig 10. Mean shock wave velocity as a function of the distance traveled for KH30
From the results presented in Figs. 2-3 and 5-10 it follows that the run-to-detonation
distance for tested low-sensitivity explosives is relatively long and it is contained within a
range of 20 to 40 mm. A sustained pressure pulse estimated for the resin block boundary
(Fig. 1) has the amplitude of about 20 GPa and it is reduced considerably by the resin layer.
For comparison, the run distance to detonation of RDX-based formulations changed from 2
to 20 mm for the pressure pulse of 5 to 9 GPa and duration about 2.5 s at the boundary of
tested explosive [2].
5.
SUMMARY
Analysis of the experimental results presented in the work enable us to formulate the
following inferences.
1. The gap test configuration can be applied to investigate the process of buildup of
detonation in low-sensitivity explosives. Obtained results enable us to determine
the mean velocity of shock wave propagation at different depths in the explosive
charge. To increase the accuracy of the method, the measuring course should be
decrease. But applying greater amount of sensors can influence the process of
shock initiation. So, no invasion method of measuring the shock wave velocity
should be found.
2. To initiate the detonation process in tested mixtures, the shock wave with
appropriate velocity must be generated in the explosive charge. The wave with
lower velocity is diminished by the lateral rarefaction waves and the detonation
process falls.
3. The run-to-detonation distance for low-sensitivity explosives containing NTO is
relatively long for shock pressure pulse slightly higher than a threshold one.
837
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
J. WACKERLE, J. O. JOHNSON, P. M. HALLECK, Shock initiation of high-density PETN, Sixth

Symposium (International) on Detonation, White Oak 1976.
H. MOULARD, J. W. KURY, A. DELCLOS, The effect of RDX particle size on the shock sensitivity
of cast PBX formulations, Eighth Symposium (International) on Detonation, Albuquerque
1985.
M. QUIDOT, J. GROUX, Reactive modeling in shock initiation of heterogeneous explosives,
Ninth Symposium (International) on Detonation, Portland 1989.
S. A. AUBERT, et al., Gap test as a method of discriminating shock sensitivity, Ninth
Symposium (International) on Detonation, Portland 1989.
S. CUDZIO, A. MARANDA, J. NOWACZEWSKI, W. TRZCISKI, Shock initiation studies of
ammonium nitrate explosives, Combustion and Flame, 1002, 64-72 (1995).
M. SUCESKA, Test methods for explosives, Springer-Verlag, New York 1995.
Explosives, shock sensitivity tests, STANAG 4488 (Edition 1, 7/15/98).
W. A. TRZCISKI, L. SZYMACZYK, S. CUDZIO: Detonation characteristics of low-sensitivity
NTO-based explosives, 6th International Seminar New Trends in research of energetic
Materials, Pardubice 2003.
W. A. TRZCISKI, L. SZYMACZYK, Detonation performances of low-sensitivity NTO/HMX
explosives, 7th International Seminar New Trends in research of energetic Materials,
Pardubice 2004.
838
THE USE OF THE GAP TEST TO INVESTIGATE

THE SHOCK TO DETONATION TRANSITION
IN LOW-SENSITIVITY EXPLOSIVES
PART II NUMERICAL SIMULATION
Waldemar A. Trzciski
Abstract:
In the present work the process of shock to detonation transition in the gap test
configuration was modelled. The reacting explosive was approximated by a twocomponent gas-solid medium. To describe the rate of chemical transformation of the
explosive, the three-term macrokinetic equation was applied. The results of modelling
were compared with the gap test data.
Keywords:
1.
insensitive munitions, initiation of detonation
INTRODUCTION
In Ref [1], an attempt was undertaken to use the gap test system for investigation of the
process of buildup of detonation in NTO-based explosives. In this work, the process of
shock to detonation transformation in the gap test configuration is studied numerically. A
theoretical model describing the processes of shock initiation and buildup of detonation is
described. This model was earlier examined by applying in the so called wedge test [2]. In
this work the model was modified to study the detonation proceeding in the system
corresponding to the set-up used in the gap test. The results of modelling are compared with
the gap test data presented in Ref. [1].
2.
THEORETICAL MODEL
The gap test configuration used in Ref. [1] can be modelled by the one-dimensional layer
system presented in Fig. 1. At the moment when the detonation wave in the booster charge
reaches the boundary of the resin block, one-dimensional shock wave is produced in the
latter. The shock wave travels along the resin layer and reaches the layer of explosive. The
decomposition of the explosive takes place and the chemical energy is released behind the
shock wave propagating through the charge. After some distance, the wave transforms into a
full detonation wave or, if rarefaction waves scattered from the free surface of the resin layer
disperses the chemical energy, it is weakened and the detonation process fails.
839
p(t)
Resin
-x r
Fig 1.
Explosive
Diagram of the system investigated
For the mathematical description of the formulated problem, use will be made of the
integral form of the equations of unsteady one-dimensional gas-dynamic flow with plane
symmetry, written down in the Lagrangian coordinates x, t
dx dt = 0 ,
(1)
where
=
0 u
u2
0 +
2

u
= p .
u p
The symbols u, p, , in the set of Eqs. (1) denote the mass velocity, pressure, density
and the internal energy of the medium per unit mass; 0 is the initial density, symbol
denotes an arbitrary closed contour on the plane x, t.
The set of Eqs. (1) is completed by the equation of state of the medium. The equation of
state for the material of inert gap can be obtained from an expansion of the internal energy
about a known empirical relation in the p- plane. If it is, for instance, a shock adiabate
pH(), the approximate equation of state in the neighbourhood of the latter can be written in
the form
( p, v ) =
1 1
1
1
+ [ p p H ( )]
p H ( )
2
0 0
0
(2)
where 0 is the Gruneisen coefficient of the material under normal conditions.

For most condensed media, the empirical relation between the speed of a shock wave in
a medium Ds and the mass velocity us may be approximated by the equation of a straight line
Ds = a + b u s
(3)
Then, the shock adiabate of the medium is described by the equation

p H ( ) =
a 2 0
(1 b )
= 1
,
where
0
.
(4)
A description of the layer of explosive is more complex. Initially, it is a condensed

medium. In the reaction zone, it can be considered as a two-component medium composed
of the unreacted explosive and the products of reaction. Once the chemical reactions are
840
finished, it is mostly a gaseous medium. The current state of the medium is described by a
parameter (the mass fraction of reaction products). To describe the state of the mixture in
the reaction zone, the specific volume and internal energy are assumed to be additive
v = v p + (1 ) v e
(5)
= p + (1 )( e + Q )
(6)
where Q is the heat of reaction.

To describe the thermodynamic properties of the products, the JWL (Jones, Wilkins,
Lee) equation of state is assumed
( p, v ) =
R v
R v
v v
v v
pv
A 0 exp 1 B 0 exp 2
w
w R2
w R1
v0
v0
(7)
where A, B, R1, R2, are constants, v = 1/, v0 = 1/0. The equation of state for
unreacted explosive is assumed in the form (2)(4).
The mass fraction of explosive products, , is obtained from a solution of the
macrokinetic equation. We assume that the shock initiation of heterogeneous solid
explosives is modelled by a three-term, macrokinetic equation [3]
d
= I (1 ) x0 r + G1 (1 ) x1 y1 p z1 + G 2 (1 ) x2 y2 p z2
dt
(8)
where = /0-1, and I, G1, G2, x0, r, x1, y1, z1, x2, y2, z2 are constants.
The first term is connected with formation of the hot spots created by various
mechanisms (void closure, viscous heating, gas heating in the pores, friction between
particles, etc.) during the shock compression of the explosive. This term vanishes off by
substituting I = 0 when exceeds the given value ig. The second term in Eq. (8) describes
the relatively slow growth of reaction by both the inward and outward burning of isolated
hot spots. The second part of Eq. (8) is invalid when > p. The last step in the shock
initiation process is a rapid completion of the reaction as the reacting hot spots begin to
coalesce. This stage is described by the third term in Eq. (8).
The initial conditions of the problem are as follows
p = p0
= 0r
u = 0
for
xr < x < 0 ,
p = p0
= 0e
u = 0
for
0< x<,
(8)
where 0r and 0e denote the density of resin and of the explosive, respectively.
The formulation of the problem is completed by the boundary conditions in the form [4]
841
l
l
p = p max d ; t d
Dd
Dd t
for
x = xr ,
p = p0
for
x,
(9)
where pmax denotes the maximal pressure at the boundary of resin layer, Dd the
detonation velocity of donor explosive, ld the donor length, n the exponent describing the
rate of pressure decreasing.
The numerical method for solving the problem of shock initiation of the heterogeneous
explosive was based on Godunovs scheme [5] with the first order of approximation. Its
adaptation for the solution of the equation set, given in Lagrangian coordinates, of a onedimensional motion of the medium in plane symmetry has been discussed in details in
Ref. [6].
3.
NUMERICAL RESULTS
To calculate the pressure pulse loading the boundary of resin layer, the following
parameters for the donor charge were assumed: Dd = 7000 m/s, pmax = 20 GPa, ld = 30 mm
and n = 3.2. Only so-called active mass of RDX charge was taken into account [7]. Estimated
pressure profile at the resin boundary is shown in Fig. 2.
20
p [GPa]
15
10
0
0
10
12
14
t [s]
Fig 2. Pressure profile at the boudary of resin layer ( = t ld/Dd)
To investigate numerically the process of shock-to-detonation transition, the data of

equations of state are needed for considered explosive. Up to now, coefficients of the shock
adiabate, JWL equation of state or macrokinetic equation are not determined for explosive
compositions containing NTO. To compare in a qualitative manner the gap test results with
those of numerical simulation, material data of a formulation containing fine RDX are used.
Details concerning the equation of macrokinetics, the equation of state of the formulation
and its detonation products are taken from Ref. [3]. The density 0 = 1600 kg/m3, the
constants for the shock adiabate a = 2670 m/s and b = 1.33, the Gruneisen coefficient
0 = 0.97 - Ref. [8-9] were assumed for resin.
842
The calculated time-dependence of shock wave position in the explosive layer and the
velocity of shock wave as a function of the distance traveled are presented in Figs. 3 and 4,
respectively. In the case of 50 mm gap, it is seem that the initial shock wave moves with
almost constant velocity and it suddenly transforms into the detonation wave with higher
velocity. This process is not observed for the system with 60 mm inert layer. Comparison of
the experimental and calculated gap test data, i.e., the time-dependence of shock wave
position in the explosive layer, indicates a good qualitative agreement between them.
However, a quantitative coherence will be possible when the material figures of explosives
tested are determined.
80
70
Distance [mm]
60
50
40
30
20
50 mm
10
0
60 mm
0
8 10 12 14 16 18 20
Time [s]
Fig 3. Calculated time-dependence of shock wave position in the explosive layer
8000
Wave velocity [m/s]
7000
6000
5000
4000
3000
2000
50 mm
1000
60 mm
10 20 30 40 50 60 70 80
Distance [mm]
Fig 4. Calculated dependence of shock wave velocity on the distance traveled
843
The calculated pressure histories at various depths in the modelled explosive loaded
throughout the resin layer of different gap are presented in Figs. 5 and 6. In the case of
thinner resin layer, the chemical energy is released relatively early and it influences the
pressure in the reacting medium at a short distance behind the shock wave. Due to the
subsonic flow of the medium, the disturbances generated by increasing the pressure catch up
and reinforce the shock front and, thanks to this, the rate of chemical reactions increases. As
a result, the shock wave transforms into the detonation wave with its characteristic peak of
pressure. It is follows from the pressure histories, that the backward pressure jump runs
trough unreacted or only partially reacted explosive in the reverse direction with respect to
the detonation wave propagation. This phenomenon is commonly known as a retonation.
30
25
p [GPa]
20
15
10
10
6
10
t [s]
Fig 5.
x[
mm
]
30
25
20
15
5
12
14
Pressure profiles at different depths in the charge for the system with 50
mm resin layer
30
25
p [GPa]
20
15
10
25
20
5
2
10
4
x[
mm
]
15
0
5
10
t [ s]
12
14
Fig 6. Pressure profiles at different depths in the charge for the system with 60 mm
resin layer
844
When the charge is loaded by the weaker pressure pulse in the configuration with 60 mm
inert layer, the chemical energy is released at a greater distance behind the shock front and it
influence on shock front is less effective. Moreover in reality, the shock pressure is reduced
by the lateral rarefaction waves moving from the boundary of the explosive charge. Finally,
the transition of the shock wave into the detonation wave is prevented.
4.
SUMMARY
The analysis of the numerical results enables us to draw the following inferences.
1) The curve of shock wave position in an explosive charge versus time from
calculations for fine RDX parameters is in good qualitative agreement with those
from experiments for compositions with NTO.
2) The model proposed can be successfully used to simulate the shock to detonation
transition in the gap test configuration.
3) The three-term, macrokinetic equation can be used to describe actual ignition and
growth processes in low-sensitive explosives.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
W. A. TRZCISKI, L. SZYMACZYK, M. KUTKIEWICZ: The use of the gap test to investigate

the shock to detonation transition in low-sensitivity explosives Part I Experimental
approach, 8th International Seminar New Trends in research of energetic Materials,
Pardubice 2005.
W. A. TRZCISKI: Numerical simulation of the shock to detonation transition in condensed
explosives, 35th International Annual Conference of ICT, Karlsruhe 2004.
C. M. TARVER, J. O. HALLQUIST, L. M. ERICSON: Modeling short pulse duration shock
initiation of solid explosives, Eighth Symposium (International) on Detonation, Albuquerque
1985.
K. P. STANYUKOVITCH: Physics of explosion (in Russian), Nauka, Moscow 1975.
S. K. GODUNOV: Numerical solution of multi-dimensional problems of gas dynamics (in
Russian), Nauka, Moscow 1976.
R. TRBISKI, W. TRZCISKI, E. WODARCZYK: On penetration of pressure-pulse into
layered medium of diphase (liquid/gas) structure, J. Tech. Phys. 20, 4, 1979.
R. TRBISKI, W. TRZCISKI, E. WODARCZYK: Estimation of the active mass of an
explosive charge (in Polish), Engineering Transactions, 35, 3, 1987.
S. P. MARSH: LASL shock Hugoniot data, University of California Press, Berkeley 1980
J. BUCHAR, J. VOLDRICH: Terminalni balistika, Academia, Praha 2003.
845
AGING BEHAVIOR OF PROPELLANTS

Yukihiro Tsunezumi*, Miyako Akiyoshi**, Hiroshi Miya***,
and Hidetsugu Nakamura*
* Department of Applied Science for Integrated System Engineering,
Kyusyu Institute of Technology, Sensui-machi, Tobata-ku, Kitakyusyu-shi 804-8550, Japan
** National Institute of Advanced Industrial Science and Technology (AIST),
Higashi1-1-1, Tsukuba, Ibaraki, 305-8565, Japan
*** Asahi Kasei Chemicals Corporation, Engineering and Development Department ,
Oita Plant, 2620, Oaza-Sato, Oita 870-0392, Japan
Abstract
Nitrocellulose is used as a main component of propellants. The aging behavior of
various type of nitrocellulose were studied at the neighborhood of room temperature
from 25 degrees to 50 degrees by accelerating aging tests. Moreover, the aging
behavior of propellants which contained stabilizers was also investigated using four
kinds of stabilizers such as ethylcentralite, diphenylamine, 2-nitrodiphenylamine and
akardite 2 in this study. From these results, the reason of degradation of nitrocelulose
and propellant were ascribed to acid hydrolysis and peroxide formation.
Keywords:
1.
nitrocellulose, aging, propellants, degradation
INTRODUCTION
Nitrocellulose (NC) is used as a main component of the propellant. However, though it

is the self-reactive material, the degradation mechanism has not accurately been clarified.
Therefore, many accidental explosions by spontaneous ignition in the handling after
prolonged storage were reported. Though the spontaneous degradation is suppressed by
adding stabilizers such as the diphenylamine, the reaction mechanism of stabilizer remained
unclear.
There are two theories for the mechanism of propellant degradation; one is the
hydrolysis by remained acid and water in the manufacturing, and the other is the formation
and decomposition of the unstable intermediate (peroxide) by the oxidation. In this report,
deterioration mechanism and temperature dependency of the degradation of NC were
examined by accelerating aging tests. Moreover, the reaction mechanism of stabilizer were
examined from aging characteristic of the propellant which contained various stabilizers for
the purpose of suppressing spontaneous ignition and selecting of optimum stabilizer.
846
2.
EXPERIMENT
2.1
Sample
The experiment was carried out using the following four types of nitrocellose, single
base propellant (SB) and double base propellant (DB).
Table 1. Four types of nitrocellose
Sample
NC 1
NC 2
NC 3
NC 4
Nitrogen Content
Low nitrogen quantity
Low nitrogen quantity
High nitrogen quantity
High nitrogen quantity
Polymerization degree
Low polymerization level
High polymerization level
Low polymerization level
High polymerization level
The degradation depression effect and reaction mechanism were examined using
Ethylcentralite (ECL), Diphenylamine (DPA), 2-nitrodiphenylamine (2NDPA) and
Akardite2 (Ak 2) as a stabilizer.
2.2
Storage condition
The sample was put in the container in the darkroom in order to prevent the degradation
by the light. The temperature was kept constant of 25, 40 and 501degrees and humidity
about 75%.
2.3
The experimental method
The thermal analysis was carried out using differential thermal analysis equipment
TAS-200. 0.5mg of samples was sealed in aluminum container and heating rate 5deg./min.
Measurement of ion content formed by degradation was carried out using ion chromatograph
IC-5A made by Shimadzu Corp. The measurement of peroxide quantity was carried out by
the iodometric titration method. The measurement of the stabilizer quantity was carried out
using liquid chromatograph LC-5A made by Shimadzu Corp.
847
3.

3.1
Degradation of nitrocellose
For 4 kinds of NC, the heat resistance was decreased with increasing aging period.
And, heat resistance was rapidly decreased for higher temperature aging in comparison with
that under at the lower temperature in either NC. From this result, it was proven that the
degradation speed of NC depended on the ambient temperature.
at 25deg.
at 40deg.
at 50deg.
10
titime/min/r
Heat resistance time/min.
Acid content and active oxygen quantity (peroxide quantity) showed the effect on
deterioration which increased in either NC. It is confirmed that there exist a correlation
between combined quantity of acid content and active oxygen and heat resistance. And, the
linear relationship was shown on combined quantity of acid content and active oxygen and
heat resistance under the same temperature condition. From these results, it was concluded
that the degradation of NC depended for both of peroxide generation mechanism and acid
catalyst mechanism.
NC1
0
0
20
40
60
Storage time/days
Variation of Heat resistance timE with Storage time
Acid cont.and
Active Oxygen cont./%
Fig 1.
0.08
0.04
NC1
0
0
10
15
Heat resistance time /min.
Fig 2.
Relationship between Acid cont. and Active Oxygen cont

848
3.2
Degradation of single base propellant (SB)

and double base propellant (DB).
It is known that NC shows the peak of exothermic decomposition which is occurred at

about 194degrees. Fig.3 shows the results of thermal analysis of NC without aging and NC
with AK2 after aging for 42 days. Both samples show a similar thermal behavior.
0
50
100
0
NC/Ak2
Weight loss /%
Endo. T Exo.
NC 0day
50
42days
100
0
Fig 3.
100
200
Temperature /deg.
300
DTA-TG curves of SB
Fig.4 shows the result of nitrate ion formation caused by aging. The amount of ion
formation measured according to the ion chromatograph showed the remarkable increase in
the nitrate ion for the NC without stabilizer. This ascribed to nitrate ion formed by
hydrolysis of the nitro group. Amount of nitrate ion formed within storage decreased by
adding of stabilizers.
849
Amount of nitrate ion formed /ppm
60
50
40
30
NC
ECL
DPA
2NDPA Ak
20
10
0
Fig 4.
14
28
42
56
70
Storage time /days
84
Amount of nitrate ion formed in storage
Fig.5 shows the results of active oxygen content formation in propellant after aging. The
peroxide content increased for NC without stabilizer after the aging. Amount of active
oxygen formed in storage decreased by addition of stabilizer. Especially, stabilizer AK2 and
ECL had a large effect on stabilization of single base propellant.
Active oxygen content /%
10.0
10-2
8.0
6.0
NC
NC/ECL
4.0
2.0
Fig 5.
14
28
42
56
Storage time /day
70
84
Active oxygen content in propellants
850
NC/DPA
NC/Ak2
Amount of stabilizers /wt%
The stabilizer is consumed by combining with the free acid by the degradation of NC.
Therefore, the residual quantity of stabilizer becomes the index which shows the level of the
degradation of the propellant. It also becomes a criterion in the life of the propellant. Then,
figure 6 showed the stabilizer quantity in the propellant in the course of aging. Amount of
stabilizer decreased with increasing storage. On the propellant with 2NDPA, it became a
result which the stabilizer quantity increased in the aging, and this ascribed to the nitration of
DPA derivative which exists in the stabilizer as an impurity. In DB, DPA was remarkably
consumed, and it concluded us that its effect as a stabilizer was low.
3.0
2.0
1.0
0
14
28
42
56
70
84
Storage time /day
Fig 6.
ECL
DPA
2NDPA
AK
Ak
Amount of stabilizers in storage
Addition of stabilizer caused suppression of nitrate ion and active oxygen formation in
storage. On the other hand, stabilizer in propellant decreased with increasing storage time
and AK2 or ECL were more effective to prevent NC degradation.
851
4.
CONCLUSION
There have been proposed two mechanisms in propellant degradation. But, it was
concluded that the degradation of NC depended on both of peroxide generation mechanism
and acid catalyst mechanism in this experiment. Moreover, addition of stabilizer caused
suppression of nitrate ion and active oxygen formation in storage. On the other hand,
stabilizer in propellant decreased with increasing storage time and AK2 or ECL were
effective to prevent NC degradation.
REFERENCES
[1]
JUN-ICHI KIMURA: Proceedings of 1998 autumn meeting of Explosives society of JAPAN,

P63-64 (1998)
[2]
T.URBANSKI: Chemistry and Technology of Explosives, vol.2, Pergamon Press, P307-318

(1965)
852
CONTRIBUTION TO STUDIES
OF DETONATION OF A CURVED CYLINDRICAL CHARGE
Ji Vgenknecht*, Vladislav Adamk, and Zbynk Aktein**
* University of Pardubice, Department of Theory and Technology of Explosives,
530 12 Pardubice, CZ
** Research Institute of Industrial Chemistry (RIICH),
532 17 Pardubice - Semtn, CZ
Summary
In connection with previous experiments and newly gained knowledge there have been
suggested and realized experiments, which should have better involved the character of
process of the detonation wave influenced by the way of deformation of cylindrical
charge made from a special plastic explosive on RDX basis. The character of process of
the detonation wave has been researched in connection with done experiments by the
help of the numerical simulation method Finite Element Method in Lagrangian system
ANSYS/LS-DYNA 3D.
Keywords:
1.
detonation, curved cylindrical charge, numerical simulation,

ANSYS, LS-DYNA
INTRODUCTION
Plastic explosives get a wide response in a range of branches and applications. Their
major advantage consists in the possibility of adaptation of the charges shape and rather
complicated geometries. It is applied above all at charges for destructive applications, where
there are apart from classical charges often applied also shaped or shapeable cumulative
charges. [1]
Change of the charges shape is significantly projected also onto the way, which e.g.
bend of the deformed charge is subjected to the detonation transformation. As the results of
some of the experimental works prove, e.g. [2,3], the progress of detonation through the
deformed curved charge has a characteristic profile, which shows principal difference
from the progress of detonation in direct charge. Documentable outer manifestations of
detonation at generally deformed charges can be viewed as anomalous. Character of
mentioned abnormalities at curved charges consists first of all in detonation velocities tested
at discreet distances at outer and inner circumference of the charge at its detonation.
Measured detonation velocities significantly vary from nominal detonation velocity of the
explosive in given diameter of the charge. At inner circumference of the charge there is
measured lower velocity, at outer circumference the measured detonation velocities
markedly exceed the nominal value. Connection between the change in geometry and outer
manifestations of detonation is evident, but its quantification is rather difficult. There occur
changes in geometry in all three dimensions because of deformation bend of the charge.
These changes are related above all with the deformation characteristics of used plastic
explosives. In optimal case it wouldnt come to any changes in geometry in the cross-section
of the charge. Influence of the progress of detonation wave of the charges would then be
caused by relative shortening of the charge at inner circumference and its relative
853
prolongation at outer circumference together with corresponding changes of the density of

the charge in its cross-section. In practice there of course unambiguously comes also to
indispensable deformation in cross-section of the charge. It is manifested in asymmetrical
distribution of the explosives mass in the charge, which must be again in a certain way
projected in the detonation profile.
It is necessary to take into account of the resultant effect also a certain influence of used
way of initiation and eventual influence of the cover, which can be presented e.g. by
a material pad.
It was necessary to prepare experiments for documentation of this process, in order the
most essential outer manifests of detonation of curved charges could be sufficiently covered,
that would then allow competently interpreting the process of the detonation action in
dependency on applied deformation.
2.
EXPERIMENTS
2.1
Original experiment
Study of the problems has resulted from the modification of original experiments [3]. The
main output of these outputs was the detection of different detonation velocities at the
progress of detonation wave through particular parts of the curved charge at its inner and
outer surface, as it is evident and shown in table 1.
Table 1. Measured values of detonation velocities in different parts
of the curved charge at work [3]
[m.s-1]
VOD
part of
charge
average VOD [m.s-1]
forepart
of the
charge
INNER
7675,3
7600,2
7752,4
7551,0
forepart
of the
charge
OUTER
8357,1
8223,1
8314,4
8347,8
end
of the
charge
INNER
7650,9
7548,2
7552,1
7664,5
end
of the
charge
OUTER
8630,0
8625,6
8627,0
8650,9
7644,7
8310,6
7603,9
8633,4
Arrangement of the experiment, which is using available measuring technique, that is

limited for 7 scanning points, has been designed in such way, it is able to cover the entire
length if possible together with shifts of synchronization of the movement of the detonation
wave in opposite points of outer and inner surface of tested charge. It proved, that realized
experiments offer a range of substantial information though, which it is possible to conclude
basic conception about the behavior of shaped charges at their detonation, but they are
insufficient for reliable documentation of the whole process. Therefore there has been
projected a realization of new experiments, which would better catch some of the less
conclusive manifestation.
854
2.2
New experiment
2.2.1. Used explosive

As well as in earlier experiments there has been used an explosive with marking Pl-SEM
(Plastic Sheet Explosive Military) produced by RIICH for measuring and simulation. This
explosive embodies one of the best possible combination of the characteristics, especially
homogeneity, solidity and plasticity, for observation the behavior of shaped charges.
P1-SEM explosive is a special shapeable explosive a hexogen (RDX), polymeric binder
- polyisobutadien (PIB) and some of the special additives based. Used composition
guarantees a good performance to the form of sheet of thickness of several millimeters. High
filling with hexogen and its granulometry assure a low critical diameter and also generally
stable and high detonation parameters.
2.2.2. Used charges
For experiments there have been prepared screwed charges of circular cross-section with
resulting diameter of 29.3 mm. Total length of particular charges has been 450 mm. Charges
have been on the pad curved round a solid matrix into a quarter circle with inner semidiameter of 220 mm.
2.2.3. Measuring technique and instrumentation
It has been dealt about use of a proved technique of time measurement of detonation
velocities at inner and outer surface of the charge with use of very tiny wire ionizing sensors
located in exactly limited distances and positions in a surface film of the charge see closer
fig. 1.
A)
Fig 1.
B)
Localization of sensors at outer circumference of the charge

A) set of sensors
B) detail of localization of sensors
Involvement of the sensors into measuring circuit and also used instrumentation have
been identical as the original experiment. See scheme in fig. 2.
855
ionizing sensors
OSCILLOSCOPE
coaxial cable
Fig 2.
R-C coaxial
CIRCUIT cable
curved charge
initiation
Scheme of measurement adjustment
Certain changes have come at measurement adjustment. Measuring bases have been
chosen in such way, they would be concentrated into particular sections of the charges,
which should enable as detailed catch of progress of detonation wave at the surface of these
sections along the total length of the charge.
For comparison there has been documented the progress of detonation also in unshaped
straight charge. Practical realization of experiments adjustment including placing the
booster element and detonator is displayed in pic. 3.
A)
Fig 3.
B)
Examples of placing the charges for measurement

A) curved charge
B) straight charge (comparative measurement)
Results of measurement are summarized in Chapter 4 Results and Discussion
3.
NUMERICAL SIMULATION
As it has been already shown in original experiments, interpretation of real detonation

processes running inside the charge based only on their outer manifestations can be really
difficult or even speculative.
Already in past there have been at DTTX used with success numerical simulation
methods by the help of Finite Element Method (FEM) in ANSYS/LS DYNA3D [4] system at
visualization and verification of similar dynamic processes. Right space numerical
simulation allows to get in the mass of charge and put in relation known patterns of
856
running processes with investigated outer manifestations. Basic question is then correct
treatment of the model, setting the relevant material characteristics and boundary conditions.
Parameters of the explosive and the form of expansion of detonation products is
successfully defined by parameters of JWL (1) equation for given explosive.
p = A 1 R1 0 e
R1
+ B 1 R2 0 e
R2
+ E 0
(1)
where p is pressure, is density, E is internal energy, A, B, R1, R2, are material parameters.
Parameters of this explosive have been calculated based on the composition of explosive
by the help of CHEETAH [5] program. For Pl-SEM explosive there have been used
calculated parameters mentioned in table No. 2.
Table 2. JWL parameters for Pl-SEM [6]
Explosive
P1-SEM
E0
[kg/m3]
[GPa]
[GPa]
[GPa]
1501.0
8.226
631.0
7.4530
R1
R2
4.795
1.038
0.331
pJ-C
[GPa]
21.2
There have been used two accesses for solving the problem.
3.1
Model of the whole curved charge [7]
There have been simulated the whole real 3D geometry of the system (fig. 4). Complete
geometry has been used for catching the influence of the pad, which has been simulated
as a solid wall. The simulation has had following characteristics:
calculations have been processed by the help of calculation program LS-DYNA 3D

Code, Version 950e
there have been used space (3D) brick elements: 8 nodes, single-point integration,
viscous regulation of instability of hourglass type
number of elements of simulation:
339094
number of nodes of simulation:
320112
area initiation of the charge has been assumed in the x-y plain of the explosive
(zone of area initiation is represented by blue colour - fig. 4, 5)
Fig 4.
Curved charge solved

geometry
- numerical simulation model
Fig 5.
857
Detail of initiated forehead

of curved charge used mesh
In this simulation there has been at DTTX for the first time successfully proved
a suitability of setting for description of progressing of detonation wave in the explosive
model, so called burn up, which counts increasing in pressure in the element to setting
initiation limit based on explosives element compression, when it comes to blasting off
the whole element. Profitability of the used progress is connected with a good
correspondence with real way, by which it comes to propagation of detonation by explosive.
It is possible to document the realised result of numerical simulation, e.g. by a series
of pictures in pic. 6, 7 and 8.
Fig 6. Progress of detonation through the curved charge pressure profiles

in times of 4, 17,25 and 35 s sideview of contact with the pad
Fig 7.
Detail of pressure profile

of forehead of detonation wave
in contact with the pad in time
of 35 s
Fig 8.
858
Global view of a model

expansion of products of curved
charge at an absolutely rigid wall
in time of 35 s
Numerical simulation of progress of detonation through the curved charge has definitely
confirmed perpendicular of detonation wave on surface of the charge towards its longitudinal
axis, symmetry and equability of the whole process. Problem of this access has consisted in
dealing about the description of outer manifestations, which has not allowed to clear up the
real profile of the progress of detonation inside the charge, which is the right cause in certain
sense of anomalous behaviour of detonating charge. Therefore there has been used the
second access.
3.2
Model of curved charge in 1/2 cross-section
Regarding the fact, it has been confirmed, that the pad has no influence on behaviour of
detonating curved charge, it has been considered the performance of simulation of process of
detonation of curved charge with similar solution access (characteristics of the model,
description of the materials and reactions), but in cross-section in the plain of symmetry
parallel with the direction of bend of the charge. The way of cross-section is evident from
picture No. 9. Improvement of calculation platform and reducing the volume of the model
has allowed to use thicker network see fig. 10 (compared to fig. 5).
Fig 9.
4.
Fig 10. Detail of initiated forehead

of curved charge used network
Curved charge solved

geometry
- numerical simulation model

4.3
Measurement of detonation velocity
Characteristic oscilloscopic record of the measurements is described in fig. 11.
859
Fig 11. Record of pulses generated while passing through detonation wave
via particular sensors on the surface of the charge.
Processed results of performed measurements of detonation velocities in particular
sections of curved and also straight charges are in a brief form in table No. 3. For
simplification of the survey in localization of sensors at quarter circle charge with different
relative prolongation / shortening on outer and inner surface the distances between particular
sensors have been re-counted to angular shift from the forepart of bend of the charge.
Table 3. Detonation velocities measured in particular sections of curved
and straight charges
outer side
distance
from the
forepart
detonation
velocity
[mm]
inner side
straight charge
angle
distance
from the
forepart
[m.s ]
[]
[mm]
[m.s ]
[]
[m.s ]
[]
65
90
115
140
165
190
205
8012,8
8250,8
8278,1
8278,1
8888,9
8389,3
9049,8
7,73
18,43
24,37
30,31
36,25
42,20
46,96
25
75
92
100
125
132
150
7716,0
7575,8
7692,3
7598,8
7598,8
7448,8
7462,7
3,38
13,52
22,57
25,95
30,41
34,73
38,11
7761,4
7827,8
7692,3
7827,8
7575,8
7692,3
7952,3
10
25
40
55
65
75
85
230
245
285
8417,5
8547,0
8547,0
49,93
54,68
63,01
172
212
252
7561,4
7575,8
7692,3
43,52
51,9
62,71
325
9049,8
72,52
292
7561,4
73,52
-1
860
detonation
angle
velocity
-1
detonation
angle
velocity
-1
These results have been processed into a graph in fig. 12.
Fig 12. Dependence graph of changes of detonation velocities at inner and outer side
of curved charges in particular sections of charges in comparison with
an unshaped charge
inner side of a curved charge outer side of a curved charge + unshaped charge
It is well evident from graph in fig. 12 that relative reduction of detonation velocity at
inner side of the charge in comparison with nominal velocity is not very significant and it is
achieved very quickly. To the contrary there comes markedly slower to increase and
stabilization of detonation velocity at outer circumference of the charge in used geometry
and the difference in velocity resulting from used geometry is more significant. Numerical
simulation of profile of detonation wave in a cross-section of the charge has contributed to
clarification of this effect in an essential way.
4.4
Outputs of numerical simulations
The most substantial information, which has resulted from numerical simulations
according Chapter 3.2, has consisted in catching the profile of detonation wave passing
through the charge, as it is evident from figures 11 and 12.
861
Fig 13. Profile of propagating

detonation wave in the profile
of the cross-section.
Fig 14. Vectors of velocities

of propagation of detonation
wave in profile of the charge
Interpretation of these outputs allows to conclude the causes of outer manifestations of

detonation of curved charges. It is evident that the documented influence of shaping /
bending of plastic charges through a change in geometry is projected partly into mechanical
characteristics of the charges and it also partly influences compression-expansion complex
of detonation wave and detonation products. The very dynamic setting of this balance is an
evident cause of disproportional manifestation of detonation of generally deformed charges,
which is significantly projected into the process of energy release of the explosive into the
area at detonation.
5.
CONCLUSION
Performed practical and simulation experiments have brought essential knowledge to

understanding of influence of bending the charges made from plastic explosives on the
process of their detonation transformation and outer manifestation of their detonation.
Experimental knowledge from measuring of detonation of charges with circular crosssection is in a good correspondence with results published for charges with rectangular
cross-sections. A very substantial contribution for documentation of studied action has meant
a successful setting of numerical simulation using Finite Element Method in ANSYS/LS
DYNA 3D system. Results of performed simulations have enabled effectively to document
the process of detonation in profile of curved charge, which have subsequently allowed to
specify main causes of outer manifestations of detonation of studied charges.
Acknowledgement:
We would like to thank to VPCH for the help at material guarantee of the experiments.
Special thanks go under Mr. Ji Majzlk from DTTX for a professional guarantee
of experimental instrumentation.
862
REFERENCE
[1]
[2]
[3]
[4]
[5]
[6]
[7]
Special Explosive Products, publicity matter, Research Institute for Industrial Chemistry,
Synthesia Corp. Pardubice-Semtn
P.C. SOUERS ET AL.: Corner turning rib tests on LX-17, Propellants, Explosives, Pyrotechnics
23, 200-207, 1998.
J. VGENKNECHT, Z. AKTEIN, P. VVRA: Measurements of detonation velocities of formed
liner charges, Proceedings of the V. seminar on New trends in research of energetic materials,
University of Pardubice, Pardubice, Czech Republic, April 24-25, 2002.
LS-DYNA THEORETICAL MANUAL, Livermore Software Technology Corp., 1998.
L. FRIED, P.C. SOUERS, W.M. HOWARD, AND P. VITELLO: Cheetaha thermochemical-kinetics
code, software, LLNL (2000)
W.A. TRZCINSKI: Private communication, Military University of Technology, Poland 2004.
J. VGENKNECHT, V. ADAMK: Using of LS DYNA for study of detonating charges in special
geometries - part 1, Procc. 11. ANSYS Users Meeting, SVS FEM, Znojmo, I-A-10, s. 1 6,
September 2003 (in Czech)
863
APPLICATION OF LC-MSD AND GC-ECD TECHNIQUES

IN THE ANALYSIS OF POST EXPLOSION RESIDUES
OF EXPLOSIVES ON THE BASIS
OF 1,3,5-TRINITROSO-1,3,5-TRIAZINANE
R. Varga*, S. Zeman*, J. Vgenknecht*, and M. Kolovri**
* Department of Theory and Technology of Explosives, University of Pardubice,
** Institute of Forensic Science of the Slovak Police Corps,
Sklabinsk 1, SK-812 72 Bratislava, Slovak Republic
Abstract:
The paper deals with possible applications of the connection of liquid chromatography
with atmospheric pressure chemical ionization (APCI) mass spectrometry (Agilent 1100
series LS-MSD system) and gas chromatography with electron capture detector (GCECD) in the analysis of some common available or improvised explosives.
The analysis of slected types of individual explosives, 2,4,6-trinitrotoluene (TNT), 1,3,5trinitroso-1,3,5-triazinane (TMTA) and analysis of their mixtures (TNT/TMTA and
RDX/TMTA) in their post explosion residues on the basis of generated characteristic
mass spectra by using LC-MSD system is the aim of this work.
Post explosion analysis of TMTA and its explosive mixtures can be similarly
complicated, like that of the well-known plastic explosive on the basis of pentaerythritol
tetranitrate, Semtex-1A.
Keywords:
1.
Explosive, LC-APCI/MS, GC-ECD, 1,3,5-trinitroso-1,3,5-triazinane,

TMTA, trace analysis, 2,4,6-trinitrotoluene, TNT.
INTRODUCTION
Forensic identification of post-explosion residues in debris material is the major interest

of the criminalistic investigation of the bombing because it will help to find the connection
between the used explosive and the suspect [1].
Today the mass spectrometry (MS) belongs to significant methods of investigation of
explosives. The utilize of new methods of molecular ionization makes to possible to measure
the mass spectra of substances, which are considerably thermally instable or have very low
volatility [2]. The sensitivity and specificity of the mass spectrometer make it a suitable
technique for this application [1].
Currently accepted analytical methods for the determination of trace levels of explosive
residues rely almost exclusively on liquid or gas chromatography. Generally, gas
chromatography with electron capture detector or mass spectrometry is sensitive technique,
but reproducibility suffers especially from the thermally labile compounds. The evolution of
the LC-MS interface has also allowed development of sensitive and selective methods for a
variety of polar and thermally - labile compounds including explosives [3]. LC-MS, with
atmospheric pressure chemical ionization (APCI), in the negative-ion mode, was found to
provide best sensitivity and selectivity for nitroaromatic compounds [4].
864
Cassada et al. described the development of a simple method for the determination of
munitions residues and RDX-metabolites in water using atmospheric pressure ionization
(API) and electrospray ionization (ESI) techniques [3]. Liquid chromatography-atmospheric
pressure chemical ionization mass spectrometry (LC-APCI/MS) has been used for the
identification and characterization of by-products of industrial 2,4,6-trinitrotoluene (TNT) in
the work of Zhao et al. [4]. Several applications of LC-APCI/MS for the analysis of
explosives and explosive residues are described by Schreiber et al. and Varga et. al. [5,6] . A
series of explosives were studied by combined high-performance liquid chromatographymass spectrometry (HPLC-MS) in the paper of Yinon et. al. [7].
GC detectors suitable for determination of explosives is reviewed by Yinon and Zitrin [8]
include the flame-ionization detector (FID), mass spectrometer (MS), nitrogen phosphorous
detector (NPD), electron capture detector (ECD), and thermal energy analyzer (TEA). The
ECD, widely used for the determination of halogenated compounds, is less selective, but is
more sensitive for nitro aromatics than the TEA or NPD. There are described applications of
GC-ECD for the identification of traces of nitroaromatic, nitramine, and nitrate ester
explosives [9,10,11].
However, literature doesnt contain information about detection and post-explosion
analysis of 1,3,5-trinitroso-1,3,5-triazinane (TMTA or R-salt). This substance is easy
available high explosive, which can be exploitable as a part of improvised explosive
mixtures. From the point of the said fact this paper deals with possibility of applications of
the liquid chromatography with atmospheric pressure chemical ionization (APCI), mass
spectrometry (Agilent 1100 series LS-MSD system) and the gas chromatography with
electron capture detector (GC-ECD).
2.
2.1
Used Substances
In this paper the following explosives were tested:

casting military 2,4,6-trinitrotoluene (TNT)
casting system of 1,3,5-trinitroso-1,3,5-triazinane (TMTA or R-salt) and additives
casting mixture of TNT and TMTA
casting matrix of RDX, TMTA and additives
Compositions of these mixtures are summarized in the Tab.1.
Table 1. Composition of studied casting explosive mixtures
Explosive
Composition (%)
Density (g/cm3)
TNT
100 % TNT
1,56
TMTA cast
92% TMTA, 4% diphenylamine, 4% anthracene
1,47
TNT/TMTA
58% TNT, 42% TMTA
1,64
RDX/TMTA
75% RDX, 23% TMTA, 1% diphenylamine,

1% anthracene
865
1,37
The used standards for the identification of individual components of explosive mixtures
were as follows: 2,4,6-trinitrotoluene (TNT), 1,3,5-trinitroso-1,3,5-triazinane (TMTA or Rsalt), 1,3,5-trinito-1,3,5-triazinane (RDX) and pentaerythrytol tetranitrate (PETN). The
molecular structures of these explosives are shown on the Fig. 1.
TMTA, a trinitroso analogue of RDX, is the only representative of nitrosamines
interesting as a military high explosive [12,13]. The TMTA manufacture was based on
nitrosolysis of hexamethylenetetramine (or aqueous mixture of ammonia and formaldehyde)
with sodium nitrite and mineral acid (sulphuric, hydrochloric), in cooled diluted water
solution. There was no need to use concentrated nitric or sulphuric acids in the manufacture
of the explosive; the process seemed to be safe an could be successfully realized also under
improvised conditions [14].
Fig 1.
2.2
TNT
TMTA
RDX
PETN
Molecular structures of investigated explosives
Apparatus for Analyses of Samples
Liquid chromatography-mass spectrometry (LC-MS) conditions

All analyses were performed by means of Agilent 1100 series LC-MSD system with the
atmospheric pressure chemical ionization (APCI) source, in the negative-ion mode. The used
chromatographic and mass spectrometric conditions are shown below.
Chromatographic conditions:
Column:
Zorbax Eclipse XDB-C8; 4,6 x 150 mm; 5 m (Agilent Technologies)
Mobil phase: Methanol/water; 70:30, v/v, with addition of ammonium chloride and
formic acid
Flow rate:
1 ml/min
Injection vol.: 20 l
866
MS conditions:
Source: APCI
Vaporizer temp. 300 C
Mode: SCAN and SIM
Dry gas flow: 5 l/min
Ionization polarity: Negative
Drying gas temp: 330 C
Capillary voltage: 3000 V
Fragmentor: 30 V
Nebulizer: 60 psi
Gas Chromatograph with Electron Capture Detector (GC-ECD)

The standards of explosives (TNT, TMTA, RDX and PETN) were analyzed on a
Hewlett Packard 5890 II chromatograph, using a capillary column HP-5 of 5 m length, 0.53
mm inner diameter, and 0,88 m film thickness. An electron capture detector (ECD) was
used for the detection, and the carrier gas was nitrogen of 20 kPa overpressure. The sample
(1 l) was injected directly (on-column), the injector temperature was being 70 C. The
temperature regime was as follows: the column was kept at 70 C for 1 min, whereupon the
temperature was increased at a rate of 15 C/min to 150 C, and 5 C/min to 230 C, and this
final temperature was kept for 8 min.
2.3
Collecting and Treatment of Samples
Every charge of explosives was placed in agreement with diagram (Fig. 2) on the
construction. The initiation was accomplished by means of Initiator No. 8. After the
detonation, samples were taken from the right side of metal plates with methanol and
distilled water (30 ml of each solvent for one metal plate). The extracts were filtered through
a paper filter (blue ribbon) in order to remove mechanical impurities. Methanol samples
were concentrated by using weak flow of nitrogen and water samples with hot air to the
minimal volume (1-2 ml).
Fig 2.
2.4
The Construction for the collection of post explosion residues
Progress of LC-APCI/MS analyses
Stock solutions of individual explosives (TNT, TMTA, RDX and PETN) were made in
methanol p.a. Standard mixtures were prepared by diluting the stock solution to a final
concentration as necessary.
867
First the Scan negative-ion APCI mass spectra of all explosives were measured in order
to find out their major ions. Then the Selected Ion Monitoring (SIM) methods maximal for a
pair of substances (TNT/TMTA and RDX/TMTA) were formed. In SIM mode minimal three
dominant ions of each substance were measured. We also measured the standards of
explosives by means of electron capture detector in combination with gas chromatography.
Mass spectra of TNT, RDX and PETN are described in detail of the paper [6]. TMTA
features mainly adduct and fragment ions in the APCI mass spectrum. TMTA also produces
like RDX adduct ions [M+Cl]- with the isotopes 35Cl and 37Cl at m/z 209 [M+35Cl]- and m/z
211 [M+37Cl]-. Mass spectrum of TMTA doesnt contain molecular ion, but its characteristic
ions are fragments rising by the loss of 30 mass units [M-NO]- at m/z 144. Adduct ions at
m/z 318, 383 and 385 were [2M-NO]-, [2M+35Cl]- and [2M+37Cl]-. The determination of the
presence of PETN in some organic extracts from the priming charge (booster) was based on
the SIM method, which included the ions of PETN at m/z 315, 351 and 378.
3.

3.5
Results of LC-APCI/MS and GC-ECD analyses

of post explosion extracts
All organic extracts of post explosion debris of TNT, TMTA and explosive mixtures of
TNT/TMTA and RDX/TMTA by means of LC-APCI/MS in SIM mode and GC-ECD
techniques were measured. In the case of liquid chromatography the used mobile phase,
methanol-water, allowed the identification of components in organic and water extracts, too.
Results of LC-APCI/MS measuring of investigated explosives are summarized in Tab. 2.
Table 2. Results of LC-APCI/MS analyses of post explosion products
of tested explosives
LC-APCI/MS analysis of post explosion samples
Explosive
Major ions of original

explosive
Methanol extract (M)

M1
TNT
[M]-, [M-H]-, [M-NO]-
TMTA
[M+35Cl]-, [M+37Cl]-,
[M-NO]-
TNT/TMTA
RDX/TMTA
[M]-, [M-H]-,
[M-NO]- (TNT),
[M+35Cl]-, [M+37Cl]-,
[M-NO][M+35Cl]-, [M+37Cl](RDX),
[M+35Cl]-, [M+37Cl]-,
[M-NO]-
Water extract (W)
M2
W1
[M]-, [M-H]-, [M-NO]-
W2
[M-H]-
[M+35Cl]-,
[M+37Cl]-,
[M-NO]-
[M+35Cl]-,
[M+37Cl]-
[M]-,
[M-H]-
[M-H]-
[M+35Cl]-,
[M+37Cl](RDX)
868
LC-APCI/MS analysis of water and organic post explosion extracts of TNT showed that
after detonation of TNT charges stays sufficiency of non-reacted particles and thats why its
post explosion LC/MS analysis is not problematical (Fig. 3A a 3B). In GC-ECD analyses of
these post explosion extracts the traces of TNT were identified, too. The results of GC-ECD
measurements are summarized in Tab. 3.
Fig. 5A shows the chromatogram obtained by the GC-ECD technique for post explosion
residues of TNT and identified component of PETN from the priming charge of Semtex-1A.
Table 3. Results of GC-ECD analyses of methanol post explosion extracts
of investigated explosives
Explosive
Identification of original components in

post explosion extracts
Identification of priming
charge (Semtex-1A)
Samples 1
Samples 2
TNT
TNT
TNT
PETN
TMTA
TMTA
TMTA
PETN
TNT/TMTA
TNT
TNT
PETN
RDX/TMTA
RDX
RDX
PETN
LC-APCI/MS post explosion analysis of TMTA already wasnt as successful as

LC/MS identification of TNT. From the two analyzed organic extracts only one sample
contained all ions of TMTA. In the case of water samples of TMTA the situation was similar
as in the case of organic extracts. The positive LC-APCI/MS chromatograms of methanol
and water post explosion extracts of TMTA are shown on the Fig. 3C a 3D. At the samples
of post explosion residues of TMTA by using of GC-ECD technique were obtained better
results (Fig. 5B) than by using LC/MS technique. It is caused by higher sensitivity of
electron capture detector to the TMTA.
LC-APCI/MS and GC-ECD chromatograms of explosive mixtures of TNT/TMTA and
RDX/TMTA (Fig. 4A, 4C and Fig. 5C, 5D) showed that amounts of non-reacted particles
of TMTA in the post explosion debris are below the sensitivity of both instruments. Two
important factors influence the amount of post explosion residues. The first factor of the
insignificant amount of TMTA in post explosion residues of investigated explosive mixtures
is the high detonation velocity, 6,95 km/s for TNT/TMTA and 8,00 km/s for RDX/TMTA
(experimentally determined values). This small amount of post explosion residues of TMTA
has also connection with the reduced amount of TMTA in tested explosive mixtures (42% of
869
this component in TNT/TMTA and 23% in RDX/TMTA) compared to the casting charge of
TMTA, which contains till 92 % of TMTA.
Fig 3.
Comparition of LC-APCI/MS chromatograms of post explosion residues

of TNT and TMTA in water (A and D) and in methanol (B and C)
870
Fig 4.
The positive results of LC-APCI/MS analyses of post explosion residues

of mixtures of TNT/TMTA (A - TNT), (B - PETN) and RDX/TMTA
(C -RDX), (D - PETN)
High detonation velocity of these mixtures causes increasing in ideal conditions of

detonation at the expense of amount of post explosion residues. Reduced amount of
components can also complicate the analysis of TNT and RDX in these explosive mixtures
(lower response of the detector for these components after detonation of explosive mixtures).
In all samples of organic extracts of TNT/TMTA and RDX/TMTA the presence of
PETN comes from the priming charges of Semtex-1A by means of GC-ECD and LCAPCI/MS (Fig. 4B a 4D) was confirmed.
871
Fig 5.
GC-ECD chromatograms of post explosion residues of TNT (A),

TMTA (B), mixtures of TNT/TMTA (C) and RDX/TMTA (D)
872
4.
CONCLUSION
On the bases of LC-APCI/MS and GC-ECD analyses of post explosion residues of

2,4,6-trinitrotoluene (TNT), 1,3,5-trinitroso-1,3,5-triazinane (TMTA) and two explosive
mixtures TNT/TMTA and RDX/TMTA we found the following results:
1) Analysis of TNT after detonation compared with TMTA, but mainly mixtures of
TNT/TMTA and RDX/TMTA, is relatively simple. TNT is the common wellknown detectable explosive. Both tested solvents (methanol and water) were
suitable for its post explosion analysis.
2) Analysis of TMTA is becoming complicated, because compared to TNT, it
detonates with higher detonation velocity and thats why after its detonation
stays lower amount of post explosion products (superior process of detonation).
Volatility of TMTA is relatively high which can also complicate its detection
after explosion.
3) TMTA in post explosion products of explosive mixtures of TNT/TMTA and
RDX/TMTA wasnt detected. However, these samples were positive for TNT
and RDX. The results advert to the fact that by using of mixing high explosive
without presence of TMTA in their post explosion residues can be obtained. In
the analysis of investigated explosives higher sensitivity of GC-ECD technique
was confirmed, but their LC/MS analysis was simpler on the basis of known
APCI mass spectrum of individual explosives.
4) Collection of samples of post explosion residues of tested explosives was carried
out at idealized conditions (sampling done immediately from the metal plate of
the construction). In real the time between the detonation and sampling and
weather conditions can also influence the success of analyses of these explosives.
Acknowledgement
The authors would like to thank to the management of Institute of Criminology of Police
Forces of the Slovak Republic in Bratislava for making possible and supporting the
realization of this work in the Institute.
873
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
J. YINON: Forensic Mass Spectrometry, Boca Raton, Florida, CRC Press I Llc, 106, 1987.
. KOVC, D. ILAVSK, J. LEKO: Metdy kontroly technologickch procesov (Methods
of inspection of technological processes), Spektrlne metdy v organickej chmii a technolgii
(Spectral methods in organic chemistry and technology), ALFA Bratislava, 1985.
D. A. CASSANDA, S. J. MONSON, D. D. SNOW AND R. F. SPALDING: Sensitive determination
of RDX, nitroso-RDX metabolites, and other munitions in ground water by solid-phase
extraction and isotope dilution liquid chromatography-atmospheric pressure chemical
ionization mass spectrometry, Vol. 844, Journal of Chromatography A, 87-95, 1999.
XIAOMING ZHAO AND JEHUDA YINON: Characterization and origin identification of 2,4,6trinitrotoluene through its by-product isomers by liquid chromatography-atmospheric pressure
chemical ionization mass spectrometry, Vol. 946, Journal of Chromatography A, 125-132,
2002.
ANDR SCHREIBER, JRGEN EFER AND WERNER ENGEWALD: Application of spectral libraries
for high-performance liquid chromatography-atmospheric pressure ionization mass
spectrometry to the analysis of pesticide and explosive residues in environmental samples, Vol.
869, Journal of Chromatography A, 411-425, 2000.
R. VARGA, P. ULBRICH, M. KOLOVRI AND M. FUKNOV: Trace analysis of explosives
by LC-MSD technique, In: Proc. of the VII. Seminar New Trends in Research of Energetic
Materials, Pardubice, Czech Republic 2004.
J. YINON AND D. G. HWANG: High-performance liquid chromatography-mass spectrometry
of explosives, Vol. 268 (1), Journal of Chromatography, 45-53, 1983.
J. YINON AND S. ZITRIN: Modern Methods and Applications in Analysis of Explosives,
Chichester, England: John Wiley & Sons Ltd., 1996.
MARIANNE E. WALSH: Determination of nitroaromatic, nitramine, and nitrate ester explosives
in soil by gas chromatography and an electron capture detector, Talanta 54, 427-438, 2001.
R. VARGA AND P. ULBRICH: Some experience with trace analysis of post-explosion residues,
Vol. 3, Issue 4, AARMS, Mikls Zrnyi National Defence University, Budapest, Hungary, 633646, 2004.
T. F. JENKINS ET AL.: Use of snow-covered ranges to estimate explosives residues from highorder detonations of army munitions, Vol. 384, Thermochimica Acta, 173-185, 2002.
URBASKI T.: Chemie a technologie vbuin (Chemistry and technology of explosives), Vol. 3,
SNTL Praha, 1959, p. 99.
B. T. FEDOROFF, O. E. SHEFFIELD: Encyclopedia of Explosives and Related Items, Vol. 3,
PATR 2700, Picatinny Arsenal, Dover, N.J., 1966, p.C630.
M. HANUS: Characterization of castable system RDX/R-salt, Proc. of the 10th Symp. Chem.
Problems Connected Stab. Explos., Margretetorp, Sweden, 35-55, 1995.
874
INTERMOLECULAR INTERACTIONS
OF ENERGETIC COMPOUNDS IN GASEOUS AND SOLID STATES
Xue-Hai Ju and He-Ming Xiao
Department of Chemistry, Nanjing University of Science and Technology,
Nanjing 210094, P. R. China
Abstract
The first-principle methods were applied to the study of some energetic compounds in
both gaseous dimer and bulk state. The binding energies have been corrected for the basis
set superposition errors. The influences of molecular interactions on the molecular
properties were discussed. The lattice energies, electronic band structures and density of
states in the solid states were obtained. Relationship between the electronic structures
and some detonation characteristics were elucidated.
Keywords: gaseous dimer, bulk state, binding energies,
molecular properties vs. detonation characteristics
1.
INTRODUCTION
Intermolecular forces are much less than those of chemical bonding, and yet they play
significant roles in a wide range of physical, chemical and biological fields. In recent years, we
have applied the intermolecular interactions to energetic systems and obtained some
meaningful information [1-7] that will be helpful to the study and molecular design of energetic
compounds. The behavior of molecular clusters is usually between two extremes: the gas
phase and the crystal solid phase. Consequently, one can obtain valuable knowledge about the
transition of these extremes by examining the properties of clusters of large size. Clusters
containing more than two molecules behave cooperative effects [8-10], which is reflected in
changes of some properties with increase in cluster size such as the interaction intensity
increase and the frequency shift. Properly characterizing these phenomena is thus crucial to
understanding the behavior of cluster. R2N-NO2 (R=H, Me) is a simple model of nitramine
energetic compounds such as HMX and RDX that are especially used as explosives.
1,1-diamino-2,2-dinitroethylene (Fox-7) has the same molecular stoichiometry as RDX and
HMX, and was characterized as a promising high energy density material with low shock
sensitivity. 3-nitro-1,2,4-triazole-5-one (NTO) was also found to have the low sensitivity and
high explosive performance. Difluoroamine, hydrazine and H2O2/H2O are other types of
explosives. Researches already done on the above compounds have been emphasized mostly
on their monomers [11-23]. To our knowledge, there are only two reports related to the
molecular interactions of dimethylnitramine dimers using DFT method or molecule
simulation method [24,25]. The intermolecular forces control such diverse phenomena as the
875
diffusion, the aggregation and the detonation. To provide the fundamental information for
further investigation on the mixture explosives, it is particularly necessary to study the
intermolecular interaction. The aim of this paper is to investigate theoretically the structures,
the frequency shifts, the binding energies, and the cooperative effects in the clustering
processes by ab initio and DFT methods.
It is also acknowledged that the performance of an explosive depends on its solid
condition such as the crystal structure. As the intermolecular interactions influence the
strengths of some weak bonds that are referred to as initiators, the performance of an explosive
depends on both the intermolecular interaction and the crystal environment. We also study the
intermolecular interaction and its influence on the behavior of Fox-7, NTO, pentaerythritol
tetranitrate (PETN) and octanitrocubane (ONC) in their crystalline phases. Finally, we studied
the electronic band structures and the density of states of the crystals, which gives a better
understanding of the basic characteristics of these crystals.
2.
COMPUTATIONAL METHODS
The monomer and all its possible stable dimers and clusters of the entitled molecules
obtained from Chem3D software are fully optimized by the Berny method [26,27]. Frequency
calculations were performed on each optimized structure.
The interaction energy of complex is evaluated as the sum of the HF interaction energy
and the correlation interaction energy. The latter term could be determined by the
Mller-Plesset perturbation theory [28-29]. The basis sets commonly used to calculate the
energies in the above equation are far from being saturated and, hence, in any complex each
subsystem will tend to lower its energy by using the basis functions of the other subsystem.
The energies derived from the equilibrium geometry of the complex for each subsystem are
lower than those calculated at the same geometry with the basis functions of the respective
subsystem alone. This energy difference is the so-called basis set superposition error (BSSE)
that can be checked by Boys and Bernardis counterpoise procedure (CP) [30-32]. For
complex of two submolecules the BSSE is
BSSE= EA* +EB* EA*(A*B*) EB*(A*B*)
where A*B* is the optimized dimer, EA* is the single point energy of submolecule A*
adopting the optimized structure of dimer A*B*, EA*(A*B*) is the single point energy of
submolecule A* by using the whole basis sets of dimer A*B*, definition for subsystem B is
the same as for subsystem A.
The effects of cooperativity in the interaction could be estimated from various parameters
such as structural changes undergone by clustering and the shifts in some vibrational
frequencies. The more direct evaluation of the contributions of cooperativity in this paper is
obtained by comparing the interaction energy of the cluster with the pairwise interaction
energies calculated with the whole basis sets for the cluster in order to exclude BSSE [9,10].
Thus, for trimer the cooperativity contribution is,
Enopair=EABC(ABC)- EAB(ABC)- EAC(ABC)- EBC(ABC)
876
where the terms in brackets mean that the whole basis sets for the cluster is used in
calculations.
For this type of study, one should choose an appropriate basis set. Usually a substantial
size of basis set is required for an accurate description of the structures and energies of
clusters. However, the size of the clusters studied in this work excluded the use of very large
sets, and hence we employed the MP2 method with 6-31G* basis set for nitramines, 6-311G*
for hydrazine and difluoroamine, aug-cc-pVDZ for H2O2/H2O. We employ the DFT-B3LYP
method with 6-311++G** for Fox-7 and NTO. All calculations for the gaseous states are
performed with Gaussian98 program [33] at a Compaq Workstation in our laboratory.
The crystal structures obtained from the X-ray diffraction were used for the computation
of the bulks [34-37]. The hybrid density functional of B3LYP was also employed for the
calculations. The split valence 6-31G** basis set was used for Fox-7 and PETN, and 6-21G**
for NTO and ONC. Computational practice for molecule crystal showed that the energy was
essentially constant from the 6-21G level on [38]. The Monkhorst shrinking factor [39] for the
periodic calculation is 12. The threshold for energy convergence is 10-7 a.u. in both periodic
and nonperiodic calculations. The crystal lattice energy is defined as the total energy of the
bulk minus that of gaseous molecule. The basis set superposition error (BSSE) corrected
lattice energy is defined as EC=E+BSSE. The BSSE was obtained with basis functions of
the neighboring atoms up to 0.40 nm. The periodic calculations of crystals employed the
CRYSTAL98 suite of programs [40].
3.

3.1
Gaseous clusters
3.1.1. Optimized geometries

Of all the stable clusters, only the most stable structures of R2N-NO2, H2N-NH2, NHF2,
Fox-7, NTO and H2O2/H2O clusters were shown in Figure 1. Table 1 lists some optimized
geometrical parameters of the clusters. Both the monomers of Me2N-NO2 and H2N-NO2 have
CS symmetry with N-NO2 coplanar. These results are in quite good agreement with all
previous theory work or with low temperature crystalline form but differ from the
experimental gas-phase geometry of C2V symmetry [41,42]. This discrepancy is due to the fact
that the planar or pyramidal nature of amino group is sufficiently weak and capable of altering
the geometry [21]. The hydrazine monomer has C2 symmetry, which is also in agreement with
the experimental result [43]. The most stable structure of DMNA dimer is very similar to that
from the DFT result that emphasized the existence of two ON contacts by 0.308 nm [25]
instead of two additional OH contacts as in Figure 1 - 3a. Both of the ON contacts in 3a
are 0.3097 nm. The Fox-7 dimer, with C1 symmetry, possesses the exact wave-shaped layered
packing in crystal [34]. The NTO possess C2h symmetry. Compared to the geometries of
monomer, all the N-N lengths of nitramines (1, 2, 3a and 3b) decrease by 1.0~2.4 pm,
whereas those of N-O in the intermolecular interaction rings increase by 0.1~1.3 pm. The
changes of bond lengths for 1 and 3b are larger than those
877
42
0.20 O10
H6
H5
N2
0.
N1
O3
2
204
H11
O4
O9
N7
O3
H7
H12
O4
H18
H16
C14
N1
H17
N13
H6 0.2111
O12
N2
H8
N8
0.2582
H15
N10
C5
O11
H9
2
C 17
C6 H
10 N1
H11
O3
O4
0.
27
76
0.2678
C 30
N 26
C18
H23
O3
N1
H 34
C 29
H22
O4
H 32
H 33
H21
O 16
0.2701
H20 C
17
N15
N 14
O 13
H 20
H 22
0.2823
0.2812
N14
O16
H19
C 18
0.2996
0.2693
N2
0.
27
76
N 15
O13
01
0.2678
H8
0.3048
H12
H7
C5
0.2
9
H9
H24
N 25
H8
763
0 .2
0.2668 H 10
O 28
0 .2 6
O 27
3a
N2
C5
C6
42
3b
N5
H6
H11
0.2860
C2
66
0.20
N3
O7 0
.2
3
N6
H14
O10
H25 N19
0.2
H27
252
N20
H28
O9
H26
C16
14
C1
N4
O22
N17
O7
O21
C15
N18
42 2
5b
N6
O8
O23
N2
C3
N4
N1
C5
H11
O3
O2
4
0.183 O20
H10
4
0.183 H21
O9
H22
C16
N12
N15
C14
N13
O19
N17
O18
O24
H1
0.242
0. 2
5a
O8
N5
H10
N1
23
F4
F3
25
0.24
0. 24
N9
7
0.1789 H9
(0.1769)
H13
O8
H
O11 12
H10
H4 0.19
20
(0.
190
0) O5
H7
H6
0.2197
(0.2161)
0.1764
(0.1735)
H13
H12
F12
H2
F11
0.
(0.1 1847
859
)
H11
F3
F4
N1
0.2861
4
42
H6
N5
N8
H12
H2
F7
F8
2
0.
H5
0.2364
0.2531
F7
F8
H9
N7
H6
42
H4
N1
N2
0.2364
H10
0. 2
H3
H7
0.1721
O5
H
(0.1696) 1
H6
8a
O3 H
4
O2
8b
Fig. 1. Optimized structures, intermolecular distances (nm) and atomic numbering of

the clusters (only the intermolecular contact hydrogens are shown for DMNA
trimers, distances in parenthesis of H2O2/H2O clusters are at the MP2/6-311G*
level))
878
Table 1. Optimized geometries of the clusters a

Clusters Parameters and Values
R1-2 134.0(-1.6)
R7-8 134.0(-1.6)
R2-5 100.2(0.4)
R1-4 120.4(1.3)
R7-10 120.3(1.2)
R8-11 100.2(0.4)
R1-2 133.6(-1.1)
R10-11 119.2(-0.1)
R10-13 133.6(-1.1)
R1-3 119.3(0.0)
R10-12 120.1(0.6)
R13-14 145.5(0.4)
R1-4 120.2(0.7)
3a
2-1-4 116.9(0.4)
4-1-2-5 24.8(-0.8)
2-1-5 112.3(1.8)
4-1-2-6 160.3(3.8)
2-1-4 116.6(0.4)
2-1-4-3 178.3(0.2)
10-13-14 119.5(1.3)
4-1-2-5 161.8(1.4)
R1-2 133.3(-1.1)
R2-5 145.9(0.4)
R18-22 108.2(-0.3)
R14-15 133.3(-1.1)
R1-3 119.8(0.1)
R2-6 145.7(0.2)
R13-14 119.8(0.1)
R15-17 145.7(0.2)
R1-4 120.3(0.6)
R5-8 108.2(-0.3)
R14-16 120.3(0.6)
2-1-4 117.8(0.4)
2-5-8 111.7(0.7)
3-1-2-4 178.3(-0.4)
13-14-16-15 178.2(-0.5)
3b
R1-2 132.8(-1.6)
R6-10 108.1(-0.4)
R18-22 108.2(-0.3)
R1-3 120.4(0.7)
R13-14 120.2(0.5)
R25-26 132.0(-2.4)
R1-4 120.0(0.3)
R14-16 119.9(0.2)
R25-28 120.8(1.1)
R2-5 145.7(0.2)
R14-15 133.4(-1.0)
R26-30 145.9(0.4)
R2-6 146.0(0.5)
R15-17 145.8(0.3)
R30-34 108.0(-0.5)
R5-8 108.2(-0.3)
R17-20 108.2(-0.3)
1-2-6 116.5(0.7)
25-26-30 116.8(1.0)
2-6-10 111.5(1.5)
26-30-34 111.5(1.5)
13-14-15 117.7(0.3)
3-1-2-4 177.9(-0.8)
14-15-18 115.9(0.1)
13-14-16-15 177.7(-1.0)
15-17-20 111.5(1.5)
27-25-26-28 178.9(0.2)
3-1-2 117.9(0.5)
27-25-26 118.2(0.8)
R1-2 141.5(0.4)
R1-4 99.8(0.2)
R2-6 100.0(0.1)
5a
R1-2 100.1(0.0)
R5-6 101.1(0.0)
R5-7 134.1(-0.2)
R1-3 134.6(0.3)
5b b
R1-2 100.0(-0.1)
R1-3 134.8(0.5)
R3-7 123.3(1.6)
R19-25 101.1(0.4)
R1-2
R4-9 122.0(0.3)
R20-27 101.0(0.3)
R1-3 141.7(-1.9)
R1-5 138.2(-1.6)
R5-9 122.2(1.8)
2-1-4 107.5(-0.5)
4-1-2-5 156.8(3.1)
2-1-3 101.9(-0.3)
6-5-7 102.8(0.6)
2-1-3 101.9 (-0.3)

143.4(1.2)
R1-10 102.8(2.0)
879
R15-18 143.2(-0.4)
R16-19
134.0(-0.5)
R3-6 145.0(-0.1)
2-1-3-7
161.8(8.1)
2-1-4-9
150.1(-3.7)
R1-2 135.6(-0.2)
8a
R1-2 97.2(0.0)
R3-4 98.1(0.9)
R2-3 147.2(0.2)
8b
R1-2 99.4(2.2)
R2-3 147.3(0.3)
R3-4 97.1(-0.1)
1-2-3 99.2(0.1)
2-3-4 98.3(-0.8)
1-2-3-4 112.1(-0.8)
1-2-3 98.5(-0.6)
2-3-4 99.5(0.4)
1-2-3-4 123.0(11.9)
bond length in pm, bond angle and dihedral angle in degree, values in parentheses
are the differences from the corresponding monomer,
bond lengths of the other two submolecules are nearly as the same as the first submolecule
(0.1pm)
of 2 and 3a, owing to the shorter intermolecular contact distance in 1 and cooperative
effects in 3b. The lengths of the intermolecular N-O and N-H bonds for dimer 6 are elongated
in the dimerization process, with an increase up to 1.6 pm for the N3-O7 bond length. However,
all the C-N bond lengths within the intermolecular interacting rings are shortened upon
complex formation, especially for the lengths of the C-NO2 bonds in dimer 6. The length of
the NCarbonyl of dimer 7 decreases by 1.6 pm, while the intermolecularly contacted NH
length increases by 2.0 pm, CO increases by 1.8 pm. However, there is hardly any change for
the C-NO2 bonds in dimer 7. The X-NO2 (X=N, C, O) bond in nitro explosives is often
referred to as a detonation trigger [12,15,17]. The interaction causes the shortening of X-NO2
bond for clusters 1~3 and 6, especially for structure 3b. It can be speculated that the
interaction may lower the sensitivity in a certain degree. Bond lengths in the intermolecular
H-bond ring in 4 increase by 0.2~0.4 pm. The N-F lengths in 5b increase by 0.5 pm, whereas
all the other bond lengths are nearly unchanged. As compared with the geometry of H2O2, the
lengths of all the intermolecular bound O-H or O-O bonds in 8 increase by 0.2~2.2pm,
whereas the other bond lengths hardly change. It is well known that the decomposition of
hydrogen peroxide is initiated by the breakdown of the peroxide bond. The interaction causes
the elongation of the O-O bond length for all clusters. It can be speculated that the interaction
increases the explosive sensitivity to a certain degree, which is in agreement with the
experimental fact that the sensitivity increases as the percentage of water increases up to 14.0
wt% [44]. In general, shorter intermolecular contact distance or cooperative effect in the system
containing three molecules causes larger change of bond length. The bond angles and the
dihedral angles of all clusters change slightly from their corresponding monomer, which
implies that the influence of interaction on bond bending or internal rotation around single
bond is very small.
3.1.2. Vibrational frequencies and their shifts
The characterized vibrational frequencies (IR), their intensities (KM/mol) and the shifts
from those of monomers are shown in Table 2. For 1 and 3, the frequencies associated with the
N-O stretches exhibit slight red shifts with respect to those of its monomer, accompanied by
880
the increasing of intensities. The shifts demonstrate that the N-O bonds are somewhat
weakened. There are slight blue shifts in the 3242-3370 cm-1 ranges for 3, these modes are
associated with the motion of hydrogen atoms of the methyl group. For 2, the H-N-N rocking
exhibits red shift. The N-H stretch modes in the 3722~3914 cm-1 for 1, 4 and 5b exhibit some
red shifts, and the N-H wagging or rocking mode in 4 shows large red shift (-84 cm-1). The
N-H stretch mode in 5a exhibits slight blue shift owing to its long contacting distance and thus
weak interaction. The N-H asymmetrical stretch mode in 6 exhibits red shift (-31.6 cm-1). The
shifts demonstrate that the N-H bond are somewhat weakened by the intermolecular
interaction. The frequency of ca. 1224 cm-1 in 6 is assigned to the N-C-N rocking, this mode
exhibits a blue shift of ca. 8 cm-1 with large intensity as the result of large dipole moment
change. The blue shift also shows that the C-N bond is slightly strengthened upon the
formation of dimers, which is consistent with the decrease of the C-NO2 bond length caused
by intermolecular interaction. The C-O stretch mode of 7 shows large red shift. For the
water-hydrogen peroxide binary complex, frequencies of 1(Hpo) are associated with the O-H
rocking modes that exhibit large blue shifts (from 164 cm-1 to 205 cm-1) of the corresponding
frequency with respect to the one in H2O2. 2(Hpo) is assigned to the stretching of OH that is
bound by H2O, these modes exhibit very large red shifts from -164 cm-1 to -438 cm-1 with large
intensities as the results of large dipole moment changes. The red shifts also show cooperative
effects as more water molecules are added to the cluster. The shift of the bound OH stretching
of hydrogen peroxide in the binary complex is slightly more than the experimental value of
-135 cm-1 [18]. In contrast, 8s 3(Hpo) are almost the same as in monomer, since these mode
are associated with the stretches of O-H outside the interacting cyclic structure. The mode of
ca. 850 cm-1 associated with O-O stretching exhibits some red shift (within 13 cm-1) with
respect to that of H2O2. The shifts demonstrate that the O-O bonds are somewhat weakened,
which also means that the interaction increases the explosive sensitivity. 1(Aq) is assigned to
the shearing of water OH, this mode exhibits somewhat blue shifts in cluster 6b. 2(Aq) and
3(Aq) are assigned to the OH stretching in water. 2(Aq) is belong to OH in the
intermolecular cyclic structure and displays significant red shifts from 49 cm-1 to 349 cm-1
along with drastic increase in the IR intensity, whereas the unbound OH stretch band 3(Aq)
exhibits only a small red shift. It is known that the normal hydrogen bonding causes a large red
shift in the X-H (X=O, N, F or Cl) stretch frequency. Therefore, there exist H-bonding effect
in 1, 2, 4 and 6~7, but these effect are small or negligible in the other clusters.
Table 2. Vibrational frequencies, shifts and intensities
Clusters
1
2
3a
3b
4
Frequency (shifts, intensity)

N-O,str. /1884(-18, 1291)
N-N,str. /1645(+11, 527)
N-H
/3748~3914(-44~-15, 0~456)
N-N,str. /1653(+5, 226)
H-N-N,rock. /1878(-11, 564)
N-H&C-H
/3736~3869(0~+3, 0~289)
N-O,str. /1816(-16, 913)
N-C-H,rock. /1674(+5, 312)
C-H
/3242~3366(-3~+17, 0~43)
N-O,str. /1812(-20, 994)
N-C-H,rock. /1676(+7, 200)
C-H
/3248~3370(+3~+17, 0~47)
N-H,wag&rock /1009(-84, 171) N-N,str. /1220(-4, 9) N-H /3722~3829(-7~-9,
0~49)
881
5a
5b
6
N-F,str./1177(+13,250) N-H,str. /3770~3772(+6~+7, 10~14)

N-F,str. /1159(-5, 508) N-H,str. /3790~3792(-7~-9, 0~45)
N-H,wag /396(+0.9, 299)
C-N-C,rock /1224(+7.8, 898)
N-H,asym. str.
/3501(-31.6, 711)
N-H,wag /539(+6.5, 192) ring def. /765(+11.5, 191) N-O,str. /1599(-1.7, 564)
C-O str. /1802(-46.7, 2129) N-H,wag /3344(-11.4, 2663) N-H str. /3649(-7.5,
288)
1 Hpo/880 (+164,166.6) 2 Hpo/3438 (-164,329.2) 3 Hpo/3600 (-4,64.2)
4 O-O/854 (-13,65.2) 1 Aq/1721 (-3,90.0)
2 Aq/3591 (-49,84.8) 3 Aq/3700 (-24,116.3)
1 Hpo/922 (+205,130.6) 2 Hpo/3164 (+438,684.9)
3 Hpo/3615 (+11,78.7) 4 O-O/856 (-11,29.6) 1 Aq/1739 (+15,70.6)
2 Aq/3291 (-349,1034.7) 3 Aq/3693 (-31,70.3)
7
8a
8b
a
Positive/negative signs indicate blue/red shifts, respectively
3.1.3. Interaction energies and cooperative effects

Table 3 gives both the uncorrected and corrected interaction energies and the
contributions of cooperativity. There are noimaginary frequencies for all structures in Table 3,
which convicts that all the structures in Figure 1 correspond minima on their potential energy
surfaces. The binding energy of 3a is between the values of -11 kcal/mol and -5 kcal/mol
[24,25]
obtained at the SCF+E(20)
. The proportion of correlated
disp and the DFT levels respectively
interaction energies to their total interaction energies (viz. [E(MP2)- E(HF)]/E(MP2) , see
Table 3) for 1~5 and 8 is at least 18.1 percent, so it is imperative to include the electron
correlation energies into interaction energies. The discrepancies between E(MP2)C and
E(MP2) are at least 9.2 kJ/mol for 1~5 and 8b, indicating that it is also necessary to correct
the BSSE. The ZPE corrections for the interaction energies are generally less than those of
BSSE. The average binding energy for each intermolecular contact in 1, 2, 6, 7 and 8b is larger
than those of others clusters. This means that there exist stronger H-bonds in the former. Based
on the changes of binding energies upon addition of a new molecule to a cluster, the transition
from the dimer to the trimer involves the contributions of cooperativity as in the cases of 3b,
5b. The transition from the dimer to the tetramer involves large cooperative effect in 8b.
TABLE 3
Energies
Interaction energies and cooperative energies (kJ/mol) a

1
2
3a
3b
5a
5b
6
E(HF)
-40.85
-26.54 -29.42
-55.89
-23.10
-14.97
-40.59
E(MP2)
-49.82
-32.70 -48.21
-91.14
-35.96
-21.61
-56.41
(18.1)
(18.8)
(39.0)
(38.7)
(35.8)
(30.7)
(28.0)
E(HF)C
-35.43
-21.72 -18.12
-35.51
-17.92
-9.90
-21.11
E(MP2)C
-36.89
-22.45 -28.37
-53.89
-25.39
-12.37
-25.06
-18.39
-9.19
-19.22
E(MP2)C,ZPE
Enopair,MP2
a
-32.93
-20.11 -24.68
-47.27
-2.97
-2.17
-44.95
-41.04
-38.15
7
-61.14
-58.05
8a
8b
-19.86
-77.79
-31.91
-122.05
(37.8)
(36.3)
-18.02
-70.81
-26.43
-100.83
-53.66
-39.96
The values in parentheses represent [E(MP2)-E(HF)]/E(MP2)100; 6-31G* is used for

nitramines, 6-311G* for hydrazine and difluoroamine, 6-311++G** for Fox-7 and NTO
and aug-cc-pVDZ for H2O/H2O2 clusters. Binding energies for 6 and 7 are E(DFT).
882
3.2
Molecular crystals
3.2.1. Lattice energy

The calculated crystal lattice energies are 122.30, -156.14, -182.25 and -45.52 kJ/mol for
Fox-7, NTO, PETN and ONC respectively. These values become 114.06, -153.43, -139.37
and -40.55 kJ/mol when using a 50% BSSE correction [45]. The 50% BSSE corrected lattice
energy of Fox-7 is in good agreement with both the value of 113.81 kJ/mol, determined from
the heat of sublimation found by Politzer et al, [46] and the value of 115.707 or 114.739
kJ/mol, determined from semiempirical intermolecular potential energy functions by using
LMIN program [47] However, the lattice energy is much larger than the binding energy of the
most stable dimer, which indicates that both the cooperativity of intermolecular interaction
and the dispersion energy between the molecular layers play an important role on the stability
of the bulk. To analyze the contribution of the electron correlation energy to the total binding
energy, single point calculation is carried out on the dimer from the crystalline structure of
Fox-7 at the MP2/6-31G** level. It gives the binding energy of 63.63 kJ/mol, and of 52.71
kJ/mol after corrected for the BSSE. The proportion of correlation energy to the corrected
binding energy is 2.8%. So the influence of electron correlation on the relative energy of
crystalline FOX-7, such as binding energy and lattice energy, is small. The corrected lattice
energy of PETN is also in agreement with the experimental value of 121.76 kJ/mol [48].
883
3.2.2.
Electronic banding structure
The electronic band structures were shown in Fig. 2, and the corresponding total and
projected densities of states (DOS) were given in Fig. 3. As shown in Fig. 2, both the upper
valence bands (the highest occupied crystalline orbitals) and lower conduction bands (the
lowest unoccupied crystalline
Fox-7
NTO
PETN
ONC
Fig. 2. Electronic band structure of the molecular crystals

orbitals) are generally quite flat. The maximum band dispersions are 0.007 a.u. (about 0.19
eV), 0.005 a.u. (about 0.14 eV), 0.002 a.u. (about 0.05 eV) and 0.004 a.u. (about 0.11 eV) for
Fox-7, NTO, PETN and ONC, respectively. The energy gaps between the highest occupied
crystalline orbital (HOCO) and the lowest unoccupied crystalline orbital (LUCO) are 0.147
a.u. (4.0 eV), 0.143 a.u. (3.9 eV), 0.235 a.u. (6.4 eV) and 0.19 a.u. (5.2 eV) for Fox-7, NTO,
884
PETN and ONC, respectively, indicating that the crystal is an electrical insulator since the
band gap is large. The HOCO and LUCO levels are degenerate. When small changes in charge
state occur, some orbitals are stabilized and others are destabilized, which eliminates the
Fox-7
NTO
PETN
ONC
Fig. 3. Projected and total density of states (DOS)

degenerate of the HOCO and LUCO levels and could lead to minor Jahn-Teller distortions
[49,50] A better understanding of the character of these bands can be obtained by projecting
the DOS on the atom-centered orbitals. As shown in Fig. 3, the C1 atom and the nitro oxygen
atom make up the upper valence bands of Fox-7. In contrast, the lower conduction bands
consist of the nitro N and O atoms. In a word, the HOCO and LUCO are formed by
contributions from the C-NO2 group. According to the principle of frontier orbital, the C-NO2
group can be the most reactive part of the molecule, which is agreeable with the deduction
from population analysis. As the occupied bands are formed by the mixed contributions of the
C-NO2 group, there exists a strong conjugated effect in C-NO2 group. It is the conjugated
effect that makes C-NO2 more stable than the commonly used explosives such as HMX, which
also explains the low sensitivity of FOX-7. Most atoms in NTO, with the exception of C5 atom
885
and the nitro atoms, make up the upper valence bands. In contrast, the lower conduction bands
mainly consist of the nitro N and O atoms. As the occupied bands are not formed by
contributions of the C-NO2 group, which is contrary to FOX-7, the conjugation effect between
the NTO ring and nitro group is relatively weak. This explains that the sensitivity of NTO is
higher than that of TATB, although there exists C-NO2 group and strong hydrogen bonding in
both explosives. The oxygen atoms in nitro group of PETN make up the narrow lower energy
bands with small component from the ester oxygen, whereas both the nitrogen and oxygen
atoms in nitro group make up the narrow higher energy bands. Surprisingly, the carbon atoms
in ONC make up both the narrow lower energy bands and the higher energy bands, which is
different from the DOS of non-caged nitro explosives [6]. The projection of DOS indicates
that the carbon skeleton is the most reactive parts of the molecule and is thus prone to
breakdown, which is in agreement with the previous theoretical prediction of pyrolysis
mechanism of ONC [51].
4.
CONCLUSIONS
From ab initio and DFT calculations in the paper, the following conclusions can be drawn:
(1) There exist stronger H-bonds in 1, 2, 6, 7 and 8b. There exist cooperative effects in the
trimers and tetramer.
(2) The stronger the interaction is, the more the bond length in the intermolecular
contacting ring changes, and also more red shifts on the N-H or O-H vibrational mode
take place.
(3) When the X-NO2 bond is within the intermolecular region, the interaction causes the
shortening of X-NO2 bond, as in the clusters of 1~3 and 6. It can be speculated that the
interaction may lower the sensitivity in a certain degree for these clusters. On the
contrary, the O-O bonds in the water/peroxide clusters are somewhat weakened,
which means that the interaction increases the explosive sensitivity.
(4) Both the upper valence bands and lower conduction bands are generally quite flat for
the molecular crystals. The electronic band structure and its projection shows that the
C-NO2 group makes the most contributions to the frontier bands and thus can be the
most reactive part of Fox-7 and PENT, while the occupied bands are not formed by
contributions of the C-NO2 group for NTO and the carbon atoms in ONC make up
both the narrow lower energy bands and the higher energy bands.
Acknowledgements
We gratefully thank the National Natural Science Foundation of China (Grant 20173028
and 10176012) and the Postdoctoral Foundation provided by the Educational Ministry of
China for their support of this work.
886
REFERENCES
[1]
J S LI, H M XIAO, H S DONG: Int J Quantum Chem, 2000, 78, 94
[2]
X H JU, H M XIAO: J Mol Struct (Theochem), 2002, 588, 79
[3]
X H JU, H M XIAO: Chin J Chem, 2002, 20, 227
[4]
X H JU, J J XIAO, H M XIAO: J Mol Struct (Theochem), 2003, 626, 231
[5]
X H JU, H M XIAO, Q Y XIA: J Chem Phys, 2003, 119, 10247
[6]
H M XIAO, X H JU: Intermolecular Interactions in Eenergetic Systems, Science Press, Beijing,

2004
[7]
H M XIAO, X H JU, G Y FANG, L N XU: J Chem Phys, 2004, 121, 12523
[8]
Z LATAJKA, S SCHEINER: J Chem Phys, 1988, 122, 413
[9]
M J ELROD, R J SAYKALLY: Chem Rev, 1994, 94, 1975.
[10] E M CABALEIRO-LAGO, M A RIOS: J Phys Chem A, 1999, 103, 6468.

[11] O GROPEN, P N SKANCKE: Acta Chem Scand, 1971, 5, 1241
[12] H M XIAO: Molecular Orbital Theory for Nitro Compounds, National Defense Industrial Press,
Beijing, 1993
[13] S ZEMAN: J Energ Mater, 1999, 17, 305
[14] B M RICE, S V PAI, J HARE: Combust. Flame, 1999, 118, 445
[15] A DELPUECH, J CHERVILLE: Propellants and Explosives, 1978, 3, 169
[16] B G SUMPTER, D L THOMPSON: J Chem Phys, 1988, 88, 6889
[17] K N FAN, H M XIAO, Y F LI: Acta Chimica Sinica, 1989, 47, 952
[18] Y LING, Z H TANG, H M XIAO: Propellants Explosives Pyrotechnics, 1992, 17, 116
[19] D HABIBOLLAHZADEH, J S MURRAY, P C REDFERN, P POLITZER, J Phys Chem, 1991, 95, 7703
[20] J N HARRIS, K LAMMERTSMA, J Phys Chem A, 1997, 101, 1370
[21] G D SMITH, R K BHARADWAJ, D BEDROV, C AYYAGARI: J Phys Chem B, 1999, 103, 705
[22] V N GREBENNIKOV, G B MANELIS, G M NAZIN: Int Annu Conf ICT 1996, 27th ,8.1.
[23] C HILLEBRAND, M KLESSINGER, M ECKERT-MAKSIC, Z B MAKSIC: J Phys Chem, 1996, 100,
9698
[24] R BUKOWSKI, K SZALEWICZ, C F CHABALOWSKI: J Phys Chem A, 1999, 103, 7322
[25] P POLITZER, M C CONCHA, J S MURRAY: Int J Quantum Chem, 2000, 80, 184
[26] R FLETCHER, M J D POWELL: Comput J, 1963, 6, 163
[27] H B SCHLEGEL: J Comp Chem, 1982, 3, 214
[28] J S BINKLEY, J A POPLE: Int J Quantum Chem, 1975, 9, 229
[29] K RAGHAVACHARI, J A POPLE, E S REPLOGLE, M HEAD-GORDON: J Phys Chem, 1990, 94, 5579
[30] S F BOYS, F BERNARDI: Mol Phys, 1970, 19, 553
[31] A JOHNSON, P KOLLMAN, S ROTHENBERG: Theor Chem Acta, 1973, 29, 167
[32] G CHALASINSKI, M M SZCZESNIAK: Mol Phys, 1988, 63, 205
[33] M J FRISCH, G W TRUCKS, H B SCHLEGEL, ET AL: Gaussian 98 (Revision A.7), Gaussian, Inc.,
Pittsburgh PA, 1998
887
[34] U BEMM, H OSTMARK: Acta Crystallogr C, 1998, 54, 1997

[35] E A ZHUROVA, A A PINKERTON: Acta Crystallogr B, 2001, 57, 359
[36] H H CADY, A C LARSON: Acta Crystallogr B, 1975, 31, 1 864
[37] THE CAMBRIDGE CRYSTALLOGRAPHIC DATA CENTRE, CCDC127763
[38] R DOVESI, M CAUSA, R ORLANDO, C ROETTI: J Chem Phys, 1990, 92, 7402
[39] H J MONKHORST, J D PACK: Phys Rev B, 1976, 13, 5188
[40] V R SAUNDERS, R DOVESI, C ROETTI, ET AL: CRYSTAL98, University of Torino, Torino, Italy,
1998
[41] R STOLEVIK, P RADEMACHER: Acta Chem Scand, 1969, 23, 672
[42] A FILHOL, G BRAVIC, M REY-LAFON, M THOMAS: Acta Crystallogr B, 1980, 36, 575
[43] K KOHATA, T FUKUYAMA, K KUCHITSU: J Phys Chem, 1982, 86, 602
[44] R ENGELKE, S A SHEFFIELD, L L DAVIS: J Phys Chem A, 2000, 104, 6894
[45] K S KIM, P TARAKESHWAR, J Y LEE: Chem Rev, 2000, 100, 4145
[46] P POLITZER, M C CONCHA, M E GRICE, J S MURRAY, P LANE, D HABIBOLLAZADEH: J Mol
Struct (Theochem), 1998, 452, 75
[47] D C SORESCU, J A BOATZ, D L THOMPSON: J Phys Chem A, 2001, 105, 5010
[48] T R GIBBS, A POPOLATO: LASL explosive property data, University of California Press:
California, 1980: 130139
[49] R HOFFMANN: Solids and Surfaces: A Chemists View of Bonding in the Extended Structures,
VCH Publishers, New York, 1988
[50] J KORTUS, M R PEDERSON, S L RICHARDSON: Chem Phys Lett, 2000, 322, 224
[51] J ZHANG, H M XIAO: J Chem Phys, 2002, 116, 10674
888
SYNTHESIS AND APPLICATION

OF GEMINI SURFACTANT IN EMULSION EXPLOSIVE
Ye Zhiwen, Lu Chunxu, and Liu Zuliang
Nanjing University of Science & Technology,
Nanjing, 210094
Abstract:
Cationic Gemini Surfactant was synthesized from N,N-dimethyldodecyltertamine and
1,3-dibromopropane by using ethanol as solvent. After being recrystallized from the
ethyl acetate-ethanol solvent mixture, the Gemini product was qualitatively detected and
verified by IR and 1HNMR. The emulsion explosive, which uses the Gemini surfactant as
crystal modifier and auxiliary emulsifier, has smaller W/O type particle size and better
distribution, stability and explosion properties.
Keywords:
1.
gemini surfactant, synthesis, emulsion explosive, application
INTRODUCTION
Gemini surfactant is a new surfactant, which has special properties due to its special
structure. The traditional surfactant only has one hydrophilic chain and a hydrophobic chain,
while the simple Gemini surfactant possesses two hydrophilic chains and two hydrophobic
chains. Their structures are shown in Fig.1.
O
traditional surfactant
alkyl chain
Fig 1.
Gemini surfactant
O
head group
spacer
Sketch map of structure of the surfactants
Compared with the traditional surfactant, the Gemini surfactant shows good water
solubility and high surface activity (low critical micelle concentration and surface tension).
Emulsion explosive, which was prepared by emulsion technology using water solution
of inorganic salt oxidants such as AN etc as disperse phase, and hydrocarbon fuel as
continuous phase, is a W/O type water-resistant industrial explosive. The application of the
traditional surfactant in emulsion explosive has been studied extensively, e.g. using nonionic
surfactant Span-80 and polyisobutenebibutanimide as emulsifier; using sodium lauryl
phosphate as crystal modifier. It was studied in the paper that the emulsion explosive, which
was prepared using the synthesized Gemini surfactant as crystal modifier and auxiliary
emulsifier, has good storage stability and explosive properties.
889
2.
EXPERIMENTAL
2.1
Reagents and apparatus
Ammonium nitrate (CP), sodium nitrate (CP), 1,3-dibromopropane (CP),

N,N-dimethyldodecyltertamine(CP), ethanol(AR), ethyl acetate, 56# refining paraffin wax,
80# refined microcrystalline wax, H formers soled in market and distilled water.
RE-52 rotating evaporator, FT-IR infrared spectrum detector, Bruker-DRX 300MHz
superconductive nuclear magaetic resonance spectromete, FM-1 emulsifying machine,
DV-I detonation velocity measuring instrument, SEM microscope.
2.2
Synthesis of Gemini Surfactant
Take 46ml of N, N-dimethyldodecyltertamine (0.16mol.l -1), 8ml of 1,3-dibromopropane

(0.77mol.l-1) and 25 ml of anhydrous ethanol into a four-necked flask, fix the condenser tube
and stirring apparatus, and then start the power. The mixture solution was gradually warmed
to 85C and then refluxed for 48 hours. The reaction solution firstly was distilled in vacuum
and the solvent of ethanol was removed completely to afford the crude of Gemini surfactant,
which was then recrystallized for 6 to 8 times from a mixture of ethyl acetate and ethanol to
give crystal. The crystal was dried for 6 to 10 hours at 50C~60C in vacuum oven to yield
42.7 gram (0.068) of the pure of Gemini surfactant.
The equation of reaction is written as follows:
CH3
CH3
C12H25NCH3
BrCH2CH2CH2Br
CH3
CH3N (CH2)3
C12H25
2.3
NCH3 2Br
C12H25
Preparation of Emulsion Explosive
2.3.1. Preparation of hydrous phase

68% ammonium nitrate, 13% sodium nitrate and 10% water was mixed, heated and
dissolved, and then 0.1% Gemini surfactant was added. The temperature was kept at 90C
100C.
2.3.2. Preparation of oil phase
5.5% paraffin wax, 0.5% microcrystalline wax was mixed, heated and melted, and then
2% emulsifier was added, mix-melted and stirred to make them uniform. The temperature
was controlled at 90C100C.
2.3.3. Preparation of Emulsion Explosive
The aqueous phase solution dissolved completely was added to the vessel filled with the
oil described above within 30s. The emulsifying machine was then operated. At the function
of its stirring (rotating velocity is 1500rmin-1) and sheering, the emulsion was formed after
890
3min and then cooled to 55C65C. The photosensitizer of 1% H formers was added to
afford emulsion explosive.
891
3.

3.1
Structure of product
The structure of the synthesized Gemini surfactant was characterized with IR and
HNMR spectroscopy. Its IR spectrum was presented in Fig.2.
90
80
70
T(%)
60
50
40
30
20
4000
3000
2000
1000
/cm-1
Fig 2.
IR spectrum of the Gemini surfactant
From Fig. 2, it can be seen that the IR (KBr) spectrum for the synthesized Gemini
surfactant shows CH3 and >CH2 stretching vibration absorption at 2850 ~ 2980cm-1, the
absorption peaks at 1380~1350cm-1 and 720~730cm-1 demonstrates CH3 and >CH2
bending tortuose vibrations, respectively and the absorption peak at 950cm-1 is due to the
C-N vibration. From so, the IR spectrum of the Gemini surfactant is similar to fatty
hydrocarbon, less absorption and lacking of the useful absorptions for analysis of structure
(except the C-N vibration absorption peak).
7.0
6.5
Fig 3.
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
HNMR spectrum of the Gemini surfactant (CD3Cl)
In Fig. 3, it was seen that the integration of chemical shift of H protons at different
groups were agreed well with the number of proton as expected. The analytic results of the
Gemini surfactant were listed in Table 1.
892
Table 1. Analysis of 1HNMR spectrum of the cationic Gemini surfactant
Item
Carbon chains
Connective groups
CH3
CH2
NC2H2 NCH2 NC3H2 NC2H2 NCH2 N-CH3
0.89
1.34
1.83
3.55
2.75
3.83
3.36
nH
36
12
The synthesized product was the target product by the analyses of IR and 1HNMR
spectra.
3.2
Properties of application in Emulsion explosive
3.2.1. Microstructure
It was observed that the diameter of the W/O type emulsion matrix with no Gemini
surfactant added was at the range from 0.5 to 5m and poor in uniformity, while that with
Gemini surfactant added was 0.2~2m in diameter, had small particles with relatively
uniform distribution. It was because the Gemini surfactant could improve effectively the
crystal state of inorganic salts such as AN etc and make liquid-drop of aqueous solution
surpersmall to inhibit the formation of crystal and make the equilibrium move towards the
direction of dissolution. Finally the small and uniform particles were formed and the Gemini
surfactant played a role of crystal modifier.
3.2.2. Stability of Emulsion explosive
Usually the stability of emulsion explosive was characterized by the times of high-low
temperature (50C,8h ;-45C,16h) cycle test. Ten high-low temperature cycles was equal to
natural storage life of 180days. The contrast results about the tests for the stability between
emulsion explosive with Gemini surfactant added and that with no Gemini Surfactant added
was illustrated in Table 2.
Table 2. Contrast in stability of two different emulsion explosives
Gemini surfactant
High-low temperature cycle/times
Addition
27
No addition
16
From Table 2, it could be seen at the situation of same emulsion technology conditions
and usages, the times of high-low temperature cycle test for the emulsion explosive prepared
with the Gemini surfactant added was far more than that with no Gemini surfactant added,
893
which shown that the emulsion explosive with Gemini surfactant added has stronger
stability.
The stability of emulsion explosive was related close to its microstructure. The size and
distribution of particles was one of important symbols for stability of emulsion explosive.
Compared to the emulsion explosive with no Gemini surfactant added, that with Gemini
surfactant added had relative strong stability due to small size and uniform distribution. In
addition, the lipophilic part of Gemini surfactant lied to the two ends in molecule. When
being emulsified, its polar part would trended to aqueous phase to make the molecule tortile
and the two ends trended to oil phase to generate membrane thickness and form
stereo-resistant membrane. The longer the lipophilic chain was, the more favorable the
action is. The stereo-resistant membrane formed between the reactions of the particles
inhibited effectively the congregation of particles to improve the stability of emulsion
explosive. Thus, the Gemini surfactant played a role of auxiliary emulsifier.
3.2.3. Explosive properties
Two types emulsion explosive were added respectively to paper rollers(32mm20mm)
and their stability were measured. The results was illustrated in Table 3. From Table 3, it
could be been that the stability of emulsion explosive with Gemini surfactant added was
improved obviously. The emulsion explosive prepared in this way, had very small W/O type
particle. The contacting area between aqueous solution of oxidant such as AN etc and oil
phase material was big. Thus, the reaction surface was big to be useful in the process of
explosion reaction.
Table 3. Explosive properties between two types explosion explosive
Gemini
surfactant
Charging Density
/gcm-3
Detonation
velocity /ms-1
Gap Distance
/cm
No addition
Addition
1.15~1.25
1.15~1.25
4250~4350
4950~5050
5~7
7~8
4.
Brisance
/mm
14~16
17~19
DISCUSSION
(1)
Cationic
Gemini
surfactant
was
synthesized
from
N,
N-dimethyldodecyltertamine and 1,3-dibromopropane, and informed by IR
and 1HNMR spectra.
(2)
Compared with common emulsion explosive, at the situation of same usage,

the emulsion explosive prepared by adding the Gemini surfactant had small
W/O type emulsion particles, with relatively uniform distribution and good
stability of liquid membrane. The stability of emulsion explosive was
improved, the times of high-low temperature cycle tests was able to reach 27
and the explosive properties was better obviously. The Gemini surfactant
played roles of crystal modifier and auxiliary emulsifier.
894
REFERENCES
[1]
ROSEN. M. J: Geminis: A new Generation of Surfactants, CHEMTECH, 1993, 23(3): 30~33
[2]
C. X. LV ET AL.: Theory of industrial explosive, 2003.294~301
[3]
X. G. WANG: Emulsion Explosive, 1993.478
[4]
G. C. XU, Z. C. ZHEN: Discussion on stability of emulsion explosive, Explosive Materials, 1997,

26(4): 11~13
895
EFFECTS OF THE REACTION OF THE LINER CONTAINING

THE MAGNESIUM AND ALUMINUM WITH WATER
AND WATER SOLUTION OF AMMONIUM NITRATE
Adam Zakrzewski, and Zenon Wilk
Institute of Organic Industry Branch in Krupski Myn,
Zawadzkiego 1 Str, 42-693 Krupski Myn, Poland
The article presents part of research on axial -direction shaped charges - with metal
powder liners containing powders of aluminum and magnesium. We are presenting the
idea of the association of the cumulative effect with the additional energetic effect.
These effects are possible to obtain by using proper chemical constitution liners. We
present effects obtained on the blasting ground.
Keywords:
1.
shaped charges, aluminium, oil bed stimulation
INTRODUCTION
Shaped charges are mainly applicable in petroleum mining industry and military
applications. Petroleum searching and mining enterprises consume huge amounts of these
charges. Many kinds of shaped charges are used in petroleum mining industry, depending on
requirements (kind of executable work ). Hitherto employed regular shaped charges
contained monolithic liners. Presently liners made of pressured mixtures of powders of
metals (copper, tungsten, bismuth, lead) are used on wide scale. During detonation of shaped
charge - target piercing jet , and the slug are created. The slug moves after the jet with
significantly smaller velocity than the jets velocity. Slug created of metal monolithic liner
may plug the piercing created by jet. Slug formed of metal powder liner has lower abilities
for plugging perforation during getting into it. It does not execute useful work also. We
suggest such liner construction, that will make plugging the perforation impossible. The slug
created of that charge has extra work to do too. The slug created of flammable metal can
react with water or water oxidizer solutions in proper conditions and generate big amount of
gas. It may be useful for additional oil bed stimulation. The stimulation is hitherto executing
after perforating [1,2]. There are many other solutions for oil bed stimulation suggested in
literature and in patent applications [3-7].We suggest association of perforating and
stimulation. Gasses emitted in reaction of flammable metal and washer creates rifts
surrounding perforation, thus the gases preliminary works zone close to the borehole. We
suggest employment of water, or water oxidizer solution as essential for reaction of
flammable metal. Other solutions are suggested and examinated in matter literature and in
patent applications. Using special construction shaped charges, for example containing
multilayer liners is suggested among the others. That kind of liner allow for formation of the
jet which is penetrating hoisting and casing pipes and the slug moving just after the jet. The
slug is made of material which reacts with washer ingredients and giving huge amount of
gases, which (in special conditions) can elongate puncture hole causing simultaneously
rifting of perforations surroundings. Another proposition reliance on set of axial- directional
shaped charges containing liners which include layer of combustible metal in association
896
with oxidizer solution which fills free spaces of perforator gun or is positioned onto the level
of facility workings (in washer). Functioning of charge using up to present was limited to
perforating material to which the charge was directed. In this work we introduce the idea of
combination of cumulative perforating with additional oil bed stimulation.
2.
THEORY
2.1
Shaped charges with multilayer liner
The construction and the rule of the activity of the shaped charge is introduced on the
Fig 1. As it is visible the jet is created of the internal part of the liner, external part of the
liner does not participate in the process of perforating. Hitherto many research on charges
with multi-layer liners were executed. These research showed the very large efficiency of
multilayer liners. In typical construction multi-layer liners the internal layer is made of
material of the greater density then the external. Usually as a material of smaller density is
used powder of copper. We propose the use of such metals as aluminum, magnesium,
zirconium, eventually the beryllium. Worked out in IPO - branch in Krupski Myn
technology of the pressing of liners of powders of metals makes possible the pressing of
liners of any shapes and practically of every materials.
2.2
Aluminum and magnesium as a fuel in red-ox reactions
Metals from the alkaline earth family, Boron family and titanium family are able to react
with some oxidants in suitable conditions. As oxidants can also be used compounds able to
produce oxygen in consequence of their own decomposition. An example of such oxidants
are water and nitrates of alkali metals and the ammonium. In case of multilayer liner we can
use two different materials to create the slug and the jet. An example of shaped charge
construction that let for additional oil bed stimulation and multilayer liner action are given at
Fig1.
Fig 1.
An example of charge containing multilayer liner construction and created

jet together with the slug
Hitherto only the kinetic energy of jets created of the inner part of the liner was
exploited, slug moving just after the jet (created of outer part of the liner) was the problem
only (slugs material might plug the perforations perforated by the jet reducing raw material
production effectiveness) . However, if suitable material was used for making the outer part
of the liner, the material might do additional work. We propose using combustible metals
such as: aluminum, magnesium, and zirconium. Information given in literature and
termochemical data shows that those metals may react with washer and produce huge
897
amount of gases which can also rift oil bed. Rule of action of reactive outer part of the liner
is shown on Fig. 2.
Fig 2.
Scheme of working of outer (reactive) part of the liner
Perforational canal is created after shaped charge detonation. The perforational canal is
infiltrated by detonation products and liner material (slug material). After filling the puncture
by washer or other medium the additional exothermic reactions came and the gases release.
It causes creating in bed extra rifts associated with perforated canal. Below we show
chemical reactions examples and energetic effects accompanying it for the case of using
aluminum as a reactive material.
2Al. + 3H2O Al2O3 + 3H2 + 946,2kJ
(1)
2Al. + 3NH4NO3 Al2O3 + 6H2O + 3N2 + 2023,43kJ
(2)
(Energy values for ambient temperature)

The first reaction (1) may appear for regular washer (main component of such washer is
water). Higher energetic effects may be obtain by using as a washer/ medium or additional
part of a perforating gun ammonium nitrate in water solution (2).
3.
EXPERIMENTAL
In this work we show part of examinations targetted on verification if multilayer liners

containing metal (aluminum, magnesium) powder outer layer production is possible. We
show one of examinations done on firing ground of sample shaped charges containing
multilayer liner of aluminum powder outer layer.
3.1
The characterization of the multilayer liners construction
The 33.3mm diameter liners of 45conical apex angle were done, they were pressed of
metal powders, they had two layers:
1. Jet creating inner part of the liner: electrolytic copper and tungsten agglomerate
ECu/W density c.a. 12.5g/cm3, mass 16g.
2. The outer reactive part: aluminum powder (Al).
898
Total mass of liner c.a. 30g. The liners were matrix pressed . On the previously created
enhanced density (ECu/W) liner the outer aluminum layer was pressed using pressure c.a.
500-600MPa.
3.2
Constructional parameters of model shaped charges
The model shaped charges, in the form of pressed shaped pieces non confinement , with
an use of previously made multilayer liners (ECu/W + Al) were done. For comparison
purposes the model charges containing regular liners of electrolytic copper agglomerate were
done. The characterization of charges:
- outer diameter: 33.3mm (the same as for liner),
- mass of explosive: 42.5g, density: 1.72g/cm3.
For charges filling the heksoflen 3.5%FP (RDX96%, fluorinated polymer 3.5%) was
used. The charge were pressed at 250MPa pressure at ambient temperature. The sight of
multilayer liner (ECu/W+Al) and model shaped charges is exposed on Fig 3 and 4.
Fig 3. The sight of multilayer liner

(ECu/W+Al)
3.3
Fig 4. Model shaped charges (without

aluminum on the left, with liner
containing aluminum- on the right)
Research on model shaped charges
Statical examinations of model shaped charge were done on special stand putting the
charge on barrier created of St3 steel plates (altogether 36mm). Flashing effect of the
charges working after puncturing the barrier was examined. To attain it the recording of
charges detonating by means of digital camera were done. Results of recording of flashing
effect for examinated charge are shown on right side of Fig 5 (on the left side the same
charge, but without aluminum).
Fig.5.
899
Fig 5. Efects of detonation of model charges with regular liner (left)

and multilayer liner containing aluminum (right)
4.
CONCLUSIONS
In exhibited examination of rare, experimental batch of model shaped charges with

aluminum containing multilayer liners we demonstrated possibility of making such
construction. Multilayer liners can be done by matrix pressing. Such technology ensure
indispensable precision of liners production and also the most sparingly production and
without loosing raw material in substance. Outcomes of detonation recording of model
shaped charges illustrates very distincly visible flashing effect for liner containing aluminum
outer layer. It suggests that exothermic effect appeared. Further examinations of this kind of
shaped charge with multilayer liners, and also using other metal or chemical compounds
powders is useful. Continuation of those research of Institute of Organic Industry, may give
new constructions and new kinds of shaped charges in effect.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
K. GRIESGRABER: Niekonwencjonalne metody stymulacji przypywu mediw do odwiertu prochowe generatory. Proc. of Int. Conf. GEOPETROL 2002, Krakw, No 116, 2002.
. . .: - .
, 106, 2003.A. P.
Z. WILK: Badania granatw kumulacyjno-odamkowych o wymiarze 38mm z dodatkowym
efektem zapalajcym. Proc. of Conf. AMUNICJA 2002, Wydawnictwo Politechniki
Poznaskiej, 2002.
Z. WILK: Badania w zakresie technologii metalurgii proszkw i wytwarzania spiekw dla
zastosowa w adunkach kumulacyjnych do perforacji odwiertw geologicznych i dla celw
specjalnych. Work No. EMC 011400013, IPO branch in Krupski Myn, 2003.
Patent US4498367, Energy Transfer Through a Multi-layer Liner for Shaped Charges.
Patent US5259317, Hollow Charge with Detonation Wave Guide.
Patent US5690171, Wellbore Stimulation and Completion.
900
STUDY OF THERMAL AND DETONATION REACTIVITIES

OF THE MIXTURES CONTAINING
1,3,5-TRINITROSO-1,3,5-TRIAZINANE (TMTA)
Svatopluk Zeman, and Rbert Varga
Department of Theory and Technology of Explosives, University of Pardubice,
Abstract:
Thermal reactivities of 1,3,5-trinitroso-1,3,5-triazinane (TMTA), 1,3,5-trinitro-1,3,5triazinane (RDX), 2,4,6-trinitrotoluene (TNT), TNT/TMTA and RDX/TMTA mixtures
and mixtures of 1,3,5-trinitrobenzene (TNB) with TMTA and TMTA/RDX were specified
by means of differential thermal analysis with outputs evaluation by the Kissinger
method. The reactivities, expressed as Ea R-1 slopes of Kissinger relationship, correlate
with squares of detonation velocities of the corresponding explosive samples in the
sense of modified Evans-Polanyi-Semenov equation. Taking this fact it is stated that
initiation of detonation of the mixtures with TMTA content proceeds through primarily
fission of the TMTA molecule. If an ionic mechanism dominantly participates on thermal
decomposition of some studied mixtures then resulting Ea R-1values do not correlate in
the sense of the equation. This is the case of the TNT/TMTA mixtures where TNT has
acidic character toward TMTA. Also in TNB/TMTA/RDX mixture might react amine
intermediates of TMTA thermal decomposition with RDX by ionic mechanism.
Keywords:
1.
1,3,5-trinitrobenzene, 1,3,5-trinitroso-1,3,5-triazinane, 1,3,5-trinitro1,3,5-triazinane, 2,4,6-trinitrotoluene, detonation, DTA, explosives,

initiation, thermal reactivity, TMTA.
INTRODUCTION
The homolytic character and the identity of the primary fission in both the lowtemperature thermal decomposition and the detonation of energetic materials [1-8] were a
motive for Zeman et al. to use the Evans-Polanyi-Semenov equation (E-P-S) [9,10] in the
study of the chemical micro-mechanism governing the initiation of detonation of energetic
materials [2-5,8,11]. The original E-P-S equation describes a relationship between the activation
energies, E, of most substitution reactions of free radicals and the corresponding heats of
reaction, H, [9,10]:
E = B H
(1)
where B is the constant for the given homological series, being a non-dimensional
gradient [9,10]. It is valid for narrow sets of substance structures and documents that the
strength of bond being split is a decisive factor in the given reaction [9,10]. Substitution of H
by heat of explosion Q and E by activation energy, Ea, of the low-temperature thermal
decomposition has led to the first version of the modified E-P-S eqn.[2-5] in the general form:
Ea = C .Q
(2)
where Ea and C are in kJ mol-1 and (the difference between eqn. (1)) is in g mol-1 [2-5].
Eqn. (2) is applicable for the detonation transformation of energetic materials [2-5,8,11].
901
The heat of explosion in this relationship can be substituted by the square of the
detonation rate, D2, because exists definiendum [12,13]:
Q = D2 {2 (2 + 1)}1
(3)
where is the polytropy coefficient which value for high explosives ranges from 2,79 to
3,48 [12].
For exploring relationships of the type of Eq. (2) we can use the results of differential
thermal analysis (DTA) [2,8,11]. In the case of plastic and commercial (oxidizing system on
the basis of ammonium nitrate) explosives [8,11], the differential thermal analysis (DTA) with
evaluation according to Kissinger method [14] proved useful. The method plots ln (/Tmax2)
against reciprocal peak temperature, Tmax, for a series of experiments at different heating
rates, (see also Fig. 2 in the present paper). Using the slope in the Kissinger relationship,
i.e. Ea.R1, and taking into account Eq. (3), Eq. (2) was modified into the following form
[8,11]
:
Ea.R1 = b a.D2
(4)
where the D is the experimentally determined values of the rate of detonation. On the
basis of this approach it was stated [8] that initiation of the detonation of explosives based on
mixtures of a fuel and /or high explosive with an ammonium nitrate oxidizing system
proceeds primarily through decomposition of this system (this statement was also verified by
results from a study of interactions of the nitros- and nitramines with ammonium nitrate [15]).
In the present paper, the above-mentioned method and Eq. (4) have been used to study the
detonation reactivity of explosive mixtures with content of 1,3,5-trinitroso-1,3,5-triazinane
(R-salt, TMTA). In this respect, the present paper can be considered as a continuation of the
previous paper [8]. Presented topic is, at the same time, an adjacent ascent of the forensic
analysis study of the post-explosion TMTA products [16 ].
2.
2.1
Substances
The 2,4,6-trinitrotoluene (TNT), 1,3,5-trinitro-1,3,5-triazinane (RDX) and 1,3,5trinitrobenzene (TNB), all the military grade, and pure 1,3,5-trinitroso-1,3,5-triazinane
(TMTA or R-salt) were used. TMTA is a solid substance crystallizing in the form of drab
yellow needles with the melting point of 105-107C [17]; it was prepared by nitrosolysis of
hexamethylenetetramine according to the published method [21]. The TMTA is sensitive to
the trace of acidity [17-20] and in the molten state it may react with such metals as iron, copper
and aluminium [17,20].
2.2
Explosive mixtures
Explosive mixtures were prepared by using the casting technology from the used
individual explosives (TNT, TNB, TMTA and RDX). A survey of the tested explosives,
their rates of detonation and characteristics of thermal reactivities are shown in Table 1.
The procedure of casting was taken from paper [22]: a preparation of casting mixture of
73 % wt. RDX, 1 % wt. anthracene (AnT) and 1 % wt. diphenylamine (DPA) was carried
out by incorporating dry RDX into melted casting matrix (i. e. TMTA + AnT + DPA). The
matrix was melted in a boiling water bath with continuous stirring. Analogous incorporating
of TMTA into melted TNT was used in the case of the TNT/TMTA (50:50) mixture
preparing which DTA record is presented in Fig. 1.
902
Casting mixture TMTA with TNB was prepared by synchronous bringing of the two
components into stirred TNB/TMTA melt on the water bath (75-80 C); the melt was
prepared by evaporation of a solvent from solution of the mixture TNB with TMTA in
acetone. The TNB/TMTA melt had a smell of amines and formaldehyde during its
preparation. Onset of its melting is 66,3 C and corresponding peak lies at 78,4 C.
Mixture TNB/TMTA/RDX was obtained by mechanical mixing of powdered casting
mixture TNB/TMTA with RDX.
2.3
Measurement of detonation velocity
The time (discontinuous) method was used to measure the detonation velocity of casting
military TNT, casting TMTA and explosive mixtures TNT/TMTA and RDX/TMTA. This
method is based on the registration of passing shock wave given by the length part of the
charge [23]. The measured values of detonation velocities for tested explosives are given in
Table 1. The detonation velocities of pure TMTA [19] and RDX [24] were taken from
literature.
For all the samples studied the detonation velocities were also calculated by means of
the well know Kamlet & Jacobs method [25] for maximum theoretical density (TMD) of the
given explosive. Corresponding results are summarized in Table 1, too.
Table 1. Composition of studied explosives, their detonation velocities and thermal
stabilities (Ea R-1)
Detonation velocity
Sample name
Code
design.
Kissinger
slope
Ea/R (K)
2,4,6-Trinitrotoluene
TNT
15724
6,47
7,20
1,3,5-Trinitro-1,3,5-triazinane
1,3,5-Trinitroso-1,3,5-triazinane cryst.
1,3,5-Trinitroso-1,3,5-triazinane
casting
Casting mixture of 75 % RDX
+ 23 % TMTA + 1 % Anthracene
+ 1 % Diphenylamine
Casting mixture of 58 %TNT
+ 42 %TMTA
Casting mixture of 50 % TNB
+ 50 %TMTA
Casting mixture of 33,33 %TNT
+ 33,33 %TMTA + 33,33 % RDX
RDX
TMTA pure
32211
15572
8,70 a
7,50 7,80 b
8,85
7,67
TMTA cast.
12456
7,30
7,67
RDX/TMTA
20850
8,00
8,22
TNT/TMTA
15363
6,95
7,30
TNB/TMTA
13243
7,45
TNB/TMTA/
RDX
33403
7,88
(km s-1)
experimental calculated
Notes to Table 1: a) data taken from [19,24]

b) data taken from [17].
2.4
Non-isothermal differential thermal analysis (DTA)
We used a DTA 550 Ex apparatus [15,26] specially developed at the Department of Theory
and Technology of Explosives for thermal analyses of explosives. The measurements were
carried out at atmospheric pressure; the tested sample was being in direct contact with the air
atmosphere.
903
The tested sample (0,05 g) was placed in a test tube made of Simax glass, 5 mm in
diameter and 50 mm in length. The reference standard was 0,05 g aluminum oxide. We used
linear rates of temperature increase, viz. 5, 10, and 15 C min1. The results of these
measurements were evaluated by means of the software delivered with the DTA apparatus
[26]
. Examples of the corresponding DTA records are presented in Fig. 1. The results
obtained were treated using the Kissinger method [14]; the treatment for pure substances is
graphically presented in Fig. 2.
The values Ea R1 thus obtained (see in Table 1) were then plotted against the squares of
the experimental values of detonation velocities of the substances and studied mixtures (see
Fig. 3).
Fig 1.
3.
DTA records of the explosive mixture of TNT/TMTA at the heating rates

5, 10 and 15 C/min and samples mass of 0,05g
On the basis of mixture RDX with TMTA it is possible, without bigger problems, to
prepare a castable explosive with RDX content up to 75 % wt. [22]. The explosive is very
powerful, detonation parameters and performance superior to mixture RDX/TNT 75/25 [22].
Its impact sensitivity is lower than in the case of TNT [22]. Fig. 2 shows that the thermal
stability of the RDX/TMTA mixture is higher than the stability of TMTA; stabilising effect
of diphenylamine in this mixture has a positive influence on this stabilisation. On the other
hand, the same Fig. 2 documents reversed influence of thermal stress of TMTA on its
thermal reactivity (melting TMTA cast is more reactive - corresponding line in the Fig. 2 lies
starboard to line of the pure substance). This stress should be expressed also in the change of
detonation velocity of TMTA (see Table 1 and Fig. 3). The said change, of course, can lie
also in density decreasing ( = 1,37 g.cm-3) of the final TMTA charge due to its aeration by
gaseous products of TMTA decomposition.
904
-9,4
y = -20850x + 33,654
2
R = 0,9564
RDX/TMTA TMTAcast
y = -12456x + 16,78
2
R = 0,9996
-9,6
-9,8
TNT
ln (/T )
-10
y = -15724x + 16,614
2
R = 0,9956
-10,2
-10,4
-10,6
TNT/TMTA
y = -15363x + 24,958
2
R = 0,9997
RDX
TMTApure
y = -32211x + 55,242
2
R = 0,998
-10,8
y = -15572x + 23,166
2
R = 0,9999
TNB/TMTA/RDX
y = -33403x + 59,077
2
R = 0,9994
-11
-11,2
0,0016
0,0017
0,0018
0,0019
0,002
-1
TNB/TMTA
y = -13243x + 17,281
2
R = 0,9903
0,0021
0,0022
0,0023
0,0024
-1
T /K
Fig 2.
DTA results treated using the Kissinger method [14], here is the rate of
temperature increase and T is the peak temperature
85
Square of detonation velocity D2 / km2s-2
80
RDX
Line for calculated D values
75
y = 0,0011x + 42,761
2
R = 0,962
70
55
50
y = 0,0011x + 40,242
2
R = 0,9622
RDX/TMTA
65
60
Line for experimental D values
TMTAcast
TMTApure
TNB/TMTA/RDX
TMTApure
TNB/TMTA TNT/TMTA
TMTAcast
TNT calcd
TNT/TMTA
45
40
11000
TNTexperim
16000
21000
26000
Kissinger's slope E a R
Fig 3.
-1
31000
36000
/K
Modified Evans-Polanyi-Semenov equation for the studied samples
Fig. 3 documents the initiation of the detonation of explosives based on mixtures with
TMTA content proceeds primarily through decomposition of this polynitrosamine. As well
as in nitramines thermal decomposition (see Ref. 1 and references herein, namely Ref. 33)
the published values of Arrhenius parameters for this decomposition of nitrosamines in the
condensed state correspond to the primary homolysis of NN bond in their molecules [27-30].
905
Fig. 3 also shows correlation on the basis of D values calculated for maximum theoretical
crystal density (TMD) by means of well know Kamlet & Jacobs method [25].
It is logical that this calculation could not regard chemical changes in TMTA sample,
which was submitted to thermal stress (see difference between the TMTA pure and TMTA cast
data in Table. 1 and Fig. 3). The indicated relationship in Fig. 3 fully corresponds to findings
obtained by analogous study of mixtures on the basis of ammonium nitrate [8].
It follows also from Fig. 2 that mixture of TNT with TMTA (TNT/TMTA) is the most
reactive from all the studied samples. It might be caused by acidic character of TNT due to
electronic transfer (see Fig. 4) [31-33]:
Fig 4.
Acidic character of TNT due to hydrogen atom of methyl group migration
TNT behaves toward TMTA as acid although the existence of the aci-form of TNT
should be not favored [34]. This fact is in a good accordance with already mentioned
sensitivity of TMTA to the trace of acidity [17-19]. A melt of the TNT and TMTA mixture
slowly decomposes already in the temperature region near its melting point: blowholes
sporadically escape from the melt as early as at 55 C without exothermal effect (see Fig. 1)
and without smell. The melt mixture, step by step, gets brown. This effect has to manifest
oneself in outputs from DTA of the mixture it might be a reason of nonexistent correlation
of the corresponding Ea/R value in the sense of modified E-P-S equation (see Fig. 3).
The TNT data dont correlate in the sense of E-P-S equation in Fig. 3. Primary fission of
the TNT molecule in its initiation processes should lie namely in NOH bond homolysis in
molecule of its aci-form (so called trinitrotoluene mechanism) [31-33].
Analogous to TNT also TNB has a negative influence on TMTA stability. Preparation of
TNB/TMTA melt was connected with an obvious escape of formaldehyde and amines from
the melt mixture. In the case of TNT this escape wasnt registered because methyl group of
this polynitro arene is able to participate in aldol and subsequently in N-Mannich
condensations [35]. As the reactions are not radical, corresponding Ea/R value cannot correlate
in the sense of E-P-S equation (see Fig. 3).
Amines formation in the TNB/TMTA melt can be the reason of higher thermal reactivity
of TNB/TMTA/RDX mixture compared to expectation; RDX has no use for alkaline media
in which decomposition is labile [36,37]. Mechanism of the decomposition is ionic [36]. It
means that in the thermal decomposition of TNB/TMTA/RDX mixture should dominate the
ionic mechanisms and, therefore, characteristic of this decomposition (i. e. slope Ea/R)
cannot correlate with the E-P-S equation (see Fig.3).
906
4.
CONCLUSION
The initiation of detonation of the explosives based on mixtures of 1,3,5-trinitroso-1,3,5triazinane (TMTA) with 1,3,5-trinitro-1,3,5-triazinane (RDX) and/or 2,4,6-trinitrotoluene
(TNT) or 1,3,5-trinitrobenzene (TNB) proceeds primarily through decomposition of the
TMTA. Modification of the Evans-Polanyi-Semenov equation, in which the activation
energy is substituted by a slope Ea R-1 of Kissingers relationship and at the same time the
heat of reaction by square of detonation velocity, D2 is applicable to study the mechanism of
these explosive mixtures detonation.
However, thermal decomposition characteristics of TMTA mixtures do not correlate
with the equation if ionic mechanisms dominate in this decomposition. This is the case of
TNT/TMTA mixture where the TNT behaves toward TMTA as acid. Corresponding
mixtures slowly decompose already in the temperature region up to 55 C. Mixture of
TNB/TMTA/RDX is another example, in which amine intermediates of TMTA
decomposition might react with RDX by ionic reactions.
Acknowledgement
The work was carried out under the auspices of the Ministry of Education, Youth &
Sports of the Czech Republic as a part of its Research project No. MSM 0021627501.
907
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
S. ZEMAN: A Study of Chemical Micro-Mechanisms of Initiation of Organic Polynitro

Compounds, in: P. Politzer and J. Murray (Eds.), Energetic Materials, Part II, Elsevier Sci. BV,
2003, p. 25.
S. ZEMAN, M. DIMUN, . TRUCHLK: The Relationship between Kinetic Data of the LowTemperature Thermolysis and the Heats of Explosion of Organic Polynitro Compounds,
S. ZEMAN, M. DIMUN, . TRUCHLIK, V. KABTOV: The Relationship between Kinetic Data of
the Low-Temperature Thermolysis and the Heats of Explosion of Inorganic Azides,
S. ZEMAN, M. DIMUN, V. KABTOV, . TRUCHLIK: Correlation of Activation Energies of
Low-Temperature Thermolysis and Photolysis of some Fulminates with their Heats of
Explosion, Thermochim. Acta, 81 (1984) 359.
S. ZEMAN: Modified Evans-Polanyi-Semenov Relationship in the Study of Chemical
Micromechanism Governing Detonation Initiation of Individual Energetic Materials,
S. ZEMAN: Kinetic Data from Low-Temperature Thermolysis in the Study of Microscopic
Initiation Mechanism of the Detonation of Organic Polynitro Compounds, Thermochim. Acta,
49 (1981) 219.
S. ZEMAN, . GAZDA, A. TOLCOV, A. DIMUN: Dependence on Temperature of the Results of
the Vacuum Stability Test for Explosives, Thermochim. Acta, 247 (1994) 447.
S. ZEMAN, P. KOHLEK, M. MARANDA: Study of Chemical Micromechanism Governing
Detonation Initiation of Condensed Explosive Mixtures by means of Differential Thermal
Analysis, Thermochim. Acta, 398 (2003) 185-194.
N. N. SEMENOV: O nekotorykh problemakh khimicheskoy kinetiki i reaktsionnoy sposobnosti
(Some Problems of Chemical Kinetics and of Reaction Capability). USSR Acad. Sci., Moscow,
1958, pp. 41 and 101, see also German version: N. N. Semjonow, Einige Probleme der
chemischen Kinetik und Reaktionsfhigkeit. Akademie-Verlag, Berlin, 1961, pp. 27, 84.
A. A. ZAVITAS: Energy Barriers to Chemical Reactions. Why, How, How Much?, Chemtech,
ACS, Washington, 1972, p. 434.
M. KRUPKA, M. CHOVANCOV, Z. JALOV, S. ZEMAN: Description of some explosive
characteristics of plastic explosives. Part II. Nonisothermal DTA and TGA in relation to
detonation characteristics. Proc. 3rd Seminar New Trends in Research of Energetic Materials,
Univ. Pardubice, April 2000, p. 274.
D. PRICE: Dependence of Damage Effects upon Detonation Parameters of Organic High
Explosives, Chem. Revs. 59 (1959) 801.
P. A. PERSON, R. HOLMBERG, J. LEE: Rock Blasting and Explosives Engineering. CRC Press,
Boca Raton, 1994.
H. E. KISSINGER: Reaction Kinetics in Differential Thermal Analysis, Anal. Chem. 29 (1957)
1702.
S. ZEMAN, Y. SHU, Z. FRIEDL, J. VGENKNECHT: Thermal Reactivity of some Nitro- and
Nitroso-Compounds Derived from 1,3,5,7-Tetraazabicyclo[3.3.1]nonane at Contamination by
Ammonium Nitrate, J. Hazard. Mater. (2005) in press.
908
[16] R. VARGA, S. ZEMAN, J. VGENKNECHT, M. KOLOVRI: Application of LC-MSD and GCECD Techniques in the Analysis of Post Explosion Residues of Explosives on the Basis of 1,3,5trinitroso-1,3,5-triazinane, Proc. 8th Seminar New Trends in Research of Energetic
Materials, Univ. Pardubice, April 2005.
[17] L. MDARD, M. DUTOUR: dute des Proprites de la Cyclo-trimthylne-trinitrosamine, Mm.
Poudres 37 (1955) 19-24.
[18] H. TADA: Decomposition Reaction of 1,3,5-Trinitrosotriazacyclohexane by Acid, J. Am. Chem.
Soc. 82 (1960) 82 263-6.
[19] T. URBANSKI: Chemistry and Technology of Explosives, Vol. 3, Pergamon Press, Oxford, p.
121, 1986.
[20] B. T. FEDOROFF, O. E. SHEFFIELD: Encyclopedia of Explosives and Related Items, Vol. 3,
PATR 2700, Picatinny Arsenal, Dover, N.J., 1966, p. C630.
[21] W. E. BACHMANN, N. C. DENO: The Nitrosation of Hexamethylentetramine and Related
Compounds, J. Amer. Chem. Soc. 70 (1951) 2777.
[22] M. HANUS: Characterization of Castable System RDX/R-salt, in J. Hansson (Ed.), Proc. 10th
Symp. Chem. Problems Connected Stab. Explos., Margretetrop, 1995, p. 35.
[23] M. SUESKA: Test Methods for Explosives, Springer, New York, 1995, p. 91.
[24] B. M. DOBRATZ, P.C. CRAWFORD: LLNL Explosives Handbook - Properties of Chemical
Explosives and Explosive Simulants, University of California, Livermore, CA., 1985, p.19-131.
[25] M. J. KAMLET, S. J. JACOBS: Chemistry of Detonation: Simple Method of Calculation
Properties of CHNO Explosives, J. Chem. Phys., 48 (1968) 23.
[26] M. KRUPKA: Devices and Equipment for Testing of Energetic Materials, Proc. 4th Seminar
New Trends in Research of Energetic Materials, Univ. Pardubice, April 2001, p. 222.
[27] A. TALL, S. ZEMAN: Thermal Decomposition of Some Nitrosamines, Thermochim. Acta, 93
(1985) 25.
[28] S. ZEMAN, M. DIMUN: On the Study of Micromechanism Governing Detonation Initiation of
Nitramines and Nitrosamines, Propellants, Explos., Pyrotech., 15 (1990) 217.
[29] B. A. LURE, L. M. FEDOTINA, B. S. SVETLOV: Characteristics of the Thermal Decomposition
of Cyclotrimethylenetrinitrosamine and Dinitrosopiperazine, Tr. Mosk. Khim.-Tekhnol. Inst.
Mendeleeva, 104 (1979) 5.
[30] G. M. NAZIN, G. B. MANELIS: Thermal Decomposition of Aliphatic Nitrocompounds, Russ.
Chem. Revs. 63 (1994) 313.
[31] V. G. MATVEEV, V. V. DUBIKHIN, G. M. NAZIN: Soglasovanyi mekhanism razlozheniya
aromaticheskikh nitrosoedinii v gazovoy faze (Thermolysis Mechanism of the Aromatic
Nitrocompounds in the Gas Phase), Izv. Akad. Nauk SSSR, Ser. Khim., (1978) 474.
[32] S. BULUSU, T. AXENROD: Electron Impact Fragmentation Mechanism of 2,4,6-TNT Derived
from Metastable ransitions and Isotopic Labeling, Org. Mass Spectrom., 14 (1979) 585.
[33] G. B. MANELIS, G. M. NAZIN, YU. I. RUBTSOV, V. A. STRUNIN: Termicheskoe razlozhenie i
gorenie vzryvchatykh veschestv i porokhov (Thermal Decomposition and Comustion of
Explosives and Powders), Izdat. Nauka, Moscow, 1996.
[34] P. C. CHEN, W. LO: Molecular Orbital Studies of the Isomers of 2,4,6-Trinitrotoluene and some
its Thermal Decomposition Products, J. Mol. Struct. (THEOCHEM) 397 (1997) 21-32.
[35] S. M. KAY: Encyclopedia of Explosives and Related Items, Vol. 9, PATR 2700, US Army
Armament Res. & Develop. Command, Doverr, N. J., 1980, p. T249.
[36] T. URBASKI: Chjemistry and Technology of Explosives, Vol. 3, Pergamon Press, Oxford,
1986, p. 82.
[37] T. URBASKI: Chjemistry and Technology of Explosives, Vol. 4, Pergamon Press, Oxford,
1985, p. 374.
909
STUDY ON MECHANISMS OF INITIATION

OF TWO POLYNITRO ARENES
Svatopluk Zeman*, Yuanjie Shu**, and Xinfeng Wang**.
* Department of Theory and Technology of Explosives,
University of Pardubice CZ-532 10 Pardubice, Czech Republic
** Institute of Chemical Materials CAEP, P. O. Box 919-301, Mianyang,
Sichuan 621900, P. R. China
Abstract:
Each from the 2,6-bis(2,4,6-trinitrophenylamino)-3,5-dinitropyridine (PYX) and 2,4,6tris(3-methyl-2,4,6-trinitrophenylamino)-1,3,5-triazine (TMPM) molecules contains two
potential centres of the primary fission in its initiation. This fission should be realized by the
migration of -hydrogen atom to oxygen atom of ortho-nitro group (trinitrotoluene
mechanism). Taking of the relationships between impact sensitivity and 13C NMR chemical
shits of some polynitro arenes and between Mulliken charges on nitrogen atoms of primarily
reacting nitro group and onsets from differential thermal analysis of the said compounds,
with the aid of DFT-B3LYP/3-21G methods of GAUSSIAN 98/03 program, the more
probable pathways of initiation of the above-mentioned molecules are estimated.
Keywords:
1.
mechanism, initiation, 2,6-bis(2,4,6-trinitrophenylamino)-3,5-dinitropyridine,

2,4,6-tris(3-methyl-2,4,6-trinitrophenylamino)-1,3,5-triazine
INTRODUCTION
It is well-known that nitro groups are centers of reactivity in organic polynitro compounds
(see Ref. 1 and references herein). It is also generally known that the electronic configuration
and steric conditions play decisive roles within the reaction centre of the molecule during its
primary fission (in the case of polynitro arenes see Ref. 2). Both the factors can be characterized
by means of the NMR chemical shifts of key atoms in the reaction center. Therefore, it is logical
that relationships were found between the 15N NMR chemical shifts of the nitrogen atoms in the
primarily reacting nitramino groupings and the characteristics of the thermal [1,3-5], impact [1,5]
and shock [1,6] reactivities and also to the sensitivity to electric spark [1,7] of nitramines. Similar
application of the 13C NMR chemical shifts of the carbon atoms, carrying the primarily reacting
nitro groups in polynitro arenas, is possible too [1,7].
Also relationships were specified and interpreted between electron charges at nitrogen
atoms of the primarily reacting nitro group, on the one hand, and Kissinger activation energies,
Ea.R-1, of the thermal decomposition of nitramines [8] or detonation characteristics of m-dinitrobenzopolyazaarenes [9] or onsets of thermal decomposition of polynitro arenes [10], on the other.
The complexity of molecular structure of polynitro arenes (besides nitramines) makes the
problem of their primary fission somewhat complicated: if a molecule of these compounds
contains several types of substituents, it can contain several potential reaction centers [1]. It is the
case also of 2,6-bis(2,4,6-trinitrophenylamino)-3,5-dinitropyridine (PYX) or 2,4,6-tris(3-
910
methyl-2,4,6-trinitrophenylamino)-1,3,5-triazine (TMPM) [1]. The paper deals with evaluation of

the said problem of the two compounds.
2.
DATA SOURCES
2.1
Impact reactivity
Fig. 3 represents a relationship between impact sensitivity, Edr (detected by sound), and 13C
NMR chemical shifts, C, of carbon atoms carrying the primarily reacting nitro group in
molecules of polynitro arenes. It was taken from Ref. 1 and reconditioned by means of the C
and Edr values from this Ref.
2.2
Thermal stability
Fig. 4 is taken from Ref. 10 and shows relationships between electronic charges, qN, at
nitrogen atoms of the primarily reacted nitro group (calculated by means of the ab initio DFT
B3LYP/6-31G** method) and onsets, TD, of thermal decomposition from differential thermal
analysis (DTA) of polynitro arenes. These relationships, whose existence might be connected
with the electrostatic interaction of instantaneous point dipoles at the reaction centre of their
molecules, are interpreted in Ref. 10.
2.3
Quantum chemical calculation
The critical points of geometric structures (e. g. reactants, products and transition states) on
the potential energy surfaces of all possible channels of thermal decomposition (PYX-I, PYX-II
see Fig. 1 and TMPM-III, TMPM-IV see Fig. 2) were optimized employing DFT-B3LYP/321G methods of GAUSSIAN 98/03 program package [11]. Using transition state theory (TST) [12]
the rate constants were obtained which are presented in Tables 1 and 2.
O
Channel II
NO2
O
H
H
N
Channel I
O
+
Pi
O 2N
homolysis
NO2
Fig 1.
Presumed reaction mechanisms (centres) of primary fission of the 2,6-bis(2,4,6trinitrophenylamino)-3,5-dinitropyridine (PYX) molecule in initiation processes
[1]
; here Pi- is 2,4,6-trinitrophenyl whose breakaway as a radical and subsequent
reaction with other fragments of the PYX molecule gives 2,2,4,4,6,6hexanitrodifenylamine as one from the stable intermedeates [15].
911
NO2
CH2
Channel III
H
O 2N
NH
N
NO2
NO2
N
H3C
NH
CH3
H
+
O 2N
Fig 2.
O 2N
NO2
Channel IV
Presumed reaction mechanisms (centres) of primary fission of the 2,4,6-tris(3methyl-2,4,6-trinitrophenylamino)-1,3,5-triazine (TMPM) molecule
in initiation processes [1].
Table 1. Calculated data for the pathways by channels A and B in the PYX molecule.
Channel Ee(F) E0(F)
Rate constants (Forward)
143.3
158.6
A
n
Ea/RT
7.66440E+12 0.32369 1.66860E+04
7.57497E+13 0.31999 1.85487E+04
I
II
135.2
149.8
Table 2. Calculated data for the pathways by channels C and D in the TMPM molecule
Channel Ee(F) E0(F)
Rate constants (Forward)
153.7
186.3
A
n
Ea/RT
9.98716E+11 0.56300 1.66648E+04
3.31924E+13 0.35605 2.17648E+04
III
IV
139.0
177.1
Notes to tables 1 and 2:

Ee(F): Classical Forward Barrier
E0(F): Zero-point Energy Corrected Forward Barrier
Rate constants: k(T)=A*(T**n)*exp [-Ea/RT]
912
3.
DISCUSSION
Molecules of both the compounds, i. e. PYX and TMPM, contain substituents with
hydrogen atom attached at the -position to the nitro group: the migration of this hydrogen atom
to oxygen atom of ortho-nitro group is regarded to be the primary step of decomposition of the
corresponding molecules (i. e. trinitrotoluene mechanism [13,14]). Possible channels of all these
migrations in the PYX and TMPM molecules are depicted by Figs. 1 and 2.
Fig. 3 shows that in the case of PYX initiation by impact both the reaction centers in its
molecule (channels I and II) should be powered. Relatively little difference of results of the
quantum chemical calculation (see Table 1) for channels I and II bears evidence of this
possibility. For similar evaluation of the other kinds of PYX initiation (i. e. by heat, shock and
electric spark) the application of 15N NMR chemical shifts of nitrogen atoms of the primarily
reacting nitro groups will be desirable (all in the sense of Refs. 1-3). However, 2,2,4,4,6,6hexanitrodifenylamine was found in the PYX samples after their exposition by heat, on the one
hand, and by shock wave, on the other [15]; the said finding indicates participation of the splitting
also in the sense of channel II in the initiations mentioned.
Fig 3.
A relationship between impact sensitivity, Edr (detected by sound), and 13C NMR
chemical shifts, C, of carbon atoms carrying the primarily reacting nitro group in
molecules of polynitro arenes (taken from Ref. 1 and reconditioned by means of
the C and Edr values from this Ref.; numbers in brackets are positions in which
the most reactive nitro groups are bonded in the given molecules).
913
0.2
0.19
Square of charge (q N ) 2 /e2
0.18
y = 0.0006x - 0.1278
2
R = 0.911
HNS(sol)
NONA(sol)
0.16
DPM
0.1
440
DPA(sol)
DPE
TMPM (2)
E
y = 0.0005x - 0.1082
2
R = 0.9297
y = 0.0008x - 0.265
2
R = 0.9555
460
TNX
DIPS
BITNT
480
TNB
HNO
HNB
DATB
TPT
HNS
NONA
TATB
DIPSO
TNT
TPT
TENN
DATB(sol)
DPA
TATB(sol)
TNCr
0.13
0.11
TENN(sol)
TMPM(6)
0.14
0.12
PA
CTB
PAM(sol)
HNO(sol)
0.17
0.15
PAM
TPT(sol)
y = 0.0002x + 0.0799
2
R = 0.9543
DCTB
TPM
D
y = 0.0008x - 0.2843
2
R = 0.9942
TNMs
500
520
540
560
580
600
620
640
Onset of thermal decomposition T D /K
Fig 4.
Relationship between Mulliken charges, qN, on nitrogen atoms of primarily

reacting nitro group and onsets, TD, from differential thermal analysis of polynitro
arenes (taken from Ref. 10); the sol in brackets means that
the given compounds has been investigated in the solution of 1,3,5trinitrobenzene.
Fig. 4 is taken from Ref. 10. It represents relationship between electronic charges, qN, on the
nitrogen atoms of nitro groups that are primarily split off and onsets of thermal decomposition,
TD, from differential thermal analysis of polynitro arenes. The existence of the relationship
might be connected with the electrostatic interaction of instantaneous point dipoles at the
reaction centre of their molecules [10]. In the Figure the qN value of TMPM for nitro group in the
position 2 of its benzene nucleus practically correlates with line A (the experimentally
determined TD value is 502.3 K [10], the calculated one in the sense of line A is 505.3 K). The
line corresponds to substances whose thermal decomposition begins by interaction of oxygen
atom of the ortho-nitro group with the above-mentioned hydrogen atom in -position, or with
sulfur bridge heteroatom or with chlorine. It means that a pathway of the primary fission by the
channel III (classical trinitrotoluene mechanism of decomposition) of TMPM appears to be
confirmed. On the other hand, the qN value of TMPM for nitro group in the position 6 of its
benzene nucleus does not correlate neither with line E nor line B. Namely line E conjugates
data of derivatives with amino groupings in molecule (the calculated TD value in this sense is
529.6 K). Therefore, a pathway of the primary fission by the channel IV (i. e. primary
interaction of oxygen atom of ortho-nitro group with the hydrogen of amino group) of TMPM
appears to be improbable. The last statement is in a good agreement with the results of
calculation in the Table 2
914
4.
CONCLUSION
Each from the 2,6-bis(2,4,6-trinitrophenylamino)-3,5-dinitropyridine (PYX) and

2,4,6-tris(3-methyl-2,4,6-trinitrophenylamino)-1,3,5-triazine (TMPM) molecules contains two
potential centres of the primary fission in its initiation [1] which should be realized by
trinitrotoluene mechanism.
Both the reaction centers in the PYX probably participate in the first fission of its molecule
by initiation stimuli although the primary participation of picryl nitro group in the position 2
(together with amino bridge) in this process seems to be more probable.
Primary participation of nitro group in the position 2 together with the methyl group in the
position 3 on initiation processes in TMPM molecule should be unique reaction centre here.
Acknowledgement
The work was created as a part of the internal project of University of Pardubice
No. MV340005/2004 (PRMV).
915
REFERENCES:
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
ZEMAN S.: A Study of Chemical Micro-Mechanisms of Initiation of Organic Polynitro Compounds,

in P. POLITZER & J. MURRAY (Eds.), Energetic Materials, Part II, Elsevier B. V., Amsterdam 2003,
p. 25.
ZEMAN S.: The Relationship between Differential Thermal Analysis Data and the Detonation
Characteristics of Polynitroaromatic Compounds, Thermochim. Acta 41 (1980) 199.
ZEMAN S.: The relationship between the Arrhenius Parameters of the Low-Temperature hermolysis
and the 13C and 15N NMR Chemical Shifts of the Nitramines. Thermochim. Acta 202 (1992) 191.
ZEMAN S.: Analysis and Predictions of the Arrhenius Parameters of Low-Temperature Thermolysis
of Nitramines by Means of the 15N NMR Spectroscopy, Thermochim. Acta 333 (1999) 121.
ZEMAN S.: New Aspects of the Impact Reactivity of Nitramines, Propellants, Explos., Pyrotech. 25
(2000) 66.
ZEMAN S.: Relationship between Detonation Characteristics and 15N NMR Chemical Shifts of
Nitramines, J. Energet. Mater. 17 (1999) 305.
ZEMAN S., ZEMAN V., KAMENSKY Z.: Relationship between the Electric Spark Sensitivity and the
NMR Chemical Shifts of some Organic Polynitro Compounds, Proc. 28th Int. Annual Conf. ICT,
Karlsruhe 199, p. 66/1.
ZEMAN S., FRIEDL Z.: Relationship between Electronic Charges at Nitrogen Atoms of Nitro Groups
and Thermal Reactivity of Nitramines, J. Thermal Anal. Calorim. 77 (2004) 217.
ZEMAN S., HUCZALA R., FRIEDL Z.: The Study of Chemical Micromechanism Governing
Detonation Initiation of some m-Dinitrobenzopolyazaarenes, J. Energ. Mater. 20 (2002) 53.
ZEMAN S., FRIEDL Z.: Relationship between Electronic Charges at Nitrogen Atoms of Nitro Groups
and Onsets of Thermal Decomposition of Polynitro Arenes, Cent. Eur. J. Energ. Mater. 1 (2004) 3.
FRISCH M. J., TRUCKS G. W., SCHLEGEL H. B., et. al.: Gaussian 98, Revision A.7. Gaussian,
Inc., Pittsburgh PA, 1998.
EYRING H.: The activated complex in chemical reactions, J. Chem. Phys. 3 (1935) 107.
MATVEEV V. G, DUBIKHIN V. V., NAZIN G. M.: Soglasovanyi mekhanism razlozheniya
aromaticheskikh nitrosoedinii v gazovoy faze (Thermolysis Mechanism of the Aromatic
Nitrocompounds in the Gas Phase), Izv. Akad. Nauk SSSR, Ser. Khim. (1978) 474.
BULUSU S., WEINSTEIN D. I., AUTERA J. R., ANDERSON D. A., VELICKY R. W.: Deuterium Kinetic
Isotope Effect: An Experimental Probe for the Molecular Processes Governing the Initiation of
RDX, HMX and TNT, Proc. 8th Int. Symp. on Detonation, Albuquerque, NM, July, 1985.
VARGA R., ZEMAN S.: unpublished results, Univ. Pardubice, Jan. 2005.
916
COMPARISON OF CONTINOUS AND DISCONTINOUS METHOD

OF DETONATION VELOCITY MEASUREMENT IN MINING DRILLS
(VOD)
Stjepan ganec*, Zvonimir Ester** and Mario Dobrilovi**
* Minervo Ltd, Ljubljana, Slovenia
** Faculty of Mining, Geology and Petroleum Engineering,
Abstract:
Detemination of caracteristics of explosives detonation velocities in mining drills by
continuous and discontinous method performed in rnotii quary. Measurements are
performed in same mine field therefore under geologically exactly same conditions. Data
retrieved show difference between mentioned measurement methods and works and
measurement principles themselves. Detonation velocitiy measurement following
discontinuous method was performed on semi-highway track Bar Podgorica on section
Tanki rt urmani. Measurements were conducted in different length of measurement
intervals for mining drill charges, with results confirming conclusions obtained by
continuous method measurements.
Keywords:
1.
measurement, VOD, explosives
INTRODUCTION
Explosives detonation velocity is on of the important characteristics of explosives, based on

which selection of different explosives is made for mining in different rock formations
(dependant on their geological and geomechanical characteristics type, layers, tectonics, and
other). Method of measuring detonation velocity in mining drills gives clear picture, due to its
real-time conditions. Those measurements reveal differnces between manufacture-declared and
real-time detonation velocities achieved on mining sites. For explosive made on-site this method
is actually the only one, allowing corrections with ANFO, heavy ANFO and EMULSION
explosives. Sampling of explosives, shipping to laboratories and on-site testing is complex
process, with demanding time constraint.
This prevents corrections based on lab and test results, simply by works either ending or
even finished on location. For VOD measurements in rnotii quary explosives used were
ANFO explosive, heavy ANFO and powder-charges amonal explosive. Initiaton was through
milisecond electric detonators, and firing of ANFO and heavy ANFO explosives through
powder-charges explosive.
Measurements by discontinuous method in drills on semi-highway Bar Podgorica,

section Tanki rt urmani were of powder-charges explosive AMG-2, manufactured by
Booster Niki. Initiation of explosive charge was through milisecond electric detonators.
2.
GEOLOGICAL CHARACTERISTICS OF THE RNOTIE
The basis of the soil is the foraminiferic limestone of the upper and lower Palaeocene age.
The limestone is grey and of middle to thick stratification with distinctive surface karstic and
subterranean karstic phenomena. Pleistocene clay is found in holes and cracks in the funnelshaped limestone and the karstic. The general dip direction of layers is north-east at the dip
angle of 20 to 40. There are two systems of cracks. The first one consists of vertical cracks in
the direction of the layers. The other one consists of the narrower cracks perpendicular to the
layers.
3.
GEOLOGICAL CHARACTERISTICS OF SEMI-HIGHWAY TRACK
Semi-highway Bar-Podgorica, section Tanki rt urmani geological data are based on

detail geo-engineer maping of broader surrounding of the track, different rock formations and
structured test drills. Majority of the track is formed of carbon rock. Carbon complex is build in
layers and cracks organogene karstic. Carbon complex are intensly cracks with relatively
numerous meter-long and even more numerous decimeter cracks in block forms.
Surface zone sees cracks extensive in greater cavities. This sediments are noted in sprouts
and structure drills. Thickness varies from 5 to 12 meters. Ground formation of this strucures is
comprised of flysch sediments. Flysch sediments are also of layery structure and texture. This
formations show Carbon original occurances karstic sediments. General fall of layers is E-W
with 30 60 degrees orientation.
4.
MEASUREMENTS OF THE VELOCITY OF DETONATION (VOD)

IN BLASTHOLES
4.1
MREL instrument characteristics for measurement of explosives

in drills using continuous method
Measure the continuous VOD in any hole diameter, wet or dry holes, and in any type
of rock.
Measure the continuous VOD in multiple holes per blast.
Determine whether full detonation, low order detonation or failure occurred, and
where in the explosive column it happened.
Check VOD against manufacturer's specifications in full scale blasting
environments.
Determine the minimum booster size for any explosive by measuring run-up
velocities in full scale blasting environments.
Measure the timing accuracy of detonators in full scale blasting environments.

Measure the effects of water, drill cutings, and rocks, etc. trapped within the
explosive mass.
Determine the length of explosive column to use in decking operations to evaluate
the effect of stemming and drill cutting dilution, water pick-up, etc. on the explosive
run-up requirements.
Determine the correct length and type of stemming material to be used between
decks of explosives to prevent sympathetic detonation of explosive desensitization
from occuring.
4.2
Eksplomet fo 2000 instrument characteristics for measurement

of explosives in drills using discontinuous method
Method is based on measurements with optic conductor:

Time interval of measurement: 0,1 s to 10 s
Detonation velocity: up to 10.000 m/s
Work temperature: 0C do 50C
Measurement tolerance per time: up to 0,1 s
Measurement results tolerance: 3%
Measurement interval: max distance of sondae up to 10 m
Sondae (connectors) on ends of optical duplex conductor HFRB 35xx36xx are used
Instrument aplication for measurements in laboratories, on-site and in drills.
Detonation velocitiy of different explosives can be measured, as well as those of
explosive,incal explosive pirotechnical components and non-electric detonators - NONEL
tubes.
4.3
Results of measurements
4.3.1. Continuous method measurements of ANFO and heavy ANFO explosives

Measurements in rnotie quary
Fig 1.
VOD ANFO explosives
Fig 2.
VOD heavy ANFO, 20 %emulsion
4.3.2. Discontinuous method measurements of ANFO and heavy ANFO explosive

Measurements in rnotie quary
Detonation velocity measurement for ANFO and heavy ANFO explosive
of drill = 82 mm ,
h length = 14 m ,
a drill cork length = 3 m
b explosive charge length = 11 m,
L measurement interval = 5-7 m
t time measured,
VOD detonation velocity measured
1 starting sondae first optical conductor
2 ending sondae second optical
conductor
a
2
h
L
b
1
Fig 3.
Testing of explosives in blastholes with the Explomet-fo-2000 instrument

VOD ANFO and heavy ANFO explosives
Measurement result (VOD)

L=5m
t = 1349,5 s (1349,510-6 s)
VOD = 3705 m/s
Detonation velocity measurement for heavy ANFO explosive:
L=7m
t = 1758,79 s (1758,7910-6 s)
VOD = 3980 m/s
4.3.3. Discontinuous method measurement of powder-charges explosive AMG-2

Semi-highway Bar-Podgorica track measurements
of drill = 82 mm,
h length = 6-9 m,
a drill cork length =2,5- 3 m
b explosive charge length = 1,95-6,5 m,
L measurement interval = 1,75-6,1 m
t time measured,
VOD detonation velocity measured
1 starting sondae first optical conductor
2 ending sondae second optical
conductor
a
2
h
L
b
Fig 4.
Discontinuous method measurement of powder-charges explosive AMG-2
Measurement interval results

L = 1,75 m
t = 562,8 s (562,810-6 s)
VOD = 3109 m/s
L = 2,15 m
t = 584,6 s (584,610-6 s)
VOD = 3677 m/s
L = 6,10 m
t = 1485,9 s (1485,910-6 s)
VOD = 4105 m/s
5.
MEASUREMENT ANALISYS
Continuous method measuements allows controll of detonation velocity in full length of

explosive charge. From diagramme (fig. 1) detonation velocity of initiating capsule of powder
explosive reads VOD = 4321,5 m/s as starting impulse for activaion of ANFO explosive.
Diagramme shows that after starting impulse velocity decreases during activation of slower
ANFO explosive. In following 3m velocity gradually increases ( VOD = 3061,6 m/s, VOD =
3500,1 m/s, VOD = 3792,5 m/s). Detonation velocity stabilizes at 3855,0 m/s to 3804,3 m/s
following that.
Discontinuous method poduces average detonation velocity of 3705 m/s. Diagramme (fig.
2) shows detonation velocity increase VOD 4380,9 m/s, measured with continuous method in
upper part of the drill charged with 6 kg of powder explosive. Detonation velocity in, heavy
ANFO charged part of the drill measured at 3798,0 m/s to 4035,2 m/s. Discontinuous method
produced average value of detonation velocity of 3980 m/s.
Discontinuous method measurement of powder explosive AMG 2 on semi-highway track
depict increase in detonation velocity relative to increase of measurement interval. Due to
starting lesser detonation velocity and gradual increase through the drill, small measurement
interval (1,75 and 2,15 m) can not produce realistic values. Interval increase to 6,1 m narrows
the gap of real average detonation velocity declared by manufacturer.
6.
CONCLUSION
Detonation velocity measurement by both continuous and discontinuous method gives us

caracteristics parameters of explosives used, whee each method shows strond points and
disadvantages.
Continuous method allows endless intervals inside the drill. This is especially
recommandable where multi type explosives were used inside one drill, allowing monitoring in
full interval of separate explosive. This method also allows for anomailies in drill or explosive
charge to be determined.
Discontinuous method of measurement is swifter, simpler and more efficient. Method
assertains average detonation velocity in interval measured. Measurement results clearly state
that shorter measurement interval varies from real values, or that maximum detonation velocity
doesnt occure in short drills.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
DESIGN: Minervo Ljubljana: Rudarski projekt za izvajanje del pri izkorianju mineralnih surovin
in sanaciji kamnoloma , 95, 2002
DESIGN: Podjetje za vrtanje in miniranje, Ljubljana: Elaborat vrtanje in miniranja,25, 1995
DESIGN: Minervo Ljubljana, Priprava razstrelilnih meanic na mestu uporabe z mobilno enoto
"Minervo EC004 3 T", 74, 2000
DESING: MREL group of Companies, VOD Operations Manual, 72, Canada, 2000
ESTER Z., VRKLJAN D.,DEKOVI Z: Usporedba uinka miniranja emulzijskih i ANFO,
5/2000eksploziva u labaritoriju i "in situ", 20-22, Mineral Zaagreb
KRSNIK J.: Miniranje, Rudarsko-geoloko naftni fakultet Zagreb, 187, 1989
GANEC S., DOBRILOVI M., ESTER Z.: Drilling and blasting works on a highway construction on
the edge of a rock-wall, 2nd World Conference on Explosives & Blasting Technique, Prague, 503508, 2003
GANEC S., DOBRILOVI M., ESTER Z.: In situ measurment of velocity of detonation and
comparison of heavy ANFO produced in the field of usage and cartridge explosives ,7th Int..
Seminar NTREM, Pardubice , Czech Republic , part 2, 730-736, 2004
NON STANDARD METHODES

IN TESTING FIRING CURRENT OF ELECTRIC DETONATORS
Igor Zori, and Dalibor Kuhinek
University of Zagreb, Faculty of Mining, Geology and Petroleum Engineering,
Pierottijeva 6, 10000 Zagreb, CROATIA
Abstract:
For testing electrical detonators some countries have national standards, members of
the European Union have drafts or accepted standards that describe equipment minimal
electrical and mechanical properties, preparation of samples to be tested, handling and
testing procedures and finally data processing and presentation. In this paper we will
consider one alternative method for testing firing current of electrical detonators and
compare it to standard method of testing.
Keywords:
1.
electric detonator, determination of firing current
INTRODUCTION
1.1
Testing electrical detonators
Testing of characteristics of electrical detonators is done by manufacturers but also by

customers (importers). Testing of imported detonators is being done regularly according to
national standards and recommendations with the purpose of confirming manufacturer's
declarations about the characteristics of detonators. Testing is done on samples, like on all
mass-produced products, that are determined according to strictly defined procedures and
quantities and should include at least 0,5 % of the series[1] of products. One part of the
testing is non-destructive and the rest is destructive. The only completely non-destructive
testing is the testing of electrical resistance, in which testing current does not exceed 10 % of
firing current necessary for detonator activation (with probability of firing 0,01%)[2] in which
process the thermal effect on detonators can be neglected (one hundredth of power or
energy). In potentially destructive testing, even when detonators are not activated,
characteristics of detonators are changed, so they can not be used anymore in complex
mining networks.
1.2
Determination of firing current
One of the testing for the imported detonators done in Croatia in a standard manner is
determination of firing current for activation of single detonator. Testing is done according
to HRN H.D8.103 and every detonator is tested only once with the current of constant value.
Required sample for testing is 150 detonators and testing is done for intervals of frequency
of firing from 0 to 10 %, 10 to 90 % and 90 to 100 % with 50 detonators in each and every
of these three intervals. In the process of testing it is necessary to select 5 to 7 different
values of current, 3 to 5 different values of which should be chosen for the middle interval of
10 to 90 %. The main problem is determining current value that should be used. Previous
experience and data can be helpful. With appropriately chosen values of current about 50 %
925
of the detonators remain not activated but also unusable for complex mining networks. The
typical result of that type of testing is shown in Fig. 1.
frekvencija paljenja (%)
99,99
99,95
99,8
99,5
99
98
97
95
90
80
70
60
50
40
30
20
10
5
3
2
1
0,5
0,1
0,05
0,01
0,19
0,21
I1d
0,23
0,25
0,27
I1m
0,29
0,31
0,33
0,35
I1g
0,37
I (A)
Fig 1.
Interependence of currents and frequencies of firing
From this line currents are extrapolated:
2.
I1d current by which detonators never fire
I1m mean value of current that has frequency of firing of 50 %
I1g current by which detonators fire with probability of 100 %
s standard deviation of measurement results.
THEORY
In the process of testing the detonators primer mixture is heated up to the temperature
that is not sufficient for its activation. The consequence is chemical change, which results in
decreased sensitivity of the primer mixture, i.e. in increased temperature needed to initiate
the primary explosive. In order to avoid the irreversible degradation of the characteristics of
non-activated detonators the following possibility is taken into consideration: the use of
current changed according to well-known principles to the point of activation of the
detonator and the value and the standard deviation of the firing current can be calculated. If
such a testing could be done even with a smaller number of tested detonators it would still be
more suitable then the testing which is already used, regardless of the more complex system
for measurement.
926
The use of the current ramp instead of constant current is considered as the simplest
possibility, as shown in Fig. 2 b.
a)
Fig 2.
2.1.
b)
Current in dependence on time,

a) current used in standard testing
b) current ramp
Construction of the current source
For the purpose of testing a current source was constructed to generate current, the value
of which was increasing steadily from zero to maximum in predetermined time. The speed of
rise was variable within the range 0,03 to 0,6 A/ms with other characteristics [3] as follows:
Maximum outgoing current of 5 A
Maximum working voltage of 60 V
Slew rate of 10V/s
Frequency of 1MHz
The used current source consists of five units that are shown schematically in Fig. 3.
Parts of the source are:
1. Microprocessor[4] which increases state of one 8-bit register from 0 to 255
2. D/A converter[5] which converts state of register into analogue voltage change from
0 V to -10 V
3. That voltage is inverted by the operation amplifier with the increase of voltage from
0 V to +10 V
4. Output operation amplifier which amplifies the signal 3 6 times and can give
current of up to 5 A
5. Resistance with the value sufficient enough compared to the resistance of the
detonator, which decreases the influence of the change of the detonator's resistance
upon the characteristics of the change of current through the detonator.
Applied voltage on the detonator and therefore the current, depend on the ratio of resistance
RP and resistance of the detonator Rd. Power supply of the operation amplifier is up to 60V,
but the maximum voltage on the detonator is limited with:
Ud = U
Rd
Rd + RP
(1)
927
Microprocessor
Inv.
D/A Converter
OpAmp
RP
Rd
Fig 3.
3.
Main parts of the current source
EXPERIMENTS
The measurements are done according to the scheme shown in Fig. 4.
Fig 4.
The scheme of measurement
The change of current through detonator is monitored indirectly by recording the voltage
drop on the known resistance serially connected to the tested detonator Rd. Typical
oscillogram of the voltage change on the operation amplifier output and the current through
the detonator are shown in Fig. 5. The upper diagram shows the change of the voltage of the
current source, and the lower diagram the change of current through the detonator from the
beginning of the current flow to the moment of activation of the detonator (sudden decrease
of the current to 0 A).
Fig 5.
Oscillogram of the voltage on the output of the current source and the
current through the detonator
928
The constructed current source has successfully activated detonators in all conditions of
conducted measurements. Diagrams of all measurements are recorded and the obtained
results are graphically shown in Fig. 6, 7 and 8.
4.
RESULTS
The diagram in Fig. 6 shows the measured dependence of the activation time of the
detonator on the slope of the current through the detonator. The diagram in Fig. 7 shows the
dependence of the activation current of the detonator on the slope of current through the
detonator. The diagram in Fig. 8 shows the connection of the current by the activation of the
detonator to the time needed to activate the detonator.
30
25
20
t (m s) 15
10
5
0
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
K (A/m s)
Fig 6.
Dependence of activation time on slope of current through detonators
K (A/ms)
2,5000
2,0000
1,5000
I (A)
1,0000
0,5000
0,0000
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
K (A/m s)
Fig 7.
Dependence of activation current on slope of current through detonators
929
2,5000
2,0000
1,5000
I (A)
1,0000
0,5000
0,0000
0
10
15
20
25
t (ms)
Fig 8.
5.
Dependence of activation time and value of activation current of detonators
DISCUSSION
As shown in Fig.6 and 7, and as it was expected, current at which activation of detonator
occurred and the time necessary for detonator to be activated depends on the slope of current
through detonator. With the increase of current slope the time for detonator activation is
shortened, but the current value for activation is increased. Measured current value at which
detonators are activated are greater than those declared by the manufacturer. For easier
reading of measured results greater slope of current is favourable, but with smaller slope of
current results are closer to values of firing current as declared by manufacturer.
Manufacturer specifications of firing current for single detonator is 0.35 A and according to
our measurement with slope of 0.03 A/ms detonator was activated with current of 0.65 A.
The results of measurements are grouped for certain current slope, the smaller deviation of
grouping, the higher slope of current is used.
6.
CONCLUSION
Due to the variable current value through detonator it is not simple to compare activation
current of the detonator with declared firing current of single detonator. The angle of current
slope has additional influence to measurement results. To get more comparable results to
those obtained by standard methods, more measurements should be done. Grouping of the
results for certain values of current slope sets the possibility of using this very method for
standard deviation determination of characteristics of tested detonators.
Acknowledgment:
This paper has been realized by the support of the Ministry of Science Education and
Sport, The Republic of Croatia; under the project "Optimizing parameters of testing the
electrical detonator characteristics" (0195027).
930
REFERENCES
[1]
[2]
[3]
[4]
[5]
HRN H.D8.100 Electric detonator, Characteristics and technical requirements, 1987.

HRN H.D8.103 Determination of firing current of electric detonators, 1973.
BURR BROWN CORPORATION: Datasheet for high voltage, high current operational amplifier
OPA 548T, 1997
MICROCHIP TECHNOLOGY INC.: Datasheet for PIC16F84A, 1998.
ANALOG DEVICES: Datasheet for DAC08C
931
COMPARISON OF VOD MEASUREMENT RESULTS WITH

CONTINUOUS AND DISCONTINUOUS METHOD
Zvonimir Ester, Mario Dobrilovi, Darko Vrkljan
University Of Zagreb, Faculty of Mining, Geology and Petroleum Engineering,

Pierottijeva 6, 10 000 Zagreb, Croatia
Abstract:
An explosives velocity of detonation (VOD), can be used to indicate a number of
important characteristics regarding the products performance under specific field and
test conditions. A number of charasteristic and transient VOD curves have been
identified, which can be used to evaluate explosive performance, control ground
vibration amplitudes, select the correct amount and type of stemming, eliminate
explosive desensitization and evaluate primer performance. Although the primary focus
of this paper is comparison of VOD measurement results with continuous and
discontinuous method.
Keywords:
1.
explosive, VOD measurement, continius method, discontinuous method
INTRODUCTION
Detonation velocitiy is the velocity at wich the chemical reaction zone propagates
through a given explosive. It is one of the most important detonation parameters. Detonation
velocitiy of known high explosives may rich nearly 10 mm/s, the experimental
determination of the detonation velocity is not easily achived. However, when compared to
the other detonation parameters, its accomplishment represent the least complicated task.
The determination of detonation velocity is based upon the measurement of the time
interval needed for the detonation wave to travel a known distance through the explosive
being tested.
Two types of measurement method have been used: continuous and discontinuous
method.
2.
DESCRIPTION OF THE CONTINUOUS METHOD
Test were performed whit a portable, continuous VOD recorder VODMate from
Instantel.
The VODMate (fig. 1) works by providing a constant current source to drive a high
resistance length of rigid sensing rod. The sensing rod is placed in explosive cartridge, and is
then shorter and consumed by explosives detonation front. The VODMate measures the
change in resistivity of the sensing rod as it consumed to calculate velocity of detonation, at
sample rates up to 2 MHz with 14 bit resolution.
932
Fig 1.
VODMate and accesories
A example of test whit short explosive sample is presented in figure 2.

The basic characteristics of emulsion explosive ELMULEX on which is done the testing
is presented in table 1.
Table 1. Explosive Elmulex tecnical data
The density
The gas volume
The explosion energy
The explosion temperatures
Oxygen balance
Micro balls
Diameter the cartridge
Length the cartridge
Fig 2.
kg/l
l/kg
kJ/kg
C
%
%
mm
mm
1,15
925
3120
2050
+0,7
2,3
50
1100
Typical monitoring set-up of short explosive samples
933
3.
DESCRIPTION OF THE DISCONTINUOUS METHOD
Tests were performed whit a portable VOD recorder Explomet-Fo-Multimode (fig.3).

The determination of the detonation velocity is based on the ability of the optical fibre to
accept a light signal when the detonation wave arrives, and to transmit that signal to the
measuring equipment, wich enables prercise recording of the time interval between two
signal.
The Explomet-Fo-Multimode has 4 independent timers measuring the time intervals
between the illumination of 4 optical probes. One of the probes starts all the timers. Then
every triggered probes stop its timer. The results of each timer is displayed in microseconcs.
Fig 3.
Explomet-Fo-Multimode and accesories
A example of test whit short explosive sample is presented in figure 4.
Fig 4.
Typical monitoring set-up of short explosive samples
934
4.
VELOCITY OF DETONATION MEASUREMENTS

FOR SHORT EXPLOSIVE SAMPLE
The results of both measurement methods are presented in table 2.
Table 2. Measurement results
Diameter
of the
cartridge
(mm)
50
50
50
50
Velocity of detonation
Primer
Continuous method
m/s
TED
TED
TED
TED
4353
3437
3938
2787
I interval
4016
4118
3992
4008
Discontinuous method
m/s
II interval
4032
3985
4133
4028
III interval
3938
4044
4098
4103
A number of characteristic and transient VOD curves have been identified whit
instrument VODMate. On the figure 5 is presented record of test no. 1.
Fig 5.
VOD curve record of short explosive sample, test no.1.
Significant differences were found in results from different measurement methods (table
2).
5.
VELOCITY OF DETONATION MEASUREMENTS IN BLASTHOLE
Instrument VODMate is predicted for in hole detonation velocity measurement,

measurement are carry out for production blasting, in real condition, with same kind of
explosive (table 1) taken for measurement of short samples. Fig 6. show VOD shot in blast
hole 76 mm, with length of 16 m.
935
Fig 6.
VOD curve record in blasthole
Fig 7.
In hole VOD setup
936
6.
CONCLUSIONS
Results of measurement on short explosive samples show:
1) Untrustworthiness of continuous measurement method of velocity of detonation.
Measurement results for identical samples varied from 2787 do 4353 m/s
2) Results of measurement with discontinuous method are balanced in area 3938 do
4133. m/s
3) Discontinuous method is prescribed with EN EN 13631-12 Explosives for civil usesHigh explosives- Part 14: Determination of velocity of detonation.
4) Continuous method is valid for measurement of VOD in situ, in blast holes
REFERENCES
[1]
[2]
[3]
[4]
ESTER Z., VRKLJAN D.,DEKOVI Z: Usporedba uinka miniranja emulzijskih i ANFO, 5/2000
KRSNIK J.: Miniranje, Rudarsko-geoloko naftni fakultet Zagreb, 187, 1989
GANEC S., DOBRILOVI M., ESTER Z.: Drilling and blasting works on a highway construction
on the edge of a rock-wall, 2nd World Conference on Explosives & Blasting Technique, Prague,
503-508, 2003
SUESKA, M., 1995, Test Methods for Explosives. New York:Springer-Verlag,1995, 225
pp.ISBN 0-387-94555-5.
937
Title
Publisher
Chairman
Editor
In print
Pages
Issue
Edition
Publication No.
NEW TRENDS IN RESEARCH OF ENERGETIC

MATERIALS,
PROCEEDINGS of the VIII. SEMINAR, CD version
University of Pardubice
prof. Ing. Svatopluk Zeman, DrSc.
Ing. Ji Vgenknecht, Ph.D.
March 2005
939
80 CD-R
first
UP 05-34
Made in
Knechtik, DTTX, University of Pardubice
UP 05-34
ISBN 80-7194-758-X
(For printed version only - Not for CD!)

University of Pardubice - 8th International Seminar NTREM (2005)

Diunggah oleh

Informasi Dokumen

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

University of Pardubice - 8th International Seminar NTREM (2005)

Diunggah oleh

Hak Cipta:

Format Tersedia

Faculty of Chemical Technology

Department of Theory and Technology of Explosives

NEW TRENDS IN RESEARCH

Pardubice, Czech Republic

NEW TRENDS IN RESEARCH OF ENERGETIC MATERIALS

held at the University of Pardubice,

intended as a meeting of students, postgraduate students, university

March 14th, 2005

University of Pardubice, 2005

Seminar is supported by:

Chairman of the Seminar:

(DSTL, Sevenoaks, U.K.)

Members of the Committee:

(National Univ. of Singapore)

Members of the Committee:

(Explosia, Ltd., Pardubice)

CZ TEAM Ltd. UXO CLEARANCE IN CZECH REPUBLIC, PRAHA

EXPLOSIA a.s., PARDUBICE SEMTN

OZM RESEARCH, Ltd. , HROCHUV TYNEC

DISLOCATION - ASSISTED INITIATION OF ENERGETIC MATERIALS

Heming Xiao*, Zhaoxu Chen**, and Xiaojuan Xu*

A QUANTUM CHEMICAL STUDY ON THERMOLYSIS INITIATION MECHANISMS

Hans J. Pasman*, Richard Bouma**, Nikolaos Zarzalis***, and Max Weiss***

SOME NOTES ON THE FUNDAMENTALS OF E.M. INITIATION

Towards a UNIVERSAL SENSITIVITY CHARACTERISTIC?

Thomas D. (Tommy) Sewell

ATOMISTIC STUDIES OF FUNDAMENTAL PROPERTIES AND PROCESSES IN ENERGETIC

HIGH-TEMPERATURE PROPELLANT TORCH-SYSTEM FOR NON-DETONATIVE

HIGHLIGHTS OF STABILITY RESEARCH IN THE FIRST DOZEN

Janusz Wrzesiski, Jadwiga Popawska-Jach, and Andrzej Koaczkowski

LIABILITY OF AMMONIUM NITRATE EMULSIONS TO SPONTANEOUS DECOMPOSITION

COMPUTER SIMULATION OF NITRO-1,2,4-TRIAZOLES DECOMPOSITION REACTIONS

Anthony J. Bellamy*, Alessandro E. Contini*, Peter Golding**, and Stephen J. Trussell**

BOMB CALORIMETRIC STUDY

Jan Bdek, Stanisaw Cudzio, and Sylwia Pietrzyk

DETERMINATION OF POST-EXPLOSION RESIDUES

REACTIVITY BETWEEN -CL20-GAP AND -HMX-GAP INVESTIGATED BY MASS LOSS,

A.G.Anshits*, N.N.Anshits*, A.A.Deribas**, S.M.Karakhanov**, N.S.Kasatkina***, A.V.Plastinin**,

DETONATION VELOCITY OF EMULSION EXPLOSIVES WITH CENOSPHERES

SYNTHESIS AND CHARACTERISATION

G. Fischer*, G. Holl**, T. M. Klaptke*, P. Mayer*, J. J. Weigand*

PROPERTIES OF A HIGHLY FRICTION SENSITIVE DERIVATIVE OF 1,5-DIAMINO-1HTETRAZOLE (DAT): 1,5-DIAMINO-4-METHYLTETRAZOLIUM DINITRAMIDE

Alexander Gromov*, Ulrich Frter-Barth**, and Ulrich Teipel**

STUDY OF NON-ISOTHERMAL NITRIDATION OF ALUMINUM NANOPOWDERS PASSIVATED

OPTIMIZATION OF 1,1-DIAMINO-2,2-DINITROETHENE SYNTHESIS

Mikhail A. Ilyushin, and Igor V. Tselinskii

THE INFLUENCE OF THE STRUCTURE OF THE SALTS OF AZOLES

CALCULATIONS OF DENSITIES AND HEATS OF FORMATIONS OF ENERGETIC MOLECULES

CHARACTERIZATION OF DETERRED PROPELLANTS BY CLOSED VESSEL TESTS:

A MOLeCULAR DYNAMICS SIMULATION STUDY

CONCEPT OF THE SPATIAL-PERIODIC MICRO-STRUCTURES EXCITATION AT THE

Sreekumar Pisharath and Ang How Ghee

DECOMPOSITION KINETICS OF GAP BINDER

Miroslav Pospil* and Pavel Vvra**

DECOMPOSITION OF ENERGETIC MATERIALS

G. Santhosh, S. Venkatachalam, K. Krishnan, K.N. Ninan

THE PHASE STABILIZATION OF AMMONIUM NITRATE

Jonas arlauskas*, Aura Nemeikait-nien**, ilvinas Anuseviius*, Henrikas Nivinskas*,

MAMMALIAN CELL CYTOTOXICITY OF NITROAROMATIC EXPLOSIVES

APPLICATION OF THE METHOD OF GRAY INTERRELATION ANALYZING

Ulrich Teipel, and Ulrich Frter-Barth

DESIGN OF PARTICULATE ENERGETIC MATERIALS

Titica Vasile, Cristian Barbu, and Doru Safta

Hans J. Pasman, Richard Bouma, Nikolaos Zarzalis, and Max Weiss***

Anthony J. Bellamy, Alessandro E. Contini, Peter Golding, and Stephen J. Trussell

A.G.Anshits, N.N.Anshits, A.A.Deribas, S.M.Karakhanov, N.S.Kasatkina*, A.V.Plastinin,

G. Fischer*, G. Holl**, T. M. Klaptke, P. Mayer, J. J. Weigand*

Alexander Gromov*, Ulrich Frter-Barth, and Ulrich Teipel

Jonas arlauskas*, Aura Nemeikait-nien**, ilvinas Anuseviius, Henrikas Nivinskas,

L.A.Merzhievsky, P.I.Zubkov, K.A.Ten, E.R.Pruuel, L.A.Lukyanchikov, V.M.Titov,

Anthony J. Bellamy, Alistair MacCuish and Peter Golding**

Michael A. Bohn, Anton Hammerl, Kate Harris, and Thomas M. Klaptke**

Daniel Buczkowski, Adam Presz, and Bogdan Zygmunt

L. Huskov, J. rmkov, K. Ventura*, Z. Aktein, and M. tancl

T. M. Klaptke, J. J. Weigand, and G. Holl**