PEROVSKITE SYNTHESIS AND ANALYSIS USING

STRUCTURE PREDICTION DIAGNOSTIC SOFTWARE

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of the Ohio State University

By

Michael W. Lufaso, B.S.

*****

The Ohio State University

2002
Dissertation Committee:
Approved by
Professor Patrick Woodward, Adviser

Professor Anne McCoy ____________________________

Adviser
Professor James Cowan Chemistry Graduate Program
 ABSTRACT

The perovskite structure type accommodates most of the metallic ions in the

periodic table and a significant number of different anions. This compositional diversity

leads to a variety of physical properties. A software program, SPuDS (Structure

Prediction Diagnostic Software) was developed to predict the crystal structures of

perovskites, including those distorted by tilting of symmetric and Jahn-Teller distorted

octahedra. Only the composition and oxidation states of the ions were required as user

inputs. Rigid octahedra were maintained in electronically symmetric octahedra, while

distortions of the octahedral bond lengths were utilized in Jahn-Teller distorted

compositions. SPuDS calculates the optimal structure in Glazer tilt systems most often

observed experimentally and for tilt systems with interesting multiple A-cation

environments. Structure optimization occurs by distorting the structure to minimize the

global instability index. Location of the A-site cation is chosen so as to maximize the

symmetry of its coordination environment.

SPuDS has been employed in a number of useful applications, including the use

of predicted structural information to estimate physical properties of both hypothetical

compositions and those materials for which accurate structural data is not available, as a

guide for exploratory synthetic efforts, as a starting model for Rietveld refinements in the

course of structurally characterizing new materials, and for extracting the effects of

ii
octahedral tilting distortions from other structural distortion mechanisms. Structural

predictions were made for a variety of compositions with a single and multiple octahedral

cations and compared with a number of previously determined structures to illustrate the

accuracy of this approach.

Synthesis and structural refinement from x-ray and neutron powder diffraction of

Ca2MnMO6 (M = Nb, Sb, Ru) and Sr2MnMO6 (M=Nb, Sb) is reported. Various types of

cooperative Jahn-Teller distortions (CJTD) that occur in ternary (AMX3) and double

(A2MM′O6) perovskites are reviewed. Interplay between CJTD’s, octahedral tilting and

cation order is systematically examined in A2MM′O6 perovskites, where M is an active

Jahn-Teller (J-T) ion (Mn3+ or Cu2+) and M′ is an octahedrally symmetric cation.

Perovskites with a+a+a+ octahedral tilting and a single octahedral cation exhibit

interesting dielectric properties. SPuDS was also used to examine the prospects for

synthesizing new compounds in tilt systems with multiple A-cation coordination

geometries. Perovskites with a specific type of octahedral tilting (a+a+a+) coupled with A-

cation and rock-salt octahedral cation ordering were predicted to exist. A collaborative

effort to perform high-pressure high-temperature synthesis resulted in the successful

creation of four new dual A-site and octahedral cation ordered perovskites.

Characterization of the structures and dielectric properties of CaCu3Ga2Sb2O12,

CaCu3Cr2Sb2O12, CaCu3Ga2Nb2O12 and CaCu3Ga2Ta2O12 is reported.

iii
Dedicated to all who assisted my education prior to and during my

study in the Department of Chemistry at the Ohio State University.

iv
 ACKNOWLEDGMENTS

I wish to show appreciation to those who have influenced me both scientifically and
personally during my graduate studies. Recognition of all who influenced me is not
possible in the space available. However, I would like to specifically acknowledge the
following people.

I appreciate the guidance Dr. Patrick Woodward provided me throughout my study of

chemistry. His expertise positively influenced my education and his impact will last long
after I leave his group. Pat was always available to answer a question, yet allowed me to
independently discover new directions in my project. I also thank Pat for providing me
the opportunity to travel and present my research at scientific meetings.

I show gratitude to group members Paris Barnes and Hank Eng for their assistance,
support, friendship and insightful discussions both inside and outside of the laboratory.

I wish to express thanks to group members that influenced my understanding of science

and chemistry including Dr. Elena Labrador, Dr. Santhosh Nagappan Nair, Dr. Jae-Hyun
Park, Dr. Nattamai Bhuvanesh, Dr. Hiroshi Mizoguchi, Young-Il Kim, Patrick
Underwood, Josh Goldberger and Meghan Knapp. The knowledge of each of these
individuals added to my understanding of chemistry.

I am grateful to Dr. Anne McCoy for introducing me to FORTRAN 77 programming

during my stay in her group. I was an undergraduate in the summer of 1997 and was able
to perform research as a part of the National Science Foundation Research Experiences
for Undergraduates program.

I appreciate Dr. I. David Brown for providing the Accumulated Table of Bond Valence
Parameters version 2.2 file and for his helpful discussions of the bond valence method.

I wish to show appreciation to Dr. John Parise and Dr. Song-Ho Byeon for assistance
with high-pressure high-temperature synthesis of a novel tilt system of ordered
perovskites.

I thank Dr. Brett Hunter for collection of neutron powder diffraction at the Australian
Nuclear Science and Technology Organisation.

v
 VITA

November 9, 1975……………………....Born – Newark, Ohio

1998……………………………………..B.S. Chemistry, Youngstown State University.

1998 – present……………………..……Graduate Fellowship, Teaching and Research

Associate. The Ohio State University

Research Publication

1. M.W. Lufaso & P.M. Woodward, "The Prediction of the Crystal Structures of
Perovskites Using the Software Program SPuDS." Acta Cryst., B57, 725, (2001).

2. A.B. McCoy, M.W. Lufaso, M. Veneziani, S. Atrill & R. Naaman, "Reactions of

oxygen atoms with van der Waals complexes: The effect of complex formation on the
internal energy distribution in the products." J. Chem. Phys. 108, 9651, (1998).

3. M.W. Lufaso & A.B. McCoy, "Effects of complex formation on reactions of

oxygen with HCl and Ar-HCl." Chem. Phys., 187, 239, (1998).

4. T.D. McCarley, M.W. Lufaso, L.S. Curtin & R. McCarley, "Multiply Charged
Redox-Active Oligomers in the Gas Phase: Electrospray Ionization Mass Spectrometry of
Metallocenes." J. Phys. Chem. B., 102, 10078, (1998).

5. C.C. Smith, J.M. Jacyno, K.K. Zeiter, P.D. Parkanzky, C.E. Paxson, P.
Pekelnicky, J. S. Harwood, A.D. Hunter, V.G. Lucarelli, M.W. Lufaso & H. G. Cutler,
"Nitration of Cyclopentenecarboxaldehyde: Studies Toward 1-Amino-2-
Nitrocyclopentanecarboxylic Acid." Tetrahedron Letters, 39, 6617, (1998).

vi
Manuscripts in Preparation for Publication

6. M.W. Lufaso & P.M. Woodward, "The Prediction and Evaluation of the Crystal
Structures of Rock-Salt Ordered Perovskites Using the Software Program SPuDS." In
preparation.

7. M.W. Lufaso & P.M. Woodward, “Jahn-Teller Distortions, Cation Ordering and
Octahedral Tilting in Perovskites.” In preparation.

8. M.W. Lufaso, P.M. Woodward, J. E. Goldberger & B. Hunter, “Crystal Structures of

Sr2MnNbO6, Sr2MnSbO6, Ca2MnNbO6, Ca2MnRuO6 and Ca2MnSbO6.” In
preparation.

9. S.-H. Byeon, J. B. Parise, P. M. Woodward, M. W. Lufaso & A. Celestian, ”High-

Pressure High-Temperature Synthesis and Dielectric Properties of A-site and
Octahedral Cation Ordered Perovskites.” In preparation.

FIELDS OF STUDY

Major Field: Chemistry

vii
 TABLE OF CONTENTS

Page

Abstract .............................................................................................................................. .ii

Dedication ......................................................................................................................... .iv

Acknowledgments................................................................................................................v

Vita..................................................................................................................................... vi

List of Tables ..................................................................................................................... xi

List of Figures…… .......................................................................................................... xiii

Chapters:

1. Introduction..............................................................................................................1

1.1 Perovskite History............................................................................................1

1.2 Physical Properties...........................................................................................2
1.3 Synthetic Methods ...........................................................................................3
1.4 Structure Connectivity .....................................................................................5
1.5 Distortions from Cubic Symmetry...................................................................6
1.6 Tolerance Factor ............................................................................................16
1.7 Experimental Determination of Distortions...................................................17
1.8 References......................................................................................................19

2. Structure Prediction Diagnostic Software (SPuDS) ..............................................22

2.1 Stimulus for Software Development..............................................................22

2.2 Variable Reduction ........................................................................................23
2.3 Equation development for lattice parameters and atomic positions ..............25
2.4 Calculation Methods ......................................................................................27
2.5 Optimization of A-site cation position...........................................................33
2.6 Tolerance Factor Based on Bond Valence.....................................................36

viii
 2.7 Temperature Dependant Bond Valence .........................................................39
2.8 Software Development...................................................................................39
2.9 Software Operation ........................................................................................41
2.10 References .....................................................................................................42

3. Structure Prediction of Single Octahedral Cation Perovskites ..............................44

3.1 Introduction....................................................................................................44
3.2 a0a0a0 (Pm 3 m) ...............................................................................................44
3.3 a0a0c- (I4/mcm) ...............................................................................................47
3.4 a0a0c+ (P4/mbm) .............................................................................................47
3.5 a0b-b- (Imma)..................................................................................................48
3.6 a-b+a- (Pnma) ..................................................................................................48
3.7 a+a+a+ (Im 3 ) ..................................................................................................57
3.8 a-a-a- (R 3 c).....................................................................................................62
3.9 a+a+c- (P42/nmc) .............................................................................................63
3.10 a0b+b+ (I4/mmm).............................................................................................64
3.11 a0b-c+ (Cmcm) ................................................................................................67
3.12 Conclusions....................................................................................................68
3.13 References......................................................................................................69

4. Structure Prediction of Ordered and Disordered Multiple Octahedral cation

Perovskites .........................................................................................................................73

4.1 Introduction to Multiple Octahedral Cation Perovskites ................................73

4.2 a0a0a0 (Fm 3 m) ...............................................................................................78
4.3 a0a0c- (I4/m).....................................................................................................83
4.4 a-b+a- (P21/n) ...................................................................................................91
4.5 a0b-b- (I2/m) .....................................................................................................96
4.6 a-a-a- (R 3 ) .......................................................................................................96
4.7 a+a+a+ (Pn 3 )....................................................................................................98
4.7.1 CASTEP Geometry Optimization of Pn 3 ..........................................108
4.8 Conclusions...................................................................................................109
4.9 References.....................................................................................................110

5. Synthesis and Crystal Structures of Sr2MnNbO6, Sr2MnSbO6, Ca2MnNbO6,

Ca2MnRuO6 and Ca2MnSbO6 ..........................................................................................112

5.1 Introduction..................................................................................................112
5.2 Experimental ................................................................................................114
5.3 Sr2MnMO6 (M=Nb, Sb)...............................................................................115
5.4 Ca2MnMO6 (M=Nb, Sb, Ru) .......................................................................124

ix
 5.5 Conclusions..................................................................................................130
5.6 References....................................................................................................132

6. Jahn-Teller Distortions, Cation Ordering and Octahedral Tilting in Perovskites134

6.1 Introduction...................................................................................................134
6.2 Cooperative Jahn-Teller Distortions .............................................................136
6.3 Octahedral Tilting and Cation Ordering .......................................................140
6.4 Influence of octahedral tilting in AMX3 systems .........................................142
6.5 Cation Ordering, Octahedral Tilting and Jahn-Teller distortions in A2MM′X6
systems...........................................................................................................156
6.6 Orbital Ordering Schemes.............................................................................161
6.7 Modeling Using SPuDS................................................................................165
6.8 Discussion .....................................................................................................167
6.9 Conclusions...................................................................................................171
6.10 References....................................................................................................172

7. Structure Prediction, Synthesis and Dielectric Properties of Novel Dual A-Site and
Octahedral Cation Ordered Perovskites...........................................................................175

7.1 Introduction...................................................................................................175
7.2 Known Dielectric Perovskite Analysis .........................................................182
7.3 Impedance Analysis Sample Preparation......................................................186
7.4 Impedance Analysis Measurement Techniques............................................188
7.5 Instrument Calibration ..................................................................................192
7.6 Prediction of Dual Ordered perovskites........................................................193
7.7 Synthesis and Structure Refinement of Dual A, Octahedral Cation Ordered
Perovskites ....................................................................................................195
7.8 Impedance Analysis of CaCu3M2M2’O12......................................................205
7.9 Conclusions...................................................................................................210
7.10 References....................................................................................................212

Bibliography ....................................................................................................................215

Appendix A......................................................................................................................251

x
 LIST OF TABLES

Table Page

1.1 Fifteen tilt systems information for single octahedral cation perovskites .............13

2.1 Lattice parameter equations based on M-X bond distance (d) and tilt angle.........26

3.1 Experimental and SPuDS predicted structural data for untilted perovskites.........46

3.2 Additional compounds in the a-b+a- (Pnma) tilt system. ........................................51

3.3 Experimental and predicted structural information for Pnma perovskites ............56

3.4 Bond valence sums and structural data for known a+a+a+ perovskites. .................60

3.5 SPuDS modeling of CaFeTi2O6. ............................................................................64

3.6 SPuDS results for hypothetical compounds designed for a0b+b+...........................67

3.7 SPuDS results for hypothetical compounds designed for a0b-c+ ...........................68

4.1 Information for the fifteen tilt systems with rock-salt ordered octahedral cations .75

4.2 Comparison of two calculation methods for ordered a0a0a0 perovskites. ...............80

4.3 Tolerance factors for I4/m perovskites. ..................................................................83

4.4 Experimental and SPuDS optimized crystal structures for I4/m perovskites .........85

4.5 Perovskites evaluated in space group P21/n (tilt system a-a-b+). ............................92

4.6 Experimental and SPuDS optimized structures for P21/n perovskites. ..................93

4.7 Experimental and SPuDS optimized structures for I2/m perovskites.....................97

4.8 Experimental and SPuDS optimized structures for R 3 perovskites.......................98

xi
4.9 Six coordinate d(M3+-O) and d(M5+-O) for hypothetical Pn 3 structure calcs. ....101

4.10 GII and octahedral cation size difference for hypothetical Pn 3 compositions. ...105

4.11 Structure prediction calculation comparison of Pn 3 by SPuDS and CASTEP. ..108

5.1 Tolerance factors for the five synthesized A2MnMO6 compounds. ....................116

5.2 SPuDS lat. param. & fractional coordinates for Sr2MnMO6 (M=Sb, Nb)...........118

5.3 Rietveld refinement details & results for Sr2MnMO6 (M = Sb, Nb) part 1.........122

5.4 Rietveld refinement details & results for Sr2MnMO6 (M = Sb, Nb) part 2.........123

5.5 Selected bond distances and valence sums for Sr2MnSbO6 and Sr2MnNbO6........ 124

5.6 SPuDS lat. param. & fractional coordinates for Ca2MnMO6 (M = Nb, Ru, Sb). 125

5.7 Rietveld refinement details & results for Ca2MnMO6 (M = Nb, Ru, Sb) part 1 .127

5.8 Rietveld refinement details & results for Ca2MnMO6 (M = Nb, Ru, Sb) part 2 .128

5.9 Bond distances and bond valence sums for Ca2MnMO6 (M = Nb, Ru, Sb). .......129

6.1 Bond valence sums of ions in KCuF3, Ba2CuWO6, La2CuSnO6, NdSrMn2O6 ....140

6.2 Structure information for LaMnO3, KCuF3, Sr2CuWO6, and La2CuTiO6 ...........142

6.3 Bonds distances and valences (Nd-O) of NdFeO3 and NdMnO3. .......................155

6.4 Structural data of Sr2MM′O6 (M=Mn, Fe; M′=Nb, Ta, Ru, Sb)..........................159

6.5 Structural data of Ca2MM′O6 (M=Mn, Fe; M′=Nb, Ta, Ru, Sb). ........................160

7.1 Measured capacitance and loss tangent of sample holder ...................................190

7.2 Dielectric analysis and porosity corrections ........................................................193

7.3 Rietveld refinement results of x-ray data for ordered Pn 3 perovskites. .............203

7.4 Rietveld refinement results of x-ray data for disordered Im 3 perovskite ...........204

7.5 Dielectric properties of CaCu3M2M′2O12 perovskites..........................................209

xii
 LIST OF FIGURES

Figure Page

1.1 A view looking down the c axis of a0a0c- (top) and a0a0c+ ....................................10

1.2 Distribution of tilt systems with a single octahedral cation...................................15

2.1 SPuDS operational flowchart.................................................................................28

2.2 GII versus octahedral tilt angle for SrTiO3 and SrZrO3 .........................................32

2.3 Valence map contour plot of A-site cation for CaTiO3 .........................................35

2.4 Bond valence and ionic radii calculated tolerance factor ......................................38

3.1 Octahedral tilt angle versus tolerance factor for a-b+a- perovskites .......................50

3.2 A-site cation displacement in x and z versus tolerance factor for a-b+a-. ..............53

3.3 % lattice parameter error versus the bond calculated tolerance factor for a-b+a- ...54

3.4 Crystal structure and cation coordinations of CaCu3Ti4O12 ..................................58

3.5 SPuDS modeling for known compounds of tilt system a+a+a+. .............................61

3.6 SPuDS predicted GII values for PdCdCa2Ti4O12, Ca2Cd2Ti4O12 and CaFeTi2O6. .66

4.1 Rock-salt, layered, and chain octahedral cation ordering. .....................................76

4.2 Distribution of rock salt ordered tilt systems.........................................................78

4.3 Lattice error for fixed M-X and GII optimized structures in Fm 3 m.....................82

4.4 Octahedra tilting in I4/m Sr2WMgO6 .....................................................................84

4.5 Percent lattice error for I4/m perovskites...............................................................87

xiii
4.6 Lattice parameter ratio versus tolerance factor for I4/m perovskites.....................88

4.7 GII versus tolerance factor for I4/m perovskites....................................................90

4.8 Experimental and SPuDS A-site fractional positions for P21/n perovskites .........94

4.9 Experimental and SPuDS octahedral tilt angles for P21/n perovskites..................95

4.10 Crystal structure of dual A-site and octahedral cation ordered a+a+a+ (Pn 3 )........99

4.11 Bond valence sums and GII versus tolerance factor in a+a+a+ (Pn 3 )..................103

4.12 GII for multiple tilt systems vs. tolerance factor of four perovskites in a+a+a+ ...107

5.1 Rietveld refinements of four different structural models of Sr2MnSbO6 ............119

5.2 Two structural models for Rietveld refinement of NPD data for Sr2MnSbO6 ....121

5.3 Rietveld refinement fit of Ca2MnNbO6 ...............................................................126

6.1 Structure diagrams for KCuF3, Ba2CuWO6, La2CuSnO6 and NdSrMn2O6..........138

6.2 Cu-F distances vs. tolerance factor for ACuF3 (A = Na, K, Rb) series ...............143

6.3 View along the b-axis (Pnma) of LaMnO3 showing ac plane orbital ordering ...145

6.4 M-O distances vs. tolerance factor for AMnO3 and AFeO3 series ......................148

6.5 O-M-O bond angles vs. tolerance factor for AMO3 (M = Mn, Fe) series ...........149

6.6 Lattice parameters of AMnO3 and AFeO3 versus tolerance factor......................150

6.7 Lattice distortion for AMnO3, AFeO3, Ca2MnMO6 (M = Sb, Ta, Nb, Ru) .........151

6.8 A-site fractional disp. vs tol. for AMnO3, AFeO3 and A2MnGaO6 (A=La, Nd) .153

6.9 A-O bond distances for AMnO3 and AFeO3 series .............................................154

6.10 Orbital ordering schemes for Jahn-Teller ions.....................................................164

6.11 GII versus τ for AMnO3 with different orientations of long Mn-O bonds ..........166

xiv
6.12 Bond distortion parameter versus τ for A2MnMO6 (A = Sr, Ca; M = Nb, Ru, Sb,
Ta) & A2MnGaO6 (A = La, Nd)...........................................................................169

7.1 Simulated and experimental diffraction patterns of CaCu3Cr2Sb2O12 .................196

7.2 Refinement fit of CaCu3Ga2Ta2O12 in a+a+a+ (Pn 3 ) ...........................................200

7.3 Refinement fit of CaCu3Ga2Sb2O12 in a+a+a+ (Pn 3 ) ...........................................201

7.4 Refinement fit of CaCu3Ga2Nb2O12 in a+a+a+ (Im 3 ) ...........................................202

7.5 Impedance analysis Z” vs. Z’ for a+a+a+ perovskites...........................................206

7.6 Impedance analysis C vs. frequency and D vs. frequency for a+a+a+ ..................207

7.7 Capacitance versus temperature for a+a+a+ perovskites .......................................208

xv
 CHAPTER 1

INTRODUCTION TO PEROVSKITES

1.1 Perovskite History

The mineral perovskite was discovered and named by Gustav Rose in 1839 from

samples obtained in the Ural Mountains. Perovskite is named after a Russian

mineralogist, Count Lev Aleksevich von Perovski who was appointed Russian secretary

of the interior in 1842. The term perovskite was originally reserved for the mineral

CaTiO3. The first synthetic perovskites were produced by Goldschmidt (1926) of the

University of Oslo led to the use of the term perovskite as a description of a class of

compounds sharing the same general stoichiometry and connectivity found in CaTiO3.

The perovskite structure has the general AMX3 stoichiometry where A and M are cations

and X is an anion. Substitution of different ions on each site is readily accepted making

the perovskite structure type one of the most frequently encountered in solid-state

inorganic chemistry. Perovskites have held the interest of crystallographers for a

significant period of time due in part to the wide range of compositions and resulting

variety of physical properties.

The perovskite structure accommodates most of the metallic ions in the periodic

table and a significant number of different anions. Perovskites have been reported with

all naturally occurring cations in the periodic table except boron and beryllium.
1
Phosphorus and the noble gases are also not observed in the perovskite structure. The

majority of the perovskite compounds are oxides or fluorides, but the perovskite structure

is also known for the heavier halides (Hönle et al., 1988; Luaña et al., 1997), sulfides

(Clearfield, 1963), hydrides (Gingl et al., 1999), cyanides (Peschel, et al., 2000; Malecki

& Ratuszna, 1999), oxyfluorides (Carlson et al., 2000) and oxynitrides (Marchand et al.,

1991).

1.2 Physical Properties

The wide range of compositions leads to diverse physical properties in perovskite

materials. Geologists have long been interested in the perovskite structure, which is

proposed to be the most abundant mineral on earth in a high pressure form of

Mg(1-x-y)CaxFeySiO3 (Hemley & Cohen, 1992). Catalytic properties are present in a

number of perovskites including the potential automotive catalytic converter material

La(Fe0.57Co0.38Pd0.05)O3 (Nishihata et al., 2002). Solid solutions of CaTiO3-NdAlO3 are

employed as microwave resonators for third generation mobile phone base stations

(Hughes et al., 2001). Perovskites formulated from mixtures of ~89% Ba(Zn1/3Ta2/3)O3,

7.5% Ba(Ni1/3Ta2/3)O3 and 3.5% BaZrO3 are microwave dielectric materials (Davies et

al., 1997). Electrical conductivity ranges from highly insulating BaZrO3 to metallic

SrRuO3 (Bouchard & Gillson, 1972). High dielectric constant BaTiO3 is ferroelectric and

is used in capacitors (Shirane et al., 1957). Piezoelectric materials convert mechanical

(i.e. strain) energy to electrical energy and vice-versa. Perovskites with piezoelectric

properties, e.g. Pb(Zr1-xTix)O3 and Pb2ScTaO6, play a dominant role in the

electroceramics industry including applications such as ceramic phonographic cartridges,

2
sensitive microphones, powerful sonar, piezo-ignition systems, pressure gauges and

transducers, ultrasonic imaging, and relays. Such materials also serve as critical

components in a number of smart devices, which are able to sense the surrounding

environment and respond to it (Newnham, 1997; Trolier-McKinstry & Newnham, 1993;

Newnham & Ruschau, 1991). The highest superconducting transition temperature for a

non-cuprate oxide material is observed in BaPb1-xBixO3 (x ≈ 0.05 – 0.3) (Sleight et al.,

1975). High temperature cuprate superconductors adopt perovskite related crystal

structures (Cava et al., 1987; Capponi et al., 1987). Members of the manganate based

perovskite system, (Ln1-xAx)MnO3 (Ln = lanthanide ion, A = alkaline earth ion), have

been studied extensively over the past decade for their fascinating magnetic and colossal

magnetoresistive (CMR) properties (Gong et al., 1995).

1.3 Synthetic Methods

Perovskites have been synthesized by a variety of methods including standard

solid state methods, precursor routes, citrate precursors, and reactive fluxes. Standard

solid state methods are often applied using oxides, carbonates, nitrates, or acetates of the

metal cations of interest as starting materials. The powders are accurately weighed out in

the stoichiometric ratio, then mixed together and ground with a mortar and pestle.

Mixtures are calcined at a temperature high enough to decompose the starting materials

(nitrates, carbonates, and acetates). The mixtures are reground and heated at an elevated

temperature. Powder x-ray diffraction techniques are used to ascertain the purity of the

compounds. Heating cycles are repeated until the synthesis is complete, i.e. when the

peaks in the diffraction pattern no longer exhibit a change in intensities upon further

3
heating cycles. Rate of heating is important in some perovskites as the cooling rates can

have a large influence on the ordered domains of the octahedral cations. Ordered domain

sizes changed from nanometer sized to >100 nm in size by a slow cooling rate in

Ba(Mg1/3Nb2/3)O3-BaZrO3 (Akbas & Davies, 1998).

Precursor routes are utilized to minimize the loss of volatile cation species

including lead and antimony. Synthesis of (Pb1-xCax)[Fe1/2(Nb1-yTay)1/2]O3 (Park et al.,

2001) and Pb(Sc1/2Ta1/2)O3 (Dmowski et al., 2000) perovskites is reported via the

columbite route. The columbite route involves heating the non-volatile cations at a

reaction temperature at which PbO would exhibit volatility. Formation of an intermediate

composition of the non-volatile cations occurs. PbO is mixed with the intermediate

composition and reacted at a lower temperature to minimize the volatility of the PbO.

The desired final composition is formed with a minimized Pb content loss that occurs for

direct high temperature solid state synthesis.

The citrate process synthesis of perovskites occurs by calcination of amorphous

citrate complex precursors. Proper molar ratios of nitrates of the cations of the desired

perovskite composition are dissolved in water. An aqueous solution of citric acid is added

in a 1:1 molar ratio of the total metal cations. Water is removed by vacuum techniques to

yield a solid amorphous citrate precursor. Heating results in the decomposition of the

amorphous mixture and formation of the crystalline perovskite phase. A homogenous

mixture of the cations is obtained through this method and results in a lower temperature

for formation of the perovskite phase. Perovskites with large surface area, which is

4
important for catalytic properties, can often be obtained by reducing the synthesis

temperatures (Zhang et al., 1987)

Reactive fluxes are useful in obtaining a wide variety of perovskite phases.

Fluxes typically consist of a relatively low melting hydroxide or chloride typically

containing one or more of the cations in the desired composition. Cations that are not

expected to be incorporated into the structure can also be present in the flux. Hydroxide

fluxes tend to stabilize high-oxidation states by lowering the synthesis temperature. As

an example, heptavalent osmium was observed in reactive hydroxide flux synthesis of

Ba2MOsO6 (M=Li, Na) (Stitzer et al., 2002). The use of a proper flux can greatly speed

ion diffusion thereby leading to a more complete reaction at a lower temperature.

1.4 Structure Connectivity

The ideal perovskite structure has AMX3 stoichiometry and is composed of a

three dimensional network of regular corner-linked MX6 octahedra. The perovskite

topology has an efficient packing scheme. Historically the octahedral site (M) is

designated the B-site. The terms M-site and octahedral cation are used interchangeably in

this text. In the ideal case, the M-site cations are at the center of the octahedra with the A

cations centrally located in the body center of the cube formed by eight corner-linked

octahedra.

The symmetry of the ideal perovskite structure is cubic in space group Pm 3 m.

Axes formed by M-X bonds of the octahedra coincide with the crystallographic cubic

axes. The ideal structure has no variable parameters in the structure besides the lattice

parameter. Several viewpoints are used to describe the perovskite structure, with the

5
corner-sharing octahedra description most commonly used. The corner-sharing

description is the most intuitive depiction when considering distortions from ideal

symmetry.

1.5 Distortions from Cubic Symmetry

Interestingly the mineral perovskite, CaTiO3, does not adopt the aristotype cubic

structure. The symmetry of CaTiO3 is lowered from cubic (Pm 3 m, Z = 1) to

orthorhombic (Pnma, Z = 4) by a cooperative tilting of the titanium centered octahedra

(Sasaki et al., 1987). This distortion is driven by the mismatch between the size of the

cubo-octahedral cavity in the corner-sharing octahedral network and the undersized ionic

radius of the Ca2+ ion. The octahedral tilting distortion lowers the coordination number

of Ca2+ from 12 to 8 in order to reduce the tension in the remaining Ca-O bonds (Brown,

1992) and increase the lattice energy. However, there is very little perturbation of the

local octahedral coordination of the Ti4+ ion.

The parent structure type is referred to as the aristotype, and distorted perovskites

are designated hettotypes. Three different types of distortions identified include:

distortions of MX6 octahedral units, M-cation displacements within the octahedra, and

octahedral tilting distortions (Megaw, 1973).

Distortions of MX6 octahedral units often result from electronic factors. In

octahedral coordination, the 3d transition metal cations Mn3+ and Cu2+ (high-spin) have

electron configurations (t2g)3(eg)1 and (t2g)6(eg)3, respectively. Electronically this is an

unfavorable situation and the Jahn-Teller (J-T) theorem states a distortion of the

geometry occurs (Kanamori, 1960). Lowering the symmetry of the octahedron removes

6
the electronic degeneracy. A distortion in M-O bond lengths results in an energetic

stabilization. J-T distortions typically occur without a large deviation of the O-M-O bond

angles from ideal values of 90° and 180°.

Displacements of octahedral cations are often observed due to the combination of

structural influences and electronic factors. Octahedral cation displacement distortions

are typically observed in compositions with a tolerance factor greater than unity.

Oversized A-cations stretch the octahedral bonds, resulting in an increase in octahedral

volume and reduction of M-X bonding. Displacement of the octahedral cation towards

anions occurs to increase the bonding of the octahedral cation and is typically only seen

for d0 cations that undergo a second order Jahn-Teller distortion. Stretching of the

octahedral cation is observed in BaTiO3, resulting in ferroelectric properties (Sleight et

al., 1975).

Octahedral tilting is the most common type of distortion the perovskite structure

type undergoes (Woodward, 1997a). Distortion from ideal cubic symmetry of the

aristotype cubic perovskite (Pm 3 m) occurs by a practically rigid tilting of the octahedral

units while maintaining both the regularity of the octahedra and the corner-sharing

connectivity. Distortions in the octahedral angles are minor (typically less than 4° for X-

M-X). Octahedral tilting allows greater flexibility in the coordination of the A-site cation

while maintaining a regular coordination environment for the octahedral cation.

Octahedral tilting reduces the symmetry of the A-site cation coordination environment

and leads to a significant change in the A-X bond lengths. The octahedra can tilt in

7
multiple ways, each leading to a different coordination environment for the A-site

cation(s).

Classification of octahedral tilting in perovskites was investigated by considering

the possible tilting patterns coupled with the development of a standard notation to

describe octahedral tilting distortions (Glazer, 1972). Glazer describes the cases where

only two successive layers are considered. An alternative, but equally valid notation was

developed by Aleksandrov (1976). For the sake of clarity the Glazer notation will be

used. The notation describes a tilt system by rotations of MX6 octahedra about the three

orthogonal Cartesian axes, which are coincident with the three axes ([100], [010], and

[001]) of the aristotype perovskite with a cubic unit cell. Glazer notation uses symbols of

the type a*b*c*. The letters in Glazer notation indicate the relative magnitude of the

rotation about a given axis, e.g., use of the letters a, b, and c imply unequal tilts about the

x, y, and z axes. The letters do not indicate the magnitude of tilting relative to one another

in increasing fashion (i.e., octahedral tilting designated by b is not necessarily larger than

a). Matching letters indicate equal tilts about the different pseudo-cubic axes. The

superscript * is used to denote the phase of the octahedral tilting in neighboring layers. A

positive superscript would denote the neighboring octahedra tilt in the same direction (in

phase) and a negative superscript implies the tilts of neighboring octahedra tilt in the

opposite direction (out of phase). A superscript of 0 signifies no tilting about that axis.

Figure 1 illustrates the structures which correspond to tilt systems a0a0c+ and a0a0c-. The

octahedral rotations in tilt systems a0a0c+ and a0a0c- occur only about the z-axis of the

cubic perovskite. Rotation of one octahedron causes the four adjacent octahedra in the

8
same layer to rotate in the opposite direction in the same angle. From this figure one can

see that rotation of a single octahedron defines the rotation of all octahedra in the same

layer. However, lattice connectivity is such that rotations of the octahedra in the layer

above and below are not geometrically constrained to the initial rotation and can occur

either in phase (+ superscript) or out of phase (- superscript) with respect to the first

octahedral layer.

9
Figure 1.1 : A view looking down the c axis of a0a0c- (top) and a0a0c+ (bottom) with the
A-site cations shown as spheres and the M-site cations located at the center of the
octahedra.

10
 Tilting of the octahedra reduces the symmetry of the undistorted perovskite tilt

system a0a0a0. Glazer (1972) derived 23 different tilt systems, leading to 15 different

space groups obtained by inspection. Minor corrections to the space groups were

published in updated descriptions (Glazer, 1975 and Glazer & Burns, 1990). Howard and

Stokes (1998) have performed a group-theoretical analysis of simple tilt systems that can

be described in terms of basic tilts around the pseudo-cubic axes. Considering in phase or

out of phase tilting and the magnitude of the tilt angle, there are six basic component tilt

systems (a+b0c0, a-b0c0, a0b+c0, a0b-c0, a0b0c+, a0b0c-). The tilt systems for perovskites are a

linear combination of the six component tilts. Considering a repeat pattern of no more

than two neighboring octahedra decreased the number of possible tilt systems. The

different possible patterns were described by tilting vectors in a representation space. For

each tilting pattern (or vector) the necessary space group is the isotropy subgroup.

Operations that leave the vector invariant are contained in the isotropy subgroup. Eight of

Glazer’s tilt systems were found to be redundant due to the fact they impose a higher

symmetry than is required by the space group symmetry. Using this approach, there are

15 tilt systems that can occur in real crystals, each with a different space group. The 15

tilt systems with the space group, degrees of freedom, number of Wyckoff sites for each

ion and the estimated number of experimentally reported structures (obtained via a

comprehensive literature search) are shown for a single octahedral cation in table 1.1, and

those with a 1:1 rock-salt ordered cation in table 4.1. Aleksandrov and Bartoleme (2001)

recently published a very comprehensive review of octahedral tilting distortions in

11
perovskites as well as perovskite-related structures. The architecture and properties of

perovskite-like crystals, including cation- and anion-deficient phases, was examined by

Aleksandrov & Beznosikov (1999). Recently Bock & Müller (2002) reported the group-

subgroup relationships for derivatives of the perovskite structure type.

12
 Glazer Tilt Space Degrees of Wyckoff Sites
Frequency
System* Group Freedom A M X
Group A – High Symmetry Tilt Systems
0 0 0
a a a (23) Pm 3 m 1 1 1 1 21
- - -
a a a (14) R3c 3 1 1 1 24
a0a0c- (22) I4/mcm 3 1 1 2 9
a0a0c+ (21) P4/mbm 3 1 1 2 5
a0b-b- (20) Imma 6 1 1 2 6
a-b+a- (10) Pnma 10 1 1 2 119
Group B – Multiple A-Site Tilt Systems
+ + +
a a a (3) Im 3 3 2 1 1 22
a0b-c+ (17) Cmcm 10 2 1 3 6
a0b+b+ (16) I4/mmm 5 3 1 2 0
a+a+c- (5) P42/nmc 8 3 1 3 1
Group C – Transitional/Low Symmetry Tilt Systems
- - -
a a c (13) C2/c 9 1 1 2 0
a0b-c- (19) C2/m 10 1 1 3 2
a-b-c- (12) P1 18 1 2 3 1
a+b-c- (8) P21/m 18 2 2 4 3
a+b+c+ (1) Immm 9 4 1 3 0

Table 1.1 : The fifteen tilt systems, space groups, degrees of freedom, number of
independent Wyckoff sites and number of observed structures reported for single
octahedral cation perovskites with the restriction that not more than two layers show
independent tilting. *The number in parentheses corresponds to the numbering of the tilt
systems originally adopted by Glazer (1972).

13
 The distribution of tilt systems found in the literature is shown in figure 1.2 with a

total of 188 perovskites included. The structures were mainly obtained from the earlier

work of Woodward (1997b), but additions have been made and are listed in table 3.2. The

distribution shown in figure 1.2 is naturally biased by trends in scientific research (e.g.,

superconductivity, magnetoresistance), as well as the fact that the vast majority of

structure determinations are carried out at room temperature. However, figure 1.2

accurately depicts the reported distribution of octahedral tilting distortions in perovskites

at room temperature (excluding perovskites which contain multiple cations on the

octahedral site). The perovskite structure class can be divided into three groups as listed

in table 1.1. Group A are high symmetry tilt systems where all A-cation sites are

crystallographically equivalent, group B are tilt systems with multiple crystallographic

sites for the A cations and group C are low symmetry/transitional tilt systems that are

often observed as intermediates in a phase transition between two of the higher symmetry

structures. The octahedral tilting in groups A & B can be described using the notation of

Zhao (1993) by a single tilt (e.g., tilting in a0a0c- corresponds to a single tilt φ about the

cubic [001] direction, a0b-b- to a single tilt θ about the cubic [110] and a-a-a- to a tilt Φ

about the cubic [111]) or two tilts (e.g., a+a+c-) at most. The majority of the perovskite

structures belong to either group A or B, while structures that fall into group C are very

uncommon.

It is fitting that the mineral perovskite CaTiO3 adopts a distorted structure, since

distorted perovskites far outnumber undistorted cubic perovskites (Figure 1.2). In fact

the prevalence of the perovskite structure type can be directly attributed to the inherent

14
ability of the corner-sharing octahedral framework to undergo cooperative octahedral

tilting distortions in response to the size mismatch between the A and octahedral cations.

Figure 1.2 : Distribution of tilt systems among known perovskites with a single
octahedral cation.

15
 The presence and magnitude of an octahedral tilting distortion affects not only the

crystal structure, but also has a profound influence on a number of physical properties,

such as electrical conductivity, magnetic superexchange interactions, and certain

dielectric properties. For example, Ln0.7A0.3MnO3 perovskites undergo a transition from

a paramagnetic insulating state to a ferromagnetic metallic state upon cooling. This

coupled electronic/magnetic transition is of great interest due to the fact that the

magnetoresistance reaches a maximum value as the temperature approaches this

transition. Furthermore, it is known that the transition temperature can be tuned from

~350 K to below 100 K by changing the magnitude of the octahedral tilting (Hwang et

al., 1995). This remarkable sensitivity to a relatively subtle structural distortion originates

from the decrease in orbital overlap that occurs as the octahedral tilting distortion

increases (Töpfer & Goodenough, 1997). Another example of coupling between the

octahedral tilting distortion and a physical property of technological significance occurs

in perovskites used for microwave dielectric applications. Colla et al. (1993) have shown

that the sign and magnitude of the temperature coefficient of the dielectric constant is

quite sensitive to changes in the octahedral tilting distortion.

1.6 Tolerance factor

The Goldschmidt tolerance factor (Goldschmidt, 1926) is a measure of the fit of

the A-site cation to the cubic corner-sharing octahedral network. In a cubic perovskite

twice the M-X bond length is the cell edge, and the twice the A-X bond length is equal to

the face diagonal. The tolerance factor is shown in the following equation.

16
 RA + RX
τ=
2 ( RM + R X )

where the variable τ is the tolerance factor, RA, RM and RX are the ionic radii of the A-

cation, M-cation and X-anion, respectively. Ionic radii have been tabulated for ions by

Shannon (1976). Cation-anion bond distances obtained from bond valence calculation

were used in this study. The calculation method for ideal bond lengths and a comparison

to values obtained using tabulated ionic radii are shown in section 2.6.

This tolerance factor geometrical relationship is unity for a perovskite structure with

an A-site cation if the lattice is treated as an array of close-packed spheres. The tolerance

factor value is an approximate guide to the structural stability of the perovskite phase. An

upper limit of the tolerance factor is approximately 1.04, whereas a tolerance factor less

than 0.87 is near the range where the ilmenite (FeTiO3) structure type becomes more

stable compared to the perovskite structure type.

1.7 Experimental Determination of Distortions

Space group symmetry changes if octahedral tilting occurs as shown in table 1.1.

Lattice parameters (unit cell lengths and angles) are one indication of the potential tilt

system and narrow the possible space group choices. Octahedral tilting results in the

anion moving off of a high symmetry site to a lower symmetry site. Considering the

doubling of a unit cell axis, additional peaks appear at half-integral reciprocal-lattice

planes in the diffraction pattern. Rules for determining the phase of octahedral tilting in

perovskites from the systematic absences were reported by Glazer (1975).

17
 The subtle distortion results in weak peaks that are easy to miss in indexing the

peaks of the diffraction pattern. Diffraction peaks due to low electron density anions are

weak compared to those of the electron-rich cations and easily missed if instrumental

signal to noise is poor or the background is high. Peak resolution is important in

distinguishing between untilted cubic and structures with a small magnitude of octahedral

tilting. Peak splitting due to octahedral tilting that is not observed on a low resolution

diffractometer may be observed on a higher resolution instrument, resulting in a different

space group determination and structure solution. Cubic untilted perovskite structures

refined from low-resolution data may have a true symmetry which is lower when

examined on a higher resolution instrument. The literature contains many initial structure

solutions with a unit cell dimension consistent with a cubic system, which are later

experimentally determined as tetragonal, and finally orthorhombic as data resolution and

refinement software is improved.

Accurate data collection is essential in solving the correct crystal structures of

perovskites correctly. Neutron powder diffraction is useful to determine the structure but

is not available as a laboratory technique. Neutron scattering length for oxygen nuclei is

nearly the same magnitude as the cations in perovskites, in contrast to x-ray diffraction

where the ionic scattering factor for oxygen is considerably smaller than most cations.

Other techniques including transmission electron microscopy (TEM) use electrons

to scatter from the O superlattice with sufficient intensity, but the intensity of each

reflection cannot be quantified. Structural refinement from TEM is generally not

possible, but the method can be used as an aid for space group determination.

18
1.8 References

Akbas, M.A. & Davies, P.K. (1998). J. Am. Ceram. Soc. 81(3), 670-676.

Aleksandrov, K.S. (1976). Kristallografiya 21, 249-255.

Aleksandrov, K.S. & Beznosikov, B.V. (1999). Ferroelectrics 226, 1-9.

Aleksandrov, K.S. & Bartolome, J. (2001). Phase Trans. 74(3), 255-335.

Bock, O. & Müller, U. (2002). Acta Cryst. B58, 594-606.

Brown, I.D. (1992). Acta Cryst. B48, 553-572.

Burns, G. & Glazer, A. M. (1990). Space Groups for Solid State Scientists , 2nd ed.

Capponi, J.J., Chaillout, C., Hewat, A.W., Lejay, P., Marezio, M., Nguyen, N., Raveau,
B., Soubeyroux, J.L., Tholence, J.L. & Tournier, R. (1987). Europhysics Letters 3, 1301-
1307.

Carlson, S., Larsson, A.-K. & Rohrer, F.E. (2000). Acta Cryst. B56, 189-196.

Cava, R.J., van Dover, R.B., Batlogg, B. & Rietman, E.A. (1987). Phys. Rev. Lett. 58(4),
408-410.

Clearfield, A. (1963). Acta Cryst. 16, 134-142.

Colla, E.L., Reaney, I.M. & Setter, N. (1993). J. Appl. Phys. 74, 3414-3425.

Davies, P.K., Tong, J. & Negas, J. (1997). J Am. Ceram. Soc. 80(7), 1727-1740.

Gingl, F., Vogt, T., Akiba, E. & Yvon, K. (1999). J. Alloys and Compounds 282, 125-
129.

Glazer, A.M. (1972). Acta Cryst., B28, 3384-3392.

Glazer, A.M. (1975). Acta Cryst., A31, 756-762.

Goldschmidt, V.M. (1926). Naturwissenschaften 14, 477-485.

Gong, G.-Q., Canedy, C., Xiao, G., Sun, J., Gupta, A. & Gallagher, W.J. (1995). Appl.
Phys. Lett. 67, 1783-1785.

19
Hemley, R.J., Jackson, M.D. & Gordon, R.G. (1987). Phys. Chem. Minerals 14, 2-12.

Honle, W., Miller G., & Simon, A. (1988). J. Solid State Chem. 75, 147-155.

Howard, C. J., & Stokes, H. T. (1998). Acta Cryst. B54, 782-789.

Hughes H. Iddles, D.M. & Reaney, I.M. (2001). Appl. Phys. Letters 79, 2952-2954.

Hwang, H.Y., Palstra, T.T.M., Cheong, S.-W. & Batlogg, B. (1995). Phys. Rev. B.
52(21), 15046-15049.

Kanamori, J. (1960). J. Appl. Phys. 31(5), 14S-23S.

Luana, V., Costales, A. & Martin Pendas, A. (1997). Phys. Rev. B. 55(7), 4285-4297.

Malecki, G. & Ratuszna, A. (1999). Powder Diff. 14, 25-30.

Marchand, R., Laurent, Y., Guyader, J., L’Haridon, P. & Verdier, P. (1991). J. Eur.
Ceram. Soc. 8, 197-213.

Newnham, R.E. & Ruschau, G.R. (1991). J. Am. Ceram. Soc. 74, 463-480.

Newnham, R.E. (1997). MRS Bulletin May, 20-34.

Nishihata, Y., Mizuki, Akao, T., Tanaka, H., Uenishi, M., Kimura, M., Okamoto, T. &
Hamada, N. (2002). Nature 418, 164-167.

Park, H.S., Yoon, K.YH. & Kim, E.S. (2001). J. Am. Ceram. Soc. 84, 99-103.

Peschel, S., Ziegler, B., Schwarten, M. & Babel, D. (2000). Z. Anorg. Allg. Chem. 626,
1561-1566.

Sakai, T., Adachi, G., Shiokawa, J. & Shin-ike,T. (1976). Mat. Res. Bull. 11, 1295-1300.

Shannon, R.D. (1976). Acta Cryst. A32, 751-767.

Shirane, G. Danner, H., & Pepinski, R. (1957). Physical Review V105, #3, 856-860.

Sleight, A.W., Gillson, J.L., & Bierstedt, P.E. (1975). Solid St. Comm. 17, 27-28.

Stitzer, K. E., Smith, M.D. & zur Loye, H.-C. (2002). Solid State Sciences 4, 311-316.

Töpfer, J. & Goodenough, J.B. (1997). J. Solid State Chem. 130, 117-128.

20
Trolier-McKinstry, S. & Newnham, R.E. (1993). MRS Bulletin April, 27-33.

Woodward, P. M. (1997a). Acta Cryst. B53, 32-43.

Woodward, P. M. (1997b). Acta Cryst. B53, 44-66.

Zhang, H., Teraoka, Y. & Yamazoe, N. (1987). Chem. Lett. , 665-668.

Zhao, Y., Weidner, D., Parise, J.B. & Cox, D. (1993). Phys. Earth and Plan. Int. 76, 17-
34.

21
 CHAPTER 2

STRUCTURE PREDICTION DIAGNOSTIC SOFTWARE (SPuDS)

2.1 Stimulus for Software Development

The prevalence and importance of octahedral tilting distortions provide clear

motivation to develop software capable of predicting distorted perovskite crystal

structures. A method to accurately predict distortions from ideal cubic symmetry was

expected to be useful considering the number of perovskite structures refined incorrectly

in the literature. One step toward this goal was the development of the program

POTATO (Woodward, 1997a), which was used in the high-pressure-high-temperature

synthesis of two new perovskites containing monovalent silver (Park et al., 1998).

Unfortunately, POTATO cannot easily be used for structure prediction because the

required input data (the M-X bond distance, the octahedral tilt system and the magnitude

of the tilting distortion) are not known in advance of synthesis and structural

characterization. This shortcoming motivated efforts to develop a more advanced

software package capable of predicting perovskite crystal structures directly from the

composition. The fruit of this labor is a new software package entitled SPuDS (Structure

Prediction Diagnostic Software).

SPuDS – (Structure Prediction Diagnostic Software) version 1.0 was developed in

its entirety to predict structures for single octahedral cation perovskites (Lufaso &
22
Woodward, 2001). Continued development of SPuDS has resulted in additional features

enabling 1:1 ordered M-site, disordered M-site, and Jahn-Teller distorted perovskite

systems to be modeled. SPuDS was expected to have a variety of potential applications

and examples of possible functions are shown in the list below.

1. Predicted structures can be used to estimate physical (magnetic, dielectric and other)

properties of both hypothetical compositions and those materials for which accurate

structural data is not available.

2. SPuDS can be used as a guide for exploratory synthetic efforts. It should be

particularly useful for compounds with multiple cations on the A-site (e.g.,

CaCu3Ti4O12) as well as high-pressure synthesis, where access to experimental

facilities is limited.

3. Predicted structures can serve as the starting point for Rietveld refinements in the

course of structurally characterizing new materials.

4. Structures generated by SPuDS can be compared with experimentally determined

structures in order to extract the effects of octahedral tilting distortions from other

structural distortion mechanisms.

2.2 Variable Reduction

Optimization of the structure in an unconstrained manner requires determination

of the unit-cell dimensions and all free positional parameters. In high symmetry and

multiple A-cation systems (groups A and B in table 1.1) the exact number of variables

23
that must be optimized varies from 3 to 10, depending upon the tilt system. Degrees of

freedom and the number of variables optimized in the structure calculation for all of the

single octahedral tilt systems are shown in table 1.1. Examination of a wide range of

perovskites shows that the octahedral environment does not undergo a large distortion

from six nearly equal bond distances, unless the distortion originates from an electronic

driving force (such as a Jahn-Teller ion on the M-site) influencing the distortion. In order

to simplify this process SPuDS restricts the octahedra to remain rigid (six equivalent M–

X distances and all X–M–X angles equal to 90º). This seems to be a reasonable

restriction in light of the fact that most distorted perovskites show very little distortion of

the MX6 octahedra, although obviously SPuDS will not work well in systems where

octahedral distortions are expected unless the distorted octahedra option is selected.

Additional computational methods were required to model systems with non-symmetric

octahedra. Subsequent developments featured in SPuDS version 2.0 include the

capability to model Jahn-Teller distortions in the single and ordered octahedral cation tilt

system a-b+a- (Pnma) and the ordered octahedral cation tilt system a0a0c- (I4/m).

The assumption that the octahedra remain symmetric, together with space group

and symmetry restrictions, leads to a significant reduction in the number of variables to

be optimized. Once this restriction is in place the full crystal structure can be generated

from two variables: the size of the octahedron and the magnitude of the octahedral tilting

distortion. Additional degrees of freedom must be taken into consideration in those tilt

systems where the A cation does not sit on a fixed position (e.g,. a-b+a-) or when

octahedral distortions cannot be avoided (a+a+c-). The reduction of the number of

24
variables is useful in reducing the computational time required in the structure

optimization.

2.3 Equation development for lattice parameters and atomic positions

Symmetry information associated with each space group and the atomic positions

determined by the tilt equations are needed in order to determine the bond lengths and

generate a complete crystallographic description of the structure. Space groups,

approximate unit-cell size, cation and anion positions of each tilt system have been

previously derived (Woodward, 1997b). Lattice parameters are based on the linear

distance between octahedral cations and decrease as the tilt angle increases. Equations for

determining the X positions and lattice parameters based on tilt angle for Glazer (1972)

tilt systems a+a+a+, a-a-a-, a0b+b+ and a-b+a- were taken as derived by O’Keeffe & Hyde

(1977). Equations for the remaining tilt systems for a single octahedral cation were

derived geometrically as a function of the octahedral tilt angle and are listed in table 2.1.

Equations for lattice parameter and fractional coordinate equations for 1:1 ordered M-site

cations were derived in a similar fashion; however, two different sized octahedra were

used. In some tilt systems the equations are quite complex and lengthy and are therefore

presented in appendix A. Lattice parameter equations based on M-X bond distance (d)

and tilt angle for the ten tilt systems with a single octahedral cation modeled in SPuDS

are shown in the table 2.1.

25
 Space Lattice Parameters
Glazer Tilt
Group (d = M-X bond distance)
a0a0a0 (#23) Pm 3 m a = 2d

a0a0c- (#22) I4/mcm a = 2d 2 cosφ

c = 4d

a0a0c+ (#21) P4/mbm a = 2d 2 cosφ

c = 2d
a=2 2d
0 - -
a b b (#20) Imma b = 4d cosθ
c = 2d 2 cosθ
a = 4d cosθ
a0b-c+ (#17) Cmcm b = 2d(cosθ + 1)
c = 2d(cosθ + 1)
a = 2d(1+ cosθ)
a0b+b+ (#16) I4/mmm
c =4d cosθ
a = 8 d cosΦ
a-a-a- (#14) R3c
c = 48 d
a = d 8(2 + cos 2 ω / 3
a+b-b- (#10) Pnma b = d 48 /(1 + 2 sec 2 ω )
c = 8 d cosω
a = 2d{cosφ+sinφ -cosθ[sinφ-cosφ]}
a+a+c- (#5) P42/nmc
c = 4dcosθ
a+a+a+ (#3) Im 3 a = d(8cosΦ+4)/3

Table 2.1 : Lattice parameter equations based on M-X bond distance (d) and tilt angle.
The angle φ is the octahedral tilt about the cubic [001], θ is octahedral tilt angle about the
cubic [110], Φ is the octahedral tilt angle about the cubic [111] and ω is the octahedral tilt
angle about the cubic [0 1 1].

26
2.4 Calculation Methods

SPuDS optimizes the structure by incrementally changing the tilt angle and

evaluating the stability of the resulting structure (as described below) at each step. The

initial optimization is coarse to allow a wide range of structures to be calculated quickly

and the tilt angle increment is finer for successive (3-5) optimization routines resulting in

a determination of the optimal tilt angle. A flowchart describing the operation of SPuDS

is shown in figure 2.1. Details of the calculation algorithm are presented in the following

sections.

27
 User input: composition and oxidation states

Calculate M-X bond distance optimizing

octahedral cation bond valence sum

Tilt octahedra; retain structure that

minimizes the GII

If A cation has a free positional parameter, optimize

A cation position via valence mapping

Decrease tilt angle and valence mapping increment

size; Repeat until optimized structure (lowest GII)

Compare tilt systems using GII

Output complete structural and stability description

of lowest GII or user's choice of tilt system

Figure 2.1 : SPuDS operational flowchart.

28
 The size of the octahedron and the optimum magnitude of the octahedral tilting

distortion are calculated utilizing the bond valence model, which is used to quantitatively

describe bonding in ionic solids (Brown, 1978). The bond valence, sij, associated with

each cation-anion interaction is calculated using the equation below,

[( Rij − d ij ) / B ]
s ij = e
where dij is the cation-anion distance. The B parameter is empirically determined, but can

often be treated as a universal constant with a value of 0.37. Rij is empirically determined

for each cation-anion pair based upon a large number of well-determined bond distances

for the cation-anion pair in question. Values of Rij for oxides and fluorides can be found

in the literature (Brown & Altermatt, 1985; Brese & O'Keeffe, 1991). Atomic valences,

Vi(calc), of the A and M cations and X anion are calculated according to the equation

below by summing the individual bond valences (sij) about each ion.

Vi (calc) = ∑j sij

Six nearest neighbor anions are used for the octahedral cation, six nearest

neighbor cations for the X anion and twelve nearest neighbor anions for the A-site cation

are used in the calculation of the bond valence sums. No assumption is made about the

coordination number of the A-site cation (valences for twelve A-X interactions are

calculated in all cases), but the contribution to the atomic valence sum becomes smaller

as the A-X bond distance increases. The octahedral cations remain at fixed positions in

29
all space groups generated by simple tilting of the MX6 octahedra. The M-X bond

distance, which determines the size of the octahedron, is calculated so as to optimize the

bond valence sum of the octahedral cation. The X-M-X bond angles of the MX6

octahedra remain ideal (90°) and the M-X bond distances are held constant in the

calculations (with the exception of the tilt system a+a+c-, which is discussed in section

3.9). The valence sum of the A-site cation is varied by changing the magnitude of the

octahedral tilting distortion.

The optimized structure is one where the difference between the calculated bond

valence sum and the formal valence (equal to its oxidation state) of each ion is

minimized. This value, which is termed the discrepancy factor, di, (Rao et al., 1998), is a

measure of the lattice strains present in the compound. The discrepancy factor is

calculated according to the equation below,

di = Vi (ox) − Vi (calc)
where Vi(ox) is the formal valence and Vi(calc) is the calculated bond valence sum for the ith

ion. The overall structure stability is determined by comparing the calculated bond

valence sums with the ideal formal valences. This quantity is referred to as the global

instability index (GII) (Salinas-Sanchez et al., 1992) and is calculated according to

equation below.

∑ (d )
N
2
i
GII = i=1
N

30
 Variables involved in the GII equation are discrepancy factor (di) and N, which is

the number of atoms in the asymmetric unit. During the optimization process the

octahedral tilt angle is stepped incrementally, and the individual A-X and M-X bond

distances, discrepancy factors and global instability index are calculated at each step.

After the first optimization process, the procedure is repeated using smaller tilt angle

increments in order to minimize the GII. The stability of perovskite compositions with

different atoms, symmetry, tilt systems and structure can be evaluated by comparing the

GII. The GII value is typically <0.1 valence units (v.u.) for unstrained structures and as

large as 0.2 v.u. in a structure with lattice induced strains. Crystal structures with a GII

greater than 0.2 v.u. are typically found to be unstable, and reports of such structures are

usually found to be incorrect (Rao et al., 1998).

An illustrative example of the relationship between the GII and octahedral tilting

is to consider the two perovskite compositions SrTiO3 and SrZrO3. The bond valence sum

calculated tolerance factors are τ = 1.00 and 0.95 for SrTiO3 and SrZrO3, respectively. No

octahedral tilting is observed in the crystal structure of SrTiO3 (Hutton & Nelmes, 1981),

while SrZrO3 undergoes octahedral tilting (Kennedy et al., 1999). SPuDS calculations of

SrTiO3 show a continuous increase in the GII for increased octahedral tilting about a

single axis. SPuDS calculations for SrZrO3 show an ideal tilt angle is observed in which

the GII is minimized, driven by the bonding requirements of the A-cation. GII versus

octahedral tilt angle for SrTiO3 and SrZrO3 is shown in the figure 2.2.

31
 0.6 SrTiO3 SrZrO3
Tilt Angle = 0° Tilt Angle = 14.9°
0.5 t = 1.00 t = 0.95

0.4

0.3 A-site cation A-site cation

32
under-bonded over-bonded
Ideal Octahedral
0.2 Tilt Angle

Global Instability Index

0.1

0.0
0 5 10 15 20 25
Octahedral Tilt Angle

Figure 2.2 : GII versus octahedral tilt angle for SrTiO3 and SrZrO3 in tilt system a0a0c-.
2.5 Optimization of A-site cation position

In certain tilt systems (e.g., a-b+a- and a+a+c-) the A-cation position has one or

more free positional parameters, so that the A-X distances are not uniquely determined by

the tilt angle. This introduces additional degrees of freedom to the optimization process.

In these tilt systems the position of the A-site cation was optimized according to the

following procedure.

1. The octahedral tilt angle is adjusted in order to minimize the GII with the A-site
cations located at their highest symmetry positions (in the center of the cube defined
by the eight surrounding octahedral cations).
2. Each A-X bond valence is treated as a vector quantity. The magnitude of each
valence vector is set equal to the valence of that particular bond, and the direction of
the valence vector is set parallel to the bond.
3. The twelve A-X valence vectors are summed and the position of the A-site cation is
adjusted in order to minimize the magnitude of the resultant vector.
4. The octahedral tilt angle is adjusted again in order to minimize the GII for the new A-
site cation position.
5. Steps 2-4 are repeated until both the GII and the A-site valence vector sum are
minimized.

This optimization approach, designated valence mapping, weights the shorter

bonds more heavily in determining the A-site cation position. This has the effect of

moving the A-site cation to the most symmetrical coordination environment available

within the distorted anion framework. This environment optimizes the lattice energy and

would be expected for most ions. Ions with a stereoactive electron lone pair are an

obvious exception. A contour plot of the calcium bond valence sum over a range of

33
fractional positions in the orthorhombic Pnma structure of CaTiO3 is shown in figure 2.3.

This clearly shows the accuracy of this approach to positioning the A-site cation within

the tilted octahedral framework.

34
Figure 2.3 : Valence map contour plot of A-site cation for CaTiO3 in tilt system a-b+a-
(space group Pnma). ∆X and ∆Z are the differences in the fractional position from the
high symmetry position located at (½,¼,½). The valence of A-site cation is shown as the
free positional parameters are varied while holding the octahedral tilt angle at 14.60°. The
open circle is the SPuDS predicted position and the filled square is the literature position.

35
2.6 Tolerance Factor Based on Bond Valence

Shannon (1976) has tabulated ionic radii for a variety of coordination

environments and oxidation states. The tolerance factor equation requires the use of

twelve coordinate radii, but unfortunately twelve coordinate values are limited in

tabulated radii (Shannon, 1976). Therefore, ionic radii from lower coordination numbers

are typically extrapolated to twelve coordinate for use in the tolerance factor calculation.

The ionic radii calculated tolerance factor uses twelve coordinate A-site cation radii, six

coordinate M cation and two coordinate X anion radii and as described in section 1.6.

Alternatively, one can use the bond valence model to calculate the ideal A-X and

M-X bond distances. A symmetric octahedral site ion is assumed with 1) twelve

equidistant A-X bonds, each contributing a bond valence of 1/12th of the ideal A-site

oxidation state, and 2) six equidistant M-X bonds contributing a bond valence of 1/6th of

the ideal octahedral site oxidation state. These A-X and M-X bond distances are then

substituted in place of the sum of the ionic radii used in the equation below in order to

calculate a bond valence based tolerance factor.

d AX
τ=
2d MX

where τ is the tolerance factor and dAX and dMX are the ideal A-X and M-X bond

distances, respectively. SPuDS uses both the ionic radii and the bond valence parameters

separately to calculate the tolerance factor (whenever possible). The use of the bond

36
valence parameters for the ionic radii does not require any assumption of the coordination

environment, only the oxidation state and coordination number are required. A

comparison of ionic radii calculated tolerance factor and bond valence calculated

tolerance factor is shown in figure 2.4 for single octahedral cation perovskites. The bond

valence tolerance factor is generally smaller than the ionic radii calculated tolerance

factor. The ionic radii calculated tolerance factor is included in SPuDS for comparison to

previous investigations. However, all further references to tolerance factor will

correspond to the bond valence tolerance factor, unless noted.

An undersized A-site cation results in a tolerance factor less than unity.

Magnitude of the octahedral tilting distortion is related to the tolerance factor with more

octahedral tilting occurring for a composition with a smaller tolerance factor. The

tolerance factors are calculated in the software program SPuDS (Lufaso & Woodward,

2001) using bond valence parameters obtained from the Accumulated Table of Bond

Valence Parameters version 2.2 (Brown, 2002).

37
Figure 2.4 : Bond valence and ionic radii calculated tolerance factor with the solid line
representing tIonic = tBond Valence.

38
2.7 Temperature Dependant Bond Valence Parameters

SPuDS version 1.0 calculated crystal structures for ambient temperature structures

only. For this reason the predicative ability will be less accurate at temperatures above

and below room temperature. Temperature-dependent bond valence parameters based

oxidation state and coordination were described by Brown, Dabkowski & McCleary

(1997). The capability to change the bond valence parameters based on temperature was

added to the software program. As ions in different coordination and oxidation state

undergo different rates of change in size, the capability to model different temperatures

was expected to be useful in understanding phase transitions that occur in perovskites.

Default values are in place and the software user has the option to change the temperature

dependent bond valence parameters to correspond with the particular system of interest.

2.8 Software Development

SPuDS source code was written in its entirety and shares no code from other

sofware programs. The numerical calculations were written in FORTRAN 77. SPuDS

version 1.0 was coded for a text based interface and compiled using a MS-DOS based

compiler. The progression of the Microsoft® Windows® based operating systems led to

a decline in the support and compatibility of DOS based programs. A text based user

interface was cumbersome and prone to errors in user input. Use of several MS-DOS

based FORTRAN 77 compilers showed the creation of different executable program files

with the same source code led to an incompatibility of the MS-DOS based program in

Microsoft® Windows® NT. Potential usage of the SPuDS software in the scientific

39
community required stable operation on a variety of Microsoft® Windows® platforms,

including Windows NT. A search was undertaken for a software platform that would

increase the potential software user base, program stability, and the user friendliness of

the interface.

Microsoft® Visual Basic® was chosen as the software platform for the creation

of a graphical user interface (GUI). Creation of dynamic linked libraries (DLL’s) from

the modular FORTRAN 77 code allowed the Visual Basic® created GUI to call the

numerically efficient FORTRAN 77 coded subroutines. The source code contains

approximately 16700 lines for the FORTRAN code and 1000 for the Visual Basic code.

The end result of these efforts is a user friendly GUI based software program fully

compatible with all current (2002) Microsoft Windows based platforms. The SPuDS

software program executes the entire optimization procedure in approximately one

minute on an Intel® Pentium® class processor using the Microsoft® Windows®

operating system, for which the downloadable version of the software is compiled. The

software program was also compiled and executed on other platforms including several

types of Unix-based operating systems for the text based user input version 1.0.

An interesting extension of the capabilities of SPuDS would be to include

octahedral tilting in combination with cation ordering (A2MM'X6, A3M2M'X9, AA'M2X6),

Jahn-Teller distortions, anion-vacancy ordering (LnAM2O5, LnA2M3O8, A2M2O5) and

intergrowth phases (Ruddlesden-Popper, Aurivillius and Dion-Jacobson phases). The

software program SPuDS is available by contacting the author (current e-mail

40
mlufaso@chemistry.ohio-state.edu) or by free download at

http://www.chemistry.ohio-state.edu/~mlufaso/spuds/index.html.

2.9 Software Operation

SPuDS calculates structures for the six high symmetry and four multiple A-site

tilt systems for single octahedral cation perovskites. SPuDS does not calculate structural

information for the five single octahedral cation low symmetry/transitional tilt systems.

SPuDS calculates structures for rock-salt ordered octahedral cation in six tilt systems;

a+a+a+, a-b+a-, a-a-a-, a0b-b-, a0a0c- (space groups Pn 3 , P21/n, R 3 , I2/m, I4/m and Fm 3 m

respectively). Disordered octahedral cation structures are available for tilt systems a0a0c-

(I4/mcm) and a-b+a- (Pnma). Jahn-Teller distortions of the octahedral cations are available

in single M-site a-b+a- (Pnma) and ordered a-b+a- (P21/n) and a0a0c- (I4/m). The software

program SPuDS requires only the composition and oxidation state of each ion as its

input. Two calculation methods, 1) fixed M-X bond distance and 2) GII optimized

structure, are available for untilted perovskites. A tilt system is chosen to display the

results in a text box inside the program.

Output generated by SPuDS contains information including the space group,

lattice parameters, atomic coordinates, atomic valence sums, individual bond valences

and distances, tolerance factor, unit cell volume, octahedral tilt angles, X-M-X bond

angles and GII for each of the evaluated tilt systems. Comparison of SPuDS structures

may require examination of equivalent symmetry positions for the atoms. A conversion

often used is from the non-standard setting Pbnm to the standard setting Pnma.

Conversion of an anion position to an equivalent Wyckoff positions on the same site is

41
needed for comparison of experimental and predicted structures. In some structures a

shift of the unit cell (e.g. ½ in the x direction) is necessary for comparison of the SPuDS

predicted and experimental crystal structure. The International Tables for

Crystallography, Vol. A (1995) contains all the symmetry information necessary to

determine equivalent atoms.

SPuDS also has a feature useful for most aspects of solid state crystal structure

solution with the incorporation of an option to calculate the bond valence sum of cations

from experimental crystal structures. Selection of the cation-anion pair and number of

bonds is user selected. Bond distances from the experimental crystal structure are input

into the text boxes. Clicking “Calculate Bond Valence Sum” displays the bond valence

sum for the cation. This feature is particulary useful in checking the feasibility of

experimental crystal structure bond distances.

2.10 References

Brese, N. E. & O'Keeffe, M. (1991). Acta Cryst. B47, 192-197.

Brown, I.D (1978). Chem. Soc. Rev. 7, 359-376.

Brown, I.D. & Altermatt, D. (1985). Acta Cryst. B41, 244-247.

Brown, I. D., Dabkowski, A. & McCleary, A. (1997). Acta Cryst. B53, 750-761.

Glazer, A.M. (1972). Acta Cryst., B28, 3384-3392.

Hutton, J. & Nelmes, R.J. (1981). J. Phys. C.: Solid State Phys. 14, 1713-1736.

International Tables for Crystallography (1995). Vol. A, Space Group Symmetry, 3rd
ed., edited by T. Hahn. Dordrecht: Kluwer Academic Publishers.

42
Kennedy, B.J., Howard, C. J. & Chakoumakos, B.C. (1999). Phys. Rev. B. 59(6), 4023-
4027.

Lufaso, M.W., & Woodward, P.M. (2001). Acta Cryst. B57, 725-738.

O'Keefe, M., & Hyde, B.G. (1977). Acta Cryst. B33, 3802-3813.

Park, J.-H., Woodward, P.M. & Parise, J.B. (1998). Chem. Mater. 10, 3092-3100.

Rao, G.H., Barner, K. & Brown, I.D. (1998). J. Phys.: Condens. Matter 10, L757-L763.

Salinas-Sanchez, A., Garcia-Munoz, J.L., Rodriguez-Carvajal, J., Saez-Puche, R. &

Martinez, J.L. (1992). J. Solid State Chem. 100, 201-211.

Shannon, R.D. (1976). Acta Cryst. A32, 751-767.

Woodward, P. M. (1997a). J. Appl. Cryst. 30, 206-207.

Woodward, P. M. (1997b). Acta Cryst. B53, 44-66.

43
 CHAPTER 3

STRUCTURE PREDICTION OF SINGLE OCTAHEDRAL CATION PEROVSKITES

3.1 Introduction

A large number of crystal structures are available in the literature to compare the

experimental crystal structure determinations to the results of SPuDS calculations. In the

following sections the ten tilt systems SPuDS calculates are evaluated by comparing to

known crystal structures. If no structural examples are known, hypothetical compositions

are explored and compositions with a low GII examined.

3.2 a0a0a0 (Pm 3 m).

The undistorted cubic perovskite structure has one degree of freedom (the cell

edge) to satisfy both the A-X and M-X bonding requirements. In most cases these two

bonding interactions will not be perfectly matched, and the equilibrium structure will

represent a compromise. To evaluate the nature of this compromise the cell edge is

calculated for several room temperature cubic oxide and fluoride perovskites using three

different approaches:(a) optimizing the M-cation valence, (b) optimizing the A-cation

valence, (c) minimizing the GII.

44
 The results are compared with observed cell edges in table 3.1. Experimental

crystal structures used were KMgF3 (Zhao et al., 1996), KNiF3 (Kijima et al., 1983),

KZnF3 (Buttner et al., 1988), KCoF3 (Kijima et al., 1981), KFeF3 (Miyata et al., 1983),

BaLiF3 (Zhao et al., 1996), KMnF3 (Kijima et al., 1983), RbCaF3 (Hutton & Nelmes,

1981), KTaO3 (Zhurova et al., 1995), SrGeO3 (Shimizu et al., 1970), BaNbO3 (Svensson

et al., 1990), BaMoO3 (Brixner, 1960), SrVO3 (Rey et al., 1990), BaSnO3 (Smith &

Welch, 1960), KUO3 (Dickens & Powell, 1991), BaZrO3 (Roth, 1957), SrTiO3 (Hutton et

al., 1981) and SrMoO3 (Liu et al., 1992). The results show, in general, the M-X bonding

dictates the length of the cell edge. This is particularly true when the A-X interactions are

highly ionic and the tolerance factor is much larger than unity. This result is not too

surprising considering the fact that highly ionic A-X bonds would be expected to show

greater flexibility. In general it will be energetically more favorable to compress the A-X

bonds than to stretch the M-X bonds.

45
 Literature M-Cation A-Cation GII
Valence Valence Minimized
Sum Sum
Optimized Optimized
Formula Tolerance a (Å) a (Å) a (Å) a (Å) GII
Factor (v.u.)
KMgF3 1.04 3.95 3.97 4.12 4.03 0.108
KNiF3 1.03 4.01 4.01 4.12 4.05 0.081
KZnF3 1.02 4.06 4.05 4.12 4.08 0.046
KCoF3 1.01 4.07 4.09 4.12 4.10 0.017
KFeF3 1.00 4.12 4.11 4.12 4.11 0.004
BaLiF3 1.00 4.00 4.05 4.17 4.04 0.009
KMnF3 0.98 4.19 4.21 4.12 4.18 0.062
RbCaF3 0.97 4.45 4.50 4.35 4.45 0.095

KTaO3 1.09 3.99 3.97 4.32 4.04 0.368

SrGeO3 1.04 3.80 3.80 3.93 3.86 0.199
BaNbO3 1.03 4.09 4.06 4.17 4.11 0.157
BaMoO3 1.02 4.04 4.07 4.17 4.11 0.140
SrVO3 1.02 3.84 3.87 3.93 3.89 0.093
BaSnO3 1.01 4.12 4.11 4.17 4.14 0.084
KUO3 1.01 4.30 4.28 4.32 4.29 0.027
BaZrO3 1.00 4.19 4.16 4.17 4.16 0.018
SrTiO3 1.00 3.91 3.93 3.93 3.93 0.006
SrMoO3 0.97 3.97 4.07 3.93 4.03 0.186

Table 3.1 : Experimental and SPuDS predicted global instability index, lattice
parameters (optimized via M-cation valence sum, A-cation valence sum and minimized
GII) for cubic untilted perovskite fluorides and oxides.

46
3.3 a0a0c- (I4/mcm)

Most of the compounds that adopt the a0a0c- and a0a0c+ tilt system do so at non-

ambient temperatures. CsAgF3 and RbAgF3 (Odenthal & Hoppe, 1971) structure

determinations were performed at room temperature in I4/mcm, but no perovskite oxides

are reported in the literature at ambient temperature. The lack of perovskite oxides that

adopt the a0a0c- tilt system at ambient conditions prevents rigorous evaluation of SPuDS

predictive capabilities in these tetragonal systems. CsAgF3 and RbAgF3 were evaluated

using SPuDS, however the AgF6 octahedra exhibit a significant distortion in Ag-F bond

lengths and SPuDS is not currently designed to handle this type of distortion. In fact it

may well be that the tetragonal distortion of the octahedra stabilizes the tilting distortion,

rather than the other way around.

3.4 a0a0c+ (P4/mbm)

Only CsDyBr3 crystallizes as a perovskite in P4/mbm at room temperature

(Hohnstedt & Meyer, 1993). SPuDS calculations for the CsDyBr3 composition is not

possible due to the lack of bond valence parameters available for a Dy(II)-Br atom pair.

The lack of perovskite oxides that adopt a0a0c+ tilt system at ambient conditions prevents

us from rigorously evaluating SPuDS predictive capabilities in these tetragonal systems.

The calculation methods employed in SPuDS cannot determine which of the two tilt

systems a0a0c- and a0a0c+ are more favorable, since all nearest neighbor distances are

equivalent.

47
3.5 a0b-b- (Imma)

An additional phase transition was recently determined in SrZrO3 (Howard et al.,

2000b), thus there are six known representatives of the a0b-b- tilt system (Imma). The A-

site cation is Ba in four of the six structures. A large and polarizable A-site cation

together with a tolerance factor intermediate between Pnma and Pm 3 m appear to

stabilize the a0b-b- tilt system. Simulations of a variety of compounds were performed,

but the a-b+a- and a-a-a- invariably had a lower calculated GII. The a-b+a- tilt system has a

lower GII due to the additional degrees of freedom and a-a-a- has a lower GII due to the

symmetric collapse of the octahedra around the A-site cation. Clearly additional

calculation methods are needed in order to understand the stability of this tilt system in

more detail.

3.6 a-b+a- (Pnma).

As shown in figure 1.2 the most common space group for simple perovskites is

Pnma, corresponding to the a-b+a- tilt system. This structure is also commonly referred to

as the GdFeO3 structure. The true crystallographic cell has Z = 4, with a ≈ c ≈ √2 ap and

b ≈ 2ap (where ap is the unit cell edge length for the undistorted cubic perovskite). The

a-b+a- tilt system maximizes the A-X covalent bonding and minimizes the repulsive A-X

overlap (Woodward, 1997b). This distortion is most common when the ionic radii based

tolerance factor becomes less than 0.98 or when the A-site cation becomes relatively

electronegative (e.g., Ca2+). The symmetry of the Pnma space group is such that there

are ten variables (five fractional coordinates corresponding to the oxygen positions, two

48
for the A-site cation, and three variables defining the size of the orthorhombic unit cell),

which must be specified in order to completely describe the crystal structure.

In order to assess the accuracy of SPuDS, structures were predicted for a

collection of structurally characterized Pnma perovskites. The structural information was

tabulated and the experimental and calculated octahedral tilt angles are plotted versus the

bond valence tolerance factor in figure 3.1. The literature octahedral tilt angle was

calculated from the reported fractional coordinates of the anion sites. Except for

compositions where the tolerance factor approaches unity SPuDS is able to predict the tilt

angle with a high degree of accuracy. Compounds experimentally determined to

crystallize in the Pnma (or a non-standard setting of Pnma) space group were tabulated

by Woodward (1997b). Additional compounds that crystallize in the a-b+a- tilt system are

listed in table 3.2.

49
Figure 3.1 : Octahedral tilt angle obtained from the optimized SPuDS structure versus
the bond valence tolerance factor for tilt system a-b+a-.

50
 Compound Reference
CaIrO3 Sarkozy et al. (1974)
AOsO3 (A=Ca, Sr) Shaplygin & Lazarev (1976)
AAlO3 (A=Lu, Yb) Anan'eva et al. (1978)
AVO3 (A=Dy, Gd) Pickardt et al. (1988)
ARuO3 (A=La, Pr) Kobayashi et al. (1994)
PrGaO3 Marti et al. (1994)
AAlO3 (A=Eu, Gd, Er, Tb) Shishido et al. (1995)
SrPrO3 Hinatsu & Itoh (1996)
YCoO3 Mehta et al. (1997)
ScCrO3 Park & Parise (1997)
SrHfO3 Kennedy et al. (1999b)
LaGaO3 Howard & Kennedy (1999)
ScAlO3 Ross (1999)
CaPbO3 Yamamoto et al. (1999)
ANiO3 (A=Dy, Eu, Gd) Alonso et al. (1999)
NdGaO3 Vasylechko et al. (1999)
AMnO3 (A=Er, Dy, Ho, Y) Alonso et al. (2000)
LaMO3 (M=Ho, Er, Tm, Yb, Lu) Ito et al. (2001)
CeMO3 (M=Tm, Yb, Lu) Ito et al. (2001)
PrMO3 (M=Yb, Lu) Ito et al. (2001)

Table 3.2 : Compounds in the a-b+a- (Pnma) tilt system not included in reference
Woodward (1997b).

51
 The A-site cation positions calculated by SPuDS are compared to literature

positions in a graph of the A-site cation displacement (in terms of absolute displacement

from the center of the simple cubic unit cell) versus tolerance factor in figure 3.2. The A-

site cation shift in Pnma perovskites is driven by the tilting of the octahedra. The

equilibrium position of the A-site cation within the tilted octahedral framework

represents the most symmetric coordination environment that can be attained for the A-

site cation. The valence vector mapping approach attempts to replicate this placement of

the A-site cation. First, note that SPuDS displaces the A-site cation off the undistorted

location in the same direction as observed experimentally. In general the prediction is

accurate, but at tolerance factors near unity SPuDS tends to underestimate the magnitude

of the A-site cation movement, while for large tilt angles the shift of the A-site cation is

slightly overestimated. The lattice parameters are dependent upon the magnitude of the

tilting distortion and the M-X bond distances. The accuracy of lattice parameter

prediction is demonstrated in a plot of percent error in lattice parameter vs. tolerance

factor (see figure 3.3). The average percent error in the lattice parameter for perovskites

in the a-b+a- tilt system is 1.0% for a, 0.9% for b and 0.8% for c lattice parameter.

52
Figure 3.2 : A-site cation displacement from high symmetry location in x and z obtained
from the optimized SPuDS structure versus the bond valence tolerance factor for tilt
system a-b+a-.

53
Figure 3.3 : Percent lattice parameter error [100×(SPuDS pred. – exp.)/exp.] versus the
tolerance factor for tilt system a-b+a-.

54
 To give a better feel for the absolute accuracy of the approach, predicted

structures were examined in detail for four specific perovskites: GdFeO3 (Marezio et al.,

1970), CaTiO3 (Sasaki et al., 1987), SrZrO3 (Kennedy et al., 1999a) and CaSnO3 (Vegas

et al., 1986). Each of these perovskites crystallizes in the Pnma space group, and the

experimental and predicted structural information is shown for comparison in table 3.3.

The experimentally determined structures in the literature are often solved using the

nonstandard space group Pbnm or Pcmn. The atomic positions from the original

experimental structures were converted to Pnma space group and to an equivalent

Wyckoff atomic position as calculated in SPuDS to allow easier comparison between

experimental and SPuDS predicted structures. The calculations performed using SPuDS

illustrate the accuracy of this approach in predicting the structural trends in the a-b+a- tilt

system. When examined in this fashion, one can see that in an absolute sense the

fractional coordinates are determined more accurately than the lattice parameters. This

can be attributed to the well-known fact that lattice parameters are very sensitive to

distortions of the octahedra.

55
 Literature Atomic Fractional Coordinates
Formula A(4c) x A(4c) z O(4c) x O(4c) z O(8d)x O(8d)y O(8d)z
GdFeO3 0.563 0.516 -0.033 0.400 0.302 0.051 0.696
CaTiO3 0.536 0.507 -0.016 0.429 0.289 0.037 0.711
SrZrO3 0.524 0.504 -0.013 0.427 0.285 0.035 0.716
CaSnO3 0.512 0.499 -0.010 0.426 0.285 0.037 0.713
SPuDS Atomic Fractional Coordinates
GdFeO3 0.568 0.524 -0.021 0.394 0.299 0.053 0.697
CaTiO3 0.534 0.511 -0.016 0.425 0.286 0.038 0.712
SrZrO3 0.534 0.511 -0.011 0.426 0.285 0.037 0.713
CaSnO3 0.526 0.508 -0.008 0.434 0.282 0.033 0.717
Literature Tilt Angle and Lattice Parameters
Magnitude Lattice Parameters (Å)
Formula Tilt Angle (º) a b c
GdFeO3 20.30 5.611 7.669 5.349
CaTiO3 14.96 5.442 7.640 5.380
SrZrO3 13.69 5.817 8.171 5.796
CaSnO3 14.31 5.681 7.906 5.532
SPuDS Tilt Angle and Lattice Parameters
GdFeO3 20.13 5.587 7.723 5.352
CaTiO3 14.60 5.499 7.688 5.379
SrZrO3 14.36 5.817 8.136 5.694
CaSnO3 12.80 5.765 8.082 5.668

Table 3.3 : Experimental and predicted structural information for typical perovskites
crystallizing in the Pnma space group.

56
3.7 a+a+a+ (Im 3 )

Perovksites which undergo the a+a+a+ octahedral tilting distortion crystallize in

the cubic space group Im 3 . The octahedra tilt in phase an equivalent amount about each

of the three cubic axes. The general formula for structures found in this tilt system is

A′A″3M4O12. The A-site cations are at fixed positions in this tilt system, with the A' and

A" having two different coordination environments. The A' cation is at an icosahedral

site with twelve equidistant anions and its coordination environment remains symmetric

and contracts rather slowly as the octahedra tilt. The A″ cation is on a square planar site

that contracts much more rapidly as the octahedral tilt angle increases. A Jahn-Teller ion

is well adapted for this coordination environment, and all known examples of a+a+a+

perovskites contain either Cu2+ or Mn3+ on the A" site. The coordination environments of

the both A-site cations are shown in figure 3.4.

57
Figure 3.4 : The crystal structure of CaCu3Ti4O12, showing the (top) octahedral
environment of Ti4+, (middle) icosahedral environment of Ca2+, and (bottom) square
planar environment of Cu2+.

58
 Perovskites in this tilt system are often synthesized under high pressure.

Perovskites of the formula CaCu3M4O12 (M = Ge, Mn, Ti, Ru) have been synthesized

(Ozaki et al., 1977), (Bochu et al., 1979), (Deschanvres et al., 1967), (Labeau et al.,

1980), respectively. A sodium containing perovskite NaMn7O12 (Marezio et al., 1973)

has also been made. Recently it has been shown that the perovskite CaCu3Ti4O12 exhibits

fascinating behavior as a dielectric material. A rather high dielectric constant of

approximately 12,000 at 1 kHz has been recently observed and is nearly constant from

room temperature to 573 K (Subramanian et al., 2000). The dielectric constant lowers

nearly 100 fold near 100 K, with no apparent structural transition. The origin of this

effect and its mechanism are still not well understood (Ramirez et al., 2000).

The CaCu3M4O12 (M = Ge, Mn, Ti, Ru) formulas were evaluated with SPuDS and

the bond valence sums, GII, lattice parameters, oxygen positions, octahedral tilt angle and

synthesis conditions are in table 3.4. The optimal GII in calculated tilt systems are

illustrated in figure 3.5. Comparing across all of the calculated tilt systems, it is observed

that the lowest GII is observed in the a+a+a+ tilt system. From this data one can see that

the global instability index for CaCu3Ti4O12 is substantially lower than for the M = Ge,

Mn, Ru structures. This observation helps to rationalize the fact that CaCu3Ti4O12 is the

only compound of the four that can be synthesized at atmospheric pressure. These results

also provide some insight regarding the values of GII, which might be expected to result

in successful high-pressure synthetic attempts.

59
 Bond Valence Sums (v.u.)
M Ca Cu M O
Ge 2.35 1.93 4.00 2.01
Mn 2.33 1.94 4.00 2.01
Ru 1.93 2.02 4.00 2.00
Ti 2.01 2.00 4.00 2.00

M GII a (Å) O(y) O(z) Tilt Angle (°) Synthesis

(v.u.) Pressure (kbar)
Ge 0.084 7.265 0.2977 0.1910 20.7 50-70
7.202 0.3012 0.1859 20.7
Mn 0.078 7.283 0.2979 0.1908 20.8 50
7.241 0.3033 0.1822 20.9
Ru 0.019 7.472 0.3050 0.1795 24.2 2
7.421 - - -
Ti 0.003 7.426 0.3036 0.1818 23.5 Ambient
7.391 0.3038 0.1786 23.5

Table 3.4 : Bond valence sums, global instability index, lattice parameter, oxygen free
positional parameters, octahedral tilt angle and synthesis pressure for known
CaCu3M4O12 (M=Ge, Mn, Ru, Ti) perovskites reported in the a+a+a+ tilt system. Values in
italics represent experimentally observed values.

60
Figure 3.5 : SPuDS modeling for known compounds of tilt system a+a+a+. The calculated
GII for tilt system a+a+a+ are given in table 3.4. The calculated GII for the next lowest tilt
system (a0b-c+) are (0.237,CaCu3Ti4O12), (0.232, CaCu3Ru4O12), (0.266, CaCu3Mn4O12),
(0.268, CaCu3Ge4O12).

61
3.8 a-a-a- (R 3 c).
The a-a-a- tilt system crystallizes in the trigonal space group R 3 c with three

degrees of freedom. This space group has a single crystallographic site for each of the A,

M, and X ions. Known compositions in the a-a-a- tilt system with exclusively lanthanum

as the A-site cation are LaNiO3 (García-Munoz et al., 1992), LaCuO3 (Demazeau et al.,

1972), LaAlO3 (Howard et al., 2000a), LaCoO3 (Thornton et al., 1986) and LaGaO3

(Howard & Kennedy, 1999). The ionic radii tolerance factor is greater than unity (1.003,

1.014, 1.017, 1.011 and 0.973 respectively) for four out of the five compounds, so that by

simple arguments one might expect an octahedral tilting distortion only for LaGaO3. In

contrast, the bond valence tolerance factor is less than unity (0.946, 0.945, 1.012, 0.970

and 0.956 respectively) for four of the five compounds. Thus evaluation using bond

valences instead of ionic radii helps to explain why these compounds undergo octahedral

tilting distortions in the first place (except for LaAlO3). Calculations were carried out on

these compounds and compared to experimentally determined structures. The octahedral

tilt angle predicted by SPuDS follows a smooth curve over a range of tolerance factors, as

one would expect. However, when the experimentally determined crystal structures are

examined, a clear relationship between tilt angle and tolerance factor does not appear.

The octahedral tilt angle in these compounds appears independent of the geometry-based

tolerance factor. This result seems to indicate that in the a-a-a- tilt system something other

than the valence requirements of the A-site cation drives the octahedral tilting distortion

(such as distortions of the octahedra).

62
3.9 a+a+c- (P42/nmc).

The symmetry for Glazer tilt system a+a+c- is correctly described in the tetragonal

space group P42/nmc rather than the original assignment of Pmmn (Leinenweber &

Parise, 1995). A vector proof has demonstrated distortions of the octahedra are necessary

to retain corner-sharing connectivity in the tilt system a+a+c- (Woodward, 1997a).

Howard and Stokes (1998) subsequently confirmed this conclusion. Thus the restriction

that the octahedra remain rigid cannot be strictly applied in this tilt system. Furthermore,

the situation is complicated by the fact that the octahedra can distort in a number of

different ways. The approach to this problem was to determine the positions of the

X1(8f) and one of the X2(8g) atoms directly from the tilt angles (there are two distinct tilt

angles). The two free positional parameters of the X3 anion, also at Wyckoff position 8g,

are varied and each of the cis-X-M-X bond angles are calculated. The positional

parameters for the third X anion are assigned the values in which the sum of the

difference between each cis-X-M-X bond angle and 90º is a minimum. Using this

approach for determining the position of the X3 anion, the M-X(8g') bond length is

altered from its ideal value. The calculation method has the effect of allowing a

distortion of the octahedral bond lengths, while retaining approximately 90° X-M-X

angles.

Evaluation of SPuDS accuracy for a+a+c- perovskites is difficult due to the fact

that this tilt system is very uncommon. The only perovskite synthesized in the a+a+c- tilt

system thus far is CaFeTi2O6 (Leinenweber & Parise, 1995). It was synthesized under

high-pressure-high-temperature conditions. The unit cell volumes, global instability

63
indexes, bond distances and valence sums, and O-Ti-O bonds angles for CaFeTi2O6 are

shown in table 3.5. The Fe is somewhat under-bonded, while the Ca is over-bonded. The

SPuDS calculations show the a+a+c- tilt system has the lowest GII of any of the calculated

tilt systems. The GII of each tilt system for the modeling of CaFeTi2O6 is shown in

figure 3.6.

SPuDS Literature Bond Distances (Å) SPuDS Literature

GII (v.u.) 0.108 0.129 Ti(8c)-O(8f) (x2) 1.97 1.97
a (Å) 7.59 7.52 Ti(8c)-O(8g) (x2) 1.97 1.94
c (Å) 7.48 7.55 Ti(8c)-O(8g) (x2) 1.98 1.97
3
Unit Cell Volume (Å ) 430.9 426.7
Bond Valence Sums (v.u.) Fe(2a)-O(8f) (x4) 2.07 2.10
Fe(2a) 1.90 1.79 Fe(2a)-O(8g) (x4) 3.18 3.15
Fe(2b) 1.78 1.85 Fe(2a)-O(8g) (x4) 2.86 2.84
Ca(4d) 2.26 2.30 Fe(2b)-O(8f) (x4) 3.30 3.22
Ti(8c) 3.91 4.05 Fe(2b)-O(8g) (x4) 2.73 2.80
O(8f) 1.97 1.97 Fe(2b)-O(8g) (x4) 2.11 2.08
O(8g) 2.03 2.14
O(8g) 1.96 2.00 Ca(4d)-O(8f) (x4) 2.77 2.72
O-Ti-O Bond Angles (º) Ca(4d)-O(8f) (x2) 2.34 2.33
O(8f)-Ti-O(8g) 90.0 89.1 Ca(4d)-O(8g) (x2) 2.48 2.46
O(8f)-Ti-O(8g)' 89.9 89.0 Ca(4d)-O(8g) (x2) 2.46 2.51
O(8g)-Ti-O(8g)' 89.8 89.8 Ca(4d)-O(8g) (x2) 3.42 3.39

Table 3.5 : SPuDS predicted and experimental GII, lattice parameters, unit cell volume,
bond valence sums, bond distances and O-Ti-O bond angles for CaFeTi2O6.

3.10 a0b+b+ (I4/mmm).

There are no known examples of simple AMX3 perovskites that crystallize in tilt

system a0b+b+ (I4/mmm). The tilt system has three A-site cation Wyckoff sites (2a, 2b and

4c) and five degrees of freedom. The three different A-site cation positions enable a large

64
number of A-site cation combinations to be evaluated. The A-site cation located on the

2a position has four short and eight long A-X bonds, the 2b position has eight short and

four long A-X bonds and the 4c position has four short, four medium and four long A-X

bonds after octahedral tilting has occurred. The 2a position has a square planar

coordination, so that an atom such as Ni2+, Cu2+, Pd2+, or Pt2+, which are known to adopt

a square planar coordination, will be atoms most likely to occupy this site. A variety of

atoms were inserted at the other A-site positions in order to obtain a low global instability

index. Hypothetical structures were evaluated and the bond valence sums, GII and

octahedral tilt angle of the three most favorable compositions are shown in table 3.6. The

global instability index for each of the tilt systems that SPuDS calculates is shown for

PdCdCa2Ti4O12 in figure 3.6. The GII for a0b+b+ is lower than the other tilt systems, but

there is not a large difference between the GII of the next lowest tilt system, a+a+c-. The

small difference in stability between the two lowest tilt systems implies that there is not a

significant driving force to adopt the a0b+b+ tilt system, though it is worthwhile to note

that a+a+c- is also a very uncommon tilt system. Furthermore, one must not forget that the

formula PdCdCa2Ti4O12 in a0b+b+ is not necessarily the most stable phase for that

combination of elements. For example, the stable phase CaTiO3 plus other phases may

form instead.

65
Figure 3.6 : SPuDS predicted GII values for PdCdCa2Ti4O12, Ca2Cd2Ti4O12 and
CaFeTi2O6 in each of the calculated tilt systems. CaFeTi2O6 was synthesized
(Leinenweber & Parise, 1995) under high-pressure high-temperature conditions and the
structure solved in the a+a+c- tilt system. PdCdCa2Ti4O12 and Ca2Cd2Ti4O12 are
hypothetical compounds designed to adopt tilt system a0b+b+ and a0b-c+, respectively. The
bond valence sums, GII and tilt angle for Ca2Cd2M4O12 (M=Ti, Ru, Ge) and
PdCdCa2M4O12 (M=Ti, Os, Ru) are given in table 3.6 and 3.7, respectively.

66
 Bond Valence Sums (v.u.)
Compound A1(4c) A2(4c) M X(8e) X(8f) X(8g) GII (v.u.) Tilt Angle (º)
CaCdGe2O6 2.12 1.90 4.00 2.01 2.01 1.99 0.052 11.5
CaCdTi2O6 2.00 1.99 4.01 2.00 2.03 1.98 0.017 16.6
CaCdRu2O6 1.96 2.02 4.02 1.99 2.03 1.98 0.025 17.7

Table 3.6 : Calculated bond valence sums, GII and octahedral tilt angle for hypothetical
compounds designed to adopt the a0b+b+ tilt system.

3.11 a0b-c+ (Cmcm)

Examples of perovskite oxide structures that have been reported in this space

group are all high temperature polymorphs, including SrZrO3 (970-1100 K) (Kennedy et

al., 1999a), NaNbO3 (793-848 K) (Darlington & Knight, 1999ab), NaTaO3 (773-843 K)

(Kennedy et al., 1999d) and CaTiO3 (1380-1500 K) (Kennedy et al., 1999c). There are

ten degrees of freedom for this tilt system and two A-site cations at Wyckoff position 4c

with a slightly different coordination. A structure that might crystallize in this tilt system

would most likely have two A-site cations of similar ionic radius. Using this approach, a

wide variety of hypothetical structures were evaluated and the three most promising

compositions are shown in table 3.7, with the most promising composition shown in

figure 3.6. The calculated global instability indexes were similar for several nearby tilt

systems. Hence there is not a large structural driving force to stabilize this tilt system and

multiphase mixtures may well be more stable than a single-phase a0b-c+ perovskite. The

small difference in GII between the different tilt systems provides insight as to why this

tilt system is not observed at ambient temperature and only a few compounds are

observed in a higher temperature range.

67
 Bond Valence Sums (v.u.)
Compound A(2a) A(2b) A(4c) M X(8h) X(16n) GII (v.u.) Tilt Angle (º)
PdCdCa2Ti4O12 1.97 1.97 2.04 4.00 2.00 2.00 0.016 17.8
PdCdCa2Os4O12 1.96 1.97 2.05 4.00 2.00 2.00 0.019 17.6
PdCdCa2Ru4O12 2.03 1.93 2.01 4.00 2.00 2.00 0.017 18.7

Table 3.7 : Calculated bond valence sums, GII and octahedral tilt angle for hypothetical
compounds designed to adopt the a0b-c+ tilt system.

3.12 Conclusions

The software program SPuDS has been developed for predicting the structures of

perovskite compounds. The optimization procedure is based on the bond valence

method, and requires only the composition as user input. Predictions for existing

compounds confirm the validity of this approach. SPuDS could be useful for a variety of

purposes, such as evaluating the stability and properties of new perovskite materials,

and/or generating accurate starting models for structure refinements.

SPuDS is capable of predicting fractional coordinates for members of the a-b+a-

(Pnma) and a+a+a+ (Im 3 ) tilt systems, as well as undistorted perovskites, with a high

degree of accuracy. The prediction of unit cell parameters is not quite as good as the

predictions of atomic fractional coordinates, due to the effects of octahedral distortions,

but the predicted values are consistently within 1% of the observed values. For the a+a+a+

tilt system the GII calculated by SPuDS appears to correlate with the pressure required

for phase stabilization and successful synthesis. Clear cut conclusions cannot be drawn

for intermediate tilt systems [a0a0c-, a0a0c+, a0b-b-, a-a-a-] between a-b+a- and a0a0a0 due to

68
the relatively small number of representatives in these tilt systems. However, it appears

that distortion mechanisms other than octahedral tilting must be taken into account in

order to fully understand these systems. The complete absence of compounds that adopt

the a0b+b+ and a0b-c+ tilt systems under ambient conditions is a consequence of the fact

that the A-site coordination environments in these structures are not sufficiently distinct

to effectively stabilize A-site cation ordering.

3.13 References

Alonso, J.A., Martínez-Lope, M.J., Casais, M.T., Aranda, M.A.G. & Fernández-Díaz,
M.T. (1999). J. Am. Chem. Soc.,121, 4754-4762.

Alonso, J.A., Martínez-Lope, M.J., Casais, M.T. & Fernández-Díaz, M.T. (2000). Inorg.
Chem., 39, 917-923.

Anan'eva, G. V., Ivanov, A. O., Merkulyaeva, T. I. & Mochalov, I. V. (1978).

Kristallografiya, 23, 200-202.

Bochu, B., Deschizeaux, M.N. Jourbet, J.C., Collomb, A., Chenavas, J. & Marezio, M.,
(1979). J. Solid St. Chem., 29, 291-298.

Brixner, L.H. (1960). J. Inorg. Nucl. Chem., 14, 225-230.

Buttner, R. H. & Maslen, E. N., (1988). Acta Cryst., C44, 1707-1709.

Darlington, C.N.W. & Knight, K.S. (1999). Physica B, 266, 368-372.

Darlington, C.N.W. & Knight, K.S. (1999). Acta Cryst., B55, 24-30.

Demazeau, G., Parent, C., Pouchard, M. & Hagenmuller, P. (1972). Mat. Res. Bull., 913-
920.

Deschanvres, A., Raveau, B. & Tollemer, M.F. (1967). Bull. Chem. Soc. Fr., 4077-4078.

Dickens, P.G. & Powell, A. (1991). J. Mater. Chem., 1, 137-138.

García-Muñoz, J.L., Rodríguez-Carvajal, J., Lacorre, P. & Torrance, J.B. (1992). Phys.
Rev. B, 46, 4414-4425.

69
Glazer, A.M. (1972). Acta Cryst., B28, 3384-3392.

Glazer, A.M. (1975). Acta Cryst., A31, 756-762.

Hinatsu, Y. & Itoh, M. (1996). J. Solid St. Chem., 132, 337-341.

Hohnstedt, C. & Meyer, G. (1993). Z. Anorg. Allg. Chem., 619, 1374-1358.

Howard, C.J., & Kennedy, B.J. (1999). J. Phys.: Condens. Matter, 11, 3229-3236.

Howard, C.J., Kennedy, B.J. & Chakoumakos, B.C. (2000). J. Phys.: Condens. Matter,
12, 349-365.

Howard, C.J., Knight, K.S., Kennedy, B.J. & Kisi, E.H. (2000). J. Phys.: Condens.
Matter, 12, L677-L683.

Hutton, J. & Nelmes, R.J. (1981). J. Phys. C: Solid State Phys., 14, 1713-1736.

Hutton, J., Nelmes, R.J. & Scheel (1981). Acta Cryst., A37, 916-920.

Ito, K., Tezuka, K. & Hinatsu, Y. (2001). J. Solid State Chem. 157, 173-179.

Kennedy, B.J., Howard, C. J. & Chakoumakos, B.C. (1999). Phys. Rev. B., 59, 4023-
4027.

Kennedy, B.J., Howard, C. J. & Chakoumakos, B.C. (1999). Phys. Rev. B., 60, 2972-
2975.

Kennedy, B.J., Prodjosantoso, A.K. & Howard, C. (1999). J. Phys. Condens. Matter, 11,
6319-6327.

Kijima, N., Tanaka, K. & Marumo, F. (1981). Acta Cryst., B37, 545-548.

Kijima, N., Tanaka, K. & Marumo, F. (1983). Acta Cryst., B39, 557-561.

Kobayashi, H., Nagata, M., Kanno, R. & Kawamoto, Y. (1994). Mater. Res. Bull., 29,
1271-1280.

Labeau, M., Bochu, B., Jourbet, J.C. & Chenavas, J. (1980). J. Solid St. Chem. 33, 257-
261.

Leinenweber, K. & Parise, J. (1995). J. Solid St. Chem., 114, 277-281.

Liu, G., Zhao, X. & Eick, H.A. (1992). J. Alloys Compd., 187, 145-156.

70
Marezio, M., Dernier, P.D., Chenavas, J. & Jourbet, J.C. (1973). J. Solid St. Chem., 6,
16-20.

Marezio, M., Remeka, J. P. & Dernier, P.D. (1970). Acta Cryst., B26, 2008-2022.

Marti,W., Fischer,P., Altorfer, F., Scheel, H.J. & Tadin, M. (1994). J. Phys.:Condens.
Matter ,6, 127-135.

Mehta, A., Berliner, R.& Smith, R.W. (1997). J. Solid. St. Chem., 130, 192-198.

Miyata, N., Tanaka, K. & Marumo, F. (1983). Acta Cryst., B39, 561-564.

Odenthal, R.-H. & Hoppe, R. (1971). Monatsh. Chem., 102, 1340-1350.

Ozaki,Y., Ghedira, M., Chenavas, J., Jourbert, J.C. & Marezio, M. (1977). Acta Cryst.,
B33, 3615-3617.

Park, J.-H. & Parise, J.B. (1997). Mater. Res. Bull., 32, 1617-1624.

Pickardt, J., Schendler, T. & Kolm, M. (1988). Z. Anorg. Allg. Chem., 560, 153-157.

Ramirez, A.P., Subramanian, M.A., Gardel, M., Blumberg, G., Li, D., Vogt, T. &
Shapiro, S.M. (2000). Solid State Commun., 115, 217-220.

Rey, M.J., Dehaudt, PH., Jourbet, J.C., Lambert-Andron, B., Cyrot, M. & Cyrot-
Lackmann, F. (1990). J. Solid State. Chem., 86, 101-108.

Ross, N.L. (1998). Phys. Chem. Miner., 25, 597-602.

Roth, R.S. (1957). J. Res. Natl. Bur. Stand., 58, 75-88.

Sarkozy, R.F., Moeller, C.W. & Chamberland, B.L. (1974). J. Solid State Chem., 9, 242-
246.

Sasaki, S., Prewitt, C.T.& Bass, J.D. (1987). Acta Cryst., C43, 1668-1674.

Shaplygin, I. S. & Lazarev, V. B. (1976). Zh. Neorg. Khim., 21, 2326-2330.

Shimizu, Y., Syono, Y. & Akimoto, S. (1970). High Temp. High Press., 2, 113-120.

Shishido, T., Nakagawa, S., Yoshikawa, A., Horiuchi, H., Tanaka, M., Sasaki, T. &
Fukuda, T.(1995). Nippon Kagaku Kaishi, 9, 697-702.

Smith, A.J. & Welch, A.J.E. (1960). Acta Cryst., 13, 653-656.

71
Subramanian, M.A, Li, D., Duan, N., Reisner, B.A., & Sleight, A.W. (2000). J. Solid
State Chem. 151, 323-325.

Svensson, G. & Werner, P.-E. (1990). Mat. Res. Bull., 25, 9-14.

Thornton, G., Tofield, B.C. & Hewat, A.W. (1986). J. Solid State Chem. 61, 301-307.

Vasylechko, L., Matkovskii, A., Savytskii, D., Suchocki, A. & Wallrafen, F. (1999). J.
Alloys Compd., 291, 57-65.

Vegas, A., Vallet-Regí, M., González-Calbet, J.M. & Alario-Franco, M.A. (1986). Acta
Cryst. B42, 167-172.

Woodward, P.M. (1997a). Acta Cryst., B53, 32-43.

Woodward, P.M. (1997b). Acta Cryst., B53, 44-66.

Yamamoto, A., Khasanova, N. R., Izumi, F., Wu, X.-J., Kamiyama, T., Torii, S. &
Tajima, S. (1999). Chem. Mater. 11, 747-753.

Zhao, C., Feng, S., Chao, Z., Shi, C., Xu, R. & Ni, J. (1996). Chem. Commun., 14, 1641-
1642.

Zhurova, E. A., Zavodnik, V. E. & Tsirel'son, V. G. (1995). Kristallografiya, 40, 816-

823.

72
 CHAPTER 4

STRUCTURE PREDICTION OF ORDERED AND DISORDERED

MULTIPLE OCTAHEDRAL CATION PEROVSKITES

4.1 Introduction to Multiple Octahedral Cation Perovskites.

A significant number of AMX3 compositions containing a single octahedral cation

have been experimentally investigated. Introduction of multiple octahedral cations greatly

increases the number of possible perovskite compositions, making it one of the most

frequently encountered in solid-state inorganic chemistry. An equal (1:1) ratio

substitution of two different octahedral cations is often reported in crystal structures

leading to an A2MM′X6 stoichiometry. Perovskites with A2MM′X6 stoichiometry have

either an ordered or disordered arrangement on the octahedral cation sublattice.

In a disordered octahedral cation arrangement, the space group symmetry is the

same as in the single octahedral cation case. The site occupancy is 1/2 for each of the two

octahedral cations. A disordered arrangement is typically observed for cations of similar

ionic size and charge. Crystallographic techniques including x-ray and neutron powder

diffraction examine only the long range order of the octahedral cations. Short range

ordering of the octahedral cations is also possible.

Ordering of the octahedral cations is enhanced by a large difference in size and

charge in the octahedral cations. Ordering is incomplete in some structures and most

73
often encountered when the charge difference between M and M′ is two or less and with

octahedral ions of similar ionic size (Anderson et al., 1993). An ordered cation

arrangement induces a change in the size of the unit cell and symmetry of the space group

compared to single octahedral cation tilt systems (Woodward, 1997a). Space groups

resulting from a combination of octahedral tilting and cation ordering are shown in table

4.1. A larger number of degrees of freedom are available in cation ordered structures.

Obtaining a good starting model for structural refinement is important (and non-trivial) in

the commonly observed tilt system a-b+a- (space group = P21/n) with three anions located

on the general Wyckoff site.

Three types of ordering of the octahedral cations are 1) rock-salt type (111 planes

of like atoms), 2) layered (001 planes of like atoms) or 3) chains (110 planes of like

atoms). Rock-salt type ordering leads to a structure where each M-cation has six M′-

cations neighbors. Considering only the octahedral cation sublattice, it resembles the

NaCl “rock-salt” structure type. Layered ordering leads to a structure with alternating

layers of MO2 and M′O2, thus each M cation has four “equatorial” M neighbors and two

“axial” M′ neighbors. Chain ordering in the perovskite structure leads to a single axis

chain of MO units surrounded by four chains of M′O. Diagrams of rock salt, layered, and

chain octahedral cation ordering are shown in figure 4.1. Rock-salt type ordering is the

predominant type of octahedral cation ordering observed for 1:1 ratios of octahedral

cations. SPuDS version 2.0 includes the capability to calculate optimized structures for

perovskites with a rock-salt ordered octahedral cation distribution, in high symmetry tilt

systems a0a0a0, a0a0c-, a-a-a-, a0b-b-, a-b+a-, and the multiple A-site tilt system a+a+a+.

74
 Glazer Tilt Space Degrees of Wyckoff Sites
Frequency
System* Group Freedom A M X
Group A - High Symmetry Tilt Systems
0 0 0
a a a (23) Fm 3 m 2 1 2 1 48
- - -
a a a (14) R3 7 1 2 1 2
a0a0c- (22) I4/m 5 1 2 2 12
a0a0c+ (21) P4/mnc 5 1 2 2 0
a0b-b- (20) I2/m 11 1 2 2 8
a-a-b+ (10) P21/n 16 1 2 3 70
Group B - Multiple A-Site Tilt Systems
a+a+a+ (3) Pn 3 4 2 2 1 4
a0b-c+ (17) C2/c 15 2 2 3 0
a0b+b+ (16) P42/nnc 7 3 2 2 0
a+a+c- (5) P42/n 12 3 2 3 0
Group C - Transitional/Low Symmetry Tilt Systems
- - -
a a c (13) F1 30 4 8 12 0
0 - -
a b c (19) I1 30 4 8 12 3
- - -
a b c (12) F1 30 4 8 12 0
+ - -
a b c (8) P1 30 4 8 12 1
a+b+c+ (1) Pnnn 12 4 2 3 0

Table 4.1 : The fifteen tilt systems, space groups, degrees of freedom, number of
independent Wyckoff sites and number of observed structures reported for 1:1 rock-salt
ordered perovskites with the restriction that not more than two layers show independent
tilting. *The number in parentheses corresponds to the numbering of the tilt systems
originally adopted by Glazer (1972).

75
Figure 4.1 : The crystal structures of rock-salt (top) and layered (middle) and chain
(bottom) octahedral cation ordering.

76
 Structural reports are widely available in the literature in which to evaluate the

accuracy of the SPuDS software. Figure 4.2 shows the distribution of tilt systems and is

meant to show the general trends in the structurally well characterized ordered

perovskites with a 1:1 rock-salt ordered octahedral cations. The total number of

structurally characterized perovskites with a single octahedral cation is 188, while that of

1:1 ordered perovskites is 148. The total number of A2MM′X6 perovskites is bound to

exceed the number of AMX3 perovskites at some point. Additional structures are

constantly being reported in the literature. Crystal structures with an assigned space

group and lattice parameters reported were not included in figure 4.2 unless the atomic

positions were also listed. The actual distribution number of compounds is expected to be

higher; however, structure refinement of the cation ordered structures is more difficult

than for the single octahedral cation compositions.

77
 a0a0a0
a0a0c-
Other
a0b-b-
a-a-a-

a-b-c-

a-b+a-

Figure 4.2 : Distribution of tilt systems with a 1:1 rock salt ordered octahedral cation
arrangement.

The six tilt systems for which SPuDS calculates structures are evaluated by

comparing to experimental crystal structures reported in the literature in the following

sections. If no structural examples are reported in the literature, hypothetical

compositions are explored and compositions with a low GII are presented.

4.2 a0a0a0 (Fm 3 m).

A diagram of untilted (a0a0a0) A2MM′O6 perovskite with a rock-salt type ordered

octahedral cation is shown in figure 4.1 (top). The structure has Fm 3 m symmetry and

there are two degrees of freedom in the structure, the lattice parameter and the x

fractional coordinate of the anion. Numerous examples are available in the literature in

which to model the untilted octahedral cation ordered cubic perovskites.

78
 Five compositions with accurate structure refinements (Woodward, 1997b),

Sr2AlNbO6, Sr2AlTaO6, Ba2ScTaO6, Ba2InTaO6 and Ba2ScBiO6, were chosen for

modeling. Two calculation schemes are available for the ordered untilted system. First is

the standard method in which the M-X bond distances are fixed by optimizing the M-X

bond distances to have a length in which the bond valence sum is equal to the oxidation

state. The fixed M-X bond distance method is not accurate in the prediction of the lattice

parameters for structures with a tolerance factor larger than unity. Oversized A-site

cations stretch the M-X bonds slightly, resulting in a larger lattice parameter due to

octahedral expansion. Stretching of the octahedra is observed in the single octahedral

cation perovskite BaTiO3 (Sleight et al., 1975) and a similar structural effect is expected

to be observed in multiple octahedral cation provskites. An option was added to the

SPuDS software to allow the M-X bond distance to vary. The optimized structure is taken

as the one in which the GII is minimized. The lattice parameter is initially calculated with

optimized M-X distances. The fractional coordinate is held constant and the lattice

parameter optimized by obtaining a minimized GII. A comparison of selected

experimental and SPuDS predicted structures for the cubic 1:1 ordered perovskites is

shown in table 4.2.

79
 Literature
Compound Tol. Factor O(24e) x a (Å)
Sr2AlNbO6 1.020 0.245 7.7858
Sr2AlTaO6 1.018 0.246 7.7866
Ba2ScTaO6 1.019 0.256 8.2200
Ba2InTaO6 1.006 0.258 8.2814
Ba2ScBiO6 0.985 0.246 8.3660

SPuDS Fixed M-X SPuDS GII Optimized M-X

Compound O(24e) x a (Å) GII O(24e) x a (Å) GII
Sr2AlNbO6 0.243 7.7098 0.164 0.245 7.7788 0.106
Sr2AlTaO6 0.243 7.7278 0.144 0.244 7.7878 0.094
Ba2ScTaO6 0.257 8.1858 0.160 0.258 8.2538 0.103
Ba2InTaO6 0.260 8.2918 0.043 0.261 8.3078 0.029
Ba2ScBiO6 0.249 8.4658 0.119 0.248 8.4178 0.085

Table 4.2 : Comparison of free position and lattice parameters for fixed M-X and GII
optimized calculation methods for a0a0a0 (Fm 3 m) perovskites.

A measure of the accuracy of the structure prediction is the percent lattice error.

The percent lattice error is defined as 100*(apred - aexp)/aexp where apred is the lattice

parameter from SPuDS calculation and aexp is from the experimental lattice parameter

determined from a diffraction experiment. Lattice error for the two calculation methods

are shown in figure 4.3. The fixed M-X distance calculation method shows a trend toward

an increasing negative percent lattice error as the tolerance factor increases. Fixed M-X

distance calculations for the larger tolerance factor compositions are more negative

because of the underestimation of the lattice parameter. The source of underestimation

for the lattice parameter is the slight stretching of the octahedra by the oversized A-

cations. A better modeling of the lattice parameter was obtained by implementation of the

80
calculation routine to optimize the GII of the structure by allowing the M-X bond

distances to vary. The error of the GII optimized structure is smaller than the fixed M-X

and is typically less than 0.5% for Fm 3 m perovskites.

81
 1.05
Optimize GII - Vary M-X

1.04
Fixed M-X

1.03

Tolerance Factor
1.02
1.01
1.00
0.99
0.98
1.5%

1.0%

0.5%

0.0%

-0.5%

-1.0%

-1.5%

-2.0%

% lattice error

Figure 4.3 : Lattice error for fixed M-X and GII from the SPuDS optimized structures
versus tolerance factor for tilt system a0a0a0 (sp. grp. Fm 3 m).
82
4.3 a0a0c- (I4/m)

Cation ordering of a 1:1 ratio of two different cations combined with single axis

(a0a0c-) antiphase tilting results in the I4/m space group. Known compounds and the

tolerance factors are shown in the table 4.3. The source of the experimental crystal

structures are Sr2CoMoO6 (Viola et al., 2002), Sr2CrTaO6, Sr2GaSbO6, Sr2GaTaO6,

Sr2WMgO6, Sr2VTaO6 (Woodward, 1997b), Ba2FeWO6 (Azad et al., 1995), Sr2NiWO6

(Iwanaga et al., 2000) and A2CuMO6 (A = Sr, Ba; M = W, Te) (Iwanaga et al., 1999).

A view of the octahedral tilting of the most tilted composition, Sr2WMgO6, is

shown in figure 4.4. The cation positions are A (0, ½, ¼), M (0, 0, 0), M′ (½, ½, 0), X1

(0, 0, z) and X2 (x, y, 0). Literature and SPuDS calculation results for the lattice

parameters and free fractional positions are shown along with the percent error in lattice

parameters and fractional coordinate difference in table 4.4.

Symmetric M′-site Jahn-Teller M′-site

Formula τ Formula τ
Ba2FeWO6 1.028 Ba2CuTeO6 1.042
Sr2CrTaO6 0.991 Ba2CuWO6 1.042
Sr2GaTaO6 0.990 Sr2CuTeO6 0.983
Sr2NiWO6 0.989 Sr2CuWO6 0.983
Sr2VTaO6 0.986
Sr2GaSbO6 0.984
Sr2CoMoO6 0.982
Sr2WMgO6 0.979

Table 4.3: Experimentally observed I4/m perovskite tolerance factors for a symmetric
and Jahn-Teller distorted M′ site.

83
 Figure 4.4 : Octahedral tilting in the experimental crystal structure of I4/m
Sr2WMgO6. The bottom view is looking down the c-axis and illustrates the out of phase
tilting.

84
 Experimental
Formula τ Lattice a Lattice c O(4e) z O(8h) x O(8h) y
Ba2WFeO6 1.028 5.755 8.125 0.237 0.235 0.241
Sr2CrTaO6 0.991 5.574 7.882 0.250 0.265 0.235
Sr2GaTaO6 0.990 5.578 7.896 0.235 0.282 0.229
Sr2NiWO6 0.989 5.559 7.918 0.255 0.289 0.227
Sr2VTaO6 0.986 5.598 7.941 0.250 0.274 0.226
Sr2GaSbO6 0.984 5.548 7.888 0.254 0.279 0.217
Sr2CoMoO6 0.982 5.565 7.948 0.259 0.290 0.230
Sr2WMgO6 0.979 5.579 7.938 0.237 0.271 0.211

SPuDS
Formula Lattice a Lattice c O(4e) z O(8h) x O(8h) y
Ba2WFeO6 5.738 8.115 0.236 0.236 0.236
Sr2CrTaO6 5.583 7.936 0.250 0.275 0.224
Sr2GaTaO6 5.587 7.948 0.250 0.277 0.223
Sr2NiWO6 5.588 7.955 0.259 0.288 0.230
Sr2VTaO6 5.595 7.974 0.251 0.282 0.219
Sr2GaSbO6 5.600 7.992 0.249 0.282 0.215
Sr2CoMoO6 5.603 8.011 0.262 0.299 0.225
Sr2WMgO6 5.583 7.936 0.239 0.278 0.199

Compound % Lattice % Lattice O(4e) z O(8h) x O(8h) y F.

Error a Error c F. Diff. F. Diff. Diff.
Ba2WFeO6 -0.29 -0.12 -0.001 0.001 -0.005
Sr2CrTaO6 0.17 0.69 0.000 0.010 -0.011
Sr2GaTaO6 0.15 0.65 0.015 -0.005 -0.006
Sr2NiWO6 0.55 0.53 0.004 -0.001 0.003
Sr2VTaO6 -0.06 0.41 0.001 0.008 -0.007
Sr2GaSbO6 0.94 1.31 -0.005 0.003 -0.002
Sr2CoMoO6 0.69 0.79 0.003 0.010 -0.005
Sr2WMgO6 -0.29 -0.12 -0.001 0.001 -0.005

Table 4.4 : Experimental (top), SPuDS predicted (middle) lattice parameters and free
fractional coordinates for I4/m perovskites. The cation at the origin is shown in bold and
listed first in the formula. The lattice error and anion fractional coordinate difference is
shown (bottom).

85
 Lattice error is calculated according to the equation 100 *(lattice predicted -

lattice experiment)/(lattice experiment). SPuDS calculations typically overestimate the

lattice parameters as shown in figure 4.5. The c-axis is typically over predicted by a

larger amount than the a-axis. Structures remain more pseudocubic [c/(√2a) = 1] in the

lattice parameters compared to calculations as shown in the c/[√2a] ratio versus tolerance

factor trend shown in figure 4.6. Nonetheless, the c/(√2a) ratio does increase with

decreasing tolerance factor as predicted. There seems to be a consistent trend showing a

small contraction of the M-O/M′-O bonds parallel to the c-axis, which is another

mechanism for increasing the valence of the A-cation.

86
 1.03
% Lattice Error a

% Lattice Error c

1.02
1.01

Tolerance Factor
1.00
0.99
0.98
0.97
1.40

1.20

1.00

0.80

0.60

0.40

0.20

0.00

-0.20

-0.40

% Lattice Error

Figure 4.5 : Percent lattice error for the optimized SPuDS structure of a0a0c- (I4/m)
perovskites. Triangles are % lattice error in (a) and squares are % lattice error in (c).

87
 1.03
Experimental M(III)-M'(V)
Experimental M(II)-M'(VI)

1.02
SPuDS

1.01
Tolerance Factor
1.00
0.99
0.98
0.97
1.014

1.012

1.010

1.008

1.006

1.004

1.002

1.000

0.998

0.996

c/(√2a)

Figure 4.6 : Lattice parameter ratio [c/(√2a)] versus tolerance factor for I4/m perovskites.
Open diamonds represent SPuDS predicted structures, filled triangles Sr2M3+M5+O6, and
filled squares Sr2M2+M6+O6 experimental crystal structures.

88
 A comparison of compounds crystallizing in the I4/m space group (except

Ba2WFeO6, τ = 1.028 & GII = 0.238) for the GII in each of the calculated rock salt

ordered tilt systems is shown in figure 4.7. Calculations indicate there are several tilt

systems with a lower GII than a0a0c-. The tilt system a-a-a- will typically have the lowest

GII, but is less often observed than a-b+a-. Once the GII falls below a certain point, the

GII does not appear to be an accurate predictor.

The calculated GII does not take into account the coordination preference of the

A-site. A structure with a few very short A-X bonds may have a low GII, however the A-

site prefers symmetric coordination and an even distribution of bond lengths. The a0a0c-

tilting appears to be stable over a fairly narrow range of GII (<0.033) and tolerance factor

(0.982 < τ < 0.991) appears to be requirements to crystallize in the a0a0c- with the

exception of Ba2WFeO6 with τ = 1.028 and GII = 0.238.

89
 0.25
#3 a+a+a+ (Pn3)
#10 a-b+a- (P21/n)
#14 a-a-a- (R-3)
#20 a0b-b- (I2/m)
#22 a0a0c- (I4/m)
0.2 #23 a0a0a0 (Fm3m)

0.15

experimentally observed I4/m perovskites.

GII

90
0.1

0.05

0
0.965 0.970 0.975 0.980 0.985 0.990 0.995
Tolerance Factor

Figure 4.7: GII versus tolerance factor for SPuDS optimized crystal structures of selected
4.4 a-a-b+ (P21/n)

The octahedral cation ordered equivalent of the GdFeO3 structure (space group =

Pnma, tilt system = a-a-b+) has space group P21/n and is the most prevalent ordered tilt

system. The octahedral cations are located on fixed positions M(2c) (0, ½, 0) and M′(2d)

(½, 0, 0). Perovskites evaluated are listed with the reference containing the crystal

structure in table 4.5. Three example compounds: La2NiRuO6 (Seinen et al., 1987),

La2PtCoO6 (Ouchetto et al., 1997), and La2LiSbO6 (López et al., 1992) were chosen to

illustrate the accuracy of the SPuDS structure calculations. Detailed experimental and

SPuDS predicted structural data for P21/n perovskites is shown in table 4.6. Comparison

of literature and SPuDS predicted A-site fractional positions for a-b+a- P21/n perovskites

is shown in figure 4.8. A comparison of literature and SPuDS predicted octahedral tilt

angles for a-b+a- P21/n perovskites is shown in figure 4.9. Calculation results are very

accurate and can be useful for starting models in structure refinement of P21/n

perovskites.

91
 Compound Reference
Ba2USrO6 Groen & Ijdo, 1987
Ca2TaAlO6 Woodward, 1997b
Ca2TaCrO6 Woodward, 1997b
Ca2TaFeO6 Woodward, 1997b
Ca2TaGaO6 Woodward, 1997b
Ca2TaScO6 Woodward, 1997b
Ca2UCaO6 van Duivenboden & Ijdo, 1986
La2CoIrO6 Seinen et al., 1987
La2HfMgO6 Woodward, 1997b
La2LiSbO6 Lopez et al., 1997
La2MgIrO6 Currie et al., 1987
La2NiIrO6 Currie et al., 1987
La2NiRuO6 Seinen et al., 1987
La2PtCoO6 Ouchetto et al., 1997
La2PtMgO6 Ouchetto et al., 1997
La2PtNiO6 Ouchetto et al., 1997
La2PtZnO6 Ouchetto et al., 1997
La2ZnIrO6 Seinen et al., 1987
Na2ZrTeO6 Woodward, 1997b
Nd2PtCoO6 Ouchetto et al., 1997
Nd2TiMgO6 Groen et al., 1987
Pr2PtCoO6 Ouchetto et al., 1997
Sr2UCaO6 Groen & Ijdo, 1987
Sr2YRuO6 Battle & Macklin, 1984

Table 4.5 : Perovskites structurally evaluated in space group P21/n (tilt system a-a-b+).
The octahedral cation at Wyckoff position 2c is listed first in the formula and the cation
at Wyckoff position 2d listed second.

92
Compound a (Å) b (Å) c (Å) β (°)
La2NiRuO6 Literature 5.5688 5.5984 7.8764 90.18
SPuDS 5.5099 5.6507 7.8875 90.01
La2PtCoO6 Literature 5.5722 5.6459 7.8906 90.00
SPuDS 5.5512 5.7434 7.9787 89.99
La2LiSbO6 Literature 5.6226 5.7199 7.9689 89.80
SPuDS 5.5548 5.7550 7.9888 90.02
Compound A(x) A(y) A(z) O(4e)x O(4e)y O(4e)z
La2NiRuO6 Literature 0.508 0.539 0.251 0.219 0.196 -0.045
SPuDS 0.513 0.540 0.250 0.215 0.207 -0.040
La2PtCoO6 Literature 0.514 0.545 0.249 0.204 0.216 -0.058
SPuDS 0.519 0.555 0.251 0.198 0.211 -0.047
La2LiSbO6 Literature 0.510 0.544 0.254 0.213 0.195 -0.042
SPuDS 0.520 0.557 0.249 0.210 0.198 -0.048
Compound O(4e)'x O(4e)'y O(4e)'z O(4e)"x O(4e)"y O(4e)"z
La2NiRuO6 Literature 0.294 0.715 -0.038 0.422 -0.014 0.250
SPuDS 0.295 0.717 -0.040 0.420 -0.012 0.245
La2PtCoO6 Literature 0.291 0.697 -0.045 0.403 -0.028 0.255
SPuDS 0.292 0.702 -0.047 0.406 -0.017 0.255
La2LiSbO6 Literature 0.301 0.718 -0.046 0.413 -0.022 0.239
SPuDS 0.306 0.713 -0.048 0.404 -0.017 0.242

Table 4.6 : Comparison of literature and SPuDS predicted lattice parameters and free
fractional positions for La2NiRuO6, La2PtCoO6 and La2LiSbO6 perovskites (space group
P21/n) with tilt system a-b+a-.

93
 0.55

0.54

0.53

A (x)
0.52

0.51

0.50

0.49
0.84 0.86 0.88 0.90 0.92 0.94
Tolerance Factor

0.62
0.61
0.60
0.59
0.58
0.57
A (y)

0.56
0.55
0.54
0.53
0.52
0.51
0.84 0.86 0.88 0.90 0.92 0.94
Tolerance Factor

0.260

0.255

0.250
A (z)

0.245

0.240

0.235
0.84 0.86 0.88 0.90 0.92 0.94
Tolerance Factor

Figure 4.8 : Comparison of literature and SPuDS predicted A-site fractional positions for
a-b+a- P21/n perovskites. Open squares are SPuDS predicted values and filled squares are
experimental tilt angles.

94
 0.94
0.92 0.90
Tolerance Factor
0.88
0.86
0.84
29.0

27.0

25.0

23.0

21.0

19.0

17.0

15.0

13.0

Tilt Angle (°)

Figure 4.9 : Comparison of literature and SPuDS predicted octahedral tilt angle for a-b+a-
P21/n perovskites. Open squares are SPuDS predicted values and filled squares are
experimental tilt angles.

95
4.5 a0b-b- (I2/m)

Octahedral cation ordering and octahedral tilting about the [110] axis of the

aristotype perovskite occurs in the a0b-b- space group. Compositions that crystallize in

I2/m often contain Bi on the octahedral site. Indication of this tilt system in the x-ray

diffraction data is by an absence of possible primitive reflections (i.e. those with hkl, k +

k + l ≠ 2n). Known compositions are Ba2MBiO6 (M = Bi, Sb) (Thornton & Jacobson,

1978), BaLaCoRuO6 (Kim & Battle, 1995) and Ba2MBiO6 (M = Nd, Pr, Tb) (Harrison et

al., 1995). Several of the compositions display incomplete ordering of the chemical and

charge order of the octahedral cations. The Ba2TbBiO6 composition exhibits cation

ordering with intermediate oxidation states for Tb (3+/4+) and Bi (3+/5+) (Harrison et

al., 1995). The Ba2PrBiO6 displays octahedral site mixing of approximately 25% and

mixed oxidation states for Pr (3+/4+) and Bi (3+/5+) (Harrison et al., 1995).

A comparison of the lattice parameters and free fractional coordinates for

perovskites crystallizing in I2/m is shown in table 4.7. The trivalent cation is located at

the origin (0, 0, 0) and Bi at (½, ½, 0). SPuDS predicts a β = 90° in each calculation.

The octahedra tilting distortion is about the 110 axis of the octahedron and is

perpendicular to the c-axis. A distortion of the octahedra must occur to obtain a β less

than 90°.

96
 Compound a (Å) b (Å) c (Å) beta (°)
Ba2TbBiO6 Lit. 6.1104 6.0813 8.5922 89.97
SPuDS 6.0584 6.2451 8.5679 90.00
Ba2PrBiO6 Lit. 6.2011 6.1583 8.6968 89.92
SPuDS 6.1264 6.3950 8.6640 90.00
Ba2NdBiO6 Lit. 6.1776 6.1366 8.6686 89.80
SPuDS 6.1054 6.3483 8.6344 90.00

Compound A (x) A(z) O(4i) x O(4i) z O(8j) x O(8j) y O(8j) z

Ba2TbBiO6 Lit. 0.002 0.249 0.936 0.259 0.742 0.254 -0.033
SPuDS 0.011 0.252 0.912 0.260 0.740 0.259 -0.044
Ba2PrBiO6 Lit. 0.004 0.250 0.936 0.259 0.742 0.254 -0.033
SPuDS 0.017 0.254 0.894 0.266 0.734 0.265 -0.053
Ba2NdBiO6 Lit. 0.003 0.251 0.940 0.263 0.736 0.260 -0.032
SPuDS 0.015 0.253 0.899 0.264 0.736 0.263 -0.050

Table 4.7 : Experimental and SPuDS optimized lattice parameters and free fractional
coordinates for selected a0b-b- (I2/m) compositions.

4.6 a-a-a- (R 3 )

Compositions that crystallize in R 3 are Ba2Bi3+Bi5+O6 (Cox & Sleight, 1979) and

Ba2YbBiO6 (Harrison et al.,1995). Similar to most structures crystallizing in I2/m, both

contain Bi on the octahedral site. Typically the a-a-a- tilt system is found as an

intermediate phase at high temperatures. SPuDS free fractional coordinates for

Ba2BiYbO6 is shown in table 4.8. Atoms on fixed positions are Bi (0, 0, 0) , Yb (½, ½, ½)

and Ba (¼, ¼, ¼).

97
 Compound a (Å) alpha (°) O(x) O(y) O(z)
Ba2BiYbO6 Lit. 6.0252 60.037 -0.221 -0.269 0.242
SPuDS 6.1068 59.050 -0.184 -0.305 0.244

Table 4.8 : Comparison of experimental and SPuDS optimized lattice parameters and
free fractional coordinates for Ba2BiYbO6 crystallizing in titl system a-a-a- (R 3 ).

4.7 a+a+a+ (Pn 3 )

The single octahedral cation perovskite series CaCu3M4O12 (M=Ge, Mn, Ti, Ru)

crystallizes in the a+a+a+ (Im 3 ) tilt system are presented in section 3.7. Two distinct

cation environments are present on the A-site for CaCu3Ti4O12. Ordering of the A-site

cations is always exhibited for the a+a+a+ tilt system. This tilt system is of importance

because the compound CaCu3Ti4O12 (Subramanian et al., 2000) exhibits fascinating

dielectric behavior, as described in section 3.7. This tilt system with cation ordering on

the A and M-site simultaneously was of interest to see if it was possible to tune the

dielectric properties with octahedral cation order.

Synthesis of novel perovskite compositions is one goal of the applications of

SPuDS. Calculations using SPuDS was attempted to predict possible novel and stable

compositions in space group Pn 3 , which has no known compounds that crystallize in the

a+a+a+ tilt system. In addition, there are no reports in the literature for a perovskite with

order on the A-site and rock salt order on the M-site simultaneously. The hypothetical

dual A-site and octahedral cation ordered structure is based on the octahedral tilting of

CaCu3Ti4O12, combined with a rock salt ordered arrangement of the octahedral cations.

Figure 4.10 shows the cation coordination environments in the crystal structure.

98
Figure 4.10 : Pn 3 structure with ordering of two A-cation sites and two octahedral
cations. White spheres are Ca, black spheres in a square planar environment are Cu, light
and dark octahedra are MO6 and M′O6 in a formula CaCu3M2M′2O12.

99
 Prediction of compounds that would favorably crystallize in this tilt system begins

by identifying potential octahedral cations. An increased ionic size difference between

the two octahedral cations leads to an increased probability of ordering due to the

difference in preferred M-X bond lengths. An increased charge difference leads to an

increased probability of cation ordering compared to isovalent ions due to the energetic

penalty of placing two highly charged ions as nearest octahedral cation neighbors. An

examination of likely candidates was undertaken using a 1:3 Ca:Cu A-site cation and

combinations of 3+:5+ cations for the octahedral cations. Reduction in the possible

number of ion combinations is obtained by considering the ease of synthetic oxidation

state attainment and the coordination preference. Selection of pentavalent ions (6) and

trivalent ions (9) lead to 54 (1:1) octahedral cation combinations. Compositions with a

1:1 ratio of trivalent (Al3+, Co3+, Cr3+, Ga3+, Ni3+, Fe3+, Rh3+, Sc3+, In3+) to pentavalent

(V5+, Ru5+, Nb5+, Ir5+, Ta5+, Sb5+) octahedral cations were chosen for structure prediction.

Octahedral M-O bond distances are shown in the table 4.9 for the selected ions.

100
 M3+ d(M-O) (Å) M5+ d(M-O) (Å)
Al 1.876 V 1.870
Co 1.893 Ru 1.967
Cr 1.980 Nb 1.978
Ga 1.986 Ir 1.983
Ni 1.996 Ta 1.987
Fe 2.015 Sb 2.009
Rh 2.049
Sc 2.105
In 2.158

Table 4.9 : Six coordinate M-O distances for trivalent and pentavalent ions used in the
hypothetical Pn 3 SPuDS structure calculations.

Hypothetical structures were generated using the ion combinations with SPuDS in

space group Pn 3 . SPuDS calculations use the assumption of a bond valence sum of the

M-site equal to the ideal oxidation state. Pentavalent ion radii decreases in size in the

order Sb5+> Ta5+> Ir5+> Nb5+> Ru5+>V5+, therefore the tolerance factors of compounds

containing these ions on the M-site follows the order Sb5+< Ta5+< Ir5+< Nb5+< Ru5+<V5+.

Bond valence sums of Ca, Cu, and O for CaCuM3+2M5+2O12 are shown in the figure 4.11

(top) for the trivalent ions, with the pentavalent ion following the tolerance factor order

listed above. Bond valence sums of the single M-site ions (Ge, Mn, Ti, Ru) compositions

are shown for comparison. Bond valence sum of Ca is farthest from ideal due to small

contribution to overall formula unit and thus a small contribution to the GII. Bond

valence sums of Ca, Cu, and O are shown simultaneously versus tolerance factor and the

GII versus tolerance factor are shown in the figure 4.11 (bottom) to show the ideal

tolerance factor (tilt angle) for Pn 3 perovskites. The octahedral tilting plays a very

101
significant role in this tilt system. The ideal tolerance factor is ~0.87. The structure

becomes less stable due to the overbonding of the Cu2+ and underbonding of the Ca2+ for

τ < 0.87, while the opposite is true for τ > 0.87.

102
 2.5 Al(III) Co(III) Ca
2.4 Cr(III) Fe(III)
Ga(III) In(III)
2.3 Ni(III) Rh(III)
Bond Valence Sum

2.2 Sc(III) Single M

2.1 O
2.0

1.9
Cu
1.8

1.7

1.6

1.5
0.82 0.83 0.84 0.85 0.86 0.87 0.88 0.89 0.90 0.91
Tolerance Factor

0.12
Al(III) Co(III)
0.10 Cr(III) Fe(III)
Ga(III) In(III)
0.08 Ni(III) Rh(III)
Sc(III) Single M
GII

0.06

0.04

0.02

0.00
0.82 0.83 0.84 0.85 0.86 0.87 0.88 0.89 0.90 0.91
Tolerance Factor

Figure 4.11 : Bond valence sums of Ca, Cu, O of the SPuDS optimized structures shown
simultaneously versus tolerance factor (top). The GII versus tolerance factor of the
SPuDS optimized structures for Pn 3 perovskites (bottom).

103
 The GII of CaCu3Ti4O12 occurs at the intersection of the bond valence sums of

Ca2+, Cu2+, and O2-. At this point the bond valence sum of each of the ions is very close

to the ideal value. Synthesis of CaCu3Ti4O12 occurs at ambient pressure, while synthesis

of the CaCu3M4O12 (M = Ge, Mn, Ru) requires elevated pressure. Synthesis pressure

required for successfully synthesis was found to have a qualitatively direct dependence

on the GII as described in section 3.7. The less stable structure with a larger GII value

required a higher synthesis pressure to obtain the desired phase. The hypothetical

compositions CaCu3Cr2Ru2O12, CaCu3Cr2Sb2O12, CaCu3Ga2Sb2O12 have a GII most

likely to be synthesized under ambient pressure. The predicted GII and difference in

octahedron size is shown in table 4.10.

104
 Compound GII (v.u.) d(M(III)-O)-d(M(V)-O) (Å)
CaCu3Cr2Ru2O12 0.007 -0.013
CaCu3Rh2V2O12 0.009 -0.180
CaCu3Ga2Ru2O12 0.010 -0.019
CaCu3Cr2Nb2O12 0.013 -0.002
CaCu3Cr2Ir2O12 0.016 0.003
CaCu3Ga2Nb2O12 0.016 -0.008
CaCu3Co2Sb2O12 0.018 0.116
CaCu3Cr2Ta2O12 0.018 0.007
CaCu3Ga2Ir2O12 0.019 -0.003
CaCu3Ni2Ru2O12 0.021 -0.039
CaCu3Ga2Ta2O12 0.021 0.001
CaCu3Sc2V2O12 0.022 -0.236
CaCu3Fe2Ru2O12 0.026 -0.048
CaCu3Ni2Nb2O12 0.027 -0.028
CaCu3Al2Sb2O12 0.027 0.133
CaCu3Fe2V2O12 0.029 -0.146
CaCu3Ni2Ir2O12 0.029 -0.023
CaCu3Cr2Sb2O12 0.030 0.029

Table 4.10 : GII and octahedral cation size difference for low GII hypothetical Pn 3
compositions.

The global instability index (GII) versus tolerance factor for the SPuDS calculated

tilt systems is shown in the figure 4.12 and indicates the stability of the compositions in

the a+a+a+ tilt system in comparison with the GII obtained for other tilt systems. The large

difference in GII indicates a large driving force to adopt the a+a+a+ tilt system, over

competing single phase perovskite structures. GII values for A′A3M4O12 compositions

stabilized by high-pressure high-temperature synthesis, with a GII ~ 0.09 and less, are

shown in table 3.4. Prediction of the GII for ordered compositions resulted in many

compositions with a GII below 0.09, indicating numerous potentially stable

105
compositions. Combinations of octahedral ions that are known to undergo ordering in

other perovskite compositions are considered to be favorable choices for ordering in

Pn 3 . Hypothetical compositions should have a reasonably low GII to have a reasonable

chance of successful synthesis. A difficulty in synthesis of these perovskites is the

competition of other simple and stable phases, including the binary and ternary oxides.

Four compositions were chosen for high-pressure high-temperature synthesis, described

in section 7.7, based (a) the driving force for cation ordering (b) a reasonably low GII.

and (c) a lack of competing high pressure phases.

The composition with smallest valence for Ca is the most likely candidate for

replacement of Ca with larger Sr. The composition SrCu3In2Sb2O12 has a GII of 0.055

v.u., which is unlikely to be stable at ambient pressure. Stabilization may occur at higher

pressures and an interesting study would be to see if a relatively large ion (Sr2+) can be

stabilized on the icosahedral site.

106
 0.862
CaCu3Ga2Nb2O12

0.861
0.860
CaCu3Ga2Ta2O12

0.859
Tolerance Factor
0.858
0.857
CaCu3Cr2Sb2O12

0.856
0.855
CaCu3Ga2Sb2O12

0.854
0.853
0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

GII

Figure 4.12 : GII versus tolerance factor for four perovskites in the Pn 3 space group.
Tilt systems are filled circles are a0a0a0, cross a0b-b-, square a0a0c-, triangle a-a-a-, plus a-
b+a-, and diamond a+a+a+.

107
4.7.1 CASTEP Geometry Optimization of Pn 3

A test of the accuracy of the SPuDS calculation routine was done by performing

higher level calculations with Materials Studio CASTEP. Geometry optimization (CGA,

quality=ultrafine, BFGS optimization) using a high level calculation is suited for the Pn 3

space group due to its cubic structure and many atoms on fixed positions. The unit cell is

relatively large (~10 Å) and geometry optimization requires several days computation

time on a standard desktop Intel based PC with a Red Hat Linux operating system.

SPuDS atomic positions were used as starting positions in the Materials Studio

calculations. Results are shown in table 4.11 and comparison to the experimental lattice

parameters are made in section 7.7. The structure did not undergo a significant change in

the higher level calculations, suggesting the SPuDS calculations are quite accurate.

CASTEP Materials Studio

a (Å) Ox Oy Oz
CaCu3Ga2Nb2O12 7.5076 0.2504 0.4250 0.5547
CaCu3Ga2Sb2O12 7.4966 0.2520 0.4265 0.5538
CaCu3Ga2Ta2O12 7.5464 0.2479 0.4236 0.5556
SPuDS
a (Å) Ox Oy Oz
CaCu3Ga2Nb2O12 7.4682 0.2506 0.4296 0.5549
CaCu3Ga2Sb2O12 7.5064 0.2484 0.4278 0.5560
CaCu3Ga2Ta2O12 7.4794 0.2499 0.4291 0.5552

Table 4.11 : SPuDS optimized and geometry optimization in CASTEP Materials Studio
optimized lattice parameter and free fractional coordinate comparison of Pn 3
compositions.

108
4.8 Conclusions

The software program SPuDS has been developed for predicting the structures of

perovskite compounds including multiple octahedral cations. Analysis of the results

indicates an accurate modeling of known perovskites. Prediction of novel compositions

was performed for the a+a+a+ tilt system (Pn 3 ), which has no reported examples. Dual

A-cation and rock salt ordering has never been observed, but was predicted to be stable

according to GII values obtained from SPuDS calculations. Optimization of the amount

of octahedral tilting is necessary in order to satisfy the coordination preferences of the

icosahedral and square planar A-site cations. The Ca-O bonds of the icosahedral site (Ca)

shorten less rapidly than the Cu-O bonds with increased octahedral tilting. A delicate

balance exists where the coordination of each A-site is satisfied, which is the near the

tolerance factor of CaCu3Ti4O12.

SPuDS calculated structure GII values are useful in determining the stability of

the composition in any particular tilt system. Comparing the GII for the untilted and tilted

compositions, a large GII implies the structure is unstable, and perhaps is likely to

undergo a distortion such as octahedral tilting to obtain a more suitable coordination

environment for each of the ions. Accuracy of cubic structures with a large GII solved

with low-resolution data need to be regarded with caution. Structures reported in the

literature with a large GII that were refined from low resolution x-ray diffraction data

may need to be reexamined with higher resolution diffraction data to check the original

solution correctness. A careful inspection of structures reported in cubic disordered

109
Pm 3 m or ordered Fm 3 m would likely result in the determination that some of the

structures undergo octahedral tilting and have lower symmetry structures. SPuDS

provides a useful and reliable starting model for Rietveld refinements, which simplifies

the structure solution process for compositions with many degrees of freedom.

4.9 References

Anderson, M.T., Greenwood, K.B., Taylor, G.A. & Poeppelmeier, K.R. (1993). Prog.
Solid St. Chem. 22, 197-233.

Azad, A.K., Eriksson, S.-G., Mellergård, A., Ivanov, S.A., Eriksen, J. & Rundlöf, H.
(2002). Mat. Res. Bull. 37, 1797-1813.

Battle, P.D. & Macklin, W. (1984). J.Solid State Chem. 52, 138-145.

Cox, D.E. & Sleight, A.W. (1979). Acta Cryst. B35, 1-10.

Currie, R.C., J.F., Vente, Frikkee, E. & Ijdo, D.J.W. (1995). J. Solid State Chem. 116,
199-204.

Glazer, A.M. (1972). Acta Cryst. B28, 3384-3392.

Groen, W.A. & Ijdo, D.J.W. (1987). Acta Cryst. C43, 1033-1036.

Groen, W.A., van Berkel, F.P.F.V. & Ijdo, D.J.W. (1986). Acta Cryst. C42, 1472-1475.

Iwanaga, D., Inaguma, Y. & Itoh, M. (1999). J. Solid State Chem. 147, 291-295.

Kim, S.H. & Battle, P.D. (1995). J. Solid State Chem. 114, 174-183.

López, M.L., Veiga, M.L., Rodriguez-Carvajal, J., Fernandez, F., Jerez, A. & Pico, C.
(1992). Mat. Res. Bull. 27, 647-654.

Lufaso, M.W., & Woodward, P.M. (2001). Acta Cryst. B57, 725-738.

Marchand, R., Laurent, Y., Guyader, J., L'Haridon, P. & Verdier, P. (1991). J. Eur.
Ceram. Soc., 8, 197-213.

110
Ouchetto, K., Archaimbault, F., Choisnet, J. & Et-Tabirou, M. (1997). Mat. Chem. Phys.,
51, 117-124.

Seinen, P.A., van Berkel, F.P.F., Groen, W.A. & Ijdo, D.J.W. (1987) Mat. Res. Bull., 22,
535-542.

Sleight, A.W., Gillson, J.L. & Bierstedt, P.E. (1975). Solid St. Commun. 17, 27-28.

Thorton, G. & Jacobson, A. (1976). Mat. Res. Bull. 11, 837-842.

van Duivenboden, H.C., Ijdo, D.J.W. (1986). Acta Cryst. C42, 523-525.

Viola, M.C., Martinez-Lope, M.J., Alonso, J.A., Velasco, P., Martinez, J.L., Pedregosa,
J.C., Carbonio, R.E.& Fernandez-Diaz, M.T. (2002). Chem. Mater. 14, 812-818.

Woodward, P.M. (1997a). Acta Cryst. B53, 32-43.

Woodward, P.M. Ph.D. Dissertation, Oregon State University, Corvallis, OR 1997b.

111
 CHAPTER 5

SYNTHESIS AND CRYSTAL STRUCTURES OF Sr2MnNbO6,

Sr2MnSbO6, Ca2MnNbO6, Ca2MnRuO6 AND Ca2MnSbO6

5.1 Introduction

Substitutions on the M-site of a perovskite can be achieved and multiple

octahedral cation ordering types (e.g. layered M-site, rock-salt M-site, multiple A-site

environments) have been observed. Cation disorder on a single site is often observed in

perovskites with multiple ions similar in size, charge, and coordination preference.

Bonding requirements of the A-site cations in perovskites drives the occurrence of

octahedral tilting. The octahedral tilting occurs without a significant disturbance of the

coordination of the M-site ions. Octahedral bond distances and angles remain symmetric

after tilting, unless there is an electronic driving force present to cause a distortion. In

octahedral coordination, the 3d transition metal cation Mn3+ has the electron

configuration (t2g)3(eg)1. Electronically this is an unfavorable situation and the Jahn-Teller

(J-T) theorem states a distortion from octahedral symmetry should occur (Kanamori,

1960). Lowering the symmetry of the octahedron removes the electronic degeneracy. A

distortion in Mn-O bond lengths results in an energetic stabilization; however, significant

distortions of the O-Mn-O bond angles are not typically observed.

112
 Crystal structures of La2GaMnO6 and Nd2GaMnO6 indicate disorder of Ga and

Mn on the M-site (Cussen et al., 2001). Interestingly, a coherent Jahn-Teller distortion is

present in Nd2GaMnO6, which is effectively absent in La2GaMnO6. Relatively subtle

differences in octahedral tilting are proposed to be the source of the unexpectedly distinct

crystal chemistry of these two compounds. Investigation of the perovskite series

A2MnMO6 (A =Ca, Sr; M = Ru, Sb, Ta, Nb), in which various degrees of octahedral

tilting, cation ordering, and Jahn-Teller distortions are possible may provide additional

insight into the influence of octahedral tilting and cation ordering on the development of

cooperative Jahn-Teller distortions.

Determination of accurate crystal structures, particularly the oxygen positions, is

necessary for a detailed understanding of the static and dynamic orbital ordering. Neutron

diffraction is a useful technique for accurate determination of oxygen positions. Accurate

coordination of the octahedral cations is then extracted from the neutron powder

diffraction data using the Rietveld method. The coordination environment of the M-site

can probed for any cooperative ordering of the Jahn-Teller distortions. Orientation of the

long M-O bond is an indirect way of determining the orientation of the occupied eg

orbitals.

A search of the literature for accurate A2MnMO6 crystal structures was

unsuccessful for Sr2MnNbO6, Sr2MnSbO6, Ca2MnNbO6, Ca2MnRuO6 and Ca2MnSbO6.

Samples of these compounds were synthesized and the structures determined from

neutron powder diffraction data in order to better understand the interactions between

cation ordering, Jahn-Teller distortions, and octahedral tilting.

113
5.2 Experimental

Polycrystalline samples of Sr2MnNbO6, Sr2MnSbO6, Ca2MnNbO6, Ca2MnRuO6

and Ca2MnSbO6 were synthesized by standard solid state reaction techniques.

Stoichiometric amounts of CaCO3 (Mallincrokdt), SrCO3 (MCB), Mn2O3 (Aldrich 99%),

Nb2O5 (Alfa 99.9%), Sb2O3 (Cerac 99%), and RuO2 (Cerac) were accurately weighed out

and intimately mixed by grinding in an agate mortar and pestle. Reactants were fired in

air as a powder in high form alumina crucibles. The crucibles were covered to minimize

potential cation volatility at the elevated synthesis temperature. Each sample was

reground in an agate mortar and pestle between each firing. Sample purity was checked

after each heating cycle by x-ray diffraction. Heating cycles continued until the sample

attained equilibrium and no change in the weakest peaks observed in laboratory x-ray

diffraction.

Reactants were initially heated to 900-950 °C for 10-12 hours to decompose the

carbonates. Final annealing temperature, balanced heating and cooling rates were

Sr2MnNbO6 (1415 °C, 5 °C min-1), Sr2MnSbO6 (1375 °C, 3 °C min-1), Ca2MnMO6 (M =

Sb, Nb, Ru) (1385 °C, 3 °C min-1). Each polycrystalline sample was dark black in color.

After the final annealing cycle x-ray diffraction data was collected for each sample on a

Bruker D8 Advance diffractometer equipped with a Ge (111) monochromator, incident

Soller slits, and PSD detector. Diffraction patterns for phase purity analysis were

collected at ambient temperature with monochromatic Cu Kα1 radiation. Data collection

proceeded with a 0.0144° 2θ step size in the angular range 15° <2θ < 100° with a 2

second count time. Sample purity and single phase verification was checked by

114
performing Rietveld (1969) refinements of the x-ray diffraction data using Topas

refinement software (Bruker AXS, 2000).

Accurate fractional coordinates of the oxygen atoms are required to obtain

detailed information about the octahedral cation coordination in perovskites.

Improvement in x-ray diffraction techniques have resulted in improved sensitivity to the

weak peaks associated with octahedral tilting in perovskites. However, accurate

refinement of x-ray diffraction data is still challenging in perovskites due to the relatively

weak scattering power of oxygen and high degree of pseudosymmetry that is invariably

present. In order to obtain accurate information on the oxygen positions, neutron powder

diffraction data were collected at the Australian Nuclear Science and Technology

Organisation (ANSTO) under ambient conditions. The wavelength of the medium

resolution data collection was 1.32137 Å in the angular range 2.154° < 2θ< 120°.

Samples (10 g each, except 7 g Ca2MnRuO6) were analyzed in vanadium containers.

Rietveld structural refinement of the neutron data was carried out using GSAS (Larson,

1990) program suite. A brief history of each compound and the refinement methodology

and results from this study are presented in the following sections.

5.3 Sr2MnMO6 (M=Nb, Sb)

A tolerance factor less than unity normally correlates with the observance of

octahedral tilting. Tolerance factors (τ) for the compositions synthesized in this study are

calculated assuming a disordered arrangement of M-site cations in the software program

SPuDS (Lufaso & Woodward, 2001). Quaternary perovskites with a tolerance factor near

0.98 often crystallize in the a0a0c- tilt system (as discussed in the previous chapter), while

115
perovskites with a lower tolerance factor commonly crystallize in the a-b+a- tilt system

(Lufaso & Woodward, 2001).

Compound τ
Sr2MnNbO6 0.987
Sr2MnSbO6 0.977
Ca2MnNbO6 0.933
Ca2MnRuO6 0.936
Ca2MnSbO6 0.924

Table 5.1 : Tolerance factors for the five A2MnMO6 compounds synthesized in this
study.

The six coordinate ionic radii for Mn3+(0.79 Å) and Sb5+(0.74 Å), Nb5+(0.78 Å),

Ta5+(0.78 Å), and Ru5+(0.71 Å) are similar. Compounds with similar compositions that

might be expected to be isostructural to Sr2MnMO6 (M=Sb, Nb) based on ionic sizes of

the cations are Sr2MnMO6 (M=Ru, Ta).The compositions Sr2MnMO6 (M=Ta, Nb) were

first reported with a cubic unit cell with a = 3.986 Å and 3.959 Å for M = Ta, Nb

respectively (Kupriyanov & Filil'ev, 1963). Sr2MnSbO6 was reported in the tetragonal

crystal system (a = 7.86 Å, c = 8.08 Å) with a disordered arrangement of M-cations

(Blasse, 1965). A magnetic study of Sr2MnSbO6 indicated Mn was present in only one

oxidation state, however did not include detailed structural information (Brach et al.,

1982). Refinement of Sr2MnSbO6 in the obscure (for perovskites) space group I4mm with

a final Rwp value of 9.1% was reported (Politova et al., 1990). Sr2MnSbO6 was reported

to be a semiconductor ferroelectric with M-cation ordering of Mn and Sb in space group

116
I4mm or I4/mmm (Foster et al., 1997). The latter study reported an unusual distortion of

the MnO6 octahedron in Sr2MnSbO6 with four short bonds, one medium bond and one

long bond (Foster et al., 1997). The authors suggest the assignment of the oxygen

positions may differ significantly from the value reported, which is due to the difficulty

in refining accurate fractional coordinates for the oxygen atoms. No electronic driving

force is present to provide a 4+1+1 type of octahedral distortion. Refinement of x-ray

diffraction data of Sr2MnSbO6 in the disordered M-site space group I4/mcm indicated

some Sb vacancy in a previous analysis (Woodward, 1997). Rietveld refinements of

neutron powder diffraction data of Sr2MnTaO6 (Woodward, 1997) and Sr2MnRuO6

(Woodward, 2002) indicated I4/mcm symmetry with disordered arrangements of Mn-Ta

and Mn-Ru on the M-site. Sr2MnMO6 (M=Ta, Ru) features single axis anti-phase

octahedral tilting, a0a0c- in Glazer’s notation (1972).

Indexing of the Sr2MnMO6 (M=Sb, Nb) diffraction patterns resulted in a

tetragonal unit cell with dimensions approximately equal to a= √2ap and c = 2ap, where ap

is the ideal primitive cubic perovskite lattice parameter. The reflection conditions

observed in the diffraction pattern are in agreement with those expected for space group

I4/mcm (and others). Disorder on a single site is more probable for cations that share a

similar charge and ionic size. Diffraction peaks associated with M-site cation order were

not observed in the x-ray or neutron powder diffraction data collected for any of the

compositions.

Refinement of the x-ray and neutron data in space group I4mm and I4/mmm using

fully ordered M-site atomic coordinates from Politova et al. (1990) as a starting model

117
resulted in an incorrect fitting of peak intensity for the (011) peak. A disordered I4mm

model was attempted and the order of the M-site refined. Based on previous structural

reports (Woodward, 1997) the structural model based on I4/mcm space group was also

implemented. The I4/mcm structural model of Sr2MnMO6 (M= Sb, Nb) was obtained

from calculations of disordered M-site with symmetric M-O bond distances using

SPuDS. The atomic positions are A (0, ½, ¼), M (0, 0, 0), O(4a) (0, 0, ¼ ), O(8h) (x, x +

½, 0). SPuDS predicted lattice parameters and free fractional coordinates are shown in

table 5.2. Fractional coordinates for all of the atoms from the SPuDS structure prediction

was combined with the lattice parameters obtained from the peak indexing for use as the

starting model.

a (Å) c (Å) O(8h) x

Sr2MnNbO6 5.5943 7.9727 0.2189
Sr2MnSbO6 5.6159 8.0484 0.2089

Table 5.2 : SPuDS optimized free fractional coordinates and lattice parameters for
Sr2MnMO6 (M=Sb, Nb).

Rietveld refinement was executed using Topas (Bruker AXS, 2000). Lattice

parameters, 2θ zero point error, scale factor and the background were refined. Profile

parameters (i.e. U, V, W) were refined next. Structural parameters were refined including

the atomic coordinates and isotropic thermal vibration parameters. Experimental (circles)

and calculated (line) diffraction patterns for the four structural models implemented in the

Rietveld refinement of x-ray diffraction data of Sr2MnSbO6 is shown in figure 5.1. The

difference pattern is shown below the experimental data and calculated pattern.

118
I4mm with a fully ordered cation distribution (Rwp = 13.658)
125,000 Sr2MnSbO6-I4mm 100.00 %

100,000

75,000

50,000

25,000

-25,000

-50,000

15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

I4/mmm with a fully ordered cation distribution (Rwp = 10.796)

125,000 Sr2MnSbO6-I4/mmm 100.00 %

100,000

75,000

50,000

25,000

-25,000

15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

I4mm with a disordered cation distribution (Rwp = 5.059)

125,000 Sr2MnSbO6-I4mm-disorder 100.00 %

100,000

75,000

50,000

25,000

15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

I4/mcm with a disordered cation distribution (Rwp = 5.051)

125,000 Structure 100.00 %

100,000

75,000

50,000

25,000

15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

Figure 5.1 : Rietveld refinement of x-ray diffraction data of 4 different structural models
for Sr2MnSbO6. At the bottom of the figure the difference plot (Iobs – Icalc) is shown.

119
 The Rwp for the x-ray diffraction data refinement in the disordered I4mm (Rwp =

5.058) and I4/mcm (Rwp = 5.051) are similar in value. The refinement of x-ray diffraction

data of Sr2MnSbO6 in space group I4/mcm resulted in positive isotropic thermal

parameters (Beq) for each atom Beq(Sr) = 1.40(5), Beq(Mn/Sb) = 2.08(6), Beq(O(4a)) =

7.5(3), and Beq(O(8h)) = 5.5(2).

The two most promising models from the refinement of the x-ray diffraction data,

the disordered I4/mcm and disordered I4mm, were implemented in the Rietveld

refinement of the neutron powder diffraction data using GSAS (Larson, 1990). Lattice

parameters, 2θ zero point error, scale factor and the background were refined. Profile

parameters (i.e. U, V, W) were refined next. Structural parameters were refined including

the atomic coordinates and isotropic thermal vibration parameters. The (112) and (200)

reflections at approximately 27° 2θ were not fit well in the I4mm model. Refinement of

the neutron powder diffraction in the disordered M-site I4mm (Rwp = 5.49) resulted in a

significantly worse fit than in the I4/mcm model (Rwp = 3.09). Experimental neutron

diffraction patterns (×) and calculated (line) diffraction patterns for the disordered I4mm

and I4/mcm structural models of Sr2MnSbO6 are shown in figure 5.2.

120
Figure 5.2 : Comparison of two structural models I4mm (Rwp = 5.49) (top) and I4/mcm
(Rwp = 3.09) (bottom) for Rietveld refinement of neutron powder diffraction data for
Sr2MnSbO6. At the bottom of the figure the difference plot (Iobs – Icalc) is shown.

121
 Refinement of neutron powder diffraction data of Sr2MnSbO6 and Sr2MnNbO6

indicated that I4/mcm symmetry with a disordered arrangement of M-site cations is the

model that provides the best fit to the experimental data. Therefore, the space group

I4/mcm was used in the Rietveld refinement of the neutron powder diffraction data. The

results are shown in the tables 5.3 and 5.4.

Space Group I4/mcm (#140) Sr2MnNbO6 Sr2MnSbO6

a (Å) 5.6119(6) 5.5553(2)
c (Å) 7.927(1) 8.0548(3)
Volume (Å3) 249.65(2) 248.58(2)
Rp (profile) 2.91 2.48
Rwp (weighted profile) 3.73 3.09
Zero Error (°) -0.0822 -0.0610
Reduced χ2 4.006 2.425
Total Refined Variables 24 24
Minimum 2θ (°) 6 10
Maximum 2θ (°) 137 135
Number of reflections 112 111
Profile Function pseudo Voigt pseudo Voigt
(GSAS type 1) (GSAS type 1)
Gaussian U 504(47) 432(39)
Gaussian V -383(65) -546(55)
Gaussian W 598(20) 603(17)
Background Function Chebyschev Chebyschev
polynomial polynomial
(12 terms) (12 terms)

Table 5.3 : Rietveld refinement details and results for Sr2MnSbO6 and Sr2MnNbO6.

122
Atom Site x y z Occupancy Ux100 Å2
Sr 4b 0 ½ ¼ 1 1.63(6)
Mn 4c 0 0 0 0.5 0.4(1)
Nb 4c 0 0 0 0.5 0.4(1)
O1 4a 0 0 ¼ 1 2.4(2)
O2 8h 0.2248(4) 0.7248(4) 0 1 1.28(6)

Atom Site x y z Occupancy Ux100 Å2

Sr 4b 0 ½ ¼ 1 1.41(8)
Mn 4c 0 0 0 0.5 -0.5(2)
Sb 4c 0 0 0 0.5 -0.5(2)
O1 4a 0 0 ¼ 1 2.1(1)
O2 8h 0.2200(3) 0.7200(3) 0 1 1.40(5)

Table 5.4 : Atomic coordinates, site occupancy and isotropic thermal parameters for
Sr2MnNbO6 (top) and Sr2MnSbO6 (bottom).

Selected bond distances and bond valence sums are shown in table 5.5. Bond

valence sums were calculated with SPuDS using bond distances obtained from the

refinements from neutron data. Bond valence sums (BVS) for Sr2+ are very close to the

ideal values, whereas BVS for the M-site have a larger difference from the ideal values.

Refinements indicated cation disorder, thus the bond distances are an average of all of the

coordination environments around that crystallographic site. In such instances it is

expected that the bond valence sums will deviate somewhat from their ideal values.

123
 Sr2MnSbO6 Sr2MnNbO6
Bond Distances (Å) M-O(4a) 2.0137(1) ×2 1.9818(4) ×2
M-O(8h) 1.9781(3) ×4 1.9937(4) ×4
Sr-O(4a) 2.7777(1) ×4 2.8060(3) ×4
Sr-O(8h) 2.653(1) ×4 2.670(2) ×4
Sr-O(8h) 2.982(2) ×4 2.946(2) ×4
Bond Valence Sums (v.u.) Sr2+ 2.00 1.95
Mn3+ 3.23 3.23
M5+ 5.28 4.85
O(4a)2- 2.00 2.00
O(8h)2- 2.13 1.99

Table 5.5 : Selected bond distances and valence sums from the refinements of neutron
powder diffraction for Sr2MnSbO6 and Sr2MnNbO6.

5.4 Ca2MnMO6 (M=Nb, Sb, Ru)

Ca2MnNbO6 was reported in space group Pbnm, a non standard setting of Pnma.

(Kruth et al., 1998). Indexing of the diffraction peaks indicated orthorhombic symmetry,

with a unit cell a = 21/2ac, b = 2ac, c = 21/2ac where ac is the typical (~4Å) lattice parameter

of a cubic perovskite. The structural model of Ca2MnMO6 (M= Sb, Nb, Ru) was obtained

from SPuDS calculations of a disordered M-site Pnma perovskites with symmetric M-O

bond distances. The lattice parameters and free fractional positions are shown in table

5.6.

124
 Ca2MnNbO6 Ca2MnRuO6 Ca2MnSbO6
a (Å) 5.5619 5.5472 5.6040
b (Å) 7.7512 7.7367 7.7922
c (Å) 5.4073 5.4008 5.4256
Ca(x) 0.5450 0.5425 0.5519
Ca(z) 0.5148 0.5139 0.5175
O1(x) -0.0137 -0.0130 -0.0157
O1(z) 0.4149 0.4171 0.4086
O2(x) 0.2902 0.2893 0.2928
O2(y) 0.0426 0.0414 0.0457
O2(z) 0.7074 0.7086 0.7043

Table 5.6 : SPuDS optimized lattice parameters and free fractional coordinates for
Ca2MnMO6 (M = Nb, Ru, Sb).

Rietveld refinement was performed using GSAS program suite. Lattice

parameters, 2θ zero point error, scale factor and the background were refined. Profile

parameters (i.e. U, V, W) were refined next. Structural parameters were refined including

the atomic coordinates (Ca first, followed by O) and isotropic thermal vibration

parameters. A representative fit with experimental (×) and calculated fit (line) to the

diffraction data is shown in the figure 5.3 for Ca2MnNbO6. Refinement details and results

are shown in the tables 5.7 and 5.8.

125
Figure 5.3 : Rietveld refinement of powder neutron diffraction data of Ca2MnNbO6. The
experimental points (x) and calculated (line) diffraction patterns and the difference
pattern. At the bottom of the figure the difference plot (Iobs – Icalc) is shown.

126
 Space Group Pnma (# 62) Ca2MnNbO6 Ca2MnRuO6 Ca2MnSbO6
a (Å) 5.5635(4) 5.4207(3) 5.5549(5)
b (Å) 7.6996(5) 7.5757(4) 7.6941(6)
c (Å) 5.4530(4) 5.3346(3) 5.4591(4)
Volume (Å3) 233.59(4) 219.07(2) 233.32(4)
Rp (profile) 2.14 2.42 2.35
Rwp (weighted profile) 2.60 2.92 2.85
Zero Error (°) -0.0675 -0.0507 -0.0673
Reduced χ2 2.804 2.374 1.990
Total Refined Variables 31 31 31
Minimum 2θ (°) 6 10 7
Maximum 2θ (°) 138 138 138
Number of reflections 449 415 447
Profile Function pseudo Voigt pseudo Voigt pseudo Voigt
(GSAS type 1) (GSAS type 1) (GSAS type 1)
Gaussian U 713(76) 578(40) 659(68)
Gaussian V -364(100) -609(64) -689(100)
Gaussian W 569(29) 624(22) 651(32)
Background Function Chebyschev Chebyschev Chebyschev
polynomial polynomial polynomial
(12 terms) (12 terms) (12 terms)

Table 5.7 : Rietveld refinement of neutron diffraction data details and results for
Ca2MnMO6 (M=Ru, Nb, Sb).

127
Atom Site x y z Occ. Ux100 Å2
Ca 4c 0.5459(6) ¼ 0.5100(9) 1 1.53(7)
Mn 4b ½ 0 0 0.5 0.0(1)
Nb 4b ½ 0 0 0.5 0.0(1)
O1 4c -0.0226(4) ¼ 0.4175(5) 1 1.03(6)
O2 8d 0.2947(3) 0.0425(2) 0.7039(3) 1 1.02(4)

Atom Site x y z Occ. Ux100 Å2

Ca 4c 0.5425(5) ¼ 0.5058(9) 1 0.69(5)
Mn 4b ½ 0 0 0.5 0.7(1)
Ru 4b ½ 0 0 0.5 0.7(1)
O1 4c -0.0181(4) ¼ 0.4208(5) 1 0.66(5)
O2 8d 0.2939(3) 0.0398(2) 0.7050(3) 1 0.46(3)

Atom Site x y z Occ. Ux100 Å2

Ca 4c 0.5469(9) ¼ 0.511(1) 1 1.58(9)
Mn 4b ½ 0 0 0.5 0.4(4)
Sb 4b ½ 0 0 0.5 0.4(4)
O1 4c -0.0257(6) ¼ 0.4135(7) 1 1.03(8)
O2 8d 0.2942(5) 0.0434(3) 0.7031(5) 1 1.16(6)

Table 5.8 : Atomic coordinates, occupancy and isotropic thermal parameters for
refinements of neutron powder diffraction data of Ca2MnMO6 (M = Nb, Ru, Sb).

128
 Selected bond distances and bond valence sums are shown in table 5.9. Bond

valence of the A-site cation includes only the first coordination sphere (ten A-O bonds),

which includes only cation to oxygen bonds with a distance smaller than the shortest A-

site to M-site bond distance. SPuDS calculated BVS using the bond distances obtained

from the refinements of neutron data assumed formal oxidations states Ca(2+), Mn(3+)

and M(5+). Bond valence sums for Ca2+ are very close to the ideal values for M=Nb, Sb,

while an increase in the BVS of Ca2+ is observed for M = Ru. Bond valence sums for the

M-site have a larger difference from the ideal values. Refinements indicated cation

disorder, thus the bond distances are an average of all of the coordination environments

around that site.

Ca2MnNbO6 Ca2MnRuO6 Ca2MnSbO6

Bond Distances (Å) M-O(4c) 1.9807(7)×2 1.9429(6)×2 1.9858(9)
M-O(8d) 2.005(2)×2 1.953(2)×2 2.011(3)
M-O(8d) 2.008(2)×2 1.956(2)×2 2.003(3)
Bond Distances (Å) Ca-O(4c) 3.202(4) 3.073(4) 3.225(6)
Ca-O(4c) 2.453(4) 2.424(4) 2.433(6)
Ca-O(4c) 2.362(5) 2.299(5) 2.352(7)
Ca-O(4c) 3.145(5) 3.076(5) 3.168(7)
Bond Distances (Å) Ca-O(8d) 2.371(4)×2 2.341(3)×2 2.366(5) ×2
Ca-O(8d) 2.687(3) ×2 2.621(3)×2 2.691(4) ×2
Ca-O(8d) 2.627(4)×2 2.603(4)×2 2.617(5) ×2
Bond Valence Sums Ca2+ 1.98 2.23 2.02
(v.u.) Mn3+ 3.16 3.58 3.14
M5+ 4.75 5.23 5.13
O(4c)2- 2.09 2.30 2.14
O(8d)2- 1.93 2.17 2.01

Table 5.9 : Selected bond distances and bond valence sums for Ca2MnMO6 (M = Nb, Ru,
Sb) from refinements of neutron powder diffraction data.

129
 Oxygen nonstoichiometry was observed in Ca2MnNbOγ (5.86 < γ < 6.00)

depending on post-reaction heat treatment at different oxygen pressures (Kruth et al.,

1998). High O2 pressures were observed to give γ = 6.00, while final heat treatments in

air at 1400 °C resulted in γ = 5.87. Lattice parameters showed little change in a and b,

with a small increase in c with decreasing γ. Linear interpolation and use of the lattice

parameter observed in the neutron data used in this study results in an estimated γ = 5.91

and x-ray data results in an estimated γ = 5.88 for the sample synthesized at 1385 °C. The

isotropic thermal vibration parameters from the neutron refinements were small (Uiso =

0.001) which may indicate some non-stoichiometry is present. The occupancy of the M-

site was refined, and resulted in a physically unreasonable value larger than the site

allows. Refinement of the O occupancies did not result in any significant deviation from

full site occupancy. A small degree of oxygen non-stoichiometry is possible, however,

Rietveld refinements of the x-ray and neutron diffraction data do not indicate the

presence of a significant deviation from the stoichiometric formulas. Determination of

the oxygen content in the samples by thermogravimetry or magnetic measurement was

not undertaken.

5.5 Conclusions

Samples of Sr2MnNbO6, Sr2MnSbO6, Ca2MnNbO6, Ca2MnRuO6 and Ca2MnSbO6

were synthesized and crystal structures were obtained from neutron powder diffraction

data using the Rietveld method. Diffraction peaks associated with M-site cation order

were not observed in either the X-ray or neutron powder diffraction data, consistent with

130
the absence of long range Mn/M cation order. Fractional positions of structures generated

in the software program SPuDS were used as starting positions in the Rietveld

refinements. Starting atomic positions agree well with the final refined fractional

coordinates. Octahedral tilting about a single axis (a0a0c-) occurs for Sr2MnMO6 (M=Nb,

Sb), whereas octahedral tilting about multiple axes (a-b+a-) occurs for Ca2MnMO6

(M=Nb, Sb, Ru). These findings are in-line with expectations based on the geometric

tolerance factors.

A negative isotropic thermal parameter was observed for the M-site in

Sr2MnSbO6 from refinements of the neutron diffraction data, while the isotropic thermal

parameter was positive from refinements of x-ray diffraction data. The neutron scattering

cross-sections (in units of fm) used in the GSAS refinements were 0.470 for Ca, 0.707 for

Sr, -0.373 for Mn, 0.557 for Sb, 0.705 for Nb and 0.703 for Ru. The disordered sites are

an average of the two neutron scattering sections, i.e. 0.092 Mn/Sb, 0.166 for Mn/Nb and

0.165 for Mn/Ru. The negative thermal parameters observed in the refinement of neutron

powder diffraction data for Sr2MnSbO6 may be due in part to the relatively weak

scattering of the disordered Mn/Sb distribution.

Bond valence sums of A-site cations agree closely with the ideal value, with a

slight overbonding in Ca2MnRuO6. Bond valence sums of M-site cations are complicated

by the two atoms disordered on a single crystallographic site. Diffraction techniques

show the average coordination environment over all of the crystallites in the sample. The

BVS of Mn3+ is greater than three in each case, indicating a slight compression of the

Mn-O bonds. The BVS for Nb5+ is less than the ideal value of 5 in both Sr2MnNbO6

131
(4.85 v.u.) and Ca2MnNbO6 (4.75 v.u.). The BVS for Sb5+ is greater than the ideal value

of 5 in both Sr2MnNbO6 (5.28 v.u.) and Ca2MnNbO6 (5.13 v.u.).

Structural data indicate a symmetric coordination environment for the M-site ions

in the Ca2MnMO6 (M=Sb, Nb, Ru) series. Nearly equivalent M-O bond distances are

observed for Sr2MnNbO6, while a slight Jahn-Teller distortion along the c-axis of I4/mcm

is observed in Sr2MnSbO6. Backgrounds with wavy shapes are characteristic of short-

range correlations (Palacin et. al, 1993) and were observed in the neutron powder

diffraction patterns. Implications of the coordination of the M-site and the state of the

Jahn-Teller distortion in this series are discussed in chapter 6.

5.6 References

Blasse, G. (1965). J. Inorg. Nucl. Chem. 27, 993-1003.

Brach, B.Y., Bobrysheva, N.P. & Chezhina, N.V. (1982). Zh. Neorg. Khim. 27, 2471-
2473.

Bruker AXS, 2000 - User Manual, Bruker AXS, Karlsruhe, Germany.

Blundell, S.J., Coldea, A.I. & Singleton, J. (2001). J. Am. Chem. Soc., 123, 1111-1122.

Foster, M.C., Nielson, R.M. & Abrahams, S.C. (1997). J. Appl. Phys. 82, 3076-3080.

Glazer, A.M. (1972). Acta Cryst. B28, 3384-3392.

Kruth, A., Tabuchi, M., Guth, U.& West, A.R. (1998). J. Mater. Chem. 8, 2515-2520.

Kupriyanov, M.F. & Filip'ev, V.S. (1963). Sov. Phys. Crystall. 8, 278-283.

Larson, A.C. & von-Dreele, R.B., General Structure Analysis System (GSAS), Los
Alamos National Laboratories, 1990.

Lufaso, M.W. & Woodward, P.M. (2001). Acta Cryst. B57, 725-738.

132
Paolasini, L., Caciuffo, R., Sollier ,A., Ghigna, P. & Altarelli, M. (2002) Phys. Rev. Lett.
88, 106403-1:106403-4.

Politova, E.D, Kaleva, G.M., Danilenko, I.N., Chuprakov, V.F., Ivanov, S.A. &
Venevtsev, Yu. N. (1990). Izvestieiia Akademii nauk SSSR. Neorganicheskie materialy
26, 2352-2356.

Rietveld, H.M. (1969). J. Appl. Cryst., 2, 65-71.

Woodward, P.M. Ph. D. Dissertation, Oregon State University, Corvallis, OR, 1997.

Woodward, P. M. et al. (2002). To be published.

133
 CHAPTER 6

JAHN-TELLER DISTORTIONS, CATION ORDERING AND

OCTAHEDRAL TILTING IN PEROVSKITES

6.1 Introduction
Perovskites containing Jahn-Teller ions (i.e. Cu2+, Mn3+) on the octahedral site

exhibit interesting physical properties including charge ordering, spin ordering and

cooperative Jahn-Teller distortions. In octahedral coordination, the 3d transition metal

cations Mn3+ and Cu2+ (high-spin) have electron configurations (t2g)3(eg)1 and (t2g)6(eg)3,

respectively. The Jahn-Teller (J-T) theorem states this is an electronically unfavorable

situation and a distortion of the geometry of the octahedron occurs to provide energetic

stabilization by removing the electronic degeneracy (Kanamori, 1960). A distortion

where two M-X bonds shorten and two M-X bonds lengthen, denoted as Q2, is observed

in orthorhombic multiple-axis (i.e. a-b+a-) tilted perovskites. A distortion of four bonds

shortening and two bonds lengthening, denoted as Q3, is observed in tetragonal single-

axis tilted (i.e. a0a0c-) oxide perovskites.

Experimentally the charge and spin configurations have been investigated by

neutron and electron diffraction techniques. Direct observance of the orbital ordering is

not easily accomplished using neutron or electron diffraction. Quantitative convergent

beam electron diffraction techniques have been used to identify the occupied dz2 orbitals

134
in LaMnO3 (Jiang et al., 2002). Resonant x-ray scattering uses incident photons tuned

near the K-absorption edge of the element undergoing orbital ordering to directly observe

orbital ordering (Murakami et al., 1998a; Paolasini, et. al, 2002). Dipole resonant x-ray

scattering studies confirmed the presence of orbital ordering of Mn 3d orbitals in

LaMnO3 (Murakami et al., 1998b) and that charge and orbital ordering occurred jointly

in Pr0.5Ca0.5MnO3 (Zimmermann et al., 2001). The orientation of the long M-X bonds in

the crystal is in agreement with orbital ordering results obtained from x-ray resonant

scattering experiments. Structures solved using diffraction techniques provide an indirect

measure of the orbital ordering. A quantitative measure of the magnitude of the Jahn-

Teller distortion is given by the octahedral distortion parameter (∆d) defined in the

equation below where d is the mean M-X bond distance and dn are the individual M-X

bond distances (Alonso et al., 2000).

2
1  [d − (d )] 
∆d =  ∑ n 
6  n=1, 6 (d ) 

In this chapter, octahedral tilting and the presence or absence of orbital and

chemical ordering is examined in perovskites containing J-T octahedral site cations.

Systematic comparisons are made to crystal structures with a symmetric octahedral ion of

approximately the same ionic size to separate the effects of octahedral tilting and the

cooperative Jahn-Teller distortions (CJTD). Crystal structures refined from neutron

powder diffraction data give accurate structural information on the oxygen positions and

were utilized when possible. The influence of cation ordering, octahedral tilting and

CJTD factors on the structural features in perovskites is described in the sections below.

135
6.2 Cooperative Jahn-Teller distortions

Cooperative Jahn-Teller distortions describe the orbital ordering arrangement and

occur to minimize the elastic energy (Goodenough, 1998). The CJTD in KCuF3 is

complex due to the numerous ways in which the crystals may form stacking errors. The

type and concentration of these stacking faults can influence the space group

determination, which makes assignment of a space group difficult. Consequently the

structure of KCuF3 has been reported with space group symmetry I4/mcm (Hutchings,

1969; Buttner et al., 1990), P4/mbm (Hutchings, 1969) and P212121 (Hidaka et al., 1998).

Crystallization of KCuF3 in space group I4/mcm with three distinct Cu-F bond lengths

and a CJTD of a long and short Cu-F bonds alternating in the ab plane was reported by

Buttner et al. (1990). Viewing perpendicular to the ab plane there are two observed

orientations of CuF2 layers. The structure is said to be d-type when the ordering of long

and short Cu-F bonds is in phase between neighboring layers, whereas, when the

orientations of the long and short Cu-F bonds are rotated 90° from one layer to the next,

the structure is designated as a-type (Hutchings et al., 1969). KCuF3 structure exists in

both forms (even in the same crystal), with the a-type favored.

Octahedral cation ordering arrangements designated as rock-salt, plane ordering

and chain ordering were presented in chapter 4. Structure diagrams for symmetric

octahedron structures are shown in figure 4.1. Ordered perovskites containing a J-T ion

and an octahedrally symmetric ion have a manifestation of the CJTD that varies with

changes in the cation ordering. Crystal structures of A2CuMO6 (A=Sr, Ba; M=W, Te)

136
were found to have I4/m space group symmetry (octahedral tilting a0a0c-) with rock salt

ordering of Cu2+ and M6+ (M = Te, W) (Iwanaga et al., 1999). The long Cu-O bonds are

all oriented parallel to the c-axis. La2CuSnO6 possesses an unusual layered arrangement

of Sn4+ and Cu2+ with a large J-T distortion about the latter ion (Anderson et al., 1991).

Charge ordering occurs near 50 K for SrNdMn2O6 resulting in a structure with chains of

Mn3+O6 and Mn4+O6 (Woodward et al., 1999). Figure 6.1 shows an a-type CJTD as well

as the most common CJTD for each type of cation ordering discussed above.

137
 KCuF3

Ba2CuWO6

La2CuSnO6

NdSrMn2O6

Figure 6.1 : Structure diagrams for a-type (KCuF3), single axis rock salt (Ba2CuWO6),
plane ordering (La2CuSnO6) and chain ordering (NdSrMn2O6) of octahedral cations.
Long M-X bonds are shown as thick and dark gray.

138
 Each ion has a bond valence sum (BVS) near the ideal oxidation state in

thermodynamically stable perovskite phases (Lufaso & Woodward, 2001). The need to

maintain a reasonable bond valence sum for the anion of the long M-X bond plays an

important role in determining the favored CJTD pattern. A large J-T distortion of the

octahedron causes of loss of bonding to the anion of the resulting long M-X bond.

Experimental crystal structures show that this anion typically forms a short bond to the

remaining M-site ion in its coordination environment in order to maintain its bond

valence sum. Two distinct anion sites are present for F in the KCuF3 structure type. One

anion F(8h) forms a short and long Cu-F bond while the other F(4b) forms two medium

length Cu-F bonds. Each environment results in a satisfactory bonding arrangement and

bond valence sum. Bond valence calculations of the structure of KCuF3 show the short

Cu-F(8h) bond (1.889Å) contributes 0.45 v.u. (valence units), the long Cu-F(8h) bond

(2.253 Å) contributes 0.17 v.u. and four K-F bonds (2.859 Å) contribute a total of 0.38

v.u. for a BVS = 1.00 v.u. for the F(8h). Two medium length Cu-F(4b) bonds (1.962Å)

contribute a total of 0.74 v.u. (valence units) and four K-F bonds (2.859 Å) contribute

0.32 v.u. for a BVS = 1.06 v.u. for the F(4b). Once the anion coordination is taken into

consideration it becomes apparent why the long bonds do not all line up in the same

direction.

Octahedral site cation ordered compounds have the additional requirements that

the coordination needs to be satisfied for both octahedral site ions. Bonding of two

different octahedral site cations in A2CuMO6 compositions occurs by orienting the long

139
Cu-O bonds along a single axis and forming a shorter bond to the neighboring octahedral

site cation. For example, O(4c) in Ba2CuWO6 has a long bond (2.418 Å) to Cu, a short

bond to W (1.900 Å) and four bonds to Ba (2.794 Å). The O(8d) has a much shorter

bond to Cu (2.011 Å), but slightly longer bonds to W (1.933 Å) and Ba (2×2.833 Å +

2×3.013 Å). In La2CuSnO6 an increased bond valence sum contribution to the O of the

long Cu-O bonds occurs by formation of a shorter Sn-O bond. The charge ordered

structure of NdSrMn2O6 has three distinct Mn sites. It should be noted that this structure

results from a low temperature, constrained refinement. Bond valence sums of the ions in

the four structures are listed in table 6.1.

KCuF3 Ba2CuWO6 La2CuSnO6 NdSrMn2O6

A-site 1.09 2.48 2.90, 3.00, 2.75, 2.93 2.59, 2.57, 2.63, 2.62
Oct. site (J-T) 1.98 1.90 2.05, 2.03 3.49, 3.49
Oct. site (Sym.) - 5.92 3.98, 3.98 4.05
X (long) 1.00 2.19 1.96, 1.91 2.14, 2.14
X 1.06 2.10 1.93, 1.95, 1.96, 2.04, 2.05, 2.20, 2.22,
1.92, 2.04, 2.18 2.18, 2.18

Table 6.1 : Bond valence sums (v.u) of ions in experimental crystal structures of KCuF3,
Ba2CuWO6, La2CuSnO6 and NdSrMn2O6.

6.3 Octahedral Tilting and Cation Ordering

Perovskites with a tolerance factor greater than unity often exhibit no octahedral

tilting. When the size of the A-cation is undersized for the corner-sharing MO6 octahedra,

an octahedral tilting distortion often occurs (Woodward, 1997b). Lowering the symmetry

140
from cubic (Pm 3 m) takes place and the various combinations of octahedral tilting and

resulting space groups were described previously (Glazer, 1972; Glazer, 1975;

Woodward, 1997a). Non-tilted cubic structures represent a small percentage of the total

perovskites reported in the literature and greater than half of the reported structures of

single octahedral site cation perovskites crystallize in a-b+a- tilt system with space group

Pnma (Lufaso & Woodward, 2001). Cooperative tilting of the undistorted octahedra

occurs to shorten A-X bonds and thereby improve the coordination environment and

bonding of the A-site cation.

Compositions with multiple octahedral site cations crystallize in either an ordered

or disordered arrangement. Ordering of the octahedral site cations in perovskites changes

the space group symmetry. The space group symmetries that result from the combination

of rock salt ordering and octahedral tilting have been reported previously (Woodward,

1997a). Factors favoring cation ordering on the octahedral site include a large difference

in size (greater than 0.1 Å) and/or ionic charge (greater than two) (Anderson et al., 1993).

Perovskites with a symmetric octahedral site cation have an octahedral coordination in

which each M-X bond is of similar length, whereas the octahedral site coordination

environments are varied for compounds containing a Jahn-Teller ion. The M-X bond

lengths, tolerance factors, ∆d, symmetries and tilting axes for a representative cross

section [LaMnO3 (Rodriguez-Carvajal et al., 1998), KCuF3 (Buttner et al., 1990),

Sr2CuWO6 (Itoh et al., 1999), and La2CuTiO6 (Palacin et al., 1993)] of perovskites

containing a CJTD are shown in table 6.2. Structural trends show a larger octahedral

distortion occurs for Cu2+ than Mn3+ containing perovskites.

141
 La2CuTiO6 LaMnO3 KCuF3 Sr2CuWO6
M-X (l) (Å) ×2 2.033 2.178 2.253 2.320
M-X (m) (Å) ×2 2.022 1.968 1.962 1.950
M-X (s) (Å) ×2 2.007 1.907 1.889 1.950
∆d (x10-4) 0.27 33.1 59.7 70.8
Symmetry Pnma Pnma I4/mcm I4/m
Tilt System a-b+a- a-b+a- a0a0c- a0a0c-

Table 6.2 : M-X bond lengths, tolerance factors, ∆d, symmetries and tilting axes for a
representative cross section of perovskites containing a formal Jahn-Teller ion
(La2CuTiO6, LaMnO3, KCuF3 and Sr2CuWO6)

6.4 Influence of octahedral tilting in AMX3 systems.

The compositions ACuF3 (A = Na, K, Rb) were refined in I4/mcm and exhibits an

a-type CJTD. No octahedral tilting occurs for ACuF3 (A=K, Rb) and the structure retains

F-Cu-F angles of 90º (Rb - Kaiser et al., 1990; K - Buttner et al., 1990), whereas NaCuF3

(τ = 0.918) exhibits significant octahedral tilting. NaCuF3 has four similar coordination

environments for Cu with strongly distorted average F-Cu-F angles (91º, 94º and 105º)

(Kaiser et al., 1990). Three distinct Cu-F bond distances (average Cu-F bond distances

presented for NaCuF3) are shown for ACuF3 (A = Na, K, Rb) in figure 6.2. The three

distinct Cu-F bond distances are similar for A = Na, K; however, the larger Rb (τ =

1.089) causes a stretching of the octahedron and the longest bond (weakest) undergoes

the largest lengthening.

142
 ∆dCu

120.0

100.0

80.0

60.0

40.0

20.0

0.0
1.10
Rb

1.05
K

Tolerance Factor
1.00
0.95
Na

0.90
2.40

2.30

2.20

2.10

2.00

1.90

1.80

Cu-F Bond Distance (Å)

Figure 6.2 : Cu-F distances vs. tolerance factor for the ACuF3 series (squares). The ∆dCu
is represented by open circles. The symbols are in order of increasing tolerance factor (A
= Na, K, Rb).

143
 The well studied series AMnO3 (A=La, Pr, Nd, Dy, Tb, Ho) crystallize in tilt

system a-b+a- with Pnma symmetry based on results from Rietveld refinements of neutron

powder diffraction data (Rodriguez-Carvajal et al., 1998; Alonso et al., 2000). All space

group settings were converted to the standard Pnma (or Pbnm in tables 6.4 and 6.5) to

facilitate comparison between compositions. A CJTD is observed crystallographically via

an alternating arrangement of long-short bonds in the ac plane. The ordered arrangement

of the long bonds is shown looking down the b-axis of LaMnO3 in figure 6.3.

144
Figure 6.3 : View down the b-axis (Pnma) of the crystal structure of LaMnO3 showing
the ac plane orbital ordering. Long Mn-O bonds are depicted by thick black bonds,
whereas thin lines are used for the short Mn-O bonds.

145
 Perovskites containing the same A-site cation and substituting the Jahn-Teller

active Mn3+ with an element of similar ionic radius have similar tolerance factors and

similar octahedral tilting magnitudes. The high-spin crystal radius for both Fe3+ and Mn3+

is 0.785 Å (Shannon, 1976). The approximate crystal structure in the absence of a Jahn-

Teller active ion can be obtained by replacing Mn3+ with octahedrally symmetric Fe3+.

Examination of AFeO3 (A = La, Pr, Nd, Sm, Eu, Gd, Tb, Tb, Er, Dy, Ho, Y, Tm, Yb, Lu)

(Marezio et. al, 1970) and AMnO3 (A=La, Pr, Nd, Dy, Tb, Ho) series confirms that both

sets of compounds crystallize with the same tilt system and space group. The Mn-O bond

distances versus the tolerance factor are shown in figure 6.4, while O-Mn-O bond angles

versus the tolerance factor are shown in figure 6.5 (top). Mn-O bond distances and O-

Mn-O bond angles undergo a larger distortion for a smaller tolerance factor (larger

octahedral tilt angle). The largest cis-O-Mn-O distortion from 90° is the bond angle

between the two longest Mn-O bonds, which are weaker and more susceptible to

distortion. The bond angle distortion parameter is defined in the equation below where

θXMX is the individual cis-O-Mn-O bond angle and <θXMX> is the mean of the cis-O-Mn-

O bond angles less than 90°.

2
1  [θ XMX ( n ) − < θ XMX >] 
∆θ XMX = ∑ 
3  n =1,3 < θ XMX > 
Figure 6.5 (bottom) shows the bond angle distortion is not completely due to the

octahedral tilting in the AFeO3 and AMnO3 series. A comparison of lattice parameters for

the AMnO3 (A=La, Pr, Nd, Dy, Tb, Ho) and AFeO3 is shown in figure 6.6. Since both

AMnO3 and AFeO3 have approximately the same degree of tilting (for the same A-site

146
cation), a comparison of the two reveals the influence of the CJTD on the lattice

parameters. Comparing AMO3 with Fe3+ in place of Mn3+ show on average a 4.1%

decrease [a(Fe) < a(Mn)] in a, 2.8% increase [b(Fe) > b(Mn)] in b and 0.4% increase

[c(Fe) > c(Mn)] in c. In the absence of tilting the a and c axes expand equally, while the b

axis would contract in response to the CJTD. The b axis is smaller for AMnO3 because of

the slightly shorter bonds formed to increase the BVS in response to the decrease in

bonding from the long Mn-O bonds in the ac plane. The multiple axis tilting (a-b+a-)

results in the largest change to the a lattice parameter, while the c axis changes the least.

In phase tilting along the b axis results in the long Mn-O bond orientation to shift more

along the a axis compared to the c axis as shown in figure 6.3. The lattice distortion index

(D) is a gauge in the distortion of the lattice from pseudocubic symmetry and is defined

in the equation below, where an is the individual lattice parameter converted to the length

of the equivalent primitive unit cell (ap ~ 4 Å) (a/√2, c/√2 and b/2) and <a> is the mean

of the converted lattice parameters.

1 an − a
∆D = ∑
3 n =1,3 a
100

The effect of the CJTD on the lattice parameters is illustrated in the plot of D

versus tolerance factor shown in figure 6.7.

147
 ∆ dM

60.0

50.0

40.0

30.0

20.0

10.0

0.0
0.95
∆dMn

∆dFe
La

0.94
0.93
Pr

Tolerance Factor
0.92
Nd

0.91
0.90
Ho Tb

0.89
Dy

0.88
2.25

2.20

2.15

2.10

2.05

2.00

1.95

1.90

1.85

M-O Bond Distance (Å)

Figure 6.4 : M-O distances vs. tolerance factor for AMnO3 and AFeO3 series. The filled
symbols are M=Mn and open symbols are M=Fe. ∆dM (M=Fe, Mn) vs. tolerance factor
are connected with lines to guide the eye.

148
 Dy Ho Tb Nd Pr La
95.0

94.0
O-M-O bond angle (°)

93.0

92.0

91.0

90.0
0.88 0.89 0.90 0.91 0.92 0.93 0.94 0.95
Tolerance Factor

Dy Ho Tb Nd Pr La
3.5

3.0

2.5
∆ O-M-O (×10 )
4

2.0

1.5

1.0

0.5

0.0
0.88 0.89 0.90 0.91 0.92 0.93 0.94 0.95
Tolerance Factor

Figure 6.5 : O-M-O bond angles vs. tolerance factor for AMO3 (M = Fe, Mn) (top). ∆O-
M-O (×104) vs. tolerance factor for AMO3 (M = Fe, Mn) (bottom). Filled symbols are M
= Mn and open symbols M = Fe.

149
 0.95
0.94
La

0.93
Pr

Tolerance Factor
0.92
Nd

0.91
0.90
Ho Tb

0.89
Dy

0.88
5.90

5.80

5.70

5.60

5.50

5.40

5.30

5.20

5.10

Lattice parameter (Å)

Figure 6.6 : Filled symbol represent lattice parameters for experimental crystal structures
of AMnO3. Open symbol represent lattice parameters for experimental crystal structures
AFeO3. Lattice parameters are represented by diamonds (a), circles (b/√2) and triangles
(c ).

150
 0.95
La
La

0.94
Ru
Sb Ta Nb

0.93
Pr

Nd

Tolerance Factor
0.92
Nd

0.91
0.90
Ho Tb

0.89
Dy

0.88
5.0%

4.0%

3.0%

2.0%

1.0%

0.0%

D (Lattice Distortion)

Figure 6.7 : Orthorhombic distortion for the AMnO3, AFeO3, Ca2MnMO6 (M = Sb, Ta,
Nb, Ru). Symbols are defined as black diamonds (AMnO3), open diamonds (AFeO3),
black squares (A2MnGaO6 A = Nd, La), open circles [Ca2MnMO6(M = Ru, Sb)], closed
circles [Ca2MnMO6(M = Nb, Ta)].

151
 The A-site cation displacement off of the high symmetry site is significantly

larger in the x fractional coordinate direction for the AMnO3 series than it is for the

AFeO3 series. This displacement occurs because of the movement of the A-site cation

into a cavity formed by the combination of octahedral tilting and the Jahn-Teller

distortion of the octahedra. The A-site displacement from the high symmetry position

versus tolerance factor for AMO3 (M=Fe, Mn) is shown in figure 6.8. A-O bond lengths

are very similar for the first six distances, while the 7th and 8th distances are shorter and

the 9th to 12th (weakly-bonding) bonds are longer in AMnO3 compared to AFeO3. The

majority of the A-O bonding is contained in the first six bonds, e.g. 80.4% and 82.0% of

the Nd bond valence sum of 2.96 v.u. and 2.93 v.u. occurs from the first six Nd-O bonds

of NdMnO3 and NdFeO3 respectively. Twelve Nd-O bond distances for NdMnO3 and

NdFeO3 are shown in table 6.3 illustrating the similarity in coordination environments for

the A-site cations even with the Jahn-Teller distortion and significantly different lattice

parameters. The A-O distances for the AMO3 (M=Fe, Mn) series of the shortest eight A-

O distances are shown in figure 6.9. The distribution of A-O distances shows very little

perturbation in response to the CJTD.

152
 AMnO3 x AMnO3 z
X AFeO3 x AFeO3 z
0.09
A2MnGaO6 (A = La, Nd) x A2MnGaO6 (A = La, Nd) z
Ca2MnMO6 (M=Nb, Ta) x Ca2MnMO6 (M=Nb, Ta) z
0.08
Ca2MnM'O6 (M' = Ru, Sb) x Ca2MnM'O6 (M' = Ru, Sb) z
0.07

0.06

0.05

0.04

153
0.03

0.02

0.01 Z

A-cation Displacement (fract. coord)

0.00
0.88 0.89 0.90 0.91 0.92 0.93 0.94 0.95
Tolerance Factor

A2MnGaO6 (A=La, Nd). The symbol description order is the same as figure 6.7.
Figure 6.8 : The A-site fractional displacement from high symmetry position in Pnma is
shown versus tolerance factor for AMO3, AFeO3, Ca2MnMO6 (M = Sb, Ta, Nb, Ru), and
 La

0.94
0.93
Pr

0.92
Tolerance Factor
Nd

0.91
0.90
Ho Tb

0.89
Dy

0.88
2.80

2.70

2.60

2.50

2.40

2.30

2.20

A-O Bond Distance (Å)

Figure 6.9 : Experimental A-O bond distances (8 shortest) for AMnO3 (filled) and
AFeO3 (open). Symbols containing × or + indicate a multiplicity of two (see table 6.3).

154
 Nd-O NdFeO3 NdFeO3 NdMnO3 NdMnO3
Distance (Å) BVS % (Å) BVS %
Number
1 2.34 18.0% 2.35 17.6%
2 2.38 16.3% 2.39 16.0%
3 2.38 16.3% 2.39 16.0%
4 2.44 13.7% 2.46 13.0%
5 2.60 8.9% 2.60 8.9%
6 2.60 8.9% 2.60 8.9%
7 2.73 6.3% 2.63 8.3%
8 2.73 6.3% 2.63 8.3%
9 3.17 1.9% 3.16 2.0%
10 3.24 1.6% 3.47 0.9%
11 3.42 1.0% 3.55 0.7%
12 3.42 1.0% 3.55 0.7%

Table 6.3 : Nd-O bonds distances and valence for experimental crystal structures of
NdFeO3 (BVS = 2.93 v.u.) and NdMnO3 (BVS = 2.96 v.u.).

In both series the decreasing tolerance factor (increased tilting) leads to a larger

distortion of the O-M-O angles. However, this trend is much more pronounced in the

AMnO3 series due to the J-T distortion. Pnma perovskites become unstable as the

octahedra tilting increases (τ < 0.88). Soft chemistry synthesis and/or high pressure

synthesis is required to stabilize the perovskite phase for manganites with an A-site ion

smaller than Tb3+(Alonso et al., 2000). A hexagonal structure type (space group P63cm)

with trigonal bipyramid coordination for Mn observed for AMnO3 as the ionic radius of

the A-site cation becomes very small (τ < 0.88). A hexagonal structure type (space group

P63/mmc) with a trigonal bipyramid coordination for Fe is observed for for InFeO3 (τ =

0.85) (Giaquinta et al., 1994). Trigonal bipyramidal coordination has no electronic

degeneracy of the highest occupied molecular orbital (HOMO) for Mn3+, thus the J-T

155
distortion adds to the destabilization of the perovskite phase. Only AMnO3 compounds

synthesized by conventional solid state techniques were used in this study.

6.5 Cation Ordering, octahedral tilting and J-T distortions in A2MM′X6 systems

The compositions A2CuMO6 (A=Ba, Sr; M = W, Te) have a strong driving force

for rock salt cation ordering due to the oxidation state difference of four between Cu2+

and Te6+/W6+. Additionally, they exhibit out-of phase tilting about the c-axis (a0a0c-). The

long Cu-O bonds are oriented parallel to the c-axis inducing a large tetragonal distortion.

Lattice parameter ratios c/(√2a) for A=Ba (τ = 1.042) are 1.098 and 1.092 and for A=Sr

(τ = 0.983) are 1.096 and 1.102 for M = W and M = Te, respectively. An octahedral

distortion ∆dCu = 70.8×10-4 (M = W) and 85.0×10-4 (M = Te) is observed for A=Sr,

whereas ∆dCu = 79.9×10-4 (M = W) and 84.8×10-4 (M = Te) when A = Ba. It is interesting

to note that octahedral tilting occurs even for Ba2CuWO6 and Ba2CuTeO6 where the

tolerance factor is much larger than unity. The tilt angle is slightly larger in the Sr

analogs.

In La2CuSnO6 the charge difference of two, combined with a large J-T distortion

about Cu2+, and the optimal degree of tilting, stabilizes a layered ordering of octahedral-

site cations. Layered octahedral site ordering is unusual and a narrow stability range of

octahedral tilting (τ = 0.917) exists for this structure type. In this structure the long Cu-O

bonds are all aligned effectively in a single direction (along the a-axis of P21/m), similar

to the Ba2CuWO6 structure. Additional layered compounds of A2CuSnO6 with A = Pr

156
and Nd, as well as La2CuZrO6 have been synthesized under high-pressure high-

temperature conditions (Azuma et al., 1998).

Upon going from La2CuSnO6 (τ = 0.917) to La2CuTiO6 (τ =0.945) (Palacin et al.,

1993) the difference in the radii of the octahedral cations actually increases (rTi = 0.75 Å,

rSn = 0.83 Å & rCu = 0.87 Å), yet the long range cation ordering disappears, as does the

CJTD. The octahedral site contains a random mixture of Cu2+ and Ti4+ with regular M-O

bond distances (Å) of 2.033 (×2), 2.022 (×2), and 2.007 (×2) and a small ∆dCu =

0.27×10-4. This example illustrates how sensitive the long range order of the J-T

distortion is to changes in octahedral tilting and/or cation ordering.

The crystal structures of A2MnGaO6 (A=La, Nd) were determined from Rietveld

refinement of high-resolution neutron powder diffraction data (Cussen et. al, 2001).

Octahedrally symmetric Ga3+ and the J-T Mn3+ are disordered over a single

crystallographic site (as in La2CuTiO6), resulting in a dilution of the magnitude of the

Jahn-Teller distortion. Surprisingly, the refinements show that the coordination of the

disordered octahedral site in each compound is quite different. Nd2MnGaO6 has a

considerable CJTD (∆d = 12.7×10-4), albeit reduced compared to AMnO3, with M-O

bond lengths (Å) of 1.970 (×2), 2.101 (×2), 1.935 (×2). In contrast, the octahedral site

coordination in La2MnGaO6 has symmetric M-O bond lengths (Å) of 1.990 (×2), 1.982

(×2), 1.978 (×2) (∆d = 0.06×10-4). This does not necessarily imply the complete absence

of a J-T distortion in La2GaMnO6, rather it is more likely that J-T distortions are present

locally but the orientation of these distortions is disordered. Bond angles (O-Mn-O) in

La2MnGaO6 and Nd2MnGaO6 have bond angle distortion parameters (∆θXMX = 0.37×10-4

157
and 0.16×10-4) comparable to LaMnO3 and NdMnO3 (∆θxbx = 0.11×10-4 and 0.39×10-4).

While it is true that the A-site cations lanthanum and neodymium have slightly different

twelve coordinate ionic radii, 1.5 Å and 1.41 Å respectively (Shannon, 1976), Nd-O and

La-O bonding are expected to be similar. Thus the presence of a CJTD in Nd2GaMnO6,

but not in La2MnGaO6, appears to originate from the larger magnitude of octahedral

tilting induced by the smaller Nd3+.

In order to further investigate the behavior of Mn3+ in double perovskites,

consider the A2MnMO6 (A = Ca, Sr; M=Ru, Sb, Ta, Nb) series where changes in cation

ordering, Jahn-Teller distortions, and octahedral tilting are expected. Determination of

the presence or absence of a CJTD requires accurate crystallographic information of the

oxygen atoms, preferably obtained from refinements of neutron powder diffraction data.

A literature search was conducted for the A2MM′O6 systems with an equal ratio of

symmetric octahedral ions and Jahn-Teller active ions. Structures of Sr2MnMO6 (M = Sb,

Ru, Ta, Nb) refined from powder diffraction data crystallize in I4/mcm (a0a0c-) (M = Nb,

Sb, Lufaso & Woodward, 2002; Ta-Woodward, 1997c; Ru-Woodward, 2002). Tolerance

factors, M-O bond lengths, distortion parameters, lattice parameters and orientations of

the long M-O bonds are shown in table 6.4. Structures of Ca2MnMO6 (M = Ta,

Woodward, 1997c; Nb, Sb, Ru, this study) have a disordered arrangement of octahedral

site cations and are refined in tilt system a-b+a- (space group Pnma). Tolerance factors,

M-O bond lengths, distortion parameters, lattice parameters and orientations of the long

M-O bonds are shown in table 6.5. In all cases a Jahn-Teller active ion and an

octahedrally symmetrical ion are disordered on the octahedral site.

158
 Sr2MnRuO
Sr2FeRuO6 Sr2FeSbO6 Sr2MnSbO6
6
a (Ǻ) 5.5379 5.4545 5.6132 5.5536
b (Ǻ) 5.5429 - 5.5973 -
c (Ǻ) 7.8782 7.9333 7.9036 8.0528
Tilt system a-b-b- a0a0c- a-b+a- a0a0c-
M-O (Ǻ) 1.970 ×2 1.983 ×2 2.006 ×2 2.013 ×2
M-O (Ǻ) 1.967 ×2 1.947 ×4 1.995 ×2 1.979 ×4
M-O (Ǻ) 1.967 ×2 - 1.994 ×2 -
∆d (x10-4) 0.01 1.04 0.07 0.67
Symmetry I2/c I4/mcm P21/n I4/mcm
Tolerance factor 0.9874 0.9898 0.9775 0.9773
Reference Battle et al., Woodward Cussen et
This Study
1989 et al., 2002 al., 1997

*
Sr2FeTaO6 Sr2MnTaO6 Sr2FeNbO6 Sr2MnNbO6
a (Ǻ) 5.6204 5.6135 5.6084 5.6098
b (Ǻ) 5.6161 - 5.6082 -
c (Ǻ) 7.9266 7.9510 7.9642 7.9237
Tilt system a-b+a- a0a0c- a-b+a- a0a0c-
M-O (Ǻ) 1.998 ×2 1.988 ×2 - 1.981 ×2
M-O (Ǻ) 1.994 ×2 1.996 ×4 - 1.993 ×4
M-O (Ǻ) 1.988 ×2 - - -
∆d (x10-4) 0.05 0.04 - 0.08
Symmetry Pbnm I4/mcm Pbnm I4/mcm
Tolerance factor 0.9841 0.9839 0.9868 0.9866
Cussen et Woodward Tezuka et
Reference This Study
al., 1997 1997c al., 2000

Table 6.4 : Lattice parameters, tilt system, M-O bond distances, octahedral distortion
parameters, space group, tolerance factor and reference for Sr2MM′O6 (M = Mn, Fe; M′ =
Nb, Ta, Ru, Sb). Monoclinic angles are Sr2FeRuO6 (β = 90.11°) and Sr2FeSbO6 (β =
90.01°). *The crystal structure of Sr2FeNbO6 was refined from powder x-ray diffraction
data.

159
 Ca2FeRuO6 Ca2MnRuO6 Ca2FeSbO6 Ca2MnSbO6
a (Ǻ) - 5.3371 5.5283 5.4598
b (Ǻ) - 5.4236 5.4389 5.5549
c (Ǻ) - 7.5794 7.7358 7.6943
Tilt system - a-b+a- a-b+a- a-b+a-
M-O (Ǻ) (ab1) ×2 - 1.957 - 2.022
M-O (Ǻ) (ab2) ×2 - 1.955 - 1.993
M-O (Ǻ) (c) ×2 - 1.946 - 1.988
∆d (x10-4) - 0.06 - 0.56
Symmetry - Pbnm P21/n Pbnm
Tolerance factor - 0.9361 0.9241 0.9248
Battle et al.,
Reference - This Study This Study
1995

Ca2FeTaO6 Ca2MnTaO6 Ca2FeNbO6 Ca2MnNbO6

a (Ǻ) 5.4498 5.4574 5.4480 5.4532
b (Ǻ) 5.5482 5.5664 5.5517 5.5639
c (Ǻ) 7.7591 7.7166 7.7612 7.7008
Tilt system a-b+a- a-b+a- a-b+a- a-b+a-
M-O (Ǻ) (ab1) ×2 - 2.024 - 2.011
M-O (Ǻ) (ab2) ×2 - 1.989 - 2.001
M-O (Ǻ) (c) ×2 - 1.988 - 1.982
∆d (x10-4) - 0.70 - 0.36
Symmetry P21/n Pbnm P21/n Pbnm
Tolerance factor 0.9307 0.9305 0.9332 0.9330
Chakhmoura
Woodward, Woodward, dian &
Reference This Study
1997c 1997c Mitchell,
1998

Table 6.5 : Lattice parameters, tilt system, M-O bond distances, octahedral distortion
parameters, space group, tolerance factor and reference for Ca2MM′O6 (M=Mn, Fe;
M′=Nb, Ta, Ru, Sb). The monoclinic angles are β = 90.00° for Ca2FeSbO6, β = 90.11° for
Ca2FeNbO6 and β = 90.072° for Ca2FeTaO6.

160
6.6 Orbital Ordering and Cation Ordering

The orbital ordering schemes observed in ACuF3 (a and d-type) and AMnO3 (a-

type) perovskites are shown in figure 6.10a and 6.10b (neglecting octahedral tilting

distortions). Upon going from a ternary to an ordered A2MM′X6 perovskite, several

arrangements are possible. If simply every other J-T active cation is replaced with a

symmetric cation in a 1:1 rock salt ordering pattern, the ordering pattern shown in figure

6.10c results. This is the orbital ordering scheme adopted by A2CuM6+O6 compounds.

Note that now all of the long M-X bonds are oriented in the same direction. For this

reason this pattern of orbital ordering is referred to as a single axis Jahn-Teller distortion.

If both ac-plane (alternating) orbital ordering and rock salt cation ordering are imposed

the result is shown in figure 6.10d. This arrangement is unfavorable because the anions

are compressed about ½ of the symmetric M′ cations, creating two crystallographically

and chemically distinct sites for the M′ cations. No reported compound adopts such an

arrangement based on a literature search. This unfavorable arrangement can be avoided if

one reverses the orientation of the occupied eg orbitals in every other row as shown in

figure 6.10e. In this arrangement the long M-X bonds are stabilized by a shift of the

symmetric M′X6 octahedra as indicated by the arrows in figure 6.10e. Notice that in

figure 6.10e, where rock salt order is maintained, that M′X6 octahedra in neighboring

layers shift in opposite directions. This creates a lattice strain that is energetically

unfavorable. However, if the cation ordering is modified to the chain type (see figure

6.1), the symmetric M′X6 octahedra in neighboring layers can shift in the same direction

161
as shown in figure 6.10f. This is the CJTD and orbital ordering scheme reported for

LnAMn3+Mn4+O6 compounds, such as NdSrMn2O6 (Woodward et al., 1999). Clearly the

cation and orbital ordering possibilities are diverse.

162
 Layer 1 Layer 2
(a) AMX3 a-type

(b) AMX3 d-type

(c) A2CuM6+O6 type

163
(d) Ordered ac-type

(e) Pseudo NdSrMn2O6 (rock salt ordering) type

(f) NdSrMn2O6 type

Figure 6.10 : Orbital ordering schemes in (a-b) ternary AMX3 perovskites and (c-f)
quaternary, ordered A2MM′X6 perovskites. The solid lines indicate the simple cubic
perovskite unit cell, while the dotted line shows the a and c axes of the Pnma unit cell.
Only the octahedral ions are shown for clarity.

164
6.7 Modeling using SPuDS

Structure modeling with the software program SPuDS has been used to accurately

predict the crystal structures of perovskites with symmetric octahedra as shown in

chapters 3 and 4 (Lufaso & Woodward, 2001). Modifications to the software

implemented an option to model a J-T distortion of the octahedra in order to investigate

the ideal theoretical structure type for compounds containing a Jahn-Teller active ion.

SPuDS modeling of the Pnma J-T perovskites for simple or disordered octahedral site

cations has default values that are available to model a typical J-T distortion found in the

AMnO3 (A = trivalent lanthanide) series. SPuDS requires bond distance addition factors

to define the size of the Jahn-Teller distortion and bond valence sum, which is adjustable

by the user. GII versus tolerance factor values for the AMnO3 series are shown in figure

6.11. Ordering of the long Mn-O bond along the b-axis (single axis CJTD) is

considerably less stable than the experimentally observed ac plane CJTD. The difference

between the b-axis and experimental ac ordering is not large when the octahedral tilting

is small, but it increases as the tolerance factor decreases. The decrease in the GII

obtained from SPuDS calculations is interesting in that it agrees with the experimentally

observed increase in the magnitude of the J-T distortion in the AMnO3 series (see figure

6.4).

165
 0.95
La

0.94
0.93
Pr

Tolerance Factor
0.92
Nd

0.91
0.90
Ho Tb

0.89
Dy

0.88
0.19

0.18

0.17

0.16

0.15

0.14

0.13

0.12

GII

Figure 6.11 : GII versus tolerance factor of AMnO3 with two orientations of long Mn-O
bonds. Filled squares represent b-axis orientation and the open circles experimental ac
plane orientation.

166
6.8 Discussion

The Jahn-Teller theorem states a distortion of the octahedra should occur,

however it does not specify the magnitude of ∆dM. Ternary perovskites containing a J-T

ion exhibit a large CJTD with octahedra of CuF6 exhibiting a larger ∆dM compared to

MnO6 octahedra. In RbCuF3, the Rb cation causes a lengthening of the longest and

weakest Cu-F bonds resulting in a larger distortion of the octahedra than the one observed

in KCuF3. Ordered perovskites containing Cu2+, A2CuMO6 (A=Ba, Sr; M=W, Te),

contain distorted CuO6 octahedra with similar size and shape as those observed for

RbCuF3. Increased octahedral tilting leads to an expansion of the long Mn-O bonds,

coupled with a slight contraction of the four shorter Mn-O bonds. This leads to an

increase in ∆dM as the tolerance factor decreases across the AMnO3 series. Increased

octahedral bond angle distortion of the X-M-X bond angles occurs for the Pnma AMnO3

series, while no increased ∆θXMX is observed with increased tilting in the corresponding

AFeO3 series. A zero bond angle distortion is observed in untilted ACuF3 (A = Rb, K),

however the octahedrally tilted NaCuF3 is strongly distorted with an average distorted Cu

coordination ∆θXMX = 36.6×10-4.

Evidence for a CJTD is exhibited in the coordination of the octahedral site, A-site

fractional displacements, and lattice parameters in Pnma AMnO3 perovskites.

Examination of Rietveld refinements of neutron powder diffraction data show the local

environment of octahedral cations in AMnO3, A2MnMO6 (A=Sr, Ca; M = Ru, Sb, Nb, Ta)

and A2MnGaO6 (A = La, Nd). The bond distortion parameter versus tolerance factor is

167
shown in figure 6.12 for A2MnGaO6 (A=La, Nd) and A2MnMO6 (A=Sr, Ca; M = Sb, Ta,

Nb, Ru) which is reduced in comparison to the AMnO3 series shown in figure 6.4. Long

range cation order is absent in all of those compositions, but the J-T distortion is

exhibited in the crystal structures in different manners depending upon the octahedral

tilting and degree of local cation ordering.

168
 1.00
Nb Ru

0.99
Ta

0.98
Sb

0.97
Tolerance Factor
0.96
0.95
La

0.94
Sb Ta Nb Ru

0.93
Nd

0.92
14.0

12.0

10.0

8.0

6.0

4.0

2.0

0.0

Distortion Parameter ∆d M (x104)

Figure 6.12 : Octahedron distortion parameter versus tolerance factor for A2MnGaO6
(A=La, Nd) and A2MnMO6 (A=Sr, Ca; M = Sb, Ta, Nb, Ru).

169
 The Sr2MnMO6 (M=Ru, Sb) series has c/(√2a) ratios of 1.028 and 1.025,

respectively and M-O bond distances shown in table 6.4. This implies local Jahn-Teller

distortions are oriented along the c-axis. Reduced (compared to AMnO3) ∆dM values are

due in part to dilution of the Jahn-Teller ion with the symmetric ion. Medium range

cation order, proposed as regions of cation ordering along the c-axis, stabilizes the CJTD

in Sr2MnMO6 (M = Ru, Sb). Local regions of cation order in Sr2MnRuO6 were reported

from TEM studies (Woodward et al., 2002). Short range cation interactions destabilize

the CJTD in Sr2MnMO6 (M=Nb, Ta) which show no evidence for a CJTD through

examination of the lattice parameters and octahedral site coordinations.

The magnitude of the CJTD may be large in a random mixture of octahedral site

cations and is dependent on the octahedral tilting and significant enough to exhibit a

presence in Nd2MnGaO6 and absence in La2MnGaO6. A considerable increase of the

∆dMn at τ =0.942 to 0.918 occurs for A=La to A=Pr in AMnO3, then the increase occurs

much more slowly as the tolerance factor continues to decrease. Increased octahedral site

distortion observed in A2MnGaO6 (A=La, Nd) is in a similar tolerance factor range (τ

=0.948 to 0.925). The A-site displacement in Nd2MnGaO6 is intermediate between the

trend for structures with a single octahedral site Jahn-Teller distortion and structures with

a symmetric ion, whereas the La2MnGaO6 structure has an A-site displacement consistent

with a symmetric ion. Lattice parameters are useful in discerning if local orbital ordering

is present and the lattice distortion index for Ca2MnMO6 are slightly larger than for

Ca2FeMO6, which suggests the presence of a reduced CJTD and short range (local)

170
orbital ordering. The average coordination environment of the octahedral cation in

Ca2MnM5+O6 is relatively symmetric with no evidence for a CJTD.

Perovskite oxides containing Ca2+ as the A-site cation and 1:1 Cu2+:M6+

octahedral site (i.e. Ca2CuWO6, Ca2CuTeO6) have not been reported in the literature.

SPuDS modeling indicates the structure is stable in the P21/n space group (a-b+a-) and a

large Jahn-Teller distortion and large octahedral tilting would be compatible. Using the

bond distance factors (x = -0.055, y = 0.25, z = -0.12), SPUDS predicted structure

resulted in Cu-O bond distances of 2.02 Å ×2, 2.00 Å ×2, and 2.32 Å ×2. The bond

valence sums for Ca2CuWO6 were Ca2+ 2.00 v.u., Cu2+ 2.00 v.u., W6+ 6.04 v.u., O2-(4e)

2.05, 2.02 and 1.94 v.u., with a GII = 0.04 v.u. implying a reasonably stable structure.

Competition from other non-perovskite phases was reported in the case of Ca2CuWO6,

indicating the perovskite structure is not the most stable for that combination of ions

under normal synthetic conditions (Blasse, 1965). Absence of a perovskite with a CJTD,

cation ordering and large octahedral tilting distortion suggests the combination is not

stable compared to mixtures of other phases. Increased octahedral distortion of CuO6 and

large tilting are not observed.

6.9 Conclusions

The crystal chemistry of perovskites containing an active Jahn-Teller ion is

diverse. The magnitude and orientation of the observed CJTD is dependent upon not only

the cation-anion pair, but also exhibits sensitivity to local cation order and octahedral

tilting. The coordination environment of Mn3+ appears to be more sensitive than Cu2+ to

171
the transition from a ternary AMX3 perovskite to an ordered quaternary A2MM′X6

perovskite. Octahedral site cation ordering influences local ordering of the Jahn-Teller

distortions and is proposed to explain the lattice distortions observed in A2MnMO6. A

large difference in both charge and size is responsible for the observation that cation

ordering is always present in Cu2+-M6+ perovskites at room temperature. Crystallographic

information of the octahedral site coordination and lattice parameters are useful in

ascertaining the presence and orientation of a CJTD. Jahn-Teller distortions can be

enhanced by chemical pressure induced by oversized A-site cations (τ > 1). Comparison

of lattice parameters and A-site cation fractional coordinates to structures with an

octahedrally symmetric ion of approximately the same size as the J-T cation is useful in

determining the presence of a CJTD. In ternary perovskites, AMX3, increased octahedral

tilting and orbital ordering in the ac plane enhance the size of a CJTD. This is the most

stable arrangement for Pnma perovskites containing Mn3+ according to SPuDS

calculations, which demonstrated orbital ordering along a single axis results in a large GII

due to insufficient bonding to the anion of the long Mn-O bonds. Rock salt ordering of

the octahedral site cations is incompatible with ac plane ordering due to the creation of

two different sites for the octahedrally symmetric ion. Orbital ordering along a single axis

is more stable for rock salt ordered A2Cu2+M6+X6 compositions with reduced octahedral

tilting (τ > 0.98).

6.10 References

Ahtee, M., Glazer, A.M. & Hewat, A.W. (1978). Acta Cryst. B34, 752-758.

172
Alonso, J.A., Martinez-Lope, M.J., Casais, M.T. & Fernandez-Diaz, M.T. (2000). Inorg
Chem. 39, 917-923.

Anderson, M.T. & Poeppelmeier, K.R. (1991) Chem. Mater. 3, 476-486.

Anderson, M.T., Greenwood, K.B., Taylor, G.A. & Poeppelmeier, K.R. (1993). Prog.
Solid St. Chem., 22, 197-233.

Azuma, M., Kaimori, S. & Takano, M. (1998). Chem. Mater. 10, 3124-3130.

Battle, P.D., Gibb, T.C., Jones, C.W. & Studer, F. (1989) J. Solid State Chem. 78, 281-
293.

Blasse, G. (1965). J. Inorg. Nucl. Chem. 27, 993-1003.

Brown, I.D (1978). Chem. Soc. Rev. 7, 359-376.

Brown, I.D (2002), http://www.ccp14.ac.uk/ccp/web-mirrors/i_d_brown (Accessed Oct

2002).

Buttner, R.H., Maslen, E.N. & Spadaccini, N. (1990). Acta Cryst. B46, 131-138.

Cussen, E.J., Vente, J.P., Battle, P.D. & Gill, T.C. (1997). J. Mater. Chem. 7, 459-463.

Cussen, E., Rosseinski, M.J., Battle, P.D., Burley, J.C., Spring, L.E., Vente, J.F.,
Blundell, S.J., Coldea, A.I. & Singleton, J. (2001). J. Am. Chem. Soc. 123, 1111-1122.

Glazer, A.M. (1972). Acta Cryst. B28, 3384-3392.

Glazer, A.M. (1975). Acta Cryst. A31, 756-762.

Goodenough, J.B. (1998). Annu. Rev. Mater. Sci. 28, 1- 27.

Hidaka, M., Eguchi, T. & Yamada, I. (1998). J. Phys. Soc. Jpn. 67, 2488-2494.

Howard, C.J., Knight, K.S., Kennedy, B. J. & Kisi, E. (2000). J. Phys.: Condens. Matter,
12, L677-L683.

Hutchings, M., Samuelsen, E.J., Shirane, G. & Hirakawa, K. (1969). Phys. Rev., 188,
919-923.

Iwanaga, D., Inaguma, Y. & Itoh, M. (1999). J. Solid State Chem. 147, 291-295.

Jiang, B., Zuo, J. M., Chen, Q. & Spence, J. C. H (2002). Acta Cryst. A58, 4–11.

173
Kaiser,V., Otto, M., Binder, F. & Babel, D. (1990). Z. Anorg. Allg. Chem. 585, 93-104.

Kanamori, J. (1960). J. Appl. Phys. 31, 14S-23S.

Kennedy, B.J., Howard, C.J. & Chakoumakos, B.C. (1999) J. Phys.: Condens. Matter,
11, 1479-1488.

Lufaso, M.W. & Woodward, P.M. (2001). Acta Cryst, B57, 725-738.

Marezio, M., Remeika, J.P. & Dernier, P.D. (1970) Acta Cryst., B26, 2008-2022.

Murakami, Y., Hill, J. P., Gibbs, D., Blume, M., Koyama, I., Tanaka, M., Kawata, H.,
Arima, T. & Tokura, Y. (1998). Phys. Rev. Lett. 81, 582-585.

Palacin, M.R., Bassas, J., Rodriguez-Carvajal, J. & Gomez-Romero, P. (1993). J. Mater.

Chem. 3, 1171-1177.

Rodriguez-Carvajal, J., Hennion, M., Moussa, F., Moudden, A.H., Pinsard, L. &
Revcolevschi, A. (1998). Phys. Rev. B, 57, 5259-5264.

Shannon, R.D. (1976). Acta Cryst. A32, 751-767.

Sleight, A.W., Gillson, J.L. & Bierstedt, P.E. (1975). Solid St. Comm. 17, 27-28.

Tezuka, K., Henmi, K. & Hinatsu, Y. (2000). J. Solid State Chem. 154, 591-597.

Woodward, P. (1997a). Acta Cryst. B53, 32-43.

Woodward, P. (1997b). Acta Cryst. B53, 44-66.

Woodward, P.M. Ph. D. Dissertation, Oregon State University, Corvallis, OR, 1997c.

Woodward, P.M., Cox, D.E., Vogt, T., Rao, C.N.R & Cheetham, A.K. (1999). Chem.
Mater. 11, 3538-3528.

Woodward, P.M. et al. (2002). To be published.

174
 CHAPTER 7

STRUCTURE PREDICTION, SYNTHESIS AND DIELECTRIC PROPERTIES OF

NOVEL DUAL A-SITE AND OCTAHEDRAL CATION ORDERED PEROVSKITES

7.1 Introduction

Perovskites have been applied as materials in dielectric resonators, band-pass

filters, voltage-controlled oscillators and duplex filters in advanced mobile and satellite

communication systems that operate in the microwave frequency range (300 MHz to 30

GHz). Dielectric materials suitable for microwave applications require a high dielectric

constant (relative permittivity) ε (20-40), low dielectric loss (tan δ < 0.001), a near zero

(0 ± 10 ppm K-1) temperature coefficient of the resonant frequency (τf ). Thermal and

chemical stability are material property requirements for long lasting and stable usage in

devices. Obtaining a material with an optimum combination of all desired properties is a

challenge from a fundamental scientific and engineering viewpoint.

Commercial application of dielectric filters is conservatively estimated at $600 - $800

million in the U.S. and Japan (Reaney & Ubic, 2000). Materials used in commercial

applications depend on several factors including the frequency of operation of the device.

Device miniaturization and increasing interest in the gigahertz frequency range require

the use of materials with a larger dielectric constant. In many applications, higher

dielectric constant materials are favored due to the smaller amount of material required to

175
give the same resonant frequency. Low ε and high quality factor (Q) materials (e.g.

Al2O3) are favored for extremely high frequency application; otherwise the filters become

unmanageably small. Cost of raw materials and processing issues are a major factor in

the commercial market for microwave dielectric ceramics.

Impedance spectroscopy is a powerful method to characterize many of the

electrical properties of materials (Irvine et al., 1990). Impedance measured directly in the

frequency domain occurs by application of a single-frequency voltage to the materials

and measurement of the phase shift and amplitude (real and imaginary) components of

the resulting current at that frequency. Permittivity of a dielectric material has a real and

imaginary mathematical representation. The symbol ε′′ represents the imaginary part of

the permittivity and describes the energy loss from an AC signal as it passes through the

dielectric. The symbol ε′ represents the real part of the permittivity and is also called the

relative permittivity or dielectric constant. Complex permittivity is expressed in the

equation below in terms of the real, ε′, and imaginary component ε′′.

ε = (ε '+iε " )

Dielectric properties of materials are frequency dependent and a great deal of

information can be lost by not examining the variation of properties with frequency

(Hirose & West, 1996). Full characterization of a dielectric material spans the entire

frequency spectrum, a wide range from 10-4 to 1010 Hz. Capacitance inherent in

insulating materials is determined by the combination of ionic and electronic polarization

mechanisms. Electronic polarization results from deformation of the electron cloud

surrounding the atom’s nucleus. The ionic component of polarization results from

176
displacement of cations and anions in the crystal structure. Ionic polarization is

important up to infrared frequencies (~10 THz) and is related to lattice vibration.

Orientational and space charge polarization are two other factors observed in molecular

systems and multiple phase systems. In the ideal case the dielectric constant is related to

the strength of the electronic polarization and ionic oscillations, making it independent of

frequency (Reaney & Ubic, 2000). Dielectric loss in the ideal case is related to the

anharmonic terms in the material’s potential energy making it frequency dependent

(Reaney & Ubic, 2000).

Analysis of the impedance spectrum using a suitable equivalent circuit model

generates frequency dependant capacitance and dielectric loss values. Relative

permittivity is the capacitance (Cp) of the dielectric material between two parallel plates

relative to the dielectric properties of a vacuum (or air), where Cv is the capacitance of

the same thickness of air between the same two parallel plates.

Cp
ε=
Cv

Capacitance of an empty cell with electrode area (A) and electrode separation of

distance d is shown in the equation below. Dielectric permittivity of free space is ε0 =

8.854 × 10-12 F m-1. Note that capacitance increases with larger surface area and thinner

electrode separation.

ε0A
Cv =
d

177
 Dissipation factor, also known as the loss tangent and dielectric loss, is the ratio

of the energy dissipated to the energy stored in the material. Dissipation factor is that

ratio the imaginary portion of permittivity (ε′′) and the real portion of permittivity (ε′).

ε"
tan δ =
ε'
Dissipated energy is typically turned into heat from conduction of electrons flowing

through the material or through the anharmonicity of the lattice vibrations (Reaney &

Ubic, 2000). Loss has also been attributed to defects and grain boundaries in

polycrystalline materials (Nomura, 1983). Loss increases with porosity (P), which is

defined as P = 1 – D where D is the fraction of the material’s theoretical density (Penn et

al., 1997). A lower dissipation factor is optimal in order to retain maximum signal and to

prevent excessive heat generation in devices.

Losses and dielectric dispersion are divided into two categories, intrinsic and

extrinsic (Penn et al., 1997). Intrinsic losses are related to the crystal structure (e.g.

atomic masses, atomic charges and bond strengths) and the interaction of the phonon with

the ac electric field. Extrinsic losses are caused by imperfections in the crystal including

impurities, grain boundaries, porosity, oxygen vacancies, microstructural defects and

random crystallite orientation. Cation disorder in perovskites causes an increase in the

frequency width of phonon vibrations at infrared frequencies and extends into the

microwave region. The increase in frequency width is attributed to an increase in loss at

microwave frequencies (Cava, 2001). Grain boundary contribution can lead to peaks in

the dissipation factor, which typically decreases in value at higher frequencies. Dielectric

loss in sintered pellets is dominated by extrinsic factors and thus is highly dependent on

178
sintering conditions. Ordering of the octahedral cations in perovskites leads to weak

coupling and a low energy leakage between the modes, which can result in lowered

intrinsic losses and a large increase in Q (Reaney & Ubic, 2000).

The quality factor (Q) is a term often reported in the characterization of dielectric

materials. Two definitions are employed, depending on the frequency range of the

measurement. At sub-microwave frequencies the quality factor is a measure of the

efficiency or power loss. At microwave frequencies a peak occurs in the transmitted

signal amplitude at the resonant frequency and has a finite width in a microwave

dielectric system. A high Q means a narrow peak, giving high selectivity to a given

frequency. Higher selectivity enables an increased density of channels in a given

frequency band. Peak resonant frequency (fo) divided by the peak width is equal to Q or

in sub-microwave frequency measurements is equal to 1/tan δ.

1 f
Q= or o
tan δ ∆f
The figure of merit, Qf, is Q times the resonant frequency in GHz. The resonant

frequency can be changed by manipulation of the volume of the resonating body and the

dielectric constant (Reaney et al., 2001). The relationship between the resonant

frequency, volume and dielectric constant is shown in the relation below.

1
f oα
Vε r1 / 2

Frequency selection can be achieved by altering the sample volume and dielectric

constant. Higher dielectric constant materials are preferred due to the ability to make

smaller volume components, which is an important consideration when expensive raw

179
materials (e.g. Ta2O5) are required. Maximum Q values are obtained with samples with

high (>95%) percent of the theoretical density (Takata & Kageyama, 1989).

Temperature coefficient of the resonant frequency (τf), the dielectric constant (τε),

or the capacitance (τc) can be used to indicate the thermal stability of dielectric properties

of a material. The temperature coefficient of the capacitance is obtained from

measurements taken as two temperatures as shown in the equation below (Park et al.,

2001).

τ c (T2 ) − τ c (T1 )
τc =
τ c (T1 )[T2 − T1 ]

The temperature coefficient of capacitance is related to the temperature

coefficient of the dielectric constant by the following equation (Colla et al., 1993) where

α is the linear expansion coefficient.

.τ ε =τc −α

The τf is related to the temperature coefficient of permittivity (τε) and the linear

expansion coefficient (α).

τε 
τ f = − +α
2 

Over a broad temperature range the linear thermal expansion coefficient (α) in the

perovskite structure does not deviate from the range 8 × 10-6 to 1 × 10-5 K-1 (8 to 100 ppm

K-1) (Park et. al, 2001) and more typically is between 8 and 20 ppm K-1 (Colla et al.,

1993). The signs of τf and τε are opposite and the τε controls the τf neglecting the small

change in the linear thermal expansion coefficient. Microwave circuits normally have

180
some low characteristic τf, so resonator components ideally counteract the inherent drift

with a small τf (Reaney & Ubic, 2000). Obtaining a material with a zero temperature

coefficient is one of the most difficult parts in development of dielectric ceramic

materials. Many high-Q high-dielectric constant materials that are available, such as TiO2

(ε > 100, Q = 14000, τf > 427 ppm/°C), are not suitable for many applications due to the

large value of τf. Adjustment of the τε of the dielectric resonator through 0 ppm/°C with

an accuracy of 0.5 ppm K-1 is a critical component in material’s design (Nagai et al.,

1992).

Estimation of the dielectric constant can be made by a linear dependence on the

polarizability of the individual ions, which is known as the oxide additivity rule. The

relationship known as the Clausius-Mossotti equation relates the dielectric constant of a

sphere of material with volume Vm to the macroscopic polarizability αm (Bosman &

Havinga, 1963). The macroscopic Clausius-Mossotti equation holds for cubic and

isotropic materials. Tabulated polarizabilities are available in the literature for a wide

variety of ions (Shannon, 1993).

ε − 1 4πα m
=
ε + 2 3Vm
The Clausius-Mossotti equation does not predict the occurrence of extremely high

dielectric constants found in ferroelectrics. BaTiO3 is a well known example of a

ferroelectric compound and has ε > 1000. Ferroelectric compounds exhibit a spontaneous

polarization of ions at T < Tc. Ferroelectrics typically exhibit a dielectric constant peak as

a function of temperature. Relaxor ferroelectric compounds exhibit a diffuse and

dispersive phase transition over a wide temperature range (Randall & Bhalla, 1990). A

181
large increase in the dielectric constant with an increase in temperature is an indication

typical of a relaxor ferroelectric.

7.2 Known Dielectric Perovskite Analysis

Many studies on various complex perovskite oxides have been performed to

improve the understanding of the structural features on the properties. Dielectric

materials, particularly those used in the microwave frequency applications, are generally

grouped into three categories (Liu & Wu, 2001) containing: (1) high ε (80-100) (Q = <

7000), e.g. BaO-A2O3-TiO2 (A = La, Nd, etc.), (2) high-Q (Q × f (GHz) =190000-

350000, ε = 25-30), e.g. Ba(M1/3Ta2/3)O3 (M = Mg, Zn) and (3) compounds with

intermediate properties (ε =~40, Q × f (GHz) =7400-50000) e.g. Ba2Ti9O20 and

(Sn,Zr)TiO4. High-Q dielectric perovskites formulated from mixtures of ~89%

Ba(Zn1/3Ta2/3)O3, 7.5% Ba(Ni1/3Ta2/3)O3 and 3.5% BaZrO3 have ε = 29, Q > 12000 and τ

= 0 are used commercially (Davies et al., 1997). Solid solutions of CaTiO3-NdAlO3 are

also in use as microwave resonators for third generation mobile phone bases stations

(Hughes et al., 2001). Materials with a different Q and f0 are applied in different

frequency areas of microwave applications. Tantalum containing compounds are often

used for high Q applications where selectivity to a given frequency is important. The

drawback is that Ta2O5 is expensive and cheaper transition metals (i.e. niobium or

titanium) with equivalent Q and ε are preferred. Similar perovskites containing niobium

have been investigated, e.g. Ba(Mg1/3Nb2/3)O3-BaZrO3, however show higher losses

(reduced Q) than the tantalum counterpart (Akbas & Davies, 1998).

182
 A drawback in many perovskite compositions is the τε often significantly differs

from 0 ppm K-1 in compounds with a high relative permittivity. One example of fine

tuning the τε from small negative to small positive value was demonstrated in Sr3Zn1-

xMgxNb2O9, by changing the ratio of Zn to Mg (Thirumal et al., 2002). Solid solutions of

CaTiO3-NdAlO3 displays a Q = 16000 at 2.7 GHz with a ε = 45 and SrTiO3-LaAlO3

displays a Q = 35000 at 2 GHz with ε = 39 (Reaney et al., 2001). A positive τε is

displayed in LaAlO3, while the Ca, Sr-titanate perovskites display a large negative τε.

Solid solutions of CaTiO3-NdAlO3 and SrTiO3-LaAlO3 results in the capability to tune

the τε through zero. Empirical control of the τε is often achieved by forming solid

solutions and mixed phases with two or more compounds with opposites sign of τε

values.

Examination of τε for a complex perovskite series containing calcium, strontium

and barium as the A-site cation was investigated as a function of octahedral tilting

(tolerance factor) (Colla et al. 1993; Reaney et al., 1994; Park et al., 2001). Barium

containing complex perovskites typically exhibit a negative τε, whereas to calcium and

strontium analogues have a more positive τε. Occurrence of octahedral tilting and

resulting onset of structural phase transitions can cause a change in the sign of the τε from

negative to positive upon cooling or modification of the tolerance factor (Reaney et al.,

2001). Increased octahedral tilting (increasing calcium content) resulted in the τf

gradually changing from positive (120 ppm/ºC) to negative (-40 ppm/ºC) in

[(Pb1-xCax)1/2La1/2](Mg1/2Nb1/2)O3 (Liu & Wu, 2001). Tuning of the τf of a material with a

183
usable ε and tan δ is typically accomplished by formation of a solid solutions with

another material having opposite sign of τf.

Fully ordered compositions of Ba(M2+1/3M5+2/3)O3 (M = Mg, Zn) are difficult to

obtain because the cation ordering induced domain boundaries and the long anneal times

often required for complete ordering. A decrease in dielectric loss often accompanies an

increase in the octahedral cation order. Increased octahedral cation ordering was found to

be important in obtaining the high Q factor for A(M2+1/3M5+2/3)O3 (A = Ca, Sr, Ba; M2+ =

Mg, Zn, Co,Ni; M5+ = Nb, Ta) (Nomura, 1983; Molodetsky & Davies, 2001). Consider

one example, Ba(Zn1/3Ta2/3)O3, in which the Q increases from ~1000 to 14000 at 12 GHz

after extended heating (120 h) at 1350 °C to increase octahedral cation ordering

(Kawashima et al., 1983).

Additives are empirically used to modify the dielectric properties or reduce the

synthetic time required for cation ordering. Addition of small relative molar amounts of

BaZrO3 to Ba(Zn1/3Ta2/3)O3 resulted in shorter annealing times without a significant

degradation of the dielectric properties (Tamura et al., 1984; Davies & Tong, 1997).

Addition of small levels of dopants, (i.e. BaZrO3), does not universally improve the

dielectric properties. Destabilization of the 2:1 order and formation of a 1:1 ordered phase

was observed with the addition of BaZrO3 to Ba(Co1/3Nb2/3)O3 (Molodetsky & Davies,

2001). Small amounts of TiO2 have been added to Ca2/5Sm2/5TiO3-Li1/2Ln1/2TiO3 in an

attempt to the tune the τf to zero (Kim et al., 1999). The use of additives is an empirical

process and is difficult to predict the effects a priori.

184
 Synthetic temperature, duration, heating and cooling rates influence a material’s

microstructure, which is a very important influence on the dielectric properties.

Conventional heat processing results in porous pellets of larger grain size and lower

permittivity values (Takeuchi et al., 1999). Polycrystalline materials often display

frequency dependent effects associated with grain boundaries or surface layers in

combination with the intrinsic properties of the materials (Hirose & West, 1996). Most

ceramic materials have a capacitance of the grain boundary that is larger than the grain

component. Long sintering times were effective in increasing the density in many

compositions e.g. Sr(Sm1/2Ta1/2)O3 (Takata & Kageyama, 1989). Prolonged sintering at

elevated temperatures can lead to a decrease in the dielectric constant as observed in the

solid solution (La1/2Na1/2)TiO3-Ca(Fe1/2Nb1/2)O3 (Kim et al., 1997). Sintering in air at

high temperatures may be reducing to some metals. Annealing in an O2 atmosphere can

reoxidize reduced metals (e.g. Ti3+ to Ti4+). Efforts are made to obtain the least amount of

porosity in the synthesis. Sintering temperature dramatically influences the porosity.

Sintering in a temperature range from 1000 to 1650 °C resulted in alumina samples with

a porosity range from 46 to 1.5%, which shows the large effect sintering temperature has

on porosity (Steil et al., 1997).

Reported dielectric properties from sample to sample vary due to variation in the

microstructure. For example, the ATiO3 (A = Ca, Sr, Ba) series displays differences from

sample to sample are often attributed to grain size, porosity and grain boundary

impedances (Hirose & West, 1996). CaTiO3 displays a ε =162-170 (Wise et al., 2001;

185
Cho et al., 1999), SrTiO3 ε =190-290 (Wise et al., 2001; Cho et al., 1999) and BaTiO3 ε

= 1070-3500 (Uchina et al., 1980; Takeuchi et al., 1999)

Twenty-four 1:1 combinations of octahedral cations exhibited dielectric constants

in the range 14-42 with τf ranging from -75 ppm K-1 to +90 ppm K-1 (Takata &

Kageyama, 1989). Structural details were not provided and multiple phases were

reported, which tends to obscure the inherent character of the dielectric properties of the

perovskites. Sr2GaTaO6 (density = 96% of theoretical) displayed a τε of 120 ppm K-1 with

ε = 27 at 1 MHz (Takahashi et al., 1997). Perovskites with a 1:1 ordering of octahedral

cations are expected to display dielectric constants in the range (14-42). Synthesis and

dielectric characterization of novel 1:1 octahedral cation ordered perovskites is reported

in section 7.4.

7.3 Impedance Analysis Sample Preparation

Preparation of cylindrical pellets for dielectric constant measurements requires a

well defined pellet size. Ideally the density of the pellet approaches the theoretical

density. Initially the composition of interest is synthesized and purity checked by lab x-

ray diffraction. Finely ground powders of the sample are added to a stainless steel die of

diameter 0.5” and uniaxially pressed (Carver Laboratory Press – Model C) into pellets at

a pressure of between 4 and 6 tons for duration of 1 minute. Pellets that do not retain

cylindrical shape on removal from the die are reground with addition of a binding agent.

Polyethylene glycol (Union Carbide, 300 NF, FCC Grade) was added (1-2 drops) as a

binder and mixed with the powders the pressed again. Sintering of the pellets is done

186
under the same conditions as the initial synthesis of the compositions and fully

decomposes the organic polyethylene glycol component.

Determination of the volume (V) and mass (m) of the pellet is used to calculate

the measured density (ρm = m V-1) of the pellet in g cm-3. Theoretical density of the

crystal (ρc) is calculated according to the equation below, where Z is the number of

formula units per unit cell, FW is the formula weight, Na is Avogadro’s number and Vuc

is the unit cell volume.

Z × FW
ρc =
N a × Vuc

Accurate crystal structure information is necessary to calculate an accurate theoretical

density. The percent of theoretical density is defined as %ρt =100 ρm/ρc, where ρm is the

measured density and ρc is the calculated theoretical density. Density can also be

determined by the Archimedes method, but this technique was not applied in this study.

Sintering temperature can be increased slightly to obtain more dense pellets, as long as

the material retains the same crystal structure.

Fine grit SiC sandpaper was used to polish opposite circular faces of the pellet.

The surface of the circular faces of the pellet was made conductive by application of

silver paint or gold sputtering. PTFE thread sealant tape was wrapped around to cover the

sides of the pellet. The PTFE tape prevents the sputtered gold from adhering to the sides

and short circuiting the sample. Gold sputtering was performed with a 90 second coating

for each side of the pellet.

187
7.4 Impedance Analysis Measurement Techniques

One measurement technique is unable to measure the entire frequency range;

however, advances in impedance spectroscopy instrumentation have enabled

measurement from low frequency (µHz) to the low MHz range. A modified Courtney

holder (TE011 mode) resonant cavity method is one accurate method used for microwave

(1-10 GHz range) measurement of permittivity and loss (Krupka, 1988). Low loss

materials, where tan δ ≤ 0.005, require the use of precisely machined samples with

application of resonant methods for accurate determination of loss (Baker-Jarvis et al.,

1998). Dielectric measurements were performed using a Solartron SI 1260

impedance/gain-phase analyzer enabling the collection of data in the sub-microwave

frequency range 10 × 10-6 to 32 × 106 Hz. The Solartron SI 1260 impedance/gain-phase

analyzer is relatively low cost with easier sample preparation relative to analyzers in the

microwave region. Accurate information regarding the ε, τf, and slightly less accurate

data in tan δ (for low loss materials) at sub-microwave frequencies can be obtained

through the use of Solartron SI 1260 impedance/gain-phase analyzer. Measurements use

a parallel plate capacitor arrangement with a dielectric material placed between two metal

electrodes, which contact opposite faces of the pellet.

A room temperature dielectric sample holder was devised with the gold coated

sample pellet placed between two parallel metal plates. Spring tension holds the

electrodes to the sample surface. Alligator clips soldered to the BNC connectors are

clipped to wire leads from the electrodes. BNC connection of the Solartron SI 1260

connects to the intermediate sample holder.

188
 A high-temperature sample holder (up to 700 °C) consists of an alumina rod

inside of a capped alumina tube. Two metal disk electrodes are placed on the rod and

cylinder with the dielectric sample placed between the two. Sample to electrode contact is

provided by the weight of the outside tube. The BNC connectors of the Solartron SI 1260

were plugged into the bottom of the high temperature sample holder and wires connect to

the electrodes.

Data collected was taken as a function of frequency from 1 Hz to 10 MHz with

AC amplitude of 500 mV. The capacitance at a frequency of 100 kHz of various elements

of the sample holders is provided in the table 7.1 for reference.

189
 Cell C tan δ
Solartron SI 1260 Leads Only 2.1 E-14 0.02
Intermediate Cell Holder (Clips) 1.7 E-13 0.03
Room Temperature Cell 1.1 E-12 0.02
High Temperature Cell 1.5 E-12 0.009

Table 7.1 : Comparison of measured capacitance (C) and loss (tan δ) at 100 kHz in
different components of the impedance sample holder apparatus.

If the impedance of the cell was similar to the open or short circuit spectra of test

samples, nulling files were necessary to obtain accurate data. Samples with a small

surface area to volume and low measured capacitance values near that of the sample

holder required nulling files to eliminate the contribution from the sample holder. A new

nulling file is needed in the instrument setup prior to performing each type impedance

experiment. Data quality should be high and noise free with as much of the apparatus left

in place as possible. A closed circuit nulling file was created using a Cu metal disk

sputter coated with Au. An open circuit nulling file was created with the electrodes

spaced apart the same distance as when containing a sample. Variable temperature

measurements require an open circuit nulling file for each temperature range. Nulling

files in approximately five degree increments were made in the temperature range 20-110

°C. A data peak in loss shifts to ~100 kHz at temperatures >110°C, thus variable

temperature measurements were made within the aforementioned range. One closed

circuit nulling file was created and has little effect of the measurement for the highly

insulating samples in this study. Nulling files are important in accurate loss value
190
determinations. However, the capacitance versus frequency spectrum shows increased

noise associated with lower measured capacitances.

Several outside factors influence the impedance analysis experiment. Inductance

from wires in the heating elements surrounding the high temperature cell holder was

found to have an effect on the impedance measurement. Capacitance values decreased

with the furnace on and current flowing. Depressed capacitance or negative capacitance

values signify a dominant contribution of an inductive component in the measurements.

Below 140 ºC humidity induces a significant conductivity in porous samples of

yttria-stabilized zirconia (Steil et al., 1997). Humidity and condensation from H2O, with τ

= 80, have potential to influence the data. Humidity is more of an influence in ionic

conductors and not expected to influence the insulating perovskites. Control experiments

under flowing argon yielded no noticeable change in the impedance spectrum and data

for all samples were collected in air.

Fitting of the AC impedance spectra may require multiple circuit elements for an

accurate model. La-doped BaTiO3 required the use of up to three circuit elements

(Markovec et al., 2001). There is no general method for assigning the correct equivalent

circuit to an impedance spectrum and it is always possible to find multiple equivalent

circuits that fit the experimental data (Hirose & West, 1996). The physical circuit setup of

the dielectric sample holder is a simple model consisting of a resistor and capacitor in

parallel. A capacitor and resistor in series was used in the examination of the dielectric

properties of large dielectric constant perovskite CaCu3Ti4O12 and isostructural

compounds (Ramirez et al., 2000).

191
 A porosity (P) correction up to P = 0.05 can be made using the formula ε = ε

m(1+1.5P), where and ε m is the dielectric constant obtained from the measured

capacitance (Bosman & Havinga, 1963). The Bruggeman effective medium formulation

for a two phase composite has been shown to have application in perovskite systems

(Geyer et al., 1994).

2(ε meas
'
) 2 (ε 'pore ((tan δ meas ) 2 − 2 tan δ meas tan δ pore − 1) − FMε meas
'
)
ε =
'
s
FM 2 (ε meas
'
) 2 + 2UMε meas
'
ε 'pore + V (ε 'pore ) 2

FMε meas
'
tan δ meas + ε 'pore ((tan δ meas ) 2 tan δ pore + 2 tan δ meas − tan δ pore )
tan δ s =
FM (ε meas
'
) 2 − ε 'pore (((tan δ meas ) 2 − 2 tan δ meas tan δ pore − 1)

where M = 3fpore – 2, F = (tan δmeas)2 +1, U = tan δmeas tan δpore + 1, and V = (tan δpore)2 +

1. Air was assumed to be the pore fluid with with ε’pore = 1.000585 and tan δpore = 1E-5.

Application of the Bruggeman effective medium formulation was successful in achieving

dielectric constant values of a 78% dense pellet of LaAlO3 comparable to those of a

single crystal (Geyer et al., 1998) and also for porosity correction in Ba2Fe2Ti4O12

(Vanderah et al., 1995). Dielectric constant and loss corrections for samples with a higher

porosity than 22% were not reported in the literature.

7.5 Instrument Calibration

Compounds of BaTiO3 (1 kHz, 97%, Takeuchi et al., 1999), Sr2AlTaO6 (10 kHz,

77 °C, Guo et al. 1995), CaCu3Ti4O12 (100 kHz, Subramanian et al., 2000) and Al2O3 (1

kHz, Biley, 1997) have reported dielectric constants spanning a large range from ε = 9 to

greater than 10,000. Sintered pellets were prepared for use in a determination of the

192
accuracy the dielectric measurements of the Solartron instrument and sample holders.

Relative permittivity calculated using the Clausius-Mossotti equation (εcm), literature

(εlit), experimental (εm) and the Bruggeman porosity corrected experimental (εmp) values

are compared in the table 7.2. Porosity corrected dielectric constants agree well with

literature values. The porosity correction had been applied to a level of porosity that is

higher than previously reported systems and results a dielectric constant near that

reported for nearly fully dense pellets.

Compound %ρt C tan δ εcm εlit εm εmp

Al2O3 100 4.9E-12 0.06 10 9 6 6
Sr2AlTaO6 56 8.9E-12 0.14 38 12 9 31
BaTiO3 63 4.2E-10 0.04 - 3500 899 2041
CaCu3Ti4O12 70 3.6E-09 0.13 48 10286 5246 9555

Table 7.2 : Theoretical density (%ρt), capacitance (C), dielectric loss (tan δ), Clausius-
Mossotti equation relative permittivity (εcm), literature (εlit), experimentally measured (εm)
and porosity corrected (εmp) for compounds with a wide range of known dielectric
properties.

7.6 Prediction of dual ordered perovskites

An extremely large dielectric constant (ε = ~12000 at 1 kHz; ε = 10286 at 100

kHz) that is nearly constant from room temperature to 300 °C was observed in

CaCu3Ti4O12 (Subramanian et al., 2000). The giant dielectric constant displayed was not

found to be the result of underlying ferroelectric or relaxor behavior (Subramanian &

Sleight, 2002). Twin boundaries acting in a manner to create a barrier layer capacitance

were hypothesized as a possible explanation for the exceptional dielectric properties

193
displayed by CaCu3Ti4O12 (Subramanian et al., 2000). Typically elaborate processing is

required to create barrier layer capacitance, which is known to produce large dielectric

constants of ~20,000. The properties observed in CaCu3Ti4O12 do not match well with

materials exhibiting barrier layer capacitance (Subramanian & Sleight, 2002). Analysis of

the structure did not answer if the unusual properties were intrinsic or extrinsic, or a

combination of the two. Examination of the isostructural composition CaCu3Ge4O12

shows a dielectric constant of 34 at 100 kHz (Subramanian & Sleight, 2002). Appearance

of twin boundaries in single crystal studies of CaCu3Ti4O12 was a structural aspect not

present in CaCu3Ge4O12 .

In the preceding sections octahedral cation order was shown to influence the

dielectric properties of A(M2+1/3M5+2/3)O3 (A = Ca, Sr, Ba; M2+ = Mg, Zn, Co,Ni; M5+ =

Nb, Ta) perovskite compositions. The composition CaCu3Ti4O12 crystallizes in the a+a+a+

Glazer tilt system. Replacement of Ti with an equal ratio of two different cations on the

octahedral site leads to the possibility the same tilt system with octahedral cation

ordering. Ordering of octehedral cations in the a+a+a+ Glazer tilt system changes the

space group symmetry from Im 3 to Pn 3 . Synthesis and impedance analysis of multiple

octahedral cation compositions is of notable interest to ascertain if dielectric properties

similar to CaCu3Ti4O12 are present in octahedral cation ordered compositions. Different

octhahedral cation compositions enable changing of the octahedral tilting, which may

lead to a potential tuning of the dielectric properties.

SPuDS was used to predict the stability of compositions with 1:3 A-site and 1:1

octahedral cation ordering and results of the most promising compositions are shown in

194
section 4.7. Four compositions CaCu3Ga2Sb2O12, CaCu3Cr2Sb2O12, CaCu3Ga2Nb2O12 and

CaCu3Ga2Ta2O12 were chosen for high-pressure high-temperature synthesis.

7.7 Synthesis and Structure Refinement of Dual A-site and Octahedral Cation Ordered
Perovskites

Conventional solid state techniques at ambient pressure attempts failed in the

synthesis of the aforementioned perovskites. A mixture of stable phases formed instead

of a single ordered perovskite phase. The a+a+a+ tilt system displays an efficient and high

packing density for the CaCu3M4O12 perovskites, thus high-pressure high-temperature

synthesis was predicted to stabilize CaCu3M4O12 (M = Ge, Mn, Ru). High-pressure high-

temperature synthesis techniques were carried out by Dr. Song-Ho Byeon in the Dr. John

Parise group at the Center for High Pressure Research (CHiPR) at the State University of

New York - Stony Brook. These efforts led to the synthesis of cylindrical pellets of

diameter 2.7 mm and thickness between 1.3 and 2.2 mm for compositions of

CaCu3Ga2Sb2O12, CaCu3Cr2Sb2O12, CaCu3Ga2Nb2O12 and CaCu3Ga2Ta2O12. Pressure

ranged from 10-12 GPa at temperatures of 1100 °C for a synthesis duration of 1 to 3

hours.

Structural information from SPuDS was used as input in Poudrix (Laugier, 2002)

to simulate a diffraction pattern for CaCu3Ga2Cr2O12 using Cu Kα1 radiation. The

simulated diffraction is shown in figure 7.1 (top). The simulated diffraction patterns were

useful in determining if synthesis of the desired perovskite phase was accomplished.

195
 250

200

150
Counts

100

50

0
10 20 30 40 50 60 70 80 90
2Theta

250

200

150
Counts

100

50

0
10 20 30 40 50 60 70 80 90
2Theta

Figure 7.1 : SPuDS simulated (top) and experimental (bottom) x-ray powder diffraction
pattern of CaCu3Cr2Sb2O12.

196
 The diffraction pattern for CaCu3Ga2Cr2O12 was collected on a General Area

Detector Diffraction System (GADDS) by Parise and coworkers and shown figure 7.1

(bottom). Note the appearance of impurities in the experimental pattern, which limited

the capability to obtain meaningful structure results from the refinements. The perovskite

phase (Pn 3 ) is the main phase present. Selecting the impurity peaks from the GADDS

data was undertaken and attempts at impurity phase identification were unsuccessful. The

small sample size prevented collection of additional x-ray diffraction data. The dielectric

property analysis required CaCu3Ga2Cr2O12 to retain its cylindrical shape.

A small amount of sample was obtained by sanding the faces of the pellet of

CaCu3Ga2Sb2O12, CaCu3Ga2Nb2O12 and CaCu3Ga2Ta2O12 with fine grit SiC sandpaper.

X-ray diffraction data was collected on a Bruker D8 Advance Diffractometer using a zero

background Si sample holder. An automated search/match of the x-ray diffraction

patterns using the software program Eva version 5.0 (Bruker AXS, 2000) to the Powder

Diffraction File (PDF) (Faber & Fawcett, 2002) resulted in zero matches. A manual

search of the impurity peaks to diffraction patterns in the PDF containing any of the

elements in the composition was attempted and resulted in zero matches. Compounds

synthesized under high-pressure high-temperature conditions often do not match up well

with the PDF database. Indexing of the impurity peaks was undertaken to attempt to find

a unit cell of one of the impurity phases. The weak intensity and low number (<10) of

peaks resulted in an unsuccessful determination of the lattice parameters for the impurity

phase(s). Four cations and oxygen leads to the possible presence of binary, ternary,

197
and/or quaternary or mixture of impurity phases. Multiple impurity phases are possible,

which would require the determination of which impurity peaks belong to which phase.

Attempts at multiple impurity phase analysis were not successful. Peaks not indexed in

the diffraction pattern indicate impurities are present in the as synthesized compositions.

Rietveld refinements x-ray diffraction data were undertaken using the Topas

software (Bruker AXS, 2000). Regions corresponding to relatively strong peaks from

impurity phases were excluded from the refinements. The LP factor was fixed with a

value of 78 and the the oxygen atoms were fixed for refinements in Pn 3 at the SPuDS

predicted fractional coordinates [CaCu3Ga2Sb2O12 (0.2484, 0.4278, 0.5560),

CaCu3Ga2Ta2O12 (0.2499, 0.4291, 0.5552)] due to refined positions giving unreasonable

bond lengths and angles. The oxygen positions were refined for CaCu3Ga2Nb2O12.

Experimental data points, Rietveld calculated fits, and difference curves are shown in

figures 7.2 (CaCu3Ga2Sb2O12), 7.3 (CaCu3Ga2Ta2O12), and 7.4 (CaCu3Ga2Nb2O12). The

results of the Rietveld refinements are tabulated in table 7.3 (CaCu3Ga2Sb2O12 and

CaCu3Ga2Ta2O12), and table 7.4 (CaCu3Ga2Nb2O12). The three compositions displayed

different degrees of octahedral cation order. According to the refinements

CaCu3Ga2Sb2O12 compound is the most ordered, while CaCu3Ga2Ta2O12 is slightly less

ordered. A significant concentration of Ga/Ta antiphase boundaries are manifested in the

broadening of the peaks associated with ordering in GADDS data (Woodward, 2002).

The CaCu3Ga2Nb2O12 composition appears to have the least long range order of Ga and

Nb and refinement indicated disordered Ga-Nb on a single crystallographic site.

198
Refinement of CaCu3Ga2Nb2O12 was undertaken in space group Im 3 with disordered

octahedral cations.

One indication of ordering of the octahedral cations is the intensity of the (111)

peak. The absence of the (111) peak in CaCu3Ga2Nb2O12 indicates the disorder of the

octahedral cations. Caution must be taken when examining the intensity of the (111) peak

as an indication of octahedral cation ordering. The intensity is related to the difference in

electron density of the two ions on the M-site. The combination of Ga3+-Nb5+ has a

difference of only 8 electrons, whereas Ga3+-Sb5+ differ by 18 electrons, and Ga3+-Ta5+

has a difference of 40 electrons formally. A larger difference in electron density results in

a more intense (111) peak.

199
 CaCu2Ga2Ta2O12 100.00 %

70
60
50
2Th Degrees
40
30
20
1,200
1,100
1,000
900
800
700
600
500
400
300
200
100
0
-100
-200
-300

Counts

Figure 7.2 : Observed (dots) and calculated (lines) of the Rietveld refinement of
CaCu3Ga2Ta2O12 (space group Pn 3 ). Regions of the pattern containing impurity peaks
are excluded. At the bottom of the figure the difference plot (Iobs – Icalc) is shown.

200
 CaCu3Ga2Sb2O12 100.00 %

90
80
70
2Th Degrees
60
50
40
30
20
4,000

3,500

3,000

2,500

2,000

1,500

1,000

500

-500

Counts

Figure 7.3 : Observed (dots) and calculated (lines) of the Rietveld refinement of
CaCu3Ga2Sb2O12 (space group Pn 3 ). Regions of the pattern containing impurity peaks
are excluded. At the bottom of the figure the difference plot (Iobs – Icalc) is shown.

201
 CaCu3Ga2Nb2O12 100.00 %

70
60
50
2Th Degrees
40
30
20
2,500

2,000

1,500

1,000

500

-500

Counts

Figure 7.4 : Observed (dots) and calculated (lines) of the Rietveld refinement of
CaCu3Ga2Nb2O12 (space group (Im 3 ). Regions of the pattern containing impurity peaks
are excluded. At the bottom of the figure the difference plot (Iobs – Icalc) is shown.

202
 CaCu3Ga2Sb2O12 CaCu3Ga2Ta2O12
Background (Chebychev) 5th order 6th order
Sample disp. (mm) -0.077(2) -0.127(5)
Excluded Regions 27.2-27.6, 30-31,
35.5-35.8, 35.6 - 35.9
59.9-60.1
Scale 0.0000479(4) 0.0000165(3)
Cry Size Lorentzian (nm) 104(4) 90(8)
Strain L 0.1606(9) 0.27(3)
Lattice parameters - a(Å) 7.4481(2) 7.4656(5)
Beq (Ca) 0.36(38) -0.61(97)
Beq (Cu) 2.3(1) 6.5(5)
Beq (Ga) (0,0,0) 2.6(1) 4.5(3)
Beq (M5+) (½, ½, ½) 1.54(6) 4.9(2)
Beq (O) 0.8(2) 1.9(5)
Occ (Ga) (0,0,0) 0.937(7) 0.833(9)
Occ (M5+) (0,0,0) 0.063(7) 0.167(9)
Occ (Ga) (½, ½,½) 0.063(7) 0.167(9)
Occ (M5+) (½,½,½) 0.937(7) 0.833(9)
Rexp 10.31 12.10
Rwp 13.22 14.18
Rp 9.64 10.81
GOF 1.28 1.17

Table 7.3 : Rietveld refinement results of powder x-ray diffraction data for
CaCu3Ga2Sb2O12 and CaCu3Ga2Ta2O12 in space group Pn 3 .

203
 CaCu3Ga2Nb2O12
Background (Chebychev) 5th order
Sample disp. (mm) -0.074(5)
Excluded Regions 14.8-15.4, 26.4-27.8,
29.8-31, 35.5-35.9, 59.9-60.2
Scale 0.000065(1)
Cry Size Lorentzian (nm) 1.2 ×102 (1)
Strain L 0.29(3)
Lattice parameters - a(Å) 7.4700(5)
O(24g) y 0.180(2)
O(24g) z 0.300(2)
Beq (Ca) 1.7(9)
Beq (Cu) 7.5(4)
Beq (Ga, Nb) 6.4(2)
Beq (O) 3.1(3)
Rexp 9.39
Rwp 14.57
Rp 10.28
GOF 1.5

Table 7.4 : Rietveld refinement results of x-ray data for CaCu3Ga2Nb2O12 in space group
Im 3 .

204
7.8 Impedance Analysis of CaCu3M2M′2O12

Pellets were prepared for impedance analysis by the procedure described in

section 7.3. Impedance analysis was performed in the frequency range 10 Hz to 10 MHz

in the room temperature and high temperature sample holder. Impedance plane plots Z”

vs Z are shown in the figure 7.5. A partial semicircle is observed in the data, which leads

to multiple equivalent circuits that fit the spectra. The assumption that the loss component

of the impedance and the capacitive part are in parallel was used following the analysis of

Ramirez et al. (2000). Variation of the capacitance (C) and dielectric loss (D) with

frequency are shown in the figure 7.6.

Impedance analysis was performed at room and temperature up to approximately

100 ºC in order to determine the τc. Nulling files were employed in the dielectric

measurement. The capacitance versus temperature at a fixed frequency (100 kHz) is

shown in the figure 7.7.

205
 -3.0E+07

-2.5E+07

-2.0E+07

CaCu3Ga2Nb2O12 and CaCu3Ga2Ta2O12.

-1.5E+07

Z"
CaCu3Ga2Nb2O12
CaCu3Cr2Sb2O12

206
-1.0E+07 CaCu3Ga2Ta2O12
CaCu3Ga2Sb2O12

-5.0E+06

0.0E+00
0.0E+00 5.0E+06 1.0E+07 1.5E+07 2.0E+07 2.5E+07 3.0E+07

Z'

Figure 7.5 : Impedance analysis Z” vs. Z’ for CaCu3Ga2Sb2O12, CaCu3Cr2Sb2O12,

 CaCu3Ga2Nb2O12
CaCu3Cr2Sb2O12
CaCu3Ga2Ta2O12
CaCu3Ga2Sb2O12
-10
10

-11

C10
-12
10

-13
10
3 4 5 6 7
10 10 10 10 10
Frequency (Hz)

0.5
D
1.0

1.5
3 4 5 6 7
10 10 10 10 10
Frequency (Hz)

Figure 7.6 : Capacitance (C) (farads) vs. frequency and loss (D) vs. frequency for
CaCu3Ga2Sb2O12, CaCu3Cr2Sb2O12, CaCu3Ga2Nb2O12 and CaCu3Ga2Ta2O12

207
 100
90
80
70
Temperature (C)
60
50
CaCu3Ga2Nb2O12

CaCu3Ga2Sb2O12
CaCu3Ga2Ta2O12
CaCu3Cr2Sb2O12

40
30
20
1.4E-11

1.2E-11

1.0E-11

8.0E-12

6.0E-12

4.0E-12

2.0E-12

0.0E+00

Capacitance

Figure 7.7 : Capacitance versus temperature for CaCu3Ga2Sb2O12, CaCu3Cr2Sb2O12,

CaCu3Ga2Nb2O12 and CaCu3Ga2Ta2O12 perovskites. Capacitance value is obtained at a
frequency of 100 kHz.

208
 Results of the impedance analysis are shown in table 7.5. Pellets had a density of

greater than 89% for each sample. Measured capacitance values were quite small with

two of the samples having a capacitance less that 0.9 ×10-12. The measured capacitance of

the sample holders are reported in table 7.5 and the sample holders have a capacitance

greater than the capacitance of CaCu3Ga2M2O12 (M=Sb, Ta) and near the value of

CaCu3Cr2Sb2O12 and CaCu3Ga2Nb2O12. Measured capacitance values similar to that

induced by the sample holder make the nulling files critical in the impedance analysis.

Error in the impedance analysis is high and the values of the capacitance should be

regarded with caution due to the loss of sensitivity when the nulling file becomes very

important. The τc values should also be regarded cautiously due to the experimental

dependence of the nulling file. Nonetheless, these data clearly show the dielectric

constants of the four compounds shown in table 7.5 are of a much smaller magnitude than

CaCu3Ti4O12.

Compound %ρt C (F) tan δ εcm εm εmp τc (ppm K-1)

CaCu3Ga2Sb2O12 93 0.3E-12 0.04 26 11 12 3.7 ×102
CaCu3Cr2Sb2O12 90 1.2E-12 0.26 27 40 46 4.7 ×103
CaCu3Ga2Nb2O12 94 3.8E-12 0.86 25 169 183 6.0 ×103
CaCu3Ga2Ta2O12 89 0.9E-12 0.10 33 23 26 3.0 ×104

Table 7.5 : Percent of theoretical density (%ρt), capacitance (C), loss (tan δ), relative
permittivity calculated using the Classius Mossotti equation (εcm), experimental (εm),
porosity corrected experimental (εmp) and temperature dependence of the capacitance (τc)
of CaCu3Ga2Sb2O12, CaCu3Cr2Sb2O12, CaCu3Ga2Nb2O12, and CaCu3Ga2Ta2O12.

209
7.9 Conclusions

SPuDS calculations successfully predicted the stability of dual A and octahedral

cation ordered perovskites in the a+a+a+ tilt system. Compositions in which 1) octahedral

cation order was observed in other perovskite compositions and 2) a low GII were

selected for high-pressure high-temperature synthesis. Novel perovskite compositions

were predicted and structure refinements of x-ray diffraction data indicate the main

phases closely match the simulated diffraction pattern using the SPuDS optimized crystal

structure. Attempts to identify the minor impurity phases were unsuccessful. Refinements

of CaCu3Cr2Sb2O12were unsuccessful due to a large amount of impurity present.

According to the refinements, the CaCu3Ga2Sb2O12 composition has the most ordering of

the octahedral cations, while CaCu3Ga2Ta2O12 has slightly less ordering. Disordered Ga-

Nb on a single crystallographic site is present in CaCu3Ga2Nb2O12.

Impedance analysis was performed on the compounds to measure the dielectric

constant, loss tangent, and temperature coefficient of the dielectric constant. Small

sample diameter and relatively large thickness resulted in measured capacitances that

were near that of the dielectric sample holder. Significant reliance on the nulling files in

the dielectric measurement leads to potentially relatively large errors in the capacitance

values, and consequently also in the calculated dielectric constant. One possibility for the

large dielectric constant observed in CaCu3Ga2Nb2O12 is a contribution from an impurity

phase, possibly ferroelectric or relaxor ferroelectric based on the large value and large

positive τc. Impurities present in the samples, make completely reliable determination of

the dielectric properties ambiguous. Impedance analysis indicates compositions do not

210
display a dielectric constant as large as the one observed for CaCu3Ti4O12. The porosity

corrected dielectric constants for CaCu3Ga2Sb2O12, CaCu3Cr2Sb2O12 and

CaCu3Ga2Ta2O12 (ε=12-46) are near what is expected for typical perovskites based on the

oxide additivity rule.

More accurate dielectric measurements could be obtained by lowering the

inherent capacitance associated with the sample holders. Larger diameter, smaller

thickness pellets result in a larger capacitance value and more accurate dielectric data.

Changes to the high-pressure high-temperature synthesis technique to obtain a different

pellet geometry is not practical. Synthetic attempts at different pressures and

temperatures, along with off stoichiometric synthesis would be beneficial in obtaining

samples with different peak intensities of the impurity phase(s) present and could

possibly assist in determining the composition(s) of the impurity phase(s). An impurity

phase(s) present poses a larger influence on the dielectric constant, particularly in the

case where the impurity is a ferroelectric composition with a large temperature dependent

dielectric constant.

An entirely new octahedral and A-site cation ordered tilt system has been

predicted to exist and the initial compositions were successfully synthesized under high-

pressure high temperature conditions with minor impurities. Many additional

compositions are predicted by SPuDS to be stable under similar high-pressure high-

temperature synthetic conditions. A systematic examination of structural and dielectric

properties of additional compositions will perhaps provide insight into the giant dielectric

211
constant CaCu3Ti4O12 and will richen the understanding the structural chemistry of Pn 3

perovskites.

7.10 References

Akbas, M.A. & Davies, P.K. (1998). J. Am. Ceram. Soc. 81(3), 670-676.

Baker-Jarvis, J., Geyer, R.G., Grosvernor, J.H., Janezic, M.D., Jones, C.J., Riddle, B.,
Weil, C.M. & Krupka, J. (1998). IEEE Trans. Dielectrics and Electrical Insulation 5(4),
571-577.

Bruker AXS, 2000 - User Manual, Bruker AXS, Karlsruhe, Germany.

Faber, J. & Fawcett, T. (2002) Acta Cryst. B58, 325-332.

Geyer, R.G., Mantese, J. & Baker-Jarvis, J. NIST Tech. Note 1371 (1994).

Geyer, R.G, Barker-Jarvis, J., Vanderah, T.A & Mantese, J. “Advances in Dielectric
Ceramic Materials”, Ceramic Transactions, American Ceramic Society, 115-128 (1998).

Guo, R., Bhalla, A.S., Sheen, J., F.W., Ainger, Erdei, S., Subbarao, E.C. & Cross, L.E.
(1995). J. Mater. Res. 10, 18-25.

Han, B., Neumayer, D.A., Goodreau, B.H., Marks, T.J., Zhang, H. & Dravid, V.P.
(1994). Chem. Mater. 6, 18-20.

Hughes H., Iddles, D.M. & Reaney, I.M. (2001). Appl. Phys. Letters 79, 2952-2954.

Irvine, J.T.S., Sinclair, D.C. & West, A.R. (1990). Adv. Mater. 2, 132-138.

Kim, H.J., Kucheiko S., Yoon, S.-J. & Jung, H.-J. (1997). J. Am. Ceram. Soc. 80, 1316-
1318.

Kim, W.S., Kim, E.S. & Yoon, K.H. (1999). Ferroelectrics 223, 277-284.

Krupka, J., Fifth Int. Conference on Dielectric Materials, Measurements and

Applications, Canterbury, UK, 27-30 June 1988, pp. 322-325.

Laugier, J. (2002). Poudrix, http://www.inpg.fr/LMGP.

212
Levin, I, Chan, J.Y., Maslar, J.E. & Vanderah, T.A. (2001). J. Appl. Phys. 90(2), 904-
914.

Liu, C.-L. & Wu, T.B. (2001). J. Am. Ceram. Soc. 84, 1291-1295.

Makovec, D., Ule, N. & Drofenik, M. (2001). J. Am. Ceram. Soc. 84, 1273-1280.

Mathe, V.L., Patankar, K.K., Kothale, M.B., Kulkarni, S.B., Joshi, P.B. & Patil, S.A.
(2002). Pramana - Journal of Physics 58(5-6), 1105-1113.

Moledetsky, I. & Davies, P.K. (2001). J. Eur. Ceram. Soc. 21, 2587-2591.

Reaney, I.M. & Ubic, R. (2000). Int. Ceram. 1, 48-52.

Nagai, T., Inuzuka, T. & Sugiyama, M. (1992). Jpn. J. Appl. Phys. 31, 3132-3135.

Nomura, S. (1983). Ferroelectrics 49, 61-70.

Park, H.S., Yoon, K.YH. & Kim, E.S. (2001). J. Am. Ceram. Soc. 84, 99-103.

Penn, S.J., Alford, N.M., Templeton, A., Wang, X., Xu, M., Reece, M. & Scrapel, K.
(1997). J. Am. Ceram. Soc. 80, 1885-1888.

Price, T. (2001) Philos. Mag. A81(2), 501-510.

Ramirez, A.P., Subramanian, M.A., Gardel, M.,Blumberg,G., Li,D.,Vogt,T. & Shapiro,

S.M. (2000). Solid St. Commun. 115, 217-220.

Randall, C.A. & Bhalla, A.S. (1990). Jpn. J. Appl. Phys. 29(2), 327-333.

Reaney, I.M., Colla, E.L. & Setter, N. (1994). Jpn. J. Appl. Phys. 33, 3984-3990.

Reaney, I.M., Wise, P., Ubic, R., Breeze, J., Alford, McN, N., Iddles, D., Cannell, D. &
Price, T. (2001). Philos. Mag A81(2), 501-510.

Shannon, R.D. (1993). J. Appl. Phys. 73(1), 348-366.

Steil, M.C., Thevenot, F. & Kleitz, M. (1997). J. Electrochem. Soc. 144, 390-398.

Subramanian, M. A. & Sleight, A.W. (2002). Solid State Sciences 4, 347-351.

Takahashi, J., Kageyama, K.,Fulii, T., Yamada, T. & Kodaira (1997). J. Mater. Sci. 8,
79-84.

213
Takata, M. & Kageyama, K. (1989). J. Am. Ceram. Soc. 72, 1955-1959.

Takeuchi, T., Betourne, E., Tabuchi, M., Kageyama, H., Kobayashi, Y., Coats, A.,
Morrison, F., Sinclair, D.C. & West, A.R. (1999). J. Mater. Sci. 34, 917-924.

Tamura, H., Konoike, T., Sakabe, Y. & Wakino, K. (1984). Comm. Am. Ceram. Soc. , C-
59-C-61.

Thirumal, M., Jawahar, I.N., Sureddiran, K.P., Mohanan, P. & Ganguli, A.K. (2002).
Mater. Res. Bull. 37, 185-191.

Vanderah, T.A., Huang, Q., Wong-Ng, W., Chakoumakos, B.C., Goldfarb, R.B., Geyer,
R.G., Baker-Jarvis, J., Roth, R.S. & Santoro, A. (1995). J. Solid State Chem. 120, 121-
127.

Wise, P.L., Reaney, I.M., Lee, W.E., Price, T.J., Iddles, D.M. & Cannell, D.S. (2001). J.
Eur. Ceram. Soc. 21, 2629-2632.

Woodward, P.M. (2002) Private communication.

214
 BIBLIOGRAPHY

Abakumov, A.M., Shpanchenko, R.V., Antipov, E.V., Lebedev, O.I. & Van Tendeloo, G.
(1997). J. Solid State Chem. 131, 305-309.

Abrahams, S.C. & Bernstein, J. L. (1967). J. Phys. Chem. Solids 28, 1685-1692.

Abrahams, S.C., Hamilton, W.C. & Sequeira, A. (1967). J. Phys. Chem. Solids 28, 1693-
1698.

Abrashev, M.V., Popov, V.N. & Thomsen, C. (1998). J. Phys.: Condens. Matter 10,
1643-1654.

Agrestini, S., Saini, N.L., Lanzara, A., Natali, F. & Bianconi, A. (2000). Int. J. Modern
Phys. B 14(25-27), 2852-2857.

Ahtee, M. & Darlington, C.N.W. (1980). Acta Cryst. B36, 1007-1014.

Ahtee, M., Ahtee, M., Glazer, A.M., & Hewat, A.W. (1976). Acta Cryst. B32, 3243-
3246.

Ahtee, M., Glazer, A.M., & Hewat, A.W. (1978). Acta Cryst. B34, 752-758.

Ahtee, M., Glazer, A.M., & Megaw, H.D. (1972). Philos. Mag. 26, 995-1014.

Aiura, Y., Iga, F., Nishihara, Y., Ohnuki, H. & Kato, H. (1993). Phys. Rev. B. 47(11),
6732-6735.

Akbas, M.A. & Davies, P.K. (1997). J. Mater. Res. 12, 2617-2622.

Akbas, M.A. & Davies, P.K. (1998). J. Am. Ceram. Soc. 81(3), 670-676.

Aleksandrov, K.S. (1976). Kristallografiya 21, 249-255.

Aleksandrov, K.S. & Bartolome, J. (2001). Phase Trans. 74(3), 255-335.

Aleksandrov, K.S. & Beznosikov, B.V. (1999). Ferroelectrics 226, 1-9.

Aleksandrov, K.S. & Beznosikov, B.V. (1999). Ferroelectrics 226, 1-9.

215
Almaer, S.A., Battle, P.D., Lightfoot, P., Mellen, R.S. & Powell, A.V. (1993). J. Solid
State Chem. 102, 375-381.

Alonso, J.A., Cascales, C., Garcia Casado, P. & Rasines, I. (1997). J. Solid State Chem.
128, 247-250.

Alonso, J.A., Martinez-Lope, M.J., Casais, M.T. & Fernandez-Diaz, M.T. (2000). Inorg
Chem. 39, 917-923.

Alonso, J.A., Martinez-Lope, M.J., Casais, M.T., Aranda, A.G. & Fernandez-Diaz, M.T.
(1999). J. Am. Chem. Soc. 121, 4754-4762.

Alonso, J.A., Martinez-Lope, M.J., Casais, M.T., Martinez, J.L., Demazeau, G.,
Largeteau, A., Garcia-Munoz, J.L.,Munoz,A. & Fernandez-Diaz, M.T. (1999). Chem.
Mater. 11, 2463-2469.

Alonso, J.A., Martinez-Lope, M.J., Casais,M.T., Garcia-Munoz,J.L &Fernandez-Diaz,

M.T. (2000). Phys. Rev. B. 61(3), 1756-1763.

Amador, U., Hetherington, C.J.D., Moran, E. & Alario-Franco, M.A. (1992). J. Solid
State Chem. 96, 132-140.

Anan`eva, G.V., Ivanov, A. O.; Merkulyaeva, T.I. & Mochalov, I.V. (1978).
Kristallografiya 23, #1, 200-202.

Anderson, M.T. & Poeppelmeier, K.R. (1991). Chem. Mater. 3, 476-486.

Anderson, M.T., Greenwood, K.B., Taylor, G.A. & Poeppelmeier, K.R. (1993). Prog.
Solid St. Chem. 22, 197-233.

Anderson, P.W. & Hasegawa, H. (1955). Physical Review 100(2), 675-681.

Araya-Rodriguez, E., Ramos, A.Y., Tolentino, H.C.N., Granado, E. & Oseroff, S.B.
(2001). J. Mag. Mag. Mater. 233, 88-90.

Armbruster, T., Rothlisberger, F. & Seifert, F. (1990). American Mineralogist 75, 847-
858.

Arulraj, A., Santhosh, P.N., Gopalan, R.S., Guha, A., Raychaudhuri, A.K., Kumar, N. &
Rao, C.N.R. (1998). J. Phys.: Condens. Matter 10, 8497-8504.

Arulrajy, A., Gundakaramy, R., Biswasz, A., Gayathriz, N., Raychaudhuriz, A.K. & Rao,
C.N.R. (1998). J. Phys.: Condens. Matter 10, 4447-4456.

216
Atou, T., Chiba, H., Ohoyama, K. Yamaguchi, Y. & Syono, Y. (1999). J. Solid State
Chem. 145, 639-642.

Attfield, J.P. (2001). Int. J. Inorg. Mat. 3, 1147-1152.

Attfield, M.P. Ballte, P.D., Bollen, S.K. & Gibb, T.C. (1992). J. Solid State Chem. 100,
37-48.

Azad, A.K., Ericksson, S.-G., Mellergard, A., Ivanov, S.A., Eriksen, J. & Rundlof, H.
(2002). Mater. Res. Bull. 37, 1797-1813.

Azad, A.K., Ivanov, S.A., Eriksson, S.-G., Eriksen, J., Rundlof, H., Mathieu, R. &
Svedlindh, P. (2001). Mat. Res. Bull. 36, 2215-2228.

Azuma, M., Kaimori, S. & Takano, M., (1998). Chem. Mater. 10, 3124-3130.

Bacher, P., Antoine, P.,Marchand, R.,L'Haridon,P.,Laurent,Y. & Roult,G. (1988). J. Solid

State Chem. 77, 67-71.

Baker-Jarvis, J., Geyer, R.G., Grosvernor, J.H., Janezic, M.D., Jones, C.J., Riddle, B.,
Weil, C.M. &Krupka, J. (1998). IEEE T. Dielec. El. In. 5(4), 571-577.

Ball, C., Begg, B.D., Cookson, D., Thorogood, G.J. & Vance, E.R. (1998). J. Solid State
Chem. 139, 238-247.

Barman, A., Ghosh, M., Biswas, S., K.De., S. & Chatterjee (1998). J. Phys.: Condens.
Matter 10, L199-L205.

Bartram, S.F. & Fryxell, R.E. (1970). J. Inorg. Nucl. Chem. 32, 3701-3706.

Battle, P. D. & Jones, C.W. (1989). J. Solid State Chem. 78, 108-116.

Battle, P. D., Gibb, T.C., & Lightfoot, P. (1990). J. Solid State Chem. 84, 271-279.

Battle, P.D. & Macklin, W.J. (1984). J. Solid State Chem. 52, 138-145.

Battle, P.D. & Macklin, W.J. (1984). J. Solid State Chem. 54, 245-250.

Battle, P.D., Claridge, J.B., Copplestone, F.A., Carr, S.W. & Tsang, S.C. (1994). Appl.
Catalysis A: General 118, 217-227.

Battle, P.D., Gibb, T.C., Herod, A. J. & Hodges, J.P. (1995). J. Mater. Chem. 5, 75-78.

217
Battle, P.D., Gibb, T.C., Herod, A., Kim & S.H Munns, P. (1995). J. Mater. Chem. 5,
865-870.

Battle, P.D., Gibb, T.C., Jones, C.W. & Studer, F. (1989). J. Solid State Chem. 78, 281-
293.

Battle, P.D., Goodenough, J.B. & Price, R. (1983). J. Solid State Chem. 46, 234-244.

Battle, P.D., Jones, C.W. & Studer, F. (1991). J. Solid State Chem. 90, 302-312.

Belsky, A., Hellenbrandt, M., Karen, V. L. & Luksch, P. (2002). Acta Cryst. B58, 364-
369.

Bensch, W., Schmalle, H.W. & Reller, A. (1990). Solid State Ionics 43, 171-177.

Bernier, J.C., Chauvel, C. & Kahn, O. (1974). J. Solid State Chem. 11, 265-271.

Beznosikov, B.V. & Aleksandrov, K.S. (2002). J. Struct. Chem. 43(1), 172-175.

Birey, H. (1977). J. Appl. Phys. 48, 5209-5212.

Blasco, J., Ritter, C., Morellon, L., Algarabel, P.A., De Teresa, J.M., Serrate, D., Garcia,
J. & Ibarra, M.R. (2002). Solid State Sci. 4, 651-660.

Blasse, G. (1965). J. Inorg. Nucl. Chem. 27, 993-1003.

Bochu, B., Buevoz, J.L., Cheneva,J.,Collomb,A.,Jourbet,J.C. & Marezio,M. (1980). Solid

St. Comm. 36, 133-138.

Bochu, B., Chenavas, J., Jourbet, J.C. & Marezio, M. (1974). J. Solid State Chem. 11, 88-
93.

Bochu, B., Deschizeaux, M.N., Jourbet, J.C., Collomb, A., Chenavas, J. & Marezio, M.
(1979). J. Solid State Chem. 29, 291-298.

Bock, O. & Muller, U. (2002). Acta Cryst. B58, 594-606.

Bokhimi (1992). Powder Diff. 7, 228-230.

Bokov, A.A., Protsenko, N.P. & Ye, Z.-G. (2000). J. Phys. Chem. Solids 61, 1519-1527.

Booth, C.H., Bridges, F., Kwei, G.H., Lawrence, J.M., Cornelius, A.L. & NeuMeier, J.J.
(1998). Phys. Rev. B. 57(17), 10440-10454.

218
Bordet, P., Chaillout, C., Marezio, M., Huang, Q., Santoro, A., Cheong, S.W., Takagi, H.,
Oglesby, C.S. & Batlogg,B. (1993). J. Solid State Chem. 106, 253-270.

Bosman, A.J. & Havinga, E.E. (1963). Phys. Rev. 129(4), 1593-1600.

Bouchard, R.J. & Gillson, J.L. (1972). Mat. Res. Bull. 7, 873-878.

Boulay, D., Maslen, E.N., Streltsov, V.A. & Ishizawa, N. (1995). Acta Cryst. B51, 921-
929.

Bouloux, J.-C. & Galy,J. (1976). J. Solid State Chem. 16, 385-391.

Brach, B.Y., Bobrysheva, N.P. & Chezhina, N.V. (1982). Zh. Neorg. Khim. 27, 2471-
2473.

Brandle, C.D. & Fretello, V.J. (1990). J. Mater. Res. 5, 2160-2164.

Brandle, C.D. & Steinfink, H. (1970). Inorganic Chem. 10(5), 922-926.

Brauer, v.G. & Kristen, H. (1979). Z. Anorg. Allg. Chem. 456, 41-53.

Brese, N. E. & O'Keeffe, M. (1991). Acta Cryst. B47, 192-197.

Brinks, H., Fjellvag, H. & Kjeksus, A. (1997). J. Solid State Chem. 129, 334-340.

Brixner, L. (1960). J. Inorg. Nucl. Chem. 14, 225-230.

Brixner, L.H. (1960). J. Phys. Chem. Solids 64, 165-166.

Brous, J., Fankuchen, I. & Banks, E. (1953). Acta Cryst. 6, 67-70.

Brown, I. D., Dabkowski, A. & McCleary, A. (1997). Acta Cryst. B53, 750-761.

Brown, I.D (1978). Chem. Soc. Rev. 7, 359-376.

Brown, I.D. (1992). Acta Cryst. B48, 553-572.

Brown, I.D. (1992). Zeit. Fur Kristallographie 199, 255-272.

Brown, I.D. (1997). Acta Cryst. B53, 381-393.

Brown, I.D. & Altermatt, D. (1985). Acta Cryst. B41, 244-247.

Brynestad, J., Yakel, H., & Smith, G.P. (1966). J. Chem. Phys. 45, 4652-4664.

219
Burbank, R.D. (1970). J. Appl. Cryst. 3, 112-120.

Burns, G. & Glazer, A. M. (1990). Space Groups for Solid State Scientists , 2nd ed.,
Appendix A9-6. San Diego: Academic Press.

Burton, B.P. (2000). Modelling Simult. Mater. Sci. Eng. 8, 211-219.

Burton, B.P. & Cockayne, E. (1999). Phys. Rev. B. 60, 12542-12545.

Buttner, R. & Maslen, E. (1988). Acta Cryst. C44, 1707-1709.

Buttner, R.H., Maslen, E.N., Spadaccini, N. (1990). Acta Cryst. B46, 131-138.

Callaghan, A., Moeller, C.W. & Ward, R. (1966). Inorganic Chem. 5(9), 1572-1576.

Capponi, J.J., Chaillout, C., Hewat, A.W., Lejay, P., Marezio, M., Nguyen, N., Raveau,
B., Soubeyroux, J.L., Tholence, J.L. & Tournier, R. (1987). Europhysics Letters 3, 1301-
1307.

Carlson, S., Larsson, A.-K. & Rohrer, F.E. (2000). Acta Cryst. B56, 189-196.

Cava, R.J. (2001). J. Mater. Chem. 11, 54-62.

Cava, R.J., van Dover, R.B., Batlogg, B., Rietman, E.A. (1987). Phys. Rev. Lett. 58(4),
408-410.

Chai, L. & Davies, P.K. (1998). Mat. Res. Bull. 33, 1283-1292.

Chai, L., Akbas, M., Davies, P. & Parise, J. (1997). Mat. Res. Bull. 32(9), 1261-1269.

Chakhmouradian, A.R. & Mitchell, R.H. (1998). J. Solid State Chem. 138, 272-277.

Chamberland, B. & Danielson, P. (1971). J. Solid State Chem. 3, 243-247.

Chamberland, B.L. (1967). Solid State Communcations 5, 663-666.

Chan, H., Chen, T. & Harmer, M. (1988). Powder Diff. 3, 118-119.

Chan, J. Y., Levin, I., Vanderah, T.a., Geyer, R.G. & Roth, R.S. (2000). Int. J. Inorg.
Mat. 2, 107-114.

Chen, B.-H., Walker, D., Scott, B.A. & Mitzi, D.B. (1996). J. Solid State Chem. 121,
498-501.

220
Chen, C.H. & Cheong, S.-W. (1996). Phys. Rev. Lett. 76(21), 4042-4045.

Chen, J., Chan, H. & Harmer, M.P. (1989). J. Am. Ceram. Soc. 72, 593-598.

Chen, Z.-X., Chen, Y. & Jiang, Y.-S. (2002). J. Phys. Chem. B 106, 9986-9992.

Chenavas, J., Jourbet, J. & Marezio, M. (1975). J. Solid State Chem. 14, 25-32.

Choy, J.-H., Byeon, S.-H. & Demazeau, G. (1988). J. Solid State Chem. 76, 97-101.

Choy, J.H., Demazeau, G. & Hong, S.T. (1992). Jpn. J. Appl. Phys. 31, 3649-3654.

Choy, J.-H., Hong, S.-T. & Choi, K.-Y. (1996). J. Chem. Soc., Far. Trans. 92, 1051-
1059.

Choy, J.-H., Park, J.-H, Hong, S.T. & Kim, D. K. (1994). J. Solid State Chem. 111, 370-
379.

Choy, J.-H., Park, J.-H., Hong, S.-T. & Kim, D.-K. (1994). J. Solid State Chem. 111,
370-379.

Christoph, G., Larson, A., Eller,G., Purson,J.,Zahrt,J.,Penneman,R. & Rinehart,G. (1988).

Acta Cryst. B44, 575-580.

Christy, A.G., Angle, R.J., Haines, J. & Clark, S. (1994). J. Phys.: Condens. Matter 6,
3125-3136.

Ciambelli, P., Cimino, S., De Rossi, S., Faticanti, M., Lisi, L., Minelli, G., Pettiti, I.,
Porta, P., Russo, G. & Turco, M. (2000). Appl. Cat. B. 24, 243-253.

Clearfield, A. (1963). Acta Cryst. 16, 134-142.

Codfunke, E., Booij, A., Smit-Groen, V. & van Vlaanderen, P. (1997). J. Solid State
Chem. 131, 341-349.

Colla, E.L., Reaney, I.M. & Setter, N. (1993). J. Appl. Phys. 74, 3414-3425.

Collomb, A., Samara, D., Buevoz, J., Levy, J. & Jourbet, J.C. (1983). J. Mag. Mag.
Mater. 40, 75-82.

Collumb,A.,Samaras,D., Bochu,B. & Jourbet,J.C. (1977). Phys. Stat. Sol. A 41, 459-463.

Courtney, W.E. (1970). IEEE Trans. Mic. TT MTT-18, 476-485.

221
Coutures, J. & Coutures, J.P. (1984). J. Solid State Chem. 52, 95-100.

Cox, D. (1972). IEEE T. Mag. 8(1), 161-182.

Cox, D.E. & Sleight, A.W. (1979). Acta Cryst. B35, 1-10.

Cox, D.E. Shirane, G. & Frazer, B.C. (1967). J. Appl. Phys. 1967, 1459-1460.

Crama, W.J. (1981). Acta Cryst. B37, 2133-2136.

Cuffini, S.L., Macagno, V.A., Carbonio, R.E., Mela, A., Trollund, E. & Gautier, J.l.
(1993). J. Solid State Chem. 105, 161-170.

Currie, R.C., J.F., Vente, Frikkee, E. & Ijdo, D.J.W. (1995). J. Solid State Chem. 116,
199-204.

Cussen, E., Rosseinski, M.J., Battle, P.D., Burley, J.C., Spring, L.E., Vente, J.F.,
Blundell, S.J., Coldea, A.I. & Singleton, J. (2001). J. Am. Chem. Soc. 123, 1111-1122.

Cussen, E.J., Vente, J.P., Battle, P.D. & Gill, T.C. (1997). J. Mater. Chem. 7, 459-463.

Damay, F., Jirak, Z., Hervieu, M., Martin, C., Maignan, A., Raveau, B., Andre, G. &
Bouree, F. (1998). J. Mag. Mag. Mat. 190, 221-232.

Damay, F., Martin, C., Maignan, A., Hervieu, M., Raveau, Jirak, Z., Andre, G. & Bouree,
F. (1999). Chem. Mater. 11, 536-541.

Dann, S.E., Currie, D.B., Weller, M.T., Thomas, M.F. & Al-rawwas, A.D. (1994). J.
Solid State Chem. 109, 134-144.

Darlington, C.N.W. & Knight, K.S. (1999). Physica B. 266, 368-372.

Darriet, J., Bontchev, R., Dussarrat, C., Weill, F. & Darriet, B. (1993). Eur. J. Solid State
Inorg. Chem. 30, 273-286.

Davies, P.K., Tong, J. & Negas, J. (1997). J Am. Ceram. Soc. 80(7), 1727-1740.

de Gennes, P.-G. (1960). Physical Review 118(1), 141-154.

Deblieck, R. (1986). Acta Cryst. A42, 318-325.

Deblieck, R., Van Tendeloo, G., Van Landuyt, J. & Amelinckx, S. (1985). Acta Cryst.
B41, 319-329.

222
Demazeu, G., Marbeuf, A., Pouchard, M. & Hagenmuller, P. (1971). J. Solid State Chem.
3, 582-589.

Demazeu, G., Parent, C., Pouchard, M. & Hagenmuller, P. (1972). Mat. Res. Bull. 7, 913-
920.

Denoey, F., Lambert, M., Comes, R. & Currat, R. (1976). Solid St. Comm. 18, 441-444.

Derighetti, B., Drumheller, J.E., Laves, F., Muller, K.A. & Waldner, F. (1965). Acta
Cryst. 18, 557.

Deschanvres, A., Raveau, B. & Tollemer, F. (1967). Bull. Soc. Chim. Fr. 1967, 4077-
4078.

Deschizeaux, M.N., Courbet, J.C., Vegas, A., Collomb, A., Chenavas, J. & Marezio, M.
(1976). J. Solid State Chem. 19, 45-51.

Dickens, P. & Powell, A. (1991). J. Mater. Chem. 1, 137-138.

Diehl,R. & Brandt,G. (1975). Mat. Res. Bull. 10, 85-90.

Doi, Y. & Hinatsu, Y. (1999). J. Phys.:Condens. Matter 11, 4813-4820.

Doi, Y. & Hinatsu, Y. (2001). J. Phys.: Condens. Matter 13, 4191-4202.

Doi, Y., Hinatsu, Y., Oikawa, K., Shimojo, Y. & Morii, Y. (2000). J. Mater. Chem. 10,
797-800.

Doi, Y., Hinatsu, Y., Oikawa, K., Shimojo, Y. & Morii, Y. (2000). J. Mater. Chem. 10,
1731-1735.

Dunaevskii, S/M., Kurbakov, A.I., trunov, V.A., Chernyshov, D.Y., Popov, v.V.,
Chernnyshev, V.V. & Rodriguez-Carvajal, J. (1998). Physics of the Solid State 40, 1158-
1162.

Dygas, J.R. & Breiter, M.W. (1996). Electrochimica Acta 41, 993-1001.

Edwards, A. J. & Peacock, R.D. (1959). J. Chem. Soc. , 4126-4127.

Eibschutz, M. (1965). Acta Cryst. 19, 337-339.

Elemans, A., van Laar, B., van der Veen, K. & Loopstra, B. (1971). J. Solid State Chem.
3, 238-242.

223
Endoh, Y., Hirota, K., Ishihara, S., Okamoto, S., Murakami, Y., Nishizawa, a., Fukuda,
T., Kimura, H., Nojiri, H., Kaneko, K. & Maekawa, S. (1999). Phys. Rev. Lett. 82, 4328-
4331.

Englman, R. (1981). Solid St. Comm. 40, 619-622.

Evdokimov, A.A. & Men'shenina, N.F. (1982). Zh. Neorg. Khim. 27, 1208-1209.

Faber, J. & Fawcett, T. (2002). Acta Cryst. B58, 325-332.

Fabry, J., Zikmund, Z., Kania, A. & Peetricek, V. (2000). Acta Cryst. C56, 916-918.

Faget, H., Grannec, J., Tressaud, A., Rodriguez, V., Roisnel, T., Flerov, I.N. & Gorev,
M,V. (1996). Eur. J. Solid State Inorg. Chem, 33, 893-905.

Faqir, H., Chiba, H., Kikuchi, M., Syono, Y., Mansori, M., Satre, P & Sebaoun, A.
(1999). J. Solid State Chem. 142, 113-119.

Feiner, L.F., Oles, A.M. & Zaanen, J. (1997). Phys. Rev. Lett. 78(14), 2799-2802.

Fesenko, E.G., Filip'ev, V.S., Kupriyanov, M.F., Devlikanova, R., Zhavoronko, G.P. &
Ochirov, V.A. (1970). Neorganicheskie materialy 6, 800-802.

Filip'ev, V.S. & Fesenko, E.G. (1966). Sov. Phys. Crystall. 10, 532-534.

Fontcuberta, J., Martinez, B., Seffar, A., Pinol, S., Garcia-Munoz, J.L. & Obradors, X.
(1996). Phys. Rev. Lett. 76(7), 1122-1125.

Foster, M.C., Nielson, R.M. & Abrahams, S.C. (1997). J. Appl. Phys. 82, 3076-3080.

Frontera, C., Garcia-Munoz, J.L., Llobet, A., Aranda, M.A.G., Ritter, C., Respaud, M. &
Vanacken, J. (2001). J. Phys.: Condensed. Matter 13, 1071-1078.

Fu, W.T. & D.J.W., Ijdo (1997). J. Solid State Chem. 128, 323-325.

Fu, Z., Li, Z. & Li, W. (1992). Wu Li xue bao 41(6), 937-947.

Fujii, T., Takahashi, J., Shimada, S. & Kageyyama, K. (1999). J. Electroceram. 3, 387-
397.

Fukuda, K., Kitoh, R. & Awai, I. (1993). Jpn. J. Appl. Phys. 32, 4584-4588.

Galasso, F. & Pyle, J. (1962). Inorganic Chem. , 482-484.

224
Gallaso, F.S., Layden, G.K. & Flinchbaugh, D.E. (1966). J. Chem. Phys. 44, 2703-2707.

Garcia-Landa, B., Ritter, C., Ibarra, M.R., Blasco, J., Algarabel, P.A., Mahendiran, R.&
Garcia, J. (1999). Solid State Comm. 110, 435-438.

Garcia-Munoz, J.L., Rodriguez-Carvajal, J., Lacorre, P. & Torrance, J.B. (1992). Phys.
Rev. B. 46(8), 4414-4425.

Garcia-Munoz, J.L., Frontera, C., Aranda, M.A.G., Llobet, A. & Ritter, C. (2001). Phys.
Rev. B 63, 644415.

Garton, G. & Wanklyn, B.M. (1967). J. Crystal Growth 1, 164-167.

Garxia-Jaca, J., Larramendi, J.I.R., Insausti, M., Arriortua, M.I. & Rojo, T. (1995). J.
Mater. Chem. 5(11), 1995-1999.

Geguzina, G.A., Fesenko, E.G. & Devlikanova, R.U. (1976). Sov. Phys. Crystall. 20, 518.

Geller, S. & Bala,V. (1956). Acta Cryst. 9, 1019-1025.

Geller, S. & Wood, E.A. (1956). Acta Cryst. 9, 563-568.

Giaquinta D M, Davis, W.M. & zur Loye, H.-C. (1994). Acta Cryst. C50, 5-7.

Gibb, T. C. (1993). J. Mater. Chem. 3, 441-446.

Gillet, Guyot, F., Price, G.D., Tourerie, B. & Le Cleach, A.L. (1993). Phys. Chem.
Minerals 20, 159-170.

Gingl, F., Vogt, T., Akiba, E. & Yvon, K., (1999). J. Alloys and Compounds 282, 125-
129.

Glazer, A.M. (1972). Acta Cryst. B28, 3384-3392.

Glazer, A.M. (1975). Acta Cryst. A31, 756-762.

Glazer, A.M. & Megaw, H.D. (1972). Philos. Mag. 25, 1119-1135.

Goldschmidt, V.M. (1926). Naturwissenschaften 14, 477-485.

Gong, G.-Q., Canedy, C., Xiao, G., Sun, J., Gupta, A. & Gallagher, W.J. (1995). Appl.
Phys. Lett. 67, 1783-1785.

225
Gontchar, L.E., Nikiforov, A.E. & Popov, S.E. (2001). J. Mag. Mag. Mater. 233, 175-
191.

Goodenough, J.B. (1955). Physical Review 100(2), 564-573.

Goodenough, J.B. (1998). Annu. Rev. Mater. Sci. 28, 1- 27.

Goodenough, J.B. & Zhou, J.S. (1998). Chem. Mater. 10, 2980-2993.

Goodenough, J.B., Longo, J.M. & Kafalas, J.A. (1968). Mat. Res. Bull. 3, 471-482.

Greedan, J.E., Bieringer, M., Britten, J.F., Giaquinta, D.M. & zur Loye, H.-C. (1995). J.
Solid State Chem. 116, 118-130.

Groen, W.A. & Ijdo, D.J.W. (1987). Acta Cryst. C43, 1033-1036.

Groen, W.A., van Berkel, F.P.F.V. & Ijdo, D.J.W. (1986). Acta Cryst. C42, 1472-1475.

Guilhaume, N. & Primet, M. (1997). J. Catalysis 165, 197-204.

Guo, R., Bhalla, A.S., Sheen, J., F.W., Ainger, Erdei, S., Subbarao, E.C. & Cross, L.E.
(1995). J. Mater. Res. 10, 18-25.

Hamada, N., Sawada, H., Solovyev, I. & Terakura, K. (1997). Physica B. 237-238, 11-13.

Han, B., Neumayer, D.A., Goodreau, B.H., Marks, T.J., Zhang, H. & Dravid, V.P.
(1994). Chem. Mater. 6, 18-20.

Harada, D., Wakeshima, M., Hinatsu, Y., Ohoyama, K. & Yamaguchi, Y. (2000). J.
Phys.:Condens. Matter 12, 3229-3239.

Harrison, W.T.A., Reis, K. P., Jacobson, A.J., Schneemeyer, L.F. & Waszczak, J.V.
(1995). Chem. Mater. 7, 2161-2167.

Harrison, W.T.A., Reis, K.P., Jacobson, A.J., Scheemeyer, L.F. & Waszczak, J.V.
(1995). Chem. Mater. 7, 2161-2167.

Hayashi, K., Demazeua, G., Pouchard, M. & Hagenmuller, P. (1980). Mat. Res. Bull. 15,
461-467.

He, T. & Cava, R.J. (2001). J. Phys.: Condens. Matter 13, 8347-8361.

Hemley, R.J. & Cohen, R. E. (1992). Annual Review of Earth and Planetary Sciences 20,
553-600.

226
Hemley, R.J., Jackson, M.D. & Gordon, R.G. (1987). Phys. Chem. Minerals 14, 2-12.

Hepworth, M.A., Jack, K.H., Peacock, R.D. & Westland, G.J. (1957). Acta Cryst. 10, 63-
69.

Hidaka, M., Eguchi, T. & Yamada, I. (1998). J. Phys. Soc. Jpn. 67, 2488-2494.

Hilton, A.D., Barber, D.J., Randall, C.A. & Shrout, T.R. (1990). J. Mater. Sci. 25, 3461-
3466.

Hinatsu,Y., Itoh, M. & Edelstein, N. (1996). J. Solid State Chem. 132, 337-341.

Hirakawa, K. & Kurogi, Y. (1970). Supp. Progr. Theor. Phys. 46, 147-161.

Hirose, N. & West, A.R. (1996). J. Am. Ceram. Soc. 79, 1633-1641.

Hirota, K. & Endoh, Y. (1999). Acta Cryst. B55, 24-30.

Hodges, J.P., Short, S., Jorgensen, J.D., Xiong, X. & Dabrowski, B. (2000). J. Solid State
Chem. 151, 190-209.

Hohnstedt, C. & Meyer, G. (1993). Z. Anorg. Allg. Chem. 619, 1374-1378.

Holsa, J., Lastusaari, M. & Valkonen, J. (1997). J. Alloys and Compounds 262-263, 299-
304.

Honle, W., Miller G., & Simon, A. (1988). J. Solid State Chem. 75, 147-155.

Horikubi, T. & Kamegashira, N. (2000). Mater. Chem. Phys. 65, 316-319.

Horikubi, T., Mori, T., Nonobe, H. & Kamegashira, N. (1999). J. Alloys Comp. 289, 42-
47.

Horiuchi, H., Saito, A., Shishido, T. & Fukuda, T. (1997). Kidorui 30, 214-215.

Howard, C. J. & Kennedy, B.J. (1999). J. Phys.: Condens. Matter 11, 3229-3236.

Howard, C. J., & Stokes, H. T. (1998). Acta Cryst. B54, 782-789.

Howard, C.J., Kennedy, B.J. & Chakoumakos, B.C. (2000). J. Phys.: Condens. Matter
12, 349-365.

227
Howard, C.J., Knight, K.S.,Kennedy, B.J. & Kisi, E.H. (2000). J. Phys.: Condens. Matter
12, L677-L683.

Huang, T. Parrish, W., Toraya, H., Lacorre, P. & Torrance, J. (1990). Mat. Res. Bull. 25,
1091-1098.

Hubner, K. (1970). Phys. Lett. 31a(7), 365-366.

Hughes, H., Iddles, D.M. & Reaney, I.M. (2001). Appl. Phys. Letters 79, 2952-2954.

Husson, E., Abello, L.. & Morell, A. (1990). Mat. Res. Bull. 25, 539-545.

Husson, E., Chubb, M. & Morell, A. (1988). Mat. Res. Bull. 23, 357-361.

Hutchings, M., Samuelsen, E.J., Shirane, G. & Hirakawa, K., (1969). Phys. Rev. 188,
919-923.

Hutton, J. & Nelmes, R.J. (1981). J. Phys. C.: Solid State Phys. 14, 1713-1736.

Hutton, J., Nelmes, R.J., & Scheel, R.J. (1981). Acta Cryst. A37, 916-920.

Hwang, H.Y., Palstra, T.T.M., Cheong, S.-W. & Batlogg, B. (1995). Phys. Rev. B.
52(21), 15046-15049.

Ihringer, J. (1982). Solid State Communications 41(7), 525-527.

International Tables for Crystallography (1995). Vol. A, Space Group Symmetry, 3rd
ed., edited by T. Hahn. Dordrecht: Kluwer Academic Publishers.

Irvine, J.T.S., Sinclair, D.C. & West, A.R. (1990). Adv. Mater. 2, 132-138.

Isobe, M., Kimizuka, N., Nakamura, M. & Mohri, T. (1991). Acta Cryst. C47, 423-424.

Ito, K., Tezuka, K. & Hinatsu, Y. (2001). J. Solid State Chem. 157, 173-179.

Itoh, M., Ohta, I. & Inaguma, Y. (1996). Mat. Sci. Eng. B41, 55-58.

Iwanaga, D., Inaguma, Y. & Itoh, M. (1999). J. Solid State Chem. 147, 291-295.

Iwanaga, D., Inaguma, Y. & Itoh, M. (2000). Mat. Res. Bull. 35, 449-457.

Izumiyama, Y., Doi, Y., Wakeshima, M., Hinatsu, Y., Oikawa, K., Shimojo, Y. & Morii,
Y. (2000). J. Mater. Chem. 10, 2364-2367.

228
Izumiyama, Y., Doi, Y., Wakeshima, M.,Hinatsu, Y.,Shimojo, Y. & Morii, Y. (2001). J.
Phys.: Condens. Matter 13, 1303-1313.

Jacobson, A.J., Tofield, B.C. & Fender, B.E.F. (1976). Acta Cryst. B32, 1083-1087.

Jacobson, A.J., Tofield, B.C., and Fender, B.E.F. (1972). Acta Cryst. B28, 956-961.

Jansen, M. & Klinkert, B. (1990). Mat. Res. Bull. 25, 1415-1420.

Jarabek, B.R., Grier, D.G. & McCarthy, G.J. (1996). Powder Diff. 11, 56-59.

Ji, K. & Zheng, H. (2001). J. Phys.: Condensed. Matter 13, 1079-1091.

Jiang, B., Zuo, J.M., Chen, Q. & Spence, J.C.H. (2002). Acta Cryst. A58, 4-11.

Jin, F., Endo, T., Takizawa, H. & Shimada, M. (1994). J. Solid State Chem. 113, 138-
144.

Jirak, Z., Damay, F., Hervieu, M., Martin, C., Raveau, B., Andre, G. & Bouree, F.
(2000). Phys. Rev. B. 61, 1181-1188.

Jirak, Z., Krupicka, Simsa, Z., Dlouha, M. & Vratislav, S. (1985). J. Magn. Magn. Mat.
53, 153-166.

Jirak, Z., Pollert, E., Andersen, A.F., Grenier, J.-C., Hagenmuller, P. (1990). Eur. J. Solid
State Inorg. Chem. 27, 412-433.

Jona, F., Shirane, G., Mazzi, F. & Pepinski, R. (1957). Physical Review 105(3), 849-856.

Jones, C.W., Battle, P.D. & Lightfoot, P. (1989). Acta Cryst. C45, 365-367.

Jones, G.O. & Thomas, P.A. (2002). Acta Cryst. B58, 168-178.

Jung, D.-Y. & Demazeau, G. (1995). J. Solid State Chem. 115, 447-455.

Jung, D.-Y. & Demazeau, G. (1995). Solid State Comm. 12, 963-967.

Jung, D.Y., Demazeau, G. & Choy, J.H. (1995). J. Mater. Chem. 5(3), 517-519.

Kaiser,V., Otto, M., Binder, F. & Babel, D. (1990). Z. Anorg. Allg. Chem. 585, 93-104.

Kamata, K., Nakamura, T. & Sata, T. (1975). Chem. Lett. , 81-86.

Kanamori, J. (1960). J. Appl. Phys. 31(5), 14S-23S.

229
Karppinen, M., Yamauchi, H., Ito, T., Suematsu, H. & Fukunaga, O. (1996). Mat. Sci.
Eng. B41, 59-62.

Kasuya, M., Tokura, Y., Arima, T., Eisaki, H. & Uchida, S. (1993). Phys. Rev. B. 47(11),
6197-6202.

Kawasaki, S., Takano, M., Kanno, R., Takeda, T. & Fujimori, A., (1998). Letters 67(5),
1529-1532.

Kawashima, S., Nishida, M., Ueda, I. & Ouchi, H. (1983). J. Am. Ceram. Soc. 66, 421-
423.

Kay, H.F. & Bailey, P.C. (1957). Acta Cryst. 10, 219-226.

Kennedy, B.J. & Hunter, B.A. (1998). Phys. Rev. B. 58(2), 653-658.

Kennedy, B.J., Howard, C. J. & Chakoumakos, B.C. (1999). J. Phys.: Condens. Matter
11, 1479-1488.

Kennedy, B.J., Howard, C. J. & Chakoumakos, B.C. (1999). Phys. Rev. B. 59(6), 4023-
4027.

Kennedy, B.J., Howard, C. J. & Chakoumakos, B.C. (1999). Phys. Rev. B. 60(5), 2972-
2975.

Kennedy, B.J., Prodjosantoso, A.K. & Howard, C. J. (1999). J. Phys.: Condens. Matter
11, 6319-6327.

Khanna, A. & Garault, Y. (1983). IEEE Trans. Mic. TT MTT-31, 261-264.

Khattak, C. & Cox, D. (1977). Mat. Res. Bull. 12, 463-472.

Khattak, C., Cox, D. & Wang, F.F.Y. (1975). J. Solid State Chem. 13, 77-83.

Kijima, N, Tanaka, K. & Marumo, F. (1981). Acta Cryst. B37, 545-548.

Kijima, N, Tanaka, K. & Marumo, F. (1983). Acta Cryst. B39, 557-561.

Kim, B.K, Cha, S.-B. & Park, J.H. (1999). Mat. Sci. Eng. B58, 244-250.

Kim, H.J., Kucheiko S., Yoon, S.-J. & Jung, H.-J. (1997). J. Am. Ceram. Soc. 80, 1316-
1318.

230
Kim, I-S., Nakamura, T., Itoh, M. & Inaguma, Y. (1993). Mat. Res. Bull. 28, 1029-1039.

Kim, J.S., Cheon,C., Kang, H-J., Shim,H-S.,Lee,C-H., Nam, S. & Byun, J-D. (1999).
Material Lett. 38, 294-299.

Kim, S.H. & Battle, P.D. (1995). J. Solid State Chem. 114, 174-183.

Kim, S.J., Lemaux, S., Demazeau, G., Kim, J.-Y. & Choy, J.-Y. (2001). J. Amer. Chem.
Soc. 123, 10413-10414.

Kim, S-J., Demazeau, G., Alonso, J. & Choy, J.-H. (2001). J. Mater. Chem, 11, 487-492.

Kim, W.S., Kim, E.S. & Yoon, K.H. (1999). Ferroelectrics 223, 277-284.

Kim, Y., Lee, K.-M. & Jang, H.M. (2000). J. Mater. Sci. 35, 4885-4893.

Knight,K. & Bonanos, N. (1994). J. Mater. Chem. 4, 899-901.

Knight,K. & Bonanos, N. (1995). Mat. Res. Bull. 30(3), 347-356.

Knizek, K., Jirak, Z., Pollert, E., Zounova, F. & Vratislav, S. (1992). J. Solid State Chem.
100, 292-300.

Knochenmuss, R., Reber, R., Rajasekharan, M.V. & Gudel, H.U. (1986). J. Chem. Phys.
85, 4280-4289.

Kobayashi, K.-I., Kimura, T., Sawada, H., Terakura, K. & Tokura, Y. (1998). Nature
395, 677-680.

Kobayashi, K.-I., Kimura, T., Tomioka, Y., Sawada, H., Terakura, K. & Tokura, Y.
(1999). Phys. Rev. B. 59(17), 11159-11162.

Kobayashi,H.,Nagata, M., Kanno,R. & Kawamoto,Y. (1994). Mat. Res. Bull. 29, 1271-
1280.

Kohl, V.P. (1973). Z. Anorg. Allg. Chem. 401, 121-131.

Kohl, V.P., Schultze-Rhonhof, E. & Reinen, D. (1970). Z. Anorg. Allg. Chem. 378, 129-
143.

Kohn, K, Akimoto, S., Uesu, Y. & Asai, K. (1974). J. Phys. Soc. Japan 37, 1169.

Kohn, K., Inoue, K., Horie, O. & Akimotot, S. (1976). J. Solid State Chem. 18, 27-37.

231
Koopmans, H., van de Velde, G. & Gellings, P. (1983). Acta Cryst. C39, 1323-1325.

Kozak, A., Samoel, M., Renaudin, J. & Ferey (1988). Eur. J. Solid State Inorg. Chem. 25,
15-19.

Krupka, J. Fifth Int. Conference on Dielectric Materials, Measurements and Applications,

Canterbury, UK, 27-30 June 1988, pp. 322-325.

Kruth, A., Guth, U. & West, A. (1999). J. Mater. Chem. 9, 1579-1583.

Kruth, A., Tabuchi, M., Guth, U. & West, A.R. (1998). J. Mater. Chem. 8, 2515-2520.

Kugel', K.I. & Khomskii, D.I. (1982). Sov. Phys. Usp. 25(4), 231-256.

Kupriyanov, M.F. & Filip'ev, V.S. (1963). Sov. Phys. Crystall. 8, 278-283.

Kuwahara, H., Tomioka, Y., Asamitsu, A., Moritomo, Y. & Tokura, Y. (1995). Science
270, 961-963.

Labeau, M., Bochu, B., Jourbet, J.C. & Chenavas, J. (1980). J. Solid State Chem. 33, 257-
261.

Lacorre, P., Torrance, J., Pannetier, J., Nazzal, A., Wang, P.W. & Huang, T. (1991). J.
Solid State Chem. 91, 225-237.

Lang, C. & Muller-Buschbaum (1989). Z. Anorg. Allg. Chem. 574, 169-171.

Larson, A.C. & von-Dreele, R.B., General Structure Analysis System (GSAS), Los
Alamos National Laboratories, 1990.

Laugier, J. (2002). http://www.inpg.fr/LMGP (accessed Dec 2002).

Le Bail, A., Jacoboni, C., Leblanc, M., de Pape, R., Duroy, H. & Fourquet, J. (1988). J.
Solid State Chem. 77, 96-101.

Lee, H.J., Park, H.M., Cho, Y.K., Ryu, H., Song, Y.W., Paik, J.H., Nahm, S. & Byun, J.-
D. (2000). J. Am. Ceram. Soc. 83, 2267-2272.

Leinenweber, K. & Parise J. (1995). J. Solid State Chem. 114, 277-281.

Leinenweber, K., Linton, J., Navrotsky, A., Fei, Y. & Parise J. B. (1995). Phys. Chem.
Minerals 22, 251-258.

Lenz, A. & Muller-Buschbaum, Hk. (1990). J. Less-Common Metals 161, 141-146.

232
Leporcher, P., Andre, G., Roisnel, T., Le Bihan, T. & Noel, H. (1994). J. Alloys Compds.
213, 506-508.

Levin, I, Chan, J.Y., Maslar, J.E. & Vanderah, T.A. (2001). J. Appl. Phys. 90(2), 904-
914.

Levin, I, Chan, J.Y., Maslar, J.E. & Vanderah, T.A. (2001). J. Appl. Phys. 90(2), 904-
914.

Levin, I. & Bendersky, L.A. (1999). Acta Cryst. B55, 853-866.

Levin, I., Bendersky, L.A., Cline, J.P., Roth, R.S. & Vanderah, T.A. (2000). J. Solid State
Chem. 150, 43-61.

Levin, I., Chan, Geyer, J.Y. R.G., Maslar, J.E. & Vanderah, T.A. (2001). J. Solid State
Chem. 156, 122-134.

Levin, I., Vanderah, T.A., Coutts, R. & Bell, S.M. (2002). J. Mater. Res. 17(7), 1729-
1734.

Lewis, G.V. & Catlow, C.R.A. (1985). J. Phys. C. Solid State Phys. 18, 1149-1161.

Liechtenstein, A.I., Anisimov, V.I. & Zaanen, J. (1995). Phys. Rev. B. 52(8), R5467-
R5470.

Ligny, D. & Richet, P. (1996). Phy. Rev. B 53(6), 3013-3022.

Liu, C.-L. & Wu, T.B. (2001). J. Am. Ceram. Soc. 84, 1291-1295.

Liu, D., Yao, X., Harmer, M.P. & Smyth, .M. (1992). Eur. J. Solid State Inorg. Chem, 29,
455-462.

Liu, D., Yao, X., Smyth, D.M., Bhalla, A.S. & Cross, L.E. (1993). J. Appl. Phys. 74,
3345-3356.

Liu, G. & Greedan, J.E. (1994). J. Solid State Chem. 110, 274-289.

Liu, G., Zhao, X. & Eick, H. (1992). J. Alloys and Compounds 187, 145-156.

Liu, R., Xuan, Y. & Jia, Y.Q. (1997). J. Solid State Chem. 134, 420-422.

Liu, R., Xuan, Y. & Jia, Y.Q. (1998). Mat. Chem. Phys. 57, 81-85.

233
Liu, X. & Lieberman, R.C. (1993). Phys. Chem. Minerals 20, 171-175.

Lombardo, E.A. & Ulla, M. A. (1998). Research on Chemical Intermediates 24(5), 581-
592.

Longo, J.M., Raccah, P.M., Kafalas, J.A. & Pierce, J.W. (1972). Mat. Res. Bull. 7, 137-
146.

Longo, V., Ricciardiello, F. & Minichelli, D. (1981). J. Mater. Sci. 16, 3503-3505.

Looby, J.T & Katz, L. (1954). J. Amer. Chem. Soc. 76, 6029-6030.

Loopstra, B.O. & Rietveld, H.M. (1969). Acta Cryst. B25, 1420-1421.

Lopez, M.L., Veiga, M.L., Rodriguez-Carvajal, J., Fernandez, F., Jerez, A. & Pico, C.
(1992). Mat. Res. Bull. 27, 647-654.

Luana, V., Costales, A. & Martin Pendas, A. (1997). Phys. Rev. B. 55(7), 4285-4297.

Lufaso, M.W., & Woodward, P.M. (2001). Acta Cryst. B57, 725-738.

Luhrs, C., Sapina, F., Beltran-Porter, D., Casan-Pastor, N. & Fuertes, A. (1998). J. Mater.
Chem. 8, 209-217.

Lutgert, B. & Babel, D. (1992). Z. Anorg. Allg. Chem. 616, 133-140.

Lynn, J.W., erwin, R.W., Borchers, J.A., Huang, Q., Santoro, A., Peng, J.-L. & Li, Z.Y.
(1996). Phys. Rev. Lett. 76(21), 4046-4049.

MacLean, D., Ng,H-N. & Greedman, J.E. (1979). J. Solid State Chem. 30, 35-44.

Magyari-Kope, B., Vitos, L. & Kollar, J. (2001). Phys. Rev. B 63, 104111.

Magyari-Kope, B., Vitos, L., Grimvall, G., Johansson, B. & Kollar, J. (2002). Phys.
Rev. B 65, 193107.

Magyari-Kope, B., Vitos, L., Grimvall, G., Johansson, B. & Kollar, J. (2002). Mat. Res.
Soc. Symp. Proc. 718, D6.5.1-D6.5.6.

Magyari-Kope, B., Vitos, L., Johansson, B. & Kollar, J. (2001). Acta Cryst. 57, 491-496.

Maignan, A., Martin, C., Hervieu, M. & Raveau, B. (2001). Solid State Comm. 117, 377-
382.

234
Majumdarb, S., Kumara, A., Nalinia, G. & Guru Rowa, T.N. (2001). App. Phys. Lett.
79(18), 2952-2954.

Makovec, D., Ule, N. & Drofenik, M. (2001). J. Am. Ceram. Soc. 84, 1273-1280.

Malecki, G. & Ratuszna, A. (1999). Powder Diff. 14, 25-30.

Manthiram, A., Tang, J.P. & Manivannan, V. (1999). J. Solid State Chem. 148, 499-507.

Marchand, R., Laurent, Y., Guyader, J., L'Haridon, P. & Verdier, P. (1991). J. Eur.
Ceram. Soc. 8, 197-213.

Marcos, M.D. & Attfield, J.P. (1994). J. Mater. Chem. 4, 475-477.

Marezio, M. & Dernier, P.D. (1971). Mat. Res. Bull. 6, 23-30.

Marezio, M., Dernier, P.D. & Remeika, J.P. (1972). J. Solid State Chem. 4, 11-19.

Marezio, M., Dernier, P.D., Chenavas, J. & Jourbet, J. (1973). J. Solid State Chem. 6, 16-
20.

Marezio, M., Remeika, J.P. & Dernier, P.D. (1970). Acta Cryst. B26, 2008-2022.

Marezio,M., Dernier,P.D. & Remeika,J.P. (1972). J. Solid State Chem. 4, 11-19.

Marsh, A. & Clark, C.C. (1969). Philos. Mag. 19, 449-463.

Marti, W.,Fischer,P.,Altorfer,F.,Scheel,H. & Tadin,M. (1994). J. Phys.: Condens. Matter

6, 127-135.

Martin, C., Maignan, A., Damay, F., Hervieu, M. & Raveau, B. (1997). J. Solid State
Chem. 134, 198-202.

Marx, D.T., Radaelli, P.G., Jorgensen, J.D., hitterman, R.L., Hinks, D.G., Pei, S. &
Dabrowski, B. (1992). Phys. Rev. B. 46(2), 1144-1156.

Mathe, V.L., Patankar, K.K., Kothale, M.B., Kulkarni, S.B., Joshi, P.B. & Patil, S.A.
(2002). Pramana-J. Phys. 58(5-6), 1105-1113.

McCarty, G.J., Gallagher, P.V. & Sipe, C. (1973). Mat. Res. Bull. 8, 1277-1284.

McCartyh, G.J., Sipe, C. & McIlvried,K.E. (1974). Mat. Res. Bull. 9, 1279-1284.

Medarde, M. (1997). J. Phys.: Condens. Matter 9, 1679-1707.

235
Medarde, M., Mesot, J., Lacorre, P. Rosenkranz,S., Fisher, P. & Gobrect, K. (1995).
Phys. Rev. B. 52(13), 9248-9258.

Megaw, H.D. (1968). Acta Cryst. A24, 583-588.

Megaw, H.D. & Darlington, C.N (1975). Acta Cryst. A31, 161-173.

Mehta, A., Berliner, R. & Smith, R. (1997). J. Solid State Chem. 130, 192-198.

Meneghini, C., Castellano, C., Kumar, A., Ray, S., Sarma, D.D. & Mobilio, S. (1999).
Phys. Stat. Sol. B 215, 647-652.

Meneghini, C., Cimino, R., Pascarelli, S., Mobilio, S., Raghu, C. & Sarma, D.D. (1997).
Phys. Rev. B. 56(7), 3520-3523.

Messer, C., Eastman, J., Mers, R. & Maeland, A. (1964). Inorganic Chem. 3, 776-778.

Meyer, C., Gros, Y., Bochu, B.,Collumb,A.,Chenavas,J.,Jourbet,J.C. & Marezio,M.

(1978). Phys. Stat. Sol. A 48, 581-586.

Michel, C., Moreau, J.M., Achenbach, G., Gerson, R. & James, W.J. (1969). Solid St.
Comm. 7, 701-704.

Milange, F., Caignaert, V., Mather, G., Suard, E. & Raveau, B. (1996). J. Solid State
Chem. 127, 131-135.

Millange, F., Brion, S. & Chouteau, G. (2000). Phys. Rev. B. 62(9), 5619-5626.

Milman, V., Winkler, B., White, J.A., Pickard, C.J., Payne, M.C., Akhmatskaya, E.V. &
Nobes, R.H. (2000). Int., J. Quant. Chem. 77, 895-910.

Minkiewicz, V.J., Fujii, Y. & Yamada, Y. (1970). J. Phys. Soc. Jpn. 28, 443-450.

Mitchell, J.F., Argyriou, D.N., Potter, C.D., Hinks, D.G., Jorgensen, J.D. & Bader, S.D.
(1996). Phy. Rev. B 54(9), 6172-6183.

Mitzi, D. & Liang, K. (1997). J. Solid State Chem. 134, 376-381.

Miyata, N., Tanaka, K. & Marumo, F. (1983). Acta Cryst. B39, 561-564.

Mizokawa, T., Khomskii, D.I. & Sawatzky, G.A. (1999). Phys. Rev. B. 60, 7309-7313.

Mizuno, M., Yamada, T. & Noguchi, T. (1977). Yogyo Kyokai Shi 85, 543-548.

236
Moledetsky, I. & Davies, P.K. (2001). J. Eur. Ceram. Soc. 21, 2587-2591.

Moreau, J. M., Michel, C., Gerson, R. & James, W.J. (1971). J. Phys. Chem. Solids 32,
1315-1320.

Mori, T., Aoki, K., Kamegashira, N., Shishido, T. & Fukuda, T. (2000). Mater. Letters
42, 387-389.

Morikawa, K., Iga, F., Inoue, I.H., Onari, S. & Nishihara, Y. (1994). Physica B 194-196,
1205-1206.

Moritomo, Y., Kuwahara, H., Tomioka, Y. & Tokura, Y. (1997). Phys. Rev. B. 55(12),
7549-7556.

Moritomo, Y., Xu, S., Machida, A., Akimoto, T., Nishiori, E., Takata, M. & Sakata, M.
(2000). Phys. Rev. B. 61, R7827-R7830.

Moure, C., Villegas, M., Fernandez, J.F., Tartaj, J. & Duran, P. (1999). J. Mater. Sci. 34,
2565-2568.

Muller, J., Haouzi, A., Laviron, C., Labeau, M. & Jourbet, J.C. (1986). Mat. Res. Bull.
21, 1131-1136.

Murakami, Y., Hill, J. P., Gibbs, D., Blume, M., Koyama, I., Tanaka, M., Kawata, H.,
Arima, T. & Tokura, Y. (1998). Phys. Rev. Lett. 81(3), 582-585.

Murakami, Y., Kawada, H., Kawata, H., Tanaka, M., Arima, T., Moritomo, Y. &
Tokura, Y. (1998). Phys. Rev. Lett. 80(9), 1932-1935.

Nagai, T., Inuzuka, T. & Sugiyama, M. (1992). Jpn. J. Appl. Phys. 31, 3132-3135.

Nakagawa, T. & Nomura, S. (1966). Jpn. J. Appl. Phys. 5, 578-581.

Nakamura, T. (1977). Mat. Res. Bull. 12, 815-824.

Nakamura, T. (1978). Mat. Res. Bull. 13, 1023-1030.

Nakotte, H., Laughlin, L., Kawanaka, H., Argyriou, D.N., Sheldon, R.I. & Nishihara, Y.
(1999). J. Appl. Phys. 85, 4850-4852.

Newnham, R.E. (1997). MRS Bulletin May, 20-34.

Newnham, R.E. & Meagher, E.P. (1967). Mat. Res. Bull.? 5, 549-553.

237
Newnham, R.E. & Ruschau, G.R. (1991). J. Am. Ceram. Soc. 74, 463-480.

Nishihata, Y., Mizuki, Akao, T., Tanaka, H., Uenishi, M., Kimura, M., Okamoto, T. &
Hamada, N. (2002). Nature 418, 164-167.

Nomura, S. (1983). Ferroelectrics 49, 61-70.

Novak, J. & Coufova, P. (1969). Chem. Zvesti 23, 66-72.

Obradors, X., Paulious, L., Maple, M., Torrance, J., Nazzal, A., Fontcuberta,J. &
Granados,X. (1993). Phys. Rev. B. 47(18), 12353-12356.

O'Bryan, H.M. & Thomson,J. (1974). J. Am. Ceram. Soc. 57, 522-526.

Odental, R-h. & Hoppe, R. (1971). Montsh. Chem. 102, 1340-1350.

Okazaki, A. & Suemune, Y. (1961). J. Phys. Soc. Japan 16(2), 176-183.

O'Keefe, M., & Hyde, B.G. (1976). Acta Cryst. B32, 2923-2936.

O'Keeffe, M., & Hyde, B.G. (1977). Acta Cryst. B33, 3802-3813.

O'keeffe, M., Hyde, B.G. & Bovin, J., (1979). Phys. Chem. Minerals 4, 299-305.

Onada, M. & Nagasawa, H. (1996). Solid St. Commun. 99(7), 487-491.

Onoda, M., Kuwata, J., Kaneta, K., Toyama, K. & Nomura, S. (1982). Jpn. J. Appl. Phys.
21(12), 1707-1710.

Ouchetto, K., Archaimbault, F., Choisnet, J. & Et-Tabirou, M. (1997). Mat. Chem. Phys.
51, 117-124.

Ozaki, Y., Ghedira, M., Chenavas, J., Jourbet, J. & Marezio, M. (1977). Acta Cryst. B33,
3615-3617.

Palacin, M.R., Bassas, J., Rodriguez-Carvajal, J. & Gomez-Romero, P. (1993). J. Mater.

Chem. 3, 1171-1177.

Palacin, M.R., Bassas, J., Rodriguez-Carvajal, J., Fuertes, A. & Casan-Pastor Gomez-
Romero (1994). Mat. Res. Bull. 29(9), 973-980.

Paladina, A.E. (1971). J. Am. Ceram. Soc. 54(3), 168-169.

238
Paolasini, L., Caciuffo, R., Sollier, A., Ghigna, P. & Altarelli, M. (2002). Phys. Rev. Lett.
88(10), 106403-1:106403-4.

Paris, J., McCarron, M. & Sleight, A. (1987). Mat. Res. Bull. 22, 803-811.

Park, B.H., Kang, B.S., Bu, S.D., Noh, T.W., Lee, J. 7 Jo, W. (1999). Nature 401, 682-
684.

Park, H.S., Yoon, K.YH. & Kim, E.S. (2001). J. Am. Ceram. Soc. 84, 99-103.

Park, J.-H. & Parise, J.B.(1997). Mat. Res. Bull. 32(12), 1617-1624.

Park, J.-H., Parise, J.B., Woodward, P.W., Lubomirsky, I. & Stafsudd, O. (1999). J.
Mater. Res. 14(8), 3192-3195.

Park, J.-H., Woodward, P.M. & Parise, J.B. (1998). Chem. Mater. 10, 3092-3100.

Park, J-H., Woodward, P.M., Parise, J.B., Reeder, R.J., Lubominrsky, I. & Stafsudd, O.
(1999). Chem. Mater. 11, 177-183.

Patrat, G., Brunel, M. & Bergevin, F. (1976). J. Phys. Chem. Solids 37, 285-291.

Peng, N., Irvine, J. & Fitzgerald, G. (1998). J. Mater. Chem. 8, 1033-1038.

Penn, S.J., Alford, N.M., Templeton, A., Wang, X., Xu, M., Reece, M. & Scrapel, K.
(1997). J. Am. Ceram. Soc. 80, 1885-1888.

Peschel, S., Ziegler, B., Schwarten, M. & Babel, D. (2000). Z. Anorg. Allg. Chem. 626,
1561-1566.

Peterlin-Neumaier & Steichele, E. (1986). J. Magn. Magn. Mat. 59, 351-356.

Philipp, J. B., Reisinger, D., Schonecke, M., Marx, A., Erb, A., Alff, L. & Gross, R.
(2001). App. Phys. Lett. 79(22), 3654-3656.

Pickardt, J., Schendler, T. & Kolm, M. (1988). Z. Anorg. Allg. Chem. 560, 153-157.

Plourde, J.K., Linn, D.F., O'Bryan, H.M. & Thomson, J. (1975). J. Am. Ceram. Soc. 58,
418-420.

Podlesnyak,A., Rosenkranz, S., Fauth,F., Marti,W., Scheel,H. & Furrer,A. (1994). J.

Phys.: Condens. Matter 6, 4099-4106.

Poirier, J.-P. (1997). Nature 387, 653-654.

239
Politova, E.D., Kaleva, G.M., Danilenko, I.N., Chuprakov, V.F. Ivanov, S.A., Venevtsev
& Yu. N. (1990). Izvestieiia Akademii nauk SSSR. Neorganicheskie materialy 26, 2352-
2356.

Poppelmeier, K.R., Leonowicz, M., Scanlon, J. & Longo, J. (1982). J. Solid State Chem.
45, 71-79.

Porta, P., Rossi, S., Faticanti, M., Minelli, G., Pettiti, I., Lisi, L. & Turco, M. (1999). J.
Solid State Chem. 146, 291-304.

Primo-Martin, V. & Jansen, M. (2001). J. Solid State Chem. 157, 76-85.

Radaelli, P.G., Cox, D.E., Marezio, M. & Cheong, S.-W. (1997). Phys. Rev. B 55, 3015-
3023.

Radaelli, P.G., Cox, D.E., Marezio, M., Cheong, S.-W., Schiffer, P.E. & Ramirez, A.P.
(1995). Phys. Rev. Lett. 75(24), 4488-4491.

Radaelli, P.G., Iannoe, G., Marezio, M., Hwang, Cheong, S.W., Jorgensen, J.D. &
Argyriou, D.N. (1997). Phys. Rev. B. 56(13), 8265-8276.

Radaelli, P.G., Marezio, M., Hwang, H.Y. & Cheong, S-W. (1996). J. Solid State Chem.
122, 444-447.

Ramirez, A.P., Subramanian, M.A., Gardel, M.,Blumberg,G., Li,D.,Vogt,T. & Shapiro,

S.M. (2000). Solid St. Commun. 115, 217-220.

Randall, C.A. & Bhalla, A.S. (1990). Jpn. J. Appl. Phys. 29(2), 327-333.

Rao, C.N.R., Arulraj, A., Cheetham, A.K. & Raveau, B. (2000). J. Phys.: Condens.
Matter 12, R83-R106.

Rao, G.H., Barner, K. & Brown, I.D. (1998). J. Phys.: Condens. Matter 10, L757-L763.

Rao, G.H., Brown, I.D. & Barner, K. (1999). J. Phys.: Condens. Matter 11, 8103-8109.

Ratuszna, A., Rousseau, M. & Daniel, P. (1997). Powder Diffraction 12, 70-75.

Raveau, B., Maignan, A., Martin, C. & Hervieu, M. (2000). Mat. Res. Bull. 35, 1579-
1585.

Raveau, B., Zhao, Y.M., Martin, C., Hervieu, M. & Maignan, A. (2000). J. Solid State
Chem. 149, 203-207.

240
Reaney, I.M. & Ubic, R. (2000). Int. Ceram. 1, 48-52.

Reaney, I.M., Colla, E.L. & Setter, N. (1994). Jpn. J. Appl. Phys. 33, 3984-3990.

Reaney, I.M., Qazi, I. & Lee, W.E. (2000). J. Appl. Phys. 88(11), 6708-6714.

Reaney, I.M., Wise, P., Ubic, R., Breeze, J., Alford, McN, N., Iddles, D., Cannell, D. &
Price, T. (2001). Philos. Mag A81(2), 501-510.

Redfern, S.A. (1996). J. Phys.: Condens. Matter 8, 8267-8275.

Reinen, v. D. & Weitzel, H. (1976). Z. Anorg. Allg. Chem. 424, 31-38.

Reinen, V.D. & Weitzel, H. (1976). Z. Anorg. Allg. Chem. 424, 31-38.

Reis, K., Jacobson, A. & Nicol, J. (1993). J. Solid State Chem. 107, 428-443.

Rey, M., Dehaudt, P., Jourbet, J., Lambert-Andron, B., Cyrot, M. & Cyrot-Lackmann,F.
(1990). J. Solid State Chem. 86, 101-108.

Rietveld, H.M. (1969). J. Appl. Cryst. 2, 65-71.

Rijssenbeek, J.T., Malo, S., Caignaert, V. & Poeppelmeier, K.R. (2002). J. Am. Chem.
Soc. Comm. 124(10), 2090-2091.

Ritter, C., Ibarra, M.R., Morellon, L., Blasco, J., Garcia, J. & De Teresa, J.M. (2000). J.
Phys.:Condens. Matter 12, 8295-8308.

Ritter, C., Radaelli, P.G., Lees, M., Barratt, G., Balakrishnan, G. & Paul, D, (1996). J.
Solid State Chem. 127, 276-282.

Ritter, H., Ihringer, J., Maichle, J.K., Pradl, W., Hoser, A. & Hewat, A.W. (1989). Phys.
B. - Condens. Matter 75, 297-302.

Rodriguez, E., Alvaraez, I., Lopez, M. & Veiga, M. (1999). J. Solid State Chem. 148,
479-486.

Rodriguez, E., Alvaraez, I., Lopez, M., Veiga, M., Pico, C. & Martinez, J. (2001). J.
Mater. Chem, 11, 673-677.

Rodriguez, R., Fernandez, A., Isalgue, A., Rodriguez, J., Labarta, A., Tejada, J. &
Obradors, X. (1985). J. Phys. C: Solid State Phys. 18, L401-L405.

241
Rodriguez-Carvajal, J., Hennion, M., Moussa, F., Moudden,A.H. , Pinsard, L. &
Revcolevschi, A. (1998). Phys. Rev. B. 57(6), R3189-R3192.

Rodriguez-Carvajal, J., Hennion, M., Moussa, F., Pinsard, L. & Revcolevschi, A. (1997).
Physica B. 234-236, 848-850.

Rodriguez-Martinez, L.M. & Attfield, J.P. (1996). Phy. Rev. B 54(22), R15622-R15625.

Rodriguez-Martinez, L.M., Ehrenburg, H. & Attfield, J.P. (2000). Solid State Sci. 2, 16-
Nov.

Rosov, N., Lynn, J.W., Lin, Q., Cao,G.,O'Reilly,J., Pernambuco-Wise,P. & Crow., J.E.
(1992). Phys. Rev. B. 45(2), 982-986.

Ross, N.L. (1998). Phys. Chem. Minerals 25, 597-602.

Roth, R. (1957). J.Res. Nat. Bur. Stand. 58(2), 75-88.

Rudorff, V.W., Lincke, G. & Babel, D. (1963). Z. Anorg. Allr. Chem. 320, 150-170.

Rudorff, W. & Babel, D. (1962). Die Naturwissenschaften , 230.

Rusakov, L.N., Novokhatskii, I.A., Lenev, L.M. & Savinskaya, A.A. (1964). Dokl. Akad,
Nauk SSSR 161(2), 410-412.

Saiki, A., Seto, Y., Seki, H., Ishizawa, N., Kato, M. & Mizutani, N., (1991). Nihon
Kagakkai shi (J.Chem. Soc. Jpn.) 1, 25-31.

Sakai, T., Adachi, G., Shiokawa, J. & Shin-ike,T. (1976). Mat. Res. Bull. 11, 1295-1300.

Sakuma, H., Hashizume, H. & Yamanaka, A. (1990). Acta Cryst. B46, 693-698.

Salamon, M.B. & Jaime, M. (2001). Rev. Mod. Phys. 73, 583-628.

Sales, M., Eguia, G., Quintana, P., Torres-Martinez, L.M. & West, A.R. (1999). J. Solid
State Chem. 143, 62-68.

Salinas-Sanchez, A., Garcia-Munoz, J.L., Rodriguez-Carvajal, J., Saez-Puche, R. &

Martinez, J.L. (1992). J. Solid State Chem. 100, 201-211.

Salje, E. (1977). Acta Cryst. B33, 574-577.

Salvi, S.V., Durge, N.G., Ranade, A.J. & Tendolkar, N.P. (1997). J. Mater. Sci. Lett. 16,
1929-1932.

242
Sarkozy, R.F & Chamberland, B.L. (1973). Mat. Res. Bull. 8, 1351-1360.

Sarkozy, R.F, Moeller, C.W. & Chamberland, B.L. (1974). J. Solid State Chem. 9, 242-
246.

Sarmaa, D.D., Sampathkumaranb, E.V.,Raya, S. & Nagarajanb, R. (2000). Solid State

Communications 114, 465-468.

Sasaki,S.,Prewitt,C. & Bass, J. (1987). Acta Cryst. C43, 1668-1674.

Sasaki,S.,Prewitt,C. & Liebermann,R.C. (1983). American Mineralogist 68, 1189-1198.

Sato, M., Jin, T., Hama, Y. & Uematsu, K. (1993). J. Mater. Chem. 3, 325-326.

Schaak, R. & Mallouk, T.E. (2002). Chem. Mater. , .

Schneck, J., Toledano, J.C. & Joffrin, C. (1982). Phys. Rev. B. 25(3), 1766-1785.

Seinen, P.A., van Berkel, F.P.F., Groen, W.A. & Ijdo, D.J.W. (1987). Mat. Res. Bull. 22,
535-542.

Serpil Gonen, Z., Gopalakrishnan, J., Eichhorn, B.W. & Greene, R.L. (2001). Inorg.
Chem. 40, 4996-5000.

Shannon, R.D. (1976). Acta Cryst. A32, 751-767.

Shannon, R.D. (1993). J. Appl. Phys. 73(1), 348-366.

Shaplygin, I. S. & Lazarev, V. B. (1976). Zh. Neorg. Khim. 21, 2326-2330.

Shevchemko, N.N., Lykova, L.N. & Kovba, L.M. (1974). Zh. Neorg. Khim. 19, 971-975
(528-530).

Shimizu, Y., Syono, Y. & Akimoto, S. (1970). High-P. High Press. 2, 113-120.

Shirane, G. Danner, H., & Pepinski, R. (1957). Physical Review 105(3), 856-860.

Shishido, T., Nakagawa, S., Yoshikawa, A., Horiuchi, H., Tanaka, M., Sasaki, T. &
Fukuda, T. (1995). Nihon Kagakkai shi 9, 697-702.

Slater, P., Irvine, J., Ishihara, T. & Takita, Y. (1998). J. Solid State Chem. 139, 135-143.

Sleight, A.W. & Prewitt, C.T. (1973). J. Solid State Chem. 6, 509-512.

243
Sleight, A.W. & Ward, R. (1964). Inorganic Chem. 3, 292.

Sleight, A.W., Gillson, J.L., & Bierstedt, P.E. (1975). Solid St. Comm. 17, 27-28.

Smith, A. & Welch, J. (1960). Acta Cryst. 13, 653-656.

Song, W.H., dai, J.M., Ye, S.L., Wang, K.Y., Du, J.J. & Sun, Y.P. (2001). J. Appl. Phys.
89(11), 7678-7680.

Steil, M.C., Thevenot, F. & Kleitz, M. (1997). J. Electrochem. Soc. 144, 390-398.

Stitzer, K.E., Smith, M.D. & zur Loye, H.-C. (2002). Solid State Sci. 4, 311-316.

Subias, G., Garcia, J., Blasco, J., Proietti, M. & Sanchez, M. (1999). J. Magn. Magn.
Mat. 196-197, 534-535.

Subramanian, M. A. & Sleight, A.W. (2002). Solid State Sci. 4, 347-351.

Subramanian, M. A., Li, D., Duan, N., Reisner, B.A. & Sleight, A.W. (2000). J. Solid
State Chem. 151, 323-325.

Subramanian, M.A., Ganguli, A.K. & Willmer, K. L. (1991). J. Solid State Chem. 95,
447-451.

Sugiyama, T. & Tsuda, N. (1999). J. Phys. Soc. Jpn. 68, 1306-1312.

Sun, J.Z., Krusin-Elbaum, L., Gupta, A., Xiao, G., Duncome, P.R. & Parkin, S.S.P.
(1998). IBM , .

Sung Im, Y., Hyun Ryu, K., Hong Kim, K. & Hyun Yo, C. (1997). J. Phys. Chem. Solids
58(12), 2079-2083.

Sunstrom, J.E., Ramanujachary, K.V., Greeenblatt, M. & Croft, M. (1998). J. Solid State
Chem. 139, 388-397.

Svensson, C. & Stahl, K. (1988). J. Solid State Chem. 77, 112-116.

Svensson, G. & Werner, P-E. (1990). Mat. Res. Bull. 25, 9-14.

Taguchi, H. (1995). J. Solid State Chem. 118, 367-371.

Taguchi, H. & Nagao, M. (1993). J. Solid State Chem. 105, 392-398.

244
Taguchi, H., Nagao, M. & Shimada, M. (1988). J. Solid State Chem. 76, 284-289.

Taguchi, H., Sonoda, M. & Nagao, M. (1996). J. Solid State Chem. 126, 235-241.

Takahashi, J., Kageyama, K.,Fulii, T., Yamada, T., Kodaira (1997). J. Mater. Sci. 8, 79-
84.

Takano, M.,Okita, M.,Nakayama, N., Bando,Y., Takeda, M.,Yamamoto,O. &

Goodenough, J. (1988). J. Solid State Chem. 73, 140-150.

Takata, M. & Kageyama, K. (1989). J. Am. Ceram. Soc. 72, 1955-1959.

Takeuchi, T., Betourne, E., Tabuchi, M., Kageyama, H., Kobayashi, Y., Coats, A.,
Morrison, F., Sinclair, D.C. & West, A.R. (1999). J. Mater. Sci. 34, 917-924.

Tamura, H., Konoike, T., Sakabe, Y. & Wakino, K. (1984). Comm. Am. Ceram. Soc. , C-
59-C-61.

Tanaka, K. & Marumo, F. (1983). Acta Cryst. A39, 631-641.

Tanaka, K., Konishi, M. & Marumo, F. (1979). Acta Cryst. B35, 1303-1308.

Tauber, A., Tidrow, S.C., Finnegan, R.D. & Wilber, W.D. (1996). Physica C. 256, 340-
344.

Terakura, K., Lee, J., Yu, J., Solovyev, I.V. & Sawada, H. (1999). Materials Science and
Engineering B63, 16-Nov.

Teraoka, Y., Wei, M. & Kagawa, S. (1998). J. Mater. Chem. 8(11), 2323-2325.

Tezuka, K., Henmi, K. & Hinatsu, Y. (2000). J. Solid State Chem. 154, 591-597.

Thirumal, M., Jain, P. & Ganguli, A.K. (2001). Mat. Chem. Phys. 70, .

Thirumal, M., Jawahar, I.N., Sureddiran, K.P., Mohanan, P. & Ganguli, A.K. (2002).
Mater. Res. Bull. 37, 185-191.

Thomas, N. (1996). Acta Cryst. B52, 16-31.

Thomas, N. (1996). Acta Cryst. B52, 954-960.

Thomas, N. & Beitollahi, A. (1994). Acta Cryst. B50, 549-560.

Thomas, N.W. (1989). Acta Cryst. B45, 337-344.

245
Thomas, N.W. (1997). Br. Ceram. Proc. 57, 119-134.

Thornton, G. & Jacobson, A.J. (1978). Acta Cryst. B34, 351-354.

Thornton, G., Tofield, B.C. & Hewat, A.W. (1986). J. Solid State Chem. 61, 301-307.

Thorton, G. & Jacobson, A. (1976). Mat. Res. Bull. 11, 837-842.

Tien, L.-C., Chou, C.-C. & Tsai, D.-S. (2000). J. Am. Ceram. Soc. 83, 2074-2078.

Tokunaga, M., Miura, N., Tomioka, Y. & Tokura, Y. (1998). Phys. Rev. B. 57(9), 5259-
5264.

Tomioka, Y., Asamitsu, A., Kuwahara, H., Moritomo, Y., Kasai, M., Kumai, R. &
Tokura, Y., (1997). Physica B. 237-238, 6-10.

Tomioka, Y., Asamitsu, A., Moritomo, Y., Kuwahara, H. & Tokura, Y. (1995). Phys.
Rev. Lett. 74(25), 5108-5111.

Tomioka, Y., Okuda, T., Okimoto, Y., Kumai, R. & Kobayashi, K.-I. (2000). Phys. Rev.
B. 61(1), 422-427.

Topfer, J. & Goodenough, J.B. (1997). J. Solid State Chem. 130, 117-128.

Tornero, J.D., Cano, F.H., Fayos, J. & Martinez-Ripoll, M. (1978). Ferroelectrics 19,
123-130.

Trolier-McKinstry, S. & Newnham, R.E. (1993). MRS Bulletin April, 27-33.

Troyanchuk, I.O. & Akimov, A.I. (1994). Phys. Solid State 36(8), 1240-1242.

Troyanchuk, I.O., Samsonenko, N.V., Kasper, N.V., Szymczak, H. & Nabialek, A.

(1997). J. Phys.: Condens. Matter 9, 8287-8295.

Troyanchuk, O., Kasper, N.V., Szymczak, H. & Nabialek, N.V. (1997). Low Temp. Phys.
23, 300-302.

Tsukuda, N. & Okazaki, A. (1972). J. Phys. Soc. Jpn. 33(4), 1088-1099.

Uchinao, K. & Cross, L.E. (1980). Jpn. J. Appl. Phys. 19, L171-L173.

Unoki, H. & Sakudo, T. (1967). J. Phys. Soc. Jpn. 23, 546-552.

246
Uyeda, Y., Yauoka, H. & Ueda, Y. (1998). J. Mag. Mag. Mat. 177-181, 1391-1392.

van Aken, B.B., Meetsma, A. & Palstra, T.T.M. (2001). LANL Preprint , 1-4.

van Duivenboden, H.C. & Ijdo, D.J.W. (1986). Acta Cryst. C42, 523-525.

Vanderah, T.A., Febo, W., Chan, J.C., Roth, R.S., Loezos, J.M., Rotter, L.D., Geyer,
R.G. & Minor, D.B. (2000). J. Solid State Chem. 155, 78-85.

Vanderah, T.A., Huang, Q., Wong-Ng, W., Chakoumakos, B.C., Goldfarb, R.B., Geyer,
R.G., Baker-Jarvis, J., Roth, R.S. & Santoro, A. (1995). J. Solid State Chem. 120, 121-
127.

Vasylechko,L.,Akelrud,L.,Morgenroth,W., Bismayer,U.,Matkovskii,A. & Savytskii,D.

(2000). J. Alloys and Compounds 297, 46-52.

Vasylechko,L.,Matkovskii,A.,Savytskii,D.,Suchocki,A. & Wallrafen,F. (1999). J. Alloys

and Compounds 291, 57-65.

Vegas, A. (1986). Acta Cryst. B42, 167-172.

Villafuerte-Castrejon, M.E., Estrada, M.R., Gomez-Lara, J., Duque, J. & Pomes, R.

(1997). J. Solid State Chem. 132, 1-5.

Vincent, H., Bertaut, E.F., Baur, W.H. & Shannon, R.D. (1976). Acta Cryst. B32, 1749-
1755.

Viola, M.C., Martinez-Lope, M.J., Alonso, J.A., Velasco, P., Martinez, J.L., Pedregosa,
J.C., Carbonio, R.E. & Fernandez-Diaz, M.T. (2002). Chem. Mater. 14, 812-818.

Vogt, T. & Schmahl, W.W. (1993). Europhysics Letters 24, 281-285.

Wakeshima, M., Harada, D. & Hinatsu. Y. (1999). J. Alloys Comp. 287, 130-136.

Wakiya, N., Shinozaki, K., Mizutani, N. & Ishizawa, N. (1997). J Am. Ceram. Soc. 80,
3217-3220.

Wang, K.-Y., Arcas, J., Chen, D.-X. & Hernando, A. (1997). J. Alloys and Compounds
252, L26-L28.

Whangbo, M.-H. & Koo, H.-J. (2002). Solid State Sci. 4, 335-346.

Will, G. (1969). Angew. Chem. Int. Edit. 8, 950-961.

247
Wise, P.L., Reaney, I.M., Lee, W.E., Price, T.J., Iddles, D.M. & Cannell, D.S. (2001). J.
Eur. Ceram. Soc. 21, 2629-2632.

Wise, P.L., Reaney, I.M., Lee, W.E., Price, T.J., Iddles, D.M. & Cannell, D.S. (2001). J.
Eur. Ceram. Soc. 21, 2629-2632.

Wiseman, P. & Dickens, P. (1976). J. Solid State Chem. 17, 91-100.

Wittmann, v.U., Rauser, G. & Kemmler-Sack, S. (1981). Z. Anorg. Allg. Chem. 482, 143-
153.

Wolcyrz, M., Horyn, R., Andre, G. & Bouree, F. (1997). J. Solid State Chem. 132, 182-
187.

Wollan, E.O. & Koehler, W.C. (1955). Physical Review 100(2), 545-563.

Wong-Ng, W., Roth, R.S., Vanderah, T.A. & McMurdie, H.F. (2001). J. Res. NIST 106,
1097-1134.

Wood, V.E., Austin, A.E., Collings, E.W. & Brog, K.C. (1973). J. Phys. Chem. Solids 34,
859-868.

Woodley, S.M, Catlow, C.R., Gale, J.D. & Battle, P.D. (2000). Chem. Commun. , 1879-
1880.

Woodward, P. (1997). Acta Cryst. B53, 32-43.

Woodward, P. (1997). Acta Cryst. B53, 44-66.

Woodward, P. (1997). J. Appl. Cryst. 30, 206-207.

Woodward, P.M. Ph. D. Dissertation Oregon State University, Corvallis, OR 1997

Woodward, P. & Vogt, T. (1998). J. Solid State Chem. 138, 207-219.

Woodward, P., Hoffmann, R-D. & Sleight, A.W. (1994). J. Mater. Res. 9(8), 2118-2127.

Woodward, P., Sleight, A.W. & Vogt, T. (1997). J. Solid State Chem. 131, 9-17.

Woodward, P., Vogt, T., Cox, D.E., Arulraj,A., Rao, C.N.R., Karen, P. & Cheetham,
A.K. (1998). Chem. Mater. 10, 3652-3665.

Woodward, P.M., Cox, D.E., Vogt, T., Rao, C.N.R & Cheetham, A.K. (1999). Chem.
Mater. 11, 3538-3528.

248
Xu, H.W., Iwaski, J., Shimizu, T., Satoh, H. & Kamegashira, N. (1995). J. Alloys and
Compounds 221, 274-279.

Yamada, K., Funabiki, S., Horimoto, H., Matsui, T., Okuda, T. & Ichiba, S. (1991).
Chemistry Letters , 801-804.

Yamamoto, A., Khasanova, N.R., Izumi, F., Wu, X.-J., Kamiyama, T., Torii, S. &
Tajima, S. (1999). Chem. Mater. 11, 747-753.

Yamamoto, T., Liimatainen, J., Linden, J., Karppinen, M. & Yamauchi, H. (2000). J.
Mater. Chem. 10, 2342-2345.

Yang, J. & Su, Q. (1992). J. Chin. Rare Earth Soc. 10, 259-261.

Yang, J. H., Choo, W. K., Lee, J. H. & Lee, C. H. (1999). Acta Cryst. B55, 348-354.

Yang, J., Su, Q. & Wang, H. (1992). Chin. Chem. Lett. 3(5), 403-404.

Yoon, K.H., Jung, B.J. & Kim, E.S. (1989). J. Mater. Sci. Lett. 8, 819-822.

Yoshii, K. (2000). J. Alloys Compds. 307, 119-123.

Yoshii, K. & Abe, H. (2001). J. Solid State Chem. 156, 452-457.

Zener, C. (1951). Physical Review 81(4), 440-444.

Zeng, Z., Greenblatt, M., Subramanian & Croft, M. (1999). Phys. Rev. Lett. 82, 3164-
3167.

Zeng, Z., Greenlatt, M., Sunstron IV, J.E. & Croft, M. (2000). J. Solid State Chem. 147,
185-198.

Zhang, H., Teraoka, Y., & Yamazoe, N. (1987). Chem. Lett. , 665-668.

Zhao, C., Feng, S., Chao, Z., Shi, C., Xu, R. & Ni, J. (1996). Chem. Commun. , 1641-
1642.

Zhao, Y. (1998). J. Solid State Chem. 141, 121-132.

Zhao, Y., Weidner, D., Parise, J.B. & Cox, D. (1993). Phys. Earth and Plan. Int. 76, 1-
16.

249
Zhao, Y., Weidner, D., Parise, J.B. & Cox, D. (1993). Phys. Earth and Plan. Int. 76, 17-
34.

Zheng, W., Pang, W. & Meng, G. (1998). Mater. Letters 37, 276-280.

Zhengmin, F. & Wenxiu, L. (1995). Sci. China 38(3), 309-316.

Zhengmin, F., Wenxiu, L. & Dongcai, L. (1983). Sientia Sinica 26, 835-847.

Zhong, Z. & Gallagher, P.K. (1995). J. Mater. Res. 10, 945-952.

Zhou, J.-S. & Goodenough, J.B. (2000). Phys. Rev. B. 62(6), 3834-3838.

Zhurova, E.A., Zavodnik, V.E & Tsirel'son, V.G. (1995). Crystallography Reports 40,
753-760.

Zimmermann, M., Hill, J.P., Gibbs, D., Blume, M., Casa, D., Keimer, B., Murakami,
Y.,Tomioka, Y. & Tokura, Y. (1999). Phys. Rev. Lett. 83, 4872-4875.

Zimmermann, M., Nelson, C.S., Hill, J.P., Gibbs, D., Blume, M., Casa, D., Keimer, B.,
Murakami, Y., Kao, C.-C., Venkataraman, C., Gog, T., Tomioka, Y. & Tokura, Y.
(2001). J. Mag. Mag. Mater. 233, 31-37.

Zubkov, V.G., Berger,I., Pesina, Z., Bazuev, G. & Shveiken, G. (1986). Sov. Phys. Dokl.
31(6), 459-460.

250
 APPENDIX A

A1.1 : Introduction to Lattice Parameter and Fractional Coordinate Equation Derivation.

Equations for the lattice parameters and anion fractional coordinates were derived

based on M-X, M′-X bond distances and the octahedral tilt angle. Equations were derived

for six rock-salt octahedral cation (A2MM′X6) ordered tilt systems (a0a0a0, a0a0c-, a-a-a-,

a0b-b-, a-b+a-, and a+a+a+). The tilt systems were simplified into the component tilts around

the primitive M-X axes (x, y & z) of the octahedra. Octahedral tilting around multiple

axes was represented by combinations of tilting around the axes. Table A1.1 lists the tilt

system and axis (axes) of rotation.

Tilt System Rotation Axis (Axes)

a0a0a0 none
a0 a0 c- 001
a0b-b- 011
a-a-a- 111
a-b+a- 110, 001
a+a+a+ 111

Table A1.1 : Tilt system and rotation axis (axes) of the octahedra.

Tilt system a0a0c- was represented by rotation about the 001 direction of the unit

cell (defined as the z axis of the octahedron), while tilt system a0b-b- was by rotation

251
about the direction of the vector y = z of the octahedron. The magnitude of the octahedral

tilting was represented by a primary tilt angle for the MX6 octahedron. The secondary tilt

angle for M′X6 must be opposite and with a magnitude to ensure corner sharing

requirement is met. A geometrical relationship relates the primary tilt angle (φ) for MX6

to the secondary tilt angle (θ) of the M′X6 octahedron.

The Cartesian positions of the anions after octahedral tilting of a single

octahedron were obtained by multiplication of a rotation matrix times the untilted

Cartesian positions. The unit cell is obtained by adding together the Cartesian vector

components representing the direction of the M-X and M′-X bonds of the tilted

octahedra. The fractional coordinates of the anions are obtained by taking the expressions

for the Cartesian vectors describing the anion positions and dividing by the Cartesian

vectors representing the lattice parameters.

The derivation for the A2MM′X6 in tilt system a0a0c- (space group I4/m) is

described in section A.1.2. Equation derivations of multiple axis tilt systems were

obtained similarly, but the intermediate matrices are not shown in detail due to

complexity. Only the final equations, used in the SPuDS source code, are shown for the

multiple axis tilt systems.

A.1.2 : Equation derivation for lattice parameters & fractional coordinates of a0a0c- (I4/m)

The tilt system a0a0c- can be represented by octahedral tilting about a single (z)

axis (001 direction of the unit cell) of the untilted perovskite. The unit cell dimensions

252
are approximately a = b ≈ √2ap, and c = 2ap, where ap is the primitive untilted cubic

perovskite cell length. A view showing one layer is shown in figure A1.1.

Figure A1.1: View looking down the c-axis of a0a0c- (I4/m). The black spheres are
anions, single hatch are M, and cross hatches are M′. The M-X and M′-X bonds are single
hatch and cross hatches, respectively. The Cartesian orientations of the MX6 and M′X6
octahedra are shown as x and y.

A rotation matrix about the z axis is defined below, where φ is the tilt angle of the

octahedral rotation.

 cos( φ) sin( φ) 0
 
 
rotate_001 = −sin( φ) cos( φ) 0
 
 0 0 1


The M cation is defined at the origin of the octahedron. The general Cartesian

positions prior to octahedral tilting of the X anion located along the x direction are

defined for MX6 as vectors coords_x, coords_y & coords_z and M′X6 coords_x2,

253
coords_y2 & coords_z2. Vectors coords_x and coords_x2 have vector components ai,

aj & ak and a2i, a2j & a2k in the x, y & z directions (respectively) representing the

Cartesian position of X anion relative to the M at the origin. The vector components for

the X anions along the y and z direction follow similarly and are shown below.

 ai   bi   ci 
     
     
coords_x =  aj  coords_y =  bj  coords_z =  cj 
     
 ak   bk   ck 
     

 a2i   b2i   c2i 

     
     
coords_x2 :=  a2j  coords_y2 :=  b2j  coords_z2 :=  c2j 
     
 a2k   b2k   c2k 
     

The MX6 and M′X6 octahedra are defined with three X anions located at a

distance of dmx and d2mx (respectively) along each of the three Cartesian axes (x, y & z).

The general Cartesian positions of the anions of the MX6 prior to octahedral tilting are

defined below.

 dmx   0   0 
     
     
coords_x :=  0  coords_y :=  dmx  coords_z :=  0 
     
 0   0   dmx 
     

Multiplication of the rotation matrix times the Cartesian positions of the anions

gives equations describing the anion positions after tilting φ degrees .

[cart_x_pos] = [rotate_001] [coords_x]

[cart_y_pos] = [rotate_001] [coords_y]

[cart_z_pos] = [rotate_001] [coords_z]

The cart_x_pos, cart_y_pos, & cart_z_pos vectors of the Cartesian positions of

the X anions of MX6 after octahedral tilting are shown in the following equations.

254
  cos ( φ) dmx   sin( φ) dmx   0 
     
     
cart_x_pos := − sin( φ) dmx  cart_y_pos := cos ( φ) dmx  cart_z_pos :=  0 
     
 0   0   dmx 
     

The tilt angle of the second octahedra is related to the tilt angle of the first

octahedra by the simple geometrical relationship calculated from the need to maintain

corner sharing connectivity. In tilt system a0a0c-, the secondary tilt angle θ in terms of φ,

dmx & d2mx is shown in the following equation, where dmx and d2mx are the M-X and

M′-X bond distances.

 dmx sin( φ) 
θ = -arcsin 
 d2mx 

Cartesian positions of the anion of M′O6 prior to tilting are shown in the vectors

coords_x2, coords_y2 & coords_z2.

 0 
d2mx  0   
     
    coords_z2 =  0 
   
coords_x2 =  0  coords_y2 = d2mx  
    d2mx
 0   0   
   

Multiplication of the rotation matrix times the Cartesian positions of the anions

results in equations describing the anion positions after tilting θ degrees. Substitution of

the expression for θ gives the anion positions of M′X6 in terms of φ after tilting .

[cart_x2_neg] = [rotate_001] [coords_x2]

[cart_y2_neg] = [rotate_001] [coords_y2]

[cart_z2_neg] = [rotate_001] [coords_z2]

255
  2 2   −sin( φ) dmx 
 dmx sin( φ)   
 1− d2mx  
 2  2 2   0 
 d2mx   dmx sin( φ)   
cart_x2_neg =   cart_y2_neg = 
  1−

d2mx  
   2  cart_z2_neg =  0 
 sin( φ) dmx   d2mx   
    d2mx
     
 0   0 
   

A numerical example is useful in demonstrating the way in which the individual

vectors were added to obtain the equations for the lattice parameters. Consider the

example where dmx = 2.0 Å and d2mx = 2.1 Å with a rotation of φ = 10° for the MX6

octahedron. The vectors a1, b1 & c1 and a2, b2 & c2 represent the Cartesian positions

after tilting of the X anion located 2.0 Å (MX6) and 2.1 Å (M′X6) along the x, y, and z

directions, respectively.

 1.969615506 .3472963554  0 
     
     
a1 = -.3472963554 b1 = 1.969615506 c1 =  0 
     
 0   0   2.0 
     

2.071083108 -.3472963554  0 
     
     
a2 = .3472963554 b2 =  2.071083108 c2 =  0 
     
 0   0   2.1 
     

Consider octahedral tilting in figure A1.1, where the individual components of the

octahedra are labeled. Addition of the individual M-X vector components in the x, y, & z

directions results in vectors representing the lattice parameters. Individual vector

components (i, j and k) for each octahedra are added together by connecting the vectors

from the M cation to a second M cation at the next corner of the unit cell. Note that the

vector component is positive when traversing M-X, and negative when traversing X-M.

Individual vector components in the x, y & z directions for the lattice parameters are

shown in the following equations.

ai_tot = [ai - (-a2i) + b2i - (-bi)]

256
 aj_tot = [aj - (-a2j) + b2j - (-bj)]

ak_tot = [ak - (-a2k) + b2k - (-bk)]

2 2 2 2
d2mx − dmx + dmx cos( φ)
ai_tot = cos( φ) dmx + d2mx
2
d2mx

2 2 2 2
d2mx − dmx + dmx cos( φ)
aj_tot = cos( φ) dmx + d2mx
2
d2mx

ak_tot = 0

bi_tot = [ai - (-a2i) + (-b2i) - (bi)]

bj_tot = [aj - (-a2j) + (-b2j) - (bj)]

bk_tot = [ak - (-a2k) + (-b2k) - (bk)]

2 2 2 2
d2mx − dmx + dmx cos( φ)
bi_tot = cos( φ) dmx + d2mx
2
d2mx

2 2 2 2
d2mx − dmx + dmx cos( φ)
bj_tot = − d2mx − cos( φ) dmx
2
d2mx

bk_tot = 0

ci_tot = [ci - (-c2i) + c2i - (-ci)]

cj_tot = [cj - (-c2j) + c2j - (-cj)]

ck_tot = [ck - (-c2k) + c2k - (-ck)]

ci_tot = 0

cj_tot = 0

ck_tot = 2 dmx + 2 d2mx

257
 The a, b, and c lattice parameters are equal to the magnitudes (lengths) of the

sums of the individual vector components of each unit cell edge.

a = √(ai_tot2 + aj_tot2 + ak_tot2)

b = √(bi_tot2 + bj_tot2 + bk_tot2)

c = √(ci_tot2 + cj_tot2 + ck_tot2)

2 2 2 2 2
a = 2 dmx cos( 2 φ) + 2 cos( φ) dmx 4 d2mx − 2 dmx + 2 dmx cos( 2 φ) + 2 d2mx

2 2 2 2 2
b = 2 dmx cos( 2 φ) + 2 cos( φ) dmx 4 d2mx − 2 dmx + 2 dmx cos( 2 φ) + 2 d2mx

c = 2 dmx + 2 d2mx

The crystal system is tetragonal, thus the angles between the lattice parameters

must be orthogonal. The angle is obtained by calculating the dot product of the lattice

parameter vectors to check this requirement. Each lattice parameter vector was

orthogonal to the other two.

angle_a_b = arcos[(ai_tot*bi_tot + aj_tot*bj_tot + ak_tot*bk_tot)/(a_lattice*b_lattice)]

angle_a_c = arcos[(ai_tot*ci_tot + aj_tot*cj_tot + ak_tot*ck_tot)/(a_lattice*c_lattice)]

angle_b_c = arcos[(bi_tot*ci_tot + bj_tot*cj_tot + bk_tot*ck_tot)/(b_lattice*c_lattice)]

Equations for the anion fractional position are obtained by the method described

in the following text. The general matrix [f_cart] converts from fractional to Cartesian

coordinates and is shown below.

 ai_tot bi_tot ci_tot

 
 
f_cart =  aj_tot bj_tot cj_tot

 
ak_tot bk_tot ck_tot
 

258
 The matrix [convert_fract_cartesian] has substituted values for I4/m (tilt system

a0a0c-) and is listed below.

 2 2 2 2 2 2 2 2 
 d2mx − dmx + dmx cos ( φ) d2mx − dmx + dmx cos ( φ) 
 cos ( φ) dmx + d2mx cos ( φ) dmx + d2mx 0 
 2 2 
 d2mx d2mx 
 
 
 
convert_fract_cartesian =  2 2 2 2 2 2 2 2 

 d2mx − dmx + dmx cos ( φ) d2mx − dmx + dmx cos ( φ) 
 cos ( φ) dmx + d2mx − d2mx − cos ( φ) dmx 0 
 2 2 
 d2mx d2mx 
 
 
 0 0 2 dmx + 2 d2mx 


The inverse of the [convert_fract_cartesian] matrix converts from Cartesian

coordinates to fractional coordinates when multiplied times the Cartesian position.

 2 2 2 2 2 2 2 2 
 d2mx − dmx + dmx cos ( φ ) d2mx − dmx + dmx cos ( φ ) 
 cos ( φ ) dmx + d2mx cos ( φ ) dmx + d2mx 
 2 2 
 
1 d2mx 1 d2mx 
 0 
2 2 2 2 2 2 2 2 2 2 
 d2mx − dmx + dmx cos ( φ ) d2mx − dmx + dmx cos ( φ ) 
 
 2 dmx 2 cos ( φ ) 2 + 2 cos ( φ ) dmx 2
d2mx + d2mx − dmx
2 2 dmx 2 cos ( φ ) 2 + 2 cos ( φ ) dmx 2
d2mx + d2mx − dmx
2 
 2 2 
 d2mx d2mx 
 
 
 
 2 2 2 2 2 2 2 2 
 d2mx − dmx + dmx cos ( φ ) d2mx − dmx + dmx cos ( φ ) 
inverse_convert_fract_cartesian =  
 cos ( φ ) dmx + d2mx cos ( φ ) dmx + d2mx 
 2 2 
1 d2mx 1 d2mx 
 
 − 0 
2 2 2 2 2 2 2 2 2 2 
 d2mx − dmx + dmx cos ( φ ) d2mx − dmx + dmx cos ( φ ) 
 2 2 2 2 2 2 dmx 2 cos ( φ ) 2 + 2 cos ( φ ) dmx 2 2 
 dmx cos ( φ ) + 2 cos ( φ ) dmx d2mx + d2mx − dmx d2mx + d2mx − dmx 
 2 2 
 d2mx d2mx 
 
 
 
 1 1 
 0 0 
 2 dmx + d2mx 


Multiplication of the [inverse_convert_fract_cartesian] [cartesian_z_pos]

results in the equation describing the fractional coordinate of the X anion located along

the z axis.

[x4e] = [inverse_convert_fract_cartesian][cartesian_z_pos]

 0 
 
 
 0 
x4e =  

1 
 dmx 
 
2 dmx + d2mx 

259
 Multiplication of the [inverse_convert_fract_cartesian] [cartesian_x_pos]

results in the matrix [x8h] with the fractional coordinates of the X anion located a

distance dmx along the x direction as shown below.

[x8h] = [inverse_convert_fract_cartesian][cartesian_x_pos]

  2 2 2 2 2 2 2 2 
  2 d2mx − dmx + dmx cos( φ) d2mx − dmx + dmx cos( φ) 

 dmx  −cos( φ) dmx + cos( φ) dmx sin( φ) − cos( φ) d2mx + d2mx sin( φ) 
 1  2 2
  d2mx d2mx  
− 
 2 2 2 2 2 
 d2mx − dmx + dmx cos( φ) 
 2 2 2 2 
 2 dmx cos( φ) + 2 cos( φ) dmx d2mx + d2mx − dmx 
 2 
 d2mx 
 
 
x8h =   2 2 2 2 2 2 2 2  
  d2mx − dmx + dmx cos( φ) d2mx − dmx + dmx cos( φ) 
 dmx  cos( φ) 2 dmx + cos( φ) dmx sin( φ) + cos( φ) d2mx + d2mx sin( φ) 
  2 2
 1  d2mx d2mx  

 
 2 2 2 2 2 
 2 2 d2mx − dmx + dmx cos( φ) 2 2 
 2 dmx cos( φ) + 2 cos( φ) dmx d2mx + d2mx − dmx 
 
 2 
 d2mx 
 
 
 0 

A numerical evaluation with d(1) = 2.0 Å, d(2) = 2.1 Å and phi = 10.0° of the

lattice parameters and fractional coordinates illustrates the resulting fractional coordinate

and lattice parameters. The lattice parameters were a = 5.714 Å, b = 5.714 Å and c =

8.200 Å. The a = b symmetry requirement of a tetragonal crystal system was met in the

calculation. The fractional coordinates are and X4e= (0, 0, 0.244) and X8h = (0.201,

0.287, 0). The approximate atom fractional coordinate positions for all of the rock-salt

ordered have been previously tabulated by Woodward (1997).

A1.2 : Lattice parameters and anions fractional coordinates equations for Fm 3 m (a0a0a0).

Approximate unit cell dimensions are a ≈ 2ap, where ap is the primitive unit cell

length. The cations fractional positions are M(4a) (0, 0, 0), M′(4b) (½, 0, 0) and A(8c)

(¼,¼, ¼). Anion fractional positions are X(24e) (x, 0, 0) (x ≈ ¼). The lattice parameters

260
and anion fractional coordinates as a function of the M-X [d(1)] and M′-X [d(2)] bond

distances are given below.

a = 2.*d(1) + 2.*d(2)

X24e(1) = d(1)/(2.*d(1) + 2.*d(2))

A1.3 : Lattice parameters and anions fractional coordinates equations for I4/m (a0a0c-).

Approximate unit cell dimensions are a ≈ √2ap and c ≈ 2ap, where ap is the

primitive unit cell length. The cations fractional positions are M(2a) (0, 0, 0), M′(2b) (0,

0, ½) and A(4b) (0, ½, ¼). Anion fractional positions are X(4e) (0, 0, z) (z ≈ ¼) and

X(8h) (x, y, 0) (x ≈ ¼ and z ≈ ¼). The lattice parameters and anion fractional coordinates

as a function of the M-X (d(1)) and M′-X (d(2)) bond distances and tilt angle are given

below.

a = sqrt(2.)*sqrt(2.*d(1)**2*cos(phi)**2+2.*cos(phi)*d(1)*sqrt((d(2)**2-

d(1)**2+d(1)**2*cos(phi)**2)/(d(2)**2))*d(2)+d(2)**2-d(1)**2)

c = 2.*(d(1)+d(2))

X(4e) z = 1./2.*d(1)/(d(1)+d(2))

X(8h) x = 1./2.*d(1)*(d(2)*sqrt((d(2)**2-

d(1)**2+d(1)**2*cos(phi)**2)/(d(2)**2))*cos(phi)-sqrt((d(2)**2-

d(1)**2+d(1)**2*cos(phi)**2)/(d(2)**2))*d(2)*sin(phi)+cos(phi)**2*d(1)-

cos(phi)*d(1)*sin(phi))/(2*d(1)**2*cos(phi)**2+2*cos(phi)*d(1)*sqrt((d(2)**2-

d(1)**2+d(1)**2*cos(phi)**2)/(d(2)**2))*d(2)+d(2)**2-d(1)**2)

261
X(8h) y = 1./2.*d(1)*(cos(phi)**2*d(1)+cos(phi)*d(1)*sin(phi)+d(2)*sqrt((d(2)**2-

d(1)**2+d(1)**2*cos(phi)**2)/(d(2)**2))*cos(phi)+sqrt((d(2)**2-

d(1)**2+d(1)**2*cos(phi)**2)/(d(2)**2))*d(2)*sin(phi))/(2*d(1)**2*cos(phi)**2+2*co

s(phi)*d(1)*sqrt((d(2)**2-d(1)**2+d(1)**2*cos(phi)**2)/(d(2)**2))*d(2)+d(2)**2-

d(1)**2)

A1.4 : Lattice parameters and anions fractional coordinates equations for I2/m (a-b0a-).

Approximate unit cell dimensions are a ≈ √2ap, b = √2ap, c ≈ 2ap and β ≈ 90,

where ap is the primitive unit cell length. The cations fractional positions are M(2a) (0, 0,

0), M′(2b) (½, ½, 0) and A(4i) (x, 0, z) where x ≈ ½ and z ≈ ¼. Anion fractional

positions are X(4i) (x, 0, z) and X(8h) (x, y, z) (x ≈ ¼, y ≈ ¼ and z ≈ 0). The lattice

parameters and anion fractional coordinates as a function of the M-X (d(1)) and M′-X

(d(2)) bond distances and tilt angles are given below.

a = sqrt(4.*d(1)*cos(theta)*d(2)*cos(phi)+2.*d(1)**2+ 4.*sin(theta)*

d(1)*sin(phi)*d(2)+2.*d(2)**2)

b = sqrt(2.)*(d(1)+d(2))

c = sqrt(2.*(-sin(theta)*sqrt(2.)*d(1)-

sin(phi)*sqrt(2.)*d(2))**2+(2.*d(1)*cos(theta)+2.*d(2)*cos(phi))**2)

X(4i) x = -1./2.*sqrt(2.)*d(1)*d(2)*(sin(theta)*cos(phi)-

cos(theta)*sin(phi))/(2.*d(1)*cos(theta)*d(2)*cos(phi)+

d(1)**2+2.*sin(theta)*d(1)*sin(phi)*d(2)+d(2)**2)

262
X(4i) z = 1./2.*d(1)*(d(1)+sin(theta)*sin(phi)*d(2)+cos(theta)*

d(2)*cos(phi))/(2.*d(1)*cos(theta)*d(2)*cos(phi)+d(1)**2+2.*

sin(theta)*d(1)*sin(phi)*d(2)+d(2)**2)

X(8j) x = 1./2.*d(1)*(d(1)+sin(theta)*sin(phi)*d(2)+cos(theta)*

d(2)*cos(phi))/(2.*d(1)*cos(theta)*d(2)*cos(phi)+d(1)**2+2.*

sin(theta)*d(1)*sin(phi)*d(2)+d(2)**2)

X(8j)y = 1./2.*d(1)/(d(1)+d(2))

X(8j)z = 1./4.*sqrt(2.)*d(1)*d(2)*(sin(theta)*cos(phi)-

cos(theta)*sin(phi))/(2.*d(1)*cos(theta)*d(2)*cos(phi)+d(1)**2+2.*sin(theta)*d(1)*sin(p

hi)*d(2)+d(2)**2)

A1.5 : Lattice parameters and anions fractional coordinates equations for R 3 (a-a-a-).

Approximate unit cell dimensions are a ≈ √2ap, b ≈ √2ap and c ≈ √2ap, where ap is

the primitive unit cell length. The cations fractional positions are M(1a) (0, 0, 0), M′(1b)

(½, ½, ½) and A(2c) (x, x, x) where x ≈ ¼. Anion fractional position is X(6f) (x, y, z) (x

≈ ¾, y ≈ ¾ and z ≈ ¼). The lattice parameters and anion fractional coordinates as a

function of the M-X [d(1)] and M′-X [d(2)] bond distances and tilt angle are given below.

The hexagonal setting (for distance calculations) and rhombohedral settings are shown

below.

Hexagonal

ah = sqrt(4.*d(1)**2*cos(rho)**2+4.*d(1)*cos(rho)*d(2)*sqrt((d(2)**2-

d(1)**2+d(1)**2*cos(rho)**2)/(d(2)**2))+2.*d(2)**2-2.*d(1)**2)

263
ch = sqrt(3.)*sqrt((2.*d(1)+2.*d(2))**2)

Rhombohedral

ar = ah/(2.*sin(alpha/2.))

α = 2.*ASIN(3./2.*1./(sqrt(3.+ch**2/(ah**2))))

Hexagonal

X(6f)h x = 1./3.*d(1)*(d(1)*cos(rho)*sin(rho)*sqrt(3.)+d(1)*
cos(rho)**2+d(2)*sqrt((d(2)**2-
d(1)**2+d(1)**2*cos(rho)**2)/(d(2)**2))*sin(rho)*sqrt(3.)+cos(rho)*sqrt((d(2)**2-
d(1)**2+d(1)**2*cos(rho)**2)/(d(2)**2))*d(2))/(2.*d(1)**2*cos(rho)**2+2.*d(1)*cos(r
ho)*d(2)*sqrt((d(2)**2-d(1)**2+d(1)**2*cos(rho)**2)/ (d(2)**2))+d(2)**2-d(1)**2)

X(6f)h y = 2./3.*(d(1)*cos(rho)+d(2)*sqrt((d(2)**2-
d(1)**2+d(1)**2*cos(rho)**2)/(d(2)**2)))*d(1)*cos(rho)/(2.*d(1)**2*cos(rho)**2+2.*
d(1)*cos(rho)*d(2)*sqrt((d(2)**2-d(1)**2+d(1)**2*cos(rho)**2)/(d(2)**2))+d(2)**2-
d(1)**2)

X(6f)h z = 1./6.*d(1)/(d(1)+d(2))

Rhombohedral

X(6f) x = X(6f)h x x + X(6f)h z

X(6f) y = -X(6f)h x + X(6f)h y + X(6f)h z
X(6f) z = -X(6f)h y + X(6f)h z

A1.6 : Lattice parameters and anions fractional coordinates equations for P21/n (a-a-b+).

A tilt about the 110 is followed by secondary correction tilt about the 001 axis,

similar to the one made by O’Keeffe & Hyde (1977) to maintain corner connectivity of

the octahedra. Approximate unit cell dimensions are a ≈ √2ap, b ≈ √2ap and c ≈ 2ap,

where ap is the primitive unit cell length. The cation fractional positions are M(2c) (0, ½,

0), M′(2d) (½, 0, 0) and A(4e) (x, y, z) where x ≈ ½, y ≈ ½ and z ≈ ¼. Anion fractional

positions are X(4e) (x, y, z) where (x ≈ ¼, y ≈ ¼ and z ≈ 0; x′ ≈ ¼, y′ ≈ ¾ and z′ ≈ 0; x″ ≈

½, y″ ≈ 0 and z″ ≈ ¼). The lattice parameters and anion fractional coordinates as a

264
function of the M-X [d(1)] and M′-X [d(2)] bond distances and tilt angle are given below.

Substitution of a = sqrt((2.+cos(rho)**2)/(4.+4.*cos(rho)+cos(rho)**2)), b =

sqrt((2.+cos(eta)**2)/(4.+4.*cos(eta)+cos(eta)**2)), c = cos(eta), e = sin(eta), f =

cos(rho), g = sin(rho), r = sqrt(3.), s = d(1), t = d(2) were made into the following

expressions.

X(6f) x = -[1./6.*s*(-6.*t*a*f**2-3.*t*a*c**2*f**2+t*b*e*c*r*f**2+2.*t*b*e*r*f**2-
9.*s*f*b*c-6.*t*a*c**2*f-12.*t*a*f+3.*t*b*e*c*f*g-
18.*s*f*b+t*a*c**2*f*r*g+2.*t*a*f*r*g+6.*t*b*e*f*g+2.*t*b*e*c*r+4.*t*b*e*r+2.*t*
a*c**2*r*g+4.*t*a*r*g)/(6.*s**2*b*f+t*b*c**2*s*f**2+2.*t*b*c**2*s+3.*t**2*a*c*f
+2.*s*a*f*t*c**2+s*a*f**2*t*c**2-
t*a*c*e*s*g*f+2.*t*b*c*s*f**2+3.*s**2*b*f*c+4.*t*b*c*s-s*b*f*g*t*e*c-
2.*t*a*c*e*s*g+6.*t**2*a*c+2.*s*a*f**2*t+4.*s*a*f*t-2.*s*b*f*g*t*e)]

X(6f) y = -[1./2.*s*(t*b*c**2*f**2+2.*t*b*c*f**2+t*a*c*e*r*f**2-
t*a*c*e*f*g+3.*s*f*b*c+2.*t*a*c*e*f*r+6.*s*f*b+r*t*b*c**2*f*g+2.*r*t*b*c*f*g+4.
*t*b*c+2.*t*b*c**2-
2*t*a*c*e*g)/(6.*s**2*b*f+t*b*c**2*s*f**2+2.*t*b*c**2*s+3.*t**2*a*c*f+2.*s*a*f*
t*c**2+s*a*f**2*t*c**2-t*a*c*e*s*g*f+2.*t*b*c*s*f**2+3.*s**2*b*f*c+4.*t*b*c*s-
s*b*f*g*t*e*c-2.*t*a*c*e*s*g+6.*t**2*a*c+2.*s*a*f**2*t+4.*s*a*f*t-
2.*s*b*f*g*t*e)+0.5] + 1.0

X(6f) z = [1./6.*t*s*(4.*e*s*g+8.*a*s*b-4.*t*e*g-8.*t*a*b+4.*s*b*a*f-
4.*s*b*a*f**3+2.*e*s*g*f**2-8.*s*b*a*f**2+12.*r*s*b*e*a*f-
2.*t*a*b*f**3+6.*r*s*b*e*a*f**2-2.*t*f**2*e*g+4.*t*r*e*f-
4.*t*a*b*f+2.*t*r*f**3*e-4.*t*a*b*f**2-2*t*r*a*b*f**2*g+4.*a*c**2*s*b-
4.*t*r*a*b*f*g+2.*r*s*g*c**3+4.*r*s*g*c+e*s*g*f**2*c**2+2.*a*c**3*s*b+2.*e*s*
g*c**2+r*s*b*a*c**2*e*f**3-r*s*b*a*c**2*e*f+2.*r*s*g*c*f**2-
2.*r*s*b*a*c**2*e+2.*r*s*b*c*e*a*f**3-s*f**3*b*c**3*a+a*c**3*s*b*f-
2.*s*b*c*a*f**3+2.*s*b*a*f*c-2.*s*f**3*b*a*c**2+2*a*c**2*s*b*f-
4.*s*b*a*f**2*c-4.*s*f**2*b*a*c**2+6.*t*r*b*g*a*c**2+12.*t*r*b*g*a*c-
4.*t*a*b*c+4.*b*a*t*c**3+8.*t*a*b*c**2+4.*a*s*b*c-2.*t*c**2*e*g-4.*r*s*b*a*c*e-
t*a*b*c*f**3+t*r*f**3*e*c**2+4.*t*r*b*g*a*c*f+7.*t*r*b*g*a*c**2*f+2.*t*r*b*g*a
*c**3*f-
t*r*a*b*c*f**2*g+2*t*r*e*f*c**2+t*r*a*b*c**3*f**2*g+7.*r*s*b*c*e*a*f**2+
2*t*r*a*b*f**2*g*c**2+2.*b*a*t*c**2*f**3-
t*c**2*e*g*f**2+4.*r*s*b*c*e*a*f+2.*r*s*b*a*c**2*e*f**2+4.*b*a*f**2*c**2*t-2.
*s*f**2*b*c**3*a+4.*t*a*b*c**2*f-
2.*t*a*b*c*f+2.*b*a*t*c**3*f**2+2.*b*a*t*c**3*f+r*s*g*c**3*f**2-
2.*b*a*t*c*f**2+t*c**3*a*f**3*b)]/[(-8.*t**2*c**2*a*b*s+2.*t*s**2*e*g*f**2-

265
4.*s**2*f**3*b*a*t-8.*s**2*f**2*b*a*t-6.*s**3*f**2-10.*s*f*b*c**2*t**2*a-
2*t**2*c**2*a*f**3*b*s-4.*s**2*f*t*c-2.*s*f**2*b*t**2*a*c**3-
7.*s*f**2*b*c**2*t**2*a-3.*t**3*c**2*f**2-t**2*c**3*a*f**3*b*s-
s**2*f**3*b*c**3*a*t-2.*t**2*c**3*s*f+2.*t**2*s*c**2*e*g-
7.*s**2*f**2*b*a*t*c**2-4.*t**2*c**3*a*b*s-4.*t**2*c*s*f-6.*t**3*c**2-
2.*t**2*c*f**3*s-2.*s**2*f*c**3*t-3.*s**3*f**2*c**2+2.*s**2*f*b*c**2*t*a*e*g-
12.*t*c*a*s**2*b*f+t**2*s*c**2*e*g*f**2+2.*t**2*c**2*a*s*b*f*g*e+t**2*c**2*a*
f**2*s*b*g*e+s**2*f**2*b*c**2*t*a*e*g+2.*t**2*c*a*f**2*s*b*g*e-
6.*s*f**2*b*t**2*a*c-2.*s**2*f**3*b*c*a*t+2.*s**2*f**2*b*t*a*c*e*g+4.
*s**2*f*b*t*a*c*e*g-2.*s**2*f**3*b*a*t*c**2+t*s**2*e*g*f**2*c**2-
2.*s*f*b*t**2*a*c**3-2.*s**2*f**3*t*c-s**2*f**3*t*c**3-2.*s**2*f**2*b*c**3*a*t-
t**2*c**3*f**3*s-12.*s*f*b*t**2*a*c-
10.*s**2*f**2*b*c*a*t+4.*t**2*c*a*s*b*f*g*e-6.*t*c**2*a*s**2*b*f)]

X(6f)′ x =
(1./6.*s*(2*t*b*e*r*f**2+t*b*e*c*r*f**2+3.*t*a*c**2*f**2+6.*t*a*f**2+9*s*f*b*c-
6.*t*b*e*f*g+2.*t*a*f*r*g+18.*s*f*b+6.*t*a*c**2*f+t*a*c**2*f*r*g+12.*t*a*f-
3.*t*b*e*c*f*g+4.*t*a*r*g+4.*t*b*e*r+2.*t*b*e*c*r+2.*t*a*c**2*r*g)/(6.*s**2*b*f+t
*b*c**2*s*f**2+2.*t*b*c**2*s+3.*t**2*a*c*f+2.*s*a*f*t*c**2+s*a*f**2*t*c**2-
t*a*c*e*s*g*f+2.*t*b*c*s*f**2+3.*s**2*b*f*c+4.*t*b*c*s-s*b*f*g*t*e*c-
2.*t*a*c*e*s*g+6.*t**2*a*c+2.*s*a*f**2*t+4.*s*a*f*t-2.*s*b*f*g*t*e))

X(6f)′ y = (1./2.*s*(-t*a*c*e*r*f**2+2.*t*b*c*f**2+t*b*c**2*f**2+6.*s*f*b-
2.*r*t*b*c*f*g-r*t*b*c**2*f*g+3.*s*f*b*c-2.*t*a*c*e*f*r-
t*a*c*e*f*g+4.*t*b*c+2.*t*b*c**2-
2*t*a*c*e*g)/(6.*s**2*b*f+t*b*c**2*s*f**2+2.*t*b*c**2*s+3.*t**2*a*c*f+2.*s*a*f*
t*c**2+s*a*f**2*t*c**2-t*a*c*e*s*g*f+2.*t*b*c*s*f**2+3.*s**2*b*f*c+4.*t*b*c*s-
s*b*f*g*t*e*c-2.*t*a*c*e*s*g+6.*t**2*a*c+2.*s*a*f**2*t+4.*s*a*f*t-
2.*s*b*f*g*t*e)+ 0.5)

X(6f)′ z = [1./6.*t*s*(4.*e*s*g+8.*a*s*b-4.*t*e*g-8.*t*a*b+4.*s*b*a*f-
4.*s*b*a*f**3+2.*e*s*g*f**2-8.*s*b*a*f**2+12.*r*s*b*e*a*f-
2.*t*a*b*f**3+6.*r*s*b*e*a*f**2-2.*t*f**2*e*g+4.*t*r*e*f-
4.*t*a*b*f+2.*t*r*f**3*e-4.*t*a*b*f**2-2*t*r*a*b*f**2*g+4.*a*c**2*s*b-
4.*t*r*a*b*f*g+2.*r*s*g*c**3+4.*r*s*g*c+e*s*g*f**2*c**2+2.*a*c**3*s*b+2.*e*s*
g*c**2+r*s*b*a*c**2*e*f**3-r*s*b*a*c**2*e*f+2.*r*s*g*c*f**2-
2.*r*s*b*a*c**2*e+2.*r*s*b*c*e*a*f**3-s*f**3*b*c**3*a+a*c**3*s*b*f-
2.*s*b*c*a*f**3+2.*s*b*a*f*c-2.*s*f**3*b*a*c**2+2*a*c**2*s*b*f-
4.*s*b*a*f**2*c-4.*s*f**2*b*a*c**2+6.*t*r*b*g*a*c**2+12.*t*r*b*g*a*c-
4.*t*a*b*c+4.*b*a*t*c**3+8.*t*a*b*c**2+4.*a*s*b*c-2.*t*c**2*e*g-4.*r*s*b*a*c*e-
t*a*b*c*f**3+t*r*f**3*e*c**2+4.*t*r*b*g*a*c*f+7.*t*r*b*g*a*c**2*f+2.*t*r*b*g*a
*c**3*f-
t*r*a*b*c*f**2*g+2*t*r*e*f*c**2+t*r*a*b*c**3*f**2*g+7.*r*s*b*c*e*a*f**2+2*t*r*
a*b*f**2*g*c**2+2.*b*a*t*c**2*f**3-

266
t*c**2*e*g*f**2+4.*r*s*b*c*e*a*f+2.*r*s*b*a*c**2*e*f**2+4.*b*a*f**2*c**2*t-2.
*s*f**2*b*c**3*a+4.*t*a*b*c**2*f-
2.*t*a*b*c*f+2.*b*a*t*c**3*f**2+2.*b*a*t*c**3*f+r*s*g*c**3*f**2-
2.*b*a*t*c*f**2+t*c**3*a*f**3*b)] /[(-8.*t**2*c**2*a*b*s+2.*t*s**2*e*g*f**2-
4.*s**2*f**3*b*a*t-8.*s**2*f**2*b*a*t-6.*s**3*f**2-10.*s*f*b*c**2*t**2*a-
2*t**2*c**2*a*f**3*b*s-4.*s**2*f*t*c-2.*s*f**2*b*t**2*a*c**3-
7.*s*f**2*b*c**2*t**2*a-3.*t**3*c**2*f**2-t**2*c**3*a*f**3*b*s-
s**2*f**3*b*c**3*a*t-2.*t**2*c**3*s*f+2.*t**2*s*c**2*e*g-
7.*s**2*f**2*b*a*t*c**2-4.*t**2*c**3*a*b*s-4.*t**2*c*s*f-6.*t**3*c**2-
2.*t**2*c*f**3*s-2.*s**2*f*c**3*t-3.*s**3*f**2*c**2+2.*s**2*f*b*c**2*t*a*e*g-
12.*t*c*a*s**2*b*f+t**2*s*c**2*e*g*f**2+2.*t**2*c**2*a*s*b*f*g*e+t**2*c**2*a*
f**2*s*b*g*e+s**2*f**2*b*c**2*t*a*e*g+2.*t**2*c*a*f**2*s*b*g*e-
6.*s*f**2*b*t**2*a*c-2.*s**2*f**3*b*c*a*t+2.*s**2*f**2*b*t*a*c*e*g+4.
*s**2*f*b*t*a*c*e*g-2.*s**2*f**3*b*a*t*c**2+t*s**2*e*g*f**2*c**2-
2.*s*f*b*t**2*a*c**3-2.*s**2*f**3*t*c-s**2*f**3*t*c**3-2.*s**2*f**2*b*c**3*a*t-
t**2*c**3*f**3*s-12.*s*f*b*t**2*a*c-
10.*s**2*f**2*b*c*a*t+4.*t**2*c*a*s*b*f*g*e-6.*t*c**2*a*s**2*b*f)]

X(6f)″ x = [1./3*(t*b*e*c*f**2+2.*t*b*e*f**2+t*a*c**2*f*g+2.*t*a*f*g+3.*s*g*b*c-
2.*t*b*e+4.*t*a*g-t*b*e*c+2.*t*a*c**2*g+6.*s*g*b)*s*r/(-6.*s**2*b*f-
t*b*c**2*s*f**2-2.*t*b*c**2*s-3.*t**2*a*c*f-2.*s*a*f*t*c**2-
s*a*f**2*t*c**2+t*a*c*e*s*g*f-2.*t*b*c*s*f**2-3.*s**2*b*f*c-
4.*t*b*c*s+s*b*f*g*t*e*c+2.*t*a*c*e*s*g-6.*t**2*a*c-2.*s*a*f**2*t-
4.*s*a*f*t+2.*s*b*f*g*t*e)] + 0.5

X(6f)″ y = t*c*s*(-b*c*f**2-2.*b*f**2+a*e*g*f+2.*b+b*c+2.*a*e*g)/(-6.*s**2*b*f-
t*b*c**2*s*f**2-2.*t*b*c**2*s-3.*t**2*a*c*f-2.*s*a*f*t*c**2-
s*a*f**2*t*c**2+t*a*c*e*s*g*f-2.*t*b*c*s*f**2-3.*s**2*b*f*c-
4.*t*b*c*s+s*b*f*g*t*e*c+2.*t*a*c*e*s*g-6.*t**2*a*c-2.*s*a*f**2*t-
4.*s*a*f*t+2.*s*b*f*g*t*e)

X(6f)″ z = -[1./6.*s*(2*t*a*c**3*f*s*b+8.*s*b*c*t*a+4.*t*a*c**3*s*b-
12.*s*g*b*c*t*e*a-2.*t*e*c**2*s*g+8.*t*a*c**2*s*b-6.*t*a*c**2*e*s*b*f*g-
t*e*c**2*s*g*f**2+4.*s*f**2*b*t*c*a+2.*b*f**3*a*s*t*c**2+2.*b*f**2*a*s*t*c**3
+2.*b*f**3*a*c*s*t+4.*s*b*c*t*a*f+3.*t*c**3*f**3*s+6.*s*f**3*t*c-
4.*t**2*c**3*a*b-8.*t**2*c**2*a*b-4.*t**2*c**2*e*g+4.*t**2*a*b*c-
6.*s*b*f**2*g*t*a*c*e+14.*b*f*a*t**2*c**2+2.*t**2*c**3*a*f**3*b+4.*s*f*b*c**2
*t*a+b*f**3*a*s*t*c**3+4.*b*f**2*a*t**2*c**3+4.*t**2*c**2*a*f**3*b-
3.*t*a*c**2*e*f**2*s*b*g-
2.*t**2*f**2*e*g*c**2+14.*b*f**2*a*t**2*c+4.*s*f**2*b*c**2*t*a-
18.*s*b*f*g*t*a*c*e-
2.*b*f*a*t**2*c**3+17.*b*f**2*a*t**2*c**2+38.*b*f*a*t**2*c-
2.*t**2*a*b*c*f**3+6.*s*f*t*c**3+6.*s**2*c**2+12.*s*f*t*c-4.*t**2*f**2*e*g-
4.*t**2*a*b*f**3+4.*b*f*a*t**2-8.*b*f**2*a*t**2-2.*t*e*s*g*f**2+8.*t*a*f*s*b-

267
12.*s*g*b*t*e*a*f+8.*b*f**2*a*s*t+4.*b*f**3*a*s*t-
24.*s*g*b*t*e*a+8.*t**2*a*b+16.*t*a*s*b-4.*t*e*s*g-
8.*t**2*e*g+6.*s**2*f**2+12.*s**2+3.*s**2*f**2*c**2)]/[(8.*t**2*c**2*a*b*s-
2.*t*s**2*e*g*f**2+4.*s**2*f**3*b*a*t+8.*s**2*f**2*b*a*t+6.*s**3*f**2+10.*s*f*
b*c**2*t**2*a+2.*t**2*c**2*a*f**3*b*s+4.*s**2*f*t*c+2.*s*f**2*b*t**2*a*c**3+7
.*s*f**2*b*c**2*t**2*a+3.*t**3*c**2*f**2+t**2*c**3*a*f**3*b*s+s**2*f**3*b*c*
*3*a*t+2.*t**2*c**3*s*f-
2.*t**2*s*c**2*e*g+7.*s**2*f**2*b*a*t*c**2+4.*t**2*c**3*a*b*s+4.*t**2*c*s*f+6.
*t**3*c**2+2.*t**2*c*f**3*s+2.*s**2*f*c**3*t+3.*s**3*f**2*c**2-
2.*s**2*f*b*c**2*t*a*e*g+12.*t*c*a*s**2*b*f-t**2*s*c**2*e*g*f**2-
2.*t**2*c**2*a*s*b*f*g*e-t**2*c**2*a*f**2*s*b*g*e-s**2*f**2*b*c**2*t*a*e*g-
2.*t**2*c*a*f**2*s*b*g*e+6.*s*f**2*b*t**2*a*c+2.*s**2*f**3*b*c*a*t-
2.*s**2*f**2*b*t*a*c*e*g-4.*s**2*f*b*t*a*c*e*g+2.*s**2*f**3*b*a*t*c**2-
t*s**2*e*g*f**2*c**2+2.*s*f*b*t**2*a*c**3+2.*s**2*f**3*t*c+s**2*f**3*t*c**3+2
.*s**2*f**2*b*c**3*a*t+t**2*c**3*f**3*s+12.*s*f*b*t**2*a*c+10.*s**2*f**2*b*c*
a*t-4.*t**2*c*a*s*b*f*g*e+6.*t*c**2*a*s**2*b*f)] + 0.5

A1.7 : Lattice parameters and anions fractional coordinates equations for Pn 3 (a+a+a+).

Approximate unit cell dimensions are a ≈ 2ap, where ap is the primitive unit cell

length. The cations fractional positions are M(4b) (0, 0, 0), M′(4c) (½, ½, ½), A(2a) (¼,

¼, ¼) and A′(6d) (¼, ¾, ¾). Anion fractional positions are X(4e) (x, y, z) where (x ≈ ¼, y

≈ ½ and z ≈ ½). The lattice parameters and anion fractional coordinates as a function of

the M-X [d(1)] and M′-X [d(2)] bond distances and tilt angle are given below.

a = 2./3.*sqrt(8.*d(1)*d(2)*cos(phi)+9.*d(1)**2+8.*d(1)* cos(rho)*d(2)+ 9.*d(2)**2

+4.* d(1)*cos(rho)*d(2)*cos(phi)-2.*d(1)*d(2)+12.*d(1)*sin(rho) *d(2)*sin(phi))

X(24h) x = -1./2.*d(1)*(-4.*d(2)**2*cos(rho)+4.*d(2)**2*cos(phi)-
8.*d(2)**2*cos(phi)**2+27.*d(1)**2+24.*d(1)*cos(rho)*d(2)+13.*d(2)**2+8.*d(2)**2
*cos(phi)**2*cos(rho)+14.*d(2)**2*cos(phi)*cos(rho)+12.*d(1)*cos(rho)*d(2)*cos(phi)
+36.*d(1)*sin(rho)*d(2)*sin(phi)-
6.*d(2)**2*sin(phi)*sin(rho)+24.*d(2)**2*cos(phi)*sin(phi)*sin(rho)+
24.*d(1)*d(2)*cos(phi)-6.*d(1)*d(2))/(-12.*d(1)*d(2)**2*cos(phi)-
36.*d(1)**2*d(2)*cos(phi)+9.*d(2)*d(1)**2-27.*d(1)**3+13.*d(2)**3-
18.*d(1)**2*cos(rho)*d(2)*cos(phi)-42.*d(1)*cos(rho) *d(2)**2*cos(phi)-
24.*d(1)*cos(rho)* d(2)**2*cos(phi)**2+18.*d(1)*sin(rho)*d(2)**2*sin(phi)-
54.*d(1)**2*sin(rho)*d(2)*sin(phi)-39.*d(1)*d(2)**2-
72.*d(2)**2*cos(phi)*d(1)*sin(rho)*sin(phi)+24.*d(1)*d(2)**2

268
*cos(phi)**2-36.*d(1)**2*cos(rho)*d(2)+12.* d(1)*cos(rho)*d(2)**2-
48.*d(2)**3*cos(phi)- 24.*d(2)**3*cos(phi)**2+32.*d(2)**3*cos(phi)**3)

X(24h) y = -1./2.*d(1)*d(2)*(-4.*d(2)*cos(phi)+12.*d(1)*cos(phi)-3.*d(1)-
3.*d(1)*cos(rho)*cos(phi)+7.*cos(rho)*d(2)*cos(phi)+4.*d(2)*cos(phi)**2*cos(rho)-
3.*sin(rho)*d(2)*sin(phi)-9.*d(1)*sin(rho)*sin(phi)-
13.*d(2)+3.*d(2)*sin(phi)*sqrt(3.)*cos(rho)+7.*d(2)*cos(phi)*sin(rho)*sqrt(3.)+4.*d(2)
*cos(phi)**2*sin(rho)*sqrt(3.)+3.*d(1)*sin(rho)*sqrt(3.)*cos(phi)-
12.*d(2)*cos(phi)*sin(phi)*sqrt(3.)*cos(rho)-9.*d(1)*sin(phi)*sqrt(3.)*cos(rho)-
2.*d(2)*sin(rho)*sqrt(3.)+6.*d(1)*sin(rho)*sqrt(3.)+12.*d(2)*cos(phi)*sin(phi)*sin(rho)
+8.*d(2)*cos(phi)**2-6.*d(1)*cos(rho)-2.*cos(rho)*d(2))/(-12.*d(1)*d(2)**2*cos(phi)-
36.*d(1)**2*d(2)*cos(phi)+9.*d(2)*d(1)**2-27.*d(1)**3+13.*d(2)**3-
18.*d(1)**2*cos(rho)*d(2)*cos(phi)-42.*d(1)*cos(rho)*d(2)**2*cos(phi)-
24.*d(1)*cos(rho)*d(2)**2*cos(phi)**2+18.*d(1)*sin(rho)*d(2)**2*sin(phi)-
54.*d(1)**2*sin(rho)*d(2)*sin(phi)-39.*d(1)*d(2)**2-72.*d(2)**2*cos(phi)*d(1)
*sin(rho)*sin(phi)+24.*d(1)*d(2)**2*cos(phi)**2-
36.*d(1)**2*cos(rho)*d(2)+12.*d(1)*cos(rho)*d(2)**2-48.*d(2)**3*cos(phi)-
24.*d(2)**3*cos(phi)**2+32.*d(2)**3*cos(phi)**3)+.5

X(24h) z = -1./2.*d(1)*d(2)*(-4.*d(2)*cos(phi)+12.*d(1)*cos(phi)-3.*d(1)-
3.*d(1)*cos(rho)*cos(phi)+7.*cos(rho)*d(2)*cos(phi)+4.*d(2)*cos(phi)**2*cos(rho)-
3.*sin(rho)*d(2)*sin(phi)-9.*d(1)*sin(rho)*sin(phi)-
13.*d(2)+12.*d(2)*cos(phi)*sin(phi)*sin(rho)+8.*d(2)*cos(phi)**2-6.*d(1)*cos(rho)-
2.*cos(rho)*d(2)-
4.*d(2)*cos(phi)**2*sin(rho)*sqrt(3.)+12.*d(2)*cos(phi)*sin(phi)*sqrt(3.)*cos(rho)+9.*
d(1)*sin(phi)*sqrt(3.)*cos(rho)-7.*d(2)*cos(phi)*sin(rho)*sqrt(3.)-3.*d(2)*
sin(phi)*sqrt(3.)*cos(rho)-3.*d(1)* sin(rho)*sqrt(3.)*cos(phi)+2*d(2)*sin(rho)*sqrt(3.)-
6.*d(1)*sin(rho)*sqrt(3.))/(-12.*d(1)*d(2)**2*cos(phi)-
36.*d(1)**2*d(2)*cos(phi)+9.*d(2)*d(1)**2-27.*d(1)**3+13.*d(2)**3-
18.*d(1)**2*cos(rho)*d(2)*cos(phi)-42.*d(1)*cos(rho)*d(2)**2*cos(phi)-
24.*d(1)*cos(rho)*d(2)**2*cos(phi)**2+18.*d(1)*sin(rho)*d(2)**2*sin(phi)-
54.*d(1)**2*sin(rho)*d(2)*sin(phi)-39.*d(1)*d(2)**2-
72.*d(2)**2*cos(phi)*d(1)*sin(rho)*sin(phi)+ 24.*d(1)*d(2)**2*cos(phi)**2-
36.*d(1)**2*cos(rho)*d(2)+12.*d(1)*cos(rho)*d(2)**2-48.*d(2)**3*cos(phi)-
24.*d(2)**3*cos(phi)**2+32.*d(2)**3*cos(phi)**3)+.5

A.1.8 : References

O'Keefe, M., & Hyde, B.G. (1977). Acta Cryst. B33, 3802-3813.

Woodward, P. M. (1997). Acta Cryst. B53, 32-43.

269