Combustion, Explosion, and Shock Waves, Vol. 36, No.

5, 2000

P h y s i c o c h e m i c a l P r o p e r t i e s o f A1203 P o w d e r P r o d u c e d by Explosive Synthesis A. A. Bukaemskii, 1 A. G. Beloshapko, 1 and A. P. Puzyr '2

UDC 534.222.2+621.762

Translated from Fizika Goreniya i Vzryva, Vol. 36, No. 5, pp. 119-125, September October, 2000. Original article submitted June 24, 1999. T h e p h a s e , dispersive, a n d m o r p h o l o g i c a l f e a t u r e s o f an a l u m i n u m oxide p o w d e r p r o d u c e d b y explosive synthesis a r e e x a m i n e d e x p e r i m e n t a l l y . I t is s h o w n t h a t t h e p a r t i c l e - s i z e d i s t r i b u t i o n has t h r e e d i s t i n c t m a x i m a , w h i c h a r e d u e t o different c o m b u s t i o n r e g i m e s . T h e r e l a t i o n s h i p b e t w e e n t h e sizes a n d m o r p h o l o g y o f t h e s t a r t i n g p o w d e r a n d t h e p r o d u c t is d e t e r m i n e d . C o n s i d e r a b l e a t t e n t i o n is given t o a s t u d y o f t h e u l t r a f i n e f r a c t i o n o f t h e p r o d u c t p o w d e r . T h e ultrafine particles are s h o w n to h a v e a r e g u l a r spherical s h a p e , a n d s i n t e r s a r e n o t revealed. Besides s p h e r i c a l particles, t h e s y n t h e s i z e d p o w d e r c o n t a i n s f a c e t e d crystals. X - r a y p h a s e analysis o f t h e u l t r a f i n e f r a c t i o n o f t h e s y n t h e s i z e d p o w d e r s s h o w s t h a t this f r a c t i o n c o n s i s t s o n l y o f t h e m e t a s t a b l e phases o f t h e o x i d e - - t h e 5- o r o x y n i t r i d e m o d i f i c a t i o n s . T h e 5 - m o d i f i c a t i o n differs f r o m t h a t d e s c r i b e d in t h e l i t e r a t u r e .

The existing physicochemical methods for synthesis of ultrafine powders are mostly based on the evaporation of and subsequent condensation of materials. At the last stage, an additional supply of energy can be due to various chemical reactions, for example, metal oxidation [1]. The physicochemical properties of ultrafine powders are largely determined by the particle size and method of preparation. Exactly conditions of synthesis determine the characteristics of the material produced, such as the average size, degree of agglomeration, and crystal structure of the particles. A distinctive feature of metal oxide powders produced by physicochemical methods is the polydispersity of particles. This effect was observed for powders produced by both plasmachemical synthesis [2] and electrical explosion of conductors [3]. However, these studies were focused mainly on the ultrafine fraction of the synthesized materials, which is of great scientific and practical interest. On the other hand, an extensive investigation has been pertbrmed on the 1Physicotechnical Institute of the Krasnoyarsk State University, Krasnoyarsk 660036. :Institute of Biophysics, Siberian Division, Russian Academy of Sciences, Krasnoyarsk 660036. 660

morphological structure of products from metal combustion in reactive media and their relationship with oxidation regimes [4]. In this case, in contrast, the main focus was on studying coarse particles of oxides; as regards the submicron fraction, it is only reported that it is formed in gas-phase metal combustion [4]. Explosive synthesis is described in [5]. It is shown that the synthesis product contains both ultrafine and coarse (with particle sizes larger than 1 #m) tractions. Their concentrations depend on the technological parameters of synthesis and are related to the state of aggregation of the starting metal at the shock-wave front [6]. The goal of the present work is to study the phase, dispersive, and morphological features of aluminum oxide powders, including the ultrafine powder produced by explosive synthesis.

EXPERIMENTAL

TECHNIQUES

Powders synthesized under conditions that are most favorable for production of the material in the ultrafine state have been studied most extensively [6]. In this case, the product of explosive synthesis is a highly porous white powder with large

0010-5082/00/360.5-0660 $25.00 (~) 2000 Kluwer Academic/Plenum Publishers

P h y s i c o c h e m i c a l P r o p e r t i e s o f A1203 P o w d e r (larger t h a n 1 #m) gray inclusions; its bulk density is ~0.05 g / c m 3. The dispersity of the synthesized powder particles was studied by sieve and sedimentation analyses [7] and electron microscopy. The complete particle size distribution was determined as follows. An averaged powder sample was diluted in distilled water, treated by ultrasound, and poured through a standard set of sieves (GOST 3584-53). To make the extraction of the ultrafine powder fraction more complete and decrease the amount of the liquid used, the suspension was made to settle in the lower part of the vessel and the upper, more liquid, fraction was used for pouring. This operation was repeated several times. The last pouring was performed using pure distilled water. Next, the powder was dried on sieves, and the masses of particular fractions were determined. The suspension was examined by sedimentation analysis [7]. After that, we constructed the complete particle size distribution. The morphological structure of particles of different fractions was studied by scanning and transmission microscopic techniques using a JEM-100C electron microscope with an EM-ASID-4 scanning adapter. The specific surface area of the powders was determined by the BET (Brunauer, Emmett, and Teller) method. The phase compositions of the synthesized powder and its fractions with different particle sizes were determined on a DRON-3 apparatus.

661

MAIN

RESULTS

Figure 1 shows a typical complete particle size distribution. It is evident that the distribution has three distinct maxima for mass-averaged particle sizes dm = 0.25, 22, and 360 pm. To check the correctness of sedimentation measurement results for the subnficron range and to r ~ F 0.9 0.6

-2

-1

log (din [pro])

Fig. 1. Typical particle size distribution ( F is the distribution function density).

fine the size distribution for ultrafine particles, we measured the particle diameters by electron micrographs and processed the results by statistical methods [7]. In all experiments, the particle size distributions are shown to be log-normal (a typical distribution is given in [5]). For the experiment considered, the parameters are as follows: numberaveraged size d0.5 = 70 nm and variance cr = 1.9. For log-normal distributions, the number-averaged and mass-averaged sizes are related by the formula In dm -- In d0.5 + 3 In2 a, which holds true in our case. This, in particular, indicates that the sedimentation of the powder in a dispersion liquid is adequately described by the equations used in sedimentation analysis. In the powder produced, the ultrafine fraction is separated from the large-size fraction (see Fig. 1). This simplifies the sizing of the powder particles by sedimentation methods. From Fig. 2a, it is evident t h a t the particles have a regular spherical shape, and sinters are not observed. T h e particles are gathered into chains and buildups, perhaps, under the action of electrostatic forces. Diluted in a dispersion medium (H20, C2H5OH, etc.), the ultrafine particles form a stable suspension. Besides spherical particles, the synthesized powder contains faceted crystals. Most often, the "shadow" of the object has 8 sides with similar sizes. Regular hexagons and irregular octagons whose opposite sides are parallel and equal are encountered rather rarely. For fractions with particle sizes larger than 50 #m, the particle size distribution determined by sieve analysis is also log-normal. Depending on the conditions of synthesis, a = 1.2-1.4. The morphological structure of particles with sizes larger t h a n 1 #m was studied by scanning electron microscopy. Typical photographs are given in Fig. 2b~d. The large fraction consists of "foam"-type tbrmations, hollow spherical particles, and their fragments. The appearance of a spherical shell is shown in Fig. 2c. The inner diameter of the sphere is 170 pm and its wall thickness is 20 #m (see Fig. 2d). The wall has a porous structure and consists of blocks wittl a typical size of ~ 1 0 pro. Sedimentation of the synthesis product leads to separation of particles with typical sizes of 1 50 #m. which forms a considerable gray sediment. Electron microscopy show that this fraction of the powder consisted predominantly of continuous particles with a nearly spherical shape (see Fig. 2b). The number of hollow spheres and foamed agglomerates is insignifi(:ant,. In separation by sedimentation, p a r t of the powder rose to the surface, forming a film. This fihn

662

Bukaemskii, Beloshapko, and Puzyr'

?e
I 200
nI~l

I~

Fig. 2. Photographs of particles from various fractions of the synthesized powder.

consists of hollow spherical shells with typical sizes of 10-20/~m. The shell walls are transparent in the optical frequency range (according to data of transmission electron microscopy). X-ray phase analysis of the synthesized powders on a DRON-3 apparatus with subsequent identification by the ASTM catalog shows that the powders contain the a- and (i-modifications of the oxide and the oxynitride phase Als/3-~/3Oa-xNx, where 0.22 < x < 0.5. The presence and quantitative proportion of the phases are determined by the condi-

tions of synthesis [5, 6]. X-ray phase analysis shows that the ultrafine fraction of the synthesized powders consists only of the metastable phases of the oxide ((i- or oxynitride modifications) or their mixture (depending on the conditions of synthesis). The x-ray pattern of the ultrafine powder fraction identified as the (f-phase of aluminum oxide, and data from the ASTM catalog are given in Fig. 3a. The lattice of this phase has the following parameters: a = 15.8 /~, b = 11.7 /~, and c = 7.9 ~ [8]. We note that the (i-modification is most frequently

P h y s i c o c h e m i c a l P r o p e r t i e s o f Al~_O3 P o w d e r I 200

663

I O0

100 50

200

was reported in [9, 10]. This was explained by the small grain sizes [9] and the nonstoichiometry of the oxide inside a particle [10]. For our powders, this can also be related to the presence of a certain amount of ultrafine material in the pores of large particles. We measured the specific surface area (Ssp) of the ultrafine portion of the synthesized powder. Irrespective of the phase composition, Ssp ~ 20 m2/g. Assuming that the particle shape is spherical, we calculated the surface-averaged size of the particles: d~ = 106 nm. The specific surface area of the coarse traction of the powder is much smaller than that of the ultrafine fraction: Ssp = 5 m2/g. This value is apparently determined by the typical sizes of the crystal grains and the presence of ultraflne particles in pores and not by the particle size.

O0

0 100 50

DISCUSSION OF RESULTS AND CONCLUSIONS It is reasonable to assume that each maximum in the particle size distribution (see Fig. 1) corresponds to a particular aluminum combustion regime. According to [4], gas-phase combustion is the most probable form of steady-state aluminum combustion in a high-temperature gas flow, and its occurrence depends on heat exchange with the ambient medium and the size of the starting metal particles. For large particles, a considerable part of the heat is transferred from the surface into the depth of the material. This leads to surface oxidation of the solid metal. For small particles there is no chance for a considerable oxide layer to form over the heating time. Depending on the heat exchange with the ambient medium, the following combustion mechanisms are possible. If the surface temperature does not exceed the melting point of the oxide, then, as in the previous case, surface oxidation occurs. At a surface temperature higher than the melting point of the oxide, the particles are coated with a liquid oxide l~'er. Complete oxidation of the particles proceeds faster since the rate of oxygen diffusion through the liquid oxide is higher than through the solid oxide. Partial evaporation of the oxide is possible. Combustion results in formation of continuous drops of oxide. which are similar in size to the starting particles. If the starting metal contains a dissolved gas, the latter is evolved upon heating. This leads to swelling of the shell of the liquid oxide. Combustion results in formation of hollow spherical shells of the oxide. High-rate heating of the particles can lead to overheating of the metal and rupture of the oxide shell

0
is

I,
2;

,J

I
'

is

'

5's

,,I
'

6;

20, deg Fig. 3. Results of x-ray analysis and data from the ASTM catalog for the synthesized powder stabilized in the 5-phase of the oxide (a) and in the oxynitride modification (b). encountered in powders produced by explosive synthesis. It is also formed using various grades of the starting powder over a broad range of initial explosive chamber pressures (2--7 a t m ) in various ga.~,s (air and CO2) with the masses of the explosive and the starting aluminum powder varied by a factor of 20. The x-ray pattern of the ultrafine powder fraction identified as the oxynitride phase Als/a-~/304-xN~. (Fig. 3b) shows a cubic crystal lattice with parameter a = 7.92 A. This phase is less typical of explosive synthesis and is formed in those eases where expansion of the synthesis products proceeds under less favorable cooling conditions, tot example, when the initial air pressure in the explosive chamber is 1 atm. X-ray phase analysis was used to study powders of individual sieve fractions. It is shown that when the particle size is smaller t h a n 50 #m, the powder is stabilized only in the a-phase. W i t h increase in the particle size, the portion of the metastablc & phase increases and reaches 30%. The presence of metastable phases in large particles of metal oxides

664
of the metal. In the further process, the metal burns in the gas phase with formation of submicron oxide particles. In explosive synthesis, the starting material was PAP-1 aluminum powder. It consists of flake-shaped particles whose average size is 7.2 x 10.6 x 1 #m, according to electron microscopy d a t a (volumeaveraged size 4.3 #m). The thickness of the flakes is 1 tim, and, according to the classification in [4], they can be classed as small particles. Special experiments on synthesis in media t h a t are inert toward aluminum (N2 and CO2) show t h a t at the shockwave front there is partial sintering of the aluminum powder particles. Owing to heat release due to the reaction with oxygen, the material is heated above the melting point of aluminum, and the flake shape of the particles becomes spherical. Subsequent combustion results in formation of continuous oxide particles with sizes of 5-50 #m (middle m a x i m u m in Fig. 1). The shape of the formations - - spherical or "irregular" - - is determined by the particle surface temperature during the oxidation. The formation of hollow spherical shells with a typical d i a m e t e r of 10-20 #m seems to occurs during combustion of individual particles of the aluminum powder. The first maximum of the distribution in Fig. 1 (din = 0.25 pm) corresponds to gas-phase metal combustion. This combustion regime is typical of small metal particles. The shape of the particles is determined by the characteristic t e m p e r a t u r e in the zone of their growth [11]. The regular spherical shape of the ultrafine particles indicates t h a t the temperature of the explosive synthesis is higher than the melting point of the oxide. The growth of faceted crystals occurs at temperatures lower than the melting point of the oxide. The spherical shape of the particles (see Fig. 2a) and the log-normal size distribution show that the particles grow from the oxide vapor by the mechanism of liquid-drop coalescence [1]. The combustion products of large aluminum particles have been studied extensively [4]. Hollow spherical shells were observed in the combustion of coarse (d = 3 mm) aluminum particles in an air flow [12, 13]. The formations with a foamed structure are similar in appearance to the agglomerates of magnesium oxide produced by low-temperature combustion of large (d = 3~6 mm) metal particles with loam formation [14]. The size of the oxide formations is approximately equal to the initial size of the metal particles. In explosive synthesis, large metal particles are formed during heating and sintering of the starting aluminum powder particles at the shock-wave front. In [6, 15], it is shown that temperatures sufficient for

Bukaemskii, Beloshapko, and Puzyr'

complete melting of the starting metal occur on the boundary with the explosive. In the subsequent expansion, the liquid metal layer breaks up into drops, whose size appears to correlate with the size of large particles of the synthesized oxide (third maximum in Fig. 1). The breakup process determines, according to [7], the log-normal size distribution of coarse fraction particles. The experiments were performed for metals with an initial density of about 0.4 1.0 g / c m 3. As a result, the shock-compressed molten metal contains a considerable amount of t r a p p e d gas. Evolution of this gas during expansion and subsequent combustion promotes synthesis of foamed formations and hollow spherical shells rather than continuous oxide particles. The only stable crystalline modification of" aluminum oxide is the a-phase or corundum. The remaining modifications undergo an irreversible transformation into corundum at 1300~ In our experiments, the (i- and oxynitride phases belong to the group of high-temperature anhydrous oxides [16]. The ultrafine state is characterized by stabilization of aluminum oxide in the 7-, x-, 0-, and (i-phases. For example, a mixture of ~- and (f-phases is mentioned in a description of ultrafine powders produced by electrical explosion of conductors [3] and by the plasmachemical method [2]. The position of the x-ray diffraction lines of the synthesized powder is in good agreement with the d a t a of the ASTM catalog (card No. 16-394) for the 5phase (see Fig. 3a). However, the intensities of some lines differ significantly, for example, for angles 20 = 38.90 ~ and 34.48 ~ The x-ray p a t t e r n also shows a significant peak (20 = 32.06~ a single peak that was not identified by the A S T M catalog. For thorough examination of the phase structure of the synthesized material, the powder was annealed at T = 1150~ This temperature is sufficient to initiate all phase transitions in aluminum oxide (for example, T ~ 0 = 1050~ [16]), except for transformation into the a-modification. The x-ray patterns of the starting and annealed powders are identical. Annealing does not lead to a change in the amplitudes of the diffraction lines at 20 = 38.90 ~ and 34.48 ~ and the reflection at 20 = 32.06 ~ does not disappear. In addition, the amplitudes of the "controversial" peaks were normalized by the amplitudes of the most significant lines of the (f-phase Aij = Ii/Ij, where i = 38.90 ~ 34.48 ~ and 32.06 ~ and j = 67.04 ~ 45.68 ~ and 36.52 ~ Results from processing of more than 20 x-ray patterns of powders produced in the above-mentioned ranges of technological parameters show that the values of Aij remain constant within

P h y s i c o c h e m i c a l P r o p e r t i e s o f AI~.O3 P o w d e r

665

the experimental error. The results obtained suggest that the x-ray pattern given in Fig. 3a corresponds to the 5-phase of aluminum oxide produced by explosive synthesis. The above-mentioned differences between the synthesized powder and reference data can be due to the fact that the 6-phase is meta~table and the position of the peaks on its x-ray pattern depends strongly on the production process [16]. Thus, in the ASTM catalog there are three cards (No. 16-394, No. 4-0877, and No. 20-43) for the 5-modification, which differ in both the position and amplitudes of diffraction lines. The metastability of the phases produced by explosive synthesis is determined by the high rate of quenching of the material, which is in turn due to the small size of the particles and their interaction with the high-velocity gas flow during expansion. For example, the production of the d-phase by fast cooling of the molten oxide is reported in [17]. The absence of the ~/-phase in the synthesized powders is unambiguously shown by annealing at 1150~ since the -~ --~ 5 transition occurs at a temperature of 900~ [16]. Annealing at 1300~ stabilizes the synthesized powder in the pure a-phase. Aluminum nitride (card No. 34-679) is identiffed by the ASTM catalog in the same manner as the oxynitride modification of the oxide Als/3_~/304_~N~ (card No. 18-52; Fig. 3b). For a more correct identification, the nitrogen content in the synthesized powder was determined by the Kjeldahl method. The amount of nitrogen in this case was 1.11%, and fbr a control sample of aluminum nitride, it was 37.3%. However, the percentage of nitrogen in the oxynitride phase is lower than the value indicated in the catalog (Xmi~ = 0.22). This can be attributed to the conventionality of the boundaries along the x in the A S T M catalog or to the presence of "insoluble" nitrogen in our powders. Thus, we can argue that the powders produced by explosive synthesis contain the oxynitride phase rather than aluminum nitride. The nitrogen content in the 5-modification of the oxide produced by the explosive method is 2.2%. According to the d a t a of" [18], the 5-modification can contain up to 7% nitrogen, and in the diagram of the system, the AI~O:~ A1N phase (:an be treated as the oxynitride phase. Thus, in the present work, we studied the phase, dispersive, and morphological properties of aluminum oxide powder produced by explosive synthesis. The complete particle size distribution has three distinct maxima, which are related to the different aluminum combustion regimes. The relationship

between the sizes and morphology of the starting and synthesized powders is determined. X-ray phase analysis of the ultrafine fraction of the synthesized powders shows that it consists only of the metastable oxide phases the 5- or ox3."nitride modifications. The 5-modification, which is most typical of powders produced by explosive synthesis, differs from its literature analogs. This work was supported by the INTAS Foundation (Grant No. 97-1754).

REFERENCES 1. I. D. Morokhov, L. I. Trusov, and V. N. Lapovok, Physical Phenomena in Highly Disperse Media [in Russian], I~nergoatomizdat, Moscow (1984). 2. M. I. Aivazov, V. V. Volod'ko, et al., "Forms of growth of ultrafine powders in plmsmachemical reactors," Po,vshk. Metall., No. 1, 1 5 (1981). 3. Yu. A. Kotov and N. A. Yavorskii, "Particles produced by electrical explosion of conductors," Fiz. Khim. Obrab. Mater., No. 4, 24-k29 (1978). 4. F. Pokhil, A. F. Belyaev, et al., Combustion of Metal Powders in Reactive Media [in Russian], Nauka. Moscow (1972). 5. A. G. Beloshapko, A. A. Buakemskii, and A. M. Staver, "Formation of ultradisperse compounds by shock-wave loading of porous aluminum. Investigation of the particles produced," Fiz. Goreniya I/zryva, 26, No. 4, 93-98 (1990). 6. A. A. Buakemskii, "Production of new ultradisperse materials and investigation of their properties," Candidate's Dissertation in Phys.-Math. Sci., Krasnoyarsk (1995). 7. P. A. Kouzov, Fundamentals of the Analysis of the Grain-Size Composition for Commercial Dusts and Ground Materials [in Russian], Khimiya, Leningrad (1974). 8. J. E. Bonevich and L. D. Marks, "The sintering behavior of ultrafine alumina particles," J. Mater. Res.. 7. No. 6, 1489--1500 (1992). 9. S. N. Kul'kov, V. F. Nesterenko, M. P. Bondar', et al., "Explosion activation of quench-hardened ZrO2 Y203 ceramic submicron powders," Fiz. Goreniya Vzr~va, 29, No. 6, 66-72 (1993). 10. A. G. Beloshapko, A. A. Bukaemskii, I. G. Kuz'min, and A. M. Staver, "Dynamic synthesis of zirconium dioxide powders," Fiz. Gorv.n.tya Vzryva, 29. No. 6. 78-81 (lq93). 11. V. P. Grachukho, A. M. Stepanov, and A. A. Khvatsev. "Estimation of the dispersity of metal particle combustion products," Fiz. Goreniya Vzryva, 12, No. 4, 519 530 (1976).

666

Bukaemskii,

Beloshapko,

and Puzyr'

12. G. I. Smelkov, A. A. Aleksandrov, V. A. Pekhotikov, and E. V. Grishin, "Combustion of large aluminum particles in an air flow," Fiz. Goreniya Vzryva, 14, No. 5, 33 37 (1978). 13. V. A. Arkhipov, V. A. Ermakov, and A. A. Razdobreev, "Dispersity of condensed products of combustion of an aluminum drop," Fiz. Goreniya Vzryva,

18, No. 2, 16-19 (1982). 14. M. E. Derevyaga, L. N. Stesik, and l~. A. Fedorin, "Magnesium combustion regimes," Fiz. Goreniya Vzryva, 14, No. 5, 3-10 (1978).

15. A. G. Beloshapko and A. A. Bukaemskii, "Shock adiabat of highly porous aluminum," in: Treatment of Materials by Pulsed Loading (collected papers), Novosibirsk (1990), pp. 19-21. 16. B.C. Lippens and J. J. Steggerda, "Active aluminum oxide," in: B. G. Linsen et al. (eds.), Physical and Chemical Aspects of Adsorbents and Catalysts, Academic Press, London-New York (1970). 17. P. P. Budnikov, I. A. Bulavin, et al., New Ceramics [in Russian], Stroiizdat, Moscow (1969). 18. G. V. Sanmonov, O. P. Kulik, and V. S. Polishchuk, Production and Methods of Analysis of Nitrides [in Russian], Naukova Dumka (1978).