Chemical Decomposition of Crystals of Explosive Materials 



345 



maintained at all subsequent stages so that the 

 quantum efficiency in the scanning instrument may 

 approach unity for all modes of operation. 



For reflexion operation, therefore, the total num- 

 ber of electrons which the specimen must receive 

 will be greater in the conventional instrument by 

 roughly a factor of I0-' for recording and 10^-10" 

 for direct observation at ecjual integration times of 

 observation. In the conventional instrument the 

 integration time is usually that of the eye (0.2 

 second) whereas in the scanning instrument a long 

 persistence screen is used. This gives an additional 

 factor of 10 or so in favour of the scanning instru- 

 ment which means that the total current passing 

 to the specimen during direct viewing in reflexion 

 may well be higher in the conventional instrument 

 by a factor of 10'. 



So far, only the total dosage of specimen and the 

 current passing to the specimen have been consid- 

 ered; the current density incident on the specimen is 

 also of importance. The maximum current density 

 which can be brought to bear on the specimen is 

 limited by the electron gun and will be the same in 

 both instruments. Other things being equal, there- 

 fore, the rate of transfer of information in the 

 scanning instrument in which the elements are 

 imaged sequentially, will be much lower. However, 

 for reflexion operation, the improved quantum effi- 

 ciency and longer integration time of the scanning 

 instrument offsets the reduced time efficiency, and 

 the information content in the directly observed 

 pictures will therefore be much the same. 



An additional factor which reduces the heating 

 effect is that in the scanning instrument only the 

 area actually observed is irradiated by the electron 

 beam whereas in the conventional instrument the 

 irradiated area will be appreciably larger than the 

 field of view, even with a specimen screening aper- 

 ture. Also, in the scanning instrument, the primary 

 effects of electron bombardment are reduced by the 

 use of a comparatively low accelerating voltage 

 (-15 kV). 



Results. — Silver azide melts at 230 C and detonates 

 at 350 C but at temperatures between 120 C and 

 230°C it decomposes slowly, giving off nitrogen and 

 leaving metallic silver. It is soluble in ammonium 

 hydroxide and specimens were prepared by recrys- 

 tallisation from this solution on to silver discs 

 which screw into the hot stage. 



Fig. 2 a shows a needle shaped crystal of silver 

 azide which has decomposed from one end, because 

 of the better contact with the hot plate at that end. 



Decomposition is not considered to be complete at 

 this stage, but it can be seen that the crystal is break- 

 ing up into pebbles of approximately 0.3 f^i in size. 

 Complete decomposition would reduce this size to 

 about 0.2 // which is the size most common in other 

 larger crystals heated in this way. 



It is not clear yet whether these small pebbles are 

 formed by aggregation of silver by a diffusion process 

 into preferred sizes, as found by Sawkill, or whether 

 the initial size is determined by a fine structure of 

 dislocations and decomposition to pure silver then 

 follows. It is evident, however, from later work that 

 free surfaces assist in the mechanism of decomposi- 

 tion. This is shown by the fact that the top and side 

 faces of a larger crystal decompose more rapidly 

 than does the bulk, when heated from the underside. 



A break-up of larger dimensions I -^ 2// has been 

 observed which is attributed to a polymorphic phase 

 change known to take place at 180 C (Sawkill and 

 Duke, independently, private communications). The 

 pieces produced in this way are angular in shape 

 and very different from those in Fig. 2 a. When 

 this has occurred it is difficult to decompose the 

 crystal further because of the poor thermal conduc- 

 tivity through the porous mass, and high radiation 

 losses from the large surface area. 



The second material, lead styphnate monohydrate, 

 was examined to determine the effects of dehydration 

 which takes place above 120 C. Chemical decom- 

 position at this temperature is very slow and the 

 major effects can be attributed to the loss of water 

 which is tightly bound in the material. Fig. 2 b 

 shows one stage of the dehydration process, the 

 crystallographic nature of break-up is very apparent. 



Break-up in this way by polymorphic change or 

 dehydration, which can be considered a physical 

 rather than a chemical change, has some bearing 

 on the study of reaction kinetics by pressure-time 

 curves. Since the rate of reaction normally depends 

 on surface area any mechanism not initially related 

 to the chemical change but which causes an increase 

 in surface area will assist in the later stages of the 

 reaction, and must be considered in the final analysis. 



References 



1. BowDEN, F. P. and Singh, K.. Proc. Roy. Soc. A 111, 22 



(1954). 



2. McMuLi.AN, D., Proc. Inst. Elec. Engr.i. 100 (II). 245 



(1953). 



3. Sawkill, J., Proc. Roy. Soc. A 229, 135 (1955). 



4. Smith, K. C. A. and Oatlev, C. W., Biii. J. Appl. Phys. 



6, 391-399 (1955). 



