380 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1960 



diamond, shows that we are then dealing with about 10,000 million 

 electrons per square centimeter sweeping through the crystal. They 

 are moving with a velocity of 10^ cm./sec. and each one is likely to 

 collide with the atoms of the lattice about a million times before being 

 trapped at one of the impurities. Under the influence of the applied 

 electric field, they will have migrated about 0.1 to 1 mm., before being 

 trapped. 



The passage of the electrons through the crystal can be likened to 

 billiard balls moving across a very unorthodox billiard table which has 

 not six pockets into which the balls can fall but hundreds scattered 

 at random across the baize. If, in addition, we imagine that the table 

 is hinged at one end and raised at the other, then the slope on the table 

 will be analogous to the effect of the electric field. As the pockets are 

 filled we must imagine the slope on the table to be reduced in just the 

 way that trapped charges reduce the effective electric field by building 

 up an internal space charge. 



Only a veiy small proportion of the traps are occupied in the crys- 

 tal, and even if free electrons are continually generated at the rate 

 of 10,000 million per second, it would require the best part of a year 

 to fill all the traps. The energy of the beta particles is such that some 

 of them can penetrate right through the small diamonds which we 

 have been using, but the majority of the electron hole pairs are pro- 

 duced fairly close to the electrode through which the particles enter. 

 It is obsen^ed that the counting rate, or — what is the same thing — the 

 number of free electrons contributing to the voltage pulses, is not 

 maintained at its initial level. If the experiment is done in darkness, 

 then, after 10 or 15 minutes, the counting rate has dropped to about 

 one-half where it remains steady. To explain this, let us look at the 

 factors controlling the counting rate. In a given diamond it will 

 depend on three things : 



1. The number of electrons and holes produced by each incident particle. 



2. The value of the electric field in the diamond. 



3. The length of time the electrons spend in the conduction band before 

 recombining with a positive hole. 



This length of time is called the lifetime. The longer the electron is 

 free to drift through the crystal, the larger is the output pulse ob- 

 tained. Without impurities the lifetime would be long, because re- 

 combinations between electrons and holes will seldom occur by direct 

 collision. A certain energy is required to produce the electron-hole 

 pair, and when they recombine, the energy must be dissipated in the 

 form of irradiation. For the recombination to occur, the electron and 

 hole must be within 2 A of each other. The characteristic time for 

 radiation is about 10"^ sec, which is longer than the time for which 

 the pair are close enough to interact. It is much more probable that 



