CHEMICAL INTERACTIONS AMONG DEFECTS IN Ge AND Si 609 



10^ 



10= 



10' 



10- 



102 



10 



1 



0.0030 



0.0035 0.0040 0.0046 0.0050 



l/TEMPERATURE IN DEGREES KELVIN 



0.0055 



Fig. 28 — Plots of logarithm of relaxation time versus reciprocal temperature 

 showing agreement between theory and experiment. 



of section X of 1.66 X 10 seconds. The result is in good agreement with 

 iiieoiy. 



Studies of the kind ilkistrated in Fig. 27 have been carried out in 

 samples doped to various levels and also at various temperatures. 

 Boron and indium have been used as doping agents, as well as galHum. 

 Relaxation times have been measured over the range extending from 

 about a second to hundreds of thousands of seconds. In each case straight 

 line plots were obtained and the agreement between calculated and 

 measured r's has been as good as in the example illustrated by Fig. 28. 

 Relaxation connected with dissociation has also been measured with 

 equally satisfactory results. 



Some of these data are shown in Fig. 29 where log r is plotted as a 

 function of reciprocal temperature for gallium and boron at two different 

 values of doping. The drawn curves are theoretical obtained from Fig. 20 

 while the points shown are experimental. It is seen that agreement is 

 nearly perfect. The relaxation time, true to the demands of theory, does 

 not seem to depend on the kind of acceptor used for doping, i.e., it is 

 independent of a, the distance of closest approach. 



The data in Fig. 28 actually can be used to measure the diffusivity of 



