1036 THE BELL SYSTEM TECHNICAL JOURNAL, SEPTEMBER 1956 



(F — 111 X). Fig. 7 shows the variation of surface rerombiiiation velocity 

 with F — hi X, using the experimental photoconductivity data and values 

 of y read from Fig. 5. The values of s have been divided by (X + X~'), 

 as indicated, since s/(X + X" ) is expected to be the same, at a given 

 value of ()' — In X), for all samples, so long as the distribution of fast 

 states is the same. The agreement shown in Fig. 7 is probably closer 

 than would be expected in the light of the experimental accuracy. 



Fig. 8 shows the observed dependence of dY/d8 on (F — In X) for both 

 samples, using the data of Tables III and IV, and using the photocon- 

 ducti^'ity to determine, from Fig. 7, the value of F at each point. On 

 the figiu'e the expected limiting values ( — X and X~ ) are shown for both 

 samples. Of the four asymptotes, the higher limit of (dY/d8) for the 

 At-type sample is satisfactorily reached for large negative values of Y; 

 the experimental values for the p-type sample appear to be approaching 

 the expected limit for large positive values of F, while the information 

 regarding the approach to the two lower limits is too fragmentary to do 

 more than show that the order of magnitude is as expected. Now taking 

 the data shown in Fig. 8, making use of (1) and the calculations given 

 in Tables III and IV, one calculates (82^/88) Y/(d2,/dY)s . The values 

 so found are plotted against Y in Fig. 9. Fig. 6, 7 and 9, showing the ob- 

 served variation of {dXs/dY)s , s and {d2s/d8)Y/id2s/dY)s with F, 

 furnish a complete description of the properties of the fast states at the 



200 



CO 



100 

 80 



60 



k 50 



+ 



-^40 

 30 



20 



10 



o 

 o 



o 

 o 



._5 1 



— o 



: • • 



o 



o 



^2 ? 



.o • 



o • 



o 



-2 



-1 1 



Y-mx 



Fig. 7 — Surface recombination velocity versus surface potential. The curves 

 show s/(X + X~') plotted against F — In X. Dots: p-type; circles: n-type. 



