30 



THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1953 



z 2 



>-^ 



10-3 



8 



6 



5 



4 

 3 



10" 



-0.35 -aaO -0.25 -0.20 -0.15 -0.10 -0.05 0.05 0.10 0.15 

 C.R-(C.R)n IN VOLTS 



Fig. 16 — Same as Fig. 15 for sample A, n-type. 



following was done. It was assumed that the surface was made up of 

 two parts each having the theoretical dependence of 8Vl on c.p. but 

 with the contact potential where the light effect goes through zero 

 differing by 0.05 volts. The values of 5Fl at a given contact potential 

 were averaged and plotted against c.p. The difference between this 

 curve and the original one was almost entirely a simple shift in (c.p.)o . 

 It is therefore unlikely that non-uniformities in c.p., over the surface, of 

 the order of 0.1 volts or less would change the relation between 8Vl 

 and c.p. sufficiently to be detectable. 



One can use equations (8), (10) and (21) to calculate values of 

 (Vb — Vq) for each experimental value of 8Vl and then plot these 

 values against the corresponding values of contact potential. These 

 results are shown in Fig. 18 for n-type sample A and p-type sample D. 

 (Vb — Vo) changes with c.p. as one would expect and in an approxi- 



