SURFACE PROPERTIES OF GERMANIUM 



35 



e.v. measured from the midband energy are reasonable and not incon- 

 sistent with the original assumptions. 



As mentioned in the experimental section, some data on (Ac.p.)l for 

 samples C and E have been obtained. While no complete analysis of 

 these results has been made, one can see that they are of the right 

 order of magnitude. As the specific resistance of the sample decreases, 

 the body life time r decreases. This empirical result is to be expected.^ 

 Consequently 8p for the same light intensity decreases and one would 

 expect (Ac.p.)l to decrease as it does. For sample E of course one could 

 not neglect the charge in the space layer so that the theory would be 

 more involved. 



The comparison between the contact potentials of samples A, C, D 

 and E shown in Fig. 6 can be understood in part at least. Consider first 

 the over-all result that at the same time in the cycle the c.p. for a sand- 

 blasted surface is less than for an etched surface, i.e., the work function 

 for the sand-blasted surface is greater. It is known that the surface re- 

 combination increases enormously when the surface is sand-blasted. This 

 means that either the surface trap density has increased or that the 

 distribution has changed or both in such a way as to increase surface 



-0.08 -0.06 -0.04 -0.02 



Vq in volts 

 Fig. 19 — Dependence of carrier densities in surface traps, n«o and Fm , for 



Vb = 0, and cross section of trapping St on Fo . 



