THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1953 



0.8 



20 24 28 32 

 TIME IN MINUTES 



Fig. 4 — Contact potential cycles for etched sample D. 



then given a polishing etch (CP-4). After the etch the surface was 

 washed in running distilled water of reasonable quality. The surface was 

 then dried with filter paper and kept covered until placed in the bell 

 jar. From here on the procedure was the same as before. In this case 

 cycles 1, 3 and 11 are shown. The surprising result is the similarity be- 

 tween Fig. 2 and Fig. 3. To a first approximation one is tempted to say 

 that the dipole of a Ge surface, in the bell jar atmosphere is independent 

 of past history. Within certain limits this is approximately true. There 

 are differences between Fig. 2 and Fig. 3 but they are small and probably 

 due to differences in surface treatment. Fig. 4 shows the results for 

 p-type slice, D, when this surface is etched as above. This slice and the 

 slice A were placed in the bell jar at the same time. Cycle 1, Fig. 3, was 

 taken on A, cycle 2, Fig. 4, on D and so on. Any differences between 

 the results in these two figures are to be attributed to the differences in 

 samples. They cannot very well be ascribed to the reference electrode 

 and the initial surface treatments were as nearly the same as they 

 could be made with reasonable care. By making runs of this type two 

 samples at a time, different samples and different surface treatments 

 can be intercompared. This method eliminates the shifts in the work 

 function of the reference electrode that sometimes occur from run to 

 run. Such resulte can \)e illustrated by plotting the data for different 

 samples and different surface treatments for cycles 10 or greater where 

 the resulte for successive cycles are the same. This has been done in Fig. 



