810 THE BELL SYSTEM TECHNICAL JOURNAL, JULY 1952 



EFFECT OF SURFACE TEMPERATURE DURING BOMBARDMENT 



In the preparation of photocells it was found that the surface tem- 

 perature during bombardment had a pronounced effect on the efficiency. 

 In order to study this effect it was necessary to determine the surface 

 temperature of the silicon itself. Since it was impractical to measure the 

 silicon temperature during bombardment, a calibration was made of the 

 surface temperature in terms of the temperature of the graphite heating 

 block. This calibration was carried out by tw^o platinum/platinum 

 rhodiimi thermocouples made of 5-mil wires. The fused thermojunction 

 beads were held in contact with the surfaces by miniature tungsten 

 springs. Temperature measurements ^^dth the thermojunction in contact 

 with the silicon surface were subject to error from the slightest con- 

 tamination at the point of contact. Perhaps the most difficulty was due 

 to a reaction between the platinum and silicon at temperatures above 

 400°C. 



The effect of surface temperature on the photoresponse is shown in 

 Fig. 7. It is apparent that maximum sensitivity results when the target 

 is kept at about 395°C. Perhaps by coincidence this is also the temper- 

 ature at which no Hall Effect is observable in this hyper-pure material. 



Cells prepared at temperatures above the critical value show lower 

 back resistances than those prepared at the critical temperature and 

 conversely those at temperatures below the critical value have higher 

 back resistances but a much reduced photoresponse. 



EFFECT OF TOTAL BOMBARDING CHARGE 



The 4)hotoresponsiveness improves as the total bombarding charge is 

 increased until it has reached about 600 microcoulombs per sq. cm. 

 Further bombardment produces no appreciable improvement. In certain 

 exploratory tests a total charge of about 9000 microcoulombs at 30 kv 

 has been applied. Under these severe conditions the surface may show 

 small areas having a slightly etched appearance. 



No extensive tests have been made to determine the effect of the rate 

 of application of the bombarding charge. The apparatus was designed 

 for use at a rate of about 5 microamperes per sq. cm. It is known how- 

 ever, that between the limits of about 2.5 and 10 microamperes per sq. 

 cm. the effects are subject only to the total charge or the total number 

 of ions which strike the silicon surface. 



EFFECT OF BOMBARDMENT VOLTAGE IN SPECTRAL RESPONSE 



Six spectral curves are shown in Fig. 8 which illustrate the result ob- 

 tained with the intermediate size cells over the bombardment voltage 



