POINT-CONTACT TRANSISTOR SURFACE EFFECTS 793 



pected contamination of the points used. For this reason, particular 

 attention was given to chemical processing of the beryllium copper points 

 used in the preceding experiments. These points were chemically cleaned 

 to remove oxides and unwanted contaminants, and carefully washed 

 before use. Several lots were processed at different times, and all experi- 

 ments repeated on the different lots, with no contradictory results. 



It is particularly important that the point be free from donor elements, 

 since it has been observed that that phosphor bronze points or "poi- 

 soned" beryllium copper points washed with a lithium chloride solution 

 often exhibit on superoxol etched surfaces a kind of "forming" after the 

 application of reverse bias. The symptoms of this are a sudden increase 

 in I CO which take place as the reverse bias is increased above 15-20 volts. 

 The alpha emitter current sweep shows evidence of excessive noise in 

 such a case, and it is not until the collector is given a conventional form- 

 ing pulse that this excessive noise is ehminated, and the unit becomes 

 stable in operation. 



A donorless point can be reasonably identified by the fact that elec- 

 trical pulsing, heavy or light, will not increase the initially low average 

 alpha on a superoxol-etched surface to values much above 1 .0, although 

 J CO may be increased or decreased depending on the type of condenser 

 discharge used. The beryllium copper points used were tested on super- 

 oxol-etched surface to make sure they showed no tendency to form 

 electrically. 



If high values of alpha can be found when these points are used as un- 

 formed collectors on the surfaces treated in HF or solder flux, the ques- 

 tion arises whether such values may be attributable to presence of a 

 donor element left on the surface in some mysterious way by the chemi- 

 cal treatment. If such is the case, the donor might, at high enough re- 

 \ erse bias, be responsible for an increased alpha in a manner similar to 

 that observed in connection with the forming in under bias of phosphor 

 bronze collectors on etched surfaces. Two precautions were taken in this 

 connection. No reverse bias greater than 10 volts was ever applied in- 

 tentionally to these collectors during experiments (with exception of the 

 iunit in Figure 11), and secondly, forming characteristics of both phos- 

 jphor bronze points and the beryllium copper points on this type of sur- 

 face were investigated. 



It was found that on a superoxol surface treated with HF or the solder 

 iflux, a phosphor bronze point would form to a high average a, but this 

 invariably required more forming pulses than on a superoxol etched 

 [surface. "One-shot" forming is common for a superoxol etched surface, 

 'whereas after the HF or solder-flux treatment, forming to high average 



