804 THE BELL SYSTEM TECHNICAL JOUKNAL, JULY 1956 } 



eludes any premature forming which could occur. This curve shows that 

 a low reverse emitter current (high back impedance) is associated with 

 easy forming and that a high reverse emitter current is associated with 

 hard forming. Fig. 16(b) represents data on the same group of units 

 with /c(6, — 5)//c(0, —20) plotted versus the reverse emitter current on 

 a log abscissa. It is significant to note that the figure of merit is consist- 

 ently high for units with low reverse emitter current and low for units 

 with high reverse emitter currents. It was possible to achieve this wide 

 range in reverse currents on the same material by adjusting the current 

 density in the manner summarized by Table V. In each case a high cur- 

 rent density results in the low reverse currents. 



Some other oxidizing agents may be used interchangeably with the 

 materials just discussed. A dilute nitric acid solution produces a surface 

 on which excellent diode properties are observed and good forming re- 

 sults on these surfaces. It has also been found that a treatment in potas- 

 sium cyanide results in a surface which appears to be well oxidized. 

 There are, however, some indications that certain chemical treatments 

 tend, more than others, to passivate the germanium surface to any sub- 

 sequent treatment. 



Although it has been shown that variations in the surface oxide layer 

 markedly affect the transistor made on that particular surface, varia- 

 tions in forming yield such as illustrated by the manipulator line in Fig. 

 13 are still unaccounted for. The etching procedure in the fabrication of 

 the point contact transistor has always been one of the most carefully 

 controlled steps. It therefore becomes necessary to examine the process 

 for some subtle interaction between the germanium surface and the 

 ambient to which the surface is subjected during processing. 



4.2.3 Controlled Ambient Experiments 



The experiment summarized by Fig. 17 represents a "dry box" ex- 

 periment designed to investigate the effect of ambient on the forming 

 yield. Ten germanium wafers were mounted on hermetic seal headers, 

 they were electro-etched, and then five treated for one minute in HF. 

 The wafers were rinsed in deionized water, dried for three minutes in a 

 stream of nitrogen, and placed in a nitrogen dry box where the relative 

 humidity was maintained at less than 1 per cent. One micromanipulator 

 transistor was formed on each wafer immediately and then at subsequent 

 intervals of one day, always in widely different locations on the wafer. 

 These manipulations Mere carried out inside the dry box using rubber 

 gloves so that at no time was the RH greater than 1 per cent. After two 

 days the box was opened to room air and the experiment continued. 



