554 



THE BELL SYSTEM TECHNICAL JOURNAL, MAY 1956 



16 



lem in preparing samples without precipitate. Thus consider the A~ = 

 curve. Suppose the specimen is saturated at 200°C. Then 



D^ after satura- 

 However, as the sample is cooled it will tend, 

 at first, to become supersaturated. For example it will achieve its maxi 

 mum supersaturation at about 140°C. where the minimum of the 10 



~^ - Thereafter it will return to its undersaturated state. 



could be supported, 

 lithium atoms. 



Some of these may have precipitated as the cooling process passed 

 through the minimum, so that sufficient time must be provided for the 

 process of re-solution. 



If the original saturation had taken place at 250°C, the concentration 



inl6 -3 



1 cm 



according to Fig. 5, if A" for the specimen is 10 cm~ 



tion will be 7 X 10 cm" 



cm curve appears. 



In fact at 25°C a concentration of 9.3 X 10^^ cm"^ 



whereas the solution contains no more than 7 X 10 cm" 



10 



19 



10 



n 

 I 



u 



2 10' 



18 



17 



Z 



o 



< 



cr 



(- 

 z 



LU 



u 



z 

 o 

 o 



10 



16 



10 



15 



10 



H 



10 



,13 



100 



200 300 400 500 

 TEMPERATURE IN ° C 



600 700 



Fig. 5 — Solubility of lithium in germanium as a function of temperature for 

 various gallium dopings. The external phase is an alloy of lithium and germanium. 

 The broken line is the locus of the points (circles) calculated from equation (4.1) 

 for zero gallium concentration. The values of A' and Z)+, used in applying (4.1),^ 

 correspond to the points shown by X iu the illustration. See Table II. 



