558 



THE BELL SYSTEM TECHNICAL JOURNAL, MAY 1956 



Fig. 8 — Photograph of experimental situation described schematically in 

 Fig. 7. 



to the possibility of interactions between the donor and acceptor ions 

 themselves. For example, in (2.1) direct interaction of Li^ and B" above 

 600°C may be possible, especially in view of the mobility of Li^ . Such 

 a reaction was indicated in the work of Reiss, Fuller, and Pietruszkie- 



34 



wicz. 



Fig. 9 is of assistance in understanding the nature of these observa- 

 tions. In it are shown plots of the solubility of lithium in silicon. In this 

 case the situation is similar to that involved in the germanium curves 

 of Fig. 5 because the external phase is composed of silicon and lithium 

 and is probably of the liquidus composition. It is formed by simply 

 alloying lithium to the silicon surface. In Fig. 9, Curve A, illustrates 

 how solubility depends on temperature when the silicon is undoped. 

 Curve B, unlike A, is not an experimental plot, i.e., it is not supposed 

 to represent the locus of the points through which it seems to pass. In- 

 stead it has been calculated from the theory expounded below. The points 

 themselves are experimental and represent solubility measurements on 

 silicon doped with boron to the level 1.9 X 10 cm" 



I 



