CHEMICAL INTERACTIONS AMONG DEFECTS IN Ge AND Si 557 



formed by the phosphorus layer prevents lithium from reaching the sur- 

 face, diffusion in the sample with the skin should be plane parallel with a 

 straight front (except at the rear where the skin has been lapped off and 

 lithium can leak out) as the p-n junction contour in the lower part of 

 Fig. 7 indicates. 



An acid staining technique " which reveals the junction contours should 

 then develop a picture resembling Fig. 7. The two specimens were cut 

 along their long axes and the stain applied to the newly exposed sur- 

 faces. The result has been photographed and is shown in Fig. 8 where 

 the crystal on the right has the n-skin. The p-regions show up dark and 

 the n, light. The result agrees wdth Fig. 7. 



In another experiment a crystal completely enclosed in a phosphorus 

 skin was immersed in the tin bath described in Section III. It was dis- 

 covered that lithium entered the crystal with no evident difficulty, just 

 as though the skin were absent, but once in, could not be driven out by 

 removal of the external source and continued heating. The implication is 

 clear. The built-in field has a rectifying action permitting the lithium to 

 enter the crystal but not to leave. In this sense it performs the same func- 

 tion for the mobile lithium ions as it does for holes in a p-n junction 

 diode.'' 



V. COMPLEX ION FORMATION 



In the previous text processes involving the interaction of electrons 

 and holes have been considered. In this section attention will be drawn, 



JoCn>o<xxx><^0<><>^6^^^\:SX\\«^ 



Fig. 7 — Distribution of lithium after an extended period of diffusion at a 

 temperature lower than the alloying temperature — showing leakage out of the 

 crystal in the one case (no-skin) and conservation in the other. 



