Abstract
An analysis is made of the slow and fast components found in the 203Hg and 109Cd self-diffusion profiles in Hg0.8Cd0.2Te. It is shown that arguments attributing both components to volume diffusion processes are incorrect and disagree with experimental evidence. Satisfactory fits to experimental profiles are obtained using the dislocation and grain boundary analyses of LeClaire and Rabinovitch (1981, 1982) and LeClaire in which the slow component represents volume self-diffusion and the fast component represents dislocation or grain boundary diffusion. Discussion of the evidence leads to the conclusions that (sub) grain boundaries rather than dislocations are probably responsible for the fast diffusion tails.

This publication has 17 references indexed in Scilit: