It was shown that the predictions of the Stanford University CdHgTe process simulator were in only approximate agreement with published Hg self-diffusivity data at, or near to, Hg saturation. The cation vacancy component which was involved in Hg self-diffusion, according to the simulator, was shown to be too large by a factor of least 50. At all temperatures between 350 and 500C, there was shown to be a significant contribution to Hg self-diffusivity which was independent of the Hg pressure and was not accounted for by the simulator. The correct point defect concentrations were considered with regard to Hg self-diffusion data and simulator-predicted defect concentrations.

D.Shaw: Semiconductor Science and Technology, 1997, 12[6], 761-6