Theoretical high-stress optical birefringence and piezoresistance in heavily doped germanium. Arguments against band tailing

Abstract

The gradual and delayed saturation in the high-stress optical birefringence and piezoresistance experiments on heavily doped germanium has been considered as clear evidence for band tailing. The origin of this tailing of the conduction band has been assumed to be spatial fluctuations in the donor concentration. We present here a calculation that reproduces the experimental behavior under the assumption of a random distribution of impurities, without any band tailing, in the unstressed samples. The origin of the effect is, in this treatment, electron-electron and electron—donor-ion interactions. The interactions change the relation between the applied stress and the stress-induced electron transfer between the conduction-band valleys. This change is such that the experimental anomalous behavior is reproduced.