Comparison of two models for phonon assisted tunneling field enhanced emission from defects in Ge measured by DLTS

Abstract

Deep Level Transient Spectroscopy (DLTS) was used to measure the field enhanced emission rate from a defect introduced in n-type Ge. The defect was introduced through low energy (±80 eV) inductively coupled plasma (ICP) etching using Ar. The defect, named EP0.31, had an energy level 0.31 eV below the conduction band. Models of Pons and Makram-Ebeid (1979) [2] and Ganichev and Prettl (1997) [3], which describe emission due to phonon assisted tunnelling, were fitted to the observed electric field dependence of the emission rate. The model of Pons and Makram-Ebeid fitted the measured emission rate more accurately the Ganichev and Prettl. However the model of Ganichev and Prettl has only two parameters, while the model of Pons and Makram-Ebeid has four. Both models showed a trasition in the dominant emission mechanism from a weak electron-phonon coupling below 152.5 K to a strong electron –phonon coupling above 155 K. After the application of a x2 goodness of fit test, it was determined that the model of Pons and Makram-Ebeid describes the data well, while that of Ganichev and Prettl does not.