Are interface states consistent with Schottky barrier measurements

Abstract

Many Schottky barrier heights have less dependence upon metal work function than the simple work function matching model of this interface proposed by Schottky 40 years ago. The conventional explanation for this result involves the assumption that interface states of high density change their occupancy in response to work function differences and thereby reduce the variation of Schottky barrier heights. Such states could also change their occupancy under bias variations, and transport measurements have been interpreted in this fashion. We show that these latter assumptions also imply that the Schottky barrier height should itself be strongly altered by bias - a conclusion inconsistent with the results of capacitance and current transport experiments.