Extended-state mobility and its relation to the tail-state distribution ina-Si:H

Abstract

The differing interpretations of drift-mobility data for electrons in a-Si:H have been examined. On the one hand, there are the analyses which involve the ‘‘thermalization approximation’’ (TA). One such analysis, based on the density-of-states profile deduced by Spear from field-effect data, leads to the unexpectedly high value of the extended-state mobility ${\mu }_{\mathrm{ext}=500\mathrm{}}$ ${\mathrm{cm}}^2$ ${\mathrm{V}}^{\mathrm{-}1}$ ${\mathrm{s}}^{\mathrm{-}1}$. Another analysis, which assumes an exponential distribution of tail states, results in the more commonly accepted value ${\mu }_{\mathrm{ext}=13\mathrm{}}$ ${\mathrm{cm}}^2$ ${\mathrm{V}}^{\mathrm{-}1}$ ${\mathrm{s}}^{\mathrm{-}1}$. In this work, we show that both models are inconsistent with experimental data when a more thorough examination is made of the field dependence and the activation energy of the drift mobility ${\mu }_d$ as predicted by those models. On the other hand, we examined Spear’s analysis, which does not involve the TA. We show that a thorough analysis of multiple-trapping transport in a tail-state distribution as proposed by Spear leads to more dispersion than he experimentally observed. Our examination is based on a newly developed computational procedure, which analyzes multiple-trapping transport by discretizing the continuous distribution of localized states. Apart from the relevance with respect to electron transport in a-Si:H, this work shows how a relatively simple but accurate analysis of drift mobility and transient photocurrents can be performed with a much wider applicability than the analyses based on the TA.