Computation of the energy distribution of localised states from transient photo-decay measurements on disordered semiconductors: a comparison of methods

A computer simulation technique is employed to calculate the transient photo-decay characteristics for a disordered semiconductor featuring an exponential tail of localized states plus a narrow Gaussian feature of adjustable height The resulting data are subjected to analysis via the "pre-transit" (1/I(t) t). "post-transit" (I(t) t) and "Fourier Transform" procedures It is shown that all three options can detect the presence of the Gaussian component However, even when the peak height of this becomes comparable to or less than that of the local exponential background, the pre-transit procedure consistently miscalculates its energy In contrast, the post-transit and Fourier transform procedures correctly identify this energy. and also provide improved resolution of the energy distribution and other properties of the localized states We discuss further the theory and application of the Fourier Transform procedures to the analysis of I(t) data, focusing in particular on the effect of the inevitable short and long time truncation of I(t) data in experimental measurements We show how some simple measures can be taken to extend the energy range and accuracy of the method in the determination of density-of-states distributions