Published online by Cambridge University Press: 01 February 2011
We compare the predictions of several analytical models for conductivity fluctuations in a homogeneous semiconductor containing discrete and distributed traps, using a Monte-Carlo simulation of the relevant multi – trapping (MT) transitions. The simulation directly embodies the statistical features associated with such processes, in a simple ‘model - independent’ approach, free of approximations and assumptions. We compare the results with those of several analytical approaches. In one, the noise spectrum is assumed to reflect separately, the characteristic individual release time constants of the various trapping centers in the material. In another, the trapping time into the ensemble of electron traps is taken to be the dominant time constant, and hence, in a material such as a-Si:H, where the trapping time into tail sates is of order 1ps, this is taken to imply that this component of the conductivity noise spectrum is unobservable in practice. Our own analytical approach, incorporates coupling (albeit weak) between traps, which necessarily communicate via the extended states. Preliminary results of the simulation support our thesis, and verify that the same information is contained in the real part of the modulated photoconductivity (MPC) spectrum. A ‘full Monte’ – Carlo simulation incorporating all gap states and spatial inhomogeneities is now a priority.