Trap assisted carrier generation and recombination in semiconductors can be accurately described by Shockley-Read-Hall (S-R-H) statistics [1] if transitions between deep levels can be neglected. Originally mathematical expressions for this model were derived by considering the transition rates of a small non-equilibrium carrier concentration to and from a single-level deep trap. However, this model was generalized for an arbitrary distribution of non-interacting traps in the later publications [2, 3]. The main assumptions of the S-R-H model are valid in the majority of practical problems, but this may not be the case when dealing with low quality or heavily damaged semiconductor materials. It is reasonable to suppose that the recombination process may become similar to a random walk and therefore may be treated by ad hoc methods [4] when trap densities appear to be sufficiently high at low concentration of carriers
In this work a new charge relaxation model has been proposed for description of relaxation processes, when density of carrier traps of various species significantly exceeds equilibrium density of dopants. Random-walk approach has been involved to simulate the intricate redistribution of carriers among trapping centres during relaxation of small density of excess carriers. Results of the numerical simulations, employed to describe the processes of random and repeated events of trapping and release within excess carrier system, have been verified using comparative approach to equivalent low trap/low carrier density regime which can be analytically described through solutions of the master equation system. This comparison showed good agreement between numerical and analytical solutions for equivalent relaxation regimes and illustrates the compatibility of these methods. Furthermore, more intricate numerical experiments were employed to study the behaviour of a model system in the non-linear regime.