Semiconductor parameter extraction using cathodoluminescence and genetic algorithms

This paper presents a novel approach for the extraction of five semiconductor-related parameters, that is, the absorption coefficient α, diffusion length L, normalized surface recombination velocity S, dead layer thickness Zt and the relative quantum efficiency Q directly from any theoretical/experimental steady state cathodoluminescence (CL) signal. The extraction technique, based on genetic algorithms (GA), has many advantages such as the simultaneous obtainment of near-optimum values for the five parameters (α, L, S, Zt, Q). Simulation results for an n-type GaAs defect free semi-infinite semiconductor sample show that the proposed approach is successful and a set of parameter values with error less than 12% from the nominal values, in 95% of the cases, can be obtained.