Application of the genetic algorithm to the optimized design of semimagnetic semiconductor-based spin-filters

Abstract

We have further improved our procedure for the optimization of multilayer semiconductor nanostructures, based upon diluted magnetic semiconductors, developed previously with the goal of maximizing their spin-filtering properties. The new optimization scheme relies on the application of a modern class of evolutionary algorithms for global optimization, specifically the genetic algorithm. Its fitness function is set to select the best possible spin-polarization properties within the chosen range of bias voltages and with a fixed value of external magnetic field. Numerical calculations are presented for the ZnSe/Zn1−xMnxSe based semiconductor system, and the obtained results predict an enhanced spin-diode performance over the existing designs.