Optimization of semiconductor quantum devices by evolutionary search

A novel simulation strategy is proposed to search for semiconductor quantum devices which are optimized with respect to required performances. Based on evolutionary programming, a tecnique implementing the paradigm of genetic algorithms to more complex data structures than strings of bits, the proposed algorithm is able to deal with quantum devices with preset non-trivial constraints (transition energies, geometrical requirements, etc.). Therefore, our approach allows for automatic design, thus avoiding costly by-hand optimizations. We demonstrate the advantages of the proposed algorithm by a relevant and non-trivial application, the optimization of a second-harmonic-generation device working under resonance conditions.