Genetic algorithm optimization of X-ray multilayer coatings

摘要

A simple genetic algorithm for global optimisation of the reflectivity of multilayer coatings in the extreme ultra-violet and X-ray wavelength ranges has been implemented as a software tool. The genetic algorithm identifies the best-performing multilayer among a population of solutions that evolves while random mutations are applied to the thickness of the layers. The tool is designed for maximising the reflectivity either over a wavelength range at fixed incident angle or over a range of incident directions at fixed energy. The algorithm has been preliminarily tested on two specific applications: a Pt/C multilayer for hard X-rays applications in astrophysics and cosmology and a Mo/Si coating prominent to next generation lithography at 13.5 nm. The results of the analyses are compared to the performances achievable with periodic multilayers and traditional supermirrors.