Reverse engineering from spectrophotometric measurements: performances and efficiency of different optimization algorithms

A large number of parameters is often required to describe optical dispersion laws, and it is only through the use of an appropriate global optimization procedure that an accurate thin-film index determination can be achieved. In this paper, we propose to investigate the respective performances of three different optimization algorithms, namely Simulated Annealing, Genetic Algorithm and Clustering Global Optimization and compare results with a commercial software dedicated to thin-film index determination. This study is restricted to the single-layer thin-film index determination of transparent and absorbing materials. It includes the theoretical study of simulated reflection and transmission spectra, and the experimental characterization of Ta₂O₅ and Si layers.