Analysis of sol–gel silica–titania films doped with Ag and Er using artificial neural networks

Abstract

Listen

Translate article

Abstract Sol–gel processing is a cheap and versatile method of producing silica on silicon films for planar integrated optics. It leads to high Er incorporation and to easy incorporation of Ag, that can intensify the rare-earth photoluminescence. Different heat treatments and compositions must be tested to optimise the properties of the films grown, and the annealing may lead to redistribution of the elements into different layers, including of H present in the films. We have analysed sol–gel silica films doped with Er and Ag and subject to different annealing procedures with Rutherford backscattering (RBS) and elastic recoil detection (ERDA), leading to a large quantity of complex spectra. We developed an artificial neural network (ANN) able to analyse simultaneously the RBS and ERDA spectra collected from one sample. Non-standard network architecture was necessary due to the complexity of the problem. The optimised ANN is applied to experimental data leading to results that are practically as accurate as those obtained with a conventional data analysis code.