Particle swarm optimization of Rutherford backscattering spectra

This paper presents the optimization of Rutherford backscattering spectrometry analysis for thin films using the Particle Swarm Optimization (PSO) technique. Initial simulations performed on silicon and gold samples demonstrate the baseline effectiveness of the proposed method. Subsequently, a detailed experimental study involving an iron-implanted silicon sample shows that the PSO algorithm successfully optimizes the simulation parameters to achieve close agreement with the experimental conditions. The PSO algorithm iteratively minimizes the error between the experimental and theoretical spectra. Its robustness is validated through comparisons with both experimental data and results obtained using the Levenberg–Marquardt Algorithm (LMA). The comparative analysis focuses on layer thickness, elemental composition, and detector resolution, revealing that the spectra optimized by PSO exhibit better agreement with the experimental spectra than those optimized by LMA.