LAUSR

Abstract A quantitative determination of chemical and mineral composition at the nanoscale is nowadays fundamental for the knowledge of properties of innovative composite materials. Scanning electron microscopy (SEM) equipped with energy dispersive X-ray spectrometry (EDS) is one of the most commonly employed spatially-resolved analytical methods, because of its versatility and great potential for nano-analysis. However, because of both complex architecture, texture and reduced grain sizes (micro-to-nano) in many composites, to avoid analytical errors, several effects related to electron and X-ray transport in solids must be considered. In the present work, a Monte Carlo SEM-EDS simulation approach is proposed and applied to selected micro-nanosized architectures usually found in composites. The effects of both micro-nanometric grain sizes (100 nm−20 μm) and basic geometrical shapes (cubic, hemicylindrical) of embedded features in the sample matrix (a metal matrix and a glass fibre-reinforced cementitious composite), together with a realistic SEM-EDS setup, were studied. The results evidenced a high dependence of the simulated X-ray spectra versus particles thickness and shape, beam energy and sample-to-detector configuration, which directly affect a correct analytical characterization. The Monte Carlo simulation allowed to investigate and control the physical phenomena affecting the measurement and eventually to determine the optimal SEM-EDS parameters.