X-ray microanalysis of porous materials using Monte Carlo simulations

Quantitative X-ray microanalysis models, such as ZAF or φ(ρz) methods, are normally based on solid, flat-polished specimens. This limits their use in various domains where porous materials are studied, such as powder metallurgy, catalysts, foams, etc. Previous experimental studies have shown that an increase in porosity leads to a deficit in X-ray emission for various materials, such as graphite, Cr(2) O(3) , CuO, ZnS (Ichinokawa et al., '69), Al(2) O(3) , and Ag (Lakis et al., '92). However, the mechanisms responsible for this decrease are unclear. The porosity by itself does not explain the loss in intensity, other mechanisms have therefore been proposed, such as extra energy loss by the diffusion of electrons by surface plasmons generated at the pores-solid interfaces, surface roughness, extra charging at the pores-solid interface, or carbon diffusion in the pores. However, the exact mechanism is still unclear. In order to better understand the effects of porosity on quantitative microanalysis, a new approach using Monte Carlo simulations was developed by Gauvin (2005) using a constant pore size. In this new study, the X-ray emissions model was modified to include a random log normal distribution of pores size in the simulated materials. This article presents, after a literature review of the previous works performed about X-ray microanalysis of porous materials, some of the results obtained with Gauvin's modified model. They are then compared with experimental results.