Geometry correction procedure for quantitative X-ray microanalysis of low-Z number matrices

Abstract

Low-Z element matrices (Z<10) are typical for biological materials. H, C, N and O are the main contributors enriched with elements from Na to Ca (biological elements). In biological EPMA, protein-based standards in a "bulk" form are used to assess elemental concentrations; however, the size of an unknown specimen does not always meet the criterion of "bulk specimen" in one of its dimension. Then, specimen geometry affects accurate and precise calculation of elements. Here, Monte Carlo simulation was applied to assess the effects of specimen geometry on the efficiency of X-ray emission from cellulose of increasing width (from 0.5 to 25 μm) containing "biological elements" and exposed to 10 keV electrons. The normalized intensities of emitted X-rays were plotted against specimen width and the double decay exponential function fitted. 150 simulations were performed using randomly defined concentrations of "biological elements" to estimate relative errors for uncorrected and geometrically corrected results. The contents of "biological elements" calculated from simulated intensities revealed ± 30 % of relative error if no correction was applied that decreased to ± 5 % after correction for specimen geometry. It seems that the procedure might be suitable for many biologically derived materials and polymers.