Monte Carlo simulation code for confocal 3D micro‐beam X‐ray fluorescence analysis of stratified materials

Abstract

Stratified materials are of great importance for many branches of modern industry, e.g. electronics or optics and for biomedical applications. Examination of chemical composition of individual layers and determination of their thickness helps to get information on their properties and function. A confocal 3D micro X-ray fluorescence (3D µXRF) spectroscopy is an analytical method giving the possibility to investigate 3D distribution of chemical elements in a sample with spatial resolution in the micrometer regime in a non-destructive way. Thin foils of Ti, Cu and Au, a bulk sample of Cu and a three-layered sandwich sample, made of two thin Fe/Ni alloy foils, separated by polypropylene, were used as test samples. A Monte Carlo (MC) simulation code for the determination of elemental concentrations and thickness of individual layers in stratified materials with the use of confocal 3D µXRF spectroscopy was developed. The X-ray intensity profiles versus the depth below surface, obtained from 3D µXRF experiments, MC simulation and an analytical approach were compared. Correlation coefficients between experimental versus simulated, and experimental versus analytical model X-ray profiles were calculated. The correlation coefficients were comparable for both methods and exceeded 99%. The experimental X-ray intensity profiles were deconvoluted with iterative MC simulation and by using analytical expression. The MC method produced slightly more accurate elemental concentrations and thickness of successive layers as compared to the results of the analytical approach. This MC code is a robust tool for simulation of scanning confocal 3D µXRF experiments on stratified materials and for quantitative interpretation of experimental results.

Confocal X-ray technology based on capillary X-ray optics

“…Since deconvolution is executed by the iterative procedure, a convergence of the sequence of the successive steps could be rather slow in case of very complex Brought to you by | Carleton University OCUL Authenticated Download Date | 6/15/15 10:37 AM structures. Therefore, it is necessary to increase simulation efficiency with more advanced variation reduction techniques (Czyzycki et al 2011). In order to use confocal XRF to determine the element concentration profiles in stratified materials, a direct deconvolution of the measured depth-dependent XRF intensity signal with the established response function of the spectrometer was developed.…”

Atomic spectrometry update—X-ray fluorescence spectrometry

“…In a report strictly devoted to methods for quantitative m-XRF analysis, Czyzycki et al 91 described an inverse Monte Carlo simulation code capable of modelling X-ray characteristic peak intensities and estimating the elemental composition of stratified multicomponent samples by processing the data from confocal m-XRF measurements. The results obtained by the Monte Carlo code and by using an analytical approach developed earlier by Malzer and Kanngiesser (W. Malzer and B. Kanngiesser, Spectrochim.…”

The perspective of new multi‐layer reference materials for confocal 3D micro X‐ray fluorescence spectroscopy

“…A Monte Carlo code for the simulation of a confocal 3D μXRF experiment was proposed recently. In this code, the fate of each individual photon from an incident beam is tracked from the point where it leaves the focussing optics, through sample penetration where it may undergo different interactions, until its ingress into the optics attached to the detection channel, followed by registration in the X‐ray detector.…”