Introduction

XRFMC-PyMCA is an extension for the PyMCA X-Ray Fluorescence Toolkit. It offers the user the possibility to quantify XRF data using iterative Monte-Carlo simulations, as an alternative for the Fundamental Parameter Method (FPM). Unlike the FPM, the Monte-Carlo simulations take into account secondary and higher order excitations in the sample which will lead to a more reliable quantification. The quantification procedure (based on the Levenberg-Marquardt algorithm) is explained in the schematic: the net-line elemental areas obtained by fitting the experimental spectrum with PyMCA, are compared with the elemental areas produced by a Monte-Carlo simulation determined by a number of input parameters. If the sum of the squares of the area deviations is smaller than a certain threshold, then the procedure is stopped. If not, then the sample composition is adjusted and the simulation will be performed again after which the results will be compared with the experimental elemental areas. The iteration will continue until convergence is achieved. This method also has the extra advantage that it produces a spectrum allowing direct comparison with the experimental data (see NIST SRM 1577 example).

Control over the simulations is possible through the existing PyMCA Advanced Fitting interface, by adjusting the different parameters such as beam characteristics, attenuator and matrix composition etc. The Monte-Carlo simulations themselves will be performed by an executable (written in Fortran 77 and C), spawned by PyMCA. This executable will not be provided as a part of the PyMCA package that is distributed through Sourceforge.net, but has to be requested from the author directly. All requests will be handled on a case-by-case basis.

Corresponding with the authors:

Prof. Laszlo Vincze and Tom Schoonjans, X-ray Micro-spectroscopy and Imaging research group (XMI), Department of Analytical Chemistry, Ghent University. Krijgslaan 281, B-9000 Ghent, Belgium. (E-mail: Laszlo.Vincze@ugent.be).