Development of a peak deconvoluation software for X-ray fluorescence spectra

Abstract

X-ray fluorescence (XRF) analysis is a well established non-destructive analytical method for determining the chemical composition of a sample. After recording a spectrum with an XRF spectrometer it has to be processed with deconvolution software to retrieve the net intensities of the peak representing the respective element.In this master thesis I present JPeakFit, a new Java-based deconvolution tool that comes with a graphical user interface (GUI). JPeakFit was developed with ease of use and functionality in mind. A comparison of JPeakFit with established decon- volution software packages AXIL and PyMCA illustrates the respective strengths of each tool. JPeakFit combines functionality from AXIL not available in PyMCA with a GUI of comparable ease of use as PyMCA. The state-of-the-art design pattern Model-View-Presenter was used to achieve a well structured program.This thesis explains every part of the developed software, both how it can be used as well as how it is structured programmatically. An example fit is given, where every step to fit a spectrum is explained. Finally the results of five different spectra are evaluated and compared with AXIL and PyMCA, whereby good agreement was found.