Peak Deconvolution Using Intra-Element Ratios Refined by Fundamental Parameter Method

<p>Peak deconvolution is necessary in elemental analyses by using X-ray spectrometry, in order to correct errors due to the overlapping of the emission peaks. In this report, we try to increase the accuracy to determine the compositions of samples by employing fundamental parameter (FP) method in the deconvolution process. The basic point is to determine default intra-element ratios by FP calculation. Since the ratios are dependent on composition of the sample as well as X-ray parameters such as tube voltage, anode material and the geometry of the spectrometer, we recalculate the ratios based on the composition that are calculated using default ratios. The peak intensities are refined when recalculated ratios are used. Calculated composition based on the refined intensities is also refined. Iterative calculations of peak deconvolutions and quantifications are also possible. We applied the refining method for various typical EDXRF applications that clearly show the effect and/or accuracy of ratios has strong effect in calculated peak intensities. In a spectrum of Cu-Zn alloy, Zn-Kα/Kβ ratio is strongly dependent on the composition since Zn-Kα is not strongly absorbed by Cu but Zn-Kα that has slightly higher energy than Cu-K absorption edge. A similar example is Ag-Pd alloy. Pd-K absorption edge is between Ag-Kα and Ag-Kβ Ratios of L-lines are more dependent on the composition in many cases since absorption of irradiating X-rays and secondary excitation also affect the ratios. Examples of a gold alloys are shown. Trace Pb, Bi, As analysis in Sn based alloy is a good example of the effect of the refinement. If we use the ratios that are experimentally determined on single element metal reference, peak deconvolution results zero intensity of As-Kα On the other hand, if we refine the ratios using FP method, better fitting is achieved and As-K line intensity is calculated properly.</p>