|
内容記述 |
When X-rays collide with atoms in a material, the inner-shell electrons are excited, and holes remain. Along with the electron transition from the outer shell to the hole, X-rays with energy equal to the energy difference between the initial and final electronic states are emitted. The emit-ted X-rays are known as X-ray fluorescence (XRF). XRF is a characteristic X-ray in which each element has a unique energy, rendering it useful for elemental analyses.We performed a quantitative analysis of ura-nium on a rubber plate using XRF analysis [1]. In the XRF spectrum, a small U Lα peak (13.6 keV) was observed between the strong Br Kβ (13.3 keV) and Sr Kα peaks (14.2 keV) derived from components of rubber plate. Each peak was able to be separated by multiple Gaussian fittings for quantitative analysis. However, when a large amount of strontium was present, the tail at the lower energy side of the Sr Kα peak over-lapped with the U Lα peak, and the accuracy of peak fitting using Gaussian functions was af-fected by this overlap. Such non-major peaks caused by peaks other than XRF are called satel-lite peaks or structures. This tailing of the Sr Kα peak is due to the incomplete charge collection in the silicon drift detector commonly used for XRF spectroscopy [2] and the radiative Auger effect [3]. The radiative Auger effect is caused by a potential shift within the atom due to the emission of inner-shell electrons, resulting in XRF emission at a slightly lower energy. As strontium is common in rocks and soils, solving this problem is crucial for the XRF analysis of uranium in environmental samples.In XRF analysis, peak fitting is typically per-formed using Gaussian functions to model the main peaks. In this presentation, we present a method for the quantitative analysis of uranium by fitting the U Lα peak obtained from samples containing both uranium and considerable stron-tium by approximating the tail of the Sr Kα peak with another Gaussian function [4]. With con-ventional fitting using only Gaussian functions for each main peak, the intensity of the U Lα peak was overestimated because of the Sr Kα peak tail. However, the proposed method allows for the stable quantification of uranium regard-less of the strontium concentration.This study was conducted under the auspices of the Secretariat of the Nuclear Regulation Au-thority, Japan.References[1]Yoshii H et al 2022 Spectrochim. Acta B 189 106368[2]Calzolai G et al 2018 Nucl. Instrum. Methods Phys. Res, Sect. B 417 51[3]Campbell J.L. et al 2021 Nucl. Instrum. Methods Phys. Res, Sect. B 499 77[4]Watanabe Y et al 1986 Nucl. Instrum. Methods Phys. Res, Sect. B 17 81 |
