• Spectroscopy and Spectral Analysis
  • Vol. 41, Issue 9, 2807 (2021)
Ming-bo LIU1、*, Xue-liang LIAO2、2;, Da-wei CHENG1、1; 2;, Zi-yue NI1、1; 2;, and Hai-zhou WANG1、1; 2; *;
Author Affiliations
  • 11. Central Iron and Steel Research Institute, Beijing 100081, China
  • 22. NCS Testing Technology Co., Ltd., Beijing 100081, China
  • show less
    DOI: 10.3964/j.issn.1000-0593(2021)09-2807-05 Cite this Article
    Ming-bo LIU, Xue-liang LIAO, Da-wei CHENG, Zi-yue NI, Hai-zhou WANG. An EDXRF Quantitative Algorithm Based on Fundamental Parameters and Spectrum Unfolding[J]. Spectroscopy and Spectral Analysis, 2021, 41(9): 2807 Copy Citation Text show less

    Abstract

    In the analysis of spectrum for energy dispersive X-ray fluorescence (EDXRF), if the contentof elements is not high in the sample, the peak will keep its original shape as the pure element, peeling the spectrum of pure elements is a good way. Meanwhile, the linear relationship between peak intensity and element content is reliable, and the quantitative analysis is accurate. But in constant analysis, there usually has obvious absorption and enhancement effect between elements, which lead to a different spectrum for the element in the mixed sample compared with pure element spectrum, hence, usingthe fixed shape of pure element spectrum peeling methods will have greater deviation.At the same time, the linear relationship between peak intensity and element content will be disturbed by the absorption enhancement effect.The superposition of the two factors leads to the inaccuracy of element quantification. Therefore, in constant analysis, the simple spectral peeling method is not suitable.This article introduces a full spectrum fitting quantitative algorithm based on the Fundamental Parameter (FP) method, which can fit the spectrum accuratelyandrealize the quantitative calculation at the same time. The process is as follows: firstly, the measured intensity for each line was obtained base on pure element peeling method,and which was used to estimate the content for elements in the sample, and then the theoretical intensity of each spectral line was calculated in the sample using FP method. The element content was adjusted according to the deviation of the measured value, and the two processes of “FP calculating content Adjusting” were iterated until the intensity of the calculated spectrum has no different from the measured spectrum.The obtained sample composition was used to calculate the shape of spectrum intensity for each element. In the meantime, the pure element spectrum was modified. And the steps of “peeling spectrum-estimating content-iterating FP calculation” was repeated. After that, the final element content was considered the test result. This method was used to analyze La/Ce/Pr/Nd spectra in mixed solution samples for about 250 g·L-1 based on EDXRF. The residual error between the calculated spectra and the measured spectra was reduced to 474.5, which was less than the 1 415.0[1] for using the pure element spectral stripping method alone. The method is applied to the determination of several rare earth mixed solutions with the partition content ranging from 0% to 90%, and the deviation of each element in each sample is less than 1%. For a successive test, each element's relative standard deviation (RSD) was less than 1%. The actual measurement shows that the accuracy and stability of the method are good, which can well meet the needs of the rare earth hydrometallurgy industry.
    Ming-bo LIU, Xue-liang LIAO, Da-wei CHENG, Zi-yue NI, Hai-zhou WANG. An EDXRF Quantitative Algorithm Based on Fundamental Parameters and Spectrum Unfolding[J]. Spectroscopy and Spectral Analysis, 2021, 41(9): 2807
    Download Citation