Determination of Major Elements in Small-Weight Soil and Sediment Samples by X-ray Fluorescence Spectroscopy with Pressed-Powder Pellets

The analysis of small-weight samples utilizing X-ray fluorescence spectrometry (XRF) poses a pivotal technical challenge in determining the chemical composition of valuable and scarce materials. Furthermore, the application of XRF to verify the homogeneity of reference materials has sparked debates regarding the minimum sample weight. Presently, geological samples, encompassing reference materials, predominantly adhere to a particle size of 74μm (-200 mesh). In the context of XRF analysis employing pressed-powder pellets, the conventional sample weight is approximately 4g. In this paper, 0.1g weight soil or sediment was used for pressed-powder pellet preparation. The diameter of the XRF spectrometry sample box mask was changed to 12mm, and the diameter of the field view light barrier was reduced to 10mm. Based on the previously optimized instrumental measurement conditions, we successfully established a 0.1g sample weight analytical method for the quantification of ten major elements (SiO₂, Al₂O₃, TFe₂O₃, MgO, Cao, Na₂O, K₂O, Mn, Ti, and P) utilizing wavelength dispersive X-ray fluorescence (XRF) spectroscopy, which significantly reducing the necessary sample weight. This methodology employed a diverse range of 32 geochemical reference materials, encompassing various types and content gradients. The detection limit of the 0.1g sample weight analysis method was between 14μg/g and 0.35%, and the precision (RSD, n=12) was less than 3.9%. Through comparative analysis, the results of the reference materials were all within the standard value range, and the absolute value of relative error (|RE|) was between 0 and 15.7%. There was no significant difference with the 4g sample weight analysis results (|RE| ranged from 0.3% to 28.3%). The practical sample results determined by the method of 0.1g sample and the conventional method of 4g sample are consistent. The established XRF method for the 0.1g weight sample is reliable.