Matrix effect correction method based on the main spectral parameters for rock samples in an in situ energy dispersive X-ray fluorescence analysis

Due to the differences in the matrix effects between different types of rocks, the target element contents in different rocks cannot be accurately determined by in situ X-ray fluorescence (XRF) analyses with the same parameters of the quantification procedure. We investigated the matrix effect correction methods of an in situ energy dispersive X-ray fluorescence (EDXRF) analysis for different matrix rock samples using Monte Carlo simulations (17 types of rock samples) and experimental verification (10 types of rock samples), in which both Cu and Zn were selected as target elements. The following conclusions were drawn. The matrix effect classification of the rock samples was not completely controlled by the detectable elemental composition or petrographic classification. According to the correlation between the spectral parameters and the characteristic X-ray intensity of the target elements, the matrix effect classification of rock samples could be more detailed. For rock samples in the same rock classification set and with the same target element content, the target element Kα X-ray intensity could be accurately described by the main spectral parameters, which include the scattering background at an energy interval of 4–19 keV, the Compton peak and the Rayleigh peak. Then, a matrix effect correction method was established, which allowed for the fast measurement of the target element contents of different rock samples by following the same parameters of the quantification procedure in the in situ EDXRF analysis. Simultaneously, six types of rock samples containing 3% Zn were selected as the validation set, and the relative errors of the target element content measurements in different rock samples were all less than 6% using the same parameters of the quantification procedure.