Correction of Matrix Effect in Multielemental Quantitative Analysis by X-Ray Fluorescence Spectroscopy Using the Linear Behavior in the Analytical Range of the Substitution-Dilution Method

Abstract

This paper is an analytical study of the possibility of applying the linear range of the substitution-dilution method to correct the matrix effect in quantitative analysis by X-ray fluorescence (XRF) spectroscopy. The analytical range is obtained from a series of samples prepared in the form of glass discs by substituting the unknown sample with a standard sample (substitution factor, h) including a diluent-melt. In general, the substitution-dilution method is hyperbolic in character and therefore the diluent is required to ensure linear behavior between I_f vs. h in the experimental range. The linear range is located between the concentrations of standard and unknown for each element analyzed. This linear model makes it possible to correct the matrix effect in quantitative analysis by XRF using a single multi-elemental standard for different types of samples with a complex matrix, such as geologicals and cements. The results found for Si, Ti, Al, Fe, Mn, Ca, K, and P in soil and sediment samples and Si, Fe, Al, Ca, and K in cements (white and gray) are statistically satisfactory. Thus, the mean relative standard deviation calculated for all analytes in each sample was: ±4.0% and ±5.0% in soils; ±5.0% in sediments; and ±6.0% or ±3.0% in cements, white and gray, respectively.