Examination of the reliability of X-ray powder diffraction analysis to determine mineral composition of soils

Abstract

X-ray powder diffraction (XRPD) is an effective technique for identifying and quantifying mineral types in soil. However, few studies have compared quantitative values based on XRPD with those from conventional wet chemical methods (WCMs). Here, we determined the primary mineral content in artificial mineral mixtures and 79 agricultural soils from across Japan using WCMs and two XRPD-based quantitative methods: the mineral intensity factor (MIF) and the full-pattern summation (FPS) methods performed with the powdR package for R. For artificial mixtures, the accuracy of mineral content determination (i.e., micas, quartz, K-feldspar, and plagioclase) followed the order: WCMs > FPS > MIF. For Japanese agricultural soils, the contents of each mineral were highly similar between WCMs and FPS, based on mean absolute differences and correlation coefficients. Alternatively, MIF displayed lower similarities with WCMs, likely due to preferred orientation and peak shift or overlap issues. Using the FPS method, the mica and amorphous phase contents were positively and significantly correlated with nonexchangeable K content and cation exchange capacity, respectively. Additionally, the plagioclase content was negatively and significantly correlated with clay content. Thus, the powdR-based FPS method is recommended for determining the mineral composition of soils, as it allows for a clearer and more quantitative demonstration of the relationship between individual minerals and soil properties.