Application of pXRF technique to the determination of major and trace elements in large soil datasets, advantages and limitations

Soil plays a fundamental role for both ecosystems and humans, and its characterization is an essential tool for strategic planning in areas such as agriculture, natural resource exploration and detection of potential soil contamination. To achive an effective geochemical characterization, it is necessary to study the background geochemical levels, considering the local geology of the environment and anomalous levels derived from point mineralization and anthropogenic activities such as mining or industrial activities. These background and reference levels are often published by each country or region. Although many regions in Spain have determined these levels, Castilla-La Mancha present incomplete partial information due to its extensive (79461 km2), complex geology (Variscan (meta)sedimentary and granitic domain, Alpine domain, Undeformed Mesozoic domain and Post-Alpine domain) and diverse mining activities (Almadén Hg district, Pg-Ag-Zn-Cu districts, Sb district, Campos de Calatrava volcanic field, clay, granite, Ca salts, Na salts and Diatomites mining areas). Given this scenario, the main objective of this work is to evaluate the applicability of portable x-ray fluorescence (pXRF) techniques to the determination of generic reference levels from large datasets. The proposal involves the analysis of a shortened pretreatment method to get an aliquot for analysis, and an analytical data quality study using certified reference materials (Soil-1, Soil-2, NIST-R, STSP-2, STSP-3 and STSP-4). For this purpose, the fine fraction dried in the laboratory were analyzed with a pXRF spectrometer in different modes, with “mining mode” and “soil mode”. Optimization of analysis times was carried out, setting it in 45s as the optimum time, after testing 60 and 90 seconds. Recovery percentages of major elements using “Mining mode” ranged Al, Mgand Si and “Soil mode” Ca, Fe and K, and for trace elements using “Mining mode” ranged Nb and P and using “Soil mode” As, Ba, Cr, Cu, Mn, Ni, Pb, Rb, Sn, Sr, Ti, V, Zn, Zr. The following elements, Ag, Au, Cd, Co, Hg, Mo, Sn, Ta, Th and U have been discarded as being too different from the reference materials used. In conclusion, this rapid analysis technique offers an efficient solution for the characterization of large surface areas or the sampling of large number of samples. Of the analytical modes of pXRF, the “soil mode” is the most suitable in terms of quantification, providing results that are optimally adjusted to the analyzed patterns and on a larger number of analyzed elements.