Automated mineralogy for quantification and partitioning of metal(loid)s in particulates from mining/smelting-polluted soils
Abstrakt
Topsoils near active and abandoned mining and smelting sites are highly polluted by metal(loid) contaminants, which are often bound to particulates emitted from ore processing facilities and/or windblown from waste disposal sites. To quantitatively determine the contaminant partitioning in the soil particulates, we tested an automated mineralogy approach on the heavy mineral fraction extracted from the miningand smelting-polluted topsoils exhibiting up to 1920 mg/kg As, 5840 mg/kg Cu, 4880 mg/kg Pb and 3310 mg/kg Zn.
A new generation of automated scanning electron microscopy (autoSEM) was combined and optimized with conventional mineralogical techniques (XRD, SEM/EDS, EPMA). Parallel digestions and bulk chemical analyses were used as an independent control of the autoSEM-calculated concentrations of the key elements.
This method provides faster data acquisition, the full integration of the quantitative EDS data and better detection limits for the elements of interest. We found that As was mainly bound to the apatite group minerals, slag glass and metal arsenates.
Copper was predominantly hosted by the sulfides/sulfosalts and the Cu-bearing secondary carbonates. The deportment of Pb is relatively complex: slag glass, Fe and Mn (oxyhydr)oxides, metal arsenates/vanadates and cerussite were the most important carriers for Pb.
Zinc is mainly bound to the slag glass, Fe (oxyhydr)oxides, smithsonite and sphalerite. Limitations exist for the less abundant contaminants, which cannot be fully quantified by autoSEM due to spectral overlaps with some major elements (e.g., Sb vs.
Ca, Cd vs. K and Ca in the studied soils).
AutoSEM was found to be a useful tool for the determination of the modal phase distribution and element partitioning in the metal(loid)-bearing soil particulates and will definitely find more applications in environmental soil sciences in the future. (c) 2020 The Authors. Published by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).