ABSTRACT: Iron is prone to change its form and speciation in phases. Thus, several methods have been developed to estimate iron partitioning in the mineral phases of soils. However, the accuracy of these methods to evaluate the iron contribution from minor phases, such as actinolite, almandine, biotite, chlorite, epidote, hornblende, muscovite, and Fe-diospide, remains low. Furthermore, most of the current iron speciation research is focused on bulk samples, and only applies to soil samples that are mostly composed of clays or clay minerals, without a wide-ranging evaluation of soil particles with different grain sizes. In this study, we classified several iron phases using a mineral liberation analyzer on desert soil particles with diameters ranging from silt- to fine-sand (5-20 µm, 20-45 µm, 45-63 µm, 63-75 µm, and > 75 µm). The iron containing minor phases were identified, the modal mineral abundances were determined via matching with the standard energy dispersive spectra library, and a particle size analysis was performed using mineral processing tools on each of the examined 40,000 particles. The iron partition results were ultimately established based on the standard iron concentration in the mineral phases and the modal mineral abundances.•This new method could be automated, thereby facilitating high efficiency identification of iron-containing phases that would allow, for the first time, the generation of a dataset for iron partitioning in soil particles.•This method allows the identification of minor iron phases in soil particles, and permits in situ mapping of iron mineralogy in fine sand- to silt-sized soil particles.•Not restricted by single mineral particles, this method considers multi-phase complex particles. Thus, it largely improves the accuracy for estimating the iron partition parameter.
