ABSTRACT

Selenium (Se) is a micronutrient for human with a very narrow range between deficit and toxic levels. Because selenate (SeO₄^2-) is the most prevalent form of bioavailable Se in agricultural soils, the prediction of selenate adsorption/desorption behaviors on soil constituting minerals is an important step to minimize potential health problems. This article describes a comprehensive predictive model for selenate adsorption on oxide minerals applicable for a wide range of pH, ionic strength, selenate concentration, solid/water ratio, and amorphous and crystalline adsorbents. The principle selenate surface species are a monodentate-mononuclear inner-sphere and a bidentate-binuclear outer-sphere complexes represented by >MOH + SeO₄^2- + H+ → >MOSeO₃^- + H2O and 2>MOH + SeO₄^2- + 2H+ → (>MOH₂⁺)2_SeO₄^2-, respectively, for all oxides except for ferrihydrite, which requires an additional surface species, 2>MOH + SeO₄^2- + 2H+ → (>MOH₂⁺)(>MOSeO₃^-) + H₂O. With the three reactions formulated in the extended triple layer model, we demonstrate adequate reproduction of a large number of adsorption, surface titration, and proton co-adsorption data for various selenate-oxide systems reported in literature. Calculations with the theoretically derived selenate adsorption constants indicate that, in addition to aqueous condition, crystalline states of Fe and Al oxides significantly influence the surface adsorption affinity for selenate. Thus, our developed model is useful not only for estimating the solid-liquid partitioning of selenate by soil oxide minerals, but also for assessing its response to changing aqueous conditions and ageing processes.