Sb(III) and Sb(V) retention by aptian sands: Quantitative insights into adsorption and heterogeneous redox reactions controlled by iron-containing minerals

Abstract

The retention of Sb(III) and Sb(V) by Aptian sands, which contain quartz and glauconite, was investigated as a function of Sb concentration and with 7 d and 30 d contact time. Two samples, collected along a vertical profile, were used for the experiments. They originated from the capillary fringe (S_CFr) and saturated (S_Sat) horizons and were respectively representative of oxidizing and reductive redox conditions. The Aptian sands had a lower retention capacity for Sb(V) than for Sb(III). Sb(V) distribution coefficients (R_d) ranged between 10 ± 3 L∙kg⁻¹ and 13 ± 4 L∙kg⁻¹ in S_Sat and between 11 ± 3 L∙kg⁻¹ and 19 ± 4 L∙kg⁻¹ in S_CFr, after 7 d interaction. These values almost doubled when the interaction time was increased to 30 d and, in all conditions, Sb(V) was the only species detected in the aqueous solution. The evolution of Sb(III) R_d as a function of Sb equilibrium concentration and time was more complex. When contact time was 7 d, R_d first increased from 68 ± 7 L∙kg⁻¹ ([Sb]_eq = 2.2∙10⁻⁸ M) to 547 ± 33 L∙kg⁻¹ ([Sb]_eq = 1.4∙10⁻⁷ M) and then decreased down to 323 ± 21 L∙kg⁻¹ ([Sb]_eq = 1.1∙10⁻⁶ M) for S_Sat after 7 d of interaction. Sb(III) retention by S_CFr was slightly lower: R_d increased from 15 ± 4 ([Sb]_eq = 3.4∙10⁻⁸ M) to 174 ± 13 L∙kg⁻¹ ([Sb]_eq = 7.7∙10⁻⁷ M) and then decreased down to 137 ± 10 L∙kg⁻¹ ([Sb]_eq = 2.3∙10⁻⁶ M). While Sb(V) was the only species detected in the aqueous solution at lower concentrations, Sb(III) was increasingly present when added Sb(III) concentration increased, hence pointing out to heterogeneous redox interaction between Sb(III) and the solid phase that could be successfully modeled.

Overall, this study demonstrated that Sb(III) and Sb(V) retention depended on the speciation of added Sb, on the redox reactivity of the sand samples, and on contact time. Such results are a step forward to our capacity to model, and hence quantitatively predict, Sb mobility in the natural environment.