Due to potential hazardous effects, understanding the fate of antimony (Sb) in soils is essential. In soils, Sb(V) and Sb(III) are the main oxidation states. Both Sb(V) and Sb(III) can be strongly sorbed onto iron (Fe) and aluminium (Al) hydroxides, by inner- or outer sphere surface complexation and this can be described by surface complexation modelling. It is hypothesized that Sb in soils originating from historical contamination may be less mobile than recently added Sb. The mechanism of this so-called ageing can be attributed to solid-state diffusion of Sb into the soil matrix or to incorporation into the Fe (hyd)roxide crystal lattice, both leading to Sb fixation. The aims of this study were (i) to relate time-dependent changes in the Sb mobility in soils to the Sb speciation and coordination environment in the solid phase by use of Sb K-edge XANES/EXAFS spectroscopy and (ii) to describe the solubility of Sb in soils by surface complexation modelling with the CD-MUSIC model for ferrihydrite.
Nine soil samples with a range in soil organic carbon concentration, CaCO3 , pH and amorphous Fe and Al concentration were amended with 500 mg Sb(V) kg-1 via a stock solution of KSb(OH)6. This high concentration was needed to ensure good quality of the Sb K-edge EXAFS spectra and is a good proxy for Sb-contaminated soils. Model soil hydroxides (hydrous Fe and Al oxides) were prepared and spiked with Sb(V) at a Sb to Fe/Al ratio of 0.1. The samples were incubated at 20°C for 6 months at field-moist conditions. To evaluate changes in Sb sorption over time, freshly spiked samples were compared with the 6-month aged soil samples. The concentrations of Sb in the soil pore water decreased on average by a factor of three in the aged soils compared to Sb concentrations in the pore water of freshly contaminated soils. This resulted in increased solid:liquid distribution coefficients for both the spiked soils and the Sb(V) spiked Fe hydroxide samples (ferrihydrite), but not for the Al hydroxide. In the soils, the pH remained constant, however, in 2 soils, a pH drift towards lower pH was observed. The solid phase speciation analysis showed that over time and in all samples the redox speciation of the added Sb(V) did not change. No precipitation of ferric or calcium antimonates in the soil samples occurred. Preliminary results of the EXAFS modelling showed that the added Sb(V) is adsorbed to both Fe as Al hydroxides in the soil samples and that, with ageing, the Sb(V) coordination environment changes in the second shell. These changes were observed also for the Sb(V) spiked ferrihydrite, but not for the Al hydroxide, possibly indicating that Fe hydroxides are controlling Sb ageing in soils.
At the conference, further EXAFS modelling of these results will be presented. Furthermore, the EXAFS data will be used to constrain a surface complexation model to describe Sb(V) solubility in soils.