Surface complexation of arsenic(V) to iron(III) (hydr)oxides: structural mechanism from ab initio molecular geometries and EXAFS spectroscopy

Abstract

Arsenic(V), as the arsenate (AsO4)3− ion and its conjugate acids, is strongly sorbed to iron(III) oxides (α-Fe2O3), oxide hydroxides (α–,γ–FeOOH) and poorly crystalline ferrihydrite (hydrous ferric oxide). The mechanism by which arsenate complexes with iron oxide hydroxide surfaces is not fully understood. There is clear evidence for inner sphere complexation but the nature of the surface complexes is controversial. Possible surface complexes between AsO4 tetrahedra and surface FeO6 polyhedra include bidentate corner-sharing (2C), bidentate edge-sharing (2E) and monodentate corner-sharing (1V). We predicted the relative energies and geometries of AsO4-FeOOH surface complexes using density functional theory calculations on analogue Fe2(OH)2(H2O)nAsO2(OH)23+ and Fe2(OH)2(H2O)nAsO4+ clusters. The bidentate corner-sharing complex is predicted to be substantially (55 kJ/mole) more favored energetically over the hypothetical edge-sharing bidentate complex. The monodentate corner-sharing (1V) complex is very unstable. We measured EXAFS spectra of 0.3 wt. % (AsO4)3− sorbed to hematite (α-Fe2O3), goethite(α–FeOOH), lepidocrocite(γ–FeOOH) and ferrihydrite and fit the EXAFS directly with multiple scattering. The phase-shift-corrected Fourier transforms of the EXAFS spectra show peaks near 2.85 and 3.26 Å that have been attributed by previous investigators to result from 2E and 2C complexes. However, we show that the peak near 2.85 Å appears to result from As-O-O-As multiple scattering and not from As-Fe backscatter. The observed 3.26 Å As-Fe distance agrees with that predicted for the bidentate corner-sharing surface (2C) complex. We find no evidence for monodentate (1V) complexes; this agrees with the predicted high energies of such complexes.