Comparative DFT study of inner-sphere As(iii) complexes on hydrated α-Fe2O3(0001) surface models

Abstract

The long-recognized risk to human health arising from arsenic-contaminated waters is known to be linked to partitioning reactions between arsenic and natural solid phases. Currently, the ability to predict As surface complexation is limited by the lack of molecular-level understanding of As-solid interactions. In the present study, we use density functional theory (DFT) to model mono-, bi-, and tri-dentate As(III) surface complexes on different (previously proposed) structural models for hydrated hematite, modeled as α-Fe₂O₃(0001)–water interfaces. One of the modeled hematite–water interfaces is terminated entirely by hydroxyl surface functional groups, comprised of hematite lattice oxygen atoms. The other hematite–water interface is an iron-terminated model in which the outermost oxygen functional groups are water (and water dissociation product) ligands. We report the DFT trends in adsorption energies in terms of As-hematite coordination, hematite surface geometry/stoichiometry, and oxygen functional group identity. The DFT energetics predict that a monodentate As(III) surface complex is preferred on both hematite–water structures, suggesting that the two structural models here employed do not sufficiently represent the true surface structure to reproduce the experimental observation of As(III) bidentate coordination. However, the results do elucidate fundamental concepts of interface reactivity: A key result, supported by electronic structure analysis, is that ligand oxygen functional groups cannot be treated on equal ground with true surface oxygen functional groups. For the systems modeled here the distinction between surface and ligand functional groups supersedes the differences in oxygen coordination with surface Fe. We discuss the impact of this finding on the application of bond-valence-based predictions of mineral–water reactivity, and use the results of this study to pose questions and directions for ongoing modeling efforts aimed at linking macroscopic reactivity with molecular-level understanding.