DFG The influence of thioarsenic species formation on arsenic complexation to natural organic matter

DFG PL 302/20-1

Under anoxic conditions, arsenic (As) in its trivalent form (arsenite) has been reported to be completely sequestered by binding to natural organic matter (NOM) via sulfhydryl functional groups. Laboratory experiments showed that even when conditions turned oxic, arsenite bound to S(-II)-NOM had a half-life of more than 300 days and its stability was higher than that of the more commonly investigated arsenite sorbed to Fe(III) oxyhydroxides. The global implication is that organic-rich, sulfidic environments, such as wetlands, are important quantitative As sinks.

However, all mechanistic studies carried out so far have exposed arsenite to pre-formed S(-II)-NOM, ignoring that in a system containing As(III), S(-II), and NOM, complexation of As(III) and S(-II) will form thioarsenites (H2AsIIIS-IInO3-n-, n=1-3) and thioarsenates (HAsVS-IInO4-n2-, n=1-4). Our central hypothesis is that kinetics of thioarsenic species formation in solution is faster than As(III) or S(-II) sorption to NOM and that therefore thioarsenic species determine extent and kinetics of As sorption in organic-rich, sulfidic environments. Competitive sorption to co-occurring (meta)stable iron mineral phases will also differ from known behavior of arsenite.

Due to their instability and a lack of pure standards, sorption behavior of thioarsenites is unknown, yet. For thioarsenates, no information exists about binding to NOM, but binding to different Fe(III) minerals is known to be lower than that of arsenite. We hypothesize that thioarsenates show less and slower sorption than arsenite to S(-II)-NOM, because the covalent inorganic S bonds in thioarsenates decrease the affinity for S(-II)-NOM complexation. Thioarsenate binding to Fe(III)-NOM should be lower and slower in analogy to the known lower affinity for Fe(III) mineral phases. We also hypothesize that sulfidation causes faster and more As remobilization than the previously investigated oxidation because abiotic oxidation is slow, while S-complexation with As occurs spontaneously and destabilizes As binding to NOM and Fe mineral phases.

To test our hypotheses, batch experiments will be conducted with mono- and trithioarsenate standards and an arsenite-sulfide mixture (forming thioarsenites) at pH 5, 7, and 9 with two selected NOMs (Federseemoor Peat and Elliott Soil Humic Acid; each non-reacted, S(-II)- and Fe(III)-reacted). Sorption affinity and kinetics will be determined and binding mechanisms will be identified using X-ray absorption spectroscopy. The stability of (thio)arsenic-loaded NOMs will be investigated under oxidizing but also under sulfidic conditions and preferential sorption in binary systems (combinations of iron oxyhydroxides, Fe(III)-NOM, S(-II)-NOM, and iron sulfides) will be determined.

The overall goal is further elucidating As binding mechanisms in systems containing S(-II), Fe, and NOM to help predict under which conditions As sinks will turn into As sources.