Experimental analysis of arsenic precipitation during microbial sulfate and iron reduction in model aquifer sediment reactors

Microbial SO42− reduction limits accumulation of aqueous As in reducing aquifers where the sulfide that is produced forms minerals that sequester As. We examined the potential for As partitioning into As- and Fe-sulfide minerals in anaerobic, semi-continuous flow bioreactors inoculated with 0.5% (g mL−1) fine-grained alluvial aquifer sediment. A fluid residence time of three weeks was maintained over a ca. 300-d incubation period by replacing one-third of the aqueous phase volume of the reactors with fresh medium every seven days. The medium had a composition comparable to natural As-contaminated groundwater with slightly basic pH (7.3) and 7.5 μM aqueous As(V) and also contained 0.8 mM acetate to stimulate microbial activity. Medium was delivered to a reactor system with and without 10 mmol L−1 synthetic goethite (α-FeOOH). In both reactors, influent As(V) was almost completely reduced to As(III). Pure As-sulfide minerals did not form in the Fe-limited reactor. Realgar (As4S4) and As2S3(am) were undersaturated throughout the experiment. Orpiment (As2S3) was saturated while sulfide content was low (∼50 to 150 μM), but precipitation was likely limited by slow kinetics. Reaction-path modeling suggests that, even if these minerals had formed, the dissolved As content of the reactor would have remained at hazardous levels. Mackinawite (Fe1 + xS; x ⩽ 0.07) formed readily in the Fe-bearing reactor and held dissolved sulfide at levels below saturation for orpiment and realgar. The mackinawite sequestered little As (<0.1 wt.%), however, and aqueous As accumulated to levels above the influent concentration as microbial Fe(III) reduction consumed goethite and mobilized adsorbed As. A relatively small amount of pyrite (FeS2) and greigite (Fe3S4) formed in the Fe-bearing reactor when we injected a polysulfide solution (Na2S4) to a final concentration of 0.5 mM after 216, 230, 279, and 286 days. The pyrite, and to a lesser extent the greigite, that formed did sequester As from solution, containing 0.84 and 0.23 wt.% As on average, respectively. Our results suggest that As precipitation during Fe-sulfide formation in nature occurs mainly in conjunction with pyrite formation. Our findings imply that the effectiveness of stimulating microbial SO42− reduction to remediate As contamination may be limited by the rate and extent of pyrite formation and the solubility of As-sulfides.