Kinetics of Arsenate Reduction by Dissolved Sulfide ELIZABETH A. ROCHETTE, † BENJAMIN C. BOSTICK, ‡ GUANGCHAO LI, ‡ AND SCOTT FENDORF* ,‡ Department of Natural Resources, University of New Ham pshire, Durham , New Ham pshire 03824, and Departm ent of Geologic and Environm ental Sciences, Stanford University, Stanford, California 94305-2115 Arsenic toxicity and mobility in soil and aquatic environments depends on its speciation, with reducing environments generally leading to more hazardous conditions with respect to this element. Aqueous sulfide (H 2 S or HS - ) is a strong reductant and often occurs at appreciable concentrations in reduced systems. Consequently, it may play an integral part in arsenic redox chemistry.Therefore,reactions between arsenic and sulfide may strongly influence water quality in arsenic-contaminated systems. To evaluate this possibility, we investigated the kinetics and reaction pathways of arsenate with sulfide. Arsenate reduction by hydrogen sulfide is rapid and conforms to a second-order kinetic model, having a rate constant, k ) 3.2 × 10 2 M -1 h -1 , that is more than 300 times greater at pH 4 than at pH 7. However, arsenite is not the direct reaction product. Rather, arsenic -sulfide complexes develop, including the formation of a trimeric species (H x As 3 S 6 x-3 ), that persist in solution for several days, ultimately dissociating and leading to the production of dissolved arsenite. The precipitation of orpiment is dominant only at high (20:1) S:As ratios, considering the reaction conditions used in this study (133 μM As, pH 4). Hence, models of arsenic behavior in the environment should consider abiotic reduction of arsenate by sulfide, at least under moderately acidic conditions, and the possibility of dissolved arsenic -sulfide complexes. Introduction Agricultural soils as well as stream and lake sediments may contain arsenic from historic arsenicalpesticide applications, miningpractices,and industrialactivities.Variouspathways have been proposed for arsenate (HxAsO4 x-3 )reduction,such as reduction and alkylation (1) and through respiration coupled with the oxidation oforganicmoleculesor hydrogen (2-5). Microbial activity is also thought to be indirectly or directly responsible for arsenate reduction to arsenite (HxAsO3 x-3 ) in flooded soils and sediments (6-8), with microbially stimulated reduction demonstrated in these natural materials (9-11). Nonetheless, abiotic arsenate reduction processes should not be dismissed. Sulfide in the form of H2S or HS - is frequently present at appreciable concentrations in reducing environments and is a strong reductant (see,for example,ref 12).On thebasisofstandard- state redox conditions, hydrogen sulfide (H2S) should be capable ofreducingarsenate (H2AsO4 - or HAsO4 2- )to arsenite (H3AsO3): On the basis of the free energies of formation provided by Wagman et al. (13), ΔG°rxn )-100.5 kJ/ mol. Newman et al. (4) noted that the kinetics of arsenate reduction by sulfide at pH 6.8 are slow (second-order rate constant, k pH6.8 of 1.04 M -1 h -1 ). However, other pH values were not considered,and the conversion ofsulfide from HS - to H2S below pH 7 could vastly change the interactions of sulfide with arsenate. For instance, within 33 h nearly complete conversion of arsenate to arsenite occurs at pH 4, while less than 10%occurs at pH 7with initialsulfide:arsenate molar ratios of roughly 10 4 :1 (14). Furthermore, arsenate reduction to arsenite has been observed in fjord waters at a depth where H2S is dominant over O2 (15). Once reduced,As(III)mayalso form soluble and insoluble arsenic-sulfide complexes. Several soluble arsenic-sulfide complexes,such as di- and trithioarsenite monomers as well as dimeric and trimeric arsenic-sulfur complexeshave been described or suggested during orpiment (As2S3) dissolution (16-24).Because ofthe environmentalsignificance ofarsenic redox reactions and the potential for it reacting with sulfide in acidic systems, we conducted this study to determine the kinetics and reaction pathways of arsenate with dissolved sulfide. Experimental Section Chemicals. Sodium sulfide (Na 2S‚9H2O, 99%), sodium boro- hydride (NaBH4, 98-99%),ferric ammonium sulfate [FeNH4- (SO4)2‚12H2O,99%],and N ,N -dimethyl- p -phenylenediamine sulfate [(C2H5)2NC6H4NH2‚H2SO4, 97%] were obtained from Sigma-Aldrich ChemicalCo.(St.Louis,MO).Sulfidic solutions were prepared using degassed doublydeionized water in an O2-free environment (i.e., within a N2-purged glovebox). Sodium arsenate (NaH2AsO4‚7H2O), sodium arsenite (NaAsO2), glacialacetic acid (CH3COOH),sodium acetate (CH3COONa ‚ 3H2O), Tris [(HOCH2)3CNH2],sodium hydroxide (NaOH),and concentrated hydrochloric acid (HCl) were reagent-grade chemicals. Acetate buffers (0.1 M) were used to maintain a constant pH of 4.0 or 5.0 and were prepared by combining acetic acid and sodium acetate to obtain the desired pH.Tris buffer was used to maintain a pH of 7 and was prepared by adjusting the pH of 0.1 M Tris with the addition of concentrated hydrochloric acid. Kinetic Experiments. We first explored the effect ofsulfide and arsenate concentrations on the ensuing redoxreaction. Sodium arsenate was added to 0.1 M acetate buffer to obtain a concentration of 133 μM arsenate. Buffered (pH 4, 5, or 7) arsenate solutions in glass bottles were placed in a nitrogen- purged glovebox along with a concentrated solution of sodium sulfide; dissolved sulfide was added to the buffered arsenate solutions in the glovebox to obtain initial sulfide concentrationsof13.3,26.6,133,and 266 μM sulfide,yielding sulfide:arsenate ratios of 1:10, 1:5, 1:1, and 2:1, respectively. Solution pH and Eh were determined, and then the bottles were capped and placed on a shaker at 150 rpm for ca. 96 h, after which time the samples were filtered through 0.22- μm pore filters. These experiments were performed in duplicate at 24 ( 2 °C. To obtain greater detail on the reduction of arsenate by sulfide, we investigated the reaction at pH 4 and an initial sulfide:arsenate ratio of2:1 (266:133 μM)whilesamplingmore *Correspondingauthor phone: (650)723-5238;fax: (650)725-2199; e-mail: Fendorf@stanford.edu. † University of New Hampshire. ‡ Stanford University. H 2 AsO 4 - + H 2 S + H + S H 3 AsO 3 + 1/ 8S 8 + H 2 O (1) Environ. Sci. Technol. 2000, 34, 4714-4720 4714 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 22, 2000 10.1021/es000963y CCC: $19.00  2000 American Chemical Society Published on Web 09/30/2000