Article

pubs.acs.org/est

Stimulation of Fe(II) Oxidation, Biogenic Formation, and Arsenic Immobilization by Pseudogulbenkiania Sp. Strain 2002 † ‡ † ‡ ‡ ‡ ‡ † ‡ Wei Xiu, , Huaming Guo,*, , Jiaxing Shen, Shuai Liu, Susu Ding, Weiguo Hou, Jie Ma, † and Hailiang Dong † State Key Laboratory of Biogeology and Environmental Geology, China University of Geosciences, Beijing 100083, P.R. China ‡ School of Water Resources and Environment, China University of Geosciences, Beijing 100083, P.R. China

*S Supporting Information

ABSTRACT: An anaerobic nitrate-reducing Fe(II)-oxidizing bacterium, Pseudogulbenkiania sp. strain 2002, was used to investigate As immobilization by biogenic Fe oxyhydroxides under different initial molar ratios of Fe/As in solutions. Results showed that Fe(II) was effectively oxidized, mainly forming lepidocrocite, which immobilized more As(III) than As(V) without changing the redox state of As. When the initial Fe/As ratios were kept constant, higher initial Fe(II) concentrations immobilized more As with higher Asimmobilized/Feprecipitated in biogenic lepidocrocite. EXAFS analysis showed that variations of initial Fe(II) concentrations did not change the As−Fe complexes (bidentate binuclear complexes (2C)) with a fixed As(III) or As(V) initial concentration of 13.3 μM. On the other hand, variations in initial As concentrations but fixed Fe(II) initial concentration induced the co-occurrence of bidentate binuclear and bidentate mononuclear complexes (2E) and bidentate binuclear and monodentate mononuclear complexes (1V) for As(III) and As(V)-treated series, respectively. The coexistence of 2C and 2E complexes (or 2C and 1V complexes) could contribute to higher As removal in experimental series with higher initial Fe(II) concentrations at the same initial Fe/As ratio. Simultaneous removal of soluble As and nitrate by anaerobic nitrate- reducing Fe(II)-oxidizing bacteria provides a feasible approach for in situ remediation of As-nitrate cocontaminated groundwater.

■ INTRODUCTION As(V) were effectively removed from aquatic systems by biogenic Fe(III) oxide precipitation under oxic conditions by Arsenic (As), a prevalent contaminant found in drinking 21 21 ff Gallionella ferruginea, Leptothrix ochracea, Pseudomonas sp. groundwater, a ects hundreds of millions of people via 17 consumption of As-contaminated groundwater.1 Arsenic strain GE-1, etc. and under anoxic conditions by Acidovorax sp. strain BoFeN1,16 Rhodobacter ferrooxidans strain SW2,16 and contamination in groundwater is usually a result of both 16 natural processes and anthropogenic activities. Natural strain KS, among others. processes include geothermal sources, weathering of As-bearing Because groundwater As is usually mobilized under anoxic , and release of -adsorbed As due to microbial conditions, anaerobic nitrate-reducing Fe(II) oxidizers may be beneficial for in situ remediation of As-contaminated ground- activity, whereas anthropogenic activities include mining, wood 22 1−8 water. Different Fe(III) (oxyhydr)oxide phases were formed preservation, and As-containing pesticide utilization. The 16,19,25,26 dominant inorganic species of As in aqueous environments in the presence of anaerobic Fe(II) oxidizers. More- (groundwater and surface water) are arsenate [As(V)] and over, relative to synthetic abiogenic Fe(III) minerals, biogenic 1 Fe(III) minerals are mostly characterized by cell−mineral arsenite [As(III)]. Arsenic(III) is more mobile at neutral pH 23,24 − and 25−60 times more toxic (acute poisoning) than As(V).8,9 aggregates and possibly contain cell Fe/As complexes in − 2− the presence of As(III) and As(V), which would contribute to Arsenate oxyanions (e.g., H2AsO4 and HAsO4 ) and arsenite 16 species (e.g., H AsO 0) are usually adsorbed onto the surfaces greater As immobilization. Furthermore, binding mechanisms 3 3 of As would be expected to vary depending on Fe(III) mineral of Fe(III) (oxyhydr)oxide minerals over a wide pH range, 2 forming strong inner-sphere as well as outer-sphere surface phases, including bidentate binuclear complexes ( C), bidentate − 2 complexes.10 12 mononuclear complexes ( E), and monodentate mononuclear Fe(II) oxidation by Fe(II)-oxidizing bacteria produces not only poorly crystalline Fe(III) oxyhydroxide (e.g., ) Received: February 2, 2016 but also crystalline phases (e.g., green rusts, , and Revised: May 24, 2016 − magnetite),13 20 which have the potential to coprecipitate or Accepted: May 25, 2016 adsorb As. Previous studies showed that both As(III) and Published: May 25, 2016

© 2016 American Chemical Society 6449 DOI: 10.1021/acs.est.6b00562 Environ. Sci. Technol. 2016, 50, 6449−6458 Environmental Science & Technology Article

− complexes (1V).27 29 These different As binding modes As(V) molar ratios (37.5, 75, 375, and 750 with initial As(V) contribute to the differences in As immobilization, with 2E concentration of 13.3 μM) and three initial Fe/As(V) molar and 1V complexes showing greater As immobilization than 2C ratios (37.5, 750, and 1500 with the corresponding initial As(V) complexes.6 Therefore, more investigation is required to reveal concentrations of 266.7, 13.3, and 6.67 μM) were set for Fe/ phases of biogenic Fe(III) (oxyhydr)oxides and binding fixed As(V) series and fixed Fe/As(V) series, respectively. The mechanisms of As in the presence of anaerobic Fe(II) oxidizers. higher toxicity of As(III) compared to As(V) made it difficult to Pseudogulbenkiania sp. strain 2002 has been identified as an work with such high initial concentrations of As(III). anaerobic neutrophilic Fe(II)-oxidizer, which oxidizes Fe(II) Therefore, we obtained the similar Fe/As(III) ratios by using nitrate as the electron acceptor and results in the proportionally decreasing initial concentrations of Fe(II) and formation of biogenic Fe(III) minerals.30,31 However, it is As(III). Four initial Fe/As(III) ratios (37.5, 75, 150, and 375 unclear how As immobilization occurs during the formation of with initial As(III) concentration of 13.3 μM) and three initial biogenic Fe(III) oxides induced by strain 2002. The effects of Fe/As(III) ratios (37.5, 75, and 375 with the corresponding As speciation on the formation of biogenic Fe(III) oxides and initial As(III) concentrations of 133.3, 66.7, and 13.3 μM) were the mechanism of As immobilization during the Fe(II) used for Fe/fixed As(III) and fixed Fe/As(III) series, oxidation also are not well understood. respectively. Because of these knowledge gaps, the objectives of this study For biotic control free of As, we used anoxic PIPES (10 mM, ff ff are to (i) investigate the e ects of As species on biogenic Fe(II) pH 7.0) bu er amended with Fe(II) (5 mM, FeCl2) as the sole oxidation by strain 2002; (ii) determine the effects of molar electron donor and nitrate (10 mM) as electron acceptor. Heat- ratios of initial Fe(II) relative to As(III) or As(V) on As killed controls were prepared by pasteurizing the inoculum in removal by biogenic Fe(III)-oxide minerals; (iii) reveal the an autoclave at 121 °C for 20 min. Assays were carried out mechanisms of As(III) and As(V) removal by biogenic Fe(III)- under nongrowth medium without an organic carbon source in ff oxides. an anoxic (100% N2 atmosphere) PIPES bu er (10 mM, pH 7.0) at 30 °C. Supernatant samples were anoxically taken at ■ MATERIALS AND METHODS different time intervals, filtered with 0.22 μm membrane filter, Bacterial Strain and Cultivation Conditions. Pseudogul- and analyzed for total soluble As, As species, Fe species, nitrate, benkiania sp. strain 2002 (ATCC BAA-1479; DSM 18807), an and nitrite. oxide precipitates were sampled following anaerobic nitrate-dependent Fe-oxidizing bacterium isolated repeated centrifugation (1000 rpm for 10 min) and rinsed with from a freshwater lake in Illinois,32 was retrieved from lab stock deionized water. The precipitates were then dried in the (20% glycerol at −80 °C, Geomicrobiology Lab at the China anaerobic glovebox and preserved in anaerobic amber glass University of Geosciences (Beijing)). Strain 2002 was bottles with a headspace of N2/H2 (92.5/7.5, v/v) and analyzed anaerobically cultured to the early stationary growth phase within 1 week. All experiments were conducted in duplicate, with an initial incubation in 5% (v/v) freshwater basal medium and results were reported using averages. · −1 · −1 · −1 · (0.25 g L NH4Cl, 0.6 g L NaH2PO4, 0.1 g L KCl, 0.42 g Sample Analysis. Details on the analysis of total soluble As, −1 · −1 · −1 L NaNO3, 2.52 g L NaHCO3,10mLL vitamin, and 10 As species, total Fe, nitrate, nitrite, and As K-edge X-ray mL·L−1 trace mineral solution). Contents of various absorption spectroscopy can be found in the Supporting components in vitamin and trace mineral solution were Information. Fe(II) concentration was determined spectro- detailed in the Supporting Information. Nitrate (10 mM) and metrically with a modified ferrozine assay at 562 nm.33 Samples acetate (10 mM) were used as the electron acceptor and donor, were mixed with 40 mM sulfamic acid (pH ∼ 1.8) instead of respectively. The cells were harvested by centrifugation (6000g, HCl because the sulfamic acid could react rapidly with nitrite ° 33 10 min, 30 C), washed twice with anoxic (100% N2 and prevent Fe(II) oxidation by nitrite at acidic pH. atmosphere) PIPES [piperazine-N,N-bis(2-ethanesulfonic Solid samples were examined by scanning electron acid)] buffer (10 mM, pH 7.0), and resuspended to serve as microscopy (SEM) to obtain the morphology information an inoculum for nongrowth Fe(II) oxidation experiments. All using a Zeiss Supra 35VP SEM at an accelerating voltage of 3 to chemicals used in this study were of analytical reagent grade, 10 kV, which was equipped with energy dispersive spectroscopy and all volumetric flasks and vessels were cleaned by soaking in (EDS) for chemical analysis. The biogenic solids were 10% HNO3 for at least 24 h, rinsed several times with deionized anaerobically sealed within two layers of Kapton tape in the water, and pasteurized in an autoclave at 121 °C for 20 min. anoxic glovebox and analyzed for Fe mineral phases by Experimental Setup. All experiments were carried out in synchrotron X-ray diffraction (μ-XRD), which was performed an anaerobic glovebox (Coy Laboratory Products, Grass Lake, at 10 keV (k = 0.6199 Å) on beamline BL15U at Shanghai USA) under a N2/H2 (92.5/7.5, v/v) atmosphere. Stock As(V) Synchrotron Radiation Facility (SSRF) with a Si(111) and As(III) solutions (5000 and 1000 mg/L) were prepared monochromator. The program MDI Jade 6.0 was used to · ff fi from sodium hydrogen arsenate (Na2HAsO4 7H2O, Fluka determine di raction patterns. Speci c surface area was Chemical) and sodium arsenite (NaAsO2, Fluka Chemical) determined for solid samples by Brunauer-Emmett-Teller ± ± with an initial pH of 8.4 0.1 and 10.3 0.2, respectively, (BET) N2 adsorption analysis using the Gemini VII 2390 using deionized water. To determine the effects of As(III) and (Micromeritics Instrument Corp., USA). As(V) on Fe(II) oxidation by strain 2002, different initial molar Arsenic K-edge EXAFS was recorded at 3.5 GeV and 300 mA ratios of Fe(II) to As(III) or As(V) were applied. Two different on beamline BL14W at SSRF, China. A Si(111) mono- methods were used to control the initial Fe(II)/As ratio: one chromator was used. The spectra were recorded at room fi − was to x the initial Fe(II) concentration (using 500 mM FeCl2 temperature and acquired in the energy range from 200 to as a stock solution) but vary the initial concentrations of As(III) +800 eV relative to the As K-edge. Binding modes of As(III) or As(V) (called fixed Fe/As series); the other was to fix the and As(V) to the Fe(III) (oxyhydr)oxide phases were identified As(III) or As(V) concentration but vary the initial concen- using EXAFS analysis. Data processing was performed using the trations of Fe(II) (called Fe/fixed As series). Four initial Fe/ program ATHENA, and theoretical fitting was performed using

6450 DOI: 10.1021/acs.est.6b00562 Environ. Sci. Technol. 2016, 50, 6449−6458 Environmental Science & Technology Article

Figure 1. Variation in Fe(II) concentrations as a function of incubation time under different initial Fe/(As(III) or As(V)) in the presence of strain 2002. For Fe/fixed As series, we fixed initial As(III) or As(V) concentration (13.3 μM) but varied initial concentrations of Fe(II) (between 0.5 and 5 mM for As(III)-treated series and between 0.5 and 10 mM for As(V)-treated series): (a) Fe/fixed As(III) series and (b) Fe/fixed As(V) series. For fixed Fe/As series, we fixed initial Fe(II) concentration (5 and 10 mM for As(III) and As(V)-treated series, respectively) but varied initial concentrations of As(III) or As(V) (between 13.3 and 133.3 μM for As(III)-treated series and between 6.67 and 266.7 μM for As(V)-treated series): (c) fixed Fe/As(III) series and (d) fixed Fe/As(V) series. The percentage of Fe(II) oxidation at the end of incubation is given.

ARTEMIS.34 Details for the spectra analysis can be found in reduction (Figure S1). These data support the earlier result that our previous study.6 higher As(III) concentration resulted in higher inhibitory effects on Fe(II) oxidation and nitrate reduction. However, the ■ RESULTS AND DISCUSSION inhibitory effects of As(V) were less obvious. The toxicity of As(III), especially at high concentrations, to microbial Fe(II) Oxidation by Strain 2002. For all biotic experi- metabolism also was found in biotic Fe(II) oxidation by ments, Fe(II) and nitrate concentrations decreased simulta- 17 Pseudomonas sp. strain GE-1, which showed that Fe(II) neously, and nitrite concentrations increased during Fe(II) oxidation rates decreased in As(III)-treated series. Similarly, an oxidation (Figures 1 and S1). In the controls with heat-killed obvious delay in growth of Acidovorax sp. strain BoFeN1, cells, no nitrate reduction, no Fe(II) oxidation, and no changes Rhodobacter ferrooxidans strain SW2, and strain KS was of either As concentration or speciation occurred (Figure S2). observed in the presence of high As(III) concentrations (500 The variations of Fe(II) concentrations in either the absence or μ 16 ffi presence of As can be divided into two stages: an initial rapid M). It was found that As(III) had a very high a nity for − thiol groups and therefore inhibited many enzymes that rely on decrease in Fe(II) concentration (Stage I: 0 6 h for Fe/As(III) 35 series, 0−4 h for Fe/As(V) series and As free series) and a thiol groups in critical positions. slight decrease in Fe(II) concentration (Stage II: 6−24 h for Arsenic Speciation in Liquid and Solid Phases. Arsenic Fe/As(III) series, 4−24 h for Fe/As(V) series and As free redox states did not change in all studied treatments. Arsenic series) (Figure 1). In each stage, the rate of Fe(II) oxidation speciation in solution showed that As(III) was below the depended on As speciation. For instance, at Stage I with the detection limit in As(V)-treated series, and no As(V) was same Fe/As ratio (375) and initial As concentration (13.3 μM), observed in the As(III)-treated series (Figure S3), which the overall rate of Fe(II) oxidation was either 0.969 or 0.560 indicated that strain 2002 did not induce As redox trans- mM−1·h−1, depending on whether the As was As(V) or As(III) formation under our experimental conditions, although it was in Fe/fixed As series (Table S1). These data suggest that reported that strain 2002 used As(V) as the electron acceptor 36 As(III) exerted a stronger inhibitory effect on Fe(II) oxidation coupled to phenol as an electron donor. The anaerobic than As(V), likely due to its higher toxicity to microbial nitrate-dependent Fe(II) oxidation of BoFeN1 also induced 16 metabolism. neither As(V) reduction nor As(III) oxidation. In addition to the As speciation effect, initial As No As redox transformation was observed in biogenic solid concentration affected strain 2002-mediated Fe(II) oxidation. phases. Lepidocrocite was the only biogenic crystalline phase The overall rate of Fe(II) oxidation decreased from 0.16 to 0.08 observed in our experiments (discussed below). XANES mM−1·h−1 when initial As(III) concentration increased from spectra of As(III)-loaded and As(V)-loaded biogenic lepidoc- 13.3 to 133 μM in As(III)-treated series. However, when initial rocite samples exhibited a well-resolved edge structure with an As(V) concentration increased from 6.67 to 266.7 μMin absorption maximum at 11 871.3 and 11 875.0 eV, which As(V)-treated series, the Fe(II) oxidation rate did not show any corresponds to As(III) and As(V), respectively (Figure S4). appreciable decrease (from 0.19 to 0.18 mM−1·h−1)(Table S1). This result is consistent with As-bearing lepidocrocite Similar effects of As(III) and As(V) also were found for nitrate synthesized under anaerobic conditions,29 which shows that

6451 DOI: 10.1021/acs.est.6b00562 Environ. Sci. Technol. 2016, 50, 6449−6458 Environmental Science & Technology Article

Figure 2. Biogenic Fe minerals produced during Fe(II) oxidation by strain 2002 (initial Fe/As(III) = 37.5 (a), 75 (b), 150 (c), and 375 (d) in Fe/ fixed As(III) series; initial Fe/As(III) = 37.5 (i), 75 (j), and 375 (k) in fixed Fe/As(III) series; initial Fe/As(V) = 37.5 (e), 75 (f), 375 (g), and 750 (h) in Fe/fixed As(V) series; initial Fe/As(V) = 37.5 (l), 750 (m), and 1500 (n) in fixed Fe/As(V) series; As free (o)). The lath-like structures and the irregularly shaped mass or sheet structures are indicated by red and yellow arrows, respectively. The elements in biogenic lepidocrocite are given in wt %.

As(III) was not oxidized in the presence of lepidocrocite. Ona- Arsenic speciation affected As removal by biogenic Nguema and co-workers also observed no As(III) oxidation by lepidocrocite. Higher molar ratios of Asimmobilized/Feprecipitated in lepidocrocite under anoxic conditions.37 biogenic lepidocrocite were observed in As(III)-treated series Arsenic Immobilization by Biogenic Fe Minerals. than in As(V)-treated series, indicating that biogenic During the strain 2002-induced Fe(II) oxidation, soluble As lepidocrocite removed more As(III) than As(V). The molar concentrations varied depending on initial Fe/As molar ratio ratios of Asimmobilized/Feprecipitated averaged 0.016 and 0.030 in (Figure 3). Direct As uptake by bacteria was negligible, as Fe/fixed As(III) and fixed Fe/As(III) series and 0.012 and demonstrated by stable As concentration in Fe(II)-free control 0.017 in Fe/fixed As(V) and fixed Fe/As(V) series, respectively experiments with strain 2002 (Figure S5). Both As(V) and (Table S1). Similar trends were observed for As removal by ffi biogenic ferrihydrite produced by Pseudomonas sp. strain GE- As(III) were e ciently removed from solutions, possibly either 17 through adsorption onto and/or coprecipitation with biogenic 1. Biogenic Fe(III) minerals formed from Fe(II) oxidation by Fe(III) minerals (Figure 3). In the present study, the maximum Acidovorax sp. BoFeN1 similarly showed higher As(III) immobilization than As(V) in both high and low As series.16 molar ratios of Asimmobilized/Feprecipitated in biogenic lepidocrocite, ff Furthermore, chemically synthesized lepidocrocite also showed calculated by di erences in the concentrations of soluble As and 38,40 Fe, before and after Fe(II) oxidation, achieved 0.063 ± 0.003 preferential incorporation of As(III) relative to As(V). Arsenic removal from solution depended on initial Fe/As and 0.049 ± 0.003 for fixed Fe/As series with initial Fe/As(III) ratio. In Fe/fixed As(III) series when initial Fe/As ratio = 37.5 and initial Fe/As(V) = 37.5, respectively, which are increased from 37.5 to 375 (with an increase of initial Fe(II) higher than those in chemically synthesized lepidocrocite μ 38 from 0.5 to 5 mM and an initial As concentration of 13.3 M), (0.061 and 0.022 for As(III) and As(V), respectively). The the amount of As removal increased from 12.55 ± 0.03 to 13.32 higher As/Fe in solids would be associated with coprecipitation ± 0.02 μM(Table S1). When initial Fe/As ratio increased to of As with biogenic Fe(III) minerals in this study. Jia and >75, aqueous As was removed to values below 0.0133 μM (the Demopoulos suggested that more As(V) was removed by current WHO drinking water limit), regardless of the initial As coprecipitation than adsorption, possibly due to the max- concentrations. The results are in good agreement with imization of the coordination sites during the neutralization of previous studies that showed nearly complete As removal of − 39 the acidic As(V) Fe(III) solution. Additionally, compared to up to 50 μM by biogenic Fe(III) minerals produced by strain synthetic abiogenic Fe(III) minerals, biogenic Fe(III) minerals BoFeN1 when Fe/As was >100.16,41 are mostly characterized by cell−mineral aggregates (Figure 2), The effect of initial Fe(II) concentration on As removal also which form by adsorption of Fe(III) minerals on negatively was evident by comparing the two experimental series with charged cell surfaces.23 These cell−mineral aggregates possibly differing initial Fe/As ratios. Greater As immobilization was contain cell−Fe/As complexes when formed in the presence of achieved in fixed Fe/As series than in Fe/fixed As series, As(III) and As(V), which contributes to greater As especially in experiments with lower initial Fe/As ratios (Table 16 immobilization. S1). Molar ratios of Asimmobilized/Feprecipitated were consistent with

6452 DOI: 10.1021/acs.est.6b00562 Environ. Sci. Technol. 2016, 50, 6449−6458 Environmental Science & Technology Article

Figure 3. Variations in As concentrations in the liquid phase as a function of incubation time under different initial Fe/(As(III) or As(V)) in the presence of strain 2002. For the Fe/fixed As series, we fixed the initial As(III) or As(V) concentration (13.3 μM) but varied initial concentrations of Fe(II) (between 0.5 and 5 mM for As(III)-treated series and between 0.5 and 10 mM for As(V)-treated series): (a) Fe/fixed As(III) series and (b) Fe/fixed As(V) series. For the fixed Fe/As series, we fixed the initial Fe(II) concentration (5 and 10 mM for As(III) and As(V)-treated series, respectively) but varied the initial concentrations of As(III) or As(V) (between 13.3 and 133.3 μM for As(III)-treated series and between 6.67 and 266.7 μM for As(V)-treated series): (c) fixed Fe/As(III) series and (d) fixed Fe/As(V) series. The percentage of Fe(II) oxidation at the end of incubation is given. this observation. Samples in the fixed Fe/As series had crystalline biogenic solid formed in all samples (Figure S6). ± ± relatively higher Asimmobilized/Feprecipitated ratios (i.e., 0.063 Cell lattice parameters of the As free samples (a = 3.88 0.01 0.003 and 0.049 ± 0.003 for initial Fe/As(III) = 37.5 and initial Å; b = 12.56 ± 0.02 Å; c = 3.08 ± 0.01 Å), As(III)-treated Fe/As(V) = 37.5, respectively) than those in the Fe/fixed As samples (a = 3.90 ± 0.02 Å; b = 12.60 ± 0.03 Å; c = 3.05 ± series (i.e., 0.032 ± 0.002 and 0.029 ± 0.002 for initial Fe/ 0.03 Å), and As(V)-treated samples (a = 3.89 ± 0.02 Å; b = As(III) = 37.5 and initial Fe/As(V) = 37.5, respectively) (Table 12.55 ± 0.03 Å; c = 3.10 ± 0.02 Å), obtained by the Cell S1). The likely reason was that the solutions in the fixed Fe/As Refinement using Aman(63) in Jade 6.5, were similar to series had relatively higher initial Fe(II) concentrations than synthetic lepidocrocite (a = 3.87 Å; b = 12.51 Å; c = 3.06 Å).43 those in the Fe/fixed As series. Biological Fe(II) oxidation by With ferrous chloride as the Fe(II) source, lepidocrocite was strain 2002 was nitrate dependent with a molar ratio of 0.47 for expected to preferentially form at neutral pH.44 However, reduced nitrate to oxidized Fe(II), which was close to the crystallinity decreased with an increase in As relative to Fe in theoretical stoichiometry (0.5).32 In our present study, nitrate this study (Figure S6), which mainly resulted from the was in excess relative to Fe(II) in either fixed Fe/As series or influences of As-complexes on the crystal growth of Fe/fixed As series. Thus, the rate of biological Fe(II) oxidation lepidocrocite. should be limited by the Fe(II) concentration, resulting in Scanning electron micrographs revealed that the majority of higher rates in fixed Fe/As series than in Fe/fixed As series microbial cells in As-free samples was encrusted by biogenic (Table S1). The higher rate of Fe(II) oxidation would be lepidocrocite with a lath-like morphology (length: ∼400 nm; conducive to quicker formation of biogenic lepicocrocite with width: ∼100 nm) (Figure 2o). However, in the As-bearing more active sites39,42 and therefore promote more As samples (either As(V) or As(III)), both lath-like and irregularly immobilization. This rate-dependent As immobilization has shaped mass or sheet-like structures were observed (Figure 2), been observed previously. For example, the overall faster rate of indicating the effects of As-complexes on the morphology of Fe(II) oxidation by Acidovorax sp. strain BoFeN1 relative to the biogenic lepidocrocite. It was reported that As(III) or strain KS (29.1 and 15.1 μM·h−1, respectively) resulted in an As(V) complexation led to the irregular morphologies and a 45,46 increased molar ratio of Asimmobilized/Feprecipitated from 0.002 to decrease in the mean size of magnetite particles. Moreover, 0.003 with excess nitrate.16 The initial As concentration showed high As concentrations may reduce the enzymatic activity of a slight effect on As removal. For instance, the molar ratio of strain 2002 cells and thus slow down the nucleation and immobilized As to initial As slightly decreased from 99.8 ± crystallization process of biogenic lepidocrocite. Refait and co- 0.1% (0.013 μM remaining) to 99.4 ± 0.1% (0.676 μM workers found that lepidocrocite, the product of abiotic remaining) with an increasing initial As(III) concentration from oxidation of aerated aqueous suspensions of Fe(II) oxy- 13.3 to 133.3 μMinfixed Fe/As(III) series (with initial Fe(II) in the absence of arsenate, was replaced by 6-line concentration of 5 mM) (Table S1). ferrihydrite with increasing As(V) concentration.47 Similarly, Mineralogy of Biogenic Fe-Bearing Mineral Phases. ferrihydrite was formed during the oxidation of Fe(II) by The μ-XRD analysis indicated that lepidocrocite was the only BoFeN1 at low Fe/As ratios (Fe/As = 15−50), and its relative

6453 DOI: 10.1021/acs.est.6b00562 Environ. Sci. Technol. 2016, 50, 6449−6458 Environmental Science & Technology Article

Figure 4. As K-edge k3-weighted EXAFS data of solid samples from both fixed Fe/As series and Fe/fixed As series (the left panel) as well as their Fourier Transform in the R range between 0.4 and 4.0 Å (the middle panel) and in the R range between 2.0 and 4.0 Å (right panel). Data were fit using a classical shell-by-shell analysis in ARTEMIS, including multiple scattering within the first oxygen shell. For each sample, the experimental data and fitted curves are displayed as dotted and solid lines, respectively. Fitting results are reported in Table S2. abundance systematically increased with decreasing Fe/As Å in the AsO4 tetrahedral molecule, which is consistent with the ratio.28 The biogenic goethite in the presence of Fe(II)- As−O distance for As(V) adsorption on goethite, siderite, and oxidizing bacteria Acidovorax sp. strain BoFeN1 tended to lose magnetite49 and lepidocrocite and magnetite.37,48 The first its characteristic acicular morphology with a decreased particle coordination shell surrounding As in the As(III)-treated size due to the presence of As.16 It was previously suggested samples was fit with 3.4 oxygen atoms at an As−O distance that As(V) and As(III) would restrain crystal growth during of 1.76−1.78 Å (Figure 4 and Table S2), consistent with 1.79 Å coprecipitation and a delayed crystal growth would decrease in the AsO3 pyramidal molecule and in good agreement with particle size. Additionally, our data showed that As speciation previous studies.27,37,48,49 also exerted an important effect on the morphology of biogenic The second neighbor contributions to the EXAFS data were lepidocrocite. In comparison with the As(V)-treated series, fit using As−Fe pairs at various distances and a multiple more irregularly shaped mass or sheet structures were observed scattering (MS) contribution corresponding to the As−O−O− in biogenic solids in As(III)-treated series (i.e., Fe/fixed series As path. For samples in Fe/fixed As(III) series, an As−Fe with Fe(II)/(As(III) or As(V)) = 37.5, in Figure 2a,e), possibly distance of 3.30−3.49 Å dominated the second-neighbor resulting from the higher affinity of As(III) to biogenic contribution, corresponding to bidentate binuclear corner- lepidocrocite than that of As(V). Higher As(III) loadings sharing complexes (2C). Such 2C complexes with similar As−Fe implied more As complexation, which would result in greater distances were found in As(III)/biogenic Fe minerals28 and impact on the morphologies of biogenic lepidocrocite. As(III)/abiogenic Fe minerals.6,37,50,51 For samples in fixed Fe/ EXAFS Analysis of As Immobilization Mechanisms. As(III) series with an initial Fe/As(III) ratio of 375, an As−Fe Arsenic K-edge EXAFS data exhibited weak second-neighbor distance of 3.30 Å also dominated the second-neighbor contributions for all As-containing biogenic samples, indicating contribution. However, that was not the case for initial Fe/ that As was not incorporated in the structure of any crystalline As(III) ratios between 75 and 37.5, where the second-neighbor phase but rather formed inner-sphere surface complexes on the contributions were best fit with a combination of two Fe shells surfaces (Figure 4). The theoretical fits of EXAFS spectra are at distances of 2.94−2.97 and 3.40−3.45 Å. The As−Fe shown in Figure 4, and the fitted parameters are given in Table distances of 2.94−2.97 and 3.40−3.45 Å corresponded to S2. The first coordination shell surrounding As in the As(V)- bidentate mononuclear edge-sharing complexes (2E) and 2C treated samples had 4.0 oxygen atoms at an As−O distance of complexes, respectively. The 2E complexes with an As−Fe pair 1.69 Å (Figure 4 and Table S2), corresponding closely to 1.69 at 2.97 ± 0.05 Å have been proposed in the case of As(III)/

6454 DOI: 10.1021/acs.est.6b00562 Environ. Sci. Technol. 2016, 50, 6449−6458 Environmental Science & Technology Article lepidocrocite,51 As(III)/ferrihydrite,37 and As(III)/biogenic Fe conditions.27,28,52 Consequently, we suggest that the coex- minerals.28 Moreover, the combination of 2C complexes and 2E istence of 2C corner-sharing complexes and 2E edge-sharing or complexes was suitable for the description of As(III)/ 1V corner-sharing complexes played an important role in higher 37,51 lepidocrocite EXAFS spectra. Ona-Nguema and co-workers As removal in higher initial Fe(II) series with the same Fe/As 2 2 found that C complexes at 3.40 Å and E complexes at 2.92 Å ratio. can describe the EXAFS spectra of As(III) adsorption on 37 ferrihydrite. ■ ENVIRONMENTAL IMPLICATIONS With regard to samples in the Fe/fixed As(V) series, a single As−Fe distance of 3.23−3.27 Å was observed, while in samples High As groundwater often contains elevated Fe(II) concen- from fixed Fe/As(V) series, a single As−Fe distance of 3.25− tration as a result of microbial Fe(III) reduction,1 which can 3.27 Å was observed for high initial Fe/As(V) ratios (750 and provide a soluble Fe(II) source for in situ remediation by 1500), and the coexistence of two As−Fe distances of 3.34 and Fe(II)-oxidizing bacteria. In reducing high As groundwater, 3.56 Å was obtained for a low initial Fe/As(V) ratio (75). The Fe(II) usually accounts for >70% of total soluble Fe As−Fe distance of 3.23−3.27 Å in samples from both fixed Fe/ concentrations, which is up to 29 mg/L in Bangladesh, 13.7 As and Fe/fixed As series was similar to the As−Fe distance in mg/L in West Bengal, 44.3 mg/L in Vietnam, 26.5 mg/L in 27,51,54 52 As(V)/lepidocrocite, As(V)/ferrihydrite, As(V)/goe- Cambodia, 5.9 mg/L in Inner Mongolia,58 and 6.8 mg/L in the 53 27 2 thite, and As(V)/ hematite, indicating the presence of C Yinchuan basin.59 In addition, nitrate and As cocontamination − − complexes. The As Fe distance of 3.57 3.63 Å indicated in groundwater commonly occurs in many areas around the 1 − monodentate mononuclear corner-sharing complexes ( V), world.60 64 Anaerobic nitrate-reducing Fe(II) oxidizing bacteria which was also used to describe the cases of As(V)/ 37,53 (AN-FOB) have the ability to oxidize Fe(II) using nitrate as lepidocrocite and As(V)/goethite, As(V)/ electron acceptor under anoxic conditions and produce various (anoxic),55 and As(V)/ .56 2 − Fe(III) oxide minerals, which coprecipitate or adsorb soluble Bidentate binuclear complexes ( C) were the only As Fe As.65,66 Anaerobic nitrate-reducing Fe(II) oxidizing bacteria, for complexes in samples from the Fe/fixed As series. The number example, Citrobacter freundii strain PXL1, Paracoccus ferroox- of Fe atoms coordinated with each As (N values) for the idans strain BDN-1, Pseudomonas sp. SZF15, and Candidatus bidentate binuclear complexes decreased from 1.5 to 0.5 in the − accumulibacterphosphatis TR1,22,67 69 may be expected to be case of As(III) and from 2.0 to 1.3 in the case of As(V) with an ff increase in initial Fe/As ratio from 75 to 375 (Table S2). e ective for cleaning up water contaminated with nitrate and Correspondingly, with an increase in initial Fe/As ratio from 75 As. It was reported that all respiratory nitrate-reducing bacteria are innately capable of triggering nitrate-dependent Fe(II) to 375, the molar ratios of Asimmobilized/Feprecipitated in biogenic 70 lepidocrocite decreased from 0.017 to 0.004 in the Fe/fixed oxidation, and therefore, more nitrate-reducing Fe(II) As(III) series and from 0.013 to 0.003 in the Fe/fixed As(V) oxidizing bacteria will be found and applied to the treatment series (Table S1).The results suggest that more Fe atoms are of nitrate and As cocontaminated water. coordinated with each As for the 2C complexes when associated As an AN-FOB isolate, strain 2002 grows with a high As resistance and high removal efficiency for both As(V) and with higher Asimmobilized/Feprecipitated values in lower Fe/As ratio series. However, due to the high uncertainties of N values As(III). In the case of the initial Fe/As ratios of >75, As was (±0.5), we cannot conclude that more Fe atoms coordinated efficiently removed to values below 0.0133 μM (the current with each As explain the apparent higher As removal by WHO drinking water limit), which are comparable to or better lepidocrocite. Indeed, more Fe atoms were found to be than other nitrate-reducing Fe(II)-oxidizers.16,65,67 The low coordinated with each As in As(III)-ferrihydrite37 and As(V)- initial Fe/As ratio required to lower As concentrations <0.0133 ferrihydrite27 with a higher molar ratio of As/Fe in the solid μM is of importance for As immobilization. For As- phase. contaminated water with an Fe/As ratio >75, this strain can However, for samples from the fixed Fe/As series, both 2E be used to effectively remove As down to <0.0133 μM. When and 2C complexes and both 1V and 2C complexes occurred in the Fe/As ratio is low (<75), less Fe(II) needs to be added for low initial Fe/As ratios with high F(II) concentrations. Surface effective As immobilization induced by the strain 2002 relative 2 complexation modeling of As(V) showed that the C complexes to the other strains. 2 1 were more stable than the E complexes and V complexes for As(III) is the major As species in groundwater under 27 57 As(V)/ferrihydrite at pH 7.0 and As(V)/gibbsite at pH 5.5. reducing conditions, which has been found to account for more 2 1 The instability of E complexes and V complexes could lead to than 60% of total As.63,71 This study showed that strain 2002 their transformation to 2C complexes in the solids with low 2 induced higher removal of As(III) from aqueous solution Asimmobilized/Feprecipitated. However, the transformation of E compared with As(V). Therefore, strain 2002, as a nitrate complexes and 1V complexes to 2C complexes was likely ff 2 reducing Fe(II)-oxidizer, could represent a cost-e ective and inhibited due to the limited Fe atoms for C complexes with an environmentally friendly technology for in situ remediation of increase in the As /Fe (reaching 0.063, 0.021, immobilized precipitated As and nitrate cocontaminated groundwater. and 0.049 in solid samples from fixed Fe/As series with initial Fe/As(III) ratios of 37.5 and 75 and initial Fe/As(V) ratios of ■ ASSOCIATED CONTENT 37.5, respectively), although these high Asimmobilized/Feprecipitated ratios would result from the formation of weakly crystalline *S Supporting Information As−Fe precipitates and the location of scavenged As partly 2 The Supporting Information is available free of charge on the inside the precipitates. In comparison with C complexes, fewer ACS Publications website at DOI: 10.1021/acs.est.6b00562. Fe atoms are needed to coordinate with each As atom for 2E and 1V complexes, and thus, more As could be coordinated to Detailed descriptions of water sample analysis, additional thesamenumberofFeatomsundernearlyidentical figures and tables (PDF)

6455 DOI: 10.1021/acs.est.6b00562 Environ. Sci. Technol. 2016, 50, 6449−6458 Environmental Science & Technology Article ■ AUTHOR INFORMATION (13) Chaudhuri, S. K.; Lack, J. G.; Coates, J. D. Biogenic magnetite formation through anaerobic biooxidation of Fe(II). Appl. Environ. Corresponding Author Microbiol. 2001, 67 (6), 2844−2848. *Tel.: +86-10-8232-1366; fax: +86-10-8232-1081; e-mail: (14) Kappler, A.; Newman, D. K. Formation of Fe(III)-minerals by [email protected]. Fe(II)-oxidizing photoautotrophic bacteria. Geochim. Cosmochim. Acta 2004 − Notes , 68 (6), 1217 1226. (15) Kappler, A.; Straub, K. L. Geomicrobiological cycling of iron. fi The authors declare no competing nancial interest. Rev. Mineral. Geochem. 2005, 59 (1), 85−108. (16) Hohmann, C.; Winkler, E.; Morin, G.; Kappler, A. Anaerobic ■ ACKNOWLEDGMENTS Fe(II)-oxidizing bacteria show as resistance and immobilize As during Fe(III) mineral precipitation. Environ. Sci. Technol. 2010, 44 (1), 94− The study is financially supported by the National Natural 101. Science Foundation of China (Nos. 41222020 and 41172224), (17) Xiu, W.; Guo, H.; Liu, Q.; Liu, Z.; Zou, Y. E.; Zhang, B. Arsenic the National Key Basic Research Development Program (973 removal and transformation by Pseudomonas sp. strain GE-1-Induced Program, No. 2010CB428804), the Fundamental Research ferrihydrite: co-precipitation versus adsorption. Water, Air, Soil Pollut. Funds for the Central Universities (No. 2652013028), and the 2015, 226,1−14. Fok Ying-Tung Education Foundation, China (Grant No. (18) Ona-Nguema, G.; Abdelmoula, M.; Jorand, F.; Benali, O.; Gehin,́ A.; Block, J. C.; Genin,́ J. M. R. Microbial reduction of 131017). The authors would like to thank the Shanghai γ Synchrotron Radiation Facility (Beamlines BL14W1 and lepidocrocite -FeOOH by Shewanella putrefaciens; the formation of green rust. Hyperfine Interact. 2002, 139−140 (1), 231−237. BL15U1) and its staff (Z. Jiang, S. Zhang, X. Yu, and A. Li) μ (19) Miot, J.; Benzerara, K.; Morin, G.; Kappler, A.; Bernard, S.; for allowing us to perform the EXAFS and -XRD analysis. Obst, M.; Ferard,́ C.; Skouri-Panet, F.; Guigner, J. M.; Posth, N.; et al. Courtesy review of English by Dr. Richard Wanty (USGS) and Iron biomineralization by anaerobic neutrophilic iron-oxidizing Dr. Michael Kersten (Mainz University) is specially acknowl- bacteria. Geochim. Cosmochim. Acta 2009, 73 (3), 696−711. edged. We are grateful to the Associate Editor (Dr. Daniel (20) Pantke, C.; Obst, M.; Benzerara, K.; Morin, G.; Ona-Nguema, Giammar) and three anonymous reviewers, whose comments G.; Dippon, U.; Kappler, A. Green rust formation during Fe(II) significantly improved the quality of the manuscript. oxidation by the nitrate-reducing Acidovorax sp. strain BoFeN1. Environ. Sci. Technol. 2012, 46 (3), 1439−1446. (21) Katsoyiannis, I. A.; Zouboulis, A. I. Application of biological ■ REFERENCES processes for the removal of arsenic from ground-waters. Water Res. (1) Smedley, P. L.; Kinniburgh, D. G. A review of the source, 2004, 38 (1), 17−26. behaviour and distribution of arsenic in natural waters. Appl. Geochem. (22) Li, B.; Pan, X.; Zhang, D.; Lee, D.-J.; Al-Misned, F. A.; Mortuza, 2002, 17 (5), 517−568. M. G. Anaerobic nitrate reduction with oxidation of Fe(II) by (2) Harvey, C. F.; Swartz, C. H.; Badruzzaman, A. B. M.; Keon-Blute, Citrobacter Freundii strain PXL1 − a potential candidate for N.; Yu, W.; Ali, M. A.; Jay, J.; Beckie, R.; Niedan, V.; Brabander, D.; simultaneous removal of As and nitrate from groundwater. Ecol. Eng. Oates, P. M.; Ashfaque, K. N.; Islam, S.; Hemond, H. F.; Ahmed, M. F. 2015, 77, 196−201. Arsenic mobility and groundwater extraction in Bangladesh. Science (23) Schadler,̈ S.; Burkhardt, C.; Hegler, F.; Straub, K. L.; Miot, J.; 2002, 298 (5598), 1602−1606. Benzerara, K.; Kappler, A. Formation of cell-iron-mineral aggregates by (3) Islam, F. S.; Gault, A. G.; Boothman, C.; Polya, D. A.; Charnock, phototrophic and nitrate-reducing anaerobic Fe(II)-oxidizing bacteria. J. M.; Chatterjee, D.; Lloyd, J. R. Role of metal-reducing bacteria in Geomicrobiol. J. 2009, 26 (2), 93−103. arsenic release from Bengal delta sediments. Nature 2004, 430 (6995), (24) Hao, L.; Guo, Y.; Byrne, J. M.; Zeitvogel, F.; Schmid, G.; Ingino, 68−71. P.; Li, J.; Neu, T. R.; Swanner, E. D.; Kappler, A.; Obst, M. Binding of (4) Morin, G.; Calas, G. Arsenic in soils, mine tailings, and former heavy metal ions in aggregates of microbial cells, EPS and biogenic industrial sites. Elements 2006, 2 (2), 97−101. iron minerals measured in-situ using metal- and glycoconjugates- (5) Bhattacharya, P.; Welch, A. H.; Stollenwerk, K. G.; McLaughlin, specific fluorophores. Geochim. Cosmochim. Acta 2016, 180,66−96. M. J.; Bundschuh, J.; Panaullah, G. Arsenic in the environment: (25) Miot, J.; Benzerara, K.; Morin, G.; Bernard, S.; Beyssac, O.; biology and chemistry. Sci. Total Environ. 2007, 379 (2−3), 109−120. Larquet, E.; Kappler, A.; Guyot, F. Transformation of vivianite by (6) Guo, H.; Ren, Y.; Liu, Q.; Zhao, K.; Li, Y. Enhancement of anaerobic nitrate-reducing iron-oxidizing bacteria. Geobiology 2009, 7 arsenic adsorption during mineral transformation from siderite to (3), 373−384. goethite: mechanism and application. Environ. Sci. Technol. 2013, 47 (26) Miot, J.; Benzerara, K.; Obst, M.; Kappler, A.; Hegler, F.; (2), 1009−1016. Schadler, S.; Bouchez, C.; Guyot, F.; Morin, G. Extracellular iron (7) Alam, M. S.; Wu, Y.; Cheng, T. Silicate minerals as a source of biomineralization by photoautotrophic iron-oxidizing bacteria. Appl. arsenic contamination in groundwater. Water, Air, Soil Pollut. 2014, Environ. Microbiol. 2009, 75 (17), 5586−91. 225 (11), 1−15. (27) Sherman, D. M.; Randall, S. R. Surface complexation of (8) Wu, Y.; Li, W.; Sparks, D. L. Effect of iron(II) on arsenic arsenic(V) to iron(III) (hydr)oxides: structural mechanism from ab δ sequestration by -MnO2: desorption studies using stirred-flow initio molecular geometries and EXAFS spectroscopy. Geochim. experiments and X-ray absorption fine structure spectroscopy. Environ. Cosmochim. Acta 2003, 67 (22), 4223−4230. Sci. Technol. 2015, 49 (22), 13360−13368. (28) Hohmann, C.; Morin, G.; Ona-Nguema, G.; Guigner, J.-M.; (9) Jain, C. K.; Ali, I. Arsenic: occurrence, toxicity and speciation Brown, G. E., Jr; Kappler, A. Molecular-level modes of As binding to techniques. Water Res. 2000, 34 (17), 4304−4312. Fe(III) (oxyhydr)oxides precipitated by the anaerobic nitrate-reducing (10) Dixit, S.; Hering, J. G. Comparison of arsenic(V) and Fe(II)-oxidizing Acidovorax sp. strain BoFeN1. Geochim. Cosmochim. arsenic(III) sorption onto iron oxide minerals: Implications for arsenic Acta 2011, 75 (17), 4699−4712. mobility. Environ. Sci. Technol. 2003, 37 (18), 4182−4189. (29) Wang, L.; Giammar, D. E. Effects of pH, dissolved oxygen, and (11) Catalano, J. G.; Park, C.; Fenter, P.; Zhang, Z. Simultaneous aqueous ferrous iron on the adsorption of arsenic to lepidocrocite. J. inner- and outer-sphere arsenate adsorption on corundum and Colloid Interface Sci. 2015, 448, 331−338. hematite. Geochim. Cosmochim. Acta 2008, 72 (8), 1986−2004. (30) Zhao, L.; Dong, H.; Kukkadapu, R.; Agrawal, A.; Liu, D.; Zhang, (12) Mohan, D.; Pittman, C. U. Arsenic removal from water/ J.; Edelmann, R. E. Biological oxidation of Fe(II) in reduced wastewater using adsorbents–A critical review. J. Hazard. Mater. 2007, nontronite coupled with nitrate reduction by Pseudogulbenkiania sp. 142 (1−2), 1−53. Strain 2002. Geochim. Cosmochim. Acta 2013, 119, 231−247.

6456 DOI: 10.1021/acs.est.6b00562 Environ. Sci. Technol. 2016, 50, 6449−6458 Environmental Science & Technology Article

(31) Zhao, L.; Dong, H.; Kukkadapu, R. K.; Zeng, Q.; Edelmann, R. (49) Jönsson, J.; Sherman, D. M. Sorption of As(III) and As(V) to E.; Pentrak,́ M.; Agrawal, A. Biological redox cycling of iron in siderite, green rust (fougerite) and magnetite: Implications for arsenic nontronite and its potential application in nitrate removal. Environ. Sci. release in anoxic groundwaters. Chem. Geol. 2008, 255 (1), 173−181. Technol. 2015, 49 (9), 5493−5501. (50) Manning, B. A.; Fendorf, S. E.; Goldberg, S. Surface structures (32) Weber, K. A.; Pollock, J.; Cole, K. A.; O’Connor, S. M.; and stability of arsenic(III) on goethite: spectroscopic evidence for Achenbach, L. A.; Coates, J. D. Anaerobic nitrate-dependent iron(II) inner-sphere complexes. Environ. Sci. Technol. 1998, 32 (16), 2383− bio-oxidation by a novel lithoautotrophic betaproteobacterium, strain 2388. 2002. Appl. Environ. Microbiol. 2006, 72 (1), 686−694. (51) Farquhar, M. L.; Charnock, J. M.; Livens, F. R.; Vaughan, D. J. (33) Klueglein, N.; Kappler, A. Abiotic oxidation of Fe(II) by reactive Mechanisms of arsenic uptake from aqueous solution by interaction nitrogen species in cultures of the nitrate - reducing Fe(II) oxidizer with goethite, lepidocrocite, mackinawite, and pyrite: an X-ray Acidovorax sp. BoFeN1 - questioning the existence of enzymatic Fe(II) absorption spectroscopy study. Environ. Sci. Technol. 2002, 36 (8), oxidation. Geobiology 2013, 11 (2), 180−190. 1757−1762. (34) Ravel, B.; Newville, M. ATHENA, ARTEMIS, HEPHAESTUS: (52) Waychunas, G. A.; Rea, B. A.; Fuller, C. C.; Davis, J. A. Surface data analysis for X-ray absorption spectroscopy using IFEFFIT. J. chemistry of ferrihydrite: Part 1. EXAFS studies of the geometry of Synchrotron Radiat. 2005, 12, 537−541. coprecipitated and adsorbed arsenate. Geochim. Cosmochim. Acta 1993, (35) Hughes, M. F. Arsenic toxicity and potential mechanisms of 57, 2251−2269. action. Toxicol. Lett. 2002, 133 (1), 1−16. (53) Fendorf, S.; Eick, M. J.; Grossl, P.; Sparks, D. L. Arsenate and (36) Weber, K. A.; Hedrick, D. B.; Peacock, A. D.; Thrash, J. C.; chromate retention mechanisms on goethite. 1. surface structure. − White, D. C.; Achenbach, L. A.; Coates, J. D. Physiological and Environ. Sci. Technol. 1997, 31 (2), 315 320. taxonomic description of the novel autotrophic, metal oxidizing (54) Randall, S. R.; Sherman, D. M.; Ragnarsdottir, K. V. Sorption of · bacterium, Pseudogulbenkiania sp. strain 2002. Appl. Microbiol. As(V) on green rust (Fe4(II)Fe2(III) (OH)12SO4 3H2O) and γ Biotechnol. 2009, 83 (3), 555−565. lepidocrocite ( -FeOOH): surface complexes from EXAFS spectros- − (37) Ona-Nguema, G.; Morin, G.; Juillot, F.; Calas, G.; Brown, G. E., copy. Geochim. Cosmochim. Acta 2001, 65 (7), 1015 1023. Jr. EXAFS Analysis of arsenite adsorption onto two-line ferrihydrite, (55) Wang, Y.; Morin, G.; Ona-Nguema, G.; Juillot, F.; Guyot, F.; hematite, goethite, and lepidocrocite: influence of surface structure. Calas, G.; Brown, G. E. Evidence for different surface speciation of − arsenite and arsenate on green rust: an EXAFS and XANES study. Environ. Sci. Technol. 2005, 39 (23), 9147 9155. − (38) Wan, J.; Simon, S.; Deluchat, V.; Dictor, M.-C.; Dagot, C. Environ. Sci. Technol. 2010, 44 (1), 109 15. Adsorption of As(III), As(V) and dimethylarsinic acid onto (56) Morin, G.; Ona-Nguema, G.; Wang, Y.; Menguy, N.; Juillot, F.; synthesized lepidocrocite. J. Environ. Sci. Health, Part A: Toxic/Hazard. Proux, O.; Guyot, F.; Calas, G.; Brown, G. E., Jr. Extended X-ray − absorption fine structure analysis of arsenite and arsenate adsorption Subst. Environ. Eng. 2013, 48 (10), 1272 1279. − (39) Jia, Y. F.; Demopoulos, G. P. Adsorption of arsenate onto on maghemite. Environ. Sci. Technol. 2008, 42 (7), 2361 6. (57) Ladeira, A. C. Q.; Ciminelli, V. S. T.; Duarte, H. A.; Alves, M. C. ferrihydrite from aqueous solution: Influence of media (sulfate vs M.; Ramos, A. Y. Mechanism of anion retention from EXAFS and nitrate), added gypsum, and pH alteration. Environ. Sci. Technol. 2005, density functional calculations: arsenic (V) adsorbed on gibbsite. 39 (24), 9523−9527. Geochim. Cosmochim. Acta 2001, 65 (8), 1211−1217. (40) Repo, E.; Makinen,̈ M.; Rengaraj, S.; Natarajan, G.; Bhatnagar, (58) Guo, H.; Liu, C.; Lu, H.; Wanty, R. B.; Wang, J.; Zhou, Y. A.; Sillanpaä,̈ M. Lepidocrocite and its heat-treated forms as effective Pathways of coupled arsenic and iron cycling in high arsenic arsenic adsorbents in aqueous medium. Chem. Eng. J. 2012, 180 (3), groundwater of the Hetao basin, Inner Mongolia, China: An iron 159−169. − ̈ isotope approach. Geochim. Cosmochim. Acta 2013, 112, 130 145. (41) Berg, M.; Luzi, S.; Trang, P. T.; Viet, P. H.; Giger, W.; Stuben, (59) Guo, Q.; Guo, H.; Yang, Y.; Han, S.; Zhang, F. Hydro- D. Arsenic removal from groundwater by household sand filters: geochemical contrasts between low and high arsenic groundwater and comparative field study, model calculations, and health benefits. − its implications for arsenic mobilization in shallow aquifers of the Environ. Sci. Technol. 2006, 40 (17), 5567 5573. northern Yinchuan Basin, P.R. China. J. Hydrol. 2014, 518, 464−476. (42) Rancourt, D. G.; Fortin, D.; Pichler, T.; Thibault, I. J.; (60) Fytianos, K.; Christophoridis, C. Nitrate, arsenic and chloride Lamarche, G.; Morris, R. V.; Mercier, P. H. J. Mineralogy of a natural pollution of drinking water in Northern Greece. Elaboration by As-rich hydrous ferric oxide coprecipitate formed by mixing of applying GIS. Environ. Monit. Assess. 2004, 93 (1−3), 55−67. hydrothermal fluid and seawater: Implications regarding surface (61) Glenn, S. M.; James Lester, L. An analysis of the relationship complexation and color banding in ferrihydrite deposits. Am. Mineral. between land use and arsenic, vanadium, nitrate and boron − − 2001, 86 (7 8), 834 851. contamination in the Gulf Coast aquifer of Texas. J. Hydrol. 2010, (43) Ewing, F. J. The of Lepidocrocite. J. Chem. 389 (1−2), 214−226. Phys. 1935, 3 (7), 420. (62) Hosono, T.; Nakano, T.; Shimizu, Y.; Onodera, S. I.; Taniguchi, (44) Cornell, R. M.; Schwertmann, U. The iron oxides: structure, M. Hydrogeological constraint on nitrate and arsenic contamination in properties, reactions, occurrences and uses; Wiley: New York, 2003. Asian metropolitan groundwater. Hydrol. Process. 2011, 25 (17), (45) Wang, Y.; Morin, G.; Ona-Nguema, G.; Menguy, N.; Juillot, F.; 2742−2754. Aubry, E.; Guyot, F.; Calas, G.; Brown, G. E., Jr Arsenite sorption at (63) Guo, H.; Wen, D.; Liu, Z.; Jia, Y.; Guo, Q. A review of high the magnetite−water interface during aqueous precipitation of arsenic groundwater in Mainland and Taiwan, China: Distribution, magnetite: EXAFS evidence for a new arsenite surface complex. characteristics and geochemical processes. Appl. Geochem. 2014, 41, Geochim. Cosmochim. Acta 2008, 72 (11), 2573−2586. 196−217. (46) Wang, Y.; Morin, G.; Ona-Nguema, G.; Juillot, F.; Calas, G.; (64) Li, P.; Wang, Y.; Jiang, Z.; Jiang, H.; Li, B.; Dong, H.; Wang, Y. Brown, G. E. Distinctive arsenic(V) trapping modes by magnetite Microbial diversity in high arsenic groundwater in Hetao Basin of nanoparticles induced by different sorption processes. Environ. Sci. Inner Mongolia, China. Geomicrobiol. J. 2013, 30 (10), 897−909. Technol. 2011, 45 (17), 7258−7266. (65) Upadhyaya, G.; Jackson, J.; Clancy, T. M.; Hyun, S. P.; Brown, (47) Refait, P.; Girault, P.; Jeannin, M.; Rose, J. Influence of arsenate J.; Hayes, K. F.; Raskin, L. Simultaneous removal of nitrate and arsenic species on the formation of Fe(III) oxyhydroxides and Fe(II−III) from drinking water sources utilizing a fixed-bed bioreactor system. hydroxychloride. Colloids Surf., A 2009, 332 (1), 26−35. Water Res. 2010, 44 (17), 4958−4969. (48) Manning, B. A.; Hunt, M. L.; Amrhein, C.; Yarmoff, J. A. (66) Li, B.; Tian, C.; Zhang, D.; Pan, X. Anaerobic nitrate-dependent Arsenic(III) and arsenic(V) reactions with zerovalent iron corrosion iron(II) oxidation by a novel autotrophic bacterium, Citrobacter products. Environ. Sci. Technol. 2002, 36 (24), 5455−5461. freundii strain PXL1. Geomicrobiol. J. 2014, 31 (2), 138−144.

6457 DOI: 10.1021/acs.est.6b00562 Environ. Sci. Technol. 2016, 50, 6449−6458 Environmental Science & Technology Article

(67) Kumaraswamy, R.; Sjollema, K.; Kuenen, G.; van Loosdrecht, M.; Muyzer, G. Nitrate-dependent [Fe(II)EDTA](2-) oxidation by Paracoccus ferrooxidans sp nov., isolated from a denitrifying bioreactor. Syst. Appl. Microbiol. 2006, 29 (4), 276−286. (68) Mattes, A.; Gould, D.; Taupp, M.; Glasauer, S. A novel autotrophic bacterium isolated from an engineered wetland system links nitrate-coupled iron oxidation to the removal of As, Zn and S. Water, Air, Soil Pollut. 2013, 224 (4), 1−15. (69) Su, J. F.; Shao, S. C.; Huang, T. L.; Ma, F.; Yang, S. F.; Zhou, Z. M.; Zheng, S. C. Anaerobic nitrate-dependent iron(II) oxidation by a novel autotrophic bacterium, Pseudomonas sp. SZF15. J. Environ. Chem. Eng. 2015, 3 (3), 2187−2193. (70) Carlson, H. K.; Clark, I. C.; Blazewicz, S. J.; Iavarone, A. T.; Coates, J. D. Fe(II) oxidation is an innate capability of nitrate-reducing bacteria that involves abiotic and biotic reactions. J. Bacteriol. 2013, 195 (14), 3260−3268. (71) Edmunds, W. M.; Ahmed, K. M.; Whitehead, P. G. A review of arsenic and its impacts in groundwater of the Ganges-Brahmaputra- Meghna delta, Bangladesh. Environ. Sci.: Processes Impacts 2015, 17, 1032−1046.

6458 DOI: 10.1021/acs.est.6b00562 Environ. Sci. Technol. 2016, 50, 6449−6458