- Email: [email protected]
Mobilization of arsenic on nano-TiO2 in soil columns with sulfate reducing bacteria
Environmental Pollution 234 (2018) 762e768
Contents lists available at ScienceDirect
Environmental Pollution journal homepage: www.elsevier.com/locate/envpol
Mobilization of arsenic on nano-TiO2 in soil columns with sulfate reducing bacteria* Ting Luo a, b, Li Ye a, c, Tingshan Chan d, **, Chuanyong Jing a, c, * a
State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China b School of Environmental Science and Engineering, Yancheng Institute of Technology, Jiangsu 224051, China c University of Chinese Academy of Sciences, Beijing 100049, China d National Synchrotron Radiation Research Center, 101 Hsin-Ann Road, Hsinchy Science Park, Hsinchu 30076, Taiwan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 18 October 2017 Received in revised form 6 December 2017 Accepted 8 December 2017
Arsenic (As) remediation in contaminated water using nanoparticles is promising. However, the fate and transport of As associated with nano-adsorbents in natural environment is poorly understood. To investigate the fate of adsorbed As on nano-TiO2 in changed redox condition from oxic to anoxic, we added the As(V)-TiO2 suspension in groundwater to an autoclaved soil column which inoculated a sulfate-reducing bacterium, Desulfovibrio vulgaris DP4. The dissolved As(V) in effluent increased to 798 mg/L for the biotic column and to 1510 mg/L for the abiotic control, and dissolved As(III) was observed only in biotic column. The total As (dissolved plus particulate) in the biotic column effluent (high to 2.5 mg/L) was substantially higher than the abiotic control (1.5 mg/L). Therefore SRB restrained the release of dissolved As, and facilitated the transport of particulate As. Micro-XRF analysis suggested that the nano-TiO2 with As was mainly retained in the influent front and that its transport was negligible. Our pe-pH calculation and XANES analysis demonstrated that generated secondary iron minerals containing magnetite and mackinawite mainly were responsible for dissolved As retention, and then transported with As as particulate As. The results shed light on the mobilization of adsorbed As on a nano-adsorbent in an anoxic environment. © 2017 Elsevier Ltd. All rights reserved.
Keywords: Arsenic mobilization Nano-TiO2 Sulfate reducing bacteria Soil column
1. Introduction Arsenic (As) is a redox active and toxic element that poses considerable human health risks. Recently, adsorption on nanoTiO2 has been demonstrated to be a promising technique for removing As from groundwater and industrial wastewater (Jing et al., 2009; Luo et al., 2010; Pena et al., 2006; Yan et al., 2015). However, the fate of the As, once it is removed from the water, is not well known. The pressing need to predict the fate and transport of As adsorbed onto nano-TiO2during environmental redox transition has motivated this study. The shift in the redox environment from oxidizing to reducing, as mediated by microbial activities, plays an important role in As *
This paper has been recommended for acceptance by B. Nowack. * Corresponding author. State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China. ** Corresponding author. E-mail addresses: [email protected] (T. Chan), [email protected] (C. Jing). https://doi.org/10.1016/j.envpol.2017.12.029 0269-7491/© 2017 Elsevier Ltd. All rights reserved.
biogeochemical cycling (Ye et al., 2017; Sun et al., 2016; Guo et al., 2015; Polizzotto et al., 2005). Sulfate-reducing bacteria (SRB) are ubiquitous in the anoxic subsurface, such as in groundwater systems (Burton et al., 2014; Canfield et al., 2005), and biogenic sulfide is an efficient reductant for As(V) and Fe(III) (Rochette et al., 2000; Poulton et al., 2004). Consequently, the reduced As(III) may form As-sulfide precipitates such as realgar (AsS) and orpiment (As2S3), which controls the dissolved As levels in the aquifer (Couture et al., 2014; Kumar et al., 2016a). Moreover, the formation of Fe(II)-sulfide minerals provides an additional sink for As immobilization (Lowers et al., 2007; Zhang et al., 2017). The effect of SRB on the mobility of As and the transformation of iron minerals has been extensively studied using ferrihydritecoated sand column experiments. Kocar et al. (2010) observed that SRB (Desulfovibrio vulgaris) inhibit As release from the sand column and induce the ferrihydrite transfer to magnetite (Fe3O4), green rust, and amorphous FeS. Burton et al. (2011) reported that SRB (Desulfovibrio vulgaris) enhance the As mobility in ferrihydriterich environments at pH 8, and conversely inhibit the As release at
T. Luo et al. / Environmental Pollution 234 (2018) 762e768
pH 6, due to the formation of goethite (FeOOH). In the presence of iron, As mobility may also be limited by its sorption to secondary Fe-minerals that formed in situ, including magnetite, pyrite (FeS2), greigite (Fe3S4), and Lepidocrocite (g-FeOOH) (Kumar et al., 2016b; Burton et al., 2011; Kirk et al., 2010). However, to the best of our knowledge, no publications report on the effects of SRB on the transport and potential risks of nanosized-TiO2-containing As in the natural environment. There is a lack of knowledge as to whether nano-TiO2-containing As ends up in soils or travels long distances in the subsurface, and how SRB affect this complicated process. It is critical to understand whether the current and future use of nano-TiO2 for As removal would trigger the release of As, and result in an even greater amount of As contamination due to the mobility of nanoparticles. Moreover, previous column experiments mainly adopt idealized model systems, such as sand or ferrihydrite-coated sand columns, which may not adequately represent the complex natural environment. The objectives of this research are to study the redox reactions and release of adsorbed As on nano-TiO2in the presence of SRB in soil columns. The soil column and effluent were monitored for As and Fe speciation and TiO2 transport using complimentary analytical techniques, including XANES and m-XRF, and were analyzed by thermodynamic calculations. This study provides the first evidence regarding the mobilization of adsorbed As on nano-TiO2 with SRB in natural soil under anoxic conditions. 2. Materials and methods 2.1. As(V) loading on nano-TiO2 The nano-TiO2 was prepared via hydrolysis of titanyl sulfate, as described in our previous work (Luo et al., 2010). Synthetic groundwater was prepared based on the composition of groundwater samples from naturally-occurring As sites in Shanxi, China (Luo et al., 2012): KNO3 0.02 mM, NaHCO3 6.39 mM, CaCO3 0.16 mM, MgCl2 0.87 mM, NH4Cl 0.02 mM, NaCl 2.94 mM, pH 8.0. The As(V) solution was prepared with Na2HAsO4$7H2O in a groundwater matrix. The adsorption of As(V) on TiO2 was conducted in a glovebox (100% N2). Phosphate, added as Na2HPO4, was first pre-adsorbed on the sterilized TiO2 at a loading of 750 mg/kg to facilitate microbial growth (Kocar et al., 2006). The As(V) solution was mixed with the TiO2-P solids for 24 h at pH 8.0. Then, the TiO2-As(V) suspension was centrifuged, and the TiO2-As solid was rinsed several times with deionized (DI) water. The final load of As(V) on TiO2 was approximately 390 mg/g by mass balance calculation. The TiO2-As(V) suspension with the TiO2 concentration of 0.1 g/L was prepared by dispersing the TiO2-As(V) solid into the groundwater matrix. The soluble As(V) concentration was 2.36 mg/L after the TiO2-As(V) solid was dispersed into the groundwater. The average particle size in the TiO2-As(V) suspension was 628 nm, as determined by a Zeta sizer Nano ZS (Malvern Instrument, UK). 2.2. Column experiments Soil samples were collected at 1 m depth at a site in Shanxi, China (GPS location 39 250 24.300 N, 112 520 50.000 E). The As content in soil was 8.6 mg/kg, which is within the background As level (5e10 mg/kg) of uncontaminated soils (Smedley and Kinniburgh, 2002). The components of the soil are shown in Table S1. The autoclaved soil was used for packing the columns. A sulfate-reducing bacterium identified as Desulfovibrio vulgaris DP4 was isolated from the soil (Luo et al., 2013). The SRB were grown to the late-exponential phase under anoxic conditions in a modified Baar's medium (Gherna et al., 1992) at 37 C. Cells were
763
harvested by centrifugation (6000 g, 8 min, 25 C), and then rinsed with synthetic groundwater. The SRB were subsequently added to the autoclaved soil and mixed in the glovebox. Then, 14 g of this mixture was packed into a glass column (length ¼ 10 cm, diameter ¼ 1.2 cm), resulting in a porosity of 0.54. The TiO2-As(V) suspension was delivered to the column in an upward flow using a peristaltic pump at three pore volumes (PV) per day (11 mL/min). Sulfate (MgSO4 5.5 mM) was added as an electron acceptor. The electron donor was lactate (Na-lactate10 mM). Two abiotic columns, without the addition of SRB, were also set up for comparison. The influent stream for one abiotic column was the TiO2-As(V) suspension, which was used as a control. Synthetic groundwater was used as background in the other column. The column experiments were conducted in the glovebox. Eh and pH in effluent samples were monitored over the experimental period of 51 d. A subsample of the effluent was passed through a 0.22 mm syringe filter for the analysis of dissolved As species, ferrous iron, sulfide, sulfate, and phosphate. The unfiltered effluent sample was used to analyze the total As, Ti, Fe, Ca, Na, Mg, Al, Si, and Mn. The solids were sampled for cell counts during the column experiments using an epiflourescence microscope with 40 -6-diamidino-2phenylindoledDAPIdstaining (Kocar et al., 2010). 2.3. Aqueous analysis Arsenic speciation was determined using high performance liquid chromatography atomic fluorescence spectrometry (HPLCAFS, Jitian, China). Ferrous iron and sulfide concentrations were measured using the colorimetric method Phenanthroline Method and Methylene Blue Method (Eaton et al., 1995), respectively, immediately following the sample collection. Sulfate and phosphate concentrations were determined using a DX-1100 ion chromatograph (Dionex, US) with an AS11-HC Ion Pac column. 2.4. Total metal concentrations in effluent To determine the total content of As, Ti, Fe, Ca, Na, Mg, Al, Si, and Mn in the unfiltered effluent, the samples were heated dry, followed by microwave digestion with 4 mL concentrated HNO3, 1 mL concentrated HF, and 5 mL of the sulfuric acid-ammonium sulfate solution (Zhang et al., 2007) using a microwave accelerated reaction system (CEM MARS, US). The digestion solution was then transferred to a 10 mL volumetric flask. The soil and solids were digested by the same method in a column incubated with Desulfovibrio vulgaris DP4. The total As, Ti, and Fe concentrations were determined by ICP-MS (Agilent, US). The total Mn, Al, Si, Mg, Ca, and Na concentrations were measured by ICP-OES (Perkin Elmer, US). 2.5. m-XRF and XANES analysis The solids in the bottom (influent front) and upper (end) layers of the column were collected at the end of the experiment and then freeze dried. The solid sample was deposited onto Kapton tape and analyzed by m-XRF at beamline 15U at Shanghai Synchrotron Radiation Facility (SSRF), China. m-XRF analysis was conducted with the monochromator set at 12 keV, a 10 10 mm beam, a 1-second dwell time per pixel, and a step size of 10 mm. The peak intensities for As, S, Fe, Mn, Ti, Zn, Ca, and Cu were collected at each pixel of m-XRF maps, which were 0.6 0.6 mm in size. The As and Fe K-edge XANES spectra were collected at beamline 01C1 at the National Synchrotron Radiation Research Center (NSRRC), Taiwan. An energy range of 150 to 300 eV from the K-edge of As (11867 eV) and Fe (7112 eV) was used to acquire the spectra with a 0.5 eV step size from 20 eV below to 50 eV above the absorption edge. The spectra were taken under standard NSRRC
764
T. Luo et al. / Environmental Pollution 234 (2018) 762e768
TOP-UP operation conditions (1.5 GeV and 240 mA) with a doublecrystal Si (111) monochromator. Fluorescence signals were collected using a Lytle detector, positioned at 90 to the incident beam. Spectra were acquired at cryogenic temperatures (77 K) using a helium cryostat to prevent the beam-induced oxidation of As(III) and Fe(II). Four scans were collected from each sample, inspected for overall quality, and averaged to improve the signal/noise ratio. As standard reference chemicals, Na2HAsO4$7H2O, NaAsO2, realgar (AsS), orpiment (As2S3), and arsenopyrite (FeAsS) were analyzed. Fe reference standards including magnetite (Fe3O4), mackinawite (FeS), FeAsS, FeS2, FeOOH, FeSO4, Fe2O3, and FeO were also measured. Fe reference standards were purchased in the National Standard Substances Center in China except mackinawite. Mackinawite was synthesized in lab according to a previous method (Wolthers et al., 2005). The XANES analysis procedure was described in our previous study (Luo et al., 2010). Briefly, the spectra were analyzed with the Athena program in the IFEFFIT computer package (Ravel and Newville, 2005) for a linear combination fit over the relative energy range of 20 to 30 eV. No energy shift was allowed during the fitting procedure. The weighting of single species was constrained between 0 and 1, and the sum of species was constrained to be 1. 3. Results and discussion 3.1. Arsenic release and reduction in soil columns Effluent As was monitored during the 51 d (153 PV) experiment. An experiment column was incubated with D. vulgaris DP4. There was also one abiotic control column and one background column which pumped groundwater only without TiO2-As(V) suspension (Fig. 1). The release of As can be divided into a lag phase (<40 PV, 13 d) and an acceleration phase (>40 PV). In the lag phase, dissolved effluent As(V) values were all below 100 mg/L, demonstrating the dissolved As(V) in the influent was adsorbed on soil minerals. The accumulative dissolved As(V) load in the lag phase was about 35 mg/kg, which was comparable to the As(V) adsorption capacity of the soil (32.5 mg/kg, Fig. S1). When the soil became saturated, As release accelerated, so that the dissolved As(V) in the effluent increased substantially to 798 mg/L for the biotic column and to 1510 mg/L for the control, at 153 PV. In contrast, only 11 mg/L As(V),
Fig. 1. Dissolved As(III) and As(V) in filtered samples from biotic column incubated with D. vulgaris DP4, an abiotic control with no D. vulgaris DP4, and a background column with artificial groundwater flow and no TiO2-As(V) suspension. The dashed line denoted the pore volume was 40 which displaced the soil adsorption for arsenic was saturated.
on average, were detected in the effluent of the background column, suggesting that the initial As in the soil (8.6 mg/kg) contributed little to the overall As release. Even the presence of SRB resulted in only an 8 mg/L As release from the soil to the groundwater matrix (Fig. S2). Therefore, the addition of the SRB restrained the release of dissolved As. Dissolved As(III) in the effluent was observed only from the biotic column after 3 PV, indicating that SRB induced the As(V) reduction (Luo et al., 2017). As(III) concentration was increased to 178 mg/L, accounting for 21 ± 7% of the total dissolved As in the effluent, after 120 PV in the acceleration phase (Fig. 1). In agreement with the observed As speciation, redox potential measurements superimposed on the pe-pH diagram (Fig. 2, Table S2) suggest that As(III) was the stable As species in the reducing environment (1.6 < pe < 3.6) in biotic column, whereas As(V) predominated in the abiotic control and background columns. Moreover, the phosphate concentration was determined (Figs. S3, S4, and S5), and the results showed that the phosphate initial preloading on TiO2 had a negligible influence on dissolved As(V) release. Total As concentrations in unfiltered effluent increased to 2.5 mg/L in the biotic column and 1.5 mg/L in the abiotic control at the end of the experiment (Fig. 3A), indicating that SRB enhanced the mobility of total As which included dissolved and particulate As. The total effluent As in the background column was only 11 mg/L on average. Compared with dissolved As in the effluent, the total As in the biotic column was substantially higher in the lag and accelerated phase (Fig. 3B). In contrast, dissolved and total As in the effluent were similar in abiotic control and background columns (Fig. 3C and D), indicating that total As concentration was all from dissolved As(V). The results demonstrate that SRB promoted the release of particulate As. The As mass balance showed that 4% of the total As input was passed through the column, however, only 0.2% was in dissolved form. The most interesting observation was that the addition of the SRB restrained the release of dissolved As, while facilitating the transport of particulate As in effluent (Figs. 1 and 3). The functional groups present in cells and cellular matter appear to be a potential sink of As in bioreacting columns in previous studies (Kocar et al., 2006; Huang et al., 2011; Yan et al., 2016). However, the bacteria outflow played only a negligible role for As adsorption as
Fig. 2. pEepH diagram for the AseSeFeeH2O system under experimental conditions. Pyrite and magnetite occupy the regions within the green and blue dashed lines, respectively. Realgar and mackinawite stabilize underneath the red and yellow dashed line regions, respectively. Experimental results are superimposed for dissolved As(V) (blue circle) and As(III) (green circle) in biotic column, and dissolved As(V) in abiotic control (black circle) and background column (red circle). The data was collected in different time during column experiments. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
T. Luo et al. / Environmental Pollution 234 (2018) 762e768
765
Fig. 3. Total As in unfiltered effluent of three columns (A). The total As in unfiltered effluent and dissolved As in filtered effluent of biotic column (B), abiotic control (C), and background column (D). The dashed line denoted the pore volume was 40 which displaced the soil adsorption for arsenic was saturated.
particulate As in effluent in this study. The range of initial and final cell loading was between 7.2 106 and 8.6 108 cells per gram (Table S3). About 2.1 105 cells/L on average was detected in the effluent. Our another batch adsorption test showed that the bacteria adsorption capacity was 20 1010 mg As(III)/cell and 23 1010 mg As(V)/cell (Fig. S6). According to the calculation, the amount of As(III) and As(V) adsorbed on bacteria in the effluent were about 0.42 and 0.48 mg/L, respectively. Thus, the particulate As transport in effluent was not associated with bacteria. 3.2. Sulfate and iron reduction in soil column Effluent SO2 4 in abiotic control and background columns ranged from 5.2 to 5.5 mM, which was comparable to the influent concentration of 5.5 mM (Fig. 4A). In contrast, the average effluent SO2 4 was 3.6 mM in biotic column. The loss of aqueous SO2 4 (1.9 mM) was attributed to SO2 4 reduction by SRB. As a result, sulfide levels increased to 61 mg/L (1.9 mM) at the end of the column experiment, accounting for 34% of the total SO2 4 input (Fig. 4A). The quick loss of sulfate due to the reduction and correspondingly slow production of sulfide highlight the complex As-S-Fe interactions, including the reduction of Fe(III) and As(V) with biogenic sulfide (Burton et al., 2011, 2014; Rochette et al., 2000). Coincident with the increase of sulfide, an increase in Fe(II) up to 180 mg/L after 70 PV (24 d) was observed in biotic column (Fig. 4B). Sulfide and Fe(II) were not detected in abiotic control and background columns. The total Fe in effluent increased from approximately 2.4 to 13 mg/L in inoculated column, while remaining stable in abiotic control (3.1 ± 1.1 mg/L on average) and background column (3.0 ± 1.0 mg/L on average) (Fig. 4B). Iron dissolution due to reduction resulted in a net loss of 0.08% of initial solid phase Fe in soil packed in biotic column. The correlation analysis in effluent from the biotic column found that total As was only correlated with total Fe (R ¼ 0.74, p < .01), not with other elements including Al, Ca, Mg, Mn, Na, Si, and Ti (Fig. S7).
3.3. TiO2 transport in soil column SRB had no appreciable impact on the transport of nano-TiO2, as evidenced by the comparable effluent TiO2 in biotic column (147 ± 92 mg/L) and abiotic control (126 ± 90 mg/L) (Fig. S8). About 34 ± 28 mg/L TiO2 were released from the soil in the background column. The nano-TiO2 cannot be the dominant As carrier as particulate As in effluent (Fig. 3). The average TiO2 concentration in unfiltered effluent of biotic column was only 147 ± 92 mg/L. Even if all of the TiO2 in effluent was associated with As, at the initial concentration of 390 mg-As/g-TiO2, the maximum As concentration would be only about 57 ± 35 mg/L in effluent. This As concentration is substantially less than the detected total As at mg/L level (Fig. 3). In addition, no significant correlation was found between the total As and Ti in the effluent (Fig. S7), indicating that the particulate As release might not result from the transport of nano-TiO2. The mass balance of TiO2 showed that over 99.9% TiO2 was retained in biotic column as evidenced by the negligible amount of Ti in effluent (3e450 mg/L, Fig. S8) and the considerable amounts in the solids (1943e2634 mg/kg, Table S1). Moreover, Ti contents in influent front of the column (2634 mg/kg) were greater than the effluent end (1943 mg/kg), indicating that a physical mechanism, such as straining, is mainly responsible for the TiO2 trap. 3.4. In situ column analysis with synchrotron techniques
m-XRF images showed the distribution of As, Ti, Fe, Mn, S, Zn, Ca, and Cu in the influent front of biotic column (Fig. 5). The images and correlation plot for the solid sample show a weak significant correlation between As and Ti (R ¼ 0.36, p < .01), which might be ascribed to the initial As injected including soluble As in suspension and adsorbed As on TiO2. There was no correlation between As and other elements (Fig. S9). Therefore, the majority of adsorbed As in the column was still associated with TiO2, which was mainly
766
T. Luo et al. / Environmental Pollution 234 (2018) 762e768
precipitates with As, such as orpiment and realgar, and soluble As-S complexes, such as di-and tri-thioarsenite monomers as well as dimeric and trimeric arsenic-sulfur soluble complex (Rochette et al., 2000; Couture et al., 2013; Stucker et al., 2014). Thioarsenate (H2AsO3S) and dithioarsenate (H2AsO2S 2 ) may form before As(V) was reduced to As(III) (Rochette et al., 2000). Moreover, the formation of thioarsenite (H3AsO2S) and dithioarsenite (H2AsOS 2 ) might be also occurred in the process of As(V) reduction by sulfide (Rochette et al., 2000). These soluble thio-As species, however, were not observed in the effluent of the three columns in this study. Meanwhile, both the thermodynamic pe-pH calculation (Fig. 2 and Table S2) and XANES analysis (Fig. 6) did not support the formation of realgar and orpiment under the experimental conditions. Fe(III) oxides can be reduced by the biogenic sulfide according to Eq. (1) (Poulton et al., 2004): 2Fe(OH)3 þ HS þ 5Hþ / 2Fe2þ þ S0 þ 6H2O
(1)
The produced Fe(II) can then form secondary Fe-minerals, such as pyrite, mackinawite, magnetite, and lepidocrocite (Kumar et al., 2016a,b; Burton et al., 2011; Kirk et al., 2010; Morse and Rickard, 2004). In agreement with our thermodynamic pe-pH calculation and XANES analysis, the formation of magnetite and mackinawite during the microbial sulfidogenesis has been reported in a series of iron-rich model systems (Kumar et al., 2016a,b; Kocar et al., 2010; Burton et al., 2011; Tufano and Fendorf, 2008). Previous studies indicate that the reaction of Fe(III) oxides and Fe(II) leads to magnetite formation according to Eq. (2) (Kocar et al., 2010; Saalfield and Bostick, 2009): Fe2þ þ 2Fe(OH)3 / Fe3O4 þ 2Hþ þ 2H2O SO2 4
Fig. 4. concentrations (open symbols) in filtered effluent of the three columns, and sulfide concentrations (closed symbols) in biotic column (A). Total Fe concentrations in unfiltered effluent of the three columns (open symbols), and Fe(II) concentrations (closed symbols) in filtered effluent of biotic column(B).
retained in the influent front of the soil column during the transport process. The speciation of As and Fe in biotic column was determined with XANES (Fig. 6). As(V) was the primary As species in biotic column as evidenced by its peak position at 11,874 eV (Fig. 6A) (Jing et al., 2005). As(V) (92%) and As(III) (8%) were the As species in the influent front, and no As(III) was detected in the effluent end of the biotic column (Fig. 6A, Fig. S10). As-S precipitates, such as realgar and orpiment, were not detected by XANES (Fig. 6A), in line with the pe-pH diagram predication (Fig. 2). In agreement to the pe-pH diagram, the Fe k-edge XANES analysis indicates the presence of magnetite in biotic column (Fig. 6B). The mackinawite, as a pyrite precursor, was also detected by XANES (Hunger and Benning, 2007), though it was not predicted by thermodynamic pe-pH diagram under the experimental conditions (Fig. 2). XANES results showed that magnetite (23.4%), mackinawite (21%), FeOOH (21.1%), Fe2O3 (17.7%), and FeO (16.8%) were the solid iron minerals in the influent front of the biotic column (Fig. S10, Table S4). In the effluent end, magnetite, mackinawite, FeOOH, FeO, and Fe2O3 accounted for 31.5%, 16.9%, 28.3%, 17.3%, and 6%, respectively, of the total iron minerals. 3.5. As redistribution and transport in soil The fate of As is implicitly tied to SO4 and Fe(III) reduction as mediated by SRB (Kumar et al., 2016a,b; Burton et al., 2011; Dhar et al., 2011). The biogenic sulfide can form insoluble As-S
(2)
As has a high adsorption affinity to magnetite, and therefore exhibits a strong preference for partitioning to magnetite in soil (Burton et al., 2011). Mackinawite has often been observed in the presence of microbial sulfidogenesis in an iron-rich system due to Eq. (3) (Kumar et al., 2016a,b; Kocar et al., 2010; Burton et al., 2011; Kirk et al., 2010; Morse and Rickard, 2004): Fe2þ þ HS / FeS þ Hþ
(3)
Mackinawite is also effective in immobilizing As (Kumar et al., 2016a,b; Kocar et al., 2010; Burton et al., 2011; Kirk et al., 2010). The formation of magnetite and mackinawite were observed as evidenced by XANES analysis (Fig. 6) and pe-pH diagram (Fig. 2). The color of the biotic column turned from brown to black at the end of the experiment (Fig. S11), further supporting the formation of Fe(II) minerals. In line with our results, Saalfield and Bostick (2009) found that microbial sulfidogenesis can mitigate As mobility due to the formation of secondary iron minerals. Interestingly, the total effluent As from biotic column was greater than that in abiotic control (Fig. 3). Actually, the dissolved As from biotic column was the sum of the As originally soluble in influent and those liberated from TiO2. The dissolved As attached to the in situ generated nanosized secondary iron minerals including magnetite and mackinawite (accounting for 44.4%e48.4% in solids of the biotic column, Fig. S12), and flew out of the column as particulate As. Hellige et al. (2012) also found that the reaction between sulfide and lepidocrocite can produce nanometer-scale magnetite. Kumar et al. (2016a,b) reported that nanocrystallite mackinawite was formed in the system of As-loaded zero-valent iron with the microbial sulfate reduction. The average ratio of As/Fe was 315 ± 137 mg/g in effluent in this study, which is comparable to
T. Luo et al. / Environmental Pollution 234 (2018) 762e768
767
Fig. 5. Spatial distributions of As, Ti, Fe, Mn, S, Zn, Ca, and Cu in solid sample which at influent front of biotic column. The relative concentrations of the elements are shown on the right in the colored bar scales in numbers of fluorescence count. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
Fig. 6. Normalized XANES at As (A) and Fe (B) K-edge for samples of biotic column. Solid line was experimental data, dotted line was linear combination fitting.
768
T. Luo et al. / Environmental Pollution 234 (2018) 762e768
the reported adsorption capacity of 214e334 mg/g for ferrihydrite from pH 4.6 to 9.2 (Raven et al., 1998; Dixit and Hering, 2003). 4. Conclusions The present study reveals the specific effects of SRB on the mobility of As and nano-TiO2 adsorbent in natural soil columns by excluding other confounding microbial processes. Our results indicate that SRB inhibit dissolved As release due to its re-adsorption on in situ formed secondary iron minerals including magnetite and mackinawite. Moreover, the SRB facilitate the migration of secondary iron minerals associated with As. The transport of these minerals might be due to their submicron or nanometer particle size, and therefore, lead to the As mobility and contamination in the subsurface. This finding provides insights on As biogeochemical cycling in the natural environment. Specifically, As mobility might not only be attributed to the dissolved As liberated from the solids, but also to its migration with in situ formed secondary iron minerals. Meanwhile, the nano-TiO2 released to the effluent was only 147 ± 92 mg/L with influent concentration 0.1 g/L. The nano-TiO2 with loaded As was mainly retained in the front of soil column. Therefore, the nano-TiO2 release from the soil and subsequently its the effect on As mobility were not significant as we initially thought. Acknowledgments We acknowledge the financial support of the National Basic Research Program of China (2015CB932003), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB14020201), and the National Natural Science Foundation of China (41373123, 41425016, and 21407124). The m-XRF was performed at SSRF BL15U, and XANES spectra were acquired at NSRRC BL01C1. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.envpol.2017.12.029. References Burton, E.D., Johnston, S.G., Bush, R.T., 2011. Microbial sulfidogenesis in ferrihydriterich environments: effects on iron mineralogy and arsenic mobility. Geochim. Cosmochim. Acta 75, 3072e3087. Burton, E.D., Johnston, S.G., Kocar, B.D., 2014. Arsenic mobility during flooding of contaminated soil: the effect of microbial sulfate reduction. Environ. Sci. Technol. 48, 13660e13667. Canfield, D.E., Thamdrup, B., Kristensen, E., 2005. The Sulfur Cycle. In Aquatic Geomicrobiology. Elsevier Academic Press, San Diego. €ger, D., Van Cappellen, P., 2013. Couture, R.M., Rose, J., Kumar, N., Mitchell, K., Wallschla Sorption of arsenite, arsenate, and thioarsenates to iron oxides and iron sulfides: a kinetic and spectroscopic investigation. Environ. Sci. Technol. 47, 5652e5659. €ger, D., Rose, J., Van Cappellen, P., 2014. Arsenic binding to Couture, R.M., Wallschla organic and inorganic sulfur species during microbial sulfate reduction; a sediment flow-through reactor experiment. Environ. Chem. 10, 285e294. Dhar, R.K., Zheng, Y., Saltikov, C.W., Radloff, K.A., Mailloux, B.J., Ahmed, K.M., van Geen, A., 2011. Microbes enhance mobility of arsenic in pleistocene aquifer sand from Bangladesh. Environ. Sci. Technol. 45, 2648e2654. Dixit, S., Hering, J.G., 2003. Comparison of arsenic(V) and arsenic(III) sorption onto iron oxide minerals: implications for arsenic mobility. Environ. Sci. Technol. 37, 4182e4189. Eaton, A.D., Clesceri, L.S., Greenberg, A.E., 1995. Stand Methods for the Examination of Water and Wastewater, nineteenth ed. American Public Health Association, Washington DC. Gherna, R., Pienta, P., Cote, R., 1992. American Type Culture Collection Catalogue of Bacteria and Phages, eighteenth ed. (Rockville, MD). Guo, H., Liu, Z., Ding, S., Hao, C., Xiu, W., Hou, W., 2015. Arsenate reduction and mobilization in the presence of indigenous aerobic bacteria obtained from high arsenic aquifers of the Hetao basin, Inner Mongolia. Environ. Pollut. 203, 50e59. Hellige, K., Pollok, K., Larese-Casanova, P., Behrends, T., Peiffer, S., 2012. Pathways of ferrous iron mineral formation upon sulfidation of lepidocrocite surfaces. Geochim. Cosmochim. Acta 81, 69e81.
Huang, J.H., Elzinga, E.J., Brechbuehl, Y., Voegelin, A., Kretzschmar, R., 2011. Impacts of shewanella putrefaciens strain CN-32 cells and extracellular polymeric substances on the sorption of As(V) and As(III) on Fe(III)-(Hydr)oxides. Environ. Sci. Technol. 45, 2804e2810. Hunger, S., Benning, L.G., 2007. Greigite: a true intermediate on the polysulfide pathway to pyrite. Geochem. Trans. 8, 1. Jing, C., Liu, S., Meng, X., 2005. Arsenic leachability and speciation in cement immobilized water treatment sludge. Chemosphere 59, 1241e1247. Jing, C., Meng, X., Calvache, E., Jiang, G., 2009. Remediation of organic and inorganic arsenic contaminated groundwater using a nanocrystalline TiO2-based adsorbent. Environ. Pollut. 157, 2514e2519. Kirk, M.F., Roden, E.E., Crossey, L.J., Brearley, A.J., Spilde, M.N., 2010. Experimental analysis of arsenic precipitation during microbial sulfate and iron reduction in model aquifer sediment reactors. Geochim. Cosmochim. Acta 74, 2538e2555. Kocar, B.D., Borch, T., Fendorf, S., 2010. Arsenic repartitioning during biogenic sulfidization and transformation of ferrihydrite. Geochim. Cosmochim. Acta 74, 980e994. Kocar, B.D., Herbel, M.J., Tufano, K.J., Fendorf, S., 2006. Contrasting effects of dissimilatory iron(III) and arsenic(V) reduction on arsenic retention and transport. Environ. Sci. Technol. 40, 6715e6721. Kumar, N., Couture, R.M., Millot, R., Battaglia-Brunet, F., Rose, J., 2016a. Microbial sulfate reduction enhances arsenic mobility downstream of zerovalent-ironbased permeable reactive barrier. Environ. Sci. Technol. 50, 7610e7617. Kumar, N., Millot, R., Battaglia-Brunet, F., Omoregie, E., Chaurand, P., Borschneck, D., Bastiaens, L., Rose, J., 2016b. Microbial and mineral evolution in zero valent iron-based permeable reactive barriers during long-term operations. Environ. Sci. Pollut. Res. 23, 5960e5968. Lowers, H.A., Breit, G.N., Foster, A.L., Whitney, J., Yount, J., Uddin, N., Muneem, A., 2007. Arsenic incorporation into authigenic pyrite, bengal basin sediment. Bangladesh. Geochim. Cosmochim. Acta 71, 2699e2717. Luo, T., Cui, J., Hu, S., Huang, Y., Jing, C., 2010. Arsenic removal and recovery from copper smelting wastewater using TiO2. Environ. Sci. Technol. 44, 9094e9098. Luo, T., Hu, S., Cui, J., Tian, H., Jing, C., 2012. Comparison of arsenic geochemical evolution in the Datong basin (Shanxi) and Hetao basin (Inner Mongolia), China. Appl. Geochem. 27, 2315e2323. Luo, T., Tian, H., Guo, Z., Zhuang, G., Jing, C., 2013. Fate of arsenate adsorbed on nano-TiO2 in the presence of sulfate reducing bacteria. Environ. Sci. Technol. 47, 10939e10946. Luo, T., Ye, L., Ding, C., Yan, J., Jing, C., 2017. Reduction of adsorbed As(V) on nanoTiO2 by sulfate-reducing bacteria. Sci. Total Environ. 598, 839e846. Morse, J.W., Rickard, D., 2004. Chemical dynamics of sedimentary acid volatile sulfide. Environ. Sci. Technol. 38, 131Ae136A. Pena, M., Meng, X.G., Korfiatis, G.P., Jing, C., 2006. Adsorption mechanism of arsenic on nanocrystalline titanium dioxide. Environ. Sci. Technol. 40, 1257e1262. Polizzotto, M.L., Harvey, C.F., Sutton, S.R., Fendorf, S., 2005. Processes conducive to the release and transport of arsenic into aquifers of Bangladesh. Proc. Natl. Acad. Sci. U. S. A 102, 18819e18823. Poulton, S.W., Krom, M.D., Raiswell, R., 2004. A revised scheme for the reactivity of iron (oxyhydr)oxide minerals towards dissolved sulfide. Geochim. Cosmochim. Acta 68, 3703e3715. Raven, K.P., Jain, A., Loeppert, R.H., 1998. Arsenite and arsenate adsorption on ferrihydrite: kinetics, equilibrium, and adsorption envelopes. Environ. Sci. Technol. 32, 344e349. Ravel, B., Newville, M., 2005. Athena, artemis, hephaestus: data analysis for X-ray absorption spectroscopy using IFEFFIT. J. Synchrotron Radiat. 12, 537e541. Rochette, E.A., Bostick, B.C., Li, G., Fendorf, S., 2000. Kinetics of arsenate reduction by dissolved sulfide. Environ. Sci. Technol. 34, 4714e4720. Saalfield, S.L., Bostick, B.C., 2009. Changes in iron, sulfur, and arsenic speciation associated with bacterial sulfate reduction in ferrihydrite-rich systems. Environ. Sci. Technol. 43, 8787e8793. Smedley, P.L., Kinniburgh, D.G., 2002. A review of the source, behaviour and distribution of arsenic in natural waters. Appl. Geochem. 17, 517e568. Stucker, V.K., Silverman, D.R., Williams, K.H., Sharp, J.O., Ranville, J.F., 2014. Thioarsenic species associated with increased arsenic release during biostimulated subsurface sulfate reduction. Environ. Sci. Technol. 48, 13367e13375. Sun, J., Steven, N., Brian, C., Mailloux, J., Bostick, B.C., 2016. Arsenic mobilization from sediments in microcosms under sulfate reduction. Chemosphere 153, 254e261. Tufano, K.J., Fendorf, S., 2008. Confounding impacts of iron reduction on arsenic retention. Environ. Sci. Technol. 42, 4777e4783. Wolthers, M., Charlet, L., van der Weijden, C.H., van der Linde, P.R., Rickard, D., 2005. Arsenic mobility in the ambient sulfidic environment: sorption of arsenic(V) and arsenic(III) onto disordered mackinawite. Geochim. Cosmochim. Acta 69, 3483e3492. Yan, L., Huang, Y., Cui, J., Jing, C., 2015. Simultaneous As(III) and Cd removal from copper smelting wastewater using granular TiO2 columns. Water Res. 68, 572e579. Yan, W., Wang, H., Jing, C., 2016. Adhesion of shewanella oneidensis MR-1 to goethite: a two-dimensional correlation spectroscopic study. Environ. Sci. Technol. 50, 4343e4349. Ye, L., Liu, W., Shi, Q., Jing, C., 2017. Arsenic mobilization in spent nZVI waste residue: effect of Pantoea sp.IMH. Environ. Pollut. 230, 1081e1089. Zhang, D., Guo, H., Xiu, W., Ni, P., Zheng, H., Wei, C., 2017. In-situ mobilization and transformation of iron oxides-adsorbed arsenate in natural groundwater. J. Hazard Mater. 321, 228e237. Zhang, X., Sun, H., Zhang, Z., Niu, Q., Chen, Y., Crittenden, J.C., 2007. Enhanced bioaccumulation of cadmium in carp in the presence of titanium dioxide nanoparticles. Chemosphere 67, 160e166.
Contents lists available at ScienceDirect
Environmental Pollution journal homepage: www.elsevier.com/locate/envpol
Mobilization of arsenic on nano-TiO2 in soil columns with sulfate reducing bacteria* Ting Luo a, b, Li Ye a, c, Tingshan Chan d, **, Chuanyong Jing a, c, * a
State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China b School of Environmental Science and Engineering, Yancheng Institute of Technology, Jiangsu 224051, China c University of Chinese Academy of Sciences, Beijing 100049, China d National Synchrotron Radiation Research Center, 101 Hsin-Ann Road, Hsinchy Science Park, Hsinchu 30076, Taiwan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 18 October 2017 Received in revised form 6 December 2017 Accepted 8 December 2017
Arsenic (As) remediation in contaminated water using nanoparticles is promising. However, the fate and transport of As associated with nano-adsorbents in natural environment is poorly understood. To investigate the fate of adsorbed As on nano-TiO2 in changed redox condition from oxic to anoxic, we added the As(V)-TiO2 suspension in groundwater to an autoclaved soil column which inoculated a sulfate-reducing bacterium, Desulfovibrio vulgaris DP4. The dissolved As(V) in effluent increased to 798 mg/L for the biotic column and to 1510 mg/L for the abiotic control, and dissolved As(III) was observed only in biotic column. The total As (dissolved plus particulate) in the biotic column effluent (high to 2.5 mg/L) was substantially higher than the abiotic control (1.5 mg/L). Therefore SRB restrained the release of dissolved As, and facilitated the transport of particulate As. Micro-XRF analysis suggested that the nano-TiO2 with As was mainly retained in the influent front and that its transport was negligible. Our pe-pH calculation and XANES analysis demonstrated that generated secondary iron minerals containing magnetite and mackinawite mainly were responsible for dissolved As retention, and then transported with As as particulate As. The results shed light on the mobilization of adsorbed As on a nano-adsorbent in an anoxic environment. © 2017 Elsevier Ltd. All rights reserved.
Keywords: Arsenic mobilization Nano-TiO2 Sulfate reducing bacteria Soil column
1. Introduction Arsenic (As) is a redox active and toxic element that poses considerable human health risks. Recently, adsorption on nanoTiO2 has been demonstrated to be a promising technique for removing As from groundwater and industrial wastewater (Jing et al., 2009; Luo et al., 2010; Pena et al., 2006; Yan et al., 2015). However, the fate of the As, once it is removed from the water, is not well known. The pressing need to predict the fate and transport of As adsorbed onto nano-TiO2during environmental redox transition has motivated this study. The shift in the redox environment from oxidizing to reducing, as mediated by microbial activities, plays an important role in As *
This paper has been recommended for acceptance by B. Nowack. * Corresponding author. State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China. ** Corresponding author. E-mail addresses: [email protected] (T. Chan), [email protected] (C. Jing). https://doi.org/10.1016/j.envpol.2017.12.029 0269-7491/© 2017 Elsevier Ltd. All rights reserved.
biogeochemical cycling (Ye et al., 2017; Sun et al., 2016; Guo et al., 2015; Polizzotto et al., 2005). Sulfate-reducing bacteria (SRB) are ubiquitous in the anoxic subsurface, such as in groundwater systems (Burton et al., 2014; Canfield et al., 2005), and biogenic sulfide is an efficient reductant for As(V) and Fe(III) (Rochette et al., 2000; Poulton et al., 2004). Consequently, the reduced As(III) may form As-sulfide precipitates such as realgar (AsS) and orpiment (As2S3), which controls the dissolved As levels in the aquifer (Couture et al., 2014; Kumar et al., 2016a). Moreover, the formation of Fe(II)-sulfide minerals provides an additional sink for As immobilization (Lowers et al., 2007; Zhang et al., 2017). The effect of SRB on the mobility of As and the transformation of iron minerals has been extensively studied using ferrihydritecoated sand column experiments. Kocar et al. (2010) observed that SRB (Desulfovibrio vulgaris) inhibit As release from the sand column and induce the ferrihydrite transfer to magnetite (Fe3O4), green rust, and amorphous FeS. Burton et al. (2011) reported that SRB (Desulfovibrio vulgaris) enhance the As mobility in ferrihydriterich environments at pH 8, and conversely inhibit the As release at
T. Luo et al. / Environmental Pollution 234 (2018) 762e768
pH 6, due to the formation of goethite (FeOOH). In the presence of iron, As mobility may also be limited by its sorption to secondary Fe-minerals that formed in situ, including magnetite, pyrite (FeS2), greigite (Fe3S4), and Lepidocrocite (g-FeOOH) (Kumar et al., 2016b; Burton et al., 2011; Kirk et al., 2010). However, to the best of our knowledge, no publications report on the effects of SRB on the transport and potential risks of nanosized-TiO2-containing As in the natural environment. There is a lack of knowledge as to whether nano-TiO2-containing As ends up in soils or travels long distances in the subsurface, and how SRB affect this complicated process. It is critical to understand whether the current and future use of nano-TiO2 for As removal would trigger the release of As, and result in an even greater amount of As contamination due to the mobility of nanoparticles. Moreover, previous column experiments mainly adopt idealized model systems, such as sand or ferrihydrite-coated sand columns, which may not adequately represent the complex natural environment. The objectives of this research are to study the redox reactions and release of adsorbed As on nano-TiO2in the presence of SRB in soil columns. The soil column and effluent were monitored for As and Fe speciation and TiO2 transport using complimentary analytical techniques, including XANES and m-XRF, and were analyzed by thermodynamic calculations. This study provides the first evidence regarding the mobilization of adsorbed As on nano-TiO2 with SRB in natural soil under anoxic conditions. 2. Materials and methods 2.1. As(V) loading on nano-TiO2 The nano-TiO2 was prepared via hydrolysis of titanyl sulfate, as described in our previous work (Luo et al., 2010). Synthetic groundwater was prepared based on the composition of groundwater samples from naturally-occurring As sites in Shanxi, China (Luo et al., 2012): KNO3 0.02 mM, NaHCO3 6.39 mM, CaCO3 0.16 mM, MgCl2 0.87 mM, NH4Cl 0.02 mM, NaCl 2.94 mM, pH 8.0. The As(V) solution was prepared with Na2HAsO4$7H2O in a groundwater matrix. The adsorption of As(V) on TiO2 was conducted in a glovebox (100% N2). Phosphate, added as Na2HPO4, was first pre-adsorbed on the sterilized TiO2 at a loading of 750 mg/kg to facilitate microbial growth (Kocar et al., 2006). The As(V) solution was mixed with the TiO2-P solids for 24 h at pH 8.0. Then, the TiO2-As(V) suspension was centrifuged, and the TiO2-As solid was rinsed several times with deionized (DI) water. The final load of As(V) on TiO2 was approximately 390 mg/g by mass balance calculation. The TiO2-As(V) suspension with the TiO2 concentration of 0.1 g/L was prepared by dispersing the TiO2-As(V) solid into the groundwater matrix. The soluble As(V) concentration was 2.36 mg/L after the TiO2-As(V) solid was dispersed into the groundwater. The average particle size in the TiO2-As(V) suspension was 628 nm, as determined by a Zeta sizer Nano ZS (Malvern Instrument, UK). 2.2. Column experiments Soil samples were collected at 1 m depth at a site in Shanxi, China (GPS location 39 250 24.300 N, 112 520 50.000 E). The As content in soil was 8.6 mg/kg, which is within the background As level (5e10 mg/kg) of uncontaminated soils (Smedley and Kinniburgh, 2002). The components of the soil are shown in Table S1. The autoclaved soil was used for packing the columns. A sulfate-reducing bacterium identified as Desulfovibrio vulgaris DP4 was isolated from the soil (Luo et al., 2013). The SRB were grown to the late-exponential phase under anoxic conditions in a modified Baar's medium (Gherna et al., 1992) at 37 C. Cells were
763
harvested by centrifugation (6000 g, 8 min, 25 C), and then rinsed with synthetic groundwater. The SRB were subsequently added to the autoclaved soil and mixed in the glovebox. Then, 14 g of this mixture was packed into a glass column (length ¼ 10 cm, diameter ¼ 1.2 cm), resulting in a porosity of 0.54. The TiO2-As(V) suspension was delivered to the column in an upward flow using a peristaltic pump at three pore volumes (PV) per day (11 mL/min). Sulfate (MgSO4 5.5 mM) was added as an electron acceptor. The electron donor was lactate (Na-lactate10 mM). Two abiotic columns, without the addition of SRB, were also set up for comparison. The influent stream for one abiotic column was the TiO2-As(V) suspension, which was used as a control. Synthetic groundwater was used as background in the other column. The column experiments were conducted in the glovebox. Eh and pH in effluent samples were monitored over the experimental period of 51 d. A subsample of the effluent was passed through a 0.22 mm syringe filter for the analysis of dissolved As species, ferrous iron, sulfide, sulfate, and phosphate. The unfiltered effluent sample was used to analyze the total As, Ti, Fe, Ca, Na, Mg, Al, Si, and Mn. The solids were sampled for cell counts during the column experiments using an epiflourescence microscope with 40 -6-diamidino-2phenylindoledDAPIdstaining (Kocar et al., 2010). 2.3. Aqueous analysis Arsenic speciation was determined using high performance liquid chromatography atomic fluorescence spectrometry (HPLCAFS, Jitian, China). Ferrous iron and sulfide concentrations were measured using the colorimetric method Phenanthroline Method and Methylene Blue Method (Eaton et al., 1995), respectively, immediately following the sample collection. Sulfate and phosphate concentrations were determined using a DX-1100 ion chromatograph (Dionex, US) with an AS11-HC Ion Pac column. 2.4. Total metal concentrations in effluent To determine the total content of As, Ti, Fe, Ca, Na, Mg, Al, Si, and Mn in the unfiltered effluent, the samples were heated dry, followed by microwave digestion with 4 mL concentrated HNO3, 1 mL concentrated HF, and 5 mL of the sulfuric acid-ammonium sulfate solution (Zhang et al., 2007) using a microwave accelerated reaction system (CEM MARS, US). The digestion solution was then transferred to a 10 mL volumetric flask. The soil and solids were digested by the same method in a column incubated with Desulfovibrio vulgaris DP4. The total As, Ti, and Fe concentrations were determined by ICP-MS (Agilent, US). The total Mn, Al, Si, Mg, Ca, and Na concentrations were measured by ICP-OES (Perkin Elmer, US). 2.5. m-XRF and XANES analysis The solids in the bottom (influent front) and upper (end) layers of the column were collected at the end of the experiment and then freeze dried. The solid sample was deposited onto Kapton tape and analyzed by m-XRF at beamline 15U at Shanghai Synchrotron Radiation Facility (SSRF), China. m-XRF analysis was conducted with the monochromator set at 12 keV, a 10 10 mm beam, a 1-second dwell time per pixel, and a step size of 10 mm. The peak intensities for As, S, Fe, Mn, Ti, Zn, Ca, and Cu were collected at each pixel of m-XRF maps, which were 0.6 0.6 mm in size. The As and Fe K-edge XANES spectra were collected at beamline 01C1 at the National Synchrotron Radiation Research Center (NSRRC), Taiwan. An energy range of 150 to 300 eV from the K-edge of As (11867 eV) and Fe (7112 eV) was used to acquire the spectra with a 0.5 eV step size from 20 eV below to 50 eV above the absorption edge. The spectra were taken under standard NSRRC
764
T. Luo et al. / Environmental Pollution 234 (2018) 762e768
TOP-UP operation conditions (1.5 GeV and 240 mA) with a doublecrystal Si (111) monochromator. Fluorescence signals were collected using a Lytle detector, positioned at 90 to the incident beam. Spectra were acquired at cryogenic temperatures (77 K) using a helium cryostat to prevent the beam-induced oxidation of As(III) and Fe(II). Four scans were collected from each sample, inspected for overall quality, and averaged to improve the signal/noise ratio. As standard reference chemicals, Na2HAsO4$7H2O, NaAsO2, realgar (AsS), orpiment (As2S3), and arsenopyrite (FeAsS) were analyzed. Fe reference standards including magnetite (Fe3O4), mackinawite (FeS), FeAsS, FeS2, FeOOH, FeSO4, Fe2O3, and FeO were also measured. Fe reference standards were purchased in the National Standard Substances Center in China except mackinawite. Mackinawite was synthesized in lab according to a previous method (Wolthers et al., 2005). The XANES analysis procedure was described in our previous study (Luo et al., 2010). Briefly, the spectra were analyzed with the Athena program in the IFEFFIT computer package (Ravel and Newville, 2005) for a linear combination fit over the relative energy range of 20 to 30 eV. No energy shift was allowed during the fitting procedure. The weighting of single species was constrained between 0 and 1, and the sum of species was constrained to be 1. 3. Results and discussion 3.1. Arsenic release and reduction in soil columns Effluent As was monitored during the 51 d (153 PV) experiment. An experiment column was incubated with D. vulgaris DP4. There was also one abiotic control column and one background column which pumped groundwater only without TiO2-As(V) suspension (Fig. 1). The release of As can be divided into a lag phase (<40 PV, 13 d) and an acceleration phase (>40 PV). In the lag phase, dissolved effluent As(V) values were all below 100 mg/L, demonstrating the dissolved As(V) in the influent was adsorbed on soil minerals. The accumulative dissolved As(V) load in the lag phase was about 35 mg/kg, which was comparable to the As(V) adsorption capacity of the soil (32.5 mg/kg, Fig. S1). When the soil became saturated, As release accelerated, so that the dissolved As(V) in the effluent increased substantially to 798 mg/L for the biotic column and to 1510 mg/L for the control, at 153 PV. In contrast, only 11 mg/L As(V),
Fig. 1. Dissolved As(III) and As(V) in filtered samples from biotic column incubated with D. vulgaris DP4, an abiotic control with no D. vulgaris DP4, and a background column with artificial groundwater flow and no TiO2-As(V) suspension. The dashed line denoted the pore volume was 40 which displaced the soil adsorption for arsenic was saturated.
on average, were detected in the effluent of the background column, suggesting that the initial As in the soil (8.6 mg/kg) contributed little to the overall As release. Even the presence of SRB resulted in only an 8 mg/L As release from the soil to the groundwater matrix (Fig. S2). Therefore, the addition of the SRB restrained the release of dissolved As. Dissolved As(III) in the effluent was observed only from the biotic column after 3 PV, indicating that SRB induced the As(V) reduction (Luo et al., 2017). As(III) concentration was increased to 178 mg/L, accounting for 21 ± 7% of the total dissolved As in the effluent, after 120 PV in the acceleration phase (Fig. 1). In agreement with the observed As speciation, redox potential measurements superimposed on the pe-pH diagram (Fig. 2, Table S2) suggest that As(III) was the stable As species in the reducing environment (1.6 < pe < 3.6) in biotic column, whereas As(V) predominated in the abiotic control and background columns. Moreover, the phosphate concentration was determined (Figs. S3, S4, and S5), and the results showed that the phosphate initial preloading on TiO2 had a negligible influence on dissolved As(V) release. Total As concentrations in unfiltered effluent increased to 2.5 mg/L in the biotic column and 1.5 mg/L in the abiotic control at the end of the experiment (Fig. 3A), indicating that SRB enhanced the mobility of total As which included dissolved and particulate As. The total effluent As in the background column was only 11 mg/L on average. Compared with dissolved As in the effluent, the total As in the biotic column was substantially higher in the lag and accelerated phase (Fig. 3B). In contrast, dissolved and total As in the effluent were similar in abiotic control and background columns (Fig. 3C and D), indicating that total As concentration was all from dissolved As(V). The results demonstrate that SRB promoted the release of particulate As. The As mass balance showed that 4% of the total As input was passed through the column, however, only 0.2% was in dissolved form. The most interesting observation was that the addition of the SRB restrained the release of dissolved As, while facilitating the transport of particulate As in effluent (Figs. 1 and 3). The functional groups present in cells and cellular matter appear to be a potential sink of As in bioreacting columns in previous studies (Kocar et al., 2006; Huang et al., 2011; Yan et al., 2016). However, the bacteria outflow played only a negligible role for As adsorption as
Fig. 2. pEepH diagram for the AseSeFeeH2O system under experimental conditions. Pyrite and magnetite occupy the regions within the green and blue dashed lines, respectively. Realgar and mackinawite stabilize underneath the red and yellow dashed line regions, respectively. Experimental results are superimposed for dissolved As(V) (blue circle) and As(III) (green circle) in biotic column, and dissolved As(V) in abiotic control (black circle) and background column (red circle). The data was collected in different time during column experiments. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
T. Luo et al. / Environmental Pollution 234 (2018) 762e768
765
Fig. 3. Total As in unfiltered effluent of three columns (A). The total As in unfiltered effluent and dissolved As in filtered effluent of biotic column (B), abiotic control (C), and background column (D). The dashed line denoted the pore volume was 40 which displaced the soil adsorption for arsenic was saturated.
particulate As in effluent in this study. The range of initial and final cell loading was between 7.2 106 and 8.6 108 cells per gram (Table S3). About 2.1 105 cells/L on average was detected in the effluent. Our another batch adsorption test showed that the bacteria adsorption capacity was 20 1010 mg As(III)/cell and 23 1010 mg As(V)/cell (Fig. S6). According to the calculation, the amount of As(III) and As(V) adsorbed on bacteria in the effluent were about 0.42 and 0.48 mg/L, respectively. Thus, the particulate As transport in effluent was not associated with bacteria. 3.2. Sulfate and iron reduction in soil column Effluent SO2 4 in abiotic control and background columns ranged from 5.2 to 5.5 mM, which was comparable to the influent concentration of 5.5 mM (Fig. 4A). In contrast, the average effluent SO2 4 was 3.6 mM in biotic column. The loss of aqueous SO2 4 (1.9 mM) was attributed to SO2 4 reduction by SRB. As a result, sulfide levels increased to 61 mg/L (1.9 mM) at the end of the column experiment, accounting for 34% of the total SO2 4 input (Fig. 4A). The quick loss of sulfate due to the reduction and correspondingly slow production of sulfide highlight the complex As-S-Fe interactions, including the reduction of Fe(III) and As(V) with biogenic sulfide (Burton et al., 2011, 2014; Rochette et al., 2000). Coincident with the increase of sulfide, an increase in Fe(II) up to 180 mg/L after 70 PV (24 d) was observed in biotic column (Fig. 4B). Sulfide and Fe(II) were not detected in abiotic control and background columns. The total Fe in effluent increased from approximately 2.4 to 13 mg/L in inoculated column, while remaining stable in abiotic control (3.1 ± 1.1 mg/L on average) and background column (3.0 ± 1.0 mg/L on average) (Fig. 4B). Iron dissolution due to reduction resulted in a net loss of 0.08% of initial solid phase Fe in soil packed in biotic column. The correlation analysis in effluent from the biotic column found that total As was only correlated with total Fe (R ¼ 0.74, p < .01), not with other elements including Al, Ca, Mg, Mn, Na, Si, and Ti (Fig. S7).
3.3. TiO2 transport in soil column SRB had no appreciable impact on the transport of nano-TiO2, as evidenced by the comparable effluent TiO2 in biotic column (147 ± 92 mg/L) and abiotic control (126 ± 90 mg/L) (Fig. S8). About 34 ± 28 mg/L TiO2 were released from the soil in the background column. The nano-TiO2 cannot be the dominant As carrier as particulate As in effluent (Fig. 3). The average TiO2 concentration in unfiltered effluent of biotic column was only 147 ± 92 mg/L. Even if all of the TiO2 in effluent was associated with As, at the initial concentration of 390 mg-As/g-TiO2, the maximum As concentration would be only about 57 ± 35 mg/L in effluent. This As concentration is substantially less than the detected total As at mg/L level (Fig. 3). In addition, no significant correlation was found between the total As and Ti in the effluent (Fig. S7), indicating that the particulate As release might not result from the transport of nano-TiO2. The mass balance of TiO2 showed that over 99.9% TiO2 was retained in biotic column as evidenced by the negligible amount of Ti in effluent (3e450 mg/L, Fig. S8) and the considerable amounts in the solids (1943e2634 mg/kg, Table S1). Moreover, Ti contents in influent front of the column (2634 mg/kg) were greater than the effluent end (1943 mg/kg), indicating that a physical mechanism, such as straining, is mainly responsible for the TiO2 trap. 3.4. In situ column analysis with synchrotron techniques
m-XRF images showed the distribution of As, Ti, Fe, Mn, S, Zn, Ca, and Cu in the influent front of biotic column (Fig. 5). The images and correlation plot for the solid sample show a weak significant correlation between As and Ti (R ¼ 0.36, p < .01), which might be ascribed to the initial As injected including soluble As in suspension and adsorbed As on TiO2. There was no correlation between As and other elements (Fig. S9). Therefore, the majority of adsorbed As in the column was still associated with TiO2, which was mainly
766
T. Luo et al. / Environmental Pollution 234 (2018) 762e768
precipitates with As, such as orpiment and realgar, and soluble As-S complexes, such as di-and tri-thioarsenite monomers as well as dimeric and trimeric arsenic-sulfur soluble complex (Rochette et al., 2000; Couture et al., 2013; Stucker et al., 2014). Thioarsenate (H2AsO3S) and dithioarsenate (H2AsO2S 2 ) may form before As(V) was reduced to As(III) (Rochette et al., 2000). Moreover, the formation of thioarsenite (H3AsO2S) and dithioarsenite (H2AsOS 2 ) might be also occurred in the process of As(V) reduction by sulfide (Rochette et al., 2000). These soluble thio-As species, however, were not observed in the effluent of the three columns in this study. Meanwhile, both the thermodynamic pe-pH calculation (Fig. 2 and Table S2) and XANES analysis (Fig. 6) did not support the formation of realgar and orpiment under the experimental conditions. Fe(III) oxides can be reduced by the biogenic sulfide according to Eq. (1) (Poulton et al., 2004): 2Fe(OH)3 þ HS þ 5Hþ / 2Fe2þ þ S0 þ 6H2O
(1)
The produced Fe(II) can then form secondary Fe-minerals, such as pyrite, mackinawite, magnetite, and lepidocrocite (Kumar et al., 2016a,b; Burton et al., 2011; Kirk et al., 2010; Morse and Rickard, 2004). In agreement with our thermodynamic pe-pH calculation and XANES analysis, the formation of magnetite and mackinawite during the microbial sulfidogenesis has been reported in a series of iron-rich model systems (Kumar et al., 2016a,b; Kocar et al., 2010; Burton et al., 2011; Tufano and Fendorf, 2008). Previous studies indicate that the reaction of Fe(III) oxides and Fe(II) leads to magnetite formation according to Eq. (2) (Kocar et al., 2010; Saalfield and Bostick, 2009): Fe2þ þ 2Fe(OH)3 / Fe3O4 þ 2Hþ þ 2H2O SO2 4
Fig. 4. concentrations (open symbols) in filtered effluent of the three columns, and sulfide concentrations (closed symbols) in biotic column (A). Total Fe concentrations in unfiltered effluent of the three columns (open symbols), and Fe(II) concentrations (closed symbols) in filtered effluent of biotic column(B).
retained in the influent front of the soil column during the transport process. The speciation of As and Fe in biotic column was determined with XANES (Fig. 6). As(V) was the primary As species in biotic column as evidenced by its peak position at 11,874 eV (Fig. 6A) (Jing et al., 2005). As(V) (92%) and As(III) (8%) were the As species in the influent front, and no As(III) was detected in the effluent end of the biotic column (Fig. 6A, Fig. S10). As-S precipitates, such as realgar and orpiment, were not detected by XANES (Fig. 6A), in line with the pe-pH diagram predication (Fig. 2). In agreement to the pe-pH diagram, the Fe k-edge XANES analysis indicates the presence of magnetite in biotic column (Fig. 6B). The mackinawite, as a pyrite precursor, was also detected by XANES (Hunger and Benning, 2007), though it was not predicted by thermodynamic pe-pH diagram under the experimental conditions (Fig. 2). XANES results showed that magnetite (23.4%), mackinawite (21%), FeOOH (21.1%), Fe2O3 (17.7%), and FeO (16.8%) were the solid iron minerals in the influent front of the biotic column (Fig. S10, Table S4). In the effluent end, magnetite, mackinawite, FeOOH, FeO, and Fe2O3 accounted for 31.5%, 16.9%, 28.3%, 17.3%, and 6%, respectively, of the total iron minerals. 3.5. As redistribution and transport in soil The fate of As is implicitly tied to SO4 and Fe(III) reduction as mediated by SRB (Kumar et al., 2016a,b; Burton et al., 2011; Dhar et al., 2011). The biogenic sulfide can form insoluble As-S
(2)
As has a high adsorption affinity to magnetite, and therefore exhibits a strong preference for partitioning to magnetite in soil (Burton et al., 2011). Mackinawite has often been observed in the presence of microbial sulfidogenesis in an iron-rich system due to Eq. (3) (Kumar et al., 2016a,b; Kocar et al., 2010; Burton et al., 2011; Kirk et al., 2010; Morse and Rickard, 2004): Fe2þ þ HS / FeS þ Hþ
(3)
Mackinawite is also effective in immobilizing As (Kumar et al., 2016a,b; Kocar et al., 2010; Burton et al., 2011; Kirk et al., 2010). The formation of magnetite and mackinawite were observed as evidenced by XANES analysis (Fig. 6) and pe-pH diagram (Fig. 2). The color of the biotic column turned from brown to black at the end of the experiment (Fig. S11), further supporting the formation of Fe(II) minerals. In line with our results, Saalfield and Bostick (2009) found that microbial sulfidogenesis can mitigate As mobility due to the formation of secondary iron minerals. Interestingly, the total effluent As from biotic column was greater than that in abiotic control (Fig. 3). Actually, the dissolved As from biotic column was the sum of the As originally soluble in influent and those liberated from TiO2. The dissolved As attached to the in situ generated nanosized secondary iron minerals including magnetite and mackinawite (accounting for 44.4%e48.4% in solids of the biotic column, Fig. S12), and flew out of the column as particulate As. Hellige et al. (2012) also found that the reaction between sulfide and lepidocrocite can produce nanometer-scale magnetite. Kumar et al. (2016a,b) reported that nanocrystallite mackinawite was formed in the system of As-loaded zero-valent iron with the microbial sulfate reduction. The average ratio of As/Fe was 315 ± 137 mg/g in effluent in this study, which is comparable to
T. Luo et al. / Environmental Pollution 234 (2018) 762e768
767
Fig. 5. Spatial distributions of As, Ti, Fe, Mn, S, Zn, Ca, and Cu in solid sample which at influent front of biotic column. The relative concentrations of the elements are shown on the right in the colored bar scales in numbers of fluorescence count. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
Fig. 6. Normalized XANES at As (A) and Fe (B) K-edge for samples of biotic column. Solid line was experimental data, dotted line was linear combination fitting.
768
T. Luo et al. / Environmental Pollution 234 (2018) 762e768
the reported adsorption capacity of 214e334 mg/g for ferrihydrite from pH 4.6 to 9.2 (Raven et al., 1998; Dixit and Hering, 2003). 4. Conclusions The present study reveals the specific effects of SRB on the mobility of As and nano-TiO2 adsorbent in natural soil columns by excluding other confounding microbial processes. Our results indicate that SRB inhibit dissolved As release due to its re-adsorption on in situ formed secondary iron minerals including magnetite and mackinawite. Moreover, the SRB facilitate the migration of secondary iron minerals associated with As. The transport of these minerals might be due to their submicron or nanometer particle size, and therefore, lead to the As mobility and contamination in the subsurface. This finding provides insights on As biogeochemical cycling in the natural environment. Specifically, As mobility might not only be attributed to the dissolved As liberated from the solids, but also to its migration with in situ formed secondary iron minerals. Meanwhile, the nano-TiO2 released to the effluent was only 147 ± 92 mg/L with influent concentration 0.1 g/L. The nano-TiO2 with loaded As was mainly retained in the front of soil column. Therefore, the nano-TiO2 release from the soil and subsequently its the effect on As mobility were not significant as we initially thought. Acknowledgments We acknowledge the financial support of the National Basic Research Program of China (2015CB932003), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB14020201), and the National Natural Science Foundation of China (41373123, 41425016, and 21407124). The m-XRF was performed at SSRF BL15U, and XANES spectra were acquired at NSRRC BL01C1. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.envpol.2017.12.029. References Burton, E.D., Johnston, S.G., Bush, R.T., 2011. Microbial sulfidogenesis in ferrihydriterich environments: effects on iron mineralogy and arsenic mobility. Geochim. Cosmochim. Acta 75, 3072e3087. Burton, E.D., Johnston, S.G., Kocar, B.D., 2014. Arsenic mobility during flooding of contaminated soil: the effect of microbial sulfate reduction. Environ. Sci. Technol. 48, 13660e13667. Canfield, D.E., Thamdrup, B., Kristensen, E., 2005. The Sulfur Cycle. In Aquatic Geomicrobiology. Elsevier Academic Press, San Diego. €ger, D., Van Cappellen, P., 2013. Couture, R.M., Rose, J., Kumar, N., Mitchell, K., Wallschla Sorption of arsenite, arsenate, and thioarsenates to iron oxides and iron sulfides: a kinetic and spectroscopic investigation. Environ. Sci. Technol. 47, 5652e5659. €ger, D., Rose, J., Van Cappellen, P., 2014. Arsenic binding to Couture, R.M., Wallschla organic and inorganic sulfur species during microbial sulfate reduction; a sediment flow-through reactor experiment. Environ. Chem. 10, 285e294. Dhar, R.K., Zheng, Y., Saltikov, C.W., Radloff, K.A., Mailloux, B.J., Ahmed, K.M., van Geen, A., 2011. Microbes enhance mobility of arsenic in pleistocene aquifer sand from Bangladesh. Environ. Sci. Technol. 45, 2648e2654. Dixit, S., Hering, J.G., 2003. Comparison of arsenic(V) and arsenic(III) sorption onto iron oxide minerals: implications for arsenic mobility. Environ. Sci. Technol. 37, 4182e4189. Eaton, A.D., Clesceri, L.S., Greenberg, A.E., 1995. Stand Methods for the Examination of Water and Wastewater, nineteenth ed. American Public Health Association, Washington DC. Gherna, R., Pienta, P., Cote, R., 1992. American Type Culture Collection Catalogue of Bacteria and Phages, eighteenth ed. (Rockville, MD). Guo, H., Liu, Z., Ding, S., Hao, C., Xiu, W., Hou, W., 2015. Arsenate reduction and mobilization in the presence of indigenous aerobic bacteria obtained from high arsenic aquifers of the Hetao basin, Inner Mongolia. Environ. Pollut. 203, 50e59. Hellige, K., Pollok, K., Larese-Casanova, P., Behrends, T., Peiffer, S., 2012. Pathways of ferrous iron mineral formation upon sulfidation of lepidocrocite surfaces. Geochim. Cosmochim. Acta 81, 69e81.
Huang, J.H., Elzinga, E.J., Brechbuehl, Y., Voegelin, A., Kretzschmar, R., 2011. Impacts of shewanella putrefaciens strain CN-32 cells and extracellular polymeric substances on the sorption of As(V) and As(III) on Fe(III)-(Hydr)oxides. Environ. Sci. Technol. 45, 2804e2810. Hunger, S., Benning, L.G., 2007. Greigite: a true intermediate on the polysulfide pathway to pyrite. Geochem. Trans. 8, 1. Jing, C., Liu, S., Meng, X., 2005. Arsenic leachability and speciation in cement immobilized water treatment sludge. Chemosphere 59, 1241e1247. Jing, C., Meng, X., Calvache, E., Jiang, G., 2009. Remediation of organic and inorganic arsenic contaminated groundwater using a nanocrystalline TiO2-based adsorbent. Environ. Pollut. 157, 2514e2519. Kirk, M.F., Roden, E.E., Crossey, L.J., Brearley, A.J., Spilde, M.N., 2010. Experimental analysis of arsenic precipitation during microbial sulfate and iron reduction in model aquifer sediment reactors. Geochim. Cosmochim. Acta 74, 2538e2555. Kocar, B.D., Borch, T., Fendorf, S., 2010. Arsenic repartitioning during biogenic sulfidization and transformation of ferrihydrite. Geochim. Cosmochim. Acta 74, 980e994. Kocar, B.D., Herbel, M.J., Tufano, K.J., Fendorf, S., 2006. Contrasting effects of dissimilatory iron(III) and arsenic(V) reduction on arsenic retention and transport. Environ. Sci. Technol. 40, 6715e6721. Kumar, N., Couture, R.M., Millot, R., Battaglia-Brunet, F., Rose, J., 2016a. Microbial sulfate reduction enhances arsenic mobility downstream of zerovalent-ironbased permeable reactive barrier. Environ. Sci. Technol. 50, 7610e7617. Kumar, N., Millot, R., Battaglia-Brunet, F., Omoregie, E., Chaurand, P., Borschneck, D., Bastiaens, L., Rose, J., 2016b. Microbial and mineral evolution in zero valent iron-based permeable reactive barriers during long-term operations. Environ. Sci. Pollut. Res. 23, 5960e5968. Lowers, H.A., Breit, G.N., Foster, A.L., Whitney, J., Yount, J., Uddin, N., Muneem, A., 2007. Arsenic incorporation into authigenic pyrite, bengal basin sediment. Bangladesh. Geochim. Cosmochim. Acta 71, 2699e2717. Luo, T., Cui, J., Hu, S., Huang, Y., Jing, C., 2010. Arsenic removal and recovery from copper smelting wastewater using TiO2. Environ. Sci. Technol. 44, 9094e9098. Luo, T., Hu, S., Cui, J., Tian, H., Jing, C., 2012. Comparison of arsenic geochemical evolution in the Datong basin (Shanxi) and Hetao basin (Inner Mongolia), China. Appl. Geochem. 27, 2315e2323. Luo, T., Tian, H., Guo, Z., Zhuang, G., Jing, C., 2013. Fate of arsenate adsorbed on nano-TiO2 in the presence of sulfate reducing bacteria. Environ. Sci. Technol. 47, 10939e10946. Luo, T., Ye, L., Ding, C., Yan, J., Jing, C., 2017. Reduction of adsorbed As(V) on nanoTiO2 by sulfate-reducing bacteria. Sci. Total Environ. 598, 839e846. Morse, J.W., Rickard, D., 2004. Chemical dynamics of sedimentary acid volatile sulfide. Environ. Sci. Technol. 38, 131Ae136A. Pena, M., Meng, X.G., Korfiatis, G.P., Jing, C., 2006. Adsorption mechanism of arsenic on nanocrystalline titanium dioxide. Environ. Sci. Technol. 40, 1257e1262. Polizzotto, M.L., Harvey, C.F., Sutton, S.R., Fendorf, S., 2005. Processes conducive to the release and transport of arsenic into aquifers of Bangladesh. Proc. Natl. Acad. Sci. U. S. A 102, 18819e18823. Poulton, S.W., Krom, M.D., Raiswell, R., 2004. A revised scheme for the reactivity of iron (oxyhydr)oxide minerals towards dissolved sulfide. Geochim. Cosmochim. Acta 68, 3703e3715. Raven, K.P., Jain, A., Loeppert, R.H., 1998. Arsenite and arsenate adsorption on ferrihydrite: kinetics, equilibrium, and adsorption envelopes. Environ. Sci. Technol. 32, 344e349. Ravel, B., Newville, M., 2005. Athena, artemis, hephaestus: data analysis for X-ray absorption spectroscopy using IFEFFIT. J. Synchrotron Radiat. 12, 537e541. Rochette, E.A., Bostick, B.C., Li, G., Fendorf, S., 2000. Kinetics of arsenate reduction by dissolved sulfide. Environ. Sci. Technol. 34, 4714e4720. Saalfield, S.L., Bostick, B.C., 2009. Changes in iron, sulfur, and arsenic speciation associated with bacterial sulfate reduction in ferrihydrite-rich systems. Environ. Sci. Technol. 43, 8787e8793. Smedley, P.L., Kinniburgh, D.G., 2002. A review of the source, behaviour and distribution of arsenic in natural waters. Appl. Geochem. 17, 517e568. Stucker, V.K., Silverman, D.R., Williams, K.H., Sharp, J.O., Ranville, J.F., 2014. Thioarsenic species associated with increased arsenic release during biostimulated subsurface sulfate reduction. Environ. Sci. Technol. 48, 13367e13375. Sun, J., Steven, N., Brian, C., Mailloux, J., Bostick, B.C., 2016. Arsenic mobilization from sediments in microcosms under sulfate reduction. Chemosphere 153, 254e261. Tufano, K.J., Fendorf, S., 2008. Confounding impacts of iron reduction on arsenic retention. Environ. Sci. Technol. 42, 4777e4783. Wolthers, M., Charlet, L., van der Weijden, C.H., van der Linde, P.R., Rickard, D., 2005. Arsenic mobility in the ambient sulfidic environment: sorption of arsenic(V) and arsenic(III) onto disordered mackinawite. Geochim. Cosmochim. Acta 69, 3483e3492. Yan, L., Huang, Y., Cui, J., Jing, C., 2015. Simultaneous As(III) and Cd removal from copper smelting wastewater using granular TiO2 columns. Water Res. 68, 572e579. Yan, W., Wang, H., Jing, C., 2016. Adhesion of shewanella oneidensis MR-1 to goethite: a two-dimensional correlation spectroscopic study. Environ. Sci. Technol. 50, 4343e4349. Ye, L., Liu, W., Shi, Q., Jing, C., 2017. Arsenic mobilization in spent nZVI waste residue: effect of Pantoea sp.IMH. Environ. Pollut. 230, 1081e1089. Zhang, D., Guo, H., Xiu, W., Ni, P., Zheng, H., Wei, C., 2017. In-situ mobilization and transformation of iron oxides-adsorbed arsenate in natural groundwater. J. Hazard Mater. 321, 228e237. Zhang, X., Sun, H., Zhang, Z., Niu, Q., Chen, Y., Crittenden, J.C., 2007. Enhanced bioaccumulation of cadmium in carp in the presence of titanium dioxide nanoparticles. Chemosphere 67, 160e166.