Journal of Hydrology 519 (2014) 414–422

Contents lists available at ScienceDirect

Journal of Hydrology

journal homepage: www.elsevier .com/locate / jhydrol

Coupled iron, sulfur and carbon isotope evidences for arsenic enrichmentin groundwater

http://dx.doi.org/10.1016/j.jhydrol.2014.07.0280022-1694/� 2014 Elsevier B.V. All rights reserved.

⇑ Corresponding authors. Tel.: +86 27 67883170; fax: +86 27 87481030 (X. Xie).Tel.: +86 27 67883998; fax: +86 27 87481030 (Y. Wang).

E-mail addresses: yx.wang@cug.edu.cn (Y. Wang), xjxie@cug.edu.cn (X. Xie).

Yanxin Wang a,⇑, Xianjun Xie a,⇑, Thomas M. Johnson b, Craig C. Lundstrom b, Andre Ellis c, Xiangli Wang b,Mengyu Duan a, Junxia Li a

a School of Environmental Studies & State Key Laboratory of Biogeology and Environmental Geology, China University of Geosciences, 430074 Wuhan, Chinab Department of Geology, University of Illinois at Urbana-Champaign, 1301 W. Green Street, Urbana, IL 61801, USAc Department of Geological Sciences, California State University, Los Angeles, CA 90032, USA

a r t i c l e i n f o

Article history:Received 7 February 2014Received in revised form 22 June 2014Accepted 12 July 2014Available online 29 July 2014This manuscript was handled by LaurentCharlet, Editor-in-Chief, with the assistanceof Pedro J. Depetris, Associate Editor

Keywords:Iron isotopeCarbon isotopeSulfur isotopeArsenicBiogeochemistryGroundwater

s u m m a r y

It is generally accepted that microbial processes play a key role in the mobilization and enrichment ofarsenic (As) in groundwater. However, the detailed mechanism of the metabolic processes remain poorlyunderstand. We apply isotopic measurements of iron (d56Fe vs. IRMM-14), sulfur (d34SSO4 vs. V-CDT) andcarbon (d13CDIC vs. V-PDB) to an experimental field plot in the Datong Basin, northern China. An array ofmonitoring wells was installed in a �1700-m2 plot in which high concentrations of As, ranging from 4.76to 469.5 lg/L, were detected in the groundwater. The measured range of d34SSO4 values from 10.0‰ to24.7‰ indicates the prevalence of microbial sulfate reduction within aquifers. The range of d56Fe valuesmeasured in the groundwater suggests microbial Fe(III) reduction and the occurrence of isotopicexchange between Fe(II)aq and FeS precipitation. The low d13CDIC values (up to �33.6‰) measured ingroundwater are evidences for the microbial oxidation of organic matter, which is interpreted as the lightcarbon pool within the aquifer sediments. The high As (As > 50 lg/L) groundwater, which has higherd34SSO4 and d56Fe values and lower d13C values, indicates the following: (1) microbial reduction of sulfatecauses the mobilization of As through HS� abiotic reduction of Fe(III) minerals and/or formation of As-sulfur components; and (2) direct microbial reduction of Fe(III) oxides, hydroxides and oxyhydroxidescannot increase As concentrations to greater than 50 lg/L. Re-oxidation of Fe-sufide explains how sampleC1-2 can have a high As concentration and low d34SSO4 and high d56Fe values. The results provide newinsight into the mechanism of As enrichment in groundwater.

� 2014 Elsevier B.V. All rights reserved.

1. Introduction

Arsenic (As) contamination of groundwater in sedimentaryaquifers is a major global public health crisis. Natural occurrencesof high As groundwater have been reported in several geologicallyyoung aquifers around the world, including the Ganges Plain (Indiaand Bangladesh) (Harvey et al., 2002; Islam et al., 2004; Nicksonet al., 2000), the Red River Delta (Vietnam) (Berg et al., 2007,2008; Postma et al., 2007), the Mekong River Basin (Cambodia)(Buschmann et al., 2008; Papacostas et al., 2008; Thi et al., 2010)and the Hetao Basin in China (Guo et al., 2008a,b). Similar condi-tions are also present in the Datong Basin of Shanxi Province, northChina (Wang et al., 2009; Xie et al., 2009). Arsenic concentrationsas high as 1820 lg/L have been detected in groundwater of the

Datong Basin (Xie et al., 2008), which greatly exceeds the maxi-mum limited value (10 lg/L) for drinking water recommendedby the WHO (WHO, 2004).

Aquifers with natural As contamination are characterized by Asconcentrations that vary significantly over short lateral andvertical distances (Kirk et al., 2004). Studies of As-contaminatedaquifers around the world demonstrate that As concentrations ingroundwater are controlled by complex sets of conditions and bio-geochemical processes (Appelo et al., 2002; Bose and Sharma,2002; Islam et al., 2004; Kirk et al., 2004; Nickson et al., 2000;Oremland and Stolz, 2003; Polizzotto et al., 2008). Arsenic that issorbed and sequestered on Fe oxides/hydroxides is one of the mostcommon As reservoirs in some sedimentary basins. The reductionof As(V) to As(III) results in desorption or the formation of weaklysorbing complexes. Changes in pH and redox conditions mayinduce the dissolution of Fe-oxides, hydroxides and oxyhydroxidesand/or the release of sorbed As from the mineral surfaces (Nicksonet al., 1998, 2000; Swartz et al., 2004). And the formation of

Y. Wang et al. / Journal of Hydrology 519 (2014) 414–422 415

As-sulfur compounds in high HS� concentration water promote themobilization of As (Planer-Friedrich et al., 2007; Stauder et al.,2005; Bostick et al., 2005). In addition, biotic processes can playimportant roles by either directly mediating redox reactions orchanging the redox conditions in the aquifer (Oremland andStolz, 2003). More recently, Polizzotto et al. (2008) proposed thatsurface hydrological processes and groundwater flow can signifi-cantly promote the enrichment of As in groundwater. Accordingly,the mechanism of As release may vary with location depending onthe hydrogeological conditions and (bio)geochemical processes.This variation can be accounted for by the distribution of As con-centrations in natural As-contaminated aquifers, which is crucialto understanding the generation of high As groundwater. However,this explanation is complicated by heterogeneities in sedimentlithology and variations in the hydrogeological conditions and(bio)geochemical processes within the aquifers. Therefore, thecomplex (bio)geochemical processes should be investigated byanalyzing the different hydrogeochemical constituents (includingthe isotopic compositions) of groundwater in a high As aquifer sys-tem in a relatively homogenous hydrogeological setting.

Hydrogeochemical studies of a field experimental plot(20 m � 85 m) in the Datong Basin of Shanxi Province, China mayprovide a clear understanding of the complex (bio)geochemicalprocesses that cause As mobilization. The experimental plot islocated in a groundwater discharge zone in the center of the basinand adjacent to a river. The groundwater at this site moves fromsouth to north toward the river (Fig. 1). Because the area is far frommountains, no significant mixing of different recharge waters isexpected at the study site. Studies of sediment cores from theexperimental plot indicate that the sediment consist of homoge-nous fine sands and silts. Three sand layers are present at depths

Fig. 1. Location of the experimental plot. The monitoring wells are indicated. The dire

less than 30 m below the surface. Arsenic is associated with Fe inthe sediments at this site, and the Fe in the sediments is mainlypresent as crystalline Fe(III) oxides/hydroxides (Xie et al., 2013a).Microbial Fe(III) and SO4

2� reduction is thought to be the control-ling process that causes the mobilization of As within the aquifers(Xie et al., 2013a). This interpretation is supported by the low ORP(oxidation and reduction potential) values and the high concentra-tions of Fe(II) and HS� in the groundwater (Xie et al., 2013b). How-ever, the detailed metabolic mechanisms of Fe and S and the effectson As mobilization are not clear. The relative homogeneity of thesediments and the simple hydrogeological conditions at the siteprovide excellent conditions for investigating the (bio)geochemicaleffects on As mobility.

The (bio)geochemical cycles of As are closely linked with S andFe (Fisher et al., 2008; Oremland and Stolz, 2003). Sulfate-reducingbacteria use sulfate (SO4

2�) as a terminal electron acceptor to oxi-dize organic matter and produce sulfide (HS�) (Postgate, 1984).Sulfur isotopes of sulfate (d34SSO4) are fractionated during SO4

2�

reduction, and the residual SO42� is enriched in the heavier isotope

(Böttcher and Thamdrup, 2001; Canfield, 2001; Detmers et al.,2001; Robertson and Schiff, 1994). Similarly, Fe(III)-reducing bac-teria oxidize organic matter by using ferric oxide minerals as ter-minal electron acceptors (Canfield et al., 1993; Lovley, 1997;Raiswell and Canfield, 1998) to produce dissolved Fe(II). In naturalenvironments, an Fe isotopic fingerprint appears to be associatedwith microbial Fe reduction in the form of low d56Fe values foraqueous Fe(II) (Beard et al., 2003; Percak-Dennett et al., 2011).The geochemical cycles of Fe and S are strongly linked, so studiesof Fe and S isotope systems should provide a unique view of themetabolic processes of Fe and S. Sulfur and Fe isotopes are there-fore useful tools for tracing biogeochemical processes in the

ction of groundwater flow from south to north is noted by the arrow on the map.

416 Y. Wang et al. / Journal of Hydrology 519 (2014) 414–422

near-surface environment (Beard et al., 2003; Gibson et al., 2011;Herbert and Schippers, 2008; Johnson et al., 2008a). The naturalvariations of Fe isotopes caused by abiotic or biological processesare an order of magnitude smaller than those for S. Fortunately,analytical methods have been developed to measure Fe isotopeswith precisions that are suitable for evaluating naturally occurring,mass-dependent isotope variations (Beard and Johnson, 1999). Inthis study, we report Fe and S isotopic data of groundwater froma known As-contaminated site. The isotopic data provide strongevidence that biogeochemical cycling of Fe and S significantlyaffects the enrichment of As in the groundwater. Our approach ofdemonstrating the enrichment of As in groundwater may be appli-cable to other regions with similar conditions.

2. Materials and methods

2.1. Sampling

To identify the complex biogeochemical cycling of Fe and Swithin the aquifers, an experimental plot was constructed at aknown natural As-contaminated site in the Datong Basin, northernChina. The experimental plot includes ten monitoring wells thatwere installed within the shallow aquifers, which are composedof alluvial sands. All of the monitoring wells were constructedusing PVC pipe (5 cm diameter) and screened beneath the watertable at the appropriate depths (ca. 10, 15, and 20 m below theground surface). Thirty water samples were collected from theexperimental plot for isotopic and hydrochemical analyses inAugust 2011. A map of the site with the wells indicated is shownin Fig. 1. Groundwater was collected using a peristaltic pump afterpurging the monitoring wells for 5 to 10 min. Chemical and phys-ical parameters, such as pH, ORP and temperature (T), were mea-sured using Hach Instruments portable meters. Two filtered(<0.45 lm) acidified samples (acidified to pH < 2 using ultra-pureHNO3) were collected in 50 mL HDPE bottles for the analysis ofAs and Fe concentrations and Fe isotopic ratios in the laboratory.The groundwater samples for anion concentration includingSO4

2� were stored in 50 mL HDPE bottles after filtration. Samplesfor the d34SSO4 isotope analyses were collected in 500 mL HDPEbottles after filtration, and SO4

2� was precipitated as BaSO4 imme-diately. The unfiltered d13CDIC samples were collected in 50 mlglass bottles, closed without head space using stoppers and storedin the dark until the analysis.

2.2. Analytical methodology

The concentration of HS� was determined using a spectropho-tometer (HACH, DR2800) with methylene blue assay. The con-centrations of anion in groundwater ere determined using ionchromatography (IC) (Metrohm 761 Compact IC). The trace metalion concentrations, including As and Fe, were measured byinductively coupled plasma mass spectrometry (ICP-MS) (PerkinElmer, ELAN DRC-e) at the School of Environmental Studies atthe China University of Geosciences in Wuhan. Random dupli-cates of all of the analyses were within 5% for the analyzedanion and trace elements, and the field and laboratory blankswere below the detection limits for the major and trace hydro-chemical components.

For the stable carbon isotope analysis of the DIC, the CO32� was

precipitated as BaCO3 using Ba(OH)2. The dried BaCO3 was thenconverted to CO2 at 90�C in an automated acid bath preparationsystem using H3PO4 and was measured using an isotope ratio massspectrometry (Thermo Fisher Scientific, MAT 253) at the Environ-mental Isotope Laboratory at the Institute of Karst Geology, CAGSin Guilin. The d13C value was calculated using the equation:

d ¼ Rsample

Rstandard� 1

� �� 1000 ð1Þ

where (R = 13C/12C). The carbon isotopic compositions are reportedin standard d notation, which represents the per mil deviations fromthe V-PDB standard. Replicate analyses of BaCO3 were reproducedto within ±0.2‰ (1r).

Sulfate was precipitated as BaSO4 with BaCl2. The precipitatewas recovered by centrifugation, carefully washed and dried priorto the isotope analyses. The isotope analysis of S was performedusing an isotope ratio mass spectrometer (Thermo Fisher Scientific,Delta PlusXL) coupled with an elemental analyzer (Costech) afterthe complete conversion of BaSO4 to SO2 via high-temperaturecombustion. The S isotope ratios are reported in per mil (‰) unitsas the conventional delta-notation (d). The d34S value was calcu-lated using Eq. (1) (R = 34S/32S). The d34SSO4 values are reported rel-ative to the V-CDT standard. The reproducibility of the analyses ofBaSO4 was better than 0.15‰ (1r) for d34SSO4. The measurementsof d34SSO4 were performed at the Environmental Isotope Laboratoryat the University of Arizona.

The acidified water samples for the Fe isotopic compositionmeasurements were dried and suspended with 0.4 mL 8 N HCl.The solution was then passed through an HCl-conditioned anionexchange resin (Bio-Rad, AG1-X8) for Fe purification. The matrixelements were removed by washing with 8 N HCl. Fe was elutedusing 0.5 N HCl and H2O followed by 8 N HNO3 and H2O. After sep-aration, the purified Fe fraction was evaporated to dryness and dis-solved using 2% HNO3 for the isotope analysis. After adjusting theFe concentrations in all of the samples and standards to 1.0 ppm,the Fe isotope ratio measurements were made with a multi-collec-tor inductively coupled plasma mass spectrometer (MC-ICP-MS)(Nu Instrument, Nu Plasma) equipped with a Cetac ASX-110 auto-matic sampler and a DSN-100 Desolvating Nebulizer System at theDepartment of Geology at the University of Illinois at Urbana-Champaign (UIUC).

Procedural blanks were prepared and were found to contributeto less than 0.1% of the Fe content in the samples. The Fe isotopicratios were calculated using Eq. (1) (R = 56Fe/54Fe) and are reportedin delta notation. The measured 56Fe/54Fe isotope ratio for IRMM-014 is used as the standard. Instrument mass bias is correctedfor using the sample-standard bracketing approach. Chromiumwas monitored at mass 53 to correct the interference of 54Cr on54Fe. The 57Fe/54Fe and 56Fe/54Fe ratios were measured simulta-neously, and all of the data plotted on the theoretical mass frac-tionation line (Young et al., 2002), demonstrating the absence ofisobaric interferences after correction. The average d56Fe value of77 analyses of the in-house standard (UIFe) was 0.70 ± 0.04‰.The average external precision of d56Fe of the groundwater sam-ples is approximately 0.05‰ (2SD).

3. Results

The As and Fe concentrations of the groundwater samples fromthe experimental plot range from 4.76 to 469.5 lg/L (with an aver-age value of 75.37 lg/L) and from 192.0 to 1168 mg/L (with anaverage value of 528.1 mg/L) (Table 1), respectively. Unlike previ-ous studies conducted by Nickson et al. (2000) in Bangladesh andWest Bengal, India, an apparent inverse correlation was observedbetween dissolved Fe and As (r2 = 0.78, a = 0.05) (Fig. 2A). The con-centration of SO4

2� varies between 45 mg/L and 3259 mg/L with anaverage value of 715 mg/L. It is interested to note that the concen-trations of SO4

2� in high As groundwater (As > 50 lg/L) is relativelower than those of low As groundwater (As < 50 lg/L) (Table 1).The groundwater pH ranges from 7.14 to 7.82, indicating typicalweak-alkaline conditions (Table 1). The ORP values of the ground-water vary between �36 and �255 mV, indicating the presence of

Table 1Iron and arsenic concentrations and iron, carbon and sulfur isotopic compositions of groundwater from the experimental plot, Datong Basin (high As groundwater: As > 50 lg/L;low As groundwater: As < 50 lg/L).

Sample ID pH ORP(mv) Fe (lg/L) As (lg/L) HS� (lg/L) SO42� (mg/L) d34S(SO4) ‰ d13C (DIC) ‰ d56Fe ‰

Low as groundwaterD2-2 7.47 �211 739.5 4.76 1 1160 10.8 – �1.55C2-2 7.67 �36 394.5 6.21 3 430 – �13.7 �0.37B2-2 7.33 �174 1013 6.48 4 881 11.5 �19.4 0.3A2-1 7.29 �195 828 8.27 – 577 – – –E1-2 7.30 �253 1169 8.42 22 – 10 �7.2 �0.96D2-1 7.41 �209 435 9.84 22 350 12.7 – 0.4D1-2 7.33 �216 601.5 11.0 12 2174 10.3 �6.4 0.35D1-1 7.18 �208 735 11.1 5 3259 10.1 �4.8 0.58E1-1 7.20 �248 693 11.8 14 1196 10.2 – 0.48B2-1 7.25 �194 708 11.9 6 487 12.6 �25.6 �0.52B1-2 7.20 �199 619.5 12 8 861 11.3 �21.5 0.37C2-1 7.22 �124 676.5 12.4 17 861 11.4 �22 0.19E2-1 7.18 �212 783 12.9 5 – 10 �4.9 0.02E1-3 7.43 �255 465 13.1 7 204 14.7 �17 0.85B1-1 7.20 �167 588 14.5 3 1258 10.9 �14.5 0.04A1-2 7.27 �211 862.5 14.5 9 1846 10.4 �16.3 0.21A2-2 7.14 �161 903 15.3 – 917 – – –E2-2 7.50 �218 385.5 38.1 4 592 – �6.9 1.22D2-3 7.52 �179 315 38.3 0 213 12.4 – 0.51C2-3 7.82 �64 300 41.6 – 293 – – –

High as groundwaterC1-2 7.55 �209 489 63 4 1192 11 �24.7 0.84B1-3 7.68 �196 250.5 81.6 5 149 17.6 – 0.53A2-3 7.60 �194 243 111 – 175 – – –A1-1 7.66 �174 205.5 114 3 241 18.5 �32.7 0.67E2-3 7.64 �207 213 118 4 49 – �17.9 0.35A1-3 7.70 �171 235.5 132 3 257 19 �20.9 0.71B2-3 7.62 �182 228 159 2 125 24.7 �30.4 0.36C1-3 7.77 �200 226.5 287 3 116 24.1 �33.6 0.23C1-1 7.74 �204 192 299 2 102 – �33.4 0.23D1-3 7.80 �225 204 470 9 45 – – 0.25

– Not determined.

Y. Wang et al. / Journal of Hydrology 519 (2014) 414–422 417

a reducing environment within the aquifer system. No significantdifference in ORP and pH was found between the high and lowAs groundwater (Table 1). The measured Fe isotopic compositions(d56Fe) of the groundwater range from �1.55‰ to 1.22‰ (Table 1)and have a mean d56Fe value of 0.24‰ (n = 26), which is compara-ble to the Fe isotopic values (d57Fe values vary between �1.2‰ and+1.5‰) of the reduced groundwater from Hanoi, Vietnam thatwere analyzed by Teutsch et al. (2004). It is worth noting thatthe high As groundwater has heavier mean d56Fe values than thelow As groundwater. The mean d56Fe values for the high and lowAs groundwater are 0.46‰ and 0.12‰, respectively. It is interestingto note that the Fe isotopic composition of the groundwater is neg-atively correlated with the dissolved Fe concentration in the low Asgroundwater (r2 = 0.4, a = 0.05) (Fig. 2B). Other than samples E2-2and D2-3, a positive correlation is observed between the d56Fe val-ues and the As concentrations in the low As groundwater (r2 = 0.58,a = 0.05) (Fig. 3A). For the high As groundwater, gradually isotopi-cally-lighter d56Fe values are observed with increasing As concen-trations (Fig. 3B). The S isotopic compositions (d34SSO4) of thedissolved SO4

2� range from 10.0‰ to 24.7‰ (Table 1), which is lessthan the measured d34SSO4 values for sulfate in our previous study(Xie et al., 2009). The d34SSO4 values vary between 10.0‰ and14.7‰ and have an average of 11.3‰ (n = 15) in the low As ground-water. In comparison, the high As groundwater has higher d34SSO4

values that range from 11.0‰ to 24.7‰ and have a mean of 19.2‰

(n = 6). The carbon isotope compositions (d13C) of dissolved inor-ganic carbon (DIC) have a wide range from �33.59‰ to �4.81‰

and a mean of �18.69‰. All of the high As groundwater sampleshave d13C values between �33.59‰ and �17.94‰, whereas thelow As groundwater has distinctly higher d13C values between�25.63‰ and �4.81‰ (Table 1).

4. Discussion

4.1. Carbon, sulfur and iron isotope systems in groundwater

Isotopic fractionation has been observed between Fe(II) andFe(III) in both biotic and abiotic systems (Johnson et al., 2008a,b;Rouxel et al., 2008). The biological reduction of Fe(III) mineralsforms several secondary phases, including FeCO3, and FeS, underparticular geochemical conditions in the subsurface environment.In addition, the aqueous Fe(II) produced by Fe(III) reduction canbe adsorbed onto the Fe(III) minerals. However, different biologicalreduction pathways and chemical adsorption may leave uniqueiron isotope ‘‘fingerprints’’ on the resulting Fe(II) pool and theresidual Fe(III) phases. According to Johnson and Beard (2005),the formation of biogenic magnetite and Fe-carbonate duringreduction of microbial Fe(III) oxides suggests equilibrium fraction-ation factors of �1.3‰, 0.0‰, and +0.9‰ for the Fe(II)aq-magne-tite, Fe(II)aq-siderite, and Fe(II)aq-ankerite fractions, respectively,at room temperature. Fe sulfide precipitation favors the light iso-tope with D56FeFe(II)aq-FeS values between 0.3‰ and 0.85‰ (Butleret al., 2005). In the presence of Fe-reducing bacteria, the d56Fe val-ues of Fe(II)aq are approximately 1.3‰ lower than the Fe(III) sub-strates such as ferrihydrite, hematite, and goethite (Beard et al.,2003). Due to the fluctuation of redox conditions, the producedFe(II)aq can be re-oxidized and precipitate as secondary minerals,such as Fe(II) hydroxides and Fe(III) hydroxides, that do not causesignificant isotopic fractionation between Fe(II)aq and Fe(III) pre-cipitation compared with biotic processes (Bullen et al., 2001).Crosby et al. (2005) indicated that adsorption cannot result in sig-nificant isotopic fractionation between aqueous Fe(II) and Fe(III)oxide. Johnson et al. (2005) also concluded that sorbed Fe(II) has

Fig. 2. The relationships between the Fe concentration and the As concentration (A)and the d56Fe values (B) in groundwater from the experimental plot, Datong basin.Note the overall inverse correlation and negative correlation between dissolved Feand As concentrations and the iron isotopic compositions of the groundwater,respectively, which may be related to the formation of iron sulfide during microbialreduction of ferric oxides/hydroxides and sulfate.

Fig. 3. The relationship between dissolved As and the d56Fe values in groundwater.The positive correlation between dissolved As and the d56Fe values in the low Asgroundwater can be attributed to the reduction of ferric oxides/hydroxides. Theslightly isotopically-lighter d56Fe in the high As groundwater may be due to Fe(II)aqbeing in contact with FeS under strong reducing conditions, which can result inlight iron isotopic compositions of Fe(II)aq.

418 Y. Wang et al. / Journal of Hydrology 519 (2014) 414–422

Fe isotope compositions that are similar to those of aqueous Fe(II)at equilibrium. In this study, therefore, the large Fe isotopevariations observed in the groundwater indicate widespread redoxcycling of Fe catalyzed by microbial Fe reduction within theaquifers. The prevailing consensus is that microbial reductive dis-solution of As-associated Fe(III) oxides, hydroxides and oxyhydrox-ides is the main process that controls the level of As ingroundwater (Nickson et al., 2000; Smedley and Kinniburgh,2002). In the studied aquifers, 56Fe-depleted Fe(II)aq caused bymicrobial Fe(III) reduction would have to occur in the early stageof the reduction reaction. As the Fe(II)aq becomes 56Fe-depleted,the residual Fe(III) will be enriched in the heavier isotope (Rayleighdistillation); therefore, the corresponding Fe(II)aq from this 56Fe-enriched Fe(III) pool will also be enriched in 56Fe relative to theFe(II)aq that formed in the early stage of the reduction reaction.Thus, microbial reductive dissolution of As-associated Fe(III) oxi-des/hydroxides (Xie et al., 2013a) can result in increased As con-centrations with a gradual increase of d56Fe values. A positivecorrelation between dissolved As and Fe and d56Fe values shouldthus be expected in groundwater. The observed positive correla-tion between the As and the d56Fe values in low As groundwater(Fig. 3A) indicates As release accompanying the reduction of micro-bial Fe(III) oxides/hydroxides. However, such a positive correlationbetween dissolved As and d56Fe values was not observed in thehigh As groundwater (Fig. 3B). In contrast, the d56Fe values

gradually decrease with increasing As concentrations. An experi-mental study of Fe isotope fractionation during precipitation ofFeS indicated that the kinetic Fe isotope fractionation zero-ageFeS was D56FeFe(II)aq-FeS = 0.85‰ and that the FeS in contact withFe(II)aq became progressively isotopically heavier (Butler et al.,2005). Therefore, the formation of Fe(II) sulfide in the sedimentincorporates the lighter isotope of Fe and result in the enrichmentof heavy isotope composition in residual Fe(II)aq. And then, thed56FeFe(II)aq values in high As groundwater gradually decreaseddue to the interaction between Fe(II) sulfide and Fe(II)aq. The for-mation of Fe(II) sulfide can be supported by the strong H2S odorand the high HS� concentrations in the groundwater (Table 1). Fur-thermore, FeS has been detected in the aquifer sediments from thissite (Xie et al., 2014). Accordingly, the gradually lighter d56Fe val-ues of the high As groundwater and the inverse relation betweend56Fe and As concentration (Fig. 3B) can be attributed to the forma-tion of FeS and interaction between Fe(II)aq and the produced FeSprecipitation.

According to the study conducted by Yao and Millero (1996),sulfide can directly react with Fe(III) oxides, hydroxides and oxyhy-droxides to abiotically reduce Fe(III) to Fe(II) due to high concen-trations of HS� in water. However, no significant isotopicfractionation can be expected between Fe(II)aq and the residualFe(III) minerals in this process. Therefore, the abiotic reduction ofAs-bearing Fe(III) oxides/hydroxides by HS� can partially explain

Fig. 5. The relationships between the dissolved As and the d34SSO4 (A) and d13CDIC

values (B) in groundwater. Positive and negative correlations are observed betweenthe dissolved As and the d34SSO4 and d13CDIC values, respectively, which reflect theimpacts of the microbial oxidation of organic carbon and the reduction of sulfate onthe mobilization of arsenic.

Y. Wang et al. / Journal of Hydrology 519 (2014) 414–422 419

the enrichment of As in groundwater, especially for groundwaterwith As concentrations greater than 50 lg/L. In addition, manystudies (Planer-Friedrich et al., 2007; Stauder et al., 2005; Bosticket al., 2005) demonstrated that soluble As-sulfur components suchas thioarsenates or thioarsnites are the important As species inneutral to alkaline sulfide-rich water. Accordingly, high HS� con-centrations and weak alkaline conditions indicate that these As-sulfur components were expected occurrence in the groundwaterin the studied aquifers. More importantly, thioarsenates or thioars-nites is more stable than As oxyanions in the high concentrationsof HS� groundwater. Thus, apart from As-bearing Fe(III) oxides/hydroxides reduction, the formation of thioarsnates or thioarsnitesunder high HS� concentration condition may be also an importantcontributor for the enrichment of As in groundwater at this site.

Studies of the geochemical behavior of S isotopes (d34SSO4) havecontributed greatly to the understanding of the biogeochemistry ofsurficial or near surficial environments over the last several dec-ades (Canfield, 2001; Xie et al., 2009). The fractionation factors ofmicrobial SO4

2� reduction range from�46.0‰ to 19.0‰ dependingon bacteria populations, electron-donor type and concentrationand metabolic pathways (Böttcher and Thamdrup, 2001;Canfield, 2001; Detmers et al., 2001). Small positive S isotopeenrichments have been documented for the biological oxidationof dissolved HS� (Canfield, 2001; Taylor et al., 1984). The S isotopiccompositions (d34SSO4) of the dissolved SO4

2� analyzed in thisstudy range from 10.0‰ to 24.7‰. The d34SSO4 values vary slightlyand cluster around an average of 11.3‰ (n = 15) in the low Asgroundwater (Fig. 4). The high As groundwater has a relativelywide range of d34SSO4 values that vary between 11.0‰ and 24.7‰

(Table 1 and Fig. 4). Assuming that the background d34SSO4 valuesare approximately 8–15‰ in the groundwater at Datong (Xieet al., 2009), the trend toward heavier d34SSO4 values in the highAs groundwater indicates that microbial SO4

2� reduction may haveoccurred. The reduction of SO4

2� in the high As groundwater canbe supported by the observed relatively lower SO4

2�concentrationthan that of the low As groundwater (Table 1). It is interesting tonote that a positive correlation between the d34SSO4 values andthe dissolved As concentrations has been observed (r2 = 0.91,a = 0.05) (Fig. 5A), which indicates that SO4

2� reduction accompa-nies As mobilization. As discussed above, HS� which produced bySO4

2� reduction, can directly reduce As-bearing Fe oxides/hydrox-ides and release As. Alternatively, the produced high concentrationof HS� can also lead to the formation of thioarsenates or thioarse-nites promoting the mobilization of As. Fig. 4 shows that all of the

Fig. 4. The relationship between d34SSO4 and the d56Fe in groundwater. The Fe(III)reduction zone (Zone 1) and SO4

2� reduction zone (Zone 2) can be identified basedon the d34SSO4 and d56Fe values. The Fe(III) reduction zone is characterized by lowerd34SSO4 values and a wide range of d56Fe values. However, the SO4

2� reduction zonehas a wide range of d34SSO4 values and higher d56Fe values.

samples with As concentrations <50 lg/L plot in Zone 1, which ischaracterized by a wide range of d56Fe values and d34SSO4 valueswithin the background range of d34SSO4 in this area (Xie et al.,2009). As discussed above, large Fe isotope variations in groundwa-ter are evidence for widespread redox cycling of Fe that is cata-lyzed by microbial Fe reduction. Therefore, the d56Fe and d34SSO4

values of the low As groundwater indicate that the prevalence ofmicrobial Fe(III) reduction without SO4

2� reduction cannotincrease As concentrations to greater than 50 lg/L. Other fromsample C1-2, all of the high As samples have higher d56Fe andd34SSO4 values and plot in Zone 2 in Fig. 4. The 56Fe-enrichedFe(II)aq formed during the late stage of Fe(III) reduction. The highd34SSO4 values are caused by the extensive sulfate reduction. Thisobservation indicates that extensive Fe(III) and SO4

2� reductionpromotes the release of As into the groundwater. Based on thed56Fe and d34SSO4 values and the As concentration, we concludethat: (1) the early stage of microbial reduction of Fe(III) mineralswithout the significant reduction of SO4

2� results in As concentra-tions lower than 50 lg/L; and (2) high As concentrations ingroundwater are due to the concurrent reduction of SO4

2� andFe(III) minerals. The high As sample C1-2 is characterized by lowd34SSO4 and high d56Fe values (Table 1). Because abiotic and biolog-ical oxidation of dissolved HS� cannot cause significant S isotopefractionation (Canfield, 2001; Taylor et al., 1984), the low d34SSO4

values in sample C1-2 can be attributed to the re-oxidation of Fesulfide, which has high d56Fe values due to the reaction betweenFe sulfide and Fe(II)aq (Butler et al., 2005b).

Fig. 6. The relationship between the d34SSO4 and the d13CDIC values in groundwater.The observed negative correlation between the d34SSO4 and the d13CDIC valuesconfirms the microbial oxidation of organic carbon and SO4

2� reduction.

420 Y. Wang et al. / Journal of Hydrology 519 (2014) 414–422

The measured d13CDIC values range from �33.59‰ to �4.81‰

(Table 1). The range of light d13C values measured in the ground-water from the Datong Basin requires a source of isotopically-lightcarbon. Becker and Clayton (1972) pointed out that the mostreasonable source of light carbon was organic matter. The carbonisotopic compositions of fossil organic matter fall in a range of verylight d13C values between �30‰ and �33‰ (Brocks et al., 1999).High organic carbon contents up to 2 wt% have been reported inthe Datong Basin (Wang et al., 2009). The diagenetic oxidation ofisotopically-light organic carbon can decrease the d13C values ofthe existing pore water reservoir of the DIC. In addition, similarto the S and Fe isotopic compositions, bacteria preferentially utilize12C, which decreases the d13C values of the DIC during the micro-bial mineralization of organic carbon (Clark and Fritz, 1997).Wachniew (2006) reported that the microbial decomposition oforganic matter can release DIC with isotopically-light d13CDIC val-ues ranging from �25‰ to �18‰. The low d13CDIC values in thegroundwater from the experimental plot are therefore attributedto a biogenic origin. Almost all of the high As groundwater sampleshave d13CDIC values between �17.94‰ and �33.59‰ (Fig. 5B). Thed13CDIC and d34SSO4 data indicate that microbial oxidation oforganic matter coupled to SO4

2� reduction results in the enrich-ment of As in groundwater (Fig. 5). The d13CDIC values ofapproximately �30‰ in the high As groundwater may indicatethat the DIC originated primarily from the oxidation of organicmatter. Because limestone has d13CDIC values of approximately0‰, the relatively high d13CDIC values (from �25.6‰ to �4.8‰)in the low As groundwater was possibly contributed by bothorganic and limestone-originated DIC sources.

4.2. Carbon, sulfur and iron isotopic evidence for arsenic mobilization

It is generally accepted that microbial activity plays a criticalrole in the mobilization of As from sediments (McArthur et al.,2004; Postma et al., 2007). However, the mechanisms of the met-abolic processes that promote As mobilization remain poorlyunderstood. Due to the direct and indirect interactions betweenthe C, Fe, S and As cycles, the C–S–Fe isotope system providesimportant tools to investigate the mechanisms of As mobilizationfrom aquifer sediments. The isotopic data in this study indicatethat the oxidation of organic carbon coupled to Fe(III) and SO4

2�

reduction results in As mobilization. The microbial oxidation oforganic carbon must be coupled to the reduction of an appropriateelectron acceptor. The order of oxidant consumption is O2 >Mn-oxides > NO3

� > Fe-oxides > SO42� (Borch et al., 2010). In

anoxic aquifer sediments, oxygen is severely limited as the primaryelectron acceptor, as is Mn-oxide and NO3

� (Anbar et al., 2007;Anbar and Holland, 1992). Therefore, Fe(III) and SO4

2� are poten-tially significant electron acceptors in the microbial respirationoxidation of organic matter in aquifers (Lovley et al., 1995). Theuse of SO4

2� as an electron acceptor to oxidize organic carbon issupported by the observed negative correlation between thed13CDIC values and the d34SSO4 values (r2 = 0.79, a = 0.05) (Fig. 6)because microbial organic carbon oxidation and SO4

2� reductionresult in lighter d13CDIC values and heavier d34SSO4 values. However,no significant correlation was observed between the d56Fe andd13CDIC values. Thus, we can conclude that SO4

2� experiencedmicrobial reduction, while Fe(III) experienced primarily abiotic Fereduction in the studied aquifer. As discussed above, the abioticFe reduction may be dominated by HS� direct reduction of Fe(III)oxides/hydroxides in the studied aquifers. Based on the 56Fe,d34SSO4 and d13CDIC values of the high As groundwater, it can beinferred that the HS� produced by microbial SO4

2� reductiondirectly abiotically reduces As-bearing Fe(III) oxides/hydroxidesand promotes As mobilization and enrichment in groundwater.However, the d56Fe values indicate the microbial Fe(III) reduction

release of As in groundwater with As concentrations lower than50 lg/L.

The results of our chemical extraction experiment indicatedthat crystalline Fe oxides contain much more As than amorphousFe oxide/hydroxides (Xie et al., 2013a). According to the sequenceof redox reactions (Borch et al., 2010), the reduction of SO4

2� to HS�

is followed by the reduction of crystalline Fe oxide/hydroxidessuch as a-FeOOH(s), while the reduction of amorphous Fe hydrox-ides such as Fe(OH)3(am)(s) occurs before SO4

2� reduction. Accord-ingly, the microbial reduction of As-bearing amorphous Fehydroxides such as Fe(OH)3(am)(s) results in As concentrationslower than 50 lg/L in the earlier stage. Later, microbial SO4

2� reduc-tion catalyzed HS� directly to reduce crystalline Fe oxide/hydrox-ides and generated the high As groundwater. Furthermore, theformation of thioarsenates or thioarsenites in high HS� concentra-tion groundwater is favorable for the mobilization of As.

5. Conclusions

Iron, C and S isotopic analysis of As-contaminated groundwatercan be used to constrain the biochemical processes that control Asmobilization and enrichment in groundwater. In this study, wereport the Fe, C and S isotopic compositions of a known As-con-taminated site in the Datong basin, north China. Based on theresults, we can conclude the following:

(1) High As concentrations were detected in groundwater fromthe experimental plot and range from 4.76 to 469.5 lg/L. Thelow As groundwater with lighter d56Fe values reflects partialmicrobial reduction of ferric oxides/hydroxides. The gradu-ally isotopically lighter d56Fe values in the high As ground-water can be attributed to the contact of Fe(II)aq with FeSor to HS� abiotically reducing Fe(III) oxides/hydroxides.

(2) The light d13CDIC values of groundwater between �4.8‰ and�33.6‰ must have originated from a reservoir of carbonwith very light d13C values. The source of the light d13C islikely microbial oxidation of organic carbon within the aqui-fer sediments.

(3) The groundwater with a wide range of d34SSO4 valuesbetween 10.0‰ and 24.1‰ indicates microbial sulfatereduction. The negative correlation between the d34SSO4

and d13CDIC values provides evidence of the microbial reduc-tion of sulfate that is coupled to the oxidation of organic car-bon within the aquifers.

Y. Wang et al. / Journal of Hydrology 519 (2014) 414–422 421

(4) The combined Fe, C and S isotopic data indicate that: (i) themicrobial reduction of amorphous Fe(III) oxides/hydroxidesresults in As concentrations lower than 50 lg/L; and (ii)microbially-reduced SO4

2� catalyzed HS� abiotically reducescrystalline Fe(III) oxides/hydroxides or the formation ofAs-sulfur components, resulting in groundwater As concen-trations greater than 50 lg/L.

Acknowledgements

This research was financially supported by the National NaturalScience Foundation of China (Nos. 41202168 and 41372254), theMinistry of Science and Technology of China (2012AA062602),the China Postdoctoral Science Foundation and the FundamentalResearch Fund for National Universities, China University ofGeosciences (Wuhan). The authors would like to appreciate thereviewers and editors for their helpful comments and suggestions.

References

Anbar, A.D., Duan, Y., Lyons, T.W., Arnold, G.L., Kendall, B., Creaser, R.A., Kaufman,A.J., Gordon, G.W., Scott, C., Garvin, J., Buick, R., 2007. A whiff of oxygen beforethe Great Oxidation Event? Science 317 (5846), 1903–1906.

Anbar, A.D., Holland, H.D., 1992. The photochemistry of manganese and the originof banded iron formations. Geochim. Cosmochim. Acta 56 (7), 2595–2603.

Appelo, C.A.J., Van der Weiden, M.J.J., Tournassat, C., Charlet, L., 2002. Surfacecomplexation of ferrous iron and carbonate on ferrihydrite and the mobilizationof arsenic. Environ. Sci. Technol. 36, 3096–3103.

Böttcher, M.E., Thamdrup, B., 2001. Anaerobic sulfide oxidation and stable isotopefractionation associated with bacterial sulfur disproportionation in thepresence of MnO2. Geochim. Cosmochim. Acta 63, 1653–1660.

Beard, B.L., Johnson, C.M., 1999. High precision iron isotope measurements ofterrestrial and lunar materials. Geochim. Cosmochim. Acta 63, 1653–1660.

Beard, B.L., Johnson, C.M., Skulan, J.L., Nealson, K.H., Cox, L., Sun, H., 2003.Application of Fe isotopes to tracing the geochemical and biological cycling ofFe. Chem. Geol. 195, 87–117.

Becker, R.H., Clayton, R.N., 1972. Carbon isotopic evidence for the origin of a bandediron-formation in Western Australia. Geochim. Cosmochim. Acta 36, 577–595.

Berg, M. et al., 2007. Magnitude of arsenic pollution in the Mekong and Red RiverDeltas – Cambodia and Vietnam. Sci. Total Environ. 372, 413–425.

Berg, M., Stengel, C., Trang, P.T.K., Viet, P.H., Sampson, M.L., Leng, M., Samreth, S.,Fredericks, D., 2008. Hydrological and sedimentary controls leading to arseniccontamination of groundwater in the Hanoi area, Vietnam: the impact of iron-arsenic ratios, peat, river bank deposits, and excessive groundwater abstraction.Chem. Geol. 249, 91–112.

Borch, T., Kretzschmar, R., Kappler, A., Van Cappellen, P., Ginder-Vogel, M., Voegelin,A., Campbell, K., 2010. Biogeochemical redox processes and their impact oncontaminant dynamics. Environ. Sci. Technol. 44 (1), 15–23.

Bose, P., Sharma, A., 2002. Role of iron in controlling speciation and mobilization ofarsenic in subsurface environment. Water Res. 36, 4916–4926.

Bostick, B.C., Fendorf, S., Brown, G.E., 2005. In situ analysis of thioarsenitecomplexes in neutral to alkaline arsenic sulphide solutions. Miner. Mag. 69,781–795.

Brocks, J.J., Logan, G.A., Buick, R., Summons, R.E., 1999. Archean molecular fossilsand the early rise of eukaryotes. Science 285 (5430), 1033–1036.

Bullen, T.D., White, A.F., Childs, C.W., Vivit, D.V., Schulz, M.S., 2001. Demonstrationof significant abiotic iron isotope fractionation in nature. Geology 29 (8), 699–702.

Buschmann, J. et al., 2008. Contamination of drinking water resources in theMekong delta floodplains: arsenic and other trace metals pose serious healthrisks to population. Environ. Int. 34, 756–764.

Butler, I.B., Archer, C., Vance, D., Oldroyd, A., Rickard, D., 2005. Fe isotopefractionation on FeS formation in ambient aqueous solution. Earth Planet. Sci.Lett. 236 (1–2), 430–442.

Canfield, D.E., 2001. Biogeochemistry of sulfur isotope. Rev. Miner. Geochem. 43,607–636.

Canfield, D.E., Thamdrup, B.J., Hansen, W., 1993. The anaerobic degradation oforganic matter in Danish coastal sediments: iron reduction, manganesereduction, and sulfate reduction. Geochim. Cosmochim. Acta 57, 3867–3883.

Clark, I.D., Fritz, P., 1997. Environmental Isotopes in Hydrogeology. Lewis, NewYork, p. 328.

Crosby, H.A., Johnson, C.M., Roden, E.E., Beard, B.L., 2005. Coupled Fe(II)–Fe(III)electron and atom exchange as a mechanism for Fe isotope fractionation duringdissimilatory iron oxide reduction. Environ. Scie. Technol. 39 (17), 6698–6704.

Detmers, J., Bruchert, V., Habicht, K.S., Kuever, J., 2001. Diversity of sulfur isotopefractionations by sulfate-reducing prokaryotes. Appl. Environ. Microb. 67, 888–894.

Fisher, J.C., Wallschlager, D., Planer-Friedrich, B., Hollibaugh, J.T., 2008. A new rolefor sulfur in arsenic cycling. Environ. Scie. Technol. 42, 81–85.

Gibson, B.D., Amos, R.T., Blowes, D.W., 2011. 34S/32S fractionation during sulfatereduction in groundwater treatment systems: reactive transport modeling.Environ. Sci. Technol. 45, 2863–2870.

Guo, H.M., Tang, X.H., Yang, S.Z., Shen, Z.L., 2008a. Effect of indigenous bacteria ongeochemical behavior of arsenic in aquifer sediments from the Hetao Basin,Inner Mongolia: evidence from sediment incubations. Appl. Geochem. 23,3267–3277.

Guo, H.M., Yang, S.Z., Tang, X.H., Li, Y., Shen, Z.L., 2008b. Groundwater geochemistryand its implications for arsenic mobilization in shallow aquifers of the HetaoBasin. Inner Mongolia. Sci. Total Environ. 393, 131–144.

Harvey, C.F., Swartz, C.H., Badruzzaman, A.B.M., Keon-Blute, N., Yu, W., Ali, M.A., Jay,J., Beckie, R., Niedan, V., Brabander, D., Oates, P.M., Ashfaque, K.N., Islam, S.,Hemond, H.F., Ahmed, M.F., 2002. Arsenic mobility and groundwater extractionin Bangladesh. Science 5, 1602–1606.

Herbert Jr., R.B., Schippers, A., 2008. Iron isotope fractionation by biogeochemicalprocesses in mine tailings. Environ. Sci. Technol. 42, 1117–1122.

Islam, F.S., Islam, F.S., Gault, A.G., Boothman, C., Polya, D.A., Charnock, J.M.,Chatterjee, D., Lloyd, J.R., 2004. Role of metal-reducing bacteria in arsenicrelease from Bengal delta sediments. Nature 430, 68–71.

Johnson, C.M., Beard, B.L., 2005. Biogeochemical cycling of iron isotopes. Science309 (5737), 1025–1027.

Johnson, C., Beard, B., Roden, E., 2008. Temporal variations in Fe isotopecompositions of banded iron formations record changes in the nature ofredox cycling. Geochim. Cosmochim. Acta 72. A435–A435.

Johnson, C.M., Beard, B.L., Roden, E.E., 2008. The iron isotope fingerprints of redoxand biogeochemical cycling in the modern and ancient Earth. Annu. Rev. EarthPlanet. Sci. 36, 457–493.

Johnson, C.M., Roden, E.E., Welch, S.A., Beard, B.L., 2005. Experimental constraintson Fe isotope fractionation during magnetite and Fe carbonate formationcoupled to dissimilatory hydrous ferric oxide reduction. Geochim. Cosmochim.Acta 69, 963–993.

Kirk, M.F., Holm, T.R., Park, J., Jin, Q.S., Sanford, R.A., Fouke, B.W., Bethke, C.M., 2004.Bacterial sulfate reduction limits natural arsenic contamination ingroundwater. Geology 32, 953–956.

Lovley, D.R., 1997. Microbial Fe(III) reduction in subsurface environments. FEMSMicrob. Rev. 20, 305–313.

Lovley, D.R., Woodard, J.C., Coates, J.D., Chapelle, F.H., 1995. Anaerobicbioremediation of benzene in petroleum-contaminated aquifers. Abstracts ofPapers of the American Chemical Society, 209: 98-ENVR.

McArthur, J.M., Banerjee, D.M., Hudson-Edwards, K.A., Mishra, R., Purohit, R.,Ravenscroft, P., Cronin, A., Howarth, R.J., Chatterjee, A., Talukder, T., Lowry, D.,Houghton, S., Chadha, D.K., 2004. Natural organic matter in sedimentary basinsand its relation to arsenic in anoxic ground water: the example of West Bengaland its worldwide implications. Appl. Geochem. 19 (8), 1255–1293.

Nickson, R., McArthur, J., Burgess, W., Ahmed, K.M., Ravenscroft, P., Rahman, M.,1998. Arsenic poisoning of Bangladesh groundwater. Nature 395, 338.

Nickson, R.T., McArthur, J.M., Ravenscroft, P., Burgess, W.G., Ahmed, K.M., 2000.Mechanism of arsenic release to groundwater, Bangladesh and West Bengal.Appl. Geochem. 15, 403–413.

Oremland, R.S., Stolz, J.F., 2003. The ecology of arsenic. Science 300, 939–944.Organization, W.H., 2004. Guidelines for drinking water quality, vol. 1

Recommendations, World Health Organization, Geneva.Papacostas, N.C., Bostick, B.C., Quicksall, A.N., Landis, J.D., Sampson, M., 2008.

Geomorphic controls on groundwater arsenic distribution in the Mekong RiverDelta, Cambodia. Geology 36, 891–894.

Percak-Dennett, E.M., Beard, B.L., Xu, H., Konishi, H., Johnson, C.M., Roden, E.E.,2011. Iron isotope fractionation during microbial dissimilatory iron oxidereduction in simulated Archaean seawater. Geobiology 9, 205–220.

Polizzotto, M.L., Kocar, B.D., Benner, S.G., Sampson, M., Fendorf, S., 2008. Near-surface wetland sediments as a source of arsenic release to ground water inAsia. Nature 454, 505–508.

Postgate, J.R., 1984. The Sulphate Reducing Bacteria, second ed. CambridgeUniversity Press, Cambridge, 208 p.

Postma, D., Larsen, F., Hue, N.T.M., Duc, M.T., Viet, P.H., Nhan, P.Q., Jessen, S., 2007.Arsenic in groundwater of the Red River floodplain, Vietnam: controllinggeochemical processes and reactive transport modeling. Geochim. Cosmochim.Acta 71, 5054–5071.

Planer-Friedrich, B., London, J., McCleskey, R.B., Nordstrom, D.K., Wallschlager, D.,2007. Thioarsenates in geothermal waters of yellowstone national park:determination, preservation, and geochemical importance. Environ. Sci.Technol. 41, 5245–5251.

Raiswell, R., Canfield, D.E., 1998. Sources of iron for pyrite formation in marinesediments. Am. J. Sci. 298, 219–245.

Robertson, W.D., Schiff, S.L., 1994. Fractionation of sulphur isotopes during biogenicsulphate reduction below a sandy forested recharge area in south-centralCanada. J. Hydrol. 158, 123–134.

Rouxel, O., Sholkovitz, E., Charette, M., Edwards, K.J., 2008. Iron isotope fractionationin subterranean estuaries. Geochim. Cosmochim. Acta 72 (14), 3413–3430.

Smedley, P.L., Kinniburgh, D.G., 2002. A review of the source, behaviour anddistribution of arsenic in natural waters. Appl. Geochem. 17 (5), 517–568.

Stauder, S., Raue, B., Sacher, F., 2005. Thioarsenates in sulfidic waters. Environ. Sci.Technol. 39, 5933–5939.

Swartz, C.H., Blute, N.K., Badruzzman, B., Ali, A., Brabander, D., Jay, J., Besancon, J.,Islam, S., Hemond, H.F., Harvey, C.F., 2004. Mobility of arsenic in a Bangladeshaquifer: inferences from geochemical profiles, leaching data, and mineralogicalcharacterization. Geochim. Cosmochim. Acta 68 (22), 4539–4557.

422 Y. Wang et al. / Journal of Hydrology 519 (2014) 414–422

Taylor, B.E., Wheeler, M.C., Nordstrom, D.K., 1984. Stable isotope geochemistry ofacid-mine drainage-experimental oxidation of pyrite. Geochim. Cosmochim.Acta 48 (12), 2669–2678.

Thi, H.H., Bang, S., Kim, K.W., My, H.N., Duy, M.D., 2010. Arsenic in groundwaterand sediment in the Mekong River delta, Vietnam. Environ. Poll. 158, 2648–2658.

Teutsch, N., Berg, M., von Gunten, U., Halliday, A., 2004. Iron isotope variations inreduced groundwater and in drinking water supplies: a case study of Hanoi,Vietnam. American Geophysical Union, Fall Meeting: V51A-0511.

Wachniew, P., 2006. Isotopic composition of dissolved inorganic carbon in a largepolluted river: the Vistula, Poland. Chem. Geol. 233 (3–4), 293–308.

Wang, Y.X., Shvartsev, S.L., Su, C.L., 2009. Genesis of arsenic/fluoride-enriched sodawater: a case study at Datong, northern China. Appl. Geochem. 24, 641–649.

Xie, X., Johnson, T.M., Wang, Y.X., Lundstrom, C.C., Ellis, A., Wang, X.L., Duan, M.Y.,2013a. Mobilization of arsenic in aquifers from the Datong Basin, China:evidence from geochemical and iron isotopic data. Chemosphere 90 (6), 1878–1884.

Xie, X., Wang, Y.X., Ellis, A., Li, J.X., Su, C.L., Duan, M.Y., 2013b. Multiple isotope (O, Sand C) approach elucidates the enrichment of arsenic in the groundwater fromthe Datong Basin, northern China. J. Hydrol. 498, 103–112.

Xie, X.J., Ellis, A., Wang, Y.X., Xie, Z.M., Duan, M.Y., Su, C.L., 2009. Geochemistry ofredox-sensitive elements and sulfur isotopes in the high arsenic groundwatersystem of Datong Basin, China. Sci. Total Environ. 407, 3823–3835.

Xie, X.J., Wang, Y.X., Su, C.L., Liu, H.Q., Duan, M.Y., Xie, Z.M., 2008. Arsenicmobilization in shallow aquifers of Datong Basin: hydrochemical andmineralogical evidences. J. Geochem. Explor. 98, 107–115.

Xie, X.J., Johnson, T.M., Wang, Y.X., Lundstrom, C.C., Ellis, A., Wang, X.L., Duan, M.Y.,Li, J.X., 2014. Pathways of arsenic from sediments to groundwater in thehyporheic zone: evidence from an iron isotope study. J. Hydrol. 511, 509–517.

Yao, W., Millero, F.J., 1996. Oxidation of hydrogen sulfide by hydrous Fe(III) oxidesin seawater. Mar. Chem. 52 (1), 1–16.

Young, E.D., Galy, A., Nagahara, H., 2002. Kinetic and equilibrium mass-dependentisotope fractionation laws in nature and their geochemical and cosmochemicalsignificance. Geochim. Cosmochim. Acta 66, 1095–1104.