Insights into the Surface Complexation of Dimethylarsinic Acid on Iron

Publication Date (Web): September 21, 2010 ...... characterization of arsenate (V) ion adsorption from mine waters, macraes mine, new zealand Appl. Ge...
0 downloads 0 Views 481KB Size
Environ. Sci. Technol. 2010, 44, 7802–7807

Insights into the Surface Complexation of Dimethylarsinic Acid on Iron (Oxyhydr)oxides from ATR-FTIR Studies and Quantum Chemical Calculations ADRIAN ADAMESCU, WILLIAM MITCHELL, I. P. HAMILTON, AND HIND A. AL-ABADLEH* Chemistry Department, Wilfrid Laurier University, Waterloo, Ontario N2L 3C5, Canada

Received April 11, 2010. Revised manuscript received August 26, 2010. Accepted September 10, 2010.

The surface chemistry of methylated arsenicals with ubiquitous geosorbents and industrial catalysts is poorly understood. These arsenic compounds pose both a health and an environmental risk in addition to being a challenge to the energy industry. We report herein a detailed spectroscopic analysis of the surface structure of dimethylarsinic acid (DMA) adsorbed on hematite and goethite using attenuated total internal reflectance Fourier transform infrared spectroscopy (ATRFTIR). Spectra of adsorbed DMA, DMA(ads), were collected in situ as a function of pH and ionic strength, using both H2O and D2O at 298 K in flow mode. Experimental data were complemented with DFT calculations of geometries and frequencies of hydrated DMA-iron oxide clusters. Results indicate the simultaneous formation of inner- and outer-sphere complexes with distinct spectral components. Desorption behavior of DMA due to chloride and phosphate was studied as a function of time from the decrease in the absorbance of apparent spectral features. The impact of our studies on the environmental fate of DMA in geochemical environments and the design of technologies to reduce arsenic content in fuels are discussed.

Introduction Arsenic (As) is a known carcinogen that is naturally occurring in bedrocks around the world (1). It is also one of the metal(loids) that accumulate in petroleum, shale, and coal deposits as a result of biogeochemical processes (2), and it has been found in fly ash from the combustion of solid biofuels (3). Arsenic speciation in environmental samples and in the above energy sources includes inorganic arsenic (iAs) and methylated organoarsenicals. Methylated organoarsenicals such as monomethylarsonic acid (MMA) and DMA are produced in biomethylation processes of iAs that occur under aerobic and anaerobic conditions (1). They have been detected in the leachates of landfills rich in waste containing arsenic such as glass, alloys, and semiconductors (4) and biologically pretreated municipal solid waste (5). Both compounds were also historically used as pesticides, herbicides, and defoliants on golf courses and agricultural lands (6) in the U.S. and Canada. In addition, organoarsenicals * Corresponding author phone: (519)884-0710 ext. 2873; fax: (519)746-0677; e-mail: [email protected]. 7802

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 20, 2010

including methylated forms are synthesized during the pyrolysis of oil shale (7). Under certain soil solution conditions, there is potential cycling to more toxic forms of arsenic (i.e., As in the oxidation state of +3 and inorganic forms of As) as a result of redox activity (8). Additionally, arsenicals in fossil fuels and biomass grown in As-rich soils are potent catalyst poisons hindering the optimum use and conversion of these fuels (2). Hence, methylated organoarsenicals pose both a health and an environmental risk (9) and continue to be a challenge to the energy industry. Little is known about the adsorption of methylated organoarsenicals on surfaces relevant to environmental geosorbents, materials used in pollution remediation, or catalysts employed in the petroleum industry. Cox and Ghosh (10) studied the adsorption of MMA and DMA on hydrous ferric oxide and activated alumina and assumed, from ionic strengthdependent studies and surface complexation modeling (SCM), that these arsenicals form inner-sphere complexes. Lafferty and Loeppert (11) obtained, by means of batch experiments, the adsorption isotherms of MMAIII, MMAV, DMAIII, and DMAV on iron minerals as a function of pH. They found that the degree of methylation and oxidation state of arsenic affect the binding affinities to these minerals. In another study by Jing and co-workers (12), surface complexation of MMAV and DMAV on nanocrystalline TiO2 was studied using extended X-ray absorption fine structure spectroscopy (EXAFS). The data showed the formation of inner-sphere bi- and monodentate complexes for MMAV and DMAV, respectively. Xu et al. (13) studied the mechanism of the photocatalytic degradation of MMAV and DMAV adsorbed on TiO2 under UV radiation and found that 93% of MMA photodegrades to iAs(V) within 72 h and that DMA forms iAs(V) through MMA as an intermediate. More recently, Shimizu et al. (14) utilized ATR-FTIR and EXAFS for elucidating structural information on the adsorption of MMA and DMA on amorphous aluminum oxide (AAO). They concluded that both organoarsenicals form bidentate-binuclear complexes with AAO and that increasing methyl group substitution increases As desorption from AAO. With the exception of the study by Shimizu et al. (14), the published reports on methylated organoarsenicals adsorption on Al- and Fe-containing sorbents using transmission infrared spectroscopy lack the surface sensitivity needed to characterize surface complexes (15, 16). We report herein in situ and surface-sensitive ATR-FTIR studies complemented with quantum chemical calculations on the surface interactions of DMA with hematite and goethite. DMA is the focus of our studies because it has two methyl groups and was reported to form mono- and bidentate surface complexes (vide infra). The objective of our studies is to gain insight into the nature of DMA surface complexes from pH(D)- and ionic strength (I)-dependent adsorption experiments and quantum chemical calculations. Graphs of pH-envelopes of arsenicals are conventionally constructed from batch experiments, where total adsorbed arsenic is quantified versus pH (10, 11, 14, 17). Such graphs are useful in establishing trends on the extent of adsorption under variable acidic conditions, electrolyte solution, and surface loading, which can lead to qualitative conclusions on the binding mechanism and nature of surface complexes. The latter can only be deduced from in situ surface sensitive measurements like the ones reported herein. Quantum chemical calculations performed on DMA-iron oxide clusters in the inner- and outer-sphere configurations were used to aid in the interpretation of our experimental results using ATR-FTIR. The significance of our studies is discussed in terms of their usefulness in developing surface 10.1021/es1011516

 2010 American Chemical Society

Published on Web 09/21/2010

complexation models, the geochemical conditions that increase the bioavailability of DMA, and potential technologies for arsenic removal from fuel.

Experimental and Computational Section Chemicals. Stock solutions of DMA (cacodylic acid sodium salt trihydrate, C2H6AsO2Na · 3H2O, Alfa Aesar, used as received) were prepared by dissolving the powder in concentrated NaOH (ACS grade, EMD) with continuous mechanical stirring and then lowering the pH using HCl (6 N, Ricca Chemical). Caution: DMA is highly toxic via inhalation and skin contact and is a carcinogen (18). Ionic strength of the 18 MΩ Millipore water was adjusted using KCl (99.5%, EM Science). Solutions of DMA were also prepared in D2O (99.9 atom %D, Sigma) by dissolving the powder in NaOD (40 wt % solution, Sigma) and then lowering the pD using DCl (35 wt % solution, Sigma). Stock solutions of phosphate (99%, Na2HPO4, BDH) adjusted to pH ) 7 and I ) 0.01 M were used for part of the desorption experiments. The Fe(oxyhydr)oxides used herein are hematite (R-Fe2O3, >99.9%, Nanostructured and Amorphous Materials) and goethite (RFeOOH, >99.9%, Alfa Aesar). Characterization of BET surface area, particles’ shape and size, and isoelectric points was reported earlier (19). Details on the experimental procedure for preparing thin Fe-(oxyhydr)oxide films on the ATR internal reflection element (IRE) were described in the Supporting Information (SI) of ref 19. ATR-FTIR Experiments. The collection procedure of ATRFTIR spectra as a function of pH(D) was previously described (17). Briefly, for each adsorption experiment on a freshly prepared film, spectra were collected using a HATRPlus accessory (Pike Technologies) installed in a Nicolet 8700 FTIR spectrometer (Thermo Instruments) equipped with a MCT detector. The Fe-(oxyhydr)oxide films were directly deposited on a 60°ZnSe crystal IRE (80 × 10 × 4 mm) housed in a 100 µL ATR flow cell. Single beam ATR-FTIR spectra were collected at 4 cm-1 resolution by averaging 300 scans as a function of I and pH(D). Background aqueous KCl followed by DMA solutions (1 mM) were flown at a rate of 1 mL/min across a given Fe-(oxyhydr)oxide film using Tygon tubes (0.8 mm I.D., Maserflex) and a compact pump (Masterflex L/S). Adsorption kinetic experiments were performed to confirm that spectra collected after 15 min are at equilibrium (Figure S1). Desorption behavior experiments by electrolyte and phosphate solutions were conducted at pH 7 and I ) 0.01 M. To determine the uncertainty in our measurements, experiments were repeated 3-4 times on freshly prepared films. Computational Method. Quantum chemical calculations of monodentate, bidentate binuclear, and outer-sphere DMAFe(oxyhydr)oxide clusters were performed using the Gaussian 09 program (20) running on Sharcnet (21). Structures were energy-minimized without any symmetry constraints using DFT with the B3LYP functional and the 6-31G* basis set. Frequencies of the energy-minimized structures were also calculated using Gaussian 09, and vibrational modes were visualized using ChemCraft(v.1.6). DFT calculations were performed on isolated and hydrated clusters with a net charge of zero, +1 or +2. Hydration was simulated both explicitly by adding four water molecules and implicitly using the Integral-Equation-Formalism Polarizable Continuum Model (IEFPCM) (22). In both cases structures were energyminimized as above. A scaling factor of 0.9787 was used to correct calculated frequencies for anharmonicity.

Results and Discussion ATR-FTIR Spectra of DMA(ads). Representative ATR-FTIR spectra of DMA(ads) on R-Fe2O3 and R-FeOOH are shown in Figure 1(top). The spectral range shown (700-950 cm-1) contains infrared absorbances due to the stretching vibrations

FIGURE 1. Top: ATR-FTIR absorption spectra of DMA(ads) on (a) r-Fe2O3 and (b) r-FeOOH particles as a function of pH (solid lines) and pD (dashed lines) and I ) 0.01 M KCl. Bottom: pH-envelopes of DMA adsorption on (c) r-Fe2O3 and (d) r-FeOOH particles as a function of I ) 0.001, 0.01, and 0.1 M at 298 K. Experimental data (filled markers) are the baselinecorrected ATR absorbance at 840 ( 2 cm-1, which is used to calculate SDMA. Error bars are (σ from averaging 3-4 experiments, each on a freshly prepared film. (v) of the AsO2 moiety in DMA, which is sensitive to protonation (Figure S2), H-bonding and binding to metal(oxyhydr)oxide surfaces (14). This spectral range also contains rocking vibrations of methyl groups, F(CH3) (>880 cm-1). There is no evidence for the contribution of this mode to the spectra shown in Figure 1, which was collected using 1 mM DMA(aq). This is not surprising given its relatively low intensity (compare to Figure S2). The growth of shoulders at v > 880 cm-1 was observed using [DMA(aq)] that are a factor of 8 higher than those used in our adsorption studies herein. The assignment of the observed spectral components in Figure 1 is summarized in Table 1, which was based on features observed in the IR spectra of bulk DMA in the liquid (pKa ) 6.1, Figure S2), solid acid, and salt phases (23), reported assignments of iAs(V) on Fe-(oxyhydr)oxides (24-30), and aided by results from DFT calculations described below. This assignment (detailed below) suggests the presence of innersphere complexes (monodentate and/or bidentate) and also outer-sphere complexes. Simultaneous formation of these complexes might explain the relatively large width of the bands observed in Figure 1 relative to those recorded for DMA in the bulk phases. For comparison, Catalano et al. (31) showed simultaneous inner- and outer-sphere complexes for iAs(V) on hematite using in situ resonant X-ray scattering. The intensity of components in Figure 1 increases with decreasing pH(D) consistent with the increase in the surface coverage of DMA (SDMA) with no obvious shift in frequencies. Components below 800 cm-1 are assigned to v(As-OFe) in inner-sphere DMA complexes. These values are higher than those assigned to v(As-OH) in DMA(aq) (ca. 735 cm-1). This is expected from the reported spectral analysis of adsorbed iAs(V), which showed that the binding of the As-O groups to protons causes a larger decrease in their force constant relative to Fe3+ (24). When spectra in Figure 1 are compared with that of DMA(s,salt), which is H-bonded to water molecules in its structure (23), the frequencies at 793 and 787 cm-1 also suggest the presence of As-O · · · · H groups in inner- and/or outer-sphere complexes. The latter complexes might become protonated in the pH range 5.5-4, which would give rise to low intensity shoulders at 735 cm-1 VOL. 44, NO. 20, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7803

TABLE 1. Assignment of As-O Stretching Vibrations of DMA(ads) on Fe-(Oxyhydr)oxides from ATR-FTIR Experimentsa outer-sphereb R-Fe2O3 R-FeOOH

inner-spherec

AsdO

As---Od

As-O · · · · He

As---Od

As-O · · · · He

As-OFe

877 876

840 837

793 768, 787

840 837

793 787

793, 775 787, 768

Comparison with Bulk Phases [SI and ref 23]

DMA(aq) pH 3 pD 4 pH 11 pD 10 DMA(s,salt) DMA(s)

AsdO

As---Of

As-O · · · · H

As-OH

879 879 -

-

787 -

735 731 754

831 835 841 868, 829

a Compared with calculated frequencies in Figure 3 and Table S3. b Compare to iAs(V) outer-sphere (ca. 870 cm-1) (24, 25). c Compare to iAs(V) monodentate (816-860 cm-1) (26) and/or bidentate binuclear (805-817 cm-1) (27-30). d Uncomplexed, free, and bond order of ca. 1.5 influenced by binding to the surface. e Uncomplexed, strong H-bonding, compare to solid phase. f Bond order of ca. 1.5 due to resonance.

indicating the presence of As-OH groups. These shoulders are too low in intensity to be detected using [DMA(aq)] ) 0.001 M. As shown below, the decrease in the absorbance of the 775 and 793 cm-1 components due to DMA desorption suggests that they arise from outer- and inner-sphere complexes, respectively. Components >800 cm-1 indicate the presence of stronger AsO bonds than those directly protonated or bonded to Fe3+. Given that DMA has two AsO bonds, protonation or surface binding of one group will decrease the delocalization of electrons causing an increase in the force constant of the second AsO bond. We recently quantified this effect due to protonation and organic substitution using DFT calculations on hydrated iAs(V), MMA, p-arsanilic acid (p-AsA), and DMA clusters (32). The highest degree of resonance among AsO2 groups in DMA occurs in the fully deprotonated form, which results in an intense band around 835 cm-1. Hence the 840 and 837 cm-1 components in Figure 1 are assigned to the uncomplexed (free) As---O bond in DMA(ads), whose bond order is influenced by either electron withdrawal due to surface binding of the other AsO bond, electrostatic attraction with neighboring electron-deficient sites, or strong solvation effects. This component persists in adsorption studies using D2O as a solvent (dashed lines, Figure 1) suggesting insensitivity to protons. The lower panel of Figure 1 shows that the absorbance of ca. 840 cm-1 component has stronger dependency on pH than on I (within the uncertainty of our measurements). This component is used to quantify SDMA (SI, right axis in Figure 1). Constructing pH-envelopes from the absorbance of specific spectral components can verify the assignment of these components to certain surface complexes under variable adsorption conditions. This procedure provides quantitative insight into the binding mechanism, leading to potentially more accurate modeling formulations. When taking into account the pKa of DMA (6.14) and the point of zero charge (PZC) of the Fe-(oxyhdyr)oxides used in our studies (ca. 9), one could conclude that adsorption of DMA under basic conditions (pH > 9) is unfavorable. The value of SDMA increases with decreasing pH due to the increase in electrostatic attraction between the increasing positively charged sites and DMA molecules. The maximum SDMA is observed between pH 4-5 for both films, which is about a factor of 4 higher than SDMA at pH 10. The nonzero SDMA at pH > 9 suggests the association of the component at 840 cm-1 with inner-sphere complexes. This is in contrast to the pH-envelope reported by Lafferty and Loeppert (11), who 7804

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 20, 2010

observed maximum adsorption of DMA on goethite between pH 4-5 and no adsorption at pH > 7.5. They attributed their observation to a binding mechanism different than that observed for iAs(V) and MMA, where inner-sphere bidentate surface complexes predominate. As shown below, the decrease in the absorbance of this component due to DMA desorption follows the same time profile of the component at 793 cm-1 suggesting that it arises from the predominant surface complex (i.e., outer-sphere). Moreover, components in Figure 1 at 877 and 876 cm-1 are assigned to v(AsdO) in outer-sphere complexes (Tables 1 and S3). These components are present under basic conditions and increase in intensity with lowering pH(D). The persistence of these features when adsorption studies are conducted in D2O (dashed lines) further supports this assignment. The presence of H-bonded complexes of iAs(V) adsorbed on Fe-(oxyhydr)oxide films was suggested to explain features around 870 cm-1 (24, 33). Figure S3 shows pH envelopes for DMA adsorption on Fe-(oxyhydroxide) films from the absorbance at 877 cm-1 as a function of I. Within the uncertainty of our measurements, the extent of adsorption observed therein is similar to that using the 840 cm-1 component. As shown below, quantum chemical calculations on outer-sphere DMA-iron oxide clusters give rise to frequencies >860 cm-1. Based on the above spectral analysis, desorption experiments were conducted using ATR-FTIR to further explore the existence of inner- and outer-sphere DMA(ads). Figure 2 shows the time profile of DMA desorption from R-Fe2O3 due to chloride and phosphate anions at pH 7. This figure was constructed from the absorbance of the spectral components assigned to DMA(ads), 877, 840, 793, and 775 cm-1, which were normalized to the absorbance at the same wavenumber after equilibrium was established with 1 mM DMA(aq) at pH 7 and I ) 0.01 M. This [DMA(aq)] results in SDMA ≈ 0.5 · Smax (i.e., unsaturated surface sites). Because adsorbed phosphate contributes to the component at 877 cm-1, its time-profile was not analyzed upon flowing phosphate solutions (after 30 min). When chloride is used as the desorbing agent, it results in ca. 40% reduction in SDMA, with similar kinetic behavior among the components analyzed. Flowing phosphate solutions resulted in further reduction in the intensity of the 840 and 793 cm-1 components, resulting in the desorption of >80% of initial SDMA. However, the component at 775 cm-1 leveled off after ca. 35 min, even after flowing phosphate. This clearly indicates that this component arises from strongly bonded inner-sphere

FIGURE 2. Time-profile of DMA desorption from r-Fe2O3 (SDMA≈0.5 · Smax) due to flowing Cl-(aq) and HPO42-(aq) at pH ) 7 and I ) 0.01 M KCl. Time-profile of the baseline-corrected and normalized ATR absorbance at 877, 840, 795, and 775 cm-1 assigned to v(As-O) in DMA(ads) and 1045 cm-1 assigned to v3(PO4) in adsorbed phosphate (17). Error bars (ca. ( 30%, removed for clarity) were determined from averaging 3 experiments, each on freshly prepared film. complexes, whereas the other components have major contributions from weakly bonded outer-sphere complexes. Carefully designed kinetics experiments are currently ongoing in our lab to derive relative desorption rate constants from spectral data. The component at 816 cm-1 in Figure 1(a) is absent from spectra in Figure 1(b), where one could conclude that there is a loss feature around 800 cm-1. This spectral region contains frequencies assigned to OH libration vibrations in R-FeOOH (30, 34) and frustrated rotation modes of surface water molecules (35). Hence, the signal loss in Figure 1(b) suggests a decrease in the concentration of ≡FeOH or ≡FeOH2+ sites on FeOOH films as a result of DMA adsorption through H-bonding or ligand exchange mechanisms. The fact that this component increases in intensity with decreasing pH (higher SDMA) suggests that it corresponds to protonated sites on the surface not involved in ligand exchange with DMA. Indeed, when adsorption experiments are conducted using D2O on R-Fe2O3, a reduction in the intensity of the 816 cm-1 feature is observed suggesting its sensitivity to isotopic exchange (30). This is very likely given the molecular size of DMA, which upon adsorption would turn some neighboring sites inaccessible for further adsorption and hence available for isotopic exchange. DFT Calculations on Hydrated Inner-Sphere DMA-Fe(oxyhydr)oxide Clusters. Figures S4 and S5 show energyminimized equilibrium geometries of hydrated DMA and inner- and outer-sphere DMA-Fe-(oxyhydr)oxide clusters, respectively. DFT calculations were performed for net neutral and positively charged complexes by varying the number of OH- groups on the Fe-(oxyhydr)oxide cluster. Corresponding calculations were performed on deuterated clusters for comparison with experiments carried out in D2O. Factors that affect the As-Fe interatomic distances are mainly the multiplicity, the location of the H2O ligands on the two Fe3+ atoms, and the number of hydrogen bonds to the iron oxide cluster for outer-sphere complexes. Final inner-sphere geometries with multiplicities greater than 1 and four H2O ligands in the same plane as the two Fe3+ atoms resulted in As-Fe distances [d(As-Fe))3.16-3.34 Å, Table S2). Values of d(As-Fe) in optimized outer-sphere complexes are 4.95-6.42 Å compared to 5.2-5.8 Å reported for outer-sphere iAs(V) (31). To date, there are no reports on d(As-Fe) for DMA adsorbed Fe-(oxyhydr)oxides from X-ray absorption studies. Arsenic-metal distances were reported for innersphere DMA complexes on AAO (14) and TiO2 (12) using EXAFS at pH 5. Values of d(As-Al) is 3.17 Å for a bidentatebinuclear configuration, and that of As-Ti is 3.37 Å for a monodentate configuration (12). The above distances are comparable with those calculated for the bidentate-binuclear

FIGURE 3. Correlation between experimental and calculated v(As-O) frequencies listed in Table 1 and Table S3, respectively. MD ) monodentate, BD ) bidentate binuclear, and OS ) outer-sphere. Numbers in parentheses correspond to total charge on clusters shown in Figure S5. and fully protonated iAs(V)-iron oxide cluster (3.21 Å, Figure S6 and Table S2). These distances are close to those reported by Sherman and Randall from calculations on the same cluster (3.29 Å) and from EXAFS fits (3.25-3.30 Å) (26, 31, 36). The values of d(As-OFe) in monodentate DMA clusters (A, B) are 1.70-1.73 Å compared to 1.71-1.72 Å in bidentate clusters (C, D). These values are shorter than d(As-OH) calculated for fully protonated and hydrated DMA (1.77 Å, Figure S4). This explains the higher frequency values observed experimentally and assigned to v(As-OFe) compared to v(As-OH) (Table 1). The values of d(As-Ouncomplexed) are 1.71 and 1.70 Å for monodentate DMA clusters A and B, respectively, due to H-bonding with explicit water molecules and solvation effects simulated by the IEFPCM model [compare to d(AsdO) ) 1.64 Å in isolated and fully protonated DMA (32)]. For outer-sphere cluster (E) containing fully deprotonated DMA, both As-O groups on DMA are uncomplexed and involved in extensive H-bonding resulting in d(As-O) ) 1.71 and 1.69 Å, which are close to the values calculated for hydrated and fully deprotonated DMA (Figure S4). As explained below, this type of cluster explains the high frequency v(As-O) ≈ 877 cm-1 observed experimentally. The outer-sphere cluster (F) simulates net positively charged surface site dominant under acidic conditions where DMA is fully protonated with d(As-O) ) 1.69 and 1.74 Å. Calculated values of v(As-O) are listed in Table S3 for the clusters shown in Figure S5 and were correlated with experimental values (Figure 3). For monodentate clusters, multiple components are assigned to v(As-OFe) and v(As--O-----H). The low symmetry of bidentate-binuclear clusters also results in multiple components assigned to v(As-OFe), which are red-shifted relative to those calculated for iAs(V) bidentate-binuclear clusters. This is due to sCH3 groups whose size and electron-donating abilities cause electrostatic repulsion that is greater than that for sOH groups. Hence As-OFe distances are longer in bidentate-binuclear DMA (1.71-1.72 Å) relative to those in iAs(V) (1.70 Å). Outer-sphere VOL. 44, NO. 20, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7805

complexes gave rise to v(As---O-----H) vibrations comparable to those calculated for inner-sphere complexes due to H-bonding with explicit water molecules. The fewer the number of H-bonds in the cluster, the shorter the As-O bond responsible for v(As-O) > 860 cm-1. In summary, these DFT calculations further support the experimental spectroscopic data on the simultaneous formation of inner- and outersphere surface complexes of DMA on Fe-(oxyhydr)oxides. Environmental Significance. Results reported herein are the first to examine the binding of DMA to Fe-(oxyhydr)oxides as a function of pH and I using ATR-FTIR, complemented with quantum chemical calculations to identify the structure of DMA(ads). Equations 1 and 2 show the ligand exchange reactions that are likely to occur under basic and acidic conditions, respectively, forming neutral surface complexes ≡FeOH+ 2 / ≡ FeOH + Me2 - AsO2 f [≡FeO2AsMe2] +

H2O/OH- (1) ≡FeOH+ 2 / ≡ FeOH + Me2 - AsO2H f [≡FeO2AsMe2] + H3O+/H2O

(2)

where Me)CH3 group. Inner-sphere complexes are identified by frequencies listed in Tables 1 and S3. Formation of positively charged surface complexes is also likely under acidic conditions + ≡FeOH+ 2 / ≡ FeOH + Me2 - AsO2H f [≡FeO2HAsMe2] +

H2O/OH- (3) Our data also suggest the formation of outer-sphere complexes, which are characterized by v > 860 cm-1 (Tables 1 and S3)

≡FeOH2+ /

≡ FeOH + Me2 -

[≡FeOH2+

- Me2 -

AsO2-]

AsO2- /Me2

- AsO2H f

+ [≡FeOH - Me2 - AsO2-] (4)

Integration of our results reported herein with results of published batch experiments could be used to build SCMs for DMA adsorption similar to those for other oxyanions (37). Quantum chemical calculations on a number of ligand exchange reactions are currently underway to estimate binding thermodynamics of these clusters following ref 38. Smax[DMA] was estimated on the Fe-(oxyhydr)oxides used in our experiments at pH 7 and I ) 0.01 M using ATR-FTIR and bulk experiments (SI). Average values of the maximum net absorbance at 840 cm-1 [Amax(0.0012 cm)] are 0.018(8) and 0.004(1) using R-Fe2O3 and R-FeOOH, respectively. Using eq S1, these values correspond to 3(1) × 1013 and 2(0.5) × 1013 molecules · cm-2, respectively, which are equivalent to 0.8(3) and 0.6(2) mmol As/mol Fe. From batch isotherm experiments, Lafferty and Loeppert (11) reported 50 and 0.4 mmol As/mol Fe for DMA adsorption on ferrihydrite and goethite at pH 7 and I ) 0.044 M (NaNO3), respectively. For comparison with Smax[iAs(V)] and Smax[MMA], the same group (11) reported 105 and 7 mmol As/mol Fe and 94 and 7, respectively, on the aforementioned adsorbents. The surface area of the solid substrates was not reported, and hence molar ratios could not be normalized to surface area for crosscomparison. A study by Gimenez et al. (39) reported molar ratios of 1 and 0.5 mmol As/mol Fe for Smax[iAs(V)] on hematite and goethite, respectively, at pH 7 and I ) 0.1 M NaCl. These values correspond to 2 × 1015 and 2 × 1014 molecules · cm-2 when normalized to the surface area of the solids. As discussed by Shimizu et al. (14), the trend observed in values of Smax[iAs(V)] relative to DMA for a given substrate could be explained by the differences in molecular structure and not 7806

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 20, 2010

necessarily the presence of different types of surface complexes. The more methyl groups on arsenate, the less the adsorption affinity to metal oxides, which also gives rise to slower rates of adsorption for methylated arsenicals relative to iAs. Our data suggest that under neutral conditions, small anions such Cl- (0.01 M) are capable of desorbing mostly outer-sphere DMA(ads), which is enhanced using phosphate concentrations as low as 10-4 M. Desorption of strongly bonded DMA(ads) requires higher phosphate concentrations, similar to those used in batch studies (0.1 M) (11, 40, 41). Hence, under neutral to acidic geochemical conditions with relatively high Fe and Al content and low P conditions, outersphere complexes become bioavailable, and transport of colloidal or nanosize particles with strongly bonded DMA(ads) could become an important transport mechanism. Under high P conditions, DMA becomes mobalized and readily bioavailable. Technologies aimed at removing DMA could be designed to lower the arsenic content of organic-rich fuels. For example, the As-content could be reduced by washing fuels with slurries of Fe-(oxyhdr)oxides instead of water alone, and contaminated particles could be collected, recycled, or compressed into pellets.

Acknowledgments The authors acknowledge funding from the WLU Science and Technology Endowment Program (STEP), Research Corporation Cottrell College Award, and NSERC. Acknowledgment is made to the donors of the American Chemical Society Petroleum Research Fund for support (or partial support) of this research. A.A. and W.M. have contributed equally to this work and hence are listed alphabetically.

Note Added after ASAP Publication This paper was published ASAP on September 21, 2010. A misspelled author name was corrected. The revised paper was reposted on September 28, 2010.

Supporting Information Available Adsorption kinetic studies, ATR-FTIR spectra of DMA(aq), quantification of SDMA from ATR-FTIR, and results of DFT calculations of hydrated DMA- and iAs(V)-iron oxide clusters. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Ravenscroft, R.; Brammer, H.; Richards, K. Arsenic pollution: A global synthesis; Wiley-Blackwell: Malden, MA, 2009. (2) Cramer, S. P.; Siskin, M.; Brown, L. D.; George, G. N. Characterization of arsenic in oil shale and oil shale derivatives by x-ray absorption spectroscopy. 1988, 2, 175–180. (3) Baernthaler, G.; Zischka, M.; Haraldsson, C.; Obernherger, I. Determination of major and minor ash-forming elements in solid biofuels. Biomass Bioenergy 2006, 30, 983–997. (4) Ponthieu, M.; Pinel-Raffaitin, P.; Le Hecho, I.; Mazeas, L.; Amouroux, D.; Donard, O.; Pontin-Gautier, M. Speciation analysis of arsenic in landfill leachate. Water Res. 2007, 41, 3177– 3185. (5) Huang, J.-H.; Ilgen, G.; Vogel, D.; Michalzik, B.; Hantsch, S.; Tennhardt, L.; Bilitewski, B. Emissions of inorganic and organic arsenic compounds via leachate pathway from pretreated municipal waste materials: A landfill reactor study. Environ. Sci. Technol. 2009, 43, 7092–7097. (6) Pelley, J. Common arsenical pesticide under scrutiny. Environ. Sci. Technol. A-Pages 2005, 39, 122A–123A. (7) Fish, R. H.; Walker, W.; Tannous, R. S. Organometallic geochemistry. 2. The molecular characterization of trace organometallic and inorganic compounds of arsenic found in green river formation oil shale and its pyrolysis product. Energy Fuels 1987, 1, 243–247. (8) Sierra-Alvarez, R.; Yenal, U.; Field, J. A.; Kopplin, M.; Gandolfi, A. J.; Garbarino, J. R. Anaerobic biotransformation of organoarsenical pesticides monomethylarsonic acid and dimethylarsinic acid. J. Agric. Food Chem. 2006, 54, 3959–3966.

(9) Datta, R.; Sarkar, D.; Sharma, S.; Sand, K. Arsenic biogeochemistry and human health risk assessment in organo-arsenical pesticide-applied acidic and alkaline soils: An incubation study. Sci. Total Environ. 2006, 372, 39–48. (10) Cox, C. D.; Ghosh, M. M. Surface complexation of methylated arsenates by hydrous oxides. Water Res. 1994, 28, 1181–1188. (11) Lafferty, B. J.; Loeppert, R. H. Methyl arsenic adsorption and desorption behavior on iron oxides. Environ. Sci. Technol. 2005, 39, 2120–2127. (12) Jing, C.; Xiaogunag, M.; Liu, S.; Baidas, S.; Patraju, R.; Christodoulatos, C.; Korfiatis, G. P. Surface complexation of organic arsenic on nanocrystalline titanium oxide. J. Colloid Interface Sci. 2005, 290, 14–21. (13) Zhonghou, X.; Jing, C.; Li, F.; Meng, X. Mechanisms of photocatalytical degradation of monomethylarsonic and dimethylarsinic acids using nanocrystalline titanium dioxide. Environ. Sci. Technol. 2008, 42, 2349–2354. (14) Shimizu, M.; Ginder-Vogel, M.; Parikh, S. J.; Sparks, D. L. Molecular scale assessment of methylarsenic sorption on aluminum oxide. Environ. Sci. Technol. 2010, 44, 612–617. (15) Lim, S.-F.; Zheng, Y.-M.; Chen, J. P. Organic arsenic adsorption onto magnetic sorbent. Langmuir 2009, 25, 4973–4978. (16) Ramesh, A.; Hasegawa, H.; Maki, T.; Ueda, K. Adsorption of inorganic and organic arsenic from aqueous solutions by polymeric Al/Fe modified montmorillonite. Sep. Sci. Technol. 2007, 56, 90–100. (17) Chabot, M.; Hoang, T. N.; Al-Abadleh, H. A. ATR-FTIR studies on the nature of surface complexes and desorption efficiency of p-arsanilic acid on iron-(oxyhydr)oxides. Environ. Sci. Technol. 2009, 43, 3142–3147. (18) Kenyon, E. M.; Hughes, M. F. A concise review of the toxicity and carcinogenicity of dimethylarsinic acid. Toxicology 2001, 160, 227–236. (19) Depalma, S.; Cowen, S.; Hoang, T. N.; Al-Abadleh, H. A. Adsorption thermodynamics of p-arsanilic acid on iron (oxyhydr)oxides: In-situ ATR-FTIR studies. Environ. Sci. Technol. 2008, 42, 1922–1927. (20) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A. Gaussian 09, revision a.1; Gaussian, Inc.: Wallingford, CT, 2009. (21) Shared hierarchical academic research computing network (Sharcnet). www.sharcnet.ca. (22) Mennucci, B.; Tomasi, J.; Cammi, R.; Cheeseman, J. R.; Frisch, M. J.; Devlin, F. J.; Gabriel, S.; Stephens, P. J. Polarizable continuum model (PCM) calculations of solvent effects on optical rotations of chiral molecules. J. Phys. Chem. A 2002, 106, 6102–6113. (23) Cowen, S.; Duggal, M.; Hoang, T. N.; Al-Abadleh, H. A. Vibrational spectroscopic characterization of some environmentallyimportant organoarsenicals: A guide for understanding the nature of their surface complexes. Can. J. Chem. 2008, 86, 942– 950. (24) Myneni, S. C. B.; Traina, S. J.; Waychunas, G. A.; Logan, T. J. Experimental and theoretical vibrational spectroscopic evaluation of arsenate coordination in aqueous solutions, solids, and at the mineral-water interfaces. Geochim. Cosmochim. Acta 1998, 62, 3285–3300, and references therein.

(25) Roddick-Lanzilotta, A. J.; McQuillan, A. J.; Craw, D. Infrared spectroscopic characterization of arsenate (V) ion adsorption from mine waters, macraes mine, new zealand. Appl. Geochem. 2002, 17, 445–454. (26) Loring, J. S.; Sandstrom, M. H.; Noren, K.; Persson, P. Rethinking arsenate coordination at the surface of goethite. Chem.sEur. J. 2009, 15, 5063–5072. (27) Jia, Y.; Xu, L.; Wang, X.; Demopoulos, G. P. Infrared spectroscopic and x-ray diffraction characterization of the nature of adsorbed arsenate on ferrihydrite. Geochem. Cosmochim. Acta 2007, 71, 1643–1654. (28) Glodberg, S.; Johnston, C. T. Mechanisms of arsenic adsorption on amorphous oxides evaluated using macroscopic measurements, vibrational spectroscopy, and surface complexation modeling. J. Colloid Interface Sci. 2001, 234, 204–216. (29) Harrison, J. B.; Berkheiser, V. E. Anion interactions with freshly prepared hydrous iron oxides. Clays Clay Miner. 1982, 30, 97–102. (30) Lumsdon, D. G.; Fraser, A. R.; Russell, J. D.; Livesey, N. T. New infrared band assignments for the arsenate ion adsorbed on synthetic goethite (R-FeOOH). J. Soil Sci. 1984, 35, 381–386. (31) Catalano, J. G.; Park, C.; Fenter, P.; Zhang, Z. Simultaneous inner- and outer-sphere arsenate adsorption on corundum and hematite. Geochem. Cosmochim. Acta 2008, 72, 1986–2004. (32) Adamescu, A.; Gray, H.; Stewart, K.; Hamilton, I. P.; Al-Abadleh, H. A. Trends in the frequencies of ν(AsOxHx-1) [x ) 2-4] in selected As(V)-containing compounds investigated using quantum chemical calculations. Can. J. Chem. 2010, 88, 65–77. (33) Lefevre, G. In situ Fourier-transform infrared spectroscopy studies of inorganic ions adsorption on metal oxides and hydroxides. Adv. Colloid Interface Sci. 2004, 107, 109–123. (34) Weckler, B.; Lutz, H. D. Lattice vibration spectra. Part XCV. Infrared spectroscopic studies on the iron oxide hydroxides goethite(R), akaganeite(β), lepidocrocite(γ), and feroxyhite(δ). Eur. J. Solid State Inorg. Chem. 1998, 35, 531–544. (35) Ikawa, S. I.; Maeda, S. Infrared intensities of stretching and librational bands of H2O, D2O and HDO in solids. Spectrochim. Acta 1968, 24A, 655–665. (36) Sherman, D. M.; Randall, S. R. Surface complexation of arsenic(V) to iron(III) (hydr)oxides: Structural mechanism for Ab Initio molecular geometries and EXAFS spectroscopy. Geochem. Cosmochim. Acta 2003, 67, 4223–4230. (37) Goldberg, S.; Hyun, S.; Lee, L. S. Chemical modeling of arsenic(III, V) and selenium(IV, VI) adsorption by soils surrounding ash disposal facilities. Vadose Zone J. 2008, 7, 1231–1238. (38) Kubicki, J. D.; Kwon, K. D.; Paul, K. W.; Sparks, D. L. Surface complex structures modelled with quantum chemical calculations: carbonate, phosphate, sulphate, arsenate and arsenite. Eur. J. Soil Sci. 2007, 58, 932–944. (39) Gimenez, J.; Martinez, M.; de Pablo, J.; Rovira, M.; Duroc, L. Arsenic sorption onto natural hematite, magnetite, and goethite. J. Hazard. Mater. 2007, 141, 575–580. (40) Zhang, J. S.; Stanforth, R.; Pehkonen, S. O. Irreversible adsorption of methyl arsenic, arsenate, and phosphate onto goethite in arsenic and phosphate binary systems. J. Colloid Interface Sci. 2008, 317, 35–43. (41) Jackson, B. P.; Miller, W. P. Effectiveness of phosphate and hydroxide for desorption of arsenic and selenium species from iron oxides. Soil Sci. Soc. Am. J. 2000, 64, 1616–1622.

ES1011516

VOL. 44, NO. 20, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7807