Article pubs.acs.org/est
Uranium(VI) Reduction by Iron(II) Monosulfide Mackinawite Sung Pil Hyun,†,* James A. Davis,‡ Kai Sun,§ and Kim F. Hayes† †
University of Michigan, Civil and Environmental Engineering, Ann Arbor, Michigan 48109 Lawrence Berkeley National Laboratory, Earth Sciences Division, Berkeley, California 94720 § University of Michigan, Electron Microbeam Analysis Laboratory, Ann Arbor, Michigan 48109 ‡
S Supporting Information *
ABSTRACT: Reaction of aqueous uranium(VI) with iron(II) monosulfide mackinawite in an O2 and CO2 free model system was studied by batch uptake measurements, equilibrium modeling, and LIII edge U X-ray absorption spectroscopy (XAS). Batch uptake measurements showed that U(VI) removal was almost complete over the wide pH range between 5 and 11 at the initial U(VI) concentration of 5 × 10−5 M. Extraction by a carbonate/bicarbonate solution indicated that most of the U(VI) removed from solution was reduced to nonextractable U(IV). Equilibrium modeling using Visual MINTEQ suggested that U was in equilibrium with uraninite under the experimental conditions. X-ray absorption near edge structure (XANES) and extended Xray absorption fine structure (EXAFS) spectroscopy showed that the U(IV) phase associated with mackinawite was uraninite. Oxidation experiments with dissolved O2 were performed by injecting air into the sealed reaction bottles containing mackinawite samples reacted with U(VI). Dissolved U measurement and XAS confirmed that the uraninite formed from the U(VI) reduction by mackinawite did not oxidize or dissolve under the experimental conditions. This study shows that redox reactions between U(VI) and mackinawite may occur to a significant extent, implying an important role of the ferrous sulfide mineral in the redox cycling of U under sulfate reducing conditions. This study also shows that the presence of mackinawite protects uraninite from oxidation by dissolved O2. The findings of this study suggest that uraninite formation by abiotic reduction by the iron sulfide mineral under low temperature conditions is an important process in the redistribution and sequestration of U in the subsurface environments at U contaminated sites.
■
INTRODUCTION Many United States Department of Defense and Department of Energy sites are contaminated with uranium (U) from the legacy of weapons manufacturing during the Cold War era and nuclear fuel processing.1−3 U is a redox-active actinide element, and redox conditions and speciation strongly affect the reactivity and mobility of U. Some U remediation approaches focus on converting and maintaining U in its tetravalent oxidation state because of the formation of sparingly soluble, immobile U(IV) mineral phases, such as uraninite. It is reported that under iron reducing conditions achieved either naturally or by biostimulation, U was successfully removed from the groundwater.4−6 To achieve iron reducing conditions, organic compounds such as acetate may be injected into a contaminated aquifer to stimulate microorganisms that reduce ferric oxides and hydroxides to Fe(II). However, from a longer term perspective, redox conditions may be subject to change. For example, at the Old Rifle UMTRA (Uranium Mill Tailings Remedial Action) site in Colorado, sulfate reducing conditions prevailed after acetate injection was over, following iron reducing conditions.4 Under the sulfate reducing condition, it was observed that dissolved U(VI) concentrations had an initial spike followed by decreased concentrations. Understanding abiotic processes controlling U mobility under sulfate reducing conditions is an important step toward the assessment of the long-term performance of current U remediation practices. Previous experimental studies reported partial reduction of U(VI) by a variety of iron(II) sulfide minerals, including amorphous iron sulfide, mackinawite, and pyrite.7−9 Among © 2012 American Chemical Society
various iron sulfide minerals, mackinawite, a tetragonal ferrous monosulfide (FeS1−x), is the first ferrous sulfide solid phase to form under sulfate reducing conditions.10,11 It is a precursor to other stable iron sulfide minerals, such as pyrite and greigite. In this study we examined U(VI) reactions with the ferrous monosulfide mineral mackinawite. Understanding U reactions with mackinawite is important in assessing U sequestration mechanisms by geochemical or biogeochemical processes under sulfate reducing conditions. The objective of this study was to investigate the mechanisms of U(VI) sorption reactions with mackinawite under variable pH using batch uptake data, extractions by carbonate solution, equilibrium modeling, and Xray absorption spectroscopy (XAS). Specifically, the study aimed to test the hypothesis that iron(II) sulfide mackinawite only partially reduces U(VI), with both reduction to U(IV) phases and simple adsorption of U(VI) responsible for U removal from a simple electrolyte solution in a CO2 and O2 free system. The study also sought to test the hypothesis that oxidative mobilization of U previously reacted with mackinawite would be inhibited by the presence of mackinawite.
■
MATERIALS AND METHODS Mackinawite Synthesis. Mackinawite was synthesized by mixing 2.0 L of 0.57 M Fe(II) solution with 1.2 L of 1.1 M S(-
Received: Revised: Accepted: Published: 3369
October 25, 2011 January 8, 2012 February 9, 2012 February 9, 2012 dx.doi.org/10.1021/es203786p | Environ. Sci. Technol. 2012, 46, 3369−3376
Environmental Science & Technology
Article
Anoxic Carbonate (CARB) Extraction. The 5 mL of the mackinawite suspension after the U(VI) uptake experiment was transferred using a pipetter to a 15 mL PP centrifuge tube predispensed with 5 mL of a dilute bicarbonate/carbonate (CARB) extraction solution. The CARB solution was prepared by dissolving sodium bicarbonate (28.8 mM) and sodium carbonate (5.60 mM) in deoxygenated water, following the recipe by Kohler et al. (2004),15 but modified to use anoxic water and twice as concentrated to compensate for the dilution. The sample/extractant mix was vortexed for a minute and then allowed to react for 1 h on an end-over-end rotator in the anaerobic chamber. The short extraction time was used to ensure that the extraction only extracted weakly bound U(VI) species. After 1 h, the suspension pH was measured, the suspension was filtered using the 0.1 μm nylon syringe filter (GE Cameo), and the filtered solution was measured for dissolved U concentration using inductively coupled plasmamass spectrometry. Thermodynamic Modeling. Visual MINTEQ (ver. 3.0) was used for the equilibrium modeling of the U(VI) uptake data by mackinawite under different pH conditions. The thermodynamic constants for mackinawite and U species were from visual MINTEQ version 3.0 databases. These constants were originally derived from the MINTEQA2 databases, revised using NIST version 6.0 and 7.0 databases. The U species constants used are based on the values in Guillaumont et al. (2003) by OECD/NEA.16 XAS Data Collection and Analysis. U LIII edge X-ray absorption spectra were collected at the beamlines 11-2 and 4-1 of Stanford Synchrotron Radiation Lightsource (SSRL). Wet paste samples were prepared and sent to SSRL in glass vials tightly sealed with rubber septa and Al crimp caps in a vacuum vessel. Sample preparation and loading were performed in an anaerobic chamber. Fluorescence spectra of the wet paste samples were collected using a 30-element, 13-element Ge detector or Lytle detector at the beam energy of 3.0 GeV with a beam current of 195−200 mA in a low temperature cryostat filled with liquid nitrogen. To minimize the contribution from higher order harmonics, the monochromator was detuned 30% at the highest energy position of the scans. The beam energy was calibrated using a simultaneously scanned Y standard foil. To obtain improved signal-to-noise ratios, multiple scans were collected. Data analyses were performed using FEFF8, IFEFFIT, EXAFSPAK, and SixPack.17−19 Acceptable signal channels of each scan were selected and the multiple scans were averaged after energy calibration against the Y foil K edge value of 17038 eV. Backgrounds were removed using linear fits below the absorption edge and spline fits above the edge using the SixPack code. The spectra were then converted from energy to frequency space using the photoelectron wave vector k typically in the range of 3 < k < 12. EXAFS fitting was performed using SixPack with phase and amplitude functions for backscattering paths obtained from FEFF8 calculations with crystallographic input files generated using ATOMS program. To obtain the optimal structural parameters, including the coordination number (CN) and interatomic distance (R), the Debye− Waller factor (σ2) and energy reference E0 parameter were also floated during the fitting. The many-body factor S02 was fixed at 0.9 to reduce the number of fitting parameters. The leastsquares fitting subroutine of SixPack was used to fit the sample XANES spectra using model spectra of uraninite, rutherfordine, and U(VI) solution. The goodness of the XAS fitting or the R factor of SixPack, which is the sum of squares of the remnant
II) solution prepared by dissolving ACS grade FeCl2.7H2O and Na2S.9H2O with anoxic deionized water in an anaerobic chamber filled with 4% H2 with balance N2, following the procedure described in Butler and Hayes (1998).12 All the glassware and plastic bottles were stored in the anaerobic chamber for at least 2 days before use to make sure that adsorbed O2 on the walls was removed in the anaerobic chamber. The resulting iron sulfide precipitate was allowed to age for 3 days while tightly capped and continuously mixed using a stir bar and magnetic plate. After aging, the precipitate was rinsed six times with anoxic deionized water. The rinsed iron sulfide paste was freeze-dried for three days to prevent further aging and to minimize oxidation. The dry powder was transferred and crimp sealed in a glass vial with a rubber septum and aluminum lid and stored in the anaerobic chamber until use. The resulting iron sulfide was nanocrystalline mackinawite; detailed mineralogical description of the mackinawite precipitate is given in Jeong et al. (2008).11 The synthesis method used in this study was selected after years of research.11−14 The method repeatedly gives synthetic mackinawite particles with consistent mineralogical characteristics. The resulting mackinawite sample is nanocrystalline, in contrast to amorphous or well crystalline.11 The particle size is in a several nanometer range, with a characteristic X-ray diffraction pattern. It may form micrometer-sized aggregates after freeze-drying. Freeze-drying stops the aging and the sample properties do not change for up to a year or longer when stored in a crimp sealed, airtight glass bottle in the anaerobic chamber. Batch U(VI) Uptake Measurements. U(VI) sorption experiments were performed in the anaerobic chamber (4% H2/balance N2) over the pH range between 5 and 11 with the initial U(VI) concentration of 5 × 10−5 M. The mackinawite suspension was prepared by resuspending the dry powder in anoxic, deionized water in a 5 g/L ratio. A homogeneous mixture was obtained by two days of rigorous mixing of the suspension with a magnet stirrer. Ten mL of the suspension was dispensed to 15 mL polypropylene (PP) centrifuge tubes using a pipetter. A 1000 mg/L uranyl nitrate solution (as U) was used as the primary U(VI) stock solution. Sodium chloride was added to adjust the ionic strength at 0.1 M and the pH was adjusted by adding specific amounts of 1 M HCl predetermined from a preliminary titration. The reaction tubes were tightly capped and continuously shaken on an end-over-end rotator for 48 h. After 48 h of sorption reaction, the pH of the suspension was measured. Then 5 mL of the mackinawite suspension was transferred by a pipetter to a 15 mL PP centrifuge tube for carbonate extraction. The remaining 5 mL suspension was filtered using a 0.1 μm nylon syringe filter (GE Cameo), and the filtered solution was measured for dissolved U concentration using inductively coupled plasma-mass spectrometry (ICP-MS; Perkin-Elmer Elan). Samples used for the XAS measurement were prepared at three different pH values, namely 5, 7, and 10, using basically the same procedures as the batch uptake experiments, but scaled up by using 50 mL mackinawite suspensions in 75 mL sealed glass bottles to enable the collection of enough samples and to conduct air injection experiments. The oxidation experiments were performed by injecting 1 or 5 mL air using an airtight syringe (Hamilton) into 75 mL sealed glass reaction bottles containing 50 mL mackinawite suspensions reacted with aqueous U(VI). Details are in the Supporting Information (SI). 3370
dx.doi.org/10.1021/es203786p | Environ. Sci. Technol. 2012, 46, 3369−3376
Environmental Science & Technology
Article
concentration of 5 × 10−5 M, uranium removal was complete over the entire experimental pH range between 5 and 11. Considering the reported pHpzc of mackinawite (∼7.5)20 and the changing dominant U(VI) aqueous species as a function of pH, a sorption pH edge or a sharp increase in the U(VI) uptake over a narrow pH range of 1 or 2 pH unit as a function of pH was expected in the reaction of U(VI) with mackinawite, similar to those typically observed for U(VI) adsorption by various minerals under oxic conditions. However, contrary to the expectation, no notable sorption pH edge was observed in the U(VI) uptake by mackinawite. Thermodynamic modeling, carbonate extraction of the solid bound U species, and XAS analysis were performed to further investigate the apparent pH independent U removal mechanism by mackinawite. CARB Extraction of U-loaded Mackinawite. U-reacted mackinawite samples were extracted using an anoxic 14.4 mM bicarbonate and 2.8 mM carbonate mixed solution (CARB) at pH 9.4, prepared following Kohler et al.’s method (2004)15 originally designed to extract reversibly adsorbed U(VI) from contaminated sediments, but modified for use under anoxic conditions. The extraction was intended to selectively extract the reversibly adsorbed fraction of solid-bound U(VI) species. The anoxic extraction is not expected to oxidize and dissolve U(IV) solid in the short duration of extraction used (1 h), given that even more concentrated (up to 1 M) carbonate solution did not dissolve uraninite without being first oxidized using strong oxidizing agents such as hydrogen peroxide, molecular oxygen or potassium permanganate.15,21−23 The extraction result showed that less than 10% of the U associated with mackinawite was extracted, regardless of the sorption pH (Figure 1). The result shows that CARB-extractable, reversibly and loosely bound U(VI) is a minor U species, whereas the nonextractable, irreversibly bound U phase is a major U species in the U(VI) reaction with mackinawite. The possible irreversibly bound U species may include a U(IV) solid phase, a mackinawite surface bound molecular U(IV) species, and U(VI) species incorporated into the structure of mackinawite or its oxidation products. To further investigate the operating U uptake mechanism, equilibrium modeling and U LIII edge XAS characterization of the solid phase were performed. Equilibrium Modeling of Uranium Speciation. Visual MINTEQ was used to model equilibrium U speciation under the experimental conditions. The equilibrium modeling of 5 × 10−5 M dissolved U(VI) reacting with 5 g/L mackinawite suspension in 0.1 M NaCl solution at the fixed pH of 7 yielded an equilibrium pe value of −1.629 (Eh = −96.35 mV). Under these conditions, the calculation predicted that U existed as uraninite (5 × 10−5 M), Fe(II) existed as 99.9% mackinawite, and elemental S formed in an equal molar quantity (5 × 10−5 M) to uraninite. Under the same conditions, but at pH 5, the equilibrium pe was computed to be −1.022 (Eh = −60.45 mV), with 96.3% of the initial mackinawite present as mackinawite, U as uraninite (5 × 10−5 M), and formation of elemental S (5 × 10−5 M). Under the same condition, but with pH fixed at 9, the equilibrium pe was computed to be −1.684 (Eh = −99.60 mV). Under this condition, Fe(II) existed as mackinawite (99%), U as uraninite (5 × 10−5 M), and elemental S formed in an equivalent amount to uraninite (5 × 10−5 M). Consistent with the variations in the pe value as a function of pH, the dominant dissolved U species changed from UO22+ to (UO2)3(OH)5+, and then UO2(OH)3− with increasing pH. In conclusion, thermodynamics predicted that uraninite formation was
errors or the differences between the fit and the measured data, was used to get the best fit results. A lower R value is indicative of a closer fitting to the measured data, hence better fitting. Transmission Electron Microscopy (TEM). A JEOL 2010 analytical electron microscope (AEM) operated at 200 kV was used for conventional bright-field (BF) TEM, high-resolution electron microscopy (HREM), high angle annular dark-field (HAADF, performed in scanning TEM mode) imaging, selected area electron diffraction (SAED), X-ray energy dispersive spectroscopy (XEDS), and electron energy loss spectroscopy (EELS). Holey carbon film coated 300-mesh Au grids were used for the TEM specimen preparation by dropping U-loaded mackinawite suspension on them and letting them dry in a vacuum desiccator filled with drierite in the anaerobic chamber. The prepared specimens were transferred in the vacuum desiccator to the TEM chamber to minimize possible oxygen exposure.
■
RESULTS Batch Uptake Results. U(VI) uptake by mackinawite as a function of pH is shown in Figure 1, along with the U(VI) aqueous speciation as a function of pH. At the initial uranium
Figure 1. The amounts of U(VI) removed by macknawite (filled diamond) as a function of solution pH in the reaction of 5 × 10−5 M U(VI) with a 5 g/L mackinawite suspension in 0.1 M NaCl electrolyte solution under an O2 and CO2 free atmosphere and the corresponding amounts of U extracted by CARB solution at pH 9.4 (open diamond) from the U-reacted mackinawite under the various pH condition (Upper panel). Lower panel is the initial U(VI) aqueous speciation under the experimental condition as a function of pH. U(VI) removal from solution is almost complete over the whole pH range tested, independent of the U(VI) aqueous speciation. 3371
dx.doi.org/10.1021/es203786p | Environ. Sci. Technol. 2012, 46, 3369−3376
Environmental Science & Technology
Article
Table 1. Structural Parameters Extracted from the EXAFS Analysis (CN: Coordination Number of Backscatterer Atoms around the Absorber U Atom, R: Interatomic Distance in Å Unit of the Absorber-Backscatterer Pair, σ2: Debye-Waller Factor of the Absorber-Backscatterer Pair) sample
pH 7
pH 5
pH 10
shell
description
pH 7.04
pH 5.06
pH 9.90
O
CN R σ2 CN R σ2 CN R σ2 CN R σ2
9.4(5) 2.33(5) 0.0120 4.4(4) 3.85(0) 0.0059
8.3(7) 2.32(1) 0.0108 3.6(4) 3.87(1) 0.0024
8.9(0) 2.33(2) 0.0117 4.4(3) 3.84(5) 0.0057
UIV
Oax
Oeq
U(VI) stock 1000 ppm uranyl solution
uraninite 25
X-ray diffraction. These values were used for Feff calculation. 8 2.3678 0 12 3.8666 0
uraninite synthetic (this study) 6.5(5) 2.25(5) 0.0164 2.5(2) 3.84(4) 0.0071
uraninite biogenic (this study) 7.2(8) 2.33(4) 0.0122 4.8(8) 3.83(7) 0.0105
UO2.00(s) 27
uraninite
synthetic
biogenic27
8 2.354(7) 0.0046(7) 12 3.867(4) 0.0029
8 2.347(7) 0.0103(6) 6(1) 3.846(5) 0.0055(7)
2.4(5) 1.76(6) 0.0030 4.8(3) 2.42(3) 0.0061
Figure 2. U LIII edge EXAFS spectra (left panel) and corresponding Fourier transform magnitude functions (right panel) of U(VI) reacted with mackinawite under different pH conditions along with model compound U(VI) aqueous solution and uraninite. A: Uranyl solution, B: U(VI) reacted with mackinawite at pH 5, C: at pH 7, D: at pH 10, E: uraninite model compound, and F: Feff 8.10 calculated uraninite EXAFS spectrum (Solid lines: experimental data, Dots: fitting results).
U valence state across the pH conditions, least-squares fitting of the sample XANES spectra was performed using the SixPack least-squares fitting module with the spectra of synthetic uraninite, rutherfordine, and aqueous U(VI) as component spectra. The combination of 95% reference uraninite spectrum and 5% U(VI) solution spectrum reconstructed almost 100% of the pH 7 sample spectrum, requiring no extra U component (SI Figure S1 inset). The U(VI) content is consistent with the CARB extraction results (Figure 1). Other pH condition samples were successfully simulated with the uraninite spectrum only. This result strongly suggests that the U(IV) phase in the sample was almost 100% composed of uraninite without other major U components such as U(V) species or monomeric U(IV). EXAFS Analysis of U(VI)-Reacted Mackinawite. The structural parameters of U(VI)-reacted mackinawite samples obtained from EXAFS analysis are given in Table 1. The
favorable over the whole range of pH conditions at the total concentration of 5 × 10−5 M U(VI) in equilibrium with 5 g/L mackinawite suspension. X-ray Absorption Spectroscopy Results. X-ray absorption spectroscopy was used to gather spectroscopic evidence for the solid bound U speciation in the reaction of aqueous U(VI) with mackinawite under the experimental conditions. XANES of U(VI)-REACTED Mackinawite. The XANES result for the U(VI) reacted with 5 g/L mackinawite suspension at the initial U(VI) concentration of 5 × 10−5 M shows that the U LIII edge position had the first derivative inflection point position at ca. 17171.5 eV compared to ca. 17173.5 eV of the original U(VI) solution (Supporting Information Figure S1). This observation of the edge position shift from the starting U(VI) solution to a lower energy suggests that U(VI) was reduced to U(IV) under the range of experimental pH conditions studied. For a more quantitative assessment of the 3372
dx.doi.org/10.1021/es203786p | Environ. Sci. Technol. 2012, 46, 3369−3376
Environmental Science & Technology
Article
Fe(II) sulfides have not been reported, underestimating the potential role the iron sulfide minerals may play in the redox chemistry of U under sulfate reducing conditions. This study shows reduction of dissolved U(VI) to uraninite by the Fe(II) sulfide, mackinawite. Quantification of the XANES spectra using least-squares fitting routine suggested that regardless of the reaction pH, U(VI) was reduced to U(IV). Using only the uraninite model compound spectrum, the XANES spectra of U(VI) reacted with mackinawite were well reproduced. EXAFS data fitting also suggested uraninite as the major U(IV) species. The oxidation experimental results show that mackinawite can serve as a redox buffer slowing the oxidation of uraninite when oxidants such as dissolved oxygen are introduced to the U−Fe−S system under sulfate reducing conditions (SI Table S1 and Figure S3). Results of U-reacted mackinawite oxidation showed that no detectable U release was observed from the oxygen injected reaction bottles, suggesting that mackinawite slows uraninite oxidation under the experimental conditions. An alternative explanation is that U is oxidized to U(VI) but is adsorbed by mackinawite or its oxidation products so that it is not released to solution. To test between the two possibilities, EXAFS spectroscopy was used. EXAFS analysis results of U reacted with mackinawite at pH 7 after the 1 and 5 mL air injection are given SI Table S1 and Figure S3, along with model compounds uraninite and U(VI) solution. The EXAFS structural parameters of U from the air injected bottles were very similar to those before air injection (Table 1). These results confirmed that U(IV) oxidation was inhibited in the presence of mackinawite under the experimental conditions. The injected O2 was not enough to oxidize all the mackinawite present, but enough to significantly oxidize uraninite if it was not consumed by reaction with mackinawite. The oxidation experiment was designed assuming a scenario with a low oxygen concentration, such as the one in the Rifle aquifer right after an acetate injection is over and oxygen starts to be introduced into the anoxic groundwater. The dissolved oxygen concentration will be low under such circumstances, justifying the low O2 level used in this study. Under such conditions, this study shows that uraninite is protected from oxidation in the presence of mackinawite. The changes in Fe mineralogy by reaction with U(VI) or subsequent oxidation by the injected air were not detectable by Mössbauer spectroscopy. For the same reason, the Mössbauer data are not included in this paper. Our research group performed a parallel study which focused on the oxidation of uraninite (0.5 g/L uraninite) by dissolved oxygen (0.7 mg/L) in the presence of mackinawite (5 g/L mackinawite) by continuously bubbling the uraninite/ mackinawite containing artificial groundwater with 2% PO2 gas.36 This work was intended to achieve significant up to complete oxidation of mackinawite and uraninite. The result showed that mackinawite delayed the oxidation of uraninite by consuming all the initial dissolved oxygen. Once all the mackinawite was oxidized to goethite/lepidocrocite, uraninite was oxidatively dissolved by the dissolved oxygen. The unifying conclusion of this current paper and Bi et al. (2011)36 is that when mackinawite is present in excess of uraninite, it effectively scavenges the available oxygen, and inhibits uraninite oxidation. Hua et al. (2008) reported U(VI) reaction with amorphous iron sulfide.8 They suggested U(VI) uptake by amorphous FeS by an ion exchange mechanism resulting in the release of Fe(II) to solution followed by the reduction of U(VI) to U3O8, U4O9, and UO2 according to X-ray photoelectron spectroscopic
EXAFS structural parameters are consistent with the model compound uraninite as well as the theoretical uraninite EXAFS spectrum calculated using the Feff 8.10 code and the well crystalline uraninite structure determined by X-ray diffraction.24,25 The central U(IV) has 8−9 first neighbor O atoms at the interatomic distance of 2.32−2.34 Å (Table 1 and Figure 2). These first shell structural parameters are in good agreement with uraninite values reported in the literature for either abiotic- or biotic-synthesized phases.26−29 This first shell feature is distinct from aqueous U(VI) species, which has characteristic features in the spectra from two nearest axial O atoms at 1.77 Å and six equatorial O atoms at 2.42 Å (or 4 nearer equatorial O atoms at 2.25−2.43 Å and 2 farther equatorial O atoms at 2.42−2.52 Å).30,31 The central U in the U(VI)-reacted mackinawite has a strong second shell feature composed of 3.6−4.4 U(IV) neighbors at the interatomic distances of 3.85−3.87 Å. This distinct second shell U neighbor feature is absent from the monomeric U(IV) species recently reported in a P-ligand bound biogenic U reduction products.32−34 This feature confirms that the major U phase in this study is uraninite rather than the monomeric U(IV) species. However, the second shell coordination number (CN) is less than that expected for a well crystallized, hydrothermal uraninite with 12 nearest U(IV) neighbors at 3.8666 Å within its isometric unit cell determined by XRD analysis.25 This low CN for the nearest shell U(IV) neighbors is consistent with low temperature abiotic uraninite and is considered to be due to the nanoscale nature of the uraninite precipitate.24 This conclusion is also in good agreement with recently reported biogenic uraninite samples.26−29 Considering the low second shell U CN and higher Debye−Waller factor associated with the second shell U bond compared to a well crystallized hydrothermal origin specimen, the uraninite formed is expected to have less ordered structure. XRD did not show any crystalline features from uraninite, most likely due to an insufficient amount of the uraninite formed (0.24 wt %). Detailed XRD results are given in SI Figure S2. It slightly improved the fit to include O−U−O multiple scattering path at the interatomic distance of 3.636 Å, consistent with the uraninite structure. However, the improvement was not significant enough considering the increased number of fitting parameters. Given this, the structural parameters related to the path are not included in the final results presented in Table 1. The U associated with mackinawite had slightly higher LIII Xray absorption edge jump and EXAFS oscillation amplitudes compared to the reference uraninite produced by homogeneous reduction of U(VI) by aqueous sulfide (Figure 2 and SI Figure S1). One possible explanation for this is due to self-absorption in the model compound uraninite data, even though caution was taken to minimize the effect during X-ray absorption data collection by diluting the uraninite model with boron nitride (BN). The EXAFS parameters of the U(VI)-reacted mackinawite samples after extraction with the dilute bicarbonate/ carbonate solution suggest that the remnant U phase associated with the solid phase was uraninite. The EXAFS results of oxidation experiments are given in the Supporting Information (SI Table S1 and Figure S3).
■
DISCUSSION In previous studies, only limited and partial reduction of U(VI) to U(V), or mixed U(VI) and U(IV) were reported in the reaction of U(VI) with Fe(II) sulfide minerals.7,9,35 Extensive and complete reduction of dissolved U(VI) to uraninite by 3373
dx.doi.org/10.1021/es203786p | Environ. Sci. Technol. 2012, 46, 3369−3376
Environmental Science & Technology
Article
Figure 3. TEM results of U(VI) reacted with mackinawite at 140 μM U and 1 mM mackinawite and pH 7.3. (A) BF image of uraninite nanoparticles (smaller, dark particles) associated with mackinawite particles (larger, brighter particles). (B) HAADF image showing uraninite (bright spots) association with mackianwite (dark spots). (C) SAED pattern confirming the U(VI) phase was nanocrystalline uraninite. (D) HREM lattice fringe image directly showing the nanocrystalline nature of uraninite closely associated with elemental S.
diffraction (SAED) pattern of the U-rich particles. The pattern is interpreted to be an isometric unit cell with a dimension of 5.46 Å, consistent with uraninite (UO2+x). High resolution lattice fringe imaging shows close microtextural relations between the resulting uraninite and elemental S suggesting elemental S as the oxidation product under the experimental condition (Figure 3D). Based on the findings of this study, we propose that U(VI) reduction mechanism similar to homogeneous reduction by aqueous sulfide takes place during U(VI) reduction by mackinawite, following U(VI) adsorption onto mackinawite and simultaneous release of Fe(II). This mechanism is consistent with the one proposed by Hua et al. (2008)8 for U(VI) reaction with amorphous FeS. Once sorbed to the mackinawite surface (1), either the surface U(VI) is reduced by S(-II) at the Fe(II)-depleted mackinawite surface (2) or the dissolved U(VI) is reduced by dissolved HS− (4) released by congruent dissolution of mackinawite (3), following the reactions below.
results. However, they could not conclusively verify the U(VI) reduction mechanism by amorphous FeS following U(VI) uptake. They considered reduction by aqueous Fe(II), aqueous HS−, adsorbed Fe(II), solid phase HS−, and solid phase Fe(II). They concluded aqueous Fe(II), aqueous HS−, and adsorbed Fe(II) as negligible or minor contributors to U(VI) reduction. They considered reduction by solid phase Fe(II) or HS− as the major U reduction mechanism, but could not differentiate between the two mechanisms. Transmission electron microscopic results of U reacted with mackinawite at a higher initial U concentration of 1.4 × 10−4 M and lower mackinawite loading of 1 mM are given in Figure 3. This sample was prepared by reacting dissolved U(VI) with sulfide solution and then adding ferrous solution to precipitate mackinawite. For this higher U concentration and much lower Fe/U ratio condition, TEM was successfully used to study the microstructural characteristics of U association with mackinawite. Figure 3A is a conventional bright-field (BF) image showing U distribution on the mackinawite aggregates. The smaller, dark spots are U-rich parts and the bigger, lighter parts are mackinawite. The micrograph shows that U rich particles are rather evenly distributed over the mackinawite particles. Figure 3B is a high angle annular dark-field (HAADF) image taken from the same area as imaged in the Figure 3A (rotated by 90 degrees) showing U association with mackinawite. In the HAADF image, the brighter areas are U rich and the darker areas are the lighter elemented mackinawite. Once again, the micrograph confirms the even distribution of U-rich particles over mackinawite particles. Figure 3C is a selected area electron 3374
UO2 2 + + ≡ FeS ↔ ≡ S2 − − UO2 2 + + Fe2 +
(1)
≡S2 − − UO2 2 + ↔ S0(s) − UO2(s)
(2)
FeS(s) + H2O ↔ Fe2 + + HS− + OH−
(3)
UO2 2 + + HS− ↔ UO2(s) − S0(s) + H+
(4)
dx.doi.org/10.1021/es203786p | Environ. Sci. Technol. 2012, 46, 3369−3376
Environmental Science & Technology The resulting reaction products are texturally related nanoscale uraninite and elemental S. Considering that mackinawite solubility increases with decreasing pH, mechanism (4) may work more efficiently at lower pH conditions. This is consistent with the XRD results (SI Figure S2). We propose that when mackinawite is reacted with aqueous U(VI), Fe(II) is released to solution in an exchange reaction with U(VI), followed by a reduction of U(VI) by S(-II) producing elemental S and uraninite without noticeable changes in the iron mineralogy under the experimental conditions. Bernier-Latmani et al. (2010)32 reported the occurrence of non-uraninite U(IV) species during microbial U(VI) reduction. Their study reported the inhibition of uraninite formation by Pcontaining ligands, promoting monomeric U(IV) complexes bound to biomass. Fletcher et al. (2010)33 also reported monomeric U(IV) species closely associated with carbonate and phosphate in the biogenic U(VI) reduction by Desulfitobacterium species. However, uraninite formation was the dominant mechanism of abiotic U(VI) reduction by mackinawite in the O2 and CO2-free model system used in this study. This study suggests that U(VI) reduction by the iron sulfide mineral mackinawite may be an important process occurring under certain ambient field conditions, such as at the Old Rifle site. Qafoku et al. (2009)37 reported U association with framboidal pyrite as the major U retention mechanism in aquifer sediment samples collected from naturally reduced zones occurring at this site. There is no consensus yet in the literature on the origin of framboidal pyrite formation; nor is it clear how U is associated with the pyrite framboids. However, it is well-known that nanocrystalline mackinawite is the first ferrous sulfide mineral to form under sulfate reducing conditions relevant to field conditions10,38 and is a precursor to other more stable ferrous sulfide minerals such as greigite and pyrite.39,40 Therefore U(VI) interaction with mackinawite is a possible process leading to the U association with pyrite framboids observed at the Rifle site. Furthermore, from a longterm perspective, this phase may occur as an initial stage prior to evolution to pyrite formation and with accompanying changes in U association with changing iron mineralogy. The resistance of uraninite against oxidation in the presence of sufficient mackinawite also supports the retention of U in the solid phase during oxidative transformation of mackinawite to pyrite. Therefore, the U association with mackinawite reported in this study suggests a plausible mechanism for initial U association with nanoscale mackinawite particles leading to an eventual U association with framboidal pyrite, as found at the Rifle site.
■
ACKNOWLEDGMENTS
■
REFERENCES
This research was supported by the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research (BER), Subsurface Biogeochemical Research (SBR) program (DE-FG02-09ER64803). We thank the SSRL beamline staff, including John Bargar, Juan Lezama-Pacheco, and Joe Rogers for their support in the XAS data collection. We thank Julian Carpenter, Tara Clancy, and Yuqiang Bi for their help in the XAS data collection. We also thank Marcia Torres, Ian Evans, Aaron Gooch, Carol Morris, Cindy Patty, and Joe Miklos for the safe shipping and handling of the U-containing samples. Portions of this research were carried out at the Stanford Synchrotron Radiation Laboratory, a national user facility operated by Stanford University on behalf of the U.S. Department of Energy, Office of Basic Energy Sciences. The SSRL Structural Molecular Biology Program is supported by the Department of Energy, Office of Biological and Environmental Research, and by the National Institutes of Health, National Center for Research Resources, Biomedical Technology Program.
(1) Catalano, J. G.; Brown, G. E. Jr. Analysis of uranyl-bearing phases by EXAFS spectroscopy: Interferences, multiple scattering, accuracy of structural parameters, and spectral differences. Am. Mineral. 2004, 89, 1004−1021. (2) Morrison, S. J.; Cahn, L. S. Mineralogical residence of alphaemitting contamination and implications for mobilization from uranium mill tailings. J. Contam. Hydrol. 1991, 8, 1−21. (3) Riley, R. G.; Zachara, J. M.; Wobber, F. J. Chemical Contaminants on DOE Lands and Selection of Contaminant Mixtures for Subsurface Science Research; U.S. Department of Energy, 1992. (4) Anderson, R. T.; Vrionis, H. A.; Ortiz-Bernad, I.; Resch, C. T.; Long, P. E.; Dayvault, R.; Karp, K.; Marutzky, S.; Metzler, D. R.; Peacock, A.; White, D. C.; Lowe, M.; Lovley, D. R. Stimulating the in situ activity of Geobacter species to remove uranium from the groundwater of a uranium-contaminated aquifer. Appl. Environ. Microb 2003, 69, 5884−5891. (5) Wu, W.-M.; Carley, J.; Fienen, M.; Mehlhorn, T.; Lowe, K.; Nyman, J.; Luo, J.; Gentile, M. E.; Rajan, R.; Wagner, D.; Hickey, R. F.; Gu, B.; Watson, D.; Cirpka, O. A.; Kitanidis , P. K.; Jardine, P. M.; Criddle, C. S. Pilot-scale in situ bioremedation of uranium in a highly contaminated aquifer. 1. conditioning of a treatment zone. Environ. Sci. Technol. 2006, 40, 3978−3985. (6) Wu, W.-M.; Carley, J.; Gentry, T.; Ginder-Vogel, M. A.; Fienen, M.; Mehlhorn, T.; Yan, H.; Carroll, S.; Nyman, J.; Luo, J.; Gentile, M. E.; Fields, M. W.; Hickey, R. F.; Gu, B.; Watson, D.; Cirpka, O. A.; Zhou, J.; Fendorf, S.; Kitanidis, P. K.; Jardine, P. M.; Criddle, C. S. Pilot-scale in situ bioremediation of uranium in a highly contaminated aquifer. 2. U(VI) reduction and geochemical control of U(VI) bioavailability. Environ. Sci. Technol. 2006, 40, 3986−3995. (7) Wersin, P.; Hochella, M. F. Jr.; Persson, P.; Redden, G.; Leckie, J. O.; Harris, D. W. Interaction between aqueous uranium(VI) and sulfide minerals: Spectroscopic evidence for sorption and reduction. Geochim. Cosmochim. Acta 1994, 58, 2829−2843. (8) Hua, B.; Deng, B. Reductive immobilization of uranium(VI) by amorphous iron sulfide. Environ. Sci. Technol. 2008, 42, 8703−8708. (9) Moyes, L. N.; Parkman, R. H.; Charnock, J. M.; Vaughan, D. J.; Livens, F. R.; Hughes, C. R.; Braithwaite, A. Uranium uptake from aqueous solution by interaction with goethite, lepidocrocite, muscovite, and mackinawite: An X-ray absorption spectroscopy study. Environ. Sci. Technol. 2000, 34, 1062−1068. (10) Rickard, D. Kinetics of FeS precipitation. Part I. Competing reaction mechanisms. Geochim. Cosmochim. Acta 1995, 59, 4367− 4379.
ASSOCIATED CONTENT
S Supporting Information *
X-ray absorption spectroscopy, X-ray diffraction, and oxidation results. This material is available free of charge via the Internet at http://pubs.acs.org.
■
■
Article
AUTHOR INFORMATION
Corresponding Author
*Phone: +82-42-868-3315; fax: +82-42-861-9720; e-mail:
[email protected]. Present Address
Korea Institute of Geoscience and Mineral Resources, 124 Gwahang-no, Yuseong-gu, Daejeon 305-350, Korea. 3375
dx.doi.org/10.1021/es203786p | Environ. Sci. Technol. 2012, 46, 3369−3376
Environmental Science & Technology
Article
EXAFS and electrophoretic mobility measurements. Geochim. Cosmochim. Acta 2000, 64, 2737−2749. (31) Nico, P. S.; Stewart, B. D.; Fendorf, S. Incorporation of oxidized uranium into Fe (hydr)oxides during Fe(II) catalyzed remineralization. Environ. Sci. Technol. 2009, 43, 7391−7396. (32) Bernier-Latmani, R.; Veeramani, H.; Vecchia, E. D.; Junier, P.; Lezama-Pacheco, J. S.; Suvorova, E. I.; Sharp, J. O.; Wigginton, N. S.; Bargar, J. R. Non-uraninite products of microbial U(VI) reduction. Environ. Sci. Technol. 2010, 44, 9456−9462. (33) Fletcher, K. E.; Boyanov, M. I.; Thomas, S. H.; Wu, Q.; Kemner, K. M.; Loffler, F. E. U(VI) reduction to mononuclear U(IV) by Desulfitobacterium species. Environ. Sci. Technol. 2010, 44, 4705− 4709. (34) Sivawamy, V.; Boyanov, M. I.; Peyton, B. M.; Viamajala, S.; Gerlach, R.; Apel, W. A.; Sani, R. K.; Dohnalkova, A.; Kemner, K. M.; Borch, T. Multiple mechanisms of uranium immobilization by Celluomonas sp. strain ES6. Biotechnol. Bioeng. 2011, 108, 264−276. (35) Scott, T. B.; Tort, O. R.; Allen, G. C. Aqueous uptake of uranium onto pyrite surfaces: Reactivity of fresh versus weathered material. Geochim. Cosmochim. Acta 2007, 71, 5044−5053. (36) Bi, Y.; Hyun, S. P., Kukkadapu, R.; Hayes, K. F. Oxidative dissolution of UO2 in a simulated groundwater containing synthetic nanocrystalline mackinawite. In review. (37) Qafoku, N. P.; Kukkadapu, R. K.; McKinley, J. P.; Arey, B. W.; Kelly, S. D.; Wang, C; Resch, C. T.; Long, P. E. Uranium in framboidal pyrite from a naturally bioreduced alluvial sediment. Environ. Sci. Technol. 2009, 43, 8528−8534. (38) Wolthers, M.; Van der Gaast, S. J.; Richard, D. The structure of disordered mackinawite. Am. Mineral. 2003, 88, 2007−2015. (39) Berner, R. A. A new geochemical classification of sedimentary environments. J. Sed. Petrol. 1981, 51, 359−365. (40) Benning, L. G.; Wilkin, R. T.; Barnes, H. L. Reduction pathways in the Fe-S system below 100 degrees C. Chem. Geol. 2000, 167, 25− 51.
(11) Jeong, H. Y.; Lee, J. H.; Hayes, K. F. Characterization of synthetic nanocrystalline mackinawite: Crystal structure, particle size, and specific surface areas. Geochim. Cosmochim. Acta 2008, 72, 493− 505. (12) Butler, E. C.; Hayes, K. F. Effects of solution composition and pH on the reductive dechlorination of hexachloroethane by iron sulfide. Environ. Sci. Technol. 1998, 32, 1276−1284. (13) Carpenter, J. R.; Hyun, S. P.; Hayes, K. F. Synchrotron X-ray characterization of mackinawite and uraninite relevant to bio-remediation of groundwater contaminated with uranium. Abstract B51C-0374. AGU 2010 Fall Meeting; American Geophysical Union; San Francisco, CA; December 2010. (14) Hyun, S. P.; Hayes, K. F. Feasibility of using in situ FeS precipitation for TCE degradation. J. Environ. Eng. (Am. Soc. Civ. Eng.) 2009, 135, 1009−1014. (15) Kohler, M.; Curtis, G. P.; Meece, D. E.; Davis, J. A. Methods for estimating adsorbed uranium(VI) and distribution coefficients of contaminated sediments. Environ. Sci. Technol. 2004, 38, 240−247. (16) Guillaumont, R.; Fanghanel, T.; Furger, J.; Grenthe, I.; Neck, V.; Palmer, D. A.; Rand, M. H. Update on the Chemical Thermodynamics of Uranium, Neptunium, Plutonium, Americium, and Technetium; Elsevier: Amsterdam, 2003. (17) Ankudinov, A.; Ravel, B.; Rehr, J. J. FEFF8; The FEFF Project; Department of Physics, University of Washington, 2002. (18) Newville, M. The IFEFFIT Tutorial; Consortium for Advanced Radiation Sources; University of Chicago: Chicago, 2001. (19) George, G. N.; Pickering, I. J. EXAFSPAK: A Suite of Computer Programs for Analysis of X-ray Absorption Spectra; Stanford Synchrotron Radiation Laboratory: Menlo Park, 2000. (20) Wolthers, M.; Charlet, L.; van Der Linde, P. R.; Rickard, D.; van Der Weijden, C. H. Surface chemistry of disordered mackinawite (FeS). Geochim. Cosmochim. Acta 2005, 69, 3469−3481. (21) Buck, E. C.; Brown, N. R.; Dietz, N. L. Contaminant uranium phases and leaching at the Fernald site in Ohio. Environ. Sci. Technol. 1996, 30, 81−88. (22) Mason, C. F. V.; Turney, W. R. J. R.; Thomson, B. M.; Lu, N.; Longmire, P. A.; Chisholm-Brause, C. J. Carbonate leaching of uranium from contaminated soils. Environ. Sci. Technol. 1997, 31, 2707−2711. (23) Zhou, P.; Gu, B. Extraction of oxidized and reduced forms of uranium from contaminated soils: Effects of carbonate concentration and pH. Environ. Sci. Technol. 2005, 39, 4435−4440. (24) O’Loughlin, E. J.; Kelly, S. D.; Cook, R. E.; Csencsits, R.; Kemner, K. M. Reduction of uranium(VI) by mixed iron(II)/iron(III) hydroxide (green rust): Formation of UO2 nanoparticles. Environ. Sci. Technol. 2003, 37, 721−727. (25) Wyckoff, R.W. G. Crystal Structures; John Wiley & Sons: New York, 1978. (26) Singer, D. M.; Farges, F.; Brown, G. E. Jr. Biogenic nanoparticulate UO2: Synthesis, characterization, and factors affecting surface reactivity. Geochim. Cosmochim. Acta 2009, 73, 3593−3611. (27) Schofield, E. J.; Veeramani, H.; Sharp, J. O.; Suvorova, E.; Bernier-Latmani, R.; Mehta, A.; Stahlman, J.; Webb, S. M.; Clark, D. L.; Conradson, S. D.; Ilton, E. S.; Bargar, J. R. Structure of biogenic uraninite produced by Shewanella oneidensis strain MR-1. Environ. Sci. Technol. 2008, 42, 7898−7904. (28) Sharp, J. O.; Schofield, E. J.; Veeramani, H.; Suvorova, E. I.; Kennedy, D. W.; Marshall, M. J.; Mehta, A.; Bargar, J. R.; BernierLatmani, R. Structural similarities between biogenic uraninites produced by phylogenetically and metabolically diverse bacteria. Environ. Sci. Technol. 2009, 43, 8295−8301. (29) Veeramani, H.; Alessi, D. S.; Suvorova, E. I.; Lezama-Pacheco, J. S.; Stubbs, J. E.; Sharp, J. O.; Dippon, U.; Kappler, A.; Bargar, J. R.; Bernier-Latmani, R. Products of abiotic U(VI) reduction by biogenic magnetite and vivianite. Geochim. Cosmochim. Acta 2011, 75, 2512− 2528. (30) Bargar, J. R.; Reitmeyer, R.; Lenhart, J. J.; Davis, J. A. Characterization of U(VI)-carbonato ternary complexes on hematite: 3376
dx.doi.org/10.1021/es203786p | Environ. Sci. Technol. 2012, 46, 3369−3376
