Identification of Green Rust in Groundwater - Environmental Science

Apr 23, 2009 - Green rust is a family of compounds with Fe(II),Fe(III)-hydroxide layers, ..... M. L.; Oton , J. M. Mössbauer spectroscopy study of co...
0 downloads 0 Views 2MB Size
Environ. Sci. Technol. 2009, 43, 3436–3441

Identification of Green Rust in Groundwater B . C . C H R I S T I A N S E N , * ,† T. BALIC-ZUNIC,‡ K. DIDERIKSEN,† AND S.L.S. STIPP† Nano-Science Center, Department of Chemistry, University of Copenhagen, Universitetsparken 5, DK-2100 Copenhagen Ø, Denmark, and Department of Geography and Geology, University of Copenhagen, Øster Voldgade 10, DK-1350 Copenhagen K, Denmark

Received April 22, 2008. Revised manuscript received October 21, 2008. Accepted October 29, 2008.

Green rust, a family of Fe(II),Fe(III) layered double hydroxides, is believed to be present in environments close to the Fe(II)/ Fe(III) transition zone. Attempts to identify members of this family in nature have proven difficult because the material is oxidized after only a few minutes exposure to air. In this paper, we present a sampling method for capturing green rust so it is not oxidized. We then we used the method to identify the compound in a groundwater sample taken below the water table from fractures in granite. X-ray diffraction patterns were weak, but clearly identical to those of synthetic GRCO3, the green rust family member where carbonate and water occupy the interlayer between the iron-hydroxide layers. The method was then tested on samples taken from an artesian well and a deep underground experimental station, both within the Fe(II)/ Fe(III) redox zone. In both cases, GRCO3 could be identified. Currently, transport models for predicting the behavior of contaminants in groundwater do not include parameters for green rust. This work demonstrates they should.

Introduction Green rust is a family of compounds with Fe(II),Fe(III)hydroxide layers, where the interlayers are filled with structured water, anions, such as Cl-, CO32-, SO42- (1, 2), and sometimes cations, such as Na+ (3). Green rust (GR) can be identified unambiguously with XRD. Mo¨ssbauer spectroscopy can characterize the Fe bonding environment and Fe(II)/ Fe(III) ratio, but these are not unique for green rust. Common soil minerals such as chlorite and several Fe-bearing clays generate Mo¨ssbauer parameters at temperatures around 15 °C (4, 5) that are very similar with green rust (6, 7). Thus definitive identification cannot be made by Mo¨ssbauer spectroscopy alone. The thermodynamic and redox properties, as they have been defined up to now, e.g., 8-10, suggest that green rust should be a common mineral in soils and sediments, but we found only a few isolated reports where it has been documented. After careful sample handling, Koch and Mørup (11) identified GR in a Danish water treatment sludge that had reacted with decomposing organic matter. Trolard and colleagues (12-14) collected large blocks from hydromorphic * Corresponding author e-mail: [email protected]. † Department of Chemistry, University of Copenhagen. ‡ Department of Geography and Geology, University of Copenhagen. 3436

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 10, 2009

soils in France, stored them in original groundwater, and analyzed samples with Mo¨ssbauer spectroscopy and interpreted the results as representing green rust. They reported that their XRD data showed many clay minerals but they did not present the patterns of GR. Recently, Bearcock et al. (15) used XRD to identify a GR compound with a c-axis period of 23.697 Å in coal mine drainage sediment. Green rust is known as an intermediate product when steel corrodes in reinforced concrete (16, 17), in seawater (18-20), and in water supply pipes (21). One could also assume it to be the active agent in metallic iron reactive barriers for groundwater remediation (22, 23). The solubility constants for synthetic material suggest that many groundwaters are at or over saturation with GR, but there is not enough evidence to conclude that it is common and pervasive in all natural systems where groundwater Fe(II) is high, pH is slightly alkaline, and oxidation potential is moderately reducing. It is almost certain, however, that green rust goes unnoticed in nature because of its very high reactivity. It transforms to Fe(III) phases within minutes of contact with air before it can be observed, sampled, or analyzed. Knowing the extent of green rust occurrence in soil, sediment, and groundwater is important for deciding if parameters describing its behavior should be included in contaminant transport and risk assessment modeling. Synthetic and natural material observed to date has very small particle size and therefore high surface area. This, together with high reactivity both for adsorbates (24) and as a reducing agent (e.g., reduction of Cr(VI) (25), NO2- (8) and chlorinated solvents (26)), suggests that if GR is common, it must be an important factor in the fate of toxic components in the environment. The small size suggests that it might serve as a transport vector for adsorbed contaminants such as heavy metals, organic compounds, and radioactive elements. The purpose of this paper is (i) to report on the discovery of green rust particles in groundwater from the Fe(II)/Fe(III) redox transition zone and (ii) to describe a simple and reliable method for finding and identifying this elusive mineral.

Materials and Methods Samples. Samples were collected from three locations where groundwater is known to be at or near the Fe(II)/Fe(III) transition boundary. The first sample was taken from a site in the tunnel in the SKB (Svensk Ka¨rnbra¨nslehantering AB) ¨ spo¨, Sweden, near Oskarshamn. Hard Rock Laboratory at A We chose this site because we could be sure to collect water that was uncontaminated with oxygen. Bore holes, drilled into the Precambrian granite walls, were capped with a valve to prevent entry of air. The bore hole we chose (SA2273A) was located about 320 m below ground surface. The water was allowed to run for a several minutes to ensure it was fresh from the granite fractures and not at equilibrium with the steel pipe. The second sample was taken from an artesian well (GEUS well no. 246.761) located near the town of Rønne on Bornholm, Denmark. The well was established soon after 2000. Its purpose is to flush the Cenomanian chalk and Lower Cretaceous/Upper Jurassic sand aquifer to remove dissolved nickel and iron derived from natural oxidation of pyrite (containing trace Ni) that have resulted from lowering of the water table during the 1990s. When the pumping rate decreased at the end of the decade, the water table rose again, releasing Ni and Fe. Discharge rate is a couple of liters per second, ensuring negligible change in redox state as water leaves the well. The sample was taken directly from the outlet pipe with a pipet. 10.1021/es8011047 CCC: $40.75

 2009 American Chemical Society

Published on Web 04/23/2009

TABLE 1. Ground Water Composition Where Samples Were Taken for XRD; Saturation Indices (SI; Logarithmic Scale) Were Determined Using PHREEQC (34) and the MINTEQ Database (35)a site

PCO2

pH

temp (°C)

Fe2+ (mg/L)

Fe(tot) (mg/L)

HCO3- (mg/L)

Cl-(mg/L)

SO42-(mg/L)

Bornholmb A¨spo¨c A¨spo¨d MICROBEe

10-1.5 na 10-2.7 10-2.1 Ca2+ (mg/L) 69 na 670 2540

6.2 (6.4) na 7.7 (7.7) 6.35 (7.4) Na+(mg/L) 16 na 1500 2520

8.2 na 15 17 SI halite -4.0 na -4.0 -3.6

na 0.65 0.76 0.105 SI gypsum -0.6 na -0.6 -0.4

6.2 0.76 0.79 0.109

116 na 158 30

32 na 3630 8040

120 na 270 516

Bornholmb A¨spo¨c A¨spo¨d MICROBEe

a For the Bornholm water, PCO2 was calculated from reported content of aggressive CO2; pH was derived from PCO2, aqueous concentrations and pH measured in samples (in brackets). For the MICROBE water, calculations are also shown for complete equilibration with atmospheric CO2 na ) not analyzed. b RL 246.761, range from 2002 to 2005 (29). c SA2273A, this study. d From ref 30. e KJ0050F01 from 2003 (27).

¨ spo¨ The third sample came from a deeper location in the A underground. At 450 m below the surface, the MICROBE laboratory (27) occupies an experimental niche in the Hard Rock Laboratory. At this site, a set of flow-through experiments has been installed into one of the bore holes (KJ0050F01) by Karsten Pedersen and his group (from Go¨teborg, Sweden). The water circulates from the borehole and back into the fractures again (flow-rate 0.4 L/min), through a series of sealed columns designed for studying radionuclide behavior in environments containing bacteria that thrive at anoxic conditions and at groundwater temperatures (27). Great pains had been taken to ensure that oxygen could not enter the closed circuits, so this was another good site for sampling. ¨ spo¨ hard rock For the first site in the tunnel of the A laboratory (tunnel site SA2273A), we collected a water sample and determined Fe2+ using the ferrozine method (28) buffered to pH 5.6 by Na-acetate and total Fe using atomic absorption spectrometry (AAS). For all samples, we had access to full groundwater composition from monitoring reports (27, 29, 30). These are summarized in Table 1. Two chronic problems with identifying green rust in nature, is finding enough material and preventing oxidation during sample collection and analysis. Stabilizing methods, such as mixing with glycerol (2), may cause interlayer expansion, modifying XRD patterns, and hindering identification. Green rust particles are submicrometer scale, their density in groundwater is low, and time required for filtration and transfer for analysis increases oxidation risk. We chose to experiment with a capture and stabilization method that was inspired by iron reactivity in nature using a substrate that is omnipresent in natural systems and that is quick and easy to use during sampling and analysis; attachment on a silica-bearing substrate. Silica has long been known to inhibit the transformation of Fe(III)-oxides (31) and silicates such as clay, quartz, and feldspar, are present in nearly all geological settings. Studies by Hansen and colleagues show that green rust has a high affinity for dissolved silica (32) and very recently, Si was shown to stabilize the structure of GRSO4 (33). Muscovite, a phyllosilicate mineral, and silica glass can be obtained with surfaces that are flat on the nanometer scale, as microscope slides and coverslips. They serve as excellent analogues for natural mineral surfaces while providing a substrate suitable for use in atomic force microscopy (AFM) and XRD. Within 30 s of taking each water sample, a droplet that would cover about 1 cm2 was deposited on a substrate (glass or mica) and covered with a platelet of the same material. The sample was left for a minute so particles in the solution could adhere; then the liquid was sucked away with a tissue. We tested several settling times (1/2 to 5 min), but we detected

no difference in the XRD patterns for solids present or the amount of the adsorbed material. So for practical reasons, we chose 1 minute. In well under 5 minutes from the time the water was taken from the flowing stream, the finished sample had been stored under N2. The samples were transported to the laboratory in nitrogen and stored in an anaerobic chamber. This glovebox has 96% N2 and 4% H2; it is equipped with palladium catalyst that adsorbs and removes O2 by reduction to water. We imaged the samples with AFM in the glovebox. Reproducibility tests made with synthetic GR proved that if it is dry, GR fixed on Si-bearing substrates is stable in air for more than 24 h. Thus, the powder XRD data were collected in air. Some XRD peaks representing salts were observed. Groundwater of the composition listed in Table 1 would evaporate to produce these simple salts. This was also demonstrated using a control solution prepared to mimic concentrations of Na+, Ca2+, Cl-, and SO42- in the groundwater of the MICROBE site using 0.03 M CaCl2 · 2H2O and 0.05 M Na2SO4. Substrates were prepared using these solutions to establish the relative intensities from salts that precipitated on the substrates with preferred orientation. Analytical Techniques. Powder X-ray diffraction (XRD) is a common tool for identifying very fine-grained soil minerals. We used a Philips PW3710 Bragg-Brentano diffractometer, with a Cu tube, (characteristic wavelength, λ ) 1.5418 Å), secondary graphite monochromator, and a variable divergence slit. The generator was set at 40 kV and the current, at 40 mA. Samples were analyzed in air for periods ranging from 2 to 9 h. Data were treated with the EVA software (DIFFRACplus evaluation package, 2007, Bruker AXS, Karlsruhe, Germany) and peaks were compared with our own synthetic standards and the powder diffraction files published by the International Centre for Diffraction Data (ICDD, 2007). The influence of fluorescence radiation was avoided with a secondary graphite monochromator. Atomic force microscopy (AFM) uses a sharp tip to sense the forces at the surface of a sample at local scale and produces images of morphology and structure with atomic resolution (39). We used a Digital Instruments Nanoscope Multimode IIIa equipped with a piezoelectric scanner with 12 µm maximum range in the x,y plane and sharpened Si3N4 tips with a spring constant of about 0.12 N/m. Images presented here were taken in deflection and height contact modes in the atmosphere of the glovebox. Relative humidity was about 35%.

Results and Discussion Groundwater Composition. The groundwater concentrations varied, but all contained the ions necessary to form green rust (Table 1). The high Cl- (3770-8040 ppm) from the VOL. 43, NO. 10, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3437

FIGURE 1. X-ray diffraction patterns from the natural samples and synthetic GRCO3. The peaks representing reflections from basal plane with d-spacings ∼7.6 and ∼3.8 Å are characteristic. H marks halite and GR is green rust reflections. The broad low peak centered at about 3.5 Å is from the glass substrate. MICROBE site is typical for deep saline waters originating from Litorina Sea (7500-4000 BP) (40). The structure and composition for the GR compounds expected in groundwater are not known precisely. The Fe(II)/Fe(III) ratio for both GRCl and GRCO3 can vary significantly (36) and published thermodynamic solubility constants for pure synthetic solids (37, 38) with a specific Fe(II)/Fe(III) ratio cannot be readily used to estimate saturation states. The Bornholm water is, however, very rich in iron. Although the Fe(II)/Fe(III) ratio could not be determined, these waters had a distinct smell of sulfur at the time of sampling. Water with Eh defined by H2S would be reducing and analyses showed appreciable ¨ spö water is dissolved Fe(II). The Fe oxidation state of the A well-established (e.g., ref 27). Table 1 gives the Fe concentrations determined in this study as well as reported major ion concentrations at the same outlet. For the MICROBE setup, pH and Fe concentrations are markedly lower, whereas Cl- concentration is higher. Finally, green rust formation may be induced by pH increase resulting when CO2 is released from solution. XRD and AFM Studies. Applying X-ray diffraction and atomic force microscopy methods, we confirmed GR presence in all three groundwaters. By analyzing material adsorbed on a flat substrate for 8-10 h, we collected XRD peaks of sufficient intensity to make a confident identification (Figure 1), even though there probably were only a few discrete grains. With AFM, we imaged small, thin, hexagonal platelets, the crystal form observed for synthetic GR (Figure 2) (25, 41). The XRD data presented here were collected from samples prepared on glass substrates; we see the low, broad peak at 3.5 Å representing amorphous silica (Figure 1). Glass works better as a substrate than mica because it contributes no sharp peaks of its own. However, on the atomic scale, glass is rough, so for AFM imaging, mica is a better substrate. Because of its tabular crystal habit, GR lies flat with its basal plane on the substrate and this preferred orientation concentrates all diffracted intensity in basal reflections, making them observable even at very low concentrations and enabling the accurate determination of the layer spacing. The basal spacing is diagnostic, both for differentiating GR from chlorite and the Fe-bearing clays, and for distinguishing between the various GR types. In the green rust family, the interlayer anions can replace each other. The order of preference is CO32- > SO42- > Cl3438

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 10, 2009

FIGURE 2. AFM images of green rust: (a) deflection mode image of synthetic material on a muscovite (mica) substrate. Particles adhere on their basal plane. (b) height mode image of a muscovite substrate showing a green rust particle from the Bornholm sample. It is not as smooth and fine as synthetic particles, probably because other compounds in the water inhibit growth and it may have undergone several phases of growth and dissolution during its transport. The abundant small white particles are halite, evaporated from the water. (42) but relative solution composition also plays a role. Thus, when the dissolved concentration of SO42- is much higher than CO32-, such as for corrosion of iron in seawater (19), GRSO4 is favored. In fresh water, where HCO3- is abundant, GRCO3 is expected. The dissolved concentration of Cl- must be very high for GRCl to dominate over GRCO3. Although there is still considerable uncertainty in details of the structure, composition and crystal lattice parameters for GRCl and GRCO3 (Table 2), there is a difference in d-spacing (∼7.6 Å GRCO3; ∼7.9 Å for GRCl) which makes their distinction possible. That difference, and the general dominance of the carbonate form, allowed us to confidently identify GRCO3 as the form of green rust present at all three sites. The natural groundwater samples contain varying amounts of the ions that could precipitate as soluble salts, such as halite (NaCl) and gypsum (CaSO4 · 2H2O). Although the water volume remaining when we make the sample is very small, enough remains that salt precipitates on the substrates. We see it as small white dots in Figure 2b and as small peaks in Figure 1. Relative peak intensities do not match those expected for bulk samples because the orientation the crystals choose for precipitation on the substrate is not random. Orientation on the samples prepared from groundwater is

TABLE 2. Structural Parameters for GRCO3 and GRCl GR type

a-axis (Å)

c-axis (Å)

GRCO3

3.181 3.16 3.166

21.82 22.45 22.52 22.59 22.662 22.8 22.8 23.61 (XRD) 22.54 (SAED)

3.17 3.16

GRCl 3.19 3.198

23.8 23.85 23.85 24.2 24.3

interlayer spacing (Å)

reference

7.273 7.483 7.507 7.53 7.554 7.6 7.6 7.870 (XRD) 7.513 (SAED) 7.59-7.60

43 43 44 45 46 21 47 48

7.93 7.95 7.95 8.07 8.09

49 1 45 50 51

this work

consistent with the orientation observed on samples prepared from control salt solutions. When wet, synthetic green rust exposed to air oxidizes after about 30 min. The solution turns from teal blue to army green, to dirty yellow, and finally to yellow-orange if the end product is goethite (R-FeOOH). The actual end product formed depends on solution conditions and rate of oxidation. Magnetite and lepidocrocite are other common endproducts. Dried green rust samples are stable for about an hour, but green rust attached to glass and to mica, both silicabearing substrates, does not oxidize during the 9 or 10 h required to collect the XRD patterns. It took several days before there were visible signs of oxidation in the form of a reddish cast on the glass surface. In these oxidized samples, GR peaks had disappeared but no new ones had appeared, indicating that particle size of the transformed material was very small or that it was amorphous. The ability of GR to oxidize in air provides a further check on its identity. Peaks from clay minerals such as chlorite, smectite and kaolinite, which also appear as flat hexagonal plates in AFM images, would not decrease in intensity with

time. Figure 3 shows data from a MICROBE sample that was analyzed periodically over several days. The fresh sample has clear peaks at 7.598 and 3.794 Å spacing, diagnostic of GRCO3. After 4 days (not shown), the peaks have shrunk, but are still present after 14 days (Figure 3). Dry state oxidation of GRCO3 has been shown to keep the structure of the compound, leading to a ferric-GR (52). Calibration against the peak intensities of halite as a stable phase gives a quantitative measure of decrease of GR concentration with time. However, some precaution must be taken for the changes of the salt pattern itself. An interesting aspect of salt behavior was observed in the time-resolved analysis of the salt standards. On a freshly prepared sample, salt peaks are visible and with time they grow. Although strange, this is consistent with behavior of sparingly soluble salts on surfaces exposed to air; they recrystallize (53) in the thin layer of water that adsorbs from the humidity in air (54, 55). Natural Green Rust. Considering the dynamic behavior of these samples with vanishingly small amounts of material, one might wonder if the green rust could have formed directly on the substrate during sample preparation. The crystal visible in Figure 2b is almost a micrometer in diameter. It is not as smooth and fine as the synthetic samples (25, 41), but one expects natural samples to be rough from subsequent growth and dissolution cycles and exposure to a spectrum of growth inhibitors. Its outline is roughly hexagonal, there is no evidence of spiral growth, and its thickness varies laterally from about 4 to 7 nm, corresponding to 2-4 unit cells. In a separate study of the stickiness of green rust on a series of mineral surfaces, Hansson (56) showed that particles attach more readily to some substrates than others, but the crystal size and morphology is the same no matter what they adsorb on. Particles attached immediately to substrates from the suspension. In another set of experiments, GR nucleated and grew from solution directly on the substrate. It took more than an hour to grow micrometer size crystals, and morphology was quite different than those grown freely in solution. In our case, the size and morphology of the natural particles suggest that GR was already present in the solution when we made the samples and did not grow in the time it took the substrate to dry. Thus we can say with certainty that the colloidal particles of green rust that

FIGURE 3. XRD patterns from the MICROBE sample before and after oxidation in air, compared to a sample made from salt test solutions (bottom). The fresh sample (top) shows clear peaks for GR which strongly diminish after oxidation in air for 14 days (middle). Other peaks on the diagrams are marked G for gypsum (CaSO4 · 2H2O) and H for halite (NaCl). VOL. 43, NO. 10, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3439

were captured on the silica-bearing substrates were already present in the groundwater when it was sampled. There are several implications of this work. XRD has been shown to be an effective technique for establishing GR presence because the basal reflections are diagnostic, allowing identification from other minerals. Now there is no doubt that green rust is present in groundwaters where composition spans the Fe redox boundary. Clearly, data describing green rust formation, dissolution, transport properties, and affinity for contaminants must be included in contaminant transport and risk assessment models. If the uptake capacity for the mixed Fe(II),Fe(III) compounds is not included, the behavior of contaminants will not be correctly predicted. GR forms as free, colloidal-size particles in the groundwater. Although most GR is likely to be immobilized by adhesion to silicates in fracture walls and on sediment particle surfaces (56), this work has demonstrated that some are mobile, with the potential to serve as vectors for transport of adsorbed contaminants far away from their source. The adsorbed GR can act as a pool of reactive surfaces that can interact with or sorb contaminants. The reactivity of sorbed GR with contaminants remains to be investigated. Although the particles sorbed to glass or mica surfaces are stabilized against oxidation in air, in a dry state, reactivity with other redoxsensitive elements, in solution, may be quite another story. Thus, green rust is an important constituent in the transport mechanism of contaminants such as heavy metals, toxic organic compounds, and radioactive species.

Acknowledgments We thank Peer Jørgensen, Birgit Damgaard, and Helene Almind for technical support; the NanoGeoScience group for discussion; and especially Lone Skovbjerg for constructive comments. Karsten Pedersen, Go¨teborg University, kindly ¨ spo¨ MICROBE laboratory for sampling provided access to his A from his flow-through bioreactors and Eva-Lena Tullborg arranged access to the Hard Rock Laboratory tunnel and provided groundwater analysis data. We particularly thank Eva-Lena Tullborg, John Smellie, Ignasi Puigdomenech, and Peter Wikberg for their interest and encouragement in the project. We are grateful for the comments of the anonymous reviewers. This work was mostly funded by Svensk Ka¨rnbra¨nslehantering AB (SKB) with some support from the Danish Natural Science Research Council (FNU) and the FUNMIG integrated project, under the 6. Framework EURATOM program.

Literature Cited (1) Refait, P.; Abdelmoula, M.; Ge´nin, J.-M. R. Mechanisms of formation and structure of green rust one in aqueous corrosion of iron in the presence of chloride ions. Corros. Sci. 1998, 40, 1547–1560. (2) Hansen, H. C. B. Composition, stabilization, and light absorption of Fe(II)Fe(III) hydroxy-carbonate (green rust). Clay Miner. 1989, 24, 663–669. (3) Christiansen, B. C.; Balic-Zunic, T.; Petit, P.-O.; Frandsen, C.; Mørup, S.; Geckeis, H.; Katerinopoulou, A.; Stipp, S. L. S. Composition and structure of an iron-bearing, layered double hydroxide (LDH) -Green rust sodium sulphate. Geochim. Cosmochim. Acta [Online early access]. DOI 10.1016/j.gca. 2009.03.032. (4) Ballet, O.; Coey, J. M. D.; Burke, K. J. Magnetic properties of sheet silicates; 2:1:1 layer minerals. Phys. Chem. Miner. 1985, 12, 370–378. (5) Feder, F.; Trolard, F.; Klingelhofer, G.; Bourrie, G. In situ Mossbauer spectroscopy: Evidence for green rust (fougerite) in a gleysol and its mineralogical transformations with time and depth. Geochim. Cosmochim. Acta 2005, 69, 4463–4483. (6) Gancedo, J. R.; Martinez, M. L.; Oton, J. M. Mo¨ssbauer spectroscopy study of corrosion products of iron with am3440

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 10, 2009

(7) (8)

(9)

(10)

(11) (12)

(13)

(14)

(15)

(16) (17)

(18)

(19)

(20) (21) (22) (23)

(24)

(25)

(26) (27)

(28)

monium nitrate in aqueous solutions. J. Phys. Colloq. 1976, C6, 297–299. Cuttler, A. H.; Man, V.; Cranshaw, T. E.; Longworth, G. A. Mo¨ssbauer study of green rust precipitates: I, Preparations from sulphate solutions. Clay Miner 1990, 25, 289–301. Hansen, H. C. B.; Borggaard, O. K.; Sørensen, J. Evaluation of the free energy of formation of Fe(II)-Fe(III) hydroxide-sulphate (green rust) and its reduction of nitrite. Geochim. Cosmochim. Acta 1994, 58, 2599–2608. Ge´nin, J.-M. R.; Bourrie´, G.; Trolard, F.; Abdelmoula, M.; Jaffrezic, A.; Refait, P.; Maitre, V.; Herbillon, A. Thermodynamic equilibria in aqueous suspensions of synthetic and natural Fe(II)-Fe(III) green rusts: Occurrences of the mineral in hydromorphic soils. Environ. Sci. Technol. 1998, 32, 1058–1068. Ge´nin, J.-M. R.; Refait, P.; Bourrie´, G.; Abdelmoula, M.; Trolard, F. Structure and stability of the Fe(II)-Fe(III) green rust ”fougerite” mineral and its potential for reducing pollutants in soil solutions. Appl. Geochem. 2001, 16, 559–570. Koch, C. B.; Mørup, S. Identification of green rust in an ochre sludge. Clay Miner. 1991, 26, 577–582. Trolard, F.; Ge´nin, J.-M. R.; Abdelmoula, M.; Bourrie´, G.; Humbert, B.; Herbillon, A. Identification of a green rust mineral in a reductomorphic soil by Mo¨ssbauer and Raman spectroscopies. Geochim. Cosmochim. Acta 1997, 61, 1107–1111. Trolard, F.; Bourrie´, G. Influence of green rusts on oxidoreduction sequences in soils: L’influence des oxydes de fer de type , rouilles vertes . sur les se´quences d’oxydo-re´duction dans les sols. C. R. Acad. Sci. 1999, 329, 801–806. Trolard, F.; Bourrie´, G. Structure of fougerite and green rusts and a thermodynamic model for their stabilities J. Geochem. Explor. 2006, 88, 249–251, Extended Abstracts presented at the 7th Symp. on the Geochemistry of the Earth’s Surface (GES-7). Bearcock, J. M.; Perkins, W. T.; Dinelli, E.; Wade, S. C. Fe(II)/ Fe(III) ‘green rust’ developed within ochreous coal mine drainage sediment in South Wales, UK. Mineral. Mag. 2006, 70, 731–741. Refait, P.; Drissi, S. H.; Pytkiewicz, J.; Ge´nin, J.-M. R. The anionic species competition in iron aqueous corrosion: Role of various green rust compounds. Corros. Sci. 1997, 39, 1699–1710. Kounde, B.; Raharinaivo, A.; Olowe, A. A.; Rezel, D.; Bauer, P.; Ge´nin, J.-M. R. Mo¨ssbauer characterization of the corrosion products of steel in civil works: Suspension bridge and reinforced concrete. Hyperfine Interact. 1989, 46, 421–428. Ge´nin, J.-M. R.; Olowe, A. A.; Resiak, B.; Confente, M.; RolletBenbouzid, N.; L’Haridon, S.; Prieur, D. Products obtained by microbially-induced corrosion of steel in a marine environment: Role of green rust two. Hyperfine Interact. 1994, 93, 1807–1812. Ge´nin, J.-M. R.; Olowe, A. A.; Benbouzid-Rollet, N. D.; Prieur, D.; Confente, M.; Resiak, B. The simultaneous presence of green rust 2 and sulfate reducing bacteria in the corrosion of steel sheet piles in the harbour area. Hyperfine Interact. 1991, 69, 875–878. Rodrı´guez, J. J. S.; Gonza´lez, J. E. G. Identification and formation of green rust 2 as an atmospheric corrosion product of carbon steel in marine atmospheres. Mater. Corros. 2006, 57, 411–417. Stampfl, P. P. Ein basisches eisen-II-III-Karbonat in rost. Corros. Sci. 1969, 9, 185–187. Roh, Y.; Lee, S. Y.; Elless, M. P. Characterization of corrosion products in the permeable reactive barriers. Environ. Geol. 2000, 40, 184–194. Agrawal, A.; Ferguson, W. J.; Gardner, B. O.; Christ, J. A.; Bandstra, J. Z.; Tratnyek, P. G. Effects of carbonate species on the kinetics of dechlorination of 1,1,1-trichloroethane by zero-valent iron. Environ. Sci. Technol. 2002, 36, 4326–4333. Randall, S. R.; Sherman, D. M.; Ragnarsdottir, K. V. Sorption of As(V) on green rust (Fe4(II)Fe2(III)(OH)12SO4 · 3H2O) and lepidocrocite (γ-FeOOH): Surface complexes from EXAFS spectroscopy. Geochim. Cosmochim. Acta 2001, 65, 1015–1023. Skovbjerg, L. L.; Stipp, S. L. S.; Utsunomiya, S.; Ewing, R. C. The mechanisms of reduction of hexavalent chromium by green rust sodium sulphate: Formation of Cr-goethite. Geochim. Cosmochim. Acta 2006, 70, 3582–3592. O’Loughlin, E. J.; Burris, D. R. Reduction of halogenated ethanes by green rust. Environ. Toxicol. Chem. 2004, 23, 41–48. Pedersen, K. The Microbe Framework. Site Descriptions, Instrumentation and Characterisation, SKB International Progress Report IPR-05-05; Svensk Ka¨rnbra¨nslehantering AB: Stockholm, February 2005; p 85. Gibbs, C. R. Characterization and application of FerroZine iron reagent as a ferrous iron indicator. Anal. Chem. 1976, 48, 1197– 1201.

(29) GEUS JUPITER - Hydrogeologic database. Avaible at http:// arcims.mim.dk/website/geus/dk/DB/JUPITER/viewer.htm. (30) Laaksoharju, M.; Wold, S. The Colloid Investigations Conducted ¨ spo¨ Hard Rock Laboratory during 2000-2004, SKB at the A Technical Report TR-05-20; Svensk Ka¨rnbra¨nslehantering AB: Stockholm, December 2005; p 211. (31) Cornell, R. M.; Schwertmann, U. The Iron Oxides. Structure, Properties, Reactions, Occurences and Uses. 2nd ed.; Wiley-VCN Verlag GmbH & Co: Weinheim, 2003. (32) Hendriksen, P. T.; Hansen, H. C. B. The effect of orthosilicate on the reactivity of iron(II)iron(III)hydroxide chloride (green rust), International Workshop Biogeochemical Processes Involving Iron Minerals in Natural Waters, Monte Verita, Switzerland, 2003. (33) Kwon, S.-K.; Kimijima, K.; Kanie, K.; Suzuki, S.; Muramatsu, A.; Saito, M.; Shinoda, K.; Waseda, Y. Influence of silicate ions on the formation of goethite from green rust in aqueous solution. Corros. Sci. 2007, 49, 2946–2961. (34) Parkhurst, D. L.; Appelo, C. A. J. User’s Guide to PHREEQC (version 2)sA Computer Program for Speciation, Batch-Reaction, OneDimensional Transport, And Inverse Geochemical Calculations; U.S. Geological Survey: Denver, 1999. (35) Allison, J. D.; Brown, D. S.; Novo-Gradac, K. J. MINTEQA2/ PRODEFA2sA Geochemical Assessment Model for Environmental Systems: Version 3.0 User’s Manual., U.S. EPA.: Athens, Georgia, 1990. (36) Hansen, H. C. B. Environmental Chemistry of Iron(II)-Iron(III) LDHs (Green Rusts). In Layered Double Hydroxides: Present and Future; Rives, V.; Ed.; Nova Science Publishers Inc., New York, 2001. (37) Drissi, H.; Refait, P.; Abdelmoula, M.; Ge´nin, J.-M. R. The preparation and thermodynamic properties of Fe(II)-Fe(III) hydroxide-carbonate (Green Rust 1); Pourbaix diagram of iron in carbonate-containing aqueous media. Corros. Sci. 1995, 37, 2025–2041. (38) Refait, P.; Ge´nin, J.-M. R. The oxidation of ferrous hydroxide in chloride-containing aqueous media and Pourbaix diagrams of green rust one. Corros. Sci. 1993, 34, 797–819. (39) Eggleston, C. M. High-resolution scanning probe microscopy: tip-surface interaction, artefacts, and application in mineralogy and geochemistry. In CMS Workshop Lectures Volume 7. Scanning Probe Microscopy of Clay Minerals; Nagy, K. L., Blum, A. E. Eds.; Clay Mineral Society: Aurora, CO, 1994. (40) Laaksoharju, M.; Tullborg, E.-L.; Wikberg, P.; Wallin, B.; Smellie, J. Hydrogeochemical conditions and evolution at the Aspo HRL, Sweden. Appl. Geochem. 1999, 14, 835–859. (41) Ge´nin, J.-M. R. Fe(II-III) hydroxysalt green rusts; from corrosion to mineralogy and abiotic to biotic reactions by Mo¨ssbauer spectroscopy. Hyperfine Interact. 2004, 156/157, 471–485.

(42) Miyata, S. Anion-exchange properties of hydrotalcite-like compounds. Clays Clay Miner. 1983, 31, 305–311. (43) McGill, I. R.; McEnaney, B.; Smith, D. C. Crystal structure of green rust formed by corrosion of cast iron. Nature 1976, 259, 200–201. (44) Brindley, G. W.; Bish, D. L. Green rust: a pyroaurite type structure. Nature 1976, 263, 353. (45) Lewis, D. G. Factors influencing the stability and properties of green rusts. Adv. GeoEcol. 1997, 30, 345–372. (46) Taylor, R. M. Formation and properties of Fe(II)Fe(III) hydroxycarbonate and its possible significance in soil formation. Clay Miner. 1980, 15, 369–382. (47) Su, C.; Puls, R. W. Significance of Iron(II,III) hydroxycarbonate green rust in arsenic remediation using zerovalent iron in laboratory column tests. Environ. Sci. Technol. 2004, 38, 5224– 5231. (48) Legrand, L.; Abdelmoula, M.; Ge´hin, A.; Chausse´, A.; Ge´nin, J.-M. R. Electrochemical formation of a new Fe(II)-Fe(III) hydroxy-carbonate green rust: characterisation and morphology. Electrochim. Acta 2001, 46, 1815–1822. (49) Vins, J.; Subrt, J.; Zapletal, V.; Hanousek, F. Preparation and properties of green rust type substances. Collect. Czech. Chem. Commun. 1987, 52, 93–102. (50) Dasgupta, D. R.; Mackay, A. L. β-Ferric oxyhydroxides and green rust. J. Phys. Soc. Jpn. 1959, 14, 932–935. (51) Schwertmann, U.; Fechter, H. The formation of green rust and its transformation to lepidocrocite. Clay Minerals 1994, 29, 87– 92. (52) Refait, P.; Benali, O.; Abdelmoula, M.; Ge´nin, J.-M. R. Formation of ‘ferric green rust’ and/or ferrihydrite by fast oxidation of iron(II-III) hydroxychloride green rust. Corros. Sci. 2003, 45, 2435–2449. (53) Stipp, S. L. S.; Gutmannsbauer, W.; Lehmann, T. The dynamic nature of calcite surfaces in air. Am. Mineral. 1996, 81, 1–8. (54) Bohr, J.; Wogelius, R. M.; Morris, P.; Stipp, S. L. S. Thickness and structure and the water film deposited from vapour on calcite surfaces. Geochim. Cosmochim. Acta, in Review. (55) Stipp, S. L. S.; Konnerup-Madsen, J.; Franzreb, K.; Kulik, A.; Mathieu, H.-J. Spontaneous movement of ions through calcite at standard temperature and pressure. Nature 1998, 396, 356– 359. (56) Hansson, E. Environmental application of iron-corrosion products: -a study of iron monosulfide and green rust. Ph. D. thesis University of Copenhagen 2007.

ES8011047

VOL. 43, NO. 10, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3441