Carbon Isotope Fractionation in the Reductive Dehalogenation of

Matthew Lee , Eliza Wells , Yie Kuan Wong , Joanna Koenig , Lorenz Adrian , Hans H. Richnow ... Jordi Palau , Orfan Shouakar-Stash , and Daniel Hunkel...
0 downloads 0 Views 396KB Size
Environ. Sci. Technol. 2005, 39, 5634-5641

Carbon Isotope Fractionation in the Reductive Dehalogenation of Carbon Tetrachloride at Iron (Hydr)Oxide and Iron Sulfide Minerals ‡



LUC ZWANK, MARTIN ELSNER, ANNA AEBERHARD, AND R E N EÄ P . S C H W A R Z E N B A C H Swiss Federal Institute for Environmental Science and Technology (EAWAG) and Swiss Federal Institute of Technology Zurich (ETHZ), Ueberlandstrasse 133, CH-8600 Duebendorf, Switzerland STEFAN B. HADERLEIN* Chair of Environmental Mineralogy, Center for Applied Geosciences (ZAG), Eberhard-Karls University Tuebingen, Wilhelmstr. 56, D-72074 Tuebingen, Germany

Compound-specific isotope analysis (CSIA) is used increasingly in contaminant hydrology in the attempt to assess the nature as well as the extent of in situ transformation reactions. Potentially, variations of stable isotope ratios along a contaminant plume may be used to quantify in situ degradation. In the present study, the abiotic dehalogenation of CCl4 by Fe(II) present at the surface of different iron minerals has been characterized in terms of the reaction rates and carbon isotopic fractionation (δ13C) of carbon tetrachloride (CCl4) as well as the yields and isotopic signatures of chloroform (CHCl3), one of the main transformation products. The abiotic reductive dehalogenation of CCl4 was associated with substantial carbon isotopic enrichment effects. The observed enrichment factors, , correlated neither with the surface-normalized reaction rate constants nor with the type of products formed but fell into two distinctly different ranges for the two principal groups of minerals studied. With iron (hydr)oxide minerals (goethite, hematite, lepidocrocite, and magnetite) and with siderite, the -values for CCl4 dehalogenation were remarkably similar (-29 ( 3‰). Because this value matches well with the theoretical estimates for the cleavage of an aliphatic C-Cl bond, we suggest that dissociative electron transfer to CCl4 controls the reaction rates for this group of iron minerals. Conversely, CCl4 transformation by different preparations of the iron sulfide mackinawite was accompanied by a significantly lower carbon istotopic fractionation ( ) -15.9 ( 0.3‰), possibly due to the presence of nonfractionating rate-determining steps or a significantly different transition state structure of the reaction. Isotopically sensitive branching of the reaction pathways (i.e., the effect of different product distributions on isotope fractionation of CCl4) did not play a significant role in our systems. The extensive data set presented in this study opens new perspectives * Corresponding author phone: ++49 (0)7071 297 3148; fax: ++49 (0)7071 295 139; e-mail: [email protected]. † Present address: Stable Isotope Laboratory, Department of Geology, University of Toronto; 22 Russell Street, Toronto, Ontario M5S 3B1. ‡ Present address: s&e consult, B.P.2453 L-1024 Luxembourg. 5634

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 15, 2005

toward an improved understanding of the factors that determine reaction mechanisms and isotopic fractionation of dehalogenation reactions by Fe(II) at iron containing minerals.

Introduction Quantification of in situ degradation at contaminated sites often suffers from the inability to establish reasonable mass balances based on concentration data from a limited number of sampling wells. Recently, compound-specific isotope analysis (CSIA) has evolved into a useful tool to gain additional information on the fate of groundwater contaminants (1-9) because during irreversible chemical reactions (i.e., the cleavage or formation of chemical bonds) a shift between the isotope ratios of the contaminant and its degradation product(s) may occur known as kinetic isotope fractionation. In the reductive transformation of the priority groundwater pollutant carbon tetrachloride (CCl4), for example, an initial electron transfer has been shown to be directly concerted with the cleavage of a C-Cl bond (10). Because the C-Cl bond strength is diminished in the transition state of this reaction, the rate constant for the cleavage of 12C-Cl bonds (12k) is somewhat higher than that for 13C-Cl bonds (13k) resulting in a kinetic isotope effect, KIE () 12k/13k), larger than unity. Thus, molecules containing heavy isotopes are progressively enriched in the remaining fraction of the reactant, whereas the transformation product(s) deplete in the heavy isotopes (11). In environmental studies, the isotopic fractionation of the parent compound is commonly quantified by the Rayleigh equation that was introduced by Mariotti et al. (12)

ln

(

δ13Ct + 1000 13

δ C0 + 1000

)

)

 ln f ) (R - 1)ln f 1000

(1)

where δ 13Ct and δ 13C0 are the relative isotopic compositions of the investigated compound at time t in a closed system and at time 0, respectively. R is commonly denoted as the isotopic fractionation factor,  is referred to as the isotopic enrichment factor, and f corresponds to the remaining fraction of the substrate (i.e., f ) [CCl4]t /[CCl4]0). Because the isotopic fractionation by physical processes such as nonspecific sorption (13), evaporation (7, 14, 15), or advection and dispersion (16) is in most cases negligibly small, the determination of isotopic signatures of groundwater contaminants along a flow line may allow for quantification of in situ degradation without the need for mass balancing. This approach has already been used successfully to quantify the extent of a given degradation reaction at contaminated field sites and requires that for a given reaction the enrichment factor is known and does not vary significantly between different environmental conditions (1, 17-19). However, because certain transformation reactions can be both abiotic or microbially mediated and/or involve multiple reaction pathways depending on the environmental conditions influencing the rate limiting processes, one cannot generally assume that a single and constant isotopic enrichment factor may be used for the quantification of in situ transformation processes. Thus, it is important to address the potential variability of the isotopic enrichment observed for the degradation of a given compound under various conditions. 10.1021/es0487776 CCC: $30.25

 2005 American Chemical Society Published on Web 06/22/2005

In the present study, the abiotic dehalogenation of carbon tetrachloride has been investigated. Besides its environmental relevance, CCl4 has been chosen as the most “simple” model compound for carbon isotope studies on reductive dehalogenation of aliphatic compounds because the only carbon atom of CCl4 is directly involved in any bond cleavage. Changes in carbon isotopic signatures will therefore yield “undiluted” isotope effects because there are no carbon atoms that do not participate in the transformation reactions and would, hence, complicate a quantitative interpretation of the measured fractionation. Provided that the C-Cl bond cleavage is rate-limiting for the overall reaction, the fractionation and enrichment factors, R and , determined from eq 1 may then be directly translated into an intrinsic kinetic isotope effect KIE according to

(

R-1 ) 1 +

 1000

-1

)

12

) KIE ) 13

k k

(2)

where 12k and 13k are rate constants for the cleavage of a C-Cl and 13C-Cl bond, respectively (20). The upper limit of the expected range of such kinetic isotope effects is predetermined by the nature of the chemical bond broken in the transition state of the rate-determining step. From the C-Cl force constant and atomic masses, a maximum KIE value (“Streitwieser Limit”) of around 1.057 is estimated (21, 22). In most cases, however, bond cleavage is only partially realized in the transition state. For example, in the electrochemical reduction of CCl4 a small but significant interaction between the •CCl3 and Cl- fragments prevails even in the product state (10). To obtain a rough estimate of the expected isotope fractionation without previous knowledge of the transition state and rate-determining step(s), one may consider an “average” transition state with 50% bond cleavage, corresponding to KIE values of about 1.028-1.029 or  values around -27‰ to -28‰ for CCl4. Actual values of KIE may, nevertheless, be subject to considerable variation for the following reasons. (A) Change of the C-Cl bond strength in the transition state. For the reasons discussed above, any factors that maximize C-Cl bond changes in the transition state will lead to larger kinetic isotope effects, whereas factors that minimize bond changes will lead to smaller KIE values. (B) Importance of the ratedetermining step. R and , then, only reflect the intrinsic isotope effect, KIE, if the C-Cl bond cleavage is rate-limiting. Conversely, if a nonfractionating process such as adsorption or diffusion is the slow step, then every substrate molecule that reaches the reactive site will essentially be converted and no isotopic discrimination will then be observed in the remaining substrate. In a situation where reaction rates of both the fractionating and nonfractionating processes are of similar magnitude, apparent kinetic isotope effects (AKIE) will be obtained from eq 2 that are considerably smaller than the expected intrinsic isotope effects, KIE (22). (C) Isotopically sensitive branching. If the reaction involves different pathways, then two cases may be distinguished. Either there are two or more initial parallel reactions, which may involve different bonds of the same substrate (e.g., simultaneous cleavage of one vs two C-Cl bonds in CCl4). If these parallel reactions show different kinetic isotope effects (e.g., 1.025 for one vs 1.050 for two broken C-Cl bonds), then the observed net fractionation in the CCl4 substrate will be a weighted average and can be expected to depend strongly on the product distribution. Or there is only one initial irreversible step leading to a common intermediate (e.g., dissociative electron transfer leading to •CCl3). Consecutive parallel reactions of this intermediate may then lead to different pathways associated with different product isotope ratios. However, they will not affect isotope fractionation in the substrate because the branching occurs later on in the

12

transformation, whereas kinetic isotope effects in the substrate are affected only by the first rate-determining step. Recent work (20, 24) has shown that different factors such as pH, the presence of radical scavengers (e.g., pH buffers), and surface stabilization of reaction intermediates can control the yields of products in the reductive dechlorination of CCl4 by goethite/Fe(II) and magnetite/Fe(II). However, these factors were shown to affect primarily reactions of the intermediate •CCl3 radical, in which case they may not be expected to have a major influence on isotope fractionation in the substrate CCl4 (see above). It has also been shown that the reactivity of Fe(II) sorbed to the surfaces of a range of iron minerals (siderite, hematite, lepidocrocite, goethite, magnetite, pyrite, mackinawite, and green rust) toward the reduction of various classes of contaminants spans several orders of magnitude (25). This effect has been attributed partly to differences in the sorption densities of Fe(II) on the different minerals, and partly to fundamentally different intrinsic reactivities of the Fe(II) surface sites, depending on the type of mineral. Such parameters will concern kinetic isotope effects only insofar as they can affect massdependent molecular energies, particularly vibrational energies, which depend strongly on changes in the vibrational force constant of the C-Cl bond during reaction. Conversely, most of the overall activation energy in a reaction is attributable to changes in electronic molecular energies, which are not mass-dependent; correlations of reaction rates with isotope fractionation can, therefore, generally not be expected. Hence, if observable reactivity increases owing to changes in pH and/or solution composition, then it can have several underlying reasons that may or may not influence kinetic isotope effects. First, there can be simply a higher concentration of surface sites, in which case isotope fractionation will clearly not be affected because the change in bond strengths is still just the same at each single site. Second, the intrinsic reactivity at each site may be increased. Even in this case, isotope fractionation is still not necessarily correlated with the change in activation energy (see above), but it may depend on several factors that are indirectly correlated with surface reactivity. Different specific interactions at the mineral surface may lead to a different extent of bond changes in the transition state. Also, if higher/lower intrinsic reactivity of the mineral leads to earlier ()more substrate-like)/later ()more product-like) transition states (25), then this could correspond to a lower/higher extent of C-Cl bond cleavage in the transition state. Finally, if the elementary reaction step at the surface becomes fast, then other nonfractionating steps (e.g., transport to the surface or formation of substratesurface complexes) may become comparatively slow so that the intrinsic KIE is increasingly masked in the observable AKIE (see above). Hence, different aspects of the abiotic reductive dehalogenation by surface-bound Fe(II) may potentially influence the isotopic enrichment factor of the oxidant CCl4 and thereby limit the use of CSIA in the quantification of in situ degradation. Clearly, the determination of the variability of the enrichment factor caused by the various processes is a prerequisite for a potential application in the field. To quantify this variability for different environmentally relevant iron minerals, we studied the reaction using Fe(II) sorbed to siderite, hematite, lepidocrocite, magnetite, goethite, or mackinawite that have shown different intrinsic reactivities toward halogenated compounds. Furthermore, concentrations as well as isotopic signatures of chloroform (one of the major degradation products) were determined to obtain an estimate of the importance of the different reaction pathways on the observed isotopic signatures. VOL. 39, NO. 15, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5635

TABLE 1. Characteristics of Minerals and Suspensions Used in This Study

mineral

formula

specific surface area (m2/g)

mackinawite 1 mackinawite 2 goethite aa goethite b goethite c magnetite lepidocrocite hematite siderite

FeS FeS R-FeOOH R-FeOOH R-FeOOH Fe3O4 γ-FeOOH Fe2O3 FeCO3

13 77 17 17 17 19 18 14 38

added to suspension mineral MOPS buffer Fe(II) (g/L) (mM) (mM) 4.0 0.6 5.7 3.1 3.1 2.6 2.8 3.6 1.3

0.76 0.76 5.0 1.52 1.00 1.00 1.20

2.3 1.4 1.4 1.5 1.5 1.8 1.9 1.5 0.97

measured dissolved Fe(II) (mM)

pH

calculated Fe(II)ads Fe(II)ads (mM) (µmol/m2)

1.06 0.88 0.93 1.04 1.02 1.05 0.98 0.95 0.93

7.2 7.2 7.0 7.2 7.2 7.3 7.2 7.2 7.1

1.3 0.53 0.43 0.46 0.48 0.78 0.83 0.53 0.045

25 11 9.0 9.2 9.6 16 17 11 0.90

a Data from refs 20, 25, and 26 note that the surface concentration of minerals in suspension was ca. 100 m2/L rather than ca. 50 m2/L for all other assays.

Experimental Section Reagents and Materials. Methanol (>99.9%) used for the preparation of stock solutions was obtained from Scharlau S. A. (Barcelona, Spain). Carbon tetrachloride (CCl4, >99.5%), chloroform (CHCl3, >99.5%), tetrachloroethene (PER, >99.9%) and the antibiotics kanamycinsulfate and ampiciline were purchased from Fluka (Buchs, Switzerland). Bromoform (CHBr3, >99%) was obtained from Aldrich (Steinheim, Germany). The carbon isotopic signatures of CCl4 (-38.6 + 0.2‰), CHCl3 (-45.3 + 0.2‰), and PER (-27.3 + 0.2‰) were determined using an elemental analyzer (NC2500, Thermoquest, San Jose´, CA) coupled to an isotope ratio mass spectrometer (Isoprime, VG Instruments, Manchester, U.K.). Goethite, hematite, and lepidocrocite were purchased from Bayer (Leverkusen, Germany). The synthesis of magnetite, siderite, and mackinawite has been described in a previous study (25). The two mackinawite batches have been prepared following the same procedure and differ only by their specific surface area. The crystal structures of the minerals were confirmed using X-ray diffraction (Siemens D 5000 and XDS 2000 with a Cu KR source, scan rate: 1°/min), and the specific mineral surface areas (see Table 1) were determined by N2 adsorption (Sorptomatic 1990, Fisons Instruments). Fe(II) solutions were prepared by adding 28 g (0.5 mol) of iron powder (Merck, Darmstadt, Germany) to 1 L of 1 M deoxygenated HCl. The suspension was brought to 70 °C under gentle stirring until the release of H2 ceased (2-2.5 h). The solution was taken into an oxygen-free glovebox and filtered through a 0.2 µm PTFE filter to remove the excess iron powder. The Fe(II) concentration was determined photometrically using the phenanthroline method (27). Stock solutions of CCl4 and PER (as internal standard) were prepared in degassed deionized water or anoxic methanol. The concentrations of CCl4 and PER in the aqueous stock solution were 0.24 mM and 0.12 mM, respectively. To reduce the spike volume in the case of experiments carried out in ampules, we prepared concentrated methanolic spike solutions (3 mM of CCl4 and 1.5 mM of PER). PER was chosen as the internal standard to correct for losses during the spiking of the reaction vials as well as for diffusion into the Viton stopper. PER has been selected because its physicochemical properties are similar to those of CCl4, and the compound does not undergo significant tranformation within the time frame considered here and under the given experimental conditions (data not shown). Preparation of Mineral Suspensions. Aliquots of iron oxides as well as iron hydroxides were resuspended in 500 mL of deionized water to yield a surface concentration of 50 m2/L. The suspensions were agitated for 24 h, and, after sedimentation, the minerals were rinsed several times with 5636

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 15, 2005

deionized water. The suspensions were then degassed with argon and introduced into an oxygen-free glovebox. Suspensions of oxygen-sensitive minerals (siderite, mackinawite) were prepared in the glovebox. Aliquots of the 0.5 M FeCl2 solution were added to the suspensions, and the pH was adjusted to 7.2 (1 M NaOH, HCl). Suspensions, with the exception of goethite suspensions (where the intrinsic buffer capacity was shown to be sufficient (20)), were buffered by 1 mM 4-morpholinopropanesulfonic acid (MOPS). After equilibration (>12 h), the concentration of the dissolved Fe(II) was determined using the phenanthroline method (27). If necessary, the Fe(II) concentration was increased by successive additions of FeCl2 and pH adjustments until an aqueous Fe(II) concentration of ∼1 mM was obtained. The characteristics of the mineral suspensions used for the experiments are summarized in Table 1. Transformation Experiments. The transformation experiments of CCl4 in the presence of FeS, goethite, magnetite, and lepidocrocite were carried out in 25-mL serum vials. Replicates (12-18) were prepared for each experiment by adding 3 mL of an aqueous spike solution to 21 mL of the equilibrated suspension, yielding initial concentrations of ∼10 µM. The vials were sealed with Viton stoppers, crimped, taken out of the glovebox, and agitated on a reciprocating shaker at 23 °C until they were sacrificed for analysis. Experiments with slower reaction kinetics (t1/2 g 200 h), such as with hematite and siderite, were prepared in 20-mL glass ampules. Ampules were filled in the glovebox with 20 mL of the suspensions and 100 µL of the methanolic spike solution, yielding CCl4 concentrations of ∼10 µM. To keep the suspensions anoxic, we closed the ampules with Viton stoppers before they were taken out of the glovebox. To avoid evaporative losses, we rapidly froze the liquid in the ampules in acetone cooled with liquid nitrogen before flame sealing. During the sealing process, we avoided the bursting of the ampules by piercing the Viton stopper with a needle kept in an argon-flushed hose. Similar to the serum vials, the sealed ampules were agitated on a reciprocating shaker at 23 °C in the dark. Control samples of CCl4 in deionized water were prepared for each experiment in serum vials or glass ampules, respectively. Reaction vials were sacrificed at appropriate time intervals. In a first step, the vials were opened and samples of 2.5 mL were taken for concentration determination. These samples were extracted with 5 mL n-hexane containing 10 µM CHBr3 as the internal standard in 8-mL glass vials closed with PTFEsealed screw caps on a Vortex shaker for 3 min. The extract was filled in 0.8-mL auto-sampler vials and stored at -30 °C until analysis. For the determination of the isotopic signatures of the analytes, we centrifuged the remainder of the suspensions for 10 min at 4000 rpm (Heraeus Megafuge 1.0R, Kendro, Asheville, NC). The supernatant was transferred to 8-mL glass

vials, closed with PTFE-lined screw caps, and stored without headspace at 4 °C until analysis was performed. Analytical Procedures. The concentrations of the halogenated compounds were determined on a gas chromatograph (Carlo Erba HRGC 5160) equipped with a liquid autosampler AS200 (Thermo Finnigan, San Jose´, CA) and an electron capture detector (ECD 400 with 63Ni-source, Carlo Erba) maintained at 250 °C. Separation of the analytes was achieved on a DB-624 capillary column (30 m × 0.32 mm i.d., 1.8-µm film thickness) obtained from J&W Scientific. The hexane extracts (1.5 µL) were injected on a split/splitless injector maintained at 250°. The temperature program was as follows: 2 min at 50 °C then to 175 °C with 10 °C/min, then to 250 °C with 30 °C/min and 5 min at 250 °C. The quantification was based on an eight-point calibration (010 µM). The 13C isotopic composition of the analytes was determined on a GC-C-IRMS system (Trace GC, coupled to a GC Combustion III interface and a DeltaPLUSXL isotope ratio mass spectrometer, Thermo Electron, Bremen, Germany) coupled to a purge and trap concentrator (LSC 3100, Tekmar Dohrmann, Mason, OH) equipped with a liquid auto-sampler (AquaTek 70, Tekmar Dohrmann). To achieve baseline separation of CCl4 and CHCl3 at conditions when one of the two analytes was highly concentrated compared to the second (e.g., in the beginning or at the end of the reaction, respectively), we used a polar capillary column (Stabilwax, 60 m × 0.32 mm i.d., 1-µm film thickness, Restek). The aqueous samples were purged for 11 min with an N2 flow of 40 mL/min and trapped on a VOCARB 3000 trap (Supelco, Bellefonte, PA) at room temperature. By heating the trap to 250 °C for 1 min., we thermally desorbed and transferred the analytes to a cryofocusing unit maintained at -120 °C using liquid nitrogen. These purge and trap method parameters were investigated thoroughly in order to achieve optimal extraction efficiencies and reproducible isotopic measurements (28). The carrier gas was helium at a constant pressure of 1.9 bar. The temperature program was as follows: 2 min at 45 °C, then to 100 °C with 14 °C/min, then to 210 °C with 20 °C/min and 4 min at 210 °C. The analytes were oxidized within the combustion interface maintained at 940 °C. The catalyst was regularly reoxidized after ∼40 samples. Chloroform and carbon tetrachloride of known isotopic signatures were measured under the same conditions as the unknown samples to allow for correction of the potential isotopic fractionation due to the extraction process. Isotopic enrichment factors () were determined using the linearized Rayleigh equation (eq 1) and the software Lin2D (29) that allows one to consider the uncertainties in the x and y coordinates for the calculation of the linear regression. (20)

Results and Discussion Reaction Kinetics and Isotopic Enrichment. The reaction of CCl4 with the majority of the investigated iron minerals followed pseudo first-order kinetics (Figure 1A) and was associated with a significant stable isotopic enrichment, following Rayleigh-type behavior (Figure 1B). The pseudo first-order rate constants were determined from an exponential fit of the CCl4 concentration versus time data (Figure 1). The isotopic enrichment factors, , were determined from eq 1. All of the experiments were performed with 12-18 independent reaction vials that were prepared from the same mineral suspension and were analyzed by sacrificial sampling. Reported values do, therefore, not only reflect the uncertainty of the analytical method, but they also include variations that may occur between different reaction containers. The obtained rate constants as well as the  values are given in Table 2 and Figure 2. Similar to

earlier studies with hexachloroethane (25), the first-order rate constants of CCl4 for reaction with Fe(II) at different iron minerals spanned several orders of magnitude. With the exception of the mackinawite 2 batch, the observed surface area normalized rate constants increased from Fe(II) carbonates to Fe(II) (hydr)oxides to iron sulfides, thus following the trend observed previously for hexachloroethane (25). The observed enrichment factors show that the abiotic dehalogenation reactions of CCl4 studied were associated with substantial carbon isotope fractionation. Given that there are several factors that may potentially lead to different fractionation with different minerals (see Introduction), the  values determined for nonsulfidic iron minerals (goethite, hematite (end), lepidocrocite, siderite, and magnetite) as well as for Fe(II)porphyrin in homogeneous aqueous solution are very similar. The observed range of -29 ( 3‰ compares very well with the range of -27 to -28‰ that can be estimated from the calculated Streitwieser limit for 50% C-Cl bond cleavage (see Introduction), indicating that with these minerals generally only one C-Cl bond was broken in the initial rate-determining step of the reaction. Given the differences in mineralogy, specific surface area, or Fe(II) surface site density within the set of iron (hydr)oxide studied (see Table 1), this observed range of -29 ( 3‰ appears to be independent of absolute reaction rates and product distribution (see Figure 2 and Table 2) and it might be representative of “typical” kinetic isotope effects that can be expected in reductive dehalogenation of CCl4 by Fe(II) present at iron (hydr)oxides, at least under the geochemical conditions applied here. Conversely, isotopic enrichment factors that were determined with different preparations of the iron sulfide mackinawite were much smaller (-15.9‰). The significant difference to most nonsulfidic minerals could indicate that a nonfractionation step such as formation of substratesurface complexes or diffusion to reactive sites (e.g., into interlayers) had become rate-limiting; alternatively, the C-Cl bond strength may have changed to a much lesser extent in a transition state at the FeS surface (see Introduction). The measured fractionation was identical with both mackinawite preparations despite (i) different product distributions (see below) and (ii) different reaction rates (log k′obs -1.31 vs -2.90). The first observation indicates that the CCl4 transformations in both mackinawite preparations had a common initial rate limiting step and that the product distribution was governed by consecutive reactions of a common intermediate. Hence, isotopically sensitive branching was not significant for isotope fractionation (see Introduction). The second observation suggested that the initial reaction step and the intrinsic reactivity of the surface sites were probably equivalent for the two mineral preparations and that only the concentration of the reactive surface species was smaller in the case of mackinawite 2. (If reactive species were located in interlayers, then their true concentration may not have been reflected in the BET-surface measurement.) Therefore, the enrichment factor of about -16‰ may be characteristic of reductive dechlorination of CCl4 by mackinawite. Hematite/Fe(II). During the reduction by Fe(II) at the surface of hematite, the isotopic fractionation of CCl4 changed during the experiment (see Figure 3). While the initial reaction rate doubled from 6.2 × 10-4 to 1.14 × 10-3 h-1 after a certain “conditioning” of the reacting surfaces, the isotopic enrichment factors decreased (from  ) -49.7 ( 4.9‰ to  ) -25.8 ( 0.8‰), suggesting a change in the reaction mechanism. The observed increase in reaction rates for hematite was supposedly due to a fundamental change in the quality rather than the quantity of reactive sites, in contrast to the mackinawite batches discussed above. This might indicate VOL. 39, NO. 15, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5637

FIGURE 1. Reaction kinetics and isotopic enrichment of CCl4. (A) Time courses of carbon tetrachloride concentrations (filled circles), the pseudo first-order fit (continuous line) and chloroform concentrations (open circles). The dashed line corresponds to the average concentrations of CCl4 blanks (filled triangles) sampled during the course of the reaction and used as y(0)-values for the calculation of the reaction rate constants. Note the different time scales of the x axes. (B) Rayleigh plots corresponding to the different experiments. The isotopic signature at time 0 corresponds to the average of the determined δ13C values of the blank samples. 5638

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 15, 2005

TABLE 2. Rate Constants, Isotopic Enrichment Factors, Yields, and Isotopic Signature of CHCl3 for the Reduction of CCl4 by Fe(II) at Different Iron Minerals as well as Iron(II) Porphyrin and Polysulfide kobsa (h-1)

log k′obs b (CCl4)

ERayleigh a,c (‰)

AKIEd

E∆δ13C e (‰)

CHCl3 yieldf (%)

δ13C CHCl3end (‰)g

mackinawite 1 mackinawite 2 goethite ag goethite b goethite c magnetite lepidocrocite hematite siderite

2.434 ( 0.13 0.0633 ( 4.6 × 10-3 0.019 ( 0.002 0.023 ( 0.005 0.028 ( 0.006 0.245 ( 0.013 6.9 × 10-3 ( 9 × 10-4 1 × 10-3 ( 5 × 10-5 4 × 10-4 ( 3 × 10-5

-1.31 -2.90 -3.72 -3.34 -3.26 -2.31 -3.86 -4.69 -5.09

-15.9 ( 0.3 -15.9 ( 0.2 -26.5 ( 2.8 -25.6 ( 1.3 -26.3 ( 0.9 -25.8 ( 1.0 -32.0 ( 0.5 -29.5 ( 1.0 -26.7 ( 2.9

1.016 1.016 1.027 1.026 1.027 1.026 1.033 1.030 1.027

-17.2 ( 1.0 -22.1 ( 1.1 n.d.j n.d.j n.d.j -25.8 ( 1.0 -31.9 ( 1.0 -29.4 ( 1.0 -22.0 ( 3.2

55.3 ( 1.3 29.1 ( 1.0 33h 26 31 80.7 ( 4.3 14.2 ( 0.4 17.5 ( 7.3k 33.5 ( 4.6k

-40.7 ( 0.5 -49.6 ( 0.5 n.d.j n.d.j n.d.j -42.2 ( 0.5 -62.7 ( 0.8 n.d.j n.d.j

iron(II) porphyrinh polysulfide

0.042 ( 0.002 0.145 ( 0.003

i i

-26.2 ( 0.3 -21.6 ( 2.9

1.027 1.022

n.d.j n.d.j

n.d.j n.d.j

n.d.j n.d.j

mineral or reductant

a The stated uncertainties correspond to 95% confidence intervals calculated from the linear regression of the data. b Surface area normalized pseudo first-order rate constants (h-1m-2L). k′obs ) kobs/S with S corresponding to the surface of the mineral in the suspension (e.g., (50 m2). c Determined based on eq 2. d Apparent kinetic isotope effect calculated using eq 2 (note that AKIE equals KIE if the reaction step in which fractionation takes place is rate-determining). e Determined based on the differences in δ13C values of CCl4 and CHCl3 in the beginning of the reaction, that is, ε ) δ13C(product) - δ13C(substrate). f Corresponds to the CHCl3 yield at 99% degradation of CCl4, if not stated otherwise. The uncertainties were calculated with error propagation from the standard deviation of replicate CHCl3- and CCl4- concentration measurements. g The uncertainties correspond to the absolute precision of the measurement or, where greater than (0.5‰, to the standard deviation from replicate δ13C analyses. h Data from refs 20 and 25. i Reaction in homogeneous solution. j Not determined. k Corresponds to the average CHCl3 yield for all data points. The determination of the CHCl3 yield at 99% degradation was not possible because the reactions have not been followed to completion.

FIGURE 2. Isotopic enrichment factors of CCl4 versus log kobs. Pseudo first-order rate constants) vs isotopic enrichment observed for iron sulfides and polysulfide (open circles) and for iron (hydr)oxides (filled circles); *, data from ref 20. The right-hand section shows isotopic enrichment values for reactions in homogeneous solution. that more “magnetite-like” surface species were formed as the reaction proceeded. A formation of new reactive sites of uniform reactivity at mineral surfaces from the oxidation of sorbed Fe(II) has been invoked previously (30, 31) in batch and column experiments dealing with the reduction of nitroaromatic compounds by Fe(II) present at minerals. Pecher et al. (26) also noted a pronounced difference in hematite reactivity after conditioning of the surfaces with Fe(II) and attributed this effect to a “remodeling” of reactive surface sites at the mineral. Note that, from an environmental point of view, the longterm rather than the initial conditions will determine the reactivity of hematite surfaces. Therefore, the hematite data plotted in Figure 2 represent the conditions at prolonged reaction times. Product Yields and Isotopic Enrichments. Among the different potential transformation products of CCl4 that may be formed under the experimental conditions (see Figure 4), only chloroform (CHCl3) could be quantified and isotopically characterized in our study. The average carbon isotope composition of the sum of all of the other products of CCl4 was calculated as a bulk parameter (δ13Cother) based on an isotopic mass balance. Note that, however, according to

Figure 4 this sum may include very different chemical compounds such as carbon monoxide and formate and that under the given conditions a parallel dehalogenation reaction of CHCl3 can be neglected. Although the fraction of chloroform produced varied significantly between the different minerals (see Table 2), there was no discernible correlation with observable isotope fractionation in the substrate CCl4. Consistent with the picture of only one broken C-Cl bond in the first rate-limiting step of the reaction, a significant effect of competing initial reaction steps that would be associated with different isotopic fractionation can therefore be ruled out. However, as can be seen from Tables 2 and 3, the isotopic signatures of the product CHCl3 (δ13CCHCl3) at the end of the reaction were slightly but significantly depleted compared to the isotopic composition of the initially spiked CCl4. This indicates that the isotope fractionation in CHCl3 was influenced by isotope effects in additional, consecutive reaction steps that occurred after the rate-limiting first step to the trichloromethyl radical. As a rule, if parallel reaction pathways start off from a common intermediate such as •CCl3, then heavy isotopes will tend to accumulate at the end of the pathway with the least fractionation (“isotopically sensitive branching”). Hence, one VOL. 39, NO. 15, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5639

FIGURE 3. Changes in reaction rates and isotopic fractionations during the reaction of CCl4 with Fe(II) at hematite. (A) Increase of reaction rates during the course of the reaction. (B) The change in reaction rates correlated with a significant change in the isotopic enrichment of CCl4.

FIGURE 4. Simplified degradation pathway for the reductive dehalogenation of CCl4 by Fe(II) on goethite as suggested by Elsner et al. (20). Products in rounded boxes represent measured, those in dashed boxes represent proposed intermediates.

indicator in estimating how much percent of carbon tetrachloride has been transformed to this problematic product. Environmental Significance. At first sight, the differences in isotopic enrichment of CCl4 in the reaction with Fe(II) at iron oxides and mackinawites found in this study seems confounding for a quantification of in-situ degradation of CCl4 at contaminated sites based on the measured isotope ratios in the field samples. This is because on top of potential isotope fractionation due to biodegradation, abiotic dehalogenation reactions at iron minerals will add to the overall fractionation factor measured in the field. If mass balances cannot be established and no unique and robust enrichment factor, , is available, then it will be difficult to assess whether isotopic enrichment measured in a sample is caused by a strongly fractionating reaction (e.g., with Fe(II) on oxides) that occurred to a small degree or by a weakly fractionating reaction (e.g., with mackinawite) that proceeded to a large extent. This ambiguity, however, may be resolved by conducting microcosm experiments with field sediment of the respective contaminated site with and without inhibition of microbial activity (33). In the presence of a complete mass balance, it is then straightforward to determine the appropriate enrichment factor for quantification of the corresponding field samples. Based on the insight from our study, it may even be possible to identify from this  value the nature of the most relevant iron(II) minerals ((hydr)oxides vs mackinawite) under the prevailing conditions. If the extent of isotope fractionation obtained with mackinawite in this study can also be observed with other iron sulfide minerals, then the carbon isotopic fractionation of CCl4 may be developed into a diagnostic tool to assess the reactivity of iron minerals under iron-reducing and sulfidogenic conditions, respectively. To this end, follow up studies are needed addressing also the potential effects of different geochemical conditions (pH, surface density of Fe(II, etc.) and will be carried out in our laboratories. The key to a mechanistic understanding of isotope fractionation in such surface catalyzed reactions, finally, will be future investigations that address intrinsic surface reactivity and the influence of possible slow preceding steps (diffusion/adsorption to reactive surface sites).

Acknowledgments TABLE 3. δ13C Isotopic Signatures of Chloroform (CHCl3) for the Reductive Dehalogenation of CCl4 by Fe(II) at Various Iron Minerals mineral

δ13C CCl4,start (‰)

δ13C CHCl3,end (‰)

∆δ13CCHCl3 a (‰)

δ13Cother b (‰)

mackinawite 1 mackinawite 2 magnetite lepidocrocite

-38.3 ( 0.5 -38.3 ( 0.5 -39.1 ( 0.5 -38.6 ( 0.5

-40.7 ( 0.5 -49.6 ( 0.5 -42.2 ( 0.5 -62.7 ( 0.5

-2.4 ( 1.0 -11.3 ( 1.0 -3.1 ( 1.0 -24.1 ( 1.0

-35.3 ( 1.0 -20.7 ( 4.2 -34.5 ( 1.0 -34.6 ( 1.0

a

Calculated as δ13C of CHCl3 at the end of the reaction minus δ13C of CCl4 at the beginning. b Calculated as δ13Cother ) δ13CCCl4 + ∆δ13Cother ) δ13CCCl4 - [(f(CHCl3)∆δ13CCHCl3)/f(other)]; where fCHCl3 and fother stand for the fraction of formed chloroform and other products, respectively.

can conclude from the depletion of heavy isotopes in chloroform that the steps leading from the trichloromethyl radical to CHCl3 (left branch in the lower part of Figure 4) involved a somewhat higher carbon isotope fractionation than the average of all of the alternative parallel steps (right branch in the lower part of Figure 4). Importantly, as shown by Tables 2 and 3, the lower the yield of chloroform is, the more pronounced is this isotopic depletion. (If all of the CCl4 is converted to CHCl3, then chloroform will have the isotope signature of the original CCl4 and there will be no depletion at all!) The isotopic signature of chloroform after complete conversion of CCl4 in the field can therefore be a very valuable 5640

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 15, 2005

We are grateful to Carsten Schubert for the elemental analyzer-IRMS measurements of the pure organic analytes and to Weihua Bian for experimental assitance. We particularly thank Michael Berg, Thomas Hofstetter, and Torsten Schmidt for fruitful discussions and critical comments as well as three anonymous reviewers who helped to improve the manuscript. This work was partially funded by the Swiss National Science Foundation (grant no. 21-57171.99).

Note Added after ASAP Publication Footnote d of Table 2 has been modified. The article was originally published ASAP on June 22, 2005. The corrected version was published ASAP on July 8, 2005.

Literature Cited (1) Zwank, L.; Berg, M.; Elsner, M.; Schmidt, T. C.; Schwarzenbach, R. P.; Haderlein, S. B. A new evaluation scheme for twodimensional isotope analysis to decipher biodegradation processes: Application to groundwater contamination by MTBE. Environ. Sci. Technol. 2005, 39, 1018-1029. (2) Bolliger, C.; Ho¨hener, P.; Hunkeler, D.; Ha¨berli, K.; Zeyer, J. Intrinsic bioremediation of a petroleum hydrocarbon-contaminated aquifer and assessment of mineralization based on stable carbon isotopes. Biodegradation 1999, 201-217. (3) Sherwood Lollar, B.; Slater, G. F.; Ahad, J.; Sleep, B.; Spivack, J.; Brennan, M.; Mackenzie, P. Contrasting carbon isotope fractionation during biodegradation of trichloroethylene and

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11) (12)

(13) (14) (15)

(16) (17)

(18)

(19)

toluene: Implications for intrinsic bioremediation. Org. Geochem. 1999, 30, 813-820. Sherwood Lollar, B.; Slater, G. F.; Sleep, B.; Witt, M.; Klecka, G. M.; Harkness, M.; Spivack, J. Stable carbon isotope evidence for intrinsic bioremediation of tetrachloroethene and trichloroethene at Area 6, Dover Air Force Base. Environ. Sci. Technol. 2001, 35, 261-269. Stehmeier, L. G.; Francis, M. M.; Jack, T. R.; Diegor, E.; Winsor, L.; Abrajano, T. A. Field and in vitro evidence for in-situ bioremediation using compound-specific 13C/12C ratio monitoring. Org. Geochem. 1999, 30, 821-833. Bloom, Y.; Aravena, R.; Hunkeler, D.; Edwards, E.; Frape, S. K. Carbon isotope fractionation during microbial dechlorination of trichloroethene, cis-1,2-dichloroethene, and vinyl chloride: Implications for assessment of natural attenuation. Environ. Sci. Technol. 2000, 34, 2768-2772. Hunkeler, D.; Butler, B. J.; Aravena, R.; Barker, J. F. Monitoring biodegradation of methyl tert-butyl ether (MTBE) using compound-specific carbon isotope analysis. Environ. Sci. Technol. 2001, 35, 676-681. Hunkeler, D.; Aravena, R.; Parker, B. L.; Cherry, J. A.; Diao, X. Monitoring oxidation of chlorinated ethenes by permanganate in groundwater using stable isotopes: Laboratory and field studies. Environ. Sci. Technol. 2003, 37, 798-804. Kolhatkar, R.; Kuder, T.; Philp, R. P.; Allen, J.; Wilson, J. T. Use of compound-specific stable carbon isotope analysis to demonstrate anaerobic biodegradation of MTBE in groundwater at a gasoline release site. Environ. Sci. Technol. 2002, 36, 51395146. Pause, L.; Robert, M.; Save´ant, J.-M. Reductive cleavage of carbon tetrachloride in a polar solvent. An example of a dissociative electron transfer with significant attractive interaction between the caged product fragments. J. Am. Chem. Soc. 2000, 122, 98299835. Clark, I. D.; Fritz, P. Environmental Isotopes in Hydrogeology; Lewis Publishers (CRC Press): Boca Raton, FL, 1997. Mariotti, A.; Germon, J. C.; Hubert, P.; Kaiser, P.; Letolle, R.; Tardieux, A.; Tardieux, P. Experimental determination of nitrogen kinetic isotope fractionation: Some principles; Illustration for the denitrification and nitrification processes. Plant Soil 1981, 62, 413-430. Slater, G. F.; Ahad, J.; Sherwood Lollar, B.; Allen-King, R.; Sleep, B. Carbon isotope effects resulting from equilibrium sorption of dissolved VOCs. Anal. Chem. 2000, 72, 5669-5672. Poulson, S. R.; Drever, J. I. Stable isotope (C, Cl, and H) fractionation during vaporization of trichloroethylene. Environ. Sci. Technol. 1999, 33, 3689-3694. Slater, G. F.; Dempster, H. S.; Sherwood Lollar, B.; Ahad, J. Headspace analysis: A new application for isotopic characterization of dissolved organic contaminants. Environ. Sci. Technol. 1999, 33, 190-194. Beneteau, K. M.; Aravena, R.; Frape, S. K. Isotopic characterization of chlorinated solvents - Laboratory and field results. Org. Geochem. 1999, 30, 739-753. Meckenstock, R. U.; Morasch, B.; Warthmann, R.; Schink, B.; Annweiler, E.; Michaelis, W.; Richnow, H. H. 13C/12C isotope fractionation of aromatic hydrocarbons during microbial degradation. Environ. Microbiol. 1999, 1, 409-414. Richnow, H. H.; Annweiler, E.; Michaelis, W.; Meckenstock, R. U. Microbial in situ degradation of aromatic hydrocarbons in a contaminated aquifer monitored by carbon isotope fractionation. J. Contam. Hydrol. 2003, 1-20. Richnow, H. H.; Meckenstock, R. U.; Reitzel, L. A.; Baun, A.; Ledin, A.; Christensen, T. H. In situ biodegradation determined

(20)

(21) (22)

(23)

(24)

(25)

(26)

(27)

(28)

(29)

(30)

(31)

(32)

(33)

by carbon isotope fractionation of aromatic hydrocarbons in an anaerobic landfill leachate plume (Vejen, Denmark). J. Contam. Hydrol. 2003, 1-14. Elsner, M.; Haderlein, S. B.; Kellerhals, T.; Luzi, S.; Zwank, L.; Angst, W.; Schwarzenbach, R. P. Mechanisms and products of surface-mediated reductive dehalogenation of carbon tetrachloride by Fe(II) on goethite. Environ. Sci. Technol. 2004, 38, 2058-2066. Huskey, W. P. In Enzyme Mechanisms from Isotope Effects; Cook, P. F., Ed.; CRC Press: Boca Raton, FL, 1991. Elsner, M.; Zwank, L.; Hunkeler, D.; Schwarzenbach, R. P. A new concept linking observable stable isotope fractionation to transformation pathways of organic pollutants. Environ. Sci. Technol., in press. Hunkeler, D.; Aravena, R. Evidence of substantial carbon isotope fractionation among substrate, inorganic carbon, and biomass during aerobic mineralization of 1,2-dichloroethane by Xanthobacter autotrophicus. Appl. Environ. Microbiol. 2000, 66, 4870-4876. Danielsen, K. M.; Hayes, K. F. pH dependence of carbon tetrachloride reductive dechlorination by magnetite. Environ. Sci. Technol. 2004, 38, 4745-4752. Elsner, M.; Schwarzenbach, R. P.; Haderlein, S. B. Reactivity of Fe(II)-bearing minerals towards reductive transformation of organic contaminants. Environ. Sci. Technol. 2004, 38, 799. Pecher, K.; Haderlein, S. B.; Schwarzenbach, R. P. Reduction of polyhalogenated methanes by surface-bound Fe(II) in aqueous suspensions of iron oxides. Environ. Sci. Technol. 2002, 36, 17341741. Fadrus, H.; Maly, J. Suppression of iron(III) interferences in the determination of iron(II) in water by the 1,10-phenanthroline method. Analyst; 1975, 549-554. Zwank, L.; Schmidt, T. C.; Berg, M.; Haderlein, S. B. Compoundspecific carbon isotope analysis of volatile organic compounds in the low µg/L-range. Anal. Chem. 2003, 75, 5575-5583. Graf, T. Produktion kosmogener Nuklide in Meteoriten ETH Diss. Nr. 8515, Thesis Eidgeno¨ssische Technische Hochschule Zu ¨ rich, 1988. Klausen, J.; Tro¨ber, S. P.; Haderlein, S. B.; Schwarzenbach, R. P. Reduction of substituted nitrobenzenes by Fe(II) in aqueous mineral suspensions. Environ. Sci. Technol. 1995, 29, 23962404. Heijman, C. G.; Holliger, C.; Glaus, M. A.; Schwarzenbach, R. P.; Zeyer, J. Abiotic reduction of 4-chloronitrobenzene to 4-chloroaniline in a dissimilatory iron-reducing enrichment culture. Appl. Environ. Microbiol. 1993, 59, 4350-4353. McCormick, M. L.; Bouwer, E. J.; Adriaens, P. Carbon tetrachloride transformation in a model iron-reducing culture: Relative kinetics of biotic and abiotic reactions. Environ. Sci. Technol. 2002, 36, 403-410. Ru ¨ gge, K.; Hofstetter, T.; Haderlein, S. B.; Bjerg, P. L.; Knudsen, S.; Zraunig, C.; Mosbaek, H.; Christensen, T. Characterization of predominant reductants in an anaerobic leachate-contaminated aquifer by nitroaromatic probe compounds. Environ. Sci. Technol. 1998, 32, 23-31.

Received for review August 5, 2004. Revised manuscript received May 4, 2005. Accepted May 5, 2005. ES0487776

VOL. 39, NO. 15, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5641