1,1,2,2-Tetrachloroethane Reactions with OH-, Cr(II), Granular Iron

May 3, 2007 - Elimination of HCl by OH- gave isotope fractionation in 1,1,2 ... differed by 2.3‰ ± 1.0‰ in Cr(II) systems, but were equivalent in...
0 downloads 0 Views 180KB Size
Download PDF
Environ. Sci. Technol. 2007, 41, 4111-4117

1,1,2,2-Tetrachloroethane Reactions with OH-, Cr(II), Granular Iron, and a Copper-Iron Bimetal: Insights from Product Formation and Associated Carbon Isotope Fractionation M A R T I N E L S N E R , * ,† DAVID M. CWIERTNY,‡ A. LYNN ROBERTS,‡ AND BARBARA SHERWOOD LOLLAR† Stable Isotope Laboratory, University of Toronto, 22 Russell Street, Toronto, Ontario M5S 3B1, Canada, and Department of Geography and Environmental Engineering, Johns Hopkins University, 3400 North Charles Street, Baltimore Maryland 21218

Despite widespread implementation of zero-valent iron remediation schemes, the manner and order of chemical bond cleavage in iron-mediated organohalide transformations remains imperfectly understood. We present insights from carbon isotope fractionation for the dehalogenation of 1,1,2,2-tetrachloroethane (1,1,2,2-TeCA) and 1,1,1trichloroethane (1,1,1-TCA) by various reactants. Elimination of HCl by OH- gave isotope fractionation in 1,1,2,2-TeCA of  ) -25.6‰, KIEC ) 1.02 to 1.03 per carbon center, consistent with a concerted (E2) mechanism. In contrast, 1,1,1TCA reduction by Cr(II), Fe(0), and Cu-plated iron (Cu/Fe) resulted in  ) -13.6‰ to -15.8‰ indicating the initial involvement of a single C-Cl bond (KIEC ≈ 1.03). 1,1,2,2TeCA reduction by Cr(II), Fe(0), and Cu/Fe yielded  ) -18.7‰, -19.3‰, and -17.0‰, respectively. In the two latter cases, depletion of the minor product TCE by 26‰ indicated its formation via nonreductive dehydrohalogenation. The major 1,1,2,2-TeCA reduction products, cis- and trans-DCE, differed by 2.3‰ ( 1.0‰ in Cr(II) systems, but were equivalent in Fe(0) and Cu/Fe systems. In contrast, the ratio of cis-DCE to trans-DCE concentration was 2.5 for reduction with Cr(II) and Fe(0), but ∼3.8 with Cu/Fe. Complementary isotope and concentration data therefore suggest differences in the transition state geometry and/ or reaction intermediates in each reductant system.

Introduction Despite the implementation of zero-valent iron permeable reactive barriers at many groundwater sites contaminated by chlorinated solvents (1), relatively little is known about the surface-catalyzed reductive dehalogenation reactions * Corresponding author phone: +49 89 3187 2565; fax: +49 89 3187 3361; e-mail: [email protected]; current address: GSF National Research Center for Environment and Health, Institute of Groundwater Ecology, Ingolstaedter Landstr. 1, D-85764 Neuherberg, Germany. † University of Toronto. ‡ Johns Hopkins University. 10.1021/es063040x CCC: $37.00 Published on Web 05/03/2007

 2007 American Chemical Society

occurring in these systems. An issue of great interest is the mechanism of organohalide reduction, specifically, the manner and order in which chemical bonds are broken in the chlorinated compound, as this can dictate what products are ultimately formed in iron-based treatment systems. Organohalide dehalogenation by granular iron may be either stepwise or concerted. Equations I and II in Scheme 1 show stepwise vicinal dehalogenations, which involve the cleavage of a single carbon-halogen bond in the ratedetermining step of the reaction with electron donor D. For such stepwise vicinal dehalogenations, a dissociative single electron transfer, SET, (eq I) (2-5), or an X-philic twoelectron-transfer reaction (eq II) (6) have been suggested as possible mechanisms. In contrast, two bonds are broken at the same time in concerted dehalogenation reactions. Equation III displays such a concerted mechanism as an alternative pathway for an X-philic vicinal dehalogenation, whereas eq IV shows a concerted dehydrohalogenation reaction by non-reductive elimination of HCl, as it has been proposed for certain halogenated solvents (7); also here, a stepwise dehydrohalogenation mechanism is possible, although it is not explicitly shown in Scheme 1. In the case of eqs I and II, radical or anionic intermediates are formed, possibly stabilized in the form of complexes with the iron surface or organometallic intermediates (5, 8-10). Detection of such species can potentially provide evidence of the prevailing reaction mechanism. The direct observation of such intermediates via analytical and spectroscopic methods, however, is a challenging proposition, as they are highly unstable, short-lived, and only accumulate to very low concentrations. Consequently, most mechanistic insights to date have been based on kinetic studies, which primarily yield indirect evidence pertaining to the first rate-determining step (11-14), or on evidence from product distributions, which may lend insight into the nature of possible intermediates (5, 8, 12, 15, 16). Compound-specific isotope analysis, CSIA, is a promising new method for elucidating reaction mechanisms that is complementary to more conventional approaches (24). Using gas chromatography-isotope ratio mass spectrometry (GCIRMS), isotope ratios (e.g., 13C/12C) can be determined for individual organic constituents of a reaction mixture even at low, environmentally relevant concentrations in the ppm to ppb range (19, 20). These isotope ratios can change or be fractionated during transformation because heavy isotopes (e.g., 13C) generally react more slowly than light isotopes (e.g., 12C). Consequently, the substrate becomes enriched, while the average of all products is commensurately depleted in the heavy isotopes (21). Fractionation in the substrate provides information relating to the first rate-limiting step, whereas isotope fractionation in the products may also reflect differences in the subsequent reaction steps involved in their generation. The isotope fractionation in the substrate is most informative if it can be matched against exact computational predictions of kinetic isotope effects (22, 23). In a more empirical manner, it may be compared to known characteristic ranges of fractionation for (bio)chemical reactions (24). In Scheme 1, typical values of carbon kinetic isotope effects, KIEC ) 12k/13k, as tabulated in ref 24 are given, where 12k and 13k are the reaction rates of molecules with 12C and 13C in the respective positions. The values illustrate that isotope fractionation is expected primarily in the reacting bonds. In contrast, CSIA measures the isotope fractionation as averaged over the entire reacting compound (i.e., in “bulk”). This average fractionation may, therefore, become VOL. 41, NO. 11, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4111

SCHEME 1

more pronounced if multiple bonds are simultaneously involved in a reaction. To elucidate reaction mechanisms, observable fractionation may be evaluated under the assumption of different scenarios such as those shown in Scheme 1, the results may be compared to typical literature values, and the respective scenarios can be confirmed or discarded. Such an approach has recently been used to infer the nature of the rate-limiting steps in transformations of carbon tetrachloride (10, 26), 1,2-dichloroethane (27), methyl tert-butyl ether (28), and vinyl chloride (29). In contrast, few if any studies in environmental chemistry have used measurements of product isotope ratios for mechanistic elucidation, despite their potential for providing useful information. It was the objective of the current study to use the combined information from product distribution and isotope analysis to obtain new insight into the stepwise versus concerted mechanism of organohalide dehalogenation. 1,1,2,2-Tetrachloroethane (1,1,2,2-TeCA) was chosen, as it possesses two equivalent reactive carbon centers whose C-Cl bonds can be broken in a stepwise or concerted manner (see Scheme 1). 1,1,1-Trichloroethane (1,1,1-TCA) was included because it only possesses a single reactive CCl3 group and is thus unlikely to undergo reduction involving multiple carbon centers. The reactants considered were (i) Cr(II), as a model one-electron reductant that has been postulated to reduce organohalides in a stepwise, inner-sphere fashion (14, 18), and (ii) Fe(0) and Cu/Fe, as reductants for which the mechanism is not yet precisely known but is often proposed to be stepwise (2, 4, 6). The dehydrohalogenation of 1,1,2,2TeCA to TCE by OH- was also included as an example of a transformation with concerted C-Cl and C-H bond breaking. In addition to isotope fractionation in the substrate, this studies explores for the first time isotope analysis of 1,1,2,2TeCA reaction products as an additional means to investigate intermediates and consecutive reaction steps in reductive organochloride dehalogenation. 4112

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 11, 2007

Materials and Methods Experimental Setup. A list of chemicals, as well as additional experimental and analytical details, is provided in the Supporting Information, hereafter referred to as SI. Briefly, Fisher electrolytic iron powder, 100 mesh, was pre-cleaned by acid treatment as described in Cwiertny and Roberts (30) and was either used immediately or after plating with Cu according to the procedure in Cwiertny et al. (6). Cr(II) solutions were prepared by reducing solutions of Cr(III) sulfate with zero-valent zinc as described in Kohn et al. (14). Reactors were constructed under anoxic conditions (95% N2, 5% H2). Experimental vials contained 100 mL of aqueous solution or suspension, and 150 mL of headspace. Reactions with Cr(II) were conducted entirely within the anaerobic chamber, while experiments with Fe(0) and Cu/Fe systems were performed outside the anaerobic chamber. All reactors were sufficiently mixed at all times to ensure rapid solid/ water and water/air mass transfer and were kept overpressurized by injecting aliquots of N2 gas prior to headspace and liquid sampling. Because the reaction between Cr(II) and 1,1,1-TCA was too rapid to allow for real-time sample analysis, multiple 20 mL reactors were prepared. Duplicate seriessone with 5 mL headspace and one withoutscontained between 0.7 and 12.6 mM Cr(II) in 5 mM H2SO4 at pH 2. After addition of 1,1,1TCA and rapid shaking, these systems were allowed sufficient time to react to completion. The range of Cr(II) concentrations used in each reactor resulted in a different extent of reaction progress with respect to 1,1,1-TCA conversion. Subsequent analyses therefore allowed us to monitor the isotopic fractionation in 1,1,1-TCA as a function of its degree of transformation. Duplicate experiments with and without headspace gave the same result within the 95% confidence interval of the linear regression indicating that air-water exchange was slower than the reaction and did not bias our results. Concentration and Isotope Analysis. Details pertaining to our analytical methods are provided in the SI. Briefly,

concentrations of organohalide reactants and products were determined on a gas chromatograph equipped with a flame ionization detector (GC/FID). Total uncertainty in concentrations is (5%. In the case of 1,1,1-TCA reduction, product species were identified, but not quantified. With the exception of 1,1,2,2-TeCA, all compounds were analyzed by direct headspace sampling with 200 µL sample volume. For 1,1,2,2TeCA analysis, 0.7 mL of aqueous sample was extracted with 1.0 mL pentane, 1 µL of which was subsequently analyzed via GC/FID. Toluene present in the aqueous solution of experimental systems served as an internal standard to correct for injection volume. Isotope ratios of all compounds were analyzed by direct headspace analysis on a gas chromatograph-combustion-isotope ratio mass spectrometer (Finnigan MAT 252) and are reported as δ13C values relative to the international standard VPDB: δ13C ) (R/RVPDB - 1) × 1000, where R and RVPDB are the absolute isotope ratios of compound and international standard, respectively. As discussed in more detail in the SI and shown in Figure S1, the uncertainty of each sample incorporating accuracy and reproducibility is (0.5‰ (32, 33). Calculation of Enrichment Factors () and Apparent Kinetic Isotope Effects (AKIE). Carbon isotopic fractionation factors (RC) and enrichment factors (C) were obtained according to the logarithmic form of the Rayleigh equation

ln

C (1000 + δ13C) R ) (RC - 1)‚ln f ) ln f ) ln 13 R0 1000 (1000 + δ C0) (1)

where δ13C0 and δ13C are isotope values measured at the beginning or during the course of the reaction, respectively, and f is the fraction of substrate remaining at the time in question. Enrichment factors C are reported in ‰ and describe the bulk fractionation over the whole compound. In contrast, position-specific fractionation is reported in terms of kinetic isotope effects, KIEC ) 12k/13k where 12k and 13 k are the reaction rates of molecules with 12C and 13C in the reactive position. KIE values are referred to as intrinsic if they directly reflect the isotope fractionation in an elementary chemical reaction. They are called apparent (AKIE values) if additional slow reaction steps cannot be precluded that may mask the intrinsic KIEC value (24). In the case of 1,1,2,2-TeCA enrichment factors, C can be converted into AKIEC or KIEC values according to the equation

AKIEC )

( ) 12

13

k k

)

apparent

1 1 + z‚C/1000

(2)

where z is the number of positionally equivalent reactive carbon positions between which intramolecular competition exists. In the case of a stepwise reaction z ) 2, and for a concerted reaction z ) 1 (24). 1,1,1-TCA requires a different type of correction that takes into account nonreacting carbon atoms and in simplified form is described by (24)

AKIEC )

( ) 12

13

k k

≈

apparent

1 1 + n/x‚C/1000

(3)

where n is the total number of carbon atoms of which x are located in reactive positions. In the case of 1,1,1-TCA, n ) 2 and x ) 1. Evaluation of Product Isotope Data. Product isotope ratios were fit according to a formula adapted from Melander and Saunders (25)

(

)

D(δ13C) 1 - f (/1000 + 1) 1000 + δ13C ) 1 + ‚ 1000 1-f 1000 + δ13Csubstrate,0 (4)

where δ13C are isotope values of the respective product measured during the course of the reaction, f is the fraction of substrate remaining at that time, δ13Csubstrate,0 is the initial isotope value of the substrate and  is the enrichment factor that has been determined according to eq 1. D(δ13C) is the only fitting parameter in this expression; it describes the extent to which product isotope ratios δ13C deviate from expectations based solely on  and δ13Csubstrate,0.

Results and Discussion Figures 1 and 2 show concentrations, cis-/trans-DCE product ratios, and measured isotope ratios for the different reactions of 1,1,2,2-tetrachloroethane. Concentration versus time data for the reaction with OH- is not included, as TCE was formed as the only product in 100% yield (see Figure S4 in the SI). Good molar balances indicate that all relevant products were taken into account and that detected trace products (ethane, ethylene, and coupling products) did not contribute substantially to isotopic mass balances. Regressions of 1,1,2,2TeCA isotope data according to eq 1 are given in Table 1. Model fits to 1,1,2,2-TeCA isotope ratios are also depicted in Figure 2, together with product isotope data fit according to eq 4. All R 2 values were better than 0.993 in the case of eq 1 and better than 0.96 in the case of eq 4, indicating an excellent fit of all data to the Rayleigh model. All enrichment factors are listed along with ( 95% confidence intervals in Table 1. Isotope Fractionation in the Parent Compounds. Possible Effects of Masking. As indicated in Scheme 1, typical kinetic isotope effects for reductive cleavage of aliphatic C-Cl bonds are KIEC ) 1.02-1.03, and for cleavage of a C-H bond are KIEC ) 1.01-1.03 (24). Similar values would also be expected in the experiments, provided that there are no slow reaction steps that would mask the intrinsic kinetic isotope effect, KIEC, in the observable AKIEC value. The experimental setup, in particular the solids loading and mixing speed employed in heterogeneous systems (see SI), was carefully chosen to preclude rate limitation by air-water exchange or mass transfer in solution. In homogeneous reaction with OH- and Cr(II) masking can most likely be excluded, while in the Fe(0) and Cu/Fe reaction systems, slow surface association steps remain a potential source of masking. The consistently high fractionation summarized in Table 1 suggests that slow mass transfer was insignificant in all experiments, however, and that calculated AKIEC values likely accurately reflect intrinsic kinetic isotope effects. Accordingly, results in this study are consistently reported as KIEC. Dehydrohalogenation of 1,1,2,2-TeCA. In the reaction of 1,1,2,2-TeCA with OH-, an isotopic enrichment factor of -25.6‰ ( 0.8‰ was observed, which translates into a KIEC of 1.0262 ( 0.0009 if a concerted scenario is assumed (eq 2, Table 1). This is consistent with the range of 1.02-1.03 expected for C-H and C-Cl bond cleavage, respectively (see Scheme 1). In contrast, the calculated value of KIEC ) 1.0539 ( 0.0019 based on assuming a stepwise scenario of either initial C-H or C-Cl cleavage is not compatible with such a mechanistic model. The observed isotope fractionation therefore supports the concerted nature of the base-catalyzed dehydrohalogenation of 1,1,2,2-TeCA in an E2 mechanism. Reductive Dehalogenation of 1,1,1-TCA. The smaller magnitude of isotope fractionation we measured in reductive dehalogenation of 1,1,1-TCA by Cr(II), Fe(0), and Cu/Fe (-13.6‰ to -15.8‰) is consistent with a mechanism that involves, as hypothesized, only one of the carbon centers, that is, the CCl3 group. Products detected included 1,1dichloroethane as well as ethane and ethylene. Good agreement of calculated KIEC values (1.028 to 1.032) with the literature values (1.02 to 1.03) is consistent with reductive cleavage of a single C-Cl bond during the rate-determining step. As shown in Table 1 and the graphs in Figure S2 of the VOL. 41, NO. 11, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4113

FIGURE 1. Concentrations over time data from one representative experiment of different 1,1,2,2-TeCA transformation reactions. Lines are straight connections between measured points, and symbols are as follows: diamonds (1,1,2,2-TeCA), upward triangles (transDCE), inverse triangles (cis-DCE), circles (TCE). Squares indicate molar mass balances, and solid hexagons show cis-/trans-DCE product ratios. Error bars for concentrations are (5%. SI, little difference in the measured isotopic enrichment factor for 1,1,1-TCA was observed between systems with the known inner sphere SET reductant Cr(II) (14, 17, 18), with Fe(0), for which both inner or outer sphere SET has been postulated (2, 4), and with Cu/Fe, for which involvement of a hydridelike species has been inferred (6). Previous results obtained with CCl4 (10, 26) indicated that  values were primarily influenced by the number of sp3-hybridized C-Cl bonds undergoing cleavage, in all cases the cleavage of one C-Cl bond. In contrast, the differences depending on the identity 4114

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 11, 2007

of the active reductant were more subtle, varying between -29‰ ( 3‰ with Fe oxides, -22‰ with polysulfide, and -16‰ with iron sulfides. The fact that the variations observed in this study were even smaller indicates that the identity of the active reductant had little effect on fractionation. Less likely, the electron-transfer mechanism involving the three model reductants may be different from the previously hypothesized pathways. Further systematic research is necessary to resolve this question. Reductive Dehalogenation of 1,1,2,2-TeCA. The enrichment factors of -17.0‰ to -19.3‰ measured for the reactions of 1,1,2,2-TeCA with Cr(II), Fe(0), and Cu/Fe were consistently higher than the corresponding values for 1,1,1-TCA, but were significantly smaller than the value observed for the concerted dehydrochlorination of 1,1,2,2-TeCA. If C values are converted into KIEC values under the assumption of either a concerted or a stepwise reaction scenario, the results of KIEC ) 1.017-1.020 and KIEC ) 1.035-1.040 fall to the lower and higher end, respectively, of the expected ranges (see Table 1). Mechanistic interpretations are difficult in such a situation, because neither scenario can be discarded. Accurate computational predictions would be required for definite conclusions. To this end one would have to calculate transition state hypersurfaces for the reaction of 1,1,2,2-TeCA, compute the corresponding molecular structures, calculate all contributions to the resulting kinetic isotope effects, and average these contributions over all possible reaction paths (22). As not only molecular structures but also the relevant metal surface would have to be included, such a sophisticated treatment is not feasible with the poorly characterized heterogeneous metal surfaces used in this study. However, because a concerted scenario would involve not only carbon atoms, but also the chlorine centers, further insight may possibly be expected from future measurements of chlorine isotope effects. Isotope Fractionation in the Products. In order to obtain additional mechanistic information, and also to confirm the consistency between isotope fractionation of 1,1,2,2-TeCA and its products, isotope ratios of the products were fit with eq 4 using enrichment factors obtained from 1,1,2,2-TeCA isotope data. Figure 2 shows the corresponding graphs, and Table 1 summarizes the fitting parameters D(δ13C), which describe the difference between measured and predicted product isotope ratios. The excellent fits shown in Figure 2, and the observation that all fitting parameters in reactions with OH-, Cu/Fe, and Fe(0) fall within the analytical uncertainty of (1‰ (combined uncertainty of (0.5‰ for both substrate and product isotope ratios), indicate very good consistency. Differences in Isotope Ratios between cis- and trans-DCE. Isotope ratios of cis- and trans-DCE were the same within experimental error during reduction by Fe(0) and Cu/Fe. This indicates that activation energies of all comparable reaction stepssincluding those involving possible intermediatess had the same isotopic discrimination. In contrast, product isotope ratios in the reaction with Cr(II) showed a small but significant depletion of trans-DCE compared to cis-DCE (Figure 2). This isotopic discrimination of 2.3‰ ( 1.0‰ is also expressed in the D(δ13C)-values of Table 1, where the cis-/trans-DCE product ratio of 2.4:1 ensures that the shift in D(δ13C) for trans-DCE of 1 × (-1.7‰) is isotopically balanced by a corresponding shift of 2.4 × (+0.6‰) for cisDCE. In the case of Cr(II), the pathway to trans-DCE must therefore have involved one or several steps with a significantly greater isotope fractionation than the corresponding step(s) in the formation pathway of cis-DCE. This isotopic discrimination results from differences in the stereochemistry of reactive intermediates and/or transition state structures. Kray and Castro (34) postulated the reduction of 1,1,2,2TeCA by Cr(II) involved the formation of an organometallic

FIGURE 2. Isotope ratios including combined data from duplicate experiments for the different transformation reactions of 1,1,2,2-TeCA. Symbols are as follows: diamonds (1,1,2,2-TeCA), upward triangles (trans-DCE), inverse triangles (cis-DCE), circles (TCE). Lines are fits with the parameters given in Table 1. Total uncertainty for δ13C values is (0.5‰ incorporating both accuracy and reproducibility after the method of Mancini et al. and Gray et al. (32, 33).

TABLE 1. Measured Isotope Enrichment Factors E, Kinetic Isotope Effects KIE,a Product Isotope Ratio Fitting Parameters, and cis-/trans-DCE Product Concentration Ratios in Reactions of 1,1,2,2-TeCA and 1,1,1-TCA 1,1,2,2-TeCA

elimination of HCl

reductive dehalogenation by Cr(II)

reductive dehalogenation by Fe

1,1,1-TCA reductive dehalogenation by Cu/Fe

reductive dehalogenation by Cr(II)

reductive dehalogenation by Fe

reductive dehalogenation by Cu/Fe

enrichment factor  KIEC (concerted scenario) KIEC (stepwise scenario)

-25.6‰ ( 0.8‰b -18.0‰ ( 0.5‰b

-19.3‰ ( 0.7‰b -17.0‰ ( 0.6‰b -15.8‰ ( 0.6‰b -13.6‰ ( 0.5‰b

-13.7‰c

1.0262 ( 0.0009

1.0184 ( 0.0006

1.0196 ( 0.0008

1.0173 ( 0.0006

n.a.d

n.a.d

n.a.d

1.0539 ( 0.0019

1.0374 ( 0.0012

1.0401 ( 0.0014

1.0351 ( 0.0012

1.0326 ( 0.0012

1.0279 ( 0.0010

1.0281c

TCE

 ) -25.6‰e D(δ13C) ) 0.6‰f

parameters for modeling of δ13C product data according to eq 8 with δ13Csubstrate,0 ) 0.1‰ (see Figure 1)

trans-DCE

 ) -18.0‰e  ) -19.3‰e D(δ13C) ) -1.7‰f D(δ13C) ) 0.7‰f

 ) -17.0‰e D(δ13C) ) 0.2‰f

cis-DCE

 ) -18.0‰f D(δ13C) ) 0.6‰e

 ) -19.3‰f D(δ13C) ) 0.4‰e

 ) -17.0‰f D(δ13C) ) 0.7‰e

cis-/trans-DCE product ratio

2.6 ( 0.2

2.5 ( 0.2

3.6 ( 0.3g 3.9 ( 0.3g

a Calculated according to eqs 7 and 8. b Regression through the combined data of two replicate experiments. c Regression data of only one experiment. d Not applicable; only the CCl3 group is involved in reaction of 1,1,1-TCA so that participation of both carbon centers can be excluded. e Fixed input parameter as obtained from regression of substrate (1,1,2,2-TeCA) δ13C data. f Fitting parameter of eq 8. g Data from duplicate experiments.

intermediate and subsequent cis-elimination of Cr(H2O)5Cl. It remains to be investigated using computational methods whether this mechanism can explain the observed isotopic discrimination. Despite cis- and trans-DCE representing the major products in both Cr(II) and iron-based reductant systems, the observed isotope fractionation certainly suggests differences in the geometry of transition states and/or intermediates

involved in 1,1,2,2-TeCA transformation in homogeneous and heterogeneous reductant systems. This interpretation would not have been obtained without compound-specific isotope analysis of the products of these reactions. It underscores the notion that trends in rates and product distribution are necessary, but not sufficient, evidence in inferring that products are generated via identical steps in reactions with different reductants. VOL. 41, NO. 11, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4115

Insight from Isotope Ratios of TCE. As demonstrated in Figure 1 and 2, TCE was formed as a minor product during reaction of 1,1,2,2-TeCA with Fe(0) and showed the same isotopic discrimination as that formed by base-catalyzed elimination of HCl. In particular, the characteristic isotopic discrimination of  ) -26‰ ( 1.0‰ that was observable in TCE relative to 1,1,2,2-TeCA at the beginning of the reaction suggests that the reaction producing TCE formation in the 1,1,2,2-TeCA/Fe(0) system is the same as that in the 1,1,2,2TeCA/OH- system. In previous work, Arnold and co-workers (5) noted that substantially more TCE was generated in reaction of 1,1,2,2-TeCA with Fe(0) than predicted based upon the reaction half-life for dehydrohalogenation, which is about 3 months at pH 7 (7). The authors proposed an alternative reaction pathway in which the initial step of Fe(0)-catalyzed 1,1,2,2-TeCA reduction produced an organometallic surface complex. TCE was postulated to form from this intermediate via a pathway that competed with the formation of cis- and trans-DCE as alternative products. As discussed above, isotope data from the current study provide evidence against such a scenario for TCE formation. We suggest as an alternative explanation that the parallel reduction of water in Fe(0) and Cu/Fe systems results in a steady pH increase over the duration of the 1,1,2,2-TeCA reduction reaction (30), thereby enhancing the rate of TCE formation over time. Such a scenario would produce greater concentrations of TCE than predicted from the initial pH in iron-based systems, and is supported by similar conclusions of Cwiertny, who earlier investigated the rate of 1,1,2,2-TeCA reduction in Fe(0) systems (30, 35). Although the pH was not measured directly, additional evidence for an increase in pH during the reaction is given by the fact that TCE product yields increased over the course of the reaction (see Figure S3 in the SI). This is also reflected in the observation that TCE isotope ratios became increasingly enriched toward the end of the reaction (see Figure 2). Normally, it would be expected that TCE would show a constant isotopic discrimination relative to the other products throughout the reaction of 1,1,2,2-TeCA with Fe(0), such as observed between cis- and trans-DCE in the reaction with Cr(II). However, the increasing TCE yield over time resulted in a greater fraction of TCE molecules originating from more isotopically enriched 1,1,2,2-TeCA at later times, and as a result, a less negative TCE isotope signature than expected was observed. This was most pronounced in the reaction of 1,1,2,2-TeCA with Cu/Fe, for which TCE isotopic signatures even exceeded the original isotope ratio of 1,1,2,2-TeCA (see Figure 2). Combined Insight from Product Concentrations and Isotope Data. In reactions of 1,1,2,2-TeCA with Cr(II) and Fe(0), cis-DCE and trans-DCE were formed in a relatively consistent proportion of about 2.5, whereas in the reaction with Cu/Fe the ratio was 3.5 to 4.0 (see Figure 1 and Table 1). These results diverge somewhat with previous work by Arnold and co-workers (5), who found the same cis-DCE to trans-DCE ratio of 2.5 with Cr(II), but a much larger ratio of 4.5 in Fe(0) systems. Using Marcus theory, the authors calculated that the reaction of intermediate 1,2,2,-trichloroethyl radicals would lead to a corresponding product ratio of about 2. They suggested that values significantly deviating from this ratio were indicative of different kinds of intermediates and hypothesized the involvement of organometallic intermediates in the case of Fe(0). Our results show that a high cis- to trans-DCE product ratio is not necessarily characteristic of an Fe(0) surfacecatalyzed reaction but, in this case, appears to result from the presence of Cu as an additional metal at the granular iron surface. It is possible that such a result is evidence for the involvement of the organometallic intermediate proposed by Arnold et al. in our Cu/Fe systems, suggesting that certain 4116

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 11, 2007

granular iron additives stabilize reactive intermediates such as carbon-containing radicals, carbenes, or carbanions. Accordingly, the cis-to-trans-DCE ratio of 4 observed by Arnold et al. in unamended iron systems could have resulted from impurities in their iron, as others (9) have proposed that such impurities influence both the rate and product distribution of organohalide reduction by iron. Further investigation of such a scenario will be required, however, because prior work by our group (6) that examined the reduction of 1,1,1-TCA by a suite of bimetallic reductants did not reveal trends in product formation that were consistent with a prominent role for organometallic intermediates. Our results show that the information from compoundspecific isotope analysis is complementary to insight from product studies. In particular, this study is one of the first to highlight the importance of measuring not only substrate, but also product isotope ratios. It shows that for mechanistic conclusions, it may be essential to consider product isotope fractionation, as cis- to trans-DCE concentration ratios combined with substrate isotope fractionation alone would not elucidate the subtle differences in the mechanism of product formation in 1,1,2,2-TeCA reaction with Cr(II) and granular iron-based reductants. Measuring isotope fractionation is, therefore, a promising new approach in the suite of methods to elucidate the mechanism of environmentally relevant reactions and to better predict the formation of harmful versus benign organohalide degradation products to improve remediation strategies.

Acknowledgments This work was supported by the German Research Foundation (DFG) Research Fellowship 266/1-3 to the first author and by an EPA STAR Graduate Fellowship for the second author. Additional support was provided by the NSERC Strategic Grants Program to the Stable Isotope Lab/University of Toronto and through an U.S. NSF CRAEMS grant (CHE0089168) awarded to Johns Hopkins University. We thank Bill Arnold, Christopher Cramer, and Donald Truhlar for stimulating discussions. We also thank four anonymous reviewers whose insightful comments significantly improved the quality of this manuscript.

Supporting Information Available Further experimental information, data on the accuracy of measured isotope ratios (Figure S1), exemplary discussion of error propagation, isotope regression graphs of 1,1,1trichloroethane data (Figure S2), data on trichloroethene yields during 1,1,2,2-TeCA transformation in the presence of Fe(0) (Figure S3), and concentration versus time data for the reaction of 1,1,2,2-TeCA with OH- (Figure S4). This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (1) U.S. EPA. Field Applications of In Situ Remediation Technologies: Permeable Reactive Barriers; U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response, Technology Innovation Office: Washington, DC, 2002. (2) Fennelly, J.; Roberts, L. Reaction of 1,1,1-trichloroethane with zero-valent metals and bimetallic reductants. Environ. Sci. Technol. 1998, 32, 1980-1988. (3) Balko, B. A.; Tratnyek, P. G. Photoeffects on the reduction of carbon tetrachloride by zero-valent iron. J. Phys. Chem. 1998, 102, 1459-1465. (4) Li, T.; Farrell, J. Electrochemical investigation of the rate-limiting mechanisms for trichloroethylene and carbon tetrachloride reduction at iron surfaces. Environ. Sci. Technol. 2001, 35, 35603565. (5) Arnold, W. A.; Winget, P.; Cramer, C. J. Reductive dechlorination of 1,1,2,2-tetrachloroethane. Environ. Sci. Technol. 2002, 36, 3536-3541.

(6) Cwiertny, D. M.; Bransfield, S. J.; Livi, K. J. T.; Fairbrother, D. H.; Roberts, A. L. Exploring the influence of granular iron additives on 1,1,1-trichloroethane reduction. Environ. Sci. Technol. 2006, 40, 6837-6843. (7) Roberts, A. L.; Jeffers, P. M.; Wolfe, N. L.; Gschwend, P. M. Structure-reactivity relationships in dehydrohalogenation reactions of polychlorinated and polybrominated alkanes. Crit. Rev. Environ. Sci. Technol. 1993, 23, 1-39. (8) Totten, L. A. The Use of Model and Probe Compounds to Investigate the Mechanisms of Reductive Dehalogenation Reactions. Ph.D. Thesis, Johns Hopkins University, Baltimore, 1999. (9) Ta´mara, M. L.; Butler, E. B. Effects of iron purity and groundwater characteristics on rates and products in the degradation of carbon tetrachloride by iron metal. Environ. Sci. Technol. 2004, 38, 1866-1876. (10) 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. (11) Scherer, M. M.; Balko, B. A.; Gallagher, D. A.; Tratnyek, P. G. Correlation analysis of rate constants for dechlorination by zerovalent iron. Environ. Sci. Technol. 1998, 32, 3026-3033. (12) Arnold, W. A.; Roberts, L. Pathways of chlorinated ethylene and chlorinated acetylene reaction with Zn(0). Environ. Sci. Technol. 1998, 32, 3017-3025. (13) Tratnyek, P. G.; Weber, E. J.; Schwarzenbach, R. P. Quantitative structure-activity relationships for chemical reductions of organic contaminants. Environ. Toxicol. Chem. 2003, 22, 17331742. (14) Kohn, T.; Arnold, W. A.; Roberts, A. L. Reactivity of substituted benzotrichlorides toward granular iron, Cr(II), and an iron(II) porphyrin: a correlation analysis. Environ. Sci. Technol. 2006, 40, 4253-4260. (15) Totten, L. A.; Jans, U.; Roberts, A. L. Alkyl bromides as mechanistic probes of reductive dehalogenation: reactions of vicinal dibromide stereoisomers with zero-valent metals. Environ. Sci. Technol. 2001, 35, 2268-2274. (16) Arnold, W. A.; Roberts, A. L. Pathways and kinetics of chlorinated ethylene and chlorinated acetylene reaction with Fe(0) particles. Environ. Sci. Technol. 2000, 34, 1794-1805. (17) Castro, C. E. The role of halide in the reduction of carbonium ions by chromium(II). J. Am. Chem. Soc. 1961, 83, 3262-3264. (18) Castro, C. E.; Cray, W. C., Jr. The cleavage of bonds by low valent transition metal ions. The homogeneous reduction of alkyl halides by chromous sulfate. J. Am. Chem. Soc. 1963, 85, 2768-2773. (19) 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. (20) Hunkeler, D.; Aravena, R. Determination of compound-specific carbon isotope ratios of chlorinated methanes, ethanes, and ethenes in aqueous samples. Environ. Sci. Technol. 2000, 34, 2839-2844. (21) Bigeleisen, J. The relative reaction velocities of isotopic molecules. J. Chem. Phys. 1949, 17, 675-678. (22) Truhlar, D. G. Variational Transition State Theory and Multidimensional Tunneling for Simple and Complex Reactions in the Gas Phase, Solids, Liquids, and Enzymes. In Isotope Effects

(23)

(24)

(25) (26)

(27)

(28)

(29)

(30) (31)

(32)

(33)

(34)

(35)

in Chemistry and Biology; Limbach, H., Kohen, A., Eds.; Marcel Dekker: New York, 2006. Singleton, D. A.; Schulmeier, B. E.; Hang, C.; Thomas, A. A.; Leung, S. W.; Merrigan, S. R. Isotope effects and the distinction between synchronous, asynchronous, and stepwise Diels-Alder reactions. Tetrahedron 2001, 57, 5149-5160. 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. 2005, 39, 6896-6916. Melander, L.; Saunders, W. H., Jr. Reaction Rates of Isotopic Molecules; Wiley: New York, 1980. Zwank, L.; Elsner, M.; Aeberhard, A.; Schwarzenbach, R. P.; Haderlein, S. B. Carbon isotope fractionation during reductive dehalogenation of carbon tetrachloride at iron (hydr)oxide and iron sulfide minerals. Environ. Sci. Technol. 2005, 39, 56345641. Hirschorn, S. K.; Dinglasan, M. J.; Elsner, M.; Mancini, S. A.; Lacrampe-Couloume, G.; Edwards, E. A.; Sherwood Lollar, B. Pathway dependent isotopic fractionation during aerobic biodegradation of 1,2-dichloroethane. Environ. Sci. Technol. 2004, 38, 4775-4781. Zwank, L.; Berg, M.; Elsner, M.; Schmidt, T. C.; Schwarzenbach, R. P.; Haderlein, S. B. New evaluation scheme for twodimensional isotope analysis to decipher biodegradation processes: application to groundwater contamination by MTBE. Environ. Sci. Technol. 2005, 39, 1018-1029. Chartrand, M. M. G.; Waller, A.; Mattes, T. E.; Elsner, M.; Lacrampe-Couloume, G.; Gossett, J. M.; Edwards, E. A.; Sherwood Lollar, B. Carbon isotopic fractionation during aerobic vinyl chloride degradation. Environ. Sci. Technol. 2005, 39, 10641070. Cwiertny, D. M.; Roberts, A. L. On the nonlinear relationship between kobs and reductant mass loading in iron batch systems. Environ. Sci. Technol. 2005, 39, 8948-8957. Kelsall, G. H.; House, C. I.; Gudyanga, F. P. Chemical and electrochemical equilibria and kinetics in aqueous Cr(III)/Cr(II) chloride solutions. J. Electroanal. Chem. 1988, 244, 179201. Gray, J. R.; Lacrampe-Couloume, G.; Gandhi, D.; Scow, K. M.; Wilson, R. D.; Mackay, D. M.; Sherwood Lollar, B. Carbon and hydrogen isotopic fractionation during biodegradation of methyl tert-butyl ether. Environ. Sci. Technol. 2002, 36, 1931-1938. Mancini, S. A.; Lacrampe-Couloume, G.; Jonker, H.; van Breukelen, B. M.; Groen, J.; Volkering, F.; Sherwood Lollar, B. Hydrogen isotopic enrichment: an indicator of biodegradation at a petroleum hydrocarbon contaminated field site. Environ. Sci. Technol. 2002, 36, 2464-2470. Kray, W. C.; Castro, C. E. The cleavage of bonds by low-valent transition metal ions. The homogeneous dehalogenation of vicinal dihalides by chromous sulfate. J. Am. Chem. Soc. 1964, 4603-4608. Cwiertny, D. M. Mechanistic Investigations of Granular Iron and Iron- Based Bimetallic Reductants for Organohalide Pollutant Treatment. Ph.D. Thesis, Johns Hopkins University, Baltimore, MD, 2005.

Received for review December 21, 2006. Revised manuscript received February 28, 2007. Accepted March 1, 2007. ES063040X

VOL. 41, NO. 11, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4117