Carbon and Chlorine Isotope Effects During Abiotic Reductive

Hole Oceanographic Institution, Woods Hole, Massachusetts. 02543, Department of Earth and Environmental Sciences,. University of Illinois at Chicago, ...
0 downloads 0 Views 242KB Size
Download PDF
Environ. Sci. Technol. 2007, 41, 4662-4668

Carbon and Chlorine Isotope Effects During Abiotic Reductive Dechlorination of Polychlorinated Ethanes T H O M A S B . H O F S T E T T E R , * ,†,# CHRISTOPHER M. REDDY,† LINNEA J. HERATY,‡ MICHAEL BERG,§ AND NEIL C. STURCHIO‡ Department of Marine Chemistry and Geochemistry, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543, Department of Earth and Environmental Sciences, University of Illinois at Chicago, Chicago, Illinois 60607, and Eawag, Swiss Federal Institute of Aquatic Science and Technology, 8600 Dubendorf, Switzerland

We investigated the extent and variability of C and Cl isotope fractionation during the reduction of polychlorinated ethanes to evaluate the potential use of Cl isotope analysis for the assessment of contaminant transformation in subsurface environments. Kinetic isotope effects (KIE) for C and Cl for the reductive β-elimination of 1,1,2,2tetrachloroethane (1,1,2,2-TeCA), pentachloroethane (PCA), and hexachloroethane by Cr(II) used as model reductant in homogeneous solution were compared to KIEs measured for dehydrochlorination of 1,1,2,2-TeCA and PCA. Since isotopic reactions of polychlorinated compounds are complicated by the simultaneous presence of several Cl isotopologues and intramolecular isotopic competition, we present a procedure for the determination of KIEs for Cl from the initial reactant and final product Cl isotope ratios. Despite different reaction mechanisms, that is reduction via dissociative inner-sphere electron transfer by Cr(H2O)62+ and base-catalyzed, concerted elimination, respectively, apparent KIEs for C of both pathways fall within a similar range (1.021-1.031). In contrast, KIEs for Cl are significantly higher for reductive β-elimination (1.013-1.021) than for dehydrochlorination (1.000-1.006). These results suggest that reductive transformations of polychlorinated contaminants might be identified on the basis of combined C and Cl isotope analysis.

Introduction Compound-specific analysis of stable carbon isotopes has become an increasingly important tool in assessing polychlorinated hydrocarbons in contaminated environments (1-5). Since these compounds constitute a class of widespread toxic and carcinogenic soil and groundwater con* Corresponding author phone: +41 44 632 83 28; fax: +41 44 633 11 22; e-mail: [email protected]. † Woods Hole Oceanographic Institution. ‡ University of Illinois at Chicago. § Eawag, Swiss Federal Institute of Aquatic Science and Technology, 8600 Dubendorf, Switzerland. # Current address: Institute of Biogeochemistry and Pollutant Dynamics, ETH Zurich, 8092 Zurich, Switzerland. 4662

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 13, 2007

taminants, knowledge of the extent and product distribution of their reductive transformation is essential for a risk assessment of contaminated water resources (6). Owing to a characteristic normal C kinetic isotope effect (KIEC), that is, bonds to 12C are broken faster than those to 13C, polychlorinated hydrocarbons become enriched with heavy isotope during degradation (7). The extent of 13C enrichment of a contaminant can be linked to an “apparent” kinetic isotope effect, AKIEC, at the reacting bond if effects such as isotopic dilution or intramolecular isotopic competition are accounted for (eq 1, ref 5).

AKIEC )

1 1 + λ‚C/1000

(1)

where C is the bulk isotopic enrichment factor and λ stands for correction factors for the number of reactive sites within a molecule, intramolecular competition etc. (see ref 5 for details). Determination of this enrichment behavior allows for the qualitative and quantitative assessment of contaminant transformation pathways in complex environments. Reliable C-values are, therefore, crucial for correctly inferring the extent of contaminant degradation from its isotopic composition. However, C-values can depend on the prevailing environmental conditions (e.g., since intrinsic KIEs might be masked by preceding, nonfractionating steps such as the transport of a contaminant to a reactive site in an enzyme (8)); furthermore, C-values may even vary within a suite of related compounds undergoing transformation through a common pathway (1, 5). Reported C-values for reductive dechlorination vary over a considerable range corresponding to AKIECs between 1.005 and 1.036 (5), complicating quantification of the extent of pollutant degradation (2, 3). Moreover, this variability might introduce uncertainty with regard to attempts to infer contributions of different mechanisms of abiotic and enzymatic reductive dechlorination processes (9, 10). To address the mechanistic origins of varying isotopic enrichment factors more rigorously, simultaneous investigations of multiple isotopic elements is necessary and this approach has been applied successfully for H and C isotopes (11). For polychlorinated contaminants, the comparison of C and Cl isotope signatures has been used predominantly for source identification in environmental forensics (12, 13). Therefore, an aim of this study was to provide a combined evaluation of C and Cl isotope effects for reductive transformation of polychlorinated organic contaminants. The apparent scarcity of AKIE-values for Cl isotope fractionation of polychlorinated contaminants originates from the lack of analytical methods coupling the chromatographic separation of the analytes and their online conversion to CH3Cl, which is finally analyzed in the isotope-ratio mass spectrometer (14). Exceptions are reactions where the bulk organic Cl measured in a sample corresponds to Cl in the investigated organic reactant (e.g., aerobic microbial transformation of dichloromethane (15)). Moreover, interpretation of isotope effects in compounds labeled at natural abundance is more complex in the case of Cl than in the case of C, owing to the substantially greater abundance of 37Cl (24% of total Cl) relative to 13C (1.1% of total C) (7). This increases the probability that a polychlorinated compound under investigation contains more than one atom of the heavy element and thus more than two Cl isotopologues. Thus, observed Cl isotope fractionation originates from different isotopic reaction rates for many isotopologues, and isotopic enrichment factors cannot be assessed easily from the enrichment 10.1021/es0704028 CCC: $37.00

 2007 American Chemical Society Published on Web 05/24/2007

SCHEME 1

behavior of the substrate. Calculation of Cl isotope effects from bulk δ37Cl-data of polychlorinated compounds are only feasible if both rates and products of reductive dechlorination reactions are known (e.g., enzymatic transformation of tetraand trichloroethenes in laboratory systems (16)). This situation contrasts the detailed understanding of C and Cl isotope effects reported predominantly for elimination and nucleophilic substitution reactions of monochlorinated compounds, where C-Cl bond-specific isotopic information has been obtained from measurements of chlorine kinetic isotope effects (KIECl) of the Cl leaving group (37Cl leaving group isotope effects (7, 17, 18)). Given information on an enzyme’s active site structure and computational support, C and Cl isotope effects can be interpreted in terms of transition state structures to understand mechanisms of elementary dechlorination reactions (18-21). Environmental reductions of many polychlorinated contaminants have been suggested to be initiated via X-philic pathways at the Cl atom (22) and it is unclear whether the corresponding KIECl are similar to 37Cl leaving group isotope effects. Moreover, recent analytical developments suggest that compound-specific Cl isotope analysis might become available for selected classes of polychlorinated contaminants (23-25). Therefore, we explored the magnitude of Cl isotope fractionation during the reductive transformation of polychlorinated contaminants with regard to its applicability in combination with C isotope analysis for the assessment of dechlorination pathways. In laboratory model systems, we investigated the extent of C and Cl isotope fractionation for reductive β-elimination (Scheme 1) of three polychlorinated ethanes, 1,1,2,2-tetrachloroethane (1,1,2,2-TeCA), pentachloroethane (PCA), and hexachloroethane (HCA) by Cr(II) in homogeneous solution. To compare our findings with the established literature on isotope effects in elimination reactions, we also performed reference experiments on dehydrochlorination of 1,1,2,2-TeCA and PCA (Scheme 1). Chromium(II) was chosen as model reductant since the mechanism by which it reacts with polychlorinated ethanes is well-known (10). Cr(H2O)62+ is a one-electron reductant and its reactions with chlorinated compounds take place by an inner-sphere mechanism with Cl atoms acting as bridging ligands for the electron transfer (26-28). Moreover, reduction of polychlorinated ethanes by Cr(II) was shown to yield stoichiometrically one organic reductive β-elimination product and Cl- (Scheme 1). These model reactions offer the opportunity to derive Cl isotope effects from bulk δ37Cl-data of the reaction product after complete transformation without need for compound-specific Cl isotope analysis. Since the polychlorinated ethanes have several equivalent C-Cl bonds, of which one or two are broken during dehydrochlorination and reductive β-elimination, respectively, the isotopic composition of the product after full conversion is not equal to the isotopic composition of the initial substrate. At this stage, the isotopic composition

of the organic product and of Cl- reflects the intrinsic isotope effect (18). In this paper, we develop a method to derive KIECl from full conversion experiments of polychlorinated ethanes that allows for the concurrent evaluation of C and Cl isotope fractionation of reductive dechlorination reactions. The specific objectives of this work were (i) to provide evidence for the extent and variability of C and Cl isotope fractionation upon reduction of polychlorinated ethanes, (ii) to discuss mechanistic insights from the comparison of C and Cl isotope effects for reductive β-elimination and dehydrochlorination, and (iii) to evaluate the potential use of Cl isotope analysis for the assessment of organic contaminant transformation in the environment.

Experimental Procedures and Data Evaluation Experimental Systems. A complete list of all chemicals used as well as a detailed description of the experimental systems and analytical methods can be found in the Supporting Information (SI). Briefly, reactors used for the investigation of base-catalyzed dehydrochlorination of pentachloroethane (PCA, pH 7.5 and 8.5) and 1,1,2,2-tetrachloroethane (1,1,2,2TeCA, pH 9.1) were set up as described in ref 29. Reductive β-elimination of hexachloroethane (HCA), PCA, and 1,1,2,2TeCA by Cr(II) was conducted at pH 2.2 and Cr(II) concentrations of 1.0 mM in oxygen-free solutions. All experiments were carried out in serum bottles sealed with Viton stoppers and aluminum crimp caps at 23 ( 1°C in the dark. Isotopic Analyses. The entire solution of each of the reactors used for dehydrochlorination experiments was used for 13C isotope analysis of PCA, 1,1,2,2-TeCA, and its reaction products. Compound-specific isotope analysis (CSIA) was performed on a GC-C-IRMS system coupled to a purge and trap concentrator (4). For the determination of δ13C-values of HCA, PCA, and 1,1,2,2-TeCA from experiments with Cr(II), the solution in the reactors was extracted at given time intervals with 2 mL of hexane after addition of 1 mL of 4 M NaCl solution to the aqueous solution. Extracts (1 µL) were injected via a cold on-column injector to the GC-C-IRMS system as described in the SI. All δ13C-values of the analytes are reported relative to Vienna PeeDee Belemnite (δ13CVPDB). Aqueous samples were analyzed for δ37Cl after complete conversion of HCA, PCA, or 1,1,2,2-TeCA to the reductive β-elimination or corresponding dehydrochlorination products, tetrachloroethene (PCE), trichloroethene (TCE), or dichloroethenes (DCEs), respectively. Analysis of δ37Cl from Cl- was based on its precipitation as AgCl and further transformation to CH3Cl as described in the SI. All δ37Clvalues are reported relative to Standard Mean Ocean Chloride (δ37ClSMOC). Calculation of AKIEC. Carbon isotope enrichment factors, C, of the reactants were evaluated according to eq 2 as described in ref 30,

ln(RC) ) C/1000‚ln(c) + ln(RC,0/c0)

(2)

RC ) RC,VPDB(δ13C/1000 + 1)

(3)

with

where RC equals the C isotope ratio of the reactant (13C/12C) during the reaction (initial value is RC,0) and of reference material (RC,VPDB), respectively, C is the C isotope enrichment factor, c stands for the concentration, and δ13C represents the C isotope signature of the chlorinated ethane defined as

δ13C )

(

13

C/12Csample

13

C/12CVPDB

)

- 1 ‚1000

VOL. 41, NO. 13, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

(4)

9

4663

From C-values derived for the bulk reactant molecules, bondspecific apparent kinetic isotope effects (AKIEC) can be calculated by adjusting λ in eq 1 for assumptions for the number of reactive sites in a polychlorinated ethane. Calculation of Chlorine Isotope Effects. Given that dehydrochlorination and reductive β-elimination stoichiometrically yield one defined chlorinated organic product (setting aside distinctions between cis- and trans-DCE) and Cl-, Cl isotope effects could be calculated from δ37Cl-values of reactants and products. The principles of this procedure are given below (details see SI). During dechlorination of polychlorinated ethanes, enrichment of the reactant with heavy 37Cl occurs if 35Cl elimination, proceeding with the rate constant k35, is faster than 37Cl elimination (with rate constant k37) owing to a normal kinetic isotope effect (i.e., KIECl ) k35/k37 > 1). Thus, 35Cl- is eliminated preferentially and after complete transformation, organic products are isotopically heavy compared to the reactant whereas Cl- is isotopically lighter. The extent of Cl isotope fractionation depends on the magnitude of KIECl, which itself is subject to intramolecular isotopic competition. Intramolecular isotopic competition arises from the fact that in polychlorinated ethanes several chemically equivalent C-Cl bonds are susceptible for dechlorination reactions but only one and two bonds are broken during dehydrochlorination and reductive β-elimination, respectively. Thus, the probability of encountering Cl isotope fractionation depends on the number of equivalent C-Cl bonds and the distribution of Cl isotopologues in the reactant population. For example, elimination of H35Cl during dehydrochlorination of an 1,1,2,2TeCA isotopologue containing few 35Cl atoms will lead to an isotopically more enriched TCE exhibiting a higher δ37Cl compared to a reaction of an isotopologue containing more 35Cl. Therefore, calculation of δ37Cl for the final product (i) has to take into account the abundance of different reactant Cl isotopologues, and (ii) needs to distinguish between elimination of 35Cl vs 37Cl. The abundance of an isotopologue, Pi, of a polychlorinated ethane is given from a binomial distribution

Pi )

( ) nCl,i

37 nCl,i

‚πRnCl,i‚(1 - πR)nCl,i - nCl,i 37

37

(5) 37 nCl,i

where nCl,i is the total number of Cl atoms and is the number of 37Cl-atoms per isotopologue i. πR is the relative abundance of 37Cl, which is derived from the reactants initial isotope signature, δ37ClR0, and the isotope ratio of standard material, 37Rstd.

πR )

(δ37ClR0/1000 + 1)37Rstd (δ37ClR0/1000 + 1)37Rstd + 1

(6)

Owing to large 37Cl abundance, polychlorinated ethanes always consist of a mixture of (nCl,i + 1) isotopologues which all need to be taken into account when evaluating fractionation data for Cl isotope effects. Effects of intramolecular competition, for example upon dehydrochlorination of H35Cl vs H37Cl, are described in isotopic branching ratios, θji, which quantify the share of potential C-35Cl vs C-37Cl bond cleavages. They have to be defined for 35Cl (j ) 35) vs 37Cl elimination (j ) 37) as shown for dehydrochlorination reactions in eqs 7 and 8 per isotopologue i.

θj)35 ) i

)

4664

9

37 (nCl,i - nCl,i )‚k35 37 37 (nCl,i - nCl,i )‚k35 + nCl,i ‚k37 37 (nCl,i - nCl,i )‚KIECl 37 37 (nCl,i - nCl,i )‚KIECl + nCl,i

(7)

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 13, 2007

) θj)37 i )

37 nCl,i ‚k37 37 nCl,i)‚k35 + 37 nCl,i 37 nCl,i )‚KIECl

(nCl,i (nCl,i -

37 nCl,i ‚k37

37 + nCl,i

(8)

On the basis of the isotopologue abundance, Pi, isotopic branching ratios, θji, and the number of 35Cl and 37Cl atoms j in the product, nP,i , the isotopic yields per isotopologue, jyi,j, and thus the concentrations of 35Cl and 37Cl atoms in the products, 35ClP and 37ClP, can be calculated as follows: 35

ClP )

∑

35

yi,35 +

i)1-n;

)

∑

∑

ClP )

∑

35 Pi‚θ37 i ‚nP,i

(9)

i)1-n;

∑

37

i)1-n;

)

yi,37

i)1-n; 35 Pi‚θ35 i ‚nP,i +

i)1-n; 37

35

yi,35 +

∑

37

i)1-n; 37 Pi‚θ35 ‚n i P,i +

∑

i)1-n;

yi,37

∑

37 Pi‚θ37 i ‚nP,i

(10)

i)1-n;

These numbers finally allow one to calculate the Cl isotope signatures of the product.

(

37

δ37ClP )

ClP/35ClP 37

Rstd

)

- 1 ‚1000

(11)

Agreement between calculated and measured δ37ClP can be obtained by fitting KIECl in eqs 7 and 8 for every isotopologue using spreadsheet calculation software. Note that upon reductive β-elimination, isotopic branching ratios have to account for the successive isotopic removal of two Cl atoms, thus leading to the formulation of two KIECls in θji1/j2 (where j1 and j2 stand for the first and second Cl atom eliminated from the substrate, respectively; see equations in the SI and Tables S3-S5).

Results and Discussion Extent of Carbon Isotope Fractionation in Dechlorination Reactions. Reduction of all investigated polychlorinated ethanes by Cr(II) proceeded to completion within minutes to hours and stoichiometrically yielded reductive β-elimination products. Second-order reaction rate constants, k, given in Table 1 fall within the range of values reported for other polychlorinated organic compounds (10). As is illustrated for the reduction of pentachloroethane (PCA) to trichloroethene (TCE) in Figure 1a, the reaction followed pseudo-first order behavior and was accompanied by a significant and continuous 13C enrichment of PCA (Figure 1b) described by eq 12 (7).

1-F)

(

δ13CR + 1000

δ13CR0 + 1000

)

C/1000

(12)

where F corresponds to the fractional amount of reactant conversion, δ13CR0 is the reactant’s initial δ13C-signature, δ13CR is its value during the reaction, and C represents the bulk C isotope enrichment factor of PCA reduction by Cr(II). A C-value of -14.7‰ was obtained from the linearized form of eq 12 (Figure 1c). Additionally, Figure 1b shows that the δ13C-values of the final product, δ13CP of trichloroethene (TCE), can be described well based on the C-value derived from the PCA reduction and the initial PCA signature (δ13CR0, eq 13) thus confirming that the observed isotope fractionation is pertinent for reductive β-elimination of PCA to TCE and Cl-.

a Determined at [Cr(II) -4 and 0.01 M for reductive dechlorination experiments. b Bulk C isotope enrichment factor calculated according to eq 2. c determined with eq 1, see text for assumptions total] between 5·10 on λ. d Calculations assume stepwise reduction of polychlorinated ethanes to a radical, and that dehydrochlorination is concerted; see SI for calculation details. e No distinction of cis/trans isomers. f n.d.) not determined. g Buffers: pH 8.5-9.1 TAPS, pH 7.5 MOPS.

1.0064 ( 0.0004 0.9969 ( 0.0014 0.9991 ( 0.0032 n.d. n.d. n.d. 1.64 ( 0.11 2.85 ( 0.23 3.21 ( 0.54 0.06 ( 0.01 3.37 ( 0.14 3.37 ( 0.14 1.0277 ( 0.0007 1.0279 ( 0.0005 1.0314 ( 0.0003 -27.1 ( 0.7 -27.9 ( 0.5 -30.5 ( 0.4 9.08 ( 0.01g 7.52 ( 0.01 8.49 ( 0.02 Dehydrochlorination 1,1,2,2-TeCA TCE PCA PCE PCA PCE

(3.8 ( 0.1) × 10-1 (3.2 ( 0.1) × 101 (3.2 ( 0.1) × 101

1.0183 ( 0.0013 1.0207 ( 0.0017 1.0125 ( 0.0023 n.d.f -9.04 ( 0.07 n.d. 9.15 ( 0.35 11.37 ( 0.14 5.10 ( 0.62 0.06 ( 0.01 3.37 ( 0.14 0.95 ( 0.61 1.0261 ( 0.0012 1.0303 ( 0.0006 1.0212 ( 0.0005 -12.7 ( 1.2 -14.7 ( 0.6 -10.4 ( 0.5 (9.5 ( 0.6) × 10-4 (8.7 ( 0.3) × 10-1 2.5 ( 0.2 2.18 ( 0.02 2.18 ( 0.02 2.18 ( 0.02 Reductive β-Elimination 1,1,2,2-TeCA DCEe PCA TCE HCA PCE

(‰) (‰) (‰) (‰) s-1) (M-1

(-)

reactant

product

Cl-

(-)

KIECld δ37ClSMOC AKIECc ECb ka pH product reactant

TABLE 1. Second-Order Rate Constants, k, for Reductive β-Elimination and Dehydrochlorination of 1,1,2,2-tetrachloroethane (1,1,2,2-TeCA), Pentachloroethane (PCA), and Hexachloroethane (HCA). Isotopic Information Includes C Isotope Enrichment Factors (EC) and Corresponding Apparent C Kinetic Isotope Effects AKIEC, δ37Cl Signatures of Reactants and Products and Calculated Cl Isotope Effects KIECls

δ13CP )

1 - (1 - F)C/1000+1 13 ‚(δ CR0 + 1000) - 1000 F (13)

The identical qualitative behavior was found for C isotope fractionation during the reduction of hexachloroethane (HCA) and 1,1,2,2-tetrachloroethane (1,1,2,2-TeCA) by Cr(II). The C-values for the three compounds range from -10.4‰ for HCA to -14.7‰ for PCA and were independent of reaction rates (Table 1). Since one-electron reduction of polychlorinated ethanes by Cr(II) yields radical species that are more reactive than the parent compound one can assume the first dechlorination step to be rate-determining (see below). Neglecting any influence of the Cβ-atom on C isotope fractionation and thus correcting for isotopic dilution by the second C atom (λ ) 2 in eq 1), one can assign apparent C kinetic isotope effects, AKIECs, between 1.0212 ( 0.0005 and 1.0303 ( 0.0006 ((1σ) to the reductive β-elimination of HCA, PCA, and 1,1,2,2-TeCA (Table 1). To obtain additional evidence for the extent of C isotope fractionation during C-Cl bond cleavage, we determined C-values for the base-catalyzed dehydrochlorination of 1,1,2,2-TeCA and PCA. Rate constants of these reactions (Table 1) were in good agreement with published values (6) and an example is given for PCA transformation to tetrachloroethene (perchloroethylene, PCE) in Figure 1d. Carbon isotope fractionation of reactant and polychlorinated organic product could be described quantitatively using eqs 12 and 13, and the results are shown in Figure 1e and f. In contrast to reductive β-elimination, dehydrochlorination resulted in much stronger C isotope fractionation and bulk enrichment factors, C, were approximately twice the values observed for reductive β-elimination (Figure 1 and Table 1). Again, measured δ13CP-values of PCE agree perfectly with the evolution of δ13CP calculated with the C-value that was derived from PCA enrichment. This behavior points out that the observed C-value is pertinent for the dehydrochlorination reaction. Whereas Cs for the elimination of Cl- from 1,1,2,2TeCA- and PCA dehydrochlorination at pH 9.1 and 7.5 are almost identical within experimental error (-27.1 ( 0.7‰ and -27.9 ( 0.5, Table 1), the reaction of PCA at pH 8.5 resulted in a slightly more negative C-value (-30.5 ( 0.4‰). It is unclear whether either pH or organic buffer caused the different Cs of PCA. Dehydrochlorination of polychlorinated ethanes such as PCA has been proposed to occur in a concerted fashion, which involves at least partially concurrent proton loss at the Cβ and Cl- elimination at the CR-atom (29). It seems, therefore, reasonable to derive AKIECs from observed C-values of dehydrochlorination of PCA and 1,1,2,2-TeCA without correction for isotopic dilution in the order of 1.03 (Table 1). Chlorine Isotope Fractionation. The reductive β-elimination of PCA to TCE was taken as a model reaction to verify the applicability of the Cl isotope effect calculation on the basis of the initial δ37Cl-signatures of reactants and the δ37Cl-value of polychlorinated organic products after full conversion. As is shown in Table 1, the reaction led to substantial Cl isotope fractionation, that is, TCE was +8.0‰ enriched in 37Cl compared to PCA whereas Cl- was 12.4‰ depleted in 37Cl relative to PCA. Assuming a binomial distribution of Cl isotopologues and neglecting small contributions from secondary Cl isotope effects (18), these data offer different avenues to calculate KIECl of this reaction. First, based on the final product’s δ37ClP-value of TCE using eqs S4-S27 and isotopic branching ratios, θji1/j2 (Table S4), a KIECl of 1.0207 ( 0.0017 was obtained. The identical calculation procedure can be applied using the δ37ClP-value of Cl-, which results in a KIECl of 1.0198 ( 0.0005, and which is identical to the isotope effect derived from δ37ClP of TCE within experimental error. Alternatively, an estimate of KIECl VOL. 41, NO. 13, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4665

FIGURE 1. Reductive β-elimination of pentachloroethane (PCA) to trichloroethene (TCE) by Cr(II) in homogeneous solution at pH 2.2 (panels a-c) and base-catalyzed dehydrochlorination of PCA to tetrachloroethene (perchloroethene, PCE) at pH 8.5 (panels d-f). (a) time courses of concentrations for PCA reduction to TCE (modeled lines represent pseudo-first order fits to the data), (b) measured δ13C-values of PCA and TCE and calculated C isotope enrichment of reactant and product (eqs 12 and 13), (c) linearized 13C enrichment of PCA according to eq 2, (d) time courses of concentrations for PCA dehydrochlorination to PCE (modeled lines represent pseudo-first order fits to the data), (e) measured δ13C-values of PCA and PCE and calculated C isotope enrichment of reactant and product (eqs 12 and 13), (f) linearized 13C enrichment of PCA according to eq 2. can be found from intramolecular isotopic competition at any stage of conversion using δ37ClP of the isotopically enriched and depleted products (7). In this case, KIECl equals the 37Cl isotope ratio of the products (37RTCE/37RCl-), which results in an isotope effect of 1.0196 ( 0.0002. The good agreement of Cl isotope effects calculated from different experimental data confirms the observation of a moderately large KIECl for the reductive β-elimination of PCA. For the other polychlorinated ethanes and reactions studied, all KIECl were assessed from the δ37ClP of the polychlorinated organic product. The magnitude of KIECls for reductive β-elimination of HCA, PCA, and 1,1,2,2-TeCA follow the same trend as observed for AKIEC and decrease in the order PCA > 1,1,2,2TeCA > HCA. In contrast, KIECls found for dehydrochlorination reactions were significantly lower even though they can be as large as 1.0125 for alkoxide-promoted dehydrochlorinations of ethanes (31). Such leaving group KIECls in elimination reactions are typically smaller, approximately between 1.002 and 1.008 (7, 19). Our data for the dehydrochlorination of 1,1,2,2-TeCA agree perfectly with these observations. In contrast, the KIECl for PCA dehydrochlorination was slightly inverse and close to unity within experimental error. We do not have a mechanistic explanation for this small KIECl yet. Note, however, that the KIECl-calculation for PCA dehydrochlorination strongly depends on assumptions for δ37Cl-values of the two nonreactive Cl atoms at the Cβ-atom. A shift of their δ37Cl by -1‰ can lead to an increase of Cl isotope effect of 0.003 (Figure S1). In the absence of position-specific information on Cl isotope distribution within the reactant molecules, however, this explanation for the low KIECl is speculative given that the value for 1,1,2,2-TeCA agrees well with the literature. 4666

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 13, 2007

Mechanistic Considerations. To explain the finding that reductive β-elimination and dehydrochlorination yield similar AKIEC of C-Cl bond cleavages but significantly different KIECl, it is necessary to discuss the underlying reaction mechanisms. Variability of isotope effects in dehydrohalogenation reactions is determined predominantly by its mechanisms (concerted vs nonconcerted reaction), the type and interaction of the base with the abstracted proton and/ or leaving group as well as the extent of bond cleavage in the transition state (7, 19, 21, 32). The dehydrochlorination of PCA has been proposed to react via a concerted mechanism (E2) through a transition state with some degree of E1cBcharacter (29). This interpretation is consistent with the low 37Cl leaving group isotope effects observed here and also suggests that the extent of C isotope fractionation does not originate exclusively from bonding changes at the CR during C-Cl cleavage. Experimental studies on elimination reactions using 14CR vs 14Cβ-labeled phenyl-substituted ethanes (33, 34) as well as analogous computational studies on 13C fractionation (19) indeed point out that in E2 reactions different C isotope effects arise at CR and Cβ-atom, respectively. For example, the NaOEt-EtOH-promoted elimination of para-substituted (2-phenylethyl)trimethylammonium showed small 15N leaving group isotope effects together with lower 14CR fractionation for an early transition state and if the E2-character of the reaction was promoted with increasingly electron-withdrawing β-phenyl substituents (33). It is likely that a similar interpretation applies for the dehydrochlorination of PCA shown here and, therefore, the AKIEC of 1.03 reflects the weighted average of small isotope effects at the CR and a larger one at the Cβ-atom. Furthermore, if C isotope fractionation were fully allocated to the CR-atom, the corresponding AKIEC would be in the order of 1.06. This

SCHEME 2

value is more indicative for a substitution reaction than for the elimination pathway (35) and is also inconsistent with the small KIECl. Since the KIECl for 1,1,2,2-TeCA dehydrochlorination is substantially higher compared to that of PCA, we conclude that this reaction exhibits distinct C bonding changes in the transition state, which lead to a stronger fractionation at the CR atom but a similar overall AKIEC for the entire molecule. A different interpretation applies to results from reductive β-elimination experiments. The disappearance rate of the polychlorinated ethanes upon reaction with Cr(II) is determined by the formation of polychlorinated radical species after elimination of Cr(III)Cl2+ from an alkylchromium intermediate (Scheme 2a), the latter being characteristic for the reaction of Cr(II) with alkyl halides (26-28, 36). In another addition-elimination sequence (Scheme 2b1/2), addition of a second Cr(II) ion to the radical occurs at a much higher rate than the initial C-Cl cleavage (10, 37) leading to reductive β-elimination products. According to mechanistic information on reaction kinetics, it is unlikely that bonding changes occurred at the Cβ-atom prior to the rate-limiting step of the reaction. Therefore, one can attribute the observed C isotope fractionation entirely to the CR, as was done for the calculation of AKIEC in Table 1. It has been proposed that neighboring halogen atoms, that is halogens bound to the Cβ, participate in the reactions of alkyl halides with Cr(II) and consequently, Cl isotope fractionation presumably originates from concurrent bonding changes at multiple Cl atoms (28, 36). Evidence for neighboring group participation arises from increasing reaction rates in the presence of β-halogens (Scheme 2a) and from the stereochemistry of reductive elimination, which in the case of β-Cl promotes the formation of Cr(III)Cl2+ complexes from a cis configuration (Scheme 2b1) rather than trans elimination of Cr3+aq and Cl-aq (IIb) (26, 27). Hence, we assumed that Cl isotope fractionation during reductive β-elimination of 2 Cl atoms from HCA, PCA, and 1,1,2,2TeCA is not likely to be caused by an initial KIECl for an electron transfer upon first C-Cl cleavage and a subsequent 37Cl leaving group isotope effect for the second Cl removal. Rather, the observed KIECl represents an accumulation of sequential, isotope sensitive steps within the overall dechlorination. As a consequence, the KIECls reported here for the reduction by Cr(II) are the weighted average of isotope effects for the elimination of 2 Cl atoms, which were obtained by treating the two potential KIECl in the isotopic branching ratios (eqs S21-S23, SI) identically. Note that this data treatment led to the lowest possible KIECl that can be used to quantify the observed Cl isotope fractionation between reactant and products. Assigning the low leaving group KIECl to the second Cl removal during reductive β-elimination requires the first KIECl to be as high 1.04, which does not seem realistic for the cleavage of C-Cl bonds in the light of the available experimental and computational data. Instead,

the still rather large KIECl between 1.013 and 1.020 observed in these experiments are presumably a consequence of the Cr(II)-Cl interactions. It has been proposed that Cr(II) reduces alkyl halides in an inner-sphere electron-transfer process, in which Cl acts as bridging ligand between the metal and the polychlorinated ethane ((26, 36), Scheme 2a). We speculate that this ligand-transfer mechanism, in which Cl is part of the coordination sphere of the metal, gives rise to higher Cl isotope effects. Whereas bonding changes associated with leaving group isotope effects arise from the cleavage of a C-Cl bond and formation of new bonds of Clto coordinating the water (7), the Cl bridging ligand is likely to be bound to carbon and chromium in the transition state. Since isotope effects tend to be greater if the considered element is bound to heavier atoms, (partial) Cl-Cr bonding during reductive β-elimination might be responsible for the observed large Cl isotope fractionation. Additional effects such as early/late transition state and its geometry might also be relevant but an assessment of their contribution to the measured KIECl of HCA, PCA, and 1,1,2,2-TeCA is beyond the scope of this work. Environmental Significance. The results of our study illustrate that, despite different reaction mechanisms, reductive dechlorination and dehydrochlorination of polychlorinated ethanes show almost identical AKIEC but different Cl isotope effects. Cl isotope fractionation associated with environmental redox processes can be beyond the KIECllimits established for 37Cl leaving group isotope effects as supported by KIECls of similar magnitude found during microbial perchlorate (ClO4-) reduction. KIECls between 1.013 and 1.019 for sequential reduction steps of ClO4- to Cl- were also rationalized with the significant changes of bonding to Cl (18, 38). Even though mechanistic interpretations of KIEClvariability for the reduction of different (poly)chlorinated organic compounds is not feasible yet, these observations suggest that Cl isotope effects should be applicable to elucidate alternative pathways of their transformation at contaminated sites and improve the management of affected water resources. Competing dechlorination reactions of chlorinated ethanes and ethenes performed by microbes or at surfaces of zerovalent metals in reactive barriers might give rise to different products with sometimes equal or even greater (eco)toxicological concern than the parent compound (1, 10). Given the recent development of compound-specific Cl isotope analysis for chlorinated ethenes (23, 25), these products from hydrogenolysis, reductive R-, and/or β-elimination might be distinguished on the basis of concurrent analysis of C and Cl isotopes. Clearly, more work is required to make Cl isotope analysis accessible to a broader range of contaminants and to obtain insights into the origins of Cl isotope effects of additional dechlorination pathways.

Acknowledgments This work was funded by the Swiss National Science Foundation (PA002-104965) and US NSF (grant no. OCE0550486). We thank Jakov Bolotin for performing C isotope measurements, Werner Angst and Martin Elsner for valuable comments, as well as Rene´ Schwarzenbach and Michael Sander for reviewing the manuscript.

Supporting Information Available Description of analytical and experimental procedures, calculation algorithm for chlorine isotope effect, Figure on sensitivity of KIECl of PCA dehydrochlorination. This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (1) Hunkeler, D.; Aravena, R.; Cox, E. Carbon isotopes as a tool to evaluate the origin and fate of vinyl chloride: Laboratory VOL. 41, NO. 13, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4667

(2)

(3)

(4)

(5)

(6) (7) (8) (9) (10) (11)

(12) (13)

(14)

(15)

(16)

(17)

(18) (19) (20)

(21)

experiments and modeling of isotope evolution. Environ. Sci. Technol. 2002, 36, 3378-3384. Hunkeler, D.; Aravena, R.; Berry-Spark, K.; Cox, E. Assessment of degradation pathways in an aquifer with mixed chlorinated hydrocarbon contamination using stable isotope analysis. Environ. Sci. Technol. 2005, 39, 5975-5981. Morrill, P. L.; Sleep, B. E.; Slater, G. F.; Edwards, E. A.; Sherwood Lollar, B. Evaluation of isotopic enrichment factors for the biodegradation of chlorinated ethenes using a parameter estimation model: Toward an improved quantification of biodegradation. Environ. Sci. Technol. 2006, 40, 3886-3892. Zwank, L.; Elsner, M.; Aeberhard, A.; Schwarzenbach, R. P.; Haderlein, S. B. Carbon isotope fractionation in the reductive dehalogenation of carbon tetrachloride at iron (hydr)oxide and iron sulfide minerals. Environ. Sci. Technol. 2005, 39, 56345641. 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. Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M. Environmental Organic Chemistry, 2nd ed.; John Wiley & Sons: New York, 2003. Melander, L.; Saunders, W. H. Reaction Rates of Isotopic Molecules; John Wiley & Sons: New York, 1980. Enzyme Mechanism from Isotope Effects; Cook, P. F., Ed.; CRC Press: Boca Raton, FL, 1991. Vogel, T. M.; Criddle, C. S.; McCarty, P. L. Transformations of halogenated aliphatic compounds. Environ. Sci. Technol. 1987, 21, 722-735. Arnold, W. A.; Winget, P.; Cramer, C. J. Reductive dechlorination of 1,1,2,2-tetrachloroethane. Environ. Sci. Technol. 2002, 36, 3536-3541. 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. Shouakar-Stash, O.; Frape, S. K.; Drimmie, R. J. Stable hydrogen, carbon and chlorine isotope measurements of selected chlorinated organic solvents. J. Contam. Hydrol. 2003, 60, 211-228. Drenzek, N. J.; Tarr, C. H.; Eglinton, T. I.; Heraty, L. J.; Sturchio, N. C.; Shiner, V. J.; Reddy, C. M. Stable chlorine and carbon isotopic compositions of selected semi-volatile organochlorine compounds. Org. Geochem. 2002, 33, 437-444. Holt, B. D.; Sturchio, N. C.; Abrajano, T. A.; Heraty, L. J. Conversion of chlorinated volatile organic compounds to carbon dioxide and methyl chloride for isotopic analysis of carbon and chlorine. Anal. Chem. 1997, 69, 2727-2733. Heraty, L. J.; Fuller, M. E.; Huang, L.; Abrajano, T., Jr; Sturchio, N. C. Isotopic fractionation of carbon and chlorine by microbial degradation of dichloromethane. Org. Geochem. 1999, 30, 793799. Numata, M.; Nakamura, N.; Koshikawa, H.; Terashima, Y. Chlorine isotope fractionation during reductive dechlorination of chlorinated ethenes by anaerobic bacteria. Environ. Sci. Technol. 2002, 36, 4389-4394. Shiner, V. J.; Wilgis, F. P. Heavy atom isotope rate effects in solvolytic nucleophilic reactions at saturated carbon. In Isotopes in Organic Chemistry; Buncel, E.; Saunders, W. H., Eds.; Elsevier: Amsterdam, 1992; Vol. 8, pp 239-335. Paneth, P. Chlorine Kinetic Isotope Effects on Biological Systems. In Isotope Effects in Chemistry and Biology; Kohen, A.; Limbach, H., Eds.; CRC Press/Taylor & Francis: New York, 2006; p 1074. Sicinska, D.; Rostkowski, M.; Paneth, P. Chlorine isotope effects on chemical reactions. Curr. Org. Chem. 2005, 9, 75-88. Dybala-Defratyka, A.; Rostkowski, M.; Matsson, O.; Westaway, K. C.; Paneth, P. A new interpretation of chlorine leaving group kinetic isotope effects; A theoretical approach. J. Org. Chem. 2004, 69, 4900-4905. Jia, Z. S.; Rudzinski, J.; Paneth, P.; Thibblin, A. Borderline between E1cB and E2 mechanisms. Chlorine isotope effects in base-

4668

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 13, 2007

(22)

(23)

(24)

(25)

(26) (27) (28)

(29) (30)

(31)

(32) (33)

(34)

(35)

(36)

(37) (38)

promoted elimination reactions. J. Org. Chem. 2002, 67, 177181. Buschmann, J.; Angst, W.; Schwarzenbach, R. P. Iron porphyrin and cysteine mediated reduction of ten polyhalogenated methanes in homogeneous aqueous solution: Product analyses and mechanistic considerations. Environ. Sci. Technol. 1999, 33, 1015-1020. Shouakar-Stash, O.; Drimmie, R. J.; Zhang, M.; Frape, S. K. Compound-specific chlorine isotope ratios of TCE, PCE and DCE isomers by direct injection using CF-IRMS. Appl. Geochem. 2006, 21, 766-781. Holmstrand, H.; Mandalakis, M.; Zencak, Z.; Gustafsson, O.; Andersson, P. Chlorine isotope fractionation of a semi-volatile organochlorine compound during preparative megaborecolumn capillary gas chromatography. J. Chromatogr. A 2006, 1103, 133-138. Van Acker, M.; Shahar, A.; Young, E. D.; Coleman, M. L. GC/ multiple collector-ICPMS method for chlorine stable isotope analysis of chlorinated aliphatic hydrocarbons. Anal. Chem. 2006, 78, 4663-4667. Kochi, J. K.; Singleton, D. M. Stereochemistry of reductive elimination by chromium(II) complexes. J. Am. Chem. Soc. 1968, 90, 1582-1589. Singleton, D. M.; Kochi, J. K. The mechanism of reductive elimination of vic-dihalides by chromium(II). J. Am. Chem. Soc. 1967, 89, 6547-6555. Castro, C. E.; Kray, W. C. Cleavage of bonds by low valent transition metal ions. The homogeneous reduction of alkyl halides by chromous sulfate. J. Am. Chem. Soc. 1963, 85, 27682773. Roberts, A. L.; Gschwend, P. M. Mechanisms of pentachloroethane dehydrochlorination to tetrachloroethylene. Environ. Sci. Technol. 1991, 25, 76-86. Scott, K. M.; Lu, X.; Cavanaugh, C. M.; Liu, J. S. Optimal methods for estimating kinetic isotope effects from different forms of the Rayleigh distillation equation. Geochim. Cosmochim. Acta 2004, 68, 433-442. Koch, H. F.; McLennan, D. J.; Koch, J. G.; Tumas, W.; Dobson, B.; Koch, N. H. Use of kinetic isotope effects in mechanism studies 4. Chlorine isotope effects associated with alkoxidepromoted dehydrochlorination reactions. J. Am. Chem. Soc. 1983, 105, 1930-1937. Smith, M. B.; March, J. March’s Advanced Organic Chemistry, 5th ed.; John Wiley & Sons Inc.: New York, 2001. Eubanks, J. R. I.; Sims, L. B.; Fry, A. Carbon isotope effect studies of the mechanism of the hofmann elimination-reaction of parasubstituted (2-phenylethyl-1-14C)trimethylammonium and (2phenylethyl-2-14C)trimethylammonium bromides. J. Am. Chem. Soc. 1991, 113, 8821-8829. Wright, D. R.; Sims, L. B.; Fry, A. 14C Kinetic isotope effects and kinetic studies in the syn-elimination reactions of (2-phenylethyl)dimethylamine oxides. J. Am. Chem. Soc. 1983, 105, 37143716. Matsson, O.; Dybala-Defratyka, A.; Rostkowski, M.; Paneth, P.; Westaway, K. C. A theoretical investigation of R-carbon kinetic isotope effects and their relationship to the transition-state structure of SN2 reactions. J. Org. Chem. 2005, 70, 4022-4027. Kray, W. C.; Castro, C. E. Cleavage of bonds by low-valent transition metal ions - Homogeneous dehalogenation of vicinal dihalides by chromous sulfate. J. Am. Chem. Soc. 1964, 86, 46034608. Kochi, J. K.; Mocadlo, P. E. Reactions of organic peroxides with chromium(II). Reduction of free radicals by metal ions. J. Org. Chem. 1965, 30, 1134-1141. Sturchio, N. C.; Hatzinger, P. B.; Arkins, M. D.; Suh, C.; Heraty, L. J. Chlorine isotope fractionation during microbial reduction of perchlorate. Environ. Sci. Technol. 2003, 37, 3859-3863.

Received for review February 16, 2007. Revised manuscript received April 11, 2007. Accepted April 12, 2007. ES0704028