Pathways of chlorinated ethylene and chlorinated acetylene reaction with Zn(O)

Article abstract of DOI:10.1021/es980252o

To successfully design treatment systems relying on reactions of chlorocarbons with zero-valent metals, information is needed concerning the kinetics and pathways through which transformations occur. In this study, pathways of chlorinated ethylene reaction with Zn(O) have been elucidated through batch experiments. Data for parent compound disappearance and product appearance were fit to pseudo-first-order rate expressions in order to develop a complete kinetic model. Results indicate that reductive β- elimination plays an important role, accounting for 15% of tetrachloroethylene (PCE), 30% of trichloroethylene (TCE), 85% of cis- dichloroethylene (cis-DCE), and 95% of trans-dichloroethylene (trans-DCE) reaction. The fraction of PCE, TCE, trans-DCE, and cis-DCE transformation that occurs via reductive elimination increases as the two-electron reduction potential (E₂) for this reaction becomes more favorable relative to hydrogenolysis. In the case of PCE and TCE, reductive elimination gives rise to chlorinated acetylenes. Chloroacetylene and dichloroacetylene were synthesized and found to react rapidly with zinc, displaying products consistent' with both hydrogenolysis and reduction of the triple bond. Surface area-normalized rate constants (k(SA))for chlorinated ethylene disappearance correlate well with both one-electron (E₁) and two-electron (E₂) reduction potentials for the appropriate reactions. Correlation with E₂ allows prediction of the distribution of reaction products as well as the rate of disappearance of the parent compound. To successfully design treatment systems relying on reactions of chlorocarbons with zero-valent metals, information is needed concerning the kinetics and pathways through which transformations occur. In this study, pathways of chlorinated ethylene reaction with Zn(0) have been elucidated through batch experiments. Data for parent compound disappearance and product appearance were fit to pseudo-first-order rate expressions in order to develop a complete kinetic model. Results indicate that reductive β-elimination plays an important role, accounting for 15% of tetrachloroethylene (PCE), 30% of trichloroethylene (TCE), 85% of cis-dichloroethylene (cis-DCE), and 95% of trans-dichloroethylene (trans-DCE) reaction. The fraction of PCE, TCE, trans-DCE, and cis-DCE transformation that occurs via reductive elimination increases as the two-electron reduction potential (E₂) for this reaction becomes more favorable relative to hydrogenolysis. In the case of PCE and TCE, reductive elimination gives rise to chlorinated acetylenes. Chloroacetylene and dichloroacetylene were synthesized and found to react rapidly with zinc, displaying products consistent with both hydrogenolysis and reduction of the triple bond. Surface area-normalized rate constants (k_SA) for chlorinated ethylene disappearance correlate well with both one-electron (E₁) and two-electron (E₂) reduction potentials for the appropriate reactions. Correlation with E₂ allows prediction of the distribution of reaction products as well as the rate of disappearance of the parent compound.

Full text of DOI:10.1021/es980252o

Environ. Sci. Technol. 1998, 32, 3017-3025
treatment systems, information is needed concerning the
rates and pathways through which chlorinated intermediates
are formed and subsequently degrade.
Pathways of Chlorinated Ethylene
and Chlorinated Acetylene Reaction
with Zn(0)
Zinc was selected in the present study in order to gain
insight into pathways through which one possible candidate
metal reacts with chlorinated ethylenes. Zinc has previously
been studied for use as a reductant of carbon tetrachloride
(8, 9), as well as lindane (γ-hexachlorocyclohexane), 1,1,1-
trichloroethane, and chloroform (10). Methyl parathion has
also been found to undergo reduction when contaminated
soil is mixed with zinc metal (11). Although oxidation of
zinc metal might be anticipated to release Zn(II) species,
which might be of concern in drinking water, Carraway and
co-workers (12) have recently demonstrated that dissolved
Zn2+ concentrations generated in the reaction of PCE with
Zn(0) can be minimized by adding hydroxyapatite, onto
which the Zn(II) is presumably removed via adsorption (12-
14).
Our previous work (1, 15) has shown that chlorinated
ethylene reduction by zero-valent metals, such as iron and
zinc, proceeds via two parallel pathways: hydrogenolysis
and reductive â-elimination. In the case of ethylenes bearing
an R,â pair of chlorines, reductive elimination gives rise to
substituted acetylenes. A third reaction that may occur in
these systems is the reduction of multiple bonds. This creates
the possibility that a particular intermediate may arise from
multiple pathways. For example, vinyl chloride may be
produced via hydrogenolysis of the dichloroethylenes or
alternatively by reduction of the triple bond of chloroacety-
lene. The existence of such competing reactions greatly
complicates the modeling of transformation pathways. A
scheme that accounts for reactions of chlorinated ethylenes
and other possible intermediates with zinc is detailed in
Figure 1 (1, 15). Although our prior work established the
overall plausibility of this scheme, the data reported by us
(1, 15) and others (2-4, 16-18) have not been sufficiently
detailed as to determine the rate constants associated with
each of these reactions for either Zn(0) or Fe(0). The principal
objective of this work was to measure the kinetics of
chlorinated ethylene reaction with Zn(0) through detailed
measurements in batch systems.
W I L L I A M A . A R N O L D A N D
A . L YN N R O B E R T S *
Department of Geography and Environmental Engineering,
313 Ames Hall, The Johns Hopkins University, 3400 North
Charles Street, Baltimore, Maryland 21218
To successfully design treatment systems relying on
reactions of chlorocarbons with zero-valent metals,
information is needed concerning the kinetics and pathways
through which transformations occur. In this study,
pathways of chlorinated ethylene reaction with Zn(0) have
been elucidated through batch experiments. Data for
parent compound disappearance and product appearance
were fit to pseudo-first-order rate expressions in order
to develop a complete kinetic model. Results indicate that
reductive â-elimination plays an important role, accounting
for 15% of tetrachloroethylene (PCE), 30% of trichloro-
ethylene (TCE), 85% of cis-dichloroethylene (cis-DCE), and
95% of trans-dichloroethylene (trans-DCE) reaction. The
fraction of PCE, TCE, trans-DCE, and cis-DCE transformation
that occurs via reductive elimination increases as the two-
electron reduction potential (E ) for this reaction becomes
2
more favorable relative to hydrogenolysis. In the case of
PCE and TCE, reductive elimination gives rise to chlorinated
acetylenes. Chloroacetylene and dichloroacetylene were
synthesized and found to react rapidly with zinc, displaying
products consistent with both hydrogenolysis and
reduction of the triple bond. Surface area-normalized
rate constants (kSA) for chlorinated ethylene disappearance
correlate well with both one-electron (E ) and two-
1
electron (E ) reduction potentials for the appropriate
2
A second question relates to our ability to develop
quantitative structure-activity relationships (QSARs) through
which rates and products might be predicted for the reaction
of other organohalides with metals. Johnson et al. (19)
proposed that surface area-normalized rate constants (kSA
values) could be correlated with two-electron reduction
potentials for hydrogenolysis, even though many of the
species employed in their correlation could potentially react
to a significant extent via competing reductive â-elimination.
Without detailed knowledge of the product distributions in
the experiments compiled by ref 19, it has not been possible
to properly account for the contributions of such competing
reactions. Reduction potentials for hydrogenolysis and
reductive elimination can be quite different (see Table 1 and
ref 15); failure to correctly account for the pathways by which
different species react may influence the success of such
attempted correlations. In undertaking this study, we also
sought to assess whether QSARs could be improved through
separate correlations of the two types of reductive dehalo-
genation as well as through correlation of reaction rate
constants with one-electron reduction potentials.
reactions. Correlation with E allows prediction of the
distribution of reaction products as well as the rate of
disappearance of the parent compound.
2
Introduction
The use of zero-valent metals as reductants of chloroalkanes
and chloroethylenes represents a promising new approach
for treating groundwater contaminated with such solvents.
Removal of the parent compound, however, may not be
sufficient to attain drinking water standards as partially
dechlorinated reaction products or intermediates may also
pose human health risks. For example, some moderately
persistent species, such as cis-dichloroethylene (cis-DCE) and
vinyl chloride (VC), which have been detected as products
of tetrachloroethylene (PCE) (1) and trichloroethylene (TCE)
(1-4) reduction by iron, are themselves of concern as
contaminants. The necessity of controlling effluent levels of
such species may dictate the overall design of metal-based
treatment systems (5-7). To develop process models that
could be used to optimize the design of in situ or ex situ
Experimental Section
Chem icals. The following chemicals were used as received:
tetrachloroethylene (99.9%, HPLC grade, Aldrich), trichlo-
* Corresponding author phone: (410)516-4387; fax: (410)516-8996;
e-mail: lroberts@jhu.edu.
9
S0013-936X(98)00252-1 CCC: $15.00
Published on Web 08/27/1998
1998 Am erican Chem ical Society
VOL. 32, NO. 19, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 0 1 7

using the procedure described in ref 26. The cleaned metal
was used within 24 h, for it was found that the reactivity of
the metal steadily decreased by a factor of 3 over a period
of 3 weeks, even when stored in an anaerobic chamber. The
surface area was 0.035 m2/ g as measured via Kr adsorption
using a Micromeritics Flowsorb II 2300.
Experim ental System s. Reaction bottles were filled with
deoxygenated 0.1 M NaCl/ 50 mM Tris buffer within an
anaerobic chamber. Depending on the reactivity of the
compound being studied, varying quantities of metal were
added to the flask, with higher zinc loadings being used to
increase the reaction rate of less reactive compounds.
Reactions of PCE were studied using 5 g of zinc in 125-mL
(nominal volume) glass bottles with glass stopcock adapters.
The stopcock outlets were fitted with an NMR septum
(Aldrich) to facilitate sampling while excluding oxygen. The
glass stopcocks isolated the flask contents from the potentially
sorptive NMR septum except during the brief sampling
intervals. The bottles were mixed by rotating about their
longitudinal axes on a rotator (Cole-Parmer) at 56 rpm;
incubations were conducted at room temperature (23 ( 1
°C).
Reactions of all other species were carried out in 160-mL
serum bottles sealed with Teflon-faced butyl rubber septa
(Wheaton). Serum bottles were used in preference to the
125-mL bottles used for PCE because the pressure buildup
caused by hydrogen gas evolution at high metal loadings or
over long reaction times causes the glass stopcock adapters
on the 125-mL reactors to leak. For the remaining chlorinated
ethylenes, acetylene, and ethylene, 25 g of zinc was used,
and bottles were rotated at 36 rpm. Reaction bottles used
for chloroacetylene and dichloroacetylene contained 1 and
0.5 g of zinc, respectively, and were rotated at 56 rpm. The
chlorinated acetylenes and PCE were substantially more
reactive than any of the other species. Experiments con-
ducted with PCE suggested that a possible mass transfer
limitation existed at lower mixing speeds for this species;
experiments with the most highly reactive species were thus
conducted at a high mixing rate in an attempt to overcome
this problem. Reaction rate was independent of mixing speed
between 15 and 45 rpm for the less reactive species.
Chlorinated ethylenes were introduced to the reactors
via a methanolic spike (resulting in 0.003- 0.06% methanol
by weight) and the chlorinated acetylenes via the THF solution
in which they were synthesized (resulting in 0.03% THF by
weight). Spiking solutions of acetylene or ethylene were made
by equilibrating the appropriate gas/ helium mixture with
deoxygenated buffer in a 1-L syringe (Hamilton). The syringe
was then transferred to an anaerobic chamber where aliquots
of the equilibrated solution were introduced to the reaction
flasks to achieve the desired starting concentration. Reac-
tions were performed at various initial concentrations ranging
from 3 to 80 µM for acetylene, 3-30 µM for vinyl chloride,
and 15-350 µM for the dichloroethylenes, TCE, and PCE. At
specified intervals, 1-mL samples were withdrawn from the
reactors while simultaneously injecting 1 mL of deoxygenated
buffer. This made it possible to avoid introducing headspace
to the bottle during sampling, allowing anoxic conditions to
be maintained.
FIGURE 1. Hypothesized reaction sequence (1, 15) for reduction of
chlorinated ethylenes and related compounds. Reaction numbers
correspond to rate constants given in Table 1.
roethylene (99.5+%, Aldrich), cis-dichloroethylene (97+%,
Aldrich), trans-dichloroethylene (trans-DCE, Supelco), 1,1-
dichloroethylene (1,1-DCE, 99%, Aldrich), vinyl chloride in
methanol (Chem Service, 2000 µg/ mL), 10% ethylene in
helium (Scott Specialty gas), and 1% acetylene in helium
(Scott Specialty gas). Methanolic (HPLC grade, J. T. Baker)
spiking solutions were prepared for PCE, TCE, and the DCE
isomers.
Acid solutions for cleaning metals were prepared using
deoxygenated, deionized water (Milli-Q Plus UV, Millipore).
Solutions were deoxygenated by sparging (1 h/ L) with argon,
which was purified using an in-line molecular sieve and
oxygen traps. Reaction bottles were filled with deoxygenated
solutions containing 0.1 M NaCl (99%; J. T. Baker) and 50
mM Tris buffer [tris(hydroxymethyl)aminomethane, reagent
grade; Sigma], pH 7.2. The high (relative to typical ground-
water) NaCl concentration was employed to poise the
chloride content in the system, as chloride ion is released
during reductive dechlorination. Chloride is known to
accelerate corrosion by weakening the passive oxide film on
metals (20), and its presence has been shown to increase the
rate of carbon tetrachloride reduction by iron (21).
Synthesis of Chlorinated Acetylenes. Chloroacetylene
and dichloroacetylene were synthesized from cis-DCE and
TCE, respectively. Potassium hydride (35% in mineral oil,
Aldrich) was rinsed twice with pentane (J. T. Baker) and once
with freshly distilled dry tetrahydrofuran (THF). Dehydro-
halogenation of cis-DCE or TCE was induced by potassium
hydride and a catalytic amount of methanol in dry THF as
described by ref 22. Verification of the identities of the
synthesis products was conducted via gas chromatography/
mass spectrometry (GC/ MS). The resulting solutions of the
chlorinated acetylenes in THF were used as spiking solutions.
No purification of these reportedly explosive (22-24) gases
was attempted.
Analytical Methods. Aqueous samples were transferred
to 2.5-mL (nominal volume) autosampler vials. Headspace
samples were analyzed by gas chromatography (GC) with
flame ionization detection (FID) using the system described
in ref 26. The THF solutions containing the chlorinated
acetylenes were analyzed via cold on-column GC using an
Rtx-1 column (30 m × 0.32 mm i.d., 3 µm film thickness,
Restek) and an FID detector.
Calibration curves were generated using aqueous stan-
dards (0.1 M NaCl/ 50 mM Tris, pH 7.2) of PCE, TCE, and the
DCE isomers prepared in 20-mL glass syringes. These
Metal Preparation. The zinc metal (J. T. Baker, 30 mesh)
was cleaned with 0.4% H2SO4 to remove surface oxides (25)
9
3 0 1 8 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 19, 1998

TABLE 1. Surface Area-Normalized Rate Constants for Chlorinated Ethylenes and Their Reaction Productsa
reaction
kSA (L m-2 h-1 b
)
nc
E (V)d,e
E (V)d
1
2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
PCE f TCE
PCE f dichloroacetylene
TCE f trans-DCE
TCE f cis-DCE
TCE f 1,1-DCE
TCE f chloroacetylene
trans-DCE f VC
trans-DCE f acetylene
cis-DCE f VC
cis-DCE f acetylene
1,1-DCE f VC
VC f ethylene
acetylene f ethylene
chloroacetylene f acetylene
chloroacetylene f VC
dichloroacetylene f trans-DCE
dichloroacetylene f chloroacetylene
0.3((0.01)
9
9
5
5
5
5
2
2
2
2
3
3
5
1
1
4
4
-0.358
-0.358
-0.597
-0.578
-0.840
-0.578
-0.880
-0.880
-0.898
-0.898
-0.773
-0.954
0.582
0.631
0.534
0.543
0.523
0.609
0.488
0.573
0.478
0.564
0.498
0.375
4.5((0.7) × 10-2
1.5((0.1) × 10-3
6.0((1.0) × 10-4
6.8((9.1) × 10-5
9.4((1.2) × 10-4
6.4((8.2) × 10-7
1.3((0.1) × 10-5
5.2((1.5) × 10-7
3.0((0.2) × 10-6
4.1((0.3) × 10-5
1.0((0.5) × 10-4
5.0((1.4) × 10-4
6.5((0.7)
0.50((0.3)
20.6((0.9)
4.4((0.3)
a
b
c
Reaction num bers pertain to those illustrated in Figure 1. Uncertainties represent 95% confidence lim its. Num ber of replicate experim ents
d
used to determ ine rate constants. Calculations assum e the Henry’s law constant for a radical species is the sam e as that of the hydrogenated
parent (28), {H+} ) 10-7 M, {Cl-} ) 1 m M, and that all other species are present at unit activity. Gas-phase therm odynam ic data are from refs
29-31, corrected for air-water partitioning using Henry’s law constants recom m ended by refs 32 and 33. The sm all changes in the E1 and E2 values
e
as com pared to those previously reported (15) result from differences in the source of therm odynam ic data. E1 values are for the form ation of
the appropriate radical species. Reactions shown with identical one-electron reduction potentials produce the sam e radical interm ediates.
standards were analyzed using the same method as samples.
Acetylene, ethylene, and vinyl chloride were calibrated using
gas standards (Scott Specialty Gas). The gas standards were
mixed in a 2-mL syringe (with a wetted barrel) equipped
with a three-way stopcock, one end of which was fitted with
a septum. Gas samples (200 µL) were withdrawn and injected
manually. Aqueous concentrations were calculated using
the “dimensionless” Henry’s law constants [(mol/ Lair)/ (mol/
completely in ∼8 h) was stable at pH 7.3. The solution pH
did rise slightly during the longer incubations required for
other species, however, starting at pH 7.3 and ending at 7.6.
Kinetic Modeling. Pseudo-first-order rate constants were
calculated using the software package Scientist for Windows
(v. 2.01, Micromath, Salt Lake City, UT). The relevant system
of differential equations is given by
Lwater)] determined via a modified EPICS method (27) at 60
Nj
dCi
°C (the temperature of the autosampler oven) in the buffer
solution. The values obtained were 1.1 for acetylene, 1.1 for
vinyl chloride, and 8.7 for ethylene. Due to the lack of
commercial availability of the chlorinated acetylenes and
uncertainty as to the exact yield obtained in the syntheses,
calibrations for these species could not be conducted directly.
Bottles were initially filled without headspace. As much
as 6 mL of headspace evolved during the longer incubations,
presumably attributable to the reduction of protons by the
metal to form hydrogen gas. For most species observed in
these experiments, less than 4% of the mass would have
been lost via partitioning to the headspace at 25 °C. Ethylene,
however, has a substantially higher Henry’s law constant
than the other species, and losses of over 25% of the mass
of this species into the headspace could be incurred for the
reactions of long duration. Reported aqueous ethylene
concentrations, therefore, include the calculated fraction that
partitioned into the headspace. As ethylene is stable under
the conditions employed, its partitioning into the headspace
does not affect its subsequent behavior. Partitioning of all
other species into the headspace was minimal, so reported
aqueous concentrations were not corrected.
The identities of all reaction products detected were
confirmed by GC/ MS analysis. Samples were placed in 2.5-
mL autosampler vials and shaken by hand for 3 min. A 50-
200-µL headspace sample was then withdrawn and manually
injected (via splitless injector) on a Hewlett-Packard (HP)
5890 GC equipped with an HP 5970 mass spectrometer
detector and a GS-Q PLOT column (30 m × 0.32 mm i.d.;
J&W Scientific). The THF solutions containing the chlori-
nated acetylenes were analyzed on the same GC/ MS using
the Rtx-1 column previously described.
) -( kij)Ci
(1)
(2)
∑
dt
j)1
Ni
Nm
dCj
dt
) ( kij)Ci - (
kjm)Cj
(for all j)
∑
∑
i)1
m)1
where C represents aqueous concentrations and k values are
pseudo-first-order rate constants. Equation 1 pertains to
the reaction of the parent species, i, forming all possible
daughter products, j. In eq 2, the daughter product, j, may
be produced by multiple parents, i, and can itself react to
other species, m, according to rate constants kjm
.
Given experimental data, differential rate expressions, and
initial conditions, Scientist calculates the selected parameters
(in this case, kij and the initial concentration of the parent
compound) via a least-squares fit of the data to numerically
integrated solutions of the system of rate expressions. By
computing each kij value according to the distribution of
products, the model apportions the overall rate constant for
parent compound disappearance into the contributions for
the competing reactions. When calculating the kij values for
a given species, the rate constants (kjm) for reactions of
daughter products that were detected in a given experiment
were constrained to be equal to those determined in
independent experiments in which the daughter products
were introduced as the starting material.
Results
Pseudo-first-order rate constants for each reaction shown in
Figure 1 were divided by the metal surface area per liter of
solution to obtain kSA values. These kSA values are given in
Table 1 along with the corresponding one- and two-electron
reduction potentials.
The pH of the aqueous medium was measured before
and after the reaction, using a Ross combination electrode
(Orion). The pH during the reduction of PCE (which reacted
9
VOL. 32, NO. 19, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 0 1 9

Different initial concentrations of organohalide did not
have a significant effect on the observed rate constants; thus,
the values in Table 1 represent averages of the rate constants
determined at each initial concentration. Confidence limits
were calculated by propagating the errors associated with
the model-calculated uncertainties obtained from several
different timecourses. For the majority of the reactions listed
in Table 1, the rate constants could be determined with a fair
degree of precision ((15%, based on the 95% confidence
limits). Substantially greater uncertainty arose, however, for
some of the minor reactions that result in only trace product
accumulation (e.g., reactions 5, 7, and 9).
Preliminary experiments conducted at lower mixing
speeds with PCE had suggested a dependence of reaction
rate on initial concentration (34). This effect, however, was
not observed at the higher mixing rates employed in the
present study. Competitive inhibition effects were system-
atically investigated for the two most reactive species, PCE
and TCE. For example, 15 µM PCE was added to each of six
reaction bottles. One bottle contained no competitor, and
each of the others contained a 200 µM concentration of one
of the other chlorinated ethylenes (TCE; cis-, trans-, or 1,1-
DCE; VC). Similar experiments were performed with TCE.
No significant decrease in the rate of reaction of PCE or TCE
was observed under these conditions, indicating that com-
petitive effects, if present, were too small to detect.
Carbon mass balances at the end of the experiments were
>90% of the regressed initial parent compound concentra-
tion. No losses were observed in zinc-free blanks, suggesting
that sorption to the Teflon-faced septa and losses through
needle punctures were minimal. Further experimental
results and modeling details are discussed in the following
sections.
Effect of Metal Loading. Normalization of rate constants
by metal surface area assumes that a linear relationship exists
between observed rate constant and metal loading. Results
of some investigations with iron (35, 36) have, however,
suggested that reaction rate might not be directly proportional
to metal loading. Scherer and Tratnyek (35), for example,
found that the rate of carbon tetrachloride reaction reached
a plateau as iron loading was increased. A similar relationship
between TCE reduction rate and iron loading was observed
by Gotpagar et al. (36). Such a nonlinear dependence on
metal loading was not encountered in the present study;
rather, the pseudo-first-order rate constant for parent
compound disappearance was found to increase linearly with
metal loading (data not shown), giving r2 values of 0.983 for
PCE (at zinc loadings from 6 to 56 g/ L), 0.993 for TCE, and
0.994 for acetylene (the latter two over a range of 65-240 g/ L
of zinc).
Vinyl Chloride and Acetylene. Both of these species
reacted to form ethylene. No subsequent reduction of
ethylene to ethane was observed. Pseudo-first-order rate
constants were calculated by simultaneously fitting the
disappearance of the parent compound and the production
of ethylene. It should be noted that no acetylene was
observed in the reduction of vinyl chloride. From the
measured rates of disappearance of vinyl chloride and
acetylene, substantial accumulation of acetylene would be
anticipated if it were an intermediate in the reduction of
vinyl chloride to ethylene.
Dichloroethylenes. Reaction of cis- and trans-DCE with
Zn(0) produced acetylene, vinyl chloride, and ethylene. Only
the latter two products were observed from the reaction of
1,1-DCE. For the DCE isomers, reaction proceeded so slowly
(with a t1/ 2 >125 days) at a zinc loading of 25 g/ 160 mL that
accurate determination of the rate of disappearance of the
parent compound was not feasible. Rate constants reported
in Table 1 were therefore based on the rate of appearance
of the reaction products under pseudo-zero-order conditions,
FIGURE 2. Reduction of trans-DCE (3) by 25 g of Zn(0) in 160 mL
of 0.1 M NaCl/0.05 M Tris buffer (pH 7.2). Solid lines are model fits
to reaction products acetylene (O), ethylene (9), and vinyl chloride
()). The dashed line represents the mean concentration of trans-
DCE determined by averaging the data points shown.
assuming the concentration of the parent compound to be
invariant at very low conversions.
A representative plot of experimental results obtained for
trans-DCE is given in Figure 2. The solid lines are the model
fits with rate constants for vinyl chloride and acetylene
reaction constrained to equal values determined in inde-
pendent experiments. Due to the pressure buildup from
hydrogen gas evolution, venting of the reactors was necessary
for the last few samples. This action may have resulted in
a stripping of the highly volatile ethylene and a loss of some
of the mass contained in the headspace, accounting for the
slight overprediction by the model of the measured ethylene
concentrations at the end of the experiment.
TCE and PCE. Example time courses obtained for TCE
and PCE along with the corresponding model fits are shown
in Figures 3 and 4, respectively. The pseudo-first-order rate
constants were calculated by simultaneously fitting the
disappearance of the parent compound and the appearance
of the daughter species, constraining subsequent reactions
of daughter species as previously described. Reaction of TCE
with zinc gave rise to the three dichloroethylene isomers,
vinyl chloride, acetylene, and ethylene. It is interesting to
note that even though cis-DCE is the thermodynamically
favored species (Table 1) and is often the preferred product
of microbial TCE reduction (37), trans-DCE was the pre-
dominant DCE isomer formed from TCE reduction by zinc.
Modeling the results of TCE reduction (Figure 3) again results
in overprediction of the ethylene concentration, most likely
due to the venting that occurred toward the end of the
experiment (to which we attribute the 4% loss in mass balance
observed for this experiment). The predicted concentrations
of vinyl chloride are also higher than the observed values.
This is a reflection of the high degree of uncertainty associated
with the rate constant for formation of vinyl chloride from
chloroacetylene (reaction 15; see below).
The major products of PCE reduction are TCE, trans-
DCE, and acetylene (Figure 4). Trace amounts of cis-DCE,
vinyl chloride, and chloroacetylene (not shown) were also
observed. Experiments with PCE were relatively short in
duration; as a result, no headspace evolved.
The acetylene observed in the experiments with both PCE
and TCE appeared too rapidly to originate from the reductive
elimination of the cis- or trans-DCE observed as daughter
products in these systems. Attempts to model the data via
a sequence such as PCE f TCE f DCEs f acetylene proved
unsuccessful, as previously demonstrated by Roberts et al.
9
3 0 2 0 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 19, 1998

FIGURE 5. Reduction of chloroacetylene in 0.1 M NaCl/0.05 M Tris
buffer (pH 7.2) by 1 g of Zn(0). Solid symbols (b) represent
chloroacetylene peak areas, while open symbols for acetylene (O)
and vinyl chloride ()) reflect measured concentrations. The model
fit for the chloroacetylene data represents model-calculated
concentrations.
FIGURE 3. Reduction of TCE (4) in 160 mL of 0.1 M NaCl/0.05 M Tris
buffer (pH 7.2) by 25 g of Zn(0). Major products are trans-DCE (3),
acetylene (O), and cis-DCE (2). Inset shows trace products 1,1-DCE
(1), ethylene (9), and vinyl chloride ()). The dotted line represents
dichloroethylenes was provided by computing product yields
as a function of time. For example, when TCE is the starting
material, cis- and trans-DCE (as well as 1,1-DCE) are formed.
If the observed acetylene and vinyl chloride were secondary
products derived from these DCE isomers, their yields relative
to the DCEs would increase over time. During the reduction
of TCE (Figure 3), however, the yields of acetylene and vinyl
chloride relative to cis- and trans-DCE were constant. The
yield of vinyl chloride was also constant relative to 1,1-DCE.
This suggests that the vinyl chloride and acetylene observed
during TCE reduction are produced via a highly reactive
intermediate through a pathway that operates concurrently
to hydrogenolysis of TCE to DCEs (and their subsequent
reactions). Yields of acetylene and trans-DCE (relative to
TCE) were also constant during PCE reduction (Figure 4),
again consistent with their formation through a highly
reactive intermediate or intermediates.
the observed C mass balance, while solid lines are model fits.
2
Chlorinated Acetylene Reduction. As the chlorinated
acetylenes do not make up a significant portion of the mass
balance in the reduction of PCE or TCE, reductive elimination
to such species could only be a significant pathway if these
react too rapidly to enable their accumulation. Independent
confirmation was therefore sought by synthesizing dichloro-
and chloroacetylene and measuring their rates and products
of reaction. On the basis of our attempts to model the
reduction of TCE, we anticipated that chloroacetylene would
react to form acetylene and vinyl chloride via reactions 14
and 15 (see Figure 1) if it were the intermediate responsible
for the generation of these species. Dichloroacetylene could
give these same products after first being reduced to
chloroacetylene via reaction 17; alternatively, reduction of
the triple bond could result in trans-DCE via reaction 16 or
(in principle) cis-DCE.
Analysis of the chloroacetylene solution via GC/ MS
revealed no contaminant peaks other than those expected
(unreacted cis-DCE and residual pentane from washing of
the potassium hydride). Results obtained for chloroacetylene
reaction with zinc are shown in Figure 5. Measured rates of
chloroacetylene reaction were sufficiently rapid that no
detectable accumulation would be predicted to result from
TCE reduction. The sole observed products of chloroacety-
lene reaction were acetylene and vinyl chloride, as antici-
pated. Note that at this zinc loading (1 g/ 160 mL as compared
to the 25 g/ 160 mL used for the experiments with the DCE
isomers), reaction of the cis-DCE present to these products
FIGURE 4. Reduction of PCE (() in 150 mL of 0.1 M NaCl/0.05 M Tris
buffer (pH 7.2) by5g ofZn(0). Major products are TCE (4), trans-DCE
(3), and acetylene (O). Inset shows trace products cis-DCE (2) and
vinyl chloride ()). The dotted line represents the observed C mass
2
balance, while solid lines are model fits.
(15). Such a sequence would predict curves for acetylene
that were concave upward, in sharp contrast to the observed
results. In a similar manner, the rate of appearance of vinyl
chloride from TCE reduction (Figure 3) as well as that of
trans-DCE formation from PCE reduction (Figure 4) were
too rapid to be attributable to a sequence of hydrogenolysis
reactions of the sort TCE f trans-DCE f VC or PCE f TCE f
trans-DCE.
Further evidence that the acetylene and vinyl chloride
observed in TCE reduction did not arise from reactions of
9
VOL. 32, NO. 19, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 0 2 1

TABLE 2. Percentage of Reaction Occurring via Reductive
Elimination
% reductive
elimination
∆E )
2
species
PCE
TCE
cis-DCE
trans-DCE
(Ered. elim - E
)
hydrogenolysis
15((2)a
30((4)
85((8)
95((2)
0.049
0.075b
0.086
0.085
a
b
Uncertainties represent 95% confidence lim its. Using the value
for the reduction of TCE to trans-DCE (the principal observed DCE
isom er).
tion of dichloroacetylene and only traces of chloroacetylene
would be anticipated from PCE reduction, consistent with
our observations. Dichloroacetylene reacted to form trans-
DCE, chloroacetylene, acetylene, and vinyl chloride as the
only observed reaction products. Reaction of residual TCE
was not a significant source of these products at this metal
loading. The observed acetylene concentration, however,
was much too high to arise from the sequential reaction of
dichloroacetylene f chloroacetylene f acetylene or from
the other known acetylene precursors (trans-DCE, TCE)
present. Modeling the data suggested that acetylene was
also being produced by reaction of another species. The
concentration of this hypothesized contaminant(s), as cal-
culated by the model, is given by the dotted line in Figure
6. Direct analysis of the THF solution via GC/ MS and GC/
FID revealed several contaminant peaks besides the expected
TCE and pentane, most likely arising from side reactions
during the synthesis. These peaks were not detectable via
the headspace analysis method employed in monitoring the
reaction kinetics. The largest of these peaks has a mass
spectrum consistent with that given for 1,2-dichloro-1-
methoxyethylene (38).
FIGURE 6. Reduction of dichloroacetylene in 0.1 M NaCl/0.05 M
Tris buffer (pH 7.2) by 0.5 g of Zn(0). Solid symbols (9) represent
dichloroacetylene peak areas. The model fit for the dichloroacety-
lene data represents model-calculated concentrations. Open
symbols for trans-DCE (3), acetylene (O), and vinyl chloride ()) are
measured concentrations. The chloroacetylene concentration (+)
was calculated using the relationship between peak area and
aqueous concentration developed in Figure 5. The dashed line
represents the calculated concentration of a hypothesized con-
taminant assumed to react to form acetylene.
is extremely slow. In principle, water may add to the triple
bond of chloroacetylene to form an R-halo vinyl alcohol (23).
This can rearrange to acetate after addition of a second water
molecule. Experiments in the absence of zinc, however, failed
to reveal any significant loss of chloroacetylene that might
be attributed to hydrolysis over a 24-h period.
Due to a lack of quantitative standards, concentrations
of chloroacetylene could not be directly determined. A rate
constant for chloroacetylene reaction was therefore com-
puted by simultaneously fitting the chloroacetylene peak area
and the concentrations of the reaction products as a function
of time. This was accomplished by expressing the differential
equation for the disappearance of chloroacetylene (see eq
1) in terms of peak area. An additional equation was included
in which a linear coefficient relating peak area to concentra-
tion was treated as a fitting parameter. The model fit through
the chloroacetylene peak areas shown in Figure 5 therefore
represents the aqueous chloroacetylene concentration values
calculated by the model.
Justification for this approach is provided by computing
molar response factors from the results. Using the modified
EPICS method (27), dimensionless Henry’s law constants at
60 °C in the 50 mM Tris/ 0.1 M NaCl solution were found to
be 2.4 for chloroacetylene and 3.5 for dichloroacetylene (see
below). From these Henry’s law constants, the model-derived
aqueous concentrations, and the known air:water ratio in
the vials, the total mass of chlorinated acetylene injected
onto the GC, and thus the molar FID response factor (µV-
s/ nmol injected), could be calculated at each time point. For
chloroacetylene and dichloroacetylene, the FID molar re-
sponse factors (relative to acetylene) were estimated as 1.07
and 1.12, respectively. The similarity of these response factors
to that of acetylene gives independent confirmation of the
validity of the model-derived concentrations. We note that
FID molar response factors for the chlorinated ethylenes
display a similar trend with increasing chlorination.
Results obtained for the reaction of dichloroacetylene with
zinc are displayed in Figure 6. No hydrolysis of dichloro-
acetylene was observed in the absence of zinc over a period
of 24 h. Dichloroacetylene reaction rates in the presence of
zinc were sufficiently rapid so that no detectable accumula-
Discussion
Factors Influencing Reaction Products. Based on the rate
constants given in Table 1, the fraction of the overall reaction
that occurs via reductive elimination may be calculated for
each chlorinated ethylene possessing an R,â pair of chlorine
atoms. This percentage was calculated by comparing the
rate constant for reductive elimination to that for overall
parent compound disappearance (reductive elimination +
hydrogenolysis). Error limits were obtained by propagating
the errors associated with the different rate constants. As
the results summarized in Table 2 reveal, reductive elimina-
tion is an important contributor and cannot be ignored in
pathway modeling under the conditions employed in the
present experiments. Additional studies (beyond the scope
of the present investigation) will be required to address the
impact of reaction conditions (e.g., electrolyte composition
and concentration; pH; concentration of potential hydrogen
atom donors such as methanol, THF, and natural organic
matter; temperature; metal cleaning or aging of the metal
surface) on product branching ratios.
Note that as the reductive elimination reaction becomes
more thermodynamically favorable relative to hydrogenoly-
sis, the percentage of reaction occurring through reductive
elimination increases. If we assume that the thermodynamic
data for the reductive elimination and hydrogenolysis
reactions are of comparable accuracy, it does not appear
that the reaction products are purely under thermodynamic
control. If the E2 values for hydrogenolysis and reductive
elimination were identical, a 50:50 split of reaction products
would be expected, while it is clear that even with reductive
elimination favored thermodynamically (e.g., for PCE and
TCE), only 15-30% of the reaction proceeds via this route.
Product distributions are affected not only by the com-
petition between the reductive elimination and hydro-
9
3 0 2 2 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 19, 1998

genolysis pathways for the chlorinated ethylenes but also by
the branching ratio between hydrogenolysis and reduction
of the triple bond for the chlorinated acetylenes. Only 8%
of chloroacetylene is reduced to form vinyl chloride. With
dichloroacetylene, however, 82% of the reaction proceeds
via reduction of the triple bond to form trans-DCE.
Hydrogenation (reduction of triple bonds to double bonds
by H2) is usually thought to occur via cis addition, giving rise
to cis-olefins. For example, we have found in our laboratory
that 2-butyne is reduced to cis-2-butene by iron (26). Zinc
has also been shown to promote the stereospecific reduction
of substituted alkynes to cis-alkenes in ethanol (39, 40) or
THF/ MeOH/ H2O (41); a strong preference for cis reduction
has also been observed for the zinc-copper couple in ethanol
(42). Our results with zinc, however, indicate that the
reduction of the dichloroacetylene triple bond forms trans-
DCE rather than the cis isomer. There are precedents for
reduction of substituted acetylenes to trans-olefins by certain
dissolving metal reductants. For example, reaction of sodium
metal in ammonia produces trans-olefins from substituted
acetylenes (43-45), and the zinc-copper couple produces
a mixture of the cis and trans isomers (3:5) for reduction of
phenylpropiolic acid (C6H5CtCCOOH) to cinnamic acids
(C6H5CHdCHCOOH) in an ammonium chloride-ammonia
solution (46). Zinc has also been cited for use in the
semireduction of alkynes to trans-alkenes (47).
When PCE or TCE react with zinc, several of the products
(namely, trans-DCE, vinyl chloride, and acetylene) may arise
through more than one route. For example, vinyl chloride
may be produced via hydrogenolysis (reaction 7, 9, or 11) or
via reduction of chloroacetylene (reaction 15). On the basis
of the rate constants given in Table 1, we can determine the
percentage of the trans-DCE, vinyl chloride, and acetylene
that originates from an initial reductive elimination of PCE
or TCE via reactions 2 or 6. At relatively low conversions
(<25%) of TCE, greater than 99.7% of the vinyl chloride and
acetylene observed arise from an initial reductive elimination
reaction of TCE to form chloroacetylene via reaction 6. At
longer times (92% TCE conversion), the DCE daughter
products have begun to react to form acetylene and vinyl
chloride, but >97% of vinyl chloride and acetylene present
still results from an initial reductive elimination to chloro-
acetylene. Similar results are obtained for PCE. For 92%
conversion of PCE, 95% of the trans-DCE observed is
produced via a dichloroacetylene intermediate. At shorter
times, the percentage is higher, for hydrogenolysis of TCE
has not yet contributed appreciably to the observed trans-
DCE accumulation. Virtually all of the acetylene and vinyl
chloride observed during PCE reduction arise either from
the reductive elimination of PCE to dichloroacetylene or else
from the subsequent reductive elimination of the initial
daughter product (TCE) to chloroacetylene.
Linear Free Energy Relationships. If reaction were to
occur via an outer-sphere single electron-transfer pathway,
a good correlation between one-electron reduction potential
(E1) and the relevant rate constant might be anticipated (48).
Our results indicate that rate constants do correlate well with
one-electron reduction potentials, as shown in Figure 7a.
The calculated regression equation is log(kSA) ) 8.23((1.79)
× E1 + 2.21((1.30), where stated errors represent 95%
confidence limits. The regression reported excludes the data
point for vinyl chloride (as with regressions against E2, see
below). The good correlation of rate constants with reduction
potential is consistent with electron transfer as the limiting
factor dictating chloroethylene reactivity with zinc (rather
than mass transfer, adsorption, or desorption as a rate
controlling step).
FIGURE 7. Correlation of surface area-normalized rate constants
(kSA, Table 1) (a) with one-electron reduction potentials and (b)
with two-electron reduction potentials. Numbering corresponds to
the reactions indicated in Table 1. Regression lines exclude the
data point for vinyl chloride (filled circle, reaction 12). If no error
bar is shown, it is smaller than the symbol.
any insight as to the distribution of reaction products that
will result. Separate correlation of hydrogenolysis rate
constants and rate constants for reductive elimination against
two-electron reduction potentials furnishes a means of
predicting reaction products as well as rates. Figure 7b
displays correlations of kSA versus two-electron reduction
potential (E2) values for the hydrogenolysis component of
the reaction [log(kSA) ) 53.56((15.06) × E2 - 31.85((7.86)]
and the reductive elimination component [log(kSA) ) 60.39
((15.09) × E2 - 39.58((8.97)], where stated errors again
represent 95% confidence limits. The high rate constant for
vinyl chloride reaction to ethylene (filled circle) seems to be
anomalous and was excluded from the relevant regression.
Environm ental Significance of Reductive Elim ination.
At first glance, reductive elimination of PCE or TCE may not
seem overly beneficial. For example, the subsequent reduc-
tion of dichloroacetylene produces trans-DCE, a relatively
persistent byproduct as compared to the parent species, PCE.
This reaction pathway, however, partially circumvents TCE
formation, a product that undergoes reduction in part to
1,1-DCE (a species that reacts only to vinyl chloride) and to
cis-DCE (which produces approximately three times more
vinyl chloride per mole as does trans-DCE). Mathematical
simulations of the reactions of saturated PCE and TCE
solutions with zinc reveal that cis-DCE, by virtue of being the
least reactive species in the system, represents a bottleneck
Although one-electron reduction potentials could be used
to predict the overall rate constant pertaining to disappear-
ance of a particular species, this descriptor does not provide
9
VOL. 32, NO. 19, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 0 2 3

(5) Vogan, J. L.; Gillham, R. W.; O’Hannesin, S. F.; Matulewicz, W.
H.; Rhodes, J. E. Site specific degradation of VOCs in groundwater
using zero-valent iron. Natl. Meet.sAm. Chem. Soc., Div. Environ.
Chem. 1995, 35, 800-804 (Abstr.).
(6) Yamane, C. L.; Warner, S. D.; Gallinatti, J. D.; Szerdy, F. S.; Delfino,
T. A. Installation of a subsurface groundwater treatment wall
composed of granular zero-valent iron. Natl. Meet.sAm. Chem.
Soc., Div. Environ. Chem. 1995, 35, 792-795 (Abstr.).
(7) Tratnyek, P. G.; Johnson, T. L.; Scherer, M. M.; Eykholt, G. R.
Ground Water Monit. Rem. 1997, 17, 108-114.
(8) Boronina, T.; Klabunde, K. J.; Sergeev, G. Environ. Sci. Technol.
1995, 29, 1511-1517.
(9) Warren, K. D.; Arnold, R. G.; Bishop, T. L.; Lindholm, L. C.;
Betterton, E. A. J. Hazard. Mater. 1995, 41, 217-227.
(10) Schlimm, C.; Heitz, H. Environ. Prog. 1996, 15, 38-47.
(11) Butler, L. C.; Staiff, D. C.; Sovocool, W.; Davis, J. E. J. Environ.
Sci. Health 1981, B16, 49-58.
(12) Song, H.; Kim, Y.-H.; Carraway, E. R.; Batchelor, B. Manuscript
in preparation.
(13) Xu, Y.; Schwartz, F. W.; Traina, S. J. Environ. Sci. Technol. 1994,
28, 1472-148.
(14) Chen, X.; Wright, J. V.; Conca, J. L.; Peurung, L. M. Environ. Sci.
Technol. 1997, 31, 624-631.
(15) Roberts, A. L.; Totten, L. A.; Arnold, W. A.; Burris, D. R.; Campbell,
T. J. Environ. Sci. Technol. 1996, 30, 2654-2659.
(16) Gillham, R. W.; O’Hannesin, S. F. Ground Water 1994, 32, 958-
967.
(17) Schreier, C. G.; Reinhard, M. Chemosphere 1994, 29, 1743-
1753.
(18) Muftikian, R.; Fernando, Q.; Korte, N. Water Res. 1995, 29, 2434-
2439.
(19) Johnson, T. L.; Scherer, M. M.; Tratnyek, P. G. Environ. Sci.
Technol. 1996, 30, 2634-3640.
(20) Jones, D. A. Principles and Prevention of Corrosion; Macmillan
Publishing Company: New York, 1992.
(21) Johnson, T. L.; Fish, W.; Gorby, Y. A.; Tratnyek, P. G. J. Contam.
Hydrol. 1998, 29, 379-398.
(22) Denis, J. N.; Moyano, A.; Greene, A. E. J. Org. Chem. 1987, 52,
3461-3462.
(23) Piganiol, P. Acetylene Homologs and Derivatives; Mapleton
House: New York, 1950.
to attaining the drinking water limit for vinyl chloride.
Reductive elimination of PCE reduces the amount of TCE
formed, decreasing the amounts of cis-DCE, 1,1-DCE, and
finally vinyl chloride produced.
Recent work has investigated the uncertainty associated
with barrier design based on variations in rate constants,
flow rates, and other transport processes (7, 49, 50). In these
studies, however, the branching ratio between hydrogenolysis
and reductive elimination was not explicitly known, making
it difficult to quantitatively assess the importance of reductive
elimination reactions. The results of the present study should
facilitate more definitive investigations. We can further
demonstrate the environmental significance of reductive
elimination by comparing model predictions using the rate
constants determined in this work with those obtained from
a model that assumes the disappearance of any given species
can be described by the same overall rate constant, but with
reaction only occurring through hydrogenolysis. In other
words, PCE is assumed to react solely to TCE via a rate
constant obtained by summing reactions 1 and 2, TCE is
assumed to react solely to a mixture of the three DCE isomers
(68.7% trans-DCE, 28.1% cis-DCE, and 3.2% 1,1-DCE) with
an overall rate constant equal to the sum of reactions 3-6,
and cis- and trans-DCE react only to vinyl chloride via a sum
of the rate constants for reactions 9 plus 10 or 7 plus 8,
respectively. For a saturated solution of PCE, the model that
constrains all transformations to a sequence of stepwise
hydrogenolysis reactions requires hydraulic residence times
(for a plug flow reactor) to reach the drinking water limit
for vinyl chloride (2 µg/ L, ref 51) that are 76% greater than
if reductive elimination reactions are incorporated. Similar
calculations indicate hydraulic residence times that are 41%
greater to attain the vinyl chloride drinking water limit from
a saturated TCE solution when reductive elimination reac-
tions are routed into the hydrogenolysis pathway.
Despite the production of undesirable species from
subsequent reactions of chlorinated acetylenes, we note that
if reductive elimination did not occur, all PCE or TCE initially
present would be converted to vinyl chloride. Even though
chlorinated acetylenes are toxic (22-24), their rapid rate of
reaction with zinc does not allow accumulation to high
concentrations, decreasing the likelihood that these species
will be of toxicological concern in the effluent from a metal-
based treatment scheme. This suggests that reductive
elimination, by partially circumventing the formation of
persistent, undesirable products, should be viewed as an
environmentally desirable reaction.
(24) Rutledge, T. F. Acetyleneic Compounds: Preparation and Sub-
stitution Reactions; Reinhold: New York, 1968.
(25) American Society for Metals. Metals Handbook, 9th ed.; Vol.
5sSurface Cleaning, Finishing, and Coating; ASM: Metals Park,
OH, 1982.
(26) Fennelly, J. P.; Roberts, A. L. Environ. Sci. Technol. 1998, 32,
1980-1988.
(27) Gosset, J. M. Environ. Sci. Technol. 1987, 21, 202-208.
(28) Eberson, L. Acta Chem. Scand. B. 1982, 36, 533-543.
(29) Ho, W.; Yu, Q.; Bozzelli, J. W. Combust. Sci. Technol. 1992, 85,
22-63.
(30) Taylor, P. H.; Dellinger, B.; Tirey, D. A. Int. J. Chem. Kinet. 1991,
23, 1051-1074.
(31) Wagman, D. D.; Evans, W. H.; Parker, V. B.; Schumm, R. H.;
Halow, I.; Bailey, S. M.; Churney, K. L.; Nuttall, R. L. J. Phys.
Chem. Ref. Data 1982, Suppl. 2.
Acknowledgments
This work was funded through Contract F08637 95 C6037
from the Department of the Air Force, an NSF Young
Investigator Award (Grant BES-9457260) to A.L.R., and an
EPA STAR Fellowship to W.A.A. Special thanks are extended
to Beth Carraway for sharing unpublished data concerning
the use of hydroxyapatite. The authors also wish to thank
Lisa Totten for generously providing calculations of one-
and two-electron reduction potentials for the chlorinated
ethylenes; Pei Chun Chiu for helpful discussions concerning
chlorinated acetylenes; and Bill Ball and the three anonymous
reviewers, whose comments greatly clarified and improved
this paper.
(32) Mackay, D.; Shiu, W. Y.; Ma, K. C. Illustrated Handbook of
Physical-Chemical Properties and Environmental Fate for Or-
ganic Chemicals. Vol. 3; Lewis Publishers: Chelsea, MI, 1993.
(33) Hine, J.; Mookerjee, P. K. J. Org. Chem. 1975, 40, 292-297.
(34) Arnold, W. A.; Roberts, A. L. Development of a quantitative model
for chlorinated ethylene reduction by zero-valent metals. Natl.
Meet.sAm. Chem. Soc., Div. Environ. Chem. 1997, 37, 76-77
(Abstr.).
(35) Scherer, M. M.; Tratnyek, P. G. Dechlorination of carbon
tetrachloride by iron metal: Effect of reactant concentrations.
Natl. Meet.sAm. Chem. Soc., Div. Environ. Chem. 1995, 35, 805-
806 (Abstr.).
(36) Gotpagar, J.; Grulke, E.; Tsang, T.; Bhattacharyya, D. Environ.
Prog. 1997, 16, 137-143.
(37) Mohn, W. W.; Tiedje, J. M. Microbiol. Rev. 1992, 56, 482-507.
(38) Shainyan, B. A.; Vereshchagin, A. L. Russ. J. Org. Chem. Trans.
Zh. Org. Khim. 1993, 29, 1981-1989.
(39) Aerssens, M. H. P. J.; Brandsma, L. J. Chem. Soc., Chem. Commun.
1984, 735-736.
(40) Aerssens, M. H. P. J.; van der Heiden, R.; Heus, M.; Brandsma,
L. Synth. Commun. 1990, 20, 3421-3425.
(41) Chou, W. N.; Clark, D. L.; White, J. B. Tetrahedron Lett. 1991,
32, 299-302.
(42) Clarke, A. J.; Crombie, L. Chem. Ind. 1957, 143.
Literature Cited
(1) Campbell, T. J.; Burris, D. R.; Roberts, A. L.; Wells, J. R. Environ.
Toxicol. Chem. 1997, 16, 625-630.
(2) Orth, W. S.; Gillham, R. W. Environ. Sci. Technol. 1996, 30, 66-
71.
(3) Liang, L.; Korte, N.; Goodlaxson, J. D.; Clausen, J.; Fernando, Q.;
Muftikian, R. Ground Water Monit. Rem. 1997, 17, 122-127.
(4) O’Hannesin, S. F.; Gillham, R. W. Ground Water 1998, 36, 164-
170.
9
3 0 2 4 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 19, 1998

(43) Smith, M. Dissolving Metal Reductions. In Reduction: Techniques
and Applications in Organic Synthesis; Augustine, R. L., Ed.;
Marcel Dekker: New York, 1968; pp 95-170.
(44) Evans, R. F. Modern Reactions in Organic Synthesis; Timmons,
C. J., Ed.; Van Nostrand Reinhold: London, 1970; pp 5-40.
(45) Kemp, D. S.; Vellachio, F. Organic Chemistry; Worth Publishers:
New York, 1980.
(46) Ott, E.; Barth, V.; Glemser, O. Ber. Dtsch. Chem. Ges. 1934, 67,
1669.
(47) Hudlicky, M. Reductions in Organic Chemistry; John Wiley &
Sons: New York, 1984.
terization, Containment, Remediation, and Performance in
Environmental Geotechnics; Acar, Y. B., Daniel, D. E., Eds.;
American Society of Civil Engineers: New York, 1996; pp 1608-
1621.
(50) Eykholt, G. R. Uncertainty-based scaling of iron reactive barriers.
In In Situ Remediation of the Geoenvironment; Evans, J., Reddi,
L., Eds.; American Society of Civil Engineers: New York, 1997;
pp 41-55.
(51) Pontius, F. W. J. Am. Water Works Assoc. 1998, 90, 38-46.
(48) Eberson, L. Electron-Transfer Reactions in Organic Chemistry;
Springer-Verlag: New York, 1987.
(49) Eykholt, G. R. Contaminant transport issues for reactive-
permeable barriers. In Geoenvironment 2000, Vol. 2, Charac-
Received for review March 16, 1998. Revised manuscript
received June 1, 1998. Accepted June 15, 1998.
ES980252O
9
VOL. 32, NO. 19, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 0 2 5