Reductive transformation of trichloroethene by cobalamin: Reactivities of the intermediates acetylene, chloroacetylene, and the DCE isomers

Article abstract of DOI:10.1021/es9705248

The transformations of acetylene, chloroacetylene, 1,1-dichloroethene (DCE), and cis- and trans-DCE mediated by cobalamin in the presence of titanium(III) citrate were investigated at pH 8 and 22 °C. Acetylene quantitatively reacted to ethene via vinylcobalamin as the proposed intermediate. Chloroacetylene reacted to acetylene and vinyl chloride. Proposed intermediates are ethynylcobalamin and vinylcobalamin; respectively. The principal initial reaction of chloroacetylene formed ethynylcobalamin which decomposed to acetylene. The proposition for ethynyl- and vinylcobalamin formation is based on fitting reaction models to kinetic data. Kinetic modeling suggests half-lives for ethynyl- and vinylcobalamin of 1.4 and 251 h, respectively. 1,1-Dichloroethene reacted to approximately 20% volatiles (ethane, ethane, vinyl chloride, and acetylene) and 80% unidentified nonvolatile products. cis- and trans-DCE transformed slowly end produced small yields of vinyl chloride, ethene, and ethane. The transformations of acetylene, chloroacetylene, 1,1-dichloroethene (DCE), and cis- and trans-DCE mediated by cobalamin in the presence of titanium(III) citrate were investigated at pH 8 and 22 °C. Acetylene quantitatively reacted to ethene via vinylcobalamin as the proposed intermediate. Chloroacetylene reacted to acetylene and vinyl chloride. Proposed intermediates are ethynylcobalamin and vinylcobalamin, respectively. The principal initial reaction of chloroacetylene formed ethynylcobalamin which decomposed to acetylene. The proposition for ethynyl- and vinylcobalamin formation is based on fitting reaction models to kinetic data. Kinetic modeling suggests half-lives for ethynyl- and vinylcobalamin of 1.4 and 251 h, respectively. 1,1-Dichloroethene reacted to approximately 20% volatiles (ethene, ethane, vinyl chloride, and acetylene) and 80% unidentified nonvolatile products. cis- and trans-DCE transformed slowly and produced small yields of vinyl chloride, ethene, and ethane.

Full text of DOI:10.1021/es9705248

Environ. Sci. Technol. 1998, 32, 1207-1213
because it bypasses vinyl chloride, a relatively stable car-
Reductive Transformation of
Trichloroethene by Cobalamin:
Reactivities of the Intermediates
Acetylene, Chloroacetylene, and the
DCE Isomers
cinogen (20). There are three potential links between the
two pathways: (i) R,â-dihaloelimination of cis- and trans-
DCE and (ii) R-dihaloelimination of 1,1-DCE forming acety-
lene or (iii) hydrogenation of chloroacetylene forming vinyl
chloride.
I
The “super reduced” cobalamin (cbl ) is both a powerful
reductant and a nucleophile that has been shown to react
with chlorinated substrates via single electron transfer (SET)
(12, 16), nucleophilic substitution (S 2) or addition reactions
N
‡
M A R C O S E M A D E N I ,
†
P E I - C H U N C H I U , A N D
M A R T I N R E I N H A R D *
N
(11, 21-23). SET and S 2 reactions can compete (21, 24).
SET leads to reductive bond cleavage of the carbon-chlorine
bond and the formation of a radical (25). Spectroscopic
Environmental Engineering and Science Department of Civil
Engineering Stanford University,
I
evidence has shown that reactions of cbl with alkylhalids
Stanford, California 94305-4020
can lead to relatively stable alkyl-, vinyl-, or ethynylcobal-
amins (11, 12, 16). These alkylcobalamin complexes can
decompose by reductive cleavage of the cobalt-carbon bond
with titanium(III) citrate as the reductant (16). For PCE and
TCE, involvement of titanium(III) citrate in cobalt-carbon
cleavage reactions has been inferred from the observed pH
dependence of the product distribution found in TCE
transformations (12). Reductive cleavage is facilitated by
electron-withdrawing substituents (26) and is rapid compared
to nonreductive cleavage reactions (27-31). Heterolytic
The transformations of acetylene, chloroacetylene, 1,1-
dichloroethene (DCE), and cis- and trans-DCE mediated by
cobalamin in the presence of titanium(III) citrate were
investigated at pH 8 and 22 °C. Acetylene quantitatively
reacted to ethene via vinylcobalamin as the proposed
intermediate. Chloroacetylene reacted to acetylene and
vinyl chloride. Proposed intermediates are ethynylco-
balamin and vinylcobalamin, respectively. The principal
initial reaction of chloroacetylene formed ethynylcobalamin
which decomposed to acetylene. The proposition for
ethynyl- and vinylcobalamin formation is based on fitting
reaction models to kinetic data. Kinetic modeling sug-
gests half-lives for ethynyl- and vinylcobalamin of 1.4 and
II
reductive cleavage leading to cbl and the carbanion is favored
I
over homolytic reductive cleavage which forms cbl and a
radical (26, 29, 32). Thermally or sterically induced nonre-
II
ductive cleavage of the cobalt-carbon bond yields cbl and
a radical (33, 34). The radical intermediates are rapidly
reduced followed by either â-elimination or protonation (35,
36) or, alternatively, abstract a hydrogen from hydrogen
donors (12, 16). The formation and decomposition of
intermediate cobalamin complexes from reactions of ethynes
with cbl have not been investigated extensively in the present
context.
251 h, respectively. 1,1-Dichloroethene reacted to ap-
proximately 20% volatiles (ethene, ethane, vinyl chloride, and
acetylene) and 80% unidentified nonvolatile products.
cis- and trans-DCE transformed slowly and produced small
yields of vinyl chloride, ethene, and ethane.
In this paper, we present transformation data of chlo-
roacetylene, acetylene, and the three DCE isomers, inter-
mediates of the TCE transformation process. It is shown
that chloroacetylene is a precursor for vinyl chloride and
acetylene. The data reported here provide kinetic evidence
for the formation of intermediate cobalamin complexes are
used to quantify their stability and help to explain product
distributions. The data serve as a basis for future efforts to
model the TCE transformation process.
Introduction
The capability of microbial cultures to reductively transform
chlorinated aliphatic hydrocarbon compounds has been
attributed to transition metal coenzymes, such as cobalamin
(
vitamin B12), cofactor F430, or heme (1-4). This hypothesis
is supported by studies of pure cultures that reductively
dechlorinate tetrachloroethene (PCE) and trichloroethene
(
TCE). These studies indicate corrinoids are involved in the
dehalogenation reactions (3, 5-10). To gain insight into the
chemistry of such biological dehalogenation reactions, model
systems are studied using cobalamin (cbl) and bulk reduc-
Materials and Methods
Chem icals and Reagents. All chemicals were from com-
mercial sources (Aldrich) described previously (16) and used
as received without further purification: aquocobalamin
tants such titanium(III) citrate or reduced sulfur compounds
III
(
4, 11-18). Titanium(III) citrate reduces cob(III)alamin (cbl )
(
vitamin B12a, >96%, Fluka), TCE (99+%), cis-DCE (99%),
I
I
to cbl , a strong reductant and nucleophile (19). Cbl is
capable of dechlorinating TCE to ethene (4, 12, 13). Two
separate pathways leading to ethene have been identified:
a pathway in which sequential hydrogenolysis leads directly
to ethene through the intermediates cis-, trans-, and 1,1-
dichloroethene (DCE) and vinyl chloride (4) and a pathway
which is initiated by R,â-dihaloelimination of TCE to form
chloroacetylene (13). Chloroacetylene can be transformed
by hydrogenolysis to acetylene and hydrogenation of acetyl-
ene leads to ethene (13). The latter pathway is significant
trans-DCE (98%), and 1,1-DCE (99%). Acetylene (99.6%,
Altair) was obtained as a compressed gas. The calibration
standards for vinyl chloride, acetylene, and ethene, ranging
from 1000 ppm to 10% in nitrogen, were purchased from
Scott Specialty Gases. TCE, cis-DCE, trans-DCE, and 1,1-
DCE spike solutions and a combined calibration standard
solution containing TCE and the DCE isomers were prepared
in methanol (99.9+%) and stored at -15 °C. All other reagent
stock solutions were prepared and stored as previously
described (16). Tris (ultrapure, Baker) was used as the pH
buffer.
*
Corresponding author. E-mail: reinhard@ce.stanford.edu; fax:
(
415) 725-3162.
Synthesis of Chloroacetylene. Chloroacetylene was
synthesized from cis-DCE using the procedures described
by Denis et al. (37). Chloroacetylene is a toxic gas and can
be explosive under oxic conditions (38, 39). A diazomethane
generator (MNNG diazomethane-generation apparatus, Al
‡
Eidgen o¨ ssiche Technische Hochschule (ETH Z u¨ rich), Umwelt-
naturwissenschaften, Um weltnatur- und Um weltsozialwissen-
schaften (UNS), ETH-Zentrum, HCS B6, CH-8092 Z u¨ rich, Switzerland.
†
Department of Civil & Environmental Engineering, University of
Delaware, Newark, DE 19716.
S0013-936X(97)00524-5 CCC: $15.00
Published on Web 03/17/1998

1998 Am erican Chem ical Society
VOL. 32, NO. 9, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1 2 0 7

drich) placed in a nitrogen-flushed glovebag (Atmosbag,
Aldrich) was used as a reaction apparatus A total of 10 mmol
of cis-DCE in 2 mL of tetrahydrofuran (THF, anhydrous,
substrate, {S
i
}, using eqs 2 and 3, respectively:
-d{Si}/ dt ) k′obs {Si}
(2)
(3)
9
9.9%, inhibitor-free, Aldrich) was added to excess lithium
hydride powder (95%, Aldrich) suspended in 1.5 mL of THF
in the inner reaction tube of the closed and oxygen-free
reaction apparatus. The outer tube of the reactor contained
-d{S }/ dt ) k
i
obs
where k′obs and kobs are the observed mass-based apparent
first-order and zero-order rate constants in units of inverse
hours and micromoles per hour, respectively. Using standard
statistical techniques, 95% confidence intervals were ob-
tained.
Kinetic data were evaluated using the kinetic modeling
software package, Scientist (MicroMath, Scientific Software,
Salt Lake City, UT). The models were described by sets of
ordinary differential equations (ODEs) representing the
following rate laws:
5
mL of THF. The reactor was kept on an ice/ salt slush bath
during the reaction. A total of 10 µL of methanol (anhydrous,
9.8%, Aldrich) was injected into the inner tube to initiate
9
the reaction. After 45 min, another 300 µL of methanol was
slowly injected. The reaction was terminated after 4 h and
the solution from the outer tube was transferred under
nitrogen into an amber vial and stored at -15 °C. The
chloroacetylene solution served as the standard stock solu-
tion. The chloroacetylene concentration (27 mM) was
determined using the approach described below.
Kinetic Experim ents. All experiments were conducted
at 22 ( 1 °C and pH 8.0 under anaerobic and light-excluded
conditions. The experimental procedures used were similar
to those described previously (16) except as follows: in each
-d{S }/ dt )
k′ {′ S }{C}
(4)
(5)
(6)
i
∑
j
i
-
d{IC }/ dt )
k′ {IC } -
k′′{S } {C}
i
∑
j
i
∑
j
i
6
4 mL amber borosilicate bottle, a 32 mL aqueous solution
was placed which contained 15 mM titanium(III) citrate, 10
µM cbl, and 60 mM Tris, leaving 32 mL of headspace (N
). The reaction was initiated by injecting microliter
d{P }/ dt )
k′′{S } {C} +
k′ {IC }
i
∑
j
i
∑
j
i
2
/
H
2
where k
j
j
are the second-order and first-order mass-
} are the total amounts of
the reduced catalyst, cbl , and the intermediate cobalamin
complex, respectively, and {P } is the total molar mass of
amounts of substrate spike solution into the reaction solution
and equilibrated by vigorous manual shaking of the bottles.
Preliminary tests indicated equilibration to be complete in
less than 3 min (data not shown). During the entire reaction
period, the bottles were kept on a shaker table at 400 rpm.
Conditions for all experiments were the same except for the
initial amounts of the substrate and cbl, which in some cases
were varied. Identical solutions containing the substrate
without cbl served as controls. The pH of the reaction mixture
was measured using an EA 940 Expandable Ion Analyzer
based rate constants, {C} and {IC
i
I
i
product. The subscripts i and j indicate the different
compounds and reactions involved. To obtain solutions for
a given time interval, the ODEs were solved by numerical
integration using EPISODE, an initial value ODE solver
provided by the software package Scientist.
Only volatile compounds were measured. The interme-
diate cobalamin complexes and unknown products were
obtained by fitting the postulated models to the data sets
assuming mass conservation (Table 1). The experimental
data were fitted using a least-squares minimization algorithm
(Scientist) constrained to the ODEs with the fitting parameters
(
Orion, Boston, MA). Concentrations of volatiles were
determined by sampling 50 µL aliquots of the gaseous
headspace using GC/ FID.
The analytical methods used were adapted from Burris
et al. (13). Headspace aliquots of 50 µL were injected into
an HP 5890 series II gas chromatograph (GC)/ flame ionization
detector (FID). Calibration factors were calculated based
on the total amount of compounds added to the reactor
bottle. For chloroacetylene, an indirect method was used as
described below. The freshly synthesized chloroacetylene
was identified using an HP 5890 GC/ MSD (5970 mass selective
detector) under EI mode, scanning from 40 to 200 amu, and
a 30 m × 0.32 mm DB-5 capillary column (0.25 µm film
thickness, J&W Scientific) at an isothermal temperature of
j j
k′′and k′, j ) 1, 2, 3, ..., r, for each second- or first-order
reaction r of the kinetic model m, respectively (Table 1), and
to the mass conservation equation for the catalyst C (cbl ):
I
{C} ) {C} -
∑{ICi}
(7)
0
0
where {C} is the initial amount of catalyst present in the
reactor. This mass conservation is based on the assumption
that (1) all oxidized free cobalamin (cbl ) is rapidly recycled
to cbl by titanium(III) citrate and (2) no cobalamin con-
II
I
4
0 °C. To verify chloroacetylene as a reaction intermediate
sumption reactions exist other than intermediate complex
II
III
of the TCE transformation process, 560 µmol of TCE was
reacted with 0.32 µmol of cobalamin under standard condi-
tions. The reaction mixture was analyzed by GC/ FID as
described above. Chloroacetylene was identified by retention
times comparing the reaction mixture with a control and the
freshly prepared chloroacetylene standard.
formation. Experimentally, it was found that cbl and cbl
are reduced within seconds (data not shown). Standard
deviations for the best-fit parameters k′′and k′ and overall
, were calculated by the software
j
j
2
correlation coefficients, r
m
package Scientist. The consistency of the kinetic models
was tested by incorporating the rate constants derived from
the simplest reaction model into successively more complex
reaction models.
Data Analysis. The reactions were studied in two phase
reactors consisting of a liquid and a gas volume, denoted V
and V , respectively. The data were evaluated assuming that
i) transformations occurred only in the liquid phase, (ii)
l
g
The mass-based second-order rate constants k′
j
′derived
(
from the model fits were converted to liquid-phase con-
′
′liq
equilibration between the two phases was rapid relative to
the liquid-phase reactions, and (iii) the ratio between the
centration-based rate constants kj assuming that
gas and liquid-phase concentrations, c
Henry’s law constant, H . The total amount of a compound
i, {S }, in the reactor is given by
g
/ c
l
, was given by
-d{S }/ dt ) V d[c ]/ dt
(8)
(9)
i
l
i
c
′′liq
i
-
d[c ]/ dt ) V
∑kj
[c ][C]
i
i
l
{
S } ) c (V + V / H )
(1)
i
gi
g
l
ci
where the edged brackets designate liquid-phase concentra-
tions. On the basis of eqs 1, 4, 8 and 9, relation 10 can be
derived, which relates the mass-based rate constant to the
constant expressing the reaction rate in solution:
Pseudo-first-order and zero-order substrate removals were
expressed in terms of the measured total amount of the
1
2 0 8
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 9, 1998

a
TABLE 1. Rate Laws Given as Sets of ODEs Describing the Rate-Determining Reactions of Acetylene, Chloroacetylene, and
,1-DCE Transformations Catalyzed by cblI b
1
schemes
reactions
ODEc
I
acetylene (Schem e 1)
1, 2
-d{ACT}/dt ) (k′
1
′ + k′
d{vinyl-cbl }/dt ) k′′{ACT}{cbl } - (k′
d{ETH}/dt ) (k′ + k′
-d{CAT}/dt ) (k′′ + k′
d{ethynyl-cbl }/dt ) k′
d{ACT}/dt ) k′′{CAT}{cbl } + (k′
(k′′ + k′′){ACT}{cbl }
d{VC}/dt ) k′′{CAT}{cbl }
-d{1,1DCE}/dt ) (k′ + k′
{1,1DCE}{cbl } - (k′17 + k′18){R-cbl }
2
′){ACT}{cbl }
III
I
III
2
1
, 3, 4
2
3
+ k′
4
){vinyl-cbl }
III
I
, 3, 4, 5
3
4
){vinyl-cbl } + k′
1
′{ACT}{cbl }
I
chloroacetylene (Schem es 1 and 2)
6, 7, 8
6
7
′ + k′
8
′){CAT}{cbl }
III
I
III
8
1
, 9, 10
, 2, 7, 9, 10,
8
′{CAT}{cbl } - (k′
9
+ k′10){ethynyl-cbl }
I
III
7
9
+ k′10){ethynyl-cbl }
I
-
1
2
I
6
, 12
6
I
1
,1-DCE (Schem e 3)
14, 15, 16
1
′
4
15 16
′ + k′′ ){1,1DCE}{cbl }
III
I
III
1
1
1
5, 17, 18
6, 18
4, 17
d{R-cbl }/dt ) k′
d{volatiles}/dt ) k′
d{unknown}/dt ) k′
15
′
I
III
16
′ {1,1DCE}{cbl } + k′18{R-cbl }
I
I
I
I
1
′
4
{1,1DCE}{cbl } + k′17{R-cbl }
a
b
III
Mass-based notation is dependent on reactor configuration. ACT )acetylene, ETH )ethene, vinyl-cbl )vinylcobalam in, CAT )chloroacetylene,
III
III
VC ) vinyl chloride, ethynyl-cbl ) ethynylcobalam in, R-cbl ) interm ediate cobalam in com plex, where R represents an alkyl-, alkenyl-, or alkynyl
group, unknown ) nonvolatile products, proposed interm ediate cobalam in com plex. Steady-state approxim ations were used for radical species.
c
′
j
′liq
′′
j
SCHEME 1
k
) V k f
(10)
l
c
c
where the conversion factor f is defined as
f ) 1 + H (V / V )
(11)
c
c
g
l
For the chlorinated ethenes, the H
taken from the literature (40). For acetylene, ethene, and
ethane, H values were estimated as 0.93, 7.96, and 19.88,
c
values at 22 °C were
c
respectively, using aqueous solubility data measured at 1
atm constant pressure (41). For chloroacetylene, no solubility
data was available. The H
estimated using eq 1. The total amount of chloroacetylene,
CAT}, was obtained from transformation data under condi-
c
value of chloroacetylene was
{S
tions where chloroacetylene was assumed to be quantitatively
converted into acetylene and vinyl chloride (shown below).
Gas-phase concentrations cgCAT were estimated assuming that
acetylene and chloroacetylene have the same FID response
as predicted by the kinetic model which only considers the
addition of cbl (reaction 2) as the initial transformation. To
account for the rapid appearance of ethene, the pathway
that includes SET (reaction 1) followed by reduction of the
radical (reaction 5) was postulated.
I
factors. The estimated H
The following f values were used: f
DCE) ) 1.14, f (trans-DCE) ) 1.34, f (1,1-DCE) ) 1.97, f
chloride) ) 2.01, f (acetylene) ) 1.93, and f (chloroacetylene)
2.18.
c
value of chloroacetylene was 1.2.
(TCE) ) 1.34, f (cis-
(vinyl
c
c
c
c
c
c
c
c
)
Fitting the data shown in Figure 1 to the rate laws of Table
1
resulted in the kinetic constants (converted to liquid-phase
Results and Discussion
Acetylene. The kinetic data and the model fits of the
acetylene transformation experiment is shown in Figure 1.
The proposed acetylene transformation pathway is indicated
in Scheme 1. During the first 22 h, the acetylene mass
decreased relatively rapidly by 0.28 µmol, nearly the total
amount of cbl added (0.32 µmol). After 22 h, ethene
accounted only for 11% of the acetylene transformed. The
unaccounted mass was tentatively attributed to the formation
conditions) summarized in Table 2. The initial transforma-
tion was reported to be first order in the proton concentration
(11, 12), and k′
vinylcobalamin decomposition (via reactions 3 and/ or 4) is
represented by the sum of k′ and k′. These two constants
1 2
′and k′′are expected to be pH dependent. The
3
4
cannot be evaluated individually with the available data.
Although the contribution of SET to the overall ethene
formation appears to be small (k
1 2
′′ , k′ ′) , the formation of
III
of vinyl-cbl . Thereafter, the acetylene mass decreased at
radicals in biological systems could potentially lead to
toxicological effects.
-
3
a constant rate [kobs ) (0.80 ( 0.07) × 10 µmol/ h] while the
total mass of volatiles (acetylene + ethene) remained
constant, indicating quantitative transformation of acetylene
into ethene. These observations are qualitatively consistent
Chloroacetylene. The kinetics of the chloroacetylene
transformation was studied in the presence of 0.32 µmol of
I
cbl at two different initial chloroacetylene amounts. In the
I
with a kinetic model in which cbl catalyzes the transformation
first case, initial chloroacetylene amount was 2.05 µmol, 6.4
III
I
of acetylene into ethene via vinyl-cbl as the relatively stable
times in excess of the cbl concentration. In the second case,
I
intermediate (reaction 2). The time profiles for free cbl and
the initial chloroacetylene amount was 0.54 µmol, ap-
III
I
vinyl-cbl were not measured but were calculated using the
proximately equal to the amount of cbl . The results of these
fitted k′
1
′, k′
2
′and k′
3
+ k′
4
, assuming that the missing mass was
two experiments are shown in Figures 2 and 3, respectively.
The proposed pathway is shown in Scheme 2.
III
equal to the mass of vinyl-cbl . According to this model,
I
cbl decreases rapidly during the first 22 h and reaches a
In the experiment shown in Figure 2, two distinct reaction
regimes are evident: chloroacetylene decreased rapidly for
the first 20 min and more slowly thereafter. During the rapid
chloroacetylene decline, the total mass of volatiles declined
by 0.44 µmol and then remained constant. Of this mass loss,
0.32 µmol was tentatively attributed to the formation of
III
steady state after 40 h. Vinyl-cbl can decompose into to
ethene and the free catalyst via reaction 3 or 4. Close
examination of the initial ethene formation rate, however,
suggests that a second acetylene transforming pathway exists
besides the addition of cbl to acetylene form vinyl-cbl .
Figure 1b shows model simulations that include or exclude
the SET reaction. The appearance of ethene is immediate
and not delayed by the formation of a stable intermediate
I
III
III
ethynyl-cbl (reaction 8). Acetylene and vinyl chloride were
formed without an apparent delay, suggesting a pathway
with very reactive intermediates. The unaccounted mass
VOL. 32, NO. 9, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1 2 0 9

chloride (10%). Interestingly, vinyl chloride was formed only
when chloroacetylene was present. Acetylene appears to be
III
formed via reductive cleavage of ethynyl-cbl , presumably
with titanium(III) citrate as the reductant. Three possible
initial transformation reactions can be proposed: (i) SET
and protonation (reaction 6) to yield the vinyl chloride radical,
(
ii) SET concomitant with reductive chloride cleavage to form
the ethynyl radical (reaction 7), and (iii) nucleophilic chloride
I
III
substitution with cbl to form the ethynyl-cbl (reaction 8).
III
The ethynyl-cbl may react further via reaction 9 forming
acetylene or via reaction 10 forming the ethynyl radical which
can be reduced to acetylene. The proposed vinyl chloride
formation involves two sequential SETs (reactions 6 and 12)
although other pathways are also conceivable. For instance,
III
a short-lived chlorovinyl-cbl intermediate may form which
is then reductively cleaved (not shown in Scheme 2).
The experimental data shown in Figures 2a and 3a were
fitted to the kinetic model considering three parallel (pseudo)-
second-order chloroacetylene removal reactions (reactions
III
6
, 7, and 8), the slow transformation of the ethynyl-cbl into
acetylene, and the formation of vinyl chloride via reaction
III
6
(Table 1). The reaction of ethynyl-cbl to acetylene via
reactions 9 and 10 was modeled by a pseudo-first-order
process with a rate constant of k′ + k′10. In reactions 9 and
0, titanium(III) citrate is assumed to be the reductant. The
9
1
I
time profiles of the three postulated cobalamin species, cbl ,
ethynylcobalamin, and vinylcobalamin were calculated based
on mass balance considerations, assuming rapid reduction
II
I
of cbl to cbl , assuming rapid mass transfer of the volatiles
between the liquid and gaseous phases, and incorporating
the model parameters derived from the acetylene transfor-
mation experiments (Figure 2b and 3b). To quantitatively
compare the model results, the different kinetic parameters
obtained from fitting the data were converted to liquid-phase
rate constants as described above. The kinetic constants
FIGURE 1. Reductive transformation of 0.98 µmol of acetylene in
the presence of 0.32 µmol of cobalamin, under standard conditions
[
15 mM Ti(III) citrate at pH 8]. (a) Lines represent least-mean-square
data fits constrained to the kinetic model indicated in Scheme 1
and rate laws indicated Table 1. The experimental data originate
(
converted to liquid-phase conditions) are summarized in
I
from replicate experiments. Cbl and vinylcobalamin profiles were
Table 3. Although the model was able to reflect the time
course of the different species that were measured, some
significant discrepancies are evident: (1) the initial amounts
of chloroacetylene obtained by fitting were approximately
determined from the constrained model calculations. Total volatiles
)
sum of acetylene and ethene in the reactor (liquid and gas phases).
(
b) Model fit for ethene formation with and without SET (reaction
1).
1
0% below the analytically determined amounts; (2) the rate
(
6%) is tentatively attributed to the formation of unknown
constants for the two chloroacetylene transformation ex-
periments disagreed for all reactions except the transforma-
tion of ethynylcobalamin to acetylene (k′ + k′10). These
discrepancies are unexplained but could be due to rate-
limiting gas-liquid equilibration, slow cbl regeneration or
cbl concentration-dependent decomposition of intermediate
cobalamin complexes. The reported chloroacetylene trans-
formation rates are, therefore, system dependent and
products during the initial phase of the reaction. The
formation of ethynyl-cbl , the concomitant decrease in cbl ,
and the slow formation of vinyl- cbl predicted by the kinetic
model is indicated in Figure 2b. After 20 min, the mass of
total volatiles remained constant, suggesting quantitative
conversion of chloroacetylene into acetylene and vinyl
chloride. During this time, acetylene reacted slowly to vinyl-
III
I
9
III
I
I
III
cbl as discussed above. Ethene was not detected. The final
7 6 2 1
considered “apparent”. The ratios k′ ′/ k′′ and k′ ′/ k′′ are
yields of acetylene and vinyl chloride were 66 and 10%,
respectively. The proposed pathway is consistent with data
reported by Johnson et al., who suggested the formation of
ethynyl-cbl in reactions of cbl with bromoacetylene (23).
The data shown in Figure 3a provides additional kinetic
evidence for the pathway proposed in Scheme 2. Here, the
total volatiles decreased rapidly by 0.34 µmol by the time of
the first measurement (6 min) consistent with the rapid
approximately 0.5, and a ratio of 0.025 for chloroacetylene
and acetylene, respectively, suggests chloroacetylene reacts
more readily via SET than acetylene. This is consistent with
the Marcus theory which predicts increased reduction rates
with increasing one-electron reduction potentials, i.e., degree
of halogenation (25). The formation of chloroacetylene in
cells and its formation of radicals is likely to lead to toxic
effects and could therefore be significant.
III
I
III
formation of ethynyl-cbl . After 6 min, two different reaction
I
regimes are evident. First, while chloroacetylene was present,
the total molar mass of volatiles remained constant, indicating
that chloroacetylene was quantitatively converted into
acetylene and vinyl chloride. These conditions correspond
to the phase in the above experiment where chloroacetylene
is slowly declining. After 1 h, when chloroacetylene was
completely transformed, vinyl chloride remained constant
at 0.04 µmol, while acetylene formation continued. After 5
h, traces of ethene were observed (data not shown).
Taken together, these observations indicate that chlo-
roacetylene is transformed into acetylene (66%) and vinyl
As shown in Figure 2b and 3b, cbl is rapidly bound as
III
I
ethynyl-cbl thus lowering the overall activity of cbl . Over
time, as acetylene is formed, it reacts to vinyl-cbl . Because
vinyl-cbl is more stable than ethynyl-cbl (by a factor of
120-185) with time, the former is becoming a more significant
deactivating species. The model predicts steady-state con-
ditions are reached in 25 h with residual amounts of 0.05
III
III
III
I
µmol of cbl (results not shown). This is in agreement with
the kinetic model for the acetylene transformation. At the
low chloroacetylene amounts, the steady-state conditions
are reached after approximately 60 h.
1
2 1 0
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 9, 1998

TABLE 2. Liquid-Phase Rate Constants Based on Kinetic Parameters k′
j j
′and k′ Constrained to the Kinetic Models Describing
Acetylene, Chloroacetylene, and 1,1-DCE Transformation (Standard Conditions: 15 mM Ti(III) Citrate, pH 8)
reactants and init. liquid-phase conc.
reaction path
rate constantsa,b (liquid phase)
′′liq
) (0.3 ( 0.3) × 10-3 µM-1 h-1
) (12.2 ( 1.3) × 10-3 µM-1 h-1
1
1
6.0 µM acetylenec
0.0 µM cobalam in
acetylene f ethenyl radical
acetylene f vinylcobalam in
vinylcobalam in f ethene
k
k
(
1
′
′liq
2
′
liq
+ k′4 ) ) (2.9 ( 0.3) × 10 h-1
liq
-3
k
3
′′liq
(app) ) (0.4 ( 0.2) × 10 µM h-1
-1
-1
2
1
9.4 µM chloroacetylenec
0.0 µM cobalam in
chloroacetylene f chlorovinyl radical
chloroacetylene f ethynyl radical
chloroacetylene f ethynylcobalam in
ethynylcobalam in f acetylene
chloroacetylene f chlorovinyl radical
chloroacetylene f ethynyl radical
chloroacetylene f ethynylcobalam in
ethynylcobalam in f acetylene
1,1-DCE f unknown
k
k
k
6
′
′liq
(app) ) (1.6 ( 0.5) × 10 µM h-1
-1
-1
7
′
′liq
(app) ) (1.5 ( 0.8) × 10 µM h-1
-1
-1
8
′
liq
+ k′11 ) ) (5.4 ( 1.4) × 10 h-1
(app) ) (4.9 ( 1.9) × 10-1 µM-1 h-1
(app) ) (10.3 ( 3.1) × 10-1 µM-1 h-1
liq
-1
(
k
10
′′liq
6
′
7
′
8
7
1
.7 µM chloroacetylenec
0.0 µM cobalam in
k
k
k
′liq
′liq
(app) ) (51.6 ( 20.1) × 10 µM h-1
-1
-1
′
liq
0
liq
-1 -1
(
k
+ k′11 ) ) (3.4 ( 0.6) × 10
h
1
′′liq
1
1
2.7 µM 1,1-DCEc
0.0 µM cobalam in
-3
-1 -1
k
k
k
k
k
k
k
k
k
k
) 8.8 × 10 µM
h
1
4
1,1-DCE f R-cblIII
′′liq
) (9.4 ( 2.1) × 10 µM-1 h-1
-3
1
5
′
′liq
) 2.4 × 10 µM h-1
-3
-1
1
,1-DCE f volatiles
1
′liq
1
′liq
1
′′liq
1
′′liq
6
R-cblIII f unknown
R-cblIII f volatiles
1,1-DCE f unknown
1,1-DCE f R-cblIII
-3 -1
) 7.3 × 10
h
7
) 2.5 × 10 h-1
-3
8
2
1
00.0 µM 1,1-DCEc
00.0 µM cobalam in
) 1.7 × 10 µM h-1
-3
-1
4
) 2.0 × 10 µM h-1
-3
-1
1
5
′
′liq
) 0.4 × 10 µM h-1
-3
-1
1
,1-DCE f volatiles
1
′liq
6
III
R-cbl f unknown
R-cbl f volatiles
-3 -1
) 3.6 × 10
) 3.2 × 10
h
1
7
III
′liq
-3 -1
h
1
8
a
. b Errors represent two standard deviations
The second-order kinetic param eters were converted using the corresponding factors f
c
and V
l
c
2
derived from the m odel fits. Om itted error lim its are >65%. Correlation coefficient of m odel fit: r
m
> 0.99.
FIGURE 3. (a) Reductive transformation of 0.54 µmol of chloro-
acetylene in the presence of 0.32 µmol of cobalamin, under standard
conditions [15 mM Ti(III) citrate at pH 8]. Lines represent model fits
to the averaged data of replicate experiments (Scheme 2) including
the kinetic coefficients of the acetylene transformation (Scheme
FIGURE 2. (a) Reductive transformation of 2.05 µmol of chloro-
acetylene in the presence of 0.32 µmol of cobalamin, under standard
conditions [15 mM Ti(III) citrate at pH 8]. Lines represent least-
mean-square data fits constrained to the kinetic model (Scheme
2
) including the kinetic coefficients of the acetylene transformation
1
) shown in Table 2. Total volatiles ) sum of acetylene, chloro-
(
Scheme 1) shown in Table 2. Total volatiles ) sum of acetylene,
I
acetylene, and vinyl chloride in the reactor. (b) Cbl and ethynyl-
I
chloroacetylene, and vinyl chloride in the reactor. (b) Cbl and
III
cbl profiles were determined from the model calculation based
ethynylcobalamin profiles were determined from the model cal-
culation based on the initial chloroacetylene amount of 1.82 µmol.
on the initial chloroacetylene amount of 0.49 µmol (kinetic
coefficients for acetylene transformation were 90% of the values
in Table 2).
cis- and trans-DCE. The transformation of cis- and trans-
DCE isomers was studied under standard conditions. In the
first experiment, the initial amounts of cis- and trans-DCE
VOL. 32, NO. 9, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1 2 1 1

SCHEME 2
TABLE 3. Product Distributions of the Reductive
Transformation of the DCE Isomers Catalyzed by Cob(I)Alamin
(Standard Conditions: 15 mM Ti(III) Citrate, pH 8)
final conver-
a
a
reactants and
initial amounts
intermediates yields
sion
reaction
time (h)
and products
vinyl chloride
vinyl chloride
vinyl chloride
(%)
2.2
0.9
(%)
13.0
8.8
0
0
0
0
1
3
.8 µm ol of trans-DCE
.32 µm ol of cobalam in
.8 µm ol of cis-DCE
120
120
159
.32 µm ol of cobalam in
3.7 µm ol of cis-DCE
1.4
0.5
19.5
.2 µm ol of cobalam in ethene
ethane
0.05
3.2
1.4
0.1
9.5
6.9
2.0
0.3
1
3
3.4 µm ol of trans-DCE vinyl chloride
.2 µm ol of cobalam in ethene
ethane
27.9
94.1
159
159
1
3
2.6 µm ol of 1,1-DCE
ethene
FIGURE 4. (a) Reductive transformation of 0.8 µmol of 1,1-DCE in
the presence of 0.32 µmol of cobalamin, under standard conditions
.2 µm ol of cobalam in vinyl chloride
acetylene
ethane
[
15 mM Ti(III) citrate at pH 8]. Lines represent least-mean-square
c
data fits constrained to the kinetic model (Scheme 3). The exp-
erimental data were obtained from replicate experiments. (b)
Nonvolatile unknown products, cbl and R-cbl profiles, were
determined from the model calculation based on the given initial
conditions. Total volatiles is the sum of ethene, vinly chloride,
acetylene, and ethane. R represents an alkyl, alkenyl-, or alkynyl
group.
nonvolatile
75.4
6.3
1.9
1.8
8.7
0
0
.8 µm ol of 1,1-DCE
ethene
98.5
121
I
III
.32 µm ol of cobalam in vinyl chloride
acetylene
ethane
nonvolatile
79.9c
a
If not otherwise stated, values are derived from m easurem ents at
the indicated tim e. b Values derived from m odel calculations. Total
c
m olar m ass loss derived from experim ents.
SCHEME 3
were 13.7 µmol and in the second initial amounts were 0.8
µmol. Conditions, conversion rates, products, and yields
are summarized in Table 3. Both cis-DCE and trans-DCE
transformed very slowly, mainly to vinyl chloride. After 5
days, substrate conversions were 8.8 and 13%, respectively,
with yields of vinyl chloride of 0.9 and 2.2%, respectively.
When the amounts of cobalamin, cis- and trans-DCE were
increased to 3.2, 13.7, and 13.4 µmol, respectively, small
amounts of ethene and ethane were detected in addition to
vinyl chloride. The unaccounted mass is tentatively at-
tributed to alkyl-, alkenyl-, or alkynylcobalamin intermedi-
ates. Alkyl- and vinylcobalamins have been found recently
in similar transformation studies (11). Vinyl chloride has
disappearance was first order during the first 20 h [k′obs
)
-
3
-1
(14.4 ( 1.8) × 10
h
] (Figure 5a). Thereafter, the rate
slowed, indicating catalyst inhibition, probably due to the
formation of relatively stable intermediates binding cbl.
Approximately 20% of the 1,1-DCE reacted to volatiles (Table
3). All observed products formed concurrently and without
apparent delay (data not shown) suggesting that their
pathways did not involve stable intermediate. After 121 h,
75-80% of the 1,1,-DCE was transformed into nonvolatile
products.
III
been shown to react to ethene with vinyl-cbl and ethene-
cobalamin π-complex as possible intermediates (11). Ethane
was found previously in similar experiments (11) and was
speculated to form via hydrogenation of the ethene-cobal-
amin π-complexes. Acetylene was not detected.
1
,1-DCE. The transformation of 0.8 µmol of 1,1-DCE
Since the major portion of products formed is unknown,
mechanistic conclusions are difficult. The kinetic model
shown in Scheme 3 summarizes the apparent transforma-
tions. According to this scheme, 1,1-DCE is concurrently
transformed into unknown nonvolatile products (reaction
initiated by 0.32 µmol of cbl was relatively rapid and complete
within approximately 121 h. The data and the model fits are
indicated in Figure 4, and Scheme 3 summarizes the pathway.
1
1
,1-DCE transformation was first-order rate [k′obs ) (33.9 (
-
3
-1
III
.6) × 10
h
]. Deactivation of cbl was not apparent. The
14), R-cbl compounds (reaction 15) (R represents an alkyl-,
volatile products formed were vinyl chloride, ethene, ethane,
and acetylene (Table 3) and accounted for 19% of the added
alkenyl-, or alkynyl group) and volatiles (reaction 16). The
volatiles include ethene, vinyl, chloride, and acetylene. Rapid
II
I
1
,1-DCE. When the initial amounts of cbl and 1,1-DCE were
increased to 3.2 and 12.6 µmol, respectively, 1,1-DCE
reduction of cbl to cbl is assumed. The model is indicated
in Tables 1 and 3 and Figures 4 and 5. The rate constants
1
2 1 2
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 9, 1998

Literature Cited
(
(
(
(
(
(
(
(
(
1) Mohn, W. W.; Tiedje, J. M. Microbiol. Rev. 1992, 56 (3), 482-
07.
2) Schanke, C. A.; Wackett, L. P. Environ. Sci. Technol. 1992, 26 (4),
30-833.
3) Holliger, C.; Schraa, G.; Stupperich, E.; Stams, A. J. M.; Zehnder,
A. J. B. J. Bacteriol. 1992, 174 (13), 4427-4434.
4) Gantzer, C. J.; Wackett, L. P. Environ. Sci. Technol. 1991, 25 (4),
5
8
7
15-722.
5) Schumacher, W.; Holliger, C.; Zehnder, A. J. B.; Hagen, W. R.
FEBS Lett. 1997, 409 (3), 421-425.
6) Schumacher, W.; Holliger, C. J. Bacteriol. 1996, 178 (8), 2328-
2
333.
7) Neumann, A.; Wohlfarth, G.; Diekert, G. J. Biol. Chem. 1996,
71, 16515-16519.
8) Neumann, A.; Wohlfarth, G.; Diekert, G. Arch. Microbiol. 1995,
63, 276-281.
9) Terzenbach, D. P.; Blaut, M. FEMS Microbiol. Lett. 1994, 123,
13-218.
10) Jablonski, P. E.; Ferry, J. G. FEMS Microbiol. Lett. 1992, 96, 55-
0.
2
1
2
(
(
(
(
(
(
(
(
(
(
6
11) Glod, G.; Brodmann, U.; Angst, W.; Holliger, C.; Schwarzenbach,
R. P. Environ. Sci. Technol. 1997, 31 (11), 3154-3160.
12) Glod, G.; Angst, W.; Holliger, C.; Schwarzenbach, R. P. Environ.
Sci. Technol. 1997, 31 (1), 253-260.
13) Burris, D. R.; Delcomyn, C. A.; Smith, M. H.; Roberts, A. L.
Environ. Sci. Technol. 1996, 30 (10), 3047-3052.
14) Chiu, P.-C.; Reinhard, M., Environ. Sci. Technol. 1996, 30 (6),
1
882-1889.
15) Lewis, T. A.; Morra, M. J.; Brown, P. D. Environ. Sci. Technol.
996, 30 (1), 292-300.
16) Chiu, P.-C.; Reinhard, M. Environ. Sci. Technol. 1995, 29 (3),
95-603.
1
5
17) Krone, U. E.; Thauer, R. K.; Hogenkamp, H. P. C.; Steinbach, K.
Biochemistry 1991, 30 (10), 2713-2719.
18) Krone, U. E.; Thauer, R. K.; Hogenkamp, H. P. C. Biochemistry
1
989, 28 (11), 4908-4914.
FIGURE 5. (a) Reductive transformation of 12.6 µmol of 1,1-DCE in
the presence of 3.2 µmol of cobalamin, under standard conditions
19) Schrauzer, G. N.; Deutsch, E.; Windgassen, R. J. J. Am. Chem.
Soc. 1968, 90 (9), 2441-2442.
[
15 mM Ti(III) citrate at pH 8]. Lines represent least-mean-square
(20) Pontius, F. W. J. Am. Water Works Assoc. 1995, 87, 48-58.
(21) Zhou, D.-L.; Walder, P.; Scheffold, R.; Walder, L. Helv. Chim.
Acta 1992, 75, 995-1011.
data fits constrained to the kinetic model (Scheme 3). (b) Unknown
I
III
products, cbl and R-cbl profiles, were determined from the model
calculation based on the given initial conditions. Total volatiles is
the sum of ethene, vinly chloride, acetylene, and ethane. R represents
an alkyl, alkenyl-, or alkynyl group.
(
22) Schrauzer, G. N.; Deutsch, E. J. Am. Chem. Soc. 1969, 91 (12),
341-3350.
23) Johnson, A. W.; Mervyn, L.; Shaw, N.; Smith, E. L. J. Chem. Soc.
963, 4146-4156.
3
(
1
(
(
24) Sav e´ ant, J.-M. Adv. Phys. Org. Chem. 1990, 26, 1-130.
25) Eberson, L. In Electron-Transfer Reactions in Organic Chemistry;
Springer: Berlin, 1987.
26) Zhou, D.-L.; Tinembart, O.; Scheffold, R.; Walder, L. Helv. Chim.
Acta 1990, 73, 2225-2241.
derived from the experimental data with higher initial cblI
and 1,1-DCE concentrations were smaller by a factor of 4.5-
(
6
, except for the rate constants of the R-cblIII decomposition
reactions, which were comparable. The difference in reac-
tivity and product distributions of the cis- and trans-DCE
isomers and 1,1-DCE merits further investigation. Perhaps,
(27) Finke, R. G.; Martin, B. D. J. Inorg. Biochem. 1990, 40, 19-22.
(
(
(
28) Shepherd, R. E.; Zhang, S.; Dowd, P.; Choi, G.; Wilk, B.; Choi,
S.-C. Inorg. Chim. Acta 1990, 174, 249-256.
29) Tinembart, O.; Walder, L.; Scheffold, R. Ber. Bunsen-Ges. Phys.
Chem. 1988, 92, 1225-1231.
30) Ramakrishna, D. N.; Symons, M. C. R. J. Chem. Soc., Faraday
Trans. 1 1983, 79, 269-279.
1
,1-DCE undergoes SET reactions much more readily than
the cis- and trans-DCE isomers, forming the chlorovinyl
radical which is then further reduced to vinyl chloride (as
indicated in reaction 12). Glod et al. suggest that the different
alkylcobalamins formed with 1,1-DCE react into polar
(31) Lexa, D.; Saveant, J.-M. J. Am. Chem. Soc. 1978, 100 (10), 3220-
222.
32) Martin, B. D.; Finke, R. G. J. Am. Chem. Soc. 1992, 114 (2), 585-
92.
33) Bakac, A.; Espenson, J. H. Inorg. Chem. 1987, 26, 4305-4307.
3
(
(
2
products such as acetaldehyde (11). Highly reactive C -
5
carbenes stabilized by cobalamin coordination may lead to
the formation of acetylene through a 1,2-hydrogen shift, yield
ethene by the addition of hydrogen, and may account for the
formation of water-soluble products such as acetaldehyde.
(34) Schrauzer, G. N.; Grate, J. H. J. Am. Chem. Soc. 1981, 103 (3),
541-546.
(
35) Roberts, A. L.; Totten, L. A.; Arnold, W. A.; Burris, D. R.; Campbell,
T. J. Environ. Sci. Technol. 1996, 30 (8), 2654-2659.
36) Miller, S. I.; Lee, W. G. J. Am. Chem. Soc. 1959, 81, 6313-6319.
37) Denis, J.-N.; Moyano, A.; Greene, A. E. J. Org. Chem. 1987, 52,
(
(
Acknowledgments
3
461-3462.
Funding for this study was provided by the Office of Research
and Development, U.S. Environmental Protection Agency
under agreement R-819751-01 through the Western Region
Hazardous Substance Research Center. A scholarship was
provided for Marco Semadeni from the Swiss National
Science Foundation, Bern, Switzerland. The content of this
study does not necessarily represent the views of the agencies.
We thank Cindy Krieger for reviewing the manuscript and
David Burris et al. for sharing a prepublication copy of ref
(
38) Strang, P. J.; Diederich, F. In Modern Acetylene Chemistry; VCH
Verlagsgesellschaft: Weinheim, 1995; pp 48-66.
(39) Kloster-Jenson, E. Tetrahedron 1971, 27, 33-49.
(40) Gossett, J. M. Environ. Sci. Technol. 1987, 21 (2), 202-208.
(
41) Air Liquide. Gas Encyclopaedia; Elsevier: Amsterdam, 1976.
Received for review June 12, 1997. Revised manuscript re-
ceived January 5, 1998. Accepted January 29, 1998.
1
3.
ES9705248
VOL. 32, NO. 9, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1 2 1 3