chlorine atom initiated oxidation of chlorinated ethenes: Results for 1,1-dichloroethene (H2C=CCl2), 1,2-dichIoroethene (HClC=CClH), trichloroethene (HClC=CCl2), and tetrachloroethene (Cl2C=CCl2)

Article abstract of DOI:10.1021/jp983583w

Experiments have been carried out on the oxidation of the four chlorinated ethenes H₂C=CCl₂, HClC=CClH, HClC=CCl₂, and Cl₂C=CCl₂. Reaction was initiated by continuous photolysis of Cl₂, and

Full text of DOI:10.1021/jp983583w

J. Phys. Chem. A 1999, 103, 2031-2043
2031
Chlorine Atom Initiated Oxidation of Chlorinated Ethenes: Results for 1,1-Dichloroethene
(H2CdCCl2), 1,2-Dichloroethene (HClCdCClH), Trichloroethene (HClCdCCl2), and
Tetrachloroethene (Cl2CdCCl2)
Alam S. Hasson and Ian W. M. Smith*
The School of Chemistry, The UniVersity of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom
ReceiVed: September 1, 1998; In Final Form: December 8, 1998
Experiments have been carried out on the oxidation of the four chlorinated ethenes H2CdCCl2, HClCd
CClH, HClCdCCl2, and Cl2CdCCl2. Reaction was initiated by continuous photolysis of Cl2, and product
yields were measured using Fourier transform infrared (FTIR) spectroscopy. In all cases, experiments were
performed at different concentrations of the chlorinated ethenes (down to 3 × 1013 molecule cm-3), Cl2 and
O2. In the cases of H2CdCCl2 and HClCdCCl2, the product yields were invariant with these changes, consistent
with them being determined by competitive unimolecular dissociations of the chlorinated alkoxy radical; that
is, H2ClCCCl2O (HCl2CCCl2O) (+M) f H2ClCCOCl (HCl2CCOCl) + Cl (+M) and H2ClCCCl2O (HCl2-
CCCl2O) (+M) f COCl2 + CH2Cl (CHCl2) (+M), and the subsequent formation of HCOCl and/or COCl2
from CH2Cl and CHCl2. In the case of HCl2CCHClO formed from HClCdCClH, H atom abstraction by O2
is competitive with CCl and CC bond cleavage, so the yields depended on the concentration of O2. With
C2Cl4, the only products observed were CCl3COCl and COCl2, consistent with competitive dissociations of
perchloroethyl radicals: CCl3CCl2O (+M) f CCl3COCl + Cl (+M); CCl3CCl2O (+M) f COCl2 + CCl3
(+M). However, the relative yields of CCl3COCl and COCl2 were found to depend on the initial concentration
of C2Cl4 which is incompatible with the simple, and generally accepted, mechanism. To investigate this
unexpected result further, experiments were performed on pentachloroethane (C2Cl5H). Again, the product
yields depended on the initial concentration of the chlorinated compound. In addition, product yields from
C2Cl4 were measured using two different rates of photolysis of Cl2, at three temperatures, 298, 353, and 393
K, and in the presence of added NO and added HCl. Although the experimental results could be modeled
when bimolecular reactions of perchlorinated methoxy and ethoxy radicals with C2Cl4 and HCl were included
in the reaction mechanisms, the values of the rate constants which were needed seem unrealistically large if
the currently accepted rate for the unimolecular decomposition of CCl3CCl2O is correct. This observation on
C2Cl4 is important since it casts doubt on the current view of the oxidative chemistry of this compound,
which is released in significant amounts into the atmosphere.
1. Introduction
and hydrolysis of CCl3COCl has been postulated as an
atmospheric source of trichloroacetic acid, a chemical that is
known to be phytotoxic.6 Another recognized product of C2Cl4
oxidation is COCl2, which is removed from the troposphere by
uptake into water droplets followed by hydrolysis,7,8 probably
forming HCl and CO2.9 In addition, it has been suggested that
CCl4, a species with a significant ozone depletion potential,
might be formed by the degradation of CCl3COCl in the
atmosphere.10-12
Perchloroethene is one of the most widely used chlorinated
solvents, with an annual global release of 295 ktons/year in
1992.1 Its main uses are in dry cleaning and metal degreasing.
The atmospheric lifetime of C2Cl4 has been estimated to be ca.
0.4 years,2 on the basis that the major loss is its oxidation
initiated by reaction with OH radicals.3
Cl2CCCl2 + OH (+M) f Cl2CCCl2OH (+M)
(R0d)
Cl atoms have never been directly observed in the tropo-
sphere, and their concentrations have been the object of much
speculation. Clearly, a concentration of more than about 102-
103 molecule cm-3 would be required for reaction R2d to occur
at a significant fraction of the rate of reaction R0d. Based on
atmospheric models and on the finding that the observed
abundance of C2Cl4 is consistent with its removal only by
reaction with OH radicals through reaction R0d, recent estimates
have put the average tropospheric concentration of Cl atoms at
<103 molecule cm-3 and have therefore cast doubt on the
potential significance of Cl atoms as a tropospheric oxidant.3,13,14
Other estimates, summarized by Graedel and Keene,15 suggest
that concentrations as high as 105-106 molecule cm-3 may
occur at least on occasions in the marine boundary layer.
However, because C2Cl4 combines with Cl atoms
Cl2CCCl2 + Cl (+M) f Cl3CCCl2 (+M)
(R2d)
approximately 300 times faster than with OH radicals,4 there
has been some discussion in the literature as to whether reaction
R2d can play a significant role in initiating the tropospheric
oxidation of perchloroethene. (Throughout the text of this paper,
lower case letters a, b, c, d, and d′ are used to identify elementary
reactions in the oxidation of H2CdCCl2, HClCdCClH, HClCd
CCl2, Cl2CdCCl2, and CCl3CCl2H, respectively.) Oxidation
initiated by Cl atoms (as opposed to OH radicals) is known to
result in the formation of CCl3COCl as one of its products,5
10.1021/jp983583w CCC: $18.00 © 1999 American Chemical Society
Published on Web 03/13/1999

2032 J. Phys. Chem. A, Vol. 103, No. 13, 1999
Hasson and Smith
Although trichloroethene is sometimes used industrially as a
solvent for degreasing metals and in dry cleaning whilst 1,1-
dichloroethene is employed in the manufacture of polymers and
of 1,1,1-trichloroethane, partially chlorinated ethenes only
comprise a small fraction of the total atmospheric budget of
compounds containing chlorine.15 Consequently, their atmo-
spheric chemistry is, at most, of slight environmental signifi-
cance. Nevertheless, understanding their oxidation is of funda-
mental chemical importance, and should provide insight into
the oxidation of tetrachloroethene.
path length was set at either 20 or 40 m, depending on the
concentrations of reactants used, and spectra were recorded by
coadding 32 scans at a resolution of 2 cm-1, the data collection
taking 16 s. The time scales of most experiments varied from
as short as 5 min up to 1.5 h, each mixture being photolyzed
until reaction was complete. In experiments on C2Cl5H in the
presence of NO, the period of photolysis was extended to 15 h.
The chlorinated ethenes were all supplied by Aldridge with
stated purities of H2CCCl2 (99%), HClCCClH (98%), HClCCCl2
(99%), and Cl2CCCl2 (99.5%). They and C2Cl5H (Lancaster
Synthesis, 95%) were degassed by successive pump-freeze-
thaw cycles, impurities in their vapors being spectroscopically
undetectable. In the case of HClCdCClH, a 1:1 mixture of cis-
and trans-1,2-dichloroethenes was used. Initial concentrations
of the chlorinated species in the range 3 × 1013 to 1 × 1015
molecule cm-3 were used in the experiments. The gas mixture
in the cell was made up to a total pressure of 700 Torr using
Cl2 (Matheson, 99%), O2 (BOC Ltd.), and N2 (BOC Ltd.,
“oxygen free”). In addition to these “reagent” chemicals, it was
necessary to have samples of a number of other species to
calibrate the sensitivity of the FTIR to the product molecules.
COCl2 (Argo International Ltd., 99%), CCl3COCl (Aldrich,
99.5%), and HCl (BDH, 99.99%) were commercially available.
Dichloroacetaldehyde, CHCl2CHO, was prepared by the method
of Crampton et al.24 Attempts to prepare formyl chloride,
HCOCl, by the method of Staab and Datta25 were unsuccessful,
so yields of this species were determined using the absorption
coefficients determined for similar conditions by Libuda et al.,26
assuming the Beer-Lambert law.
The concentrations of the reagent and product species present
in the reaction cell were estimated from the recorded spectra
using calibrations of absorption strength with concentration at
the following wavenumbers: H2CCCl2 (1165-1062 cm-1),
HClCCClH (864.5 and 830 cm-1), HClCCCl2 (970-870 cm-1),
Cl2CCCl2 (940-885 cm-1), C2Cl5H (782 cm-1), CHCl2COCl
(1120-1020 cm-1), CH2ClCOCl, (760-690 cm-1), CHCl2CHO
(1761 cm-1), CCl3COCl (1816 cm-1 and 775-710 cm-1),
HCOCl (1783 cm-1), COCl2 (1831 cm-1 and 885-785 cm-1),
HCl (2821 cm-1), and CO (2177 cm-1).
This paper reports a comprehensive study of the Cl atom
initiated oxidation of the four chlorinated ethenes H2CdCCl2,
HClCdCClH, HClCdCCl2, and Cl2CdCCl2 at a total pressure
of 700 Torr with N2 as the diluent gas. In all cases, the
dependence of the product yields on the initial concentration
of the chlorinated compound (down to 3 × 1013 molecule cm-3),
Cl2, and O2 was examined. The generally accepted mechanism,
which is described in detail in section 3 below, predicts product
distributions which are independent of changes in the initial
concentrations of these reagents. Because of unexpected findings
in the case of C2Cl4, its oxidation was further examined by (a)
using two different rates of photolysis of Cl2; (b) performing
experiments at three temperatures, 298, 353, and 393 K, and at
298 K at three different total pressures, 700, 140, and 50 Torr;
(c) comparing the results obtained with those from experiments
on the Cl atom initiated oxidation of C2Cl5H; and (d) observing
the effects on the oxidation of C2Cl4 of adding HCl and NO.
The experiments on the oxidation of the partially chlorinated
ethenes and on tetrachloroethene are reported and discussed in
sections 4 and 5 of this paper, respectively. A number of
previous laboratory studies have been carried out to identify
the product yields of the Cl atom initiated oxidation of H2Cd
CCl2,12,16 HClCdCClH,17 HClCdCCl2,12,18 Cl2CdCCl2,12,16,19,20,21
and C2Cl5H,12,22 although several of these studies used partial
pressures of Cl2 and the chlorinated ethene in excess of 1 Torr
and much greater than is necessary in modern experiments using
FTIR spectroscopy. The results of these studies and the
mechanistic aspects which are common to the oxidation of all
these species are described in section 3. The mechanisms for
all the chlorinated ethenes are closely related and are thought
to be well-understood. Although our results confirm previous
results and interpretation for H2CdCCl2, HClCdCClH, and
HClCdCCl2, our data demonstrate previously undetected
complications in the case of C2Cl4 and hence challenge the view
that the atmospheric chemistry of this compound is well-
understood.
To minimize heterogeneous reactions, the walls of the vessel
are coated with Teflon. A thermostated heating tape was
wrapped round the cell underneath an insulating jacket enabling
experiments to be carried out over the temperature range 298-
393 K. The internal temperature of the reactor was monitored
using a thermocouple positioned within a glass tube protruding
into the cell. Gases were handled in a Pyrex manifold and, to
facilitate rapid mixing, were admitted into the reaction cell via
a narrow drilled tube which runs the length of the chamber. N2
was always added last to bring the total pressure up to 700 Torr.
Once all of the gases had been introduced, they were left in the
dark to ensure complete mixing and equilibration to the
temperature of the cell. At higher temperatures, this period was
ca. 2 h. During this time, spectra were recorded regularly to
confirm that no significant dark reactions were taking place.
2. Experimental Method and Procedures
The experimental apparatus has been described in detail
previously.23 In brief, an FTIR spectrometer (Nicolet Magna
550) is coupled to a 35 L cylindrical reaction cell. Continuous
photolysis is provided by two lamps mounted within the cell.
Two sets of lamps were employed during the present experi-
ments. “Blacklamps” (Phillips TL40N/08, 35 W output) pro-
vided radiation down to 300 nm which overlaps strongly with
the absorption spectrum of Cl2 and results in a relatively rapid
rate of photolysis. In some experiments, the blacklamps were
replaced by conventional fluorescent tubes (Phillips TLD, 35
W output). Their output is limited to wavelengths greater than
400 nm, and consequently, they provided a much slower rate
of Cl2 photolysis.
For each system Cl2/C2ClxHy/O2/N2, several sets of product
yield measurements were taken. Product yields were measured
as a function of the initial concentration of the organic species,
with the initial concentrations of the other two reactants held
constant. This procedure was then repeated for two further sets
of experiments with the initial concentrations of Cl2 and O2
being varied in turn, and the remaining two reactant concentra-
tions unchanged. In addition, product yields were measured at
two different rates of photolysis of Cl2, corresponding to the
two types of photolysis lamps used. Using the fluorescent lamps,
The reaction cell is equipped with White cell mirrors enabling
most species to be detected at concentrations down to ca. 1012
molecule cm-3 by long path length infrared absorption. The

Oxidation of Chlorinated Ethenes
J. Phys. Chem. A, Vol. 103, No. 13, 1999 2033
TABLE 1: Previous Measurements of Product Yields from Chlorine Atom Initiated Oxidation of Chlorinated Ethenes
chlorinated ethene
possible primary products
product yields
0.98
methoda
ref
16
H2CdCCl2
CH2ClCOCl
COCl2
H2CO
HCOCl
UV/IR
CH2ClOH
CHCl2COCl
CHCl2CHClOH
CH2ClCOH
HCOCl
HClCdCClH
1.4 ( 0.05
UV/IR
17
COCl2
0.06 ( 0.008
(304 K; pO2 < 85 Torr
CHCl2OH
CO: 0.54 ( 0.05
0.90
0.91 ( 0.10
0.09 ( 0.02
0.09 ( 0.02
HClCdCCl2
Cl2CdCCl2
CHCl2COCl
vis/GC
UV/FTIR
12
18
COCl2
HCOCl
CHCl2OH
CCl3COCl
COCl2
CCl3COCl:
COCl2
CCl3COCl
COCl2
0.85 ( 0.05
0.30 ( 0.10
2.5:1 at 297 K
3.0:1 at 305 K
0.85 ( 0.05
0.25 ( 0.10
CCl4: 0.003
0.87 ( 0.11
0.27 ( 0.03
0.85 ( 0.05
0.30 ( 0.10
0.85 ( 0.09
0.30 ( 0.04
UV/GC
12
20
16
(354, 373 K; 590 Torr)
UV/IR
(295, 305 K; 6-450 Torr)
UV/IR
(298 K; 5-700 Torr)
CCl3COCl
COCl2
CCl3COCl
COCl2
UV/FTIR
21
12
22
(298 K; 700 Torr)
UV/GC
C2Cl5H
(296 K; 590 Torr)
UV/FTIR
(354, 373 K; 700 Torr)
a UV (vis) denotes continuous ultraviolet (visible) photolysis, IR denotes detection by infrared absorption, FTIR detection by Fourier transform
infrared spectroscopy, and GC detection by gas chromatography.
the time dependence of the loss of reactants and formation of
products could be observed by recording spectra every few
minutes over the course of the entire reaction. With the
blacklamps, the yields were determined from pairs of spectra
taken before illumination with the lamps and several minutes
after reaction was complete.
identical for both of these systems. Where the two carbon atoms
in the chlorinated ethene are not equivalent, Cl adds to the less
chlorinated of the two. Oxygen quickly adds to the chlorinated
ethyl radicals to form peroxy species, which are converted to
chloroethoxy radicals by reaction either with a second peroxy
radical or, if NO is present, with NO.
The product yield dependences on the initial concentrations
[C2Cl4]0 and [C2Cl5H]0 were measured at three temperatures,
298, 353, and 393 K. At 298 K, experiments were carried out
at three different total pressure, 700, 140, and 50 Torr. In
addition, product yields were measured in the Cl2/C2Cl4/O2/N2
system in the presence of added NO and HCl. All experiments
were repeated at least twice, and the final results represent an
average of at least three runs.
H2ClCCCl2, HCl2CCClH, HCl2CCCl2, Cl3CCCl2 +
O2 (+M) f H2ClCCCl2O2, HCl2CCClHO2,
HCl2CCCl2O2, Cl3CCCl2O2 (+M) (R3a,b,c,d)
H2ClCCCl2O2, HCl2CCClHO2, HCl2CCCl2O2,
Cl3CCCl2O2 + RO2 (NO) f H2ClCCCl2O,
HCl2CCClHO, HCl2CCCl2O, Cl3CCCl2O +
RO (NO2) + O2 (R4a,b,c,d)
3. Primary Reaction Mechanisms and Previous Work
The results of previous laboratory studies designed to
determine the product yields from the Cl atom initiated
oxidations of chlorinated ethenes and pentachloroethane are
summarized in Table 1. The early stages of the oxidative
mechanisms are well-established and are similar for all five
species. The Cl atoms produced by photolysis of Cl2 form
chlorinated ethyl radicals either by addition in the cases of the
chlorinated ethenes or by H-atom abstraction from pentachlo-
roethane
The chlorinated ethoxy radicals can then undergo either C-Cl
bond scission to produce an acid chloride or an aldehyde or
C-C bond scission forming HCOCl or COCl2 and chlorinated
methyl radicals.
H2ClCCCl2O, HCl2CCClHO, HCl2CCCl2O,
Cl3CCCl2O (+M) f H2ClCCOCl, HCl2CCHO,
HCl2CCOCl, CCl3COCl + Cl (+M) (R5a,b,c,d)
Cl2 + hν f 2Cl
(R1)
H2ClCCCl2O, HCl2CCClHO, HCl2CCCl2O,
Cl3CCCl2O (+M) f H2ClC + COCl2, HCl2C +
HCOCl, HCl2C + COCl2, COCl2 +
CCl3 (+M) (R6a,b,c,d)
H2CCCl2, HClCCClH, HClCCCl2, Cl2CCCl2 +
Cl (+M) f H2ClCCCl2, HCl2CCClH,
HCl2CCCl2, Cl3CCCl2 (+M) (R2a,b,c,d)
In addition, the HCl2CCClHO radical from HClCCClH can lose
the R-hydrogen atom by reaction with O2
Cl3CCCl2H + Cl f Cl3CCCl2 + HCl (R2d′)
At this point the mechanistic pathways for C2Cl4 and C2Cl5H
converge, and the remainder of the chemistry is thought to be
HCl2CCClHO + O2 f HCl2CCOCl + HO2 (R7b)

2034 J. Phys. Chem. A, Vol. 103, No. 13, 1999
Hasson and Smith
TABLE 2: Present Measurements of Product Yields from Chlorine Atom Initiated Oxidation of Chlorinated Ethenes and
Reaction Enthalpies for Dissociation of Chlorinated Ethoxy Radicalsa
chlorinated ethenes
product yields
dissociation reactions
∆rH°/kJ mol-1
H2CdCCl2
CH2ClCOCl: 0.98 ( 0.03
COCl2: 0.03 ( 0.01
HCOCl: 0.03 ( 0.02
CH2ClCCl2O f CH2ClCOCl + Cl
CH2ClCCl2O f CH2Cl + COCl2
-69
-42
carbon balance 1.01 ( 0.04
CHCl2COCl: 0.08 ( 0.02
HCOCl: 1.26 ( 0.32
HClCdCClH
CHCl2CHClO f CHCl2COCl + H
CHCl2CHClO f CHCl2 + HCOCl
CHCl2CHClO f CHCl2CHO + Cl
+39
-19
+3
COCl2: 0.02 ( 0.02
CO: 0.96 ( 0.12
HCl: 1.15 ( 0.27
carbon balance 1.20 ( 0.25
CHCl2COCl: 0.91 ( 0.03
COCl2: 0.09 ( 0.01
HCOCl: 0.09 ( 0.02
carbon balance 1.00 ( 0.04
dependent on [C2Cl4] (see text)
HClCdCCl2
Cl2CdCCl2
CHCl2CCl2O f CH2ClCOCl + Cl
CHCl2CCl2O f CH2Cl + COCl2
-52
-44
CCl3CCl2O f CCl3 + COCl2
CCl3CCl2O f CCl3COCl + Cl
-89
-55
a Thermodynamic data are taken from: (a) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F.; Kerr, J. A.; Troe, J. J. Phys. Chem. Ref.
Data 1997, 26, 521. (b) Lias, J. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin, R. D.; Mallard, W. G. J. Phys. Chem. Ref. Data 1988,
17 (suppl 1). (c) Bertrand, L.; Exsteen-Mayer, L.; Franklin, J. A.; Huybrechts, G.; Olbrechts, J. Int. J. Chem. Kinet. 1971, 3, 89. The data for the
substituted ethoxy radicals are estimated by Bertrand et al. using group additivity rules. (d) Finally, the value of ∆rH°(HOCl) is estimated from the
following theoretical work: Francisco, J. S. J. Chem. Phys. 1992, 96, 7597.
The chlorinated methyl radicals produced in reactions (R6a,b,c,d)
are oxidized to produce HOCl or COCl2 in the following
reactions:
H2ClC, HCl2C, HCl2C, Cl3C + O2 (+M) f
H2ClCO2, HCl2CO2, HCl2CO2, Cl3CO2 (+M) (R8a,b,c,d)
H2ClCO2, HCl2CO2, HCl2CO2, Cl3CO2 + RO2 (NO) f
H2ClCO, HCl2CO, HCl2CO, Cl3CO + RO (NO2) +
O2 (R9a,b,c,d)
H2ClCO, HCl2CO, HCl2CO, Cl3CO (+M) f
H2CO, HCOCl, HCOCl, COCl2 + Cl (+M) (R10a,b,c,d)
H2ClCO, HCl2CO, HCl2CO + O2 f
HCOCl, COCl2, COCl2 + HO2 (R11a,b,c)
It is these mechanisms which lead to the species identified in
Tables 1 and 2 as “primary products”.
4. Cl Atom Initiated Oxidation of H2CdCCl2, ClHCd
CHCl, and ClHCdCCl2: Results and Discussion
When each of these chlorinated ethenes was mixed with Cl2
and O2 and left in the dark, no changes in the reagent
concentration nor any products of dark reactions were observed.
Initial photochemical experiments were carried out using
blacklamps to illuminate gas mixtures containing 4 × 1014
Figure 1. Infrared spectra recorded (a) before and (b) after 30 min
photolysis of a mixture initially containing 4 × 1014 molecule cm-3 of
HClCCCl2, 4 × 1014 molecule cm-3 of Cl2, and 140 Torr of O2 diluted
to a total pressure of 700 Torr with N2.
molecule cm-3 of the chlorinated ethene, 4 × 1014 molecule
cm-3 of Cl2, and 140 Torr of O2. Typical spectra for such
experiments recorded before and after 30 min of photolysis of
HClCCCl2 are shown in Figure 1. Concentrations of products
derived from experiments on trichloroethene, 1,1-dichloroethene,
and 1,2-dichloroethene are plotted in Figure 2 against decrease
in the reagent concentration. The product yields derived from
the slopes of these lines are listed in Table 2. These data were
determined from unweighted linear regressions on plots of the
kind shown in Figure 2 which were not forced through the
origin. The cited errors correspond to two standard deviations
arising from uncertainties in the experimentally determined
concentrations. At least three experiments were performed for
each of the alkenes under a specific set of reaction conditions,
and both the values and the errors stated are the weighted
average of these separate experiments.
According to the mechanisms outlined in section 3, the
distribution of product yields from each of the three chlorinated
alkenes should be independent of (a) the rate of photolysis, (b)

Oxidation of Chlorinated Ethenes
J. Phys. Chem. A, Vol. 103, No. 13, 1999 2035
Figure 2. Yields of products plotted against loss of reagents from
experiments on (a) H2CCCl, (b) HClCCCl2, and (c) HClCCClH: (b)
CH2ClCOCl; (1) CHCl2COCl; ([) HCOCl + CO (see text). In all
cases, 140 Torr of O2 was present in the reaction mixtures.
Figure 3. Dependence of product yields on (a) rate of photolysis and
[chlorinated ethene]0 with [Cl2]0 ) 4 × 1014 molecule cm-3, and (b)
[Cl2]0 with [chlorinated ethene]0 ) 4 × 1014 molecule cm-3: (b), (O)
CH2ClCOCl from H2CCCl2; (9), (0) COCl2 from H2CCCl2; (1), (3)
CHCl2COCl from HClCCCl2; (2), (4) COCl2 from HClCCCl2. The
filled symbols represent data from experiments performed with black-
lamps, the open symbols those from experiments performed with
fluorescent lamps.
the initial concentration of the chlorinated ethene, and (c) the
initial concentration of Cl2. Figure 3 displays the results of
experiments performed with different initial concentrations of
partially chlorinated ethene and of Cl2, using 3 × 1013 to 4 ×
1014 molecule cm-3 of chlorinated ethene and 1 × 1014 to 1 ×
1015 molecule cm-3 of Cl2, in mixtures containing 140 Torr
of O2 at a total pressure of 700 Torr. Within experimental error,
the product yields for all three compounds remained the same
as the concentrations of the reagents were varied. The product
distributions were also found to be unchanged when the rate of
photolysis was reduced by a factor of ca. 100 by replacing the
blacklamps with fluorescent lamps.
Figure 4 shows the effect on the product yields of varying
the partial pressure of O2 included in the reaction mixtures from
10 to 700 Torr. As anticipated, in the oxidations of HClCCCl2
and H2CCCl2, which contain no R-hydrogen atoms, the yields
were found to be independent of [O2]. Moreover, the product
yields from these oxidations both (a) gave excellent carbon
balances within tight experimental bounds and (b) agree very
well with the limited amount of previous data available for these
reactions (cf. Table 1). In the experiments on HClCCCl2, there
was no evidence for the unimolecular decomposition products
of HCOCl, i.e., HCl + CO, on the time scale of our experiments.
In addition, the yields of formyl chloride from all of our
experiments were consistent with chlorinated methoxy radicals
undergoing only reactions R10a,b,c,d and R11a,b,c. There is
some evidence27 that the elimination of HCl from chloromethoxy
radicals may be a minor reaction channel (ca. 20% with 140
Torr of O2 present), but these radicals are formed in such low
yield in the oxidations of 1,2-dichloroethene and trichloroethene
that any HCl and CO formed would be below the detection
limit in our experiments.
The chlorinated ethoxy radical, HCl2CCClHO, formed in the
oxidation of 1,2-dichloroethene does contain an R-hydrogen
atom which can be abstracted by reaction with O2. Consequently,
changing the concentration of O2 alters the rate of reaction R7b,
and the product yields from the oxidation of HClCCClH depend
on the concentration of O2 as shown in Figure 4b. The product
yields from the oxidation of HClCCClH are in reasonable
agreement with the only previous measurement17 in that the
majority of alkoxy radicals decompose via CC bond cleavage,
even in the presence of 140 Torr of O2. No CHCl2CHO is
observed, indicating that carbon-chlorine bond scission does
not occur to any appreciable degree in CHCl2CHClO. However,
it is necessary to examine the possibility that the concentration
of this aldehyde is reduced either because it is photolyzed or
because it reacts with Cl atoms. The latter possibility is unlikely
since the reaction of Cl atoms with CHCl2CHO is about 8 times
slower than their reaction with HClCCClH.28,29 In addition,
experiments were performed to investigate the possibility of loss
of CHCl2CHO by photolysis. No such loss was observed, so it
is concluded that reaction (R5b) was of negligible importance
in our experiments.
CHCl2COCl was observed as a minor product of the oxidation
of HClCCClH. The fact that it was not observed by Sanhueza
and Heicklen17 may due to the very different initial partial
pressures in their experiments: p(Cl2) ) 0.6-6.6 Torr; p(H-
ClCCClH) ) 1.4-9.5 Torr; p(O2) ) 5.8-87.6 Torr. Moreover,
the yields of HCOCl and CO observed in the present experi-

2036 J. Phys. Chem. A, Vol. 103, No. 13, 1999
Hasson and Smith
TABLE 3: Branching Ratios for Reactions of Chlorinated
Ethoxy Radicals Formed in the Cl Atom Initiated Oxidation
of H2CCCl2, HClCCClH, and HClCCCl2
chlorinated
ethene
branching
ratio
reaction
H2CCCl2
CH2ClCCl2O f CH2ClCOCl + Cl
CH2ClCCl2O f CH2Cl + COCl2
0.98 ( 003
0.02 ( 0.003
0.15 ( 0.08
0.77 ( 0.08
HClCCClHa CHCl2CHClO f CHCl2CO + HCl
CHCl2CHClO f CHCl2 + HCOCl
CHCl2CHClO + O2 f CHCl2COCl + HO2 0.08 ( 0.02
CHCl2CCl2O f CHCl2COCl + Cl
CHCl2CCl2O f CHCl2 + COCl2
HClCCCl2
0.91 ( 0.03
0.09 ( 0.02
a The branching ratios are given for 700 Torr total pressure of
synthetic air (i.e., 140 Torr O2). They are determined from the yields
measured under these conditions.
branching ratio of (R6b) would be given by the sum of the
product yields of HCOCl and CO. The values 1.1 ( 0.2 and
0.97 ( 0.05 are obtained from the present work and from that
of Sanhueza and Heicklen, respectively, the results agreeing
within experimental error.
When averaged over all of the experiments on 1,2-dichloro-
ethene, the yield of HCl is 19 ( 10% higher than that of CO.
This is clearly not consistent with the formation of these species
solely from the decomposition of HCOCl. The partially
chlorinated alkoxy radicals CCl3CHClO and CH2ClCHClO have
been shown to decompose via the three-centered elimination
of HCl with branching ratios of 20-25%.29,30 The corresponding
reaction in the 1,2-dichloroethene oxidation is
CHCl2CHClO f CHCl2CO + HCl
(15b)
The radical produced would rapidly react with O2 and would
ultimately form CO2.
Figure 4. Dependence of product yields on [O2] with [chlorinated
ethene]0 ) [Cl2]0 ) 4 × 1014 molecule cm-3: (a) for H2CCCl2 and
HClCCCl2, where (9) represent COCl2 from H2CCCl2, (b) CH2ClCOCl
from H2CCCl2, (2) COCl2 from HClCCCl2, (1) CH2ClCOCl from
HClCCCl2, and (b) for HClCCClH where (1) represents CHCl2COCl
(s calculated) and ([) HCOCl (- - - calculated).
CHCl2CO + O2 + M f CHCl2C(O)O2 + M
CHCl2C(O)O2 + RO2 f CHCl2C(O)O + RO + O2 (R17b)
CHCl2C(O)O f CHCl2 + CO2 (R18b)
(R16b)
ments are different from those measured by Sanhueza and
Heicklen, who also deduced that CO is a primary product of
the oxidation since its concentration was proportional to the
irradiation time. They attributed this behavior to the reaction
Unfortunately, due to the interference of CO2 absorptions outside
the reaction cell, it was impossible to obtain an accurate yield
for CO2. However, it seems likely that this is the source of the
“extra” HCl observed, leading to the ratio for kR15b/kR6b ) 0.19
( 0.1. This value must be treated with caution because, before
the yield of CO2 is included, the experimental carbon balance
from these experiments is in excess of 100%.
CHCl2CHClO f CO + HCl + CHCl2 (R12b)
However, the concentrations of CO, HCl, and HCOCl produced
in the current set of experiments were found not to be
proportional to the consumption of 1,2-dichloroethene. In
addition, when the photolysis lamps were switched off, the
concentrations of CO and HCl were both observed to increase,
whilst that of HCOCl decreased by the same amount. Conse-
quently, CO would appear, at least in part, to be a secondary
product formed by the decomposition of HCOCl, and there is
no evidence for CO production from (R12b). Studies of the
oxidation of 1,2-dichloroethane30 and 1,1,1,2-tetrachloroethane31
have shown that reactions analogous to (R12b), i.e.,
Using the yields shown in Figure 4b, the branching ratio kR7b
-
[O2]/kRb′ could be determined (where kRb′ ) kR5b + kR6b + kR12b
+ kR15b ≈ kR6b + kR15b, since kR5b and kR12b are much smaller
than kR6b and kR12b). An analysis using linear regression led to
the value kR7b/kRb′ ) (1.2 ( 0.5) × 10-20 cm3 molecule-1. This
ratio was used to construct the lines shown in Figure 4b.
Combining this result with the relative yields of HCl2CCOCl
and HCOCl shown in Table 2 allows the branching ratios for
all three reactive channels for the radical HCl2CCClHO (R6b,
R7b, and R15b) to be determined at all concentrations of O2
(see Table 3).
Table 2 lists the values of ∆rH° for the possible dissociations
of the chlorinated ethoxy radicals derived from H2CdCCl2,
HClCdCClH, and HClCdCCl2. In all three of these cases, it
can be seen that the branching ratios between CC, CCl, and
where possible, CH bond scission are a reasonable reflection
of the thermochemistry. That is, in all cases, the most favored
unimolecular process is the most exothermic pathway, although
it should be appreciated that the uncertainties in the values of
CCl3CHClO f CCl3 + CO + HCl
(R13b)
CH2ClCHClO f CH2Cl + CO + HCl (R14b)
do not occur. If the CO observed by Sanhueza and Heicklen
was a consequence of the decomposition of HCOCl rather than
reaction R7b, then the yield of CO at any time would be equal
to the amount of HCOCl that had decomposed. Thus the

Oxidation of Chlorinated Ethenes
J. Phys. Chem. A, Vol. 103, No. 13, 1999 2037
∆rH° are large since the values of ∆fH° for the chlorinated
ethoxy radicals are only estimates based on group additivity
rules.34
5. Cl Atom Initiated Oxidation of Cl2CdCCl2 and
Cl3CCCl2H: Results
A major objective of the present work was to explore and
explain the kinetics and mechanism of the oxidation of C2Cl4
and C2Cl5H under a range of experimental conditions in order
to provide information that would be useful in understanding
the fate of C2Cl4 in the troposphere. Although there have been
several previous studies of the photochemical oxidation of
C2Cl4, the older studies12,16,20 used concentrations of Cl2 and
C2Cl4 much greater than in the present work, which itself uses
concentrations of C2Cl4 that are much greater than those present
in the atmosphere. As the data in Table 1 shows, these previous
studies and that of Ariya et al.21 measured product distributions
that were very similar both to one another and to yields from
experiments on C2Cl5H, which has been taken as proof that the
two mechanisms are identical after reaction R2d or R2d′. Ariya
et al.,21 who used a similar experimental technique to ours, did
employ initial partial pressures of Cl2 and C2Cl4 of the order of
a few mTorr, but it is not obvious that the product yields were
measured as the initial conditions were systematically varied.
Likewise, although product yield studies have been carried out
previously at different temperatures, no single set of experiments
has been performed to investigate whether the yields change
over a significant temperature range.
Figure 5. Examples of infrared spectra recorded (a) before and (b)
after 20 min of photolysis of a mixture initially containing 4 × 1014
molecule cm-3 Cl2, 4 × 1014 molecule cm-3 C2Cl4, and 140 Torr O2,
made up to 700 Torr total pressure with N2, and (c) before and (d)
after 2 h of photolysis of a mixture initially containing 4 × 1014
molecule cm-3 Cl2, 4 × 1014 molecule cm-3 C2Cl5H, and 140 Torr O2,
made up to 700 Torr total pressure with N2.
Part of our aim was to compare the rates at which the
concentrations of reagents and products changed with those
predicted using the modelling program FACSIMILE.32 In order
to do this, it was necessary to determine the first-order rate
constant, kp,R1, for photolysis of Cl2 by both the blacklamps
and the fluorescent lamps. To estimate kp,R1, Cl2 was photolyzed
in the presence of C2H6 and O2 and changes in the concentra-
tions of C2H6 and HCl were simultaneously measured. The rates
of C2H6 loss and HCl formation were found to be the same and
were assumed to be twice the rate of photolysis of Cl2, yielding
values of kp,R1 of 3 × 10-4 s-1 (blacklamps) and 3 × 10-6 s-1
(fluorescent lamps).
At room temperature, the time scale for reaction was found
to differ quite dramatically between C2Cl4 and C2Cl5H. For C2-
Cl4, under typical experimental conditions and using blacklamps
to photolyze Cl2, reaction was complete in less than 5 min.
However, under identical conditions, complete oxidation of the
same concentration of C2Cl5H took about 2 h. As the temper-
ature was increased from 298 to 393 K, the rate of reaction of
C2Cl4 decreased by a factor of 10. The reasons for these
observations are discussed later.
In order to follow the kinetics of the C2Cl4 oxidation, the
fluorescent lamps were used to photolyze Cl2 and initiate the
reaction. With this relatively slow rate of photolysis, C2Cl4 was
fully consumed in about 30 min. Typical spectra obtained from
the reactions of C2Cl4 and C2Cl5H are shown in Figure 5. The
temporal profiles of the reactant and product concentrations in
a typical experiment are shown in Figure 6. It is noticeable that
the rates of loss of reactant and formation of products are slower
initially than at later time intervals. It was also observed that
the product concentrations did not increase linearly with the
reduction in reagent concentrations. These findings are due to
the fact that the concentrations of Cl atoms only slowly
approached a limiting steady-state value. Although they are
removed rapidly by reaction with C2Cl4 and C2Cl5H, they are
regenerated in subsequent steps of the reaction mechanism and
the actual removal of radicals from the system is slow.
Figure 6. Comparison of the observed concentrations of C2Cl4 (2),
CCl3COCl (b), and COCl2 (9) in a typical experiment using fluorescent
lamps (kp,R1 ) 3 × 10-6 s-1) with those predicted using FACSIMILE
and the rate constants listed in Table 4, including reactions R23d and
R24d unless otherwise stated: (a) C2Cl4 (-‚‚-), C2Cl4 excluding (R26d)
and (R27d) (- - -), (b) C2Cl4 (-‚‚-), CCl3COCl (s), and COCl2 (‚‚‚).
The reaction mixture was the same as that specified in the caption to
Figure 5.
How the final product yields depended on changes in the
initial concentrations of Cl2, O2, and C2Cl4 was investigated in
three separate series of experiments. It is important to note that

2038 J. Phys. Chem. A, Vol. 103, No. 13, 1999
Hasson and Smith
Figure 8. (a) Comparison of the observed yields of CCl3COCl (b)
and COCl2 (9) for different initial concentrations of C2Cl4 with those
predicted using FACSIMILE and the rate constants listed in Table 4
(including reactions R23d and R24d): CCl3COCl (s) and COCl2 (‚‚
‚). (b) Comparison of the dependence of the observed yields of
CCl3COCl (b) and COCl2 (9) on the initial concentration of C2Cl5H
with those predicted using FACSIMILE and the rate constants listed
in Table 4 (including reactions R25d′ and R26d′): CCl3COCl (s) and
COCl2 (‚‚‚). The reaction mixtures initially contained 4 × 1014 molecule
cm-3 Cl2 and 140 Torr O2 made up to 700 Torr total pressure with N2.
Figure 7. Comparison of the observed yields of CCl3COCl (b) and
COCl2 (9) for different initial concentrations of Cl2 with those predicted
using FACSIMILE and the rate constants listed in Table 4 (including
reactions R23d and R24d), CCl3COCl (s) and COCl2 (‚‚‚): (a) using
the blacklamps so that kp,R1 ) 3 × 10-4 s-1 and (b) using the fluorescent
lamps so that kp,R1 ) 3 × 10-6 s-1 in a reaction mixture initially
containing 5.5 × 1013 molecule cm-3 C2Cl4, and 140 Torr O2, made
up to 700 Torr total pressure with N2.
the reaction mechanism represented by reactions R1-R10d does
not predict any dependence of the product yields on the initial
concentrations of any of the reagents involved, since CCl3COCl
and COCl2 are formed only in the unimolecular dissociation
reactions R5d, R6d, and R10d. Variations in [Cl2]0, the initial
concentration of Cl2, in the range 4 × 1013 to 4 × 1014 molecule
cm-3, resulted in no significant changes in the product distribu-
tions from either C2Cl4 or C2Cl5H, except at low [Cl2]0. The
results of such experiments using [C2Cl4]0 ) 5.5 × 1013
molecule cm-3 and with [O2] ) 140 Torr are shown in Figure
7. At low photolysis rates, as shown in Figure 7b, the relative
yields of CCl3COCl and COCl2 were independent of the initial
concentration of Cl2, within experimental error. However, at
the higher photolysis rates induced by the blacklamps, the yield
of CCl3COCl fell slightly as the amount of Cl2 present in the
reaction mixture was increased. No dependence of the product
yields on the O2 concentration was observed over a range of
pressures from 30 mTorr to 700 Torr for any of the initial
concentrations of Cl2 or C2Cl4 that were investigated.
To examine if the product yields depended on the initial
concentration of the chlorinated hydrocarbon, mixtures of
C2Cl4 or C2Cl5H (3 × 1013 to 1 × 1015 molecule cm-3), O2
(140 Torr), and Cl2 (4 × 1014 molecule cm-3) were irradiated
using blacklamps for times between 10 min (C2Cl4) and 2 h
(C2Cl5H) with spectra being recorded every few minutes to
observe the loss of the reactant and the formation of CCl3COCl
and COCl2. The total yields for each experiment, measured as
C atom balances, exceeded 90%, and no products other than
CCl3COCl and COCl2 were detected. The results summarized
in Figure 8a represent normalized product distributions from
these experiments. It is seen that, at relatively high initial
concentrations of C2Cl4, CCl3COCl was the major product with
a yield of ca. 90%. This value is in good agreement with
previous work21 carried out on this reaction system under similar
conditions with similar initial concentrations (see Table 1).
However, and unexpectedly, as [C2Cl4]0 was reduced in our
experiments, the yield of CCl3COCl fell and the yield of COCl2
showed a corresponding rise. At the lowest values of [C2Cl4]0
that could be used, COCl2 became the dominant product. If this
trend were to continue, then at very low C2Cl4 concentrations,
as in the atmosphere, the yield of CCl3COCl would become
virtually zero.
Experiments were also performed to check that our results
were not affected by wall reactions or by further chemistry or
photochemistry of the primary products. To test the latter
possibility, samples of CCl3COCl and COCl2 with 140 Torr O2
were irradiated, both without and with Cl2 present. In no case
was significant change observed, even with irradiation times
well in excess of those used in the main experiments. In our
reaction cell (diameter ) 7.5 cm) at a total pressure of 700
Torr, molecules would take ca. 100 s to diffuse to the vessel
walls which are covered in Teflon to reduce their activity.
Leaving both the reagents and products in the dark brought about
no discernible change, suggesting that heterogeneous chemistry
is also unimportant in our experiments.
The effect of the photolysis rate of Cl2 on the product
distribution was also investigated by measuring the product
yields using blacklamps (kp,R1 ) 3 × 10-4 s-1) and fluorescent
lamps (kp,R1 ) 3 × 10-6 s-1) in two series of experiments. First,
gas mixtures containing a low concentration of C2Cl4, 140 Torr
of O2, and between 3 × 1013 molecule cm-3 and 4 × 1014
molecule cm-3 of Cl2 were photolyzed with the fluorescent

Oxidation of Chlorinated Ethenes
J. Phys. Chem. A, Vol. 103, No. 13, 1999 2039
Figure 10. Dependence of the observed yields of CCl3COCl (b,O,.)
and COCl2 (9,0,~) from the oxidation of C2Cl4 on the initial
concentration of C2Cl4 at different temperatures: 298, 353, and 393
K, under the same experimental conditions as Figure 6.
Figure 9. Comparison of the dependence of the observed yields of
CCl3COCl (b,O) and COCl2 (9,0) on the initial concentration of C2-
Cl4 and different rates of photolysis with those predicted using
FACSIMILE and the rate constants listed in Table 4 (including reactions
R23d and R24d). The data represented by the filled symbols were
obtained using fluorescent lamps and those given by the open symbols
were obtained using the blacklamps. The line (s) shows the results of
calculations using kp,R1 ) 3 × 10-6 s-1; the line (‚‚‚) shows the results
of calculations using kp,R1 ) 3 × 10-4 s-1
.
lamps. As already discussed, at the slower rate of photolysis, a
slightly different product distribution was obtained, and it was
found to have a much reduced dependence on [Cl2]0 (see Figure
7). In addition, a second series of experiments was performed
in which a fixed [Cl2]0 of 4 × 1014 molecule cm-3 was
photolyzed with the fluorescent lamps, and the initial concentra-
tion of C2Cl4 was varied as before. The result of these
experiments are compared in Figure 9 with those in which
photolysis by the blacklamps was used to initiate oxidation of
C2Cl4. Once again, the yields of CCl3COCl and COCl2 were
found to depend on the initial concentration of C2Cl4, although
the dependence in this case was slightly less marked than at
the higher photolysis rate.
In light of the unanticipated variation of product yields from
the oxidation of C2Cl4 with initial conditions, especially the
initial concentration of C2Cl4, it was decided to investigate the
effects of temperature and total pressure on these results.
Experiments to determine the yields of COCl2 and CCl3COCl
as [C2Cl4]0 or [C2Cl5H]0 was varied were repeated at 353 and
393 K using the black lamps to photolyze Cl2. The results for
C2Cl4 are shown in Figure 10. The temperature dependence of
the yields of CCl3CCOCl and COCl2 was found to be slight,
the limit of the yield of CCl3COCl at high [C2Cl4]0 rising from
89.0% at 298 K to 93.5% at 393 K, in agreement with previous
measurements21 carried out using similar concentrations of
reactants. At each of the three temperatures which were
employed, a similar dependence of product yields on [C2Cl4]0
was observed. Figure 11 shows the results of experiments
conducted at three different total pressures. Although the
dependence of the product yields on [C2Cl4]0 persists, it is
reduced slightly as the pressure is reduced from 700 to 50 Torr,
through the range of [C2Cl4]0 which could be used in our
experiments.
Figure 11. Dependence of the observed yields of CCl3COCl (b,O,.)
and COCl2 (9,0,~) from the oxidation of C2Cl4 on the initial
concentration of C2Cl4 at different total pressures: 700, 140, and 50
Torr, under the same experimental conditions as Figure 6.
dependence of the yields of CCl3COCl and COCl2 on the initial
concentration of the reagent, in this case C2Cl5H (see Figure
8b), although the dependence of these yields on [C2Cl5H]0 was
not as great as on [C2Cl4]0. In the limit of high [C2Cl5H]0
however, the yields of CCl3COCl and COCl2 were similar to
those from C2Cl4 under similar conditions, a finding which is
in agreement with previous measurements (see Table 1).
Suspecting that CCl3CCl2O radicals might react with HCl, which
is the only product formed in this system that is not present in
the experiments on C2Cl4, a series of experiments were
performed in which HCl was added to Cl2/C2Cl4/O2/N2 mixtures.
[C2Cl4]0 was set at 5.5 × 1013 molecule cm-3 and the partial
pressure of O2 at 140 Torr, and product yields were studied in
the presence of between 3 × 1013 and 4 × 1014 molecule cm-3
of HCl. The dependence of the yields on the amount of HCl
Experiments on the chlorine atom initiated oxidation of
pentachloroethane revealed a similar, and similarly surprising,

2040 J. Phys. Chem. A, Vol. 103, No. 13, 1999
Hasson and Smith
Some experiments were also performed on the oxidation of
C2Cl4 in the presence of NO. The time scale for reaction was
much shorter. Mixtures of 4 × 1014 molecule cm-3 C2Cl4, 4 ×
1014 molecule cm-3 Cl2, 1 × 1015 molecule cm-3 NO, and 140
Torr O2, diluted to 700 Torr in N2 were converted completely
to products in about 2 h. Again, no evidence was found for the
formation of CCl3ONO or C2Cl5ONO. The yield of CCl3COCl
from this mixture was 75%, about 14% lower than in similar
experiments without NO present. A possible explanation for
this difference is that the C2Cl5O radicals formed by reaction
of C2Cl5O2 with NO contain more initial internal energy than
when the same radicals are formed from two peroxy radicals
and this accelerates their rate of unimolecular dissociation by
CC bond scission thereby increasing the yield of COCl2.
Behavior of this kind has been observed for CF3CFHO.33
6. Cl Atom Initiated Oxidation of Cl2CdCCl2 and
Cl3CCCl2H: Discussion
The observations reported in the previous section cannot be
explained simply in terms of the mechanism which was outlined
in section 3 and which is sufficient to explain our results for
the three partially chlorinated ethenes. If this mechanism held
for the oxidations of C2Cl4 and C2Cl5H, then the ratio of
products, [CCl3COCl]/[COCl2], should be determined only by
the relative rates of the unimolecular decompositions of the C2-
Cl5O radical in reactions R5d and R6d. To explain the observed
chemistry, especially the dependence of the product yields on
the initial concentrations, [C2Cl4]0 and [C2Cl5H]0, it seems that
at least one bimolecular reaction forming product must be
included in the mechanisms for the oxidation of these com-
pounds.
Figure 12. Comparison of the dependence of the observed yields of
CCl3COCl (b) and COCl2 (9) from the Cl atom initiated oxidation of
C2Cl4 on the initial concentration of HCl with those predicted using
FACSIMILE and the rate constants listed in Table 4 (including reactions
R23d, R24d, R25d′, and R26d′): CCl3COCl (s) and COCl2 (‚‚‚)
present is shown in Figure 12. As the concentration of HCl was
increased, the yield of COCl2 fell, a corresponding rise being
observed in the concentration of CCl3COCl.
Finally, as a further attempt to understand our observations
on the oxidations of C2Cl4 and C2Cl5H, we decided to examine
the kinetics and product yields in the presence of added NO. It
was hoped that it might be possible to determine the rate of
unimolecular decomposition of the CCl3O and C2Cl5O radicals
by comparison of the yields of COCl2 from reaction R10d and
CCl3COCl and COCl2 from (R5d) and (R6d) with those of CCl3-
ONO and C2Cl5ONO formed in the association reactions
In the case of the oxidation of C2Cl4, it seemed reasonable
to suppose that this species might be involved in this additional
reaction and one possible step that would influence the yields
of products in the manner that is observed is
C2Cl4 + C2Cl5O f CCl3COCl + C2Cl5 (R23d)
This reaction is thermodynamically allowed,34 being exothermic
by -128 kJ mol-1, and, if it were fast enough to compete with
the unimolecular decompositions R5d and R6d, its occurrence
would explain why higher initial concentrations of C2Cl4 result
in higher yields of CCl3COCl and decreased yields of COCl2.
The reaction
CCl3O + NO + M f CCl3ONO + M
(R19d)
C2Cl5O + NO + M f C2Cl5ONO + M (R20d)
although it was also recognized that, with NO present, C2Cl5O2
radicals would be converted to C2Cl5O radicals by reaction with
NO, rather than with another peroxy radical (see R5d).
To generate CCl3O and C2Cl5O radicals fairly directly,
experiments were performed on the Cl atom initiated oxidations
of CHCl3 and C2Cl5H. Mixtures of 4 × 1014 molecule cm-3 of
the chlorinated alkane, 4 × 1014 molecule cm-3 Cl2, 1 × 1015
molecule cm-3 NO, and 140 Torr O2 were diluted to 700 Torr
in N2 and irradiated with the fluorescent lamps. In both cases,
COCl2 was the only product observed but the yield could not
be quantified as its formation was extremely slow. Almost
certainly this was due to the NO-catalyzed recombination of
Cl atoms,
C2Cl4 + CCl3O f CCl3COCl + CCl3
(R24d)
is also exothermic, although reaction to produce C2Cl5 + COCl2
is possible.
To examine the potential importance of these reactions, we
employed the modeling program FACSIMILE.32 In Figure 6a,
we compare the observed decrease in [C2Cl4] with time with
predictions of the model using the rate constants given in Table
4, both excluding and including reactions R23d and R24d. In
the absence of these reactions, the model failed to reproduce
the experimental results, particularly the decrease in the rate of
loss of C2Cl4 which is observed towards the end of the reaction.
In these calculations, the concentration of Cl atoms continued
to rise at long times causing the rate of loss of C2Cl4 to
accelerate. On the other hand, addition of reactions R23d and
R24d to the model resulted in good agreement with experiment,
both in regard to the consumption of C2Cl4 and the rates of
formation of the major products, as shown in Figure 6. Using
the values of the first-order rate constants for the unimolecular
dissociations of C2Cl5O and CCl3O, i.e., for reactions R5d, R6d,
Cl + NO + M f ClNO + M
Cl + ClNO f Cl2 + NO
(R21)
(R22)
drastically reducing the concentration of Cl atoms present in
the system. Even after several hours of the irradiation with the
blacklamps, only 2% of the CHCl3 had reacted. Varying the
initial concentration of NO altered the overall rate of reaction,
but COCl2 remained the only observed product.

Oxidation of Chlorinated Ethenes
J. Phys. Chem. A, Vol. 103, No. 13, 1999 2041
TABLE 4: Reactions and Rate Constants Used in Modeling Calculations at 298 K
reaction
k/cm3 molecule-1 s-1 or k/s-1 a
ref
(R1)
(R-1)
Cl2 + hν f 2Cl
2Cl f Cl2
3 × 10-4 or 3 × 10-6
fitted (see text)
39
38
38
4
22
11
11
fitted (see text)
fitted (see text)
40
35
37
fitted (see text)
fitted (see text)
fitted (see text)
fitted (see text)
fitted (see text)
fitted (see text)
2.5 × 10-13
1.18 × 10-14
4.7 × 106
3 × 10-11
1.2 × 10-14
8.6 × 10-13
6 × 10-11
1 × 103
Cl + O2 f ClO2
ClO2 f Cl + O2
(R2d)
(R2d′)
(R3d)
(R4d)
(R5d)
(R6d)
(R8d)
(R9d)
(R10d)
(R23d)
(R24d)
(R25d′)
Cl + C2Cl4 f C2Cl5
Cl + C2Cl5H f C2Cl5 + HCl
C2Cl5 + O2 f C2Cl5O2
2C2Cl5O2 f 2C2Cl5O + O2
C2Cl5O f CCl3COCl + Cl
C2Cl5O f COCl2 + CCl3
CCl3 + O2 f CCl3O2
1 × 104
2.5 × 10-12
1.5 × 10-12
1 × 105
2CCl3O2 f 2CCl3O + O2
CCl3O f COCl2 + Cl
C2Cl5O + C2Cl4 f CCl3COCl + C2Cl5
CCl3O + C2Cl4 f CCl3COCl + CCl3
C2Cl5O + HCl f C2Cl5OH + Cl
CCl3O + HCl f CCl3OH + Cl
C2Cl5OH f CCl3COCl + HCl
CCl3OH f COCl2 + HCl
5 × 10-10
5 × 10-10
9 × 10-10
9 × 10-10
1 × 106
(R26d′)
1 × 106
and R10d, agreement with experiment was given by choosing
the values of kR23d and kR24d to be 5 × 10-10 cm3 molecule-1
s-1. Although these rate constants seem unreasonably high, we
point out that such values are needed in order for the bimolecular
reactions R23d and R24d to compete with unimolecular
decomposition of the chlorinated ethoxy radicals. Similar
agreement can be obtained by lowering the values for both
groups of rate constants by the same amount.
[C2Cl4]0. It appears that the decrease in CCl3COCl produced,
and the corresponding increase in COCl2, at high [Cl2]0 and
with the fast photolysis rate must be associated with the very
rapid consumption of C2Cl4 under these conditions. Although
the results of the calculations using FACSIMILE only fit the
experimental data on Figure 8a moderately well, they do
reproduce quite well the smaller effects of changing the rate of
photolysis.
Having obtained reasonably good agreement with the ex-
perimentally determined profiles of [C2Cl4] versus time, using
the extended model, attempts were made to reproduce the
measured product yields and how they depended on experi-
mental conditions. The addition of reactions R23d and R24d
also resulted in good agreement with the sets of experiments
where the initial concentrations of Cl2 (Figure 7) and of C2Cl4
(Figure 8a), and the photolysis rate of Cl2 (Figure 7) were varied.
Then, according to the model and using the rate constants listed
in Table 4, with high [C2Cl4]0 products are formed principally
through the bimolecular reactions R23d and R24d. Hence the
major product is CCl3COCl. However, at low concentrations
of C2Cl4, this reaction becomes slower and the product distribu-
tion is largely determined by the two unimolecular reactions
R5d and R6d. Our results therefore suggest that (R6d) is faster
than (R5d), a result which is consistent with the finding that
fission of the C-C bond in the pentachloroethoxy radical is
more exothermic than fission of a C-Cl bond in the CCl2O
group (see Table 2).
At first sight, our tentative explanation for the results of our
experiments on C2Cl4, particularly the variation of product yields
with [C2Cl4]0, appears to conflict with the finding that in similar
experiments with C2Cl5H the major product is again CCl3COCl
(see Figure 7b). Since there is no C2Cl4 present in these
experiments and the C2Cl5 radicals formed by abstraction of
the H atom from C2Cl5H by Cl atoms should be rapidly
converted to C2Cl5O radicals by reactions R3d and R4d, one
might expect the products to be the same as those found in the
experiments with low [C2Cl4]0. The reason why this discrepancy
might arise is given by the experiments on the oxidation of
C2Cl4 in the presence of HCl, the results of which are shown in
Figure 12. As the HCl concentration is increased, the yield of
COCl2 falls. We ascribe this observation to the reaction of
C2Cl5O radicals with HCl:
C2Cl5O + HCl f C2Cl5OH + Cl
(R25d′)
Perchloroethanol is not observed among the products and
presumably decomposes
The small effects on the product yields from C2Cl4 of varying
[Cl2]0 is well matched by the model calculations for both fast
and slow rates of photolysis. With reactions R23d and R24d
included, the relative rates of formation of CCl3COCl and COCl2
are given by the ratio of (a) the sum of the rates of reactions
R5d, R23d, and R24d to (b) the rate of reaction R6d, i.e.,
C2Cl5OH (+M) f CCl3COCl + HCl (+M)
(R26d′)
so that HCl can act catalytically to convert C2Cl5O radicals to
CCl3COCl via reactions R25d′ and R26d′. Although a similar
sequence of reactions can occur for CCl3O radicals yielding
COCl2, it will not increase the yield of COCl2 since any CCl3O
radicals that are formed are converted to COCl2 in unit yield in
any case. Once again, FACSIMILE32 was used to examine the
validity of these proposals. Using the values kR25d′ ) 9 × 10-10
cm3 molecule-1 s-1 and kR26d′ ) 1 × 106 s-1, good agreement
was obtained for the variation of product yields with [C2Cl5H]0
in experiments on Cl2/C2Cl5H/O2/N2 mixture (see Figure 7b)
and fair agreement was obtained for the variation of product
yields when HCl was added to Cl2/C2Cl4/O2/N2 mixtures (see
Figure 12).
(rate f CCl3COCl)/(rate f COCl2) )
{kR5d[C2Cl5O] + kR23d[C2Cl5O][C2Cl4] +
kR24d[C2Cl4][CCl3O]}/kR6d[C2Cl5O] (1)
and the ratio of the observed yields corresponds to an integration
of this expression over the duration of an experiment. Effects
of changing the rate of photolysis, by exchanging blacklamps
for fluorescent lamps, are also shown as a function of [C2Cl4]0
in Figure 9. Once again, the effect of the higher rate of
photolysis is seen to be an increase in the yield of COCl2 and
a decrease in that of CCl3COCl. The effect is strongest at low
At this point, we consider the magnitude of the rate constants
kR23d, kR24d, kR25d, and kR26d that have been used in the modelling

2042 J. Phys. Chem. A, Vol. 103, No. 13, 1999
Hasson and Smith
calculations. For the oxidation of C2Cl5H, we note that the model
results were insensitive to the value chosen for kR26d′ (it was
reduced by several orders of magnitude without significant
change in the computed results, since it is only needed to explain
the absence of any CCl3CCl2OH in the reaction products), and
that kR25d′ could be reduced as long as kR5d and kR6d were reduced
by the same amount. Similarly, the calculated results for the
oxidation of C2Cl4 were insensitive to the absolute values of
the rate constants for the bimolecular and unimolecular reactions
of the chlorinated alkoxy radicals, C2Cl5O and CCl3O, but were
sensitive to their relatiVe values. The values of kR23d, kR24d, and
an energized adduct
CCl3CCl2O + C2Cl4 f (CCl3CCl2OCCl2CCl2)†
(R27d)
followed by intramolecular transfer of a Cl atom and fragmenta-
tion to CCl3COCl + CCl3CCl2. For the latter process to be
effective, it would seem that its rate would have to compete
with addition of O2 to the radical and possibly with collisional
removal of the excess internal energy released as the (RCCl2-
OCCl2CCl2)† radicals are formed in reaction (R27d). Lifetimes
with regard to both these processes can be estimated. Assuming
kR25d′ that are referred to above and given in Table 4, and which
a rate constant for addition of O2 of 10-13 cm3 molecule-1 s-1
,
may seem unreasonably large, are the result of taking the values
of the rate constants for unimolecular decomposition of these
radicals that are listed in Table 4.
with 140 Torr of O2 decomposition of the RCCl2OCCl2CCl2
radical would have to occur in less than ca. 1 µs to compete
favorably with addition of O2. If production of CCl3COCl +
RCCl2 can only occur from the energized adduct formed in
reaction (R27d), then it would have to rearrange and decompose
even more rapidly, since the average time between collisions
at 700 Torr total pressure is only ca. 0.1 ns.
As the temperature was increased, the most obvious effect
on the reaction of C2Cl4 was that the time for complete reaction,
with the illumination by the blacklamps, increased from a few
minutes to ca. 1 h. This can be attributed to the fact that, at 330
K and below, the addition of Cl atoms to C2Cl4, i.e., reaction
R2d, is essentially irreversible.4 However, at higher tempera-
tures, especially above 330 K, the reaction becomes reversible
and the equilibrium starts to favor Cl + C2Cl4 under the
conditions of our experiments. As a result, the concentration of
C2Cl5 radicals is reduced and the overall reaction slows down.
Finally, we reemphasise that the rate constants for (R23d)
and (R24d) which are required to match the model predictions
to our experimental results depend on the rate constants assumed
for (R5d), (R6d), and (R10d). The latter are very uncertain. The
only attempt to measure any of these rate constants was in an
indirect study of reaction R10d by Lesclaux et al.37 who
concluded that, at 7.5 Torr, kR10d > 1 × 105 s-1. Atkinson et
al.38 suggest a value of 8 × 106 s-1 for atmospheric conditions.
However, in order to keep the rate constants for the bimolecular
reactions R23d and R24d within anything like reasonable
bounds, kR10d was set at the lower limit set by Lesclaux et al.
and kR5d and kR6d were given the values shown in Table 4. If
these rate constants are actually smaller, then those for reactions
R23d and R24d could be lowered in proportion. Perhaps the
least satisfactory aspect of the tentatiVe proposal that is made
here is that one would expect the variation of product yields
from the oxidation of C2Cl4 to show a quite marked temperature
dependence in contrast to what is found experimentally.
Whatever the absolute values of kR5d, kR6d, and kR10d, it seems
impossible that they have a similar temperature dependence to
Figure 10 shows the CCl3COCl and COCl2 yields from the
oxidation of C2Cl4 at three temperatures, 298, 353, and 393 K.
They are surprisingly insensitive to temperature. Whatever the
rate constants for the unimolecular dissociations of the C2Cl5O
radicals via reactions R5d and R6d, it is reasonable to suppose
that they have larger activation energies than the bimolecular
reactions R23d and R24d, which must have rather large rate
constants at room temperature if they are responsible for the
dependence of product yields on the initial concentration of
C2Cl4. A partial explanation for the fact that the dependence of
product yields on [C2Cl4]0 scarcely depends on temperature may
be the difference in the temperature dependence of the rate
constants for (R5d) and (R6d). Reaction R5d which, like (R23d)
and (R24d), produces CCl3COCl, is apparently slower at room
temperature than (R6d), consistent with the fact that the C-Cl
bond is stronger than the C-C bond in CCl3CCl2O (see Table
2). Consequently, the rate of (R5d) will increase more rapidly
with temperature than (R6d), thereby compensating to some
extent for the diminishing effect of reactions R23d and R24d.
k23d
.
7. Summary and Conclusions
A comprehensive study of the Cl atom initiated oxidations
of four chlorinated ethenes has been carried out. For the first
time, experiments have been performed using initial concentra-
tions of the organic species below 1014 molecule cm-3. For the
three partially chlorinated ethenes, H2CdCCl2, HClCdCHCl,
and HClCdCCl2, the product yields show no dependence on
the initial concentrations of either Cl2 or the chlorinated ethene
and only in the case of HClCdCHCl, as expected, do they show
any variation with the concentration of O2 included in the
reaction mixture.
However, under similar conditions, the product yields of
COCl2 and CCl3COCl from Cl2CdCCl2 (and from C2Cl5H)
depend on the initial concentration of the chlorinated species
and less strongly on [Cl2]0 and on the rate of photolysis of Cl2.
This is inconsistent with the accepted reaction mechanism.
Incorporation of the steps
The mild dependence of the product yields on total pressure
which is shown in Figure 11 is consistent with the product yields
depending on bimolecular as well as unimolecular reactions.
Thus, the rates of reactions R5d, R6d, and R10d may all fall as
the total pressure is lowered, whereas the rates of the bimolecular
reactions R23d and R24d would be unaffected. As a result, the
influence of the bimolecular reactions would extend to lower
total pressures resulting in a reduced dependence of the product
yields on [C2Cl4]0.
It has to be said that the hypothesis that reactions R23d and
R24d may be predominantly responsible for the formation of
products in the Cl atom initiated oxidation of C2Cl4 at relatively
high initial concentrations of C2Cl4 is not entirely satisfactory.
Unfortunately, there have been no direct studies of reactions
R23d and R24d. CH3O radicals have been reported to react
slowly with ethene and substituted ethenes,35 whilst the reaction
of CF3O radicals with C2H4 is reported to have a rate constant
CCl3CCl2O, CCl3O + C2Cl4 f
C2Cl5, CCl3 + CCl3COCl (R23d,24d)
of about 10-11 cm3 molecule-1 s-1 36
It would seem that the
.
most likely mechanism for rapid reactions giving the products
shown in (R23d) and (R24d) would be the initial formation of
CCl3CCl2O + HCl f CCl3COCl + HCl + Cl
(R25d′)

Oxidation of Chlorinated Ethenes
J. Phys. Chem. A, Vol. 103, No. 13, 1999 2043
(15) Graedel, T. E.; Keene, W. C. Global Biogeochem. Cycles 1995, 9,
into the reaction mechanisms enabled the experimental results
to be modelled reasonably well, although the values of the rate
constants that had to be assumed conflict with present indirect
evidence about the rates of these reactions. Whilst our results
provide some evidence for the above reactions, direct kinetic
studies of these processes are required to confirm their
importance. Such measurements are at present beyond the
capabilities of our apparatus.
427.
(16) Sanhueza, E.; Hisatsure, J. C.; Heicklen, J. Chem. ReV. 1976, 76,
801.
(17) Sanhueza, E.; Heicklen, J. Int. J. Chem. Kinet. 1975, 7, 589.
(18) Catoire, V.; Ariya, P. A.; Niki, H.; Harris, G. W. Int. J. Chem.
Kinet. 1997, 29, 65.
(19) Schutt, C.; Schumaker, H. J. Z. Phys. Chem. 1941, B49, 107.
(20) Mathias, E.; Sanhueza, E.; Hisatsure, J. C.; Heicklen, J. Can. J.
Chem. 1975, 52, 3852.
The tropospheric concentration of C2Cl4 is around 15 pptv.
This is several orders of magnitude below the lowest concentra-
tion used in our experiments. As a result of this, COCl2 and
not CCl3COCl may be the major product of Cl atom initiated
oxidation of C2Cl4 in the troposphere. If this is true, then
tetrachloroethene is unlikely to be a source of trichloroacetic
acid in the environment.
(21) Ariya, P. A.; Catoire, V.; Sander, R.; Niki, H.; Harris, G. W. Tellus,
Ser. B 1997, 49, 583.
(22) Mogelberg, T. E.; Sehested, J.; Nielsen, O. J.; Wallington, T. J. J.
Phys. Chem. 1995, 99, 16932.
(23) Moore, C. M.; Smith, I. W. M. J. Chem. Soc., Faraday Trans. 1995,
91, 3041.
(24) Crampton, M. R.; Lord, S. D.; Millar, R. J. Chem. Soc., Perkin.
Trans. 1997, 909.
(25) Staab, H. A.; Datta, A. P. Angew. Chem. 1964, 3, 132.
(26) Libuda, H. G.; Zabel, F.; Fink, E. H.; Becker, K. H. J. Phys. Chem.
1990, 94, 5860.
(27) Kaiser, E. W.; Wallington, T. J. J. Phys. Chem. 1994, 98, 5679.
(28) Scollard, D. J.; Treacy, J. J.; Sidebottom, H. W.; Balestra-Garcia,
C.; Laverdet, G.; Lebras, G.; McLeod, H.; Teton, S. J. Phys. Chem. 1993,
97, 4683.
(29) Atkinson, R.; Ascmann, S. M. Int. J. Chem. Kinet. 1987, 19, 1097.
(30) Wallington, T. J.; Bilde, M.; Mogelberg, T. E.; Sehested, J.; Nielsen,
O. J. J. Phys. Chem. 1996, 100, 5751.
Acknowledgment. The authors thank Archie McCulloch,
Neil Winterton, and Alastair Kerr for useful discussions and
Rachel Woolley for help with some of the experimental work
on the partially chlorinated ethenes. We also thank NERC for
financial support of our work. A.S.H. is grateful to the same
body for the award of a research studentship.
(31) Mogelberg, T. E.; Bilde, M.; Sehested, J.; Wallington, T. J.; Nielsen,
O. J. J. Phys. Chem. 1996, 100, 18399.
(32) FACSIMILE program; UKAEA Harwell Laboratory, 1987.
(33) Wallington, T. J.; Hurley, M. D.; Fracheboud, J. M.; Orlando, J.
J.; Tyndall, G. S.; Sehested, J.; Mogelberg, T. E.; Nielsen, O. J. J. Phys.
Chem. 1996, 100, 18116.
(34) Bertrand, L.; Exsteen-Mayer, L.; Franklin, J. A.; Huybrechts, G.;
Olbrechts, J. Int. J. Chem. Kinet. 1971, 3, 89.
(35) Lissi, E. A.; Massiff, G.; Villa, A. Int. J. Chem. Kinet. 1975, 7,
625.
References and Notes
(1) McCulloch, A.; Midgley, P. M. Atmos. EnViron. 1996, 30, 601.
(2) Franklin, J. Toxicol. EnViron. Chem. 1994, 46, 169.
(3) Sidebottom, H.; Franklin, J. Pure Appl. Chem. 1996, 68, 1759.
(4) Nicovich, J. M.; Wang, S.; McKee, M. L.; Wine, P. H. J. Phys.
Chem. 1996, 100, 680.
(5) Tuazon, E. C.; Atkinson, R.; Aschmann, S. M.; Goodman, M. A.;
Winer, A. M. Int. J. Chem. Kinet. 1988, 20, 241.
(6) Frank, H.; Scholl, H.; Sutinen, S.; Norokorpi, Y. Ann. Bot. Fennici
1992, 29, 263.
(7) Kindler, T. P.; Chameides, W. L.; Wine, P. H.; Cunnold, D. M.;
Alyea, F. N. J. Geophys. Res. 1995, 100, 1235.
(8) Helas, G.; Wilson, S. R. Atmos. EnViron. 1992, 26A, 2975.
(9) Ugi, I.; Beck, F. Chem. Ber. 1961, 94, 1839.
(10) Itoh, N.; Kutsuna, S.; Ibusuki, T. Chemosphere 1994, 28, 2029.
(11) Singh, H. B.; Lillian, D.; Appleby, A.; Lobban, L. EnViron. Lett.
1975, 10, 253.
(36) Niki, H.; Chen, J.; Young, V. Res. Chem. Intermed. 1994, 20, 277.
(37) Lesclaux, R.; Dognon, A. M.; Caralp, F. J. Photochem. Photobiol.
A 1987, 41, 1.
(38) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F.; Kerr, J.
A.; Rossi, M. J.; Troe, J. J. Phys. Chem. Ref. Data 1992, 21, 5235.
(39) Weng, T. J.; Ho, T. I.; Su, T. M. J. Phys. Chem. 1987, 91, 5235.
(40) DeMore, W. B.; Golden, D. M.; Hampson, R. F.; Howard, C. J.;
Kolb, C. E.; Kurylo, M. J.; Molina, M. J.; Ravishankara, A. R.; Sander, S.
P. Chemical Kinetics and Photochemical Data for Use in Stratospheric
Modelling Number 12; JPL Publication, 1997; Vol. 97-4, 1.
(41) Russell, J. J.; Seetual, J. A.; Gutman, D.; Danis, F.; Caralp, F.;
Lightfoot, P. D.; Lesaclaux, R.; Melius, C. F.; Senkan, S. M. J. Phys. Chem.
1990, 94, 3277.
(12) Huybrechts, G.; Olbregts, J.; Thomas, K. Trans. Faraday Soc. 1967,
63, 1647.
(13) Singh, H. B.; Thaker, A. N.; Chen, E. Geophys. Res. Lett. 1996,
23, 1529.
(14) Rudolph, J.; Koppmann, R.; Plass-Dulmer, Ch. Atmos. EnViron.
1996, 30, 1887.