Rate constant and mechanism of the reaction between Cl and CH3OCl at 295 K

Article abstract of DOI:10.1021/jp9611075

The reaction between Cl atoms and CH₃OCl was investigated at 295 K in both air and N₂ bath gases at total pressures between 100 and 750 Torr by the relative rate method. The rate constant of the title reaction was found to be a factor 1.07 ± 0.02 (2σ) greater than that of Cl + C₂H₆ at room temperature and independent of pressure between 100 and 750 Torr. This yields a rate constant of (6.1 ± 0.6) × 10^-11 cm₃ molecule^-1 s^-1. The products of the reaction were detected by FTIR and UV absorption spectroscopy. Analysis of Cl₂ and HCl products allowed branching ratios of 0.2 ±0.1 for HCl + CH₂OCl formation and 0.8 ± 0.2 for Cl₂ + CH₃O formation to be determined. The high rate constant implies that reaction with Cl atoms is an important loss process for CH₃OCl in the polar stratosphere.

Full text of DOI:10.1021/jp9611075

J. Phys. Chem. 1996, 100, 17191-17201
17191
Rate Constant and Mechanism of the Reaction between Cl and CH3OCl at 295 K
S. A. Carl, C. M. Roehl,† R. Mu1ller,‡ G. K. Moortgat, and J. N. Crowley*
Max-Planck-Institut fu¨r Chemie, DiVision of Atmospheric Chemistry, Postfach 3060, D-55020 Mainz, Germany
ReceiVed: April 16, 1996; In Final Form: August 14, 1996X
The reaction between Cl atoms and CH3OCl was investigated at 295 K in both air and N2 bath gases at total
pressures between 100 and 750 Torr by the relative rate method. The rate constant of the title reaction was
found to be a factor 1.07 ( 0.02 (2σ) greater than that of Cl + C2H6 at room temperature and independent
of pressure between 100 and 750 Torr. This yields a rate constant of (6.1 ( 0.6) × 10-11 cm3 molecule-1
s-1. The products of the reaction were detected by FTIR and UV absorption spectroscopy. Analysis of Cl2
and HCl products allowed branching ratios of 0.2 ( 0.1 for HCl + CH2OCl formation and 0.8 ( 0.2 for Cl2
+ CH3O formation to be determined. The high rate constant implies that reaction with Cl atoms is an important
loss process for CH3OCl in the polar stratosphere.
1. Introduction
of suitable static gas mixtures and are designated CP-FTIR. UV
absorption experiments employed pulsed laser photolysis with
diode-array detection of absorption (LP-UV). These methods
are described separately below.
The presence of high concentrations of ClO and Cl in the
polar stratosphere during "ozone hole” conditions1,2 results in
an acceleration of the rate of oxidation of CH4 and an
increasingly important role for the reaction between CH3O2 and
ClO.3 Of the two reaction channels identified for reaction 1,
one generates CH3OCl4-8 with a branching ratio of close to
25% at 200 K.5
2.1. CP-FTIR Experiments. The experimental setup
consists of a cylindrical quartz reaction vessel of ca. 1.4 m length
and 20 cm diameter to give a volume of 44 L. The reaction
vessel is surrounded by photolysis lamps, mounted radially and
parallel to the vessels′ cylindrical axis, which emit between 280
and 350 nm (Philips TL 12) or at 254 nm (Philips TLUV). The
lamps are surrounded by a polished aluminum sheath and
provide an approximately homogeneous light flux throughout
the reaction volume. Two sets of multipass optics are built into
the cell, one for reflection in the IR and the other for UV
wavelengths, providing absorption pathlengths of ca. 34 and
9.8 m, respectively. IR analysis was carried out using a Bomem
DA-08 FTIR spectrometer operated at 0.5 cm-1 resolution. A
liquid-helium-cooled Cu-Ge detector was used to detect IR
radiation between 450 and 3600 cm-1; 256 scans were coadded
over a period of ca. 7 min to provide sufficient signal/noise
ratios. Detection of UV light (D2 lamp) was with a photomul-
tiplier tube at a single wavelength selected by a 0.25 m
monochromator (1.5 nm resolution).
CH3O2 + ClO f CH3OCl + O2
f CH3O + ClOO
(1a)
(1b)
In previous work we have examined possible removal
processes for CH3OCl and found photolysis to be important9
and reaction with OH to be negligible.10
Several exothermic pathways exist for reaction 2.
∆Hr/kJ mol-1
Cl + CH3OCl
f HCl + CH2OCl
f Cl2 + CH3O
f CH3Cl + ClO
-27
-43
-41
(2a)
(2b)
(2c)
The heats of reaction were calculated using ∆Hf(CH3OCl)
) -61 kJ mol-1 (calculated from bond additivity11) and
∆Hf(CH2OCl) ) +125 kJ mol-1, obtained by comparison with
CH3OH and CH2OH. ∆Hf data were taken from ref 12.
The goal of this study was to measure the rate constant for
reaction 2 at 295 K and conduct a product study, at the same
temperature, to determine the branching to each channel (k2a/
k2, k2b/k2, and k2c/k2).
Gases are introduced into the reaction chamber as diluted
mixtures in either Ar or N2 through a piping system that ensures
good mixing. Pressures were monitored by 1000 Torr and 100
mbar capacitance manometers; most experiments were carried
out at a total pressure of 750 Torr air or N2 and at room
temperature (295 K). Calibration curves of C2H6, HCl, HCHO,
CH3OH, and CO were obtained by measuring optical absorbance
in the 44 L cell following the dosing of 1-10 mbar amounts of
accurately premixed gases (with N2) from a 6 L glass vessel
and pressurizing to 750 Torr. HCOOH, calibrated accurately
in a separate experimental setup as described by Finkbeiner et
al.,13 was determined relative to HCHO using a relative
2. Experimental Section
The experiments designed to measure the room temperature
kinetics and product formation in the reaction of Cl atoms with
CH3OCl were conducted in two separate setups, one employing
Fourier transform infrared (FTIR) and the other employing
ultraviolet (UV) absorption spectroscopy as analytical tools. The
IR experiments involved the broad band, continuous photolysis
absorbance of HCOOH (1776 cm-1) to HCHO (1746 cm-1
)
equal to 1.72 at a resolution of 0.5 cm-1. The CH3OCl IR
absorbance was calibrated relative to the UV absorbance at 250
nm, which in turn was converted to concentration using the cross
sections given in ref 9 and the optical path length of 980 cm.
The conversion of the measured absorbance to concentration
was achieved by a stripping routine using calibration spectra
with similar ((20%) absorbance. For HCl, HCHO was first
stripped from the spectrum to enable between 3 and 5 HCl
rotational lines to be simultaneously analyzed to determine the
* Correspondence to this author.
† Present address: Jet Propulsion Laboratory, California Institute of
Technology, Pasadena, CA 91109.
‡ Institut fu¨r Stratospha¨rische Chemie (ICG 1), Forschungszentrum Ju¨lich
GmbH, D-52425 Ju¨lich, Germany.
X Abstract published in AdVance ACS Abstracts, October 1, 1996.
S0022-3654(96)01107-0 CCC: $12.00 © 1996 American Chemical Society

17192 J. Phys. Chem., Vol. 100, No. 43, 1996
Carl et al.
concentration. The following IR features were used to convert
absorbance to concentration: CH3OCl, 1006 cm-1; CH3OH
1034 cm-1; C2H6, 822 cm-1; CO, 2140 cm-1; HCHO, 1746
cm-1; HCl, 2885 cm-1; ClCHO, 1783 cm-1; HCOOH, 1776
cm-1. The concentrations from the IR experiments are expected
to be accurate to within (5% for stable, easily stored, and
calibrated species (CH3OCl, CH3OH, C2H6, CO, HCHO), (10%
for HCl, and (15% for HCOOH. The calibration of ClCHO
was performed in situ and has an estimated accuracy of (25%.
Cl2 was detected by absorption at 370 nm (when Cl2 was
present in the starting mixture) or at 330 nm. The change
in concentration of Cl2 (370 nm) during photolysis of Cl2/
CH3OCl/air(N2) was discernible by this method but could not
be accurately quantified. During the photolysis of CH3OCl/air
mixtures, Cl2 (330 nm) could be sufficiently accurately measured
to provide the branching ratio data directly. Additional details
of this apparatus are found elsewhere.14,15
glass bulb. C2H6 (Linde, 99.95%), HCl (Linde, 99.8%), Cl2
(Linde 99.8%), and CO (Linde, 99.997%) were used without
further purification. CH3OH (Aldrich, 99.9%) was degassed
prior to use. HCHO was generated by gently warming
paraformaldehyde (Aldrich, 99.5%) and purified of H2O and
polymeric HCHO by passage through a glass spiral held at
-100 °C, prior to dilution to 1% and storage at a low partial
pressure (1-2 Torr) in a blackened glass bulb. ClCHO was
generated in situ by photolysis of HCHO/Cl2 mixtures and
roughly calibrated by measuring the depletion of Cl2 and HCHO.
N2 (Linde 99.999%) and synthetic air (Linde) were taken directly
from 50 L bottles.
3. Results
3.1. Relative Rate Experiments. The room temperature
rate constant for the reaction between Cl atoms and CH3OCl
was investigated using the relative rate method and FTIR
detection of CH3OCl and C2H6, which was used as reference
gas. IR spectra of mixtures comprising Cl2 ((2.5-3.0) × 1016
molecules cm-3), CH3OCl ((1.9-3.7) × 1015 molecules cm-3),
C2H6 ((2.4-3.7) × 1015 molecules cm-3), and N2 or air (750
or 100 Torr) were illuminated at 280-350 nm for short periods
of typically 3-5 s before IR spectra were obtained. This
was repeated up to 11 times until the reactants CH3OCl and
C2H6 were reduced to ca. 15% of their starting concentrations.
High concentrations of Cl2 were employed to ensure that the
CH3OCl removal by photolysis was negligible compared to its
reaction with Cl.
2.2. LP-UV Experiments. The apparatus used for the LP-
UV absorption study of reaction 2 comprised a parallelpiped
quartz reaction/absorption cell of square internal cross section
21.1 cm2 and volume 850 cm3, a pulsed laser, operating at 351
nm (XeF Eximer Laser; 25 ns pulse duration, ca. 1 nm
bandwidth) used as the photolysis source, a deuterium lamp for
long-path UV absorption measurements (White optics, l ) 3.26
m), and a 0.5 m spectrograph with a gated (gate width ) 65
µs), Peltier-cooled, diode-array detector with image intensifier
for detection. The grating of the spectrograph was positioned
to provide radiation from 235 to 350 nm over the detector face,
encompassing the main UV spectral features of CH3OCl and
Cl2. This arrangement provided a wavelength-digitized spec-
trum of effective spectral resolution 1.7 nm. To obtain a
satisfactory signal-to-noise ratio, each recorded spectrum was
derived from an average of 500 gated diode-array images and
took ca. 6 s to collect. The spectrograph was calibrated for
wavelength using spectral lines of a Hg pen-ray lamp.
Cl2 + hν f Cl + Cl
Cl + C2H6 f HCl + C2H5
Cl + CH3OCl f products
CH3OCl + hν f CH3O + Cl
(3)
(4)
(2)
(5)
The laser beam was expanded to give approximately uniform
intensity over one of the cell’s rectangular faces. Cl2 was used
as the photolytic precursor for Cl atoms (σ(Cl2)351nm ) 1.82 ×
10-19 cm2 molecule-1 12) and was admitted to the cell in a
mixture of 1% Cl2 (Linde 3.5) in N2 (Linde 5.0). The second
reactant was introduced to the cell as 1.6% CH3OCl in Ar (with
2.8% CH3OH impurity in CH3OCl) from a 6 L, blackened,
Pyrex bulb. Total cell pressures were constant during each
experiment but were varied from experiment to experiment
between 60 and 500 Torr by adding N2 (Linde 5.0). Pressures
in the cell were measured by a 1000 Torr capacitance manom-
eter. Initial concentrations of both CH3OCl and Cl2 were varied
between (0.5-1.4) × 1016 and (0.8-1.8) × 1016 molecules
cm-3, respectively. Between 0.03% and 0.06% of the Cl2 was
photolyzed per laser pulse. CH3OCl was also photolyzed to
some extent by the laser (σ(351 nm) ) 6.3 × 10-21 cm2
molecule-1 9), but its photolytic removal was calculated to be
less than 2% of its removal due to reaction with Cl. The LP-
UV experiments measured the overall change in concentrations
of the detectable reactant, CH3OCl, and direct product Cl2,
together with the stable products produced by secondary
reactions, namely, ClCHO and HCHO, following initiation of
reaction 2 by photolysis. To easily discern changes in concen-
trations, particularly of Cl2, the gaseous mixture was photolyzed
50 times (ca. 140 or 280 mJ/pulse) before a spectrum was
recorded. Changes in concentrations of relevant species were
extracted by computer fitting to known spectral profiles of the
same resolution.
Assuming that the Cl atom production is dominated by Cl2
photolysis and that all Cl reacts with CH3OCl, the approximate
rate of removal of CH3OCl due to photolysis relative to reaction
with Cl atoms, R5/R2, is, in the absence of C2H6, given by
R5/R2 ) J(CH3OCl)[CH3OCl]/2J(Cl2)[Cl2]
(i)
Here J(CH3OCl) and J(Cl2) refer to the photolysis rate
constants of CH3OCl and Cl2, respectively. The ratio
J(CH3OCl)/J(Cl2), obtained by comparing the integrated overlap
of the known CH3OCl9 and Cl2 spectra12 with the relative
emission intensity of the photolysis lamps, is equal to 0.076.
Assuming typical initial concentrations of 2.5 × 1015 and 3.0
× 1016 molecules cm-3 for CH3OCl and Cl2, respectively, we
obtain R5/R2 ) 0.3%. In the presence of approximately equal
concentrations of C2H6 and CH3OCl the loss rate of CH3OCl
due to reaction with Cl is halved, and we obtain R5/R2 ) 0.6%.
Photolysis of CH3OCl is thus treated as negligible in the
following analyses.
The relative rate method to acquire kinetic data does not
require knowledge of the absolute concentrations of either
radical or reactants but a reference compound with a well-
characterized rate constant at the temperature of interest and
an accurate measure of the fractional loss of both stable reactant
and reference molecules. In this study, C2H6 was chosen as
the reference molecule. The rate constant used for Cl + C2H6
has a recommended value12 of kref ) (5.7 ( 0.6) × 10-11 cm3
molecule-1 s-1. The relative rate constant, k2/kref, is derived
from plots of ln(DFreactant) versus ln(DFreference), where DF is
2.3. Chemicals. CH3OCl was prepared as described previ-
ously9 and stored as a ca. 2% mixture in Ar in a blackened 6 L

Reaction between Cl and CH3OCl at 295 K
J. Phys. Chem., Vol. 100, No. 43, 1996 17193
TABLE 1: Experimental Starting Conditions for the
Relative Rate Studya
[Cl] and does not contribute significantly to removal of either
C2H6 or CH3OCl.
An examination of all four experiments in O2 reveals that
the experiments in which the ratio of CH3OCl to C2H6 was
lowest (open circles in Figure 1) yield the highest curvature.
This may imply that the secondary loss of CH3OCl is tied in
with the enhanced C2H6 oxidation chemistry (RO2 radical
chemistry). Direct reaction of peroxy radicals with CH3OCl
seems however unlikely considering the negligible reactivity
of HO2 toward HOCl or CH3O2 to CH3OCl.4,5 At present, we
do not have an explanation for the unexpected behavior in C2H6/
CH3OCl/air.
bath gas
[CH3OCl]0
[C2H6]0
[Cl2]0
air
air
air
air
N2
N2
2.46
2.71
1.93
3.66
2.59
3.19
3.11
2.43
2.38
3.69
2.36
3.04
4.18
4.37
27.2
29.2
29.2
29.9
25.5
24.3
25.0
b
N2
a Concentrations are in units of 1015 molecules cm-3
at 100 Torr total pressure; all other experiments were conducted at a
total pressure of 750 Torr.
.
b Carried out
In N2, the decay rate of CH3OCl and C2H6 was enhanced by
ca. 40% over that in O2 for the same initial Cl2 concentration
and number of photolysis lamps. This is attributed to secondary
generation of Cl atoms through reactions of the sort
Cl2 + C2H5 f Cl + C2H5Cl
(6)
The data obtained in N2 were fitted using a weighted least
squares fitting routine and yield a result of 1.07 ( 0.02 (the
error represents 2σ, statistical confidence limit). Using the
recommended rate constant for Cl + C2H6,12 this yields a rate
constant of k2(295) ) 6.1 ( 0.6 cm3 molecule-1 s-1
.
3.2. Broad Band Photolysis of Cl2/CH3OCl/Air. Product
formation in the reaction between Cl and CH3OCl was
investigated by measuring the yields of several IR-active
products and Cl2 following the broad band photolysis of Cl2/
CH3OCl/air mixtures at 750 Torr and 295 K. IR spectra were
obtained prior to reaction and at various intervals following
photolysis using 4 or 12 TL12 lamps. A high concentration of
Cl2 was used as a Cl atom source, and the stability of the UV
detection (at 370 nm) was too low for modulation in the Cl2
signal to be observed. This was only possible in those
experiments in which CH3OCl was photolyzed at 254 nm in
the absence of initial Cl2, enabling it to be monitored at the
peak of its absorption spectrum at 330 nm (see section 3.4).
Typical starting concentrations were Cl2, (0.6-2.9) × 1016
molecules cm-3 (detected by UV absorption at 370 nm);
CH3OCl, (2.1-2.5) × 1016 molecules cm-3; air/N2 to 750 Torr.
CH3Cl was not detected in any experiments. As its reac-
tion with Cl atoms is too slow to cause its removal, reaction 2c
is insignificant. Analysis of the IR spectrum near 750 cm-1
(C-Cl stretch) enables us to estimate an upper limit of 1% for
CH3Cl formation. The initial reaction is therefore expected to
be a competition between channels 2a and 2b:
Figure 1. Relative rate study of the reaction between Cl and
CH3OCl. DF is the depletion factor and is defined in the text. Open
squares, experiments in 750 and 100 Torr N2. Circles, experiments in
air. Open circles, [CH3OCl]i ) 1.93 × 1015, [C2H6]i ) 3.69 × 1015
molecules cm-3. Solid circles, [CH3OCl]i ) 2.71 × 1015, [C2H6]i )
2.38 × 1015 molecules cm-3
.
the depletion factor and is defined as DF ) (initial concentra-
tion)/(final concentration).16 The initial concentrations for seven
experiments are listed in Table 1.
Figure 1 shows such a plot for three experiments carried out
in N2 (open squares) and two carried out in air (circles). The
data obtained in N2 displays the expected behavior, with all
data points lying on a single straight line with an unforced zero
intercept. The error bars were derived from a propagation of
the errors associated with each experimental point and are largest
at high conversions where the concentrations are lowest. In
contrast, the data obtained in air give strongly curved plots, with
the degree of curvature varying from experiment to experi-
ment. In air, the relative removal rate of CH3OCl to C2H6
increases with time. Indeed at long times (i.e. last point of open
circles) about twice as much CH3OCl has been removed relative
to C2H6 as compared to the N2 experiments at similar conver-
sion. This may imply that a further removal process for
CH3OCl apart from reaction with Cl becomes more important
as the degree of conversion of reactants increases or that C2H6
is reformed.
In the presence of O2 the chemistry will be dominated by
formation of peroxy radicals such as C2H5O2 and HO2 and their
subsequent reactions to form peroxidic and aldehydic products
such as C2H5OOH, CH3CHO, and H2O2. In such chemical
systems, OH radicals are likely to be generated via photolysis
of, for example, C2H5OOH or H2O2. As the reactivity of OH
toward CH3OCl is a factor 3 higher than toward C2H6 at 295
K,10 OH may contribute to the excess CH3OCl removal.
However, numerical simulations, incorporating a full set of
reactions between the various peroxy radicals, and photolysis
of photolabile species were unable to generate concentrations
of OH that were sufficient to reproduce the curvature of the
plots. [OH] typically remained 2 orders of magnitude less than
Cl + CH3OCl f HCl + CH2OCl
f Cl2 + CH3O
(2a)
(2b)
The subsequent secondary chemistry, involving CH2OCl and
CH3O from reactions 2a and 2b, respectively, is dependent on
the bath gas used. In the presence of O2 we expect the following
reactions to be important:
CH3O + O2 f HCHO + HO2
HO2 + HCHO f f f HCOOH + products
CH2OCl f HCHO + Cl
(7)
(8)
(9)
CH2OCl + O2 f products
(10)
The products detected in air were HCl, Cl2, HCHO, CO, and
HCOOH. Both CH3OH and CO were present at the beginning
of the experiment as impurities in either the bath gas or the

17194 J. Phys. Chem., Vol. 100, No. 43, 1996
Carl et al.
Figure 2. FTIR spectrum of products (upward pointing features) and
depleted reactants (downward pointing features) following 15 s of
photolysis of 2.07 × 1015 molecules cm-3 CH3OCl and 0.97 × 1016
molecules cm-3 Cl2 in air at 750 Torr total pressure.
CH3OCl. Of central importance to the subsequent analyses,
which compare rates of formation of products from channels
2a and 2b with the rate of loss of CH3OCl and with each other,
is that total chlorine is conserved, i.e., that there are no loss
processes for these molecules that result in undetected chlorine-
containing products. Therefore the chlorine balance was
calculated after about 30-50% of the CH3OCl had been
depleted. In all experiments the chlorine balance was found to
be close to 100% (variation between 102 and 107%), confirming
that all chlorine-containing products are detected. The con-
centration of CH3OCl in the darkened 44 L reactor did not
change over 30 min, implying that thermal dissociation or loss
at the reactor wall was negligible over the course of a photolysis
experiment.
Figure 2 shows a spectrum obtained after 15 s of photolysis,
in which ca. 25% of the CH3OCl has been depleted. IR features
pointing downward are reactants that are consumed; features
pointing upward are products. Although [CO2] increases with
time, it is not thought to be a product of the gas-phase chemistry
but a result of degassing from the reactor walls under the
influence of UV light.
Figure 3 shows concentration-time profiles for CH3OCl,
HCl, HCHO, CH3OH, CO, and Cl2. The Cl2 profile, although
not accurately measured, shows that [Cl2] increases during the
course of the experiment, providing the first qualitative con-
firmation that reaction 2b is taking place. Although the
conversion of CH3O to HCHO is rapid in 750 Torr of air, HCHO
reacts with the concomitantly formed HO2, and the HCHO
profile is curved from the first measurement onward. If we
make the assumption that all HCHO that is lost early in the
experiment has been converted to HCOOH (reaction 8), then
the sum of HCHO and HCOOH should equal the total amount
of HCHO generated. The HCHO, HCOOH, and HCHO +
HCOOH concentrations of experiment 5 (see Table 2) are
plotted in Figure 4. The summed HCHO + HCOOH concen-
tration shows a linear increase until about 20 s (first four data
points). After this, the removal of HCHO by Cl atoms (which
does not lead to HCOOH formation) results in curvature.
HCHO photolysis is negligible. The rate of increase of HCHO
+ HCOOH over this period is equal to (3.01 ( 0.08) × 1013
molecules cm-3 s-1. The rate of decay of CH3OCl is (3.14 (
0.32) × 1013 molecules cm-3 s-1 (both errors are 2σ), yielding
a branching ratio to HCHO + HCOOH formation of 0.96 (
Figure 3. Concentration-time profiles of reactants and products from
photolysis of Cl2/CH3OCl/air. Initial conditions are as in Figure 2.
The solid lines are numerical simulations of the data and are described
in the text.
0.10. The results of experiment 6 were analyzed in the same
manner and yield 0.89 ( 0.10. An average value from the two
experiments is thus 0.93 ( 0.14 and provides confirmation of
the quantitative conversion of CH3OCl to HCHO as predicted
by reactions 2a, 2b, 7, and 9. This suggests that the fate of
CH2OCl is adequately described by its thermal decomposition
or, if CH2OCl does not react to form HCHO, that branching to
this channel is too small to effect the overall HCHO yield.
In the above, we have assumed quantitative conversion of
HCHO to HCOOH, which implicitly assumes that any HOCH2-
OOH generated (12b) is lost, presumably in a heterogeneous
reaction (15) to give HCOOH. The following reaction scheme17,18
describes the formation of HCOOH:
HO2 + HCHO S HOCH2O2
(11)
HOCH2O2 + HO2 f HCOOH + H2O + O2 (12a)
HOCH2O2 + HO2 f HOCH2OOH + O2
2HOCH2O2 f 2HOCH2O + O2
(12b)
(13a)
(13b)
(14)
2HOCH2O2 f HCOOH + HOCH2OH
HOCH2O + O2 f HCOOH + HO2
HOCH2OOH + wall f HCOOH + H2O
(15)
Any HOCH2OOH that is not converted to HCOOH means
that the summed HCHO + HCOOH concentration is underes-

Reaction between Cl and CH3OCl at 295 K
J. Phys. Chem., Vol. 100, No. 43, 1996 17195
TABLE 2: Summary of Data Obtained in the FTIR Experimentsa
expt no.
5
2
3
4
6
7
8
bath gas
N2
N2
N2
air
air
air
2.12
0.025
2.87
3(Hg)
air
5.33
0.046
4.38
4(Hg)
1015 [CH3OCl]i
1016 [Cl2]i
2.49
2.53
1.85
2.44
2.50
2.16
2.19
0.57
1.70
2.07
0.97
1.49
12
2.42
2.90
2.97
4
1014 [CH3OH]
no. lamps
12
12
12
d[CH3OCl]/dt (A)
1012 molec cm-1 s-1
-89.0
(6.0(5)
-67.0
(4.0(5)
-26.6
(0.7(5)
-31.4
-34.2
-1.72
-4.20
(3.2(5)
(2.8(5)
(0.06(5)
(0.06(4)
d[HCl]/dt (B)b
1011 molec cm-1 s-1
101
135
3.5
5.9
(8(5)
(26(5)
(0.2(3)
(0.7(3)
d[HCHO]/dt (C)
1013 molec cm-1 s-1
3.46
3.18
1.27
(0.1(4)
(0.3(4)
(0.02(4)
d[HCHOtot]/dt (D)c
1013 molec cm-1 s-1
3.01
(0.08(4)
3.05
(0.10(4)
d[CH3OH]/dt (E)
1012 molec cm-1 s-1
17.0
(1.0(5)
13.5
(2.0(5)
6.6
(1.0(5)
-2.2
(0.6(3)
-5.3
(1.5(4)
-0.25
-0.21
(0.03(5)
(0.02(4)
d[Cl2]/dt (F)
1011 molec cm-1 s-1
8.35
27.2
(1.1(3)
2.6(3)
G/Ad
C/A
D/A
E/A
B/F
0.25
0.96
0.24
0.89
0.39
0.19
0.46
0.20
0.49
0.25
0.12
0.14
a Numbers in parentheses are the number of points through which the slope was determined. Negative sign implies depletion; errors are 2σ
statistical uncertainty. b Not corrected for production via Cl + CH3OH. c HCHOtot ) HCHO + HCOOH. d HCl production rate corrected for Cl
+ CH3OH.
consideration of the energetics of possible pathways for reaction
with O2:
CH2OCl + O2 f HO2 + CHOCl
f HCHO + ClO2
(10a)
(10b)
Reaction 10a is expected to be highly endothermic, whereas
reaction 10b is calculated as exothermic by greater than 100
kJ/mol. However, the fate of ClOO at 295 K and 760 Torr is
rapid decomposition to Cl and O2, and thus the direct thermal
decomposition of CH2OCl and reaction 10b give the same net
result. In the following analyses we therefore assume that
CH2OCl decomposes immediately to HCHO and Cl.
Rates of formation of products relative to the removal rate
of CH3OCl are listed in Table 2 along with the starting
conditions for five experiments (nos. 2-6) carried out in air
and N2. The rate of HCl formation relative to CH3OCl
decomposition had to be corrected for the initial presence of
CH3OH. In all experiments, CH3OH impurities were between
5 and 10% of the CH3OCl concentration. As CH3OH is
removed only by reaction with Cl and is not generated in the
air system, the correction was simply achieved by subtracting
the initial CH3OH decay rate from the initial HCl produc-
tion rate. The relative rate of production of HCl to loss of
CH3OCl (Table 2, experiments 5 and 6) is then 0.25 ( 0.04,
where the errors are propagated from 2σ errors in d[CH3OCl]/
dt, d[CH3OH]/dt, and d[HCl]/dt as given in Table 2.
Figure 4. Concentration-time profiles of CH3OCl, HCHO, HCOOH,
and HCHO + HCOOH in air. The solid lines are initial slopes of the
HCHO + HCOOH production rate and the CH3OCl removal rate.
timated. The IR spectra were inspected for the presence of
HOCH2OOH by comparison with the published spectra (feature
at ca. 1150 cm-1) of Su et al.19 Although no evidence for
HOCH2OOH was found, the feature at 1150 cm-1 is weak,
broad, and in a region where CH3OH and HCOOH have very
strong interfering absorptions. For this reason we cannot rule
out its presence and prefer to quote the summed HCHO +
HCOOH concentration and the 93% conversion of CH3OCl to
HCHO as a lower limit.
Figure 3 shows that, in experiments carried out in air, CO is
formed at low conversion of CH3OCl. The CO is formed in
the secondary reaction between Cl and HCHO in air,
In reaction 9 we assume that the fate of CH2OCl in air is
thermal decomposition. This assumption is based on an upper
limit to the thermal lifetime of a similar radical, CH2OOH, which
decomposes within 20 µs in 50 Torr of He at 205 K20 and
Cl + HCHO f HCl + CHO
CHO + O2 f HO2 + CO
(16)
(17)

17196 J. Phys. Chem., Vol. 100, No. 43, 1996
Carl et al.
and indicates that secondary chemistry may influence even our
first measurement point. This makes it very difficult to measure
HCl before secondary chemistry contributes to its formation.
Assuming an initial concentration of CH3OCl of 2.5 × 1015
molecules cm-3, a minimum conversion of 5% of the CH3OCl
is needed to obtain a usable signal of HCl. This simply reflects
the low branching ratio to HCl formation. That is, a 10%
branching ratio and 5% conversion of CH3OCl give ca. 1 ×
1013 molecules cm-3 of HCl, which is close to the detection
limit in our setup. This problem is compounded by the high
reactivity of Cl to HCHO, which is formed as a major product
via reaction 7 and which forms HCl with 100% efficiency. Thus,
5% conversion of CH3OCl implies a HCHO concentration
roughly 5% that of CH3OCl. This will scavenge ca. 5% of
subsequent Cl atoms and contribute 50% of HCl at this early
stage of the reaction. Use of significantly higher concentrations
of CH3OCl to avoid such effects results in saturation of the IR
signal. For these reasons, the results were checked by numerical
simulation of a model containing the reactions described in the
body of the text.
3.3. Numerical Simulations of Product Formation in Air.
The numerical simulations21 were carried out primarily in order
to assess to what extent HCl formation is influenced by the
reactions of HCHO and CH3OH with Cl at short reaction times.
The rate of photolysis of Cl2, J(Cl2), is required in such
simulations and was determined by monitoring with FTIR the
rate of formation of HCl and depletion of C2H6 when static
Cl2/C2H6 mixtures were photolyzed in air. The following
reactions are expected to take place:
formation. This is lower than the value of 24% obtained using
initial slopes (Table 2), as the formation of HCl via reactions
of Cl with HCHO and CH3OH is rigorously taken into account.
In such simulations, the decay rate of CH3OCl is controlled by
the available Cl atom concentration. This not only is governed
by the photolysis rate constant of Cl2, J(Cl2), and the Cl2
concentration but also depends on the branching ratio to
CH2OCl formation, which results in secondary Cl atom produc-
tion. Therefore not only k2a/k2 and k2b/k2 but also J(Cl2) was
optimized. The best fit value of Jfit(Cl2) was 1.65 × 10-3 s-1
.
This is lower than but within the uncertainty of our independent
J(Cl2) determinations. In the present model, HCHO is generated
in each reaction of Cl with CH3OCl and is therefore not very
sensitive to branching ratio. The HCHO and CO profiles are
however adequately reproduced using this mechanism and
branching ratio of 0.18 to HCl formation. To test the sensitivity
of the data to changes in branching ratio, values of k2a/k2 of 0.1
and 0.3 were fixed, and the J(Cl2) necessary to fit the CH3OCl
data was varied. The results of these simulations are shown in
Figure 3 as dotted lines (k2a/k2 ) 0.1, Jfit(Cl2) ) 1.8 × 10-3
s-1) and dashed lines (k2a/k2 ) 0.3, Jfit(Cl2) ) 1.6 × 10-3 s-1
)
enveloping the HCl profile.
3.4. Photolysis of CH3OCl/Air at 254 nm. Experiments
in which mixtures of CH3OCl in air were photolyzed at 254
nm were carried out with the aim of measuring simultaneously
the rate of generation of both Cl2 and HCl. The advantage of
this method is that the rates of formation of the HCl and Cl2 do
not have to be related to the loss of CH3OCl and that the
modulation of the Cl2 signal at 330 nm is particularly strong as
there is no offset due to a large initial concentration as in the
experiments described above. However, as CH3OCl absorbs
at 330 nm, its depletion contributed to the change in absorbance
upon photolysis. Therefore CH3OCl profiles measured in the
IR were used to correct the absorbance measurements at 330
nm before they could be converted to Cl2 concentrations.
Cl2 + hν f Cl + Cl
Cl + C2H6 f HCl + C2H5
C2H5 + O2 + M f C2H5O2 + M
(3)
(4)
(18)
Table 2 lists the initial slopes for HCl and Cl2 formation from
experiments 7 and 8. The averaged ratio, d(HCl)/d(Cl2), for
the first three data points (200 s) is 0.14. After this period the
ratio increases as Cl begins to react with HCHO and generate
secondary HCl. The initial value of d(HCl)/d(Cl2) corresponds
to a branching ratio for k2a/k2 ) 0.12 ( 0.03 or k2/k2 ) 0.88 (
0.20.
The measured rate of formation of HCl or depletion of C2H6
is then given by d(HCl)/dt ) -d(C2H6)/dt ) 2J(Cl2)[Cl2]i. The
initial Cl2 concentration can be determined accurately by UV
absorption spectroscopy. The HCl and C2H6 profiles thus
obtained were slightly curved due to the changing rate of Cl
atom production as the Cl2 concentration is depleted. Initial
slopes could be determined only with limited accuracy, and it
was necessary to model the HCl formation and C2H6 decay using
a numerical simulation initiated with the above [Cl2]i and [C2H6]
concentrations. On the basis of HCl production, J(Cl2) was
found to be (1.27 ( 0.06) × 10-3 s-1 when the lamps were
switched on for a nominal 3 s photolysis period. The value
obtained by analyzing the C2H6 profile was (1.13 ( 0.06) ×
10-3 s-1. For 5 s photolysis periods, values of (2.01 ( 0.12)
× 10-3 s-1 (HCl production) and (1.86 ( 0.15) × 10-3 s-1
(C2H6 decay) were obtained. The different values for J(Cl2)
for 3 and 5 s photolysis periods are due to flickering of the
lamps when switching on. The errors given in the J(Cl2) value
represent only the goodness of the fit to the HCl and C2H6
profiles. A more realistic error includes the expected uncertainty
in calibration and measurement of Cl2, C2H6, and HCl and
approaches 20%. Although of negligible influence, the pho-
tolysis rates of CH3OCl and HCHO, calculated using known
absorption cross sections and quantum yields, were included in
the numerical simulations.
3.5. CP-FTIR and LP-UV Experiments in N2. In the
absence of O2 as a scavenger for CH3O a complex radical chem-
istry evolves. The chemistry may be divided into two phases:
an initial phase where HCHO is generated rapidly relative to
HCl and a second phase where this behavior is reversed. At
the onset of the second phase, CO and ClCHO are formed. If
the reaction is allowed to proceed further, only HCl and CO2
remain as products, which are unreactive toward Cl.
The observed end products (after ca. 50% depletion of
CH3OCl) in the IR experiments were Cl2, HCl, HCHO,
CH3OH, ClCHO, and CO. Concentration-time profiles of these
species are shown in Figure 5. The following mechanism
qualitatively describes their formation:
Cl + CH3OCl f HCl + CH2OCl
f Cl2 + CH3O
(2a)
(2b)
CH2OCl f HCHO + Cl
CH3O + CH3O f CH3OH + HCHO
HCHO + Cl f HCl + HCO
(9)
(19)
(16)
The solid lines in Figure 3 show the result of fitting the
experimental CH3OCl and HCl profiles to a reaction scheme
consisting of reactions 2a, 2b, 3, 5, 7, 9, and 11-17). The
simultaneous fit to the CH3OCl and HCl data returns a branching
ratio of k2a/k2 of 0.18 ( 0.02, i.e. 18% branching to HCl

Reaction between Cl and CH3OCl at 295 K
HCO + Cl2 f ClCHO + Cl
J. Phys. Chem., Vol. 100, No. 43, 1996 17197
(20)
(21)
HCO + CH3O f CO + CH3OH
Figure 5 shows how the HCl and ClCHO production rates
increase significantly after ca. 20 s of photolysis. At this point
the CH3OCl depletion rate increases and the net production of
Cl2 is reversed as phase 2 sets in. At the same time the
formation of HCHO is partially offset by its removal by reaction
with Cl atoms, which leads, via an unknown secondary reaction,
to CO generation (reaction 21 has a very large rate constant,
but the fate of HCO is likely to be dominated by (20)).
In Table 2 are listed the rates of change of concentration for
the CP-IR experiments in N2. Unfortunately, the correction to
d(HCl) as applied in the experiments in air cannot be undertaken
here, as CH3OH is formed in reaction 19 and the d(HCl)/
d(CH3OCl) analysis is not appropriate. Some sensitivity to the
branching ratio is expected to be found in the relative rates of
HCHO production and CH3OCl depletion. According to the
above mechanism, the branching ratio k2b/k2 is given by
k2b/k2 ) -2[(d(HCHO)/d(CH3OCl) - 1]
(ii)
at low conversion of CH3OCl. However, the measured value
of d(HCHO)/d(CH3OCl) is equal to 0.45 (Table 2) and leads to
a branching ratio in excess of unity.
The CH3OH profile should also reflect branching to channel
2b in this system. The branching ratio k2b/k2 is given by
k2b/k2 ) 2d(CH3OH)/d(CH3OCl)
(iii)
From Table 2 we see that d(CH3OH)/d(CH3OCl) ) 0.2, resulting
in a value of k2b/k2 of 0.4.
Figure 5. Concentration-time profiles of CH3OCl, HCHO, Cl2, HCl,
ClCHO, CH3OH, and CO in N2 (experiment 4). Initial concentrations
are given in Table 2.
Thus we derive a branching ratio of about 0.4 from the
CH3OH growth curve, which is inconsistent with the observation
of Cl2 formation in the N2 system, and derive branching ratios
in excess of unity from the HCHO profile. This leads us to
conclude that the above mechanism describing the reactions of
CH3O and CH2OCl in N2 may be incomplete. We come back
to this point later.
In the LP-UV experiments Cl2 could be accurately measured.
Figure 6 shows a series of spectra obtained from a single, static
gaseous mixture that had been exposed to 10 sets of 40
photolysis pulses (every alternate spectrum has been omitted
for clarity). Absorption below 260 nm is approximately
representative of [CH3OCl], whereas absorption around 330 nm
is approximately representative of [Cl2]. The initial spectrum
was taken before the first laser pulse and therefore shows initial
absorption due to Cl2 and CH3OCl. The later spectra, taken at
40 pulse intervals, increasingly show structure above 280 nm
due to formation of HCHO. ClCHO was also detected in these
experiments; it has a broad, structured absorption with a
maximum, in the near UV, around 260 nm,22 but is difficult to
discern in Figure 6. Qualitatively, one may observe an overall
increase in Cl2 concentration with the number of laser pulses
and a decrease in CH3OCl concentration. These spectra were
numerically fitted to known absorption profiles of the four
molecules mentioned above9,22-24 to obtain their absolute
concentrations. Concentrations of CH3OCl, Cl2, ClCHO, and
HCHO plotted as a function of number of photolysis pulses
are shown in Figure 7. The slopes of plots, such as those shown
in Figure 7, derived from experiments where competition for
Cl atoms from HCHO and CH3OH is observed to be negligible,
may be taken to represent the changes in concentrations of Cl2,
CH3OCl, and HCHO for a single laser pulse.
Figure 6. UV spectra collected at intervals of 80 photolysis pulses
during a LP-UV experiment. The structure above 270 nm in the latter
spectra is due to absorption by HCHO.
The results from several experiments are listed in Table 3
along with initial concentrations of CH3OCl and Cl2. A simple
analysis is carried out in which the initial rate of CH3OCl
removal is compared to the initial rate of production of Cl2. A

17198 J. Phys. Chem., Vol. 100, No. 43, 1996
Carl et al.
correction is applied to take into account Cl2 removed by laser
photolysis, and the branching ratio is, to a first approximation,
given by
k2b/k2 ) (∆(Cl2) + ∆(CH3OCl)/2)/∆(CH3OCl) (iv)
yielding an average branching ratio of k2b/k2 ) 0.78 ( 0.10.
As in the FTIR experiments, the HCHO data show low
∆(HCHO)/∆(CH3OCl) ratios, in this case consistently close to
30%. Use of eq ii then yields branching ratios much in excess
of unity and in conflict with the Cl2 data.
The measurements of Cl2 production in the experiments
carried out in N2 support the conclusion of a high branching
ratio (close to 80%) to reaction 2b. The complexity of the
secondary chemistry in the absence of O2 renders numerical
modeling of the concentration profiles gained in the LP-UV
and CP-FTIR experiments very difficult and imprecise. Experi-
ments were carried out in air in the LP-UV setup but were beset
by an experimental artifact that resulted in a chlorine balance
of between 120 and 200%, depending on the initial O2 partial
pressure. We do not have an adequate explanation for this
behavior but cannot rule out that the high intensity of the laser
light may result in the generation and photodesorption of
chlorine originating from a reservoir species from the reactor
walls.
Up to now we have based our analysis on the assumption
that CH2OCl formed in reaction 2a decomposes thermally to
yield Cl and HCHO. We note however that in the N2
experiment the CH3OH and HCHO yields are then not consistent
with a branching ratio of 0.8 in channel 2b that is derived from
the d(HCl)/d(CH3OCl) and d(HCl)/d(Cl2) ratios in air and the
d(HCl)/d(Cl2) ratio in N2. Some of this may conceivably be
due to losses of CH3O at the reactor surface, and the different
HCHO/CH3OCl ratios obtained in the CP-IR (0.45) and LP-
UV experiment (0.3) may simply reflect the different surface
to volume ratios and pressures used in the two reactors. In the
N2 system, we might also regard the perturbed CH3OH and
HCHO profiles as stemming from reactions that interfere with
the CH3O self-reaction. If we assume that CH2OCl is thermally
stable, we expect the following reactions to be fast:
Figure 7. Concentration profiles of reactants and products during a
LP-UV experiment.
On the other hand, a survey of the IR spectra close to
750 cm-1 (C-Cl stretching region) revealed absorption due
only to ClCHO, and we have no evidence for ClCH2OCl
formation.
In air a thermally stable CH2OCl could react with O2:
CH2OCl + O2 f CHOCl + HO2
f HCHO + Cl + O2
(10a)
(10b)
(10c)
CH2OCl + O2 + M f ClOCH2O2 + M
As already discussed, reaction 10a is endothermic and reaction
10b has the same net effect as thermal dissociation of CH2OCl.
Reaction 10c forms a peroxy radical, which would probably
react with HO2 to form a peroxide:
ClOCH2O2 + HO2 f ClOCH2OOH + O2
(25)
This would however not affect our analysis of CH3OCl, HCl,
or Cl2 profiles.
3.6. Summary of Branching Ratio Data. The results
obtained provide a fairly consistent picture of product forma-
tion in the reaction between Cl and CH3OCl and show that
HCl formation via H atom abstraction is the minor channel.
Analysis of the data obtained by photolysis of Cl2/CH3OCl/air
mixtures was carried out by comparing initial rates of HCl
formation to decay rates of CH3OCl. A result of k2a/k2 of 0.25
is thus obtained. This value is corrected downward to about
k2a/k2 ) 0.18 using numerical simulations that account for HCl
formation via the reaction between Cl and HCHO. Analysis
of total carbonyl bond containing molecules (HCHO and
HCOOH) was carried out and showed that CH3OCl is stoichi-
ometrically converted to HCHO in air. For experiments in
which CH3OCl was photolyzed at 254 nm in air, Cl2 and HCl
profiles could be analyzed to give the ratio k2a/k2b directly. This
was found to be 0.13 and yields branching ratios of k2a/k2 )
0.12 and k2b/k2 ) 0.88. In the LP-UV experiments in N2, Cl2
profiles could be analyzed to give a branching ratio of k2b/k2 of
0.78.
CH2OCl + CH3O f CH3OCl + HCHO
f HCl + 2HCHO
(22a)
(22b)
(23)
CH2OCl + CH2OCl f 2HCHO + Cl2
CH2OCl + Cl2 f ClCH2OCl + Cl
(24)
Using a typical Cl2 concentration of about 1 × 1016 molecules
cm-3 and a typical alkyl radical + Cl2 rate constant (1 × 10-12
cm3 molecule-1 s-1), we derive a loss rate of 1 × 104 s-1 for
CH2OCl or a chemical lifetime of about 100 µs due to reaction
24, which dominates over the radical-radical reactions (22, 23)
in the CP-IR experiments. In the pulsed LP-UV experiments
the peak radical densities will be higher and the flux through
reactions 22 and 23 will increase relative to reaction 24. This
is a possible explanation for the different d(HCHO)/d(CH3OCl)
ratios in the two experiments. With the exception of reac-
tion 24 all of these reactions result in HCHO formation and
cannot account for our observations of low HCHO yields.
Should reaction 24 take place, the branching ratio of about 78%
to Cl2 formation that was derived from the ∆(Cl2)/∆(CH3OCl)
data in the LP-UV experiments in air is a lower limit to the
real value.
Although the measurements do constrain the possible branch-
ing ratio, the uncertainties in the data and in the interpretation
of the chemical mechanism do not allow us to give a precise
branching ratio for reaction 2, and we prefer to quote ap-
proximate values of k2a/k2 and k2b/k2 of 0.2 ( 0.1 and 0.8 (
0.2, respectively.
We note that a branching ratio of k2a/k2 ) 0.2 yields a site-
specific rate constant of about 1.2 × 10-11 cm3 molecule-1 s-1

Reaction between Cl and CH3OCl at 295 K
J. Phys. Chem., Vol. 100, No. 43, 1996 17199
a
TABLE 3: Results of the LP-UV Experiments in N2
expt
[CH3OCl]i/1016
[Cl2]i/1016
∆[CH3OCl]/1012
∆[Cl2]/1012
(d[HCHO])i/1012
k2b/k2
1
2
3
4
5
6
7
8
1.0
1.4
1.6
0.94
1.5
0.84
1.8
1.4
1.2
-10.6 ( 0.1
-10.3 ( 0.3
-5.90 ( 0.04
-6.57 ( 0.06
-5.43 ( 0.05
-9.98 ( 0.03
-7.00 ( 0.10
-6.43 ( 0.05
3.08 ( 0.04
2.63 ( 0.07
1.89 ( 0.08
1.82 ( 0.03
1.72 ( 0.04
1.71 ( 0.10
2.35 ( 0.05
1.88 ( 0.03
3.40 ( 0.40
3.39 ( 0.35
2.21 ( 0.35
1.89 ( 0.07
1.66 ( 0.10
3.65 ( 0.04
2.67 ( 0.30
2.50 ( 0.10
0.79
0.76
0.82
0.78
0.82
0.67
0.84
0.79
0.61
0.98
1.0
1.0
0.46
1.3
1.4
a All concentrations in units of molecules cm-3. Branching ratios (k2b/k2) derived from analysis of Cl2 growth.
CH3OCl + hv f CH3O + Cl
(5)
for the H atom abstraction. Although lower than the rate
constant for Cl + CH3OH (5.4 × 10-11 cm3 molecule-1 s-1 12
,
it is comparable to that for Cl + CH3ONO (9.5 × 10-12 cm3
molecule-1 s-1 25 and larger than that for CH3ONO2 (2.6 ×
10-13 cm3 molecule-1 s-1 26).
CH3OCl + OH f H2O + CH2OCl
CH3OCl + Cl f Cl2 + CH3O
CH3OCl + Cl f HCl + CH2OCl
CH2OCl f HCHO + Cl
(26)
(2b)
(2a)
(9)
In our previous study of the reaction between OH and
CH3OCl10 we concluded, on the basis of measured Arrhenius
parameters, that the reaction proceeded via H atom abstrac-
tion, but did not rule out the possibility of Cl atom abstrac-
tion. The present results show deactivation of the H-site and a
highly reactive Cl-site for the Cl reaction and may add sup-
port for a Cl abstraction and formation of HOCl in the OH +
CH3OCl reaction. Further evidence for a highly deactivated
H-site is provided by a recent study27 of the OH + CH3OCl
reaction, in which HOCl was measured by mass spectrometry
and identified as the major product (ca. 80% branching to HOCl
+ CH3O).
For the computation of the photolysis rates we use an updated
version of the scheme by Lary and Pyle31 using a typical
Antarctic ozone profile and a surface albedo of 0.8. Rate
constants and absorption cross sections are taken from DeMore
et al.12
The formation of nitric acid trihydrate (NAT) and ice particles
in the model depends on the prevailing temperatures and the
partial pressures of HNO3 and water vapor. The total surface
area of the NAT and ice particles was computed assuming a
monodisperse aerosol distribution as in eariler studies.3 The
observed denitrification and dehydration33 in the polar vortex
are taken into account; it is assumed that about one-quarter of
the initial HNO3 and 2 ppm H2O remain after the PSC (polar
stratospheric cloud) period.
4. Stratospheric Chemistry of CH3OCl
The interest in CH3OCl as a species that is expected to be
formed in the polar stratosphere lies in the fact that it contains
a chlorine atom and may therefore conceivably take part either
in catalytic O3 destroying cycles, as does another hypochlorite,
HOCl, or it may represent a relatively stable sink for chlorine,
which could lead to lower O3 depletion rates.
We simulate the typical chemical history of an air mass at
polar latitudes in the stratosphere over austral winter and spring,
similar to a recent study,29 but for higher altitudes. Here, the
air parcel is initialized on June 1 at 15 hPa (650 K), and it is
assumed that it descends to 80 hPa (425 K) until the end of
October in accordance with the descent indicated by observa-
Under the conditions in which the formation of CH3OCl is
maximized in the polar stratosphere (low O3, high ClO) Cl
concentrations are concomitantly high, and the high rate constant
for Cl + CH3OCl makes the reaction between Cl and CH3OCl
a potentially important removal process for CH3OCl that can
compete with photolysis. We assume, as for the reaction
between Cl atom and Cl2O, which also has a large room
tions of CH4 in the Antarctic vortex by the HALOE.34
A
linearly decreasing temperature during polar night and the
temperature increase after polar sunrise are estimated from
measurements made at the Neumayer station (71°S, 8°W) [H.
Gernandt, personal communication]. As well as approximate
temperature variations with longitude caused by dynamical
processes in the Antarctic vortex,35 a temperature variation with
an amplitude of 5 K and a period of 7 days is superimposed on
the adopted temperature trend. The modeled air parcel is
assumed to circulate within the polar vortex at an average
latitude of 75°S with sinusoidal variations in latitude, with an
amplitude of 10° and a period of 5 days to account ap-
proximately for deviations from purely zonal flow in the vortex
region.
Initial values of O3, CH4, HCl, H2O, NO, and NO2 are taken
from HALOE observations33 at 48°S and 85°E, on June 2, 1993,
when HALOE sounded early vortex conditions in austral winter
at midlatitudes during a Rossby-wave breaking event. Values
for total inorganic chlorine, Cly, are derived from the inter-
relationship of Cly with CH4;36 values for reactive nitrogen,
NOy, are derived from the interrelation of NOy with N2O.31
The initial values are O3 ) 5.0 ppmv, HNO3 ) 5.3 ppbv, H2O
) 4.5 ppmv, CH4 ) 1.0 ppmv, HCl ) 1.6 ppbv, ClONO2 )
1.2 ppbv, NO ) 2.0 ppbv, and NO2 ) 4.0 ppbv.
temperature rate constant (9 × 10-11 cm3 molecule-1 s-1 12
)
and forms Cl2, that the rate constant will display at most a weak
temperature dependence, and we use the room temperature result
from this study to examine the importance of the reaction
between Cl and CH3OCl in the stratosphere, where temperatures
may be as low as 185 K. J-values for CH3OCl are close to 5
× 10-5 s-1 at zenith angles of between 80° and 90° and 100
mbar, giving a lifetime of ca. 4 h with respect to photodisso-
ciation.9 Equal importance for the reaction between Cl and
CH3OCl is therefore achieved when Cl atom concentrations of
1 × 106 are reached. In the following modeling study we show
that removal by reaction with Cl atoms contributes to removal
of CH3OCl. This is in contrast to HOCl, which reacts relatively
slowly with Cl atoms.12
4.1. Chemical Box Model Study. We employ a chemical
box model3,28-30 including a comprehensive set of stratospheric
gas-phase and heterogeneous reactions on NAT and ice surfaces
as well as the hydrolysis of ClONO2 and N2O5 on liquid sulfate
aerosol. For the present study, we have included the formation
of CH3OCl via reaction 1 and the loss of CH3OCl via refs 9,
10, this work)

17200 J. Phys. Chem., Vol. 100, No. 43, 1996
Carl et al.
Figure 9. (A) Temporal development of the calculated CH3OCl mixing
ratios. (B) Percentage difference (smoothed with a 4 day filter) between
the CH3OCl mixing ratios of a model run neglecting the title reaction
and a run with the full reaction scheme.
the typical pattern of chlorine deactivation for Antarctic
conditions.29,36
To investigate the atmospheric significance of the title
reaction, we performed a model calculation in which it was
reglected as a loss process for CH3OCl. Under this assumption
up to 35% higher mixing ratios of CH3OCl are computed (Figure
9), showing that reaction 2 is indeed a significant loss channel
for CH3OCl.
We have shown that the reaction of CH3OCl with Cl atoms
releases photolabile Cl2 and that this reaction can take place in
the polar stratosphere. We can thus write the following catalytic
O3 destruction sequence in which HCl formed in the first step
of the CH4 oxidation (27) is activated by a heterogeneous
reaction with HOCl (30), which is formed in subsequent HOx
chemistry.3
Figure 8. Adopted temperature history (dotted line) and NAT surface
area (in µm2/cm3) (solid line) over the model period from June 1, 1993,
to October 31, 1993 (upper panel). The second panel shows the HCl
mixing ratio, the third panel the mixing ratios of ClONO2 (solid line)
and HOCl (dashed line), the fourth panel the mixing ratio of ClOx (ClOx
) Cl + ClO + 2Cl2O2) over the model period, and the lowest panel
the mixing ratio of ozone.
For the temperature development assumed here, NAT par-
ticles exist in the model for the period from mid June to the
end of September (Figure 8), in good agreement with a
climatology of PSC occurrence in the polar regions.34 With
the first occurrence of PSCs HCl is rapidly titrated against the
available ClONO2 and HOCl via reactions catalyzed by NAT
surfaces and is converted to active chlorine ClOx ()Cl + ClO
+ 2Cl2O2). The further decrease of HCl is much slower and is
controlled by the production rate of HOCl and ClONO2;27
therefore it takes more than 2 months before the HCl is
completely activated (Figure 8).
When HCl has been almost completely removed in the model
(mid August, Figure 8), it can no longer act as a sink for OH,
and thus, while HNO3 is still sequestered in PSCs, OH levels
increase dramatically. During this period, ozone depletion also
gains momentum, and, as a consequence, extremely high Cl
atom concentrations occur.29,37 For both reasons, increasing OH
and increasing Cl radical concentrations, the methane oxidation
chain is strongly enhanced in late September and, consequently,
the peak mixing ratios of CH3OCl of about 6 pptv are reached
(Figure 9).
2Cl2 + hν f 4Cl
(3)
(27)
(28)
(1a)
(2b)
(7)
Cl + CH4 + O2 f CH3O2 + HCl
2(Cl + O3) f 2(ClO + O2)
CH3O2 + ClO f CH3OCl + O2
Cl + CH3OCl f Cl2 + CH3O
CH3O + O2 f HCHO + HO2
HO2 + ClO f HOCl + O2
(29)
(30)
HOCl + HCl f Cl2 + H2O
2O3 + CH4 f HCHO + H2O + 2O2
The high Cl concentrations, however, result in formation of
HCl via reaction with CH4. Indeed, after the last PSC event
at the end of September, HCl increases rapidly from 0 to
2.7 ppbv over about a month. Such a rapid HCl buildup is
net
The net reaction is thus identical to that obtained when the
reaction between CH3O2 and ClO yields the products CH3O

Reaction between Cl and CH3OCl at 295 K
J. Phys. Chem., Vol. 100, No. 43, 1996 17201
(11) Benson, S. W. Thermochemical Kinetics; Wiley; New York, 1976.
(12) DeMore, W. B.; Sander, S. P.; Golden, D. M.; Hampson, R. F.;
Kurylo, M. J.; Howard, C. J.; Ravishankara, A. R.; Kolb, C. E.; Molina,
M. J. Chemical Kinetics and Photochemical Data for Use in Stratospheric
Modeling; JPL Publication 94-26; Pasadena, 1994.
(13) Finkbeiner, M.; Neeb, P.; Horie, O.; Moortgat, G. K. Fresenius J.
Anal. Chem. 1995, 351, 521.
(14) Moortgat, G. K.; Cox, R. A.; Schuster, G.; Burrows, J. P.; Tyndall,
G. S. J. Chem. Soc., Faraday Trans. 2 1989, 809.
(15) Raber, W.; Moortgat, G. K. In Progress and Problems in
Atmospheric Chemistry; Barker, J. R., Ed.; World Scientific Press: Sin-
gapore, 1996; Vol. 3, pp 318-373.
(16) Hsu, K. J.; DeMore, W. B. J. Phys. Chem. 1995, 99, 1235.
(17) Su, F.; Calvert, J. G.; Shaw, J. S. J. Phys. Chem. 1979, 25, 3185.
(18) Veyret, B.; Lesclaux, R.; Rayez, M.; Rayez, J.; Cox, R. A.;
Moortgat, G. K. J. Phys. Chem. 1989, 93, 2368. Burrows, J. P.; Moortgat,
G. K.; Tyndall, G. S.; Cox, R. A.; Jenkin, M. E.; Hayman, G. D.; Veyret,
B. J. Phys. Chem. 1989, 93, 2375.
(19) Su, F.; Calvert, J. G.; Shaw, J. H.; Niki, H.; Maker, P. D.; Savage,
C. M.; Breitenbach, L. D. Chem. Phys. Lett. 1979, 65, 221.
(20) Vaghjiani, G. L.; Ravishankara, A. R. J. Phys. Chem. 1989, 93,
1948.
and ClOO, as assumed in the original modeling study of Crutzen
et al.3 and confirmed in this laboratory.4 Indeed the net reaction
above is also identical to that obtained when CH3OCl is
photolyzed to CH3O + Cl.9
Therefore, as the loss processes for CH3OCl are relatively
fast and not rate limiting in any of the catalytic reaction schemes
(e.g. above) that can be written, its formation in reaction 1
should not significantly alter the role of the Cl atom initiated
methane oxidation in modifying the O3 depletion mechanism
proposed by Crutzen et al.3
This assumption could be confirmed in the box model study,
which returned practically unaltered O3 loss rates with and
without reaction 2. Indeed, further test calculations (not shown)
indicate that even completely neglecting the formation of
CH3OCl in reaction 1b (as in ref 3) does not affect the calculated
ozone loss, as already suggested by Crowley et al.9
5. Conclusions
(21) Curtis, A. R.; Sweetenham, W. P. Facsimile program. AERE
Report R-12805; HMSO: Londong, 1987.
(22) Maric, D.; Burrows, J. P.; Meller, R.; Moortgat, G. K. J. Photochem.
Photobiol. A 1993, 70, 205.
(23) Meller, R. Diploma Thesis, University of Mainz, Germany, 1991
(data measured in this laboratory).
(24) Libuda, H. G.; Zabel, F.; Fink, E. H.; Becker, K. H. J. Phys. Chem.
1990, 94, 5860.
(25) Nielsen, O. J.; Sidebottom, H. W.; Donlon, M.; Treacy, J. Chem.
Phys. Lett. 1991, 178, 163.
(26) Nielsen, O. J.; Sidebottom, H. W.; Donlon, M.; Treacy, J. Int. J.
Chem. Kinet. 1991, 23, 1095.
(27) Kukui, A.; Benter, Th.; Schindler, R. N. Data presented at the 2nd
Euroconference, “6th sensing the atmosphere”, Marbella, Spain, 1996.
(28) Mu¨ller, R.; Crutzen, P. J. J. Geophys. Res. 1993, 98, 20483.
(29) Mu¨ller, R.; Peter, Th.; Crutzen, P. J.; Oelhaf, H.; Adrian, G. P.;
Clarmann, Th. v.; Wegner, A.; Schmidt, U.; Lary, D. Geophys. Res. Lett.
1994, 21, 1427.
The reaction between Cl and CH3OCl proceeds with a room
temperature rate constant of (6.1 ( 0.2) × 10-11 cm3 molecule-1
s-1. At the same temperature, chlorine abstraction dominates
product formation (ca. 80%) with only a ca. 20% contribution
from H atom abstraction. The fate of CH3OCl in the polar
stratosphere will be determined by competition between pho-
tolysis and reaction with Cl atoms. As most reactive encounters
between Cl and CH3OCl result in Cl2 formation, both loss
processes give essentially the same net conversion of O3.
Acknowledgment. This work was funded in part by the CEC
Project EV5V-CT93-0338, CABRIS, within the CEC Environ-
ment Programme.
(30) Mu¨ller, R.; Brenninkmeijer, C. A. M.; Crutzen, P. J. A large 13CO
deficit in the lower Antarctic stratosphere due to “ozone hole” chemistry:
Part II Modelling (1996). Geophys. Res. Lett. 1996, 23, 2129.
(31) Lary, D. J.; Pyle, J. A. J. Atmos. Chem. 1991, 13, 373.
(32) Fahey, D. W.; Kelly, K. K.; Kawa, S. R.; Tuck, A. F.; Loewenstein,
M.; Chun, K. R.; Heidt, L. E. Nature 1990, 344, 321.
(33) Kelly, K. K.; Tuck, A. F.; Heidt, L. E.; Loewenstein, M.; Podolske,
J. R.; Strahan, S. E.; Vedder, J. F. Geophys. Res. Lett. 1990, 17, 465.
(34) Russell, J. M., III; Gordley, L. L.; Park, J. H.; Drayson, S. R.; Tuck,
A. F.; Harries, J. E.; Cicerone, R. J.; Crutzen, P. J.; Frederick, J. E. J.
Geophys. Res. 1993, 98, 10777.
(35) Poole, L. R.; Pitts, M. C. J. Geophys. Res. 1994, 99, 13083.
(36) Woodbridge, E. L.; Elkins, J. W.; Fahey, D. W.; Heidt, L. E.;
Solomon, S.; Baring, T. J.; Gilpin, T. M.; Pollock, W. H.; Schauffler, S.
M.; Atlas, E. L.; Loewenstein, M.; Podolske, J. R.; Webster, C. R.; May,
R. D.; Gilligan, J. M.; Montzka, S. A.; Boering, K. A.; Salawich, R. J. J.
Geophys. Res. 1995, 100, 3057.
References and Notes
(1) Anderson, J. G.; Toohey, D. W.; Brune, W. H. Science 1991, 251,
39.
(2) Solomon, S. Nature 1990, 347, 347.
(3) Crutzen, P. J.; Mu¨ller, R.; Bru¨hl, C.; Peter, T. Geophys. Res. Lett.
1992, 19, 1113.
(4) Helleis, F.; Crowley, J. N.; Moortgat, G. K. J. Phys. Chem. 1993,
97, 11464.
(5) Helleis, F.; Crowley, J. N.; Moortgat, G. K. Geophys. Res. Lett.
1994, 21, 1795.
(6) Kenner, R. D.; Ryan, K. R.; Plumb, I. C. Geophys. Res. Lett. 1993,
20, 1571.
(7) Kukui, A.; Jungkamp, T.; Schindler, R. N. Ber. Bunsen-Ges. Phys.
Chem. 1994, 98, 1298.
(8) Biggs, P.; Canosa-Mas, C. E.; Fracheboud, J. M.; Shallcross, D.
E.; Wayne, R. P. Geophys. Res. Lett. 1995, 22, 1221.
(9) Crowley, J. N.; Helleis, F.; Mu¨ller, R.; Moortgat, G. K.; Crutzen,
P. J.; Orlando, J. J. J. Geophys. Res. 1994, 99, 20683.
(10) Crowley, J. N.; Campuzano-Jost, P.; Moortgat, G. K. J. Phys. Chem.
1996, 100, 3601.
(37) Douglass, A. R.; Schoeberl, M. R.; Stolarski, R. S.; Waters, J. W.;
Russell, J. M., III; Roche, A. E.; Massie, S. T. J. Geophys. Res. 1995, 100,
13967.
JP9611075