Kinetics and mechanism of the reaction of F atoms with CH3Br

Article abstract of DOI:10.1021/jp953397f

The reaction of F atoms with CH₃Br at 296 K was studied using a pulse radiolysis/transient UV absorption spectroscopy absolute technique and a FTIR relative rate technique. The rate constant for this reaction was determined to be (4.46 ± 0.22) × 10^-11. The reaction proceeds via two channels, 69 ± 5%, via hydrogen abstraction giving CH₂Br radicals and HF, and 31 ± 5%, to give the adduct CH₃Br-F. In the FTIR system the observed rate constant was 69 ± 8% of that measured using the pulse radiolysis system because the CH₃Br-F adduct falls apart to re-form the reactants. The CH₃Br-F adduct reacts with NO, with a rate constant of (2.25 ± 0.14) × 10^-11 cm³ molecule^-1 s^-1, giving FNO as a product. There was no discernible reaction of the CH₃Br-F adduct with O₂ and an upper limit of 6 × 10^-15 cm³ molecule^-1 s^-1 was derived for this reaction. The CH₃Br-F adduct absorbs strongly at 260-340 nm. The absorption cross section at 280 nm of the CH₃Br-F adduct was (2.06 ± 0.31) × 10^-17 cm² molecule^-1. A lower limit for the equilibrium constant was [CH₃Br-F]/([CH₃Br][F]) > 5 × 10^-16 cm³ molecule^-1 at 296 K. A lower limit of 12 kcal mol^-1 is estimated for the binding energy of the F atom in the CH₃Br-F adduct. The UV absorption spectrum of the CH₂BrO₂ radical was determined; at 250 nm σ = (3.4 ± 0.9) × 10^-18 cm² molecule^-1.

Full text of DOI:10.1021/jp953397f

J. Phys. Chem. 1996, 100, 10989-10998
Kinetics and Mechanism of the Reaction of F Atoms with CH3Br
Jens Sehested,* Merete Bilde, and Trine Møgelberg
10989
Section for Chemical ReactiVity, EnVironmental Science and Technology Department,
Risø National Laboratory, DK-4000 Roskilde, Denmark
Timothy J. Wallington*
Research Staff, SRL-E3083, Ford Motor Company, P.O. Box 2053, Dearborn, Michigan 48121-2053
Ole John Nielsen*
Ford Forschungszentrum Aarchen, Ford Motor Company, Dennewartsstrasse 25, D-52068 Aachen, Germany
X
ReceiVed: NoVember 20, 1995; In Final Form: April 18, 1996
The reaction of F atoms with CH
3
Br at 296 K was studied using a pulse radiolysis/transient UV absorption
spectroscopy absolute technique and a FTIR relative rate technique. The rate constant for this reaction was
-11
determined to be (4.46 ( 0.22) × 10 . The reaction proceeds via two channels, 69 ( 5%, via hydrogen
abstraction giving CH Br radicals and HF, and 31 ( 5%, to give the adduct CH Br--F. In the FTIR system
the observed rate constant was 69 ( 8% of that measured using the pulse radiolysis system because the
CH Br--F adduct falls apart to re-form the reactants. The CH Br--F adduct reacts with NO, with a rate
constant of (2.25 ( 0.14) × 10 cm molecule s , giving FNO as a product. There was no discernible
2
3
3
3
-
11
3
-1 -1
-
15
3
-1 -1
reaction of the CH
this reaction. The CH
nm of the CH
Br--F adduct was (2.06 ( 0.31) × 10 cm molecule . A lower limit for the equilibrium
constant was [CH
3
Br--F adduct with O
2
and an upper limit of 6 × 10 cm molecule
s
was derived for
3
Br--F adduct absorbs strongly at 260-340 nm. The absorption cross section at 280
-17
2
-1
3
-16
3
-1
3
Br--F]/([CH
3
Br][F]) > 5 × 10
cm molecule at 296 K. A lower limit of 12 kcal
Br--F adduct. The UV absorption spectrum
-1
mol is estimated for the binding energy of the F atom in the CH
3
-18
2
-1
of the CH
2
BrO
2
radical was determined; at 250 nm σ ) (3.4 ( 0.9) × 10 cm molecule .
peroxy radicals (R′O2):8
-10
1
. Introduction
Recognition of the adverse impact of bromine compounds
CH BrO + NO f CH BrO + NO
2
(3a)
(3b)
2
2
2
on stratospheric ozone has led to interest in the atmospheric
chemistry of CH3Br.1 It is believed that there are two major
sources of atmospheric CH3Br: an anthropogenic source from
CH BrO + NO + M f CH BrONO + M
2
2
2
2
2
its use as a fumigant and a natural source from the oceans.
The overall source strength is estimated to be 100-150 kT/
year. It is estimated that up 30% of the CH3Br release could
CH BrO + NO + M h CH BrO NO + M (4, -4)
2 2 2 2 2 2
2
be of anthropogenic origin. The observed atmospheric con-
centration is 10-15 ppt.2
CH BrO + HO f products
(5)
(6)
The principal removal mechanism for CH3Br in the atmo-
sphere is reaction with OH:
2
2
2
CH BrO + R′O f products
2
2
2
CH Br + OH f CH Br + H O
(1)
3
2
2
There have been two previous studies of the UV spectra and
kinetics of CH2Br and CH2BrO2 radicals. The first study was
performed in our laboratory and employed the reaction of F
atoms with CH3Br as a source of CH2Br and hence CH2BrO2
The lifetime of CH3Br in the atmosphere with respect to reaction
is approximately 1.8 years.2
-5
Reaction 1 produces CH2Br
1
radicals which then add oxygen rapidly (within 1 µs) to give
6
,7
peroxy radicals:
CH Br + O + M f CH BrO + M
6
radicals. The second study was performed by Villenave and
7
Lesclaux and used the reaction of Cl atoms with CH3Br to
(2)
generate CH2Br radicals. The spectroscopic and kinetic results
from these two studies are in substantial disagreement, sug-
gesting that different radical species were present in the two
studies. The goal of the present work was to understand this
discrepancy.
2
2
2
2
As with other peroxy radicals in the atmosphere, CH2BrO2
radicals are expected to react with NO, NO2, HO2, and other
Recent results from our laboratory show that an adduct is
formed in the reaction of F atoms with CF2BrH and that this
*
Authors to whom correspondence should be addressed.
Abstract published in AdVance ACS Abstracts, June 1, 1996.
X
11
S0022-3654(95)03397-1 CCC: $12.00 © 1996 American Chemical Society

1
0990 J. Phys. Chem., Vol. 100, No. 26, 1996
Sehested et al.
adduct is in equilibrium with the reactants, F atoms and CF2-
BrH:
of UV absorption spectra, was calculated from the observed
absorbance at 260 nm due to CH3O2 radicals formed upon
radiolysis of mixtures of CH4, O2, and SF6 using σ260nm(CH3O2)
F + CF BrH h CF BrH--F
(7, -7)
-18
2
-1 9
2
2
) (3.18 ( 0.32) × 10 cm molecule . The fluorine atom
15 -3
yield was (3.0 ( 0.3) × 10 molecules cm (full dose, 1000
Less than 1% of the reaction of F atoms with CF2BrH at ambient
temperature proceeds via direct hydrogen abstraction to give
CF2Br radicals and HF.11 The formation of a short-lived
molecular complex in the reactions of Cl atoms with CH3I and
mbar SF6). The quoted uncertainty reflects propagation of a
1
0% uncertainty in σ(CH3O2) and the statistical uncertainty (2
standard deviations) of the absorbance at 260 nm.
Reagents used were as follows: O2 (ultrahigh purity), 0-110
mbar; SF6 (99.9%), 885-1000 mbar; CH3Br (>99%), 0-20
mbar; NO (>99.8%), 0-2.5 mbar; and CH4 (>99%), 0-10
mbar. O2 was supplied by L’Air Liquide. SF6, CH3Br, and
CH4 were obtained from Gerling and Holz. NO was provided
by Messer Griesheim. All reagents were used as received.
1
2
CH3Br has also been reported recently by Wine et al.
In the present work we present kinetic and spectroscopic data
which show that the reaction of F atoms with CH3Br proceeds
via two reaction channels: adduct formation and direct hydrogen
abstraction. Formation of the CH3Br--F adduct via channel 8a
explains the discrepancy between the previous studies of the
UV spectra and kinetics of CH2Br and CH2BrO2 radicals.
Spectral and kinetic information for the CH3Br--F adduct and
CH2BrO2 radical are presented herein.
2
.2. FTIR-Smog Chamber Setup. The FTIR system was
interfaced to a 140 l Pyrex reactor. Radicals were generated
by the UV irradiation of mixtures of CH3Br, CH4, CD4, and F2
in N2 or air diluent, at 296 K using 22 black lamps (760 Torr
)
1013 mbar). The loss of reactants was monitored by FTIR
F + CH Br h CH Br--F
(8a, -8a)
3
3
spectroscopy, using an analyzing path length of 26 m and an
-1
optical resolution of 0.25 cm . Infrared spectra were acquired
by expanding known volumes of reference materials into the
reactor. All reagents were supplied by Michigan Air Gas at
F + CH Br f CH Br + HF
(8b)
3
2
2
. Experimental Section
>
99% purity. Ultrahigh purity N2 or air diluents were used.
All uncertainties reported in this paper are 2 standard
The two experimental systems used in the present work have
13-15
been described in detail previously
discussed briefly here.
and will only be
deviations unless otherwise stated. Standard error propagation
methods are used to calculate combined uncertainties.
2
.1. Pulse Radiolysis Setup. A pulse radiolysis transient
UV absorption apparatus was used to study the UV absorption
spectra and kinetics of the CH2Br radical and the CH3Br--F
adduct. Radicals were generated by radiolysis of gas mixtures
at 296 K in a 1 L stainless steel reactor by a 30 ns pulse of 2
MeV electrons from a Febetron 705B field emission accelerator.
The radiolysis dose, expressed throughout this article as a
fraction of maximum dose, was varied by insertion of stainless
steel attenuators between the accelerator and the chemical
reactor. The analyzing light was obtained from a pulsed Xenon
arc lamp and reflected in the reaction cell by internal White
type optics. The optical path lengths used in this work were
3. Results
3
.1. Kinetics of the Reaction of F Atoms with CH3Br.
Radicals were generated by radiolysis of mixtures of 0.5-2.16
mbar of CH3Br and 1000 mbar of SF6:
-
SF + (2 MeV)e f F + products
(9)
(8)
6
F + CH Br f products
3
Following the radiolysis pulse a rapid increase in absorption at
80 nm was observed, followed by a slower decay. An
2
8
0 and 120 cm. The analyzing light was monitored by a 1 m
experimental transient recorded at 280 nm using 11% of the
maximum radiolysis dose, 80 cm UV path length, and a mixture
of 0.79 mbar of CH3Br and 1000 mbar of SF6 is shown in Figure
McPherson monochromator linked to a Hamamatsu R928
photomultiplyer and a LeCroy 9450A oscilloscope. The mono-
chromator was operated at a spectral resolution of 0.8 nm when
the photomultiplier was applied as the detection method. All
transients were results of single-pulse experiments with no signal
averaging.
A Princeton Applied Research OMA-II diode array was used
to measure UV absorption spectra of transient species. The
diode array was installed at the exit slit of the monochromator
in place of the photomultiplier, which was used for measuring
transient absorptions. The setup consisted of the diode array,
an image amplifier (type 1420-1024HQ), a controller (type
1
1
A. The absorption transients such as that shown in Figure
A were fitted using a first order rise expression:
1st
A(t) ) (A - A )(1 - exp(-k t)) + A
(I)
∞
0
0
where A(t), A0, and A∞ are the absorbance as function of time
1
st
t, at time zero, and at infinite time, respectively. k is the
pseudo-first order formation rate equal to k8[CH3Br]. The values
1
st
of k obtained are plotted as a function of the initial CH3Br
pressure in Figure 2. A linear least squares regression gives k8
1
421), and a conventional personal computer used for data
-11
3
-1 -1
acquisition, handling, and storage. Spectral calibration was
achieved using a Hg pen ray lamp. The spectral resolution of
the diode array spectra of the CH3Br--F adduct and the CH2-
BrO2 radical was 3.2 nm. The spectral resolution applied for
the FNO spectra was 0.8 nm.
) (4.46 ( 0.22) × 10
cm molecule s .
There is a small positive intercept in Figure 2 of (9.9 ( 6.6)
4
-1
× 10 s . This intercept can be explained by the loss of the
absorbing radical species. A dose of 0.11 was used to obtain
the data plotted in Figure 2. The decay lifetime of the
absorbance at 280 nm under these conditions was approximately
20 µs and was independent of [CH3Br]. This gives a pseudo-
SF6 was used as the diluent gas, the radiolysis of SF6 is known
to produce fluorine atoms:13
4
-1
first order decay rate of 5.0 × 10 s . Using this decay rate,
-
-11
3
-1 -1
16
SF + (2 MeV)e f F + products
(9)
k8 ) 4.46 × 10 cm molecule s , and the CHEMSIMUL
6
numerical integration program, simulated transients were cal-
culated for [CH3Br] ) 0.5 and 2.16 mbar. The calculated
transients were fitted using the first order first expression (I)
SF6 was always present in great excess to minimize the relative
importance of direct radiolysis of other compounds in the gas
mixtures. The fluorine atom yield, required for quantification
1
st
given above, and the values of k obtained are plotted in Figure

Reaction of F Atoms with CH3Br
J. Phys. Chem., Vol. 100, No. 26, 1996 10991
Figure 3. Loss of CH
3 4 2 3
Br versus CH (triangles), CF HCH (152a)
(
circles), and CD (diamonds) when mixtures containing these com-
4
pounds were exposed to F atoms in 700 Torr of air diluent.
11
)
(4.46 ( 0.22) × 10- obtained here is in good agreement
6
with the value obtained previously by Nielsen et al. of (4.5 (
-11
3
-1 -1
0
.4) × 10
The kinetics of the reaction of F atoms with CH3Br were
also studied by a FTIR relative rate technique.
.2. FTIR Relative Rate Study of the Reaction of F Atoms
cm molecule s .
3
with CH3Br. To investigate the kinetics of the reaction of F
atoms with CH3Br, relative rate studies were performed using
the FTIR system. The techniques used are described in detail
1
7,18
elsewhere.
Photolysis of molecular fluorine was used as a
source of F atoms. The kinetics of reaction 8 were measured
relative to reactions 11-13. Figure 3 shows the observed losses
Figure 1. Transient absorption at 280 nm following radiolysis of (A)
a mixture of 0.79 mbar of CH
and (B) a mixture of 2.5 mbar of CH
.32. The UV path length was 80 cm. The smooth line in A is a first
3
Br and 1000 mbar of SF
6
, dose 0.11,
3
Br and 1000 mbar of SF
6
, dose
F2 + hν f 2F
(10)
(8)
0
order fit to the experimental data. See text for details.
F + CH Br f products
3
F + CHF CH (HFC-152a) f CHF CH + HF (11)
2
3
2
2
F + CH f CH + HF
(12)
(13)
4
3
F + CD f CD + DF
4
3
of CH3Br versus those of the reference compounds following
irradiation of mixtures of CH3Br/CHF2CH3/F2, CH3Br/CH4/F2,
and CH3Br/CD4/F2 in 700 Torr of air diluent. Linear least
squares analyses gives k8/k11 ) 1.87 ( 0.14, k8/k12 ) 0.51 (
0
.03, and k8/k13 ) 0.79 ( 0.04. The reactivity of F atoms
toward CHF2CH3 and CD4 relative to that of CH4 are k11/k12 )
18
13
19
0
.25 ( 0.02 and k /k12 ) 0.69 ( 0.02. These rate constant
ratios can be combined with the determinations of k8/k11 ) 1.87
0.14 and k8/k13 ) 0.79 ( 0.04 in the present work to provide
(
values of k8/k12 ) 0.47 ( 0.05 and 0.55 ( 0.04. Errors were
propagated using conventional error analysis. It is gratifying
to note the internal consistency of the relative rate data. We
choose to quote the direct measurement of k8/k12 ) 0.51 ( 0.03
Figure 2. First order formation rate constants (filled circles) following
pulse radiolysis of mixtures of 0.5-2.16 mbar of CH
of SF plotted as a function of [CH Br]. Measurements were made at
80 nm. The slope of the straight line through the data is (4.66 (
3
Br and 1000 mbar
6
3
-11 18
as the final value. Using k12 ) 6.8 × 10
to place this
2
0
-
11
3
-1 -1
measurement on an absolute basis gives k8 ) (3.47 ( 0.20) ×
-11 3 -1 -1
.22) × 10 cm molecule
s . The hollow circles are the results
of computer simulations. See text for details.
10 cm molecule s . Potential systematic errors associated
with uncertainty in the reference rate constant could add an
additional 20% to the uncertainty range. Propagating this
2
as open circles. As seen from the figure, the decay of the
-11
3
transients explain the small intercept in Figure 2 and therefore
no correction is needed for this complication. The value of k8
additional uncertainty gives k8 ) (3.47 ( 0.72) × 10
cm
molecule- s .
1
-1

1
0992 J. Phys. Chem., Vol. 100, No. 26, 1996
Sehested et al.
Figure 4. Maximum transient absorbance at 280 nm following
radiolysis of 20 mbar (filled circles) or 2.5 mbar (hollow circles) of
Figure 5. Composite spectrum of the products following the reaction
of F atoms with CH
Villenave and Lesclaux (hollow circles) and Nielsen et al. (filled
3 2 2
Br (adduct and CH Br). Spectra of CH Br of
3 6
CH Br and 990 mbar of SF . The slope of the straight line through
7
6
the low-dose data (dose < 0.53) is 0.652 ( 0.013.
circles) are shown for comparison. The straight line in the [CH
Br--F] spectrum is an extrapolation between two spectra.
3
-
3
.3. Investigation of the Radical Yield as Function of
Radiolysis Dose. Determination of the absolute UV absorption
cross sections of the radical species formed in the reaction of F
atoms with CH3Br requires accurate quantification of the radical
concentrations. If we assume that only one absorbing species
is formed, we need to know the initial F atom yield and the
fraction of the F atoms that is converted into the absorbing
species. It is important to choose experimental conditions which
avoid unwanted radical-radical reactions such as F + CH2Br
and F + F, which could otherwise interfere with stoichiometric
conversion of F atoms into the desired radical species. To check
for unwanted chemistry, the maximum transient absorption was
measured as a function of the radiolysis dose. In Figure 4 the
maximum transient absorbance at 280 nm is plotted as a function
of the dose. As seen from the figure, the absorbance is
proportional to the radiolysis dose up to a relative dose of 0.53
for [CH3Br]0 ) 20 mbar and 0.42 for [CH3Br]0 ) 2.5 mbar.
This linearity indicates that for low radiolysis doses radical-
radical reactions and other loss mechanisms for the absorbing
species do not interfere significantly with its formation.
was 1 µs, the integration time was 1 µs, the relative dose was
.53, and the UV path length was 80 cm. To place the spectrum
0
on an absolute basis, the absorbance at 280 nm was scaled to
the observed absorption cross section determined above. The
spectrum is plotted in Figure 5. The spectra of CH2Br radicals
6
7
reported by Nielsen et al. and Villenave and Lesclaux are also
plotted in Figure 5. As seen from the figure, at wavelengths
greater than 250 nm the diode array spectrum is in reasonable
6
agreement with that attributed by Nielsen et al. to CH2Br
radicals. The reason for this apparent discrepancy at wave-
lengths below 250 nm is probably scattered light of higher
wavelengths. Measurements at wavelengths of 240 nm and
3
below are difficult in this system because CH Br absorbs at
-
19
2
-1 20
these wavelengths, σ(230nm) ) 1.5 × 10 cm molecule .
6
The Hilger and Watts monochromator used by Nielsen et al.
had a lower discrimination toward scattered light than the
presently used McPherson monochromator. We therefore
believe that the spectrum obtained in this work upon the reaction
3
of F atoms with CH Br is plausibly the same as that reported
6
by Nielsen et al.
Two initial pressures of CH3Br were used, 2.5 and 20 mbar.
The slope of the straight line through the low-dose data (dose
Clearly, the spectra obtained from the reaction of F atoms
with CH3Br and the spectrum of CH2Br radicals reported by
0.53 and lower for [CH3Br]0 ) 20 mbar and dose 0.42 and lower
7
for [CH3Br]0 ) 2.5 mbar) at 280 nm is 0.652 ( 0.013. In the
following experiments a dose of 0.53 was used to avoid
unwanted radical-radical reactions.
Villenave and Lesclaux are very different. Villenave and
7
Lesclaux used the reaction of Cl atoms with CH3Br to generate
CH2Br radicals:
An observed value of the absorption cross section at 280 nm
can be calculated using (i) the slope of the straight line through
the data points in Figure 4, 0.652 ( 0.013, (ii) the UV path
length of 80 cm, and (iii) the initial F atom yield of (3.0 ( 0.3)
Cl + CH Br f CH Br + HCl
(14)
3
2
The inconsistency between the spectra using Cl atoms and F
atoms to produce CH2Br radicals can be explained in light of
recent results from our laboratory which show that F atoms and
CF2BrH form an adduct which is in equilibrium with the
1
5
-3
×
10 molecules cm . The absorption cross section at 280
-18
2
-1
nm is (6.26 ( 0.64) × 10 cm molecule . The quoted error
includes both the statistical uncertainty from the plot in Figure
1
1
4
and uncertainty in the initial F atom yield. It is important to
reactants:
stress here that the absorption cross section derived above is
only an observed absorption cross section for a mix of radicals
and not for a specific species.
F + CF BrH h CF BrH--F
(7, -7)
2
2
3
.4. Composite Spectrum of the Radicals Formed from
The equilibrium constant was determined to be K ) (1.59 (
eq
-17
3
-1
the F + CH3Br Reaction. The UV spectrum of the species
formed in the reaction of F atoms with CH3Br was recorded
using the diode array following radiolysis of mixtures of 2.5
mbar of CH3Br and 1000 mbar of SF6. In these experiments
the time delay between the radiolysis pulse and data acquisition
0.12) × 10 cm molecule . The spectrum of this adduct is
shown in Figure 10. As seen from Figures 5 and 10 the
absorption spectrum of the CF BrH--F adduct in the region of
2
260-340 nm is similar to that of the radical products of the
reaction of F atoms with CH3Br. This suggests that a substantial

Reaction of F Atoms with CH3Br
J. Phys. Chem., Vol. 100, No. 26, 1996 10993
were simultaneously varied, and the best fit was achieved with
-
11
3
3
k8/k12 ) 0.74 ( 0.08. Using k12 ) (6.8 ( 1.4) × 10
cm
cm
-1
-1 21
-1
-11
molecule s , we arrive at k8 ) (5.0 ( 1.1) × 10
molecule- s , in good agreement with the value of (4.46 (
1
-
11
3
-1 -1
0
.22) × 10
cm molecule
s
determined using the pulse
radiolysis technique described in section 3.1.
The measurement of k8 relative to k12 is interesting for three
reasons: First, it is a complementary determination of k8 in our
system and agrees with the other pulse radiolysis determination.
Second, the fact that we get the same value of k8 strongly
indicates that the absorbance at 280 nm is attributable to some
product of the reaction of F atoms with CH3Br. If another
radiolysis product were responsible for the absorbance at 280
nm, it is extremely unlikely that this species has the same
relative reaction rate with CH4 and CH3Br as F atoms. Third,
it is interesting to note that the value of k8/k12 derived here (0.74
(
0.08) is significantly different from k8/k12 ) 0.51 ( 0.03,
derived by the FTIR relative rate method described in section
3.2. The apparent discrepancy in values of k8 determined by
the pulse radiolysis and FTIR techniques can be rationalized if
we consider two factors: first, the different time scales employed
by the two techniques; second, the mechanism for the reaction
Figure 6. Maximum transient absorption at 280 nm following pulse
radiolysis of mixtures of 0-100 mbar of CH
000 mbar of SF . The solid line is a fit to the data. See text for
details.
4 3
, 5 mbar of CH Br, and
1
6
fraction of reaction 8 proceeds via a mechanism which involves
formation of an adduct:
of F atoms with CH Br which proceeds at least in part via the
formation of an adduct which can decompose back to reactants.
3
F + CH Br h CH Br--F
(8a, -8a)
F + CH Br h [CH Br--F] f products (8a, -8a, 15)
3
3
3
3
F + CH Br f CH Br + HF
(8b)
F + CH Br f CH Br + HF
(8b)
3
2
3
2
Further evidence suggesting that an adduct is formed in the
reaction of F with CH3Br is presented in the following
subsections.
Reaction 15 represents unimolecular processes which remove
the adduct without reformation of F atoms or CH Br. Reactions
3
8a and 8b remove F atoms on the time scale of the pulse
3
.5. Determination of k(F+CH3Br) Relative to k(F+CH4).
radiolysis experiments (0-100 µs). If reaction -8a is much
faster than reaction 15, then only reaction channel 8b will
remove F atoms on the time scale of the FTIR experiments (5-
10 min). In the case of the CF2BrH--F adduct, less than 10%
of the adduct decomposition lead to products other than F and
CF2BrH. Assuming reaction 15 is negligible, the fraction of
F atoms that react via reaction channel 8b is (0.51 ( 0.03)/
(0.74 ( 0.08) ) 0.69 ( 0.08. This suggests that 31% of the F
atoms react via channel 8a and 69% via channel 8b. The
determination of k8a/k8 ) 31 ( 8% should be regarded as a
lower limit as reaction 15 may compete with reaction -8a. At
this point it is worth noting that decomposition of the CF2BrH-
The rate of the reaction of F atoms with CH3Br was measured
relative to that of F atoms with CH4. When mixtures of 5 mbar
of CH3Br, 0-100 mbar of CH4, and 1000 mbar of SF6 are
subject to pulse radiolysis, there is a competition for the F atoms
between CH3Br and CH4:
1
1
F + CH Br f products
(8)
3
F + CH f CH + HF
(12)
4
3
Figure 6 shows the observed variation of the maximum transient
absorbance at 280 nm as a function of the concentration ratio
CH4]/[CH3Br]. As seen from the figure, the maximum transient
-
F adduct proceeds essentially entirely (>90%) to regenerate
1
1
CF2BrH and F atoms. In light of the behavior of the CF2-
BrH--F adduct it seems reasonable to assume that the determi-
nation of k8a/k8 ) 31 ( 8% is not a gross underestimate. In
the following section we investigate the reaction of the complex
with NO to confirm the value for the branching determined here.
[
absorption at 280 nm decreases with increasing CH4 concentra-
tion. The decrease in absorbance at 280 nm can be rationalized
in terms of the competition between reactions 8 and 12 for the
F atoms. At high CH4 pressures, CH3 radicals are formed at
the expense of F + CH3Br reaction products. CH3 radicals do
not absorb at 280 nm. Therefore the maximum transient
absorbance decreases with increasing amounts of CH4.
The competition between reactions 8 and 12 can be used to
determine k8/k12 as follows. The data in Figure 6 can be fitted
with a three parameter expression:
3.6. Reaction of the CH3Br--F Adduct with NO. Mixtures
of 0-2.5 mbar of NO, 5-20 mbar of CH3Br, and SF6 added to
1000 mbar were radiolyzed, and the absorbance at 230 and 280
nm were recorded. The decay rate of the transient absorbance
at these wavelengths increased with increasing NO concentra-
tion. The decays were fitted with a first order decay expression,
and first order decay rates were derived. In Figure 7, first order
decay rates are plotted as a function of [NO]. The slopes of
the straight lines through the data give rate constants of (2.25
Amax ) {A
+ (AF+CH4[CH ]/((k /k )[CH Br]))}/(1 +
4 8 12 3
F+CH Br
3
-
11
[
CH ]/((k /k )[CH Br]))
( 0.08) × 10
(filled circles, 280 nm) and (2.16 ( 0.41) ×
4
8
12
3
-
11
3
-1 -1
1
0
cm molecule
s
(hollow triangles, 230 nm). The
intercepts are due to the self-reactions and cross-reactions of
the radicals.
CH2Br radicals do not absorb significantly at 280 nm.7
Therefore the decay rate determined from the absorbance decay
where Amax, AF+CH , and AF+CH Br are the observed maximum
4
3
absorbance as a function of the ratio [CH4]/[CH3Br], the
absorbance if all F atoms react with CH4, and the absorbance
if all F atoms react with CH3Br. AF+CH , AF+CH Br, and k8/k12
4
3

1
0994 J. Phys. Chem., Vol. 100, No. 26, 1996
Sehested et al.
Figure 8. Diode array spectra 100-100.2 µs after radiolysis of
Figure 7. First order decay rates at 280 nm (filled circles) and 230
nm (hollow triangles) as a function of [NO] following radiolysis of
mixtures of (A) 10 mbar and 990 mbar of SF
6
and (B) 2.5 mbar of
NO, 20 mbar of CH Br, and 980 mbar of SF . C-E are residual spectra
3 6
mixtures of 0-2.5 mbar of NO, 5-20 mbar of CH
3
Br, and 990 mbar
obtained by subtracting 22, 27, and 32% of spectrum A from spectrum
B, respectively. For clarity 0.01, 0.02, and 0.03 was subtracted from
C-E, respectively. Full-dose and 120 cm UV path length were used.
of SF
6
.
at 280 nm is the rate constant for reaction 16. Hence, k16 )
CH Br--F + NO f products (16)
3
11
3
-1
(
2.25 ( 0.08) × 10- cm molecule s-1. We choose to add
an extra 5% uncertainty due to systematic uncertainties and
-
11
3
-1 -1
hence report k16 ) (2.25 ( 0.14) × 10
cm molecule s .
It is shown below that the CH3Br--F adduct absorbs less than
the CH2Br radical at 230 nm. Therefore, we conclude that the
rate constant derived from the decay at 230 nm is close to rate
constant for reaction 17. Hence, k17 approximately equals (2.16
CH Br + NO f products
(17)
2
11
3
-1
(
0.41) × 10- cm molecule s-1. We conclude that both
the CH3Br--F adduct and the CH2Br radical react with NO with
rate constants of approximately 2.2 × 10
-
11
3
-1
cm molecule
-
1
s .
What are the products of reactions 16 and 17? By analogy
with the reaction of other alkyl radicals with NO we expect
that the product of reaction 17 is CH2BrNO. The likely products
of reaction 16, if we regard CH3Br--F as a molecular complex
with the F atom loosely bound, are FNO and CH3Br.
Figure 9. Maximum transient absorbance at 310.5 nm following
radiolysis of mixtures of 10 mbar of NO and 990 mbar of SF
6
(filled
squares) and 2.5 mbar of NO, 20 mbar of CH Br, and 977.5 mbar of
3
SF (circles). The straight lines are linear regressions of the filled data
6
points.
To investigate the formation of FNO from reaction 16, two
chemical systems were subject to pulse radiolysis: (i) a mixture
of 10 mbar of NO and 990 mbar of SF6 and (ii) a mixture of 20
mbar of CH3Br, 2.5 mbar of NO, and 997.5 mbar of SF6. In
reaction mixture i the following reactions are important:
27, and 32% of spectrum A from spectrum B are shown. It is
evident from the figure that the FNO yield in reaction mixture
ii is 27 ( 5% of that in reaction mixture i.
Two corrections are needed before a final value of k8a/k8 can
be derived. First, some F atoms are lost because a full radiolysis
dose was used to derive the spectrum of mixture ii. In Figure
-
SF + (2 MeV)e f F + products
(9)
6
9
the maximum transient absorbance at 310.5 nm is plotted as
a function of the radiolysis dose following radiolysis of reaction
mixtures i and ii. The maximum transient absorbance is
proportional to the dose over the entire range tested for reaction
mixture i. This indicates that F atoms are converted quantita-
tively into FNO. This is not the case for reaction mixture ii.
The full-dose absorbance (0.0304) falls below the extrapolation
of the low-dose data (0.0355). This indicates that radicals are
lost by secondary chemistry. The amount of FNO expected if
secondary chemistry was absent relative to the amount of FNO
actually formed is 0.0355/0.0304 ) 1.17. We estimate the
uncertainty on this correction to be 10%. The corrected yield
of FNO is then ) (0.27 ( 0.05) × (1.17 ( 0.12) ) 0.32 (
0.07.
F + NO + M f FNO + M
(18)
The transient absorbance 100 µs after the electron pulse is shown
in Figure 8A. The formation of FNO is complete (k18 ) 5.1 ×
-12
3
-1 -1 22,23
10
cm molecule
s
) 100 µs after the radiolysis pulse.
The two peaks at 308 and 311 nm are well-known FNO
2
4
features. The absorbance following the radiolysis of reaction
mixture ii is shown in Figure 8B. Clearly, FNO is also formed
in system ii. It seems reasonable to conclude that FNO is
formed via reaction 16. The ratio of the amount of FNO formed
in experiments ii and i gives us the ratio of k8a/k8. The FNO
yield was determined by subtraction of spectrum A from B. In
Figure 8 the residual spectra following the subtraction of 22,

Reaction of F Atoms with CH3Br
J. Phys. Chem., Vol. 100, No. 26, 1996 10995
3 2
Figure 10. Spectra of the CH Br--F adduct (this work) and the CF -
Figure 11. Reciprocal of the decay half-life at 280 nm plotted as a
function of the maximum transient absorbance at 280 nm. The straight
line is a linear regression.
1
1
BrH--F adduct. The straight line in the [CH
extrapolation between two spectra.
3
Br--F] spectrum is a
The second correction stems from the fact that 1.4% of the
F atoms react directly with NO via reaction 18. The value of
plotted in Figure 10, and the absorption cross sections are listed
in Table 1. Also listed in Table 1 is the uncertainty calculated
by a conventional error propagation method using 15% uncer-
tainty in σ(observed), k /k , and σ(CH Br). As seen from
1
.4% was calculated from the initial concentrations of CH3Br
-12
3
(
20 mbar) and NO (2.5 mbar) and k18 ) 5.1 × 10
cm
-1
-1 22,23
-11
3
-1 -1
molecule
s
and k8 ) 4.46 × 10 cm molecule s .
8a
8
2
Figure 10 the absorption spectrum of the CH3Br--F adduct is
Application of this final correction gives k8a/k8 ) 0.31 ( 0.07.
This value is in excellent agreement with the value of 0.31 (
1
1
similar to the spectrum reported for the CF2BrH--F adduct.
0
.08 derived in section 3.5. The value determined from the
3.8. Decay of the CH Br--F Adduct. Figure 1B shows an
3
reaction of the CH3Br--F adduct with NO could potentially be
lower than the real value, if the reaction gives products other
than FNO and CH3Br. However, it seems likely that reaction
experimental absorption transient recorded at 280 nm following
radiolysis of a mixture of 2.5 mbar of CH Br and 1000 mbar
3
of SF . In Figure 11, 1/t is plotted as a function of the
6
1/2
1
6 gives 100% FNO and CH3Br because the F atom in the
maximum transient absorbance at 280 nm. The maximum
absorbance was varied by varying the radiolysis dose. As seen
from Figure 11, the decay rate increased linearly with radical
concentration (higher absorbance). This indicates that radical-
radical reactions such as
adduct is expected to be only weakly bound and thus removed
easily by NO. While the determination of k8a/k8 ) 0.31 ( 0.07
in this section and that of k8a/k8 ) 0.31 ( 0.08 given in section
3
.5 should be strictly viewed as lower limits, the excellent
agreement between these two independent methods suggests that
these determinations are close to the actual value of k8a/k8.
Hence, we choose to report a value of k8a/k8 of {(0.31 ( 0.07)
CH Br--F + CH Br f products
(19)
(20)
(21)
3
2
+
(0.31 ( 0.08)}/2 ) 0.31 ( 0.05.
.7. UV Absorption Spectrum of the CH3Br--F Adduct.
3
CH Br--F + CH Br--F f products
3 3
We are now in a position to calculate the UV absorption
spectrum of the CH3Br--F adduct using three pieces of informa-
tion: (i) the observed UV spectrum determined in this work,
CH Br--F + F f products
3
(
ii) the branching k8a/k8 ) 0.31 ( 0.05, and (iii) the UV
absorption spectrum of CH2Br radicals reported by Villenave
and Lesclaux. The formula σ([CH3Br--F]) ) (σ(observed) -
0
tion cross section at a given wavelength of the CH3Br--F adduct,
σ(observed) is the observed spectrum following the reaction of
F atoms with CH3Br (see Figure 5), and σ(CH2Br) is the CH2-
Br spectrum reported by Villenave and Lesclaux (see Figure
5
dominate removal of the CH3Br--F adduct. The slope of the
7
5
straight line through the data in Figure 11, (8.9 ( 0.7) × 10
-1
.69σ(CH2Br))/0.31 was used where σ(CH3Br--F) is the absorp-
s , is a measure of the speed of such radical-radical reactions.
However, it is a measure of a mixture of reactions 19-21, and
we will not pursue this further. The intercept in Figure 11 is
4
-1
(0.85 ( 1.24) × 10 s and contains important information
regarding the equilibrium between F atoms, CH3Br, and the CH3-
Br--F adduct.
7
). The spectrum of the adduct derived using this approach is
TABLE 1: Selected UV Absorption Cross Sections with Uncertainties for the CH
3
Br--F Adduct
20
2
σ × 10 (cm molecule-1)
20
2
σ × 10 (cm molecule)
Br--F) ∆σ(CH
wavelength (nm)
σ(CH
3
3
Br--F)
wavelength (nm)
σ(CH
3
Br--F)
3
∆σ(CH Br--F)
2
2
2
2
2
2
2
3
3
30
40
50
60
70
80
90
00
10
140
640
400
390
265
205
238
310
319
270
190
320
330
340
350
360
370
380
390
400
945
645
418
263
156
100
73
140
95
63
40
22
16
13
8
1125
1280
1775
2055
2140
1830
1300
51
29
5

1
0996 J. Phys. Chem., Vol. 100, No. 26, 1996
Sehested et al.
There are two possible first order loss mechanisms for the
CH3Br--F adduct. These are via reactions -8a and 15.
F + CH Br h CH Br--F f products
(8a, -8a, 15)
(8b)
3
3
F + CH Br f CH Br + HF
3
2
Decomposition via reaction -8a gives F atoms and CH3Br. The
F atoms then react again with CH3Br; 31% re-form the adduct,
while 69% do not re-form the adduct. Hence, the observed
decay rate (intercept in Figure 11) could be interpreted as being
6
9% of k-8a. The intercept in Figure 11 gives k-8a < 2.1 ×
4
-1
4
-1
1
0 ((ln 2)/0.69) s , i.e., <2.1 × 10 s . This is slower than
the decomposition rate of the CF2BrH--F adduct of (9 ( 3) ×
0 s determined by Bilde et al., suggesting that the CH3-
5
-1
11
1
Br--F adduct is more strongly bound than the CF2BrH--F adduct.
Decomposition via reaction 15 does not regenerate F atoms,
and the intercept in Figure 11 gives an upper limit of k15 <
Figure 12. Spectrum of CH
spectrum of CH BrO
2
BrO
2
determined in this work. The
4
-1
4
-1
2
.1 × 10 ln 2 s ) 1.5 × 10 s . This can be compared
7
2
2
reported by Villenave and Lesclaux (filled
with the analogous decomposition rate of the CF2BrH--F adduct
circles) is shown for comparison. The insert shows the absorbance
following radiolysis of a mixture of 5 mbar of CH Br, 50 mbar of O
and 945 mbar of SF (upper curve). The two low-absorbance curves
in the insert are the absorbance due to FO (maximum at 230 nm) and
the CH Br--F adduct (maximum at 290 nm). See text for details.
4
-1
which is (7.4 ( 0.7) × 10 s . The CH3Br--F adduct is more
stable than the CF2BrH--F adduct, both toward decomposition
into F atoms and the parent compound and other products.
k8a can be calculated from k8a/k8 ) 0.31 ( 0.05 and the overall
3
2
,
6
2
3
-
11
3
rate constant of reaction 8, k8 ) (4.46 ( 0.22) × 10
molecule s . k8a ) (1.4 ( 0.2) × 10 cm molecule
s . This rate is, within the error limits, equal to the rate for
cm
Lesclaux is 1.05 × 10- cm molecule s . Using the
absorbance shown in the insert in Figure 12, the UV absorption
spectrum of CH BrO can be determined. Three corrections
12
3
-1 -1 7
-1
-1
-11
3
-1
-1
-
11
3
the addition of F atoms to CF2BrH of (1.5 ( 0.5) × 10 cm
2
2
molecule- s . The similarity in their formation kinetics and
1
-1
2 2
are necessary to extract the spectrum of CH BrO from the
observed spectrum. (i) The spectrum has to be corrected for
the absorption due to the CH3Br--F adduct. The absorbance at
UV spectra (see Figure 10) suggests that the adducts formed
by the reaction of F atoms with CF2BrH and CH3Br are similar
in nature.
2
80 nm of the CH3Br--F adduct in this system 33 µs after the
electron pulse is approximately 0.03 absorbance units; see Figure
B. The UV absorption spectrum of the CH3Br--F adduct given
A lower limit for the equilibrium constant for the CH3Br--F
adduct can be calculated from the k8a, and k-8a values using
1
-
16
3
-1
in Figure 10 was scaled to an absorbance of 0.03 at 280 nm to
give the result shown in the insert in Figure 12 which was then
subtracted from the observed absorbance spectrum. (ii) Also
shown in the insert in Figure 12 is the absorbance of FO2 formed
in the system using the absorption cross sections reported by
Keq ) k8a/k-8a > 6 × 10
cm molecule . This should be
compared with the equilibrium constant for the CF2BrH--F
-
17
3
-1
adduct of (1.6 ( 0.1) × 10 cm molecule . The CH3Br--F
adduct is more stable than the CF2BrH--F adduct.
3
.9. Reaction of CH3Br--F Adduct with O2 and the
2
5
Ellermann et al. The fraction of F atoms that is converted
Spectrum of CH2BrO2. Three mixtures containing 5 mbar of
CH3Br, 950 mbar of SF6, and either 0, 50, or 110 mbar of O2
were subject to pulse radiolysis (32% of maximum dose), and
the transient absorption at 280 nm was recorded. The decay
half-lives of the absorbance transients were 6.0, 6.8, and 6.8
µs. The shorter half-life in the absence of O2 could indicate
that the CH3Br--F adduct reacts with CH2Br radicals but does
not react, or reacts only slowly, with O2. If we assume a 10%
error limit on the decay half-lives above, the upper limit for
the rate constant of the reaction of the CH3Br--F adduct with
O2 is {(ln 2)/(6.8 µs) - (ln 2)/(6.8 + 0.68 µs)}/(110-50 mbar)
-13
3
into FO2 can be calculated using k(F+O2) ) 1.9 × 10
-1 -1
cm
molecule- s , k8 ) 4.46 × 10
1
-1 25
-11
cm molecule s , and
3
the initial concentrations of CH3Br, 5 mbar, and O2, 50 mbar.
The fraction of F atoms that is converted into FO2 radicals is
4
%. The fraction of F atoms that forms CH2Br radicals is then
0
.69 × 0.96 ) 0.66. (iii) Finally, we need a correction due to
the self-reaction of the CH2BrO2 radical. Using a F atom yield
1
5
-3
at full dose of 3 × 10 molecules cm , 53% of full dose, a
conversion factor for the F atoms into CH2BrO2 radicals of 0.66,
and the self-reaction rate constant for CH2BrO2 radicals reported
7
-
15
3
-1 -1
by Villenave and Lesclaux, we calculate that 7% of the CH -
)
6 × 10
cm molecule s .
2
BrO2 radicals are lost in the 33 µs between radical formation
and radical detection.
The spectrum can now be calculated from the corrected
The CH2BrO2 radical is expected to be formed upon radiolysis
of mixtures of CH3Br/O2/SF6 by the reaction of CH2Br radicals
with O2:
1
5
-3
absorbance, the initial fluorine atom yield of 3 × 10 cm ,
53% of maximum dose, the UV path length of 80 cm, and the
fraction of F atoms converted into CH2Br radicals, 0.66, using
the following formula: σ(CH2BrO2) ) absorbance/(80 cm × 3
CH Br + O + M f CH BrO + M
(22)
2
2
2
2
The UV absorption spectrum of CH2BrO2 was recorded using
the diode array 30-35 µs after radiolysis (53% of maximum
dose) of mixtures of 5 mbar of CH3Br, 50 mbar of O2, and 945
mbar of SF6. The UV path length was 80 cm. The absorption
is shown in the insert in Figure 12. The reason for delaying
the data acquisition 30-35 µs after the pulse is to allow time
for the rapid decay of the CH3Br--F adduct. The decay of the
CH2BrO2 radical is expected to be slow since the self-reaction
rate constant of CH2BrO2 radicals determined by Villenave and
1
5
-3
× 10 cm × 0.66 × 0.53). The spectrum of the CH2BrO2
radical is shown in Figure 12, and selected absorption cross
sections are listed in Table 2. The spectrum of CH2BrO2
7
reported by Villenave and Lesclaux is shown in Figure 12 for
comparison. It is seen from the figure that the spectrum
determined here is in excellent agreement with the spectrum of
7
Villenave and Lesclaux. We estimate the overall uncertainty
on the absorption cross section of CH2BrO2 to be 25%.

Reaction of F Atoms with CH3Br
J. Phys. Chem., Vol. 100, No. 26, 1996 10997
4. Discussion and Conclusion
TABLE 2: Selected UV Absorption Cross Sections for the
CH BrO Radical
2 2
The kinetics and mechanism of the reaction of F atoms with
2
0
20
σ × 10
σ × 10
26,27
CH3Br have been studied previously by Iyer and Rowland,
wavelength (cm molecule-1) wavelength (cm molecule-1)
2
2
6
21
Nielsen et al., and Wallington et al. Iyer and Rowland were
the first to study this reaction and reported that at 283 K in the
presence of 3000 Torr of SF6 diluent the reaction proceeds
predominately via hydrogen atom abstraction with a small
(0.27%) contribution by a direct displacement mechanism to
give CH3F and Br. The rate constant for hydrogen abstraction
(nm)
2 2
σ(CH BrO )
(nm)
2 2
σ(CH BrO )
2
30
40
50
60
345
377
341
261
270
280
290
300
172
100
52
2
2
2
24
-11
3
reported by Iyer and Rowland was (6.1 ( 0.7) × 10
cm
Several potential reactions could interfere with the determi-
nation of the absorption spectrum of CH2BrO2 above:
molecule- s . Nielsen et al. used experimental techniques
similar to those employed here and reported an overall rate
constant for the reaction of F with CH3Br at 298 K in 1000
1
-1
6
F + CH Br/CH Br--F f products
(23)
(24)
2
3
-11
3
-1 -1
mbar of SF6 diluent of (4.5 ( 0.9) × 10 cm molecule
s
and an upper limit of 3% for the CH3 radical yield. Nielsen et
al. assumed that the reaction proceeded via hydrogen atom
abstraction. Wallington et al. used a relative rate technique
CH Br + CH Br f products
2
2
6
2
1
to derive k(F+CH3Br)/k(F+CH4) ) 0.44 ( 0.03 at 295 K in
CH Br + CH Br--F f products
(25)
2
3
7
2
00 Torr of air diluent. In the present work we show that at
96 K in 1000 mbar of SF6 diluent reaction 8 proceeds with an
-11
3
-1
CH BrO + CH Br--F f products
(26)
(27)
overall rate constant of (4.46 ( 0.22) × 10 cm molecule
2
2
3
-1
s
with 31% of the reaction giving an adduct and 69% giving
CH2Br + HF. In 700 Torr of air diluent we measure k(F+CH3-
Br)/k(F+CH4) ) 0.51 ( 0.03. The absolute value of k8
measured here is consistent with that reported by Nielsen et
F + CH BrO f products
2
2
By analogy to the reaction of other alkyl radicals with O28,10
the rate constant for the addition of O2 to the CH2Br radical is
6
al. but, for unknown reasons, is significantly lower than that
2
7
measured by Iyer and Rowland. The value of k(F+CH3Br)/
k(F+CH4) determined using the FTIR method is slightly (16%)
-
12
3
-1 -1
expected to be of the order of 3 × 10
cm molecule s .
Using this rate constant, the lifetime for CH2Br radicals with
respect to reaction with O2 is 0.27 µs. This should be compared
with a lifetime for CH2Br radicals and the CH3Br--F adduct
with respect to loss via self-reaction and cross reaction of
approximately 5 µs. Hence, reactions 24 and 25 are expected
to be insignificant in the reaction system. The lifetime for the
formation of CH2Br radicals and CH3Br--F adducts is 0.18 µs;
thus, the influence of reaction 23 is negligible. Using the decay
lifetime for F atoms and the formation lifetime of CH2BrO2
21
larger than the value reported by Wallington et al. This may
reflect a slight underestimation of the CH3Br loss in the previous
work. In the present work two additional reference compounds
were used (HFC-152a and CD4), and the results from these
references are more consistent with a ratio of k(F+CH3Br)/
k(F+CH ) ) 0.51 ( 0.03 than with a ratio of 0.44 ( 0.03.
4
Accordingly, we recommend that the present measurements
supersede our previous determination.
The novel aspect of the present work is that it shows that
reaction 8 proceeds via a two channel mechanism:
-
10
3
-1 -1
radicals and k27 ) 10
integration program, we estimate the loss of reactants to be
0.7% due to reaction 27. The significance of reaction 26,
cm molecule
s
in a numerical
F + CH Br h CH Br--F
(8a, -8a)
3
3
<
however, cannot be determined, since both the CH3Br--F adduct
and the CH2BrO2 radical will be present in the system for several
microseconds. Reaction (26) could be important. However,
the excellent agreement between the spectrum of CH2BrO2
radicals reported here and the spectrum of Villenave and
Lesclaux suggests that reaction 26 is not important.
F + CH Br f CH Br + HF
(8b)
3
2
We have used this mechanism and the CH Br spectrum of
2
7
Villenave and Lesclaux to derive the spectrum of the CH Br-
3
-F adduct. This spectrum is similar to the spectrum reported
previously for the CF2BrH--F adduct shown in Figure 11. In
addition, the rate constants for addition of F atoms to CH3Br,
The goal of the present series of experiments is not to
remeasure the absorption spectrum of CH2BrO2 radicals but
rather to understand the difference between the spectrum
reported previously from our laboratory and that of Villenave
-
11
3
-1 -1
(
0
1.4 ( 0.2) × 10
cm molecule s , and CF2BrH, (1.5 (
-
11
3
-1 -1
.5) × 10
cm molecule s , are, within the experimental
uncertainties, indistinguishable. These reaction rates were
measured both absolute and using relative techniques with
consistent results. The stability of the CH3Br--F adduct is
greater than that of the CF2BrH--F adduct. The equilibrium
7
and Lesclaux. The results reported herein show that, contrary
to our initial expectations, the reaction of F atoms with CH3Br
does not proceed via a simple hydrogen atom abstraction
mechanism but instead proceeds via two channels of which one
produces an adduct which absorbs strongly in the UV. In the
-16
3
constant in the CH3Br system is k8a/k-8a > 5 × 10
cm
-1
molecule , while the analogous equilibrium constant in the CF2-
-17
3
-1 11
BrH system is (1.6 ( 0.1) × 10 cm molecule . It seems
reasonable to assume that the entropy change in the formation
of the CH3Br--F and CF2Br--F adducts is comparable. Hence,
it can be concluded from the ratio of the equilibrium constants
that the F atom in the CH3Br--F adduct is bound by at least 2
6
previous CH2BrO2 spectrum reported from this laboratory the
absorption measured a few microseconds after the electron pulse
were used to determine the spectrum of CH2BrO2. This
spectrum includes absorption by the adduct because the adduct
is not scavenged by O2 and does not decay within a few
microseconds. Figure 12 shows that when suitable corrections
are applied to account for the formation of the CH3Br--F adduct,
the spectral data from the pulse radiolysis experiments can be
reconciled with that of Villenave and Lesclaux.
-1
kcal mol more than the CF2BrH--F adduct. The estimated
4
lower limit of the bond energy obtained from k < 2.2 × 10
-
8a
-1
-1
13 -1
s
and A-8a > 10
s
is Ea-8a > 12 kcal mol .
All evidence in the spectrokinetic study of the products of
the reaction of F atoms with CH3Br and CF2BrH points toward

1
0998 J. Phys. Chem., Vol. 100, No. 26, 1996
Sehested et al.
formation of F adducts with the parent species. The question
is: What is the nature of these species? Since F atoms form
adducts with CH3Br and CF2BrH, it seems reasonable to suggest
that in the adduct the F atom is bound to the bromine atom in
the parent compound. One experimental observation appears
to be in conflict with this conclusion; radiolysis of SF6 in the
presence of CF3Br or CCl3Br does not give significant absorp-
tion at 280 nm. Two potential explanations can be given for
this. First, perhaps the formation of the adduct requires the
presence of a Br atom and a H atom in the molecule. Therefore
no adduct is formed from the reaction of F atoms with CF3Br
or CCl3Br. Second, perhaps the equilibrium constant for the
equilibrium between F atoms and R-Br is shifted so much
toward the left in the case of CF3Br and CCl3Br that we cannot
observe the adduct formation in these cases. From the stability
of the CH3Br--F and CF2BrH--F adducts, the stability of the
complex seems to decrease when fluorine is present in the
molecule. This trend is consistent with the inductive effect of
the F atoms, reducing the electron density associated with the
Br atom and hence reducing the adduct stability. This observa-
tion supports the suggestion made above that it is a lower
equilibrium constant in the case of CF3Br and CCl3Br that leads
to the failure in observing an adduct. From the absence of any
detectable absorption and assuming that the CF3Br--F and CCl3-
Br--F adducts have absorption cross sections at 280 nm which
are comparable to those of CH2Br--F and CF2BrH--F, i.e., 1 ×
References and Notes
(1) Scientific Assessment of Ozone Depletion; Global Ozone Research
and Monitoring Project, Report No. 25; World Meteorological Organiza-
tion: 1991.
(2) Report on Concentrations, Lifetimes, and Trends of CFCs, Halons,
and Related Species; NASA Reference Publication 1339; National Aero-
nautics and Space Administration: Washington, DC, 1994.
(3) Mellouki, A.; Talukdar, R. K.; Schmoltner, A. M.; Gierczak, T.;
Mills, M. J.; Solomon, S.; Ravishankara, A. R. J. Geophys. Res. 1992, 19,
2059.
(
4) Chichinin, A.; Teton, S.; Le Bras, G.; Poulet, G. J. Atmos. Chem.
992, 24, 2413.
5) Zhang, Z.; Saini, R. D.; Kurylo, M. J.; Huie, R. E. J. Geophys.
Res. 1992, 24, 2413.
6) Nielsen, O. J.; Munk, J.; Locke, G.; Wallington, T. J. J. Phys. Chem.
1991, 95, 8714.
(7) Villenave, E.; Lesclaux, P. Chem. Phys. Lett. 1995, 236, 376.
8) Lightfoot, P. D.; Cox, R. A.; Crowley, J. N.; Destriau, M.; Hayman,
G. D.; Jenkin, M. E.; Moortgat, G. K.; Zabel, F. Atmos. EnViron. 1992,
6A, 1805.
9) Wallington, T. J.; Dagaut, P.; Kurylo, M. J. Chem. ReV. 1992, 92,
1
(
(
(
2
(
667.
(10) Orlando, J. J.; Tyndall, G. S.; Wallington, T. J. J. Phys. Chem.
996, 100, 7026.
11) Bilde, M.; Sehested, J.; Møgelberg, T.; Wallington, T. J.; Nielsen,
O. J. J. Phys. Chem. 1996, 100, 7050.
12) Wine, P. H. Paper presented at the National ACS meeting in
1
(
(
Chicago, August 1995.
(13) Sehested, J. Ph.D. Thesis, Risø-R-804, 1994.
(
14) Hansen, K. B.; Wilbrandt, R.; Pagsberg, P. ReV. Sci. Instrum. 1979,
-17
2
-1
-18
3
50, 1532.
1
0
cm molecule , we provide estimates of 1 × 10
cm
(
(
15) Wallington, T. J.; Japar, S. M. J. Atmos. Chem. 1989, 9, 399.
16) Rasmussen, O. L.; Bjergbakke, E. Risø-R-2430, Risø National
-1
molecule for the upper limits of the equilibrium constants for
the CF3Br--F and CCl3Br--F adducts at 296 K.
Laboratory: Roskilde, Denmark, 1984.
Further experimental studies of the reactions of F atoms with
CF3Br and CCl3Br at low temperatures, the reaction of F atoms
with other brominated compounds and theoretical calculations
of the nature of the complexes are needed to provide more
information on the structure of the adduct formed.
(17) Wallington, T. J.; Hurley, M. D. Chem. Phys. Lett. 1992, 189, 437.
(18) Wallington, T. J.; Hurley, M. D.; Shi, J.; Maricq, M. M.; Sehested,
J.; Ellermann, T.; Nielsen, O. J. Int. J. Chem. Kinet. 1993, 25, 651.
(19) Wallington, T. J.; Hurley, M. D. Chem. Phys. Lett. 1992, 193, 84.
(20) DeMore, W. B.; Golden, D. M.; Hampson, R. F.; Kurylo, M. J.;
Howard, C. J.; Ravishankara, A. R.; Kolb, C. E.; Molina, M. J. JPL Publ.
994, 94-26.
During the course of the present experiments Wine and co-
1
workers reported evidence for the formation of adducts in the
(21) Wallington, T. J.; Hurley, M. D.; Shi, J.; Maricq, M.; Sehested, J.;
Nielsen, O. J.; Ellermann, T. Int. J. Chem. Kinet. 1993, 25, 651.
reactions of Cl atoms with CH3Br and CH3I.12 It is possible
(22) Sehested, J. Int. J. Chem. Kinet. 1994, 26, 1023.
that adduct formation is a common mechanism of the reaction
of halogen atoms with haloalkanes. Further experimental and
computational work is needed to explore this possibility.
(23) Wallington, T. J.; Ellermann, T.; Nielsen, O. J.; Sehested, J. J. Phys.
Chem. 1994, 98, 2346.
(24) Burley, J. D.; Miller, C. E.; Johnston, H. S. J. Mol. Spectrosc. 1993,
1
58, 377.
Acknowledgment. We thank Steve Japar (Ford Motor Co.)
for a careful reading of the manuscript, and Paul Wine (Georgia
Institute of Technology) for helpful discussions regarding the
stability of adducts formed from the reaction of halogen atoms
with CH3Br and CH3I.
(25) Ellermann, T.; Sehested, J.; Nielsen, O. J.; Pagsberg, P.; Wallington,
T. J. Chem. Phys. Lett. 1994, 218, 287.
(
(
26) Iyer, R. S.; Rowland, F. S. J. Phys. Chem. 1981, 85, 2488.
27) Iyer, R. S.; Rowland, F. S. J. Phys. Chem. 1981, 85, 2493.
JP953397F