Rate coefficients for the reactions CH3 + Br2 (224-358 K), CH3CO + Br2 (228 and 298 K), and Cl + Br 2 (228 and 298 K)

Article abstract of DOI:10.1002/kin.20505

Rate coefficients for the reactions of CH₃ + Br₂ (k₂), CH₃CO + Br₂ (k₃), and Cl + Br₂ (k₅) were measured using the laser-pulsed photolysis method combined with detection of the product Br atoms using resonance fluorescence. For the reactions involving organic radicals, the rate coefficients were observed to increase with decreasing temperature and within the temperature range explored, were adequately described by Arrhenius-like expressions: k₂ (224-358 K) = 1.83×10^-11 exp(252/T) and k₃ (228-298 K) = 2.92×10^-11 exp(361/T) cm ³ molecule^-1 s¹. The total, temperature-independent uncertainty for each reaction (including possible systematic errors in Br₂ concentration measurement) was estimated as ～7% for k₂ and 10% for k₃. Accurate data on k5 was obtained at 298 K, with a value of 1.88×10^-10 cm³ molecule^-1 s^-1 obtained (with an associated error of 6%). A limited data set at 228 K suggests that k₅ is, within experimental uncertainty, independent of temperature.

Full text of DOI:10.1002/kin.20505

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Rate Coefficients for the
Reactions CH3 + Br2
(224–358 K), CH3CO + Br2
(228 and 298 K), and Cl +
Br2 (228 and 298 K)
V. KHAMAGANOV, J. N. CROWLEY
Division of Atmospheric Chemistry, Max-Planck-Institut fu¨r Chemie, Postfach 3060, 55020 Mainz, Germany
Received 1 February 2010; revised 23 March 2010; accepted 25 March 2010
DOI 10.1002/kin.20505
Published online 25 June 2010 in Wiley InterScience (www.interscience.wiley.com).
ABSTRACT: Rate coefficients for the reactions of CH3 + Br2 (k2), CH3CO + Br2 (k3), and Cl +
Br2 (k5) were measured using the laser-pulsed photolysis method combined with detection of
the product Br atoms using resonance fluorescence. For the reactions involving organic radi-
cals, the rate coefficients were observed to increase with decreasing temperature and within
the temperature range explored, were adequately described by Arrhenius-like expressions: k2
(224–358 K) = 1.83 × 10−11 exp(252/T) and k3 (228–298 K) = 2.92 × 10−11 exp(361/T) cm3
molecule−1 s1. The total, temperature-independent uncertainty for each reaction (including
possible systematic errors in Br2 concentration measurement) was estimated as ∼7% for k2
and 10% for k3. Accurate data on k5 was obtained at 298 K, with a value of 1.88 × 10−10 cm3
molecule−1 s−1 obtained (with an associated error of 6%). A limited data set at 228 K sug-
ꢀ
C
gests that k5 is, within experimental uncertainty, independent of temperature. 2010 Wiley
Periodicals, Inc. Int J Chem Kinet 42: 575–585, 2010
(sum of OH and HO2) and also stable nitrates (PAN,
CH3C(O)O2NO2) [1–5]. In a recent publication [6],
we presented primary photodissociation quantum yield
measurements of acetone based on a scavenging tech-
nique, whereby the initially formed organic radicals
were converted to Br atoms, which were subsequently
detected (R1)–(R3).
INTRODUCTION
Our major motivation to study the reactions of CH3
and CH3CO radicals and Cl atoms with Br2 is related
to the role of acetone photolysis in atmospheric chem-
istry. The atmospheric photodissociation of acetone at
wavelengths above ≈300 nm results in formation of
CH3 and CH3CO radicals, which react with oxygen
to form organic peroxy radicals and eventually HOx
CH3C(O)CH3 + hν → CH3 + CH3CO
CH3 + Br2 → CH3Br + Br
(R1)
(R2)
Correspondence to: J. N. Crowley; e-mail: john.crowley@
mpic.de.
c
ꢀ
2010 Wiley Periodicals, Inc.
CH3CO + Br2 → CH3C(O)Br + Br (R3)

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
576
KHAMAGANOV AND CROWLEY
The photolysis of Cl2 in the presence of Br2 was used
as reference Br atom source for these studies:
concentrations measured by long-path UV-absorption
spectroscopy. The reaction of Cl with Br2 (the simplest
system analytically) was studied at 298 K as a test of
the method.
Cl2 + hν → 2Cl
(R4)
(R5)
Cl + Br2 → ClBr + Br
EXPERIMENTAL
Based on heats of formation [7,8], (R2) and (R3) are
exothermic by ≈105 kJ mol−1 and (R5) is exothermic
Various features of the pulsed laser photolysis/
resonance ßuorescence (PLP/RF) experimental setup
(Fig. 1) have been described previously [6,9], and
only a brief summary is given here. The central
component of the experimental setup is a 500-cm3,
thermostated (224–358 K), cylindrical photoreac-
tor/RF cell constructed of quartz. Photolysis light was
provided by an excimer laser (Lambda-Physik Lextra
50) operated at 193 or 248 nm, a YAG laser (Quantel
Brilliant-B) operated with fourth harmonic generation
at 266 nm, or a frequency-doubled, YAG-pumped
dye-laser tuned to 300 or 308 nm. The laser entrance
and exit windows were purged by a small ßow of
N2 (≤5 sccm each) to keep them clean. Br was
detected using resonance ßuorescence with photon
counting as described by Khamaganov et al. [6] in
by 27 kJ mol−1
.
In the experiments of Khamaganov et al. at 248 and
266 nm [6], detailed kinetic information on reactions
(R2) and (R3) was not necessary for the derivation
of quantum yields as there were no competing pro-
cesses for Br formation. At longer wavelengths, how-
ever, (e.g., at 308 nm), it was suggested that Br could
be formed via reactions of triplet acetone with Br2,
necessitating a more detailed kinetic analysis of the
chemical system and thus accurate rate constants for
(R2) and (R3) at temperatures as low as 228 K. As low
temperature kinetic data were not available for either
(R2) or (R3), a series of experiments was conducted
using the pulsed laser photolysis method combined
with resonance ßuorescence detection of Br, with Br2
Figure 1 Experimental setup. IF = 185 nm interference Þlter, PD = photodiode, Pen Ray = low pressure “Pen Ray” Hg-lamp,
VUV PMT = solar blind photomultiplier tube, FC = gas-mixing and ßow control, J = Joule-meter, and MCS = multichannel
scaler. The PMT-Box car detection axis (used for laser-induced ßuorescence studies) was not operated in this work.
International Journal of Chemical Kinetics DOI 10.1002/kin

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
RATE COEFFICIENTS FOR THE REACTIONS CH3 + Br2, CH3CO + Br2, AND Cl + Br2
577
which details of the photomultiplier, optical Þlters,
and electronics are given. At a time resolution of
2 μs, a detection sensitivity of ∼6 × 107 Br cm−3
was achieved by coadding 2000–3000 scans, usually
at 10 Hz. Precise measurement of the concentration
of Br2 was achieved using a monochromator–diode
array setup coupled to a multipass absorption cell with
892-cm optical path length. The absorption cell was
located upstream of the photolysis reactor and was
at room temperature. The concentration of Br2 was
determined by online optical density measurements
before and after each experimental run at wavelengths
between ∼320 and 460 nm and least-squares Þtting
to reference spectra of Br2 [10]. With this method,
we conservatively estimated the uncertainty in the
measured Br2 concentration to be ≈5%, mostly related
to errors in the cross section. The concentration of
acetone was monitored by absorption at 185 nm [9].
All ßows were regulated by mass ßow controllers
(MKS Instruments) and well mixed in a glass mani-
fold before entering the absorption cell. The pressure
in the reactor was measured by a 100-Torr capaci-
tance manometer (MKS Instruments) and adjusted to
60 Torr by throttling the pump downstream. The typ-
ical ßow rates in the cell were 500–800 cm3 (STD)
min−1, which, combined with a laser repetition rate of
between 5 and 10 Hz, ensured that a fresh mixture was
irradiated with each laser pulse. The approximate laser
ßuence was monitored using an energy meter placed
behind the photoreactor.
Reactions (R2) and (R3) are exothermic by ≈105 kJ
mol−1, which is sufÞcient to result in population of the
spin-excited state of Br, the difference in energy be-
tween the Br(2P1/2) and Br(2P3/2) spin states being
44 kJ mol−1 [11]. For reaction between CH3 and Br2,
Kovalenko and Leone [12] suggest that 42% of the
available energy is found in vibrations of the alkyl-
halide fragment. Translational energy has been mea-
sured to account for a further 22% [13] or 56% [14]
of that available. The energy available for formation of
Br(2P1/2) is thus between ≈2 and 38 kJ mol−1, imply-
ing that spin-excited Br atoms should not play a sig-
niÞcant role. There are no data on the product energy
distribution for the CH3CO + Br2 reaction, though ex-
periments in which a few Torr of H2 were added to the
usual CH3C(O)CH3/Br2 mixture prior to photolysis at
248 nm revealed no difference in the Br atom resonance
ßuorescence proÞles. As H2 is an efÞcient quencher of
Br(2P1/2) [15], this result suggests that Br(2P1/2) need
not be further considered in our analysis.
AG, Mu¨nster, Germany; 99.999%), H2 (Linde, Pullach,
Germany; 99.999%), and Cl2 (4.76% mixture in N2;
Air Liquide, Du¨sseldorf, Germany) were used straight
from the bottles. Gases required for the resonance ßuo-
rescence lamp were He (Westfalen AG, 99.999%) and
CH4 (Messer, Sulzbach, Germany; 99.995%).
RESULTS AND DISCUSSION
CH3 + Br2 (k2)
The reaction between CH3 and Br2 was initiated by the
193-nm photolysis of CH3C(O)CH3 in the presence of
Br2. At 193 nm, the CH3CO radicals from (R1) are
formed in highly excited states and decompose instan-
taneously to CH3 + CO, the net result being formation
of two CH3 radicals and one CO per photon absorbed
[16–19]. There is also some formation of H atoms (and
an organic radical fragment) with a low branching ratio
(≈0.04) [18,20].
CH3C(O)CH3 + hν (193 nm) → 2CH3 + CO (R1a)
→ H + CH3C(O)CH2
(R1b)
As the formation of CH3 dominates, we initially ignore
channel (1b), though we return to this later. In the pres-
ence of Br2, the instantaneously formed CH3 radicals
react to form Br with the following biexponential rate
expression, describing the kinetics of their formation
and decay:
k2ꢁ
k2ꢁ − kd
ꢁ
d
2
[Br]t = 2[CH3]0 ×
× (e−k t − e−k t ) (E1)
where
k2ꢁ = k2[Br2] + c
(E2)
[Br]t is the time-dependent Br atom concentration and
k2 is the rate coefÞcient for reaction of CH3 with Br2.
[CH3]0 is the initial radical concentration formed in
the 193-nm laser pulse and kd is a Þrst-order loss rate
constant, representing transport of Br atoms from the
viewing zone (the only signiÞcant removal process for
Br in the system). The term c represents Þrst-order
losses of CH3 other than reaction with Br2. This ex-
pression is valid as long as second-order components
(e.g., reactions of CH3 radicals with each other or re-
actions of Br with CH3) are negligible. This condi-
tion is readily fulÞlled by keeping the concentration of
CH3 low relative to Br2. Typical concentrations of Br2
were ≈(5–50) × 1014 molecules cm−3, whereas initial
Acetone and Br2 (both Aldrich, Munich, Germany;
≥99.5%) were puriÞed using several freeze-pump-
thaw cycles at liquid nitrogen temperature and stored in
blackened glass bulbs as mixtures in N2. N2 (Westfalen
International Journal of Chemical Kinetics DOI 10.1002/kin

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
578
KHAMAGANOV AND CROWLEY
radical densities of approximately 1011 CH3 cm−3 were
generated from ≈5 × 1014 CH3C(O)CH3 cm−3 with
laser ßuences of <1 mJ cm−2. Experiments in which
the CH3C(O)CH3 and Br2 concentrations were held
constant and the laser ßuence was varied between 0.2
and 0.6 mJ cm−2 resulted in the same decay constant
for Br, conÞrming that secondary chemistry can be
neglected at these low radical densities.
Expression (E1) also assumes that (initially hot)
CH3 radicals formed in the 193-nm photolysis of
CH3C(O)CH3 are thermally equilibrated before react-
ing with Br2. Following Kovalenko and Leone [12], and
assuming the same vibrational energy transfer rate co-
efÞcient for O2 and N2, we estimate that vibrationally
hot CH3 radicals will be quenched in <10 μs at pres-
sures of 60 Torr N2. This is orders of magnitude smaller
than the millisecond time scale for reaction of CH3 with
Br2 and removing the Þrst 10 μs of data from the Br
proÞles did not change the value of k2ꢁ returned by the
Þt of the data to (E1).
Figure 3 Plots of k2ꢁ versus [Br2]. The solid lines are least-
squares Þts to the data according to expression (E2). The
error-bars (2σ) on each data point are generally smaller than
the symbol size.
Raw data obtained at 358 K and at two different
Br2 concentrations (4.02 × 1013 and 2.53 × 1014
molecules cm−3) are displayed in Fig. 2 along with
least-squares Þts using expression (E1). In the absence
of CH3C(O)CH3, no Br signal was observed indicat-
(essentially the volume of overlap between excimer
laser, resonance lamp emission, and focal point of col-
limating lens). The uncertainties quoted for k2ꢁ are 2σ
as returned by the Þt routine. According to (E2), a plot
of k2ꢁ versus [Br2] should be a straight line of slope k2.
Figure 3 displays values of k2ꢁ obtained at four different
temperatures (228, 258, 298, and 358 K) and various
Br2 concentrations. The slopes from these data sets re-
turn values of k2 = 5.50 0.02, 4.80 0.04, 4.20
ing that Br2 does not absorb signiÞcantly at 193 nm
193 nm
Br2
(σ
= 6.88 × 10−22 cm2 molecule−1) [21].
The values of k2ꢁ from least-squares Þtting expres-
sion (E1) to the two data sets displayed are 1613 64
and 9076 122 s−1 for the low and high Br2 concen-
trations, respectively. The Þrst-order loss term for Br
(kd) was about 60 s−1 in both cases, which is consis-
tent with diffusion from the ßuorescence-viewing zone
0.02, and 3.50
0.04 cm3 molecule−1 s−1 at 228,
258, 298, and 358 K, respectively, indicating a weak,
negative dependence on temperature. The uncertainties
quoted (2σ, statistical only) are of the order of a per-
cent, underlining the high precision of these data. The
intercept of these Þt lines (at [Br2] = 0) is not zero but
varies from 290 69 s−1 at 228 K to 185 35 s−1 at
358 K. This intercept can be partially attributed to dif-
fusive removal of CH3 from the ßuorescence-viewing
zone. To a Þrst approximation, the diffusion coefÞcient
varies inversely with the square root of the reduced
mass of diffusing molecule and bath gas and we ex-
pect that CH3 should diffuse through N2 about 50%
more rapidly than Br. This would result in losses of
≈100 s−1 for CH3, which is less than measured by a
factor of about 2. A further possibility for CH3 loss is
reaction with O2 present in the gas mixture (pure N2 at
60 Torr). The rate coefÞcient for reaction between CH3
and O2 (to form CH3O2) has a room temperature rate
coefÞcient at 60 Torr of ≈4 × 10−13 cm3 molecule−1
s
−1. A loss rate of 100 s−1 would thus require the
presence of impurity O2 at a volume-mixing ratio of
200 ppmv, which exceeds the stated impurity levels in
the N2 cylinder by a factor of 20. In addition, the rate
Figure 2 Experimental Br traces obtained at 358 K and
at two different concentrations of Br2. The solid lines are
least-squares Þts according to expression (E1).
International Journal of Chemical Kinetics DOI 10.1002/kin