Temperature-dependent kinetics of the gas-phase reactions of OH with Cl2, CH4, and C3H8

Article abstract of DOI:10.1021/jp047340h

The reactions of OH with molecular chlorine (reaction 1), methane (reaction 2), and propane (reaction 3) have been studied experimentally using a pulsed laser photolysis/pulsed-laser-induced fluorescence technique over wide ranges of temperatures (297-826, 298-1009, and 296-908 K, respectively) and at pressures between 6.68 and 24.15 kPa. The rate coefficients obtained for reactions 1-3 demonstrate no dependence on pressure and exhibit positive temperature dependences that can be represented with modified three-parameter Arrhenius expressions within their corresponding experimental temperature ranges: k ₁ = 3.59 × 10^-16 T^1.35 exp(-745 K/T) cm³ molecule^-1 s^-1, k₂ = 3.82 × 10^-19 T^2.38 exp(-1136 K/T) cm³ molecule^-1 s^-1, and k₃ = 6.64 × 10 ^-16 T^1.46 exp(-271 K/T) cm³ molecule ^-1 s^-1. For the OH + Cl₂ reaction, the potential energy surface has been studied using quantum chemical methods, and a transition-state theory model has been developed on the basis of calculations and experimental data. Model predictions suggest OH + Cl₂ → HOCl + Cl as the main channel of this reaction. The model results in the expression k₁ = 1.35 × 10^-16 T^1.50 exp(-723 K/T) cm³ molecule^-1 s^-1 for the temperature dependence of the reaction 1 rate coefficient extrapolation outside the experimental range to low temperatures down to 200 K and to high temperatures up to 3000 K. A temperature dependence of the rate coefficient of the HOCl + Cl → OH + Cl₂ reaction has been derived on the basis of the experimental data, modeling, and thermochemical information.

Full text of DOI:10.1021/jp047340h

10464
J. Phys. Chem. A 2004, 108, 10464-10472
Temperature-Dependent Kinetics of the Gas-Phase Reactions of OH with Cl2, CH4, and
C3H8
Mikhail G. Bryukov,† Vadim D. Knyazev,‡ Slawomir M. Lomnicki,† Cheri A. McFerrin,† and
Barry Dellinger*,†
Chemistry Department, Louisiana State UniVersity, Baton Rouge, Louisiana 70803, and Research Center for
Chemical Kinetics, Department of Chemistry, The Catholic UniVersity of America, Washington, D.C. 20064
ReceiVed: June 18, 2004; In Final Form: August 11, 2004
The reactions of OH with molecular chlorine (reaction 1), methane (reaction 2), and propane (reaction 3)
have been studied experimentally using a pulsed laser photolysis/pulsed-laser-induced fluorescence technique
over wide ranges of temperatures (297-826, 298-1009, and 296-908 K, respectively) and at pressures
between 6.68 and 24.15 kPa. The rate coefficients obtained for reactions 1-3 demonstrate no dependence on
pressure and exhibit positive temperature dependences that can be represented with modified three-parameter
Arrhenius expressions within their corresponding experimental temperature ranges: k1 ) 3.59 × 10-16 T1.35
exp(-745 K/T) cm3 molecule-1 s-1, k2 ) 3.82 × 10-19 T2.38 exp(-1136 K/T) cm3 molecule-1 s-1, and k3 )
6.64 × 10-16 T1.46 exp(-271 K/T) cm3 molecule-1 s-1. For the OH + Cl2 reaction, the potential energy
surface has been studied using quantum chemical methods, and a transition-state theory model has been
developed on the basis of calculations and experimental data. Model predictions suggest OH + Cl2 f HOCl
+ Cl as the main channel of this reaction. The model results in the expression k1 ) 1.35 × 10-16 T1.50
exp(-723 K/T) cm3 molecule-1 s-1 for the temperature dependence of the reaction 1 rate coefficient
extrapolation outside the experimental range to low temperatures down to 200 K and to high temperatures up
to 3000 K. A temperature dependence of the rate coefficient of the HOCl + Cl f OH + Cl2 reaction has
been derived on the basis of the experimental data, modeling, and thermochemical information.
I. Introduction
burning has the potential to generate products such as phosgene,
molecular chlorine, polychlorinated dibenzodioxins, polychlo-
rinated dibenzofurans, and polychlorinated biphenyls,1-3 which
are more hazardous than the originally incinerated waste.
Accurate kinetic modeling of these processes is essential for
understanding their mechanisms and for use of these mecha-
nisms as tools of prediction and control.
Reaction 1 plays a significant role in the formation of the
hazardous emissions from combustion of chlorine-containing
compounds1-4 as, for example, it may be important in the control
of the HCl to Cl2 ratio. Molecular chlorine is an effective
chlorinating agent and a powerful oxidant, whereas the larger
bond dissociation energy in HCl can result in significant
sequestration of chlorine. In addition, Cl2 is difficult to scrub
from the flue gas, because of its relatively low water solubility.
Therefore, the emissions of Cl2 should be controlled and
minimized during the combustion process.2 Despite reaction 1
being potentially important in these processes, the rate co-
efficient of this reaction has not been studied at high temper-
atures.
Reaction 1 has been previously studied at low temperatures,5-9
and the results of these studies are in agreement with each other
within reported experimental errors. The importance of this
reaction for atmospheric chemistry is described in the work of
Gilles et al.9 The current work represents the first experimental
study of reaction 1 carried out at high temperatures in the range
from 297 K (to compare with the earlier studies) to 826 K, which
is relevant to combustion and thermal processes.
The hydroxyl radical plays a key role in the chemistry of
combustion and atmospheric processes. Numerical simulation
of the complicated chemistry of these phenomena requires a
large database of accurately determined reaction rate coefficients
in a wide range of temperatures and pressures. This article is
devoted to the experimental and theoretical temperature-
dependent kinetic investigation of the gas-phase reaction of the
hydroxyl radical with molecular chlorine
OH + Cl2 f products
(1)
over a wide temperature range.
Molecular chlorine is produced in large quantities. According
to a conservative estimate reported by the U.S. Environmental
Protection Agency, the annual production of molecular chlorine
is about 107 metric tons just in the United States. The largest
users of molecular chlorine are companies that manufacture
ethylene dichloride and other chlorinated solvents, poly(vinyl
chloride) resins, chloroflurocarbons, and propylene oxide. As
a result of this activity, large quantities of wastes with chlorine-
containing compounds are incinerated, subject to thermal
degradation, or accidental fires. For the incineration of industrial
wastes to be more efficient and less hazardous, it is necessary
to understand all aspects of this process. This understanding is
especially important for chlorine-containing compounds, whose
* To whom correspondence should be addressed. E-mail: BarryD@
LSU.edu.
To accurately determine the rate coefficients of hydroxyl
radical reactions, experiments must be performed with suf-
† Louisiana State University.
‡ The Catholic University of America.
10.1021/jp047340h CCC: $27.50 © 2004 American Chemical Society
Published on Web 10/28/2004

Gas-Phase Reactions of OH
J. Phys. Chem. A, Vol. 108, No. 47, 2004 10465
ficiently low initial concentrations of hydroxyl radicals, thus
ensuring the absence of any complications due to possible fast
secondary reactions. For this purpose, we have constructed a
pulsed laser photolysis/pulsed-laser-induced fluorescence ap-
paratus combined with a heatable, slow-flow reactor. This
technique has excellent sensitivity to hydroxyl radicals (vide
infra), which allows one to perform OH reaction rate measure-
ments with the necessary low initial OH concentrations. Two
reactions whose rate coefficients are relatively well-established
(see refs 10-16 and references therein), reactions 2 and 3, were
tube (Electron Tubes Limited: P10N-01W) using a UV band-
pass filter (308 nm peak transmission) to reduce scattered light.
The signal from the photomultiplier was amplified (Electron
Tubes Limited: transimpedance amplifier A1) and then recorded
by an oscilloscope (Tektronix: TDS 714L digitizing oscil-
loscope). A signal induced by the probe beam in a photodiode,
Ph2, triggered the oscilloscope. The oscilloscope allows one to
obtain the integrated voltage averaged for a desired number of
pulses (typically 150) of the lasers at single time delay. The
value, St, received from the integration of the average voltage
signal is a sum of two components: the integral from the
OH + CH4 f H2O + CH3 (298-1009 K)
OH + C3H8 f products (296-908 K)
(2)
(3)
average laser-induced fluorescence signal of OH radicals, SOH
,
(this term is proportional to the absolute concentration of OH)
and the integral from the average scattered light signal, Ssc. The
average integrated voltage for the scattered light was measured
directly in the absence of OH radicals. For this measurement,
the photolysis laser was not triggered during the accumulation
of the scattered light signal. Kinetic information was obtained
by varying the time delay between pulses of the pump and the
probe lasers in the desired time interval. We used a pulse
generator (Stanford Research System: DG535 digital delay/
pulse generator) to trigger the excimer laser and the Nd:YAG
laser with different time delays.
also studied in order to validate the experimental apparatus used
here for the first time after its construction. (Numbers in
parentheses indicate the experimental temperature ranges of the
current study.)
II. Experimental Section
II. A. Experimental Apparatus. A schematic of the experi-
mental apparatus and the optical test cell (reactor) are depicted
in Figure 1a and b, respectively. This pulsed laser photolysis/
pulsed-laser-induced fluorescence apparatus is similar to the
one described in ref 17. An ArF excimer laser (LAMBDA
PHYSIK: COMPEX 102) and an Nd:YAG (Quanta-Ray: DCR-
2a) pumped, frequency doubled, tunable pulsed dye laser
(Quanta-Ray: PDL-3) were employed to generate the pump and
probe beams, respectively.
Pulsed, unfocused 193-nm radiation (10 Hz) from the excimer
laser was directed through two iris diaphragms, D1 and D2, and
a fluorescence-free UV fused silica window, W1, into a quartz
tubular six-way cross that served as the base of the optical test
cell. The pump (photolysis) beam was responsible for creation
of the hydroxyl radicals. In most experiments, hydroxyl radicals
were formed following the 193-nm laser photolysis of nitrous
oxide producing excited O(1D) atoms followed by rapid reaction
with molecules of water14,17,18
We developed software using Visual Basic 6.0 to obtain the
temporal profiles automatically. Kinetic data (St values with
corresponding reaction time delays t and Ssc value) were
collected in the computer for the subsequent data processing.
Ssc was subtracted from St measured at different reaction time
delays and one fixed concentration of molecular substrate to
obtain a relative OH concentration temporal profile SOH ) St
- Ssc. The systematic uncertainty in the experimentally obtained
SOH values (the deviation from linearity) was defined using
characteristics of the accuracy stated by the manufacturers for
the photomultiplier, amplifier, and oscilloscope. The maximum
of the systematic deviation of the OH concentration detection
system from linearity has been estimated as 3.0%.
Helium was used as the carrier gas in all experiments
presented in this article. We utilized four separately controlled
gas flows to prepare a reaction gas mixture. Three of them were
the main flow of helium, the flow of nitrous oxide diluted with
helium (200-1000 times), and the flow of helium containing
water vapor. Water vapor was added to the third flow of helium
by bubbling helium though water at controlled pressure and
temperature in a thermostabilized saturator maintained at a
temperature ∼2 K below ambient temperature. We employed
three calibrated mass-flow controllers (SIERRA Instruments,
Inc., model 810C) with appropriate flow ranges to establish the
values of the three gas flows and maintain their stability during
the performance of the experiments. At selected times during
the experiments, we controlled values of these three flows by
measuring rates of pressure increases in the calibrated volumes
located upstream from the gas inlet of the optical test cell. These
pressure measurements were performed with capacitance ma-
nometers (MKS Instruments, model 626A13TAE, range 1000
Torr or model 622A12TAD, range 100 Torr). The fourth flow
of undiluted molecular substrate was manually regulated by a
metering valve (Swagelok, type SS-SS4). This flow was also
determined by measuring the rate of the pressure increase in
the calibrated volume before and after obtaining each OH
temporal profile. A mean value of these two flow measurements
was taken to calculate molecular substrate concentration in the
reaction zone. The independence of the measured flows on the
surface-to-volume ratio of the calibrated volume was verified
N2O + hν(193 nm) f N2 + O(1D)
O(1D) + H2O f 2OH
(4)
(5)
When CH4 or C3H8 was present in the reaction mixture, OH
could potentially be produced via reactions of O(1D) with one
of these species that can occur simultaneously with reaction 5.
In selected experiments, we produced hydroxyl radicals by the
193-nm laser photolysis dissociation of H2O only ([N2O] ) 0,
also see ref 11) to avoid possible effects from the thermal
decomposition of N2O at high reaction temperatures and to
verify that the results were independent of the method of OH
generation.
Pulsed, unfocused radiation (10 Hz) from the probe laser was
directed using two flat laser (UV enhanced, aluminum coated)
mirrors, M1 and M2, and passed through two iris diaphragms,
D3 and D4, and a fluorescence-free UV fused silica window,
W2, placed at a Brewster angle into the detection zone of the
optical test cell.
After the pulse of the excimer laser, the pulse of the probe
laser induced excitation of hydroxyl radicals. OH was excited
at approximately 282 nm via the A2Σ+-X2Π (1-0) transition
followed by observation of fluorescence from the (1-1) and
(0-0) bands at 308-316 nm (see, for example, refs 19, 20).
The fluorescent radiation was monitored with a photomultiplier

10466 J. Phys. Chem. A, Vol. 108, No. 47, 2004
Bryukov et al.
to ensure the absence of interference from heterogeneous
absorption and desorption processes on the walls of this volume.
All flows were premixed and directed though the reactor. The
total flow rate ranged between 5.0 and 25 STP cm3 s-1. The
value of the total flow was set depending on total pressure and
temperature in the reaction zone to ensure slow-flow conditions,
such that each photolysis pulse-initiated reaction within a locally
fresh gas mixture, eliminating potential complications that might
result from accumulation of the photolysis or reaction products.
The composite flow conditioned the reactor for several minutes
prior to data collection, thereby minimizing any effects due to
reactant adsorption on the reactor walls and stabilizing the
established experimental conditions. The maximum systematic
uncertainty in the ratio of the measured molecular substrate flow
rate to the total flow rate of the reaction gas mixture has been
evaluated as 1.5% of this ratio value.
Pressure in the reactor was monitored with the capacitance
manometers described earlier. A pressure measurement inlet was
located (see Figure 1b) at the main axis, ∼15 mm downstream
of the photolysis zone in the quartz tubular six-way cross. The
maximum systematic uncertainty in the reaction pressure, P,
was evaluated as 0.7% of this pressure value based on the finite
accuracies stated by the manufacturers for the devices used in
pressure monitoring.
The reactor was resistively heated. A heater, thermoinsulator,
two temperature controllers (series CN77000), and Chromel-
Alumel thermocouples were supplied by OMEGA Engineering,
Inc. We employed two Chromel-Alumel thermocouples to
control the temperature within the heated zone. The first 0.813-
mm diameter thermocouple was fixed at the main axis, ∼5 mm
downstream from the photolysis zone of the quartz tubular six-
way cross (see Figure 1b). The temperature field inside the
heated zone with the desired temperature in the detection zone
(reaction temperature) was stabilized by using this thermocouple
installed with the temperature controller connected to the reactor
heater. The temperature field inside the heated zone was
measured (calibrated) with the second movable 0.508-mm
diameter Chromel-Alumel thermocouple and referenced to the
readout of the first thermocouple. The maximum total uncer-
tainty in the measurements of the reaction temperatures, T, did
not exceed 0.5% of T and was considered as the systematic
source of the experimental errors because of its behavior
revealed during the calibration and experiments.
The molecular concentration of each reactant in the reaction
zone was calculated by means of the multiplication of three
values: the total concentration in the reaction zone derived from
measured pressure and temperature using the ideal gas law, the
molecular partial concentration in the flow carrying the reactant,
and the value of the ratio of the flow carrying the reactant to
the total gas flow rate. Typical reaction mixtures used in these
experiments consisted of the following molecular concentrations:
N2O, 3.6 × 1012-1.5 × 1014 or 0.0; H2O, 4.0 × 1014-2.1 ×
1015; He, 5.06 × 1017-3.99 × 1018; Cl2, 0.0-5.07 × 1015; CH4,
0.0-3.89 × 1016; C3H8, 0.0-3.49 × 1015 in units of molecule
Figure 1. a. Schematic drawing of the experimental apparatus. b. Two-
dimensional schematic drawing of the optical test cell (reactor). The
probe beam is perpendicular to the plane of the figure and intersects
the pump (photolysis) laser beam in the center of the reactor.
II. B. Reaction Rate Measurements and Data Processing.
All experiments to measure the rate coefficients of reactions
1-3 were conducted under pseudo-first-order kinetic conditions
with a large excess of molecular substrate [Cl2], [CH4], [C3H8]
. [OH]0. An estimated initial concentration of OH radicals in
the detection zone was in the range 1.0 × 1010-3.0 × 1011
molecule cm-3, at least two orders of magnitude smaller than
the lowest molecular substrate concentration. We estimated OH
initial concentrations on the basis of a photolysis laser pulse
intensity measured as it entered the reactor, published UV
absorption coefficients of N2O,21 our own experimental evalu-
ation of the absorption cross section of N2O at 193 nm
performed at different temperatures, and on the assumption that
the maximum possible total yield of OH formation was equal
to 2.0 per absorption of one photon by N2O in the presence of
H2O. The initial OH concentrations produced from the direct
cm-3
.
The chemicals utilized in this study had the following stated
minimum purities (and were supplied by): He, 99.999% (The
BOC Group, Inc.); Cl2, 99.999% (Matheson Tri-Gas, Inc.); CH4,
99.999% (Matheson Tri-Gas); C3H8, 99.993% (Matheson Tri-
Gas); N2O, 9.98% mixture of 99.99% purity in 99.999% He
(The BOC Group); H2O, A.C.S. reagent grade (Aldrich).
Analyses of Cl2, CH4, and C3H8 samples showed only trace
levels of impurities, and these small concentrations had a
negligible effect on the observed OH decay rates.

Gas-Phase Reactions of OH
J. Phys. Chem. A, Vol. 108, No. 47, 2004 10467
Figure 3. Examples of experimentally obtained k′ vs [Cl2] depend-
ences.
Figure 2. Examples of relative OH temporal profiles obtained under
the following conditions: helium buffer gas, total pressure P ) 20.10
kPa (150.8 Torr), temperature T ) 365 K, [N2O] ) 3.8 × 1013 cm-3
,
[H2O] ) 1.59 × 1015; [MS] ) [Cl2] ) 0, 5.11 × 1014, 9.09 × 1014,
1.54 × 1015, 1.76 × 1015, 2.38 × 1015, 2.52 × 1015, 3.42 × 1015, 3.75
× 1015 cm-3 for profiles 1-9, respectively.
photolysis of H2O were estimated by comparing the SOH values
obtained in the presence of N2O to those obtained in the absence
of N2O at different concentrations of H2O and N2O and
temperatures. The values of the initial concentrations of OH
radicals used in the experiments to determine the rate coef-
ficients of reactions 1-3 (listed in Tables 1, 2) are somewhat
overestimated and should be understood as upper limits to
[OH]0. Exact knowledge of the initial OH concentrations is not
needed for the determination of rate coefficients, because the
experiments were conducted under pseudo-first-order conditions.
The evaluated detection sensitivity for OH radicals ranged from
1 × 108 to 5 × 108 molecule cm-3 depending on experimental
conditions.
Examples of temporal profiles of the set of experiments are
shown in Figure 2 as a plot of ln(SOH) versus t. The initial
reaction time value, t0, in all OH temporal profiles was not less
than 0.2 ms. This time delay followed the photolysis of the
reaction mixture and was sufficient for the completion of the
reaction 5 and for rotational and vibrational equilibration of OH
radicals to the Boltzmann distribution.18,22
We analyzed each set of experiments at first assuming first-
order kinetic behavior of OH decay
Figure 4. Examples of experimentally obtained k′ vs [CH4] (part a)
and k′ vs [C3H8] (part b) dependences.
ln(SOH) ) constant - k′t
(I)
the OH temporal profile directly measured at zero concentration
of the molecular substrate.
where
k′ ) ki[MS] + k0
(II)
However, under some experimental conditions, the OH
temporal profile detected at [MS] ) 0 and then plotted in [t,
ln(SOH)] coordinates displayed a reproducible positive curvature.
By varying the reaction pressure, temperature, total flow rate,
and diameter of the photolysis beam, we found that the
appearance of this curvature was mainly caused by the process
of the OH diffusion out of the detection zone. All sets of the
experiments with such behavior of OH background loss were
also treated by data processing that is similar to that described
in refs 23-26. This method of data processing allows one to
analyze and decrease the possible effect of the non-first-order
kinetic behavior of the OH background loss on the measured ki
value. During the first stage of data processing, each OH
temporal profile recorded in the presence of the molecular
substrate was point-by-point divided into the common interval
of time delays by the OH temporal profile recorded in the
absence of molecular substrate, in order to obtain OH back-
ground loss-corrected temporal profiles. These background
loss-corrected temporal profiles were then treated by data
is the effective first-order rate coefficient, ki is the bimolecular
rate coefficient of the reaction under study (i ) 1, 2, 3), [MS]
([MS] ) [Cl2], [CH4], [C3H8]) is the molecular substrate
concentration, and k0 is the effective first-order rate coefficient
of OH decay due to possible OH reactions with background
impurities in the buffer gas and N2O, H2O precursors, OH
diffusion, and flow out of the detection zone (OH background
loss). The effective first-order rate coefficient values, k′ values,
were obtained from linear least-squares fits of the experimental
values of ln(SOH) to eq I. (These least-squares fits and all next
least-squares fits were performed with no weighting of the data
points.) Examples of measured k′ versus molecular substrate
concentration dependences are presented in Figures 3 and 4a,b
(for Cl2, CH4, and C3H8 substrates, respectively). We determined
the bimolecular rate coefficient, ki, from the slope of the least-
squares straight line drawn through the k′ versus [MS] data
points including the (0, k0) point, where k0 was derived from

10468 J. Phys. Chem. A, Vol. 108, No. 47, 2004
Bryukov et al.
TABLE 1: Conditions and Results of Experiments to
Measure Rate Coefficients of the Reaction of Hydroxyl
Radicals with Molecular Chlorine
TABLE 2: Conditions and Results of Experiments to
Measure Rate Coefficients of the Reactions of Hydroxyl
Radicals with Methane and Propane
c
[Cl2] range/
Ib/mJ [OH]0/1010
k1 /10-14
molecule-1
cm3 s-1
OH + CH4 f H2O + CH3 (2)
1014 molecule pulse-1 molecule
[CH4] range/ Ib/mJ [OH]0/1010
k2 /10-15
c
no.a T/K P/kPa
cm-3
cm-2
cm-3
1015 molecule pulse-1 molecule molecule-1
1
2
3
4
5
6
7
8
9
297 13.41 4.84-50.7
297 13.40 4.05-48.3
330 6.70 4.58-41.7
330 6.72 4.63-33.1
365 20.10 5.11-37.5
404 13.41 3.05-49.1
404 13.40 3.11-19.1
447 13.41 5.94-20.6
447 13.41 5.93-20.7
13
2.6
12
2.8
11
11
2.6
11
2.6
11
11
12
12
4.3
12
13
2.6
10
2.1
13
13
12
30
6.0
19
4.4
9.6
17
4.0
12
2.8
7.0
7.3
4.3
4.3
5.9
16
19
3.8
16
6.25 ( 0.31
6.38 ( 0.21
9.25 ( 0.67
9.28 ( 0.26
13.72 ( 0.44
19.72 ( 0.33
19.00 ( 0.38
26.32 ( 0.96
25.2 ( 1.3
35.7 ( 1.7
36.9 ( 2.3
49.7 ( 3.0
50.7 ( 4.0
60.3 ( 2.6
59.0 ( 3.6
87.1 ( 4.0
87.7 ( 5.0
125.8 ( 2.5
124.7 ( 7.4
131.0 ( 7.9
121.5 ( 4.1
131.3 ( 6.9
no.a T/K P/kPa
cm-3
cm-2
cm-3
cm3 s-1
1
2
3
4
5
6
7
8
9
298 13.46 3.41-20.5
298 13.43 3.51-20.7
299 13.43 5.04-29.5
299 13.43 5.11-29.9
341 6.71 5.86-18.7
341 6.71 6.05-19.1
408 13.43 3.02-38.9
408 13.43 3.00-38.1
507 13.43 2.55-21.6
507 13.43 2.56-21.7
11
3.6
11
4.1
10
2.6
10
4.3
12
3.3
13
13
3.3
14
6.1
2.0
27
6.48 ( 0.19
6.29 ( 0.34
6.75 ( 0.12
6.78 ( 0.14
15.19 ( 0.57
15.22 ( 0.62
36.7 ( 1.2
37.1 ( 1.3
108.5 ( 3.7
106.9 ( 3.6
243.7 ( 8.7
482 ( 12
481.9 ( 7.5
803 ( 19
1246 ( 42
1232 ( 43
1421 ( 43
1687 ( 72
1670 ( 62
10
8.5
2.2
7.7
3.3
7.3
2.0
2.5
11
2.9
3.2
16
3.8
6.3
12
10 501 13.41 3.71-43.5
11 501 13.42 3.21-20.2
12 562 6.70 3.76-16.5
13 562 6.70 3.79-16.3
14 620 13.40 2.78-17.6
15 620 13.39 3.18-16.8
16 719 13.42 1.44-10.1
17 719 13.42 1.43-10.7
10
11* 606 24.15 2.05-12.5
12
13
706 13.41 0.422-6.47
706 13.41 0.427-6.41
14* 807 13.43 0.121-4.00
15* 908 13.41 0.271-2.70
16* 908 13.41 0.271-2.70
11
2.6
18 810 10.74
2.4-11.2
17* 961 6.71 0.333-1.260 13
19 810 10.74 2.39-11.1
20* 825 10.73 1.19-5.68
21* 825 21.46 3.79-10.9
22* 826 13.39 0.71-6.3
3.4
3.8
7.6
4.4
18* 1009 21.46 0.230-1.33
19* 1009 21.46 0.236-1.33
11
3.6
3.8
OH + C3H8 f products (3)
a Experiment number. [N2O] ) 0 was in experiments marked with
*. b Photolysis laser intensity. c Error limits represent 2σ statistical
uncertainties only. Maximum estimated systematic uncertainty is 5%
of the rate coefficient value (see text).
[C3H8] range/ Ib/mJ [OH]0/1010
k3 /10-12
c
1014 molecule pulse-1 molecule
molecule-1
cm3 s-1
no.a T/K P/kPa
cm-3
cm-2
cm-3
1* 296 12.07
3.78-20.3
1.73-32.7
3.28-33.5
3.25-34.9
1.86-18.0
1.34-11.9
1.17-4.69
8.6
13
13
4.1
12
8.9
4.8
13
14
4.8
13
12
2.9
1.0
3.7
6.7
2.1
2.1
2.7
2.4
3.0
7.7
2.6
2.7
15
1.083 ( 0.055
1.081 ( 0.024
1.487 ( 0.034
1.453 ( 0.044
2.156 ( 0.092
3.37 ( 0.15
4.91 ( 0.22
6.65 ( 0.20
7.41 ( 0.30
7.37 ( 0.38
8.37 ( 0.20
10.23 ( 0.31
10.36 ( 0.19
2
3
4
5
6
7
8
9
296 12.01
339 12.09
339 12.09
407 6.71
495 12.05
605 6.71
processing described in the previous paragraph. The linear
least-squares fit method with two floated parameters was ap-
plied in both stages to obtain k′) ki[MS] (in this case) and
ki values. The point with (0, 0) coordinates was always in-
cluded in the data points set of the third stage. For the same set
of experiments, the application of data processing with the
background loss correction stage always yielded a lower value
for ki than the value obtained by applying the data pro-
cessing without the background loss correction stage; how-
ever, the difference between these two values never exceeded
4% of their average value. Only in those cases where this
difference was 1.0% or more of the ki value, the ki value obtained
by the data processing with the background loss correction stage
was taken as the result of its measurement for the subsequent
consideration, instead of the ki value obtained assuming first-
order kinetic behavior of OH decay. Therefore, a systematic
uncertainty in the experimentally obtained rate coefficients,
caused by unaccounted-for deviation of OH background loss
from the first-order kinetic behavior, was evaluated as 1% of ki
and taken into account in the evaluation of the systematic error
of the measured bimolecular rate coefficients (vide infra).
II. C. Experimental Results. The kinetic study of the
reactions of OH with Cl2, CH4, and C3H8 was performed over
the temperature ranges of 297-826, 298-1009, and 296-908
K and at pressures between 6.7 and 21.46, 6.71 and 24.15, and
6.68 and 24.13 kPa, respectively. The initial concentration of
OH radicals was varied by changing the photolysis laser
intensity, the concentration of N2O, and the concentration of
H2O. Conditions and results of experiments to determine the
values of the rate coefficients of reactions 1-3 are presented
in Tables 1 and 2. The rate coefficients demonstrate no
dependence on pressure or initial concentration of OH radicals
within the experimental ranges. The observed pressure inde-
pendence was anticipated, because the mechanisms of reactions
1-3 were expected to be those of atom abstraction (see next
705 12.05 0.460-2.88
765 6.68 0.453-4.39
10 765 6.68 0.482-3.83
11 805 24.13 0.498-3.65
12* 908 12.05 0.429-4.00
13* 908 12.05 0.408-4.12
3.6
a Experiment number. [N2O] ) 0 was in experiments marked with
*. b Photolysis laser intensity. c Error limits represent 2σ statistical
uncertainties only. Maximum estimated systematic uncertainty is 5%
of the rate coefficient value (see text).
section). The fact that the rate coefficients are independent of
the initial OH radical concentration indicates an absence of any
influence of potential secondary reactions on the kinetics of OH
radicals, as can be expected by the low values of [OH]0 (1.0 ×
1010-3.0 × 1011 molecule cm-3) estimated as upper limits.
Similarly, the observed absence of any correlation between the
measured rate coefficients and the photolysis laser intensity
indicates that the potential effects of reactions between OH and
the products of the laser photolysis of molecular substrates are
negligible. At the highest temperatures used in the study of
reaction 1, the absence of any potential effects of thermal
decomposition of molecular chlorine was verified by measuring
rate coefficients at different pressures and bulk flow velocities
varied by a factor of approximately 2.
Temperature dependence values of the bimolecular rate
coefficients for reaction 1-3 are displayed in Figures 5 and 6
in Arrhenius coordinates. The rate coefficients exhibit positive
temperature dependence that can be represented with the
following modified three-parameter Arrhenius expressions

Gas-Phase Reactions of OH
J. Phys. Chem. A, Vol. 108, No. 47, 2004 10469
substrate, ratios of the calibrated volumes, reaction pressure,
temperature, time, and St and Ssc values, were subdivided into
statistical and systematic categories according to their physical
nature or the behavior observed during the performance of the
experiments. The evaluation of potential systematic errors was
mainly based on the finite accuracy of the equipment. The
uncertainties in the measured experimental parameters were
propagated to the final values of the uncertainties in the
measured rate coefficients, using different mathematical pro-
cedures for propagating systematic and statistical uncertainties.27
The error limits of the experimentally obtained rate coefficients
listed in Tables 1 and 2 represent the 2σ level of statistical
(random) uncertainty only. The evaluated maximum of the
systematic uncertainty in the experimental rate coefficients is
5% of the value.
Figure 5. Temperature dependence of the rate coefficient for reaction
1 displayed in Arrhenius coordinates. Closed circles (B) represent the
current experimental data; open square (0) represents ref 5; open
triangle up (4) represents ref 6; open diamond (]) represents ref 7;
open hexagons (") represent ref 8; open circles (O) represent ref 9.
The solid line (-) represents the modified Arrhenius fit of eq III; the
dashed line (- - -) represents eq VI (see text).
III. Potential Energy Surface and Transition-State
Theory Model of Reaction 1
III. A. Potential Energy Surface Study. The potential
energy surface (PES) of reaction 1, that of OH with Cl2, was
studied using quantum chemical approaches. Two methods were
used for the optimization of molecular structures and calculation
of vibrational frequencies: density functional BH&HLYP28,29
and QCISD,30 both with the aug-cc-pVDZ basis set.31 The
version of the BH&HLYP functional implemented in Gaussian
9832 was used which, as described in the Gaussian manual, is
different from that of ref 28. The sets of vibrational frequencies
obtained using BH&HLYP and QCISD methods are similar to
each other (Table 1S in the Supporting Information). The
Gaussian 9832 program was used in all PES calculations. In
addition, high-level single-point energy calculations were carried
out for the PES stationary points using the QCISD(T)30 method
with two basis sets, aug-cc-pVDZ and aug-cc-pVTZ.31 All
QCISD(T) results lead to the same qualitative conclusions
regarding the mechanism of reaction 1; energy values quoted
in the text henceforth are those obtained in the QCISD(T)/aug-
cc-pVTZ//QCISD/aug-cc-pDTZ calculations and include vibra-
tional zero-point energies unless stated otherwise. The results
of the PES study are summarized in Table 3 and Figure 7, and
the detailed information is given in the Supporting Information
(Table 1S).
The reaction path leading from the OH + Cl2 reactants to
the HOCl + Cl products has an energy barrier (transition
state TS1) and two shallow van der Waals minima on both
the reactants (vdW1) and the products (vdW2) sides of the
barrier (-6.5 relative to the reactants and -10.1 kJ mol-1
relative to the products). The energy barrier height was
overestimated at both the BH&HLYP (39 kJ mol-1) and the
QCISD (29 kJ mol-1) levels, as can be seen by comparing
these values with the 11.6 kJ mol-1 experimental activation
energy obtained if the experimental k1(T) dependence is fit
with a two-parameter Arrhenius expression. The use of the
higher-level QCISD(T) single-point calculations reduced
the barrier height to more realistic values in the 11.1-16.6 kJ
mol-1 range (see Table 3 for individual values corresponding
to each combination of basis set and geometry optimization
method).
Figure 6. Temperature dependences of the rate coefficients for
reactions 2 and 3 displayed in Arrhenius coordinates. Reaction 2:
Closed squares (9) represent the current experimental data; open
triangles up (4) represent ref 10; open squares (0) represent ref 11;
open triangles down (3) represent ref 12; open hexagons (") represent
ref 13; the solid line (-) represents the result of the fit of the current
experimental data only (eq IV). Reaction 3: closed circles (B) represent
the current experimental data; open circles (O) represent ref 14; closed
triangles up (2) represent ref 15; open diamonds (]) represent ref 16;
the dashed line (- - -) represents the result of the fit of the current
experimental data only (eq V).
within their corresponding experimental temperature ranges
k1 ) 3.59 × 10-16 T1.35 exp(-745 K/T) cm3 molecule-1 s-1
(297-826 K) (III)
k2 ) 3.82 × 10-19 T2.38 exp(-1136 K/T) cm3 molecule-1 s-1
(298-1009 K) (IV)
k3 ) 6.64 × 10-16 T1.46 exp(-271 K/T) cm3 molecule-1 s-1
(296-908 K) (V)
Maximum deviations of our experimental rate coefficient values
from the rate coefficient values calculated using these param-
etrized expressions are 7.0% (expression III), 4.8% (expression
IV), and 2.5% (expression V). Error limits of the parameters in
expressions III-V are not reported, because these parameters
bear no physical meaning.
An attempt was made to locate a transition state (PES saddle
point) for the reaction channel leading from OH + Cl2 to the
HCl + ClO products. In these calculations, a two-dimensional
PES scan (Cl-Cl and O-H, starting from the HOCl + Cl
product valley) at the BH&HLYP/aug-cc-pVDZ level and three-
The sources of error in the measured experimental physical
parameters, such as total flow rate, flow rate of the molecular
dimensional scan (Cl-Cl, O-H, and O-Cl) at the B3LYP29,33
aug-cc-pVDZ level were performed. The searched-for transition
/

10470 J. Phys. Chem. A, Vol. 108, No. 47, 2004
Bryukov et al.
TABLE 3: Energies of Reactants, Products, and Stationary Points on the PES of Reaction 1 Obtained in Quantum Chemical
Calculationsa
speciesb
vdW2
BH&HLYP/aug-cc-pVDZ optimization
method
HOCl + Cl
HCl + ClO
vdW1
vdW3
IP1
TS1
TS2
BH&HLYP/aug-cc-pVDZ
QCISD(T)/aug-cc-pVDZ
QCISD(T)/aug-cc-pVTZ
15.11
-8.17
4.45
-8.04
-30.86
-26.50
-5.63
-7.30
-6.18
5.80
-16.65
-6.70
-14.11
-33.87
-29.58
10.39
-15.68
-3.19
39.13
11.12
13.91
49.62
26.34
44.55
QCISD/aug-cc-pVDZ optimization
QCISD/aug-cc-pVDZ
QCISD(T)/aug-cc-pVDZ
QCISD(T)/aug-cc-pVTZ
experimentalc
-6.87
-9.52
3.51
-27.62
-6.44
-7.50
-6.47
-14.41
-18.80
-6.61
-37.12
-41.13
-34.48
-13.44
-17.08
-4.20
29.22
13.85
16.55
9.63
34.53
30.69
31.67
-30.05
-24.40
9.7 ( 3.3
-29.2 ( 3.5
a Energy values are given in kJ mol-1 relative to OH + Cl2 and include zero-point vibrational energy (ZPE). b See text for the description of the
individual PES stationary points. c Ref 35 for products (∆H0°) and the result of transition-state theory fit of experimental k1(T) dependence for TS1.
potential formation of the HCl + ClO products can be neglected
in the OH + Cl2 reaction.
III. B. Rate Coefficient Calculations. To provide means for
an extrapolation of the experimental k1(T) dependence to
temperatures outside the experimental range, a transition-state
theory model of reaction 1 was created in the current work. In
the initial approximation to the properties of the reaction
transition state, geometry, and vibrational frequencies obtained
at the QCISD/aug-cc-pVDZ level, calculations were used
together with the experimental properties of the reactants and
products.35 Rate coefficient values were calculated using the
classical transition-state theory formula (see, for example, ref
36).
The reaction transition state has one torsional degree of
freedom. The PES of this torsion, obtained using a relaxed scan
Figure 7. Potential energy surface of reaction 1 obtained in quantum
of the H-O-Cl-Cl dihedral angle has two unequal maxima
chemical calculations (see text). QCISD(T)/aug-cc-pVTZ//QCISD/aug-
(7.5 and 28.1 kJ mol-1). In the rate constant calculations, the
cc-pDTZ energies are used for the plot (solid horizontal lines, -).
partition function of this torsional degree of freedom was
calculated by summing over the energy levels evaluated via a
numerical solution of the Schro¨dinger equation. The FGH1D
program37 of Johnson was used in the energy level calculations.
Existence of the low-lying excited electronic states35 of OH
(139.7 cm-1) and Cl (882.5 cm-1) was taken into account in
the rate coefficient calculations.
Dotted lines (‚‚‚) indicate the experimental values of ∆H0° for HOCl
+ Cl and HCl + ClO (ref 35) and the energy barrier height for
TS1 obtained in transition-state theory modeling of the experimental
data.
state could not be found. The three-dimensional scan produced
a PES ridge without a saddle point separating the OH + Cl2
and the HCl + ClO valleys.
However, a reaction path leading from the van der Waals
complex corresponding to the HOCl + Cl products (vdW2) to
HCl + ClO was found using both the BH&HLYP and QCISD
geometry optimizations. This path leads from vdW2 through
the structure denoted as IP1, a transition state TS2, and a van
der Waals complex vdW3 corresponding to the HCl + ClO
products. The IP1 structure (Table 1S in the Supporting
Information) has energy that is 7.7 kJ mol-1 below that of HOCl
+ Cl and resembles another shallow-well van der Waals
complex. The molecular structure optimization procedure
implemented in Gaussian 9832 reported IP1 as a PES minimum;
however, a BH&HLYP-level scan of the Cl-Cl distance leading
from IP1 to vdW2 indicated that the path between these two
structures is barrierless. Therefore, it is likely that IP1 represents
an inflection point on the reaction PES, not a local minimum.
The TS1 PES saddle point represents the dynamic bottleneck
of the OH + Cl2 f HOCl + Cl reaction. Any further chemical
transformation of the exit van der Waals complex (vdW2) can
be expected to largely favor dissociation into HOCl + Cl via a
very loose34 transition state with densities of states that are much
larger than those of the tighter PES saddle points TS1 and TS2.
In addition, the energy of the transition state for the pathway
leading to the HCl + ClO products (TS2) is higher than that of
the HOCl + Cl products by 28.2 kJ mol-1, indicating that the
The energy barrier of reaction 1 was adjusted to provide a
match with the experimental k1(T) dependence. A combined set
of data formed by the results of the current study and those of
Gilles et al.9 (included to represent the low-temperature k1(T)
dependence) was used in the fitting. The resultant model
reproduces the experimental k1(T) dependence very well, with
the calculated room-temperature preexponential factor differing
from the experimental one by a factor of 1.4. An adjustment of
the lowest vibrational frequency of the transition state (from
171 to 242 cm-1) brought the calculated temperature dependence
into agreement with the experimental one. The final adjusted
value of the reaction energy barrier is 9.6 kJ mol-1
.
The rate coefficient values extrapolated via modeling can be
represented by the modified Arrhenius expression
k1 ) 1.35 × 10-16 T1.50 exp(-723 K/T) cm3 molecule-1 s-1
(200-3000 K) (VI)
with deviations from the calculated values of less than 4%
between 200 and 3000 K.
Information on reaction 1 obtained in the current work can
be used in combination with the known thermochemistry to
derive the temperature dependences of the rate coefficient of

Gas-Phase Reactions of OH
J. Phys. Chem. A, Vol. 108, No. 47, 2004 10471
the reverse process, that of the reactions of abstraction of a
chlorine atom from HOCl by Cl:
means for the extrapolation of the k1(T) temperature dependence
to outside the experimental temperature range. The PES of
reaction 1 obtained in quantum chemical calculations demon-
strates good agreement with experimental results. The vibrational
frequencies of the transition state (TS1) obtained at both the
BH&HLYP and QCISD levels yield realistic preexponential
factors; only a minor adjustment (see section III. B.) was needed
to bring the model to complete agreement with the experimental
k1(T) dependence. The QCISD(T)/aug-cc-pVTZ level energy
calculations also provide reasonable agreement with the ex-
perimental values of the reaction enthalpy and barrier height,
with deviations of 5-6 and 4-7 kJ mol-1, respectively (see
Table 3). Theoretical model predictions suggest OH + Cl2 f
HOCl + Cl as the main channel for reaction 1, in agreement
with the room-temperature evaluation of the branching ratio for
the OH reaction with Cl2 presented in the work of Loewenstein
et al.7 The agreement between experimental and computational
results provides support for further use of the computational
techniques applied here for the treatment of other reactions of
similar types.
Cl + HOCl f OH + Cl2 (-1)
The transition-state theory model of reaction 1 created in the
current study results in the k-1(T) dependence that can be
represented by the expression
k-1 ) 2.76 × 10-16 T1.39 exp(177 K/T) cm3 molecule-1 s-1
(200-3000 K) (VII)
with deviations from the calculated values of less than 7%.
Here, thermochemical properties from ref 35 were used. The
uncertainty in expression VII originates, primarily, from that
in the enthalpy of reaction 1. The uncertainty factor can be
calculated for any temperature using the van’t Hoff factor with
the cumulative reaction enthalpy uncertainty of 3.3 kJ mol-1
.
The bimolecular rate coefficient values for reaction -1 calcu-
lated from the temperature dependence of expression VII are
consistent with the recommendation given in ref 38 for Cl +
HOCl f products within reported uncertainties in a common
temperature interval.
Acknowledgment. This research was supported by the
Patrick F. Taylor Chair Foundation. The authors would like to
thank Dr. V. L. Orkin for helpful discussion and advice and
Mr. Christian Boussert for manufacturing the quartz and Pyrex
parts of the equipment.
IV. Discussion
Reactions 2 and 3 have been studied in the present work to
validate the experimental apparatus employed for the first time
after its construction. Hydrogen atom abstraction from hydro-
carbons by hydroxyl radicals plays a fundamental role in the
chemistry of atmospheric and combustion processes; for this
reason, reactions 2 and 3 had been extensively studied previ-
ously. These earlier results are in general agreement with each
other. Reviews of these data can be found in ref 13 for reaction
2 and in ref 16 for reaction 3, and they are not repeated here.
The rate coefficients obtained in the current study are in
agreement with these earlier measurements (see Figure 6; data
are exemplified by refs 10-13 and 14-16 for reactions 2 and
3, respectively). For example, over common temperature ranges,
the maximum deviations between the rate coefficient values
calculated using modified three-parameter Arrhenius expressions
IV and V and those calculated using the expressions recom-
mended in refs 11 and 12 for reaction 2 and in refs 14 and 15
for reaction 3 are the following: 5.4% and 2.5% for reaction 2
and 5.1% and 1.3% (of the average rate coefficient value) for
reaction 3. Such a good agreement with previously obtained
data gives us confidence in the measurements performed with
the experimental apparatus employed in this work.
All previous studies of the reaction of OH + Cl2 have been
performed at low temperatures (see Figure 5). Data reported in
refs 5-7 were measured at room temperature only, and they
are in agreement with our data obtained at room temperature
within experimental errors. Experimental temperature-dependent
kinetic investigations of reaction 1 were carried out in two
works.8,9 Our data are in agreement with both of these studies
over the common temperature ranges, within reported experi-
mental errors. Nevertheless, only data from ref 9 were included
to represent the low-temperature k1(T) dependence for the
purposes of rate coefficient modeling. This preference was given
to the results of ref 9 because of the wider temperature range
used, better purity of molecular chlorine sample, and the
experimental conditions free from potential contributions of the
OH heterogeneous reactions.
Supporting Information Available: Supplement including
the results of the quantum chemical calculations (Table 1S).
This material is available free of charge via the Internet at http://
pubs.acs.org.
References and Notes
(1) Chang, W.-D.; Senkan, S. M. EnViron. Sci. Technol. 1989, 23, 442.
(2) Procaccini, C.; Bozzelli, J. W.; Longwell, J. P.; Smith, K. A.;
Sarofim, A. F. EnViron. Sci. Technol. 2000, 34, 4565.
(3) Procaccini, C.; Bozzelli, J. W.; Longwell, J. P.; Sarofim, A. F.;
Smith, K. A. EnViron. Sci. Technol. 2003, 37, 1684.
(4) Aizava, T.: Kamimoto, T.; Tamaru, T. Appl. Opt. 1999, 38, 1733.
(5) Leu, M. T.; Lin, C. L. Geophys. Res. Lett. 1979, 6, 425.
(6) Ravishankara, A. R.; Eisele, F. L.; Wine, P. H. J. Chem. Phys.
1983, 78, 1140.
(7) Loewenstein, L. M.; Anderson, J. G. J. Phys. Chem. 1984, 88, 6277.
(8) Boodaghians, R. B.; Hall, I. W.; Wayne, R. P. J. Chem. Soc.,
Faraday Trans. 2 1987, 83, 529.
(9) Gilles, M. K.; Burkholder, J. B.; Ravishankara, A. R. Int. J. Chem.
Kinet. 1999, 31, 417.
(10) Vaghjiani, G. L.; Ravishankara, A. R. Nature (London) 1991, 350,
406.
(11) Dunlop, J. R.; Tully, F. P. J. Phys. Chem. 1993, 97, 11148.
(12) Gierczak, T.; Talukdar, R. K.; Herndon, S. C.; Vaghjiani, G. L.;
Ravishankara, A. R. J. Phys. Chem. A 1997, 101, 3125.
(13) Bonard, A.; Daele, V.; Delfau, J.-L.; Vovelle, C. J. Phys. Chem. A
2002, 106, 4384.
(14) Droege, A. T.; Tully, F. P. J. Phys. Chem. 1986, 90, 1949.
(15) Talukdar, R. K.; Mellouki, A.; Gierczak, T.; Barone, S.; Chiang,
S.-Y.; Ravishankara, A. R. Int. J. Chem. Kinet. 1994, 26, 973.
(16) Kozlov, S. N.; Orkin, V. L.; Huie, R. E.; Kurylo, M. J. J. Phys.
Chem. A 2003, 107, 1333.
(17) Taylor, P. H.; D’Angelo, J. A.; Martin, M. C.; Kasner, J. H.;
Dellinger, B. Int. J. Chem. Kinet. 1989, 21, 829.
(18) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F., Jr.; Kerr,
J. A.; Rossi, M. J.; Troe, J. J. Phys. Chem. Ref. Data 1997, 26, 1329.
(19) Wollenhaupt, M.; Carl, S. A.; Horowitz, A.; Crowley, J. N. J. Phys.
Chem. A 2000, 104, 2695.
(20) D’Ottone, L.; Campuzano-Jost, P.; Bauer, D.; Hynes, A. J. J. Phys.
Chem. A 2001, 105, 10538.
(21) Okabe, H. Photochemistry of Small Molecules; Wiley: New York,
1978.
(22) Silvente, E.; Richter, A. C.; Hynes, A. J. J. Chem. Soc., Faraday
Trans. 1997, 93, 2821.
(23) Tully, F. P.; Golgsmith, J. E. M. Chem. Phys. Lett. 1985, 116, 345.
The theoretical simulation of reaction 1 was performed in
this work to assess possible reaction pathways and to provide

10472 J. Phys. Chem. A, Vol. 108, No. 47, 2004
Bryukov et al.
(24) Tully, F. P.; Droege, A. T.; Koszykowski, M. L.; Melius, C. F. J.
Chem. Phys. 1986, 90, 691.
(25) Orkin, V. L.; Huie, R. E.; Kurylo, M. J. J. Phys. Chem. 1996, 100,
8907.
(26) Kurylo, M. J.; Orkin, V. L. Chem. ReV. 2003, 103, 5049.
(27) Bevington, P. R. Data Reduction and Error Analysis for the
Physical Sciences; McGraw-Hill: New York, 1969.
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.;
Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi,
I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M.
W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,
M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision A.7; Gaussian,
Inc.: Pittsburgh, PA, 1998.
(28) Becke, A. D. J. Chem. Phys. 1993, 98, 1372.
(33) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(34) Benson, S. W. Thermochemical Kinetics, 2nd ed.; John Wiley and
Sons: New York, 1976.
(35) Chase, M. W., Jr. J. Phys. Chem. Ref. Data 1998, Monograph 9.
(36) Johnston, H. S. Gas-Phase Reaction Rate Theory; The Ronald
Press: New York, 1966.
(29) Lee, C. T.; Yang, W. T.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
(30) Pople, J. A.; Head-Gordon, M.; Raghavachari, K. J. Chem. Phys.
1987, 87, 5968.
(31) Kendall, R. A.; Dunning, T. H., Jr.; Harrison, R. J. J. Chem. Phys.
1992, 96, 6796.
(32) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick,
(37) Johnson, R. D. I. FGH1D, a PC/Windows program for computing
vibrational levels for nonharmonic potentials; National Institute of Standards
and Technology: Gaithersburg, MD 20899.
(38) 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. JPL Publ. 1997, 4, 29.