Vibrational relaxation of OH by oxygen atoms

Article abstract of DOI:10.1016/j.cplett.2005.08.131

The collisional removal of OH(v = 1) by O(³P) atoms is investigated. OH is generated by 193 nm photolysis of H₂O ₂, and O(³P) atoms are generated by a microwave discharge in O₂ diluted in Ar. OH(v = 0 and 1) concentrations are monitored by laser-induced fluorescence vs. the time after the photolysis laser pulse. From comparison of these concentrations with kinetic simulations, the room-temperature total removal rate constant for OH(v = 1) in collisions with O(³P) is determined to be (3.9 ± 0.6) × 10 ^-11 cm³ molecule^-1 s^-1. This value is slightly larger than the OH(v = 0)-O(³P) reaction rate constant, but the difference is within the experimental uncertainty.

Full text of DOI:10.1016/j.cplett.2005.08.131

Chemical Physics Letters 415 (2005) 1–5
www.elsevier.com/locate/cplett
Vibrational relaxation of OH by oxygen atoms
Ani Khachatrian, Paul J. Dagdigian *
Department of Chemistry, Zanvyl Krieger School of Arts and Sciences, The Johns Hopkins University, 3400 North Charles Street (Remsen B42),
Baltimore, MD 21218-2685, USA
Received 27 May 2005; in final form 22 August 2005
Available online 19 September 2005
Abstract
The collisional removal of OH(v = 1) by O( P) atoms is investigated. OH is generated by 193 nm photolysis of H
3
3
2
2
O , and O( P) atoms
are generated by a microwave discharge in O diluted in Ar. OH(v = 0 and 1) concentrations are monitored by laser-induced fluorescence
2
vs. the time after the photolysis laser pulse. From comparison of these concentrations with kinetic simulations, the room-temperature
3
ꢀ11
3
ꢀ1 ꢀ1
total removal rate constant for OH(v = 1) in collisions with O( P) is determined to be (3.9 ± 0.6) · 10
cm molecule s . This value
3
is slightly larger than the OH(v = 0)–O( P) reaction rate constant, but the difference is within the experimental uncertainty.
ꢀ
2005 Elsevier B.V. All rights reserved.
1
. Introduction
studied; the IUPAC subcommittee for gas phase data
evaluation [12] and the NASA panel [13] recommend
ꢀ11
3
The OH radical is an important species in the earthÕs
kv = 0-(OH–O) = (3.5 ± 0.4) and (3.3 ± 0.7) · 10
cm
ꢀ
1
ꢀ1
atmosphere and other environments. An understanding
of its collisional vibrational relaxation rates is important
in modeling OH chemistry in the upper atmosphere [1,2].
Collisional vibrational relaxation of OH in a wide range
of vibrational levels by stable collision partners has been
investigated [3–10]. Investigation of collisional relaxation
of vibrationally excited OH by oxygen atoms, present at
significant concentration in the upper atmosphere, is also
relevant. This energy transfer process is interesting from
a fundamental point of view [11] since it involves two
open-shell species and occurs through formation and decay
molecule
s
at 298 K, respectively.
3
Collisional removal of OH(v P 1) by O( P) can occur by
both chemical reaction and vibrational relaxation. In EPR
studies, Spenser and Glass [14] report values of
ꢀ
10
3
ꢀ1 ꢀ1
(1.45 ± 0.25) and (1.05 ± 0.53) · 10
cm molecule
s
3
for OH(v = 1)–O( P) vibrational relaxation and chemical
reaction rate constants, respectively. Marschall et al. [15]
3
have investigated collisional removal of OH(v) by O( P)
through laser-induced fluorescence measurements of time-
dependent OH(v) concentrations after 248 nm photolysis
1
of ozone and subsequent reaction of O( D) with water to
of a transient HO complex (see Fig. 1).
OH(v = 0) can be removed in collisions with oxygen
atoms through the reaction
form vibrationally excited OH. They reported a prelimin-
2
ꢀ11
3
ꢀ1 ꢀ1
ary value kv = 2(OH–O) = 4.6 · 10
cm molecule
s
(quoted in [16]) for the room-temperature OH(v = 2)–
3
O( P) total removal rate constant. Varandas [17] has
2
3
OHðX PÞ þ Oð PÞ ! H þ O
2
computed rate constants for reactive and non-reactive
ꢁ
DH ð0 KÞ ¼ ꢀ16:29 ꢂ 0:09 kcal=mol.
ð1Þ
3
OH(v)–O( P) collisions at several temperatures through
trajectory calculations on the lowest HO potential energy
2
Because of its importance in atmospheric and combustion
chemistry, the rate of reaction 1 has been extensively
surface (PES).
We present here an investigation of the collisional re-
3
moval of OH(v = 1) by O( P). Oxygen atoms are prepared
*
in a flow system by a microwave discharge in O diluted in
Corresponding author. Fax: +1 410 516 8420.
2
E-mail address: pjdagdigian@jhu.edu (P.J. Dagdigian).
argon. Hydroxyl is generated by 193 nm photolysis of
0
009-2614/$ - see front matter ꢀ 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2005.08.131

2
A. Khachatrian, P.J. Dagdigian / Chemical Physics Letters 415 (2005) 1–5
2
0
0
vapor above a 95% solution is ꢃ90% H O , with the
2
2
OH(v=1) + O
OH(v=0) + O
remainder H O [22]. Based on the 193 nm H O absorption
2 2 2
cross section and quantum yields for the OH + OH and
H + O (a)
2
H + HO dissociation channels [12], we estimate 0.8% dis-
2
sociation of H O and an initial OH concentration of
2
2
1
1
ꢀ3
7
· 10 molecules cm . H O was deemed to be a more
suitable precursor than the commonly employed HNO
2
2
3
-
-
-
20
40
60
H + O (X)
2
1
precursor since the latter also yields at 193 nm O( D)
23,24], which reacts with hydrogen-containing compounds
to form OH(v), causing a cascade contribution to
OH(v = 1)]. Photolysis at 248 nm would provide a cleaner
source of OH, but the absorption cross sections are too low
13] to allow detectable photolytic production of OH(v = 1)
[
˜
2
HO (A A’)
2
[
[
2
˜
HO (X A”)
2
at the required low partial pressures of the precursor.
OH(v = 0 and 1) were detected vs. time after photolysis
by pulsed laser fluorescence excitation on the A–X (1,0)
Q (1) and (2,1) Q (2) lines and detection in the (1,1) and
2
3
Fig. 1. Schematic energy diagram of the OH(X P) + O( P) ! H + O
2
reaction.
1
1
(2,2) bands, respectively. The frequency-doubled output
of an optical parametric oscillator (Continuum Sunlite
EX) was employed for excitation. The probe beam counter-
propagated along the photolysis laser beam and was intro-
duced through the opposite sidearm of the apparatus. The
delay between the photolysis and probe lasers was set by a
digital delay generator which was controlled by the data
acquisition computer. A run consisted of recording the
fluorescence intensity (10-shot average) while cycling
through a grid of delays 10 times [20 times for OH(v = 1)
with the microwave discharge on].
H O added to the flow, and OH(v = 0 and 1) concentra-
2
2
tions are followed by laser-induced fluorescence. The mea-
sured time-dependent OH(v) concentrations are compared
with full kinetic modeling of species concentrations in or-
3
der to extract the OH(v = 1)–O( P) total removal rate con-
stant kv = 1(OH–O). This method of studying vibrational
relaxation of OH is challenging because of the very small
(
ꢃ1%) v = 1 to v = 0 population ratio of OH formed in
the photolysis of H O [18,19] but avoids problems in mod-
2
2
eling the population of OH(v = 1) from collisional cascade
from higher OH vibrational levels.
3. Results
2
. Experimental
The total removal rate constant kv = 1(OH–O) was deter-
These experiments were performed in a modified appa-
ratus previously employed for collisional energy transfer
studies [20,21]. Oxygen atoms were generated by passing
mined by comparison of the temporal decays of OH(v = 0
and 1) concentrations with the results of kinetic modeling,
as described below. Figs. 2 and 3 present measured
OH(v = 0 and 1) concentration profiles, as detected by
laser-induced fluorescence on given rotational transitions,
with the microwave discharge off and on, respectively.
The OH(v = 0) concentration with the discharge off is seen
to decay very slowly, mainly from diffusion out of the laser
excitation/detection zone. By contrast, [OH(v = 0)] decays
more rapidly with the discharge on, due to reaction with
a 5% mixture of O in Ar at a total pressure of 2.2 Torr
2
through a 2450 MHz microwave discharge (80 W). The
gas mixture flowed (120 sccm total, measured with cali-
brated flow meters) at an estimated velocity of 150 cm/s
through a 3.1 · 6.3 cm rectangular section mounted on a
1
m spectrometer, employed with wide slits (4 nm spectral
resolution) to isolate fluorescence in a given vibrational
band. Surfaces downstream of the discharge were coated
with fluorocarbon wax to inhibit wall recombination. Pres-
sures were measured with a capacitance monometer
3
O( P) atoms. This decay rate was used with a literature
3
value [12] of kv = 0(OH–O) to estimate the O( P) concentra-
tion, as described below. The decay of [OH(v = 1)] with the
(
MKS).
The photolytic OH precursor H O was added
discharge off results mainly from vibrational relaxation by
H O and is used to estimate the k
(OH–H O ) total re-
2 2
2
2
2
2
v = 1
(
2.5 ± 1 mTorr) neat through a teflon needle valve 10 cm
moval rate constant. Finally, kv = 1(OH–O) was determined
by analysis of [OH(v = 1)] decay with the discharge on, tak-
ing into account removal by H O and other species.
upstream of the fluorescence detection zone, and OH was
generated by 193 nm photolysis using radiation from an
excimer laser (Lambda Physik COMPex 102). The H O
partial pressure was determined by monitoring the needle
valve opening with the pressure rise into the O /Ar flow.
Typical laser energies were 10 mJ in a 1.2 cm diameter
beam in the apparatus. Concentrated liquid H O
(
2
2
The OH(v = 0 and 1) profiles displayed in Figs. 2 and 3
were taken at the same photolysis laser energy (± 1 mJ),
and Ar, O , and H O partial pressures (constant to
2
2
2
2
2
2
3
± 0.001 Torr). The stability of [O( P)] was monitored by
the OH(v = 0) decay rate with the microwave discharge
on; this decay rate was constant to <0.5% on a given day
2
2
P95%) was prepared by pumping on 30% reagent. The

A. Khachatrian, P.J. Dagdigian / Chemical Physics Letters 415 (2005) 1–5
3
a
and varied by 65% between days. The baseline was deter-
mined by runs with the excimer laser blocked. The spike at
t = 0 in Figs. 2b and 3b is due to scattered light from the
excimer laser. After an initial buildup due to rotational
thermalization, the concentrations are seen to decay
exponentially.
v=0
0
0
0
.3
.2
.1
0
Comparison of the fluorescence intensities plotted in
Figs. 2 and 3 shows that the OH concentration drops by
a factor of ꢃ3 when the microwave discharge is turned
on. This is due to loss of H O from a slow chain reaction
2
2
3
involving O( P) atoms, HO , OH, and H atoms. With the
2
b
0.012
v=1
photolysis laser off and the discharge on, no OH laser fluo-
rescence was observed, and [OH(v = 0)] is <1% of the con-
centration with the laser on.
0
0
.008
.004
0
We have simulated this decomposition by kinetic model-
ing in order to estimate species concentrations in the flow
3
with the microwave discharge on and O( P) atoms present.
Table 1 presents the processes included the model (10 spe-
cies and 29 processes). Rate constants were taken from the
IUPAC compilation [12] and other sources [4,5,7,25,26].
OH vibrational levels were considered separate species in
order to model the vibrational relaxation, discussed below.
This modeling qualitatively, but not quantitatively, repro-
duces the observed decomposition of H O . For Ar, O ,
0
50
100
150
200
photolysis – probe delay (
µsec)
Fig. 2. Time-dependent concentrations of (a) OH(v = 0) and (b)
OH(v = 1) vs. the time between the photolysis and probe laser pulses,
with the microwave discharge off. The partial pressures of the added
reagents were: Ar, 2.1; O
and (b) represent sums over 16 and 13 runs, respectively. The solid lines
show least squares fits of the decay profiles to exponential functions.
2
2
2
and H O partial pressures 2.1, 0.1, and 0.0025 Torr and
2
2
an assumed 8% O dissociation fraction, we compute the
2
, 0.1; H
2
O
2
0.0025 ± 0.001 Torr. The plots in (a)
2
ꢀ
3
following concentrations (in molecule cm ) after 0.2 s
1
5
3
14
reaction time: [O ] = 3.1 · 10 , [O( P)] = 4.1 · 10 ,
2
1
3
12
[
1
H O ] = 5.6 · 10 ; [H O] = 7.0 · 10 ; [HO ] = 1.2 ·
2
2
2
2
1
0
10
13
11
0 ; [OH] = 2.8 · 10 ; [H] = 2.2 · 10 ; [H ] = 9.4 · 10 .
2
a
0.08
0.06
0.04
0.02
0
3
v=0
It can be seen that, aside from O( P), the free radical pre-
dicted to have the highest concentration is the H atom.
This model predicts 29% H O decomposition, significantly
2
2
less than observed, and a higher OH concentration than in-
9
ꢀ3
ferred (<7 · 10 molecules cm ) from our observations.
Nevertheless, these results are useful in taking account of
the decomposition of H O in computing kv = 1(OH–O)
2
2
and were employed to estimate species concentrations just
before the photolysis laser pulse. Inclusion of the possible
1
effects of O( D), O (a), or O in the flow did not signifi-
2
3
b
v=1
cantly affect the H O decomposition computed by our ki-
2
2
netic model.
0
.004
.002
0
The OH(v = 0 and 1) decays were fitted (solid lines
through measured concentrations in Figs. 2 and 3) to expo-
nential functions in a nonlinear least squares procedure
over 18 6 t 6 200 ls to determine decay rates. The decay
of [OH(v = 0)] with the discharge off (Fig. 2a) was fitted
to a first-order diffusion loss rate of 250 ± 60 s . This loss
was included for all the free radicals in kinetic modeling to
fit OH(v) decays after the photolysis laser pulse. A decay
0
ꢀ
1
0
50
100
150
sec)
200
4
ꢀ1
rate of (1.13 ± 0.01) · 10 s
time-dependent OH(v = 0) concentration with the dis-
was determined for the
photolysis – probe delay (
µ
Fig. 3. Time-dependent concentrations of (a) OH(v = 0) and (b)
OH(v = 1) vs. the time between the photolysis laser and probe laser
pulses, with the microwave discharge on. The partial pressures of the
added reagents were: Ar, 2.1; O , 0.1; H O 0.0025 ± 0.001 Torr. The plots
2 2 2
in (a) and (b) represent sums over 22 and 37 runs, respectively. The solid
lines show least squares fits of the decay profiles to exponential functions.
charge on (Fig. 3a). The assumed O dissociation fraction
2
was varied in the kinetic model to match this decay rate,
3
14
ꢀ3
and we estimate [O( P)] = (4.1 ± 0.2) · 10 atom cm at
the laser excitation/detection zone, with the dissociation
fraction given above; the quoted error includes fitting

4
A. Khachatrian, P.J. Dagdigian / Chemical Physics Letters 415 (2005) 1–5
Table 1
3
Kinetic model for simulating concentrations [species included: Ar, O
2
, O( P), OH(v = 0), OH(v = 1), H
O
2 2
, H
2
O, H, HO
2
, H
2
]
3
ꢀ1 ꢀ1
s )
Reaction
k
298 (cm molecule
Reference
3
ꢀ11
O( P) + OH(v = 0) ! H + O
2
3.5 · 10
3.9 · 10
1.7 · 10
5.8 · 10
1.7 · 10
6.7 · 10
1.1 · 10
1.5 · 10
1.6 · 10
4.2 · 10
5.2 · 10
5.6 · 10
7.2 · 10
2.4 · 10
1.1 · 10
1.0 · 10
6.0 · 10
5.4 · 10
4.3 · 10
6.9 · 10
4.0 · 10
1.3 · 10
1.9 · 10
4.0 · 10
1.5 · 10
[12]
This work
[12]
[12]
[12]
[12]
[12]
[12]
[12]
3
a
ꢀ11
ꢀ15
ꢀ11
O( P) + OH(v = 1) ! products
3
O( P) + H
O
2 2
! OH(v = 0) + HO
! OH(v = 0) + O
! H O + HO
! H O + H
! H O + O
OH(v) + OH(v) ! H
HO + HO
! H
H + H
! OH(v = 0) + H
+ HO
+ O
OH(v = 0) + OH(v = 0)
O + O
2
3
O( P) + HO
2
2
ꢀ12
ꢀ15
ꢀ10
ꢀ12
ꢀ12
ꢀ14
ꢀ15
ꢀ12
ꢀ11
ꢀ12
ꢀ16
ꢀ33
ꢀ33
ꢀ32
ꢀ30
ꢀ31
ꢀ32
ꢀ13
ꢀ11
ꢀ11
ꢀ10
OH(v) + H
OH(v) + H
2
O
2
2
2
2
2
OH(v) + HO
2
2
2
2
O + O
2
2
2
O
2
+ O
2
2
O
2
2
O
[25]
!
H
2
2
H + HO
2
! H
2
2
[12]
!
!
H
2
H + OH(v) ! H
2
+ O
[26]
[26]
[25]
[12]
[12]
[12]
[12]
[5]
O + O + M ! O
H + H + M ! H
H + O
+ M ! HO
H + OH(v) + M ! H
OH(v) + OH(v) + M ! H
HO + HO
+ M ! H
OH(v = 1) + O
2
+ M
2
+ M
+ M
O + M
2
2
2
O
2 2
+ M
+ O + M
! OH(v = 0) + O
O ! OH(v = 0) + H
! OH(v = 0) + H
2
2
2
O
2
2
2
2
OH(v = 1) + H
OH(v = 1) + H
2
2
O
[7]
2
O
2
2
O
2
This work
[29]
[4]
[4]
Assumed
Assumed
OH(v = 1) + H ! OH(v = 0) + H
OH(v = 1) + H
! OH(v = 0) + H
OH(v = 1) + Ar ! OH(v = 0) + Ar
ꢀ
ꢀ
ꢀ
15
15
11
ꢀ
2
2
1 · 10
1 · 10
5 · 10
OH(v = 1) + OH ! OH(v = 0) + OH
11
OH(v = 1) + HO
2
! OH(v = 0) + HO
2
3.5 · 10
a
The branching between reaction and vibrational relaxation was assumed to follow the prior distribution (see text).
uncertainties, but not the uncertainty in kv = 0(OH–O) em-
[OH(v = 1)] with the discharge on (Fig. 3b). Through com-
parison of this decay rate with variation of kv = 1(OH–O) in
our kinetic model with concentrations of H O decomposi-
ployed [12] in the model.
3
The [O( P)] determined from the OH(v = 0) decay rate is
2
2
slightly different from a value computed by just considering
the diffusion loss and reaction 1, because of the effect of
several OH-producing reactions in our kinetic model. It
tion products taken from above, we derive a value
ꢀ11
3
ꢀ1 ꢀ1
kv = 1(OH–O) = (3.9 ± 0.6) · 10
cm molecule
s
[1r
uncertainty, including the uncertainty in the literature
value [12] for kv = 0(OH–O)]. The major contributions to
[OH(v = 1)] decay can be estimated by multiplying
OH(v = 1) removal rate constants (Table 1) by concentra-
is well known [27] that a microwave discharge in O also
2
creates some metastable O (a). This species would be ex-
2
pected to de-activate OH(v = 1) at a rate comparable to
3
4
ꢀ1
3
that for O (X), which is much slower than for O( P) atoms.
tions to yield first-order decay rates (in 10 s ): O( P),
1.59; H, 0.32; H O , 0.23; O , 0.04; H O, 0.01. It can be
2
The vibrational relaxation of OH by H O has not been
2
2
2
2
2
2
previously reported. The decay of [OH(v = 1)] with the dis-
seen that the determined rate of removal of OH(v = 1) by
3
charge off was employed to estimate the rate constant. A de-
O( P) is somewhat dependent upon the assumed value for
3
ꢀ1
3
cay rate of (4.45 ± 0.13) · 10 s
was determined for
[H]. Since [O( P)] and kv = 1(OH–O) are determined by
[
OH(v = 1)] with the discharge off (Fig. 2b). By comparison
the OH(v = 0 and 1) decay rates, respectively, the ratio of
kv = 1(OH–O) to kv = 0(OH–O) is better determined than
kv = 1(OH–O) itself; we have determined the ratio
kv = 1(OH–O) to kv = 0(OH–O) to equal 1.1 ± 0.1 Hence,
kv = 1(OH–O) is found to be slightly larger than
kv = 0(OH–O), but the difference is within the experimental
uncertainty.
of this decay rate with kinetic modeling results, the
OH(v = 1)–H O vibrational relaxation rate constant
2
2
k
(OH–H O ) was estimated to equal (4.0 ± 1.0) ·
v = 1
11
2
2
ꢀ
3
ꢀ1 ꢀ1
1
0
cm molecule
s , or approximately twice that for
relaxation by H O [7]. The quoted uncertainty includes the
2
uncertainty in [H O ]. This OH(v = 1)–H O relaxation rate
2
2
2
2
was used in our kinetic modeling.
The observed decay of the OH(v = 1) concentration
4. Discussion
with the microwave discharge on (Fig. 3b) is primarily
3
determined by collisional removal by O( P) atoms, with a
The present study provides slight support to the previ-
ous experimental studies [14,15] and a single-surface classi-
cal trajectory study [17] that the removal of OH(v = 1) by
significant contribution due to H O and H atoms. A decay
2
ꢀ1
2
4
rate of (2.19 ± 0.16) · 10 s
was determined for

A. Khachatrian, P.J. Dagdigian / Chemical Physics Letters 415 (2005) 1–5
5
3
O( P) atoms is enhanced by vibrational excitation. Our
[3] I.W.M. Smith, M.D. Williams, J. Chem. Soc., Faraday Trans. 2 (81)
1985) 1849.
4] K.J. Rensberger, J.N. Jeffries, D.R. Crosley, J. Chem. Phys. 90 (1989)
174.
5] J.A. Dodd, S.J. Lipson, W.A.M. Blumberg, J. Chem. Phys. 95 (1991)
5752.
(
total collisional removal rate constant is significantly lower
than that determined in an EPR study [14] and somewhat
lower than the value reported for the collisional removal
of OH(v = 2) [15]. The single-surface classical trajectory
study of Varandas [17] predicts a similar small enhance-
ment of the total removal rate constant in OH(v)–O( P)
collisions vs. the OH vibrational quantum number.
Since the collisional removal of OH with O( P) atoms is
mediated by the initial formation of an HO complex, it
[
[
2
[6] M.J. Dyer, K. Knutsen, R.A. Copeland, J. Chem. Phys. 107 (1997)
7809.
3
[
7] E. Silvente, R.C. Richter, A.J. Hynes, J. Chem. Soc., Faraday Trans.
3 (1997) 2821.
9
3
[
8] J. Lacoursi e´ re, M.J. Dyer, R.A. Copeland, J. Chem. Phys. 118 (2003)
1661.
2
would be expected that enhancement in k (OH–O) by
[9] D.C. McCabe, S.S. Brown, M.K. Gilles, R.K. Talukdar, I.W.M.
Smith, A.R. Ravishankara, J. Phys. Chem. A 107 (2003) 7762.
10] L. DÕOttone, D. Bauer, P. Campuzano-Jost, M. Fardy, A.J. Hynes,
Phys. Chem. Chem. Phys. 6 (2004) 4276.
11] I.W.M. Smith, J. Chem. Soc., Faraday Trans. 87 (1991) 2271.
[12] R. Atkinson, D.L. Baulch, R.A. Cox, J.N. Crowley, R.F. Hampson,
R.G. Hynes, M.E. Jenkins, M.J. Rossi, J. Troe, Atmos. Chem. Phys.
v
OH vibrational excitation would result from an enhanced
formation rate of the complex. The rate of formation of
the complex is determined by the strongly attractive
[
[
2
2
0
0
0
~
~
X A and A A HO PESs. Hence, the enhancement in
2
kv(OH–O) with an increasing v should reflect the depen-
dence of the PESs on the OH internuclear separation in
the entrance channel.
4
(2004) 1461. Available from: http://www.iupac-kinetic.ch.cam.a-
c.uk/index.html .
[
13] S.P. Sander, A.R. Ravishankara, D.M. Golden, C.E. Kolb, M.J.
Kurylo, R.E. Huie, V.L. Orkin, M.J. Molina, G.K. Moortgat, B.J.
Finlayson-Pitts, Chemical kinetics and photochemical data for use in
atmospheric studies. Evaluation no. 14, Jet Propulsion Laboratory,
Pasadena, CA, 2003. Available from: <http://jpldataeval.jpl.nasa.
gov>.
3
Collisional removal of OH(v P 1) by O( P) atoms oc-
curs by both chemical reaction and vibrational relaxation.
Since this process proceeds through formation and decay
of a transient HO complex (see Fig. 1), we can make an
2
estimate of the branching ratio for chemical reaction vs.
vibrational relaxation through a statistical theory, namely
the prior distribution [28]. Summing over the accessible
[
[
14] J.E. Spenser, G.P. Glass, Int. J. Chem. Kinetics 9 (1977) 97, 111.
15] J. Marschall, K.S. Kalogerakis, R.A. Copeland, Laboratory Mea-
surements of OH(v = 2) Collisional Reactivation by Oxygen Atoms,
American Geophysical Union Spring 2001 meeting, paper SA31A-21.
16] R. Robertson, G.P. Smith, Chem. Phys. Lett. 358 (2002) 157.
17] A.J.C. Varandas, Chem. Phys. Lett. 396 (2004) 182.
18] A. Jacobs, K. Kleinermanns, H. Kuge, J. Wolfrum, J. Chem. Phys. 97
3
ꢀ
1
rovibrational molecular levels (in both the X R and a D
g
[
[
[
electronic states of O ), the prior statistical model predicts
2
branching ratios of 85 and 15% for chemical reaction and
collisional vibrational relaxation, respectively, in OH-
(
1983) 3162.
3
(
v = 1) + O( P) collisions. It would be interesting to com-
[19] A.U. Grunewald, K.-H. Gericke, F.J. Comes, J. Chem. Phys. 89
(1988) 345.
pare the rate constants k (OH–O) and predicted branching
v
[20] B. Nizamov, X. Yang, P.J. Dagdigian, M.H. Alexander, J. Phys.
ratios with the results of multi-surface theoretical treat-
ments of the collision dynamics.
Chem. A 106 (2002) 8345.
[
[
[
21] X. Tan, P.J. Dagdigian, Chem. Phys. Lett. 375 (2003) 532.
22] S.L. Manatt, M.R.R. Manatt, Chem. Eur. J. 10 (2004) 6540.
23] A.A. Turnipseed, G.L. Vaghjiani, J.E. Thompson, A.R. Ravishan-
kara, J. Chem. Phys. 96 (1992) 5887.
Acknowledgements
[
[
24] T.L. Myers, N.R. Forde, B. Hu, D.C. Kitchen, L.J. Butler, J. Chem.
Phys. 107 (1997) 5361.
25] D.L. Baulch, C.J. Cobos, R.A. Cox, C. Esser, P. Frank, Th. Just, J.A.
Kerr, M.J. Pilling, J. Troe, R.W. Walker, J. Warnatz, J. Phys. Chem.
Ref. Data 21 (1992) 411.
26] W. Tsang, R.F. Hampson, J. Phys. Chem. Ref. Data 15 (1986) 1087.
27] M.A.A. Clyne, W.S. Nip, in: D.W. Setser (Ed.), Reactive Interme-
diates in the Gas Phase, Academic Press, New York, 1979, p. 2.
28] R.D. Levine, R.B. Bernstein, Molecular Reaction Dynamics and
Chemical Reactivity, Oxford University Press, New York, 1987.
We appreciate the comments of two anonymous referees
and the encouragement of Millard Alexander. This re-
search has been supported by the Air Force Office of Scien-
tific Research under Grant No. FA9550-04-1-0103.
[
[
References
[
[
[
1] S. Adler-Golden, J. Geophys. Res. 102 (1997) 19969.
2] S.R. Conway, M.E. Summers, M.H. Stevens, J.G. Cardon, Geophys.
Res. Lett. 27 (2000) 2613.
[29] P.W. Barnes, P. Sharkey, I.R. Sims, I.W.M. Smith, Faraday Discuss.
113 (1999) 167.