The ultraviolet photodissociation of jet-cooled ClO and BrO radicals

Article abstract of DOI:10.1063/1.1448282

A molecular beam source for jet-cooled ClO and BrO radicals was developed using an electric discharge in order to study the UV photodissociation dynamics of these species. In the photodissociation of ClO, the wavelength dependence of the dynamics was observed. In the BrO experiments, the bond dissociation energy was directly measured.

Full text of DOI:10.1063/1.1448282

The ultraviolet photodissociation of jet-cooled ClO and BrO radicals
Peng Zou, Hahkjoon Kim, and Simon W. North
Citation: The Journal of Chemical Physics 116, 4176 (2002); doi: 10.1063/1.1448282
View online: http://dx.doi.org/10.1063/1.1448282
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/116/10?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Photodissociation of the BrO radical using velocity map ion imaging: Excited state dynamics and accurate D 0 0 (
Br O ) evaluation
J. Chem. Phys. 124, 134304 (2006); 10.1063/1.2173265
The UV photodissociation dynamics of ClO radical using velocity map ion imaging
J. Chem. Phys. 123, 174303 (2005); 10.1063/1.2083487
Photodissociation of CSCl 2 at 235 nm: Kinetic energy distributions and branching ratios of Cl atoms and CSCl
radicals
J. Chem. Phys. 117, 1123 (2002); 10.1063/1.1480272
C–Cl bond fission, HCl elimination, and secondary radical decomposition in the 193 nm photodissociation of allyl
chloride
J. Chem. Phys. 116, 2763 (2002); 10.1063/1.1433965
Resonance enhanced multiphoton ionization/time-of-flight measurements of the velocity and internal energy
content of thermal and photochemical methyl radical sources
Rev. Sci. Instrum. 68, 2031 (1997); 10.1063/1.1148093
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.70.241.163 On: Sun, 21 Dec 2014 02:21:48

JOURNAL OF CHEMICAL PHYSICS
VOLUME 116, NUMBER 10
8 MARCH 2002
The ultraviolet photodissociation of jet-cooled ClO and BrO radicals
Peng Zou, Hahkjoon Kim, and Simon W. Northa)
Department of Chemistry, Texas A&M University, College Station, Texas 77842
͑Received 23 October 2001; accepted 11 December 2001͒
The photodissociation dynamics of jet-cooled ClO and BrO radicals have been studied using
resonance-enhanced multiphoton ionization with time-of-flight mass spectrometry. Four asymptotic
electronic channels are energetically allowed for ClO photodissociation at 235 nm, although the
dominant products observed at 235 nm are Cl(2P3/2) and O(1D2). Polarization-dependent
time-of-flight profiles indicate that this channel has an anisotropy parameter of ␤ϭ1.2Ϯ0.2.
Evidence for O(3PJ) products was detected in coincidence with Cl(2P1/2) fragments, and we
estimate that this channel constitutes Ͻ3% at this wavelength. In the photodissociation of BrO at
355 nm we observe two asymptotic electronic state channels, Br(2P3/2)ϩO(3P2) and Br(2P1/2
)
ϩO(3P2) in a 0.75:0.25 ratio. Both channels have an anisotropy parameter of ␤ϭ1.5Ϯ0.1. Based
on the measured asymptotic velocities of both channels, we have directly determined the Br–O bond
dissociation energy to be 55.8Ϯ1.0 kcal/mol, providing a heat of formation for the BrO radical at
298 K of 29.7Ϯ1.0 kcal/mol in good agreement with previous spectroscopic determinations.
© 2002 American Institute of Physics. ͓DOI: 10.1063/1.1448282͔
I. INTRODUCTION
Photolysis at wavelengths longer than 263 nm can only result
in O(3PJ) fragments ͑channels 2 and 3͒. The formation of
these fragments from the bound levels of the A 2⌸i state is
due to predissociation via several repulsive electronic states.
On the basis of rotational linewidths, predissociation life-
times on the order of a picosecond have been reported.11 Ab
initio calculations suggest that the principal candidates for
states which predissociate the A 2⌸i state are the 1 4⌺ϩ,
2 4⌺Ϫ, and 3 2⌸ states.10–12
The importance of chlorine and bromine in the catalytic
destruction of stratospheric ozone has been well established.1
Halogen oxides, in particular ClO and BrO radicals, repre-
sent important intermediate species in catalytic cycles of
ozone destruction in the middle and upper stratosphere.2,3
The coupling of ClO and BrO cycles to produce free bro-
mine and chlorine atoms via the reaction
There have only been two direct experimental studies of
the UV photodissociation of the ClO radical. Davis and Lee
studied the photolysis at 248 nm using molecular beam neu-
tral time-of-flight ͑TOF͒ and concluded that channel 2 was
the dominant channel.13 A nonlimiting anisotropy parameter
͑␤ϭ1.2͒ was determined for this channel, suggesting that a
perpendicular transition contributes significantly to the oscil-
lator strength at this wavelength. The authors also observed a
minor contribution from the Cl(2PJ)ϩO(3PJ) channel
͑Ͻ3%͒. Schmidt et al. subsequently measured the velocity
distributions of the O(1D2) atom fragments using 2ϩ1
resonance-enhanced multiphoton ionization ͑REMPI͒ from
thermal ClO photodissociation in the wavelength range of
237–270 nm.14 Insufficient resolution precluded a determi-
nation of the Cl(2P3/2)/Cl(2P1/2) branching ratio, and the
role of the minor O(3PJ) channel was not discussed although
the authors reported an O(1D2) yield of unity throughout the
wavelength range. It is unclear a priori how the fragment
anisotropy and branching ratios will be affected as a function
of wavelength in the continuum region. The potential roles of
higher-lying states reached via perpendicular transitions have
been predicted by theory and their influence on the dissocia-
tion dynamics is untested.
BrOϩClO→BrϩClϩO2
is estimated to be responsible for about 20%–25% of the
lower stratosphere ozone depletion with even moderate mix-
ing ratios of bromine.4
There are several pathways which convert ClO to Cl
atoms including UV photolysis that accounts for 2%–3% of
the total loss in the stratosphere.5 The role of ClO photodis-
sociation is also important in the interpretation of secondary
photochemistry in experimental studies of ClONO2 and
͑1͒
Cl2O photodissociation, where ClO is
a
primary
photofragment.6,7
The UV absorption spectrum of ClO consists of a vibra-
tional progression that begins at 316 nm and terminates near
263 nm.8 Based on the convergence of the vibrational struc-
ture Durie and Ramsay reported a Cl–O bond dissociation
energy of 63.427Ϯ0.008 kcal/mol.9 A broad featureless con-
tinuum extends from 263 to near 220 nm. In this continuum
region there are four thermodynamically accessible pathways
for ClO photodissociation,
h␯
2
3
ClO X 2⌸ ͒→Cl P ͒ϩO P ͒
͑2͒
͑3͒
͑4͒
͑5͒
͑
͑
͑
⍀
3/2
J
2
3
→Cl P ͒ϩO P ͒
͑
͑
1/2
J
In contrast to ClO, the thermodynamics of BrO radicals
are not particularly well established. Until as recently as
1997, the NASA panel had recommended a value of 26Ϯ5
kcal/mol for the BrO heat of formation.15 Since BrO radicals
are involved in many stratospheric chemical reactions, an
accurate value for the BrO heat of formation is critical for
2
1
→Cl P ͒ϩO D ͒
͑
͑
3/2
2
2
1
→Cl P ͒ϩO D ͒.
͑
͑
1/2
2
a͒
Electronic mail: north@mail.chem.tamu.edu
0021-9606/2002/116(10)/4176/8/$19.00
4176
© 2002 American Institute of Physics
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.70.241.163 On: Sun, 21 Dec 2014 02:21:48

J. Chem. Phys., Vol. 116, No. 10, 8 March 2002
Photodissociation of ClO and BrO
4177
determining these energetics of reactions and heats of forma-
tion for other bromine-containing species. The heat of for-
mation of BrO has been measured either by measuring ki-
netic rate constants of BrO related reactions16,17 or by
measuring the BrO dissociation energy spectroscopically.8,18
Although these results have overlapping experimental uncer-
tainty ranges, there is clearly a discrepancy between early
kinetic measurements and spectroscopic measurements.
However, recent measurements of the enthalpy of reaction
for the decomposition of BrONO2 ͑Ref. 19͒ and the equilib-
rium process involving BrONO2ϩH2O,20
BrONO2→BrOϩNO2 ,
͑6͒
͑7͒
BrONO2ϩH2O↔HOBrϩHONO2 ,
indirectly support the spectroscopic values.21
The UV absorption spectrum of BrO consists of a diffuse
vibrational progression that ranges from 280 to 380 nm. The
absorption band has well-defined vibrational structure, and is
assigned to the A 2⌸3/2 X 2⌸3/2 electronic transition.8
Birge–Sponer analysis of the BrO vibrational structure by
Durie and Ramsay yielded a Br–O bond dissociation energy
of 55.2Ϯ0.8 kcal/mol.
There are two thermodynamically accessible pathways
for BrO photodissociation,
FIG. 1. Cross-sectional view of the discharge molecular beam source.
h␯
2
3
BrO X 2⌸ ͒→Br P ͒ϩO P ͒
͑8͒
͑9͒
͑
͑
͑
⍀
3/2
J
2
3
→Br P ͒ϩO P ͒.
͑
͑
1/2
J
or two laser pulses. The resulting ions were detected using a
two-stage Wiley–McLaren time-of-flight mass spectrometer
͑TOF-MS͒ in either ‘‘noncore-sampled’’ or ‘‘core-sampled
mode’’ using delayed pulsed extraction.25–28
Theoretical calculation of the low-lying excited states of BrO
shows that the bound A 2⌸3/2 state is crossed by several
22
2
2
2
2⌺Ϫ, ⌺ϩ, ⌸, ⌬ repulsive states. Rotational bandwidth
analysis by Wheeler et al.23 indicate that the ϭ12 and
In the ClO radical experiments, argon was bubbled
through Cl2O maintained at Ϫ64 °C in a dry-ice/chloroform
slush resulting in a molecular beam of 5% Cl2O in approxi-
mately 800 Torr Ar. ClO radicals were produced when the
mixture passed through a discharge source mounted in front
of the pulsed valve nozzle, and Cl2O was converted to ClO
and Cl. A schematic diagram of the source is shown in Fig. 1.
A large negative potential ͑Ϫ800 V͒ is applied to the cathode
and provides a discharge, with the electron current traveling
toward the valve orifice during the beam pulse. Typical peak
currents during pulsed valve operation were 500 ␮A. The
pulsed discharge source is base on previous design with
slight modification.29,30 In order to minimize interference
from charged species, a deflection plate, biased at Ϫ150 V,
was placed directly beyond the discharge source. In the BrO
radical experiments, molecular oxygen was bubbled through
Br2 maintained at Ϫ10 °C in resulting in a molecular beam
of 10% Br2 in approximately 800 Torr O2 . BrO radicals
were produced by the recombination of Br and O atom aris-
ing from the discharge. Cooper et al. were also able to pro-
duce a jet-cooled beam of ClO using a discharge of 5% Cl2
and 5% O2 in Ar.31 In those studies the ClO beam was char-
acterized with a rotational temperature of 15 K based on
simulations to the observed 2ϩ1 REMPI spectra. In order to
determine the bond dissociation energy and angular distribu-
tion photofragments, the temperature of the molecular beam
must be characterized. Characterization of the beam tempera-
ture was estimated by measuring jet-cooled 2ϩ1 REMPI
v
v
Ј
Ј
ϭ7 levels have predissociation lifetimes of 1.3Ϯ0.2 and 1.7
Ϯ0.2 ps, respectively. Wilmouth et al.18 measured the rota-
tional level dependent lifetime for the ϭ12 and ϭ7
v
v
Ј
Ј
levels, and the values are in the order of magnitude of pico-
seconds. To the best of our knowledge, there have been no
direct measurements of the fragments arising from BrO pho-
todissociation at any wavelength, and very little is known
about the predissociation dynamics and product state branch-
ing ratios.
We have developed a molecular beam source for jet-
cooled ClO and BrO radicals using an electric discharge in
order to study the UV photodissociation dynamics of these
species. The paper reports the results of these studies, using
resonance-enhanced multiphoton ionization with time-of-
flight mass spectrometry, with two distinct objectives in
mind. In the photodissociation of ClO we are primarily in-
terested in the wavelength dependence of the dynamics in
order to compare to recent theoretical work. In the BrO ex-
periments our focus is on direct measurement of the bond
dissociation energy.
II. EXPERIMENT
The molecular beam apparatus and experimental setup
has been described in detail elsewhere and only a brief sum-
mary is presented here.24 A pulsed molecular beam was col-
limated by a conical skimmer and intersected by either one
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.70.241.163 On: Sun, 21 Dec 2014 02:21:48

4178
J. Chem. Phys., Vol. 116, No. 10, 8 March 2002
Zou, Kim, and North
spectra of CO using the (B X) Q-branch transition around
230 nm.32,33 The CO beam was prepared using a discharge
through a 5% mixture of acetone and N2 carrier gas at a total
pressure of 800 torr. We typically observed CO rotational
temperatures of ϳ10 K with negligible vibrational excita-
tion.
The chlorine atoms ͓Cl(2P3/2) and Cl(2P1/2)͔ were
probed using 2ϩ1 REMPI transitions in the region of 235
nm34 and the O(3P2) atoms were probed using 2ϩ1 REMPI
transitions near 226 nm.35 The laser pulses were generated
pumping a dye laser with the third harmonic of a Nd:YAG
laser operating at 10 Hz, using either Coumarin 480 or Cou-
marin 450 dyes for the Cl and O transitions, respectively. In
the ClO photodissociation experiment at 235 nm, the dye
laser was used for both pump and probe pulses. The BrO
photodissociation a beam splitter picked off a fraction of the
355 nm beam prior to entering the dye laser and this beam
was overlapped with the probe beam from dye laser.
The ions were detected by a set of dual Chevron micro-
channel plates. A 400 MHz digital oscilloscope was used to
collect noncore sampled TOF spectra interfaced to a Pentium
PC. The core-sampled data were acquired using a multichan-
nel scalar with 5 ns bin width.
Cl2O was synthesized by the method of Cady.36 Cl2 was
collected on prebaked HgO ͑Aldrich͒ powder and the reac-
tion was allowed to run overnight at 195 K. The product was
then purified by vacuum distillation, and the purity ͑Ͼ90%͒
was checked by UV absorption spectroscopy. The Br2 and
O2 were used from commercial sources without further pu-
rification.
FIG. 2. The closed circles are experimental data taken with the discharge
source on ͑top͒ and off ͑bottom͒. Shown beneath the discharge on data is a
simulation of the predicted cored REMPI-TOF profiles of Cl(2P3/2) pro-
duced from the 235 nm photodissociation of ClO including chlorine atoms
in the beam. The solid arrows indicate the identity of the peaks in the TOF
spectra. The dashed arrows indicate the arrival times of the 35Cl(2P3/2
formed in coincidence with O(3PJ). The shaded peaks represent Cl(2P3/2
atoms produced by discharge.
)
)
III. RESULTS AND ANALYSIS
A. ClO photodissociation at 226 nm
Shown in Fig. 2 are experimental core-sampled TOF
spectra for Cl(2P3/2) fragments acquired using light polar-
ized along the detector axis ͑␹ϭ0°͒ with the discharge on
͑top panel͒ and off ͑bottom panel͒. The discharge off spec-
trum can be fitted using the forward convolution ͑FC͒
technique,24 and the derived speed and angular distributions
are consistent with Cl2O photodissociation at 235 nm.6,37 We
estimate that, based on the relative absorption cross sections
of Cl2O and ClO, the discharge results in Ͼ70% conversion
of Cl2O to ClO. The features that appear with the discharge
on arise from ClO photodissociation and from entrained Cl
atoms. In the case of diatomic molecule photodissociation
the partitioning of the available energy involves only elec-
tronic excitation and translation of the photofragments. The
total translational energy, ET , is thus given by
response function. We have adopted an instrument response
function of a Gaussian ͑60 ns full width at half-maximum͒
based on the width of the entrained atomic species. Peaks
corresponding to entrained chlorine atoms that are also pro-
duced in the discharge are included ͑indicated by the shaded
region͒. The contributions from the 35ClϩO(1D) and 37Cl
ϩO(1D) channels are indicated in Fig. 2 by a set of combs
above the respective peaks. The simulation includes contri-
butions only from O(1D2) coincident fragments, neglecting
the minor Cl2O photolysis contribution. The dashed arrows
indicate the predicted arrival times corresponding to forma-
tion of coincident O(3PJ) fragments. No evidence for chan-
nel 2 is observed, and we estimate that this channel contrib-
utes less than 2% to the total Cl(2P3/2) signal based on the
signal-to-noise ratio of the TOF spectra.
The Cl(2P1/2) signal is extremely weak indicating that
channels 3 and 5 make small contributions at this wave-
length. The small signal size precluded collection of core-
sampled data. The noncore-sampled TOF spectra for
Cl(2P1/2) taken with the laser polarization parallel to the
flight axis ͑upper panel͒ and perpendicular to the flight axis
͑lower panel͒ are shown in Fig. 3. The observed TOF profiles
are significantly wider than predicted by the presence of only
channel 5. The FC simulations to the data therefore, consist
of two components for each isotope; a contribution from the
ETϭEh␯ϪD0 Cl–O͒ϪEClϪEOϩEparent
,
␯,J,e
͑10͒
͑
0
e
e
where Eh␯ is the energy of the dissociation photon,
D00(Cl–O) is the bond dissociation energy of ClO ͑63.4 kcal/
mol͒, Eie is the electronic energy of the ith atomic fragment,
and Eparent is the internal energy of the parent molecule prior
␯,J,e
to dissociation. Only the distribution of initial parent energy
results in a broadening of the translational energy distribu-
tion. The solid line in the top panel of Fig. 2 is a forward
convolution simulation to the data including an instrumental
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.70.241.163 On: Sun, 21 Dec 2014 02:21:48

J. Chem. Phys., Vol. 116, No. 10, 8 March 2002
Photodissociation of ClO and BrO
4179
TABLE I. Fragment anisotropy parameters and quantum yields arising from
the photodissociation of ClO at 235 nm.
Channel
⌽
␤
Cl(2P3/2)ϩO(3PJ)
Cl(2P1/2)ϩO(3PJ)
Cl(2P3/2)ϩO(1D2)
Cl(2P1/2)ϩO(1D2)
a
a
0.02Ϯ0.01
0.96Ϯ0.02
0.02Ϯ0.01
1.1Ϯ0.3
1.2Ϯ0.2
Ͼ0
aNot observed in the TOF spectra.
Cl(2P3/2) channels, the measurements are insensitive to the
changes of Cl(2P3/2)/Cl(2P1/2) line strengths. This includes
a minor correction for the entrained atomic signal that was
determined based on simulations of the TOF spectra. The
relative contributions of channels 2–5 are given in Table I.
Our branching ratios are consistent with the results of Davis
and Lee at 248 nm, who reported a O(1D2)/O(3PJ) ratio of
0.97:0.03. Those authors concluded that channel 3 was more
significant than channel 2 based on a slightly better fit to
their TOF spectra.
In order to determine the anisotropy parameters for the
asymptotic channels two types of experiments were per-
formed. The first involves recording noncored TOF spectra
using three experimental geometries. The projection of the
velocity distribution onto the flight axis for each speed,
,
v
corresponding to a specific channel is given by43,44
FIG. 3. Noncore sampled REMPI-TOF profiles of Cl(2P1/2) produced from
the 235 nm photodissociation of ClO. The closed circles are the experimen-
tal data and the solid lines are forward convolution fits. The top and bottom
panels correspond to horizontal ͑␹ϭ90°͒, and vertical ͑␹ϭ0°͒ experimental
geometries, respectively. The dashed lines are the contributions from
Cl(2P1/2) atoms formed in coincidence with the O(3PJ) fragments. The
dashed-dotted lines are the contribution from both entrained Cl atoms and
from Cl(2P1/2) atoms formed in coincidence with the O(1D) fragments.
ϱ
1
vz
f
,␹͒ϭ
1ϩ␤P cos ␹͒P
d ,
͑11͒
͑vz
͑
v
͵
ͫ
ͩ ͪ
ͬ
2
2
2
v
͉ ͉
z
v
v
where ␤ is the spatial anisotropy parameter, P2(␹) is the
second Legendre polynomial, and ␹ is the angle between the
electric field vector of the laser beam and the flight axis. The
best fit simulations to the data are shown in Fig. 4 for
Cl(2P3/2). The simulation includes a contribution from en-
trained chlorine atoms. This unavoidable contamination fa-
cilitates analysis since it provides an internal calibration of
the ClO dissociation signal for different geometries, assum-
ing that there is no alignment of the entrained atoms. A mi-
nor contribution from Cl2O photodissociation ͑discharge
off͒ is included in simulating the experimental data. The fits
are for shown for ␤ϭ1.2. An alternative method involves
measuring the difference in signal intensity of core-sampled
data between horizontal ͑␹ϭ90°͒ and vertical ͑␹ϭ0°͒
polarization.45,46 Both methods resulted in similar anisotropy
parameters, although we prefer analysis of the noncore
sampled data which is less sensitive to laser power and beam
deviations since it depends only on shape and not the inten-
sity of the profiles. All channels observed exhibit strong an-
isotropy, indicative of prompt dissociation. Since the disso-
ciation wavelength used in the ClO experiments is in the
continuum region of the A–X band, the dissociation lifetime
is sufficiently short to neglect the rotational depolarization of
the ClO anisotropy.
O(3PJ) coincident channel ͑dashed lines͒ and a contribution
from Cl(2P1/2) atoms in coincidence with O(1D2) ͑dashed-
dotted lines͒ which includes the entrained atoms. The sum of
both contributions is indicated by the solid line. Although the
relative contributions of channels 3 and 5 are difficult to
quantify due to uncertainty in the magnitude of the entrained
atom signal, we estimate that the channels have approxi-
mately equal yields. Attempts to fit the data without inclu-
sion of the O(1D2) channel, resulted in substantially poorer
fits.
Determination of the relative branching of the four
asymptotic electronic channels requires the Cl(2P1/2)/
Cl(2P3/2) branching ratio. This ratio was determined from
integrated 2ϩ1 REMPI scans of the Cl(2P1/2) and Cl(2P3/2
)
transitions near 235 nm. The detection efficiency of the
Cl(2P1/2) and Cl(2P3/2) transitions at 235.205 and 235.336
nm, respectively, have been well-established by calibration
to the spin–orbit measurements of HCl dissociation at 193
nm by Wittig and co-workers.38 Using the results of Ref. 38,
the previously published values for the relative
Cl(2P3/2)/Cl(2P1/2) line strengths of 0.8 ͑Refs. 39–41͒ has
been adjusted to 0.6.42 Based on the calibrated sensitivity of
both transitions we find that the Cl(2P1/2)/Cl(2P3/2) branch-
ing ratio is 0.04. Given the small ratio of Cl(2P1/2) to
2
The ground and first excited state of ClO are ⌸i and,
therefore, excitation involves a parallel transition ͑⌬⍀ϭ0͒. A
contribution from the higher spin–orbit excited component
of the ground state, ⌸1/2 to the ⌸1/2 component of the A
state also involves a parallel transition. Ab initio calculations
have predicted that these are the dominant transitions to the
2
2
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.70.241.163 On: Sun, 21 Dec 2014 02:21:48

4180
J. Chem. Phys., Vol. 116, No. 10, 8 March 2002
Zou, Kim, and North
for Cl(2P1/2) fragments taken with ␹ϭ90° and ␹ϭ0° are
different, indicating velocity anisotropy. The FC simulations
for the O(3PJ) coincident channel were generated using the
best fit anisotropy ␤ϭ1.1Ϯ0.3 ͑dashed line͒. The 3 2⌸i state,
which is also thought to be responsible for the predissocia-
tion of the bound levels of the A 2⌸i state, correlates to
ground state oxygen atoms. However, it is unlikely that this
state is reached directly since the vertical excitation energy
greatly exceeds the energy of a 235 nm photon.10,12 Excita-
tion to the A 2⌸i state followed by coupling to the 3 2⌸i
state beyond the Franck–Condon region is thus necessary. A
one-dimensional Landau–Zener curve crossing calculation
predicts that a small adiabatic splitting, assuming a crossing
location near the asymptote ͑4.2 eV͒,10 would result in a
large wavelength-independent nonadiabatic curve crossing
probability. Toniolo et al. find that the yield for O(3P)
should be 2%, from the A 2⌸ states crossing primarily to the
3 2⌸ and 1 2⌬ states. This is very close to the experimental
O(3P) yield measured in the present study and Davis and
Lee’s result.13 The perpendicular contribution to the
Cl(2P1/2)ϩO(3PJ) channel, which may be substantial, could
arise from numerous repulsive states accessible in the
Franck–Condon region. Since O(3P) formation represents a
minor product channel, these states do not contribute signifi-
cantly to the total oscillator strength. However, a small os-
cillator strength from these states can make significant con-
tribution to the O(3P) minor channel and may account for
the 30% perpendicular transition observed in the experi-
ments. A positive anisotropy ͑␤Ͼ0͒ for the O(1D2)
ϩCl(2P1/2) channel contribution produced a slightly better
fit although there is insufficient velocity resolution for quan-
tification.
FIG. 4. Noncore sampled REMPI-TOF profiles of Cl(2P3/2) produced from
the 235 nm photodissociation of ClO. The closed circles are the experimen-
tal data and the solid lines are forward convolution fits. The middle, top, and
bottom panels correspond to magic angle ͑␹ϭ54.7°͒, horizontal ͑␹ϭ90°͒,
and vertical ͑␹ϭ0°͒ experimental geometries, respectively. The dashed lines
in the bottom panel are the contributions from the 35Cl and 37Cl isotopes.
B. BrO photodissociation
Shown in Fig. 5 are experimental TOF spectra for
O(3P2) fragments acquired with the discharge on and the
355 nm dissociation laser on ͑top panel͒ and off ͑bottom
panel͒. The features that appear with both 355 and 225 nm
present arise from BrO photodissociation. No signal of BrO
dissociation was observed when either the discharge was
turned off or the Br2 flow was turned off. Two fast compo-
broad continuum. According to ab initio calculations only a
minor contribution to the oscillator strength arises from the
2
2⌺, ⌬, and 4⌺ excited states which result from perpendicu-
lar transitions ͑⌬⍀ϭϮ1͒.10 However, the observation of
nonlimiting anisotropy parameters suggests that perpendicu-
lar transitions may constitute nearly one-third of the oscilla-
tor strength for channels 3 and 4, and perhaps channel 5. It is
unlikely that this deviation from limiting anisotropy can be
due to either lifetime or rotational depolarization effects. Our
anisotropy parameter for channel 4 is identical to the mea-
surement of Davis and Lee.13 Orr-Ewing and co-workers pre-
dict that the only excited state correlating to excited state
oxygen atoms, 2 2⌺ϩ, reached via a perpendicular transition,
could play a significant role in the dissociation. And this state
may have a strong oscillator strength and make a significant
contribution to ClO photochemistry below 240 nm.12 Toniolo
et al., however, predict that this state should contribute only
a fraction of 0.02 of the oscillator strength of the A 2⌸
state,10 which is not consistent with the experimental result
of Davis and Lee and our measurement.
nents can be observed which correspond to the Br(2P3/2
)
ϩO(3P2) ͑channel 8͒ and Br(2P1/2)ϩO(3P2) ͑channel 9͒
indicated by combs above the TOF spectra. The solid line in
the top panel of Fig. 5 is a forward convolution simulation to
the data including the instrumental response function which
is close to the intrinsic resolution limit of the apparatus ͑25
ns full width at half maximum͒. The minor difference ͑1 ns͒
between O atoms arising from the individual BrO isoto-
pomers has been included in the simulation.
The slow component in center-of-mass frame which is
centered at 6.45 ␮s and observed with the 355 nm laser on
and off, can be attributed to both entrained oxygen atoms
produced by the discharge source and the photodissociation
of O3 , formed via recombination, at 226 nm. Miller et al.47
and Syage48 have performed ozone photodissociation experi-
ments at 226 nm. Syage observed the O(3P2) fragment
shows obvious bimodal velocity distribution, with anisotropy
Figure 3 shows that the noncore sampled TOF spectra
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.70.241.163 On: Sun, 21 Dec 2014 02:21:48

J. Chem. Phys., Vol. 116, No. 10, 8 March 2002
Photodissociation of ClO and BrO
4181
FIG. 6. REMPI-TOF profiles of O(3P2) produced from the 355 nm photo-
dissociation of BrO. The top and bottom panels correspond to vertical ͑␹
ϭ0°͒, and horizontal ͑␹ϭ90°͒ experimental geometries, respectively.
FIG. 5. The closed circles are experimental data taken with the discharge
source on. The upper set of data corresponds to both 355 and 225 nm. The
solid line is a forward convolution simulation of the O(3P2) data including
contributions from both Br(2P1/2) and Br(2P1/2) coincident fragments. The
lower data set corresponds to only 225 nm.
Figure 6 shows ␹ϭ90° and ␹ϭ0° noncore sampled TOF
spectra for O(3P2) fragments. The best fit anisotropy ␤ϭ1.4
Ϯ1.0 for both the Br(2P3/2) and Br(2P1/2) coincident chan-
nels is shown. The quantum yield of Br(2P3/2)ϩO(3P2) and
Br(2P1/2)ϩO(3P2) is 0.83Ϯ0.10 and 0.17Ϯ0.10, respec-
tively, from the forward convolution fitting. Unlike ClO
photodissociation at 235 nm, which occurs promptly in
the continuum region of UV absorption spectrum, the predis-
sociation lifetime of BrO is significantly longer and may
result in depolarization of the photofragment angular distri-
bution. In order to determine the intrinsic anisotropy param-
eters for BrO photodissociation, the rotational depolarization
must be taken into account.
parameters 1.1Ϯ0.1 and 0.4Ϯ0.2 for fast and slow compo-
nents, respectively. Wilson et al.49 observed a similar speed-
dependent anisotropy in ozone photodissociation at 226 nm
with slightly different anisotropy parameters of 1.31Ϯ0.20
and 0.67Ϯ0.05 for the fast and slow components, respec-
tively. In order to minimize the effect of ozone photodisso-
ciation, the polarization of the probe beam was horizontal
with respect to the flight axis for all the spectra taken. There
the slow component in the BrO TOF spectra is a combination
of O atoms from the discharge and photodissociation of
ozone. Since the signal intensity depends inversely on the
speed of the photofragment under core-sampling conditions,
there is only a negligible contribution from the fastest frag-
ments arising from ozone dissociation and this does not af-
fect the fitting of BrO signal.
The total translational energy from the forward convolu-
tion fitting for Br(2P3/2)ϩO(3P2) channel is 24.7Ϯ1.0 kcal/
mol and for the Br(2P1/2)ϩO(3P2) channel is 14.0Ϯ1.0
kcal/mol. The difference is constrained to be equal to the
known Br atom spin–orbit energy of 10.5 kcal/mol.50 The
Br–O bond energy D00(Br–O) can be derived by using Eq.
͑10͒, where Ehv is the 355 nm photon energy, and assuming
that the parent internal energy is neglegible for molecular
beam source. The calculated Br–O bond energy is 55.8 kcal/
mol, which is in excellent agreement with Durie and Ram-
say’s result, 55.2 kcal/mol.8 The heat of formation of BrO at
298 K can also be derived from this bond energy. The result
is 29.7Ϯ1.0 kcal/mol, in good agreement with the spectro-
scopic results.8,18
Recent ab initio calculations of low-lying excited states
of BrO show that there are only four low-lying excited states
which have vertical excitation energy less than 4.5 eV.22
Among them, only the A 2⌸ state is a bound state correlating
to Br(2P3/2)ϩO(1D2) and has a parallel transition dipole
moment. The remaining states are all repulsive in nature cor-
relating to Br(2P3/2)ϩO(3PJ). The possibility that these
states contribute to the oscillator strength at 355 nm is mini-
mal considering the vertical excitation energies for these
states are greater than 4 eV according to theoretical calcula-
tions, which is well above the 355 nm photon energy. There-
fore, the intrinsic photofragment angular distribution should
follow the limiting case for a parallel transition with an an-
isotropy parameter of 2.0, and the observed nonlimiting an-
isotropy parameter of 1.4Ϯ1.0 indicates a significant rota-
tional depolarization. Since the 355 nm is right on top of
͑4,0͒ band of A 2⌸ X 2⌸ transition, the predissociation
lifetime of the ϭ4 level of BrO A 2⌸ state can be calcu-
v
Ј
lated using Eq. ͑12͒,51
2
2
1ϩ␻ ␶
␤
ϭ␤
,
2
͑12͒
͑
␶͒
͑
␶ϭ0͒
2
1ϩ4␻ ␶
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.70.241.163 On: Sun, 21 Dec 2014 02:21:48

4182
J. Chem. Phys., Vol. 116, No. 10, 8 March 2002
Zou, Kim, and North
ACKNOWLEDGMENTS
where ␶ is the predissociation lifetime, ␻ is the rotation an-
gular velocity, ␤(␶) is the observed anisotropy parameter,
The authors would like to thank Dr. John Bevan for his
useful comments regarding discharge beam sources. The
technical assistance of Bridget Schmitz is also gratefully ap-
preciated. Support for this project was provided by the Texas
Research Endowment Program and a Research Enhancement
Grant from Texas A&M University.
␤
is the intrinsic anisotropy parameter. For simplicity,
(␶ϭ0)
we have assumed that all the rotational sublevels have same
predissociation lifetime. The lifetime derived from Eq. ͑12͒
strongly depends on the average angular velocity, ␻, of BrO,
which is determined by the rotational distribution of the BrO
radicals. We have calculated the lifetimes for different rota-
tional temperatures. Lifetime values range from 1.1 to 0.61
ps for rotational temperatures from 5 to 15 K. Therefore, we
1 M. J. Molina and F. S. Rowland, Nature ͑London͒ 249, 810 ͑1974͒.
2 R. R. Garcia and S. Solomon, J. Geophys. Res., ͓Atmos.͔ 99, 12937
͑1994͒.
estimated that the upper limit of the ϭ4 level of BrO
v
Ј
A 2⌸ state was 1.1 ps.
3 S. Solomon, R. R. Garcia, F. S. Rowland, and D. J. Wuebbles, Nature
͑London͒ 321, 755 ͑1986͒.
Wheeler et al.23 and Wilmouth et al.18 have investigated
the predissociation lifetimes of the ͑12,0͒ and ͑7,0͒ transi-
tions of the A 2⌸ X 2⌸ band of BrO by measuring the
lifetime spectra broadening. They also reported that the ab-
sorption bands for all the vibrational levels, except ͑12,0͒
and ͑7,0͒ transitions, are quite diffuse, prohibiting a lifetime
analysis. Therefore, if perpendicular contributions to the ex-
4 P. O. Wennberg et al., Science 266, 398 ͑1994͒.
5 S. R. Langhoff, L. Jaffe, and J. O. Arnold, J. Quant. Spectrosc. Radiat.
Transf. 18, 227 ͑1977͒.
6 P. Zou, B. Schmidt, T. Nguyen, J. Park, and S. W. North, J. Phys. Chem.
͑in press͒.
7 C. M. Nelson, T. A. Moore, M. Okumura, and T. K. Minton, Chem. Phys.
207, 287 ͑1996͒.
8 R. A. Durie and D. A. Ramsay, Can. J. Phys. 36, 35 ͑1958͒.
9 J. A. Coxon and R. A. Ramsay, Can. J. Phys. 54, 1034 ͑1976͒.
10 A. Toniolo, M. Persico, and D. Pitea, J. Chem. Phys. 112, 2790 ͑2000͒.
11 W. H. Howie, I. C. Lane, S. M. Newman, D. A. Johnson, and A. J. Orr-
Ewing, Phys. Chem. Chem. Phys. 1, 3079 ͑1999͒.
12 I. C. Lane, W. H. Howie, and A. J. Orr-Ewing, Phys. Chem. Chem. Phys.
1, 3087 ͑1999͒.
citation are negligible at 355 nm, the shorter lifetime of
ϭ4 level, comparing with the results of ϭ12 and 7 levels,
calculated from our measured anisotropy parameters are con-
sistant with previous estimates of the predissociation life-
times.
v
Ј
v
Ј
13 H. F. Davis and Y. T. Lee, J. Phys. Chem. 100, 30 ͑1996͒.
14 S. Schmidt, Th. Benter, and R. N. Schindler, Chem. Phys. Lett. 282, 292
͑1998͒.
15 W. B. DeMore, S. P. Sander, D. M. Golden, R. F. Hampson, M. J. Kurylo,
C. J. Howard, A. R. Ravishankara, C. E. Kolb, and M. J. Molina, ‘‘Chemi-
cal kinetics and photochemical data for use in stratospheric modeling,’’
Evaluation Number 11, NASA JPL Publ. 1994, No. 94-26.
16 Y. Bedjanian, G. Le Bras, and G. Poulet, Chem. Phys. Lett. 266, 233
͑1997͒.
IV. CONCLUSIONS
The UV photodissociation dynamics of ClO and BrO
radicals have been studied using REMPI-TOF. We find that
the dominant channel in the photodissociation is Cl(2P3/2
)
ϩO(1D2). The anisotropy of this channel suggests that it
derives from the A 2⌸i state but that contributions from the
2 2⌺ϩ state may be nonnegligible. Channels 3 and 5 are also
observed but are much less significant, constituting
ϳ3%–4% of the products. The Cl(2P1/2)ϩO(3PJ) channel
arises from a predominately parallel transition and the wave-
length invariance of this channel’s yield suggests that it
arises from adiabatic dynamics arising from a weak interac-
tion of the repulsive 3 2⌸i state with the A-state consistent
with recent theoretical predictions.10,12 The similarity of the
results to a previous study at 248 nm suggests that the prod-
uct branching ratios and contributions of excited states are
not strongly dependent on wavelength throughout the con-
tinuum region from 220 to 260 nm.
17 M. K. Gilles, A. A. Turnipseed, J. B. Burkholder, and A. R. Ravishankara,
Chem. Phys. Lett. 272, 75 ͑1997͒.
18 D. M. Wilmouth, T. F. Hanisco, N. M. Donahue, and J. G. Anderson, J.
Phys. Chem. A 103, 8935 ͑1999͒.
19 J. J. Orlando and G. S. Tyndall, J. Phys. Chem. 100, 19398 ͑1996͒.
20 D. R. Hanson, A. R. Ravishankara, and E. R. Lovejoy, J. Geophys. Res.,
͓Atmos.͔ 101, 9063 ͑1996͒.
21 The recommended BrO heat of formation is now 30Ϯ2 kcal/mol; W. B.
DeMore, S. P. Sander, D. M. Golden, R. F. Hampson, M. J. Kurylo, C. J.
Howard, A. R. Ravishankara, C. E. Kolb, and M. J. Molina, ‘‘Chemical
kinetics and photochemical data for use in stratospheric modeling,’’
Evaluation Number 12, NASA JPL Publ. 1997, No. 97-4.
22 Y. Li, J. S. Francisco, and K. A. Peterson, J. Chem. Phys. 113, 8556
͑2000͒.
23 M. D. Wheeler, S. M. Newman, T. Ishiwata, M. Kawasaki, and A. J.
Orr-Ewing, Chem. Phys. Lett. 285, 346 ͑1998͒.
24 W. S. McGivern, R. Li, P. Zou, and S. W. North, J. Chem. Phys. 111, 5771
͑1999͒.
The photodissociation dynamics of BrO have been stud-
ied at 355 nm for the first time using REMPI-TOF. We find
that both excited state and ground state bromine atoms are
produced at this wavelength with a quantum yield of 0.83
Ϯ0.10 and 0.17Ϯ0.1, respectively. The intrinsic anisotropy
parameter of the photofragments suggest excitation to the
A 2⌸i state dominates the oscillator strength at this wave-
length. Based on the measured asymptotic velocity of both
channels we have directly determined the Br–O bond disso-
ciation energy of 55.8Ϯ1.0 kcal/mol, providing a heat of
formation for the BrO radical at 298 K of 29.7Ϯ1.0 kcal/
mol. The rotational depolarization of the photofragment an-
gular distribution allows an estimation of the upper limit of
25 W. C. Wiley and I. H. McLaren, Rev. Sci. Instrum. 26, 1150 ͑1955͒.
26 H. J. Hwang, J. Griffiths, and M. A. El-Sayed, Int. J. Mass Spectrom. Ion
Processes 131, 265 ͑1994͒.
27 R. Ogorzalek-Loo, H.-P. Haerri, G. E. Hall, and P. L. Houston, J. Chem.
Phys. 90, 4222 ͑1989͒.
28 J. A. Syage, J. Chem. Phys. 105, 1007 ͑1996͒.
29 D. T. Anderson, S. Davis, T. S. Zwier, and D. J. Nesbitt, Chem. Phys. Lett.
258, 207 ͑1996͒.
30 M. C. van Beek and J. J. ter Meulen, Chem. Phys. Lett. 337, 237 ͑2001͒.
31 M. J. Cooper, T. Diez-Rojo, L. J. Rogers, C. M. Western, M. N. R. Ash-
fold, and J. W. Hudgens, Chem. Phys. Lett. 272, 232 ͑1997͒.
32 M. M. Ahern, D. A. Steinhurst, and M. A. Smith, Chem. Phys. Lett. 300,
681 ͑1990͒.
33 K. M. Beck, K. A. H. German, and W. P. Hess, Chem. Phys. Lett. 256, 297
͑1996͒.
34 S. Arepalli, N. Presser, D. Robie, and R. J. Gordon, Chem. Phys. Lett. 118,
88 ͑1985͒.
predissociation lifetime of ϭ4 level of 1.1 ps.
v
Ј
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.70.241.163 On: Sun, 21 Dec 2014 02:21:48

J. Chem. Phys., Vol. 116, No. 10, 8 March 2002
Photodissociation of ClO and BrO
43 R. N. Zare, Mol. Photochem. 4, 1 ͑1972͒.
4183
35 D. J. Bamford, L. E. Jusinski, and W. K. Bischel, Phys. Rev. A 34, 185
͑1986͒.
44 R. N. Dixon, J. Chem. Phys. 85, 1866 ͑1986͒.
36 G. H. Cady, Inorg. Synth. 5, 156 ͑1957͒.
45 H. J. Hwang, J. Griffiths, and M. A. El-Sayed, Int. J. Mass Spectrom. Ion
Processes 131, 265 ͑1994͒.
37 Y. Tanaka, M. Kawasaki, Y. Matsumi, H. Fujiwara, T. Ishiwata, L. J.
Rogers, R. N. Dixon, and M. N. R. Ashfold, J. Chem. Phys. 109, 1315
͑1998͒.
46 W. S. McGivern, R. Li, P. Zou, T. Nguyen, and S. W. North, Chem. Phys.
249, 237 ͑1999͒.
38 J. Zhang, M. Dulligan, and C. Wittig, J. Chem. Phys. 107, 1403 ͑1997͒.
39 Y. Matsumi, P. K. Das, and M. Kawasaki, J. Chem. Phys. 97, 5261 ͑1992͒.
40 J. Cao, H.-P. Loock, and C. X. W. Qian, Can. J. Chem. 72, 758 ͑1993͒.
41 R. Iyanage, Y. Yang, S. Hasimoto, R. Gordon, and R. W. Field, J. Chem.
Phys. 103, 6811 ͑1995͒.
47 R. L. Miller, A. G. Suits, P. L. Houston, R. Toumi, J. A. Mack, and A. M.
Wodtke, Science 265, 1831 ͑1994͒.
48 J. A. Syage, J. Chem. Phys. 105, 1007 ͑1996͒.
49 R. J. Wilson, J. A. Mueller, and P. L. Houston, J. Phys. Chem. A 101, 7593
͑1997͒.
42 M. Ahmed, D. Blunt, D. Chen, and A. G. Suits, J. Chem. Phys. 106, 7617
͑1997͒.
50 R. L. Kelly, J. Phys. Chem. Ref. Data 16, Suppl. 1, 1336 ͑1987͒.
51 G. E. Busch and K. R. Wilson, J. Chem. Phys. 56, 3638 ͑1972͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.70.241.163 On: Sun, 21 Dec 2014 02:21:48