The ultraviolet photodissociation of Cl2O at 235 nm and of HOCl at 235 and 266 nm

Article abstract of DOI:10.1063/1.476682

The primary photochemistry of gas phase dichlorine monoxide (Cl₂O) and of hypochlorous acid (HOCl) following excitation at 235 nm has been investigated using photofragment ion imaging to obtain the recoil velocity and angular distributions of the ground (²P_3/2) and spin-orbit excited (²P_1/2) atomic chlorine products. In the case of Cl₂O, both Cl spin-orbit products exhibit angular distributions characterized by an anisotropy parameter, β=1.2±0.2, consistent with previous interpretations of the ultraviolet (UV) absorption spectrum of Cl₂O which associate the broad intense absorption feature peaking at λ～255nm with excitation to a (bent) dissociative state of ¹B₂(C_2v) symmetry. The recoil velocity distributions of the two Cl spin-orbit products are markedly different. The ground state atoms (which constitute >90% of the total Cl atom yield) are partnered by ClO fragments carrying significantly higher average levels of internal excitation. The slowest Cl atoms are most readily understood in terms of three body fragmentation of Cl₂O to its constituent atoms. These findings are rationalized in terms of a model potential energy surface for the 1¹B₂ state, which correlates diabatically with ClO(X) radicals together with a spin-orbit excited Cl atom, with efficient radiationless transfer to one (or more) lower energy surfaces at extended Cl-O bond lengths accounting for the dominance of ground state Cl atom fragments. The image of the ground state Cl atoms resulting from photolysis of HOCl at 235 nm is consistent with parent excitation via a transition for which the dipole moment is closely aligned with the Cl-O bond, followed by prompt dissociation (β=1.7±0.2) with the bulk of the excess energy partitioned into product recoil. Such conclusions are consistent with the results of laser induced fluorescence measurements of the OH(X) products resulting from 266 nm photodissociation of HOCl which reveal OH(X) products in both spin-orbit states, exclusively in their zero-point vibrational level, and carrying only modest levels of rotational excitation (well described by a Boltzmann distribution with T_rot～750±50K).

Full text of DOI:10.1063/1.476682

The ultraviolet photodissociation of Cl 2 O at 235 nm and of HOCl at 235 and 266 nm
Yoshiki Tanaka, Masahiro Kawasaki, Yutaka Matsumi, Hisashi Fujiwara, Takashi Ishiwata, Leon J. Rogers,
Richard N. Dixon, and Michael N. R. Ashfold
Citation: The Journal of Chemical Physics 109, 1315 (1998); doi: 10.1063/1.476682
View online: http://dx.doi.org/10.1063/1.476682
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/109/4?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Cl2O photochemistry: Ultraviolet/vis absorption spectrum temperature dependence and O(3P) quantum yield at
193 and 248 nm
J. Chem. Phys. 134, 204310 (2011); 10.1063/1.3592662
C–Cl bond fission dynamics and angular momentum recoupling in the 235 nm photodissociation of allyl chloride
J. Chem. Phys. 121, 11016 (2004); 10.1063/1.1812757
On the ultraviolet photodissociation of H 2 Te
J. Chem. Phys. 121, 9389 (2004); 10.1063/1.1799572
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
The ultraviolet photodissociation dynamics of hydrogen bromide
J. Chem. Phys. 110, 281 (1999); 10.1063/1.478063
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:
128.42.202.150 On: Mon, 24 Nov 2014 12:19:42

JOURNAL OF CHEMICAL PHYSICS
VOLUME 109, NUMBER 4
22 JULY 1998
The ultraviolet photodissociation of Cl O at 235 nm and of HOCl at 235
2
and 266 nm
Yoshiki Tanakaa) and Masahiro Kawasaki
Department of Molecular Engineering, Kyoto University, Kyoto 606-01, Japan
Yutaka Matsumi
Solar Terrestrial Environmental Laboratory, Nagoya University, Toyokawa 442, Japan
Hisashi Fujiwara and Takashi Ishiwata
Faculty of Information Science, Hiroshima City University, Hiroshima 731-31, Japan
Leon J. Rogers, Richard N. Dixon, and Michael N. R. Ashfold
School of Chemistry, University of Bristol, Bristol BS8 1TS, United Kingdom
͑
Received 19 February 1998; accepted 16 April 1998͒
The primary photochemistry of gas phase dichlorine monoxide (Cl O) and of hypochlorous acid
2
͑
HOCl͒ following excitation at 235 nm has been investigated using photofragment ion imaging to
2
obtain the recoil velocity and angular distributions of the ground ( P3/2) and spin-orbit excited
(
2
P1/2) atomic chlorine products. In the case of Cl O, both Cl spin-orbit products exhibit angular
2
distributions characterized by an anisotropy parameter, ␤ϭ1.2Ϯ0.2, consistent with previous
interpretations of the ultraviolet ͑UV͒ absorption spectrum of Cl O which associate the broad
2
intense absorption feature peaking at ␭ϳ255 nm with excitation to a ͑bent͒ dissociative state of
1
B (C ) symmetry. The recoil velocity distributions of the two Cl spin-orbit products are markedly
2
2v
different. The ground state atoms ͑which constitute Ͼ90% of the total Cl atom yield͒ are partnered
by ClO fragments carrying significantly higher average levels of internal excitation. The slowest Cl
atoms are most readily understood in terms of three body fragmentation of Cl O to its constituent
2
1
atoms. These findings are rationalized in terms of a model potential energy surface for the 1 B2
state, which correlates diabatically with ClO(X) radicals together with a spin-orbit excited Cl atom,
with efficient radiationless transfer to one ͑or more͒ lower energy surfaces at extended Cl-O bond
lengths accounting for the dominance of ground state Cl atom fragments. The image of the ground
state Cl atoms resulting from photolysis of HOCl at 235 nm is consistent with parent excitation via
a transition for which the dipole moment is closely aligned with the Cl-O bond, followed by prompt
dissociation (␤ϭ1.7Ϯ0.2) with the bulk of the excess energy partitioned into product recoil. Such
conclusions are consistent with the results of laser induced fluorescence measurements of the
OH(X) products resulting from 266 nm photodissociation of HOCl which reveal OH(X) products
in both spin-orbit states, exclusively in their zero-point vibrational level, and carrying only modest
levels of rotational excitation ͑well described by a Boltzmann distribution with T ϳ750Ϯ50 K͒.
rot
©
1998 American Institute of Physics. ͓S0021-9606͑98͒01328-2͔
4
,5,15
INTRODUCTION
Cl atoms, ClO and OH radicals, through the ClO and
‘
‘shoulder’’ at ␭ϳ300 nm.
Comparison with ab initio
10,12
theory
suggests that the dominant absorption at these
x
1
˜
1
wavelengths is associated with the 2 AЈϪX AЈ transition,
HO cycles, all play a role in the catalyzed destruction of
x
1
1
˜
1,2
but that the 1 AЉϪX AЈ perpendicular transition will con-
tribute some oscillator strength at the longer excitation wave-
lengths. Experimental measurements of the alignment of the
OH(X) fragments arising in the 248 and 266 nm photolysis
of HOCl both indicate that these fragments arise following
global stratospheric ozone. Though dichlorine monoxide,
Cl O, is currently thought to play at most a minor role in
2
stratospheric chemistry, its hydrate, hypochlorous acid
7
8
3
͑
HOCl͒, is an important chlorine reservoir species. Conse-
quently, the near UV spectroscopy and photochemistry of
HOCl has received considerable attention, both
1
prompt dissociation of an excited state of AЈ symmetry,
experimental and theoretical,10–14 since its photochemical
stability directly affects its atmospheric lifetime and thus its
importance in atmospheric chemistry. HOCl exhibits a broad
diffuse absorption at wavelengths ␭Ͻ360 nm, with a maxi-
4–9
though time-of-flight ͑TOF͒ measurements of the partner Cl
atoms formed in HOCl photodissociation at wavelengths
ϳ235 nm gave somewhat less conclusive results.9
The spectroscopy and photochemistry of Cl O have
received rather less attention. It has a bent (C ) ground
2
v
2
Ϫ19
2
Ϫ1
mum at ϳ250 nm (␴ϳ4ϫ10
cm molecule ) and a
state geometry, with r (Cl-O)ϭ1.696 Å and ЄClOCl
0
ϭ110.9°.1
6,17
Cl O exhibits a continuous absorption spec-
2
a͒On leave from Graduate School of Environmental Science, Hokkaido Uni-
versity, Sapporo 060, Japan.
trum, with maxima at ϳ530, 420, 255 and 170 nm, beyond
0021-9606/98/109(4)/1315/9/$15.00
1315
© 1998 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:
28.42.202.150 On: Mon, 24 Nov 2014 12:19:42
1

1316
J. Chem. Phys., Vol. 109, No. 4, 22 July 1998
Tanaka et al.
which comes the first Rydberg origin, at 162.9 nm.4,15,18–22
ization probability f(Cl)/f(Cl )ϭ0.85Ϯ0.1 as reported by
*
Liyanage et al.38 A few similar one color experiments
Cl O photolysis finds use as a convenient source of ClO
2
3
radicals and Cl atoms in laboratory experiments relevant to
atmospheric chemistry, but detailed, quantum state resolved
studies of its primary photochemistry and the accompanying
photodissociation dynamics remain rare. Early studies of
searching for O( P ) atom products from Cl O photolysis
were performed also; the requisite wavelength in this case
was 225.65 nm, appropriate for 2ϩ1 REMPI involving the
2
2
3
1
3
4
3
39
O(2p 3p ; P ←2p ; P ) transition.
2
2
2
3–27
21,28
40
Cl O photolysis,
2
reviewed elsewhere,
showed reac-
Cl O was prepared by the method of Cady, in which a
2
tion ͑1͒ to be the dominant fragmentation pathway at all near
UV excitation wavelengths but also suggested that one or
more of channels ͑2͒–͑4͒ must be responsible for the minor
O atom yield identified when photolyzing at the shorter UV
wavelengths,
low pressure of Cl gas (Ͻ50 Torr) was flowed slowly over
2
dried HgO dispersed on glass beads held at room tempera-
ture. The effluent gas was collected in a trap maintained at
liquid N temperature. Residual Cl in the condensed product
2
2
was removed by vacuum distillation at Ϫ112 °C. The purity
2
of the final Cl O sample was verified by UV absorption
2
Cl Oϩh␯→Cl͑ P ͒ϩClO͑X͒ ␭thresholdϽ840 nm ͑1͒
2
J
spectroscopy.15 HOCl was prepared by mixing Cl O ͑5 Torr͒
2
2
3
→
→
→
2Cl͑ P ͒ϩO͑ P ͒ ␭thresholdϽ292 nm ͑2͒
with H O ͑13 Torr͒ and leaving to equilibrate. Given a value
J
J
2
31,41
0
.082Ϯ0.010 for the equilibrium constant K5 ,
3
Cl ϩO͑ P ͒ ␭thresholdϽ710 nm
͑3͒
͑4͒
2
J
1
Cl ϩO͑ D͒ ␭thresholdϽ335 nm.
H OϩCl Oꢀ2HOCl,
͑5͒
2
2
2
Note that the long wavelength limits quoted correspond to
thermodynamic thresholds,2
9–31
we estimate that the HOCl partial pressure in this mixture is
and make no allowance for
ϳ1.5 Torr. Both the Cl O sample and the HOCl ͑as the
any possible energy barriers in these various dissociation
channels. More recently, Okumura, Minton and
2
HOCl/Cl O/H O equilibrium mixture͒ were seeded in Ar or
2
2
2
8,32
Kr ͑typical mixing ratio ϳ2%͒ and introduced into the ion
imaging spectrometer in the form of a pulsed molecular
beam with the backing pressure held at ϳ600 Torr. The
measured images were insensitive to carrier gas, encouraging
us to conclude that there is negligible clustering in the mo-
lecular beam. It was noticed that the signal level would de-
teriorate with time—an effect attributable to surface cata-
lyzed decomposition in the delivery line leading to the
pulsed valve. To counteract this, a flow system was used to
ensure that the sample mixture immediately behind the valve
was continually being refreshed.
co-workers
have studied the photolysis of jet-cooled
Cl O molecules at three excimer laser wavelengths ͑193, 248
2
and 308 nm͒, using traditional photofragment translational
spectroscopy techniques with mass spectrometric detection
ϩ
ϩ
ϩ
of products with m/z 16 (O ), 35 (Cl ), 51 (ClO ) and 70
ϩ
(
Cl ). They confirm Cl-O bond fission ͓channel ͑1͔͒ as the
2
sole primary fragmentation pathway following excitation at
08 nm. 248 nm photolysis results in ClO fragments with a
3
broad, apparently bimodal, spread of internal energies, and
some three body dissociation. This latter atomization process
͑
2͒ is found to be dominant when Cl O is photolyzed at 193
2
2
8
Two different ion imaging apparatus, one in Kyoto, the
other in Bristol, were used in the present experiments. The
Kyoto apparatus, which follows the original design of Chan-
dler and Houston,33 has been described in detail elsewhere.
nm; the authors suggest that these products arise via the
concerted eliminations ͑3͒ and ͑4͒ and subsequent fragmen-
tation of the highly vibrationally excited Cl products.
2
42
Here we report state-specific and angle and velocity re-
solved measurements of the Cl atomic products resulting
Essential details of the Bristol apparatus have also been re-
ported previously.3
7,43
It incorporates one innovative feature
from photolysis of jet-cooled HOCl and Cl O molecules at
2
designed to improve the image fidelity—namely, the use of
software to establish, on a shot by shot basis, which of the
individual ion events in the image evident on the phosphor
screen mounted behind the microchannel plates ͑MCPs͒
mounted at the end of the TOF tube satisfy user selected
thresholds for minimum size and intensity, and to determine
their centroids. The final image is built up by accumulating
these centroids, while the laser frequency is scanned repeat-
edly across the Doppler profile of the relevant Cl/Cl* transi-
tion. As in all ion imaging experiments, the three-
dimensional ͑3D͒ fragment velocity distribution is detected
as a two-dimensional ͑2D͒ projection. Given the assumption
of axial symmetry around the direction of the electric vector
wavelengths around 235 nm using the technique of photo-
fragment ion imaging, first introduced by Chandler and
3
3,34
Houston.
The technique provides a direct two-
dimensional ͑2D͒ visualization of the laboratory-frame pho-
tofragment velocity distributions and, as such, provides a
particularly powerful route to elucidating the full 3D center-
of-mass frame photofragment velocity distributions.
EXPERIMENT
A single excimer pumped dye laser, the output of which
was frequency doubled in BBO, was used to generate the
necessary radiation at wavelengths ϳ235 nm required both
for photodissociating the Cl O and HOCl precursors and to
effect the two-photon resonance enhanced multiphoton ion-
of the dissociation laser radiation, we employ a filtered back-
2
projection method4
2,44
to extract the required 3D velocity
ization ͑REMPI͒ of the Cl atom photofragments. As
distribution from the observed 2D images. The velocity ͑ra-
dial͒ scale of the resulting images was calibrated by imaging
the Cl fragments arising in the well-documented 235.336 nm
photolysis of ICl.37 Figure 1͑d͒ shows the equatorial slice
through the reconstructed 3D image of these Cl atom frag-
3
5–37
2
previously,
ground Cl( P3/2) and spin-orbit excited
2
Cl*( P1/2) atomic photoproducts were probed using wave-
lengths of 235.336 and 235.205 nm, respectively. Cl/Cl*
branching ratios were estimated by adopting the relative ion-
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:
128.42.202.150 On: Mon, 24 Nov 2014 12:19:42

J. Chem. Phys., Vol. 109, No. 4, 22 July 1998
Tanaka et al.
1317
ϩ
FIG. 1. Cl ion images, displayed as equatorial slices through the recon-
structed 3D distributions, from the photodissociation of Cl O ͑a͒ and ͑b͒,
2
HOCl ͑c͒ and ICl ͑d͒ at 235 nm. The detected species ͑Cl or Cl*͒ is indi-
cated on each image. In each case the E vector of the photolysis laser
radiation is aligned vertically, as shown by the arrow. Note that because of
the limited available grey scale it has proved convenient to display the
ϩ
strongest image ͑a͒ with deceptively faint intensity, and that the Cl signal
in Fig. 1͑c͒ contains contributions from Cl atoms arising from photolysis of
both Cl O and HOCl.
2
FIG. 2. Center-of-mass translational energy distribution, P(E ), of ͑a͒ the
T
2
2
ground state Cl( P3 ) products and ͑b͒ the spin-orbit excited Cl*( P1/2)
/2
ments; the width of this image ͑ϳ60 kJ molϪ1 ͓full width at
products arising in the 235 nm photodissociation of Cl O. The maximum
fragment kinetic energies allowed by energy conservation following disso-
ciation via channels ͑1͒ and ͑2͒ are indicated.
2
half maximum ͑FWHM͔͒ at a Cl fragment translational en-
Ϫ1
ergy E ϭ213 kJ mol ͒ gives a measure of the prevailing
T
instrumental resolution. Note that the image radius actually
scales linearly with velocity, so the energy resolution should
branching ratio of 13Ϯ2. The Cl and Cl* photofragment
be E dependent. Thus, for example, for Cl fragments with
T
images shown in Figs. 1͑a͒ and 1͑b͒ are the equatorial slices
through the 3D velocity distributions that have been ex-
tracted from the raw 2D projections, after symmetrizing, us-
ing the filtered back-projection method. The Cl and Cl* im-
ages show essentially identical angular distributions relative
to the direction of the E vector of the photolysis laser radia-
tion, but careful inspection shows their relative sizes ͑i.e., the
Cl and Cl* recoil velocity distributions͒ to be different; the
ground state Cl atoms are formed with less average kinetic
energy.
Ϫ1
E ϭ100 kJ mol ͑closer to the mean Cl atom kinetic en-
T
ergy arising in the 235 nm photolysis of Cl O and HOCl͒,
2
Ϫ1
the FWHM energy resolution will be ϳ40 kJ mol . For
consistency, all of the images shown in this work were re-
corded in Kyoto, but equivalent results for Cl O were ob-
2
tained in Bristol also.
OH(X) photofragments resulting from the linearly polar-
ized 266 nm photodissociation of HOCl ͑in a slowly flowing
HOCl/Cl O/H O equilibrated mixture, at room temperature
2
2
and a total pressure of 0.1–0.2 Torr͒ were detected by laser
induced fluorescence ͑LIF͒ on the AϪX ͑0,0͒ band at 308
nm using apparatus in Hiroshima. Full details of the experi-
mental procedure and methods of data analysis are reported
Center-of-mass ͑CM͒ translational energy distributions,
P(E ), for the Cl and Cl* atom products are shown in Fig.
T
2
. Also marked on these plots are the maximum possible
kinetic energy releases for chlorine atoms arising via chan-
4
5
elsewhere.
nels ͑1͒ and ͑2͒ given the photon energy and the respective
2
9,30
bond dissociation energies D (Cl-OCl)ϭ141
and
0
Ϫ1 46
RESULTS
D0(Cl-O)ϭ265 kJ mol
.
In constructing these, we as-
sume the counter fragment to have a mass appropriate for
ClO. We recognize the inaccuracy of such a procedure for all
Cl fragments that arise via the three body fragmentation pro-
cess ͑2͒ and thus caution against reading too much into the
Photodissociation of Cl O at 235 nm
2
Cl and Cl* atomic fragments produced from Cl O pho-
2
tolysis at wavelengths ϳ235 nm were detected by 2ϩ1
REMPI. The measured Cl signal intensity ͑at 235.336 nm͒
was about 15 times that for the Cl* peak ͑measured at
detailed form of the lowest energy part of these P(E ) dis-
T
tributions. Clearly, however, and notwithstanding the limited
energy resolution of the experiment, the ClO fragments part-
nering the ground state Cl atoms are formed with significant
͑and a broad spread of͒ internal excitation. The Cl* atom
235.205 nm͒ for the same incident laser pulse energy which,
when allowing for the different ionization detection efficien-
38
cies for the two spin-orbit states, leads to a ͓Cl͔/͓Cl*͔
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:
128.42.202.150 On: Mon, 24 Nov 2014 12:19:42

1318
J. Chem. Phys., Vol. 109, No. 4, 22 July 1998
Tanaka et al.
FIG. 3. Comparison of product translational energy distributions resulting
from Cl O photolysis at 248 nm ͑solid line, obtained by monitoring ions
2
with m/z 51, after Ref. 28͒ and at 235 nm ͑by imaging the Cl atom prod-
ucts͒. The shaded area represents the Cl* yield, while the filled points indi-
cate the total Cl atom yield obtained by summing the deduced P(E ) dis-
T
tributions for Cl and Cl* atoms, weighted in the ratio 13:1. For the purposes
of this comparison the horizontal scale is (h␯ϪE ). Note that much of the
T
present signal appearing at energies in excess of D (Cl—O)—indicated by
0
the vertical arrow—is attributed to the three body dissociation ͑2͒; any cor-
responding products arising in the 248 nm photolysis of Cl O will not ap-
2
pear in the solid line distribution because of the detection method used
͑Refs. 28 and 32͒.
products exhibit an even broader E distribution ͓Fig. 2͑b͔͒
but, relative to the Cl products, the center of gravity of the
Cl* distribution falls at much higher translational energies
T
FIG. 4. Angular distributions of ͑a͒ Cl and ͑b͒ Cl* atom yield ͑integrated
over all recoil velocities͒ resulting from 235 nm photolysis of Cl O. The
2
solid lines are best-fits to the data in terms of Eq. ͑6͒ with ␤ϭ1.2 and ␤
ϭ1.1, respectively.
͑
i.e., lower ClO internal energies͒.
Okumura, Minton and co-workers2
8,32
have reported
P(E ) spectra, obtained by conventional TOF methods fol-
T
lowed by electron bombardment and mass spectrometric de-
tection, for the ClϩClO products resulting from 248 nm pho-
products arising via channel ͑2͒. We return to further discus-
sion of these points, and of the likely nature of the ClO
internal excitation, below.
tolysis of Cl O. These they have chosen to fit, arbitrarily, in
2
terms of three Gaussian distributions centered at
͗
ET
͘
ϭ74,
Figure 4 shows plots of the angular distributions of the
Cl and Cl* products arising in the 235 nm photolysis of
Ϫ1
150 and 214 kJ mol , respectively. Such mass spectromet-
ric measurements are, of necessity, non-spin-orbit state se-
Cl O. The angular distribution of the photofragment velocity
vector v is given by
2
lective, though we note that Chichinin47 has reported that
48
18%Ϯ2% of the total Cl atom yield arising in the 248 nm
I͑v,␪͒ϭf͑v͒.
͕
1ϩ␤P ͑cos ␪͒
͖
/4␲,
͑6͒
2
photolysis of Cl O is in the spin-orbit excited Cl* state.
2
Clearly, we can compare this earlier P(E ) data with an
where f(v) is the speed distribution of the fragment, ␪ is the
angle between the polarization vector of the photolysis laser
radiation and the fragment recoil velocity vector, ␤ is the
T
appropriately weighted sum of the distributions shown in
Figs. 2͑a͒ and 2͑b͒. Figure 3 shows such a comparison,
where we have chosen to use as the horizontal scale (h␯
ϪE ), i.e., to offset the experimentally derived P(E ) dis-
anisotropy parameter and P (cos ␪) is the second order Leg-
2
endre polynomial. From Fig. 4 we obtain best-fit values ␤
ϭ1.2Ϯ0.1 for both the total Cl and the total Cl* product flux
distributions, with no discernible dependence upon v. Such a
finding contradicts the results of the earlier 248 nm photoly-
sis study by Moore et al.,32 who deduced the following best-
fit ␤ parameters for the three assumed subgroups within their
T
T
tributions by an amount equal to the difference in the 235
and 248 nm photon energies and then align the spectra in
terms of the ClO internal energy. The distribution obtained at
248 nm shows some similarities with the present data ͑after
summing over both spin-orbit states͒, despite the difference
in photolysis wavelength; in both cases the distribution ap-
pears bimodal and the ClO fragments are deduced to be
formed in a wide spread of internal energy states, with little
probability of being formed internally ‘‘cold.’’ We note the
seemingly very different relative importance of the ‘‘slow’’
component in the two distributions but this is, in part, an
artefact since the 248 nm distribution was generated by
P(E ) distribution: fast, ␤ϭ0.7Ϯ0.2; intermediate, ␤ϭ1.5
T
Ϯ0.2; and slow, ␤ϭ1.2Ϯ0.2. We do not believe that this
apparent discrepancy can be attributed to the low energy
resolution of the present imaging apparatus since we can
easily distinguish between overlapping Cl angular distribu-
tions with ␤ϭ1.2 and 1.7 ͑from 235 nm photolysis of Cl O
2
and HOCl, respectively͒—see below. The 225.65 nm pho-
ϩ
transforming TOF data recorded on the m/z 51 (ClO ) mass
tolysis of Cl O was also studied briefly since this is a suitable
2
3
channel only, and thus includes no contribution from slow
wavelength for inducing 2ϩ1 REMPI of any O( P ) pho-
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:
128.42.202.150 On: Mon, 24 Nov 2014 12:19:42

J. Chem. Phys., Vol. 109, No. 4, 22 July 1998
Tanaka et al.
1319
FIG. 5. Kinetic energy distribution ͓solid points, labeled ͑b͔͒ of the ground
FIG. 6. Angular distribution of that fraction of the Cl atom product yield
state Cl atoms resulting from 235 nm photolysis of an HOCl/Cl O/H O
Ϫ1
2
2
with KEϾ75 kJ mol
HOCl/Cl O/H O mixture ͑open circles͒. The solid line shows a best-fit to
resulting from 235 nm photolysis of the
mixture derived from the image shown in Fig. 1͑c͒. The crosses show the
corresponding Cl atom kinetic energy distribution obtained using a pure
2
2
the experimental data in terms of Eq. ͑6͒ with ␤ϭ1.7.
Cl O sample ͓labeled ͑a͔͒, arbitrarily scaled ͑vertically͒ so as to account for
2
Ϫ1
most of the signal at KEϳ45 kJ mol
.
fragment as found experimentally in the case of HOCl pho-
7
8
tolysis at 248 and 266 nm ͑and see below͒, and predicted
tofragments that may result. Signal levels were low, in part
because of the weakness of the parent absorption at this
wavelength. The resulting O atom image was small ͑indicat-
ing little translational energy release͒ and, notwithstanding
the poor signal to noise ratios, showed little obvious anisot-
ropy.
theoretically.14
The angular distribution of the Cl fragments resulting
from 235 nm photolysis of HOCl, shown in Fig. 6, is well
described by an anisotropy parameter ␤ϭ1.7Ϯ0.2, signifi-
cantly higher than for the Cl atoms from Cl O photolysis at
2
this wavelength. Note that, even though this plot shows the
angular dependence of just the faster part of the entire Cl
Ϫ1
Photodissociation of HOCl at 235 nm
atom signal ͑KEϾ75 kJ mol ), this value is surely a lower
limit given the presumed contribution of some Cl signal from
As shown in Fig. 1͑c͒, an additional higher velocity,
parallel component is evident in the ground state Cl atom
image observed from the 235 nm photolysis of the equilib-
rium mixture of HOCl, Cl O and H O. This we associate
the photolysis of Cl O ͑see Fig. 5͒. This high value is con-
2
10,12
1
sistent with theoretical expectations
ϪX AЈ parallel transition is the dominant contributor to the
that the 2 AЈ
˜
1
2
2
near UV absorption of HOCl at these wavelengths, and ac-
with Cl atoms resulting from HOCl photolysis,
9
cords with the suggestions of Bell et al., who observed the
2
HOClϩh␯→Cl͑ P ͒ϩOH͑X͒,
J
Doppler profile of Cl atoms formed in the 235 nm photolysis
of HOCl and concluded that the dissociation process must be
␭thresholdϽ518 nm ͑Ref. 49͒,
͑7͒
1
occurring on the steeply repulsive 2 AЈ potential energy sur-
while the inner component is attributable to photolysis of
Cl O. Since the Cl atoms from these two different precursors
2
face, resulting in fast Cl atoms with a relatively narrow ki-
netic energy distribution.
have different counter fragments we choose to plot the dis-
tribution of Cl atom kinetic energies ͓not the center-of-mass
Photodissociation of HOCl at 266 nm
P(E ) distribution͔ in Fig. 5. The previously established ki-
T
netic energy distribution of the ground state Cl atoms result-
In parallel experiments, described more fully
elsewhere,4 the internal state population distribution within
the OH(X ⌸) fragments resulting from 266 nm photolysis
5
ing from 235 nm photolysis of Cl O ͓Fig. 2͑a͔͒ is replotted
2
2
on the appropriate energy scale ͓crosses, curve marked ͑a͔͒;
its vertical scaling has been chosen arbitrarily so as to ac-
of HOCl ͑in the HOCl/Cl O/H O equilibrium mixture͒, and
2
2
Ϫ1
count for most of the shoulder apparent at KEϳ45 kJ mol
.
their recoil velocity distributions, were investigated via mea-
surement and analysis of the OH(A–X) LIF excitation spec-
trum at ϳ308 nm. Only OH(X)vϭ0 fragments are detected;
no indication of any vibrationally excited OH(X) products
could be found. Figure 7 shows the OH(X)vϭ0 rotational
state population distributions, plotted in the form of Boltz-
mann plots, for each spin-orbit state ͑⍀ϭ3/2 and 1/2͒. The
individual data points were obtained by measuring the inten-
sities of unblended P or R lines ͑iϭ1, 2 for the ⍀ϭ3/2 and
The profile mapped out by the data points labeled ͑b͒ is taken
to be the translational energy distribution of the ground state
Cl atoms resulting from 235 nm photolysis of HOCl, with an
underlying contribution from the inevitable Cl O impurity. It
2
Ϫ1
peaks at 90Ϯ10 kJ mol , with an energy width of
ϳ40 kJ mol ͑FWHM͒. As mentioned previously, an en-
Ϫ1
ergy spread of this magnitude is largely attributable to the
limited resolution of the ion imaging experiment. The corre-
sponding center-of-mass translational energy of the OHϩCl
i
i
1/2 components, respectively͒, correcting for any variations
in the photolysis and/or probe laser intensities, and dividing
by the appropriate rotational line strength as listed by Dieke
and Crosley.50 As found previously, the OH(X)vϭ0 rota-
Ϫ1
fragments (280Ϯ40 kJ mol ) accounts for all of the photon
energy over and above that required to break the Cl—O
bond, implying little internal excitation in the partner OH(X)
7,8
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:
128.42.202.150 On: Mon, 24 Nov 2014 12:19:42

1320
J. Chem. Phys., Vol. 109, No. 4, 22 July 1998
Tanaka et al.
assigned the absorption maxima evident at ϳ255 and
˜
1
ϳ170 nm in terms of transitions from the X A ground state
1
1
1
to the 1 B state and the 2 B state, respectively; the shoul-
2
1
der on the long wavelength side of the intense 255 nm peak
1
˜
1
is most likely due to the 1 B ←X A transition. Table I
1
1
lists relevant configurations and the resulting states that can
derive therefrom in the limit of linear (Dϱh) and bent (C2v)
geometries. All of the calculated transitions can be under-
stood in terms of electron promotion from one of the higher
occupied nonbonding or weakly bonding orbitals in the
ground state configuration to orbitals that are antibonding
with respect to Cl—O stretching. The present experimental
findings support the theoretical predictions. The transition
1
1
FIG. 7. Semilogarithmic plot of rotational state populations versus rota-
tional energy for the OH(X, ⍀ϭ1/2͑ϩ͒ and 3/2(᭹), vϭ0͒ fragments re-
sulting from HOCl photodissociation at 266 nm. The solid lines show best-
fits to the data, and correspond to Boltzmann temperatures of 800 (⍀
ϭ3/2) and 770 K (⍀ϭ1/2).
moment associated with a B ← A is y-polarized in the
2
1
molecular frame, i.e., parallel to a line connecting the two Cl
1
atoms. If the equilibrium bond angle in the B excited state
2
is the same as in the ground state ͑110.9°͒, then the limiting
␤
value in the case of an instantaneous dissociation would be
given by ␤ϭ2P (cos ␹), where ␹ϭ(180–110.9)°/2 is the
2
tional state population distributions are well represented by
Boltzmann distributions with ‘‘temperatures’’ of 800Ϯ70 K
angle between ␮ and the recoil velocity vector, v. This
yields a value ␤ϭ1.0, somewhat smaller than that observed
experimentally for both the Cl and Cl* product forming
͑for the ⍀ϭ3/2 component͒ and 770Ϯ50 K (⍀ϭ1/2), with
the two spin-orbit components populated in the ratio ⍀
ϭ3/2/⍀ϭ1/2)ϳ2:1. Clearly, as at 235 nm, the OH(X) frag-
channels. Thus we conclude that Cl O photolysis at 235 nm
2
1
1
does indeed proceed via a B ← A excitation followed by
2
1
ments resulting from 266 nm photolysis of HOCl are inter-
prompt fragmentation on a potential energy surface ͑or sur-
faces͒, the topology of which tends to encourage some small
opening of the ЄClOCl bond angle ͑reducing ␹ from 34.5 to
ϳ31° would result in ␤ϭ1.2͒.
Ϫ1
EINT
͗ ͘
nally ‘‘cold,’’ with ϳ550 cm only. Analysis of the
Doppler profiles of, for example, the P (5) and Q (4) tran-
1
1
sitions within the OH(A–X) ͑0,0͒ band indicate that the OH
fragments resulting from the 266 nm photodissociation of
From Figs. 2 and 3 we see that less than half of the
Ϫ1
HOCl are formed with high recoil velocity (4100 m s ) and
excess energy ͓h␯ϪD (Cl—OCl)͔ is released into product
0
an
angular
anisotropy
well
described
by
translational energy. The lowest excited state of ClO ob-
4
5
␤
ϭ1.3Ϯ0.1 —broadly paralleling observations at 235 nm
2
served
(
experimentally,
the
A ⌸
state
1
˜
1
46
Teϭ31749.6 cmϪ1), lies too high in energy to be formed
and suggesting that the 2 AЈϪX AЈ transition still domi-
nates the parent absorption at 266 nm.
as a primary product in the 235 nm photolysis of ground
state Cl O molecules. This is the only excited state of ClO
2
DISCUSSION
that can arise from the ␲*←␲ electron promotion, but
51
Langhoff has predicted the presence of a weakly bound
Photodissociation of Cl O at 235 nm
2
4
Ϫ
Ϫ1
⌺
state of ClO with T ϳ18 000 cm , arising from a
e
The key observations requiring explanation are the an-
gular anisotropy of the recoiling Cl fragments, the nature of
the energy disposal amongst the fragments, and the sensitiv-
ity of this energy disposal to the spin-orbit state of the Cl
atom. Nickolaisen et al.21 have calculated vertical excitation
␴*←␲* electron promotion. The present P(E ) spectra
T
͓Figs. 2͑a͒ and 2͑b͔͒ show no obvious indication for any
involvement of such a ClO product state and, henceforth,
electronic excitation of the ClO fragment ͑other than the
ground state spin-orbit splitting͒ is not considered further as
a possible sink for that fraction of the excess energy not
appearing in the form of product translation.
˜
1
energies and oscillator strengths from the X A ground state
1
3
͑
and the lowest energy 1 B1 excited state͒ of Cl O to a
2
number of excited singlet and triplet states and, as a result,
Dissociation from a bent geometry will induce a large
TABLE I. Configurations and resulting terms relevant for discussion of the valence states of Cl O in the limit
2
of linear (Dϱh) and bent (C2v) geometries. The final column provides a complete listing of all spin-orbit
components in C2v symmetry relevant for constructing the correlation diagram shown in Fig. 9. Singlet com-
ponents are shown in bold, while the superscripts are appropriate to the case of reduced (C ) symmetry.
s
Configuration
Configuration
(
D
)
Terms (Dϱh)
(C
)
Terms (C2v)
Components (C2v)
ϱh
2v
4
4
u
1
ϩ
2
2
2
2
1
1
.
..␲ ␲
..␲ ␲ ␴
⌺g
...a b a b
A1
aЈ
g
1
2
2
1
4
g
3
u
1
1⌸u ; 3⌸u
1⌸g ; 3⌸g
2
2
2
1
1
1
1
1
B1 ; 3B1
.
.
...a1b2a2b1a1
a1Ј ,a2Љ,b1Љ ,b2
Ј
g
1
2
2
2
1
1
1
A1 ; 3A1
A2 ; 3A2
.
..a b a b a
1 2 2 1
...a1b2a2b1a1
aЈ ,aЉ ,bЉ ,bЈ
1
2
1
2
3
g
4
u
1
g
2
2
1
2
..␲ ␲ ␴
a1Ј ,a2Љ ,b1Љ ,b2
Ј
2
1
2
2
1
1
1B2 ; 3B2
.
..a b a b a
1 2 2 1
aЈ ,aЉ ,bЉ ,bЈ
1
2
1
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:
28.42.202.150 On: Mon, 24 Nov 2014 12:19:42
1

J. Chem. Phys., Vol. 109, No. 4, 22 July 1998
Tanaka et al.
1321
torque on the ClO fragment; a simple impact parameter
model based on the ground state geometry and assuming
Ϫ1
͗
͘
ET ϭ150 kJ mol would predict that the most populated
state of the ClO partner would have rotational quantum num-
ber Nm.p.ϳ80 and a mean rotational energy EROT
͗
͘
Ϫ1
Ϫ1
ϳ4000 cm (48 kJ mol ). This is sufficient to account for
some, but by no means all, of the observed spread of internal
energies in the ClO fragment. Neither does it provide any
obvious explanation for the complete absence of fast (E
Ͼ250 kJ mol ) products in the channel leading to ground
T
Ϫ1
state Cl atoms. However, the realization that the ClO(X)
fragments are likely to be formed with significant rotational
excitation does provide a possible explanation for some of
the chlorine atom product flux seen at low E , at energies
T
which would require the partner ClO fragment to carry
32
EINTϾD (Cl—O). Moore et al. have calculated that the
0
centrifugal barrier to dissociation of highly rotational excited
Ϫ1
ClO products could be as high as 30 kJ mol , though, as
1
FIG. 8. Schematic potential energy surface for the dissociative 1 B excited
2
Fig. 2 shows, even a barrier of this size is not sufficient to
account for all of the slow Cl atoms in terms solely of chan-
nel ͑1͒. As shown below, the slowest Cl atoms arise from the
three body dissociation ͑2͒.
Product vibration is the other contributor to the observed
high levels of ClO internal excitation. Vibrational excitation
of the ClO product is to be expected, simply on Franck–
Condon grounds, given that the Cl—O equilibrium bond
state of Cl O. The filled circle represents the vertical Franck–Condon region
2
˜
starting from Cl2O(X), vϭ0, while the ellipse indicates the portion relevant
to excitation at 235 nm. Representative trajectories leading to Cl atoms,
together with ClO(X) fragments with little ͑A͒ and extensive ͑B͒ vibrational
excitation are shown, as is one leading to the three atomic products ͑C͒. It is
assumed that there are many surface crossings or near crossings in the
shaded region at long range which can mediate spin-orbit relaxation ͑see the
text and Fig. 9͒.
length in Cl O is longer than that of the ground state ClO
2
radical by 0.126 Å. We suggest that ClO product vibration is
͑
concerted͒ three body dissociations ͑C͒. In this picture, dis-
sociation channel ͑2͒ is just a natural extension of channel
1͒, thus explaining the observation that all Cl ͑and Cl*͒
1
further enhanced by the nature of the B excited state po-
2
tential energy surface and, in what follows, we present a
plausible explanation for the observed high levels of product
vibrational excitation, the proposed involvement of the three
body fragmentation process ͑2͒, and the strikingly different
energy disposals in the product channels leading to ground
͑
velocity subgroups show the same angular anisotropy. ͓Such
would be unlikely if, as suggested previously,28 any signifi-
cant fraction of the observed Cl atom yield arose from sec-
ondary photolysis of the primary ClO(X) fragments.͔ The
relative probabilities of branching into channels ͑1͒ and ͑2͒,
͑
Cl͒ and spin-orbit excited (Cl*) atoms.
The rationale begins with a schematic potential energy
and into the various ClO(X) product states associated with
v
1
surface for the excited 1 B2 state of Cl O, shown in the
channel ͑1͒, will be determined by the detailed topology of
the potential energy surface. We note that the present experi-
ments suggest a much larger relative yield of slow Cl atoms
2
form of a contour plot, as a function of the two Cl—O bond
lengths, in Fig. 8. For simplicity, we assume a constant bond
angle similar to that of the ground state ͑110.9°͒—an as-
sumption supported by the foregoing analysis of the mea-
sured ␤ parameter for forming Cl and Cl* products. The
filled circle indicates the Franck–Condon region for excita-
tion from the zero-point level of the ground state, while the
shaded ellipse indicates the portion relevant for
␭photϭ235 nm. Our positioning of the ellipse ͑which repre-
sents the starting point for trajectories or wave packet propa-
gations representing the dissociation process of current inter-
arising from the three body dissociation ͑2͒ than do the ear-
lier measurements at 248 nm.2
8,32
Such is consistent with the
above picture in as much that the excess energy provided by
a 248 nm photon is less, the ellipse representing the range of
starting configurations will be less high up on the steeply
repulsive wall, the dissociating trajectories will start with
less momentum along the diagonal and be more prone to
‘‘veer off’’ into one of the Cl—OCl exit channels.
Finally, we turn to consider the observation that the en-
ergy disposal in the ClO fragment is sensitive to the spin-
orbit state of the partner Cl atom: the P(E ) distributions in
Figs. 2͑a͒ and 2͑b͒ show that internally ‘‘cold’’ ClO(X) frag-
ments are formed exclusively with spin-orbit excited (Cl*)
atoms. Cl atom spin-orbit selectivity has also been observed
in the UV photodissociation of Cl CO, ClNO and CCl4.
In these cases, as here, the spin-orbit selectivity is most
1
˜
1
est͒ is guided by the knowledge that the 1 B ←X A
2
1
transition involves electron promotion to a Cl—O antibond-
ing orbital and that the absorption maximum lies well to the
red of the 235 nm photolysis wavelength. Clearly, trajecto-
ries starting from the ellipse on this model surface have suf-
ficient energy to be able to sample all regions of configura-
tion space involving extension of one or other or both of the
Cl—O bonds. Thus we should expect a complete spectrum
of energy disposals, ranging from ClϩClO fragments with
little vibrational energy ͑illustrated by trajectory A͒, through
those with substantial product vibration ͑B͒, to the direct
T
52
53
2
readily explained in terms of one ͑or more͒ surface crossings
in a selected area of configuration space. In the case of Cl O,
the correlation diagram shown in Fig. 9 indicates that the
1 B excited state is actually the first excited state of Cl O
2
1
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:
128.42.202.150 On: Mon, 24 Nov 2014 12:19:42

1322
J. Chem. Phys., Vol. 109, No. 4, 22 July 1998
Tanaka et al.
ever, the vast majority of trajectories ͑e.g., B and C͒ will
pass into the region of crossings; if the nonadiabatic coupling
is strong then most will cross to lower surface͑s͒ and evolve
to ground state Cl atoms and, given that all such trajectories
involve some elongation of both Cl—O bonds, vibrationally
excited ClO(X) fragments. Such a picture could also accom-
modate the observation of a larger Cl*/Cl product branching
ratio when Cl O is photolyzed at the longer wavelength of
2
2
48 nm47 in as much that, again, in the language of Fig. 8,
photoexcitation will place the starting configuration point for
the trajectories nearer the saddle point, the subsequent mo-
mentum along the diagonal will be less and thus a larger
fraction of the dissociating molecules may evolve along the
exit valleys leading to Cl*ϩClO without encountering the
region of the ‘‘seam.’’ Clearly, further studies of the wave-
length dependence of the Cl*/Cl branching ratio should
prove fruitful as, too, would measurements of the quantum
state population distributions in the ClO(X) partner.
Photodissociation of HOCl at 235 and 266 nm
The measured anisotropy parameters for the Cl atoms
resulting from 235 nm photolysis of HOCl (␤ϭ1.7Ϯ0.2),
and that deduced from analysis of the Doppler line shapes of
the OH(X) fragments resulting from photolysis at 266 nm
(␤ϭ1.3Ϯ0.1),45 are both consistent with previous
FIG. 9. Diagram showing the parent (Cl O)–product (ClϩClO) correla-
2
tions for bent (C ) geometries. The energetic ordering of the electronic
10,11,14
s
assignments
of the intense UV absorption of HOCl
states of Cl O is taken from Ref. 21. For clarity of display, the parent and
2
1
1
˜
centered at ϳ250 nm in terms of the 2 AЈϪX AЈ transition
arising as a result of electron promotion from an in-plane
nonbonding p orbital on the Cl atom to a ␴*Cl—O antibonding
orbital. The measured ␤ values are less than the limiting
value of ϩ2. This could indicate that the transition moment
is aligned some 18° off the axis of the breaking Cl—O bond
or may reflect an underlying contribution to the overall ab-
product sides of the diagram are plotted on different energy scales. States of
aЈ and aЉ symmetry ͑in C ͒ are joined by solid and dashed lines, respec-
s
tively.
that correlates diabatically with the spin-orbit excited Cl*
atom. The energetic ordering of the electronic states of the
Cl O parent shown in this figure follows the ab initio results
1
˜
1
2
sorption cross section from the 1 AЉϪX AЈ transition
21
of Nickolaisen et al., though our conclusion regarding the
diabatic correlation between the 1 B excited state and spin-
which is predicted to be more important at longer excitation
1
10,11,14
2
wavelengths.
Indeed, recent analyses of the Doppler
orbit excited Cl* atoms is insensitive to any reordering of the
line shapes of OH(X) fragments resulting from the 355 nm
photolysis of HOCl ͑which are best fit using an anisotropy
parameter ␤ϳϪ0.8͒ confirm the involvement of this perpen-
3
3
3
lower lying, closely spaced A , B and A states, or to the
2
2
1
1
1
relative energetic ordering of the 1 B2 state and the A2
state with which it is calculated to be in near resonance.
Thus we see that the asymptotic products of the model
potential energy surface considered in Fig. 8 should in fact
be Cl*ϩClO(X), and that ground state Cl atom products
45
dicular transition at longer excitation wavelengths. Alterna-
tively, at least in the case of the Cl product measurements,
the reduced ␤ value may simply reflect an overlapping con-
tribution from the less anisotropic recoil velocity distribution
͑which we observe to dominate in the 235 nm photodissocia-
of the Cl atoms from Cl O photolysis ͑characterized by ␤
2
tion of Cl O͒ must arise as a result of radiationless transfer
2
ϳ1.2͒. The production of OH(X) fragments exclusively in
their ground vibrational state, and their modest rotational ex-
citation, are both explicable in terms of Franck–Condon
1
from the 1 B surface to one or more of the myriad of lower
2
lying surfaces indicated in the correlation diagram ͑Fig. 9͒.
Note that there will be many such surfaces: the various com-
1
models and a potential energy surface for the 2 AЈ state
3
2
binations of the ground state atoms ͓O( P )ϩ2Cl( P )͔
J
J
which is steeply repulsive along the Cl—O stretching coor-
correlate with no fewer than 27 singlet states of Cl O, plus
7,8
2
dinate but fairly isotropic in the bending coordinate.
54 triplets and a further 27 quintet states. The model trajec-
tories sketched in Fig. 8 even allow us to speculate on the
region͑s͒ of configuration space where this transfer occurs.
The shaded area at larger Cl—O bond lengths in Fig. 8
shows a plausible region of many surface intersections,
which is consistent with experimental observations. Trajec-
tories such as A, involving only one elongated bond are un-
affected, and evolve to spin-orbit excited Cl* atoms and ClO
partners carrying little vibrational excitation, as before. How-
ACKNOWLEDGMENTS
M.N.R.A. is grateful to the EPSRC and NERC for finan-
cial support, in the form of equipment grants, a Senior Re-
search Fellowship ͑M.N.R.A.͒, and a studentship ͑L.J.R.͒, to
K. N. Rosser and Dr. B. J. Whitaker ͑University of Leeds͒
for their invaluable contributions toward the design and
implementation of the Bristol ion imaging experiment, and to
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:
128.42.202.150 On: Mon, 24 Nov 2014 12:19:42

J. Chem. Phys., Vol. 109, No. 4, 22 July 1998
Tanaka et al.
1323
2
2
3
4
Dr. A. J. Orr-Ewing for many helpful discussions. T.I. is
grateful for a Hiroshima City University Grant for Special
Academic Research ͑General Studies͒, M.K. is grateful to
the Heiwa-Nakajima Foundation for support, while both
M.N.R.A. and M.K. are grateful to the Japan Society for
Promotion of Science, the British Council and the Royal So-
ciety for the Japan–UK Co-operative program which made
this collaboration possible.
W. Finkelnburg, H. J. Schumacher, and G. Stieger, Z. Phys. Chem. Abt. B
5, 127 ͑1931͒.
H. J. Schumacher and R. V. Townend, Z. Phys. Chem. Abt. B 20, 375
1933͒.
F. H. C. Edgecomb, R. G. W. Norrish, and B. A. Thrush, Proc. R. Soc.
London, Ser. A 243, 24 ͑1957͒.
N. Basco and S. K. Dogra, Proc. R. Soc. London, Ser. A 323, 401 ͑1971͒.
S. P. Sander and R. R. Friedl, J. Phys. Chem. 93, 4764 ͑1989͒.
C. M. Nelson, T. A. Moore, M. Okumura, and T. K. Minton, J. Chem.
Phys. 100, 8055 ͑1994͒.
S. Abramowitz and M. W. Chase, Jr., Pure Appl. Chem. 63, 1449 ͑1991͒.
M. W. Chase, Jr., C. A. Davies, J. R. Downey, Jr., D. J. Frurip, R. A.
McDonald, and A. N. Syverud, J. Phys. Chem. Ref. Data 14, Suppl. 1
͑1985͒.
1
͑
25
2
2
2
6
7
8
2
9
1
30
R. P. Wayne, Chemistry of Atmospheres ͑Clarendon, Oxford, 1994͒.
R. P. Wayne, G. Poulet, P. Biggs, J. P. Burrows, R. A. Cox, P. J. Crutzen,
2
G. D. Hayman, M. E. Jenkin, G. Le Bras, G. K. Moortgat, U. Platt, and R.
N. Schindler, Halogen Oxides: Radicals, Sources and Reservoirs in the
Laboratory and in the Atmosphere ͑European Commission, Luxembourg,
Office for Official Publications of the European Communities, 1996͒, and
references cited therein.
3
3
1
2
R. P. Thorn, Jr., L. J. Stief, S.-C. Kuo, and R. B. Klemm, J. Phys. Chem.
100, 14,178 ͑1996͒, and references therein.
T. A. Moore, M. Okumura, and T. K. Minton, J. Chem. Phys. 107, 3237
͑
1997͒.
3
M. J. Molina, T.-L. Tso, L. T. Molina, and F. C.-Y. Wang, Science 238,
33
34
D. W. Chandler and P. L. Houston, J. Chem. Phys. 87, 1445 ͑1987͒.
A. J. R. Heck and D. W. Chandler, Annu. Rev. Phys. Chem. 46, 335
1
253 ͑1987͒.
4
5
L. T. Molina and M. J. Molina, J. Phys. Chem. 82, 2410 ͑1978͒.
E. A. Mishalanie, C. J. Rutkowski, R. S. Hutte, and J. W. Birks, J. Phys.
Chem. 90, 5578 ͑1986͒.
R. Vogt and R. N. Schindler, J. Photochem. Photobiol., A 66, 133 ͑1992͒.
A. J. Bell, P. R. Pardon, C. G. Hickman, and J. G. Frey, J. Chem. Soc.,
Faraday Trans. 86, 3831 ͑1990͒.
C. G. Hickman, N. Shaw, M. J. Crawford, A. J. Bell, and J. G. Frey, J.
Chem. Soc., Faraday Trans. 89, 1623 ͑1993͒.
A. J. Bell, S. A. Boggis, J. M. Dyke, J. G. Frey, R. Richter, N. Shaw, and
M. Tabrizchi, J. Chem. Soc., Faraday Trans. 91, 17 ͑1994͒.
G. Hirsch, P. J. Bruna, S. D. Peyerimhoff, and R. J. Buenker, Chem. Phys.
Lett. 52, 442 ͑1977͒.
͑
1995͒.
35
36
37
38
S. Arepalli, N. Presser, D. Robie, and R. J. Gordon, Chem. Phys. Lett.
18, 88 ͑1985͒.
1
6
7
V. Skorokhodov, Y. Sato, K. Suto, Y. Matsumi, and M. Kawasaki, J.
Phys. Chem. 100, 12,321 ͑1996͒.
L. J. Rogers, M. N. R. Ashfold, Y. Matsumi, M. Kawasaki, and B. J.
Whitaker, Chem. Phys. Lett. 258, 159 ͑1996͒.
8
9
0
R. Liyanage, Y. A. Yang, R. Gordon, and R. W. Field, J. Chem. Phys.
101, 6811 ͑1995͒.
3
9
1
M. Roth, C. Maul, and K.-H. Gericke, J. Chem. Phys. 107, 10582 ͑1997͒,
and references therein.
G. H. Cady, Inorg. Synth. 5, 156 ͑1957͒.
M. J. Molina, T. Ishiwata, and L. J. Molina, J. Phys. Chem. 84, 821
͑1980͒.
Y. Sato, Y. Matsumi, M. Kawasaki, K. Tsukiyama, and R. Bersohn, J.
Phys. Chem. 99, 16,307 ͑1995͒.
L. J. Rogers, M. N. R. Ashfold, Y. Matsumi, M. Kawasaki, and B. J.
Whitaker, J. Chem. Soc., Faraday Trans. 92, 5181 ͑1996͒.
T. Budinger and G. T. Gullberg, IEEE Trans. Nucl. Sci. 21, 2 ͑1974͒.
H. Fujiwara and T. Ishiwata, J. Phys. Chem. ͑in press͒.
P. W. McLoughlin, C. R. Park, and J. R. Wiesenfeld, J. Mol. Spectrosc.
4
4
0
1
1
1
1
1
1
2
3
4
S. Nambu, K. Nakata, and S. Iwata, Chem. Phys. 135, 75 ͑1992͒.
S. Nambu and S. Iwata, J. Phys. Chem. 96, 2103 ͑1992͒.
H. Guo, J. Phys. Chem. 97, 2602 ͑1993͒.
4
4
2
3
A. R. Offer and G. G. Balint-Kurti, J. Chem. Phys. 101, 10416 ͑1994͒;
1
04, 563 ͑1996͒; Chem. Phys. Lett. 247, 173 ͑1995͒.
1
5
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 Modelling,
Evaluation No. 12, JPL Publication 97-4 ͑1997͒.
4
4
4
4
5
6
1
1
6
7
G. E. Herberich, R. H. Jackson, and D. J. Miller, J. Chem. Soc. A 1966,
162, 307 ͑1993͒, and references therein.
3
36.
4
4
4
7
8
9
A. J. Chichinin, Chem. Phys. Lett. 209, 459 ͑1993͒.
R. N. Zare, Angular Momentum ͑Wiley, New York, 1988͒.
M. R. Wedlock, R. Jost, and T. R. Rizzo, J. Chem. Phys. 107, 10344
͑1997͒.
G. H. Dieke and D. R. Crosley, J. Quant. Spectrosc. Radiat. Transf. 2, 97
͑1962͒.
S. R. Langhoff, unpublished results reported in Ref. 28.
C. Maul, T. Haas, K.-H. Gericke, and F. J. Comes, J. Chem. Phys. 102,
3238 ͑1995͒.
M. Kawasaki, K. Suto, Y. Sato, and Y. Matsumi, J. Phys. Chem. 100,
19853 ͑1996͒.
M. Nakata, M. Sugie, H. Takeo, C. Matsumara, T. Fukuyama, and K.
Kuchitsu, J. Mol. Spectrosc. 86, 241 ͑1981͒.
C.-L. Lin, J. Chem. Eng. Data 21, 411 ͑1976͒.
1
1
8
9
H.-D. Knauth, H. Alberti, and H. Clausen, J. Phys. Chem. 83, 1604
5
0
͑1979͒.
2
2
0
1
J. B. Nee, J. Quant. Spectrosc. Radiat. Transf. 46, 55 ͑1991͒.
S. L. Nickolaisen, C. E. Miller, S. P. Sander, M. R. Hand, I. H. Williams,
and J. S. Francisco, J. Chem. Phys. 104, 2857 ͑1996͒.
F. Motte-Tollet, M.-P. Ska, G. M. Marston, I. C. Walker, M. R. F. Siggel,
J. M. Gingell, L. Kaminski, K. Brown, and N. J. Mason, Chem. Phys. Lett.
51
5
2
22
5
3
2
75, 298 ͑1997͒.
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:
28.42.202.150 On: Mon, 24 Nov 2014 12:19:42
1