Photodissociation dynamics of the CH2Cl radical: Ion imaging studies of the Cl+CH2 channel

Article abstract of DOI:10.1063/1.1400130

Photodissociation of the chloromethyl radical, CH₂Cl to chlorine atom and methylene, which can serve as a prototype for halogenated methyl radicals, was described. Ch₂Cl was produced in a molecular beam using pulsed pyrolysis. Cl and CH₂ products were detected by laser ionization and their velocity and angular distributions were determined using the ion imaging technique. The available energy was partitioned into the translational degrees of freedom. It was found that the CH₂Cl generated in a molecular beam via pulsed pyrolysis exhibits continuous, unstructured absorption throughout the investigated region, with Cl atoms as major products.

Full text of DOI:10.1063/1.1400130

Photodissociation dynamics of the CH 2 Cl radical: Ion imaging studies of the Cl+CH 2
channel
V. Dribinski, A. B. Potter, A. V. Demyanenko, and H. Reisler
Citation: The Journal of Chemical Physics 115, 7474 (2001); doi: 10.1063/1.1400130
View online: http://dx.doi.org/10.1063/1.1400130
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/115/16?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Product branching ratios in photodissociation of phenyl radical: A theoretical ab
initio/Rice–Ramsperger–Kassel–Marcus study
J. Chem. Phys. 136, 234305 (2012); 10.1063/1.4726455
Photodissociation of CH 3 Cl , C 2 H 5 Cl , and C 6 H 5 Cl on the Ag(111) surface: Ab initio embedded cluster
and configuration interaction study
J. Chem. Phys. 132, 074707 (2010); 10.1063/1.3322289
Photodissociation of the geometric isomers of 1,2-dibromoethylene
J. Chem. Phys. 131, 154304 (2009); 10.1063/1.3246824
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 photodissociation of the vinyl radical ( C 2 H 3 ) at 243 nm studied by velocity map imaging
J. Chem. Phys. 110, 4248 (1999); 10.1063/1.478307
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:
137.112.220.85 On: Wed, 10 Dec 2014 18:27:02

JOURNAL OF CHEMICAL PHYSICS
VOLUME 115, NUMBER 16
22 OCTOBER 2001
Photodissociation dynamics of the CH Cl radical:
2
Ion imaging studies of the Cl¿CH channel
2
V. Dribinski, A. B. Potter, A. V. Demyanenko, and H. Reisler
Department of Chemistry, University of Southern California, Los Angeles, California 90089-0482
͑
Received 13 June 2001; accepted 16 July 2001͒
The photodissociation of the chloromethyl radical, CH Cl, to chlorine atom and methylene is
2
examined following excitation at selected wavelengths in the region 312–214 nm. CH Cl is
2
produced in a molecular beam by using pulsed pyrolysis. Cl and CH products are detected by laser
2
ionization and their velocity and angular distributions are determined by using the ion imaging
technique. The spectrum obtained by monitoring the Cl fragment yield as function of photolysis
wavelength shows that throughout this wavelength region Cl atoms are major products. With
3
12–247 nm photolysis, the angular distributions are typical of a perpendicular transition (␤
3
2
˜
ϭϪ0.7) and the main products are CH (X B )ϩCl( P3/2). The available energy is partitioned
2
1
preferentially into the translational degrees of freedom. ‘‘Hot band’’ transitions are prominent in this
region even in the molecular beam indicating that the geometries of the ground and excited states
of CH Cl must be very different. With 240–214 nm photolysis, the angular distributions are typical
2
2
2
of a parallel transition (␤ϳ1.2), and the predominant products are Cl( P ) and Cl( P1/2), with
3
/2
1
CH2(
internal energy of CH2(
Krylov presented in the accompanying paper enables the assignment of the perpendicular and
˜a A ) as the main cofragment. A large fraction of the available energy is partitioned into
1
1
˜
a A ). Comparison with the ab initio calculations of Levchenko and
1
2
2
2
2
parallel transitions predominantly to 1 A � 1 B and 2 B � 1 B excitations, respectively, and
1
1
1
1
both upper states are probably repulsive in the C–Cl coordinate. The electronic states of the
products obtained via these two transitions are in agreement with the predictions of a simple diabatic
state correlation diagram based on the calculated vertical energies of the upper states. © 2001
American Institute of Physics. ͓DOI: 10.1063/1.1400130͔
I. INTRODUCTION
many of their excited states are dissociative or predissocia-
tive, giving rise to multiple dissociation pathways.
As shown by Levchenko and Krylov in the accompany-
The photoinitiated chemistry of free radicals is of great
importance in atmospheric and environmental chemistry,1
but its study is hindered by experimental and theoretical dif-
ficulties. Halogenated methyl radicals are important in strato-
spheric chemistry, since they can provide an additional
source of photolytic halogen atoms, but their absorption
spectra in regions where the actinic flux is high are still un-
known. Molecules and radicals with halogen substituents of-
ten possess low-lying valence states associated with the pro-
motion of an electron from a lone pair centered on the
halogen to an antibonding orbital, thereby weakening the
carbon-halogen bond. In addition, radicals usually have low
ionization potentials, and therefore they have Rydberg states
that lie in the same energy region as their lowest valence
states. Because Rydberg states are often correlated with ex-
cited electronic states of the products, the interactions be-
tween the Rydberg and valence states and their dependence
on internuclear distances are of crucial importance to the
3
ing paper, determining the relative order of Rydberg and
valence excited states of CH X radicals and the associated
2
electronic transition dipole moments requires high level of
theory. The energies of the excited states depend sensitively
on such factors as how strongly the halogen lone pair is
bound, the energetics and nature of the relevant molecular
orbitals, and the Franck–Condon factors for absorption.
In this paper we describe the first experimental results on
the photodissociation of the chloromethyl radical, CH Cl,
2
which can serve as a prototype for halogenated methyl radi-
cals. CH Cl is produced as a primary product in the pyrolysis
2
and photolysis of CH ClX molecules, where X is a halogen,
2
4
and in the reaction of Cl or F with CH Cl. To date the only
3
reported electronic absorption is a strong, unstructured band
centered at ϳ200 nm ͑where the cross section is 1.45
Ϫ17
2
4
ϫ10
cm ͒, which extends to ϳ220 nm. No absorption
2
measurements at wavelengths longer than 220 nm have been
reported, although the existence of an excited electronic state
around 5 eV has been suggested based on analysis of micro-
dynamics. An intriguing aspect concerning the low-lying
Rydberg states is the extent of their penetration into the core
region. For example, if the nodes in the radial Rydberg wave
function lend a significant antibonding character to regions
of the wave function associated with specific bonds, this pen-
etration will influence the photodissociation dynamics. Radi-
cals often dissociate to more stable species, and therefore
5
wave spectra; i.e., an absorption band at ϳ250 nm.
The photochemistry of CH2Cl is interesting from another
perspective. The radical has several dissociation channels
with very similar thermochemical thresholds:6
–13
0021-9606/2001/115(16)/7474/11/$18.00
7474
© 2001 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:
37.112.220.85 On: Wed, 10 Dec 2014 18:27:02
1

7475
J. Chem. Phys., Vol. 115, No. 16, 22 October 2001
Photodissociation of CH Cl
2
CH Cl→ClϩCH , ⌬Hϭ391Ϯ5 kJ molϪ1
,
temperature calibration curve obtained in a separate experi-
ment by using an optical pyrometer. The pyrolysis efficiency
is estimated by comparing the REMPI signals of iodine frag-
2
2
HϩCHCl, ⌬Hϭ395Ϯ5 kJ molϪ1
,
H ϩCCl, ⌬Hϭ382Ϯ5 kJ molϪ1
ments obtained by photodissociation of the CH ClI precursor
2
,
2
in the molecular beam with the heating of the pyrolysis tube
turned on and off. As the temperature of the nozzle is in-
creased, the I atom signal from the photodissociation of
HClϩCH, ⌬Hϭ384Ϯ5 kJ molϪ1
.
The main uncertainties in calculating the enthalpy
changes in the decomposition derive from the heat of forma-
CH ClI diminishes. However, above a certain temperature,
2
8
–13
the efficiency of production of CH Cl starts to decrease as a
2
tion of the CH Cl radical.
Thus, the relative dissociation
2
Ϫ1
result of both its thermal decomposition, and the increasing
contribution of higher-threshold dissociation pathways in
thresholds are known to a better accuracy ͑ϳϮ2 kJ mole ͒.
The first two channels involve simple bond-fission reactions
ending in two free radicals, and on the ground electronic
state they are likely to evolve without significant barriers.
Because their thresholds are so similar, the branching ratio
between them is important. The last two channels are ex-
pected to have significant activation barriers. In this paper
CH ClI. For example, fragments such as Cl atoms and HI are
2
produced in the pyrolysis source. Under our experimental
conditions, a large depletion ͑more than a factor of 2͒ of the
iodine photofragment signal is observed, while at the same
time the amount of thermal Cl atoms does not exceed 1% of
the total Cl signal.
After the supersonic expansion, the radicals are skimmed
twice and introduced into the reaction chamber. In order to
minimize background signals from the photodissociation of
we report results solely on the ClϩCH channel.
2
In our experiments, we use pulsed pyrolysis to prepare
the radicals in a molecular beam, and the ion imaging tech-
nique for photofragment detection. We clearly identify two
distinct electronic transitions in the region 312–214 nm, and
CH ClI, the experiments are carried out at the leading edge
2
3
of the molecular pulse exiting the nozzle. With He as the
carrier gas, the heavy, iodine-containing species arrive at the
reaction chamber delayed by approximately 30 ␮s with re-
spect to the light species, as a result of velocity slip.15 Thus,
the leading edge of the molecular pulse does not contain a
based on the ab initio results we assign their upper states as
2
2
1
A and 2 B . Both are dissociative, and give rise to Cl
1 1
atom as a major product. We argue that the geometries of the
2
2
ground 1 B state and the first excited 1 A state must be
1
1
substantially different, as evidenced by strong ‘‘hot band’’
absorption. We also comment on the electronic states of the
CH cofragment and the spin–orbit state populations of the
significant amount of CH ClI. When Ar is used as a carrier
2
gas, there is no velocity slip, and substantial background
subtraction of signals from the photodissociation of the pre-
cursor is necessary, degrading the signal to noise ratio in the
experiment.
2
Cl atoms. We end by discussing a state correlation diagram
relevant to the dissociation to ClϩCH , and the possible par-
2
2
ticipation of the 2 A Rydberg state.
1
B. Photodissociation and detection of products
II. EXPERIMENTAL METHODS
A. Radical production
The dissociation of CH Cl is examined at selected wave-
2
lengths in the region 311.8–214 nm. Radical photolysis and
product detection are achieved by using the frequency-
doubled output of a Nd:YAG or excimer-laser pumped dye
laser systems. One system is used for photolysis ͑pump͒ and
the other to probe the products. The pump and probe lasers
are both linearly polarized, and introduced coaxially at a
right angle to the molecular beam. The polarizations of the
two lasers are fixed parallel to each other and kept perpen-
dicular to the plane defined by the molecular beam and the
laser propagation axis. The time delay between the pump and
probe lasers is kept at 0–5 ns.
CH Cl radicals are produced by thermal decomposition
2
of chloroiodomethane (CH ClI) in a pulsed pyrolysis source.
2
The choice of the precursor is dictated by the energetics of
dissociation of CH ClI on the ground electronic state. Ac-
2
6
–13
cording to the known thermochemical data,
dissociation channel is CH ClϩI (⌬H ϭ242 kJ/mol). The
next channel, CHClϩHI, has ⌬H ϭ337 kJ/mol, but is ex-
the lowest
ؠ
2
ؠ
pected to have a significant activation barrier, while the third
ؠ
channel is CH IϩCl (⌬H ϭ367 kJ/mol). Therefore, it
2
should be possible to optimize the conditions in the pulsed
pyrolysis source to maximize production of CH Cl.
Velocity map imaging combined with resonance en-
hanced multiphoton ionization ͑REMPI͒ is used for product
2
The experiments are performed using a pulsed molecular
beam ͑0.5–1.0% CH ClI seeded in 2.0 atm of He or Ar͒. The
gas mixture is prepared by passing the carrier gas over
detection. Our ion-imaging arrangement has been described
2
in detail elsewhere.1
6,17
In brief, it consists of an ion-
CH ClI ͑Fluka, у97% purity͒ kept at Ϫ29 °C ͑o-xylene
acceleration stage, a 60-cm long drift tube, and a CCD cam-
era that monitors a phosphor screen coupled to a MCP de-
tector. The ion-optics consists of a repeller, an open extractor
electrode ͑50 mm hole͒, and an open ground electrode ͑25
mm hole͒. At an optimal voltage ratio of the repeller and
extractor plates, all ions with the same initial velocity are
focused onto the same spot on the MCP detector.
2
slush͒ or 0 °C ͑ice͒. This mixture is expanded through a
pulsed piezoelectric valve ͑10 Hz repetition rate͒ with an
attached SiC tube nozzle ͑1 mm i.d.͒, and our design is simi-
lar to that of Chen and co-workers.14 The nozzle has a 1-cm
long resistively heated zone.
In our experiments, the SiC tube nozzle is heated to a
temperature of ϳ1800 K which results in removal of I atoms
2
We use 2ϩ1 REMPI to detect Cl( P3/2) (␭
2
18
from CH ClI with an efficiency of more than 50%. The tem-
ϭ235.337 nm) and Cl( P1/2) (␭ϭ235.205 nm). Other Cl-
2
perature of the nozzle was estimated from a voltage/
atom detection wavelengths in the region 235–240 nm have
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:
137.112.220.85 On: Wed, 10 Dec 2014 18:27:02

7476
J. Chem. Phys., Vol. 115, No. 16, 22 October 2001
Dribinski et al.
smaller intensities and are only used for calibration purposes.
Due to the Doppler width of the Cl absorption band, the
probe frequency is scanned over the entire bandwidth to en-
sure that the images include all product velocities.
3
˜
CH (X B ) radicals are monitored by 2ϩ1 REMPI at
2
1
1
9
␭
ϭ311.80 nm, as proposed by Hudgens and co-workers.
Although the REMPI spectrum has not been assigned, it
shows clear structures with the largest peak at 311.80 nm,
and the intermediate state is assigned as a Rydberg 3p
1
9
1
state. No REMPI scheme for the detection of CH2( ˜a A )
1
is known, and no REMPI signal from CH Cl could be de-
2
tected, because of the rapid dissociation of the radical in the
wavelength region studied here. A weak REMPI signal is
observed at 302–318 nm, which is a result of two-photon
excitation, probably to a Rydberg state.
C. Data analysis
Two types of background signals are subtracted from the
images obtained in two-color pump and probe experiments:
͑i͒ the signal obtained from the molecular beam pulse, but
with the pump laser off accounts for background generated
by the probe laser alone; ͑ii͒ the signal obtained with both
pump and probe lasers off accounts for the ‘‘dark noise’’ of
the CCD camera. In addition the signals obtained with the
pyrolysis tube heating turned on and off are always com-
pared.
˜
3
2
FIG. 1. Abel transformed images of CH (X B ) ͑top͒ and Cl( P3/2) ͑bot-
2
1
The resultant images are normalized to the relative sen-
sitivity of the pixels in the CCD camera and deconvoluted
from the response function of the data acquisition system.
The inverse Abel transform is performed on the normalized
signal to extract the 3D distributions that give rise to the 2D
˜
3
tom͒ obtained from CH Cl photodissociation at 266 nm. CH (X B ) and
2
2
1
2
Cl( P3/2) were detected by (2ϩ1) REMPI at 311.8 nm and 235.34 nm,
respectively. The size of each image is adjusted to fill the detector plane. The
electric field vector of the photolysis laser is parallel to the vertical direction
of the image plane.
1
6,17,20
projections.
for each radial distance from the center ͑after calibrating
each image for conversion from distance to speed͒. Each
speed distribution is then converted to the translational en-
ergy distribution of the specific fragment, and the total c.m.
III. RESULTS AND ANALYSIS
Two electronic transitions are clearly identified in our
3
˜
experiments. In the region 311.8–247 nm, CH (X B ) and
2
1
translational energy distribution, E , is finally derived by
2
t
Cl( P3/2,1/2) products derive from dissociation via a perpen-
scaling according to the mass ratio dictated by momentum
dicular electronic transition in CH Cl. At wavelengths
shorter than about 240 nm the Cl( Pjϭ3/2,1/2) products ex-
hibit angular distributions that are characteristic of a parallel
transition.
2
2
and energy conservation. For Cl( P ) atoms, the E distri-
2
3/2
t
bution can be directly converted to internal energy distribu-
tion of the matching CH fragment. The situation is more
2
3
˜
complicated for the CH (X B ) product, since neither a
2
1
Figure 1 depicts typical Abel transformed images of
single rovibrational state nor the total state population is
2
3
˜
Cl( P ) and CH (X B ) obtained with 266 nm photolysis.
2
3
/2
2
1
monitored ͑see below͒. Small signals from Cl( P ), which
1
/2
In this figure, as in all other images, the electric field vector
of the photolysis laser is parallel to the vertical direction of
the image plane. Both images have an angular distribution
characteristic of a perpendicular transition. The effective re-
coil anisotropy parameter ␤ is Ϫ0.7Ϯ0.1. Analysis of the
Ϫ1
2
18
is 881 cm higher in energy than ground state Cl( P3/2),
are also observed at all wavelengths, and their associated
2
energy distributions are similar to those of Cl( P3/2).
The widths of the translational energy distributions for
Cl correspond to a range of internal excitations in the CH2
co-fragment. Note in particular the two major peaks in this
velocities shows that the Cl and CH fragments match by
linear momentum.
2
Ϫ1
distribution, which are approximately 3000 cm apart, and
Figures 2 and 3 show the total photoproduct center-of-
mass ͑c.m.͒ translational energy distributions obtained by
3
˜
may correspond to the ground X B and the first excited
˜
1
1
Ϫ1 7
a A (T ϭ3147 cm ) states of CH . In Figs. 2 and 3,
2
3
1 e 2
˜
monitoring Cl( P ) and CH (X B ) at dissociation wave-
3
/2
2
1
arrows indicate the maximum allowed translational energy of
lengths 280, 266, and 247 nm. These distributions are de-
rived from the speed distributions of each fragment. The
max
the products, Et , calculated for photolysis of CH Cl in its
2
ground vibrational state:
2
3
˜
speed distributions of Cl( P3 ) and CH (X B ) are first ob-
/2
2
1
E mt axϭhvϪD0 .
͑1͒
tained from the transformed images by integrating the signal
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:
37.112.220.85 On: Wed, 10 Dec 2014 18:27:02
1

7477
J. Chem. Phys., Vol. 115, No. 16, 22 October 2001
Photodissociation of CH Cl
2
˜
3
2
FIG. 2. CH (X B ) ͑top panel͒ and Cl( P ) ͑bottom panel͒ signals ob-
2
1
3/2
Ϫ1
tained following 266 nm dissociation of CH Cl ͑37 594 cm ͒ plotted as a
2
function of the total c.m. photofragment translational energy, E . The details
t
of the transformation of the Abel transformed images to total c.m. transla-
tional distributions are given in the text. The long arrows show the energy
˜
3
2
FIG. 3. CH2(X B1) ͑top panel͒ and Cl( P3/2) ͑bottom panel͒ signals ob-
Ϫ1
threshold calculated using D ϭ391 kJ/mol. The short arrow in the Cl en-
tained following 247 nm ͑40 486 cm ; solid lines͒ and 280 nm ͑35 714
0
1
cmϪ1; dotted lines͒ dissociation of CH2Cl plotted as a function of the total
ergy distribution shows the energy threshold for production of CH2(
˜
a A ).
1
c.m. photofragment translational energy, E , as described in Fig. 2. The
t
arrows indicate the energy thresholds calculated using D ϭ391 kJ/mol.
0
The complete energy balance for CH Cl dissociation is
2
given by
wavelengths in the range 312–247 nm. Another possible
source of fragments with excess kinetic energy is two-photon
E ͑CH Cl͒ϩhvϭD ϩE ϩE ͑CH ͒ϩE ͑Cl͒, ͑2͒
int
2
0
t
int
2
int
where hv is the photon energy, D is the dissociation thresh-
dissociation of CH Cl. However, the products of such disso-
0
2
old of the CH Cl radical, and E (CH Cl), E (CH ), and
ciation ͑i͒ should have considerably larger translational ener-
gies since absorption of a second photon at these wave-
lengths would impart additional energy of more than 30 000
2
int
2
int
2
E (Cl) are the internal energies ͑rotational, vibrational, and
int
electronic͒ of the respective reactant and fragments. The
Ϫ1
value of D was derived from the enthalpies of formation of
cm , and ͑ii͒ must show higher power dependence on pump
0
CH Cl, CH , and Cl to be 391Ϯ5 kJ/mol (32 690
laser energy. Neither was observed in our experiments, and
this rules out the presence of signal from two-photon disso-
ciation.
2
2
Ϫ1 6–13
Ϯ420 cm ).
As seen from Figs. 2 and 3, large fractions
max
of the products have translational energies exceeding Et
.
Products with excess translational energy can arise from
In order to confirm that products with excess kinetic en-
ergy result from dissociation of vibrationally excited radi-
cals, we carried out experiments with Ar as a carrier gas.
the dissociation of vibrationally ‘‘hot’’ CH Cl radicals ͓i.e.,
2
E (CH Cl)Þ0͔. The experiments described above were
int
2
2
performed using He as a carrier gas. Based on previous
Figure 4 shows an image of Cl( P3/2) obtained from the
2
1,22
work,
effective rotational cooling of CH Cl radicals is
dissociation of CH Cl at 266 nm in Ar carrier gas and de-
2
2
expected in the pyrolysis source, but He is not efficient in
relaxing the vibrational excitations. Thus, the high energy
tails of the observed product distributions are attributed
mainly to the dissociation of vibrationally excited ͑‘‘hot’’͒
tected at 235.34 nm, and Fig. 5 shows the corresponding c.m.
translational energy distribution. Similar results were ob-
tained at 247 nm. Compared with the He experiments, two
main differences are observed. First, the translational energy
distribution becomes narrower with the largest peak at high
CH Cl radicals. A similar behavior is observed at other
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:
137.112.220.85 On: Wed, 10 Dec 2014 18:27:02

7478
J. Chem. Phys., Vol. 115, No. 16, 22 October 2001
Dribinski et al.
2
2
FIG. 4. Abel transformed image of Cl( P ) obtained from CH Cl photo-
FIG. 6. Abel transformed image of Cl( P ) obtained from CH Cl photo-
3/2 2
3
/2
2
2
dissociation at 266 nm using Ar as a carrier gas. Cl( P3/2) was detected by
2ϩ1) REMPI at 235.34 nm. The electric field vector of the photolysis
laser is parallel to the vertical direction of the image plane.
dissociation at 235.34 nm in a one-color experiment. The electric field vec-
tor of the photolysis laser is parallel to the vertical direction of the image
plane.
(
ation in CH Cl by using a carrier gas that is 1:7 CF :He. The
results, however, were similar to those obtained with Ar as
the carrier gas.
energies corresponding to the dissociation of vibrationally
2
4
cold CH Cl. The short arrows in Fig. 5 indicate the energy
2
thresholds of steps that are discernible in the Cl energy dis-
tribution obtained in the Ar carrier. These steps are separated
The widths of the energy distributions of the CH frag-
2
Ϫ1
ments in Figs. 2 and 3 require comment. Each CH fragment
by ϳ800 cm , which is the frequency of the C–Cl stretch in
2
2
2
can correlate with either Cl( P ) or Cl( P1/2) and, there-
CH Cl. Second, the relative intensity of the low energy peak
3/2
2
1
fore, the translational energy distribution for each particular
͓
which is likely to contain CH2(
˜
a A )͔ is larger. In the ex-
1
Ϫ1
3
CH (␯,J) should consist of two narrow peaks 881 cm
˜
2
periments with Ar, we were unable to detect CH (X B ) by
2
1
apart, though in the present case the peak correlated with
REMPI. This may be a result of the much smaller signal
levels obtained in general when using Ar as a carrier gas in
the pyrolysis source, and/or the smaller fraction of
2
Cl( P1/2) would be small. As seen in Figs. 2 and 3, the
translational energy distributions of CH obtained in our ex-
2
periments are much broader than the experimental resolution
3
1
˜
CH (X B ) relative to CH ( ˜a A ) produced in the dissocia-
2
1
2 1
(
⌬E/Eϳ0.04). This is especially clear in the 247 nm distri-
tion of ground state CH Cl. We cannot rule out, however, the
2
bution, which has a smaller ‘‘hot band’’ component. A pos-
sible cause for such broad distributions is the simultaneous
possibility that the REMPI signals that correspond to Cl at-
oms with the lowest translational energies derive also from
detection of multiple rotational levels of CH encompassed
2
other sources, since we do not have direct diagnostics for
in the bandhead at 311.8 nm. This conclusion is supported by
calculations of the approximate positions of the rotational
1
CH2(
˜
a A ). We attempted to improve the vibrational relax-
1
lines of the 0–0 vibronic band of the CH electronic transi-
2
tion ͑Appendix͒. The rotational constants for the ground
electronic state of CH are very close to those of the 3p
2
Rydberg state ͑the intermediate state in the 2ϩ1 REMPI
detection scheme͒,19 and therefore several rotational transi-
tions have almost the same energy. As shown in the Appen-
dix, for the QQ rotational branch ͑⌬Nϭ0, ⌬Kϭ0͒, transi-
tions involving rotational levels up to Nϭ6, Kϭ5 have
Ϫ1
transition energies that differ by less than 10 cm . Because
the upper electronic state is broadened by intramolecular
interactions,19 many of these levels are detected simulta-
neously. At the same time, the rotational energies of these
Ϫ1
levels can cover a range of up to 2000 cm , and thus the
resulting translational energy distributions of the detected
CH products can be broad.
2
When the dissociation wavelength is tuned to the region
2
40–214 nm, the angular distributions of the products
2
FIG. 5. Cl( P ) signals obtained following 266 nm dissociation of
change qualitatively. Shown in Fig. 6 is an image of
Cl( P3 ) obtained in a one-color experiment at 235.34 nm.
The angular distribution of Cl( P3/2) is typical of a parallel
electronic transition in CH Cl, as are all the other distribu-
3
/2
Ϫ1
2
CH Cl (37 594 cm ) using Ar ͑solid line͒ or He ͑dotted line͒ carrier gas
plotted as a function of the total c.m. photofragment translational energy,
E . The long arrows indicate the energy thresholds for the production of
2
/2
2
t
˜
3
1
2
CH (X B ) and CH ( ˜a A ) from vibrationally ‘‘cold’’ CH Cl ͑see Fig. 2͒.
The short arrows show the positions of ‘‘hot band’’ peaks which are resolved
in the Ar experiments.
2
1
2
1
2
2
tions of Cl( P3/2,1/2) in the 240–214 nm photolysis region.
Figure 7 shows the c.m. translational energy and the an-
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:
37.112.220.85 On: Wed, 10 Dec 2014 18:27:02
1

7479
J. Chem. Phys., Vol. 115, No. 16, 22 October 2001
Photodissociation of CH Cl
2
FIG. 7. Total c.m. translational energy ͑solid lines͒ and angular anisotropy
2
2
parameter ͑squares͒ distributions of Cl( P ) ͑upper panel͒ and Cl( P1/2)
3
/2
͑
lower panel͒ obtained in one-color experiments at 235.34 nm and 235.21
FIG. 8. Total c.m. translational energy ͑solid lines͒ and angular anisotropy
parameter ͑squares͒ distributions of Cl( P3/2) obtained from the dissociation
nm, respectively. Arrows show the energy thresholds for production of
2
˜
3
1
CH (X B ) and CH ( ˜a A ).
2 1
2
1
of CH Cl at 214 nm ͑upper panel͒ and 219.5 nm ͑lower panel͒ and detected
2
by (2ϩ1) REMPI at 235.34 nm. Arrows show the energy thresholds for
˜
3
1
production of CH (X B ) and CH ( ˜a A ).
2 1
2
1
isotropy parameter distributions obtained by monitoring
2
2
Cl( P ) and Cl( P1/2) in one-color experiments at 235.34
3
/2
Ϯ0.1. We conclude that the observed spectrum at 240–235
nm is a combination of excitations via parallel and perpen-
dicular transitions, each giving rise to Cl product with a dif-
ferent translational energy distribution. The fraction of the
parallel component increases gradually as the wavelength is
decreased from 240 nm to 214 nm.
nm and 235.21 nm, respectively. For a fast, parallel transition
in CH Cl, the maximum anisotropy parameter ␤ϭ2 will be
2
obtained when the transition dipole moment is parallel to the
2
2
C–Cl bond. However, in both Cl( P3 ) and Cl( P1/2), ␤ is
/2
considerably smaller than 2, and in addition it is speed de-
pendent, i.e., products with higher translational energies have
smaller values of ␤ than those with lower translational ener-
gies. At 219.5 and 214 nm, however, ␤ is characteristic of a
In Figs. 7 and 8 we also mark the energy thresholds for
the production of the ground and first excited states of CH .
2
3
˜
Very little Cl is observed that corresponds to CH (X B ).
parallel transition at all speeds where the Cl population is
2
1
Ϫ1
This, along with the fact that we are unable to detect
significant (E Ͼ2000 cm ) as seen in Fig. 8. We note that
t
3
˜
the speed and ␤ distributions are the least reliable at energies
CH (X B ) by REMPI at these wavelengths even though
2
1
Ϫ1
below ϳ2000 cm for the following reasons: ͑i͒ Errors
the Cl signal is quite strong, indicates that in excitation via
1
from the Abel transformation accumulate in the low velocity
region of the image ͑see, for example, Fig. 6͒; ͑ii͒ the prod-
the parallel transition, CH2(
product.
˜a A ) is probably the major
1
uct signal at low E is quite low; ͑iii͒ a small, unknown
We note that at 214 and 219.5 nm, there are only few
fragments with translational energies exceeding significantly
Et , and no change is observed when the He carrier gas is
t
contamination sometimes gives rise to a narrow feature with
Ϫ1
max
negative ␤ in the region 1500–2000 cm . The error bars on
Ϫ1
the measured values of ␤ at E Ͼ2000 cm are typically
replaced by Ar. Hence, we conclude that the parallel transi-
t
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:
137.112.220.85 On: Wed, 10 Dec 2014 18:27:02

7480
J. Chem. Phys., Vol. 115, No. 16, 22 October 2001
Dribinski et al.
tion is not as severely affected by ‘‘hot band’’ excitation as is
the perpendicular transition.
The experiments in the 240–214 nm photolysis region
were performed under the same pyrolysis conditions as those
at 312–247 nm, and therefore we do not expect that sources
other than CH Cl would contribute significantly to the Cl or
2
CH signals. We conclude that Cl products with positive and
negative values of ␤ result from dissociation following par-
IV. DISCUSSION
2
A. Identification of the radical precursor
allel and perpendicular transitions, respectively, in CH Cl.
2
Since the results reported here are based exclusively on
monitoring photofragments, it is essential to verify, before
attempting an interpretation, that the precursor of all the ob-
B. Electronic transitions in CH Cl
2
served products is CH Cl. The absence of a REMPI peak of
1. Perpendicular transition
2
molecular CH Cl in this wavelength region is attributed to
the fast dissociation ͑␤ϭϪ0.7 and 1.2 for the perpendicular
and parallel transitions, respectively͒. We note, however, that
2
Analysis of the Cl and CH angular distributions indi-
cates that in the region between 311.8 nm ͑3.98 eV͒ and 247
2
nm ͑5.02 eV͒ the excited electronic state͑s͒ of the CH Cl
2
weak 2ϩ1 REMPI peaks corresponding to CH Cl cations
2
radical is ͑are͒ accessed via a perpendicular transition. At all
wavelengths in this region, the angular distributions of both
2
3
are observed upon excitation at ϳ376 and 302–318 nm,
and are attributed to two photon excitations to Rydberg
CH and Cl products can be fitted with a negative recoil
2
3
,24
states.
anisotropy parameter, ␤ϭϪ0.7Ϯ0.1. This implies fast, di-
For the products observed via the perpendicular transi-
rect dissociation on a repulsive region of the potential energy
tion ͑312–247 nm͒, the identification of CH Cl as the pre-
3,5,25–27
2
surface. As established previously,
CH Cl is planar in
2
cursor is based on several observations. First, both
its ground state and has C2v symmetry. Since the ground
state is of B1 symmetry, the presence of a perpendicular
transition indicates excitation to either A or A electronic
3
2
˜
CH (X B ) and Cl( P3/2) photofragments are detected un-
2
1
der the same experimental conditions ͑He carrier gas͒, exhib-
iting identical angular distributions and the same maximum
value of linear momentum.
2
1
states, whose transition dipole moments lie either in-plane or
out-of-plane, respectively, but perpendicular to the Cl recoil
direction.
Second, no other significant sources of the CH and Cl
2
pair are expected to be present in the leading region of the
Figures 2, 3, and 5 show clearly that with He carrier gas,
molecular pulse. As mentioned above, parent CH ClI mol-
2
dissociation of vibrationally excited CH Cl contributes sig-
2
ecules are largely absent because of velocity slip. The ther-
nificantly to the observed CH and Cl signals, especially at
2
mochemistry of CH ClI dictates that CH I and CHCl would
2
2
wavelengths longer than 260 nm. The fraction of fragments
max
be only minor pyrolysis products under our experimental
with translational energies greater than Et is much larger
conditions, and they would lead to Cl and CH products that
2
than expected had the vibrationally excited radicals in the
molecular beam been excited with the same efficiency as the
unexcited radicals. The absorption spectrum can only be ra-
tionalized by postulating a very large increase in the Franck–
Condon factors for ‘‘hot band’’ excitation due to the vibra-
tional motions in the radical. Therefore, we argue that the
ground and excited electronic state geometries must be rather
different. Note that even in Ar some of the Cl fragments have
are not momentum matched.
Last, although only a small signal from ground state CH2
is detected in the experiments with Ar carrier gas, the
2
Cl( P /2) fragments exhibit a similar structure in their energy
3
distributions to that obtained with He carrier gas; i.e., two
Ϫ1
broad peaks separated by about 3000 cm , which corre-
3
1
˜
sponds to the energy separation of the X B and ˜a A states
1
1
max
of CH . In Ar, the translational energy associated with most
energies exceeding Et . This energy distribution is struc-
2
of the Cl products is smaller than the maximum calculated
tured, with the bands separated by the C–Cl stretch energy in
for dissociation of vibrationally ‘‘cold’’ CH Cl ͑see below͒.
CH Cl. Apparently, this mode is not cooled efficiently even
2
2
In addition, the same angular anisotropy parameter is ob-
tained for Cl produced with He and Ar carrier gases.
in Ar or CF :He. We conclude that the ground and electroni-
4
cally excited states of CH Cl must have quite different C–Cl
2
Likewise, we believe that CH Cl is the main precursor of
the Cl( P3/2,1/2) products from the parallel transition at 240–
14 nm. The energy distributions of the Cl fragments can be
well fit by assuming that the precursor is ground state
bond lengths. The strong propensity to ‘‘hot band’’ excitation
suggests also that the excited state may be bound in the C–Cl
coordinate, with excitation accessing mainly the repulsive
wall. ‘‘Hot bands’’ resulting from excitation of other vibra-
2
2
2
Ϫ1
CH Cl, and that a large fraction of the Cl atoms are associ-
tional modes, such as the out-of-plane bend ͑ϳ390 cm ͒,
2
1
25,26
ated with singlet CH2(
˜
a A ) cofragments ͑Figs. 7 and 8͒.
cannot be excluded.
1
The angular distributions of the Cl atoms change gradually
as the dissociation wavelength decreases from 247 nm,
where a perpendicular transition is observed, to shorter
wavelengths. In the region 240–235 nm the value of ␤ is
speed dependent, ranging from ϳ0 at high Cl speeds to 1.2 at
low speeds, suggesting contributions from both parallel and
perpendicular transitions. At 219.5 and 214 nm, the parallel
The definitive assignment of the symmetry of the excited
state is obtained from the ab initio calculations of Levchenko
and Krylov.3 Using the EOM–CCSD method
28
with
6-311͑3ϩ,3ϩ͒G͑3df,3pd͒ basis set, they have assigned the
lowest excited states of CH Cl, and calculated their vertical
2
excitation energies and electronic transition moments ͑see
2
9
Table II in Ref. 3͒. The first allowed transition is to the
2
2
transition dominates, and the ␤ parameter is largely speed
1 A state, and there are no low-lying states of A symme-
1
2
Ϫ1
independent for E Ͼ2000 cm
.
try. The calculated value of the vertical excitation energy for
t
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:
137.112.220.85 On: Wed, 10 Dec 2014 18:27:02

7481
J. Chem. Phys., Vol. 115, No. 16, 22 October 2001
Photodissociation of CH Cl
2
2
the 1 A state is 4.92 eV, which corresponds to an excitation
are different. We assign the mixed transition in this region to
1
2
2
wavelength of 252 nm. This is where we expect the maxi-
mum in the absorption cross section, as well as the largest
product yield originating in vibrationally unexcited radicals.
This is confirmed by the results shown in Figs. 2 and 3.
excitations to the 1 A and the 2 B upper states. However,
1 1
at the shorter wavelengths studied here, the allowed transi-
2
tion to the 2 A (3s) state can also give rise to products with
1
a negative ␤. According to the calculations, the electronic
2
While at wavelengths у280 nm most of the CH and Cl
transition dipole moment to the 2 A state is two orders of
2
1
signals are attributed to ‘‘hot band’’ excitation, at 266 nm the
contribution of fragments with translational energies exceed-
ing Et is smaller, and this fraction decreases further at 247
magnitude smaller than the maximum value for excitation to
2
3
2
2
the 2 B state, but the 2 B � 1 B absorption cross sec-
1
1
1
max
tion in the 240–214 nm wavelength region reported here is
more than an order of magnitude smaller than the maximum
nm. In addition, by monitoring the Cl fragments while scan-
ning the photolysis energy we find a broad maximum in the
4
value. Also, differences in Franck–Condon factors between
2
Cl( P3/2,1/2) yield at 250 nm.
the transitions are not included in the transition dipole mo-
2
The ab initio calculations also show that the vertical
ment calculations. Thus, excitation to the 2 A1 state can
2
electronic transition dipole moment to the 1 A state is sig-
contribute to the reduction in the measured value of ␤ at the
shorter wavelengths. More detailed studies of the product
energy and angular distributions in the region 245–210 nm
are needed in order to separate the contributions of the pos-
sible excited states.
1
2
nificantly larger than that to the 2 A (3s) Rydberg state,
1
whose vertical maximum lies at 5.54 eV ͑224 nm͒. However,
the calculations do not take into account the vibrational
Franck–Condon factors and thus provide only upper limits.
2
Note that the 2 A state has a Rydberg character, and there-
In summary, the main absorption features responsible for
the products observed in the 312–214 nm region derive from
1
fore is expected to have a planar geometry similar to that of
the cation. Thus, its Franck–Condon factors for excitation
2
2
two valence states, the 1 A and the 2 B states, giving rise
1
1
from ‘‘cold’’ CH Cl may be more favorable than those for
to a perpendicular and a parallel transition, respectively.
2
2
2
the 1 A state, and its contribution, especially at the shortest
Some contribution from the 2 A (3s) Rydberg state cannot
1
1
wavelengths that still exhibit a negative ␤ parameter, cannot
be ruled out ͑see below͒.
be ruled out.
C. Photodissociation dynamics and product
electronic states
2
. Parallel transition
2
Cl( P3/2,1/2) fragments with large positive values of ␤
Further insight into the dissociation dynamics can be ob-
tained from the internal state distributions of the fragments.
are observed at 240–214 nm, and derive mainly from a par-
allel transition corresponding to an excited state of B sym-
metry in the C2v point group. Levchenko and Krylov have
found that the lowest excited B state has a vertical excita-
tion energy of 6.28 eV ͑corresponding to 197 nm excitation͒
and a large electronic transition dipole moment. This theo-
retical result agrees well with a previous measurement of the
absorption spectrum of CH Cl radical that shows a very
strong unstructured absorption band centered at 200 nm.
1
2
2
In the wavelength region of the 1 A � 1 B transition,
1
1
3
2
2
˜
CH (X B )ϩCl( P ) is the main dissociation channel,
and CH2(
˜a A ) and Cl( P ) appear only as minor prod-
1 1/2
2
1
3/2
1
1
2
2
2
ucts. In the region of the 2 B � 1 B transition, both
1
1
2
2
Cl( P
) and Cl( P ) are produced in comparable
1/2
3/2
1
amounts, and they are better correlated with the CH2(
1
˜a A )
cofragment.
2
4
The calculation of the vertical energies of the electronic
states allows us to compare our observations to a diabatic
state correlation diagram, as shown in Fig. 9. Several points
are noteworthy regarding this diagram. First, it does not take
into account possible curve crossings along the dissociation
pathways, and therefore should only be used as a reference
frame for comparison with the experimentally observed
product branching ratios. Second, the order of the states is
Thus, we believe that the assignment of the Cl fragments
2
2
with positive ␤ to the 2 B � 1 B transition is secure. Ac-
1
1
cording to the calculations, this state corresponds to ␲*
CCl
�
␲CCl excitation, i.e., promotion of a bonding C–Cl ␲ elec-
tron to an antibonding ␲* orbital, which would lend a repul-
sive nature to the potential energy surface in the C–Cl coor-
dinate, and is also known to give rise to strong transitions.
The maximum observed value of the ␤ parameter is
smaller than the maximum value of 2.0. There can be several
possible reasons for a reduction in ␤ other than dissociation
that is slower than a rotational period. For a parallel transi-
tion, these include simultaneous excitation of parallel and
perpendicular transitions, out-of-plane motions during the
dissociation, and tangential, in-plane, nuclear motions that
3
based on vertical excitation energies. If the geometries of
the excited states are very different, the order of the adiabatic
origins of the states may change. Third, the diagram is con-
structed only in C2v symmetry. If the symmetry is lowered to
C due to nonplanarity, avoided crossings would change the
s
correlations. Consequently, the differences between the ob-
servations and the predictions based on this diagram can only
be used to suggest possible interactions between different
states of the radical.
lead to ejection of products in a direction that is not parallel
to the dissociating bond.3
0–33
One or several of these factors
may contribute to the reduction in ␤ in CH Cl photodisso-
2
As seen from the diagram, both the ground and the first
ciation.
3
˜
As mentioned earlier, in the 240–235 nm photolysis re-
gion contributions from both parallel and perpendicular tran-
sitions are clearly discerned, because the speed distributions
of the Cl fragments corresponding to the two excited states
excited states of CH Cl correlate with CH (X B )
2
2
1
2
ϩCl( P3/2). This is in good agreement with the experimen-
tal results where ground-state products dominate. The paral-
2
lel transition, which involves the 2 B state, should give rise
1
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:
137.112.220.85 On: Wed, 10 Dec 2014 18:27:02

7482
J. Chem. Phys., Vol. 115, No. 16, 22 October 2001
Dribinski et al.
ergy surfaces are known, the probabilities of nonadiabatic
transitions should be evaluated as well.
Despite the potential complications in interpreting the
state correlation diagram, it is safe to suggest that the two
electronic states examined in this study give rise to direct
2
dissociation dynamics. The 1 A state involves promotion
1
of an electron from the singly occupied ␲*CCl orbital to an
antibonding ␴* orbital, which should weaken and lengthen
CCl
the C–Cl bond and give rise to repulsion in this coordinate.
This is evidenced by the large kinetic energy release in the
products, especially at wavelengths where ‘‘cold’’ reactants
are preferentially excited ͑e.g., 247 nm͒. In experiments with
Ar, where the final Cl product energy distributions are not as
broadened by the rovibrational distribution of the parent
CH Cl radical, we observe a large narrowing of the first,
2
high-energy, peak ͑Fig. 5͒. With Ar, most of the Cl products
produced at 266 and 247 nm have translational energies
close to the maximum allowed by the energetics. This large
kinetic energy release points toward a direct dissociation
mechanism taking place on the electronically excited state at
these wavelengths.
As discussed above, the parallel transition must also be
repulsive in the C–Cl coordinate, as confirmed by the ab-
sorption spectrum which is unstructured. For this transition,
photodissociation partitions significant fraction of the energy
FIG. 9. Diabatic correlation diagram for the dissociation of CH Cl to
1
2
in internal excitation of the CH2(
˜
a A ) product. This must
1
CH ϩCl. The vertical energies and assignments of the excited electronic
2
reflect either specific motions along the dissociation coordi-
nate, or geometrical differences between the CH moiety in
states of CH Cl are taken from Ref. 3. The excited states energies of CH are
2
2
taken from Ref. 7.
2
2
1
the 2 B state of CH Cl and the free CH (
˜
a A ) radical.
1
2
2
1
1
to CH2(
˜
a A ), also in agreement with our interpretation of
1
V. SUMMARY
the experimental Cl energy distributions at 219–214 nm.
Significant amounts of Cl fragments correlating energetically
with triplet CH are observed only in the region 240–235 nm
Based on our experimental data and the ab initio calcu-
lations presented in the accompanying paper, the following
conclusions regarding the photochemistry of the CH Cl radi-
cal at 312–214 nm can be reached.
2
2
2
where the 1 A and 2 B states are both excited, and they
2
1
1
2
2
most likely result from the 1 A � 1 B transition. In the
1
1
͑
1͒ CH Cl generated in a molecular beam via pulsed
parallel transition, both spin–orbit states or Cl are signifi-
cantly populated, while according to the correlation diagram
2
pyrolysis exhibits continuous, unstructured absorption
throughout the investigated region, giving rise to Cl atoms as
major products.
2
Cl( P1/2) should be preferentially produced. This, however,
is not surprising since the Cl spin–orbit states are separated
Ϫ1
͑
2͒ The lowest electronically excited state is assigned as
by only 881 cm , and long-range exit channel interactions
1 2
A in C symmetry, and is reached via a perpendicular
1
may lead to their mixing.34
2v
transition. Dissociation occurs on the repulsive part of the
The agreement in the major features of the observed
product branching with a simple diabatic state correlation
diagram does not necessarily mean that the products are
formed directly from each of these electronic states and that
surface crossings are unimportant. The excited radical may
experience surface crossings and nonadiabatic interactions
along the dissociation pathway that finally converge to the
same products predicted ‘‘diabatically.’’ A good example is
3
˜
potential energy surface terminating in CH (X B )
2
1
2
ϩCl( P3/2) as a major channel. The equilibrium geometry of
the excited state must be very different from that of the
ground state; in particular, the C–Cl bond length is different.
‘‘Hot band’’ absorption is particularly strong at wavelengths
Ͼ260 nm. The products are born with substantial transla-
tional energy.
the case of ClO dissociation35–38 ͑which is isoelectronic with
CH Cl͒, where a bound excited state correlating with O( D)
products is crossed in the exit channel by several repulsive
states, terminating in O( P) products.
In order to unravel the dissociation dynamics, more de-
tailed ab initio calculations are required. Such calculations
must be carried out not only in the vertical excitation region,
but also near the equilibrium geometry and in the region of
large internuclear separation. Once the adiabatic potential en-
͑3͒ The strongest transition is assigned as the 2 B
2
1
1
2
� 1 B parallel transition, and the upper state is repulsive.
2
1
This transition begins to dominate at ϳ240 nm, and the ab-
3
sorption cross section increases towards 200 nm. The major
products are CH2( ˜a A )ϩCl( P1/2,3/2), and CH2( ˜a A ) is
produced with broad internal excitation. ‘‘Hot band’’ absorp-
tion is much less important for this transition, indicating a
1
2
1
1
1
smaller difference in geometry between the ground and
2
2 B states.
1
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:
137.112.220.85 On: Wed, 10 Dec 2014 18:27:02

7483
J. Chem. Phys., Vol. 115, No. 16, 22 October 2001
Photodissociation of CH Cl
2
TABLE I. Calculated rotational energies for selected ͑N,K͒ levels of the two
Since CH is a near-prolate top in both the ground and
the 3p-Rydberg states, the energy levels for a prolate top are
2
˜
3
photon transition to the 3p Rydberg state of CH (X B ).
2
1
used in the calculations of the rotational energies with Bavg
⌬␻rot /cmϪ1
Erot /cmϪ1
͑
N,K͒
ϭ0.5(BϩC). The rotational selection rules for a two-photon
͑
͑
0,0͒
1,0͒
2,0͒
1,1͒
3,0͒
2,1͒
4,0͒
3,1͒
4,1͒
5,0͒
5,1͒
2,2͒
3,2͒
4,2͒
5,2͒
3,3͒
4,3͒
5,3͒
6,3͒
4,4͒
5,4͒
6,4͒
5,5͒
6,5͒
0.0
Ϫ0.8
Ϫ2.4
Ϫ0.3
Ϫ4.9
Ϫ2.0
Ϫ8.1
Ϫ4.4
Ϫ7.6
Ϫ12.1
Ϫ11.7
Ϫ0.5
Ϫ2.9
Ϫ6.2
Ϫ10.2
Ϫ0.6
Ϫ3.8
Ϫ7.8
Ϫ12.7
Ϫ0.5
Ϫ4.5
Ϫ9.4
Ϫ0.2
Ϫ5.1
0
15
44
81
89
transition are: ⌬Nϭ0, Ϯ1, Ϯ2; ⌬Kϭ0, Ϯ1, Ϯ2. If we con-
Q
sider only transitions with ⌬Nϭ0, ⌬Kϭ0 ͑ Q branch͒, then
͑
͑
͑
͑
͑
͑
͑
͑
͑
͑
͑
͑
͑
͑
͑
͑
͑
͑
͑
͑
͑
͑
the rotational contribution to the total rovibronic transition
energy is given by
111
148
155
215
222
289
310
355
414
488
687
746
821
910
1212
1286
1375
1884
1973
⌬
␻ ϭ͑BЈ ϪBЉ ͒N͑Nϩ1͒
rot
avg avg
2
ϩ͓͑AЈϪBЈ ͒Ϫ͑AЉϪBЉ ͔͒K .
avg
avg
Table I shows ⌬␻rot for transitions originating from selected
N, K͒ levels, along with their rotational energies. Clearly,
͑
Ϫ1
many rotational levels have ⌬␻rot within 10 cm . Taking
into account the fact that the transitions are broadened by
intramolecular interactions,19 we expect simultaneous excita-
tion of these levels with the detection at 311.8 nm.
1
B. J. Finlayson-Pitts and J. N. Pitts, Jr., Chemistry of the Upper and Lower
Atmosphere: Theory, Experiments, and Applications ͑Academic, New
York, 2000͒.
C. Sandorfy, Top. Curr. Chem. 86, 91 ͑1979͒.
2
3
S. V. Levchenko and A. I. Krylov, J. Chem. Phys. 115, 7485 ͑2001͒,
following paper.
4
P. B. Roussel, P. D. Lightfoot, F. Caralp, V. Catoire, R. Lesclaux, and W.
Forst, J. Chem. Soc., Faraday Trans. 87, 2367 ͑1991͒.
Y. Endo, S. Saito, and E. Hirota, Can. J. Phys. 62, 1347 ͑1984͒.
S. G. Lias, J. E. Bartmess, J. F. Liebman, J. L. Holmes, R. D. Levin, and
5
6
2
The participation of the 2 A (3s) state in the dissocia-
1
W. G. Mallard, Gas-Phase Ion and Neutral Thermochemistry ͓J. Phys.
Chem. Ref. Data, Suppl. 17, 1 ͑1988͔͒.
NIST Chemistry WebBook: NIST Standard Reference Database No. 69
tion is not yet clear, as is the role of possible surface cross-
ings. More theoretical work on the ground and excited state
potential energy surfaces is clearly needed. On the experi-
mental front, a search of other dissociation channels is now
in progress.
7
͑
February 2000 Release͒; http://webbook.nist.gov/chemistry/
8
J. J. DeCorpo, D. A. Bafus, and J. L. Franklin, J. Chem. Thermodyn. 3,
25 ͑1971͒.
1
9
0
M. Weissman and S. W. Benson, J. Phys. Chem. 87, 243 ͑1983͒.
E. Tschuikow-Roux and S. Paddison, Int. J. Chem. Kinet. 19, 15 ͑1987͒.
J. L. Holmes and F. P. Lossing, J. Am. Chem. Soc. 110, 7343 ͑1988͒.
J. A. Seetula, J. Chem. Soc., Faraday Trans. 92, 3069 ͑1996͒.
J. A. Seetula, Phys. Chem. Chem. Phys. 2, 3807 ͑2000͒.
1
11
ACKNOWLEDGMENTS
1
1
1
2
3
4
Support by the National Science Foundation and the Do-
nors of the Petroleum Research Fund, administered by the
American Chemical Society is gratefully acknowledged. The
authors benefited greatly from discussions with Anna Krylov,
Sergei Levchenko, and Pavel Jungwirth.
D. W. Kohn, H. Clauberg, and P. Chen, Rev. Sci. Instrum. 63, 4003
͑
1992͒.
15
E. Kolodney and A. Amirav, Chem. Phys. 82, 269 ͑1983͒.
Th. Droz-Georget, M. Zyrianov, H. Reisler, and D. Chandler, Chem. Phys.
Lett. 276, 316 ͑1997͒.
16
17
M. Zyrianov, A. Sanov, Th. Droz-Georget, and H. Reisler, J. Chem. Phys.
1
10, 10774 ͑1999͒.
APPENDIX: REMPI SPECTROSCOPY
18
19
C. E. Moore, Atomic Energy Levels v.I ͑NSRDS-NBS, 1971͒.
K. K. Irikura, R. D. Johnson III, and J. W. Hudgens, J. Phys. Chem. 96,
3
˜
OF THE CH „X B … RADICAL
2
1
6131 ͑1992͒.
20
In order to calculate the positions of the rotational lines
R. N. Bracewell, The Fourier Transform and Its Applications ͑McGraw-
Hill, New York, 1986͒.
K. Mikhaylichenko, C. Riehn, L. Valachovic, A. Sanov, and C. Wittig, J.
Chem. Phys. 105, 6807 ͑1996͒.
M. R. Cameron and S. H. Kable, Rev. Sci. Instrum. 67, 283 ͑1996͒.
A. V. Demyanenko, V. Dribinski, A. B. Potter, and H. Reisler ͑unpub-
lished͒.
R. D. Johnson III, J. Chem. Phys. 96, 4073 ͑1992͒.
M. E. Jacox and D. E. Milligan, J. Chem. Phys. 53, 2688 ͑1970͒.
L. Andrews and D. W. Smith, J. Chem. Phys. 53, 2956 ͑1970͒.
F. Temps, H. Gg. Wagner, and M. Wolf, Z. Naturforsch., A: Phys. Sci. 47,
3
˜
of CH (X B ) it is necessary to know the rotational con-
2
1
21
stants for the ground and excited electronic states of the radi-
3
39
22
23
cal. For the ground B state, the rotational constants are
1
AЉϭ73.811 cmϪ1, BЉϭ8.45 cm , CЉϭ7.184 cm . The ro-
tational constants for the excited 3p-Rydberg state of CH2
are not known. However, since the geometry of this state is
Ϫ1
Ϫ1
24
25
26
expected to be close to that of the ground state of the cation
ϩ
2
27
28
29
CH , the rotational constants for the 3p-Rydberg state can
6
͑
3
60 ͑1992͒.
a͒ H. Koch, H. Jørgen, Aa. Jensen, and P. Jørgensen, J. Chem. Phys. 93,
345 ͑1990͒; ͑b͒ J. F. Stanton and R. J. Bartlett, ibid. 98, 7029 ͑1993͒.
be evaluated using the same geometry. Geometry optimiza-
tion was performed with the Q-Chem program package us-
4
0
4
1
ing the DFT method with B3LYP functionals in a 6-311
ϩ,ϩ͒G͑3df,3pd͒ basis set.42,43 The values obtained ͓r͑CH͒
ϭ1.097 Å, HCH angleϭ141.36°͔ agree well with previous
The axis system in the C2v notation used here is the one customary in
spectroscopy, cf. J. M. Hollas, High Resolution Spectroscopy ͑Wiley, New
York, 1998͒, in which the x axis is out-of-plane, and the z axis is along the
C–Cl bond. Note that in the accompanying paper ͑Ref. 3͒, a different axis
system, common in ab initio calculations, is used in which the y axis is
out-of-plane. However, a nonstandard table of C2v characters is used in
͑
4
4
calculations. The corresponding rotational constants are
Ϫ1
Ϫ1
Ϫ1
AЈϭ73.883 cm , BЈϭ7.783 cm , CЈϭ7.042 cm
.
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:
137.112.220.85 On: Wed, 10 Dec 2014 18:27:02

7484
J. Chem. Phys., Vol. 115, No. 16, 22 October 2001
Dribinski et al.
3
7
Ref. 3 in order to keep the spectroscopic assignments the same as those
customary in spectroscopy.
C. Jonah, J. Chem. Phys. 55, 1915 ͑1971͒.
S. Yang and R. Bersohn, J. Chem. Phys. 61, 4400 ͑1974͒.
H.-P. Loock, J. Cao, and C. X. W. Qian, Chem. Phys. Lett. 206, 422
I. C. Lane, W. H. Howie, and A. J. Orr-Ewing, Phys. Chem. Chem. Phys.
1, 3087 ͑1999͒.
3
3
3
0
38
39
40
A. Toniolo, M. Persico, and D. Pitea, J. Chem. Phys. 112, 2790 ͑2000͒.
M. E. Jacox, J. Phys. Chem. Ref. Data, 1 ͑1994͒, Monograph No. 3.
J. Kong et al., J. Comput. Chem. 21, 1532 ͑2000͒.
P. J. Stephens, F. J. Devlin, C. F. Chabalowski, and M. J. Frisch, J. Phys.
Chem. 98, 11623 ͑1994͒.
M. J. Frisch, J. A. Pople, and J. S. Binkley, J. Phys. Chem. 80, 3265
͑1984͒.
T. Clark, J. Chandrasekhar, and P. V. R. Schleyer, J. Comput. Chem. 4, 294
͑1983͒.
1
2
41
͑1993͒.
33
A. V. Demyanenko, V. Dribinski, H. Reisler, H. Meyer, and C. X. W. Qian,
J. Chem. Phys. 111, 7383 ͑1999͒.
J. Cao, Y. Wang, and C. X. W. Qian, J. Chem. Phys. 103, 9653 ͑1995͒.
H. Floyd Davis and Y. T. Lee, J. Phys. Chem. 100, 30 ͑1996͒.
S. Schmidt, Th. Benter, and R. N. Schindler, Chem. Phys. Lett. 282, 292
42
3
3
3
4
5
6
43
44
͑1998͒.
P. R. Bunker, Can. J. Phys. 72, 871 ͑1994͒.
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:
37.112.220.85 On: Wed, 10 Dec 2014 18:27:02
1