Photodissociation of carbonic dibromide at 267 nm: Observation of three-body dissociation and molecular elimination of Br2

Article abstract of DOI:10.1063/1.1509063

The photodissociation of Br₂CO at 267 nm was studied by ion velocity imaging and TOF-MS method. Bimodal kinetic energy distributions of bromine atoms were found. In general, the results indicate that the transition dipole moment of Br₂

Full text of DOI:10.1063/1.1509063

Photodissociation of carbonic dibromide at 267 nm: Observation of three-body
dissociation and molecular elimination of Br 2
Dadong Xu, Jianhua Huang, Joseph S. Francisco, Jaron C. Hansen, and William M. Jackson
Citation: The Journal of Chemical Physics 117, 7483 (2002); doi: 10.1063/1.1509063
View online: http://dx.doi.org/10.1063/1.1509063
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/117/16?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Photodissociation dynamics of C Br 4 at 267 nm by means of ion velocity imaging
J. Chem. Phys. 125, 133311 (2006); 10.1063/1.2213260
Velocity map imaging study of BrCl photodissociation at 467 nm: Determination of all odd-rank ( K = 1 and 3)
anisotropy parameters for the Cl ( P 3 ∕ 2 0 2 ) photofragments
J. Chem. Phys. 124, 094305 (2006); 10.1063/1.2168149
One photon dissociation and multiphoton dissociative ionization of bromochlorodifluoromethane ( CF 2 BrCl ) at
267 nm region
J. Chem. Phys. 119, 3661 (2003); 10.1063/1.1591728
Ultraviolet photodissociation of bromoform at 234 and 267 nm by means of ion velocity imaging
J. Chem. Phys. 117, 2578 (2002); 10.1063/1.1491877
Dynamics of vibrationally mediated photodissociation of CH 3 CFCl 2
J. Chem. Phys. 115, 6418 (2001); 10.1063/1.1404392
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:
131.111.164.128 On: Sun, 21 Dec 2014 02:48:14

JOURNAL OF CHEMICAL PHYSICS
VOLUME 117, NUMBER 16
22 OCTOBER 2002
Photodissociation of carbonic dibromide at 267 nm:
Observation of three-body dissociation and molecular elimination of Br2
Dadong Xu and Jianhua Huang
Department of Chemistry, University of California, Davis, California 96516
Joseph S. Francisco and Jaron C. Hansen
Department of Chemistry and Department of Earth and Atmospheric Sciences, Purdue University,
West Lafayette, Indiana 47907
William M. Jacksona)
Department of Chemistry, University of California, Davis, California 96516
͑Received 29 April 2002; accepted 2 August 2002͒
The photodissociation of Br2CO around 267 nm has been studied by time-of-flight mass
*
spectroscopy and ion velocity imaging. The atomic ͑Br and Br ͒ and molecular products (Br2 and
BrCO͒ are detected via multiphoton ionization with the same laser. The results show that the
molecule dissociates into ͑1͒ Br͑fast͒ϩBr͑slow͒ϩCO via an asynchronously concerted three-body
decay process for both ground and spin–orbit excited bromine atoms, ͑2͒ BrCO(A)ϩBr, and ͑3͒
Br2ϩCO, the molecular elimination channel. The translational energy distributions of bromine
atoms from reaction ͑1͒ are bimodal. For both spin–orbit states the anisotropy parameters differ
clearly for slow and fast bromine atoms, where the ␤ values for slow bromine atoms decrease
relative to those for fast atoms. The ␤ values for the Br2 elimination channel almost reach the low
limit of Ϫ1. Taking into account the translational energy and angular distributions of these reactions,
an asynchronously concerted decay mechanism could be proposed for the three-body dissociation.
It is concluded that the transition dipole moment is in the direction perpendicular to the CvO bond
in the initial excitation, which is also consistent with all the observations for reactions ͑2͒ and ͑3͒.
© 2002 American Institute of Physics. ͓DOI: 10.1063/1.1509063͔
C–Br coordinate.10 This suggests that the photodissociation
I. INTRODUCTION
of Br2CO should perhaps be dominated by C–Br fission.
By comparison, the photodissociation of phosgene
(Cl2CO) has received considerable attention11–15 in the past
decades, not only due to its importance in atmospheric chem-
istry, but also due to its being an excellent model molecule
for studying three-body dissociation dynamics. In a three-
body decay process involving two identical bond breakages
there are usually three different cases, in which the second
bond scission occurs at different times relative to the first
one.16 One extreme is a synchronously concerted decay, in
which the two bonds break almost simultaneously and the
kinetic energy and angular distributions of the two identical
fragments are indistinguishable. An example of this case is
the photodissociation of s-Tetrazine, C2N4H2→N2ϩ2HCN,
at 248 and 551 nm.17,18 Another extreme is a sequential, or
stepwise, decay, in which a bimodal kinetic energy distribu-
tion usually arises from the dissociation and there is no an-
gular distribution coupling between the two identical frag-
ments. In this case the lifetime of the larger fragment from
the first bond scission is much longer than its rotational pe-
riod. Examples of this case are the dissociation of acetone19
and dimethyl sulfoxide.20 The third case is asynchronously
concerted decay where a bimodal energy distribution is usu-
ally found and the angular distributions of the two fragments
are coupled. Dissociation of azomethane at 351 nm ͑Ref. 21͒
fits in this category. These classifications also apply to three-
body dissociation involving breakage of two different bond
The release of bromine-containing species into the atmo-
sphere has raised concern regarding their potential impact on
stratospheric ozone abundances.1 Bromine released into the
atmosphere in the form of bromine atoms plays significant
roles in the bromine catalytic cycles for the removal of
ozone.2,3 Consequently, it is important to understand how
bromine is released from bromine-containing organic com-
pounds. Methyl bromide, CH3Br, is a major bromine-
containing compound that contributes half of the total or-
ganic bromines.4,5 Bromoform, CHBr3 , which accounts for
only 3% of the total bromine, accounts for at least 9% of the
reactive bromine in the stratosphere.6 Dvortsov et al. found
that bromoform contributes more inorganic bromine to the
stratosphere than all the combined sources of halons and me-
thyl bromide.7 The mechanism for how bromine is released
from bromoform is not clear at all. Gas phase studies that
have examined how bromoform is oxidized by OH and Cl
species in the atmosphere have shown that at least one bro-
mine atom is released followed by the production of carbonic
dibromide, Br2CO.8,9 However, there are no studies that
have characterized the atmospheric consequences of Br2CO.
Recently, theoretical studies of excited states of Br2CO show
that the low-lying excited states are predissociative along the
a͒
Author to whom correspondence should be addressed. Electronic mail:
wmjackson@ucdavis.edu
0021-9606/2002/117(16)/7483/8/$19.00
7483
© 2002 American Institute of Physics
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
131.111.164.128 On: Sun, 21 Dec 2014 02:48:14

7484
J. Chem. Phys., Vol. 117, No. 16, 22 October 2002
Xu et al.
sition. If it did occur before the jet expansion, Br2 should be
cooled down in the molecular beam. As a result, the image of
Br2 should be very small and centered in the middle of the
image since Br2 would only have little transverse speed in
the beam due to the supersonic expansion. The image of Br2
presented in the next section indicates no such component at
all. In addition, photodissociation of Br2 has been studied at
265 nm ͑Ref. 27͒ and the energy and angular distributions of
types, where the three products are usually different. A rather
complete compilation of photoinduced three-body decay can
be found in Ref. 16. In a most recent study on phosgene at
235 nm,22 where the dependence of the anisotropy parameter
␤ on velocity is monitored in addition to the measurement of
velocities of both Cl(2P3/2) and Cl(2P1/2), Einfeld et al. con-
cluded that the dissociation of phosgene could be explained
by a sequential decay.
*
Br and Br are totally different from the present results on Br
In this paper, we provide details on the photodissociation
mechanisms of Br2CO at 267 nm using ion velocity imaging
and time-of-flight mass spectrometry ͑TOF-MS͒ methods.
The results suggest that Br2CO undergoes asynchronously
*
and Br . This is further evidence that there was no appre-
ciable decomposition of Br2CO before the laser strikes the
molecular beam and that the products we observed have re-
sulted from photodissociation of Br2CO.
concerted three-body decay where Br(2P3/2) and Br(2P1/2
)
Other molecules, which can be dissociated to produce
fragments with known kinetic energy distributions, are used
in the molecular beam to calibrate the ion imaging apparatus.
In our case ethyl bromide is used to obtain the calibration
factor.28 One of the advantages of ion imaging is that once
the apparatus is calibrated at a particular voltage setting on
the ion optics the same calibration factor applies to all other
species since the radius ͑R͒; i.e., the size of the image is an
explicit function of the recoil kinetic energy (ET), R
ϰ(ET)1/2, of the fragment. The modification introduced by
Eppink and Parker,29 termed velocity map imaging, has made
great improvements upon the energy resolution of this tech-
nique.
are released from both dissociation steps. The formation of
an electronic excited state, BrCO(A), is also proposed. Fur-
thermore, we report the first evidence of the Br2 elimination
channel in the photodissociation of Br2CO. No studies to
date on X2CO ͑where XϭF, Cl, and Br͒ molecules have re-
ported the direct observation of this channel in the UV re-
gion.
II. EXPERIMENT
The ion velocity imaging apparatus has been described
in detail elsewhere.23 Briefly, the apparatus consists of three
stainless steel chambers. The first chamber is the molecular
beam source chamber that is pumped with a 400-l/s molecu-
lar pump. A background pressure of 5ϫ10Ϫ6 Pa is achieved
in this chamber. A piezoelectric pulsed valve with a 0.25-mm
nozzle orifice is used to produce the pulsed molecular beam
in these studies. The second chamber is the reaction chamber
containing the ion optics, and it is pumped by a 500-l/s tur-
bomolecular pump. The pressure in this chamber is typically
4ϫ10Ϫ5 Pa during the experiments. The laser beam enters
the reaction chamber at 90° to the molecular beam in the
center of the ion optics. The ion optics collect and focus ions
into a flight tube that is perpendicular to the plane defined by
the laser and molecular beams. The flight tube, the third
chamber, is pumped by a turbopump that keeps the pressure
in the TOF-MS in the 8ϫ10Ϫ7 Pa range. A pulsed reactant
beam is prepared by expanding a Br2CO/He ͑1%͒ mixture of
200 Torr through the pulsed valve into the vacuum chamber.
The Br2CO is prepared by the synthesis outlined
previously.24,25 The sample is purified by trap-to-trap distil-
lation. The purity of the sample is checked by Fourier trans-
form IR ͑FTIR͒ and UV absorption measurements. No im-
purities are detected. Photodissociation is initiated by using a
UV laser at 267 nm obtained by doubling the 534-nm output
of a Spectra Physics Pro 230 Nd-YAG pumped dye laser
͑Coumarin 500͒. The 267-nm laser light ͑ϳ0.2 mJ/pulse͒ is
focused on the molecular beam with a quartz lens with a
focal length fϭ50 cm. The ground, Br(2P3/2), and excited,
Br(2P1/2), state atoms are probed via 2ϩ1 resonance-
enhanced multiphoton ionization ͑REMPI͒ at 266.65 and
266.71 nm, respectively.26 The dye laser wavelength was
scanned to cover the Doppler profile of Br atoms. The ex-
periments are run at a repetition rate of 30 Hz.
III. RESULTS AND DISCUSSION
The dissociation channels that are energetically possible
for Br2CO at 267 nm are
Br2COϩh␯267 nm→COϩ2Brϩ57.3 kcal/mol
→BrCOϩBrϩ57.6 kcal/mol
͑1͒
͑2͒
͑3͒
→COϩBr2ϩ103.5 kcal/mol.
The heats of reaction for the above reactions are based on
recent ab initio studies of the heats of formation of Br2CO
and BrCO using a complete basis set30 and the NIST
webbook.31 The heats of reactions ͑1͒ and ͑2͒ are for the
ground state of Br(2P3/2) and BrCO, respectively. The ex-
cited state Br(2P1/2), which is 10.5 kcal/mol higher in energy
than the ground state, is also detected in the experiments.
A. TOF studies
Figure 1 shows the TOF spectra observed when the laser
is tuned on and off the resonance line of the bromine atom.
The observed m/e peaks are due to masses 79 and 81 for
Brϩ and a small abundance of 107 and 109 for BrCOϩ, as
well as 158, 160, and 162 for Br2ϩ. The double peaks for
Brϩ in the TOF spectra arise from the two isotopes of bro-
mine, 79Br and 81Br, which have roughly similar natural
abundances. As shown in Fig. 1, the Brϩ signal is dramati-
cally increased when the laser is resonant with the electronic
state of the bromine atom. This is due to the REMPI of the
bromine atoms produced from reactions ͑1͒ and ͑2͒.
The decomposition of Br2CO in the gas line or pulsed
valve can be eliminated based on the following facts. The
most likely products would be COϩBr2 from the decompo-
The presence of BrCOϩ in the TOF-MS was quite a
surprise. Any ground-state BrCO produced from reaction ͑2͒
should decompose into CO and another Br atom almost si-
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:
131.111.164.128 On: Sun, 21 Dec 2014 02:48:14

J. Chem. Phys., Vol. 117, No. 16, 22 October 2002
Photodissociation of carbonic dibromide
7485
nm͒,34,35 where the excited molecule should have a fairly
long lifetime. This would allow the molecule to efficiently
absorb another photon to become ionized. The proposed
mechanism can be further confirmed by analyzing the kinetic
energy release and angular distribution for each channel.
B. Radical channel: BrCO„A…¿Br
As pointed out in the previous section, BrCO should be
formed in the electronic excited BrCO(A) state to explain
how it lives long enough to undergo MPI given the 57.6
kcal/mol of available energy for the ground state BrCO(X).
We can also discount the possibility that BrCOϩ is formed
via multiphoton dissociative ionization. By analogy to ClCO,
whose first excited state lies at about 1 eV above the ground
state,36 the observed BrCOϩ should be formed from
BrCO(A) state via multiphoton ionization.
FIG. 1. Time-of-flight spectra at 267 nm taken around the wavelength for
the detection of Br(2P3/2) via 2ϩ1 REMPI. Similar spectra can be obtained
2
*
for Br ( P1/2).
Figure 2 shows the raw two-dimensional ͑2D͒ image of
BrCOϩ, the translational energy distribution P(ET) and the
anisotropy parameter distribution ␤(ET) as a function of ET .
The image clearly shows that the dissociation is fast and the
initial excitation is parallel in nature. The maximum total
translational energy reaches at ϳ16 kcal/mol. Assuming the
BrCO(A) is ϳ23 kcal/mol above the ground state, as it is in
ClCO, it may be inferred that the well depth of this state
could be around 19 kcal/mol along the Br–C bond. It is
likely that some available energy will be consumed to excite
the C-O bond, leading to a binding energy lower than 19
kcal/mol for the BrCO(A) state. This is certainly not an un-
reasonable value for the BrCO(A) state. Similar or much
higher binding energies are found in excimer molecules
where the ground state is unstable and the excited state is
bound. For example, the first excited-state binding energies
of the LiHg and NaCd excimers are about 24 kcal/mol ͑Ref.
37͒ and 20 kcal/mol ͑Ref. 38͒. In a laser-assisted reaction
study of XeϩCl2 , Kohel and Keto39 found from the fluores-
multaneously because of the weakness of the Br-CO bond
and the large amount of excess energy. Ab initio calculations
indicate that ground-state BrCO is very weakly bound ͑only
0.3 kcal/mol for the Br-CO bond energy͒.30 However, the
first excited state should be more stable by analogy with
ClCO. The first excited state of ClCO radical is strongly
bound and is about ϳ1 eV above the ground state.32 Simi-
larly, a real excited state of BrCO(A) should have a fairly
long lifetime, and we think that the observed BrCOϩ is pro-
duced from this state via multiphoton ionization. One possi-
bility is that this ion is produced by multiphoton dissociative
ionization of Br2CO involving three photons at 267 nm, the
total energy of which is above the ionization potential of
Br2CO ͑IPϭ10.8 eV͒. In the TOF spectrum, there is no evi-
dence for Br2COϩ. However, electron impact and chemical
ionization mass spectrometer experiments33 on Br2CO indi-
ϩ
cate that Br2 ions are not observed over the energy range
from 20 to 50 eV, but Br2COϩ ions are. This shows that
Br2COϩ is a stable ion. On the other hand, in our experi-
ment, the 267-nm laser energy is about 0.2 mJ/pulse and the
focal length of the lens is 50 cm. Under these conditions, it
can be estimated that the rate of the one-photon production
of Br atoms through reaction ͑2͒ is at least two orders of
magnitude higher than the rate of formation of these atoms
via three-photon dissociative ionization of Br2CO. Further-
more, as shown in the next section, the similar angular dis-
tributions of BrCOϩ and Brϩ suggest that both BrCO and Br
result from the same excitation. If one assumes that dissocia-
tive ionization of Br2CO produces BrCOϩϩBrϩeϪ and Br
is subsequently ionized via REMPI, then a total of six pho-
tons will be needed to produce Brϩ. If the REMPI process
converts all of the Br atoms to Brϩ ions, then the Brϩ signal
should be about the same size as the BrCOϩ signal since an
equal amount of Br and BrCOϩ would be produced from the
dissociative ionization. In fact, as shown in Fig. 1, the Brϩ
signal is about 600 times stronger than BrCOϩ. A possible
explanation for this large difference in signal intensity is that
ionization of Br atoms via REMPI is far more efficient than
the ionization of BrCO(A) radicals.
*
cence of the XeCl excimer around 308 nm that average
vibrational energy E of 1.1 eV and vibrational state
v
Ј
͗
͘
v
Ͼ100 can be observed for this excited state, which means
*
the well depth of XeCl is much deeper than 1 eV. High
level calculations are needed to confirm the well depth of the
BrCO(A) state.
In the photodissociation of phosgene14–16,22 the radical
channel forming ClCO has not been observed, although the
binding energy for the ClCO(X) is 16.7 kcal/mol,14 which is
much greater than the value of 0.3 kcal/mol for BrCO(X).
The well depth for ClCO(A) is also probably deeper than
that for BrCO(A). The fact that ClCO has not been observed
can probably be attributed to the fact that the dissociation of
phosgene was done around 235 nm, where there is more
available energy. It could also be due to very low ionization
efficiency for ClCO(A), which may dissociate too rapidly
after absorbing a 235-nm photon to be ionized.
The anisotropy parameter ␤(ET) is almost constant
throughout the translational energy ET region and is close to
the theoretical limit from the structure calculations in the
latter section. This limit is reduced from 2 because Br2CO is
nonlinear and there is an inherent angle between the product
recoil and transition dipole moment. The ␤ values indicate
that the lifetime of the initial excited state of Br2CO is very
ϩ
For the Br2 signal, we note that Br2 has a weak absorp-
tion band in the 267-nm region and the addition of a second
photon puts the molecule up in the Rydberg region ͑near 133
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:
131.111.164.128 On: Sun, 21 Dec 2014 02:48:14

7486
J. Chem. Phys., Vol. 117, No. 16, 22 October 2002
Xu et al.
ized. We therefore suggest that BrCO(A) must be formed
only with Br atoms in the ground state.
C. Atomic channel: Br„fast…¿Br„slow…¿CO
2
Both bromine atoms Br(2P3/2) and Br ( P1/2) are de-
*
tected via 2ϩ1 resonance enhanced multiphoton ionization
using the transitions identified in Ref. 26. The small differ-
ence of 0.06 nm in detection wavelengths should have no
effect on the dissociation dynamics. The images, energy, and
*
angular distributions for Br and Br are shown in Figs. 3͑a͒
and 3͑b͒, respectively. It should be noted that the transla-
tional energy shown in the figure is for the bromine atoms,
not the total kinetic energy, in the center-of-mass frame. Ob-
viously, the energy distribution in Fig. 3͑b͒ is bimodal for
*
Br . There is also a shoulder at 6 kcal/mol for Br in Fig.
3͑a͒, although it is less obvious. The bimodal distribution
indicates that the cleavage of the two Br–C bonds produces
*
bromine atoms ͑both Br and Br ͒ at different speeds, which
implies that the dissociation is not a synchronous or symmet-
ric process.
The translational energy distribution shown in Fig. 3͑a͒
for ground-state Br contains some contribution from reaction
͑2͒, as discussed in Sec. III B. The energy distributions of Br
from reactions ͑1͒ and ͑2͒ are reproduced in Fig. 4͑a͒, where
the latter ͑dotted line͒ is deduced from the distribution shown
in Fig. 2 by applying the mass factor M(BrCO)/M(Br2CO)
to the energy axis. This factor is determined by the momen-
tum conservation of the two products, m1 1ϭm2 2 . The ␤
v
v
distributions from Figs. 2 and 3 are reprinted in Fig. 4͑b͒ to
make it convenient for comparison. Using the following
equations, we can estimate the relative contribution of Br
*
from reaction ͑2͒ by assuming that Br and Br from reaction
͑1͒ have similar ␤ parameters as a function of translational
energy:
P1ϫ␤1ϩP2ϫ␤2ϭ␤,
P1ϩP2ϭ1,
͑4͒
͑5͒
FIG. 2. Two-dimensional image of BrCO, total translational energy distri-
bution P(ET), and angular distribution ␤(ET). The arrow indicates the laser
polarization direction.
and therefore
P ϭ ␤Ϫ␤ ͒/ ␤ Ϫ␤ ͒,
͑6͒
͑
͑
2
2
1
1
where P1 is the contribution of Br with slow velocity from
reaction ͑1͒, and P2 is the contribution of Br from reaction
͑2͒. Here ␤ is the composite anisotropy parameter shown as
short compared to its rotational period. The short lifetime
will prevent the molecule from absorbing more photons be-
fore it breaks apart. This is in accordance with our earlier
proposal that multiphoton dissociative ionization can be dis-
counted. The ␤ parameters can be used to separate the con-
tributions of Br from reactions ͑1͒ and ͑2͒ by assuming that
only Br(2P3/2) is formed with BrCO(A) in reaction ͑2͒. The
Br(2P1/2) and BrCO(A) may correlate with a higher molecu-
lar Br2CO excited state that has no crossing with the Br2CO
*
the thick line in Fig. 4͑b͒; ␤1 is the thin line for Br and ␤2
is the dotted line for Br from reaction ͑2͒ in Fig. 4͑b͒. Using
the average ␤’s from 1 to 3 kcal/mol to determine P2 a value
of 0.4Ϯ0.1 is obtained, and from Eq. ͑5͒, P1ϭ0.6Ϯ0.1. The
deconvolution is shown in Fig. 4͑a͒. This ratio ͑0.4/0.6͒ is a
measure of the contributions from ͑2͒ and ͑1͒ to the produc-
tion of ground-state Br(2P3/2). We note that this ratio is con-
siderably different from the relative intensities of BrCOϩ to
Brϩ in Fig. 1. As stated earlier, the relative intensities on the
TOF spectrum depend strongly on the ionization efficiencies
of the two species and, hence, cannot be used to predict the
relative contributions.
(1 1A ) in the first excited state. This first singlet excited
Љ
*
state of Br2CO results from the nBr,O,Br→␲
transition.10
C–O
Earlier we estimated that the well depth for the BrCO(A)
2
*
state is Ͻ19 kcal/mol. The energy of Br ( P1/2) is 10.5
kcal/mol above the energy of Br(2P3/2), and if it is formed
with BrCO(A), then the well depth of this state will be less
than 8.5 kcal/mol, which seems too low to stabilize the radi-
cal against dissociation so that it can subsequently be ion-
*
The ␤(ET) distributions for Br and Br shown in Fig.
4͑b͒ demonstrate the same trend and merge together at ET
Ͼ4 kcal/mol. At lower energies ␤ is larger for Br than for
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:
131.111.164.128 On: Sun, 21 Dec 2014 02:48:14

J. Chem. Phys., Vol. 117, No. 16, 22 October 2002
Photodissociation of carbonic dibromide
7487
2
FIG. 3. Two-dimensional images, translational energy, and angular distributions for ͑a͒ Br(2P3/2) atoms and ͑b͒ Br ( P1/2) atoms. The arrow indicates the
*
laser polarization direction.
*
Br as a result of additional contribution from channel ͑2͒,
The difference between Br2CO and Cl2CO is that for Cl2CO
the ␤ parameter goes down to zero in the low kinetic energy
portion for both spin–orbit states. This suggests that the sec-
ond bond cleavage occurs at a later time when the initial
alignment is washed out through rotation. The dissociation is
described by a sequential decay.22 The C–Cl bond strength in
ClCO is 16.7 kcal/mol, which is substantially stronger than
the C–Br bond in BrCO. This deep well may help ClCO to
survive its rotation, which results in an isotropic angular dis-
tribution for the slow Cl.
which has a fairly constant ␤(ET) distribution. It is clear that
␤(ET)’s for the fast components of Br and Br are larger
*
than those for the slow components. The nonvanishing
anisotropy distributions suggest both Br–C bonds break in a
fraction of rotational periods of Br2CO and BrCO, and
the angular distribution of the second bromine is still coupled
with the angular distribution of the first bromine. While the
two bromine atoms are formed with different speeds,
this three-body decay can be best described as an asynchro-
nously concerted reaction. It is interesting to note that each
*
of the two bond cleavages produces both Br and Br . But
D. Molecular elimination channel: Br2¿CO
it is not feasible to distinguish the contributions from
each of the combinations, namely, Br͑fast͒ϩBr͑slow͒,
ϩ
Br2 is also observed in the TOF spectrum shown in
*
*
*
Br͑fast͒ϩBr ͑slow͒, Br ͑fast͒ϩBr͑slow͒, and Br ͑fast͒
Fig. 1. It is likely that this molecule is ionized through mul-
tistep REMPI via its Rydberg state. The image, translational
energy, and angular distributions of Br2 are shown in Fig.
*
ϩBr ͑slow͒.
In the photodissociation of phosgene at 235 nm ͑Ref. 22͒
ϩ
*
bimodal energy distributions are also found for Cl and Cl .
5. As discussed in Sec. II, bromine molecule is produced
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:
131.111.164.128 On: Sun, 21 Dec 2014 02:48:14

7488
J. Chem. Phys., Vol. 117, No. 16, 22 October 2002
Xu et al.
FIG. 4. ͑a͒ Translational energy distribution of ground Br. Solid line: de-
picted from Fig. 3͑a͒ which consists of contributions from channels ͑1͒ and
͑2͒. Dotted line: Br contribution from channel ͑2͒, which is converted from
Fig. 2 using momentum conservation. A factor of 0.4 has been applied to the
self-normalized P(ET). This factor is obtained using the ␤(ET) values
shown in ͑b͒. See text Sec. III C for details. Dashed line: contribution to Br
from channel ͑1͒, which is the difference of the solid line and dotted line. ͑b͒
Anisotropy parameter distributions ␤(ET). Solid heavy line: composite ␤
distribution consisting of contributions from channels ͑1͒ and ͑2͒, depicted
from Fig. 3͑a͒. Dotted line: ␤ distribution of reaction ͑2͒, depicted from Fig.
*
2. Solid thin line: ␤ distribution for excited Br , depicted from Fig. 3͑b͒.
from the photodissociation of Br2CO, and any decomposi-
tion in the gas line or in the pulsed valve forming Br2 can be
ruled out. The image of Br2 shows that the recoil direction of
the products is perpendicular to the laser polarization direc-
tion, in sharp contrast to the other two channels described in
Secs. III B and III C of this paper. The translational energy
distribution shows a Gaussian-type function and the ␤ pa-
rameter goes to Ϫ0.9 near the peak region. These distribu-
tions suggest that the molecule dissociates on an excited re-
pulsive surface, rather than from the ground-state surface.
The available energy for this channel amounts to 103.5
kcal/mol, and the maximum ET only reaches 50 kcal/mol.
This means that Br2 and CO would have internal energy in
the range of 54–103.5 kcal/mol. The dissociation energy of
ground-state Br2 is 46 kcal/mol, which is below the mini-
mum internal energy of the products. The lowest excited
state of CO (a 3⌸) lies at 139 kcal/mol, which is higher than
the maximum available energy for this channel. On the other
hand, there are two excited states of Br2 (A 3⌸1u and
B 3⌸0uϩ) at 40 and 46 kcal/mol. Although there should be
some excitation in the CO(X) vibration since the CvO
double bond in Br2CO changes to a triple bond in CO, we
expect some Br2 molecules to be formed in the excited triplet
states on the basis of the available energy and absorption
spectrum. The absorption spectrum of Br2CO ͑Ref. 40͒
shows that the absorption at 267 nm is located in the tail of
the absorption band of Br2CO where state mixing of singlet
FIG. 5. Two-dimensional image, total translational energy, and angular dis-
tributions of the Br2 channel. The arrow indicates the laser polarization
direction. The oscillation of ␤(ET) at low ET is probably noise generated
from the back-projection procedure.
and triplet states may occur. While intersystem crossing is
likely to happen, especially with the help of the strong spin–
orbit interaction of Br, it is also probable that mixed states
are initially excited and some of the dissociation occurs im-
mediately along the triplet surfaces.
Molecular elimination of H2 from formaldehyde H2CO
has been studied in great detail near the threshold.41,42 It is
found that the anisotropic angular distributions are retained
as long as 10–100 ns following the excitation of the mol-
ecule. During this period of time, the molecule undergoes
many rotations before dissociation. Another finding is that
fragments with both positive and negative ␤ parameters can
be produced from the same rotational transition of different
bands. It is not yet clear what gives rise to the different
angular distributions. In the present case of Br2CO, negative
␤(ET) distributions are found in the wavelength region be-
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:
131.111.164.128 On: Sun, 21 Dec 2014 02:48:14

J. Chem. Phys., Vol. 117, No. 16, 22 October 2002
Photodissociation of carbonic dibromide
7489
try during dissociation and Br2 exits at a direction perpen-
dicular to the transition moment direction, which will give
rise to negative ␤ values close to the low limit of Ϫ1.
Similar results are obtained in a recent photodissociation
study of 1,2-dibromoethylene BrCHvBrCH ͑Ref. 44͒ using
photofragment translational spectroscopy ͑PTS͒. For Br at-
oms the bimodal energy distribution composes of a fast part
with large positive ␤ϭ1.5 and a slow part with small positive
␤ϭ0.4. The authors proposed an asynchronously concerted
mechanism for this three-body decay. The Br2 channel is also
found with a negative ␤ϭϪ0.5. Although the molecules in-
volved in this and the present studies are totally different, the
results are strikingly similar. The only exception is that no
BrC2H2 is found, in contrast to the detection of BrCO in the
present experiment. This is understandable if one considers
the different experimental methods employed. In PTS the
neutral products have to travel some distance before they
reach the detector. The typical time of flight is from 50 to
100 ␮s for the fast fragments. This requires that the neutral
products survive this time of flight in order to be detected.
For molecules like BrCO or an excimer whose excited state
is bound and whose ground state is unstable or weakly bound
there is no chance to detect these molecules by conventional
PTS method. In ion imaging experiments where the products
are ionized right after they are formed, such short-lived spe-
cies can be stabilized via ionization and then detected.
FIG. 6. Schematic of the structure of Br2CO and the direction of its transi-
tion dipole moment ͑␮͒. Geometry is optimized at the B3LYP/6-311G(d,p)
level. C2v symmetry should be preserved during the formation of Br2 ,
which exits at a direction perpendicular to ␮.
*
tween the two wavelengths for Br and Br 2ϩ1 REMPI. It is
well known that H2 elimination occurs over a steep barrier
from the vibrationally excited ground electronic state of
formaldehyde.43 The present results suggest Br2 elimination
takes place very quickly on excited repulsive surfaces.
To facilitate the understanding of current results for all
the observed channels, a simple ab initio calculation using
density functional theory at the B3LYP/6-311G(d,p) level
was performed. Figure 6 shows the structure of Br2CO
ground state. Also shown in the figure are the recoil direc-
tions of products ͑Br and Br2) and the proposed orientation
of the transition dipole moment. In this excitation the transi-
tion dipole moment should be perpendicular to the CvO
bond. The Br–C–Br bond angle in the ground state Br2CO is
112.8°. The angular distributions indicate that dissociation to
each channel is very fast. If the excited-state structure is not
distorted from that of the ground state or the very short life-
time of the excited state prevents the molecule from rear-
ranging, the angular distributions for all the channels can be
fully understood.
IV. CONCLUSIONS
In summary, the photodissociation of Br2CO at 267 nm,
studied by ion velocity imaging and TOF-MS methods, indi-
cates that the dissociation follows the mechanisms shown
below:
Br2COϩh␯267 nm→Br͑fast)ϩBr͑slow)ϩCO
͑8͒
͑9͒
→BrCO A͒ϩBr
͑
→Br2ϩCO.
͑10͒
The anisotropy parameter ␤ can be expressed as a func-
tion of the angle between the transition dipole moment and
the recoil direction of the product,
Bimodal kinetic energy distributions of bromine atoms
are found. The angular distributions ␤(ET) for the fast and
slow velocity distributions are coupled, which suggests that
the three-body decay is an asynchronously concerted pro-
cess. The data indicate that dissociation occurs on a repulsive
state as indicated by the close-to-limit ␤ values. Both
ground-state and excited-state Br atoms are produced with
fast and slow speeds through this channel. A second channel
produces BrCO(A) radical in the exited state with almost
constant ␤ parameters that merge with those from Eq. ͑8͒ at
the high-energy end. The Br2 elimination from Br2CO is
observed for the first time for the class of X2CO systems in
the UV region. The Gaussian-shaped energy distribution and
the negative ␤(ET) distribution suggest that this channel oc-
curs on a repulsive surface and that triplet surfaces may be
involved. The results clearly indicate that the transition di-
pole moment of Br2CO should be perpendicular to the CvO
bond at 267 nm.
␤ϭ2 P cos ␪͒ ϭ3 cos2 ␪Ϫ1.
͑7͒
͑
͗
͘
2
During the dissociation, the bromine atoms exit at an
angle around 34° from the transition dipole moment direc-
tion. Putting this angle into Eq. ͑7͒, one gets ␤ϭ1, which is
*
in good agreement with the ␤ values for the fast Br and Br
components, and for the radical channel ͑2͒. During this first
dissociation step, the force between the departing Br and
BrCO will cause BrCO to rotate counterclockwise since the
center-of-mass of BrCO lies to the Br side. This rotation will
increase the angle between the second Br exit direction and
the transition dipole moment direction, resulting in decreased
␤ values. But the nonvanishing ␤ values for the slow part of
the energy distribution indicate that the second bond fission
is also very fast. An asynchronously concerted decay seems
the most suitable model to describe this three-body decay. As
for the Br2 channel, the molecule may maintain C2v symme-
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:
131.111.164.128 On: Sun, 21 Dec 2014 02:48:14

7490
J. Chem. Phys., Vol. 117, No. 16, 22 October 2002
Xu et al.
21 S. W. North, C. A. Longfellow, and Y. T. Lee, J. Chem. Phys. 99, 4423
͑1993͒.
ACKNOWLEDGMENTS
22 T. Einfeld, A. Chichinin, C. Maul, and K.-H. Gericke, J. Chem. Phys. 116,
2803 ͑2002͒.
This work was supported by NSF ͑Grant No. CHE-
9713929͒ and NASA ͑Grant No. NAG54711͒.
23 W. M. Jackson and D. D. Xu, J. Chem. Phys. 113, 3651 ͑2000͒.
24 M. J. Parkington, T. A. Ryan, and K. R. Seddon, J. Chem. Soc., Dalton
Trans. 2, 257 ͑1997͒.
1 World Meteorological Organization Global Ozone Research and Monitor-
ing Project. Report No. 20; Scientific Assessment of Stratospheric Ozone.
Appendix: AFEAS Report, Vol. 2, Chap. 6, 1989.
2 S. C. Wofsy, M. B. McElroy, and Y. L. Yang, Geophys. Res. Lett. 2, 215
͑1975͒.
25 J. Overend and J. C. Evans, Trans. Faraday Soc. 55, 1817 ͑1959͒.
26 S. Arepalli, N. Presser, D. Robie, and R. J. Gordon, Chem. Phys. Lett. 117,
64 ͑1985͒.
27 Y.-J. Jee, M. S. Park, Y. S. Kim, Y.-J. Jung, and K.-H. Jung, Chem. Phys.
Lett. 287, 701 ͑1998͒.
3 Y. L. Yung, J. P. Pinto, R. T. Watson, and S. P. Sander, J. Atmos. Sci. 37,
339 ͑1980͒.
28 W. M. Jackson, D. D. Xu, R. J. Price, K. L. McNesby, and I. A. McLaren,
in Imaging in Chemical Dynamic, edited by A. G. Suits and R. E. Con-
tinetti ͑American Chemical Society, Washington, D.C., 2000͒, Chap. 7.
29 A. T. J. B. Eppink and D. H. Parker, Rev. Sci. Instrum. 68, 3477 ͑1997͒.
30 D. A. Dixon, K. A. Peterson, and J. S. Francisco, J. Phys. Chem. A 104,
6227 ͑2000͒.
4 W. Reifenhaeuser and K. G. Heumann, Chemosphere 24, 1293 ͑1992͒.
5 R. A. Rassmussen and M. A. K. Kholil, Geophys. Res. Lett. 11, 433
͑1984͒.
6 W. T. Sturges, D. E. Oram, L. J. Carpenter, S. A. Penkett, and A. Engel,
Geophys. Res. Lett. 27, 2081 ͑2000͒.
7 V. L. Dvortsov, M. A. Geller, S. Soloman, S. M. Schauffler, E. L. Atlas,
and D. R. Blake, Geophys. Res. Lett. 26, 1699 ͑1999͒.
8 W. B. DeMore, J. Phys. Chem. A 100, 5813 ͑1996͒.
9 K. G. Kambanis, Y. G. Lazarou, and P. J. Papagiannakopoulos, J. Phys.
Chem. A 101, 8496 ͑1997͒.
31 NIST chemistry webbook, http://webbook.nist.gov/chemistry
32 J. S. Francisco and M. M. Maricq, Adv. Photochem. 20, 79 ͑1995͒.
33 Unreported experiments in J. S. Francisco’s laboratory.
34 P. Venkateswarlu, Can. J. Phys. 47, 2525 ͑1969͒.
35 D. I. Austin, R. J. Donovan, A. Hopkirk, K. P. Lawley, D. Shaw, and A. J.
Yencha, Chem. Phys. 118, 91 ͑1987͒.
10 Y. Li and J. S. Francisco, J. Chem. Phys. 113, 1807 ͑2000͒.
11 J. Heicklen, J. Am. Chem. Soc. 87, 444 ͑1965͒.
12 H. Okabe, J. Chem. Phys. 66, 2058 ͑1977͒.
36 J. S. Francisco and A. N. Goldstein, Chem. Phys. 128, 367 ͑1988͒.
37 D. Gruber, M. Musso, L. Windholz, M. Gleichmann, B. A. Hess, F. Fuso,
and M. Allegrini, J. Chem. Phys. 101, 929 ͑1994͒.
13 A. I. Chichinin, Chem. Phys. Lett. 209, 459 ͑1993͒.
14 C. Maul, T. Haas, K.-H. Gericke, and F. J. Comes, J. Chem. Phys. 102,
3238 ͑1995͒.
38 C. Angeli, M. Persico, M. Allegrini, G. De Filippo, F. Fuso, D. Gruber, L.
Windholz, and M. Musso, J. Chem. Phys. 102, 7782 ͑1995͒.
39 J. Kohel and J. W. Keto, J. Chem. Phys. 113, 10551 ͑2000͒.
40 J. C. Hansen, S. D. Belair, A. Kemeny, and J. S. Francisco ͑unpublished͒.
41 T. J. Butenhoff, K. L. Carleton, and C. B. Moore, J. Chem. Phys. 92, 377
͑1990͒.
15 C. Maul, C. Dietrich, T. Haas, and K.-H. Gericke, Phys. Chem. Chem.
Phys. 1, 1441 ͑1999͒.
16 C. Maul and K.-H. Gericke, Int. Rev. Phys. Chem. 16, 1 ͑1997͒.
17 R. M. Hochstrasse and D. S. King, J. Am. Chem. Soc. 97, 4760 ͑1975͒.
18 L. Nahon, P. Morin, M. Larzilliere, and I. Nenner, J. Chem. Phys. 96, 3628
͑1992͒.
42 K. L. Carleton, T. J. Butenhoff, and C. B. Moore, J. Chem. Phys. 93, 3907
͑1990͒.
19 S. W. North, D. A. Blank, J. D. Gezelter, C. A. Longfellow, and Y. T. Lee,
J. Chem. Phys. 102, 4447 ͑1995͒.
43 C. B. Moore and J. C. Weisshaar, Annu. Rev. Phys. Chem. 34, 525 ͑1983͒.
44 Y. R. Lee, C. C. Chou, Y. J. Lee, L. D. Wang, and S. M. Lin, J. Chem.
Phys. 115, 3195 ͑2001͒.
20 D. A. Blank, S. W. North, D. Stranges, A. G. Suits, and Y. T. Lee, J. Chem.
Phys. 106, 539 ͑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:
131.111.164.128 On: Sun, 21 Dec 2014 02:48:14