193 nm photolysis of vinyl bromide: Nascent product distribution of the C2H3Br→C2H2 (vinylidene)+HBr channel

Article abstract of DOI:10.1063/1.1382812

The internal energy content of fragment HBr and C₂H₂ following photolysis of the precursor, vinyl bromide, at 193 nm was determined using time-resolved Fourier transform spectroscopy (TR FTS). Data taken 1 μs after the laser photolys

Full text of DOI:10.1063/1.1382812

1
93 nm photolysis of vinyl bromide: Nascent product distribution of the C 2 H 3 Br → C
H 2 (vinylidene)+HBr channel
2
Dean-Kuo Liu, Laura T. Letendre, and Hai-Lung Dai
Citation: The Journal of Chemical Physics 115, 1734 (2001); doi: 10.1063/1.1382812
View online: http://dx.doi.org/10.1063/1.1382812
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/115/4?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Photodissociation of vinyl cyanide at 193 nm: Nascent product distributions of the molecular elimination channels
J. Chem. Phys. 130, 044307 (2009); 10.1063/1.3065986
Molecular elimination in photolysis of fluorobenzene at 193 nm: Internal energy of HF determined with time-
resolved Fourier-transform spectroscopy
J. Chem. Phys. 121, 8792 (2004); 10.1063/1.1802537
Branching ratios of C 2 products in the photodissociation of C 2 H at 193 nm
J. Chem. Phys. 114, 9821 (2001); 10.1063/1.1370942
Three-center versus four-center elimination in photolysis of vinyl fluoride and vinyl bromide at 193 nm: Bimodal
rotational distribution of HF and HBr (v5) detected with time-resolved Fourier transform spectroscopy
J. Chem. Phys. 114, 7396 (2001); 10.1063/1.1343079
I. Three-center versus four-center HCl-elimination in photolysis of vinyl chloride at 193 nm: Bimodal rotational
distribution of HCl (v7) detected with time-resolved Fourier-transform spectroscopy
J. Chem. Phys. 114, 160 (2001); 10.1063/1.1328736
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.70.241.163 On: Mon, 22 Dec 2014 17:14:29

JOURNAL OF CHEMICAL PHYSICS
VOLUME 115, NUMBER 4
22 JULY 2001
193 nm photolysis of vinyl bromide: Nascent product distribution
of the C H Br\C H „vinylidene…¿HBr channel
2
3
2
2
Dean-Kuo Liu, Laura T. Letendre, and Hai-Lung Dai
Department of Chemistry, University of Pennsylvania, Philadelphia, Pennsylvania 19104-6323
͑
Received 17 August 2000; accepted 9 May 2001͒
The internal energy content of the photofragments HBr and C H from the minor channel of the
2
2
photolysis of C H Br with 193 nm radiation has been measured using time-resolved infrared Fourier
2
3
transform IR emission spectroscopy with 0.5 ␮s resolution. Vibrational level population and the
rotational population of the HBr fragment are determined from 1.0 ␮s following the photolysis until
complete HBr relaxation. The nascent distribution of HBr is extrapolated, from a collision
quenching model with a Boltzmann distribution, to be at 8690 and 7000 K for the vibration and
rotation respectively. The product vibrational energy distribution supports a reaction mechanism
based on the 3-centered HBr elimination process yielding vinylidene and HBr. The nascent internal
energy of vinylidene is deduced to be 24 kcal/mol. Vinylidene isomerizes to acetylene and the
acetylene emission bands, ␯ , ␯ ϩ␯ and ␯ , are detected. © 2001 American Institute of Physics.
3
4
5
5
͓
DOI: 10.1063/1.1382812͔
I. INTRODUCTION
dissociation dynamics. Time-resolved Fourier transform
spectroscopy ͑TR FTS͒, on the other hand, has the ability to
measure the vibration–rotation energy content of the nascent
The photodissociation dynamics of small molecules have
provided a fruitful testing ground for unimolecular reaction
theories. Photodissociation reactions can now be character-
ized with quantum state resolution and examined with ab
initio theoretical calculations leading to a detailed under-
standing of the mechanism of unimolecular reactions. Inter-
ests for recent studies include determining the potential en-
ergy surface and transition state in these reactions,
understanding the energy flow in the molecule and how the
energy affects reactivity.
Photodissociation reactions of halogenated ethylene
molecules are of specific interest. These molecules are in-
triguing because their dissociation may proceed via several
energetically plausible dissociation pathways. The dissocia-
tion mechanism strongly depends on the extent and identity
of the halogen substitution which influence both the potential
energy surface and transition state of the fragmentation pro-
cess. The studies of these molecules are particularly interest-
ing since their reaction dynamics can be experimentally
probed from the energy distribution of the dissociation prod-
ucts.
5–7
photofragments through the IR emission.
The time-resolved IR emission spectra can be acquired
by a Fourier transform spectrometer operated in step–scan
8
mode. In this technique the movable mirror is temporarily
held at a fixed position while the transient interferometric
signal is collected, averaged, and stored. After data collec-
tion, the mirror of the FTIR interferometer moves to the next
position and the process is repeated. When the mirror scan is
completed, the signal intensity over the entire mirror position
and time range is available, and the interferogram at a spe-
cific time can be extracted from the 3D data array and Fou-
rier transformed. The result is a frequency spectrum for each
time slice. Currently, the time resolution of this technique is
limited by the detector response time which in the IR is
9,10
around 10 ns.
Previous studies have indicated that the photodissocia-
tion channels of vinyl bromide ͑VBr͒ at 193 nm are domi-
11,12
nated by the following reactions:
2
C2H3Br→C2H3ϩBr
*
͑2P1/2͒ or Br͑ P3/2͒,
͑1͒
͑2͒
͑3͒
Photodissociation dynamics of vinyl halides (C H X),
2
3
→
→
HCCH͑acetylene͒ϩHBr,
which are the simplest halogen substituted ethylene mol-
ecules, have been studied by several laboratories using mo-
lecular beam and time-of-flight mass spectrometry tech-
niques to measure the energy distribution of photofragments
HX and C H ͑either as acetylene or vinylidene͒ or X and
H CC:͑vinylidene͒ϩHBr.
2
Reaction ͑1͒ is suggested to result from the C–Br bond rup-
ture in an electronically excited state of VBr.1
2,13
This reac-
2
2
1
,2
C H . The heat of formation, branching ratio, and reaction
tion was used for the production of vibrationally excited vi-
nyl radicals in our study of the vinyl vibrational modes.14
Reactions ͑2͒ and ͑3͒ are assumed to occur on the ground
state surface subsequent to internal conversion following
2
3
pathways can be determined from measured speed distribu-
tions of each species. Interestingly, results from time-of-
flight molecular beam experiments are in conflict with earlier
observations made in matrix isolated environments3 in
terms of the reaction branching ratio. Despite their superior
capacity in product identification, molecular beam tech-
niques are less sensitive to the internal energy distribution of
the photofragments which is crucial for understanding the
,4
13
electronic excitation. There are two possible mechanisms
for the elimination of HBr from the VBr precursor. The first,
reaction ͑2͒, is the 4-centered ͑␣␤͒ elimination channel that
involves bromine from the ␣ carbon combining with the hy-
drogen from the ␤ carbon. The second, reaction ͑3͒, is the
0021-9606/2001/115(4)/1734/8/$18.00
1734
© 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:
30.70.241.163 On: Mon, 22 Dec 2014 17:14:29
1

J. Chem. Phys., Vol. 115, No. 4, 22 July 2001
Photolysis of vinyl bromide
1735
FIG. 1. The relevant energy levels of the fragments resulting from the
photolysis of vinyl bromide with 193 nm radiation. All energies are in
kcal/mol.
FIG. 2. The time-resolved IR emission spectra following photodissociation
of vinyl bromide at 193 nm. The spectra are taken with 50 laser shots
averaged per mirror position, 4 cm spectral and 500 ns time resolutions.
Ϫ1
The emission positions of four major products C H , C H , Br*, and HBr
are described in the text. The arrows indicate the nine vibrational modes of
the vinyl radical. The dashed lines mark the acetylene emission peaks.
2
3
2
2
3
-centered ͑␣␣͒ elimination where the hydrogen and the bro-
mine from the same ␣ carbon are separated from the VBr
parent molecule. Because of the low energy barrier ͑ϳ2.6
1
5
kcal/mol͒ along the isomerization coordinate, the vi-
nylidene product of reaction ͑3͒ is expected to quickly
isomerize to the more energetically stable acetylene form. As
both channels eventually result in the same set of products, it
is difficult to ascertain which is the dissociation channel from
product identification alone. The two mechanisms, however,
should give distinctly different product energy distributions
as the two transition states are distinctly different.
inserted from the top of the cell and into the center region of
a set of spherical mirrors. The mirrors were setup according
to the Welsh cell design for collecting emission. To prevent
particle deposition on the cell windows Ar was flowed over
them while a constant cell pressure of 4 Torr was maintained.
Emission signal from the cell was collected perpendicu-
lar to the photolysis beam by a pair of Potassium Chloride
lenses (f4) and focused into a FTIR spectrometer ͑Bruker
IFS-88͒ which ran in step–scan mode. The signal passed
through the Michelson Interferometer and was detected by a
mercury cadmium telluride ͑MCT͒ ͑EG&G͒ detector with a
response time of 500 ns and spectral response of 800 to 8000
The relevant thermal chemistry of these dissociation
pathways along with the energy of the transition states for
3-center and 4-center HBr elimination from photodissocia-
tion of VBr at 193 nm irradiation are illustrated in Fig. 1.
Most recently, Raff et al. used classical trajectory calcula-
Ϫ1
tions to model the dissociation pathways from the excited
cm . The slower detector, instead of the faster 10 ns detec-
and ground state potential surfaces.1
6,17
Their results pre-
tor, was used because of its higher sensitivity so that higher
spectral resolution could be achieved. The time resolution is
sufficient for resolving the collision quenching process dis-
cussed later. The signal was amplified and then digitized with
the Bruker 8 bit PAD82a transient digitizing board to obtain
the interferogram. The interferograms were Fourier trans-
formed using a Mertz correction, zero filled eight times and
apodized with a Blackman–Harris 3 term correction. The
spectra were taken with 50 laser shots averaged per mirror
dicted that HBr is formed almost solely via a 3-center elimi-
nation mechanism as this channel has a lower transition state
barrier than that of the 4-centered elimination.
Here we report the vibrational and rotational energy dis-
tribution of the HBr fragment after photodissociation of the
precursor vinyl bromide acquired with the TR FTS tech-
nique. The observed nascent product energy distribution al-
lows us to deduce details of reaction dynamics, such as the
branching ratio of the reaction channels and the transition
state structure, and to compare with several theoretical dy-
namical model calculations.
Ϫ1
position, 4 cm spectral and 500 ns intervals. For higher
resolution analysis of emission from HBr, 100 laser shots
Ϫ1
were averaged at each mirror position, with 1 cm spectral
and 500 ns temporal resolution.
Vinyl Bromide ͑Aldrich͒, purity Ͼ99.3%, was used
without further purification. Argon buffer gas ͑Spectra
Gases͒ was of 99.99% purity. Several experiments using VBr
purified through several freeze, pump, and thaw cycles were
performed, and no difference in the resultant spectra was
observed.
II. EXPERIMENT
The experimental procedure used here for probing IR
emission from excited photofragments following photodisso-
ciation of VBr was described in detail previously.
an excimer laser ͑Lambda–Physik model EMG 201 MSC͒
run at 20 Hz, was used to produce a 193 nm photolysis pulse
1
4,18
Briefly,
2
with 30 mJ pulse energy in a rectangular beam of c.a. 1 cm .
III. RESULTS AND ANALYSIS
The laser beam excited the precursor which flowed through a
gas cell. The body of the cell was made of aluminum and has
four view ports. A gas mixture of Ar premixed with VBr in a
ratio of 80:1, filled the cell through a shower head nozzle
Figure 2 shows selected time slices of the emission spec-
trum following VBr photodissociation. Emission from four
photofragments are observed, namely, C H , Br, C H
2
3
2
2
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
30.70.241.163 On: Mon, 22 Dec 2014 17:14:29
1

1736
J. Chem. Phys., Vol. 115, No. 4, 22 July 2001
Liu, Letendre, and Dai
because of the sharp cutoff of the MCT detector sensitivity in
this region. Rotationally resolved HBr emission peaks are
Ϫ1
observed within the range of 1900 to 2800 cm . In this
paper we will focus on the HBr elimination channels through
the state resolved analysis of the HBr emission.
A. HBr emission
Selected time slices at 0.5, 1, 2, 3, 4, 5, and 6 ␮s from
the HBr emission spectrum are shown in Fig. 3. The rovibra-
tional line positions of the HBr emission from the first six
vibrational levels were calculated using molecular constants
from Rank et al.19 and are unambiguously assigned in the
spectra. The complexity of the spectra reflects the emission
from both of the equally naturally abundant isotopes of bro-
mine, 79Br and Br. The difference in the rotational con-
81
Ϫ3
Ϫ1
stants of these isotopes is only 2.6ϫ10 cm and therefore
the isotope splitting of the transitions cannot be resolved in
Ϫ1
the current 1 cm resolution experiments. For this reason
the average rotational constant of the two isotopes was used
for the calculation.
During the 500 ns rise time of the detector approxi-
mately 25 Ar/HBr collisions occur under the experimental
conditions ͑4 Torr Ar, 400 mTorr VBr͒. Also, in the FT emis-
sion spectra the noise is proportional to the overall signal
incident on the detector, consequently in early time slices
FIG. 3. The time-resolved IR emission spectra of the HBr fragment from the
photodissociation of vinyl bromide. The spectra are taken with 100 laser
shots averaged per mirror position, 1 cm spectral and 500 ns time resolu-
Ϫ1
tions.
͑Ͻ1 ␮s͒ the high level of background noise prohibits the
assignment of the HBr peaks. For this reason the nascent
internal energy distributions for HBr cannot be measured de-
finitively. The earliest time slice to which the rovibrational
distributions can be confidently analyzed is 1.0 ␮s. The na-
scent distribution, defined at tϭ0, can, however, be extrapo-
lated from the later time energy distributions.
͑
acetylene͒, and HBr. The part of the emission spectrum rel-
1
4
evant to the vinyl radical has been reported previously, the
peaks marked with arrows are attributed to the nine vibra-
tional modes of the vinyl radical. The small peak at 3685
cm is due to the 2P3/2� P transition of Br*. Peaks
Ϫ1
2
1/2
Ϫ1
marked with the dashed line at 3294, 1342, and 750 cm are
assigned to the ␯ , ␯ ϩ␯ , and ␯ modes, respectively, of
Figure 4 shows the IR emission spectrum for HBr at 1
␮s. The entire spectral range of the HBr emission is shown in
3
4
5
5
vibrationally excited acetylene. The ␯ peak is diminished
5
FIG. 4. The time-resolved IR emission
spectra of the HBr photodissociation
product collected 1.0 ␮s after the 193
nm laser excitation of VBr. The emis-
sion spectrum is shown along with the
corresponding simulated spectrum ͑in-
verted͒. A line spectrum with rovibra-
tional assignments is depicted above
the spectrum. ͑a͒ The entire HBr emis-
Ϫ1
sion range of 1900 cm through 2800
Ϫ1
cm is shown. ͑b͒ Emission from the
Ϫ1
range of 2100 through 2300 cm is
enlarged for clarity. The dashed line
indicates the rotational transitions
from the highest assigned vibrational
level (vϭ6) after the photodissocia-
tion, which were found to match the
observed lines marked by *. For com-
parison, the inverted simulation spec-
trum did not include the vϭ6 transi-
tions but only transitions from vϭ5
and below.
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:
30.70.241.163 On: Mon, 22 Dec 2014 17:14:29
1

J. Chem. Phys., Vol. 115, No. 4, 22 July 2001
Photolysis of vinyl bromide
1737
Figure 4͑a͒. For clarity, emission peaks from 2100 to 2300
Ϫ1
cm are enlarged in the inset, where the positions of rota-
tional transitions from the highest assigned vibrational quan-
tum number (vϭ6) are marked with a line and found to
match the observed lines marked with an asterisk. The cal-
culated spectrum, for comparison, is generated by applying a
Gaussian line shape to each of the calculated rovibrational
transitions. To correctly fit the experimental spectrum it was
Ϫ1
found that a full width at half maximum of 1.2 cm was
needed for each line. This is in reasonable agreement with
Ϫ1
the nominally 1 cm resolution set for the FT spectrometer,
and taking into consideration the isotopic splitting.
The intensity of the calculated lines was varied to fit the
experimentally observed spectra. Peaks from the vϭ7→6
transitions can also be assigned although they are not in-
cluded in the simulation since their intensity is quite weak.
FIG. 5. The vibrational temperature of fragment HBr obtained from a plot
of ln͓population(v)͔ vs the HBr vibrational level energy at time slices 1.0
and 1.6 ␮s, respectively, after photolysis of VBr at 193 nm. The value at
zero time slice is determined from the collisional quenching model de-
scribed in the text. The uncertainty of the population of each point is smaller
than Ϯ10%.
͑Peaks from vϭ8→7, or higher vibrational levels, emission
cannot be clearly identified.͒
B. Nascent energy distribution
ergy content from extrapolation of the later time results,
since direct measurement is not plausible at the desirable
spectral resolution. A seven-level system, with each level
representing a vibrational state, is used to model the inte-
grated HBr IR emission decay. The set of rate equations are
as follows:
The relative rotational and vibrational state populations
of HBr can be extracted from the integrated peak intensities,
I(
j,v)
. We have found that both the rotational and vibrational
population distributions can be approximately described by
the Boltzmann distribution. To begin with, following the
rovibrational assignment of each spectral line, the rotational
temperature of each vibrational level can be determined by a
linear plot of ln(I(j,v))/(2jϩ1) versus rotational state energy
for each vibrational state at each specific time following the
dissociation.
dN /dtϭkvϩ1,v͓M͔Nvϩ1Ϫkv,vϪ1͓M͔Nv , vϭ1Ϫ6.
v
͑1͒
In these equations M represents the combined concentration
of all quenchers, N is the population of vibrational state v.
v
The relative population of the emitting vibrational state
obs
obs
kj,k is the quenching rate coefficient for collisions between
HBr (vϭj) and all of the quenchers combined resulting in
HBr (vϭk). Calculations were performed allowing the
kv,vϪ1 rate constants to vary with the vibrational quantum
Nv is related to the intensity, Iv , by N ϰIv /A(v→v
v
Ϫ1), where A is the Einstein coefficient. Here the set of A
coefficients previously reported for vibrational quantum
numbers up to five is used.20 The coefficient for the 6→5
n
number, with kv,vϪ1ϭkϩkЈv . Previous studies have found
transition was extrapolated by using a second power polyno-
mial from the first five coefficients. The overall population,
in some cases the relaxation rate increases with the vibra-
tional quantum number.21 Furthermore, it is safe to assume
that the rates of ⌬vу1 transitions are negligible in compari-
son with the ⌬vϭϪ1 transition rates. kv,vϪ1͓M͔ is far
larger than the radiative decay rate, therefore the latter is not
included in the treatment of the decay of the vibrationally
excited population. The set of equations ͑1͒ can be solved
numerically to fit the vibrational populations defined by the
respective temperatures at 1.0 and 1.6 ␮s. A nonlinear least
squares fit shows that the quenching constants have a very
weak dependence on the vibrational quantum number ͑i.e.,
nϭ0.02͒. This is probably because the differences in energy
gaps between the HBr vibrational levels are small and the
primary collider is monoatomic, which allows only the V-T
channel.
in relative magnitude, for a specific vibrational state N is
v
then attained by summing the intensities of the emission
from each rotational level of this vibrational state and nor-
malized by the A coefficient. The relative population of the
excited vibrational levels can then be determined by dividing
the population of each vibrational level by the total excited
population. Plotting the ln(N ) at a specific time as a function
v
of vibrational energy gives the vibrational temperature. As
shown in Fig. 5, the plots for HBr fragments at 1.0 and 1.6
␮s after photolysis results in vibrational temperatures of
4098Ϯ370 and 3139Ϯ210 K, respectively.
During the time period following the dissociation HBr
has many collisions with the Ar buffer gas, the other photo-
fragments, and residual precursor molecules, facilitating re-
laxation of the internal energy of the molecule eventually to
a 300 K Boltzmann distribution. This relaxation process is
evidenced by the rotational temperature drop in the vϭ1
vibrational state of c.a. 600 K during the time period from
Equation ͑1͒ can also be solved analytically when we
treat all of the quenching rates as only one parameter of ␣
ϭkv,vϪ1͓M͔. Using the Laplace transform method, the con-
secutive equations can be solved as:
1
.0 to 1.6 ␮s. Similar trends are also observed in other vi-
brational states.
A model is employed to determine the nascent HBr en-
6
␣t͒iϪv
͑
0
eϪ␣t vϭ1Ϫ6.
N ϭ
Ni
͑2͒
v
ͭ
͚
ͮ
iϭv
͑iϪv͒!
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.70.241.163 On: Mon, 22 Dec 2014 17:14:29

1738
J. Chem. Phys., Vol. 115, No. 4, 22 July 2001
Liu, Letendre, and Dai
FIG. 6. The population of individual vibrational levels of HBr determined from the collisional quenching model analysis of the time-resolved emission spectra
after 1.0 ␮s. The time evolution of the vibrational temperature is shown in the figure inset.
0
each vibrational level, weighted by the population of that
vibrational level reveals a nascent rotational temperature of
N is the concentration of level i immediately following pho-
i
todissociation of VBr. The quenching rate, ␣, can be deter-
mined from the exponential decay of N at vϭ6. Equation
2͒ reduces to a simple exponential decay. Using the ratio of
measured vibrational intensities of vϭ6 between times 1.0
and 1.6 ␮s, ␣ is determined to be 3.65ϫ10 s . HBr has over
0 times as many collisions with argon as with other photo-
7
000 K or 14 kcal/mol for HBr.
The average energy content of the C H product can also
6
͑
2
2
be determined from the emission spectrum. Three peaks
Ϫ1
6
1
identified at 3289, 1342, and 750 cm in the spectrum are
attributed to the asymmetric CH stretching mode, combina-
tion band, and CH bending mode of acetylene, respectively.
Emission from the 3-centered elimination product, vi-
nylidene, even if it exists as a nascent product, is not ob-
served at 1 ␮s most likely due to the low barrier to isomer-
ization leading to acetylene and corresponding short ͑ps͒
1
fragments or precursor molecules. It is assumed in our cal-
culations that collisions with other molecules are at least as
effective as argon collisions for the vibrational relaxation of
HBr. From this rate equation model the cross-section for
HBrϩM collisional energy transfer is estimated to be 0.44
2
15
Å . As expected, this number is significantly smaller than the
lifetime of vinylidene. By analyzing the anharmonic shift
2
hard sphere cross section of 7.9 Å of HBr and Ar.
of the acetylene peaks the average vibrational energy acety-
lene can be determined. The asymmetric CH stretching peak
The relative population of each vibrational state as a
function of time obtained from the model calculation is
shown in Fig. 6. The overall time evolution of the vibrational
temperature is shown as an inset in the same figure.
Ϫ1
frequency of acetylene is redshifted at 1 ␮s by 34 cm from
the fundamental transition. The anharmonicity constant for
Ϫ1
this transition is 27.41 cm . This broad feature is then most
Using the nascent vibrational energy distribution, the av-
erage vibrational energy content of HBr as a nascent reaction
likely a mixture of the fundamental band and the vϭ2 to
vϭ1 band ͑redshifted from the fundamental by 55 cm ͒. It
Ϫ1
product is determined to be
the average rotational energy can be calculated for each vi-
brational level. The sum of the average rotational energy for
͗
Evib͘ϭ17.3 kcal/mol. Similarly,
is then estimated that the average vibrational quantum num-
ber of the excited acetylene is about 1.5 quanta in the asym-
metric CH stretching mode. Analysis of the anharmonic shift
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.70.241.163 On: Mon, 22 Dec 2014 17:14:29

J. Chem. Phys., Vol. 115, No. 4, 22 July 2001
Photolysis of vinyl bromide
1739
in the emission from the bending mode yields consistent en-
ergy content. Assuming that the energy in different modes is
in thermal equilibrium, maintained by collisions, we estimate
that acetylene has at least 14 kcal/mole of vibrational energy
at 1.0 ␮s after vinyl bromide dissociation.
IV. DISCUSSION
A. Reaction mechanism
FIG. 7. The geometry of the equilibrium structure of the VBr molecule, and
the transition states of the 3-centered and 4-centered HBr elimination chan-
nels, from Ref. 16 respectively.
The HBr elimination reactions ͑2͒ and ͑3͒ could proceed
in a variety of ways. The VBr is initially excited to a number
of closely spaced purely repulsive potential energy
1
3,22
surfaces.
Direct dissociation from a barrierless repulsive
distribution of the vibrational energy through a 3-centered
elimination mechanism.26 The modification to the traditional
PST entails using vibrational frequencies from the transition
state for the PST calculation with the idea that the energy is
partitioned late in the exit channel when the VX molecule is
in the transition state geometry on the potential energy sur-
face.
surface could occur, or the initial excitation could be fol-
lowed by internal conversion to the ground state leaving the
VBr with significant excess vibrational energy. Since the mo-
lecular elimination of HBr involves the breaking of two
bonds, excess vibrational energy would be required for this
concerted elimination channel. It is believed, therefore, that
the elimination of HBr occurs from the ground state of VBr
after internal conversion.13 The two possible mechanisms for
the elimination of HBr on the ground state potential energy
surface of VBr would result in different C H products; vi-
The vibrational energy distribution of HBr should also
reflect the transition state structure in the dissociation and
therefore be different for the 3 versus 4-centered elimination
mechanism. This effect can be illustrated intuitively by ex-
amining the Franck–Condon overlap of the H–Br internu-
clear distance in the transition state of excited VBr onto the
HBr bond length as a function of vibrational energy, which
should give the final vibrational distribution of the
fragment.27 The Rydberg–Klein–Rees ͑RKR͒ method is
used to reconstruct the ground state potential energy surface
of HBr. By comparing the internuclear distance at the clas-
sical turning point for a particular vibrational level with the
internuclear distance of the hydrogen and bromine in the
different photodissociation transition states a qualitative pre-
diction for the vibrational distribution of the HBr product can
be obtained. The transition state internuclear H–Br bond dis-
tance should correlate with the HBr vibrational level having
the highest population after the dissociation. The population
distribution of the various vibrational levels can then be de-
scribed by a Gaussian function centered at this most probable
vibrational level.27 The width of the distribution will depend
on the amount of energy in excess of the dissociation energy.
From bond distances of the transition state structures cal-
culated by Abrash et al.16 the internuclear H–Br distance for
the 4-centered transition states is found to be 2.09 Å whereas
in the 3-centered transition state the H–Br distance is only
1.73 Å. Figure 7 shows the calculated transition state geom-
etries. The Franck–Condon overlap depicts that the
4-centered elimination reaction would give a vibrational en-
ergy distribution centered at vϭ6 and extending to approxi-
mately vϭ12 quanta of HBr stretching. However, the experi-
mental results do not show populated levels higher than v
ϭ7. The observed population distribution is biased toward
the low vibrational levels, which is more in line with the
3-centered elimination reaction which depicts the largest
population at vϭ2, with levels up to vϭ6 having substantial
population.
2
2
nylidene for the 3-centered and acetylene for the 4-centered
elimination channels.
The product vibrational state distribution also yields in-
formation about the dissociation mechanism. Dissociation
from the purely repulsive electronically excited state of VBr
would happen via a fast fragmentation with little time for
redistribution of the internal energy. The hallmark of fast
impulsive dissociation reactions such as these is a strongly
inverted vibrational distribution. On the other hand, dissocia-
tion from the ground electronic state following internal con-
version provides ample time for energy redistribution and a
statistical distribution of energy in the products is likely. Fur-
thermore, we would expect the product state distribution to
reflect the structure of the transition state of the reaction thus
aiding in the determination of the reaction mechanism.
Studies monitoring the product distribution of the hydro-
gen halide fragments from the 193 nm dissociation of Vinyl
Chloride ͑VCl͒ and Vinyl Fluoride ͑VF͒, for instance, have
shown that the vibrational population peaks at low quanta of
HX stretch, supporting the idea of dissociation from the
ground potential surface following internal conversion.2
VCl vibrational populations up to vϭ6 were identified with
a bimodal rotational distribution corresponding to 3-centered
28
3,24
͑high J͒ and 4-centered ͑low J͒ elimination channels. The
3-centered elimination channel was found to be almost seven
times more probable, and gave a vibrational distribution
peaking at vϭ1 and extending to vϭ6. This finding is con-
sistent with results from VCl isotopic studies showing that
the 3-centered elimination channel is dominant in the HCl
5
elimination. Similarly, the VF product distributions peak at
vϭ1 although the VF photodissociation reactions showed a
higher degree of vibrational energy deposited in the HF
product as emission up to vϭ8 was observed.
The HX vibrational distribution from the VCl and VF
findings have been successfully modeled using modified
phase space theory ͑PST͒ to predict the expected statistical
A quantitative calculation of the vibrational level distri-
bution has been performed by Raff et al.17 using classical
trajectory calculations on the ground state PES at 6.44 eV of
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.70.241.163 On: Mon, 22 Dec 2014 17:14:29

1740
J. Chem. Phys., Vol. 115, No. 4, 22 July 2001
Liu, Letendre, and Dai
isomerization process, a number of collisions has taken place
and the vibrational and rotational energy content of acetylene
would decrease. The collision model in Sec. III depicts a
decrease of more than half of the HBr vibrational energy
from tϭ0 to 1.0 ␮s ͑Fig. 6͒. It is therefore quite reasonable
to find the acetylene vibrational energy at more than 14 kcal/
mol at 1 ␮s.
V. CONCLUSION
Time-resolved Fourier transform IR emission spectros-
copy with 0.5 ␮s resolution allowed the determination of the
internal energy content of fragment HBr and C H following
2
2
photolysis of the precursor, vinyl bromide, at 193 nm. Data
taken 1 ␮s after the laser photolysis reveals excited HBr with
vibrational levels of at least up to vϭ6 populated.
A collisional quenching model is coupled to the emission
spectra at 1.0 ␮s and later times following photolysis for
extrapolating the nascent vibration–rotation distribution for
HBr. The nascent population fits a Boltzmann distribution for
both the vibration and rotation with a temperature of 8690
and 7000 K respectively.
FIG. 8. The normalized vibrational population distribution for the HBr
product of 193 nm vinyl bromide photodissociation reactions. The theoret-
ical prediction of the HBr distribution from Ref. 17 is also shown.
VBr. The PES was determined by using a global analytic
potential energy hypersurface that was fit to the results of ab
initio electronic structure calculations. They conclude that
for energies far above the reaction barrier for any of the
elimination pathways, dynamical rather than energetic con-
siderations determine the most favorable elimination mecha-
nism. As predicted by the intuitive Franck–Condon argu-
ment based on the transition state structure, the calculation
does show that the HBr vibrational distribution is biased to-
ward the lower vibrational levels for the 3-centered channel,
Fig. 8. The HBr vibrational distribution, however, has not
been found for the 4-centered calculated mechanism.
The energy distribution is consistent with the dissocia-
tion model where the electronically excited VBr first under-
goes internal conversion to the highly vibrationally excited
ground state and then the energy is partitioned in the exit
channel of the reaction through a 3-centered transition state
giving vinylidene and HBr. The vinylidene product should
have about 24 kcal/mol of internal energy ͑or 64 kcal/mol
above the acetylene zero point energy͒. At this energy vi-
nylidene isomerizes to acetylene, resulting in acetylene
emission detected at 3294 (␯ ), 1342 (␯ ϩ␯ ), and
Excellent agreement between experimental and theoreti-
cal distributions is found and shown in Fig. 8, supporting the
3
4
5
Ϫ1
7
50 cm (␯ ).
5
3
-centered elimination mechanism. These results also show
ACKNOWLEDGMENTS
that the modified PST calculation gives an excellent indica-
tion of the proper dissociation mechanism and that it is not
necessary to invoke impulsive models with energetic con-
straints to describe vinyl halide photodissociation.
This work is supported by the Basic Energy Sciences,
U.S. Department of Energy, through Grant No. DEFG 02-
8
6ER 134584. One of the authors ͑L.T.L.͒ acknowledges the
receipt of a NASA Earth System Science Fellowship ͑Ref.
No. ESS/98-0000-0117͒. D.K.L. acknowledges support from
the Ministry of Education, Taiwan, Republic of China. The
authors are grateful to Professor Lionel Raff for stimulating
discussions and providing the data in Fig. 8 prior to its pub-
lication ͑Ref. 17͒.
B. Energy partitioning in acetylene and HBr
The energy available for partitioning into the in-
ternal degrees of freedom of the VBr photofragments is
1
22 kcal/mol for the HBr and acetylene products. Consider-
ing the theoretically calculated15 and experimentally
1
1
5
M. Ahmed, D. S. Peterka, and A. G. Suits, J. Chem. Phys. 110, 4248
determined zero point energy of vinylidene to be about 40
kcal/mol above the acetylene zero point energy, about 82
kcal/mol is available for the HBr and vinylidene products.
Analysis of HBr spectral features reveals 17 kcal/mol of vi-
brational energy in up to 6 quanta of stretching vibration and
͑
͑
1999͒; A. Fahr, P. Hassanzadeh, and D. B. Atkinson, Chem. Phys. 236, 43
1998͒; I. R. Slagle, J.-Y. Park, M. C. Heaven, and D. Gutman, J. Am.
Chem. Soc. 106, 4356 ͑1984͒.
A. M. Wodtke, E. J. Hintsa, J. Somorjai, and Y. T. Lee, Isr. J. Chem. 29,
2
3
4
3
83 ͑1989͒.
S. A. Abrash, Ph.D. dissertation, University of California at Berkeley,
987.
1
4 kcal/mol in rotational energy. Time of flight results of
1
2
Wodtke et al. allow determination of the average relative
translational energy of the acetylene and HBr products from
the flight times of these molecules which is found to be 27
kcal/mol. Subtracting all of these energies, approximately 24
kcal/mol of energy is available for the internal excitation of
the vinylidene. At this energy, vinylidene should isomerize to
acetylene on a time scale much faster than the experimental
time resolution. Assuming this to be the case, the initial
acetylene excitation would be 64 kcal/mol. By 1 ␮s after the
S. A. Abrash, M. T. McMahon, and R. W. Zehner, J. Phys. Chem. 98,
11909 ͑1994͒.
5
6
7
8
D. J. Donaldson and S. R. Leone, Chem. Phys. Lett. 132, 240 ͑1986͒.
T. R. Fletcher and S. R. Leone, J. Chem. Phys. 88, 4720 ͑1988͒.
S.-R. Lin and Y. P. Lee, J. Chem. Phys. 111, 9233 ͑1999͒.
G. V. Hartland, W. Xie, H.-L. Dai, A. Simon, and M. J. Anderson, Rev.
Sci. Instrum. 63, 3261 ͑1992͒.
L. T. Letendre, H.-L. Dai, I. A. McLaren, and T. J. Johnson, Rev. Sci.
Instrum. 70, 18 ͑1999͒.
C. D. Pibel, E. Sirota, J. Brenner, and H.-L. Dai, J. Chem. Phys. 108,
1297, ͑1998͒.
9
10
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:
30.70.241.163 On: Mon, 22 Dec 2014 17:14:29
1

J. Chem. Phys., Vol. 115, No. 4, 22 July 2001
Photolysis of vinyl bromide
1741
1
1
19
A. M. Wodtke, E. J. Hintsa, J. Somorjai, and Y. T. Lee, Isr. J. Chem. 29,
83 ͑1989͒.
D. H. Rank, U. Fink, and T. A. Wiggins, J. Mol. Spectrosc. 18, 170 ͑1965͒.
H. J. Babrov, J. Chem. Phys. 40, 831 ͑1964͒.
R. J. Malins and D. W. Setsa, J. Chem. Phys. 73, 5666 ͑1980͒; D. J.
Douglas and C. B. Moore, ibid. 70, 1769 ͑1979͒.
J. Park, R. Bersohn, and I. Oref, J. Chem. Phys. 93, 5700 ͑1990͒.
S.-R. Lin, S.-C. Lin, Y.-C. Lee, Y.-C. Chou, I.-C. Chen, and Y.-P. Lee, J.
Chem. Phys. 114, 7396 ͑2001͒.
20
3
1
1
1
1
2
3
4
5
21
H. Katayanagi, N. Yonekuira, and T. Suzuki, Chem. Phys. 231, 345
1998͒.
G. J. Mains, L. M. Raff, and S. A. Abrash, J. Phys. Chem. 99, 3532
1995͒.
͑
22
23
͑
L. T. Letendre, D. K. Liu, C. D. Pibel, J. B. Halpern, and H. L. Dai, J.
Chem. Phys. 112, 9209 ͑2000͒.
24
D. J. Donaldson, D. G. Watson, and J. J. Sloan, Chem. Phys. 68, 95
Y. Osamura, H. F. Schaefer III, S. K. Gray, and W. H. Miller, J. Am.
Chem. Soc. 103, 1904 ͑1981͒; T. Carrington, Jr., L. M. Hubbard, H. F.
Schaefer III, and W. H. Miller, J. Chem. Phys. 80, 4347 ͑1984͒; K. M.
Ervin, J. Ho, and W. C. Lineberger, ibid. 91, 5974 ͑1989͒; Ph. Halvick, D.
Liotard, and J. C. Rayez, Chem. Phys. 177, 69 ͑1993͒.
͑
1982͒.
2
2
2
5
6
7
Y. Huang, Y. A. Yang, G. X. He, and R. J. Gordon, J. Chem. Phys. 99,
752 ͑1993͒.
2
D. J. Donaldson, D. G. Watson, and J. J. Sloan, Chem. Phys. 68, 95
͑1982͒; D. J. Donaldson and J. J. Sloan, J. Chem. Phys. 82, 1873 ͑1985͒.
R. Schinke, Photodissociation Dynamics ͑Cambridge University Press,
Cambridge, 1993͒.
R. Rydberg, Z. Phys. 73, 376 ͑1931͒; O. Klein, ibid. 76, 226 ͑1932͒; A.
Rees, Proc. Physic. Soc. 59, 998 ͑1947͒.
16
S. A. Abrash, R. W. Zehner, G. J. Mains, and L. M. Raff, J. Phys. Chem.
9
9, 2959 ͑1995͒.
1
1
7
8
A. Rahaman and L. M. Raff, J. Phys. Chem. ͑in press͒.
G. V. Hartland, D. Qin, and H. L. Dai, J. Chem. Phys. 102, 8677 ͑1995͒;
28
1
07, 2890 ͑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:
30.70.241.163 On: Mon, 22 Dec 2014 17:14:29
1