Rotational and vibrational energy distributions of HCl produced by three- and four-center eliminations of HCl from halogenated ethanes

Article abstract of DOI:10.1016/S0009-2614(99)00499-6

Rotational and vibrational energy distributions of HCl from CF₃CClFH and CClF₂CH₃ have been measured using a 2+1 resonantly enhanced multiphoton ionization (REMPI) technique. In the case of the three-center elimination from CF₃CClFH, the HCl products are rotationally and vibrationally cold. On the other hand, hot HCl products are produced in the case of the four-center elimination from CClF₂CH₃. The energy partitioning in the HCl eliminations has been discussed on the basis of structural change along the intrinsic reaction coordinate (IRC) or transition state structures obtained by ab initio molecular orbital (MO) calculations.

Full text of DOI:10.1016/S0009-2614(99)00499-6

25 June 1999
Ž
.
Chemical Physics Letters 307 1999 48–54
Rotational and vibrational energy distributions of HCl produced
by three- and four-center eliminations of HCl from halogenated
ethanes
b
Atsushi Yokoyama a,), Toshiyuki Takayanagi
a
Department of Materials Science, Japan Atomic Energy Research Institute, Tokai, Ibaraki 319-1195, Japan
b
AdÕanced Science Research Center, Japan Atomic Energy Research Institute, Tokai, Ibaraki 319-1195, Japan
Received 12 October 1998; in final form 17 February 1999
Abstract
Rotational and vibrational energy distributions of HCl from CF3CClFH and CClF2CH3 have been measured using a
Ž
.
2q1 resonantly enhanced multiphoton ionization REMPI technique. In the case of the three-center elimination from
CF3CClFH, the HCl products are rotationally and vibrationally cold. On the other hand, hot HCl products are produced in
the case of the four-center elimination from CClF2CH3. The energy partitioning in the HCl eliminations has been discussed
Ž
.
on the basis of structural change along the intrinsic reaction coordinate IRC or transition state structures obtained by ab
Ž
.
initio molecular orbital MO calculations. q 1999 Elsevier Science B.V. All rights reserved.
1. Introduction
nation, that is, typically a few tens of kJrmol for the
three-center elimination while 160–210 kJrmol for
the four-center elimination. In spite of this substan-
tial difference, the average translational energy is not
so much different between the three- and four-center
eliminations of HCl from the saturated chlorinated
hydrocarbons. This means that a smaller fraction of
the exit barrier is converted to the relative transla-
tional energy of the products in the four-center elimi-
nation than in the three-center elimination, and that
the internal energies of the products are expected to
be larger for the four-center elimination than for the
three-center elimination.
Elimination of HCl from chlorinated hydrocar-
bons is an important dissociation pathway in uni-
molecular dissociation of that in the ground elec-
w x
tronic state 1 . Usually, the elimination of HCl
occurs through three- or four-center transition states,
depending on the molecular structure and the activa-
tion energies of both transition states.
Translational energy distributions released from
the three- and four-center eliminations of HCl have
been measured for several molecules using
w
x
photofragment translational spectroscopy 2–5 . The
exit barrier for the three-center elimination is gener-
ally much smaller than that for the four-center elimi-
Several workers have studied internal state distri-
butions of HCl produced by the elimination of HCl
from several kinds of chloroethylenes following in-
ternal conversion from the initially prepared pp)
)
Corresponding author. E-mail: jun@beam.tokai.jaeri.go.jp
0009-2614r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S0009-2614 99 00499-6

(
)
A. Yokoyama, T. TakayanagirChemical Physics Letters 307 1999 48–54
49
w x
state to the ground state. Berry 6 measured vibronic
an advantage for the determination of the transition
state of the HCl elimination, because the isomeriza-
tion of the molecules does not take place. We em-
state distributions of HCl from CH 2CHCl,
CH2CDCl, CH2CCl2 , cis- and trans-CHClCHCl,
and CHClCCl2 using a chemical laser technique. He
concluded that the distributions were non-statistical,
and proposed a bootstrap reaction dynamic model to
explain the non-statistical distributions. Donaldson
Ž
.
ployed infrared multiphoton excitation IRMPE for
preparing highly vibrationally excited molecules. The
average excess energy above the transition state of
the molecule excited by the IRMPE is generally
w x
Ž
.
and Leone 7 have measured the vibrational energy
small 80–130 kJrmol . Therefore, employment of
this excitation technique is advantageous for the
elucidation of energy partitioning of exit barrier.
distributions of HCl from CH2CHCl and trans-
CHClCHCl using time-resolved FTIR emission spec-
troscopy. The measured distributions are in good
agreement with Berry’s distributions. However, they
concluded that the distributions were statistical from
the calculation with the HCl vibrational frequency at
the transition state instead of that of free HCl. Gor-
2. Experimental and ab initio calculation proce-
dures
w x
don et al. 8 measured the rotational distributions of
HCl from CH2CHCl using a 2q1 REMPI tech-
Y
Ž
.
nique. The rotational distributions of HCl Õ s1, 2
The experimental apparatus used in this work is
w
x
were the Boltzmann-like distributions, while the dis-
the same as that in Ref. 13 . The CTEFE and CDFE
molecules in supersonic molecular beams were dis-
sociated at 9.294 and 10.504 mm, respectively, by a
CO2 laser Lumonics TEA-840 . The CO2 laser
light, the pulse shape of which was a 200 ns spike
followed by a 2 ms tail, was focused with a ZnSe
Y
Ž
.
tribution of HCl Õ s0 consisted of two compo-
nents with different rotational temperatures. Similar
distributions were also observed for three isomers of
dichloroethylene 9,10 . Several speculations have
been done for the origin of these two different
Ž
.
w
x
w x
Ž
.
distributions: At first, Gordon and co-workers 8
lens fs30 cm . The molecules were irradiated at
30, 430 and 1000 Jrcm2 of the laser fluence at the
focus. The HCl product was probed by a 2q1
REMPI technique combined with time-of-flight mass
spectrometry. The rotational state distributions of
HCl were determined by measuring the REMPI spec-
proposed the possibility of the contributions of the
three- and four-center eliminations of HCl. Later,
they proposed another mechanism that the elimina-
w
x
tion of HCl was vibrationally adiabatic 11 . Sato et
al. discussed the origin of the distributions on the
basis of the competition of the three- and four-center
eliminations. They suggested the importance of 1,2-
shifts of H and Cl atoms for the elimination of HCl
1
X
X
1
Y
Y
Ž
.
qŽ
.
tra of the two-photon F D Õ , J § X S Õ , J
w
x
transition 14,15 . The probe laser light was pro-
duced by frequency doubling of 480–505 nm light
w
x
Ž
.
from 1,1-dichloroethylene 5,9 . A theoretical work
supports that the 1,2-shifts of the H and Cl atoms are
energetically possible for the three isomers of
from a Nd:YAG Laser Continuum Powerlight7010
Ž
pumped dye laser Lambda Physik SCANMATE-
2EY . The light was focused with a quartz lens
fs30 cm . The output of the light was typically 1
mJ. The probe laser pulse, the duration of which was
;10 ns, was fired at 1.2 or 4.3 ms after the dissocia-
tion laser pulse.
In order for understanding the difference in dy-
namics between the three- and four-center elimina-
tions, Ab initio MO calculations were done at the
MP2 FC level of theory. Wadt and Hay’s effective
core potentials 16 accompanied with split valence
.
w
x
Ž
.
dichloroethylene 12 . These H and Cl atom migra-
tions complicate the dissociation pathways for
chloroethylenes, and make the understanding of the
energy disposal to the products from the three- and
four-center eliminations of HCl difficult.
In this Letter, the characteristics of the internal
energy distributions of the HCl product are examined
for the three- and four-center eliminations from 2-
Ž
.
Ž
.
w
x
chloro-1,1,1,2-tetrafluoroethane CTEFE and 1-
Ž
.
Ž
.
chloro-1,1-difluoroethane CDFE , respectively.
Compared with the elimination of HCl from
chloroethylene, the chlorofluoroethane systems have
plus polarization functions ECPDZP were used as
basis functions. All calculations were carried out
using a Gaussian 92 program 17 .
w
x

(
)
50
A. Yokoyama, T. TakayanagirChemical Physics Letters 307 1999 48–54
3. Results
3.1. Three-center HCl elimination from CTEFE
Ž
.
The IR multiphoton dissociation IRMPD of
2
Ž
.
Ž
CTEFE was done at high 1010 Jrcm and low 30
Jrcm2 peak fluence. In Fig. 1 are shown REMPI
.
spectra for HCl produced by the three-center elimi-
nation of HCl from CTEFE at the high fluence:
CF3CHClF ™ CF3CFqHCl .
1
Ž .
Ž
.
At the low fluence the REMPI spectrum of the 0, 1
Ž
.
and 1, 2 bands was not recorded because of low
signal intensity. The rotational state distributions of
Ž
.
Ž
.
HCl P Erot are shown in Fig. 2. The P Erot ’s were
obtained from the measured rotational line intensities
1
X
X
1
Y
Y
Ž
.
qŽ
.
of the two-photon F D Õ , J § X S Õ , J
Fig. 2. Rotational energy distributions of HCl from CF3CHClF at
Y
Y
Ž .
Ž .
b
Ž
.
Ž
B
.
Ž .
'
a
high and
low fluences; v Õ s0,
Õ s1, and
ÕY s2. The solid lines indicate the best fits to the Boltzmann
distribution.
transition. Since the intensity loss of the rotational
lines occurs in this transition, the line intensities of
the measured spectra were corrected using experi-
mental correction factors for the rotational line
Y
w
x
Ž
strength 15 . The relative population of HCl Õ s
1, JY was determined using the vibrational correc-
tion factor 18 for the sensitivity difference between
1, 1 and 0, 0 band. Since the vibrational correc-
tion factor for the 1, 2 band has not been reported,
the relative population of HCl Õ s2, J was sim-
ply determined by scaling the rotational line intensi-
.
w
Ž
x
Ž
.
.
Ž
.
Y
Y
Ž
.
Ž
.
ties of the 1, 2 band with the average ratio of the
Ž
.
rotational line intensities of the 1, 1 band to those
Fig. 1. Intensity corrected REMPI spectrum of HCl from
CF3CHClF at high fluence.
Ž
.
of the 0, 1 band and with the ratio of the Franck–

(
)
A. Yokoyama, T. TakayanagirChemical Physics Letters 307 1999 48–54
51
s0 : ÕY s1 : ÕY s2 at high CO2 laser fluence and
1.0 : 0.52"0.32 for ÕY s0 : ÕY s1 at low fluence.
Table 1
²
:
Average rotational energies Erot of HCl produced by the three-
center HCl elimination from CF3CHClF
²
: Ž .
Erot kJrmol
Õ
3.2. Four-center HCl elimination from CDFE
high fluence
low fluence
0
1
2
23"5
5.4"0.8
5.9"0.8
9"2
6"2
–
The IRMPD of CDFE was done at peak fluence
of 430 Jrcm2. Fig. 3 shows the REMPI spectrum of
HCl produced by four-center elimination of HCl
from CDFE:
CClF2CH3 ™ CF2 5CH2 qHCl .
2
Ž .
Ž
.
Ž
.
Condon factor for the 0, 1 band to that for the 1, 2
Ž
.
Ž
.
The P Erot ’s are shown in Fig. 4. Only the P Erot
for ÕY s2 can be well fitted to a Boltzmann distribu-
band. The solid lines in Fig. 2 indicate the best fits to
the Boltzmann distributions. The average rotational
energies at the high and low fluences are listed in
Table 1. The relative populations of the vibrational
levels obtained by the integration of the rotational
distributions are 1.0 : 0.29"0.12 : 0.12"0.05 for ÕY
tion with an average energy of 5.4"1.4 kJrmol as
Y
Ž
.
shown in Fig. 4. The fit of the P Erot for Õ s1 to
Ž
.
a Boltzmann distribution is poor, and the P Erot for
ÕY s0 could not be fitted to a Boltzmann distribu-
Ž
.
tion. Since all P Erot ’s increase with the rotational
energy in the range of the measured rotational lines,
the relative populations of the vibrational levels could
not be calculated by summing up the relative popula-
tions of all significant rotational levels. The average
relative population ratio of HCl at different vibra-
tional states by summing up the relative populations
over all measured rotational levels is 1.0:4.5"
1.3:92"26 for ÕY s0:ÕY s1:ÕY s2. Therefore, the
yields of each vibrational state of HCl should be
ÕY s0-ÕY s1-ÕY s2.
Fig. 4. Rotational energy distributions of HCl from CClF2CH3;
Y
Y
Fig. 3. Intensity corrected REMPI spectrum of HCl from
CClF2CH3.
Õ s0, B Õ s1, and ' ÕY s2. The solid lines indicate
Ž
v
. Ž .
Ž .
the best fits to the Boltzmann distribution.

(
)
52
A. Yokoyama, T. TakayanagirChemical Physics Letters 307 1999 48–54
kJrmol 1. Therefore, most of the exit barrier is
.
4. Discussion
released as the internal energy of the HCl and
CH2CF2 fragments. Since the population of the HCl
at the vibrational state higher than ÕY s2 could not
be determined in this study, the average vibrational
energy of HCl cannot be deduced. The average
vibrational energy, however, is expected to be at
least 67 kJrmol, because the vibrational population
The internal state distributions of HCl are remark-
ably different for the three- and four-center elimina-
tion reactions studied in this work. In the case of the
three-center elimination of HCl from CTEFE, the
Y
Ž
.
P Erot ’s for Õ s0 to 2 are well reproduced by
Boltzmann distributions, and the vibrational popula-
tions are ÕY s0)ÕY s1)ÕY s2. On the other
Y
Y
Ž
.
Ž
of HCl Õ s2 should be much larger than HCl Õ
.
hand, in the case of the four-center elimination from
s0, 1 . Statistical partitioning of the exit barrier to
all degrees of freedom of the fragments results in the
monotonically decreasing vibrational populations
with increasing vibrational quantum numbers. The
average vibrational energy of HCl with the statistical
vibrational distribution is 6.3 kJrmol. Even if we
use the vibrational frequency of HCl at the transition
state according to Donaldson and Leone’s analysis
Y
Ž
.
CDFE, the P Erot ’s for Õ s0 and 1 could not be
fitted to a Boltzmann distribution, and the vibrational
populations are ÕY s0-ÕY s1-ÕY s2.
In the case of the three-center HCl elimination,
the exit barrier is small: For the elimination from
CTEFE, the exit barrier was estimated to be ;16
kJrmol from the product translational energy distri-
w x
w x
bution 3 . Moreover, most of the barrier energy is
7 , the average energy is only 22 kJrmol. There-
w
x
released as translation 2,3 . Thus, the internal en-
ergy of HCl comes mainly from excess energy above
the exit barrier. A small conversion fraction of the
exit barrier to the vibrational states of HCl should
indicate that the HCl and CF3CF are almost formed
at the transition state. In fact, the H–Cl distance at
fore, the observed vibrational distribution indicates
that the energy partitioning to HCl is non-statistical
for the four-center elimination of HCl, and more
energies are partitioned to HCl vibration.
The favorable partitioning of the exit barrier to
HCl vibration is due to dynamics beyond the transi-
tion state and can be qualitatively explained by
structural change along IRC shown in Fig. 5, where
the length along IRC is represented by s. The H–Cl
˚
the transition state is calculated to be 1.38 A, which
is only 0.11 A longer than the normal HCl in con-
trast with the H–Cl distance 1.77 A of HCl pro-
duced by the four-center elimination.
˚
˚
Ž
.
˚
distance drastically changes from 1.77 A at ss0 to
The average rotational energy in the ÕY s0 state
decreases with decreasing laser fluence. This behav-
ior is explained by the change of the average excess
energy. In an IRMPE the average excitation energy
of a molecule before dissociation is determined by
the competition of the excitation process to upper
vibrational levels above the dissociation limit with
the dissociation process. The excitation rate is pro-
portional to laser intensity, which is proportional to
laser fluence if the laser pulse width is constant in
such case as this experiment. Therefore, the average
excess energy is expected to be higher at higher
fluence.
˚
1.35 A at ss0.9 with the potential energy decrease
of 80.4 kJrmol in early stage of reaction. Since
other coordinates such as the distance between the
center-of-masses of HCl and CH2CF2 and the C–C
distance change less than the H–Cl distance in this
stage, the potential energy released in this stage is
expected to be converted mainly to HCl vibration.
w
x
Kato and Morokuma 22 did an IRC analysis more
quantitatively for four-center elimination of HF from
CH3CH2 F. In their calculation, the potential energy
1 The exit barrier is calculated using the heat of formation of
In the case of four-center elimination, the exit
barrier, which is typically 160–210 kJrmol, is much
higher than that for the three-center elimination, and
only 20–30% of the exit barrier is released as the
relative translational energy. In the HCl elimination
from CDFE, the average relative translational energy
Ž
w
x.
Ž
CH3CClF2 y529.7 kJrmol 19 , that of CH2CF2 y335 kJrmol
w
x. x.
Ž
w
20 , that of HCl y92 kJrmol 20 , and the activation energy
Ž
w
x. w x
289 kJrmol 21 . Sudbø et al. 2 calculated to be 230 kJrmol
by taking the heat of formation of CH3CClF2 to be y490
kJrmol, which was obtained by the additivity rule. The exit
Ž
.
barrier 162 kJrmol obtained using an ab initio MO method at
MP2rECPDZP level of theory supports our estimation using the
Ž
.
Ž
50 kJrmol is only 27% of the exit barrier 184
w x
experimental heat of formation of CH3CClF2 cited in Ref. 19 .

(
)
A. Yokoyama, T. TakayanagirChemical Physics Letters 307 1999 48–54
53
Ž
.
The comparison of P Erot ’s for the four-center
elimination with that for the three-center elimination
shows that HCl produced by the four-center elimina-
tion is rotationally excited more than that produced
by the three-center elimination. The angle u between
the H–Cl bond and the line joining the center-of-
masses of HCl and CH3CClF2 changes at sF0.9 as
shown in Fig. 5. This change will produce torque
and give rotationally excited HCl molecules.
5. Conclusions
The rotational and vibrational energy distributions
of HCl produced by the four- and three-center elimi-
nations were measured. The distributions are remark-
ably different in two types of reactions: In the case
of the four-center elimination, the vibration and rota-
tion of HCl is excited significantly as a result of
dynamical effect during dissociation. This can be
explained on the basis of the structural change along
IRC beyond the transition state. On the other hand,
in the case of the three-center elimination, the vibra-
tional and rotational distributions are statistical, and
most of energy comes from the excess energy above
the transition state. This is because the exit barrier is
small and because the C–Cl distance is small at the
transition state.
Ž .
Fig. 5. Energy and structural change of CClF2CH3 along IRC s .
The insertion in the upper figure indicate the definition of struc-
tural parameters.
Acknowledgements
released from ss0 to ss0.5 is converted to kinetic
energy along the IRC and energy transfer between
IRC and the HF vibration takes place in the early
We wish to thank Mr. Ginji Fujisawa and Mr.
Kazukiyo Ebata for their assistance in the experi-
ment and Dr. Keiichi Yokoyama for his advice on
the ab initio calculations.
Ž
.
stage 0.5-s-0.83 . They estimated that two-thirds
of the IRC energy is converted to the HF vibration in
this region. If we assume that the energy transfer
between IRC and the HCl vibration is similar to the
case of the HF elimination, the HCl vibrational
energy amounts to ;53.6 kJrmol. If we consider
References
w x
1
P.J. Robinson, K.A. Holbrook, Unimolecular Reactions, Wi-
ley-Interscience, London, 1971.
Ž
the difference between the experimental 184
.
Ž
.
kJrmol and calculated 162 kJrmol exit barrier,
the estimated HCl vibrational energy becomes ;61
kJrmol. Although this estimation is rough, it can be
explained qualitatively that high vibrational excita-
tion in HCl occurs during dissociation, and this is in
good agreement with the present experimental re-
sults.
w x
2
Aa.S. Sudbø, P.A. Schulz, Y.R. Shen, Y.T. Lee, J. Chem.
Ž
.
Phys. 69 1978 2312.
w x
3
A. Yokoyama, K. Yokoyama, G. Fujisawa, J. Chem. Phys.
Ž
.
100 1994 6487.
w x
4
K. Yokoyama, G. Fujisawa, A. Yokoyama, J. Chem. Phys.
Ž
.
102 1995 7902.
w x
5
K. Sato, S. Tsunashima, T. Takayanagi, K. Yokoyama, G.
Ž
.
Fujisawa, A. Yokoyama, Chem. Phys. Lett. 232 1995 357.

(
)
54
w x
A. Yokoyama, T. TakayanagirChemical Physics Letters 307 1999 48–54
Ž
.
w
x
6
w x
7
M.J. Berry, J. Chem. Phys. 61 1974 3114.
15 Y. Xie, P.T.A. Reilly, S. Chilukuri, R.J. Gordon, J. Chem.
Ž .
Phys. 95 1991 854.
16 W.R. Wadt, P.J. Hay, J. Chem. Phys. 82 1985 284.
Ž
.
D.J. Donaldson, R. Leone, Chem. Phys. Lett. 132 1986
240.
P.T.A. Reilly, Y. Xie, R.J. Gordon, Chem. Phys. Lett. 178
w
w
x
x
Ž
.
w x
8
17 M.J. Frisch, G.W. Trucks, M. Head-Gordon, P.M.W. Gill,
Ž
.
1991 511.
M.W. Wong, J.B. Foresman, B.G. Johnson, H.B. Schlegel,
M.A. Robb, E.S. Replogle, R. Gomperts, J.L. Andres, K.
Raghavachari, J.S. Binkley, C. Gonzalez, R.L. Martin, D.J.
Fox, D.J. Defrees, J. Baker, J.J.P. Stewart, J.A. Pople, Gauss-
ian 92, Revision C.4, Gaussian, Inc., Pittsburgh, PA, 1992.
w x
9
K. Sato, Y. Shihira, S. Tsunashima, H. Umemoto, T.
Takayanagi, K. Furukawa, S. Ohno, J. Chem. Phys. 99
Ž
.
1993 1703.
w
w
w
w
w
x
10 G. He, Y.A. Young, Y. Huang, R.J. Gordon, J. Phys. Chem.
Ž
.
w
x
97 1993 2186.
18 W.R. Simpson, T.P. Rakitzis, S.A. Kandel, A.J. Orr-Ewing,
x
Ž .
R.N. Zare, J. Chem. Phys. 103 1995 7313.
11 Y. Huang, Y.-A. Yang, G. He, S. Hashimoto, R. Gordon, J.
Chem. Phys. 103 1995 5476.
12 J.-F. Riehl, D.G. Musaev, K. Morokuma, J. Chem. Phys. 101
Ž
.
w
w
x
x
Ž
.
19 G.S. Buckley, A.S. Rodgers, J. Phys. Chem. 87 1983 126.
20 S.W. Benson, Thermochemical Kinetics, Wiley-Interscience,
x
Ž
.
1994 5942.
New York, 1976.
x
w
x
13 K. Sato, S. Tsunashima, T. Takayanagi, G. Fujisawa, A.
21 P. Cadman, A.W. Kirk, A.F. Trotman-Dickenson, J. Chem.
Ž
.
Ž .
Soc., Faraday Trans. I 72 1976 996.
Yokoyama, Chem. Phys. Lett. 245 1995 432.
x
w x Ž .
22 S. Kato, K. Morokuma, J. Chem. Phys. 73 1980 3900.
14 R. Callaghan, S. Arepalli, R.J. Gordon, J. Chem. Phys. 86
Ž
.
1987 5273.