Collisional excitation of CO2(0000) vibrationless state by the highly translationally excited HCN produced from the photodissociation of s-triazine

Article abstract of DOI:10.1016/S0009-2614(98)00808-2

The collisional excitation of CO₂(00⁰0) state by quenching of highly translationally excited HCN has been studied with high-resolution diode laser absorption spectroscopy. The translationally hot HCN is produced from the photodissociation of s-triazine (1,3,5-triazine) at 266 nm. The average translational energy of HCN is measured to be 18±1 kcal/mol, which corresponds to 84% of available energy for the HCN fragment. The fraction of available energy partitioned into the rotational excitation is 15%. A little excitation in the HCN bending vibration is observed. The nascent rotational excitation in the 00⁰0 levels of CO₂, due to collisions with the translationally hot HCN (from s-triazine), is accompanied by a substantial enhancement in the translational energy. The excitation in the translational recoil velocities appears to increase with increasing J in 00⁰0 vibrational state.

Full text of DOI:10.1016/S0009-2614(98)00808-2

4
September 1998
Chemical Physics Letters 293 Ž1998. 383–390
0
ž
/
Collisional excitation of CO 00 0 vibrationless state by the
2
highly translationally excited HCN produced from the
photodissociation of s-triazine
Jeunghee Park )
Department of Chemistry, Korea UniÕersity, 208 Seochangdong, Jochiwon, Chungnam 339-700, South Korea
Received 20 February 1998; in final form 11 May 1998
Abstract
0
The collisional excitation of CO Ž00 0. state by quenching of highly translationally excited HCN has been studied with
2
high-resolution diode laser absorption spectroscopy. The translationally hot HCN is produced from the photodissociation of
s-triazine Ž1,3,5-triazine. at 266 nm. The average translational energy of HCN is measured to be 18"1 kcalrmol, which
corresponds to 84% of available energy for the HCN fragment. The fraction of available energy partitioned into the
rotational excitation is 15%. A little excitation in the HCN bending vibration is observed. The nascent rotational excitation in
0
the 00 0 levels of CO , due to collisions with the translationally hot HCN Žfrom s-triazine., is accompanied by a substantial
2
enhancement in the translational energy. The excitation in the translational recoil velocities appears to increase with
0
increasing J in 00 0 vibrational state. q 1998 Elsevier Science B.V. All rights reserved.
1
. Introduction
Despite numerous interest for collisional relaxation phenomena, the detailed mechanism for the collisional
energy transfer between the translationally hot molecule and a bath gas collision partner has not yet been
established at the quantum-state resolved level. Recently, high-resolution Žinfrared. IR diode laser devices have
been used to probe the vibrational, rotational and translational energy of bath molecules, thereby providing
details of the energy transfer process that have not in general been available using other experimental methods.
In this work, the quantum resolved details in the collisional relaxation of translationally hot HCN by a CO bath
2
is investigated using the IR diode laser probe technique.
It is well established that s-triazine Ž1,3,5-triazine. yields three HCN fragments following the absorption of a
photon. In the past years, the photodissociation of s-triazine has been of great interest because of the high
symmetry of the dissociation process, each excited s-triazine molecule yielding three HCN molecules. Ondrey
and Bersohn measured the translational energies of the HCN fragments produced at the excitation wavelength of
2
48 and 193 nm by TOF technique w1x. They suggested that the initial 3-fold symmetry would be retained during
)
Corresponding author. E-mail: parkjh@tiger.korea.ac.kr
0
009-2614r98r$ - see front matter q 1998 Elsevier Science B.V. All rights reserved.
PII: S0009-2614Ž98.00808-2

3
84
J. ParkrChemical Physics Letters 293 (1998) 383–390
dissociation and that dissociation must take place from the ground state of the parent molecule, and is therefore
likely to generate nearly equal speeds in all fragments. They measured that the HCN molecules have an average
kinetic energy of 10 kcalrmol Ž41% of the available energy. at 248 nm but only 2 kcalrmol Ž11% of the
available energy. at 193 nm. Goates et al. observed the vibrationally excited HCN molecules produced from the
photodissociation of s-triazine at 193 nm by IR fluorescence spectroscopy w2x. Recently, Huber and his
coworkers reported the average translational energy of HCN produced from the photodissociation of s-triazine
at several ultraviolet ŽUV. wavelengths by using the molecular beam photofragment translation spectroscopy,
and argued that the dissociation mechanism is a two-step concerted process rather than simultaneous one-step
process w3x. They measured the translational energies of HCN molecules to be about 55% and 70% of available
energy at 193 nm and other wavelengths Ž248, 275, 285 and 295 nm., respectively, which are quite different
from the result reported by Ondry and Bersohn. Therefore, it is necessary to determine the translational-state
distribution of HCN before undertaking the collisional energy transfer involving HCN molecule. A high-resolu-
tion diode laser was employed to probe the state-specific energy deposition in the HCN produced from the
photodissociation of s-triazine.
The basic experimental scheme can be illustrated by the following equations:
l
sytriazineqhn
Ž
266 nm
.
™3HCN
Ž
mn p, J,V
.
Ž
HCN production
.
l
l
HCN
Ž
mn p, J,V
.
qhn ;3.1 mm
Ž
.
™HCN
Ž
mn pq1, J"1,V
.
Ž
Diode laser probe ,
.
where m, n and p are the quantum numbers for the symmetric stretching Žn ., bending Žn . and asymmetric
1
2
stretching Žn . vibrational modes. J and l denote, respectively, the rotational and vibrational angular
3
momentum quantum numbers and V denotes the translational velocity along the direction of the diode beam.
The nascent translational-, rotational- and vibrational-state distributions of HCN were probed in the P branch of
the strongly allowed antisymmetric stretching band at ;3.1 mm. The experimental results as described in the
following sections show that the HCN molecule carries high translational energy such as 18"1 kcalrmol.
High-resolution single-collision study on the energy transfer between the translationally excited HCN
molecule produced from the photodissociation of s-triazine and CO bath molecule were performed. The IR
2
diode laser probe technique is also used to measure the rotational and translational excitation of vibrationally
ground CO2 resulting from the collisions with the translationally hot HCN. It provides the information in
exquisite detail about the energy transfer process from translationally hot triatomic molecule to the quantum
states of CO2 bath acceptor molecules. The basic experimental approach is described by the following
equations:
sytriazineqhn
Ž
266 nm
.
™3HCN
Ž
HCN production ,
.
l
HCNqCO ™HCNqCO
Ž
mn p, J,V
.
Ž
Collisionalquenching
.
,
2
2
l
l
CO2
Ž
mn p, J,V
.
qhn
Ž
;4.3 mm
.
™CO2
Ž
mn pq1, J"1,V
.
Ž
Diode laser probe .
.
The nascent rotational populations and translational recoils of CO within the ground vibrational manifold
2
are probed in the P branch of the strongly allowed antisymmetric stretching band at 4.3 mm.
2
. Experimental
The diode laser absorption technique has been described in detail previously and only a brief description is
given here w4–6x. The photolysis laser is a Continuum Nd:YAG 266 nm laser. For measuring the nascent-state
distributions of HCN produced by photodissociation, a 3.1 mm cw diode laser ŽM u¨ tek. is propagated collinearly
with the UV laser beam through a 2.7 m Pyrex cell containing less than 8 mTorr. All measurements were made
with Nd:YAG laser repetition rates of 0.1;0.15 Hz in order to ensure the pumping of residual gas. For the

J. ParkrChemical Physics Letters 293 (1998) 383–390
385
collision experiment, a 4.3 mm cw diode laser ŽM u¨ tek. is propagated co-linearly with the UV laser beam
through a 2.0 m Pyrex cell containing approximately 13 mTorr of a flowing 1:4 gas mixture of s-triazine and
CO . The IR light intensity transmitted through the cell is measured using a liquid nitrogen-cooled InSb
2
photodetector which, together with its amplifier, has a rise time of approximately 450 ns. In order to insure that
the detected IR light from a single laser mode, the transmitted IR is passed through a monochromator before
being focused onto the photodetector. The absorption signals are recorded and averaged over 1000 cycles of the
modulated diode laser by a LeCroy 9354A digital oscilloscope.
Transient IR absorption signals are collected with the diode laser locked to the center of the relevant
absorption line. The signal from this detector is used as the reference for an SR510 lock-in amplifier ŽStanford
Research Systems.. This arrangement allows the diode laser frequency to be locked to any absorption line of
interest. The transient absorption signal is input into two channels on the digital storage oscilloscope, one
ac-coupled and the other dc-coupled. The ac-coupled channel records D I, the change in IR absorption while the
dc-coupled channel represents IqD I, the absolute intensity of the transient of IR signal. For each excimer shot,
the ac-coupled signal is divided by the dc-coupled signal to yield D IrŽIqD I., which can be algebraically
rearranged to give ŽD IrI., the desired quantity on a shot-to-shot basis without the use of a pair of matched
detectors. Transient signals are typically averaged over ;70 excimer shots fired. Additionally, a reference line
technique is used for all population measurements in order to account for change in the signal intensity due to
long-term drifts in the system.
Active locking of the diode laser frequency to a HCN or CO2 absorption line is accomplished using a
separate reference cell. A fraction Ž;10%. of the diode beam is sent through a reference cell and a
monochromator and detected by an additional InSb detector. The IR absorption signal from the reference
detector is input to a lock-in amplifier, the output of which is used as an error signal sent to the diode laser
controller, thus keeping the diode laser frequency locked to the absorption transition.
For the line-width measurements, the diode laser frequency was locked to an IR fringe of a Laser Photonics
Confocal Etalon ŽModel SP 5945. using a 1 kHz modulation frequency. The modulation amplitude was roughly
y1
the width of a fringe, 0.002 cm , and precise diode laser frequency control was accomplished by computer-
controlled rotation of a scanner plate internal to the etalon. Transient absorption signals were obtained while the
etalon fringe was over the absorption line of interest. The excimer laser was fired at the center of the modulation
cycle in order to sample only a single diode laser frequency for a given etalon fringe position.
s-Triazine ŽAldrich, 97%. was purified by several freeze Ž77 K.–pump–thaw cycles. Research grade CO2
ŽMatheson, 99.995%. was used without further purification.
3
. Results and discussion
3
.1. Translational, rotational and Õibrational energy distribution of HCN produced from the photodissociation
of s-triazine
0
0
y1
The Doppler profile of the HCNŽ00 0, Js23.™HCN Ž00 1, Js22. absorption line at 3238.20 cm is
shown in Fig. 1a. The Doppler profiles were measured at 450 ns after the excimer laser pulse. The pressure of
s-triazine was 4 mTorr and the repetition rate of photolysis laser was 0.1 Hz. The full width at half maximum
0
0
ŽFWHM. averaged over this line and other rotational lines such as HCNŽ00 0, Js22.™HCNŽ00 1, Js21.
0
0
y1
and HCNŽ00 0, Js20.™HCNŽ00 1, Js19. is 0.040"0.002 cm . However, when 8.5 mTorr of s-triazine
is used, the top of Doppler profile becomes less flat and its FWHM is reduced by a few percent, indicating the
Doppler profile measured even at 4 mTorr may not reveal the nascent velocity distribution of HCN. The
1
1
1
1
Doppler profiles of HCNŽ01 0, Js14.™HCNŽ01 1, Js13. and HCNŽ01 0, Js17.™HCNŽ01 1, Js16.
absorption lines have been measured for 4 mTorr and 8.5 mTorr of s-triazine, showing the same FWHM as
y1
1
1
0
.040"0.002 cm . The Doppler profile of HCNŽ01 0, Js14.™HCNŽ01 1, Js13. at 8.5 mTorr is shown

3
86
J. ParkrChemical Physics Letters 293 (1998) 383–390
0
0
Fig. 1. Ža. Nascent Doppler profile line shapes of the HCNŽ00 0, Js23.™HCNŽ00 1, Js22. absorption line produced by photodissocia-
y1
1
1
tion of 4 mTorr s-triazine. The FWHM is 0.04 cm . Žb. The Doppler profile of the HCNŽ01 0, Js14.™HCNŽ01 1, Js13. absorption
y1
line produced by photodissociation of 8.5 mTorr s-triazine. The FWHM is 0.04 cm . The Doppler profiles Ža. and Žb. were measured at
4
50 ns after the excimer laser pulse and the repetition of Nd:YAG 266 nm laser was 0.1 Hz.
0
in Fig. 1b, which is similar to Fig. 1a. The pressure effect on the Doppler profiles of 00 0 rotational lines is due
to the thermal HCN quenched from other highly excited ro-vibrational states. We took the averaged FWHM for
0
0
1
1
y1
several HCNŽ00 0.™HCNŽ00 1. and HCNŽ01 0.™HCNŽ01 1. transition lines as 0.040"0.002 cm
.
If the velocity distribution is isotropic, which cannot be checked under our collinear pump and probe beams
alignment, the velocity distribution can be obtained from the first derivative of the Doppler profile. The average
translational energy of HCN is calculated to be 18"1 kcalrmol, which is consistent with 17 kcalrmol reported
by Huber and coworkers for the 248 nm photodissociation of s-triazine. The reaction enthalpy for the
dissociation is known as 43.0 kcalrmol, so the fraction of available energy partitioning into translational state is
8
4%.
0
0
Typical time-resolved signal observed while monitoring the HCNŽ00 0, Js20.™HCNŽ00 1, Js19.
y1
transition line at 3248.48 cm
is shown in Fig. 2a. The IR diode laser light absorbed in the Õ™Õq1
transition is proportional to the population difference between Õ and Õq1 levels:
2
N
Ž
Õ, J
.
yN
Ž
Õq1, J"1
.
A
Ž
D IrIo
.
r X R n ,
1
Ž .
J
where D I is the change in the amount of IR light transmitted following the reaction, I is the total IR intensity
o
at frequency n, X is the H o¨ nl–London factor given by Ž Jq1.rŽ2 Jq1. for R branch lines and JrŽ2 Jq1.
J
0
0
Fig. 2. Ža. Time-resolved absorption signals observed while monitoring the HCNŽ00 0, Js20.™HCNŽ00 1, Js19. transition line at
y1
3
248.48 cm . The pressure of s-triazine is 4 mTorr and the repetition of Nd:YAG laser was 0.1 Hz. Žb. Boltzmann plot of the nascent
0
rotational distribution of HCNŽ00 0. produced by the photodissociation of s-triazine. The line is the best fit to a rotational temperature of
1
750 K. The NJ is the experimentally determined number of molecule produced in a given rotational state of HCN.

J. ParkrChemical Physics Letters 293 (1998) 383–390
387
for P branch lines and R is the transition matrix element. In the present experiment, the upper-state population
has been determined to be negligible compared to the lower-state populations. Thus the observed transient
absorption signals reflect the temporal behavior of the population in the lower ro-vibrational level only. The
nascent population of ro-vibrational states was obtained by measuring the absorption signal amplitude at 500 ns
after the excimer laser pulse. In order to confirm that the observed transient absorption signals come from single
photon absorption of s-triazine, the absorption signals have been measured as a function of laser intensity,
showing a good linearity. The rotational Boltzmann plot of the nascent rotational-state distribution of
0
HCNŽ00 0. is shown in Fig. 2b. The best fit to a Boltzmann distribution gives a temperature of T s1750 K.
R
0
The average of several experimental runs provides T s1750"100 K for the 00 0 state and T s1350"100
K for the 01 0 state. The rotational-state distribution has been measured for Js17–25 for 00 0 state, and
Js11–19 for 01 0 state. Since the translational energy distribution of 00 0 and 01 0 states are about same, the
rotational temperature can be expected to be similar, which is consistent with the results. The rotational
temperature of HCN is ;5 times of room temperature and the average rotational energy is 3.2"0.2 kcalrmol.
Thus the fraction of available energy partitioned into the rotational excitation is 15%.
Since the translational- and rotational-state distributions for the 00 0 and 01 0 states of HCN were measured,
the relative populations of these two vibrational levels were determined by measuring relative absorption
amplitudes for some rotational levels in the 00 0 and 01 0 states. The population ratio of 01 0 to 00 0 state is
.4. The population in other vibrational states was found to be negligible. Therefore, about 70% of HCN
R
R
1
0
1
0
1
0
1
0
1
1
0
0
fragments is in the vibrational ground state, and an average vibrational energy is estimated to be 0.5"0.1
kcalrmol above the HCN zero point level, which is close to the residual energy calculated by subtracting the
average translational energy and rotational energy from the available energy of HCN. Table 1 lists the average
translational, rotational and vibrational energies of the HCN molecules produced by the photodissociation of
s-triazine at 266 nm.
)
The UV absorption spectrum of s-triazine in the region of 238–326 nm shows two overlapping n™p
1
1
Y
transitions; a very weak one which is due to the forbidden doubly degenerated A ™ E and a much stronger
1
1
1 Y
one which is due to the symmetry allowed A ™ A w7x. The transition dipole moment for the transition
1
2
1
1 Y
A ™ A is perpendicular to the molecular plane of s-triazine. It is known that s-triazine is simply
1
2
decomposed to three HCN molecules. There are two likely mechanisms for the photodissociation of s-triazine.
The one is a symmetric simultaneous single process and the other one is a sequential process. In the case of
single step dissociation, three HCN fragments have same kinetic energy in the center-of-mass frame of
s-triazine molecule, and in the case of sequential reaction two different translational energies are expected.
Recently, Pai et al. calculated the stable points and the transition states along the reaction path for a
concerted triple and a stepwise decomposition mechanisms of s-triazine by using non-local density functional
methods ŽNDFT. and ab initio methods w8,9x. All methods predict that the low-energy pathway to decomposition
is a concerted triple dissociation and that a weakly bound cyclic ŽHCN. clusters exists on the reaction path.
3
Osamura et al. also concluded that the triple dissociation of s-triazine occurs on the ground-state potential
surface and is a one-step concerted reaction from the ab initio calculations w10x. All these calculations for the
geometry and nuclear motion of the transition state show that the bending mode of HCN should be excited in
HCN molecules after the dissociation, which is consistent with our result. Unfortunately, the translational
Table 1
Average energy partitioning of HCN
hn
07.7
EAVAL
3²ET :
54 Ž84%.a
3²ER :
3²EV :
1
64.5
9.7 Ž15%.
1.5 Ž2%.
EAVAL shn-dissociation energy, ²E :, ²E : and ²E : are the average total translational, rotational and vibrational energy, respectively.
T
R
V
All energies are in unit of kcalrmol.
a
The percentage of available energy.

3
88
J. ParkrChemical Physics Letters 293 (1998) 383–390
energy distribution derived from the Doppler profile of ro-vibrational absorption line cannot give a definite
evidence for the concerted dissociation mechanism. Thus a sequential process could occur on similar time scale,
as suggested by Huber and his coworkers w3x.
3
.2. Energy transfer from translationally hot HCN to CO2
Transient populations of CO molecules in Js62, 72, 74, 78, 80 and 82 states of the ground vibrationless
2
level were monitored via IR diode laser absorption. A typical transient signal is shown in Fig. 3a, where the
0
0
y1
absorption on the CO Ž00 0, Js74.™CO Ž00 1, Js73. line at 2274.99 cm
is plotted against the time
2
2
relative to the excimer laser pulse. The early, linear rise in the signal represents predominantly single collision
excitation of CO to Js74 by translationally hot HCN. The total pressure of the 1r4 s-triazinerCO mixture
2
2
is adjusted to be 13 mTorr such that the mean time between ambient gas collisions is 2.3 ms, which is much
slower than the detector response time. The fast rise component of the signals is due to nascent rotational
excitation of CO by fast HCN. The nascent rotational distribution is shown in Fig. 3b. As described below, the
2
transient linewidths for Js62, 74 and 80 were also measured. Because the linewidth increases as a function of
rotational level J, one must integrate over the line shape in order to obtain rotational-state populations. All line
shapes fit exactly to the Gaussian functions, the signal at the line center was therefore corrected by the
linewidth. The nascent populations of Js72, 78 and 82 rotational states are corrected by the interpolated
linewidth from the measured one of Js62, 74 and 80 on the assumption that the FWHM increases linearly
with increasing J. The resulting rotational-state distributions fit a Boltzmann distribution with the temperature
of 1300"130 K, based on data for the states Js62–82.
The recoil velocities onto the probe beam axis were determined from the measurements of the Doppler
broadened line shape. All of the measured line shapes fit well to Gaussian forms, and the FWHM of Doppler
profile for the absorption transitions Js62, 74 and 80 are 0.011"0.005, 0.0125"0.005 and 0.0140"0.005,
respectively, as shown in the inset of Fig. 3b. Note that these linewidths are measured at ts700 ns after the
excimer laser pulse. The linewidth of a transition probing CO molecules with a room temperature velocity
2
y1
distribution is 0.00425 cm . Clearly CO molecules which are rotationally excited in high J states by the
2
collisions with fast HCN possess a large amount of translational energy and the translational recoil velocity
increases substantially as J increases from 62 to 82.
0
0
y1
Fig. 3. Ža. The transient signal observed while monitoring the CO Ž00 0, Js74.™CO Ž00 1, Js73. transition line at 2274.99 cm
.
2
2
The total pressure of the 1r4 s-triazinerCO mixture is 13 mTorr. Žb. Rotational population distribution of CO in high J levels of the
0 0 state produced by translationally excited HCN. The inset is the FWHM of the CO2 00 000 1 transition as a function of rotational
2
2
0
0
0
0
level J. The line widths are measured at ts700 s after the laser pulse.

J. ParkrChemical Physics Letters 293 (1998) 383–390
389
A
sim ple billiard ball m odel predicts
a
translational energy transfer of D E
2
s 2 M
MCO2
r
Ž
MHCN qM
.
E0, where MHCN and MCO2 are the HCN and CO mass and E is the
HCN
CO2
2
0
translational energy of hot HCN w11x. For a hard-sphere collision between HCN and CO , 47% of the HCN
2
translational energy may be transferred. This equation roughly predicts only the magnitude of the translational
0
recoil, thus a detailed trajectory calculation needs to explain the linewidth dependence of the 00 0 states. A
similar rotational dependence was observed in the transient linewidth measurement for the collisional excitation
of CO by hot atoms w11x. The breathing ellipsoid model was used to model the collisional energy transfer
2
between hot hydrogen atom and CO w12x. This model can be examined for the collision between hot HCN and
2
CO as a future work.
2
The collisional energy transfer from the vibrationally excited pyrazine to bath CO was studied by Flynn and
2
his coworkers w4–6x. Chesko et al. reported that pyrazine can be dissociated at a significant yield of HCN which
can recoil and produce the excited CO .by the collisions w13x. Flynn and coworkers carried out the experiment
2
to probe HCN evolution and found that the extent of dissociation is small at 1 ms delay time used in their
experiments and thus the energy transfer is not affected significantly w14x. The translational energy of HCN
produced from the s-triazine measured in this work is about same as that of HCN produced from the
photodissociation of the pyrazine at 248 nm measured by Chesko et al. However, the transient absorption of
rotationally excited CO produced by translationally hot HCN in this study ŽFig. 3a. show the faster rise rate
2
than that of the CO2 excited by vibrationally excited pyrazine in the work of Flynn and his coworkers.
Furthermore, the translation recoils of CO produced by translationally hot HCN are different from those of
2
CO2 in the Flynn and his coworkers’ work. Therefore, we also conclude that in the case of pyrazine, the
collisional CO excitation by HCN is insignificant.
2
4
. Conclusions
The collisional quenching of translationally excited HCN by CO were studied with high-resolution diode
2
laser absorption spectroscopy. It is first time to study the collisional energy transfer between the translationally
excited triatomic molecule and bath molecule at a quantum-state resolved level. The translationally hot HCN are
produced from Nd:YAG 266 nm photodissociation of s-triazine. The translational- and rotational-state distribu-
0
1
tions in 00 0 and 01 0 vibrational states of HCN were probed at 450 ns following the excitation by laser pulse.
On the assumption of isotropic velocity distribution, the average translational energy HCN is estimated to be
1
3
8"1 kcalrmol, which is 84% of available energy of HCN fragment. The average rotational energy of HCN is
.2"0.2 kcalrmol, thus the fraction of available energy partitioned into the rotational excitation is 15%. A
0
little excitation in the bending vibration is observed. The nascent rotational excitations in the 00 0 vibrational
levels of CO produced by translationally hot HCN are accompanied by a substantial enhancement in the CO
2
2
translational energy, and the translationally recoil velocity increases substantially with rotational level. A
detailed trajectory calculation needs to explain the experimental results.
Acknowledgements
This work was performed at Columbia University and supported by the Korea Ministry of Education
Foundation ŽBSRI-97-3432.. I gratefully acknowledge Prof. Flynn for support of the facility and a variety of
useful suggestions.
References
w1x G.S. Ondrey, R. Bersohn, J. Chem. Phys. 81 Ž1984. 4517.
w2x S.R. Goates, J.O. Chou, G.W. Flynn, J. Chem. Phys. 81 Ž1984. 4521.
w3x T. Gejo, J.A. Hrrison, J.R. Huber, J. Phys. Chem. 100 Ž1996. 13941.

3
90
J. ParkrChemical Physics Letters 293 (1998) 383–390
w4x A.S. Mullin, J. Park, J.Z. Chou, G.W. Flynn, R.E. Weston Jr., Chem. Phys. 175 Ž1993. 52.
w5x A.S. Mulin, C.A. Michaels, G.W. Flynn, J. Phys. Chem. 102 Ž1995. 6032.
w6x C.A. Michaels, A.S. Mullin, G.W. Flynn, J. Phys. Chem. 102 Ž1995. 6682.
w7x A. Bolovinos, P. Tsekeris, J. Philis, E. Pantos, D. Andritsopoulos, J. Mol. Spectrosc. 103 Ž1984. 240.
w8x S.V. Pai, C.F. Chabalowski, B.M. Rice, J. Phys. Chem. 100 Ž1996. 5681.
w9x S.V. Pai, C.F. Chabalowski, B.M. Rice, J. Phys. Chem. 100 Ž1996. 15368.
w10x Y. Osamura, M. Unno, K. Hashimoto, J. Am. Chem. Phys. 109 Ž1987. 1370.
w11x J.F. Heshberger, S.A. Hewitt, S.K. Sarkar, G.W. Flynn, R.E. Weston Jr., J. Chem. Phys. 91 Ž1989. 4636.
w12x T. Kreutz, G.W. Flynn, J. Chem. Phys. 93 Ž1990. 452.
w13x J.D. Chesko, D. Stranges, A.G. Suits, Y.T. Lee, J. Chem. Phys. 103 Ž1995. 6290.
w14x C.A. Michaels, H.C. Tapalian, Z. Lin, E.T. Sevy, G.W. Flynn, in press.