Direct measurement of eigenstate-resolved unimolecular dissociation rates of HOCI

Article abstract of DOI:10.1063/1.474227

Double-resonance overtone excitation prepares HOCl molecules in single rovibrational states above the unimolecular dissociation threshold in the ground electronic state. Detecting the OH dissociation fragments allows us to observe which reactant states li

Full text of DOI:10.1063/1.474227

Direct measurement of eigenstate-resolved unimolecular dissociation rates of HOCI
M. R. Wedlock, R. Jost, and T. R. Rizzo
Citation: The Journal of Chemical Physics 107, 10344 (1997); doi: 10.1063/1.474227
View online: http://dx.doi.org/10.1063/1.474227
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/107/23?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
A classical trajectory study of the intramolecular dynamics, isomerization, and unimolecular dissociation of HO2
J. Chem. Phys. 139, 084319 (2013); 10.1063/1.4818879
The barrier height, unimolecular rate constant, and lifetime for the dissociation of HN 2
J. Chem. Phys. 132, 064308 (2010); 10.1063/1.3310285
Vibrationally resolved rate coefficients and branching fractions in the dissociative recombination of O 2 +
J. Chem. Phys. 122, 014302 (2005); 10.1063/1.1825991
State-to-state unimolecular reaction dynamics of HOCl near the dissociation threshold: The role of vibrations,
rotations, and IVR probed by time- and eigenstate-resolved spectroscopy
J. Chem. Phys. 111, 7359 (1999); 10.1063/1.480058
State selected unimolecular dissociation of HOCl
J. Chem. Phys. 107, 3730 (1997); 10.1063/1.474729
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:
130.88.90.110 On: Sat, 20 Dec 2014 20:16:13

LETTERS TO THE EDITOR
The Letters to the Editor section is divided into four categories entitled Communications, Notes, Comments, and Errata.
Communications are limited to three and one half journal pages, and Notes, Comments, and Errata are limited to one and
three-fourths journal pages as described in the Announcement in the 1 July 1997 issue.
COMMUNICATIONS
Direct measurement of eigenstate-resolved unimolecular dissociation
rates of HOCI
M. R. Wedlock
´
´
´ ´
Laboratoire de chimie physique moleculaire (LCPM), Ecole Polytechnique Federale de Lausanne, CH-1015
Lausanne, Switzerland
R. Jost
Grenoble High Magnetic Field Laboratorya) MPI and CNRS BP 166 F-38042 Grenoble Cedex 9, France
T. R. Rizzo
´
´
´ ´
Laboratoire de chimie physique moleculaire (LCPM), Ecole Polytechnique Federale de Lausanne, CH-1015
Lausanne, Switzerland
͑Received 21 August 1997; accepted 8 October 1997͒
Double-resonance overtone excitation prepares HOCl molecules in single rovibrational states above
the unimolecular dissociation threshold in the ground electronic state. Detecting the OH dissociation
fragments allows us to observe which reactant states lie above or below the dissocation threshold
and determine that threshold to be 19 290.3Ϯ0.6 cmϪ1 . Dissociation rates from single,
well-characterized eigenstates of the parent molecule exhibit fluctuations of more than an order of
magnitude. © 1997 American Institute of Physics. ͓S0021-9606͑97͒04047-6͔
I. INTRODUCTION
levels of 6␯ ͑OH stretch͒. The vibrational density of states
1
of HOCl at this energy is approximately 0.1 per cmϪ1 ,and
The most rigorous tests of unimolecular reaction rate
theories come from experiments in which one specifies com-
pletely the initial state of the reactant molecule and detects
individual state͑s͒ of the products. Such state-to-state experi-
ments have the potential to critically test both ab initio cal-
culations of molecular potential energy surfaces as well as
techniques for predicting the dynamics on such surfaces.
Several groups have measured unimolecular dissocation
rates from single eigenstates of a highly excited reactant
molecule on the ground potential energy surface, either by
linewidth analysis of transititions to quasibound states as in
the case of D2CO,1,2 HCO3–5, and DCO,6,7 or by spectro-
scopically probing the decay of the excited reactant, as in the
the overtone spectra presented herein show no perturbations
due to coupling to other vibrational levels. We can therefore
consider the dissociation process to be the decay of a pure
6␯ level into the dissociative continuum. We measure the
1
dissociation rate directly by time-resolved laser-induced
fluorescence detection of the OH products. Our results indi-
cate that the unimolecular dissociation rates exhibit large
fluctuations even with such well-defined vibrational eigen-
state preparation.
II. EXPERIMENTAL APPROACH
The experimental approach for these double-resonance
overtone spectroscopy experiments, shown schematically in
Fig. 1, has been described previously.15,16 Briefly, an infrared
laser pulse, generated by difference frequency mixing the
output of a narrowband ͑0.02 cmϪ1 Nd:YAG pumped dye
laser with the single mode output of the YAG in a LiNbO3
crystal, is used to prepare molecules in single rovibrational
of HFCO8 and CH3O,9,10 or by spectroscopic detection of the
dissociation fragments as in the case of HN3 .11,12 One theme
that has been emerging from these studies is that at this level
of detail, the unimolecular dissociation rates depend on the
vibrational character of the reactant state, either in a mode
specific manner8 or from the statistical nature of the highly
excited eigenstate.2,10,13,14
We report here direct time-resolved measurements of the
unimolecular dissociation of HOCl molecules prepared in
single eigenstates near threshold with carefully controlled
vibrational character and completely defined rotational quan-
tum numbers. By using double-resonance vibrational over-
tone excitation, we put HOCl molecules in single rotational
states containing two quanta of ␯ ͑OH stretch vibration͒.
1
Approximately 15 ns later, a second Nd:YAG pumped dye
laster ͑0.02 cmϪ1 bandwidth͒ promotes these preselected
molecules to single rovibrational states of 6␯1 . If the excited
rovibrational state is above the energy needed to break the
O–Cl bond, the molecule will dissociate to produce OH ϩ
Cl. We detect the overtone absorption by monitoring the OH
dissociation fragments via laser-induced fluorescence ͑LIF͒
in the A–X band using the frequency doubled output of a
third Nd:YAG pumped dye laser. The delay between the
second and third lasers is variable but typically set to ϳ15
a͒
The Grenoble High Magnetic Field Laboratory is ‘‘Laboratoire conven-
tionne aux universites UJF et INP de Grenoble.’’
10344
J. Chem. Phys. 107 (23), 15 December 1997
0021-9606/97/107(23)/10344/4/$10.00
© 1997 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:
130.88.90.110 On: Sat, 20 Dec 2014 20:16:13

Letters to the Editor
10345
FIG. 1. Energy level schematic for 2←0, 6←2 double-resonance overtone
excitation used to access single rovibrational eigenstates in the 6␯ band of
1
HOCl.
ns. The frequencies of the dye lasers are calibrated using a
Burleigh WA-4500 pulsed wave meter. As shown in Fig. 1,
we denote the rotational quantum numbers in the ground
vibrational state with double primes and those in the inter-
mediate state with single primes. The quantum numbers for
the dissociating molecules are unprimed.
FIG. 2. Series of double-resonance vibrational overtone excitation spectra of
the 6␯1←2␯ band of HOCl. Rotational states in the 2␯ intermediate level
1
1
preselected in the first step have K ϭ3 and J ϭ6–15. The total OH fluo-
Ј
Ј
rescence is recorded as a function of the frequency of the dissociating laser.
The dotted lines indicate the progression of the QP (J ),QQ (J ), and
Ј
Ј
3
3
Because the 2␯ band of HOCl is sufficiently sparse,17
QR (J ) transitions from each intermediate state.
1
Ј
3
we can pick out single rovibrational states from a low-
pressure sample at room temperature. Moreover, the high
resolution of the state-selection laser allows us to resolve the
chlorine isotopes in the first step, and hence we observe vi-
brational overtone spectra of single rotational states of a
single isotopic species. The data presented herein pertain ex-
clusively to the 35Cl isotope.
We prepare HOCl by simply mixing Cl2O, which is syn-
thesized according to the method of Schack and Lindahl,18
with H2O in a 1:1 ratio. Equilibrium is rapidly attained be-
tween these two species and HOCl, with the latter having a
fractional concentration of 14%.19 The resultant mixture
flows from a glass bulb through a glass fluorescence cell at
approximately 40–50 mTorr as measured by a capacitance
manometer. The total OH fluorescence emission passes
through a visible-blocking colored glass filter and is imaged
onto a UV sensitive photomultiplier tube. The resulting sig-
nal is integrated and transferred to a 486 computer via a
CAMAC interface.
obtained by preparing molecules in single rovibrational
states of 2␯ in the first step and then scanning the frequency
1
of the second laser while collecting the total OH fluorescence
emission induced by the third laser. For the data shown in
this figure, the infrared state selection laser prepares HOCl
molecules in 2␯ with K ϭ3, and J ranging from 6 to 15,
Ј
Ј
1
and the laser induced fluorescence probe laser excites the
Q1(1) line of the OH products. Because each of these spec-
tra that terminate on the 6␯ vibrational level originates from
1
a single rotational state in 2␯1 , one observes only three tran-
sitions: QP (J ),QR (J ), and a weaker QQ (J ) transition
Ј
Ј
Ј
3
3
3
that decreases rapidly in intensity with increasing J .The fact
Ј
that there are no ⌬KϭϮ1 transitions indicates that the tran-
sition moment is parallel to the A axis. The simplicity of the
spectra testifies to the state selectivity of the double-
resonance technique and makes the assignment of the spectra
straightforward. Moreover, the lack of splitting of these
bands indicates that the final states reached are not perturbed.
Although the transition frequencies shown Fig. 2 appear as
expected for a parallel band of a prolate symmetric rotor, as
is discussed below, the transition intensities are not.
III. RESULTS AND DISCUSSION
A. Photofragment spectra and assignments
Figure 2 shows a series of double resonance vibrational
overtone excitation spectra in the 6␯1←2␯ band of HOCl
1
J. Chem. Phys., Vol. 107, No. 23, 15 December 1997
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:
130.88.90.110 On: Sat, 20 Dec 2014 20:16:13

10346
Letters to the Editor
B. Dissociation energy
Because we measure vibrational overtone action spectra
by detecting the OH dissociation product of the highly ex-
cited molecules, only transitions to final states that are above
the dissociation threshold will appear in the spectrum. This
allows us to precisely determine the dissociation energy for
the O–Cl bond in HOCl. In Fig. 2, the spectra resulting from
excitation out of the K ϭ3;J Ͼ7 intermediate states con-
Ј
Ј
tain all three expected transitions, QP (J ), QQ (J ), and
Ј
Ј
3
3
QR (J ), although the transition intensities are not as ex-
Ј
3
pected. The spectrum resulting from excitation via the K
Ј
ϭ3; J ϭ7 intermediate state contains only the QQ3(7) and
Ј
QR3(7) transitions that reach final states with Kϭ3 and J
ϭ7 and 8, respectively. The disappearance of the QP3(7)
transition indicates that the Kϭ3; Jϭ6 rotational level in
the 6␯ vibrational level is below the dissociation threshold
1
and does not dissociate. We confirm this by measuring the
spectrum obtained by excitation via the K ϭ3; J ϭ6 inter-
Ј
Ј
mediate state in which we observe only the QR3(6) transition
to the Kϭ3, Jϭ7 final state. We therefore conclude that the
dissociation threshold lies somewhere between the energies
of the 6␯ Kϭ3, Jϭ6 rotational state and the Kϭ3, Jϭ7
1
state.
We determine the dissociation threshold more precisely
by measuring spectra in other K manifolds. In the Kϭ2
manifold, for example, we find that the dissociation thresh-
old occurs between the Jϭ14 and Jϭ15 final states. The
interleaving of energy levels in the Kϭ2 and Kϭ3 mani-
folds allows us to place the dissociation threshold between
the energy of the Kϭ3, Jϭ7 state ͑19 290.84 cm -1͒, which
leads to dissociation and the state with Kϭ2, Jϭ14
͑19 289.66 cmϪ1͒, which does not dissociate. We therefore
FIG. 3. Dissociation rates of HOCl in the 6␯ vibrational band. The upper
panel shows the change in laser-induced fluroescence from the OH disso-
ciation product as a function of the delay between the overtone pump and
1
LIF probe lasers for the 6␯ Kϭ3; Jϭ8 and Kϭ3; Jϭ9 states of HOCl.
1
Exponential fits to the early portion of the data are shown as solid lines. In
the lower panel dissociation rates are plotted versus the amount of excess
energy above the dissociation threshold. Rates for final states with Kϭ3 and
Jϭ7–22 are shown. The dotted vertical lines indicate excess energies at
which new OH product rotational levels become energetically available.
From left to right the dotted lines are, respectively, the energetic thresholds
determine the dissociation energy to be 19 290.3Ϯ0.6 cmϪ1
.
The final state energies are determined by adding the excita-
tion wave number of the second laser to the eigenvalue of the
intermediate state, as determined from the infrared spectra of
2␯ of Cavazza and co-workers.20
1
2
2
for the
uct states.
⌸ ⌸3/2(Nϭ3) OH prod-
3/2(Nϭ2),2⌸1/2(Nϭ1),2⌸1/2(Nϭ2) and
C. Dissociation rates
The upper panel of Fig. 3 shows the early portion of the
time evolution of the unimolecular dissociation obtained by
fixing the frequencies of the state-selection laser and the
overtone excitation laser to produce parent HOCl molecules
in two different initial states, fixing the probe laser on the
Q1͑1͒ transition of the OH product, and scanning the time
delay between the overtone pump laser and the LIF probe
laser. As the HOCl molecule decays producing OH frag-
ments in the probed level, the LIF intensity increases with a
rate that reflects the unimolecular dissociation rate of the
parent. The total pressure in the fluorescence cell for these
experiments was 50 mTorr. At this pressure, the hard sphere
gas kinetic collision rate is approximately 5ϫ105 per sec-
ond, corresponding to one collision every 2 ␮s. Rotational
relaxation of parent molecules could conceivably occur as
much as 20 times the hard sphere rate or one rotational state
changing collision every 100 ns. We therefore estimate dis-
sociation rates by a single exponential fit of the early portion
As previously noted, transition intensities in the spectra
of Fig. 2 are not as expected for a parallel band of a near
prolate symmetric top. For example, the R transition arising
from the K ϭ3; J ϭ8 level ͑Fig. 2͒ appears to be anoma-
Ј
Ј
lously weak when compared to the P and Q transitions from
the same intermediate state. This results from collecting
spectra with a short ͑ϳ15 ns͒ delay between the dissociation
and probe lasers. If a molecule falls apart slowly compared
to this short delay, its intensity will appear smaller than ex-
pected. Indeed, the dissociation rate for the Kϭ3; Jϭ9 final
state is substantially smaller than that for the other two states
reached via the K ϭ3; J ϭ8 intermediate state ͑i.e., Jϭ7
Ј
Ј
and Jϭ8.) This distortion of spectral intensites occurs
throughout the set of spectra shown in Fig. 2. In measure-
ments with longer delays between the overtone pump laser
and the probe laser, all the transitions approach their ex-
pected intensity.
J. Chem. Phys., Vol. 107, No. 23, 15 December 1997
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:
130.88.90.110 On: Sat, 20 Dec 2014 20:16:13

Letters to the Editor
10347
of the curves. These should be of sufficient accuracy to show
the strong dependence of the dissociation rate on the initial
state of the parent molecule, although it is less accurate for
the slower decays. In a full publication to follow, a complete
kinetic analysis of the pressure-dependent curves will fit both
the rising and falling portions.
From the two time evolution curves shown in Fig. 3, one
can see that the unimolecular dissociation rate of HOCl mol-
ecules from the state with Kϭ3; Jϭ8 is about 15 times
faster than that at the next higher rotational level, Kϭ3; J
ϭ9. The lower panel shows the disociation rate as a function
of energy for the states with Kϭ3 and J ranging from 7 to
22. It is clear that over the 225 cmϪ1 range shown in this
figure, the dissociation rate varies sharply and nonmonotoni-
cally with energy and angular momentum. The sharp
changes in dissociation rate are not directly correlated with
the availability of new rotational levels for the OH product.
Dashed vertical lines in the figure represent the energies at
which new OH product rotational states become energeti-
cally accessible ͑see the figure caption͒.
Not shown on the plot of Fig. 3 are the dissociation rates
for HOCl molecules with other K states. We have measured
the dissociation rates of HOCl parent molecules from a few
states with Kϭ0 and find them to be as much as four times
faster than the fastest rate measured for molecules with K
ϭ3. We also observe, however, that Kϭ0 lines are per-
turbed by mixing with two other zeroth-order vibrational
states. It is not yet clear whether the higher rates in Kϭ0 are
evidence of an inverse dependence on the K quantum num-
ber or the result of mixing with other vibrational modes that
couple more strongly to the dissociative continuum. If the
latter is true, the variations that we observe at Kϭ3 might
result from small components of other vibrational states with
intensities too small for us to observe spectroscopically but
and variations of as much as a factor of 15 for states differing
in J by only one. These fluctuations do not vary monotoni-
cally with increasing energy and do not directly depend on
the opening of new product channels. Our initial measure-
ments show a strong negative K dependence to the dissocia-
tion rate. There is also some evidence of an increase in the
dissociation rate for states that have mixtures of other zeroth-
order vibrational modes.
With eigenstate-resolved unimolecular dissociation mea-
surements, a molecule as small as HOCl should provide criti-
cal tests both of calculations of potential energy surfaces as
well as of the dissociation dynamics on such surfaces.
ACKNOWEDGMENTS
´
We gratefully acknowledge the support of the Ecole
´ ´
Polytechnique Federale de Lausanne ͑EPFL͒ and the Fonds
National Suisse through Grant No. 20-45808.95. Professor
R. Jost would like to particularly thank the EPFL for a vis-
iting professorship that supported his stay in Lausanne dur-
ing which this work was performed. We also thank Arnaud
Allanic for providing us with Cl2O from which we synthe-
sized HOCl and L. Fusina for providing the transition fre-
quencies and vibrational term values for 2␯1.
1 W. F. Polik, D. R. Guyer, and C. B. Moore, J. Chem. Phys. 92, 3453
͑1990͒.
2 W. F. Polik, D. R. Guyer, W. H. Miller, and C. B. Moore, J. Chem. Phys.
92, 3471 ͑1990͒.
3 G. W. Adamson, X. Zhao, and R. W. Field, J. Mol. Spectrosc 160, 11
͑1993͒.
4 D. W. Neyer, X. Luo, I. Burak, and P. L. Houston, J. Chem. Phys. 102,
1645 ͑1995͒.
5 J. D. Tobiason, J. R. Dunlop, and E. A. Rohlfing, J. Chem. Phys. 103,
1448 ͑1995͒.
6 C. Stock, X. N. Li, H. M. Keller, R. Schinke, and F. Temps, J. Chem.
Phys. 106, 5333 ͑1997͒.
7 H. M. Keller, M. Stumpf, F. Schroder, C. Stock, F. Temps, R. Schinke, H.
J. Werner, C. Bauer, and P. Rosmus, J. Chem. Phys. 106, 5359 ͑1997͒.
8 Y. S. Choi and C. B. Moore, J. Chem. Phys. 97, 1010 ͑1992͒.
9 A. Geers, J. Kappert, F. Temps, and J. W. Wiebrecht, J. Chem. Phys. 99,
2271 ͑1993͒.
that dissociate significantly more rapidly than the pure 6␯
1
eigenstates. Measurements in progress on the other K mani-
folds will help clarify the dependence on K as well as the
effect of vibrational state mixing.
10 S. Dertinger, A. Geers, J. Kappert, J. Wiebrecht, and F. Temps, Faraday
Discuss. Chem. Soc. 102, 31 ͑1995͒.
IV. CONCLUSIONS
11 B. R. Foy, M. P. Casassa, J. C. Stephenson, and D. S. King, J. Chem.
Phys. 90, 7037 ͑1989͒.
Double-resonance vibrational overtone excitation allows
us to obtain eigenstate-resolved spectra of molecules in
highly excited vibrational states. By applying this technique
to the HOC1 molecule we precisely determine the dissocia-
12 B. R. Foy, M. P. Casassa, J. C. Stephenson, and D. S. King, J. Chem.
Phys. 92, 2782 ͑1990͒.
13 W. H. Miller, R. Hernandez, C. B. Moore, and W. F. Polik, J. Chem. Phys.
93, 5657 ͑1990͒.
tion energy for the O–Cl bond to be 19 290.3Ϯ0.6 cmϪ1
.
14 R. Hernandez, W. H. Miller, C. B. Moore, and W. F. Polik, J. Chem. Phys.
99, 950 ͑1993͒.
Measurements of the dissociation threshold for different K
manifolds at a range of J values are consistent, implying that
centrifugal barriers to dissociation do not have a significant
impact on our determination.
We also directly measure unimolecular dissociation rates
for single, well-characterized eigenstates of HOCl lying near
the dissociation threshold. These measurements show wide
variations in the dissociation rate over a range of 225 cmϪ1
15 B. Kuhn, O. V. Boyarkin, and T. R. Rizzo, Ber. Bunseges. Phys. Chem.
101, 339 ͑1997͒.
16 X. Luo and T. R. Rizzo, J. Chem. Phys. 94, 889 ͑1991͒.
17 F. Cavazza, G. D. Lonardo, R. Escribano, L. Fusina, P. C. Gome, and J.
Ortigoso, J. Mol. Spectrosc. 159, 395 ͑1993͒.
18
C. J. Schack and C. B. Lindahl, Inorg. Nucl. Chem. Lett. 3, 397 ͑1967͒.
19 M. -L. Junttila, W. J. Lafferty, and J. B. Burkholder, J. Mol. Spectrosc.
164, 583 ͑1994͒.
20 L. Fusina ͑personal communication͒.
J. Chem. Phys., Vol. 107, No. 23, 15 December 1997
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:
130.88.90.110 On: Sat, 20 Dec 2014 20:16:13