3732
Letters to the Editor
ening at ⌬Ͻ0.05 cmϪ1. This value allows us to place a
stack shown in Fig. 2͑a͒ and then scan the dye laser corre-
sponding to the second step of the double resonance excita-
tion process, we obtain the spectrum shown in Fig. 2͑b͒. This
action spectrum corresponds to excitation to the 7OH vibra-
tional level from the selected intermediate state͑s͒ and is ob-
tained by monitoring the yield of OH(Nϭ8,2⌸3/2) fragments
from the unimolecular dissociation of HOCl as the frequency
of the 2OH→7OH excitation laser is varied. As is clear from
the figure, we are able to resolve individual rotational tran-
sitions of this parallel band with our ϳ0.1 cmϪ1 resolution
dye laser. Thus, even though the resolution of our OPO pre-
vents complete state selection in the 2OH intermediate level,
lower limit on the dissociation lifetime of HOCl from the
7
level at ⌬tϾ1.06ϫ10Ϫ10 s (⌬tϭ1/2⌬). We can
OH
compare this result with predictions of RRKM theory where
the dissociation rate is given by k(E)ϭN](EϪE0)/h(E).
In the above expression N] corresponds to the number of
open channels at the transition state, h is Plank’s constant,
is the density of states of the reactant molecule and E0 is the
threshold energy for reaction. As is well known, application
of this expression to a barrierless potential, such as in HOCl,
is difficult due to the problems associated with locating the
transition state and thus calculating N](EϪE0). Typically in
these situations the transition state is determined by applying
variational methods to locate the region of the potential en-
ergy surface where the reactive flux is a minimum. For our
current purposes, however, it suffices to take N](EϪE0)
ϭ1, which gives a lower limit for the RRKM rate. As we
show below, even this lower limit RRKM rate is substan-
tially larger than that obtained from the measurements. Apart
from N](EϪE0), the other quantity required for the rate
calculation is the density of states, which using currently
available spectroscopic data,12͑b͒ we determine to be
ϭ0.2 states/cmϪ1 at the excitation energy. Using these val-
ues in the RRKM expression gives a lower limit estimate for
the unimolecular rate of: kRRKM(E)Ͼ1/h(E)ϭ1.5
the final rotational states of the 7
band are cleanly re-
OH
solved in these experiments allowing a state selective study
of the dissociation dynamics. By parking the OPO on various
other Ka stacks of the 2OH spectrum shown in Fig. 2͑a͒ and
then scanning the dye laser, we can effectively decompose
2OH→7
excitation spectrum into its Ka components.
Figure 2͑c͒ shows the results of selecting transitions belong-
OH
P
ing to the Q3 stack in the intermediate state.
Although extensive spectroscopic analysis exist for the
lower vibrational levels of HOCl,12 the 7 band has not
OH
been previously analyzed. The rotational assignment shown
in Fig. 2͑b͒, is based on a analysis using an asymmetric rotor
fitting program13 and has been subsequently verified using
combination differences of the well characterized 2 in-
ϫ1011 sϪ1
, which corresponds to a lifetime of 6.7
OH
termediate state. In these experiments for confirming spectral
assignment, we fix the dye laser ͑corresponding to the sec-
ond step of the double resonance scheme͒ on a particular
ϫ10Ϫ12 s. The difference between this lifetime and that es-
timated from the linewidth measurements suggests that the
fundamental assumption of RRKM theory, that energy ran-
domizes on a time scale much faster than the rate of disso-
feature of the 7 band that we are interested in assigning,
OH
and then scan the wavelength of the IR laser to determine
ciation, appears not to hold for the 7 vibrational state of
OH
which states of the 2 level are connected to it through
dipole selection rules. The lines appearing in the resulting
spectra can easily be assigned using known spectroscopic
HOCl. Incomplete vibrational mode mixing in the molecule
is consistent with the fairly regular spectrum observed. We
have also tried to determine the dissociation rate of the
OH
constants for the 2 level and from this information, the
HOCl͑7 ͒ level directly by varying the time delay be-
OH
OH
assignment of the spectral feature in the 2OH→7OH spectra
can then be deduced. From a preliminary analysis of the
action spectra, such as the ones shown in Figs. 2͑b͒–2͑c͒, we
have obtained a set of approximate spectroscopic constants
for the band origin and rotational parameters of the HOCl35
tween the excitation and probe lasers, but have found the rate
to be faster than the time resolution afforded by our nano-
second laser system.
In addition to measuring the linewidths, we have also
examined the reaction dynamics by monitoring the OH frag-
ment internal state distributions resulting from the unimo-
lecular dissociation. Figure 3, shows the product rotational
(7
)
OH
band. These are ϭ21 709.07 cmϪ1
ϭ15.156 cmϪ1, Bϭ0.503 cmϪ1, and Cϭ0.487 cmϪ1
,
A
.
2
An examination of the linewidths associated with the
resonances appearing in Kaϭ1 subband ͓i.e., Figs. 2͑b͔͒ in-
dicates that they are limited by the linewidths of our dye
laser (ϳ0.1 cmϪ1). For example, the feature indicated by the
arrow in Fig. 2͑b͒, which corresponds to the Jϭ12, Kaϭ1,
Kcϭ11, 12(7OH)←Jϭ13, Kaϭ1, Kcϭ12, 13(2OH) reso-
nance, has a linewidth of 0.11 cmϪ1 ͑FWHM͒. We are un-
able to resolve the asymmetry splitting associated with this
transition or the splitting due to the Cl35/Cl37 isotopes.
Within our resolution, the widths of the resonances in the
Kaϭ1 manifold appear to be independent of the J quantum
number of the parent molecule spanning a narrow range of
values around 0.11 Ϯ 0.01 cmϪ1 ͑FWHM͒. Taking into ac-
count the influence of Doppler broadening and the finite
spectral width of the excitation laser, we estimate an upper
limit contribution to the linewidth from homogeneous broad-
state distribution for the OH ⌸3/2 manifold obtained from
dissociation of the Jϭ12 state of HOCl͑7 ͒ for various
OH
values of Ka . These measurements clearly reveal that the
position and amplitude of the oscillations appearing in the
OH product state distribution are strongly dependent on the
rotational quantum numbers of the initial state. From the
Doppler widths associated with the laser induced fluores-
cence spectra of the OH fragment, it is possible to obtain an
estimate for the heat of formation, ⌬H00 of HOCl. So far, we
have not determined the branching ratio between the two
possible spin–orbit states of the chlorine fragment which are
separated by ϳ 882 cmϪ1. However, if we assume that the
highest energy rotational state of the OH fragment that is
detected, Nϭ11, correlates exclusively with the production
2
of chlorine atoms in their lowest spin–orbit state, the P3/2
state, then the linewidth associated with transitions appearing
J. Chem. Phys., Vol. 107, No. 9, 1 September 1997
130.239.20.174 On: Sun, 23 Nov 2014 03:48:15