2678
J. Chem. Phys., Vol. 116, No. 7, 15 February 2002
Kwon et al.
calculation indeed predicts fast conversion to INT1 ͑with its
barrier height of only 3.7 kcal/mol͒ and to some aldehyde
intermediates undergoing unfavorable high-energy processes,
respectively.20
ciation, as evidently manifested in the comparison to the
prior analysis. The fast and random decomposition behavior
of the short-lived, energy-rich intermediate is also quite con-
sistent with the equally populated ⌳-doublet and spin-orbit
components. Due to the poor fit in the statistical population
analysis the relative contribution of each mechanism is not
determined except for the fact that the reaction takes place
mainly through a direct mechanism, rather than a complex-
forming process, as clearly shown in Fig. 2. Similar bimodal
distributions have rather been observed in the exothermic
For the O(3P)-allyl system, therefore, it is highly prob-
able that the counterpart fragment C3H4 of the observed OH
product is believed to be allene through the intermediate
INT1 after taking into account the factors of reaction en-
thalpy, barrier height and the number of intermediates in-
volved. It is worthwhile obtaining the energy-specific rate
constants for the unimolecular decomposition pathways of
the energized INT1 using the RRKM model. At our available
energy of 82 kcal/mol, the rate constants for acrolein, allene,
methylacetylene and cyclopropene channels are 8.1ϫ1012,
1.2ϫ1010, 1.5ϫ106, and 2.4ϫ105 sϪ1, respectively, which
reveals the branching ratio and further supports the major
acroleinϩH channel and the alleneϩOH channel. The pre-
diction is also quite consistent with recent theoretical calcu-
lations for the reaction of O(3P) with halogenated alkyl
radicals,21 in which the energy-rich association intermediates
similar to INT1 in our system were formed and the major
decomposition product was determined by the aforemen-
tioned factors.
O(1D)-saturated hydrocarbon reactions in which the low-N
Љ
regime with low vibrational excitation ͑less than 0.1 for
P1 /P0͒ is well characterized by the Boltzmann-type distri-
bution assuming the long-lived alcohol intermediate, while
the high-N regime is described in terms of the short-lived
Љ
insertion process.23–28 Such vibrationally cold and statistical
low-N components stand in sharp contrast to our non-
Љ
statistical low-N distribution with high vibrational excita-
Љ
tion (P1 /P0ϭ0.75Ϯ0.11). However, the minor contribu-
tion, if any, from a long-lived complex embedded in the low-
N regime cannot be ruled out in this work.
Љ
In summary the nascent rovibrational distributions of the
OH product from the newly observed exothermic reaction of
O(3P)ϩC3H5→C3H4ϩOH were studied. To the best of our
knowledge the atom-radical reaction dynamics was first
probed through the combination of crossed beams and LIF
techniques. The distributions showed significant bimodal in-
ternal excitations without spin-orbit and ⌳-doublet propensi-
ties. With the aid of ab initio and statistical calculations the
reactive scattering may be explained by two competing non-
statistical dynamical mechanisms: a major direct abstraction
process and a minor indirect short-lived complex-forming
addition process.
To better characterize the available energy partitioning
into rovibrational states, the observed bimodal distributions
are compared to those predicted by the prior statistical
theory. For polyatomic systems the prior distributions are
determined by the available volumes in the phase space of
the two fragments, and are constrained only by energy con-
servation. The prior calculations predict much higher rota-
tional temperatures ͑4800 K (F1) and 4710 K (F2) for
v
Љ
ϭ0, and 2920 K (F1) and 2840 K (F2) for ϭ1͒ and a
v
Љ
smaller fraction (P1 /P0ϭ0.33) of population partitioned
into ϭ1 compared to the observed experimental distribu-
v
Љ
One of the authors ͑J.H.C.͒ wishes to thank Professor H.
Reisler for her invaluable help in the development of the
crossed beam apparatus, and the MOE and KBSI programs.
J.H.P. and H.L. wish to acknowledge the support of BK21
fellowships. This work was financially supported by the
CRM-KOSEF ͑2001͒.
tions. The surprisal analysis is also performed. However, the
rotational surprisals for the low- and high-N components fit
Љ
very poorly except for the large negative slope in the low-N
Љ
regime indicating that the low rotational components are
highly populated in comparison to the expected statistical
distribution. The disagreement suggests that the statistical
picture is not suitable to describe the reactive atom-radical
processes and that the nascent internal distributions with bi-
modal features show clear dynamical biases.
The observed bimodal internal distributions are quite un-
usual in cases of reactions of ground-state atomic oxygen.
The common reaction mechanism of O(3P) with saturated
hydrocarbon has been known to proceed through collinear
direct abstraction pathways resulting in the vibrationally hot
and rotationally cold distributions of the OH product.22
1 D. W. Kohn, H. Clauberg, and P. Chen, Rev. Sci. Instrum. 63, 4003
͑1992͒.
2 I. Fischer, Chimia 54, 96 ͑2000͒.
3 H. J. Deyerl, I. Fischer, and P. Chen, J. Chem. Phys. 110, 1450 ͑1999͒ and
references therein.
4 D. Stranges, M. Stemmler, X. Yang, J. D. Chesko, A. G. Suits, and Y. T.
Lee, J. Chem. Phys. 109, 5372 ͑1998͒ and references therein.
5 M. Yamaguchi, J. Mol. Struct. THEOCHEM 365, 143 ͑1996͒.
6 I. R. Slagle, J. R. Bernhardt, D. Gutman, M. A. Hanning-Lee, and M. J.
Pilling, J. Phys. Chem. 94, 3652 ͑1990͒.
Highly populated low-N components in Fig. 2 might be
7 T. Baer and W. L. Hase, Unimolecular Reactions Dynamics ͑Oxford Uni-
versity Press, New York, 1996͒.
Љ
dominated by such direct abstraction dynamics yielding the
observed high vibrational partitioning ratios and low rota-
tional temperatures. In contrast the extraordinarily hot rota-
tional distribution observed in this work implies that some
fraction of the reactants is sampled to proceed through an
addition complex such as INT1 in the potential energy sur-
face. However, such an intermediate is not likely to be long-
lived enough to allow the available energy to be statistically
partitioned among all the degrees of freedom prior to disso-
8 K. A. Holbrook, M. J. Pilling, and S. H. Robertson, Unimolecular Reac-
tions, ͑Wiley, New York, 1996͒.
9 H. K. Kim, H. C. Kwon, J. H. Park, Y. S. Choi, and J. H. Choi, Bull.
Korean Chem. Soc. 20, 1441 ͑1999͒; J. H. Park, H. Lee, H. C. Kwon, H.
K. Kim, Y. S. Choi, and J. H. Choi ͑unpublished͒.
10 R. Zhang, W. J. Zande, M. J. Broikowski, and R. N. Zare, J. Chem. Phys.
94, 2704 ͑1991͒; J. Han, X. Chen, and B. R. Weiner, Chem. Phys. Lett.
332, 243 ͑2000͒.
11 Y. Hurwitz, P. S. Stern, R. Naaman, and A. B. McCoy, J. Chem. Phys. 106,
2627 ͑1997͒.
130.70.241.163 On: Sat, 20 Dec 2014 01:24:35