Conclusions

In this article we report quantum scattering calculations on the
reactive system H + O_{2}
OH + O for total angular
momentum *J*=0, 1, 2, and 5. To facilitate our calculations we
performed them on a parallel computer using the Coriolis coupled method
of Goldfield and Gray[49].

From the shape of the *J*=0 reaction probability we infer that the
reaction proceeds through two
different reaction mechanisms. In the first reaction mechanism, at low
energies (*E* < 1.25 eV), the reaction probably goes through an
intermediate HO_{2}-complex. This gives rise to very sharp
resonances. These resonances take a long time to resolve (
150 000 a.u.). At higher energies (*E* > 1.25 eV), a second non-complex
forming (direct) reaction path opens up, leading to enhanced reaction
probability. The broader resonances in this energy range are caused by
the reactants that do form a complex. This is consistent with earlier
findings in the
literature[16,25,26,27,32,33,34].
The exact reason for the broadening of the resonances can not be
obtained from the present work.

Comparing the *J*=0 to the *J*=1 results we see a big decrease in the
reaction probability at higher energies. Examining the *J*=1 reaction
probabilities, we note that the difference is primarily caused by the
reaction probabilities for which the wave function starts in the
or in the
substate. The difference is
tentatively explained by realizing that when the wave function starts
substate, the complex is initially in a ``planar''
configuration, in which the O_{2} molecule rotates in the plane of the 3
atoms. On the other hand, when the wave function starts in the
or
substate, the complex is initially in a
``perpendicular'' configuration (see Sec. IVB). In the
first configuration the reaction probability is enhanced, because it is
possible for the H atom to ``hit'' one of the O atoms directly, opening
up the ``direct'' pathway. This explanation is consistent with the
findings of Bronikowski *et al.*[16]. The
differences between the
calculation and the
calculation we tentatively explain by centrifugal
barrier effects[85] (i.e., change in the *J*(*J*+1) term
in the Hamiltonian).

Going from *J*=1 to *J*=2 to *J*=5, the above mentioned reaction mechanism
and explanations for the reactivity of the wave packet in different initial
states are still valid. This leads us to believe that there might be a way to
devise an approximate method to calculate cross sections based on only a
small number of calculations with *J*>0. Furthermore, we see that the
reaction probability is strongly dependent on the initial state of the wave
packet. This means that ,
which is not a good quantum number, does
not scramble very quickly. However, looking at the
distributions
for *J*=5, it is clear that no
states can be left out of the
calculation. This might change, when we choose a different embedding for the
calculations, i.e., when we take the O_{2} bond to be the BF
*z*-axis[89] or when we perform the calculation in product
coordinates. However, it is not clear beforehand, whether this will make a
difference and the fact that the resonance features for higher *J* are less
pronounced than at *J*=0 suggests that it may not. We are currently
investigating this issue[84].

Currently, we are performing calculations for *J*>5 to check whether
the models and explanations given in this article still hold at higher
*J*. Ultimately, we want to use our calculations to get cross sections
for this reaction. Furthermore, we want to devise an accurate
approximate method for performing calculations for *J*>0 based on only
a few *J*>0 calculations[84].

The authors wish to thank Dr. R. T Pack for sending us his results and for useful discussions. We wish to thank Dr. J. Z. H. Zhang for useful discussions. Further, we wish to thank Dr. G. C. Groenenboom for critically reading the manuscript and useful comments. Partial support for this research comes from NSF grant CHE-9526658. We also acknowledge generous grants of computer time from the Cornell Theory Center, the Argonne High Performance Computing Research Facility and the High Performance Computing Facility at Wayne State University.