A quasiclassical trajectory study of the OH+CO reaction

Abstract

We present a quasiclassical trajectory study of the OH+CO reaction using a potential surface that has been derived from ab initio calculations. Among quantities that have been studied are cross sections for reaction and for HOCO complex formation, cross sections associated with reaction from excited vibrational and rotational states, product energy partitioning and CO2 vibrational-state distributions, HOCO lifetime distributions, and thermal and state-resolved rate constants. We also present the results of Rice–Ramsberger–Kassel–Marcus (RRKM) calculations, using the same potential-energy surface, of HOCO lifetimes and of reactive and complex formation rate constants. The trajectory results indicate that the dominant mechanism for reaction involves complex formation at low energies. However, a direct reaction mechanism is responsible for half the reactive cross section at higher energies. This leads to a rate constant that is weakly temperature dependent at low temperatures, and becomes strongly temperature dependent at high temperature. Our trajectory results agree with measured rates over a wide range of temperatures, but the trajectory results at low temperatures are dominated by classical ‘‘leak’’ through zero-point barriers, so this agreement may be somewhat fortuitous. Rate constants for nonreactive processes such as OH(v=1) deactivation by CO that are controlled by the HOCO formation step are well above experiment (factor of 6), while rate constants for processes such as CO(v=1) reaction with OH that are controlled by decay of HOCO into H+CO2 are much closer (factor of 2). This suggests that the entrance channel barrier on our surface is too loose while the exit barrier is accurate. The error in the entrance channel barrier is studied using RRKM, and it is found to be due to an incorrect out-of-plane bend potential in the analytical surface used. Modifying the potential so that it is more consistent with ab initio calculations leads to greatly improved HOCO formation rates. The RRKM rate constant for HOCO formation is in good agreement with trajectories, but the RRKM reactive rate constant is well below either the trajectory value or experiment. This reactive rate is found to be very sensitive to tunneling through the exit channel barrier, so that by changing the barrier frequency from the fitted surface value to one which is more consistent with ab initio calculations, overall rates in agreement with experiment are obtained. Our trajectory cross section for highly translationally and rotationally excited OH is over an order of magnitude smaller than a value reported by Wolfrum. HOCO lifetimes are in the 0.4–2.0 ps range, about equal to RRKM theory near threshold but dropping more slowly with energy. Both trajectory and RRKM lifetimes are below the values measured by picosecond methods. Product energy partitioning is dominated by conversion of the exit channel barrier energy into product translation. Product rotational excitation is strongly controlled by angular momentum constraints. Product vibrational distributions are colder than statistical, with a propensity towards simultaneous bend–stretch excitation and not pure stretch excitation.