A coupled states distorted wave study of the O(3P)+H2 (D2, HD, DH) reaction

Abstract

In this paper, the coupled states distorted wave (CSDW) method is used to study the quantum reactive collision dynamics of O(3P)+H2 → OH+H and its D2, HD, and DH counterparts. The potential surface used is the sum of a LEPS potential (due to Johnson and Winter) and a correction factor which raises the barrier for H+OH → HO+H exchange to a realistic value. Full basis set convergence of the CSDW transition probabilities is established at low energies where tunneling dominates the dynamics, which means that the calculated cross sections should be exact except for errors introduced by the CS approximation, and the latter are expected to be less than 30%. The results presented for all four isotopes include: reaction probabilities as a function of energy E and total angular momentum J, total and state to state integral cross sections (including an analysis of product state distributions), and thermal and state resolved rate constants. Comparison of the results with those of several previous dynamical calculations on the same or similar surfaces is made and the accuracy of the approximations made in those calculations is assessed. For example, the product rotational distributions predicted by vibrationally adiabatic distorted wave theory are found to be quite close to what we calculate, although the absolute magnitudes of the cross sections are quite different. Comparison with the results of quasiclassical trajectory calculations indicates good agreement of the reactive cross sections well above the classical threshold, but not of the rate constants (because of tunneling) or of isotope ratios. Wigner corrected conventional transition state theory is very inaccurate in predicting rate constants, but a method which uses collinear exact quantum (CEQ) transmission coefficients to correct transition state theory does quite well. Variational transition state theory estimates of the rate constants and isotope ratios are also quite good, with the CSDW results generally bracketed by results obtained using the least action ground state (LAG) and small curvature ground state (SCTSAG) tunneling approximations. Comparison with experimental rate constants and isotope ratios is studied, and we find that the CSDW results are just outside the experimental error bars in all cases. The present results on the corrected Johnson and Winter surface are found to be slightly less accurate than the best estimates of rate constants and isotope ratios obtained using the modified POLCI surface of Walch and co-workers, suggesting that the latter surface is more accurate.