Efficient calculation of low energy statistical rates for gas phase dissociation using umbrella sampling.

Monte Carlo (MC) simulations can be used to compute microcanonical statistical rates of gas phase dissociation reactions. Unfortunately, the MC approach may suffer from a slow convergence and large statistical errors for energies just above the dissociation threshold. In this work, umbrella sampling is proposed as a device to reduce the statistical error of MC rate constants. The method is tested by computing the classical dissociation rate for the reaction [H5O2+]* --> H2O + H3O(+) over the range of internal energy 38 < E < or = 100 kcal/mol. Comparing with other literature methods, it is found that umbrella sampling reduces the computational effort by up to two orders of magnitude when used in conjunction with a careful choice of sampling distributions. The comparison between MC rate constants and classical Rice-Ramsperberg-Kassel harmonic theory shows that anharmonicity plays an important role in the dissociation process of the Zundel cation (H5O2+) at all energies.