Sensitivity of molecular dynamics unimolecular rate calculations to defects in the potential-energy surface

We report a detailed examination of the reasons why classical trajectory calculations of the rate of isomerisation of methyl isocyanide to methyl cyanide fail, by at least an order of magnitude, to match the observed results. We conclude that the internal motions in the CH,NC molecule are essentially chaotic, and the discrepancy stems from an inadequate degree of coupling between the vibrational motions on the assumed potential-energy surface and the reaction coordinate. Some consequences of these flaws are quite unacceptable: that the rate of decay is lower the greater the average potential energy; or that there is a severe departure from the random-gap law for the first few thousand wavenumbers above threshold. The difficulties impeding a full and satisfactory calculation of the rate of a unimolecular reaction of a polyatomic molecule by classical trajectory methods are outlined.