Login / Signup

Rate constants from instanton theory via a microcanonical approach.

Sean R McConnellAndreas LöhleJohannes Karwounopoulos
Published in: The Journal of chemical physics (2018)
Microcanonical instanton theory offers the promise of providing rate constants for chemical reactions including quantum tunneling of atoms over the whole temperature range. We discuss different rate expressions, which require the calculation of stability parameters of the instantons. The traditional way of obtaining these stability parameters is shown to be numerically unstable in practical applications. We provide three alternative algorithms to obtain such stability parameters for non-separable systems, i.e., systems in which the vibrational modes perpendicular to the instanton path couple to movement along the path. We show the applicability of our algorithms on two molecular systems: H2 + OH → H2O + H using a fitted potential energy surface and HNCO + H → NH2CO using a potential obtained on-the-fly from density functional calculations.
Keyphrases
  • machine learning
  • density functional theory
  • molecular dynamics
  • molecular dynamics simulations
  • deep learning
  • big data
  • risk assessment
  • artificial intelligence