Cargando…

Correcting Rate Constants from Anharmonic Molecular Dynamics for Quantum Effects

[Image: see text] Anharmonicity can greatly affect rate constants. One or even several orders of magnitude of deviation are found for obtaining rate constants using the standard rigid-rotor harmonic-oscillator model. In turn, reactive molecular dynamics (MD) simulations are a powerful way to explore...

Descripción completa

Detalles Bibliográficos
Autores principales: Schmalz, Felix, Kopp, Wassja A., Kröger, Leif C., Leonhard, Kai
Formato: Online Artículo Texto
Lenguaje:English
Publicado: American Chemical Society 2020
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7016917/
https://www.ncbi.nlm.nih.gov/pubmed/32064385
http://dx.doi.org/10.1021/acsomega.9b03383
Descripción
Sumario:[Image: see text] Anharmonicity can greatly affect rate constants. One or even several orders of magnitude of deviation are found for obtaining rate constants using the standard rigid-rotor harmonic-oscillator model. In turn, reactive molecular dynamics (MD) simulations are a powerful way to explore chemical reaction networks and calculate rate constants from the fully anharmonic potential energy surface. However, the classical nature of the dynamics and the required numerical efficiency of the force field limit the accuracy of the resulting kinetics. We combine the best of both worlds by presenting an approximation that pairs anharmonic information intrinsic to classical MD with high-accuracy energies and frequencies from quantum-mechanical electronic structure calculations. The proposed scheme is applied to hydrogen abstractions in the methane system, which allows for the benchmarking of rate constants corrected by our approach against experimental rate constants. This comparison reveals a standard deviation of factor 2.6. Two archetypes of possible failure are identified in the course of a detailed investigation of the CH(3)(•) + H(•) → CH(2)(2•) + H(2) reaction. From this follows the application range of the method, within which the method shows a standard deviation of factor 2.1. The computational efficiency and beneficial scaling of the method allow for application to larger systems, as shown for hydrogen abstraction from 2-butanone by HO(2)(•).