Premium
A metropolis Monte Carlo method for computing microcanonical statistical rate constants
Author(s) -
Dumont Randall S.
Publication year - 1991
Publication title -
journal of computational chemistry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.907
H-Index - 188
eISSN - 1096-987X
pISSN - 0192-8651
DOI - 10.1002/jcc.540120313
Subject(s) - monte carlo method , inefficiency , statistical physics , microcanonical ensemble , dynamic monte carlo method , computer science , inversion (geology) , quasi monte carlo method , monte carlo molecular modeling , metropolis–hastings algorithm , hybrid monte carlo , algorithm , markov chain monte carlo , mathematics , physics , statistics , canonical ensemble , paleontology , structural basin , economics , biology , microeconomics
Abstract A new method of computing microcanonical statistical rate constants is presented. The method utilizes the Metropolis Monte Carlo algorithm in a manner which circumvents some of the numerical inefficiency associated with other Metropolis and “shot‐gun” Monte Carlo based schemes. It is therefore expected to be useful in studies of many degree of freedom systems where numerical efficiency is crucial. Optimization of the method efficiency with respect to its adjustable parameters is examined in detail, both theoretically and in a numerical study of the T‐shaped Ar 3 “inversion” process. The energy dependence of the T‐shaped Ar 3 inversion rate is studied in a sample application of the method. An application to full three dimensional Ar 3 will be presented in a future study.