Modeling of configurational transitions in atomic systems
R.S. Berrya,
B.M. Smirnovb aDepartment of Chemistry, University of Chicago, 5735 South Ellis Ave., Chicago, Illinois, 60637, USA bJoint Institute for High Temperatures, Russian Academy of Sciences, ul. Izhorskaya 13/19, Moscow, 127412, Russian Federation
Configurational transitions in atomic systems, i.e. transitions that change the system’s geometric structure, include chemical reactions in gases, transitions between aggregate states of a polyatomic system, i.e. the phase transitions, and nanocatalytic processes. These transitions are analyzed from the standpoint of the behavior of the system on its effective Potential Energy Surface (PES), so that the transition results from passage between different local minima of the PES. It is shown that the density functional theory (DFT) is suitable in principle for the analysis of complex atomic systems, but based on contemporary computer codes, this method is not suitable even for simple atomic systems, such as heavy atoms or metal clusters. Next, a statistical determination of energetic parameters of atomic systems does not allow to analyze the dynamics of configuration transitions; in particular, the activation energy of a chemical process differs significantly from the height of a barrier which separates the atomic configurations of the initial and final states of the transition. In particular, the statistical models, including DFT, give a melting point for clusters with a pairwise atomic interaction that is twice that from dynamic models which account for thermal motion of atoms. Hence the optimal description of configurational transitions for atomic systems may be based on joining the DFT methods for determination the PES of this system with molecular dynamics, to account for thermal motion of atoms.