Metastability, Spectra, and Eigencurrents of the Lennard-Jones-38 Network

Abstract

We develop computational tools for spectral analysis of stochastic networks representing energy landscapes of atomic and molecular clusters. Physical meaning and some properties of eigenvalues, left and right eigenvectors, and eigencurrents are discussed. We propose an approach to compute a collection of eigenpairs and corresponding eigencurrents describing the most important relaxation processes taking place in the system on its way to the equilibrium. It is suitable for large and complex stochastic networks where pairwise transition rates, given by the Arrhenius law, vary by orders of magnitude. The proposed methodology is applied to the network representing the Lennard-Jones-38 cluster created by Wales's group. Its energy landscape has a double funnel structure with a deep and narrow face-centered cubic funnel and a shallower and wider icosahedral funnel. Contrary to the expectations, there is no appreciable spectral gap separating the eigenvalue corresponding to the escape from the icosahedral funnel. We provide a detailed description of the escape process from the icosahedral funnel using the eigencurrent and demonstrate a superexponential growth of the corresponding eigenvalue. The proposed spectral approach is compared to the methodology of the Transition Path Theory. Finally, we discuss whether the Lennard-Jones-38 cluster is metastable from the points of view of a mathematician and a chemical physicist, and make a connection with experimental works.

arXiv comment24 pages, 10 figures

References in corpus (10)

• Global Optimization by Basin-Hopping and the Lowest Energy Structures of Lennard-Jones Clusters Containing up to 110 Atoms
• The double-funnel energy landscape of the 38-atom Lennard-Jones cluster
• Phase changes in 38 atom Lennard-Jones clusters. I: A parallel tempering study in the canonical ensemble
• Thermodynamics and the Global Optimization of Lennard-Jones clusters
• Flows in Complex Networks: Theory, Algorithms, and Application to Lennard-Jones Cluster Rearrangement
• Simulating structural transitions by direct transition current sampling: the example of LJ38
• Metastability and low lying spectra in reversible Markov chains
• Six out of equilibrium lectures
• Canonically invariant formulation of Langevin and Fokker-Planck Equations
• Computing the Asymptotic Spectrum for Networks Representing Energy Landscapes using the Minimal Spanning Tree

Cited by in corpus (3)

• Jump Markov models and transition state theory: the Quasi-Stationary Distribution approach
• Modeling aggregation processes of Lennard-Jones particles via stochastic networks
• Testing Lennard-Jones Clusters for Optimality