Hamiltonian dynamics, nanosystems, and nonequilibrium statistical mechanics

TL;DR

This paper reviews recent advances in nonequilibrium statistical mechanics using Hamiltonian dynamical systems theory, emphasizing relaxation, transport, and entropy production, with applications to nanosystems and connections to fluctuation theorems.

Contribution

It introduces large-deviation dynamical relationships linking microscopic chaos properties to macroscopic transport and entropy production, extending fluctuation theorem concepts.

Findings

Relaxation rates derived from Liouville's equation.

Transport coefficients related to Pollicott-Ruelle resonances.

Fractal and singular properties of nonequilibrium states.

Abstract

An overview is given of recent advances in nonequilibrium statistical mechanics on the basis of the theory of Hamiltonian dynamical systems and in the perspective provided by the nanosciences. It is shown how the properties of relaxation toward a state of equilibrium can be derived from Liouville's equation for Hamiltonian dynamical systems. The relaxation rates can be conceived in terms of the so-called Pollicott-Ruelle resonances. In spatially extended systems, the transport coefficients can also be obtained from the Pollicott-Ruelle resonances. The Liouvillian eigenstates associated with these resonances are in general singular and present fractal properties. The singular character of the nonequilibrium states is shown to be at the origin of the positive entropy production of nonequilibrium thermodynamics. Furthermore, large-deviation dynamical relationships are obtained which relate the transport properties to the characteristic quantities of the microscopic dynamics such as the Lyapunov exponents, the Kolmogorov-Sinai entropy per unit time, and the fractal dimensions. We show that these large-deviation dynamical relationships belong to the same family of formulas as the fluctuation theorem, as well as a new formula relating the entropy production to the difference between an entropy per unit time of Kolmogorov-Sinai type and a time-reversed entropy per unit time. The connections to the nonequilibrium work theorem and the transient fluctuation theorem are also discussed. Applications to nanosystems are described.