We consider a one-dimensional lattice system of unbounded, real-valued spins with arbitrarystrong, quadratic, finite-range interaction. The first area of interest concerns the equivalence of the grand canonical ensemble and the canonical ensemble: on the level of thermodynamic functions, on the level of observables, and on the level of correlations. More precisely, in the thermodynamic limit (size N of the system goes to infinity), we show that the free energy, expectation of intensive observable, and correlation of two intensive functions are the same for the grand canonical ensemble and canonical ensemble. The second area of interest concerns the decay of correlations and uniqueness of infinite-volume Gibbs measure of the canonical ensemble. It is shown that the correlations of the canonical ensemble decay exponentially plus a volume correction term. As a consequence, we verify a conjecture that the infinite-volume Gibbs measure of the canonical ensemble is unique on the one-dimensional lattice, extending results that are known for the case of weak interaction. The third area of interest concerns the logarithmic Sobolev inequality (LSI). It is shown that the canonical ensemble satisfies a uniform LSI. The LSI constant is uniform in the boundary data, the external field and scales optimally in the system size. We deduce the LSI by combining two different methods, the two-scale approach and the Zegarlinski method. The last area of interest concerns the hydrodynamic limit. We deduce the hydrodynamic limit of Kawasaki dynamics. The main ingredients are uniform LSI and decay of correlations for the canonical ensemble. The proof is based on a method invented by Grunewald, Otto, Villani and Westdickenberg.

A self-contained, mathematical introduction to the driving ideas in equilibrium statistical mechanics, studying important models in detail.

Introduction to applications and techniques in non-equilibrium statistical mechanics of chaotic dynamics.

Stemming from the IHP trimester "Stochastic Dynamics Out of Equilibrium", this collection of contributions focuses on aspects of nonequilibrium dynamics and its ongoing developments. It is common practice in statistical mechanics to use models of large interacting assemblies governed by stochastic dynamics. In this context "equilibrium" is understood as stochastically (time) reversible dynamics with respect to a prescribed Gibbs measure. Nonequilibrium dynamics correspond on the other hand to irreversible evolutions, where fluxes appear in physical systems, and steady-state measures are unknown. The trimester, held at the Institut Henri Poincaré (IHP) in Paris from April to July 2017, comprised various events relating to three domains (i) transport in non-equilibrium statistical mechanics; (ii) the design of more efficient simulation methods; (iii) life sciences. It brought together physicists, mathematicians from many domains, computer scientists, as well as researchers working at the interface between biology, physics and mathematics. The present volume is indispensable reading for researchers and Ph.D. students working in such areas.

Equilibrium and Non-equilibrium Statistical Mechanics is a source-book of great value to college and university students embarking upon a serious reading of Statistical Mechanics, and is likely to be of interest to teachers of the subject as well. Written in a lucid style, the book builds up the subject from basics, and goes on to quite advanced and modern developments, giving an overview of the entire framework of statistical mechanics. The equilibrium ensembles of quantum and classical statistical mechanics are introduced at length, indicating their relation to equilibrium states of thermodynamic systems, and the applications of these ensembles in the case of the ideal gas are worked out, pointing out the relevance of the ideal gas in respect of a number of real-life systems. The application to interacting systems is then taken up by way of explaining the virial expansion of a dilute gas. The book then deals with a number of foundational questions relating to the existence of the thermodynamic limit and to the equivalence of the various equilibrium ensembles. The relevance of the thermodynamic limit in explaining phase transitions is indicated with reference to the Yang-Lee theory and the Kirkwood-Salsburg equations for correlation functions. The statistical mechanics of interacting systems is then taken up again, with reference to the 1D and 2D Ising model and to the spin glass model of disordered systems. Applications of the Mean field theory are worked out, explaining the Landau-Ginzburg theory, which is then followed by the renormalization group approach to phase transitions. Interacting systems in the quantum context are referred to, addressing separately the cases of interacting bosons and fermions. The case of the weakly interacting bosons is explained in details, while the Landau theory for fermi liquids is also explained in outline. The book then goes on to a modern but readable account of non-equilibrium statistical mechanics, explaining the link with irreversible thermodynamcs. After an exposition of the Boltzmann equations and the linear response theory illustrated with reference to the hydrodynamic model, it explains the statistical mechanics of reduced systems, in the context of a number of reduction schemes. This is followed by an account of the relevance of dynamical chaos in laying down the foundations of classical statistical mechanics, where the SRB distributon is introduced in the context of non-equilibrium steady states, with reference to which the principle of minimum entropy production is explaned. A number of basic fluctuation relations are then worked out, pointing out their relation to irreversible thermodynamics. Finally, the book explains the relevance of quantum chaos in addressing foundational issues in quantum statistical mechanics, beginning with Berry’s conjecture and then going on to an exposition of the eigenstate thermalization (ETH) hypothesis, indicating how the latter is relevant in explaining the processes of equilibriation and thermalization in thermodynamic systems and their sub-systems.

The problem of deriving irreversible thermodynamics from the re versible microscopic dynamics has been on the agenda of theoreti cal physics for a century and has produced more papers than can be digested by any single scientist. Why add to this too long list with yet another work? The goal is definitely not to give a gen eral review of previous work in this field. My ambition is rather to present an approach differing in some key aspects from the stan dard treatments, and to develop it as far as possible using rather simple mathematical tools (mainly inequalities of various kinds). However, in the course of this work I have used a large number of results and ideas from the existing literature, and the reference list contains contributions from many different lines of research. As a consequence the reader may find the arguments a bit difficult to follow without some previous exposure to this set of problems.

The structure of the theory ofthermodynamics has changed enormously since its inception in the middle of the nineteenth century. Shortly after Thomson and Clausius enunciated their versions of the Second Law, Clausius, Maxwell, and Boltzmann began actively pursuing the molecular basis of thermo dynamics, work that culminated in the Boltzmann equation and the theory of transport processes in dilute gases. Much later, Onsager undertook the elucidation of the symmetry oftransport coefficients and, thereby, established himself as the father of the theory of nonequilibrium thermodynamics. Com bining the statistical ideas of Gibbs and Langevin with the phenomenological transport equations, Onsager and others went on to develop a consistent statistical theory of irreversible processes. The power of that theory is in its ability to relate measurable quantities, such as transport coefficients and thermodynamic derivatives, to the results of experimental measurements. As powerful as that theory is, it is linear and limited in validity to a neighborhood of equilibrium. In recent years it has been possible to extend the statistical theory of nonequilibrium processes to include nonlinear effects. The modern theory, as expounded in this book, is applicable to a wide variety of systems both close to and far from equilibrium. The theory is based on the notion of elementary molecular processes, which manifest themselves as random changes in the extensive variables characterizing a system. The theory has a hierarchical character and, thus, can be applied at various levels of molecular detail.