Thermodynamic transformations of nonequilibrium states (Q1938801)

Normally in equilibrium statistical mechanics, it is assumed that, in order to obtain the typical value of macroscopic observables and their fluctuations, it is not necessary to solve any equation of motion and the calculations can be performed using the Gibbs distribution. The nonequilibrium states are imagined as the stationary states of systems in contact with different reservoirs and/or under the action of external (electric) fields. Contrary to equilibrium states, there are currents (electrical, heat, mass, etc.) through the system whose macroscopic behavior is encoded in transport coefficients like the diffusion coefficient, the conductivity or the mobility. In this case, one can not bypass an analysis of the dynamical properties of the system: the Gibbs distribution (its calculation is a most challenging task) is replaced by the invariant distribution for the microscopic dynamics. This approach has some disadvantages and the authors of this paper want to reach a nonequilibrium stationary state by constructing analogues of thermodynamic potentials to extract the typical macroscopic behavior of the system as well as the asymptotic probability of the fluctuations. This can be implemented without the explicit knowledge of the stationary ensemble and is known as macroscopic fluctuation theory based on an extension of Einstein equilibrium fluctuation theory. The authors develop this theory, here. An open system in contact with boundary reservoirs and/or under the action of an external field is considered the situation in which the reservoirs and field vary with time driving the system from a state to another one is analyzed. A thermodynamic principle asserts that reversible transformations are accomplished by a sequence of equilibrium states and are well approximated by quasistatic transformations in which the variations of the environment are very slow. The authors show that this principle can be derived from the assumption that during the transformation the system is in a local equilibrium state. They consider as well a transformation between nonequilibrium stationary states; define a renormalized work performed by a given transformation that is obtained by subtracting from the total energy exchanged the energy needed to maintain the (nonequilibrium) stationary state; prove that the renormalized work satisfies a Clausius inequality; and show that also for nonequilibrium states quasi-static transformations are optimal, in the sense that they minimize the renormalized work. The paper is divided into seven sections. In Section 2, there is a formulation of the basic assumptions for the thermodynamic description of driven diffusive systems. Section 3 presents a discussion about the case of equilibrium states and how the Clausius inequality can be obtained. Section 4 presents the case of nonequilibrium states and their transformations. Section 5 considers a system with a general time dependent forcing. The last two sections exemplify the theory discussed in this paper by considering, respectively, the case of stochastic lattice gases, giving also a microscopic definition of work, and Langevin dynamics.