Exponential rate of convergence in current reservoirs (Q2515519)

The authors of this article study an interacting particle system that serves as model for Fick's first law of diffusion, which proportionates the current to the gradient of the concentration. The studied particle system is a continuous-time Markov process with state space \(\{0, 1\}^{[-N, N]}\). The dynamics is described by the sum of the generators of a nearest-neighbor symmetric exclusion process and a certain birth-and-death process. In the latter process, births and deaths occur in a neighborhood of \(-N\) and \(N\), respectively. The main result of the article is that the exponential rate of convergence to the stationary measure is of order \(N^2\). To solve this out-of-equilibrium problem, the authors ``relied on stochastic inequalities and coupling methods, thus reducing the problem to that of bounding the extinction time of the set of discrepancies between two coupled evolutions.''