Large Deviations for the SSEP with slow boundary: the non-critical case

TL;DR

This paper establishes a large deviations principle for the empirical measure of the symmetric simple exclusion process with slow boundary reservoirs, analyzing subcritical and supercritical boundary interaction regimes and their impact on hydrodynamic equations.

Contribution

It extends large deviations analysis to the SSEP with slow boundary reservoirs, covering new regimes and deriving corresponding rate functions.

Findings

Rate function matches previous results for subcritical regime

Supercritical regime shows mass conservation with super-exponentially small boundary exchange probability

Hydrodynamic equations differ between subcritical and supercritical regimes

Abstract

We prove a large deviations principle for the empirical measure of the one dimensional symmetric simple exclusion process in contact with reservoirs. The dynamics of the reservoirs is slowed down with respect to the dynamics of the system, that is, the rate at which the system exchanges particles with the boundary reservoirs is of order $n^{-\theta}$, where $n$ is number of sites in the system, $\theta$ is a non negative parameter, and the system is taken in the diffusive time scaling. Two regimes are studied here, the subcritical $\theta\in(0,1)$ whose hydrodynamic equation is the heat equation with Dirichlet boundary conditions and the supercritical $\theta\in(1,+\infty)$ whose hydrodynamic equation is the heat equation with Neumann boundary conditions. In the subcritical case $\theta\in(0,1)$, the rate function that we obtain matches the rate function corresponding to the case $\theta=0$ which was derived on previous works (see \cite{blm,flm}), but the challenges we faced here are much trickier. In the supercritical case $\theta\in(1,+\infty)$, the rate function is equal to infinity outside the set of trajectories which preserve the total mass, meaning that, despite the discrete system exchanges particles with the reservoirs, this phenomena has super-exponentially small probability in the diffusive scaling limit.