Quenched non-equilibrium central limit theorem for a tagged particle in the exclusion process with bond disorder (Q731670)

This paper deals with a classical problem in statistical mechanics : Proving that the dynamical of a single particle in a mechanical system is well approximated on a large scale by a Brownian motion. In the spirit of the seminal work of Kipnis and Varadhan, who proved an invariance principle in the simple symmetric exclusion process (extended to interacting particle systems whose generators satisfy a gradient condition), the authors adapt a recent quenched central limit theorem for random conductance model (from Sidoravicius and Sznitman) to one dimensional exclusion processes evolving in an environment, by adding a rule telling that a jump is suppressed whenever a particle decides to jump over a site already occupied. They first prove (under a strict ellipticity and mild moments conditions) a quenched non-equilibrium central limit theorem for the density field. Together with a similar result for the current, they get a central limit theorem for the position of a tagged particle starting from a configuration in which particles are distributed according to a Bernoulli product measure associated to a smooth density profile, for almost all environment. Most of their strategy is to use a functional transformation of the empirical measure to control this non-gradient model via a gradient associated one.