Condensation of non-reversible zero-range processes (Q1731126)

The problem of the metastability phenomenon occurs in many different models related to probability theory and statistical physics. Among them there are: the system of sticky particles, low-temperature ferromagnetic spin systems, stochastic partial differential equations, random perturbations of dynamical systems. Thanks to the works of Bovier, Eckhoff, Gayrard, and Klein (BEGK) we know that the potential theoretic notions, such as the capacity and the equilibrium potential can be connected with metastable behavior of the system and the quantities such as hitting probability and transition time. BEGK framework is called the potential theoretic approach and has many important theoretical and practical applications. Usually, the successive metastable transitions are described as a limiting Markov chain. Beltran and Landim proposed a framework known as the martingale approach. The investigations of metastability of non-reversible dynamics based on the potential theoretic approach or the martingale approach are complicated and challenging. This is caused by two reasons: (i) the estimation of the capacity between metastable valleys requires an analysis of the flow structure dynamics; (ii) in the non-reversible cases the estimation of the mean jump rate between metastable valleys is very difficult. \par The author focuses on the condensation phenomena for a class of nonreversible zero-range processes on a fixed finite set. The author is able to prove the Markovian behavior of the condensed site of the corresponding zero-range processes. He gives a generalized version of the Dirichlet and Thomson principles, which can be applied in the asymptotic analysis of capacity, and establishes a general method to deduce the mean jump rate from the capacity estimate. \par The whole paper is divided into 8 sections. After the introduction (Section 1), we have the description of zero-range processes in Section 2, and the presentation of the main results in Section 3. Section 4 shows the adjoint dynamics and proves the sector condition. Section 5 reviews the flow structure and refers to the generalized Dirichlet-Thomson principle. The following section presents the general framework for the quantitative analysis of metastability of non-reversible processes. In Sections 7 and 8, the author gives approximating objects playing a central role in the proof presented in Section 6.