Tunneling of the Kawasaki dynamics at low temperatures in two dimensions (Q2261592)

The authors have introduced a new general method that allows the description of the asymptotic evolution of one-parameter families of continuous-time Markov chains. So far, the proposed method has been successfully applied in two situations: for zero-range dynamics on a finite set which exhibits condensation and for random walks evolving among random traps. This paper is a continuation of their so far done work [J. Stat. Phys. 140, No. 6, 1065--1114 (2010; Zbl 1223.60061); J. Stat. Phys. 149, No. 4, 598--618 (2012; Zbl 1260.82063)] and shows the derivation of some estimates needed in the proof of the convergence, in the zero-temperature limit, of the two-dimensional Kawasaki dynamics for the Ising model in a large cube to a Brownian motion. It also illustrates the interest of another method proposed in the literature by applying it in a simple context. Some steps were previously done in this direction by the derivation of the asymptotic behavior of continuous-time Markov chains evolving on a fixed and finite state space imposing only one simple condition on the jump rates. Now it is time to apply the so far obtained results to the Kawasaki dynamics for the Ising model on a fixed two-dimensional square with periodic boundary condition. The paper shows that for a fixed square and a fixed number of particles the asymptotic behavior of the dynamics among the ground states converges to a Brownian motion when the length of the square and the number of particles increase with the inverse of the temperature. There are several papers about theoretic approach to metastability in the framework of a fixed and finite state space Markov chain. In this paper, we have an approach where the formula of the jump probabilities written through eigenvectors of the generator (expressed only in terms of the capacities, which can be estimated using the Dirichlet and the Thomson variational principles) is replaced. The new approach allows to prove the convergence of the process by solving a martingale problem. The article is divided into seven sections. After the introduction, we have in Section 2 necessary notations and some important preliminaries with presentation of Theorem 2.1. Then, in Section 3, there is a sketch of the proof for this theorem. Section 4 shows the evolution of the Markov process \(\{\eta^\beta_t: t\geq0\}\) between two consecutive visits to the ground states \(\{\eta^x: x\in \Lambda_L\}\). The next section shows that at very low temperatures, in the time scale \(e^\beta\), much smaller than the time scale of an excursion between ground states, the process \(\eta_t\) evolves as a continuous-time Markov chain whose state space consists of subsets. In Sections 6 and 7, the proofs of the main results are presented.