One-dimensional Ising models with long range interactions: cluster expansion, phase-separating point (Q2450872)

Consider a one-dimensional ferromagnetic Ising model on \(\mathbb Z\) with long-range interactions \(J(n) = n^{-2+\alpha}\) between spins of different signs, where \(n\) is the distance between the spins, and \(\alpha\) is a nonnegative parameter. For any configuration of spins on \(\mathbb Z\) taking values \(-1\) to the left of \(-L\) and \(+1\) to the right of \(+L\) (inhomogeneous boundary condition), there is an algorithm to uniquely identify a so-called phase-separating point of the configuration, which is a vertex on the shifted lattice \((1/2) + \mathbb Z\) located between \(-L-(1/2)\) and \(L+(1/2)\). The first result of the paper states that for \(\alpha = 0\), if the temperature is sufficiently low, the phase-separating point rescaled by \(1/L\) converges in distribution to a random variable with an explicit non-degenerate (not uniform) density on the interval \([-1,1]\). In particular, the phase-separating point exhibits macroscopic fluctuations. The second result states that for \(0<\alpha<(\log 3)/(\log 2) - 1\), if the temperature is sufficiently low, the phase-separating point rescaled by \(1/L^{1-(\alpha/2)}\) converges in distribution to a centered Gaussian random variable. In particular, the fluctuations of the phase-separating point are mesoscopic around zero. The proofs are based on a cluster expansion. The upper bound constraint on \(\alpha\) in the second statement is a technical condition. It is expected that the result remains true for all \(\alpha>0\).