Aging and intermittency in a p-spin model of a glass

TL;DR

This paper investigates the aging and intermittent heat flow in a non-random Ising spin model, revealing energy bursts, domain-based fluctuations, and temperature-dependent dissipation, providing insights into glassy dynamics.

Contribution

It introduces a numerical analysis of heat flow intermittency in a non-quenched spin model, linking bursts to record energy fluctuations and domain structures, advancing understanding of glassy aging.

Findings

Energy leaves the system in intermittent bursts called quakes.

The rate of energy dissipation decreases inversely with age.

Heat flow has temperature-independent and Arrhenius components.

Abstract

We numerically analyze the statistics of the heat flow between an aging system and its thermal bath, following a method proposed and tested for a spin-glass model in a recent Letter (P. Sibani and H.J. Jensen, Europhys. Lett.69, 563 (2005)). The present system, which lacks quenched randomness, consists of Ising spins located on a cubic lattice, with each plaquette contributing to the total energy the product of the four spins located at its corners. Similarly to our previous findings, energy leaves the system in rare but large, so called intermittent, bursts which are embedded in reversible and equilibrium-like fluctuations of zero average. The intermittent bursts, or quakes, dissipate the excess energy trapped in the initial state at a rate which falls off with the inverse of the age. This strongly heterogeneous dynamical picture is explained using the idea that quakes are triggered by energy fluctuations of record size, which occur independently within a number of thermalized domains. From the temperature dependence of the width of the reversible heat fluctuations we surmise that these domains have an exponential density of states. Finally, we show that the heat flow consists of a temperature independent term and a term with an Arrhenius temperature dependence. Microscopic dynamical and structural information can thus be extracted from numerical intermittency data. This type of analysis seems now within the reach of time resolved micro-calorimetry techniques.