On thermalization of a system with discrete phase space

TL;DR

This paper studies how a discrete phase space system thermalizes when changing from an initial to a final temperature, revealing that relaxation times depend on energy gaps and differ between cooling and heating, with results confirmed in a 1D Ising model.

Contribution

It demonstrates the non-universal relation between cooling and heating time scales and extends analysis to complex systems like the 1D Ising model.

Findings

Cooling can be faster than heating for large energy gaps.

Relaxation times depend on the eigenvalues of the stochastic matrix.

Both Kullback-Leibler divergence and Fisher information reveal similar thermalization dynamics.

Abstract

We investigate the thermalization of a stochastic system with discrete phase space, initially at equilibrium at temperature $T_i$ and then termalizing in an environment at temperature $T_f$ , considering both cases $T_i > T_f$ and $T_i < T_f$. For the simple case of a system with constant energy gaps, we show that the relation between the time scales of the cooling and heating processes is not univocal, and depends on the magnitude of the energy gap itself. Specifically the eigenvalues of the corresponding stochastic matrix set the time scales of the relaxation process and for large energy gaps the cooling process is found to exhibit the shortest relaxation times to equilibrium while the heating process is found to be faster at all scales for small gaps. We consider both the Kullback-Leibler divergence and the Fisher information and its related quantities to quantify the degree of thermalization of the system. In the intermediate to long time regime both quantities are found to bear the same type of information concerning the rate of thermalization, and follow the ordering predicted by the dynamic eigenvalues. We then consider a more complex system with a more intricate stochastic matrix, namely a 1D Ising model, and confirm the findings on the existence of two regimes, one in which cooling becomes faster than heating. We make contact with a previous work where an harmonic oscillator was used as working fluid and the heating process was always found to be faster than the cooling one.