Particle density probability distribution function and center symmetry breaking in finite density lattice gauge theories

TL;DR

This paper investigates phase transitions in finite density lattice gauge theories using probability distribution functions, addressing symmetry-related issues and proposing a method to avoid the sign problem, validated through numerical simulations.

Contribution

It introduces a novel approach to compute probability distribution functions at finite density by leveraging center symmetry, overcoming zero-value problems and the sign problem.

Findings

Probability distribution functions can be computed at finite density.

The method effectively avoids the sign problem in simulations.

Numerical results confirm the approach's validity in U(1) gauge theory.

Abstract

We study the nature of the phase transition at high temperature and high density in lattice gauge theories by focusing on the probability distribution function, which represents the probability that a certain density will be realized in a heat bath. The probability distribution function is obtained by constructing a canonical partition function by fixing the number of particles from the grand partition function. However, if the Z3 center symmetry, which is important for understanding the finite temperature phase transition of SU(3) lattice gauge theory, is maintained on a finite lattice, the probability distribution function is always zero, except when the number of particles is a multiple of 3. For U(1) gauge theory, this problem is more extreme. The probability distribution becomes zero when the particle number is not zero. In this study, we find a solution to this problem and propose a method of avoiding the sign problem, which is an important problem at finite density, using the center symmetry. This problem is essentially the same as the problem that the expectation value of the Polyakov loop is always zero when calculating with finite volume. In the case of U(1) lattice gauge theory with heavy fermions, numerical simulations are actually performed, and we demonstrate that the probability distribution function at a finite density can be calculated by the method proposed in this study.