Residual Entropy of Ice: A Study Based on Transfer Matrices

TL;DR

This paper uses transfer matrix methods to analyze the residual entropy of ice Ih and ice Ic, providing new insights into their relationship and properties of related two-dimensional ice models.

Contribution

It introduces a transfer matrix approach for ice Ih and Ic, proving the residual entropy of ice Ih is not less than that of ice Ic, and explores properties of these matrices.

Findings

Residual entropy of ice Ih is at least that of ice Ic.

Transfer matrices M and M' reveal properties linked to residual entropy.

The work offers a new effective description for 2D ice models.

Abstract

Residual entropy of ice systems has long been a significant and intriguing issue in condensed matter physics and statistical mechanics. The exact solutions for the residual entropy of realistic three-dimensional ice systems remain unknown. In this study, we focus on two typical realistic ice systems, namely the hexagonal ice (ice Ih) and cubic ice (ice Ic). We present a transfer matrix description of the number of ice-ruled configurations for these two systems. First, a transfer matrix $M$ is constructed for ice Ic, where each element is the number of ice-ruled configurations of a hexagonal monolayer under certain condition. The product of $M$ and its transpose corresponds to a bilayer unit in ice Ih lattice, therefore is exactly a transfer matrix for ice Ih. Making use of this, we simply show that the residual entropy of ice Ih is not less than that of ice Ic in the thermodynamic limit, which was first proved by Onsager in 1960s. Furthermore, we find an alternative transfer matrix $M'$ for ice Ih, which is based on a monolayer periodic unit. Some interesting properties of $M$, $MM^T$ and $M'$ are illustrated, specifically the summation of all elements, the element in the first row and first column, and the trace. Each property is equivalent with the residual entropy of a two-dimensional ice model. Our work rediscovers the relationship between the residual entropies of ice Ih and ice Ic, and provides an effective description for various two-dimensional ice models.