Phase diagram and critical properties of a two-dimensional associating lattice gas

TL;DR

This study thoroughly investigates the phase diagram of a two-dimensional associating lattice gas model, revealing a critical line and discontinuous phase transitions, and compares findings with exact solutions, clarifying its water-like anomaly limitations.

Contribution

It provides a comprehensive analysis of the phase behavior of the symmetric ALG model using transfer matrix and Monte Carlo methods, clarifying the nature of phase transitions and critical points.

Findings

Identification of a critical line ending at a critical point.

Discontinuous transition between disordered fluid and low-density liquid.

Phase diagram similar to Husimi lattice solution, but limited in water-like behavior.

Abstract

We revisit the associating lattice gas~(ALG) introduced by Henriques \textit{et al.} [PRE 71, 031504 (2005)] in its symmetric version. In this model, defined on the triangular lattice, interaction between molecules occupying nearest-neighbor sites depends on their relative orientation, mimicking the formation of hydrogen bonds in network-forming fluids. Although all previous studies of this model agree that it has a disordered fluid (DF), a low-density liquid (LDL) and a high-density liquid (HDL) phase, quite different forms have been reported for its phase diagram. Here, we present a thorough investigation of its phase behavior using both transfer matrix calculations and Monte Carlo (MC) simulations, along with finite-size scaling extrapolations. Results in striking agreement are found using these methods. The critical point associated with the DF-HDL transition at full occupancy, dentified by Furlan and coworkers [Phys.~Rev.~E~{\bf 100}, 022109 (2019)] is shown to be one terminus of a {\it critical line} separating these phases. In opposition to previous simulation studies, we find that the transition between the DF and LDL phases is always discontinuous, similar to the LDL-HDL transition. The associated coexistence lines meet at the point where the DF-HDL critical line ends, making it critical-end-point. Overall, the form of the phase diagram observed in our simulations is very similar to that found in the exact solution of the model on a Husimi lattice. Our results confirm that, despite the existence of some waterlike anomalies in this model, it is unable to reproduce key features of the phase behavior of liquid water.