Pressure-constant Monte Carlo simulation of phase I of solid CO2 up to 10 GPa at T = 200 K using Kihara potential model

TL;DR

This study extends Monte Carlo simulations of solid CO2 (phase I) under high pressure at 200 K using a modified NPT protocol with the Kihara potential, providing P-V relations up to 10 GPa.

Contribution

It introduces a modified NPT simulation protocol for solid CO2 to accurately determine lattice constants under pressure, addressing challenges due to molecular shape.

Findings

P-V relation consistent with previous theories

Molar volume ~0.5% smaller than experimental data

Modified protocol enables equilibrium lattice constant calculation

Abstract

This article is a continuation of the past three papers, arXiv:1711.04976 (2017), arXiv:1809.04291 (2018), and arXiv:2006.09673v2 (2020), in which configurations of the molecules around a vacancy in solid CO2 with the Pa3 structure (phase I) were calculated by the Monte Carlo (MC) simulation technique at T < 200 K and at a nominal pressure of P = 1 atm using lattice constants determined in reference to experimental data. For the intermolecular potential, the Kihara model of a rod-shape core with zero diameter was used. For theoretical consistency, however, the lattice constant should be determined by a pressure-constant MC simulation, i.e. by the NPT simulation. It was anticipated that the NPT simulation, successful for monoatomic molecular fluids, might not work straightforwardly for solid CO2 because of the non-spherical molecular shape and interactions. In fact, it was found in the present work that the protocol of the standard NPT simulation had to be modified for solid CO2 to achieve an equilibrium lattice constant, given an external pressure, within practical computer time and capacity. The modified protocol was used to calculate the P-V relation at 1 bar and from 2 to 10 GPa at which a structural phase transition is known to take place. The temperature was fixed at 200 K. As a result, the calculated P-V relation was similar to those in previous theoretical works, the molar volume being roughly 0.5 % smaller than experimental data.