Comparison of different water models from ambient to supercritical conditions: A Monte Carlo simulation and molecular Ornstein–Zernike study

Abstract

A molecular Ornstein-Zernike study of popular models for water and methanol Analysis of the hydrogen-bonded structure of water from ambient to supercritical conditions Gibbs ensemble Monte Carlo simulation of the properties of water with a fluctuating charges model Structural, thermodynamic, and dielectric properties of three polarizable and two nonpolarizable water models are compared with experimental data at four different thermodynamic states from ambient to supercritical conditions. Pair-correlation functions and thermodynamic data are obtained from Monte Carlo simulations, performed both on the (N,V,T) and ͑N,p,T͒ ensembles. The dielectric constants are determined with the molecular Ornstein-Zernike theory. It is found that the polarizable models can reproduce the experimental structure considerably better than the nonpolarizable ones at the high-temperature states. In particular, the elongation of the hydrogen bonds with increasing temperature, which is observed by neutron diffraction measurements, in only seen in the simulations with the polarizable potential models. On the other hand, the polarizable models fail to describe the correct temperature dependence of the thermodynamic properties. Although at ambient conditions they overestimate both the density and the dielectric constant of the system, around the critical temperature they result in 10%-50% lower densities than the experimental values. The obtained magnitude of the internal energies as well as the dielectric constants are also considerably smaller than their experimental values at these thermodynamic state points. The results of this study point out the need of new polarizable water models which, besides the reasonable reproduction of the experimental pair-correlation functions, are also able to describe the dependence of the thermodynamic properties on the temperature and pressure.

Charge-on-spring polarizable water models revisited: From water clusters to liquid water to ice

“…113 Gen- erally, polarizable water models overestimate the heat capacity. 114 The isothermal compressibility T is well reproduced while the thermal expansion coefficient ␣ is overestimated by our models.…”

Multi-particle sampling in Monte Carlo simulations on fluids: efficiency and extended implementations

“…These states were also considered by Předota et al [3,4] and they were therefore chosen for the sake of comparison. With respect to the complexity of the problem, both physical and computational, and the novelty of cutoff smoothing, in the following we will first demonstrate correctness of the proposed implementation of the MPM method by presenting the results for the thermodynamic and structural properties and their comparison with those reported in [3,4,18]. Then the problem of efficiency of the method and its assessment will be addressed.…”

“…Following Jedlovszky and Richardi [18], we will refer to these states as A and C, respectively. These states were also considered by Předota et al [3,4] and they were therefore chosen for the sake of comparison.…”

“…The number of iteration loops was about 12 for the MPM MC and 8 for the conventional MC. It is worth recalling that Jedlovszky and Richardi [18] used much softer criterion to terminate the iteration, m = 3, 11 In the NVT simulations we evaluated both the common thermodynamic properties, the internal energy and the dielectric constant, and the relevant structural properties, the site-site correlation functions g ij and the dipole-dipole correlation functions G l , l = 1, 2 [16]. The latter functions, called also 'local g factors', provide information on the mutual alignment of dipoles; moreover, they may also be used as an alternative route to the dielectric constant, see e.g.…”

“…It seems necessary first to recall that in both references, [4] and [18], the authors report error estimates only for the internal energy but not for other properties. It is also necessary to remark that the latter results seem to be subject of rather large errors.…”