How important is the dielectric constant in water modeling? Evaluation of the performance of the TIP4P/ ε \varepsilon force field and its compatibility with the Joung-Cheetham NaCl model
Łukasz Baran, Cosmin A. Dicu-Gohoreanu, Luis G. MacDowell · 2025
Read How important is the dielectric constant in water modeling? Evaluation of the performance of the TIP4P/ ε \varepsilon force field and its compatibility with the Joung-Cheetham NaCl model by Łukasz Baran, Cosmin A. Dicu-Gohoreanu, Luis G. MacDowell (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-00153.txt.
Introduction
Baran, Dicu-Gohoreanu, and MacDowell test how much one electric property of water matters when choosing a computer model of it, in a 2025 paper. Simulating watery solutions needs simple stand-in models that still get the physics right. A key property, the dielectric constant, says how water screens electric fields, and models differ in whether they are tuned to match it.
The team scored rival models on a broad benchmark of water and ice traits. The model tuned to the dielectric constant matched a leading rival overall and stayed accurate for that property across pressures and heats. It also predicted salt solution densities without extra fudging. A rival handled phase borders better, so the choice depends on the job.
Catalog record
| CREATOR | Łukasz Baran, Cosmin A. Dicu-Gohoreanu, Luis G. MacDowell |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 62 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-00153
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
How important is the dielectric constant in water modeling? Evaluation of the performance of the TIP4P/ ε \varepsilon force field and its compatibility with the Joung-Cheetham NaCl model Łukasz Baran, Cosmin A. Dicu-Gohoreanu, Luis G. MacDowell Abstract Efficient large scale computer simulations of complex aqueous solutions require the use of accurate but simple empirical force fields for water. However, the growing complexity of these systems is evidencing the difficulties to describe solution properties without due account of polarization. Conflicting strategies to remedy this problem are parametrizing water force fields to the dielectric constant or charge scaling of the solvated ions. In this work we compare results from TIP4P/ ε \varepsilon and OPC models, which are parametrized with an emphasis in the prediction of the dielectric constant, with results from TIP4P/2005, which is closer in spirit to the charge scaling strategy. The overall performance of the models is rated according to the Vega-Abascal benchmark. Our results show that TIP4P / ε /\varepsilon and TIP4P/2005 perform equally good, with the OPC model lying significantly behind. TIP4P/ ε \varepsilon can predict bulk phase properties (transport properties, thermal expansion coefficients, densities) of both liquid water and ice polymorphs, as well as surface tensions, with an accuracy very similar to TIP4P/2005, while performing very well for dielectric constants over a wide range of pressures and temperatures. On the other hand, TIP4P/2005 provides a better description of phase boundaries, including liquid-vapor and freezing transitions. However, the accurate prediction of dielectric constants allows TIP4P/ ε \varepsilon to describe solution densities for model ions targeted to their crystal and melt properties. This is achieved without the need to rescale charges, modify the Lorentz-Berthelot rule or tune the ion’s Lennard-Jones parameters. Our findings hinge on the significance of dielectric constants as a target property and cast some doubt on the need for the appealing concept of charge scaling. keywords electrolyte solutions † † affiliation: Department of Theoretical Chemistry, Institute of Chemical Sciences, Faculty of Chemistry, Maria Curie-Sklodowska University in Lublin, Lublin, Poland. † † affiliation: Departamento de Química Física, Facultad de Ciencias Químicas, Universidad Complutense, Madrid, 28040, Spain. † † email: [email protected] † † affiliation: Instituto de Química Física Blas Cabrera, CSIC , C/Serrano 119, 28006 Madrid, Spain † † affiliation: Departamento de Química Física, Facultad de Ciencias Químicas, Universidad Complutense, Madrid, 28040, Spain. † † email: [email protected] 1 Introduction Water is essential for the existence of life on Earth. A simple molecule, yet exhibiting numerous anomalies, such as a negative expansion coefficient below 4 ∘ C, a sharp increase of response functions upon cooling, a decrease of viscosity with the increase in pressure, and many others Gallo et al. 2016 . This unique behavior makes water a perfect testbed for the study of a wide range of fundamental physical principles in statistical mechanics. Therefore, great efforts have been devoted to the development of accurate and meaningful force fields ever since the advent of computer simulations Ouyang and Bettes 2015 . On physical grounds, it is clear that the design of an all-purpose model which can adapt depending on the environment requires dealing with the polarizability issue, if not exactly, at least approximately. Unfortunately, polarizable models such as the BK3 Kiss and Baranyai 2013 , HBP Jiang et al. 2016 , iAMOEBA Wang et al. 2013 give only a modest improvement, but their computational cost makes them inefficient for large system sizes. On the other hand, the very active line of research based on machine learned potentials can do no better than the underlying approximations Zhang et al. 2018 , and some recent studies for state of the art density functionals provide very disappointing results Montero de Hijes et al. 2023 ; Montero de Hijes et al. 2024 . Further improvement along this line requires a training set obtained from expensive ab-initio calculations dealing accurately with electronic correlations, but alas, neural networks appear to exhibit here significant limitations to model the ab-initio potential energy surface for a wide range of different chemical environments Zhai et al. 2023 . Due to these difficulties, model development is still often directed to the design of simple and tractable force fields, based on intuition and basic principles of chemistry. As a result, one often neglects polarizability altogether, and describes the average molecular polarity with a simple distribution of fixed point charges embedded into a spherical core. This is the case of the celebrated SPC model Berendsen et al. 1987 . Here, two positive partial charges sitting on hydrogen positions are neutralized by a negative charge located at the center of a simple Lennard-Jones site, which accounts for repulsion and dispersion of the oxygen atom. Alternatively, the negative charge can be shifted along the bisector axis of the HOH angle, as proposed by Bernal-Fowler already in 1933 Bernal and Fowler 1933 , leading to the so-called TIP4P family of water models. Jorgensen et al. 1983 Despite their simplicity, SPC and TIP4P models do a rather good job at describing the qualitative behavior of liquid water at ambient conditions, but the TIP4P family has been shown to be more robust, and remains qualitatively correct over a much wider range of thermodynamic conditions Sanz et al. 2004 ; Abascal et al. 2007 . Particularly, the TIP4P/2005 model provides a good qualitative description of the phase diagram, including the coexistence between several solid phases, as well as a quantitative description of the temperature of maximum density, the orthobaric liquid densities, molar volumes and surface tensions Abascal et al. 2005 ; Abascal et al. 2007 . Of course, neglecting polarizability comes at a price. For the particular case of the TIP4P/2005 model, the accuracy of liquid phase densities is at the cost of a rather poor description of the dielectric constant, which is about 30% too low compared to experiments Vega and Abascal 2011 . Alternatively, it often occurs that models providing a good description of the dielectric constant, such as SPC/E or TIP3P Berendsen et al. 1987 ; Jorgensen et al. 1983 , yield a qualitatively incorrect description of the phase diagram Vega and Abascal 2011 . A simple way out of this problem was suggested by Leontyev and Stuchebrukhov some time ago Leontyev and Stuchebrukhov 2009 ; Leontyev and Stuchebrukhov 2011 . These authors argued that the electric field generated by point charges on a molecule is screened by the electronic polarizability of the surrounding molecules. Therefore, the ”effective” charges, q_{\rm eff} that are required to reproduce an accurate force field are smaller than the ”true” point charges of the molecule, q_{\rm true} by a factor equal to the high frequency refractive index, n_{\infty} , i.e. q_{\rm eff}=\frac{q_{\rm true}}{n_{\infty}} . Consequently, models predicting a good force field will yield too low a molecular dipole moment when measured in terms of the effective charge. However, assuming the true charge is a factor n_{\infty} larger than the effective charge, and noting water’s refractive index is ca. n_{\infty}=1.33 , brings the dipole moment of usual force fields ( \approx 2.2-2.4 D) in close agreement with results of ab-initio calculations ( \approx 3.0 D). This appealing idea has received great attention and has become known as the scaled charge paradigm Kann and Skinner 2014 ; Kirby and Jungwirth 2019 ;