Assimilating Radial Distribution Functions To Build Water Models with Improved Structural Properties

Area of Science:

• Computational chemistry
• Molecular modeling
• Physical chemistry

Background:

• Accurate molecular simulations require precise force fields.
• Existing water models like TIP3P and TIP4P/2005 have limitations in predicting structural properties.
• Radial distribution functions (RDFs) are key descriptors of molecular structure.

Purpose of the Study:

• To develop an improved parametrization methodology for water models.
• To enhance the accuracy of structural property predictions for water.
• To investigate the performance of Buckingham potentials in water models.

Main Methods:

• Extended the ForceBalance code to target any radial distribution function (RDF).
• Implemented a new objective function based on the mean squared difference (MSD) between experimental and simulated RDFs.
• Replaced Lennard-Jones potentials with Buckingham potentials in TIP3P and TIP4P/2005 models.

Main Results:

• Achieved 93% and 98% lower MSDs in the OO RDF for TIP3P-Buckingham and TIP4P-Buckingham models, respectively.
• Significantly reduced errors in predicting water's entropy, from 11% to 3% (TIP3P) and 2% (TIP4P/2005).
• Demonstrated improved predictive power for non-fitted properties, especially for TIP3P.

Conclusions:

• The Buckingham potential offers a superior description of water's structural properties compared to the Lennard-Jones potential.
• The new RDF-targeting methodology provides systematic force field optimization.
• Buckingham potentials are computationally efficient alternatives to adding interaction sites for improving water models.