The PHAST 2.0 Force Field for General Small Molecule and Materials Simulations
Adam Hogan1, Logan Ritter1, Brian Space1
1Department of Chemistry, North Carolina State University, Raleigh, North Carolina 27695, United States.
Abstract:
Classical, empirical molecular simulation has become increasingly important in chemistry due to its ability to accurately model and resolve experimental phenomena on the atomic scale. Still, many challenges remain including obtaining subkilojoules per mole accuracy while maintaining speed, computational efficiency, and transferability to novel and heterogeneous chemistries. Further, a distinct lack of systematic progress on force fields over the last decades is shown to be due to a lack of a comprehensive, systematic approach rather than inherent deficiencies in rationally chosen potential energy surfaces. Indeed, to achieve these goals, it has become necessary to move beyond the highly approximate mean field Lennard-Jones equation, which is the backbone of many modern general purpose force fields. Lennard-Jones based force fields are both state point dependent and overfit to a large number of atom types to try and reproduce bulk thermophysical benchmarks; this fitting comes at the cost of physical grounding and transferability. Here, a new general purpose force field, PHAST 2.0, is presented where the parameters are fit solely to electronic structure data; PHAST is validated on the same kinds of experimental bulk data with which current Lennard-Jones based force fields are parametrized. PHAST has the emergent property of generality and transferability that is aided by including explicit many-body polarization. It is also designed to work with many-body dispersion models in a modular fashion. Being constructed from the fundamental interactions allows for force field elaboration for specific molecules or materials of interest without reparametrization. Several atom typing schemes, as well as an implicit polarization version, are explored. PHAST 2.0 has accuracy in line with the best common general purpose force fields used today with very minimal atom typing and thermodynamic state point independence, maintaining reduced complexity and enhanced transferability. It is argued this approach leads straightforwardly to bespoke, one-off PHAST force fields for a chemistry of interest, where additional training is warranted for additional accuracy.
More Related Videos
05:37Rapid in-silico Battery Electrolyte Electrochemical Reaction Generation using 3T-VASP Multi-Scale Energy Minimization
Published on: August 22, 2025
10:52Multiscale Sampling of a Heterogeneous Water/Metal Catalyst Interface using Density Functional Theory and Force-Field Molecular Dynamics
Published on: April 12, 2019
Related Concept Videos
Two-Dimensional Force System
Three-Dimensional Force System
