APT a next generation QM-based reactive force field model

Author： Rappé A. K. Bormann-Rochotte L. M. Wiser D. C. Hart J. R. Pietsch M. A. Casewit C. J. Skiff W. M.

ISSN： 1362-3028

Source： Molecular Physics, Vol.105, Iss.2-3, 2007-01, pp. : 301-324

Disclaimer: Any content in publications that violate the sovereignty, the constitution or regulations of the PRC is not accepted or approved by CNPIEC.

Previous Menu Next

Abstract

Modelling reactivity at the nanoscale is a major computational challenge. Both reactive force field and combined QM-MM methodologies have been and are being developed to study reactivity at this boundary between molecules and the solid state. There have been more than 1500 publications since the mid-1990s, on combined QM-MM methodologies. Limitations in current models include the distortional characteristics of force field potential terms, the smooth transit from one potential surface to another, rather than surface hopping, and the blending of electrostatics between QM and MM portions of a QM-MM model. Functional forms, potential surface coupling terms, and parameterization strategies for the Approximate Pair Theory (APT), a next generation reactive force field model, are described. The APT model has been developed to correct a number of limitations in current reactive force field models as well as providing a foundation for a next generation QM-MM model. Chemical bonding concepts are used to develop fully dissociative bond stretch, bend, torsion, and inversion valence terms. Quantum mechanics also provides functional forms for potential surface coupling terms that permit a general description of reactivity from hydrogen bonding, through non-classical carbocations and cracking, to olefin polymerization, oxidation, and metathesis. Van der Waals, electrostatic, and metallic bonding models also derive from quantum mechanical resonance. Finally, Pauli Principle-based orthogonality provides a way to electrostatically couple the QM and MM portions of a QM-MM model that will support arbitrarily large basis sets.