A general intermolecular force field based on tight-binding quantum chemical calculations.

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Molecular Modeling

Background:

• Accurate prediction of intermolecular interactions is crucial for understanding molecular behavior.
• Existing methods often struggle with diverse chemical structures or require extensive parameterization.
• Developing efficient and accurate potential energy functions remains a significant challenge.

Purpose of the Study:

• To present a novel black-box procedure for generating molecule-specific intermolecular potential energy functions.
• To enable the treatment of non-covalently bound complexes and aggregates with arbitrary chemical structures.
• To provide a versatile computational tool for molecular aggregation and alignment studies.

Main Methods:

• Utilizes quantum chemical (QC) information from the extended tight-binding semi-empirical scheme (GFN-xTB).
• Incorporates density-dependent Pauli repulsion, penetration, electrostatics, D4 dispersion, Drude oscillators for polarization, and charge-transfer terms.
• Requires minimal empirical parameters (one element-specific, ~20 global) for elements up to radon (Z=86).

Main Results:

• The method accurately predicts intermolecular energies and equilibrium distances for standard benchmark sets.
• Demonstrates versatility with examples of charged systems and structures containing hundreds of atoms.
• The generated potential energy functions show high accuracy for various molecular interactions.

Conclusions:

• The developed black-box procedure offers an accurate and versatile approach for generating molecule-specific intermolecular potentials.
• The method, implemented in a stand-alone code, facilitates efficient global minimum energy searches for molecular aggregation.
• This advancement provides a valuable tool for computational studies involving complex molecular systems.