Accuracy of Quantum Mechanically Derived Force-Fields Parameterized from Dispersion-Corrected DFT Data: The Benzene...

Area of Science:

• Computational Chemistry
• Quantum Mechanics
• Materials Science

Background:

• Accurate description of intermolecular interactions is crucial for molecular simulations.
• Standard force fields often struggle with precise predictions, especially for non-bonded interactions.
• Density Functional Theory (DFT) functionals are widely used but their performance varies.

Purpose of the Study:

• To assess the reliability of popular dispersion-corrected DFT functionals in describing two-body interaction potential energy surfaces (IPESs).
• To develop a robust protocol for deriving force field parameters from quantum mechanical data.
• To evaluate the performance of DFT functionals for aromatic interactions using the benzene dimer as a benchmark.

Main Methods:

• Utilized the automated Picky procedure for iterative parametrization of intermolecular force fields against quantum mechanical data.
• Employed classical Monte Carlo and Molecular Dynamics simulations with derived force fields on systems up to 1000 molecules.
• Calculated condensed phase properties and compared structural and dynamic results with experimental data.

Main Results:

• CAM-B3LYP-D3 functional showed the best performance in describing the benzene dimer's IPES.
• The multilevel approach successfully generated force field parameters from quantum mechanical data.
• Assessed the quality of DFT functionals based on their ability to reproduce experimental condensed phase properties.

Conclusions:

• CAM-B3LYP-D3 is a reliable DFT functional for describing aromatic interactions.
• The developed methodology offers a robust protocol for deriving accurate force field parameters from QM data.
• This approach is valuable for systems where standard force fields yield inaccurate predictions.