A supervised fitting approach to force field parametrization with application to the SIBFA polarizable force field

Area of Science:

• Computational Chemistry
• Molecular Modeling
• Physical Chemistry

Background:

• Force field parametrization is essential for accurate molecular simulations.
• Traditional methods often face stability issues due to nonlinear fitting problems.
• Polarizable force fields, like SIBFA, require robust parameterization strategies.

Purpose of the Study:

• To introduce a supervised, semiautomated approach for force field parameter fitting.
• To apply and validate this method for the SIBFA polarizable force field.
• To refine parameters for specific molecular fragments and their metal complexes.

Main Methods:

• Utilized the I-NoLLS interactive, nonlinear least squares fitting program.
• Employed a supervised, semiautomated fitting strategy to maintain physical parameter ranges.
• Generated reference data using high-level ab initio calculations with an aug-cc-pVTZ basis set.

Main Results:

• Successfully obtained refined parameters for H2O, formamide, and imidazole, including their Mg(2+) complexes.
• Demonstrated that interactive fitting mitigates stability problems common in force field parametrization.
• Achieved a more accurate SIBFA parametrization compared to previous efforts.

Conclusions:

• The described supervised, semiautomated fitting approach is effective for polarizable force fields.
• This method offers improved stability and accuracy in force field parameter development.
• Advances in computational hardware enable more precise ab initio calculations for better parametrization.