Variable atomic radii for continuum-solvent electrostatics calculation.

Area of Science:

• Computational chemistry
• Molecular modeling
• Physical chemistry

Background:

• Continuum-solvent models are crucial for electrostatics calculations.
• Existing models often use fixed atomic radii, limiting accuracy.
• The interlocking-sphere model relies on accurate solute cavity definitions.

Purpose of the Study:

• To develop an improved method for defining solute cavities in electrostatics calculations.
• To create a variable-radii model where atomic sizes adapt to their chemical environment.
• To enhance the accuracy of solvation energy predictions.

Main Methods:

• Approximating molecular electron density using Gaussian functions.
• Optimizing Gaussian function parameters against ab initio quantum calculations.
• Determining atomic radii based on electron density cutoffs specific to each element.
• Validating the model against experimental solvation energies for 31 small neutral molecules.

Main Results:

• The variable-radii model demonstrated improved solvation energy predictions compared to fixed-radii models (e.g., Bondi radii).
• The model showed particular benefit for quantum mechanics/Poisson-Boltzmann calculations.
• Significant improvements were observed for molecules possessing large dipole moments.

Conclusions:

• The developed method offers a more chemically realistic representation of atoms in molecular simulations.
• This approach enhances the predictive power of continuum-solvent electrostatics calculations.
• The variable-radii model provides a flexible and accurate alternative for defining solute cavities.