Polarizable Embedding without Artificial Boundary Polarization.

Area of Science:

• Computational chemistry
• Molecular modeling
• Quantum mechanics

Background:

• Polarizable embedding (PE) models are crucial for simulating molecular systems.
• Common PE models can suffer from unphysical boundary polarization, affecting accuracy.
• Cutoff models like the droplet model have known limitations in PE calculations.

Purpose of the Study:

• To introduce a fully self-consistent polarizable embedding (PE) model that eliminates unphysical boundary polarization.
• To investigate the limitations of existing cutoff models in PE calculations.
• To provide a more physically accurate description of polarization in molecular systems.

Main Methods:

• Developed a novel PE model incorporating the minimum-image convention (MIC) for induced electrostatics.
• Applied the PE with MIC (PE-MIC) model to study solute-solvent and biomolecular systems.
• Compared PE-MIC results with traditional cutoff models, analyzing induced dipoles, electrostatic potentials, and optical properties.

Main Results:

• The PE-MIC model effectively prevents unphysical boundary polarization.
• Demonstrated that inaccuracies in cutoff models depend on system properties, cutoff method, and distance.
• Quantified the impact of unphysical polarization on electrostatic and optical properties.

Conclusions:

• The PE-MIC model offers a significant improvement for accurate molecular simulations.
• Highlights the critical need to address boundary artifacts in PE calculations.
• The findings guide the selection of appropriate models and parameters for reliable computational studies.