A Local Variant of the Conductor-Like Screening Model for Fragment-Based Electronic-Structure Methods

Area of Science:

• Computational chemistry
• Quantum chemistry
• Molecular modeling

Background:

• Ab initio methods are increasingly applied to larger molecular systems.
• Subsystem methods divide large systems into smaller, manageable units.
• Implicit solvent models like conductor-like screening model (COSMO) incorporate environmental effects.

Purpose of the Study:

• To develop a computationally efficient COSMO variant for subsystem electronic structure calculations.
• To address the computational bottleneck posed by COSMO in large-scale simulations.
• To improve the treatment of solvent effects in biomolecular modeling using methods like frozen density embedding (FDE).

Main Methods:

• Implementation of an alternative COSMO variant within the Amsterdam Density Functional (Adf) program suite.
• Exploitation of the subsystem nature of the electronic description for computational savings.
• Comparison of different density and surface charge update schemes.
• Assessment of a single tuning parameter's effect on accuracy and performance.

Main Results:

• The developed COSMO variant dramatically reduces computational cost for solvent modeling.
• The accuracy of the traditional COSMO description is retained.
• The performance is evaluated across various update schemes and parameter choices.

Conclusions:

• The new COSMO variant offers a significant speed-up for subsystem calculations involving implicit solvent models.
• This approach is particularly beneficial for large biomolecular systems.
• The method provides an accurate and computationally feasible alternative for environmental effects in electronic structure calculations.