Multigrid-Based Methodology for Implicit Solvation Models in Periodic DFT.

Area of Science:

• Computational Chemistry
• Materials Science
• Quantum Mechanics

Background:

• Continuum solvation models are crucial for studying environmental effects in Density Functional Theory (DFT).
• Solving the Generalized Poisson Equation (GPE) is key for incorporating solvation contributions.
• Multigrid methods offer efficiency for solving partial differential equations but face memory challenges in DFT codes.

Purpose of the Study:

• To implement an accelerated multigrid solver-based approach for solvation effects within the Vienna ab initio Simulation Package (VASP).
• To develop an efficient and transferable implicit solvation model, VASP-MGCM (VASP-Multigrid Continuum Model).

Main Methods:

• Implementation of an accelerated multigrid solver for the GPE within VASP.
• Development of an efficient sparse matrix product algorithm.
• Application of the VASP-MGCM model to calculate solvation free energies.

Main Results:

• The VASP-MGCM model demonstrates high performance, outperforming serial multigrid solvers.
• Calculated solvation free energies for neutral and ionic species show good agreement with experimental data.
• Solvation energies for molecules adsorbed on metallic surfaces also align with experimental and other continuum model results.

Conclusions:

• The VASP-MGCM model provides an efficient and accurate treatment of solvation effects in DFT.
• This approach overcomes previous memory limitations associated with multigrid methods in DFT codes.
• VASP-MGCM offers a valuable tool for studying environmental influences on molecular and surface systems.