Implementation and Optimization of the Embedded Cluster Reference Interaction Site Model with Atomic Charges

Area of Science:

• Computational Chemistry
• Physical Chemistry
• Theoretical Chemistry

Background:

• The accurate prediction of solvation free energies is crucial for understanding chemical processes.
• The embedded cluster reference interaction site model (EC-RISM) combines quantum mechanics with integral equation theories.
• Previous implementations of EC-RISM required extensive parameterization and computational resources.

Purpose of the Study:

• To implement and optimize the EC-RISM method for calculating aqueous solvation free energies.
• To investigate various computational settings to determine optimal performance.
• To assess the accuracy of the optimized EC-RISM implementation against established databases.

Main Methods:

• Implementation of the EC-RISM method, integrating quantum mechanical calculations with 3D-reference interaction site model (3D-RISM).
• Systematic investigation of computational parameters including buffer, grid space, basis set, charge model, water model, and closure relation.
• Neglect of small point charges from the 3D-RISM solvent distribution to reduce computational overhead without compromising accuracy.

Main Results:

• The optimized EC-RISM method achieved root-mean-square deviations (RMSDs) of 5.70, 6.32, and 6.44 kJ/mol on the MNSOL[a], MNSOL, and FreeSolv databases, respectively.
• Alternative settings yielded RMSDs of 5.22, 6.08, and 6.63 kJ/mol.
• The performance was comparable to or better than other methods, except for those with fitting parameters like COSMO-RS.

Conclusions:

• The implemented and optimized EC-RISM method provides accurate solvation free energies in water.
• Neglecting small point charges is a viable strategy to reduce computational cost.
• The optimized EC-RISM method presents a competitive approach for solvation free energy calculations.