Publications Repository - Gdańsk University of Technology

The importance of incorporating solvent polarization effects into the modeling of solvation processes has been well-recognized, and therefore a new generation of hybrid quantum mechanics/molecular mechanics (QM/MM) approaches that accounts for this effect is desirable. We present a fully self-consistent, mutually polarizable QM/MM scheme using the AMOEBA force field, in which the total energy of the system is variationally minimized with respect to both the QM electronic density and the MM induced dipoles. This QM/AMOEBA model is implemented through the Q-Chem/LibEFP code interface and then applied to the evaluation of solute-solvent interaction energies for various systems ranging from the smallest water dimer to neutral and ionic solutes (NH3-, NH4+, CN-) surrounded by increasing numbers (up to 100) of water molecules. In order to analyze the obtained interaction energies, we also utilize an energy decomposition analysis (EDA) scheme which identifies contributions from permanent electrostatics, polarization and van der Waals (vdW) interaction for the interaction between the QM solute and the solvent molecules described by AMOEBA. This facilitates a component-wise comparison against the full QM references where the corresponding energy components are obtained via a modified version of the absolutely localized molecular orbitals (ALMO)-EDA. The results show that the present QM/AMOEBA model can yield reasonable solute-solvent interaction energies for neutral and cationic species, which, however, highly relies on the delicate balance between insufficiently favorable permanent electrostatics and softened vdW interaction. For anionic solutes where the charge penetration effect becomes more pronounced, the QM/MM interface turns out to be unbalanced. These results are consistent with and further elucidate our findings in a previous study using a slightly different QM/AMOEBA model (J. Chem. Phys. 145, 124106 (2016)), and the implications on future improvement of this model are also discussed.

Authors

• Yuezhi Mao,
• Yihan Shao,
• dr hab. inż. Jacek Dziedzic link open in new tab ,
• Chris-Kriton Skylaris,
• Teresa Head-Gordon,
• Martin Head-Gordon