Atomic Decomposition of the Electrostatic Solvation Energy Using the Interacting Quantum Atoms Approach

  1. D. Suárez 1
  2. F. Jiménez-Grávalos 1
  3. N. Díaz 1
  4. E. Francisco 1
  5. A.M. Pendás 1
  1. 1 Universidad de Oviedo
    info

    Universidad de Oviedo

    Oviedo, España

    ROR https://ror.org/006gksa02

Proceedings:
11th Congress on Electronic Structure: Principles and Applications (ESPA 2018)

Year of publication: 2018

Pages: 156

Congress: Congress on Electronic Structure: Principles and Applications (ESPA 2018).17-19 July (11º. 2018. Toledo)

Type: Conference paper

Abstract

Basing on the first- and second-order density matrices, the interacting quantum atoms (IQA)method[1] decomposes the total energy of a molecular system in terms of one- and two-center(atomic) contributions within the context of the quantum theory of atoms in molecules. IQA is ableto process wavefunctions derived from several electronic-structure methods (HF, CCSD, CASSCF),but also admits DFT charge densities by defining ad hoc additive exchange-correlation energies.[2] IQA quantifies both intra- and intermolecular interactions, characterizing unambiguously manyaspects of chemical bonds and intermolecular forces. Here we incorporate electrostatic continuumsolvent effects into the IQA energy decomposition. To this end, the interaction between the soluteelectrostatic potential and the solvent screening charges as defined within the COSMO solvationmodel[3] is now included in a new version of the PROMOLDEN code, allowing thus to apply IQA incombination with COSMO-QM methods as well as to partition the electrostatic solvation energy intoeffective atomic contributions. To test the robustness of this approach, we carry out COSMO-HF/aug-cc-pVTZ calculations in aqueous solution followed by IQA calculations on hundreds of neutraland ionic solutes extracted from the MNSol database. The computational results reveal a detailedatomic mapping of electrostatic solvation energy that is useful to assess to what extent the solvationenergy can be decomposed into atomic and group contributions of various parts of a solute molecule,as generally assumed by empirical methodologies that estimate solvation energy and/or logP valuesfrom multilinear regression models.