Field-Extremum Model for Short-Range Contributions to Hydration Free Energy

Abstract

The performance in describing hydration free energies of a broad class of neutral, cationic, and anionic solutes is tested for the recently proposed FESR (Field-Extremum Short-Range) implicit solvation model for interactions between the solute and nearby water molecules, as taken in conjunction with the previously developed SS(V)PE (Surface and Simulation of Volume Polarization for Electrostatics) dielectric continuum model for long-range interactions with bulk water. The empirical FESR model mainly describes solute-water hydrogen bonding interactions by correlating them with the maximum and minimum values of the electric field produced by the solute at the surface of the cavity that excludes solvent. A preliminary report showed that, with only four adjustable parameters, the FESR model, in conjunction with SS(V)PE, can produce hydration energies comparable to the best analogous efforts in the literature that utilized many more parameters. Here, the performance of the FESR model is more fully documented in several respects. The dependence on the underlying quantum mechanical method used to treat the internal electronic structure of the solute is tested by comparing uncorrelated Hartree-Fock to correlated density functional calculations and by comparing a modest sized to a large basis set. The influence of cavity size is studied in connection with an isodensity contour construction of the cavity. The sensitivity of the results to the parameters in the FESR model is considered, and it is found that the dependence on the electric field strength is quite nonlinear, with an optimum exponent consistently in the range of 3 to 4. Overall, it is concluded that the FESR model shows considerable utility for improving the accuracy of implicit models of aqueous solvation.

Self-consistent continuum solvation (SCCS): The case of charged systems

“…The performance of the SCCS model can now be discussed in comparison with other continuum solvation models. In particular the IEF-PCM model, implemented in GAUSSIAN 03 and GAUSSIAN 09, the SMx family, the SS(V)PE model, as well as its combination with the FESR correction 41,42 are considered here. Results from SMx, 5,6 SS(V)PE, 42 and SS(V)PE+FESR 42 represent a very good point of comparison, since they are based on a closely similar database.…”

“…Values for SS(V)PE and SS(V)PE+FESR are from Ref. 42 and correspond to 51 cations and 60 anions from the Minnesota Solvation Database. Values for PCM / g03, PCM / g09, and SCCS come from this study and correspond to a reduced data set of 51 cations and 55 anions, but without any explicit water molecules.…”

Advances in molecular quantum chemistry contained in the Q-Chem 4 program package

“…Unfortunately, the analytic energy gradient for such a construction has yet to be developed, and at present, carefully parameterised, bond-connectivitydependent atomic radii can surpass the accuracy of an isodensity cavity when it comes to computing solvation free energies [280]. (Recent attempts to incorporate nonelectrostatic terms into the isodensity construction show great promise for high accuracy with minimal parameterisation [281,282], but these corrections are not yet available in Q-CHEM. )…”

The charge-asymmetric nonlocally determined local-electric (CANDLE) solvation model

“…We consider this far too restrictive as it precludes applications to solutes that combine sites with different charges. Recently, Pomogaeva and coworkers 18 introduce a correction for hydration free energies that depend on the electric field at the surface of the cavity to account for asymmetries due to hydrogen bonding. However, their approach is based on sharp atomic-sphere cavities and is not self-consistent.…”