Abstract:The solvation parameter model uses six descriptors identified as excess molar refraction, E, dipolarity/polarizability, S, overall hydrogen-bond acidity, A, overall hydrogen-bond basicity, B, McGowan's characteristic volume, V, and the gas-liquid partition constant on hexadecane at 25 °C, L to model the distribution of neutral compounds in biphasic systems. Abraham's version of this model uses all six descriptors with two separate linear free energy relationship models for the transfer of compounds from a gas phase to a condensed phase and between condensed phases. Goss proposed a modification to this model that uses a single calibration model regardless of the physical state for each phase and five of the descriptors employed in Abraham's model (E descriptor is eliminated). The capability of Abraham's model and the Goss-modified model to characterize the contribution of intermolecular interaction to retention for gas and reversed-phase liquid chromatographic systems and distribution in liquid-liquid partition systems is evaluated using the WSU compound descriptor database. These more accurate values for the Abraham descriptors have not been utilized previously for the evaluation of the Goss-modified model and should be more capable of discerning subtle differences in model performance. It is shown that model quality defined by statistical parameters favors Abraham's model over the Goss-modified model with differences in model quality greater for systems in which Abraham's model indicates a significant contribution from electron lone pair interactions and for systems in which one phase is a solvent containing perfluoroalkyl substituents. There is a small systematic difference for the terms describing the combined contributions of cavity formation and dispersion interactions and for interactions of a dipole-type. The contribution of hydrogen-bonding interactions is virtually identical for the two models. The model intercepts are generally different and potentially assigned to a larger contribution from lack-of-fit for the Goss-modified model. Although the Abraham model descriptors have been routinely employed for applications using the Goss-modified model the possibility that Goss-model specific descriptors should be employed was evaluated. Using the Solver method and Goss-model specific calibration models for chromatographic and liquid-liquid partition systems a new set of Goss-specific descriptors was calculated for 28 varied compounds. These descriptors show good statistical agreement with the Abraham descriptor values with an average deviation of 0.009, -0.003, -0.004, and -0.023, respectively, for the S, A, B, and L descriptors, corresponding to a relative absolute deviation in percent of 2.2 %, 3.9 %, 4.3 %, and 1.2 %, respectively.

Estimation of Molecular Linear Free Energy Relation Descriptors Using a Group Contribution Approach

Estimation of Molecular Linear Free Energy Relationship Descriptors by a Group Contribution Approach. 2. Prediction of Partition Coefficients

Extended atom-based and bond-based group contribution descriptor and its application to melting point prediction of energetic compounds

Alchemical prediction of hydration free energies for SAMPL

Using molecular similarity to construct accurate semiempirical electron structure theories

Theoretical Evaluation of Radii of Atoms in Molecules and their Dependence on Atomic Partial Charge

Pairwise Additivity and Three-Body Contributions for Density Functional Theory-Based Protein-Ligand Interaction Energies

An Efficient Fragment-Based Approach for Predicting the Ground-State Energies and Structures of Large Molecules.

Application of a General Computer Algorithm Based on the Group-Additivity Method for the Calculation of Two Molecular Descriptors at Both Ends of Dilution: Liquid Viscosity and Activity Coefficient in Water at Infinite Dilution

ADME Evaluation in Drug Discovery. 4. Prediction of Aqueous Solubility Based on Atom Contribution Approach

From Intermolecular Interaction Energies and Observable Shifts to Component Contributions and Back Again: A Tale of Variational Energy Decomposition Analysis.

An adaptive distance-based group contribution method for thermodynamic property prediction.

Evaluation of the Goss-modified solvation parameter model for the characterization of biphasic systems and descriptor assignments

Probing non-covalent interactions with a second generation energy decomposition analysis using absolutely localized molecular orbitals.

Free Energy Methods for the Description of Molecular Processes

Molecular Dynamics Fingerprints (MDFP): Machine Learning from MD Data To Predict Free-Energy Differences

Physics-based Machine Learning to Predict Hydration Free Energies for Small Molecules with a minimal number of descriptors: Interpretable and Accurate

Efficient interatomic descriptors for accurate machine learning force fields of extended molecules

Prediction of organic homolytic bond dissociation enthalpies at near chemical accuracy with sub-second computational cost

Combining Molecular Dynamics and Machine Learning to Predict Self-Solvation Free Energies and Limiting Activity Coefficients

Automated relative binding free energy calculations from SMILES to ΔΔG