Quantification and localization of molecular hydrophobicity

Hydrophobic effects play an essential role for a variety of chemical processes. These effects fall into two categories: on one hand the poor solubility of nonpolar compounds which is termed hydrophobic hydration and is quantified by the free energy of hydration. On the other hand association of nonpolar molecules or molecular fragments in an aqueous environment is termed hydrophobic interaction and the free energy of transfer (2-phase system, 1-Octanol/water) is used to quantify this distribution process. In this work the molecular-free-energy-density concept is presented at first. It is based on the representation of the free energy of solvation as a energy density on the solvent accessible surface. In addition to the physical basis of this concept, a model for this energy density is presented which allows the localization of hydrophobic regions on the molecular surface. The energy density is chosen in such a way that integration of surface fragments yields a partial free energy of transfer. For a series of halogen substituted sucrose derivatives an excellent correlation was established between the relative sweetness of this series of compounds and their related partial free energy of transfer. The main focus of this work is the development of a new model for the free energy surface density to arrive at a physically sound localization of molecular hydrophobicity. For the old model only global information, for instance experimentally determined values of the free energy of transfer, was considered. The three-dimensional free energy density (3D-FED) offered to account for local, thermodynamic information during the parameterization of the new model. The interaction part of this 3D-FED was approximated by an appropriate molecular interaction field which was generated by the program Grid. Such interaction fields were then expanded in terms of surface-based functions. Furthermore, the parameters of the model were fitted to local and global reference data simultaneously. Therefore, the model not only renders excellent predictive power for both the free energy of hydration and transfer, it also yields a plausible localization of hydrophobic regions on the molecular surface. The statistical significance of the correlation mentioned above (halogen substituted sucrose derivatives) was further improved