Interactions of the b-cyclodextrin (b-CD) ligand with Na ? , Cu ? , Mg 2? , Zn 2? , and Al 3? cations were investigated using density functional theory modeling. The objective of this study was to give insight into the mechanism of cation complexation. Two groups of conformers were found. The first group preserved the initial orientation of glucopyranose residues inside the b-CD ligand. The mutual orientation of glucopyranose residues was strongly affected by the cation in the second group of conformers. The system interaction energy was decomposed into electrostatic (ES), Pauli and orbital contributions using the Ziegler-Rauk energy partitioning scheme. The total electrostatic energy, i.e., the sum of ES energy and polarization energy, is the dominating term in the interaction energy. In vacuum, the complexes formed with Al 3? were found to be more stable than with di-and monocations. The vacuum stability sequence was changed in aqueous solution.

Charge sensitivity analysis (CSA) was extended to AMBER force-field resolution. The effective electronegativity and hardness data were found using evolutionary algorithms. Four model hardness matrices based on the classical electrostatic, Mataga-Nishimoto, Ohno, and Louwen-Vogt interpolation formulae were considered. Mulliken population analysis and electrostatically derived charges (CHELPG) were taken into account. It was demonstrated that the Ohno interpolation formula gives the best fit to Mulliken charges. For all molecules from the training set and all model hardness matrices, Mulliken charges were reproduced more accurately than CHELPG charges, indicating their good transferability from system to system. The effective electronegativities and hardnesses obtained were further verified by applying CSA to molecules from a validation set that was different from the training set. The correlation between CSA and Mulliken charges was of the same quality as that obtained for the training set.

A novel, neural network controlled, dynamic evolutionary algorithm is proposed for the purposes of molecular geometry optimization. The approach is tested for selected model molecules and some molecular systems of importance in biochemistry. The new algorithm is shown to compare favorably with the standard, statically parametrized memetic algorithm.