In this dissertation, a new cluster-continuum solvation modeling method is presented and applied. The most important feature of this method is a novel approach for the generation of clusters, in which only important solvent molecules are added to the solute. This approach is based on the concept of functional group solvation spheres. Solvation sphere radii are defined by analyzing radial distribution functions of solvent molecules around solute functional groups, which enables the elimination of molecules unperturbed by the solute. Also, the presented method implements a procedure for simple, effective and unbiased generation of geometries with a selected number of solvent molecules. This is achieved by sorting the solvent molecules according to their interaction energy. The method was tested on two model systems, using water as the solvent. The patterns of solvent-solute bonding for neutral and charged molecules were identified by analysis of obtained geometries, while the calculated Boltzmann weighted Gibbs energies gave values of thermodynamic quantities with an average deviation of 1 kcal/mol when compared to experimental data. After that, the method was used for the investigation of acetyl-chloride hydrolysis reaction mechanisms.