The landscape of a potential energy surface is marked by chemically interesting features. Hills and valleys correspond to transition states and reactive intermediates; the deepest valley gives the most stable configuration. Mapping these features for individual molecules and for the interactions between molecules is one of the goals of computational chemistry.

The dispersion energy is a weak attractive force in intermolecular interactions. Dispersion energy results from a purely quantum mechanical effect, in which instantaneous multipoles on one molecule induce multipoles on another. Among neutral atoms or molecules that lack permanent multipole moments, the dispersion interaction is the principal attractive force. Dispersion also plays a significant role in the interaction between molecules with diffuse π clouds. This interaction is often difficult to capture with standard computational chemistry methods, so a comparison of the results obtained with various methods is itself important.

This work presents explorations of the potential energy surface of clusters of atoms and of the interactions between molecules. First, structures of small aluminum clusters are examined and classified as ground states, transition states, or higher-order saddle points. Subsequently, the focus shifts to dispersion-dominated π-π interactions when the potential energy surfaces of benzene, substituted benzene, and pyridine dimers are explored. Because DNA nucleotide bases can be thought of as substituted heterocycles, a natural extension of the substituted benzene and pyridine investigations is to model paired nucleotide bases. Finally, the success of the dispersion studies inspires the development of an extension to the computational method used, which will enable the dispersion energy to be modeled - and the potential energy surface explored - in additional chemical systems.