Physical chemists are drawn to studies of fundamental physical and chemical phenomena using model systems that lay bare these phenomena, and enable their study uncluttered by complicating aspects that are peripheral to the central problem of interest. One of the enduring themes of our research has been the use of molecular clusters as model systems in which to study solvent effects and hydrogen bonding, particularly involving nature’s solvent water. Molecular clusters are collections of two to tens of molecules which are held together by intermolecular forces rather than chemical bonds. These molecular clusters can be formed in the gas phase by expanding a high pressure reaction mixture containing the components of interest (e.g., ‘solute’ and ‘solvent’) into vacuum, thereby cooling the molecules to low internal temperatures (within a few degrees of absolute zero) and forming (solute)_n-(solvent)_m molecular clusters. These molecular clusters enable the study of certain aspects of solvation in exquisite detail, building up the bulk phase one molecule at a time. We have used a range of laser-based spectroscopy techniques to determine the ultraviolet and infrared spectra of solute-(solvent)_n molecular clusters as a function of cluster size and composition. Many of our early studies used benzene as a prototypical aromatic solute, probing the interactions of benzene with a range of polar and non-polar solvents, including H₂O, CH₃OH, HCl, CHCl₃, and CCl₄.

Benzene-(H₂O)_n clusters with n=1-9 were studied in great detail. Benzene, as a non-polar aromatic, is relegated to the surface of the (H₂O)_n clusters, reminiscent of the immiscible nature of these solvents in the bulk. Because the benzene molecule sits on the surface of the (H₂O)_n cluster, it only weakly perturbs the H-bonding network of the ‘n’ H₂O molecules, forming a π H-bond between one of the surface OH groups and the π cloud of benzene. Infrared spectra of these clusters as a function of water cluster size provided some of the first size-selected spectra of the (H₂O)_n clusters, which changed from π-bound monomer and dimer to H-bonded cycles (n=3-5), to 3-dimensional networks (n=6,7), and then to cubes (n=8) and expanded cubes (n=9). This was the first ‘sighting’ of the water octamer, proving that its most stable structure was indeed that of a molecular ‘ice cube’, as predicted by theory. Two isomeric forms were detected, of S₄ and D_2d symmetry, distinguished from one another by slight differences in their interactions with benzene, and assigned based on their unique infrared spectral signatures. These studies were followed by analogous studies of benzene-(methanol)_m and mixed benzene-(H₂O)_n-(CH₃OH)_m clusters, each with characteristic OH stretch infrared absorptions that were signatures of the H-bonding architecture of the cluster.

Most molecules with biological relevance incorporate many H-bonding sites. Since biological processes occur in water, it was natural to extend our studies of solute-(H₂O)_n clusters to other biologically relevant molecules that have H-bonding sites for attachment of the water. We have studied a large number of such prototypical biological ‘building blocks’, and the first steps of their solvation by water. In many cases, water forms H-bonded bridges between donor and acceptor sites, reminiscent of the bound water that is found prevalent in X-ray structures of proteins. We have determined the spectroscopic signatures of many such bridges, and examples were found where the presence of the water bridge influenced the conformations of the solute, when it has such conformational flexibility.