Vibrational spectroscopy is one of the most powerful techniques in chemistry due to its ability to report direct information on the geometries and nuclear motions of complex molecules. At thermal equilibrium, rearrangements between different, or equivalent, molecular structures help to define the reactivity of the system. On the ultrafast timescale, these chemical exchange reactions may cause features in the vibrational spectrum to become averaged. The Bloch equations have been used previously to quantitatively predict the rates of such reactions. This approach has been quite successful in NMR spectroscopy where exchange on the microsecond timescale is often sufficient to cause peak coalescence. The application of the Bloch equations to IR spectroscopy has been debated due to the presence of other contributions to the vibrational lineshape that may prevent an accurate description of the relevant dynamics. Two dimensional infrared spectroscopy (2D-IR) is a time resolved ultrafast technique that is capable of directly measuring the kinetics of chemical exchange at thermal equilibrium where there is no net change in the populations of reactants and products. This work examines two model systems that display dynamic IR peak coalescence. A group of mixed valence dimers of trinuclear ruthenium clusters have exhibited a wide range of IR coalescence that is sensitive to solvent, temperature, and ligand substitution. No electron exchange was observed by 2D-IR during the timescale required for peak coalescence. The two isomers of a square pyramidal ruthenium dithiolene compound also result in highly coalesced spectra. The equilibrium populations, as well as the extent of peak coalescence, are strongly dependent on temperature and solvent. The experimental results of this second project were largely inconconclusive; however, future work with density functional theory calculations looks promising for revealing more information on these dynamics. Ultimately, the application of the Bloch equations to IR spectra is not generally reliable. While this type of analysis may still yield accurate results in some special cases, it is very difficult to distinguish legitimate exchange induced coalescence from similar broadened lineshape features, and this approach should be avoided.