xviii, 724 p. : ill. (some col.) / Structural trends applicable to crystal engineering were studied in three classes of thiourea-based compounds. The aim of the study was to identify, predict, and ultimately design reliable single-molecule structural features, which could then be used to engineer crystals with desirable properties. In one class of compounds, this goal was achieved: N-alkyl and N-aryl derivatives of N,N'-bis(3-thioureidopropyl)piperazine adopted an identical conformation in the solid state, which resulted in near-identical crystal packing. A second class of closely related compounds, N-substituted tris(2-thioureidoethyl)amines, showed no such reliability in the solid state, likely because the parent structure lacked hydrogen-bonding functionalities sufficient to control intramolecular structure. In the third class of compounds that we studied, 1-benzoyl-3-(2-pyridyl)thioureas, substitution patterns were often predictive of molecular conformation; however, these intramolecular trends did not lead to recognizable crystal packing motifs. Nevertheless, certain physical properties observed in this last class of compounds--color, solubility, and often crystallinity--were conformer-specific, interestingly without any apparent relevance to crystal lattice structure. Solution-state and solid-state conformational trends in these 1-benzoyl-3-(2-pyridyl)thioureas have been documented, and speculations as to the source of color in one of the two observed conformations have been noted. / Committee in charge: Shih-Yuan Liu, Chairperson; Kenneth M. Doxsee, Advisor; Michael M. Haley, Member; Catherine J. Page, Member; Paul J. Wallace, Outside Member