Timothy P. Hanusa
Solvent-free reactions are a characteristic feature of mechanochemistry, which uses mechanical action (e.g., impact and friction) to generate the energy needed for chemical reactions. The cracking of crystals by grinding can fracture bonds and create radicals in ways that do not occur in a solvent, and the outcome of mechanochemical reactions need not be (are often are not!) the same as in solution. Solid-state syntheses provide the opportunity to investigate new compounds not otherwise obtainable. This might be because available solvents interfere with the interaction of the reagents, or because solvent molecules may bind irreversibly to the product, and change its structure and reactivity.
We are particularly interested in low-coordinate metal complexes, as they can display higher reactivity than solvated species. The tris(allyl)aluminum species Al[A´3] (A´= 1,3-(SiMe3)2C3H3)), for example, can be made mechanochemically in a few minutes of grinding, and reacts immediately with oxygen-containing organics, unlike the much slower reactions of solvated species.
Control of stereochemistry in solid-state organometallic synthesis is also under investigation. For example, the group 15 compounds M[A´]3 (M = As, Sb) exist as both symmetric (C3) and asymmetric (C1) diastereomers, but mechanochemical synthesis increases the amount of the C1 diastereomer. This appears to be related to the particular lattice structures of the solid state starting materials, and provides a unique handle for designing stereocontrolled reactions specifically for mechanochemical initiation.
New Ligand Design. Main-group metals (groups 1, 2, 12, 13) and the lanthanides usually have only one or at most two readily available oxidation states, which limits the range of chemistry possible from their metal complexes. We are developing ligands that put metals into unusual coordination environments, potentially overcoming this restriction. For instance, a set of two cyclopentadienyl (Cp) ligands, when bridged with a silyl group [R2SiCp2]2-, necessarily carries a dinegative charge. If a phosphonium bridge is used instead (i.e., [R2P+Cp2]-), the ligand is uninegative, and when bound to a group 2 metal, the metal coordination environment is similar to that of a trivalent lanthanide. Using this approach, we have made a calcium amido complex that is structurally similar to a samarium analogue.
We are investigating other ligand types besides cyclopentadienyl (indenyls, allyls) so as to extend the concept to a wider range of applications, including catalysis with earth-abundant main-group elements (alkali/alkaline-earth metals).
Computational investigations of bonding, structure, and dispersion effects in inorganic/organometallic systems. Unusual structural arrangements often require computational analysis in order to generate a complete picture of their origins and consequences for structure and reactivity. Computational modeling of cation-pi bonding, such as between K+ and the double bonds in the allyl ligands in the complex K[ZnA´3], has demonstrated that the interaction can be ≥ 100 kJ mol-1 in strength, and that it also contributes to the templating of the ligands, resulting in the highly symmetric (C3 axis) complex. Studies of the effect of cation-pi interactions on reactivity are ongoing.