As chemists, we know of ligands as a convenient tool for manipulating the electronic properties of a metal center. Specifically, they are typically thought of as a means to shape the energetic and spatial properties of metallic frontier orbitals, and to stabilize some of the more exotic metal oxidation states. However, to perceive ligands in this way is an unnecessarily limited view; this realization occurred to me during the isolation of an interesting terminal manganese nitride phthalocyanine complex. We were initially studying the complex as a model compound for nitride coupling, hoping to glean some insight into the mechanism of ammonia oxidation for the development of better catalysts for a direct ammonia fuel cell. As with manganese nitride salen complexes, we expected the reactivity at the terminal nitrogen to be tunable by accessing different redox events. However, our characterization suggested that this was not the case, and in fact all of the frontier orbital density appeared to be on macrocyclic ring. We found this bizarre, and proceeded to isolate single crystals of all five accessible oxidation states, and indeed, all of the redox processes seemed to originate from the ring, and the bond and reactivity of the nitride did not seem to change significantly with oxidation state. Additionally, the oxidation states on the ring were exotic, with some displaying anti-aromatic character, and others unexpected changes in bond length and bond angle. This bizarre complex appeared to be an example of an inert metal core stabilizing unusual oxidation states of a redox-rich ligand, rather than an inert ligand stabilizing unusual oxidation state of a redox-rich metal.
Around the same time, a colleague of mine, Tim Carroll, was designing model complexes for VPO, an industrial catalyst utilized in the production of maleic anhydride. Tim demonstrated that one of his model complexes was capable of performing H-atom transfer through a P=O bond, with the metal center acting as a sterically-isolated electron reservoir, but not a direct participant in the H-atom transfer. It had been proposed computationally that this was precisely how VPO was carrying out the 14 e– oxidation of n-butane to maleic anhydride, although the specifics of how it is capable of performing such a demanding reaction with high selectivity remains a mystery worth studying. Both of these projects changed the way I perceive what a ligand is: Ligands are not merely a stabilizing entity used to manipulate the orbitals of metals, they themselves are can be the most interesting part of a metallic complex, with the metal functioning as the ancillary component. I suspect that a new and exciting frontier in organometallic chemistry will be designing and manipulating ligand orbitals through the choice of different metal centers, and designing catalysts that perform their catalytic function through the ligand, rather than the metal. This could clear the way for a new regime of complexes with well-defined catalytic behavior that utilize the rich electronic profile of metals to stabilize exotic, catalytically-useful oxidation state of the ligand. An apparent advantage of this strategy would be that the reactivity of early main group elements is generally more predictable than metal centers, as seems to be evidenced by VPO’s extremely good selectivity. Ultimately, this could be an interesting pathway for organometallic chemists to explore, and I hope to find more examples in the future.