Redox Champ: NHC Ipso, Facto

Bond density difference between electron-rich and -poor NHCs
Source: Cavallo, JACS  ASAP

On the surface, N-heterocyclic carbenes (NHCs) seem like dream ligands. They have strong, neutral dative electron pairs, steric bulk, and enough variants now exist that chemists can "dial in" chirality, control R-group rotation, or add them as salts.
Could we possibly improve them any further?

A new JACS ASAP answers: of course! Professor Luigi Cavallo, and KAUST / U. de Salerno coworkers, like to examine NHC ligands computationally. Using Gaussian '09 molecular modeling software, the authors toss in a variety of electron-donating (ex: NMe2, Me) or withdrawing (ex: Br, NO2) substituents on the aryl rings of mock NHC ligands. They then virtually "staple" them onto Grubbs, Grela, and Ir-COD catalysts, and calculate the redox potentials of the Ru and Ir metal centers.

These data trend like you might expect: electron-withdrawing groups on the NHC increase the redox potential, while donating groups reduce it. Well, what's driving these effects? The authors initially suspect σ/π donicity (that's a $5.00 word, right there), the ability of the NHC-M bond to directly influence metal properties. However, they note something odd: there's precious little change in the bond lengths between complexes, only 0.005-0.05Å; by contrast, in carbon-carbon bonding, moving from alkane to alkene shaves off 0.2Å! Thus, they suspect that the NHCs' influence isn't coming through their main bond.

Well, if it's not electronics, then what's changing these redox properties? Enter the 'ipso effect,' a long-observed, yet little-invoked, mechanism for ligand-to-metal charge transfer. The arene carbon directly connected to the heteroatom rubs up against the metal d orbitals, and charge can move through space. Perhaps the best known non-NHC systems to exhibit this effect are the highly-active Buchwald biaryl phosphines, which show ipso interactions with palladium and gold, among other metals. Cavallo and coworkers compare electron density maps (the trippy blue-and-red models, above R) between cationic and neutral metal states, and note a huge red splotch of π-to-d donation (see d) from the NHC orbitals to the metal.

NHC electronics lower intermediate energies
Source: Cavallo, JACS ASAP

But wait, there's more! The authors apply this model to previous observations in Ru metathesis chemistry, then extend the metaphor and tackle Pd biaryl coupling. As shown in the graphic (left), they calculate a ~3 kcal / mol difference for the first catalytic intermediate, depending on the electronic environment of the NHC used. For a throwaway closing line, they let us in on a final trade secret: this effect could be used to stabilize lots of high-valent metal species...do I hear any calls for nickel (IV) chemistry???