Abstract and subjects
The study of the mechanism of the Noyori-Ikariya asymmetric transfer hydrogenation of ketones spans nearly three decades of investigations. Whereas the early part of the catalytic cycle being the hydride transfer is now well-understood, the later part being the proton transfer is still ambiguous. Specifically, the source of the proton can be the N-H functionality of the catalyst and/or the O-H functionality of the reagent/solvent, leading to two conceptually different catalytic cycles or even their combination. For three popular reagents/solvents typically used in the method, namely, propan-2-ol, 5:2 HCO2H-NEt3, and water, either the source of the proton is presently unknown or the evidence is presented partially by only one approach-experimental or computational. The present work eliminates this ambiguity by means of various molecular dynamics simulation methods such as ab initio, quantum mechanics/ molecular mechanics, and path integral to include quantum tunneling effects. Here, we show that for the archetypal (S)-RuH[(R,R)-Tsdpen](mesitylene) catalyst complex, the source of the proton in propan-2-ol is the catalyst's N-H functionality, whereas in more acidic water, binary 5:2 HCO2H-NEt3, or neat formic acid, the source of the proton is the reagent/solvent. Thus, depending on the nature of the reagent/solvent, the catalyst's ligand can be either chemically non-innocent or chemically innocent in the Noyori-Ikariya reaction, which opens opportunities for outer-sphere homogeneous catalyst design.