“Synthesis of Quaternary Carbon Centers via Hydroformylation” Sun, X.; Frimpong, K.; Tan, K. L. J. Am. Chem. Soc. 2010, 132, 11841. DOI: 10.1021/ja1036226
Just as the advent of protecting groups opened up new chemical space accessible by current synthetic techniques, so too did the advent of the directing group. Both however come with the downside of often requiring additional synthetic steps for the installation and removal of these groups. For example, the past decade or so has seen a number of efforts at using directing groups such as phosphines to affect the course of hydroformylation reactions. I should mention here that people are so interested in the hydroformylation of alkenes in part because it is such an industrially important reaction, with 9 million tons of aldehyde products being produced in this way per year (ref). The challenges with this reaction are selectivity, one of the foremost issues being that terminal alkenes preferentially give linear products. Shown below is a 2001 example from the Leighton group where a phosphine directing group on allylic ethers yields the branched hydroformylation product (Markovnikov addition), whereas the linear product (anti-Markovnikov) would be favored in the absence of the directing group:
The Up-Side: this method offers access to substrates not previously available through hydroformylation. The Down-Side: it is hard to imagine an industrially-relevant product containing that specific phosphine moiety, so it would undoubtedly have to be cleaved, making the overall process highly atom un-economical. This latter point could be addressed by somehow using the phosphine group in a catalytic, instead of stoichiometric, fashion. This is exactly what the research group of Kian Tan at Boston College has been up to lately. They have developed what they term a “scaffolding ligand” that coordinates to the organometallic catalyst as well as rapidly and reversibly (two important things for this type of catalysis) forms covalent bonds with alcohols. In doing so it brings the substrate (in blue below) and catalyst (in red below) in close proximity and influences the course of the reaction.
In what I think is the most impressive recent employment of this ligand, the Tan group was able to use benzazaphospholidine 1 to form quaternary stereocenters (another big challenge for hydroformylation reactions):
Due to problems with isolating the resulting beta-hydroxy aldehyde products, in their report they oxidize the reaction mixture and isolate the beta-hydroxyl carboxylic acids instead:
If PPh3 is used in place of their benzazaphospholidine ligand, a <2:98 branched:linear ratio is observed, showing what an efficient job their ligand does at completely reversing the selectivity of this process. The substrate scope is one big limitation of this methodology, as it is limited to substrates like that shown above possessing an aromatic ring as the second substituent on the alkene. With alky substitution in that position, the branched:linear ratio drops. It is notable that only primary allylic alcohols were investigated in this study, as they avoid complications with syn:anti selectivity as was seen in the above example from the Leighton group. One thing at a time I suppose.
The Tan group have also applied this ligand to the hydroformylation of homoallylic alcohols, allylic sulfonamides, and trisubstituted allylic alcohols. In addition they have recently developed a ligand capable of affecting enantioselective hydroformylation.
I would be remiss without mentioning the fact that the Tan group are not the only ones in this specific field. For a brief overview, see this recent highlight. Overall this new type of catalysis gets me really psyched, and I’m sure we’ll be seeing it extended into numerous different types of reactions in the future.