This weeks Mega Chemist was Yugoslavian born Tomislav Rovis, who since 2008 has held the John K. Stille chair at Colorado State University. As a human biology graduate, despite his minimal background in synthetic organic chemistry, Rovis had a productive Ph.D with Prof. Mark Lautens at Toronto with six journal publications before moving to Harvard. As an NSERC PDRF he investigated the use of C2 symmetric Cu(II) complexes as chiral lewis acids under the supervision of the illustrious Prof. David Evans. His publications kept building up, with a couple more sneaking out of Toronto with his name on thanks to the continuation of his work by non-other than the late Keith Fagnou. In 2000 Rovis started his independent career at Colorado State, where he remains to date.
A slightly extended search online about Rovis led me to discover koofers.com, a website where you can rank your professor! Rovis receives a solid 4 out of 5 stars as ranked by his undergraduate students, but his class sure as hell sounds tough.
Rovis has diverse research interests, but as I said before rhodium and NHC catalysis seem to take up most of his time. His most significant independent work is arguably that on the asymmetric Stetter reaction. Here are a couple of papers should you fancy it: One and Two.
This weeks paper was chosen in a little bit of a rush. My inability to read off the screen with any ease, and a lack of printer at home (where I am slowly writing my thesis and NOT my blog) means once chosen, the paper becomes THE one. As we did a bit on NHCs a couple of weeks ago I wanted to head away from that, so instead we are heading into the ever so trendy corner of C-H activation.
The paper I grabbed on pyridone synthesis was chosen as the headline was ‘improved catalyst architecture for Rh (III) C-H activation’. This sounded like the kind of thing I am into; the more we understand catalysts and their ligands, the less we should have to screen conditions for different substrates, and thus alleviate what often limits the wide spread uptake of otherwise brilliant chemistry. Anyway, my hopes of a complex rational design of new ligands were soon shattered. As is often the case, it turns out bigger is (mostly) better; never the less this paper throws up a couple of interesting mechanistic surprises – it is not all about concerted metalation deprotonation!
The main body of the paper is an extension of previous work on the formation isoquinilones from benzamides using Rh catalysis (scheme 1, A). Rovis reports the use of a bulkier ligand in contrast to earlier work, and is utilising acrylamides rather than benzamides, and thus is generating pyridones (scheme 1, B).
A direct comparison of the new catalyst system vs. [RhCp*Cl2]2 generally shows enhanced regioselectivity (X and Y) for the bulkier ligand. A couple of the prettiest products are shown, highlighting the scope of the substrates (scheme 2).
Importantly for contemporary (green) methodology and the scale up a chemical reactions, Rovis showed that if reaction was undertaken in air it was possible to cut the loading of the copper co-oxidant from 210% to 20%. Fantastic. More so, reaction proceeded without the addition of any copper when performed under an atmosphere of O2, and with the addition of 4 equivalants of NaOAc – I know which one I would rather get rid of. The reported yield was only 40% and no reaction time was given, relative or otherwise, but things like this always cheer me up a bit.
Onto the mechanism; although this only made the SI it gets a mention here (scheme 3) as Rovis used the beautifully named angelic acid (well, its’ N-methyl amide analogue) to demonstrate that a cis-beta-proton is required for reaction to proceed. Unfortunately he did loose a methyl group from angelic methyl amide in the SI, giving the much less exquisitely named crotonic acid!
Support for the concerted metalation deprotonation (CMD) mechanism often cited in Rh and Pt C-H activation is often demonstrated through KIE studies. Further investigation into the mechanism, looking for conformation of the turnover limiting step (TLS) Rovis found through comparison of proteo and deutero N-methyl-cinnamide a kinetic isotope effect of 2.2. This suggested that C-H activation – the formal insertion of rhodium into the C-H bond – was the limiting step, but Rovis was not so confident he didn’t investigate further.
It follows that if C-H activation is the TLS, then increasing the acidity of this proton facilitating C-H activation should increase the rate of reaction. To investigate this a competition experiment between N-methyl-cinnamide and the penta-fluorinated cyclopently analogue (which has a much more acidic beta proton) was undertaken, and threw up a surprising result: only pyridone 1 was formed (scheme 4).
A subsequent Hammet study was undertaken, and a potential break in the plot, with electron rich and electron poor substrates undergoing slower reaction than electron neutral species suggested (I have just learned) a change in mechanism. A further KIE study (scheme 5) showed significant differences in the KIE for electron neutral and electron deficient systems. A KIE value of 1.2 suggested that for electron deficient systems the C-H activation is not in fact the TLS, supporting a change in mechanism.
So although we did not get a beautifully hand crafted and exquisitely designed new ligand, we did learn about an important shift in mechanism depending on the electronics of the substrate. This new understanding can only help direct us that little bit more when it comes to determining new reaction conditions and reagents for what will be a continually expanding smorgasbord (stolen from KC himself) of C-H activation reactions.