B.R.S.M. Help! I'm trapped in a molecule factory!

25Mar/123

Named Reactions 2: Break It Down Now

Physics Nobel Laureate, legendary teacher and all-round cool guy Richard Feynman once said: “[There’s a] difference between knowing the name of something and knowing something”. This is true in a whole range of fields, and we’ve probably all seen enough students confidently assert that a particular step is “just a simple named reaction”, only to completely crumble when asked the mechanism or conditions. Still, I think named reactions are a great way of learning some really important chemistry that can then be applied to many other things. A chemist who knows, say, the fifty most common named reactions and a decent chunk of basic theory will be in a good position to take a guess at the mechanism of most things they encounter.  They're also a useful conversational shorthand if you want to convey how something works without reaching for a pen and paper. Very few reactions are so obscure and ‘out there’ that they’re not at least conceptually related to things we all know well. For example, I set this as part of a group problem session I ran last week:[1]

I know at least one reader will immediately recognise the rather obscure named reaction that starts this fairly wild cascade (his name begins with See), but for the average first or second PhD year student, things appear quite daunting at first glance. However, if we take a step back there are plenty of familiar elements in the reaction conditions here. Copper iodide and an amine base – that puts us into Sonogashira-type Cu-acetylide forming territory, giving us a nucleophile, and formaldehyde and the same secondary amine gives us a good iminium electrophile à la Mannich. Combining these two in the predictable fashion leads to a propargylic amine. From here we need to somehow add hydride to the top alkyne, and eliminate the dicyclohexylamine, which obviously isn't in the product. Initially, it's not so obvious how to do this, and Grossman’s rule is a must. As you can draw a nice 6-membered transition state for this, I like to think of this as a retro-ene-type reaction to lose an iminium ion, although I’m not sure what the literature consensus is. Finally, the allene formed undergoes a Myers-Saito reaction,[2] 1,5-hydrogen abstraction gives a nice stable captodative radical, and this recombines with the benzylic radical to give the final compound with only slight loss of ee.[3]

So, although not every step in a total synthesis can (or should) be a named reaction, I think that they provide a useful pool of information for breaking down more complex sequences.

Etc

  1. Full answer in Chem. Commun., 2012, 48, 2549-2551. This may go some way to explaining my massive unpopularity at group meetings.
  2. The Myers-Saito is just a Bergman, but with one alkyne swapped for an allene. The result is that the temperature barrier for cyclisation is much lower, and Nature occasionally uses this in place of the Bergman as the warhead in some of the enediyne antibiotics such as neocarzinostatin:
  3. This is because the intramolecular recombination of radicals is fast compared to the radical inversion which causes racemisation. See the paper for more discussion.

 

Comments (3) Trackbacks (0)
  1. Hey BRSM!

    Thanks for the track-back…but I feel I must clarify: if you hark back to Crabbe’s early papers (1981, I think?), he does a boatload of mechanistic studies, and surmises that the hydride transfer is actually copper-mediated. In essence, the Cu coordinates to the amine and alkyne simultaneously, the hydride pops off, “jumps” to the copper, and then to the alkyne.

    This model has had some support from further reaction development; when you use Cu, Au, or Ag catalysts with chiral ligands, you can control allene geometry (P vs. M).


Leave a comment


+ 3 = eight

No trackbacks yet.