Monday, February 15, 2010

Fabulous methylene functionalizations

ResearchBlogging.orgM. Christina White from UIUC has again hit Science with her direct C-H functionalization chemistry in the January 29th issue (doi: 10.1126/science.1183602). As an alternate route to traditional functional group modification, White's group, and those of groups pursuing C-H functionalization (which I mentioned in my previous post) seek to "streamline" synthesis by cutting out unnecessary steps and going right in for the kill at the otherwise "inert" C-H bond.  Badass. 

In this paper, selective methylene C-H oxidation is achieved with substoichiometric amounts of peroxide, acetic acid, and the same catalyst as published in the same journal in 2007 for tertiary C-H bond activation, Fe(S,S-PDP) - nice environmentally friendly conditions.  This time the catalyst is selective for secondary C-H bonds when tertiary positions are unfavored due to additional sterics or a nearby electron withdrawing group (EWG).  This was indicated in one of the substrates in the 2007 Science paper, so it's not a huge surprise nor is it fishy that the selectivity is suddenly "different." 

Normally, tertiary C-H bonds (tertiary = 3 C's attached to the C) are the easiest for this catalyst to cleave as they are the most electron-rich C-H bonds of the molecule, but the catalyst is surprisingly - and predictably - selective for certain secondary C-H bonds, as demonstrated with a substrate scope table and a few complex molecules to boot.  The C-H bond that is oxidized:
  • Is the furthest from any EWG, which would obviously deactivate the bond by reducing electron density
  • Furthest from bulky carbon substituents (i.e. dimethylene)
  • Next to an sp2 hybridized substituent (including a cyclopropyl group) or an atom with lone pairs (i.e. ethereal oxygen)
These are all very intuitive factors to drive selectivity.  One of the tables neatly summed them up with examples, so I'm reproducing it here (crappily), complete with White's signature office-wall yellow:

Check out this exquisite example of applying the methodology to a more complex substrate - the method predicts the sites of oxidation well.  

Picky, skeptical scientist that I am, I'm a wee bit bothered by the lack of integrations on the 1H NMR spectra and the obvious alteration of the HMQC spectra  in the Supporting Information (it looks like someone dragged around the big spraypainter in Paint - at least when I viewed it on a Macbook Pro).  I suppose beautification is important, but why don't I get to see the dirty specks??  Anyway, there are 2 different sets of conditions to perform this reaction, one of which is sometimes better than the other - either a slow addition with 25 mol% catalyst and the peroxide (~1h) via syringe pump, or 3 iterative additions with 5 mol% catalyst and the peroxide in each, dropwise.

There's been ANOTHER paper recently wherein the authors from Penn State, Y. Feng & G. Chen, report a direct functionalization of a methylene C-H bond in Angew. Chem. Int. Ed. (doi: 10.1002/anie.200905134), "Total Synthesis of Celogentin C by Stereoselective C H Activation."  Celogentin C is a bicyclic peptide active against tubulin polymerization isolated from Celosia argentea.  

The parts of the synthesis I'm summarizing install the part of the molecule that is dark green, and the methylene indolylation occurs at the bright green bond.   By using an iodoindole and a palladium catalyst with a temporary palladium-coordinating group to direct the reaction to the specific C-H bond, the reaction proceeds in good yield and excellent regio- and stereoselectivity.  The chemistry is inspired by the strong precedent (same substrate, same conditions) by Corey in Org Lett in 2006 (doi: 10.1021/ol061389j) where the same bond was either arylated or acetylated.  In this paper, the difference is the iodoindole, which is prepared from tryptophan.  Tryptophan is protected, nitrated, the nitro group reduced, and a Sandmeyer reaction applied to convert to iodine.  The C-C bond is then formed via palladium-catalyzed coupling at the methylene beta to the carbonyl.  The phthalate protecting group is used though azide is needed there later on in the synthesis because the indolylation shut down in the presence of the azide. 

Presumably, the above intermediate forms according to the authors, and it is this species that performs oxidative addition with the C-I bond and reductive elimination to produce the coupling.  I don't like the idea of the C-H insertion happening immediately, but perhaps the slowness of this step is why 2 equivalents of this coupling partner is optimal.  Oxidative addition into the iodide, especially considering the presence of the silver salt, SHOULD theoretically be first, but if Corey proposed the reverse order, then, well.

For more awesome chemistry like this, check out a recent review in Chemistry, doi:  10.1002/chem.200902374, published in memory of Keith Fagnou. 

  • Chen, M., & White, M. (2010). Combined Effects on Selectivity in Fe-Catalyzed Methylene Oxidation Science, 327 (5965), 566-571 DOI: 10.1126/science.1183602
  • Chen, M., & White, M. (2007). A Predictably Selective Aliphatic C H Oxidation Reaction for Complex Molecule Synthesis Science, 318 (5851), 783-787 DOI: 10.1126/science.1148597
  • Feng, Y., & Chen, G. (2009). Total Synthesis of Celogentin C by Stereoselective CH Activation Angewandte Chemie International Edition DOI: 10.1002/anie.200905134


Constructive criticism welcome; criticism for judgement's sake, not.