Friday, February 21, 2014

Rhodium Gets It Done

Interesting, informative intermediates from Rh(I) silylation
In this week's Science, Hartwig and Cheng disclose some pretty swanky C-H activation chemistry. By combining Rh(I) dimer, a really bulky diphosphine ligand, a cheap, bulky hydrosilane, and a hydrogen "acceptor," the pair turn relatively distant aromatic bonds into highly functional silicon handles. In certain cases, tricks like capping a carbazole with a bulky TIPS group direct the silylation para - all the way to the other side of the aromatic ring!

The authors quickly point out that this silylation runs at "low" temps (80 deg C), uses fairly cheap commercially-available reagents, and occurs with almost reversed selectivity to the corresponding direct borylation conditions. But my favorite part comes from a deep dive into the Supporting Information. Far from the discussion of academic over-publishing we've had for the past few days, Hartwig and Cheng really sculpt a great paper here: Stability studies. Reactivity differences (Si vs. B). Cross-couplings. Preliminary mechanistic details.

As always, there's tons more to do. Getting out of the glovebox with a more stable Rh precursor, or translating the reactivity to an earlier metal (a tall order!). Deeper mechanistic studies would certainly show the way. Even more tantalizing? Using single-enantiomer versions of the bulky ligands to incorporate some chiral-at-silicon synthons. I can't wait to see the rest of this story.

5 comments:

  1. Nice paper, I guess the big improvement is the orthogonal methods for transforming C-Si and C-B bonds into OH, Br, I etc. That and the move away from difluorodisilanes as the silylating agents.

    The actual regioselectivity of the silylation step is still pretty much based on sterics, similar to C-H borylation.

    ReplyDelete
  2. Check out the NMRs given in the supp... they are quite scary! Many look like unpurified mixtures, and a lot of peaks are partially integrated to give the right numbers. I now wonder about the yields/ratios reported... Page S70 is especially awful, wth is that 13C!

    ReplyDelete
    Replies
    1. The important question: Is it there, beyond a doubt? There is no physical law that says everything is completely separable, that certain NMR experiments will always give you good integrations or that the compounds are always stable enough to provide extremely pure material. As long as it is there, it is an interesting discovery.

      Delete
    2. Its called C-F coupling....

      Delete
    3. I do not know where you hallucinated C-F coupling on page S70 since there are no fluorine in the molecule. Besides, coupling does not alter integrations in 1H NMR...

      Delete