Skip to main content

Chemical Biology

A New Way to Make Azides

I wanted to mention this paper that’s out in Nature, especially since I was mentioning azide/alkyne click chemistry the other day. If you’re using that system in any sort of chemical diversity sense, you’ve run into problems on the azide end. There are not a whole lot of commercially available azides out there (although definitely more than there used to be before this reaction became popular!) And while there are ways to prepare them from amines, none of them are wildly appealing. For a while there was an imidazolium reagent that looked promising, but it has some stability problems. Then there’s triflyl azide, but no one enjoys that stuff, either: you have to make it fresh and keep it dilute, and it’s a rather slow reaction that doesn’t always complete.

The current paper, from the Shanghai Inst. for Organic Chemistry and Barry Sharpless at Scripps, describes the use of fluorosulfonylazide, which is generated in solution in situ, as a fast and high-yielding reagent for azide formation from primary amines. The paper describes its use on a very wide variety of structures, with all sorts of complex functional groups, and indeed extends it to 96-well plate format for generating azide libraries. They prepared 1224 azides, half of which are previously unknown in the literature, and show that they’re ready for Cu-catalyzed click chemistry after sitting for at least six months in solution, in the plate wells.

So this looks like first a good way to make specific azides for chemical biology applications, including direct azide formation on very complex substrates and secondly an entry into azide diversity sets, which have really never existed at this scale. Given the number of primary amines out there, though, they sure can now. . .

16 comments on “A New Way to Make Azides”

  1. Wavefunction says:

    I am as intrigued by the plate-based chemistry as by the methodology. A lot of indicators are pointing toward more nanomole scale, high-throughput experimentation in chemistry and this seems to be another data point on the way there.

  2. milkshake says:

    it would be better to develop some alternate route to FSO2N3. The starting fluorosulfonyl dimethylimidazolium reagent is very dear (1000 USD buys you 75mmol of the stuff at Aldrich) and making it by yourself, while not too laborious, involves use of MeOTf and SO2F2. I would hate scaling this up – and I did work with MeOTf on this scale before.

    1. G2 says:

      They described the synthesis using Tf2O on page S34 of the supporting info …

      1. milkshake says:

        no they did not – it is a different reagent, please check the structure again. The reagent you need is made by 2 step process (2-Me imidazole +SO2F2, then methylation with MeOTf), it is described in their preceding Angewandte paper

        1. G2 says:

          Sorry, my fault: mixed up CF3 with F.

          And Aldrich surprisingly seems to be the cheapest source, however you have to hurry up – delivery date is mid Dec. for smaller quantites when ordered in Europe from Aldrich.

    2. BK says:

      Milkshake, in my last position, we were making a compound that prescribed 1 mole of MeOTf (yes, your eyes did not deceive you, one MOLE of MeOTf), which was a 10 molar excess (or 10 eq.) for the reaction in question. We were not looked at favorably by QA and QC because we wanted to change it for safety sake but the process was validated for commercial production. A colleague found that reaction to be just as reactive at 1.5 eq. To this day, he and I were the only ones to did that process and we are the only two people that know it is effective at a smaller excess. While I never scaled down to 1.5 eq., I would not be surprised if he violated his ethics and GDP training to use that lower excess and just scrap the remaining 8.5 eq. until he can validate the new process that had numerous safety updates.

  3. Patrick Lam says:

    Chan-Lam Reaction is another simple alternative to make aryl azides- copper catalyzed N-arylation of sodium azide with aryl pinacolboronate at RT (Guo et al, TL, 2007, 3525 and Aldrich et al, Synlett, 2010, 1441.

    1. milkshake says:

      sodium azide + copper catalyst = very nasty risk. Lead azide is incompatible with brass, copper because traces of copper azide sensitize already very touchy primary explosive to a point of instability. Heavy metal azides like Hg, Ag azides readily blow up even in wet slurry.

    2. the chan chan man says:

      Are you the Lam of the Chan-Lam?

  4. Patrick Lam says:

    Yes, I am the Lam of the Chan-Lam Reaction (Syn. 2011, 829). I am also the group leader/co-inventor for the discovery of ELIQUIS® which has the highest sales of all small molecule drugs in 2018 (😃

  5. Anon3 says:

    Barry Sharpless at Scripps? Apparently no more, if you have a look at the author affiliations listed in the paper.

    1. electrochemist says:

      According to this press release, he will work at the Shanghai Institute of Organic Chemistry “at least one month every year for five years.”

    2. Fighting Butane says:

      He’s here, don’t you fret.

      1. milkshake says:

        The decoherence is manifest.

  6. Krusty says:

    Azides need to be handled in flow

  7. Korora says:

    Good news for Crazy Harry, then!

Comments are closed.