Skip to main content

Chemical News

Ripping Out Amines

I’ve been meaning to write about this new synthetic organic chemistry paper, because it’s just so weird. That adjective probably needs a little explanation. The next few paragraphs will try to provide that; then we’ll get down to the paper itself.

Probably a good fraction of the readership here has had at least the sophomore-organic level of exposure to the field. Even if you haven’t, the landscape can be described in general terms. The different ways to form and break bonds in molecules fall into broad classes, and most organic chemistry classes are structured along those lines. If you try to memorize two semesters of organic chemistry as if it were a huge pile of random facts you are making a lot of unnecessary work for yourself. The reactions themselves fall into families whose actual mechanisms are conceptually related, and it’s a lot easier to keep track of things that way.

To pick one classic example, you can form a negatively charged carbon much more easily if it’s next to an electron-withdrawing group like a carbonyl or sulfonyl, for example, and those carbanions can go on to attack other “electrophilic” groups. So there’s a whole list of such reactions, many of which are named after 19th- and early-20th-century chemists, but they all start with carbon center that’s easily deprotonated, getting a full negative charge on it and turning around and reacting with functional groups that have more of a positive charge on one of their atoms.

There’s a big family of reactions that go through positively-charged carbons. Fully or partially charged oxygen and nitrogen atoms have their own things going. Outside of positive and negative charges, there’s a whole group of ring-forming cycloaddition reactions where several bonds’ worth of electrons sort of snap-snap-snap around in a circle (and some rearrangement reactions that work more or less the same way). There’s a big family of “free radical” reactions that involve single electrons on carbons and other atoms instead of the electron pairs you see in the charged mechanisms, and so on.

But what all this does is give an organic chemist a familiar landscape to work in. There are many transformations of molecules that (even if you’ve never seen them before), just look fairly sensible because of all the others you’ve learned. You’d maybe need a little time to draw the exact “arrow-pushing” mechanism of the reaction, but just looking at it you’re willing to believe that it’s OK.

And then there are some that just look bizarre. I’m a fan of those, because they promise stuff beyond the mental framework that I’ve gotten used to since I took sophomore organic back in ’81. When these appear, it’s like suddenly discovering while you’re driving that there’s a new shortcut road that you didn’t know about, one that connects two neighborhoods that you’ve always thought of as being far apart.

At right is the current bizarre reaction. Conceptually, it’s an amine deletion: an NH group is ripped out of a ring or chain and replaced by a new carbon-carbon bond. It’s safe to say that we haven’t had anything quite like that! It works through functionalizing the amine with a reagent that forms an N-N bond and then breaks down to an isodiazene, a rather exotic species that would then rather just turn into nitrogen gas and vamoose. When it does, it leaves behind free radicals on the two carbons that used to have the amine bonds, and those (doubtless shocked by this weird turn of events) gladly embrace each other to form the new C-C bond. Why yes, that is the way I tend to think about reaction mechanisms, why do you ask?

As the authors note, this means that a wide-ranging amine-forming reaction like reductive amination can now serve as a precursor step for a carbon-carbon bond forming reaction. That means a lot, because making amines is relatively easy, while making C-C bonds is relatively hard. The reaction seems to be very tolerant of other functional groups in the molecule, which is good news. You do want to have radical-stabilizing groups on at least one of the two carbons involved, ideally.

My first thought on seeing this reaction was that I wished it ran (conceptually) in the opposite direction: I would very much like something that turned an alkane chain into a secondary amine, or a cyclopentane into a piperidine. But the more I think about this one, the more it grows on me. There are some carbon frameworks that are going to be a lot easier to get to via this route than by any others, and a lot of diversity than can come in via amine-forming reactions between a variety of partners. I hope I can get the chance to try out this transformation with my own hands!

43 comments on “Ripping Out Amines”

  1. bruce says:

    “My first thought on seeing this reaction was that I wished it ran (conceptually) in the opposite direction: I would very much like something that turned an alkane chain into a secondary amine, or a cyclopentane into a piperidine.”

    Selectivity would be an issue, for anything larger than propane.

    1. Anonymous says:

      Schmidt reaction? I used to know more than sophomore chemistry, but they made me not be not do organic synthesis anymore. Not if I wanted a paycheck. 🙁

  2. Walter Sobchak says:

    I misread the head line as “Ripping Out Animes” and I wondered if you had a thing about Japanese movies.

    As Emily Litella used to say: “Never Mind’

  3. sgcox says:

    I wonder if it can be utilised to make all those wonderful terpenes we see in Nature.
    In scale, if possible.

  4. John Wayne says:

    I just read every word of this paper and skimmed the extensive supplemental section. This is almost unbelievably strong work. It presents a useful and new reaction, details the current scope of the reaction, and uses solid physical organic chemistry techniques to comment on the mechanism and how they interact with the limitations of the method. I am very impressed. Congratulations to everybody involved.

    1. luysii says:

      Amen. A classic example of thinking outside the box. Bravo

  5. Ash says:

    I remember an old chemical degradation method from the heroic days of structure determination through degradation that I am blanking out on, but I think that too involved a rather unexpected disappearance of nitrogen.

  6. Sarah H says:

    Great post! I loved the description of grouping chemical reactions into categories– and your “conceptual” description of what is going on in the skeletal nitrogen excision. If you had taught organic chemistry, I might have become a medicinal chemist rather than evolving into a crystallographer.

  7. CET says:

    With the important caveat that I don’t have access to the paper itself, I am both very impressed and more than a little surprised. I’d have written the idea off as a selectivity and functional group tolerance nightmare that would only work in a couple of carefully selected examples.

    It’s good to be wrong about these things sometimes!

  8. Barry says:

    Running this is reverse is effectively the (long-known) Beckmann rearrangement.
    Ripping out Nitrogen is conceptually akin to Barton’s “winch reaction” making a C=C bond from the azine

  9. a says:

    Also note: 2nd year assistant professor, only with startup money (and talented U Chicago postdoc/students to be sure)

  10. tt says:

    Really cool and fundamentally different reaction. Closest analogy I can think of (and it’s not really close) is like a cheletropic elimination, but one that generates radicals that form a bond. Big congrats to the Levin Lab…crazy strong start for a new PI. Paper of the year (so far).

  11. Some idiot says:

    As a process chemist, I say hats off! A very nice bit of work, particularly for getting to structures that are a bit weird…!

    My first thought (again as a process chemist) would be as to how acceptable the “nitrogen donor” would be on scale from a safety point of view… But there is a lot of carbon on the heteroatoms, so it could well be ok… Well worth looking at though! A DSC on the stuff would be interesting… And even if there was a concern there, it may well be that it would be well suited to flow…

  12. Kazoochemist says:

    Very nice work. It reminds me of the Romberg-Backlund reaction.

    1. Kazoochemist says:

      rAmberg, not Romberg. I hate spell-check. I am ignoring the umlaut.

  13. Project Osprey says:

    There is a similarity here with the Kishner cyclopropane synthesis. Ejection of N2 from a ring, forming a diradical which then cyclises. The end product is totally different though.

    I enjoy these “so much is the same, except the outcome” comparisons (can anyone else thing of anything close?)

  14. rhodium says:

    Clever and useful. Can you convert carbonyls to dioxaranes and blow out carbon dioxide?

  15. Siddharth Dasgupta says:

    >and those (doubtless shocked by this weird turn of events) gladly embrace each other to form the new C-C bond. Why yes, that is the way I tend to think about reaction mechanisms, why do you ask?

    That brings up a related but amusing mental image of Marty Semmelhack’s class in 1981!

  16. Iosef K. says:

    Am I the first one to mention the similarities with J. Org. Chem. 2017, 82, 4677 (which is ref. 22 in the current paper)? I know we’re standing on the shoulders of giants but this is a bit too close of a precedent IMO, especially given the journals these respective papers are in. Someone cynical could whisper the words “rip-off” here…

    1. RealChemist says:

      I agree with you.

  17. Andy says:

    This is chemistry, nothing is new, but if I had the choice between making Kennedy’s reagent and sulfuryl azide, I know which one I would choose.

    1. Dr Dad says:

      Leaving aside that it’s a multiple step approach, you’re really not kidding: sulfuryl diazide is described as “exceedingly explosive” …

      There is always the even older-school approach of using Angeli’s salt, also cited by the paper:, if you can get your hands on it!

      I think the better criticism here is that the scope of this paper is pretty limited because isodiazenes are limited. Pemetrexed is nice, but I don’t know of any other APIs with a bibenzyl linkage…

      1. Andy says:

        Huh. I’d never heard of Agneli’s salt. Fascinating, thank you.

  18. PeterC says:

    The shortcut comment reminds me of the sensation I get in London sometimes when I discover a short walking route between two tube stations I’d previously thought were far apart on the tube map.

  19. Anon says:

    Derek: This is really very clever reaction. If one were to look this as cheletropic extrusion (of Nitrogen!) then I can think of other reaction that could do the same. For e.g. tetra hydro thiophenes sequentially could be brominated (at C alpha to S), oxidize the Sulphur to sulfones, and then treatment of the resultant intermediate with base results in proton abstract (next to C-alpha, the other one), kick off the bromine to form epi-sulfone that cheletropically extrude sulfur dioxide to make cyclobutene!

    1. KazooChemist says:

      Look up the Ramberg-Backlund reaction I mentioned above.

  20. RealChemist says:

    It is shocking that no one even mentioned this work, which reported a exactly same reaction.
    The reagents used in the JOC paper are much cheaper and more readily available than the Nature paper.
    The authors clearly did a fantastic job hiding this work in the references.
    People need to a better job when reviewing manuscripts.

    1. Hap says:

      Scope’s a whole lot narrower, though — only bibenzyls are shown in the JOC paper. It needs to be referenced, if it isn’t, but it’s not the same – it might work on similar substrates, but it’s not shown (which makes me think it doesn’t).

      1. RealChemist says:

        So Nature is a Journal to demonstrate a broader substrate scope for an old method?
        Maybe that is true given the papers published by MacMillan or Baran.

        1. Hap says:

          The broader use is kind of a biggie, though – if you can use it on cyclic systems rather than just bibenzyls, it’s a significantly more useful. (Some of the examples in this paper are bibenzyls, and they seem like they can be made in other ways.) Also suspect the sulfonyl azide reagent is not friendly (possibly explosive) – it’s likely cheaper, but will likely limit use (though I can’t see this as a process method anyway?).

          A lot of the big methods (azide-alkyne cycloaddition, for example) were known but either had narrower scope or no good method of implementation. Having broad scope and implementation makes a method actually (rather than potentially) useful.

          1. RealChemist says:

            I do not understand what you mean by “only bibenzyls are shown in the JOC paper”. You probably should check Table 4 of the JOC paper. I am not surprised that you did not see that result, as that is what I mean by “People need to a better job when reviewing manuscripts.”

            Btw, in case you do not how to run the reactions in the JOC paper (I am sure you do not), all you need to do is to mix NaN3 and SO2Cl2 with the amines. No need to make the explosive sulfonyl azide.

          2. Fake Chemist says:

            @Realchemist –

            I think you can’t see the forest for the trees. A single step prep that’s broadly functional group tolerant isn’t just a competitive method, it lets you think about the net reaction in a totally different way. We could argue about whether anyone wants to isolate sulfamoyl azides, or about the particulars of the heteroarene tolerance, but at the end of the day this paper has captivated people because of the way it’s framed.

            Maybe Lu and coworkers could have called their method “nitrogen deletion” but they chose “Thermal Rearrangement of Sulfamoyl Azides” instead. Or maybe no one would have bought it because of their conditions.

            Organocatalysis as an idea flourished once it was named, even though Hajos-Parish among many others was already known.

    2. Derek Lowe says:

      They refer to it as another skeletal-editing method, but with undesirable features. And I think that’s right: sulfonyl azide is not a reagent that’s going to make any route popular, and that alone could explain why the earlier paper is not better known.

      1. RealChemist says:

        @ Fake Chemist

        I like your ID.

        It is so pathetic that nowadays publishing on Nature or Science is more about how to frame your work than how innovative the real work is.

        It is a shame to compare this work with organocatalysis. At least, early work of organocatalysis did not just simply expand the scope of some old reactions.

    3. Nick K says:

      It is remiss of the authors not to cite the following, definitive article by Dervan and others on 1,1-diazenes, their isolation and decomposition:

      1. Dr Dad says:

        They did cite this related paper from that series:

        With a journal like Nature you can’t cite everything you may want to, they probably decided to go with the isolation report.

  21. Barry says:

    kudos for the Hammett study. But more questions remain. E.g. can it make vicinal quaternary carbons?

  22. Oliver says:

    I really enjoyed reading the Levin paper, no doubt it has a great pitch and the advance is significant and interesting. This was before I noticed the earlier, brilliant JOC article – which essentially establishes the same mechanistic manifold and has an impressive substrate scope along with some preliminary mechanistic insights. Does the Levin paper deserve to be in Nature because it is better conveyed, and at its essence, only introduces a more practical and mild method? That is a discussion for another day. However what IS undisputed and disturbing is the manner in which the JOC paper was dubiously cited! The authors (deliberately?) relegate the JOC citation to a flippant statement that no one will even notice. It is painful that the introductory figure of the Nature paper mentions much weaker literature precedence but brazenly leaves out the most important JOC precedence. This is what makes people who are not in the ‘elite club’ rightfully cynical and suspicious of the quality of peer/editorial review happening at the ‘elite journals’. At the very least, the editor and reviewers could have ensured that the earlier work is discussed properly and put into context. Sadly, it increasingly appears that these ‘elite journals’, publication in which secures careers and fame, consider scholarship secondary to showmanship.

    1. Paul says:

      It’s definitely disturbing and disappointing to see that it appears they left out this crucial reference in the schemes but merely mentioned it in passing. Nevertheless this is an excellent paper but I would have changed execution. You don’t want to be known as a snake oil salesman.

  23. Albert says:

    I can imagine the poor JOC paper has been rejected by JACS, ANGEW, Chem. Sci. etc. and landed where it is, it’s common for people from a less-known institute and not a good seller of what they accomplish.

  24. BTDT says:

    Very nice work. I’m curious how similar the mechanim of this is to the Ramberg–Bäcklund reaction.

    1. Barry says:

      Extruding dinitrogen evokes the Barton thiadiazole “winch” reaction more than it does the Ramberg-Backlund. See last week’s discussions of each of these above. But this Levin reaction may be able to build vicinal quaternary carbons. Neither of the other extrusion reactions can do that.

  25. Louis says:

    Does anyone have the ref? The link to the paper is broken.

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.

This site uses Akismet to reduce spam. Learn how your comment data is processed.