Skip to Content

Chemical News

Diazos On Demand

I don’t think that anyone really likes diazomethane. Organic chemists like what it can do (cyclopropane formation, cycloadditions, fast, clean methyl ester formation, etc.), but the compound itself is very hard to have warm feelings for. It’s unstable on storage, and thus has to be prepared fresh. That preparation is fairly tedious, and you have to pay close attention to what you’re doing (unlike this guy), because the diazomethane product can and will decompose violently (and even explode) if you mishandle it. “Mishandle” includes, but is not limited to, making it too concentrated, exposing it to friction or rough surfaces, the presence of metals, etc. It’s also quite volatile and very toxic, explosions aside. That goes even for the less explosive form, trimethylsilyldiazomethane, which is stable enough to be an article of commerce but which is sadly just as deadly as the original item.

This would explain why you don’t see an awful lot of plain diazo compounds in the literature. The stabilized ones like ethyl diazoacetate are pretty common, but when’s the last time you heard about diazopropane? Or cyclopropyldiazomethane? Right, probably never, and that’s largely because the higher diazos, being less volatile, are no fun to prepare because you can’t distill them over in ether like you can the parent compound. By the time you’ve purified them, there’s not much left.

This new paper, though, could change that. A group at Montréal has adapted a solution-phase prep (silver oxide and potassium carbonate) to flow chemistry, which is the natural home for these kinds of compounds. (Update: see also this recent paper from the Ley group – another flow approach). Running hydrazones in dichloromethane through a column of these reagents (mixed with Celite) gives you diazo compounds out the other end. But you can tell that these are lively compounds – when they tried this on acetone hydrazone at room temperature, they got a colorless stream and plenty of gas bubbles, because the 2-diazopropane had already decomposed. Lowering the temperature to -20C and cranking up the flow rate, though, gave a red solution which is the real thing, ready to use.

They used this setup to produce a whole range of basically unknown diazo compounds (they’re yellow, orange, red, and purple in solution), and took them directly into esterification, cycloaddition, and cyclopropanation reactions. You’re not going to let these things sit around; unstabilized diazo formation is going to be an on-demand process no matter how it’s done. But it looks like these compounds are now open to experimentation in a way that they’ve never been before.

29 comments on “Diazos On Demand”

  1. A Nonny Mouse says:

    Link to the paper?

    1. Derek Lowe says:

      In there now. Dang it, I was sure I’d pasted that in, but something else came up first. Thanks. . .

  2. A Nonny Mouse says:

    I see in the references that the Ley group just beat them to it! Surprised that Victor is still working.

    1. Derek Lowe says:

      Added a link to that one as well.

      1. A Nonny Mouse says:

        By “still working” , I meant the UK universities’ policy of throwing people out at 65 (as they did to Barton!). Maybe this has changed.

  3. Jim Hartley says:

    From your first link: “That diazo group is looking for an excuse to revert back to nitrogen gas, which process comes with an inevitable no-substitutions side order of kaboom.” Love it.

    1. David Edwards says:

      If you want to see some truly stellar examples of Derek’s oeuvre, then the place to visit is “Things I Won’t Work With”. That entire section has a loyal fan base, one which has been clamouring for Derek to put the section in a book. For some reason, when you introduce Derek to substances with particularly hideous attributes, it ignites a literary spark in him that really does deserve a wider audience. 🙂

      The piece on Dioxygen Difluoride is a particularly good place to start, if you haven’t already seen it, though most people first encounter this section through the infamous “Sand Won’t Save You This Time” post, which probably has more hits than all his other posts combined.

      Though when he mentioned the incident with the non-standard Diazomethane still and the propane torch … that one should be marketed as a laxative. Along with another “How Not To Do It” incident involving testing for the presence of ether peroxide deposits, by shaking the bottle next to one’s ear. Chemistry is probably the big discipline where you do not want people bringing out their inner Darwin Award candidate as in these cases, because the results could involve nasty levels of collateral damage (read: innocent bystander fatalities). Which merely magnifies the respect one has for those people who work with perilous materials without undue mishaps, and continue doing so for numerous years. Luck tends to run out after a while, whereas diligence is more persistent. A view I suspect many here will concur with.

      Indeed, one thought arising from the new diazo preps in that paper, is that some of them are likely to be a lot worse in both the toxicity and shrapnel departments than the already troublesome Diazomethane. Which is why the flow chemistry methods will probably remain the preferred methods of pressing them into service, even if easier non-flow preps materialise in the future, because the flow methods allow consumption of the potentially evil reagent within moments of synthesis. Provided said flow methods yield much more benign products of course, but the mere fact that the authors cited a med-chem application means that this is the case there at least.

  4. David Edwards says:

    Interesting comment in that first paper:

    “Because of the diazoalkanes’ enhanced nucleophilicity, we envisaged that they could be good partners in Michael-induced ring-closure (MIRC) reactions with α,&#946-unsaturated carbonyls. This one-step transformation proceeded in good yields in the presence of various Michael acceptors, thus producing highly substituted cyclopropanes (Scheme 5). The gem-dimethyl cyclopropane motif, for instance,is of particular interest in the pharmaceutical industry, and versatile methods for preparation of this building block are lacking.[19]”

    NOTE: If my attempt to produce the Greek letters using HTML entities above fails, then that should read “alpha, beta-unsaturated carbonyls”.

    I suspect Derek will tell us more about the reasons gem-dimethyl-cyclopropane is being investigated in the pharma world elsewhere, but I can see why Derek’s eyebrows were raised upon encountering this paper. 🙂

    I’m also intrigued by the fact that these reagents are strongly coloured. Presumably, the diazo group has effects upon transmitted light similar to those observed in compounds with an -N=N- moiety integrated therein, a good number of which were synthesised as azo dyes in the past. If the hideous toxicity and explosion risks weren’t a hindrance in this matter, they would make fairly spectacular educational reagents, presuming that the reactions they’re used in involve conspicuous colour changes, though performing the reactions on camera will get around that problem in environments lacking the blast shields, etc. Any chemistry student worthy of the name should be jumping at the chance to watch something like this, especially when the commentary includes such details as “a number of hitherto difficult or impossible to synthesise highly substituted cyclopropanes, became amenable to synthesis in relatively mild conditions and in good yield, courtesy of this deep purple diazo reagent”! Makes me wish I’d chosen chemistry instead of software development in my younger days …

    Not to mention the cited “first nucleophilic addition on a difluorocyclopropene ever reported”.

    Who’s betting Derek is thinking of rigging this up in his own lab right now, with a range of med chem targets in sight? 🙂

    1. Hap says:

      Use someone else’s flow system first.

        1. A Nonny Mouse says:

          Phoenix were doing this for many years with no problems (though they used dizald for making it).

      1. Some idiot says:

        Classic!!! 🙂

        1. Tourettes of Chemistry says:

          Very interesting – Rh(I) diazo insertions among other things leap to mind as being enabled to a greater extent. Good stuff.

          It is noteworthy that flow to blow is just a three letter countdown…..

          [I]Caveat emptor[I]

  5. AVS-600 says:

    Aside from the isolation challenge, another difficulty with higher diazo compounds is that their larger carbon skeletons make all sorts of new decomposition pathways possible; in other words, they are usually less stable than diazomethane itself (!)

  6. Mark Thorson says:

    When I took organic chem, I was advised diazomethane was a powerful allergen. The professor said he had a grad student who had to switch to another major because she couldn’t set foot in the building because of the vapor in the air.

    By the way, the fuse on the captcha is way too short. I think it’s less than 10 seconds. I have to write a comment in a separate program, then rush to post. If there are many comments in a thread, I can’t post at all.

    1. Ted says:

      You can just refresh the captcha before you go to post – it shouldn’t erase any of your writing. Click on the icon just to the right of the equation (where the two arrows are sniffing each other’s behind).


      1. David Edwards says:

        A more serious problem is this … if you decide you want to rewrite part of your post, for some reason, the site locks out the ability to cut and paste text from the comment box once a certain amount of time has elapsed. Which means that if you want to edit and proofread your post thoroughly beforehand (I’m still cringing at some of the errors I let through in the past here), you have to write your post offline in a Notepad type application or other text editor first, then refresh the page, copy and paste your post from the text editor, then fill in the other details. All the usual text editing keys are locked out if you try editing directly in the comment box, after about 5 minutes has elapsed. Needless to say, I took that route to post this, just t be sure.

  7. Azetidine says:

    “Solution” is not a phase. The phase is liquid.

  8. Pitips says:

    Listen, when i started going hardcore on students, my life got easier. Plenty o’ work, none to question me. It frankly helped me build my chemical empire as it is today. Sure, the students are working at arbys today, but who cares. They were dumb, i did all the work.

  9. Dan DB says:

    Quick spelling note: Montréal not Montreál; the name is French, not Spanish 🙂

  10. Gatekeeper says:

    Flow chemistry is now the perfect niche for low-creativity academics. You’ll see more and more of these “idea- and concept-exhausted” professors taking shelter under the flow umbrella. Pathetic to say the least…

    1. A Nonny Mouse says:

      The Ley group in Cambridge has just been given a large grant from an Indian (his personal money he told me- the only billionaire that I know!) as the company that he was previously CEO/owner of has now adopted several of their published procedures with substantially increased yields and throughput.

    2. james says:

      Chemistry is both a science and a tool. No need to throw shade. Advancement of the science through new methods (flow) are inevitable and welcome.

  11. ScientistSailor says:

    I’m not afraid of diazomethane, just don’t distill it, make solution, decant, use. Nice trick for the other ones, though.

    I wonder if I could just pack a pipet or ISCO cartridge with the Ag2O mixture and flow through with air…

  12. Anonymous says:

    I’ve used quite a few diazos. My preferred precursor in N-nitrosomethyl urea (NMU) and the “freeze – decant” method of separation. The frozen aq phase is thawed and acidified as it still contains some CH2N2. I have stored excess CH2N2 over KOH in the fridge for a few days at most; then it gets destroyed. I think some early experimentals on cyclopropanations (by Doering, I think), read something like, “Prepare several small [10 or 25 mL] flasks of CH2N2 … as they have a tendency to explode before or during use.” I’ve also observed polymethylene (polyethylene) in some reactions; the FIRST time, I had no idea what was going on. Stockcock grease? (I almost never use grease.) ?? That’s why distillation is sometimes useful (depending on when / where the polymerization occurs).

    I never directly distilled CH2N2 but I have entrained it in a stream of N2 (using 18 gauge teflon tubing as cannulae) to bubble directly into a solution of reactant.

    I’m trying not to repeat items from the older (2008, 2009) discussions, but there is a JOC, 1980, describing a LARGE SCALE distillation of CH2N2. I think it’s this one, by Milos Hudlicky: I’ve always been comfortable with my own uses of CH2N2 but I would NOT want to do that distillation!

    1. David Edwards says:

      One of the papers citing that work you linked to, is a paper bearing the title Titration of Nonstabilized Diazoalkane Solutions by Fluorine NMR, by Victor L. Rendina and Jason S. Kingsbury, The Journal of Organic Chemistry, 2012 77 (2), 1181-1185. I suspect that paper contains much that would find its way into the TIWWW section here. 🙂

      Hmm, apparently involves using deuterated trichloromethane and 2-fluorobenzoic acid. The former will probably be a woefully expensive reagent.

      1. Anonymous says:

        “… deuterated trichloromethane and 2-fluorobenzoic acid. The former will probably be a woefully expensive reagent.” In small quantities, deuterated trichloromethane is around 15 cents per mL; at 1.5g/mL that’s ~10 cents per gram. Despite containing THREE chlorine atoms, chlorine being a toxic element, I’m willing to man up and use it.

        (I’m not sure if DE was being sarcastic or not.)

        1. David Edwards says:

          No, I wasn’t being sarcastic. My understanding is that a pure deuterated reagent is expensive, because it’s expensive to separate out deuterium in the first place. Unless there’s been a technological development that’s drastically reduced the cost last time I paid attention to the process.

          I was certainly expecting it to cost a lot more than $100 per kg.

          Mind you, wind the clock back, and there was a time when rapamycin was $1,000 per gram. A kilo was worth the same as eight Lamborghini Gallardos at that time. I gather it’s somewhat less expensive now, but one website I’ve just visited to check still quotes $1,440 for 5 grams. Still a lot of money to pay for a small quantity of substance.

          In the case of substances that require isotope separation, that isotope separation step, which in many cases is tedious and energy hungry, adds a lot to the cost. If you want pure calcium-48, for example, you’re looking at paying $500,000 for just 2 grams. Likewise, this website informs us that if you want some Californium-252, that’s $27 per microgram. One gram will set you back $27,000,000. Prices like that make gold look like dirt.

          With this sort of price background, I was expecting pure deuterated reagents to be pretty pricey.

  13. Anonymous says:

    Thanks, RE, for clarifying. Many readers here are organickers, but not all are. “Deuterated trichloromethane” is just deuterochloroform CDCl3, the most common NMR solvent used by organickers. We usually have a couple bottles of 99.8% CDCl3 on the shelf and it’s always on hand in the department or company stockroom. It is that common and important. D20 (deuterium oxide) is even cheaper (on a molar basis) and used as an NMR solvent for water soluble compounds (proteins and other biologicals). (And CDCl3 is made from D20.)

    In the paper you cited, CDCl3 is being used as the reaction solvent, not as a reagent. Many of us will run reactions in ~0.7 mL of CDCl3 in an NMR tube to easily follow what’s going on (or run a larger reaction in 1-5 mL of CDCl3 in a flask or vial and take aliquots for direct NMR analysis).

    D20 is produced on an industrial scale because it is needed in the nuclear energy industry. Compared to many other isotopes, deuterium is dirt cheap.

Leave a Reply

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

Time limit is exhausted. Please reload CAPTCHA.