Org Prep Daily

January 30, 2012

Potassium hydride self-ignition

Filed under: lab destruction — milkshake @ 1:40 pm

I had a rather bad fire last Friday. I was washing a large jacketed glass reaction vessel used for polymer scale-ups, after pouring the reaction mixture out, and a tiny particle of potassium hydride (from this poorly quenched reaction) that was adhering to the bottom of the reaction flask ignited just as I was giving the flask a proper acetone rinse. So I had a flaming flask in my hands + burning hands + flaming sink in front + a whole bunch of wash bottles ablaze next to me (plastic wash bottles peeing their burning solvents around…) A colleague promptly put the fire out with a mid-sized CO2 fire extinguisher before the flames spread any further. There was no damage to the lab, my fingers or the reaction mixture but it was a pretty scary situation – considering how fires in organic labs can get out of control so fast.

Potassium hydride pyrophoric nature is well documented in the literature; from my limited experience I would say KH is quite comparable to potassium metal in its tendency to flame up. But there are some aspects that make KH more treacherous than K metal: KH in paraffin or mineral oil is docile and only when the oil or wax is washed off the pyrophoric nature becomes apparent. Also, the KH appearance (a grayish-white powder) is less dramatic than shiny low-melting globules of K metal and one cannot easily guess whether KH is fully consumed or quenched by the sediment appearance if the reaction produces inorganic precipitate of its own. Also, I noticed that some alcohols react with KH in THF surprisingly sluggishly while reaction of other alcohols is prompt – I believe the solubility of the K-alkoxide in THF plays a role and the KH particles may get coated by a poorly soluble material and laze about the bottom – and then at some later point flame up when least expected.

Since K-alkoxides have significant reactivity advantages over Na and Li alkoxides in alkylation reactions[2], and since the easy-to-handle KH formulation in paraffin wax is now commercially available, it is likely that KH will get used increasingly more often in place of NaH. Despite its innocuous appearance KH is less tame than NaH;  having unreacted KH excess present in the reaction mix makes it prone to auto-ignition during the workup if the reaction was not quenched with care.

Note 1: I was impressed how good is CO2 extinguisher for large solvent fires – and it leaves no mess behind. I don’t think a dry powder extinguisher would have worked nearly as well.

Note 2: Taber et. al.: Tet. Letters 51 (2010), 3545-6


  1. Are you still shocked? I don’t think I’ve ever seen you type accented like this before.

    Comment by Sili — January 30, 2012 @ 4:06 pm

    • Well when this screwup happened I was setting up a filtration of the poorly quenched reaction mixture precipitated with ether, with the unreacted KH in an open plastic bucket with 10 liters of ether/THF mix in a semi-opened hood right next to the flaming sink… That bucket of precipitation mix luckily did not catch fire or I would not have the hood, and possibly the lab to work in today. Afterwards we dumped lots of isopropanol and ethanol into that bucket of KH/ether mix, put a lid on and N2 needle through to blanket it and we let it sit over weekend, to make sure that one wouldn’t self-ignite on us during the filtration.

      And all this fun to be had with just 300mg of KH.

      Comment by milkshake — January 30, 2012 @ 4:34 pm

      • Uh… 300mg does not equal a tiny little particle. This story sounds fishy.

        Comment by vanitylicenseplate — January 30, 2012 @ 5:23 pm

        • 300mg of KH was the load in the entire react mixture in a 5L flask… I added about 0.65g of 50% paraffin-KH to the reaction mix and some of it probably got consumed. What was adhering to the bottom must have been just a tiny speck. The paraffin-KH tends to be a bit more granular and slower to go in than the mineral oil 30% KH suspension

          Comment by milkshake — January 30, 2012 @ 5:41 pm

          • Would you recommend the Taber-style paraffin-KH, then? I guess it’s pretty important to wash away the paraffin when it’s safely in the reaction flask under inert gas? The only lab fire I had during my PhD happened exactly the same way as yours – tiny bit of KH in poorly quenched flask I was cleaning out set off the acetone in/around our sink. I’m sure the reaction only had 100mg KH or so (30% in oil)… it doesn’t take much.

            Comment by BRSM — February 26, 2012 @ 9:15 am

          • I have been using Taber’s KH/wax sold by Aldrich and liked it very much (despite the fire). I believe it is prepared from H2 and K metal in a pressure reactor, using molten paraffin as a solvent. I am emphatically against washing off the paraffin or oil from KH, I would rather deal with little wax present in my reaction mix than work with naked KH powder. The waxed stuff is very nice and you dont need to weigh out or store KH/wax under nitrogen, just common dessicator for storage and weighting out into a test tube to be dumped into a dried flask in one go… The techniques similar that you would use for LAH or NaH. The only difference is in quench – you have to be absolutely sure it is gone and decomposed before you pour the reaction mix out from a 3-necked flask.

            Comment by milkshake — February 26, 2012 @ 5:20 pm

    • Well, I dont want to be an A** now to give you a lecture on how to avoid and all. It was a bad accident and it just happens. However, I would strongly recommend now though, especially if you are working in industry or so, that before you clean any of such reactors, usually one should go ahead and wash off all the material using the solvent used for the reaction and then only take the assembly out (Inspite if somebody has washed it once it wouldn’t harm doing it again..!!! and try to use the solvent like ethyl acetate to quench hydrides as they are non inflammable and would take the hydrides off the reactors. I pray it wont happen again but one need to be prepared for such things again…(I am saying it from my experience here) DO’NOT forget to thank the person who used the FIRE extinguisher lol…!!! it was his courage that saved you and lab too….!!! 😉

      Comment by IronMan — July 30, 2012 @ 6:15 pm

  2. Sounds like a close one. Glad you’re safe!

    Comment by Arjun — January 30, 2012 @ 4:21 pm

  3. I was recently exposed to the dangers of dry KH. Worked previously with very colleagues who routinely rinsed their NaH and KH with hexanes (or pentane) prior to use. The procedure was performed in a glove box. I had not *seen* them do it however: this made the difference.

    Took about 100 mg of the KH suspension in mineral oil in the glove box, rinsed with hexanes and decanted; weighed out the dry KH in the glove box. Placed on weighing paper, intending to take it out of the box and dump directly in the flask. It lasted about 10 seconds in the atmosphere as a dry powder before bursting into flame. Thankfully the fire was contained to the glovebox chamber and no damage was done. Moral of the story was to charge the flask in the glove box and not expose dry KH to the atmosphere.

    Comment by James — January 30, 2012 @ 11:27 pm

    • Uh, what did you expect?

      Comment by Mike — January 31, 2012 @ 2:11 pm

      • Well, I was expecting to have a little more than 10 seconds leeway; I suppose I was expecting something more like LAH, which is also pyrophoric but on a longer time scale.

        Comment by James — January 31, 2012 @ 9:57 pm

        • KH is nothing like LAH. The fact that it is packaged in the same fashion as alkali metals should have been a dead giveaway.

          Comment by Mike — February 1, 2012 @ 6:32 pm

  4. Hi Milkshake,

    Glad to hear that the accident had no serious consequences (maybe S.t Barbara, protecting chemists and fireworks manufacturers was looking over you 🙂 ). Maybe overhead stirring during the quench will be of help in eliminating the risk of further accidents like this one, and now the urge to find an alternative to diethyl ether should be clear…
    My best wishes of a rapid resolution of the safety problems in this synthesis.

    Comment by processchemist — January 31, 2012 @ 5:03 am

    • we definitely need to find alternative solvent(s) for precipitating our polymers and to figure out a convenient way of drying them on scale, and it will be tricky. We can’t bring diethylether-based workup to a GMP process facility…

      Comment by milkshake — January 31, 2012 @ 12:08 pm

      • I always tried acetone and ethanol for polymer ppt first…obvious chemist is obvious, but that’s my first go-to.

        Comment by opsomath — January 31, 2012 @ 3:37 pm

        • these polymers are very hydrophilic; ether works best unfortunately. One can use some other ether solvents but then the precipitate either comes out muddy and it is hard to filter, or the precipitate is too soluble and hygroscopic, turning soggy on a Buchner funnel and so on. Possibly one can use isopropanol but then the Buchner has to be chilled and the filtration setup has to be done under positive pressure (to deal with the high viscosity of 2-PrOH and to keep the moisture out) and drying a gell-like product precipitate soaked with 2-propanol takes a long time…

          Comment by milkshake — January 31, 2012 @ 4:04 pm

          • I’ve been making a lot of dextran derivatives lately. Ugh. Lyophilization is the only way to go for the really hydrophilic stuff.

            Comment by opsomath — January 31, 2012 @ 4:14 pm

          • this is quite interesting – by chance have you done any phosphate cross-linking on those dextrans, in water, with cyclic sodium trimethaphosphate as a cross-linker?

            Comment by milkshake — January 31, 2012 @ 4:48 pm

          • No, I haven’t, but that is a good idea for something else I’m working on. 🙂 I am mainly carboxylating and adding other functional groups off of those.

            Comment by opsomath — January 31, 2012 @ 5:02 pm

          • I have seen papers about starch cross-linking with Na3-trimetaphosphate (in alkaline water, with added Ca2+ as a catalyst) and I thought this would be a mild and bio-cleavable way to do it for other systems too.

            Comment by milkshake — January 31, 2012 @ 5:10 pm

          • You can try DOE, with some binary mixture of solvents (two levels of composition)… in case of precipitation/crystallization this method can suggest unimmaginable conditions that can give good results.

            Comment by processchemist — January 31, 2012 @ 6:07 pm

          • By my experience in handling PEG, we often disslove PEG in dichloromethane then drop-wise add to MTBE, the ratio of DCM and MTBE is important to afford easy filtration precipitates and decent yield.

            Comment by Barry — February 29, 2012 @ 1:28 pm

          • so what ratio worked best for you – and on what scale have you been doing PEG chemistry?

            Comment by milkshake — February 29, 2012 @ 3:49 pm

          • The ration of PEG/DCM/MTBE (1g/5ml/100ml) often works well for me. We did 100 grams scale in lab, and used equipments like schlenk filter to filter precipitates under nitrogen.

            Comment by Barry — March 1, 2012 @ 1:58 pm

  5. Pretending that the polymer is unknown (I have experience with the monomer and catalyst system) you can try hexane (+/- a carbon penatne or heptane). The precipitations work very well in those non-solvents. If its GMP you need try using nito-benzene.

    Comment by Curryworks — February 1, 2012 @ 12:02 pm

  6. Milk Shake: Glad all ended peacefully! My question was why diethyl ether was preferred and not MTBE ot Di-isopropyl ether? No one can change the Murphy’s law, but my point was the later two solvents have low vapour pressure in comrarison to ether. Just a thought…

    Comment by Sham — February 1, 2012 @ 12:47 pm

    • iPr2O is not actually very process friendly because the peroxides form in it fast and are known to precipitate out/evaporate to generate a contact-explosive solid residue in the drum whereas peroxides from Et2O are oily. (If you submit a process paper with iPr2O to Org. Process. R&D journal you can expect the editor throwing a hissing fit about iPr2O.) We employ tBuOMe on scale but the physical form of the polymer precipitated from it is not nearly as lovely as from Et2O and the drying takes longer. And the solvent reeks of mint and camphor salve…

      Comment by milkshake — February 1, 2012 @ 1:10 pm

  7. I was always under the impression that these fires were most often caused by potassium metal in the KH, a common contaminant.

    Comment by synthon — February 14, 2012 @ 1:46 pm

    • I don’t know, the Aldrich stuff in 50% paraffin wax is pretty whitish when you wash it (rather than the more common gray) so if there is some K in it there is not much of it. The other reason why I think there is not much K is that I used KH in the presence of some easy-to-reduce groups that are definitely ruined with K metal, an they survived just fine

      Comment by milkshake — February 14, 2012 @ 3:20 pm

      • Fair enough. My experience was with the material in mineral oil. I currently use this material as the first step in a really nice Claisen dimerization and have had no problems to date.

        Comment by synthon — February 14, 2012 @ 3:50 pm

        • But, by all means, never grow overconfident. That’s when shite happens! Happy Claisens!

          Comment by HPCC — February 14, 2012 @ 11:21 pm

  8. The real beast is NaK alloy! That stuff scares the crap out of me. Hey Milkshake, this is off topic, but I was wondering if you knew any procedures for selective o-acylation of phenols in the presence of alcohols (for example, para-hydroxyphenethyl alcohol). Regular acyl chlorides (eg benzoyl chloride) give me a mixture of products w/pyridine or Hunig’s base. I had success with diphenylcarbamoyl chloride in neat pyridine, which only acylated the phenolic position, but it wasn’t suitable for my purposes later in the synthetic scheme. Would be honored if you responded! Cheers

    Comment by NorthwesternChemist — February 16, 2012 @ 8:37 pm

    • I would maybe try DCC coupling – mixing the carboxylic acid with the phenol in acetonitrile at room temp, without any base or catalyst, then throwing in DCC solid (1 equivalent), stirring for a bit and then filtering off the dicyclohexylurea sideproduct few hours later and evaporating the filtrates.

      K-Na: So far I did not have the pleasure but our polymer bosses who are not afraid of working with potassium mirror shuddered and crossed themselves when I proposed K-Na alloy to them (to use it in place of K metal). I also heard of people who managed to burn down their hood to ashes with just few grams the stuff.

      Comment by milkshake — February 16, 2012 @ 8:52 pm

      • Hi milkshake

        Thanks for responding! I’ve done DCC couplings in the past, but with 10-20 mol % DMAP. I recall reading that DMAP abolishes the chemoselectivity. I will try it with just DCC.

        Cheers! keep up the great blog!

        Comment by NorthwesternChemist — February 17, 2012 @ 1:56 am

        • Yeah, DMAP is great if you specifically want to esterify an alcohol. Thank you for reading!

          Comment by milkshake — February 17, 2012 @ 10:18 am

  9. Hey milkshake, completely off topic, but how stable is tosyl azide? I’ve seen several preps that prepare it on >10g scale and isolate the neat product without a hitch. However I’m wondering if there’s more than meets the eye in those preps (one of them used DCM for extraction, I shudder at that thought). The stuff is commercially available as a 15% solution in toluene, so I guess it can’t be TOO jumpy…

    Comment by student — February 18, 2012 @ 2:52 am

    • I never worked with this reagent myself, people often make it fresh before the diazo transfer and store it in fridge. It is not supposed to be too bad but for a very large scale I would maybe use a less sensitive analogue, if I remember correctly p-acetamidosulfonyl azide is meant to be safer. Btw, you should avoid chlorinated solvents when working with azide salts.

      Comment by milkshake — February 18, 2012 @ 3:54 am

      • I have never worked with tosyl azide myself, but for one it used to be available commercially (Aldrich says it is discontinued, but we had a bottle of it, I believe in the freezer without argon) and for another it was used in our workgroup before with no problems that I remember, so it should not be that bad to handle. As for the preparation of it however, I heard a story that happened a few years before I joined my workgroup, where 15 g of tosyl azide were prepared which exploded in the rotavap. It might be that it wasn’t watched too closely and/or a mistake was made during evaporation, but personally I would be very cautious if I had to prepare that stuff in anything over a 1 gram scale. Fortunately the person who made the azide got out of it with no injuries, but that was just a lot of luck that she wasn’t that closeby when it happened.

        Comment by Xing — February 25, 2012 @ 8:15 pm

        • azides blowing up on rotovap – this is quite typical mishap when someone makes the mistake of combining dichloromethane with sodium azide and produces diazidomethane: It goes boom when you concentrate it down

          Comment by milkshake — February 25, 2012 @ 9:57 pm

  10. Yeah definitely avoiding the prep that uses DCM (and they put it on the rotavap too! not sure how they managed to keep their fingers…). I’m actually going to use this in click chemistry rather than as a diazo transfer reagent. But yes you’re right, probably best to prepare it fresh and use it right away. The excess stuff I’ll probably either quench or store under Ar in the freezer. Btw, would the Staudinger reduction be a good way to quench these reactive azides, or is that a bad way to go about it?

    Comment by student — February 18, 2012 @ 2:24 pm

    • Fokin from Scripps has done lots of click chemistry with sulfonyl azides, and the produced sulfonyltriazoles do interesting ring-opening reaction to diazo compounds that could be used for Rh(II)-catalyzed cyclopropanations…
      Phosphine sounds reasonable if you need mild quench conditions. Otherwise (if your product can tolerate it) I would use just excess of ammonia.

      Comment by milkshake — February 18, 2012 @ 10:05 pm

      • Are there any refs to methods for destryong tosyl azide?
        I have some in the fridge and reading this makes me think its best if I dispose of it before anyone comes across it.

        Comment by FleaTamer — February 28, 2012 @ 6:33 am

        • I would just add it into waste drum with non-halogenated solvents. It is gong to be OK in solution as long as there are no heavy metals or base+halocarbons

          Comment by milkshake — February 28, 2012 @ 11:24 am

  11. Btw, Klapötke et al. recently published a study about the stability of imidazolesulfonyl azides, doi: 10.1021/jo202264r

    Comment by JH — February 23, 2012 @ 8:28 pm

  12. I am doing some nucleotide chemistry for the first time and trying to find the best way to purify these compounds when I have hundreds of mgs of material. One issue is that these compounds are so water soluble there is no way to remove salts by extraction. RP-HPLC isn’t so economical in this case either. I see references in the literature to activated charcoal chromatography? Anyone have experience with this or any other ideas? Would be much appreciated.

    Comment by PotStirrer — March 7, 2012 @ 4:04 am

    • Are these compounds oligos – or just small molecule nucleotides? Because if the Mw is high enough you could do dialysis or better yet hollow fiber ultrafiltration (we use ultrafiltration for de-salting and cleaning up some very polar charged polymers in water and it works like a charm).

      By the way, HPLC is not really that expensive if you have it in house and run it with methanol-water (and you get the benefit of UV detector and fraction collector). But for crude de-salting you can also absorb your stuff on a short column of C18 silica, wash the salts out in straight water, then elute your material off the column with water+tert-butanol 7:3 and lyophilize – this water-tBuOH mix is a bit viscous for regular HPLC use but it works really well with lyo

      Comment by milkshake — March 7, 2012 @ 5:07 am

  13. Thanks Milkshake. These are small molecule single nucleotides, so ultrafiltration doesn’t sound like an option. It’s true that MeOH-H2O on an HPLC isn’t so expensive, but can be time consuming with large quantities of material. I was looking for some sort of relatively quick method that could be done in a single run. I am curious about the H2O-tBuOH mixture. Why is it superior for lyophilization?

    Comment by PotStirrer — March 7, 2012 @ 11:23 am

    • the tBuOH-water mix freezes up really nicely and it does not mess up lyo vacuum – no need to concentrate the solution on rotovap before lyophilization. It also has a decent dissolving power for many organic molecules and it has low toxicity, we are using it for drug formulations

      Comment by milkshake — March 7, 2012 @ 12:01 pm

  14. I hope u r safe and recovering now. wish u a speedy recovery.

    Comment by pashu — March 8, 2012 @ 12:17 pm

  15. Hi Milkshake,
    i would like to ask you a question.
    i’m running Suzuky coupling between aryl iodides and different boronic acids, but on a solid support (quite particular).
    Is there any way to suppress proto-deiodination?The best conditions for me are Pd2(dba)3, DMF-H2O 4:1, and Na2CO3, 50-80 C.
    Many thanks as always.

    Comment by madforit — March 15, 2012 @ 2:42 pm

    • de-iodination can be sometimes a difficult problem to minimize, depending on the substrate structure – I once side-stepped this Pd-promoted de-iodination by using the corresponding bromo compound instead (the bromo reaction was slightly slower but cleaner) . I wonder if you can go to anhydrous conditions with using cesium carbonate as a base. I would try anhydrous DMF and I would also try peroxide-free anhydrous THF as a solvent with Cs2CO3 (the Cs carbonate will not be fully dissolved in the mix but this should not be a problem.) You may need to do a phosphine ligand screen starting with Pd(PPh3)4 (then dppe, dppp, dppb, dppf, xanphos…), I remember that dehalogenation and homo-aryl coupling sideproduct formation tends to be very dependent on the used phosphine ligand. Also using Molanders Aryl trifluoroborato salts (made from boronic acid in one step: ArB(OH)2 + KHF2 => K(+) ArBF3(-) ) might be worth trying, I think some of them are commercially available. Sorry that I cannot give you a better answer.

      Comment by milkshake — March 15, 2012 @ 7:26 pm

  16. Thank you for your exhaustive reply, no need to sorry, it’s more than enough.
    The main problem is that the environment it’s quite strange (resin bound long polypeptide chain) so i don’t know where to start, i’m gonna try just with DMF, Pd(PPh3)4, Cs2CO3 an see if something happen.
    Thanks again.

    Comment by madforit — March 16, 2012 @ 11:43 am

  17. Hi Milkshake/everyone. We plan to do a Noyori asymmetric transfer hydrogenation of an ynone. It is a published procedure (SynLett 2009, 17, 2801-2802 and Chem. Eur. J. 2011, 17, 6964-6972). Both procedures and others that I have read describe mixing the active Noyori catalyst with the ynone. I assume this means they prepared the catalyst in a separate step beforehand and that in situ activation of the precatalyst could result in loss of enantiopurity. I also assume that the catalyst is quite sensitive to air. Are we out of luck if we don’t have a glove box or can this be prepared in the hood? I have no experience in this area. BTW, I did see your 2006 post describing a Noyori transfer hydrogenation, Milkshake, but figure we should just follow the literature procedure on this one. Thanks!

    Comment by PotStirrer — March 30, 2012 @ 4:48 am

    • From my (limited) experience using Noyori Ru-Cymene + TsDPEN ligand catalyst system you won’t need a glovebox because it is fairly robust – the catalyst oxidizes relatively slowly and is not sensitive to moisture. The precatalyst and ligand you can weight out in air, when you combine them in solution you would use about the same techniques like working with Pd(0) tetrakis – just with minimum precaution that a medicinal chemist wold take (sparging solvents with a needle stucked through septa, with Ar balloon, on ultrasonic bath, using reasonably purified solvents) and you should be able to do this in the hood. But I would definitely pre-form the active catalyst complex. I have done it by mixing the RuCl2-cymene precatalyst with 2 equivs of TsDPEN ligand in dichloromethane and heating it for a while to 40C (the mixture turns from dark brown to light tan-yellow) but since just one amine ligand is needed for Ru and the other one works only as a base to mop up liberated HCl equivalent, you can also use just 1 equiv of TsDPEN and some other base.

      Comment by milkshake — March 30, 2012 @ 11:31 am

  18. Indeed, pre-formation of the catalyst is safer. I made the Ru(p-cymene)(Ts-DPEN) by heating the Ru precursor with 1.02 equiv Ts-DPEN and 2.0 equiv Et3N in refluxing iPrOH for 45 minutes under N2 atmosphere (about 0.5 M). The heterogeneous suspension becomes a solution upon heating and eventually starts spitting out the desired catalyst. Cool to 0 C, filter, wash with minimal amounts of cold iPrOH, and you should get 85+% yield of Noyori’s catalyst. Worked wonders in our case to reduce a Bischler-Napieralski imine to the beta-carboline in 98% ee using Et3N and formic acid as reductant in DMF (r.t.)

    Comment by HPCC — April 2, 2012 @ 1:08 pm

  19. Hi, I’m contacting you because I read your comment in another blog about Ninhydrin Stain. I’ve tested the cadmium spray after dipping my plate in ninhydrin, and the colour becomes really stronger. What is the principle of this reaction? Do you have the article reference about this? Thank you very much!

    Comment by Aline — April 21, 2012 @ 10:53 pm

    • I don’t really know. I only presume that the formed blue dye, which is supposedly a mixture of ninhydrine imines (the main one being a ninhydrine imine of 2-amino-1,3-indadione) has good metal coordinating properties and coordination to metals could stabilize developed ninhydrine stain from fading, or provide even more colorful stain. I remember vaguely that I was reading a book with stain recepies going back to old-fashioned paper chromatography and they had TLC detection of aminoacids with ninhydrine followed by a metal solution spray (apoart from cadmium they also used copper if i remember correctly) but I don’t have any reference to give you. Why don’t you spend some time on Google or find some books in library

      Comment by milkshake — April 22, 2012 @ 1:38 pm

  20. Hey Milkshake, quick question about methyl triflate. Just how dangerous is this methylating agent? I know it’s supposed to be something like 10^4 times more reactive than methyl iodide, which I suppose puts it in the same league as TMS-diazomethane. Sucks that I’m going to be working with large quantities too since I need to run a few large scale reactions.

    On an unrelated note, any experience making unsymmetrical sulfonylureas? There’s a JOC paper (2003, 68, 115-119), but I wonder if it will work with indoles and diphenylamines.

    Comment by student — May 2, 2012 @ 12:16 am

    • I only made very few sulfuryl diamides myself, and I noticed that reagents like Et2NSO2Cl do not give clean reactions with amines. I think the method you reference is much better and more versatile.

      Methyl triflate is probably one of nastiest reagents that you encounter in the lab: It is a very hard electrophile which makes it far more mutagenic than stuff like methyl iodide. Dimethyl sulfate is a famously nasty reagent and I expect methyl triflate to be worse because it has much lower boiling point. MeOSO2F is even lower boiling and that reagent (“magic methyl”) put several chemists on life support because of the lung edema. So I suggest you work with it very carefully, double-glove, move your rotovap into the hood and decontaminate everything at the end with a mix of concentrated ammonia with alcohol or acetone. I worked with Me-triflate on large scale several times and was really afraid of it.

      Since it looks like you will be using Me-triflate repeatedly, and the genotoxicity of alkylating agents is a cumulative exposure problem, maybe you should look into alternative strong methylating agents. Me3O+ BF4- is commercially available and it is a stable a solid, and it is quite inexpensive. So I would definitely look into using trimethyl oxonium as the imidazole activating agent. There is also the corresponding triethyloxonium tetrafluoroborate, I made it myself by an OrgSyn procedure (from Et2O, BF3.Et2O and epichlorohydrine) but triethyloxonium (unlike the trimethyl oxonium) is very hygroscopic and does not store well – it usually decomposes after few weeks (both as a solid and a solution) whereas trimethyloxonium storability is reportedly very good. And by the way, oxonium salts and triflate esters are not compatible with THF – they polymerize it. The preferred solvent is DCM

      Comment by milkshake — May 2, 2012 @ 12:41 pm

      • Thanks so much for the safety warnings milkshake, I’ll definitely be respectful of methyl triflate. An unrelated question if you don’t mind — I’ll be doing a large scale(ish) Suzuki coupling. There’s no way around a column, but that I’m not too concerned about since I’ve run this reaction at least 7 times. The issue is, I always have a lot of leftover boronic acid since I use 2 equivalents, and that probably makes up 30% of the crude weight. If I could wash clean the boronic acid I’d definitely save a lot solvent (and time) with the column. Any good tricks for washing out the boronic acid? I run the reaction in DME and my product is very insoluble in water, so that won’t be an issue. Thanks again.

        Comment by student — May 2, 2012 @ 7:20 pm

        • I don’t have any experience with removing boronic acids from the mix, I suppose since your reaction mix after Suzuki is alkaline leftover boronic acid should partition into aqueous phase, especially if you use a less polar solvent like ether for the extraction. Boronic acids complex with polyols so maybe you can add some Rochelle salt or sucrose into the aqueous wash

          Comment by milkshake — May 3, 2012 @ 2:23 pm

  21. Hi Milkshake, I need to use tributyltin hydride for a palladium catalyzed reduction. Just wondering what the proper procedure for quenching the tributyltin residues are, since the paper doesn’t mention anything about it. Thanks.

    Comment by student — May 20, 2012 @ 11:42 pm

  22. Hi, Could you refer to some scalable procedure for Suzuki coupling of heteroaryl boronic acids with aryl bromides using normal Pd Catalysts ( not one of those exotic POPd s that combiphos sells)?


    Comment by Christiaan — May 30, 2012 @ 8:04 pm

    • I have had fairly good luck with plain Pd(PPh3)4 and K2CO3 in refluxing dioxane-water mixture. The only stipulation is that dioxane should be free of peroxides and Pd catalyst unoxidized, ( = bright lemon yellow, from a good source like Strem, stored under Ar in a freezer). An example is here:
      ( I think it was not optimized for low Pd loading, I weighed out quite a bit and it worked well on first try).

      There are several alternate variants, depending on what is troubling your Suzuki. For example some pyridineboronic acids love to hydrolytically de-borylate to parent aromatic very fast, some other Suzukis suffer from parasitic dehalogenation of aryl iodide (accompanied by biaryl formation from boronic acid). If possible I would go with aryl bromide and maybe pinacolate ester (or the ArBF3(-)K(+) salt from Molander work) which has the advantage of generating a steady low concentration of boronic acid gradually, in situ which limits some of the sidereactions.
      if one of the substrates is base sensitive, you can substitute aqueous K2CO3 for KHCO3 (or NaHCO3) in a closed pressure flask – Suzuki typically still works with bicarbonate but is somewhat slower. Or you can use Cs2CO3 with THF with a more active catalyst like S-Phos /Pd2(dba)3 combo, that one I remember using at 40C (but Buchwald-type hindered phosphine frequently fail with aryl iodides and strongly coordinating substrates – in that case Dppf/Pd2(dba)3 would be probably a better choice). Also I run once a strictly anhydrous Suzuki succesfully, as my heteroaryl chloride was moisture sensitive and rapidly crapped up under typical conditions, I did it with Pd (PPh3)4 in dioxane and with dried K2CO3 and added 4A activated sieves, but the reaction required a microwave heating to 130C in this anhydrous system)

      Comment by milkshake — May 30, 2012 @ 9:42 pm

      • I personally really loved the Pd(OAc)2 / PCy3 or PtBu3 system developed by Greg Fu, for room-temperature couplling of (even sterically-hindered) aryl bromides and triflates with boronic acids. Needs anhydrous THF, 1 mol% Pd, and about 5 equiv flame-dried KF. Be careful though, when adding KF, the whole thing is extremely exothermic!! I would recommend adding the KF portionwise with 15 ºC water bath cooling. Fu developed these couplings at 1.0 M concentration, but I used 0.4 M dilution, and that also helped take care of the tremendous heat kick.

        Sorry I don’t have a reference anymore since I’m not working in a lab right now.

        Comment by HPCC — June 5, 2012 @ 1:37 pm

  23. Thanks. I will try and update the outcome.

    I wonder if you have a cheaper and reliable source for HRMS service. I just do not want to go through the pain of getting the elemental analysis on my greasy stuff.

    Thanks in advance.

    Comment by Christiaan — June 7, 2012 @ 6:44 pm

    • Please suggest any simple catalyst(s) (chiral/achiral) for the intermolecular sttetter reaction of aldehydes to unsaturated esters. I am either getting the xotic triazole based ones or dreadful cyanides…… there any thing simpler to get this reaction to work on multiple grams would be appreciated. Thanks, again.

      Comment by Christiaan — June 7, 2012 @ 9:54 pm

      • I have no experience with Stetter reaction, never run it myself.
        Cyanides are not a problem – they are easy to work with, I run reactions using 10+ grams of NaCN or TMSCN without issue. You just don’t pour the solution on your hands (NaCN aqueous solutions slowly permeate skin so if you spill some on you you need to remove the contaminated clothes) and you do not acidify a cyanide reaction obviously. If you need to destroy cyanides before pouring it to waste, some Fe(II) salt plus ammonia or bicarbonate produces a slurry thats pretty good for binding cyanides, and you just let it sit in a beaker overnight. Alternatively CuSO4 + calcium hydroxide (the copper(II) mix will release some dicyane gas (CN)2 so you need to do it in the hood)

        Comment by milkshake — June 8, 2012 @ 11:00 am

  24. Hi milkshake: how nasty are aromatic nitriles? MSDS warns of eye irritation, but seems to be generic as a class of compounds…..any experience?
    Looking for a commercially available , cheap (<$3/gram), solid @ rt, soluble in diethyl ether not-too-toxic nitrile with no phenol, carboxylate or amine groups (i.e. remains in organic phase when ether soln extracted w/acid or base).
    4-bromo benzonitrile looks good, but those eye irritant warnings worry me….

    Comment by psl — June 12, 2012 @ 1:17 pm

    • PS: this is for an undergrad lab on the ~0.5g scale, so nasties common to you & I are more problematic for them.

      Comment by psl — June 12, 2012 @ 1:18 pm

      • I dont have experience with designing freshman lab experiments, irritant though is a very low-level hazard – almost any chemical including common solvents will have it. Quickly looking on Aldrich website, I noticed isophthalnitrile (1,3-dicyanobenzene) is a room-temp solid and very cheap

        Comment by milkshake — June 12, 2012 @ 5:34 pm

  25. Any suggestions on how to get any C-C bond formation using the ketone of beta-ketoester? I am struggling with Wittig, reformatsky, …..etc presumably due to the keto-enol tautomerization. How to keep this in ketone form and get your reactions to happen at the ketone center.

    Comment by A1 — June 20, 2012 @ 10:50 pm

    • generally speaking, you can’t (the enol form is pretty acidic). A reasonable synthetic alternative in many cases is to make enol triflate and employ it as electrophile in Negishi, Suzuki, Sonogshira, Stille, Heck etc. There is a recent review in Tetrahedron, ASAP:

      Comment by milkshake — June 21, 2012 @ 12:07 am

  26. Thanks for pointing to that important review. I will check. How ever, mine is a benzylic C-( nucleophile) and the C=O of the keto ester, which should give = bond conjugated to aromatic ring…not to that of ester side. I did not know that there exists protecting groups for active methylenes……ofcourse, we will have to do a desulforization of thiketal later on….I do not have much options left as I do not want to protect the methylene with X2 halogens…….

    Comment by A1 — June 22, 2012 @ 12:23 am

    • I am sorry but your synthesis scheme sounds half-baked. (The fixes that you are proposing are not going to work.) Try to join a different project and find a more competent research advisor.

      Comment by milkshake — June 22, 2012 @ 12:54 am

  27. I am struck with it for now as I want MS before I apply for PhD programs of different schools. I brought this problem on to me!

    Comment by A1 — June 24, 2012 @ 6:02 pm

    • if you are really working on this in the lab and you are trying to finish it so that you can get a degree, probably the best option would be to find someone who gives total synthesis classes, sit in on his classes and then approach him, explain you situation and ask him to fix your proposal. Then spend some time on Scifinder trying to find the closest possible methodology precedents for each substrate that you are using, and so on.

      Comment by milkshake — June 24, 2012 @ 6:26 pm

  28. Dear MS,

    I need your help. I am planning to use LiH instaed of NaH as the anion generated would be more soluble in THF and takes less solvent volume.

    Is LiH is allowable on scale up reactions?.

    Thanks and regards


    Comment by marto — July 4, 2012 @ 11:24 pm

    • I think LiH is considerably less pyrophoric than NaH but I don’t have any experience working with LiH. What are you trying to deprotonate – and would BuLi or LiN(TMS)2 work in your case?

      Comment by milkshake — July 4, 2012 @ 11:46 pm

  29. We see you comment on inthepipeline, but whe no more blogs.

    WE want your wisdom O shaker of milk!

    Comment by psl — July 25, 2012 @ 2:42 pm

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Blog at

%d bloggers like this: