Skip to main content

How Not to Do It

When Reagents Attack!

Well, since I was just talking about a reagent that can potentially take off without warning, I wanted to solicit vivid experiences from the crowd. What’s a compound that you’ve made that did something violently unexpected? I can recall making some para-methoxybenzyl chloride in grad school (for a protecting group; I was running out of orthogonal protecting groups by that time). It’s not hard – take the benzyl alcohol and some conc. HCl and swoosh ’em around. But the product you get by that method isn’t the cleanest thing in the world, and on storage, well. . .a vial of it blew out in my hood after the acid had had a chance to work on it.
My most vivid reagent-gone-bad story is probably this one; that’s a time I literally came down counting fingers. What other things have you had turn on you?

78 comments on “When Reagents Attack!”

  1. Mike says:

    In graduate school I was doing a Birch reduction in cooled liquid ammonia. After the reaction had been going steadily for a few hours, it suddenly within seconds changed color and rapidly warmed. As I was standing a few feet from it, the entire contents of the two liter reaction flask then immediately geysered out leaving a coating of dry material on the roof of my hood. The flask was completely empty. I could smell nothing but ammonia for the next day or so.

  2. Travis says:

    My first semester of grad school I was doing an ortho-formylation of a phenol using MgCl2, Et3N and paraformaldehyde. After heating the mixture to reflux, paraformaldehyde can be seen repolymerizing on the walls of the condenser. My first time doing a large scale-up I used a narrower-bore condenser than usual for some reason and during the reflux the condenser must have plugged. The resulting pressure blew off the condenser and sprayed my entire fume hood with a sticky, canary-yellow polymer.
    Fortunately no one was hurt but it took 5 hours to scrape off the gunk from the hood.

  3. My worst reagent experience was a pretty innocuous one: Fe/AcOH
    I was doing a reduction of an aromatic nitro compound and Fe/AcOH was the only one that worked half-decent.
    The procedure called for mixing the nitro compound with AcOH, heating to 50C, then adding the iron powder.
    I have the solution heated, ready to add the iron powder. I add the iron in one charge and then put the condenser back on. I watch the temperature and it goes 50C..60C..90C..115C in about 10 seconds. I go to remove the heating mantle and WOOSH….
    The condenser popped off and I was sprayed with a nice mixture of acetic acid, iron powder and partially reduced nitro compound.
    I quickly rinsed by face off (safety glasses for the win!) since acetic acids stings when it hits your lips.
    I had visions of my skin coming off in sheets, but luckily no damage was done. I had to use abs. ethanol to remove the waxy nitro compound from my skin (including in my ears). I’m sure my liver appreciated the exposure.
    Lesson learned! Never try and save a reaction, just shut the hood and walk away. You can always clean it up later.

  4. anonymous says:

    I made an unexpectedly pyrophoric compound (a silicon analogue of bis-vinyl zinc(II)) which quite surprised me.

  5. LittleGreenPills says:

    In my limited synthetic exposure I have had three experiences:
    1. As an inexperienced undergrad I picked up an old 4L bottle of diisopropyl ether that had a lot of some sort of grey material in the bottom. The instant I picked it up it bubbled like a bottle of champagne. Fortunately it was stoppered with a cork, which shot out also like a champagne cork. Still not sure how I did not drop the bottle.
    2. Working up a LAH reduction on a very rainy night, I was left holding a flask crusted with a white solid. As I held it I noticed a strange black spot that was growing. Then noticed that it was surrounded by a red ring. As it dawned on me that it was burning, in a state of panic I had to decide, do I put it back in the hood with the L of THF from the reaction or leave it out? I put it at the far end of the hood closed the sash and watched.
    3. Trying to rearrange an allylic alcohol. I added H2SO4 to a flask of the alcohol. It began to bubble, then smoke, then whistle, and finally with a little poof sound, a puff of smoke came out and I had a flask of expensive charcoal.
    I now do very little synthesis.

  6. Opps says:

    I can’t forget the multiple times I have seen someone’s mineral oil heating bath go up in flames upon high heating. “Opps, I though that was a silicon oil bath.”

  7. baluba says:

    I recently moved to process chemistry after grad school and postdocs and soon came to realize how many things are HIGHLY underestimated in academic labs and “small scale” environments. Many simple work up (quenching acid chlorides for instance) can go very badly even moving to say 100g batch size. And NO training in process safety is available at any stage of grad school careers!

  8. AlMe3 says:

    Trimethylaluminum-mediated amidation of an ester on 100g scale: I quenched the reaction by quickly adding a few mLs of ammonium chloride solution and it shot up like a volcano (on the day of a lab safety audit no less). The next time I ran the reaction (on the same scale), I found that pouring the reaction mixture into rapidly stirring, cold ammonium chloride solution only resulted in a fizzle.

  9. p says:

    baluba makes a great point about scale up. The “best” reaction I ever had was a Simmons-Smith. I did the reaction on 50mg at 0 C and it went beautifully. I then scaled up to 2 grams, still at 0 C, and blew a flask, addition funnel and ice bath to smithereens (pun intended). My advisor saw me after clean up and said, “You know, I was thinking about your reaction. When you scale up you probably ought to cool it to -78”. Yep.
    But, yeah, scaling up creates hazards out of stuff you don’t think will be all that bad.

  10. MnO2 says:

    I once added activated MnO2 to a secondary benzylic/propargylic alcohol and the flask shot flames out of it with such force that it made the sound of a train whistle for several seconds. After the fireworks were over and I cleaned out my pants, I decided to TLC the reaction mixture. To my surprise the reaction was complete and clean enough to proceed with a workup and column. The best part of the story was that I achieved 75% isolated yield of the desired ketone after purification.

  11. Calvin says:

    Ozonolyis followed by oxidative work up with peracetic acid. Ozonolysis fine, work up not so much. Needed a bit of heat and a decent amount of peracetic acid to get it going which was fine on mg scale but on gram scale was pretty hard to judge. Monitor temp inside the reaction and in the oil bath and cool v quickly when there was a difference netween the two. One time I was a bit slow and it exploded out the top of the flask and covered the inside of the hood in stick goo. I managed to get the sash down in time to prevent myself getting covered in it but it did make a mess of the hood. Some of it is probably still there to this day.

  12. canchem says:

    First reaction I was ever instructed to do involved alkylation of an alcohol using NaH as base. I was told to wash the NaH with hexanes. I did, but thinking the supernatant was what I wanted, dumped the grease into my reaction and left about 4g of this funny grey powder in an erlenmeyer.
    That alkylation reaction didn’t work. A reaction did take place when I tossed some dil. HCl into that erlenmeyer before washing it. Kept my eyebrows, but only barely.

  13. NJBiologist says:

    Does a loose rhesus monkey count as “a reagent that can potentially take off without warning”? ‘Cause that will get your heart rate up pretty fast….

  14. processchemist says:

    A 4-nitrophenylglycoside unexpectedly exploded: it happened during the recrystallization of a second crop.

  15. anon the II says:

    Many years ago, when I left big pharma to try out biotech (or was I leaving NJ for NC), my new boss was emphatic that we save money whenever possible. My idea of saving money was different from one of my new coworkers who had spent a few years in the fields during the Cultural Revolution. One day, after rotavaping off thionyl chloride from a reaction, he poured the contents of the rotavap trap into a bottle for the next time he needed the “expensive” reagent. Oh yeah, he tightened the cap on the bottle real tight. Who knows how much pressure was in that bottle when it finally gave way in the back of his hood. Luckily, he was at the sink and I was at my desk when it went. The largest piece of glass we found was maybe a couple of mm’s long. The former bottle of hydrazine that was sitting next to the recovered prize helped form a nice white smoke in the aftermath of the event.

  16. Doubja says:

    A few years ago I wanted to convert a secondary amine to the corresponding methyl carbonyl carbamate using methyl isocyanatoformate (CAS# 5843-42-5). I found a bottle in an adjacent lab that was a couple of years old according to our inventory system. It came from a lab where I trusted the guys to take care of their chemicals. The Aldrich website and the label on the bottle made me expect a white powder. The actual content of the 5 g bottle appeared to be a liquid. I didn’t think too much about it and fortunately unsuccessfully tried (and I am not one of those lab whimps) to unscrew the lid. Luckily the next thing I tried was INSIDE the hood. I started tapping the edge of the vial against an edge thinking the thread got dirty and stuck. After the second tap the bottle (thankfully only a small 5 g bottle) exploded in my hands with a bang that made my ears ring and the guys from the lab next door to run over to see if I had killed myself. Needless to say I disappointed them. Aside from the shock I stayed unharmed. I couldn’t even find any shards from the vial which to this day freaks me out. In hindsight the change in physical appearance should have given me a hint and set off a couple of alarm bells. Fortunate for me I got to live and learn. I wont bore you with the chemistry that went on in the vial but the experts in the audience will immediately figure out what happened to the reagent.

  17. NI3 says:

    Not an accident. But once we made a large batch of nitrogen triiodide (safe when wet, contact explosive when dry) for a high school demonstration. Before we could set it off on our own with a long meter stick with a feather on the end, a fly landed on the the dry powder – KABOOM. Caught everyone (especially the fly) by surprise

  18. anchor says:

    This story goes long back to a time when we accidentally made active manganese dioxide. We carried out exhaustive oxidation of a natural product (several kilos) with KMnO4. None of us anticipated that the byproduct manganese dioxide would be super reactive enough. The filtration of the heterogeneous mixture was smooth but the pad of manganese dioxide along with residual organic caught fire and had us all surprised! After putting off fire and exercising caution we stored the manganese oxide in several containers for disposal. The manganese dioxide worked like charm for selective allylic/benzylic oxidation for several months. I have used other variants of manganese dioxide (active) that is sold commercially these days, but I reckon that the one we made accidentally was the best!

  19. Yazeran says:

    Well my most unexpected chemical reaction was way back in high school when I was fiddeling with a variety of redox reactions (read ways to create home made fireworks). I had read through our chemistry textbook and found that the entry on amonium nitrate mentioned that a significant part of the world production went to military use…
    Not surprised, I worked out that NH4NO3 liberates excess oxygen, and that could be used to burn things. I also looked up the melting point and found that it melts before it decomposes, and by accident the melting point is similar to that of plain sucerose.
    Bingo I thought, the perfect mixture for making solid rocket fuel.
    So I carefully worked out the exact ratio of amonium nitrate to sugar and slowly melted the amonium nitrate in my parents kitchen. As soon as it had melted, I removed it from the stove and slowly added the sugar while stirring. Initially it went ok, but after a few seconds the mixture became first brownish yellow, and then gradually darker untill it was almost black while iot starterd to smoke. As it dit this, it also became more and more sticky, and I was in a bit of a loss as to how I was to explain a completely gunked pot to my mom.
    I did a fast thinking and rememberred that both the amonium mitrate and the sugar is water suluble, so quickly added some water to quency/clean while the substance was hot and likely to be still disovlvable.
    What I didn’t count on was that as soon as the first drop of water hit the mixture, the complete batch exploded (it didn’t detonate fortunately, it only burned extremely fast).
    I can still with dread remember how the window in front of me suddenly looked to be blurred as the hot gasses from the reaction literally made a meter and a half jet up from the pot. Incidently I must have added just a bit too much sugar, as afterwards most of the kitchen was covered in extremely fine black particles (soot) which took more than an hour top clean up as I had to wash almost all surfaces to remove all traces.
    The pot you say? It was completely clean afterwards!
    I only told my parents about that incident some 10 years afterwards…
    Yours Yazeran

  20. Seth says:

    I was working with an old bottle of thioacetic acid, when I went to syringe some out the needle aided in the degassing of what I presumed to be H2S. This resulted in the thioacetic acid being redistributed all over the hood and me. The smell was something on the order of a skunk and dog poo, which my wife didn’t really care for.

  21. Superted says:

    Not exactly a reaction gone bad, but… we used to have a small kugelrohr kit that we used to distill products from Rh and Ir catalysed reactions we were running. I used to use conc. sulfuric acid and hydrogen peroxide to clean it which was always entertaining, but on one occasion the exotherm resulted in the entire mixture erupting out, just as I was cleaning up in the rest of the hood. Result: labcoat sleeves a mess, and scars on one arm that are still visible today, 20 years later.

  22. okemist says:

    I would like to thank everyone for these posts. They contain valuble information. Two of my worst experiences were:
    1. after cooling a hot distillation of ether and opening it to air, the mixture was reheated and it exploded. I learned to always work in an inert atmosphere, it take a lot longer when you scale to large vessles, but it is worth the patience.
    2. 800g of LiAlH pellets in a crys dish ignited just with static, we had used dailey for years with out problem. opened a tank of lN2 to smother it but it would not extinguish. finally buried it in cat litter (clay) about 100 lbs did the trick. Thank you all again for sharing your experiences.

  23. newton says:

    Back at my postdoc, I prepared a 50g batch of MoOOPH to oxidize thioethers. The material looked beautiful as I collected it by filtration, but as I scooped it out of the funnel it went up in smoke and gave off a not-so-surprisingly awful smell. The resulting cloud and odor chased the entire floor out of the building. Luckly the lab had windows that still opened (probably not a good idea, but probably saved a few years of my life).
    Another reference for the prep of MoOPH said that if all the solvents are not removed from the intermediate steps the product could be pyrophoric. Wish I read that one first.

  24. ProteinChemist says:

    Back in my first year of grad school I was working in a lab that needed a lanthanum oxide solution (not sure why now). To get it soluble, it was recommended at the time that I use concentrated nitric acid. Not thinking about the exothermic nature of the reaction, I combined 200g of lanthanum oxide and about 150mL of nitric acid in a 1L erlenmeyer. It looked a lot like those science fair volcanoes, except that sure wasn’t baking soda. Needless to say, next time I dissolved it a bit at a time and over ice.

  25. ScientistSailor says:

    Not me, but in the next lab over, a grad student was doing a large-scale Diels-Alder with butadine in a sealed tube. When it blew, destroyed everything in the hood and even pushed the walls of the hood out. Fortunately, the lab was empty at the time.

  26. James says:

    I was making bis TMS peroxide through the DABCO-peroxide adduct. I had filtered the white complex and left the bench for a minute to retrieve something before treating with trimethylsilyl. In the meantime I heard a terrible hissing noise and came back to find the adduct, blackened and smoking, consuming my filter flask. Not the kind of compound that should be left out in air for long.

  27. John Wayne says:

    Heh … I’ve got a few of these:
    1. In graduate school I was making an enolate with sodium sand based on a pretty old procedure. I had the sodium suspended in solvent in the round bottom flask and adding the ketoester in solution slowly via addition funnel. It all added, but nothing seemed to happen, so I let the reaction warm up a little … when the whole thing exothermed it shot the additional funnel through my manifold and covered the hood in little bits of sodium and solvent. Fortunately, I saw the internal temperature starting to spike and closed my hood sash in time; the only part of the addition funnel I ever found was the teflon stopcock.
    2. I had just performed a reaction wherein I had washed sodium hydride with hexanes repeatedly to get most of the oil out of it. Everything went well, and the reaction was proceeding, but the hexane washes needed to be quenched. I added water via wash bottle slowly, got impatient, and it got away from me right as the professor I was teaching for walked into the lab to talk to me. He walks up to a graduate student putting out a reasonably sized fire with a towel and his gloved hands and says, “I can come back if you’re busy.” I got the fire out just at that moment and replied, “My schedule just cleared up; what can I do for you?”

  28. MTK says:

    Fire, Flood, Explosion.
    The unholy trinity of chemistry.
    I’ve had ’em all.

  29. gippgig says:

    Did anyone figure out what it was in that para-methoxybenzyl chloride that did the actual exploding? Are there any compilations of the chemistry behind these unexpected explosions etc.? It seems like an interesting and useful topic to study.

  30. Neil says:

    @Yazeran – I had almost the exact same thing happen to me, while in high school. A friend and I were mixing up our own solid-rocket fuel (to re-fuel our spend model rockets), and were also in the kitchen at the time.
    We were heating the mixture in a small metal pot, stirring constantly. My friend noted that a small amount of smoke was present, when the whole thing blew up on us!
    Anyone who knows the old ammonium nitrate and sugar bit will also know of its use as a smoke powder…and smoke it did! My friends kitchen was more or less instantly filled with a dense sweet-smelling smoke. His mom was outside at the time, but ran screaming into the house when she saw the billowing clouds of smoke emanating from the kitchen door (which my friend had thrown open to help clear the mess).
    We were lucky, however, and did our cooking in the sink! After a good hour or so the kitchen was clear and clean.

  31. Anonymous says:

    #19 reminds me of a very similar reaction that surprised me by working precisely as advertised – I never would have dreamed of trying it, but the greybeard doing the serious chemistry on our team swore by it.
    As with Yazeran, we were making a rocket propellant, but we specifically wanted to nitrate the (proprietary) polymer we wer using. So, mix polymer with the right amount of NH3NO4 and heat to just shy of the melting point. Result, a molten eutectic that isn’t hot enough for ammonium nitrate to decompose outright, but close enough for individual nitro groups to jump over to the energetically-favorable carbon right next door. Wait for the ammonia to stop bubbling out of the mix, cool, and wash. Or, when we were planning to use ammonium nitrate as an oxidizer, just arrange for there to have been a sufficient surplus in the first place.
    Worked just fine (and yes, we were doing this in a shielded test cell in a bunker designed for experimental rocket propellant work).
    I suspect something similar may have been going on in the nitrate/sugar melt incident, save that the result would have been crude sucrose octanitrate. “…hitherto proved of little practical value as an explosive on account of its extreme sensitivness and liability to spontaneous decomposition” (Hoffman and Hawes, JACS, 1919).

  32. Jonathan says:

    I wonder if possibly my finest moment as a young PhD student back in the 1990s counts? As a pharmacologist, I didn’t really have to worry about blowing up the lab so much as accidentally poisoning myself with the fruits of someone else’s reactions. We did roll our own radioimmunoassays though, which involved heating and stirring the buffer to get the gelatin to dissolve. For reasons I can’t quite remember, I had been making up my buffer in a lab on the second floor, despite the radioactive lab being on the ground floor. The floor in-between was where the ice machine was located, and I stopped on the way down to fill up a large tray with the necessary ice to perform the assay. Only having two hands (and presumably no functioning brain), I decided that the easiest way to get everything back to the lab was to carry the bottle of buffer in the tray of ice. Did I mention that the bottle was hot? Or that it was made of glass?
    I think I made down a flight of stairs before a loud bang and lots of tiny bits of glass announced to me that negative thermal expansion had done its thing. Bloody good job I was wearing glasses!

  33. Anon says:

    When I was a graduate student, a post-doc across the hall decided to purify ~100g of impure TBS-Cl. This is a solid at RT and purified by vacuum distillation (with a gently warmed short-path condenser).
    The Post-doc had not done it before and chilled the condenser with water. He got ~10g over before the distillation stopped…. so he cranked up the temperature… WAY UP because it wouldn’t distill….. You can probably guess what happened.
    The condenser had sealed off with solid TBS-Cl, so the vacuum was not longer applied to the still pot. By cranking up the temperature (on a now sealed system), the pressure built until the still-head popped out explosively venting a massive amount of hot TBS-Cl vapor that created clouds through his lab bay. I only saw the aftermath of clouds, shattered Still-head, water (from the condenser) and a spray of TBS-Cl all over the interior of the hood. Fortunately, he was at his desk when the “eruption” occurred so no one was hit with flying glass.

  34. anon says:

    Where’s milkshake? Go back in his blog and he has had more excitement with energetic molecules than anyone I ever heard of. Mostly in Prague,as I recall.

  35. Gene says:

    Not in a lab, but how about turning a 40-foot garbage compactor container barrel shaped (think of Mrs Puff after Spongebob crashes the boat again)? I’m an environmental regulator and got called out to the rail loading yard here because, “A container has exploded.” When I got there the container was on its side next to the train and the fire department had three 2 1/2″ hoses blasting into it full force with heavy smoke pouring out. So I started collecting info on what had happened.
    Trash and garbage here go to transfer stations where it’s dumped into pits and compacted into 40-foot, heavily reinforced containers which go on trucks to the loading yard where they are put on trains to the landfill. This train had been loaded with containers and was getting ready to depart soon, so one of the yardbirds was walking the train ensuring everything was in good shape for the trip. As he said, “I was about two cars past that one when I heard a loud boom behind me. I turned around and it was all blowed up and falling off the car.” Luckily, it was top loaded, if it had been the one on the bottom it would have been a whole lot more interesting.
    After much investigation (by another agency, I deal with what goes to sewers) here’s what was discovered: A few days before an employee at a heat-treating/metal finishing/welding company was cleaning up and discarded what he thought was waste to the dumpster. They regularly put waste into the sealable metal tins that welding rods come in. This dumpster had gone where dumpsters go and its contents had been compacted into the container along with many tons of other stuff, much of it wet, then sealed. It turns out that the “waste” was granular CaC2, so as the CaC2 was getting wet in the container it was producing C2H2, which at some point reached the right concentration/pressure/ignition source/incompatible substance and went BOOM! It blew the container barrel shaped, blew the doors off (or they fell off due to distortion) and knocked the container off the train car.
    That energetic enough?

  36. Hap says:

    Mine was dumb but with few consequences. I was reducing oligo(oxyethylene) azides with Pd/C and H2 balloons (getting my product in 15% yield due to my lack of competence via secondary amine formation), and I added Pd/C to a 25 mL flask with hydrogen. Lots of sparks, but no flame or explosion. Fortunately.
    Two grad students when I was in school were moderately injured by ozonolyses gone bad. Both were competent and smart, but apparently their reactions weren’t quenched properly and some of the ozonides remained. Both had nice metal attachments to their wrists for awhile afterwards, and at least one had a shattered stir plate to remember the incident by.
    35: Acetylene’s already been around that particular block here – see this post for an example.

  37. Dmitry B says:

    LAH is always dangerous. Especially when you reduce carboxylic acids (such as cyclopropane oc cyclopentane carboxylic acids) . It starts to react violently after it heated to reflux.
    also 2-(Chloromethyl)pyridines are very irritant. gloves are useless in this case.

  38. Anonymous says:

    Not me but some wanker at a big pharma Co. here in NJ (I won’t name it but please be my guest) decided to do a large scale displacement of an aryl or alkyl halide with NH3 in a pressure bottle. Sealed that bugger up, applied heat and went away. The blast was heard 2 floors up!!!
    So why wasn’t this wanker fired??? Well, simply put, he was one of the favorites….still there today LOL.

  39. This one didn’t happen to me, but it happened at the pharma company I was working for.
    A technician was repeating a reaction (epoxidation I believe) using t-butyl hydroperoxide. This was on scale mind you (100-200 mL of a 70% aqueous solution).
    The procedure called for the solution of peroxide to be dried through an azeotropic distillation with toluene. In the end, too much solvent was lost and the epoxide exploded. I had a chance to go into the lab after the fire was put out.
    The best way to describe it was that the hood looked like those pictures of bombed storefronts in Algeria during the war with France. Soot everywhere. The sash was blown out. The metal frame of the hood was pushed out a good foot or so, pulling the pipes out of the wall at the same time. All the windows in the lab were shattered but the wire mesh held them in place.
    The young lady ended up with third-degree burns and a great deal of glass injury.
    Overall a pretty horrible accident.

  40. Anonymous says:

    Here’s one experience that I had while working in a university lab many years back EH?? LOL
    Someone was drying toluene with Na in a big ass 1-2L triple neck flask in the hood. I believe the rubber hose slipped off the condenser and so there was frantic activity to get it back on securely. Unfortunately, the condenser snapped (at that weak point) and water got into the flask. All I can say is there was a very rapid mass exodus. The fire was put out after about 10 mins but the hood looked like someone had sprayed 2 gallons of black paint in it. Luckily no one was hurt.

  41. Hap says:

    39: Why did someone think azeotropic distillation of a peroxide in toluene was a good idea? I’m not blaming the tech, more the people who set up the procedure – it just seems like heating a peroxide at 111C seems like a good way to find out (among other things) how many glass pieces you can embed in various lab surfaces. I don’t think I need to find that out.

  42. Dr. Demented says:

    We used to get pretty regular (small) fires upon adding dry Pd/C to a solution of MeOH and substrate for hydrogenations. Then someone got the bright idea of switching to DeGussa-type (50% water) and we haven’t had a fire since.
    Along these lines – know how you can tell if your Raney Ni is any good? Put a little bit on a paper towel…wait 1 min…is the towel on fire?…yes?…you’ve got good Raney Ni!

  43. @41
    Yeah, it certainly wasn’t the technician’s fault. She was following a procedure and may have let the solvent get low, but she clearly wasn’t informed of the risks.
    I think they did the azeotropic distillation because any trace of water in the reaction messed up the yield and/or purity. Azeotropic distillations are easily scaleable, that’s why they went with it.
    I’m pretty sure out process safety people approved the distillation procedure. However, I’m also sure that it was based on the idea that the concentration of the peroxide in toluene never got above a certain %.
    Again, it was more of a lack of information than anything.
    This reminds me of another pharma accident I witnessed. It was an oxidation of a sulfide to a sulfone using hydrogen peroxide. The reaction was scaled up from 1g, to 5g, to 50g, then to 500g.
    The last reaction was done in a 22L RB with over a 1L of 30% H2O2 and didn’t go as planned. Typically exothermic run-away. The scientist reported that the temperature slowly climbed, ice was applied, the temperature continued to climb, then the normally white suspension started to get “puffs” of gray solids in it before the RB burst. A good friend has a noticeable scar on his neck from a chunk of glass. Luckily that was the worst of the injuries.

  44. G. says:

    and how I learn to be patient.
    I was trying to do a strange acetylenisation of an bromopyridine in a microwaves seal tube at 150°C. Pressure build up for sure, does it explode ? not yet, the µw vials stand strong. Does it react ? hum … i didn’t took a TLC of the raw and did not really wait till the mixture cool down, and follow the reported work-up : 6M HCL quench. WOOOoosh. Next day my lab coat was missing a sleeves.
    Two little stories that I have witness.
    1) when I was undergrad in this sugar lab, a grad student did THE classic reaction with thiophenol to protect sugar from épimerisation. The 3rd time she tryede that was on 50g scale, the same time as the hotplate went mad. The Kaboom was not so impressive but I took half a day off 15 min after that. I got the worst headach ever.
    2) Some times ago one of my lab mate run a reaction to get R-dithiocarboxilic acid from R-CH2Cl. Using in-situ formed MeOS. in exces. He somehow quenched this reaction with acetone. We, i think, characterized thioacétone just by the horrendious smell it produce.

  45. milkshake says:

    Some common nitrite esters and chloroformates are underestimated decomposition risk. When improperly stored at room temperature over long period of time, they can pressurise themselves with CO2 in autocatalytic fashion and burst the bottle when you pick them up in the storage room or will erupt into a geyser into your face when you unscrew the cap. Isopropyl, allyl and benzyl chloroformates are particularly prone to this decomposition.
    Low-molecular weight acetylenes are quite endothermic and most people are unaware of this. I saw 1-bromopropyne erupting into flames and a column of soot next hood after it was accidentally exposed to air (the stuff is lacrymatory and toxic as hell too: 1,2-Diiodoacetylene was being developed as a chemical warfare agent during WWI but was never fielded because it could not be handled safely)
    Another nasty thing is an old rusty tank with anhydrous HF. HF tanks are made of steel and steel is not supposed to react with HF except that sometimes (depending on the grade of used steel) steel actually does react with HF, very slowly, producing hydrogen gas. Some old low-pressure tanks of anhydrous HF were found over-pressurized way above their rated pressure because of this problem.

  46. Richard Friary says:

    Ozonolysis of cyclooctatetraene monoepoxide, followed by ineffective treatment with dimethyl sulfide. After work-up and evaporation of CH2Cl2, the experimenter had 45 min. to begin chromatography. Otherwise, the crude product turned black and issued a column of black smoke. As long as the chromatography was begun promptly, the yield of dialdehyde was reasonable. See, I. Ernest, Angew. Chem., Int. Ed., in the mid 1970s.

  47. gippgig says:

    #45 – Which “common nitrite esters” are decomposition risks? Some alkyl (i.e., butyl) nitrites are used as recreational drugs (poppers). Is there a possible hazard here?

  48. Yazeran says:

    30, Neil.
    Actually the strange thing was that apart from the initial small smoke and bubbles from the mixture, no smoke was created when the thing went off, it was an almost completely clean combustion (apart from those pesky small soot particles which were such a nuisance to clean up). Were it not for those, no trace of the complete incident would have existed 1 second later…

  49. bcpmoon says:

    I second the call for scale-up 101 for every PhD student, I am still amazed that I passed uni unscathed.
    Two stories:
    1. Once in while the 2L THF/Ether-stills had to be cleaned when the Na-gunk started to accumulate. I quenched the deposits with iPrOH then EtOH then MeOH and finally water until only (ha) some white stuff remained. I removed this with a brush at the sink and at the same moment when I noticed some glinting embedded, there was a whoosh and the world went yellow in a big fireball. Never let your guard down, especially during work-up.
    2. A LAH-reduction was insufficiently vented in the neighboring lab, but thats what hoods are for, right? But that hood had two openings, venting at the top and at the bottom. Guess which was open… H2 accumulated at the top, found a ignition source (its quite ressourceful in this regard) aaaaaaaand: BOOM.
    Hood completely destroyed, doors off their hinges, windows cracked. No one hurt and miraculously none of the 5000 solvent bottles on the shelves (university lab) broken.
    I also recommend Trevor Kletz book “What went wrong” and the the aptly titled successor “Still going wrong”.

  50. scientistbymistake says:

    I was reducing an anthraquinone down to the anthracene using powdered zinc in an ammonia solution. It was at reflux, and as I watched waiting for the colour of the liquid component to change from deep red to pale yellow, whoosh – off it went all at once, blowing the condenser into the top of my hood where it smashed, and spraying the mixture all over the place!
    As I flinched away from it a drop found its way between my glasses and my face, straight in to my eye. Cue a trip to hospital after much eye washing, having the pH of my eye tested by a doctor (stick a piece of pH paper in an already sore eye!), and abrasions to the surface of my eye (cornea?). It was pretty sore for quite a few days.
    Blast shield etc was employed for the next run, and the scale up etc, all of which went without a single hitch. I remain mystified about why it ‘did one’ all at once like that.

  51. Calvn says:

    Following on from your story bcpmoon I remember a couple of events from my PhD. In the lab 2 floors below me a student switched on the THF still but inexplicably failed to notice that the condenser wasn’t even attached to the flask and the fumehood (a recirculating one) wasn’t on. 30 minutes later it all went kaboom quite impressively (nobody hurt but a few temporarily deaf). 2 floors above I felt the floor hop and I knew it was time to get the hell outta there.
    A few weeks later, a post-doc in another lab decided that he’d run a pretty big falsh column using pet ether/ Et2O but because he was short of space he’d run it in the hood next to his reaction which was cooking at 200oC. More kaboom, more red faces.
    Ahh the joys of academic research

  52. RB Woodweird says:

    Many accounts of kaboom here are due to aged reagents, but I recall vividly the time I was astonished by a brand new bottle from Aldrich. I had been reducing a double bond on a multigram scale in a Parr shaker. I had synthesized the precursor, and the Pd/C was a giant ancient bottle which might have been the original lot ever made. After several runs, I used up the last of the reagent. My PI rent his garments, gnashed his teeth, shaved his eyebrows, knocked out one of his front teeth, and eventually bought me a new bottle. I obtained the shiny new trophy and proceeded with the usual: add solid alkene to Parr bottle, add catalyst to Parr bottle, add ethanol to Parr bottle. This time, as I was standing at the bench, looking down into the Parr bottle as I was pouring ethanol from a gallon jug into it, I noticed something I had not seen in previous reactions: a roiling orange-red ball of flame boiling up from the depths of the bottle like it was chasing Bruce Willis down a tunnel. I just managed to jerk my head out of the path of the jet of fire.
    A few minutes later, flames all out, I went to run my fingers through my hair in one of those involuntary gestures of relief and came away with a handful of toasted matted hair ash.
    I have still a healthy respect for the power of fresh catalysts of all kinds.

  53. Anonymous says:

    Just read your original 2004 article, I was making phenyl sulfenyl chloride too in my first year of postgrad, however, I made it by cleaving diphenyldisulphide. The round bottom flask lived in the fridge under Ar for the remainder of my PhD without any consequences (thankfully)..
    However, I remember getting the flask weighed after the synthesis and accidently knocking off the rubber bung for no more than 2 seconds. This has caused a stench that cleared the lab for the best part of an hour.. You live and learn.

  54. Kevin says:

    Sodium-Mercury amalgam… used to in the system of ferrocene. One batch just exploded one day.

  55. OldLabRat says:

    As an undergrad, I worked in a lab making diphenyl ethers from diphenyl iodonium salts. The latter reagents were made from iodyl sulfate and halobenzenes. We usually ran these on small scale cooled to about -20C. A 10umol test reaction with an o-Cl-nitrobenzene blew out a 1/2″ thick, wire-reinforced Lexan hood sash, two blast shields, and destroyed a large Schlenk apparatus on the opposite bench with glass shards from the hood reaction embedded in the stainless steel monkey bars.
    Fortunately, my lab supervisor and I managed the few steps out the door before the boom. The rapid appearance of iodine was a nice hint.

  56. JJT says:

    I was distilling propargyl alcohol from potassium carbonate, see Note 2 of Organic Syntheses, Vol. 64, p. 182 (1986), when the flask blew up right in front of me. Miraculously I was unhurt. My hood was damaged; bottles were knocked off the shelf in my lab and adjacent labs. Accelerating rate calorimetry (ARC) studies by our safety group showed that a mixture of propargyl alcohol and potassium carbonate spontaneously auto-heat beginning at ca. 70 C to generate enormous pressure. The ARC stainless steel sphere containing the mixture ruptured.

  57. Anon says:

    Quenching an ancient canister of DIBAL solution, after removing the ball valve I added a drop of water to what I thought was a near empty container and was treated to a glorious tower of flame that had to be 3 feet tall. I was able to quickly take care of it with the fire extinguisher, but was a little shaky the rest of the day to be sure.

  58. Hap says:

    45: When I was in undergrad (early 90s), the previous explosion of a bottle of ethyl chloroformate carried without a carrier (and killing the tech who had it) from hydrolysis overpressure was the given justification for the “no carrying bottles without a bottle carrier” rule.
    One of the grad students my year had an incident with Na/K alloy, but I’ve described it here at least twice already.

  59. Anonymous says:

    A coworker had an explosion caused by the (unreported in a procedure) shock-sensitive nature of the TMS-enolate of nitromethane. A distillation bumped and upon removing the heating bath, it blew.

  60. Ken Rubenstein says:

    In grad school, I was making a saturated solution of sodium methoxide by tossing small chunks of sodium into methyl alcohol. I went out to lunch and came back to find nice crystals of NaOMe coating the bottom of the flask. I poured off the liquid, set the flask on a cork ring and added some water to dissolve the crystals. Much to my shock and dismay, the crystals dissolved to reveal a sizable chunk of sodium. I backpedaled furiously and soon heard a really loud boom. I returned to find the bottom of the flask, looking like a watch glass, sitting on the ring, and no trace of wreckage. The remaining glass had simply powdered. I never did that again.

  61. Jason says:

    This was in the mid 1960’s in Rockland County, New York. I was ten years old, between 5th and 6th grade, enjoying my summer by experimenting with home-made rockets and pyrotechnics. I had done many successful experiments, but what got my chemistry set taken away was this: I mixed equal amounts of sulfur and powdered swimming pool chlorine (mostly calcium hypochlorite) by pouring the mixture back and forth between two pieces of paper. Then I poured it into an aluminum foil tube. There was 1 or 1.5 ounces in there. I held the ‘rocket engine’ up to my eye, peering into it to see how full it was. Provoked only by the heat of my hand, it burst into a jet of flame, as it was designed to do, but directed unfortunately at my eye ball. I screamed, and explained to my father what had happened and he picked me up and thrust my head under the kitchen water faucet for several minutes, then sped to the emergency room. My right eyebrow and lashes were completely gone, and some of my hair. My skin was bleached white from cheek to forehead. I had to wear a black eye patch for the first few weeks of 6th grade, like a pirate. It gave me a reputation I quite enjoyed. I recovered fully, no scars or eye damage at all, because of my father’s quick action, the doctor said. Ultimately I became a mathematician, not a chemist. I still make small July 4th fireworks for fun, but carefully now.

  62. Nick K says:

    Superb comments! Perhaps Derek could publish it in a book as there is a wealth of really valuable safety info here.
    A couple of my own, rather shameful safety anecdotes: using calcium hydride to dry DMF. To dispose of the residue I carefully cool the flask in ice water and added ethanol very slowly. A mild fizz results. Said flask is then left for a few hours, by which time all hydrogen evolution seems to be over. Flask then placed under a jet of water – POW! A huge geyser of grey slush and steam erupts and splatters the ceiling.
    Work-up of a lrge-scale acyloin condensation with sodium metal and trimethylchlorosilane – after a scary delayed exotherm the reaction proceeds normally, producing a large amount of a strange purple powder. The reaction mixture is carefully filtered under argon and the filtrate is evaporated and distilled to obtain the product. Fine, but how to dispose of the purple solid? Again, I quench it under argon with ethanol. A few hours later the residue is tossed in small portions into the sink with water running freely. Yikes! Abundant hydrogen produced! A tiny speck of sodium metal then ignites the hydrogen, producing a huge fireball which burns the ceiling.

  63. Secondaire says:

    I’ve never had any problems with the PMB-Cl prep – I Immediately add anhydrous K-carb and freeze it to prevent any acid shenanigans. I wonder why yours blew…
    I did a lot of aminohaloborane chemistry in graduate school, especially the Sugasawa modification of the Friedel-Crafts acylation, where one complexes anilines with BCl3-Me2S before adding some acylating agent – it directs ortho very nicely. I used to run them in big, three-necked flasks so I could sparge with argon during the first part, and then I’d simply stopper up the side neck.
    Yeah, one must quench that carefully. At the end, you pour aqueous HCl down the condenser. My reaction was on a large scale, and not *quite* at room temp yet, and I was rushed for time, and added a the HCl a little too fast; resulting exotherm went absolutely out of control, blew the stopper out the side of the flask, and vomited pea-soup-green nastiness all over the back of my hood. My friend (coming for dinner) came to my lab to find me practically inside the hood scrubbing away, with everything reeking like Me2S…
    I had a couple of those reactions go absolutely awry during the quench step, especially if the mixture wasn’t homogenous. I also had an exploding Dieckmann condensation (beware 95% NaH!), and at least three separatory funnel neutralizations that resulted in either a) stoppers being launched 30 feet, or b) me getting coated from head-to-toe with the stuff (one of the latter involved HCl + formaldehyde. Ugh).
    My boyfriend had an old Na still quench quite literally turn into a jet engine of flame. He also lit his arm on fire while making LiNH2 because a piece of lithium bounced out of the reaction mixture and onto his sleeve.
    I also saw a guy light his *entire* hood on fire with t-BuLi.

  64. D.J. says:

    Non-chemist here with two thoughts:
    I keep thinking this would be a show to put on FOX like many of theirs from the 1990s.
    Also, where does the term ‘reagent’ come from? What would agents be in chemistry? This then leads off to the philosophical questions of agency in chemistry….

  65. As an undergrad I sucked 5 M NaOH into my mouth and could not eat properly for a week. I also had solid bromine in a vacuum tube explode and had to evacuate. Result: I became a theorist.

  66. Secondaire says:

    Forgot to share one more. The absolute WORST exploding reagent story I’ve EVER heard of happened in my graduate lab on a day I was sick. I simply got an alarming phone call at 6PM involving the phrase “a lot of blood.”
    During a prep of a high-MW organic azide (containing some impurities), the thing blew on the rotovap. It had been synthesized over a hundred times before without any incident, and this one time, there was a spark and kaboom. The hood was open at the time, and the entire rotovap was utterly pulverized – half the glassware ended up in my labmate’s arms and he had to go to the hospital.
    I saw the damage the next day. Tiny, sharp shards of glassware had been blown like, ten feet in every direction. How at least three people didn’t lose an eye from the incident baffles me sometimes.

  67. opps says:

    In grad school I remember a guy chewing tobacco in lab, and spitting into the garbage can. A couple of seconds later it started on fire. That was unexpected!

  68. opps says:

    In grad school I remember a guy chewing tobacco in lab, and spitting into the garbage can. A couple of seconds later it started on fire. That was unexpected!

  69. Anonymous says:

    PMB-Cl is the one I recall vividly. It was a commercial Aldrich bottle not a home made batch. I took it from the fridge and set it in the back of my hood to warm to RT. As soon as I cracked the cap there was a powerful release of gas and the contents erupted into a foaming goo. I came this close to wetting myself. I have not used PMB-Cl since. There is always another protecting group I’ll choose to go for before ever using it again.

  70. Design Monkey says:

    Old bottle of acryloyl chloride spontaneously polymerized and blew itself up in fridge.
    Cleanup required gas masks.

  71. newnickname says:

    The hazards and instability of propiolaldehyde was documented in a few places: avoid bases, avoid amines, avoid metals, etc.. Glass seems to be OK. 🙂
    The Org Synth, Coll IV, 813 prep is essentially a Jones Oxidation of commercial propargyl alcohol. I made it several times but the major loss was in the final redistillation. I made a large batch (5 L pot) to get approx 500 mL of colorless crude and distilled in small portions, as needed. The stored material (brown bottle; fridge) was pretty good by NMR and turned pale yellow only very slowly. After distilling several small batches as needed over many weeks, the “last” batch didn’t go too well.
    I warmed the 500 mL bottle to room temp, poured out around 50 mL to distill, recapped the 500 mL bottle and let it stand while getting the distillation set up. A short while later, the bottle blew up, sending tiny bits of glass everywhere and breaking some glassware in the hood.
    It was probably exothermic polymerization that pressurized and burst the bottle but I have no idea what set it off. I knew never to put anything in the bottle (not even a pipet, I always decanted what I needed). Maybe it was a cosmic ray.
    There was an explosion down the hall from me one day. The local CSIs (chemical scene investigators) figured out that the guy had put a can of ether (Et2O) on top of the lab oven. Maybe it’s the same guy who went to work near #38.

  72. Gibbon1 says:

    70. Old bottle of acryloyl chloride spontaneously polymerized and blew itself up in fridge.
    Reminds me, couple of years after I graduated high school, our chemistry teacher noticed a 30 year old 1000ml bottle of picric acid stashed away on the top shelf of the store room. Likely part of the original lab supply. And also completely crystallized. Bomb squad called, nothing happened. Several other bottles were found at other schools in the district which lead to general purging of unneeded supplies.

  73. alex says:

    I made a Hofmann rearrangement of an aromatic amide on a 4 l scale using bleach. The first batch went well when freshly prepared bleach (Cl2 into NaOH) was used. The next batch I used a stock of industrial bleach. At 10°C the reaction mixture started to foam and suddenly a violent detonation destroyed the whole 4 l sized apparature – flask, dropping funnel and reflux condenser leaving only smashed glasparticles of

  74. Dickweed Jones says:

    Two “associates” were told to run a BBr3 deprotection. When they pulled the 1 L bottle out of the fridge it started smoking. No biggie. At this point they had the choice of:
    a. Calling their supervisor
    b. Putting the bottle in a giant crystallizing dish, half filled with DMSO and smashing it with a hammer.
    Safety had quite a time with this one.
    And exactly how you word that memo?
    Good lord…

  75. Hap says:

    Whiskey Tango Hotel Foxtrot, over? That was the first thing that came to mind because:
    a) there was a bongXXXXwater pipe in the room and they had partaken of it multiple times,
    b) both of them had received severe head traumas not long before,
    c) they wanted to see what happened with Swern oxidations under ambient conditions but didn’t have any oxalyl chloride lying around,
    or d) other….?
    I assume drug tests were administered or requested soon (if not immediately) after.

  76. Anonymous says:

    I worked at an environmental lab for a time. We had a refrigerator where we kept extracts for 5 years with no problems. These extracts were using Hexane as a solvent. One day, early in the morning before almost anyone had shown up, the hexane discovered an ignition source. This converted a perfectly harmless refrigerator into a 28 cubic foot fuel-air explosive device.
    The door of the refrigerator was blown straight down the aisle between the hoods and the benches, and when it hit the edge of a bench, the bench cut right through the steel of the door.
    The ensuing fire was hot enough that it melted the copper and aluminum of the heat exchange coils in the fridge. The blast not only removed the door of the fridge, it blew all the ceiling tiles in the room out, cleared all the walls in the building of pictures and paintings, blew the plastic wall mounted mailboxes off the walls and cracked the skylight above the room.
    No one was hurt. However, the lab was completely trashed. What the explosion and fire didn’t wreck, the sprinkler system did. Had it waited 10 minutes to occur, someone would have been working about 8 feet from the fridge, and likely would have been killed when the door made its land speed record attempt.

  77. Anonymous says:

    Re #16 – Doubja: Not a chemist by trade, but I’m going to take a wild guess and say someone opened it on a humid day and it hydrolyzed on you, forming a urea derivative of some kind and CO2 gas.

  78. Sverdalov says:

    It occurred to me while reading this that I do actually have one of these stories. My background is Chem-E. We were doing an acid hydrolysis of biomass derived carbohydrates at large scale (50 gallon glass lined reactor) at ~140C. The student whose experiment it was wanted samples during the run, which we (I) took from a 1/4″ dip tube in the reactor. So while taking a sample, the flow stops, response: open the valve more. The line had gotten plugged and the plug let loose upon further opening of the valve, spraying boiling sulfuric over me and the wall. If I hadn’t believed in safety goggles before, that would have done it. Luckily, a sink was handy and I got the acid off before too much damage occurred. Acid burns went away (much to my surprise) after a few weeks. Ah the fun of scaleup. I would say the most important thing is know your system / reaction. If I had known that we would get solids forming mid reaction, I would have built the sampling rig differently!

Comments are closed.