Skip to Content

Trimethylaluminum Explosion in Massachusetts

Word has come of a bad industrial accident in the town of North Andover, about 25 to 30 miles north/northwest of Boston. There’s a Dow facility there, the Advanced Materials division, and what makes this particularly bad is that they had an explosion and death there as recently as 2013.

That one appears to have been due to trimethylindium, and today’s accident is said to be trimethylaluminum. My fellow chemists will be pursing their lips – these organometallics are among the most notorious fireballs in all of chemistry. Trimethylaluminum earns the respect of everyone who handles it – and if it doesn’t, then there’s always next time. Any exposure to oxygen, moisture, and a whole list of other things will produce billowing masses of flame, a wildly exothermic reaction driven (as with thermite) by the formation of the corresponding way-down-there metal oxide. Honestly, I would far rather handle t-butyllithium on a regular basis, its reputation and track record and all, than neat TMA.

The compound is bad enough on a bench scale, and has caused many a face to go pale behind the lab goggles (and if you don’t have something in front of your eyes when you’re working with this kind of reagent, you’re a fool). On the industrial scale. . .well, that’s something else again. There are all sorts of sealed systems, pumps, and reactors to keep trouble at bay, but as these events show, trouble doesn’t rest. It’s far too early to say what went wrong here, and reports of injuries are confused (4? 5?). But something went seriously wrong. Again.

34 comments on “Trimethylaluminum Explosion in Massachusetts”

  1. Handles

    These rockets contain triethylaluminium, not trimethyl, but you get the idea:
    https://youtu.be/mkFU7o3IAaM

  2. Chrispy

    Maybe not a “Things I Won’t Work With” but close!

  3. Anon

    Definitely a “Things I Won’t Squirt Into My Eyes”

  4. Pennpenn

    I would think in some cases I think the problem is for those that don’t respect Trimethylaluminum is that there might not be a next time, at least not for them.

  5. Shion Arita

    Is TMA really that bad? I’ve used it as a solution in toluene on several occasions, and I was always really careful with it, but is there any reason to be any more cautious with it than other pyrophorics like diethyl zinc or tbuli?

    1. wonkychemist in reply to Shion Arita

      I’ve used it neat and in solution (in solution probably a couple hundred times). In solution isn’t bad at all. Neat I actually had an easier time with it than with tBuLi. It didn’t seem that bad, but this was small scale (less than a milliliter), and in dry Colorado air. Still not something I was excited to do.

      1. Lindsay in reply to wonkychemist

        There’s still oxygen in that Colorado air though, right?

  6. steve

    I think the common name for trimethylaluminum is donaldtrumpite – much more destructive than thermite and known to spontaneously combust when there is even the slightest hint of oxygen in the room.

    1. Donald Trump in reply to steve

      Steve I can see the village has found it’s idiot….

      1. Hap in reply to Donald Trump

        “I can see the village has found it’s idiot.”

        Yes, yes it has.

        The next eleven months (and maybe beyond) have “Request for mental bleach, aisle 11” written all over them.

  7. Nick K

    Shion Arita: My experience with both trimethylaluminum and t-BuLi is that that they are equally pyrophoric, but the former is more dangerous as it seems to burn more exothermically. TMA is terrifying and I’m glad I’ll never have to work with it again.

  8. Flatland

    @Shion,

    As someone who has used both TMA as a solution and as the neat liquid, I can say the neat liquid is a whole different level of nasty. Imagine a highly volatile liquid (BP 120C or so) that spontaneously combusts and generates choking clouds of abrasive, lung destroying particles.

    As an undergrad, I got to see the power of neat TMA and exactly how bad things can get. A postdoc was quenching an empty TMA cylinder in the lab. The procedure called for washing/dilution with a high boiling solvent (heptane or toluene) in the glovebox, adding a stirbar, opening outside the glovebox and waiting for the fuming to stop, followed by slow addition of IPA, EtOH, MeOH, then water. The postdoc was in a bit of a rush that day, and after the fuming stopped, went right for the water bottle. I was behind a closed door with those little windows in them, and saw the white flash and I felt the concussive force. I walked into a completely white lab, with a partially hairless and eyebrowless postdoc. This from an empty cylinder. I have an unabiding respect for this stuff.

    PS, want to check the atmosphere quality in a glove box, but don’t have a fancy O2 or water sensor? Open some TMA, immediate vigorous fumes indicate more than sub ppm water or O2.

  9. Chenbo Wang

    I second Flatland. In my grad school we had a glovebox that developed a slight leak. The oxygen sensor was bad so we did not detect it until someone opened a cylinder of neat TMA in the box and it caught fire. I do not recall how the person put out the fire inside glovebox, though.

  10. Isidore

    It was reported that one of those injured in Andover has died from burns.

    1. North Andover Resident in reply to Isidore

      That’s awful. Where was that reported? I haven’t seen any mention of it….

      1. Isidore in reply to North Andover Resident

        I saw it earlier while googling the accident, it was reported in some local paper’s website, but I cannot find it now, so one hopes it was a mistake and nobody is dead.

        1. Sara in reply to Isidore

          No one died. The death was from 2013 and was unrelated to the chemicals they were working with yesterday.

  11. silane

    Shion Arita: Trimethylaluminum is bad. As stated by others it is volatile and exceptionally pyrophoric. When it burns it burns with the heat of a gasoline fire but putting it out is difficult since smothering just presses the pause button and water only makes things worse. The solutions of pyrophoric materials are formulated so that they are less pyrophoric. That is to say they are only pyrophoric under the right conditions. I have had exposed solutions of trimethylaluminum and t-butyllithium and sometimes they do not light up and sometimes they do. I have never had neat trimethylaluminum not light up when exposed to air. Caution should be exercised with most chemistry but you have less wiggle room with neat pyrophoric materials. It is all in the technique since they are same for using dilute solutions of pyrophoric materials as it is with neat pyrophoric materials. You will just know sooner when you make a mistake with a neat pyrophoric material. I personally do not like the organozinc materials. The neat materials are just as touchy about technique but the flames are sootier than organoaluminum or organolithium.

  12. David Newman

    No need to use Al or Li compounds to check out a dry box. Titanium tetrachloride is just as effective and you obtain very pure titanium dioxide deposited everywhere!!!

    When I think of what I did as a grad student in the very early 1960s in the UK; distilling phosphines under nitrogen and watching the small flame outside of he condenser drip nozzle to see when the phosphine was distilling, or performing syntheses using liquid HCN. None of these would be permitted today but we learned how to use dangerous materials under careful supervision.

  13. Anchor

    Derek, I agree with you t-BuLI is a saint in comparison with Tri(m)ethylaluminium!

  14. Me

    I concur: in my first month of PhD, I needed to use neat AlMe3 – it ate through the SureSeal when I put my nitrogen line into it. Luckily the post-doc was watching me like a hawk. He told me (calmly!) to close the sash on my fume hood and walk away. I did just that – he turned up the nitrogen pressure and luckily the needle didn’t fall out of the bottle, so he quickly sealed it up and dumped it.

    Second time I used it was 12 years later, and I soiled myself when I put the needle into the SureSeal (it was fine).

  15. Anchor

    This is safety related question and that is, how do you dispose of those residual organoaluminum agents stored over Tol., hexane etc.?

    1. Me in reply to Anchor

      check out the post by flatland.

      Basically: very carefully. Without a glove box I’d probably go for -78 degrees under a nitrogen line with the solvent washes described there.

      But since I worked in industry for many years, I’d place it in the waste cupboard at which point it becomes somebody else’s problem!

  16. dave w

    The pyrophoric metal alkyls have their uses, as in situations where you really, really need to set fire to something very rapidly.
    The classical example is the startup of a liquid oxygen rocket – you have this sudden flow of very wet, very cold liquids into the combustion chamber, and it would be Dangerous if they were not ignited quite promptly… current practice is to use a mixture of triethyl aluminum and triethyl boron as a “starting fluid”, which reacts on contact with the LOX and provides a nice hot flame for the incoming fuel (e.g. kerosene) to encounter.
    (Usually this material is contained in sealed cartridges with rupture disks, prepared in a special facility – these can be loaded into a fixture on the rocket engine without direct exposure of the reactive fluid; as the engine starts up, the starting mixture is automatically dispensed into the chamber.)

  17. I used the 1M Sure seal toluene solution couple times (there is a easy to do Weinreb’s procedure for mild aminolysis of esters to amides, promoted with stoechiometric Me3Al) and I would not mind using it again, in small volumes. But handling it neat, on process scale – thank you, they aren’t paying me enough.

    1. Hap in reply to milkshake

      Considering what you have done (that you’ve written about), that’s really saying something.

      1. Nick K in reply to Hap

        Seconded. If even Milkshake won’t touch it you know it’s really dangerous.

  18. Daniel Shenai

    I have worked this in another plant at Midland, Mi, in which I analyzed the impurity of metals to the lowest. In order to avoid any hazardous outcome, we used gasoline or Kerosene first, then mixed with Isopropylamine which neutralized the severity of heavy smoke. Since TMA is the one of pyrophoric materials (along with others like TMI), we performed in a glove bag that filled with Nitrogen. After prepared the TMA sample from the hood, we transferred the samples and analyzed in AAS or ICP-OES or ICP-MS. Again since we still considered the sample might react when opening the covers in air, we performed this under glove bag with Nitrogen. One quick note about this is that the TMA sample bottle’s temperature is very hot, but after placing in a glass tray, the temperature turns normal but yet warm.

  19. DDTea

    This thread makes me admire producers of methyl aluminoxane, an activator/cocatalyst for metallocene polymerization catalysts. Its synthesis entails partial hydrolysis of TMA. On a process scale. To the best of my knowledge, only Albemarle and Chemtura do this.

  20. gippgig

    Off topic but may be of interest:
    Cooper Barrett’s Guide to Surviving Life on Fox tonight (Jan. 10) at 8:30 EST – “The guys decide to participate in a drug trial…”

  21. InfMP

    “In 1952 we synthesized in the pilot
    plant of our Institute our first 20 kg
    batch of triethylaluminum—a material
    that could hitherto be prepared only in
    small quantities.
    I have mentioned quite deliberately
    that we synthesized the triethylaluminum
    on the kilogram scale, since this
    leads me to comment on present day
    conditions. Nowadays, if one were to
    work in a research institute like ours
    under conditions that were normal in
    1952, the institute’s director would be in
    court faster than at the patent office.”
    Angew. Chem. Int. Ed. 2003, 42, 5000–5008

  22. Dan

    There is a cute experiment using triethylaluminum to make chemiluminescent clouds:
    http://www.google.com/patents/US3576752
    It is described in this article (German) http://onlinelibrary.wiley.com/doi/10.1002/ciuz.19930270609/abstract
    Maybe not the right experiment for a popular science magazine, but I’ve always wanted to try it.
    If anyone is fearless enough, please post it on youtube!

Leave a Reply

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

Solve the math problem. * Time limit is exhausted. Please reload CAPTCHA.