During thesis defense in Prague, one professor asked me: “The experimental procedure on page 64 is describing an explosion - please would you care to elaborate more on this?” I found myself explaining that the material, racemic glycidol, was made only once for the project and there was, uh, a mishap as described – and I got enough material to carry forward so I had no urgency to re-do the preparation under more controlled conditions… What I did not tell them then was that the volcano which erupted three years before was set up with blessings of a man working for the said professor.
I needed glycidol and I couldn’t afford to buy it from Aldrich - so I made a whole bunch of chloropropane diol by hydrolysis of epichlorohydrine and I was going to treat it with methoxide. As I couldn’t find a reliable lit procedure for my particular substrate and being lab-naive, I asked a faculty dude about a general chlorohydrine ring-closing conditions. He suggested that I should use chloroform as a solvent for the reaction, and run it at 0C. “But isn’t chloroform supposed to react with methoxide?” I asked him. “Sure it will - that’s the beauty of the system, the solvent reaction will mop-up any unreacted methoxide so it won’t ruin your epoxide. We in sugar chemistry are using this trick all the time”.
Alrighty, then: I dissolved the chlorohydrine thing - all 136 grams of it - in chloroform 0.5L and slowly added 1.5 mol of methoxide solution. The addition was not exothermic; in fact the reaction seemed calm like a lamb. About 3 hours later I removed the cooling bath and as I was walking away suddenly a thermometer flew by. There was a whooshing sound while the geyser commenced - from the flask neck where the addition funnel was just moments before… I slammed the hood sash down and bolted for the door. Later I found that there was enough left in the flask to proceed with the workup (alongside with the cleanup). The chloroform/methoxide trick did not work as advertised though: there was lot of methoxypropane diol in my glycidol and I had to do a fractional vacuum distillation on Vigreaux to obtain the pure stuff, in 8.5% yield.
Few years later I visited the lab and looked into that hood and sure enough, the dried up white cake of salts - round and pretty like a giant camembert - was still there, hanging from the ceiling…