Chemistry Lab Demonstrations: Upsidedown Thunderstorm

*For more cool stories, pictures, and videos of chemistry demonstrations, click here*

SN2 lab today.  Preparation of (2-methylphenoxy)acetic acid from o-cresol and sodium chloroacetate.  NaOH deprotonates cresol.  Add sodium chloroacetate and reflux.  SN2 reaction occurs.  Acidify, collect the precipitate, and recrystallize from water.  Unfortunately, today’s lab took a really long time.  I’m not sure why, but it did.

Anyway, I looked around for cool SN2 demos… and there really aren’t any.  Sure, I can do the Finkelstein at different concentrations and show that one precipitates NaCl faster than another because it is more concentrated.  Woo hoo.   That’s not very exciting.  So I ultimately decided that today’s lab just isn’t going to have anything to do with the SN2 reaction.

I decided to do the upsidedown thunderstorm.  That’s my name.  I’ve seen ‘thunderstorm in a test tube,’ but that’s not a very flashy name, now is it?

Here’s the deal:  Add a layer of con’c sulfuric acid to a test tube (use all proper precautions for handling con’c acid!).  Slowly and carefully add ethanol down the side to create an ethanol/acid biphase (if you leave it sit too long, the acid and ethanol will mix, so don’t let that happen).  Add a few crystals of finely-ground potassium permanganate.  The reaction occurs, and evolves a gas.  The gas bubbles through the ethanol layer and looks like upsidedown rain.  After a few seconds, “lightning” appears at the phase boundary, and the sulfuric layer becomes cloudy and purple (the storm clouds…).  The ‘storm’ continues for several minutes.  Manganese waste should probably not be thrown down the drain, so if you do this, dispose of it according to local regulations.

There appears to be a few different processes leading to ‘lightning’ formation.  Permanganate is converted to the exposive Mn(VII) oxide.  Additionally, ozone is created.  The ozone oxidizes the ethanol resulting in combustion of the organic material.  Wikipedia explains:

Concentrated sulfuric acid reacts with KMnO4 to give Mn2O7, which can be explosive … Potassium permanganate and sulfuric acid react to produce some ozone, which has a high oxidizing power and rapidly oxidizes the alcohol, causing it to combust. As a similar reaction produces explosive Mn2O7, this should only be attempted with great care. An approximate equation for the ozone formation is shown below.

6 KMnO4(aq) + 9 H2SO4(aq) → 6 MnSO4(aq) + 3 K2SO4(aq) + 9 H2O(l) + 5 O3(g)

When I was practicing the demo, I wanted to see if I could scale it up some to make it more visible than just in a test tube.  I tried it with about 20 mL each of the acid and alcohol in a beaker.  The detonations were sufficiently energetic (and perhaps the ethanol layer was not tall enough) that the ethanol caught fire.  I had a cute little ‘sterno’ burner going.  For the actual demo, I switched to an Erlenmeyer flask which helped.  No fire this time.

Storm In A Test Tube !


  1. This is a great demo! Still, I would be very careful scaling up this reaction. Also, it’s better to use something like a graduated cylinder rather than a beaker because the surface area between the two plates is much smaller–this limits the possibility of a runaway. Be sure to use big crystals of permanganate.

    Having done this demo numerous times, I was always uncomfortable scaling it up above 20 or so mL ethyl alcohol. I am under the impression that it is Mn2O7 that actually causes the thunderstorm appearance, not ozone (although it could be a combination of ozone and the heptaoxide). I know that both highly concentrated ozone and dimanganese heptaoxide both spontaneously detonate with organics. Having worked with both ozone and Mn2O7, I can say that I’d rather work with ozone.

    Curiously enough, Wikipedia claims that Mn2O7 is a red crystalline solid. I never obtained a red crystalline product, mine was always a nice green which fumed red HMnO4.

    • I actually noticed that bigger crystals seemed to correlate with bigger ‘detonations,’ so I stuck with smaller crystals. I agree with the grad cylinder approach. It was a thought I had right before I started, but by that point, I was all in. If I do it again for an audience, I will do that.

      I also think it is a combination of ozone/heptaoxide which causes detonation. I think it is the hetpaoxide which physically detonates, perhaps assisted by the ozone?

      Thanks for the tips.

      • You had a faster reaction rate with larger chunks with less relative surface area? Hm, I always had mine go smoother with bigger (as in 5 mm large chunks) permanganate crystals.

        Manganese heptaoxide can spontaneously detonate. I know this because I have experienced this. It will also detonate with a loud report when it touches any easily oxidized organic like acetone, MEK, alcohols. It is relatively stable in carbon tet, but that too has a limited shelf life.

        Mn2O7 is an oily green, iridescent, red-fuming, ozone-reeking liquid.
        I’ve made it in small quantities on quite a few occasions. The last time I ever made it (or ever will) I did with a fellow enthusiast, and we followed a literature preparation. Unfortunately, it was rather hot and humid out and the material exploded in our faces. Luckily, we escaped with minor burns since we were wearing protective gear (face shield over top goggles, also an apron). It is possible to isolate it.

        I think what happens in the demo is that the M2O7 formed dissolves in the sulfuric until it reaches some critical concentration where it reacts at the interface between the alcohol and the sulfuric.

        I recommend you to my friend Wilco’s page. He has done some small scale experiments that will definitely impress your students.

        If you’ve not ever been to his website, I highly recommend you check it out.

        Hopefully this isn’t a double post–the last one didn’t go through for some reason.

        • You had a faster reaction rate with larger chunks with less relative surface area?

          Not faster… bigger, larger, more violent detonations that threatened to erupt out of the test tube and all over the place. Smaller crystals gave more controlled ‘lightning’ flashes that behaved nicely. At least that was my experience the 4 or 5 times I’ve ever run this demo.

          Thanks for the link. I’ll definitely give it a read.

  2. No video of the fire?

  3. Cool demo. Ozonolysis rxns are typically performed in MeOH or some combination of MeOH and CH2Cl2 – usually at -78. If O3 reacted so violently with EtOH it should with MeOH as well. The Mn2O7 is probably doing the oxidation/explosions.

  4. Hey! It’s the greatest experiment which I have seen.
    I will have a competition about Science Show.
    Do you know Science Show?
    We do the laboratory on the stage, show to the audience.
    Is this the same experiment as the one I found on
    Thanks for the good experiment like this.

    P.S. am Thai student, so if I write the wrong grammar I’m sorry about that.

  5. Thanks for the write up on this demo, I’d some how managed to miss it previously.

  6. sean gallyout says:

    can you give the chemical formula for this whole reaction , you gave formula for reaction between only KMno4 and sulphuric please tell me all formulas for chemical reactions that occur in this experiment.

  7. This experiment caused the end of my chemistry lab back in 1980. I was 12 at the time and really into science. When my parents weren’t home I made all kind of funny stuff with potassium chlorate, potassium permanganate and the likes. The “thunder in a test tube” was a great success. The only trouble was that the fumes turned the stone bath tub I was doing the experiment in black! No amount of scrubbing helped so I had to confess to my parents that it was my experiment that caused the tub trouble. I had to throw all my stuff away and promise never to do these experiments again. Can’t wait till my son is old enough so we can do it together though…

Leave a Reply to Rain Cancel

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