We're wanting to generate hydrogen to fill balloons and possibly make flammable soap bubbles.

Our plan is to react aluminum metal with water, in a solution of NaOH (lye) to catalyze the reaction by dissolving the transparent "rust" that forms on the surface of aluminum metal.

But we expect this reaction to produce heat. If it produces too much we'll have problems, because the water might boil (adding heavy water vapor to our output) as well as giving us a hot caustic mess which might splatter, break the reaction vessel, etc.

So we are first going to try to figure how much heat to expect by doing small scale experiments and carefully monitoring the temperature. We're using rubber gloves and goggles while doing these tests. We're also doing the tests near a water tap; if we get any lye on us (or even think we have) we'll flush it off with lots of running water.

For the first test, we made a solution of 10g of NaOH in 200ml of water, and cut 1cm squares of aluminum from a pop can. Then we put 20ml of lye water into each of three flasks. We had to wait for it to cool, since just dissolving the NaOH in water generated a fair amount of heat.

We added one square of aluminum to the first flask, two to the second, and four to the third (doubling). If the test had succeeded (but not produced too much heat) we planned to do a second round with 8, 16, and 32 square cm of aluminum. Instead we saw:

...almost nothing. There were no visible signs of a reaction, no sound (hissing, bubbling) and little or no heat generated.

We considered several possibilities:

• We were wrong about the reaction
• There was something wrong with our lye water
• There was something wrong with our aluminum
• There was something wrong with our technique

So made three balls of aluminum foil (small, medium, and large) and added one to each flask. They were supposed to contain about the same amount of aluminum as the squares in the flasks, but we did not weigh or measure so it might have been way off.

This time, we got some reaction!

We got clear bubbles and temperature change, but not the data we wanted.

• Since we were just trying to verify the reaction and our materials, we hadn't weighed the foil.
• Our temperature reading were pretty erratic (the graph above shows a rolling average and is still pretty jiggly). We had several ideas about this:
  • The 20 ml of solution, which had been enough to cover the 1cm squares, wasn't covering the aluminum balls so the amount of aluminum that was able to react was changing as they moved around
  • The IR thermometer we were using gave changing reading (possibly depending on if it hit the liquid, the foil, or the flask?)
  • With so little liquid, the heat being produced could change the temperature suddenly
  • The wind may have been blowing the heat from x4's flask towards x2's, and from x2's flask towards x1's.
• The problem was with the aluminum?
  • Maybe aluminum cans aren't really made of aluminum?
  • Maybe it is aluminum, but it's coated with plastic or something?
  • Maybe we didn't have enough aluminum squares to start with?

So we tried again with weighed-out bits of aluminum foil in 50ml of the same NaOH solution, using glass chemistry thermometers with a good separation between the flasks. Now the graphs are starting to look more like we expected:

We had a few minor issues with the data collection; the thermometers disagreed slightly when we tested them in the same liquid, and for the first several reading we were misreading one of the thermometers (it's marked every 2°C, not every 1°) but we quickly got these sorted out.