Science Demonstrations

I found an interesting paper on the web by Lueddecke, et al.¹ that had a classroom demonstration of the effect of CO₂ on the greenhouse effect. It sounded so simple. Hah! So many things to go wrong, and all the more if you don’t read the paper carefully. I’m still working on how to do this with minimal equipment so I can use this as a demonstration for grades 3–5 (and maybe even a bit younger). It was originally done for high school students. I need it to be foolproof so I can do it without looking like a fool and without ending up making the cardinal sin of having to explain away the bad results. That would send all the wrong messages to children trying to understand how to do science.

You’ll have to read the whole paper for the full experimental protocol. But in order to do this in a small space, I tried to use to 1 pint jars which formerly held spaghetti sauce. I flooded one with CO₂ from a little dust-blower cylinder. The temperature in the jar immediate drops due to the volume expansion of the gas. No problem, I thought, since in the paper it describes a temperature drop from the endothermic reaction (vinegar + baking soda) used to create CO₂. I covered the jars with plastic wrap and Matthew, my 8 year-old, helped me by recording the temperatures every minute for 25 minutes. There was no significant difference in the heating rate; the CO₂ jar stayed about 0.7F lower than the control (plain air) over the whole period. The actual temperature difference fluctuated a bit, first increasing faster in the CO₂ jar, but then slowing again so we started about 0.7F below and ended about 0.9F below and in the middle got as close as 0.4F below.

I repeated using vinegar + baking soda to produce the CO₂ and switched from manual readings of the temperature using a pair of indoor-outdoor thermometers to using Dallas Semiconductor DS18S22 sensors and digitemp to collect data and rrdtool to plot it in real time. Same problem, but with a bigger temperature drop after I added a tablespoon or so of baking soda.

Time to read the paper again (when all else fails, read the directions).

First, covering the container was a bad idea. Seems that Lueddecke found that and printed a nice warning that I glossed over.

Second, just how much CO₂ does that tablespoon of baking soda produce? Well, if you assume the baking soda (food grade) is near 100% NaHCO₃, then it produces about 5 liters of CO₂. This is based on 3 tablespoons = 1 teaspoon and 1 teaspoon weighs about 5.7g (stolen from a chemistry exercise by Dr. Walt Volland). Use the ideal gas law, PV=nRT and the molar weight of sodium bicarbonate of 84g. In any event, once you work through the math, you get about 1.5 liter per teaspoon. Since my jars were 1 pint jars (about 1/2 liter), I produced more CO₂ than the jar would hold (about 9 times more) and since the reaction is endothermic, I just made the jar cold without any benefit of increased heating rate.

Third, I didn’t do the suggested control of testing the heating rate in the two before adding the baking soda. The idea is to make sure when turn on the lamp, they heat at the same rate.

So, I set up with one jar having vinegar, one empty, and let them come to equilibrium before starting. Hmm, one of the jars is cooler than the other. The probes has already been calibrated and were known to be consistent to within 0.1C, but the temperature difference was very consistently 0.2–0.3C different. My wife suggested evaporative cooling in the jar with vinegar, but I had already ruled that out since it was cooler in the "air" jar. Oops, I had labelled them incorrectly when I cleaned the jars and reassembled everything. It was evaporative cooling in the vinegar jar. So I put a little vinegar in both and watched both of them cool to the same 0.2–0.3C below room temperature (I have four probes, two are sitting on the table to measure the room temperature near the jars).

Finally, I tried again. Add 1/4 teaspoon of baking soda and start the experiment. The CO₂ jar showed an immediate temperature drop, but only about 0.3C. For the first eight minutes, the air jar appeared to be heating faster. At that point, the heating rate for the air jar appears to slow while the CO₂ jar continued upward. By the end of 20 minutes, the CO₂ jar was nearly 0.5C warmer. Jonathan, my 6 year-old, walked past the CO₂ jar which showed a dip in temperature for 3 minutes before it began rising again. I think this was just extra air currents. At the end of 55 minutes, the temperature difference was up to 0.8C at which point I turned off the lamp.

I think the initial slow rise in the CO₂ jar during this last experiment was due to extra evaporative cooling directly on the sensor. Because my jars are small (compared to the 2.5 gallon aquariums used by Lueddecke), I get splashing onto the probes when adding the baking soda. I think this causes the heating rate to appear lower in the CO₂ jar because I’m initially evaporating liquid off the sensor which cools it somewhat even as the temperature in the jar is rising. I either need larger containers or else a shield to prevent the splashing. Or maybe just mount the sensor higher, but the problem is it really needs to be down in the CO₂ layer. Remember, the goal is to make this simple to execute and understand. The more complex the protocol, the more the message will get lost….

So, what did I learn?

1. Bigger lamps are good. My 60W bulb is much smaller than Lueddecke’s 150W heat lamp.
2. Bigger containers are good, they hold more CO₂ which should enhance the effect.
3. Don’t make more CO₂ than the container will hold. It just causes extra cooling (from the reaction) with no benefit.
4. Evaporative cooling is an important effect which needs to be minimized. Protect your probes.

Lastly, Lueddecke et al. used larger containers with a bed of dark material on the bottom and a small dish for the reactants. Glass aquariums are nice, but hard to store and are fragile. I want to try this with plastic contains around 1 gallon in size. It doesn’t really matter if they are not fully transparent, at least not for the experiment itself. I’m not sure how much the psychological effect of being able to see into the containers is worth. What I need to mostly is to work out a protocol that gives me that 2C temperature differential in 10 minutes. If I can do it with the indoor-outdoor thermometers, that’s worth something since it makes the equipment list more accessible to grades 3–5.

¹ Lueddecke, Susann B., et al., Journal of Geoscience Education, v49 n3, May 2001, pp 274–279.