Water boils at 100ºC and the ice in your iced latte is at 0ºC. These are facts that we think we know about water: it boils at 100ºC and it melts at 0ºC. A sharp observer may point out that these are pressure dependent and that if we were at the top of a mountain, the water would boil at a slightly lower temperature (I once had a student argue that this was a good reason to only ever drink green tea at high altitude). But if we are at ground level and it is a normal day, we will be fairly certain that the water for our coffees would boil at 100ºC and ice would form at 0ºC.
Yet these ‘facts’ hide some complicated physics and some oddities about our planet. Pure water, that is, water without any impurities in a clean vessel (such as a clean, scratch free glass) does not boil at 100ºC but at temperatures significantly higher than that. Nor does pure water freeze at 0ºC but at temperatures significantly below that. These are phenomena known as superheating and supercooling respectively and, if you are observant, you could see them occasionally in your coffee cup. To see why, and how, we need to think a bit more about how water freezes.
If you fill an ice cube tray with water and put it in your freezer, you would expect ice cubes to start forming at about 0ºC. We expect the freezing temperature to be the same as the melting temperature, that is the temperature at which the ice cubes would melt. And yet, if you make the water very pure (even distilled water would be a start) and put that in a clean, defect free container (such as a clean glass jar) in the freezer, the freezing process will not begin until much lower temperatures. It’s because the water has to crystallise and change state from a liquid to a solid and to start this process, there needs to be a seed, a surface on which the ice can form. Called a “nucleation site”, this seed could be a piece of dust, a small impurity in the water, a scratch on the surface of the container holding the water, or in fact anything that allows the bonds of ice to start to form. The same is true at the other end of the temperature scale. When the liquid water turns into steam, nucleation sites are needed so that the gas bubbles can start to form at those sites. In the absence of impurities in the water, the water will not boil until temperatures high above 100ºC.
Fortunately in tap water, or in your super-filtered water that you make your coffee with, there are plenty of such nucleation sites so the water boils and freezes at roughly the temperatures you’d expect them to. The same is not true however for clouds in the sky where some (high altitude) clouds have been shown to contain water droplets that are at -35ºC, well below the “freezing temperature”. Exactly why this occurs is still puzzling and a topic of research, but when you stop and think about it, how would you actually measure this temperature? If you supercooled a cup of water and then put a thermometer into it, the thermometer would provide a nucleation site and the water would immediately freeze. How can you measure the water’s temperature without a thermometer?
Recently a study reported in Physical Review Letters used a laser to measure the diameter of a series of supercooled liquid droplets by determining the energy of a resonance that depended on the droplet’s size. To calculate the temperature of the droplet, the authors then used the principle that as water evaporates, the droplet from which it is evaporating will become colder at the same time that it shrinks in size. Measuring the size of the droplet allowed them to calculate the evaporative loss and therefore the temperature of the drop. They double checked this new technique by measuring (with the same laser) the energy of a particular atomic bond in water that has a known temperature dependence (at higher temperatures). The temperature determined from the drop’s size corresponded with the extrapolation of the energy of this atomic bond and so the team were fairly confident that they had measured liquid water to very cold temperatures indeed. In fact, the authors suggested that it was still possible to have liquid water at 230.6±0.6 K which, in more every-day units corresponds to -42.55ºC, well below the nominal ‘freezing point’.
So pure, liquid, water can get very cold indeed. But could you ever see this in your coffee cup? Although you may like to try some experiments with freezing ultra-pure water, it is easier to see the phenomenon of superheating in your coffee. However, given the possibility of an accident, it may be safer to watch the effect on the video below. The idea is that if you put very pure water in a clean cup into a microwave, it is possible to superheat it well above 100ºC without it boiling, because there are no nucleation sites in the cup or the water on which the steam bubbles could start to form. When you take the cup out and put a nucleation site in (perhaps a spoon or maybe even instant coffee granules), the water will boil suddenly as a result of those new nucleation sites and can even explode. Obviously if you were anywhere near the water when this happened you could get seriously burnt and so it is probably safer to watch the Mythbusters do it with their robotic arm. Enjoy the video, enjoy your coffee, preferably far from superheated: