In this column, Cecil explains the function of nucleation sites in the formation of CO2 bubbles. He also mentions their imporatnace in the formation of other things, such as snowflakes. What would happen if water was boiling in a container so perfect that no such sites existed. For that matter, what would happen to a snowflake if no dust existed in the atmosphere to accommodate its formation?
You can get superheated water, water which is hotter than 100 Celsius but remains precariously a liquid. This often happens in a microwave oven; when the water is taken out, any slight bump or jar can cause it to tip over that precarious balance, and a lot of water explosively boils at once.
What about the snow and rain question? If the atmosphere gets fully saturated, but there is no dust or anything else on which droplets or snowflakes would form, what would happen?
This can also happen to filter coffee, so that you think it’s at a drinkable temperature, but when you tip it into your mouth it is in fact hotter than 100C, and it boils in your mouth, and you spit it into the sink, but later an entire layer of mucus membrane sloughs off the roof of your mouth. YMMV.
It is also possible to have supercooled water. It is below 0 °C, but still a liquid. Any disturbance, such as striking your windshield, will cause it to immediately freeze, as in freezing rain.
About the snowflake, this is thought to be very common in the atmosphere. However, there is still a bit of dust (usually having some soluble ammonium sulfate) to get a droplet to form first. As Rhubarb pointed out, the drops remain supercooled, that is, they exist below the freezing point 32 F or 0 degrees C. Some of these ‘supercooled’ droplets often get carried up very high in the atmosphere to where they spontaneously freeze at about -40 F (also the same as -40 C). In other words, there are often two nucleation events for a snowflake (actually, ‘snow crystal’) to form: the one to get the droplet, and the one to get the crystal. When the second one does not occur, the droplet freezes when it reaches the height at which the air is -40 F.
What happens after a droplet freezes at -40 F? Probably most of these become a type of ice crystal called a ‘bullet rosette’. I mentioned these odd crystals in my previous post “Snow crystals: uniformity is not enough”.
Jon
Plastic bottles like the ones shown in the supercooling video have LOTS of imperfections, and I freeze water in such bottles regularly. How did they supercool the water for the videos?
Have you actually studied the inside surface of such a bootle under sufficient magnification to show the kind of imperfections required to serve as nucleation sites?
I suspect the insides of, at least some, such bottles are actually close to perfectly smooth, and that all you need is to use distilled water like the description for the video says.
Supercooling a few milliliters of water by 5 or 6 degrees F requires neither perfectly smooth containers nor distilled water. I haven’t seen this video, so I don’t know how much water they had, but a fellow called Dorsey did extensive studies of supercooling about 70 years ago, and he showed that it was actually very hard to get 5-10 ml of dirty river water to freeze within 5-6 degrees F of the freezing point. I don’t have his report handy, so these numbers are based on my (often faulty) memory, though I think they are conservative.
Atmospheric scientists have looked far and wide to find materials that can freeze water with 1-2 degrees C of freezing. Last I saw, the best ‘nucleant’ was ineffective above about -1 C. Silver iodide, a material often used for cloud seeding, is very good for freezing water, but even this is ineffective above about -3 C. The point is, it is hard to get water to freeze near the freezing point. Supercooling is the rule, not the exception.
Jon