~ A break in an unusually warm Hungarian winter allowed Monika Landy-Gyebnar to indulge in some frozen bubble making. Quite cold weather is needed because glycerine in the bubble solution depresses the freezing point below zero Celsius.
Images ©Monika Landy-Gyebnar, shown with permission
"I used dishwasher liquid (one for hand washing dishes), some glycerine and water. Nothing special.
The bubbles took about 3-4 minutes to freeze when it was coldest outside. Later as the temperature rose the freezing became much slower. Of course it took a lot of patience as many bubbles popped before freezing.
Many people have tried it in Hungary during the cold days and in the east where it was much colder (even -18°C was reached) it was easier. One girl blew the bubbles on snow surface and that made it freeze really quickly. I had no snow when it was cold, and tried many surfaces, both the grass and the lid of our waste bin worked well.
I’ve seen a lot of frozen bubble pics from really cold locations (Canada, Finland etc.) and the fern pattern on those bubbles differs a lot from our results in less cold weather. The colder it was the ferns became smaller and more even shaped (small 4-point star shapes were common). Of course I don't know if the mixture was similar or the type they used was the cause of the pattern difference, but in temperatures below -20°C the patterns were very similar despite very different locations and bubble blowers."
Water molecules need a nucleus* – dust, a snow machine protein, bacteria or other particles on which to start the phase change to ice. Once a tiny crystal forms, it spreads across the thin bubble sheet. Growth is fastest at crystal edges. It is dendritic, treelike, along the hexagonal crystal directions to give the characteristic patterns. The complex but overall symmetric shapes of snowflakes are similarly produced.
*Thought, would slightly dusty bubble mixture start to freeze faster?
**At very low temperatures no nuclei are needed - there is spontaneous nucleation.