You can imagine, then, what a difficult decision it was to come back to New Zealand to start a post-doctoral fellowship in a city without an ice rink. And despite the promises of our plucky mayor, my skates still hang unused in the closet.
While we wait for the rink to be built, it’s worth discussing what makes ice skating possible in the first place. One explanation commonly offered is that as water is denser than ice, the pressure of a skate blade will melt the ice underneath, creating a lubricating layer of water on which the skater glides.
However, this doesn’t explain why ice is slippery to walk on, whether or not skates are worn. And sure enough, it turns out that the pressure generated by a skate blade (a few hundred atmospheres) is only enough to lower the freezing point by about 3 degrees C — how do people skate at 4 degrees below zero?
In fact, ice is slippery because it is covered by a thin layer of water even at atmospheric pressure. The first person to suggest that this might be the case was the hugely influential English scientist Michael Faraday in the 1850s. He was intrigued by the fact that ice cubes tend to stick together when brought into contact and guessed (correctly) that this was because the cubes were covered by a layer of water too thin to see, but which freezes when the ice cubes touch.
These days we can measure this thin liquid layer using a variety of modern techniques: at one degree below zero, it turns out that the thickness of this layer on ice is about 100 nanometres, or one ten millionth of a metre, enough to make ice very slippery.
We have also discovered that this behaviour is not unique to water — at temperatures close to their melting points, most solids will be covered by a thin layer of their melt. Bridget Ingham, a colleague of mine at Industrial Research Ltd, recently returned from the Australian Synchrotron with x-ray measurements of the thickness of the liquid layer that forms on indium nanoparticles close to their melting points.
The presence of this molten layer below the melting point also explains why it is impossible to superheat most solids. As the temperature of a solid approaches the melting point, this liquid layer simply grows and grows, until eventually the whole solid disappears at the melting point.
This is in contrast to liquids, which can often be supercooled well below their freezing point. Freezing rain is another phenomenon that Canadians are more familiar with than Wellingtonians — this occurs when supercooled droplets of rain freeze on impact with the ground.
Despite the lack of facilities in Wellington for the study of ice skating, we do have a small company here called Beaglehole Instruments, which makes devices that can measure the thickness of these liquid films. This company was spun off from Victoria University in the 1990s after several physicists there became interested in thin premolten liquid layers. These physicists made some of the first measurements of the thin liquid layer that forms on a metal close to its melting point.
Nanostructures have an annoying habit of melting well below their normal melting temperature, so in nanoscience today, we are very interested in substances that don’t premelt. You can imagine that it is pretty difficult to keep a 10 nanometre ice crystal stable when ice usually forms a 100 nanometre liquid layer at —1 C. And it is the same with metallic nanocrystals, which obviously can’t be used in a nanoelectronic device if they are going to melt when the device heats up after it is switched on.
Hence the interest of another colleague of mine, Nicola Gaston, in very small gallium particles that for some reason are able to remain solid hundreds of degrees above the melting temperature of normal gallium. It seems that there is still quite a bit to be understood about surface melting.
p.s. The first lucky reader to guess which player is me in the photo, or alternatively, upon which part of my anatomy my ice hockey scar resides, gets a free skating lesson when the Wellington ice rink gets built.