It’s well-known that many liquid metals can be cooled below their freezing point. This is, scientists assume, due to dense and symmetric, but non-periodic ordering within the liquid. This theory implies that the freezing point of supercooled metal liquids can be controlled, just like crystallization can be induced by a template – all it takes is something that stabilizes the short-range order of the liquid. In the latest issue of Nature researchers from Grenoble published an impressive proof of this concept, not only showing that phase transistion temperatures depend on several variables, but also revealing the structure forced on the liquid.
The material in question is an eutectic mixture of gold and silicon, which melts at 636 Kelvin. The mixture forms from about 200 namometer large gold clusters on a silicon surface. Upon heating silicon atoms migrate into the cluster until the eutectic composition is reached and the cluster melts at 636 K. This is where it gets interesting, because, as you already may have guessed, when cooled the drop doesn’t re-solidify at 636 K. Instead the freezing point depends on the surface and how the liquid is treated after melting. If the temperature is reduced again immediately, the liquid solidifies at 563 K, but if the eutectic mixture is kept more than 40 degrees above its melting point for some time and then cooled, it stays liquid down to 513 K. That’s 120 K below the melting point. And it gets better. As the researchers point out, the composition of the mixture changes during cooling: Si is frozen out, so the drop is enriched in gold. At 513 K the mixture actually has a solidus Temperature over 360 K higher. You couldn’t supercool water to this extent, because absolute zero would get in the way. Amazing!
The researchers then had a look at the boundary layer where the liquid meets the solid. And that’s where the action is. Apparently the gold slightly distorts the silicon surface, and in turn the Au atoms of the liquid are forced into a pentagonal structure that is even denser than the normal gold crystal. Pentagonal arrangements are present in many liquids, and the researchers conclude that surface arrangements that stabilize such pentagonal clusters may be a universal trick to induce supercooling in liquids.
Supercooled liquids are highly sought-after in material science, where researchers strive to discover ever new crystal structures. So this discovery will certainly come handy to engineers of all sorts, but what I like best about this paper is how exotic it all is. The phase transition temperatures are reproducible over many cycles, and that means this is not like supercooled liquid as we used to know it, always just one knock away from suddenly freezing. This is rather like an entirely new kind of phase change (albeit not in the strict thermodynamic sense).
And this, I suspect, is just the beginning. There are few areas in chemistry as versatile as surface chemistry. Even in macroscopic surface chemistry, researchers barely, well, scratched the surface. There is no telling what other odd effects chemists will discover next.
Schülli, T., Daudin, R., Renaud, G., Vaysset, A., Geaymond, O., & Pasturel, A. (2010). Substrate-enhanced supercooling in AuSi eutectic droplets Nature, 464 (7292), 1174-1177 DOI: 10.1038/nature08986