When gold is heated really quickly, it seems it remains solid at temperatures far beyond the point where it should have become liquid, a new study in Nature has found.
The term for when a solid stays a solid at or beyond its melting point is superheating. Most materials can be superheated only in a short range after that point, before they promptly melt. Scientists used to think this range was fixed because of a limit called the entropy catastrophe.
Entropy is a measure of disorder in a system. When you heat a substance, its entropy increases (among other attributes). Previously, scientists thought that if you heated a crystal to about three-times its melting temperature, it wouldn’t be able to stay a solid any more: it’d have to melt because its atoms would have become too disordered.
Entropy catastrophe
In 1948, an American chemist named Walter Kauzmann flipped this. He found that when he continuously cooled a liquid to under its melting point but at the same time prevented it from crystallising, the amount of entropy in the liquid would be less than that in a crystal of the same material beyond a specific temperature — which shouldn’t be possible. This came to be called the Kauzmann paradox.
Four decades later, Hans-Jörg Fecht from Germany and William Johnson from the US flipped this once more. They reported that when a solid was superheated to around three-times its melting point, it would eventually possess more entropy than its liquid form beyond a particular temperature, which is another impossibility. This temperature was called TEC, where EC stood for “entropy catastrophe”.
Both these outcomes are “catastrophic” because of the second law of thermodynamics, which states that in an isolated system evolving spontaneously, the entropy can’t decrease over time. For two phases at the same temperature and pressure, the phase with higher entropy is (loosely speaking) the more disordered one. As the German physicist Rudolf Clausius interpreted this law, the entropy of an isolated system can’t spontaneously decrease — yet that is what the entropy of a solid being higher than that of a liquid implies.
The “catastrophe” is thus a warning that extrapolating to those problematic temperatures in the Kauzmann and Fecht-Johnson experiments doesn’t enjoy thermodynamic legitimacy. It’s also a sign that something happens before those temperatures to prevent the impossible outcomes.
Heat it quickly
For example, Kauzmann found that the liquid would either crystallise first or that it would turn into a glass well before it reached the “catastrophe” temperature. This avoidance is why every ordinary piece of glass you come across — like the one on your windows, say — forms around a glass‑transition temperature that’s noticeably higher than the problem temperature. Similarly, a crystal melts long before its “catastrophe” temperature or simply vaporises.
The new study with gold explores what happens to these expectations when the metal is heated very quickly.
Understanding the limit of how much heat a solid can imbibe without changing its phase (i.e. turning from solid to liquid) is important for engineers to design materials that work in extreme environments, such as on planets with brutal atmospheres or in facilities that manufacture them using punishing physical conditions.
As with a lot of research of this type, the new study used a simple process but wasn’t possible to conduct until now because the technologies required have only just become accessible. The researchers, from Germany, Italy, the UK, and the US, used powerful laser pulses to heat gold films about 50 nm thick. They used the lasers in order to heat the gold rapidly, without giving it time to disintegrate or liquefy. Each pulse lasted only 45 femtoseconds and was just 400 nanometres long.
Then the team used a technique called high-resolution inelastic X-ray scattering to determine the gold atoms’ temperature. A device produced and emitted streaks of X-ray radiation, which struck the gold atoms and scattered off only a few picoseconds after they’d been heated. By measuring the changes in energies of those X-rays and the directions in which they emerged from the nanofilms, the team could deduce how fast the atoms were moving, and from that figure out the temperature. (The temperature of a material is simply the average kinetic energy of its constituent particles.)
Older results stay
Thus the team found that solid gold superheated to 14-times its melting point — leagues beyond the three-times limit — remains solid for a few trillionths of a second, which is a significantly long time in the microscopic realm. The X-ray diffraction patterns revealed the atoms were still arranged in the ordered pattern typical of solid crystals.
According to the researchers, the rapid heating could overtake the effects that came with heating more slowly. This isn’t a gimmick so much as a signal that if a material is heated rapidly enough, there may not actually be an “entropy catastrophe”. The ultrashort laser pulses made sure the gold atoms didn’t have time to “relax” before the X-ray instrument came on, revealing the nanofilm to have been solid even at a temperature where melting was expected to be unavoidable.
In fact, when the researchers calculated the nanofilms’ entropy in conditions where the films lacked the time to expand due to heating, they found that the films themselves could never reach the classical catastrophe temperature.
The findings challenge materials scientists’ core assumptions about how matter behaves at extreme conditions. At the same time, they don’t invalidate the work of Kauzmann and Fecht and Johnson: the latter two assumed the material they worked with could expand when heated whereas the new study didn’t allow for that possibility.
Nonetheless, the implications could go beyond the earth. For example, certain substances may be able to survive in the cores of planets or on stars in a particular phase for longer than what models have predicted. Such details may come to light when scientists apply the technique in this experiment to more materials.
Published – July 24, 2025 11:52 am IST
