Overheated gold remains hard after the expected melting point

The study suggests that the heating, which lasted only trillionths of a second, raised the temperature of the gold sample to 19,000 K without melting it.

Scientists say that they heated solid gold for a fleeting moment to a temperature multiplied 14 times by its melting point before it melts - narrowly bypassing the laws of thermodynamics.

Physicists have thought for decades that it is impossible to heat gold and other solid materials more than three times before their melting point without melting them. If the materials exceeded this threshold while remaining solid, the result would be an "entropy catastrophe", a theoretical situation in which the solid phase of the material would have more entropy or disorder than the liquid phase. This would be a violation of the second law of thermodynamics.

But in an article published today in Nature, scientists describe the use of a laser to quickly heat gold to 14 times the melting point - well above the theoretical limit - within a few trillion fractions of a second before it melts and explodes. Gold atoms reach this temperature before they have time to expand and become disordered, so the result does not violate any laws of thermodynamics, says Tom White, a physicist at the University of Nevada at Reno and co-author of the study.

I measure the temperature

If they are heated very carefully, extremely clean samples of some solid materials can be "overheated", reaching temperatures exceeding their melting points without melting. Silver, for example, usually melts at 961.8 ºC, but under the right conditions it can be heated to 986 ºC and remain solid. Understanding overheating is relevant for such subjects as the study of molten planetary interiors.

In 1988, physicists calculated the theoretical overheating limit, which is that no material can be heated more than three times at its melting point and remain solid. In addition to this Limit, a solid substance will have a higher disorder than a liquid. (To understand why this will be a problem, imagine the smooth, hard ice of the rink melting in a puddle of liquid water, which is somehow even calmer.) Attempts to observe overheated solids were hampered by the difficulty of measuring temperature at intervals shorter than nanoseconds.

The team controlled this by heating a 50-nanometer gold sample with a laser until the gold ions began to shake from the kinetic energy. The researchers then took X-ray photons on a gold sample and measured the shift of their energy as they bounced off the material to conclude that the temperature of the gold was within several thousand degrees Celsius.

Using this measurement method, the team noticed that the laser can overheat gold, which has a melting point of 1,337 Kelvin (1064 ºC), up to about 19,000 Kelvin within a few trillion fractions of a second.

Unreliable narrators

In the peer-reviewed reports that the authors shared with the Nature news team, the reviewers recognized the importance of the technique, but widely asked the authors about the reliability of the temperature measurement. One difficulty: the measurement is statistically limited. Only a few X-ray photons - ten out of every ten billion - bounce off the discovered gold.

Arthur Tamm, a computational physicist from the University of Tartu in Estonia, who reviewed the article, points out another problem: reducing photons may not carry accurate information about the temperature of the sample if the atoms in the gold sample are not in a stable thermal equilibrium, a state in which the temperature is balanced. Tamm says that the experiment provides valuable data, but he doubts how the temperature is displayed. "Do we exceed the melting point by 10-14 times before melting? I don't think so," he says.

The team recognized the limitations of the method, but claimed that the result was statistically reliable, and that gold reached balance relatively quickly. White says that his team has found preliminary signs that silver also exceeds the predicted entropy-collapse threshold.

For authors who were interested in how quickly gold would heat up and did not want to break any records, the events came as a surprise. Looking at the data, they were confused, White says: shouldn't the gold have melted? "I just googled: "How hot can something be before it melts?"

If the find can be further confirmed, the search result will require updating.