Replication of high-temperature superconductor comes up empty

On Monday, the journal Nature released a report from Nanjing University researchers that had attempted to replicate an earlier paper that described a compound that superconducted at room temperature and relatively moderate pressures. Despite persuasive evidence that they've produced the same chemical, the team indicates they see no sign of superconductivity, even down to extremely low temperatures.

The failure will undoubtedly raise further questions about the original research, which came from a lab that had an earlier paper on superconductivity retracted.

Bold claims

The work is part of a growing body of literature on metals complexed with hydrogen. These hydrogen-rich chemicals can form at ambient conditions, but added pressure can force additional hydrogen atoms into the structure. The resulting high-pressure compounds have ingredients—spare electrons from the metal, light nuclei from the hydrogen—that are thought to favor the formation of Cooper pairs from the electrons, a key ingredient in superconductivity. And a number of hydrogen-rich chemicals have been found to superconduct at over 200 K (-75° C) if the pressure is high enough.

In 2020, the lab run by Ranga Dias at the University of Rochester reported a carbon-hydrogen-sulfur compound formed at extreme pressures could superconduct at room temperature. But the results were controversial, partly because it wasn't clear that the paper included enough information for anyone else to produce the same conditions and because Dias was uncooperative when asked to share experiment data.

Eventually, it became apparent that the team had used undocumented methods of obtaining some of the data underlying the paper, and it was retracted. But Dias continued to claim that the superconductivity was present. (There's a good overview of the controversy on the American Physical Society website.)

Despite Nature retracting one of Dias' papers, the journal published another paper on superconductivity from his group. In this case, a lutetium-hydrogen chemical doped with nitrogen was reported to superconduct at room temperature but at much lower pressures, which could allow it to be tested with somewhat less specialized equipment. Given the history, the claim was greeted with an even higher degree of skepticism than the earlier paper.

Trying again

Given the details in the paper, a group at Nanjing University attempted to form the same compound that reportedly superconducts. And all indications are that they have. Data taken using X-rays at room temperature indicates that the structure of their hydrogen-lutetium-nitrogen compound is very, very similar to the structure reported by Dias' group. And the Nanjing team could see the same signals in Raman spectroscopy, which identifies the vibrational frequencies associated with a chemical's bonds.

The compound even showed the same changes in color that occurred at high pressures. All indications are that they're working with the same chemical that Dias' group generated for its paper.

The differences came when the chemical's electrical resistance and magnetic behavior were measured. Superconductors have a critical temperature at which they experience a sudden shift. As a superconducting chemical is cooled down, resistance to current drops suddenly, and the magnetic behavior changes at this point. When the chemical was tested in Nanjing, it showed no sign of any transitions—all measurements with the chemical showed smooth curves instead of sharp drops.

This was true even though the Chinese researchers tested the chemical down to temperatures of just two Kelvin.

This doesn't definitively show that the initial report is wrong. The nitrogen is present in small quantities as a dopant and may not have a significant influence on the structural measurements done in this new report. The Nanjing team also indicates that it's distributed a bit unevenly throughout the chemicals they tested, which might potentially explain the differences in the behavior of the chemical.

The rapid publication of this report, however, should put pressure on Dias' group to cooperate to try to sort out any differences that might explain the lack of replication. The degree to which they cooperate will probably say a lot about where we can expect things to go next.

Nature, 2023. DOI: 10.1038/s41586-023-06162-w  (About DOIs).