A result last year hailed as a breakthrough in physics also generated skepticism that has now escalated into angry recriminations. Researchers said they had made the first superconductor that works at room temperature, a long-sought goal. But Jorge Hirsch, a physicist at the University of California (UC), San Diego, attacked some of the evidence, particularly a set of magnetic measurements. He says his requests to see the underlying data have been rebuffed by the authors for nearly a year. And, last month in a peer-reviewed paper, he charged that the results are “probably fraudulent.”
Ranga Dias, an applied physicist at the University of Rochester, who with his colleagues made the room-temperature superconductivity claim, rejects Hirsch’s allegations. He asserts that Hirsch isn’t an expert in high-pressure physics and that he has a history of claiming that the Nobel Prize–winning “BCS theory” underlying superconductivity is incorrect. Dias says Hirsch relentlessly badgers superconductivity researchers. “Hirsch is a troll,” Dias says. “We are not going to feed this troll” by providing the data.
Superconductivity is normally seen only at temperatures well below 200 K, or –73°C. But several research groups working with hydrogen-rich compounds called hydrides have claimed that they became superconductors between 200 K and 250 K when squeezed to intense pressures. Dias and his team went further. They reported that by adding a bit of carbon to precursors that make H3S, a known hydride superconductor, they were able to create a carbon sulfur hydride (CSH) material that pushed the superconducting temperature up to 287 K (nearly 15°C), the temperature of a cool room. The result, published in the 14 October 2020 issue of Nature, generated worldwide acclaim.
Some scientists attempted to replicate or extend the finding, without much success. And Hirsch and others raised concerns. Like other superconductors, CSH showed a characteristic plunge in electrical resistance as it dropped below the “critical temperature” (Tc) and became a superconductor. To confirm that a material superconducts, however, physicists also look for a second telltale indicator, known as the Meissner effect, in which the material expels magnetic fields below Tc.
Measuring the Meissner effect hasn’t been possible in hydrides because they are formed in minute amounts inside a high-pressure device called a diamond anvil cell (DAC), which normally is made from magnetic materials. So, hydride researchers have instead evaluated a property known as AC susceptibility, a measure of how much a material becomes magnetized in an applied magnetic field. In Dias’s Nature paper, CSH’s AC susceptibility dropped sharply at Tc, consistent with the interpretation that the material was expelling magnetic fields.
But the data also show that as the material cools below Tc, the AC susceptibility rises again. That’s a behavior not usually seen in superconductors, Hirsch argues, though others say the behavior has been seen in other superconductors under high pressure.
Hirsch also contends some of the AC susceptibility data for CSH look suspiciously similar to other data that are now in question, from a 2009 Physical Review Letters paper on superconductivity in europium under high pressure. One of that study’s authors, James Hamlin, a physicist now at the University of Florida, who participated in the europium study as a graduate student, recently determined, he says, that “there are alterations to the data.” The study’s senior author, James Schilling, an emeritus physicist at Washington University in St. Louis, says he shares that concern. Now, some of the co-authors are redoing the measurements, and if they don’t hold up, the team will retract the paper, Hamlin says.
The first author of the europium paper, Mathew Debessai, now with Intel Corporation, was responsible for the AC susceptibility measurements, and also carried out those measurements for Dias’s CSH work. And a data trace from the CSH paper looks “remarkably similar” to one from the europium paper, Hirsch contends. Schilling agrees parts of the traces do look similar, but he can’t say why. Debessai declined to respond to the questions about the data but via email wrote that he will post a formal response on the arXiv preprint server.
In October and November 2020, Hirsch emailed Dias requests for raw data. Dias replied that he would not provide the data for reasons including that his team was working to patent the work, and his lawyer advised against releasing the data. By then, Hirsch had raised concerns about the CSH data in a preprint, which was published in Nature in August. In his email to Hirsch, Dias wrote, “Given that you have an active comment on our work, we consider such a request would not be reasonable.”
Frustrated, Hirsch requested the data from Nature and the National Science Foundation (NSF), which funded the work. On 30 August, Nature appended an editor’s note to Dias’s paper saying: “The editors of Nature have been alerted to undeclared access restrictions relating to the data behind this paper. We are working with the authors to correct the data availability statement.” NSF and the University of Rochester both tell Science they cannot comment on possible investigative matters.
Then, last month in Physica C :Superconductivity and its Applications, Hirsch wrote that Dias’s anomalous AC susceptibility result, combined with seeming irregularities in the europium data, and his struggles to get the data on CSH, fed his doubts. “I suggest that one possible explanation is that [the CSH finding] is the result of data manipulation and alteration such as was described … for [europium],” Hirsch wrote.
Alexander Goncharov, a physicist at the Carnegie Institution for Science, won’t go that far, but he thinks Hirsch’s concerns about AC susceptibility measurements are fair. “I tend to believe there is a problem with the [CSH] paper,” says Goncharov, who tried repeatedly to synthesize CSH using Dias’s recipe and failed.
Given the questions, several scientists say Dias should make his data public. “I am unhappy that Dias is supposedly not cooperating with researchers who are questioning his data,” says Marvin Cohen, a theoretical physicist at UC Berkeley. Schilling is blunt: “I told Dias to give [Hirsch] the raw data, for heaven’s sake.”
Dias and others say they don’t trust Hirsch to appraise the data fairly. “Unfortunately, sometimes he is not objective,” says Vasily Minkov, a chemist at the Max Planck Institute for Chemistry who synthesizes hydride superconductors and says Hirsch has cherry-picked data from the Max Planck experiments for his critiques. Hirsch calls such critiques “completely unfounded.” Hirsch counters that his belief that the BCS theory is incorrect “does not mean I am ‘biased’ or not ‘impartial.’ It means that I am motivated to scrutinize carefully the experimental evidence and judge it on its merits, as opposed to assuming it is likely to be right because BCS theory predicts it to be right, as everybody else does.”
Several new results have only deepened the mystery. In a preprint posted on 30 September on arXiv, Goncharov’s group reports synthesizing CSH under high pressure—using a recipe different from Dias’s—and observing a crystalline structure similar to the one Dias reported. Dias sees the result as vindication, saying Goncharov’s material has the “exact structure” his team reported. But Goncharov is more cautious. His team didn’t test whether its CSH sample was superconducting. But in unpublished work, Minkov says he and his colleagues synthesized the same CSH structure as Dias’s and found it doesn’t superconduct above the Tc of H3S. Minkov says H3S may be responsible for superconductivity in his CSH sample. “We couldn’t see any effect of carbon,” he says.
Hirsch, meanwhile, is mounting a broader attack: on claims for superconductivity in any hydride. In a preprint posted on 4 October on Research Square, Minkov and a team led by Mikhail Eremets, a physicist also at Max Planck, reported remaking diamond anvil cells without magnetic materials and testing large samples of two hydrides, H3S and LaH10. The result: the first ever evidence of the Meissner effect in hydrides, which the team calls “unambiguous evidence” that superconductivity in hydrides is real. Hirsch disagrees, calling the analysis “deeply flawed” in a preprint he and a colleague posted on 14 October on the arXiv server.
Only one thing seems certain to emerge from the controversy over room-temperature superconductivity: more heat.