On an autumn day last year, Isaac Silvera of Harvard University received a phone call from his postdoc Ranga Dias. “It’s shining!” an excited Dias blurted out. The “it” was a 10-μm-diameter sample of solid hydrogen inside a high-pressure diamond-anvil cell (DAC), shown in the photograph.
That shininess could mean that the hydrogen had become an atomic metal. Ever since Eugene Wigner and Hillard Huntington predicted in 1935 that hydrogen under high pressure would become metallic, generations of physicists have chased after the elusive phase. Silvera and Dias are convinced that they’ve finally caught up to it, but many in their field express serious reservations.
The DAC setup was in a cryostat at 83 K. Dias had been monitoring the sample through a microscope attached to the cryostat as he increased pressure. At the relatively low pressure of 250 GPa, the sample was transparent, as expected for insulating solid molecular hydrogen. Somewhere above 350 GPa, a pressure that rivals that at Earth’s core, the sample turned black, indicating a phase transition. Silvera and Dias suggest that the hydrogen is semiconducting at that pressure.
Dias kept increasing the pressure, and eventually the sample appeared to have become reflective, an expected signature of metallic hydrogen. In the photograph, the hydrogen sample appears reddish and is surrounded by a lighter ring, which Silvera and Dias identify as the rhenium gasket that encloses the hydrogen between the two anvils of a high-pressure diamond-anvil cell. Based on the frequency of a diamond vibrational mode measured with Raman scattering, Silvera and Dias estimate that they’ve reached 495 GPa.
They found reflectance around 0.9, at various frequencies and at both 83 K and 5 K. A fit of the reflectivity data to the Drude model of a free-electron gas suggests a carrier density of 1 per atom, consistent with atomic hydrogen.
Since last October, when Silvera and Dias submitted the paper detailing their results, they’ve kept their sample at liquid-nitrogen temperature, untouched except for one IR reflectance experiment to firm up their data. Given that previous claims for metallic hydrogen didn’t survive scrutiny, Silvera wanted his sample kept safe until after their paper was accepted for publication.
Their now-published paper has met with great skepticism from some in the community, especially from those specializing in DAC measurements. “This is absolutely not convincing,” says Paul Loubeyre of France’s Alternative Energies and Atomic Energy Commission. “The pressure claim is false. The reflectivity measurement presented here cannot be considered as a signature of metal hydrogen.”
Those sentiments are seconded by another DAC expert, Eugene Gregoryanz at the University of Edinburgh. The authors “present wishful thinking,” he says. “It’s absolutely an outstanding example of the refereeing process gone totally wrong.”
Silvera defends the experimental results. “We would not have submitted a paper if we were not highly confident of our observation,” he says. He adds that critics should try to falsify or confirm his and Dias’s results rather than dismiss them out of hand.
Marcus Knudson and Mike Desjarlais of Sandia National Laboratories are more enthusiastic; in 2015 they used shock compression to elicit a liquid–liquid insulator to metal transition in deuterium (see Physics Today, September 2015, page 12). They would like Silvera to better nail down the actual measured pressure and to determine whether the phase is liquid or solid. Recent theoretical work has pointed to a metallic liquid ground state for hydrogen at the highest pressures. But the Sandia researchers both call the results exciting. “I don’t think I have any real concerns or doubts that they are producing metallic hydrogen,” Knudson says.
Nonetheless, Knudson and Desjarlais view the skepticism as a healthy response. “I see that as a good thing,” says Desjarlais. “My hope is that we’ll see several more experiments from other groups attempting to reproduce the results.”
That’s one point everyone can agree on. “We tell exactly what we did to get the high pressures,” Silvera says. “We hope that others will be able to again produce such pressures and help to understand the properties of metallic hydrogen.”
In the meantime, he and Dias plan to attempt Raman scattering measurements on their hydrogen sample. The Raman signal would be extremely weak because light should penetrate metallic hydrogen by only a few nanometers. But if successful, Raman spectroscopy could definitively identify the sample as solid metallic hydrogen.
Failing a positive Raman diagnosis, Silvera will try x-ray scattering and look for the characteristic interference that would indicate a crystal structure. Hydrogen scatters x rays exceptionally poorly, but the high intensities and microfocusing available at synchrotron x-ray sources may just enable Silvera to accomplish the trick.
Gregoryanz too concurs that other measurements should and will be done. “Sooner or later,” he says, “it will be all cleared up.” (R. P. Dias, I. F. Silvera, Science, in press.)
