In 2024, building on LaH10's properties, other scientists predicted a better candidate in the same family of clathrate hydrides: LaSc2H24. Like LaH10, it is a hydrogen-rich material whose crystal structure consists of metallic atoms sitting inside cages of hydrogen ions. Calculations gave it a critical temperature of 43°C at a pressure of 167 gigapascals—slightly lower pressure, dramatically higher temperature. What the calculations could not explain was why two scandium atoms made such a difference.

A team from Jilin University in Changchun and Zhejiang University in Hangzhou, including some of the same researchers behind the 2024 prediction, has now answered that question. Yanming Ma, a co-author of both studies, and his colleagues traced the limitation of plain LaH10 to an anisotropy—an unevenness—in its electronic band structure. The band structure describes the ranges of energy electrons can and cannot occupy in a solid. In these materials, hydrogen atoms sit at three inequivalent sites, forming structurally different anisotropic layers. That unevenness opens two distinct superconductivity channels, each with its own critical temperature, and the weaker of the two holds the material back.

The crystal structure of LaSc2H24 differs from its parent in detail: lanthanum sits fully enclosed in a cage of 30 hydrogen atoms, while scandium atoms sit partially enclosed in cages of 24 hydrogen ions.

The decisive player turns out to be scandium's 3d electrons. The team found that these electrons fill two roles at once. Their 3d orbitals strongly overlap with the surrounding hydrogen cages, and they reshape the Fermi surface—the map of quantum states occupied by the material's electrons—so that certain scandium-hydrogen-scandium bonds are favored over others. The result, as the authors put it, is an "Sc-induced gap unification": the two superconductivity channels merge into one, eliminating the weak link that caps LaH10's critical temperature.

To test whether the 3d electrons truly matter, the researchers tried substituting calcium or magnesium for scandium. Calcium has an empty 3d orbital; magnesium has none at all. Both resulting materials were chemically stable, but neither was superconducting. The comparison indicates that scandium's 3d electrons are essential to shaping the Fermi surfaces and, through them, to superconductivity itself.

The findings remain theoretical in the sense that LaSc2H24's superconductivity was predicted computationally rather than measured in a lab, and the required pressure of 167 gigapascals still rules out practical applications today. Still, the group is optimistic. "This suggests 3d transition metal doping as a promising strategy for engineering isotropic room-temperature superconductivity in high-pressure hydrides," they write—turning a one-off prediction into a design principle for future materials.