Scientists Just Cracked the Code for Finding New Superconductors

Superconductors, materials that carry electricity with no resistance at certain temperatures, have always been a game of chance. Out of countless possible combinations of elements, only a few yield the elusive quantum effect, and most require chilling to near absolute zero.

But now, an international team led by Aalto University’s Professor Päivi Törmä has shown how machine learning can cut through the noise, filtering billions of possibilities to pinpoint promising candidates.

Superconductors are the backbone of technologies ranging from quantum computers to maglev trains. Yet finding them has been notoriously slow.

“Over the decades, researchers have recognized over 7,000 superconductors, but mostly serendipitously,” Törmä explained. “The process of identifying possible materials is so computationally heavy that, in fact, researchers have only been able to predict the viability of about 20 of these theoretically.”

In 2023, the first global coordinated effort at accelerating discovery was established through the formation of the SuperC consortium. In a recent proof-of-concept study published in the journal Physical Review Research, the authors show how machine learning can be used to pre-screen materials before performing costly quantum calculations.

It led them to two new superconductors: YRu₃B₂ and LuRu₃B₂, in which the electrons weave flat bands into a cage-like setting called a kagome lattice, whose geometric design is based on traditional Japanese basket weaving.

Using the algorithm to flag candidates, collaborators at Rice University synthesized the compounds under the guidance of Professor Emilia Morosan. Rigorous testing confirmed their superconductivity, marking a milestone in the consortium’s mission to find a room-temperature superconductor by 2033.

Room-temperature superconductors would transform energy use worldwide.

“Superconductive materials that can operate at room temperature would forever change the way we consume energy,” said Törmä. “If such a material could replace regular conductors in applications like computers and data centers, global energy consumption could be slashed and the heat footprint of the ICT sector vastly reduced.”

The discovery is just the beginning. By combining machine learning with quantum geometry, SuperC hopes to push screening capacity into the billions of materials.”

“Our method uses machine-learning-based pre-screening followed by targeted calculations on the promising candidates. This approach will greatly speed up superconductor discovery in the future,” Törmä said.

The race for a room-temperature superconductor is no longer just about finding the right combination of elements, but algorithms, geometry, and international collaboration. And if the consortium succeeds, it could weave a different future for global energy, one composed of kagome lattices.