King's College London

King’s College London physicists and partners have developed a new approach that explains why a class of superconductors can function at high temperatures.

Focusing on cerium superhydride (CeH9), the breakthrough which is published in npj Computational Materials identifies the missing ingredient behind the material’s superconductivity, and reveals it can function at a temperature twice as high as previously predicted.

The researchers believe that by incorporating this ingredient into a more complete theory, this work lays the foundation for a computational search for room temperature superconductors.

Running at a zero-resistance state, superconductors can power electric technologies with almost 0% of energy loss from heat caused by resistance, making them ideal for addressing the world’s growing energy demands. Whilst they currently run in technologies such as MRI scanners, and despite hundreds of materials with superconductivity properties being discovered, the vast majority require extremely low temperatures to function – at least -196C, making them expensive and impractical for widespread use.

Physicists have been searching for decades for room temperature superconductors. One class of materials that have been experimentally proven to function at the highest temperature on record are hydrogen rich compounds, in particular LaH10, the decahydride of Lanthanum, a rare-earth metal. It realises a superconducting state approaching 250 Kelvin (K) – around -23C. However, LaH10 is highly impractical, as it only functions under extreme pressures - comparable to conditions at the Earth’s core.

While LaH10’s superconductivity is widely understood, the theory fails dramatically in describing other hydride superconductor candidates that function at lower, more practical pressures. The King’s team and researchers from the University of Cambridge, Vienna University of Technology, and the Université catholique de Louvain aimed to find a new, better theory.

Dr Yao Wei, former PhD researcher at King’s, said, “We picked one of the most challenging compounds in the hydride class - cerium superhydride (CeH9). Its superconductivity was proven in a 2019 experiment for lower pressures than in any other superhydride, but state-of-the-art theory failed miserably to describe it.”

While it had been accepted that superconductivity in CeH9 arises from the way the lattice structure of the crystal vibrates and produces phonons – waves carrying heat and sound – which interact with electrons, a key missing piece of the puzzle had been overlooked.

What the scientists discovered was that alongside the phonon-electron interactions, electron-electron interactions – or electron scattering – actually held the key to the better-than-expected superconductivity.

Understanding CeH9 presents a major challenge due to its substantial electronic complexity. It contains more than sixty electrons per formula unit, including very heavy electrons.

Dr Jan Tomczak, Senior Lecturer in Physics, explains, “Contrary to electrons in say copper, Ce-borne electrons hesitate to move through the solid because the repulsion between them is especially large. Imagine electrons in copper acting as a fluid like water, the electrons in Ce are more alike to viscous honey - the scattering or friction between electrons slows them down.”

“Whilst superconductivity mediated by phonons has been understood since 1957, what is special here is that we finally understand how beneficial electron scattering is to superconductivity.”, added Dr Siyu Chen, former PhD student at the University of Cambridge.

In essence, this scattering reduces the electrons’ energy, and the more negatively charged, low-energy electrons, the more the nuclei's’ positive charges are shielded, causing nuclei to repel each other less.

Samuel Poncé, Professor at Université catholique de Louvain, said, “The atomic lattice of the crystal can be compared to an array of masses connected with springs - these springs have now become softer, facilitating vibrations. Low-energy electrons and soft phonons are the key ingredients to the superconductivity, and we just got more of both!”

Indeed, when the team took into account the effect of electronic scattering —a challenging quantum many-body problem—on both electrons and phonons, the 50% discrepancy between experimental findings and previous theoretical modelling, was essentially eliminated. The new theory reproduces the transition temperature within 1%.

The team believe this framework is transferrable to many other systems and could be applied to predict phonon-based superconductors that run at even higher temperatures. It could also be used to look at different crystal structures – different coordinations of the hydrogen network for instance – to improve other aspects of phonon-mediated superconductivity, such as reducing the high pressures under which these superconductors need to operate.

The team also believe their framework could advance the role of machine learning in finding superconducting materials. As Dr Tomczak said, “Our computational tool could help simulate synthetic data on different crystal structures and chemical compositions – to build a data set on which to train neural networks. ML could then be used to find optimal solutions to the temperature and pressure challenges, finetuning structures and combinations, and helping us work out which direction to take.”

Dr Yao Wei said “Whilst experimental observation remains the definitive test of superconductivity, there is too much chemical and structural freedom to synthesise all possible materials and check them for superconductivity in the lab.

“Our work establishes a versatile and predictive computational tool that could speed up the exploration and discovery of promising phonon-mediated superconductors functioning at high temperatures and lower pressures.”