The search for materials that can conduct electricity at room temperature without losing energy is one of the greatest and most consequential challenges of modern physics: loss-free power transmission, more efficient motors and generators, more powerful quantum computers, cheaper MRI devices. Hardly any other material discovery has the potential to change so many areas of technology and everyday life at the same time. An international research team with the participation of Christoph Heil from the Institute of Theoretical and Computational Physics at Graz University of Technology (TU Graz) is now presenting a systematic approach to finding such materials. In a perspective article in the journal Proceedings of the National Academy of Sciences (PNAS) – a strategy paper that assesses the current state of research and sets out future directions – the 16 authors state that there are no fundamental physical laws that rule out superconductivity at ambient temperature.
New research results provide optimism
The researchers emphasise that, under the right conditions, superconductivity is an almost universal property of non-magnetic metals and not a rare phenomenon. As impressive proof of the recent progress in this field, the strategy paper refers to an accompanying study that was carried out as part of the same research programme and appears in the same issue of PNAS; namely that researchers at the University of Houston set a new record with a process known as pressure quenching. The mercury-based compound Hg-1223, the record holder for superconductivity at normal pressure since 1993, was first cooled to near absolute zero and simultaneously exposed to a pressure of up to 300,000 times the normal atmospheric pressure. As a result, the critical temperature at which the material becomes superconducting rose from 133 Kelvin to up to 151 Kelvin. After the pressure was quickly released, the increased critical temperature was maintained and thus represented the highest transition temperature ever measured at ambient pressure. This effect persisted for a fortnight after the experiment and was reproduced in five different samples.
For the international team of authors, this result exemplifies a new dynamic in superconductivity research. In order to translate these advances into new materials in a targeted manner, the team has identified two central tasks: a prediction challenge and an engineering challenge. The first task is to significantly improve computer-aided models. In the future, they will not only predict whether a material can become superconducting, but also whether it can actually be produced – a gap that previous models have barely addressed. The aim is to systematically search through large combinations of chemical elements in order to identify promising candidates for industrially producible superconductors.
From serendipity to strategic search
The second challenge lies in the purposive manipulation of materials. Physical influences such as extreme pressure, targeted doping, nanostructures or ultrashort light pulses could be used to artificially generate or significantly amplify superconducting states. The researchers therefore suggest that potential superconductors should be regarded as so-called quantum metamaterials. These are specifically designed material systems in which superconducting properties are not determined by the chemical composition alone, but by the interaction of precisely designed structures on the nanoscale.
A central component of this strategy is the close integration of theory and experiment. In the future, computer models will determine the direction of new experiments, while experimental results will flow directly back into the improvement of theoretical models. This should make the search for new superconductors much more efficient than the previous trial-and-error principle.
Theory, experiment and AI: Pulling together in a coordinated way
"In recent years, we have made enormous progress in the computer-aided simulation of realistic materials," says Christoph Heil. "Today, we can carry out ab-initio calculations on superconductivity in the nanometre range – in other words, on length scales that are actually accessible in experiments. Just a few years ago, we were limited to much smaller unit cells in the angstrom range – that's a difference of around a power of ten." According to Heil, this opens up completely new possibilities: "If we combine these precise calculations with machine learning and artificial intelligence, we can now search the huge space of possible material combinations much more efficiently and accurately than ever before. This is precisely the core of our approach: to link theory, simulation and experiment more closely in order to systematically pursue the path to practically usable superconductors."
The strategy paper therefore ends not with a summary, but with an appeal to the entire research community in physics, chemistry and materials science. The aim is to join forces worldwide and, based on modern AI and simulation methods, to systematically push the limits of superconductivity towards room temperature. In addition to Christoph Heil, researchers from Harvard, Cambridge, MIT, the University of Houston, Columbia University, the University of California, the University at Buffalo, the Carnegie Institution for Science, Travertine Labs and the Enterprise Science Fund of Intellectual Ventures were involved in the publication.
