Scientific computing contributes to superconductivity

Super-fast levitating trains and lossless electrical wiring are among the realistic future applications of superconductivity – meaning the absence of electrical resistance in certain materials.

The existence of superconductivity has been known for more than a hundred years, but for a long time had mostly academic attention since the phenomenon was only observed at extremely low temperatures close to the absolute zero of minus 273 degrees Celsius. This changed in the 1980s, when several ceramic materials designed for the purpose displayed superconductivity at minus 183 degrees Celsius. The findings triggered a race to develop materials able to superconduct at even “warmer” temperatures, and to understand the phenomenon better.

A new method for describing the electronic structure of superconductors has been developed through a European project with use of High-Performance Computing (HPC).

“We are a long way from solving everything, but we can do things that we could not do before, like predicting the superconducting critical temperature and examining the way in which the superconducting state competes with other states like ferromagnetism,” says Mark van Schilfgaarde, Professor at King’s College London and head of the project.

In this context, the term critical temperature refers to a distinct feature for superconductors. While in ordinary metallic conductors, resistance decreases gradually as temperature is lowered, a superconductor has a point in temperature where the resistance drops suddenly to zero. The critical temperature is specific for each type of material.

Models rely on assumptions

For scientists and industry looking to develop new superconductors, it is of great value to be able to predict the critical temperature and other key properties of a candidate material. Thereby, the labor and economic cost for synthesizing materials which would not be relevant can often be avoided. However, the theory behind superconductivity, to a large extent involving events at the atomic level, in other words quantum physics, is so complex that such predictions have traditionally been very uncertain.

Mark van Schilfgaarde and his colleagues overcame these limitations through a combination of new theory and raw computing power.

Through the Partnership for Advanced Computing in Europe (PRACE), the project was granted 40,000,000 core hours on the Joliot-Curie – Rome supercomputer, hosted by GENCI at CEA, France. Several national research and education networks (NRENs) have contributed with the necessary data transfer.

“A big problem in the field of superconductivity research is that it is so complicated that it almost entirely relies on models,” explains van Schilfgaarde. “This is not to denigrate the work of those who create these models – these people are brilliant physicists who have created intricate and ingenious theories for explaining superconductivity. However, the models all rely on assumptions that we have no basis for knowing whether they are correct or not, which has led to a lot of confusion in the literature and a lack of consensus about the mechanisms behind superconductivity.”

A step towards a systematic theory

The new method has now been used to study several superconductors, using the power of HPC resources provided by PRACE to solve the complex quantum mechanical equations involved. One example involved the investigation of the critical temperature of iron selenide. In bulk, the critical temperature of this material is 8.5 Kelvin (so 8.5 degrees above the absolute zero), but recent experiments showed that if a monolayer of it was deposited on top of strontium-titanate, the critical temperature jumped to around 100 Kelvin. The reason for this jump in critical temperature was originally thought to be assistance from electron-phonon interactions, but van Schilfgaarde’s calculations indicated otherwise:

“Our calculations showed that the real reason for this jump in critical temperature is due to a small change in what is known by modelers as the Hubbard Hamiltonian parameter, which causes a massive change in the incoherence of the electrons and thus the critical temperature. We showed in our calculations that if you tweaked this parameter for the material in bulk, the same increase in temperature was achieved.”

The team from King’s College London have now studied enough systems to know that their theory can describe many kinds of superconductors. The method represents a first step towards developing a systematic theory of superconductivity, which will provide a powerful tool in the further search for the holy grail in superconductivity: materials which display superconductivity at room temperature.

The text is inspired by the article “Unconventional Superconductivity”published on the PRACE website.