The Only Game in Town

Using RIXS to Understand the World’s Thinnest Superconductor

July 13, 2021

Can we engineer superconductivity? Steven Johnston, associate professor of physics, builds models to explain what’s going on in materials at the most fundamental level, including recent experiments that uncovered the secrets of the world’s thinnest superconductor.

Johnston is part of an international collaboration studying single sheets of iron selenium only one atom thick, which have superconducting properties that don’t exist in bulk versions of the same material. An ultrathin sample carrying electric current with virtually no resistance could be a boon for fields like computing and electrical transportation. For Johnston, the findings have the added bonus of seeing how his work can make the most of new scientific tools.

While the neutron has long been an excellent tool to probe a material, it won’t work in the one-atomic-layer scenario.

"What’s unique is we’re probing a monolayer of iron selenium," Johnston explained. "If you try to do that with a neutron you can’t, because a neutron doesn’t interact very strongly with matter. And so when you scatter neutrons off the film and substrate (base layer), the chances that it interacts with this single atomic layer of the film are basically zero."

That’s not the case with resonant inelastic x-ray scattering, or RIXS, which the team used to study a single layer of iron selenium (FeSe) on a substrate of strontium titanate (SrTiO₃).

"With resonant x-ray scattering, you essentially tune the x-ray’s energy to an absorption resonance that’s unique to the iron," Johnston said. "So you’re really saying ‘I want to make excitations on iron, and only iron.’ If you’re doing these sort of thin film measurements, RIXS is the only game in town."

These systems aren’t new to Johnston, whose group has been researching them since he came to UT in 2014. His role in the project is a testament to the power of networking and shared expertise: he joined at the invitation of Riccardo Comin from the Massachusetts Institute of Technology, whom he met at the University of British Columbia when Johnston was a postdoctoral fellow and Comin was a graduate student.

"He actually approached me because I have expertise in RIXS and in this monolayer system and he wanted help understanding the experimental data," Johnston said.

Johnston in turn recruited colleague Thomas Maier of Oak Ridge National Laboratory to work on theoretical calculations, along their postdoctoral associate Seher Karakuzu.

But what, exactly, were they trying to figure out?

In superconductors, electrons overcome their natural repulsion to one another and form pairs. Exactly why they do this in materials like FeSe isn’t completely clear, but it happens at a point called the transition temperature (T_c). An added challenge is getting electrons to pair up at room temperature (293 kelvin) or even higher, which broadens a material’s application potential. Part of working out these questions has to do with understanding an electron’s properties, including spin.

At specified energy levels electrons have a particular spin, the same as they have a particular mass and charge. As Johnston explained, there’s a distribution of spin excitations as a function of energy, and there are optimal places where you want to put that weight according to models of unconventional superconductivity.

"If you have more spin excitations at energy x, you tend to get a higher T_c than if you put it at energy y," he said. "Is that’s what happening in this monolayer system? Is this a case of putting the spins where you need them to be in order to get a higher T_c, or is there something else going on? The key question is trying to determine what are the spin excitations in iron selenium doing."

As many scientists believe they drive electron pairing in superconductors, Johnston said "understanding the detailed spin excitation spectrum is really important."

In this work, published in Nature Communications, scientists were measuring low-energy spin excitations and trying to figure out how they were reshaped somewhere between the bulk material and the single-layer sample on a substrate, changing from a poor superconductor to an extremely efficient one. Johnston, Maier, and Karakuzu built models to try to understand what happened.

"The hope is that if you understand why T_c is becoming enhanced in this monolayer maybe you can start playing games to make it even higher by choosing different substrates or different geometries to boost the Tc levels," Johnston said. "The ultimate goal is to determine if we can engineer superconductivity. That’s why this is really interesting: you’re taking a really poor superconductor and putting it on top of something else. All of a sudden you get 10 times’ enhancement of T_c. If you can do that for 8 kelvin iron selenium, maybe you can do this for a 90 kelvin copper-oxide superconductor and get up to 900 kelvin."

To do so would help answer what Johnston calls an age old question in superconductivity: getting the highest transition temperature imaginable, which could usher in a host of new applications.

More information on this research is available from the MIT Materials Research Laboratory, which led the project with Brookhaven National Laboratory.