The case of the curious superconductor
Superconducting strontium ruthenate is a conundrum. It does not behave like a conventional metallic superconductor, nor is it like its close cousins, the famous cuprates that superconduct at relatively high temperatures.
Neutron studies provide the best tool to unravel its intriguing behaviour, which may emerge from specific magnetic interactions.
Strontium ruthenate (Sr2RuO4) is a mixed metal oxide with a layered crystal structure that is very similar to those of the mixed metal cuprates shown to be superconductors at temperatures above that of liquid nitrogen. Superconductors are materials that can conduct electricity without resistance and thus loss of energy. They are thus of great interest for power transmission and other technological applications. Until the cuprates were discovered in the 1980s, superconductivity was seen in just a few metals, and only at temperatures not much above absolute zero.
The standard explanation for superconductivity is that the electrons responsible for electrical conduction couple into pairs, the interaction between them being mediated by the vibrations of the crystal lattice so that they surf through it unhindered. It was soon realised that the cuprates behaved differently. The crystal lattice was not involved; instead, complex magnetic interactions between electrons in the structure’s copper-oxide layers were key (electrons have spin and therefore a magnetic moment). Physicists are still unravelling the details of the pairing mechanism.
As soon as high-temperature superconductivity was discovered, researchers began to look for other compounds that might behave similarly. In the 1990s, strontium ruthenate was shown to be superconductor too – but only below an extremely low temperature of 1.5K. It looked as though the mechanism behind its superconductivity might be different yet again.
To investigate further required exploring the nature of the superconducting pairs. In superconductors identified so far, the spins of the electrons in each pair couple in opposite directions, so have no magnetic moments. The pairs, therefore, do not respond to a magnetic field, and the magnetic response, or susceptibility, goes to zero at zero temperature. Above zero, however, the thermal energy will cause the pairs to start to break up into single electrons (each with a magnetic moment), and so the susceptibility rises. Measuring the susceptibility of a sample as it changes from a conducting to a superconducting state thus offers a way of probing the pairing behaviour. Two experimental methods for measuring susceptibility can be used: nuclear magnetic resonance (NMR) and polarised neutron scattering (PNS).
Results from the first experiment (green) compared with the results at 1T from  (blue), with modelled temperature dependencies and the structure of Sr2RuO4.
Measuring unpaired electrons
NMR works by applying a combination of powerful magnetic fields and radio-frequencies to probe how an atomic nucleus with a spin is affected by its immediate electronic environment in a material; it can also detect any small magnetisation of conducting electrons caused by the magnetic field, and so measure the susceptibility. Neutrons have spin, so can also probe the magnetic behaviour of materials. PNS involves using a neutron beam in which the neutron spins are aligned in a given direction (polarised). To determine the amount of magnetisation and thus susceptibility, the polarised beam is passed through the sample, with a magnetic field applied both parallel and antiparallel to the polarised beam. The effect of any induced magnetisation is revealed after scattering as a slight increase in the proportion of neutrons polarised parallel to the magnetic field when compared to those polarised antiparallel to it.
Early NMR and PNS studies on strontium ruthenate (with the magnetic field applied along the layers) could not detect any drop in susceptibility in the superconducting state. It remained constant down to a temperature of zero. The results therefore suggested that the superconducting pairs retain a magnetic moment and that their spins are therefore parallel, unlike in a similar cuprate. Strontium ruthenate seemed to behave more like helium-3 when in its low-temperature superfluid state (a quantum state similar to superconductivity but showing zero friction) in which the two electrons are in the same spin states. However, more recent NMR measurements have challenged this picture. They showed that the susceptibility dropped in the superconducting phase after all, and that the previous NMR measurements had suffered from sample heating. We recently carried a second PNS experiment at an even lower magnetic field and at extremely low temperatures. We found that the susceptibility goes down to about 1/2, so most likely stays finite. This might be because some of electron pairs retain a magnetic moment – or it might be due to impurities in the sample. There are several possible reasons, and we hope to carry out experiments to explore them further.
Instrument used : Three-axis spectrometer IN20
Article from: A.N. Petsch et al.. Phys. Rev. Lett., 2020, 125, 217004; doi: 10.1103/PhysRevLett.125.217004.
 K. Ishida et al., Nature (London), 1998, 396, 658.
 J. A. Duffy et al., Phys. Rev. Lett., 2000, 85, 5412.
 A. Pustogow et al., Nature (London), 2019, 574, 72.
 A. T. Rømer et al., Phys. Rev. Lett., 2019, 123, 247001.