For 40 Years, Theories of High-Temperature Superconductivity Assumed the Crystal Was the Same All the Way Through. Warwick Chemists CT-Scanned One Atom by Atom and Found a Patchwork of Two Structures Separated by Walls 150 Nanometers Wide.

A team of chemists at the University of Warwick has looked inside a bulk crystal of a copper-oxide superconductor in three dimensions for the first time and found that it is not a single, uniform structure at all. The crystal is a patchwork of two subtly different atomic arrangements, and the boundaries between them are roughly 150 nanometers wide, hundreds of times broader than a textbook domain wall. The result, published September 17 in Physical Review Letters, cuts against an assumption that has underpinned most theories of high-temperature superconductivity since the cuprates were discovered in 1986.

Cuprates are the family of materials that first showed electricity could flow with zero resistance at temperatures far above those of conventional superconductors, though still well below room temperature. Why they do it remains one of the biggest open problems in physics. Nearly every attempt at an answer starts from the same picture: the crystal lattice that the electrons move through is the same everywhere in the sample, so whatever is going on in one corner is going on in all of them. Evie Ladbrook, Jon Wright and Mark Senn set out to test that picture directly in a single crystal of La1.675Eu0.2Sr0.125CuO4, a member of the so-called 214 family of lanthanum cuprates.

The tool was scanning three-dimensional X-ray diffraction at the European Synchrotron Radiation Facility in Grenoble, France, which recently completed a 150-million-euro upgrade that makes its beam bright enough to map the atomic arrangement inside a solid, voxel by voxel, in the way a hospital CT scanner maps tissue. The technique reconstructs which crystal structure sits at each point in the sample and in which orientation. Rather than the expected single low-temperature orthorhombic phase, the maps showed orthorhombic regions threaded with broad tetragonal wall-like zones. On cooling further, the pattern inverted: fine orthorhombic stripes appeared embedded inside a tetragonal matrix. The two structures differ only in how the oxygen octahedra around each copper atom tilt, but that tilt controls how electrons hop between copper sites.

"For 40 years, the working assumption has been that these materials are the same all the way through, and nearly all the theory is built on that picture," said Senn, who led the study. "We've shown it doesn't hold. The crystal is fundamentally patchworked and textured, with unusually wide boundary regions that likely work against superconductivity rather than just sitting alongside it." He added that the finding "might explain why some materials perform better than others and means some existing bulk measurements will need to be reinterpreted and future theoretical models built that incorporate this structural complexity."

The boundaries matter because they are so wide that they behave less like a dividing line and more like a third structure in their own right, occupying a substantial fraction of the crystal's volume. Any measurement that averages over the whole sample, which is most of them, has been blending the properties of three environments and reporting them as one. The Warwick team suspects the same hidden texture is common across the broader cuprate family and possibly in related materials being pushed toward superconductivity under extreme pressure.

The method may end up being the more lasting contribution. Scanning 3D X-ray diffraction at this resolution was not possible before the ESRF upgrade, and it can be turned on almost any crystalline material to ask whether its internal structure is what everyone has assumed. The Warwick group's answer for cuprates is that it was not.