All high-temperature superconductors consist of parallel planes of copper oxide, with other elements sandwiched in between these layers. The copper atoms lie on a square lattice and the charge is carried by "holes" sitting on oxygen sites. Previous X-ray scattering measurements on yttrium barium copper oxide (YBCO) superconductors revealed spectra containing diffuse features that were attributed to the formation of stripes in the copper oxide planes. Many physicists believe that these stripes serve as channels along which the super-current can flow.

Jörg Stempfer from the Max Planck Institute for Solid-State Research in Stuttgart and co-workers have now found that these features have their origins in oxygen defects instead (Phys. Rev. Lett. 93 157007). The German team performed its measurements on single crystals of YBCO doped with calcium at the HASYLAB synchrotron radiation source at the DESY Research Centre in Germany and the Advanced Photon Source (APS) at the Argonne National Laboratory in the US. With energies of around 100 keV, the X-rays were able to penetrate thick samples and probe the bulk properties of the samples.

Stempfer and colleagues observed an ordered superstructure with a periodicity of four unit cells in materials that contained oxygen defects, or vacancies, but not in samples that did not contain oxygen defects. Moreover, the formation of the superstructure depended on the oxygen concentration, but not on the charge carrier concentration, which suggests that stripes are not responsible for the diffuse features observed in the X-ray spectra.

Meanwhile in similar experiments Zahirul Islam, who is based at the APS at Argonne, and colleagues have observed "nanodomains" of displaced copper, barium and oxygen atoms in YBCO crystals. According to the US team, the presence of these domains indicates that an ordered pattern of oxygen vacancies forms, leading to the same superstructure seen by Stempfer and colleagues (Phys. Rev. Lett. 93 157008). Such oxygen ordered superstructures were first predicted by Didier de Fontaine and co-workers at the University of California at Berkeley in 1990.

The defects lead to strains in the crystal, which make it inherently inhomogeneous, whereas many model of high-temperature superconductors assume that they are homogeneous. The results from both groups are likely to upset supporters of "stripe theory" but could help explain the origin of several, as yet unexplained, spectral features observed in experiments performed by other teams.