News Release

Stripes or no stripes?

Peer-Reviewed Publication

Max-Planck-Gesellschaft

Systems of many interacting quantum particles can organize themselves in ways that continue to fascinate physicists. Examples are electrons on the surface of certain semiconductors that show the integer and fractional quantum Hall effects, and electrons in bulk copper oxide materials that exhibit high temperature superconductivity. Elucidating the organizing principles of these quantum states remains a fundamental challenge that occupies some of the best minds in theoretical physics. Experimentalists, on the other hand, continue to discover new ground states that add to this challenge. A recent one is the "stripe" phenomenon: In some copper oxides, at least a fraction of the electrons spontaneously form stationary, one-dimensional arrays ("stripes") instead of participating in the superconducting state. (Related observations have also recently been reported in quantum Hall systems.) There is currently an intense debate about whether the striped state is essential for the mechanism of high temperature superconductivity, or whether it represents a competing state and is therefore actually detrimental to superconductivity. In the 19 May issue of Science Magazine, an international team of scientists at the Max Planck Institute for Solid State Research, Princeton University (USA) and the Centre d'Energie Atomique (France) reports strong experimental evidence favoring the latter scenario.

Both the striped phase and high temperature superconductivity are ultimately due to strong Coulomb interactions between the electrons. But while the mechanism of high temperature superconductivity is still mysterious, the origin of the striped phase is better understood: Here, the electrons are at least partially localized, with their relative distance maximized for a given density. Their mutual (repulsive) Coulomb interaction is therefore minimized, and the resulting energy gain provides the "incentive" for the material to enter the striped state. Experimentally, the striped state is best detected by neutron diffraction. Specifically, the magnetic moments of the localized electrons order in an antiparallel fashion on the copper atoms (red arrows in the figure). As neutrons carry a magnetic moment themselves, scattering of neutrons from a material in a striped phase produces a characteristic magnetic diffraction pattern. This is how the static striped phase was discovered in a specific copper oxide material, a derivative of the original La-Sr-Cu-O system that earned Bednorz and Müller the 1987 Nobel Prize in Physics. The striped state is closely similar to the so-called antiferromagnetic state that is ubiquitously observed in insulating cuprates (see figure). However, in contrast to the two-dimensional antiferromagnetic state where all electrons are localized, the magnetic order in the striped phase is weakened along sets of lines which provide room for mobile electrons. The striped state can therefore be metallic and perhaps even superconducting.



The upper panel shows the crystal structure of YBa2Cu3O6+x investigated in the report by Bourges et al. Magnetic copper atoms are highlighted in red, oxygen atoms are green. The middle panel shows the antiferromagnetic structure of insulating YBa2Cu3O6+x. The arrows are the spins of electrons localized on the copper atoms. The lower panel shows the magnetic structure of the proposed striped phase. Bourges et al. present evidence that striped phases, though present in some copper oxide superconductors, are not responsible for high temperature superconductivity.

Full size image available through contact

Figure:

Since the original discovery by Bednorz and Müller, many other high temperature superconductors have been found, with crystal structures that share only one common element: two-dimensional sheets consisting of copper and oxygen (see figure). In order to uncover the mechanism of high temperature superconductivity one has to look for experimental features that are common to all of these. A recent report by a group of U.S. scientists of evidence for stripes in another family of superconductors, Y-Ba-Cu-O, (Mook et al., Nature 395, 580, (1998)) has therefore generated much speculation that high temperature superconductivity is essentially a one-dimensional phenomenon that is nucleated within a stripe and spreads between stripes only when it is already well developed. Given the layered crystal structure, a mechanism based on one-dimensional physics would be rather counterintuitive.

What the U.S. scientists had observed -- again by neutron scattering -- was a magnetic pattern not unlike that seen in the La-Sr-Cu-O system but not as straightforward to interpret: In order to be consistent with an interpretation based on stripes, one had to assume that the stripes are not static but fluctuating. The new and much more detailed work reported in Science shows that this interpretation is untenable. Rather, at least in those materials that have the highest superconducting transition temperatures, the pairing and spin correlations in the copper oxide planes appear to have a fully two-dimensional symmetry and are closely linked to the antiferromagnetic state, as one might have expected based on the crystal structure.

What, if not stripes, is responsible for superconductivity in the copper oxides? There is increasing evidence that two-dimensional spin excitations are essential. These were already reported last year (Fong et al., Nature 398, 588 (1999)) and mapped out in much more detail in the present work. They appear to be capable of organizing the electrons into Cooper pairs, analogous to phonons in ordinary superconductors. While the spin fluctuation pairing scenario is gaining wider currency, a full microscopic theory as powerful as the famous "BCS" theory for ordinary superconductors is not yet within grasp. Nontheless, the neutron scattering work represents a giant step forward towards a phenomelogical understanding of the unusual electronic state that supports high temperature superconductivity.

###

Published: May 19, 2000

Contact: Prof. Dr. Bernhard Keimer
Max Planck Institute for Solid State Research, Stuttgart, Germany
Phone: 49-711-689-1650
Fax: 49-711-689-1632
e-mail: keimer@kmr.mpi-stuttgart.mpg.de



Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.