Room temperature superconductors¿materials that conduct electricity perfectly¿ continue to elude scientists. For now, the closest things available are high-critical temperature (Tc) superconductors. Unlike superconducting metal alloys, which must remain within a few degrees of absolute zero in order to display their resistance-free electron flow, high-Tc superconductors can operate at temperatures around 77 kelvins. But more than just the operating temperature distinguishes these two types of superconductors. According to a report published today in the journal Nature, the electrons in high-Tc superconductors actually behave differently from those in the conventional variety.

S¿amus Davis of the University of California, Berkeley, together with Shin-ichi Uchida of the University of Tokyo and colleagues used a scanning tunneling microscope to probe the behavior of Bi-2212, a high-Tc superconductor that belongs to a class of compounds known as BSCCOs. (BSCCOs are named for their constituent atoms: bismuth, strontium, calcium, copper and oxygen.) Specifically, the team studied single crystals of so-called underdoped Bi-2212, which has fewer electrons¿or more holes¿in its framework lattice. Theorists had predicted that such a material might reorganize the extra holes so as to achieve minimal energy, leaving some regions doped, or superconducting, and others undoped, or non-superconducting. The new observations bore that out. Indeed, the researchers found superconducting islands measuring a mere three nanometers across floating in an electronically distinct background sea (see image). In a conventional superconductor, in contrast, the cloud of conducting electrons is remarkably uniform. According to Davis, the new observations suggest that the electronic structure of even a single crystal high-Tc superconductor made by the best technology available will still exhibit such "granular superconductivity."

The authors note that their findings provide "a new and unconventional context in which to view the underdoped copper oxides." They cannot, however, distinguish among various microscopic mechanisms that may govern the material's behavior. But as Jan Zaanen of Leiden University, The Netherlands, notes in an accompanying commentary, "studies like these are revealing amazing diversity in the behavior of these mystery electron systems at the nanoscale."