For more than 20 years, physicists have been unable to explain exactly how and why high-temperature superconductivity only seems to occur in a special group of mostly copper-based compounds. Now, scientists in Japan have discovered a completely new type of high-temperature superconductor based on iron that could provide physicists with new ways of studying the phenomenon — and shed new light on an important mystery of condensed-matter physics.

Superconductivity is the complete absence of electrical resistance in a material and is observed in many materials when they are cooled to below their superconducting transition temperature (Tc). Superconductivity relies on getting electrons to overcome their mutual electrical repulsion and form “Cooper pairs” that then travel unheeded through the material. In the Bardeen–Cooper–Schrieffer (BCS) theory of low-temperature superconductivity, the electrons are held together as a result of their interactions with lattice vibrations (called phonons) in the material.

However, the BCS theory can not explain the behaviour of high-temperature superconductors, discovered in 1986, which have transition temperatures as high as 138 K. These “cuprates” consist of parallel planes of copper oxide in which the copper atoms lie on a square lattice and where the charge is carried by “holes” sitting on oxygen sites. Each copper atom has an unpaired electron, and hence a magnetic moment or “spin”, and some researchers believe that it is the coupling between these spins that gives rise to superconductivity in these materials.

Superconductor below 26 K

Now, Hideo Hosono of the Tokyo Institute of Technology and colleagues have discovered the first iron-based high-temperature superconductor, which has no electrical resistance at temperatures below 26 K (J. Am. Chem. Soc. 130 3296).

The crystalline material is called LaOFeAs comprises layers of lanthanum and oxygen sandwiched between layers of iron and arsenic — and is doped with fluoride ions. The researchers expect that the Tc of 26 K could be further increased by modifying the material by applying pressure, for example.

Not phonon-mediated

Preliminary studies of the material suggest that its superconductivity is not phonon-mediated, as expected from the classic “BCS theory”, but could be similar to that thought to exist in the high-temperature cuprates.

“One would think superconductivity to be phonon-mediated in this material — like in low-temperature superconductors,” said Kristjan Haule, a theorectical physicist at Rutgers University in the US, whose team is also working on understanding this material. “However, we performed density functional calculations that suggest that the Tc would be around 1 K at most if phonons were indeed responsible.”

Haule's team have calculated that the undoped LaOFeAs compound should be a very bad metal at low temperatures and is almost an insulator (arXiv: 0803.1279). “This is another strong indication that the superconductivity is not mediated by phonons, which requires a good metallic state with coherent carriers,” Haule told physicsworld.com.

Weak-coupling theories

Indeed, this bad metal state resembles lightly-doped high-temperature superconductors, he explained. According to the team, this means that weak-coupling theories — such as spin fluctuations — suggested in the past (and which failed) to describe cuprates might now be useful explaining superconductivity in LaOFeAs compounds. Preliminary experimental results from Hosono's group appear to agree with these findings.

The new superconductor is also proof that superconductivity is not limited to copper oxides and a few other compounds based on uranium, cerium and plutonium. Although superconductivity is destroyed by high magnetic fields, the discovery shows that it can even exist in a strongly magnetic material like iron when the iron is surrounded by other suitable atoms (in this case, arsenic). Moreover, effects related to the orbital properties of electrons, usually neglected in cuprates, can play an important role too.

Haule believes that the new class of superconductors might be technologically important but much more research is still needed before this can be said with any certainty.