For two decades, the search for superconductors that worked at high temperatures was restricted to copper. Now a new family of high-temperature superconductors based on iron has been discovered - and materials scientists are wondering what other combinations of elements they might try. So far the quest is guided by luck as much as by judgement, so researchers are hoping the new materials will provide clues to the recipe for high-temperature superconductivity.

Superconductors can carry electric current with no resistance - a property which could transform electric power generation if it were possible to maintain at room temperature. Since 1986, the field has been dominated by ceramic copper-oxide materials (cuprates), where copper-oxygen planes carry electrons between surrounding layers of various elements. Though an operating temperature of 138K (at ambient pressure) was established in 1995, progress since then has been frustratingly slow. Most other types of superconductors only work at temperatures close to absolute zero.

Materials developed this year by Japanese and Chinese researchers, however, consisting of iron-arsenic compounds sandwiched between rare earth oxides doped to provide extra charge carriers, are superconductors up to a promising 55K, though they have yet to work at the boiling point of liquid nitrogen. 'It's superconductivity in places you've never thought of,' says David Larbalestier, a materials scientist working at Florida State University's national high magnetic field laboratory.

The first of the new compounds, LaOFeAs doped with fluorine, was reported in February 2008 by Hideo Hosono and colleagues at the Tokyo Institute of Technology [1]. It was a superconductor at 26K, and higher if put under pressure. Subsequently, Xianhui Chen's team at the University of Science and Technology of China, in Hefei, replaced the rare earth lanthanum with samarium (SmO_1-x  F_xFeAs) and got the critical temperature up to 43K [2]. Zhong-Xian Zhao's group at Beijing's National Laboratory for Superconductivity, China, have since shown that similar compounds with other rare earths are superconductors above 50K; the samarium material under pressure holds the current 55K record [3].

Schematic crystal structure of LaOFeAs © Takahash et al, Nature

The crystal structures of materials made so far are very similar, but a good deal of tinkering with the [rare earth][oxygen][dopant][transition metal][group 5 element] formula is expected. 'You can try replacing just about every element in the formula,' says Pengcheng Dai, a physicist at the University of Tennessee, Knoxville. The rare earths have varied through lanthanum, samarium, praseodymium, neodymium, cerium and gadolinium. The arsenic has been replaced with phosphorus, the iron with nickel, and even the dopant has been changed from fluorine, providing extra electrons, to strontium or oxygen vacancies, providing holes - though these experiments have not reached such high critical superconducting temperatures (T_c).

Whether the iron-based superconductors work in the same way as traditional cuprates is unclear. In both systems, it's suspected that electrons pair up to travel through FeAs or CuO₄ layers unimpeded, though not via the phonons (lattice vibrations) implicated in the conventional theory of low temperature superconductivity. Dai's team published research on 28 May showing that the undoped LaOFeAs is antiferromagnetic - rows of iron ions are magnetised in opposite directions - but this property disappears as the material is doped and superconductivity kicks in [4]. That is rather like the behaviour of the cuprates; on the other hand, the iron compounds also conduct electricity at room temperature - unlike the cuprates which are insulators.

'Theorists tend to think they can predict things, but the best superconductors will really only be found by a systematic study of this class of materials,' says Dai. Next steps for experimenters are to grow single crystals or thin films of the iron-based materials, which are harder to make than the cuprates. Chen says that incorporating multilayers of iron arsenide into a structure could also boost the critical temperature, as has worked with traditional superconductors.

There's much excitement too about moving into different areas of the periodic table to find more high-temperature superconductors - but no way to direct a search. 'The observation of a high T_c in LaOFeAs by electron doping was by chance,' says Hosono. His team were looking for transparent semiconductors and picked out the compounds which had been made - undoped - in the 1990s by Wolfgang Jeitschko and colleagues at the University of Munster, Germany. Could many more high-temperature superconducting systems join the cuprates and the pnictides? 'The game has just begun,' says Dai.

Richard Van Noorden

Interesting? Spread the word using the 'tools' menu on the left.