This week in the arxiv: superconductivity update

Summer writing and travel are eating my blogging time a bit, and I've also agreed to write the occasional nano-related blurb for the ACS. While my posting rate has taken a hit, science has continued to march forward, with a lot of exciting new preprints concerning (relatively) high temperature superconductivity. Here's a sampling....

arxiv:0805.4463 - Matsumoto et al., Superconductivity in undoped T' cuprates with T_c over 30 K
This paper is a perfect example of why materials growers are (unfortunately often unsung) heroes in this field. The authors have come up with a new method for growing cuprate compounds of the form T'Re₂CuO₄, where T'Re is a rare earth from the series (Pr, Nd, Sm, Eu, Gd). Historically these compounds were found to be antiferromagnetic insulators - no superconductivity. In this new work the authors argue that these old results were due to interstitial oxygen leading to pair-breaking. Instead, with the new growth + annealing technique, these compounds are found to exhibit superconductivity with transition temperatures as high as 30 K. These subtleties are why one should always be very careful when looking at suggested compositions in new compounds....

arxiv:0805.4630 - Rotter et al., Superconductivity at 38 K in the iron arsenide (Ba_1-xK_x)Fe₂As₂
This is the first paper I've seen (though I may have missed one) that reports superconductivity in a compound related to the new iron arsenide systems but with two iron arsenide layers per unit cell rather than one. Back in the heyday of the cuprates, the same sort of thing happened - people went from compounds with single copper oxide planes to those with multiple planes per unit cell, and transition temperatures went up. Once again we see how rich the materials landscape can be. Update: as anon. in the comments pointed out, this isn't actually the 2-layer version of the compound. Rather, it's analogous to the so-called "infinite layer" version. My mistake.

arxiv:0806.0063 - Wang et al., Very high critical field and superior J_c-field performance in NdO_0.82F_0.18FeAs with T_c of 51 K
Other exciting features of the new iron arsenide superconductors are their extremely high critical fields and critical currents. If the transition temperatures could be raised a bit (say past 77 K) and the compounds could be made in wire form (certainly not easy in the cuprates; unlikely to be simple in these either since like the cuprates they are brittle), this could be a huge deal for high field magnets and other applications of superconductivity.

arxiv:0805.4616 - Chen et al., The BCS-like gap in superconductor SmFeAsO_0.85F_0.15
arxiv:0806.0249 - Matano et al., Spin-singlet superconductivity with multiple gaps in PrO_0.89F_0.11FeAs
These two papers examine two related compounds with different techniques, trying to figure out how the charge carriers in these iron arsenides pair up to form the Cooper pairs that make up the superconducting condensate state. In the former, measurements of Andreev reflection (a process where an electron in a normal metal approaches a superconductor, two electrons actually cross into the superconductor, and a hole is "retroreflected" back into the normal metal, leading to a pronounced feature in the conductance of the metal/superconductor interface) strongly suggest that the samarium compound acts like an ordinary BCS superconductor. That is, each Cooper pair has zero angular momentum (s-wave pairing); this implies that the superconducting gap is uniform in momentum space, with no nodes. In contrast, the cuprates exhibit d-wave pairing, with a superconducting gap that has a four-lobe structure in momentum space and that goes to zero along four particular crystallographic directions.

The second paper uses NMR measurements of the Pr compound to argue instead that there are multiple gaps, and further that the pairing symmetry is p-wave (which has been seen in superfluid ³He and in strontium ruthenate). At first glance, these two results seem to disagree, though (a) they are talking about different materials, and (b) the Andreev measurements are particularly sensitive to the surface, while the NMR measurements are nontrivial to interpret, at least for nonexperts. Well, this is the fun part - stay tuned, and we'll see how this shakes out.