Magnetism and spin - fluctuation induced superconductivity in ruthenates . Talk presented at the SNS ’ 97 Conference , Cape Cod , 1997

Layered and pseudocubic Ru-based perovskites have been the subject of considerable recent attention, due to their unusual magnetic properties and the discovery of superconductivity in one member of the family, Sr2RuO4. From a magnetic point of view, interest derives from the stable ferromagnetism in SrRuO3, gradually disappearing to a non-magnetic phase upon substituting Sr with isovalent Ca, a very unusual kind of behavior for 3d perovskites. On the superconducting side, interest was stimulated by theoretical conjectures and experimental indications that Sr2RuO4 might be a p-wave superconductor. We report first-principles LSDA calculations for ferromagnetic SrRuO3, antiferromagnetic Sr2YRuO6, non-magnetic CaRuO3, and superconducting Sr2RuO4. In all cases, magnetic properties are well reproduced by the calculations. Anomalous properties are explained in terms of simple TB models and Stoner theory. An important result is that O bears sizable magnetic moments and plays an important role in the formation of the magnetic states. Based on these calculations, we have built a model for the q-dependent Stoner interaction, which we consequently applied to Sr2RuO4 to estimate superconducting and mass-renormalization electron-paramagnon coupling constants. We found that spin-fluctuation induced p-wave superconductivity is possible in Sr2RuO4. The estimated critical temperature, specific heat and susceptibility renormalizations are all in good agreement with experiment. The recent discovery of superconductivity in the layered ruthenate, Sr2RuO4 [1] has generated new interest in Ru-based perovskites. At first glance this material seems analogous to the high-Tc cuprates. For instance, it has a similar crystal structure (it is isostructural with La2CuO4) and is apparently close to a magnetic instability (SrxCa1−xRuO3 and Sr2RuYO6 are ferroand antiferromagnetic, respectively). On the other hand, the more we learn about ruthenates, the less similar to cuprates they seem. While initial interest was largely related to the similarity to the high-Tc materials, now it is more that ruthenPreprint submitted to Elsevier Preprint 1 February 2008 ates are deemed interesting per se, and, at least in their magnetic properties they are more variegated and probably more interesting than cuprates. Furthermore, it appears that superconductivity in Sr2RuO4 can hardly be understood without a good understanding of magnetism in ruthenium perovskites in general. Thus this paper naturally breaks into two parts. First, we discuss magnetism in ruthenates, specifically antiferromagnetic Sr2RuYO6, ferromagnetic SrRuO3, and paramagnetic CaRuO3. We will show that despite the wide range of magnetic properties, they all are governed by a simple Stoner-type mechanism, which manifests itself differently depending on crystal structure. We then shall show how closeness to a ferromagnetic instability can produce a triplet superconductivity in Sr2RuO4 and explain its normal-state transport properties. We shall also discuss what is currently maybe the most intriguing question in the theory of superconductivity in Sr2RuO4, namely why the experiment shows finite electronic density of states at zero energy (in NMR and specific heat experiments) at as low as 0.3Tc.