The superconducting proximity effect in strong ferromagnets.

摘要

Superconductivity and ferromagnetism are similar in that they are both phases of matter with ordered electronic spins. In their simplest form, superconducting Cooper pairs are composed of a spin-up and spin-down electron in a spin-singlet with total spin zero. In a ferromagnet, the exchange energy between electrons tends to align the electronic spins into a single spin state. Thus, the two phases are generally incompatible with each other and few materials exist which display both properties. The most straightforward way to study the competition of the two order parameters, therefore, is to put a superconducting material (S) and a ferromagnetic material (F) in electrical contact and measure the properties of the electrons that diffuse between the materials: this is known as the proximity effect. As the energy scale of superconductivity---set by the critical temperature, Tc---is generally much lower than that of ferromagnetism---set by the Curie temperature, TCurie---it is the superconducting order parameter which changes most dramatically.; When a Cooper pair enters a ferromagnet, the spin-up and spin-down electrons are affected oppositely by the presence of the exchange field. In the most simple case, this should lead to the Cooper pair acquiring a non-zero center-of-mass momentum which will result in a spatially varying superconducting density. In response, the superconducting density---and therefore related superconducting phenomena---will oscillate as a function of ferromagnet thickness, dF. This oscillation has been observed in the form of oscillating critical temperatures, Tc, of S/F bilayers and oscillating critical currents, Jc, of S/F/S Josephson junctions, with varying degrees of qualitative and quantitative theoretical agreement. Recent theoretical predictions have greatly increased both the breadth of expected phenomena and the number of critical materials parameters. As both Tc and Jc measurements are particularly sensitive to boundary conditions and provide relatively little quantitative information per sample, more discerning measurement techniques are needed for definitive tests and to narrow the possible range of parameter-space and advance the field.; We measure the tunneling density of states (DOS) of superconductor/strong ferromagnet thin-film bilayers as a function of F-layer thickness, dF, with planar Al2O3 tunnel junctions. The DOS gives us much more direct information about the superconducting order parameter than either Tc or Jc and is not nearly as sensitive to differences in boundary conditions. By measuring the DOS systematically as a function of dF we can observe the spatial evolution of the order parameter and therefore distinguish between the many different possible theoretical predictions in the field. Here, we report measurements made on Nb/CoFe, Nb/Ni, and Nb/CuNi bilayers. By fitting the DOS as a function of rip to the most recent theoretical models, we have determined that the most important parameters needed to explain our data are the exchange field, Eex, and spin-orbit scattering. GammaSO. Further, we have measured an anomalous sub-gap structure that is not predicted by the current theoretical model. Based on its peculiar behavior in a magnetic field, we propose that this feature might be the result of long-range triplet superconducting correlations, which have not been previously observed in these materials.