No Arabic abstract
The spin valve effect for the superconducting current based on the superconductor/ferromagnet proximity effect has been studied for a CoO_x/Fe1/Cu/Fe2/Cu/Pb multilayer. The magnitude of the effect $Delta T_c$ = T_c^{AP} - T_c^{P}, where T_c^{P} and T_c^{AP} are the superconducting transition temperatures for the parallel (P) and antiparallel (AP) orientation of magnetizations, respectively, has been measured for different thicknesses of the Fe1 layer d_{Fe1}. The obtained dependence of the effect on d_{Fe1} reveals that $Delta T_c$ can be increased in comparison with the case of a half-infinite Fe1 layer considered by the previous theory. A maximum of the spin valve effect occurs at d_{Fe1} ~ d_{Fe2}. At the optimal value of d_{Fe1}, almost full switching from the normal to the superconducting state when changing the mutual orientation of magnetizations of the iron layers Fe1 and Fe2 from P to AP is demonstrated.
The Andreev current and the subgap conductance in a superconductor/ insulator/ ferromagnet (SIF) structure in the presence of a small spin-splitting field show novel interesting features (A. Ozaeta et al., Phys. Rev. B 86, 060509(R), 2012). For example, the Andreev current at zero temperature can be enhanced by a spin-splitting field h, smaller than the superconducting gap, as has been recently reported by the authors. Also at finite temperatures the Andreev current has a peak for values of the spin-splitting field close to the superconducting gap. Finally, the differential subgap conductance at low temperatures shows a peak at the bias voltage eV = h. In this paper we investigate the Andreev current and the subgap conductance in SFF structures with arbitrary direction of magnetization of the F layers. We show that all aforementioned features occur now at the value of the effective field, which is the field acting on the Cooper pairs in the multi-domain ferromagnetic region, averaged over the decay length of the superconducting condensate into a ferromagnet. We also briefly discuss the heat transport and electron cooling in the considered structures.
Competition between superconducting and ferromagnetic ordering at interfaces between ferromagnets (F) and superconductors (S) gives rise to several proximity effects such as odd-triplet superconductivity and spin-polarized supercurrents. A prominent example of an S/F proximity effect is the spin switch effect (SSE) observed in S/F/N/F superconducting spin-valve multilayers, in which the superconducting transition temperature T$_c$ is controlled by the angle $phi$ between the magnetic moments of the F layers separated by a nonmagnetic metallic spacer N. Here we present an experimental study of SSE in Nb/Co/Cu/Co/CoO$_x$ nanowires measured as a function of bias current flowing in the plane of the layers. These measurements reveal an unexpected dependence of T$_c(phi)$ on the bias current: T$_c(pi)$--T$_c(0)$ changes sign with increasing current bias. We attribute the origin of this bias dependence of the SSE to a spin Hall current flowing perpendicular to the plane of the multilayer, which suppresses T$_c$ of the multilayer. The bias dependence of SSE can be important for hybrid F/S devices such as those used in cryogenic memory for superconducting computers as device dimensions are scaled down to the nanometer length scale.
Superconducting spin valves based on the superconductor/ferromagnet (S/F) proximity effect are considered to be a key element in the emerging field of superconducting spintronics. Here, we demonstrate the crucial role of the morphology of the superconducting layer in the operation of a multilayer S/F1/F2 spin valve. We study two types of superconducting spin valve heterostructures, with rough and with smooth superconducting layers, using transmission electron microscopy in combination with transport and magnetic characterization. We find that the quality of the S/F interface is not critical for the S/F proximity effect, as regards the suppression of the critical temperature of the S layer. However, it appears to be of paramount importance in the performance of the S/F1/F2 spin valve. As the morphology of the S layer changes from the form of overlapping islands to a smooth case, the magnitude of the conventional superconducting spin valve effect significantly increases. We attribute this dramatic effect to a homogenization of the Green function of the superconducting condensate over the S/F interface in the S/F1/F2 valve with a smooth surface of the S layer.
We have investigated CuNi/Nb/CuNi trilayers, as have been recently used as the core structure of a spin-valve like device [J. Y. Gu et al., Phys. Rev. Lett. 89, 267001 (2002)] to study the effect of magnetic configurations of the CuNi layers on the critical temperature, Tc, of the superconducting Nb. After reproducing a Tc shift of a few mK, we have gone on to explore the performance limits of the structure. The results showed the Tc shift we found to be quite close to the basic limits of this particular materials system. The ratio between the thickness and the coherence length of the superconductor and the interfacial transparency were the main features limiting the Tc shift.
We investigate the subgap transport properties of a S-F-Ne structure. Here S (Ne) is a superconducting (normal) electrode, and F is either a ferromagnet or a normal wire in the presence of an exchange or a spin- splitting Zeeman field respectively. By solving the quasiclassical equations we first analyze the behavior of the subgap current, known as the Andreev current, as a function of the field strength for different values of the voltage, temperature and length of the junction. We show that there is a critical value of the bias voltage V * above which the Andreev current is enhanced by the spin-splitting field. This unexpected behavior can be explained as the competition between two-particle tunneling processes and decoherence mechanisms originated from the temperature, voltage and exchange field respectively. We also show that at finite temperature the Andreev current has a peak for values of the exchange field close to the superconducting gap. Finally, we compute the differential conductance and show that its measurement can be used as an accurate way of determining the strength of spin-splitting fields smaller than the superconducting gap.