Upper critical fields in normal metal-superconductor-normal metal trilayers

Abstract

The role of spin orbit interaction in superconducting proximity effect is an area of intense research effort. Recent theoretical and experimental works investigate the possible role of spin-orbit interaction in generating spin-triplet pair correlations. In this work, we present an experimental survey of thin normal metal-superconductor-normal metal trilayers with Nb superconductor and Al, Ti, Cu, Pt, Ta, and Au normal metals, along with single layers of Nb as reference. We aim to probe the role of spin-orbit interaction and resistivity on the normal metal proximity effect through measurements of the upper critical field. We find that the upper critical fields of the trilayers are lower than that of a single layer Nb reference sample, and that the trilayers with higher resistivity metals, Ti, Pt, and Ta, behave as 2-dimensional superconductors. At low applied in-plane magnetic fields and temperatures close to the zero field transition temperature, we find a possible deviation from 2-dimensional to 3-dimensional behavior in the Ti and Pt trilayers. We also find that compared to single layer Nb films, all of our trilayers show a greater suppression of critical temperature during rotation from an in-plane to an out-of-plane applied magnetic field, with the greatest suppression observed in trilayers with Au or Al. This suppression of the critical temperature under field rotation might appear analogous to the colossal spin valve effect that can be achieved in systems with ferromagnetic materials; however, in our trilayers, only conventional orbital screening contributions to the suppression are present and the additional suppression is not present in the absence of applied magnetic field.

Fig. 1

Electrical measurements of a 25 nm single layer niobium film for (a) normalized resistance versus temperature at 0, 1, and 2 T out-of-plane magnetic field; (b) normalized resistance versus in-plane magnetic field; and (c) normalized resistance versus out-of-plane magnetic field. Resistance versus magnetic field measurements presented in (b) and (c) were taken at constant temperatures from 2 K in 0.5 K increments. Lines connecting the data are guides.

Fig. 2

Temperature dependance of the upper critical magnetic field, for (a) 25 and (b) 55 nm thick, single layer Nb films and for (c) Al–Nb–Al, (d) Ti–Nb–Ti, (e) Cu–Nb–Cu, (f) Ta–Nb–Ta, (g) Pt–Nb–Pt, and (h) Au–Nb–Au trilayer films. The lines represent fits to the Ginzburg–Landau theory, Eq. 1. The best fit parameters are given in Table 1.

Fig. 3

Film dimensionality and effective thickness of the superconductor. (a) Extracted film dimensionality parameter, , for the fits presented in Fig. 2. In the Ginzburg-Landau theory, indicates a 2-dimensional superconductor and a 3-dimensional superconductor. (b) Schematic representation of the effective thickness of the superconductor, , based on the comparison of fitted and extracted between the single layer and trilayer samples.

Fig. 4

Crossover from 3-dimensional towards 2-dimensional superconductivity at low applied in-plane magnetic fields in the (a) Ti–Nb–Ti and (b) Pt–Nb–Pt trilayers. The critical temperature is determined at each applied field value by measuring resistance while sweeping temperature at a fixed in-plane magnetic field. The dashed line represents Ginzburg-Landau theory, Equation 4, fit to the lowest field data only ( T) with fixed to 1, the 3-dimensional limit.

Fig. 5

Ostensible analogy with the colossal spin-valve effect. (a) The extracted change in critical temperature, , between in-plane and out-of-plane applied magnetic fields. The trilayers show a greater suppression of compared to the single layer Nb films. (b) The difference in between the trilayers and 25 nm single layer Nb film, , for the Au and Al trilayers.