Electromagnetic penetration depth in superconducting alloys containing magnetic impurities

Measurements of the attenuation of an axial magnetic field by cylindrical, thin, superconducting films were made by using a SQUID-based magnetometer. The sample films were between 40 and 85 A thick. They were composed of indium and of indium containing a small amount of gadolinium. Each film was made by flash-evaporating pellets and quench-condensing the film onto a substrate. The sector of the cylindrical substrate not facing the evaporator was covered by a heavy indium film. Correction terms were calculated numerically to transform the attenuation measurements for the finite-length sample into equivalent values for a hypothetical, infinitely long sample which has the sample film evaporated all the way around the substrate. The film thickness was determined from the temperature dependence of the film resistance between 77 and 140 K. The electron mean free path was determined from the normal-state resistance at low temperatures. The electromagnetic penetration depth lambda, the reduced penetration depth lambda(sigmanTco)1/2, and the London penetration depth lambdaL were determined from the corresponding magnetic field attenuation of the infinitely long sample, the film thickness, and other parameters. The results were compared with calculations based on Shiba's theory b) the Abrlkosov-Gortkov limit (E = 1). The best results were obtained o for the thickest sample film, which was made of pure indium. This sample had lambda = 774 ± 39 X,lambda(sigmanTco)1/2 = -0.1595 ± 3%, and LambdaL = 374 ± 19 X n co at zero temperature. The expressed uncertainty does not include contrihut ions from the uncertainties in published values of the resistivity-mean free-path parameter p£ and the ideal coherence length [o. Our value of lambda(sigmanTco)1/2 was found to be 27% larger than the theoretical value of 0.120. 11 co The results indicate that thicker samples should be investigated.U of I OnlyThesi