Bonding of a niobium wire t ~ a niobium I thin film

A method for bonding a niobium wire to a niobium thin film i s described The bonds are to be used as superconducting connections between wire-wound gradiometers and thin-film coupling coils on oc SOUIOS. The method is characterized by two steps. Firstly. the hardness of the niobium wire is reduced by a heat treatment. Secondly, the niobium film is covered with a thin layer of palladium to prevent it from oxidizing. Superconducting bonds were realized using an ultrasonic bonding technique. We tested the bonds and measured superconductivity (to a sensitivity level of 6 x IO-'' n) with currents up to 80 PA, which is equivalent lo 10 times the dynamic range of the oc soulo systems. Even at 80pA, limited by the measuring set-up, the critical current of the bonds is not reached. Present SQUID magnetometers are most often equipped with thin-film DC SQUIDS to which external signals can be coupled via spiral-shaped thin-film input coils [I]. Usually a signal is transferred to such a SQUID by means of superconducting wire. Therefore a proper connection between wire and thin-film input coil has to be realized. This superconducting contact can he made with lead [2] or a lead alloy [3]. In this communication an alternative approach is described. We present a method to bond a niobium wire to a niobium thin film. The annealing of the wire is first considered and after that the bonding and a test experiment. In order to obtain a highly ductile niobium bonding wire, it is first processed in z vacuum annealing step. The niobium bonding wire has a diameter of 50 pm after its insulation layer is removed by sulphuric acid. About 90cm of the wire is hung in a vacuum chamber in a vertical V-shape with an additional weight of 0.2 g in the centre, The chamber is pumped to a pressure lower than Pa (typically 3 x IO-& Pa). Then a DC current of 0.45 to 0.50A is fed through the wire. We turn on the current in about 20 s. The temperature of the wire follows the current according to the brightness. A stable temperature is thus quickly reached and results from a dissipation in the wire of 90 to 100 W for 1 m length. The wire is heated for typically 5 min. If the heating exceeds 6 min the niobium may recrystallize. This effect was clearly observed after 10 min of heating. We tried t On leave from: Institute ol Low Temperature and Structure Research, Polish Academy of Science, POB 937, 50.950 Wrodaw, Poland. 0957-0233/91/111121+02 $03.50