Abstract
Superconductor-insulator-superconductor (SIS) structures have been studied from both theoretical and experimental aspects. The theoretical study identifies factors that may cause degradation of device behaviour and links them with techniques of junction fabrication. It was shown that oxygen contamination of superconducting films of niobium is the most likely cause of device degradation. As a result clean UHV system has been designed and built in which niobium deposition by electron beam (e-beam) evaporation can be performed in a region completely enclosed in a liquid air cooled shroud. This work has shown that niobium films can be deposited by e-beam evaporation with a transition temperature close to the bulk value of 9.3K. It was also found that SIS devices could be prepared but only if the insulator was prepared by oxidation of an aluminum layer which was also deposited at UHV. It was found that the best results for SIS behaviour could be obtained by oxidation of an aluminum film with a thickness of 55 Angstroms. It was also found that SIS devices could not be obtained by simply oxidising the first layer of niobium. This latter finding was thought to be due to the dissolution of oxygen from the oxide into the second niobium layer which not only degraded the stoichiometry of the insulator but also adversely affects the superconducting properties of the second electrode. An empirical model has been developed to explain the behaviour of the observed I-V characteristics of SIS junctions. This model has been used to interpret barrier height data and assess junction quality. It has also been found useful for understanding device characteristics reported in the literature.