Influence of Oxide Layers on the Microwave Surface Resistance of Superconducting Niobium.
Fred Lawrence Palmer, Ph.D.
Cornell University 1988
Abstract
We have measured the influence of oxide layers grown at room temperature on the microwave surface resistance of superconducting niobium by comparing the quality factors of oxidized and unoxidized cavities. We have also investigated the microwave and surface properties that are altered when oxide layers are heated to temperatures near 300°C. Auger spectroscopy was used to determine a procedure for producing clean oxygen-free surfaces by ultra-high-vacuum firing the niobium at 1100-1400°C. A special vacuum furnace was then constructed so that the 8.6 Ghz. niobium cavities could be fired at these temperatures and then taken to a cryostat for testing without any exposure to air. These cavities exhibited residual resistances of 6-12 nOmega. Exposure of niobium to oxygen at room temperature is known to create a ~ 15A thick layer consisting mostly of Nb
2O
5. This layer changed the temperature dependent BCS resistance by less than 2%, and changed the low temperature residual resistance by less than 1.5 nOmega. Several cavities were heated to 325° for 10 minutes to demonstrate that the "clean" cavities were not contaminated with oxygen. Heating an oxidized cavity causes the oxide layer to dissolve into the first ~ 1000 A of the bulk metal. The resulting oxygen-rich layers had ~ 20% less BCS resistance due to the reduced mean free path, and had residual resistances of 40 n Omega or more. The "clean" cavities were unaffected by heating, demonstrating that they were actually oxide-free. Several other measurements were made on oxidized cavities that had been heated to temperatures near 300
o. The RF critical temperature of these cavities was reduced by ~5%; the energy gap was unaffected (+/- 1%); and the dependence of residual resistance on ambient magnetic field during cooldown was increased from ~0.4muOmega/gauss to ~1muOmega /gauss. Comparison of these cavities with the XPS data of Kirby et al. showed that the increase in residual resistance did not correlate with the presence of any specific oxide, but appeared to correlate with the concentration of dissolved oxygen near the surface.
For a printed version send e-mail to:
preprint@mail.lns.cornell.edu
