Electrical Contact Resistance and Device Lifetime Measurements of Au-RuO₂-Based RF MEMS Exposed to Hydrocarbons in Vacuum and Nitrogen Environments

Electrical Contact Resistance (ECR) measurements are reported for RF micro-electromechanical switches with Au-RuO₂ contacts, situated within an ultrahigh vacuum system equipped with in situ oxygen plasma cleaning capabilities. Two studies are reported, each involving a comparison of the ECR in vacuum and nitrogen environments for measurements performed immediately after cleaning. The first study reports measurements of initial resistance (resistance measured upon first time closure) versus pressure as dodecane gas is admitted to the chamber. A significant increase is observed at pressures in vacuum as low as 10⁻⁵ torr, (P/P_sat < 10⁻⁴) consistent with earlier reports involving repetitive cycling of macroscopic switches in partial pressures of hydrocarbons in nitrogen. Somewhat unexpectedly, however, the resistance only doubles, even for pressures sufficiently high as to result in full monolayer condensation. In a second study, switch lifetimes in vacuum (10⁻⁸–10⁻⁹ torr) and nitrogen gas environments are compared, for switches operated immediately afterward, or alternatively left open for a number of days before operation. Although it was expected that vacuum would reduce and/or prevent contamination of the electrical contact surfaces, no enhancement or extension of lifetime was observed: Continuous operation of a switch in a nitrogen environment immediately after plasma cleaning was in fact the only procedure observed to indefinitely prolong device lifetime. The results suggest that (1) Hydrocarbon reaction products, but not mobile physisorbed hydrocarbons themselves, are responsible for increasing ECR by orders of magnitude and (2) Repetitive cycling motion of a clean switch in nitrogen inhibits formation of physisorbed hydrocarbon contaminants on the contacts, while vacuum levels far superior to 10⁻⁹ torr are required to prevent contamination.