Impact of Environmental Conditions on the Contact Physics of Gold Contact RF Microelectromechanical Systems (MEMS) Switches.

Abstract

RF MEMS switch technology is poised to create a new generation of devices capable of vastly outperforming current mechanical and semiconductor switching technology. Despite the efforts of top industrial, academic, and government labs, commercialization of RF MEMS switches has lagged expectations. This dissertation focuses on issues associated with switch contact physics. Understanding the failure mechanisms for metal contact switches is a complex challenge. There is strong interplay between variables such as mechanical creep, deformation, contact heating, contact asperity size, real contact area, and current flow leading to the eventual failure of the switch. Stiction failures moreover are highly sensitive to ambient conditions and absorbed film layers at the switch contact. The experiments in this thesis seek to isolate individual failure mechanisms and tie them to the physics driving that behavior through correlation of experimental data and theoretical modeling. Four experiments in controlled environments were performed: 1) the impact of cryogenic temperatures on RF MEMS contacts, 2) a correlation between experimental data and theoretical modeling for gold asperity creep at room and cryogenic temperatures, 3) a power law relationship between contact resistance and time dependent creep, and 4) the pressure dependence of switch closure. Cryogenic temperatures were used to isolate contaminant film effects. Contaminant films were found to have less mobility at 77 K, and contact resistance measurements showed that the film could be reduced on the contact surface through mechanical cycling and high temperatures at the gold asperities. It was also noted at cryogenic temperatures that the choice of atmosphere was important. A nitrogen atmosphere at liquid nitrogen temperature produced variable contact resistance as the condensed liquid boiled off the switch contacts. Data was correlated with a single asperity creep model to show that change in contact resistance as a function of time is related to the creep of gold asperities at the contact interface. The change in contact resistance over time can be described by a power law relation derived from the single asperity creep model that takes into account the surface topography, material characteristics, and contributions from additional sources such as adsorbed film layers. Additionally, it was shown that the creep mechanism was temperature dependent and that creep was significantly decreased at cryogenic temperatures. A drop in pressure as a result from cryogenic temperatures was observed to create switch bounce at closure. This was explored in a set of room temperature experiments which established the onset of bounce at 60 Torr. The results were and correlated to the damping coefficient and the ratio of the damping force to the electrostatic force of actuation. This work contributes to the field of contact physics and MEMS switch technology by firmly establishing the role of creep in contact mechanics and quantifying its time and temperature dependent impact on contact resistance.