A Physics-Based Predictive Modeling Framework for Dielectric Charging and Creep in RF MEMS Capacitive Switches and Varactors

Abstract

In this paper, we develop a physics-based theoretical modeling framework to predict the device lifetime defined by the dominant degradation mechanisms of RF microelectromechanical systems (MEMS) capacitive switches (i.e., dielectric charging) and varactors (i.e., creep), respectively. Our model predicts the parametric degradation of performance metrics of RF MEMS capacitive switches and varactors, such as pull-in/pull-out voltages, pull-in time, impact velocity, and capacitance both for dc and ac bias. Specifically, for dielectric charging, the framework couples an experimentally validated theoretical model of time-dependent charge injection into the bulk traps with the Euler-Bernoulli equation for beam mechanics to predict the effect of dynamic charge injection on the performance of a capacitive switch. For creep, we generalize the Euler-Bernoulli equation to include a spring-dashpot model of viscoelasticity to predict the time-dependent capacitance change of a varactor due to creep. The new model will contribute to the reliability aware design and optimization of the capacitive MEMS switches and varactors.