No abstract

Reliability issues currently hamper the commercialization of capacitive RF MEMS switches. The most important failure mode is parasitic charging of the dielectric of such devices. In this paper we present an improved analytical model that enables us to calculate and understand the effect of insulator charging on the behavior of capacitive RF MEMS switches, and to describe the way they fail, and their reliability. Emphasis is placed on a shift of the pull-out voltage to predict failures. Tests with capacitive RF MEMS switches have been performed that validate the most important features of the model.

One of the most important reliability problems in micro-electromechanical systems (MEMSs) is stiction, the adhesion of contacting surfaces due to surface forces. After reviewing the known physical theory, and the measurement method commonly used to investigate stiction, we present a model that can be used to investigate the sensitivity of MEMS to stiction. It quantitatively predicts the surface interaction energy of surfaces in contact. Included in the model is the roughness of the contacting surfaces and the environmental conditions (humidity and temperature). This is done by describing the surface interaction energy as a function of the distance between the surfaces. This distance is not a unique number, but rather a distribution of distances. It is shown that, if we know this distribution, we can calculate the surface interaction energy. The model is suitable for the prediction of forces due to capillary condensation and molecular interactions.

This paper presents a comprehensive review of the reliability issues hampering capacitive RF MEMS switches in their development toward commercialization. Dielectric charging and its effects on device behavior are extensively addressed, as well as the application of different dielectric materials, improvements in the mechanical design and the use of advanced actuation waveforms. It is concluded that viable capacitive RF MEMS switches with a great chance of market acceptance preferably have no actuation voltage across a dielectric at all, contrary to the ‘standard’ geometry. This is substantiated by the reliability data of a number of dielectric-less MEMS switch designs. However, a dielectric can be used for the signal itself, resulting in a higher Con/Coff ratio than that one would be able to achieve in a switch without any dielectric. The other reliability issues of these devices are also covered, such as creep, RF-power-related failures and packaging reliability. This paper concludes with a recipe for a conceptual ‘ideal’ switch from a reliability point of view, based on the lessons learned.