Influence of the substrate on the lifetime of capacitive RF MEMS switches

Abstract

We show for the first time that the substrate can influence the lifetime of capacitive RF MEMS switches. We demonstrate that the influence of the substrate should not be ignored. The influence of the environment on the lifetime of a switch is different when it is fabricated on two different substrates. We also present that a switch actuated with a DC voltage lower than the pull-in voltage can pull-in after some time. The goal of the performed experiment was to emphasize the charging of the substrate. The presented results help to understand the substrate charging problem.

Simple and Robust Air Gap-Based MEMS Switch Technology for RF-Applications

“…The main cause of failure for capacitive switching devices is stiction due to interposer dielectric charging of the high-k dielectric between the bridge and bottom electrode. Since there is no interposer dielectric in the device presented here, the only remaining source of charging is substrate charging [2]. Lifetime measurements with 50V unipolar actuation, 100Hz and 50% duty cycle in N 2 as depicted in Figure 16 showed no failure during 10 8 cycles, after which the measurement was stopped.…”

“…On top of this layer a 3μm thick PW-1530 polyimide sacrificial layer is spin-coated and photolithographically patterned. After the 250°C crosslink bake, 500nm deep pits which define the dimples below the membrane are dry etched in the polyimide (2). Next a 10 μm thick nickel film is electroplated through a photoresist mold (3).…”

“…By removing the dielectric from the active area of capacitive switches, many of the original failure mechanisms are removed. However, substrate charging may still remain an issue, as demonstrated in [2] [3]. At the same time, removing the typically thin high-k dielectric from the capacitive device reduces the capacitance ratio, which is an important figure of merit of an RF-MEMS capacitive switch.…”

Degradation of RF MEMS capacitive switches

“…Positive charge will shift the C-V curve in the positive direction, negative charge will shift it in the negative direction. However, while a built-in voltage explains the shifting of a C-V curve, it does not explain the narrowing observed in [23,[40][41][42][43]. Three causes can account for this narrowing effect: Of these three causes, the third one is not very likely, as the spring constant is mainly a function of the size and geometry of the springs and its material properties, which both do not change significantly during stress tests.…”

“…An explanation for this narrowing of the C-V curve was proposed by Rottenberg et al [35] based on the assumption that the charge is injected inhomogeneously. Later publications of the same group [41][42][43][44] reported more about the conditions under which the C-V curve will narrow due to inhomogeneous charge injection. However, there is a second possible explanation for the narrowing effect: a reduction of the gap due to mechanical degradation of the springs will also narrow the C-V curve [44,45].…”