Skip to Main Content
This paper presents for the first time the effects of moisture on radio frequency microelectromechanical systems (MEMS) switches packaged entirely inside a flexible, organic polymer [namely, liquid crystal polymer (LCP)]. Moisture tests were administered at 100degC and 100% relative humidity to evaluate long-term exposures and at 85degC and 85% relative humidity to evaluate short-term exposures. The effect of the moisture was quantified by before and after S-parameter measurements, by a weight gain analysis, and by visual inspection. Both global and localized bonding techniques were investigated to compare the best-case scenario to a more practical case. The effects of an 18 mum thick copper layer on both sides of the package were studied as well as the size of the bonding contact area. It was found that many packages that passed the Military Standard 883 G, Method 1014.12 for seal quality were unable to provide adequate protection from moisture. This indicates that the requirements for MEMS devices is more rigorous than the Military Standard. This standard is commonly quoted in literature as a metric for qualifying polymer packaging techniques. This paper demonstrates the necessity for proper testing of MEMS devices in a moist environment. It has been determined that for the bonding methods presented in this paper, an LCP packaged MEMS switch could potentially survive 7-10 h in jungle conditions, 5-7 weeks in ambient conditions, or 1.4-1.8 years in desert conditions.