MEMS ultra low leak detection methods: a review

Abstract

PurposeThe purpose of this paper is to review traditional hermeticity test methods when applied to typical micro‐electro‐mechanical systems (MEMS) cavity volumes and to propose potential solutions.Design/methodology/approachStandards for traditional testing have been applied to typical MEMS cavity volumes and the resulting issues of range and sensitivity discussed. In situ test structures have been designed and fabricated with access to the internal cavities to allow characterisation of the structures as a function of pressure.FindingsThe ultra low leak rates necessary to guarantee hermeticity of MEMS cannot be measured using traditional methods. Optical test methods are possible although in situ test structures currently provide the greatest sensitivity. A portfolio of test techniques is required to allow accurate hermeticity testing of MEMS.Research limitations/implicationsThis paper provides a starting point for further investigation into several methods of MEMS hermeticity testing.Originality/valueThis paper provides a review of the limitations of traditional testing and proposals for future testing as the trend towards smaller volume packaging continues.

The Limits of Hermeticity Test Methods for Micropackages

“…Further, the physical detection limits of the methods described in the standards are too high for our application, that is, a package with a leak rate lower than the lowest detectable rate may still let in too much moisture over time. This statement only echoes what has been found in other engineering disciplines (3)(4)(5). Most of the recent work on hermetical packaging has been concerned with microelectromechanical systems (MEMS) where the focus is on the entrance of air inside a cavity with a low internal pressure.…”

Vacuum microdevices

“…Additionally, different gas particles as well as moisture diffuse into MEMS microcavity through microchannels and microcracks in bonding area or in microchamber walls. Moisture is the most deadly factor, because it may condense on moving elements of MEMS and cause their damage (sticking, braking, cracking, corrosion) [4]. To sum up, a few phenomena could be listed, which make it hard or even impossible to achieve high/ultra high vacuum in microcavity:…”

Hermeticity of eutectic bond layers for sensor packages on wafer-level

“…The most commonly used test method is the helium fine leak test described in the Military Standard MIL-STD-883 TM 1014. However, the minimum allowed cavity volume of 0.01 cm3 and the minimum detectable leak rate of approximately 10-11 atm cm3 S-I make the helium fine leak test unsuitable for typical MEMS devices [5]. Another test described in MIL-STD-883G is fine leak testing using Kr-85 radioactive tracer gas.…”

“…Additionally the fabrication of the test structures by standard surface micromachining processes and an easy evaluation by a simple electrical measurement make this method highly applicable for industry. A good overview containing the methods presented here and beyond is also given in [5], [6] and [12]. As measuring the Q-factor has proven to be an adequate method for monitoring Hermeticity, this paper focuses on a novel MEMS resonator test structure to evaluate the potential of eutectic bond layers for hermetical wafer level encapsulation.…”