3.8 Proceedings Paper

Design and Demonstration of Highly Miniaturized, Low Cost Panel Level Glass Package for MEMS Sensors

出版社

IEEE
DOI: 10.1109/ECTC.2017.283

关键词

glass; MEMS; sensors; package; panel; cost; hermeticity; feedthrough; bonding

向作者/读者索取更多资源

This paper describes an ultra-thin, low cost 3D glass sensor packaging platform for near-hermeticity with novel feedthrough and encapsulation technologies. Glass panels of thicknesses ranging from 50 mu m to 300 mu m are used which limits overall form factor to < 0.7 mm. A process flow for fabrication of cavity/embedded sensor packages is described with demonstration three unique fundamental technologies. Vertical electrical feedthroughs are demonstrated using a low-cost conductive Transient Liquid Phase Sintering (TLPS) paste in a high throughput process. Lateral electrical feedthroughs embedded in polymer trenches are proposed for higher reliability, better coplanarity, reduced vulnerability to chemical corrosion and lower parasitics. Finally, four different adhesive polymers are explored to demonstrate a low temperature glass-glass panel bonding technique. Samples bonded at fixed conditions using the four polymers showed sufficiently high bond strength (> 10 MPa) and Dow Chemical's Benzocyclobutene (BCB) 14-P005 is found to be the best candidate for panel level glass-glass bonding. Modelling of the proposed three-layer glass packaging platform was performed in COMSOL Multiphysics. Results show a maximum deformation of about 2.3 mu m - 2.5 mu m in the BCB and GX-92 bonded package and the least average internal stress of 6.40 MPa in the BCB bonded package. The complete manufacturing cycle starting from cavity formation on bare glass to final 3D assembly to form the lidded/open cavity package including singulation is panel based, enabling significant cost reduction (depending on die dimensions and panel size) compared to ceramic and other substrate technologies.

作者

我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。

评论

主要评分

3.8
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

暂无数据
暂无数据