4.5 Article

Ultra Deep Reactive Ion Etching of High Aspect-Ratio and Thick Silicon Using a Ramped-Parameter Process

期刊

JOURNAL OF MICROELECTROMECHANICAL SYSTEMS
卷 27, 期 4, 页码 686-697

出版社

IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JMEMS.2018.2843722

关键词

Deep reactive ion etching; DRIE; ultra-high aspect-ratio; through wafer DRIE; DRIE with ramped-parameter recipe; through silicon via (TSV); hole and trench DRIE

资金

  1. Micro Autonomous Systems & Technology Program of the Army Research Lab [W911NF-08-2-0004]

向作者/读者索取更多资源

This paper reports an advanced deep reactive ion etching (DRIE) process for realizing ultra-deep (>500-mu m) and ultra-high aspect-ratio (UHAR) silicon structures (AR > 40 for 1-mm through-trench etch, AR approximate to 80 for 500-mu m through-trench etch, and AR > 20 for 500-mu m through-hole etch), with straight sidewalls across a wide range of feature sizes. The challenges of making such structures are overcome by continuously ramping critical parameters of the Bosch DRIE process throughout the process, including the 380-kHz bias power during etch step, the etch/passivation step duration, and the chamber pressure. The masking material capable of enduring the long DRIE process is also discussed; 10-mu m and 25-mu m wide trenches are etched to a depth of >750-mu m and >1000-mu m, respectively, in 1-mm-thick silicon wafers with straight sidewall profiles and flat trench bottoms. Deeper trenches are expected to be etched beyond a 1-mm thick wafer with thicker and/or higher selectivity masking materials. We have also demonstrated etching of circular holes of diameters as small as 25-mu m to a depth of >500-mu m, and potentially with 10-15 mu m diameter holes. This advanced DRIE process offers opportunities for applications ranging from through-silicon via in 3-D CMOS integration to emerging micro- and mesa-scale microelectromechanical system applications that demand ultra-deep and UHAR DRIE.

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

4.5
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