材料科学
表面微加工
蚀刻(微加工)
薄脆饼
碳化硅
激光器
飞秒
表面粗糙度
硅
光电子学
微电子机械系统
机械加工
深反应离子刻蚀
表面光洁度
各向同性腐蚀
光学
复合材料
反应离子刻蚀
制作
冶金
病理
物理
替代医学
医学
图层(电子)
作者
You Zhao,Yulong Zhao,Lukang Wang
标识
DOI:10.1016/j.sna.2020.112017
摘要
Pressure sensors that can work over 500 ℃ are widely needed in aerospace and petrochemical fields, while silicon based high temperature pressure sensors are difficult to survive such high temperature. Silicon carbide (SiC) shows better mechanical and electrical properties than silicon, which is a feasible candidate to replace silicon for manufacturing high temperature pressure sensor that can work over 500 ℃. However, because of the ultra-high hardness and excellent chemical inertness of SiC, it is difficult to pattern structures on SiC wafer by common etching methods, especially deep etching. Femtosecond laser (fs-laser) is a practicable tool for SiC micromachining as reported in many literatures, but few of them concerns on SiC deep etching in fabricating high temperature pressure sensor’s sensitive diaphragm. In this paper, blind holes with diameter of 1200 μm and depth of 270 μm were fabricated on a 350 μm-thick 4H-SiC wafer by fs-laser, which forms an 80 μm-thick membrane as sensitive diaphragm for pressure sensor. Experiments were conducted to study shape deviation, machining accuracy, edge morphology, sidewall steepness as well as surface roughness of the machined structure. Testing results show that the size errors of machined blind holes are 2.02 % in diameter and 0.38 % in depth, the sidewall steepness is good with an inclination angle of 1.7°, and surface of the sensitive diaphragm is smooth with an average roughness of 320 nm. However, elliptical shape and over etching happens to the fabricated sensitive diaphragm because of inappropriate laser scanning mode, uneven energy distribution and laser irradiation delay. Generally, research shows that fs-laser micromachining is a feasible method for SiC deep etching in fabricating sensitive diaphragm of high temperature pressure sensor.
科研通智能强力驱动
Strongly Powered by AbleSci AI