期刊
出版社
ELSEVIER SCIENCE INC
DOI: 10.1016/j.precisioneng.2022.06.003
关键词
Brittle material; Silicon wafer; Surface characterization; Surface roughness; Subsurface damage; Diamond wire sawing
类别
资金
- National Natural Science Foundation of China [12104074]
- China Postdoctoral Science Foundation [2020M683233]
- Chongqing Special Postdoctoral Science Foundation [XmT20200021, XmT20200043]
- Innovation and Technology Commission (ITC) of the Government of the Hong Kong Special Administrative Region, China [GHP/142/19SZ]
A theoretical model is developed to accurately and quickly determine the surface roughness and subsurface damage depth of silicon wafers by considering fracture width. The model integrates digital image processing to extract fracture parameters and shows good agreement with experimental results. This image processing-based model provides a reasonable approach for estimating damages in silicon wafers during wire sawing.
The damages of silicon wafer in diamond wire sawing have a significant impact on its mechanics, economy, and use properties, which should be evaluated accurately and quickly. A theoretical model is developed to determine the surface roughness Rz (ten-point mean height) and subsurface damage (SSD) depth by the fracture width of silicon wafer. The model takes into account the scratch groove direction and material pile-up effect of silicon wafer. A digital image processing method is integrated into the model to extract the fracture parameters of silicon wafer. Many silicon wafers are processed under different slicing parameters, for which the fracture parameters and surface roughness Rz are measured by confocal microscopy, and the SSD depth is measured with the method of cross-section scanning microscopy. The scratch-induced material pile-up height is evaluated by nanoscratching. The calculated and measured fracture parameters, surface roughness Rz, and SSD depth are analyzed by contrast. The result shows that the model can accurately and quickly determine the surface roughness Rz and SSD depth with an average relative error of less than 12.0% in 20 s. The distribution characteristics of fractures, surface roughness, and SSD are analyzed. The result shows that more fractures have relatively smaller fracture width, depth, length, or related SSD depth. The image-processing-based model would be a reasonable approach for estimating the damages in silicon wafers during wire sawing.
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。
推荐
暂无数据