Nondestructive Detection of Micropipes and Inclusions in 4H-Sic Substrate by Laser-Based Optical Surface Inspection System

The performance of 4H-SiC devices is greatly affected by “killer” defects (defects that affect yield) that form during crystal growth. Along with improving growth technologies that reduce defect density, inspection technologies that can identify and locate defects have become a critical condition for manufacturing processes. In this paper, the defects of micropipes and inclusions in 4H-SiC substrate are analyzed using Candela CS, SICA and Eagle, which are the mainstream SiC defect detection devices in the market, and the imaging characteristics of micropipes and inclusions in different detection environments are explained. Among them, Candela CS can effectively distinguish micropipes and inclusions by using the obvious morphological characteristics of Scatter Normal channels and Specular reflection channels. Eagle can effectively identify micropipes and distinguish them from inclusions by using the stress birefringence phenomenon of micropipes under bright-field polarized light. According to the imaging characteristics of these two defects, the analysis formula can be established and the defect distribution can be plotted. However, there is no difference in the imaging of micropipes and inclusions in the bright field and PL mapping in SICA, and the key to distinguish micropipes and inclusions in 4H-SiC substrate detection is the birefringence effect of micropipes under the action of polarized light.