Critical Defect Detection at 3nm Technology Node: Enhanced Detection Using DUV Optical Inspection Technology With AI-Based Algorithm

In this paper, we present the results of a study conducted using cutting-edge deep-ultraviolet (DUV) brightfield wafer inspection technology, enhanced by a new artificial intelligence (AI) based image enhancer algorithm. The study focused on the most advanced design rule technology and demonstrated a reduction of up to 2.8x in the number of defects that are sent for Scanning Electron Microscope (SEM) review, while maintaining an equal or better Capture Rate of key Defects of Interest (DOI) compared to the benchmark inspection recipe. These results highlight the potential value of the AI-based algorithm in achieving tighter defect control and enabling higher yield, all while minimizing the Fab Cost of Ownership.