부산대학교 도서관

SEM review is increasingly being utilized for defect analysis and classification. Root cause analysis typically required both detailed defect imaging and elemental analysis. The use of SEM ADC facilitates the procedure but requires both high re-detection rate and high confidence in the classification. Results are presented for a defect review SEM with ADC, the SEMVision CX. The key ADC parameters of re-detection rate, accuracy and purity are established for several layers. The results provide sufficiently high confidence to incorporate the defect review SEM as part of the inspection flow along with optical and manual classification. Various aspects of the implementation facilitate the defect identification and analysis including: consistent ADC, automated EDX for bare wafer, and the use of a "killer index" which assigns a likelihood of yield impact to a given defect. The implementation has helped to rapidly identify process tool issues. The defect review SEM capabilities of automated defect review and classification are being utilized in the manufacturing environment, which requires high turn around time and reliable information.