Three-dimensional simulation of FIB-SEM image series from layout data for automatic FIB endpoint detection
Three-dimensional simulation of FIB-SEM image series from layout data for automatic FIB endpoint detection
Tuesday, October 6, 2026: 3:40 PM
Summary:
FIB-SEM processes for semiconductor samples, such as cross-section milling and TEM lamella preparation, have now been automated to the point where process steps that do not require continuous monitoring and frequent immediate corrections by a human operator can be performed reliably without user intervention. However, research is still ongoing into how the operator can be replaced by automatic milling endpoint determination to ensure precise target preparation. At present, there is no solution that can be universally applied to a wide variety of different semiconductor devices. Since the internal structures of every device produced are based on a mask layout, it is obvious to use that layout as a reference for automatic endpoint detection. We are presenting a new approach for generating synthetic 3D FIB-SEM data from CAD layout files to serve as reference data for precise automatic determination of the target preparation endpoint.
FIB-SEM processes for semiconductor samples, such as cross-section milling and TEM lamella preparation, have now been automated to the point where process steps that do not require continuous monitoring and frequent immediate corrections by a human operator can be performed reliably without user intervention. However, research is still ongoing into how the operator can be replaced by automatic milling endpoint determination to ensure precise target preparation. At present, there is no solution that can be universally applied to a wide variety of different semiconductor devices. Since the internal structures of every device produced are based on a mask layout, it is obvious to use that layout as a reference for automatic endpoint detection. We are presenting a new approach for generating synthetic 3D FIB-SEM data from CAD layout files to serve as reference data for precise automatic determination of the target preparation endpoint.