Wafer Inspection Optical Modes Tuned by IC Layout Simulation

The patent creates virtual copies of wafers and defects through 3D modeling and simulation. Instead of repeatedly inspecting real wafers, the system generates virtual wafer models with embedded defect models and uses optical simulation to predict inspection outcomes. This copying approach maintains measurement precision while eliminating the time cost of physical inspection iterations.

The system performs preliminary optimization actions by simulating optical parameter effects on virtual defects before actual inspection. The defect model and optical simulation engine evaluate potential optical parameter combinations in advance, identifying optimal settings without requiring repetitive real-world inspections. This preliminary action resolves the contradiction by preparing optimal parameters computationally before physical inspection occurs.

The patent introduces an intermediary simulation engine that mediates between optical parameters and defect detection outcomes. Instead of directly testing multiple parameter combinations on real wafers, the simulation engine acts as an intermediary, computationally evaluating how different optical parameters would affect defect visibility. This intermediary layer simplifies the complexity by providing a virtual testing environment that predicts real-world performance without physical complexity.

The system systematically changes optical parameters in the simulation environment to evaluate their impact on defect detection. The optical simulation engine varies parameters such as wavelength, numerical aperture, and polarization in virtual settings, allowing comprehensive evaluation of detection capability across multiple parameter combinations without the complexity of physical reconfiguration. This parameter exploration in silico maintains high detection capability while reducing operational complexity.

The patent creates a universal optimization framework where a single simulation engine can evaluate optical parameters across multiple defect types and process conditions. The system models various defect categories (particles, pattern defects, overlay errors) and fabrication variations within the same virtual environment, allowing the identification of optical parameter sets that maintain precision across diverse scenarios. This multi-functional simulation approach resolves the contradiction by enabling one system to serve multiple inspection needs simultaneously.

The system dynamically adapts the simulation models to reflect changing process conditions and defect types. Rather than static optimization for a single defect category, the simulation engine can reconfigure virtual defect models and process parameters to match current inspection needs. This dynamic modeling capability maintains detection precision while improving adaptability, as the simulation environment can be rapidly reconfigured without physical system changes.