High-accuracy vectorial Hopkins lithography modeling with full incident-angle Mask3D awareness

Incorporating three-dimensional mask topography (Mask3D) effects under different incident angles is essential for accurate aerial image computation in advanced lithographic modeling. State-of-the-art Hopkins imaging methods fall short in handling Mask3D effects across all incident angles, as they often rely on the assumption of a constant mask transmittance function within divided source sectors, leading to inaccurate imaging predictions. This paper presents high-accuracy vectorial Hopkins lithography modeling enabled by full incident-angle Mask3D awareness. With given outputs of a variable-separated Mask3D model in the form of a series expansion that decouples the source-dependent and spatial-coordinate-dependent terms, the proposed method introduces these source-related terms as an additional dimension into the transmission cross coefficient (TCC), leading to a vectorial Hopkins formulation that fully accounts for the variation of Mask3D effects across all incident angles. A TCC decomposition algorithm is then applied to generate the optical kernels along with their corresponding weighting coefficients, thereby establishing a complete and efficient Hopkins imaging workflow. Simulation results demonstrate that the proposed method can effectively capture the Mask3D effects under all incident angles, achieving superior imaging accuracy and higher computational efficiency compared to existing methods, and is well suited for full-chip lithographic simulations.