Optimal Light Source Selection in SMO Simulation: A Multi-criteria Approach

Sustaining Moore’s Law requires advancements in lithographic Resolution Enhancement Techniques (RET), particularly Source Mask Optimization (SMO). Parametric SMO optimizes standard source shapes, such as dipole and annular by adjusting parameters like sigma and illumination angle to enhance Depth of Focus (DOF). However, when multiple source shapes achieve maximum DOF, additional lithographic metrics must guide selection. This research evaluates five source selection algorithms based on Edge Placement Error (EPE), Mask Error Enhancement Factor (MEEF), Process Variation (PV) Band, Normalized Image Log-Slope (NILS), and DOF. Results show that Algorithm C, balancing MEEF and NILS, offers superior lithographic performance, compromising 2.71% in MEEF and 18.31% in NILS compared to optimal values. By integrating multi-criteria optimization, this study advances lithographic techniques for scaling metal and gate pitches, improving print fidelity and yield in advanced semiconductor manufacturing.