Predictive Numerical Modeling of Stealth Dicing Process for Different Wafer Pre-Thin Thicknesses

Stealth Dicing (SD), a cutting-edge technology crucial for the semiconductor industry, enables precise wafer separation into individual dies by inducing subsurface damage and cracks using infrared light. The optimal process sequence, SD before grinding, includes a pre-thinning step that enhances efficiency and reduces defect risk during SD. However, our previous work shows variations in wafer pre-thin thickness impact temperature and stress distribution during laser absorption, influencing SD performance. This study as an extension from previous work to systematically evaluates the influence of wafer pre-thin thickness, correlating simulation findings with in-house experimental data. Furthermore, SD recipe was optimized through the simulation model and validated with experimental data to address the non-separation issue. This developed model for SD process can be used to predict and optimize the process recipe with different wafer pre-thin thickness.