Fan-out wafer molding process optimization using cure-dependent viscoelastic modeling and characterizations

Due to the significant volume change of molding compounds during assembly processes, warpage is the top issue to be overcome in the application of the fan-out wafer form. Minimizing warpage of the fan-out wafer during the molding process, though challenging, is essential for ensuring good assembly process yields. In this paper, we propose a predictive modeling approach to accurately address warpage evolution and to effectively mitigate warpage by optimizing molding process conditions. This approach establishes a viscoelastic-based material model accounting for cure-dependent viscoelastic relaxation behavior of molding compound. The material model can be generated by material characterization data using differential scanning calorimetry (DSC) and dynamic mechanical analysis (DMA). This cure-dependent viscoelastic model is then incorporated into finite element (FE) analysis to simulate warpage evolution over the entire compression molding curing process, including compression molding cure (CMC) and the subsequent post molding cure (PMC). The simulation results are compared to in-line warpage measurement data to validate the integrated process modeling approach. This study found that the finite element method using cure-dependent viscoelastic model is highly suitable for the predictions of fan-out wafer level packaging warpage. Furthermore, optimized curing condition was found to effectively improve warpage performance to enhance throughput, process yield, and package reliability results. Warpage concerns can be minimized by good predictions to determine solution, reduce design cycle time and cost significantly.