Process emulation for predicting die shift and wafer warpage in wafer reconstitution

2017 
Wafer reconstitution is a vital process for serving as a buffer to decouple the processing developments between IC fabrication and electronics packaging. By this approach, the IC packaging is then independent from the chip processing. However, such a process brings numerous mechanical loadings during molding and curing phases. Without carefully planning, failures such as die-shifting and excessive wafer warpages are frequently reported and it induces problems for subsequent processing. In this work, it is desired to examine the key factor of die-shift and wafer warpage by performing both fluidic mold flow and solid thermo-mechanical analyses, as well as essential material characterizations. Preliminarily, the die-shift problem is deduced as interaction of fluid load, thermal expansion, shrinkage of molding compound and viscoelastic effect. To have a deeper insight, simplified fluid model and finite element analyses have been constructed to mimic the entire Recon process. For mold flow analysis, a simplified 1-D viscous flow analytical model is adapted. It aims to find the relationship between the molding parameters and the achieved velocity and pressure fields for calculating the possible drag and shear forces acting on dies for causing shift. On the other hand, after molding, the stress and deformation of the entire curing process is then performed by finite element method. The possible die shift and final warpage are then examined step by step to evaluate the contribution from each step and even each processing parameter. Both simplified 2-D axisymmetric and 3D models are performed and analyzed. The preliminary analysis results indicate that the thermal stress during curing is the current dominating factor. Related parameters such as the properties of compounds and carriers and the process parameters such as the curing temperature and duration could be the major controlling factors.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    11
    References
    5
    Citations
    NaN
    KQI
    []