Optimization of 2.2D Underfill Process by Novel Methodology and Direct Observation of Capillary Underfill Process

The process of capillary underfill encapsulation encounters productivity issue of long filling time that draws additional manufacturing costs. Since the chip size becomes larger, multiple passes are required for fill and fillet passes to fill up the cavity and create the sealing joint. Traditionally, to obtain an adequate dispensing design (pass route and order) for the capillary process to fill up the package without obvious voids, a long waiting time between passes is set to ensure that the underfill moving front for each pass to halts dispensing before the next pass. Based on our direct underfill observation experiments, the underfill dispensing waiting time is unnecessarily too long. In this work, to reduce the cycle time of underfill process without extensive and tedious trial and error, a novel methodology for optimizing dispensing strategy is proposed. A well-validated model-equivalent bump group (EBG) model of 3D CAE modeling for capillary underfill process is applied to predict the optimal waiting time between passes for dispensing design. Moreover, in-situ observation of underfill process experiment is employed to verify the methodology. It shows that when the optimized waiting time for each pass is applied, the propagation of the underfill moving front by EBG model is like that of the experiment with a long waiting time for each pass. This methodology can optimize the underfill process to reduce cost and speed up production time compared to the conventional try and error underfill process.