A Methodology for Identifying Capillary Underfill Optimized Material Properties

Complex semiconductor device architecture, requires memory, processing, and other electronics components packaged together leading to larger package size and more challenging stress distribution for reliability performance. Modeling can help understand and address the semiconductor packaging challenges, also by introducing novel material modeling techniques which assist in fine tuning the material properties required for reliable packaging. These new techniques are suitable for applications of Capillary Underfill (CUF) materials that are a family of materials which provide support to solder connections and help improve the fatigue life of advanced packages. This work proposes a method to create freely designable virtual viscoelastic material models. These material models are incorporated into a finite element modeling strategy of a flip chip Ball Grid Array (BGA) package. A comprehensive Design of Experiment (DOE) using the finite element model covers the design space for possible CUF materials. A Response Surface Method (RSM) connects material properties input to the model response predicted by the finite element analysis. The effect of different material properties on the model outputs are studied. As a final step, optimized CUF material properties are proposed based on output objectives.