Search for a command to run...
ABSTRACT At WLPS 2023 the initial work done under an iNEMI project was presented with plans for the next steps in the investigation of RDL (redistribution layers) adhesion measurements. This presentation will cover the additional work done for peel testing of 20, 10, and 5 micron (μm) wide traces, plus the ongoing Finite Element (F.E.) Modeling used to simulate the actual shear and peel test data, plus extrapolate for RDLs of various sizes and geometry ratios. As mentioned last year, in 2019 iNEMI published a technology roadmap that included challenges and unknowns. One of those challenges was determining the actual adhesion strength of RDLs to materials used within WLPs (wafer level packages), PLP (panel level packages), and on advanced packages and/or substrates with fine pitch traces. These fine lines are typically exposed to the manufacturing environment, which can result in the degradation of adhesion between the RDL traces and other materials, such as the encapsulant, die, and other dielectrics. It has been assumed that the adhesion properties measured with an older IPC test method, IPC-TM-650, for Cu (Copper) traces as small as 1mm (1000μm) and larger would also apply to thinner and fine traces 20μm wide and smaller. There were concerns about characterizing the adhesion strength at the relevant length and small scales needed to duplicate WLP applications. The test method(s) developed may be further affected by flaws and features that are not significant in larger trace patterns. In addition, the traditional test method(s) such as peel and shear, do have challenges, including, but not limited to, the risk of trace breakage and tool load detection limitations for traces below 20μm. This updated presentation will briefly review the test methods considered for the iNEMI project, the test samples designed with 20, 10 and 5μm features, and the adhesion shear strength data previously presented. New peel test data obtained from multiple contributors for 20, 10 and 5 μm will be presented along with the F.E. modeling developed to duplicate the actual shear and peel test data. In addition, the results of the modeling used to simulate smaller RDL features in width and ratios (width to thickness) will be presented and discussed for future considerations.