An Experimental and Computational Study on Moisture Induced Epoxy Swelling in Non-hermetic Optoelectronic Packages

Abstract

Moisture induced epoxy swelling is a potential failure mechanism in nonhermetic packages. Epoxy materials used in the package absorb moisture and swell in a relatively humid environment. This will result in hygroscopic stresses in the material that can eventually lead to failure. The coefficient of hygroscopic swelling (CHS) is a material property that characterizes moisture induced swelling in the material. It is defined as the ratio of hygroscopic strain to the moisture concentration in the material. Prior research investigated the measurement of CHS experimentally using techniques such as thermo mechanical analysis (TMA) (Ardebili et al., 2003, “Hygroscopic Swelling and Sorption Characteristics of Epoxy Molding Compounds Used in Electronic Packaging,” IEEE Trans. Compon. Packag. Technol., 26(1), pp. 206–214; Mckague et al., 1978, “Swelling and Glass Transition Relations for epoxy Matrix Material in Humid Environments,” J. Appl. Polym. Sci., 22, pp. 1643–1654.), Moiré interferometry (Han et al., 2003, “Measurement of the Hygroscopic Swelling Coefficient in Mold Compounds Using Moire Interferometer Experimental Techniques,” IEEE Trans. Compon. Packag. Technol., 27(4), pp. 40–44), and digital image correlation (DIC) (Park and Zhang, 2007, “Investigation of Hygroscopic swelling of Polymers in Freezing Temperature,” ASME International Mechanical Engineering Congress and Exposition). Some of these studies recommended investigation of improved measurement techniques, while others made some procedural assumptions that may not be applicable for all materials. One of the goals of this study was to investigate an improved technique for CHS measurement and helps to better understand the various factors that affect the measurement. The DIC technique was used to measure the moisture swelling of the epoxy material considered for use in the package. Moisture loss during the measurement results in a change in moisture concentration in the sample. While it may be thought that the moisture loss during the DIC scan can be assumed negligible due to the short test time compared with other methods, this assumption did not hold well for the current epoxy material. The ramp rate chosen for the test affects the moisture loss. It introduces a level of nonuniformity in temperature and moisture distribution in the sample. A suitable value that takes into account both of these effects was determined. The moisture loss measured during the DIC scan was accounted for in the CHS computation. In addition to this, the temperature and concentration dependence of the CHS was determined. Results indicated that the temperature and concentration effects are small for the current test material within the test temperature range. The moisture loss found during the DIC measurement leads to a nonuniform moisture distribution in the sample. This was characterized by using experimental and computational methods and the effect on the measurement was determined.