Reliability Modeling of Microelectronic Interconnections in Long-term Applications

Abstract

Solder alloy materials are used to form mechanical and electrical connections between printed circuit boards and electronic components. Enhancing the reliability of electronic assemblies is mostly dependent on the reliability of the interconnected solder joints. Therefore, it is crucial to estimate the reliability and fatigue behavior of solder joints under realistic operating conditions. Mechanical and thermal cyclic stresses are the most common factors that lead to failures in solder joints. Aging is another factor that can change the mechanical and fatigue properties of solder joints. One of the most popular solder alloys that are utilized in fabricating solder joints is SAC 305 (96.5 % tin, 3% silver, and 0.5% copper). Adding bismuth (Bi) to SAC-based solder materials could lead to significant evolutions in their fatigue and mechanical properties. Modeling the reliability and the mechanical properties of the SAC-based solder joints under several levels of cycling stress and different aging conditions is the main objective for this dissertation. Instron 5948 Micromechanical Tester is used to perform accelerated shear fatigue and shear tests in order to measure the solder joints performance. Several experimental parameters were utilized in the dissertation, which include: Six levels of stress amplitude (16, 20 and 24MPa ’SAC305’ 22, 26 and 28MPa ’SAC-Q’), four levels of aging time (2, 10, 100 and 1000 hrs), and three levels of aging temperature (50, 100 and 150ᵒC). In the first study, the effcet of different stress amplitudes and aging time conditions on the fatige properties of SAC 305 solder joints with OSP surface finish are investegated. The aging temperature effects on the fatigue and shear properties of SAC 305 solder joints with OSP surface finish are examined in the second study. The effect of aging time on the fatigue and shear properties of SAC305 and SAC-Q (SAC-based with bismuth) solder joint iii materials cyled at different stress amplitudes are demonstared in the third study. A Two-parameter Weibull distribution is constructed for each experimental combination to assess the reliability of solder alloys with different conditions. A prediction model is built to assess the fatigue life at different stress amplitudes, aging times, and aging temperatures. The evolutions in the hysteresis loop, plastic strain and the inelastic work at different conditions are determined. Coffin-Manson and Morrow energy models are implemented to estimate the fatigue life as a function of plastic strain and inelastic work per cycle, respectively. Arrhenius model is utilized to describe the effect of aging temperature on the fatigue and shear properties of the solder joints. The artificial neural networks (ANNs) technique is applied to estimate the fatigue life at different experimental combinations. In study four, a new methodology is proposed to determine the optimal process parameters for reliability data by identifying different reliability indices. The fatigue performance of the solder joints under accelerated conditions of the thermal cycling environment is considered as validation case study. Surface finishes, solder paste alloys, and solder sphere materials are the process parameters for this study. Those indices are the scale parameter from the Weibull distribution, the signal to noise ratio, the earlier failure probability and normalized mean-variance. After determining those indices which are considered as the performance response, the multi-response optimization problem is initiated. Fuzzy logic will be then utilized to solve the multi-response problem. The study results indicate that significant reductions in the reliability and shear strength are observed when the loading level, aging time or aging temperature are increased. Fatigue properties (plastic strain and inelastic work) are increased when the stress amplitude or aging levels are increased. Adding Bi into SAC-based solder alloy leads to significantly enhance the solder joints fatigue resistance and decrease the amount of degradation in the solder joints reliability.