Diffusion driven microstructural evolution and its effect on mechanical behavior of SnAgCu solder alloys
Abstract
In microelectronic assemblies, the fatigue life of solder joints is a critical element of the overall product reliability. Fundamental to understanding the fatigue life of solder joints is the creep behavior of the solder alloys. This is since the alloys experience homologous temperatures ranging from 0.4-0.8T m during field use conditions. Specifically, the popular Pb-free candidates, SnAgCu solder alloys, evolve in their microstructure, and as a result in their mechanical behavior, with time under field use conditions. The challenge in developing predictive models of the mechanical behavior of these solder joints is twofold. The first challenge is in developing microstructural descriptors to characterize the evolution of the microstructure in these alloys under conditions that induce aging. The second challenge is in connecting the microstructural evolution to a corresponding evolution in the mechanical behavior. Traditionally, constitutive laws of creep deformation describe a relation between stress, strain, strain rate and temperature at a given "state" of the material. Hence, material models for an alloy that is subjected to continuos evolution in microstructure can only be indirectly characterized using traditional viscoplastic constitutive laws since the connection to microstructure has to be made through the state variables. Furthermore, as the solder joint size shrinks, pad finish, reflow cooling rate, coarsening due to faster diffusion become additional critical factors in determining the creep behavior. During reflow, the joints undergo non-equilibrium solidification with Sn undercooling considerably "locking" the Ag3Sn and Cu6Sn5 precipitates in the matrix. With time, the microstructure evolves into a thermodynamically stable state. The primary force of phase transition/evolution is the "diffusion potential" between the meta-stable (unaged) state to a stable (aged) state. The research described in this thesis is based on the hypothesis that there exists a relation between the microstructure and the mechanical behavior observed in SnAgCu solder alloys. As a preliminary effort to the main thrust of the thesis, the mechanical behavior variation of the solder joints is studied as a function of surface finish, reflow cooling profile and alloy composition. Diffusion being the critical factor for secondary phase evolution, microstructures were next analyzed as a function of aging. Specifically, two microstructural descriptors are considered in this study: the size of the Ag3Sn inter metallic particles and the size of the βSn dendrite. Three dimensional phase growth and coarsening of the microstructure is extensively characterized with the help of stereological techniques applied on two-dimensional micrographs. The thesis lastly focuses on addressing the effect of mechanical behavior of SnAgCu alloys as a function of the microstructure by developing functions to relate the creep model parameters to underlying microstructural parameters. The versatility of the creep model is demonstrated via the ability to predict creep behavior of hypo-eutectic SnAgCu joints with an unknown process, design and aging history, but with known microstructural state. For this work, more than one hundred and forty eight mechanical creep and steady state strain rate experiments were performed apart from at least around thirty six metallographic experiments that involved the use of electron microscopy techniques. The developed microstructurally adaptive creep model is validated rigorously in three ways. First to demonstrate the model's ability to fit to data used in developing the model, second in predicting both the microstructure as well as the creep behavior at aging conditions not used to develop the model and lastly by predicting the behavior of a solder joint consisting of a different Ag composition relative to those used for model development. In all three cases, the predictions of creep behavior are shown to be in statistical agreement with the experimental observations.
Degree
Ph.D.
Advisors
Dayananda, Purdue University.
Subject Area
Inorganic chemistry|Mechanical engineering|Materials science
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.