Behavior of Sn-Ag-Cu Solder Bumps During Thermal Cycling
E. Tsepi (TU Delft - Mechanical Engineering)
M.J. Santofimia Navarro – Mentor (TU Delft - Team Maria Santofimia Navarro)
Lara Barros Rebouças – Mentor (NXP Semiconductors)
C. Kwakernaak – Graduation committee member (TU Delft - Team Maria Santofimia Navarro)
H. Beladi – Graduation committee member (TU Delft - Team Maria Santofimia Navarro)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
The present study investigates the microstructural evolution of Sn-Ag-Cu (SAC) solder joints subjected to thermal cycling, with emphasis on recrystallization and grain orientation. Thermal cycling is the main failure source in reliability tests of electronic devices, yet its microstructural effects remain insufficiently understood. Although numerous studies have examined solder joint fatigue using numerical modeling, the early microstructural changes are often overlooked. To address this gap, the present work characterizes experimentally stages of thermal-cycling-induced microstructural evolution.
The evolution of SAC solder joints was characterized using Electron Backscatter Diffraction (EBSD) on two sets of samples. (i) joints thermally cycled for an increasing number of cycles and subsequently cross sectioned, and (ii) pre-cross-sectioned joints analyzed after consecutive rounds of cycling. Two distinct recrystallization mechanisms, namely primary and continuous dynamic, were identified, stabilizing after three thermal cycling stages. Grain rotation toward the [001]Sn-substrate angle and activation of slip systems 4, 6, and 10 suggest a strong link to the initial reflow texture. Moreover, IMC coarsening which facilitates grain boundary unpinning and promotes crack initiation, was also observed. Recrystallized grains exhibited a decrease in Young's Modulus, likely associated with orientation effects.
Calculation of Geometrically Necessary Dislocation (GND) density and Stored Energy Density (SED), with the aim to link EBSD and numerical modeling, while adequate for a first estimation, showcases the need for improved integration methods. Finally, the use of pre-cross-sectioned samples for in situ study of microstructural evolution is not recommended due to stress relaxation effects.