Authored By:
Hannah Fowler, Sukshitha Achar Puttur Lakshminarayana, Sui Xiong Tay, Yifan Wu, Ganesh Subbarayan, John Blendell, Carol Handwerker
Purdue University
Indiana, USA
Summary
The low melting temperature of eutectic Sn-Bi alloys (139°C) makes eutectic Sn-Bi a suitable low-temperature alternative to high-Sn, lead-free solders in electronics packaging. The low eutectic temperature allows for a peak reflow temperature of 180°C rather than the 240°C reflow temperature required for Sn-Ag-Cu (SAC) alloys. Lower reflow temperatures reduce warpage-induced solder joint defects. However, the strain-rate sensitivity of Sn-Bi alloys results in poor drop-shock performance despite having high reliability in thermal cycling.
As reported in the literature, microalloying with Sb and Ag has shown improved ductility and lower strain-rate sensitivity that are known to enhance drop-shock performance and result in a more reliable joint. In this paper, we will compare the bulk microstructure of eutectic Sn-Bi with Sb and Ag additions to solder joints on Cu substrates and ENIG (electroless nickel immersion gold) surface finishes. Our results demonstrate how small changes in composition from the substrate or surface finish alter the microstructure in this system and the impact that heterogenous microstructures can have on the mechanical properties of Sn-Bi solder joints.
Conclusions
Small Sb additions to eutectic Sn-Bi solder alloys increase ductility without decreasing alloy strength. Determining the mechanisms behind these beneficial mechanical attributes is important for better alloy development and to understand how to optimize these alloys for better drop-shock reliability. The first step in determining mechanisms is the quantitative characterization of how the microstructure changes with alloy composition. The presence of Sn dendrites in these alloys and the lack of SnSb IMC particles in the 0.5 wt% Sb alloys suggest the NIST Sn-Bi phase diagram is an accurate representation of the Sn-Bi system.
If Sb is still soluble in SnBi at 0.5 wt%, this means that the increase in ductility of SnBi must be primarily caused by solute-related effects such as solid solution strengthening or solute segregation to the phase boundaries rather than by precipitation hardening or precipitate effects. Exactly how these solutes increase the ductility still must be determined from deeper microstructure investigation including interrupted shear testing. Adding Sb at concentrations above 0.5 wt% results in SnSb precipitates, but the solid solution effects are still present. As Sb addition increases, the SnSb precipitates increase in size and increase in quantity. As previously stated, Sakuyama suggested that these larger SnSb particles are the reason why the ductility of SnBi decreased above 5 wt% Sb addition.3 Preventing these larger SnSb particles from forming seems to be critical for maintaining improved ductility from Sb addition to eutectic SnBi.
The next step in quantification will be to determine the volume fraction of Sn dendrites relative to eutectic and the spatial distribution of the SbSn IMC particles. To obtain the needed statistics, both will require analyzing substantially more solder joints and cross sections than done up to this point. These will be accompanied by nanoindentation testing to study the impacts of Sb and Ag addition on the strain-rate sensitivity and creep resistance of eutectic Sn-Bi alloys and may reveal information on the effects of microstructural heterogeneity. In addition, interrupted monotonic shear testing and fatigue testing will be used to study the microstructural evolution of these alloys after different amounts of deformation. These tests will also show the effects of Sb and Ag alloying on the ductility and strength of these alloys as a solder joint.
Initially Published in the SMTA Proceedings
|
