|
|
|
|
|
Solder Joint Embrittlement Mechanisms, Solutions and Standards
Materials Tech |
|
Authored By:Mike Wolverton Raytheon Dallas, Texas USA SummaryThe change to lead-free solders in electronic assemblies created a need to replace tin-lead solderable termination finishes with materials such as pure tin or soft gold, on electronic components and substrates. Gold presented a risk of solder joint embrittlement, which could reduce the joint mechanical durability. Several case studies were run, because solder joint embrittlement prevention requires a clear understanding of the materials and mechanisms of embrittlement. We confirmed two known mechanisms and verified two other ways in which gold finishes can degrade solder joints. The known mechanisms are, 1) A gold layer dissolves from one side of a surface mount joint and precipitates AuSn4 compound onto the opposing termination and 2) The gold fully dissolves from a surface mount termination, but it results in an excessive gold weight percentage in the joint. The other ways are, 3) A manual soldering process temperature is high enough to dissolve some nickel along with the gold and 4) A slow dissolving, hard gold surface finish incompletely dissolves during plated-through-hole soldering and solid state diffusion forms an AuSn2 compound layer. Close-up and cross-sectional images with SEM/EDS compositional information are shown for each case. A table of solder and gold volumes, which produce 3.0 and 4.0 weight percent gold, is provided. The embrittlement problems and their reliability solutions are discussed (over twenty literature references). The data suggests how to improve gold plating requirements, for solder joint embrittlement prevention, in solder assembly industry standards. ConclusionsThe present paper describes four case studies of solder joint embrittlement, three with Sn63Pb37 and one with SnAg3.7 solder alloy. Solutions are given in each case. Applicability of the solutions to lead-free alloys and standards is addressed. In the first case, a gold-tin compound was shown to precipitate onto a surface mount lead that had no gold on it. The substrate had a thick gold that was supposed to be capped with nickel, but the perimeter of the thick gold had not been capped, causing the gold to dissolve profusely during soldering. The case was illustrative of the common embrittlement mechanism of gold dissolution followed by AuSn4 intermetallic compound formation and precipitation onto an interface, as well as into the solder joint bulk. A simple redesign of the nickel cap solved it. In the second case, a surface mount soldered connector pin and the substrate pad both had too much gold, and the solder volume was too small, resulting in too high of a gold weight percent. We measured the gold content in the solder joint by Scanning Electron Microscopy with Energy Dispersive Spectroscopy, showing that one does not have to always rely on calculations. A 32 run experiment optimized the gold removal processes, the solder volume and three temperature settings for the unique manufacturing equipment. Validation cross-sections showed success. While the first and second cases demonstrated commonly known mechanisms, the third and fourth cases provided some different information. In the third case, manual soldering of a surface mount connector was performed. Solder joint cracking occurred during electrical test. A multifunctional team ran a cause-effect investigation, and found four causes. The manual soldering temperature was too high, found by microstructural analysis. Gold and nickel embrittlement resulted in (Au0.45 Ni0.55)Sn4 intermetallic compound, with a morphology different from AuSn4. We reduced the soldering temperature with a better heat sinking setup. Also, we improved the flux and used a connector with thinner gold. In the fourth case, a hard gold surface finish on a pin caused a solder joint crack in a plated through hole application. Cross-sectional analysis revealed that in the fillet area, not all of the gold was dissolved from the pin. AuSn4 at first was attached to the residual gold layer. AuSn2 formed by diffusion, resulting in a duplex AuSn2/AuSn4 layer which broke from the gold. Gold hardeners have extremely slow dissolution rates in solder, compared with pure gold. The hardening of the gold slowed the gold dissolution rate, resulted in residual gold thickness after solder joint solidification, caused the AuSn2 formation, and was causal to the cracking. Hard gold needs complete dissolution. Regarding industry and standards opportunities, the case study findings and solutions are applicable to lead-free alloys. We demonstrated how gold and palladium weight percent limits can be converted to maximum finish thicknesses, aiding solderability and wire bondability in assembly, as well as solder joint reliability in the field. Initially Published in the IPC Proceedings |
|
Comments
|
|
|
|
|
|
|
|
|