Formation and resettlement of (AuxNi1-x)Sn-4 in solder joints of ball-grid-array packages with the Au/Ni surface finish

Citation
Ce. Ho et al., Formation and resettlement of (AuxNi1-x)Sn-4 in solder joints of ball-grid-array packages with the Au/Ni surface finish, J ELEC MAT, 29(10), 2000, pp. 1175-1181
Citations number
10
Categorie Soggetti
Apllied Physucs/Condensed Matter/Materiales Science
Journal title
JOURNAL OF ELECTRONIC MATERIALS
ISSN journal
03615235 → ACNP
Volume
29
Issue
10
Year of publication
2000
Pages
1175 - 1181
Database
ISI
SICI code
0361-5235(200010)29:10<1175:FARO(I>2.0.ZU;2-R
Abstract
The interfacial reactions between eutectic PbSn solder and the solder ball pads with the Au/Ni surface finish were studied. Solder joints subjected to up to three repeated reflow-and-aging treatments were examined. For the re flow, the peak reflow temperature was 225 degrees C, and the reflow time wa s 115 s. Each aging process was performed at 160 degrees C for 500 h. After the first reflow, all the Au would disappear from the interface, and forme d many (AuxNi1-x)Sn-4 particles inside the solder joints. The value of x wa s between 0.99 and 0.75. In addition, there was a thin layer of Ni3Sn4 (1.4 mu m) at the interface. After one reflow and one subsequent aging, most of the (AuxNi1-x)Sn-4 would relocate from inside the solder joint to the inte rface, and the value of x for (AuxNi1-x)Sn-4 at the interface decreased to 0.45. This (AuxNi1-x)Sn-4 resettlement process repeated itself for addition al reflow-aging cycles. More reflow-aging treatments, however, made the mic rostructure of (Au0.45Ni0.55)Sn-4 at the interface become more non-planar. It was shown that gravitational effect was not the driving force for the re settlement of(AuxNi1-x)Sn-4. It is proposed that the driving force is for ( AuxNi1-x)Sn-4 to seek Ni at the interface so that it can become more Ni-ric h. In other words, the driving force is lowering the Gibbs energy of (AuxNi 1-x)Sn-4 by dissolving more Ni. A decomposition-diffusion mechanism is prop osed to explain what happened. Kinetic rationales for this rapid resettleme nt of (AuxNi1-x)Sn-4 at such a low temperature were also discussed.