We are considering the use of Immersion Silver for a PCB surface finish. We currently use Immersion Gold surface finish, which has served us well. However, recent discussions on the cracking of corner joints on large pin count BGA packages, with the suspected cause being the inherent weakness of the intermetalic gold / solder joint, has caused us some concern. Has anyone any experience using Immersion Silver on double-sided boards? This process uses an OSP coating to provide an acceptable shelf life for the boards, however the OSP is burned away during the first pass through an IR Oven. We would like to allow our suppliers flexibility in their SMT Assembly/ Planning processes, but are concerned about the maximum time exposure of the silver without its OSP protection between first and secondary side assembly. Should we be overly concerned about this issue? We would like to have a one to two-week window, if that is possible and maintain good reflow / wetting quality. Comments?
1 If you're talking AlphaLevel process (?): � Good solderability, as HASL, even no-cleans bite well. � Planarity, uniform deposit, no bridgin� of FP. � Doesn't store as well as NiAu ... LT 6 months sealed � Very sensitive to touch, more sensitive than OSP, if that�s possible!!! Finger "acid" breaks the OSP seal, then the silver tarnishes very quickly. � Doesn't like multipass, obviously. Experiment prior to double side process. Similar to issue that (White Tin PCB - Hand soldering - Mark Miller 14:25:47 01/25/2000) spoke about a week or ago on SMTnet regarding immersion tin. 2 If you're not talking AlphaLevel process (?): ask the base material supplier (not the fab) for electromigration studies carried out by reputable independent labs. 3 Corners of large BGAs see the greatest amount of flex during thermal cycles. So it�s reasonable that you�d see your cracking there. And cracking might be expected for various reasons, not the least of which is brittle connections, caused by tin/gold intermetallics, but I hesitate to go that way. I wonder: � What is the concentration of gold in the solder connection? � Is the gold in solution? Or is it hanging around the pads, that makes one gluteus maximus ugly (brittle) connection. � If a uniform amount of gold and solder are in all of the solder balls, why aren�t more balls cracking than just the corners? � Could the cracking be totally a function of the flexing of the BGA substrate? Could the temperature ramp rates be flexing the interposer so much that the corner balls would crack of the balls were made of conductive silly putty? � Where is the cracking occurring? Top er bottom of the balls? When we see corner ball cracking, it�s in the tops of voided-up balls.