I was chatting recently with Indium Corporation's Mike Fenner about Sn/Cu alloys for solder paste. Mike has many years of experience and I value his opinions greatly. Here is a summary of his thoughts.
Mike, there has been some buzz in the industry about using SnCu alloys for solder paste. We all know that some of these alloys have been quite successful in wave soldering. But, what about the alloys being used in a solder paste for reflow?
1) Let us recall that, going back over 50 years, the electronics industry has always had a need for a reliable higher temperature alloy (compared to eutectic tin/lead.) Interestingly, Sn/Ag was chosen over Sn/Cu due to its superior mechanical and physical properties. Users were, and are, the avionics, military, automotive, aerospace, and medical industries.
Sn/Ag has always been more expensive than Sn/Cu. So, the question, "Can we use Ag-free alloys like Sn/5Sb, Sn/0.X Cu?" has been asked since the alloy was developed. The conclusion is always no.
2) When our industry moved to compulsory Pb-free, all the great and the good of the industry reevaluated the potential alloy choices. These suppliers, users, universities, and industry associations did not choose Sn/Cu, but Sn/Ag with a small amount of Cu added; what we now refer to as SAC alloys. Since Sn/Ag produces (ultimately) a SAC alloy when soldered to Cu, this is, essentially, the same alloy for the PCB people.
Nowhere does it say Sn/Cu is better than Sn/Ag/Cu. Actually the opposite.
4) The proponents of Sn/Cu solder paste suggest that it can be reflowed with the same thermal profile as SAC alloys. This is simply not the case; SAC melts at about 217°C, whereas Sn/Cu melts at 227°C. In addition, although there is much data for Sn/Cu wave soldered (through-hole) joints, there is no long term test data for Sn/Cu reflowed (SMT) solder joints.
5) The principal advantage of the popular Sn/Cu alloy in wave soldering is that it does not dissolve the solder pot, impellors, and other equipment. This is due to the addition of trace amounts of Ni to the alloy. Ge is also typically added as an anti-oxidant, to reduce drossing. These trace elemental additions also produce grain modifications, which is why the joints look relatively bright compared SAC alloys. These are user benefits for the wave process, but all are irrelevant in reflow as they are not in-service benefits to the reflow user.
6) The stress on a through-hole solder joint is relatively low. This is in complete contrast to the very small, unreinforced, single-sided SMT joint. A through-hole solder joint is much larger and stronger than an SMT joint. The joint size and structure is completely different in these two processes. Additionally, in wave soldering, the component attachment is made like a rivet, with the component lead going right though the board and out the other side, supported on each side by a solder joint. In essence, there are two "anchors" per termination. Any reliability data on Sn/Cu has been developed exclusively from though-hole soldering. So, for the same thermal cycling regime, of course a through-hole joint will be intrinsically stronger, due to the structural mechanics. The proper response to any claims about long-term reliability would be to assure that you are comparing through-hole to through-hole and SMT to SMT.
7) When customers evaluate solder pastes, which criteria determine their final selection? Most often the answer is some combination of the following performance realms: material storage and handling, printing, component retention, slump, response-to-pause, wetting, voiding, head-in-pillow, etc. The flux/vehicle formulation is predominantly responsible for all of these characteristics. Performance in these realms has very little to do with the solder alloy. Additionally, the principle cost of most solder pastes is the manufacturing process. The alloy is typically a secondary cost (although this might not be the case with a high percentage of precious metal).
Summary SMT is not wave soldering. If you are looking for cost savings in your SMT reflow process, implementing a Sn/Cu solder paste may be a very expensive way of achieving those results.
Dr. Ron: Thanks Mike for your insight!