A complete analysis of the manufacturing process can aid in troubleshooting solderability problems.

Did you ever work really hard to tackle a problem, only to find that the area you were focusing on had little or nothing to do with the problem in the first place? The Organic Solderability Preservative (OSP) process can fall victim to this type of narrow analysis. The question of “Which OSP is best?” is sometimes the wrong question. Often, we need to look outside the OSP chemical process tank. Before we can get to the correct answer, we need to expand our view of the entire process, from OSP coating through assembly, to be sure we are asking the right questions.

A 2006 Intel Corporation study1 that focused on OSPs cited only minor differentiation in performance among a number of widely used OSP processes. A key focus point of the study was on OSP’s well-known problem, hole-fill solderability and the differentiations in hole-fill yields that have been reported. What drove the differentiation? The two most significant sources of yield problems were deviations among PCB fabricators on the details of how the OSP coating was applied (differences in the boring, but critical, old surface finishing details relating to proper metal cleaning and activation) and poor choices of fluxing chemistry at the assembly side of the process.

Can we expand and provide some meaningful information on the two key findings above? Let’s start on the manufacturing side. Many of us already know that new families of high temperature OSPs have emerged to better cope with the stresses of lead-free reflow. Despite variations in the specific chemistry, all the popular OSP coating materials can be chemically categorized as azoles. Important differences in the specifics of those azoles may exist from supplier to supplier, but it turns out that all of them are sensitive to the state of the copper surface as it enters the OSP process tank. Specifically, these azoles love to initiate and to build dense coatings on incoming copper with liquid films at specific pHs. From this, swings in pH of incoming films can result in major shifts in OSP coating thickness and uniformity. I’m betting that few people realize that the pH range of supposedly “high quality” rinse water can vary from below a pH of 4 to somewhere around a pH of 8.

According to the report, Intel, “did not see significant performance differences between OPS chemistries... [but] did see significant variation between different PCB suppliers using the same chemistry... some suppliers were found to skip certain steps specified in the chemistry provider’s recipe.”

We don’t want to shift all the blame on the PCB manufacturers, but I can say that it is all too common for a chemical process to be wedged into a piece of process equipment that was originally built for another purpose. As an example, a process line missing a cleaning step would be prone to solder mask residues and potentially prone to assembly yield losses, so it’s important to pay attention to these details.

The second conclusion of the study relates to utilizing the proper fluxing chemistry in the assembly environment. There is some debate about what exactly happens to OSP coatings when they see extreme heat. From my experience and research, the evidence indicates that the coating is not lost from the surface, but it does physically change as it goes through more and more heat excursions. Our findings of several OSP finishes are that the coating hardens and compacts through these heat cycles. After heat exposure, OSP coatings are much harder to strip than when freshly applied (note that even after nitrogen reflow, they are harder to strip, so it is likely not a result of oxidation). Also, we found that the copper surface under the coating does oxidize to some degree. This certainly puts more pressure on the flux.

We certainly don’t have enough time to review all the details of flux chemistry, but we can review one key point of the study. One major conclusion involved identifying a well-designed flux chemistry that enables the flux to actively work at dissolving the OSP coating and copper oxide throughout the full duration of the lead-free wave solder heat cycle. Intel cautioned against fluxes which were thoroughly volatilized early in the preheat stage of the cycle. Fluxes with properly selected materials will volatilize at different stages of the process, keeping the flux active and working for a longer time. Remember the difficulty in stripping the OSP coating that has seen multiple surface mount heat cycles? Some OSP users call for lower OSP thicknesses for just this reason.

So, the point here is that upstream and downstream analysis of the manufacturing process is often required to properly review and to troubleshoot a final finish problem, particularly with finishes like OSP. One expert’s insignificant process variable may end up being the most significant factor in solving a processing problem. In the PCB business, it’s always best to expand your view to encompass the entire process. PCD&F

REFERENCE

2006 Lead Free Wave Soldered Board Assembly and Test Study Result, WG 2006 1011, George Arrigotti, Intel.

John Swanson is director of Final Finishes and Interconnect Technologies at MacDermid Inc; This email address is being protected from spambots. You need JavaScript enabled to view it..