The HIP epidemic proliferates, but the cure must not create new problems.
![Pb-Free Lessons Learned Pb-Free Lessons Learned](http://circuitsassembly.com/cms_bu/images/stories/column_icons/pbfreelessons.gif)
Earlier this year I reported on the head-in-pillow epidemic affecting many assembly lines running Pb-free reflow processes (“HOP-ping Mad,” July 2008). I postulated that a combination of circumstances was occurring simultaneously to create this plague: The higher surface-area-to-volume ratios of progressively smaller paste deposits and BGA balls create a higher proportion of exposed (readily oxidizable) exterior solder to (protected) interior solder, which accelerates flux consumption; reflow processes with extended pre-liquidus thermal exposures that also help to exhaust the flux; and Pb-free oxide films that are more difficult for fluxes to break through than SnPb ones. Since those initial observations, reports of HIP defects continue to increase, and not just in Pb-free soldering. My completely unscientific, statistically insignificant polls indicate the incidence of HIP in SnPb processes is also rising.
![Image Image](http://circuitsassembly.com/cms_bu/images/stories/ArticleImages/0812/0812pbfreefig1.jpg)
When we remove Pb-free alloy from the HIP equation, we’re basically left with flux exhaustion as the root cause. No-clean fluxes are designed to activate and deactivate under certain time-temperature conditions. If their thermal exposure limits are exceeded, their effectiveness wanes. Whether it’s from running too hot in the soak zone, taking too long to reach liquidus temperatures, or exposing too much surface area to heat and airflow – either from package warpage or questionable stencil design – spent flux is spent flux, and there’s no reviving it. The first line of defense in SnPb HIP should be profile adjustment to limit the thermal energy input, and in the majority of cases, it works. But what if it doesn’t?
Luckily, solder paste manufacturers have a solution at the ready – what I like to refer to as hybrid solder pastes. These products are a blend of flux designed for Pb-free solder paste combined with SnPb solder powder. When we bring together the more thermally tolerant (Pb-free) flux with the SnPb powder, voilà! A SnPb solder paste that really can take the heat, so to speak.
These hybrid pastes were introduced in 2007 to address mixed metals soldering, primarily for SAC 305/405 BGA balls on SnPb assemblies. The logic behind the blend is straightforward and with obvious advantages: Fluxes devised for Pb-free soldering would better facilitate wetting to Pb-free surfaces than those designed for SnPb alone, and these fluxes are sufficiently thermally robust to perform in the top quartile of the SnPb peak temperature range typically associated with mixed metals assembly. Henkel was one of the first to introduce this hybrid solder paste, and according to Dr. Brian Toleno, global director of technical services, the blends were devised specifically to provide increased activity over a larger thermal window. He explains that “the increased activity over a wider profile range permits assemblers to push SnPb’s traditional top-end temperature and time above liquidus limitations, thereby extending the conventional process of solder joint formation to accommodate the differences in the alloys.”
Since those first few formulations hit the market early last year, many suppliers have followed suit. Nearly every major player now offers a paste that blends SnPb solder powder with a flux designed for Pb-free. Although the original intent of these products was to facilitate mixed metals soldering, they do provide a seemingly handy solution for problems related to flux exhaustion. Toleno adds, “In the case of head-in-pillow, the increased activity is more effective at penetrating the oxide films, and can substantially increase the probability of forming good, reliable solder joints.” With many of last year’s mixed metals dilemmas now resolved, some suppliers are specifically marketing hybrid pastes as tools to help mitigate HIP defects.
When the solder paste supply base introduced these products, I cautioned the higher temperature flux might not be a foolproof drop-in lower temperature SnPb flux. Three performance attributes concerned me, and since then a fourth point of interest has emerged. The initial concerns centered on electrochemical reliability, voiding and intrusive reflow. The latest concern is the inclusion of halogenated materials in flux products.
As I described in June 2007, the two most notable risks of hybrid pastes include lower electrical reliability and increased voiding rates (“Lead Spread,” June 2007). A Pb-free flux designed to run in a hotter reflow process may not provide the desired electrical reliability when run in a SnPb process. Recall no-clean pastes are designed to activate and deactivate under certain ranges of thermal energy. Running a higher temperature paste may prevent it from burning out prematurely in a hot process, but if it runs on the cool side, it may not be fully deactivated in the reflow cycle, thereby posing a potential reliability threat in service. In the context of mixed-metals systems, the paste flux is exposed to peak temperatures above 220°C for at least 30 sec., relatively close to its design specification. If broadly deployed as a SnPb alternative, however, the probability it is exposed to smaller amounts of thermal energy in the reflow process is much higher. To err on the side of safety, I would suggest subjecting the product to SIR or electrochemical migration tests when processed under the coolest anticipated SnPb profile – with an adequate safety margin – before it is introduced into production.
Voiding concerns remain relatively the same whether the paste is used as a vehicle to solder mixed metals systems or to address HIP. For the most part, voids are created when the volatile materials in the flux portion of the paste cannot outgas before the metal melts. A flux designed for Pb-free processing can volatilize and vent until 217°C when used with SAC alloys, but its outgassing paths will close off at 183°C if used with SnPb solders. A faster/cooler profile may permit more gasses to be trapped than a slower/hotter profile, but a paste’s behavior with respect to voiding can be formulation-dependent and difficult to predict. The best thing an assembler can do is run the paste candidate at the fastest/coolest thermal profile they expect to use in production, measure the voiding rates and decide if they are acceptable.
The third concern was pin-in-paste processing. If using intrusive reflow processes, examine the ability of the paste overprints to pull back to the PTHs. We know different pastes have different sweet spots in the reflow window with respect to pull back, and users should make sure the candidate hybrid paste will support pin-in-paste without leaving random solder balls on the board. And again, we’re not just looking at the high end of the SnPb window like we were with mixed metals; we’re looking at the whole window if we want to qualify these products for general production use.
Finally, the newest worry: halogenated materials in the flux. When I mention halogenated materials, I mean non-ionic halogenated compounds, not the ionic halides we are accustomed to in the world of flux. (Halides make great flux activators because they are keen at reducing metal oxides, and our industry has long-established methods of identifying and communicating their presence in soldering fluxes.) Here, I’m referring to those compounds often used as flame-retardants in all types of electronics components, most notably printed circuit boards, molding compounds and connectors. They are sometimes used in small quantities in soldering chemistries to improve a flux’s survivability under extended thermal stresses. They can be found in both SnPb and Pb-free fluxes, but they are not necessarily required for successful soldering. Some fluxes have them; some don’t. Some have such a small amount that they squeak under the 1500 ppm threshold and are therefore considered halogen-free. Given the thermal sustainability demands put on Pb-free fluxes, they are slightly more likely to contain halogens than their SnPb counterparts. If the assembler has no concerns about halogenated materials, this point is entirely moot. But looking down the road at green initiatives and potential environmental legislation surrounding halogenated compounds, qualifying a new process chemistry that contains these materials may be a decision that could backfire in a few years’ time. And let’s not overlook the public relations implications: Who wants to be the target of a one-sided Internet video that opines a product is not “green enough,” but neglects to mention the often irreplaceable safety factor the halogens provide?
The lesson to be learned with hybrid solder pastes is to ask a few key questions prior to making the switch:
Can the Pb-free flux provide the necessary electrochemical reliability when processed on the relatively faster, cooler profiles associated with SnPb reflow?
Will the new SnPb paste create more voids than the incumbent one?
Will the solder coalesce properly in the intrusive reflow process?
Does the product contain any materials that are politically controversial or under consideration for near-term elimination?
Hybrid pastes now appear to address two widespread issues for assemblers: improving BGA soldering in both mixed metals and SnPb systems. They show promise for improving yield and reliability, but should not be implemented without due consideration and appropriate testing. As engineers, we must make sure that when we fix one problem, we don’t create another. In other words, look before we leap.
Chrys Shea has 20 years’ experience in electronics manufacturing and is founder of Shea Engineering; chrys@sheaengineering.com.