By Keith Gurnett and Tom Adams
First of a two-part series
Lead-free solders, which were relative novelties a few years ago, are now widely understood, although not always in a positive sense. The most commonly used lead-free solders are known to be more brittle than lead-tin solders, and there have been suggestions recently that there may be large numbers of failures related to lead-free solders in the next several years, both in civilian and non-civilian applications.
Prof. Nihal Sinnadurai is a well-known industry expert in reliability assurance methodology. He advises manufacturers on strategies for achieving high reliability, and also appears frequently as an expert witness in cases where the absence of reliability may have caused harm. He has looked closely into the characteristics, behavior, and manufacturability of lead-free solders, and here offers his best answers to some of the most profound questions: Will there be large numbers of lead-free related failures? How can these failures be avoided? Can lead-free solders in the long run achieve the reliability of leaded solders?
Despite mandates for the use of lead-free solders in Europe and in Asia, Sinnadurai recognizes that many U.S. manufacturers would prefer not to use lead-free solders at all. He thinks, though, that the nearly universal replacement of leaded solders with lead-free solders is inevitable, that components that military and aerospace users procure from commercial sources will virtually all be available only in lead-free forms, and that manufacturers would be better to find ways to work with lead-free solders.
What he is suggesting, however, is not a matter of accepting an inferior technology: he has looked closely at lead-free solders themselves and has outlined methodologies for making them as reliable as tin-lead solders. Lower reliability is not an inherent characteristic of lead-free solders. Tin-silver-copper SAC solders, which are the most widely used replacements for tin-lead solder, have, for example, better thermal cycling reliability than tin-lead.
A similar evolution
“Almost every comment that one would make about lead-free, you would have made about lead-based solders when lead-based solders were going through the evolution and rebirth into surface-mounting technology,” Sinnadurai observes. He recalls a key speaker at a conference in Europe during the early days of surface-mount technology who proclaimed, “Surface-mounting technology will never happen. It is always going to be ceramic technology.”
Even under European RoHS, military and aerospace applications have many exemptions from the use of lead-free solders, he notes, but the same pressure will be exerted in Europe as in the U.S.: increasing numbers of components will be available only in lead-free versions because the largest markets for components are commercial markets, and because component makers generally do not find it worthwhile to keep turning out two versions of the same component. Tin-lead versions of an existing component, if available at all, will become more and more expensive. A typical example of this requirement is the Lockheed C-130 turboprop aircraft, which first flew in 1956 and is still flying. More than 2,500 were built and still require replacement parts, some of which can only be replaced by reworking existing units.
Likelihood of significant failures
Sinnadurai says he believes that the likelihood of significant equipment failures attributable to “lead-free” assembly is quite high over the next several years, and that these failures will happen in all types of applications—military, aerospace, telecommunications, medical, and consumer. This will be primarily due to the high probability of mixed lead and lead-free arising from non-declared changes of component terminations without proper change notification by component suppliers. Eventually, he predicts, lead-free assembly will attain the same long-term reliability that tin-lead solders now have. But he anticipates that there will be a period of 10 to 15 years during which manufacturers will go through the learning curve of materials and process improvements to give lead-free solders that high level of reliability.
In theory, one would think that assembled electronics systems and their associated reliability statistics could be neatly divided between tin-lead and lead free. Reliability studies that have been done in the past have been based on the assumption that a particular board will contain only pure lead-free solder, Sinnadurai says, but the reality today is that a board is very likely to harbor a mixture of lead-free and tin-lead solders.
“One of the problems the world is going to face...is that the assemblies, and the components that one uses in the assemblies, will not necessarily be pure lead-free,” Sinnadurai explains. “They will almost certainly be ‘hybridized’ (mixed lead and lead-free) because they will not be cleanly, purely lead-free. Component suppliers switch from lead-based to lead-free without submitting change notices and without changing the part numbers. And they are completely untraceable as to which is lead-free and which is lead-based, without routine incoming analysis of termination materials. That was one of the problems highlighted by assembly houses when I audited them.”
Failures on hybridized boards
Many failures on “hybridized” (i.e. lead and lead-free) boards will occur simply because of thermal differences—a component whose leads are finished with lead-free solder, for example, being reflowed at the tin-lead temperature of 230 degrees Celsius. “Then of course you will not have adequate melting and wetting, or you may have melting on one side and not the other,” Sinnadurai says. ”Such joints will not be inspectable when concealed under surface-mount components.”
Failures on hybridized boards may also occur because of differences in other materials, such as the molding compound: lead-free components use molding compounds that will survive 260 C, a temperature at which the molding compound on a tin-lead component might fail.
A third site for failures is the joint. “Amongst the things that contribute to failures in joints are the growth of intermetallics, embrittlement of the joint, and then eventual fracture of the joint,” Sinnadurai says. “Now, growth of intermetallics and embrittlement stem from inadequate cool-down process in the reflow. For example, if you do a very, very long ramped cool-down process, almost certainly you will convert the material to being more brittle. So you need to leave it more amorphous, and therefore less rigid, less brittle, and less likely to fail.” Less rigid, less brittle solder joints are critical, for example, aboard the AH-64 Apache attack helicopter, where normal shock and vibration become extreme during the launching of its enormous fire power.
The culture of cheapness
The equipment failures that Sinnadurai expects to occur in future years resulting from poor qualification practices will include conspicuous failures that affect large numbers of people—failures of electrical transmission systems, for example, or of operations networks at large airports. The fundamental reason behind these failures is the worldwide preoccupation with cheap solutions and inadequate reliability assurance. “I believe that we will suffer the penalty for wanting everything to be cheap,” Sinnadurai adds. “And the world will pay the price over the next 5 to 10 years if we continue to do things cheaply.” Along with cheap, he warns, the buyer of a system may get nasty surprises.
Today, Sinnadurai points out, manufacturers are taking a lot of shortcuts to avoid paying higher prices and to avoid taking the time to do the job properly. “That same culture has now pervaded the high-reliability business,” he explains. “And what is happening now is that we are putting into networks shoddy products, or unproven products. [This is] the tip of the iceberg and within the next five years we will see network crashes.”
Researchers, Sinnadurai says, report specific research aspects with lead-free solders—how they’ve discovered a nuance that applies to this or that lead-free material. These tend to be the papers presented at technical conferences. “However, when it comes to practical solutions in production the requirement is for whole product solutions, real situations on the factory floor, variability from one manufacturing site to another and solutions that are reproducible many times—then, the major problem is that the theoretical nuances do not translate into real production where the real practical problems arise.”
A key to cheapness is outsourcing manufacturing to low labor cost countries, but this works only to a point, Sinnadurai says. “If they want to continue to cut prices, they will continue to erode the controls that are needed to make products properly.
OEMs retain liability
“An equipment manufacturer may choose to outsource everything,” Sinnadurai adds, “but the liability remains at home. The responsibility for delivering reliable products to the customer remains with the original equipment manufacturer. The liability cannot be outsourced to a cheap assembly house but with the company that puts their badge on the product.”
There are many ways in which outsourcing can go wrong. Sinnadurai cited an OEM who outsources its assembly to low-cost assemblers in Asia and stated that their strategy is to change the assembly house every two years, and seek a cost reduction of 5 percent from each new assembly house. “I said to him, ‘Well, you can only do that maybe twice, before other assembly houses get wise to your intentions and will not then deal with you. Those who will do business with you will simply deliver garbage, because they will take many shortcuts to achieve your constant cost cutting.’”
The problem addressed
The “problem” with lead-free solders, in Sinnadurai’s view, is that the solders are not inherently prone to failures, and that they will become as reliable as tin-lead solders. The advantage that tin-lead solder has is half a century of development.
For military and aerospace manufacturers, as for the rest of the world, the trick is to bring lead-free solders up to speed, to do the work that is needed to characterize them so extensively that high reliability is more easily achieved. While this is true in the broad sense, engineers who need to achieve high reliability with lead-free solders cannot wait for this gradual evolution to take place.
In Part 2 of this article, Sinnadurai offers four strategies to achieve lead-free reliability now.
