Originally Published MEM Fall 2009
Originally Published MEM Fall 2009
Damage-Free Alloy Conversion for Medical Components
As suppliers rid themselves of traditional lead solders, how do medical device manufacturers cope with the changes? The answer is very carefully.
(click to enlarge) Sonoscan acoustic images showing delaminations (red) on lead fingers, and type of damage consistent with improper alloy conversion methods.
The introduction of the RoHS (Restriction of Hazardous Substances) Directive by the European Union (EU) on July 1, 2006, had a significant effect on electronics. It meant that traditional tin-lead (SnPb) solder could no longer be used in most segments of electrical and electronics manufacturing. However, some segments were given exemptions and thus could continue to use SnPb. The medical device and monitoring and control equipment segments are still able to use SnPb.
These exemptions were granted because it was recognized that the systems involved were typically employed in critical situations. There were no long-term data on the robustness of Pb-free solders and thus no way to judge the risk of sudden electrical failure that could result from the use of Pb-free solders. The exemptions are not permanent. They expire, depending on the industry subsegment involved, in 2012 or 2018. The exemptions may, of course, be extended beyond those dates.
The RoHS Directive applied to items manufactured in, or imported into, the member states of the EU. But the supply and sales networks of the global electronics industry are so broadly intertwined that limiting the use of Pb in the EU had global effect.
For a time, the manufacturers of electronic components sold Pb-free components to EU assemblers and SnPb components to assemblers in other areas—with the expectation that the second group of assemblers would not export products into the EU.
During the same time, some component manufacturers gave new catalog numbers to the Pb-free components, but some did not. Component users, therefore, had no way to know whether a particular lot was Pb-free. Such uncertainty faded because component manufacturers gradually stopped making SnPb components and turned instead to making only Pb-free components. Today very few SnPb components are still available, and it is hard to tell whether those that are available are from current manufacture or are from stocks made before RoHS.
(click to enlarge) A fixture holds 14 plastic IC packages about to be dipped into the SnPb bath.
Photo courtesy of CORFIN INC.
A result of RoHS is that makers of medical devices no longer have sources for the SnPb-coated components that they are permitted to use. A medical manufacturer that requires component X can still buy it, but the terminations on the component are not coated with the SnPb that would be compatible with the manufacturer’s processes. Instead, a manufacturer purchases the needed component with whatever coating is available (pure tin or an alternative finish, such as Pb-free). Then the manufacturer removes the unsuitable finish on the terminations and applies the SnPb finish required by the processes.
Manufacturers of medical electronics could avoid this complication by switching to components that accommodate pure tin finish. This would also require a redesign of products so that they can be exposed to higher temperatures typically needed for Pb-free solders. However, there are additional difficulties a manufacturer would have to overcome, such as higher-temperature reflow and the risk of sudden equipment failure from tin whisker growth.
Tin whiskers have received much publicity since the introduction of the RoHS Directive in 2006, but they have been a recognized problem since 1951. In March 1986, tin whiskers were the subject of a warning from FDA’s Office of Regulatory Affairs. The warning described tin whiskering as “a little-known phenomenon,” and gave this profile:
Tin whiskers are metal filaments [that] grow from tin. They are extremely thin, 1–2 µm typically, and grow as straight, kinked, or spiraled single crystals of tin. They can reach a length of 9 mm (3⁄8 in.) and carry 10 mA of current before burning up. The electrical resistance of a tin whisker 3 mm (1⁄8 in.) long is about 50 Ω. Because of their current-carrying ability and low electrical resistance, whiskers are a threat to electronic circuits.
Tin whiskers grow only from pure tin—not from tin alloys such as SnPb.
Because of these risks, medical electronic equipment makers have chosen to continue to use components with SnPb coating on the terminations for many of their products. To do so, they purchase components with pure tin on the terminations and perform alloy conversion to replace it with SnPb. In doing so, they avoid not only the risk of tin whisker growth, but also the hazards associated with the required higher-processing temperature for Pb-free solders.
Not So Simple
To remove pure tin coating on terminations, a component is dipped into a flowing bath of SnPb. The pure tin is washed away, and the SnPb replaces it. The conversion process sounds like a task that could be performed manually, but this would be an oversimplification. The temperature of the wash must be rigorously controlled, as must the dwell time and the immersion depth. Without such controls, alloy conversion can result in short-term or long-term failure of the component. If the component is overheated, it may fail immediately or, even worse, suffer internal damage that will cause an eventual field failure. If pure tin is left on a termination—usually close to the body of the part—tin whiskers can develop and contribute to part failure.
To avoid such problems, successful alloy conversion involves very precise control of three factors: the depth to which the part is immersed, the temperature of the SnPb, and the dwell time.
Solder Dip Standards
(click to enlarge) Two JEDEC-style trays of plastic IC packages are imaged acoustically by the scanning transducer at center.
Photo courtesy of SONOSCAN INC.
Photo courtesy of CORFIN INC.
Tech America (formerly The Information Technology Association of America) recently published the GEIA-STD-0006 standard, “Requirements for Using Solder Dip to Replace the Finish on Electronic Part Pieces.”
Implementation of STD-0006 to avoid damage to components is voluntary. The standard can be employed, for example, by an end-user that has procured parts for which the termination finish needs to be changed without causing damage. A distributor that receives a customer purchase order specifying that the termination finish needs to be converted from Pb-free to SnPb can also use it.
The standard provides guidelines about the key parameters that need attention during alloy conversion. For example, it requires that robotic equipment be capable of controlling the dwell time, in the preheat leading up to dipping and during the dipping process, to within 0.1 second. It also specifies that the immersion depth of a part in the solder be controlled to within 0.1 mm. The standard spells out the requirements for the dry baking of components, the use of flux, and the requirements after dipping for cool down, drying, cleaning, and inspection.
The standard requires that parts be imaged by acoustic microscopy both before and after the conversion process. For accept-reject criteria, the standard refers to Section 6 of J-STD-020D, “Moisture/Reflow Sensitivity Classification for Nonhermetic Solid State Surface Mount Devices,” but adds items specific to hot-solder dipping. For example, delamination in a die attach is permitted to expand by no more than 10% of the total die attach area as a result of hot-solder dipping. Exceeding this degree of expansion might be acceptable for some commercial products, but it would typically be a cause for rejection in medical devices.
Success of alloy conversion depends on procedures carried out before and after the hot-solder dipping process. For the past few years, the alloy conversion process has been completely automated, with robotic equipment employed to achieve the needed precision. One supplier converts tens of thousands of Pb-free components per day. To ensure that conversion processes do not damage the components, acoustic microscopes are employed to perform nondestructive acoustic imaging to inspect the internal features of the part.
Acoustic microscopes make nondestructive images by pulsing ultrasound from a moving transducer into the part. The ultrasound is reflected by material interfaces inside the part—an interface between the mold compound and the lead frame, for example. The reflections, or echoes, that have the highest amplitude come from gap-type features such as delaminations, voids, and cracks. The transducer scans the part and performs its pulse-echo function several thousand times per second. There may be thousands to millions of locations from which a return echo is collected. Each of these echoes provides the data for one pixel in the acoustic image of the interior of the part.
Any internal damage that occurs during alloy conversion is likely to be located near the edge of the part, where the terminations extend from the body. The damage might, for example, consist of the delamination of the mold compound from the lead frame. Or it might take some other form, such as a crack in the mold compound.
A troubling fact is that such internal damage does not usually result in immediate electrical failure. Unless the damage is relatively massive, the affected part may pass the assembler’s electrical tests. In service, however, the delamination is likely to grow. It may open a pathway from the exterior to the die, thus permitting moisture and contaminants to reach the die and the interconnections on the die. Acoustic microscopy was used in defining the details of STD-0006 because it nondestructively images and analyzes the many types of internal damage that, although initially innocuous, are capable of causing eventual electrical failure.
An acoustic image (shown on page 36) depicts one type of damage that can occur if the alloy conversion process is not sufficiently precise. It shows the interior of four plastic-encapsulated devices. The white and gray features are (from center out) the center, the die, the die paddle, and the lead fingers. The red areas mark where the return echoes have the highest amplitude (i.e. delamination points). In this example the lead finger is delaminated from the epoxy. The delaminations reach the outside surface of the part, permitting moisture and contaminants to travel some distance to the die without further expansion. This is one type of damage that can occur if the alloy conversion process is not sufficiently precise.
The exact parameters (solder bath, the dwell time, etc.) for a specific part are determined by referring to the manufacturer’s specifications.
A typical example would be a plastic-encapsulated integrated circuit (IC). When the part first arrives at an alloy converter service, it is visually inspected. If there is any question about its exposure to humidity, the component is baked according to J-STD-033. Next, vacuum grippers or titanium tweezers pick it up and move it to a tank holding liquid flux. A pump keeps the flux in constant motion. As flux is consumed in the process, or as the thinner (ultrafiltered deionized water) evaporates, new flux or thinner is pumped in to maintain the activity level and the specific gravity value.
After the part’s terminations are dipped in the flux, the excess flux is removed by an air blast. Next, a forced hot-air preheater activates the flux and dehydrates it.
The next step is to introduce the part to the dynamic SnPb soldering wave. There are 300 lb of Sn63Pb37 mixture solder in a pot, which is continually pumped upward under a blanket of nitrogen. The plastic-encapsulated IC is held at the proper angle and its terminations are immersed in the SnPb wave for the proper duration. The dipping angle and immersion time are determined, respectively, by the part geometry and its thermal sensitivity.
(click to enlarge) Figure 1. Robotically controlled immersion depth avoids damage to components.
Photo courtesy of CORFIN INC.
The goal here is to immerse the entire length of the terminations, and to just make contact with the edge of the body without damaging the part. Robotic control ensures very tight tolerance on the dipping depth as well as the temperature and the dwell. This level of control makes it possible to avoid damage to the body of the component while flushing off the alloy from the full length of the terminations. If any of the pure tin remains on the termination, tin whiskering is still a risk. If the part is immersed too deeply, internal thermal damage may occur. The diagram in Figure 1 shows desirable and undesirable dipping depths.
The tin that is flushed off of the terminations mixes in with the SnPb alloy in the solder pot. However, because tin is depleted more rapidly than lead in the solder pot, it does not alter the Sn63Pb37 mixture. In fact, even though pure tin is being added as the original terminations are removed, it is actually necessary to add more pure tin to the pot periodically to maintain the balance.
After dipping, a cleaning process that uses an ultrafiltered hot water wash is performed. The temperature of the water is lower than the temperature of the solder. In the author’s experience, it is best to allow a pause of only 5–10 seconds between the two steps—just enough time for the part to cool down to the temperature that the part’s manufacturer suggests for the hot-water wash. STD-0006 requires only that parts coming out of hot-solder dipping be washed within one hour. The problem is that the flux can solidify in that time and become much harder to remove. Manual or semiautomated alloy conversion processes tend to encounter this problem because they allow parts to accumulate so they can all be washed at once.
After washing, parts are dried by an air blast. Next, an acoustic microscope images some or all of the parts. Acoustic imaging provides nondestructive assurance of quality and ensures that the alloy conversion process has not compromised the parts.
Many parts used for medical devices, both active and passive, are initially manufactured with pure tin coating on the terminations. Occasionally parts that do have an SnPb coating need to be replaced because the coating has degraded during storage in an inappropriate environment. In addition, Pb-free coatings such as SAC (tin-silver-copper) or tin-bismuth are sometimes used.
The two most frequent items that require solders are pacemakers and defibrillators. Failure of these implantables could result in significant pain or death. As such, the devices require high reliability in the conversion process.
A bedside monitoring device may not have the same level of risk as an implantable device, but malfunctions caused by electrical failure can be just as severe. Some bedside devices dispense medication; a failure in such a device could be catastrophic for the patient.
The RoHS Directive is a good example of how efforts to reduce lead have had unintended consequences for medical device and other mission-critical applications. Medical device makers must now take an extra step to make terminations of device electronics compatible with their processes. In doing so, OEMs must ensure that the conversion process is handled with expertise and that service providers have appropriate process checks in place to ensure quality.
The author would to acknowledge the contributions of Don Tyler of Corfin Industries and of Steve Martell of Sonoscan Inc. in preparing this article.
Tom Adams is a consultant for Sonoscan Inc. (Elk Grove Village, IL). He can be reached at firstname.lastname@example.org. ■
Copyright ©2009 Medical Electronics Manufacturing