Real Time with… SMTAI 2020: Reflections on the Technical Conference

Estimated reading time: 7 minutes

It was unavoidable. This year’s SMTA International Conference and Exposition went virtual! The organising team did a fine job in difficult circumstances to present an interactive show and technical conference. Having entered the virtual environment, online navigation was logical and straightforward. An all-inclusive searchable directory gave descriptions and contact information of the exhibiting companies, as well as an introduction to special events and awards and a listing of technical presentations.

The extensive conference programme included over 90 half-hour technical papers grouped into categories: advanced packaging technologies, flux solder and adhesive, harsh environment applications, lead-free and low-temperature soldering technologies, manufacturing excellence, PCEA topics, substrates, PCB technologies, and technical innovations. Each presentation was packaged as a pre-recorded entity, capable of being streamed on-demand, with the facility to pause and rewind as required; this was very helpful for checking details and taking a second look at points of interest.

Although there was no provision for live Q&A, each screen had a field into which questions could be typed for the presenter to respond to in due course. And while the event was scheduled for a nominal three days, the conference presentations would be accessible for three weeks afterward—a generous opportunity to attend many more sessions at leisure without the risk of suffering information overload from trying to cram them in all at once.

Indeed, there are certainly several further sessions I will endeavour to take in over the next couple of weeks, but I decided to check out a few in which I was particularly interested. My selection of papers had a genuine international flavour, including from the U.S., India, Germany, and Japan. I will work my way through the rest of the list and choose some more to review later.

Dr. Kunal Shah, LiloTree

Dr. Kunal Shah, president and chief scientist at LiloTree, discussed the effect of solderable finishes on high-frequency insertion loss and made a convincing case for the elimination of nickel as a barrier layer in the established electroless-nickel immersion gold (ENIG) finish. As a result of its magnetic properties and comparatively low conductivity, the loss associated with nickel increased with increasing signal frequency making it unsuitable for use in 5G, radio-frequency, and microwave applications.

Currently available alternatives include direct immersion gold (DIG) and electroless-palladium-based finishes, in combination with immersion gold (EPIG) or autocatalytic gold (EPAG), and relied on substantial thicknesses of precious metal and were less cost-effective. A proprietary nano-engineered passivation treatment for copper had been demonstrated to inhibit diffusion and restrict intermetallic growth, enabling the use of a significantly thinner immersion gold coating and consequent cost saving with no adverse effect on reliability.

Moreover, the treatment reduced the roughness of the copper surface and gave an additional improvement in signal integrity. Extensive testing of surface resistance, solderability, and solder joint durability after aging had been carried out by independent laboratories and demonstrated enhanced reliability, as well as reduced insertion loss.

Dr. Morgana Ribas, MacDermid Alpha Electronics Solutions

A fascinating lesson in solder metallurgy was delivered by Dr. Morgana Ribas, senior R&D manager at the India Research Centre of MacDermid Alpha Electronics Solutions. She discussed the evolution of low-temperature lead-free solder from simple eutectic tin-bismuth alloys of the early 2000s, through higher-reliability modifications to elegant present-day versions as packaging and assembly technologies progressed. She emphasised the importance of choosing the right combination of elements to achieve a balance between melting behaviour and mechanical properties.

Failure in surface-mount assemblies could result from thermal expansion mismatch, vibration, thermal shock, and mechanical shock. Small additions of minor alloying elements could dramatically influence properties, and she illustrated the effects of silver on the drop-shock and thermal cycling characteristics of eutectic tin-bismuth, and the effect of copper on its ductility.

Dr. Ribas explained how solid solution strengthening, diffusion modification, precipitate strengthening, and grain refinement could be combined to engineer the features of commercial solders. The technology drivers for low-temperature solders were reduced material cost, long-term reliability, reduced labour cost, reduced energy cost, and reduced substrate and component warpage.

With many practical examples, some interesting metallurgical observations, and laboratory results, she described the development of the fourth generation tin-bismuth alloy, compared its performance with earlier versions and with SAC305 as a reference, and demonstrated how multi-component variants could achieve lower processing temperatures without detriment to their thermo-mechanical performance.

Dr. Britta Schaftsteller, Atotech

Besides the quality of its technical content, the presentation of Dr. Britta Schaftsteller, global product manager for selective finishing in Atotech Germany, was particularly remarkable in that it came close to reality with the recording placing her in front of the screen as if she was speaking in a live lecture theatre. Her topic was ENIG corrosion: “When it gets critical, and how is the status in the PCB industry?”

Dr. Schaftsteller began with a brief review of the ENIG (electroless nickel immersion gold) process flow: surface preparation, activation, nickel deposition, and gold deposition. Because the gold immersion process was a displacement reaction, it involved the dissolution of a small amount of the nickel-phosphorus electroless deposit, which tended to occur preferentially at grain boundaries. Although the ENIG finish had been established in the market for several decades, there were still some concerns regarding its reliability in soldered joints.

She continued by listing a number of factors that could influence ENIG corrosion, illustrated with various examples. The formulation of the immersion gold electrolyte had a significant effect, and she demonstrated how this had evolved from early pure-immersion chemistries, where “spreader corrosion” was a typical effect.

A second generation, with some autocatalytic constituents, showed occasional instances of “single bite” corrosion. The latest generation, with a higher proportion of autocatalytic components, gave no detectable corrosion. When was corrosion a critical problem?

Dr. Schaftsteller showed examples of solder dewetting and aluminium wire-bond peel-off, caused by surface corrosion. How could corrosion be detected and categorised? She also described microscopic techniques, a standardised test panel, and a classification scheme that gave an industry baseline for evaluation of gold electrolytes by PCB fabricators.

Discussing the deposition mechanism of third-generation gold chemistries, Dr. Schaftsteller described a mixed reaction whereby an initial deposit formed by immersion acted as a catalytic surface for a subsequent reduction reaction. Using the systematic approach, no corrosion effects had been observed.

Dr. Tetsuya Sasamura, Uyemura

Although 5G promises data speeds of up to 100 gigabits per second and is set to be up to 100 times faster than 4G, its super-high operating frequencies have set challenges to providers of PCB substrates and finishes. A presentation reviewed above discussed low-loss alternatives to the ubiquitous ENIG. Dr. Tetsuya Sasamura, final finish researcher with Uyemura in Japan, presented a further option for consideration, with a discussion of the deposit properties of electroless gold/palladium/gold or IGEPIG—another acronym to add to a growing list).

Electroless nickel electroless palladium immersion gold (ENEPIG) had proved a very reliable solderable and wire-bondable finish, but lower nickel thickness was required to minimise signal loss at 5G frequencies, and a no-nickel finish would be preferable, especially on designs that had conductor spacing of 20 microns.

Dr. Sasamura compared insertion loss values for normal-nickel ENEPIG, thin-nickel ENEPIG, electroless palladium immersion gold (EPIG), direct immersion gold (DIG), and various layer thicknesses of IGEPIG. Then, he evaluated solder joint reliability for various solder alloys with IGEPIG. It had been observed that the thickness of the palladium layer in IGEPIG affected both insertion loss and solder joint reliability. As the palladium thickness increased, the proportion of palladium in the intermetallic increased, causing a reduction in solder joint reliability.

Also, because of the skin effect, insertion loss increased. The solder joint reliability of IGEPIG was comparable with that of ENEPIG for most solder alloys, increasing if nickel was present in the solder, whereas that of ENEPIG decreased when the solder contained copper. Additionally, IGEPIG showed better wire-bonding reliability than normal-nickel ENEPIG.

Conclusion

My initial experience of a virtual conference was very positive. The format has many benefits, particularly the on-demand streaming facility and information-gathering potential. I particularly liked the production style of the Atotech presentation: the illusion of a live conference session and not just a disembodied voice talking over a PowerPoint, or a thumbnail video of the speaker tucked away inconspicuously in the corner of the screen.

Obviously, what was missing was the community—the people you meet at live events, the professional and personal interactions, the informal discussions over a coffee, and the exchange of trade gossip and scandal over a beer later. But a big thank you to SMTA for all your efforts. It was an enjoyable and most informative experience.