2020 EIPC Winter Conference, Day 1
In mid-February, there were two major attractions in the Blijdorp area of Rotterdam, the main port city in the Dutch province of South Holland—the Rotterdam Zoo and the EIPC Winter Conference. The Rotterdam Zoo is open every day of the year, but the premier event of the season was the 2020 EIPC Winter Conference on February 13–14, which attracted around 90 delegates from a dozen European countries—as well as a few from North America—to an outstanding learning and networking experience for members of the PCB community. The programme of more than 20 presentations and panel discussions, together with a table-top and poster exhibition, covered the theme “The Needs for Next-Generation Electronic Devices and Changes in Fabrication Solutions for PCBs, PCBAs, Materials, and Technologies.”
EIPC president Alun Morgan welcomed all to the conference and gratefully acknowledged the support of the sponsors. He invited members to adopt the revised statutes of the association, and these were received very positively. Morgan also took the opportunity to remind everyone that the ECWC15 World Convention would be held at the end of November 2020 in Hong Kong and Shenzhen and that there was scope for additional papers to be submitted.

“It’s 2020. What shall I talk about?” he continued. He reviewed predictions that had been made 10 years previously and some of the scientific discoveries and medical advances made during the last decade. It had been said that there would be no need for futurists to predict the future because progress in AI would make them redundant.
Time-honoured as the opening presentation was Walt Custer’s eagerly-awaited Business Outlook on the global electronics industry, with emphasis on Europe. Unfortunately, Walt was unable to travel to the conference, but he sent his best regards and delegated the delivery to Alun Morgan, who did a fine job as his deputy.

Although this edition of Custer’s business outlook had been compiled before coronavirus had distorted the 2020 situation, the fact remained that global manufacturing growth had reached bottom based on data from purchasing managers’ indices (PMIs). Tariffs, trade disputes, and the U.K.’s withdrawal from the European Union were key issues, and geopolitical concerns remained very significant. However, Custer’s observation that most sectors of the world electronic supply chain were now expanding slowly had to be qualified in the context of quarantine restrictions in China.
Custer considered the PMIs to be “useful but sobering leading indicators,” and figures for Europe had shown a continuing contraction toward the end of 2019—the only notable area of growth being in medical electronics. Production in the European automotive industry was down because of uncertainty about the future of diesel power, and mil-aero revenues were flat, as were volume consumer-electronics markets. However, there were substantial growth forecasts for 5G handsets and infrastructure.
World PCB production had been almost static for the last three years. Based on 2018 figures, the $74.5 billion total split by geographical area was China 54% and Taiwan and South Korea 10% each. Europe only represented 3.1% against North America’s 4.2%. The number of PCB manufacturers in Europe had declined to 202, with the majority of the revenue being generated in the German-speaking countries. Michael Gasch had predicted that Europe would close 2019 with a loss of 10% against 2018. “Hope for the best, prepare for the worst.”
Hans Friedrichkeit believed that there was a glimmer of hope in the German industry because the downturn in incoming orders had slowed, and the automotive industry was likely to stabilise. “It looks as if the bottom of the economic valley of tears has been passed.”
Custer’s opinion was that the world market had reached the bottom and was beginning to improve, although the first quarter of 2020 could be significantly influenced by the consequences of coronavirus. Trade disputes had impacted long-standing regional alliances and U.S.-driven impulsive market actions remained as problems. He believed that electronic assembly might shift globally to countries less affected by tariffs and that many good new electronic products were on the horizon. In particular, 5G offered exciting opportunities for the next decade.
The following presentation was Dr. Hayao Nakahara’s outlook on PCBs in automotive electronics. Dr. Nakahara apologised that he was unable to attend in person and asked EIPC Technical Director Tarja Rapala-Virtanen to deliver the presentation on his behalf.
There was a continuing steady 5–6% annual increase in PCB usage as a consequence of the increasing functionality of advanced driver-assistance systems (ADAS). The growth in electric-motor-driven cars was slower than had been predicted, and despite the hype about autonomous vehicles, it was unlikely that completely driverless cars (Level 5) would be generally available before 2030. “How can Level-5 cars make decisions in the streets of Saigon, Hanoi, and Jakarta where millions of motorbikes cut in front of you all the time?”
Sales of electric vehicles were still constrained by the price and life expectancy of lithium batteries. The typical cost in 2018 was $200 per kilowatt-hour. It was forecast that this could be reduced to less than $100 by 2023, which would certainly stimulate the market. The world sales of new light vehicles were 90.3 million units, but the world’s largest market, China, continued to shrink as a consequence of reduced tax rebates, a general economic downturn, and increasing sales of used cars.
Reliability of automotive electronics was a fundamental consideration, with 200–300 sensors per car, and security remained an important issue, particularly in the context of connected vehicle technologies Dr. Nakahara estimated the automotive share of world PCB output to be 11% in 2109. The value of PCBs per car ranged from $30–40 at the low end to $100–150 at the high end, averaging $50–70 per car, and it had been estimated that the electronics content of a car might rise to 50% of its total value. “To replace bumpers and headlamps will not be cheap anymore in the future.”
Mustafa Özkök, global product manager with Atotech, discussed the manufacturing challenges that 5G and millimetre-wave technology imposed upon the PCB industry, in terms of materials, processes, and design considerations. He explained that the main driver for 5G was the massive forecast increase in the number of connected devices and the quantity of mobile data. For example, an autonomous car was expected to create 4,000 GB per day. Millimetre-wave would drive new hardware, and the wireless infrastructure would need the enormous geographic density of small cells in order to achieve connectivity of everything. The move to 5G would enable high-frequency, high-speed data transfer, big volume data traffic, and high-density connection.
Investment in 5G infrastructure was forecast to exceed $300 billion by 2025 with a large proportion related to upgrades of data centres and public clouds, which would demand substantial growth in the production of servers, routers, and switches. New-generation substrate materials would be required for PCBs and ICs, and designs would regard signal integrity, impedance and capacitance, thermal management, and EMI shielding as critical issues.
Dielectrics formulated for high-speed, high-frequency operation, mixed dielectric builds, and advanced adhesion promoters would all contribute to minimising signal attenuation. Advances in metallisation and plating processes could offer compatibility with hybrid builds and give uniformity of copper distribution. Final finishes designed to minimise signal loss and withstand aggressive environments and extended storage would be required. Many of these material attributes already existed or were in development. Thermal management and new options for EMI shielding were also important considerations.
The session on inkjet and other coating technologies was moderated by EIPC board member Emma Hudson. Her first speaker was Dr. Andreas Albrecht from Cicor Group in Switzerland, with an enlightening presentation entitled “Printed Electronics: Pushing the Limits.” Discussing the principles of fabricating printed electronics devices, Dr. Albrecht described how a functional material—which might be a metal, a conductive polymer, a dielectric, or a sensor material—was converted from the bulk state to nanoparticles dispersed in a solvent mixture or polymer lacquer to form a printable functional ink. He stressed that Cicor did not make inks themselves, but were happy to cooperate with several specialist ink manufacturers and to evaluate and select products to suit their own specific requirements.
Most delegates were familiar with inkjet printing techniques, but aerosol jet printing was emerging as an alternative contactless direct-write method of producing fine features on a wide range of substrates, which could be used for creating conductors, active and passive components, actuators and sensors.
Dr. Albrecht explained the basic procedure. Ultrasound was used to break the ink into droplets, the size of which was governed by ultrasonic power and temperature, and an add-back of solvent ensured stability. A deposition head focused a jet of droplets, carried as an aerosol in a stream of nitrogen, onto the workpiece at a separation between 1–5 mm, as a 10-micron spot. The workpiece was mounted on a stage capable of movement in up to five axes and heated to dry the ink. The deposition head could itself be programmed for free-form writing if required. Pneumatic atomisation was a more complicated alternative to ultrasonic.
Compared with inkjet printing, aerosol jet printing offered some technical advantages: higher resolution, higher printing thickness, and a larger material portfolio. Because of the large and variable stand-off distance, it was capable of printing increasingly complex devices on three-dimensional surfaces. Dr. Albrecht also reported that Cicor had invested in a technology centre for printed electronics at their Bronschhofen site in Switzerland, where a team of application engineers was working on the development and industrialisation of new additive manufacturing processes and connection technologies.
Dr. Albrecht’s introduction to aerosol jet printing provided a fascinating vision of an alternative procedure, but inkjet technology had become adopted as a selective digital imaging process in the printed circuit industry and recently established as an effective means of applying solder mask.
Chris Wall, technical director of Electra Polymers, gave a detailed description of the development and processing of inkjettable solder mask and the benefits of its use in PCB manufacturing. He went through the logical sequence of stages in the development of a new ink: understanding the inkjet application process, identifying the formulation constraints for an inkjet solder mask, identifying and sourcing suitable candidate raw materials, and meeting solder mask performance requirements and external compliance requirements.
Wall summarised the types of head available: drop-on-demand piezo, recirculating and non-recirculating, with examples of each, and described how the head type influenced droplet size and resolution. He further illustrated the way droplets were generated, their shape in flight, and the consequences of satellite formation.
When it came to formulating a jettable solder mask ink, Wall reviewed the constraints of raw material suitability, resin viscosity, pigment particle size, and filler content, and discussed the effects of surface tension and contact angle of the resulting ink formulation. Curing mechanisms were generally a combination of preliminary UV pinning and a final thermal cure. He explained resin and photoinitiator chemistry in sufficient depth to convey the principles without confusing the non-chemists in the audience and described how new products were formulated, evaluated, and submitted for independent testing for compliance to standards, such as UL 94V0, RoHS, IPC-SM840 and NASA outgassing, as well as automotive standards and customer-specific requirements.
Wall reviewed the benefits of inkjet solder mask as compared with liquid photoimageable products, which lay principally in a shorter fully-additive process route and no solvent emissions or developer effluent, and discussed the characteristics of image edge-profiles. The inkjet process enabled custom print strategies for specific designs, such as selective thickness control, multilayer deposition, and different surface textures.
Recognised for his consistently informative and entertaining presentations, Don Monn from Taiyo America gave his update on the progress of inkjet solder mask over the five years since 2015, without standing still. He focused on practical aspects of image definition and geometry, with particular reference to solder dams and solder mask defined pads, using a whole series of “then” and “now” cross-section photographs of actual solder mask features.
Monn demonstrated how edge definition and flow-out could now be controlled to maintain the physical height of dams such that whereas a nominal 5-mil dam in 2015 would slump to a height of barely 0.5 mils and an effective width of almost 10 mils, the equivalent feature now would have a height of almost 2 mils, while maintaining a width of almost exactly 5 mils with a sharply defined edge.
Taiyo America had submitted samples of their inkjet-printed solder mask for third-party testing and measurement of thickness over copper and over laminate, adhesion of solder mask dams, of solder mask over copper and of legend over mask, resistance to electroless nickel immersion gold, to solvents and to solder, and for print quality and registration. Monn showed comprehensive test results confirming compliance with standards in all respects. Of particular note were thickness measurements proving that coating thickness was maintained at track edges, and the results of thermal cycling tests: 1000 cycles -40°C +125°C, 500 cycles -40°C +140°C, and 100 cycles -65°C +125°C, with no deterioration of adhesion, surface hardness, or dielectric strength. “It’s a technology worth investigating!” was Monn’s closing comment.
Continuing on the inkjet theme, Uwe Altmann reviewed the evolution of the technology within Orbotech, from their first inkjet for PCB applications in 1999, their joint venture with the subsequent acquisition of a new system in 2007, to the present day with over 520 legend-printing machines installed worldwide. He discussed the key benefits of Orbotech’s patented DotStream Pro technology, which featured multi-LED based UV for perfect ink-drop pinning, and superior depth-of-focus for the best printing on challenging topographies.
Their NAT technology demonstrator had been used for several years in collaboration with a leading ink vendor to develop inkjet solder mask, resulting in the launch at the recent IPC APEX EXPO of the new Orbotech NEOS solder mask inkjet printer, offering a shorter, simpler and eco-friendly alternative to traditional methods. In addition to their proven DotStream Pro technology, the NEOS featured structural printing technology, which optimised feature sharpness with unique 3D algorithms and included automatic calibration procedures and consistent drop volume control.
The afternoon session began with a roundtable on roadmapping and standardisation, moderated by Tarja Rapala-Virtanen, with short presentations from Steve Payne, Alun Morgan, and Emma Hudson followed by a panel discussion.
Steve Payne, manager of European operations for iNEMI, introduced the organisation as an industry-led research and development consortium that roadmapped the future technology requirements of the global electronics industry. He explained that the iNEMI roadmap had become recognised as an important tool for defining the state-of-the-art in the electronics industry as well as identifying emerging and disruptive technologies. There were 500 participants globally. The organic PCB chapter drew on information from product emulator chapters, which defined the future needs for high-end systems, aerospace, automotive, office and computer, and portable and wireless. It had identified development needs in HDI technology, microvia plating, modelling tools for embedded active and passive components, improved layer registration, fine-line imaging, alternatives to back-drilling, cycle time reduction for rigid and flexible circuits, and Industry 4.0 adoption. Key drivers were miniaturisation with increased functionality, increasing data volumes and transfer speeds, higher-speed processors, increasing complexity of components, form and flexibility of interconnects, and environmental considerations.
Payne also listed current collaborative projects aimed at eliminating gaps in the technology evolution to satisfy product sector needs, commenting that, 10 years hence, 5G would probably have been superseded by 6G and that artificial intelligence might be the next boundary to cross.
Alun Morgan, technology ambassador for Ventec International Group, suggested revising the ANSI grade definitions for thermosetting laminates as used by UL, IEC, and IPC. The current grading system was based on the NEMA LI 1 Standard, first published in 1965, which classified materials based on their chemistry.
Morgan commented that, in the modern world of electronics, the performance of the PCB and the printed circuit assembly was all that mattered, that much of the performance was contributed by the base material, and that the designer only cared about the performance. The varnish building block definitions used in the current UL 746E standard requested that they are listed by chemistry. In fact, the building blocks were not defined as allowable chemistries, and some building blocks fell into multiple categories, which led to confusion. And the building blocks were not defined as to the minimum and maximum loadings.
The way forward, Morgan proposed, was to define base materials less on what they were made of but more on their performance. As an example, he gave IPC-4103, which characterised materials only on their Dk and Df performance at certain frequencies without concern for the chemical composition. Morgan recommended that once UL listed as to their safety characteristics, the base materials be put in common categories where the printed board shop must test one material out of the same group for PCB approval. And if the base material qualified as FR-4.1, it no longer mattered that a secondary resin was present from the board point of view.
Emma Hudson, technical consultant and expert on electronics industry standards and certification, summarised the current and recent standards development work of IEC TC 91, and the upcoming proposals from UL—particularly those related to solder limits, which were recognised parameters for PCBs, solder resists and metal-clad laminates. Solder limits were meant to represent the soldering processes the PCB would be exposed to during component assembly operations. UL was looking to change the term to “assembly soldering process” to provide a better description of what was meant by the parameter. The proposed default condition would be six cycles of IPC-TM-2.6.27 T260 reflow, and the reflow profile would be considered representative of wave and selective soldering. Follow-up service inspectors would check the maximum reflow temperature and the number of soldering cycles that the printed circuit assembly was exposed to, but would not be checking any other details of the reflow profile.
There was no plan to apply the new assembly soldering process limits to existing board types, but only to new evaluations, once the proposal had been accepted by the UL Standards Technical Panel and added to UL 796, UL 746E, UL 796F, and UL 746F. Hudson stressed that there was no guarantee at this stage that the proposal would be accepted. Not surprisingly, it was Emma Hudson’s topic that attracted the majority of questions from the audience.
The session on reliability and environmental technology was moderated by EIPC board member John Fix, director of marketing and sales with Taiyo America. His first presenter was Martin Cotton, project facilitator with the HDP User Group, discussing the final report of the project evaluating the effect of moisture conditioning on several high-frequency Dk and Df test methods using a variety of laminate materials. Earlier work had found that differences in moisture conditioning contributed up to a 43% difference in measured Df values, for the same laminate material, depending upon the test method used.
The objective of the final phase of the project had been to evaluate the effect of varying moisture conditioning on Df and Dk measurements for a series of different low-Df, mid-Df and high-Df laminate materials—half of which were halogen-free—to determine the sensitivity of each test method to different moisture levels and what test frequency range or ranges were most affected by the moisture present within a laminate material.
The results were presented as an extensive series of scatter plots and graphs, which demonstrated that the test method coupon design had a significant impact on the measured results. Trace-Conductor coupons with plane layers on both sides showed no detectable difference in Dk and Df that might be correlated with moisture conditioning. The moisture conditioning methodology used for the Z-axis and in-plane test methods was sufficient to show a significant impact on loss due to moisture conditioning. The effect of moisture conditioning on the Dk and Df of laminate materials was shown to vary depending upon the specific laminate material being tested, and the HDP User Group recommended that users of these high-frequency test methods include in their test runs a base-line laminate material as a control. The test results showed that the Dk and Df values for the as-received and dry baked samples with internal plane layers were not always very close, mainly because of uncertainties in measuring the exact moisture content in such samples.
Dr. Anna Graf, OEM marketing and application engineering specialist with Isola in Germany, discussed high-temperature stable base materials for e-mobility applications. She commented that there were still several challenges to be addressed, specifically driving range, charging time, and high battery cost. How could the challenges be approached? Dr. Graf explored the basic electrical architecture of electric and hybrid-electric vehicles and concluded that improvements were needed in system efficiency, power density, and miniaturisation.
New materials were required for power electronics. Silicon carbide was a more effective material than silicon but operated at higher temperatures. Ceramic-based direct bonded copper substrates had excellent thermal properties but were expensive and required specialist fabrication techniques. A high-temperature-stable FR4-like laminate could provide a cost-effective alternative.
Dr. Graf described a German government BMBF-funded project named HELP, with 15collaborating companies working to develop reliable and cost-effective high-temperature electronics for e-mobility based on PCBs made from high-temperature-resistant resin systems. The objective was an organic-based printed circuit material capable of operating at 175°C with a peak temperature of 200°C and increased temperature cycling resistance. In collaboration with the consortium, Isola had developed a halogen-free, next-generation automotive high-reliability laminate solution for high power and voltage applications that required extreme thermal stability. It was a glass-reinforced material with very low thermal expansion. Dr. Graf did not disclose the resin chemistry but made it clear that it was not epoxy. The material had processing characteristics similar to FR-4 and properties similar to polyimide.
Additional information and further discussion of base materials with high-temperature reliability came from Volker Klafki of Technolam in Germany. He began with a review of laminate developments and reliability expectations since the 1990s, up to recently-emerging requirements in connection with e-mobility when extreme anti-CAF performance and resistance to thermal ageing were demanded. The ongoing theme appeared to be, “What is sufficient now will very likely be insufficient in the future.”
Klafki summarised the criteria currently used to classify reliability: withstanding the thermal stresses of assembly, reliability during thermal cycling, and no degradation under high-temperature storage, defining end-of-life as the point at which electrical and mechanical properties had declined by 50% from their initial values. He discussed the significance of relative thermal index with reference to UL standards, and CAF resistance at high temperatures. He also described qualification procedures, properties, and process guidelines for two high-reliability proprietary laminates developed for high-temperature applications, with very low thermal expansion and superior CAF resistance.
“Quo vadis flame retardants? How can we meet ever more stringent performance and sustainability demands?” were questions posed by Dr. Adrian Beard from Clariant in Switzerland. In his presentation, he gave an overview of the many challenges facing the flame retardants industry and some solutions that were in the pipeline.
Dr. Beard explained that the regulatory and environmental pressure on halogenated flame retardants continued to grow, and flame retardants continued to be added to the candidate list for “substances of high concern.” The RoHS directive was once again under revision, and more substance restrictions might happen. The European Ecodesign Directive would restrict the use of halogenated flame retardants in electronic displays as of 2021. In Sweden, a tax on household appliances and electronics, clearly not based on science, penalised not only halogenated flame retardants but also additive phosphorus flame retardants that had good environmental and health profiles. It was clear that not only stricter chemical legislation but also the trend to non-chemical legislation could drive the transition to halogen-free flame retardants in many areas.
Dr. Beard explained that Clariant were members of the Phosphorus, Inorganic, and Nitrogen Flame Retardants Association (PINFA)—a global group of flame retardant manufacturers and users committed to fire safety and improving the health and environmental profiles of their products. Clariant themselves produced a range of flame retardants based on aluminium diethylphosphinate for advanced electronic materials, and these had achieved the highest sustainability standards. “Be prepared. Flame retardants under scrutiny usually do not come as surprises,” he stated.
The subject of brominated flame retardants continues to be a matter of regulatory controversy. Tetra-bromo-bisphenol-A (TBBPA) has traditionally been used as a reactive component in the manufacture of brominated epoxy resins. Product designers have long presumed that TBBPA is fully reacted in laminates although, there has been little public data documenting this or defining potential “free TBBPA” within the epoxy polymer.
Sergei Levchik, new product development manager at ICL-IP America—a member of North America Flame Retardant Alliance—reported a study analysing commercial laminates to determine the concentration unreacted TBBPA flame retardant in printed wiring boards.
Laminate samples selected to cover a wide Tg range—from 140–200°C—were obtained from major suppliers to the electronics supply chain. In addition, halogen-free laminates were tested as controls. Samples were prepared according to EPA Method 3545A (pressurized fluid extraction), and EPA Method 8321B (high-performance liquid chromatography, coupled with thermospray mass spectrometry and ultraviolet detector). The test method was independently validated with respect to TBBPA detection and quantification, and two control samples of TBBPA-free laminates were used.
It was shown that a representative set of FR-4 laminates provided by three suppliers did not contain free TBBPA after curing. It was also shown that high-end laminates did not contain free TBBPA. These analyses demonstrated that TBBPA could be used effectively and safely to enhance the flame retardant performance of printed wiring boards.
Dr. Beard’s presentation concluded the first day’s conference session, and delegates boarded buses for a lesson in automated logistics and smart technology in a tour of Hutchison Ports Europe Container Terminal (ECT) Rotterdam, one of the most advanced container terminals in Europe, situated at the Maasvlakte, directly on the North Sea.

Approaching 20 million 20-foot equivalent container units are handled every year, with most of the loading and stacking done by autonomous robotic cranes and computer-controlled automated guided vehicles. It is an enormous and amazing operation.

The final event of a long day was a splendid networking dinner at the castle pavilion in the Sparta Stadium, home of the oldest football club in the Netherlands.

After the meal, many would-be football stars invaded the pitch for team photographs in the goal-mouth before returning to the conference hotel for rest and relaxation in preparation for the second day’s programme.

As ever, I am extremely grateful to Alun Morgan for allowing me to use his photographs. —Pete Starkey
Editor’s note: Click here for Pete’s recap on Day 2 of the 2020 EIPC Winter Conference.