Innovative Electronics Manufacturing Research Centre 4th Annual Conference

Loughborough, UK, 2nd September 2009

Henry Ford CollegeNotwithstanding the predictably miserable English late-summer weather, delegates and speakers from all over the world gathered enthusiastically at Henry Ford College, University of Loughborough for the fourth annual conference of IeMRC, the Innovative Electronics Manufacturing Research Centre. They were welcomed by Industrial Director Professor Martin Goosey who explained the objectives of IeMRC as the UK’s internationally recognised provider of world-class electronics manufacturing research, focused on sustaining and growing high value manufacturing in the UK by delivering innovative and exploitable new technologies, highly skilled people and strategic value to the electronics industry. Since its inception in 2004, IeMRC had funded research in 22 UK Universities ranging from 6 month feasibility studies to multi-institutional industrially-led flagship projects. Support for a further five years had been approved by the Engineering and Physical Sciences Research Council of the UK government, and new projects were being identified for commencement in March 2010.

Peter SandbournePeter Sandbourn from the Centre for Advanced Life Cycle Engineering at the University of Maryland made the opening presentation, on the subject of Making Business Cases for Sustaining Electronic Systems, subtitled “Managing obsolescence, counterfeits, lead-free, and other nasty supply chain problems”. Quoting legendary US baseball star Mickey Mantle “If I knew I was going to live this long, I’d have taken better care of myself.” Sandbourn posed some searching questions about sustaining safety-critical electronic systems, aircraft for example: “How do I build airplanes and sustain them for 30 years when I can only buy the components for two years?” Pointing out that in military systems it is not unusual for over 70% of the electronic parts to be obsolete before the first system is installed, and that system life cycles continue to increase, the B52 bomber being an example whose initially projected lifetime had been 30 years, back in 1955, and would probably be still in service in 2040, he commented that such products had no control over their supply chain for key components due to their low production volumes and were subject to DMSMS (diminishing manufacturing sources and material shortages)-type obsolescence. Present-day sustainment culture tended to be focused on rewarding management organisations for fire fighting, rather than fire avoidance, and breaking this reactive culture was difficult. If no serious event had occurred, a business case might be needed before management would be moved to action, and engineers tended not to be good at making business cases. Mitigation of parts obsolescence did not stop obsolescence from taking place, only managing it when it happened, and system sustainment problems would not be solved with just a technology solution, or by standardization of design rules and interfaces.

Edward Horsley, a PhD student from University of Sheffield, reviewed recent advances in the technology of piezoelectric transformers for power electronics applications. High-voltage step-up piezoelectric transformers had already been commercialized, mainly for driving fluorescent backlighting of liquid crystal displays, but step-down versions presented more complex challenges. Horsley explained the principles of four examples of piezoelectric transformer: Rosen, Radial-Mode, Thickness-Mode and Thickness-Shear-Mode, and described the relationship between the physical characteristics of the radial-mode device and its electrical performance, by way of equivalent circuit models which enabled transformers to be analytically designed for specific switch-mode power supply applications, giving substantial savings in size, weight, power density and EMI over equivalent magnetic devices.

Prof David CardwellDavid Cardwell, Professor of Superconductor Engineering at the University of Cambridge, gave a presentation entitled Bulk High Temperature Superconductors for High Field Applications, illustrated with live demonstrations of magnetic levitation accompanied by dramatic stage-smoke effects from the liquid nitrogen he used to cool his superconductor samples. In his context, “high temperature” referred to the boiling point of liquid nitrogen, 77degK, (-196degC), as opposed to that of liquid helium, 4degK. The only class of superconductor material suitable for 77degK operation was large-grain, boundary-free Rare-Earth Barium Copper Oxide, the rare-earth options being Yttrium, Samarium or Neodymium. The Yttrium-based material YBa2Cu3O7 (YBCO), processed in the form of large single grains by melt processing and peritectic solidification, offered the best short-term opportunity. Professor Cardwell discussed the bulk microstructure of YBCO, and explained the effect of second-phase inclusions on the flux-pinning capability of the material. Methods had now been developed to incorporate nano-scale inclusions into the bulk structure, and a 10-200nm nano-phase based on RE2Ba4CuMOy had been shown to form effective artificial flux-pinning sites. Although these materials required further refinement in microstructural homogeneity, they appeared extremely promising for improved critical-current-density bulk superconductors. There were potential applications in magnetic bearings, flywheel energy storage, motors, generators and NMR imagers.

Prof Johan LiuNanomaterials and Processes for Potential Electronics Packaging Applications was the title of the presentation from Professor Johan Liu of the Chalmers University of Technology in Gothenburg, Sweden. He discussed current research into nano-particle lead-free solders, nano-composites for heat dissipation, nano-particle conductive adhesives for flexible die-attach, and carbon nano-tube bumps for ultra-fine flip-chip interconnection. A particularly interesting characteristic of nano-particle solders was the lowering of melting temperature, by as much as 10degC in the case of Sn-3.0Ag-0.5Cu alloy. Professor Liu’s team had developed techniques using spark erosion in a liquid dielectric to generate solder nano-particles and had studied both pure nano-particle pastes and bimodal pastes containing mixtures of micro-particles with added nano-particles which had the effect of controlling grain growth and hindering dislocations. Heat dissipation nano-composites had been developed using electrospinning techniques to produce metallic filaments which were then impregnated at high pressure and high vacuum to yield fibrous metal networks in polymer matrices with remarkably good thermal conductivity, which could be tailored to specific applications. Work on conductive die-attach adhesives had resulted in flexible materials with micro- and nano-fillers which could be cured at room temperature with latent curing agents based on cobalt-organic complexes. Professor Liu reviewed developments in flip-chip interconnection using carbon nano-tube bumps, which could be formed in precisely pre-defined patterns, and transfer techniques utilising patterned isotropic conductive adhesive. In a separate application, carbon nano-tube based cooling fins had been produced in pitches down to 2 microns.

Continuing the nano-technology theme, Dr Samjid Mannan from Kings College London reported progress in the development of nano-particle-stabilized solder materials for high-reliability applications. The aim of the project was to incorporate inert nano-particles into solder at the reflow stage, to harden it and to stabilize its microstructure. The chosen nano-particles had a silica core approximately 90nm diameter, with a solder-wettable gold or palladium shell, and successful techniques had been developed for the deposition of the metallic shell. Dispersion of the nano-particles presented some difficulties, and best results had been obtained by mixing the particles into the flux before the addition of solder. The printability of the solder pate had not been affected, and incorporation of nano-particles into solder had already been partially successful, at 0.1 volume %, the objective being 1%. Compression testing had shown an increase in flow stress, and reliability testing was in progress.

General viewFrom Feasibility to Spin-out: Dr Alex Robinson from University of Birmingham described how molecular photoresists had been developed, from basic research through technology towards commercial realisation, with a step-by-step guide to structuring and implementing a business plan. Photolithographic imaging accounted for approximately 40% of the overall manufacturing time of microelectronic devices, with possibly 40 to 60 different patterns required in the fabrication of a processor chip. The Moore’s Law trends in feature sizes were pushing conventional polymeric resists beyond their limits: chemically amplified molecular resists offered a solution to the resolution/line-width roughness/sensitivity (RLS) trade-off for next-generation lithography. Resists based on derivatised Fullerenes (C60 bucky-balls) could be polymerised by electron-beam exposure, and sensitivity limitations could be overcome by chemical amplification using photoacid generators such as poly[(phenyl glycidyl ether)-co-formaldehyde]. Dr Robinson explained how line-width roughnesses as low as 2nm had been achieved on sparse resolutions of 12nm and dense resolutions of 20nm, and how other parameters including pattern transfer, stability to process conditions and ageing exposure on industry-standard tools had been quantified in establishing the commercial feasibility of molecular photoresists.

Dr Ricardo ValerdiMoving away from manufacturing processes and into cost modelling, Dr Ricardo Valerdi from the Lean Advancement Initiative at Massachusetts Institute of Technology discussed the economic impact of re-use on systems engineering. He likened cost estimation to crystal-ball gazing and considered cost modelling to be a black art. All models were wrong, but some of them were useful! Referring to the Department of Defense Handbook MIL-HDBK-881A, he posed the questions: How is systems engineering defined in a defence systems context, why measure it, and how do you quantify complexity? He introduced the concept of the Constructive Systems Engineering Cost Model (COSYSMO), used to estimate the number of person-months it would take to staff systems engineering resources on hardware and software projects, helping organisations plan activities including development, integration and test. ANSI/EIA 632 was used as the reference standard for data collection and work breakdown structure. COSYSMO addressed the first four phases of the systems engineering life cycle: conceptualisation, development, operational test and evaluation, and transition to operation. The COSYSMO calculation was based on a combination of four size drivers and fourteen cost drivers, of which eight were application factors and six team factors, and the model contained calibration data from more than 50 projects provided by major aerospace and defence companies. including Raytheon, Northrop Grumman, Lockheed Martin, SAIC, General Dynamics, and BAE Systems. Re-use of systems engineering work products could reduce the expected systems engineering effort for the development of a new system, although the benefits were limited to related domains and did not scale linearly. In a worked example, Dr Valerdi demonstrated a saving by re-use of 30 person-months on a project which would otherwise have taken an estimated 129 person-months.

Dr Linda Newnes, Head of Costing Research at the University of Bath gave a presentation entitled Modelling Costs for Through Life Availability, her stated aim being to provide methods and tools for managing through-life costing from concept design to end-of-life disposal. Amongst the examples she quoted to highlight the importance of cost modelling, the National Audit Office had determined that the twenty largest Ministry of Defence projects were on average 96 months late and £205M over budget. Addressing the challenge of helping industry to make informed decisions in estimating through-life costs, she tackled the complex issue of uncertainty modelling, explaining the differences between epistemic and aleatory uncertainty in calculating probabilities, and demonstrating how the principles could be applied across the supply chain to enable more transparent decision making under uncertainty and risk.

General viewLeading authority on electronics packaging, Professor Ricky Lee from Hong Kong University of Science and Technology gave an enlightening presentation on emerging trends and challenges of high-density packaging and three-dimensional IC integration.
He described a series of 3D packaging concepts: flip chips on silicon chip carrier with through-silicon vias as interposer (C2S), stacked multiple flip chips with through-silicon vias as vertical interconnection (C2C), embedded wafer level packaging (C2W), and wafer bonding for 3D integration (W2W). Discussing the scope of through-silicon via technologies, he described techniques for TSV formation by laser drilling or deep reactive-ion etching, and went on to explain interfacial structures and via plugging by copper electroplating. Other issues included 3D flip-chip stacking, re-distribution and bumping, wafer thinning and handling. Professor Lee examined the technology roadmap of the Semiconductor 3D Equipment & Materials Consortium, EMC3D, an international consortium with the mission to rapidly develop a cost-effective and manufacturable through-silicon via technology for 3D chip stacking and MEMS integration. Finally, he stressed the importance of design tools: “The technology pull is real, but the 3D designs won’t happen without the tools!”

The concluding presentation of the wide-ranging conference programme came from Professor Marc Desmulliez of Heriot Watt University. who discussed the results, opportunities and challenges originating from the Grand Challenge Project: 3D Mintegration, an EPSRC-funded project with the objectives of developing radically new ways of thinking for the end-to-end design, processing, packaging, integration and testing of complete 3D miniaturised/integrated products, to generate sustainable energy-efficient products and manufacturing processes and to reduce the UK’s vulnerability to globally-strategic supply chains. 23 companies were involved in a market-driven initiative aimed at next generation automotive, aerospace, telecommunications, medical and consumer products, with tangible deliverables.
In addition to academic outputs, personnel development and knowledge transfer activities, industrial take-ups already included a microprobe demonstrator, optical metrology techniques, microfluidic designs for blood separation and a health and usage monitoring microsystems demonstrator. The project continued to encourage agile thinking processes in a new generation of applications-orientated product and process designers and developers

General viewIn his closing remarks, Professor Paul Conway, Academic Director of IeMRC, thanked presenters and delegates, reviewed the successes of the past year and called for proposals for future projects, whilst offering guidance on selection of topics and presentation of submissions.

A truly outstanding conference, superb organisation, an excellent venue, the highest calibre of presenters and an attentive full-house audience, what more could we ask? Maybe the sun will shine next year…

Pete Starkey
First published by I-Connect007 and reproduced with their permission