ICT Evening Seminar - "Strategies for Lead-Free Transition".
21st April 2004
Peter Starkey welcomed delegates to the seminar, which was devoted to the theme of lead-free, and its implications for bare board manufacturers and electronics manufacturing service providers. He introduced the first speaker, Dr. Martin Goosey, Research Fellow at Rohm and Haas Electronic Materials Europe Ltd, who talked about the imminent legislation on soldering and assembly, and how it will affect the industry.
Europe produces 6.5 million tonnes of electronic waste per annum and 90% goes to landfill. Landfill availability will change in the UK in the coming months, with the number of sites being dramatically reduced. Worryingly, the whole recycling infrastructure is currently underdeveloped. That is the physical situation, but the EC WEEE Directive covers 10 sectors and the RoHS Directive applies to them all except medical equipment. The RoHS Directive seeks to reduce the environmental impact of WEEE by restricting the use of certain hazardous materials which may be used in electronic equipment manufacturing, and as such lead, mercury, cadmium, hexavalent chromium, polybrominated biphenyls and two specific polybrominated diphenyl ethers will be banned from 1st July 2006.
It is the removal of lead that has the biggest impact. Lead is widely used, and the new alloys introduced to replace it behave differently. Materials and processes will have to change, products will need to be retested, and test procedures will have to be re-written. So what are the alternatives? It seems that conductive adhesives are some way off yet, so the industry will have to use alternative lead-free materials for PCB etch resists, solderable finishes, metallisation on components, device assembly and soldering.
There are alternative alloys, but none appear to offer all the processing benefits of conventional tin-lead solder. Viable alternatives have been identified, and the most promising alloys for widespread use have melting points well above tin/lead. Leading these is Sn-Ag-Cu (SAC), which has a melting range of 217º C. Sn-Bi is widely used in Japan.
However, there are areas of concern; these include substrate stability, mechanical and thermal; component degradation; fluxes and cleaning; effects on other organic materials;
Equipment capability – reflow profiles etc. and inspection, which would need different criteria; rework & repair - which alloy to use.
Martin looked at the concerns about laminates, which include warping, degradation, delamination, failure of interconnects in multilayer plated through holes. FR4 is OK, but CEM-3 is not. But what about unfilled dielectrics for sequential build applications? Higher Tg materials are needed to minimise effects of thermal expansion. Higher Tg materials are more stable, but they can be more expensive, and the desmear process might need some modification.
Concerning finishes, obviously tin/lead HASL board finishes will no longer be possible, but the alternatives are well established and increasingly popular, and include tin, nickel-gold, silver, and OSP. However, lead-free solders do not wet OSP coated copper as well as tin/lead eutectic solder. On the other hand, HASL can be made lead-free, as CEMCO recently demonstrated so convincingly.
Martin then went on to look at reflow. For example, typical reflow temperatures for alloys melting at around 215º C may be as high as 230-240º C. Reflow profiles will change dramatically, process windows will be narrower. Nitrogen inerting of the soldering process offered one way to broaden the process window. Consideration must be given to the modification of existing equipment, can they tolerate the higher temperatures; and what about the energy costs?
Lead-free is already well established. The Japanese are committed to lead-free, with or without legislation, and as one example Matsushita, based in Europe as well as in Japan, made all electronic components completely lead-free by April 2003.
Dr Robert Horsley is the Process Development Engineering Advisor at Celestica, Kidsgrove, and is immensely qualified to be so. He talked in great detail about the lead-free process development at Celestica. He examined the core lead free programme, and how it has been developed over the years.
Celestica want to qualify a lead-free assembly process by Q2 2004, and are well-placed it seems to meet that target. They commenced activities back in 1999, and by 2001 had in place a lead-free global team, who started assembling lead free products in 2003. Celestica brought together a team of some 25 scientists and chemist, from their centres in the UK, North America, Singapore, Italy and the weekly conference call was the way in which they discussed progress. Celestica became involved in NEMI, HDPUG,and are members of Soldertec, the SMART Group, CMAP and CECI.
Robert then looked at their findings with OSP and ENIG solderable finishes together with Sn-Pb reflow and SAC reflow. He discussed the work done with components, the thermal shock tests carried out on boards. Celestica had two programmes, the RIA1 was initiated using high Tg FR4 6-layer boards; they had 330 boards assembled, with surface finishes including OSP, ENIG and immersion silver. Components included CBGA, QFP, BGA etc. He talked through a flow chart showing the assembly route. Some of his comments included the note that nitrogen is good for reflow on OSP, and SAC better than Sn-Cu.
Test processes included ageing, vibration, plus visual inspection. Lead-free solder joint roughness was looked at in terms of voiding; OSP boards gave increased voiding. He then showed some results of inter-metallic types on ENIG boards.
The RIA2 programme covered SMT, wave soldering and rework. Splitting the focus of attention, they used the 405 alloy in Europe and North America, whilst Singapore and Italy used the 305. Here they produced 12 layer boards, 8” x 10”, 0.093” thick, with OSP, ENIG and immersion Ag. This was to determine a lead-free process compatible with PCB laminate. 400 boards were made, with an IST coupon, and IPC 9701 test method was used to assess reliability. Celestica will produce the results later this year, but Robert said that they had noted a big problem noted with barrel cracking. They had assessed several different laminates, but had used one PCB fabricator to make all the test panels.
Celestica are building lead-free products in Asia. In Europe and North America there is a lot of work being done with prototypes; the alloy of preference seems to be the SAC based 405, with 305 gaining increased interest. In conclusion, he talked about the implications for components, and the EMS implementation challenges; Celestica have an on-going programme of lead-free development to increase their knowledge of the subject, and that of their customers.
In answers to questions, Robert said that Celestica had calculated that the increased energy costs in moving to lead-free would cost the group some $ 8-10 million a year. The costs to Celestica of the lead-free project could not be given, he felt, but they would reflect the commitment given to the matter by a professional company investing in time and money on a subject, the enormity of which seems to have eluded some. Interestingly, not all of their customers had been committed to the lead-free project, but some had become closely involved.
It seems that the lead-free game is a ‘pull’. Who will ‘pull’ first? The convergence of compliant component demand with compliant component availability with compliant customer demand is possibly a timebomb, but whatever the situation, Celestica are ready to meet their customers’ demands.
Peter Starkey then thanked both speakers for their fascinating talks, and also expressed, on behalf of the many delegates attending, their thanks to Rohm & Haas for their kind hospitality and customary over-catering.
John Ling
Associate Editor
Circuit World