The move to 450nm: if, why and whenJul 23rd, 2012 | By admin | Category: Electronic Design
By Lou Covey
Editorial Director, New Tech Press
A successful trade show is one where industry news seems to point to a positive future. As such Semicon West 2012 was successful. The conference was abuzz with announcements supporting the transition to 450mm wafer production and extreme ultraviolet (EUV) methodologies. Intel announced a 10 percent equity investment to speed up ASML 450mm tool development, KLA Tencor announced QA technology for 450mm and Cymer announced that companies were buying their EUV source tech already. Several analysts claimed that in a couple of years, the foundation will be established for a complete transition by 2020 at the latest.
But out of all the positive vibes there were some questions. In the chicken-and-egg category was whether EUV will be production ready before 450mm tools. In the is-this-trip-necessary category, whether 450mm is even necessary if EUV is productive at 300mm.
KLA-Tencor is firmly in the “450mm is necessary” camp. The company launched the the Surfscan SP3 for 300mm defect and surface quality characterization requirements just last year and followed up with an announcement of the version for 450mm at Semicon this year.
Amir Azordegan, the head of marketing for surface scan systems at KLA-Tencor, said the company has booked seven orders for the new product and is in negotiation with several other customers. He identifies the interest in the product as a harbinger of the wide acceptance of 450mm. “Deep UV blanket wafer inspection systems enable the process to create better wafers. ” he said. “Without clean process tools we won’t get to 450mm.” (See complete interview at New Tech Press)
At Cymer, however, they are not as optimistic about the rapid acceptance of 450mm within the decade. Nigel Farrar, vice president for marketing and lithography at Cymer, said dealing with light sources, particularly EUV, will precede acceptance of 450mm. “The industry is risk adverse and is reluctant to make two major changes at the same time. Once EUV is mature and stable we’ll see the migration to 450mm.” Farrar sees the migration to 450mm beginning as late as 2018. (See complete interview at New Tech Press)
Cymer is working concurrently on EUV technology as well as 193nm immersion technology (they announced a focus drilling tool for that process at Semicon) for 300mm wafer production, both of which will migrate relatively easily to 450mm.
But all of that is based on everything going as planned. EUV technology was anticipated as being an industry standard by 2007. Here in 2012 we are seeing EUV technology rolling out only in “pre-preproduction” levels, according to Chris Mack, a recognized and award winning authority on microlithography. (See complete interview at New Tech Press)
Mack said that Cymer has provided the light sources for five of the 6 tools ASML has shipped to customers, but those tools are producing wafers at the rate of 5 per hour. The original spec was that the pre-production tool would produce 60 wafers an hour. So today we are at only pre-preproduction levels. Cymer has promised to upgrade the tools to reach that pre-production number by the end of the year to achieve the level of 60 wafers, which is probably enough to allow these customer to develop a process for 14nm, he estimated.
Mack said, however, there was no reason to link the success of EUV to 450mm acceptance since there were several alternatives to EUV that are further along, including 193nm immersion lithography. “Frankly, I’m a little surprised at the interest in 450mm recently. To me it represents a feeling of desperation on the part of the fabs looking at the 10nm node and beyond. The number of companies considering that node is practically zero. No one can figure out how to make chips at that level profitably. In the past, increasing the wafer size lowers the cost of equipment but I’m not sure that will work this time.”
Mack posited that moving to 450mm before 2020 may not be necessary if the new techniques are adopted at 300mm. “If everything goes flawlessly to plan, it would be possible to imagine these technologies in use by 2015. History shows, however, it rarely goes according to plan.”