DALLAS - Governments in Europe and Japan are spending lavishly to bankroll the development of extreme-ultraviolet lithography systems.
Various European Community countries will put about $100 million into EUV-related programs over the next few years, participants at the inaugural International EUV Lithography Symposium here said last week.
Tokyo has stepped up to the challenge as well, organizing a cooperative extreme-UV source program and bringing together stepper rivals Canon Inc. and Nikon Corp. at an unprecedented level of cooperation.
Indeed, at the dawn of commercial EUV manufacturing, "Europe, led by ASML, is in the lead position, but Japan is scaling up rapidly," said Chuck Gwyn, an Intel Corp. employee who has been program manager at EUV LLC (Livermore, Calif.) since its inception in 1997. The private industry consortium is in the process of winding down its pioneering proof-of-concept work.
Both Nikon and Canon are expected to have beta EUV machines ready by late 2005, Gwyn said, in roughly the same time frame as ASML, the Dutch-based lithography vendor.
Relatively little federal money is going into EUV in the United States, though the Defense Advanced Research Projects Agency and the National Institute of Standards and Technology are providing some backing. The state of New York, by promising about $280 million over five years, convinced International Sematech to locate its EUV mask-substrate and photoresist centers in Albany.
Backed by its dozen chip-making members, Sematech will sink $40 million into the Albany center. And with the EUV LLC closing up shop, Sematech will step up to coordinate development by the vendor infrastructure and provide some financial support to equipment and material vendors, said Kevin Kemp of Motorola Inc., who is program manager for EUV lithography at Sematech.
By one estimate, more than $1 billion has been spent thus far in bringing EUV lithography to today's rudimentary state. Several more billions will flow before costs are recouped late in this decade, when EUV scanner shipments may exceed 100 units per year.
The keen interest in this next-generation lithography technology was evident here last week. Attendance at the three-day symposium exceeded the expectations of sponsor Sematech. More than half of the 310 participants were from outside the United States.
The carrot of public money is driving EUV development in Europe. Germany, for example, will spend about $50 million between 2001 and 2004, investing in R&D programs that involve German-based makers of the lasers, mirrors, mask substrates, metrology equipment and other components of the EUV infrastructure, said European participants here. For German companies, government support often amounts to 40 percent of development funds money that does not have to be repaid directly.
Japan's universities and private companies have already organized into an EUV association, and the government is bringing together laser makers Komatsu, Ushio and Gigaphoton into a cooperative EUV source program. Those three companies will develop the powerful lasers needed to generate the EUV radiation, said Yasuhiro Horiike, a University of Tokyo professor.
Elsewhere in Asia, South Korea's Ministry of Industry and Energy has formed a 10-year EUV program that kicks off next month. The Korean goal is to develop an alpha tool by 2008, said University of Tokyo's Horiike.
One fear is that the very complexity of EUV lithography will make it unaffordable. A European conference goer said that the Concorde supersonic airplane the symbol of a technical success that failed commercially often comes up as a chilly reminder at EUV meetings in Europe.
Peter Silverman, Intel's director of lithography capital equipment, raised the cost issue several times in his keynote address. "Margins in the semiconductor industry are miserable, and costs for 130-nm technology on 200-mm wafers are near the limit that the industry can afford," Silverman declared. With much of the equipment and materials industry facing "very hard times," Silverman said the commercial evolution of EUV requires that the big chip makers "step up and show their interest" by placing orders for EUV beta tools.
Many expect EUV scanners to be priced in the $50 million range initially, and Silverman warned that "EUV lithography must be affordable or it simply won't be adopted." Some scoffed at Silverman's call for EUV scanners in the $20 million range, with a maximum throughput of 120 wafers per hour.
"There are limits to the cost of EUV scanners that are very real," said Malcolm McGeoch, chief executive officer at Plex LLC (Brookline, Mass.), a 10-year-old company developing a gas-discharge method for generating the EUV radiation.
McGeoch, who earlier in his career developed technology for the Star Wars defense program, argued that laser-based systems will prove much too expensive, and will generate excessive heat in the process. Gas-discharge systems will win out, he argued, because the cost of ownership is likely to be one-third that of laser-based systems.
Ewe Stamm, president of Xtreme Technologies (Gottingen, Germany), a 30-person joint venture between Jena Optics and Lambda Physik, said his company is evaluating both approaches and will make a decision on which path to pursue in 2004.
Gwyn, the EUV LLC director, said there are several ways to improve the throughput of EUV scanners. The most obvious is to boost the output power of the source by an order of magnitude, and Gwyn said that over the past two years this has been done, albeit only to 10 to 20 W.
The sensitivity of the resist must be improved, and in his keynote address Silverman mentioned a target of 2 millijoules of sensitivity for commercial EUV resists another goal considered lofty by participants at the symposium. Gwyn said throughput can be increased in step with the reflectivity ratio for the 11 multilayer coated mirrors, which form the condenser and reflective-optics subsystems. A 1 percent gain in the reflectivity of each mirror will boost the power of the light at the wafer by 10 percent, he noted.
In addition, conference goers said that much work remains to be done on condenser design, collecting more of the photons generated by the laser source.
The symposium comes as the EUV LLC consortium, backed by six big chip makers with the help of three national labs, winds down. Glenn Kubiak, EUV-lithography program manager at Sandia National Laboratories, said about 20 Sandia scientists will continue to work on the engineering test stand in Livermore, and others will do contract research for the EUV LLC. But most of the scientists assigned to EUV lithography issues are back at work on Sandia's core responsibilities in nuclear weapons and homeland security projects. The Lawrence Livermore and Lawrence Berkeley Labs also contributed to the lithography project.