Industry funds DOE lithography development
11/01/1997
Industry funds DOE lithography development
In hopes of enhancing the competitiveness of the US semiconductor industry and ensuring the future safety and effectiveness of the nuclear arsenal via lithography, Frederico Pe?a, secretary of the Department of Energy (DOE), announced a Cooperative Research and Development Agreement (CRADA) between DOE and the Extreme Ultra-Violet Limited Liability Co. (EUV-LLC). The EUV-LLC is a private IC industry consortium, presently composed of Intel, AMD, and Motorola, but open to new US members willing to pay the minimum of $5 million. Over the next three years, EUV-LLC will pay $250 million to the DOE to support 100% of the costs of developing EUV lithography at the Virtual National Laboratory (VNL), a superposition of groups at Lawrence Berkeley, Livermore, and Sandia National Laboratories under the direction of Dr. Richard Freeman.
EUV lithography (see "EUV Lithography" in July and August Solid State Technology) is an experimental method of projecting images with resolution below 130 nm using 13-nm wavelength radiation from a laser-produced plasma. Because of the short wavelength, conventional transmissive optics does not work, and the all-reflective optical system requires special coatings and superb precision, both demonstrated at VNL during a previous CRADA. The major problems of building a prototype exposure system will be addressed during the new three-year CRADA.
The split of the money between the labs has been predetermined: 45% to Sandia, 45% to Lawrence Livermore, and 10% to Lawrence Berkeley. Berkeley`s contribution will primarily be characterization of optical components using the lab`s Advanced Light Source (the world`s brightest source of EUV radiation). Since this CRADA represents something of a "swords-to-plowshares" transition, headcounts are not expected to increase at any of the labs.
 |
This wafer was patterned on the world`s first integrated laboratory research system capable of printing functioning microelectronics devices using EUV lithography. The tool was assembled at Sandia National Laboratories.
Secretary Pe?a emphasized in his remarks that the Clinton Administration is committed to supporting the IC industry, including the semiconductor equipment industry. Continuing progress at the rate required by Moore`s Law was an appropriate new mission for the National Laboratories in the post-Cold War era, according to Pe?a. The DOE also has an internal interest in more powerful computers because of its need to certify the safety of the nuclear arsenal through computation, rather than bomb testing. Nevertheless, Congress declined to support the EUV lithography program, requiring private-sector funding.
Gordon Moore, chairman emeritus of Intel, spoke of the need to go beyond optics to continue the 107-fold reduction in the price of ICs, which has driven the Silicon Revolution. "Optics has done wonderfully, but it will hit a wall at 130 nm in about ten years," averred Moore. "We need to go beyond optics, and while a lot remains to be done, the National Laboratories have proved the EUV Lithography concept." (See figure.)
In later discussion, Pe?a admitted that there would be no direct return to taxpayers for the millions of dollars invested in EUV before the creation of the EUV-LLC, but Jerry Sanders, chairman and CEO of AMD, emphasized that large taxes would be paid by US IC companies if they were successful in the future. "Industry revenue is predicted to be $300 billion in the year 2000, the more of that in the US, the more taxes," said Sanders, "and this partnership also preserves a national treasure, the Virtual National Laboratory."
Sunlin Chou of Intel reported that the US currently leads in EUV, but he expects parallel and competing efforts to arise in Europe and Asia. In response to a question, Chou admitted that there are multiple options for post-optical lithography, which will be developed in parallel to the EUV program by SEMATECH and others. "We will consult with SEMATECH. Much of the infrastructure is common among the different methods," said Chou.
Apparently repealing the law that bears his name, Dr. Moore admitted that the industry will not be able to fabricate 5% of predicted semiconductor demand using EUV lithography in 2007, "But the capability will be ready to install."
EUV lithography requires many challenging innovations. The EUV-LLC will contract with vendors for components and rely on some equipment vendor (yet to be named) to integrate it into a production tool. The highly accurate optical surfaces will be fabricated by Tinsley Inc., according to Robert Aronno. (At press time, Silicon Valley Group planned to acquire Tinsley, but the deal had not yet been concluded.) ISI Inc. will develop the high-precision maglev stages, according to Larry Thompson, ISI VP of R&D. The critical multilayer coatings will most likely be applied by 3M, and the laser driver for the radiation source will be built by TRW Inc. and Northrop-Grumman, according to Moore. Building the lithography tools could become a tremendous business, with sales above $25 billion/year predicted for 2005.
The EUV consortium may be a natural follow-on to the 1? x-ray lithography consortium, which expired on September 30, 1997. Thus, investing in this new venture seems appropriate to C. Joseph Mogab of Motorola, which is a key participant - along with IBM - in the 1? x-ray enterprise. At $250 million, EUV seems a bargain compared to the $1 billion or so spent on x-ray in the last 15 years. Still, $250 million may be only a down payment on the total cost of changing to a lithography capable of extending down to 50 nm.
Notably absent were representatives of Ultratech Stepper and SVG Lithography, the two US companies most likely to be capable of integrating the various EUV technologies into production-worthy step-and-scan systems. Ultratech, in particular, had announced previously a corporate strategy of investing in R&D to enter the critical lithography market in the EUV era. While only US companies are welcome in the EUV-LLC, that consortium seems willing to transfer production of the exposure systems to whomever commits to build them in the US and deliver them initially only to EUV-LLC members.
"We`d certainly like to have more players," Moore commented. We`ve already approached other US companies. Asking a company to increment its R&D expenditures in the middle of a budget year is always difficult. This is an expensive program. We hope that we`ll get more (participants) as we start showing favorable results."
Future lithography is always controversial, and the transition to post-optical lithography is full of risk. The creation of EUV-LLC and its CRADA with VNL preserves EUV lithography as an option, which would have disappeared otherwise through lack of federal support. Other options include 1? x-ray, SCALPEL and direct-write e-beam and ion projection lithographies and perhaps imaging interferometric optical lithography. Each is likely to be pursued by different companies and consortia. SEMATECH will hold a meeting in Colorado Springs in hopes of achieving consensus as to which method deserves a clear track. Even if consensus is reached, tremendous difficulties and controversies remain ahead. The risks and potential rewards range into the hundreds of billions of dollars. For now, the announcement in San Jose appears to be a step forward. - M.D.L., E.K.