WILTON, Conn. Citing the extreme conservatism of semiconductor-manufacturing managers, Silicon Valley Group (SVG) has proposed a comprehensive road map for optical lithography at its Lithography Division here. The proposal would bridge the gap between current systems and the extreme-ultraviolet (EUV) litho systems being developed by the EUV LLC consortium.
The company backed the proposal with the introduction of its Micrascan 193, a "production-worthy" 193-nanometer argon-fluoride excimer laser step-and-scan system.
Micrascan is being used as a platform for a 157-nm flouride excimer laser stepper designed as a bridge to next-generation EUV lithography. Silicon Valley Group is confident that it will be able to roll out a production 157-nm system by the third quarter of 2000.
The argument for evolving optical technology is simply that it is already accepted by the industry. In addition, changes at the manufacturing level are "extremely painful," said Bill Hightower, SVG's president. "We have aligned our lithography road map with what our customers have been telling us they want, and if we start hearing something different, then we are prepared to change it," he added.
While competing technologies such as Lucent Technologies' Scalpel system are an important addition to the semiconductor industry's tool set, "cost of ownership" considerations will continue to make optical systems the mainstream technology well into the EUV region of the spectrum, Hightower said.
Assuming a similar purchase price across all offerings, semiconductor line managers will look chiefly at wafer throughput to determine the cost of ownership, and that is where Hightower believes SVG will have an edge. The Micrascan 193 will be able to process fifty 200-mm-diameter wafers per hour, and follow-on systems will use more power laser sources to increase the throughput. "Scalpel will not be able to keep pace," Hightower said.
The argument for an industrywide commitment to optical methods depends on developing EUV capability, which is still being researched. When that capability goes into production, however, SVG expects to have evolved its optical capability for a smooth transition. SVG engineers are developing 157-nm capability to fill what the company sees as the "90-nm node gap," said James McClay, vice president of the 157-nm advanced program.
"Last year, 157 nm was not even on anyone's road map. People felt there was no source and no transmission medium available, so why bother," McClay said. But, for the past two years, SVG has been working with partners to develop both missing elements.
The critical part of the puzzle is an optical-transmission medium suitable for very short wavelengths. The company has developed lenses for the Micrascan 193 that use calcium fluoride. The next step will be to reproduce the entire optical system using that material, which would form the optics for a 157-nm system.
That, in turn, will require a reliable source of the exotic material. McClay said several companies are working with SVG to put that supply in place.