To: StanX Long who wrote (13158 ) 1/28/2004 11:52:49 PM From: StanX Long Respond to of 95865 Immersion could extend to 22-nm node, says TSMC exec Silicon Strategies 01/28/2004, 5:00 AM ET LOS ANGELES -- Taiwan Semiconductor Manufacturing Company Ltd. plans to introduce immersion lithography for critical layers of the 65-nm node, starting in the middle of 2005, said Burn Lin, senior director of the micropatterning technology division at TSMC. Lin added that with alternative liquids the technology could extend 193-nm wavelength lithography to the 22-nm manufacturing process node. As announced in December, TSMC has ordered a 193-nm immersion scanner from Dutch lithography vendor ASML Holding NV (Veldhoven, Netherlands). Plans call for the 193i system to be installed in September at TSMC's Fab 12, a 300-mm facility. TSMC (Hsinchu, Taiwan) initially plans to use the immersion tool for the polysilicon gate formation and contact layers at the 65-nm node. It is expected to enter what TSMC calls "risk production" in mid-2005. While some companies plan to use "dry" 193 scanners for the 65-nm node, Lin said "our thinking is that if we have the tool, we would be foolish not to use it." In an interview here at the International Sematech immersion lithography workshop, Lin said a dry 193-nm tool provides a depth of focus of about 180-nm, while an immersion tool with the same 0.85 NA (numerical aperture) lens is expected to have a depth of focus of about 380-nm. That will allow a much better process window to deal with any lens aberration, tilting or wafer flatness issues, he said. Lin said he believes 193-nm immersion lithography can be extended much further than earlier expected. Catadioptric lenses, which mix refractive and reflective optics, will further boost the depth of focus. Also, new liquids are being studied that extend the refractive index to the 1.6 regime, compared with 1.43 for purified water. The catadioptric lenses and new immersion fluids, could combine to push the numerical aperture to 1.6 or so. That would make 157-nm scanners with immersion techniques unnecessary, he said, and could also push out extreme ultraviolet (EUV) lithography so far that it might never be introduced, largely for cost reasons. "The 193 immersion extensions are much less of a change in the infrastructure than the 157-nm or EUV systems would be. If we can come up with a fluid with a refractive index of 1.6, we can forget about 157 immersion, because that fluid would take 193 immersion into the 22-nm territory. And that makes the EUV people very unhappy," said Lin. Lin is considered an early pioneer in immersion lithography. He worked on the approach during a 22-year career at IBM Corp., and has promoted immersion lithography for use at TSMC, which he joined in 2000. During a morning session on Tuesday (January 27) at the Sematech-organized workshop, Bruce Smith of the Rochester Institute of Technology, and Will Conley, a Motorola assignee to Sematech, separately reported on early investigations into new immersion fluids. Doping water with phosphoric acid appears to be the most promising, Lin said.