Open, Honest Forum for Everyone
Aaron Hand, Managing Editor -- Semiconductor International, 5/1/2005
I don't know if I was just paying more attention this time, or if there really were more contentious words to be said at this year's SPIE Microlithography conference. Perhaps it's the stress of all that looming line edge roughness, but more likely it's the fear of getting burned once again after putting countless amounts of time, money and effort into developing a technology just to have it dumped onto a passing garbage truck.
It seemed that speakers and attendees alike were speaking a bit more openly about their opinions on the subject of extreme ultraviolet (EUV) lithography as the heir apparent to the optical throne. Much was said behind the scenes about a faltering faith that it can make it in at a workable price point, although there were only a few brave souls willing to make public statements to that effect.
Fabian Pease, a professor of electrical engineering at Stanford University who has been involved in countless efforts — some successful, some not — to develop the next greatest lithography technique, included EUV on a long list of lithography techniques not likely to succeed (see "Economics Are Not on the Side of EUV Lithography"). Privately, after his plenary talk, he joked that EUV was a good way to keep the nuclear physicists busy with something other than nuclear physics.
Grant Willson, a professor of chemical engineering at the University of Texas at Austin who has been instrumental in making possible the production of advanced chips, gave a keynote speech on Monday of SPIE. He also won this year's Fritz Zernike Award for Microlithography. But later in the week he was fuming. He was angry with one reporter for (aside from spelling his name wrong) quoting him as saying, "No one will ever print a wafer for profit with EUV." In fact, according to Willson, he'd gotten in trouble for saying such things in the past, so he told the audience this time that he was not going to say, "No one will ever print a wafer for profit with EUV." Semantics aside, the audience knew what he was trying to get across.
And, despite his anger about being quoted saying such things, by Thursday night he was again publicly proclaiming that he thinks there's been an overstatement of immersion lithography's capabilities, and that there will likely never be a way to economically do EUV.
Although repeating his words in print may find me in equally hot water with Willson, the point is that the industry needs more, not less, open communication about whether a technology really has what it takes to succeed. Plenty of companies are still reeling from the lost investments in 157 nm lithography, and surely will be for years to come. They're still trying to figure out how to pay the bills left behind by 157 development, and aren't feeling quite so eager to throw more money at EUV, despite Intel's urgings.
Michael Goldstein of Intel , speaking at the Thursday night panel — "Can NGL Crest the 32 nm Summit: Where Is the Window of Opportunity for Emerging Lithography? — not surprisingly made his support of EUV clear. "EUV will be adopted," he asserted. And although he was perhaps a bit more subtle in his jabs at competing technologies, his message was plain when he said, "It's important that we stay on this schedule, and have as few diversions as possible." He referred to that pesky distraction known as 157 nm lithography, and was undoubtedly warning that the industry try to keep its mind off of other competing NGLs such as nanoimprint lithography.
Addressing Willson's apparent ire at the articles that had been written earlier in the week, Scott Mackay, co-chairing the NGL panel, commented that the speakers had nothing to worry about because the night's panel was to be an open, relaxed forum, so surely no reporter would be writing about it. Perhaps my uninvitation got lost in the mail. I looked around the packed room, noting the unwavering interest in this topic despite the late hour and the lateness in a grueling week of everything lithography, and wondered why a journalist would not also be interested in such a discussion. And why is it so hard to believe that the reading public might actually be interested in an honest, open discussion about the perceived realities of their future?
I'm not saying that it's time to pull the plug on EUV development, but the industry needs to continue to question authority, and be willing to discuss in an open forum when they think things aren't working.
-
Nanoimprint knows no limits. It is not limited by diffraction or NA
(optical and EUV are). It is not limited by repulsion between electrons or
secondary electron straggling (EUV is affected by latter). Molecular Imprints'
Imprio 250 reports to have &10 nm 3s overlay.
Templates exhibit self-cleaning, which addresses defect concerns. Throughput is
expected to be a few to tens of WPH, but this can be improved with larger
templates. Add to that the fact that cost is extremely low at the wafer and
template (which can also be nanoimprinted) level, and it is clear why this
technology should be very attractive to the cost-sensitive semiconductor
industry.
gl - 9/2/2005 12:16:00 AM CDT -
Immersion is less disruptive than EUV. Prior to immersion becoming a candidate, EUV was least disruptive among the known NGL possiblities, even though an infrastructure change was necessary. Even today, the EUV light source is not yet a done deal. On the other hand for immersion, the only changes required are higher index for immersion fluid and higher index for final lens. Viable candidates have been proposed or demonstrated for both cases (JSR's Solonx and Burnett's selection of n=1.9 materials). Polarization control is already built into immersion tools (e.g., POLANO). With these materials, an ultimate half-pitch of 30 nm is the limit. As tool throughput goes through continuous improvement, double exposure processing with a 200 WPH tool is still easily justified, allowing a half-pitch down to 15 nm.
GD - 8/7/2005 11:46:00 PM CDT -
EUV development fails to heed fundamental lessons and learnings from X-ray and electron-beam lithography. Fortunately, immersion lithography came through in time to save the 45 and 32 nm nodes.
Ed - 6/1/2005 12:17:00 AM CDT
Score One for EUV
10/19/2007Is EUV the Next 157 nm Lithography?
03/31/2006NGL Fights Through Economic Adversity
02/09/2010
