193-nm development hampered by contamination

05/01/1998

193-nm development hampered by contamination

As the world lithography research community congregated at the SPIE Microlithography Symposium in Santa Clara, CA, the aggressive drive to commercialize 193-nm technology has reached some new milestones, but also turned up some thorny problems that could affect plans to bring it into manufacturing in 2001.

Perhaps the most challenging of the problems is the issue of resist outgassing and subsequent contamination of optical elements. Ron Kunz of MIT-Lincoln Lab, a leading site of 193-nm research, indicated that the outgassed material is a complex mixture of low molecular weight photo-products, which can be oxidized and bonded chemically to the lens surfaces by UV light. The problem has affected several developmental exposure tools, which have required extensive cleaning. Rumors are also circulating of lenses having been ruined by contamination from low-energy activation resists, which are the most susceptible to the problem.

Stepper lenses have always been vulnerable to resist outgassing, and cleaning procedures of one type or another are standard, but the issue takes on added significance at 193 nm because of the extremely close tolerances of the calcium fluoride optics. Even a few nanometers of buildup (which would occur in a year`s production even if today`s outgas levels can be drastically reduced) can result in light scattering, image flare, aberrations, and a decrease in ultimate resolution - an especially difficult problem when 193-nm systems will be asked to begin their work at device geometries substantially smaller than the 193-nm wavelength, and extend a generation or two beyond.

"It`s a serious issue, and it needs to be looked at carefully," said Mordy Rothschild, a program leader at Lincoln Lab. "We`re working at Lincoln Lab to quantify the amounts; it may vary from resist to resist, depending on the formula. Some outgas more than others."

John Bruning, president of lens maker Tropel Inc., Fairport, NY, added, "It`s happened with i-line for years; some deep UV resists are more prone to it than others. You have to either clean the lens once in a while, or add a window to protect it. But either way, it becomes a component of the optical system." Rothschild was skeptical that cleaning procedures could be practical in a field situation.

A long-standing issue with 193 nm is the calcium fluoride optical material itself. While new attention to manufacturing methods has greatly reduced the problems of internal absorption of the light and subsequent damage to the lens, production of the material is still not a trivial task. "There are definitely a number of grades of fused silica and calcium fluoride that meet the needs, and others that don`t," said Rothschild. "There`s significant variation across the grades. Depending on the grade, when it comes to producing large quantities, you have to qualify it on a batch-by-batch basis. It`s a high-risk item, and there`s not much safety margin - less than today`s 248-nm materials."

On the upside, Kunz suggests that while the first commercially available resists are likely to be marginal, the resist industry should be able to quickly refine them into production-worthy formulations.

In another step forward, researchers from the University of Texas and DuPont Photomasks Inc. said they had printed 80-nm line-and-space patterns using a 193-nm exposure tool at SEMATECH. Numerical aperture on the tool was 0.06; sigma was less than 0.3. The process utilized a chromeless etched quartz phase-shift mask produced at DuPont`s Round Rock, TX, facility, and an amorphous polyolefin resist that has been in development for three years with funding from Japanese resist firm JSR. - M.D.L., P.D.