High-Index Fluids Look to 2nd-Generation Immersion

The first generation of immersion lithography has barely made it into a commercial production environment (see "Chipmakers Immerse Themselves in 193 Wet ," Semiconductor International , February 2005), but researchers have been nonetheless turning considerable attention to the coming generations — with higher numerical apertures (NAs), fluids with higher refractive indexes, and other high-index materials.

Placing a liquid between the final lens element and the wafer, thereby sending the laser beam through an element with a higher refractive index than air, does not improve the resolution of a printed image in itself, but it enables the use of lenses with higher NAs. The high-NA optics, in turn, bring improved resolution. With air as the imaging medium, the NA limit is just below 1.0, but with deionized water's refractive index of 1.44, the NA limit rises to just under 1.4. Water is expected to keep immersion lithography going to the 65 nm node. But higher-index fluids, affectionately referred to as "flavored water," will be needed to take immersion to the 45 nm node and perhaps beyond, enabling the abandonment of 157 nm lithography and easing the delay of next-generation lithography (NGL) techniques.

With the ability to achieve higher NAs correlating directly with the refractive index of the fluid, researchers are looking at fluids with indexes of at least 1.6 for their second-generation choices. A refractive index like that gives a 193 nm lithography tool an effective wavelength close to 116 nm. A third-generation fluid of 1.8 or higher could take that effective wavelength down to ~107 nm.

JSR Corp. (Mie, Japan) announced late last year that it had successfully demonstrated 32 nm line and space patterns (Figure) with a 193 nm immersion lithography system and its high-index solution, developed based on its SOLOnX technology. JSR revealed more details about its fluid research at SPIE Microlithography last month, noting that its researchers had studied several fluids that meet the criteria of 1.6+ refractive index, with one of the highest being 1.664. But the refractive index is by no means the only consideration for second-generation fluids. Beyond also looking at the fluid's absorbance at 193 nm (which needs to be low to achieve the appropriate transparency), there is a slew of other issues involved (Table ).

JSR has demonstrated 32 nm lines and spaces with immersion lithography using a fluid with a refractive index of 1.64. Such high-index fluids will be required for second-generation immersion schemes.

JSR began studying high-index fluids in earnest after the Microlithography conference last year, where many people were asking about the availability of next-generation fluids, according to Katsuhiko Hieda, manager of the Electronic Materials Division at JSR. Plenty of candidates have a high refractive index, he said, but some react easily with resists, and some absorb too much. Water doped with salts or acids is one group of materials that could be considered for their high refractive index, but they have shown to have lower transparency. Also, salts inherently break up and react with the resist and/or topcoat; and acids will attack the tool and lens elements, Hieda noted. Alcohol and its derivatives (e.g., glycerol) also have a high refractive index, but again have lower transparency, and also higher viscosity. Simple hydrocarbons have high transparency at 193 nm, but an insufficient refractive index. So JSR's studies focused on organic fluids. "Fortunately, we found a very ideal material soon," Hieda said.

The material, dubbed HIF-001, shows a stable absorption level, showing no significant interaction with the photoresist film, with or without a topcoat. At a 193 nm wavelength, the refractive index is 1.64 and transparency is M98%/mm. It is thermally and chemically stable, and shows no damage to the lens or other tool parts. In cooperation with Canon, using a two-beam interferometer exposure tool and a high-resolution ArF resist, JSR was able to achieve 32 nm lines and spaces without using a topcoat. The estimated NA in this case was 1.5.

JSR just recently began collaboration work with toolmakers, Hieda said, so it will still take a few years before the fluid could be commercialized for immersion lithography production. Until then, there are many reasons for the industry to stick with water and smaller-NA lenses to reach the predicted 65 or even 45 nm capabilities.

For additional information on lithography, go to www.semiconductor.net/lithography.