by Christian Wagner, Ron Kool, Theo Modderman, Hans Jasper, ASM; and Kurt Ronse, IMEC
Immersion arrived on the lithography roadmap shortly after the first immersion images became available in the fall of 2003 . In addition to enabling a >50% increase in depth of focus (DOF) over dry lithography, immersion also offers increased critical dimension (CD) control, as well as the potential for mask and process simplification because of the larger DOF budget. With polarized illumination, the resolution limit for 0.93NA immersion systems has been extended to the 55nm half-pitch and, besides enlarged DOF, the water between the lens and the wafer enables the jump to lens designs having an NA>1.
Although water boosts DOF and enables higher resolution in the immersion lithography process, the challenge of introducing water is increased defectivity, which has been a primary concern of the industry since the advent of the technology. Significant progress in overcoming processing defects has been made in the past 18 months. Defects have been reduced to well below 50 per 300mm wafer for several state-of-the-art processes. Champion data as low as three defects per wafer have been communicated by TSMC . Consistency of results and process maturity will be key R&D topics for the coming months, with the goal being to bring the number of defects per wafer consistently below 10 during 2006.
IMEC’s Immersion Affiliation program on 193nm immersion lithography involves the participation of more than two-dozen equipment and materials suppliers, besides integrated circuit manufacturers. Some of the R&D activities undertaken by this program include development and optimization by all major manufacturers of immersion-specific resists and top coats. Leading photoresist suppliers are developing new resist solutions optimized for immersion. In order to support cost-effective processes, the target processes are topcoat-less, resist-only processes on the long term, and processes using a developer solvable top-coat on the short term.
In the IMEC 193nm immersion program, the major emphasis is on trying to understand the fundamental mechanisms behind the defect formation process. By identifying the root causes, solutions and workarounds can be tested. Several immersion-specific defect types have been removed during the last 12 months. Average defect counts on a 300mm wafer are below 50, with champion wafers having <10 immersion specific defects.
During the last year, progress has also been made in the IMEC program on overlay, bringing immersion overlay to the level of dry performance.
Besides enhancing DOF, the key benefit of immersion is allowing an NA>1. During SPIE 2006, imaging results of a 1.2NA exposure tool were presented. Among results showing an overview of 42/45/50nm dense lines printed at dipole/c-quad/annular illumination using polarized light, most notably, a DOF >900nm has been reached at a k1 of 0.265. The technology nodes that are accessible by 1.2NA are given in the table.
Improving the resolution is not only a question of NA, but also of more advanced illumination modes. There has been a tremendous advance in customizing illumination pupils per application in recent years. Starting with annular illumination, additional methods include dipole, quasar, C-quad, 5-pole, HexaPole, and more specific modes, optimizing both DOF and exposure latitude per application.
In order to make maximum use of the 1.2NA, polarized illumination is required. Flexible polarization pupil fill can be obtained by redirecting the polarization state of the laser to get the desired polarization state at full transmission. A variety of pupil shapes are made possible with X or Y polarization and pupil shapes with both X and Y polarization.
Immersion made substantial progress during 2005. With the advent of 1.2NA exposure tools, immersion lives up to its NA>1 promise and supports ArF lithography down to 45nm. Defectivity and overlay numbers, comparable to those of dry ArF lithography, indicate the technology is rapidly maturing and will soon be ready for introduction in volume manufacturing.
Twinscan is a trademark of ASML.
1. B. Streefkerk, et al., Extending Optical Lithography with Immersion, SPIE Vol. 5377, pages 285-304, 2004
2. “TSMC: Immersion Litho Nearly Production Ready,” Electronic News, 2/22/2006
Christian Wagner received his PhD in quantum optics from Universitat Munchen in 1994, and is senior product manager at ASML, ph 31/40-268-2975; e-mail Christian.Wagner@asml.nl.
Ron Kool received his masters and PhD in mechanical engineering from Delft U. of Technology. He is a director of product marketing at ASML Netherlands B.V.
Theo Modderman received his degree in physics from the Delft U. of Technology in 1986 and is a product development manager at ASML Netherlands B.V.
Hans Jasper received his degree in physics from the U. of Technology Eindhoven in 1977 and is a senior system engineer at ASML Netherlands B.V.
Kurt Ronse received his masters and PhD degrees in electrical engineering at the Catholic U. of Leuven, Belgium, and is director of the Advanced Lithography Program at IMEC.