Photolithography at wavelengths below 200 nm

Summary form only given. The mainstream patterning technology in microelectronics has been projection photolithography. It has successfully sustained the continuous shrinking of device dimensions through the 1-/spl mu/m barrier to the 0.25 /spl mu/m employed today in advanced circuits, through a combination of wavelength reduction, increased numerical aperture of stepper optics, and resolution enhancing techniques such as phase shifting masks. The next major change will probably entail a shift to 193-nm lithography, possibly followed by a further change in wavelength to 157 nm. The obstacles in the way of implementing these changes are mainly related to materials: optical materials and coatings for the optical elements and for the photomasks, and photoresists for patterning and pattern transfer. In addition, high-repetition-rate, robust lasers operating at these wavelengths must be engineered. This paper reviews recent results on the performance of fused silica and calcium fluoride as optical materials, as well as data on coatings, photomasks, and pellicles. The alternatives for photoresists are presented, including a discussion on the limitations of each approach. We report on high-resolution imaging at 193 and 157 nm, with and without resolution-enhancing techniques. A critical assessment of all these results indicates that photolithography at 193 and 157 nm, while still facing significant challenges, has the potential of extending current lithographic practices to patterning at least 130-nm, and possibly sub-100-nm devices.