Immersion interference lithography was used to pattern gratings with 22-nm half pitch. This ultrahigh resolution was made possible by using 157-nm light, a sapphire coupling prism with index 2.09, and a 30-nm-thick immersion fluid with index 1.82. The thickness was controlled precisely by spin-casting the fluid rather than through mechanical means. The photoresist was a diluted version of a 193-nm material, which had a 157-nm index of 1.74. An analysis of the trade-off between fluid index, absorption coefficient, gap size and throughput indicated that, among the currently available materials, employing a high-index but absorbing fluid is preferable to using a highly transparent but low-index immersion media.

We have measured the transparencies of a number of candidate resist materials for 157 nm, with an emphasis on determining which chemical platforms would allow resists to be used at maximum thicknesses while meeting requirements for optical density. Although ideal imaging is usually obtained at an optical density between 0.1 and 0.3 and values in excess of 0.5 can often result in nonvertical wall profiles, we chose to arbitrarily choose 0.4 as the maximum tolerable optical density. Using this analysis, our findings show that all existing commercially available resists would need to be Ͻ60 nm thick, whereas specialized hydrocarbon resists could be made ϳ100 nm thick, and new resists based on hydrofluorocarbons, siloxanes, and/or silsesquioxanes could be engineered to be used in thicknesses approaching 200 nm. We also assess the tradeoff between these thicknesses and what current information exists regarding defects as a function of resist thickness.

Liquid immersion lithography, especially at 193 nm, is a serious candidate for extending projection optical lithography to the 65 nm node and beyond. This article reviews the status of this technology, the potential pitfalls that it may still encounter, and also the potential to extend it to 157 nm and to higher-index liquids. At 193 nm, no fundamental obstacles have been found yet, although defect control and materials compatibility must still be worked out. At 157 nm, significant progress has been made in developing suitable liquids. The next hurdle is to increase their refractive index, in order to make the transition in wavelengths cost-effective.