Mark Melliar-Smith Chief Executive Officer,Molecular Imprints Inc.
1807 West Braker Lane, Bldg C-100, Austin, Texas 78758, USA
Small feature imprint lithography has existed for several years. The original technique involved the use of a patterned template which is impressed onto a thermo plastic material, and with the combination of heat and pressure, the pattern in the template was transferred to the substrate. Compact disks were one on the early applications for the technology. Recently the technique has been significantly improved with the development of Step and Flash Imprint Lithography (S-FIL) This technique was invented by at the University of Texas, and involves deposition of a low viscosity monomer on the substrate, lowering a template into the fluid which then flows into the patterns of the template. Following this fill step, the monomer is exposed to UV light to cross link it and convert it into a solid, and the template is removed leaving the solid pattern on the substrate. The process is shown schematically in Figure 1 and has been recently commercialized by Molecular Imprints Inc.
The purpose of this paper is to describe the recent advances in S-FIL technology with particular reference to its application for silicon integrated circuits. In this respect the technology has become a viable contender for the “Next Generation Lithography” in competition with EUV. Although this paper, and this conference, are focused on this application, it is worth noting that S-FIL imprint technology can also be used in a wide variety of other advanced applications, many of which are only commercially viable given the availability of low cost (sub $2M) tools capable of delivering sub 50nm features. These application include the use of photonic band gap crystals to enhance LED efficiency and brightness, patterned media for disk drives, polarizers for projection optical engines and a wide variety of semiconductor electronic and photonic devices. For most of these applications the capital cost of 193nm immersion lithography is commercially untenable, even assuming that the required resolution limits could be reached.