Lith Experts Explore Options for 32 nm Half-Pitch

The promise of continued lithography advances seem to be coming from several different directions these days. For our latest technology webcast, “Lithography at 32 nm: Finding the Right Solution,”  four industry experts gathered to debate the merits of high-index immersion, double patterning, extreme ultraviolet (EUV), nanoimprint, e-beam and optical maskless lithography techniques. Kurt Ronse, director of advanced lithography at IMEC (Leuven, Belgium); Bryan Rice, immersion lithography program manger at Sematech (Austin, Texas); Grant Willson, Rashid Engineering Regent Chair at the University of Texas at Austin; and Hank Smith, professor of electrical engineering at the Massachusetts Institute of Technology (MIT, Cambridge, Mass.), presented the latest developments in these technology areas. Although all of these techniques show considerable potential to one degree or another, not one of them is without significant challenges for fulfilling the requirements for high-volume production of critical layers at the 32 nm half-pitch.

Double patterning with 193 nm water-based immersion lithography has been gaining considerably more favorable attention in the past year. IMEC has been experimenting with a number of typical cells, finding great success with flash patterns. For logic, however, some problems have surfaced. There is some tendency for necking, Ronse pointed out, and exposure steps must also compensate for line-end shortening.

EUV lithography offers a considerable relaxation of the k₁ factor, but further improvements are needed in terms of source power in combination with good debris mitigation, Ronse noted. “And also the optics lifetime still needs to be proven in real life. Furthermore, infrastructure needs to be established in terms of EUV resists and EUV reticles.”

Sematech's Rice presented possibilities in high-index immersion lithography. Although second-generation fluids with an index of refraction ~1.65 are commercially available, third-generation materials (with an index of ~1.8) are as yet unavailable. The second key materials question for extending to high-index immersion is the lens material itself. According to a Sematech-sponsored study of materials, the one material that stands out as potentially extending all the way up to ~1.7 NA is lutetium aluminum garnet (LuAG). Although a manufacturer is working on creating production-worthy LuAG, production will take a number of years to complete — an obvious deterrent, Rice noted.

Although double patterning shows promise for bridging the gap to the next lithography solution, the increased steps and hit in throughput make it an expensive proposition. (Source: Sematech)

Willson, who described the merits of nanoimprint lithography, conceded that it is probably the least likely to be ready for commercial application at the 32 nm half-pitch. “It's probably not going to happen in that timeframe unless a major change occurs in the level of support that's being provided to it,” he said. “The problem is, for example, one needs defect-free templates, and that is not something that's going to be developed spontaneously by small companies and university laboratories.” That has to be demonstrated by the sort of consortium activity that's being funded by Sematech to make defect-free EUV templates. “I don't know how to initiate that, but certainly everything that we can see is extremely promising.”

MIT's Smith said that he thinks the probability of e-beam lithography addressing the 32 nm node is essentially zero. “I think that the same can be said for EUV. I think they have a very serious shot noise problem, not to mention all the other problems.”

Ronse sees double patterning as an expensive technology that will nonetheless buy time for alternatives. “And I think EUV definitely needs some more time to develop and to resolve the remaining issues, but I don't consider it impossible that it is going to happen, if you look at the progress that has been made in the past several years.”

Unlike the others, Rice is confident that a 1.55 NA system will be available in time with materials solutions for high-index fluids, but not so for 1.7 NA systems because of the difficulty in finding high-index immersion fluids. “One of the problems is that there's not been enough investment in high-index materials development because the money's going to other technologies — or, other technology (singular). It's the same problem for all the competing technologies. Many of the problems are actually tractable, but the industry has not focused on those problems.”

Find more information on lithography.