Litho Factions Search for Resources

What lithography techniques will the industry use to mass produce critical layers at 32 nm? Of course, if anybody truly had the answer to that question, they could probably get a little rich. But there seems to be very little in the way of answers. Or rather, perhaps there's too much in the way of answers. It's an extremely difficult task that lies ahead, and the debate rages on about who should get the dollars and other resources demanded to make their technologies successful.

Molecular Imprints Inc. (Austin, Texas), for one, wouldn't mind having a little more of the green thrown its way to prove out its capabilities in nanoimprint lithography (certainly, the same would be true for other imprint companies). In a recent podcast interview with Senior Editor Alexander Braun, Molecular Imprints CEO Mark Melliar-Smith insisted that manufacturers were attracted to imprint lithography — with a particular pull from NAND flash and hard disk drives — because there simply are no alternatives that can meet the resolution requirements in the time frame needed.¹

But questions are still swirling around nanoimprint acceptance, particularly with regard to defectivity. As one listener commented after reviewing the podcast interview with Melliar-Smith, "The big final question for nanoimprint to pass is whether it can tolerate particles on the wafer. It is not only a question of defects on the stamp/template. Everything can be perfectly arranged except for a single particle on the wafer — would the imprint process be tolerant? How much would not get imprinted?"¹

Meanwhile, as Bryan Rice, immersion lithography program manager at Sematech (Albany, N.Y.), has repeatedly contended, funds are sorely needed to make the progress needed in high-index immersion lithography materials as well. Although the industry seems to be converging on water-based immersion combined with some sort of double patterning scheme for 32 nm, after that, "everything gets dark," Rice said during a presentation last month at the Strategic Materials Conference (SMC) in Half Moon Bay, Calif.²

Making high-index immersion lithography work for the 22 nm node and perhaps beyond will require developments in three keys materials areas: a lens material with a refractive index (n) of >1.9, an immersion fluid of n=1.8, and a photoresist with n>1.9. The particularly difficult area seems to be the lens material, which has been narrowed down on most accounts to lutetium aluminum garnet (LuAG). A key issue here is that the industry needs to make a go/no-go decision on LuAG (which may ultimately be too dark to be useful) by the end of March — not much time left to get all of the work done that will be necessary to prove its worth (or not). Fluids and resists face similarly daunting challenges.³

This past October at the Immersion Symposium in Keystone, Colo., somebody asked what it would take to accelerate development of high-index lens materials by a factor of 2. Peter Krüll of optics maker Schott stood up in the audience and gave an answer of $5M. A relative drop in the bucket compared to what extreme ultraviolet (EUV) lithography has received to try to get past its long list of significant challenges. Perhaps it's time to write the check. Anyone?

As Rice noted in his SMC presentation, the way forward for lithography will be difficult, regardless of the choice of technology. With the challenges being so, well, challenging, it seems like the solution would be to really give all solutions their fair shake. Throw a few million dollars in the direction of each alternative to make sure it won't indeed work for the industry. We might get lucky and find out that one of them actually will work.