A few years ago, when lithographers happily realized that immersion techniques would extend ArF lithography in a significant way, the expectation arose that high-index fluids could be invented to extend immersion capabilities even further.
It all seemed to simple.
Now, high-index immersion technology is facing an uphill battle. Chemists at DuPont and elsewhere are making progress with fluids that could replace water in between the lens and the wafer. A liquid with a 1.65 index of refraction (water is about 1.44) would be needed to extend immersion to 35-nm resolution.
The bigger challenge is the bottom lens element, which also needs to be a high-index material. ASML chief scientist Bill Arnold has said 1.9 is the index needed for the bottom element in order to extend immersion. Certain ceramic-based materials, such as Spinell, and LuAG, a rare earth material, are being studied. But they have serious problems with birefringence, one of the reasons why the whole 157-nm lithography effort was killed off.
John Burnett, a scientist at the National Institute of Standards and Technology (NIST) has surveyed the formidable technical issues with various materials that could be used as the bottom lens element. Not only are the technical challenges, including birefringence, pretty horrible. The cost of creating the infrastructure for making these materials for such a relatively small market as scanners is another huge issue.
Even if the material could be identified, and the infrastructure created, there is the timing issue. If a bottom lens element with a 1.9 index could be identified and brought to use for scanners, the need for such extended scanners in fabs would be in the 2012 time frame, when the 22-nm generation comes to volumes. That means chip companies would need these high-index scanners by 2010 to get their processes ready.
That is why ASML, Canon, and Nikon all have publicly backed away from high-index solutions on their roadmaps.
There is the possibility that these companies have programs back in the cloakrooms somewhere. Perhaps they are keeping the fires burning as insurance against EUV failures.
The SPIE Advanced Lithography conference, beginning Feb. 25th in San Jose, Calif., would be a good time to hear about research work that could lead to a high index lens element. Without that, EUV is the only viable high-volume option.