Litho Experts Choose Immersion and EUV

Semiconductor manufacturers and their suppliers seem to be in agreement with the latest International Technology Roadmap for Semiconductors (ITRS) outlook for lithography technologies, according to survey results from the invitation-only Litho Forum in May in Vancouver. After hearing the latest results regarding the readiness of various candidate litho technologies, survey respondents indicated a preference for 193 nm immersion lithography for 45 nm manufacturing, and for extreme ultraviolet lithography (EUVL) at 32 nm.

To get the details of the results out to the rest of public, Semiconductor International held a webcast last month in cooperation with Sematech. Results were presented by Bernie Roman, Litho Forum program chairman; and Michael Lercel, Litho Forum conference chairman and director of lithography at Sematech. A question and answer session followed, with additional panelists including Janice Golda, director of lithography capital equipment development at Intel; and Will Conley, a member of the Advanced Optical Lithography Group at Freescale Semiconductor.

Asked to choose which primary lithography technology their company would employ or support for manufacturing in 2009, ~40% chose single-exposure 193 nm immersion for both gates and contacts. Alternatives were seen as double-exposure 193 nm immersion and dry 193 nm lithography. For 32 nm in 2012, ~37% chose EUVL, with interest also in double-exposure 193 nm immersion. For 22 nm manufacturing in 2015, EUVL became a more overwhelming choice, with ~60%.

Survey results showed a preference for EUV lithography by 2012, related to 32 nm half-pitch producton. (Source: Sematech)

The leading concern expressed by those choosing 193 immersion lithography for 2009 was the technology's extendibility. However, respondents showed surprisingly low interest in next-generation immersion — including fluids, resists and/or lens materials with higher refractive indexes — which would extend immersion's capabilities. Roman indicated surprise at the lack of support survey respondents showed for high-index immersion lithography. "In view of the tremendous industry activity going on for high-index materials, it would imply that there's more planning for this by the end users, and that didn't seem to come out, at least so much in the survey at the forum," he said. However, there was a bit more support for high-index immersion from respondents of a pre-forum survey that was sent only to end users, he pointed out.

"I think the reason that it didn't have a better showing is because I just don't think there's been enough information that's been presented," Conley said. "I also think that a number of the tool companies have been fairly open about whether or not they're going to pursue some kind of a system that just incorporates new immersion fluids. And there's concerns about cost; there's concerns about processability; there's concerns about what kind of radiation stability these materials have. So over the coming months, and probably over the remainder of the year, we'll see a lot more information about how these materials are progressing, and whether or not they can be incorporated."

Extendibility seemed to be the major reason that people supported EUVL for production in 2012 and beyond. For this technology, the top concern was cost of ownership (CoO), although there seemed to be some disconnect between chipmakers and suppliers on this point. "Exposure tool suppliers seem to think that's not as big an issue," Roman noted during the first presentation of the results at the Litho Forum. During the webcast Q&A, he elaborated: "In general, I think you could say from the tool point of view, the suppliers are clearly closer to the day-to-day issues on dealing with many of the things that are highlighted here in red. And they know the particular engineering solutions that are in place to deal with those things."

Golda added, "If you drill down to below CoO to exposure tool throughput and really source lifetime and the mask availability, those are the key factors that go into the cost of ownership. And I think what you're seeing here is what people anticipate will be the state of technology for the readiness by 2012. And there has been good progress in each of those areas, so people see that moving forward."

During the Litho Forum, there was considerable talk and interest in a variety of double-exposure techniques, but perhaps not enough yet to make the cut. "Certainly people recognize that double exposure is now a viable option," Lercel said. "But if you also look closely at the results, especially among the end users, there's a serious concern about many of the technical issues involved in implementing it. And especially the cost, and that was really something reflected in the survey results."

Find more information on lithography.