IBM drives chipmaking process beyond current limitations

San Jose (CA) - IBM researchers claim to have found a way to use current deep-ultraviolet (DUV) lithography techniques for the production of 32nm chips and smaller. While The limits of DUV have been pushed out several times, the semiconductor industry plans to make a costly switch to an extreme ultraviolet (EUV) process in 2009. IBM's discovery could delay the transition for at least one more product generation.

Moore's Law is considered to be the main innovation driver for the semiconductor industry, and according to Intel, alive and kicking at least until 2015. But Intel's confidence often is met with doubt as the observation by Intel co-founder Gordon Moore that transistor counts in microchips double every 18 to 24 months appears to be increasingly difficult to achieve. Every new product generation requires companies to dedicate more human and financial resources and drives development and production cost to astronomic levels.

One of key investments is targeted at lithography, which is the process used to print circuits onto computer chips. Chip lithography is very similar to traditional silk-screening, but semiconductor firms use light instead of ink to print microscopic circuits onto their products. DUV, which uses 193 nm light sources, has been in place since 1995, but was believed to run out of steam by 2003 or 2004. Back in 1997, Intel, Motorola and AMD created the EUV Limited Liability Corporation to develop an extreme ultraviolet process (EUV) to replace DUV at the 100 nm production level. Today, first chips are built in 65 nm and the industry generally believes that EUV may be necessary to be implemented for 32 nm semiconductors, which are scheduled to arrive in 2009.

IBM now claims that it may be able to push out the introduction of EUV even further. Researchers of the firm claim to have developed 29.9 nm structures. A "high-index immersion" variant promises to be able to provide a path for extending Moore's Law further, thus buying the industry time before having to make the transition to EUV.

"We believe that high-index liquid imaging will enable the extension of today's optical lithography through the 45- and 32-nanometer technology nodes," said Mark Slezak, technical manager of JSR Micro, a firm that developed the 29.9 nm structures with IBM. "Our industry faces tough questions about which lithography technology will allow us to be successful below 32 nanometers. This new result gives us another data point favoring the continuation of optical immersion lithography."

IBM's and JSR Micro's findings are especially important, as it is not just very expensive to implement, but because the industry is still facing serious challenges in making EUV a reality. While Intel said in 2004 that it installed world's first commercial EUV lithography tool and set up an EUV mask pilot line - a milestone in producing smaller circuits with EUV - the technology especially appears to not only lack a stable and powerful EUV source.

Also, the 13.5 nm light of EUV is absorbed by all materials - including air - including the glass that traditional lenses are made from. Operating in vacuum, EUV lithography relies on high-quality, and exotic all-reflective optics instead of transmissive optics. Another concern are defects in EUV photomasks, which are prone to slight defects and currently expand under the heat generated by the EUV printing process beyond tolerance levels. Intel addresses this issue through a partnership with Corning, which is expected into a technology development that delivers high-volume EUV photomasks by 2009.

IBM's solution is based on an "interference immersion lithography" test apparatus, which the company calls "Nemo." The devices uses two intersecting laser beams to create light-and-dark interference patterns with spacings closer than can be made with current chip-making apparatus, IBM explained. The company plans to use Nemo for further research that may push DUV even further. "Now that IBM's new result shows a path for extending optical lithography, high-index lens materials must be developed to enable its commercial viability," the firm said.