A research project at the University of Texas, funded by semiconductor consortium Sematech and backed by Texas photomask manufacturer Dupont Photomask Inc, reckons it’s achieved a breakthrough in process technology that is capable of achieving 0.08 micron feature sizes and smaller, something not featured on the Semiconductor Industry Association’s roadmap until the year 2009. Most efforts to date aimed at etching smaller chip designs onto silicon have centered around post- optical lithography techniques using x-rays, electron beams or extreme ultraviolet lasers. But the Texas team, led by the university’s Dr Grant Willson, used current photo lithography equipment to etch 0.08 micron features using an etched quartz phase shift photomask produced by Dupont, in combination with photoresist chemicals that interact with the light source to pattern the image onto the semiconductor wafer. Dupont used its own proprietary techniques to fashion the photomask, while the university team spent three years developing the photoresist chemical, an amorphous polyoelfin. I didn’t believe it could be done at first, said Willson, professor of chemisty and chemical engineering at the University. It really works better than my wildest imaginings, and it appears that the process latitude is there to generate smaller features yet. Although still unproven, the approach offers the possibility of an extension of the life of current optical-based lithography techniques, avoiding some of the huge costs associated with such a move. The big semiconductor manufacturers have already embarked upon such investments. Intel Corp, Advanced Micro Devices Inc and Motorola Inc, along with defense and telecommunications giant TRW Inc, last year began a $250m, three- year project to investigate the use of extreme ultraviolet lasers (CI No 3,245), while IBM Corp and the Japanese chipmakers have been investigating x-rays, and Lucent Technologies Inc has been looking at the use of electron beams. Other breakthroughs will be needed before 0.08 micron chips can be brought to market, but the lithography problem was the most immediate to be faced. The features were generated at Sematech’s Austin facilities. The team presented a paper on the results at a meeting in Santa Clara on Tuesday.
