Outlook improving for 193-nm material

07/01/1998

Outlook improving for 193-nm material

A recent worldwide survey by the Arthur D. Little consulting firm has found a generally improving outlook for production and availability of high-purity calcium fluoride (CaF2), a critical material for 193-nm optics, but the survey`s leader says it is too soon to know whether shortages of the material will limit widespread rampup of 193-nm technology.

Chipmakers are putting heavy pressure on wafer stepper developers for 193-nm equipment, which will likely be needed for production of 0.15- or 0.13-?m geometries. Current R&D tools have very small exposure fields and lack a number of features needed for development of full manufacturing processes; major stepper companies are currently eyeing initial shipments of 193-nm systems late this year or early in 1999. This will provide a scant two-year window for development before 0.15-?m production, and four years before 0.13-?m.

CaF2 will be used for key optical elements, especially in refractive systems; until now there has been no demand for ultra-pure versions of the material, and its availability has been one of the biggest lingering question marks affecting 193 nm.

"It`s too soon to tell whether CaF2 shortages will limit the widespread ramp up of 193-nm lithography," said Irwin C. Winkler, senior program manager for semiconductors at Arthur D. Little, Cambridge, MA. Winkler headed the three-month study. "At present, CaF2 shortages are affecting the rate of 193-nm lithography development. But the industry is changing rapidly and there is nothing fundamental, either technical or commercial, to suggest that shortages are inevitable. In my opinion, the biggest challenge is in convincing small companies, i.e. the crystal growers, to make investments in capacity to meet the anticipated demand."

Currently, the most urgent need is for production of large-diameter, high-quality single crystalline material, with high purity, low stress birefringence, and low index inhomogeneity, said Winkler. "The outlook for additional capacity to produce this grade of material is actually quite good. Several of the manufacturers we spoke with were in the process of planning, or at least seriously considering, investments for 193-nm lithography optical material, and this was true for both the crystal growers and the suppliers of synthetic CaF2 powder, (which is) used as the feedstock for the crystal furnaces."

Winkler said the CaF2 industry currently has about six crystal growers and two or three suppliers of raw materials. "Some are further along than others in developing material suitable for 193-nm lithography, but the differences are not huge," he said. Most crystal growers have been designing furnaces and growing crystals for many years, he added, with some purifying their own feedstock. Two are US-based, one European, and three are in Japan, where HDTV cameras provide a small market for large crystals. Winkler noted that Japanese stepper maker Canon owns one of the crystal growers, while rival Nikon is conducting joint development with another. ASM Lithography works closely with its optical supplier, which has ties to the European grower. SVG Lithography, which makes heavier use of reflective technology and has less reliance on CaF2, has been less assertive.

Ramp-up of raw material production "may require some new process development, but it appears to be primarily an investment issue," said Winkler. While producers of materials and crystals are currently not able to reliably produce materials that meet spec at acceptable yields, Winkler commented, "I believe ramp-up can still take place quickly once the decision to invest has been made and furnace designs are locked in. There are no technical show-stoppers. The uncertainty is in bringing up and maintaining yields."

The study team analyzed financial incentives for entry in the CaF2 business using an Arthur D. Little model. The model took into account the number and type of optics needed for illumination systems, projection systems, and lasers, and likely prices for each type of blank, as well as market size, growth, life cycle, and manufacturing cost. "We found that it could be a very profitable business," said Winkler. However, no other compelling commercial applications were found. -P.N.D.