Manufacturing CaF2 Lenses
Ruth DeJule, Associate Editor -- Semiconductor International, 2/1/1999
C alcium fluoride (CaF2) optical components for ArF 193 nm and F2 157 nm lithography must be manufactured from high-purity grade synthetic crystalline material that is free from trace impurities that lead to high absorption, fluorescence and/or color center formation. The most common technique for the growth of CaF 2 is the Bridgeman or Stockbarger method. Unlike "dipping" a seeded crystal as in silicon growth, a mechanical elevator system slowly lowers molten CaF2 from the hottest part of a vacuum furnace into a lower temperature region, where the molten material solidifies into a crystal. Once the crystal is grown, it is slow-cooled to room temperature to minimize strain. The entire process can take from a few to several weeks.
Yields are the main manufacturing concern for CaF2 , said Dennis Cope, business manager at Bicron (Solon, Ohio). It is an inherently difficult material to process. Hygroscopic, soft and prone to chipping, CaF2 fabrication time is typically 50% to 100% longer than for glass. So while the size and specifications for CaF2 projection optics have been achieved, improved process control is needed in the production of these crystals. Currently, yields obtained are below what Bicron has achieved in its other growth operations, but Cope believes that yields 2-3X higher than current levels can be attained. Parallel efforts in growing larger diameter crystals are also expected to boost supply.
|
|
|
DUV optical components coated with anti-reflective coatings for 157 and 193 nm optimize transmission performance. |
The formation of CaF2 DUV optics such as windows, lenses, beamsplitters and prisms require state-of-the-art polishing techniques that produce surfaces typically having 1/10 wave flatness at 193 nm with <20/10 scratch-dig surface quality. Because CaF 2 is a fairly soft crystal, polishing techniques that eliminate surface sleeks or fine scratches are necessary to decrease surface roughness, which can lead to scattering, subsurface damage and microscopic flaws in a coated surface.
To optimize lens transmission performance of many DUV optics, anti-reflective coatings are necessary. Typically coated on one or both sides of a lens, an AR coating reduces per surface reflection losses, and hence increases DUV transmission. Coatings at 193 and 157 nm, however, frequently consist of complex multilayers rather than traditional low-index, one-quarter wave designs. Coatings, such as narrow "V," have low reflectance properties over a limited DUV wavelength region, depending on the refractive index (n) of the material being coated and the coating materials used. AR coatings designed for both 157 and 193 nm typically reduce reflectance from 4% to 5% to <0.2% to 0.5%.
In addition to enhancing transmission performance, AR coatings also serve to protect the optical material. At 193 nm, certain grades of fused silica substrates can significantly decrease the optic's lifetime to <400 million pulses due to degradation of the fused silica substrate. Coatings developed by Acton Research (Acton, Mass.) for 193 nm optics designed for normal incidence use on CaF2 have withstood one billion pulses at 15 mJ/cm2 at 400 Hz continuously without any significant change in optical performance and only a slight discoloration of the coating, said Glen Callahan, optics sales manager.
With growing interest in 157 nm lithography, companies such as Acton Research are developing specialty AR coatings. Thus far, they have successfully demonstrated coatings for 157 nm optics, which have withstood exposure to 50 million pulses at an average of 3 mJ/cm2 . Long-term testing is planned for the near future.
Research efforts such as those at MIT's Lincoln Laboratory (Lexington, Mass.) led by Mordechai Rothschild and Vladimir Liberman and funded by International SEMATECH (San Jose, Calif.) endeavor to improve testing of excimer-grade CaF2 optics under actual working conditions. The most important conclusion drawn from the testing to date is that initial material measurements do not necessarily correlate with extended-use performance, creating concern and the need for establishing long-term characteristics.
CaF2 technology has advanced rapidly
over the past year as improved purification methods, growth control and
annealing technologies have made significant strides. However,
significant issues must be resolved before CaF2
becomes a production-worthy technology.