Intel is prime mover to fund EUV litho project

04/01/1997

Intel is prime moverto fund EUV litho project

A newly formed private corporation, led by Intel, has been formed to oversee and fund cooperative development of extreme UV (EUV) lithography for the 0.13-micron generation of ICs. Known as the EUV Limited Liability Corp., the company hopes to have an alpha-level machine built within 36 months; process development could then begin about five years from now.

The bold move is indicative of the deep concern over litho availability felt by aggressive chipmakers such as Intel and Samsung. While today`s leading-edge 248-nm tools will handle quarter-micron work and probably 0.18-micron as well, there are no clear winners beyond that point and all the candidates (193 nm, x-ray, e-beam, EUV) have major weaknesses.

Plans are still being finalized for the enterprise, said Chuck Gwyn, a former Sandia National Laboratory researcher who is now working from an Intel office as program director for the EUV LLC. However, the general scheme involves funding and project management by the EUV LLC, with development work being done by multiple lithography suppliers and a "virtual national laboratory" composed of researchers from Sandia and the Lawrence Livermore and Lawrence Berkeley national labs. Work on EUV, which uses 13-nm radiation and multicoated resonant reflective optics, has been underway for several years at the Integrated Manufacturing Technology Center at Sandia`s Livermore, CA, unit.

"The concept is that by working together, in a very tightly coupled way, we will realize an alpha machine much faster than if we were working separately," said one source close to the project. "We`re still looking at all the options, but the EUV LLC is negotiating business agreements with the stepper companies. The EUV LLC represents the semiconductor manufacturers, who are ultimately the customers for the virtual national lab and the stepper companies." No information was available on what other firms are participating.

All funding for the project will be private, coming largely from semiconductor makers who join the EUV LLC. No firm budget numbers were available, but estimates of the cost of developing a future-generation litho system (including exposure tool, masks, resists, and illumination source) run well into the hundreds of millions of dollars. A key issue for EUV is optics: in order to reflect the 13-nm radiation while reducing the circuit pattern image, aspheric surfaces must be coated with dozens of layers of material, with roughness controlled to within a few angstroms. Because the optics tend to be inefficient, and multiple reflections are needed to reduce the circuit pattern`s image, any EUV resist will have to be very fast. Masks will use the same reflective technology as the optics, and will require a new paradigm for maskmaking.

If EUV can, in fact, be ready for the 0.13-micron generation, it would cast a significant pall over the future of 193-nm lithography, which has generally been perceived as the successor to 248 nm. Intel`s Joe Langston recently stated that he doubts 193 will be ready in time for Intel`s 0.18-micron process development, and there would be little incentive to try and eke out 0.13-micron production from 193 nm if EUV were ready. However, taking EUV out of the science project stage and into fabs is a decidedly nontrivial job. Among optical engineers, there is no consensus that the EUV scheme is achievable, let alone economically viable. - P.N.D.