Designers of semiconductor lithography equipment have made some impressive strides over the last few years. But soon, they're going to be victims of their own success. Scientists at the National Institute of Standards and Technology in Gaithersburg, Md., say that designers will face some challenges as they use calcium-fluoride lenses at lithography scales under 157 nm.
Advances in lithography drive the Moore's Law phenomenon, where chip complexity doubles every 18 months or so. Current state-of-the-art production processes use deep ultraviolet lasers at 248 and 193 nm to image circuits with critical features as small as 130 to 150 nm. To jump to the next level, these designers will need to develop imaging systems using 157-nm light to achieve 70-nm feature sizes.
All of the optics involved in 157-nm lithography are made from calcium fluoride, as it is one of the only materials that is transparent at this scale. NIST physicist John Burnett, though, recently showed that calcium fluoride is inherently bifringent in the deep ultraviolet. Its crystal refracts light differently, depending on the polarization of the light. In other words, the material can't keep up with the industry's innovation.
As a result, designers will have to account for this intrinsic bifringence in their work. Calcium-fluoride lenses will not focus properly at this scale without careful control of the light as it enters the lens. This can be quite a challenge, as most "stepper" or optical lithography lenses contain about 20 such lens elements.
Modeling software used to design such systems is being modified to account for these effects, but it won't be ready until October. Meanwhile, NIST is investigating the use of mixed crystals to compensate for the bifringence effect.
For details, point your browser to www.nist.gov.
Designers of semiconductor lithography equipment have made some impressive strides over the last few years. But soon, they're going to be victims of their own success. Scientists at the National Institute of Standards and Technology in Gaithersburg, Md., say that designers will face some challenges as they use calcium-fluoride lenses at lithography scales under 157 nm.
Advances in lithography drive the Moore's Law phenomenon, where chip complexity doubles every 18 months or so. Current state-of-the-art production processes use deep ultraviolet lasers at 248 and 193 nm to image circuits with critical features as small as 130 to 150 nm. To jump to the next level, these designers will need to develop imaging systems using 157-nm light to achieve 70-nm feature sizes.
All of the optics involved in 157-nm lithography are made from calcium fluoride, as it is one of the only materials that is transparent at this scale. NIST physicist John Burnett, though, recently showed that calcium fluoride is inherently bifringent in the deep ultraviolet. Its crystal refracts light differently, depending on the polarization of the light. In other words, the material can't keep up with the industry's innovation.
As a result, designers will have to account for this intrinsic bifringence in their work. Calcium-fluoride lenses will not focus properly at this scale without careful control of the light as it enters the lens. This can be quite a challenge, as most "stepper" or optical lithography lenses contain about 20 such lens elements.
Modeling software used to design such systems is being modified to account for these effects, but it won't be ready until October. Meanwhile, NIST is investigating the use of mixed crystals to compensate for the bifringence effect.
For details, point your browser to www.nist.gov.