There are significant theoretical advantages to using arrays of high resolution microlenses for photolitography. Some of these derive from the fact that optical aberrations scale with lens size and are therefore small for microlenses. Microlenses are extremely reproducible, can be fabricated from fused silica producing good quality ultraviolet lenses, and their curved surfaces are atomically smooth as they are not formed by polishing. Sensors can be integrated into this optical system to improve the focus of such a system in real time. Adaptive control could allow to improve the focus of such a system in real time. Adaptive control could allow the microlens array to correct for nonplanar topography on a semiconductor wafer, for example. Using an array of microlenses, a large exposure field can be printed with uniform quality. Field flatness problems can be greatly reduced in such designs. By exploiting massive parallelism, high throughput can be achieved. Thus it is possible to eliminate photomasks for short turn-around production runs as is desirable in ASIC fab lines or multichip module fabrication. We propose to study the feasibility of such an autofocus system, towards the end of building a bench model prototype. Applications include: Manufacture of integrated circuits; manufacture of photomasks; manufacture of flat panel displays; inspection of photomasks; inspection of semiconductor wafers. Anticipated
Benefits: If successful, the project will result in an improved lithography and inspection for semiconductor wafer and photomask manufacturing. This will help to improve or maintain national competitiveness in the semiconductor and related industries.
