High-aspect-ratio photolithography for MEMS applications

High-aspect-ratio photolithography using a commercially available positive photoresist and a conventional contact mask aligner with standard UV light source is described, A multiple coating process is developed to obtain a photoresist thickness up to 23 mu m while maintaining a smooth photoresist surface, Intimate contact between the mask and wafer is found to be most critical for high-resolution photolithography, Vacuum contact is found to be well-suited for this purpose. Additionally, edge bead removal is found to be of significant importance for intimate contact between the mask and the substrate, Prebake, exposure, and development conditions are optimized for resolution and aspect ratio, Maximum prebake temperature still allowing the photoresist to be developed is found to be the optimal temperature for obtaining high resolution, Prebake time distribution is optimized for avoiding residual stress in the photoresist, as well as maintaining high resolution, when multiple coating is applied, Minimum linewidth and spacing of 3.5 mu m and 2.5 mu m, respectively, and a maximum aspect ratio of 7.7 have been achieved in a photoresist thickness of 23 mu m. Postbake improves the chemical resistance to subsequent processes, for example, electroless nickel plating using the photoresist as a mold. However, postbake also causes pattern distortion, which can be severe at times. Therefore, optimal process and design conditions for minimizing the pattern distortion have been studied, In addition to the basic process conditions, a contrast enhancement material (GEM) is investigated for improving pattern resolution and found to be effective, especially when photolithography is performed on a wafer with topographical steps, Lateral resonant comb drive actuators and polygon mirror microscanners have been fabricated using this photolithography process in conjunction with electroless nickel plating. The photoresist used in this study is to be discontinued because of its health hazard, and a substitute photoresist will be provided by the same vendor. We have performed a preliminary test of this photoresist and have found that our finding also applies to the new photoresist, although the process conditions for the new resist have not been optimized. [130]