KMPR Photoresist Process Optimization Using Factorial Experimental Design

Abstract

MicroChem Corp. and Nippon Kayaku Co., Ltd. have jointly developed a new, photoimageable coating composition for use in fabricating a wide variety of micro-electromechanical systems (MEMS) and optical MEMS. This chemically amplified thick-film photoresist, known as KMPR, has many of the same high-aspect ratio photo imaging properties as solvent developable SU-8, but can be developed in a conventional aqueous alkaline developer (TMAH) and readily stripped from the wafer.Single-factor experimental methods rarely find the optimum performance of a photoimageable coating efficiently and often fail to detect the interactions so common in complicated photochemical systems. By combining the settings of several factors simultaneously in an array (experimental design), it is possible to separate out the main effects and their interactions.1This paper presents the optimization of KMPR for contact alignment and near diffraction-limited, thick film reduction stepper applications. A Sequence of Taguchi style, half-factorial experimental designs was used to evaluate the process parameters for a 5 and 50 Mm thick film exposed on a contact aligner followed by a third half-factorial design for a 10 Mm thick film of KMPR photoresist exposed on a 365 nm, 0.40 N.A. stepper. Independent factors, such as: soft bake time, focus, exposure dose, initiator activation time (post-exposure bake time), developer time and delay time were evaluated