A Method for Optical Proximity Correction of Thermal Processes: Orthogonal Functional Method

Abstract

Pattern reduction has created a great deal of interest in finding effective methods to reduce the feature sizes of microelectronic and data-storage devices. These methods are divided between top-down approach such as photolithography and bottom-up approach such as self-assembly. For below 32 nm node technology, top-down approach has obstacles such as diffraction-limited resolution and high cost of ownership and bottom-up approach has obstacles such as the insufficient support of processes and mass production. Thermal treatment is a new process extension technique using current-day lithography equipment and chemically amplified resists (CARs). In the lithography process, thermal processes are softbake (SB), post-exposure bake (PEB), and thermal reflow. The purpose of SB is to remove excess solvent after spin coating, relieve strain in the solid film, and provide better adhesion to the substrate. The purpose of PEB is to reduce the standing wave effect and, thus, increase linewidth control and resolution. The purpose of thermal reflow is to reduce the pattern size by using thermal heating at temperatures about the glass transition temperature of the resist after development. These three kinds of thermal processes are essentially the same for heat treatment, but they have different effects on CD. Hence, it is required to understand mechanism behaviors that drive photo resist image and to deal with optical proximity effects (OPEs) due to thermal processes. OPE is quite severe as the critical dimensions (CDs) shrink down to the sub-30 nm patterns. Although the simulation parameters are not used to analyze the chemical phenomena of thermal processes, the CD bias after thermal reflow can be predicted in the linear system. Through the optical proximity correction (OPC), the distorted image from OPE can be manipulated back to the original design. However, the thermal affects of CD is the non-linear system, so that the prediction of OPE is not easy. Hence, the understanding of mechanistic behaviors that drive photo resist image can make the physically corrected resist model, and achieve the best prediction of resist images across multiple process conditions. One of the most critical issues for sub-50-nm patterning is patterning a fine contact hole (C/ H). The resolution performance of contact hole patterns is lower than that of line and space patterns because the depth of focus (DOF) for patterning contact holes is insufficient due to the low aerial-image contrast. The resist reflow process is a good method due to its simplicity without the additional complex process steps and due to its efficient technique