(360ba) Modeling of Photoresist Pattern Formation in EUV Lithography Through Molecular Dynamics Simulations

As competition for micro-processes of semiconductor circuits intensifies, there is a growing need for securing a photoresist suitable for an ultra-fine line width and technology for reduction of RLS trade-off. In recent years, a polymer that introduces acid cleavable moiety (ACM) that is decomposed by reacting with acid has been actively developed. Under ultra-fine line width conditions, the radius of gyration (Rg) of each polymer chain has a size similar to that of the line width. Therefore, it is necessary to control the length of the polymer chain at the exposed/unexposed interface in consideration of chain decomposition. Thus, we explore ACM introduction strategies that can reduce the line edge roughness of the pattern while maintaining other performance properties. First of all, the Coarse-grained molecular dynamics (CGMD) simulation platform is established to precisely control various parameters such as the position of the chain scission in the mesoscopic level. This model can demonstrate the molecular weight distribution of the interface according to the presence or absence of ACM and calculate structural homogeneity of residual polymer interfaces after the process. After the principle and feasibility of the chain scission strategy are verified through CGMD simulation, all-atomic molecular dynamics (AAMD) simulation can be performed to confirm the specific interaction between the polymer resin and the developer to be modeled. An atomistic structure of the interface is back-mapped from the CGMD structure. This model verifies the interaction parameters between polymer-solvent, and the change of solubility according to the properties of ACM. Through this, we explore the optimal polymer-solvent candidate group. This approach guides further optimization to design a high-performance photoresist that overcomes the RLS trade-off by utilizing the polymer main chain scission and we anticipate that It can produce large-scale data by establishing an efficient simulation progress platform.