Analysis of line edge roughness using probability process model for chemically amplified resists

Line edge roughness (LER) in chemically amplified resists (CARs) is understood to be fluctuations in the acid catalyzed reaction that determine molecular solubility and developer percolation. Two probability processes that cause LER are modeled: one is local acid generation with diffusion process, and the other is the main reaction and developer percolation process. A typical fluctuation size in these processes is significantly larger than the molecular size and depends on various parameters, such as acid concentration, diffusion length, molecular size, protection ratio and variation, and image conditions. Careful scaling for these parameters is required to reduce LER with certain shrinking design feature sizes. Our results suggest various factors that dominate LER in several resist systems. Reliable metrology for LER requires a sufficient sample size due to the random nature of origin. High contrast exposure characteristics of the dissolution rate in CARs today are also explained with the model.