Estimation of diffusion lengths of acid and quencher in chemically amplified resist on the basis of extreme ultraviolet exposure results

Line- and space-type patterns with a half pitch (hp) of 32–65nm were printed with the extreme ultraviolet microexposure tool at the Lawrence Berkeley National Laboratory using the positive-tone chemically amplified resist MET-1K, and the diffusion lengths of the acid and quencher in the resist were estimated by fitting the calculated shapes of resist patterns to the experimental results. Simulations of the resist process employed an acid/quencher mutual-diffusion model. It was found that not only acid diffusion but also quencher diffusion had a considerable effect on the fidelity of hp 32–45nm patterns. When the diffusion lengths were assumed to be 40nm for the acid and 30nm for the quencher, the calculated shapes of resist patterns agreed well with the experimental results for both line- and space-type patterns. If quencher diffusion were not considered, the acid diffusion length would be underestimated, which would make it difficult to explain the shapes of various types of resist patterns using the same acid diffusion length. In order to fabricate various types of hp 32nm patterns, the diffusion lengths of both the acid and the quencher should be less than 10nm.