Simulation of the combined effects of polymer size, acid diffusion length, and EUV secondary electron blur on resist line-edge roughness

Abstract

Device shrinking combined with material manipulation under various process conditions becomes a difficult task if specific optimization conditions should be met. Nanolithography is limited by effects as line-edge and line-width roughness (LER and LWR respectively) and secondary electron blur (SEB). Simulation studies could show the direction of solving design for manufacturing problems. In the current article a simulation methodology is presented, based on the concept of stochastic modeling of exposure, material, and process aspects of lithography and pattern transfer with plasma etching in order to get information about the evolution of critical dimensions (CD), LER and LWR in the layout. The study reveals that under certain process conditions, the effect of acid diffusion on LER is more important than the one of SEB, although both deal with blurring, because acid diffusion is supposed to extend in longer radius. However, when resists of low degree of polymerization are used, SEB should also be considered explicitly since the deteriorations from blurring on both LER and CD are enhanced due to the graining nature of the material. In any case, etching smoothing effects of high frequency LER components should be considered in terms of CD loss.

Line edge roughness transfer during plasma etching: modeling approaches and comparison with experimental results

“…The etching of the resist sidewall by the trimming process is approximated by an isotropic etching process: The resist profile or surface is moving under an isotropic etching rate [26][27][28][29][30] . The level set method, which is based on the notion of the implicit functions, is used for the evolution of the surfaces 31,32 .…”