Background: The metal-oxide resist (MOR) is a promising type of nonchemically amplified resist (CAR) for EUV lithography. This family of resists shows some advantages over the conventional CARs. Even though a prior MOR model exists, no documented references for the application described could be found in open literature to develop a physically rational lithographic model that can accurately simulate and predict lithographic results for these resists.

Aim: Increase the fundamental understanding of this class of resists by creating a model that also considers these aspects of the resist.

Approach: Model the metal-oxo clusters of the MOR as nanoparticles with an effective radius that operate under excluded volume.

Results: We show the possibility to include both the packing noise effect, as well as accurate roughness (characteristics) predictions by utilizing power spectral density (PSD) plots.

Conclusions: Varying the calibrated model parameters has a clear effect on the overall roughness of the resist lines and is reflected in the PSD behavior. In contrast to experimental data, changing the resist film thickness did not result in a change in PSD behavior.