Comparison of different approaches for the simulation of topography evolution during lithography development

Different algorithms for simulating topography evolution are compared in 2D and 3D, using rates for a lithography development process as a benchmarking example. The methods studied are the cell removal, the string, and the fast-marching algorithm. Issues considered are the convergence of the extracted critical dimensions of resist layers with increasing resolution of the simulation grid and the computation time and its dependence on the resolution. Furthermore, it is shown that a slicewise 2D simulation of topography evolution for a 3D structure is not capable of correctly representing the evolving 3D shape of the resist.