Improvement of feature-scale profile evolution in a silicon dioxide plasma etching simulator using the level set method

Abstract

We present a three-dimensional simulator of silicon dioxide etching in a fluorocarbon plasma process. Explicit parametrization of the surface is currently one of the most frequently used methods to evolve the etched surface according to the equipment simulation results. These techniques update the coordinates of the vertices and need to add and/or remove faces to keep an accurate surface representation. These processes can introduce errors and produce unrealistic results, especially in complex structures. In this paper we prove the effectiveness of our level set (LS) implementation to evolve the etched surface according to etching models, resulting in a fully operational plasma etching simulator. The LS implementation is based on a surface reconstruction algorithm from scattered points enabling the simulation of complex topological changes such as coalescing or splitting of contiguous regions. Additionally, our algorithm is based on the sparse field method for reducing computational time of the surface evolution process and it is perfectly suited to be used with the Anetch software package. Finally, several structures are simulated and an experimental result is used to compare and validate the effectiveness of the simulator we have developed.