1.4 Simulator Implementation

In order to simulate topography modifications a method which is able to describe geometric deformations over time is required. The initial topography surface $ \mathcal{S}\left(t=0\right)$ is extracted from the desired geometry and is known to the simulator. After a specific time for processing $ t_{process}$, the surface is modified under surface velocities $ V\left(\vec{x}\right)$ acting in the surface normal direction for all points $ \vec{x}\in\mathcal{S}\left(t\right)$. With the initial surface, surface velocities, and processing time known, the final surface $ \mathcal{S}\left(t=t_{process}\right)$ can be found. It was shown by Sethian [165], [185] that this type of topography problem for microelectronics simulations can be solved using the LS method. A LS simulator for traditional semiconductor processes was subsequently created at the Institute for Microelectronics, TU Wien, by Ertl [50]. This simulator serves as the basis for the simulations and models presented in this document. All models are incorporated within the original simulator, allowing for these techniques to be used together with traditional processing techniques in subsequent steps in order to generate a desired device structure.

In this section, a summary of the topography simulator which is based on the LS framework from [50] will be described. For a more detailed description of the LS functions, refer to [50]. This section serves to summarize the implemented simulator by first introducing the LS framework, including the sparse field method for the surface description and the HRLE data structure. Subsequently, the calculations of the surface rate and velocity field required for topography movement are explained.


L. Filipovic: Topography Simulation of Novel Processing Techniques