Prev: 1.3.2 Equipment Simulation

1.4 Outline of the Thesis

At the time when I started my work at the Institut für Mikroelektronik in December 1996, Dr. Ernst Strasser had developed a set of basic algorithms for cellular topography simulation [73]. It included a comprehensive comparison of different approaches for surface movement algorithms and the introduction of the cellular geometry representation, visibility testing algorithms, and surface advancement, as well as a set of basic models for isotropic and directional etching and deposition, reactive ion etching and so called hemispherical deposition for ballistic transport determined processes. Furthermore his thesis also included algorithmic considerations about calculation of surface normals, not implicitly given in the cellular representation, detection of unconnected regions of material as well as of voids, and derivations for handling complex ballistic transport models including reflection, redeposition and passivation layer formation.

Still many aspects required and justified the continuation of the work in order to consolidate the physical fundamentals of the modeling and to push the simulator to a level of general applicability. Surface normal, unconnected material, and void detection parts had to be extended to three-dimensional implementations of the introduced algorithms. The generation of geometries then was restricted to a few hard-coded examples and had to be extended to a process-step-based solid modeling directly including layout data. The first step ``validation'' implementation of the applied algorithms had proved to be suitable and very robust, but still showed a considerable margin and need for algorithmic optimization, and finally the models had to be extended from their rather phenomenological level to a more physically based modeling.

At the very same time a request for the simulation of ${\rm TiN}$ sputter deposition from SONY Corporation, Atsugi Technology Center, one of the industrial partners of our institute, proved the presence of industrial need for topography simulation, intensifying the enthusiasm I put into this work.

With this thesis in hand I have now covered many -- of course not all -- of the issues already formulated in Section 1.3.1 and all of the aspects stated above. In Chapter 2 the basics of the cellular approach are recapitulated and followed by the details of the algorithmic optimization achieved for the surface movement. Chapter 3 introduces the process-step-based geometry editing together with the basic topography models and is expanded to layout inclusion in Chapter 4. Etching and deposition models are continued in the next three chapters with detailed studies on resist development (Chapter 5), low-pressure processes and physical vapor deposition (PVD, Chapter 6), as well as chemical vapor deposition (Chapter 7). Furthermore the integration of simulations on reactor and feature scale for low-pressure (Section 6.1.3) as well as for high-pressure processes (Section 7.4) will be addressed in the corresponding chapters. Summarizing considerations and a view to future trends in Chapter 8 close the thesis.

The different chapters of the thesis cover a wide range of simulations from single process optimization to integrated simulation flows including layout information. The presented cellular approach shows a way to fulfill all prerequisites for performing these simulations at an appropriate level of accuracy and within reasonable CPU times with one and the same program. Therefore let's start with the explanation of the fundamental principles of the cellular structuring element algorithm.

Prev: 1.3.2 Equipment Simulation