12.2 Overview of Topography Simulation

Figure 12.1: Overview of the general simulation flow of topography simulations.
\begin{figure}\centering \setlength{\unitlength}{0.9mm} \begin{picture}(95,130)(...
...00){\line(0,1){70}}
\put(95, -30){\vector(-1,0){15}}
\end{picture} \end{figure}

Before immersing into the details of the simulation, a typical CVD (chemical vapor deposition) process is discussed [111]. The chemical reaction can happen in the gas phase, where it is called a homogeneous reaction, or at the boundary between gas and wafer, where it is called a heterogeneous reaction. Homogeneous reactions are not desired in CVD processes since they lead to the formation of clusters and thus to defects in the film. The heterogeneous reactions, however, yield high quality films and run selectively in areas where the temperature is high enough. During heterogeneous reactions the surface of the substrate often acts as a catalyst. Concerning CVD processes, the following chemical reactions are discerned: pyrolysis (heterogeneous decomposition reaction), reduction, oxidation, and hydrolysis.

The whole deposition or etching process, including the reactor scale processes which were not considered, can be divided into several parts.

Here the boundary layer is the small region above the wafer surface where the velocity of the flow of the gas is between zero at the surface and the velocity of the flow in the convective zone. Operating the equipment the temperature, the pressure and the amount of species entering the reactor can be adjusted. These three parameters influence the transport near the wafer surface and the surface deposition reactions which are discussed in two of the following sections.

The feature scale simulation of deposition and etching processes can be divided into three main steps. The according simulation flow of a typical topography simulation is shown in Figure 12.1.

The rest of this chapter and Chapter 13 are devoted to the simulation of these main steps. Before we go into details, we recall some chemical facts.

The coupling between gas concentrations $ c_i$ within the reactor space and the fluxes $ \Gamma_i$ towards the surface is given by the relation

$\displaystyle \Gamma_i = {\nu_i c_i\over4}, \qquad\textrm{where}\qquad \nu_i = \sqrt{\frac{8 k T}{\pi m_i}}.$ (12.1)

Here $ \nu_i$ is the average molecular velocity of gas species $ i$ with a molar mass $ m_i$.

Relating pressure, volume, and temperature, the state equation of the ideal gas is

$\displaystyle P V = n R_0 T,
$

where $ P$ denotes pressure, $ V$ volume, $ n$ the number of moles, $ R_0=8.31447\,\mathrm{J / (mol \cdot K)}$ the universal gas constant and $ T$ absolute temperature. Using $ n=N/N_A$, where $ N$ is the number of molecules and $ N_A$ the Avogadro constant, and introducing the Boltzmann constant $ k$ as $ k:=R_0/N_A$, the state equation can also be written as

$\displaystyle P V = N k T,
$

where $ k=1.38065\cdot{10}^{-23}\,\mathrm{J/K}$. In the form

$\displaystyle P = k C T$ (12.2)

the state equation relates pressure, temperature, and concentration. As the unit of pressure, $ \mathrm{Torr}$ is usually used by equipment manufacturers. In SI units this is $ 1\,\mathrm{Torr} =
133.322\,\mathrm{Pa}$.

The classic Langmuir adsorption model [111] describes how gas adsorption on a solid surface depends on pressure via the equation $ C_i=saf/(1+af)$, where $ C_i$ is the concentration of adsorbed gas molecules, $ a$ a constant, $ f$ the gas fugacity, and $ s$ the concentration of potential adsorption sites on the surface.

Clemens Heitzinger 2003-05-08