previous up next contents
Prev: 5.3.1 Lithography Modeling Up: 5.3 Coupling with Lithography Next: 6. Low-Pressure Etching and


5.3.2 Contact Hole Printing

The following example is an extract from a comprehensive study of DUV printing of a single contact hole over planar and stepped topographies. For three different geometries, three types of illumination, namely, coherent illumination, partially coherent illumination, and quadrupole illumination have been simulated for best focus and defocus situations. In the doctoral thesis of Heinrich Kirchauer [31] the complete description of the aerial image simulation and the differences for the various illumination techniques is given together with the comprehensive illustration of the 18 combinations for the different geometries, illumination types and focus situations. Furthermore the influence of the different illumination schemes on the final resist profile is discussed in detail by means of rigorous exposure simulation. Within the scope of topography simulation, it is sufficient to demonstrate the basic applicability of the development simulation for the extraction of resist profiles, without a detailed differentiation of the implications of the various illumination techniques. Therefore only the results for coherent illumination with best focus conditions will be shown for three different topographies.

The simulation domain for the examples was set to 1.0$\mu\mathrm m$$\times$1.0$\mu\mathrm m$$\times$0.7$\mu\mathrm m$. A quadratically shaped mask with side length of 0.25$\mu\mathrm m$ was centered over a planar silicon substrate with a refractive index of $n_{\rm Si}=1.68 + j3.58$ (see Fig. 5.6, left), a dielectric oxide step with $n_{\rm SiO_2}=1.508$ (middle), and a reflective amorphous silicon step with $n_{\rm a-Si}=1.69 + j2.76$ (right). The stepper wavelength was 248nm and a non-bleaching DUV resist was chosen with $n_{\rm Resist}=1.65 + j0.02$. The numerical aperture $\mathit{NA}$ was fixed to 0.5.

Figure 5.6: Exposure and development simulation over a planar substrate (left), a dielectric step (middle), and a reflective step (right).
\begin{figure}\psfrag{xlab}[][][0.7]{1.0 $\mu$m}
\psfrag{ylab}[][][0.7]{0.5 $\mu...
...s[width=0.32\textwidth]{eps-dev/REdevelReflCohe.eps}
}}
\end{center}\end{figure}

A comparison of the simulations shown in Fig. 5.6 exhibits a smaller opening in the developed photo-resist for the stepped topographies. The effective diameter of the contact hole depends on the non-planarity of the wafer topography. The dependence is stronger for the reflective amorphous silicon step than for the dielectric oxide step. The standing waves evolve regularly above the planar substrate and get distorted by the reflections over the steps. The resist profiles are more conformal along the reflective silicon step in comparison with the dielectric step.

The complexity of the resist profiles evolving from the integrated lithography simulation is comparable with the benchmark examples from above. The lithography examples show the same formation of standing waves and account for reflections on the step. The evolving profiles are strongly irregular, which is especially pronounced for the reflective silicon step, and make high demands on the robustness of the topography simulator. Together with the benchmark examples, the integrated lithography/development simulations underline the excellent performance of the cellular approach for such complex and irregular structures.

previous up next contents
Prev: 5.3.1 Lithography Modeling Up: 5.3 Coupling with Lithography Next: 6. Low-Pressure Etching and


W. Pyka: Feature Scale Modeling for Etching and Deposition Processes in Semiconductor Manufacturing