next up previous
Next: 5.2.2 Device Simulation Up: 5.2 Application Example Previous: 5.2 Application Example


5.2.1 A Low-Voltage, Low-Power Technology

To demonstrate the abilities of the described tools, they were used to develope an advanced CMOS process. It is based on an twin well technology with trench isolation and a $0.5 \mu m$ transistor's channel length. Careful channel engineering allows to reduce short channel effects and a TiN barrier layer is used to make a reliable contact to the shallow source/drain junctions.

Only two metal layers are simulated, but this number could be extended typically to six. The vias use a tungsten pad layer. The not shown metal levels could be formed in the same way as METAL2, but a CMP process step should be done to smooth the oxide surface prior to metal deposition. This would usually be necessary in the lowest ones, as normally the design rules of the upper layers are quite large (they are used as power distribution).

In Figure 5.3 the simulation flow controller for hierarchical and user-friendly process recipe edition is displayed. The simulated circuit is a 5-stage ring oscillator the layout of which is shown in Figure 5.15. For the two-dimensional process simulations a cut over the last inverter was performed (see also Figure 5.15). This circuit though simple, contains the most important elements in a CMOS integrated circuit: Both PMOS and NMOS transistors and the isolation structure between them.

Figure 5.3: The simulation flow controller: Specification the use of layout layer P-WELL as mask for the P-well high energy implant.
\begin{figure}
\centerline{\epsfig{file=DSflowController.eps,width=0.85\linewidth}}
\end{figure}

The most relevant intermediate results from the simulation flow are shown in Figures 5.4 to 5.11. Note that the x-direction in these results is not the same as in the layout of Figure 5.15. The x-direction in the two-dimensional simulations has not layout based coordinates, but refers to distances to the first point of the cut-line. So whatever the direction of the cut, the results will be in the x-axis for mask ticks and y-axis for depth (direction perpendicular to the wafer plane).

Figure 5.4: Substrate material: p-type (100)-oriented silicon (the picture also shows a 40nm screen oxide layer formed afterwards).
\begin{figure}
\vspace{0.35cm}
\centerline{\epsfig{file=DStechP0geom.eps,width=0.70\linewidth}}\end{figure}

Figure 5.5: N-well and P-well implants after annealing.
\begin{figure}
\vspace{0.15cm}
%
\centerline{\epsfig{file=DStechP1dopCol.eps,width=0.77\linewidth}}
\vspace{-0.25cm}
\end{figure}

Figure 5.6: Net-doping profile after threshold adjustment implants (the simulation grid is also shown).
\begin{figure}
\vspace{0.1cm}
%
\centerline{\epsfig{file=DStechP1cdopCol.eps,width=0.82\linewidth}}\end{figure}

Figure 5.7: Net-doping profile after trench etch with reactive ion-etching and channel-stop implants.
\begin{figure}
\vspace{0.25cm}
%
\centerline{\epsfig{file=DStechP2dopCol.eps,width=0.845\linewidth}}\end{figure}

Figure 5.8: Geometry after gate oxidation, gate polysilicon deposition and sidewall spacer formation.
\begin{figure}
\vspace{0.1cm}
\centerline{\epsfig{file=DStechP3geom.eps,width=0.90\linewidth}}
\vspace{-0.1cm}
\end{figure}

Figure 5.9: Net-doping profile after source/drain implants.
\begin{figure}
\vspace{0.5cm}
%
\centerline{\epsfig{file=DStechP4dopCol.eps,width=0.94\linewidth}}
\vspace{-0.21cm}
\end{figure}

Figure 5.10: Geometry after oxidation, contact hole etch, first metal deposition and etching
\begin{figure}
\vspace{0.1cm}
\centerline{\epsfig{file=DStechP5geom.eps,width=0.75\linewidth}}
\vspace{-0.3cm}
\end{figure}

Figure 5.11: Structure (final) after intermetal oxide deposition, via hole etch, second metal deposition, second metal etching and passivation.
\begin{figure}
\vspace{0.5cm}
\centerline{\epsfig{file=DStechP6geom.eps,width=0.97\linewidth}}
\vspace{-0.05cm}
\end{figure}


next up previous
Next: 5.2.2 Device Simulation Up: 5.2 Application Example Previous: 5.2 Application Example
Rui Martins
1999-02-24