next up previous
Next: 5.3.2 The VISTA Simulation Up: 5.3 Optimization Previous: 5.3 Optimization

5.3.1 Process Simulation

The input files for the device simulation were generated by simulation of the complete manufacturing process using process simulators for deposition, implantation, diffusion, and etching. This was necessary to find the optimum thickness of the deposited epi-layers and to investigate the effects of repeated diffusion steps after deposition and implantation for each layer and to create as realistic input files as possible.

Table 5.1 shows the process flow and data for the modified device structure. Deposition and stripping steps were performed by the simulator sketch. The simulator animpl was used for implantation steps. For diffusion simulation TSuprem-4 was taken.

The sequence of simulation steps is not exactly the same as the sequence of process steps in reality in order to reduce demand on computational resources. To keep the simulation domain as small as possible and to reduce the number of grid points for the implantation and diffusion steps the homogeneously doped substrate was added after the last epi-layer has been processed. This was necessary because the version of TSuprem-4 used for simulation of diffusion steps did not allow more than 20000 grid points.

During each deposition of an epi-layer heating of the underneath epi-layers occurs. To account for the effects of this heating an additional diffusion step has been introduced into the simulation flow after the the first epitaxial deposition for each layer except the first.


Table 5.1: Process flow for the modified device structure
# SFC specification
1 substrate P, 1.5 . 1015 cm- 3
2 deposition, n-epi standard P, 4.15 . 1015 cm- 3, thickness 5.2 $ \mu$m
layer 1    
3 deposition, n-epi P, 2.8 . 1013 cm- 3, thickness $ \mu$m
4 deposition, n-bubble mask  
5 implantation, n-bubble P, 2 . 1012 cm- 2, 120 keV
6 strip, n-bubble mask  
7 deposition, p-bubble mask  
8 implantation, p-bubble B, 2 . 1012 cm- 2, 80 keV
9 strip, p-bubble mask  
10 diffusion 1150o C, 1 min
layer 2    
11 deposition, n-epi P, 2.8 . 1013 cm- 3, thickness $ \mu$m
12 diffusion, deposition heating 1110o C, 1 min
13 deposition, n-epi P, 2.8 . 1013 cm- 3, thickness $ \mu$m
14 deposition, n-bubble mask  
15 implantation, n-bubble P, 2 . 1012 cm- 2, 120 keV
16 strip, n-bubble mask  
17 deposition, p-bubble mask  
18 implantation, p-bubble B, 2 . 1012 cm- 2, 80 keV
19 strip, p-bubble mask  
20 diffusion 1150o C, 1 min
layer 3    
21 deposition, n-epi P, 2.8 . 1013 cm- 3, thickness $ \mu$m
22 diffusion, deposition heating 1110o C, 1 min
23 deposition, n-epi P, 2.8 . 1013 cm- 3, thickness $ \mu$m
24 deposition, n-bubble mask  
25 implantation, n-bubble P, 2 . 1012 cm- 2, 120 keV
26 strip, n-bubble mask  
27 deposition, p-bubble mask  
28 implantation, p-bubble B, 2 . 1012 cm- 2, 80 keV
29 strip, p-bubble mask  
30 diffusion 1150o C, 1 min
layer 4    
31 deposition, n-epi P, 2.8 . 1013 cm- 3, thickness $ \mu$m
32 diffusion, deposition heating 1110o C, 1 min
33 deposition, n-epi P, 2.8 . 1013 cm- 3, thickness $ \mu$m
34 deposition, n-bubble mask  
35 implantation, n-bubble P, 2 . 1012 cm- 2, 120 keV
36 strip, n-bubble mask  
37 deposition, p-bubble mask  
38 implantation, p-bubble B, 2 . 1012 cm- 2, 80 keV
39 strip, p-bubble mask  
40 diffusion 1150o C, 1 min
layer 5    
41 deposition, n-epi P, 2.8 . 1013 cm- 3, thickness $ \mu$m
42 diffusion, deposition heating 1110o C, 1 min
43 deposition, n-epi P, 2.8 . 1013 cm- 3, thickness $ \mu$m
44 deposition, n-bubble mask  
45 implantation, n-bubble P, 8 . 1011 cm- 2, 120 keV
46 strip, n-bubble mask  
47 deposition, oxide SiO2, thickness 0.85 $ \mu$m
48 deposition mask  
49 implantation, p- B, 1 . 1013 cm- 2, 80 keV
50 strip mask  
51 deposition mask  
52 implantation, p+ B, 5 . 1014 cm- 2, 45 keV
53 strip mask  
54 strip oxide  
55 oxidation, gate oxide thickness 50 nm
56 p-diffusion 1100o C, 200 min
    1100o C $ \rightarrow$ 800o C, 150 min
57 deposition, gate poly SiO2, P, 1 . 1015 cm- 3, thickness 0.5 $ \mu$m
58 deposition, source doping mask  
59 implantation, source doping As, 5 . 1015 cm- 2, 120 keV
60 deposition SiO2, thickness $ \mu$m
61 As diffusion 800o C $ \rightarrow$ 1000o C, 20 min
    1000o C $ \rightarrow$ 800o C, 100 min
62 deposition, source contact mask  
63 etch, recess  
64 deposition, source contact Al, thickness $ \mu$m

Figure 5.8: Concentration of the phosphorus doping of the modified device.
\resizebox{12cm}{!}{
\includegraphics[width=12cm,angle=90]{eps/dmos-np-100V-phosphorus.col.ps}}

Figure 5.9: Concentration of the boron doping of the modified device.
\resizebox{12cm}{!}{
\includegraphics[width=12cm,angle=90]{eps/dmos-np-100V-boron.col.ps}}

Fig. 5.8 and Fig. 5.9 show the geometry of the simulated device and the resulting phosphorus and boron concentrations, respectively, of the modified device. The center of the additional vertical n-doping is located at x = 15 $ \mu$m as can be seen in Fig. 5.8. Because of the phosphorus implantation and diffusion in each epi-layer the doping concentration in the vertical direction in the additional doping is not constant. Small variations of the doping concentration in vertical direction are visible in Fig. 5.8 and the epi-layer structure consisting of 5 layers is clearly discernible.

The boron channel doping extends in lateral direction from the source contact to the additional vertical p-doping at the gate contact edge. The center of the additional vertical p-doping is located at x = 11 $ \mu$m. Again the epi-layer structure is clearly visible from the small doping concentration variations in vertical direction.


next up previous
Next: 5.3.2 The VISTA Simulation Up: 5.3 Optimization Previous: 5.3 Optimization
Martin Rottinger
1999-05-31