In this section we want to show how to obtain faster simulation than the previous section. First of all we simulate the MESFET device by means of the simplified MEP model. This is done by the following script
# Silicon MESFET test-1 # created on 27 feb.2005, J.M.Sellier # modified on 05 mar.2005, J.M.Sellier # Simplified MEP model MATERIAL SILICON TRANSPORT MEP ELECTRONS FINALTIME 5.0e-12 TIMESTEP 0.001e-12 XLENGTH 0.6e-6 YLENGTH 0.2e-6 XSPATIALSTEP 90 YSPATIALSTEP 20 # Energy relaxation time # ====================== TAUW 0.4e-12 # Definition of the doping concentration # ====================================== DONORDENSITY 0. 0. 0.6e-6 0.2e-6 1.e23 DONORDENSITY 0. 0.15e-6 0.1e-6 0.2e-6 3.e23 DONORDENSITY 0.5e-6 0.15e-6 0.6e-6 0.2e-6 3.e23 ACCEPTORDENSITY 0. 0. 0.6e-6 0.2e-6 1.e20 # ACCEPTORDENSITY 0. 0. 0.6e-6 0.2e-6 1.e23 # ACCEPTORDENSITY 0. 0.15e-6 0.1e-6 0.2e-6 3.e23 # ACCEPTORDENSITY 0.5e-6 0.15e-6 0.6e-6 0.2e-6 3.e23 # Definition of the various contacts # ================================== CONTACT DOWN 0.0 0.6e-6 INSULATOR 0.0 CONTACT LEFT 0.0 0.2e-6 INSULATOR 0.0 CONTACT RIGHT 0.0 0.2e-6 INSULATOR 0.0 CONTACT UP 0.1e-6 0.2e-6 INSULATOR 0.0 CONTACT UP 0.4e-6 0.5e-6 INSULATOR 0.0 CONTACT UP 0.0 0.1e-6 OHMIC 0.0 3.e23 1.e20 CONTACT UP 0.2e-6 0.4e-6 SCHOTTKY -0.8 CONTACT UP 0.5e-6 0.6e-6 OHMIC 1.0 3.e23 1.e20 NOQUANTUMEFFECTS MAXIMINI # SAVEEACHSTEP LATTICETEMPERATURE 300. STATISTICALWEIGHT 100 MEDIA 500 OUTPUTFORMAT GNUPLOT # end of MESFET test-1
Then we copy the density, energy and potential files with the following names
density_start.xyz energy_start.xyz potential_start.xyzand run the following script
# Silicon MESFET test-1 # created on 27 feb.2005, J.M.Sellier # modified on 05 mar.2005, J.M.Sellier # Fast Monte Carlo method MATERIAL SILICON TRANSPORT MC ELECTRONS FINALTIME 1.e-12 TIMESTEP 0.001e-12 XLENGTH 0.6e-6 YLENGTH 0.2e-6 XSPATIALSTEP 90 YSPATIALSTEP 20 # Energy relaxation time # ====================== TAUW 0.4e-12 # Definition of the doping concentration # ====================================== DONORDENSITY 0. 0. 0.6e-6 0.2e-6 1.e23 DONORDENSITY 0. 0.15e-6 0.1e-6 0.2e-6 3.e23 DONORDENSITY 0.5e-6 0.15e-6 0.6e-6 0.2e-6 3.e23 ACCEPTORDENSITY 0. 0. 0.6e-6 0.2e-6 1.e20 # ACCEPTORDENSITY 0. 0. 0.6e-6 0.2e-6 1.e23 # ACCEPTORDENSITY 0. 0.15e-6 0.1e-6 0.2e-6 3.e23 # ACCEPTORDENSITY 0.5e-6 0.15e-6 0.6e-6 0.2e-6 3.e23 # Definition of the various contacts # ================================== CONTACT DOWN 0.0 0.6e-6 INSULATOR 0.0 CONTACT LEFT 0.0 0.2e-6 INSULATOR 0.0 CONTACT RIGHT 0.0 0.2e-6 INSULATOR 0.0 CONTACT UP 0.1e-6 0.2e-6 INSULATOR 0.0 CONTACT UP 0.4e-6 0.5e-6 INSULATOR 0.0 CONTACT UP 0.0 0.1e-6 OHMIC 0.0 3.e23 1.e20 CONTACT UP 0.2e-6 0.4e-6 SCHOTTKY -0.8 CONTACT UP 0.5e-6 0.6e-6 OHMIC 1.0 3.e23 1.e20 NOQUANTUMEFFECTS MAXIMINI # SAVEEACHSTEP # Load Electron Initial Data LEID LATTICETEMPERATURE 300. STATISTICALWEIGHT 100 MEDIA 500 OUTPUTFORMAT GNUPLOT # end of MESFET test-1
As the reader can note from the following script, we only need picosecond for the Monte Carlo method. This is done because we load the initial conditions from the previous MEP simulation. This is a very fast way to obtain fast Monte Carlo simulations. We report, in the following some interesting results.