Author Affiliations
Silicon Hall: Micro/Nano Manufacturing Facility, Faculty of Engineering and Applied Science, Ontario Tech University, 2000 Simcoe St N, Oshawa, Ontario L1G 0C5, Canadashow less
Fig. 1. Schematic of fabrication set-up. Figure reproduced with permission from ref.
6, Elsevier.
Fig. 2. Filtered reflectance data.
Fig. 3. Filtered transmittance data.
Fig. 4. Model validation with experimental transmittance.
Fig. 5. Refractive index as a function of wavelength.
Fig. 6. Extinction coefficient as a function of wavelength.
Fig. 7. PUMA model validation with analytical refractive index (n); R_Puma: Simulation results when only reflectance was input; B_Puma: Simulation results when both reflectance and transmittance were inputs.
Fig. 8. PUMA model validation with analytical extinction coefficient (k).
Fig. 9. PUMA model validation with analytical extinction coefficient (k).
Fig. 10. PUMA model evaluation with experimental transmittance.
Fig. 11. PUMA model evaluation with experimental reflectance.
Fig. 12. Flowchart for deep learning algorithm.
Fig. 13. Deep Learning Model developed.
Fig. 14. Comparison of model-predicted extinction coefficient with analytical values.
Fig. 15. Comparison of model-predicted refractive index with analytical values.
Fig. 16. Absorption regions for fabricated silicon thin film.
Fig. 17. Tauc’s Plot for determining optical bandgap.
K_Mean absolute error | 0.000418969 | K_Mean squared error | 2.6281E-07 | Sum_K.sq error | 9.93422E-05 | R2 | 0.987741346 |
|
Table 1. Prediction of extinction coefficient (k).
N_Mean absolute error | 0.023172839 | N_Mean squared error | 0.000574442 | Sum_N.sq error | 0.217139095 | R2 | 0.867649736 |
|
Table 2. Prediction of refractive index (n)
| Frequency (kHz) | Pulse duration (ps) | Temperature (°C) | Sample 1 | 600 | 150 | RT | Sample 2 | 900 | 150 | 200 | Sample 3 | 1200 | 150 | RT | Sample 4 | 1200 | 150 | 600 | Sample 5 | 1200 | 5000 | RT |
|
Table 3. Manufacturing parameters.
Sample | MSE of Transmittance | MAE of k | MAE of n | MSE of k | MSE of n | R2 value of k | R2 value of n | 1 | 1.91E–09 | 5.47E–04 | 8.33E–02 | 4.77E–07 | 1.39E–02 | 0.98 | 0.97 | 2 | 9.65E–11 | 7.74E–04 | 9.70E–02 | 1.00E–06 | 1.50E–02 | 0.97 | 0.97 | 3 | 1.87E–09 | 6.67E–04 | 1.04E–01 | 6.55E–07 | 2.20E–02 | 0.97 | 0.96 | 4 | 6.44E–10 | 1.04E–03 | 2.22E–01 | 1.79E–06 | 6.86E–02 | 0.90 | 0.88 | 5 | 3.97E–10 | 7.48E–04 | 1.05E–01 | 9.58E–07 | 1.79E–02 | 0.97 | 0.97 |
|
Table 4. Validation of proposed methodology.
Semiconductors | Crystal structure | Eg (T=300K)
| Type of band gap | Si | Diamond | 1.12 | Indirect | a-Si:H | Amorphous | 1.7 to 1.8 | Indirect | SiC(α)
| Wurtzite | 2.9 | Indirect |
|
Table 5. Band gap information for various silicon structures
7