Thermal Growth Oxide Layer Thickness Detection Using Terahertz Time-Domain Spectroscopy Combined with Deep Learning

Table 1. Finite-difference time-domain simulation parameters
View table
Table 1. Finite-difference time-domain simulation parameters
Parameter Value
Time 100 ps
Frequency 0.3‒1.0 THz
Thickness of TC 300 $μ$ m
Refractive index of TC 5.0
Thickness of TGO 1.0‒10.5 $μ$ m

Table 2. TGO layer thicknesses of physical samples
View table
Table 2. TGO layer thicknesses of physical samples
Number Thickness of TGO /μm
1 3.0
2 4.0
3 5.2
4 5.8
5 6.2
6 7.2
7 8.0
8 8.5
9 9.2
10 10.3

Table 3. Simulation data training errors of LSTM model
View table
Table 3. Simulation data training errors of LSTM model
Thickness /μm Mean relative error /% Thickness /μm Mean relative error /%
1 122.5 6 16.7
2 64.9 7 12.3
3 39.7 8 11.8
4 22.8 9 12.4
5 19.5 10 10.0

Table 4. Training errors of simulation data
View table
Table 4. Training errors of simulation data
Thickness /μm Mean relative error /% Thickness /μm Mean relative error /%
1.0 20.1 6.0 5.9
1.5 13.3 6.5 2.9
2.0 18.8 7.0 3.4
2.5 11.0 7.5 4.1
3.0 8.1 8.0 5.5
3.5 6.5 8.5 6.1
4.0 7.7 9.0 2.9
4.5 4.8 9.5 3.4
5.0 6.3 10.0 4.5
5.5 4.8 10.5 3.6

Table 5. Training errors of physical sample data
View table
Table 5. Training errors of physical sample data
Thickness /μm Mean relative error /%
3.0 9.4
4.0 7.9
5.2 8.7
5.8 6.3
6.2 5.7
7.2 4.1
8.0 3.7
8.5 4.2
9.2 2.8
10.3 3.9

Tools

Get Citation

Copy Citation Text

Chenhao Xue, Jianhui Ma, Guang Yang, Jingqi Tong, Jiyuan Zhao. Thermal Growth Oxide Layer Thickness Detection Using Terahertz Time-Domain Spectroscopy Combined with Deep Learning[J]. Laser & Optoelectronics Progress, 2025, 62(17): 1730004

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites