Borophene-Ge2Sb2Te5 (GST)-Based Refractive Index Sensor: Numerical Study and Behaviour Prediction Using Machine Learning.

Autor: Sorathiya, Vishal, Soni, Umangbhai, Vekariya, Vipul, Golani, Jaysheel, Almawgani, Abdulkarem H. M., Alhawari, Adam R. H.
Předmět:
Zdroj: Plasmonics; Jun2024, Vol. 19 Issue 3, p1211-1226, 16p
Abstrakt: We proposed the numerical studies and machine learning prediction for the multilayered Borophene-GST-silica-Ag-based refractive index sensor for the 1.3–1.5 µm wavelength range. The top layers of the proposed sensor incorporate Ag gratings. The proposed structure utilizes a mode based on surface plasmonic resonance, which has been created using a finite element model through computational analysis. Resonance is observed across a range of refractive index values from 1 to 2.5, which primarily aligns with the refractive indices of key biomolecules such as haemoglobin, saliva, urine, and cancerous cells. The proposed structure has also been investigated for the different physical parameters such as material height, grating space, the incident wave's wide incident angle, and the GST material's phase. We demonstrate the modulation of reflectance values in response to varying phases of the GST material (amorphous and crystalline), which can be controlled through external temperature changes. The distribution of the electric and magnetic fields of the structure is also provided in order to examine the field distribution across the grating and other layers of material. This sensor offers a wide angle of stability (0 to 80°) for the respective resonating points. An artificial neural network is employed to analyze the simulated data and make predictions regarding the behaviour of the structure. We have found the R2 = 0.97 for the proposed artificial neural network (ANN) model. The ANN model of this structure also presented the values of mean square error (MSE), mean absolute error (MAE), and root mean square error (RMSE) for different epoch values. Results shown in this research can help to sense the wide range of biomolecule samples whose refractive index ranges from 1 to 2.5. The wide wavelength range makes this device applicable for developing wideband infrared biosensors. [ABSTRACT FROM AUTHOR]
Databáze: Complementary Index