Tunable graphene-based metamaterial thermal absorber design for thermal sensing applications with behaviour prediction using machine learning

Abstract : This investigated Graphene Monolayer Nanostructure as a Narrow Band (GMNNB) absorber. Which used AuAl 2 as a substrate and SiO 2 as a resonator material with graphene monolayer-based structure. It presents the narrow band of 10 nm with more than 97.81 % absorption peak at 1880 nm wavelength. Also, a narrow band region 0.5 μm to 0.6 μm. To make the process faster, we utilized a machine learning approach based on the K-Nearest Neighbors (KNN) model which reduces the process time by 1/8 then the conventional method. A low mean squared error (4.12652 x 10 -5 ) and high R 2 values (ranging from 0.944 to 0.986) in the prediction indicate that the model is very accurate and generalizable in estimating GMNNB structure. By varying the resonator thickness, substrate thickness, and structure breadth to target certain absorption wavelengths, the GMNNB structure may be tuned to operate as a metamaterial absorber and sensor. Its exceptional selectivity makes it valuable for various applications like optical filtering, gas/chemical sensing, stealth tech, and thermal energy harvesting.