Graphene-based metasurface solar absorber design for the visible and near-infrared region with behavior prediction using Polynomial Regression

The broadband graphene absorber designs for a visible and near-infrared (NIR) regime is developed in the study. The gold resonator is placed over an MgF 2 substrate which is followed by a gold base layer. The graphene spacer is sandwiched between the substrate and resonator layer due to its remarkable conductivity to increase the absorption response. The results are presented for the wavelength spanning from 0.2 µm to 1.2 µm. The average absorption for all three designs is compared and the best design is identified. The proposed structures are also validated by varying various parameters such as resonator thickness, substrate thickness, plus shape width, and needle-point width, and in the process, the absorptance response is improved. The result clearly shows the shift in absorptance response by varying the geometrical parameters. The proposed needle-point shaped solar absorber design can be an elementary block for visible and NIR photovoltaic devices whose utilization can be managed in applications related to energy harvesting. A novel part of this research is using polynomial regression models to predict the absorption capacity at a varying wavelength for variations in angle, resonator thickness, substrate thickness, and needle-point width. Experimental results show that polynomial regression models can predict the values absorption capacity with high accuracy (R 2 score).