Graphene-based multilayer metasurface solar absorber with parameter optimization and behavior prediction using Long Short-Term Memory model

Most of the solar absorbers are absorbing visible spectrum energy and are not focused on the ultraviolet spectrum and infrared spectrum energies. Here we propose a solar absorber that absorbs most of the energy of the available solar spectrum including visible spectrum and ultraviolet spectrum energies. The multilayer metasurface solar absorber design is analyzed and its results are compared with the single-layer metasurface solar absorber design. Both design results are also compared with AM 1.5 solar spectral irradiance plot for a wavelength range of 0.2 μm–1.5 μm. The graphene is used as a spacer to increase the absorption response of the proposed structure. The efficiency of multilayer design is high in both visible spectrum regions and ultraviolet spectrum regions compared to single-layer design. The parameter optimization for substrate thickness and resonator thickness is analyzed for improvement of absorption of the proposed design. The design results in the form of absorption, a normalized electric field. The proposed solar absorber with its high efficiency in visible and ultraviolet regions can be used in solar energy harvesting devices. The novel part of this research is predicting absorption values for assorted variations in substrate thickness, resonator thickness for forthcoming wavelengths using a Long Short-Term Memory model. The results of the experiments show that prediction may accurately anticipate absorption values, reducing simulation time and resource requirements. The visual comparison of simulated and predicted values is also presented in the paper.