Iron-based multilayer materials for a machine learning-optimized graphene broadband solar absorber for sustainable energy applications

Solar absorber, as a new technology in the solar energy system, can also help reduce CO2 through thermal conversion. In the current research, we investigated a tri-layer absorber using machine-learning optimization and then proved that it is polarization-insensitive as well. The current absorber mainly uses the iron material with the cylinder and four squares, and it is named the Iron-based Cylinder Four-squares Shapes Solar Absorber (ICFSA). The three layers with the iron-based materials, including the pure iron (Fe), the magnetite (Fe3O4), and the ferric oxide (Fe2O3) for resonator-substrate-based construction, and the graphene is also included. With 93.98% near-perfect rate for the overall range (2800 nm), the developed ICFSA design can be explored 88.27%, 95.02%, 94.47%, and 93.87% for UV – MIR air regions. Linear regression method as the ML fabrication in the ICFSA design, and then showing the above 0.9 R 2 value and the polarization-insensitive indication included with 0 – 80 incidence angle variation. Design development, the layer construction, the AM 1.5, the parametric analysis, and the comparison table were also included to develop the presentation of the current ICFSA work. As a new technology of the solar absorber design, the developed ICFSA design can be used for thermal conversion uses, mainly including solar heating, space heating, domestic water heating, the industrial heating and so on.