Graphene-based solar absorber's potential for improved solar thermal energy conversion with optimization using machine learning

This article introduces a new concept of solar thermal absorber that has a Metal-Oxide Perovskite Solar Absorber (MOPSA), which is an Fe2O₃ resonator doped with a MAPbI₃ active layer on a Sn substrate. The Sn substrate, known for excellent thermal conductivity and stability, is highly compatible with the MAPbI₃ layer, which absorbs light in an exceptional manner at nearly all wavelengths along the visible and near-infrared spectrum of sunlight. Then, in addition, the Fe2O₃ resonator, based on localized surface plasmon resonances, will further amplify solar light absorption and energy conversion into thermal energy. At impressive mean absorption efficiencies of 94.51 % and 95.60 % for a wide span of 200–1600 nm of AM1.5 solar irradiation, MOPSA exhibits outstanding average absorption efficiencies under rigorous simulations and optimization. Efficiency was increased by using machine learning (LWR), which allowed for quicker optimization with fewer simulations while preserving high prediction accuracy. With its high solar-to-thermal energy conversion efficiency, quick temperature response, and long-term durability, the improved MOPSA design is a promising system for solar thermal applications. The goal of this research is to create low-carbon, sustainable energy technologies that will help a range of businesses decarbonize.