Deep learning empowering design for selective solar absorber

Abstract The selective broadband absorption of solar radiation plays a crucial role in applying energy. However, despite being decade-old technology, the rapid and precise designs absorbers spanning from spectrum to infrared region remain significant challenge. This work develops high-performance design paradigm that combines deep learning multi-objective double annealing algorithms optimize multilayer nanostructures for maximizing spectral minimum radiation. Based on design, we experimentally fabricate designed absorber demonstrate its photothermal effect under sunlight. exhibits exceptional (calculated/measured = 0.98/0.94) low average emissivity 0.08/0.19). has potential result annual energy savings up 1743 kW h/m 2 areas with abundant resources. Our study opens powerful method solar-thermal harvesting manipulation, which will facilitate their broad applications other engineering applications.