Genetic algorithm-driven design of NIR-Reflective transparent colored multilayers for enhanced radiative cooling

Abstract

Passive radiative cooling is a promising path to tackle worsening energy crisis and global warming. Despite advancements in cooling mechanisms, material design, preparation technologies, and practical applications, the traditional white or silver appearance fails to meet both aesthetic and functional requirements, and the lack of transparency limits their applicability in scenarios where optical clarity is crucial. In this work, a genetic algorithm is employed to optimally design a dielectric/metal/dielectric/metal/dielectric stacked multilayer structure as a near-infrared (NIR) reflector, which is integrated with a plain glass substrate and an infrared high-emission PDMS layer on the outermost layer to form a transparent-colored radiative cooler (TCRC). By integrating a Fabry-Perot resonant cavity within the NIR reflector, we achieved customization of TCRC with varying colors. A grey TCRC exhibits optimal performance with visible transmissivity of 0.63, high NIR reflectivity of 0.88, and atmospheric transparency window emissivity of 0.95, all demonstrating angular independence (<60°). In outdoor experiments during midday, TCRC achieves a room temperature reduction of 17.6 °C compared to the original glass. Additionally, TCRC exhibits an extraordinary potential for building energy-saving in most climate zones. This work provides a valuable reference for the further development of radiative cooling and the design of metamaterials.