Machine-Learning-Enabled Wood with Nanopump Functionalization for Solar Interfacial Evaporation

The increasing scarcity of freshwater resources has driven the development of solar interfacial evaporation technology, the effectiveness of which depends on the design of photothermal materials. Herein, a machine learning approach was combined with material design to develop Fe-Co Prussian blue analogue (PBA)-derived carbide/wood composite photothermal materials for desalination via solar interfacial evaporation. Through a bimetallic synergistic effect, the Fe-Co ions enhanced the wood's light absorption and stability, and also improved water transport via a “nanopump” effect, enabling the composite material to achieve a high evaporation rate of 2.807 kg m −2 h −1 under 1 sun. Notably, the desalinated seawater from the material was experimentally confirmed to meet World Health Organization (WHO) drinking-water standards. The machine learning model was applied to optimize Prussian blue analogue selection and the composite material design, providing a data-driven route for developing photothermal materials.