Data-Driven Accelerated Discovery Coupled with Precise Synthesis of Single-Atom Catalysts for Robust and Efficient Water Purification.

The development of advanced catalysts frequently employs trial-and-error methods and is lack of highly controlled synthesis, resulting in unsatisfactory development efficiency and performance. Here we propose a data-driven prediction coupled with precise synthesis strategy to accelerate the development of single-atom catalysts (SACs) for efficient water purification. The data-driven approach enables the rapid screening and prediction of high-performance SACs from 43 metals-N structures comprising transition and main group metal elements, followed by validation and structural modulation for improved performance through a highly controllable hard-template method. Impressively, a well-designed Fe-SAC with a high loading of Fe-pyridine-N sites (~3.83 wt %) and highly mesoporous structure, exhibits ultra-high decontamination performance (rate constant of 100.97 min g), representing the best Fenton-like activities for sulfonamide antibiotics to date. Furthermore, the optimized Fe-SAC shows excellent robust environmental resistance and cyclic stability with almost 100 % degradation efficiency of sulfonamide antibiotics for 100-h continuous operation. Density functional theory calculations reveal that Fe-pyridine-N sites can reduce the energy barrier of intermediate O* formation, the rate-determining step, resulting in highly selective generation of singlet oxygen. The integration of data-driven method with precise synthesis strategy provides a novel paradigm for the rapid development of high-performance catalysts for environmental field as well as other important fields including sustainable energy and catalysis.