Electronic-structure regulation of graphdiyne-supported dual-atom catalysts drives efficient urea synthesis from CO2 and N2

Abstract

The electrocatalytic coupling of carbon dioxide (CO₂) and nitrogen (N₂) to obtain urea offers a sustainable route to valorize carbon and close the nitrogen cycle under ambient conditions. However, the design and screening of electrocatalysts with both high activity and selectivity remain major challenges in this field. Hence, density functional theory (DFT) combined with machine learning (ML) was employed to design graphdiyne-supported heteronuclear dual atom catalysts (TMCu@GDY) for urea electrosynthesis. A new four-step screening strategy was proposed, identifying three highly active urea electrocatalysts (MoCu@GDY, WCu@GDY, and NbCu@GDY). Among them, WCu@GDY exhibits the best intrinsic performance with a low limiting potential of −0.61 V. The d-band centers emerge as a robust activity descriptor for the change in the free energy of the potential-determining step (PDS) in urea formation (R² = 0.84), which is consistent with the tunable orbital hybridization at the dual site. Qualitative and quantitative selectivity analyses further show that WCu@GDY achieves near-quantitative selectivity for urea while strongly suppressing a series of side reactions. ML results indicate that the electronic structures (electron affinity, valence electrons, electronegativity and d-band center) of the other metal atom on GDY are the key factors affecting the performance of catalysts; thus, the catalytic activity can be tuned by changing the factors. Collectively, this work delivers a generalizable, data-guided blueprint for accelerating urea-catalyst discovery and offers transferable principles for other multi-molecule electrosynthetic reactions.