Electrochemical NO-to-NH₃ conversion on TM@NiN₂ single-atom catalysts: a DFT and machine learning investigation

Electrocatalytic reduction of nitrogen monoxide (NO) to ammonia (NH₃) offers a win-win solution for environmental remediation and chemical production. The key to realizing this technology lies in designing catalysts with superior performance. This study employs a combined density functional theory (DFT) and machine learning (ML) approach to systematically screen the nitric oxide reduction reaction (NORR) performance of 26 single-atom catalysts (TM@NiN₂, where TM = 3d, 4d, 5d). Through a multi-step screening protocol evaluating stability, NO adsorption, activity, and selectivity, Zn@NiN₂ is identified as the most promising candidate, exhibiting an ultra-low limiting potential (U_L) of 0 V. ML results reveal that high Q_s values and appropriate ε_d positions jointly determine NORR activity. Electronic structure analysis further reveals hybridization between Zn-3d orbitals and NO-2p orbitals, facilitating donor-acceptor interactions for NO activation. Concurrently, Bader analysis indicates the Zn site acts as an electron transfer mediator, directing electrons from the NiN₂ substrate to the reaction intermediate, thereby promoting NORR. To more accurately evaluate the activity of Zn@NiN₂, we explicitly consider the effects of solvent, pH, and electrode potential. Under these conditions, it achieves a record-low U_L of 0 V (vs. RHE). This work not only identifies an exceptional NORR catalyst but also provides guidelines for the rational development of electrocatalysts for NORR and related electrochemical reactions.