Unraveling the Reaction Mechanism and Activity of Grain Boundary Engineered M2C MXenes Toward Urea Electrosynthesis by High-Throughput DFT Calculations and Machine Learning.

Electrochemical urea synthesis from carbon and nitrogen sources holds significant potential for addressing environmental and energy challenges, yet complex reaction networks and competitive side reactions hinder the rational design of efficient electrocatalysts. Here, we performed high-throughput density-functional theory calculations on 36 M2C MXenes containing grain boundaries (GBs) to identify promising candidates for urea electrosynthesis. According to the well-defined screening criteria encompassing thermodynamic stability, adsorption behavior, reaction activity, and selectivity, six M2C GBs were identified as optimal electrocatalysts. Mechanistic analysis reveals that urea synthesis activity is primarily governed by C─N coupling and subsequent hydrogenation, which can be quantitatively described by the adsorption free energies of *CO (ΔG (*CO)) and *NCON (ΔG (*NCON)), respectively. Further, a machine learning (ML) model based on the Gradient Boosting Regression (GBR) algorithm demonstrated that the metallic electronegativity, the charge transfer to *CO, and the intrinsic charge transfer of metal atoms are the three key factors collectively modulating ΔG (*CO) and ΔG (*NCON). This work proposes a high-throughput framework for exploring GB-based electrocatalysts, elucidates the reactivity trends, and identifies the activity descriptors for urea synthesis, thereby advancing the discovery and rational design of electrocatalysts for urea production.