Unraveling Surface Reconstruction of Copper Alloy Catalysts During CO₂ Electroreduction in Gas Diffusion Electrode-Based Electrolyzers

The electrochemical CO₂ reduction reaction (CO₂RR) has attracted significant interest for the long-term storage of intermittent renewable energy and net-zero carbon emissions. For economic profitability, electrocatalyst design is necessary for selective production of value-added multi-carbon (C₂₊) chemicals. In CO₂RR electrocatalysts, various heteroatoms have been arranged in Cu to steer the reaction pathway toward specific C₂₊ products. However, surface reconstruction-induced atomic rearrangements change the initial alloy active sites during CO₂RR and prevent steering the reaction pathway in the intended direction. For the synthesis of efficient CO₂RR electrocatalysts, it is necessary to understand catalyst reconstruction behaviors. It is reported how reconstruction changes the morphology, facets, and oxidation state of mono-metallic catalysts during CO₂RR. However, for alloy catalysts, it is more challenging to elucidate the reconstruction mechanism due to chemical complexity of multi-elemental system. Furthermore, most reconstruction studies were conducted in conventional H-cell with low-rate CO₂RR. Therefore, the knowledge gained from the mechanism studies is difficult to apply directly to the practical reaction conditions, where gas diffusion electrodes (GDEs) are inevitably introduced for rapid CO₂ mass transport.
Herein, we present a universal principle for understanding and predicting the reconstruction of metal alloy catalysts during CO₂RR in GDE-based electrolyzers. Through thermochemical computation, we constructed a thermodynamic elemental map using oxophilicity and atomic miscibility between Cu and heteroatoms (X) as descriptors to systematically analyze the reconstruction behaviors of Cu alloys. By categorizing Cu-X alloys into four sections, we selected Ag, Fe, Zn, and Pd as representative X elements to investigate the reconstruction behaviors. Cu-X catalyst layers were fabricated on polytetrafluoroethylene (PTFE) fiber-based gas diffusion layers (GDLs). Cross-sectional transmission electron microscopy (TEM) with a focused ion beam (FIB) enabled to observe the changes in atomic arrangement and elemental clustering over the catalyst surface after CO₂RR. As a result, we confirmed that alloying immiscible X with Cu (X: Ag or Fe) induced the growth of Cu-rich nanoparticles with lattice defects; in contrast, alloying miscible X with Cu (X: Zn or Pd) maintained the atomically flat surface and homogeneous element distribution. Furthermore, real-time structural analysis revealed that the dynamic evolution of adparticles occurred via the dissolution-redeposition mechanism with the coalescence of the particles during CO₂RR. With density functional theory (DFT) calculation, we elucidated that miscibility and adsorbate-modulated oxophilicity are the determining factors in the evolution of adparticles. Furthermore, we demonstrated a strong correlation between reconstructed surface structure and *CO spillover efficiency, controlling the selectivity of ethanol (C₂H₅OH)/ ethylene (C₂H₄) as well as C₂/C₁ products. Finally, we developed a methodology to control the reconstruction behaviors in operando condition. This enabled a dramatic improvement in C₂H₅OH selectivity and its structure-performance relationship is discussed.