Hao Ming Chen

Autonomous experimentation offers a time-efficient route towards accelerated catalyst discovery, but current practices often rely on black-box models that are difficult to interpret. Here we demonstrate the paradigm of interpretable autonomous exploration of a 12-dimensional space in the electrocatalytic hydrogen evolution reaction by integrating an interpretable decision-making artificial intelligence with a fully automated electrochemical platform. Guided by a human’s chemistry-informed hypothesis, a chemistry-infused bandit is established to reflect the locally quadratic approximation between synthetic conditions and electrocatalytic activities near the explored optimum region. Not only does such a decision-making bandit lead to the discovery of previously unreported, earth-abundant electrocatalysts at least comparable to Pt-based ones, the bandit also reveals model-derived hypothesis for how synthetic conditions affect the electrocatalysts’ performance near the identified local optima as autonomous inquiry proceeds, which is otherwise untenable from black-box ones. Incorporating chemistry-informed hypothesis into the decision-making bandit is generally applicable to any chemistry problem and offers a framework for autonomous investigation, advancing hypothesis generation for catalyst’s design principles beyond simple optimization.

Electrocatalytic nitrate reduction reaction ( e‐ NO ₃ RR) offers a promising route to valorize nitrate‐contaminated wastewater while enabling sustainable ammonia production; however, how local coordination asymmetry regulates Cu‐site evolution and reaction pathways across different pH environments remains insufficiently understood. Herein, three‐ and four‐nitrogen‐coordinated Cu single‐atom motifs, denoted as Cu─N ₃ and Cu─N ₄ , are constructed as well‐defined pre‐catalyst models to elucidate how the initial Cu coordination environment directs pH‐dependent active‐site reconstruction. Operando X‐ray absorption spectroscopy and wavelet‐transform EXAFS reveal that Cu─N ₄ readily undergoes rapid reconstruction under acidic, neutral, and alkaline conditions, generating low‐coordinated, electron‐enriched Cu species while concomitantly inducing Cu─Cu aggregation and excessive Cu─site reduction. In contrast, Cu─N ₃ displays more regulated site evolution, with limited activation under acidic conditions but progressive formation of low‐coordinated Cu─N _x active structures under neutral and alkaline conditions. In situ ATR‐FTIR further verifies that the dynamically evolved Cu─N ₃ sites enhance nitrate adsorption, facilitate N─O bond activation, and promote the formation of hydrogenated nitrogen intermediates. These findings establish coordination‐governed site evolution as a design principle for balancing productive Cu─site activation and metal‐centered over‐reduction, providing molecular‐level guidance for the design of pH‐adaptive NO ₃ RR catalysts.

Electrocatalytic urea oxidation reaction (UOR) offers a low‐thermodynamic‐potential alternative to the oxygen evolution reaction, yet the true reactive state of nickel and a quantitative electronic descriptor governing activity remain unresolved. Here, we construct a well‐defined catalyst–support interaction platform that enables dynamic regulation of the Ni 3d electronic structure through π–π coupling between nickel phthalocyanine (NiPc) and carbon supports with tunable curvature. Crucially, operando x‐ray absorption spectroscopy, Kβ x‐ray emission spectroscopy, and high‐energy‐resolution fluorescence‐detected XAS unambiguously demonstrate that enhanced UOR activity originates from dynamic redistribution of Ni 3d orbitals rather than oxidation‐state evolution, thereby resolving a long‐standing controversy in Ni‐based catalysis. The operando results further confirm that Ni ²⁺ remains the catalytically active state throughout the reaction and identify the occupancy of the 3dz ² orbital as a quantitative electronic descriptor governing intrinsic activity. NiPc supported on 2 nm carbon nanotubes achieves a turnover frequency of 1.67 s ⁻¹ at 1.45 V versus RHE, outperforming its planar and weakly interacting counterparts, in which strong interfacial interaction induces a high‐spin Ni(II) configuration with a nearly half‐filled 3dz ² orbital. This study establishes operando‐validated frontier orbital engineering via catalyst–support interactions as a rational strategy for advanced electrocatalyst design.

Electrocatalytic urea oxidation reaction (UOR) offers a low‐thermodynamic‐potential alternative to the oxygen evolution reaction, yet the true reactive state of nickel and a quantitative electronic descriptor governing activity remain unresolved. Here, we construct a well‐defined catalyst–support interaction platform that enables dynamic regulation of the Ni 3d electronic structure through π–π coupling between nickel phthalocyanine (NiPc) and carbon supports with tunable curvature. Crucially, operando x‐ray absorption spectroscopy, Kβ x‐ray emission spectroscopy, and high‐energy‐resolution fluorescence‐detected XAS unambiguously demonstrate that enhanced UOR activity originates from dynamic redistribution of Ni 3d orbitals rather than oxidation‐state evolution, thereby resolving a long‐standing controversy in Ni‐based catalysis. The operando results further confirm that Ni ²⁺ remains the catalytically active state throughout the reaction and identify the occupancy of the 3dz ² orbital as a quantitative electronic descriptor governing intrinsic activity. NiPc supported on 2 nm carbon nanotubes achieves a turnover frequency of 1.67 s ⁻¹ at 1.45 V versus RHE, outperforming its planar and weakly interacting counterparts, in which strong interfacial interaction induces a high‐spin Ni(II) configuration with a nearly half‐filled 3dz ² orbital. This study establishes operando‐validated frontier orbital engineering via catalyst–support interactions as a rational strategy for advanced electrocatalyst design.

Electrochemical CO2 reduction toward multi-carbon products is often attributed to adjacent Cu sites that enable *CO coupling, yet whether designed Cu–Cu motifs persist under reaction conditions remains unclear. Here, we investigate a series of trinuclear Cu catalysts with similar initial coordination structures but different steric rigidity in the secondary coordination shell. Operando Raman and X-ray absorption spectroscopies reveal that catalysts with insufficient steric constraint undergo asymmetric Cu–Cu distortion under cathodic bias, leading to heterogeneous *CO adsorption behavior and enhanced C1 pathways. In contrast, sterically stabilized catalysts preserve symmetric Cu–Cu configurations and sustain favorable *CO–*CO coupling toward C2+ formation. These findings demonstrate that geometric persistence of neighboring Cu sites under operating conditions, rather than the initial structure alone, governs catalytic selectivity in CO2 electroreduction and highlights the importance of operando characterization for identifying active-state structures.

High-entropy alloys are often presumed stable and intrinsically beneficial for electrocatalysis. This Comment argues that in situ structural evolution can diminish configurational entropy, necessitating rigorous identification of active phases under operando conditions before attributing catalytic performance to high-entropy effects.

Morphology control has long been a widely adopted strategy in catalyst design, based on the assumption that geometric and electronic structures are inseparable in determining catalytic behavior. This perspective revisits that assumption by highlighting how catalyst structures dynamically evolve under realistic reaction conditions. Traditional structural descriptors, including local structural metrics such as coordination number and morphology-related features such as facet exposure, are commonly derived from the as-synthesized state and may not reflect the actual active configurations during reactions. Evidence from photocatalytic, thermocatalytic, and electrocatalytic systems reveals that catalysts undergo significant structural and electronic transformations in response to illumination, thermal input, and applied potential. These findings challenge the conventional static structure–activity paradigm and call for a revised framework that integrates morphology control with in situ characterization techniques. By tracking structural, electronic, and compositional changes in real time, morphology can be redefined not as a fixed geometric label but as an evolving platform that connects initial design to dynamic catalytic function.