Mo Yan

Growing evidence shows that heterogeneous thermochemical redox reactions can proceed through coupled electrochemical half-reactions at a mixed potential. Yet how the electrical double layer (EDL), which sets the local reaction environment, regulates these reactions remains poorly understood, especially on supported catalysts where neighboring double layers can overlap. Here we develop a minimal Au–Pd model coupling Poisson–Boltzmann electrostatics to Frumkin-corrected Butler–Volmer kinetics, with anion oxidation on Au and cation reduction on Pd. First, we show that the EDL itself cannot be neglected: for Au and Pd electronically connected but spatially separated, so that their double layers do not overlap, resolving the EDL substantially changes the predicted kinetics relative to an EDL-absent description. Second, comparing a physical mixture, a carbon-supported Janus configuration, and alternating Au–Pd stripes reveals distinct rates arising from EDL overlap. The effects of EDL on the mixed potential and mixed current are consolidated into an analytical expression. These results establish the EDL and its overlap as key, tunable factor linking catalyst structure to the kinetics of heterogeneous thermochemical reactions.

Heterogeneous catalysis optimization relies on the coordinated selection of catalyst composition, synthesis procedure, and reaction conditions, yet the empirical links between these variables and catalytic performance remain scattered across individual studies. Here, we show that an interpretable machine learning (ML) framework can decode this complexity by integrating disparate literature data. We curate, to our knowledge, the first unified dataset from 200 papers spanning three decades of CO ₂ -to-methanol synthesis over Cu/ZnO-based catalysts, and derive a subset of apparent activation energy ( E _app ). Using the dataset, we optimize six ML models via Bayesian hyperparameter tuning and apply regularization strategies to mitigate overfitting. We then use SHAP to quantify the non-linear impact of catalyst properties, synthesis procedures, and reaction conditions on methanol selectivity, CO ₂ conversion, and methanol space-time yield. Crucially, this data-driven approach uncovers previously obscured operating windows for achieving high methanol productivity at milder pressures than conventionally assumed. Extending the ML framework to the E _app kinetic subset further identifies regime-sensitive signatures. A controlled experiment over commercial catalysts corroborates the positive space velocity– E _app relationship inferred from ML models. These results show how interpretable ML built on curated legacy experiments can support operating-window selection, hypothesis-driven kinetic experiments, and practical optimization of mature catalytic technologies.

Understanding how chemical energy dissipates as heat in non-equilibrium redox systems is a fundamental problem in physical chemistry.

Ethanol synthesis from CO ₂ and H ₂ is currently a hot topic of research in heterogeneous catalysis. However, the underlying mechanism has not yet been clearly elucidated. We propose that the synthesis involves several other reactions on the electrode surface, as evident by the formation of a mixed potential. In this study, we prepared solid catalysts, Ru/CoO _x and CuPd/C, as the nano‐anode and nano‐cathode components and carried out CO ₂ hydrogenation in the presence of small amounts of water to facilitate ion conduction. Ethanol was produced along with CO, CH ₄ , and methanol at reaction temperatures below 473 K in the presence of both the catalysts. The activation energy for ethanol formation with Ru/CoO _x was found to be 25 ± 5 kJ mol ⁻¹ , which was much lower than those for the formations of CO (44 ± 4 kJ mol ⁻¹ ), CH ₄ (77 ± 7 kJ mol ⁻¹ ), and methanol (81 ± 9 kJ mol ⁻¹ ). In addition, the production of CO was promoted by the encapsulated water. These results also suggest a mixed‐potential‐driven mechanism for the synthesis of ethanol.

Gold is an effective catalyst for low‐temperature CO oxidation, yet its mechanism remains debated. Among the various proposed mechanisms, increasing attention has been given to a possible electrochemical pathway in recent studies. Here, experimental evidence is reported that CO oxidation at room temperature in electrolytes proceeds via coupled electrochemical CO oxidation and oxygen reduction half‐reactions, i.e., the mixed‐potential‐driven mechanism. A model system is designed with deposited Au nanoparticles (NPs) and nitrogen‐doped reduced graphene oxide (NrGO) on carbon paper acting as spatially separated electrodes in a single electrolysis cell. After shorting the two electrodes, the reaction current and mixed potential of the overall CO oxidation are measured without any externally applied bias. Independent current–potential curves for each half‐reaction are also obtained under identical experimental conditions. The intersection points of these curves, representing the predicted mixed potential and net current, showed good agreement with the values measured in the short‐circuited system. These results provide clear evidence that CO oxidation in aqueous media proceeds via a mixed‐potential‐driven mechanism, analogous to corrosion. This finding provides strong support for the hypothesis that low‐temperature CO oxidation over‐supported Au catalysts in the presence of water vapor proceeds via a similar electrochemical pathway in gas‐phase heterogeneous catalysis.

Electrochemical CO ₂ reduction (CO ₂ RR) presents a sustainable approach to converting greenhouse gases into valuable chemicals. However, achieving precise product selectivity remains challenging. This study investigates the influence of applied potential and catalyst morphology on CO ₂ RR selectivity using a machine learning (ML) framework. Unlike most existing ML applications which primarily rely on open databases, we collect 378 experimental data points from literature focusing on Cu and Cu‐derived catalysts in H‐type cells. Various ML models are trained using both numerical and categorical parameters, with the artificial neural network model demonstrating the highest accuracy in predicting Faradaic efficiencies of key CO ₂ RR products. The SHapley Additive exPlanations (SHAP) analysis identifies applied potential as the primary determinant of product distribution. Additionally, it reveals synergistic effects between catalyst morphology and size, influencing pathways toward CO, C ₁ , and C ₂₊ products. These insights offer guidance for the rational design of Cu‐based CO ₂ RR systems by optimizing operational parameters to enhance multi‐carbon product selectivity. The study introduces a novel dataset and a data‐driven methodology that integrates experimental findings with ML techniques, advancing the mechanistic understanding of CO ₂ RR and providing a scalable approach for catalyst discovery and optimization.

Mixed-potential-driven catalysis is expected to be a distinctive heterogeneous catalytic reaction that produces products different from those produced by thermal catalytic reactions without the application of external energy. Electrochemically, the mechanism is similar to that of corrosion. However, a theory that incorporates catalytic activity as a parameter has not been established. Herein, we report the theoretical framework of mixed-potential-driven catalysis, including exchange currents, as a parameter of catalytic activity. The mixed potential and partitioning of the overpotential were determined from the exchange current by applying the Butler–Volmer equation at a steady state far from equilibrium. Mixed-potential-driven catalysis is expected to open new areas not only in the concept of catalyst development but also in the field of energetics of biological enzymatic reactions.