Machine-Learning-Guided Genetic Inverse Design of Single-Atom Electrocatalysts for CO2 Reduction

Machine learning-optimized bioinspired N-doped carbon-wrapped trimetallic oxides for an efficient oxygen evolution reaction.

Despite the significant advancement in the development of a wide range of nitrogen-doped multi-metal oxide-based electrocatalysts for the oxygen evolution reaction (OER), there is a need to investigate the individual contributions of each metal node and the specific role of the N-source in OER performance. However, conventional experimental approaches are often labor-intensive, time-consuming, and inefficient for decoupling the complex synergistic interactions within multi-metallic systems, thereby limiting the development of efficient electrocatalysts. To address this challenge, herein we have employed machine learning (ML) optimization to understand the individual contributions of each metal in the multi-metallic system and N-sources, enabling both the screening and designing of efficient OER electrocatalysts. We have fabricated trimetallic FeCoZn squarate MOFs (FCZ-Sq MOFs), and ML models were systematically employed to optimize and identify the optimal (Fe/Co/Zn) metal node ratio to design highly efficient MOF-based electrocatalysts. The ML-optimized MOF was subsequently decorated with different bio-inspired N-sources (purine (Pu), pyridine (Py), and xanthine (Xn)) and wrapped with polydopamine (PDA), and a second-stage ML optimization was employed to elucidate and select the most effective nitrogen source. The ML-optimized materials were subjected to calcination to form N-doped carbon-coated trimetallic oxides (NC@FCZ-Ox). Among the ML-optimized materials, the Pu-derived catalyst (NCPu@FCZ-Ox) has shown enhanced electrocatalytic activity by exhibiting low overpotential (270 mV at 10 mA cm-2), low onset potential (1.40 V vs. RHE), and Tafel slope (74 mV dec-1) compared to NCPy@FCZ-Ox (1.42 V, 320 mV), NCXn@FCZ-Ox (1.44 V, 340 mV), FCZ-Sq MOF (1.47 V, 355 mV) and NF (1.60 V, 430 mV). Importantly, this work not only demonstrates the effectiveness of ML-assisted optimization in accelerating catalyst discovery but also provides a fundamental understanding of structure-performance relationships in multi-metallic and N-doped systems. To the best of our knowledge, this is the first study to report the impact of ML in precisely optimizing and screening the best metal nodes and N-sources for catalyst design in sustainable energy applications.

Machine Learning Reveals Role-Asymmetric Electronic Cooperation in Fe-Based Dual-Atom Catalysts.

Dual-atom catalysts (DACs) provide a powerful platform for oxygen electrocatalysis, yet rational design remains limited by the lack of transferable mechanistic principles. Machine learning (ML) has the potential to address this gap, yet its role in mechanistic discovery remains largely underexplored despite its wide use in catalyst screening. Here, using extended phthalocyanines (M1M2-ePc), we establish an integrated DFT-ML-experiment framework that maps catalytic performance onto an interpretable electronic landscape. Screening 81 DFT-computed and 360 ML-predicted metal pairs identifies FeM-ePc as a promising bifunctional catalyst family. Notably, SHapley Additive exPlanations (SHAP) analysis highlights the importance of electronic background and the key role of the secondary metal in regulating catalytic activity. First-principles calculations further uncover a cooperative dual-descriptor mechanism, in which d-band center and charge transfer jointly govern bifunctional activity. Combining LASSO with SISSO yields compact analytical formulas that quantitatively reproduce ηORR and ηOER, providing interpretable descriptors for DACs. Guided by these findings, FeCo-ePc-L with atomically dispersed Fe-Co sites was synthesized to experimentally examine the ML-guided prediction. This work highlights the utility of interpretable ML for mechanistic discovery in DACs by revealing role-asymmetric electronic cooperation between paired metal centers.

Machine learning-guided genetic algorithm for accelerated screening of Ni–Fe-based OER catalysts and validation by plasma-assisted electrodeposition

Developing efficient nickel–iron-based Oxygen Evolution Reaction (OER) catalysts via plasma-assisted electrodeposition holds great promise for the green hydrogen economy due to its compatibility with large-scale industrial manufacturing. However, the vast catalyst design space remains largely unexplored due to the inefficiency of traditional trial-and-error investigation. Herein, we develop a Machine Learning-guided Genetic Algorithm (ML-GA) paradigm with two complementary modes: exploration and exploitation. The exploration mode prioritizes population diversity while maintaining promising predicted performance to broadly sample the chemical space, while the exploitation mode focuses on high-performance regions to identify promising catalysts. Guided by this framework, NiFe/NiS catalysts were prioritized and synthesized via plasma-assisted electrodeposition, which demonstrated remarkable OER activity with low overpotentials of 219 mV and 315 mV at current densities of 10 mA cm −2 and 1000 mA cm −2 , respectively. This work demonstrates the effectiveness of the ML-GA approach in overcoming data limitations and accelerating the design of high-performance OER catalysts.

Toward accelerated electrocatalyst design: synergistic integration of DFT, machine learning, and microkinetic modeling.

Emerging opportunities in physics-informed machine learning, graph neural networks, generative artificial intelligence, active learning, and autonomous closed-loop DFT-ML-MKM workflows are discussed as promising directions for accelerating the discovery of next-generation electrocatalysts with enhanced activity, selectivity, and long-term stability.

Machine learning-guided discovery of γ-Al2O3 as defect-rich structural ensemble governs single-atom Pt catalysis

Machine-Learning-Guided Design of Pd–Au Electrocatalysts for the Oxygen Reduction Reaction

Machine learning has significantly reduced the computational power necessary to estimate the free energy of adsorption of key reaction intermediates on a diverse range of catalytic surfaces. Nevertheless, translating this computational capability into the discovery and experimental validation of catalysts necessitates targeting specific questions where this faster computation can yield the most significant impact. Using the electrochemical oxygen reduction reaction (ORR) as a model reaction due to its known linear scaling relationships between key catalytic intermediates, we demonstrate that the Open Catalyst Project’s machine-learning-based calculations of adsorption energies can inform experimental catalytic research. The primary challenge we addressed was the structural effect of the ORR, wherein the higher Miller index facets of Pt exhibit diminished ORR activity in comparison to Pt(111). The Open Catalyst Project’s rapid relaxation energy calculations enabled us to screen a large number of bimetallic materials for each key intermediate of the ORR over a wide range of crystal facets. The Open Catalyst Project was able to identify that PdAu3 can overcome the negative structural effects observed on the higher Miller index facets of Pt for the ORR. Electrochemical experimentation via rotating disk electrode linear sweep voltammetry and Tafel slope analysis revealed that polycrystalline PdAu3 nanoparticles exhibit an improved onset potential for the ORR compared to commercial polycrystalline Pt/C. Thus, this study demonstrates an example of how machine-learning-driven materials discovery can accelerate catalyst design while simultaneously offering an experimental solution to the persistent challenge of high ORR overpotential.