Aug 2026· Advances in Materials· pp.
e74833
· 0 citations· 74 references
Medicine
Abstract
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.
Recent studies have demonstrated that precise modulation of the coordination environment around metal centers can significantly influence the catalytic activity, inspiring interest in the rational design of efficient CO2 reduction reaction (CO2RR) catalysts through coordination engineering. In this work, a series of single-atom catalysts (SACs) with asymmetric coordination environments were systematically constructed based on heteroatom-doped graphene frameworks, aiming to elucidate the structure-activity relationships governing CO2RR performance. By integrating density functional theory (DFT) calculations with machine learning (ML) approaches, the limiting potential (UL) was employed as the target descriptor. EXtreme Gradient Boosting Regression (XGBR) and Sure Independence Screening and Sparsifying Operator (SISSO) algorithms were adopted to construct a black-box predictive model. An interpretable descriptor model, achieving coefficients of determination of R2 = 0.96 and R2 = 0.95, respectively, validated the reliability and generalizability of the established models. The results revealed that, compared to conventional M-N4 SACs with symmetric coordination environments, the introduction of asymmetric coordination could further enhance the CO2RR catalytic activity of Co-, Fe-, and Mn-based SACs, with notable performance variations across different metal centers. Four SACs with superior catalytic potential, namely, Co-N3O, Co-N3S, Fe-N4O2, and Mn-N4N3, were identified. Electronic structure analysis indicated that coordination environment modulation drove an upshift of the metal center d band toward the Fermi level, strengthening interactions with key reaction intermediates and optimizing adsorption energies to enhance catalytic activity. The DFT-ML strategy established herein provides a systematic framework for the rational design of SACs toward efficient CO2RR and offers generalizable principles for coordination engineering in electrocatalysis.
Unknown authors· ACS Applied Materials and In...· 0 citations
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.
Swetarekha Ram, Shalini Tomar, S. Bhattacharjee· Chemical Communications· 0 citations
Conventional dual-atom catalysts (DACs) for the nitrogen reduction reaction (NRR) primarily rely on adjacent metal sites to achieve synergistic effects. However, this configuration often suffers from inherent limitations such as steric hindrance, restricted active site density, and difficulty in finely tuning electronic interactions between metal centers, which collectively hinder catalytic performance under practical conditions. To address these challenges, we propose a novel class of bimetallic electrocatalysts featuring non-adjacent dual-atom sites anchored on N-doped graphene (denoted as TMATMB@NGII), in contrast to the widely studied adjacent-site counterparts (TMATMB@NGI). Through a combined approach of density functional theory (DFT) calculations and machine learning (ML), we systematically predict 784 transition metal combinations and identify V-Os@NGII as a highly promising catalyst, while Mn-Cd@NGII is selected as a representative ML-predicted candidate for complete-pathway validation. Mechanistic investigations reveal an unprecedented electron "storage-feedback" mechanism, which differs fundamentally from the conventional synergistic effects observed in adjacent-site systems and provides a new paradigm for understanding long-range electronic interactions in dual-atom catalysis. Our work not only introduces a non-adjacent dual-site architecture but also establishes an integrated DFT-ML framework for the rational design of advanced electrocatalysts beyond traditional coordination geometries.
Molecular design of an active electrocatalyst can determine the primary product. Precise selection of the active metal center influences the adsorption of reactants, intermediates, and product selectivity and ultimately affects the efficiency of the process. Here, we predict a potent catalyst, namely, Fe@C6N6, with enhanced activity and low overpotential for CO2 reduction reactions (CO2RRs) utilizing the density functional theory (DFT) approach, coupled with machine-learning-based model building. A series of 24 transition-metal single-atom catalysts (SACs) anchored on the C6N6 monolayer were screened, among which 19 exhibit effective CO2 activation. Employing density functional theory (DFT), a systematic study on reaction pathways demonstrates that 13 catalysts favor CH4 formation, while the remaining 6 exhibit selectivity toward CH3OH production. Additionally, 9 among the 19 catalysts are further screened out as they preferentially promote the competing hydrogen evolution reaction (HER). The active SACs demonstrate strong CO2RR activity, with Fe@C6N6 emerging as the most promising catalyst with the lowest limiting potential (0.39 V). To further establish atomic–property relationships, a machine-learning (ML) regression model was developed to predict adsorption energies using statistically significant descriptors and identify key factors governing CO2 adsorption. Furthermore, a classification model developed using the elemental characteristics of the metal atoms shows excellent correlation with the DFT results across the dataset. This combined DFT and machine-learning approach provides mechanistic insights for the rational development of efficient SACs for CO2 electroreduction.
Unknown authors· ACS Applied Energy Materials· 0 citations
The rational design of high-performance single-atom catalysts (SACs) requires precise optimizing of both the metal identity and its local coordination environments. In this study, a combined DFT, machine learning (ML), and experiment approach is employed to systematically screen ethane direct dehydrogenation (EDH) performance across 204 B/C/N/O/P/S-doped graphene-supported V/Cr/Ni/Pd SACs. Key findings reveal that asymmetric coordination structures, particularly Cr-CN3 with localized charge polarization at the Cr–C/N sites, exhibited lower C–H bond cleavage barriers, achieving improved EDH activity compared to symmetric counterparts (e.g., Cr-C4). Among seven ML algorithms tested, the GBRT algorithm shows the best predictive performance. GBRT-based models were subsequently developed, effectively predicting both the stability (R2 = 0.898) and catalytic activity (R2 = 0.848) of Pd SACs. From 51 candidate structures, the Pd-CB3 SAC was identified as the most promising, with a predicted higher activity. Interpretable ML analysis via Lasso regression further yielded a universal descriptor, φ′, which shows strong correlation with EDH activity across Pt/V/Cr/Ni/Pd SACs, providing a general design principle for SACs development. Finally, experimental synthesis and evaluation of DFT-screened Cr-CN3 and Ni-C3B SACs, ML-predicted Pd-CB3 SAC, and the reference symmetric Cr-C4 SAC confirmed that the asymmetric Cr-CN3 SAC shows the highest C2H6 conversion and the lowest deactivation rate constant (kd = 0.02 h–1).
Yajie Cao, Zhongkui Zhao, Shaojia Song et al.· ACS Catalysis· 0 citations
A genome-inspired materials intelligence framework (GIMI) for inverse design in high-dimensional compositional spaces is proposed, enabling targeted exploration of complex compositional space and accelerating the discovery of high-performance catalysts.
Chen Zhu, H. Yang, Haifeng Wang et al.· ACS Catalysis· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
