Interpretable Machine Learning Framework Deciphers the Role of Local Environment of High‐Entropy Intermetallic Compounds for Alkaline Hydrogen Evolution Reaction

Machine Learning Unveils Isolated‐Surrounded Pt Motifs in High‐Entropy Alloys for Superior Low‐Temperature Ammonia Oxidation

The sluggish kinetics of the ammonia oxidation reaction constitute a critical bottleneck in the development of low‐temperature direct ammonia fuel cells. High‐entropy alloys (HEAs), owing to their diverse active sites, have emerged as promising catalysts. However, their vast compositional space makes traditional quantum chemical screening prohibitively expensive. In this study, we provide a computational proof of concept showing that the classical d ‐band center theory fails to predict the ammonia oxidation activity of quinary HEAs and exhibits a negligible correlation with the energy barrier of the rate‐determining step. To address this limitation, we employed the Sure Independence Screening and Sparsifying Operator (SISSO) method to construct a transparent and interpretable symbolic descriptor, achieving excellent predictive accuracy ( R 2  = 0.981) within the range of the training data. This descriptor extends beyond simple single‐electron parameters by integrating the synergistic effects of electron‐donating ability, lattice stiffness, and local electronegativity perturbations. This data‐driven approach reduces computational costs by several orders of magnitude relative to exhaustive density functional theory (DFT) screening and identifies an “isolated‐surrounded” geometric configuration as a highly active site that significantly enhances intrinsic catalytic activity from a thermodynamic perspective. Crucially, the predicted motif should be interpreted within the hydrazine‐mediated thermodynamic framework used here, and its thermodynamic superiority may shift if alternative kinetic pathways dominate under operating conditions. This structural motif promotes efficient NN coupling while suppressing site poisoning. Overall, this study provides a coordination chemistry‐based blueprint for the rational design of next‐generation catalysts with reduced platinum‐group‐metal content and offers a theoretical framework for future experimental validation.

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

Machine Learning Guided Discovery of Corundum High Entropy Oxides

Early thinking in the field of high entropy oxides (HEOs) emphasized their likely abundance, with combinatorial arguments hinting at a myriad of new materials. The experimental reality has proven more challenging: the stability of HEOs cannot be straightforwardly predicted based on ionic radii, lattice geometry, and charge-balancing considerations alone. In this work, we employ machine learning interatomic potentials (MLIPs) to predict the synthesizability of HEOs of the form $A_2$O$_3$ derived from a selection of trivalent cations. From nearly 500 possible compositions, we identify 16 promising candidates for experimental validation with solid-state and combustion synthesis. We discover three new HEOs in the corundum structure, including (Al,Cr,Fe,Rh,Sc)$_2$O$_3$, and one novel cation-ordered phase, (Al,Fe,Ga,Sc)$_2$O$_3$. By far the most common synthesis outcome was a mixture of competing phases, sometimes involving redox reactions. Our results also reveal profound synthesis method dependence for the final product, where qualitatively equivalent outcomes between the two synthesis methods were only observed for 3 of the 16 tested compositions. We conclude that the occurrence rate of HEOs is far rarer than initially believed and that machine learning approaches can effectively guide us to the"needle in the haystack".

Interpretable data-driven discovery of high-performance multielement metal oxide electrocatalysts for oxygen evolution reaction from a small hybrid dataset

ABSTRACT One of the most formidable challenges in materials chemistry is the rational design of functionalities capable of dramatically enhancing performance. However, it is well-known that the discovery of promising materials often requires several decades of continuous trial-and-error. Herein, we show an interpretable data-driven framework for the discovery of multielement metal oxide oxygen evolution reaction (OER) electrocatalysts in alkaline media within a substantially shorter timeframe. This framework was trained on a hybrid dataset comprising only 557 data, consisting of the curated literature and our own experimental results. Furthermore, this framework was specifically designed to enable extrapolative materials discovery, including the exploration of elemental combinations absent from the training database. Consequently, from a large material search space of approximately 3 million candidates, we identified a promising unconventional quinary oxide composed of V, Ni, W, Rh, and Ru that exhibits, in 0.1 M KOH, an exchange current density approximately 20 times higher than that of IrO2. This work serves as a proof-of-concept, demonstrating that the rational design of high-performance electrochemical functionalities from an extensive candidate space can be achieved using a small hybrid dataset combined with an interpretable data-driven approach.

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

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.

Machine Learning-Guided Discovery of PGM-Lean High-Entropy Alloys for Efficient Solar Hydrogen Evolution

High-entropy alloys (HEAs) have emerged as a powerful materials platform for electrocatalysis due to their tunable surface energetics, structural stability, and diverse local atomic environments arising from multielement interactions. Composed of several principal elements in near-equimolar ratios, HEAs leverage high configurational entropy to stabilize single-phase solid solutions, while surface heterogeneity creates new catalytic motifs in which collective electronic and geometric effects yield activities exceeding those of pure metals. These attributes make HEAs particularly promising for the hydrogen evolution reaction (HER). Platinum-group-metal-containing HEAs (PGM-HEAs) exhibit exceptional HER activity and durability in acidic environments; however, their reliance on multiple precious metals limits large-scale deployment. Recent efforts demonstrate that partial substitution with earth-abundant transition metals can significantly reduce noble-metal content without sacrificing performance. Despite this progress, the atomic-scale origins of HER activity in HEAs—specifically the geometric and electronic descriptors governing optimal hydrogen binding—remain insufficiently understood. Here, we present an integrated density functional theory (DFT) and machine learning (ML) framework for discovering PGM-lean HEAs optimized for HER. Equimolar binary and ternary alloys are constructed from a nine-metal design space (Pt, Pd, Ir, Rh, Fe, Co, Ni, Mo, W), yielding 120 unique compositions. Representative surface configurations are selected using Kennard–Stone sampling and modeled as close-packed FCC(111) and BCC(110) slabs. Hydrogen adsorption energetics, and local active-site descriptors are computed using DFT and analyzed using ML models to establish structure–property relationships. Benchmark calculations on elemental FCC(111) surfaces reproduce expected periodic trends, with hydrogen adsorption free energies ranging from −0.46 eV on Ni(111) to −0.24 eV on Pt(111) at 1/4 monolayer coverage, consistent with d-band theory. Coverage effects are quantified by comparing 2×2×7 and 3×3×7 slab models, revealing stabilization of H* by approximately 0.02 eV at lower coverage. These results validate the computational methodology and establish a robust foundation for screening multicomponent HEA surfaces. The combined DFT–ML framework enables the rational identification of cost-effective, high-performance HEA electrocatalysts while providing fundamental insight into active-site chemistry in complex alloy systems. All computational data will be made available upon request to promote transparency and reproducibility.