Validating ΔΔG Selectivity Descriptor for Electrosynthesis of H2O2 from Oxygen Reduction Reaction.

Understanding selectivity trends is a crucial hurdle in the developing innovative catalysts for generating hydrogen peroxide through the two-electron oxygen reduction reaction (2e-ORR). The identification of selectivity patterns has been made more accessible through the introduction of a newly developed selectivity descriptor derived from thermodynamics, denoted as ΔΔG introduced in Chem Catal. 2023, 3(3), 100568. To validate the suitability of this parameter as a descriptor for 2e-ORR selectivity, we utilize an extensive library of 155 binary alloys. We validate that ΔΔG reliably depicts the selectivity trends in binary alloys reported for their high activity in the 2e-ORR. This analysis also enables the identification of nine selective 2e-ORR catalysts underscoring the efficacy of ΔΔG as 2e-ORR selectivity descriptor. This work highlights the significance of concurrently considering both selectivity and activity trends. This holistic approach is crucial for obtaining a comprehensive understanding in the identification of high-performance catalyst materials for optimal efficiency in various applications.

Data-driven discovery of electrocatalysts for CO2 reduction using active motifs-based machine learning.

The electrochemical carbon dioxide reduction reaction (CO2RR) is an attractive approach for mitigating CO2 emissions and generating value-added products. Consequently, discovery of promising CO2RR catalysts has become a crucial task, and machine learning (ML) has been utilized to accelerate catalyst discovery. However, current ML approaches are limited to exploring narrow chemical spaces and provide only fragmentary catalytic activity, even though CO2RR produces various chemicals. Here, by merging pre-developed ML model and a CO2RR selectivity map, we establish high-throughput virtual screening strategy to suggest active and selective catalysts for CO2RR without being limited to a database. Further, this strategy can provide guidance on stoichiometry and morphology of the catalyst to researchers. We predict the activity and selectivity of 465 metallic catalysts toward four expected reaction products. During this process, we discover previously unreported and promising behavior of Cu-Ga and Cu-Pd alloys. These findings are then validated through experimental methods.

Accelerating the Search for New Solid Electrolytes: Exploring Vast Chemical Space with Machine Learning-Enabled Computational Calculations.

Discovering new solid electrolytes (SEs) is essential to achieving higher safety and better energy density for all-solid-state lithium batteries. In this work, we report machine learning (ML)-assisted high-throughput virtual screening (HTVS) results to identify new SE materials. This approach expands the chemical space to explore by substituting elements of prototype structures and accelerates an evaluation of properties by applying various ML models. The screening results in a few candidate materials, which are validated by density functional theory calculations and ab initio molecular dynamics simulations. The shortlisted oxysulfide materials satisfy key properties to be successful SEs. The advanced screening method presented in this work will accelerate the discovery of energy materials for related applications.

Facet-Defined Strain-Free Spinel Oxide for Oxygen Reduction.

Exposing facet and surface strain are critical factors affecting catalytic performance but unraveling the composition-dependent activity on specific facets under strain-controlled environment is still challenging due to the synthetic difficulties. Herein, we achieved a (001) facet-defined Co-Mn spinel oxide surface with different surface compositions using epitaxial growth on Co3O4 nanocube template. We adopted composition gradient synthesis to relieve the strain layer by layer, minimizing the surface strain effect on catalytic activity. In this system, experimental and calculational analyses of model oxygen reduction reaction (ORR) activity reveals a volcano-like trend with Mn/Co ratios because of an adequate charge transfer from octahedral-Mn to neighboring Co. Co0.5Mn0.5 as an optimized Mn/Co ratio exhibits both outstanding ORR activity (0.894 V vs RHE in 1 M KOH) and stability (2% activity loss against chronoamperometry). By controlling facet and strain, this study provides a well-defined platform for investigating composition-structure-activity relationships in electrocatalytic processes.

Electrochemical Hydrogen Peroxide Synthesis from Selective Oxygen Reduction over Metal Selenide Catalysts.

Se-based nanoalloys as an emerging class of metal chalcogenide with tunable crystalline structure, component distribution, and electronic structure have attracted considerable interest in renewable energy conversion and utilization. In this Letter, we report a series of nanosized M-Se catalysts (M = Cu, Ni, Co) as prepared from laser ablation method and screen their electrocatalytic performance for onsite H2O2 generation from selective oxygen reduction reaction (ORR) in alkaline media. A flexible control on 2e-/4e- ORR pathway has been achieved by engineering the alloying component. Moreover, through a feedback loop between theory and experiment an optimized scaling relationship between oxygenated ORR intermediates has been discovered on cubic Cu7.2Se4 nanocrystals, that is, the ensemble effect of isolated Cu component destabilizes O* binding while the ligand effect of Se to Cu fine-tunes the binding strength of OOH*, leading to a superb H2O2 selectivity above 90% over a wide potential window even after 1400 potential cycles.

Atomic Structure-Free Representation of Active Motifs for Expedited Catalyst Discovery.

To discover new catalysts using density functional theory (DFT) calculations, binding energies of reaction intermediates are considered as descriptors to predict catalytic activities. Recently, machine learning methods have been developed to reduce the number of computationally intensive DFT calculations for a high-throughput screening. These methods require several steps such as bulk structure optimization, surface structure modeling, and active site identification, which could be time-consuming as the number of new candidate materials increases. To bypass these processes, in this work, we report an atomic structure-free representation of active motifs to predict binding energies. We identify binding site atoms and their nearest neighboring atoms positioned in the same layer and the sublayer, and their atomic properties are collected to construct fingerprints. Our method enabled a quicker training (200-400 s using CPU) compared to the previous deep-learning models and predicted CO and H binding energies with mean absolute errors (MAEs) of 0.120 and 0.105 eV, respectively. Our method is also capable of creating all possible active motifs without any DFT calculations and predicting their binding energies using the trained model. The predicted binding energy distributions can suggest promising candidates to accelerate catalyst discovery.