Fritz Haber Institute of the Max Planck Society

Interlayer engineering is widely used to improve charge storage in layered transition metal dichalcogenides, yet most studies rely on nanosized materials where the effects of interlayer expansion and particle downsizing are intertwined. Here, molecular pillaring is translated to bulk molybdenum disulfide (MoS2) using a top-down strategy. Chemical pre-reduction with butyllithium enables exfoliation and restacking in the presence of hexanediammonium (HDA) molecules, forming a pillared MoS2-HDA structure with an expanded interlayer spacing of 0.98 nm while preserving the bulk particle morphology and specific surface area. Electrochemical analysis reveals that improved rate capability primarily originates from the chemical activation associated with the pre-reduction step rather than from interlayer expansion itself. Operando X-ray diffraction and electrochemical dilatometry show that bulk MoS2 undergoes solvent co-intercalation in diglyme electrolyte, leading to pronounced lattice expansion and electrode swelling. In contrast, pillared MoS2-HDA suppresses solvent cointercalation despite its larger interlayer spacing, demonstrating that interlayer expansion alone does not dictate solvent co-intercalation in layered electrodes.

The electrochemical behavior of Nb–Ni binary phases in terms of electrocatalytic activity and stability under oxygen evolution reaction (OER) conditions was investigated. The outcomes of electrochemical studies in an alkaline electrolyte, supported by extensive characterization of electrode materials before and after electrochemical experiments, were provided. OER activities using the intermetallic compounds Nb7Ni6 and NbNi3 as electrocatalyst precursors were compared with those obtained with the solid solution of NbxNi1-x (x = 0.09) and a reference Ni foil under the same conditions. While Nb7Ni6 deactivates gradually during benchmarking stability test (chronopotentiometry at current density of 10 mA cm–2 for 2 h), NbNi3 possesses remarkable OER activity and keeps it stable not only under mild conditions of benchmarking experiment but also during long-term operation at elevated current densities (50, 100, and 200 mA cm–2). To understand the difference in chemical behavior and the OER performance of Nb–Ni compounds, extensive characterization was supported by carrying out quantum chemical calculations, shedding light on the charge distribution and chemical bonding in the investigated compounds.

The field of catalysis currently lacks a structured repository for experimental data, and the publication of machine-readable datasets remains uncommon. To address this gap and support FAIR (Findable, Accessible, Interoperable, and Reusable) data principles, we introduce a new plugin within the NOMAD platform for managing and publishing heterogeneous catalysis data, the nomad_catalysis plugin. This plugin enables the upload of structured experimental data and metadata with built-in visualization and alignment to the community-developed vocabulary Voc4Cat, ensuring long-term interpretability. In addition to facilitating efficient data sharing, the catalysis app offers intuitive search functionality, enabling researchers to quickly identify relevant catalytic reactions, catalyst materials, reaction conditions and kinetic properties. This infrastructure lays the foundation for advanced data analytics and machine learning applications, supporting more efficient and reproducible catalyst development.

Extending throughput capabilities and advanced sampling approaches are strongly accelerating catalyst discovery, increasingly performed within automated self-driving laboratories. The larger the tractable design spaces become though, the more questionable is the lasting value of individual optimal catalysts that are identified in black-box searches. Here, we demonstrate a sparse sampling approach that combines search efficiency with chemically interpretable insight into the topology of the design space. 1 Applied to the non-oxidative propane dehydrogenation reaction, it readily finds paretooptimal multi-promoter formulations that exceed the present industry reference in both yield toward the desired commodity product propylene and catalyst longevity. At the same time, it explains this superior performance in terms of individual promoter effects and synergistic promoter interactions. The latter interactions are missed in prevalent empirical single-promoter studies and are shown here as a key element toward further performance gains expected upon insight-motivated future modifications of the design space.

Effective kinetic models of heterogeneous catalytic processes are an indispensable tool for reactor design, optimization, and control. Under the assumption of using functional forms like power laws, model parameters are traditionally fitted to kinetic data measured along local line scans. A local line scan involves systematically varying one individual reaction parameter, such as a reactant concentration or temperature, at a time. This approach typically involves numerous separate kinetic measurements and is susceptible to the uncertainty of these line scans in determining the model's parameters. Here, we explore the use of profile reactors in combination with a fully automated adaptive design approach for an efficient identification of effective kinetic models. Originally developed to provide operando information along the axis of tubular reactors, profile reactors provide a complex line scan that encapsulates kinetic information across all reaction conditions probed along the tube. The proposed Model-Driven Adaptive Design with Profiles algorithm harnesses this extensive dataset to strategically guide the selection of initial reaction conditions for subsequent profile reactor measurements. This approach ensures that each line scan provides maximally complementary information, thereby significantly enhancing the efficiency and accuracy of kinetic model identification.

Abstract The deployment of artificial intelligence (AI) is transforming the scientific fields central to interdisciplinary catalysis research. By enabling more effective use of data, AI (including simpler machine learning and data science tools) holds great promise for accelerating discoveries. However, progress has so far been modest, largely due to the lack of standardized, machine‐readable, and openly shared catalysis data. This perspective, accounting for community insights emerging at conferences, analyses the underlying reasons for these challenges and proposes solutions to a future whereFAIR data management becomes an integral part of research in catalysis. In the short‐term, we deem that mandatory FAIR data depositing prior to scientific publications along with consensualized top‐down guidelines on data sharing powered by ease‐to‐use tools can make the necessary step change happen to catalyse data as key resource in our community.

Tip-enhanced Raman spectroscopy (TERS) is a powerful method for chemically identifying single molecules on surfaces. Although the mechanisms of the Raman signal enhancement have long been proposed, their precise evaluation and understanding remain challenging. This article introduces two key techniques to obtain single-molecule Raman signals with high reproducibility using TERS : (1) electromagnetic enhancement by a strong plasmonic field confined at atomic-scale structures of the tip apex, namely picocavities, and (2) chemical enhancement by forming an atomic point contact between the tip and the target molecule. These techniques have enabled TERS measurements on advanced and challenging samples, including molecular hydrogen physisorbed on metal surfaces and organic molecules chemisorbed on a non-plasmonic semiconductor substrate. Thus, characterizing and controlling plasmonic junctions at the atomic scale would further expand the TERS targetability.