Themed collection Advancing energy-materials through high-throughput experiments and computation

Non-precious metal catalysts for acidic OER typically require a high concentration of activity-promoting elements, e.g., Mn. We describe the high throughput discovery of quinary oxide catalysts with low Mn concentration via mixing with Sb, Sn, and Ti.

Catalytically dynamic NSP-CF@NCW electrode engineered by strategic integration of 3D Se and P-fused NSP microflakes with CF cubes docked NCW was studied as a trifunctional electrocatalyst for urea sensing and urea-assisted water splitting.

Generalization performance of machine learning models: (upper panel) generalization from small ordered to large disordered structures (SQS); (lower panel) generalization from low-order to high-order systems.

Machine-learning molecular dynamics provides predictions of structural and anharmonic vibrational properties of solid-state ionic conductors with ab initio accuracy. This opens a path towards rapid design of novel battery materials.

We demonstrate a link between workflow management and instrument automation tools, effectively bridging “trust” from tracking data provenance with automated “control” of experiments. We illustrate our approach using a battery cycling case study.

High-throughput screening and subsequent batch synthesis identified and characterized novel zeolitic imidazolate frameworks gels.

Prediction of crystal structures with desirable material properties is a grand challenge in materials research. We deployed graph theory assisted structure searcher and combined with universal machine learning potentials to accelerate the process.

A comprehensive inert-gas workflow for combinatorial aging studies gives insight into the intrinsic stability of hybrid perovskites under relevant aging conditions.

High-throughput MBE with cyclical growth and in situ etch increases experimental throughput by approximately 6× and substrate utilization by >40×.

The formation energy of gold nanoclusters could be predicted quickly by deep learning.

Introduction of an SQL and Python-based tool for managing research data from acquisition to publication. The method enables FAIR-compatible data management, minimizes user interaction, and provides customizability for diverse research domains.

This study identifies suitable coating materials that can prevent the electrode–electrolyte interfacial reaction to remove the obstruction in all-solid-state batteries composed of LiFePO₄ and sulfide solid electrolytes.

Active learning based on literature and experimental data enabled the discovery of highly active novel catalysts for ammonia synthesis. Pathway analysis implies that these activities have been achieved by both structural and electronic promotion.

This study highlighted the effectiveness of AI-driven multiobjective Bayesian optimization for electrocatalysis, accelerating the search for active and stable compositions for the acidic oxygen evolution reaction by 17x.

Li-ion batteries store electrical energy by electrochemically reducing Li ions from a liquid electrolyte in a graphitic electrode.

High-throughput synthesis was used to fabricate ternary Fe_xCo_yLa_z-based aerogel electrocatalysts for stoichiometric assessment. This work suggests a feasible way to find water-splitting non-precious metal electrocatalysts.

We propose a novel HTCS accelerated inverse design in a very large materials space combining the benefits of generative modeling, computationally efficient machine learning surrogate and high-quality physics-based simulation.

We have implemented a new reinforcement learning method able to rationally design unique metamaterial structures, which change shape during operational conditions. We have applied this to design nanostructured silicon anodes for Li-ion batteries.