Aug 2026· ACS Nano· Vol 20 33, pp.
23367-23380
· 0 citations· 34 references
Medicine
Abstract
The introduction of weakly solvating additives (WSAs) into aqueous electrolytes holds significant potential in promoting the desolvation kinetics of Zn2+ and reducing polarization. However, the current strategy of WSA screening primarily depends on trial-and-error experiments and theoretical calculations. This study reveals that, among various molecular features, the electrostatic potential minimum (ESPmin) exhibits the strongest correlation with the desolvation activation energy (Ea), demonstrating that ESPmin could serve as an effective descriptor for screening WSAs. Herein, we propose a data-driven screening strategy based on ESPmin, in which ESPmin can be accurately predicted directly from molecular structure using a graph convolutional neural network (GCN) model. Assisted by this strategy, several promising WSAs, including tetrahydropyran, methanol, and acetone, were successfully identified. As a proof of concept, tetrahydropyran was selected as the additive for systematic studies. Remarkably enhanced cycling stability with a lifespan of over 2400 h and low overpotential was demonstrated for the symmetric cell with tetrahydropyran. The proposed data-driven strategy enables the rapid screening of WSAs directly from molecular structures, offering an efficient pathway for the exploration of high-performance electrolyte additives.
Metal organic frameworks (MOFs) have emerged as promising electrode materials for supercapacitor (SC) due to their high surface areas, tunable porosity, and redox active sites. However, the vast chemical space of MOFs leads to millions of possible structures, makes experimental trial and error discovery inefficient. This review provides a focused perspective on how density functional theory (DFT) and machine learning (ML) are enabling the accelerated discovery and rational design of MOF-based SC electrodes. Key insights from DFT are discussed in relation to three critical performance descriptors as electrical conductivity, electrochemical and structural stability, and redox activity. In parallel, recent advances in ML-driven screening are reviewed, covering the development and use of large-scale MOF databases, descriptor engineering strategies, and predictive model architectures. Case studies demonstrating the successful integration of DFT and ML for identifying high-performance MOFs are highlighted in the review. Finally, the current limitations are analysed, including the discrepancy between idealized computational models and real polycrystalline electrodes, intrinsic trade-offs between conductivity and stability, and the need for interpretable and physics-informed ML models. Overall, this review outlines a computational roadmap for the rapid discovery and optimization of next-generation MOF-based SC electrodes. This is the comprehensive review on ML-driven prediction of electrochemical performance in MOF based electrode for SC applications. It integrates DFT-calculated electronic descriptors with ML models to discover hidden structure-property relationships. It identifies critical data gaps, model transferability issues, lack of dynamic ion-transport modelling in current studies. It proposed a multi-fidelity active learning framework combining DFT, ML and experiments for accelerated MOF discovery. It outlines standardized database protocols and explainable AI strategies to guide future high-performance MOF design.
Achal Siddharth Fulmali, H. Panda· Journal of Materials Science...· 0 citations
Anion exchange membrane water electrolyzers (AEMWEs) are promising for hydrogen production, yet their performance is bottlenecked by the alkaline hydrogen evolution reaction (HER) with sluggish kinetics induced by high water dissociation barriers and imbalanced H*/OH* adsorption-desorption. Herein, interpretable machine learning (ML) is exploited as a core tool for precise catalyst structure optimization, guiding the fabrication of a Ru2Ni3-carbon nanotubes (CNTs) hybrid catalyst. The ML-engineered catalyst exhibits high HER activity, with an ultra-low overpotential of 14 mV at 10 mA cm–2 and a Tafel slope of 34.3 mV dec–1. When integrated into an AEMWE with a NiFe-LDH anode, the system achieves 1.86 V at 1 A cm–2 (80 °C, no iR correction) and maintains stability for 200 h. Experimental and theoretical studies confirm that the ML-tailored Ru2Ni3-CNTs synergy modulates d-band centers, reduces reaction barriers, and optimizes intermediate adsorption, highlighting ML’s pivotal role in rational electrocatalyst design for advanced AEMWEs.
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
Aqueous zinc batteries (AZBs) lack a stable anion-derived solid electrolyte interphase (SEI) on the Zn anode, resulting in severe competition between Zn deposition and the hydrogen evolution reaction (HER). A conventional in-shell co-solvent coordinates strongly with Zn2+, displacing coordinated water and weakening Zn2+-anion interactions. This introduces a critical trade-off between HER suppression and anion-derived SEI formation. Here, we propose an out-of-shell co-solvent strategy that weakens Zn2+-H2O interactions, thereby enhancing Zn2+-anion interactions. To screen an optimal candidate, machine learning molecular dynamics (MLMD) was employed, achieving a ∼104-fold acceleration over ab initio molecular dynamics (AIMD) without sacrificing accuracy, and identifying N,N-dimethylacetamide (DMAC) from 28 candidates. In situ spectroscopic characterization further reveals that DMAC reconstructs the solvation environment, which facilitates desolvation and mitigates the formation of the inherently anion-lean interface. Consequently, this strategy promotes anion-derived SEI formation, synergistically suppressing HER. The DMAC electrolyte exhibits high Coulombic efficiency in Zn∥Cu cells (99.3% over 950 cycles) and long-term stability in Zn∥I2 full cells (12,000 cycles). Beyond demonstrating a rational electrolyte design, this work illustrates that MD simulations reform the traditional closed loop from material regulation to performance feedback, while ML integration accelerates screening. For bulk-interfacial solvation structure discrepancies, a feedback loop founded on dynamic interfacial processes regulates MLMD parameters, enabling more precise performance regulation.
Yaxin Ru, Feng Wang, Xiaoyu Yu et al.· Journal of the American Chem...· 0 citations
Metal–organic frameworks (MOFs) have emerged as versatile platforms for heterogeneous catalysis, owing to their structural tunability, well-defined active sites, and tailorable pore environments. Achieving efficient hydrogenation of complex molecules under mild conditions, however, remains challenging because the catalytic performance is highly sensitive to the local electronic and structural microenvironment of active sites. The stepwise hydrogenation of dicyclopentadiene (DCPD), which involves multiple unsaturated bonds and competing reaction pathways, provides a representative model system for elucidating these structure–activity relationships.
Over the past decade, MOF-based catalysts, including pristine frameworks, single-atom catalysts (SACs), and nanocluster/nanoparticle-loaded systems, have shown significant potential for hydrogenation reactions. Nevertheless, rational catalyst design remains difficult because framework topology, electronic structure, defect distribution, and metal–support interactions are strongly coupled across multiple length scales. As a result, establishing generalizable design principles through empirical approaches alone can be challenging.
In this Account, we present an integrated strategy for the rational design of MOF-based catalysts for DCPD hydrogenation by combining computational chemistry with artificial intelligence (AI). This framework bridges atomistic-level mechanistic understanding with synthesis optimization, enabling precise control of the catalytic performance.
At the atomic and electronic levels, density functional theory (DFT) calculations and descriptor-based analyses reveal key mechanistic factors governing hydrogen activation, charge redistribution, and reaction energetics. These studies guide the rational design of frustrated Lewis pairs (FLPs) in pristine MOFs through the optimization of charge distributions for heterolytic H2 cleavage. Extending this concept to SACs, electronic structure engineering demonstrates how metal–support interactions and d-band modulation regulate the adsorption behavior and reaction pathways. At larger scales, investigations of nanometer clusters and nanoparticle-loaded systems further highlight the critical role of interfacial electronic structures, including electron transfer and hydrogen spillover, in enhancing the catalytic activity.
Complementing these mechanistic insights, machine learning (ML) and large language model (LLM) driven approaches are employed to address the complexity of MOF synthesis. Data-driven models efficiently explore high-dimensional synthetic parameter spaces, enabling defect engineering and multiobjective optimization of catalytic properties. Furthermore, the integration of LLMs with retrieval-augmented generation (RAG) supports closed-loop synthesis frameworks in which the experimental design, data analysis, and knowledge extraction are autonomously coordinated. These AI-assisted methodologies significantly accelerate catalyst optimization and the precise regulation of active-site structures.
Through this integrated computational and AI-assisted strategy, MOF-based catalysts achieve complete conversion and high selectivity for DCPD hydrogenation under mild conditions. More broadly, this work establishes a generalizable paradigm in which mechanistic modeling and data-driven methodologies are synergistically integrated to guide the rational design of heterogeneous catalysts.
Yi-Fan Zhang, Zuoshuai Xi, Zhimeng Liu et al.· Accounts of Materials Resear...· 0 citations
CO2 capture and storage using amine-based solvents is a widely explored strategy in the literature aimed at mitigating the environmental impact associated with large-scale fossil fuel combustion. In the present work, four amines with distinct basicity levels were modeled in fifteen solvents with dielectric constants ranging from 1.88 (hexane) to 111 (formamide), in order to assess which implicit solvation approach, PCM, CPCM, or SMD, provides co-solvation results in closer agreement with theoretical and experimental data reported in the literature. A comprehensive analysis of implicit solvation effects was conducted, examining both the structural consequences of CO2 capture and the variations in stabilization energies associated with the formation of the zwitterionic intermediate. The results indicate that the SMD solvation model exhibits trends more consistent with literature data, owing to its sensitivity to local solute–solvent interactions, particularly hydrogen bonding. Notably, the SMD parameterization incorporates hydrogen-bond acidity and basicity descriptors derived from the Abraham solvation model, enabling correlation analyses between these parameters and the thermodynamic and solvation quantities. These findings provide deeper insight into the fundamental role of co-solvation in CO2 capture by amine-based solvents, particularly in reducing the free energy of stabilization of the zwitterionic state. Furthermore, this study identifies the implicit solvation approach that, when combined with DFT, yields result most consistent with established theoretical and experimental benchmarks reported in the literature. All calculations were performed at the DFT CAM-B3LYP/6–311++G(d,p) level of theory.
Jonathan de Brito Brum, José Walkimar de Mesquita Carneiro, L. D. da Costa· Journal of Molecular Modelin...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
