Minisymposium

This is the age of separation. Conventional separation methods are energy-intensive and environmentally unsustainable, making adsorption-based techniques a promising alternative. Covalent organic frameworks (COFs) have emerged as highly tunable materials with modular structures, high surface areas, and tailored functionalities, making them ideal for gas separation. However, identifying optimal COFs remains challenging due to the vast number of possible structures. Computational screening provides an efficient solution but relies on standardized datasets of computation-ready experimental structures. While databases like CoRE-MOF offer valuable resources for MOFs, COFs remain underrepresented and lack standardized curation protocols. To address this, we developed the CURATED COFs database, which integrates refinement protocols and automated DFT-based structure optimization via the AiiDA workflow engine. Hosted on the Materials Cloud platform, CURATED COFs ensures open access, reproducibility, and transparency. This talk presents the CURATED COFs database, explores its curation workflows, and highlights its applications, demonstrating how systematic data curation can accelerate COF-based materials discovery.

Machine learning (ML) has emerged as a powerful tool for accelerating catalyst discovery by integrating computational and experimental data. While ML excels at pattern detection in large datasets, many catalyst studies rely on limited experimental data. Our approach combines ML with first-principles calculations to extract insights from small experimental datasets. We train a complex ML model on a large computational library of transition-state energies and complement it with simple linear regression models fitted to experimental data. This method allows us to explore the catalytic activity of monolayer bimetallic catalysts for ethanol reforming, identifying key reactions and predicting promising compositions.In another study, we applied ML-driven molecular dynamics and metadynamics simulations to investigate the oxygen evolution reaction (OER) on pristine and Ni-doped BaTiO3. Using a neural network potential, we captured dynamic mechanistic details, revealing the impact of nickel doping on BaTiO3, a perovskite oxide synthesized from earth-abundant precursors.

The transition to a sustainable energy economy hinges on advanced battery technologies that combine high energy density, rapid charging capabilities, and environmentally responsible materials. Nevertheless, ultrafast dynamics in battery cathodes remain poorly understood despite their crucial role in energy storage performance.In Mn or Ni based cathode materials, Jahn-Teller active centers induce structural distortions that fundamentally alter charge transport properties. LiMn₂O₄ (LMO) exemplifies this behavior, featuring a ferrodistortive tetragonal ground state that undergoes an order-disorder transition at 290K, where pseudorotations continuously reorient the Jahn-Teller axis. Electron correlation is accurately captured through tuned exact exchange fractions based on GW approximation to determine precise polaron migration barriers. Complex simulations combining Ab Initio Molecular Dynamics with high-level hybrid functional Nudged Elastic Band calculations are orchestrated using PerQueue, a modular workflow manager, to investigate polaron formation and migration across LMO's complex configurational landscape. Analysis reveals how pseudorotations influence polaron transport pathways in both ferrodistorted and thermally disordered states. By elucidating these structure-property relationships, our work provides critical mechanistic understanding for engineering improved battery materials with enhanced charge mobility.

Water electrolysis is a key enabler for meeting the global energy demands by sustainably producing molecular hydrogen and oxygen. While significant progress has been made in developing catalysts for the hydrogen evolution reaction and oxygen evolution reaction, the performance of electrochemical cells remains limited by the activity, selectivity, and durability of existing catalysts. This talk will highlight advances in the modeling of electrochemical materials and interfaces under realistic operating conditions to accelerate the discovery and optimization of electrocatalytic cells.