Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

MM: Fachverband Metall- und Materialphysik

MM 4: Materials for the Storage and Conversion of Energy

MM 4.8: Vortrag

Montag, 17. März 2025, 12:00–12:15, H22

Machine Learning Interatomic Potentials for Studying Hydrogen Storage in TiCr2 Laves Phase Alloys — •Pranav Kumar, Blazej Grabowski, and Yuji Ikeda — Institut für Materialwissenschaft, Universität Stuttgart, Pfaffenwaldring 55, 70569 Stuttgart, Germany

Efficient hydrogen storage is a crucial challenge for the widespread absorption of hydrogen as a clean energy carrier. TiCr2 Laves phase alloys, with their exceptional hydrogen absorption and diffusion properties, represent a promising class of materials for this purpose. This study employs a multiscale computational strategy, integrating density functional theory (DFT) and machine learning interatomic potentials (MLIPs), to investigate hydrogen behavior in TiCr2 Laves phases. Key areas of focus include hydrogen binding energies, diffusion pathways, and the complex effects of hydrogen clustering, which influence storage capacity and transport. Notably, our findings reveal possible equilibrium hydride structures consistent with experimental observations, offering valuable validation of the computational models. By leveraging large-scale DFT simulations within an active learning framework, we develop accurate MLIPs that enable efficient exploration of the phase space across diverse hydrogen concentrations. These MLIPs bridge atomic-scale simulations, facilitating precise calculations of hydrogen diffusion coefficients and shedding light on the mechanisms of hydrogen mobility. The insights derived from this work enhance our understanding of hydrogen-metal interactions and provide critical information for optimizing hydrogen storage materials.

Keywords: Machine learning interatomic potential; Laves Phase; Hydrogen Storage; Diffusion