Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
O: Fachverband Oberflächenphysik
O 56: Metal Substrates II
O 56.3: Vortrag
Mittwoch, 20. März 2024, 15:30–15:45, MA 043
Crystal growth of alkali and alkali earth metal on the transition metal surfaces — •Yuanyuan Zhou1 and Chunye Zhu2 — 1Technical University of Denmark, Copenhagen, Denmark — 2Guangdong University of Technology, Jieyang, China
Electrochemical nitrogen reduction is an attractive alternative to the Haber-Bosch process for making ammonia. The promising electrochemical system that produces large amounts of ammonia is the Li- and Ca-mediated process, which has achieved nearly 100% selectivity [Li et al. Joule 6, 1-19 (2022), Fu et al. Nat. Mater. (2023)]. However, the formation of the Li/Ca dendrites reduce the selectivity and deteriorate the stability of the Li- and Ca-mediated process. Therefore, a prerequisite for avoiding the dendrite formation is an in-depth understanding towards the growth mechanisms in an atomistic level.
We coupled our developed replica-exchange grand-canonical (REGC) [Zhou et al. Phys. Rev. B. 100, 174106(2019), Zhou et al. Phys. Rev. Lett. 128, 246101(2022)] method with machine-learned interatomic potentials (MLIPs) [Schütt et al. PMLR 139, (2021) ] to simulate larger length scale and time scale. The MLIPs is constructed in an autonomous active learning during the REGC simulations. We investigated the Li/Ca growth on the transition metal surface (Cu and Fe) using REGC MLIPs-accelerated molecular dynamics. This framework vividly showcased growth process of Li/Ca on Cu/Fe, pinpointing the origin of different growth behavior between Li and Ca.
Keywords: metal crystal growth; replica-exchange grand-canonical; ab initio molecular dynamics; machine learning interatomic potentials