Parametrization of Gaussian approximation potential for the global optimization of magnesium clusters Mg<sub><i>N</i></sub> (<i>N</i> ≤ 100)
A two-stage GridSearch combined with active learning was employed to optimize GAP model parameters for Mg clusters, enabling reliable structural predictions in the extrapolative domain Mg<sub><i>n</i></sub>, <i>n</i> > 50. Global optimization using the parameterized GAP model revealed energetically favorable Mg<sub>51</sub>-Mg<sub>53</sub> clusters, showing early onset of pyramidal core formation previously reported only from Mg<sub>54</sub>. Global optimization identified new global minima candidates for the "magic" Mg<sub>59</sub>, Mg<sub>69</sub>, Mg<sub>74</sub> and Mg<sub>99</sub> clusters. The presence of hcp-like motifs does not significantly influence structural stability in clusters with Mg<sub><i>n</i></sub>, <i>n</i> < 100, as no structural differences were observed between GM "magic" clusters and others of similar size.
Authors
- Sergey V. Panteleev (ORCID: https://orcid.org/0000-0001-8780-5642)
- Stanislav K. Ignatov (ORCID: https://orcid.org/0000-0002-6039-3905)
- Igor Petukhov (ORCID: https://orcid.org/0009-0000-2365-4857)
- Ilya S. Steshin (ORCID: https://orcid.org/0000-0002-3330-716X)
Institutions
- Volga State University of Technology (RU)
- N. I. Lobachevsky State University of Nizhny Novgorod (RU)
Publication Details
- Journal
- Physical Chemistry Chemical Physics
- Published
- 2025-01-01
- DOI
- https://doi.org/10.1039/d5cp02189h
- Citations
- 2
- Primary Topic
- Machine Learning in Materials Science
- Type
- article
- Field-Weighted Citation Impact
- 1.04
Funders
- National Science Foundation