Programme

P.102 -- Posters, VCTP-50

Date: Thursday, 7 August 2025

Time: 08:30 - 10:00

Searching for carbon cluster isomers using machine-learning potential

Nguyễn Duy Huy*, Nguyễn Hải Phong

VNU University of Science, Vietnam National University, Hanoi

We utilize the recently developed machine-learning potential, namely GAP-20, to predict the low-energy isomers of twelve-atom C clusters. While the GAP-20 aligns well with density-functional theory calculations for most isomers, accurately describing geometric structures and average C-C bond lengths, it significantly underestimates energies related to cage-like structures, leading to incorrect ground state predictions. Notably, the Jahn-Teller distortion associated with monocyclic rings is absent. These intriguing findings prompt further investigation. Additionally, GAP-20 identifies two novel low-energy isomers composed of multicyclic rings, challenging our existing knowledge and opening exciting avenues for future research. Overall, GAP-20 provides valuable insights into the complex landscape of carbon clusters by illuminating these multicyclic configurations.

Presenter: Nguyen Duy Huy