Continuous Metaheuristic Optimization of Carbon Nanoclusters: Global Minima and Finite-Temperature Thermodynamic Stability

Accurate structure prediction requires thorough energy landscape exploration, yet exhaustive search remains computationally intractable. While diverse strategies ranging from basin hopping to genetic algorithms have been implemented to map these landscapes, particle swarm optimization (PSO) has proven to be a particularly powerful approach for high-dimensional search. This capability is especially critical for systems like carbon nanostructures, where competing covalent and noncovalent interactions generate exceptionally rugged potential energy surfaces (PESs) densely populated with local minima. In this study, we systematically map the PES of carbon clusters (C3–C20) using PSO coupled with semi-empirical methods, benchmarking structural and energetic accuracy against density functional theory (DFT) and the machine-learned GAP-20 potential. We examine the impact of initialization schemes and sampling strategies within the PSO framework, incorporating tailored inertia schedules and a self-consistent field (SCF) restart protocol to mitigate non-convergence and premature trapping. Our approach successfully traces the structural evolution from linear chains (C3–C6) and cyclic intermediates (C7–C9) to polycyclic networks (C10–C19), culminating in the spontaneous, unbiased formation of a 3D canonical cage at C20. These results demonstrate the exploratory power of PSO and highlight the trade-offs between semi-empirical efficiency, DFT benchmark precision, and GAP-20 accuracy in cluster discovery.

Authors

Institutions

Publication Details

Journal
ChemRxiv
Published
2026-11-17
DOI
https://doi.org/10.26434/chemrxiv.15012403/v1
Primary Topic
Advanced Chemical Physics Studies
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Continuous Metaheuristic Optimization of Carbon Nanoclusters: Global Minima and Finite-Temperature Thermodynamic Stability

Alejandro Santana‐Bonilla, Carla de Tomás, Arden Tsang, Chris Lorenz et al.
ChemRxiv
Advanced Chemical Physics Studies
preprint

Continuous Metaheuristic Optimization of Carbon Nanoclusters: Global Minima and Finite-Temperature Thermodynamic Stability

Alejandro Santana‐Bonilla, Carla de Tomás, Arden Tsang, Chris Lorenz, Yao Wei, Anastasia Kuchmina
preprint en

Abstract

Accurate structure prediction requires thorough energy landscape exploration, yet exhaustive search remains computationally intractable. While diverse strategies ranging from basin hopping to genetic algorithms have been implemented to map these landscapes, particle swarm optimization (PSO) has proven to be a particularly powerful approach for high-dimensional search. This capability is especially critical for systems like carbon nanostructures, where competing covalent and noncovalent interactions generate exceptionally rugged potential energy surfaces (PESs) densely populated with local minima. In this study, we systematically map the PES of carbon clusters (C3–C20) using PSO coupled with semi-empirical methods, benchmarking structural and energetic accuracy against density functional theory (DFT) and the machine-learned GAP-20 potential. We examine the impact of initialization schemes and sampling strategies within the PSO framework, incorporating tailored inertia schedules and a self-consistent field (SCF) restart protocol to mitigate non-convergence and premature trapping. Our approach successfully traces the structural evolution from linear chains (C3–C6) and cyclic intermediates (C7–C9) to polycyclic networks (C10–C19), culminating in the spontaneous, unbiased formation of a 3D canonical cage at C20. These results demonstrate the exploratory power of PSO and highlight the trade-offs between semi-empirical efficiency, DFT benchmark precision, and GAP-20 accuracy in cluster discovery.

ChemRxiv
King's College London (GB), Czech Academy of Sciences (CZ), Institute of Physics of the Slovak Academy of Sciences (SK)
Advanced Chemical Physics Studies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon