The research, carried out by Mohammad Fathianpour, Hossein Farrokhpour, and Kiamars Eskandari of Isfahan University of Technology, systematically examined magnesium clusters containing between two and twenty-one atoms. Their goal was to close a stubborn gap in the cluster literature. Decades of theoretical and experimental work had already established which magnesium clusters are unusually stable — the so-called magic clusters — and measured their binding energies, but the physical reasons why these particular sizes stand out remained murky. Previous studies largely tracked how average binding energies changed with cluster size without dissecting the actual interactions between individual atoms. The Iranian team set out to decompose those interactions atom by atom, pair by pair, to expose the electronic origins of stability.

The researchers optimized the geometry of each cluster using two different density functional theory methods, B3LYP and M06-2X, each paired with the 6-31+G(df) basis set, and confirmed that every optimized structure represented a true minimum on the potential energy surface through vibrational frequency analysis. The geometries themselves tell a story of gradual densification. The magnesium dimer is barely bound, with the two atoms separated by a wide 3.90 angstroms, held together only by weak van der Waals forces. As atoms are added, the bonds shorten: the three-atom cluster forms an equilateral triangle, the four-atom cluster adopts a compact tetrahedron with the shortest bonds among its neighbors, and by nine atoms a triangular prism core emerges that persists as a structural motif through larger sizes. Average bond lengths fall steadily as clusters grow, a signature of the slow march from molecular weak binding toward genuine metallic cohesion.

Stability analysis using three classic energetic descriptors — average binding energy per atom, stepwise binding energy, and the second difference of the total energy — confirmed earlier findings that specific cluster sizes punch above their weight. The tetramer Mg4, the decamer Mg10, and the twenty-atom Mg20 consistently emerged as the most robust candidates, with Mg7, Mg13, Mg15, and Mg17 showing size-dependent, more equivocal signs of stability. These sizes have long been associated with magic numbers in the jellium model, which treats metal clusters as electrons sloshing in a smooth positive background, filling shell-like energy levels. But the jellium picture alone could not explain the whole story, which is where the new study’s quantum atom analysis comes in.

The team’s central innovation was applying the Interacting Quantum Atoms scheme, a real-space energy partitioning method that splits the total electronic energy of a cluster into self-energies of individual atoms and interaction energies between atom pairs. Each self-energy contains the kinetic energy of the electrons in an atomic basin plus the classical electrostatic potential energy and the quantum exchange-correlation energy within that basin. Likewise, every pair of atoms shares a classical interaction and an exchange-correlation interaction. By tracking how these components evolved with cluster size, the researchers could finally see the ledger of gains and losses that determines whether a cluster is stable — something no simple binding energy plot can reveal.

The results were striking. As clusters grew, the average self-energy of each atom actually rose — meaning individual atoms became internally less stable — while the average interatomic interaction energy became increasingly stabilizing. In other words, the growing cohesion of a magnesium cluster is not a matter of atoms becoming more comfortable in their own basins; it is a matter of increasingly favorable quantum interactions between them. And for the magic clusters Mg4, Mg10, Mg17, and Mg20, this pattern was amplified to a remarkable degree. Their atoms showed more pronounced self-energy destabilization, but this was more than compensated by stronger interatomic exchange-correlation energy. The classical electrostatic component of interatomic interaction was actually destabilizing, meaning that the entire stabilizing role rests on the quantum exchange-correlation term — the energy credit atoms earn for sharing their electrons. For these magic sizes, the exchange-correlation stabilization peaks sharply and swamps the classical penalty.

Corroborating this picture, the quantum theory of atoms in molecules provided independent measures of electron sharing. The average delocalization index — the number of electrons shared between neighboring atomic basins — increased with cluster size and showed distinct local maxima precisely at Mg4, Mg10, Mg13, Mg17, and Mg20. Meanwhile, the average localization index, counting electrons confined within a single atom, fell correspondingly. In the four-atom cluster, electron delocalization jumped from a mere 1.29 percent in the dimer to 6.23 percent per atom. Analysis of the bond critical points between magnesium atoms showed rising electron density and increasingly negative Laplacians with cluster size, with maxima at Mg4, Mg10, and Mg17. Taken together, these topological indicators confirmed that the magic clusters are those in which electrons spread most effectively across the atomic framework, weaving the atoms into a coherent quantum whole.

The natural bond orbital analysis added a third, complementary perspective rooted in chemical bonding concepts. By comparing the energy of each cluster’s ideal Lewis structure with its actual self-consistent electronic structure, the researchers quantified how much stabilization comes from electron delocalization beyond classical localized bonds. This delocalization energy per atom grew with cluster size, again peaking at Mg4, Mg7, Mg10, Mg13, and Mg20. Second-order perturbation analysis revealed stronger donor-acceptor orbital interactions — charge flowing from occupied Lewis orbitals into unoccupied non-Lewis orbitals — at these same sizes. The convergence of three independent electronic analyses on the same handful of cluster sizes gave the team confidence in assigning magic character, though the integrated assessment classified only Mg4, Mg10, and Mg20 as robust magic clusters, with the others deemed descriptor-dependent.

Perhaps the most conceptually significant finding concerns what the results say about the nature of magnesium bonding itself. Combining the exchange-correlation energies with the electron density analyses showed that the character of magnesium-magnesium interaction evolves continuously with cluster size: small clusters are held together primarily by dispersion forces, the same weak attractions that bind noble gas atoms, but as clusters grow the rising exchange-correlation contribution and densifying bond critical points reveal an increasing dose of weak covalent character — genuine electron sharing. This gradual transition from van der Waals bonding toward diffuse, weak covalent bonding is the microscopic mechanism behind the emergence of metallic behavior in magnesium, a transition that photoelectron spectroscopy experiments have observed but that has been difficult to pin down theoretically at the level of individual atom pairs.

Beyond resolving a decades-old puzzle in cluster science, the work carries practical weight. Metal clusters are the seed particles of nanoscience, with applications ranging from catalysis to optical materials, and magnesium clusters in particular serve as benchmark systems for understanding how metallic bonding is born. The demonstration that quantum exchange-correlation energy — the fingerprint of electron delocalization — is the decisive stabilizing force offers a new design principle: to build exceptionally stable nanoclusters, one should engineer geometries that maximize electron sharing across the structure. The study also delivers a methodological lesson, showing that no single energetic descriptor can reliably crown a cluster as magic; only the convergence of structural, energetic, and electronic evidence, from IQA partitioning through QTAIM topology to NBO delocalization, provides a defensible verdict. As computational chemistry continues to sharpen its tools, the humble magnesium cluster has now revealed its innermost secret: stability at the nanoscale is not about atoms holding themselves together, but about electrons choosing to be shared.

Subject of Research: Electronic origins of stability in magic magnesium clusters analyzed via Interacting Quantum Atoms and electron delocalization methods

Article Title: Electronic origins of stability in magic Mg clusters: Interacting quantum atoms and electron delocalization perspectives

Article References: Fathianpour, M., Farrokhpour, H., & Eskandari, K. (2026). Electronic origins of stability in magic Mg clusters: Interacting quantum atoms and electron delocalization perspectives. Results in Chemistry, 30, Article 103851. https://doi.org/10.1016/j.rechem.2026.103851

Image Credits: AI Generated

DOI: 10.1016/j.rechem.2026.103851

Keywords: magnesium clusters, magic clusters, Interacting Quantum Atoms, electron delocalization, QTAIM, natural bond orbital, density functional theory, jellium model, van der Waals bonding, cluster stability, nanoscience, quantum chemistry