Structures of selected H2O adsorption configurations in clusters (left) and monolayers (right). The ice structure was generated by global optimization. The labels indicate the adsorption energy and number of hydrogen bonds for the acceptor (A) and donor (D) types. The color code is shown in Figure 1. —physics.chem-ph
The binding energy (BE) of adsorbates on interstellar dust particles critically controls adsorption, desorption, diffusion, and surface reactivity, and therefore has a major impact on astrochemical models of star and planet forming regions.
Recent computational studies have increasingly reported complete distributions of BE rather than a single representative value, but these distributions are typically derived for either bare particle surfaces or thick water ice mantles. In this study, we bridge these situations by systematically investigating the BE distribution of water on partially and fully iced dust particle surfaces.
Machine learning interatomic potentials (MLIP) based on graph neural networks are used to model water adsorption onto Mg-terminated (010) surfaces of graphene and forsterite, representing carbonaceous and silicate particles, respectively. This model allows extensive sampling of adsorption sites on water clusters, monolayers, and bilayers produced under both crystalline (thermal treatment) and amorphous (low temperature) growth conditions.
At less than monolayer coverage, the chemistry of the underlying particles strongly influences both the ice morphology and the binding energy, with Mg-O interactions on the silicate surface producing particularly deep binding sites. After monolayer coating, the adsorption on both substrates is dominated by hydrogen bonds within the ice, reducing the influence of the particulate material.
Across all coverages, the amorphous ice structure systematically shifts the BE distribution toward stronger binding compared to crystalline ice, introducing highly stable defects and pocket sites. These results demonstrate that the BE distribution in submonolayer to few-layer ice regimes is wide-ranging and highly surface-dependent, providing physically motivated input for next-generation astrochemical models that incorporate surface heterogeneity.
Anant Vaishnav, Niels M. Mikkelsen, Mi Andersen
Subjects: Chemical Physics (physics.chem-ph); Galaxy Astrophysics (astro-ph.GA)
Quote: arXiv:2602.11050 [physics.chem-ph] (or arXiv:2602.11050v1 [physics.chem-ph] for this version)
https://doi.org/10.48550/arXiv.2602.11050
focus to learn more
Post history
Posted by: Anant Vaishnav
[v1] Wednesday, February 11, 2026 17:19:52 UTC (13,790 KB)
https://arxiv.org/abs/2602.11050
astrobiology,
