Bridging surface dynamics and phase transitions in CsPbI 3 perovskite with machine-learning force fields and dispersion interactions
In the rapidly evolving field of machine learning (ML) for atomistic simulations of materials and molecules, achieving transferability across different systems remains a key challenge. In this contribution, we demonstrate a case of transfer learning in which an ML model trained exclusively on surface slab data of the cubic ( α ) phase of CsPbI 3 perovskite can reproduce the key features of the dynamic behavior for different phases of the bulk material. Using an equivariant message-passing network (MACE) architecture, we construct an efficient ML force field trained on a dataset obtained from molecular dynamics simulations of a CsPbI 3 slab. The resulting model yields stable and reliable trajectories for both slab and bulk, reproducing the experimentally observed phase transitions of CsPbI 3 . We demonstrate that pairwise dispersion methods (TS/D3) mispredict the correct phase of CsPbI 3 bulk at ambient conditions, while more advanced approaches (MBD-NL/rVV10) recover both the experimentally observed phase sequence with increasing temperature (orthorhombic–tetragonal–cubic) and the corresponding phase-transition temperatures. Through a local analysis of the rotational driving forces acting on PbI 6 octahedra, we show that this discrepancy can be traced to the strong sensitivity of Cs and I atomic forces to the treatment of van der Waals (vdW) dispersion. Overall, our results demonstrate that surface–slab datasets augmented with an accurate treatment of vdW dispersion provide a robust foundation for constructing transferable machine-learning force fields that consistently describe both surface and bulk behavior across multiple length scales.
Authors
- Alexandre Tkatchenko (ORCID: https://orcid.org/0000-0002-1012-4854)
- Anton Charkin-Gorbulin (ORCID: https://orcid.org/0009-0002-2449-0043)
- Iryna Knysh (ORCID: https://orcid.org/0000-0001-8420-9431)
- Igor Poltavsky (ORCID: https://orcid.org/0000-0002-3188-7017)
- Claudio Quarti (ORCID: https://orcid.org/0000-0002-5488-1216)
- David Beljonne (ORCID: https://orcid.org/0000-0001-5082-9990)
Institutions
- University of Mons (BE)
- University of Luxembourg (LU)
Publication Details
- Journal
- Physical Review Materials
- Published
- 2026-11-16
- DOI
- https://doi.org/10.1103/qcvb-y84n
- Primary Topic
- Machine Learning in Materials Science
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Waalse Gewest
- Fonds National de la Recherche Luxembourg
- Fonds De La Recherche Scientifique - FNRS
- European Research Council