Aug 2026· Journal of Physical Chemistry Letters· Vol 17 32, pp.
9192-9199
· 0 citations· 45 references
Medicine
TL;DR
This study developed an efficient framework integrating density functional theory with machine learning methods to predict surface energies in vdW crystals, by combining structural characteristics with elemental properties and found that the generative adversarial network achieved the best performance.
Abstract
Surface energy is a fundamental physical quantity that governs the stability and properties of van der Waals crystals, yet accurate estimation remains challenging due to the limitations of experimental and first-principles approaches. Herein we developed an efficient framework integrating density functional theory with machine learning methods to predict surface energies in vdW crystals. By combining structural characteristics with elemental properties, we trained several models and found that the generative adversarial network achieved the best performance (R2 = 96.97%, MSE = 1.693). Leveraging this model, we predicted surface energies for ∼800 vdW crystals, ranging from 0.67 to 42.47 meV/Å2. Further feature and bonding analysis revealed surface energy is significantly influenced by interlayer distance, atomic volume, and periodic elemental properties. Our study provides theoretical insights and a cost-effective, high-accuracy pathway for predicting surface energies, facilitating the design of 2D nanosheets and heterostructures.
This work demonstrates how recent foundational machine learning interatomic potentials (MLIPs) trained at the r$^2$SCAN level can be leveraged to improve the agreement of formation energies with experiment, reducing the mean absolute error by more than 40% relative to GGA without requiring any additional DFT calculation.
Timo Reents, Marnik Bercx, Giovanni Pizzi· 0 citations
Foundation machine-learning interatomic potentials (MLIPs) enable atomistic simulations at substantially lower computational cost than first-principles methods, but their reliability across structural geometries remains insufficiently understood. Here, we construct a density-functional-theory dataset of ZrO2 configurations spanning bulk, slab, particle, neck, and atomically thin wire environments motivated by an experimentally observed ZrO2 desintering process involving neck thinning and atomic wire formation. We first benchmark 26 pretrained MLIPs and observe pronounced geometry-dependent degradation in zero-shot predictions. Without any training, after only reference-energy alignment, the best zero-shot model (ORB-V3) reaches energy and force root-mean-square errors of 6 meV/atom and 197.3 meV/{\AA}, respectively, with the largest force errors in neck and wire configurations. We then compare zero-shot inference, fine-tuning, and training from scratch strategies. Fine-tuning yields lower energy and force errors than training from scratch, while both require comparable wall-clock time. Geometry-specific fine-tuning improves in-domain accuracy but frequently produces negative transfer to other structural classes, whereas mixed-geometry fine-tuning reduces cross-geometry errors. Evaluations of elastic and vibrational properties, surface energies, and neck dynamics further show that rankings based on average energy and force errors do not universally predict property-level behavior. These results demonstrate that geometry-diverse target data and independent physical validations are necessary when adapting foundation MLIPs to low-coordination (ionic) nanostructures.
P. Zanineli, B. Focassio, G. R. Schleder· 0 citations
This work not only establishes a pioneering paradigm for interpretable ML-driven force field refinement but also provides the first feature engineering solution incorporating chemical, physical, and structural information specifically designed for the machine learning of energetic molecular crystals.
Qi He, Pengju Wang, Xudong He et al.· Molecules· 0 citations
Metallenes have appealing properties, but stabilizing them in a monolayer phase poses challenges for their synthesis. A recent experiment showed that the van der Waals squeezing method can stabilize certain metallenes in a MoS2 sandwich. This pioneering work motivates systematic studies, but such studies are experimentally impractical, while first-principles modeling remains prohibitive. Here, armed with universal machine-learning interatomic potentials, we constructed 1620 metallene sandwich heterostructures containing 6 different sandwich layers and 45 metals. We performed phonon calculations, which revealed 1208 dynamically stable structures. We found that transition-metal dichalcogenides, particularly MoSe2, are highly effective in stabilizing metallenes. Specifically, buckled hexagonal and honeycomb crystal lattices exhibit the greatest stability. We further evaluated the thermal stability of selected heterostructures with density-functional theory molecular dynamics simulations at room temperature. By uncovering the physical and chemical factors governing the stabilization of metallenes, our results provide systematic insights to guide and accelerate synthesis for future applications.
Machine-learned interatomic potentials (MLIPs) have become an increasingly important tool for molecular dynamics (MD) simulations, enabling near quantum-mechanical accuracy at significantly reduced computational cost. Recent studies indicate that the Graph Atomic Cluster Expansion (GRACE) neural network architecture delivers strong performance in materials chemistry. In this work, we assess the GRACE architecture for the prediction of potential energy surfaces for organic molecules and introduce GRACE-OFF (GRACE Organic Force Field). GRACE models of varying depth (one-layer and two-layer) and size (small, medium, large) are trained on the SPICE v2.0 data set. We validate the resulting models using a variety of benchmarks. These include single-point energy and force predictions, torsional energy profiles, condensed phase properties of organic liquids and water (thermodynamic properties, self-diffusion coefficients, radial distribution functions, and temperature-dependent water density), as well as the stability of biomolecular MD simulations for gas-phase Ala15 and solvated crambin. For the single-molecule benchmarks (single point energies and forces, torsional energy profiles), the one-layer models showed only mediocre performance, whereas the two-layer models outperformed the MACE-OFF models to which we compare. For the condensed phase properties, the two-layer models gave consistently better results than the MACE-OFF family of MLIPs. For water and hexane, the GRACE-OFF models also beat the much more expensive small UMA/OMol25 (S) model. The two-layer GRACE-OFF models accurately reproduce experimental water radial distribution functions and predict water densities in close agreement with experimental data over a temperature range from 270 to 330 K. Benchmarks demonstrate that GRACE-OFF achieves higher MD performance than comparable MACE-OFF models in both single and double precision. This establishes GRACE-OFF as an accurate and computationally efficient foundation potential for routine simulations of organic liquids and biomolecular systems.
Anna Katharina Picha, Johannes Karwounopoulos, Linus C. Erhard et al.· Journal of Chemical Theory a...· 1 citation
Layering two-dimensional (2D) materials into van der Waals bilayers provides an effective method to achieve innovative quantum states and tunable electronic properties. Exploring the extensive configurational space resulting from various layer combinations, twist angles, and stacking patterns presents a significant computing challenge. As the volume of accessible datasets expands, training machine learning models on conventional hardware may ultimately become excessively expensive, prompting the advancement of quantum machine learning (QML) methodologies for materials discovery. This study presents a quantum-enhanced machine learning framework for predicting the bandgaps of van der Waals bilayers utilizing quantum simulators. We start with conventional feature selection to ascertain the most important descriptors affecting electronic properties. The properties are subsequently encoded into quantum states utilizing parameterized quantum circuits. We evaluated three quantum feature maps, namely ZFeatureMap, ZZFeatureMap, and PauliFeatureMap, and observed task-dependent performance. For the classification task, ZFeatureMap with two repetitions gave the strongest quantum-classifier performance, whereas for Quantum Support Vector Regressor (QSVR) bandgap regression, PauliFeatureMap with one repetition achieved the lowest average root-mean-square error (RMSE). These results indicate that feature-map choice should be optimized separately for classification and regression rather than being interpreted as the universal superiority of a single feature map. Using these feature-map comparisons, we employ a QSVR and a variational quantum regressor to predict the bandgaps of 1850 semiconducting vdW bilayers, illustrating that QML may attain significant predictive accuracy despite limited training data. Overall, the results demonstrate the feasibility of a simulator-based QML workflow for materials-property prediction. The quantum classifiers are competitive with the classical support vector classifier (SVC) baseline for zero-gap/non-zero-gap classification, whereas the classical Support Vector Regression (SVR) achieves the lowest RMSE in the regression task. Therefore, the present study should be interpreted as an exploratory benchmark of hybrid classical-quantum models for van der Waals (vdW) bilayers rather than as evidence of a definitive quantum performance advantage.
C. Chowdhury, L. Giordano· Journal of Chemical Physics· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
