Machine-learning-accelerated materials discovery has yielded large numbers of computationally stable compounds, yet many remain experimentally unrealized, underscoring a persistent gap between prediction and synthesis. Here, we introduce a hierarchical screening framework that combines PBE-based thermodynamic stability, efficient dynamical-stability screening enabled by universal machine-learning interatomic potentials, and SCAN-based thermodynamic refinement. Applying this protocol to the 894 stable materials previously reported in Sci. Data 9, 302 (2022), we first curate 603 unique structures, of which only 298 remain thermodynamically stable on the complete PBE phase diagrams, demonstrating the critical role of competing phases in stability assessment. Dynamical screening then identifies 166 materials stable under both harmonic-phonon and finite-temperature molecular dynamics criteria, and SCAN phase diagrams further narrow this set to 109. Finally, by combining decomposition enthalpy with chemical-space completeness, we prioritize 25 candidates as high-confidence targets for experimental synthesis. This work provides a practical protocol for translating stability predictions into experimentally actionable synthesis targets, closing a key gap in machine-learning-driven materials discovery.
Machine learning (ML) is transforming materials discovery by enabling rapid prediction of properties that previously required computationally expensive first-principles calculations. Yet most current ML models remain fundamentally limited to zero-temperature descriptions, learning static lattice energies while neglecting the thermodynamic effects that govern materials behaviour at finite temperature. Because phase stability, functional response, and performance are governed by free-energy landscapes rather than static energies alone, this limitation represents a major barrier to predictive materials design under realistic operating conditions. In this Perspective, we argue that developing thermodynamics-informed ML constitutes one of the most important and least explored frontiers in materials discovery. We examine the fundamental shortcomings of energy-based models, highlighting the essential roles of entropy and anharmonicity in determining free energies and materials functionality. We review emerging strategies, including machine-learned interatomic potentials and hybrid ML-statistical mechanics frameworks, while identifying key challenges related to data availability, transferability, and thermodynamic consistency. Building on these advances, we outline a roadmap for thermodynamics-informed ML centred on direct free-energy learning, entropy-aware representations, and adaptive sampling across temperature. We highlight the transformative opportunities this paradigm offers for energy materials and argue that the next generation of ML models must move beyond static energy predictions towards a thermodynamic description of materials behaviour under realistic operating conditions.
Pol Benítez, Cibr'an L'opez, Claudio Cazorla· 0 citations
Crystal structure databases curated by high-throughput density functional theory calculations typically serve as the starting point for computational materials discovery efforts. Thermodynamic stability data, such as formation energies and the energy above the convex hull, are important quantities to guide the search for novel materials, enabling filtering for (meta)stable structures. Here, we present the thermodynamic stability of the fully open-source, reproducible, and experimentally focused Materials Cloud three-dimensional crystals database (MC3D). We compare against two other DFT databases, the Open Quantum Materials Database (OQMD) and the Materials Project (MP), as well as against experimental formation enthalpies. We then demonstrate how recent foundational machine learning interatomic potentials (MLIPs) trained at the r$^2$SCAN level (specifically, we test PET-OMATPES here) 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. Our results validate and extend the established practice of combining PBEsol geometries with meta-GGA energies to the era of foundational MLIPs. Finally, we train classical machine learning models to further correct the formation energies in a delta-learning framework, where we use the information-rich latent features of the foundational MLIP. These models further reduce the mean absolute error below 50 meV/atom, bringing it down to values comparable with the experimental uncertainty itself. Notably, compared to purely compositional features, the latent features (combined with carefully tuned regularization) simultaneously reduce the prediction error and limit the impact of the learned corrections on the relative phase stability.
Timo Reents, Marnik Bercx, Giovanni Pizzi· 0 citations
Phase diagrams encode the thermodynamic equilibria that govern alloy processing, but finite-temperature construction remains slow because candidate phases must be identified and their Gibbs free energies evaluated accurately. We report a machine-learning workflow that couples the crystal generator MatterGen with a fine-tuned MatterSim interatomic potential to expand the candidate phase space and compute temperature-dependent phase stability with accuracy approaching density functional theory. As demonstrated for the Li-Ga-Sn ternary system, the workflow constructs 0 and 300 K Gibbs phase-equilibrium diagrams and predicts a temperature-induced switch near the 3Li-2Ga-2Sn composition from the {LiGaSn, LiGa, Li8Sn3} assemblage to {LiGaSn, LiGa, LiSn} at approximately 230 K. X-ray diffraction of two synthesized compositions supports the predicted room-temperature assemblages. The approach offers a practical route for scalable finite-temperature phase-diagram construction and thermodynamic screening of intermetallic systems.
Chen Su, Jie Lu, Yuchen Fu et al.· Journal of Physical Chemistr...· 0 citations
The prediction of crystal structures is a key challenge in chemistry and materials science, but evolutionary crystal structure prediction (CSP) remains computationally expensive because it relies on repeated \textit{ab initio} relaxations and energy ranking. Machine learning interatomic potentials (MLIPs) can accelerate CSP, yet their use is limited by the need for large training sets and by the difficulty of choosing which candidate structures should be labeled by density functional theory (DFT). Here we introduce a self-consistent, foundation-model-assisted CSP workflow that combines evolutionary search with adaptive data selection and fine-tuning. Starting from a pretrained MLIP, the algorithm rapidly explores configuration space while iteratively selecting compact, representative, and physically relevant subsets of structures for DFT labeling, thereby reducing redundant calculations and improving a system-specific potential. We apply the method to the chemically complex Ca--Fe--Ni ternary system. The workflow reproduces the known low-pressure convex hull and enables efficient high-pressure exploration. It predicts a previously unreported compound, Ca$_6$FeNi, which becomes thermodynamically stable above 100~GPa. These results show that foundation-model-based, data-efficient CSP can greatly reduce computational cost while preserving accuracy and enabling the discovery of new materials in complex multicomponent systems.
N. Chtchelkatchev, M. Magnitskaya, R. Ryltsev· 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.
Metal phosphosulfides have emerged as unique multifunctional materials, but they present unique synthesis challenges compared to more established material classes such as oxides and nitrides. As a consequence, experimental development and theoretical understanding of phosphosulfides have focused on individual compounds rather than on accelerated broad-range exploration. In this work, we first evaluate the synthesizability and band gaps of 909 hypothetical ternary phosphosulfides by density functional theory. We find 19 previously unknown thermodynamically stable compounds, including the first Si- and Ge-based phosphosulfides. For rapid band gap prediction, we then develop a multi-fidelity machine learning model to translate semilocal density functional theory band gaps into experimentally calibrated band gaps. Importantly, we extend the accelerated material development workflow to the experimental domain by demonstrating a route to high-throughput synthesis and characterization of virtually any phosphosulfide material system. The method is based on thin-film combinatorial libraries and yields over 100 unique compositions in each experiment, enabling us to synthesize four distinct phosphosulfide compounds in only four combinatorial experiments without prior synthesis recipes and without compromising on material quality. Thus, we argue that accelerated materials development workflows combining theory, artificial intelligence, synthesis, and characterization can be viable even for experimentally challenging inorganic materials.
J. Sanz Rodrigo, Nicholas A. Kryger-Nelson, Lena A Mittmann et al.· Small· 0 citations