arXiv subjects

Hai-Chen Wang

Publications and source records attributed to Hai-Chen Wang.

10 recordsLinked to original sources

arXiv2026

High-throughput study of electrical conductivity in ordered metals

We present a computational framework that integrates machine learning with high-throughput ab initio calculations to screen over 2.8 million compounds for metallic transport. We identify several intermetallic candidates with predicted high conductivities comparable to that of aluminum (36.59 x $10^6$ S/m). We perform full electron-phonon coupling calculations for the top-performing materials, yielding results in good agreement with available experimental data. Our analysis reveals that while the noble metals (Ag, Au, Cu) possess a conductivity that remains difficult to surpass due to their unique electronic structure and low scattering, compounds like LiBePt2 can achieve comparable performance by utilizing valence electrons from light elements to shift high-scattering d-states beneath the Fermi level. This study not only identifies novel high-performance conductors but also demonstrates the predictive power of combining statistical learning with detailed ab initio calculations.

cond-mat.mtrl-sci↗

arXiv2026

Machine Learning Materials Properties by Encoding Orbital-Projected Density of States

Graph neural networks have become the dominant machine-learning architecture for predicting materials properties from crystal structures. Yet the initialization of atomic node features has received comparatively little attention, and conventional approaches rely on static elemental descriptors that carry no information about the quantum-mechanical electronic environment of each atom in its crystalline host. Here we show that augmenting atomic node representations with site-projected orbital density of states (pDOS) fingerprints, computed directly from density functional theory calculations, yields systematic and substantial improvements in predictive performance.These representations are fused with Pettifor elemental embeddings at each atomic site before message passing. For the superconducting critical temperature $T_c$ and the optical dielectric constant $ε_{\infty}$,the pDOS augmentation reduces prediction errors by 22.9% and 27.9%, respectively, relative to the elemental-descriptor baseline. These improvements are comparable to those achieved by doubling the training-set size. The gains are, however, contingent on training-set size. For the magnetic exchange energies of Heusler compounds, a substantially smaller dataset, the improvement is reduced,indicating that pDOS augmentation is most effective when the training data exceeds the length of the pDOS feature vector. We introduce an interpretable spectral attention-gating mechanism that reveals that the model autonomously learns to prioritize the orbital channels and energy windows most physically relevant to each target property. These results establish pDOS-augmented graph nodes as a broadly applicable strategy for infusing first-principles electronic-structure knowledge into graph networks, opening a practical route to high-accuracy property prediction in data-scarce regimes.

cond-mat.mtrl-sci↗

arXiv2026

AI-Driven Expansion and Application of the Alexandria Database

We present a novel multi-stage workflow for computational materials discovery that achieves a 99% success rate in identifying compounds within 100 meV/atom of thermodynamic stability, with a threefold improvement over previous approaches. By combining the Matra-Genoa generative model, Orb-v2 universal machine learning interatomic potential, and ALIGNN graph neural network for energy prediction, we generated 119 million candidate structures and added 1.3 million DFT-validated compounds to the ALEXANDRIA database, including 74 thousand new stable materials. The expanded ALEXANDRIA database now contains 5.8 million structures with 175 thousand compounds on the convex hull. Predicted structural disorder rates (37-43%) match experimental databases, unlike other recent AI-generated datasets. Analysis reveals fundamental patterns in space group distributions, coordination environments, and phase stability networks, including sub-linear scaling of convex hull connectivity. We release the complete dataset, including sAlex25 with 14 million out-of-equilibrium structures containing forces and stresses for training universal force fields. We demonstrate that fine-tuning a GRACE model on this data improves benchmark accuracy. All data, models, and workflows are freely available under Creative Commons licenses.

cond-mat.mtrl-sci↗

arXiv2025

Generative AI for Crystal Structures: A Review

As in many other fields, the rapid rise of generative artificial intelligence is reshaping materials discovery by offering new ways to propose crystal structures and, in some cases, even predict desired properties. This review provides a comprehensive survey of recent advancements in generative models specifically for inorganic crystalline materials. We begin by introducing the fundamentals of generative modeling and invertible material descriptors. We then propose a taxonomy based on architecture, representation, conditioning, and materials domain to categorize the diverse range of current generative AI models. We discuss data sources and address challenges related to performance metrics, emphasizing the need for standardized benchmarks. Specific examples and applications of novel generated structures are presented. Finally, we examine current limitations and future directions in this rapidly evolving field, highlighting its potential to accelerate the discovery of new inorganic materials.

cond-mat.mtrl-sci↗

arXiv2025

Universal Machine Learning Interatomic Potentials are Ready for Phonons

There has been an ongoing race for the past several years to develop the best universal machinelearning interatomic potential. This progress has led to increasingly accurate models for predictingenergy, forces, and stresses, combining innovative architectures with big data. Here, we benchmarkthese models on their ability to predict harmonic phonon properties, which are critical for under-standing the vibrational and thermal behavior of materials. Using around 10 000 ab initio phononcalculations, we evaluate model performance across various phonon-related parameters to test theuniversal applicability of these models. The results reveal that some models achieve high accuracyin predicting harmonic phonon properties. However, others still exhibit substantial inaccuracies,even if they excel in the prediction of the energy and the forces for materials close to dynamicalequilibrium. These findings highlight the importance of considering phonon-related properties inthe development of universal machine learning interatomic potentials.

cond-mat.mtrl-sci↗

arXiv2025

High-throughput study of kagome compounds in the AV3Sb5 family

The kagome lattice has emerged as a fertile ground for exotic quantum phenomena, including superconductivity, charge density waves, and topologically nontrivial states. While AV3Sb5 (A = K, Rb, Cs) compounds have been extensively studied in this context, the broader AB3C5 family remains largely unexplored. In this work, we employ machine-learning-accelerated, high-throughput density functional theory calculations to systematically investigate the stability and electronic properties of kagome materials derived from atomic substitutions in the AV3Sb5 structure. We identify 36 promising candidates that are thermodynamically stable, with many more close to the convex hull. Stable compounds are not only found with a pnictogen (Sb or Bi) as the C atom but also with Au, Hg, Tl, and Ce. This diverse chemistry opens the way to tune the electronic properties of the compounds. In fact, many of these compounds exhibit Dirac points, Van Hove singularities, or flat bands close to the Fermi level. Our findings provide an array of compounds for experimental synthesis and further theoretical exploration of kagome superconductors beyond the already known systems.

cond-mat.mtrl-sci↗

arXiv2025

The Maximum $T_c$ of Conventional Superconductors at Ambient Pressure

The theoretical maximum critical temperature ($T_c$) for conventional superconductors at ambient pressure remains a fundamental question in condensed matter physics. Through analysis of electron-phonon calculations for over 20,000 metals, we critically examine this question. We find that while hydride metals can exhibit maximum phonon frequencies of more than 5000 K, the crucial logarithmic average frequency $ω_\text{log}$ rarely exceeds 1800 K. Our data reveals an inherent trade-off between $ω_\text{log}$ and the electron-phonon coupling constant $λ$, suggesting that the optimal Eliashberg function that maximizes $T_c$ is unphysical. Based on our calculations, we identify Li$_2$AgH$_6$ and its sibling Li$_2$AuH$_6$ as theoretical materials that likely approach the practical limit for conventional superconductivity at ambient pressure. Analysis of thermodynamic stability indicates that compounds with higher predicted $T_c$ values are increasingly unstable, making their synthesis challenging. While fundamental physical laws do not strictly limit $T_c$ to low-temperatures, our analysis suggests that achieving room-temperature conventional superconductivity at ambient pressure is extremely unlikely.

cond-mat.supr-con↗

arXiv2023

Two-Dimensional Noble Metal Chalcogenides in the Frustrated Snub-Square Lattice

We study two-dimensional noble metal chalcogenides, with composition {Cu, Ag, Au}2{S, Se, Te}, crystallizing in a snub-square lattice. This is a semi-regular two-dimensional tesselation formed by triangles and squares that exhibits geometrical frustration. We use for comparison a square lattice, from which the snub-square tiling can be derived by a simple rotation of the squares. The mono-layer snub-square chalcogenides are very close to thermodynamic stability, with the most stable system (Ag2Se) a mere 7 meV/atom above the convex hull of stability. All compounds studied in the square and snub-square lattice are semiconductors, with band gaps ranging from 0.1 to more than 2.5 eV. Excitonic effects are strong, with an exciton binding energy of around 0.3 eV. We propose the Cu (001) surface as a possible substrate to synthesize Cu2Se, although many other metal and semiconducting surfaces can be found with very good lattice matching.

cond-mat.mtrl-sci↗

arXiv2023

Symmetry-based computational search for novel binary and ternary 2D materials

We present a symmetry-based exhaustive approach to explore the structural and compositional richness of two-dimensional materials. We use a ``combinatorial engine'' that constructs potential compounds by occupying all possible Wyckoff positions for a certain space group with combinations of chemical elements. These combinations are restricted by imposing charge neutrality and the Pauling test for electronegativities. The structures are then pre-optimized with a specially crafted universal neural-network force-field, before a final step of geometry optimization using density-functional theory is performed. In this way we unveil an unprecedented variety of two-dimensional materials, covering the whole periodic table in more than 30 different stoichiometries of form A$_n$B$_m$ or A$_n$B$_m$C$_k$. Among the found structures we find examples that can be built by decorating nearly all Platonic and Archimedean tesselations as well as their dual Laves or Catalan tilings. We also obtain a rich, and unexpected, polymorphism for some specific compounds. We further accelerate the exploration of the chemical space of two-dimensional materials by employing machine-learning-accelerated prototype search, based on the structural types discovered in the exhaustive search. In total, we obtain around 6500 compounds, not present in previous available databases of 2D materials, with an energy of less than 250~meV/atom above the convex hull of thermodynamic stability.

cond-mat.mtrl-sci↗

arXiv2022

Large-scale machine-learning-assisted exploration of the whole materials space

Crystal-graph attention networks have emerged recently as remarkable tools for the prediction of thermodynamic stability and materials properties from unrelaxed crystal structures. Previous networks trained on two million materials exhibited, however, strong biases originating from underrepresented chemical elements and structural prototypes in the available data. We tackled this issue computing additional data to provide better balance across both chemical and crystal-symmetry space. Crystal-graph networks trained with this new data show unprecedented generalization accuracy, and allow for reliable, accelerated exploration of the whole space of inorganic compounds. We applied this universal network to perform machine-learning assisted high-throughput materials searches including 2500 binary and ternary structure prototypes and spanning about 1 billion compounds. After validation using density-functional theory, we uncover in total 19512 additional materials on the convex hull of thermodynamic stability and ~150000 compounds with a distance of less than 50 meV/atom from the hull. Combining again machine learning and ab-initio methods, we finally evaluate the discovered materials for applications as superconductors, superhard materials, and we look for candidates with large gap deformation potentials, finding several compounds with extreme values of these properties.

cond-mat.mtrl-sci↗