arXiv subjects

Yu Kumagai

Publications and source records attributed to Yu Kumagai.

At least 19 recordsLinked to original sources

arXiv2026

Atomic-scale phase-field modeling with dopants: Stochastic self-consistent harmonic approximation with fractional site occupancy

Phase-field modeling has achieved great success in predicting pattern formation in materials, such as the formation of ferroelectric domains. However, because it is typically based on continuum mechanics, conventional phase-field modeling cannot be straightforwardly applied to atomic-scale pattern formation, such as dopant segregation and vacancy ordering, which are driven by chemical potentials. Here, we extend the phase-field concept to the atomic scale by formulating the free energy of atomic systems within stochastic self-consistent harmonic approximation (SSCHA) theory to allow fractional site occupations. Our methodology enables us to directly calculate the derivative of the free energy with respect to site occupation and thereby obtain the chemical potential, successfully reproducing Ag distributions in bulk Cu as well as the resulting lattice expansion. Furthermore, we applied our methodology to investigate dopant segregation around a {\Sigma}5(310)[001] Cu grain boundary doped with Ag atoms. We found that Ag atoms preferentially segregate at the vertices of the triangular motif of a grain boundary. As the number of dopants increases, excess Ag atoms segregate near the vertices and then at the bottom sites of the triangular motif. This study extends the phase-field concept to discrete atomic systems, enabling the identification of preferential dopant-segregation sites and thereby visualizing atomic-scale pattern formation.

cond-mat.mtrl-sci↗

arXiv2026

Atomic-scale phase-field modeling for 2D ferroelectrics including non-Gaussian fluctuations

Atomic-scale phase-field modeling extends the phase-field framework down to the level of individual atoms by treating the probability density of atomic vibrations as a field variable and constructing a corresponding free-energy functional from this atomic field together with interatomic potentials. In this way, the framework has the potential to visualize local thermodynamic states with atomic-level resolution, just as conventional phase-field modeling has served as a computational microscope for free energy, stress, and related quantities at mesoscopic scales. However, existing formulations mainly assume Gaussian probability distributions for atomic vibrations, which limits their applicability to more complex and heterogeneous systems such as surfaces. In this work, we generalize the atomic-scale phase-field methodology by extending the free-energy functional to include non-Gaussian fluctuations. We apply this approach to monolayer SnTe in the NVT ensemble and show that the predicted equilibrium polarization is in better agreement with molecular dynamics simulations and that the ferroelectric-to-paraelectric phase transition is successfully reproduced. Furthermore, by decomposing the entropy on a per-atom basis, we visualize atomically resolved maps of local entropy and find that Sn atoms contribute more strongly to the entropy than Te atoms at high temperature, which is a driving factor of the ferroelectric-to-paraelectric phase transition. These results broaden the applicability of phase-field approaches to a wider range of atomic systems and suggest a route toward an atom-resolved theory of phase transitions based on high-resolution thermodynamics.

cond-mat.mtrl-sci↗

arXiv2026

Machine Learning Approaches to Point Defects in Non-Metallic Materials: A Review of Methods

We review recent machine-learning (ML) approaches for point defects in non-metallic materials, with an emphasis on defect formation energies. Existing studies largely fall into two categories: direct ML models that predict defect energetics from local structural representations, and machine-learning potentials (MLPs) that approximate the defect-containing potential energy surface. We summarize key achievements as well as persistent bottlenecks, emphasizing that dataset quality often dominates practical model performance. We further identify charged-defect formation energies as a central frontier, where Fermi-level alignment, finite-size corrections, and long-range electrostatics must be handled carefully and consistently to enable meaningful comparisons and transferable predictions across different materials.

cond-mat.mtrl-sci↗

arXiv2026

Alternating Target-Path Planning for Scalable Multi-Agent Coordination

The concurrent target assignment and pathfinding (TAPF) problem extends multi-agent pathfinding (MAPF) by asking planners to allocate distinct targets and collision-free paths to agents. Prior work on TAPF has relied exclusively on Conflict-Based Search (CBS), which tightly couples target assignment and pathfinding, resulting in compute-intensive, non-scalable solutions. In contrast, we propose an iterative refinement framework that decouples target assignment from pathfinding. Our framework builds on modern, fast, suboptimal MAPF solvers, such as LaCAM. Specifically, within a given time budget, it repeatedly solves MAPF for the current target assignment, identifies bottleneck agents via MAPF feedback, and refines the assignment. Empirical results show that feedback-driven reassignment loop is effective, enabling our framework to scale well beyond the reach of the state-of-the-art CBS-based solver while maintaining decent solution quality. This represents a solid step toward practical, large scale TAPF suitable for real-world setups.

cs.AI↗

arXiv2026

ZEBRA-Prop: A Zero-Shot Embedding-Based Rapid and Accessible Regression Model for Materials Properties

Large language models (LLMs) exhibit substantial potential across diverse scientific disciplines, including materials science. A property prediction framework, ZEBRA-Prop (Zero-Shot Embedding-Based Rapid and Accessible Regression Model for Materials Properties), is presented here as an extension of LLM-Prop. In contrast to LLM-Prop, which requires task-specific fine-tuning of the LLM, ZEBRA-Prop eliminates fine-tuning, thereby reducing computational cost and enabling rapid model training. The framework employs MatTPUSciBERT, an LLM specialized for materials science, to enhance predictive capability. Multiple textual embeddings are incorporated through a learnable weighting mechanism, which alleviates the context-length constraints inherent in LLM-Prop and facilitates effective integration of diverse textual representations. Evaluation is conducted using two datasets: the TextEdge dataset (approximately 140,000 entries) and an in-house dataset (approximately 2,000 entries) derived from the Materials Project database, with physical properties obtained from first-principles calculations. The predictive performance of ZEBRA-Prop is close to that of LLM-Prop, while the training time is reduced by approximately 95%. The performance improvements are attributable to three principal factors: domain-specific LLM utilization, diversified textual descriptions, and systematic text preprocessing. ZEBRA-Prop constitutes a scalable and computationally efficient framework for materials property prediction and supports accelerated materials discovery, particularly under limited computational resources.

cond-mat.mtrl-sci↗

arXiv2026

Charting the Landscape of Oxygen Ion Conductors: A 60-Year Dataset with Interpretable Regression Models

Oxygen ion conductors are indispensable materials for such as solid oxide fuel cells, sensors, and membranes. Despite extensive research across diverse structural families, systematic data enabling comparative analysis remain scarce. Here, we present a curated dataset of oxygen ion conductors compiled from $84$ experimental reports spanning $60$ years, covering $483$ materials. Each record includes activation energy ($E_a$) and prefactor ($A$) derived from Arrhenius plots, alongside detailed metadata on structure, composition, measurement method, and data source. When the original papers derive these using an erroneous Arrhenius equation $\sigma_T=A\exp{\left(-\frac{E_a}{RT}\right)}$, where ($\sigma_T$ is the oxygen ion conductivity at temperature $T$ and $R$ is the gas constant), we replotted these using the correct one, $\sigma_{T}T=A\exp{\left(-\frac{E_a}{RT}\right)}$. To illustrate how the database can be used, we constructed interpretable regression models for predicting oxygen ionic conductivity. Two symbolic regression models for E_a and A suggest that oxygen ion transport is primarily governed by local coordination environment and the electrostatic interactions, respectively. This dataset establishes a reliable foundation for data-driven discovery and predictive modeling of next-generation oxygen ion conductors.

cond-mat.mtrl-sci↗

arXiv2025

On-demand phase-field modeling: Three-dimensional Landau energy for HfO2 through machine learning

The unexpected emergence of ferroelectricity in HfO2 at reduced dimensions has attracted considerable attention, as it provides a pathway toward the realization of ultrasmall ferroelectric devices. Ab initio calculations suggest that this effect arises from a unique mode coupling, in which an antipolar displacement mode stabilizes a robust polar distortion. Based on these insights, Landau-Devonshire energy models have been proposed using such lattice modes as order parameters. However, most existing models are limited to a simplified one-dimensional model because of the computational cost of ab initio calculations and the limitations of conventional Landau polynomials. Here, we constructed a three-dimensional Landau-Devonshire potential for HfO2 by employing the tetragonal, antipolar, and polar modes as coupled order parameters, based on the latest machine-learning technologies. We generated a large-scale dataset of energies over a three-dimensional structural space, with the computational cost drastically reduced through the use of machine-learning interatomic potentials, and trained a multilayer perceptron (MLP) to learn the relationship between the order parameters and the energy. The energy predicted by the MLP successfully captures the characteristic coupling behavior whereby the antipolar modes induce the polar mode. Furthermore, by extending this MLP-based Landau potential to a position-dependent functional, that is, to a phase-field modeling framework, we revealed that the polarization magnitude in thin films decreases compared with the bulk state, while the critical strain required for the onset of spontaneous polarization increases due to surface effects. This study presents a new framework for the on-demand construction of Landau energy and phase-field modeling using the latest machine-learning techniques, enabling multiscale analysis of complex ferroelectric phenomena.

cond-mat.mtrl-sci↗

arXiv2025

Deep Learning-Based Extraction of Promising Material Groups and Common Features from High-Dimensional Data: A Case of Optical Spectra of Inorganic Crystals

We report an interpretation method for deep learning models that allows us to handle high-dimensional spectral data in materials science. The proposed method uses feature extraction and clustering analysis to categorize materials into classes based on similarities in both spectral data and chemical characteristics such as elemental composition and atomic arrangement. As a demonstration, we apply this method to an atomistic line graph neural network (ALIGNN) model trained on first-principles calculation data of 2,681 metal oxides, chalcogenides, and related compounds for optical absorption spectrum prediction. Our analysis reveals key elemental species and their coordination environments that influence optical absorption onset characteristics. The method proposed herein is broadly applicable to the classification and interpretation of diverse spectral data, extending beyond the optical absorption spectra of inorganic crystals.

cond-mat.mtrl-sci↗

arXiv2025

Oxygen-deficiency-driven phase segregation enables enhanced hole transport in amorphous tellurium oxides

Amorphous oxide semiconductors allow scalable electronics, yet high-mobility p-type counterparts remain rare because O-$2p$ valence bands are typically deep and spatially localized. Motivated by recent reports of unusually high hole mobilities in oxygen-deficient \ce{Se}-doped amorphous tellurium oxides ($a$-\ce{TeO$_x$}), we investigated $a$-\ce{TeO$_x$} with and without \ce{Se} doping using machine-learning-accelerated ab initio molecular dynamics with hybrid-functional defect calculations. We find that oxygen depletion drives nanoscale segregation into interpenetrating $a$-Te and $a$-TeO$_2$ domains with distinct roles: Te vacancies in oxide-like/interfacial environments supply holes, while transport is mediated by percolating Te-$5p$ pathways within the $a$-Te subnetwork. Upon doping, we theoretically verify that Se preferentially incorporates into the $a$-Te domains enhancing connectivity. This preference is nontrivial without explicit modeling given that \ce{Se} shares similar electronic structure with both Te and O. We further find that reducing the oxygen content can likewise enhance hole conductivity. Finally, using amorphous SeO$_x$, we show that domain segregation persists in other amorphous chalcogen oxides, suggesting a transferable route to achieving higher-mobility p-type amorphous oxides.

cond-mat.mtrl-sci↗

arXiv2025

Machine Learning Prediction of Charged Defect Formation Energies from Crystal Structures

Recent advances in materials informatics have expanded the number of synthesizable materials. However, screening promising candidates, such as semiconductors, based on defect properties remains challenging. This is primarily due to the lack of a general framework for predicting defect formation energies in multiple charge states from structural data. In this Letter, we present a protocol, namely data normalization, Fermi level alignment, and treatment of perturbed host states, and validate it by accurately predicting oxygen vacancy formation energies in three charge states using a single model. We also introduce a joint machine-learning model that integrates defect formation energies and band-edge predictions for virtual screening. Using this framework, we identify 89 hole-dopable oxides, including BaGaSbO, a potential ambipolar photovoltaic material. Our protocol is expected to become a standard approach for machine-learning studies on point defect formation energies.

cond-mat.mtrl-sci↗

arXiv2025

Physics-Based Factorized Machine Learning for Predicting Ionic Dielectric Tensors

Considerable effort continues to be devoted to the exploration of next-generation high-\k{appa} materials that combine a high dielectric constant with a wide band gap. However, machine learning (ML)-based virtual screening has remained challenging, primarily due to the low accuracy in predicting the ionic contribution to the dielectric tensor, which dominates the dielectric performance of high-\k{appa} materials. We here propose a joint ML model that predicts Born effective charges using an equivariant graph neural network, and phonon properties using a highly accurate pretrained ML potential. The ionic dielectric tensor is then computed analytically from these quantities. This approach significantly improves the accuracy of ionic contribution. Using the proposed model, we successfully identified 38 novel high-\k{appa} oxides from a screening pool of over 8,000 candidates.

cond-mat.mtrl-sci↗

arXiv2025

Atomic-scale phase-field modeling with universal machine learning potentials

Atomic-scale phase-field modeling formulates the probability densities of atomic vibrations as Gaussian distributions and derives a free energy functional using variational Gaussian theory and interatomic potentials. This framework permits per-Gaussian decomposition of the free energy, providing a description of local thermodynamic states with atomic resolution. However, existing formulations are limited to classical pairwise interatomic potentials, restricting their applicability to specific materials and compromising quantitative accuracy. In this work, we extend the atomic-scale phase-field methodology by incorporating universal machine learning interatomic potentials, thereby generalizing the free energy functional to many-body systems. This extension enhances both the accuracy and transferability of the approach. We demonstrate the method by applying it to bulk copper under NVT and NPT ensembles, where the predicted pressures and equilibrium lattice constants show excellent agreement with molecular dynamics simulations, validating the theoretical framework. Furthermore, we apply the method to {\Sigma}5(310)[001] grain boundaries in copper, enabling the visualization of local free energy distributions with atomic-scale resolution. The results reveal a pronounced free energy concentration at the grain boundary core, capturing the thermodynamic signature of the interface. This study establishes a versatile and accurate framework for atomic-scale thermodynamic modeling, significantly broadening the scope of phase-field approaches to include complex materials and defect structures.

cond-mat.mtrl-sci↗

arXiv2025

Investigation of Hole Dopability in Oxygen $2p$-Dominated Bands

The development of $p$-type oxide semiconductors remains impeded by the inherently low-lying valence-band maximum (VBM) dominated by O-2$p$ states. A prevailing approach to mitigate this limitation is to elevate the VBM by introducing cation states that hybridize with O-2$p$ orbitals or lie energetically above the O-2$p$ level. Nevertheless, the $p$-type oxides reported to date exhibit limited hole mobilities. To expand the search space, it is essential to accurately understand the intrinsic difficulty of introducing holes into O-2$p$-dominated bands. Accordingly, we evaluated 845 oxides to identify those in which holes can be doped into O-2$p$-dominated bands. Our high-throughput screening revealed CaCdO$_2$ as the only promising exemplar, in which the VBM is slightly hybridized with deep-lying Cd-3$d$ states. Our screening suggests that hole doping into O-2$p$-dominated bands is extremely difficult and thus reinforces the effectiveness of the traditional ``VBM-raising strategy.''

cond-mat.mtrl-sci↗

arXiv2024

Exploring Intrinsic and Extrinsic $p$-type Dopability of Atomically Thin $β$-TeO$_2$ from First Principles

Two-dimensional (2D) $β$-TeO$_2$ has gained attention as a promising material for optoelectronic and power device applications, thanks to its transparency and high hole mobility. However, the underlying mechanism behind its $p$-type conductivity and dopability remains unclear. In this study, we investigate the intrinsic and extrinsic point defects in monolayer and bilayer $β$-TeO$_2$, the latter of which has been experimentally synthesized, using the HSE+D3 hybrid functional. Our results reveal that most intrinsic defects are unlikely to contribute to $p$-type doping in 2D $β$-TeO$_2$. Moreover, Si contamination could further impair $p$-type conductivity. Since the point defects do not contribute to $p$-type conductivity, we propose two possible mechanisms for hole conduction: hopping conduction via localized impurity states, and substrate effects. We also explored substitutional $p$-type doping in 2D $β$-TeO$_2$ with 10 trivalent elements. Among these, the Bi dopant is found to exhibit a relatively shallow acceptor transition level. However, most dopants tend to introduce deep localized states, where hole polarons become trapped at Te's lone pairs. Interestingly, monolayer $β$-TeO$_2$ shows potential advantages over bilayers due to reduced self-compensation effects for $p$-type dopants. These findings provide valuable insights into defect engineering strategies for future electronic applications involving 2D $β$-TeO$_2$.

cond-mat.mes-hall↗

arXiv2024

Native defects and $p$-type dopability in transparent $\beta$-TeO$_2$: A first-principles study

Although $\beta$-TeO$_2$ is a promising $p$-type transparent conducting oxide (TCO) due to the large optical gap ($\sim$ 3.7 eV) and a light effective hole mass, its hole dopability still remains unexplored. In this work, electronic structure of $\beta$-TeO$_2$ and its point defects are investigated using the HSEsol functional with the band-gap-tuned mixing parameter. Our calculations reveal that $\beta$-TeO$_2$ exhibits a significant difference between the fundamental and optical band gaps because lower energy optical transitions are dipole forbidden. Additionally, it has a low hole effective mass, especially in-plane. The point defect calculations show that $\beta$-TeO$_2$ is intrinsically an insulator. From systematic calculations of the trivalent dopants as well as hydrogen, Bi doping is suggested as the best candidate as an acceptor dopant. This work paves the way for the material design of the $p$-type $\beta$-TeO$_2$.

cond-mat.mtrl-sci↗

arXiv2023

Finite-Size Corrections to Defect Energetics along One-Dimensional Configuration Coordinate

Recently, effective one-dimensional configuration coordinate diagrams have been utilized to calculate the line shapes of luminescence spectra and non-radiative carrier capture coefficients via point defects. Their calculations necessitate accurate total energies as a function of configuration coordinates. Although supercells under periodic boundary conditions are commonly employed, the spurious cell size effects have not been previously considered. In this study, we have proposed a correction energy formalism and verified its effectiveness by applying it to a nitrogen vacancy in GaN.

cond-mat.mtrl-sci↗

arXiv2023

Structural Properties of Two-Dimensional Strontium Titanate: A First-Principles Investigation

Motivated by the experimental synthesis of two-dimensional (2D) perovskite materials, we study the stability of 2D SrTiO$_3$ from first principles. We find that the TiO$_6$ octahedral rotations emerge in 2D SrTiO$_3$ with a rotation angle twice that in the 3D bulk. The rotation angle decreases significantly with the film thickness, reflecting the strong interlayer coupling that is absent in the conventional 2D materials. Using the molecular dynamics simulations, the cubic-like phase is found to appear above 1000 K that is much higher than the transition temperature of 3D SrTiO$_3$.

cond-mat.mtrl-sci↗

arXiv2022

Switchable Electric Dipole from Polaron Localization in Dielectric Crystals

Ferroelectricity in crystals is associated with the displacement of ions or rotations of polar units. Here we consider the dipole created by donor doping ($D^+$) and the corresponding bound polaron ($e^-$).A dipole of 6.15 Debye is predicted, from Berry phase analysis, in the Ruddlesden-Popper phase of ${\rm Sr_3Ti_2O_7}$. A characteristic double-well potential is formed, which persists for high doping densities. The effective Hubbard $U$ interaction can vary the defect state from metallic, a two-dimensional polaron, through to a zero-dimensional polaron. The ferroelectric-like behavior reported here is localized and distinct from conventional spontaneous lattice polarization.

cond-mat.mtrl-sci↗