東北大学研究者紹介

研究者詳細

ホーム

日本語 English

ジヤン　ソンフン

Jang Seonghoon

所属

未踏スケールデータアナリティクスセンター　ソーシャルインテグレーション研究部門

職名

准教授

学位

博士（工学）（東京大学）
修士（工学）（東北大学）

researchmap

https://researchmap.jp/jang.seonghoon

J-GLOBAL ID

202401019316906255

ORCID

https://orcid.org/0000-0001-6026-636X

論文 32

Digital hydrogen platform (DigHyd): a rigorously curated database for hydrogen storage materials empowered by AI-assisted literature mining

Seong-Hoon Jang, Di Zhang, Xue Jia, Hung Ba Tran, Linda Zhang, Ryuhei Sato, Yusuke Hashimoto, Toyoto Sato, Kiyoe Konno, Shin-ichi Orimo, Hao Li

Applied Physics A　2026年10月

DOI： 10.1007/s00339-026-09903-6 　
Physics‐Grounded Materials Artificial Intelligence for Reliable Materials Discovery

Yuhang Wang, Qian Wang, Seong‐Hoon Jang, Hao Li

Advanced Functional Materials　2026年9月3日

DOI： 10.1002/adfm.78119 　

ISSN：1616-301X 1616-3028

詳細を見る詳細を閉じる

<jats:title>ABSTRACT</jats:title> <jats:p>Artificial intelligence (AI) is transforming materials discovery, yet conventional data‐driven approaches often suffer from limited interpretability, poor extrapolation, and inconsistency with physical laws. Since materials behavior is fundamentally governed by thermodynamics, kinetics, electronic structure, transport processes, and operating environments, the next generation of materials intelligence must move beyond correlation‐based prediction toward physics‐grounded reasoning. In this Perspective, we systematically discuss Physics‐Grounded Materials AI (PhysMat AI) as a unifying perspective for integrating physical knowledge into materials intelligence through five complementary roles: physics as prior knowledge, descriptors, constraints, verifiers, and infrastructure. Using representative examples from catalysis, solid‐state electrolytes in solid‐state battery, and hydrogen‐storage materials, we illustrate how physical principles guide data representation, model reasoning, validation workflows, and knowledge management. We further present how AI agents can leverage these physics‐aware components to perform mechanism‐guided discovery within physically feasible search spaces. Finally, we outline a developmental roadmap from physics‐aware AI to physics‐reasoning AI and ultimately physics‐autonomous AI. Looking forward, materials intelligence should evolve from predictive models toward autonomous scientific systems capable of integrating physical reasoning, multiscale simulations, experimental validation, and continuous knowledge updating for reliable materials discovery.</jats:p>
Building a Physics-Aware AI Ecosystem for Solid-State Hydrogen Storage Materials

Seong-Hoon Jang, Yiwen Yao, Chuanyu Liu, Linda Zhang, Di Zhang, Xue Jia, Hung Ba Tran, Eric Jianfeng Cheng, Ryuhei Sato, Yusuke Ohashi, Toyoto Sato, Yusuke Hashimoto, Mark D. Allendorf, Nongnuch Artrith, Marcello Baricco, Andreas Borgschulte, Darren P. Broom, Ang Cao, Benjamin Wei Jie Chen, Lixin Chen, Ping Chen, Eun Seon Cho, Stefano Deledda, Zhao Ding, Martin Dornheim, Michael Felderhoff, Yaroslav Filinchuk, George E. Froudakis, Mingxia Gao, Thomas Gennett, Zaiping Guo, Ikutaro Hamada, Jason Hattrick-Simpers, Bjørn C. Hauback, Michael Hirscher, Torben R. Jensen, Baohua Jia, Hyoung Seop Kim, Takahiro Kondo, Kentaro Kutsukake, Xiao-Yan Li, Tongliang Liu, Piao Ma, Jianfeng Mao, Rana Mohtadi, Hyunchul Oh, Mark Paskevicius, Chris J. Pickard, Astrid Pundt, Anibal J. Ramirez-Cuesta, Hiroyuki Saitoh, Kaihang Shi, Aloysius Soon, Chenghua Sun, Chris Wolverton, Hiroshi Yabu, Weijie Yang, Zhenpeng Yao, Xuebin Yu, Jianxin Zou, Shouyi Hu, Panpan Zhou, Xi Lin, Zhigang Hu, Zhenhao Zhou, Pengfei Ou, Jiayu Peng, Shin-ichi Orimo, Hao Li

ACS Energy Letters　2026年8月14日

DOI： 10.1021/acsenergylett.6c01856 　
Competing hydrogenation pathways to metastable CaH6 revealed by machine learning potential molecular dynamics 査読有り

Ryuhei Sato, Peter I. C. Cooke, Maélie Caussé, Hung Ba Tran, Seong Hoon Jang, Di Zhang, Hao Li, Shin-ichi Orimo, Yasushi Shibuta, Chris J. Pickard

Physical Review B　2026年8月5日

DOI： 10.1103/pb7t-5bj1 　
Digital battery platform: extending the dynamic database of solid-state electrolytes to a diversified electrolyte database with ion migration models 査読有り

Qian Wang, Hanghui Liu, Seong Hoon Jang, Di Zhang, Ryuhei Sato, Kazuaki Kisu, Yusuke Hashimoto, Eric Jianfeng Cheng, Shin-ichi Orimo, Hao Li

Nano Materials Science　2026年7月

DOI： 10.1016/j.nanoms.2026.07.001 　

ISSN：2589-9651
Charting the Landscape of Oxygen Ion Conductors: A 60-Year Dataset with Interpretable Regression Models 査読有り

Seong-Hoon Jang, Shin Kiyohara, Hitoshi Takamura, Yu Kumagai

Scientific Data　2026年4月1日

DOI： 10.1038/s41597-026-07100-x 　

ISSN：2052-4463
A unified descriptor framework for hydrogen storage capacity and equilibrium pressure in interstitial hydrides 査読有り

Seong-Hoon Jang, Di Zhang, Xue Jia, Hung Ba Tran, Linda Zhang, Ryuhei Sato, Yusuke Hashimoto, Yusuke Ohashi, Toyoto Sato, Kiyoe Konno, Shin-ichi Orimo, Hao Li

Chemical Science　2026年

DOI： 10.1039/D6SC03089K 　
Discovering new materials knowledge from “old data” 査読有り

Yuhang Wang, Qian Wang, Seong-Hoon Jang, Eric Jianfeng Cheng, Hao Li

Chemical Communications　2026年

DOI： 10.1039/D6CC01716A 　
Digital materials ecosystem: from databases to AI agents for autonomous discovery 査読有り

Di Zhang, Xue Jia, Yuhang Wang, Heng Liu, Qian Wang, Seong-Hoon Jang, Daksh Shah, Songbo Ye, Hung Ba Tran, Hao Li

Chemical Science　2026年

DOI： 10.1039/D5SC09229A 　
“DIVE” into hydrogen storage materials discovery with AI agents 査読有り

Di Zhang, Xue Jia, Hung Ba Tran, Seong Hoon Jang, Linda Zhang, Ryuhei Sato, Yusuke Hashimoto, Toyoto Sato, Kiyoe Konno, Shin-ichi Orimo, Hao Li

Chemical Science　2026年
出版者・発行元： arXiv
DOI： 10.1039/D5SC09921H 　

詳細を見る詳細を閉じる

Data-driven artificial intelligence (AI) approaches are fundamentally transforming the discovery of new materials. Despite the unprecedented availability of materials data in the scientific literature, much of this information remains trapped in unstructured figures and tables, hindering the construction of large language model (LLM)-based AI agent for automated materials design. Here, we present the Descriptive Interpretation of Visual Expression (DIVE) multi-agent workflow, which systematically reads and organizes experimental data from graphical elements in scientific literatures. We focus on solid-state hydrogen storage materials-a class of materials central to future clean-energy technologies and demonstrate that DIVE markedly improves the accuracy and coverage of data extraction compared to the direct extraction by multimodal models, with gains of 10-15% over commercial models and over 30% relative to open-source models. Building on a curated database of over 30,000 entries from 4,000 publications, we establish a rapid inverse design workflow capable of identifying previously unreported hydrogen storage compositions in two minutes. The proposed AI workflow and agent design are broadly transferable across diverse materials, providing a paradigm for AI-driven materials discovery.
Mott insulating state of IrO6 honeycomb monolayer structured in ilmenite-type oxide superlattice 査読有り

Masamichi Negishi, Kohei Fujiwara, Seong-Hoon Jang, Yi-Feng Zhao, Shun Sasano, Ryo Ishikawa, Naoya Shibata, Daisuke Shiga, Hiroshi Kumigashira, Yuto Nakamura, Hideo Kishida, Yukitoshi Motome, Atsushi Tsukazaki

Physical Review Materials　2025年8月12日

DOI： 10.1103/gz4v-8mds 　
Dissimilar Diffusion Mechanisms of Li+, Na+, and K+ Ions in Anhydrous Fe-Based Prussian Blue Cathode 査読有り

Dan Ito, Seong-Hoon Jang, Hideo Ando, Toshiyuki Momma, Yoshitaka Tateyama

Journal of the American Chemical Society　2025年7月23日

DOI： 10.1021/jacs.5c05274 　
Dual influence of protonation on Li-ion transport in garnet solid electrolytes: A first-principles study 査読有り

Feye-Feng Lu, Huu Duc Luong, Seong-Hoon Jang, Randy Jalem, Yoshitaka Tateyama, Hong-Kang Tian

Journal of Power Sources　2025年2月

DOI： 10.1016/j.jpowsour.2024.235906 　
GoodRegressor: A General-Purpose Symbolic Regression Framework for Physically Interpretable Materials Modeling

Seong-Hoon Jang

2025年
出版者・発行元： arXiv
DOI： 10.48550/ARXIV.2510.18325 　

詳細を見る詳細を閉じる

Symbolic regression offers a promising route toward interpretable machine learning, yet existing methods suffer from poor predictability and computational intractability when exploring large expression spaces. I introduce GoodRegressor, a general-purpose C++-based framework that resolves these limitations while preserving full physical interpretability. By combining hierarchical descriptor construction, interaction discovery, nonlinear transformations, statistically rigorous model selection, and stacking ensemble, GoodRegressor efficiently explores symbolic model spaces such as $1.44 \times 10^{457}$, $5.99 \times 10^{124}$, and $4.20 \times 10^{430}$ possible expressions for oxygen-ion conductors, NASICONs, and superconducting oxides, respectively. Across these systems, it produces compact equations that surpass state-of-the-art black-box models and symbolic regressors, improving $R^2$ by $4\sim40$ %. The resulting expressions reveal physical insights, for example, into oxygen-ion transport through coordination environment and lattice flexibility. Independent ensemble runs yield nearly identical regressed values and the identical top-ranked candidate, demonstrating high reproducibility. With scalability up to choices without interaction terms, GoodRegressor provides a foundation for general-purpose interpretable machine intelligence.
Physically interpretable descriptors drive the materials design of metal hydrides for hydrogen storage 査読有り

Seong-Hoon Jang, Di Zhang, Hung Ba Tran, Xue Jia, Kiyoe Konno, Ryuhei Sato, Shin-ichi Orimo, Hao Li

Chemical Science　2025年
出版者・発行元： arXiv
DOI： 10.1039/D5SC07296D 　

ISSN：2041-6520 2041-6539

詳細を見る詳細を閉じる

Designing metal hydrides for hydrogen storage remains a longstanding challenge due to the vast compositional space and complex structure-property relationships. Herein, for the first time, we present physically interpretable models for predicting two key performance metrics, gravimetric hydrogen density $w$ and equilibrium pressure $P_{\rm eq,RT}$ at room temperature, based on a minimal set of chemically meaningful descriptors. Using a rigorously curated dataset of $5,089$ metal hydride compositions from our recently developed Digital Hydrogen Platform (\it{DigHyd}) based on large-scale data mining from available experimental literature of solid-state hydrogen storage materials, we systematically constructed over $1.6$ million candidate models using combinations of scalar transformations and nonlinear link functions. The final closed-form models, derived from $2$-$3$ descriptors each, achieve predictive accuracies on par with state-of-the-art machine learning methods, while maintaining full physical transparency. Strikingly, descriptor-based design maps generated from these models reveal a fundamental trade-off between $w$ and $P_{\rm eq,RT}$: saline-type hydrides, composed of light electropositive elements, offer high $w$ but low $P_{\rm eq,RT}$, whereas interstitial-type hydrides based on heavier electronegative transition metals show the opposite trend. Notably, Be-based systems, such as Be-Na alloys, emerge as rare candidates that simultaneously satisfy both performance metrics, attributed to the unique combination of light mass and high molar density for Be. Our models indicate that Be-based systems may offer renewed prospects for approaching these benchmarks. These results provide chemically intuitive guidelines for materials design and establish a scalable framework for the rational discovery of materials in complex chemical spaces.
Predicting Room‐Temperature Conductivity of Na‐Ion Super Ionic Conductors with the Minimal Number of Easily‐Accessible Descriptors 査読有り

Seong‐Hoon Jang, Randy Jalem, Yoshitaka Tateyama

Advanced Energy and Sustainability Research　2024年12月

DOI： 10.1002/aesr.202400158 　
Proton Intercalation into an Open‐Tunnel Bronze Phase with Near‐Zero Volume Change 査読有り

Kosuke Kawai, Seong‐Hoon Jang, Yuta Igarashi, Koji Yazawa, Kazuma Gotoh, Jun Kikkawa, Atsuo Yamada, Yoshitaka Tateyama, Masashi Okubo

Angewandte Chemie　2024年10月26日

DOI： 10.1002/ange.202410971 　

ISSN：0044-8249 1521-3757

詳細を見る詳細を閉じる

<jats:title>Abstract</jats:title><jats:p>Managing safety and supply‐chain risks associated with lithium‐ion batteries (LIBs) is an urgent task for sustainable development. Aqueous proton batteries are attractive alternatives to LIBs because using water and protons addresses these two risks. However, most host materials undergo large volume changes upon H<jats:sup>+</jats:sup> intercalation, which induces intraparticle cracking to accelerates parasitic reactions. Herein, we report that Mo<jats:sub>3</jats:sub>Nb<jats:sub>2</jats:sub>O<jats:sub>14</jats:sub> bronze exhibits reversible H<jats:sup>+</jats:sup> intercalation (200 mAh g<jats:sup>−1</jats:sup>) with a Coulombic efficiency of 99.7 % owing to near‐zero volume change and solid‐solution‐type phase transition. Combination of experimental and theoretical analyses clarifies that rotation and shrinkage of open tunnels, which consist of flexible corner‐sharing Mo/NbO<jats:sub><jats:italic>n</jats:italic></jats:sub> polyhedra, relieve local structural distortions upon H<jats:sup>+</jats:sup> intercalation to suppress intraparticle cracking. The prototype full cell of an aqueous proton battery with a Mo<jats:sub>3</jats:sub>Nb<jats:sub>2</jats:sub>O<jats:sub>14</jats:sub> anode operates stably over 1000 charge/discharge cycles. This study reveals the importance of implementing distortion‐relieving voids in host materials to reduce volume change upon charge/discharge.</jats:p>
Exchange interactions in the rare-earth magnets A2PrO3 (A=alkali metals) 査読有り

Seong-Hoon Jang, Yukitoshi Motome

Physical Review B　2024年10月10日

DOI： 10.1103/PhysRevB.110.155124 　
Exploring rare-earth Kitaev magnets by massive-scale computational analysis 査読有り

Seong-Hoon Jang, Yukitoshi Motome

Communications Materials　2024年9月18日

DOI： 10.1038/s43246-024-00634-w 　
Computational discovery of stable Na-ion sulfide solid electrolytes with high conductivity at room temperature 査読有り

Seong-Hoon Jang, Randy Jalem, Yoshitaka Tateyama

Journal of Materials Chemistry A　2024年

DOI： 10.1039/D4TA02522A 　
Metallic ruthenium ilmenites: First-principles study of MgRuO₃ and CdRuO₃ 査読有り

Jang, S.-H., Motome, Y.

AIP Advances　14　(1)　2024年1月1日

DOI： 10.1063/5.0185801 　

ISSN：2158-3226
Multiobjective Solid Electrolyte Design of Tetragonal and Cubic Inverse-Perovskites for All-Solid-State Lithium-Ion Batteries by High-Throughput Density Functional Theory Calculations and AI-Driven Methods 査読有り

Jalem, R., Tateyama, Y., Takada, K., Jang, S.-H.

Journal of Physical Chemistry C　127　(35)　2023年9月7日

DOI： 10.1021/acs.jpcc.3c02801 　

ISSN：1932-7455 1932-7447
EwaldSolidSolution: A High-Throughput Application to Quickly Sample Stable Site Arrangements for Ionic Solid Solutions 査読有り

Jang, S.-H., Jalem, R., Tateyama, Y.

Journal of Physical Chemistry A　127　(27)　2023年7月13日

DOI： 10.1021/acs.jpca.3c00076 　

ISSN：1520-5215 1089-5639
3．第一原理計算を基軸とした高イオン伝導度固体電解質のハイスループット探索と記述子抽出査読有り

Yoshitaka TATEYAMA, Seong-Hoon JANG, Randy JALEM

Denki Kagaku　91　(2)　150-154　2023年6月5日
出版者・発行元： The Electrochemical Society of Japan
DOI： 10.5796/denkikagaku.23-fe0012 　

ISSN：2433-3255

eISSN：2433-3263
キタエフスピン液体の物質設計招待有り査読有り

張成燻, 求幸年

個体物理　57　(11)　613-784　2022年11月
High-Throughput Data-Driven Prediction of Stable High-Performance Na-Ion Sulfide Solid Electrolytes 査読有り

Jang, S.-H., Tateyama, Y., Jalem, R.

Advanced Functional Materials　32　(48)　2022年11月

DOI： 10.1002/adfm.202206036 　

ISSN：1616-3028 1616-301X
Electronic and magnetic properties of iridium ilmenites 査読有り

Jang, S.-H., Motome, Y.

Physical Review Materials　5　(10)　2021年10月20日

DOI： 10.1103/PhysRevMaterials.5.104409 　

ISSN：2475-9953
Vortex creation and control in the Kitaev spin liquid by local bond modulations 査読有り

Jang, S.-H., Kato, Y., Motome, Y.

Physical Review B　104　(8)　2021年8月23日
出版者・発行元： American Physical Society ({APS})
DOI： 10.1103/PhysRevB.104.085142 　

ISSN：2469-9969 2469-9950
Computational design of f -electron Kitaev magnets: Honeycomb and hyperhoneycomb compounds A2PrO3 (A= alkali metals) 査読有り

Jang, S.-H., Sano, R., Kato, Y., Motome, Y.

Physical Review Materials　4　(10)　2020年10月30日
出版者・発行元： American Physical Society ({APS})
DOI： 10.1103/PhysRevMaterials.4.104420 　

ISSN：2475-9953
Materials design of Kitaev spin liquids beyond the Jackeli–Khaliullin mechanism 査読有り

Yukitoshi Motome, Ryoya Sano, Seonghoon Jang, Yusuke Sugita, Yasuyuki Kato

Journal of Physics: Condensed Matter　32　(40)　404001-404001　2020年6月30日
出版者・発行元： IOP Publishing
DOI： 10.1088/1361-648x/ab8525 　

ISSN：0953-8984

eISSN：1361-648X
Antiferromagnetic Kitaev interaction in f-electron based honeycomb magnets 査読有り

Seong-Hoon Jang, Ryoya Sano, Yasuyuki Kato, Yukitoshi Motome

Physical Review B　99　(24)　2019年6月10日
出版者・発行元： American Physical Society (APS)
DOI： 10.1103/physrevb.99.241106 　

ISSN：2469-9950

eISSN：2469-9969
Structural, elastic, and polarization parameters and band structures of wurtzite ZnO and MgO 査読有り

S.-H. Jang, S. F. Chichibu

Journal of Applied Physics　112　(7)　2012年10月1日
出版者・発行元： AIP Publishing
DOI： 10.1063/1.4757023 　

ISSN：0021-8979

eISSN：1089-7550

詳細を見る詳細を閉じる

Ab initio calculations were carried out to predict lattice constants, elastic stiffness constants, spontaneous polarization, piezoelectric constants, and band structure of virtually wurtzite (wz)-MgO. The ground-state properties for both wz-ZnO and wz-MgO were computed using the pseudopotential-planewave method in conjunction with the local density approximation adding the Hubbard parameter to density functional theory. From the results of fitting to reliable in-plane and out-of-plane lattice constants for strain-free and perfectly pseudomorphic wz-MgxZn1−xO alloys, the elastic stiffness constant C33 of the alloy system is revealed to deviate from Vegard's law. The validity of other calculated results for virtually wz-MgO is discussed based on the physical meaning and accuracy, making a comparison with the results shown in previous reports.

︎全件表示 ︎最初の5件までを表示

講演・口頭発表等 1

高並列計算によるイオン性固溶体の構造探索－計算の完全自動化への取り組み招待有り

張成燻

スーパーコンピュータ「富岳」成果創出加速プログラムDDCoMS公開シンポジウム　2024年3月27日

共同研究・競争的資金等の研究課題 2

Building a Design Basis for Oxide Electrolytes toward Next-Generation All-Solid-State Batteries

JANG SEONGHOON

2026年4月1日～ 2028年3月31日
Designing Stable and Active TMXA/TMXB Heterostructures for Cost-Effective Electrocatalytic Hydrogen Generation and Utilization

JANG SEONGHOON

2025年7月31日～ 2027年3月31日

学術貢献活動 6

査読者:Nature Communications

学術貢献活動種別: 査読等
査読者:Energy and AI

学術貢献活動種別: 査読等
査読者:Communications Materials

学術貢献活動種別: 査読等
査読者：ACS Applied Energy Materials

学術貢献活動種別: 査読等
青年編集委員会メンバー：AI Agent

学術貢献活動種別: 審査・学術的助言
青年編集委員会メンバー：AI for Materials

学術貢献活動種別: 審査・学術的助言

︎全件表示 ︎最初の5件までを表示