Updated on 2026/08/07

Information

 

写真a

 
TSUJIKAWA KOTA
 
Organization
Kyushu University Platform of Inter/Transdisciplinary Energy Research Assistant Professor
Title
Assistant Professor
External link

Papers

  • Mitigating proton trapping in cubic perovskite oxides via ScO<sub>6</sub> octahedral networks

    Tsujikawa, K; Hyodo, J; Fujii, S; Takahashi, K; Tomita, Y; Shi, N; Murakami, Y; Kasamatsu, S; Yamazaki, Y

    NATURE MATERIALS   2025.8   ISSN:1476-1122 eISSN:1476-4660

     More details

    Language:English   Publisher:Nature Materials  

    Advances in electrochemical devices have been primarily driven by the discovery and development of electrolyte materials. Yet the development of high-performance and chemically stable proton-conducting oxide electrolytes remains a challenge due to proton trapping and the resulting trade-offs between ionic carrier concentration and conductivity in doped oxides. Here we demonstrate that cubic perovskite oxides with heavy Sc doping can overcome these limitations. BaSn<inf>0.3</inf>Sc<inf>0.7</inf>O<inf>3–δ</inf> and BaTi<inf>0.2</inf>Sc<inf>0.8</inf>O<inf>3–δ</inf> are found to exceed the technological threshold of a total proton conductivity of 0.01 S cm<sup>−1</sup> for fuel cell electrolytes at 300 °C. The structural stability of BaSn<inf>0.3</inf>Sc<inf>0.7</inf>O<inf>3–δ</inf> is further validated under harsh chemical and fuel cell conditions. Molecular dynamics simulations using a machine learning force field illustrate rapid proton diffusion pathways along the ScO<inf>6</inf> octahedral network, effectively mitigating proton trapping, while protons are preferentially associated with Sc. Lattice softness is proposed as a primary design descriptor for increasing Sc content in perovskite oxides and developing high-performance electrolytes for electrochemical devices.

    DOI: 10.1038/s41563-025-02311-w

    Web of Science

    Scopus

    PubMed

Research Projects

  • Accelerating Materials Discovery via a Mutual Interpretation Support System for Researchers and Machine Learning

    Grant number:23K19187  2023 - 2024

    Japan Society for the Promotion of Science  Grants-in-Aid for Scientific Research  Grant-in-Aid for Research Activity start-up

    Tsujikawa Kota

      More details

    Authorship:Principal investigator  Grant type:Scientific research funding

    In this study, we developed a mutual interpretation support system that facilitates collaboration between researchers and machine learning models to accelerate materials discovery. Traditionally, the selection of training data has relied heavily on the researcher’s experience, often leading to biased or inefficient exploration. Our system addresses this by enabling balanced sampling based on descriptor space, allowing for the construction of accurate predictive models with a limited number of data points. We evaluated the system using proton-conducting oxides as a model case and compared three different sampling strategies. The results demonstrated that our proposed method efficiently identified high-conductivity compositions with fewer experimental iterations. This outcome highlights the system’s potential to significantly enhance the speed and effectiveness of materials development by promoting synergistic interaction between human expertise and AI-driven exploration.

    CiNii Research