Fig. 1 Bitter coil Fig. 1 Bitter coil 94Poster Award NomineeWater-cooled Bitter Magnet for Measurement in High Magnetic Fields Semiconductors, magnetic materials, and superconductors exhibit interesting physical properties under conditions, such as low temperatures, high pressures, and high magnetic fields, as well as under combined conditions. Water-cooled magnets, superconducting magnets, and pulsed field magnets capable of generating high magnetic fields over 10 tesla are used for the measurements. NIMS has developed water-cooled Bitter coils to generate steady [1] and pulsed magnetic fields. The Bitter coil is shown in Fig. 1. The Bitter plate is a disk-shaped copper alloy plate with slits for coiling, holes for fastening stud bolts against electromagnetic force, and holes for cooling. Bitter plates are assembled into a solenoid shape with insulating sheets to form a Bitter coil. To achieve the desired magnetic field, an optimal design must be made, and the current density distribution, temperature distribution, magnetic field distribution, and stress distribution due to electromagnetic force in the coil must be evaluated, and an actual coil must be manufactured. By applying the Bitter coil to the pulsed field magnet, it is expected that repetitive magnetic fields can be generated by taking advantage of the water cooling. The discovery of high-performance solid-state electrolytes (SSEs) is critical for advancing all-solid-state batteries (ASSBs). However, traditional methods like ab initio molecular dynamics (AIMD), while accurate, are computationally expensive for large-scale screening. This study demonstrates the use of Machine learning force fields (MLFFs)[1] to predict the ionic conductivity of Li8SeN2, a promising new SSE, and Li10GeP2S12 (LGPS), a benchmark material. The MLFF, trained on first-principles data, accurately predicts ionic conductivity and ion migration pathways, showing excellent agreement with experimental results for LGPS. For Li8SeN2, MLFF simulations reveal their ionic mobility, providing critical insights into its structure and diffusion mechanisms. A major advantage of MLFFs is their ability to handle large systems and long timescales, making them ideal for material screening. In this work, MLFF enabled efficient simulation of Li8SeN2 across various conditions, yielding precise predictions at a fraction of the cost of AIMD. The flexibility to incorporate corrections for long-range interactions further enhances MLFF’s accuracy in complex systems. By significantly reducing computation time, MLFFs accelerate the valid of new SSE materials, allowing P5-27Shinji Matsumoto and Yasutaka Imanaka Center for Basic Research on Materials, National Institute for Materials Science (NIMS) [1] T. Asano et al., IEEE Trans. Supercond., 16, 965 (2006). P5-28Accelerating the Valid of Solid Electrolyte Candidates Through Machine Learning Force Fields: A Case Study on Li8SeN2 and Li10GeP2S12 Meiqi ZHANG, Yen-Ju Wu, Yukinori KOYAMA, Masao ARAI, and Yibin XUCenter for Basic Research on Materials, National Institute for Materials Science (NIMS) researchers to focus resources on the most promising candidates. [1] R. Jinnouchi, F. Karsai, C. Verdi, R. Asahi, and G. Kresse, J. Chem. Phys., 152, 234102 (2020).
元のページ ../index.html#94