Presentation Information
[WBP1-11]Data-Driven Optimization of Liquid-Phase-Assisted Process for Engineering High-Jc Coated Conductor.
*Shunta Ito1, Ibuki Kato1, Tomoya Horide1, Yutaka Yoshida1 (1. Nagoya University (Japan))
Keywords:
YBCO thin films,Liquid-phase-assisted
[Introduction]
REBa2Cu3Oy(REBCO) is a promising material for coated conductor applications, primarily due to its high critical temperature (Tc) and excellent performance in high magnetic fields. However, a key challenge is the difficulty of simultaneously achieving strong in-field properties and low fabrication costs. Improvements in in-field properties have been pursued through the introduction of artificial pinning centers (APC). To lower manufacturing costs, increasing the deposition rate is crucial. In order to address this issue, researchers are exploring deposition techniques that use liquid-phase assistance. These methods require precise control of various process parameters to ensure the liquid phase forms and remains stable. While several successful cases have been reported based on the knowledge and experience of individual researchers, these findings are often difficult to reproduce across different systems. This is because process parameters are specific to each deposition system. In this study, we aimed to establish a clearer relationship between process parameters and the thin-film growth mechanism in liquid-phase-assisted PLD. By analyzing the resulting surface morphology and properties of YBCO+BMO films under various deposition conditions, we sought to develop a process map.
[Method]
Thin films were fabricated on IBAD-MgO substrates by PLD using targets with varied Y:Ba:Cu ratios. The surface morphology and precipitates of the fabricated films were examined using SEM and EDX. Based on the features obtained from surface observations, the thin-film samples were grouped according to their fabrication conditions, and these groupings were used as the basis for constructing a process map. To build the process map, we employed a data science method called Label Propagation, a machine learning technique. We used the four-probe method to evaluate the superconducting properties of the films through transport measurements.
[Results and Discussion]
In the samples prepared in this study, the characteristic surface morphology of a liquid-phase-assisted process, such as Ba-Cu-O precipitates and grooves, was observed in samples with specific film deposition conditions. Based on the presence or absence of these characteristic surface structures, the samples were grouped, and an estimated process map was constructed using data science methods. The process map shows that thin films with a surface structure derived from liquid-phase growth are expected under specific conditions of composition and substrate temperature. Therefore, it may be possible to achieve high Jc via the liquid-phase-assisted method by optimizing the conditions, focusing on this region. However, the Tc and Jc values for these initial samples are lower compared to typical YBCO+BMO samples grown using the liquid-phase-assisted method. An ideal liquid-phase-assisted process is expected to improve both crystal orientation and Tc due to the low degree of supersaturation in liquid-phase growth. Therefore, further optimization is required. At the conference, we will present a detailed report on the process map developed using the data science methods along with the surface morphology and properties of the samples. We will also report on the optimization of process parameters for the liquid-phase-assisted method using this map.
REBa2Cu3Oy(REBCO) is a promising material for coated conductor applications, primarily due to its high critical temperature (Tc) and excellent performance in high magnetic fields. However, a key challenge is the difficulty of simultaneously achieving strong in-field properties and low fabrication costs. Improvements in in-field properties have been pursued through the introduction of artificial pinning centers (APC). To lower manufacturing costs, increasing the deposition rate is crucial. In order to address this issue, researchers are exploring deposition techniques that use liquid-phase assistance. These methods require precise control of various process parameters to ensure the liquid phase forms and remains stable. While several successful cases have been reported based on the knowledge and experience of individual researchers, these findings are often difficult to reproduce across different systems. This is because process parameters are specific to each deposition system. In this study, we aimed to establish a clearer relationship between process parameters and the thin-film growth mechanism in liquid-phase-assisted PLD. By analyzing the resulting surface morphology and properties of YBCO+BMO films under various deposition conditions, we sought to develop a process map.
[Method]
Thin films were fabricated on IBAD-MgO substrates by PLD using targets with varied Y:Ba:Cu ratios. The surface morphology and precipitates of the fabricated films were examined using SEM and EDX. Based on the features obtained from surface observations, the thin-film samples were grouped according to their fabrication conditions, and these groupings were used as the basis for constructing a process map. To build the process map, we employed a data science method called Label Propagation, a machine learning technique. We used the four-probe method to evaluate the superconducting properties of the films through transport measurements.
[Results and Discussion]
In the samples prepared in this study, the characteristic surface morphology of a liquid-phase-assisted process, such as Ba-Cu-O precipitates and grooves, was observed in samples with specific film deposition conditions. Based on the presence or absence of these characteristic surface structures, the samples were grouped, and an estimated process map was constructed using data science methods. The process map shows that thin films with a surface structure derived from liquid-phase growth are expected under specific conditions of composition and substrate temperature. Therefore, it may be possible to achieve high Jc via the liquid-phase-assisted method by optimizing the conditions, focusing on this region. However, the Tc and Jc values for these initial samples are lower compared to typical YBCO+BMO samples grown using the liquid-phase-assisted method. An ideal liquid-phase-assisted process is expected to improve both crystal orientation and Tc due to the low degree of supersaturation in liquid-phase growth. Therefore, further optimization is required. At the conference, we will present a detailed report on the process map developed using the data science methods along with the surface morphology and properties of the samples. We will also report on the optimization of process parameters for the liquid-phase-assisted method using this map.