AI Analyzes 2,860 Structures... New Design Principle Boosts Hydrogen Cell Performance by 4.4 Times [Reading Science]

Joint Research Team Led by Seoul National University Develops Air Electrode with Balanced Ion and Electron Transport
Featured as the Cover Article in "Nature Energy"

A new design principle that significantly enhances the performance of the "reversible solid oxide cell (RSOC)"—a device that produces hydrogen and converts it back into electricity—has been elucidated by analyzing 2,860 atomic interface structures using artificial intelligence (AI). By designing the air electrode to allow ions and electrons to move in a balanced manner, the research team improved fuel cell performance by 2.6 times and water electrolysis performance by 4.4 times compared to conventional electrodes.

The College of Engineering at Seoul National University announced on August 11 that Jungwoo Han, Professor at the Department of Materials Science and Engineering, and his research team, in collaboration with Junyoung Park, Professor at Sejong University, Seonju Song, Professor at Chonnam National University, and Ryan O'Hayre, Professor at Colorado School of Mines in the United States, have developed a new composite air electrode with a balanced ionic and electronic conductivity. The results were published in the international journal Nature Energy as the cover article for the July issue.

Cover of the July issue of Nature Energy. Provided by the research team

The RSOC serves as both an electrolyzer that produces hydrogen by splitting water and a fuel cell that generates electricity using hydrogen, all within a single device. By storing surplus power from renewable energy sources such as solar and wind as hydrogen and converting it back into electricity when needed, the RSOC is attracting attention as an energy storage technology.

One of the critical factors that determines performance is the air electrode, where oxygen ion exchange reactions occur. Conventionally, a mixed conductor that transmits both ions and electrons has been used in tandem with GDC, which only conducts oxygen ions. However, the electronic conductivity of GDC is about 1,000 times lower than its ionic conductivity, creating a bottleneck in electron migration. As a result, the oxygen reaction was confined to the narrow interface between the two materials, making it difficult to fully utilize the entire electrode.

AI-Based Reversible Solid Oxide Fuel Cell Simulation. Analyzed 2,860 interface structures using machine learning to elucidate oxygen ion generation, migration, and charge transfer characteristics at the atomic level. Provided by the research team

Instead of GDC, the research team applied the proton conductor "BCZYYb7111," which features a comparatively balanced ability to transport ions and electrons, and combined it with the bismuth-layer structured mixed conductor "GCCCO."

With the new air electrode, the maximum power density reached 7.08 W/cm² in fuel cell mode and 7.88 A/cm² at 1.3 V in electrolysis mode. These results correspond to 2.6 and 4.4 times higher performance, respectively, compared to the conventional LSCF-GDC air electrode. Even when BCZYYb7111 replaced GDC in other air electrode materials, the fuel cell performance improved by 38–129%, and the electrolysis performance by 50–104%.

2,860 Atomic Interfaces Analyzed Using AI … Identifying the Cause of Performance Improvement

To determine the cause of enhanced performance, the research team built an AI-based multiscale simulation by integrating machine learning interatomic potentials (MLIP), molecular dynamics (MD), and density functional theory (DFT). This allowed them to explore 2,860 interface structures, which had been challenging to analyze on such a large scale with traditional computational methods.

Analysis showed that in the new air electrode, "oxygen vacancies," which permit oxygen ion migration, form easily and increase the mobility of oxygen ions. At the interface, a strong built-in potential that accelerates both ion and electron movement was observed. The team confirmed at the atomic level that these factors jointly promote oxygen ion exchange reactions.

Research team photo. (From left) Wonjun Lee, Ph.D. student in the Department of Materials Science and Engineering at Seoul National University; Jinwook Moon, Integrated Master’s and Doctoral Program student in the Department of Materials Science and Engineering at Seoul National University; Seonju Song, Professor in the Department of Advanced Materials Engineering at Chonnam National University, corresponding author; Junyoung Park, Professor in the Department of Nano New Materials Engineering at Sejong University, corresponding author; Jungwoo Han, Professor in the Department of Materials Science and Engineering at Seoul National University, corresponding author. Provided by Seoul National University

Particularly, this research is significant not only for enhancing the performance of specific materials but also for establishing a design criterion that the balance between ionic conductivity and electronic conductivity is essential for RSOC air electrodes. The team expects that this principle can be leveraged to boost both green hydrogen production through electrolysis and power generation efficiency in fuel cells.

Professor Han emphasized, "By systematically exploring 2,860 interface structures through AI-based multiscale simulation, we quantitatively elucidated the atomic-level mechanism by which composite air electrode interfaces enhance RSOC performance. The new design criterion—that ionic and electronic conductivities must be balanced—will serve as a starting point for future material selection and cell design."

Wonjun Lee, PhD student, and Jinwook Moon, an integrated master's and doctoral student of the Department of Materials Science and Engineering at Seoul National University, also participated in the study, which was supported by the Ministry of Science and ICT and the National Research Foundation of Korea.

This content was produced with the assistance of AI translation services.