Researchers from Tohoku University in Japan have compiled and analyzed a comprehensive dataset of oxygen ion conductors, materials crucial for various energy technologies. The team, led by Seong-Hoon Jang, Shin Kiyohara, Hitoshi Takamura, and Yu Kumagai, has curated data from 84 experimental reports spanning six decades, encompassing 483 different materials. Their work, published in the journal Nature Communications, aims to provide a systematic foundation for the discovery and development of next-generation oxygen ion conductors.
Oxygen ion conductors are vital components in solid oxide fuel cells, sensors, and membranes. Despite extensive research, a comprehensive dataset for comparative analysis has been lacking. The researchers addressed this gap by compiling a dataset that includes activation energy and prefactor values derived from Arrhenius plots, along with detailed metadata on structure, composition, measurement methods, and data sources. Notably, they corrected erroneous Arrhenius equations in some original papers to ensure data accuracy.
To demonstrate the utility of their dataset, the researchers constructed interpretable regression models for predicting oxygen ionic conductivity. They found that two symbolic regression models for activation energy and prefactor suggest that oxygen ion transport is primarily governed by the local coordination environment and electrostatic interactions, respectively. This insight could guide the design of more efficient oxygen ion conductors.
The dataset and models provide a reliable foundation for data-driven discovery and predictive modeling in the energy sector. For example, in the development of solid oxide fuel cells, understanding and optimizing oxygen ion conductivity can lead to more efficient and cost-effective energy conversion devices. Similarly, in sensors and membranes, improved materials can enhance performance and durability. The researchers’ work offers a valuable resource for scientists and engineers working on these and other energy technologies.
This article is based on research available at arXiv.