In the realm of energy research, a team of scientists from various institutions, including the University of Sheffield, University of York, and the United Kingdom Atomic Energy Authority, has made significant strides in understanding radiation damage in a high-temperature superconductor material called YBa$_2$Cu$_3$O$_{7-δ}$ (YBCO). Their work, published in the journal npj Computational Materials, focuses on improving the predictive modeling of radiation damage in this material, which is crucial for evaluating the performance of high-temperature superconducting (HTS) tapes in compact fusion reactors.
The researchers employed advanced machine-learning techniques to develop interatomic potentials that can accurately predict radiation damage across a wide range of oxygen stoichiometries in YBCO. Traditional empirical models have struggled with this variability, but the new machine-learned models, specifically an Atomic Cluster Expansion (ACE) potential and a tabulated Gaussian Approximation Potential (tabGAP), have shown remarkable accuracy. These models were validated against Density Functional Theory (DFT) calculations, ensuring their reliability in describing atomic-scale collision processes.
Using these models, the team conducted molecular dynamics simulations of 5 keV cascades, which are events where an atom is displaced and creates a cascade of further displacements. The simulations revealed enhanced peak defect production and recombination compared to previous empirical models, indicating different cascade evolution. Importantly, the researchers found that total defect production depends only weakly on the oxygen stoichiometry, suggesting that radiation damage processes in YBCO are robust across different compositions.
The study also delved into fusion-relevant 300 keV cascade simulations, which showed the formation of amorphous regions with dimensions comparable to the superconducting coherence length. This finding aligns with electron microscopy observations of neutron-irradiated HTS tapes, further validating the predictive power of the machine-learned interatomic potentials.
For the energy sector, particularly in the development of fusion reactors, these findings are significant. High-temperature superconductors like YBCO are critical components in compact fusion reactors, and understanding their behavior under radiation damage is essential for ensuring their performance and longevity. The machine-learned interatomic potentials developed in this research provide a powerful tool for predicting radiation damage, enabling engineers to design more resilient and efficient HTS tapes for future fusion energy systems.
Source: npj Computational Materials
This article is based on research available at arXiv.