Insights Into Radiation Damage in YBa 2 Cu 3 O 7-δ From Machine Learned Interatomic Potentials Journal: Superconductivity

Ashley Dickson, Niccolò Di Eugenio, Federico Ledda, Daniele Torsello, Francesco Laviano, Flyura Djurabekova, Jesper Byggmästar, Mark R. Gilbert, Duc Nguyen-Manh, Erik Gallo, Antonio Trotta, Davide Gambino, Samuel T. Murphy · 2025

Read Insights Into Radiation Damage in YBa 2 Cu 3 O 7-δ From Machine Learned Interatomic Potentials Journal: Superconductivity by Ashley Dickson, Niccolò Di Eugenio, Federico Ledda, Daniele Torsello, Francesco Laviano, Flyura Djurabekova, Jesper Byggmästar, Mark R. Gilbert, Duc Nguyen-Manh, Erik Gallo, Antonio Trotta, Davide Gambino, Samuel T. Murphy (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-24430.txt.

Introduction

Ashley Dickson and five colleagues simulate radiation damage in a fusion material. The material is YBCO, a high-temperature superconductor used in tapes. Compact fusion reactors will bombard those tapes with neutrons. Older atomistic models handle only the perfect form of the compound. Real conductors lack some oxygen, and that lack changes behavior.

Machine-learned models lift that limit. Two recent ones are used, built by cluster expansion and by Gaussian approximation. Both match quantum calculations for energies and forces. Cascade simulations then follow collision damage directly. Peak defect output exceeds the older model's predictions. Total defect output depends only weakly on oxygen content. High-energy cascades leave amorphous pockets sized near the coherence length. That matches microscope images of irradiated tapes.

Catalog record

CREATORAshley Dickson, Niccolò Di Eugenio, Federico Ledda, Daniele Torsello, Francesco Laviano, Flyura Djurabekova, Jesper Byggmästar, Mark R. Gilbert, Duc Nguyen-Manh, Erik Gallo, Antonio Trotta, Davide Gambino, Samuel T. Murphy
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE62 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2512-24430

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Insights Into Radiation Damage in YBa 2 Cu 3 O 7-δ From Machine Learned Interatomic Potentials Journal: Superconductivity Ashley Dickson, Niccolò Di Eugenio, Federico Ledda, Daniele Torsello, Francesco Laviano, Flyura Djurabekova, Jesper Byggmästar, Mark R. Gilbert, Duc Nguyen-Manh, Erik Gallo, Antonio Trotta, Davide Gambino, Samuel T. Murphy Abstract Accurate prediction of radiation damage in YBa 2 Cu 3 O 7-δ (YBCO) is essential for assessing the performance of high-temperature superconducting (HTS) tapes in compact fusion reactors. Existing empirical interatomic potentials have been used to model radiation damage in stoichiometric YBCO, but fail to describe oxygen-deficient compositions, which are ubiquitous in industrial Rare-Earth Barium Copper Oxide conductors and strongly influence superconducting properties. In this work, we demonstrate that modern machine-learned interatomic potentials enable predictive modelling of radiation damage in YBCO across a broad range of oxygen stoichiometries, at a greater fidelity than previous empirical models. We employ two recently developed approaches: an Atomic Cluster Expansion (ACE) potential and a tabulated Gaussian Approximation Potential (tabGAP). Both machine-learned models are shown to accurately reproduce Density Functional Theory (DFT) energies, forces, and threshold displacement energy distributions, providing a quantitatively reliable description of atomic-scale collision processes. Molecular dynamics simulations of 5 keV cascades predict significantly enhanced peak defect production and recombination relative to a previous empirical potential, indicating qualitatively different cascade evolution. Moreover, these new machine learning models generate increased proportions of copper and oxygen vacancies compared to previous models, in direct agreement with experimental observations. By explicitly varying oxygen deficiency, we further show that total defect production exhibits only a weak dependence on stoichiometry, providing new insight into the robustness of radiation damage processes in oxygen-deficient YBCO. Finally, fusion-relevant 300 keV cascade simulations reveal amorphous regions with characteristic dimensions comparable to the superconducting coherence length, consistent with electron microscopy observations of neutron-irradiated HTS tapes. These results establish machine-learned interatomic potentials as powerful, computationally efficient tools for uncovering radiation-damage physics in YBCO, enabling predictive simulations across technologically relevant compositions and irradiation conditions. 1 Introduction HTS magnets deliver high field strengths at relatively high temperatures and so offer great promise for use in compact fusion reactors [ 54 ] . The functional materials in these magnets are typically the Rare Earth Barium Copper Oxide ceramics, of which YBa 2 Cu 3 O 7-δ (YBCO) is the prototypical example. During reactor operation they are subject to bombardment with high energy neutrons from the fusion plasma. Consequently, it is essential to understand how these neutrons will damage the structure of these HTS materials and how this affects their superconducting properties. Neutrons transfer their energy to the lattice primarily through elastic collisions, producing high-energy PKAs that induce a sequence of atomic collisions and displacing atoms from their lattice sites to create defects, a process called a collision cascade [ 41 ] . These cascades proceed through three characteristic stages [ 41 ] . First, is the ballistic phase, during which the PKA transfers energy through a series of rapid, high-energy elastic collisions. This is followed by the heat-spike phase, where the dense cascade core rapidly rises in temperature, allowing partial local melting and subsequent recrystallization. Finally, as the region cools, the remaining primary damage may anneal out over longer timescales. In the absence of a fusion-relevant neutron source, prediction of the extent and type of damage observed in YBCO tapes represents a major challenge [ 54 ] . Modelling and simulation thus play a crucial role in predicting the microstructural evolution during radiation-damage cascades [ 23 , 55 ] . The predictive power of modelling is further enhanced when complemented with experimental investigations using neutron irradiation [ 59 ] and proxy techniques such as light ion irradiation [ 50 , 63 ] . MD allows the simulation of primary radiation damage with atomic resolution, however, these simulations require an interatomic potential to model the interactions between atoms in the material. For YBCO in particular, development of such a potential is associated with a number of challenges: for instance, the copper environments (Cu1 and Cu2, see figure 1 a for a structural diagram of YBCO) have differing charge states [ 36 ] , it exhibits considerable non-stoichiometry, and the layered anisotropic structure offers numerous degrees of freedom. Previous potentials for YBCO [ 1 , 9 , 23 ] have been employed to understand the properties of bulk YBCO as well as its response to irradiation [ 23 , 55 , 14 ] . However, the Buckingham-Coulomb formalism employed by Gray et al . cannot be used to study YBa 2 Cu 3 O 7-δ where \delta\neq 0 as this would create an unphysical charge imbalance. Most industrially-produced HTS tapes utilise inherently sub-stoichiometric YBCO. Further, HTS tapes also accumulate additional damage over their oxygen sublattice during their operational lifetime [ 37 , 52 , 60 , 25 , 38 ] . Capturing this oxygen-deficient regime is, therefore, essential for the development of an accurate and reliable model. MLPs have gained significant attention in recent years due to their unique combination of flexibility, computational efficiency, and near-DFT accuracy [ 66 , 44 , 2 , 26 , 16 ] . Their adaptability makes them well-suited for modelling complex materials, including systems with multiple polymorphs or variable stoichiometries [ 66 , 44 , 2 ] . Notably, MLPs have already demonstrated success across a wide range of material classes, from single elements [ 6 , 43 ] , to multi-element alloys [ 26 ] , and even complex oxides [ 66 ] . In our previous work [ 19 ] we developed 4 MLPs for YBCO. For the purposes of this work, we only employ and compare two of these potentials: the ACE [ 15 ] and the tabGAP [ 3 , 8 ] MLPs (available at [ 12 ] and [ 13 ] , respectively) to explore radiation damage in YBCO. These potentials were selected due to their computational efficiency, which allows them to used to study large scale radiation damage events. As outlined in previous work, all MLPs demonstrate considerable improvements in accuracy over the Gray potential [ 23 ] , while also enabling simulation of non-stoichiometric YBCO. In this work, we explore TDEs compared to DFT results [ 14 ] and to the Gray potential [ 23 ] , investigate the curvature of the potential energy surface with respect to displacement, and validate the accuracy of our potentials for the liquid phase of YBCO (as the heat spike of a cascade is analogous to recrystallization of hot liquid [ 40 ] ). After this initial validation, we explore radiation damage in YBa 2 Cu 3 O 7-δ where 0\leq\delta\leq 0.4 . Finally, we perform large-scale, fusion-relevant collision cascades in YBCO at an oxygen concentration typical of commercial tapes and generate HRTEM images of the resulting damage to show that the amorphous cores formed from the cascades are on the order of the coherence length, accordant with experimental observations [ 20 , 29 , 21 ] . 2 Methods 2.1 Calculating forces This study employs a mixture of simulations employing interatomic potentials and electronic structure DFT to represent the forces between the atoms, these are detailed in this section. 2.1.1 Interatomic potentials All simulations using interatomic potentials were carried out with

Read the complete text