Large-scale density functional theory (DFT) simulations of lanthanide and actinide oxide materials

Abstract

Complex lanthanide and actinide materials play an important role within the nuclear fuel cycle, due to their thermal and chemical stability. However, the high toxicity and radioactivity of these materials, limits experimental investigations into their properties. Computational methods such as density functional theory (DFT), is an effective method for accurately determining their characteristics. Although conventional DFT methods are limited to calculations to just tens of atoms by unfavourable (cubic) scaling of computational effort with the number of atoms. However, the development of linear-scaling DFT methods within the ONETEP package, allows simulations with thousands of atoms.

To accurately simulate materials containing heavy, f-block elements, it is key to include complex electronic effects in the description of the electrons. Such effects are included in the calculations by representing the core lanthanide or actinide electronic states with an effective core potential or pseudopotential. This eliminates the core states and describes the valence electrons with pseudo wavefunctions with fewer nodes, reducing computational cost. The description of the electronic ground state of these systems, which may be strongly correlated, is further improved by a Hubbard correction. This correction is aimed at localising the f-electrons, producing more accurate band gap energies.

Lanthanide and actinide oxides are susceptible to radiation damage from ionising radiation, particularly from fission product nuclei and α-particles. A fission process in (235U) can produce up to 60,000 defects over a 7 μm range for a heavy fission product such as barium. An α-decay event in 241Am can produce ∼200 defects over 15 μm for the α-particle and ∼1500 defects over 20 nm from the recoil atom. Other radiative processes (β, γ, neutrons) may also act to displace atoms in the structure. Defect formation influences the physical properties of lanthanide and actinide oxide materials, degrading their thermal and mechanical performance as a nuclear fuel.

Exploiting the linear-scaling performance of ONETEP, large-scale simulations of defect containing lanthanide and actinide oxide materials were performed. Of interest are the mechanical properties and performance of these materials as they degrade, changing their composition over time. The chemical aging processes and radiation damage introduce defects into the material, for which the formation energies are predicted. Future investigations could then extend this work to explore the transport mechanisms of intrinsic and extrinsic defects in the material.

Exploiting the linear-scaling performance of ONETEP, large-scale simulations of defect containing lanthanide and actinide oxide materials were performed. Of interest are the mechanical properties and performance of these materials as they degrade, changing their composition over time. The chemical aging processes and radiation damage introduce defects into the material, for which the formation energies are predicted. Future investigations could then extend this work to explore the transport mechanisms of intrinsic and extrinsic defects in the material.

Defects were introduced into simulation cells of increasing size, for which the electronic structure are explored, and their formation energies calculated. In general, the metal vacancy is found to be the most unfavourable intrinsic defect, where the thorium vacancy in ThO2 becomes more unstable with a smaller charge. For CeO2, the simulation cell size and position (octahedral or oxygen-edge) of the oxygen and cerium interstitial defects affects the formation energy and charge state. The peroxide and hydroxyl species can be created in plutonium by introducing oxygen or hydrogen along the <111> direction. Depending on the position of the hydrogen in PuO2, the radical (octahedral site), anion (substitution) or cation (oxygen-edge) species was formed. Hydrogen substitution onto the plutonium or oxygen site was found to be less favourable than the substitution of oxygen on the plutonium site or plutonium on the oxygen site.

The large simulation cells allowed for Frenkel pairs to be introduced along the <100>, <110> and <111> directions, with increasing separation between the vacant and interstitial site. For CeO2, placing the oppositely charged vacancy and interstitial site close to one another or a periodic image can stabilise the Frenkel pair, leading to lower defect formation energies. Oxygen Frenkel pairs are favoured over the metal Frenkel pairs, with both generally being more stable when compared to formation energies calculated at infinite dilution. Similarly, the bound Schottky defects created along the <100>, <110> and <111> directions are more favourable compared to infinite dilution. Initial simulations with ThO2 suggested that the Schottky defect along <110> direction as being most stable. Increasing the supercell size of CeO2 and PuO2 to at least 324 atoms changes the energetic ordering to favour the <111> direction for the most stable Schottky defect.

More information

Identifiers

ORCID for Nabeel Anwar: orcid.org/0000-0003-1221-9256

ORCID for Chris Skylaris: orcid.org/0000-0003-0258-3433

Catalogue record

Export record