Effect of cationic chemical disorder on defect formation energies in uranium-plutonium mixed oxides

TL;DR

This study investigates how cationic chemical disorder influences defect formation energies in uranium-plutonium mixed oxides using advanced computational methods, revealing local environment effects and proposing a new interaction model.

Contribution

It introduces a systematic approach to evaluate defect energies considering local disorder, and develops an interaction model for BSD formation energies in (U,Pu)O_2.

Findings

Cation shells within 7 Å significantly affect defect energies.

BSD formation energy varies by up to 0.97 eV depending on local composition.

DFT+U results align well for U-rich environments, less so for Pu-rich ones.

Abstract

At the atomic scale, uranium-plutonium mixed oxides (U,Pu)O_2 are characterized by cationic chemical disorder, which entails that U and Pu cations are randomly distributed on the cation sublattice. In the present work, we study the impact of disorder on point-defect formation energies in (U,Pu)O_2 using interatomic-potential and Density Functional Theory (DFT+U) calculations. We focus on bound Schottky defects (BSD) that are among the most stable defects in these oxides. As a first step, we estimate the distance R_D around the BSD up to which the local chemical environment significantly affects their formation energy. To this end, we propose an original procedure in which the formation energy is computed for several supercells at varying levels of disorder. We conclude that the first three cation shells around the BSD have a non-negligible influence on their formation energy (R_{D} = 7.0 \{AA}). We apply then a systematic approach to compute the BSD formation energies for all the possible cation configurations on the first and second nearest neighbor shells around the BSD. We show that the formation energy can range in an interval of 0.97 eV, depending on the relative amount of U and Pu neighboring cations. Based on these results, we propose an interaction model that describes the effect of nominal and local composition on the BSD formation energy. Finally, the DFT+U benchmark calculations show a satisfactory agreement for configurations characterized by a U-rich local environment, and a larger mismatch in the case of a Pu-rich one. In summary, this work provides valuable insights on the properties of BSD defects in (U,Pu)O_2, and can represent a valid strategy to study point defect properties in disordered compounds.