Hello, A new project on the LCRC cluster has been requested. Please forward the information on to the LCRC Allocation sub-committee. Applicant's name: Zhigang Mei Applicant's institution: ANL Applicant's division: NE Project Name: CeO2-voids-bubbles Project title: Void and Gas Bubble Formation in Doped Ceria: Role of Oxygen Defects and Fission Gases Diffusion Associated funding: LDRD Other Systems: CNM-Carbon, LCRC-fusion Science: Cerium dioxide (CeO2) is an important material with a wide range of technological applications. It is used as an electrolyte in solid oxide fuel cells (SOFCs), as a catalytic converter in the automotive industry or as a model material for PuO2 in nuclear energy applications. It has a fluorite structure (space group Fm3m [1]) and can develop a complex pattern of defects, depending on temperature and oxygen pressure [2]. Its extended hypo-stoichiometric domain was experimentally confirmed by Bevan and Kordis [3]. A review of oxygen diffusion in oxides, including surface and grain boundary effects, is available in Mayer et al. [4]. Dieckmann [5] published a detailed presentation of the basic concepts related to point defects and transport in non-stoichiometric oxides, including ceria. Considerable effort was devoted to improving the oxygen ion conductivity in ceria by optimizing the doping elements and concentrations. In the case of mixed oxide (MOX) nu clear fuels (UO2+x, PuO2-x), the point defects that survive the initial damage due to irradiation form extended defects, such as vacancy clusters, dislocation loops and voids. These nano-scaled defects impact the microstructure of the fuel, leading to degradation of its thermal and mechanical properties. In order to fully understand and to be able to accurately predict the microstructural evolution under irradiation, elucidating the underlying formation mechanisms of the extended defects is very important. The current goal of studying CeO2 is to understand the formation and evolution of nano-scaled defects in irradiated oxide nuclear fuels, including vacancy clusters, lattice voids, dislocation loops, and gas bubbles. There is essentially a multi-scaled problem, since the size of these defects covers a wide range length scale, from a few Angstroms to a few hundred of nanometers. Meanwhile, the formation of defects evolves diffusion of point defects over a long period of time, typically in the nanosecond-microsecond regime. A number of ion beam implantation experiments on CeO2 have been conducted with common fission gas species: Kr and Xe [6, 7]. Under different ion irradiation conditions, such as irradiation temperature, ion dose and energy, dislocation loops, voids and gas bubbles form at different stages. However, the formation and growth of these defects are still unclear. For example, completely different void size distributions, with void diameter from 0.5m to 5nm, were ob served for 1MeV Kr irradiated 5% La doped CeO2 and 25% La doped CeO2 at 600 °C [6]. Defects and their diffusion in CeO2 are essential to understand this observation. Critical questions relevant to the formation of void and bubble in irradiated CeO2 include: 1) what is the driving force for Xe and other fission gas to form clusters; 2) how does the size affect the stabilities of lattice voids and gas bubble; 3) how does Xe and vacancy diffusivities relate to the diffusivity of Xe clusters. Additionally, the formation and growth of dislocation loops in CeO2 was found to be strongly affected by temperature and non-stoichiometry. Due to the limitation of current experimental technique, it is difficult to track the position of oxygen atom and investigate the nucleation of dislocation loop from atomistic scale using in-situ TEM. So far the microscopic understanding of the formation of dislocation loop under irradiation is still not clear. The formation mechanism for these nano-sc aled defects are of critical importance to understand and further improve the irradiation tolerance of nuclear fuel materials and can only be answered by using appropriate computer simulation techniques to analyze and enhance the experimental information. The purpose of this study is to provide a united view of the effect of non-stoichiometry and temperature on the formation and evolution of nano-scaled defects, i.e., defect clusters, voids and fission gas bubbles, in irradiated ceria using existing experimental data, multiscale modeling and computer simulations. Project description: We will use DFT to calculate the formation of point defects and clusters. The effect of temperature and oxygen pressure on the stabilities of defects will be explored. We will use nudged elastic band (NEB) method to investigate the migration barrier of point defect, defect clusters and fission gas in doped ceria. The predicted barrier will be used in a kMC framework to study the long-time oxygen and fission gas diffusivities. In parallel to DFT calculations, we will explore point-defect clustering and formation of voids and gas bubbles in ceria using accelerated molecular dynamics technique, such as metadynamics. The formation of defect clusters evolves diffusion of point defects over a long period of time, typically in the nanosecond-microsecond regime. These accelerated molecular dynamics method will been used to link the MD-observed defect configurations with the experimentally accessible radiation-induced extended defects. Due to the strongly correlated nature of the f electrons, more accurate methods, such as the LDA+U and hybrid density functional methods are required to overcome the limitations of conventional DFT method. These new methods require one order of magnitude longer computational time. Highly efficient DFT codes, such as VASP, will be utilized for DFT calculations. To go beyond point defects, we will use highly parallel MD code LAMMPS to study the formation and migration of defect clusters, voids and gas bubbles in realistic time scale, nanosecond-microsecond regime. These simulations need significant amount of computational time. We will employ DFT calculations to study the formation energies and stabilities of point defects and point defect clusters in non-stoichiometric and rare-earth doped ceria using a 3×3×3 unit cell of ceria with around 300 atoms. These defect formation energies will be used to estimate the defect concentrations based on point defect model. The most computationally intensive calculations are focused on the study of oxygen defects, defect clusters and fission gas diffusion in rare-earth doped ceria. We will explore the several effects on the migration energy in doped ceria: 1) role of doping elements, 2) doping concentration, 3) type of point defects, 4) morphology of defect clusters; 5) cation ordering around diffusion atoms; and 6) role of fission gas. We expect the computational requirement for the proposed study to be substantial. For a typical DFT-NEB calculation of oxygen diffusion barrier in ceria using a supercell with ~100 atoms, it requires ~1000 core hours for a five -image NEB calculations. By considering all the six effects on the diffusion, we will calculate ~150 coverage-dependent diffusion barriers. Thus we estimate ~150,000 core hours are needed for all the DFT calculations. With predicted migration barriers, a parallel kinetic Monte Carlo code SPPARKS will be adopted to study the long-time oxygen and fission gas diffusivities in doped ceria. The diffusivity of oxygen and fission gas will be investigated with respect to doping concentration and temperature. A total of ~500 coverage-dependent diffusivity calculations are required. The simulation system will be based on a 20×20×20 unit cell of ceria to cover all the doping concentration we are interested. Our test shows that a minimum 500,000 KMC steps are required to achieve good convergence for diffusivities, requiring ~100 core hours for each diffusivity calculation. Therefore we estimate ~50,000 core hours are needed for the KMC calculations. On the other hand, MD simulations with several initial fission gas concentrations, temperatures, and gas pressures will be used to explore on the nucleation of voids and gas bubbles in pure ceria. A relative large cell size is critical for simulating voids and gas bubbles. A computational cell with 200,000-500,000 atoms is required for setting up the pressure in the system. The most common simulation time in this study (10 ns) requires up to 20 million simulation time steps that will be performed for several gas concentrations, temperatures and gas pressure. To cover all the effects, it requires ~12 coverage-dependent MD simulations. Our benchmark calculations show that ~20,000 cores hours are needed for one typical MD run. We estimate ~250,000 core hours are needed for all the MD calculations. The total amount of CPU time needed to finish all the DFT, KMC and MD calculations for this project will be approximately 450,000 core hours. We expected three people will involve in this project. Project URL: Requested allocation: 440000 Q1: 110000 Q2: 110000 Q3: 110000 Q4: 110000 Justification: The requester has used undetermined amount hours of their initial startup project. In addition to approving an initial amount, please specify a Category and Subcategory for this project. For a list of the current selection of approved categories, please see: https://wiki.lcrc.anl.gov/wiki/Processes/Categories Once the Allocation committee has approved the project, please go to the Project Management page to create it: https://accounts.lcrc.anl.gov/projects.php Thank You, The LCRC Accounts System