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: Multilayer_nitride Project title: Atomic-scale design of radiation-tolerant multilayer coatings for nuclear applications Associated funding: LDRD, DOE Other Systems: CNM-Carbon Science: The international program on reduced enrichment for research and test reactors (RERTR) encourages reducing the 235U enrichment in nuclear fuels to values below 20 wt%. The most promising candidate is a U-Mo alloy with 6-10 wt% Mo. The addition of Mo is necessary in order to stabilize the high temperature γ phase of uranium at room temperature [Hofman, 21st RERTR, Sao Paulo, 1998]. In the dispersion type of fuels, U-Mo fuel particles with averaged szie about 70 μm are dispersed in an Al matrix. The fuel meat is then enclosed by Al claddings which results in a fuel plate. During in-pile irradiation tests, the growth of an inter-diffusion layer (IL) between the dispersed U-Mo fuel particles and the surrounding Al matrix strongly limits the fuel’s performance [Kim, JNM 425 (2011) 181]. The irradiation induced IL not only decreases the thermal conductivity of the fuel, but also responsible for the anomalous swelling of the fuel plates. One major goal of U-Mo development is finding solutions to suppress this IL. Among these, modification of the Al matrix or U-Mo alloys by adding alloying elements has been proposed. An enhancement of the in-pile performances has been obtained by adding Si to the Al matrix [Park, JNM 374 (2008) 442]. It is suggested that a silicon rich layer formed at the interface between U-Mo and Al prevents the conventional U-Mo/Al inter-diffusion during irradiation. However, an excessive plate swelling can be still observed in very aggressive irradiation conditions. Using a similar idea, the application of a diffusion barrier by coating U-Mo particles might be more efficient in retarding the inter-diffusion. Heavy ion irradiation experiments [Jungwirth, JNM 434 (2013) 296] suggest that Si coating do provide a good protection of U-Mo fuel by reducing the formation of IL. Meanwhile, nitride coatings have been widely used to protect surfaces from diffusion against Al. Recently ZrN was tested as diffusion b arrier for UMo/Al dispersion fuel [Leenarers, JNM 440 (2013) 220]. However, experiments by Krusin-Elbaum et al. [Thin Sold Films 104 (1983) 81] show that ZrN barrier fails due to the decomposition of ZrN by Al. A potential solution for avoiding reaction between ZrN and Al is by deposition of multilayer coatings on U-Mo fuel particles. Since AlN is relative stable with respect to both Al and ZrN, AlN can be used as a second layer of coating on top of ZrN. It has been demonstrated that materials with nanoscale architectures, such as nano-layered Cu-Nb composites, are both thermally stable and offer improved performance under irradiation [Demkovicz, PRL 100 (2008) 136102]. Coating U-Mo fuel particle with ZrN/AlN multilayers could be a solution to reduce the fuel swelling by suppressing Al diffusion and to further improve the irradiation tolerance of the fuel. During last several decades the understanding of the atomic-scale origins of material behavior under irradiation has been significantly advanced due to the improvements in experimental techniques such as high-resolution transmission electron microscopy, as well as computation modeling techniques such as density functional theory and classical potential molecular dynamics. This knowledge provides a foundation for the emergence of atomic-scale design of materials for radiation resistance. In contrast to conventional try-and-error development approach, atomic-scale design aims to achieve superior radiation response by purposely manipulating composition and microstructure to control the behavior of radiation-induced defects. It relies on modeling to determine the impact of these modifications on engineering-level material behavior. Atomic-scale modeling will be used to accelerate the improvement of existing materials and design new materials. In this work, we are going to use cla ssical potential molecular dynamics to investigate the performance of multilayered nitride coating under irradiation, therefore to design novel multilayer coatings for U-Mo dispersion fuel. The purpose of the propose work is to study the stability of the multilayer nano-composite under irradiation. Depending on the number of multilayers deposited, the thickness of the individual layer varies from a few nanometers to a few hundred of nanometer. Interfaces are efficient sinks and recombination sites for radiation-induced point defects, so it is possible to improve the radiation resistance of the coating by increasing their interface area per unit volume, for example, by reducing the thickness of individual layer. Interfaces with differing atomic structures may exhibit different sink strength, diffusivities, mechanical properties and susceptibilities to embrittlement [Demkovicz, MRS bulletin 35 (2010) 992]. At the same time, interfaces increase the free energy of a material and could be the weakest microstructure link, limiting its overall lifetime. For enhanced performance under irradiation, it is therefore not sufficient for a material to contain a large number o f interfaces. Therefore, tailoring the interfaces with nano-scale design is the key of this study to improve the fuel performance. Project description: We use atomistic MD simulations to investigate one of the interface-governed behaviors exhibited by multilayered nanocomposite of ZrN and AlN, i.e., their substantially enhanced resistance to radiation damage compare to pure ZrN or AlN. Results from MD simulation will be compared to experimental ion irradiation studies of multilayered AlN/TiN nanocomposite [Milsosavljevic, J. Phys. D: Appl. Phys. 43 (2010) 65302]. The interatomic potentials for the nitrides are based on the modified embedded atom method (MEAM). However, there is no ternary potential available for ZrN-AlN. We will use the so-called force matching method as implemented in potfit [Brommer, Modelling Simul Mater Sci Eng 15 (2007) 295] to develop such ternary potential. To build a potential suitable for the simulation of a complicated ternary system in a wide ranges of pressure and temperature, we have to prepare an extensive set of reference configurations representing all kinds of inter esting structures. Each configuration is a small structure model corresponding to one of the possible states of the ZrN-AlN system. VASP will be used to calculate energy, force and stress of the reference configurations with atoms up to 250. The total computational time for DFT calculations will be about 30000 CPU-hours. Model ZrN-AlN bilayers are constructed in the experimentally observed orientation relation [Milosavljevic, J Phys. D: Appl. Phys. 43 (2010) 065302], in which a ZrN {111} plan neighbors an AlN {0001} plane and an interfacial ZrN <112> direction lies parallel to an AlN <2110> direction. The constituent ZrN and AlN layers are 4.5 nm thick, have periodic boundaries in the interface plane, and terminate in free surfaces away from the interfaces. The ZrN/AlN multilayer nanocomposite can be created in a similar way by stacking 4.5 nm-thick ZrN and AlN crystalline slabs, in which each ZrN or AlN layer contains up to 100000 atoms (22.5×22.5×4.5nm). The layer thickness, number of layers and interface orientation can be controlled in deposition experiments, which will be investigated independently by atomistic simulations. The ZrN-AlN+ZBL potential is used to model collision cascades in in cubic ZrN, hexagonal AlN, and in the vicinity of ZrN-AlN interfaces formed in the multilayer composites. Cascades are initiated by giving a selected atom-the primary knock-on atom (PKA)-a large kinetic energy. The Kinetic energies given to PKAs are chosen to be representative of Ar+ implantation experiments carried out on ZrN-AlN nanocomposites, where the energy of the implanted ions was 200 keV. In the multilayer model systems PKAs are selected from the central layer. For the collision cascade simulations, the first few pico-seconds require extreme fine time step to capture the recombination of point defects, and then nano-second time scale simulations are necessary to reach a thermal equilibrium for the defect structure. By combining all the factors to be studied, we will expect about 25 production jobs. Our initial test shows that a single job needs about 25000 CPU-hours (256 CPU×96 hours). We therefore r equest the total computational time of 650000 CPU-hours. Project URL: Requested allocation: 640000 Q1: 160000 Q2: 160000 Q3: 160000 Q4: 160000 Justification: The used molecular dynamics code LAMMPS has been demonstrated to be cable of simulating millions of atoms using thousands of CPU efficiently. The detailed benchmarks of the parallel efficiency of lammps on different platform can be found from the website of the author http://lammps.sandia.gov/bench.html. With the typical number of CPU (256 CPU) in our MD simulations, the parallel efficiency of LAMMPS on Blues will be above 70%. 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