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: Alper Kinaci Applicant's institution: ANL Applicant's division: NST Project Name: VO2_thermal Project title: Evaluation of thermal transport in bulk and nanostructured VO2 Associated funding: This work is funded by a Strategic Partnership Project (previously known as Work-for-Others) sponsored by the Toyota Motor Engineering & Manufacturing North America (TEMA). Other Systems: Carbon cluster at Center for Nanoscale Materials (CNM) at Argonne: 522,000 core hours Science: Metal to insulator transition (MIT) refers to a sharp switch in the electrical conductivity of a material. This phenomenon can be introduced by electron localization [1, 2], carrier doping and electron-phonon coupling [3]. The actual driver of this transition is still in debate in different materials since the drastic change in conductivity is accompanied by a distortion in the lattice (and generally a phase transformation) [4]. Among materials possessing MIT, VO2 is heavily studied in terms of its electronic properties since its transition occurs close to room temperature. The semiconductor phase below 340K has a monoclinic structure while the metallic phase above 340K has a rutile structure [5]. During the transformation between these phases electrical conductivity changes about 4 orders of magnitude [6]. In contrast to electrical conductivity, the thermal transport in VO2 has not been given the same attention. Practically, the electronic part of the thermal conductivity maybe estimated from Wiedemann-Franz law. Accordingly, we will focus on the theoretical estimation of the phonon part of the thermal transport. In order to characterize lattice heat transport in different phases of VO2, we will employ molecular dynamics (MD) simulations. The predictive power of MD calculations depends heavily on the empirical potential parameters. Several empirical potentials for VO2 exist in the literature; however, they were parametrized to represent bulk structural properties and oxidation thermodynamics of tetragonal VO2 [7], and fail to reproduce DFT phonon and elastic properties. In this proposal, we will develop an empirical potential that can represent phonon dispersions, elastic properties, surface and defect formation energies of the two VO2 phases. The Buckingham potential plus coul omb interactions form is selected to represent the interactions. The training set, which will include the aforementioned properties, will be generated from density functional theory calculations. The parameter fitting will be accomplished with a genetic algorithm global optimizer. The resultant interatomic potential will be used in molecular dynamics calculations to evaluate lattice thermal conductivity of bulk and nanostructured VO2. The direct method, in which a temperature gradient is imposed on the system [8], and Green Kubo method, in which the heat current is obtained from equilibrium fluctuations of energy [9, 10], will be utilized to calculate thermal conductivity. With the same procedure, we will systematically evaluate the effects of substitutions, point defects, and film thickness on the thermal transport of VO2 phases. The results of these calculations will guide the experimental efforts and ultimately lead to effective tuning of the thermal transport in VO2-base d MIT materials. [1] P. W. Anderson, Phys. Rev. 109, 1492 (1958). [2] N. F. Mott, “Metal-Insulator Transitions” (Taylor and Francis, London 1990). [3] G. Gruner, Rev. Mod. Phys. 60 1129 (1988). [4] A. Perucchi, L. Baldassarre, P. Postorino and S. Lupi, J. Phys.: Condens. Matter 21, 323202 (2009). [5] R. M. Wentzcovitch, W. W. Schulz and P. B. Allen, Phys. Rev. Lett. 72, 3389 (1994). [6] F. J. Morin, Phys. Rev. Lett. 3, 34 (1959). [7] B. Jeon, C. Ko, A. C.T. van Duin, and S. Ramanathan, Surface Science, 606, 512 (2012). [8] P. K. Schelling, S. R. Phillpot, and P. Keblinski, Phys. Rev. B 65, 144306 [9] M. S. Green, J. Chem. Phys. 22, 398 (1954). [10] R. Kubo, J. Phys. Soc. Japan 12, 570 (1957). Project description: The calculations will involve the items below. (1) Density functional theory (DFT) calculations will be employed to obtain the training set that will be used to parameterize the empirical potential. This set will include the phonon dispersions, elastic constants, and surface and vacancy formation energies. The surface energies will be computed using large slab configurations with at least 2 slab thicknesses. We will evaluate 5 low index surfaces. For the vacancies, we will compute both vanadium and oxygen vacancy formation energies. All these calculations will be done for both VO2 phases. (2) A Buckingham potential plus a Coulomb potential will be parameterized to fit the DFT training data generated in (1). In parameterization, a genetic algorithm global optimizer is used in conjunction with local optimizers e.g. simplex and conjugate gradient. We estimate 20 trials before the parameter set converges to give a reasonable representation of the training set. (3) The parametrized potential will be used to test the convergence of thermal conductivity in pristine VO2 phases. From previous experience, we estimate that 30,000-60,000 atoms will be sufficient to capture enough long wavelength phonons that are important for thermal transport. Each thermal transport calculation will be repeated 4 times starting from different initial conditions (i.e. different initial velocity distributions) to obtain a reasonable ensemble average. The simulation time step will be 1 fs. Depending on the method of evaluation (direct method or Green-Kubo), simulations should continue about 1 to 5 million steps (corresponding to 1 to 5 ns) before calculation of thermal conductivity. (4) In the bulk forms, the effect of vacancies (oxygen and vanadium) and metal substitutions (i.e. to be treated as mass differences) will be investigated with 3 different concentrations in both phases. The time step, total number of time steps and number of different initial conditions for each system will be same as the ones in (3). Both monoclinic and tetragonal phases of VO2 will be considered. (5) Lattice thermal conductivity of thin films of VO2 will be investigated. The film thicknesses and sizes will be set according to experimental needs. Here we estimate thermal conductivity calculations for 3 different film thicknesses. The number of atoms in these systems can be larger than bulk systems and will be adjusted depending on the experimental needs. Apart from number of atoms, a similar set up will be used for the calculations as given in (4) and (5) The DFT calculations will be performed using the plane wave code VASP, with GGA-PBE exchange correlation functional, projector augmented wave (PAW) treatment of core electrons and Hubbard correction (U). The VASP code is tested in Blues with a 350 electron system and 65%-80% scaling efficiency is obtained 64 to 128 cores compared to 16 cores. In Fusion where the code has been tested for systems having up to 7000 electrons, the code showed 70%-80% scaling efficiency for 64 to 128 cores compared to 8 cores. Given this performance, the VASP code has the efficiency to simulate large number configurations in reasonable amount of time. For the optimization process, we will use a multi-objective genetic algorithm toolbox which is developed at University of Illinois Urbana-Champaign. The code has a 100% scaling efficiency in single node (8 cores on Fusion and 16 cores on Blues). Molecular dynamics calculations will be carried out with LAMMPS code. This code is tested on Fusion for a ~13000 atom system using a Buckingham + point charge force field. A scaling efficiency of 87% is obtained for 64 cores compared to 8 cores. Preliminary testing shows that an MD simulation of 30,000 atoms and 1 million time steps require 30 hours on 64 cores. Therefore, we estimate that thermal conductivity calculations require 2000 core-hours on average. The time estimates for the various parts of our proposal considering the computational efficiencies are listed below. (1) a- Evaluation of phonon dispersions and elastic constants: 10 calculations x 640 core hours each = 6,400 b- Evaluation of vacancy (vanadium and oxygen) formation energies: 10 calculations x 640 core hours each = 6,400 c- Evaluation of surface energies: 20 calculations x 1280 core hours each = 25,600 (2) Force field optimization: 20 calculation x 2,000 core hours each = 40,000 (3) Thermal conductivity calculations in bulk phases: 8 calculations x 2,000 core hours each = 16,000 (4) Thermal conductivity calculations on defected systems: 108 calculations x 2,000 core hours each = 216,000 (5) Thermal conductivity calculations in VO2 thin films: 24 calculations x 5,000 core hours each = 120,000 The total request is 430,400 core-hours. Dr. Alper Kinaci and Dr. Maria Chan will be responsible for carrying out the simulations. Industry partnership: The project is funded by TEMA and is part of an ongoing partnership between Argonne and Toyota in the experimental and computational investigations of novel ways to control thermal transport Project URL: Requested allocation: 50000 Q1: 0 Q2: 0 Q3: 0 Q4: 50000 Justification: Storage requirements: 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