471505 High-Throughput Prediction of Finite Temperature Free Energies of Solids

Tuesday, November 15, 2016: 3:51 PM
Golden Gate 4 (Hilton San Francisco Union Square)
Christopher J. Bartel1, Ann M. Deml2,3, Samantha L. Miller1, Alan W. Weimer4, Stephan Lany2, Charles B. Musgrave1, Vladan Stevanovic2,3 and Aaron Holder1,2, (1)Chemical & Biological Engineering, University of Colorado, Boulder, CO, (2)Energy Sciences, National Renewable Energy Laboratory, Golden, CO, (3)Metallurgical and Materials Engineering, Colorado School of Mines, Golden, CO, (4)Chemical & Biological Engineering, University of Colorado at Boulder, Boulder, CO

High-throughput, theory-driven computational materials design and discovery has become an essential tool in materials science and engineering. The data-driven and cost-effective framework for accelerated discovery introduced by the Materials Genome Initiative (MGI) has transformed the scale and rate of materials development by exploiting the predictive ability of quantum chemical computational methods. In turn, the drive to rapidly predict, screen, and optimize materials using first-principles calculations has led to large datasets and the construction of open-source databases that are populated primarily by low-cost density functional theory (DFT) total energy calculations (inherently performed at 0 K). Although these datasets are an invaluable resource, their predictive ability at finite temperatures is limited and current methods for evaluating the missing temperature dependencies are computationally prohibitive, limiting the success of MGI approaches for existing and emerging high temperature applications. In this report, we show that compound free energies of formation can be predicted with small errors at finite temperatures of up to at least 1800 K, and at the same computational expense as a DFT total energy calculation.

Our broadly applicable method for evaluating temperature-dependent compound free energies of formation employs readily computable material-specific descriptors and utilizes existing datasets of more than 50,000 inorganic compounds to enable rapid computational prototyping at elevated temperatures. We build upon prior work that has demonstrated the use of total energy calculations to approximate the compound enthalpy of formation at 298 K with mean absolute error relative to experiment (MAE) = 0.054 eV/atom with fitted elemental-phase reference energies (FERE). Our prediction requires only the DFT total energy, the FERE correction, and two minimal-computational-cost material-specific predictors to compute compound free energies of formation with MAE < 0.1 eV/atom up to at least 1800 K. At 1000 K, the MAE between our prediction and experiment is 0.039 eV/atom for oxides, 0.057 eV/atom for nitrides, 0.055 eV/atom for selenides, 0.049 eV/atom for arsenides, 0.044 eV/atom for phosphides, and 0.037 eV/atom for sulfides. These results bridge the temperature gap between existing MGI inspired computational approaches, which have historically been limited to low temperature, and high temperature materials applications, providing the equivalent of tabulated thermochemical data for more than 50,000 inorganic materials.

Extended Abstract: File Not Uploaded