Combining the AFLOW GIBBS and Elastic Libraries for efficiently and robustly screening thermo-mechanical properties of solids

Abstract

Thorough characterization of the thermo-mechanical properties of materials requires difficult and time-consuming experiments. This severely limits the availability of data and it is one of the main obstacles for the development of effective accelerated materials design strategies. The rapid screening of new potential systems requires highly integrated, sophisticated and robust computational approaches. We tackled the challenge by surveying more than 3,000 crystalline solids within the AFLOW framework with the newly developed "Automatic Elasticity Library" combined with the previously implemented GIBBS method. The first extracts the mechanical properties from automatic self-consistent stress-strain calculations, while the latter employs those mechanical properties to evaluate the thermodynamics within the Debye model. The new thermo-elastic library is benchmarked against a set of 74 experimentally characterized systems to pinpoint a robust computational methodology for the evaluation of bulk and shear moduli, Poisson ratios, Debye temperatures, Grüneisen parameters, and thermal conductivities of a wide variety of materials. The effect of different choices of equations of state is examined and the optimum combination of properties for the Leibfried-Schlömann prediction of thermal conductivity is identified, leading to improved agreement with experimental results than the GIBBS-only approach.

29 pages, 20 panel figures, 23 tables

References in corpus (5)

• Necessary and Sufficient Elastic Stability Conditions in Various Crystal Systems
• Intrinsic Correlation between Hardness and Elasticity in Polycrystalline Materials and Bulk Metallic Glasses
• High-Throughput Computational Screening of thermal conductivity, Debye temperature and Grüneisen parameter using a quasi-harmonic Debye Model
• A RESTful API for exchanging Materials Data in the AFLOWLIB.org consortium
• First principles search for $n$-type oxide, nitride, and sulfide thermoelectrics

Cited by in corpus (9)

• High-entropy high-hardness metal carbides discovered by entropy descriptors
• SISSO: a compressed-sensing method for identifying the best low-dimensional descriptor in an immensity of offered candidates
• Machine learning modeling of superconducting critical temperature
• Using the Callaway model to deduce relevant phonon scattering processes: The importance of phonon dispersion
• The molybdenum-titanium phase diagram evaluated from ab-initio calculations
• Automated coordination corrected enthalpies with AFLOW-CCE
• AFLOW-QHA3P: Robust and automated method to compute thermodynamic properties of solids
• PINK: physical-informed machine learning for lattice thermal conductivity
• Machine learning-driven elasticity prediction in advanced inorganic materials via convolutional neural networks