High-throughput Prediction of High-pressure Materials Properties for the AFLOWLIB Database
Abstract
High-throughput first principles methods and machine learning will be used to predict the behavior of materials at high pressure.Pressure and temperature properties have already been calculated for over 2000 materials in the AFLOWLIB database using the AGL Debye model and AEL elastic constants algorithms implemented in AFLOW, and more data is being generated continuously. This data set will be used to screen for candidate materials and to calibrate machine learning algorithms. More detailed investigations of the high pressure phonon properties for promising materials will be carried out using the recently implemented quasi-harmonic extension to APL. Particular properties of interest include pressure induced structural phase transitions and high pressure stabilization of particular crystal phases, pressure induced metallization and superconductivity, and pressure e ects on lattice thermal conductivity and measures of hardness.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Nov 23, 2016
- Source ID
- N000141612583
Entities
People
- Stefano Curtarolo
Organizations
- Duke University
- Office of Naval Research
- United States Navy