Search NASA⌕ Search

Engineering topics

Goyal, Anuj

Publications and source records attributed to Goyal, Anuj.

5 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

LaMnO 3 Dopants for Efficient Thermochemical Water Splitting Identified by Density Functional Theory Calculations

While ceria is the standard material for two-step water splitting, perovskites are emerging as viable alternatives. In this work, based on the orthorhombic LaMnO 3 supercell, we substitute Li Na K Rb Mg Ca Sr Ba on the A-sites (La sites) and Al Ga In Mg Zn on the B-sites (Mn sites) at a concentration of 37.5%. The range of temperature and oxygen partial pressure at which each composition is stable is predicted. For compositions that are stable in relevant temperature and pressure ranges, the oxygen vacancy formation energies are determined for all of the oxygen vacancy site positions available in the computational supercell. Mg, Ca, Sr, and Ba A-site-substituted LaMnO 3 and Al and In B-site-substituted LaMnO 3 meet these two criteria for candidates in solar-thermal water splitting applications. Finally, oxygen vacancy formation energy can also be controlled by adjusting the doping strategy.

08 HYDROGEN↗
DOE OSTI2023

Defect graph neural networks for materials discovery in high-temperature clean-energy applications

We present a graph neural network approach that fully automates the prediction of defect formation enthalpies for any crystallographic site from the ideal crystal structure, without the need to create defected atomic structure models as input. Here we used density functional theory reference data for vacancy defects in oxides, to train a defect graph neural network (dGNN) model that replaces the density functional theory supercell relaxations otherwise required for each symmetrically unique crystal site. Interfaced with thermodynamic calculations of reduction entropies and associated free energies, the dGNN model is applied to the screening of oxides in the Materials Project database, connecting the zero-kelvin defect enthalpies to high-temperature process conditions relevant for solar thermochemical hydrogen production and other energy applications. The dGNN approach is applicable to arbitrary structures with an accuracy limited principally by the amount and diversity of the training data, and it is generalizable to other defect types and advanced graph convolution architectures. In conclusion, it will help to tackle future materials discovery problems in clean energy and beyond.

97 MATHEMATICS AND COMPUTING↗