Predicting the Heat of Hydride Formation by Graph Neural Network - Exploring the Structure-Property Relation for Metal Hydrides
Само за регистроване кориснике
2022
Аутори
Batalović, KatarinaRadaković, Jana
Paskaš Mamula, Bojana
Kuzmanović, Bojana
Medić-Ilić, Mirjana
Чланак у часопису (Објављена верзија)
Метаподаци
Приказ свих података о документуАпстракт
Theoretical tools or structure-property relations that enable the prediction of metal hydrides are of enormous interest in developing new hydrogen storage materials. Density functional theory (DFT) is one such approach that provides accurate hydride formation energies, which, if complemented with vibrational zero-point energy and other contributions, provides accurate hydride formation enthalpies. However, this approach is time consuming and, therefore, often avoided, hindering the modeling of experimental behavior. The recent implementation of graph neural networks (GNN) in materials science enables fast prediction of crystal formation energy with a DFT accuracy. Starting from the MatErials Graph Network (MEGNet), transfer learning is applied to develop a model for predicting hydride formation enthalpy based on the crystal structure of the starting intermetallic. Excellent accuracy is achieved for Mg-containing alloys, allowing the screening of the Mg-Ni-M ternary intermetallics. In a...ddition, data containing matching experimental properties and crystal structure of metal hydrides are provided, enabling future development.
Кључне речи:
DFT / discovery / hydrogen storage / intermetallic compounds / learning based prediction / machine learning / metal hydride / mg / Mg2Ni / Mg3MnNi2 / progress / tiИзвор:
Advanced Theory and Simulations, 2022, 5, 9, 2200293-Финансирање / пројекти:
- Ministry of Education, Science, and Technological Development of the Republic of Serbia
Напомена:
- Preprint version available at: https://dx.doi.org/10.2139/ssrn.4055259
- Data can be found at: https://vinar.vin.bg.ac.rs/handle/123456789/11282
Повезане информације:
- Повезани садржај
https://vinar.vin.bg.ac.rs/handle/123456789/11282
DOI: 10.1002/adts.202200293
ISSN: 2513-0390
WoS: 000819550400001
Scopus: 2-s2.0-85133161452
Институција/група
VinčaTY - JOUR AU - Batalović, Katarina AU - Radaković, Jana AU - Paskaš Mamula, Bojana AU - Kuzmanović, Bojana AU - Medić-Ilić, Mirjana PY - 2022 UR - https://vinar.vin.bg.ac.rs/handle/123456789/10348 AB - Theoretical tools or structure-property relations that enable the prediction of metal hydrides are of enormous interest in developing new hydrogen storage materials. Density functional theory (DFT) is one such approach that provides accurate hydride formation energies, which, if complemented with vibrational zero-point energy and other contributions, provides accurate hydride formation enthalpies. However, this approach is time consuming and, therefore, often avoided, hindering the modeling of experimental behavior. The recent implementation of graph neural networks (GNN) in materials science enables fast prediction of crystal formation energy with a DFT accuracy. Starting from the MatErials Graph Network (MEGNet), transfer learning is applied to develop a model for predicting hydride formation enthalpy based on the crystal structure of the starting intermetallic. Excellent accuracy is achieved for Mg-containing alloys, allowing the screening of the Mg-Ni-M ternary intermetallics. In addition, data containing matching experimental properties and crystal structure of metal hydrides are provided, enabling future development. T2 - Advanced Theory and Simulations T1 - Predicting the Heat of Hydride Formation by Graph Neural Network - Exploring the Structure-Property Relation for Metal Hydrides VL - 5 IS - 9 SP - 2200293 DO - 10.1002/adts.202200293 ER -
@article{ author = "Batalović, Katarina and Radaković, Jana and Paskaš Mamula, Bojana and Kuzmanović, Bojana and Medić-Ilić, Mirjana", year = "2022", abstract = "Theoretical tools or structure-property relations that enable the prediction of metal hydrides are of enormous interest in developing new hydrogen storage materials. Density functional theory (DFT) is one such approach that provides accurate hydride formation energies, which, if complemented with vibrational zero-point energy and other contributions, provides accurate hydride formation enthalpies. However, this approach is time consuming and, therefore, often avoided, hindering the modeling of experimental behavior. The recent implementation of graph neural networks (GNN) in materials science enables fast prediction of crystal formation energy with a DFT accuracy. Starting from the MatErials Graph Network (MEGNet), transfer learning is applied to develop a model for predicting hydride formation enthalpy based on the crystal structure of the starting intermetallic. Excellent accuracy is achieved for Mg-containing alloys, allowing the screening of the Mg-Ni-M ternary intermetallics. In addition, data containing matching experimental properties and crystal structure of metal hydrides are provided, enabling future development.", journal = "Advanced Theory and Simulations", title = "Predicting the Heat of Hydride Formation by Graph Neural Network - Exploring the Structure-Property Relation for Metal Hydrides", volume = "5", number = "9", pages = "2200293", doi = "10.1002/adts.202200293" }
Batalović, K., Radaković, J., Paskaš Mamula, B., Kuzmanović, B.,& Medić-Ilić, M.. (2022). Predicting the Heat of Hydride Formation by Graph Neural Network - Exploring the Structure-Property Relation for Metal Hydrides. in Advanced Theory and Simulations, 5(9), 2200293. https://doi.org/10.1002/adts.202200293
Batalović K, Radaković J, Paskaš Mamula B, Kuzmanović B, Medić-Ilić M. Predicting the Heat of Hydride Formation by Graph Neural Network - Exploring the Structure-Property Relation for Metal Hydrides. in Advanced Theory and Simulations. 2022;5(9):2200293. doi:10.1002/adts.202200293 .
Batalović, Katarina, Radaković, Jana, Paskaš Mamula, Bojana, Kuzmanović, Bojana, Medić-Ilić, Mirjana, "Predicting the Heat of Hydride Formation by Graph Neural Network - Exploring the Structure-Property Relation for Metal Hydrides" in Advanced Theory and Simulations, 5, no. 9 (2022):2200293, https://doi.org/10.1002/adts.202200293 . .