Data-driven Design of New Mg-based Hydride Materials – A Synergy of Experiments and DFT
Аутори
Batalović, KatarinaRadaković, Jana
Kuzmanović, Bojana
Medić-Ilić, Mirjana
Paskaš Mamula, Bojana
Конференцијски прилог (Објављена верзија)
Метаподаци
Приказ свих података о документуАпстракт
Hydrogen absorption/desorption is one of the key processes underlying many clean energy applications, such as thermal energy storage, hydrogen storage, hydrogen compression, and nickel-metal hydride batteries. For all those applications fast and reliable characterization of new materials, and in particular, information regarding energetics of hydride formation reaction is of main interest. In the last decades, DFT (density functional theory) approach showed good predictive potential for the ground state properties and calculation of hydride formation energies. Recently, MEGNet implementation of graph neural networks showed promising results for fast and reliable prediction of formation energies for molecules and crystals. Here, we consider the development of a machine learning model based on the available DFT predicted structures and experimentally measured hydride formation enthalpies. The proposed model is capable to predict hydride formation behavior for a wide variety of intermetal...lic compounds and distinguish the behavior of the polymorphs. In particular, based only on the crystal structure of the starting intermetallic compound, we were able to predict hydride formation enthalpy with accuracy comparable to DFT calculated values. Further, we demonstrate the application of this model for proposing new materials in Mg-Ni-M compound space with the desired enthalpy for hydrogen storage.
Извор:
COIN2022 - Contemporary Batteries and Supercapacitors - International Symposium : Program and Book of Abstracts, 2022, 49-49Издавач:
- Belgrade : Faculty of Physical Chemistry, University of Belgrade
Напомена:
- COIN2022 - Contemporary Batteries and Supercapacitors - International Symposium ; June 1-2, 2022 ; Belgrade, Serbia
Колекције
Институција/група
VinčaTY - CONF AU - Batalović, Katarina AU - Radaković, Jana AU - Kuzmanović, Bojana AU - Medić-Ilić, Mirjana AU - Paskaš Mamula, Bojana PY - 2022 UR - https://vinar.vin.bg.ac.rs/handle/123456789/11287 AB - Hydrogen absorption/desorption is one of the key processes underlying many clean energy applications, such as thermal energy storage, hydrogen storage, hydrogen compression, and nickel-metal hydride batteries. For all those applications fast and reliable characterization of new materials, and in particular, information regarding energetics of hydride formation reaction is of main interest. In the last decades, DFT (density functional theory) approach showed good predictive potential for the ground state properties and calculation of hydride formation energies. Recently, MEGNet implementation of graph neural networks showed promising results for fast and reliable prediction of formation energies for molecules and crystals. Here, we consider the development of a machine learning model based on the available DFT predicted structures and experimentally measured hydride formation enthalpies. The proposed model is capable to predict hydride formation behavior for a wide variety of intermetallic compounds and distinguish the behavior of the polymorphs. In particular, based only on the crystal structure of the starting intermetallic compound, we were able to predict hydride formation enthalpy with accuracy comparable to DFT calculated values. Further, we demonstrate the application of this model for proposing new materials in Mg-Ni-M compound space with the desired enthalpy for hydrogen storage. PB - Belgrade : Faculty of Physical Chemistry, University of Belgrade C3 - COIN2022 - Contemporary Batteries and Supercapacitors - International Symposium : Program and Book of Abstracts T1 - Data-driven Design of New Mg-based Hydride Materials – A Synergy of Experiments and DFT SP - 49 EP - 49 UR - https://hdl.handle.net/21.15107/rcub_vinar_11287 ER -
@conference{ author = "Batalović, Katarina and Radaković, Jana and Kuzmanović, Bojana and Medić-Ilić, Mirjana and Paskaš Mamula, Bojana", year = "2022", abstract = "Hydrogen absorption/desorption is one of the key processes underlying many clean energy applications, such as thermal energy storage, hydrogen storage, hydrogen compression, and nickel-metal hydride batteries. For all those applications fast and reliable characterization of new materials, and in particular, information regarding energetics of hydride formation reaction is of main interest. In the last decades, DFT (density functional theory) approach showed good predictive potential for the ground state properties and calculation of hydride formation energies. Recently, MEGNet implementation of graph neural networks showed promising results for fast and reliable prediction of formation energies for molecules and crystals. Here, we consider the development of a machine learning model based on the available DFT predicted structures and experimentally measured hydride formation enthalpies. The proposed model is capable to predict hydride formation behavior for a wide variety of intermetallic compounds and distinguish the behavior of the polymorphs. In particular, based only on the crystal structure of the starting intermetallic compound, we were able to predict hydride formation enthalpy with accuracy comparable to DFT calculated values. Further, we demonstrate the application of this model for proposing new materials in Mg-Ni-M compound space with the desired enthalpy for hydrogen storage.", publisher = "Belgrade : Faculty of Physical Chemistry, University of Belgrade", journal = "COIN2022 - Contemporary Batteries and Supercapacitors - International Symposium : Program and Book of Abstracts", title = "Data-driven Design of New Mg-based Hydride Materials – A Synergy of Experiments and DFT", pages = "49-49", url = "https://hdl.handle.net/21.15107/rcub_vinar_11287" }
Batalović, K., Radaković, J., Kuzmanović, B., Medić-Ilić, M.,& Paskaš Mamula, B.. (2022). Data-driven Design of New Mg-based Hydride Materials – A Synergy of Experiments and DFT. in COIN2022 - Contemporary Batteries and Supercapacitors - International Symposium : Program and Book of Abstracts Belgrade : Faculty of Physical Chemistry, University of Belgrade., 49-49. https://hdl.handle.net/21.15107/rcub_vinar_11287
Batalović K, Radaković J, Kuzmanović B, Medić-Ilić M, Paskaš Mamula B. Data-driven Design of New Mg-based Hydride Materials – A Synergy of Experiments and DFT. in COIN2022 - Contemporary Batteries and Supercapacitors - International Symposium : Program and Book of Abstracts. 2022;:49-49. https://hdl.handle.net/21.15107/rcub_vinar_11287 .
Batalović, Katarina, Radaković, Jana, Kuzmanović, Bojana, Medić-Ilić, Mirjana, Paskaš Mamula, Bojana, "Data-driven Design of New Mg-based Hydride Materials – A Synergy of Experiments and DFT" in COIN2022 - Contemporary Batteries and Supercapacitors - International Symposium : Program and Book of Abstracts (2022):49-49, https://hdl.handle.net/21.15107/rcub_vinar_11287 .