Erika Michela Dematteis

The development of efficient and sustainable hydrogen storage materials is a key challenge for realizing hydrogen as a clean and flexible energy carrier. Among various options, metal hydrides offer high volumetric storage density and operational safety, yet their application is limited by thermodynamic, kinetic, and compositional constraints. In this work, we investigate the potential of machine learning (ML) to predict key thermodynamic properties—equilibrium plateau pressure, enthalpy, and entropy of hydride formation—based solely on alloy composition using Magpie-generated descriptors. We significantly expand an existing experimental dataset from ~400 to 806 entries and assess the impact of dataset size and data augmentation, using the PADRE algorithm, on model performance. Models including Support Vector Machines and Gradient Boosted Random Forests were trained and optimized via grid search and cross-validation. Results show a marked improvement in predictive accuracy with increased dataset size, while data augmentation benefits are limited to smaller datasets and do not improve accuracy in underrepresented pressure regimes. Furthermore, clustering and cross-validation analyses highlight the limited generalizability of models across different material classes, though high accuracy is achieved when training and testing within a single hydride family (e.g., AB₂). The study demonstrates the viability and limitations of ML for accelerating hydride discovery, emphasizing the importance of dataset diversity and representation for robust property prediction.

The development of efficient and sustainable hydrogen storage materials is a key challenge for realizing hydrogen as a clean and flexible energy carrier. Among various options, metal hydrides offer high volumetric storage density and operational safety, yet their application is limited by thermodynamic, kinetic, and compositional constraints. In this work, we investigate the potential of machine learning (ML) to predict key thermodynamic properties—equilibrium plateau pressure, enthalpy, and entropy of hydride formation—based solely on alloy composition using Magpie-generated descriptors. We significantly expand an existing experimental dataset from ~400 to 806 entries and assess the impact of dataset size and data augmentation, using the PADRE algorithm, on model performance. Models including Support Vector Machines and Gradient Boosted Random Forests were trained and optimized via grid search and cross-validation. Results show a marked improvement in predictive accuracy with increased dataset size, while data augmentation benefits are limited to smaller datasets and do not improve accuracy in underrepresented pressure regimes. Furthermore, clustering and cross-validation analyses highlight the limited generalizability of models across different material classes, though high accuracy is achieved when training and testing within a single hydride family (e.g., AB2). The study demonstrates the viability and limitations of ML for accelerating hydride discovery, emphasizing the importance of dataset diversity and representation for robust property prediction.

The hexagonal structure of LiBH ₄ at room temperature can be stabilised by substituting the BH ₄ ⁻ anion with I ⁻ , leading to high Li-ion conductive materials.

Hydrogen is an efficient energy carrier that can be produced from renewable sources, enabling the transition towards CO2-free energy. Hydrogen can be stored for a long period in the solid-state, with suitable alloys. Ti-rich TiFe0.90 compound exhibits a mild activation process for the first hydrogenation, and Ti(Fe,Mn)0.90 substituted alloys can lead to the fine tuning of equilibrium pressure as a function of the final application. In this study, the crystal structure of TiFe(0.90-x)Mnx alloys (x = 0, 0.05 and 0.10) and their deuterides has been determined by in-situ neutron diffraction, while recording Pressure-Composition Isotherms at room temperature. The investigation aims at analysing the influence of Mn for Fe substitution in Ti-rich Ti(Fe,Mn)0.90 alloys on structural properties during reversible deuterium loading, which is still unsolved and seldom explored. After activation, samples have been transferred into custom-made stainless-steel and aluminium alloy cells used for in-situ neutron diffraction experiments during deuterium loading at ILL and ISIS neutron facilities, respectively. The study enables remarkable understanding on hydrogen storage, basic structural knowledge, and support to the industrial application of TiFe-type alloys for integrated hydrogen tank in energy storage systems by determining the volume expansion during deuteration. Furthermore, the study demonstrates that different contents of Mn do not significantly change the volumetric expansion during phase transitions, affecting only the deuterium content for the {\gamma} phase and the cell evolution for the \b{eta} phase. The study confirms that the deuterated structures of the {\gamma} phase upon absorption, \b{eta} and {\alpha} phase upon desorption, correspond to S.G. Cmmm, P2221 and Pm-3m, respectively.

Industrial and public interest in hydrogen technologies has risen strongly recently, as hydrogen is the ideal means for medium to long term energy storage, transport and usage in combination with renewable and green energy supply. In a future energy system, the production, storage and usage of green hydrogen is a key technology. Hydrogen is and will in future be even more used for industrial production processes as a reduction agent or for the production of synthetic hydrocarbons, especially in the chemical industry and in refineries. Under certain conditions material based systems for hydrogen storage and compression offer advantages over the classical systems based on gaseous or liquid hydrogen. This includes in particular lower maintenance costs, higher reliability and safety. Hydrogen storage is possible at pressures and temperatures much closer to ambient conditions. Hydrogen compression is possible without any moving parts and only by using waste heat. In this paper, we summarize the newest developments of hydrogen carriers for storage and compression and in addition, give an overview of the different research activities in this field.