Interpretable AI reveals key atomic traits for efficient hydrogen storage in metal hydrides

Hydrogen fuel is a clean energy option, but the main barrier to its wider use is efficient storage. Storing hydrogen requires either extremely high-pressure tanks or extremely low temperatures, meaning storage itself uses a lot of energy. This is why metal hydrides, which can store hydrogen more efficiently, are such a promising option.

New digital platform advances hydrogen research

To accurately predict performance metrics hydrogen storage Materials Researchers from Tohoku University used a newly created data infrastructure: the Digital Hydrogen Platform (DigHyd). DigHyd brings together more than 5,000 carefully selected experimental records from the literature, supported by an artificial intelligence language model. Job published in the magazine Chemical Science.

Using this vast database, the researchers systematically examined physically interpretable models and discovered that the fundamental features of the atom are:atomic masselectronegativity, molar density and ionic occupancy factor become key descriptors. Other researchers can use this as a tool to guide the materials design process without having to go through a lengthy process of trial and error in the lab to find the right materials. potential candidates.

“This white-box regression model not only makes accurate predictions, but also maintains full physical interpretability,” explains Hao Li, Distinguished Professor at the Advanced Institute for Materials Research (WPI-AIMR) at Tohoku University. “This means it is transparent, unlike traditional black-box machine learning approaches where it is unclear how the model calculated the final answer.”

This transparency allows scientists to define design strategies because the model shows mathematically simple but clearly interpretable expressions for performance targets. By matching fundamental atomic-scale properties with measurable storage behavior, the models provide a clear and chemically intuitive picture of how material composition controls hydrogen uptake and release.

Key findings and future directions

The study also revealed a fundamental trade-off that drives the current landscape of metal hydrides. Compounds consisting of light electropositive elements have a high hydrogen capacity, but give a low equilibrium pressure at room temperaturewhile those based on heavier transition metals release hydrogen more easily, but at the cost of capacity. Remarkably, beryllium-based alloys have proven to be unique systems capable of balancing these conflicting characteristics by combining high storage density and suitable thermodynamic stability.

In addition to identifying promising candidates, this work establishes a methodology for accelerating discovery in the field of energetic materials research. The descriptor-based framework offers a new paradigm for combining data analysis with physical understanding, providing a scalable and transparent framework for the design of hydrogen storage materials.

This approach can be extended to more complex alloys and porous structures, paving the way for the development of safe, efficient and high-performance hydrogen storage systems that will support the transition to clean, carbon-neutral energy technologies.