Student name: Mohammad Altaf

RUN-EU institution: Technological University of the Shannon: Midlands Midwest, Ireland

Abstract

The last decade has been the hottest in 125,000 years, with carbon dioxide levels reaching their highest in 2 million years. The planet is losing 1.2 trillion tons of ice annually, and extreme weather events are becoming increasingly severe. If fossil fuel-based energy consumption continues at its current rate, many regions may become too hot for habitation by the end of the century. To combat against climate change, several Sustainable Development Goals (SDGs) were to shift toward cleaner alternatives. Hydrogen presents a promising solution due to its high calorific value of 140 MJ/kg, zero carbon emissions, and versatility across various applications. However, hydrogen storage remains a major challenge, as it is highly explosive and requires either large, high-pressure tanks or cryogenic temperatures. These challenges can be solved through solid-state hydrogen storage which offers a safer and more efficient alternative by storing hydrogen in interstitial sites within metal hydrides such as magnesium. However, it has challenges related to kinetics, stability, and cyclability. This study aims to solve these challenges through experimentation and use of Artificial Neural Networks (ANN) to predict the optimal composition for efficient hydrogen storage. In experimentations, suitable catalyst, additives and elemental substitution will be identified and added in magnesium to boost its storage performance. Additionally, suitable polymers will be used to provide air stability. The data from the experimentation will further be used to frame and train ANN model, that can predict the best composition for the composite for safe and efficient hydrogen storage.