Frontier Research Institute for Interdisciplinary Sciences
Tohoku University

In recent years, the increasing demand for efficient and sustainable resource extraction has highlighted the need for advanced separation technologies. This research aims to develop a novel membrane platform to extract helium (0.3%) from natural gas with over 90% cost efficiency. Helium is vital for Japan's electronics industry, consuming 8% of the global supply. Traditional extraction methods, such as cryogenic distillation, are unsustainable and costly, leading to high helium prices (~$80 per liter).

We propose an innovative approach that integrates microporous crystal design with advanced machine learning algorithms to enhance material performance screening. The primary goal is to develop metal-organic framework (MOF) membranes with precision-engineered pore structures to maximize helium selectivity. By leveraging machine learning, we aim to elucidate the intricate functionality-structure relationship of MOF membranes, significantly improving the efficiency of material discovery and optimization.

To achieve this, we will synthesize MOF-membranes by crafting tunable, cross-linkable networks for enhanced MOF integration, optimizing adsorption and permeation. This involves adjusting solution viscosity, surfactant levels, and coating processes to refine membrane thickness, resulting in a solution-processable MOF-based membrane with permanent micropores. Once synthesized, the MOF-membranes will be tested to validate the helium separation coefficient and efficacy of the materials. We will use molecular dynamics simulations with a machine-learning-trained force field (MLFF) to predict and evaluate the relationships between structure, properties, adsorption, and separation in the MOF membranes. This will involve creating an integrated database incorporating experimental and computational data related to helium separation performance.

This research will contribute to understanding gas separation technology and pave the way for developing efficient and cost-effective helium extraction methods. By systematically integrating machine learning with MOF membrane design, this research will revolutionize helium separation efficiency, driving technological advancements and applications in various industries.