Akash K Ball, Heather J Kulik
Metal-organic frameworks (MOFs) with cage-like narrow openings and large internal cavities have shown potential for addressing outstanding challenges such as water harvesting and purification. However, a fundamental understanding of water structure and dynamics within these MOFs is essential to achieving high water flux. Here, we take a computational approach to determine the effects of pore confinement and hydrophilicity on water density, hydrogen bonding, diffusivity, and permeability in 78 cage-like MOFs that are predicted by literature-trained machine learning models to be stable in humid/wet conditions, including under water submersion. Across this set, we observe significant variations in water density, hydrogen-bond network, dielectric environment, and diffusion driven by confinement and hydrophilicity of the MOFs. We show that confinement exerts a strong influence on water structure, diffusion, and intrinsic permeability with hydrophilicity exerting a secondary influence. Finally, we establish design principles to achieve exceptional water permeability within cage-like MOFs without compromising the ionic selectivity.