Decoding local framework dynamics in the ultra-small pore MOF MIL-120(Al) CO2 sorbent with Machine Learned Potentials

Metal-organic frameworks (MOFs) with ultra-small pores offer an optimal environment to effectively capture guest molecules such as CO2. Subtle local dynamics of their frameworks, either throughout reorientation of functional groups grafted to the organic linkers or those present in their inorganic nodes, is expected to play a major role in their sorption behaviors. Here, we combine density-functional theory (DFT) with a purpose-trained machine-learned potential to systematically investigate the local dynamics of the bridging hydroxyl groups, {\mu}2-OH groups present in the prototypical ultra-small pore MOF MIL-120(Al), reported recently as an attractive CO2 sorbent. We identified six MOF configurations associated with distinct {\mu}2-OH orientations with relatively low interconversion energy barriers (0.07-0.19 eV per unit cell) suggesting that all these states can be observed experimentally at room temperature. We demonstrated that our MLP achieves near-DFT-level fidelity, reproducing the energy barriers and phonon spectra of the empty MOF, and accurately predicting CO2 adsorption geometries depending on the {\mu}2-OH orientations with CO2 adopting either parallel or perpendicular alignment to the pore axis, which in turn governs the adsorption energetics. This work establishes that a reliable description of the local structure, such as reorientation/flipping of bridging hydroxyl groups, is a key feature to gain an accurate description of the guest locations and energetics in ultra-small pore MOFs.