Improving transferable force-field for MOFs using classical density functional theory

Metal-organic frameworks (MOFs), porous materials composed of metallic centers and organic ligands, have become a versatile class of materials due to their customizable design and the flexibility of their pore functionalization. These materials have high porosity and high specific surface area, which has led to numerous applications in gas storage, separations, and catalysis. The traditional computational method for predicting gas adsorption in MOFs is Grand Canonical Monte Carlo (GCMC). The agreement between the results obtained via GCMC and experimental data is directly related to the choice of the force field for the solid structure. Another method for computing adsorption in MOFs with lower computational cost is the classical Density Functional Theory (cDFT). The cDFT is a theoretical framework that describes the spatial distribution of fluid molecules within a porous material by minimizing the free energy functional of the density field. This work focused on estimating a new set of Lennard-Jones parameters for solid-solid interactions to predict adsorption in MOFs that outperform traditional general force fields, utilizing experimental data of methane adsorption in MOF-5, MOF-177, MOF-200, and MOF-205 at 298 K across a pressure range of 0.5 to 85 bar, and 3D-cDFT to calculate theoretical isotherms. The solid-solid interaction parameters were estimated simultaneously for all MOFs, directly minimizing the objective function using the Sequential Least Squares Programming (SLSQP) method. Using the parameters estimated from methane adsorption for another adsorbate, carbon dioxide, it remains possible to improve the prediction of experimental data using GCMC, showing the transferability of the force field to other systems. The proposed strategy for obtaining solid-solid parameters outperforms the conventional DREIDING force field in the analysed cases, improving the adsorption predictions while maintaining computational efficiency.