Computational investigations of Bio-MOF membranes for uremic toxin separation

Developing new and efficient methods as an alternative to hemodialysis is important due to the challenges associated with poor efficiency of membranes and long dialysis sessions. Recently, metal organic frameworks (MOFs) have attracted interest in the membrane community due to their tunable physical and chemical properties. However, their potential in uremic toxin separations is still unknown and it is not practical to test each synthesized MOF for uremic toxin separations. The main objective of this study is to computationally assess membrane-based uremic toxin separation performances of 60 bio-compatible MOFs (bio-MOFs). Combining grand canonical Monte Carlo (GCMC) and equilibrium molecular dynamics (EMD) simulations, we predicted urea, creatinine, and water permeabilities of bio-MOFs and their membrane selectivities for urea/water and creatinine/water separations. Results showed that OREZES, a carboxylate-based MOF exhibited the highest membrane selectivity (347.94) for urea/water separation whereas BEPPIX, an amino-based MOF gave the highest creatinine/water selectivity (1.5 x 105) at infinite dilution and 310 K. Guest-guest and host-guest interaction energies for uremic toxins were also computed during EMD simulations and van der Waals interactions were found to be much stronger than the coulombic interactions. We finally examined the effect of MOF's flexibility on the predicted membrane performance and membrane selectivities of bio-MOFs for urea/ water separation significantly enhanced when the structural flexibility was considered in simulations. Our results will be a guide for further studies to design novel bio-MOF membranes for uremic toxin separations.