• Dražević, Emil
• Szymcyk, Anthony
• Ghoufi, Aziz

Abstract

<p>In this work we have examined a computational approach in predicting the interactions between uncharged organic solutes and polyamide membranes. We used three model organic molecules with identical molecular weights (100.1 g/mol), 4-aminopiperidine, 3,3-dimethyl-2-butanone (pinacolone) and methylisobutyl ketone for which we obtained experimental data on partitioning, diffusion and separation on a typical seawater reverse osmosis (RO) membrane. The interaction energy between the solutes and the membrane phase (fully aromatic polyamide) was computed from molecular dynamics (MD) simulations and the resulting sequence was found to correlate well with the experimental rejections and sorption data. Sorption of the different organic solutes within the membrane skin layer determined from attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR) nicely agreed with interaction energies computed from molecular simulations. Qualitative information about solute diffusivity inside the membrane was also extracted from MD simulations while ATR-FTIR experiments indicated strongly hindered diffusion with diffusion coefficients in the membrane about 10^-15 m²/s. The computational approach presented here could be a first step toward predicting rejections trends of, for example, hormones and pharmaceuticals by RO dense membranes.</p>

Topics

• phase
• experiment
• simulation
• molecular dynamics
• molecular weight
• diffusivity
• ketone
• Fourier transform infrared spectroscopy