Untitled

We present a comparative study of the filtration performance of one ceramic and two organic nanofilters. The results of a detailed experimental and theoretical investigation of the rejection of single salts (NaCl, NaNO3, and CaCl2) and multi-electrolyte mixtures (NaCl/CaCl2/HCl and NaNO3/NaCl/CaCl2) by a loose ceramic TiO2 nanofilter (Membralox SCT) are compared with similar results obtained using two organic nanofilters (Desal 5 and NF200). We modelion transport in nanoscale pores using a hybrid Hindered Electro-Transport Theory (HETT) that accounts for the charge and size of ions. Theoretical ion rejections for multi-electrolyte mixtures are obtained by solving numerically the Hindered Transport extended Nernst-Planck ion flux Eqs. using a computer simulation program, NanoFlux, that we have developed to simulate transport in the nanofiltration range. The ion rejection in binary and ternary mixtures is reasonably well predicted by the theory, over a wide range of pH and relative salt mole fraction, from the single salt results, using an appropriate weighting and interpolation scheme.

By using NanoFlux and its internal database, the rejection performance of different membranes can be compared directly, under the same operating conditions, for different salt compositions. In contrast to previous HETT approaches, our method can be used to study feed mixtures of up to 11 ions. We also treat the effective membrane thickness as an adjustable parameter (on par with the effective membrane charge density) and employ a new choice of ionic radii (Pauling crystal radius). We propose that NanoFlux can be the kernel of a standardized combined experimental and theoretical method for characterizing, simulating, and optimizing the performance of charged nanofilters. This simulation tool attempts to fill in a gap in methods for predicting nanofiltration performance that has arisen due to the complexity of the fundamental ion transport processes and has probably slowed the industrial development of nanofiltration.