Analytical framework for nanofiltration processes with imperfect rejection and its application to curcumin concentration

Organic solvent nanofiltration (OSN) performance is influenced by several factors, including the solvent, membrane, transmembrane pressure, system size, flowrate, and desired end concentration. System design and optimization of such a complex process requires a unified framework so that the effect of each influence can be compared quantitatively in a single plot. To that effect, a mathematical framework was derived by functionally relating the target quantities (solute yield and solvent recovery) interconnected by the mass balances of each species. The plot obtained indicates solute-solvent selectivity, solute yield, solvent recovery, closeness to the saturation concentration, and a comparison of system sizes; hence when plotted for varied operating conditions and solvent-membrane systems, such a plot is instructive to improved system design. In the current paper, this framework is illustrated for OSN implemented for curcumin extract concentration. Two membranes (Evonik Puramem Performance, PMP, and Evonik Puramem Selective, PMS) were compared at various pressures on a single plot. The key output values obtained are solute yield, solvent recovery, system size comparison, and permeate bulk solute concentration. It was shown using the framework that when operated at 10 bar, the PMS membrane has a higher solute yield (180%); however, it requires a larger membrane area (25%) to reach saturation, compared to PMP. Finally, the effect of concentration polarization on the separation is demonstrated using the framework.