Characterisation of organic solvent nanofiltration membranes in multi-component mixtures: Phenomena-based modelling and membrane modelling maps

Organic solvent nanofiltration (OSN) is a promising, energy saving membrane separation technology for the purification and separation of organic liquids. Challenges for process design include the selection of a suitable membrane and a suitable solvent (mixture) in the process. To solve these challenges, decisions are currently taken based on costly and time-consuming OSN membrane screening experiments. Besides the large associated experimental effort, they often fail to identify the best suitable combination of OSN membrane and solvent (mixture) due to their non-targeted nature. Therefore, modelling of the underlying permeation mechanism to predict solvent fluxes/solute rejections and, more important, to generate insight into the transport, is recommended. In this paper, a phenomena-based model for multi-component permeation through polymeric OSN membranes based on solution–diffusion, pore-flow and mutual coupling terms is presented. Instead of developing a very detailed model, the objective of the model is to attain a fixed set of binary interaction parameters for a given OSN membrane, solvents and solutes in analogy to vapour–liquid equilibria (VLE) parameter sets. Moreover, the processing of insights gained by the model is demonstrated, leading to a suggestion for a separation improvement, either being focused on the membrane or on the applied solvent (mixture). As a conceptual graphical tool, membrane modelling maps (MMM) are introduced. MMM highlight the shares of the permeation phenomena and give recommendations for rejection improvement in dependence of the solute and the applied solvent (mixture). As an experimental basis for parameter estimation and model evaluation, permeability and rejection data in binary and ternary mixtures of toluene, n-hexane and 2-propanol for Starmem™122 from a previous publication is used.