When the membrane is not enough: A simplified membrane cascade using Organic Solvent Nanofiltration (OSN)

2013 
Abstract Organic Solvent Nanofiltration (OSN) is a relatively new molecular separation technology used for separating solutes present in an organic solvent. Although there are many potential applications in industrial processes, obtaining a sharp separation between two molecules which are both in the nanofiltration range from 100 to 2000 g mol −1 remains a challenge. This is because the differences in rejection of the solutes by the membrane are often insufficient for them to be separated in a single filtration stage. Membrane cascades can meet this challenge and have potential for process intensification (PI) allowing more sustainable membrane units. The membrane cascade concept was proposed in the early 1940s, but to date several implementation challenges such as control difficulties have presented major hurdles and only few experimental data have been reported. Here we present a simplified-control cascade process that not only overcomes previous implementation issues, but also minimizes the product loss and maximizes the purity. The process operates using a single high pressure pump as the primary pressure source, and has no need for a buffer tank between membrane stages. The process was tested on an organic solution of polyethyleneglycol 400 (PEG-400) and PEG-2000 in acetonitrile which is a challenging model, as the flexible chains make it difficult to obtain 100% rejection of any PEG. Polyimide and polybenzimidazole membranes were screened for the cascade process. Using this process it was possible to increase the final yield of PEG-2000 from 59% to 94%, and the solvent-consumption/productivity ratio was also reduced. The process has been analyzed using a model developed to predict the cascade performance. It was found experimentally that the recycle ratio, as well as the trans-membrane pressure, affect the process yield significantly, in agreement with the model predictions. Model analysis and experimental validation are presented.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    52
    References
    42
    Citations
    NaN
    KQI
    []