1. Effect of the manufacturing conditions on the structure and performance of thin-film composite membranes
- Author
-
Arcadio Sotto, Lore Poelmans, Johan Vanneste, Bart Van der Bruggen, Ruixin Zhang, and Xiaolin Wang
- Subjects
Materials science ,Polymers and Plastics ,Scanning electron microscope ,Composite number ,General Chemistry ,Casting ,Surfaces, Coatings and Films ,Membrane ,Thin-film composite membrane ,Polyamide ,Materials Chemistry ,Wetting ,Composite material ,Curing (chemistry) - Abstract
Received 4 August 2011; accepted 22 November 2011DOI 10.1002/app.36542Published online 24 February 2012 in Wiley Online Library (wileyonlinelibrary.com).ABSTRACT: A systematic investigation of the influenceof the manufacturing conditions on the structure and per-formance of thin-film composite (TFC) membranes is pre-sented for polyamide (PA) supported by poly(ethersulfone) (PES). The TFC membranes were composed of anultrathin PA layer synthesized by interfacial polymeriza-tion on top of a porous PES support layer formed byimmersion precipitation. For the PES support layer, therole of the wetting pretreatment, initial casting film thick-ness, and relative air humidity were studied. Assuming astrong correlation between the thermodynamics and thehydrodynamics of the casting process, we derived newinsights from scanning electron microscopy images andthe experimental data. In view of optimization of the fluxthrough the membranes, a wetting pretreatment should beavoided. Important polymer savings were obtained with-out a loss of performance through a decrease in the cast-ing thickness in combination with the use of a verysmooth support. Last but not least, a high air humidityduring casting was found to inhibit the formation of a dense,defect-freeskinlayer.ForthePA layer, the interfacial poly-merization method, the drying method, and the curing timewere studied. The clamping of the membrane in a framewith one side in contact with the piperazine (PIP) solutionand the other side to the air yielded the highest membraneflux and rejection with the lowest use of PIP and trimesoyl-chloride solution. Because of the absence of a uniform PIPsolution layer for some drying methods, nodular PA struc-tures could be observed in the macrovoids of the underlyingPES layer because of hexane intrusion; this resulted in a dra-matic decrease in the flux. Moreover, the omission of thedrying step did not result in a significant loss of perform-ance and enhanced the ease of operation. Finally, a curingtime of 8 min was found to be optimal.
- Published
- 2012