Characterisation of PES hollow fibre membranes

• Main Author: Fang, Rongquan
• Other Authors: Guo Ningqun
• Format: Final Year Project
• Language: English
• Published: 2009
• Subjects: DRNTU::Engineering::Environmental engineering::Water treatment
• Online Access: http://hdl.handle.net/10356/17179
• Institution: Nanyang Technological University

Polyethersulfone (PES) hollow fibre membranes were fabricated using Dimethylacetate (DMAc) as a solvent, Diethylene glycol (DEG) as a non-solvent additive and water as a strong coagulant. The purpose of adding DEG was to increase the casting solution’s viscosity and thereby suppressing the formation of macrovoids. Three batches of fibres with a total of 16 samples with varying fabrication conditions were produced with the same PES/DMAc/DEG ratio (PES/DMAc/DEG ratio of 17/43/40 in wt %). PES hollow fibres were then produced using the dry-jet wet spun process. Studies were done on the effects of fabrication parameters (i.e. air gap length, bore fluid and casting solution pressure, bore fluid composition, bore fluid flow rate and take up speed) on the properties (i.e. the mechanical strength, selectivity and water flux) of the fibres. Results showed that an increase in air gap length and bore fluid flow rate results in a decrease in wall thickness of the fibres. For example with a constant bore fluid flow rate of 0.8 litres/hour and a decrease in air gap from 3cm to 0.5cm, the wall thickness increased from 0.5664mm to 0.5942mm for sample 4 and 5 respectively. The result also showed that take up speed may have an influence on the wall thickness. A higher take up speed may reduce the wall thickness as it stretches the fibre while it is still in the coagulation bath. Macrovoid formation depended largely on the viscosity of the casting solution. Generally an increase in viscosity will suppress the formation of macrovoids. This result is extensively shown in the literatures. However the results obtained from this report shows that there is evidence that bore fluid composition does have an effect on the likelihood of macrovoid formation. As shown in the results, an increase in water concentration in the bore fluid which causes a dense skin on the lumen, prevented macrovoids from forming. SEM images show that the inner surface of samples 1 to 6 did not have a dense skin layer that samples A to G had. Samples 1 to 6 developed macrovoids and samples A to G did not despite the fact that these two batches had very similar viscosity. It is suggested that the pores on the inner surface of samples 1 to 6 allowed the bore fluid to penetrate into the fibre matrix, which was still forming, and caused localized reservoirs of coagulation bath that formed the macrovoids. An interesting and unexpected result from the 1st and 3rd batch showed that the air gap length may have had an influence to the burst pressure of the fibres. It seems that there is a possibility of an ideal air gap length of 3cm for fibres to possess maximum burst pressure. However this paper could not confirm that an ideal air gap exists as the influence of air gap length on burst pressure could not be isolated in the samples produced. Further tests with carefully controlled fabrication parameters need to be done in order to confirm this finding.