Microsoft Word - 18patella.docx CHEMICAL ENGINEERING TRANSACTIONS VOL. 60, 2017 A publication of The Italian Association of Chemical Engineering Online at www.aidic.it/cet Guest Editors: Luca Di Palma, Elisabetta Petrucci, Marco Stoller Copyright © 2017, AIDIC Servizi S.r.l. ISBN 978-88-95608- 50-1; ISSN 2283-9216 Modification of Hollow Fibre Membranes through Electron Beam Irradiation Minh N. Nguyen, Nathalie Karpel Vel Leitner, Benoit Teychené Institut de Chimie des Milieux et des Matériaux de Poitiers (UMR CNRS 7285) École Nationale Supérieure d'Ingénieurs de Poitiers (ENSIP), Université de Poitiers. 1 rue Marcel Doré, Bâtiment 1, 86022 POITIERS CEDEX, France. benoit.teychene@univ-poitiers.fr The capacity for secondary effluent filtration of hollow fiber polypropylene membranes grafted with several zwitterionic molecules and multi-walled carbon nanotubes was evaluated. In particular, all the modifications result in significant improvement of filtration performances. The anti-fouling properties of the membrane had been increased without damaging the membrane as observed from FTIR, TGA/DTA analyses. The loading of MWCNT increased both the resistance of the membrane to organic matters fouling and permeate quality thank to the adsorption on carbon nanotubes. However, upon irradiation of PP-MWCNT membranes, a proportion of CNT was degraded and some of its adsorptive characteristics deteriorated, although the membrane’s filtration capability was still very impressive. Among the zwitterionic molecules, the best performers were the cysteine-modified membrane, followed by phosphocholine and DMAEMA. 1. Introduction The swift expansion of the global population and economies has led to the occurrence of water scarcity and water quality degradation in many parts of the world. In this context, the research and development of efficient water treatment technologies is needed more than ever to meet the world’s growing demand for fresh and clean water. Among the existing treatment technologies, membrane-based processes have become the most robust technique to treat polluted water. Indeed, thanks to their properties (various pore sizes, hydrophilicity…) membrane are able to simultaneously separate a wide variety of pollutants (bacteria, viruses and organic compounds) from many water sources without using toxic or harmful chemicals (Mulder, 1996) . However, membranes suffer from fouling, which refers to the accumulation of rejected matter at the membrane surface. Fouling strongly impacts treatment operating cost by increasing cleaning procedure frequency, and consequently reducing membrane life time (Cha et al. 2000; Chen & Fane, 2008). Many studies were dedicated to understand and mitigate membrane fouling. As intensively reported, a sustainable membrane for water treatment should have a hydrophilic surface in order to avoid the adsorption of foulant species (Liu et al. 2009; Ochoa, 2003; Zhao et al. 2013). The most common method to improve membrane hydrophilicity is to blend membrane polymer with polyvinylpyrrolidone (PVP). However, many works report on the leakage of PVP during membrane cleaning due to chlorine (Causserand et al. 2015) . Therefore, other routes to improve membrane hydrophilicity and fouling resistance are needed. The present study investigates the modification of commercial membranes by electron beam irradiation in order to improve filtration performances. Electron beam irradiation is a versatile technique to induce cross linking in polymers (Le Moel et al. 2986; Zhu & Xu, 2006; Nasef et al. 2006; Schulze et al. 2010 & 2013). Indeed, irradiation induced changes in polymer matrix due to radical-involved reactions, recombination without the need of strong chemicals for pre-treatment and consequently might be considered as a “green routes” to modify membranes materials. In this study, three zwitterionic molecules were grafted on the polypropylene hollow fibre membranes induced by electron beam (EB) irradiation. It had been reported that zwitterionic molecules could effectively alleviate membrane fouling (Lin et al. 2014; Zhao et al. 2013). In addition, the modification of hollow fibre with multi-walled carbon nanotubes (MWCNT) was tested and the impact of electron beam irradiation on the MWCNT-loaded membrane was investigated. MWCNT is known for its DOI: 10.3303/CET1760049 Please cite this article as: Nguyen M., Karpel Vel Leitner N., Teychene B., 2017, Modification of hollow fiber membranes through electron beam irradiation, Chemical Engineering Transactions, 60, 289-294 DOI: 10.3303/CET1760049 289 adsorptive properties towards organic foulants (Lu & Su, 2007; Pan & Xing, 2008; Wang et al. 2009), thus preventing cake formation at the membrane surface. 2. Materials and methods 2.1 Feed water characteristics and filtration procedure Filtration tests were carried out with a secondary effluent (SE) sampled from Saint Julien l’Ars wastewater treatment plant (June 2016, Poitiers, France). The effluent was prefiltered using a 200 µm sieve to remove coarse materials and stored in dark at 4 °C for a maximum of 10 days. All the experiments were conducted at room temperature (approximately 20 °C). The total organic carbon (TOC) content of the secondary effluent was approximately 7 mgC/L. Prior to every filtration test, the permeability of the pristine or modified hollow fibre membranes was evaluated with the flow step procedure by varying permeate flux from 1 to 4 mL/min using Milli-Q water (20 °C). Filtration tests were carried out using a fully automated lab-scale filtration unit using home-made membrane modules (consisting of 2 fibres 20 cm in length; 15 cm2 of surface area) according to the procedure described elsewhere (Touffet et al. 2015). The filtration was operated in dead end inside-out mode at 80 L/h/m2 (2 mL/min) for 45 minutes, interspersed by membrane backwash (outside-in mode using permeate solution) at 127 L/h/m2 (4 mL/min) for 2 minutes. The filtration tests duration was set for 24 h or until reaching a maximum pressure of 1.5 bars in order to avoid membrane bursting. The extent of fouling was characterized by the cake fouling resistance RC (m -1), which could be calculated using Eq(1). J= ∆Pμ20° C( Rm+Rc ) (1) In which J is the permeate flux (L/h/m2), ΔP is the transmembrane pressure (bar), µ(20°C) is the water viscosity (bar.s) and Rm and RC are the membrane hydraulic resistance and cake fouling resistance (m -1), respectively. 2.2 Membranes and modification procedures The hollow fibre membranes were purchased from ACCUREL®, (PP300/1200 Membrana, 3M GmbH). The membranes consisted of polypropylene fibre with an inner diameter of 1.8 mm, a wall thickness of 300 µm and a nominal pore size of 0.2 µm. According to the manufacturer, the permeability of the membrane module was 3060 L/h/bar and the bubble pressure was 1.17 bar. Membranes were modified using L-Cysteine, DMAEMA, phosphocholine and MWCNT followed by electron beam irradiation. The respective chemical structures of the investigated molecules are given in Table 1. The three graft molecules were purchased from Sigma (Aldrich, France) and the MWCNTs (O.D: 10−20 nm and 10−30 µm in length) were purchased from CheapTube (USA). Table 1: Properties of the investigated molecules Molecule Full product name Formula MW (Da) Log P (ACD predicted values / ChemSpider) Phosphocholine Phosphocholine chloride calcium salt tetrahydrate 183.143 −4.99 Cysteine L-Cysteine > 99.5% 121.158 0.23 DMAEMA 2-(Dimethylamino)ethyl methacrylate 157.210 1.50 For grafting with hydrophilic molecules, prior to irradiation, the membranes were filtrated with and submerged in the respective 10 mg/L graft molecule solutions (dissolved in Milli-Q) for 24 hours. The filtration with graft molecules was performed to enhance the interactions between the graft molecules and the inner surface of the membrane module. In contrast, membranes containing MWCNT (5 mg/m2) were produced according to the previous reported procedure (Gallagher et al. 2013). This methodology allows producing stable MWCNT mats on the membrane lumen. The pristine and EB membranes were submerged in Milli-Q water only. Electron beam irradiation was performed on a semi-industrial electron accelerator (3 MeV Van de Graaff, VIVIRAD, France). In the irradiation process, the selected membrane modules were concealed inside plastic bags filled with nitrogen gas. Subsequently, the bags were placed on a conveyor and delivered through the 290 electron beam source (2600 kV, 195 μA) at a speed of 0.6 cm/s. The corresponding irradiation dose was set at 10 kGy. It has to be noted that the polypropylene membrane cannot withstand an irradiation dose above 50 kGy (Makuuchi & Cheng, 2011). 2.3 Analytical tools The membrane surfaces were characterized by ATR-FTIR (Thermo Nicolet NEXUS, GMI Inc.) and X-ray microtomography (Viscom X8050-16, France). The thermal properties of the materials were investigated through Differential Thermogravimetric Analysis (TGA/DTA) (SDT Q600, TA Instruments, France). Organic matter rejection during filtration tests were analysed by Liquid Chromatography − Organic Carbon Detection (LC-OCD). The LC-OCD system contained a 0.5 μm size-exclusion chromatography (SEC) column (ReproSil 200, France). The mobile phase consisted of 20 mmol L−1 of Na2HPO4 at pH 6.8. The flow rate was set at 1 mL/min1. 3. Results and discussions 3.1 Characterization of the modified hollow fibre membranes As shown in TGA results (Figure 1-a), the degradation of the pristine PP membrane occurs between 200 °C and 400 °C. As expected, adding MWCNT onto the membrane lumen did not change the thermal degradation of the membrane, and results show that MWCNT thermal degradation occurred between 400 °C and 600 °C. Figure 1-b shows that a thin MWCNT layer was deposited on the membrane inner surface. The membranes grafted with zwitterionic molecules (PP-CYSTE is shown Figure 1-a) show similar TGA results as the pristine, demonstrating that little polymer deterioration was induced by the low irradiation dose used (10kGy). Figure 1: (a) TGA results obtained for the pristine, PP-CYSTE membrane and PP-MWCNT (5 mg/m2) membranes with and without irradiation. (b) 3D cross-section of hollow fibre obtained by X-ray microtomography of a PP-MWCNT (5 mg/m2) membrane dried at room temperature. Surface cracking comes from MWCNT layer shrinkage during drying. The FTIR spectra for the internal surface (Figure 2-a) of the irradiated module without molecule grafting confirmed the chemical changes in the PP, probably due to oxidation resulting in the formation of carbonyl or amide groups (Mirabedini et al. 2004). The doublet at 1631 and 1659 cm−1 might be related to the carbonyl groups. The O−H stretching vibrations were represented by two peaks at 3184 and 3358 cm−1 (data not shown). These peaks decreased in intensity seen for the modules in the presence zwitterionic molecules, suggesting that the presence of these molecules reduce the occurrence of functional groups formed by EB. However, as shown in Figure 2-a, the variation between membranes in the presence of different zwitterion molecules was not significant. In contrast, strong changes were observed on the external surface of irradiated HF membranes (figure 2-b). As indicated, four main changes were observed at 1409, 1425, 1631 and 1659 cm−1, respectively. These changes might be attributed to the formation of hydrophilic groups induced by irradiation and the grafting of the zwitterions. Interestingly, the presence of different graft molecules varied the intensity of these bands observed for the outer surface. The bands increase follows this order: EB-only, phosphocholine, cysteine and DMAEMA. According to Log P values reported in table 1, DMAEMA is the most hydrophobic compound; consequently, it is more susceptible to adsorption on the hydrophobic polypropylene membrane before irradiation, thus the band intensity is highest because the grafting was most efficient with DMAEMA. 291 Figure 2: ATR-FTIR spectra obtained for the modified membranes on the internal surface (a) and external surface (b). 3.2 Filtration performances of modified membranes As shown, in Figure 3-a&b, the transmembrane pressure strongly increases during the filtration of SE on pristine PP membranes. As reported, the maximum pressure equals to 1.5 bars was reached in only 320 minutes corresponding to a filtered volume of 274 L m−2 (7 cycles). These results demonstrate that membrane backwash are not enough efficient to remove fouling and irreversible fouling layer quickly accumulate in the membrane lumen (Nasef et al. 2006, Ye et al. 2005) . Adding MWCNT inside the membrane lumen strongly reduce the TMP increase and allow longer filtration time. Indeed, the maximum pressure allowed by the filtration system was reached after nearly 861 L m−2 of filtration corresponding to 22 cycles. As previously reported, MWCNT might protect the membrane surface from pore blockage by adsorption phenomena and might also structure fouling layer to avoid highly compressible organic matter layer formation at higher pressure (Gallagher et al. 2013, Wang et al. 2009). Irradiation of the pristine membrane has also a beneficial effect of TMP increase. In early stage of filtration, the TMP quickly increases and then levelled off. The maximum TMP was observed after 18 cycles (704 L m−2) indicating that the slight change observed on the irradiated membrane has a strong impact on membrane resistance to fouling. This might be attributed to a possible change in materials hydrophilicity after electron beam irradiation, due to the formation of hydrophilic functional groups (C=O, O-H, COOH) on membrane surface. Irradiation on the PP-MWCNT membrane has further positive effect in the fouling resistance of the membrane (25 cycles of filtration). Similarly, modification of the membrane with zwitterion molecules shows significant improvement in membrane filtration performances. The best performers were the cysteine-modified membranes with an average filtration volume of 821 L m−2, followed by phosphocholine (704 L m−2) and DMAEMA (626 L m−2). Figure 3: Transmembrane pressure (TMP) increase during filtration of SE through the pristine and modified hollow fibre membranes - (a) with MWCNT with and without irradiation at 10 kGy, and (b) with zwitterionic molecules. Inside-out filtration in dead end mode at a constant flux of 52 L/h/m2). 292 Figure 4: LC-OCD analysis of the SE feed waters, permeate obtained on pristine membrane and cysteine grafted membrane filtration. Despite the slight change in internal surface modification revealed by ATR-FTIR analysis, hollow fibre membranes undergone electron beam irradiation have a strong resistance to cake fouling during SE filtration. According to Table 1, cysteine has a Log P value close to zero, thus the best performing membranes might be grafted with neither hydrophilic nor hydrophobic molecules. In addition to TMP increase the quality of the permeate solutions was also monitored using LC-OCD analysis. An example of LC-OCD chromatogram was provided in Figure 4 and all results obtained by TOC peak integration for macromolecules were reported in Table 2. As seen, the SE organic matters consist of macromolecular (HMW) compounds with an elution time of around 6 min and low molecular weight (LMW) compounds exhibiting an elution time of about 14 min. As shown, the pristine membrane efficiently removed the macromolecular compounds but has no impact on LMW compounds content. In contrast, PP-Cysteine membrane slightly rejected the HMW compounds but better rejected the LMW. The rejection of HMW compounds obtained on all investigated membranes is reported in Table2. As shown, the lowest rejection was obtained for the PP-DMAEMA membranes with only 15 % of rejection. The highest was obtained for the MWCNT w/o EB membranes with 81 % of rejection. This high value is attributed to the strong adsorption capacity of MWCNT deposited on the membrane. Interestingly, the irradiation MWCNT exhibited similar rejection as the pristine membrane, indicating that irradiation strongly decreased the adsorption capability of CNT. Table 2: Macromolecule rejection efficiency obtained from TOC peak integration Molecule Pristine PP- WATER PP- PCHOL PP- DMAEMA PP- CYSTE MWCNT w/o EB MWCNT with EB Efficiency (%) 40.6 23.0 30.2 15.2 36.5 81.1 40.5 Despite their strongly improved resistance to fouling, PP-Cysteine membranes showed similar rejection compared to the pristine membranes (Table 2). This results suggest that hollow fiber irradiation in presence of Cysteine molecules led to the development of an efficient membrane material. 4. Conclusions The hollow fiber PP modules were successfully modified in the presence of zwitterionic graft molecules and MWCNT. All the modifications result in significant improvement at secondary effluent filtration performance. At low irradiation dose (10 kGy), the anti-fouling properties of the membrane had been increased without damaging the membrane, as seen from FTIR, TGA/DTA analysis and TC measurements for macromolecules. Interestingly, the changes in FTIR spectra were correlated to the Log P values of the investigated zwitterionic molecules. The best performer among the investigated zwitterionic molecules was the Cysteine compounds leading to strong fouling reduction and to macromolecular rejection upholding. The formation of MWCNT mats at the membrane lumen increased both the membrane fouling resistance and permeate quality thank to the adsorption of carbon nanotubes. This unique geometry allow to stabilize the CNT layer indicating the membrane-MWCNT could be backwashed without losing the MWCNT benefit effect. However, upon irradiation of MWCNT-HF membranes, a proportion of CNT was degraded and some of its adsorptive characteristics deteriorated, although the membrane’s filtration capability was still very impressive. 293 Reference Causserand, C., Pellegrin, B., & Rouch, J.-C., 2015, Effects of sodium hypochlorite exposure mode on PES/PVP ultrafiltration membrane degradation. Water Research, 85, 316–326. Cho, J., Amy, G., & Pellegrino, J., 2000, Membrane filtration of natural organic matter: Factors and mechanisms affecting rejection and flux decline with charged ultrafiltration (UF) membrane. Journal of Membrane Science, 164(1–2), 89–110. Gallagher, M. J., Huang, H., Schwab, K. J., Fairbrother, D. H., & Teychene, B., 2013, Generating backwashable carbon nanotube mats on the inner surface of polymeric hollow fiber membranes. Journal of Membrane Science, 446, 59–67. Le Moel, A., Duraud, J. P., & Balanzat, E., 1986, Modifications of polyvinylidene fluoride (PVDF) under high energy heavy ion, X-ray and electron irradiation studied by X-ray photoelectron spectroscopy. Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms, 18(1–6), 59–63. Lee, E. K., Chen, V., & Fane, A. G., 2008, Natural organic matter (NOM) fouling in low pressure membrane filtration — effect of membranes and operation modes. Desalination, 218(1–3), 257–270. Lin, P., Ding, L., Lin, C.-W., & Gu, F., 2014, Nonfouling Property of Zwitterionic Cysteine Surface. Langmuir, 30(22), 6497–6507. Liu, F., Xu, Y.-Y., Zhu, B.-K., Zhang, F., & Zhu, L.-P., 2009, Preparation of hydrophilic and fouling resistant poly(vinylidene fluoride) hollow fiber membranes. Journal of Membrane Science, 345(1–2), 331–339. Liu, F., Zhu, B.-K., & Xu, Y.-Y., 2006, Improving the hydrophilicity of poly(vinylidene fluoride) porous membranes by electron beam initiated surface grafting of AA/SSS binary monomers. Applied Surface Science, 253(4), 2096–2101. Lu, C., Su, F., 2007, Adsorption of natural organic matter by carbon nanotubes. Separation and Purification Technology, 58(1), 113–121. Makuuchi, K., Cheng, S., 2011, Radiation processing of polymer materials and its industrial applications. Hoboken, N.J: Wiley. Mirabedini, S. M., Rahimi, H., Hamedifar, S., & Mohsen Mohseni, S., 2004, Microwave irradiation of polypropylene surface: a study on wettability and adhesion. International Journal of Adhesion and Adhesives, 24(2), 163–170. Mulder, M., 1996, Basic Principles of Membrane Technology. Springer Netherlands. Nasef, M. M., Zubir, N. A., Ismail, A. F., Khayet, M., Dahlan, K. Z. M., Saidi, H., Sulaiman, N. A., 2006,. PSSA pore-filled PVDF membranes by simultaneous electron beam irradiation: Preparation and transport characteristics of protons and methanol. Journal of Membrane Science, 268(1), 96–108. Ochoa, N., 2003, Effect of hydrophilicity on fouling of an emulsified oil wastewater with PVDF/PMMA membranes. Journal of Membrane Science, 226(1–2), 203–211. Pan, B., Xing, B., 2008, Adsorption Mechanisms of Organic Chemicals on Carbon Nanotubes. Environmental Science & Technology, 42(24), 9005–9013. Schulze, A., Maitz, M. F., Zimmermann, R., Marquardt, B., Fischer, M., Werner, C.,Thomas, I., 2013, Permanent surface modification by electron-beam-induced grafting of hydrophilic polymers to PVDF membranes. RSC Advances, 3(44), 22518. Schulze, A., Marquardt, B., Kaczmarek, S., Schubert, R., Prager, A., & Buchmeiser, M. R., 2010, Electron Beam-Based Functionalization of Poly(ethersulfone) Membranes. Macromolecular Rapid Communications, 31(5), 467–472. Touffet, A., Baron, J., Welte, B., Joyeux, M., Teychene, B., & Gallard, H., 2015, Impact of pretreatment conditions and chemical ageing on ultrafiltration membrane performances. Diagnostic of a coagulation/adsorption/filtration process. Journal of Membrane Science, 489, 284–291. Wang, X., Tao, S., & Xing, B., 2009, Sorption and Competition of Aromatic Compounds and Humic Acid on Multiwalled Carbon Nanotubes. Environmental Science & Technology, 43(16), 6214–6219. Zhao, Y.-F., Zhu, L.-P., Yi, Z., Zhu, B.-K., & Xu, Y.-Y., 2013, Improving the hydrophilicity and fouling- resistance of polysulfone ultrafiltration membranes via surface zwitterionicalization mediated by polysulfone-based triblock copolymer additive. Journal of Membrane Science, 440, 40–47. 294