Nanoporous pvdf based membranes for lonic dye removal from aqueous solutions and filtration machanisms therein

長庚大學化工與材料工程學系 博士論文 Department of Chemical and Materials Engineering Chang Gung University Doctoral Dissertation 具奈米微孔的聚偏二氟乙烯及其複合薄膜用於離子性染料的 去除及其分離機制的探討 Nanoporous PVDF-based Membranes for Ionic Dye Removal from Aqueous Solutions and Filtration Mechanisms Therein 指導教授：呂幸江 博士 Advisor: Shing-jiang Jessie Lue, Ph.D 研究生：鄭氏德芳 Graduate Student: Tran Thi Tuong Van 中華民國 108 年 12 月 December 2019 Acknowledgements I would like to extend acknowledgements to many people who so generously assisted me during my Ph.D studying time First and foremost, truly deep gratitude goes to Professor Shingjiang Jessie Lue, my dedicated and expert advisor Since my first day in graduate school, she gave me continuing academic support and offered me so many wonderful opportunities Under her guidance, I learned how to define a research problem, conduct scientific research and finally publish the result I also benefited a lot from joining in various related works of the laboratory In addition, participating in conferences brought me eye-opening and fascinating experiences Special thanks go to Professors of Chemical and Materials Engineering Department who provided me plenty of precious knowledge and warm encouragement I am very appreciative to nice and helpful officers at every section in Chang Gung University Sincere thanks go to willing generations of labmates in Microcontamination & Membrane Separation Laboratory who enthusiastically helped me in lab works and many other things In particular, deep thanks come to Doctor S Rajesh Kumar who supplied me invaluable advices during my writing process I am so thankful to my Taiwanese and international friends for sharing with and supporting me during classes, experiments and life Thank you all for turning my time in CGU into unforgettable memories ever Sincere gratitude goes to my managers and colleagues at Institute of Environmental Science, Engineering and Management, Industrial University of Ho Chi Minh City in Vietnam who facilitated my Ph.D study and sent me many kind encouragements Last but not least, I would like to wholeheartedly thank my dear family and parents for endless love and support This thesis is dedicated to them -iii- 摘要 本文探討了各種有機染料在聚偏氟乙烯（PVDF）超濾膜上的過濾行為，發現 Donnan 排斥機制在染料分離中為主要分離機制。PVDF 膜對兩性離子若丹明 B（RhB，1.1%）的截留率極低，而對陽離子亞甲基藍（MB，45.6%）的截留 率中等；但對陰離子染料，鉻黑 T（EBT，83.5%）和萘酚藍的截留率高。黑 色（NBB，89.1%）。兩性離子或相同電荷的混合染料溶液（RhB-MB，RhBNBB 和 EBT-NBB）顯示出的染料去除率與其成分的單一染料溶液相似。相 反，由於分子間的相互作用，帶相反電荷的染料（MB-NBB）的混合溶液形 成微米級的聚集體，導致染料幾乎完全被去除、並提高了滲透率。在多次循 環過濾過程中證明了膜過濾性能的穩定性。接下來，本文研究了氧化石墨烯 （GO）和聚乙烯基吡咯烷酮（PVP）添加劑在 PVDF 混成膜的個別和加成作 用。結果顯示 PVP 的主要作用是增強膜的孔隙率，而 GO 的顯著貢獻是降低 了膜表面電荷。含有 GO 和 PVP 的 PVDF 膜在膜孔隙率，水接觸角和 Zeta 電 位特性方面表現出累加效應，並使 PVDF 的結晶型態由轉變至較親水的 型。在 pH 值為 4-10 的純水和染料溶液（NBB 和 MB-NBB）中，濾液的滲透 率從純 PVDF（P），GO 混合 PVDF（PG），PVP 混合 PVDF（PP）和 GOPVP 混合 PVDF（PPG）膜逐漸增加。其中，GO 和 PVP 的引入分別使滲透率 增加了幾倍和幾十倍。由於 Donnan 排斥機理，所有膜均表現出高於 80%的染 料排斥率，與微弱的染料吸附能力。由於 GO 和 PVP 的累加作用，PPG 膜具 有出色的高滲透性（953.0–1353.0 L m-2 h-1 MPa-1）和傑出的脫色率（83.2– 91.4%），本研究證實這種複合 PVDF 膜在實際染料廢水處理應用中開了一條 可行的路。 關鍵字：離子染料；聚偏二氟乙烯；氧化石墨烯；聚乙烯吡咯烷酮；Donnan 排除；分子間相互作用；加乘作用；膜通透性 -iv- Abstract In this thesis, filtration behaviors of various organic dyes across a poly(vinylidene fluoride) (PVDF) ultrafiltration membrane were explored The Donnan exclusion mechanism played a major role in dye separation The PVDF membrane had extremely low rejection for zwitterionic rhodamine B (RhB, 1.1%) and medium rejection for cationic methylene blue (MB, 45.6%), but high rejections for anionic dyes, eriochrome black T (EBT, 83.5%) and naphthol blue black (NBB, 89.1%) The mixed zwitterionic or same-charged dye solutions (RhB-MB, RhB-NBB and EBT-NBB) exhibited dye removal rates similar to their constituents’ single dye solutions Conversely, the mixed solution of opposite-charged dyes (MB-NBB) formed micron-sized aggregates due to intermolecular interaction, leading to almost complete dye removal and enhanced permeance value The stable membrane filtration performance was demonstrated during multi-cycle filtration Next, individual and simultaneous effects of graphene oxide (GO) and poly(vinyl pyrrolidone) (PVP) additives on PVDF based membranes were investigated It was demonstrated that the principal role of PVP was to enhance membrane porosity and -to- phase transformation The noticeable contribution of GO was to lower membrane surface charge PVDF membrane containing GO and PVP showed additive effects in membrane porosity, polar crystallization, water contact angle and zeta potential properties In both pure water and dye solutions (NBB and MB-NBB) at pHs 4-10, the filtrate permeance increased from the pure PVDF (P) to GO blended PVDF (PG), PVP blended PVDF (PP) and GO-PVP blended PVDF (PPG) membranes Therein, GO and PVP introduction provided permeance increases of several times and several tens of times, respectively All membranes exhibited high dye rejections of  80.0% due to the Donnan exclusion mechanism The membranes showed minimal dye adsorption capacities Owing to the additive effects of GO and PVP, the PPG membrane achieved outstanding high permeance (953.0 – 1353.0 L m-2 h-1 MPa-1) and excellent dye removal efficiencies (83.2 – 91.4%), which opened an avenue for application of this composite PVDF membrane to actual dye wastewater treatment Keywords: Ionic dyes; Poly(vinylidene fluoride); Graphene oxide; Poly(vinyl pyrrolidone); Donnan exclusion; Intermolecular interaction; Additive effect; Membrane permeance -v- Table of Contents Recommendation Letter from the Thesis Advisor Dissertation Oral Defense Committee Certification Acknowledgements iii 摘要 iv Abstract v Table of Contents vi List of Figures ix List of Tables xiii Chapter Literature Review 1.1 Dye removal from wastewater 1.1.1 Dye structure and classification 1.1.2 Wastewater containing dyes 1.1.3 Treatment of dye wastewater 1.2 Membrane processes 11 1.2.1 Introduction to membrane processes 11 1.2.2 Membrane modifications 13 1.2.3 Pressure-driven membrane processes for dye removal 16 1.2.4 PVDF membranes for dye removal and their modifications 18 1.3 Rationale of this research 20 1.4 Objectives and scopes of this research 22 Chapter Materials and Methods 25 2.1 Materials 25 2.1.1 Commercial materials 25 2.1.2 GO preparation 28 -vi- 2.2 Membrane preparation 28 2.3 Characterization 29 2.3.1 GO and PVP characterization 29 2.3.2 Membrane characterization 30 2.3.3 Dye characterization and dye concentration evaluation 31 2.4 Membrane performance experiment 32 2.4.1 Membrane filtration system 32 2.4.2 Filtration performance evaluation 34 2.4.3 Static adsorption 34 2.4.4 Multi-cycle filtration 35 Chapter Filtration Mechanisms of Binary Dye Mixtures Using a PVDF Membrane 36 3.1 Membrane characterization 36 3.2 Membrane filtration performance for single dyes 39 3.3 Membrane filtration performance for binary dye mixtures 43 3.3.1 RhB-MB mixture 43 3.3.2 EBT-NBB mixture 46 3.3.3 RhB-NBB mixture 46 3.3.4 MB-NBB mixture 47 3.4 Membrane adsorption performance 50 3.5 Stability of membrane performance 50 Chapter GO and/or PVP Blended PVDF Membranes for Enhanced Dye Filtration 55 4.1 GO and PVP characteristics 55 4.2 Membrane morphology and pore size distribution 58 4.3 Membrane chemical composition 65 4.4 Membrane hydrophilicity and surface charge 72 4.5 Membrane filtration performance 73 -vii- 4.5.1 Pure water 73 4.5.2 NBB dye 76 4.5.3 MB-NBB dye mixture 83 4.6 Membrane adsorption performance 84 Chapter Conclusions and Future Recommendations 89 5.1 Conclusions 89 5.2 Future recommendations 90 References 92 Appendix A Dye concentration evaluation 112 Appendix B Curriculum Vitae 116 -viii- List of Figures Fig 1-1 Schematic representation of (a) a membrane process and (b) a two-phase system separated by a membrane [59] 11 Fig 1-2 Fabrication of composite membranes through phase inversion process and main effects of nanomaterials on final products [62] 16 Fig 1-3 Pressure-driven membrane processes classified by pore size, target species and operating pressure [61] 17 Fig 1- Investigated membrane and dye subjects in this thesis 24 Fig 2-1 (a) Photograph and (b) schematic diagram of the membrane filtration unit 33 Fig 3-1 FESEM micrographs of the PVDF membrane showing (a, c) differently magnified top surfaces, (b, d, e) overall, enlarged upper and lower cross-sections 37 Fig 3-2 (a) Pore size distribution, (b) FTIR spectrum, (c) surface zeta potential over a pH range and (d) water contact angles at different drop ages of the PVDF membrane 39 Fig 3-3 (a) Dye rejection, TOC rejection and (b) permeance performance of the PVDF membrane with single dyes 39 Fig 3-4 Dye rejection, TOC rejection and permeance performance of the PVDF membrane with binary dye mixtures of (a) RhB-MB, (b) EBT-NBB, (c) RhB-NBB and (d) MB-NBB 44 Fig 3-5 Absorbance spectra of feed and permeate after 120 of filtration with binary dye mixtures of (a) RhB-MB, (b) EBT-NBB, (c) RhB-NBB and (d) MB-NBB 45 Fig 3-6 Proposed separation mechanism for MB-NBB mixture on the PVDF membrane 49 Fig 3-7 Photographs of the pristine and adsorbed PVDF membranes with dyes 51 Fig 3-8 (a) Dye rejection and (b) permeance performance in multi-cycle filtration process for single dyes on the PVDF membrane 52 -ix- membrane characteristics, such as the surface charge and filtration behavior towards charged solutes Consequently, more investigation is needed for a comprehensive understanding of GO and PVP roles in blended membranes 1.4 Objectives and scopes of this research The aim of this research is to evaluate and develop nanoporous PVDF-based UF membranes for ionic dye removal To achieve this, the thesis focus on two objectives, namely exploring filtration behaviors of binary dye mixtures across a PVDF membrane and clarifying roles of GO and PVP additives in enhancing PVDF-based membrane characterization and performance Accordingly, the scope of research activities includes as follows: Filtration behavior of binary dye mixtures across a PVDF membrane - Investigate membrane filtration performance for four single dyes of different charges, including rhodamine B (RhB), methylene blue (MB), eriochrome black T (EBT) and naphthol blue black (NBB) - Explore membrane filtration performance for various binary dye solutions and compare to those of the single dye solutions - Propose dye filtration mechanisms - Perform static adsorption and multi-cycle operation to validate membrane reusability GO and PVP additives for enhancing PVDF-based membrane characterization and performance - Investigate individual and simultaneous effects of GO and PVP additives on PVDF membrane characteristics and dye filtration performance - Examine and discuss changes in morphological structure, chemical composition, hydrophilicity, surface charge and pure water permeance between the pristine and blended PVDF membranes - Test and compare filtrate permeance and rejection efficiency of the composite membranes for single dye (NBB) and dye mixture (MB-NBB) -22- - Explore the influence of pH on the dye filtration process - Perform static adsorption experiment on the composite membranes - Propose additive interaction between GO and PVP in the blended membrane The thesis structure is divided into five chapters Chapter provides a comprehensive literature background of the research problem, including dye removal from wastewater and membrane processes especially using PVDF-based membranes The rationale, objectives and scopes of the research are also included Chapter presents details about starting materials, membrane preparation method, characterization techniques, filtration unit and experimental descriptions Chapter discusses the filtration behaviors of binary dye mixtures using a pure PVDF UF membrane Filtration performance of dye mixed solutions are compared to those of single dye solutions and filtration mechanisms are proposed Membrane stability is studied over multi-cycle filtration Chapter deals with PVDF membrane modifications by GO and/or PVP additives Variations in membrane characteristics, and dye filtration and adsorption performance are examined and compared to the pure PVDF membrane Additive interactions between GO and PVP in the composite PVDF membrane are proposed Investigated membrane and dye subjects in Chapters and are briefly summarized in Fig 1-4 Chapter draws major conclusions from experimental findings and propose future recommendations for further investigation in this area of research -23- Fig 1- Investigated membrane and dye subjects in this thesis -24- Chapter Materials and Methods 2.1 Materials 2.1.1 Commercial materials Graphite powder (99%, < 20µm), sulfuric acid (H2SO4, 95.0-98.0%), phosphoric acid (H3PO4, 85%), hydrochloric acid (HCl, 37 %), sodium hydroxide (NaOH, 99.8%), poly(vinylidene fluoride) (PVDF, average Mw ~534,000 by gel permeation chromatography), polyvinylpyrrolidone (PVP, average Mw ~40,000), 1-methyl-2-pyrrolidinone (NMP, 99.9%), rhodamine B (RhB, ≥95%, HPLC grade), methylene blue (MB, ≥95%) and naphthol blue black (NBB, BioReagent) were purchased from Sigma-Aldrich, St Louis, MO, USA Eriochrome black T (EBT, Reagent European Pharmacopoeia) was bought from Fisher Scientific GmbH, Schwerte, Germany Detailed characteristics regarding the polymers (PVDF and PVP), solvent (NMP) and dyes (RhB, MB, EBT and NBB) are shown in Table 2-1 and 2-2, respectively Potassium permanganate (KMnO4, 99%) was acquired from Nihon Shiyaku Industries Ltd., Osaka, Japan Hydrogen peroxide (H2O2, 35.0%) was obtained from Showa Chemical Co., Tokyo, Japan Nonwoven fabric (grade 3706, made from polyethylene terephthalate) was supplied from Ahlstrom, Helsinki, Finland Deionized (DI) water was produced using a double distiller -25- Table 2-1 Characteristics of experimental polymers and solvent Chemical Molecular Molecular Density Glass transition Melting point Boiling point Solubility formula weight (at 25 C) temperature (Tg) (Tm) (Tb) in water PVDF (CH2CF2)n ~534,000 Da 1.7 g cm-3 −38 C 171 C – Insoluble PVP (C6H9NO)n ~40,000 Da 1.2 g cm-3 67-86 C 158-169 C – Solublea [107] [107] – -24 C 202 C Soluble NMP a C5H9NO 99.1 Da 1.0 g cm-3 Molecular structure PVP is readily soluble in cold water and it is possible to prepare PVP free-flowing solutions of concentrations up to 60% [108] -26- Table 2-2 Characteristics of experimental dyes Dye RhB Molecular Molecular Molecular formula weight size C28H31ClN2O3 479.0 Da 18.0Å Class Charge Acidity constant, pHa pKa Zwitterionic – 3.7 [110] 4.4 Cationic +1 3.8 [112] 5.5 Anionic -1 6.2 [114] 5.7 Anionic -2 4.5, 8.3b [96] 6.2 [109] MB EBT C16H18ClN3SxH2O C20H12N3O7SNa 319.9 Da 14.5 Å (anhy) [111] 461.4 Da 15.5 Å [113] NBB C22H14N6Na2O9S2 616.5 Da 21.0 Å [115] a Natural pH value of 20 mg L-1dye solution (in DI water) b pKa1 and pKa2 values referred from a synonym Acid Black -27- Molecular structure 2.1.2 GO preparation GO sheets were synthesized using a modified Hummers’ method [116, 117] A preblended mixture of concentrated H2SO4/H3PO4 (9:1) was added to a mixture of graphite/KMnO4 (1:6), producing a slight exothermic reaction The resultant suspension was magnetically stirred at 50 C for 12 h After the reaction completion, the suspension was cooled to ambient temperature and poured slowly into DI water ice The resulting suspension was then added with H2O2 to remove residual KMnO4 and turned brilliant yellow, showing a high oxidation level of graphite The final suspension was left to stand for 12 h; the supernatant was then separated from the precipitate Afterwards, the precipitate was washed three times with HCl and repeatedly with DI water until the supernatant pH became neutral For each wash, the precipitate was dispersed, centrifuged at 10,000 rpm and the supernatant was decanted During the washing process, the graphite oxide underwent exfoliation, resulting in the thickening of the GO solution and ultimately forming a GO gel Finally, the GO gel was vacuum-dried at 60 C for 24 h to obtain GO solid 2.2 Membrane preparation Four experimental membranes  pure PVDF (P), GO blended PVDF (PG), PVP blended PVDF (PP) and PVDF containing PVP and GO (PPG)  were fabricated using a phase inversion technique Their casting solution components are presented in Table 2-3 First, 0.06 g of GO (0.2 wt.%) was dispersed in NMP (24 g) solvent The GO solution was then ultra-sonicated in an ultrasonic bath for h to achieve complete GO exfoliation Next, 0.6 g of PVP (2 wt.%) and g of PVDF (20 wt.%) were imported into the prepared GO solution The mixed solution was stirred at 60 °C for 12 h until homogeneity NMP was chosen as the solvent for its good thermodynamic stability and fast solid-liquid demixing in water coagulant bath [118] The PVDF concentration of 20 wt% in NMP was selected to ensure a well-dissolved dope solution [119, 120] and porous structure formation with small pore diameter in the ultimate membrane -28- [119, 121] Later, the solution was left to stand until air bubbles were removed Afterwards, the solution was cast onto a nonwoven fabric using a casting knife with a 200-µm gap at a 15-mm s-1 speed under ambient conditions The nonwoven fabric was employed as a supporting substrate to prevent shrinkage of the PVDF based functional layer and to increase mechanical strength of the resultant membrane [122, 123] The cast membrane was immediately submerged into a non-solvent (DI water) bath at room temperature After complete phase inversion, the formed membrane was transferred into a new DI water bath and kept overnight to remove residual solvent and additives Finally, the membrane was vacuum-dried at 60 °C overnight before further characterization and filtration tests For the GO blended, PVP blended and pure PVDF membranes, similar preparation protocols were performed but without GO or/and PVP addition Table 2-3 Casting solution components of the prepared membranes, g (mass ratio in wt.%) Membrane NMP PVDF GO PVP P 24 (80) (20) - - PG 24 (80) (20) 0.06 (0.2) - PP 24 (80) (20) - 0.6 (2) PPG 24 (80) (20) 0.06 (0.2) 0.6 (2) 2.3 Characterization 2.3.1 GO and PVP characterization The synthesized GO was characterized by analysis techniques including electron microscopy, Raman spectroscopy, X-ray diffraction (XRD), etc The morphology and elemental composition of the GO were investigated using a field emission scanning electron microscope equipped with energy dispersive X-ray spectrometer (FESEM/EDX, SU8220, Hitachi, Tokyo, Japan) The morphology of the GO was further observed under a transmission electron microscope (TEM, JEM1230, JEOL, Tokyo, Japan) The Raman spectrum of the GO was -29- collected by a Raman spectrometer (UniDRON, UniNanoTech Co Ltd, Gyeonggi-Do, Korea) The XRD pattern of the GO was recorded using an X-ray diffractometer (D2 phaser, Bruker, Karlsruhe, Germany) The X-ray photoelectron spectroscopy (XPS) spectra of the GO and PVP were obtained by an X-ray photoelectron spectrometer (VG Microtech MT-500, Thermo Fisher Scientific Inc., Waltham, MA, USA) The functional groups of the GO and PVP were identified using a Fourier transform infrared spectrometer (FTIR, Bruker Tensor 27, Bruker, Karlsruhe, Germany) The zeta potential of the GO and PVP was determined by a zeta potential and particle size analyzer (Zetasizer Nano-ZS, Malvern, Worcestershire, UK) 2.3.2 Membrane characterization The microstructure and elemental composition of the membranes were investigated using FESEM/EDX analysis The thickness of the membranes was determined using a thickness gauge (Elcometer model 345, Elcometer, Manchester, UK) The volume porosity of the membranes was measured using the gravimetric method [123], with Galwick (surface tension of 15.9 dyn cm-1) as a wetting liquid The surface area (Brunauer, Emmett and Teller (BET) method) and pore volume (Barrett-Joyner-Halena (BJH) method) of the membranes were evaluated by N2 gas adsorption/desorption isotherm measurement at -196 C on an accelerated surface area and porosimetry system (ASAP2020, Micromeritics, Norcross, GA, USA) The pore size distribution and mean flow pore size of the membranes were measured by capillary flow porometry (CFP, CFP-1500AE, PMI, Ithaca, NY, USA) The surface chemical composition and functionality of the membranes were obtained using XPS, FTIR and ATR (Attenuated total reflectance)-FTIR analyses, respectively FTIR analysis was performed on (thin) PVDF-based films – membrane without nonwoven support, and ATR-FTIR analysis was done for PVDFbased membrane surfaces – membrane including nonwoven support The water contact angle of the membranes was recorded by a drop shape analysis system (DSA, G10-MK2, Kruss GmbH, Hamburg, Germany) The surface charge of the membranes at various pH values was determined -30- by a streaming current electrokinetic analyzer (SurPASS, Anton Parr GmbH, Graz, Austria) equipped with an adjustable gap cell 2.3.3 Dye characterization and dye concentration evaluation The zeta potential and particle size analyzer was utilized to identify the zeta potential of the dye solutions pH measurement was performed using a pH meter (PH500, Clean Instruments, New Taipei, Taiwan) Two quantitative analytical methods were applied to evaluate the concentration of dyes in aqueous solutions The first technique was based on dye absorption of visible light measured through a UV-Vis spectrophotometer (V650, Jasco, Tokyo, Japan) Afterwards, the unknown dye concentration (in mg L-1) was calculated from a previously determined calibration curve of concentration versus absorbance or first order (1st) derivative Second, an additional analytical parameter expressing the dye concentration was chosen as total organic carbon (TOC, in mg C L-1) The TOC content was identified by using a TOC analyzer (Torch, Teledyne Tekmar, Mason, OH, USA) employing oxidation of carbonaceous materials in a high-temperature combustion furnace followed by nondispersive infrared detection of the CO2 product Hereafter, simply “dye concentration” in mg L-1 and “TOC concentration” in mg C L-1 are used when referring to the results attained from these two examination methods The spectrophotometric absorbance of each individual dye solution (RhB, MB, EBT or NBB) was obtained at its corresponding maximum wavelength (λmax) of 553 nm, 654 nm, 534 nm and 618 nm, respectively (Fig A-1) Furthermore, Fig A-2 represents the four calibration curves of these single dye solutions established at their λmax Meanwhile, the concentration determination of each dye in its binary system was undertaken by 1st derivative spectrophotometry in light of the considerable spectral overlap among the two single dyes and their mixture (Figs A-3a, A-4a and A-5a) In this thesis, the data of a zero-order derivative spectrum obtained by scanning from 780 nm to 380 nm with a band width of nm at a scan -31- speed of 400 nm min-1 was differentiated to give a first-order derivative spectrum using a sevenpoint data filter and third-order Savitzky-Golay polynomial The spectra of the zero or firstorder derivative values of the two single dyes together with their mixture versus wavelength (Figs A-3, A-4, A-5 and A-6) demonstrated the best wavelength for the accurate analysis of each separate dye in its mix, which was selected using a zero-crossing technique [124] In detail, the RhB and MB contents in their binary solution were examined by 1st derivative spectrometry at 505 nm and 682 nm, respectively Similarly, EBT and NBB were examined by 1st derivative at 474 nm and 537 nm, RhB and NBB by 1st order derivative at 474 nm and 645 nm, and MB and NBB by zero derivative at 695 nm and 419 nm, respectively Figs A-7, A-8, A-9 and A-10 show the calibration graphs of each component in these four binary dye systems built by corresponding derivative values at as-determined wavelengths The dye concentration of a mixture was taken by adding the as-determined dye concentration of its two components together For TOC measurements, only the overall TOC concentration was obtained for either single dye solutions or binary dye mixtures 2.4 Membrane performance experiment 2.4.1 Membrane filtration system The membrane filtration experiments were performed through a lab-scale cross-flow filtration unit Photograph and schematic diagram of the membrane filtration system are presented in Fig 2-1 The feed water was held in a 1-L reservoir and pumped into the membrane cell with an effective area of 6.16 cm2 The operational conditions were maintained at 0.49 MPa transmembrane pressure and 0.30 Lmin-1 feed flow rate at room temperature (25 ± C) A new membrane sample was employed for each test and fixed into the membrane cell before operating the filtration system The feed water and permeate samples were collected every 30 during the experimental duration of 120 The data was collected at a filtration time of 120 to ensure steady-state operation All the filtration tests were repeated at least twice to -32- ensure reproducibility, and the average values were reported (a) (b) Fig 2-1 (a) Photograph and (b) schematic diagram of the membrane filtration unit In the first research content of this thesis on dye filtration mechanisms of a pure PVDF membrane, four different zwitterionic (RhB), cationic (MB) and anionic (EBT, NBB) dyes were investigated In the first stage, the filtration performance of the PVDF membrane for the four single dyes was tested In the next stage, the two-component dye solutions were supplied to the filtration module in order to investigate membrane performance In detail, four different mixtures of zwitterionic dye (RhB-MB and RhB-NBB), co-charged dyes (EBT-NBB) and counter-charged dyes (MB-NBB) were selected The feed water of both the single and binary dye systems was prepared at a total dye concentration of ca 20 mg L-1 in DI water, and the concentration ratio for every binary mixture was 1:1 In the second research content on GO and/or PVP blended PVDF membranes for enhancing dye filtration, four PVDF based membranes including the P, PG, PP and PPG membrane were explored The filtration behavior of the membranes was tested on DI water, NBB dye solution and MB-NBB mixed solution of ca 20 mg L-1 DI water, NBB and MB-NBB solutions had natural pH values of 5.5 ± 0.1, 5.8 ± 0.1 and 5.7 ± 0.2, respectively To investigate the pH effect on membrane performance, pH of the feed water was either kept at its natural pH or adjusted at 4.0 ± 0.2, 7.0 ± 0.2 or 10.0 ± 0.2 using NaOH or HCl -33- 2.4.2 Filtration performance evaluation After the filtration test, separation performance evaluations of the membrane for individual or binary mixtures of dyes were made in terms of filtrate permeance and rejection rates The permeance P (L m-2 h-1 MPa-1), of either the pure water or dye solutions, were calculated by Eq 2-1, where V (L) is the volume of the permeate flowing across the membrane of an effective area A (m2) in the time period t (h) and at the operating transmembrane pressure p (MPa) The rejection R (%), of dyes either singly or constituently in dual-dye mixtures, was quantified using Eq 2-2, where Cf and Cp are the concentrations of dyes in the feed and the permeate, respectively The dye mixture rejection was based on the rejection of the total dye concentrations before and after filtration 𝑉 (Eq 2-1) 𝑃 = 𝐴×𝑡×∆𝑝 𝑅= (𝐶𝑓 −𝐶𝑝 )×100 (Eq 2-2) 𝐶𝑓 2.4.3 Static adsorption Static adsorption experiments of single and dual dye mixtures were performed to evaluate the membrane performance comprehensively Membranes of 6.16 cm2 area were submerged into dye solutions (50 ml, ca 20 mg L-1 and natural pH) and shaked at 100 rpm at room temperature (25 ± C) for 24 h The dye concentration remaining in the solution was recorded at h intervals The membrane adsorption capacity (mg m-2 or mg g-1) was determined as a function of time All the adsorption tests were carried out twice and the average data were reported to ensure reproducibility After adsorption, the used membranes were rinsed with DI water The adsorbed membranes were then dried in a vacuum oven before further characterization -34- 2.4.4 Multi-cycle filtration In order to evaluate the pure PVDF membrane performance stability, long-term filtration experiments with cycles were conducted for the single and binary dye systems During the 1st cycle with a prolonged filtration time of h, the evolution of rejection and permeance values were recorded The 2nd and 3rd cycles of 2-h duration were aimed to initially assess the membrane reusability After every cycle, the used membrane was cleaned by flushing the membrane filtration unit with DI water for The membrane was then employed for the next cycle After the 8-h stability test, the used membrane was only wiped off the remaining dye feed (hereafter called “used membrane”) or cleaned using DI water (hereafter called “cleaned membrane”) The used and cleaned membranes were then dried in a vacuum oven before further characterization -35- Chapter Filtration Mechanisms of Binary Dye Mixtures Using a PVDF Membrane 3.1 Membrane characterization The as-prepared pure PVDF membrane formed a typical microporous structure, as depicted in the FESEM micrographs (Fig 3-1) The top surface of the membrane comprises numerous nearly circular pores of different sizes ranging from several to ~200 nm The membrane cross-section possesses a thickness of ~60 µm and an asymmetric morphology The selective top layer is made of small voids, and the thick sublayer is composed of parallel fingerlike macro voids Most of the finger-like cavities are observed to grow in width and length towards the sponge-like bottom region of the membrane Most PVDF-based UF membranes fabricated by non-solvent induced phase separation methods in the literature [74, 125] also indicated highly asymmetric membrane structures similar to this work Quantitative evaluation of the active pores (flow-through voids) was additionally conducted by a capillary flow method based on measuring the flow of a fluid through the membrane sample [126] The pore size distribution of the PVDF membrane is presented in Fig 3-2a There is a broad distribution of pore sizes ranging from 50 to 160 nm The majority of the pores (62.2%) lie in the size range of 60-80 nm, and each remaining pore size range occupies a small distribution portion (0.5-8.7%) -36- ... successful removal of around 97% for congo red dye and over 70% for reactive black dye Their permeation flux reached around 64 L m-2 h-1 MPa-1 for congo red and 93 L m-2 h-1 MPa-1 for reactive black -1 8-. .. membranes -9 - Process and material Dye Dye removal (in mg L-1) performance Ref Membrane filtration PES/MCNC membrane, at bar Direct red 16 (50) 8 9-9 9% [55] 4. 1-1 9.5 L m-2 h-1 PES-Fe3O4-MDA membrane,... 52 -ix- Fig 3-9 Dye rejection and permeance performance in multi-cycle filtration process for binary dye mixtures on the PVDF membrane: (a) RhB-MB, (b) EBT-NBB, (c) RhB-NBB and (d) MBNBB