Controllable Growth of Zeolitic Imidazolate Framework Composite Membranes for Gas Separation

Transcription

1 Controllable Growth of Zeolitic Imidazolate Framework Composite Membranes for Gas Separation Ezzatollah Shamsaei M. Eng. (Chemical) A thesis submitted for the degree of Doctor of Philosophy at Monash University in 2016 Department of Chemical Engineering

2 Copyright notice Ezzatollah Shamsaei (2016). I certify that I have made all reasonable efforts to secure copyright permissions for third-party content included in this thesis and have not knowingly added copyright content to my work without the owner's permission. i

3 Dedicated to: My parents, my sisters and brothers and my beloved wife whom I love the most because they never lose faith in me and give me endless support and encouragement through my whole life. ii

4 Abstract As an emerging class of hybrid organic-inorganic nanoporous material, metal organic frameworks (MOFs) with tunable pore size and chemistry are very attractive for integration into membranes and thin films for gas separation applications. Zeolitic imidazolate frameworks (ZIFs), a subfamily of MOFs, are known for their permanent porosity and exceptional chemical and thermal stability. Among a number of available ZIF materials, ZIF-8 is particularly interesting owing to its relative ease of synthesis as well as its great potential in separating small gas molecules. However, the progress on the fabrication of ZIF-8 membranes with satisfactory gas separation performance is very limited and there is no report of ZIF membranes being used in industrial scale so far. Therefore, development of simple and more effective methods to fabricate high quality ZIF molecular sieving membranes with high gas selectivity is still required. The new processing approaches require the advantages of being rapid, reproducible, scalable, and economically and environmentally viable while simultaneously producing high quality ZIF membranes. The ultimate goal of this PhD research program is to address challenges that hinder the facile synthesis of supported-zif membranes in a reproducible and scalable manner. In this thesis, three new approaches are demonstrated to potentially address these challenges. First, a novel scalable strategy of using vapor phase to chemically modify the polymer support for ZIF membrane fabrication is developed. Such surface modification enabled fast formation of a continuous ZIF-8 ultrathin layer after only 3 minutes. The resulting ZIF-8 membranes exhibited exceptional H2 permeance as high as mol m -2 s -1 Pa -1 with high H2/N2 and H2/CO2 selectivities (9.7 and 12.8, respectively). Next, based on the chemical vapor modification, a simple, effective, and environmentally friendly method is described for the fabrication of high-quality ZIF-8 membranes with controllable positioning on a polymer substrate in aqueous solution. The ZIF-8 membrane iii

5 exhibited a propylene permeance of mol m 2 s 1 Pa 1 and excellent selective permeation properties; after post heat-treatment, the membrane showed ideal selectivities of C3H6/C3H8 and H2/C3H8 as high as 27.8 and 2259, respectively. The new synthesis approach holds promise for further development of the fabrication of high-quality polymer-supported ZIF membranes for practical separation applications. Finally, a new concept for the use of onedimensional material (e.g. CNT) as nano-scaffolds and pseudo-seeds for the fabrication of molecular sieving membranes supported on a porous substrate is introduced. To demonstrate the potential for universal applicability of the proposed pseudo-seeding and nano-scaffolding method, ZIF-8/CNTs membranes were prepared on both polymeric and inorganic substrates. At 25 C and 1 bar, the ideal separation selectivities of H2/CO2, H2/N2, H2/CH4, H2/C3H6, and H2/C3H8 are 14, 18, 35, 52.4 and 950.1, respectively, with H2 permeance as high as mol m 2 s 1 Pa 1. This high hydrogen permselectivity combined with its mechanically reinforced structure shows that the ZIF-8/CNT membrane is a promising candidate for hydrogen separation and purification. Finally, it is anticipated that the novel strategies developed in this research may be further developed for the fabrication of other MOF and zeolite molecular sieve membranes. iv

6 Declaration This thesis contains no material which has been accepted for the award of any other degree or diploma at any university or equivalent institution and that, to the best of my knowledge and belief, this thesis contains no material previously published or written by another person, except where due reference is made in the text of the thesis Signature: Print Name: Ezzatollah Shamsaei Date: 20/09/2016 v

7 List of Publications Journal Publications: 1. Shamsaei, E.; Lin, X.; Wan, L.; Tong, Y.; Wang, H. A One-Dimensional Material as a Nano-Scaffold and a Pseudo-Seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes. Chem Commun. 2016, 52 (95), Shamsaei, E.; Lin, X.; Low, Z.-X.; Abbasi, Z.; Hu, Y.; Liu, J. Z.; Wang, H. Aqueous Phase Synthesis of ZIF-8 Membrane with Controllable Location on an Asymmetrically Porous Polymer Substrate. ACS Appl. Mater. Interfaces 2016, 8, Shamsaei, E.; Low, Z.-X.; Lin, X.; Mayahi, A.; Liu, H.; Zhang, X.; Zhe Liu, J.; Wang, H. Rapid Synthesis of Ultrathin, Defect-Free ZIF-8 Membranes Via Chemical Vapour Modification of a Polymeric Support. Chem. Commun. 2015, 51, Shamsaei, E.; Low, Z.-X.; Lin, X.; Liu, Z.; Wang, H. Polysulfone and Its Quaternary Phosphonium Derivative Composite Membranes with High Water Flux. Ind. Eng. Chem. Res. 2015, 54, Oral and Poster Presentations: 1. E. Shamsaei, X. Lin, Z-X. Low, Z. Abbasi, Y. Hu, J.Z. Liu, H. Wang. Development of Metal Organic Framework Composite Membranes for Improved Gas Separation Properties (Received Best Oral Presentation Award by European Membrane Society (EMS)). PERMEA & MELPRO Conference, Prague - Czech Republic, May E. Shamsaei, K. Wang, Z-X. Low, X. Lin, H. Wang. Enhanced water permeation through nanoporous polymer membranes (Received Best Poster Presentation Award by Membrane Society of Australasia (MSA)). 4th Membrane Society of Australasia Early Career Researcher Symposium at Geelong Victoria 19th to 21st of November E. Shamsaei, Z-X. Low, H. Wang. High flux polysulfone-based ultrafiltration membrane. Chemeca, Perth, Western Australia, Australia, 28th September to1st October Other journal publications during enrolment: 1. Lin, X.; Kim, S.; Zhu, D.M; Shamsaei, E.; Xu, T.; Fang, X; Wang, H. Preparation of porous diffusion dialysis membranes by functionalization of polysulfone for acid recovery. J. Membr. Sci. 2016, 524, Abbasi, Z.; Shamsaei, E.; Leong, S.; Ladewig, B.; Zhang, X.; Wang, H. 2016, Microporous Mesoporous Mater. 2016, 236, Zhao, Z.; Shamsaei, E.; Feng, Y.; Song, J.; Wang, H.; He, L. J. Membr. Sci. 2016, 518, 1 9. vi

8 4. Lin, X.; Shamsaei, E.; Kong, B.; Liu, J. Z.; Zhao, D.; Xu, T.; Xie, Z.; Easton, C. D.; Wang, H. J. Membr. Sci. 2016, 510, Lin, X.; Shamsaei, E.; Kong, B.; Liu, J. Z.; Hu, Y.; Xu, T.; Wang, H. J. Membr. Sci. 2016, 502, Wan, L.; Wei, J.; Liang, Y.; Hu, Y.; Chen, X.; Shamsaei, E.; Ou, R.; Zhang, X.; Wang, H. RSC Advances. 2016; 6, Lin, X.; Shamsaei, E.; Kong, B.; Liu, J. Z.; Xu, T.; Wang, H. J. Mater. Chem. A 2015, 3, Feng, Y.; Shamsaei, E.; Davies, C. H. J.; Wang, H. Mater. Chem. Phys.2015, 167, Low, Z.-X.; Liu, Q.; Shamsaei, E.; Zhang, X.; Wang, H. Membranes 2015, 5, Moslehyani, A.; Mobaraki, M.; Ismail, A. F.; Matsuura, T.; Hashemifard, S.A., Othman, M.H.D., Mayahi, A., DashtArzhandi, M.R., Soheilmoghaddam, M.; Shamsaei, E. React Funct Polym. 2015, 95, Mayahi, A.; Ilbeygi, H.; Ismail, A. F.; Jaafar, J.; Daud, W. R. W.; Emadzadeh, D.; Shamsaei, E.; Martin, D. J. Chem. Technol. Biotechnol. 2015, 90, Shamsaei, E.; Nasef, M. M.; H. Saidi; Yahaya, A. H. Radiochim. Acta 2014,102, vii

9 Acknowledgment I would like to express my gratitude to my supervisor, Professor Huanting Wang for his great support and guidance throughout my Ph.D. study. I would also like to thank Dr Zhe Liu, my associate supervisor, Assoc Prof Xiwang Zhang, Prof Jianfeng Yao and Dr Kun Wang for their help. Thanks to the members in our group including Dr Xiaocheng Lin, Dr Ze-Xian Low, Dr Yi Feng, Dr Dongbo Yu, Dr Jing Wei, Dr Soo Kwan Leong, Dr Seungju Kim, Li Wan, Ranwen Ou, Xiaofang Chen, Yan Liang, Yaoxin Hu, Kang Liu, Dr Kha Tu, Zahra Abbasi, Dr Huacheng Zhang, and Prof Yuping Tong for their kind help. Appreciation also goes to, Dr Meng Na, Huiyuan Liu for their continuous support and companionship during my study. To my friends in the department, Dr Tahereh Hosseini, Shahrouz Taranejoo, Soroush Shakiba, and Sajjad Asadi who encouraged me and supported me during the PhD journey. I also wish to express my gratitude to the staffs in the Department of Chemical Engineering, especially Kim Phu, Jill Crisfield and Lilyanne Price for their help during my study. Special thanks to Sally El Meragawi and Joanne Tanner for proof reading my thesis and helping me with my English. I would like to acknowledge the financial support from Monash University. Without this funding the research would not have been possible. I would like to thank my beloved mother and father. Thanks to my dear sisters and brothers to support me and believe me, without you I couldn t achieve any success in my life. Finally, I would like to thank my beautiful wife, Parisa, my best friend and best wife ever. You have supported me and encouraged me during all difficult moments of my work. Without you and your encouragement I couldn t survive these past three and half years. viii

10 Contents ABSTRACT... III DECLARATION... V LIST OF PUBLICATIONS... VI ACKNOWLEDGMENT... VIII CONTENTS... IX LIST OF FIGURES... XI LIST OF TABLES... XV NOMENCLATURE... XVI INTRODUCTION... 1 BACKGROUND AND CHALLENGES... 1 RESEARCH AIMS... 6 THESIS STRUCTURE AND CHAPTER OUTLINE... 7 REFERENCES... 9 LITERATURE REVIEW OVERVIEW POLYMERIC MEMBRANES INORGANIC MEMBRANES METAL ORGANIC FRAMEWORKS (MOFS) MOF Materials and Fabrication Synthesis of MOF-based membranes Supported MOF membranes MOF-based mixed-matrix membrane CONCLUSION AND PERSPECTIVES REFERENCES RAPID SYNTHESIS OF ULTRATHIN, DEFECT-FREE ZIF-8 MEMBRANES VIA CHEMICAL VAPOUR MODIFICATION OF POLYMERIC SUPPORT OVERVIEW INTRODUCTION EXPERIMENTAL Materials Synthesis of BPPO membrane and its EDA-vapour modification Growth of ZIF-8 Thin Film on modified BPPO Supports Pure water flux and molecular weight cut off (MWCO) measurements Gas permeation experiments Characterization RESULTS AND DISCUSSION Membrane support Fast in Situ Seeding ix

11 3.4.3 Supported ZIF-8 membrane Single gas performance SUMMARY REFERENCES AQUEOUS PHASE SYNTHESIS OF ZIF-8 MEMBRANE WITH CONTROLLABLE LOCATION ON AN ASYMMETRICALLY POROUS POLYMER SUBSTRATE OVERVIEW INTRODUCTION MATERIALS AND METHODS Chemicals Sample preparation Characterization Single gas permeation test Measurements of the support pore size RESULTS AND DISCUSSION Membrane support Supported ZIF-8 membrane Membrane prepared by conventional contra-diffusion Single gas performance Effects of activation temperature on the ZIF-8 membranes SUMMARY REFERENCES ONE-DIMENSIONAL MATERIAL AS NANO-SCAFFOLD AND PSEUDO-SEED FOR FACILITATED GROWTH OF ULTRATHIN, MECHANICALLY REINFORCED MOLECULAR SIEVING MEMBRANES OVERVIEW INTRODUCTION MATERIALS AND METHODS Chemicals Polydopamine modification of CNTs Preparation of ZIF-8/CNTs membrane on porous AAO disk Characterization Gas permeation experiments RESULTS AND DISCUSSION PDA-coated CNTs ZIF-8/CNT membrane CNTs coverage level Mechanical and structural stability of the ZIF-8/CNT membranes Synthesis time Single gas performance Universal applicability SUMMARY REFERENCES CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK CONCLUSIONS RECOMMENDATIONS FOR FUTURE WORK x

12 List of figures Figure 1-1. Olefin/paraffin experimental upper bound based on pure gas permeation data. Symbols: ( ) 100 C; ( ) 50 C; ( ) 35 C; ( ) 30 C; ( ) 26 C [7] Figure 2-1. Schematic representation of various types of membranes [4] Figure 2-2. Relationship between the slope (n) of the upper bound and the difference between the kinetic diameters of the gas pairs [12] Figure 2-3. Loss of selectivity occurred in polymeric membranes with increasing partial pressures of CO2 [13]. Open points, pure gas. Closed points, mixed gas Figure 2-4. a) Chemical structure of zeolite, b) Primary building unit of zeolite structure [28].. 20 Figure 2-5. Silicon atoms are positioned at the intersections and linked by lines. (a) Sodalite cage; (b) zeolite A, the sodalite cages are connected to each other by double 4-membered rings and form an α-cage indicated by circle; (c) zeolite Y, the sodalite cages are linked by double 6-membered rings and organized as in the diamond framework [27] Figure 2-6. Comparison of the effective pore sizes of various zeolites and the kinetic diameters of common gas molecules [35] Figure 2-7. Crystalline structure for three commonly used MOFs for separations Figure 2-8. Section of the crystal packing diagram of ZIF-8 [57] Figure 2-9. Wire-frame model and ball-and-stick model of the crystal structure of (A) ZIF-L and (B) ZIF-8 [67] Figure SEM images of (a) ZIF-L nanoflakes, (b) ZIF-8 nanoparticles [67] Figure Demonstration of UiO-66: (a) secondary building units (SBUs), (b) BDC ligand; (c) crystal model, (d) a simplified form [74]; (e) SEM images of UiO-66 powders [73] Figure Schematic diagram of in-situ synthesis of pure UiO-66 membranes supported on porous hollow fibers [73] Figure Fabrication methods for MOF-based membranes Figure Scheme of the fabrication methods for continuous MOF membranes [87] Figure (a) Diffusion cell for contra-diffusion preparation of ZIF-8 film and (b) the schematic synthesis of ZIF-8 films on the sides of the nylon substrate by contra-diffusion of Hmim and Zn 2+ through the porous nylon [56] Figure Synthesis illustration for the CuBTC@PSF membrane by in situ method and layer by layer crystal deposition [99] Figure H2/CH4 (a) and H2/N2 (b) separation factors as a function of H2 permeance for ZIF-8 membranes in [100] as compared to those in the literature Figure Schematic diagram of constituting layers of a supported-zif-8 membrane [102].. 37 xi

13 Figure SEM images of: a) the ZIF-90 seed-layer, b) top view and c, d) cross section views of the polycrystalline ZIF-90 membrane after secondary growth [104] Figure a) Schematic representation of oriented synthesis of HKUST-1 crystals controlled by surface functionalization [75]. b) A simplified model of anchoring a typical MOF-5 building unit to a carboxylic acid-terminated self-assembled organic monolayers (SAM) [76] Figure Scheme of the morphology and chemical structure of a MOF/PVDF membrane prepared by a non-activation method [96] Figure (a) Optical images of the (1) pristine PAN hollow fiber, (2) hydrolyzed PAN hollow fiber and (3) Cu3 (BTC)2 PAN hollow fiber membrane. (b),(c) proposed involved chemical reactions of the PAN hollow fiber [97] Figure Schematic illustration of (a) Zeolite Membranes on polymer zeolite MM hollow fiber supports [111] and (b) trinity MOF membranes preparation [112] Figure Scheme of pressure-assisted preparation of HKUST-1 layer on the surface of PVDF hollow fibers substrate [113] Figure Schematic diagram of a typical MOF membrane and MOF based MMM Figure 3-1(a) Schematic diagram of UF membrane fabrication via phase inversion, (b) Experimental setup of vapour-phase EDA modification process Figure 3-2. Schematic diagram of the preparation of BPPO polymer-supported ZIF-8 membrane Figure 3-3. FTIR ATR spectra of untreated BPPO support, BPPO modified with EDA-vapour for 16 h (MBPPO-16), MBPPO-16 supported ZIF-8 layer (ZIF-8-MBPPO-16), and synthesized ZIF- 8 powder Figure 3-4. TGA curves (under air flow) of (1) untreated BPPO support, (2) MBPPO-16, (3) ZIF- 8-BPPO, (4) ZIF-8-MBPPO-16, and (5) synthesized ZIF-8 powder Figure 3-5. Pure water flux and pore size of BPPO membranes as a function of exposure time to EDA vapour Figure 3-6. XRD pattern (a), TEM image and SAED pattern (inset) (b) and nitrogen sorption isotherm (c) of as-synthesized ZIF-8 nanocrystals Figure 3-7. SEM images of (a) cross-section and (b) surface of the ZIF-8-MBPPO Figure 3-8. EDS line scan across ZIF-8-MBPPO-16 cross-section for the zinc atoms Figure 3-9. SEM images of the (a) surface and (b) cross section of the BPPO support, (c) surface of the ZIF-8-BPPO, (d) surface of the ZIF-8-MBPPO-4, (e) surface of the ZIF-8-MBPPO-10, (f) surface and (g, h) cross-section of the ZIF-8-MBPPO Figure XRD patterns of the membranes and simulated ZIF-8 powder Figure FTIR ATR spectra of (1) untreated BPPO support, BPPO modified with EDA-vapour for (2) 4 h (MBPPO-4), (3) 10 h (MBPPO-10), (4) 16 h (MBPPO-16), (5) ZIF-8-MBPPO-16, (6) synthesized ZIF-8 powder xii

14 Figure Single gas permeances of ZIF-8-MBPPO-16 as a function of kinetic diameter of gas molecule Figure 4-1. Digital photograph of a home-made contra-diffusion cell Figure 4-2. Schematic diagram of gas permeation set-up Figure 4-3. Schematic diagram of the preparation of a BPPO polymer supported ZIF-8 membrane using chemical vapour modification and subsequent contra diffusion synthesis Figure 4-4. FTIR ATR spectra of the untreated BPPO support, BPPO modified with EDA-vapor (BPPO-EDA), BPPO-EDA supported ZIF-8 layer (ZIF-8-BPPO-EDA), and ZIF-8 powder Figure 4-5. Thermogravimetric analysis (under air flow) of untreated BPPO and BPPO-EDA substrates Figure 4-6. SEM images of untreated BPPO (a), vapour-phase-eda-modified BPPO (BPPO-EDA) (b), ZIF-8@BPPO-EDA grown for 60 min (c), 90 min (d), and 120 min (e, f) Figure 4-7. SEM images of ZIF-8 membranes grown for 2 h (a, b, c), 4 h (d, e, f), 6 h (g, h, i) via conventional contra-diffusion method using untreated BPPO substrate Figure 4-8. XRD patterns of ZIF-8@BPPO-EDA membranes as a function of growth time Figure 4-9. SEM images of ZIF-8@BPPO-EDA grown for 120 min at different magnifications. Cross-sectional view: (a, b, c); Top view: (d, e, f) Figure EDS line scan across ZIF-8@BPPO-EDA-120 cross-section for the zinc atoms Figure SEM images of cross-section of ZIF-8 membrane synthesized by contra-diffusion (at the reaction time of 2 h) after modification of the BPPO via immersion of the polymer in 60% EDA aqueous solution at room temperature for 3 h Figure Single gas permeances (a) and ideal selectivities (b) as a function of kinetic diameter of gas molecules of the ZIF-8 membranes grown for 120 min and activated at 120 C (ZIF8@BPPO-EDA ) and at 150 C (ZIF8@BPPO-EDA ) Figure Comparison of C3H6 permeability and C3H6/C3H8 selectivity of the membranes in the present work with previously reported membranes. Closed and open symbols indicate separation data obtained from single and binary gas permeation analysis, respectively. Hexagon: inorganic supported ZIF-8 membranes [10]; pentagon: ZIF-8 mixed matrix membranes [15]; triangle: polymer membranes [42]; circle: carbon membranes [43]; star: polymer supported ZIF-8 membranes in this study Figure SEM images of ZIF-8@BPPO-EDA-120 membranes after activation at 150 ºC (a, b, c) and 200 ºC (d, e, f). (a, d, e) top view and (b, c, f) cross-sectional view Figure Room-temperature propylene/propane permeation properties of ZIF-8 membranes grown for 120 min (ZIF-8@BPPO-EDA-120) as a function of activation temperatures Figure Heat-induced cross-linking of BPPO substrate Figure FTIR ATR spectra of the BPPO support, BPPO support after being heated at 150 ºC (BPPO-150) for 2h under air, EDA-vapour-modified BPPO (BPPO-EDA), EDA-vapour-modified BPPO after being heated at 150 ºC (BPPO-EDA-150) for 2h under air xiii

15 Figure (a) FTIR spectra and (b) XRD patterns of membranes (grown for 120 min) as a function of activation temperature ( C) Figure 5-1. Schematic illustration of the preparation process of ZIF-8/CNT membrane through deposition of modified CNTs on the support, followed by a contra-diffusion synthesis Figure 5-2. (a) Photos of water dispersibility and the corresponding TEM images of CNTs (I) and polydopamine-coated CNTs (II), XRD patterns (b) and FTIR spectrum (c) of CNTs (I) and polydopamine-coated CNTs (II), (f) schematic illustration of the coated CNT and the chemical structure of polydopamine Figure 5-3. SEM (a, b) and optical (c) images of pristine (a, c1-c3) and modified (b, c4-c6) CNTsdeposited on AAO. Detailed experimental: deposited pristine CNTs from (c1, c4) 1 ml (c2, c5) 3 ml and (c3, c6) 6 ml mother solution. Pristine CNTs mother solution: 10 mg CNTs in 200 ml DDI water Figure 5-4. SEM images of modified CNTs-deposited on AAO (a), ZIF-8/CNT membranes grown for 5 min (b), 30 min (c), and for 60 min (d, e), and XRD patterns of ZIF-8 membranes as a function of synthesis time (f) Figure 5-5. SEM images of ZIF-8 film prepared on (a, b, c) bare AAO and on (d, e, f) AAO deposited with pristine CNTs. Zinc side: b, e; Hmim side: c, f. Synthesis time: 60 min Figure 5-6. SEM images of bare AAO (a), as-prepared samples with insufficient (b, c), and excess (d, e, f) deposition of modified CNTs on AAO before (b, d) and after (c, e, f) contra-diffusion synthesis. The inset in f is a high magnification cross-sectional view. Detailed experimental: deposited CNTs from (b) 1 ml and (d) 6 ml mother solution respectively Figure 5-7. SEM images (a, b) and XRD pattern (c) of the ZIF-8/CNTs membrane after sonication for 2 h Figure 5-8. Optical image of the free standing ZIF-8/CNT hybrid membrane floated in the sodium hydroxide solution (a) and SEM images of Cross-sectional view (b, c) and surface edge (d) of the corresponding free standing membrane Figure 5-9. Single gas permeances of several gases through the ZIF-8/CNT membrane at 20 C and different feed pressures Figure FTIR ATR spectra of the AAO support deposited with modified CNTs, the ZIF- 8/CNTs membranes as a function of synthesis time, and ZIF-8 powder Figure Single gas permeances of several gases through the ZIF-8/CNT membrane at 20 C and 1 bar as a function of the kinetic diameter. The inset shows the ideal gas selectivity for H2 over other gases Figure SEM images of (a, b) bare PES, (c, d) PES with deposited modified CNTs (6 ml of mother solution) and (e, f) as prepared membrane after contra-diffusion synthesis (1 h) Figure (a) XRD patterns and (b) FTIR spectra of pristine PES, supported ZIF-8/CNT membrane and ZIF xiv

16 List of tables Table 2-1. Primary membrane companies and polymeric membrane materials [9] Table 3-1. Single gas permeances and ideal selectivities for the composite membranes at 25 ⁰C and 1 bar Table 3-2. Single gas permeances and ideal selectivities at 25 ⁰C and 1 bar of 3 tested ZIF-8- MBPPO-16 membranes showing the reproducibility of membrane synthesis and testing Table 3-3. Comparison of gas permeation properties (H2 permeance, H2/N2 and H2/CO2 selectivity) of ZIF-8 membranes on inorganic and polymeric supports reported in recent literature Table 4-1. Single gas permeances and ideal selectivities for the ZIF-8@BPPO-EDA-x-y (x: crystallization time (min); y: heat treatment temperature (⁰C)) composite membranes at 25 ⁰C and 1 bar Table 4-2. Comparison of gas permeation properties of the ZIF-8@BPPO-EDA composite membrane in this work with other ZIF-8 membranes in the literature Table 5-1. Single gas permeances and ideal selectivities for the ZIF-8@CNTs-t (t: crystallization time (min), hybrid membranes at 20 ⁰C and 1 bar. E shows the sample prepared with an excess use of CNTs (6ml of the mother solution) Table 5-2. Comparison of the synthesis parameters (time and temperature) and gas permeation properties of the ZIF-8/CNTs hybrid membrane in this work with other ZIF-8 membranes from the recent literature Table 5-3. Single gas permeances and ideal selectivities of three ZIF-8/CNT-60 membrane samples tested at 25 ⁰C and 1 bar xv

17 AAO: Anodized aluminum oxide Nomenclature ATR: Attenuated total reflectance BDC: 1, 4-benzenedicarboxylic acid BET: Brunauer-Emmett-Teller BPPO: Bromomethylated poly(2,6-dimethyl-1,4-phenylene oxide) CNF: Carbon nanofiber CNT: Carbon nanotubes CuBDC: Copper 1, 4-benzenedicarboxylate DDI: Double-deionized water DMF: Dimethylformamide EDA: Ethylenediamine EDX: Energy-dispersive X-ray spectroscopy FTIR: Fourier transform infrared HKUST: Hong Kong University of Science and Technology Hmim: 2-methylimidazole IRMOF: Isoreticular metal organic framework LbL: Layer-by-layer LPE: Liquid-phase epitaxy xvi

18 LTA: Linde type A MIL: Matériaux de l'institut Lavoisier MMM: Mixed matrix membranes MOF: Metal organic frameworks MWCNT: Multiwall carbon nanotubes MWCO: Molecular weight cut-offs NMP: 1-methyl-2-pyrrolidinone PAN: Polyacrylonitrile PBI: Polybenzimidazole PCP: Porous coordination polymers Pd: Palladium PDA: Polydopamine PDMS: Polydimethylsiloxane PEG: Polyethylene glycol PEI: Polyetherimide PES: Poly (ether sulfone) PIM: Polymers of intrinsic microporosity PSf: Polysulfone PVDF: Polyvinylidene fluoride xvii

19 RO: Reverse osmosis RT: Room temperature RTD: Rapid thermal deposition SAED: Selected-area electron diffraction SAM: Self-assembled monolayer SBU: Secondary building units SEM: Scanning electron microscopy SOD: Sodalite TEM: Transmission electron microscopy TGA: Thermogravimetric analyses TMP: Transmembrane pressure TOC: Total organic carbon TPIM-1: Triptycene-based ladder polymer Tris: Tris (hydroxymethyl)aminomethane UF: Ultrafiltration UiO: University of Oslo XRD: X-ray diffraction ZIF: Zeolitic imidazolate frameworks 1D: One-dimensional xviii

20 Chapter 1 Introduction Introduction Background and Challenges Membranes have found an essential role in chemical technology and are used in a wide variety of applications such as water purification and desalination, wastewater treatment, food and dairy, medical and chemical production. The main characteristic of a membrane that is utilized is its ability to regulate the permeation rate of a chemical species across the membrane. In separation applications, the aim is to allow one component of a mixture to freely pass through the membrane, while inhibiting permeation of other components. Therefore, a membrane is principally a thin interface that controls the permeation of chemical species in contact with it [1]. In particular, membrane-based separation methods are gaining increasing importance for energy efficient gas separations and other molecular separations [2]. Currently, almost all of gas separation membranes used commercially are polymeric. Polymer membranes can be readily processed into a number of forms and modules (e.g. hollow fibers or anisotropic spiral-wound) and still maintain their separation performance. However, polymer membranes have a number of limitations such as short lifespans, low thermal and chemical stabilities [3], and a separation efficiency that is restricted by the well-known trade-off relationship between selectivity and permeability of the polymer-based membranes. More selective polymers are generally less permeable and vice versa [4, 5]. Based on literature data, Robeson quantified the gas separation performance of a number of polymers for the separation of O2/N2 and CO2/CH4 and constructed the so-called upper bound trade-off line between permeability and selectivity [6]. Similar results were observed for the C3H6/C3H8 separation by Koros and coauthors [7]. The limitations of polymers in terms of the upper bound 1

21 Chapter 1 Introduction line has also been addressed by Freeman [8]. The logical extension of these earlier studies lies in the development of new membrane materials that go beyond this trade-off limit. Figure 1-1. Olefin/paraffin experimental upper bound based on pure gas permeation data. Symbols: ( ) 100 C; ( ) 50 C; ( ) 35 C; ( ) 30 C; ( ) 26 C [7]. Ceramic membranes and zeolites are of particular interest for gas separation due to their advantages over polymers which include high thermal and chemical stabilities and promising separation efficiency. However, these materials are more expensive, more complex in preparation, less reproducible, and have lower mechanical resistance as compared to polymer membranes [9]. Carbon membranes have also been considered as promising non-traditional materials for separation applications under high temperatures and harsh chemical environments. Then again, carbon-based membranes have poor mechanical properties and the scale-up to industrial size is problematic [10]. As emerging hybrid class of organic-inorganic nanoporous materials, metal organic frameworks (MOFs) are very attractive option for integration into membranes and thin films for a wide range of industrial applications including membrane-based gas separations. Their welldefined pore structures can be rationally designed by the interplay of their building blocks, i.e. 2

22 Chapter 1 Introduction metal ions and organic linkers, and their pore chemistry is readily tunable via a variety of methods [11]. MOF synthesis is less energy intensive as compared to zeolites. For example, unlike zeolites, fabrication of most MOFs is conducted at relatively low temperature and pressure conditions, without the use of structure-directing agents, which eliminates the calcination step required in the synthesis of zeolites. Furthermore, MOFs possess larger pore volumes and lower density than zeolites, which make them more attractive for the preparation of composite membranes [12]. Several reports have discussed the synthesis and applications of MOF materials [13, 14]; however, relatively few reports exist addressing MOF membranes. This inequality is due to challenges associated with the fabrication of MOF membranes. As the MOF layer does not hold adequate mechanical strength as a self-supporting membrane, it is essential to be prepared on a mechanically strong, porous support. Fabrication of supported-mof membranes is accordingly challenging due to the difficulty of directing nucleation and crystal growth onto the surface and poor MOF-to-substrate adhesion. MOF films/membranes are prepared by growing a thin MOF layer on a substrate via two general techniques, i.e. in situ (direct) growth and secondary (seeded) growth. The in situ synthesis is a simple method that allows for simultaneous nucleation, deposition and crystal growth by immersing the substrate in a MOF precursor solution. Seeded growth involves the synthesis and anchoring of seed crystals on substrates, followed by their crystallization. Although secondary growth requires additional steps, seeded growth has been noted to more effectively induce controlled MOF growth on the porous support. The first MOF film was prepared in 2005 by selectively anchoring MOF-5 particles at the carboxylate-terminated areas of the self-assembled monolayer (SAM) on the surface of a dense substrate [15]. Similarly, an oriented growth of MOFs on SAM-functionalized metal substrates 3

23 Chapter 1 Introduction has been reported [16]. In 2007, fabrication of microporous MOF film was investigated on porous substrates [17]. It was found that the support surface properties and the synthesis route were essential factors that affect the MOF film density. The growth of much denser MOF membranes on a porous substrate was then achieved through a seeded method [18]. However, none of these pioneering studies reported gas separation results, indicating it is critical yet still challenging to fabricate a compact polycrystalline MOF membrane. Unlike MOF films which are typically used for applications such as sensors, MOF membranes utilized in gas separation applications require well-structured grain boundary and absence of pinholes or defects, minimizing the nonselective intercrystalline diffusion. Very intimate contact between the MOF layer and the support is also critical as to provide the sufficient mechanical stability when operating under harsh environment in commercial applications. In 2009, the first MOF-based membrane was reported for gas separation. Continuous MOF-5 membrane was prepared on a modified alumina disk by a solvothermal synthesis. Shortly thereafter, continuous membranes of ZIF-7 and ZIF-8 were prepared on alumina and titanium substrates, respectively, by a microwave-assisted solvothermal preparation method. The absence of macroscopic defects in these reports were confirmed by pressure dependent gas permeation measurements. These reports demonstrated the feasibility of constructing supported MOF membranes for gas separations and other molecular discriminations. To date, a number of porous organic (e. g PVDF, PES, and nylon) and inorganic (alumina, silica, and porous ZnO) materials have been used as supports for the fabrication of MOF membranes. Several innovative methods have been reported on the synthesis of MOF thin films and membranes including contra-diffusion method [19], liquid-phase epitaxy [20], rapid thermal deposition (RTD) [21], layer-by-layer deposition of crystals [22], and interfacial microfluidic processing [23]. Even though much progress on the construction of supported-mof membranes has been achieved, there is 4

24 Chapter 1 Introduction still much research to be conducted for the facile fabrication of high quality MOF membranes before robust synthesis strategies can be developed. It is also noted that there are no reports of supported- MOF membranes being employed as separating membranes on the industrial scale so far. Zeolitic imidazolate frameworks (ZIFs), a subclass of metal organic frameworks (MOFs), are porous crystalline hybrid materials consisting of imidazolate ligands (Im) bridging tetrahedral metal ions (e.g., Zn, Co) [24]. They closely resemble the topologies of zeolites, due to the M-Im- M (M = Zn, Co) bond angle of 145, which is close to the T-O-T (T = Al, Si, P) angle ( ) in zeolites [14, 25]. ZIFs show properties that combine the attractive features of both MOFs and zeolites such as tunable pore size and chemistry, large internal surface area, and relatively good thermal and chemical stability [26, 27]. These properties make ZIFs excellent candidates for the fabrication of molecular sieving membranes for gas separation [28, 29]. Thus, fabrication of supported-zif membranes has drawn a lot of attention in recent years [11]. Among a number of available ZIF materials, ZIF-8 is particularly interesting owing to its relatively facile synthesis procedure as well as its great potential in separating small gas molecules. ZIF-8 is made of zinc ions bridged by 2-methylimidazolate ligands, forming the sodalite-related structure with a large cavity (11.6 Å) and small aperture size (3.4 Å). Several fabrication methods for ZIF membranes have been published which can be categorized into two groups: direct growth and secondary growth [30, 31]. Nagaraju et al. [22] and Cacho-Bailo et al. [32] grew ZIF-8 on a porous polysulfone substrate using direct growth. However, although the direct synthesis is a simple method that allows for simultaneous nucleation, deposition and crystal growth, it is not very effective in preparing continuous ZIF membranes due to limited heterogeneous nucleation sites on the substrate [33]. Alternatively, Ge at al. [34] used secondary seeded growth to fabricate a continuous ZIF-8 film on an asymmetrically porous poly(ether sulfone) substrate. Secondary 5

25 Chapter 1 Introduction seeded growth has been shown to effectively induce controlled ZIF growth on polymer supports, but the resulting ZIF layer often suffers from weak adhesion to the support, leading to membrane delamination. Furthermore, only a limited number of studies reported ZIF-8 membranes with satisfactory gas separation performances [31]. Therefore, development of facile and more effective methods to fabricate high quality ZIF molecular sieving membranes with high gas selectivity is still required. The new processing approaches require the advantages of being rapid, reproducible, scalable, and economically and environmentally viable while simultaneously producing high quality ZIF membranes. Research Aims The overall aim of this project is to develop novel methods for fabrication of ultrathin ZIF-8 molecular sieve membranes with high selectivity performances. The specific aims are summarised as: Growing ZIF-8 on polymer substrate to achieve high-quality membranes at low cost. Developing a novel scalable strategy of using vapour phase to chemically modify the polymer support for ZIF membrane fabrication. Investigating the role of the modifier in the mechanism of the growth. Developing a simple, effective, and environmentally friendly method for the fabrication of high-quality ZIF-8 membrane with controllable placement on a polymer substrate in aqueous solution. Utilizing one-dimensional material as a nano-scaffold and pseudo-seed for the fabrication of molecular sieving membranes supported on a porous substrate, inspired by the success of the 6

26 Chapter 1 Introduction nano-scaffolding technique in tissue engineering where biocompatible nanofibers are used as nano-scaffold to promote tissue growth and provide mechanical support [35]. Thesis structure and chapter outline This thesis is organised into six sections; the overview of each chapter is summarized below. Chapter 1 (Introduction) contains a brief background for the motivation of this thesis, the research aims and an outline of the thesis structure. In chapter 2 (literature review), a short introduction to the structure and chemistry of several MOFs is provided. Then focus is refined to the currently pursued strategies of supported-mof membrane fabrication, associated challenges and reported approaches addressing these problems. Chapter 3 provides detailed experimental procedures for the rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of a polymeric support. The results of the experiments are presented, analysed and discussed. The gas permeation test results for the membranes are also given. This chapter has been published as a journal article: E. Shamsaei, Z.-X. Low, X. Lin, A. Mayahi, H. Liu, X. Zhang, J. Z. Liu, H. Wang, Chem. Commun. 2015, 51, Chapter 4 demonstrates a simple, scalable, and environmentally friendly route for controllable fabrication of continuous, well-intergrown ZIF-8 on a flexible polymer substrate via contradiffusion in conjunction with chemical vapor modification of the polymer surface. The results of experiments are presented, analyzed and discussed. The gas permeation test results for the membranes are also given. 7

27 Chapter 1 Introduction This chapter has been published as a journal article: E. Shamsaei, X. Lin, Z.-X. Low, Z. Abbasi, Y. Hu, J. Z. Liu, H. Wang, ACS Appl. Mater. Interfaces 2016, 8, Chapter 5 presents a new concept for using one-dimensional material as a nano-scaffold and pseudo-seed for the fabrication of molecular sieving membranes supported on a porous substrate. This chapter provides detailed experimental procedures to utilize one-dimensional (1D) materials such as carbon nanotubes (CNTs) and carbon nanofibers (CNFs) to form a porous nano-scaffold and pseudo-seed layer on the porous substrate for facilitated growth of ultrathin ZIF membranes with mechanically reinforced structures. The results of experiments are presented, analyzed and discussed. The gas permeation test results for the membranes are also given. This chapter has been submitted as a journal article: E. Shamsaei, X. Lin, L. Wan, Y. Tong, H. Wang, Chem Commun. 2016, 52 (95), Chapter 6 summarizes the major findings of this thesis, and recommendations are presented for the future work. 8

28 Chapter 1 Introduction References [1] R.W. Baker, Membrane Technology and Applications, 3 ed., John Wiley & Sons, Ltd, [2] N.N. Li, A.G. Fane, W.W. Ho, T. Matsuura, Advanced membrane technology and applications, John Wiley & Sons, [3] S. Qiu, M. Xue, G. Zhu, Metal organic framework membranes: from synthesis to separation application, Chem. Soc. Rev., 43 (2014) [4] L.M. Robeson, Correlation of separation factor versus permeability for polymeric membranes, J. Membr. Sci., 62 (1991) [5] S.A. Stern, Polymers for gas separations: the next decade, J. Membr. Sci., 94 (1994) [6] L.M. Robeson, The upper bound revisited, J. Membr. Sci., 320 (2008) [7] R.L. Burns, W.J. Koros, Defining the challenges for C3H6/C3H8 separation using polymeric membranes, J. Membr. Sci., 211 (2003) [8] B.D. Freeman, Basis of permeability/selectivity trade-off relations in polymeric gas separation membranes, Macromolecules, 32 (1999) [9] H. Strathmann, L. Giorno, E. Drioli, Introduction to membrane science and technology, Wiley- VCH Weinheim, [10] S. Lagorsse, F. Magalhaes, A. Mendes, Carbon molecular sieve membranes: sorption, kinetic and structural characterization, J. Membr. Sci., 241 (2004) [11] M. Shah, M.C. McCarthy, S. Sachdeva, A.K. Lee, H.-K. Jeong, Current status of metal organic framework membranes for gas separations: promises and challenges, Ind. Eng. Chem. Res., 51 (2012) [12] H.B. Tanh Jeazet, C. Staudt, C. Janiak, Metal-organic frameworks in mixed-matrix membranes for gas separation, Dalton Trans., 41 (2012) [13] H. Furukawa, K.E. Cordova, M. O Keeffe, O.M. Yaghi, The Chemistry and Applications of Metal-Organic Frameworks, Science, 341 (2013). 9

29 Chapter 1 Introduction [14] K.S. Park, Z. Ni, A.P. Côté, J.Y. Choi, R. Huang, F.J. Uribe-Romo, H.K. Chae, M. O Keeffe, O.M. Yaghi, Exceptional chemical and thermal stability of zeolitic imidazolate frameworks, Proc. Natl. Acad. Sci. U. S. A., 103 (2006) [15] S. Hermes, F. Schröder, R. Chelmowski, C. Wöll, R.A. Fischer, Selective nucleation and growth of metal organic open framework thin films on patterned COOH/CF3-terminated selfassembled monolayers on Au(111), J. Am. Chem. Soc., 127 (2005) [16] E. Biemmi, C. Scherb, T. Bein, Oriented growth of the metal organic framework Cu3 (BTC) 2 (H2O)3.H2O tunable with functionalized self-assembled monolayers, J. Am. Chem. Soc., 129 (2007) [17] M. Arnold, P. Kortunov, D.J. Jones, Y. Nedellec, J. Kärger, J. Caro, Oriented crystallisation on supports and anisotropic mass transport of the metal organic framework manganese formate, Eur. J. Inorg. Chem., 2007 (2007) [18] J. Gascon, S. Aguado, F. Kapteijn, Manufacture of dense coatings of Cu3(BTC)2 (HKUST-1) on α-alumina, Microporous Mesoporous Mater., 113 (2008) [19] J. Yao, D. Dong, D. Li, L. He, G. Xu, H. Wang, Contra-diffusion synthesis of ZIF-8 films on a polymer substrate, Chem. Commun., 47 (2011) [20] O. Shekhah, R. Swaidan, Y. Belmabkhout, M. du Plessis, T. Jacobs, L.J. Barbour, I. Pinnau, M. Eddaoudi, The liquid phase epitaxy approach for the successful construction of ultra-thin and defect-free ZIF-8 membranes: pure and mixed gas transport study, Chem. Commun., 50 (2014) [21] M.N. Shah, M.A. Gonzalez, M.C. McCarthy, H.-K. Jeong, An unconventional rapid synthesis of high performance metal organic framework membranes, Langmuir, 29 (2013) [22] D. Nagaraju, D.G. Bhagat, R. Banerjee, U.K. Kharul, In situ growth of metal-organic frameworks on a porous ultrafiltration membrane for gas separation, J. Mater. Chem. A, 1 (2013) [23] A.J. Brown, N.A. Brunelli, K. Eum, F. Rashidi, J. Johnson, W.J. Koros, C.W. Jones, S. Nair, Interfacial microfluidic processing of metal-organic framework hollow fiber membranes, Science, 345 (2014)

30 Chapter 1 Introduction [24] O.M. Yaghi, M. O'Keeffe, N.W. Ockwig, H.K. Chae, M. Eddaoudi, J. Kim, Reticular synthesis and the design of new materials, Nature, 423 (2003) [25] K. Yamamoto, Y. Nohara, Y. Domon, Y. Takahashi, Y. Sakata, J. Plévert, T. Tatsumi, Organic inorganic hybrid zeolites with framework organic groups, Chem. Mater., 17 (2005) [26] B. Chen, Z. Yang, Y. Zhu, Y. Xia, Zeolitic imidazolate framework materials: recent progress in synthesis and applications, J. Mater. Chem. A, 2 (2014) [27] N. Rangnekar, N. Mittal, B. Elyassi, J. Caro, M. Tsapatsis, Zeolite membranes a review and comparison with MOFs, Chem. Soc. Rev., 44 (2015) [28] J. Yao, H. Wang, Zeolitic imidazolate framework composite membranes and thin films: synthesis and applications, Chem. Soc. Rev., 43 (2014) [29] V.M.A. Melgar, J. Kim, M.R. Othman, Zeolitic imidazolate framework membranes for gas separation: A review of synthesis methods and gas separation performance, J. Ind. Eng. Chem., 28 (2015) [30] Y. Liu, N. Wang, J.H. Pan, F. Steinbach, J.r. Caro, In situ synthesis of MOF membranes on ZnAl-CO3 LDH buffer layer-modified substrates, J. Am. Chem. Soc., 136 (2014) [31] H.T. Kwon, H.-K. Jeong, In situ synthesis of thin zeolitic imidazolate framework ZIF-8 membranes exhibiting exceptionally high propylene/propane separation, J. Am. Chem. Soc., 135 (2013) [32] F. Cacho-Bailo, B. Seoane, C. Téllez, J. Coronas, ZIF-8 continuous membrane on porous polysulfone for hydrogen separation, J. Membr. Sci., 464 (2014) [33] W. Li, Q. Meng, X. Li, C. Zhang, Z. Fan, G. Zhang, Non-activation ZnO array as a buffering layer to fabricate strongly adhesive metal organic framework/pvdf hollow fiber membranes, Chem. Commun., 50 (2014) [34] L. Ge, W. Zhou, A. Du, Z. Zhu, Porous polyethersulfone-supported zeolitic imidazolate framework membranes for hydrogen separation, J. Phys. Chem. C, 116 (2012)

31 Chapter 1 Introduction [35] R.M.A. Domingues, S. Chiera, P. Gershovich, A. Motta, R.L. Reis, M.E. Gomes, Enhancing the biomechanical performance of anisotropic nanofibrous scaffolds in tendon tissue engineering: reinforcement with cellulose nanocrystals, Adv. Healthcare Mater., 5 (2016)

32 Chapter 2 Literature Review Literature Review Overview Membranes can be categorized according to different viewpoints. The simplest classification is by nature, i.e. synthetic or biological membranes. Membrane in this thesis only refers to synthetic membranes, excluding all biological structures. Based on the type of materials, synthetic membranes can be divided into inorganic (e.g. ceramic, silica, metals and oxides) membranes, organic (polymeric or liquid) and composite membranes. Polymeric membranes, owing to their relative ease of processing, low cost and large area per volume, have been developed for a vast variety of industrial applications like ultrafiltration (UF), reverse osmosis (RO) and gas separation. Inorganic membranes compared to organic counterparts have the advantages of high chemical and thermal stability, ease of cleaning when fouling, and well-defined, stable pore structure. Composite membranes, herein defined as combination of different materials (e.g. organic/inorganic and organic/organic) in the form of membranes, have been explored in the recent years. Owing to the synergistic effects of the materials, composite membranes have the potential for solving the trade-off between permeability and selectivity in conventional membranes [1, 2]. In this dissertation, polymeric and inorganic membranes were selected as the base substrate and metal organic frameworks (MOFs) were selected as active layer to prepare composite membranes, aiming to attain high performance membranes for gas separation. Another classification of membranes is in terms of morphology or structure - for example solid synthetic membranes can be divided into two types of membrane structures, symmetric and asymmetric (anisotropic) membranes. Figure 2-1 schematically represents the various types of 13

33 Chapter 2 Literature Review membranes. A membrane structure may also fit more than one of these classes. For instance, a membrane may have electrical charges and be microporous with an asymmetrical structure. In terms of configurations, membranes can be categorized into hollow fiber membranes, spiral membranes, tubular membranes and flat sheet membranes [3]. Figure 2-1. Schematic representation of various types of membranes [4]. Polymeric Membranes Polymeric membranes, owing to their relative ease of processing, low cost and large area per volume, have been developed for a vast variety of industrial applications like ultrafiltration (UF), reverse osmosis (RO) and gas separation. By the influential discovery of Loeb and Sourirajan for making highly permeable asymmetric RO membranes, membrane-based separation transformed from a research laboratory to the industrial applications [5]. The technique involves precipitation 14

34 Chapter 2 Literature Review of a casting solution, consisting of one or more polymers in a proper solvent/solvents, by immersion in a non-solvent (water) bath [6]. The membrane introduced by these workers was the first integrally skinned asymmetric cellulose acetate [7]. The water flux of such membranes in RO mode was 10 folds more than that of any other available membrane, which made RO applications feasible. Subsequently, other viable processes for microfiltration, ultrafiltration and electrodialysis were all established. Early gas separation membranes were also adapted from the Loeb and Sourirajan phase separation approach [8]. Although a large number of polymers have been studied to date as potential membrane materials, only a handful of polymers, as presented in Table 2-1, have found actual application in commercially viable processes [9, 10]. Properties such as cross-linking, functional groups, glass transition temperature and degree of crystallinity result in membranes with different properties and applicability. A robust separation membrane renders mechanical, thermal, and chemical stability, good film-forming properties, absence of micro-defects and aging in selective layer (reduction of flux in time). It stands to reason that the most of the technically used membranes (including support membranes for hybrid membranes) are, and will continue to be, made from organic polymers. This is mainly due to the inherently inexpensive and reproducible nature of polymer membrane synthesis, as well as offering an efficient separation in a range of systems. From desalination to biomedical applications, polymer membranes have demonstrated to be not only valuable and efficient, but also, as in the case of membranes utilized for hemodialysis, outperform their biological counterparts [11]. 15

35 Chapter 2 Literature Review Table 2-1. Primary membrane companies and polymeric membrane materials [9]. In spite of all of the usefulness polymer membranes serve, some hurdles still exist for their applications in a number of relevant gas separation. The most widely documented is known as Robeson s upper bound [12]. Due to the trade-off relationship between permeability and selectivity, polymer membranes generally suffer an upper bound limitation. This upper bound suggests that polymers with large segmental mobility and high sorption exhibit high permeability but a low selectivity, and vice versa [11]. The severity of this effect was found to be proportionally correlated to an increase in the difference between the kinetic diameters of the two permeating gas molecules in the binary mixture. This is graphically illustrated in Figure 2-2 for binary system from the list of H2, He, CH4, N2, O2, and CO2. 16

36 Chapter 2 Literature Review Figure 2-2. Relationship between the slope (n) of the upper bound and the difference between the kinetic diameters of the gas pairs [12]. Robeson s trade-off curve, however, is not the only prevention to the widespread polymer membrane usage. Another major hindrance to the commercialization of polymeric membranes is physical aging, which is instability of their permeability over time. Polymer aging, though still not fully understood, it possibly occurs as a result of molecular rearrangements, depending on chain mobility [13]. Plasticization is another major phenomenon that has extensive practical effects in membrane based gas separation. Plasticization is also based upon molecular chain rearrangements of a polymer in the presence of highly condensable molecules such as CO2. High pressure performance data for polymer membranes are typically anticipated from low pressure experiments. However, in the presence of highly condensable CO2, data achieved by the extrapolation to relevant high pressure applications are uncertain [14]. The plasticizing effect of CO2 causes the experimental deviation from theoretical extrapolations. The plasticizing effect results in a 17

37 Chapter 2 Literature Review significant loss in selectivity, as shown in Figure 2-3 the CO2/CH4 separation factor (mixed-gas selectivity) in a triptycene-based ladder polymer (TPIM-1) dropped 60% in the mixed gas system when CO2 partial pressure increases from 2 bar to 10 bar [13]. Figure 2-3. Loss of selectivity occurred in polymeric membranes with increasing partial pressures of CO2 [13]. Open points, pure gas. Closed points, mixed gas. Despite their inherent drawbacks, polymer membranes will undoubtedly be used to separate gas mixtures, but in the longer term, novel membranes made from other materials such as ceramics or organic/inorganic composites are likely to be required [10]. Inorganic Membranes Since the prediction of the upper limit of performance (the upper bound) for polymer membranes for gas separation in the early 1990s, only a few new materials could have separations above this upper limit. The required performance level for most of practical applications is above this limit of performance. 18

38 Chapter 2 Literature Review The efficiency of polymer membranes declines over time due to thermal instability, chemical degradation, compaction and fouling. As a result of this inadequate thermal stability and vulnerability to abrasion and also chemical attack, polymer membranes are not viable in separation processes where high temperature reactive gases are encountered. All these weaknesses associated with polymeric membranes triggered a shift of interest to inorganic membranes. Owing to well-known chemical and thermal stabilities and usually having higher gas permeability as compared to polymer membranes, inorganic membranes are increasingly being investigated for the separation of gas mixtures [15-17]. Inorganic membranes can be basically divided into porous and dense (non-porous) membranes. Ceramic membranes, such as silica, alumina, glass, and titania and porous metals, such as silver and stainless steel are examples of commercially available porous inorganic membranes. These membranes usually exhibits high gas flux, but low selectivity [18]. Dense inorganic membranes possess high selectivity in separating certain gases. For instance palladium (Pd)-metal based membranes are highly hydrogen selective and their alloys have been widely explored as potential membrane materials [19-22]. Recently, various techniques have been attempted to combine the high flux characteristics of porous membranes with outstanding selectivities of the dense membranes by supporting dense membranes on porous supports [23-25]. In addition, along with novel materials, other new fabrication techniques are being devised and developed to make thinner selective layers and/or narrower pore-sized membranes [26]. However, given the nature of this thesis, not all permutations of the inorganic membranes are related and necessary to discuss. The main focus herein, therefore, will be on microporous crystalline alumina-silicate membranes known as zeolites. 19

39 Chapter 2 Literature Review Zeolite membranes have open frameworks made of corner-shared TO4 (T = Al, Si, P) tetrahedra with well-defined structures and channels of molecular dimensions. Figure 2-4 illustrates chemical structure and primary building unit of zeolite structure. The specific geometry of zeolite pores enables discriminating guest molecules based on their size and/or shape. This feature results in employing zeolites as molecular sieves [27]. The formation of tetrahedral metallic (T-atoms)-oxygen clusters result in the defining feature of zeolites; their crystalline, rigid, and porous framework. Zeolite frameworks can be generally classified into ultra large, large, medium, and small pore materials. The number of tetrahedrons that makes the rings determine the pore size. Ultra large pore frameworks have 14-, 18- or 20-membered rings, large structures have 12- membered rings, medium pore zeolites have 10-membered rings, and small ones have 6-, 8-, or 9- membered rings [29]. A representative illustration of some zeolite structures can be seen in Figure 2-5. To date, more than 40 unique natural types of zeolites from volcanic sources, and at least 150 different synthetic types are available. However, the commonly used zeolites for purification and gas separation applications are limited to a handful of synthetic ones [30]. Figure 2-4. a) Chemical structure of zeolite, b) Primary building unit of zeolite structure [28]. 20

40 Chapter 2 Literature Review Figure 2-5. Silicon atoms are positioned at the intersections and linked by lines. (a) Sodalite cage; (b) zeolite A, the sodalite cages are connected to each other by double 4-membered rings and form an α-cage indicated by circle; (c) zeolite Y, the sodalite cages are linked by double 6-membered rings and organized as in the diamond framework [27]. Zeolitic membranes can be synthesized in pure phase (self-supported or symmetric membranes) or on a variety of supports (supported or asymmetric membranes) [31]. Generally, the preparation of self-supported symmetric zeolitic membranes introduces several drawbacks related to the mechanical stability, dimension, and to the lack of uniform thickness. Therefore, as opposed to most polymeric membranes, they are typically synthesized on either momentary support (to be removed after the synthesis) such as Teflon sleeve and silver plate, or permanent (to form zeolite composite membranes) such as active silica and porous ceramics [29]. In addition to their versatility, these microporous materials are also characterized by their outstanding thermal and hydrothermal stability. Khodakov et al. [32] reported that zeolite types A, X and Y were stable up to about 965 C, and more recently zeolite Y membranes was found to remain unchanged in its pore volume following 4 hours in steam at 788 C [33]. These outstanding thermal and chemical stabilities are not expected when using polymer membranes and perfectly demonstrate why these materials are necessary compliment to the polymer membranes. However, zeolite membranes are not without their own set of challenges. As mentioned before, due to the 21

41 Chapter 2 Literature Review finite number of membered rings by which zeolitic structures can be formed, there are zeolites with a limited and discontinuous array of available pore sizes. Therefore, zeolites of right pore sizes may not be always available for separating gas mixtures of certain sizes [34]. The effective pore sizes of some common zeolites and the kinetic diameters of several gases are shown in Figure 2-6. Apart from the limited pore size and its chemical tailorability, the major disadvantage of zeolite membranes is the high cost of production that hinders their wider applications in gas separations. Zeolite membranes have also proven very difficult to make free of defects, and this is more pronounced as the membrane size increases, while with the polymer membranes fabrication is easy and reproducibility is very high. Metal organic frameworks (MOFs), a rather new type of hybrid materials comprising of inorganic and organic moieties in solid crystalline lattices, have the potential to overcome some of the issues facing materials for membranes in gas separation. Metal organic Frameworks (MOFs) Metal organic frameworks (MOFs) or porous coordination polymers (PCPs), are hybrid porous crystalline materials formed by the coordination of metal ions/clusters and organic ligands. MOF materials are relatively new and the systematic studies of their synthesis and applications only initiated about two decades ago. With the pioneering work carried out by Hoskins [36, 37], Zaworotko [38], Moore [39, 40], and Yaghi [41] in early 1990s, these materials were soon known to be capable of integrating with designed structural, electrical, catalytic, optical, and magnetic properties by an appropriate selection of metal ions and organic ligands [42]. Many scientists, subsequently, joined this dynamic field and various synthetic approaches have been devised and developed to form MOFs with different pore sizes, crystal structures and surface chemistry [43, 44]. 22

42 Chapter 2 Literature Review Figure 2-6. Comparison of the effective pore sizes of various zeolites and the kinetic diameters of common gas molecules [35]. Given the broad range of metal ions and linkers available, MOFs usually have a number of unique features, such as exceptional large surface area, unusual adsorption affinities, tunable pore sizes, structure diversity and facilely chemical tailorability [45]. These features make MOFs very attractive for applications beyond the conventional areas of porous materials including molecule storage and adsorption [46, 47], catalysis [48, 49], delivery [50] and separation [51, 52]. Similar to the zeolites, MOFs can also be utilized for membrane constructions. However, due to the relatively poor mechanical strength (brittleness) [53], MOFs are typically formed on the support to attain continuous membranes or used as filler to obtain mixed matrix membranes (MMMs) [54, 55]. Different from the zeolitic membranes, where their pore structure can be only available (by removing the surfactants) after a sintering process at elevated temperatures, the activation process of MOF membranes is conducted at fairly lower temperatures. Furthermore, because of the organic 23

43 Chapter 2 Literature Review nature of the linkers in MOF materials they can interact with the polymer support and create a stronger MOF-to-support adhesion, which in general enhances the selectivity [56] MOF Materials and Fabrication Over 20,000 different MOFs with typically high surface area values, range from 1000 to 10,000 m 2 g -1, have been reported and studied in the past decade [50]. However, like zeolites, only a handful of MOF types are used in membrane separation. This is because a number of considerations such as fabrication, activation, pore size, diffusivity, and solubility of the MOF membranes should be taken into account when choosing a MOF type. In this section, we only introduce MOFs that are widely studied for MOF-based membranes. The crystalline structure of these MOFs are shown in Figure 2-7. Figure 2-7. Crystalline structure for three commonly used MOFs for separations. Yaghi s group has first reported the synthesis of zeolitic imidazolate frameworks (ZIFs) by copolymerization of Zn (II) or Co (II) with imidazolate-type links [57]. Several synthesis methods, afterwards, have been developed to obtain ZIF-8 in a large quantity, implementing simpler and greener strategies [58-60]. They closely resemble the topologies of zeolites, due to the M-Im-M (M = Zn, Co) bond angle of 145, which is close to the T-O-T (T = Al, Si, P) angle in zeolites. One of the most common and extensively studied member of ZIFs is ZIF-8, which consist of zinc metal ions linked by 2-methylimidazole (Hmim). It has a sodalite (SOD) topology with a large 24

44 Chapter 2 Literature Review cavity of 11.6 Å (Figure 2-8) accessible through the theoretical small aperture (six-membered ring window) of 3.4 Å (Figure 2-8) [57]. Owing to its permanent porosity, high thermal and chemical stability, and excellent solvent resistance, ZIF-8 has attracted intensive interest in materials science and chemistry [61]. Due to its small pore size (3.4 Å), ZIF-8 membranes gained considerable interest for hydrogen separation. Zhang et al. [62] recently found that due to the flexibility of the organic linker the effective pore size of ZIF-8 is in fact in the range of 4.0 to 4.2 Å, which is considerably bigger than the XRD-derived value (3.4 Å). This opened up new opportunities for ZIF-8 for separations that could not be economically accomplished by traditional microporous materials such as synthetic zeolites. Very recently Li et al. [63] reported that propylene ( 4 Å) diffuses in ZIF times faster than propane ( 4.3 Å) due to its unexpected molecular sieving effects and consequently Pan et al. [64] synthesized ZIF-8 membranes capable of separating propylene from propane mixtures effectively. In addition, the surface structure and hydrophobic pore of ZIF-8 likely repels water molecules and prevent the attack of ZnN4 units and decomposition of the framework [57, 65], making ZIF-8 moisture resistant and a potential candidate for separating humidified gas mixtures. Figure 2-8. Section of the crystal packing diagram of ZIF-8 [57]. ZIF-L with a leaf-shaped morphology is comprised of zinc nitrate and Hmim, i.e. same building blocks as ZIF-8 (Figure 2-9, Figure 2-10). It has a two-dimensional layered framework with cushion-shaped cavities between layers with a dimension of 9.4 Å 7.0 Å 5.3 Å [66]. The 25

45 Chapter 2 Literature Review two-dimensional layers in ZIF-L are bridged by hydrogen bonds, unlike ZIF-8 in which the two neighboring sod layers are bridged by Hmim (Figure 9). The layers are parallel to each other, 3.97 Å apart and stacked along the c direction [66] shows the crystal structures of ZIF-L and ZIF-8. Figure 2-9. Wire-frame model and ball-and-stick model of the crystal structure of (A) ZIF-L and (B) ZIF-8 [67]. Figure SEM images of (a) ZIF-L nanoflakes, (b) ZIF-8 nanoparticles [67]. Attributed to the strong interactions between Hmim and CO2 molecules and the unique cushion-shaped cavities, ZIF-L exhibits superior CO2 adsorption capacity (0.94 mmol g 1 ) and 26

46 Chapter 2 Literature Review CO2/CH4 adsorption selectivity (7.2) as compared with other ZIFs with large cages. ZIF-L crystals with its unique cushion-shaped cavity and two-dimensional leaf-like morphology can be useful as a kind of potential gas separation membrane/nanocomposite membrane materials [68, 69]. Wang et al. [70] prepared ZIF-L membranes with two different orientations along their layered porous structure, i.e., b-oriented and c-oriented membranes, and investigated their gas separation properties. The c-oriented membrane due to the unique pore systems was found more favorable than the b-oriented membrane respecting selectivity for H2/CO2 and H2/N2. UiO-66 (UiO stands for University of Oslo), characterized by very high surface area and with an unprecedented thermal stability, is the first generation of zirconium-based MOF that was recently presented by Lillerud et al. [44] This robust, 3-dimensional porous structure is formed by connecting hexanuclear zirconium clusters, as secondary building units (SBUs), with a commonly available bridging ligand (1, 4-benzenedicarboxylic acid) (BDC) (Figure 2-11). Its outstanding thermal and mechanical stability has been attributed to its high degree of strong coordination of Zr O metal moieties to the organic linkers [71]. Furthermore, revealed by repetitive hydration/dehydration tests, porous UiO-66 switches reversibly between its dehydroxylated and hydroxylated versions that further increase its thermal stability. The UiO-66 has a degradation temperature over 500 C and is resistance to several chemicals, and its crystalline structure remains unchanged even after exposure to tons of external pressure [44]. Considering all these outstanding features, therefore, it is expected that membranes constructed with UiO-66 would achieve various promising applications including molecular separations. As estimated from crystallographic data, the aperture size of UiO-66 is about 6.0 Å and thus capable of separating water molecules ( 2.8 Å) from hydrated ions ( Å) [72]. Additionally, due to the specific interaction between 27

47 Chapter 2 Literature Review hydroxylated Zr6 cluster in the framework and CO2, UiO-66 can preferentially adsorb CO2 over other gases [73]. Figure Demonstration of UiO-66: (a) secondary building units (SBUs), (b) BDC ligand; (c) crystal model, (d) a simplified form [74]; (e) SEM images of UiO-66 powders [73]. Although a number of pure MOF membranes, such as HKUST-1 [75], MOF-5 [76], ZIF-8 [77], MIL-53 [78], have been made and analyzed for gas separation, preparing continuous defectfree UiO-66 membranes on porous supports by direct solvothermal synthesis is more challenging due to the slow growth kinetics of this MOF [73]. UiO-66-type MOF thin films was initially grown on flat silicon substrates using solvothermal approach [79] and on electrodes by electrochemical deposition [80]. Very recently, Li and his co-workers successfully fabricated continuous UiO-66 polycrystalline membranes on alumina hollow fiber supports via an in situ solvothermal synthesis approach (Figure 2-12) [73]. The obtained UiO-66 membrane exhibited high separation performance for multivalent ion rejection (e.g., 98.0% for Mg 2+, 99.3% for Al 3+, and 86.3% for Ca 2+ ) based on size-exclusion mechanisms with moderate permeance of 0.14 L m 2 h 1 bar 1 and good permeability (0.28 L 28

48 Chapter 2 Literature Review m 2 h 1 bar 1 μm) in water desalination [73]. The membranes showed excellent recyclability due to exceptional chemical stability of the UiO-66 material, which can be reasonably promising for sea water desalination. Considering the wide-ranging of ligands used in UiO-66-type MOFs, it can be expected that more pure UiO-66 membranes to be synthesized in the near future [74]. Figure Schematic diagram of in-situ synthesis of pure UiO-66 membranes supported on porous hollow fibers [73]. Besides the ZIF-8, ZIF-L and UiO-66, the MIL-101, MIL-53(Al), CAU-1, CAU-1-NH2, Bio- MOF-14, Bio-MOF-13, Bio-MOF-1, ZIF-69, ZIF-22, ZIF-7, MOF-71, ZIF-78, ZIF-95, ZIF-90, IRMOF-3, CuBTC etc. have also been utilized for constructing the MOF-based membranes. The thermal and water stability of the employed MOFs are very essential for its applications. Some of MOFs (MOF-5, IRMOF-3, CuBTC, etc.) are not stable in water or moisturized environments and therefore these MOF membranes are not appropriate for use in humid or aqueous environments [81-83]. Although MIL series of MOFs exhibit the high water and thermal stability, the properties of the framework are changed upon the adsorption of water [84, 85]. ZIF series of MOFs, on the other hand, usually display an excellent water and thermal stability and therefore these MOF 29

49 Chapter 2 Literature Review membranes have found a wide separation applications, such as organic solvent nanofiltration, pervaporation, gas separation with steam, and dye water solution separation [86]. Figure Fabrication methods for MOF-based membranes Synthesis of MOF-based membranes Different methods have been reported for the fabrication of continuous MOF membranes on various substrates. Figure 2-13 summarizes the fabrication methods of MOF-based membranes. Considering the diffusion direction of the organic linker and metal ion during crystallization, the synthesis procedures can be divided into three general categories: hydrothermal or solvothermal method, contra-diffusion and interfacial synthesis method, and liquid phase epitaxy (layer by layer or step-by-step) method (Figure 2-14). In hydrothermal or solvothermal method, the substrate is directly immersed in a solution of organic linker and metal ion (growth solution), so the same diffusion direction of metal ion and linker is expected during the MOF film formation. This method can be subdivided into in situ growth and secondary growth (seeded growth). In in situ growth the MOF layer is fabricated on the substrate. Due to insufficient heterogeneous nucleation on substrate and poor adhesion between the MOF layer and the substrate, support surface modification is usually introduced to create the functional groups on the substrate that can be combined with metal ions or linkers. Hydrothermal or solvothermal treatment can then be applied to form MOF layer on the modified substrate. In 30

50 Chapter 2 Literature Review contrast, in the secondary growth procedure, the substrate is initially coated by a seeded layer before immersing into the precursor solution for hydrothermal or solvothermal treatment. The seeded layer can be accomplished by either physical absorption (rubbing or dip coating) or by the reaction between the substrate and precursor reactants. Compared with other methods, hydrothermal or solvothermal method demonstrates to be applicable in a wider range. However, it also has some drawbacks. The membrane synthesis process can be complex due to the employed seeding or modification for improving the heterogeneous nucleation density on the substrates. In addition, in this method a large amount of precursor solution is usually wasted due to the homogenous formation of the MOF crystals in bulk solution and it is difficult to be scaled up. Figure Scheme of the fabrication methods for continuous MOF membranes [87]. In contra diffusion method, the substrate physically separates metal ions and ligand molecules where they diffuse in opposite direction and crystallize at the interface. Since the crystallization takes place at the interface the method is known as a self-limiting growth. This is because the 31

51 Chapter 2 Literature Review diffusion of precursor molecules is faster at the remaining defects than the already formed layer, which is beneficial for a continuous, defect-free MOF film synthesis. This unique feature also helps the formation of a film with uniform thickness. Owing to its simplicity and effectiveness in the fabrication of ZIF films, this method has attracted many researchers attention, mainly in the field of gas separation via molecular sieving. Yao et al. [56] prepared continuous ZIF-8 membrane on a porous nylon with a moderate gas separation performance. Wasting the precursor solution is less pronounced in this method as the metal ion and ligand solution is separated and the excess solutions can be recycled. The method also can be employed in a membrane module directly and thus displaying a superior scalability [88]. However, only a few MOF membranes, such as ZIF-7 [89], ZIF-71 [90], ZIF-8 [56], and Cu-BTC [91], have been fabricated by this method. In MOF membrane synthesis by liquid phase epitaxy method [92, 93], the metal ion solution and organic ligand solution are separately prepared. The substrate is immersed in one precursor solution and after rinsing by the solvent, for removing the excess precursor, it is soaked into the other precursor solution for MOF crystallization. Due to the non-consecutive crystallization, precise control over the MOF film thickness is achievable, which is one of the outstanding advantage of this method. Although the method also offers a simple, mild and controllable synthesis approach, it is usually used for fabricating the MOF film rather than MOF membrane as the continuous, defect-free MOF layer is difficult to achieve [87]. For MOF-based MMMs [94], the MOF particles are dispersed into polymer solution to make the casting solution. The casting solution is then casted on the surfaces of a clean glass plate to form the free standing MMMs or on the substrate to form the supported MMMs. The supported MMMs usually possess higher permeability because they are usually thinner than their free standing counterparts. 32

52 Chapter 2 Literature Review Supported MOF membranes Owing to their relative ease of processing, low cost and large area per volume, polymeric materials are attractive to be used as support in fabricating MOF-based membranes especially for industrial separation application. Furthermore, due to the favourable chemical interaction between the polymer and the MOF s organic ligand, growth of MOF films on flexible polymeric substrates is principally achievable. Hatton and co-workers for the first time demonstrated the feasibility of growing MOF on polymer substrate [95]. They grew MIL-47 on the surface of polyacrylonitrile (PAN) using microwave assisted solvothermal synthesis. There are several challenges in the fabrication of polymer-supported MOF membranes. First, the polymer substrates should retain the thermal stability and good solvent resistance as the substrates in the synthesis of MOF membranes are usually soaked into the organic solvent, such as the methanol or DMF, at high temperature [96]. Second, the polymer materials usually swell in organic solvents and delamination or cracking of MOF layers can occur as a results of the shrinking of the polymers if the MOF layers are formed on the swelling polymer substrates [97]. Therefore, it is required to reduce the swelling of the polymer substrates as much as possible. Third, due to the relative poor mechanical stabilities of MOFs, the MOF layer tend to fall off from the elastic polymeric substrates when using it for separation. This is due to the low stability and shrinkage of the polymer under hydrostatic compression [96, 97]. However, several MOF membranes have been successfully fabricated on polymer substrates. The polymer substrates used for the construction of MOF membranes can be classified into flat polymer substrates and hollow fiber polymer substrates. Yao et al. [56] prepared for the first time a polymer-supported ZIF-8 membrane on the porous nylon membrane via contra-diffusion method (Figure 2-15). The employed nylon acts as the substrate for growing ZIF-8 film. This strategy creates a concentration gradient of the ligand and 33

53 Chapter 2 Literature Review metal ion near the surface of the substrate, which promotes the heterogeneous formation of ZIF-8 on the substrate [98]. ZIF-8 film was formed on both side of the substrate with different morphologies due to the different local molar ratio of Hmim and zinc nitrate. At the overall Hmim/zinc ion molar ratio of 8, because of slower diffusion of organic ligand in nylon substrate as compared to the metal ion, the Hmim/ zinc ion molar ratio at the zinc nitrate side of the substrate was close to zero, resulting in large ZIF-8 crystals of nm. While, on the ligand side the local molar ratio of Hmim/zinc ion should be larger than the overall designed molar ratio of 8, generating the ZIF-8 films made up of nanocrystals. ZIF-8 membranes prepared at room temperature for 72 h showed a moderate ideal selectivity of hydrogen over nitrogen (4.3) with a large hydrogen permeance of mol m 2 s 1 Pa 1. The aqueous solution synthesis of ZIF- 8 films on a porous nylon substrate was further conducted with Hmim/zinc ion stoichiometric ratio and the addition of ammonium hydroxide solution [98]. Figure (a) Diffusion cell for contra-diffusion preparation of ZIF-8 film and (b) the schematic synthesis of ZIF-8 films on the sides of the nylon substrate by contra-diffusion of Hmim and Zn 2+ through the porous nylon [56]. Nagaraju et al. [99] grew ZIF-8 and CuBTC on a porous asymmetric ultrafiltration polysulfone using in situ (direct) growth followed by the layer by layer deposition at room temperature (Figure 2-16). Among the prepared ZIF-8@PSF and CuBTC@PSF membranes, 34

54 Chapter 2 Literature Review CuBTC@PSF permeance dropped drastically after 7 crystallization cycles with an enhanced H2/CO2 and H2/C3H6 selectivity of around 7.2 and 5.7, respectively. Figure Synthesis illustration for the CuBTC@PSF membrane by in situ method and layer by layer crystal deposition [99]. More recently, Cacho-Bailo et al. [100] grew a 35 μm thick ZIF-8 layer on commercial porous polysulfone by an alternating synthesize procedure using solutions that produced nano- and micrometer-sized particles. A high excess of sodium formate (NaCOOH) was used in the synthesis solution as ligand-deprotonator. Adding different ligand/metal ratios enabled the formation of nano- or micro-sized ZIF-8 crystals. The prepared ZIF-8 membranes were able to separate H2/CH4 and H2/N2 mixtures with high separation factors of 10.5 and 12.4, respectively. These values, as 35

55 Chapter 2 Literature Review shown in Figure 2-17, were the highest among the previously reported selectivities for ZIF-8 membranes on polymer substrates. Figure H2/CH4 (a) and H2/N2 (b) separation factors as a function of H2 permeance for ZIF-8 membranes in [100] as compared to those in the literature. Although the in situ synthesis is a simple method that allows for simultaneous nucleation, deposition and crystal growth, it is not very effective in preparing continuous ZIF membranes due to limited heterogeneous nucleation sites on the substrate. Alternatively, Ge at al. [101] used secondary seeded growth to fabricate a continuous ZIF-8 film on an asymmetrically porous polyethersulfone substrate. The prepared 7.2 μm thick ZIF-8 layer showed good affinity with the PES substrate and displayed molecular sieving separation. At 333 K and 150 KPa, the H2 permeance reached about mol m 2 s 1 Pa 1 with the ideal separation factors of 10.8, 9.7, 10.7, and 9.9 for H2/O2, H2/Ar, H2/CH4, and H2/N2, respectively. Long-term hydrogen permeance and H2/N2 separation performance show the stable permeability of the derived membranes. Longterm hydrogen permeance as well as H2/N2 separation analysis showed the stable permeability for the prepared ZIF-8 membranes. 36

56 Chapter 2 Literature Review Secondary seeded growth has been shown to effectively induce controlled ZIF growth on the polymer support, but the resulting ZIF layer often suffers from weak adhesion to the support, leading to membrane delamination. To enhance the ZIF-to-substrate adhesion strength, Barankova et al. [102] employed a tailor-made porous polyetherimide/zinc oxide mixed-matrix as the substrate. A non-woven polyester material was first coated with a mixture solution of polyetherimide (PEI, Ultem 1000) and zinc oxide nanoparticles (Figure 2-18). After a polishing step, a combination of rubbing and dip-coating was used for placing ZIF-8 seeds on the PEI/ZnO substrate. Since the zinc ions in ZnO can act as a secondary source of metal for ZIF-8 formation, the PEI/ZnO mixed matrix was beneficiary for the fabrication of continuous ZIF-8 crystals with enhanced adhesion. After 36 hours of secondary growth at 45 C a dense ZIF-8 layer with a thickness of 1.5 μm was prepared. The ZIF-8 membranes were dried by applying solvent exchange technique to avoid formation of cracks. The achieved hydrogen permeation of mol m 2 s 1 Pa 1 was about 4 times higher than that reported for polyethersulfone-supported ZIF- 8 membrane due to its thinner ZIF layer (1.5 μm versus 7.2 μm in [101]), but two folds less than the one synthesized on a nylon substrateby the contra-diffusion method [56]. However, the membrane showed a relatively high H2/C3H8 ideal selectivity of 22.4 [102]. Figure Schematic diagram of constituting layers of a supported-zif-8 membrane [102]. 37

57 Chapter 2 Literature Review Hollow fiber polymer membranes are advantageous over flat sheets due to their large membrane surface area per unit volume [103]. These fibers are readily used to produce modules with large membrane areas exceeding 1000 m 2 m -3 of module volume [104]. Brown et al. [104] prepared a hollow fiber polymer-supported MOF membrane for the first time by growing ZIF-90 on poly (amide-imide) Torlon hollow fiber in a technologically scalable low-temperature synthesis procedure. Torlon was chosen as an appropriate polymeric substrate for separation applications due to its high pressure endurance (up to 2000 psia) with no plasticization, chemical resistance and ease of processing. A dense layer of uniform ZIF-90 seed crystals was deposited on the surface of the hollow fiber by dip-coating technique. A continuous 5 μm thick (Figure 2-19) polycrystalline ZIF-90 membranes were obtained at 65 C for 4 h by secondary growth. The membrane showed a CO2/CH4 selectivity of 1.5 which was interestingly lower than the CO2/N2 selectivity of 3.5. This behavior was attributed to the fact that ZIF-90 and other ZIF materials are known to have high CO2 adsorption capacities, and typically also adsorb CH4 more strongly than N2. Figure SEM images of: a) the ZIF-90 seed-layer, b) top view and c, d) cross section views of the polycrystalline ZIF-90 membrane after secondary growth [104]. 38

58 Chapter 2 Literature Review Figure a) Schematic representation of oriented synthesis of HKUST-1 crystals controlled by surface functionalization [75]. b) A simplified model of anchoring a typical MOF-5 building unit to a carboxylic acid-terminated self-assembled organic monolayers (SAM) [76]. To strengthen the adhesion between the support and MOF layer in inorganic-supported MOF membranes, support surface modification with organic ligands [105, 106] or organosilane molecules (Figure 2-20) have been widely applied [75, 76, 107, 108]. However, the process of the surface modification with linkers is generally time-consuming [87]. In order to eliminate this step and yet achieve a MOF membrane with excellent adhesion, Li et al. [96] reported a new method applying a non-activation ZnO array as a buffering layer on polyvinylidene fluoride (PVDF) hollow fibers (Figure 2-21). To produce the non-activation (NA) ZnO array, 2-methyl-imidazole, zinc nitrate hexahydrate and sodium formate were employed. After the successful growth of NA- ZnO array on the PVDF hollow fiber, the crack-free and uniform MOF (HKUST-1, ZIF-7 and ZIF-8) membranes were fabricated by directly immersing the NA-ZnO/ PVDF into the MOF precursor for crystallization. The prepared MOF/PVDF membranes possess excellent hollow fiber structures and displayed exceptional hydrogen permselectivity. For the ZIF-7/PVDF membrane, the ideal separation factors for H2/CO2 and H2/N2 were and 20.27, respectively, with high H2 permeance of mol s 1 m 2 Pa 1. Due to these properties the NA-ZnO array is an excellent buffering layer for fabricating MOF membranes, and the prepared ZIF/PVDF 39

59 Chapter 2 Literature Review membranes are potentially promising candidates for industrial hydrogen separation. Furthermore, the strong MOF-to-substrate adhesion was confirmed by ultrasonic treatment test in which no crystal was exfoliated from the hollow fiber after 60 min of sonication. Figure Scheme of the morphology and chemical structure of a MOF/PVDF membrane prepared by a non-activation method [96]. Though the flexibility of polymeric substrates can be advantageous over brittle inorganic materials, highly flexible polymers are not desirable for fabrication of polymer-supported MOF membranes. In other words, to enhance the performance of a MOF membrane supported on a polymeric substrate, the flexibility of the polymer substrate must be reduced to avoid MOF layer cracking. Polyacrylonitrile (PAN) hollow fibers is a good choice to meet this requirement (low flexibility) in addition to its commendable adhesion [97]. PAN with a chain of carbon connected to each other is a hard polymer. More importantly, it has abundant nitrile groups ( CN) which can crosslink together upon heating and dramatically increase the stiffness and mechanical and chemical stability of the polymer [109, 110]. Li et al. prepared a continuous well intergrown ZIF- 8 and Cu3 (BTC)2 membrane on PAN hollow fibers by solvothermal treatment. Dehydrogenation, cyclization and crosslinking reactions (Figure 2-22) significantly improved the support stiffness and its compression strength. The prepared Cu3 (BTC)2 PAN composite membrane achieved a 40

60 Chapter 2 Literature Review high hydrogen permeance of mol m 2 s 1 Pa 1 and good H2/CO2 separation factor of 7.14 for binary mixture. Figure (a) Optical images of the (1) pristine PAN hollow fiber, (2) hydrolyzed PAN hollow fiber and (3) Cu3 (BTC)2 PAN hollow fiber membrane. (b),(c) proposed involved chemical reactions of the PAN hollow fiber [97]. Ge et al. [111] imbedded zeolite crystals in the polymer hollow fiber to make polymer zeolite mixed-matrix hollow fiber membranes and used as substrate for the growth of a zeolite layer (Figure 2-23a). The imbedded zeolite crystals act as seeds for the growth of zeolite membrane and also anchor the zeolite layer to the polymer support to improve the zeolite membrane adhesion. A similar strategy was applied to fabricate MOF membranes on polymer hollow fiber substrates with strong adhesion and good separation performance (Figure 2-23b) [112]. The MOF crystals were first blended with polydimethylsiloxane (PDMS) in the presence of catalyst and cross-linking agent. The resultant solution then was deposited on the surface of PSf hollow fiber by drop coating. The obtained polymer-mof composite provides seeds for MOF growth, increase the adhesion of the MOF layer, enhances the gas separation performance, and reduces the mass transfer resistance compared with MMMs or polymer membranes. Second, the PSf hollow fiber with the PDMS/MOF layer was placed into the Teflon autoclave filled with the MOF precursor to crystallize. The 41

61 Chapter 2 Literature Review obtained 20 μm thick trinity Cu3(BTC)2 membrane exhibited an excellent separation performance with hydrogen permeance of mol m 2 s 1 Pa 1 and N2/CO2 and H2/CO2 ideal selectivity of and 7.23, respectively. The membrane also maintained excellent performance under different pressures (from 1 bar to 3 bar), helpful for industrial applications [112]. Figure Schematic illustration of (a) Zeolite Membranes on polymer zeolite MM hollow fiber supports [111] and (b) trinity MOF membranes preparation [112]. Reducing the growth temperature and minimizing the use of organic solvents are the two critical factors for the scalable fabrication of continuous MOF layer on polymer hollow fibers. Mao et al. [113] developed a pressure-assisted growth strategy to fabricate compact HKUST-1 films on PVDF hollow fiber in 40 min at room temperature (Figure 2-24). First, a mesoporous copper hydroxide nanostrands (CHNs) layer was deposited on PVDF hollow fiber by filtering, to serve as the copper source. Then, without turning off the pump, the PVDF substrate with CHN layer was immersed into a linker ethanol/water solution (volume ratio of 1:1) for MOF layer growth. The fabricated HKUST-1/PVDF membrane showed a good separation performance with 42

62 Chapter 2 Literature Review high hydrogen permeance of mol m 2 s 1 Pa 1 and H2/CH4, H2/N2, H2/CO2 separation factors of 5.4, 6.5, 8.1, respectively [113]. Figure Scheme of pressure-assisted preparation of HKUST-1 layer on the surface of PVDF hollow fibers substrate [113]. Contra-diffusion synthesis has also been applied for the scale-up synthesis of the hollow fiber polymer-supported MOF membranes. Based on contra-diffusion synthesis concept, Brown et al. [88] recently reported a methodology, interfacial microfluidic membrane processing (IMMP), for the scalable fabrication of molecular sieving MOF membranes on polymeric hollow fibers. The method also enable the control over the ZIF membrane position by employing an oil/ water system, in which crystals grow at the interfaces between the two immiscible solvents. Different from the 43

63 Chapter 2 Literature Review previously reported methods, the MOF precursor solution flew within the hollow fiber bore. Therefore, the MOF membrane can be directly produced on the hollow fiber substrate that was installed into the module and excess reactants can be readily recycled into the hollow fiber bore. The high quality ZIF-8 membranes made in water-octanol system by IMMP method could achieve H2/C3H8 ideal selectivity of more than 600 with the permeance of H2 around mol m 2 s 1 Pa 1 and demonstrated an excellent stability within 35 days of operation. Similarly, Biswal et al. [103] developed a simple and scalable room temperature interfacial approach for growing ZIF-8 and CuBTC. The method allowed the MOF growth on either the outer or inner side of a polybenzimidazole (PBI) based hollow fiber support surface in a controlled manner. An immiscible pair of low boiling solvents (isobutyl alcohol (IBA), CHCl3 and water) was employed for the fabrication of MOF@membrane composite. As compared to other high boiling solvents such as octanol, which is usually used for the interfacial fabrication of MOFs [88, 114], the employed solvents have the benefit of the easy exclusion of these solvents from the MOFs and membrane with a mild activation procedure MOF-based mixed-matrix membrane MOF-based mixed matrix membranes (MMMs) consisting of a dispersed MOF nanocrystals in a polymer matrix are another important family of MOF membranes. Unlike supported MOF membranes, MMMs need not form the continuous MOF layer. Figure 2-25 schematically compares supported MOF membranes with MOF-based MMMs. MMMs have increasingly attracted the attention of researchers over the last decades because they usually combine the advantages of polymers and MOFs [ ]. 44

64 Chapter 2 Literature Review Figure Schematic diagram of a typical MOF membrane and MOF based MMM. A good dispersion of the filler with an excellent interaction with the polymer chains (composite interface) is extremely important for the optimum MMM performance [119]. Due to the organic linker of MOFs, a good interaction between MOF and polymer matrix is usually achieved. This can result in a good interfacial contact between MOF and the polymer and reduce the micro-gaps, which is usually an issue in inorganic/polymer MMMs. Therefore a higher optimal filler loading is achievable in MOF-based MMMs compared with inorganic (e.g. zeolite or silica)- based MMMs, which usually have the optimal loading below 10 wt. % [120]. However, the void spaces between the polymer and filler still exist for some MOF-based MMMs. To overcome this challenge and minimize the interfacial defects and also strengthen the adhesion between the polymer phase and dispersion phase, enrichment of interfacial interaction between continuous phase and inorganic phase of the MOF, inhibition of particle agglomeration and proper choice of MOF/polymer pair are important aspects to take into account [120, 121]. Additionally, the separation performance of gas mixture is usually affected by the diffusivity and solubility of the gas components in MMMs [122]. In most of the polymer membranes the diffusivity selectivity term favours permeation of the smaller gas molecule, H2, while the solubility selectivity term favours permeation of the more condensable component, CO2 [123]. This causes the challenge of 45

65 Chapter 2 Literature Review separating H2/CO2 mixtures using polymer membranes, which remains undissolved in the case of MMMs [124]. The gas permeability in MMMs is relatively low as compared with continuous supported MOF membranes. Since MMMs possess a lot of advantages such as flexibility, easy of processing, relatively large permeability (as compared to polymer membranes) and low cost, many recent studies have focused on the development of the MOF-based MMMs. In the following section, we will discuss the MOF-based MMMs for gas separation. Ordonez et al. [125] prepared ZIF-8/Matrimid MMMs with loadings up to 80 wt. %.These loadings, as mentioned, are much greater than the usual loadings attained with selected zeolite materials. To examine the quality of the ZIF-8/Matrimid MMMs, gas permeation experiments were carried out for C3H8, CH4, N2, O2, CO2, H2 and gas mixtures of CO2/CH4 and H2/CO2. The permeability initially increased with increasing the ZIF-8 loading, however at loadings above 40 wt. %, the permeability of all gases decreased. It is known that nanoparticles can interrupt chain packing in glassy polymers, leading to an increase in the polymer free volume and its permeability [126, 127]. For the work done by Ordonez et al., the addition of ZIF-8 nanocrystals to the Matrimid polymer matrix results in an increase in the polymer chain-to-chain distance, creating more polymer free volume. Loadings above 50 wt. % increased selectivities which demonstrated the influence of the ZIF additive and a transition from a polymer-governed to a ZIF-8-governed gas transport process, where at higher loadings the ZIF-8 sieving effect becomes dominant. Sonication is commonly used for a homogenous dispersion of MOF particles in the polymer matrix. It was shown that sonication causes substantial changes in the structure, shape and size distribution of ZIF-8 nanocrystals dispersed in an organic solvent in membrane processing [128]. 46

66 Chapter 2 Literature Review However, data obtained from powder X-ray diffraction and nitrogen physisorption indicated that losses in microporosity and crystallinity are minor. In a parallel study, 60 nm ZIF-8 nanocrystals with surface area of m 2 g 1 were added into a Matrimid polymer via the solution mixing method [129]. An excellent dispersion of nanocrystals, up to 30 wt. % loading, was obtained with a good interfacial interaction between nanocrystals and polymer matrix, as confirmed by SEM and gas sorption studies. Single gas analysis were conducted for CH4, N2, O2, CO2 and H2. The gas permeability considerably increased by increasing the ZIF-8 loading while the selectivity remained relatively unchanged as compared to the pristine polymer membrane. The porous nature of ZIF-8 was also utilized to enhance the permeability of the polybenzimidazole (PBI) membrane [124, 130, 131]. The addition of ZIF-8 into the polymer matrix resulted in a hundred times enhancement in hydrogen permeability of the resultant MMMs, which reached about Barrer without any significant reduction in H2/CO2 selectivity (12.3) compared to the pure polymer membrane [124]. ZIF-8 was also used in combination with zeolite in polysulfone (PSf) membranes for the separation of CO2/N2, CO2/CH4, O2/N2, and H2/CH4 gas mixtures [132]. However, it was shown that the ZIF-8 and silicalite-1 combination, ZIF-8/S1C-PSF MMM, could not outperform either S1C-PSF or ZIF-8- PSF MMMs towards CO2/N2 and CO2/CH4 gas mixtures. This was attributed to the relatively large S1C crystals which might not be intercalated between small ZIF-8 nanocrystals (ca. 100 nm). Silicalite-1 also has a lower affinity towards CO2. Conclusion and perspectives In summary, membranes provide an eco-friendly and energy efficient alternative to traditional separation protocols. MOF materials, due to their tunable pore size and chemistry, are promising 47

67 Chapter 2 Literature Review candidates for constructing membranes capable of separating small gases (e.g. hydrogen) from other larger gases (e.g. N2, CH4) and also gas mixtures (e.g., C3H6/C3H8) that other materials such as polymer-based membranes only show low separation performances. ZIFs are particularly attractive for membrane-based gas separations due to their high thermal and chemical stability. Despite several synthesis method being developed, the highly reproducible fabrication of MOF membranes is particularly challenging owing to the difficulty of directing nucleation and crystal growth onto the surface and the tendency for growth into unfavourably large crystals and thick MOF layers. Various synthesis methods targeted at addressing these challenges have been discussed. To date, a number of metal organic frameworks with diverse structures have been synthesized that can be assembled into membranes for many commercially challenging separation applications. Hence, intensive research efforts for developing facile MOF membrane fabrication will continue to be critical and many new MOF membranes are expected to be reported. It is also important to note that due to similar mechanisms of crystallizing of MOFs and zeolites, the key techniques for fabrication zeolite membranes have been commonly modified to fabricate MOF membranes. However, considering that the coordination chemistry of MOFs is basically different from the covalent chemistry of zeolites, novel fabrication methods are expected to afford utilizing distinctive features of MOFs for not only traditional gas separations but also for the new area of enantioselective and chiral separations. The new processing approaches require advantages of being rapid, reproducible, scalable, and economically and environmentally viable and at the same time produce high quality of MOF membranes. 48

68 Chapter 2 Literature Review References [1] P. Goh, B. Ng, W. Lau, A. Ismail, Inorganic nanomaterials in polymeric ultrafiltration membranes for water treatment, Sep. Purif. Rev., 44 (2015) [2] F. Peng, L. Lu, H. Sun, Y. Wang, J. Liu, Z. Jiang, Hybrid organic-inorganic membrane: solving the trade-off between permeability and selectivity, Chem. Mater., 17 (2005) [3] L.M. Camacho, L. Dumée, J. Zhang, J.-d. Li, M. Duke, J. Gomez, S. Gray, Advances in membrane distillation for water desalination and purification applications, Water, 5 (2013) [4] R.W. Baker, Membrane Technology and Applications, 3 ed., John Wiley & Sons, Ltd,, [5] L. Sidney, S. Srinivasa, Sea water demineralization by means of an osmotic membrane, in: saline water conversion-ii, American Chemical Society, 1963, pp [6] C. Charcosset, 1 - Principles on membrane and membrane processes, in: C. Charcosset (Ed.) Membrane processes in biotechnology and pharmaceutics, Elsevier, Amsterdam, 2012, pp [7] P. Pandey, R. Chauhan, Membranes for gas separation, Prog. Polym. Sci., 26 (2001) [8] P.K. Gantzel, M. Ulrich, Method and apparatus for gas separation by diffusion, in, Google Patents, [9] Y. Yampolskii, Polymeric gas separation membranes, Macromolecules, 45 (2012) [10] R.W. Baker, Future directions of membrane gas separation technology, Ind. Eng. Chem. Res., 41 (2002) [11] M. Ulbricht, Advanced functional polymer membranes, Polymer, 47 (2006) [12] L.M. Robeson, Correlation of separation factor versus permeability for polymeric membranes, J. Membr. Sci., 62 (1991) [13] R. Swaidan, B. Ghanem, E. Litwiller, I. Pinnau, Physical aging, plasticization and their effects on gas permeation in rigid polymers of intrinsic microporosity, Macromolecules, 48 (2015) [14] Y. Xiao, B.T. Low, S.S. Hosseini, T.S. Chung, D.R. Paul, The strategies of molecular architecture and modification of polyimide-based membranes for CO2 removal from natural gas a review, Prog. Polym. Sci., 34 (2009)

69 Chapter 2 Literature Review [15] H. Du, J. Li, J. Zhang, G. Su, X. Li, Y. Zhao, Separation of hydrogen and nitrogen gases with porous graphene membrane, J. Phys. Chem. C, 115 (2011) [16] S. Hashim, A. Mohamed, S. Bhatia, Oxygen separation from air using ceramic-based membrane technology for sustainable fuel production and power generation, Renewable Sustainable Energy Rev., 15 (2011) [17] C. Yacou, S. Smart, J.C.D. da Costa, Long term performance cobalt oxide silica membrane module for high temperature H 2 separation, Energy Environ. Sci., 5 (2012) [18] A.J. Burggraaf, L. Cot, Fundamentals of inorganic membrane science and technology, Elsevier, [19] A. Iulianelli, S. Liguori, A. Vita, C. Italiano, C. Fabiano, Y. Huang, A. Basile, The oncoming energy vector: hydrogen produced in Pd-composite membrane reactor via bioethanol reforming over Ni/CeO2 catalyst, Catal. Today, 259 (2016) [20] T. Peters, M. Stange, P. Veenstra, A. Nijmeijer, R. Bredesen, The performance of Pd Ag alloy membrane films under exposure to trace amounts of H2S, J. Membr. Sci., 499 (2016) [21] M. Llosa, I. Azkarate, E. Fernandez, D. Pacheco, Thin palladium membranes for hydrogen separation, (2013). [22] Ø. Hatlevik, S.K. Gade, M.K. Keeling, P.M. Thoen, A. Davidson, J.D. Way, Palladium and palladium alloy membranes for hydrogen separation and production: History, fabrication strategies, and current performance, Sep. Purif. Technol., 73 (2010) [23] K. Watenabe, M. Yuasa, T. Kida, Y. Teraoka, N. Yamazoe, K. Shimanoe, High performance oxygen permeable membranes with an asymmetric structure using Ba0. 95La0. 05FeO3 δ Perovskite Type Oxide, Adv. Mater., 22 (2010) [24] S.S. Kim, N. Xu, A. Li, J.R. Grace, C.J. Lim, S.-K. Ryi, Development of a new porous metal support based on nickel and its application for Pd based composite membranes, Int. J. Hydrogen Energy, 40 (2015)

70 Chapter 2 Literature Review [25] N. Al-Mufachi, N. Rees, R. Steinberger-Wilkens, Hydrogen selective membranes: A review of palladium-based dense metal membranes, Renewable Sustainable Energy Rev., 47 (2015) [26] S. Gopalakrishnan, Y. Yoshino, M. Nomura, B.N. Nair, S.-I. Nakao, A hybrid processing method for high performance hydrogen-selective silica membranes, J. Membr. Sci., 297 (2007) 5-9. [27] D. Bougeard, K.S. Smirnov, Modelling studies of water in crystalline nanoporous aluminosilicates, Phys. Chem. Chem. Phys., 9 (2007) [28] D. Georgiev, B. Bogdanov, K. Angelova, I. Markovska, Y. Hristov, Synthetic zeolites Structure, classification, current trends in zeolite synthesis, in: Review. Economics and Society development on the Base of Knowledge : Int. Sci. Conf. VII, Stara Zagora, [29] A. Tavolaro, E. Drioli, Zeolite membranes, Adv. Mater., 11 (1999) [30] S. Sircar, A.L. Myers, Gas separation by zeolites, Handbook of Zeolite Science and Technology, (2003) [31] T. Bein, Synthesis and applications of molecular sieve layers and membranes, Chem. Mater., 8 (1996) [32] Y.S. Khodakov, I. Mikheikin, V. Nakhshunov, V. Shvets, V. Kazanskii, K.M. Minachev, Investigation of the structure and thermal stability of zeolites of types A, X, and Y, containing trivalent chromium, Bull. Acad. Sci. USSR, Div. Chem. Sci., 18 (1969) [33] J. García-Martínez, M. Johnson, J. Valla, K. Li, J.Y. Ying, Mesostructured zeolite Y high hydrothermal stability and superior FCC catalytic performance, Catal. Sci. Technol., 2 (2012) [34] M. Shah, M.C. McCarthy, S. Sachdeva, A.K. Lee, H.-K. Jeong, Current status of metal organic framework membranes for gas separations: promises and challenges, Ind. Eng. Chem. Res., 51 (2012) [35] E.E. McLeary, J.C. Jansen, F. Kapteijn, Zeolite based films, membranes and membrane reactors: Progress and prospects, Microporous Mesoporous Mater., 90 (2006)

71 Chapter 2 Literature Review [36] B.F. Hoskins, R. Robson, Infinite polymeric frameworks consisting of three dimensionally linked rod-like segments, J. Am. Chem. Soc., 111 (1989) [37] B. Hoskins, R. Robson, Design and construction of a new class of scaffolding-like materials comprising infinite polymeric frameworks of 3D-linked molecular rods. A reappraisal of the zinc cyanide and cadmium cyanide structures and the synthesis and structure of the diamond-related frameworks [N (CH3)4][CuIZnII (CN)4] and CuI [4, 4', 4'', 4'''-tetracyanotetraphenylmethane] BF4.C6H5NO2, J. Am. Chem. Soc., 112 (1990) [38] S. Subramanian, M.J. Zaworotko, Porous solids by design:[zn (4, 4 -bpy)2 (SiF 6)] ndmf, a single framework octahedral coordination polymer with large square channels, Angew. Chem., Int. Ed. Engl., 34 (1995) [39] D. Venkataraman, G.B. Gardner, S. Lee, J.S. Moore, Zeolite-like behavior of a coordination network, J. Am. Chem. Soc., 117 (1995) [40] G.B. Gardner, D. Venkataraman, J.S. Moore, S. Lee, Spontaneous assembly of a hinged coordination network, (1995). [41] O.M. Yaghi, G. Li, H. Li, Selective binding and removal of guests in a microporous metal organic framework, Nature, 378 (1995) [42] S. Li, F. Huo, Metal organic framework composites: from fundamentals to applications, Nanoscale, 7 (2015) [43] M.J. Katz, Z.J. Brown, Y.J. Colón, P.W. Siu, K.A. Scheidt, R.Q. Snurr, J.T. Hupp, O.K. Farha, A facile synthesis of UiO-66, UiO-67 and their derivatives, Chem. Commun., 49 (2013) [44] J.H. Cavka, S. Jakobsen, U. Olsbye, N. Guillou, C. Lamberti, S. Bordiga, K.P. Lillerud, A new zirconium inorganic building brick forming metal organic frameworks with exceptional stability, J. Am. Chem. Soc., 130 (2008) [45] S.L. James, Metal-organic frameworks, Chem. Soc. Rev., 32 (2003) [46] M.P. Suh, H.J. Park, T.K. Prasad, D.-W. Lim, Hydrogen storage in metal organic frameworks, Chem. Rev., 112 (2011)

72 Chapter 2 Literature Review [47] H. Wu, Q. Gong, D.H. Olson, J. Li, Commensurate adsorption of hydrocarbons and alcohols in microporous metal organic frameworks, Chem. Rev., 112 (2012) [48] M. Yoon, R. Srirambalaji, K. Kim, Homochiral metal organic frameworks for asymmetric heterogeneous catalysis, Chem. Rev., 112 (2011) [49] J. Gao, L. Bai, Q. Zhang, Y. Li, G. Rakesh, J.-M. Lee, Y. Yang, Q. Zhang, Co6 (μ3-oh)6 cluster based coordination polymer as an effective heterogeneous catalyst for aerobic epoxidation of alkenes, Dalton Trans., 43 (2014) [50] H. Furukawa, K.E. Cordova, M. O Keeffe, O.M. Yaghi, The chemistry and applications of metal-organic frameworks, Science, 341 (2013). [51] K. Sumida, D.L. Rogow, J.A. Mason, T.M. McDonald, E.D. Bloch, Z.R. Herm, T.-H. Bae, J.R. Long, Carbon dioxide capture in metal organic frameworks, Chem. Rev., 112 (2011) [52] J.-R. Li, J. Sculley, H.-C. Zhou, Metal organic frameworks for separations, Chem. Rev., 112 (2012) [53] J.C. Tan, A.K. Cheetham, Mechanical properties of hybrid inorganic-organic framework materials: establishing fundamental structure-property relationships, Chem. Soc. Rev., 40 (2011) [54] X. Guo, H. Huang, Y. Ban, Q. Yang, Y. Xiao, Y. Li, W. Yang, C. Zhong, Mixed matrix membranes incorporated with amine-functionalized titanium-based metal-organic framework for CO2/CH4 separation, J. Membr. Sci., 478 (2015) [55] J. Yu, Y. Pan, C. Wang, Z. Lai, ZIF-8 membranes with improved reproducibility fabricated from sputter-coated ZnO/alumina supports, Chem. Eng. Sci., 141 (2016) [56] J. Yao, D. Dong, D. Li, L. He, G. Xu, H. Wang, Contra-diffusion synthesis of ZIF-8 films on a polymer substrate, Chem. Commun., 47 (2011) [57] K.S. Park, Z. Ni, A.P. Côté, J.Y. Choi, R. Huang, F.J. Uribe-Romo, H.K. Chae, M. O Keeffe, O.M. Yaghi, Exceptional chemical and thermal stability of zeolitic imidazolate frameworks, Proc. Natl. Acad. Sci. U. S. A., 103 (2006)

73 Chapter 2 Literature Review [58] K. Kida, M. Okita, K. Fujita, S. Tanaka, Y. Miyake, Formation of high crystalline ZIF-8 in an aqueous solution, CrystEngComm, 15 (2013) [59] Q. Shi, Z. Chen, Z. Song, J. Li, J. Dong, Synthesis of ZIF 8 and ZIF 67 by Steam Assisted Conversion and an Investigation of Their Tribological Behaviors, Angew. Chem., 123 (2011) [60] Y. Pan, Y. Liu, G. Zeng, L. Zhao, Z. Lai, Rapid synthesis of zeolitic imidazolate framework- 8 (ZIF-8) nanocrystals in an aqueous system, Chem. Commun., 47 (2011) [61] J. Yao, H. Wang, Zeolitic imidazolate framework composite membranes and thin films: synthesis and applications, Chem. Soc. Rev., 43 (2014) [62] C. Zhang, R.P. Lively, K. Zhang, J.R. Johnson, O. Karvan, W.J. Koros, Unexpected Molecular sieving properties of zeolitic imidazolate framework-8, J. Phys. Chem. Lett., 3 (2012) [63] K. Li, D.H. Olson, J. Seidel, T.J. Emge, H. Gong, H. Zeng, J. Li, Zeolitic imidazolate frameworks for kinetic separation of propane and propene, J. Am. Chem. Soc., 131 (2009) [64] Y. Pan, T. Li, G. Lestari, Z. Lai, Effective separation of propylene/propane binary mixtures by ZIF-8 membranes, J. Membr. Sci., 390 (2012) [65] P. Küsgens, M. Rose, I. Senkovska, H. Fröde, A. Henschel, S. Siegle, S. Kaskel, Characterization of metal-organic frameworks by water adsorption, Microporous Mesoporous Mater., 120 (2009) [66] R. Chen, J. Yao, Q. Gu, S. Smeets, C. Baerlocher, H. Gu, D. Zhu, W. Morris, O.M. Yaghi, H. Wang, A two-dimensional zeolitic imidazolate framework with a cushion-shaped cavity for CO2 adsorption, Chem. Commun., 49 (2013) [67] Z.-X. Low, J. Yao, Q. Liu, M. He, Z. Wang, A.K. Suresh, J. Bellare, H. Wang, Crystal transformation in zeolitic-imidazolate framework, Cryst. Growth Des., 14 (2014) [68] Z.-X. Low, A. Razmjou, K. Wang, S. Gray, M. Duke, H. Wang, Effect of addition of twodimensional zif-l nanoflakes on the properties of polyethersulfone ultrafiltration membrane, J. Membr. Sci., 460 (2014)

74 Chapter 2 Literature Review [69] Z. Zhong, J. Yao, Z.-X. Low, R. Chen, M. He, H. Wang, Carbon composite membrane derived from a two-dimensional zeolitic imidazolate framework and its gas separation properties, Carbon, 72 (2014) [70] Z. Zhong, J. Yao, R. Chen, Z. Low, M. He, J.Z. Liu, H. Wang, Oriented two-dimensional zeolitic imidazolate framework-l membranes and their gas permeation properties, J. Mater. Chem. A, 3 (2015) [71] H. Wu, T. Yildirim, W. Zhou, Exceptional mechanical stability of highly porous zirconium metal organic framework UiO-66 and its important implications, J. Phys. Chem. Lett., 4 (2013) [72] E. Nightingale Jr, Phenomenological theory of ion solvation. Effective radii of hydrated ions, J. Phys. Chem., 63 (1959) [73] X. Liu, N. Keser Demir, Z. Wu, K. Li, Highly water stable zirconium metal-organic framework UiO-66 membranes supported on alumina hollow fibers for desalination, J. Am. Chem. Soc., (2015). [74] Z. Hu, D. Zhao, De facto methodologies toward the synthesis and scale-up production of UiO- 66-type metal-organic frameworks and membrane materials, Dalton Trans., 44 (2015) [75] E. Biemmi, C. Scherb, T. Bein, Oriented growth of the metal organic framework Cu3(BTC)2(H2O)3 H2O tunable with functionalized self-assembled monolayers, J. Am. Chem. Soc., 129 (2007) [76] S. Hermes, F. Schröder, R. Chelmowski, C. Wöll, R.A. Fischer, Selective nucleation and growth of metal organic open framework thin films on patterned COOH/CF3-terminated selfassembled monolayers on Au(111), J. Am. Chem. Soc., 127 (2005) [77] H. Bux, F. Liang, Y. Li, J. Cravillon, M. Wiebcke, J. Caro, Zeolitic imidazolate framework membrane with molecular sieving properties by microwave-assisted solvothermal synthesis, J. Am. Chem. Soc., 131 (2009) [78] C. Scherb, A. Schödel, T. Bein, Directing the structure of metal organic frameworks by oriented surface growth on an organic monolayer, Angew. Chem., Int. Ed., 47 (2008)

75 Chapter 2 Literature Review [79] M. Miyamoto, S. Kohmura, H. Iwatsuka, Y. Oumi, S. Uemiya, In situ solvothermal growth of highly oriented Zr-based metal organic framework UiO-66 film with monocrystalline layer, CrystEngComm, 17 (2015) [80] I. Stassen, M. Styles, T. Van Assche, N. Campagnol, J. Fransaer, J. Denayer, J.-C. Tan, P. Falcaro, D. De Vos, R. Ameloot, Electrochemical film deposition of the zirconium metal organic framework UiO-66 and application in a miniaturized sorbent trap, Chem. Mater., 27 (2015) [81] K.A. Cychosz, A.J. Matzger, Water stability of microporous coordination polymers and the adsorption of pharmaceuticals from water, Langmuir, 26 (2010) [82] P.M. Schoenecker, C.G. Carson, H. Jasuja, C.J.J. Flemming, K.S. Walton, Effect of water adsorption on retention of structure and surface area of metal organic frameworks, Ind. Eng. Chem. Res., 51 (2012) [83] F. Gul-E-Noor, B. Jee, A. Poppl, M. Hartmann, D. Himsl, M. Bertmer, Effects of varying water adsorption on a Cu3(BTC)2 metal-organic framework (MOF) as studied by 1 H and 13 C solidstate NMR spectroscopy, Phys. Chem. Chem. Phys., 13 (2011) [84] J. Liu, P.K. Thallapally, B.P. McGrail, D.R. Brown, J. Liu, Progress in adsorption-based CO2 capture by metal-organic frameworks, Chem. Soc. Rev., 41 (2012) [85] N.C. Burtch, H. Jasuja, K.S. Walton, Water stability and adsorption in metal organic frameworks, Chem. Rev., 114 (2014) [86] K. Zhang, R.P. Lively, M.E. Dose, A.J. Brown, C. Zhang, J. Chung, S. Nair, W.J. Koros, R.R. Chance, Alcohol and water adsorption in zeolitic imidazolate frameworks, Chem. Commun., 49 (2013) [87] W. Li, Y. Zhang, Q. Li, G. Zhang, Metal organic framework composite membranes: Synthesis and separation applications, Chem. Eng. Sci., 135 (2015) [88] A.J. Brown, N.A. Brunelli, K. Eum, F. Rashidi, J. Johnson, W.J. Koros, C.W. Jones, S. Nair, Interfacial microfluidic processing of metal-organic framework hollow fiber membranes, Science, 345 (2014)

76 Chapter 2 Literature Review [89] H.T. Kwon, H.-K. Jeong, In situ synthesis of thin zeolitic imidazolate framework zif-8 membranes exhibiting exceptionally high propylene/propane separation, J. Am. Chem. Soc., 135 (2013) [90] K. Huang, Q. Li, G. Liu, J. Shen, K. Guan, W. Jin, A ZIF-71 hollow fiber membrane fabricated by contra-diffusion, ACS Appl. Mater. Interfaces, 7 (2015) [91] N. Hara, M. Yoshimune, H. Negishi, K. Haraya, S. Hara, T. Yamaguchi, Metal organic framework membranes with layered structure prepared within the porous support, RSC Adv., 3 (2013) [92] O. Shekhah, R. Swaidan, Y. Belmabkhout, M. du Plessis, T. Jacobs, L.J. Barbour, I. Pinnau, M. Eddaoudi, The liquid phase epitaxy approach for the successful construction of ultra-thin and defect-free ZIF-8 membranes: pure and mixed gas transport study, Chem. Commun., 50 (2014) [93] O. Shekhah, H. Wang, S. Kowarik, F. Schreiber, M. Paulus, M. Tolan, C. Sternemann, F. Evers, D. Zacher, R.A. Fischer, C. Wöll, Step-by-step route for the synthesis of metal organic frameworks, J. Am. Chem. Soc., 129 (2007) [94] B. Seoane, J. Coronas, I. Gascon, M.E. Benavides, O. Karvan, J. Caro, F. Kapteijn, J. Gascon, Metal organic framework based mixed matrix membranes: a solution for highly efficient CO2 capture?, Chem. Soc. Rev., 44 (2015) [95] A. Centrone, Y. Yang, S. Speakman, L. Bromberg, G.C. Rutledge, T.A. Hatton, Growth of metal organic frameworks on polymer surfaces, J. Am. Chem. Soc., 132 (2010) [96] W. Li, Q. Meng, X. Li, C. Zhang, Z. Fan, G. Zhang, Non-activation ZnO array as a buffering layer to fabricate strongly adhesive metal organic framework/pvdf hollow fiber membranes, Chem. Commun., 50 (2014) [97] W. Li, Z. Yang, G. Zhang, Z. Fan, Q. Meng, C. Shen, C. Gao, Stiff metal organic framework polyacrylonitrile hollow fiber composite membranes with high gas permeability, J. Mater. Chem. A, 2 (2014)

77 Chapter 2 Literature Review [98] M. He, J. Yao, L. Li, Z. Zhong, F. Chen, H. Wang, Aqueous solution synthesis of ZIF-8 films on a porous nylon substrate by contra-diffusion method, Microporous Mesoporous Mater., 179 (2013) [99] D. Nagaraju, D.G. Bhagat, R. Banerjee, U.K. Kharul, In situ growth of metal-organic frameworks on a porous ultrafiltration membrane for gas separation, J. Mater. Chem. A, 1 (2013) [100] F. Cacho-Bailo, B. Seoane, C. Téllez, J. Coronas, ZIF-8 continuous membrane on porous polysulfone for hydrogen separation, J. Membr. Sci., 464 (2014) [101] L. Ge, W. Zhou, A. Du, Z. Zhu, Porous polyethersulfone-supported zeolitic imidazolate framework membranes for hydrogen separation, J. Phys. Chem. C, 116 (2012) [102] E. Barankova, N. Pradeep, K.-V. Peinemann, Zeolite-imidazolate framework (ZIF-8) membrane synthesis on a mixed-matrix substrate, Chem. Commun., 49 (2013) [103] B.P. Biswal, A. Bhaskar, R. Banerjee, U.K. Kharul, Selective interfacial synthesis of metal organic frameworks on a polybenzimidazole hollow fiber membrane for gas separation, Nanoscale, 7 (2015) [104] A.J. Brown, J. Johnson, M.E. Lydon, W.J. Koros, C.W. Jones, S. Nair, Continuous polycrystalline zeolitic imidazolate framework 90 membranes on polymeric hollow fibers, Angew. Chem., Int. Ed., 51 (2012) [105] M. Arnold, P. Kortunov, D.J. Jones, Y. Nedellec, J. Kärger, J. Caro, Oriented crystallisation on supports and anisotropic mass transport of the metal organic framework manganese formate, Eur. J. Inorg. Chem., 2007 (2007) [106] M.C. McCarthy, V. Varela-Guerrero, G.V. Barnett, H.-K. Jeong, Synthesis of zeolitic imidazolate framework films and membranes with controlled microstructures, Langmuir, 26 (2010) [107] S. Hermes, D. Zacher, A. Baunemann, C. Wöll, R.A. Fischer, Selective growth and MOCVD loading of small single crystals of MOF-5 at alumina and silica surfaces modified with organic self-assembled monolayers, Chem. Mater., 19 (2007)

78 Chapter 2 Literature Review [108] E. Biemmi, A. Darga, N. Stock, T. Bein, Direct growth of Cu3 (BTC)2 (H2O)3. xh2o thin films on modified QCM-gold electrodes: Water sorption isotherms, Microporous Mesoporous Mater., 114 (2008) [109] W. Li, Z. Yang, G. Zhang, Q. Meng, Heat-treated polyacrylonitrile (pan) hollow fiber structured packings in isopropanol (IPA)/water distillation with improved thermal and chemical stability, Ind. Eng. Chem. Res., 52 (2013) [110] L.I.B. David, A.F. Ismail, Influence of the thermastabilization process and soak time during pyrolysis process on the polyacrylonitrile carbon membranes for O2/N2 separation, J. Membr. Sci., 213 (2003) [111] Q. Ge, Z. Wang, Y. Yan, High-performance zeolite NaA membranes on polymer zeolite composite hollow fiber supports, J. Am. Chem. Soc., 131 (2009) [112] W. Li, G. Zhang, C. Zhang, Q. Meng, Z. Fan, C. Gao, Synthesis of trinity metal-organic framework membranes for CO2 capture, Chem. Commun., 50 (2014) [113] Y. Mao, J. Li, W. Cao, Y. Ying, L. Sun, X. Peng, Pressure-assisted synthesis of HKUST-1 thin film on polymer hollow fiber at room temperature toward gas separation, ACS Appl. Mater. Interfaces, 6 (2014) [114] F. Cacho-Bailo, S. Catalán-Aguirre, M. Etxeberría-Benavides, O. Karvan, V. Sebastian, C. Téllez, J. Coronas, Metal-organic framework membranes on the inner-side of a polymeric hollow fiber by microfluidic synthesis, J. Membr. Sci., 476 (2015) [115] E.V. Perez, K.J. Balkus, J.P. Ferraris, I.H. Musselman, Mixed-matrix membranes containing MOF-5 for gas separations, J. Membr. Sci., 328 (2009) [116] T. Yang, Y. Xiao, T.-S. Chung, Poly-/metal-benzimidazole nano-composite membranes for hydrogen purification, Energy Environ. Sci., 4 (2011) [117] B. Seoane, J. Coronas, I. Gascon, M.E. Benavides, O. Karvan, J. Caro, F. Kapteijn, J. Gascon, Metal-organic framework based mixed matrix membranes: a solution for highly efficient CO2 capture?, Chem. Soc. Rev., 44 (2015)

79 Chapter 2 Literature Review [118] B. Zornoza, C. Tellez, J. Coronas, J. Gascon, F. Kapteijn, Metal organic framework based mixed matrix membranes: An increasingly important field of research with a large application potential, Microporous Mesoporous Mater., 166 (2013) [119] H.B. Tanh Jeazet, C. Staudt, C. Janiak, Metal-organic frameworks in mixed-matrix membranes for gas separation, Dalton Trans., 41 (2012) [120] B. Zornoza, A. Martinez-Joaristi, P. Serra-Crespo, C. Tellez, J. Coronas, J. Gascon, F. Kapteijn, Functionalized flexible MOFs as fillers in mixed matrix membranes for highly selective separation of CO2 from CH4 at elevated pressures, Chem. Commun., 47 (2011) [121] R. Lin, L. Ge, L. Hou, E. Strounina, V. Rudolph, Z. Zhu, Mixed matrix membranes with strengthened MOFs/polymer interfacial interaction and improved membrane performance, ACS Appl. Mater. Interfaces, 6 (2014) [122] J.G. Wijmans, R.W. Baker, The solution-diffusion model: a review, J. Membr. Sci., 107 (1995) [123] R.W. Baker, B.T. Low, Gas separation membrane materials: a perspective, Macromolecules, 47 (2014) [124] T. Yang, G.M. Shi, T.-S. Chung, Symmetric and asymmetric zeolitic imidazolate frameworks (ZIFs)/polybenzimidazole (PBI) nanocomposite membranes for hydrogen purification at high temperatures, Advanced Energy Materials, 2 (2012) [125] M.J.C. Ordoñez, K.J. Balkus Jr, J.P. Ferraris, I.H. Musselman, Molecular sieving realized with ZIF-8/Matrimid mixed-matrix membranes, J. Membr. Sci., 361 (2010) [126] T.C. Merkel, B.D. Freeman, R.J. Spontak, Z. He, I. Pinnau, P. Meakin, A.J. Hill, Sorption, transport, and structural evidence for enhanced free volume in poly(4-methyl-2-pentyne)/fumed silica nanocomposite membranes, Chem. Mater., 15 (2003) [127] S. Matteucci, V.A. Kusuma, S.D. Kelman, B.D. Freeman, Gas transport properties of MgO filled poly(1-trimethylsilyl-1-propyne) nanocomposites, Polymer, 49 (2008) [128] J.A. Thompson, K.W. Chapman, W.J. Koros, C.W. Jones, S. Nair, Sonication-induced Ostwald ripening of ZIF-8 nanoparticles and formation of ZIF-8/polymer composite membranes, Microporous Mesoporous Mater., 158 (2012)

80 Chapter 2 Literature Review [129] Q. Song, S.K. Nataraj, M.V. Roussenova, J.C. Tan, D.J. Hughes, W. Li, P. Bourgoin, M.A. Alam, A.K. Cheetham, S.A. Al-Muhtaseb, E. Sivaniah, Zeolitic imidazolate framework (ZIF-8) based polymer nanocomposite membranes for gas separation, Energy Environ. Sci., 5 (2012) [130] T. Yang, T.-S. Chung, Room-temperature synthesis of ZIF-90 nanocrystals and the derived nano-composite membranes for hydrogen separation, J. Mater. Chem. A, 1 (2013) [131] T. Yang, T.-S. Chung, High performance ZIF-8/PBI nano-composite membranes for high temperature hydrogen separation consisting of carbon monoxide and water vapor, Int. J. Hydrogen Energy, 38 (2013) [132] B. Zornoza, B. Seoane, J.M. Zamaro, C. Téllez, J. Coronas, combination of MOFs and zeolites for mixed-matrix membranes, ChemPhysChem, 12 (2011)

81 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support Rapid Synthesis of Ultrathin, Defect-Free ZIF-8 Membranes via Chemical Vapour Modification of Polymeric Support Overview In this chapter, ultrathin ZIF-8 membranes with a thickness of around 200 nm were prepared by chemical vapour modification of surface chemistry and nanopores of asymmetric bromomethylated poly (2, 6-dimethyl-1, 4-phenylene oxide) (BPPO) substrate. The resulting ZIF-8 membranes exhibited exceptional H2 permeance as high as mol.m -2.s -1.Pa -1 with high H2/N2 and H2/CO2 selectivities (9.7 and12.8, respectively). This chapter has been reformatted from the following published manuscript: Shamsaei, E., Low, Z.X., Lin, X., Mayahi, A., Liu, H., Zhang, X., Liu, J.Z. and Wang, H. Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of a polymeric support. Chemical Communications, 2015, 51(57), pp Introduction Zeolitic imidazolate frameworks (ZIFs), a subclass of metal organic frameworks (MOFs), are hybrid porous crystalline materials composed of metal ions (e.g., Zn, Co) bridged by imidazolates [1, 2]. Remarkably, they exhibit permanent porosity and relatively high chemical and thermal stability [3, 4], which make them very promising 62

82 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support candidate materials for numerous applications such as catalysis [5], molecular separation [6], chemical sensing [7], gas adsorption and storage [8]. In particular, the preparation of MOFs into membranes and thin films is desirable for gas separation [9]. The welldefined porous structures of ZIFs allow them to achieve gas separation with high selectivity via molecular sieving. ZIF membranes are prepared by growing a thin ZIF layer on porous substrate via two general techniques, i.e. in situ (direct) growth and secondary (seeded) growth [10-12]. In situ growth, a method used for direct growth of a ZIF layer on a bare porous substrate, has been widely studied for the fabrication of ZIF membranes. However, due to limited heterogeneous nucleation sites on the support and poor compatibility between ZIF and support, this method may result in defective ZIF films with intercrystalline voids [13, 14]. Chemical modification can effectively overcome the issue by providing anchors to ligands or metal ions. Compared to other methods such as microwave-assisted solvothermal synthesis [15], rapid thermal deposition (RTD) [16], layer-by-layer deposition of crystal [17], and liquid-phase epitaxy (LPE) [18], chemical modification not only provides a faster and energyefficient route but also indirectly improves the thermal stability and chemical resistance of the composite. So far, most supports for growing a ZIF layer are ceramic-based materials such as alumina. To favour ZIF formation and adhesion, these supports are functionalized with organosilane molecules or other functional groups with amine group [12, 19-24]. Caro and co-workers prepared a continuous ZIF-90 membrane by using 3- aminopropyltriethoxysilane (APTES) as a covalent linker [23]. The amine groups of the APTES were shown to react with the aldehyde groups of the ZIF precursor and promote the nucleation and growth of the ZIF-90 at these fixed sites on the surface of the porous ceramic substrate. A similar surface modification method and its influence on 63

83 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support heterogeneous nucleation have been also demonstrated with other MOF materials by other groups [25-27]. However, the poor scalability and high cost of the inorganic materials limit the applications of ZIF membranes [14, 28-31]. Growing ZIF on a polymer substrate, on the other hand, has great potential to achieve high-quality membranes at a low cost. To date, there have been only a few reports on successful in situ growth of ZIF on a polymer surface [13, 14, 17, 18]. To effectively use a polymer membrane as a substrate, the polymer membrane needs to be chemically modified. Li et al. successfully grew a continuous and well integrated ZIF-8 layer on a polyvinylidene fluoride (PVDF) substrate treated in ammonia or ethanediamine solution [14]. The same group also successfully prepared ZIF-8-polyacrylonitrile (PAN) membrane by hydrolysing PAN substrate to produce deprotonated carboxyl groups [32]. The modified polymer substrates exhibited increased stiffness due to cross-linking [14, 32]. Nevertheless, solution phase chemical modification for the preparation of polymersupported ZIF membranes can cause uneven swelling and adversely affect the polymeric membrane morphology and separation performance [33, 34]. Also, the contaminants in chemical modifier solution increase with substrate modification cycle and need to be treated before reuse. Furthermore, the substrates after modification need to be rinsed and dried before growing ZIF. At a lab scale, fresh modifiers can be used and the substrate can be left at room temperature until completely dry; but at larger scale, these steps need to be improved on the basis of production cost and time. Here we report a novel scalable strategy of using vapour phase ethylenediamine (EDA) to chemically modify the polymer support for ZIF membrane fabrication. An asymmetric bromomethylated poly(2,6-dimethyl-1,4-phenylene oxide) (BPPO) ultrafiltration membrane was fabricated by non-solvent induced phase separation, and used as the support for growing a thin ZIF-8 layer via a rapid in situ route after chemical 64

84 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support vapour modification. BPPO is a common polymer with a high glass transition temperature, high mechanical strength and excellent hydrolytic stability [35]. It attains superior membrane formation and functionalizable characteristics due to the abundant highly reactive -CH2Br groups. EDA-vapour modification of BPPO results in the reduction of pore size of support and also provides a large number of nucleation sites. When combined with the rapid, in situ growth of ZIF, a submicron-thin and defect-free ZIF-8 membrane with high gas separation performance can be attained. Experimental Materials BPPO was kindly supplied by Tianwei Membrane Co. Ltd., Shandong of China. Zinc acetate dihydrate (Zn(CH3COO)2.2H2O, 98%), 2-methylimidazole (Hmim, C4H6N2, 99%), ethylenediamine (EDA, 99.5%) and ammonium hydroxide solution (NH3, 28 30% aqueous solution) were purchased from Sigma-Aldrich, Australia and used as received. Methanol (absolute) was purchased from Merck, Australia Synthesis of BPPO membrane and its EDA-vapour modification The flat sheet BPPO support ultrafiltration membranes were fabricated via nonsolvent induced phase inversion (also known as the immersion precipitation technique). Dope solution was prepared by dissolving 18 wt. % of BPPO in NMP at around 25 C for 24 h with mechanical stirring at 200 rpm. The homogenous solution was left stagnant until no bubbles were observed. Subsequently, the polymer solution was cast on a cleaned glass plate using a casting knife (Paul N. Gardner Co., Inc. USA) with a gap of 150 µm at room temperature (21 23 C) and % humidity and immediately immersed in a coagulation bath of deionized water (Figure 3-1a). After peeling off from 65

85 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support the glass plate, the membranes were removed from the coagulation bath, washed and kept in water bath for at least one day to thoroughly remove the residual solvents. The thickness of the prepared membrane was about 70 μm. The vapour-phase EDA modification was conducted in a custom-made container as illustrated in Figure 3-1b. 20 ml of EDA was allowed to vaporize, and stabilized for 1 h. Based on the Antoine equation [36], the EDA vapour pressure at 25 C was estimated to be 12.0 mm of Hg and the air of the closed chamber consists of 1.6 % v/v EDA vapour. The support membranes were quickly placed inside the containment with the top layer exposed and suspended above the EDA solution. After surface modification at room temperature for 4-16 h, the surface modified membranes were removed from the containment and immediately washed with pure water to completely remove the residual EDA. The resultant membranes were denoted as MBPPO-4, MBPPO-10, and MBPPO- 16, where numbers show EDA exposure time. Figure 3-1(a) Schematic diagram of UF membrane fabrication via phase inversion, (b) Experimental setup of vapour-phase EDA modification process. 66

86 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support Growth of ZIF-8 Thin Film on modified BPPO Supports Modified BPPO supports were immersed vertically in the solution of zinc acetate dehydrate (0.22 g) in 9.6 g methanol and sonicated for 3 min to fix the Zn 2+. A solution of Hmim (0.164 g) in 9.6 g methanol was added to the above solution followed by dropwise addition of ammonia hydroxide solution (0.12 g) and the mixture was then ultrasonically treated for another 3 min. After crystallization, the composite membranes were washed with methanol and dried. The ZIF-8 nanocrystals were also separated from the solution by centrifugation and washed several times with methanol, and dried at 60 C overnight. For comparison, BPPO was also used for ZIF-8 membrane growth under the same condition Pure water flux and molecular weight cut off (MWCO) measurements Pure water flux of the membranes was determined at room temperature (21 23 C) using a Sterlitech HP4750 dead-end stirred cell (Sterlitech Corporation, USA) with an inner diameter of 49 mm and an effective membrane area of 14.6 cm 2. The cell has a volume capacity of 300 ml and is attached to a 5.0 L dispensing vessel. To attain stable flux data, each membrane was first pre-compacted at 150 kpa for about 60 min, and then the pure water flux was measured at a trans-membrane pressure drop 100 kpa. Pure water flux was measured constantly by collecting the permeate on a digital balance (PA2102C, Ohaus) interfaced with a computer. The data from the balance was logged to a computer using a program written in LabVIEW. Polyethylene glycol (PEG) with a molecular weight of 10, 20, 35, 100, 200 and 300 kda (analytical grade, Sigma-Aldrich) was dissolved in deionized water to prepare 1 g L -1 aqueous solutions for the estimation of MWCO and solute rejection. Rejection measurements were performed at a pressure of 100 kpa. 20 ml of permeate was collected. The permeate and feed solution were both 67

87 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support diluted by 10 times and then the concentration of each solution was measured via a total organic carbon analyser (TOC-LCSH, Shimadzu, Japan). The PEG rejection was calculated from the measured feed (Cf) and permeate (Cp) concentrations by R = (1 C p C f ) 100 All results represent average values for at least three repeated experiments with less than ±5% deviation. The pore size of the membrane was defined as the hydrodynamic diameter of PEG. The hydrodynamic radius of PEG can be calculated from the MWCO of the membrane by the following equation [37]: Solute Radius (nm) = MW.03 where MW is the lowest molecular weight of the PEG molecule which has a rejection of 90% in the ultrafiltration measurements Gas permeation experiments The gas permeation test is carried out as previously reported [13]. The composite membranes were attached to a stainless steel stand with pore size ~200 nm, which was fixed in a sample holder with Torr Seal epoxy resin (Varian). The film was dried at 100 C for 2 h to remove H2O. Gas permeation tests were performed at 20 C for pure H2, CO2 and N2. Between each measurement, the system was evacuated for 30 min prior to introduction of the next gas. The pressure increase of the permeate stream was measured and the permeance Pi of each gas calculated by: P i = N i P i A 68

88 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support where Ni is the permeating flow rate of component i (mol/s); ΔPi is the transmembrane pressure difference of component i (Pa), and A is the membrane area (m 2 ). The ideal selectivity Sij is defined as the ratio of the two permeances Pi and Pj Characterization Fourier Transform Infrared (FTIR) spectra of the membranes were recorded using an attenuated total reflectance (ATR) FTIR (Perkin Elmer, USA) in the range of cm -1 at an average of 32 scans with a resolution of 4 cm -1. Thermogravimetric analyses (TGA) were carried out on a SETARAM (TGA 92) device from 30 to 800 C at a heating rate of 10 C min -1 under air flow. Scanning electron microscopy (SEM) (FEI Nova NanoSEM 450) with a X-ray detector (Bruker Nano GmbH, Germany) was used for imaging the surface and cross-sectional morphologies of membranes. Energydispersive X-ray sepectroscopy (EDS) line-scan analysis of the membrane samples was conducted using EDX equipped in Nova NanoSEM 450 (Quantax 400 X-ray analysis system, Bruker, USA). The membranes were fractured in liquid nitrogen, fixed on stubs with double-sided carbon tape and then sputter coated with roughly 2 nm iridium (Ir) layer to ensure good electrical conductivity. The images were recorded at an accelerating voltage of 5 kv with different magnifications. Transmission electron microscopy (TEM) micrographs were obtained using a JEOL JEM- 2100F instrument operating at 200 kv. Selected-area electron diffraction (SAED) patterns were taken using the same instrument. The ZIF-8 samples were dispersed on a copper-supported carbon grid for TEM observation. Powder X-ray diffraction (XRD) patterns were measured using a Miniflex 600 diffractometer (Rigaku, Japan) with Cu Kα radiation (15 ma and 40 kv) at a scan rate of 2 min -1 with a step size of The XRD studies were carried out at room temperature. The crystallite size was estimated by using the Scherrer equation: 69

89 d = Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support kλ β cos θ where β is the full-width at half-maximum in radian for the peak (011); k is Scherrer constant (0.94); λ is the X-ray wavelength, nm for Cu Kα; θ is the angle of the diffraction peak in degree. Nitrogen (N2) adsorption desorption isotherms were measured using physisorption analyser (Micromeritics ASAP 2020, USA) at liquid nitrogen temperature (77 K). All the samples were degassed at 100 C for 12 h prior to analysis. Figure 3-2. Schematic diagram of the preparation of BPPO polymer-supported ZIF-8 membrane. Results and Discussion The method developed in this work involves two steps, as shown in Figure 3-2. First, the physicochemical properties of the top layer of BPPO support were modified by using EDA-vapour. Then, ZIF-8 was grown inside the pores and on the surface of the support via a rapid, in-situ seeding method [38]. The modified BPPO is denoted as MBPPO-X (X: modification reaction time). EDA-vapour modification is surface-limiting with minimal swelling effect to the polymer support, and this process can be carried out at room temperature [33, 34]. Unlike liquid phase modification, EDA is directly reusable 70

90 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support in the vapour modification. Therefore, amine functional groups for the coordination with Zn 2+ ions can be covalently attached on the top layer of BPPO support without affecting the sublayer structure. For formation of ZIF-8 layer, the vapour-phase-eda-modified BPPO support was immersed vertically in the solution of zinc acetate dehydrate (0.22 g) in 9.6 g methanol and sonicated for 3 min. A solution of Hmim (0.164 g) in 9.6 g methanol was added to the above solution, followed by dropwise addition of ammonia hydroxide solution (0.12 g) and the mixture was then ultrasonically treated for another 3 min. The final precursor solution had a Zn: Hmim: NH3: CH3OH molar composition of 1: 2: 2: 300 and was kept constant in our study. After crystallization at room temperature, the ZIF-8/BPPO composite membranes were washed with methanol and dried. The following simultaneous amination and crosslinking reactions may take place during the surface modification [39]. R CH2Br + NH2CH2CH2NH2 R CH2NHCH2CH2NH2 + HBr (1) R CH2Br + R CH2NHCH2CH2NH2 R CH2NHCH2CH2NHCH2R + HBr (2) Reactions (1) and (2) are typical amination and crosslinking reactions. Reaction (1) takes place where one end of a diamine molecule reacts with bromine groups ( CH2Br). Reaction (2) occurs where EDA reacts at both ends of the diamine molecule to form either interchain or intrachain crosslinks Membrane support Pure water flux and PEG rejection of the membranes were analysed to evaluate the influence of the chemical modification on the membrane permeability in correlation with the change of membranes surface microstructure after EDA-vapour modification. There was a pronounced drop in pore size from 17.5 nm for BPPO to 11.5 nm for the membrane following 16 hours EDA-vapour treatment (MBPPO-16). It is worth noting that the 71

91 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support values of the pure water flux and the pore size show that all the membranes fall in the ultrafiltration range. Membranes with such porosity are desirable as support for ZIF membranes since they provide a platform for the growth of ZIFs with no resistance or interruption to gas permeation. Figure 3-3. FTIR ATR spectra of untreated BPPO support, BPPO modified with EDAvapour for 16 h (MBPPO-16), MBPPO-16 supported ZIF-8 layer (ZIF-8-MBPPO-16), and synthesized ZIF-8 powder. The chemical reaction between the BPPO support and EDA vapour during the modification was determined using the FTIR-ATR technique and the results are shown in Figure 3-3. The pristine BPPO has IR bands at around 586 cm -1 and 633 cm -1, which are attributed to the benzyl bromide ( CH2Br) groups (C Br stretching). After EDAvapour modification, these bands almost disappear and a new broad band in the range of ~ cm -1 emerges, which is ascribed to the N H stretching and confirms the amination of BPPO. The ZIF-8 characteristic band at 421 cm -1 (Zn N stretching) is observed in the BPPO-supported ZIF-8 membrane (ZIF-8-MBPPO-16). Furthermore, TGA results (Figure 3-4) suggest higher degradation temperature for main chains of BPPO after EDA vapour-phase modification, which could be attributed to the presence 72

92 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support of partial crosslinking in the modified BPPO substrate. TGA measurements of the untreated BPPO support show a two-step degradation pattern primarily due to the weight loss commencing at ~ 240 C associated with the degradation of bromomethyl side groups ( CH2Br) followed by the decomposition of the aromatic main chains at ~ 468 C. In contrast, MBPPO shows a slight mass loss (about 4 wt. %) at lower temperatures due to the loss of absorbed water. The loading amounts of ZIF-8 can be roughly estimated from the zinc oxide (ZnO) residue at 800 C in TG curves. The results show the ZIF-8 loading amounts increase from 1.9 % for ZIF-8-BPPO to 13.8 % for ZIF-8-MBPPO-16. These results clearly show the importance of EDA modification in the growth of ZIF-8 on the supporting membranes. This was further verified by reduced water flux and increased rejection of the support after the modification, since the crosslinking causes tightening of the polymer network which increases the membrane dimensional stability and reduces the pore size of BPPO substrate (Figure 3-5). This reduces the flexibility of the polymeric substrate, which is favourable for reducing ZIF layer cracking [9]. Figure 3-4. TGA curves (under air flow) of (1) untreated BPPO support, (2) MBPPO- 16, (3) ZIF-8-BPPO, (4) ZIF-8-MBPPO-16, and (5) synthesized ZIF-8 powder. 73

93 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support Figure 3-5. Pure water flux and pore size of BPPO membranes as a function of exposure time to EDA vapour Fast in Situ Seeding For ZIF-8, the nucleation rate controls the crystallization process [40, 41], which is crucial to the particle sizes. Due to the generation of localized extremely high temperatures and pressures, the fast in situ seeding method [38], employed in this study, results in a high nucleation rate and subsequently in small-sized crystals. The introduction of ammonium hydroxide, in addition, can deprotonate organic ligands and thereby accelerate ligand exchange reactions, resulting in an even higher nucleation rate and consequently in a smaller final crystal size. As shown in Figure 3-6a, the XRD pattern of the particles collected after 1h seeding is exactly same as the simulated SODtype ZIF-8 structure, which confirms the formation of pure crystalline ZIF-8 phase. The average crystal size was ~ 20 nm estimated from the full width at half maximum of the (011) peak using the Scherrer s equation. The formation and size of the ZIF-8 crystals were further confirmed by TEM, as shown in Figure 3-6b. The diffraction rings of the different planes, shown in inset of Figure 3-6b, are in good agreement with the XRD peaks of ZIF-8. Spherical particles of ~20 nm observed in the TEM image was also 74

94 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support consistent with crystallite size obtained from XRD patterns. It should be noted that the small ZIF-8 nanocrystals are easily damaged in the high energy of the electron beam of a TEM [42]. Type I nitrogen sorption isotherms (Figure 3-6c) were observed representing the microporous nature of the as-synthesized ZIF-8 crystals. The second step (at P/Po > 0.8) observed in the isotherm with an obvious adsorption desorption hysteresis loop is attributed to interparticle mesopores. The micropore volume of the ZIF-8 nanocrystals is 0.74 cm³/g, and the BET and Langmuir surface areas are 1146 and 1715 m²/g, respectively. Figure 3-6. XRD pattern (a), TEM image and SAED pattern (inset) (b) and nitrogen sorption isotherm (c) of as-synthesized ZIF-8 nanocrystals. 75

95 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support Figure 3-7. SEM images of (a) cross-section and (b) surface of the ZIF-8-MBPPO Supported ZIF-8 membrane After the substrate modification, an ultrathin ZIF-8 membrane of about 200 nm is formed on top of the polymer support (Figure 3-7). From the membrane cross section (Figure 3-7a), it can also be observed that a fraction of ZIF-8 crystals formed inside the porous polymer support. This can be due to the diffusion of the EDA vapour into the polymer sublayer which creates additional active sites for ZIF-8 nucleation. Energydispersive X-ray spectroscopy (EDS) line-scan analysis confirmed the existence of ZIF- 8 within the support sublayer as zinc was detected up to ~150 nm underneath the membrane surface (Figure 3-8). Growth of the crystals partially inside the support could improve the membrane structural integrity. The top-view image (Figure 3-7b) shows that the support surface was covered entirely with a continuous and compact ZIF-8 layer without any visible defects such as pinholes or cracks. This very thin, dense and defectfree ZIF-8 membrane on the support with large pore size is desirable for highperformance gas separation. By contrast, separate ZIF-8 crystals and crystal islands formed if the support surface was not modified with EDA before ZIF-8 crystallization (Figure 3-9). This difference in the morphology between ZIF-8-MBPPO and ZIF-8- BPPO provides strong evidence of the important role of EDA modification in growing a thin and compact ZIF-8 layer. The intensity of the XRD (Figure 3-10) peaks of ZIF-8 76

96 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support in ZIF-8-MBPPO is much higher than those in ZIF-8-BPPO, indicating that the ZIF-8 layer has higher crystallinity and better surface coverage. Benzyl bromine groups ( CH2Br) in the BPPO are readily transformed to primary amine groups ( NH2) in the modification process. The obtained primary amine groups can subsequently react with Zn 2+, as reported by Liu et al. [43], where stable zinc complexes are formed from Zn 2+ coordinated with monoamine. The FTIR results also show that primary amine groups generated by the EDA treatment have been consumed in the reaction with Zn 2+ during ZIF-8 nucleation step where the broad band ascribed to the N H stretching disappeared (Figure 3-3). Figure 3-8. EDS line scan across ZIF-8-MBPPO-16 cross-section for the zinc atoms. On the basis of experimental results described above, it is clear that EDA can act as a covalent link between the ZIF-8 crystals and support, providing a large number of nucleation sites for the growth of the ZIF-8 layer. The existence of a strong interaction between ZIF-8 and MBPPO support was also confirmed by TGA results (Figure 3-4) in which a considerable shift from 310 C to 358 C is observed between the degradation peaks of MBPPO and ZIF-8-MBPPO; whereas the BPPO and its counterpart ZIF-8-77

97 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support BPPO membrane show only a slight difference in their degradation peaks. To gain an insight into the effect of EDA on the growth of ZIF-8 on the BPPO support, ZIF-8- MBPPO-4 and ZIF-8-MBPPO-10 were prepared and their SEM images were compared to that of ZIF-8-MBPPO-16 (Figure 3-9). These images reveal that the shorter EDA exposure resulted in ZIF-8 membranes with larger defects and pinholes. This can be explained by considering the multiple roles of EDA. By increasing the EDA exposure time, the reaction extent increases, as confirmed by FTIR (Figure 3-11), resulting in aforementioned a higher number of nucleation sites (amine groups). In addition, ethylenediamination induces a change in the membrane microstructure. The effect of pore size of porous support on the formation of ZIF membranes was previously reported [44, 45]. Increasing the EDA modification time leads to increased crosslinking and reduced substrate pore size, which is essential for the formation of a thin and defect-free ZIF-8 membrane. The corresponding pore size of MBPPO-16 is 11.4 nm. 78

98 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support Figure 3-9. SEM images of the (a) surface and (b) cross section of the BPPO support, (c) surface of the ZIF-8-BPPO, (d) surface of the ZIF-8-MBPPO-4, (e) surface of the ZIF-8-MBPPO-10, (f) surface and (g, h) cross-section of the ZIF-8-MBPPO

99 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support Figure XRD patterns of the membranes and simulated ZIF-8 powder. Figure FTIR ATR spectra of (1) untreated BPPO support, BPPO modified with EDA-vapour for (2) 4 h (MBPPO-4), (3) 10 h (MBPPO-10), (4) 16 h (MBPPO-16), (5) ZIF-8-MBPPO-16, (6) synthesized ZIF-8 powder Single gas performance To further evaluate the quality of the ZIF-8 membranes, single gas (H2, N2 and CO2) permeation experiments were carried out, and the results are summarized in Table 3-1. ZIF-8-BPPO without modification showed no gas separation property because no 80

100 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support continuous ZIF-8 film was formed. The result also confirms the formation of a continuous and compact ZIF-8 film on the modified BPPO substrate when it is sufficiently aminated. Lower amination time led to a low-quality ZIF-8 layer on the BPPO substrate (ZIF-8-MBPPO-4 and ZIF-8-MBPPO-10). ZIF-8-MBPPO composite membranes prepared with longer amination times show excellent gas selectivities. ZIF- 8-MBPPO-16 exhibited H2/CO2 and H2/N2 ideal selectivities of 12.8 and 9.7, respectively; it also had H2 permeance as high as mol.m -2.s -1.Pa -1 (Figure 3-12). Furthermore, all H2 permeances and ideal selectivities of H2/CO2 and H2/N2 are similar for the membranes obtained from different synthesis batches (Table 3-2), indicating the good reproducibility of the reported synthesis strategy. Table 3-1. Single gas permeances and ideal selectivities for the composite membranes at 25 ⁰C and 1 bar. Sample Permeance (10-7 mol.m -2.s - Ideal selectivity 1.Pa -1 ) H2 H2/N2 H2/CO2 BPPO MBPPO ZIF-8@BPPO ZIF-8@MBPPO ZIF-8@MBPPO ZIF-8@MBPPO

101 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support Table 3-2. Single gas permeances and ideal selectivities at 25 ⁰C and 1 bar of 3 tested ZIF-8-MBPPO-16 membranes showing the reproducibility of membrane synthesis and testing. Sample Permeance (10-7 Ideal selectivity mol.m -2.s -1.Pa -1 ) H2 H2/N2 H2/CO2 Original membrane Membrane Membrane This membrane is amongst the best ZIF-8 membranes reported previously (Table 3-3). For instance, at a similar H2/N2 selectivity, the ZIF-8 membrane prepared in this study had two orders of magnitude higher H2 gas permeance than those prepared by the epitaxy method [18, 46]. Figure Single gas permeances of ZIF-8-MBPPO-16 as a function of kinetic diameter of gas molecule. 82

102 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support It is worth mentioning that the N2 permeance is higher than CO2 permeance despite CO2 having a smaller kinetic diameter than N2. This behaviour can be attributed to the peculiar structure of ZIF-8 and to the combined linear structure and permanent dipole moment of CO2 [6, 47]. The small CO2 permeance may also be related to partially coordinated organic ligand (Hmim) molecules present in the ZIF-8 crystals as it was previously shown that this molecule is able to strongly coordinate with CO2 gas [47]. The similar behaviour was also observed in other ZIF materials [48-51]. The high permeance for hydrogen as well as high hydrogen selectivity in this study is due to the ultrathin ZIF layer (~ 200 nm) and the absence of pinholes or defects, which are attributed to vapour phase amination of asymmetrically porous BPPO support. Another possible reason for the high permeance may be the highly porous and asymmetric structure of the support, which minimises the overall hydraulic resistance of the permeate flow through the membrane structure. The formation of a thin and compact ZIF-8 layer in this study is essentially attributed to the enriched heterogeneous nucleation density and the reduction of the support pore size via EDA-vapour crosslinking. Thus, this work demonstrates a new strategy that can be applied to the formation of polymer-supported ZIF membranes through rational design and chemical modification of the thin polymer supports. 83

103 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support Table 3-3. Comparison of gas permeation properties (H2 permeance, H2/N2 and H2/CO2 selectivity) of ZIF-8 membranes on inorganic and polymeric supports reported in recent literature. Permeance Selectivity (10-7 mol/m 2.s.Pa) Support Synthesis Method Thickness (µm) T [ C] H 2 N 2 CO 2 H 2/N 2 H 2 /CO 2 Ref. Polymeric support BPPO Surface chemistry and pore structure modification ~0.3 Room This work PVDF Nylon ~30 Room [14] NR * 4.3 NR [13] Nylon NR 4.6 NR [52] Torlon Interfacial microfluidic ~ NR NR NR NR [53] membrane processing (IMMP) PES Secondary growth NR 9.9 ** NR [54] Chemical modification Contradiffusion Contradiffusion PSf In situ followed by layer-bylayer NR 1.06 NR 3.8 [17] PAN Surface chemical modification NR NR 0.44 NR 6.85 ** [32] 84

104 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support Permeance Selectivity (10-7 mol/m 2.s.Pa) Support Synthesis Method Thickness (µm) T [ C] H 2 N 2 CO 2 H 2/N 2 H 2 /CO 2 Ref. Polymeric support PSf In situ NR 12.4 ** NR [55] Inorganic support Alumina/PTFE Seeded growth Alumina Hot support hollow seeding NR 9.4 NR [56] [57] fiber Alumina hollow fiber Alumina tube Repeated growth APTES and cycling precursors [58] 2 Room [59] Alumina Surface chemical modification [12] γ-al 2O 3 AAO Surface chemical modification Fast in situ seeding and secondary growth 20 Room [10] [38] Al 2O 3 tube Titania Repeated synthesis Direct synthesis NR NR 20.7 [60] [61] Alumina LBL*** ~ [18] *NR: Not reported; **Mixed gas separation; ***LBL: Layer-by-Layer 85

105 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support Summary In summary, we have successfully prepared a compact, ultra-thin ZIF-8 layer on an asymmetric polymeric substrate by chemical vapour modification of the surface chemistry and pore structure. The thickness of the membrane fabricated by simultaneous modification of surface chemistry and pore structure is one of the thinnest ever reported. In addition, we have also demonstrated the influence of the surface microstructure and chemical composition of the polymer substrate on the formation of a continuous ZIF-8 layer. Vapour-phase EDA has been used to simultaneously tailor the chemical nature and pore size of the surface of BPPO for the successful growth of ZIF-8 membrane. The EDA treatment produced a large number of nucleation sites and modified the BPPO pore structure, promoting the formation of a thin ZIF-8 layer. The ZIF-8 membrane exhibits ideal selectivities (H2/CO2: 12.8; H2/N2: 9.8) and permeance ( mol.m -2.s -1.Pa - 1 ) which is among the highest reported so far. The proposed chemical vapour modification followed by fast in situ synthesis provides a rapid, convenient and effective route for preparing thin yet continuous and defect free ZIF membranes on the surface of polymeric substrates. 86

106 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support References [1] O.M. Yaghi, M. O'Keeffe, N.W. Ockwig, H.K. Chae, M. Eddaoudi, J. Kim, Reticular synthesis and the design of new materials, Nature, 423 (2003) [2] H. Deng, C.J. Doonan, H. Furukawa, R.B. Ferreira, J. Towne, C.B. Knobler, B. Wang, O.M. Yaghi, Multiple functional groups of varying ratios in metal-organic frameworks, Science, 327 (2010) [3] N. Stock, S. Biswas, Synthesis of metal-organic frameworks (MOFs): routes to various MOF topologies, morphologies, and composites, Chemical reviews, 112 (2011) [4] I. Ahmed, S.H. Jhung, Composites of metal organic frameworks: Preparation and application in adsorption, Materials Today, 17 (2014) [5] C. Chizallet, S. Lazare, D. Bazer-Bachi, F. Bonnier, V. Lecocq, E. Soyer, A.-A. Quoineaud, N. Bats, Catalysis of transesterification by a nonfunctionalized metal organic framework: Acido-basicity at the external surface of ZIF-8 probed by FTIR and ab initio calculations, J. Am. Chem. Soc., 132 (2010) [6] L. Dumee, L. He, M. Hill, B. Zhu, M. Duke, J. Schutz, F. She, H. Wang, S. Gray, P. Hodgson, L. Kong, Seeded growth of ZIF-8 on the surface of carbon nanotubes towards self-supporting gas separation membranes, J. Mater. Chem. A, 1 (2013) [7] G. Lu, J.T. Hupp, Metal organic frameworks as sensors: a ZIF-8 based Fabry Pérot device as a selective sensor for chemical vapors and gases, J. Am. Chem. Soc., 132 (2010) [8] Q. Liu, Z.-X. Low, L. Li, A. Razmjou, K. Wang, J. Yao, H. Wang, ZIF-8/Zn 2 GeO 4 nanorods with an enhanced CO 2 adsorption property in an aqueous medium for photocatalytic synthesis of liquid fuel, J. Mater. Chem. A, 1 (2013) [9] J. Yao, H. Wang, Zeolitic imidazolate framework composite membranes and thin films: synthesis and applications, Chem. Soc. Rev., 43 (2014) [10] Y. Liu, N. Wang, J.H. Pan, F. Steinbach, J.r. Caro, In situ synthesis of mof membranes on ZnAl-CO3 LDH buffer layer-modified substrates, J. Am. Chem. Soc., 136 (2014)

107 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support [11] H.T. Kwon, H.-K. Jeong, In situ synthesis of thin zeolitic imidazolate framework ZIF-8 membranes exhibiting exceptionally high propylene/propane separation, J. Am. Chem. Soc., 135 (2013) [12] M.C. McCarthy, V. Varela-Guerrero, G.V. Barnett, H.-K. Jeong, Synthesis of zeolitic imidazolate framework films and membranes with controlled microstructures, Langmuir, 26 (2010) [13] J. Yao, D. Dong, D. Li, L. He, G. Xu, H. Wang, Contra-diffusion synthesis of ZIF- 8 films on a polymer substrate, Chem. Commun., 47 (2011) [14] W. Li, Q. Meng, C. Zhang, G. Zhang, Metal organic framework/pvdf composite membranes with high H2 permselectivity synthesized by ammoniation, Chemistry-A European Journal, (2015). [15] H. Bux, F. Liang, Y. Li, J. Cravillon, M. Wiebcke, J. Caro, Zeolitic imidazolate framework membrane with molecular sieving properties by microwave-assisted solvothermal synthesis, J. Am. Chem. Soc., 131 (2009) [16] M.N. Shah, M.A. Gonzalez, M.C. McCarthy, H.-K. Jeong, An unconventional rapid synthesis of high performance metal organic framework membranes, Langmuir, 29 (2013) [17] D. Nagaraju, D.G. Bhagat, R. Banerjee, U.K. Kharul, In situ growth of metalorganic frameworks on a porous ultrafiltration membrane for gas separation, J. Mater. Chem. A, 1 (2013) [18] O. Shekhah, R. Swaidan, Y. Belmabkhout, M. du Plessis, T. Jacobs, L.J. Barbour, I. Pinnau, M. Eddaoudi, The liquid phase epitaxy approach for the successful construction of ultra-thin and defect-free ZIF-8 membranes: pure and mixed gas transport study, Chem. Commun., 50 (2014) [19] E. Biemmi, C. Scherb, T. Bein, Oriented growth of the metal organic framework Cu3 (BTC) 2 (H2O)3. x H2O tunable with functionalized self-assembled monolayers, J. Am. Chem. Soc., 129 (2007) [20] S. Hermes, D. Zacher, A. Baunemann, C. Wöll, R.A. Fischer, Selective growth and MOCVD loading of small single crystals of MOF-5 at alumina and silica surfaces modified with organic self-assembled monolayers, Chem. Mater., 19 (2007)

108 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support [21] E. Biemmi, A. Darga, N. Stock, T. Bein, Direct growth of Cu3 (BTC)2 (H2O)3. xh2o thin films on modified QCM-gold electrodes: Water sorption isotherms, Microporous Mesoporous Mater., 114 (2008) [22] H. Guo, G. Zhu, I.J. Hewitt, S. Qiu, Twin copper source growth of metal organic framework membrane: Cu3 (BTC)2 with high permeability and selectivity for recycling H2, J. Am. Chem. Soc., 131 (2009) [23] A. Huang, W. Dou, J.r. Caro, Steam-stable zeolitic imidazolate framework ZIF-90 membrane with hydrogen selectivity through covalent functionalization, J. Am. Chem. Soc., 132 (2010) [24] C. Hou, Q. Xu, J. Peng, Z. Ji, X. Hu, (110) Oriented ZIF-8 thin films on ITO with controllable thickness, ChemPhysChem, 14 (2013) [25] S. Hermes, F. Schröder, R. Chelmowski, C. Wöll, R.A. Fischer, Selective nucleation and growth of metal organic open framework thin films on patterned COOH/CF3-terminated self-assembled monolayers on Au(111), J. Am. Chem. Soc., 127 (2005) [26] O. Shekhah, H. Wang, S. Kowarik, F. Schreiber, M. Paulus, M. Tolan, C. Sternemann, F. Evers, D. Zacher, R.A. Fischer, C. Wöll, Step-by-step route for the synthesis of metal organic frameworks, J. Am. Chem. Soc., 129 (2007) [27] E. Biemmi, C. Scherb, T. Bein, Oriented growth of the metal organic framework Cu3(BTC)2(H2O)3 xh2o tunable with functionalized self-assembled monolayers, J. Am. Chem. Soc., 129 (2007) [28] D. Vasanth, G. Pugazhenthi, R. Uppaluri, Fabrication and properties of low cost ceramic microfiltration membranes for separation of oil and bacteria from its solution, J. Membr. Sci., 379 (2011) [29] M.A. Carreon, S. Li, J.L. Falconer, R.D. Noble, Alumina-supported SAPO-34 membranes for CO2/CH4 separation, J. Am. Chem. Soc., 130 (2008) [30] X. Zhang, Y. Liu, S. Li, L. Kong, H. Liu, Y. Li, W. Han, K.L. Yeung, W. Zhu, W. Yang, J. Qiu, New membrane architecture with high performance: ZIF-8 membrane supported on vertically aligned ZnO nanorods for gas permeation and separation, Chem. Mater., 26 (2014)

109 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support [31] X. Zhang, Y. Liu, L. Kong, H. Liu, J. Qiu, W. Han, L.-T. Weng, K.L. Yeung, W. Zhu, A simple and scalable method for preparing low-defect ZIF-8 tubular membranes, J. Mater. Chem. A, 1 (2013) [32] W. Li, Z. Yang, G. Zhang, Z. Fan, Q. Meng, C. Shen, C. Gao, Stiff metal organic framework polyacrylonitrile hollow fiber composite membranes with high gas permeability, J. Mater. Chem. A, 2 (2014) [33] L. Shao, C.-H. Lau, T.-S. Chung, A novel strategy for surface modification of polyimide membranes by vapor-phase ethylenediamine (EDA) for hydrogen purification, Int. J. Hydrogen Energy, 34 (2009) [34] C.H. Lau, B.T. Low, L. Shao, T.-S. Chung, A vapor-phase surface modification method to enhance different types of hollow fiber membranes for industrial scale hydrogen separation, Int. J. Hydrogen Energy, 35 (2010) [35] T. Xu, D. Wu, L. Wu, Poly (2, 6-dimethyl-1, 4-phenylene oxide)(ppo) a versatile starting polymer for proton conductive membranes (PCMs), Prog. Polym. Sci., 33 (2008) [36] B. Zhao, R.W. Fu, M.Q. Zhang, B. Zhang, W. Zeng, M.Z. Rong, Q. Zheng, Analysis of gas sensing behaviors of carbon black/waterborne polyurethane composites in low concentration organic vapors, J. Mater. Sci., 42 (2007) [37] D. Möckel, E. Staude, M.D. Guiver, Static protein adsorption, ultrafiltration behavior and cleanability of hydrophilized polysulfone membranes, J. Membr. Sci., 158 (1999) [38] M. He, J. Yao, Z.-X. Low, D. Yu, Y. Feng, H. Wang, A fast in situ seeding route to the growth of a zeolitic imidazolate framework-8/aao composite membrane at room temperature, RSC Adv., 4 (2014) [39] X. Tongwen, Y. Weihua, A novel positively charged composite membranes for nanofiltration prepared from poly(2,6-dimethyl-1,4-phenylene oxide) by in situ amines crosslinking, J. Membr. Sci., 215 (2003) [40] S.R. Venna, J.B. Jasinski, M.A. Carreon, Structural evolution of zeolitic imidazolate framework-8, J. Am. Chem. Soc., 132 (2010)

110 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support [41] B. Seoane, J.M. Zamaro, C. Tellez, J. Coronas, Sonocrystallization of zeolitic imidazolate frameworks (ZIF-7, ZIF-8, ZIF-11 and ZIF-20), CrystEngComm, 14 (2012) [42] J. Cravillon, R. Nayuk, S. Springer, A. Feldhoff, K. Huber, M. Wiebcke, Controlling zeolitic imidazolate framework nano-and microcrystal formation: Insight into crystal growth by time-resolved in situ static light scattering, Chem. Mater., 23 (2011) [43] P. Liu, Z.-X. Lian, J. Cai, Crystal structures of two monoamine-coordinated zinc complexes obtained via solid-vapor reaction, J. Chem. Crystallogr., 36 (2006) [44] Z. Xie, J. Yang, J. Wang, J. Bai, H. Yin, B. Yuan, J. Lu, Y. Zhang, L. Zhou, C. Duan, Deposition of chemically modified α-al2o3 particles for high performance ZIF-8 membrane on a macroporous tube, Chem. Commun., 48 (2012) [45] J. Yang, Z. Xie, H. Yin, J. Wang, J. Xu, J. Wang, J. Lu, D. Yin, Y. Zhang, Simultaneous surface and pore structure modification for growth of ZIF-8 membranes on coarse macroporous tube, Microporous Mesoporous Mater., 198 (2014) [46] O. Shekhah, M. Eddaoudi, The liquid phase epitaxy method for the construction of oriented ZIF-8 thin films with controlled growth on functionalized surfaces, Chem. Commun., 49 (2013) [47] G. Xu, J. Yao, K. Wang, L. He, P.A. Webley, C.-s. Chen, H. Wang, Preparation of ZIF-8 membranes supported on ceramic hollow fibers from a concentrated synthesis gel, J. Membr. Sci., (2011) [48] A. Huang, Y. Chen, N. Wang, Z. Hu, J. Jiang, J. Caro, A highly permeable and selective zeolitic imidazolate framework ZIF-95 membrane for H2/CO2 separation, Chem. Commun., 48 (2012) [49] C. Yuan, Q. Liu, H. Chen, A. Huang, Mussel-inspired polydopamine modification of supports for the facile synthesis of zeolite LTA molecular sieve membranes, RSC Adv., 4 (2014) [50] N. Wang, Y. Liu, Z. Qiao, L. Diestel, J. Zhou, A. Huang, J. Caro, Polydopaminebased synthesis of a zeolite imidazolate framework ZIF-100 membrane with high H2/CO2 selectivity, J. Mater. Chem. A, 3 (2015)

111 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support [51] X. Dong, K. Huang, S. Liu, R. Ren, W. Jin, Y.S. Lin, Synthesis of zeolitic imidazolate framework-78 molecular-sieve membrane: defect formation and elimination, J. Mater. Chem., 22 (2012) [52] M. He, J. Yao, L. Li, Z. Zhong, F. Chen, H. Wang, Aqueous solution synthesis of ZIF-8 films on a porous nylon substrate by contra-diffusion method, Microporous Mesoporous Mater., 179 (2013) [53] A.J. Brown, N.A. Brunelli, K. Eum, F. Rashidi, J. Johnson, W.J. Koros, C.W. Jones, S. Nair, Interfacial microfluidic processing of metal-organic framework hollow fiber membranes, Science, 345 (2014) [54] L. Ge, W. Zhou, A. Du, Z. Zhu, Porous polyethersulfone-supported zeolitic imidazolate framework membranes for hydrogen separation, J. Phys. Chem. C, 116 (2012) [55] F. Cacho-Bailo, B. Seoane, C. Téllez, J. Coronas, ZIF-8 continuous membrane on porous polysulfone for hydrogen separation, J. Membr. Sci., 464 (2014) [56] P. Hu, Y. Yang, Y. Mao, J. Li, W. Cao, Y. Ying, Y. Liu, J. Lei, X. Peng, Room temperature synthesis of ZIF-8 membranes from seeds anchored in gelatin films for gas separation, CrystEngComm, (2015). [57] K. Tao, L. Cao, Y. Lin, C. Kong, L. Chen, A hollow ceramic fiber supported ZIF- 8 membrane with enhanced gas separation performance prepared by hot dip-coating seeding, J. Mater. Chem. A, 1 (2013) [58] G. Xu, J. Yao, K. Wang, L. He, P.A. Webley, C.-s. Chen, H. Wang, Preparation of ZIF-8 membranes supported on ceramic hollow fibers from a concentrated synthesis gel, J. Membr. Sci., 385 (2011) [59] K. Huang, Z. Dong, Q. Li, W. Jin, Growth of a ZIF-8 membrane on the inner-surface of a ceramic hollow fiber via cycling precursors, Chem. Commun., 49 (2013) [60] Y. Zhu, Q. Liu, J. Caro, A. Huang, Highly hydrogen-permselective zeolitic imidazolate framework ZIF-8 membranes prepared on coarse and macroporous tubes through repeated synthesis, Sep. Purif. Technol., (2015). 92

112 Chapter 3 Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of polymeric support [61] H. Bux, F. Liang, Y. Li, J. Cravillon, M. Wiebcke, J.r. Caro, Zeolitic imidazolate framework membrane with molecular sieving properties by microwave-assisted solvothermal synthesis, J. Am. Chem. Soc., 131 (2009)

113 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Aqueous Phase Synthesis of ZIF-8 Membrane with Controllable Location on an Asymmetrically Porous Polymer Substrate Overview In chapter 3, ultrasonication was utilized to generate intense, localized heating to promote ZIF-8 growth. Although the prepared membrane showed excellent separation properties, such ultrasonication-assisted method may not be applicable for large-scale membrane sample preparation. In this chapter, we have demonstrated a simple, scalable, and environmentally friendly route for controllable fabrication of continuous, well-intergrown ZIF-8 on a flexible polymer substrate via contra-diffusion method in conjunction with chemical vapour modification of the polymer surface. The combined chemical vapour modification and contra-diffusion method resulted in controlled formation of a thin, defect-free and robust ZIF-8 layer on one side of the support in aqueous solution at room temperature. The ZIF-8 membrane exhibited propylene permeance of mol m -2 s -1 Pa and excellent selective permeation properties; after post heat-treatment, the membrane showed ideal selectivities of C3H6/C3H8 and H2/C3H8 as high as 27.8 and 2259, respectively. The new synthesis approach holds a promise for further development for the fabrication of high-quality polymer-supported ZIF membranes for practical separation applications. This chapter has been reformatted from the following published manuscript: Shamsaei, E., Lin, X., Low, Z.X., Abbasi, Z., Hu, Y., Liu, J.Z. and Wang, H.. Aqueous phase 94

114 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate. ACS Applied Materials & Interfaces, 2016, 8(9), pp Introduction Zeolitic imidazolate frameworks (ZIFs), a subclass of metal organic frameworks (MOFs), are porous crystalline hybrid materials consisting of imidazolate ligands (Im) bridging tetrahedral metal ions (e.g., Zn, Co) [1]. They closely resemble the topologies of zeolites, due to the M-Im-M (M = Zn, Co) bond angle of 145, which is close to the T-O-T (T = Al, Si, P) angle ( ) in zeolites [2, 3]. ZIFs show properties that combine the attractive features of both MOFs and zeolites such as tunable pore size and chemistry, large internal surface area and relatively good thermal and chemical stability [4, 5]. These properties make ZIFs excellent candidates for the fabrication of molecular sieving membranes for gas separation [6-8]. ZIF-8 membranes, for example, have been reported to be capable of molecularly discriminating propylene (~4.0 Å) from propane (~ 4.3 Å) since the effective pore aperture size of ZIF-8 falls in the range of Å (larger than its crystallographic value of 3.4 Å, owing to the swaying effect of the ligands) [2, 9-11]. ZIFs have been widely used to fabricate the so-called mixed matrix membranes (MMMs, consisting of pre-synthesized ZIF particles dispersed in a polymeric matrix) to afford a solution to go beyond the Robeson's upper-bound trade-off limit of the polymeric membranes [6, 12-15]. While MMMs have been shown to enhance the permeation properties of polymeric membranes, further enhancements were made using in-situ synthesized ZIF membranes [16]. For example, ZIF-8 supported membranes prepared by Pan et al. [10] showed superior propylene/propane permselectivity (permeability of propylene up to 200 barrers and a propylene/propane separation factor up to 50) compared to those of ZIF-8 MMMs, as for instance the ZIF-8/ 6FDA DAM 95

115 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate polyimide (a propylene permeability of 56.2 barrer and propylene/propane ideal selectivity of 31.0) reported by Zhang et al. [15]. To date, several synthesis methods have been reported for the formation of ZIF films on various substrates [6, 17]. In particular, polymer-supported ZIF membranes are of great interest as they potentially combine the advantages of both polymer membranes (e.g. easy processing and low cost) and ZIFs (e.g. high selectivity). In principle, the growth of ZIF films on flexible polymeric substrates can be easily achieved due to favourable chemical interaction between the polymer and the organic ligand of ZIFs. Nagaraju et al. [18] and Cacho-Bailo et al. [19] grew ZIF-8 on a porous polysulfone using in situ (direct) growth. However, although the in situ synthesis is a simple method that allows for simultaneous nucleation, deposition and crystal growth, it is not very effective in preparing continuous ZIF membranes due to limited heterogeneous nucleation sites on the substrate [20]. Alternatively, Ge at al. [21] used secondary seeded growth to fabricate a continuous ZIF-8 film on an asymmetrically porous polyethersulfone substrate. Secondary seeded growth has been shown to effectively induce controlled ZIF growth on the polymer support, but the resulting ZIF layer often suffers from weak adhesion to the support, leading to membrane delamination [22]. The surface modification has been commonly used to functionalize the support, thereby promoting heterogeneous nucleation and enhancing the ZIF-to-substrate adhesion strength [23-25]. Very recently, we successfully developed a new strategy of vapour phase modification to introduce amine groups and reduce surface pore sizes of the polymer support; such surface modification enabled fast formation of a continuous ZIF-8 ultrathin layer in the presence of ammonium hydroxide (as a deprotonating agent) under sonication for only 3 minutes [23]. However, the sonication-induced crystallization offers limited control over the ZIF-8 crystal sizes and intergrowth, and thus membrane properties. Our group was one of the first groups to report 96

116 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate contra-diffusion synthesis, which self-limits growth of ZIF films on porous substrate, and has great potential to offer better control in the membrane fabrication [11, 26-28]. However, the growth of ZIF films via contra diffusion method depends on the surface properties and porous structure of support; the formation of ZIF layer on both sides of support or within porous channels of support has been reported [26, 27]. To achieve better control over the membrane position, Brown et al. recently introduced an interfacial synthesis approach [29]. The control over the membrane position relies on employing an oil/ water system, in which crystals grow at the interfaces between the two immiscible solvents. The resulting ZIF-8 membrane exhibited high gas separation performance with H2/C3H8 and C3H6/C3H8 separation factors as high as 370 and 12, respectively. In this work, we report a simple, effective and environmental friendly method for the fabrication of high-quality ZIF-8 membrane with controllable location on a polymer substrate in aqueous solution. Our synthesis method is based on contra-diffusion (CD) concept in conjunction with chemical vapour modification (hereafter chemical vapour modification-contra diffusion method). A flat sheet asymmetric bromomethylated poly (2, 6-dimethyl-1, 4-phenylene oxide) (BPPO) ultrafiltration membrane was prepared via phase inversion and employed as the support for growing a thin ZIF-8 layer via contra-diffusion after modification. We have selected BPPO for ZIF-8 growth because of its outstanding membrane formation and mechanical properties as well as excellent hydrolytic stability [30]. It can also be easily functionalized and crosslinked due to the abundant highly reactive CH2Br groups. Using vapour-phase ethylenediamine (EDA), we have previously shown that amine functional groups can be covalently attached selectively on the top layer of the support without affecting the sublayer structure [23]. The presence of the covalent link (amine groups) can be a driving factor for maintaining a high concentration of the metal ions selectively near the support surface. When combined with the slow diffusion of the ligand in contra 97

117 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate diffusion process, the approach can lead to well-controlled crystal growth in the vicinity of the support surface. This results in the formation of thin, defect-free and robust ZIF-8 layer on one side of the support at room temperature without the addition of deprotonating agents, which has proven to be challenging when using other reported synthesis procedures [11, 26, 31, 32]. Materials and Methods Chemicals BPPO, BPPO was kindly supplied by Tianwei Membrane Co. Ltd., Shandong of China. Ethylenediamine (EDA, 99.5%),1-methyl-2-pyrrolidone (NMP, 99.5%), zinc acetate dihydrate (Zn(CH3COO)2.2H2O, 98%) and 2-methylimidazole (Hmim, C4H6N2, 99%) were purchased from Sigma-Aldrich, Australia and used as received. Methanol (absolute) was purchased from Merck, Australia. The water used for the experiments was purified with a water purification system (Milli- Q integral water purification system, Merck Millipore) with a resistivity of 18.2 MΩ/cm. Distilled water was obtained from a laboratory water distillation still (Labglass Aqua III) Sample preparation BPPO support ultrafiltration membranes were prepared via non-solvent induced phase separation at room temperature [33]. The casting solution was prepared by dissolving 15 wt. % of BPPO in NMP for 12 h with mechanical stirring at 200 rpm. The homogenous solution was left to degas for 10 h before use. Subsequently, the solution was cast on a clean glass plate using an adjustable micrometre film applicator (Paul N. Gardner Co., Inc. USA) with a gap of 200 μm at room temperature (22 ± 2 C) and immediately immersed in a coagulation bath of deionized water. After peeling off from the glass plate, the membranes were removed from the bath, washed and kept in fresh deionized water (DI) for at least one day to thoroughly remove the residual solvents. 98

118 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate The vapour-phase EDA modification was conducted in a custom-made container according to method developed in chapter 3 [23]. In brief, 20 ml of EDA was allowed to vaporize, and stabilized for 1 h. The support membranes were quickly placed inside the containment with the top layer exposed and suspended above the EDA solution. After surface modification at room temperature for 16 h, the surface modified membranes were removed from the containment and immediately washed with pure water to completely remove the residual EDA. The resultant membrane were denoted as BPPO-EDA. Figure 4-1. Digital photograph of a home-made contra-diffusion cell. For preparation of the BPPO supported ZIF-8 membranes, the modified BPPO supporting membrane was cut into 32 mm diameter discs, which were then mounted on a home-made setup (Figure 4-1), where the zinc acetate solution and Hmim solution were separated by the supporting membrane. Zinc acetate solution was prepared by dissolving 0.09 g of Zn (CH3COO)2.2H2O (0.5 mmol) in 20 ml of deionized water, and Hmim solution was prepared by adding g of Hmim (8 mmol) in 20 ml of deionized water. The designed Hmim: Zn 2+ molar ratios in the system 99

119 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate was 16 and was kept constant in our study. After crystallization at room temperature (22 ± 2 C) for min, the membrane samples were taken out and rinsed with DI water several times. Finally, the composite membranes were dried in ambient conditions for 24 h, followed by heating at C for 2 h before tests. The resulting samples were denoted as ZIF-8-BPPO-EDA-t-T, where t and T denote the crystallization time and heat treatment temperature, respectively Characterization Fourier Transform Infrared (FTIR) spectra of the membranes were recorded using an attenuated total reflectance (ATR) FTIR (Perkin Elmer, USA) in the range of cm -1 at an average of 20 scans with a resolution of 4 cm -1. Scanning electron microscopy (SEM; FEI Nova NanoSEM 450) with an X-ray detector (Bruker Nano GmbH, Germany) was used for imaging the surface and cross-sectional morphologies of membranes. Energy-dispersive X-ray spectroscopy (EDS) line-scan analysis of the membrane samples was conducted using EDX equipped in Nova NanoSEM 450 (Quantax 400 X-ray analysis system, Bruker, USA). The membranes were fractured in liquid nitrogen, fixed on stubs with double-sided carbon tape and then sputter coated with roughly 2 nm iridium (Ir) layer to ensure good electrical conductivity. The images were recorded at an accelerating voltage of 5 kv with different magnifications. Thermogravimetric analyses (TGA) were carried out on a SETARAM (TGA 92) device from 30 to 800 C at a heating rate of 10 C min -1 under air flow. Powder X-ray diffraction (XRD) patterns were measured using a Miniflex 600 diffractometer (Rigaku, Japan) with Cu Kα radiation (15 ma and 40 kv) at a scan rate of 2 min -1 with a step size of The XRD studies were carried out at room temperature. 100

120 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Single gas permeation test Figure 4-2. Schematic diagram of gas permeation set-up. The single gas permeation of composite membranes was measured using the pressure rise method [34]. The schematic of the single gas permeation setup is shown in Figure 4-2. To measure the gas permeation flux, the composite membrane (16 mm diameter disc) was attached to a porous stainless steel holder (pore size ~200 nm) using epoxy resin (Torr seal, Varian), and then placed inside a larger Pyrex tube and connected to a sensitive pressure transducer (MKS 628B Baratron) and a vacuum line. The effective remaining membrane area was 1 cm 2. For each single gas measurement, the pure single gas was fed to one side (feed) of the membrane while the other side (permeate) of the membrane was under vacuum. Since the feed side was at ambient pressure, a pressure difference of 1 atm was maintained between the permeate side and the feed side during permeation measurements. After allowing enough time to achieve a steady state conditions, the permeate side was shut off from vacuum and the pressure build-up of the permeating gas was measured by the pressure transducer and continuously recorded in a computer. To accomplish a 101

121 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate single test, the pressure was allowed to reach a few Torr. To repeat an analysis, the permeate side was evacuated again and then shut off from vacuum so as to record the pressure rise. All the gas permeation tests were performed at room temperature. The molar flow rate of the permeating gas was calculated based on the recorded pressure. The permeance, Pi, of each gas was calculated according to the following equation, Pi= (V/RTAΔp) (dp/dt) where, V is the volume of the permeate side that was obtained by calibration using a bubble flowmeter (m 3 ), R is the ideal gas constant (m 3 Pa K 1 mol 1 ), T is the temperature (K), p is the pressure difference across the membrane (Pa), A is the effective membrane area (m 2 ), and dp/dt is the rate of pressure rise in the permeate side (Pa/s). The ideal selectivity Sij is defined as the ratio of the two permeances Pi and Pj. Permeation data are average values recorded from at least three samples, which were prepared from different batches Measurements of the support pore size Polyethylene glycol (PEG) with a molecular weight of 300, 200, 100, 35, 20, and 10 kda (analytical grade, Sigma-Aldrich) was dissolved in deionized water to prepare 1 g L -1 aqueous solutions for the estimation of molecular weight cut off (MWCO) and solute rejection. Rejection analysis were performed at a pressure of 100 kpa. 20 ml of permeate was collected. The permeate and feed solution were both diluted by 10 times and then the concentration of each solution was measured via a total organic carbon analyser (TOC-LCSH, Shimadzu, Japan). The PEG rejection was calculated from the measured feed (Cf) and permeate (Cp) concentrations by R = (1 Cp/Cf)

122 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate All results represent average values for at least three repeated experiments with less than ±5% deviation. The pore size of the membrane was defined as the hydrodynamic diameter of PEG. The hydrodynamic radius of PEG can be calculated from the MWCO of the membrane by the following equation [35]: Solute Radius (nm) = MW 0.03 where MW is the lowest molecular weight of the PEG molecule which has a rejection of 90% in the ultrafiltration measurements. Results and Discussion Membrane support Figure 4-3 illustrates the synthesis of dense and defect-free polymer supported ZIF-8 membrane using chemical vapour modification-contra diffusion method. Figure 4-3. Schematic diagram of the preparation of a BPPO polymer supported ZIF-8 membrane using chemical vapour modification and subsequent contra diffusion synthesis. 103

123 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate As illustrated in the figure (step (1)), the surface chemistry and pore size of the top layer of the BPPO are modified by using EDA-vapour. Substitution of bromide functional group with amine groups during EDA-vapour modification was confirmed by FTIR (Figure 4-4). Figure 4-4. FTIR ATR spectra of the untreated BPPO support, BPPO modified with EDA-vapor (BPPO-EDA), BPPO-EDA supported ZIF-8 layer (ZIF-8-BPPO-EDA), and ZIF-8 powder. Upon modification, the peak at 586 cm -1 and 633 cm -1 attributed to the benzyl bromide ( CH2Br) groups (C Br stretching) almost disappear and a new broad band in the range of cm -1 emerges, which is attributed to the N H stretching and confirms the amination of BPPO. TGA results (Figure 4-5) also show that when the temperature is below 150 ºC, BPPO-EDA have more weight loss (~4%) than BPPO due to higher moisture residual in the membrane as a result of the introduction of hydrophilic amine groups. Additionally, the decomposition of EDA initiates at ~ 180 C, which is higher than the boiling point (117 C) of the EDA, indicating that there is an interaction between the EDA molecules and BPPO. Furthermore, SEM images (Figure 4-6a, b) show an obvious decrease in the size of the nanopores at the top surface of the membrane after EDA vapour-phase modification. The reduction in the pore size of the support (from 25.5 to 15 nm) after its modification was further confirmed by TOC analysis (see measurements of the support pore size in section 4-3-5). The changes in the surface microstructure can be attributed to 104

124 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate the partial cross-linking effect of the EDA, since the crosslinking causes tightening of the polymer network which reduces the pore size of the BPPO substrate. In addition, the final decomposition of the BPPO-EDA in the TGA results (Figure 4-5) is much slower than BPPO, which indicates a higher thermal stability due to partial crosslinking of the BPPO substrate. Note that partial crosslinking reduces the flexibility of the polymer support, which is favourable for avoiding ZIF layer cracking [6]. Figure 4-5. Thermogravimetric analysis (under air flow) of untreated BPPO and BPPO-EDA substrates. 105

125 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Figure 4-6. SEM images of untreated BPPO (a), vapour-phase-eda-modified BPPO (BPPO-EDA) (b), ZIF-8@BPPO-EDA grown for 60 min (c), 90 min (d), and 120 min (e, f). 106

126 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Supported ZIF-8 membrane The ZIF-8 membrane is formed on the pre-treated support by applying contra diffusion synthesis, in which the metal precursor solution and ligand (Hmim) solution are separated by the modified BPPO substrate (step (2) in Figure 4-3), at room temperature for various crystallization times. As demonstrated in chapter 3, the direct heterogeneous nucleation and growth of a dense ZIF-8 layer on untreated BPPO surface was unsuccessful (Figure 4-7). In fact, due to the fast diffusion of zinc ions through the pores of the unmodified support, crystallization occurs constantly inside the support channels, where there can be a high concentration of the reactant solutions, until the entire path through the ZIF-8 layer becomes plugged. A similar phenomenon was observed by Hara et al. in preparing copper-benzene tricarboxylate (Cu-BTC) or ZIF-8 membranes using porous α-alumina capillary substrate by applying a typical contra diffusion method [28, 32]. Figure 4-7. SEM images of ZIF-8 membranes grown for 2 h (a, b, c), 4 h (d, e, f), 6 h (g, h, i) via conventional contra-diffusion method using untreated BPPO substrate. 107

127 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Figure 4-8. XRD patterns of ZIF-8@BPPO-EDA membranes as a function of growth time. After modification of the BPPO with EDA-vapour before contra diffusion synthesis, a compact ZIF-8 layer was selectively formed on only one side (pre-treated side) of the support. Figure 4-6 and Figure 4-8 show the SEM images and XRD patterns of the ZIF-8 membranes grown for different lengths of time. As shown in Figure 4-6c and Figure 4-8, a large amount of ZIF-8 crystals with clear facets is observed on the modified support even after 60 min of the contra diffusion synthesis at room temperature. Nanopores of the skin layer of the support are still observable through inter-crystalline gaps between the ZIF-8 crystals in the high magnification image (inset in Figure 4-6c). With increasing the reaction time to 90 min, a dense and continuous ZIF-8 layer is formed on the modified BPPO skin layer, as shown in Figure 4-6e. Eventually, after 120 min of reaction, a layer of well intergrown ZIF-8 crystals with rhombic dodecahedron 108

128 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate morphology and a thickness of about 2 µm fully covered the support surface without any visible defects such as pinholes or cracks (Figure 4-6e, f and Figure 4-9). There are only few studies reporting such a thin continuous polycrystalline ZIF-8 film [24, 36] and most of the membranes prepared by the conventional in situ methods are too thick (in the range of tens of micrometers), showing lower gas flux through the membranes [11, 37]. The continuous thin ZIF-8 membranes remained unchanged even with further growth, demonstrating the selflimiting crystal growth, in which the crystals continue to grow only if the metal ions and the ligand molecules are in contact. Figure 4-9. SEM images of ZIF-8@BPPO-EDA grown for 120 min at different magnifications. Cross-sectional view: (a, b, c); Top view: (d, e, f). 109

129 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Figure EDS line scan across ZIF-8@BPPO-EDA-120 cross-section for the zinc atoms. Another important observation is that unlike the conventional contra diffusion method in which the crystals grow along the whole thickness of the support, ZIF-8 crystals can be observed only at the very outermost section of the EDA-vapour-modified BPPO support (Figure 4-6f). Energy-dispersive X-ray spectroscopy (EDS) line-scan analysis (Figure 4-10) further confirmed the presence of ZIF-8 within the support sublayer as zinc was detected up to about 1 µm underneath the support surface. This means that the heterogeneous nucleation and crystal growth occur on the skin layer of the pre-treated support, which is contrary to what was observed for the untreated BPPO, where the nucleation and crystal growth happened all along the support channels with a non-continuous film, if any, on the surface. As explained in our previous work, EDA can simultaneously create amino groups, which can coordinate to the free zinc ions and provides a large number of nucleation sites; and reduce the substrate pore size induced by its crosslinking 110

130 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate effect [23]. In the present work, therefore, the reduction in the surface pore size and the coordination interaction between the support surface and zinc ions can lead to a decrease of the diffusion rate of Zn 2+ and provide a relatively high precursor concentrations at the support surface and restricting the reaction zone in this vicinity (Figure 4-3). This high precursor concentration and the large number of previously formed nucleation sites result in the faster and thinner crystal growth in the vicinity of the support skin layer as compared to the slow and undirected crystal growth along the channels of untreated BPPO Membrane prepared by conventional contra-diffusion Figure 4-11 presents ZIF-8 membranes synthesized by contra diffusion (at the reaction time of 2h) after modification of the BPPO via immersion of the polymer in 60% EDA aqueous solution at room temperature for 3h. It is obvious that the crystals are grown along the whole thickness of the polymer. As shown, ZIF-8 crystals not only block the polymer micro-channels but also grow within the whole porous structure of the polymer as the interface between the ZIF-8 and the polymer matrix is hardly distinguishable. This is because the immersion of the polymeric support in the nucleophilic diamine solution can result in an extremely high degree of modification (a large number of nucleation sites) within the bulk polymer and the subsequent formation of ZIF-8 crystals when applying contra diffusion synthesis. This shows that crystal formation within the whole polymer support is unavoidable when modifying the BPPO via solution immersion method. Instead, as already shown, contra-diffusion method in conjunction with vapour phase modification of the support offers more degrees of freedom in directing the formation of ZIF-8 membranes. It is worth mentioning that growing MOFs inside the pores of the support can be very attractive for molecular separations and such membranes have been shown to outperform mixed matrix membranes (MMM) for organic solvent nanofiltration (OSN) applications, as demonstrated by 111

131 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Livingston et al [16, 37, 38]. Although the resulting ZIF-8 membrane (Figure 4-11) was too brittle to allow permeation experiments, this work demonstrates a new methodology for in situ growth of ZIFs predominantly inside the supports, which needs to be improved further for practical applicability. Figure SEM images of cross-section of ZIF-8 membrane synthesized by contra-diffusion (at the reaction time of 2 h) after modification of the BPPO via immersion of the polymer in 60% EDA aqueous solution at room temperature for 3 h Single gas performance To further evaluate the quality of the obtained ZIF-8 membranes, single-component gas permeation experiments were conducted, and the results are summarized in Figure 4-12 and Table

132 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Figure Single gas permeances (a) and ideal selectivities (b) as a function of kinetic diameter of gas molecules of the ZIF-8 membranes grown for 120 min and activated at 120 C (ZIF8@BPPO-EDA ) and at 150 C (ZIF8@BPPO-EDA ). Table 4-1. Single gas permeances and ideal selectivities for the ZIF-8@BPPO-EDA-x-y (x: crystallization time (min); y: heat treatment temperature (⁰C)) composite membranes at 25 ⁰C and 1 bar. Sample ID Permeance (10-7 mol.m -2.s - 1.Pa -1 ) Ideal selectivity H 2 C 3H 6 H 2/CO 2 H 2/N 2 H 2/CH 4 H 2/C 3H 8 C 3H 6/C 3H 8 BPPO ±19.5 ±16.5 ±0.1 ±0.2 ±0.2 ±0.2 ±0.2 BPPO-EDA ±19.0 ±14.5 ±0.6 ±0.4 ±0.3 ±0.1 ±0.3 ZIF-8@BPPO- EDA ± ± ± ± ± ± ±0.3 ZIF-8@BPPO- EDA ± ± ±1.0 5 ± ± ± ±0.4 ZIF-8@BPPO- EDA ± ± ± ± ± ± ±4.4 ZIF-8@BPPO- EDA ± ± ± ± ± ± ±3.6 ZIF-8@BPPO- EDA ± ± ± ± ± ± ±

133 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate In comparison, single gas permeation of the untreated BPPO and its vapour phase modified counterpart were tested. Due to their large pores (pore size of 25.5 and 15 for untreated and treated supports, respectively), none of these membranes showed any obvious gas selectivity. However, upon modification, H2 permeance was decreased by more than half when compared to the untreated support. This is attributed to the reduced pore size of the support induced by partial crosslinking effect of the EDA modification, which increases the support dimensional stability and surface tightness. This also lessens the flexibility of the polymeric substrate, which is beneficial for reducing ZIF layer cracking [24, 25]. Membranes started to display molecular sieve performance, favouring the smaller molecules, with a moderate H2 permselectivity after 90 min of the crystal growth (H2/C3H8 ideal selectivity of 32.9 compared to Knudsen diffusion selectivity of 4.7). However, no clear C3H6/C3H8 ideal separation selectivity was observed at this point, indicating the presence of grain boundary micro-defects. As crystallization time is extended, more pronounced molecular sieving behaviour with an obvious increase in the propylene/ propane ideal selectivity can be observed which reaches as high as 16 after 120 min reaction time. It is worth mentioning here that except for a few membranes [11, 29, 32, 39, 40], majority of ZIF-8 membranes reported so far have not shown any decent C3H6/ C3H8 selectivity [41]. Figure 4-13 depicts C3H6/ C3H8 separation performance of ZIF-8 membranes developed in this study in comparison to those reported in the literatures. As can be seen, our membranes not only overtake polymeric and ZIF-8 mixed-matrix membranes in terms of C3H6/C3H8 selectivity and C3H6 permeance they are also amongst the best ZIF-8 membranes previously reported (Table 4-2). For instance, the high quality ZIF-8 membranes made in water-octanol system by interfacial microfluidic membrane processing (IMMP) method could achieve H2/C3H8 ideal selectivity of more than 600 and the permeance of H2 around mol m -2 s -1 Pa -1 [29]. While in the current 114

134 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate study, the H2/C3H8 selectivity and H2 permeance are considerably enhanced, with H2/C3H8 selectivity of and H2 permeance of mol m -2 s -1 Pa -1. The enhancement in permeance in this study is in agreement with the reduction in thickness of the ZIF-8 layer (~2 versus ~9 µm in [29]) and also the highly porous and asymmetric structure of the support, which minimizes the overall hydraulic resistance of the permeate flow through the membrane structure; whereas the enhanced selectivity is mainly due to the improved membrane quality, such as the well-structured grain boundary and absence of pinholes or defects. Figure Comparison of C3H6 permeability and C3H6/C3H8 selectivity of the membranes in the present work with previously reported membranes. Closed and open symbols indicate separation data obtained from single and binary gas permeation analysis, respectively. Hexagon: inorganic supported ZIF-8 membranes [10]; pentagon: ZIF-8 mixed matrix membranes [15]; triangle: polymer membranes [42]; circle: carbon membranes [43]; star: polymer supported ZIF-8 membranes in this study. 115

135 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Table 4-2. Comparison of gas permeation properties of the ZIF-8@BPPO-EDA composite membrane in this work with other ZIF-8 membranes in the literature. Support Synthesis method/ Synthesis media Synthesis T[ C] ZIF thickness /μm Permeance (10-7 mol/m 2.s.Pa) H2 C3H6 Selectivities Ref Polymeric supports BPPO Contra diffusion in conjunction with chemical vapour modification (CD-CVM)/ H2O Room ~ Η2/Ν2 5.1 Η2/CΟ2 9.1 Η2/CΗ4 833 This work Η2/C3Η8 16 C3Η6/C3Η8 Torlon Interfacial microfluidic membrane processing (IMMP)/ Room ~ H2/C3H8 9.2 C3H6/C3H8 [29] C8H17OH- H2O PAN Double-zincsource method/ Room _ 26 H2/C3H8 [44] H2O-NaOH PSf In situ growth followed by LBL deposition/ CH3OH- NaOH Room ~ Η2/CΟ2 [18] 116

136 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Support Synthesis method/ Synthesis media Synthesis T[ C] ZIF thickness /μm Permeance (10-7 mol/m 2.s.Pa) H2 C3H6 Selectivities Ref Polymeric supports PEI/ZnO secondary seeding method/ H2O _ 22.4 H2/C3H8 [45] PES Metal based gel deposition/ CH3OH _ 22.7* Η2/Ν2 5.2* Η2/CΟ2 [46] PES Microfluidic/ Room ~ _ 18.3* [36] CH3OH- NaCOOH Η2/Ν2 2.6* Η2/CΟ2 17.2* Η2/CΗ4 PSf In situ synthesis / CH3OH- NaCOOH Room 35 2 _ 12.4* Η2/Ν2 10.5* Η2/CΗ4 [19] PVDF Chemical modification/ CH3OH- NaCOOH 110 ~ _ 14.3 Η2/Ν Η2/CΟ2 [25] PES Secondary growth/ CH3OH _ 9.9 Η2/Ν2 9.1 Η2/CΗ4 [21] 9.7 Η2/Ar 10.8 Η2/O2 117

137 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Support Synthesis method/ Synthesis media Synthesis T[ C] ZIF thickness /μm Permeance (10-7 mol/m 2.s.Pa) H2 C3H6 Selectivities Ref Polymeric supports PAN Nylon CH3OH Contradiffusion/ Contradiffusion/ H2O-NH4OH Room _ 4.3 Η2/Ν2 [26] Room _ 4.3 Η2/Ν2 [27] BPPO Chemical modification/ CH3OH-NH3 Room ~ _ 9.7 Η2/Ν Η2/CΟ2 [23] Inorganic supports α- alumina Contradiffusion/ _ C3Η6/C3Η8 [31] CH3OH α- alumina Contradiffusion/ CH3OH Η2/Ν Η2/C3Η8 [32] 59 C3Η6/C3Η8 α- alumina Secondary growth/ _ * C3Η6/C3Η8 [10] H2O α- alumina Secondary growth/ CH3OH- NaCOOH ~1 _ 15 Η2/CΗ4 300 Η2/C3Η8 [47] 118

138 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Support Synthesis method/ Synthesis media Synthesis T[ C] ZIF thickness /μm Permeance (10-7 mol/m 2.s.Pa) H2 C3H6 Selectivities Ref Inorganic supports α- alumina Microwaveassisted seeding and secondary growth/ 100 W for 1.5 min 1.5 _ * C3Η6/C3Η8 [48] CH3OH- NaCOOH α- alumina Contradiffusion - based in situ method/ _ ~0.2 50* C3Η6/C3Η8 [11] CH3OH- NaCOOH α- alumina Rapid thermal deposition (RTD)/ N,Ndimethyl acetamide (DMA)- H2O _ * C3Η6/C3Η8 [40] α- alumina Contradiffusionbased in situ method/ 120 ~1 _ ~ * C3Η6/C3Η8 [49] CH3OH- NaCOOH Inorganic supports Alumina hollow fiber Hot support seeding/ CH3OH- NaCOOH _ 9.2 Η2/Ν2 5.4 Η2/CΟ2 [50] 119

139 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate 10.8 Η2/CΗ4 7.1 Η2/O2 γ- alumina Surface chemical modification/ CH3OH- NaCOOH _ 10 Η2/Ν2 4.2 Η2/CΟ [51] Η2/CΗ4 Titania Direct synthesis/ 100 ~ _ 11.6 Η2/Ν2 [52] CH3OH- NaCOOH 4.5 Η2/CΟ Η2/CΗ4 *Gas mixture separation. Highly improved grain boundary structure, on the other hand, can be related to the confinement effects of the crystals within the porous support as the confinement can increase the compactness of the grain boundary structure [49]. Another possible reason for the high quality grain boundary structure can be the well-intergrown ZIF-8 crystals with no preferred orientation as a result of the aqueous synthesis. It was found that water, being less acidic, in comparison with organic solvents can more easily deprotonate the organic ligand on the growing surface, leading to growth occurring in more directions, resulting in a better crystals intergrowth and formation of denser ZIF membranes [39, 53]. Attributed to the formation of high heterogeneous nucleation density in the vicinity of the support surface, the EDA vapour modification is helpful for the controlled synthesis of thin, defect-free and reproducible ZIF-8 membranes. In summary, the enhanced gas permeation properties strongly suggest that the chemical vapour modification-contra diffusion method 120

140 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate provides a new route for preparing high quality ZIF-8 membranes having superior grain boundary structure as compared to those prepared by other methods. Figure SEM images of ZIF-8@BPPO-EDA-120 membranes after activation at 150 ºC (a, b, c) and 200 ºC (d, e, f). (a, d, e) top view and (b, c, f) cross-sectional view Effects of activation temperature on the ZIF-8 membranes Finally, we investigated the effects of activation temperature on the morphology and performance of ZIF-8 membranes. Membranes (grown for 120 min) were further exposed to 150 and 200 ºC for 2 h under oxidative conditions (air). For the membranes activated at 150 ºC, a compact well-intergrown ZIF-8 grains of rhombic dodecahedral shape with no defects (i.e. pinholes or cracks) in the entire membrane surface can be observed (Figure 4-14). A very intimate contact between ZIF-8 and the support is also observed in the cross section view of the membranes as the interface between ZIF-8 and the support is hardly distinguishable. Figure 4-15 shows the room-temperature C3H6/C3H8 permeation properties of ZIF-8 membranes activated at different temperatures. 121

141 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Figure Room-temperature propylene/propane permeation properties of ZIF-8 membranes grown for 120 min (ZIF-8@BPPO-EDA-120) as a function of activation temperatures. As shown in the figure, by increasing the activation temperature from 120 to 150 ºC the C3H6/C3H8 ideal selectivity is significantly increased, with minimal effect on the permeance. This unique behaviour indicates that the ZIF-8@BPPO-EDA membrane becomes even denser with probably more compact grain boundary structure when activated at a higher temperature. This positive result is likely attributed to the fact that the BPPO and BPPO-EDA form crosslinking structure by heating (Figure 2-16) [54], which can further increase the stability and tightness of the support. This was further supported by the 20 (±4) % reduction observed in the hydrogen permeability of the BPPO-EDA substrate upon heating at 150 C for 2 h. 122

142 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Figure Heat-induced cross-linking of BPPO substrate. Since a fraction of ZIF-8 is formed within the support porous structure, it may subsequently enhance the interfacial interaction between the ZIF-8 and the support and also increase the compactness of the grain boundary structure, thereby minimizing the non-selective intercrystalline diffusion and leading to improved separation performance. Thermally induced cross-linking reaction was verified by FTIR result (Figure 4-17) where bands attributed to the benzyl bromide (CH2Br) and amine groups for BPPO and BPPO-EDA, respectively, disappeared upon their thermal treatment at 150 ºC. Another possible reason for the observed improvement in the membrane selectivity activated at a higher temperature can be the complete removal of residual solvent molecules from the ZIF-8 layer [55]. However, further analysis is required to fully understand the effect of annealing temperature on the grain boundary and subsequent gas performance. Further increasing the activation temperature up to 200 ºC results in a significant increase in the propylene permeance (more than one order of magnitude) with a drop in C3H6/C3H8 ideal selectivity from 27.8 to 4.5, indicating the grain boundary structure of the membranes was compromised. While the XRD patterns in Figure 4-18 indicate that the ZIF-8-BPPO-EDA samples did not undergo obvious structural alterations in the studied temperature range compared to the simulated ZIF-8 pattern, the FTIR shows a drop in the intensity of the Zn-N peak for the membrane activated at 200 C. The activation process at the elevated temperature caused some degradation 123

143 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate of the ZIF-8 layer (Figure 4-14d, e, and f). However, as can be seen from the cross sectional view (Figure 4-14f), the degradation was apparently restricted to the membrane surface, resulting in still reasonable C3H6/C3H8 ideal selectivity of 4.5 (compared to the Knudsen propylene/propane selectivity of 1.02). Similar degradation behaviour of ZIF-8 membranes upon activation process at high temperatures were recently reported and was correlated to the corrosion action of water or methanol on ZIF-8 grains [49, 56]. These results indicate that the activation temperature plays a critical role in determining the gas permeation properties of the ZIF-8 membranes. However, an elaborate choice of activation conditions (temperature, duration, and environment) is required in order to achieve ZIF-8 membranes with the best performance. Figure FTIR ATR spectra of the BPPO support, BPPO support after being heated at 150 ºC (BPPO-150) for 2h under air, EDA-vapour-modified BPPO (BPPO-EDA), EDA-vapour-modified BPPO after being heated at 150 ºC (BPPO-EDA-150) for 2h under air. 124

144 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate Figure (a) FTIR spectra and (b) XRD patterns of ZIF-8@BPPO-EDA membranes (grown for 120 min) as a function of activation temperature ( C). It should be noted that the propylene/propane selectivity obtained in this study is the best ever obtained for ZIF-8 membrane on polymeric supports, but it is still lower than those obtained with inorganic-supported ones. For example, alumina-supported ZIF-8 membrane prepared by an in situ counter-diffusion method exhibited both higher C3H6/C3H8 selectivity ( 50) and propylene permeability ( mol m -2 s -1 Pa -1 ) than the best membrane obtained in this work with the C3H6/C3H8 selectivity and propylene permeability of 27.8 and mol m -2 s -1 Pa -1, respectively [11]. However, since the contra diffusion method is dependent on the surface properties and porous structure of support, the performance of the membranes could be further improved by investigating the effect of the modification reaction condition. Optimizing activation conditions could also further improve the performance of the ZIF-8 membranes. In addition, it is worth mentioning that the synthesis procedure developed here offers a number of advantages over those reported in the literature. First of all, almost all of previously reported ZIF membranes were made by using organic solvents (e.g. methanol, dimethyl formamide, octanol) and/or alkaline additives (e.g. sodium formate, ammonia) under non-ambient conditions (i.e. high temperature, high pressure) [11, 27, 52, 57]. In many other cases the use of seed crystals and a long reaction time were unavoidable [26, 39, 47, 58]. The high quality ZIF-8 membranes in this study were made in aqueous solution under ambient conditions (room temperature, atmospheric 125

145 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate pressure) in a relatively short period of time (less than 2 h), without any additives or seed crystals. The synthesis process requires significantly smaller amounts of metal salt and organic ligand reagents. For example, 41-90% savings in the usage of the reagents (per cm 2 of permeable area) could be achieved compared to ZIF-8 membranes fabricated by the microfluidic experimental approach [36]. Summary In summary, we reported a novel strategy, contra-diffusion based synthesis in conjunction with vapour modification, for room temperature synthesis of high-quality ZIF-8 membranes on an asymmetric polymeric substrate in aqueous solution. The ZIF-8 membranes have shown excellent gas permeation properties (e.g. propylene/propane ideal selectivity of 16 with propylene permeance of mol m -2 s -1 Pa -1 ), intensely indicating impressively enhanced membrane microstructure (in particular enhanced grain boundary structure). More importantly, by increasing the activation temperature from 120 to 150 ºC, the propylene/propane selectivity was further increased (almost two-fold), without compromising the high permeance of propylene, indicating the important role of thermal activation conditions (in particular activation temperature) in microstructures of ZIF-8 membranes. With efficient synthesis conditions, the strategy developed here provides an effective and environmentally friendly route for preparing high-quality ZIF membranes on the surface of polymeric support. 126

146 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate References [1] O.M. Yaghi, M. O'Keeffe, N.W. Ockwig, H.K. Chae, M. Eddaoudi, J. Kim, Reticular synthesis and the design of new materials, Nature, 423 (2003) [2] K.S. Park, Z. Ni, A.P. Côté, J.Y. Choi, R. Huang, F.J. Uribe-Romo, H.K. Chae, M. O Keeffe, O.M. Yaghi, Exceptional chemical and thermal stability of zeolitic imidazolate frameworks, Proc. Natl. Acad. Sci. U. S. A., 103 (2006) [3] K. Yamamoto, Y. Nohara, Y. Domon, Y. Takahashi, Y. Sakata, J. Plévert, T. Tatsumi, Organic inorganic hybrid zeolites with framework organic groups, Chem. Mater., 17 (2005) [4] B. Chen, Z. Yang, Y. Zhu, Y. Xia, Zeolitic imidazolate framework materials: recent progress in synthesis and applications, J. Mater. Chem. A, 2 (2014) [5] N. Rangnekar, N. Mittal, B. Elyassi, J. Caro, M. Tsapatsis, Zeolite membranes a review and comparison with MOFs, Chem. Soc. Rev., 44 (2015) [6] J. Yao, H. Wang, Zeolitic imidazolate framework composite membranes and thin films: synthesis and applications, Chem. Soc. Rev., 43 (2014) [7] V.M.A. Melgar, J. Kim, M.R. Othman, Zeolitic imidazolate framework membranes for gas separation: A review of synthesis methods and gas separation performance, J. Ind. Eng. Chem., 28 (2015) [8] C. Zhang, W.J. Koros, Zeolitic imidazolate framework-enabled membranes: challenges and opportunities, J. Phys. Chem. Lett., 6 (2015) [9] C. Zhang, R.P. Lively, K. Zhang, J.R. Johnson, O. Karvan, W.J. Koros, Unexpected molecular sieving properties of zeolitic imidazolate framework-8, J. Phys. Chem. Lett., 3 (2012) [10] Y. Pan, T. Li, G. Lestari, Z. Lai, Effective separation of propylene/propane binary mixtures by ZIF-8 membranes, J. Membr. Sci., 390 (2012) [11] H.T. Kwon, H.-K. Jeong, In situ synthesis of thin zeolitic imidazolate framework ZIF-8 membranes exhibiting exceptionally high propylene/propane separation, J. Am. Chem. Soc., 135 (2013)

147 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate [12] L.M. Robeson, Polymer membranes for gas separation, Curr. Opin. Solid State Mater. Sci., 4 (1999) [13] L.M. Robeson, The upper bound revisited, J. Membr. Sci., 320 (2008) [14] S. Sorribas, P. Gorgojo, C. Téllez, J. Coronas, A.G. Livingston, High flux thin film nanocomposite membranes based on metal organic frameworks for organic solvent nanofiltration, J. Am. Chem. Soc., 135 (2013) [15] C. Zhang, Y. Dai, J.R. Johnson, O. Karvan, W.J. Koros, High performance ZIF-8/6FDA- DAM mixed matrix membrane for propylene/propane separations, J. Membr. Sci., 389 (2012) [16] J. Campbell, J.D.S. Burgal, G. Szekely, R.P. Davies, D.C. Braddock, A. Livingston, Hybrid polymer/mof membranes for Organic Solvent Nanofiltration (OSN): Chemical modification and the quest for perfection, J. Membr. Sci., 503 (2016) [17] Y. Hu, J. Wei, Y. Liang, H. Zhang, X. Zhang, W. Shen, H. Wang, Zeolitic imidazolate framework/graphene oxide hybrid nanosheets as seeds for the growth of ultrathin molecular sieving membranes, Angew. Chem., Int. Ed., 55 (2015) [18] D. Nagaraju, D.G. Bhagat, R. Banerjee, U.K. Kharul, In situ growth of metal-organic frameworks on a porous ultrafiltration membrane for gas separation, J. Mater. Chem. A, 1 (2013) [19] F. Cacho-Bailo, B. Seoane, C. Téllez, J. Coronas, ZIF-8 continuous membrane on porous polysulfone for hydrogen separation, J. Membr. Sci., 464 (2014) [20] W. Li, Q. Meng, X. Li, C. Zhang, Z. Fan, G. Zhang, Non-activation ZnO array as a buffering layer to fabricate strongly adhesive metal organic framework/pvdf hollow fiber membranes, Chem. Commun., 50 (2014) [21] L. Ge, W. Zhou, A. Du, Z. Zhu, Porous polyethersulfone-supported zeolitic imidazolate framework membranes for hydrogen separation, J. Phys. Chem. C, 116 (2012) [22] B.R. Pimentel, A. Parulkar, E.k. Zhou, N.A. Brunelli, R.P. Lively, Zeolitic imidazolate frameworks: next generation materials for energy efficient gas separations, ChemSusChem, 7 (2014)

148 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate [23] E. Shamsaei, Z.-X. Low, X. Lin, A. Mayahi, H. Liu, X. Zhang, J.Z. Liu, H. Wang, Rapid synthesis of ultrathin, defect-free ZIF-8 membranes via chemical vapour modification of a polymeric support, Chem. Commun., 51 (2015) [24] W. Li, Z. Yang, G. Zhang, Z. Fan, Q. Meng, C. Shen, C. Gao, Stiff metal organic framework polyacrylonitrile hollow fiber composite membranes with high gas permeability, J. Mater. Chem. A, 2 (2014) [25] W. Li, Q. Meng, C. Zhang, G. Zhang, Metal organic framework/pvdf composite membranes with high H2 permselectivity synthesized by ammoniation, Chem. - Eur. J., 21 (2015) [26] J. Yao, D. Dong, D. Li, L. He, G. Xu, H. Wang, Contra-diffusion synthesis of ZIF-8 films on a polymer substrate, Chem. Commun., 47 (2011) [27] M. He, J. Yao, L. Li, Z. Zhong, F. Chen, H. Wang, Aqueous solution synthesis of ZIF-8 films on a porous nylon substrate by contra-diffusion method, Microporous Mesoporous Mater., 179 (2013) [28] N. Hara, M. Yoshimune, H. Negishi, K. Haraya, S. Hara, T. Yamaguchi, Metal organic framework membranes with layered structure prepared within the porous support, RSC Adv., 3 (2013) [29] A.J. Brown, N.A. Brunelli, K. Eum, F. Rashidi, J. Johnson, W.J. Koros, C.W. Jones, S. Nair, Interfacial microfluidic processing of metal-organic framework hollow fiber membranes, Science, 345 (2014) [30] T. Xu, D. Wu, L. Wu, Poly (2, 6-dimethyl-1, 4-phenylene oxide) (PPO) a versatile starting polymer for proton conductive membranes (PCMs), Prog. Polym. Sci., 33 (2008) [31] N. Hara, M. Yoshimune, H. Negishi, K. Haraya, S. Hara, T. Yamaguchi, Thickness reduction of the zeolitic imidazolate framework-8 membrane by controlling the reaction rate during the membrane preparation, J. Chem. Eng. Jpn., 47 (2014) [32] N. Hara, M. Yoshimune, H. Negishi, K. Haraya, S. Hara, T. Yamaguchi, Diffusive separation of propylene/propane with ZIF-8 membranes, J. Membr. Sci., 450 (2014)

149 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate [33] E. Shamsaei, Z.-X. Low, X. Lin, Z. Liu, H. Wang, Polysulfone and its quaternary phosphonium derivative composite membranes with high water flux, Ind. Eng. Chem. Res., 54 (2015) [34] Y. Yan, M.E. Davis, G.R. Gavalas, Preparation of zeolite ZSM-5 membranes by in-situ crystallization on porous.alpha.-al2o3, Ind. Eng. Chem. Res., 34 (1995) [35] D. Möckel, E. Staude, M.D. Guiver, Static protein adsorption, ultrafiltration behaviour and cleanability of hydrophilized polysulfone membranes, J. Membr. Sci., 158 (1999) [36] F. Cacho-Bailo, S. Catalán-Aguirre, M. Etxeberría-Benavides, O. Karvan, V. Sebastian, C. Téllez, J. Coronas, Metal-organic framework membranes on the inner-side of a polymeric hollow fiber by microfluidic synthesis, J. Membr. Sci., 476 (2015) [37] J. Campbell, R.P. Davies, D.C. Braddock, A.G. Livingston, Improving the permeance of hybrid polymer/metal-organic framework (MOF) membranes for organic solvent nanofiltration (OSN) - development of MOF thin films via interfacial synthesis, J. Mater. Chem. A, 3 (2015) [38] J. Campbell, G. Szekely, R.P. Davies, D.C. Braddock, A.G. Livingston, Fabrication of hybrid polymer/metal organic framework membranes: mixed matrix membranes versus in situ growth, J. Mater. Chem. A, 2 (2014) [39] Y. Pan, Z. Lai, Sharp separation of C2/C3 hydrocarbon mixtures by zeolitic imidazolate framework-8 (ZIF-8) membranes synthesized in aqueous solutions, Chem. Commun., 47 (2011) [40] M.N. Shah, M.A. Gonzalez, M.C. McCarthy, H.-K. Jeong, An unconventional rapid synthesis of high performance metal organic framework membranes, Langmuir, 29 (2013) [41] M. Shah, M.C. McCarthy, S. Sachdeva, A.K. Lee, H.-K. Jeong, Current status of metal organic framework membranes for gas separations: promises and challenges, Ind. Eng. Chem. Res., 51 (2012) [42] R.L. Burns, W.J. Koros, Defining the challenges for C3H6/C3H8 separation using polymeric membranes, J. Membr. Sci., 211 (2003)

150 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate [43] J.-i. Hayashi, H. Mizuta, M. Yamamoto, K. Kusakabe, S. Morooka, S.-H. Suh, Separation of ethane/ethylene and propane/propylene systems with a carbonized BPDA-pp'ODA polyimide membrane, Ind. Eng. Chem. Res., 35 (1996) [44] P. Neelakanda, E. Barankova, K.-V. Peinemann, Polymer supported ZIF-8 membranes by conversion of sputtered zinc oxide layers, Microporous Mesoporous Mater., 220 (2016) [45] E. Barankova, N. Pradeep, K.-V. Peinemann, Zeolite-imidazolate framework (ZIF-8) membrane synthesis on a mixed-matrix substrate, Chem. Commun., 49 (2013) [46] P. Su, W. Li, C. Zhang, Q. Meng, C. Shen, G. Zhang, Metal based gels as versatile precursors to synthesize stiff and integrated MOF/polymer composite membranes, J. Mater. Chem. A, 3 (2015) [47] H. Bux, A. Feldhoff, J. Cravillon, M. Wiebcke, Y.-S. Li, J. Caro, Oriented zeolitic imidazolate framework-8 membrane with sharp H2/C3H8 molecular sieve separation, Chem. Mater., 23 (2011) [48] H.T. Kwon, H.K. Jeong, Highly propylene-selective supported zeolite-imidazolate framework (ZIF-8) membranes synthesized by rapid microwave-assisted seeding and secondary growth, Chem. Commun., 49 (2013) [49] H.T. Kwon, H.-K. Jeong, Improving propylene/propane separation performance of zeoliticimidazolate framework ZIF-8 membranes, Chem. Eng. Sci., 124 (2015) [50] K. Tao, L. Cao, Y. Lin, C. Kong, L. Chen, A hollow ceramic fiber supported ZIF-8 membrane with enhanced gas separation performance prepared by hot dip-coating seeding, J. Mater. Chem. A, 1 (2013) [51] Y. Liu, N. Wang, J.H. Pan, F. Steinbach, J.r. Caro, In situ synthesis of MOF membranes on ZnAl-CO3 LDH buffer layer-modified substrates, J. Am. Chem. Soc., 136 (2014) [52] H. Bux, F. Liang, Y. Li, J. Cravillon, M. Wiebcke, J.r. Caro, Zeolitic imidazolate framework membrane with molecular sieving properties by microwave-assisted solvothermal synthesis, J. Am. Chem. Soc., 131 (2009) [53] S. Tanaka, T. Shimada, K. Fujita, Y. Miyake, K. Kida, K. Yogo, J.F. Denayer, M. Sugita, T. Takewaki, Seeding-free aqueous synthesis of zeolitic imidazolate framework-8 membranes: How 131

151 Chapter 4 Aqueous phase synthesis of ZIF-8 membrane with controllable location on an asymmetrically porous polymer substrate to trigger preferential heterogeneous nucleation and membrane growth in aqueous rapid reaction solution, J. Membr. Sci., 472 (2014) [54] D. Wu, R. Fu, T. Xu, L. Wu, W. Yang, A novel proton-conductive membrane with reduced methanol permeability prepared from bromomethylated poly (2, 6-dimethyl-1, 4-phenylene oxide)(bppo), J. Membr. Sci., 310 (2008) [55] Q. Song, S.K. Nataraj, M.V. Roussenova, J.C. Tan, D.J. Hughes, W. Li, P. Bourgoin, M.A. Alam, A.K. Cheetham, S.A. Al-Muhtaseb, E. Sivaniah, Zeolitic imidazolate framework (ZIF-8) based polymer nanocomposite membranes for gas separation, Energy Environ. Sci., 5 (2012) [56] Y. Pan, W. Liu, Y. Zhao, C. Wang, Z. Lai, Improved ZIF-8 membrane: Effect of activation procedure and determination of diffusivities of light hydrocarbons, J. Membr. Sci., 493 (2015) [57] M.C. McCarthy, V. Varela-Guerrero, G.V. Barnett, H.-K. Jeong, Synthesis of zeolitic imidazolate framework films and membranes with controlled microstructures, Langmuir, 26 (2010) [58] S.R. Venna, M.A. Carreon, Highly permeable zeolite imidazolate framework-8 membranes for CO2/CH4 separation, J. Am. Chem. Soc., 132 (2009)

152 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes Overview In chapter 4, we combined the vapour phase modification strategy, described in chapter 3, with a simple contra-diffusion method for the scalable fabrication of polymer-supported ZIF-8 membranes in aqueous solution at room temperature. The method offers an environmentally friendly route for controllable fabrication of high-quality polymer-supported ZIF membranes, but the development of a simpler and more versatile method that can construct ZIF layer on various substrates without substrate modification is desired. In this chapter, we utilize one-dimensional (1D) materials such as carbon nanotubes (CNTs) to form a porous nano-scaffold layer on the porous substrate for facilitated growth of ultrathin ZIF membranes with mechanically reinforced structures. CNTs with surface enriched coordination sites create a uniform pseudo-seeding matrix layer that facilitates rapid nucleation and crystal growth during membrane formation. The submicron-thick ZIF-8 hybrid membrane ( nm), consisting of a CNT network integrated in a ZIF-8 matrix, exhibits good mechanical and structural stability whose performance is amongst the best ZIF membranes studied so far. This chapter has been reformatted from the following 133

153 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes submitted manuscript: Shamsaei, E., Lin, X., Wan, L., Tong, Y., and Wang, H. One-dimensional material as nano-scaffold and pseudo-seed for facilitated growth of ultrathin, mechanically reinforced molecular sieving membranes. Chem Commun. 2016, 52 (95), Introduction Zeolitic imidazolate frameworks (ZIFs), Membrane-based separation methods are gaining increasing importance for energy-efficient gas separations and other molecular discriminations [1]. Among many porous materials, metal-organic frameworks (MOFs), owing to their permanent porosity with a diverse structure, chemistry and relative ease of preparation, have been extensively investigated for their application as molecular sieving membranes. MOF membranes not only offer an alternative to overcome the traditional permeability-selectivity trade-off of polymer membranes [2-4], but also they are seen as promising tools in the new area of enantioselective and chiral separations [5]. The major requirement for large-scale, efficient separations is to develop scalable preparation of robust, ultrathin and defect -free MOF membranes. It is well known that the highly reproducible fabrication of such membranes is particularly challenging owing to the difficulty of directing nucleation and crystal growth onto the surface and the tendency for growth into unfavourably large crystals and thick MOF layers. Among numerous fabrication methods being developed, the seeded method, in which the porous substrates are coated with MOF crystals for secondary crystal growth, has been proven to be one of the most effective ways to grow MOF membranes [6-8]. However, seeding route often relies on the MOF features and the support surface chemistry and pore size distribution, requiring elaborate preparation of the crystal seed layer. The use of nanosheets as seeds, which allows for seeding of substrates with large pores and rough surfaces, has been recently investigated as a practical solution to the problems associated with the conventional seeding routes; such a strategy has been demonstrated to be effective for controllable 134

154 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes growth of ultrathin molecular sieving membranes [9, 10]. For instance, our group has very recently applied two dimensional ZIF-8/ graphene oxide (2D ZIF-8/GO) hybrid nanosheets as seeds to the synthesis of ultrathin ZIF-8 membranes with a thickness of around 200 nm and superior CO2/N2 selectivity [11]. However, the efficient synthesis of the nanosheet seeds remains a key task for adoption of such a seeding technique for fabrication of molecular sieving membranes from a wide variety of MOF structures. In this chapter, a new concept for using one-dimensional material as nano-scaffold for fabrication of molecular sieving membranes supported on a porous substrate is described, inspired by the great success of nano-scaffolding technique in tissue engineering where biocompatible nanofibers are used as nano-scaffold to promote tissue growth and provide mechanical support [12]. We propose to utilize one-dimensional (1D) materials such as carbon nanotubes (CNTs) and carbon nanofibers (CNFs) to form a porous nano-scaffold layer on the porous substrate for facilitated growth of ultrathin MOF membranes with mechanically reinforced structures. Owing to their high surface area, outstanding mechanical strength, and thermal and chemical stability, 1D carbon materials (CNTs and CNFs) have been widely employed for the synthesis of novel membranes both as direct filters and effective reinforcing fillers [13, 14]. A number of methods, such as surface oxidation and coating, have been introduced for the chemical modification and dispersion of these 1D carbon nanomaterials, making it possible to prepare a variety of carbon based composite materials [15]. Particularly, hybrid composites based on CNTs and MOFs have recently attracted enormous interest due to their unique synergistic effects [16]. Several MOF- CNT hybrids, such as MOF-5/CNT [17, 18], HKUST-1/CNT [19], ZIF-67/CNT [20], ZIF-8/CNT [21, 22], have been reported, clearly showing the feasibility of using 1D carbon materials as nanoscaffold for MOF growth and mechanical toughening. 135

155 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes To demonstrate our new concept, polydopamine (PDA)-coated CNTs are utilized as nanoscaffold for construction of ultrathin MOF membranes on porous substrates. The high-surface- area CNTs provide superior contact with MOF precursors during crystallization; and the surface functionalization with PDA endows CNTs with the metal-chelating ability of catecholamine moieties present in the PDA structure [23]. It is anticipated that the PDA-coated CNTs functions as the so-called pseudo-seeds to bind MOF building blocks onto various surfaces. These flexible, highly-dispersible hybrid nanotubes allow for the uniform pseudo-seeding of substrates via a simple coating method, without substrate modification and/or the complicated seeding processes usually needed in conventional membrane preparation. Importantly, this new concept has great potential for further development as a versatile platform technique for the scalable fabrication and mechanical reinforcement of ultrathin molecular sieve membranes. Zeolitic imidazolate framework-8 (ZIF-8), Zn (Hmim)2 (Hmim = 2-methylimidazolate), was selected as an example to illustrate our new technique. ZIFs possess the attractive features of both zeolites and MOFs such as tunable pore size and chemistry, high chemical and thermal stability, and as such are increasingly being explored for water treatment, chemical sensors, and gas separation applications [24-27]. Owing to its small aperture size ( 3.4 Å) and sodalite-related structure, defect-free ZIF-8 membranes effectively separate H2 (molecular size 2.9 Å) from larger molecules such as CO2 (3.3 Å), N2 (3.6 Å), CH4 (3.8 Å), C3H6 (4 Å), and C3H8 (4.3 Å), and therefore they are promising candidates for hydrogen separations. 136

156 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes Materials and Methods Chemicals All chemicals were used as received. Anopore aluminium oxide (AAO, GE Healthcare Life Sciences; diameter: 25 mm, pore size: 0.1 µm) membranes, with annular polypropylene ring were used as support. The polyethersulfone (PES) membrane with diameter of 47 mm and 0.03 micron pore size was provided by Sterlitech Corporation. Carbon nanotubes, (multi-walled O.D. I.D. L nm 5-10 nm μm, purity 95%) were purchased from Sigma Aldrich. Dopamine hydrochloride (98 %, Sigma-Aldrich) and tris(hydroxymethyl)aminomethane (Tris, ACS reagent, 99.8%, Sigma-Aldrich) were used for CNT coating layer modification. Zinc nitrate hexahydrate (Zn(NO3)2.6H2O, 98%) and 2-methylimidazole (Hmim, C4H6N2, 99%) supplied from Sigma-Aldrich were used for ZIF-8 preparation. Methanol (absolute) was purchased from Merck, Australia. The water used for the experiments was purified with a water purification system (Milli- Q integral water purification system, Merck Millipore) with a resistivity of 18.2 MΩ/cm Polydopamine modification of CNTs CNTs were modified with polydopamine by an ethanol-mediated oxidative dopamine polymerization process as reported previously [28]. Briefly, 10 mg of CNTs was dispersed in a solution of water (15 ml) and ethanol (20 ml) under sonication, followed by the addition of 40 mg dopamine. 10 ml of Tris aqueous solution (25 mm) was then added with magnetic stirring. After 2 h of reaction at room temperature, the modified CNTs were refined from the solution by centrifugation (5000 rpm for 5 min) and washed with deionized water three times. The modified CNTs were dispersed by 200 ml fresh deionized water and kept as mother solution. 137

157 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes Preparation of ZIF-8/CNTs membrane on porous AAO disk The above mother solution was sonicated for 60 min to yield a stable and uniformly dispersed nanotubes. 3 ml of the solution was then deposited onto the AAO porous membranes by vacuum filtration. Contra-diffusion method was conducted to fabricate the ultrathin ZIF-8/CNT membranes [29]. The CNTs deposited supports were mounted on a homemade setup, where the zinc nitrate solution and Hmim solution were separated by the supporting membrane. The zinc nitrate solution, prepared by dissolving g of Zn(NO3)2.6H2O I n 20 ml of methanol, was added to the CNTs deposited side of the support and Hmim solution, prepared by adding g of Hmim in 20 ml of methanol, was immediately added to the other side of the support. The designed Hmim: Zn 2+ molar ratios in the system was 8 and was kept constant throughout the study. After crystallization at room temperature (20 ± 2 C) for up to 60 min, the obtained membranes were thoroughly rinsed with methanol and dried at 50 C for 6 h. For comparison, ZIF-8 membranes were also prepared on a bare AAO and on a pristine CNTs deposited AAO, following the same preparation method described above for 60 min growth Characterization Powder X-ray diffraction (XRD) patterns were recorded at room temperature using a Miniflex 600 diffractometer (Rigaku, Japan) with Cu Kα radiation (15 ma and 40 kv) at a scan rate of 2 min 1 and a step size of Scanning electron microscopy (SEM; FEI Nova NanoSEM 450) was used for imaging the surface and cross-sectional morphologies of membranes. The membranes were sputter coated with roughly 2 nm iridium (Ir) layer to ensure good electrical conductivity. Transmission electron microscopy (TEM) images were taken by a FEI Tecnai G2 T20 operated at an accelerating voltage of 200 kv. Fourier Transform Infrared (FTIR) spectra of the samples were 138

158 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes taken by an attenuated total reflectance (ATR) FTIR (PerkinElmer, U.S.A.) at an average of 20 scans with a resolution of 4 cm Gas permeation experiments The single gas permeation of hybrid membranes was measured using the pressure rise method [30], as described in detail in chapter 4. The membrane samples were attached to a porous stainless steel holder using a vacuum sealant (Torr seal, Varian). The gas permeation tests were performed at room temperature (20 ± 2 C) on H2, CO2, N2, and CH4. The pressure rise of the permeating gas was measured using a pressure transducer (MKS 628B Baratron). Results and Discussion PDA-coated CNTs The method developed in this work is quite simple, comprising three steps as illustrated in Figure 5-1. Figure 5-1. Schematic illustration of the preparation process of ZIF-8/CNT membrane through deposition of modified CNTs on the support, followed by a contra-diffusion synthesis. 139

159 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes CNTs were first coated by a layer of PDA in a freshly prepared dopamine solution via a very simple process as reported elsewhere [28]. The CNTs coated with PDA show considerably better water dispersibility as compared to the pristine CNTs (Figure 5-2a), indicating an obvious change in surface properties. In high resolution transmission electron microscopy (HRTEM) images (Figure 5-2a), a layer of lower contrast, surrounding the CNTs can be easily identified, demonstrating a conformal PDA coating on CNTs. Fourier transform infrared spectroscopy (FTIR) analysis further confirmed that PDA was successfully coated on CNTs (Figure 5-2c). Upon coating with PDA, new absorption bands at around 1633 and 3438 cm -1 appear, which are assigned to the aromatic rings and catechol OH groups [28], respectively, confirming the successful deposition of PDA on the surface of the CNTs. For the minimal changes in the intrinsic properties of CNTs, specifically its flexibility and diameters, an extremely thin coating layer (< 2 nm) of PDA was formed on the CNTs. The introduction of thin PDA coating on the CNTs made no difference in the X-ray diffraction patterns (XRD) (Figure 5-2b) as compared with the pristine CNTs. After the successful conformal coating, a dilute suspension of polydopamine-coated CNTs was simply vacuum-filtered onto a porous substrate such as an anodized aluminum oxide (AAO) disk. The homogeneity of the deposited layer is guaranteed by high water dispersibility of the nanotubes (Figure 5-3). Interestingly, PDA-coated carbon nanotubes tend to recline with ultimate overlap as they accumulate onto the substrate, producing an ultrathin film with maximal mechanical integrity. A nanoporous structure is also created by the interstices between the CNTs within the film. These can have an important implication for the next step for successful fabrication of continuous ultrathin ZIF-8 membrane with reinforced microstructure. Finally, the ZIF-8/CNT membrane was synthesized via a simple room-temperature contra-diffusion method [31]. 140

160 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes Figure 5-2. (a) Photos of water dispersibility and the corresponding TEM images of CNTs (I) and polydopamine-coated CNTs (II), XRD patterns (b) and FTIR spectrum (c) of CNTs (I) and polydopamine-coated CNTs (II), (f) schematic illustration of the coated CNT and the chemical structure of polydopamine ZIF-8/CNT membrane Well-grown ultrathin ZIF-8 membranes were prepared in only one hour of synthesis time. As shown in the top-view scanning electron microscope (SEM) images (Figure 5-4) the support surface was fully covered with well-grown and compact ZIF-8 nanocrystals devoid of any defects such as pinholes. From the cross-sectional view, an ultrathin well intergrown ZIF-8 layer of nm is observable. Another important observation is that the interface between ZIF-8 layer and the AAO is hardly distinguishable and also the CNTs are completely embedded within the ZIF-8 crystal matrix, indicating tight contacts between ZIF- 8, CNTs, and the support. This exceptionally thin, compact and defect-free ZIF-8 layer on the porous substrate is expected to favour gas separations. 141

161 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes Figure 5-3. SEM (a, b) and optical (c) images of pristine (a, c1-c3) and modified (b, c4-c6) CNTsdeposited on AAO. Detailed experimental: deposited pristine CNTs from (c1, c4) 1 ml (c2, c5) 3 ml and (c3, c6) 6 ml mother solution. Pristine CNTs mother solution: 10 mg CNTs in 200 ml DDI water. Interestingly, ZIF-8 layer formed only on one side (i.e. nanotubes-deposited side) of the support, leaving the other side almost intact and only some scattered single crystals are attached on the AAO channel walls. This is very important from both synthesis and application points of view. In the synthesis of supported MOF membranes by conventional modification or seeded growth one side of the support is often required to be shielded [32], making MOF membrane synthesis complicated, otherwise MOF layer grows on both sides and/or within the support porous structure, causing an undesirable increase in the overall hydraulic resistance [33]. In the present strategy, the strong chelating activity of catechol components in PDA and the attractive hydrophobic interaction between organic ligands and aromatic units of PDA, in addition to the high surface area of the CNTs, all represent an ideal reaction environment for the rapid 142

162 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes heterogeneous nucleation and growth of ZIF-8 on CNT-deposited side of the substrate, affording unique ultrathin ZIF-8 hybrid membrane. A high degree of crystallinity was revealed by XRD analysis, and the ZIF-8 membrane patterns match well with those of the simulated ZIF-8 powder (Figure 5-4f). Figure 5-4. SEM images of modified CNTs-deposited on AAO (a), ZIF-8/CNT membranes grown for 5 min (b), 30 min (c), and for 60 min (d, e), and XRD patterns of ZIF-8 membranes as a function of synthesis time (f). On sharp contrast, distinct ZIF-8 nanocrystals and crystal islets were notable on both sides and within the support if the support surface was not deposited with the modified CNTs (Figure 5-5a-c). In fact, due to the uninterrupted fast contra-diffusion of metal ions and ligand molecules through the large pores of the support and the week interaction between the species and the support, a non-continuous ZIF-8 layer is formed. The situation is similar when the pristine CNTs are attempted (Figure 5-5d-f). Filtration of non-modified CNTs suspension usually results in a deposition of huge agglomerates and bundles on the substrate as a result of their poor aqueous dispersibility and high tendency to bundle up [34]. Subsequently, such a non-uniform deposition does not effectively contribute to the formation of a continuous MOF film; as seen in Figure 5-5d, scattered crystals along with islands composed of a mixture of CNTs and ZIF-8 crystals is 143

163 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes observable. It is also important to note that ZIF-8 and CNTs are phase-separated as opposed to the fully embedded CNTs in ZIF-8 when PDA-coated CNTs are used (Figure 5-5e). On the other hand, the homogenous and sufficient coverage of the entire surface is also an important factor in preparing a continuous ZIF-8 layer. Figure 5-5. SEM images of ZIF-8 film prepared on (a, b, c) bare AAO and on (d, e, f) AAO deposited with pristine CNTs. Zinc side: b, e; Hmim side: c, f. Synthesis time: 60 min CNTs coverage level The coverage level is easily regulated, with nano-scale precision, through the suspension volume filtered and concentration of the CNTs [35]. When PDA-coated CNTs coverage is insufficient (see materials and methods), only non-continuous and imperfect growth of ZIF-8 on the support can be obtained (Figure 5-6). Meanwhile, using an excess volume of the suspension (see materials and methods) results in a membrane with rough surface morphology after the contradiffusion synthesis (Figure 5-6). However, high-magnification top view and cross-sectional SEM images reveal that the rough morphology at the top is indeed composed of an underlying compact and defect-free intercalating layer. The excessive volume of the suspension results in a multi-layer CNTs network, which by generating a nanoporous structure with high surface area per unit volume 144

164 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes provides an ideal medium for ZIF-8 crystallization. However, the crystallization rate on the CNTs closer to the surface of AAO is much faster due to the presence of higher ligand concentration [31]. Subsequently, owing to the self-limiting crystal growth mechanism in the contra-diffusion process [36], the underlying CNT network is sealed prior to further growth across the CNTs film, leaving an ultrathin continuous and compact ZIF-8/CNT membrane with a rough and larger size ZIF-8/ CNTs composite on the surface. This mechanism is further confirmed by the fact that increasing synthesis time affects neither the membrane morphology nor its thickness. This nanoscale surface morphology can provide a better contact between the feed gas and the hybrid membrane, potentially leading to a higher gas permeation through the membrane [37]. Indeed, the resulting membrane exhibits higher gas permeances (shown below). Figure 5-6. SEM images of bare AAO (a), as-prepared samples with insufficient (b, c), and excess (d, e, f) deposition of modified CNTs on AAO before (b, d) and after (c, e, f) contra-diffusion synthesis. The inset in f is a high magnification cross-sectional view. Detailed experimental: deposited CNTs from (b) 1 ml and (d) 6 ml mother solution respectively. A long reaction time was noted for the preparation of ZIF-8 layer on the supports directly modified with PDA and the reported MOF membranes had thicknesses in the rage of tens of microns [38], adversely affecting gas flow through the membranes. In our study The formation of 145

165 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes considerably thinner (more than two order of magnitude) membrane in a shorter synthesis time (1 hr. versus 1 day in reference [38]) is attributed to the considerably larger nucleation area available on the PDA-coated CNTs deposited support resulted from the subsequent larger PDA content, considering the high surface area of the deposited CNTs as compared to the corresponding bare support, as well as to the self-terminating contra-diffusion synthesis process. These results explicitly demonstrate the essential role of PDA-coated CNTs hybrid and that a sufficient and uniform coverage of nanotubes on the support is required for the formation of ultrathin and compact ZIF-8 layer on support Mechanical and structural stability of the ZIF-8/CNT membranes Remarkably, the ultrathin hybrid membrane is composed of CNTs and finely intergrown small sized ZIF-8 crystals ( 50 nm), in which the CNTs act as a nano-scaffold for the construction of a densely packed reinforced ZIF-8 film. Such a reinforced nanostructure is anticipated to enhance the mechanical integrity of the ZIF-8 layer, which is one of the central concerns for practical applications. Indeed, the mechanical integrity test using the ultrasonication method revealed that the ZIF-8 layer retained its morphology without substantial degradation, deconstruction or detachment, even after 2 h of intensive ultrasonication (Figure 5-7), strongly demonstrating the exceptional mechanical integrity of the hybrid ZIF-8/CNT membranes. Figure 5-7. SEM images (a, b) and XRD pattern (c) of the ZIF-8/CNTs membrane after sonication for 2 h. 146

166 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes Figure 5-8. Optical image of the free standing ZIF-8/CNT hybrid membrane floated in the sodium hydroxide solution (a) and SEM images of Cross-sectional view (b, c) and surface edge (d) of the corresponding free standing membrane. The integrity of the CNT/ZIF-8 hybrid membrane was further tested by soaking the supported membrane in a weak sodium hydroxide solution (NaOH). The base solution selectively etched and removed the AAO support and an ultrathin free standing ZIF-8/CNT hybrid membrane floated in the solution (Figure 5-8). ZIF-8/CNTs membranes also showed a greater structural stability against strong electron beams as compared to pure ZIF-8 films. While pure ZIF-8 films underwent structural deformation and local cracking immediately upon exposure to electron beam irradiation (Figure 5-5), the ZIF-8/CNTs membranes retained their structure and integrity as shown in high magnification SEM images (the inset in Figure 5-4d). The structural robustness of the ZIF-8 hybrid membranes upon exposure to electron beam, could be attributed to the ability of the integrated CNTs to dissipate electrostatic charges and subsequently preserves the structure. The enhancement in the structural stability of MOF materials after integration with carbon materials was also observed for other MOF/carbon systems such as MOF-5/CNTs and HKUST-1/GO [18, 39]. This experimental observation not only validates the high mechanical and structural stability of the hybrid membrane, it also suggest that the method explained here can lead to a versatile route for the development of ultrathin free-standing MOF/CNTs hybrid membranes. Carbon nanotubes 147

167 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes (CNTs), having low wettability, low thermal expansion coefficient (CTE) and high gas adsorption selectivity [40], are considered as an ideal reinforcing material and the resulting free-standing MOF hybrid should be able to retain good thermal and mechanical stability under practical applications Synthesis time To gain further insight into the membrane formation mechanism, different synthesis times were conducted. Figure 5-4 shows the XRD patterns and SEM images of the ZIF-8 films formed in different growth times. CNTs entirely embedded with phase-pure ZIF-8 crystals can be observed in just 5 min of contra diffusion synthesis (Figure 5-4b), strongly demonstrating the presence of a perfect environment on the support surface for the rapid heterogeneous nucleation and crystal growth. Remaining interstitial gaps within ZIF-8 embedded CNTs are observable in the high magnification image. However, an obvious reduction in the size and population of these nanopores is noted when the synthesis time was increased to 30 min (Figure 5-4c). This was also evidenced by the observed reduction in the H2 permeance of the membrane. Eventually, after 60 min, the adjacent ZIF-8 crystals grow together and form a uniform and compact ZIF-8 membrane (Figure 5-4d). The absence of cracks and defects was further confirmed by transmembrane pressure-dependent permeation measurements (Figure 5-9). XRD and FTIR spectroscopy were also employed to track the growth process of ZIF-8 on the surface of CNTs deposited substrates. As shown in Figure 5-4f and Figure 5-10, the characteristic peaks of ZIF-8 increase with increasing synthesis time, further confirming the complete coverage of ZIF-8 on the substrate to form a uniform membrane structure. Our proposed mechanism is that the rapid crystal growth on the surface of the nanotubes progressively bridge the interstices between the homogenously deposited CNTs and soon acts as an operative barrier between the metal and the ligand precursor solutions 148

168 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes and effectively confines the synthesis within the nanotubes matrix. Once the CNTs interstitial spaces have been fully sealed through the entire film, termination of the reaction is achieved as there would be no more reagents (ZIF-8 building blocks) in contact to react with each other. Figure 5-9. Single gas permeances of several gases through the ZIF-8/CNT membrane at 20 C and different feed pressures Single gas performance Following its successful synthesis, the supported ZIF-8/CNT hybrid membranes were further investigated for their single component permeation properties for H2, CO2, N2, CH4, C3H6, and C3H8. Owing to its large diameter (~ 100 nm) straight channels, AAO support shows very high and relatively close permeances for all the gases, indicating its minimal effect on the overall membrane separation properties. Surprisingly, an increase in the permeances of all gases was noticed upon homogeneous deposition of modified nanotubes. For example, H2 permeance of blank AAO increased from 1267 to mol m 2 s 1 Pa 1 when a moderate amount (See materials and methods) of the nanotubes were deposited on its surface. This is attributed to the increased porosity 149

169 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes and surface area per unit volume of the CNT-deposited support and its subsequent superior contact with the feed gases. Figure FTIR ATR spectra of the AAO support deposited with modified CNTs, the ZIF- 8/CNTs membranes as a function of synthesis time, and ZIF-8 powder. Figure 5-11 represents the single gas permeances across the ZIF-8/CNT membrane versus the molecular diameter of the permeating gas at room temperature and atmospheric pressure. As shown in the Figure and Table 5-1, there is an explicit cut-off between H2 permeance and CO2 ( and mol m 2 s 1 Pa 1, respectively) and other larger gases, indicating that the ZIF-8/CNT hybrid membrane displays high hydrogen permselectivity. The ideal selectivity of H2 over CO2, N2, CH4, C3H6, and C3H8 are 14, 18, 35, 52.4 and which considerably surpass their corresponding Knudsen coefficients (4.7, 3.7, 2.8, 4.6 and 4.7). This ideal separation factors are in good agreement (except for H2/CO2) with those obtained from ZIF-8 membrane prepared on PDA directly modified alumina support. However, the ZIF-8 hybrid membrane synthesized in this study has more than two orders of magnitude higher H2 permeance than those synthesized on PDA-modified alumina disc. The observed higher H2/CO2 selectivity in this study (14 vs 10.3 in 150

170 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes reference [38]) can be attributed to the presence of CNTs network integrated in ZIF-8 matrix. An enhanced CO2 adsorption capacity and higher CO2/N2 selectivity was previously observed by integration of CNTs into ZIF-8 as compared to pure ZIF-8 [21, 22]. This was further confirmed when analysing ZIF-8 membranes prepared with an excess use of CNTs, in which a decrease in the permeance of CO2 was observed. Surprisingly, the membrane also showed an enhanced H2 permeance (304 (± 8.717) 10 7 mol m 2 s 1 Pa 1 ) and greater H2 selectivity over CO2 (16.3 ± 0.87), which could be attributed to the resultant new nano-scale surface morphology, which can allow better contact between the ZIF-8 crystals and highly mobile H2 molecules [37]. Further, the developed ultrathin ZIF-8/CNT hybrid membranes show superior hydrogen permselectivity as compared to the recently reported high quality ZIF-8 membranes such as, GO [11], TiO2 [37], and EDA assisted ZIF-8 ultrathin membranes (Table 5-2) [41]. These results show that the strategy developed in this study is a powerful approach for increasing the hydrogen permeance without compromising the selectivity. Figure Single gas permeances of several gases through the ZIF-8/CNT membrane at 20 C and 1 bar as a function of the kinetic diameter. The inset shows the ideal gas selectivity for H2 over other gases. 151

171 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes Table 5-1. Single gas permeances and ideal selectivities for the ZIF-8@CNTs-t (t: crystallization time (min), hybrid membranes at 20 ⁰C and 1 bar. E shows the sample prepared with an excess use of CNTs (6ml of the mother solution). Sample ID Permeance (10-7 mol/ m 2.s.Pa) Selectivity H2 CO2 Η2/CΟ2 Η2/Ν2 Η2/CΗ4 Η2/C3Η8 C3Η6/C3Η8 Bare AAO ±60.2 ±55.8 ±0.1 ±0.1 ±0.1 ±0.1 ±0.1 AAO-CNTs ±63.3 ±58.8 ±0.1 ±0.1 ±0.2 ±0.3 ±0.1 AAO- ZIF /CNTs-5 ±40.6 ±10.1 ±0.2 ±0.1 ±0.4 ±0.4 ±0.1 AAO- ZIF /CNTs-30 ±25.4 ±8.7 ±0.4 ±0.4 ±0.6 ± ± AAO- ZIF /CNTs-60 ±7.5 ±1.1 ±1.1 ±0.9 ±1.0 ±118.2 ±2.1 AAO- ZIF /CNTs-60- ±9.2 ±1.2 ±0.8 ±0.8 ±1.5 ±98.8 ±1.8 E The outstanding hydrogen permselectivity in this study can be due to the ultrathin ZIF-8 layer (~200 nm) which is defect-free, densely packed and reinforced within a network of CNTs. Another possible reason for the clear cut-off between H2 and larger gases can be the presence of the CNTs 152

172 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes network in the matrix of ZIF-8. It was previously shown that the presence of carbonaceous materials such as GO can effectively slow down the permeance of larger gases through the ZIF-8 membrane by reducing the nonselective intercrystalline defect pathways and also by its possible constriction effect on the flexibility of ZIF-8 lattices, thus restricting larger molecules to go into the pores [42]. Since the PDA modification of CNTs and their subsequent vacuum deposition are readily controllable, the preparation of ZIF-8/CNT membrane was found to be highly reproducible (Table 5-3). Table 5-2. Comparison of the synthesis parameters (time and temperature) and gas permeation properties of the ZIF-8/CNTs hybrid membrane in this work with other ZIF-8 membranes from the recent literature. Membrane/ Growth Facilitator Synthesis T ( C)/t (hr) H2 Permeance (10-7 mol/ m 2.s.Pa) Η2/ CΟ2 Selectivity Η2/ Η2/ Ν2 CΗ4 Η2/ C3Η8 ~ 2 150/ [43] 30 85/ [44] Less than 1 Ref RT/ _ [37] ZIF-8/Nil 16 RT/ _ 3.7 [29] ZIF-8/EDA modified surface 0.2 RT/3 min [41] ZIF-8/EDA modified surface MOF Thickness (μm) ZIF-8/APTES functionalized α-alumina particles ZIF-8/PDAcoated surface ZIF-8/APTES functionalized TiO2 layer ZIF-8/ZnO nanorods ZIF-8/PDAcoated surface ~2 RT/ [33] 6 100/ _ [45] 20 85/ [38] 153

173 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes ZIF-8@GO/ PDA-coated surface 20 85/ [42] ZIF-8/ZnAl / a 10 a 12.1 _ [46] CO3 LDH buffer layers ZIF-8/Nil / _ 370 [36] ZIF-8/ 2D ZIF-8/GO hybrid nanosheets ZIF-8/ZIF-L seed crystals ZIF-8/PDAmodified CNTs 0.1 RT/ [47] 3.5 RT/6 ~ 6000 Barrer b ~ 4.5 ~ 6.5 ~ 6 _ [48] Less RT/ This than 0.2 work a) Mixture separation factor; b) 1 Barrer = mol m m 2 s 1. Table 5-3. Single gas permeances and ideal selectivities of three ZIF-8/CNT-60 membrane samples tested at 25 ⁰C and 1 bar. Membrane Permeance H2/ CO2 Average Standard deviation (10-7 mol/ m 2.s.Pa) selectivity selectivity of selectivity H2 CO2 M M M Universal applicability Finally, to demonstrate the potential for universal applicability of our pseudo-seeding method, ZIF-8/CNTs membrane was prepared on a porous polymeric substrate. In particular, a commercial 154

174 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes porous polyethersulfone filtration membrane (PES, 0.03 µm, 47 mm, STERLITECH) was selected as the support material. A densely packed ZIF-8/CNTs hybrid membrane was formed on only CNT deposited side of the polymer within 1h (Figure 5-12), following the same preparation method described above for AAO supported ZIF-8/CNTs membrane but having replaced the AAO substrate with PES. The XRD patterns of the PES supported ZIF-8/CNTs match well with those of the simulated ZIF-8 powder (Figure 5-13). FTIR analysis further confirmed the presence of a thin ZIF-8 layer on the support as the characteristic peaks of both ZIF-8 and PES are detected (Figure 5-13b). A long synthesis time ( 16h) and formation of ZIF-8 layer on both side of the substrate through the same contra-diffusion method were noticed when a bare polymer substrate was used in our previous study [31]. Figure SEM images of (a, b) bare PES, (c, d) PES with deposited modified CNTs (6 ml of mother solution) and (e, f) as prepared membrane after contra-diffusion synthesis (1 h). 155

175 Chapter 5 One-dimensional Material as Nano-scaffold and Pseudo-seed for Facilitated Growth of Ultrathin, Mechanically Reinforced Molecular Sieving Membranes Figure (a) XRD patterns and (b) FTIR spectra of pristine PES, supported ZIF-8/CNT membrane and ZIF-8. Summary In conclusion, in the present study we have successfully synthesized a new ZIF-8/CNT hybrid membrane with outstanding hydrogen permselectivity. The fabrication method is simple, reproducible, and adaptable that consist of deposition of PDA-coated CNTs on the supports followed by contra-diffusion synthesis. PDA and CNTs jointly provide an ideal environment for the rapid heterogeneous nucleation and growth of ZIF-8 on the deposited support. It yields ultrathin yet defect free and reinforced ZIF-8 hybrid membranes whose performance is amongst the best ZIF membranes studied so far. At 25 C and 1 bar, the ideal separation selectivities of H2/CO2, H2/N2, H2/CH4, C3H6, and C3H8 are 14, 18, 35, 52.4 and 950.1, respectively, with H2 permeance as high as mol m 2 s 1 Pa 1. We have also observed that the H2 permeance could be even higher by altering the CNTs content in the ZIF-8 matrix and subsequent surface morphology. This high hydrogen permselectivity combined with its mechanically reinforced structure recommend the developed ZIF-8/CNT membrane as a promising candidate for hydrogen separation and purification. Finally, we anticipate that one-dimensional carbonaceous materials assisted crystallization strategy may be further adapted for the fabrication of other MOF and zeolite molecular sieve membranes. 156