Synthetic Design and Development of Sterically-Protected Hydroxide-Conducting Polymers for Energy Conversion Devices

by Andrew Gordon Wright B.Sc. (Hons), Simon Fraser University, 2012

Thesis Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

in the Department of Chemistry Faculty of Science

 Andrew Gordon Wright 2016 SIMON FRASER UNIVERSITY Fall 2016

Approval

Name: Andrew Gordon Wright Degree: Doctor of Philosophy (Chemistry) Title: Synthetic Design and Development of Sterically- Protected Hydroxide-Conducting Polymers for Energy Conversion Devices Chair: Dr. Roger G. Linington Associate Professor Examining Committee:

Dr. Steven Holdcroft Senior Supervisor Professor

Dr. Andrew J. Bennet Supervisor Professor

Dr. Vance E. Williams Supervisor Associate Professor

Dr. Robert A. Britton Internal Examiner Professor

Dr. Patric Jannasch External Examiner Professor Department of Chemistry Lund University

Date Defended/Approved: October 06, 2016

ii

Abstract

The production of renewable energy conversion devices is crucial in reducing greenhouse gas emissions and sustaining the energy required for future generations. However, most energy conversion devices currently available have high costs, which greatly slow down any transition from non-renewable combustion devices. The most promising low-cost, renewable energy conversion devices are based on anion- conducting membranes, such as those found in hydrogen fuel cells, water electrolyzers, redox flow batteries, and electrodialysis. Unfortunately, the current lifetime of such devices is too short for wide-spread adoption.

The main issue is the instability of the alkaline anion exchange membrane towards caustic hydroxide. While a significant amount of research has been on demonstrating materials that have longer lifetimes, little work has been concentrated on investigating the degradation pathways on small molecule model compounds. By understanding the chemistry behind their weakness, materials can be specifically designed to counter such pathways. This then leads towards specifically designed polymers with high endurance. The development towards permanently-stable, alkaline anion exchange membranes is the focus of this thesis.

Throughout this thesis, new model compounds are developed and extensively characterized. Using new stability tests, the degradation pathways are identified and the stability is quantitatively compared. Novel polymers are then prepared, which are designed to mimic the highest stability small molecule compounds. Steric hindrance is found to be the most promising method towards durable cationic polymers. From Chapter 2 to Chapter 5, the prepared materials become more and more resistant to hydroxide, demonstrating development in the correct direction.

Keywords: alkaline anion exchange membrane; hydroxide-conducting polymer; steric hindrance; benzimidazolium; fuel cell; organic chemistry

iii

Dedication

For Mom & Dad, whose love and support allow me to pursue my own dreams

iv

Acknowledgements

I am grateful to Prof. Steven Holdcroft for allowing me to be a part of his lab over the years. Thanks to his guidance and advice and giving me the freedom to investigate new ideas, I have been able to accomplish more than I ever thought possible. His enthusiasm and optimism for science has always motivated me to do more research. I would also like to thank Prof. Andrew Bennet and Prof. Vance Williams for their constructive feedback and advice as supervisory committee members.

I would like to thank present and previous Holdcroft Group members and co- authors for their suggestions and assistance during my research, especially Thomas Weissbach, Dr. Jiantao Fan, Benjamin Britton, Bryton Varju, Elizabeth Kitching, and Dr. Hsu•Feng Lee. I would specifically like to thank Mr. Owen Thomas and Dr. Tim Peckham for mentoring me when I started in this lab and helping me get onto the right track towards success.

I would like to give a special thanks to the SFU Department of Chemistry faculty and staff that helped me along the way. Specifically, Dr. Andrew Lewis and Yilin Zhang for NMR assistance, Hongwen Chen for MS assistance, Bruce Harwood for glassware repair and modification, and Anthony Slater for electronics repair. I would also like to thank Prof. Daniel Leznoff and Dr. Jeffrey Ovens for training and assistance on single crystal XRD.

Lastly, my mother, father, brother, and sister have always been there to support me throughout my life. Despite never being in the same place for more than eight years, they always made me feel at home. My lovely wife, Deena, helps me be better and get the most out of life. Without them, I would not be the man I am today.

v

Table of Contents

Approval ...... ii Abstract ...... iii Dedication ...... iv Acknowledgements ...... v Table of Contents ...... vi List of Tables ...... ix List of Figures...... x List of Schemes ...... xv List of Acronyms ...... xvii

Chapter 1. Introduction ...... 1 1.1. Energy Conversion Devices...... 2 1.2. Anion Exchange Membranes ...... 5 1.3. Poly(benzimidazole) ...... 10 1.4. Thesis Overview ...... 16

Chapter 2. Water-Insoluble and Hydroxide-Stable Ionenes ...... 18 2.1. Introduction ...... 18 2.2. Experimental ...... 20 2.2.1. Materials ...... 20 2.2.2. Synthesis ...... 20 2.2.3. Membrane Preparation ...... 29 2.2.4. Degree of Methylation...... 29 2.2.5. Ion-Exchange Capacity ...... 30 2.2.6. Ion-Exchange Procedure ...... 30 2.2.7. Electrochemical Impedance Spectroscopy ...... 30 2.2.8. Water Uptake (Wu) ...... 31 2.2.9. Degradation Procedure ...... 31 2.2.10. Deuterium-Exchange Experiment ...... 32 2.2.11. Variable Temperature 1H NMR Spectroscopy ...... 32 2.3. Results and Discussion ...... 33 2.3.1. Synthesis ...... 33 2.3.2. Ionic Conductivity and Water Uptake ...... 37 2.3.3. Hydroxide Stability ...... 39 2.3.4. Atropisomerization ...... 43 2.4. Conclusion ...... 45

Chapter 3. Studying HMT-PMBI on Larger Scales ...... 46 3.1. Introduction ...... 46 3.2. Experimental ...... 48 3.2.1. Materials ...... 48 3.2.2. Synthesis ...... 49 3.2.3. Membrane Preparation ...... 54

vi

3.2.4. Degree of Methylation Determination ...... 55 3.2.5. Ion-Exchange Capacity (IEC) ...... 55 3.2.6. Water Uptake ...... 55 3.2.7. Dimensional Swelling...... 56 3.2.8. Electrochemical Impedance Spectroscopy ...... 56 3.2.9. Anion Concentration ...... 57 3.2.10. Mechanical Strength ...... 57 3.2.11. Chemical Stability ...... 58 3.2.12. Assembly of the Catalyst-Coated Membrane ...... 58 3.2.13. AEMFC Operation ...... 60 3.2.14. Water Electrolyzer Operation ...... 60 3.2.15. AEMFC Power Density Error Analysis ...... 61 3.3. Results and Discussion ...... 61 3.3.1. Up-Scaled Monomer Synthesis ...... 61 3.3.2. Polymerization and Post-Functionalization ...... 64 3.3.3. Hydroxide Ion Conductivity ...... 69 3.3.4. Physical Properties of HMT-PMBI Incorporating Various Ions ...... 72 3.3.5. Mechanical Strength ...... 75 3.3.6. Ex Situ Stability to Hydroxide Ions ...... 76 3.3.7. In Situ Fuel Cell Operation ...... 83 3.3.8. In Situ Water Electrolysis Operation ...... 87 3.4. Conclusion ...... 88

Chapter 4. Attempted Synthesis of Triphenylbenzene- Poly(benzimidazolium) ...... 90 4.1. Introduction ...... 90 4.2. Experimental ...... 91 4.2.1. Materials ...... 91 4.2.2. Synthesis ...... 92 4.3. Results and Discussion ...... 97 4.3.1. Monomer Synthesis ...... 97 4.3.2. Condensation Reactions of Triphenylisophthalic Acid ...... 99 4.3.3. Ortho-Substituted Phenyl-Protecting Groups ...... 102 4.4. Conclusion ...... 106

Chapter 5. Poly(phenylene)-Protected Benzimidazoliums ...... 108 5.1. Introduction ...... 108 5.2. Experimental ...... 109 5.2.1. Materials ...... 109 5.2.2. Synthesis ...... 110 5.2.3. Intrinsic Viscosity ...... 121 5.2.4. Membrane Preparation ...... 122 5.2.5. Water Uptake ...... 122 5.2.6. Electrochemical Impedance Spectroscopy ...... 122 5.2.7. Degradation Tests of Model Compounds ...... 123 5.2.8. Identification of Model Compound Degradation Products ...... 124 5.2.9. Polymer Alkaline Stability ...... 124

vii

5.2.10. Mechanical Strength ...... 125 5.2.11. Single Crystal X-ray Diffraction ...... 125 5.2.12. Density Functional Theory (DFT) Calculations ...... 125 5.3. Results and Discussion ...... 126 5.3.1. Model Compound Synthesis ...... 126 5.3.2. Model Compound Stability ...... 129 5.3.3. Single Crystal XRD Structures and DFT Calculations ...... 135 5.3.4. Monomer and Polymer Synthesis ...... 141 5.3.5. Polymer Properties and Stability ...... 147 5.4. Conclusion ...... 151

Chapter 6. Conclusions and Future Work ...... 152 6.1. Conclusions ...... 152 6.2. Future Work ...... 153

References ...... 157

Appendix A. Supporting Information for Chapter 2 ...... 172

Appendix B. Supporting Information for Chapter 3 ...... 174

Appendix C. Supporting Information for Chapter 4 ...... 182

Appendix D Supporting Information for Chapter 5 ...... 185

viii

List of Tables

Table 2.1. Properties of HMT-PMBI-OH− at varying degrees of methylation...... 38 Table 3.1. Diffusion coefficients at infinite dilution of anions and the respective anion conductivity and anion concentration of HMT- PMBI...... 74 Table 3.2. Mechanical properties of HMT-PMBI membranes compared to that of Nafion 212...... 76 Table 5.1. Properties of the model compounds based on experimental data and DFT calculations...... 137

ix

List of Figures

Figure 1.1. Examples of cationic moieties that have been used in literature (guanidinium,71 DABCO,72 sulfonium,73 piperidinium,74 pyrrolidinium,75 imidazolium,76 cyclopropenium,77 ruthenium,78 and phosphonium79). The counter-anions are not shown for clarity...... 6 Figure 1.2. The structures of some recently published AEMs (P1,88 P2,91 P3,155 P4,154 P5,167 P6,168 P7,97 and P8169) that were tested for alkaline stability. The counter-anions are not shown for clarity...... 9 Figure 1.3. The general chemical structure of PBI and m-PBI...... 11 Figure 1.4. Chemical structures of O-PDMBI and mes-PDMBI...... 16 Figure 2.1. Structures of mes-PDMBI and HMT-PDMBI ...... 33 1 Figure 2.2. H NMR spectrum (400 MHz, DMSO-d6) of the sublimed product from the reaction of HTMA and DAB in PPA at 200 °C...... 35 Figure 2.3. 1H NMR spectra of HMT-PMBI in iodide form with various dm% in DMSO•d6. The arrows show the direction of increasing dm% (66, 73, 80, 89, 92, and >96 dm%). Mixed arrows show increased peak height followed by decreased peak height as the dm% increases, likely due to the formation of repeat unit b in the polymer which is then converted into repeat unit c...... 37 Figure 2.4. Ionic conductivity plot versus the percent degree of methylation for HMT-PMBI-OH- measured at 22 °C...... 39 Figure 2.5. Stacked 1H NMR spectra of 92% dm HMT-PMBI after exposure to 2 M KOD/CD3OD/D2O at 60 °C for up to 159 h in a polypropylene tube...... 40 1 Figure 2.6. Stacked H NMR spectra of a 2 M KOD/CD3OD/D2O solution after heating to 60 °C for up to 170 h in a polypropylene tube (control experiment)...... 41 Figure 2.7. Stacked 1H NMR spectra regions and corresponding chemical structures of 92% dm HMT-PMBI (a) initially in its cast, iodide form, (b) after 89 h in 2 M KOD/CD3OD/D2O at 60 °C to exchange the polymer to the deuterium form, and (c) after 90 h of the deuterium-exchanged polymer being subjected to 2 M KOH/CH3OH/H2O at 60 °C conditions to return the polymer to its original hydrogen-based form. The anions are not shown for clarity...... 42 1 Figure 2.8. Aromatic regions of stacked H NMR spectra (500 MHz, CD3OD) of (a) deuterium-exchanged 92% dm HMT-PMBI, (b) hydrogen- exchanged 92% dm HMT-PMBI, and (c) DMMB in iodide form...... 42 Figure 2.9. Racemization of HMTA in solution as observed by 1H NMR -1 spectroscopy (500 MHz) in DMSO-d6 at 20, 50, and 80 g L concentrations...... 44

x

Figure 2.10. Superimposed variable temperature 1H NMR spectra (500 MHz) of -1 HMTA in DMSO•d6 at 50 g L concentration...... 45 Figure 3.1. Photograph of the resulting precipitated HMT-PBI in water after reacting HTME with DAB with a concentration of HMTE in Eaton’s reagent of (a) 25 g L-1 and (b) 40 g L-1...... 65 Figure 3.2. Photograph of the resulting precipitated HMT-PBI (25 g L-1) in (a) room temperature water and (b) ice-water...... 66 Figure 3.3. Photograph of >55% dm HMT-PMBI-I- cast (~45 g) from methanol using (a) a hot plate to dry the solvent in air, showing the bottom of the membrane with a layer of glass, and (b) a vacuum oven to quickly dry the solution to a porous material...... 68 Figure 3.4. Ionic conductivity as measured by EIS of 89% dm HMT-PMBI, initially in OH- form, at 95% RH and 30 °C in air over time, where the inset shows the expanded, 0–60 min, region...... 70 Figure 3.5. (a) Arrhenius plot of ion conductivity of 89% dm HMT-PMBI membranes in mixed carbonate form at various temperatures and RH, in air, and (b) the corresponding calculated activation energy at a given RH...... 71

Figure 3.6. (a) Volume dimensional swelling (Sv) versus water uptake (Wu), including a dashed trendline which excludes K2CO3, Na2SO4, and KF and (b) directional dimensional swelling (Sx, Sy, or Sz) for 89% dm HMT-PMBI after being soaked in various 1 M ionic solutions and washed with water. Sz represents the out-of-plane swelling...... 73 Figure 3.7. Mechanical properties of 89% dm HMT-PMBI membranes in different ion forms, such as the as-cast form (I-, dry), chloride- exchanged wet and dry forms, and hydroxide-exchanged wet form, measured in air...... 76 Figure 3.8. Selected regions of the solution 1H NMR spectrum (500 MHz, DMSO-d6) of the degraded model compound MeB after exposing to 4 M KOH/CH3OH/H2O at 80 °C for 7 days. The chemical structure represents one of the suggested isomers that are formed...... 78 Figure 3.9. (a) Measured chloride ion conductivity and (b) relative percent of benzimidazolium remaining of 89% dm HMT-PMBI membranes after 7 days of soaking in 2 M KOH at various temperatures; (c) measured chloride ion conductivity and (d) relative percent of benzimidazolium remaining of 89% dm HMT-PMBI membranes after 7 days immersion in NaOH solutions of increasing concentration at 80 °C. Membranes were first reconverted to the chloride form for conductivity measurements and then the benzimidazolium remaining was determined from their 1H NMR spectra. The open diamonds refer to the initial samples...... 79

xi

Figure 3.10. The corresponding stacked solution 1H NMR spectra (400 MHz, DMSO•d6) of 89% dm HMT-PMBI in chloride form after membranes were soaked for 7 days in 2 M KOH at various temperatures (as labeled). The arrows show where new resonances are observed...... 80 Figure 3.11. The corresponding stacked solution 1H NMR spectra (400 MHz, DMSO•d6) of 89% dm HMT-PMBI in chloride form after membranes were soaked for 7 days in various concentrations of NaOH (as labeled) at 80 °C...... 81

Figure 3.12. Photograph of the dissolved HMT-PMBI membranes in DMSO-d6 after 7 days in various NaOH concentrations (0.0 – 6.0 M) at 80 °C...... 82 1 Figure 3.13. Stacked solution H NMR spectra (400 MHz, DMSO-d6) of 89% dm HMT•PMBI after 7 days of membrane immersion in 6 M NaOH at room temperature, after being exchanged to chloride form...... 83 Figure 3.14. Measured applied potentials over time for an AEMFC incorporating HMT•PMBI membrane and ionomer, operated at 60 °C, with humidified H2 being supplied to the anode, at the current density shown. At 60 h, the AEMFC was shut-down, left idle for 5 days at room temperature, and restarted back to 60 °C. Between 70–91 h, the cathode was run using air (CO2-containing); otherwise, it was operated with humidified O2...... 84 Figure 3.15. AEMFC polarization and power density curves after various operational times for an HMT-PMBI-based device operated at 60 °C and with H2/O2 at 100% RH supplied to anode/cathode unless otherwise noted, where (a) shows the performance before, during, and after switching the cathode supply from O2 (51 h) to air (75 h) and then back to O2 (94 h), (b) shows the power density at 0, 51, and 94 h, and (c) shows the variable temperature performance after an initial 109 h of operation at 60 °C. The CCMs were pre-conditioned from the chloride form by soaking in 1 M KOH for 7 days followed by 7 days in water...... 85 Figure 3.16. AEMFC performance of FAA-3 and HMT-PMBI devices operated under zero backpressure at 60 °C and with H2/O2 at 100% RH. Both CCMs were pre-conditioned by soaking in 1 M KOH for 24 h and then operated for 45 min before measurement...... 86 Figure 3.17. Measured potential over time for a water electrolysis test of FAA-3 (20 mA cm-2) and HMT-PMBI (25 mA cm-2) devices, where the flowing electrolyte was 1 M KOH at 60 °C for up to 195 h, at which point the still-functional electrolyzer was shut down. At 144 h, the current was stopped, the cell was allowed to re-condition with the same electrolyte and temperature, and then restarted...... 87 Figure 4.1. Chemical structure of TRIP-PDMBI in hydroxide form...... 91 1 Figure 4.2. The H NMR spectrum (500 MHz, DMSO-d6) of a side product from the base hydrolysis of 3...... 98

xii

1 Figure 4.3. H NMR spectrum (500 MHz, DMSO-d6) of the products from the reaction of 4 with o-phenylenediamine in PPA...... 100 1 Figure 4.4. H NMR spectrum (500 MHz, DMSO-d6) of the isolated products from reacting compound 3 in sulfuric acid at room temperature for 10 min...... 102 1 Figure 4.5. H NMR spectrum (500 MHz, DMSO-d6) of 7, isolated from the attempted synthesis of 2c...... 104 1 Figure 4.6. H NMR spectrum (500 MHz, CD2Cl2) of 10...... 105 1 Figure 4.7. H NMR spectrum (500 MHz, DMSO-d6) of the precipitated product from the reaction of 10 with o-phenylenediamine in Eaton’s reagent at 140 °C, after washing with ethyl acetate (EtOAc)...... 106 1 Figure 5.1. H NMR spectrum (400 MHz, DMSO-d6) of isolated BrPhB from the Suzuki-Miyaura coupling of BrB with phenylboronic acid in aqueous-THF...... 129 Figure 5.2. Chemical structures of the four ortho-protected model benzimidazolium compounds examined for hydroxide stability, where X- is the counter-anion...... 130 Figure 5.3. Measurement of remaining starting material over time for the dissolved model compounds (0.02 M) in 3 M NaOD/CD3OD/D2O at 80 °C as determined by 1H NMR spectroscopy...... 131 1 Figure 5.4. H NMR spectra of HB (0.02 M) in 3 M NaOD/CD3OD/D2O taken after dissolution (“0 h”) as well as HB (0.02 M) in pure CD3OD (without NaOD/D2O)...... 131 1 Figure 5.5. H NMR spectra of BrB (0.02 M) in 3 M NaOD/CD3OD/D2O after heating at 80 °C for the specified duration...... 132 1 Figure 5.6. H NMR spectra of MeB (0.02 M) in 3 M NaOD/CD3OD/D2O after heating at 80 °C for the specified duration...... 133 1 Figure 5.7. H NMR spectra of PhB (0.02 M) in 3 M NaOD/CD3OD/D2O after heating at 80 °C for the specified duration...... 134 Figure 5.8. X-ray crystal structures of model compounds in their iodide form (ellipsoids set at 50% probability) alongside the dihedral angles measured (A represents the 2-phenyl plane and B represents the benzimidazolium plane). Only one of the two unique BrB structures is shown for clarity and PhB co-crystallized with H2O (where the hydrogen atoms of H2O are not shown)...... 136 Figure 5.9. Reaction profiles for the two hydroxide-mediated degradation pathways (de-methylation and ring-opening) for HB, MeB, and PhB. The dotted lines represent the higher energy, TS2,cis, ring- opening degradation pathway. No barrier was found between IS1 and IS2...... 138

xiii

Figure 5.10. The ESI mass spectrum of the isolated organic degradation products from MeB after 240 h in 3 M NaOD/CD3OD/D2O at 80 °C...... 139 Figure 5.11. The ESI mass spectrum of the MeB before the degradation test...... 140 Figure 5.12. The ESI mass spectrum of the isolated organic degradation products from PhB after 240 h in 3 M NaOD/CD3OD/D2O at 80 °C...... 140 Figure 5.13. The ESI mass spectrum of the PhB before the degradation test...... 141 1 Figure 5.14. A selected region in the H NMR spectrum (400 MHz, DMSO-d6) of the product from the attempted synthesis of PPB2...... 143 Figure 5.15. (a) Mass spectrum (ESI-MS) of by-products synthesized from the preparation of 14 by Suzuki-Miyaura coupling of 12 with 4•chlorophenylboronic acid, and (b) mass spectrum (ESI-MS) of pure 14...... 145 Figure 5.16. A Huggins-Kraemer plot of PPB3 in NMP calculated from the measured viscosities at 25.0 °C for various concentrations (c)...... 148 Figure 5.17. A stress-strain curve of PPMB (iodide form) under ambient conditions (21 °C, 42% RH) with a cross-head speed of 5.00 mm min-1...... 149 1 Figure 5.18. H NMR spectra of PPMB (chloride form) in DMSO-d6 before (“initial”) and after being subjected to either 1 M or 2 M KOHaq. at 80 °C for 168 h...... 150 Figure 5.19. Relative percent remaining benzimidazolium of PPMB over time in 1 2 M KOHaq. at 80 °C for up to 168 h, as measured by H NMR spectroscopy of the membranes in DMSO-d6...... 150 Figure 6.1. Benzimidazolium structures with varying R1, R2, and R3 groups that could be theoretically prepared...... 153

xiv

List of Schemes

Scheme 1.1. A schematic of an alkaline anion exchange membrane fuel cell (AAEMFC) and alkaline water electrolyzer, with the relevant stoichiometries of oxidant and reductant shown...... 4 Scheme 1.2. Two possible degradation routes for trimethylammonium cations...... 7 Scheme 1.3. Literature synthetic routes used to prepare m-PBI...... 12 Scheme 1.4. Deprotonation and alkylation procedure for functionalization of m- PBI...... 13 Scheme 1.5. Synthetic route to PDMBI in iodide form...... 14 Scheme 1.6. Degradation routes of benzimidazolium hydroxide...... 15 Scheme 2.1. Schematic of a polymer with pendant cations and a polymer in which the cation is part of the backbone...... 19 Scheme 2.2. The synthetic route used to prepare HMTA...... 34 Scheme 2.3. Condensation polymerization of the diacid monomer HMTA and DAB to produce HMT-PBI...... 34 Scheme 2.4. Functionalization of HMT-PBI to produce HMT-PDMBI in iodide form...... 36 Scheme 2.5. Controlled, two-step methylation procedure of HMT-PBI to produce HMT•PMBI in iodide form...... 36 Scheme 3.1. Scaled-up synthetic route to HMTE, showing the average yield and standard deviation for multiple batches of each reaction. The amounts of reagents and products shown represent the total amount used over the multiple syntheses performed...... 62 Scheme 3.2. Scaled-up synthetic route to HMT-PMBI, showing the average yield and standard deviation for multiple batches of each reaction. The amounts of reagents and products shown represent the total amount used over the multiple syntheses performed. The last step shows the yield after optimization...... 67 Scheme 3.3. Possible degradation pathways for the model compound MeB in hydroxide solution: (a) nucleophilic displacement, (b) ring- opening/C2-hydroxide attack, followed by (c) hydrolysis of the amide...... 77 Scheme 4.1. Synthetic route used to prepare monomer 4...... 98 Scheme 4.2. Attempted synthetic route to TRIP-PBI...... 99 Scheme 4.3. Attempted reaction of 4 with o-phenylenediamine in PPA...... 100 Scheme 4.4. Proposed pathways for the reaction of 4 with o-phenylenediamine in PPA...... 101 Scheme 4.5. The synthetic route attempted to prepare monomers 2b and 2c...... 103

xv

Scheme 4.6. Synthetic route to the mesityl-protected carboxylate ester 10...... 105 Scheme 5.1. Condensation reaction of 9 with o-phenylenediamine to produce 11...... 127 Scheme 5.2. Synthetic route to phenyl-protected benzimidazole 13 and phenyl- protected benzimidazolium PhB...... 128 Scheme 5.3. The two degradation pathways for benzimidazolium hydroxides (ring•opening and de-methylation)...... 138 Scheme 5.4. Synthetic routes attempted to prepare various poly(phenylene) polymers bearing pendant benzimidazoles...... 142 Scheme 5.5. Synthetic route used to prepare monomers 14 and 15...... 144 Scheme 6.1. A possible mechanism for deuterium-exchange at the 4- and 7- positions of benzimidazoliums...... 155

xvi

List of Acronyms

° degree(s) µm micrometer (micron) 13C carbon-13 1H proton A ampere(s) Å angstrom(s) AAEMFC alkaline anion exchange membrane fuel cell AcOH acetic acid AEM anion exchange membrane AEMFC anion exchange membrane fuel cell aq. aqueous atm atmosphere (unit) bipy 2,2’-bipyridine C Celsius c concentration CCM catalyst-coated membrane cm centimeter(s) COD 1,5-cyclooctadiene cP centipoise d doublet (NMR spectroscopy) Da dalton(s) DABCO 1,4-diazabicyclo[2.2.2]octane dba dibenzylideneacetone DCM dichloromethane dd doublet of doublets (NMR spectroscopy) DFT density functional theory DI water deionized water dL deciliter(s) dm degree of methylation DMAc N,N-dimethylacetamide DMF N,N-dimethylformamide

xvii

DMSO dimethylsulfoxide

DMSO-d6 deuterated dimethylsulfoxide - DX- diffusion coefficient for the anion X E° standard electrode potential

Ea activation energy EIS electrochemical impedance spectroscopy eq equivalent(s) ESI electrospray ionization EtOAc ethyl acetate EtOH ethanol FEP fluorinated ethylene propylene copolymer g gram(s) GDL gas diffusion layers GPa gigapascal(s) h hour(s) HDPE high-density poly(ethylene) IEC ion-exchange capacity IS intermediate state IUPAC International Union of Pure and Applied Chemistry kcal kilocalorie(s) kg kilogram(s) kPa kilopascal(s) L liter(s) LDA lithium diisopropylamine m multiplet (NMR spectroscopy) M molar m mass m/z mass-to-charge ratio mA milliampere(s) Me methyl MeI methyl iodide MeOH methanol meq milliequivalent(s)

xviii

mg milligram(s) MHz megahertz min minute(s) mL milliliter(s) mm millimeter(s) mmol millimole(s) mol mole(s) MPa megapascal(s) mS millisiemens MS mass spectrometry mW milliwatt(s) n/a not applicable NMP N-methyl-2-pyrrolidinone NMR nuclear magnetic resonance OCV open circuit voltage PBI poly(benzimidazole) Ph phenyl + pH hydrogen ion concentration (-log10[H ])

Pmax maximum power density PPA polyphosphoric acid ppm parts per million pt-1 per point PTFE poly(tetrafluoroethylene) RH relative humidity rt room temperature s singlet (NMR spectroscopy) s second(s)

Sv volume dimensional swelling

Sx x-direction dimensional swelling (length)

Sy y-direction dimensional swelling (width)

Sz z-direction dimensional swelling (thickness) t triplet (NMR spectroscopy) T temperature

xix

t1/2 half-life TCE 1,1,2,2-tetrachloroethane THF tetrahydrofuran TS transition state TW terrawatt(s) V volt(s) W watt(s) wt. weight

WU water uptake (by mass) XRD x-ray diffraction δ chemical shift ΔG Gibbs free energy ε dielectric constant η viscosity

ηrel relative viscosity

ηsp specific viscosity λ hydration number (number of water molecules per ionic charge) σ ionic conductivity

xx

Chapter 1.

Introduction

Energy conversion devices are one of the key technologies for the development of clean and renewable energy systems. The global energy consumption has been predicted to grow from 13.5 terawatts (TW) in the year 2001 to 27 TW in 2050 and 47 TW in 2100.1,2 However, increased energy consumption requires greater use of primary energy sources, whether renewable, such as solar, wind, hydro, and geothermal or non-renewable, such fossil fuels (oil, coal, gas) and fission (nuclear power).3 While fossil fuel reserves will last for at least 100 years,4 the increased use of fossil fuels has led to increased greenhouse gas (CO2) emissions, which persist for 2000 years or more.1,5,6 To meet future energy demands without increasing the average global temperature by 2 °C, carbon-free sources, such as those from solar, wind, and hydro, are necessary.7–9

The predicted amount of energy required from renewable sources for the year 2 2050 to minimize CO2 emissions is estimated to be 10 TW, which would represent 37% of the total energy production. However, in 2004, renewable energy sources (wind, solar, hydro, geothermal, and bioenergy) accounted for 6.9% of the total primary energy production, which only increased to 9.3% in 2014. Additionally, while wind and solar accounted for the least amount of energy in 2014 (1.6%),8 they are predicted to be the main sources of renewable energy in the future.9,10 For this to be realized, it is inevitable that a method of energy storage and/or grid extension is required, as both solar and wind are intermittent energy sources.11,12

1

1.1. Energy Conversion Devices

There are numerous technologies that can be used for electrochemical energy storage. For example, lead-acid batteries have been demonstrated as electrochemical energy storage devices for wind and solar energy.13 However, scaling the production of conventional battery technologies, such as lithium batteries, has been considered problematic.14 A more relevant technology for large scale storage are redox flow batteries, which use an ion-selective polymer membrane between two redox electrolytes and have high efficiencies of 75-85%.15 Due to their bulk size, however, flow batteries are unlikely to be used in automotive applications.16 Instead, hydrogen fuel cells are more appropriate for vehicles, which are regarded as being superior to battery vehicles in terms of mass, cost, efficiency, refueling time, and greenhouse gas emissions when the driving range is over 160 km.17,18

The electrochemical energy conversion of intermittent electricity into the chemical 19–21 bond of hydrogen (H2) has been considered a realistic method for energy storage. Similar to gasoline, hydrogen is an efficient energy carrier that can be transported and 22 quickly used to refill hydrogen fuel cells but does not increase CO2 emissions. While there are a number of ways to produce hydrogen, such as by steam reformation,23–26 a completely carbon-free, renewable energy system would require the following four components:

1. Energy Capturer and Converter (solar/wind energy  electricity)

2. Water Electrolyzer (electricity  H2)

3. Storage for H2 (compression and transportation)

4. Hydrogen Fuel Cell (H2  electricity)

While a transition to such a hydrogen economy would require sociopolitical acceptance, as the consumers are themselves investors,27 it is undisputed that hydrogen fuel cell and water electrolyzer technologies possess some of the highest energy conversion efficiencies. As fuel cells and water electrolyzers do not have moving parts, they are not limited by Carnot efficiencies, resulting in theoretical maximum efficiencies of up to 83%28,29 and 95%,30–32 respectively. For storage, where hydrogen can be stored as a compressed gas, liquefied, adsorbed onto porous substrates, or covalently bound to

2

metal alloys, technologies are currently ongoing extensive research.33–37 Lastly, the energy capturer and converter, such as a solar cell, has the lowest efficiency, with scalable solar cells most likely having efficiencies close to 10%.38,39 When all four components of the renewable system are considered, the projected total energy efficiency is 1•2%,19,40 similar to that of photosynthesis.41 As such, it is imperative that each component is low-cost, scalable, efficient, and long-lived.

Unfortunately, current hydrogen fuel cells, which operate under acidic conditions, require expensive Pt catalysts for optimal performance, due to a delicate balance of activity, selectivity, and stability.42–47 These utilize the transport of protons (H+) through an ion-selective separator, known as the proton exchange membrane. While there is a significant amount of research in reducing the Pt-loading48–50 or moving entirely to non-Pt catalysts51–54 for proton exchange membrane fuel cells, the simplest method would be to operate the fuel cell under alkaline conditions. A number of non-Pt catalysts, such as Ni and Ag, have been demonstrated to work equally well to or better than that of Pt when operated under alkaline conditions.54–57 As such, by replacing the proton exchange membrane with an alkaline separator, such as an anion exchange membrane (AEM), the charge carrier becomes hydroxide (OH-) and the overall cost of the system substantially decreases. A schematic of an alkaline AEM fuel cell (AAEMFC) and alkaline water electrolyzer are shown below in Scheme 1.1.

3

Scheme 1.1. A schematic of an alkaline anion exchange membrane fuel cell (AAEMFC) and alkaline water electrolyzer, with the relevant stoichiometries of oxidant and reductant shown.

4

The alkaline separator for both AAEMFCs and water electrolyzers is responsible for preventing the oxidant (H2 at the anode) and reductant (O2 and H2O at the cathode) from contacting each other while providing selective anion transport (OH-) in order to maintain charge balance.58 It should also be electrically insulating, such that the electron flow can be used externally to do useful work. The alkaline separator used in the 1970s was composed of aqueous KOH in an inert matrix.59 More recently, this type of separator has been prepared by doping acidic polymers with KOH.60–62 This ensures high hydroxide conductivities and low initial costs. However, the main problem of using KOH- doped systems is that exposure to CO2 from any external source (such as air) results in carbonate precipitates (KHCO3 and K2CO3), which blocks electrolyte pathways over 63,64 time. As such, it is mainly suited to non-CO2 environments, such as the NASA space shuttle orbiter, providing lifetimes of >15,000 h.58,65

It was first demonstrated by Fauvarque et al. in 2001 that hydroxide-exchanged AEMs could instead be used as the separator in alkaline fuel cells, thus preventing precipitate formation over time.66 However, the chemical instability of the AEM was problematic, despite this being well documented beforehand when AEMs were applied in electrodialysis (water treatment) applications.67,68 To this day, the chemical instability of AEMs is the major drawback for alkaline energy conversion devices.69 In fact, in 2014, Lewis et al. stated that the fabrication of “[a] highly conductive and stable anion- exchange membrane would be a major innovation for conventional alkaline water electrolysis and solar-driven alkaline electrolysis.”70 For the reasons previously given, the development of such a material is the focus of this thesis. The specific thesis topics and goals will be summarized later in Section 1.4.

1.2. Anion Exchange Membranes

The alkaline anion exchange membrane (AEM), as previously discussed, transports hydroxide anions across but not cations. For this to occur rapidly (i.e., high conductivity), the anion must be dissociated from an immobilized cation. A number of cations have been investigated for AEM applications, such as the examples listed in Figure 1.1.

5

Figure 1.1. Examples of cationic moieties that have been used in literature (guanidinium,71 DABCO,72 sulfonium,73 piperidinium,74 pyrrolidinium,75 imidazolium,76 cyclopropenium,77 ruthenium,78 and phosphonium79). The counter-anions are not shown for clarity.

6

For now, the cation can be a quaternary nitrogen (ammonium,66,72,80–121 guanidinium,71,122–127 pyrrolidinium,75,111,128 piperidinium,74 imidazolium,76,90,129–139 benzimidazolium,140–150 and triazolium151–153), quaternary phosphorus (phosphonium79,111,154–157), tertiary sulfur (sulfonium73,158), metal complex (ruthenium78 and cobalt159), or tertiary carbon (cyclopropenium77). Of these cations, ammonium is most often chosen due to its ease of functionalization on a polymer, low cost, and moderate stability.

However, the alkaline stability of most cations is poor due to the strong nucleophilicity of hydroxide.160 For example, ammonium cations generally degrade by either two mechanisms: (i) β-Hofmann elimination and (ii) direct nucleophilic displacement.58,161 As shown in Scheme 1.2, β­Hofmann elimination occurs by the initial removal of a β­proton and rearrangement, resulting in the loss of trimethylamine. While the polymer backbone is unaffected, the loss of the cation would result in a decrease in ionic conductivity and therefore decrease the fuel cell performance. Instead, if the trimethylammonium is attached at the benzylic position, such that no β­hydrogen is present, direct nucleophilic displacement occurs, resulting in the loss of the cationic charge and methanol.

Scheme 1.2. Two possible degradation routes for trimethylammonium cations.

7

It is also important to consider the stability of the cation and polymer to which it is covalently linked to, as one species. For example, it has been demonstrated that functionalizing hydroxide-stable, ether-containing polymers with hydroxide-stable cations typically results in AEMs that are not alkaline stable.162–164 This is a result of the changing charge distribution such that the cation causes the ether groups to be susceptible to hydrolysis. As such it is preferable to either have the cation far from heteroatoms in the backbone or to have none at all, such that the backbone is entirely linked by C•C bonds.161,162 Examples of recent cationic polymers tested under alkaline conditions are shown in Figure 1.2.

Including the previously described chemical stability requirements, AEMs used in AAEMFCs and water electrolyzers have additional criteria that need to be satisfied:58,69,161,165,166

1. High chemical stability 2. High anionic conductivity 3. Good mechanical stability 4. Good solubility properties 5. Electrically insulating 6. Impermeable to gases 7. Scalable synthesis and low cost

While there are a large number of factors that affect the ionic conductivity of AEMs, such as morphology, cation type, mechanical properties, relative humidity, and temperature, the main property that can be easily manipulated by the synthetic/polymer chemist is the ion-exchange capacity (IEC). The IEC of a polymer can be theoretically calculated or experimentally measured and is in units of meq g-1. It describes the number of milliequivalents (or mmol) of charge or cation per gram of polymer (including the counter ion). Increasing the number of cations per repeat unit of polymer will increase the IEC. In general, as the IEC increases, the conductivity also increases, as there will be more anions and they will be more closely packed. This is the main motivation for developing polymers with high IECs.

8

Figure 1.2. The structures of some recently published AEMs (P1,88 P2,91 P3,155 P4,154 P5,167 P6,168 P7,97 and P8169) that were tested for alkaline stability. The counter-anions are not shown for clarity.

9

As the membrane supports the cathode and anode layers (see Scheme 1.1), it needs to have appropriate mechanical properties to prevent deformation/cracking from temperature, relative humidity (RH), or pressure fluctuations. Unfortunately, while mechanical properties can be improved by using specific polymer backbones that have exceptional mechanical properties, the mechanical properties are generally inversely proportional to IEC. With high IECs, AEMs absorb significantly more water (described by mass water uptake, Wu), thereby making the polymer more plasticized and fragile. The most common method used to obtain high IEC materials with low Wu is to crosslink high IEC materials, as the crosslinks will prevent the material from excessively swelling.93,170– 173 However, this makes the polymer insoluble in solvents and not easily processed. Additionally, AAEMFC and water electrolyzers typically incorporate the cationic polymer into the cathode and anode when spray-coated with the catalyst in order to decrease the contact resistance between the electrodes and the AEM (in which case, the polymer is described as the ionomer).47,174 As such, the most desirable cationic polymer would be soluble in the catalyst solvent, such as methanol, have high conductivity at low water uptakes, and have no crosslinks, allowing it to be used as the membrane and ionomer. A material that could satisfy all of the described criteria would not only be useful for AAEMFCs69,165,166 and alkaline water electrolyzers,175 but also for redox-flow batteries176 and water purification applications.177

1.3. Poly(benzimidazole)

One of the most interesting polymers in literature is poly(benzimidazole) (PBI). PBIs were first discovered by E. I. du Pont de Nemours and Company in 1955,178 with the general structure shown in Figure 1.3. However, the most well-known PBI derivative, named m-PBI (poly(2,2’-(m-phenylene)-5,5’-bibenzimidazole)), was described by Vogel and Marvel in 1961.179 This type of polymer exhibits excellent mechanical properties, producing tough membranes as well as fibers, high thermal stability of >400 °C, and high chemical resistance.180 As such, this polymer has been used in numerous applications, such as aerospace laminates,181 fire-proof fibers,180 and proton exchange membranes.182

10

Figure 1.3. The general chemical structure of PBI and m-PBI.

There have also been numerous synthetic routes used to prepare m-PBI, as shown in Scheme 1.3. The original route used by Vogel and Marvel was a two-step melt condensation of 3,3’-diaminobenzidine (DAB) with diphenyl isophthalate. The first step involves melting the monomers at temperatures up to 290 °C, resulting in a high viscosity mixture with significant foaming, due to the production of phenol and water.183 The mixture is then cooled, ground to a powder, and reheated up to 500 °C to increase the molecular weight (described by viscosity measurements). E. Choe found that if a phosphorus catalyst was added, such as 1% triphenyl phosphite ((PhO)3P), the melt condensation reaction could be performed with the lower cost monomer isophthalic acid to produce m-PBI in one step by heating to 400 °C for 1 h.184,185

11

Scheme 1.3. Literature synthetic routes used to prepare m-PBI.

A third, and potentially easier, method was developed by Iwakura et al., involving a solution polycondensation reaction of the monomers in polyphosphoric acid (PPA).186 While the reaction time is typically longer (>8 h), the temperature is significantly decreased to 200 °C. Additionally, the anhydrous phosphorus-containing solvent, although also viscous, acts as both the catalyst and dehydrating/anti-foaming agent, as it absorbs the produced water. Pouring the solution into water thereby directly results in fibrous polymer.

12

The main disadvantage of PBI derivatives is their poor solubility in organic solvents, which limits their use in certain fields.181 Typically, PBIs are only partially soluble in dimethylsulfoxide (DMSO), N-methyl-2-pyrrolidinone (NMP), and N,N•dimethylacetamide (DMAc) at room temperature, requiring high temperature or pressure to be used for complete solubilization. Consequently, a significant amount of research has focused on functionalization of the nitrogen moieties to prevent hydrogen bonding between polymer chains and improve solubility. For example, in 1990, Sansone described a method for deprotonation of m-PBI followed by alkylation, as shown in Scheme 1.4.187 Using a strong base (lithium hydride), m-PBI can be deprotonated to form the anionic form, which is soluble in NMP. By subsequently adding an alkyl halide, the anionic nitrogen sites become functionalized. Similar methods have been utilized by others in the following two decades to examine the effects of different N-functional groups.140,188–193

Scheme 1.4. Deprotonation and alkylation procedure for functionalization of m-PBI.

It is important to note that while the previous functionalization reactions were optimized for the preparation of 50% N-substituted PBI (only 2 of 4 nitrogens functionalized) to remove all N-H protons, excess alkylating agent was generally used. As a result, the degree of N•substitution yielded was most often higher than 50%. This means that a proportion of the nitrogens were quaternized and thus cationic, although this was not considered detrimental. Only Hu et al. in 1993 purposely used excess iodomethane in two steps in an attempt to prepare 100% methylated m-PBI (named PDMBI, Scheme 1.5) in order to demonstrate it as a water-soluble polymer.140

13

Unfortunately for Hu et al., while the isolated 90% functionalized PDMBI was found to be highly soluble in DMSO and DMSO/water mixtures, it was not soluble in water.

Scheme 1.5. Synthetic route to PDMBI in iodide form.

It took nearly two decades after this work for it to be realized that PDMBI could be potentially used as an AEM. In March 2011, both Henkensmeier et al.142 and Holdcroft et al.141 submitted their work on the anionic conductivity of PDMBI. Both reported similar membrane properties and iodide and bromide conductivities, which was possible due to its insolubility in water. Unfortunately, both also discovered that it was not possible to convert PDMBI into hydroxide form, which is desired for AAEMFC applications. When the iodide form of the membrane was soaked in hydroxide solution to exchange the anion, the membrane was unstable, breaking into pieces. While Henkensmeier et al. hypothesized several degradation routes, Holdcroft et al. experimentally identified the degradation product of a similar benzimidazolium hydroxide model compound. It was observed that the degradation product was an amide, which did not further degrade at room temperature. This suggested that upon exchange of the iodide counter-ion to hydroxide, the C2 position undergoes nucleophilic attack by hydroxide, as shown in Scheme 1.6.

14

Scheme 1.6. Degradation routes of benzimidazolium hydroxide.

In 2012, both Henkensmeier et al.143 and Holdcroft et al.144 demonstrated new methylated PBI derivatives, shown in Figure 1.4, in order to overcome the previously discussed hydroxide instability. Henkensmeier et al. demonstrated that O-PDMBI was more stable than PDMBI in alkaline solutions. This was suggested to be due to the increased electron-donation from the ether, which decreases the positive charge at the C2 position. However, at higher temperatures or higher hydroxide concentrations, O•PDMBI was not stable in hydroxide form. On the other hand, Holdcroft et al. demonstrated unprecedented hydroxide stability with mes-PDMBI, where it was stable in 2 M KOH at 60 °C for 10 days. It was suggested that the greatly improved stability was a result of the increased dihedral angle of the mesitylene group with respect to the benzimidazolium of 83° compared to 58° for a C2-phenyl-substituted benzimidazolium. The 2,6-dimethyl groups of the mesitylene were ideally positioned above and below the C2 carbon, acting as steric hindering groups for nucleophilic attack. Unfortunately, mes•PMDBI, which has a high IEC of 4.5 meq g-1, is soluble in water. This means that it is not usable in AAEMFCs or water electrolyzers without blending or modification.

15

Figure 1.4. Chemical structures of O-PDMBI and mes-PDMBI.

1.4. Thesis Overview

The objective of this thesis is to develop a hydroxide-conducting membrane that is stable at elevated pH and temperature for extended periods of time. In the literature, unfortunately, the term “stable” has often been used ambiguously, as it generally relates to highly specific experimental conditions. For new published materials to stand out from previous ones deemed “stable”, researchers have had to move from describing “stable” materials to “highly stable” and now “ultra stable”. Additionally, it is important to note that while it is difficult to compare the stabilities of different published materials, there have been no reports of a cationic polymer that is actually 100% stable at high temperature (>80 °C) and high pH (>6 M KOH). As such, there is still motivation to develop materials with better alkaline stability.

At the present, there is no specific target for how stable a material should be ex situ. However, it is expected that moving towards materials with higher ex situ stability should generally lead to materials with better in situ stability. Through the course of this thesis, different materials with increasing alkaline stability are developed, which is verified by comparing them under identical experimental conditions. Additionally, the ability to use steric hindrance as a means of preventing degradation is the founding principle used in this work.

16

Chapter 2 will discuss the synthesis and preliminary ex situ measurements of a novel, water-insoluble, and hydroxide-conducting poly(benzimidazolium) that exhibits good chemical resistance. Its unique properties will be discussed and a new method for demonstrating alkaline stability will be presented. Chapter 3 will demonstrate the feasibility of processing the previously discovered polymer on a larger scale, looking at improving and up-scaling the monomer and polymer synthesis, characterizing its chemical and mechanical properties in various anion various forms, temperatures, and humidities, and applying it as a membrane and ionomer in an AAEMFC and alkaline water electrolyzer. Under more extreme alkaline conditions, the degradation product is isolated, providing the basis for further improvements in the stability of benzimidazoliums.

Chapter 4 will discuss a new sterically-protecting moiety that could improve the stability of benzimidazoliums. While the synthesis of the desired polymer was unsuccessful, the many synthetic routes tested provide insight into potential ways of developing new polymer backbones with improved stability. These ideas are then put into practice in Chapter 5, where a new polymer structure is successfully prepared. Numerous small molecule compounds are also prepared. Under equal and highly caustic conditions, different protecting groups are assessed and quantitatively ranked in stability. To explain the observed differences, single crystal x-ray diffraction and density functional theory are used.

Lastly, Chapter 6 will discuss the future work for the continued development towards new hydroxide-conducting polymers for energy conversion devices that will ultimately be immune to degradation from hydroxide anions.

17

Chapter 2.

Water-Insoluble and Hydroxide-Stable Ionenes

The work described in this chapter is reported in part in: Wright, A. G.; Holdcroft, S. ACS Macro Lett. 2014, 3 (5), 444–447. This article can be accessed free-of-charge at http://pubs.acs.org/articlesonrequest/AOR-HzfmXmrFpbcWuSM2ifG9.

Dr. Andrew Lewis performed the variable-temperature 1H NMR spectroscopy. Andrew Wright performed the rest of experimental work in this chapter.

This work was financially supported by the Natural Sciences and Engineering Research Council of Canada (NSERC).

2.1. Introduction

The development of stable anion exchange membranes (AEMs) has typically focused on functionalizing commercial polymers with pendant cations, as previously shown in Section 1.2. However, polymers which have cations within the repeating backbone are rare (Scheme 2.1). In 1968, Eisenberg et al. described similar polymers, which contain ionic amines (i.e., ammoniums) in the backbone, as ionenes, which IUPAC has also similarly termed.194 One of the few examples of ionenes is methylated PBI. As discussed in Chapter 1, mes-PDMBI demonstrated high alkaline stability, showing no significant degradation in 2 M KOH at 60 °C for 10 days. However, due to its water solubility, it was not possible to measure its pure conductivity in a membrane.

18

Scheme 2.1. Schematic of a polymer with pendant cations and a polymer in which the cation is part of the backbone.

The solid-state polymer structure of electrolytes is important for the ionic conductivity of a material, which can be controlled by numerous variables, such by the temperature and solvent used during casting of the membrane as well as by the repeating structure and pendant chain length.195,196 However, little is known about the degree of ion conduction in ionene materials. In contrast to pendant cations, which can aggregate to form channels,197 materials such as PDMBI are highly rigid and are unlikely to form such channels.

In the pursuit of an ionene with high alkaline stability, similar to that of mes•PDMBI, yet insoluble in water, a new PBI derivative was prepared in which additional hydrophobic groups were added. The addition of neutral hydrophobic groups decreases the overall IEC of the material, potentially leading to a membrane that is insoluble in water. This would allow the hydroxide conductivity (σ), water uptake (Wu), and alkaline stability to be measured. If it exhibits good properties, this would later be suitable for in situ AAEMFC and water electrolyzer testing. Additionally, Henkensmeier et al. also found that increasing the distance between cationic groups improves overall stability.146

19

2.2. Experimental

2.2.1. Materials

All chemicals were obtained from Caledon Laboratories Ltd. unless otherwise stated. Mesitoic acid (98%) and 1,4-phenylenediboronic acid (97%) were purchased from Combi-Blocks, Inc.. Potassium deuteroxide solution (40 wt. % in D2O, 98 atom% D), 3,3’-diaminobenzidine (>99% by HPLC), dichloromethane (HPLC grade), and chloroform (HPLC grade) were purchased from Sigma-Aldrich. Tetrakis(triphenylphosphine)palladium(0) (99%) was purchased from Strem Chemicals Inc.. Dimethyl sulfoxide (anhydrous, packed under argon) and 1,4-dioxane (99+%) were purchased from Alfa Aesar. Methylene chloride-d2 (D, 99.9%), dimethyl sulfoxide•d6 (D,

99.9%), CD3OD (D, 99.8%), and deuterium oxide (D, 99.9%) were purchased from Cambridge Isotope Laboratories, Inc.. Degradation experiments were performed in 15 mL polypropylene conical tubes (BD Falcon). 3,3’-diaminobenzidine was purified according to literature procedures.198 Deionized water was purified using a Millipore Gradient Milli-Q® water purification system at 18.2 M cm. 1H NMR and 13C NMR were obtained on a 400 or 500 MHz Bruker AVANCE III running IconNMR under TopSpin 2.1 instruments and the residual solvent resonances for DMSO•d6, CDCl3, CD2Cl2, and 1 CD3OD were set to 2.50 ppm, 7.26 ppm, 5.32 ppm, and 3.31 ppm for H NMR spectra, 13 respectively, and 39.52 ppm and 77.16 ppm for C NMR spectra of DMSO•d6 and

CDCl3, respectively.

2.2.2. Synthesis

Preparation of 3-bromomesitoic acid (BMA)

Compound BMA was synthesized as given in a previous literature procedure.199 More specifically, mesitoic acid (39.41 g, 240 mmol) was dissolved in 560 mL of glacial

20

acetic acid. A separate solution containing bromine (12.3 mL, 239 mmol) and 160 mL of glacial acetic acid was then added and the resulting solution was stirred for 2 h at room temperature. The solution was then poured into 3 L of stirring, distilled water and the precipitate was filtered. After washing the white solid with water, the solid was recrystallized twice from ethanol/water and the collected solid was dried at 70 °C under vacuum, resulting in BMA (43.1 g, 74%) as white needles. 1H NMR (500 MHz, 13 DMSO•d6, ppm) δ: 13.34 (s, 1H), 7.09 (s, 1H), 2.32 (m, 6H), 2.19 (s, 3H). C NMR

(125 MHz, DMSO•d6, ppm) δ: 170.47, 138.54, 135.56, 133.30, 132.64, 130.39, 124.89, 23.88, 21.32, 19.16.

Preparation of methyl 3-bromomesitoate (BME)

Compound BME was synthesized using a generalized methylation procedure from literature.200 More specifically, powdered potassium hydroxide (14.82 g, 264 mmol) was vigorously stirred in dimethyl sulfoxide (300 mL) at room temperature for 30 min. A solution containing BMA (43.1 g, 177 mmol) dissolved in dimethyl sulfoxide (150 mL) was added to the previous solution. After 15 min of stirring, iodomethane (16.3 mL, 262 mmol) was added and stirred for 2 h. The mixture was then poured into a stirring solution of potassium hydroxide (10.0 g) in 3 L of ice-water. The precipitate was filtered, thoroughly washed with distilled water, briefly dried under vacuum at 80 °C (melt), and cooled back to room temperature to yield BME (42.1 g, 93%) as a colourless crystal. 1 H NMR (500 MHz, DMSO•d6, ppm) δ: 7.11 (s, 1H), 3.85 (s, 3H), 2.33 (s, 3H), 2.26 (s, 13 3H), 2.15 (s, 3H). C NMR (125 MHz, DMSO•d6, ppm) δ: 168.85, 138.92, 133.59, 133.30, 132.94, 130.01, 124.46, 52.24, 23.44, 20.84, 18.59.

21

Preparation of dimethyl 2,2’’,4,4’’,6,6’’-hexamethyl-p-terphenyl-3,3’’-diester (HMTE)

In an argon-purged 1 L round-bottom flask with stirbar and condenser, BME (25.00 g, 97.2 mmol), 1,4-phenylenediboronic acid (8.16 g, 49.2 mmol), 1,4-dioxane

(500 mL), 2 M K2CO3 (156 mL), and aliquat 336 (6 drops) were added. The mixture was sparged with argon for 20 min and tetrakis(triphenylphosphine)palladium(0) (0.198 g, 0.2 mol%) was added. After heating at reflux for 22 h under argon, the solution was cooled to 80 °C and poured into a stirring, 55 °C solution of ethanol (800 mL)-water (1000 mL). The mixture was slowly cooled to room temperature and the resulting precipitate was filtered, washed with water, and dried. The solid was purified by flash chromatography on silica with chloroform. The collected and dried product was recrystallized in hexanes and dried under vacuum at 110 °C, resulting in HMTE (12.61 g, 1 60%) as small white crystals. H NMR (500 MHz, CDCl3, ppm) δ: 7.15 (s, 4H), 7.01 (s, 13 2H), 3.93 (s, 6H), 2.33 (s, 6H), 2.05-2.02 (m, 12H). C NMR (125 MHz, CDCl3, ppm) δ: 171.09, 139.79, 139.05, 137.70, 137.68, 133.44, 132.79, 132.78, 132.28, 132.26, 129.46, 129.16, 129.15, 52.05, 21.02, 20.98, 19.55, 18.17, 18.13.

Preparation of 2,2’’,4,4’’,6,6’’-hexamethyl-p-terphenyl-3,3’’-dicarboxylic acid (HMTA)

In a 100 mL round-bottom flask with stirbar were added HMTE (12.00 g, 27.9 mmol) and concentrated sulfuric acid (75 mL). The mixture was vigorously stirred

22

for 30 min at room temperature, where all of the solid was dissolved. The solution was then poured into stirring distilled water (2 L) and stirred for 15 min. The precipitate was filtered, thoroughly washed with water, and dried under vacuum at 110 °C. The collected solid was moved into a 250 mL round-bottom flask and stirred in concentrated sulfuric acid (100 mL) for 45 min. The fully dissolved solution was then poured into stirring distilled water (2 L) and stirred for 15 min. After filtering the precipitate, washing thoroughly with water, and drying under vacuum at 110 °C, HMTA (10.92 g, 97%) was collected as an off-white powder and used without further purification. 1H NMR

(500 MHz, DMSO•d6, ppm) δ: 7.18 (s, 4H), 7.04 (s, 2H), 2.28 (s, 6H), 1.99 (s, 6H), 1.97 13 (s, 6H). C NMR (125 MHz, DMSO•d6, ppm) δ: 171.08, 139.05, 138.56, 136.03, 136.00, 133.87, 132.02, 131.00, 130.98, 129.24, 128.73, 20.50, 20.44, 19.06, 17.79, 17.73.

Preparation of poly[2,2’-(2,2’’,4,4’’,6,6’’-hexamethyl-p-terphenyl-3,3’’-diyl)-5,5’- bibenzimidazole] (HMT-PBI)

In a 500 mL, 3-neck round-bottom flask with a CaCl2 drying tube, glass stopper, and argon inlet, was added HMTA (10.0003 g, 24.85 mmol), 3,3’-diaminobenzidine (5.3240 g, 24.85 mmol), and Eaton’s reagent (400 mL). The vigorously stirred mixture was heated to 120 °C for 30 min under argon flow and then increased to 140 °C for 1 h. The solution was then poured into 3 L of stirring distilled water to precipitate the polymer. The material was filtered, thoroughly washed with water, and then stirred in 3 L of distilled water containing potassium carbonate (200 g) for 2 days at room temperature. The material was filtered again, washed with water, boiling water, and then with acetone, and dried under vacuum at 110 °C, resulting in HMT-PBI (13.8 g, 102%) as light brown, paper-like-textured solid. The 1H NMR spectrum was taken by dissolving HMT-PBI in 1 warm DMSO•d6 with a few drops of 40 wt% KOD (in D2O). H NMR (500 MHz,

DMSO•d6, ppm) δ: 7.62 (m, 2H), 7.37 (m, 2H), 7.23 (s, 4H), 7.07 (m, 2H), 6.96 (m, 2H), 2.08-1.83 (m, 18H).

23

Preparation of HMT-PDMBI-I- (>96% dm HMT-PMBI-I-)

A 250 mL, 3-neck round-bottom flask with septum, condenser, and glass stopper was purged with argon and HMT-PBI (2.00 g, 3.67 mmol), anhydrous dimethyl sulfoxide (128 mL), and lithium hydride (several spatula tips) were added. The mixture was heated to 70 °C for 45 min under argon and cooled back to room temperature. Additional lithium hydride (several spatula tips) was added and then mixture was heated again to 70 °C for 45 min. The dark brown solution was cooled to room temperature and iodomethane (4.0 mL, 64.3 mmol) was added, resulting in immediate precipitation. The mixture was heated to 70 °C for 30 min which re-dissolved the precipitate. Additional iodomethane (6.0 mL, 96.4 mmol) was added and the solution was stirred at 70 °C under argon for 20 h. The solution was cooled to room temperature and poured into stirring distilled water (1.5 L). Potassium iodide (5.0 g) was added to the mixture, the precipitate was filtered, and washed with water. The solid was dried at 80 °C under vacuum, resulting in >96% dm HMT-PMBI-I- (3.02 g, 96%) as dark red solid. As the solid still contained a small amount of impurities, a small portion of this red solid was dissolved in hot

0.2 M KOHaq, solid impurities were filtered off, and the polymer was precipitated from the filtrate by addition of potassium iodide. This solid was filtered, washed with water, and dried under vacuum at 60 °C, resulting in light yellow-coloured product, whose 1H NMR 1 spectrum was taken. H NMR (500 MHz, DMSO•d6, ppm) δ: 8.74 (m, 2H), 8.37 (m, 4H), 7.45 (m, 6H), 4.03 (s, 6H), 3.98 (s, 6H), 2.16 (m, 12H), 1.86 (s, 6H).

Preparation of ~50% dm HMT-PMBI

24

In a 500 mL round-bottom flask with stirbar was added HMT-PBI (10.00 g, 18.36 mmol). A solution containing potassium hydroxide (2.35 g) in 7.2 mL of water was added to the polymer followed by 250 mL of dimethyl sulfoxide. The mixture was heated to 70 °C in air. An additional 50 mL of DMSO was added followed by a solution of potassium hydroxide (1.92 g) dissolved in 5.5 mL of water while at 70 °C. After 30 min, the mixture was cooled to room temperature and vacuum filtered into a clean round- bottom flask. While vigorously stirring the solution, iodomethane (2.75 mL, 44.17 mmol) was rapidly added and manually stirred for 3 min due to the immediate precipitate that formed. The mixture was poured into 3 L of stirring water, the solid was collected, and washed with water and acetone. The solid was moved to 3 L of water containing potassium iodide (20.00 g) and stirred at room temperature for 1 h. The solid was collected again and washed with water and acetone. The solid was stirred in 2 L of acetone for 3 days, collected, washed with acetone, and dried under vacuum at 80 °C to yield a fine, brown powder of 52% dm HMT-PMBI-I- (8.85 g, 84%). 1H NMR (500 MHz,

CD2Cl2, ppm) δ: 8.20-8.01 (m, 1.19H), 7.97-7.44 (m, 5.01H), 7.41-7.04 (m, 6.00H), 4.20- 3.86 (m, 0.41), 3.85-3.31 (m, 5.92H), 2.29-1.51 (m, 18.42H).

Preparation of 66-92% dm HMT-PMBI-I-

In a round-bottom flask, 52% dm HMT-PMBI-I- was dissolved in dichloromethane in air (1.5 g of polymer per 25 mL dichloromethane). A small excess of iodomethane was added for the desired degree of methylation and the flask was capped with a glass stopper. The mixture was heated to 30 °C for 16-18 h. Depending on the degree of methylation, the material was purified differently. For example, the 66% dm polymer was purified by evaporation of the solvent by dynamic vacuum and the resulting film was soaked in acetone, collected, and dried under vacuum at 80 °C to yield a stiff, dark brown film. The 92% dm polymer was purified by precipitation into acetone, filtration, and drying under vacuum at 80 °C to yield light brown fibers. The procedure can also be repeated using a different starting dm%, such as the synthesis of the 89% dm from the 66% dm polymer.

25

Preparation of 2,2'-(2,2'',4,4'',6,6''-hexamethyl-p-terphenyl-3,3''-diyl)- bisbenzimidazole (HMT-B)

In a 200 mL 3-neck round-bottom flask, with a CaCl2 drying tube, stopper, and argon inlet, was added HMTA (1.5005 g, 3.728 mmol), o-phenylenediamine (0.8060 g, 7.453 mmol), and Eaton’s reagent (50 mL). While under an argon stream, the mixture was stirred at 120 °C for 15 min. The mixture was then slowly poured into a stirring solution of aqueous ammonium hydroxide (120 mL NH4OH, 28% aq., with 380 mL water). The precipitated solid was then collected and washed with water. The solid was added to a 200 mL water solution containing 10.00 g potassium carbonate and stirred for 17 h at room temperature. The solid was collected, washed with water, methanol, and water. The solid was dried under vacuum at 100 °C, resulting in HMT-B (1.86 g, 91%) as an off-white powder. The 1H NMR spectrum, shown in Figure A1, was taken by 1 dissolving HMT-B in warm DMSO•d6 with 5 drops of 40 wt% KOD (in D2O). H NMR

(500 MHz, DMSO•d6, ppm) δ: 7.47-7.33 (m, 4H), 7.21 (s, 4H), 7.02-6.90 (m, 2H), 6.84- 13 6.70 (m, 4H), 2.05-1.94 (m, 12H), 1.78-1.69 (m, 6H). C NMR (125 MHz, DMSO•d6, ppm) δ: 161.85, 161.83, 147.00, 146.98, 140.20, 139.11, 137.86, 136.67, 135.54, 135.49, 134.35, 134.30, 130.05, 128.77, 117.69, 116.34, 21.33, 21.27, 20.99, 19.49, 19.42.

26

Preparation of 2,2'-(2,2'',4,4'',6,6''-hexamethyl-p-terphenyl-3,3''-diyl)-bis(1,3- dimethylbenzimidazolium) iodide (HMT-MB)

In a 50 mL 3-neck round-bottom flask with a condenser, septum, and stopper attached, was added HMT-B (0.2000 g, 0.366 mmol). The flask was evacuated and refilled with argon, followed by addition of anhydrous dimethyl sulfoxide (8 mL) and a spatula tip of lithium hydride. The mixture was heated to 70 °C and stirred for 30 min. The mixture was cooled to room temperature and iodomethane (0.23 mL, 3.70 mmol) was added. The temperature was increased to 70 °C and additional iodomethane (0.23 mL, 3.70 mmol) was added. After stirring for 40 min at 70 °C, the mixture was poured into 200 mL ice-water. The solid was collected by filtration and air-dried to yield HMT-MB (0.22 g, 70%) as a brown solid. The solid could be further purified to an off- 1 white powder by trituration in acetone. H NMR (400 MHz, DMSO•d6, ppm) δ: 8.29-8.06 (m, 4H), 7.89-7.69 (m, 4H), 7.53-7.33 (m, 6H), 3.92 (d, J = 6.1 Hz, 12H), 2.18 (d, J = 13 19.2 Hz, 6H), 2.11 (s, 6H), 1.79 (d, J = 16.9 Hz, 6H). C NMR (100 MHz, DMSO•d6, ppm) δ: 149.86, 149.84, 141.07, 141.03, 140.07, 140.04, 137.72, 137.69, 137.43, 136.30, 136.27, 131.57, 131.55, 130.04, 130.02, 129.48, 126.80, 118.03, 113.92, 113.89, 32.33, 32.33, 20.85, 18.79, 17.86, 17.84. The 1H NMR spectrum of HMT-MB is shown in Figure A2.

27

Preparation of 2,2'-dimesityl-5,5'-bibenzimidazole (DMB)

In a 100 mL 3-neck round-bottom flask with a CaCl2 drying tube, stopper, and argon inlet was added mesitoic acid (1.0008 g, 6.10 mmol), 3,3’-diaminobenzidine (0.6517 g, 3.04 mmol), and Eaton’s reagent (47 mL). While under an argon stream, the mixture was heated at 120 °C for 15 min and then poured into 800 mL water. The resulting precipitate was collected, washed with dilute aqueous K2CO3 and water. The filtrate was neutralized by addition of K2CO3, resulting in additional precipitate, which was collected and washed with water. The combined solids were dried under vacuum at 100 °C to yield DMB (1.42 g, 99%) as off-white flaky solid. The 1H NMR spectrum, shown in Figure A3, was taken by dissolving DMB in warm DMSO•d6 with 5 drops of 1 40 wt% KOD (in D2O). H NMR (500 MHz, DMSO•d6, ppm) δ: 7.69 (s, 2H), 7.44 (d, J = 8.1 Hz, 2H), 7.15 (d, J = 8.0 Hz, 2H), 6.84 (s, 4H), 2.27 (s, 6H), 2.06 (s, 12H). 13C NMR

(125 MHz, DMSO•d6, ppm) δ: 161.66, 147.65, 145.63, 137.64, 137.18, 135.92, 133.63, 127.97, 117.62, 115.92, 114.28, 21.44, 21.08.

Preparation of 2,2'-dimesityl-1,1',3,3'-tetramethyl-5,5'-bibenzimidazolium iodide (DMMB)

In a 100 mL 3-neck round-bottom flask attached with a septum, condenser, and stopper was added DMB (0.4995 g, 1.06 mmol). The flask was flushed with argon and anhydrous dimethyl sulfoxide (20 mL) was added, followed by lithium hydride (0.07 g, 8.81 mmol). While under argon, the mixture was heated at 80 °C for 30 min and then

28

cooled to room temperature. Iodomethane (0.66 mL, 10.6 mmol) was added and the mixture was heated to 70 °C. At 70 °C, additional iodomethane (0.66 mL, 10.6 mmol) was added and stirred for 1 h. The mixture was poured into 200 mL water, resulting in a small amount of precipitate. After filtering the mixture, the aqueous layer was washed with toluene and ethyl acetate. The remaining aqueous layer was then evaporated to a red oil at 86 °C. The oil was then dissolved in a minimal amount of acetone and precipitated into ethyl acetate twice. The resulting solid was washed with ethyl acetate and dried under vacuum at 80 °C to yield DMMB (0.43 g, 52%) as a light brown powder. 1 H NMR (500 MHz, DMSO•d6, ppm) δ: 8.66 (s, 2H), 8.36-8.29 (m, 4H), 7.29 (s, 4H), 3.93 (s, 6H), 3.89 (s, 6H), 2.43 (s, 6H), 2.11 (s, 12H). The 1H NMR spectrum of DMMB is shown in Figure A4.

2.2.3. Membrane Preparation

The polymers were cast by first dissolving 0.20 g of polymer in 12 mL of hot DMSO. The resulting solution was filtered into a clean, flat Petri dish and allowed to dry at 86 °C for 48 h in air. The resulting transparent, brown films were removed by addition of water, peeling the films off of the glass, and transferring them into deionized water for at least 48 h before the ion exchange steps. The films were approximately 50 microns thick.

2.2.4. Degree of Methylation

For this calculation only, the 1H NMR spectra were baseline corrected using the “Full Auto (Polynomial Fit)” function found in MestReNova 6.0.4. The degree of methylation (dm) was calculated by first setting the integration area between 4.300- 1 - 3.780 ppm to 12.00H for the H NMR spectra of HMT-PMBI-I in DMSO•d6. This represents the N-methyl groups for the charged benzimidazolium groups. The 3.780- 3.500 ppm area was then integrated, whose value is “x”, representing the N-methyl resonances of the uncharged benzimidazole groups. Equation 1, shown below, was then used to calculate the dm%.

29

( ) ( ) (1)

2.2.5. Ion-Exchange Capacity

The ion-exchange capacity (IEC) in the hydroxide form of HMT-PMBI was calculated from the dm% using Equation 2, shown below.

[ ( )] ( ) ( ) (2) ( ) [ ( )] [ ( )]

where is the percent fraction of the degree of methylation ( ), is the mass

of one repeating unit in 100% dm HMT-PMBI-OH- ( ), and is the mass of

one repeating unit in 50% dm HMT-PMBI-OH- ( ).

2.2.6. Ion-Exchange Procedure

The wet iodide-form film was soaked in 300 mL of 1 M KOH solution for 48 h at room temperature in the presence of air. The membrane was then transferred into 300 mL of deionized water which was exchanged with fresh deionized water multiple times over at least 5 days before conductivity and water uptake measurements were taken.

2.2.7. Electrochemical Impedance Spectroscopy

A piece of the wet, hydroxide-converted film was cut into a small piece (approximately 6 x 10 x 0.05 mm3) and sandwiched between two PTFE blocks with two platinum electrodes on opposite sides of one of the blocks and a cavity in the center of the blocks (in-plane measurement). The central cavity was filled with enough deionized water to cover the film during the measurement and the ionic resistance (Rp) was taken

30

from a best fit of Randles equivalent circuit model, using a Solartron SI 1260 impedance/gain-phase analyzer. The ionic conductivity (σ) was then calculated from Equation 3, as shown below.

( ) (3)

where L is the length, in mm, between the two platinum electrodes (length of cavity), Rp is the resistance, in , calculated from Randles circuit, T is the thickness of the film, in mm, and W is the width of the film between the electrodes, in mm. For each polymer, four measurements of four different pieces were taken (16 measurements per polymer) and the standard deviation was used as the uncertainty.

2.2.8. Water Uptake (Wu)

The wet hydroxide-exchanged film was placed between two kimwipes to remove surface water and its mass was quickly measured (Ww). The film was dried at 100 °C under vacuum for at least 18 h and its dry mass was quickly measured (Wd). The mass water uptake was calculated as shown in Equation 4 below.

( ) (4)

Four different films were measured for each polymer and the standard deviation was used as the uncertainty.

2.2.9. Degradation Procedure

In a polypropylene tube was added 54.0 mg of 92% dm HMT-PMBI-I-. In a graduated cylinder was added 1.00 g of 40% wt. KOD (in D2O) and then diluted to

3.5 mL with CD3OD (2 M KOD/CD3OD/D2O solution). This basic mixture was then added to the polymer and capped. The mixture was heated at 60 °C for 159 h, where samples 1 were periodically taken for analysis of their H NMR spectra (500 MHz, CD3OD). The

31

polymer dissolved within 30 min after heating and no precipitate was observed over the 159 h. A control experiment was also run using the exact same conditions but without the polymer.

2.2.10. Deuterium-Exchange Experiment

In a polypropylene tube was added 92% dm HMT-PMBI-I- (46.1 mg) followed by a 3.5 mL 2 M KOD/CD3OD/D2O solution (1.00 g of 40% wt. KOD in D2O which was diluted to 3.5 mL with CD3OD). After heating the solution for 89 h at 60 °C, the solution was pipetted into a stirring solution of deionized water (100 mL) containing 30.01 g of potassium iodide. The precipitate was collected and washed with H2O. A small amount of D2O was used to wash the solid and approximately one-third of the solid was 1 dissolved in CD3OD for analysis of their H NMR spectra. The remaining solid was washed with H2O again and transferred into a clean polypropylene tube, followed by an additional 3.5 mL of 2 M KOH/CH3OH/H2O solution (1.00 g of 40% wt. KOH in H2O which was diluted to 3.5 mL with CH3OH). The solution was heated to 60 °C for 90 h and the solution was then pipetted into a stirring solution of deionized water (100 mL) containing 30.01 g of potassium iodide. The precipitate was collected, washed with H2O 1 and a small amount of D2O, and all of the solid was dissolved in CD3OD for H NMR spectroscopic analysis. No precipitate was observed during the heating procedure.

2.2.11. Variable Temperature 1H NMR Spectroscopy

The variable temperature 1H NMR spectra were acquired on a 500 MHz Bruker AVANCE III using 5 mm TXI probe, BCU-05 chiller, and BVT-3000 temperature control unit calibrated with ethylene glycol. Compound HMTA, at 50 g L-1 concentration in 1 DMSO•d6 inside of an NMR tube, was placed in a ceramic turbine and H NMR spectra were acquired at 25, 50, 75, 101, 125, and 148 °C, with manual shimming performed at each temperature. The residual DMSO•d6 resonance was set to 2.50 ppm.

32

2.3. Results and Discussion

2.3.1. Synthesis

2,2ʺ,4,4ʺ,6,6ʺ-hexamethyl-p-terphenylene (HMT) was chosen to replace the mesitylene group in mes-PDMBI to increase the hydrophobicity of the backbone and render the polymer water-insoluble, while maintaining steric C2-protection of the cation, as shown in Figure 2.1. This also increases the distance between the two adjacent cations, which has been shown to increase the thermal and chemical stability of this class of polymers.146

Figure 2.1. Structures of mes-PDMBI and HMT-PDMBI

To prepare the diacid monomer HMTA, Suzuki-Miyaura coupling was first used to synthesize the methyl-protected diester HMTE, as shown in Scheme 2.2. The carboxylic acid functional groups were converted to esters prior to the Suzuki-Miyaura in order to improve the purification process. Instead of using carboxylic acids, esters maintained high solubility in organic solvents and HMTE was able to crystallize in hexanes, thus producing high purity product. This was then hydrolyzed to the monomer by dissolving HMTE in concentrated sulfuric acid and pouring the solution into water to precipitate HMTA.

33

Scheme 2.2. The synthetic route used to prepare HMTA.

Using one of the common condensation polymerization methods previously discussed in Section 1.3 for the preparation of PBIs from diacid monomers,186 HMTA was mixed with 3,3ʹ-diaminobenzidine (DAB) in polyphosphoric acid (PPA), as shown in Scheme 2.3. However, HMTA appeared insoluble in the mixture, even at 180 °C. Therefore, the temperature was further increased to 200 °C for 30 min, where the solid dissolved. During this time, however, sublimed crystals formed on the upper walls of reaction vessel. Analysis of the crystals by 1H NMR spectroscopy (Figure 2.2) revealed pure 2,2ʺ,4,4ʺ,6,6ʺ-hexamethyl-p-terphenylene, which must have formed due to the double decarboxylation of HMTA. This effect has also been shown in literature to occur on mesitoic acid when dissolved in sulfuric acid at elevated temperature.201

Scheme 2.3. Condensation polymerization of the diacid monomer HMTA and DAB to produce HMT-PBI.

34

1 Figure 2.2. H NMR spectrum (400 MHz, DMSO-d6) of the sublimed product from the reaction of HTMA and DAB in PPA at 200 °C.

Due to the low solubility of HMTA in PPA and potential decarboxylation that could occur at high temperatures, the solvent was changed to Eaton’s reagent, which is a mixture of 7.7% wt. phosphorus pentoxide in methanesulfonic acid, as it has been shown to have similar properties to PPA but is much less viscous.202,203 As such, when HMTA was mixed with DAB in Eaton’s reagent, all products were soluble at 140 °C and required no mechanical stirring. After 1.5 h at 140 °C, the mixture was poured into water, successfully producing the desired polymer HMT-PBI as fibrous solid. The yield of 102% was most likely due to residual water and acid remaining in the polymer.

HMT-PBI was found to be insoluble in all solvents tested as well as in DMSO, DMF, and NMP, which are often used to dissolve PBIs.190 However, it was found to be soluble in basic DMSO (containing LiH or KOH), which allowed for complete methylation of the nitrogen groups (Scheme 2.4). The ionene in iodide form was successfully produced and drop-cast from DMSO at 86 °C to yield strong and flexible membranes. However, when the membrane was soaked in hydroxide solution to exchange iodide to hydroxide and then transferred into pure water, the membrane dissolved. This suggested that the ion-exchange capacity (IEC) of the polymer in hydroxide form of 3.14 meq g-1 was still too high, even though this was significantly lower than -1 mes•PDMBI in hydroxide form (IECOH- of 4.5 meq g ) .

35

Scheme 2.4. Functionalization of HMT-PBI to produce HMT-PDMBI in iodide form.

As such, the IEC needed to be decreased even further. One method would be making a new polymer by replacing the HMT moiety to a more hydrophobic one, such as a tetraphenyl. Instead, a second method would be the partial methylation of PBI, as has been shown on other PBI derivatives.140,142,143 However, literature methylation procedures typically used a one-pot reaction in LiH-NMP, which requires accurate stoichiometry of reagents (including volatile iodomethane) under an inert environment with heating, rendering methylation difficult to control. In order to control the degree of methylation more reproducibly, a two-step, scalable, air-insensitive methylation procedure was developed wherein the first methylation (requiring DMSO/KOH/H2O) takes only 3 min at room temperature to yield the neutral polymer, 50% dm HMT-PMBI, as shown in Scheme 2.5. However, it was preferable to slightly over-methylate the polymer in order to ensure complete conversion of anionic sites.

Scheme 2.5. Controlled, two-step methylation procedure of HMT-PBI to produce HMT•PMBI in iodide form.

The 50% dm HMT-PMBI is insoluble in DMSO but soluble in chloroform and dichloromethane (DCM). Controlled, partial methylation of the remaining basic nitrogens, performed in DCM using controlled amounts of iodomethane at 30 °C over 16−18 h, led to 66, 73, 80, 89, and 92% degree of methylation (dm). The polymers are soluble in DMSO, thus aiding analysis of 1H NMR spectra (Figure 2.3). HMT-PMBI consists of four

36

randomly distributed units: a, b, and c, where b represents two structural isomers (see Scheme 2.5). 50% dm HMT-PMBI is 100 mol% unit a; whereas 100% dm HMT•PMBI is 100 mol% unit c. Integration of the N-methyl resonances in the 1H NMR spectra provides quantification of dm% (using Equation 1 from Section 2.2.4). All partially methylated polymers, including 92% dm, were insoluble in water in both their iodide and hydroxide forms. Cast from DMSO solutions, they formed strong, flexible, transparent brown films. The 89% dm HMT-PMBI derivative was soluble in methanol in the iodide and hydroxide form, insoluble in anhydrous ethanol, but readily soluble in ethanol/water mixtures.

Figure 2.3. 1H NMR spectra of HMT-PMBI in iodide form with various dm% in DMSO•d6. The arrows show the direction of increasing dm% (66, 73, 80, 89, 92, and >96 dm%). Mixed arrows show increased peak height followed by decreased peak height as the dm% increases, likely due to the formation of repeat unit b in the polymer which is then converted into repeat unit c.

2.3.2. Ionic Conductivity and Water Uptake

The iodide was exchanged for hydroxide by soaking the films in 1 M KOH for 48 h at room temperature, followed by soaking in deionized water (DI water) for 120 h with repeated exchanges of water. The DI water was exposed to air and thus

37

carbonated. While the presence of atmospheric CO2 is known to convert the hydroxide form to a mixed hydroxide, carbonate, and bicarbonate form82 and that specialized setups are required to characterize the polymer in its hydroxide form,89 the polymer is labeled HMT-PMBI-OH− for discussion.

The ionic conductivities (σ) measured by electrochemical impedance spectroscopy (EIS) for HMT-PMBI-OH− derivatives are listed in Table 2.1 (see Section 2.2.7 for method). The conductivity exponentially increased with increasing dm%, as shown in Figure 2.4, from 0.10 ± 0.03 mS cm−1 for 66% dm to 9.7 ± 0.6 mS cm−1 for

82,89,166 92% dm. These are likely to be much higher in the absence of atmospheric CO2. For example, Yan and Hickner found that the majority of hydroxide is converted to bicarbonate in air,89 and Disabb-Miller et al. found that the bicarbonate diffusion coefficient of the mobile ion approached that of the dilute solution limit at hydration numbers (λ) above 20.204 With dilute solutions, the diffusion coefficients of hydroxide and bicarbonate are 5.3 × 10−5 and 1.2 × 10−5 cm2 s−1, respectively.205 Assuming that the 92% dm HMT-PMBI-OH- was in bicarbonate form, the hydroxide conductivity would be 4.4 times larger, that is, 43 mS cm−1, which is of the same order of magnitude as for other imidazolium- and benzimidazolium-based polymers.90,132,133,144

Table 2.1. Properties of HMT-PMBI-OH− at varying degrees of methylation.

a -1 b c -1 d dm (%) IEC (meq g ) Wu (wt. %) λ σ (mS cm ) ~50 0.0 n/ae n/ae n/ae 66 1.1 29 ± 4 15 0.10 ± 0.03 73 1.5 36 ± 3 13 0.45 ± 0.06 80 2.0 42 ± 3 12 1.4 ± 0.2 89 2.5 80 ± 20 18 6.1 ± 1.2 92 2.7 180 ± 50 37 9.7 ± 0.6 ~100 3.1 n/af n/af n/af aDegree of methylation as determined by 1H NMR spectroscopy (Equation 1). bHydroxide ion-exchange capacity as calculated from 1H NMR spectroscopy (Equation 2). c - H2O/OH mole ratio. dIonic conductivity at 22 °C when fully hydrated. eCould not be cast from DMSO due to its insolubility. fWater-soluble material.

38

Figure 2.4. Ionic conductivity plot versus the percent degree of methylation for HMT- PMBI-OH- measured at 22 °C.

2.3.3. Hydroxide Stability

The hydroxide stability of 92% dm HMT-PMBI dissolved in 2 M KOD/CD3OD/D2O at 60 °C over 7 days (see Section 2.2.9 for details) was monitored; these conditions are known to strongly accelerate degradation.133 The stability test was carried out in polypropylene tubes as silica glass leads to precipitates in strongly basic conditions.132 Samples were extracted periodically and analyzed by 1H NMR spectroscopy (Figure 2.5). The chemical shifts of the resonances associated with the polymer were unchanged over the 7 days. However, several of the aromatic resonances slowly vanished after 63 h, suggestive of deuterium exchange, as opposed to degradation. Additionally, it is important to note that several new resonances appeared in all of the degradation test spectra and continued to grow over time at 1.27 ppm (broad singlet), 1.23 ppm (sharp singlet), and 0.83 ppm (multiplet). These resonances were found to arise from the polypropylene tube, as observed in a control experiment shown in Figure 2.6, an observation that does not appear to have been reported by others.

39

Figure 2.5. Stacked 1H NMR spectra of 92% dm HMT-PMBI after exposure to 2 M KOD/CD3OD/D2O at 60 °C for up to 159 h in a polypropylene tube.

40

1 Figure 2.6. Stacked H NMR spectra of a 2 M KOD/CD3OD/D2O solution after heating to 60 °C for up to 170 h in a polypropylene tube (control experiment).

To prove that deuterium exchange was occurring, the same experiment was carried by dissolving the polymer for 4 days in KOD/CD3OD/D2O (deuterium-containing mixture) and precipitating the polymer for analysis of its 1H NMR spectrum. Then, the deuterium-exchange polymer was dissolved for 4 days in KOH/CH3OH/H2O (hydrogen- containing mixture), precipitated, and also analyzed by 1H NMR spectroscopy (see Section 2.2.10 for details). The stacked spectra of 92% dm HMT-PMBI (Figure 2.7) show the disappearance of two aromatic resonances due to deuterium-exchange, which are fully recovered after hydrogen-exchange. The location of deuterium-exchange was validated by comparison to two small molecule analogues DMMB, as shown in Figure 2.8, and HMT-MB (Figure A2).

41

Figure 2.7. Stacked 1H NMR spectra regions and corresponding chemical structures of 92% dm HMT-PMBI (a) initially in its cast, iodide form, (b) after 89 h in 2 M KOD/CD3OD/D2O at 60 °C to exchange the polymer to the deuterium form, and (c) after 90 h of the deuterium-exchanged polymer being subjected to 2 M KOH/CH3OH/H2O at 60 °C conditions to return the polymer to its original hydrogen-based form. The anions are not shown for clarity.

1 Figure 2.8. Aromatic regions of stacked H NMR spectra (500 MHz, CD3OD) of (a) deuterium-exchanged 92% dm HMT-PMBI, (b) hydrogen-exchanged 92% dm HMT-PMBI, and (c) DMMB in iodide form.

42

Since no new NMR spectra resonances were formed from 92% dm HMT-PMBI over 7 days, it can be assumed that steric C2-protection of the cationic group has been successful at negating ring-opening degradation. As the 92% dm HMT-PMBI-OH− is not fully methylated, the uncharged N-methyl groups were used as an internal standard for 1H NMR integration relative to the cationic N-methyl groups. Integration of these resonances indicates the dm% varies by <1% over the 7 day degradation experiment and 8 day deuterium-exchange experiment, thus providing evidence that there is also negligible nucleophilic displacement occurring, which showcases the remarkable hydroxide stability of this polymer.

2.3.4. Atropisomerization

An interesting aspect of the HMT group is the atropisomerization it exhibits in solution at room temperature, as shown in Figure 2.9. With a concentration of 50 mg of -1 HMTA per 1.0 mL of DMSO-d6 (50 g L ), the NMR resonances at 1.98 ppm appear as a doublet of doublets but in fact are two overlapping sets of two singlets derived from the inner 2, 2ʺ, 6, and 6ʺ methyl groups from hindered rotation around the central phenyl ring. The chemical shifts are further influenced by concentration, suggestive of aggregation. For example, with a concentration of 20 g L-1 and below, the methyl resonances appeared as a triplet. Likewise, this effect is also observed using variable temperature 1H NMR of HMTA (50 g L-1) between 25 and 148 °C, as shown in Figure 2.10. At 25 °C, atropisomerization is slower than the 500 MHz 1H NMR time-scale but faster at 50 °C, going from an apparent doublet of doublets to one doublet resonance. Upon increasing the temperature, the two resonances overlap at 101 °C and continue to move apart at 148 °C, explained by the average angle of preference around the central phenyl ring. HMTE also exhibits atropisomerization based on the additional alkyl NMR resonances than would be expected for a symmetrical molecule. Atropisomerization may be important in choosing the temperature used to cast or spray-coat HMT-PMBI-OH− as this may influence the solid state conformation of the polymer, free volume, and various physical properties. In this work, membranes were cast only from DMSO at 86 °C.

43

Figure 2.9. Racemization of HMTA in solution as observed by 1H NMR spectroscopy -1 (500 MHz) in DMSO-d6 at 20, 50, and 80 g L concentrations.

44

Figure 2.10. Superimposed variable temperature 1H NMR spectra (500 MHz) of HMTA -1 in DMSO•d6 at 50 g L concentration.

2.4. Conclusion

In summary, a novel alkaline anion exchange ionene was synthesized, named HMT-PMBI. Due to the sterically hindering methyl groups crowding the C2 position, the polymer showed exceptionally high stability in accelerated alkaline degradation conditions as well as possessing solubility in alcohol-based solvents, giving it potential for application in fuel cells and water electrolyzers. When fully methylated, the polymer is soluble in water. By using a new, air-insensitive, and scalable methylation procedure, the IEC of PBI materials can be easily controlled to produce water-insoluble HMT-PMBI, which requires a degree of methylation of ≤92%. As the degree of methylation increases, the conductivity exponentially increases, reaching conductivities similar to that of pendant-type cationic polymers. These materials have also been shown to comprise of an atropisomeric unit, which can be further investigated, as can the possibility of crosslinking via the remaining basic nitrogens.

45

Chapter 3.

Studying HMT-PMBI on Larger Scales

The work described in this chapter is reported in part in: Wright, A. G.; Fan, J.; Britton, B.; Weissbach, T.; Lee, H.-F.; Kitching, E. A.; Peckham, T. J.; Holdcroft, S. Energy Environ. Sci. 2016, 9 (6), 2130–2142.

Andrew Wright, Dr. Hsu-Feng Lee, Elizabeth Kitching, and Dr. Tim Peckham performed the synthesis. Elizabeth Kitching performed the small molecule degradation test. Thomas Weissbach performed the membrane casting. Dr. Jiantao Fan performed the ex situ polymer characterization (σ, Wu, swelling, mechanical strength, alkaline stability) and Benjamin Britton performed the in situ characterization (AAEMFC and water electrolyzer).

This work was financially supported by the Natural Sciences and Engineering Research Council of Canada (NSERC). This work made use of the 4D LABS shared facilities supported by the Canada Foundation for Innovation (CFI), British Columbia Knowledge Development Fund (BCKDF), Western Economic Diversification Canada (WD), and Simon Fraser University (SFU).

3.1. Introduction

Of the required criteria for an AEM to be used in AAEMFC and water electrolyzer applications (Section 1.2), the preliminary measurements in Chapter 2 of HMT-PMBI suggested that 3 of 7 were met. HMT-PMBI demonstrated good stability, similar to that of mes-PDMBI, had reasonably good hydroxide conductivity, and good solubility properties. However, the other criteria were not tested, such as scalability, nor was it demonstrated in situ as either a membrane or ionomer.

46

The long-term in situ stability of a cationic polymer that can act as both an anion- exchange membrane and ionomer would represent a significant advance in AEMFC and water electrolysis research. Such a material could serve as a benchmark material, allowing the effect of radical species on AEMs to be probed, for example, so as to form the basis for the development of accelerated durability tests.69 Furthermore, a chemically stable, high-conductivity anion-exchange ionomer that is resistant to CO2 impurities is required to assess the function and stability of novel alkaline catalysts. While AEMFC stability has been shown to some extent at 50 °C,91 higher temperatures are required to 86 69 increase hydroxide conductivity and improve CO2 tolerance. Additionally, a benchmark AEM material would require its synthesis to be scalable as well as possess a wide-range of properties, such as good mechanical properties, high anionic conductivity, low water uptake, low dimensional swelling, and high chemical stability.

In Chapter 3, it will be demonstrated that HMT-PMBI meets all 7 criteria as an AEM and ionomer. The synthesis of HMT-PMBI is up-scaled to show the versatility and reproducibility of its modified synthesis. Additionally, large scale synthesis allows for further extensive characterization of ionene-based membranes and elucidation of properties which are not currently understood. Every synthetic step is addressed for high yield and high purity, particularly the post-functionalization steps, where the yield is improved by >20% compared to the previous chapter. The 89% dm HMT-PMBI polymer, possessing, a theoretical OH- IEC of 2.5 mmol g-1, was chosen for extensive study due to the balance of high conductivity and low water uptake. The tensile strength and elongation at break are compared to commercial proton-exchange ionomer materials. Water uptake, dimensional swelling, and conductivity of various anions are determined and the upper limit of stability is found using extensive degradation tests. By utilizing the material as both the membrane and ionomer in an alkaline AEMFC and water electrolyzer for more than 4 days, HMT-PMBI is the first reported cationic polymer to operate between 60 and 90 °C, incorporating shutdowns, restarts, and exposure to CO2 during the operational cycle.

47

3.2. Experimental

3.2.1. Materials

Chemicals were purchased from Sigma Aldrich unless otherwise noted. Acetic acid (glacial) and potassium iodide (99.0%) were purchased from Caledon Laboratories Ltd.. Mesitoic acid (98%) and 1,4-phenylenediboronic acid (97%) were purchased from Combi-Blocks, Inc.. Ethanol (anhydrous grade) was purchased from Commercial Alcohols. Potassium hydroxide (ACS grade, pellets) was purchased from Macron Fine Chemicals. Dimethylsulfoxide (spectrograde), potassium carbonate (99.0%), potassium chloride (ACS grade), sodium bicarbonate (ACS grade), and hexanes (ACS grade) were purchased from ACP Chemicals Inc.. Dichloromethane (ACS grade, stabilized), silica (230-400 mesh, grade 60), sodium dithionite, acetone (ACS grade), methanol (ACS grade), and sodium chloride (ACS grade) were purchased from Fisher Scientific. Chloroform (ACS grade) and sodium hydroxide (ACS grade) were purchased from BDH. Activated charcoal (G-60) and hydrochloric acid (ACS grade) were purchased from Anachemia. Tetrakis(triphenylphosphine)palladium (99%) was purchased from Strem

Chemicals. Dimethylsulfoxide•d6 (99.9%-D), chloroform-D (99.8%-D), and methylene chloride-d2 (99.9%-D) were purchased from Cambridge Isotope Laboratories, Inc.. Nuclear magnetic resonance (NMR) spectra were obtained on a 400 or 500 MHz Bruker AVANCE III running IconNMR under Top Spin 2.1. The residual 1H NMR solvent resonances for DMSO•d6, CDCl3, and CD2Cl2 were set to 2.50 ppm, 7.26 ppm, and 13 5.36 ppm, respectively. The residual C NMR solvent resonances for DMSO•d6 and

CDCl3 were set to 39.52 ppm and 77.16 ppm, respectively. All NMR solutions had a solution concentration between 20 and 80 g L-1. The conductivity measurements under controlled humidity and temperature in CO2-containing air were collected in an Espec SH-241 chamber. The 5 L reactor used was a cylindrical jacketed flask (all glass), allowing the temperature to be controlled by a circulation of oil around the reactant mixture, which was generally a different temperature than the measured internal (reactant mixture) temperature. Eaton’s reagent was prepared prior to polymerization by stirring P2O5 (308.24 g) in methanesulfonic acid (2.5 L) at 120 °C under argon until dissolved, where it was then stored in sealed glass bottles until needed. Deionized water (DI water) was purified to 18.2 MΩ cm using a Millipore Gradient Milli-Q® water

48

purification system. MeB (2-mesityl-1,3-dimethyl-1H-benzimidazolium iodide) was synthesized according to literature.149

3.2.2. Synthesis

Up-scaled preparation of 3-bromomesitoic acid (BMA)

To the 5 L reactor was added glacial acetic acid (1.0 L) followed by mesitoic acid (225.1 g, 1.37 mol). The circulator temperature was set to 28.0 °C and mechanical stirred at 140 rpm. More glacial acetic acid was added (1.0 L) and stirred for approximately 30 min until the mesitoic acid fully dissolved. Bromine (100 mL, 1.95 mol) was added slowly over 5 min followed by glacial acetic acid (500 mL) to rinse down the sides. After 10 min, the internal temperature was observed to be 10 °C. The red mixture was stirred for 50 min, whereupon the internal temperature returned to approximately 25 °C. The mixture was then transferred by liquid transfer pump and PTFE tubing into 9 L of stirring distilled water (3 x 4 L beakers). The foamy precipitate was collected by vacuum filtration (requiring multiple funnels to collect all solid), compressed with a wide spatula to better dry the solid, and washed with water until white (~2 L total). The cakes were transferred to a 4 L beaker. The solid was recrystallized from approximately 2750 mL of 60% ethanol by boiling and then cooling to room temperature. The fluffy needles were collected by vacuum filtration, washed with room temperature 33% ethanol, and thoroughly dried at 80 °C under vacuum. This resulted in approximately 320 g of white needles. Two of these batches were combined and recrystallized a second time in 2700 mL of 55% ethanol, collected, washed with 33% ethanol, and dried under vacuum at 80 °C to yield 577.5 g of BMA as white fluffy needles (86.6%). 1H NMR

(500 MHz, DMSO•d6, ppm) δ: 13.33 (s, 1H), 7.09 (s, 1H), 2.33 (s, 3H), 2.32 (s, 3H), 2.19 13 (s, 3H). C NMR (125 MHz, DMSO•d6, ppm) δ: 170.02, 138.09, 135.11, 132.86, 132.19, 129.95, 124.45, 23.44, 20.88, 18.71. This procedure was repeated once more. The resulting data is shown in Table B1.

49

Up-scaled preparation of methyl 3-bromomesitoate (BME)

Potassium hydroxide pellets (36.0 g, 0.64 mol) were ground with a mortar and pestle to a fine powder and added to a 1 L round-bottom flask followed by DMSO (360 mL). The mixture was vigorously stirred for 30 min. BMA (104.4 g, 0.43 mol) was separately dissolved in DMSO (360 mL) and then added to the basic DMSO mixture, stirring for 15 min at room temperature. Iodomethane (40 mL, 0.64 mol) was then slowly added to the mixture (exothermic, temperature was kept below 40 °C) and then stirred closed for 2 h at room temperature. The mixture was then poured into 5 L of stirring ice- water and left stirring at room temperature until all of the ice melted. The precipitate was collected by vacuum filtration, thoroughly washed with water, and dried under vacuum at room temperature for at least 24 h to produce 106.1 g of BME as white crystals (96.4% 1 yield). H NMR (400 MHz, DMSO•d6, ppm) δ: 7.12 (s, 1H), 3.85 (s, 3H), 2.34 (s, 3H), 2.26 (s, 3H), 2.15 (s, 3H). The above procedure represents a “1.0 Scale”. For repeated syntheses, “2.0 Scale” represents the procedure being performed twice simultaneously and the final collected precipitates being combined prior to drying. The resulting data is shown in Table B2. The 1H NMR spectra of BME for each reaction performed are shown in Figure B1.

Up-scaled preparation of dimethyl 2,2’’,4,4’’,6,6’’-hexamethyl-p-terphenyl-3,3’’- diester (HMTE)

To the 5 L reactor was added 1,4-dioxane (2.9 L), BME (150.0 g, 0.58 mol), 1,4- phenylenediboronic acid (48.4 g, 0.29 mol), and 2 M K2CO3 (950 mL). The reactor was

50

connected to a water-cooled condenser and the mixture was degassed by bubbling argon through a needle sub-surface for 1 h. The needle was removed and Pd(PPh3)4 (1.80 g, 0.27% mol per BME) was added under a flow of argon. The circulator temperature was set to 105 °C and the solution was mechanically stirred at 280 rpm for 22 h, where the internal temperature read 89 °C. The dark yellow solution was then cooled to 60 °C and transferred by liquid transfer pump equally into 4 x 4 L beakers, each containing boiling and stirring 43% ethanol (2.62 L, aq.). The mixtures were stirred until they reached room temperature. The dark grey precipitates were collected by vacuum filtration and washed with water. The solid was dissolved in dichloromethane (1.0 L), washed with water (300 mL), and passed through a thick silica pad (~300 g). More dichloromethane (~4 L) was used to flush the silica and the filtrate was then evaporated by rotary evaporation to a pale yellow solid. The solid was then recrystallized in hexanes (5 L) by boiling until dissolved and cooling to -14 °C overnight. The white crystals were collected by vacuum filtration, washed with hexanes (400 mL), and dried under vacuum at 110 °C to yield 68.0 g of HMTE as fluffy, pure-white crystals (54% 1 yield). H NMR (400 MHz, CDCl3, ppm) δ: 7.15 (s, 4H), 7.00 (s, 2H), 3.92 (s, 6H), 2.33 13 (s, 6H), 2.03 (dd, J = 9.0, 4.3 Hz, 12H). C NMR (100 MHz, CDCl3, ppm) δ: 171.12, 139.82, 139.07, 137.72, 133.46, 132.82, 132.81, 132.30, 132.28, 129.48, 129.18, 129.17, 52.07, 21.04, 21.00, 19.57, 18.19, 18.15. The above procedure represents a “1.0 Scale”. For repeated syntheses, “Scale” represents an appropriately scaled version of all reactants and solvents by that factor. The amount of catalyst used was lowered for each subsequent reaction. The resulting data is shown in Table B3. The 1H NMR spectra of HMTE for each reaction performed are shown in Figure B2.

Purification of 3,3’-diaminobenzidine (DAB)

This procedure was modified from the literature.198 A 2 L Erlenmeyer flask was filled with distilled water. The water was boiled while bubbling with argon. The bubbling of argon was stopped and then, using an inverted glass funnel, a low flow of argon was kept over the solution for the next steps. The as-received DAB (25.0 g) was added to the

51

boiling water and stirred until dissolved. While boiling the solution, sodium dithionite (0.50 g) was added and stirred for 15 min. Activated charcoal (3.00 g) was then added and boiled for 30 min. The mixture was then quickly vacuum filtered through a hot funnel, producing a colorless filtrate. Argon was flowed through the filter flask and it was then kept sealed in the dark for 18 h. The resulting precipitate was then collected by vacuum filtration, washed with water, and quickly dried under vacuum at 100 °C. The collected recrystallized DAB was stored in the dark under argon until use. The purification process was repeated on more as-received DAB from different companies and the resulting data is shown in Table B4.

Up-scaled preparation of poly[2,2’-(2,2’’,4,4’’,6,6’’-hexamethyl-p-terphenyl-3,3’’- diyl)-5,5’-bibenzimidazole] (HMT-PBI)

To a 1 L, 3-neck round-bottom flask was added HMTE (20.0000 g, 46.5 mmol), recrystallized DAB (9.9535 g, 46.5 mmol), and Eaton’s reagent (800 mL, self-prepared).

Argon was flowed into the flask and out through a CaCl2 drying tube throughout the reaction. This mixture was heated at 120 °C until fully dissolved. After heating at 120 °C for 30 min, the temperature was increased to 140 °C for 20 min. The black solution was then slowly poured into distilled water (3.0 L) with manual stirring to break up the dense fibrous solid that formed. The solid was collected by vacuum filtration on glass fiber and washed with distilled water (1.5 L). The solid was then transferred to fresh distilled water (3.5 L) and the pH was adjusted to ~10 by addition of potassium carbonate (~70 g). The mixture was stirred overnight at room temperature. The solid was collected again, washed with water, boiling water, and acetone, and dried for at least 24 h at 100 °C to yield 26.0 g of HMT-PBI as fibrous solid (103% yield). The overestimated yield is likely due to trace water and acid in the fibrous solid, which will be later discussed. For 1 H NMR spectroscopy, HMT-PBI (~13.0 mg) was dissolved in DMSO•d6 (0.65 mL) by 1 addition of KOD (5 drops of KOD 40 wt% in D2O) and heating. H NMR (400 MHz,

DMSO•d6+KOH/D2O, ppm) δ: 7.76-7.51 (m, 2H), 7.51-7.32 (m, 2H), 7.32-7.15 (m, 4H),

52

7.16-7.02 (m, 2H), 7.02-6.83 (m, 2H), 2.27-1.91 (m, 12H), 1.91-1.70 (m, 6H). For repeated syntheses, the same method above was used and the resulting data is shown in Table B5. The 1H NMR spectra of HMT-PBI for each reaction performed are shown in Figure B3.

Up-scaled preparation of ~50% dm HMT-PMBI

To two separate 1 L, 3-neck round-bottom flasks was each added HMT-PBI (30.00 g, 55.1 mmol), DMSO (800 mL), and potassium hydroxide in water (14.00 g KOH in 35 mL H2O). Each was vigorously stirred and heated at 70 °C for 16 h closed. The viscous dark red/brown mixtures were cooled to room temperature and both were combined by decanting into one 2 L beaker. While manually stirring the mixture with a glass rod, iodomethane (21.0 mL, 337 mmol) was added (exothermic). The dark- coloured mixture was stirred for approximately 5 min until the mixture became a chunky, pale brown sludge. The mixture was then poured equally into 4 x 4 L beakers, each containing distilled water (3 L). To each beaker was then added potassium iodide (5.0 g) and briefly stirred with a glass rod. The precipitate was collected by vacuum filtration and washed with water. The collected cakes were transferred to a clean 4 L beaker and the cakes were beaten to a powder using a metal spatula. This wet solid was then stirred for 16 h in acetone (3 L) with potassium iodide (15.0 g). The solid was collected by vacuum filtration and washed with acetone. The yielded cakes were added to a 1 L container and beaten again to a powder. The solid was dried under vacuum at 80 °C for at least 24 h yielding 58.2 g of 53.7% dm HMT-PMBI as a pale brown powder (88.9% yield). 1H NMR

(400 MHz, CD2Cl2, ppm) δ: 8.28-7.45 (m, 6.03H), 7.44-7.09 (m, 6.00H), 4.46-3.87 (m, 0.91H), 3.87-3.39 (m, 5.61H), 2.33-1.97 (m, 11.71H), 1.97-1.70 (m, 7.39H). The amount of iodomethane used, the precipitation solvent, and amount of potassium iodide used was varied in subsequent reactions and the resulting data is shown in Table B6. The 1H NMR spectra of ~50% dm HMT-PMBI for each reaction performed are shown in Figure B4. The dm% was calculated using Equation 1 (from Section 2.2.4).

53

Up-scaled preparation of >55% dm HMT-PMBI

To a 1 L round-bottom flask containing dichloromethane (300 mL) was added ~50% dm HMT-PMBI (34.00 g, 51.4% dm) followed by additional dichloromethane (400 mL). The solid was broken up inside with a spatula and the mixture was stirred for 1.5 h until mostly dissolved. Iodomethane (13.0 mL, 209 mmol) was added and the mixture was stirred for 18 h closed at 30 °C. The precipitate was broken up with a spatula and the stirring was continued for 3 h at room temperature. The solvent was evaporated at 44 °C by dynamic vacuum, leaving a strong solid film stuck to the inner glass wall. Methanol was added and heated to dissolve the polymer and then transferred to a large, flat glass dish, using additional methanol to rinse all of the contents into the large dish. The solvent was evaporated in air at room temperature and then under vacuum at 100 °C, yielding one thick 45.6 g brown film of 90.2% dm HMT-PMBI (97.2% yield). The 1H NMR spectra were taken of washed and dried membranes. 1H NMR

(400 MHz, DMSO•d6, ppm) δ: 8.97-7.66 (m, 6.15H), 7.66-7.04 (m, 6.00H), 4.30-3.78 (m, 9.57H), 3.78-3.50 (m, 1.16H), 2.44-1.49 (m, 17.88H). For repeated syntheses, the polymer was purified by different methods, such as precipitation into acetone rather than evaporation of dichloromethane. Additionally, if a lower than desired dm% was yielded, such as 86% dm instead of 89% dm, the same procedure could be repeated using DMSO as the solvent and a stoichiometric amount of iodomethane at 30 °C for 18 h. The resulting synthetic data is shown in Table B7. The dm% was calculated using Equation 1.

3.2.3. Membrane Preparation

HMT-PMBI (89% dm, iodide form, 3.5 g) was dissolved in DMSO (46.67 g) by stirring and gently heating for 12 h. The polymer solution was vacuum filtered through glass fiber at room temperature, cast on a levelled glass plate using a K202 Control Coater casting table and a doctor blade (RK PrintCoat Instruments Ltd) and stored in an oven at 85 °C for at least 12 h. The membrane peeled off the glass plate upon immersion in distilled water. After soaking the membrane in distilled water (2 L) for 24 h, the membrane was dried under vacuum at 80 °C for 24 h.

54

3.2.4. Degree of Methylation Determination

The degree of methylation (dm) for polymers possessing >55% dm was calculated as previously described in Chapter 2. Specifically, using the baseline corrected (MestReNova 9.0.1, “Full Auto Polynomial Fit”) 1H NMR spectrum of >55% dm

HMT-PMBI (400 MHz, DMSO•d6), the integration region 4.30-3.78 ppm was set to 12.00H and the respective integration for 3.78-3.50 ppm was calculated to be x. The dm% was then calculated using Equation 1. Also using Equation 1, the dm% for 1 ~50% dm HMT-PMBI was calculated from its H NMR spectrum (400 MHz, CD2Cl2), where the integration of 4.46-3.87 ppm was set to 12.00H and the respective integration for 3.87-3.39 ppm was calculated to be x.

3.2.5. Ion-Exchange Capacity (IEC)

- IECX-, representing the IEC of HMT-PMBI in a specific X anion form, was calculated using Equation 2 as previously described in Chapter 2, where is the molar mass (g mol•1) of one repeat unit of 100% dm HMT-PMBI (including the mass of - •1 two X counter ions), and is the molar mass (g mol ) of one repeat unit of 50% dm HMT-PMBI.

3.2.6. Water Uptake

HMT-PMBI membranes were soaked in corresponding 1 M aqueous MX solutions at room temperature for 48 h (exchanged twice), where MX represents KF,

KCl, KBr, KI, Na2SO4, KOH, KHCO3, or K2CO3. The membranes were washed with deionized water (DI water) several times and soaked in DI water for another 48 h at room temperature (with three exchanges of water). A fully hydrated (wet) membrane was removed and weighed (Ww) immediately after excess water on the surface was removed with tissue paper. The hydrated membrane was dried under vacuum at 40 °C to a constant dry weight (Wd). The water uptake (Wu) was calculated using Equation 4, in Chapter 2.

55

3.2.7. Dimensional Swelling

The procedure for determining dimensional swelling was analogous to determining water uptake wherein the wet dimensions (Dw) and dry dimensions (Dd) were measured. The percent directional dimensional swelling (Sx, Sy, and Sz) was calculated by using Equation 5 below,

( ) (5)

where Sx represents the dimensional swelling in the x-direction (length-direction) and is calculated from Dw and Dd, which represent the dimensions in the x-direction of the wet and dry membrane, respectively. Sy and Sz represent the dimensional swelling in the y- and z-directions (width and thickness), respectively, using their respective Dw and Dd values in the y- and z-directions. The percent volume dimensional swelling (Sv) was calculated using Equation 6 below,

( ) (6)

where Vw and Vd represent wet and dry volumes, determined from the products of the x-, y-, and z-dimensions Dw and Dd, respectively.

3.2.8. Electrochemical Impedance Spectroscopy

The ionic conductivity was measured and calculated as previously described in Chapter 2 using Equation 3. Unless otherwise noted, the conductivity was measured for fully hydrated membranes at ambient temperature (~22 °C) and in air.

For measuring mixed hydroxide/bicarbonate/carbonate ionic conductivities under controlled temperature and relative humidity (RH) conditions, two membranes were first soaked in argon-degassed 1 M KOH for 48 h. The membranes were then washed with degassed DI water for 24 h. After the surface water was removed with tissue paper, the membrane was mounted on a two-point probe inside a pre-set humidity chamber (Espec

56

SH-241) and left to equilibrate in air for 16 h. At a given humidity, the temperature was increased in 10 °C increments, and the membrane equilibrated for 30 min before measuring the resistance. When the humidity was changed, the cell was allowed to equilibrate for 2 h before the first measurement. The average of the two membrane conductivities was used.

3.2.9. Anion Concentration

The anion concentration, [X-], in HMT-PMBI membranes was calculated using Equation 7 below,

[ ] (7)

where IECX-, Wd, and Vwet are the IEC, dry weight, and wet volume of HMT-PMBI in the - - - 2- X form, respectively. The anion concentration in OH , HCO3 , and CO3 forms were - calculated using the IEC of HMT-PMBI in the pure HCO3 form.

3.2.10. Mechanical Strength

The membranes were die-cut to a barbell shape using a standard ASTM D638-4 cutter. The mechanical properties of the membranes were measured under ambient conditions on a single column system (Instron 3344 Series) using a crosshead speed of 5 mm min-1. The determined tensile strength, Young’s moduli, and elongation at break were averaged over at least three samples. The error reported is the standard deviation. To convert from the as-cast iodide form to the chloride form, the membrane was soaked in 1 M NaCl for 48 h (exchanged twice), soaked in DI water for 48 h (with multiple exchanges), and dried at 80 °C under vacuum for 16 h before cutting. To convert from the as-cast iodide form to the hydroxide form, the membrane was soaked in 1 M KOH for 48 h (exchanged twice) and soaked in DI water for 48 h (with multiple exchanges), where it was then cut and measured in its wet (fully hydrated) state.

57

3.2.11. Chemical Stability

A model compound of the polymer, 2-mesityl-1,3-dimethyl-1H-benzimidazolium iodide (MeB) (83 mg), was dissolved in a 2.0 mL mixture of 10 M KOHaq with 3.0 mL methanol (resulting in a 4 M KOH solution). The mixture was heated to 80 °C for 7 days. After cooling to room temperature, a red solid was collected by filtration, washed with water, and dried under vacuum at 60 °C. The solid was dissolved in DMSO•d6 and analyzed by 1H NMR spectroscopy.

Prior to testing the chemical stability of HMT-PMBI in various ionic solutions, the as-cast iodide form was converted to the chloride form by soaking the membrane in 1 M NaCl for 48 h (exchanged twice) and then in DI water for 48 h (multiple fresh exchanges). The membrane was subjected to degradation tests using various conditions in closed HDPE containers. Following the degradation test, the membrane was exchanged back to chloride form by soaking in 1 M NaCl for 48 h (exchanged twice) and then in DI water for 48 h (multiple fresh exchanges). The ionic conductivity of this membrane in its wet form was measured and the membrane was dried at 50 °C for 16 h. -1 Membrane pieces were dissolved in DMSO•d6 (25 g L concentration) and analyzed by 1H NMR spectroscopy. The relative amount of benzimidazolium remaining was calculated from the 1H NMR spectroscopic data using Equation 8 below,

Remaining (%) ( ) (8)

where y represents the integration area between 6.00-4.35 ppm relative to 12.00H for the 9.20-6.30 ppm area for the sample of interest and z represents the integration area between 6.00-4.35 ppm relative to 12.00H in the 9.20-6.30 ppm region for the initial sample.

3.2.12. Assembly of the Catalyst-Coated Membrane

An HMT-PMBI membrane (89% dm) was first exchanged from iodide to chloride form by immersion in 1 M KCl for 7 days followed by soaking in DI water for two days

58

with one fresh exchange of DI water half-way through. The chloride form HMT-PMBI was dissolved in MeOH to form a 10 wt% ionomer dispersion. Separately, a catalyst mixture was prepared by adding water followed by methanol to commercial carbon- supported Pt catalyst (46.4 wt% Pt supported on graphitized C, TKK TEC10E50E). The ionomer dispersion was added drop-wise to the catalyst mixture while the solution was rapidly stirred. This resulting catalyst ink contained 1 wt% solids in solution and a 3:1 - (wt/wt) MeOH:H2O ratio. The solids comprised of 15 wt% HMT-PMBI-Cl and 85 wt% Pt/C. 15 wt% HMT-PMBI-I- catalyst ink was similarly produced from the iodide form. 25 wt% FuMA-Tech FAA-3 (Br-) catalyst ink was similarly produced using commercial ionomer dispersion (FAA-3, 10 wt% in NMP).

To form the catalyst-coated membrane (CCM), a membrane was fixed to a vacuum table at 120 °C. The HMT-PMBI-Cl- membrane thickness used for the AEMFC testing was 34 ± 2 µm; for the water electrolysis, it was 43 ± 2 µm. The commercial membrane (FAA-3, Br-) thickness used for the AEMFC testing was 20 µm; for the water electrolysis, it was 50 µm. The prepared catalyst ink was applied using an ultrasonicating spray-coater (Sono-Tek ExactaCoat SC) to create a 5 cm2 electrode area with cathode/anode catalyst loadings of 0.4/0.4 mg Pt cm-2 for the AEMFC or 0.5/0.5 mg Pt cm-2 for the water electrolyzer. The HMT-PMBI-Cl- CCM was then immersed in 1 M KOH in a sealed container for 7 days and DI water for 7 days in a sealed container. FAA-3 CCMs were non-operational after this process and the FAA-3 CCM was instead immersed in 1 M KOH for 24 h. For comparison, the HMT-PMBI-I- CCM were exchanged in 1 M KOH for 24 h. Gas-diffusion layers (GDL, Sigracet 24BC) were applied to the electrodes and gasketing of a specific thickness was chosen to achieve a 20-30% GDL compression. The resultant assembly was torqued to 2.26 N m. Alignment and adequate compressions were confirmed by using a pressure-sensitive film (Fujifilm Prescale LLLW). For water electrolysis, CCMs were mounted in a fuel cell hardware modified for alkaline electrolyte stability, which included Ti flow fields. CCMs were laminated, and a 50 µm thick Ti screen (60% open area) was applied to the electrodes, with gasketing that was sufficient to provide a zero gap between the flow field, Ti screen, and the electrode.

59

3.2.13. AEMFC Operation

The resultant HMT-PMBI (OH-, previously in Cl-) AEMFC was conditioned at

100 kPaabs and 60 °C under 100% RH and H2/O2 and subsequently operated at

300 kPaabs. The potential and resistances measured stabilized for current densities >100 mA cm-2 after 8 h operation. The CCM was conditioned by running multiple, slow polarization sweeps. The current was increased stepwise from open circuit voltage (OCV) at a rate of 10 mA cm-2 per 5 min up to a 0.15 V cut-off. Over the operation time of the AEMFC, multiple sets of polarization data were taken at 5 min pt-1 from OCV at 5 or 10 mA cm-2 intervals to a 0.2 V cut-off, with additional 1 min pt•1 steps at 2 mA cm-2 intervals between 0-20 mA cm-2 in order to resolve the kinetic polarization region. Multiple fuel cells were constructed and tested under different conditions. In the fuel cell test presented here, the fuel cell was subjected to a 5-day shutdown after 60 h of operation, and then restarted for an additional 10 h. At 70 h of total operation, CO2- containing air was used as the gas feed for 21 h (70-91 h time period) rather than pure

O2 in order to examine fuel cell operation using ambient air, before returning the gas feed back to pure O2. During the following 109-114 h operational time using pure O2, the temperature was increased to 70 °C for 1 h and then increased in 5 °C intervals to 90 °C. The same conditioning procedures and conditions were used for FAA-3 CCMs - and HMT-PMBI-I CCMs for comparison. For operation using 100 kPaabs, polarization data was taken at 5 s pt-1.

3.2.14. Water Electrolyzer Operation

The water electrolyzer used a 1 M KOH circulating electrolyte flow heated to 60 °C, separately supplied to the anode and cathode at a rate of 0.25 mL min-1. The electrolyte was circulated for 1 h prior to electrolysis to ensure the polymer within the CCM was converted to the hydroxide form. 20 mA cm-2 was drawn from the FAA-3 based cell and 25 mA cm-2 was drawn from the HMT-PMBI based cell, using a Solartron SI 1260. The experiment was terminated for FAA-3 cells when the applied potential reached 3 V or fell below 1.2 V, corresponding to two different modes of cell failure, as reported in literature.206 The hydrogen evolution reaction and oxygen evolution reaction current density attributable to the Ti screens was measured ex situ in 1 M KOH

60

potentiodynamically, and accounted for <0.1 mA cm-2 in this potential range. Rates of hydrogen and oxygen gas evolution were also observed by inspection.

3.2.15. AEMFC Power Density Error Analysis

All potential data taken at 5 min pt-1 in the polarization curves representing 47.7, 48.9, and 49.2 mW cm-2 were paired according to current density (equal fluxes) and subjected to a paired student t-test with α = 0.005, rejecting the Ho that difference = 0 in each comparison, statistically determining all three curves representing the three mentioned power densities were indeed different, as the p-values calculated were <0.005 for each pair. This gives a >99.5% confidence that successive improvements between each polarization curve occurred and therefore that the maximum power densities are different. Additionally, when five consecutive polarization curves were taken, the average power density of 45.2 mW cm-2 had a standard deviation of ± 0.2 mW cm-2, demonstrating that the peak power values of 47.7, 48.9, and 49.2 mW cm-2 are significantly different.

3.3. Results and Discussion

3.3.1. Up-Scaled Monomer Synthesis

The scaled-up synthesis of the monomer HMTE was performed on multiple smaller scale setups, in either a 5 L reactor or a 1 L round-bottom flask. The average yields and standard deviations for multiple syntheses are shown in Scheme 3.1. The quantities shown are the total used (u) or the yield obtained (y). The synthetic route used mesitoic acid (900 g, 5.5 mol) which was readily brominated to yield BMA in high yield (87.4 ± 1.1%, 4 runs, 5 L scale, Table B1 in Appendix B). This yield was significantly higher than the original reported 60% yield199 and our previously reported yield of 74% (Chapter 2). The reason for the increase in yield was found due to using excess bromine rather than a stoichiometric amount, as dibromination was not observed to be significant. The reaction was also found to be endothermic; addition of bromine would lower the internal temperature from 25 °C to <10 °C within 15 min. As such, if high concentrations of mesitoic acid in acetic acid were used (such as 18% wt. mesitoic acid), addition of

61

bromine would result in precipitation/freezing of the mixture, which led to difficulties in purification. To remedy this, only 10% wt. mesitoic acid was used and the temperature of the circulating fluid around the mixture was set to 28 °C, resulting in a minimum internal temperature of 10 °C during the reaction and a final temperature 25 °C after 50 min. To transfer the solution out of the 5 L reactor, a liquid transfer pump with PTFE tubing was used to add the mixture to water, resulting in precipitated product, but which also contained ~10% mesitoic acid by 1H NMR spectroscopy. This suggests that even more bromine could theoretically be used and/or the reaction time could be increased for a further improvement in yield but this may begin to produce dibrominated mesitoic acid. Instead, recrystallization was used to purify the material. One recrystallization in ethanol/water decreased the amount of mesitoic acid to ~2% and a second recrystallization resulted in high purity BMA as fluffy white needles.

Scheme 3.1. Scaled-up synthetic route to HMTE, showing the average yield and standard deviation for multiple batches of each reaction. The amounts of reagents and products shown represent the total amount used over the multiple syntheses performed.

The second step involved methylation of BMA to BME, which was achieved in near-quantitative yield (98.8 ± 1.1%, 11 runs, 1 L scale, Table B2). This reaction was not

62

performed in the 5 L reactor due to the potential difficulty of transferring the mixture after the reaction, which contains powdered potassium hydroxide. Instead, it was performed in multiple 1 L round-bottom flasks. The addition of iodomethane to the reaction was found to be exothermic, and was therefore slowly added, such that the internal temperature did not exceed 40 °C. After the reaction, pouring the DMSO mixture into ice-water resulted in crystallization of the desired product as colourless blocks, thus requiring no additional purification; only trace DMSO (~2% by 1H NMR spectroscopy, Figure B1) was found in the isolated product.

The third step involved Suzuki-Miyaura coupling of BME with 1,4•phenylenediboronic acid to form monomer HMTE, which was reproduced in 57 ± 2% yield (Table B3). The 5 L reactor was used for this reaction, which required the reaction to be repeated 8 times. The amount of catalyst, Pd(PPh3)4, was varied in each run, initially starting with 0.54% mol catalyst per BME. The amount of catalyst was decreased to as low as to 0.08% mol catalyst per BME without any reduction in yield, suggesting that even less could be used. High purity HMTE was obtained from every batch, as judged by their indistinguishable 1H NMR spectra (Figure B2, Appendix B).

In contrast to the original synthesis of HMT-PBI in Chapter 2 where monomer HMTE was first hydrolyzed to its diacid form, HTMA, and then polymerized, polymer HMT-PBI was obtained directly from monomer HMTE. The original hydrolysis route involved dissolving HMTE in concentrated sulfuric acid and precipitating into water. However, as the polymerization to HMT-PBI involves Eaton’s reagent, the hydrolysis was found to occur in situ, thus eliminating the need for a specific hydrolysis step, and reducing the monomer synthesis by one step. The overall yield for monomer HTME from mesitoic acid was improved from the previously of 40% to 49%, an overall increase of 23%.

The second monomer, DAB, which was purchased from more than one company and having different purities, was recrystallized prior to every polymerization in order to ensure uniform DAB purities and thus reproducible polymers (Table B4).

63

3.3.2. Polymerization and Post-Functionalization

The polymerization and post-functionalization steps were all performed in 1 L round-bottom flasks due to the high viscosity of the solutions and the ease of cleaning, as the flasks had to be manually scrubbed to remove residual polymer on the walls. The synthesis of HMT-PBI was first optimized on small scale for concentration of HMTE, as higher concentrations of monomer require less Eaton’s reagent and thus produce less waste. In Chapter 2, the synthesis of HMT-PBI involved reacting 1.0 g HMTA with DAB in 40 mL Eaton’s reagent (25 g L-1). Reacting HMTE at the same concentration and poured into water resulted in paper-like precipitate, as shown in Figure 3.1. However, when 1.0 g HMTE was reacted at a higher concentration, with only 25 mL of Eaton’s reagent (40 g L-1), one tough, continuous fiber was formed. The fiber contained significant amounts of Eaton’s reagent within the fibers. When both polymers were functionalized to 90% dm HMT-PMBI, they both produced strong and flexible membranes with similar hydroxide conductivities. However, the fibrous polymer was very difficult to dissolve, which formed gels above 2% wt. solutions in DMSO. Due to the difficulty of handling the fibrous material, 25 g L-1 concentration was chosen for the up- scaled synthesis. Additionally, as the precipitation of the hot solution into water required a cool-down time before filtration, the precipitation was also tested in ice-water, as shown in Figure 3.2. Unfortunately, precipitation into ice-water of a 25 g L-1 solution resulted in formation of a thick fiber, similar to the previous 40 g L-1 reaction. As such, the up-scaled polymer was precipitated in room temperature water, requiring several hours for the mixture to cool prior to purification.

64

(a) (b) Figure 3.1. Photograph of the resulting precipitated HMT-PBI in water after reacting HTME with DAB with a concentration of HMTE in Eaton’s reagent of (a) 25 g L-1 and (b) 40 g L-1.

65

(a) (b) Figure 3.2. Photograph of the resulting precipitated HMT-PBI (25 g L-1) in (a) room temperature water and (b) ice-water.

HMT-PBI was produced with a theoretical yield of 102 ± 2% (26 runs, 1 L scale), as shown in Scheme 3.2. The over-estimated yield is most likely due to residual acid present in the polymer fibers. While there were small differences in colour and thickness of the precipitated polymer from batch to batch (Table B5), the 1H NMR spectra (Figure B3) show no visible differences in the expected resonances. Due to the negligible variability of each batch, batches were combined, blended, and ground into a saw-dust-like powder using a 700 W blender and mixed together in a 12 L vessel (the 673 g of powder occupied ~10 L volume). This ensures that all post-functionalized batches have the same molecular weight.

66

Scheme 3.2. Scaled-up synthetic route to HMT-PMBI, showing the average yield and standard deviation for multiple batches of each reaction. The amounts of reagents and products shown represent the total amount used over the multiple syntheses performed. The last step shows the yield after optimization.

The following partial methylation procedure of HMT-PBI to ~50% dm HMT-PMBI was varied throughout the repeated syntheses in order to optimize the yield and decrease the purification time. For example, the average yield of batches 1–4 and 5–11 were 81 ± 2% and 88.7 ± 0.5%, respectively (see Table B6). This significant increase in yield was due to the addition of potassium iodide in the purification of the polymer in batches 5–11. For example, when the polymer was precipitated from DMSO into water, its amphiphilic nature caused it to partially dissolve, due to its over-methylation of 55 ± 2% dm. The addition of potassium iodide prevents the polymer from dissolving without potentially exchanging the counter-ion, for example, to chloride, if sodium chloride was used instead. This lowers the time needed to filter the polymer for batches 5–11, which also possessed significantly less solvent impurities than the initial batches (Figure B4), and is likely due to the ability to better wash the filtered polymers. An unassigned 1 resonance in all of the H NMR spectra of ~50% dm HMT-PMBI (400 MHz, CD2Cl2) polymers was observed at 0.13 ppm, which may be due to methylated silicates arising from the hot KOH–DMSO solution that etches the glass walls of the reactors. This suggests that non-glass reaction vessels would perform better for repeated large scale batches.

67

The final synthetic step was the controlled methylation of ~50% dm to >55% dm HMT-PMBI in dichloromethane using iodomethane, which was achieved in near- quantitative yield (98.1 ± 1.1%) over a range of dm%. The original procedure involved the precipitation of the ionene from the dichloromethane solution into acetone, but this procedure led to a 20•30% loss in yield. Instead, evaporation of the dichloromethane in the round-bottom flask by rotary evaporation resulted in nearly quantitative yield for all degrees of methylation. As it was often difficult to remove the polymer from the round- bottom flask without breaking the flask (due to the strength of the polymer), those with >80% dm were dissolved in methanol and transferred into a large glass dish. If the dish was gently heated on a hot plate in air to evaporate the solvent, the cast membrane would stick to the bottom of the glass, expand slightly when nearly dried, and shatter the glass dish, peeling a sheet of glass off with it (see Figure 3.3a). Instead, placing the methanol-containing dish into an 80 °C vacuum oven and vacuum drying the solvent quickly yielded a porous and flexible material without breaking the dish (Figure 3.3b).

(a) (b) Figure 3.3. Photograph of >55% dm HMT-PMBI-I- cast (~45 g) from methanol using (a) a hot plate to dry the solvent in air, showing the bottom of the membrane with a layer of glass, and (b) a vacuum oven to quickly dry the solution to a porous material.

While a number of polymer batches of >55% dm HMT-PMBI were prepared in order to show the extent of control and reproducibility, only those polymers with 89% dm

68

were subjected to characterization and stability tests. The choice of 89% dm arises from the previous work in Chapter 2 in which we showed that this dm%, and hence IEC, provided membranes with balanced ionic conductivity, water uptake, and mechanical strength.

The overall synthetic yield, based on the starting mesitoic acid to >55% dm HMT- PMBI was 42 ± 3%, which is high for a six-step synthesis. Each step showed high reproducibility in terms of yield, as well as purity. The synthesis of 617 g of 55 ± 2% dm HMT-PMBI, which corresponds to 1.03 mol repeat units, demonstrates the versatile scale-up of this synthetic route.

3.3.3. Hydroxide Ion Conductivity

The as-cast 89% dm HMT-PMBI-I- membrane was converted to hydroxide form by immersion into 1 M KOH for 48 h, followed by washing with argon-degassed water several times. The ionic resistance in air was then measured immediately (~3 min after exposure to air) over time by electrochemical impedance spectroscopy (EIS). The initial conductivity of 23 mS cm-1 decreased rapidly upon exposure to air and leveled off at 8.1 mS cm-1 after ~40 min, as shown in Figure 3.4. This effect was previously discussed in Chapter 2 and is attributed to rapid conversion of hydroxide to a mixed hydroxide/bicarbonate/carbonate form upon exposure to atmospheric CO2.

69

Figure 3.4. Ionic conductivity as measured by EIS of 89% dm HMT-PMBI, initially in OH- form, at 95% RH and 30 °C in air over time, where the inset shows the expanded, 0–60 min, region.

Accordingly, after 16 h equilibration in air, the mixed carbonate conductivity was measured at various temperatures and relative humidity (RH), which followed Arrhenius- type behavior, as shown in Figure 3.5a. The highest conductivity was measured at 95% RH and 90 °C to be 17.3 mS cm-1. Based on the Arrhenius expression (Equation 9 below), the conductivity (σ) is dependent on the pre-exponential factor A, the activation energy Ea, and the temperature T. Changes in relative humidity will affect the activation energy due to changes in swelling, ion concentration, and ion diffusivity as well as the pre-exponential factor, which is related to the dielectric constant.207 Using the collected data in Figure 3.5a, the activation energies (Ea) were calculated at each humidity level using Ea = -mR, where m represents the slope of the linear regression for ln σ vs. 1000/T and R represents the universal gas constant (8.314 J mol•1 K•1). Between 70–95% RH, -1 89 Ea was calculated to be 25–26 kJ mol , which is typical for bicarbonate AEMs. However, as the RH was decreased below 60%, the activation energy linearly increased, as shown in Figure 3.5b, possibly due to the loss of accessible cationic sites that are immobilized in the backbone. This suggests that an RH of at least 60% is required to hydrate the polymer for unhindered bicarbonate conduction.

70

(9)

Figure 3.5. (a) Arrhenius plot of ion conductivity of 89% dm HMT-PMBI membranes in mixed carbonate form at various temperatures and RH, in air, and (b) the corresponding calculated activation energy at a given RH.

71

3.3.4. Physical Properties of HMT-PMBI Incorporating Various Ions

The water uptake (Wu), volume dimensional swelling (Sv), and directional swelling (Sx, Sy, and Sz) were measured for 89% dm HMT-PMBI after soaking for 48 h in various 1 M ionic solutions, to exchange the anion, and washed with water for 48 h. The resulting water uptake and swelling are shown in Figure 3.6. Figure 3.6a illustrates a proportional relationship between dimensional swelling and water uptake for the monovalent anions, with the exception of the fluoride ion form. Dimensional swelling increased in the order of I- < Br- < F- < Cl- < OH-. This unusual behavior of the fluoride form is more clearly observed in plots of directional swelling (Figure 3.6b), where KF results in significant anisotropic swelling. The fluoride form swells by almost three times more in each in-plane direction compared to the out-of-plane (Sz) whereas the other halogens exhibit minor increases in thickness relative to in-plane swelling. The relative decrease in swelling in the thickness direction of the fluoride ion form is similar to that of 2- 2- the bivalent anions (CO3 and SO4 ), which have the ability to ionically-crosslink the polymer. The observed anisotropic dimensional swelling of the fluoride ion form may be due to the anisotropic orientation of the polymers, i.e., aligning parallel to the in-plane direction, due to the slow evaporation process during casting, but this would require further study for validation.

72

Figure 3.6. (a) Volume dimensional swelling (Sv) versus water uptake (Wu), including a dashed trendline which excludes K2CO3, Na2SO4, and KF and (b) directional dimensional swelling (Sx, Sy, or Sz) for 89% dm HMT-PMBI after being soaked in various 1 M ionic solutions and washed with water. Sz represents the out-of-plane swelling.

The conductivity of wet membranes of each ion form is shown in Table 3.1. The highest conductivity was observed for membranes ion-exchanged using KOH solution; exchange with KCl produced a membrane with the second highest conductivity. The conductivity differences between KOH, KHCO3, and K2CO3 were larger than expected,

73

as each ion is known to equilibrate to a similar mixed carbonate form in air. A possible reason is due to the mechanical changes that occur due to different swelling behavior in the various ionic solutions, as previously shown in Figure 3.6.

Table 3.1. Diffusion coefficients at infinite dilution of anions and the respective anion conductivity and anion concentration of HMT-PMBI.

a 5 2 -1 b -1 - c Solution Dx- (x 10 cm s ) σx- (mS cm ) [X ] (M) KF 1.48 6.2 ± 0.2 1.86 ± 0.10 KCl 2.03 7.5 ± 0.4 1.7 ± 0.2 KBr 2.08 4.2 ± 0.6 1.89 ± 0.09 KI 2.05 0.87 ± 0.01 2.04 ± 0.07 KOH 5.27 10.0 ± 1.2 1.51 ± 0.07d d KHCO3 1.19 3.8 ± 0.4 1.57 ± 0.09 d K2CO3 0.92 2.0 ± 0.2 1.69 ± 0.12 a 205 Literature data for diffusion coefficients of anions (Dx-) in aqueous solution at 25 °C. bAnion conductivity (σ) of 89% dm HMT-PMBI membranes in water after anion exchange in 1 M solutions at room temperature. cAnion concentration in HMT-PMBI at room temperature. d - As these ion forms were not protected from atmospheric CO2, the IEC of the HCO3 was used in their calculation of [X-].

The differences in conductivity of membranes containing different anions does not correlate to the corresponding diffusivity coefficient of the anion at infinite dilution, D, listed in Table 3.1. For example, the diffusivity coefficient at infinite dilution is similar for Cl-, Br-, and I-, but the conductivity decreases in the order of Cl- > Br- > I-. This trend is likely due to differences in water uptake and differences in dimensional swelling of the membranes, the effect of an anion possessing different dissolution enthalpies,208 and the fact that the anions are far from being at infinite dilution – the anion concentration in the volume of the membrane is in the order of 1.5–2 M (Table 3.1), which means an even greater concentration in the ionic channels. This brings into question the validity of using D values to estimate hydroxide conductivities based on mixed carbonate forms (ratio of 3.8) or chloride forms of the polymer, which have been previously used in the literature to draw comparisons between anions.78,89,130,204

The ability of the hydroxide ion to convert to the mixed carbonate form makes the measurement of the hydroxide conductivity form unreliable. As a result, measurements pertaining to degradation tests and mechanical properties of HMT-PMBI after exposure

74

to different aggressive conditions were reconverted to the chloride form, as the chloride form exhibits the next closest properties to the hydroxide ion form, yet is stable in air.

3.3.5. Mechanical Strength

Tensile strength, elongation at break, and the Young’s modulus were measured for 89% dm HMT-PMBI membranes using the as-cast iodide, chloride-exchanged, or hydroxide-exchanged form (Figure 3.7). At least three measurements were performed on each form and their properties are tabulated in Table 3.2. The tensile strength of the as-cast HMT-PMBI membrane (I-, dry) was measured to be 64.7 ± 0.3 MPa, equivalent to the high performance polymer, m-PBI (65 MPa),209 which has a similar backbone. However, the elongation of 89% dm HMT-PMBI (97 ± 13%) is higher than m-PBI (2%)209 by two orders of magnitude. From these data, HMT-PMBI is viewed as being exceptionally strong and flexible for an ionic solid polymer electrolyte. The tensile strength of the dry polymer decreased when the iodide was exchange for chloride form, and furthermore decreased when in a wet state, which is attributed to the increased water uptake and dimensional swelling of the chloride forms, as previously shown. However, when the membrane was in its mixed hydroxide form, it had a lower tensile strength and Young’s modulus of 12 ± 5 MPa and 19 ± 2 MPa, respectively, yet a higher elongation at break of 150 ± 30%. Nevertheless, the wet chloride and mixed hydroxide forms possessed comparable mechanical properties to the commercial ion-exchange membrane, Nafion 212, illustrating that HMT-PMBI exhibits robust mechanical properties, potentially suitable for fuel cell or water electrolyzer applications.

75

Figure 3.7. Mechanical properties of 89% dm HMT-PMBI membranes in different ion forms, such as the as-cast form (I-, dry), chloride-exchanged wet and dry forms, and hydroxide-exchanged wet form, measured in air.

Table 3.2. Mechanical properties of HMT-PMBI membranes compared to that of Nafion 212. Mechanical Nafion 212 Nafion 212 HMT-PMBI HMT-PMBI HMT-PMBI Propertyb,c (d)a (w)a (I-, d) (Cl-, d) (Cl-, w) Tensile 23.9 19.4 64.7 (± 0.3) 50 (± 2) 33 (± 3) Strength (MPa) Elongation at 136 119 97 (± 13) 76 (± 10) 63 (± 5) Break (%) Young’s 270 200 1070 (± 160) 940 (± 40) 230 (± 30) modulus (MPa) aLiterature data210 for a membrane. bd = dry form; w = fully hydrated form. cMechanical properties for HMT-PMBI membranes (89% dm) were measured three times and the standard deviations are shown. The chloride form was produced by exchanging the as-cast iodide membranes in 1 M NaCl.

3.3.6. Ex Situ Stability to Hydroxide Ions

The model compound, 2-mesityl-1,3-dimethyl-1H-benzimidazolium (MeB)

(Scheme 3.3), was subjected to 4 M KOH/CH3OH/H2O at 80 °C for 7 days in order to observe the 1H NMR spectrum of the degradation product without the complicated side- products from deuterium exchange. The precipitated, dark red-coloured degradation product was collected and analyzed by 1H NMR spectroscopy, as shown in Figure 3.8.

76

The main degradation pathways for imidazoliums reported in literature are nucleophilic displacement (Scheme 3.3a)86 and ring-opening degradation (Scheme 3.3b)133 followed by hydrolysis (Scheme 3.3c). The spectrum of the degraded product suggests that the only degradation route is through ring-opening degradation (which will be further verified in Chapter 5). There appears to be more than one isomer, which results in multiple resonances within a given area. For example, the N–H resonances appear between 5.4– 4.4 ppm but only one quartet resonance was expected. Two quartets are instead observed, suggesting that conjugation through the amide bond locks rotation on the NMR time scale, observing trans- and cis-like compounds simultaneously. The conjugation is also the likely reason for the dark red colour of this product.

Scheme 3.3. Possible degradation pathways for the model compound MeB in hydroxide solution: (a) nucleophilic displacement, (b) ring-opening/C2- hydroxide attack, followed by (c) hydrolysis of the amide.

77

1 Figure 3.8. Selected regions of the solution H NMR spectrum (500 MHz, DMSO-d6) of the degraded model compound MeB after exposing to 4 M KOH/CH3OH/H2O at 80 °C for 7 days. The chemical structure represents one of the suggested isomers that are formed.

The hydroxide stability of 89% dm HMT-PMBI was examined under high temperature and high pH conditions in order to determine the upper limit of stability. Membranes were first examined for stability in 2 M KOH at 20, 40, 60, and 80 °C for 7 days. The anion conductivity, as shown in Figure 3.9a, was stable up to 40 °C but decreased at 60 °C by 9% and at 80 °C by 19%. To determine if the resulting loss in conductivity was due to chemical degradation, the 1H NMR spectra of each sample was collected (Figure 3.10). For the spectra of membranes exposed to 2 M KOH at 60 °C and lower, no chemical change was observed. This agrees with previous findings in Chapter 2 wherein HMT-PMBI was subjected to a 2 M KOH, 60 °C test in methanol, and where no degradation was observed after 8 days. The 9% decrease in conductivity is therefore attributed to a morphological change in the membrane, similar to a conditioning process, and not to chemical degradation.

78

Figure 3.9. (a) Measured chloride ion conductivity and (b) relative percent of benzimidazolium remaining of 89% dm HMT-PMBI membranes after 7 days of soaking in 2 M KOH at various temperatures; (c) measured chloride ion conductivity and (d) relative percent of benzimidazolium remaining of 89% dm HMT-PMBI membranes after 7 days immersion in NaOH solutions of increasing concentration at 80 °C. Membranes were first reconverted to the chloride form for conductivity measurements and then the benzimidazolium remaining was determined from their 1H NMR spectra. The open diamonds refer to the initial samples.

79

Figure 3.10. The corresponding stacked solution 1H NMR spectra (400 MHz, DMSO•d6) of 89% dm HMT-PMBI in chloride form after membranes were soaked for 7 days in 2 M KOH at various temperatures (as labeled). The arrows show where new resonances are observed.

Exposure of the membrane to 2 M KOH at 80 °C for 7 days revealed minor changes in the 1H NMR spectrum of HMT-PMBI, i.e., at 7.2, 5.5–4.6, 3.2–2.6, and 2.4 ppm. Similar to the degradation of MeB, the resonances shift up-field to similar positions as those for the degraded model compound. In particular, two small resonances emerge in the 5.5–4.6 ppm region, representative of the characteristic N–H group formed by ring-opening degradation. By comparing the integration of the 6.00– 4.35 ppm region relative to 12.00H corresponding to the aromatic region, the extent of chemical degradation was quantified using Equation 8 (Section 3.2.11). In the event that 100% ring opening degradation of the polymer occurred, the 6.00–4.35 ppm should integrate to 2.00H. As such, the remaining quantity of benzimidazolium relative to the initial spectrum can be plotted, as shown in Figure 3.9b.

The relative amount of benzimidazolium remaining is unchanged for the 60 °C test, which quantitatively verifies that there is no chemical degradation within this 7 day time period. At 80 °C, the amount of benzimidazolium remaining decreases from 100% to 94%. While this 6% degradation may be approaching the numeric uncertainty in this method, it appears to be consistent with the qualitative changes observed in the NMR spectra. However, it is unclear whether the 19% decrease in conductivity is solely

80

related to the minor chemical degradation or if it is a combination of chemical degradation and conditioning.

In a modified degradation study, 89% dm HMT-PMBI was immersed in solutions of increasing NaOH concentration at 80 °C for 7 days. The resulting measured conductivities, after reconverting to the chloride form, as well as the percent benzimidazolium remaining (calculated based on 1H NMR spectra in Figure 3.11) are shown in Figure 3.9c and Figure 3.9d, respectively. The conductivity of the membranes exposed to 0.5 M and 1.0 M NaOH at 80 °C did not significantly change over the 7 day period, demonstrating its stability against hydroxide ion attack. However, immersion into solutions above 1 M NaOH results in a decrease in conductivity. Analysis of the 1H NMR spectra indicated significant degradation is observable (Figure 3.9d), where the amount of remaining benzimidazolium reaches a plateau of 40% for the 5.0 and 6.0 M NaOH treatments. This may due to the inability of hydroxide to permeate any further into the increasingly hydrophobic membrane which is induced after degradation. Similar to the prior degradation experiment, 94% benzimidazolium remained for the 2 M NaOH treatment, suggesting that there is no significant difference between 2 M NaOH and 2 M KOH degradation tests at 80 °C for 7 days.

Figure 3.11. The corresponding stacked solution 1H NMR spectra (400 MHz, DMSO•d6) of 89% dm HMT-PMBI in chloride form after membranes were soaked for 7 days in various concentrations of NaOH (as labeled) at 80 °C.

81

Over the 7 days at 80 °C, all membranes subjected to NaOH solutions (0.5 M to 6.0 M) remained intact and flexible. However, the initially yellow-coloured membrane became darker in colour commensurate with the NaOH concentration (Figure 3.12). The retention of the physical properties of the membrane suggests ring-opening degradation does not result in extensive backbone cleavage, which would occur if amide hydrolysis continued, as is observed with methylated m-PBI.141 It can be presumed that the degradation is retarded after ring-opening degradation. The red-shifted colour, as was observed with the fully ring-opened model compound, appears to follow the same trend as the percent of remaining benzimidazolium (Figure 3.9d), suggesting that this ring- opening degradation is the major, and possibly only, degradation pathway occurring.

Figure 3.12. Photograph of the dissolved HMT-PMBI membranes in DMSO-d6 after 7 days in various NaOH concentrations (0.0 – 6.0 M) at 80 °C.

When the membrane was subjected to 6 M NaOH at room temperature for 7 days, no chemical degradation was apparent (Figure 3.13). By using Equation 8, the amount of benzimidazolium remaining was calculated to be 98%, which quantitatively implies no significant degradation. Similar to the previously mentioned conditioned process, the membrane that was treated in 6 M NaOH had an increased anionic conductivity from 10.1 ± 0.4 mS cm-1 initially to 12.0 ± 0.4 mS cm-1. This suggests that the material is inert in 6 M NaOH at room temperature, representative of exceptional ex situ hydroxide resilience.

82

1 Figure 3.13. Stacked solution H NMR spectra (400 MHz, DMSO-d6) of 89% dm HMT•PMBI after 7 days of membrane immersion in 6 M NaOH at room temperature, after being exchanged to chloride form.

3.3.7. In Situ Fuel Cell Operation

The first method that was used to measure in situ stability was by operating a fuel cell with 89% dm HMT-PMBI as both the membrane and ionomer (Benjamin Britton performed these experiments). As the ionomer is in close proximity to the catalysts, a long lifetime would represent high chemical stability of the ionomer.206 Additionally, a stable fuel cell would also demonstrate that the membrane has sufficient mechanical stability, as well as good permeability properties, such as low gas cross-over and electrical conductivity.

As HMT-PMBI was prepared in iodide form, it was initially unclear whether trace amounts of iodide counter-ion would interact with the platinum catalysts during operation. As literature reports have shown that chloride has no significant impact on fuel cell performance,211 89% dm HMT-PMBI membranes were first exchanged to chloride form by soaking in KCl solution and washing with water (see Section 3.2.12 for details). Chloride-form HMT-PMBI was then dissolved in methanol and spray-coated with platinum on carbon (Pt/C) onto both sides of a chloride-form HMT-PMBI membrane to prepare the catalyst-coated membrane (CCM). The chloride-form CCM was then soaked in 1 M KOH for 7 days to exchange to hydroxide form and then washed in water

83

for 7 days to remove residual salts and impurities from the membrane as well as from the catalyst layers, potentially improving the lifetime.87,212

The HMT-PMBI CCM was conditioned by operating it in an AEMFC at low potentials for 8 h, at which point the potential and resistance reached a steady-state.

The AEMFC was then operated at 60 °C, with humidified (100% RH) H2 gas supplied to the anode side and humidified (100% RH) O2 gas supplied to the cathode side (denoted as H2/O2 at 100% RH). At certain points in time, polarization curves were taken and the measured potential at various current densities is shown in Figure 3.14.

Figure 3.14. Measured applied potentials over time for an AEMFC incorporating HMT•PMBI membrane and ionomer, operated at 60 °C, with humidified H2 being supplied to the anode, at the current density shown. At 60 h, the AEMFC was shut-down, left idle for 5 days at room temperature, and restarted back to 60 °C. Between 70–91 h, the cathode was run using air (CO2-containing); otherwise, it was operated with humidified O2.

The AEMFC was not only stable for 100 h of operation, but was also stable to a cold restart (i.e., shutdown, cool-down for 5 days, and re-equilibration to full function) at the 60 h mark. Additionally, the AEMFC was also stable to prolonged CO2 exposure (70- 91 h mark), where it returned to its original potential thereafter, as more clearly shown in Figure 3.15a. From the polarization data shown in Figure 3.15b for the AEMFC after 0, 51, and 94 h of operation, the power densities were calculated to be 47.7, 48.9, and 49.2 mW cm-2, respectively. This improvement in power density over time for an AEMFC at 60 °C is, to the best of our knowledge, the first of its kind (see Section 3.2.15 for error analysis). This demonstrates that HMT-PMBI, as both a membrane and ionomer, has excellent alkaline stability for operation of an AEMFC at 60 °C.

84

Figure 3.15. AEMFC polarization and power density curves after various operational times for an HMT-PMBI-based device operated at 60 °C and with H2/O2 at 100% RH supplied to anode/cathode unless otherwise noted, where (a) shows the performance before, during, and after switching the cathode supply from O2 (51 h) to air (75 h) and then back to O2 (94 h), (b) shows the power density at 0, 51, and 94 h, and (c) shows the variable temperature performance after an initial 109 h of operation at 60 °C. The CCMs were pre-conditioned from the chloride form by soaking in 1 M KOH for 7 days followed by 7 days in water.

85

After operating the fuel cell at 60 °C for 109 h, the temperature was increased up to 90 °C and polarization data were recorded. For 5 °C temperature increments from 70- -2 90 °C, Pmax values were 54.5, 55.9, 57.4, 64.4, and 62.5 mW cm , respectively (Figure 3.15c). The reduction in peak power density between the 85 °C and 90 °C data is attributed to mass transport losses. The overall peak power density increased 31% between 60 and 85 °C. Improvements in the high-temperature operation were consistent over multiple fuel cell tests, which is noteworthy for membrane-based alkaline AEMFCs.

To compare an HMT-PMBI-based fuel cell with that of a commercial-type AEMFC, CCMs were similarly prepared using FuMATech FAA-3 as membrane and ionomer (which has an undisclosed structure). When the cell containing this commercial membrane and ionomer was subjected to the same pre-conditioning as HMT-PMBI (1 M KOH for 7 days followed by 7 days in water), the AEMFC based on commercial materials was non-operational due to degradation. However, if the commercial fuel cell was first conditioned using a 1 M KOH soak for 24 h and used without a water wash, the -2 cell was fully operational, reaching a Pmax of 430 mW cm , as shown in Figure 3.16, which agrees well with literature.213 However, the cells containing the commercial membrane and ionomer could not be subjected to the shutdown, restart, and exposure to CO2-containing air, nor operated at higher temperatures without rapid degradation. When HMT-PMBI fuel cells were also conditioned using the latter method (by soaking -2 the iodide form CCM in 1 M KOH for 24 h), they yielded a similar Pmax of 370 mW cm .

Figure 3.16. AEMFC performance of FAA-3 and HMT-PMBI devices operated under zero backpressure at 60 °C and with H2/O2 at 100% RH. Both CCMs were pre-conditioned by soaking in 1 M KOH for 24 h and then operated for 45 min before measurement.

86

3.3.8. In Situ Water Electrolysis Operation

The second method used to verify the stability of HMT-PMBI in situ was by incorporating it as a CCM (membrane and ionomer) into a water electrolyzer, operated with 1 M KOH electrolyte at 60 °C (Benjamin Britton also performed these experiments). Similar to the AEMFC, this water electrolyzer rigorously tests not only chemical stability but also mechanical stability,206 due to the switching differential pressures as a result of bubble formation. The voltage at a given current density was monitored over time for both the commercial (FAA-3) and HMT•PMBI based water electrolyzer cells, as shown in Figure 3.17.

Figure 3.17. Measured potential over time for a water electrolysis test of FAA-3 (20 mA cm-2) and HMT-PMBI (25 mA cm-2) devices, where the flowing electrolyte was 1 M KOH at 60 °C for up to 195 h, at which point the still- functional electrolyzer was shut down. At 144 h, the current was stopped, the cell was allowed to re-condition with the same electrolyte and temperature, and then restarted.

Under similar experimental conditions, the cell based on commercial materials became inoperable after 9.5 h at 20 mA cm-2. During the 9.5 h, the average potential was 2.16 ± 0.04 V. End-of-life was most often represented by the absence of gas evolution and a substantial drop in potential (below 1.23 V), which we believe resulted from membrane degradation and electrical shorting.206 The relatively short lifetime of the commercial material under these conditions was reproducible, and repeated on individual cells, 11 times. The average operational time of three cells prepared using FAA-3 membranes and ionomer was 16.2 h.

87

In contrast, water electrolyzers fabricated from HMT-PMBI were operable at 25 mA cm-2 for >144 h, with an overall applied potential of 2.4 ± 0.1 V (this voltage is a typical applied potential for AEM electrolysis).112,113 HMT-PMBI cells demonstrated improved performance compared to the commercial cell. Operation of the cell was stopped after 144 h for evaluation, during which electrolyte flows and temperature were maintained. In the example shown, the cell was stopped for 50 h before restarting, and electrolysis continued for an additional 51 h at a potential 2.4 ± 0.1 V, whereupon the still-operational cell was shut down. The total time in situ was 245 h, representing a minimum of >20 times longevity versus our benchmark commercial membrane. The observed re-conditioning between the two periods of operation suggests that the catalyst was subject to poisoning from feed water impurities rather than material degradation, as trace impurities have been shown to strongly affect catalysis.214,215

3.4. Conclusion

The feasibility of scaled-up preparation of HMT-PMBI was demonstrated through the synthesis of 617 g of high purity polymer in 42 ± 3% overall yield. Each step was synthetically improved and shown to be highly reproducible; a synthetic step was eliminated. The HMT-PMBI dimethylated to 89% and cast as membranes were exceptionally strong and flexible when dry, as demonstrated through the measured tensile strength, elongation at break, and Young’s modulus. In their fully hydrated chloride form, the mechanical properties were superior to commercial Nafion 212 membrane, which is remarkable in the context of AEMs, given that membranes were cast from solvents which contained no additives or crosslinking agents. A wide range of properties, including dimensional swelling, water uptake, and conductivity for various ion forms is presented. The activation energy for the ionic conductivity of HMT-PMBI in mixed hydroxide/bicarbonate form in air was constant above 60% RH but increased when the RH was reduced to <60% RH.

Ex situ stability tests demonstrated the exceptional stability of HMT-PMBI under various hydroxide concentrations and temperatures; for example, no significant degradation in 1 M NaOH at 80 °C or 6 M NaOH at room temperature was observed after 168 h. HMT-PMBI was examined for in situ stability as the membrane and ionomer

88

in an AEMFC and water electrolyzer. For the AEMFC at 60 °C, the polymer demonstrated >100 h of operation at various current densities, which improved during operation, despite modulating the cathode feed between pure O2 and CO2-containing air and the first reported fully-restored restart of an AEMFC. AEMFCs based on the material achieved high power densities of 370 mW cm-2, comparable to commercial AEMFCs. However, HMT-PMBI cells demonstrated increased material stability, resulting in substantially more stable operation and longer lifetimes. For example, in water electrolyzers, an HMT-PMBI-based cell, using 1 M KOH electrolyte at 60 °C, was operated for 195 h without any drop in performance, whereas comparable cells based on the commercial materials became inoperable after only 16 h.

Collectively, the in situ and ex situ stability of HMT-PMBI, together with its ease of synthesis, mechanical properties, solubility in selective solvents, and insolubility in water make it a benchmark alkaline anion-exchange membrane and ionomer, providing motivation for further study in a wide range of energy applications, e.g., redox flow and metal–air batteries. Additionally, access to a versatile and useable hydroxide-conducting polymer will facilitate further research into AEMFCs and water electrolyzers, including the investigation of novel catalysts, effects of CO2, and impact of free radical formation on the lifetime of AEMFCs and electrolyzers.

89

Chapter 4.

Attempted Synthesis of Triphenylbenzene- Poly(benzimidazolium)

Andrew Wright performed all of the experimental work in this chapter.

4.1. Introduction

The work in Chapter 3 demonstrated the chemically-resilient, cationic polymer HMT•PMBI. In ex situ degradation experiments, HMT-PMBI showed no observable degradation in 1 M KOH at 80 °C and at 6 M KOH at ~25 °C. However, the polymer began noticeably degrading at 80 °C with concentrations of hydroxide ≥2 M. This suggests that while the stability of poly(benzimidazolium)s has been substantially increased, further improvements should be possible. From the small molecule degradation test, the main degradation product was the ring-opened benzimidazolium. This suggests that the methyl-protecting groups (on the mesitylene) are not effective enough for negating hydroxide C2 attack. Therefore, if the methyl group is changed to a bulkier and more sterically hindering group, the stability should be further improved.

There are a number of functional groups that are bulkier than methyl groups, such as ethyl/propyl/butyl or methoxy. Methoxy would also provide an electron-donating group, which would improve the stability of the C2 carbon.143,146 However, introducing heteroatoms near the cation would likely promote hydrolysis of the methoxy group.163,164 While longer alkyl chains could provide greater steric hindrance on average, they are also flexible. The flexibility of the linear alkanes may result in temporary rotation away from the C2 position, intermittently opening the C2 position for hydroxide attack. Instead, a more ideal functional group would be phenyl, as it would provide greater hindrance than that of a methyl group216 while also being significantly more rigid than alkyl groups.

90

For comparison to the previous PBI derivatives, the ideal polymer structure would be that of TRIP-PDMBI, as shown in Figure 4.1. This polymer would be directly comparable to mes-PDMBI, in which mesitylene was replaced with 2,4,6•triphenylbenzene. Each C2 position would be ortho to two phenyl groups and the theoretical hydroxide IEC would be 3.18 meq g-1, similar to that of HMT-PDMBI. Once prepared, stability of TRIP-PDMBI could be compared to that of mes-PDMBI, allowing for a direct comparison of the steric hindering power of phenyl groups compared to methyl groups. Several synthetic methods will be discussed in an attempt to prepare such a material.

Figure 4.1. Chemical structure of TRIP-PDMBI in hydroxide form.

4.2. Experimental

4.2.1. Materials

All chemicals were obtained from Sigma-Aldrich, Alfa Aesar, Strem Chemicals Inc., BDH Chemicals, Fisher Chemical, Combi-Blocks, or ACP Chemicals and were reagent grade unless otherwise stated. Potassium hydroxide was purchased from Macron Fine Chemicals. Deuterated solvents were purchased from Cambridge Isotope Laboratories, Inc.. 3,3’-diaminobenzidine was purified as discussed in Chapter 3. 2,6- dibromobenzoic acid (8) was prepared according to literature procedure.217 Deionized water (DI water) was purified using a Millipore Gradient Milli-Q® water purification system. 1H NMR and 13C NMR spectra were obtained on either a 400 MHz or 500 MHz Bruker AVANCE III running IconNMR under TopSpin 2.1 instruments. Electrospray ionization mass spectrometry (ESI-MS) was performed using a Bruker MicrOTOF in positive-mode.

91

4.2.2. Synthesis

Preparation of dimethyl 5-amino-2,4,6-tribromoisophthalate (1)

Dimethyl 5-aminoisophthalate (11.55 g, 55.2 mmol) was dissolved in a mixture of 330 mL deionized water and 110 mL concentrated hydrochloric acid. Bromine (9.3 mL, 180 mmol) was added drop-wise over 1 h with an addition of acetone (10 mL) halfway through. After stirring for an additional hour at room temperature, the precipitate was filtered, washed with water, and dried, resulting in 1 (21.15 g, 86%) as an off-white 1 13 powder. H NMR (500 MHz, DMSO•d6, ppm) δ: 6.12 (s, 2H), 3.90 (s, 6H). C NMR

(125 MHz, DMSO•d6, ppm) δ: 165.62, 143.80, 137.42, 104.59, 98.69, 53.31. ESI-MS m/z + + 1 calcd for C10H9Br3NO4 [M +H]: 445.806, found 445.805. The H NMR spectrum of 1 can be found in Appendix C, as Figure C1.

Preparation of dimethyl 5-amino-2,4,6-triphenylisophthalate (2)

In a 500 mL round-bottom flask, 1 (9.6768 g, 21.7 mmol), phenylboronic acid

(11.8911 g, 97.5 mmol), ethanol (44 mL), benzene (220 mL), and 20.98 g Na2CO3·H2O in 88 mL deionized water were mixed and degassed using a freeze, pump, thaw method twice. Tetrakis(triphenylphosphine)palladium(0) (0.260 g, 0.225 mmol) was added and the mixture was refluxed for 72 h under argon. The organic layer was collected, dried over magnesium sulfate, and solvent removed under vacuum. The solid was dissolved in minimal chloroform, passed through a silica column to remove catalyst, and dried to a

92

yellow solid. This solid was recrystallized in boiling acetone, vacuum filtered, and dried 1 to yield 2 (7.18 g, 76%) as a white flaky solid. H NMR (500 MHz, DMSO•d6, ppm) δ: 7.51 – 7.45 (m, 4H), 7.44 – 7.38 (m, 2H), 7.36 – 7.26 (m, 7H), 7.21 – 7.16 (m, 2H), 4.12 13 (s, 2H), 3.07 (s, 6H). C NMR (100 MHz, DMSO•d6, ppm) δ: 167.91, 141.82, 137.60, 135.65, 134.53, 129.44, 128.93, 128.84, 128.05, 127.81, 127.31, 124.44, 124.08, 51.23. + + 1 ESI-MS m/z calcd for C28H24NO4 [M +H]: 438.170, found 438.173. The H NMR spectrum of 2 is shown in Figure C2.

Preparation of dimethyl 2,4,6-triphenylisophthalate (3)

This was procedure was adapted from a generalized literature procedure.218 In a 100 mL round-bottom flask, 2 (8.00 g, 18.3 mmol), 32 mL acetic acid, and 3.65 g concentrated sulfuric acid were mixed. A solution of sodium nitrite (1.92 g, 27.8 mmol) dissolved 16 mL deionized water was added to the mixture drop-wise over 15 min and then stirred for an additional 15 min at room temperature. This solution was poured into a solution of FeSO4•7 H2O in 320 mL N,N-dimethylformamide. After several minutes, this solution was poured into 1200 mL of deionized water, filtered, washed with water, and 1 dried, resulting in 3 (7.21 g, 93%) as an off-white powder. H NMR (500 MHz, DMSO•d6, ppm) δ: 7.54 – 7.35 (m, 14H), 7.23 (d, J = 6.2 Hz, 2H), 3.24 (s, 6H). 13C NMR (125 MHz,

DMSO•d6, ppm) δ: 168.01, 140.08, 138.80, 137.67, 136.94, 132.60, 130.15, 128.64, + + 128.53, 128.19, 128.11, 127.95, 51.76. ESI-MS m/z calcd for C28H23O4 [M +H]: 423.159, found 423.159. The 1H NMR spectrum of 3 is shown in Figure C3.

93

Preparation of 2,4,6-triphenylisophthalic acid (4)

In a 500 mL round-bottom flask, 3 (7.00 g, 16.6 mmol) was refluxed for 96 h in a solution of 200 mL ethanol, 150 mL water, and 14.82 g of potassium hydroxide (264 mmol). The solution was cooled, filtered, and the filtrate acidified with concentrated hydrochloric acid to yield precipitate. The precipitate was washed with 2 M hydrochloric acid and refluxed in a solution of 10.5 g potassium hydroxide in 150 mL water overnight. The mixture was cooled, filtered, and acidified with concentrated hydrochloric acid to yield 4 (1.45 g, 22%) as a light-pink powder. Recrystallization in acetic acid resulted in 1 pure white crystals. H NMR (500 MHz, DMSO•d6, ppm) δ: 12.84 (s, 2H), 7.54 (d, J = 7.0 Hz, 4H), 7.45 (t, J = 7.3 Hz, 4H), 7.43 – 7.36 (m, 6H), 7.36 – 7.32 (m, 2H). 13C NMR

(125 MHz, DMSO•d6, ppm) δ: 169.08, 139.36, 138.38, 137.34, 136.03, 134.28, 130.11, 129.41, 128.54, 128.46, 128.43, 127.89, 127.76, 127.61. ESI-MS m/z calcd for + + 1 C26H19O4 [M +H]: 395.128, found 395.136. The H NMR spectrum of 4 is shown in Figure C4.

Preparation of 11-phenylindeno[2,1-b]fluorene-10,12-dione (5) and 6- phenylindeno[1,2-a]fluorene-7,12-dione (6)

(5) (6)

In a 25 mL round-bottom flask was added 3 (0.25 g, 0.63 mmol) followed by 2.0 mL of concentrated sulfuric acid, resulting in immediate dissolution and a colour

94

change to dark red. After 10 min of stirring at room temperature in air, the mixture was transferred into 125 mL of stirring water, resulting in yellow precipitate. The solid was filtered and dried under vacuum at 60 °C, resulting in 0.20 g of bright yellow solid. The solid was then recrystallized in ethanol to yield a mixture of 5 and 6 (0.06 g, 27%) as 1 fluffy yellow solid. H NMR (500 MHz, DMSO•d6, ppm) δ: 8.85 (d, J = 7.6 Hz, 0.59H), 8.33 (s, 0.37H), 7.93 (dd, J = 13.9, 7.4 Hz, 1.35H), 7.74 – 7.27 (m, 11.60H). ESI-MS m/z + + 1 calcd for C26H15O2 [M +H]: 359.107, found 359.107. The H NMR spectrum of 5 and 6 is shown in Figure 4.4.

Preparation of dimethyl 2'-amino-2,2''-dimethyl-m-terphenyl-4',6'-dicarboxylate (7)

In a 50 mL round-bottom flask, 1 (0.9667 g, 2.2 mmol), o-tolylboronic acid

(1.0330 g, 7.6 mmol), ethanol (4.4 mL), benzene (22 mL), and 2.12 g Na2CO3·H2O in 8.8 mL deionized water were mixed and degassed using a freeze, pump, thaw method. Tetrakis(triphenylphosphine)palladium(0) (0.030 g, 1 mol% per 1) was added and the mixture was refluxed for 96 h under argon. The organic layer was collected, dried over magnesium sulfate, and solvent removed under vacuum. The resulting brown oil was dissolved in 10 mL of hexane and 4 mL of methanol and then cooled in the fridge, allowing the two solvents to phase separate. The hexane layer was removed and 10 mL of ethanol was added to the methanol solution and left to evaporate in air to half of its 1 volume, resulting in crystals of 7. H NMR (500 MHz, DMSO•d6, ppm) δ 7.55 (d, J = 2.9 Hz, 1H), 7.35 – 7.23 (m, 6H), 7.04 – 6.98 (m, 2H), 3.75 (d, J = 6.1 Hz, 2H), 3.50 (d, J = 3.2 Hz, 6H), 2.04 (d, J = 3.1 Hz, 6H). The 1H NMR spectrum of 7 is shown in Figure 4.5.

95

Preparation of methyl 2,6-dibromobenzoate (9)

Powdered potassium hydroxide (1.13 g, 20.1 mmol) was stirred in dimethylsulfoxide (35 mL) for 1 h in a 100 mL round-bottom flask. 2,6-dibromobenzoic acid (8, 5.00 g, 17.9 mmol) was separately dissolved in dimethylsulfoxide (35 mL) and added to the basic mixture. After stirring for 30 min at room temperature, iodomethane (1.50 mL, 24.1 mmol) was added and stirred for 1 h while capped at room temperature. The yellow solution was then poured into stirring ice-water (1 L), resulting in fine white precipitate. The solid was collected by filtration, washed with water, and dried under 1 vacuum to yield 9 (4.39 g, 83%) as fluffy white powder. H NMR (500 MHz, DMSO•d6, ppm) δ: 7.75 (d, J = 8.1 Hz, 2H), 7.36 (t, J = 8.1 Hz, 1H), 3.91 (s, 3H). 13C NMR

(125 MHz, DMSO•d6, ppm) δ: 165.99, 136.88, 132.75, 131.74, 118.88, 53.14. The 1H NMR spectrum of 9 is shown in Figure C5.

Preparation of methyl 2,6-dimesitylbenzoate (10)

Into a 25 mL Schlenk tube was added 9 (1.0009 g, 3.41 mmol), 2,4,6•trimethylbenzeneboronic acid (2.7890 g, 17.0 mmol), tripotassium phosphate

(5.0537 g, 23.8 mmol), tris(dibenzylideneacetone)dipalladium(0) (Pd2(dba)3, 0.1562 g, 10 mol% Pd per 9), bis[(2-diphenylphosphino)phenyl] ether (DPEPhos, 0.2208 g, 12 mol% DPEPhos per 9), and molecular sieves (4 Å, 0.66 g). The tube was evacuated

96

and refilled with argon. Additional Pd2(dba)3 (0.0735 g) and DPEPhos (0.1002 g) were added, followed by toluene (18 mL). The tube was then capped with a Teflon cap and heated to 110 °C for 22 h. The brown mixture was poured into dichloromethane (150 mL) and filtered through a pad of silica, flushing with additional dichloromethane. The eluent was evaporated to a yellow solid. The solid was triturated in 25 mL methanol at room temperature to yield a white powder, which was collected by filtration and recrystallized in methanol-ethanol solution (40 mL). The crystals were collected and dried under vacuum at 80 °C to yield 10 (0.43 g, 34%) as white crystalline sheets. 1 H NMR (500 MHz, CD2Cl2, ppm) δ 7.54 (t, J = 7.6 Hz, 1H), 7.12 (d, J = 7.6 Hz, 2H), 13 6.90 (s, 4H), 3.11 (s, 3H), 2.31 (s, 6H), 2.02 (s, 12H). C NMR (125 MHz, CD2Cl2, ppm) δ: 169.55, 139.73, 137.43, 137.15, 136.84, 134.83, 130.32, 128.79, 128.16, 51.53, 21.39, 20.69. The 1H NMR spectrum of 10 is shown in Figure 4.6.

4.3. Results and Discussion

4.3.1. Monomer Synthesis

The diacid monomer 4 was prepared as shown in Scheme 4.1. The starting material, dimethyl 5-aminoisophthalate, was chosen due to its low cost and protected carboxylic acids. It was tribrominated in high yield to produce 1. Suzuki-Miyaura coupling of 1 with phenylboronic acid was catalyzed in good yield with Pd(PPh3)4 to produce 2. A one-pot diazotization and hydrodediazoniation of 2 yielded the sterically-hindered diester 3. The base hydrolysis of 3 to produce the diacid monomer 4 was performed in two steps, due to the low solubility of 3. Refluxing 3 in KOH/EtOH/H2O for 96 h, filtering off the insoluble solid, and acidifying the filtrate to precipitate carboxylic acids yielded a mixture of products, such as half mono-methyl and ethyl esters of 3. Further refluxing the mixture of products in aqueous KOH for >18 h and cooling to room temperature resulted in complete precipitation of the mono-ethyl ester, shown in Figure 4.2. Filtering the ester off and acidifying the filtrate resulted in precipitation of high purity diacid monomer 4.

97

Scheme 4.1. Synthetic route used to prepare monomer 4.

1 Figure 4.2. The H NMR spectrum (500 MHz, DMSO-d6) of a side product from the base hydrolysis of 3.

98

4.3.2. Condensation Reactions of Triphenylisophthalic Acid

In an attempt to synthesize the poly(benzimidazole) TRIP-PBI, as shown in Scheme 4.2, monomer 4 and DAB were heated from 140 to 200 °C in polyphosphoric acid. The solids were fully dissolved but pouring the solution into water produced black powder. Analysis using 1H NMR spectroscopy revealed no benzimidazole N-H resonance was present (which is typically observed at ~13 ppm). The powder was dissolved in DMSO and cast in a glass dish at 86 °C in an attempt to prepare a membrane. Instead, small obsidian shards were formed, indicative of very low molecular weight material. Repeating this method with diester 3 instead of 4 also did not result in any benzimidazole-containing polymer.

Scheme 4.2. Attempted synthetic route to TRIP-PBI.

To understand the reaction that was occurring, monomer 4 was instead heated with o-phenylenediamine, which should produce the small molecule analogue of TRIP•PBI, as shown in Scheme 4.3. However, as previously, a very dark-coloured solid was produced. The 1H NMR spectrum of the products is shown in Figure 4.3, which consists of a large number of aromatic resonances and no N-H resonance (expected at ~13 ppm).

99

Scheme 4.3. Attempted reaction of 4 with o-phenylenediamine in PPA.

1 Figure 4.3. H NMR spectrum (500 MHz, DMSO-d6) of the products from the reaction of 4 with o-phenylenediamine in PPA.

The aromatic resonances in the 1H NMR spectrum at ~8.9 ppm and the dark- colour of the products were indicative of highly conjugated products. In 1953, Taylor and Kalenda demonstrated that C-C bond formation could occur under such conditions, as phenanthridine could be produced by heating 2•formamidobiphenyl in PPA at 140 °C.219 The first step of this type of reaction has been proposed to be the condensation of PPA with an acyl group, thus activating the acyl position.220 Due to the ortho-substituted phenyl groups around the carboxylic acids on compound 4, it would be theoretically possible that the reaction occurring is an intramolecular condensation-cyclization, which would occur by two pathways proposed in Scheme 4.4. Pathway (1) is the intra- molecular cyclization to produce fluorenone products. Further cyclization with the second acyl group would produce an indeno[2,1•b]fluorenedione (5) and indeno[1,2-

100

a]fluorenedione (6). Compounds 5 and 6 could then also condense with o•phenylenediamine to produce a mixture of fluoren-9-imine compounds (of which only three of many are shown). Pathway (2) would initially involve amide formation with o•phenylenediamine followed by the cyclization reactions.

Scheme 4.4. Proposed pathways for the reaction of 4 with o-phenylenediamine in PPA.

The first synthesis of indeno[2,1-b]fluorene-10,12-dione (analogous to compound 5 without the phenyl substituent) was only recently reported by Romain et al. in 2015,221 which was prepared in a similar manner. This was done by heating 4,6•diphenylisophthalate to 120 °C in methanesulfonic acid, resulting in intramolecular bicyclization. This suggests that the likely products from compounds 3 or 4 should follow pathway (1). To test this hypothesis, compound 3, the methyl ester of compound 4, was dissolved in concentrated sulfuric acid at room temperature, stirred for 10 min, and then poured into water, resulting in yellow precipitate. After purification, the isolated product was analyzed by 1H NMR spectroscopy, as shown in Figure 4.4. The large number of aromatic resonances is similar to that of the previous tests in PPA and still appeared to be more than one product. Surprisingly, mass spectrometry revealed an m/z consistent with that of compounds 5 and 6 (which are constitutional isomers), indicating that bicyclization occurs not only at room temperature but is also very fast. As such, it would not be possible to prepare TRIP-PBI using diester 3 or diacid 4.

101

1 Figure 4.4. H NMR spectrum (500 MHz, DMSO-d6) of the isolated products from reacting compound 3 in sulfuric acid at room temperature for 10 min.

4.3.3. Ortho-Substituted Phenyl-Protecting Groups

Given that the cyclization of compounds 3 and 4 occurred due to the un- substituted ortho-positions of the ortho-phenyl groups, it was hypothesized that if the phenyl protecting groups were also ortho-substituted with methyl groups, intramolecular cyclization would not be possible. This would then allow time for the benzimidazole to properly form. As such, the synthesis of two monomers (2b, 2c) was attempted, as shown in Scheme 4.5.

102

Scheme 4.5. The synthetic route attempted to prepare monomers 2b and 2c.

Unfortunately, the Suzuki-Miyaura coupling of 1 using Pd(PPh3)4 to synthesize either monomer 2b or 2c was unsuccessful, most likely due to the very sterically- hindered positions of the bromo groups. Compound 2c should be easier to prepare than 2b due to having only one ortho methyl group on the phenyl protecting groups. As such, its synthesis was further investigated using more specialized catalyst-ligand systems, such as DPEPhos and SPhos (Scheme 4.5) with Pd2(dba)3, as they have been used in literature to successfully prepare sterically-hindered tetraortho-substituted biaryl compounds.222,223 Despite using those systems, compound 2c was not significantly observed. After purification and recrystallization of the crude organics from the reaction to synthesize 2c, compound 7 was isolated in high purity, as seen in its 1H NMR spectrum (Figure 4.5). The structure of 7 demonstrates that dehalogenation occurs at the 2-position, which is ortho-positioned to both carboxylates. Dehalogenation has been shown in literature to occur for palladium-catalyzed couplings and has often been attributed to the protic solvent used.224–228 However, as the reactions of 1 with DPEPhos and SPhos were performed without aqueous solvent in toluene and under air-free conditions (with and without molecular sieves), it is currently unknown why this occurs. Based on the 1H NMR data of 7, it is also worth noting that 7 possesses two atropisomers, as all expected singlet resonances appear as two singlets. One

103

atropisomer would represent a structure where both ortho-methyl groups are directed to the same side and the second atropisomer would have opposite facing methyl groups.

1 Figure 4.5. H NMR spectrum (500 MHz, DMSO-d6) of 7, isolated from the attempted synthesis of 2c.

As 2b and 2c could not be synthesized to confirm nor deny the previous hypothesis that ortho-substituents would prevent intramolecular cyclization, a new mono- carboxylate derivative was instead prepared, as shown in Scheme 4.6. Compound 9 was chosen to mimic 1 but have significantly less steric hindrance around the bromo groups. Suzuki-Miyaura coupling of 9 with mesitylboronic acid to produce 10 was unsuccessful when Pd(PPh3)4 was used, returning the starting material after the reaction. Instead, when Pd2(dba)3 was used with DPEPhos, 10 was successfully synthesized in high purity, as shown in Figure 4.6.

104

Scheme 4.6. Synthetic route to the mesityl-protected carboxylate ester 10.

1 Figure 4.6. H NMR spectrum (500 MHz, CD2Cl2) of 10.

Compound 10 was then reacted with o-phenylenediamine (1:1 mol ratio) in Eaton’s reagent at 140 °C. Pouring the red solution into water produced yellow precipitate. Unfortunately, the 1H NMR spectrum of the precipitate, shown in Figure 4.7, did not contain any N-H resonance at 13 ppm, signifying that no significant amount of benzimidazole was synthesized. No starting material was present and there were no resonances observed at ~9 ppm chemical shift, indicating that no imine-type by-products formed (which occurred previously due to intramolecular cyclization). The alkyl region contained approximately thirty different singlet resonances, whereas the starting material contained only three, suggesting a large number of compounds formed. The most likely

105

explanation is the formation of numerous amide compounds, which as discussed in Section 3.3.6, can have more than one isomer. This suggests that the first step in the condensation reaction is successful in synthesizing the amide. However, the second condensation to produce the benzimidazole fails, most likely due to the high activation barrier for the second amino group to position properly for nucleophilic attack of the amide carbonyl. When the yellow product mixture was placed in polyphosphoric acid and heated to higher temperatures (180 °C) and then poured into water, black solid was yielded with very broad 1H NMR spectroscopic resonances and no indication of benzimidazole formation. Unfortunately, due to the very bulky ortho-phenyl groups, the benzimidazoles could not be prepared by acid condensation reactions of carboxylate esters at high temperatures.

1 Figure 4.7. H NMR spectrum (500 MHz, DMSO-d6) of the precipitated product from the reaction of 10 with o-phenylenediamine in Eaton’s reagent at 140 °C, after washing with ethyl acetate (EtOAc).

4.4. Conclusion

A phenyl-protected diacid/diester was prepared in good yield and high purity, which is structurally similar to the mesitylene diacid monomer used to prepare mes•PDMBI. Unfortunately, the subsequent polycondensation reaction was unsuccessful in producing the desired polymer TRIP-PBI. Instead, intramolecular

106

cyclization was the dominating reaction and could only be hindered by replacing the phenyl groups with even bulkier o-tolyl and mesityl groups. Attempts to prepare the benzimidazole from the condensation of an ester bearing two ortho-mesityl groups with o-phenylenediamine in Eaton’s reagent or polyphosphoric acid, however, were futile. The very bulky protecting groups now prevented the condensation/cyclization benzimidazole formation. Ultimately, the preparation of phenyl-protected benzimidazoles by condensation of ortho-aryl esters or acids does not appear possible.

107

Chapter 5.

Poly(phenylene)-Protected Benzimidazoliums

The work described in this chapter is reported in part in: Wright, A. G.; Weissbach, T.; Holdcroft, S. Angew. Chem. Int. Ed. 2016, 55 (15), 4818–4821.

Thomas Weissbach performed the DFT calculations. Andrew Wright, with the assistance of Dr. Jeffrey S. Ovens, performed the single crystal XRD measurements and refinements. Andrew Wright performed the remaining experimental work in this chapter.

This work was financially supported by the Natural Sciences and Engineering Research Council of Canada (NSERC). Computing resources were provided by West- Grid (www.westgrid.ca) and Compute Canada Calcul Canada (www.computecanada.ca).

5.1. Introduction

The work described in Chapter 4 found that the condensation reaction of ortho- phenyl acid or ester monomers with o-phenylenediamine was unsuccessful. This was due to either intramolecular cyclization of the o-aryl group with the acyl position to form fluorene derivatives or stopped at the initial amide condensation product if the o-aryl was o-mesityl. As poly(benzimidazole)s (PBI) are all prepared by condensation reactions of acid/ester monomers, it would not be possible to prepare a PBI derivative with phenyl- protecting groups. As such, if it was possible to prepare a phenyl-protected benzimidazole, it would likely require a different polymer backbone altogether and would not be directly comparable to mes-PDMBI or HMT-PMBI.

108

To be able to compare the methyl-protected benzimidazolium with a phenyl- protected benzimidazolium, it was desired to only prepare small molecule model compounds and subject them to identical, solution-based, degradation tests. Using the previous synthetic routes tested, a new method for preparing the phenyl-protected benzimidazole will be discussed. The key change in the synthesis involves formation of the benzimidazole first followed by functionalization with the o-phenyl groups. This prevents issues of fluorene formation and reduces the initial sterics during the critical benzimidazole formation step. Once prepared, several benzimidazolium model compounds are subjected to a newly designed, quantitative, alkaline degradation test. The compounds are then analyzed by single crystal x-ray diffraction (XRD) and density functional theory (DFT) calculations in order to find any correlation in hydroxide stability with steric properties, such as dihedral angle.

Using this new versatile synthetic method, several polymers are envisioned that would have highly resilient backbones with high IECs. One of the polymers is successfully prepared and a number of its properties are analyzed, demonstrating unprecedented stability for a benzimidazolium-containing polymer.

5.2. Experimental

5.2.1. Materials

All chemicals were obtained from Sigma Aldrich and were ACS reagent grade unless otherwise stated. Mesitoic acid (98%) and 1,3-dibromobenzene (98%) were purchased from Combi-Blocks. Phenylboronic acid (98+%) and anhydrous dimethylsulfoxide (99.8+%) were purchased from Alfa Aesar. 4-chlorophenylboronic acid (98%) was purchased from Ark Pharm, Inc.. Hydrochloric acid (37%, aq.), hexanes, potassium carbonate, diethyl ether, sodium chloride, and potassium chloride were purchased from ACP Chemicals Inc.. Potassium hydroxide was purchased from Macron Fine Chemicals. Methanol, ethyl acetate, acetone, and dichloromethane were purchased from Fisher Chemical. Tetrakis(triphenylphosphine)palladium(0) (99%) and bis(cyclooctadiene)nickel(0) (98+%) were purchased from Strem Chemicals Inc.. Sodium hydroxide, magnesium sulfate, dimethyl sulfoxide (DMSO), and chloroform were

109

purchased from BDH Chemicals. Ethanol was purchased from Commercial Alcohols. Basic aluminum oxide (Brockmann I, 50-200 μm, 60 Å) was purchased from Acros

Organics. Eaton’s reagent (7.7 wt% P2O5 in methanesulfonic acid) was prepared in advance by dissolution of P2O5 under argon atmosphere in methanesulfonic acid at

120 °C and then stored in glass at room temperature until needed. Dimethyl sulfoxide-d6

(D, 99.9%) and methanol-d4 (D, 99.8%, CD3OD) were purchased from Cambridge

Isotope Laboratories, Inc.. Sodium deuteroxide (30 wt% in D2O, 99 atom% D) and

1,1,2,2-tetrachloroethane-d2 (D, 99.5%, C2D2Cl4) were purchased from Sigma Aldrich. 1H NMR and 13C NMR spectra were obtained on a 500 MHz Bruker AVANCE III running

IconNMR under TopSpin 2.1 and the residual solvent resonances for DMSO•d6, CD3OD, 1 and C2D2Cl4 were set to 2.50 ppm, 3.31 ppm, and 5.36 ppm for their H NMR spectra, 13 respectively, and 39.52 ppm for the C NMR spectra in DMSO•d6. Deionized water (DI water) was used from a Millipore Gradient Milli-Q® water purification system at 18.2 MΩ cm. Electrospray ionization mass spectrometry (ESI-MS) was performed using a Bruker micrOTOF in positive-mode.

5.2.2. Synthesis

Preparation of 2,6-dibromobenzoic acid (8)

2,6-dibromobenzoic acid was synthesized according to literature procedure.217 More specifically, dry tetrahydrofuran (400 mL) in a 2-neck round-bottom flask under argon was cooled to 0 °C in an ice-water bath. n-Butyllithium (2.5 M in hexanes, 117 mL) was added followed by a slow addition of diisopropylamine (45 mL) at 0 °C. The mixture was stirred for 45 min and then cooled to -78 °C in a dry ice/acetone bath. While at this temperature, 1,3-dibromobenzene (50.0 g, 0.212 mol) was added drop-wise over 5 min and stirred for 1 h. Dry ice was then closed in a glass container and connected by syringe into the reaction mixture. After bubbling the mixture with the evolved gaseous

CO2 for 45 min, pieces of dry ice were added into the mixture. The mixture was allowed

110

to slowly warm up to room temperature. Aqueous sodium hydroxide (0.5 M) and ethyl acetate (400 mL) were added until all of the solid was dissolved. The aqueous layer was washed with ethyl acetate and then acidified with concentrated hydrochloric acid until approximately pH=1. The precipitate was dissolved in fresh ethyl acetate, washed with brine, dried over MgSO4, filtered, and the solvent was evaporated at 40 °C under vacuum. The oil was cooled to room temperature, resulting in crystallization. The solid was boiled in hexanes (2 L) for 1 h and then cooled to room temperature. The solid was collected, washed with hexanes, and dried under vacuum at 80 °C, resulting in 8 (39.3 g, 1 66%) as an off-white powder. H NMR (500 MHz, DMSO•d6, ppm) δ: 7.70 (d, J = 8.1 Hz, 13 2H), 7.29 (t, J = 8.1 Hz, 1H). C NMR (125 MHz, DMSO•d6, ppm) δ: 166.87, 138.67, 131.87, 131.65.

Preparation of 2-(2,6-dibromophenyl)-1H-benzimidazole (11)

In a 200 mL, 3-neck round-bottom flask with a CaCl2 drying tube, stopper, and argon inlet was added 8 (17.00 g, 60.7 mmol), o-phenylenediamine (6.57 g, 60.8 mmol), and Eaton’s reagent (136 mL). The mixture was heated to 120 °C under argon until fully dissolved. The mixture was then heated at 150 °C for 45 min. The mixture was poured into water (3.3 L) and neutralized to pH=7 using potassium hydroxide and potassium carbonate. The resulting precipitate was collected and washed with water. The solid was dried under vacuum at 90 °C to yield 11 (20.79 g, 97%) as an off-white powder. 1H NMR

(500 MHz, DMSO•d6, ppm) δ: 12.87 (s, 1H), 7.84 (d, J = 8.1 Hz, 2H), 7.69-7.56 (m, 2H), 13 7.44 (t, J = 8.1 Hz, 1H), 7.30–7.20 (m, 2H). C NMR (125 MHz, DMSO•d6, ppm) δ: + 149.71, 134.33, 132.95, 131.77, 124.51, 122.15. ESI-MS m/z calcd for C13H9Br2N2 [M++H]: 350.913, found 350.908.

111

Preparation of 2-(2,6-dibromophenyl)-1-methyl-1H-benzimidazole (12)

In a 500 mL round-bottom flask was added powdered potassium hydroxide (6.18 g, 110 mmol) followed by DMSO (120 mL) and was vigorously stirred at room temperature for 30 min. A solution of 11 (20.00 g, 56.8 mmol) in DMSO (120 mL) was then added to the basic DMSO solution and stirred closed for 45 min at room temperature. Iodomethane (3.9 mL, 62.7 mmol) was then added and stirred for 45 min at room temperature. The mixture was poured into water (2.0 L) containing potassium hydroxide (10.0 g). Diethyl ether (500 mL) was added and the mixture stirred until fully dissolved. The organics were collected by decantation. The process was repeated by using additional diethyl ether (2 x 150 mL) and the combined organics were washed with water, brine, and water, dried over MgSO4, filtered, and the solvent was evaporated. Drying under vacuum at room temperature resulted in 12 (18.20 g, 88%) as pale brown 1 flakes. H NMR (500 MHz, DMSO•d6, ppm) δ: 7.89 (d, J = 8.2 Hz, 2H), 7.69 (dd, J = 23.2, 7.8 Hz, 2H), 7.49 (t, J = 8.1 Hz, 1H), 7.32 (dt, J = 28.6, 7.7 Hz, 1H), 3.58 (s, 3H). 13 C NMR (125 MHz, DMSO•d6, ppm) δ: 150.80, 142.30, 134.88, 133.43, 132.64, 131.92, + 124.77, 122.73, 122.01, 119.52, 110.66, 29.89. ESI-MS m/z calcd for C14H11Br2N2 [M++H]: 364.928, found 364.924.

Direct preparation of 2-(2,6-dibromophenyl)-1,3-dimethyl-1H-benzimidazolium (BrB) iodide from 12

In a 50 mL round-bottom flask was added 12 (4.00 g, 11.4 mmol) followed by dichloromethane (20 mL). Once fully dissolved, iodomethane (3.4 mL, 54.6 mmol) was

112

added and the closed mixture was stirred 30 °C for 18 h. The solvent was evaporated at 44 °C by dynamic vacuum and diethyl ether (50 mL) was added. The solid was collected by vacuum filtration, washed with diethyl ether, and dried at 80 °C under vacuum, 1 yielding BrB (5.21 g, 94%) as off-white powder. H NMR (500 MHz, DMSO•d6, ppm) δ: 8.29–8.20 (m, 2H), 8.14 (d, J = 8.2 Hz, 2H), 7.92–7.84 (m, 2H), 7.79 (t, J = 8.2 Hz, 1H), 13 3.98 (s, 6H). C NMR (125 MHz, DMSO•d6, ppm) δ: 147.36, 136.96, 132.88, 131.14, + + 127.83, 124.64, 123.17, 114.16, 32.47. ESI-MS m/z calcd for C15H13Br2N2 [M ]: 378.944, found 378.945.

Direct preparation of 2-(2,6-dibromophenyl)-1,3-dimethyl-1H-benzimidazolium (BrB) iodide from 11

In a pressure tube was added 11 (1.41 g, 4.01 mmol), sodium hydroxide (0.16 g, 4.00 mmol), methanol (20 mL), and iodomethane (1.2 mL, 19.3 mmol). The tube was sealed and the mixture was heated to 110 °C overnight with stirring. After cooling to room temperature, the solvent was evaporated to a yellow solid. Acetone was added, heated, and solid collected. The filtrate was evaporated, the solid recrystallized in ethanol, re-precipitated in THF, collected, and dried under vacuum at 60 °C, resulting in BrB (0.55 g) as white powder. The previously collected solid from acetone was recrystallized in acetone, resulting in BrB (0.30 g) as off-white needles (combined yield of 0.85 g, 41%).

113

Preparation of 2-([m-terphenyl]-2'-yl)-1-methyl-1H-benzimidazole (13)

In a 500 mL round-bottom flask was added 12 (8.00 g, 21.9 mmol), benzeneboronic acid (8.00 g, 65.6 mmol), 1,4-dioxane (240 mL), and 2 M K2CO3 (aq.) (80 mL). The mixture was bubbled with argon for 15 min and then tetrakis(triphenylphosphine)palladium(0) (106 mg, 0.4% mol per 12) was added. The mixture was heated to 104 °C for 18 h and then poured into hot 33% ethanol (1.2 L, aq.). The resulting mixture was bubbled with air for 5 min until the solution became black and the mixture was cooled to room temperature while stirring. The resulting precipitate was collected and washed with water. The grey solid was dissolved in dichloromethane:ethyl acetate (1:1 vol.) and filtered through basic alumina by rinsing with the same solvent mixture. The filtrate was evaporated. The resulting solid was recrystallized from methanol and dried under vacuum at 80 °C to yield 13 (4.30 g, 55%) as colourless 1 crystals with a faint yellow tint. H NMR (500 MHz, DMSO•d6, ppm) δ: 7.75 (t, J = 7.7 Hz, 1H), 7.57 (d, J = 7.8 Hz, 2H), 7.52–7.44 (m, 1H), 7.32–7.26 (m, 1H), 7.19–7.05 (m, 12H), 13 3.13 (s, 3H). C NMR (125 MHz, DMSO•d6, ppm) δ: 151.29, 143.02, 142.15, 140.11, 134.51, 130.32, 129.26, 128.37, 127.98, 127.33, 127.06, 121.80, 121.39, 118.94, + + 109.95, 29.75. ESI-MS m/z calcd for C26H21N2 [M +H]: 361.170, found 361.172.

114

Preparation of 2-([m-terphenyl]-2'-yl)-1,3-dimethyl-1H-benzimidazolium (PhB) iodide

In a 50 mL round-bottom flask was added 13 (3.00 g, 8.32 mmol) and dichloromethane (25 mL). Once fully dissolved, iodomethane (2.6 mL, 41.8 mmol) was added and the mixture was stirred closed at 30 °C for 17 h. The mixture was evaporated at 40 °C by dynamic vacuum and the resulting solid was briefly stirred in diethyl ether (50 mL). The solid was collected by vacuum filtration, washed with diethyl ether, and dried under vacuum at 80 °C, yielding PhB (4.09 g, 98%) as an off-white powder. 1 H NMR (500 MHz, DMSO•d6, ppm) δ: 8.08 (t, J = 7.8 Hz, 1H), 7.92–7.87 (m, 2H), 7.85 (d, J = 7.8 Hz, 2H), 7.70–7.62 (m, 2H), 7.36–7.25 (m, 6H), 7.20–7.13 (m, 4H), 3.53 (s, 13 6H). C NMR (125 MHz, DMSO•d6, ppm) δ: 149.25, 144.07, 137.90, 134.09, 130.64, 130.25, 128.97, 128.60, 127.96, 127.20, 117.30, 113.41, 32.42. ESI-MS m/z calcd for + + C27H23N2 [M ]: 375.186, found 375.187.

Preparation of 2-(4,4''-dichloro-[m-terphenyl]-2'-yl)-1-methyl-1H-benzimidazole (14)

115

In a 1 L round-bottom flask was added 12 (14.64 g, 40.0 mmol), 4•chlorophenylboronic acid (24.80 g, 159 mmol), 1,4-dioxane (366 mL), and 2 M

K2CO3 (aq.) (132 mL). The mixture was bubbled with argon for 15 min and then tetrakis(triphenylphosphine)palladium(0) (0.46 g, 1% mol per 12) was added. The mixture was heated to 104 °C for 19 h. As the solution was cooling to room temperature, the mixture was bubbled with air for 15 min until the solution colour darkened. The organics were collected by addition of ethyl acetate (600 mL) and washed with water, brine, and water. After drying the organic phase over magnesium sulfate, the solution was evaporated at 55 °C by rotary evaporation to yield an orange-coloured oil. The crude mixture was purified by flash chromatography on basic alumina using 1:2 vol ethyl acetate:hexanes. The collected solid was then washed with hexanes (400 mL) to yield off-white solid. This solid was recrystallized once in ethanol/water and three times in ethyl acetate/hexanes. Drying under vacuum at 100 °C yielded 14 (3.61 g, 21%) as a 1 white powder. H NMR (500 MHz, DMSO•d6, ppm) δ: 7.77 (t, J = 7.8 Hz, 1H), 7.59 (d, J = 7.8 Hz, 2H), 7.51 (d, J = 8.1 Hz, 1H), 7.36 (d, J = 7.3 Hz, 1H), 7.22 (d, J = 8.5 Hz, 13 4H), 7.18–7.09 (m, 6H), 3.17 (s, 3H). C NMR (125 MHz, DMSO•d6, ppm) δ: 150.82, 142.14, 141.88, 138.88, 134.61, 132.17, 130.67, 130.29, 129.68, 128.16, 127.27, + + 122.17, 121.75, 119.14, 110.31, 29.95. ESI-MS m/z calcd for C26H19Cl2N2 [M +H]: 429.092, found 429.093.

Preparation of 2-(4,4''-dichloro-[m-terphenyl]-2'-yl)-1,3-dimethyl-1H- benzimidazolium (15) iodide

In a 25 mL round-bottom flask was added 14 (1.20 g, 2.80 mmol) followed by dichloromethane (5 mL). The mixture was stirred at room temperature until the solid fully

116

dissolved. Iodomethane (0.52 mL, 8.35 mmol) was added and the mixture was capped. The mixture was stirred at 30 °C for 67 h, resulting in some precipitate. The solvent was evaporated to a white solid and diethyl ether (100 mL) was added. After trituration at room temperature, the solid was collected and washed with diethyl ether. The solid was dried under vacuum at 80 °C to yield 15 (1.54 g, 96%) as an off-white powder. 1H NMR

(500 MHz, DMSO•d6, ppm) δ: 8.09 (t, J = 7.8 Hz, 1H), 7.93 (dd, J = 6.3, 3.1 Hz, 2H), 7.86 (d, J = 7.8 Hz, 2H), 7.69 (dd, J = 6.3, 3.1 Hz, 2H), 7.35 (d, J = 8.5 Hz, 4H), 7.20 (d, 13 J = 8.5 Hz, 4H), 3.58 (s, 6H). C NMR (125 MHz, DMSO•d6, ppm) δ: 148.53, 142.88, 136.61, 134.20, 133.58, 130.72, 130.67, 129.91, 129.03, 127.33, 117.12, 113.62, 32.57. The 1H NMR spectrum and 13C NMR spectrum of 15 are shown in Appendix D, as Figure D1 and D2, respectively.

Preparation of poly(4,4''-[2'-(1-methyl-1H-benzimidazol-2-yl)-m-terphenylene]) (PPB3)

To a 250 mL round-bottom flask was added 14 (1.9974 g, 4.65 mmol) and 2,2’•bipyridyl (1.7078 g, 10.9 mmol). The flask was capped with a septum. Using a needle through the septum, the flask was evacuated and refilled three times with argon. Bis(1,5-cyclooctadiene)nickel(0) (2.9414 g, 10.7 mmol) was then added by removing the septum and quickly recapping. The flask was evacuated and refilled three times with argon again. Anhydrous dimethylsulfoxide (130 mL) was then added and the mixture was heated at 80 °C while stirring for 19 h. The mixture was then poured into 1:1 vol

H2O:conc. HCl (1.0 L) and stirred for 30 min, causing the colour to change from black to white. The precipitate was collected by vacuum filtration over a glass frit and washed with water. The solid was stirred in a solution of potassium carbonate (10 g in 250 mL water) for 30 min. The solid was collected by vacuum filtration and washed with water followed by acetone. The solid was then stirred in acetone (200 mL) for 30 min. The

117

solid was collected by vacuum filtration, washed with acetone, and dried at 100 °C under 1 vacuum, yielding PPB3 (1.58 g, 95%) as a fluffy white solid. H NMR (500 MHz, C2D2Cl4, ppm) δ: 7.37–6.05 (m, 15H), 2.52 (s, 3H).

Preparation of poly(4,4''-[2'-(1,3-dimethyl-1H-benzimidazolium-2-yl)-m- terphenylene]) (PPMB) iodide

In a 50 mL round-bottom flask was added PPB3 (1.00 g) followed by 1-methyl-2- pyrrolidinone (25 mL). The mixture was stirred and heated for 2 h at 80 °C until the polymer was fully dissolved. The mixture was then cooled to room temperature. Iodomethane (1.7 mL) was added and the mixture was stirred closed with a glass stopper at room temperature for 15 h. The solution was then poured slowly into stirring ethanol (600 mL). The resulting fibrous precipitate was collected and washed with ethanol. The solid was dried under vacuum at 80 °C, resulting in PPMB (1.40 g, 100%) as a red, brittle, film-like solid. The 1H NMR spectrum was taken of the cast, water- washed, and 80 °C vacuum-dried membrane in its iodide form (see casting section for 1 method). H NMR (500 MHz, DMSO•d6, ppm) δ: 8.20–7.98 (m, 1H), 7.96–7.76 (m, 4H), 7.70–7.61 (m, 2H), 7.60–7.36 (m, 4H), 7.32–7.08 (m, 4H), 3.57 (s, 6H).

Preparation of 2-phenyl-1-methyl-1H-benzimidazole (16)

Powdered potassium hydroxide (2.24 g, 39.9 mmol) was added to a 250 mL round-bottom flask and vigorously stirred in DMSO (65 mL) for 30 min. A solution of 2•phenylbenzimidazole (4.11 g, 21.2 mmol) in DMSO (65 mL) was added to the basic

118

DMSO solution and the mixture stirred for 45 min closed at room temperature. Iodomethane (1.4 mL, 22.5 mmol) was then added and the mixture stirred for 45 min. The mixture was then poured into a stirring solution of water (1.0 L) containing potassium hydroxide (5.0 g). Diethyl ether (300 mL) was then added and stirred until both layers were transparent. The organic layer was decanted and the same process was repeated with additional diethyl ether (2 x 200 mL). The combined organics were washed with water, brine, water, dried over magnesium sulfate, filtered, and evaporated at 44 °C under dynamic vacuum to yield 16 (3.74 g, 85%) as a pale brown powder. 1 H NMR (500 MHz, DMSO•d6, ppm) δ: 7.86 (dd, J = 7.8, 1.7 Hz, 2H), 7.69 (d, J = 7.9 Hz, 13 1H), 7.64–7.52 (m, 4H), 7.33–7.22 (m, 2H), 3.88 (s, 3H). C NMR (125 MHz, DMSO•d6, ppm) δ: 152.98, 142.47, 136.57, 130.15, 129.60, 129.28, 128.63, 122.32, 121.90, 118.98, 110.53, 31.64.

Preparation of 2-phenyl-1,3-dimethyl-1H-benzimidazolium (HB) iodide

Dichloromethane (20 mL) was added to 16 (3.00 g, 14.4 mmol) in a 50 mL round- bottom flask and stirred until fully dissolved. Iodomethane (2.7 mL, 43.4 mmol) was added and the mixture was stirred at 30 °C closed for 17 h. The solvent was evaporated at 45 °C using dynamic vacuum and diethyl ether was added. The solid was collected by vacuum filtration, washed with diethyl ether, and dried under vacuum at 40 °C, yielding 1 HB (4.59 g, 91%) as an off-white powder. H NMR (500 MHz, DMSO•d6, ppm) δ: 8.15 (dd, J = 6.2, 3.1 Hz, 2H), 7.93 (d, J = 7.0 Hz, 2H), 7.88–7.73 (m, 5H), 3.91 (s, 6H). 13 C NMR (125 MHz, DMSO•d6, ppm) δ: 150.29, 132.91, 131.68, 130.76, 129.42, 126.61, + + 120.96, 113.39, 32.85. ESI-MS m/z calcd for C15H15N2 [M ]: 223.123, found 223.124.

119

Preparation of 2-mesityl-1H-benzimidazole (17)

In a 200 mL 3-neck round-bottom flask, attached with a CaCl2 drying tube, glass stopper, and argon inlet, was added mesitoic acid (13.34 g, 81.2 mmol), o•phenylenediamine (8.79 g, 81.3 mmol), and Eaton’s reagent (136 mL). Under argon flow, the mixture was heated to 120 °C for 15 min. The mixture was stirred for an additional 15 min at 140 °C and the mixture was then poured into distilled water (3.3 L). The mixture was neutralized to pH=7 by addition of potassium hydroxide and potassium carbonate. The resulting precipitate was collected by vacuum filtration, washed with water, and dried under vacuum at 90 °C, yielding 17 (18.93 g, 98.6%) as a white 1 powder. H NMR (500 MHz, DMSO•d6, ppm) δ: 12.51 (s, 1H), 7.61–7.51 (m, 2H), 7.23– 13 7.16 (m, 2H), 6.99 (s, 2H), 2.31 (s, 3H), 2.06 (s, 6H). C NMR (125 MHz, DMSO•d6, ppm) δ: 151.26, 138.36, 137.13, 128.89, 127.95, 121.50, 20.76, 19.70.

Preparation of 2-mesityl-1,3-dimethyl-1H-benzimidazolium (MeB) iodide

Powdered potassium hydroxide (2.24 g, 39.9 mmol) was added to a 250 mL round-bottom flask and vigorously stirred in DMSO (65 mL) for 30 min. A solution of 17 (5.00 g, 21.2 mmol) in DMSO (65 mL) was added to the basic DMSO solution and the mixture stirred for 45 min closed at room temperature. Iodomethane (1.4 mL, 22.5 mmol) was then added and the mixture stirred for 45 min. The mixture was then poured into a stirring solution of water (1.0 L) containing potassium hydroxide (5.0 g). Diethyl ether (300 mL) was then added and stirred until both layers were transparent. The organic layer was decanted and the same process was repeated with additional diethyl ether (2 x 150 mL). The combined organics were washed with water, brine, water, dried over

120

magnesium sulfate, filtered, and evaporated at 44 °C under dynamic vacuum to yield a viscous pale yellow oil. Dichloromethane (25 mL) was added to the oil and stirred until fully dissolved. Iodomethane (6.0 mL, 96.4 mmol) was added and the mixture was stirred at 30 °C closed for 18 h. The solvent was evaporated at 44 °C using dynamic vacuum and diethyl ether (150 mL) was added. The solid was collected by vacuum filtration, washed with diethyl ether, and dried under vacuum at 80 °C, yielding MeB (6.81 g, 82%) 1 as an off-white powder. H NMR (500 MHz, DMSO•d6, ppm) δ: 8.19-8.09 (m, 2H), 7.82- 7.75 (m, 2H), 7.26 (s, 2H), 3.83 (s, 6H), 2.40 (s, 3H), 2.05 (s, 6H). 13C NMR (125 MHz,

DMSO•d6, ppm) δ: 149.82, 143.01, 138.57, 131.59, 129.07, 126.74, 117.15, 113.84, + + 32.16, 20.95, 18.85. ESI-MS m/z calcd for C18H21N2 [M ]: 265.170, found 265.171.

5.2.3. Intrinsic Viscosity

Four separate solutions of PPB3 at various concentrations in NMP (2-5 mg mL-1) were first prepared by gently heating the mixtures until fully dissolved. The solutions were then filtered through 0.45 μm PTFE syringe filters. The viscosity (η) of each solution was then measured using a temperature-controlled (25.0 °C) RheoSense, Inc. μVisc viscometer equipped with a 0.2-100 cP sensor. The settings were set to “AUTO” except for the shear rate, which was set to 5000 s-1. Prior to each measurement, the solution was allowed to thermally equilibrate for 5 min. Four measurements were taken for each concentration and averaged. The relative viscosity (ηrel) and specific viscosity

(ηsp) were then calculated for the measured concentrations using Equation 10 and Equation 11, respectively.

(10)

(11)

A plot of the Huggins (ηsp/c) and Kraemer ([ln(ηrel)]/c) parameters versus concentration (c) can then be produced. The average of the two y-intercepts, calculated

121

from the extrapolation of the Huggins-Kraemer linear regressions, represents the intrinsic viscosity.

5.2.4. Membrane Preparation

PPMB in its iodide form (1.5 wt% in DMSO) was evenly spread in a flat Petri dish and dried at 86 °C in air for 24 h. DI water was added to the dish and the membrane was peeled off of the glass. The membrane was soaked in DI water for 24 h, 1 M NaCl for 48 h, and DI water for 24 h at room temperature. The membrane was then dried at 80 °C under vacuum. This chloride-exchanged membrane was then recast as 1.5 wt% PPMB in DMSO as previously described. This chloride-cast membrane was then peeled off the glass using DI water and soaked in DI water for at least 24 h before use in the subsequent experiments.

5.2.5. Water Uptake

Pieces of a PPMB membrane in chloride form were soaked in 1 M KOH for 48 h followed by multiple fresh exchanges of DI water over 48 h. After removing the surface water with a kimwipe, the wet (hydrated) mass was measured (Ww). The pieces were soaked in 1 M NaCl (with one fresh exchange in between) for 4 days followed by several DI water exchanges for 24 h. The membrane pieces were then dried under vacuum at

50 °C and weighed to yield the dry mass (Wd). The water uptake (Wu) was then calculated from the average of four samples and the standard deviation was used as the uncertainty using Equation 4, previously described in Chapter 2.

5.2.6. Electrochemical Impedance Spectroscopy

Membrane pieces of PPMB (chloride-form) were soaked in 1 M KOH for 48 h at room temperature followed by soaking in DI water with multiple fresh exchanges over an additional 48 h at room temperature under ambient atmosphere (CO2 containing). The ionic conductivity was then measured and calculated as previously described in Chapter 2 using Equation 3. The conductivity was measured for four fully hydrated membranes, four times each, at ambient temperature (~22 °C).

122

5.2.7. Degradation Tests of Model Compounds

Each model compound in its iodide form (HB, BrB, MeB, and PhB) was dissolved in 3 M NaOD/CD3OD/D2O (prepared by diluting 2.05 g of 30 wt% NaOD (in

D2O) with CD3OD to 5.0 mL) inside PTFE containers, such that the final concentration of each model compound in the solution was 0.02 M. Once fully dissolved, ~0.6 mL of the solution was removed and analyzed by 1H NMR spectroscopy (“0 h” spectrum). The tightly-closed PTFE containers were then heated in an oven at 80 °C and samples were removed at certain points in time for analysis of their 1H NMR spectra. The spectra were all baseline-corrected using the “Full Auto (Polynomial Fit)” function found in MestReNova 9.0.1.

The percent remaining of PhB and MeB over time (from the 1H NMR spectra) was calculated using Equation 12. This formula involves the integration of an aryl resonance that does not overlap with any other resonances, including any that would appear from degradation products, relative to the total aryl region, which includes all degradation aryl protons. For PhB and MeB, only two aryl protons are deuterium- exchanged under these degradation conditions, which are the 4- and 7-position protons of the benzimidazolium, and are completely exchanged for deuterium by 68 h. For example, the total number of aryl protons of MeB at 0 h is 6H but decreases to 4H for the 68-240 h spectra.

( ) ( ) (12)

where nt is the number of expected protons in the aromatic region (17 and 6 for PhB and

MeB, respectively, at 0 h and 15 and 4 for 68 h and higher, respectively), xt is the integration value for the 8.13-8.02 ppm region relative to the integration of the total aryl region, yt, at 8.31-6.26 ppm for PhB (for MeB, xt and yt are the integration regions of

7.87-7.75 ppm relative to 8.20-6.23 ppm, respectively), and n0, x0, and y0 represent the 0 h values.

123

5.2.8. Identification of Model Compound Degradation Products

After the previously mentioned 3 M NaOD/CD3OD/D2O, 80 °C, 240 h degradation test of MeB and PhB, each solution was cooled to room temperature. The organic degradation products were then isolated using the following method (PhB as the example):

The PhB mixture was acidified with dilute, aqueous hydrochloric acid until the pH was neutral, resulting in precipitate. Diethyl ether was added to fully dissolve the precipitate and the organic layer was washed with water three times, dried over MgSO4, filtered, and evaporated at 40 °C using a dynamic vacuum. The resulting residue was then analyzed by mass spectrometry. The starting PhB material was also analyzed by mass spectrometry for comparison.

5.2.9. Polymer Alkaline Stability

Membrane pieces of PPMB (chloride-form) were initially soaked in 1 M KOH for 48 h at room temperature followed by soaking in DI water with multiple fresh exchanges over an additional 48 h at room temperature under ambient atmosphere. One of the membrane pieces was then soaked in 1 M NaCl for 48 h (with one fresh exchange in between) and DI water for 24 h (multiple fresh exchanges). After drying at 50 °C under 1 vacuum, the piece was dissolved in DMSO-d6 and analyzed by H NMR spectroscopy (represents the “initial” spectrum). Other membrane pieces (~25 mg each) were immersed in either 30 mL of aqueous 1 M or 2 M KOH in closed FEP containers inside an 80 °C oven for up to 168 h. These membrane pieces were removed at certain points in time (such as 48, 96, and 168 h) and then transferred into 1 M NaCl or KCl, soaking for at least 48 h (with one fresh exchange in between) followed by soaking in DI water for at least 24 h (with multiple fresh exchanges). These membrane pieces were then dried under an argon stream and part of each sample was dissolved in DMSO-d6 and analyzed by 1H NMR spectroscopy. The 1H NMR spectra of the 168 h test samples were baseline-corrected using the “Full Auto (Polynomial Fit)” function found in MestReNova 9.0.1.

124

The extent of degradation was calculated using Equation 12, where xt represents the integration of the 8.56-8.01 ppm region relative to the aryl region, yt, of 8.56-

5.88 ppm, and n is equal to 1 (as there is no significant deuterium present); x0 and y0 represent the same region integrations but for the “initial” spectrum only.

5.2.10. Mechanical Strength

A PPMB (iodide form) membrane, that was previously washed with water and dried under vacuum at 80 °C, was cut into a barbell shape using an ASTM D638-4 cutter. The 40 μm thick sample was pulled apart at both ends at a rate of 5.00 mm min-1 on a single column system (Instron® 3344 Series) until broken under ambient conditions (21 °C, 42% RH). The measured force at each point was then used to calculate the stress. The Young’s modulus was calculated from the slope of a linear regression in the 0.5%-2.0% strain region.

5.2.11. Single Crystal X-ray Diffraction

The four different model compounds (HB, BrB, MeB, and PhB) were crystallized in their iodide forms (see Table D1 for the crystallization method). The single-crystal x- ray crystallography was performed on a Bruker SMART APEX II system with an APEX II CCD detector and a tunable graphite crystal monochromator. The detector was placed at 5.0 cm from each crystal and measured under ambient conditions. The data was collected and processed using the APEX2 Suite followed by structural refinements using ShelXle.229 The collected crystal data is tabulated in Table D1 and CCDC 1439721- 1439724 contains the supplementary crystallographic data for this paper. The data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/getstructures. The structural figures were prepared using Mercury.230

5.2.12. Density Functional Theory (DFT) Calculations

The DFT calculations were performed by Thomas Weissbach. Electronic structure calculations for the model compounds as well as the degradation of HB, MeB,

125

and PhB were performed using Gaussian G09,231 B3LYP density functional theory (DFT), and Polarizable Continuum Model (PCM) using the in G09 integrated Integral Equation Formalism (IEFPCM) with water as solvent (ε = 78.36). Pre-optimization was performed using 6-31G(d) basis set. Final calculations were done using 6-311++G(2d,2p) basis set, tight convergence criteria and no symmetry. Structures of reagents, intermediate structures (IS) and products (P) were optimized to energy minimum; transition states (TS) were optimized using the G09 implemented Berny algorithm, having one imaginary frequency (i.e., passing through an energy maximum saddle point). Intermediates of HB were confirmed by calculating the intrinsic reaction coordinates. Frequency analysis was performed using a temperature of 298.15 K. Reaction free energy (ΔG) and reaction free energy barrier (ΔG‡) are given with respect - to the sum of the reagent free energy: benzimidazolium cation + 2 OH for the addition- - elimination reaction and benzimidazolium cation + OH for the SN2 reaction.

5.3. Results and Discussion

5.3.1. Model Compound Synthesis

To find suitable conditions to prepare a bromo-protected benzimidazole, methyl 2,6-dibromobenzoate (9), prepared in Chapter 4, was reacted with o-phenylenediamine in Eaton’s reagent at various temperatures (Scheme 5.1). When the reaction mixture was heated to 110 °C and poured into water, only starting material 9 was observed, signifying good chemical stability. When the mixture was heated to 140 °C for 15 min followed by heating to 150 °C for 15 min, the precipitate contained a product mixture, based on 1H NMR spectroscopy, of approximately 9:1 11:9. Thus, when 9 was heated at 150 °C for 1 h, the desired product 11 was quantitatively isolated. Using the same conditions on the carboxylic acid version of 9 (8), 11 was also successfully isolated in high purity.

126

Scheme 5.1. Condensation reaction of 9 with o-phenylenediamine to produce 11.

The second main step for preparing the phenyl-protected benzimidazole is the Suzuki-Miyaura coupling of the bromo groups with phenylboronic acid. As the N-H proton on the benzimidazole could become de-protonated due to the base present in the coupling reaction, it was first protected by either monomethylation or dimethylation of the benzimidazole (Scheme 5.2).

127

Scheme 5.2. Synthetic route to phenyl-protected benzimidazole 13 and phenyl- protected benzimidazolium PhB.

Dimethylation of 11 to produce the dibromo-protected benzimidazolium model compound BrB allowed for a subsequent direct reaction to the final desired model compound PhB. The Suzuki-Miyaura coupling of BrB with phenylboronic acid in an aqueous-THF mixture resulted in a low yield of PhB, as the mono-phenylated intermediate compound BrPhB precipitated from the solvent mixture. BrPhB could be isolated in high purity, as shown in Figure 5.1. To improve the solubility of this intermediate during the coupling reaction, ethanol was also added, greatly increasing the yield of PhB. This was important in demonstrating that PhB could be synthesized by this

128

new route of forming a benzimidazole followed by addition of aryl-protecting groups, contrary to the numerous attempts in Chapter 4. However, after the 25 h Suzuki-Miyaura coupling reaction, BrPhB was still observed (>10%), making the purification of the ethanol/water-soluble PhB difficult. Additionally, the counter-ion of PhB was unknown, - - - - as a number of salts (i.e., OH , HCO3 , B(OH)4 , and I ) were present in the reaction mixture.

1 Figure 5.1. H NMR spectrum (400 MHz, DMSO-d6) of isolated BrPhB from the Suzuki-Miyaura coupling of BrB with phenylboronic acid in aqueous-THF.

In order to obtain high purity PhB, a second synthetic route, involving the initial monomethylation of 11 to produce 12, was used (Scheme 5.2). By performing a Suzuki- Miyaura coupling of 12 followed by a later second methylation allowed for the synthesis of PhB with only one expected counter-ion (I-). The coupling reaction of 12 was achieved in moderate yield, due to a number of side products that formed (which will be later identified and discussed in Section 5.3.4). The methylation of 13 successfully produced desired phenyl-protected benzimidazolium, PhB (iodide form), in high purity.

5.3.2. Model Compound Stability

The hydroxide stability of four similar benzimidazolium, each bearing either ortho- positioned hydrogen atoms (HB), bromine atoms (BrB), methyl groups (MeB), or phenyl

129

groups (PhB) was then tested. Each of the four model compounds (Figure 5.2) was subjected to the same accelerated hydroxide stability test, which involved dissolution of the model compound (0.02 M) in 3 M NaOD/CD3OD/D2O (7:3 wt. CD3OD:D2O). The solution was deuterated in order to analyze the mixtures by 1H NMR spectroscopy, as hydrogen-containing solvents would produce large overlapping signals. Methanol was used to solubilize all starting materials and products from the reaction, ensuring no precipitates formed over the course of the reaction. The solutions were heated to 80 °C for up to 240 h. Aliquots were intermittently extracted and analyzed by 1H NMR spectroscopy, allowing the extent of degradation to be quantified using Equation 12 (Section 5.2.7), as plotted in Figure 5.3.

Figure 5.2. Chemical structures of the four ortho-protected model benzimidazolium compounds examined for hydroxide stability, where X- is the counter- anion.

130

Figure 5.3. Measurement of remaining starting material over time for the dissolved model compounds (0.02 M) in 3 M NaOD/CD3OD/D2O at 80 °C as determined by 1H NMR spectroscopy.

Compound HB began degrading immediately after its dissolution in the basic solution at room temperature and was fully degraded to its amide product by time of the first measurement (Figure 5.4), demonstrating extreme lability of unprotected benzimidazoliums in strongly alkaline media.

1 Figure 5.4. H NMR spectra of HB (0.02 M) in 3 M NaOD/CD3OD/D2O taken after dissolution (“0 h”) as well as HB (0.02 M) in pure CD3OD (without NaOD/D2O).

131

BrB appeared to be inert at room temperature, but fully degraded after 17 h when the temperature was raised to 80 °C (Figure 5.5). The chemical shift transition of the resonance from 4.00 ppm to 3.90 ppm in the 1H NMR spectrum suggests that new dimethylated benzimidazoliums are produced, which form from the nucleophilic displacement of bromide for hydroxyl groups, as well as amide products, which appear at 3.0-2.7 ppm.

1 Figure 5.5. H NMR spectra of BrB (0.02 M) in 3 M NaOD/CD3OD/D2O after heating at 80 °C for the specified duration.

There was degradation for MeB and PhB, as shown in Figure 5.6 and Figure 5.7, respectively. Quantification of their degradation followed exponential decay, indicative of a pseudo-first order reaction. By fitting the data to exponential functions, the half-life (t1/2) for MeB and PhB at 80 °C in those solutions was calculated. The rate of degradation of

PhB (t1/2 3240 h) was ~7 times slower than that of MeB (t1/2 436 h), which represents the highest alkaline stability for a benzimidazolium hydroxide reported to date.

132

1 Figure 5.6. H NMR spectra of MeB (0.02 M) in 3 M NaOD/CD3OD/D2O after heating at 80 °C for the specified duration.

133

1 Figure 5.7. H NMR spectra of PhB (0.02 M) in 3 M NaOD/CD3OD/D2O after heating at 80 °C for the specified duration.

It is important to note that significant deuterium-exchange occurs at the 4- and 7- positions of the benzimidazolium and on all methyl (-CH3) groups. The 4- and 7-positions exchanged for deuterium readily, such that the 1H NMR spectra had an integration decrease of 2H in aromatic region after 65 h of the degradation test. The methyl groups exchanged slowly over time to various amounts of deuterium (i.e. -CH2D, -CHD2, and 1 •CD3), resulting in a decrease in their H NMR spectroscopic signatures as well as their resonances changing from singlets to multiplets. For example, the integration of the N- methyl protons (at 3.9 ppm) on MeB (Figure 5.6) decreased to ~20% of its original value after 240 h (i.e., ~80% of the hydrogen exchanged for deuterium). The ortho-methyl groups (at 2.07 ppm) of MeB exchanged for deuterium slightly faster (87% exchanged for deuterium) while the deuterium exchange at the para•methyl group (at 2.4 ppm) was slower (57% exchange). The deuterium-exchange for the N-methyl hydrogens (at 3.6 ppm) of PhB (Figure 5.7) was similar, with 75% deuterium-exchange after the 240 h

134

- experiment. Additionally, deuterated methoxide (CD3O ) was not observed to attack the C2 position.

5.3.3. Single Crystal XRD Structures and DFT Calculations

To investigate the origin of the stability differences between the C2-protected benzimidazolium small molecules, single crystals were grown and characterized by XRD. Furthermore, the structures were compared to those found using DFT calculations. Each compound was crystallized in its iodide form. The methods for their crystallization as well as their relevant crystal data can be found in the Appendix D, Table D1. Refined crystal structures are shown in Figure 5.8. Using the crystal structures, the dihedral angles between the benzimidazolium plane and that of the C2-substituted phenyl plane as well as the shortest distance between the C2 carbon and iodide, were measured (Table 5.1).

135

Figure 5.8. X-ray crystal structures of model compounds in their iodide form (ellipsoids set at 50% probability) alongside the dihedral angles measured (A represents the 2-phenyl plane and B represents the benzimidazolium plane). Only one of the two unique BrB structures is shown for clarity and PhB co-crystallized with H2O (where the hydrogen atoms of H2O are not shown).

136

Table 5.1. Properties of the model compounds based on experimental data and DFT calculations.

- Compound Solid-State Solution-State C2-I t1/2 Dihedral Anglea Dihedral Angleb (Å)c (h)d HB 54.40/55.02 62 3.704 < 0.1 70.45/73.26 5.497 BrBe 88 < 10 73.08/81.92 5.587 MeB 79.21/83.77 86 4.743 436 PhB 65.03/68.58 71 6.218 3240 aMeasured between the benzimidazolium and C2 phenyl planes in the iodide form from XRD below 90°. bDFT calculated solution structures. cThe shortest C2 carbon–iodide distance(s) for the iodide-form X-ray structures. d The half-life of the compound dissolved in 3 M NaOD/CD3OD/D2O at 80 °C. eBrB (XRD) possessed two unique structures within one unit cell.

The solid-state dihedral angles within each molecule were unique for each quadrant owing to the non-planarity of the benzimidazolium ring. BrB possessed the largest variation of dihedral angles, and also possessed two molecular structures in its unit cell, leading to eight different dihedral angles. The average dihedral angles increased in the order HB < PhB < BrB < MeB. As this trend does not follow the trend in half-life in strong base, the dihedral angle alone cannot be used as a measure of hydroxide stability. However, the C2 carbon-iodide distance does match the trend in half•life, with the longer distance translating to a longer half-life. The exception to this trend is BrB, as its protecting bromine groups are strongly susceptible to nucleophilic displacement.

To compare the hydroxide stability differences between HB, MeB, and PhB, DFT was used to calculate the energy barriers and states along two possible degradation pathways, which are graphically displayed in Figure 5.9 (Thomas Weissbach performed these calculations). The overall reaction for each pathway is shown in Scheme 5.3. The first pathway represents the nucleophilic addition–elimination reaction of hydroxide on the C2 carbon of the benzimidazolium, resulting in the amide “ring-opened” product. The second pathway represents the nucleophilic substitution of hydroxide with the N-methyl carbon, resulting in a 2-substituted-1-methylbenzimidazole, which is termed as “de- methylation” degradation.

137

Figure 5.9. Reaction profiles for the two hydroxide-mediated degradation pathways (de-methylation and ring-opening) for HB, MeB, and PhB. The dotted lines represent the higher energy, TS2,cis, ring-opening degradation pathway. No barrier was found between IS1 and IS2.

Scheme 5.3. The two degradation pathways for benzimidazolium hydroxides (ring•opening and de-methylation).

As observed in Figure 5.9, the nucleophilic addition–elimination reaction on the

C2 carbon of the benzimidazolium leads to the formation of an intermediate state (IS1) after overcoming the first transition state (TS1). HB has a reaction free energy barrier ǂ -1 (ΔG ) of 10.6 kcal mol for TS1, which is considerably lower in energy compared to MeB (22.9 kcal mol-1), and is similar to findings of Long and Pivovar.232 As ΔGǂ is greatest for

TS1, the higher the energy for this rate-limiting step, the slower the ring-opening degradation. As such, MeB should have improved stability over that of HB, which is in good agreement with experimental observation. PhB is even more resistant to ring- opening degradation, consistent with the larger ΔGǂ (24.2 kcal mol-1).

138

The second transition state (TS2) may proceed by one of two ways depending on the orientation of the two N-methyl groups (TS2,trans or TS2,cis) and results in two different configurational isomers of the amide product. The 1H NMR spectra of degraded MeB reveal numerous amide products being formed, as only two alkyl resonances are expected for a single isomer configuration in the 3.0–2.0 ppm resonance region. Degraded products were also isolated (see Section 5.2.8 for method) and analyzed by mass spectrometry (Figure 5.10). The starting material was highly pure by mass spectrometry (Figure 5.11) whereas only the amide products were observed for the degraded species. This agrees well with the degradation test previously shown in Section 3.3.6 (Figure 3.8). Various amounts of deuterium exchange on the methyl groups was also observed, as previously discussed in Section 5.3.2.

Figure 5.10. The ESI mass spectrum of the isolated organic degradation products from MeB after 240 h in 3 M NaOD/CD3OD/D2O at 80 °C.

139

Figure 5.11. The ESI mass spectrum of the MeB before the degradation test.

However, when the same process was performed on the isolated PhB degradation products, two products were observed (Figure 5.12) that were not present in the mass spectrum prior to the degradation test (Figure 5.13). The ring-opened amide was present alongside the de-methylated product, which is the first observation of its kind for an alkali-degraded benzimidazolium hydroxide.

Figure 5.12. The ESI mass spectrum of the isolated organic degradation products from PhB after 240 h in 3 M NaOD/CD3OD/D2O at 80 °C.

140

Figure 5.13. The ESI mass spectrum of the PhB before the degradation test.

DFT calculations indicated that the activation energies of de-methylation differ ǂ only slightly between HB, MeB, and PhB (ΔG of TSSN2 of 27.4, 26.9, and •1 27.3 kcal mol , respectively). As TS1 is generally significantly lower than TSSN2, the de- methylation product is usually not observed. However, the substantial increase in the ǂ ΔG of TS1 for PhB has decreased the energetic advantage of ring-opening degradation over that of de-methylation, with a difference of only 3.1 kcal mol-1. While the effects of methanol were not considered in the DFT calculations, the estimated differences between the degradation rate and mechanism of individual model compounds are in good agreement between DFT and experiment.

5.3.4. Monomer and Polymer Synthesis

Motivated by the stability of PhB, various monomers were prepared and attempted in polymerizations to produce polymers having poly(phenylene) backbones with pendant benzimidazoles, as shown in Scheme 5.4. Due to the versatility of the Suzuki-Miyaura reaction, it was used to prepare both monomers, such as dichloro monomer 14, as well as polymers, such as PPB2. In each case, the benzimidazole moiety is ortho-positioned to two phenyl groups, providing an ideal repeating structure of phenyl-protected benzimidazoles.

141

Scheme 5.4. Synthetic routes attempted to prepare various poly(phenylene) polymers bearing pendant benzimidazoles.

142

For the preparation of PPB2, 12 was reacted with a diboronate pinacol ester in toluene/2 M K2CO3 (aq) at 100 °C for 22 h ([Pd] = Pd(PPh3)4). Over the course of the reaction, a white precipitate formed, which was analyzed by 1H NMR spectroscopy (Figure 5.14). While the integration ratio of the aromatic region compared to the alkyl region was consistent with the structure of PPB2, the large number of resonances suggested that only oligomers had formed. Additionally, a singlet at 3.88 ppm was indicative of the doubly dehalogenated species (16). This suggests that the low molecular weight was not only due to the precipitation from the solution, but also due to dehalogenation during the reaction, which explains the large product distribution. This may also explain why the previous synthesis of the phenyl-protected small molecule 13 was only produced in moderate yield.

1 Figure 5.14. A selected region in the H NMR spectrum (400 MHz, DMSO-d6) of the product from the attempted synthesis of PPB2.

As the dehalogenation was potentially a result of the steric hindrance at the bromo-position and/or the lone pair of the benzimidazole nitrogen, Suzuki-Miyaura coupling of 12 with 4-chlorophenylboronic acid was used to prepare 14, shown in Scheme 5.5. This takes advantage of the selective reactivity of Pd catalysts with bromo groups over that of chloro groups, as many Pd complexes are unable to undergo oxidative addition with chloroaryl compounds.233,234 14 has chloro groups at positions that have significantly less steric hindrance, which should be substantially more reactive

143

for higher molecular weight polymers. The monomer was successfully prepared using a catalytic amount of Pd(PPh3)4 but was isolated in low yield, as the monomer had to be purified by a combination of column chromatography and multiple recrystallizations. To understand the side products from the reaction, mass spectrometry was used on a mixture of impurities collected by chromatography (Figure 5.15a). A number of dechlorinated and debrominated species were observed, revealing not only why 13 and 14 could only be prepared in low yield by Suzuki-Miyaura coupling but also why the preparation of PPB2 produced only oligomers.

Scheme 5.5. Synthetic route used to prepare monomers 14 and 15.

144

(a)

(b) Figure 5.15. (a) Mass spectrum (ESI-MS) of by-products synthesized from the preparation of 14 by Suzuki-Miyaura coupling of 12 with 4•chlorophenylboronic acid, and (b) mass spectrum (ESI-MS) of pure 14.

Nonetheless, the preparation of high purity 14 (Figure 5.15b), which was also used to prepare monomer 15 (Scheme 5.5), allowed for Ni-based polymerization reactions. As opposed to the Pd-based reactions, such as with Pd(PPh3)4, which are generally unreactive with aryl chlorides, the oxidative addition of aryl chlorides with Ni(0)

145

is more facile.235–239 Kende et al. and Colon et al. demonstrated that aryl halides could be used to form aryl-aryl compounds in high yield as well as aromatic polymers by using catalytic amounts of Ni(II) catalyst (such as NiCl2) in the presence of excess Zn metal.

The Zn metal reduces the starting Ni(II) complex as well as the formed NiCl2 after the coupling reaction back to the active Ni(0) complex in situ.240–242 Unfortunately, when a similar procedure was used on 14 and 15 in DMF at 80 °C (NiBr2/NiCl2, Zn, bipy/PPh3), no polymer was observed. The main products observed were generally dehalogenated species, showing that the oxidative addition was successful but aryl-aryl formation was not. Colon et al. suggested that one reason for this type of product is due to the reaction of trace water, which is more likely to occur with 15, as it is a charged species and likely absorbs atmospheric water.241 The second explanation was that aryl-Ni complexes are long-lived, such that when the product mixture is precipitated into aqueous solution, the dehalogenated product forms. As such, a second method was attempted instead to prepare the polymers, involving the use of a stoichiometric excess of bis(1,5- 243 cyclooctadiene)nickel(0) (Ni(COD)2), known as Yamamoto coupling.

Yamamoto coupling has been an effective method for the preparation of numerous poly(phenylene)s,243–248 as it can be used to produce homopolymers and is performed in polar, aprotic solvents (such as DMF and DMAc), which are usually required to be able to solubilize stiff-chain polymers.249 This is in contrast to Suzuki- Miyaura coupling reactions, which generally require two monomers (a dibromo monomer and a diboronic ester monomer) with equally high purities and are performed with two solvents (such as toluene and water) that are not able to solubilize such polymers. When

Yamamoto coupling was performed on 14 with excess Ni(COD)2 and bipy in DMF at 80 °C, the polymer PPB3 was successfully synthesized. However, when the polymer was fully methylated to produce PPMB and cast from DMSO, only pieces of cracked membrane resulted, suggesting that the molecular weight was not high enough. Polymerization of 15 to directly obtain PPMB was also successful using this method and when cast from DMSO, produced pale yellow membranes that were tough and flexible. This suggested that the molecular weight was improved by using this method. However, when the membrane was immersed in hydroxide solution to exchange the anion to hydroxide form and then washed with water, the membrane broke into pieces.

146

As the membrane of PPMB (produced from 15) broke into pieces rather than dissolved, it suggested that the molecular weight was still too low. Additionally, the initial counter-ion of PPMB was also unknown, as the polymer was purified by an acid wash to remove the nickel. When monomer 12 was used in a Yamamoto coupling in DMF, no precipitate was produced when the solution was added to water and suggesting no polymer PPB1 formed. As previously discussed for the Suzuki-Miyaura coupling of 12, ortho-positioned halogens do not appear to form the desired coupled products.

As PPB3 was the only neutral polymer that could be synthesized, the solvent of the reaction was changed in an attempt to improve the molecular weight. When 14 was reacted by Yamamoto coupling in DMSO as opposed to DMF, the polymer PPB3 was prepared in near quantitative yield (95%) as a white fluffy powder. It showed minimal solubility in most solvents tested, including DMSO itself. The only solvents tested that fully dissolved the polymer were NMP and 1,1,2,2-tetrachloroethane (TCE), which produced very viscous and colourless solutions. Complete methylation of PPB3 in NMP with iodomethane quantitatively yielded PPMB which could be dissolved in DMSO to produce viscous solutions and cast into strong, pliable membranes. Unlike previously, when the membrane was soaked in hydroxide solution to exchange the anion to hydroxide form followed by immersion into water to wash the remaining salts out, the membrane was mechanically stable and colourless. This demonstrates that using DMSO in the Yamamoto coupling successfully produced a high molecular weight poly(phenylene) with pendant benzimidazoles. However, it is unclear why DMSO was better than DMF. Recent reports have used DMSO in Yamamoto couplings108,250 which contrasts an early report in 1971 suggesting that DMF was the only satisfactory solvent for Ni(COD)2 reactions, as DMSO would result in the rapid decomposition of the nickel reagent.251

5.3.5. Polymer Properties and Stability

As the neutral polymer, PPB3, prepared by Yamamoto coupling of 14, was sparingly soluble in typical gel-permeation chromatography solvents, the molecular weight was estimated using viscosity measurements of the polymer in NMP. Using a Huggins-Kraemer plot, as shown in Figure 5.16, the intersecting y-intercept represents

147

the intrinsic viscosity of the polymer. The calculated intrinsic viscosity for PPB3 was 2.10 dL g-1. If the intrinsic viscosity of PPB3 dissolved in NMP behaved similarly to that -4 of m-PBI in DMF, which has Mark–Houwink constants a = 0.75 and Kw = 3.2 × 10 dL g•1,252 PPB3 would have a molecular weight of 129,000 g mol-1. This would represent a high molecular weight polymer having 360 repeat units.

Figure 5.16. A Huggins-Kraemer plot of PPB3 in NMP calculated from the measured viscosities at 25.0 °C for various concentrations (c).

Complete methylation of PPB3 with iodomethane produced PPMB in its iodide form. As a membrane, PPMB was strong and flexible, possessing a high tensile strength of 72 MPa, elongation at break of 49%, and Young’s modulus of 1.29 GPa, as shown in Figure 5.17. When compared to HMT-PMBI (from Chapter 3), the tensile strength of PPMB is higher but the elongation at break is lower. In its hydroxide form, the colourless -1 and transparent film possessed an ion-exchange capacity (IECOH-) of 2.56 meq g , which is similar to the IEC of the optimal 89% dm HMT-PMBI (2.51 meq g-1). In its fully hydrated state and in air, PPMB exhibited a mixed hydroxide/carbonate conductivity of 13.2 ± 1.4 mS cm-1 (at 22 °C), which is twice the conductivity of 89% dm HMT-PMBI and possesses a similar water uptake of 81 ± 10%. This suggests that a pendant structure is better for hydroxide conduction as opposed to ionene-type materials.

148

Figure 5.17. A stress-strain curve of PPMB (iodide form) under ambient conditions (21 °C, 42% RH) with a cross-head speed of 5.00 mm min-1.

After immersion of the membrane in 1 M or 2 M KOH at 80 °C for 168 h, only 1.7% and 5.3% degradation was observed, respectively (Figure 5.18), which is unprecedented for a benzimidazolium-containing polymer. A plot of the stability of PPMB in 2 M KOH at 80 °C over time is shown in Figure 5.19, which did not appear to follow first-order kinetics. As this is unlike the small molecule stability tests, it can be presumed that this is due to the distinct differences between homogeneous and heterogeneous degradation experiments.

149

1 Figure 5.18. H NMR spectra of PPMB (chloride form) in DMSO-d6 before (“initial”) and after being subjected to either 1 M or 2 M KOHaq. at 80 °C for 168 h.

Figure 5.19. Relative percent remaining benzimidazolium of PPMB over time in 2 M 1 KOHaq. at 80 °C for up to 168 h, as measured by H NMR spectroscopy of the membranes in DMSO-d6.

150

5.4. Conclusion

Using a new synthetic method in which the benzimidazolium is formed prior to functionalization with the o-aryl groups, aryl-protected benzimidazoliums were successfully prepared. Through examination of benzimidazolium hydroxide model compounds, XRD, and DFT calculations, correlations between their properties with stability of four C2-protecting groups have been rationalized. Changing the 2,6•substituents of the C2-phenyl decreased the hydroxide stability in the order of phenyl > methyl > bromo > hydrogen. Modification of the synthetic route also allowed access to a new class of poly(phenylene)s, in which the benzimidazolium is pendant. This allows the polymer backbone itself to serve as the protecting group, such that the benzimidazolium has ortho-disubstituted phenylenes. This polymer provided exceptional stability in alkaline solutions at 80 °C, high molecular weight, and exceptional mechanical properties. Moreover, a versatile synthetic route is presented that facilitates further investigations of numerous other C2-protecting groups.

151

Chapter 6.

Conclusions and Future Work

6.1. Conclusions

In Chapter 2 and Chapter 3, HMT-PMBI was found to be a promising candidate as a benchmark cationic polymer for membrane and ionomer applications. The polymer backbone was based on poly(benzimidazole), allowing it be prepared by polycondensation of a specifically designed diester with 3,3’-diaminobenzidine. Using an original methylation procedure, HMT-PMBI could be prepared with a controlled degree of methylation and a target ion-exchange capacity. This made it possible to prepare membranes with high ionic conductivities that remain insoluble in water. Akin to m-PBI, HMT-PMBI exhibited excellent mechanical properties. It showed high stability in alkaline conditions, with no degradation in 2 M KOH at 60 °C, and demonstrated solid performance in AAEMFC and water electrolyzer applications at 60 °C for >100 h. Additionally, the ability to scale the synthesis up while improving the yield allows this material to meet almost every criteria desired for an ideal cationic polymer. Furthermore, this material is the first example of an ionene tested in energy conversion applications. Unfortunately, at even greater hydroxide concentrations, HMT-PMBI was susceptible to degradation.

As such, Chapter 4 and Chapter 5 focused on the development of new sterically- protecting groups for benzimidazoliums, in which phenyl-protecting groups were found to be more than 7 times more resistant to hydroxide than that of methyl-protecting groups. With such a substantial improvement in stability and a method for preparing the poly(phenylene) polymer PPMB which bears pendant benzimidazoliums, a membrane with greater chemical stability was fabricated. It was found that the synthesis of a poly(benzimidazole)-backbone with phenyl-protecting groups was not possible due to

152

either intramolecular cyclization or severe steric hindrance. As such, this moved the polymer structure from that of an ionene to the more classical pendant cationic polymer structure, where it also exhibited improved ex situ hydroxide conductivity of over nearly twice that of HMT-PMBI at the same IEC. PPMB was insoluble in water and also mechanically robust. However, phenyl-protected benzimidazoliums were not found to be immune to hydroxide attack. Through small molecule degradation tests, such materials degrade by a combination of ring-opening and de-methylation. This means that the phenyl group is still insufficient at sterically-protecting the C2 position from nucleophilic hydroxide attack, allowing for the future development of materials with further improved stability.

6.2. Future Work

For the future development of stable cationic materials, benzimidazolium compounds provide a solid foundation to build on. They can be easily functionalized at monomer or polymer stages and are relatively inexpensive starting materials. When combined with the ability to prepare almost any sterically-protected benzimidazolium using the new Suzuki-Miyaura coupling route discovered, the possibilities are endless. However, the degradation products from the phenyl-protected benzimidazolium provide insight into the best directions to move in. For example, one of the two degradation products was due to C2 hydroxide attack. This suggests that the phenyl-protecting groups are not sterically-protecting enough. As such, it could be envisioned that bulkier steric groups (R1 groups, shown in Figure 6.1) would provide benzimidazoliums with better stability. One example would be using o-mesityl groups.

Figure 6.1. Benzimidazolium structures with varying R1, R2, and R3 groups that could be theoretically prepared.

153

However, it is important to note that moving to bulkier steric groups is not only synthetically challenging, as demonstrated in Chapter 4 and Chapter 5, but also decreases the IEC of the material. By choosing larger and larger R1 groups, the IEC will drop to a point that hydroxide conductivity begins to suffer. Based on the IEC, conductivity, and water uptake of the polymers prepared in this work, materials with IECs between 2.5 and 2.7 meq g-1 provided optimal balance of properties. As it is simple to calculate the theoretical IEC of a polymer structure, it would be important to ensure potential materials have IECs capable of being in this range.

The second possible direction for this work is based on the second degradation product that formed, which was due to de-methylation. Based on the DFT calculations from Chapter 5, it was found that the transition state was unaffected by changes of the methyl or phenyl-protecting groups, suggesting that the only method for preventing this is by moving to different R2 groups, as shown in Figure 6.1. In this case, it important to keep in mind that only specific R2 groups that are stable should be chosen. For example, a longer alkyl chain, such as ethyl, would potentially introduce new degradation pathways, such as β­Hofmann elimination. This significantly limits the possibilities, such as using benzyl R2 groups. However, similar to the first suggestion, it becomes increasingly difficult to synthesize not only small molecules but also polymers with such functional groups. This would also lower the IEC and potentially cause the conductivity to eventually suffer.

The third direction with regards to benzimidazoliums is the ability to functionalize the 4- and 7-positions of the benzimidazolium (R3 groups). With both mes-PDMBI and HMT-PMBI, these positions rapidly exchanged with deuterium when under alkaline conditions, suggesting the reversible formation of a zwitterion intermediate, as shown in Scheme 6.1. If the material continuously forms zwitterions in conversion device applications, this would lead to a systematic decrease in overall hydroxide conductivity, as a number of the hydroxide anions would be temporarily immobilized. Additionally, the intermediate could potentially react with electrophilic impurities, resulting in irreversible products. By pre-functionalizing these positions of the benzimidazolium, this may not only improve the conductivity of the materials but also the overall stability, as it is unknown what effect this would have on the nucleophilic displacement degradation.

154

Scheme 6.1. A possible mechanism for deuterium-exchange at the 4- and 7-positions of benzimidazoliums.

A different avenue of synthetic research could be directed towards benzimidazolium’s parent molecule, imidazolium. Due to its nearly identical structure, the same ideas discovered for benzimidazolium would apply to imidazolium, which can also be easily prepared and functionalized. In fact, Coates et al. recently demonstrated that imidazolium small molecules could also be sterically protected using the same ortho- dimethyl protecting groups to prepare very inert cations.253 Additionally, Long and Pivovar, using DFT calculations, found that similarly protected imidazoliums should be more durable than that of benzimidazoliums.232 As imidazolium is an intrinsically smaller cation than benzimidazolium, polymers containing imidazolium would also have theoretically higher IECs and should be excellent candidates for further study.

Lastly, outside of synthetic development, it should not be forgotten that all quantified degradation tests in this work and most often in literature were performed ex situ. This is mainly due to convenience and simplified conditions, allowing for easier replication of results. For example, high base concentrations (>2 M) are used to accelerate the degradation rate, such that long periods of time are not required. Additionally, non-aqueous solvents are also typically used to be able to dissolve the materials and products for spectroscopic analysis over time. Unfortunately, these

155

conditions are often unrealistic and extreme. For the field of alkaline energy conversion devices to move forward, it is necessary to establish a robust ex situ degradation test that is comparable to in situ degradation. Failure to do so could result in materials with poor ex situ stability being overlooked when they may in fact possess high in situ stability. This is also vital in being able to compare materials developed between other research groups, as there is currently no universal quantified degradation test.

156

References

(1) Lewis, N. S.; Nocera, D. G. Proc. Natl. Acad. Sci. 2006, 103 (43), 15729–15735.

(2) Hoffert, M. I.; Caldeira, K.; Jain, A. K.; Haites, E. F.; Harvey, L. D. D.; Potter, S. D.; Schlesinger, M. E.; Schneider, S. H.; Watts, R. G.; Wigley, T. M. L.; Wuebbles, D. J. Nature 1998, 395 (6705), 881–884.

(3) Panwar, N. L.; Kaushik, S. C.; Kothari, S. Renew. Sustain. Energy Rev. 2011, 15 (3), 1513–1524.

(4) Shafiee, S.; Topal, E. Energy Policy 2009, 37 (1), 181–189.

(5) Maier-Reimer, E.; Hasselmann, K. Clim. Dyn. 1987, 2 (2), 63–90.

(6) Wigley, T. M. L.; Richels, R.; Edmonds, J. A. Nature 1996, 379 (6562), 240–243.

(7) Jacobson, M. Z.; Delucchi, M. A. Energy Policy 2011, 39 (3), 1154–1169.

(8) Moriarty, P.; Honnery, D. Energy Policy 2016, 93, 3–7.

(9) Jones, G. A.; Warner, K. J. Energy Policy 2016, 93, 206–212.

(10) Jacobson, M. Z.; Delucchi, M. A.; Bazouin, G.; Bauer, Z. A. F.; Heavey, C. C.; Fisher, E.; Morris, S. B.; Piekutowski, D. J. Y.; Vencill, T. A.; Yeskoo, T. W. Energy Environ. Sci. 2015, 8 (7), 2093–2117.

(11) Steinke, F.; Wolfrum, P.; Hoffmann, C. Renew. Energy 2013, 50, 826–832.

(12) Lewis, N. S. Science 2016, 351 (6271), 353.

(13) Perrin, M.; Saint-Drenan, Y. M.; Mattera, F.; Malbranche, P. J. Power Sources 2005, 144 (2), 402–410.

(14) Scrosati, B.; Garche, J. J. Power Sources 2010, 195 (9), 2419–2430.

(15) Skyllas-Kazacos, M.; Chakrabarti, M. H.; Hajimolana, S. A.; Mjalli, F. S.; Saleem, M. J. Electrochem. Soc. 2011, 158 (8), R55–R79.

157

(16) Skyllas-Kazacos, M.; Kazacos, G.; Poon, G.; Verseema, H. Int. J. Energy Res. 2010, 34 (2), 182–189.

(17) Thomas, C. E. Int. J. Hydrog. Energy 2009, 34 (15), 6005–6020.

(18) Gröger, O.; Gasteiger, H. A.; Suchsland, J.-P. J. Electrochem. Soc. 2015, 162 (14), A2605–A2622.

(19) Yilanci, A.; Dincer, I.; Ozturk, H. K. Prog. Energy Combust. Sci. 2009, 35 (3), 231– 244.

(20) Luo, X.; Wang, J.; Dooner, M.; Clarke, J. Appl. Energy 2015, 137, 511–536.

(21) Budischak, C.; Sewell, D.; Thomson, H.; Mach, L.; Veron, D. E.; Kempton, W. J. Power Sources 2013, 225, 60–74.

(22) Conte, M.; Iacobazzi, A.; Ronchetti, M.; Vellone, R. J. Power Sources 2001, 100 (1–2), 171–187.

(23) Vaidya, P. D.; Rodrigues, A. E. Chem. Eng. J. 2006, 117 (1), 39–49.

(24) Rydén, M.; Lyngfelt, A. Int. J. Hydrog. Energy 2006, 31 (10), 1271–1283.

(25) Palo, D. R.; Dagle, R. A.; Holladay, J. D. Chem. Rev. 2007, 107 (10), 3992–4021.

(26) Haryanto, A.; Fernando, S.; Murali, N.; Adhikari, S. Energy Fuels 2005, 19 (5), 2098–2106.

(27) Wüstenhagen, R.; Wolsink, M.; Bürer, M. J. Energy Policy 2007, 35 (5), 2683– 2691.

(28) Barbir, F.; Gómez, T. Int. J. Hydrog. Energy 1997, 22 (10), 1027–1037.

(29) Barbir, F. In PEM Fuel Cells; Academic Press: Burlington, 2005; pp 17–32.

(30) Grigoriev, S. A.; Porembsky, V. I.; Fateev, V. N. Int. J. Hydrog. Energy 2006, 31 (2), 171–175.

(31) Marshall, A. T.; Sunde, S.; Tsypkin, M.; Tunold, R. Int. J. Hydrog. Energy 2007, 32 (13), 2320–2324.

(32) Ayers, K. E.; Anderson, E. B.; Capuano, C.; Carter, B.; Dalton, L.; Hanlon, G.; Manco, J.; Niedzwiecki, M. ECS Trans. 2010, 33 (1), 3–15.

(33) Züttel, A. Mater. Today 2003, 6 (9), 24–33.

158

(34) Sakintuna, B.; Lamari-Darkrim, F.; Hirscher, M. Int. J. Hydrog. Energy 2007, 32 (9), 1121–1140.

(35) Jena, P. J. Phys. Chem. Lett. 2011, 2 (3), 206–211.

(36) Durbin, D. J.; Malardier-Jugroot, C. Int. J. Hydrog. Energy 2013, 38 (34), 14595– 14617.

(37) Zacharia, R.; Rather, S. ullah. J. Nanomater. 2015, 2015, 1–18.

(38) Scharber, M. C.; Sariciftci, N. S. Prog. Polym. Sci. 2013, 38 (12), 1929–1940.

(39) Deng, Y.; Peng, E.; Shao, Y.; Xiao, Z.; Dong, Q.; Huang, J. Energy Environ. Sci. 2015, 8 (5), 1544–1550.

(40) Christopher, K.; Dimitrios, R. Energy Environ. Sci. 2012, 5 (5), 6640–6651.

(41) Benemann, J. R. Int. J. Hydrog. Energy 1997, 22 (10), 979–987.

(42) Scholta, J.; Berg, N.; Wilde, P.; Jörissen, L.; Garche, J. J. Power Sources 2004, 127 (1–2), 206–212.

(43) Peighambardoust, S. J.; Rowshanzamir, S.; Amjadi, M. Int. J. Hydrog. Energy 2010, 35 (17), 9349–9384.

(44) Wang, Y.; Chen, K. S.; Mishler, J.; Cho, S. C.; Adroher, X. C. Appl. Energy 2011, 88 (4), 981–1007.

(45) Holton, O. T.; Stevenson, J. W. Platin. Met. Rev. 2013, 57 (4), 259–271.

(46) Kraytsberg, A.; Ein-Eli, Y. Energy Fuels 2014, 28 (12), 7303–7330.

(47) Holdcroft, S. Chem. Mater. 2014, 26 (1), 381–393.

(48) Antolini, E.; Zignani, S. C.; Santos, S. F.; Gonzalez, E. R. Electrochimica Acta 2011, 56 (5), 2299–2305.

(49) Rabis, A.; Rodriguez, P.; Schmidt, T. J. ACS Catal. 2012, 2 (5), 864–890.

(50) Keeley, G. P.; Cherevko, S.; Mayrhofer, K. J. J. ChemElectroChem 2016, 3 (1), 51– 54.

(51) Jasinski, R. Nature 1964, 201 (4925), 1212–1213.

(52) Zagal, J. H.; Griveau, S.; Silva, J. F.; Nyokong, T.; Bedioui, F. Coord. Chem. Rev. 2010, 254 (23–24), 2755–2791.

159

(53) Serov, A.; Artyushkova, K.; Atanassov, P. Adv. Energy Mater. 2014, 4 (10), 1301735.

(54) Strickland, K.; Miner, E.; Jia, Q.; Tylus, U.; Ramaswamy, N.; Liang, W.; Sougrati, M.-T.; Jaouen, F.; Mukerjee, S. Nat. Commun. 2015, 6, 7343.

(55) Gamburzev, S.; Petrov, K.; Appleby, A. J. J. Appl. Electrochem. 2002, 32 (7), 805– 809.

(56) Lu, S.; Pan, J.; Huang, A.; Zhuang, L.; Lu, J. Proc. Natl. Acad. Sci. 2008, 105 (52), 20611–20614.

(57) Zhuang, Z.; Giles, S. A.; Zheng, J.; Jenness, G. R.; Caratzoulas, S.; Vlachos, D. G.; Yan, Y. Nat. Commun. 2016, 7, 10141.

(58) Varcoe, J. R.; Slade, R. C. T. Fuel Cells 2005, 5 (2), 187–200.

(59) McLean, G. F.; Niet, T.; Prince-Richard, S.; Djilali, N. Int. J. Hydrog. Energy 2002, 27 (5), 507–526.

(60) Xing, B.; Savadogo, O. Electrochem. Commun. 2000, 2 (10), 697–702.

(61) Jensen, J. O.; Aili, D.; Hansen, M. K.; Li, Q.; Bjerrum, N. J.; Christensen, E. ECS Trans. 2014, 64 (3), 1175–1184.

(62) Aili, D.; Jankova, K.; Han, J.; Bjerrum, N. J.; Jensen, J. O.; Li, Q. Polymer 2016, 84, 304–310.

(63) Cifrain, M.; Kordesch, K. V. J. Power Sources 2004, 127 (1–2), 234–242.

(64) Gülzow, E.; Schulze, M. J. Power Sources 2004, 127 (1–2), 243–251.

(65) Khalidi, A.; Lafage, B.; Taxil, P.; Gave, G.; Clifton, M. J.; Cézac, P. Int. J. Hydrog. Energy 1996, 21 (1), 25–31.

(66) Agel, E.; Bouet, J.; Fauvarque, J. F. J. Power Sources 2001, 101 (2), 267–274.

(67) Bauer, B.; Strathmann, H.; Effenberger, F. Desalination 1990, 79 (2), 125–144.

(68) Sata, T.; Tsujimoto, M.; Yamaguchi, T.; Matsusaki, K. J. Membr. Sci. 1996, 112 (2), 161–170.

(69) Varcoe, J. R.; Atanassov, P.; Dekel, D. R.; Herring, A. M.; Hickner, M. A.; Kohl, P. A.; Kucernak, A. R.; Mustain, W. E.; Nijmeijer, K.; Scott, K.; Xu, T.; Zhuang, L. Energy Environ. Sci. 2014, 7 (10), 3135–3191.

160

(70) McKone, J. R.; Lewis, N. S.; Gray, H. B. Chem. Mater. 2014, 26 (1), 407–414.

(71) Zhang, Q.; Li, S.; Zhang, S. Chem. Commun. 2010, 46 (40), 7495–7497.

(72) Faraj, M.; Elia, E.; Boccia, M.; Filpi, A.; Pucci, A.; Ciardelli, F. J. Polym. Sci. Part Polym. Chem. 2011, 49 (15), 3437–3447.

(73) Zhang, B.; Gu, S.; Wang, J.; Liu, Y.; Herring, A. M.; Yan, Y. RSC Adv. 2012, 2 (33), 12683–12685.

(74) Marino, M. G.; Kreuer, K. D. ChemSusChem 2015, 8 (3), 513–523.

(75) Gu, F.; Dong, H.; Li, Y.; Sun, Z.; Yan, F. Macromolecules 2014, 47 (19), 6740– 6747.

(76) Guo, M.; Fang, J.; Xu, H.; Li, W.; Lu, X.; Lan, C.; Li, K. J. Membr. Sci. 2010, 362 (1–2), 97–104.

(77) Jiang, Y.; Freyer, J. L.; Cotanda, P.; Brucks, S. D.; Killops, K. L.; Bandar, J. S.; Torsitano, C.; Balsara, N. P.; Lambert, T. H.; Campos, L. M. Nat. Commun. 2015, 6, 5950.

(78) Zha, Y.; Disabb-Miller, M. L.; Johnson, Z. D.; Hickner, M. A.; Tew, G. N. J. Am. Chem. Soc. 2012, 134 (10), 4493–4496.

(79) Gu, S.; Cai, R.; Luo, T.; Chen, Z.; Sun, M.; Liu, Y.; He, G.; Yan, Y. Angew. Chem. Int. Ed. 2009, 48 (35), 6499–6502.

(80) Matsui, K.; Tobita, E.; Sugimoto, K.; Kondo, K.; Seita, T.; Akimoto, A. J. Appl. Polym. Sci. 1986, 32 (3), 4137–4143.

(81) Xiong, Y.; Fang, J.; Zeng, Q. H.; Liu, Q. L. J. Membr. Sci. 2008, 311 (1–2), 319– 325.

(82) Yanagi, H.; Fukuta, K. ECS Trans. 2008, 16 (2), 257–262.

(83) Hibbs, M. R.; Hickner, M. A.; Alam, T. M.; McIntyre, S. K.; Fujimoto, C. H.; Cornelius, C. J. Chem. Mater. 2008, 20 (7), 2566–2573.

(84) Hibbs, M.; Fujimoto, C.; Cornelius, C. ECS Trans. 2009, 19 (30), 89–97.

(85) Hibbs, M. R.; Fujimoto, C. H.; Cornelius, C. J. Macromolecules 2009, 42 (21), 8316–8321.

(86) Slade, R. C. T.; Varcoe, J. R. Solid State Ion. 2005, 176 (5–6), 585–597.

161

(87) Varcoe, J. R.; Slade, R. C. T.; Yee, E. L. H. Chem. Commun. 2006, No. 13, 1428– 1429.

(88) Varcoe, J. R.; Slade, R. C. T.; Lam How Yee, E.; Poynton, S. D.; Driscoll, D. J.; Apperley, D. C. Chem. Mater. 2007, 19 (10), 2686–2693.

(89) Yan, J.; Hickner, M. A. Macromolecules 2010, 43 (5), 2349–2356.

(90) Chen, D.; Hickner, M. A. ACS Appl. Mater. Interfaces 2012, 4 (11), 5775–5781.

(91) Li, N.; Leng, Y.; Hickner, M. A.; Wang, C.-Y. J. Am. Chem. Soc. 2013, 135 (27), 10124–10133.

(92) Li, N.; Wang, L.; Hickner, M. Chem. Commun. 2014, 50 (31), 4092–4095.

(93) Wang, L.; Hickner, M. A. Polym. Chem. 2014, 5 (8), 2928–2935.

(94) Zhu, L.; Pan, J.; Wang, Y.; Han, J.; Zhuang, L.; Hickner, M. A. Macromolecules 2016, 49 (3), 815–824.

(95) Dang, H.-S.; Weiber, E. A.; Jannasch, P. J. Mater. Chem. A 2015, 3 (10), 5280– 5284.

(96) Jannasch, P. ECS Trans. 2015, 69 (17), 369–375.

(97) Weiber, E. A.; Meis, D.; Jannasch, P. Polym. Chem. 2015, 6 (11), 1986–1996.

(98) Tsai, T.-H.; Ertem, S. P.; Maes, A. M.; Seifert, S.; Herring, A. M.; Coughlin, E. B. Macromolecules 2015, 48 (3), 655–662.

(99) Pandey, T. P.; Sarode, H. N.; Yang, Y.; Yang, Y.; Vezzù, K.; Noto, V. D.; Seifert, S.; Knauss, D. M.; Liberatore, M. W.; Herring, A. M. J. Electrochem. Soc. 2016, 163 (7), H513–H520.

(100) Arges, C. G.; Wang, L.; Jung, M.; Ramani, V. J. Electrochem. Soc. 2015, 162 (7), F686–F693.

(101) Parrondo, J.; Jung, M. J.; Wang, Z.; Arges, C. G.; Ramani, V. J. Electrochem. Soc. 2015, 162 (10), F1236–F1242.

(102) Wang, J.; Zhao, Z.; Gong, F.; Li, S.; Zhang, S. Macromolecules 2009, 42 (22), 8711–8717.

(103) Zhou, J.; Unlu, M.; Vega, J. A.; Kohl, P. A. J. Power Sources 2009, 190 (2), 285– 292.

162

(104) Lee, K. M.; Wycisk, R.; Litt, M.; Pintauro, P. N. J. Membr. Sci. 2011, 383 (1–2), 254–261.

(105) Tanaka, M.; Koike, M.; Miyatake, K.; Watanabe, M. Macromolecules 2010, 43 (6), 2657–2659.

(106) Tanaka, M.; Koike, M.; Miyatake, K.; Watanabe, M. Polym. Chem. 2011, 2 (1), 99–106.

(107) Yokota, N.; Ono, H.; Miyake, J.; Nishino, E.; Asazawa, K.; Watanabe, M.; Miyatake, K. ACS Appl. Mater. Interfaces 2014, 6 (19), 17044–17052.

(108) Yokota, N.; Shimada, M.; Ono, H.; Akiyama, R.; Nishino, E.; Asazawa, K.; Miyake, J.; Watanabe, M.; Miyatake, K. Macromolecules 2014, 47 (23), 8238–8246.

(109) Zhang, M.; Liu, J.; Wang, Y.; An, L.; Guiver, M. D.; Li, N. J. Mater. Chem. A 2015, 3 (23), 12284–12296.

(110) Mohanty, A. D.; Ryu, C. Y.; Kim, Y. S.; Bae, C. Macromolecules 2015, 48 (19), 7085–7095.

(111) Meek, K. M.; Elabd, Y. A. Macromolecules 2015, 48 (19), 7071–7084.

(112) Cao, Y.-C.; Wu, X.; Scott, K. Int. J. Hydrog. Energy 2012, 37 (12), 9524–9528.

(113) Wu, X.; Scott, K. J. Power Sources 2012, 214, 124–129.

(114) Faraj, M.; Boccia, M.; Miller, H.; Martini, F.; Borsacchi, S.; Geppi, M.; Pucci, A. Int. J. Hydrog. Energy 2012, 37 (20), 14992–15002.

(115) Pan, J.; Li, Y.; Zhuang, L.; Lu, J. Chem. Commun. 2010, 46 (45), 8597–8599.

(116) Ong, A. L.; Saad, S.; Lan, R.; Goodfellow, R. J.; Tao, S. J. Power Sources 2011, 196 (20), 8272–8279.

(117) Pan, J.; Chen, C.; Zhuang, L.; Lu, J. Acc. Chem. Res. 2012, 45 (3), 473–481.

(118) Wang, J.; Wang, J.; Zhang, S. J. Membr. Sci. 2012, 415–416, 205–212.

(119) Choi, J.; Byun, Y.-J.; Lee, S. Y.; Jang, J. H.; Henkensmeier, D.; Yoo, S. J.; Hong, S.-A.; Kim, H.-J.; Sung, Y.-E.; Park, J.-S. Int. J. Hydrog. Energy 2014, 39 (36), 21223–21230.

(120) Ren, X.; Price, S. C.; Jackson, A. C.; Pomerantz, N.; Beyer, F. L. ACS Appl. Mater. Interfaces 2014, 6 (16), 13330–13333.

163

(121) Yang, Y.; Knauss, D. M. Macromolecules 2015, 48 (13), 4471–4480.

(122) Wang, J.; Li, S.; Zhang, S. Macromolecules 2010, 43 (8), 3890–3896.

(123) Kim, D. S.; Labouriau, A.; Guiver, M. D.; Kim, Y. S. Chem. Mater. 2011, 23 (17), 3795–3797.

(124) Qu, C.; Zhang, H.; Zhang, F.; Liu, B. J. Mater. Chem. 2012, 22 (17), 8203–8207.

(125) Lin, X.; Wu, L.; Liu, Y.; Ong, A. L.; Poynton, S. D.; Varcoe, J. R.; Xu, T. J. Power Sources 2012, 217, 373–380.

(126) Sherazi, T. A.; Zahoor, S.; Raza, R.; Shaikh, A. J.; Naqvi, S. A. R.; Abbas, G.; Khan, Y.; Li, S. Int. J. Hydrog. Energy 2015, 40 (1), 786–796.

(127) Sajjad, S. D.; Liu, D.; Wei, Z.; Sakri, S.; Shen, Y.; Hong, Y.; Liu, F. J. Power Sources 2015, 300, 95–103.

(128) Lan, C.; Fang, J.; Guan, Y.; Zhou, H.; Zhao, J. J. Power Sources 2015, 295, 259–267.

(129) Ye, Y.; Elabd, Y. A. Macromolecules 2011, 44 (21), 8494–8503.

(130) Deavin, O. I.; Murphy, S.; Ong, A. L.; Poynton, S. D.; Zeng, R.; Herman, H.; Varcoe, J. R. Energy Environ. Sci. 2012, 5 (9), 8584–8597.

(131) Page, O. M. M.; Poynton, S. D.; Murphy, S.; Ong, A. L.; Hillman, D. M.; Hancock, C. A.; Hale, M. G.; Apperley, D. C.; Varcoe, J. R. RSC Adv. 2012, 3 (2), 579–587.

(132) Lin, B.; Dong, H.; Li, Y.; Si, Z.; Gu, F.; Yan, F. Chem. Mater. 2013, 25 (9), 1858– 1867.

(133) Gu, F.; Dong, H.; Li, Y.; Si, Z.; Yan, F. Macromolecules 2014, 47 (1), 208–216.

(134) Lin, B.; Chu, F.; Ren, Y.; Jia, B.; Yuan, N.; Shang, H.; Feng, T.; Zhu, Y.; Ding, J. J. Power Sources 2014, 266, 186–192.

(135) Liu, Y.; Wang, J.; Yang, Y.; Brenner, T. M.; Seifert, S.; Yan, Y.; Liberatore, M. W.; Herring, A. M. J. Phys. Chem. C 2014, 118 (28), 15136–15145.

(136) Si, Z.; Qiu, L.; Dong, H.; Gu, F.; Li, Y.; Yan, F. ACS Appl. Mater. Interfaces 2014, 6 (6), 4346–4355.

(137) Yang, Y.; Wang, J.; Zheng, J.; Li, S.; Zhang, S. J. Membr. Sci. 2014, 467, 48–55.

(138) Weiber, E. A.; Jannasch, P. J. Membr. Sci. 2015, 481, 164–171.

164

(139) Lu, D.; Wen, L.; Nie, F.; Xue, L. RSC Adv. 2016, 6 (8), 6029–6037.

(140) Hu, M.; Pearce, E. M.; Kwei, T. K. J. Polym. Sci. Part Polym. Chem. 1993, 31 (2), 553–561.

(141) Thomas, O. D.; Soo, K. J. W. Y.; Peckham, T. J.; Kulkarni, M. P.; Holdcroft, S. Polym. Chem. 2011, 2 (8), 1641–1643.

(142) Henkensmeier, D.; Kim, H.-J.; Lee, H.-J.; Lee, D. H.; Oh, I.-H.; Hong, S.-A.; Nam, S.-W.; Lim, T.-H. Macromol. Mater. Eng. 2011, 296 (10), 899–908.

(143) Henkensmeier, D.; Cho, H.-R.; Kim, H.-J.; Nunes Kirchner, C.; Leppin, J.; Dyck, A.; Jang, J. H.; Cho, E.; Nam, S.-W.; Lim, T.-H. Polym. Degrad. Stab. 2012, 97 (3), 264–272.

(144) Thomas, O. D.; Soo, K. J. W. Y.; Peckham, T. J.; Kulkarni, M. P.; Holdcroft, S. J. Am. Chem. Soc. 2012, 134 (26), 10753–10756.

(145) Lin, X.; Liang, X.; Poynton, S. D.; Varcoe, J. R.; Ong, A. L.; Ran, J.; Li, Y.; Li, Q.; Xu, T. J. Membr. Sci. 2013, 443, 193–200.

(146) Henkensmeier, D.; Cho, H.; Brela, M.; Michalak, A.; Dyck, A.; Germer, W.; Duong, N. M. H.; Jang, J. H.; Kim, H.-J.; Woo, N.-S.; Lim, T.-H. Int. J. Hydrog. Energy 2014, 39 (6), 2842–2853.

(147) Wright, A. G.; Holdcroft, S. ACS Macro Lett. 2014, 3 (5), 444–447.

(148) Lee, J. Y.; Lim, D.-H.; Chae, J. E.; Choi, J.; Kim, B. H.; Lee, S. Y.; Yoon, C. W.; Nam, S. Y.; Jang, J. H.; Henkensmeier, D.; Yoo, S. J.; Kim, J.-Y.; Kim, H.-J.; Ham, H. C. J. Membr. Sci. 2016, 514, 398–406.

(149) Wright, A. G.; Weissbach, T.; Holdcroft, S. Angew. Chem. Int. Ed. 2016, 55 (15), 4818–4821.

(150) Wright, A. G.; Fan, J.; Britton, B.; Weissbach, T.; Lee, H.-F.; Kitching, E. A.; Peckham, T. J.; Holdcroft, S. Energy Environ. Sci. 2016, 9 (6), 2130–2142.

(151) Mudraboyina, B. P.; Obadia, M. M.; Allaoua, I.; Sood, R.; Serghei, A.; Drockenmuller, E. Chem. Mater. 2014, 26 (4), 1720–1726.

(152) Sood, R.; Zhang, B.; Serghei, A.; Bernard, J.; Drockenmuller, E. Polym. Chem. 2015, 6 (18), 3521–3528.

(153) Liu, L.; He, S.; Zhang, S.; Zhang, M.; Guiver, M. D.; Li, N. ACS Appl. Mater. Interfaces 2016, 8 (7), 4651–4660.

165

(154) Noonan, K. J. T.; Hugar, K. M.; Kostalik, H. A.; Lobkovsky, E. B.; Abruña, H. D.; Coates, G. W. J. Am. Chem. Soc. 2012, 134 (44), 18161–18164.

(155) Ye, Y.; Stokes, K. K.; Beyer, F. L.; Elabd, Y. A. J. Membr. Sci. 2013, 443, 93–99.

(156) Yan, X.; Gu, S.; He, G.; Wu, X.; Zheng, W.; Ruan, X. J. Membr. Sci. 2014, 466, 220–228.

(157) Womble, C. T.; Coates, G. W.; Matyjaszewski, K.; Noonan, K. J. T. ACS Macro Lett. 2016, 5, 253–257.

(158) Hossain, M. A.; Jang, H.; Sutradhar, S. C.; Ha, J.; Yoo, J.; Lee, C.; Lee, S.; Kim, W. Int. J. Hydrog. Energy 2016, 41 (24), 10458–10465.

(159) Gu, S.; Wang, J.; Kaspar, R. B.; Fang, Q.; Zhang, B.; Coughlin, E. B.; Yan, Y. Sci. Rep. 2015, 5, 11668.

(160) Mohanty, A. D.; Bae, C. J. Mater. Chem. A 2014, 2 (41), 17314–17320.

(161) Couture, G.; Alaaeddine, A.; Boschet, F.; Ameduri, B. Prog. Polym. Sci. 2011, 36 (11), 1521–1557.

(162) Fujimoto, C.; Kim, D.-S.; Hibbs, M.; Wrobleski, D.; Kim, Y. S. J. Membr. Sci. 2012, 423–424, 438–449.

(163) Arges, C. G.; Ramani, V. Proc. Natl. Acad. Sci. 2013, 110 (7), 2490–2495.

(164) Mohanty, A. D.; Tignor, S. E.; Krause, J. A.; Choe, Y.-K.; Bae, C. Macromolecules 2016, 49 (9), 3361–3372.

(165) Merle, G.; Wessling, M.; Nijmeijer, K. J. Membr. Sci. 2011, 377 (1–2), 1–35.

(166) Hickner, M. A.; Herring, A. M.; Coughlin, E. B. J. Polym. Sci. Part B Polym. Phys. 2013, 51 (24), 1727–1735.

(167) Hibbs, M. R. J. Polym. Sci. Part B Polym. Phys. 2013, 51 (24), 1736–1742.

(168) Miyake, J.; Fukasawa, K.; Watanabe, M.; Miyatake, K. J. Polym. Sci. Part Polym. Chem. 2014, 52 (3), 383–389.

(169) Zhang, Z.; Wu, L.; Varcoe, J.; Li, C.; Ong, A. L.; Poynton, S.; Xu, T. J. Mater. Chem. A 2013, 1 (7), 2595–2601.

(170) Zhou, J.; Ünlü, M.; Anestis-Richard, I.; Kohl, P. A. J. Membr. Sci. 2010, 350 (1– 2), 286–292.

166

(171) Gu, S.; Cai, R.; Yan, Y. Chem. Commun. 2011, 47 (10), 2856–2858.

(172) Luo, Y.; Guo, J.; Wang, C.; Chu, D. Electrochem. Commun. 2012, 16 (1), 65–68.

(173) Wu, L.; Pan, Q.; Varcoe, J. R.; Zhou, D.; Ran, J.; Yang, Z.; Xu, T. J. Membr. Sci. 2015, 490, 1–8.

(174) Carmo, M.; Doubek, G.; Sekol, R. C.; Linardi, M.; Taylor, A. D. J. Power Sources 2013, 230, 169–175.

(175) Carmo, M.; Fritz, D. L.; Mergel, J.; Stolten, D. Int. J. Hydrog. Energy 2013, 38 (12), 4901–4934.

(176) Kreuer, K.-D. Chem. Mater. 2014, 26 (1), 361–380.

(177) Krol, J. J.; Wessling, M.; Strathmann, H. J. Membr. Sci. 1999, 162 (1–2), 145– 154.

(178) Brinker, K. C.; Robinson, I. M. Polybenzimidazoles. US2895948 A, July 21, 1959.

(179) Vogel, H.; Marvel, C. S. J. Polym. Sci. 1961, 50 (154), 511–539.

(180) Chung, T.-S. J. Macromol. Sci. Part C 1997, 37 (2), 277–301.

(181) Neuse, E. W. In Synthesis and Degradation Rheology and Extrusion; Advances in Polymer Science; Springer Berlin Heidelberg, 1982; pp 1–42.

(182) Wainright, J. S.; Wang, J.-T.; Weng, D.; Savinell, R. F.; Litt, M. J. Electrochem. Soc. 1995, 142 (7), L121–L123.

(183) Marvel, C. S.; Vogel, H. A. Polybenzimidazoles and their preparation. US3174947 A, March 23, 1965.

(184) Choe, E.-W. J. Appl. Polym. Sci. 1994, 53 (5), 497–506.

(185) Jones, D. J.; Rozière, J. J. Membr. Sci. 2001, 185 (1), 41–58.

(186) Iwakura, Y.; Uno, K.; Imai, Y. J. Polym. Sci. A 1964, 2 (6), 2605–2615.

(187) Sansone, M. J. N-substituted polybenzimidazole polymer. US4898917 A, February 6, 1990.

(188) Sansone, M. J.; Gupta, B.; Forbes, C. E.; Kwiatek, M. S. Sulfalkylation of hydroxyethylated polybenzimidazole polymers. US4997892 A, March 5, 1991.

(189) Gieselman, M. B.; Reynolds, J. R. Macromolecules 1992, 25 (18), 4832–4834.

167

(190) Pu, H.; Liu, G. Polym. Int. 2005, 54 (1), 175–179.

(191) Klaehn, J. R.; Luther, T. A.; Orme, C. J.; Jones, M. G.; Wertsching, A. K.; Peterson, E. S. Macromolecules 2007, 40 (21), 7487–7492.

(192) Kumbharkar, S. C.; Kharul, U. K. Eur. Polym. J. 2009, 45 (12), 3363–3371.

(193) Kumbharkar, S. C.; Kharul, U. K. J. Membr. Sci. 2010, 357 (1–2), 134–142.

(194) Rembaum, A.; Baumgartner, W.; Eisenberg, A. J. Polym. Sci. [B] 1968, 6 (3), 159–171.

(195) Moore, R. B.; Martin, C. R. Macromolecules 1989, 22 (9), 3594–3599.

(196) Yang, Y.; Holdcroft, S. Fuel Cells 2005, 5 (2), 171–186.

(197) Fujimura, M.; Hashimoto, T.; Kawai, H. Macromolecules 1981, 14 (5), 1309– 1315.

(198) Baessler, K.; Schubert, H. Process for the purification of crude 3,4,3’,4’- tetraaminodiphenyl. US4433168 A, February 21, 1984.

(199) Beringer, F. M.; Sands, S. J. Am. Chem. Soc. 1953, 75 (14), 3319–3322.

(200) Avila-Zárraga, J. G.; Martínez, R. Synth. Commun. 2001, 31 (14), 2177–2183.

(201) Schubert, W. M. J. Am. Chem. Soc. 1949, 71 (8), 2639–2644.

(202) Eaton, P. E.; Carlson, G. R.; Lee, J. T. J. Org. Chem. 1973, 38 (23), 4071–4073.

(203) Ueda, M.; Sato, M.; Mochizuki, A. Macromolecules 1985, 18 (12), 2723–2726.

(204) Disabb-Miller, M. L.; Zha, Y.; DeCarlo, A. J.; Pawar, M.; Tew, G. N.; Hickner, M. A. Macromolecules 2013, 46 (23), 9279–9287.

(205) Haynes, W. M. CRC Handbook of Chemistry and Physics, 94th Edition; Taylor & Francis Limited, 2013.

(206) Leng, Y.; Chen, G.; Mendoza, A. J.; Tighe, T. B.; Hickner, M. A.; Wang, C.-Y. J. Am. Chem. Soc. 2012, 134 (22), 9054–9057.

(207) Petrowsky, M.; Frech, R. J. Phys. Chem. B 2009, 113 (17), 5996–6000.

(208) Marino, M. G.; Melchior, J. P.; Wohlfarth, A.; Kreuer, K. D. J. Membr. Sci. 2014, 464, 61–71.

168

(209) Li, Q.; He, R.; Jensen, J. O.; Bjerrum, N. j. Fuel Cells 2004, 4 (3), 147–159.

(210) Dai, H.; Zhang, H.; Luo, Q.; Zhang, Y.; Bi, C. J. Power Sources 2008, 185 (1), 19–25.

(211) Ong, A. L.; Inglis, K. K.; Whelligan, D. K.; Murphy, S.; Varcoe, J. R. Phys. Chem. Chem. Phys. 2015, 17 (18), 12135–12145.

(212) Danilovic, N.; Subbaraman, R.; Strmcnik, D.; Paulikas, A. P.; Myers, D.; Stamenkovic, V. R.; Markovic, N. M. Electrocatalysis 2012, 3 (3–4), 221–229.

(213) Britton, B.; Holdcroft, S. J. Electrochem. Soc. 2016, 163 (5), F353–F358.

(214) Su, H.; Bladergroen, B. J.; Linkov, V.; Pasupathi, S.; Ji, S. Int. J. Hydrog. Energy 2011, 36 (23), 15081–15088.

(215) Strmcnik, D.; Li, D.; Lopes, P. P.; Tripkovic, D.; Kodama, K.; Stamenkovic, V. R.; Markovic, N. M. Top. Catal. 2015, 58 (18–20), 1174–1180.

(216) Kohler, E. P.; Blanchard, L. W. J. Am. Chem. Soc. 1935, 57 (2), 367–371.

(217) Heiss, C.; Marzi, E.; Schlosser, M. Eur. J. Org. Chem. 2003, 2003 (23), 4625– 4629.

(218) Wassmundt, F. W.; Kiesman, W. F. J. Org. Chem. 1995, 60 (6), 1713–1719.

(219) Taylor, E. C.; Kalenda, N. W. J. Am. Chem. Soc. 1954, 76 (6), 1699–1700.

(220) So, Y.-H.; Heeschen, J. P. J. Org. Chem. 1997, 62 (11), 3552–3561.

(221) Romain, M.; Chevrier, M.; Bebiche, S.; Mohammed-Brahim, T.; Rault-Berthelot, J.; Jacques, E.; Poriel, C. J. Mater. Chem. C 2015, 3 (22), 5742–5753.

(222) Yin, J.; Rainka, M. P.; Zhang, X.-X.; Buchwald, S. L. J. Am. Chem. Soc. 2002, 124 (7), 1162–1163.

(223) Walker, S. D.; Barder, T. E.; Martinelli, J. R.; Buchwald, S. L. Angew. Chem. Int. Ed. 2004, 43 (14), 1871–1876.

(224) Mahanty, J. S.; De, M.; Kundu, N. G. J. Chem. Soc. [Perkin 1] 1997, No. 17, 2577–2580.

(225) Ghosez, L.; Franc, C.; Denonne, F.; Cuisinier, C.; Touillaux, R. Can. J. Chem. 2001, 79 (11), 1827–1839.

169

(226) Handy, S. T.; Bregman, H.; Lewis, J.; Zhang, X.; Zhang, Y. Tetrahedron Lett. 2003, 44 (3), 427–430.

(227) Navarro, O.; Kelly, R. A.; Nolan, S. P. J. Am. Chem. Soc. 2003, 125 (52), 16194– 16195.

(228) Navarro, O.; Marion, N.; Oonishi, Y.; Kelly, R. A.; Nolan, S. P. J. Org. Chem. 2006, 71 (2), 685–692.

(229) Hübschle, C. B.; Sheldrick, G. M.; Dittrich, B. J. Appl. Crystallogr. 2011, 44 (Pt 6), 1281–1284.

(230) Macrae, C. F.; Bruno, I. J.; Chisholm, J. A.; Edgington, P. R.; McCabe, P.; Pidcock, E.; Rodriguez-Monge, L.; Taylor, R.; Streek, J. van de; Wood, P. A. J. Appl. Crystallogr. 2008, 41 (2), 466–470.

(231) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian∼09 Revision D.01; Gaussian Inc. Wallingford CT, 2009.

(232) Long, H.; Pivovar, B. J. Phys. Chem. C 2014, 118 (19), 9880–9888.

(233) Grushin, V. V.; Alper, H. Chem. Rev. 1994, 94 (4), 1047–1062.

(234) Grushin, V. V.; Alper, H. In Activation of Unreactive Bonds and Organic Synthesis; Murai, P. S., Alper, H., Gossage, R. A., Grushin, V. V., Hidai, M., Ito, Y., Jones, W. D., Kakiuchi, F., Koten, G. van, Lin, Y.-S., Mizobe, Y., Murai, S., Murakami, M., Richmond, T. G., Sen, A., Suginome, M., Yamamoto, A., Eds.; Topics in Organometallic Chemistry; Springer Berlin Heidelberg, 1999; pp 193–226.

(235) Foà, M.; Cassar, L. J. Chem. Soc. Dalton Trans. 1975, No. 23, 2572–2576.

(236) Fahey, D. R. J. Am. Chem. Soc. 1970, 92 (2), 402–404.

(237) Parshall, G. W. J. Am. Chem. Soc. 1974, 96 (8), 2360–2366.

170

(238) Tsou, T. T.; Kochi, J. K. J. Am. Chem. Soc. 1979, 101 (21), 6319–6332.

(239) Tsou, T. T.; Kochi, J. K. J. Am. Chem. Soc. 1979, 101 (25), 7547–7560.

(240) Kende, A. S.; Liebeskind, L. S.; Braitsch, D. M. Tetrahedron Lett. 1975, 16 (39), 3375–3378.

(241) Colon, I.; Kelsey, D. R. J. Org. Chem. 1986, 51 (14), 2627–2637.

(242) Colon, I.; Kwiatkowski, G. T. J. Polym. Sci. Part Polym. Chem. 1990, 28 (2), 367– 383.

(243) Yamamoto, T.; Morita, A.; Miyazaki, Y.; Maruyama, T.; Wakayama, H.; Zhou, Z. H.; Nakamura, Y.; Kanbara, T.; Sasaki, S.; Kubota, K. Macromolecules 1992, 25 (4), 1214–1223.

(244) Kreyenschmidt, M.; Uckert, F.; Müllen, K. Macromolecules 1995, 28 (13), 4577– 4582.

(245) Yamamoto, T.; Sugiyama, K.; Kanbara, T.; Hayashi, H.; Etori, H. Macromol. Chem. Phys. 1998, 199 (9), 1807–1813.

(246) Yurteri, S.; Cianga, I.; Degirmenci, M.; Yagci, Y. Polym. Int. 2004, 53 (9), 1219– 1225.

(247) He, B.; Tian, H.; Geng, Y.; Wang, F.; Müllen, K. Org. Lett. 2008, 10 (5), 773–776.

(248) Rosen, B. M.; Quasdorf, K. W.; Wilson, D. A.; Zhang, N.; Resmerita, A.-M.; Garg, N. K.; Percec, V. Chem. Rev. 2011, 111 (3), 1346–1416.

(249) Ballauff, M. Angew. Chem. Int. Ed. Engl. 1989, 28 (3), 253–267.

(250) Miyahara, T.; Miyake, J.; Matsuno, S.; Watanabe, M.; Miyatake, K. RSC Adv. 2015, 5 (62), 50082–50086.

(251) Semmelhack, M. F.; Helquist, P. M.; Jones, L. D. J. Am. Chem. Soc. 1971, 93 (22), 5908–5910.

(252) Yuan, Y.; Johnson, F.; Cabasso, I. J. Appl. Polym. Sci. 2009, 112 (6), 3436– 3441.

(253) Hugar, K. M.; Kostalik, H. A.; Coates, G. W. J. Am. Chem. Soc. 2015, 137 (27), 8730–8737.

171

Appendix A.

Supporting Information for Chapter 2

Unpublished NMR Spectra for New Compounds:

1 Figure A1. The H NMR spectrum (500 MHz, DMSO-d6 + 5 drops of 40 wt% KOD in D2O) of HMT-B.

1 Figure A2. The H NMR spectrum (400 MHz, DMSO-d6) of HMT-MB.

172

1 Figure A3. The H NMR spectrum (500 MHz, DMSO-d6 + 5 drops of 40 wt% KOD in D2O) of DMB.

1 Figure A4. The H NMR spectrum (500 MHz, DMSO-d6) of DMMB.

173

Appendix B.

Supporting Information for Chapter 3

Table B1. Yield of BMA for each reaction performed. Reaction # Yield (g) Yield (%) 1 587.6 88.2 2 577.5 86.6

Table B2. Yield of BME and respective reaction scale for each reaction performed. Reaction # Scale Yield (g) Yield (%) 1 1.0 106.1 96.4 2+3 2.0 218.4 99.2 4+5 2.0 214.8 97.5 6+7 2.0 219.2 99.5 8+9 2.0 218.9 99.4 10+11 2.2 246.3 99.4

Table B3. Yield of HMTE, reaction scale, and amount of Pd(PPh3)4 used for each reaction performed. Pd(PPh ) used (per Reaction # Scale 3 4 Yield (g) Yield (%) mol% BME) 1 0.80 2.92 g (0.54%) 55.6 55.3 2 1.00 1.80 g (0.27%) 68.0 54.1 3 1.00 1.70 g (0.25%) 70.8 56.4 4 1.00 1.51 g (0.22%) 72.1 57.4 5 1.00 1.01 g (0.15%) 72.2 57.5 6 1.00 0.79 g (0.12%) 68.9 54.9 7 1.00 0.53 g (0.08%) 71.9 57.3 8 1.17 0.77 g (0.10%) 86.3 58.7

174

Table B4. Yield of DAB after recrystallization and the respective appearance. DAB used Reaction # Companya Yield (g) Yield (%) Appearance (g) 1 1 50.0 45.0 90.0 white/sandy sheets 2 1 50.0 42.8 85.6 white/sandy sheets long-pointed 3 2 50.0 42.7 85.4 sandy-sheets 4 2 50.0 42.3 84.6 large sandy-sheets 5 2 50.0 42.5 85.0 small sandy sheets very large pointed 6 2 50.0 42.6 85.2 sandy sheets (glass shards) largest pointed 7 2 25.0 19.4 77.6 yellowish sheets (glass shards) afrom which company the DAB was purchased from. Company 1 refers to TCI America and the DAB was received with 98% purity. Company 2 refers to Kindchem (Nanjing) Co., Ltd and the DAB was received with 98% purity.

175

Table B5. Yield of HMT-PBI with the respective DAB batch used for each reaction performed. Reaction # DAB batch Yield (g) Yield (%) Appearance/Notes 1 1 26.0 102.8 off-white fibrous solid 2 1 26.9 106.3 off-white fibrous solid 3 1 25.9 102.4 off-white fibrous solid 4 1+2 25.8 102.0 off-white fibrous solida 5 2 27.2 107.5 off-white fibrous solida 6 2 25.6 101.2 off-white fibrous solid 7 2 26.2 103.5 thick off-white fibersb 8 2+3 25.6 101.2 thick white fibrous solid 9 3 25.5 100.8 thick white fibrous solid 10 3 25.4 100.4 thick white fibrous solid 11 3 26.2 103.5 thick white fibrous solid 12 3+4 25.8 102.0 thick white fibrous solid 13 4 25.7 101.6 thick white fibrous solid 14 4 27.3 107.9 thick white fibrous solidc 15 4 24.7 97.6 thick white fibrous solidc 16 4 25.3 100.0 thick white fibrous solid 17 4+5 25.5 100.8 thin white fibrous solid 18 5 25.5 100.8 thin white fibrous solid 19 5 25.7 101.6 thin white fibrous solid 20 5 25.4 100.4 thin white fibrous solid 21 5+6 25.6 101.2 thin white fibrous solid very thick white fibrous 22 6 26.2 103.5 solid very thick white fibrous 23 6 25.9 102.4 solid very thick white fibrous 24 6 26.4 104.3 solid very thick white fibrous 25 6+7 25.5 100.8 solid very thick white fibrous 26 7 25.7 101.6 solid aturned partially yellow after being left in air overnight when wet in acetone (not immediately dried). bpolymer precipitated into ice-water rather than room temperature water. cthese two samples were likely accidently mixed when collecting the solid, as they were performed side-by- side. Their average yield is 102.8%, which matches the total overall yield.

176

Table B6. Yield of ~50% dm HMT-PMBI, reaction scale, amount of iodomethane (MeI) used, and calculated dm% from 1H NMR spectroscopy data for each reaction performed. MeI amount Reaction # Scale Yield (g) Yield (%) dm Appearance used (mL) 1 0.83 13.8 44.0 79.3 55.6% pale brown powdera 2 1.00 18.5 53.3 83.3 51.4% pale brown powderb 3 1.00 20.0 52.7 79.7 54.9% pale yellow powderb 4 1.00 22.0 54.4 83.1 53.8% pale yellow powderc 5 1.00 22.0 58.6 88.2 55.3% pale brown powder 6 1.00 22.0 59.6 88.6 56.7% brown powder 7 1.00 21.0 58.2 88.9 53.7% pale brown powder 8 1.00 21.0 59.3 88.5 56.2% brown powder 9 1.00 21.0 58.7 89.1 54.5% pale brown powder 10 1.00 21.0 58.6 87.6 56.0% brown powder 11 1.00 21.0 59.1 90.2 53.9% pale brown powder apolymer was precipitated into water and no potassium iodide was used in the purification process. bpolymer was precipitated into methanol and no potassium iodide was used in the purification process. cpolymer was precipitated into methanol with potassium iodide. No potassium iodide was used in the acetone purification step.

Table B7. Yield of >55% dm HMT-PMBI, amount of iodomethane (MeI) used, reaction time, and dm% as calculated by 1H NMR spectroscopy data for each reaction performed. MeI amount Reaction time Reaction # Initial dm Final dm Yield (g) Yield (%) used (mL) (h) 1a 10.5/0.6 16/19 55.6% 89.2% 31.6 70.6 2b 13.0 18 53.4% 89.4% 37.6 82.2 3 13.0 21 51.4% 90.2% 45.6 97.2 4 13.0 18 54.9% 88.9% 44.2 98.4 5 11.0 17 54.3% 86.5% 43.0 96.8 6 9.0 17 53.8% 82.8% 42.7 98.4 7c 18.0 90 55.3% 97.0% 47.2 99.7 athis reaction followed a two-step methylation process. The first methylation was performed in DCM for 16 h and the second methylation in DMSO for 19 h. The polymers were collected by precipitation. bpolymer was collected by precipitation into acetone containing potassium iodide. cthe initial MeI amount was 13.0 mL but was increased to 18.0 mL after 48 h and continued for a total of 90 h.

177

NMR Spectra:

1 Figure B1. Stacked H NMR spectra (400 MHz, DMSO-d6) of BME for each reaction performed.

178

1 Figure B2. Stacked H NMR spectra (400 MHz, CDCl3) of HMTE for each reaction performed.

179

1 Figure B3. Stacked H NMR spectra (400 MHz, DMSO-d6 + 5 drops of 40 wt% KOD in D2O) of HMT-PBI for each reaction performed.

180

1 Figure B4. Stacked H NMR spectra (400 MHz, CD2Cl2) of ~50% dm HMT-PMBI for each reaction performed.

181

Appendix C.

Supporting Information for Chapter 4

Unpublished NMR Spectra of New Compounds:

1 Figure C1. The H NMR spectrum (500 MHz, DMSO-d6) of 1.

1 Figure C2. The H NMR spectrum (500 MHz, DMSO-d6) of 2.

182

1 Figure C3. The H NMR spectrum (500 MHz, DMSO-d6) of 3.

1 Figure C4. The H NMR spectrum (500 MHz, DMSO-d6) of 4.

183

1 Figure C5. The H NMR spectrum (500 MHz, DMSO-d6) of 9.

184

Appendix D

Supporting Information for Chapter 5

Unpublished NMR Spectra of New Compounds:

1 Figure D1. The H NMR spectrum (500 MHz, DMSO-d6) of 15.

13 Figure D2. The C NMR spectrum (125 MHz, DMSO-d6) of 15.

185

Table D1. Crystal structure data for each of the four model compounds that were crystallized in their iodide forms.

HB BrB MeB PhB·H2O by slow cooling of a by slow evaporation by slow evaporation by slow evaporation crystallization solution of HB in of MeB in H2O of PhB in of BrB in H2O under method EtOH from reflux to under ambient EtOAc/EtOH under ambient conditions RT in air conditions ambient conditions colour and shape colourless block colourless needle colourless sheet colourless block

refined formula C15H15I1N2 C15H13Br2I1N2 C18H21I1N2 C27H23I1N2·H2O formula weight 350.203 507.995 392.284 520.414 (g mol-1) crystal dimensions 0.211 x 0.288 x 0.077 x 0.117 x 0.092 x 0.148 x 0.523 x 0.254 x 0.176 (mm3) 0.388 0.224 0.354 radiation Cu Kα Mo Kα Cu Kα Mo Kα wavelength (Å) 1.54178 0.71073 1.54178 0.71073 crystal system monoclinic monoclinic orthorhombic monoclinic

space group P21/c P21/c Pbca P21/n a (Å) 11.7933(3) 11.5893(4) 9.72290(10) 12.6275(6) b (Å) 8.2391(2) 19.6291(7) 11.59830(10) 13.5185(7) c (Å) 15.6325(3) 15.4023(5) 32.4106(4) 14.1586(7) α (°) 90 90 90 90 β (°) 111.2550(10) 106.7010(10) 90 99.1810(10) γ (°) 90 90 90 90 V (Å3) 1415.63 3356.02 3654.92 2385.98 Z 4 8 8 4 T (K) 296(2) 296(2) 299(2) 296(2)

ρcalcd (g cm-3) 1.643 2.011 1.426 1.449 μ (mm-1) 17.633 6.663 13.719 1.362

2θmax (°) 145.16 64.356 133.174 53.110 observed 2636 6677 2854 3973 reflectionsa

Rint 0.0516 0.0337 0.0428 0.0208 Ra 0.0412 0.0367 0.0381 0.0309 wRa 0.1047 0.0650 0.1063 0.0733 goodness of fit 1.042 1.000 1.041 1.111 CCDC # 1439721 1439723 1439722 1439724 a for Io > 2σ(Io).

186