DESIGN OF ANION EXCHANGE MEMBRANES FOR ELECTROCHEMICAL CARBON DIOXIDE REDUCTION TO CARBON MONOXIDE by Angelica Reyes B.Sc., The University of Alberta, 2015 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Chemical and Biological Engineering) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) November 2020 © Angelica Reyes, 2020 ii The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the thesis entitled: Design of Anion Exchange Membranes for Electrochemical Carbon Dioxide Reduction to Carbon Monoxide submitted by Angelica Reyes in partial fulfillment of the requirements for the degree of Master of Applied Science in Chemical and Biological Engineering Examining Committee: Curtis P. Berlinguette, Professor, Chemical and Biological Engineering, UBC Supervisor Előd Gyenge, Professor, Chemical and Biological Engineering, UBC Supervisory Committee Member Savvas G. Hatzikiriakos, Professor, Chemical and Biological Engineering, UBC Supervisory Committee Member iii Abstract Electrochemical CO2 reduction technologies provide a platform for transforming renewable electricity, water, and waste CO2 into synthetic building blocks (e.g., CO) and chemicals. However, CO2 electrolyzers are not yet available in the market due to poor efficiencies at high reaction rates. Emerging CO2 electrolyzers containing an anion exchange membrane (AEM) offer promise in overcoming this challenge, but limited design principles exist for AEMs in CO2 electrolysis application. This thesis reports on my investigation of the designs of AEMs that improve CO2 electrolyzer performance for CO production. I first interrogate two analogous AEM designs to determine the effect of AEM functional group on product selectivity, cell voltage, and stability. This relationship between the membrane and CO2 electrolyzer performance has not yet been established at reaction rates greater than 10 mA/cm2. I demonstrate that an imidazolium-based AEM achieves a higher cell performance than a trimethylamine analog at reaction rates up to 100 mA/cm2. I find that the low water uptake (i.e., water content) and improved chemical stability of an imidazolium group are contributing factors that lead to increased performance for CO production. I explore further how AEM water uptake and thickness, and cathode hydrophobicity impact the amount of water transported to the cathode. Water serves as a proton source for the CO2 reduction reaction to CO, but excess water at the cathode (i.e., flooding) can block the pores of catalytic sites and impede mass transport of reactants and products. Theoretical simulations predict that flooding will occur at reaction rates greater than 750 mA/cm2, but this thesis demonstrates that cathode flooding is an issue at merely 200 mA/cm2. I find that thin, low water uptake AEMs paired with hydrophobic cathodes mitigate cathode flooding and improve product selectivity and cell voltage for CO production. iv Lay summary Fuels and chemicals consumed globally are primarily produced by burning fossil fuels, and this process releases harmful carbon dioxide (CO2) into the atmosphere. A cleaner alternative to producing chemical feedstocks is by converting renewable solar electricity, water, and CO2 in a reactor called a CO2 electrolyzer. This technology, however, is not yet commercially available because it is difficult to achieve high conversion efficiencies. One of the important components in CO2 electrolyzers that affect efficiency is the anion exchange membrane, which is a solid polymer material that facilitates the transport of ions and water across the electrolyzer. This thesis demonstrates the design of membranes that enable high conversion efficiencies of CO2 into useful products. v Preface Chapter 3 is based on my unpublished work under the supervision of Prof. Curtis P. Berlinguette. Fabrication of electrodes, membrane characterization, and electrochemical testing in CO2 electrolyzers were carried out by myself. Synthesis of ionomers and NMR experiments were performed by Dr. Yang Cao in the Berlinguette group. The fabrication of ionomers into membrane films were carried out by Ryan P. Jansonius in the Berlinguette group. Chapter 4 is adapted from the paper “Managing Hydration at the Cathode Enables Efficient CO2 Electrolysis at Commercially Relevant Current Densities”, ACS Energy Lett. 2020, 5, 1612-1618.1 The work was supervised by Prof. Curtis P. Berlinguette. Membrane characterizations and electrochemical experiments were carried out by myself. Synthesis of ionomers, electrochemical surface area measurements, and NMR experiments were conducted by Benjamin P. Mowbray and Dr. Yang Cao in the Berlinguette group. The fabrication of ionomers into membrane films were carried out by Ryan P. Jansonius in the Berlinguette group. Cathodes were prepared by Jacky Chau in the Berlinguette group. SEM and EDX experiments were carried out by David J. Dvorak in the Berlinguette group. COMSOL modelling was simulated by Danika G. Wheeler in the Berlinguette group. The manuscript was written by myself, Ryan P. Jansonius, and Benjamin P. Mowbray with contributions from Prof. Curtis P. Berlinguette. vi Table of contents Abstract ................................................................................................................................... iii Lay summary ........................................................................................................................... iv Preface ...................................................................................................................................... v Table of contents ..................................................................................................................... vi List of tables............................................................................................................................. ix List of figures ............................................................................................................................ x List of schemes ........................................................................................................................ xii List of abbreviations .............................................................................................................. xiii Acknowledgements ............................................................................................................... xvii Dedication ............................................................................................................................ xviii Chapter 1: Introduction ........................................................................................................... 1 Chapter 2: Literature review ................................................................................................... 3 2.1 Advantages of the zero-gap architecture ....................................................................... 3 2.2 Review of membranes in zero-gap CO2 electrolyzers ................................................... 3 2.3 Ion transport mechanisms in AEMs .............................................................................. 4 2.3.1 Anion transport through AEMs ............................................................................... 4 2.3.2 Cation transport through AEMs ............................................................................... 6 2.4 Water transport mechanisms in anion exchange membranes ......................................... 6 2.5 Characterization of anion exchange membranes ........................................................... 8 2.5.1 Ionic conductivity ................................................................................................... 8 2.5.2 Ion exchange capacity ........................................................................................... 10 2.5.3 Water uptake ......................................................................................................... 11 2.6 Challenges with AEMs in CO2 electrolyzers ............................................................... 11 2.6.1 Crossover of reactant CO2 and cations ................................................................... 12 2.6.2 Mechanical and chemical stability ......................................................................... 13 2.6.3 Ohmic and interfacial losses .................................................................................. 13 2.7 Review of state-of-art performance of AEM-based CO2 electrolyzers ......................... 14 2.8 Thesis objectives ........................................................................................................ 15 2.9 Thesis organization .................................................................................................... 16 vii Chapter 3: Linking structure and properties of anion exchange membranes to gas-fed CO2 electrolysis performance in a zero-gap reactor ..................................................................... 17 3.1 Introduction ................................................................................................................ 17 3.2 Results and discussions .............................................................................................. 18 3.2.1 Design of zero-gap CO2 electrolyzer and MEA materials ...................................... 18 3.2.2 Characterization of AEMs ..................................................................................... 20 3.2.3 CO2RR performance of AEMs in a zero-gap reactor ............................................. 22 3.3 Conclusion ................................................................................................................. 24 3.4 Experimental methods ................................................................................................ 25 3.4.1 Electrode preparation ............................................................................................ 25 3.4.2 Membrane characterization ................................................................................... 25 3.4.3 Product analysis .................................................................................................... 26 3.4.4 Electrochemical impedance spectroscopy measurements ....................................... 27 Chapter 4: Managing hydration at the cathode enables efficient CO2 electrolysis at commercially relevant current densities ................................................................................ 28 4.1 Introduction ................................................................................................................ 28 4.2 Results ....................................................................................................................... 30 4.2.1 Design of zero-gap CO2 electrolyzer and MEA materials ...................................... 30 4.2.2 MEA materials influence water flux and CO2RR performance .............................. 32 4.3 Discussions ................................................................................................................ 37 4.3 Conclusions ................................................................................................................ 40 4.4 Experimental methods ................................................................................................ 41 4.4.1 Materials ............................................................................................................... 41 4.4.2 Electrode preparation ............................................................................................ 41 4.4.3 Ionomer synthesis ................................................................................................. 42 4.4.4 Membrane fabrication ........................................................................................... 44 4.4.5 Membrane characterization ................................................................................... 45 4.4.6 Electrochemical measurements and product analysis ............................................. 46 4.4.7 Electrochemical impedance spectroscopy measurements ....................................... 47 4.4.8 Electrochemical surface area measurements .......................................................... 48 viii Chapter 5: Conclusions and future directions ...................................................................... 53 5.1 Conclusions ................................................................................................................ 53 5.2 Future directions ........................................................................................................ 53 References ............................................................................................................................... 56 Appendices .............................................................................................................................. 62 Appendix A: Chapter 3 supplementary material .................................................................... 62 Appendix B: Chapter 4 supplementary materials .................................................................. 63 ix List of tables Table 4.1 Doctor blade spacing and annealing conditions to fabricate ionomer solutions into membranes. .............................................................................................................................. 48 x List of figures Figure 1.1 Schematic of a gas-fed zero-gap CO2 electrolyzer. .................................................... 2 Figure 2.1 Ion transport between the cathode (labeled “C”) and anode (labeled “A”) for an anion exchange membrane (AEM), cation exchange membrane (CEM), and bipolar membrane (BPM) under reverse bias. ...................................................................................................................... 4 Figure 2.2 Anion transport in the AEM depends on the structure and water content of the membrane, and may be a combination of Grotthus, diffusion, and surface site hopping mechanisms. Figure adapted from Grew and Chiu.28 .................................................................. 5 Figure 2.3 Net water flux to the cathode (JH2O,net) is the sum of diffusion (JDIFF), electro-osmotic drag (JEOD), and back convection (JBC) processes........................................................................ 7 Figure 2.4 (a) Water diffusivities (DH2O) and (b) electro-osmotic drag (ηEOD) coefficients of anion exchange membranes as a function of different water content (λ) at different temperatures. Figures adapted from Li and coworkers.49 .................................................................................. 8 Figure 3.1 (a) Schematic of the zero-gap reactor used in this study and (b) operating conditions for CO2RR experiments. ........................................................................................................... 19 Figure 3.2 (a) Structures of AEMs synthesized in this study and (b) fabricated with the same thickness. ................................................................................................................................. 20 Figure 3.3 The ion exchange capacity (IEC) of the chloride form of PSMIM and PSTMA. Error bars represent the standard deviations for four independent samples. ........................................ 21 Figure 3.4 (a) The liquid water uptake (WU) in Cl– form and (b) OH– (σ) through-plane conductivity of PSTMA and PSMIM. Error bars represent the standard deviations for four independent samples. ................................................................................................................ 22 Figure 3.5 (a) The Faradaic efficiency for CO production (FECO) and (b) cell voltage (Ecell) in the applied current density (j) of 0.75, 1, 5, 10, 25 mA/cm2 using a gas-fed zero-gap CO2 electrolyzer containtaining PSTMA (green) or PSMIM (blue). ................................................. 23 Figure 3.6 The CO Faradaic efficiency (FECO) for (a) PSMIM and (b) PSTMA during an applied current density (j) of 25 and 100 mA/cm2 for 5000s. ................................................................. 24 xi Figure 4.1 (a) Expanded and (b) cross-sectional diagrams of the gas-fed zero-gap CO2 reactor used in this study. The membrane electrode assembly consists of a nickel foam anode, anion exchange membrane, and Ag/polytetrafluoroethylene (PTFE) on carbon paper as the cathode. . 31 Figure 4.2 (a) The water flux to the cathode (JH2O,cathode) after applying current densities (j) of 25, 50, 100, and 200 mA/cm2 for MEAs with 55-μm thick HWUM (green) and LWUM (orange). (b) Total cell voltage (Ecell) (c) CO faradaic efficiency (FECO), and (d) H2 faradaic efficiency (FEH2) at applied current densities (j) of 25, 50, 100, and 200 mA/cm2. Error bars represent the standard deviation for three independent experiments. ............................................................................ 34 Figure 4.3 Effect of LWUMs with different thicknesses (20, 40 and 55 μm) on the flux of water to the cathode (JH2O,cathode) and CO2RR performance. (a) JH2O,cathode after applying a current density (j) of 200 mA/cm2. (b) CO faradaic efficiencies (FECO), (c) H2 faradaic efficiencies (FEH2) and (d) total cell voltage (Ecell) at j of 25, 50, 100, and 200 mA/cm2. Error bars represent the standard deviation for three independent experiments. ........................................................ 35 Figure 4.4 Effect of cathode hydrophobicity (manipulated by varying the PTFE content from 6-29 wt%) on the flux of water to the cathode (JH2O,cathode) and CO2RR performance. (a) JH2O,cathode after applying a current density (j) of 200 mA/cm2. (b) CO faradaic efficiencies (FECO), (c) H2 faradaic efficiencies (FEH2) and (d) total cell voltage (Ecell) at j of 25, 50, 100, and 200 mA/cm2. Error bars represent the standard deviation for three independent experiments. ......................... 36 Figure 4.5 (a) Illustration of the water mass balance at the AEM/cathode interface that accounts for water from the humidified cathode inlet CO2 stream (JH2O,inlet); expelled at the cathode outlet (JH2O,outlet); consumed from CO2RR (JH2O,reduction); and from the anode that permeated through the membrane (JH2O,net). (b) JH2O,net is the sum of diffusion (JDIFF), electro-osmotic drag (JEOD), and back convection (JBC) processes. .............................................................................................. 38 Figure 4.6 1H NMR of A1 in acetone-d6 at 400 MHz and ambient temperature. ....................... 49 Figure 4.7 1H NMR of B1 in acetone-d6 at 400 MHz and ambient temperature. ....................... 50 Figure 4.8 1H NMR of A2 in acetone-d6 at 400 MHz and ambient temperature. ....................... 51 Figure 4.9 1H NMR of B2 in acetone-d6 at 400 MHz and ambient temperature. ....................... 52 xii List of schemes Scheme 4.1 Synthesis of intermediate copolymer poly(4-vinylbenzyl chloride-co-styrene), and anion exchange polymer poly[(4-vinylbenzyl-1,2,4,5-tetramethylimidazolium chloride)-co-styrene] .................................................................................................................................... 49 xiii List of abbreviations ” inch °C degrees Celsius A contact area AC alternating current AEM anion exchange membrane Ag silver BPM bipolar membrane C celsius CEM cation exchange membrane Cl– chloride cm centimetre CO carbon monoxide CO2 carbon dioxide CO2RR carbon dioxide reduction reaction CO32– carbonate cps counts per second DH2O water diffusivity DI deionized e− electron Ecell cell voltage EDX energy dispersive x-ray EIS electrochemical impedance spectroscopy EtOH ethanol F faraday constant FE faradaic efficiency xiv Fm molar flow rate FTIR Fourier-transform infrared spectroscopy Fv volume flow rate GC gas chromatography GDE gas diffusion electrode h hour H+ protons H2 hydrogen H2O water HCO3– bicarbonate HCOO– formate HER hydrogen evolution reaction HWUM high water uptake membrane Hz hertz I total current IEC ion exchange capacity j current density JBC back convection flux JDIFF diffusion flux JEOD electro-osmotic drag flux JH2O,cathode excess flux of water to cathode JH2O,inlet water flux entering cathode inlet JH2O,net net flux of water JH2O,outlet water flux exiting cathode outlet JH2O,reduction water consumed by reduction reactions K kelvin K+ potassium ion xv K2CO3 potassium carbonate KHCO3 potassium bicarbonate KNO3 potassium nitrate KOH potassium hydroxide L length LWUM low water uptake membrane m mass mA milliampere md dry mass MEA membrane electrode assembly MeOH methanol meq milliequivalent mg milligrams mh hydrated mass MHz megahertz mm millimetre mmol millimol mV millivolt Mw molar mass Ni nickel nk number of electrons exchanged for product k NMR nuclear magnetic resonance O2 oxygen OH– hydroxide P pressure Pa pascal ppm parts per million xvi PSMIM poly[(4-vinylbenzyl-1-methylimidazolium chloride)-co-styrene PSTMA poly[(4-vinylbenzyl-trimethylammonium chloride)-co-styrene Pt platinum PTFE polytetrafluoroethylene QAPPT quaternary ammonia poly(N-methyl-piperidine-co-p-terphenyl) R ideal gas constant RHE reversible hydrogen electrode Rmem membrane resistance Rs electrolyte resistance S Siemens s seconds SEM scanning electron microscope T temperature V volt V volt wt% weight percent WU water uptake XRF x-ray fluorescence ηEOD electro-osmotic drag coefficient λ water content μA microampere μL microliter μm micrometre σ ionic conductivity Ω ohms xvii Acknowledgements I would like to thank my supervisor Prof. Curtis P. Berlinguette for allowing me to learn valuable skills and lessons in the group, and for opening doors of opportunities for my future career. Thank you as well to my committee members Prof. Előd Gyenge and Prof. Savvas G. Hatzikiriakos for their helpful feedback on my thesis. Thank you to all the Berlinguette group members for your support, specially to Dr. Yang Cao who helped kickstart my membrane project. Thank you to my brilliant scientists/friends Ryan and Ben M. for being my dream teammates and for making difficult science projects enjoyable. Thank you to my empowered gals Rebecca, Val, and Aiko for all our adventures and nourishing conversations that I will treasure forever. Pam, Tasha, Karen, Isabel, thank you for your constant support. You have inspired and motivated me in the best possible ways. To the greatest roommate I’ve had Rebecka, thank you for being my partner-in-crime in this rollercoaster grad school journey. To my parents, thank you for sacrificing your comfortable lives in the Philippines and restarting from nothing to provide a better life for our family in Canada. This one is for you. Ate Elaine and Princess, thanks for celebrating my achievements, no matter how big or small they may be. Glad to have you both as my sisters/best friends/favourite people to annoy for life. xviii Dedication Dedicated to God and my family Psalm 34:81 Chapter 1: Introduction Balancing the intermittent supply of renewable energy and the fluctuating demands for electricity necessitates scalable energy storage technologies.2 Electrochemical CO2 reduction is a compelling route for storing renewable electricity while recovering value from energy-intensive carbon capture technologies. CO2 electrolyzers transform electrons, captured CO2, and water into synthetic building blocks (e.g., CO) and commodity chemicals (e.g., ethanol).3 This technology offers the opportunity for the meaningful abatement of greenhouse gas emissions by electrifying the manufacturing of fossil fuel-derived chemicals.4 Electrochemical CO2 reduction could also leverage the infrastructure already established from existing supply chains for distribution and handling of products generated from the CO2 reduction reaction (CO2RR), thereby facilitating the widespread adoption of this technology. Despite these advantages, there remain technological barriers that need to be overcome for CO2 electrolyzers to be economically viable and be deployed at industrial scale. A key part to addressing this challenge is the design of CO2 electrolyzers that can operate at high current densities (j > 200 mA/cm2), high faradaic efficiencies (FE > 90%), and low cell voltages (Ecell < 3 V) for extended periods of time (>1 year).5,6 Amongst the different electrolyzers proposed to drive the CO2RR, we find the zero-gap design to be particularly promising.3 A zero-gap CO2 electrolyzer consists of porous cathodes and anodes that sandwich a solid polymer electrolyte (i.e., membrane; Figure 1.1). Many zero-gap CO2 electrolyzers that utilize an anion exchange membrane (AEM) have shown superior performance compared to other types of membranes.7,8 In an AEM-based zero-gap CO2 electrolyzer, water provides the necessary protons to convert CO2 to a CO2RR product at the cathode/AEM interface. Water can be delivered to the cathode by humidifying the cathode feedstock and/or from the anolyte that permeates through the AEM. Anions generated at the cathode (e.g., OH–) migrate 2 through the AEM and are oxidized at the anode, where the oxygen evolution reaction occurs. These ion and water transport processes through AEMs have been examined extensively in zero-gap water electrolyzers and fuel cells, but far less so in the context of zero-gap CO2 electrolyzers. This thesis serves to teach how the structure (e.g., functional group) and physical attributes (e.g., thickness) of the AEM impact the flow of ions and water through the membrane and, ultimately, CO2RR performance (i.e., selectivity, energy efficiency, stability). Figure 1.1 Schematic of a gas-fed zero-gap CO2 electrolyzer.3 Chapter 2: Literature review 2.1 Advantages of the zero-gap architecture Zero-gap CO2 electrolyzers effectively have “zero” gap between the porous cathodes and anodes because the two electrodes are each pressed against the membrane (5-100 μm). This reactor design minimizes the total cell voltage by decreasing the distance for ion transport (i.e., ohmic resistance). Zero-gap CO2 electrolyzers can also achieve high current densities (≥100 mA/cm2) by delivering gaseous CO2 to the cathode as a means of overcoming the mass transport limitations that arise in aqueous feedstocks.3,9 The efficiency of zero-gap CO2 electrolyzers depends on the choice of electrocatalyst (e.g., Ag-coated carbon paper), operating conditions (e.g., CO2 flow rate, cell temperature), and membrane. This thesis focuses on the effect of the membrane on the conversion of CO2 to a single product using a Ag cathode, which catalyzes the desirable CO2RR to CO and the competing hydrogen evolution reaction (HER): CO2RR: 2CO2 + 2H2O + 4e– → 2CO + 4OH– E1 = -0.11 V (vs. RHE) (Eq. 2.1) HER: 4H2O + 4e– → 2H2 + 4OH– E2 = 0.00 V (vs. RHE) (Eq. 2.2) 2.2 Review of membranes in zero-gap CO2 electrolyzers The membrane is an important component in zero-gap CO2 electrolyzers because it impacts the local pH at the cathode, modulating selectivity for CO2RR products and energy efficiency. Membranes facilitate ion and water transport between the cathode and anode while preventing reactant and/or product crossover.9–12 Aside from anion exchange membranes (AEMs),8,12–14 cation exchange membranes (CEMs)15–18 and bipolar membranes (BPMs)9,19 are also being actively used in zero-gap CO2 electrolyzers (Figure 2.1a). CEMs transport cations (e.g., H+) from the anode to cathode, AEMs transport anions (e.g., OH–) from the cathode to anode, and BPMs transport H+ to the cathode and OH– to the anode from the dissociation of H2O at the AEM/CEM 4 interface under a reverse bias. Zero-gap CO2 electrolyzers containing AEMs have demonstrated higher conversion efficiencies and energy efficiencies than CEM and BPM systems (Figure 2.1b). AEMs may be more suitable than CEMs and BPMs likely because the concentration of H+ at the catalyst surface is minimized, thereby suppressing the competing HER.20 Figure 2.1 Ion transport between the cathode (labeled “C”) and anode (labeled “A”) for an anion exchange membrane (AEM), cation exchange membrane (CEM), and bipolar membrane (BPM) under reverse bias. 2.3 Ion transport mechanisms in AEMs 2.3.1 Anion transport through AEMs AEMs contain hydrophilic positively-charged groups (e.g., imidazolium) that can be anchored as side chains or directly onto the hydrophobic polymer backbone (e.g., polystyrene). The presence of both hydrophobic and hydrophilic components in the AEM contributes to a phase segregated microstructure, forming percolated ionic domains that facilitate the flow of anions from the cathode to the anode.21–23 Anion transport through these hydrophilic domains relies on the water content in the AEM and this topic remains to be an active area of research.24–26 Previous reports describe ion transport through AEMs as a combination of the Grotthuss, diffusion, and surface site hopping mechanisms (Figure 2.2).27,28 5 Figure 2.2 Anion transport in the AEM depends on the structure and water content of the membrane, and may be a combination of Grotthus, diffusion, and surface site hopping mechanisms. Figure adapted from Grew and Chiu.28 Grotthuss mechanism is one of the dominant modes of anion (e.g. OH–) transport through the AEM.22,27,29,30 In this mechanism, OH– are shuttled through the hydrogen bond network of water molecules by the formation and cleavage of covalent O–H bonds with neighbouring molecules. The diffusion mechanism is operative when there is a concentration or electrical potential gradient across the membrane. Surface site hopping (also known as the vehicular mechanism) involves the movement of OH– ions from one side chain of the AEM to another through continuous overlapping first solvation shells of cationic functional groups.31 The contribution of each transport mechanism depends on the structure and water content of the AEM.30,31 Both surface site hopping and diffusion mechanisms require unbound water within the membrane. In a membrane with low hydration levels, the ionic conductivity of the AEM is limited 6 because the hydrophilic domains are small, thus, restricting the diffusion and surface site hopping mechanisms. In a highly hydrated membrane, the ionic conductivity is expected to be higher because water-filled channels expand and anion transport will occur through diffusion mechanisms (i.e. Grotthuss and diffusion mechanisms). 2.3.2 Cation transport through AEMs An ideal AEM should exhibit a high permselectivity by facilitating anion transport between the anode and cathode while excluding the flux of co-ions (e.g., K+). Migration of cations in AEM-based CO2 electrolyzers is undesirable because salt can precipitate at the cathode and limit mass transport to/from catalytic sites (further discussed in Section 2.7.1). There are two mechanisms for co-ion transport through the AEM: i) diffusion due to a difference in concentration gradient across the membrane and/or ii) electromigration due to an externally applied electric field.32,33 In electric field-driven migration, cations from the anolyte are electrostatically attracted to the negatively-charged cathode. Undesirable cation transport through the AEM can be minimized by optimizing the cell operating conditions (e.g., applied current or anolyte concentration) and through materials design.34–36 In a materials design perspective, AEMs that contain a high concentration of positively-charged functional groups, for example, generally decrease cation transport through electrostatic exclusion.37–39 This exclusion effect, however, decreases if increasing the concentration of positively-charged functional groups significantly leads to a high membrane water content.40 2.4 Water transport mechanisms in anion exchange membranes Water transport through the AEM is another important factor that influences the performance of the CO2RR as is the case in alkaline fuel cells.41–44 Water is a reactant for the desired CO2RR, but excess amounts of water present at the cathode could hinder CO2 diffusion to 7 the catalyst and promote the undesired HER.45 There are two methods for delivering water to the CO2RR catalyst in a zero-gap reactor: i) through the humidified CO2 feedstock and/or ii) from the anolyte that permeates through the AEM. In the latter case, the net flux of water (JH2O,net) to the cathode can be described by diffusion (JDIFF), electro-osmotic drag (JEOD), and back convection (JBC) processes (Figure 2.3).44,46,47 Water diffuses from the anolyte to the cathode CO2RR catalyst because of the liquid concentration gradient between the aqueous anode environment and the gas-fed cathode. Water is also driven from the cathode to the anode through electro-osmotic drag as solvated anions migrate toward the anode. Back convection of water from the cathode to the anode is driven by the liquid pressure gradient across the membrane.48 Figure 2.3 Net water flux to the cathode (JH2O,net) is the sum of diffusion (JDIFF), electro-osmotic drag (JEOD), and back convection (JBC) processes. The rate of each water flux depends primarily on the transport properties of the AEM. For example, water diffusivity (DH2O) and electro-osmotic drag (ηEOD) coefficients both increase for membranes that exhibit a higher water content or uptake (λ; Figure 2.4).49–51 Previous reports also highlight that the back convection of water from the cathode to the anode increases with thinner 8 membranes.44,47,52 It is therefore important to design AEMs with suitable properties (e.g., water uptake and thickness) in order to deliver the optimal amount of water to the CO2RR catalyst. Figure 2.4 (a) Water diffusivities (DH2O) and (b) electro-osmotic drag (ηEOD) coefficients of anion exchange membranes as a function of different water content (λ) at different temperatures. Figures adapted from Li and coworkers.49 2.5 Characterization of anion exchange membranes I highlight in this section the three key characterization techniques for measuring the performance and chemical stability of AEMs: ionic conductivity, ion exchange capacity, and water uptake. 2.5.1 Ionic conductivity Ionic resistances contribute to the total cell voltage of CO2 electrolyzers. Ionic conductivity can be measured along the plane (i.e., in-plane) or through the thickness of the membrane (i.e., through-plane).53,54 Conductivity values from these two measurements are not equal because of anisotropic resistances that arise from the polymer casting method or use of support materials in the membrane (e.g., reinforcements).55–57 Through-plane measurements are more difficult to implement than in-plane measurements because this technique requires other resistances (e.g., contact resistance from electrical leads and connections, interfacial resistance between the 9 electrode and the membrane) to be minimized or accounted for.54,58 The through-plane conductivity, however, is more relevant because anions in a CO2 electrolyzer migrate between the cathode to the anode through the thickness of the AEM. Through-plane membrane resistances are typically determined using an ex-situ or in-situ two-probe electrochemical impedance spectroscopy (EIS) approach, and I use both methods in this thesis. For ex-situ measurements, a membrane is sandwiched between two electrodes (e.g., Pt sheets) and this conductivity cell is either submerged in a water bath or placed in a controlled humidity chamber to ensure that the water content in the membrane remains constant throughout the experiment. In-situ conductivity measurements are performed using the same conditions as an operating zero-gap cell. For both ex-situ and in-situ diagnosis, an AC signal (ca 10-100 mV) is applied to the cell in a wide frequency range (10 MHz to 1 Hz) to obtain a Nyquist plot. The anion conductivity (σ in S/cm) can be calculated from the resistance of the membrane (Rmem in Ω) obtained from the high-frequency data in the Nyquist plot, the contact area between the electrodes (A in cm2) and the thickness of the membrane (L in cm): 𝜎 =𝐿𝑅𝑚𝑒𝑚 ⋅ 𝐴 (Eq. 2.3) Current protocols for ex-situ conductivity measurements assume that OH– is the sole charge carrier, however, AEMs contain a mixture of OH– and less mobile HCO3– and CO32– ions under CO2 electrolysis experiments. Anion conductivities collected in-situ are more representative of the conditions in a CO2 electrolyzer, but requires considerable time and material (e.g., catalysts, gas diffusion layers) to reproducibly gather the required data. 10 2.5.2 Ion exchange capacity The ion exchange capacity (IEC) provides clues about the suitability of newly developed AEMs for electrochemical CO2 reduction before testing for ionic conductivity or performance in an electrolyzer. IEC (in meq/g) is defined as the number of exchangeable functional groups per dry weight of polymer, and is typically reported for AEMs in Cl– form (i.e. Cl– is the mobile counter-ion in the membrane). To determine the IEC, the membrane is pre-treated in an aqueous electrolyte (e.g., 1.0 M KOH) as a means of exchanging the Cl– ions in the membrane with the counter-ions in the solution (e.g., OH–). The concentration of Cl– ions in the solution can be determined using titration (e.g., Mohr method, acid/base) or an ion selective electrode.59,60 Spectroscopy techniques (e.g., FTIR, NMR) can also be used to approximate IEC of the ionomer in powder form.59,61 There are variabilities between IECs determined using each procedure due to inaccurate measurement of weakly or strongly basic functional groups in the AEM that may not dissociate depending on the pH of the pre-treatment solutions.59,62 Titration-based methods are the most common techniques for determining the IEC. However, a significant source of error for this method arises from using colour change as an indicator for the equivalence point. FTIR and NMR spectroscopy can determine the theoretical total number of functional groups in the membrane, but this value is generally different from the number of functional groups that participate in ion exchange. In this thesis, I utilize a Cl– selective electrode to determine IEC, but this approach may underestimate the true IEC value because weakly basic functional groups may not be detected. IEC measured using a Cl– selective electrode is calculated from the concentration of Cl– in solution ([Cl–] in ppm) obtained from a calibration curve, the mass of the pre-treatment aqueous KNO3 11 solution (mKNO3 in g), molar mass of Cl– (MCl in g/mol) and the dried mass of the membrane (md in g): 𝐼𝐸𝐶 = [𝐶𝑙−] ⋅ 𝑚𝐾𝑁𝑂3𝑀𝐶𝑙 ⋅ 𝑚𝑑 (Eq. 2.4) 2.5.3 Water uptake It is also important to characterize the amount of water in the AEM because water facilitates anion transport through the AEM, but high water contents can compromise the mechanical integrity of the AEM due to excessive swelling.63 The water content of the AEM is related to the water uptake property, which is described as the % increase in the mass of the AEM when it is fully hydrated in liquid or water vapor (mh) relative to when the membrane is completely dry (md).64 Water uptake of AEMs are typically reported in the Cl– form and when the membrane is hydrated with liquid water. In this measurement, AEM in Cl– form is immersed in deionized water for 24 hours while replacing the solution at least three times over this duration to ensure excess ions are removed from the membrane. Water from the surface of the AEM is removed with tissue immediately before weighing the membrane in an analytical balance to yield mh. Complete drying of the membrane is typically performed in a dry, sealed environment (e.g. a glass container containing desiccants) for a specific duration (e.g. 5 days) before weighing the dried mass of the membrane. Water uptake values are then calculated as follows: 𝑊𝑈 = 𝑚ℎ − 𝑚𝑑𝑚𝑑× 100% (Eq. 2.5) 2.6 Challenges with AEMs in CO2 electrolyzers Early reports of AEM-containing CO2 electrolyzers used membranes that were originally developed for other electrochemical applications: electrodialysis, water electrolyzers, and fuel 12 cells.15,17,65,66 As such, I highlight in this section the major operational challenges that have been encountered when using AEMs in CO2 electrolyzers. 2.6.1 Crossover of reactant CO2 and cations The amount of free CO2 available to the catalyst decreases due to the reaction between the CO2 feed and CO2RR-generated OH– (Eq. 2.6 and 2.7). CO2 + OH– → HCO3– (Eq. 2.6) CO2 + HCO3– → CO32– + H2O (Eq. 2.7) The generated HCO3– and CO32– from these reactions could also crossover the AEM from the cathode to the anode electrolyte, where they can be oxidized to O2 and CO2 (Eq. 2.8 and 2.9). HCO3– → ½O2 + H2O + 2e– + 2CO2 (Eq. 2.8) CO32– → ½O2 + 2e– + CO2 (Eq. 2.9) A recent report shows that crossover of CO2 through the AEM exceeds the amount of CO2 converted to the target product.67 Negatively charged CO2RR products (e.g., HCOO–) are also readily transported across the positively-charged AEM while neutral products (e.g., MeOH, EtOH) are able to crossover through diffusion.68 This situation implicates the need for separation units downstream of the CO2 electrolyzer to recover valuable CO2RR products and a high purity O2 stream. Co-ion transport of K+ from the anolyte (typically aqueous KHCO3 or KOH solution) to the cathode forms KHCO3 and K2CO3 crystals when combined with the HCO3– and CO32– ions at the cathode. These reactions are undesirable because salt precipitates at the cathode block the pores of the gas diffusion electrode and accelerate the accumulation of excess water at the cathode (i.e., flooding).45,69 Both of these phenomena impede the transport of reactant CO2 to the catalytic sites 13 and limit the surface area available for CO2RR. The use of pure water or diluted solution (e.g. 10 mM KHCO3) at the anode reduces the formation of CO32– and HCO3– crystals, but compromises ionic conductivity and increases the cell potential. 2.6.2 Mechanical and chemical stability Another major operational challenge is the mechanical and chemical stability of AEMs in CO2 electrolyzers. Due to the high water uptake (up to 80%) of the membrane, AEMs have limited mechanical strength due to swelling.63,70 The chemical stability of AEMs primarily synthesized for other electrochemical applications are also not necessarily suitable for the conditions in a CO2 electrolyzer. Narayan and coworkers, for example, utilized a commercially-available electrodialysis AEM that is chemically stable up to a pH of 10.71 However, AEM-based CO2 electrolyzers operate at a higher pH (>11) than electrodialysis systems because OH– ions are electrolytically generated at the cathode.19,67,72,73 2.6.3 Ohmic and interfacial losses High cell voltages arise due to interfacial and ohmic losses in AEM zero-gap CO2 electrolyzers. Zhuang and coworkers demonstrate state-of-art CO2RR performance using a catalyst-coated AEM and gas diffusion layers impregnated with anion exchange ionomers.8 This configuration likely minimizes the interfacial resistances in the cell, however, many CO2 electrolyzers use bare AEMs with catalyst-coated gas diffusion layers. Our research group also recently demonstrated using an analytical reactor that the ohmic losses of the membrane during operation of the reactor are a significant contribution to the overall cell potential.74 High ohmic losses occur, in part, HCO3– and CO32– ions displace OH– ions as charge carriers in the AEM and these ions are less mobile than OH–.62 14 2.7 Review of state-of-art performance of AEM-based CO2 electrolyzers There are only a few reports of AEM-based zero-gap cells that achieve high current densities (≥200 mA/cm2) while maintaining high selectivities (>90%) and low cell voltages (≤3 V).7,8 Incompatibility issues between the operational requirements in a CO2 electrolyzer and the chemical properties of many existing AEMs is a root cause for the limited cell performance. Significant improvements in CO2RR performance (i.e., energy efficiency, conversion efficiency, and stability) have since been realized with the recent developments of AEMs tailored for gas-phase CO2 electrolysis experiments, but there remains room to innovate new membrane technology to bring these systems closer to commercial viability. Sustainion®, an N-methylimidazolium-functionalized styrene polymer, marked the first viable AEM designed for gas-fed CO2 electrolyzers.75 Sustainion® features a high OH– conductivity (100 mS/cm at 80ºC), contains a high IEC (2.5 meq/g), exhibits a high WU (80%), and is thin (50 μm).76 Masel and coworkers maintained the high conductivity of Sustainion® in the CO2 electrolyzer by supplying a humidified gaseous CO2(g) feed and circulating an aqueous anolyte at the anode. This electrolyzer configuration reached commercially-relevant current densities (200 mA/cm2) while sustaining >95% CO selectivity at 3 V for 3800 h. Many studies have since used Sustainion® in gas-fed CO2 electrolyzers;12,77–79 however, salt precipitation at the cathode occurs when a concentrated electrolyte is supplied at the anode (≥0.5 M KOH or KHCO3).72 Moreover, Sustainion® is prone to crossover of CO2RR products (e.g., ethanol), particularly at the high current densities.77 Our group also recently demonstrated high ohmic losses across the membrane in a zero-gap reactor utilizing a Sustainion® for CO2RR.74 A quaternary ammonia poly(N-methyl-piperidine-co-p-terphenyl) (QAPPT) was introduced by Zhuang and coworkers as another candidate AEM material for gas-fed CO2 15 electrolyzers.8 QAPPT exhibits a high OH– conductivity (137 mS/cm1 at 80ºC) and is chemically stable in basic solutions.8,80 The high conductivity of QAPPT eliminates the need to humidify the CO2 feed or the use of an ionically-conductive anolyte. With an operating cell temperature of 60ºC, the QAPPT-containing CO2 electrolyzer demonstrated the highest current density reported to date (500 mA cm–2) with a F.E. for CO >90% and an Ecell of 3 V. This result encourages further research on how QAPPT can mitigate the other challenges for AEMs in CO2 electrolyzers (e.g. CO2RR product crossover from the cathode to the anode). The two case studies involving Sustainion® and QAPPT draw attention to the importance of synthesizing AEMs with properties (e.g. conductivity, chemical stability, and water uptake) that match the requirements in a CO2 electrolyzer. These properties of an AEM are related to the type of cation functional group embedded in the polymer matrix. While there have been reports that attempt to correlate the functional group with the transport and chemical properties of AEMs and CO2RR performance, these studies were conducted at current densities remote from commercial operating conditions (j < 10 mA/cm2).14,81 2.8 Thesis objectives The overall objective of this thesis is to understand the influence of the AEM physical and chemical structure (i.e., ionic group concentration, thickness, and functional group) on electrochemical CO2 reduction to CO. Specifically, the objectives are: A. To determine the chemical properties (ion exchange capacity, water uptake, and ionic conductivity) of AEMs fabricated in-house with different types of functional groups, concentration of functional groups, and thicknesses. B. To evaluate the performance of the AEMs in a zero-gap CO2 electrolyzer for CO production using Faradaic efficiency (as a measure of product selectivity) and cell voltage 16 C. To design a diagnostic tool for studying the water flux to the cathode of different AEMs and cathode materials 2.9 Thesis organization In Chapter 3, I study how CO2RR performance at j up to 100 mA/cm2 is affected when incorporating an imidazolium or trimethylamine cation groups in the AEM. I characterized the properties of these AEMs (IEC, water uptake, thickness, and anion conductivity) and correlated these values to the performance in a gas-fed CO2 electrolyzer. In Chapter 4, I investigate further the imidazolium-functionalized AEM by varying the water uptake and thickness of the membrane. The goal of this chapter is to diagnose how these properties of the AEM impact the water flux to the cathode and, ultimately, CO2RR performance. Water serves as the proton source for CO2RR, but excess water present at the cathode block the pores of the cathode and impede mass transport of reactants and products.72 I designed a series of experiments to empirically track the water flux to the cathode while varying AEM thicknesses and water uptakes, and cathode hydrophobicity. Chapter 5 outlines the main conclusions of this thesis and future directions towards improvement in experimental set-up and characterization techniques, and additional materials design consideration to minimize cathode flooding. 17 Chapter 3: Linking structure and properties of anion exchange membranes to gas-fed CO2 electrolysis performance in a zero-gap reactor 3.1 Introduction There is a growing interest in the use of anion exchange membranes (AEMs) for gas-fed CO2 electrolysis, yet there are merely two reports that demonstrate >90% Faradaic efficiencies for CO (FECO) and cell voltages (Ecell) ≤3 V at current densities (j) ≥200 mA/cm2.8,76 These studies highlight the use of two AEM designs: i) an N-methylimidazolium-functionalized styrene polymer (commercially known as Sustainion®) and; ii) a quaternary ammonia poly(N-methyl-piperidine-co-p-terphenyl) (QAPPT). There has been considerable improvement in cell performance using these membranes, but there remains room to innovate new AEMs for CO2 electrolysis application. This situation provides the imperative to develop design principles that link AEM structure to CO2 reduction reaction (CO2RR) activity. AEMs facilitate the flow of ions and water to and from the CO2RR cathode. These transport properties of the AEM are intrinsically linked to the type of cation functional group, amongst other parameters of the membrane including concentration of functional group, polymer backbone, and membrane thickness. While there have been reports that attempt to establish the relationship between AEM functional group and CO2RR performance, these studies were conducted at low j (<10 mA/cm2).14,17,81,82 This knowledge gap is striking because the catalytic environment for CO2RR is known to be particularly sensitive at higher j, where the cathode pH is higher and ion and water transport rates are faster.83 In this Chapter 3, I study the effect of cation functional groups (imidazolium and trimethylamine) on CO2RR performance in a gas-fed CO2 electrolyzer at j up to 100 mA/cm2. Our results show that an AEM functionalized with an imidazolium group facilitates a consistently higher (~10-25%) FECO than a trimethylamine analog under the same operating conditions. We 18 were able to resolve these differences by fabricating two AEMs with either an imidazolium or trimethylamine group while maintaining the same polymer backbone, fabrication conditions, membrane thickness, and concentration of functional groups. We assert that AEMs functionalized with an imidazolium group achieves a higher FECO than the trimethylamine analog presumably because it exhibits a lower water uptake, which is important in minimizing salt precipitation at the cathode. We further show that an imidazolium group is more chemically stable than a trimethylamine at higher current densities (j = 100 mA/cm2). This work demonstrates the importance of designing AEMs that are chemically stable and exhibit lower water uptakes for improved CO2RR performance. 3.2 Results and discussions 3.2.1 Design of zero-gap CO2 electrolyzer and MEA materials CO2RR experiments in this study were performed in a gas-fed zero-gap electrolyzer (Figure 3.1) similar to a previous report.9 The reactor consists of end plates, flow field plates, gaskets, and the membrane electrode assembly (MEA). Two stainless steel end plates served as inlets and outlets for the reactants and products. The titanium cathode and stainless steel anode flow field plates were machined with serpentine channels that are 1.5-mm wide and 1.5-mm deep with 1-mm landings. The flow plates sandwiched the MEA (active area = 2 × 2 cm2) and served to facilitate the delivery of reactants and products to and away from the reactive interface. The MEA comprised a silver-coated cathode gas diffusion electrode (GDE; 2 × 2 cm2), an AEM (3 × 3 cm2), and a nickel foam anode (2 × 2 cm2). The AEM component of the MEA was manipulated as described below, and both the cathodes and anodes were unchanged for all experiments. The gaseous CO2 was humidified by flowing the inlet stream through a water bath (heated at 35 ºC) at 11 or 20 sccm before entering the cathode; the anode was supplied with 1.0 M KOH at 10 or 15 19 mL/min (Figure 3.1b). Chronopotentiometry experiments were performed by applying a constant j of 0.75, 1, 5, 10, or 25 mA/cm2 to the entire cell for 1500s using a potentiostat. In-line gas chromatography (GC) analyses of the cathode outlet gas were taken after 800s from the start of electrolysis to determine the FECO. Figure 3.1 (a) Schematic of the zero-gap reactor used in this study and (b) operating conditions for CO2RR experiments. Cathodic GDEs were prepared to contain the same silver catalyst and Nafion binder composition in order to correlate differences in CO2RR performance to the change in functional group of the AEM. To prepare the GDEs, I used an airbrush to manually spray coat a Toray carbon paper (TGP-H-120) with a catalyst ink mixture consisting of fixed amounts of commercially available silver nanopowder and Nafion® dispersed in a 1:1 (v/v) solvent mixture of isopropanol and water. The GDEs were coated with a constant silver loading of ~1.5 mg/cm2 and X-ray fluorescence measurements (XRF) confirmed a constant silver signal intensity of ~65 cps for each of the GDEs (Table A3.1).84 The Sustainion-inspired AEMs in this study consisted of a polystyrene backbone functionalized with either an imidazolium or a trimethylammonium group. The polystyrene 20 backbone was prepared by radical polymerization of styrene (m) with 4-vinylbenzyl chloride (n), and then the benzylic chloride was substituted with either a 1-methylimidazolium or a trimethylammonium ionic group in a second reaction. The synthesized polymers are poly[(4-vinylbenzyl-1-methylimidazolium chloride)-co-styrene (PSMIM) and poly[(4-vinylbenzyl-trimethylammonium chloride)-co-styrene (PSTMA) (Figure 3.2a). Solid ionomer powders were subsequently dissolved in 5 wt% ethanol and were cast onto a glass substrate by doctor blade coating to achieve AEMs with a thickness of ~60 μm (Figure 3.2b). The thickness of the dry membranes was confirmed at three separate locations using a digital micrometre (Mitutoyo) with a variance of ± 1 μm. Figure 3.2 (a) Structures of AEMs synthesized in this study and (b) fabricated with the same thickness. 3.2.2 Characterization of AEMs The IECs of PSTMA and PSMIM samples in chloride form (i.e., Cl– is the counter-ion attached to the functional groups) were collected as a measure of the functional group concentration in the AEM. PSTMA and PSMIM resulted in the same IECs of 2.32 ± 0.06 and 2.34 ± 0.12 mmol/g, respectively (Figure 3.3), which was expected for AEMs synthesized with the same m:n. 21 Figure 3.3 The ion exchange capacity (IEC) of the chloride form of PSMIM and PSTMA. Error bars represent the standard deviations for four independent samples. I characterized the water content of the AEM and ion conductivity to understand how different ionic groups impact these properties of the membrane. The water content of the AEM, defined by water uptake (WU), was gravimetrically measured by quantifying the increase in the mass of each membrane when it is fully hydrated with liquid water relative to when the membrane is dry (Figure 3.4a). PSTMA achieved a higher WU value (114 ± 6%) than PSMIM (66 ± 13%), which is an expected trend that corroborates with observations made in literature comparing the same functional groups.85–88 I also performed ex-situ electrochemical impedance spectroscopy (EIS) measurements of PSTMA and PSMIM in OH– form to determine the ionic conductivities of the membranes (Figure 3.4b). PSTMA achieved a slightly higher conductivity (98 ± 14 mS/cm) than PSMIM (82 ± 6 mS/cm), which again aligns with observations in literature that report the higher OH– conductivity of AEMs decorated with a quaternary amine than imidazolium functional groups.85,86 This result is also expected given that PSTMA has a higher water content in the membrane (i.e., water uptake), which facilitate the efficient transport of anions from the cathode to the anode. 22 Figure 3.4 (a) The liquid water uptake (WU) in Cl– form and (b) OH– (σ) through-plane conductivity of PSTMA and PSMIM. Error bars represent the standard deviations for four independent samples. 3.2.3 CO2RR performance of AEMs in a zero-gap reactor I tested the performance of PSMIM and PSTMA in a zero-gap reactor in order to investigate the relationship between structure of the AEMs and CO2RR activity. Both AEMs achieved similar FECO at j ≤ 1 mA/cm2, but PSTMA resulted in lower cell performance (i.e., higher Ecell and lower FECO) at higher j than PSMIM (Figures 3.5a and 3.5b). This result is unexpected because the higher ionic conductivity of PSTMA should result in lower ohmic resistance and, thus, lower Ecell. However, PSTMA also exhibits a high WU and this property is typically linked to a decrease in the ability of the AEM to exclude the flow of co-ions such as K+ (i.e., permselectivity).34 I hypothesize therefore that the limited CO2RR performance of PSTMA is due to increased K+ crossover from the anode, which can react with CO2 at the cathode. This reaction yields salt precipitates (e.g., K2CO3) that block the pores of the cathode GDE resulting in decreased catalytic sites available for CO2RR, and poor mass transport of reactants and products.45,72 An inspection of the electrolyzers after CO2RR experiments show that PSTMA resulted in more salt precipitates at the cathode flow field than PSMIM, which supports my hypothesis that the higher rates of K+ crossover through PSTMA limits CO2RR activity (Figure A3.1). 23 Figure 3.5 (a) The Faradaic efficiency for CO production (FECO) and (b) cell voltage (Ecell) in the applied current density (j) of 0.75, 1, 5, 10, 25 mA/cm2 using a gas-fed zero-gap CO2 electrolyzer containtaining PSTMA (green) or PSMIM (blue). I investigated further if the difference in CO2RR performance between PSTMA and PSMIM is also related to differences in the stability of the cation functional groups. The stability of PSTMA and PSMIM was tested by applying a constant j of 25 or 100 mA/cm2 for 5000 s (Figures 3.6a and 3.6b). At j of 25 mA/cm2, PSTMA achieved a stable, yet lower (~75%) formation for CO compared to PSMIM (>85%). While PSMIM was able to maintain a FECO of ~25% at higher j of 100 mA/cm2, PSTMA resulted in effectively zero CO production (FECO < 2%). The lower CO2RR performance of PSTMA could be due, in part, to the nucleophilic attack from OH– ions that are electrochemically generated at the cathode.85,86 An inspection of the AEMs after CO2 electrolysis shows PSMIM remained transparent while a visibly brown tint on PSTMA indicates chemical degradation of the membrane. 24 Figure 3.6 The CO Faradaic efficiency (FECO) for (a) PSMIM and (b) PSTMA during an applied current density (j) of 25 and 100 mA/cm2 for 5000s. 3.3 Conclusion AEMs offer promise in reaching commercially relevant CO2 electrolysis performance, but there remains room to innovate new AEMs tailored for CO2RR. Our work establishes the first relationship between the structure of AEMs and CO2RR activity at j up to 100 mA/cm2. Synthesizing membranes that differ only in the type of cation functional group illuminate the importance of using an imidazolium cation group instead of a trimethylamine to boost CO2RR performance (i.e., higher CO selectivities, lower cell voltage, and improved cell stability). We also demonstrate that an imidazolium group is more chemically stable than a trimethylamine analog, especially at the high current densities relevant to commercial operation (j = 100 mA/cm2). Our results further show that AEMs exhibiting a high water uptake are susceptible to excessive salt precipitation at the cathode, which is detrimental to CO2RR performance. This study motivates the work described in Chapter 4, where a detailed investigation on how AEMs with different water uptakes and thicknesses modulate the amount of water transported to the cathode and, ultimately, CO2RR performance. 25 3.4 Experimental methods Ethanol (100%, ACS grade) was purchased from Greenfield Global Inc. Nafion 117 solution (5 wt% in a mixture of lower aliphatic alcohols and water) and silver nanopowder (trace metal basis, >99%) were purchased from Sigma-Aldrich. Carbon paper (TGP-H-120) and PTFE gaskets (0.01”) were purchased from Fuel Cell Store. Nickel foam gas diffusion layer material was purchased from MTI (EQ-BCNF-16m). 3.4.1 Electrode preparation I prepared the Ag cathode catalyst inks by sonicating 32 mg of silver nanopowder, 800 μL of deionized water, 800 μL of isopropyl alcohol, and 60 μL Nafion 117 solution similar to a previous report.9 I coated eight 4-cm2 carbon papers by using an airbrush technique and subsequently dried using an infrared lamp89 Each electrode was prepared to have a Ag mass loading of ~1.5 mg/cm2 after spray coating. I prepared anodes by cutting nickel foam to size (2 × 2 cm2). Nickel foam was cleaned using a standard cleaning procedure of sonication in acetone and then deionized water. A fresh cathode, anode, and AEM were used for each electrolysis test. 3.4.2 Membrane characterization Ion exchange capacity I soaked four samples of each membrane (in Cl– form) in individual containers containing 30 g of 1.0 M KNO3 (m1 M KNO3) for 24 h to exchange chloride counter ions to nitrate. The concentration of chloride ([Cl–] in ppm) released in the solution was determined by using an ion-selective electrode (Cl– ISE, Cole Parmer on Accumet AE 150 pH meter, Fisher Scientific). The membranes were then soaked in 1.0 M KCl for 24 h to exchange the nitrate counter ions back to chloride. I thoroughly washed each sample with DI water before soaking in DI water for an additional 24 h. The DI water was replenished three times over this period. The membrane samples 26 were then dried at room temperature for 24 h, and the mass of the dried membranes (md in g) were recorded using an analytical balance. I calculated the ion exchange capacity (IEC) of the membranes using the following expression, where MCl is the molar mass of chloride (g/mol). 𝐼𝐸𝐶 = [𝐶𝑙–]⋅𝑚1𝑀 𝐾𝑁𝑂3𝑀𝐶𝑙⋅𝑚𝑑 (Eq. 3.1) Water uptake Four samples of each membrane were soaked in 1.0 M KCl at room temperature for 24 h. The membranes were thoroughly washed with DI water to remove excess KCl from the membrane and soaked in DI water for an additional 24 h at room temperature. The water was replaced with fresh DI water three times over this period. The membrane samples were then removed from the water, surface water was gently removed with a paper tissue, and the mass of each hydrated membrane (mh in g) was immediately measured using an analytical balance. The hydrated membranes were then dried at room temperature for 24 h, and the mass of the dried membranes (md in g) were recorded. The water uptake (WU) values were calculated as follows: 𝑊𝑈 (%) = 100% ×𝑚ℎ−𝑚𝑑𝑚𝑑 (Eq. 3.2) 3.4.3 Product analysis CO2RR experiments were performed using a potentiostat (CH instruments 660D with a picoamp booster) through a two-electrode set-up at ambient temperature and pressure. The CO2 electrolyzer cathode outlet stream passed through a condenser before entering the gas chromatograph (GC, Perkin Elmer; Clarus 580) to avoid liquid damage of the equipment. The GC was calibrated at different gas concentrations as previously reported.90 The flow rate of CO2 entering the cell was measured using a CO2 mass flow controller (AALBORG) and was used as the volumetric flow rate in Equation. 3.3. The faradaic efficiency (FE) of a gaseous product k was determined in accordance with Equation 3.3. 27 𝐹𝐸 = 𝑛𝑘𝐹𝑥𝑘𝐹𝑚𝐼 (Eq. 3.3) where nk is the number of electrons exchanged for CO2RR, F is Faraday’s constant (96,485 C/mol), xk is the mole fraction of the gas k in the gaseous mixture determined from the peak areas from the GC peak, Fm is the molar flow rate in mol/s, and I is the total current in A. The molar flow rate is derived from the volume flow rate Fv by the relation: 𝐹𝑚 = 𝑃𝐹𝑣/𝑅𝑇 (Eq. 3.4) with P being the atmospheric pressure in Pa, R the ideal gas constant of 8.314 J/mol K, and T is the temperature in K. 3.4.4 Electrochemical impedance spectroscopy measurements Ex-situ electrochemical impedance measurements (EIS) were performed using a 2-probe technique by immersing a conductivity cell in a beaker containing 1.0 M KOH. The conductivity cell consists of two perforated stainless steel cathodes (surface area = 2.01 cm2) that serve to sandwich the membrane. EIS measurements were performed over a frequency range of 10 MHz to 1Hz and a sinusoidal amplitude of 10 mV. The electrolyte resistance (Rs) was subtracted from the measured total resistance that included both the membrane resistance (Rmem) and Rs. The OH– conductivity of the membranes was calculated from Rmem (in Ω) obtained from the high-frequency data in the Nyquist plot, the contact area between the electrodes (A in cm2) and the thickness of the membrane (L in cm): 𝜎 =𝐿𝑅𝑚𝑒𝑚⋅𝐴 (Eq. 3.5) 28 Chapter 4: Managing hydration at the cathode enables efficient CO2 electrolysis at commercially relevant current densities Adapted from: Reyes, A.; Jansonius, R.P.; Mowbray, B.A.W., Cao,Y.; Wheeler, D.G.; Chau, J.;Dvorak, D.J.; Berlinguette, C.P. ACS Energy Lett. 2020, 5, 1612–1618 4.1 Introduction There are many electrochemical reactor designs proposed to selectively reduce CO2 at high reaction rates and energy efficiencies,91–93 yet there are no low temperature CO2 electrolyzers available to the market. A range of techno-economic studies have highlighted that a commercially relevant CO2 electrolyzer needs to drive the CO2 reduction reaction (CO2RR) at current densities (j) ≥ 200 mA/cm2 while maintaining high faradaic efficiencies (FEs > 90%) at low cell voltages (Ecell < 3 V).5,94,95 Zero-gap CO2 reactors (i.e., reactors that position the porous electrodes in contact with an anion exchange membrane) offer promise towards reaching these performance metrics;8,12,76 however, selecting suitable materials and operating conditions to achieve this performance remains a challenge. In this Chapter, I highlight the importance of water management in the design of CO2 electrolyzers. Water is critical to facilitating ion transport between the anode and cathode and is a proton source for CO2RR, yet excess water can starve the cathode of CO2 and promote the competing hydrogen evolution reaction (HER). This accumulation of excess water at the cathode (i.e., “flooding”) is especially important at the high j relevant to commercial operation. While flooding is known to exist in other CO2 electrolyzer types,45,69,96,97 it is generally not explicitly considered in a zero-gap design, which is more aligned with commercial electrochemical cells.3 Our experiments show that cathode flooding is an issue at j values < 200 mA/cm2, and is responsible for substantial decreases in product selectivity and electrolyzer energy efficiency. This observation illuminates the importance of water management in a CO2 electrolyzer. 29 A zero-gap CO2 electrolyzer contains a membrane electrode assembly (MEA) consisting of porous anode and cathode gas diffusion electrodes (GDEs) in contact with opposite sides of an ion-exchange membrane. In gas-fed CO2 reactors, the mass transport limitations inherent to aqueous feedstocks are obviated by continuously delivering gaseous CO2 to the cathode, which then diffuses through the GDE for reaction at the membrane/catalyst interface.3,9,12 Anion exchange membranes (AEMs) facilitate the flow of OH– and water between the anode and the cathode, and have been used to achieve record-setting CO2RR performance.7,8 While the zero-gap configuration is known to enable these high CO2 conversion rates, there are currently no empirical studies on cathode flooding in zero-gap reactors, and very little has been reported on how water delivery to the cathode influences CO2RR performance in any reactor architecture.69,76,98 This knowledge gap is striking considering that water is a reactant for all CO2RR reactions, and how it is delivered to the cathode will almost certainly affect CO2RR chemistry. Herein, I experimentally quantify the extent of cathode flooding in zero-gap CO2 reactors and demonstrate a materials design approach to manage water effectively in these electrolyzers. I first show that cathode flooding is indeed an issue in zero-gap reactors, and that this flooding occurs at j values of merely 200 mA/cm2. Importantly, I was able to link this cathode flooding to decreases in faradaic efficiency for CO production (FECO), and increases in Ecell and faradaic efficiency for H2 (FEH2). I drew these conclusions by designing a gravimetric method to measure water flux to the cathode during electrolysis. Cathode flooding occurs when the water transported from the anode to the cathode is in excess (i.e., >5 mg/cm2⋅h) of the required amount of water for CO2RR. I discovered that flooding can be mitigated by using thin (≤40 μm) membranes with low water uptake, in tandem with a hydrophobic cathode. Effective water management yields more 30 selective and efficient CO2 electrolysis at commercially-relevant current densities (≥100 mA/cm2) and therefore must be considered in CO2 electrolyzer design. 4.2 Results 4.2.1 Design of zero-gap CO2 electrolyzer and MEA materials All CO2 reduction experiments in this study were performed in a zero-gap reactor configuration with a MEA containing an AEM sandwiched between a silver cathode GDE and a nickel foam anode (Figure 4.1a; see section 4.4.6 for detailed description).9 The electrolyzer consists of: (i) stainless steel end plates (containing reactant and product inlet and outlets); (ii) flow field plates to deliver CO2 to the reactive site; and (iii) an MEA with a 2×2 cm2 active area. Gaseous CO2 delivered to the cathode was humidified by flowing the inlet stream through a water bath (heated at 50 ºC) while the anode was supplied with 1.0 M KOH (Figure 4.1b). Chronopotentiometry experiments were performed by applying a constant j of 25, 50, 100, or 200 mA/cm2 for 25 min each while in-line gas chromatography (GC) analyses of the cathode outlet gas were taken after 800 s of electrolysis to determine the FECO and FEH2 (See section 4.4.6 for details). I used this electrolyzer configuration to test a series of custom-made AEMs and GDEs and measured the resulting CO2RR performance metrics and water flux to the cathode. 31 Figure 4.1 (a) Expanded and (b) cross-sectional diagrams of the gas-fed zero-gap CO2 reactor used in this study. The membrane electrode assembly consists of a nickel foam anode, anion exchange membrane, and Ag/polytetrafluoroethylene (PTFE) on carbon paper as the cathode. The Sustainion-inspired AEMs in this study consisted of a hydrophobic polystyrene backbone functionalized with hydrophilic tetramethylimidazolium ionic groups.7,61,75,76,99 Using this molecular platform, we were able to vary the water uptake of the membrane by manipulating the ionic group concentration. The polystyrene backbone was prepared by radical polymerization of styrene (m) with 4-vinylbenzyl chloride (n) and then the benzylic chloride was substituted with the 1,2,4,5-tetramethylimidazolium ionic group in a second reaction. Adjusting the ratio of the styrene (m) to imidazolium-containing 4-vinylbenzyl (n) monomers used in the polymerization step yielded membranes with an m:n ratio of 1:1 and 3:1. Ionomers were then cast from the reaction solution into membranes of desired thicknesses (20 μm, 40 μm, and 55 μm) by doctor blade coating (Table 4.1). The ion exchange capacity (IECs) of each membrane was recorded as a measure of imidazolium group concentration. Membranes made from ionomers with m:n values of 1:1 and 3:1 exhibited IECs of 2.13 + 0.09 mmol/g and 1.43 + 0.05 mmol/g, respectively. I also measured 32 the relative mass of water each membrane absorbed, denoted as water uptake (WU). The WUs of the AEMs were found to increase with IEC: the membrane made from 1:1 (m:n) ionomer yielded a WU value of 144 + 8%, while the membrane from 3:1 (m:n) ionomer yielded a WU value of 43 + 4% (Figure BA4.1). I therefore define membranes made from 1:1 and 3:1 (m:n) ionomers to be high water uptake membranes (HWUM) and low water uptake membranes (LWUM), respectively. Cathode GDEs of varying hydrophobicity were prepared by modulating the amounts of polytetrafluoroethylene (PTFE) in the catalyst ink. These GDEs were prepared by ultrasonically spray coating catalyst inks containing a constant amount of commercially-available silver nanopowder and varying amounts of PTFE (6, 17, and 29 wt%) onto carbon paper. A silver loading of ~1 mg/cm2 was obtained regardless of PTFE content and was confirmed by a constant X-ray fluorescence (XRF) Ag signal intensity of ~55 cps (Figure B4.2).84 4.2.2 MEA materials influence water flux and CO2RR performance I next designed experiments to show how the MEA materials (i.e., AEM ionomer composition, AEM thickness and GDE hydrophobicity) influence water flux and performance in a CO2 electrolyzer. I used a water trap filled with anhydrous desiccant (Drierite) placed at the cathode outlet to capture expelled liquid and gaseous water (JH2O,outlet in mg/cm2⋅h) while CO2 electrolysis was performed at a constant j value (25, 50, 100, or 200 mA/cm2) for 25 min. The water trap was weighed immediately before and after electrolysis and I took the change in mass of the trap divided by the electrolysis time (25 min) and active area (4 cm2) to correspond to JH2O,outlet (in mg/cm2⋅h). I also measured the flux of water supplied to the cathode inlet (i.e., before the humidified CO2 enters the cell; JH2O,inlet) using the same technique after flowing 100 sccm of humidified CO2 through the cell for 25 min. The change in mass of the water trap was used to calculate JH2O,inlet. I define the excess flux of water to the cathode (JH2O,cathode) according to Eq. 4.1: 33 JH2O,cathode = JH2O,outlet - JH2O,inlet (Eq. 4.1) Using this water-trap method for measuring water flux across an MEA in a zero-gap CO2 electrolyzer, I quantified JH2O,cathode for reactors containing 55-μm thick HWUMs and LWUMs at each j value of 25, 50, 100, or 200 mA/cm2. The HWUM yielded consistently positive JH2O,cathode values that increased with current density (Figure 4.2a). The experiments with the LWUM, however, yielded negative JH2O,cathode values from 25 to 100 mA/cm2, but sharply increased to 11.8 mg/cm2⋅h at 200 mA/cm2. These results show that LWUM can suppress JH2O,cathode at lower j, but promotes flooding at 200 mA/cm2. I also measured the CO2RR performance (i.e., FECO, FEH2 and Ecell) of the MEA materials, and found that the choice of membrane (i.e., HWUM and LWUM) had minimal effects at j ≤100 mA/cm2. At higher current densities (200 mA/cm2), however, the HWUM yielded an FECO of >80% while the LWUM yielded a much lower FECO of ~50% (Figure 4.2c). These differences correlate with higher FEH2 values (15.2%) with the LWUM compared to the HWUM (<1%) (Figure 4.2d). Moreover, the Ecell values were consistently higher with the LWUM, particularly at the highest measured j of 200 mA/cm2 (Figure 4.2b). 34 Figure 4.2 (a) The water flux to the cathode (JH2O,cathode) after applying current densities (j) of 25, 50, 100, and 200 mA/cm2 for MEAs with 55-μm thick HWUM (green) and LWUM (orange). (b) Total cell voltage (Ecell) (c) CO faradaic efficiency (FECO), and (d) H2 faradaic efficiency (FEH2) at applied current densities (j) of 25, 50, 100, and 200 mA/cm2. Error bars represent the standard deviation for three independent experiments. Given that membranes with high water uptake (i.e., >100%), suffer from poor mechanical strength due to swelling,63,70 we sought to improve the performance of LWUM by decreasing the thickness of the membrane from 55 μm down to 40 and 20 μm. I quantified JH2O,cathode as a function of LWUM thickness and found that JH2O,cathode values measured at 200 mA/cm2 decreased with decreasing thicknesses (Figure 4.3a). The best CO2RR performance (i.e., highest FECO and lowest Ecell) was observed for the thinnest membranes at 200 mA/cm2 (Figure 3b-d). 35 Figure 4.3 Effect of LWUMs with different thicknesses (20, 40 and 55 μm) on the flux of water to the cathode (JH2O,cathode) and CO2RR performance. (a) JH2O,cathode after applying a current density (j) of 200 mA/cm2. (b) CO faradaic efficiencies (FECO), (c) H2 faradaic efficiencies (FEH2) and (d) total cell voltage (Ecell) at j of 25, 50, 100, and 200 mA/cm2. Error bars represent the standard deviation for three independent experiments. For the last set of experiments, I determined how JH2O,cathode and CO2RR performance are affected by the hydrophobicity of the cathode. To illustrate these effects, I tested cathodes with lower PTFE loadings (6 wt% PTFE) and higher PTFE loadings (29 wt% PTFE) relative to the cathodes in the above experiments (17 wt% PTFE). The lowest value of JH2O,cathode at 200 mA/cm2 (1.5 mg/cm2⋅h) was observed with the highest PTFE cathodes (Figure 4.4a). These high PTFE cathodes also yielded the best CO2RR performance. The differences in cathode hydrophobicities did not affect CO2RR at j <50 mA/cm2, but differences at 100 mA/cm2 were statistically meaningful, and even more substantial at 200 mA/cm2 (Figure 4b-d). A higher PTFE content at 36 the cathode (44 wt%) in tandem with the thinnest membrane (20 μm) did not lead to further improvements in the Ecell and FECO. Figure 4.4 Effect of cathode hydrophobicity (manipulated by varying the PTFE content from 6-29 wt%) on the flux of water to the cathode (JH2O,cathode) and CO2RR performance. (a) JH2O,cathode after applying a current density (j) of 200 mA/cm2. (b) CO faradaic efficiencies (FECO), (c) H2 faradaic efficiencies (FEH2) and (d) total cell voltage (Ecell) at j of 25, 50, 100, and 200 mA/cm2. Error bars represent the standard deviation for three independent experiments. Finally, we constructed a 1D model to support our empirical investigation of water flux to the cathode. This model was constructed in COMSOL (see Appendix A4 for details) and calculated the flux of H2O by diffusion (JDIFF), back-convection (JBC), and electro-osmotic drag (JEOD) from the anode/AEM interface to the cathode flow channel (See Figure B4.11 for details). We used this simple model to determine the influence of membrane thickness and cathode hydrophobicity on these individual fluxes (i.e., JDIFF, JBC and JEOD) that comprise the net flux of water across the 37 AEM (JH2O,net). Our model found membrane thickness to have the most dramatic influence on the flux of water to the cathode (i.e., JBC increases for thinner membranes), supporting our empirical observation that thinner membranes can mitigate flooding in a zero-gap electrolyzer. 4.3 Discussions In this work, I show that cathode flooding occurs in gas-fed CO2 reactors at modest current densities (j ≤200 mA/cm2). Accumulation of water at the cathode is undesirable because it can block pores in the GDE and impede the transport of gaseous CO2RR reagents and products to and away from the catalytic sites. These problems lead to poor electrolyzer performance because CO2-limited active sites favour the parasitic HER that occurs at a lower onset potential than the desired CO2RR.45 It is therefore important to manage the flux of water to the cathode in order to realize efficient CO2 electrolysis in zero-gap reactors. Inspired by fuel cell studies,41,44 I performed water mass balance experiments under controlled operating conditions (i.e., cell temperature, cell compression, flow rates and relative humidity of CO2 and anolyte feedstock) to resolve how JH2O, cathode is affected by j, AEM water uptake and thickness, and GDE hydrophobicity. Under steady-state operation, the water mass balance at the membrane/cathode interface is expressed in Eq. 4.2, JH2O,oulet - JH2O, inlet = JH2O,net - JH2O,reduction (Eq. 4.2) where JH2O,outlet is the water flux from the cathode outlet, JH2O,inlet is the water flux to the cathode inlet, JH2O,net is the net flux of water permeating from the anode through the membrane, and JH2O,reduction is the water consumed by the reduction reactions at the cathode (Figure 4.5a). Combining Equations 4.1 and 4.2 yields the relationship between the reported JH2O,cathode values and JH2O,net and JH2O,reduction: JH2O,cathode = JH2O,net - JH2O,reduction (Eq. 4.3) 38 I show that JH2O,reduction was effectively the same for each zero-gap configuration at the same j (Tables A4.1 and A4.2); thus, I infer that JH2O,cathode is directly correlated to JH2O,net under the same operating conditions. Figure 4.5 (a) Illustration of the water mass balance at the AEM/cathode interface that accounts for water from the humidified cathode inlet CO2 stream (JH2O,inlet); expelled at the cathode outlet (JH2O,outlet); consumed from CO2RR (JH2O,reduction); and from the anode that permeated through the membrane (JH2O,net). (b) JH2O,net is the sum of diffusion (JDIFF), electro-osmotic drag (JEOD), and back convection (JBC) processes. The JH2O,net in zero-gap reactors can be described by three key processes: i) diffusion (JDIFF); ii) electro-osmotic drag (JEOD); and iii) back convection (JBC) (Figure 4.5b).47 JH2O,net = JDIFF - JEOD - JBC (Eq. 4.4) For our experiments, JDIFF describes water transport from the anode to the cathode driven by the water concentration gradient between the aqueous anode and the gas-fed cathode. JEOD represents water transported from the cathode to the anode due to the flow of solvated anions,46 and JBC describes water repelled by the hydrophobic cathode towards the anode (as well as differences in water permeation through the membranes).52,100 The individual contributions of JH2O,net (i.e., JDIFF, JEOD, JBC) are functions of the water transport properties of the MEA components (i.e., gas 39 diffusion electrodes and the membrane).101 It is for this reason that we designed AEMs with varying water uptakes and thicknesses, and cathodes with varying hydrophobicities. Our studies show that AEM water uptake has a profound effect on JH2O,cathode and, in turn, CO2RR performance, particularly at high current densities. At 200 mA/cm2, LWUM yields a ~30% lower FECO while also requiring an additional 350 mV to drive CO2 electrolysis compared to HWUM (Figure 4.2b, 3c). We assert that these differences are due to cathode flooding, which is more significant for LWUM than HWUM at 200 mA/cm2 (Figure 4.2a). The shift in product selectivity from CO to H2 for LWUM at 200 mA/cm2 (Figure 4.2c, 4.2d) is also consistent with cathode flooding, where a high concentration of H2O would favour HER over CO2RR.45 From these experiments, I define that cathode flooding occurs when the flux of water delivered from the anode exceeds the required amount of water for CO2RR by more than 5 mg/cm2⋅h (i.e., JH2O,net - JH2O,reduction = JH2O,cathode >5 mg/cm2⋅h). Cathode flooding occurs with LWUM, presumably because the electroosmotic drag coefficient of the membrane (η) is reduced (i.e., lower JEOD) for membranes with lower water uptakes.49 By reducing the thickness of the LWUM, I found that the CO2RR performance can be significantly improved (i.e., a 30% increase in FECO and a 450 mV decrease in Ecell). Thinner membranes seemingly act to mitigate flooding (Figure 4.3a, 4.3c), presumably because of a higher back convection flux (JBC). This assertion is aligned with predictions from our 1D numerical model of the MEA (see Appendix A for more details). Fuel cell studies have also shown that thinner membranes decrease the water flux from the anode to the cathode.52 While multiphysics modelling of a gas-fed CO2 electrolyzer predicts that membrane thickness is not expected to impact FECO,98 our empirical data show that decreasing the membrane thickness does indeed improve FECO at high current densities (200 mA/cm2). Decreasing the thickness of an LWUM could be a more 40 desirable design strategy to mitigate cathode flooding than using high water uptake membranes because HWUMs exhibit poor mechanical stability.70 Finally, increasing the cathode hydrophobicity also improved CO2RR performance by suppressing cathode flooding, evidenced by the lower JH2O,cathode (Figure 4.4a) and minimal FEH2 (Figure 4.4c). I observed that the most hydrophobic cathode (29 wt% PTFE) yielded statistically meaningful improvements in CO2RR performance (Figure 4.4b, 4.4d) at j ≥100 mA/cm2. I assert that the cathode with 29 wt% PTFE is able to alleviate cathode flooding because the back convection flux (JBC) is higher for more hydrophobic GDEs.100 Our results again align with fuel cell literature that shows hydrophobic cathodes decrease the water flux from the anode to the cathode.44 I note that catalyst layer morphology and ECSA were not significantly influenced by PTFE content (see Figures A4.7 through A4.10 for details). 4.3 Conclusions Efficient electrocatalytic reduction of CO2 in a zero-gap reactor requires a proton source, which is usually water. Providing too little water (which starves the electrode of a reagent) or too much water (which prevents CO2 from reaching the catalyst) to the cathode compromises the energy efficiency and selectivity for CO2RR products. Our investigations provide the first empirical evidence that shows the factors affecting the flux of water to the cathode (JH2O,cathode). Moreover, water mass balance experiments demonstrate that cathode flooding is an issue at j ≤200 mA/cm2, which is much less than values as high as 750 mA/cm2 that have been previously predicted.98 We also show that using thin, low water uptake membranes with hydrophobic cathodes suppresses flooding, even at high current densities (200 mA/cm2). The water management strategies that we present in this study aim to inform the design of membranes and GDEs to optimize water management in zero-gap CO2 electrolyzers. 41 4.4 Experimental methods 4.4.1 Materials Reagents were used as received unless otherwise specified. Styrene (99%, stabilized with TBC) and 1,2,4,5-tetramethylimidazole (>98%) were purchased from TCI America. 4-Vinylbenzyl chloride (90%, stabilized with TBC), 2,2′-azobis(2-methylpropionitrile) (98%), toluene (ACS grade), methanol (ACS grade), chlorobenzene (ACS grade), and diethyl ether (ACS grade) were obtained from Sigma Aldrich. Ethanol (100%, ACS grade) was purchased from Greenfield Global Inc. PTFE dispersion (PTFE DISP 30, 60% in non-ionic surfactant), carbon paper (Sigracet 39BC), and PTFE gaskets (0.01”) were purchased from Fuel Cell Store. Nickel foam gas diffusion layer material was purchased from MTI (EQ-BCNF-16m). Sustainion X37-50 Grade RT anion exchange membrane was purchased from Dioxide Materials. The reference electrodes used for ECSA measurements were Ag/Ag Cl (3M NaCl) purchased from Bioanalytical Systems Inc. Styrene (99%, stabilized with TBC) and 1,2,4,5-tetramethylimidazole (>98%) were purchased from TCI America. 4-Vinylbenzyl chloride (90%, stabilized with TBC), 2,2′-azobis(2-methylpropionitrile) (98%), toluene (ACS grade), methanol (ACS grade), chlorobenzene (ACS grade), and diethyl ether (ACS grade) were obtained from Sigma Aldrich. 1H NMR spectra were collected on a 400 MHz Bruker Avance 400inv spectrometer with TopSpin 2.1 software at ambient temperature. 4.4.2 Electrode preparation We prepared the Ag cathode catalyst inks by sonicating 0.1575 g of silver nanopowder, 7.5 mL of deionized water, 7.5 mL of isopropyl alcohol, and 0.12, 0.42, or 0.84 mL of PTFE DISP 30 (60 wt% diluted to 5 wt%). We then coated the catalyst inks on eight 4-cm2 carbon papers using an automated ultrasonic spray coater (ExactaCoat, SONO-TEK) equipped with 1-Watt ultrasonic 42 nozzle operated at a frequency of 120 Hz. The catalyst ink delivery rate during spray coating was fixed at 0.6 mL/min. Each electrode was prepared to have a Ag mass loading of ~1 mg/cm2 after spray coating, which was confirmed by X-ray fluorescence spectroscopy (XRF) using a calibration between Ag mass loading and corresponding XRF values. We then prepared anodes by cutting nickel foam (pressed at 100 psi before cutting) to size (2 × 2 cm2). Nickel foam was cleaned using a standard cleaning procedure of sonication in acetone and then deionized water. 4.4.3 Ionomer synthesis Materials and reagents: Reactions carried out under nitrogen utilized standard Schlenk line and inert atmosphere techniques. 1H NMR spectra were collected on a 400 MHz Bruker Avance 400inv spectrometer with TopSpin 2.1 software at ambient temperature. Samples for NMR analysis were prepared by dissolving ~10 mg of dried solid with the selected solvent and transferring the sample into a 5 mm borosilicate NMR tube. Compounds A1 and B1 were dissolved in acetone-d6 while A2 and B2 were dissolved in methanol-d4. Chemical shifts reported in ppm were calibrated against the residual solvent signals δ 2.09 for acetone-d6 and δ 3.34 for methanol-d4. Preparation of A1: Inhibitor free styrene (20.60 g, 19.78 mmol) and 4-vinylbenzyl chloride (4-VBC) (29.62 g, 19.41 mmol) were prepared by washing the reagents in a 250 ml separatory funnel three times with 100 ml of 0.5 M KOH until the aqueous phase became colourless, indicating the complete removal of the catechol inhibitor. The mixture was further washed three times with 50 ml of distilled water and once with 100 ml NaCl (24%, w/w) before drying over anhydrous MgSO4 for 2 hours. Styrene and 4-VBC were filtered and transferred to a 250 ml round-bottom flask along with 40 ml of chlorobenzene. The flask was purged for 30 minutes with N2 before the initiator AIBN (0.540 g, 0.00329 mol) dissolved in 10 ml of 43 chlorobenzene was added using a syringe. The reaction mixture was stirred and heated to 65 °C with an oil bath and the polymerization proceeded for 96 h under a constant N2 flow. The resulting product was precipitated in large access of MeOH (2 L) and 45.89 g (90%) of a white solid was collected and dried under vacuum. Yields of >90 % were routinely achieved for batch sizes up to 60 g. 1H NMR (400 MHz, acetone-d6, ppm) δ: 7.39-6.66 (m, 9H), 4.64 (m, 2H), 1.89-1.58 (m, 6H). Preparation of B1: The procedures described for A1 were repeated to make B1, adjusting only the relative amounts of reagents used. Styrene (45.40 g, 43.59 mmol) and 4-VBC (23.45 g, 15.37 mmol) were washed free of inhibitor and combined with AIBN (0.550 g, 3.36 mmol) in chlorobenzene (~50 ml) and the reaction was left to proceed for 96 h under constant N2 flow. The product was isolated by adding the viscous reaction solution dropwise to 2 L of MeOH under constant stirring and the resulting precipitate was collected by vacuum filtration. The brittle white solid [65.12 g (94%)] was collected and dried under vacuum.1H NMR (400 MHz, acetone-d6, ppm) δ: 7.40-6.67 (m, 19H), 4.63 (m, 2H), 1.92-1.51 (m, 12H). Preparation of A2: 11.520 g of A1 was added to 2.583 g (0.020.64 mmol) of 1,2,4,5-tetramethylimidazole (TMIM) in a 100 ml round-bottom flask with 16.242 g and 10.962 g of anhydrous solvents EtOH and toluene respectively. The flask was purged for 30 minutes with N2 and heated to 78 °C over 2 hours under constant stirring until the contents dissolved completely and the solution turned clear. The temperature was reduced to 55 °C and the reaction was left to proceed for 72 hours before it was removed from the heat source. Once cooled to room temperature, the viscous, pale-yellow solution was transferred to a polypropylene bottle for storage. Rinsings from the reaction flask were precipitated with diethyl ether and dried before 1H NMR analysis, which indicated approximately 50% substitution of the benzylic chloride groups, 44 based on the integration of distinct substituted and unsubstituted benzylic proton signals. Unreacted TMIM was observed as sharp singlets at similar chemical shifts as the broad peaks of TMIM groups incorporated into the polymer. 1H NMR (400 MHz, methanol-d4, ppm) ẟ: 7.05-6.53 (m, 9H), 5.22 (m, 1H), 4.52 (m, 1H), 3.68 (m, 3H), 2.55 (m, 3H), 2.23 (m, 3H), 2.06-1.48 (m, 15H). Preparation of B2: The procedures described for A2 were repeated to synthesize B2. 13.395 g of B1 was reacted with 5.114 g (41.18 mmol) of TMIM in a mixture of anhydrous EtOH (22.170 g) and toluene (13.592 g) for 72 hours. The resulting viscous solution was transferred to a separate container and the residual material in the reaction flask was precipitated, dried, and a 1H NMR spectrum was collected. 1H NMR (400 MHz, methanol-d4, ppm) δ: 7.06-6.50 (m, 9H), 5.25 (m, 1H), 4.56 (m, 1H), 3.70 (m, 3H), 2.55 (m, 3H), 2.28 (m, 3H), 2.08 (m, 3H), 1.45 (m, 6H). 4.4.4 Membrane fabrication The membranes were prepared by directly casting each reaction solution onto borosilicate substrates by doctor blade coating. Borosilicate substrates (8” x 4” x ⅛”, McMaster-Carr) were rinsed with ultrapure water, dried in air, then cleaned with ethanol immediately prior to casting the ionomer dispersion. We placed the substrate onto the coating table (MSK-AFA-II, MTI), and set the doctor blade spacing depending on the desired thickness of the membrane (Table S1). The ionomer dispersion was poured (~8 mL) onto the substrate within ~5 mm of the doctor blade, and the blade was actuated at a speed of 18 mm/s. The membrane was immediately placed in an oven preheated at a targeted temperature and duration depending on the membrane (Table 4.1) to remove the solvent and improve the mechanical properties of the membrane. Membrane thickness was measured upon removal from the oven using an inside micrometre accurate to +1 μm (293-344-30, Mitutuyo). The membrane was subsequently delaminated from the borosilicate substrate 45 by immersing in 1.0 M KOH solution (for >20 min), which caused the membrane to swell, soften, and separate from the rigid substrate. The membrane films were stored in 1.0 M KOH for at least 24 hours prior to testing in a reactor to exchange the chloride counter ions with hydroxide. We stored completed membranes in 1.0 M KCl after casting for long-term storage and ex-situ characterization. 4.4.5 Membrane characterization Ion exchange capacity I soaked four samples of each membrane (in Cl– form) in individual containers containing 30 g of 1 M KNO3 (m1 M KNO3) for 24 h to exchange chloride counter ions to nitrate. The concentration of chloride ([Cl–] in ppm) released in the solution was determined by using an ion-selective electrode (Cl– ISE, Cole Parmer on Accumet AE 150 pH meter, Fisher Scientific). The membranes were then soaked in 1.0 M KCl for 24 h to exchange the nitrate counter ions back to chloride. I thoroughly washed each sample with DI water before soaking in DI water for an additional 24 h. The DI water was replenished three times over this period. The membrane samples were then dried at room temperature for 24 h, and the mass of the dried membranes (md in g) were recorded using an analytical balance. I calculated the ion exchange capacity (IEC) of the membranes using the following expression, where MCl is the molar mass of chloride (g/mol). 𝐼𝐸𝐶 = [𝐶𝑙–]⋅𝑚1𝑀 𝐾𝑁𝑂3𝑀𝐶𝑙⋅𝑚𝑑 (Eq. 4.5) Water uptake Four samples of each membrane were soaked in 1.0 M KCl at room temperature for 24 h. The membranes were thoroughly washed with DI water to remove excess KCl from the membrane and soaked in DI water for an additional 24 h at room temperature. The water was replaced with fresh DI water three times over this period. The membrane samples were then removed from the 46 water, surface water was gently removed with a paper tissue, and the mass of each hydrated membrane (mh in g) was immediately measured using an analytical balance. The hydrated membranes were then dried at room temperature for 24 h, and the mass of the dried membranes (md in g) were recorded. The water uptake (WU) values were calculated as follows: 𝑊𝑈 (%) = 100% ×𝑚ℎ−𝑚𝑑𝑚𝑑 (Eq. 4.6) 4.4.6 Electrochemical measurements and product analysis The electrolyzer consists of end plates, flow field plates, gaskets, and the MEA. We used a fresh Ag spray-coated cathode, Nickel foam anode, and AEM for each electrolysis test. Two stainless steel (SS 316) end plates served as inlets and outlets for the reactants and products. A titanium (grade 2) cathode and stainless steel (SS 316) anode flow field plates were machined with serpentine channels that are 1.5 mm wide and 1.5 mm deep with 1 mm wide landings. The flow plates sandwiched the MEA (geometric active area = 2 × 2 cm2) and served to facilitate the delivery of reactants and products to and away from the reactive interface. The MEAs comprised a silver-coated cathode GDE (2 × 2 cm2), an AEM (3 × 3 cm2), and a nickel foam anode (2 × 2 cm2). The cathode and AEM components of the MEAs were manipulated as described below, and the anodes were unchanged for all experiments. The entire assembly was compressed by stainless steel end plates using eight, evenly spaced fasteners. A constant torque of 3 N⋅m was applied to each fastener to provide a fixed compression to the MEA and to minimize interfacial contact resistances. The gaseous CO2 was humidified by flowing the inlet stream through a water bath (heated at 50 ºC) at 100 sccm by a CO2 mass flow controller (GFC17, AALBORG) before entering the cathode; the anode was supplied with 1.0 M KOH at 15 mL/min using a peristaltic pump (43205K11, McMaster-Carr). Chronopotentiometry experiments were performed by applying a constant j of 25, 50, 100, or 200 mA/cm2 to the entire cell for 25 min using a power supply (2280S-32-6, 47 Keithley). A Mandel gas chromatograph (SRI-8610C) was used to analyze the cathode outlet gas. Complete CO2 electrolysis performance characterization took place over 2 hours of sustained electrolysis (i.e., 25 minutes each at 25, 50, 100, and 200 mA/cm2). Current density is expressed as the total current divided by the geometric surface area of the electrodes (4 cm2). The faradaic efficiency (FE) of a gaseous product k was determined in accordance with this equation: 𝐹𝐸 = 𝑛𝑘𝐹𝑥𝑘𝐹𝑚𝑙 (Eq. 4.7) where nk is the number of electrons exchanged, F is Faraday’s constant (F = 96,485 C/mol), xk is the mole fraction of the gas k in the gaseous mixture analyzed, Fm is the molar flow rate in mol/s, and I is the total current in A. The molar flow rate is derived from the volume flow rate Fv by the relation 𝐹𝑚 = 𝑝𝐹𝑣/𝑅𝑇, with p being the atmospheric pressure in Pa, R the ideal gas constant of 8.314 J/mol K, and T the temperature in K. 4.4.7 Electrochemical impedance spectroscopy measurements Electrochemical impedance measurements (EIS) were conducted in the flow cell setup using conditions selected to closely-mirror CO2 electrolysis experiments. These measurements were performed using a CH instruments potentiostat (660D) over a frequency range of 0.03 Hz to 1kHz, sinusoidal amplitude of 5 mV, and an operating voltage of -1.75 V. We used the following conditions: i) humidified CO2 gas (85% relative humidity, see above) was fed to the cathode at a rate of 100 sccm; ii) 1 M KOH was flowed through the anode flow plate at a rate of 15 mL/min. A Ni foam anode and a 17 wt % PTFE GDE cathode were used. 48 4.4.8 Electrochemical surface area measurements The double-layer capacitance (Cdl) of the prepared cathodes was determined ex-situ by measuring the capacitive charging current at different scan rates in a non-Faradaic potential window using a conventional 3-electrode electrochemical setup, with the sample immersed in 1 M H2SO4. A 1 × 1 cm2 square of the prepared cathode was used as a working electrode with a Pt wire counter and a Ag/AgCl reference. Cyclic voltammograms were collected at scan rates of 25, 50, 75, 100, 150, and 200 mV/s between 0.2 V and 0.3 V vs Ag/AgCl and the capacitive current was measured at 0.25 V on the third scan. The slope of the straight line resulting from a plot of current versus scan rate yielded the double-layer capacitance. Each 1 × 1 cm2 sample was tested three times, and three 1 × 1 cm2 squares were prepared for each cathode; Cdl values were reported as the mean of triplicate measurements. Table 4.1 Doctor blade spacing and annealing conditions to fabricate ionomer solutions into membranes. Membrane Doctor blade spacinga Annealing conditions Thickness (m:n) in °C/mmHg μm LWUM (3:1) 0.010 70/500 20 0.028 70/500 40 0.035 70/500 55 HWUM (1:1) 0.030 165/ambient 55 ameasured from the surface of the glass substrate 49 Scheme 4.1 Synthesis of intermediate copolymer poly(4-vinylbenzyl chloride-co-styrene), and anion exchange polymer poly[(4-vinylbenzyl-1,2,4,5-tetramethylimidazolium chloride)-co-styrene] Figure 4.6 1H NMR of A1 in acetone-d6 at 400 MHz and ambient temperature. 50 Figure 4.7 1H NMR of B1 in acetone-d6 at 400 MHz and ambient temperature. 51 Figure 4.8 1H NMR of A2 in acetone-d6 at 400 MHz and ambient temperature. 52 Figure 4.9 1H NMR of B2 in acetone-d6 at 400 MHz and ambient temperature. 53 Chapter 5: Conclusions and future directions 5.1 Conclusions This thesis highlights the judicious designs of anion exchange membranes for achieving efficient CO2 reduction reaction (CO2RR) to CO in gas-fed electrolyzers. Chapter 3 demonstrates the link between membrane functional groups and CO2RR selectivity, voltage, and stability. To establish this relationship, two membranes functionalized with imidazolium or a trimethylamine groups were successfully synthesized and fabricated with the same thickness, polymer backbone, and ionic group concentration. The CO2RR performance to CO is improved with the use of an imidazolium group because this membrane exhibits a low water uptake, which is an important property for minimizing undesired salt formation at the cathode. An imidazolium group was also shown to be more stable than a trimethylamine at high current densities. Chapter 4 shows different membrane and cathode design strategies to optimize water management in CO2 electrolyzers. This chapter presents a method for quantifying water flux to the cathode and provides empirical evidence that AEM-based CO2 electrolyzers are susceptible to flooding at current densities lower than theoretically predicted. Chapter 4 shows that the water uptake and thickness of the membrane, and hydrophobicities of the cathode can be used to tune CO2RR selectivity and voltage. In this study, the optimal water management strategy is a thin, low water uptake membrane paired with a hydrophobic cathode. 5.2 Future directions Chapter 3 examines the CO2-to-CO performance of AEMs containing imidazolium or trimethylamine functional groups, but both membranes resulted in limited CO selectivities at high current densities (100 mA/cm2). We can establish new structure-activity relationships at even 54 higher current densities by extending the experiments designed in Chapter 3 to include cation chemistries with improved alkali stability, such as completely substituted imidazolium groups or quaternary ammonium cations with long n-alkyl groups.85,102–104 Future work should use NMR characterization of the membranes or expelled cathode liquid after CO2 electrolysis experiments to confirm stability of cation groups at higher current densities. Our experimental set-up used in Chapter 4 for quantifying the amount of water flux to the cathode can be improved in many ways. In our current setup, the liquid expelled from the cathode outlet is collected using a desiccating agent but can be replaced with a differential pressure gauge. This previous setup is also problematic because liquids exiting the cathode cannot be further analyzed for CO2RR liquid products or presence of ions (e.g. K+). Using a cold trap to capture the cathode drain for further analyses with ICP-MS can provide clues on how water and salt crossover impact product distribution at the cathode. Dry CO2 gas should also be used instead of a humidified feed to provide a more accurate representation of water flux from the anode to the cathode. CO2 electrolysis experiments in this thesis were performed at short periods of time (2 hours), but future work should focus on extending cell testing to better understand how water management at the cathode impacts the stability of the system. Electrochemical experiments in this thesis did not account for gas (e.g., CO, CO2, and H2) permeation through the AEM, but future work should characterize this property as well. This thesis shows that the membrane and cathodic gas diffusion electrode parameters can optimize salt and water crossover from the anode to the cathode. A nickel foam anode was kept unchanged throughout the experiments in this thesis, but other anode designs with tunable parameters (e.g. hydrophilicity, porosity, and electrode composition) can be beneficial in modulating ion and water transport mechanisms. It has been reported in fuel cells that high PTFE 55 coating (up to 40 wt%) on the anodic gas diffusion layer and microporous layers minimize water crossover from the anode to the cathode.105 Modulating the carbon and ionomer compositions at the anode has also shown to improve the balance of water in fuel cells.43 Future work should focus on systematically engineering the anode and providing design principles that increase CO2RR activity at the cathode. 56 References (1) Reyes, A.; Jansonius, R. P.; Mowbray, B. A. W.; Cao, Y.; Wheeler, D. G.; Chau, J.; Dvorak, D. J.; Berlinguette, C. P. Managing Hydration at the Cathode Enables Efficient CO2 Electrolysis at Commercially Relevant Current Densities. ACS Energy Lett. 2020, 1612–1618. (2) Cook, T. R.; Dogutan, D. K.; Reece, S. Y.; Surendranath, Y.; Teets, T. S.; Nocera, D. G. Solar Energy Supply and Storage for the Legacy and Nonlegacy Worlds. Chem. Rev. 2010, 110 (11), 6474–6502. (3) Weekes, D. M.; Salvatore, D. A.; Reyes, A.; Huang, A.; Berlinguette, C. P. Electrolytic CO2 Reduction in a Flow Cell. Acc. Chem. Res. 2018, 51 (4), 910–918. (4) De Luna, P.; Hahn, C.; Higgins, D.; Jaffer, S. A.; Jaramillo, T. F.; Sargent, E. H. What Would It Take for Renewably Powered Electrosynthesis to Displace Petrochemical Processes? Science 2019, 364 (6438). https://doi.org/10.1126/science.aav3506. (5) Jouny, M.; Luc, W.; Jiao, F. General Techno-Economic Analysis of CO2 Electrolysis Systems. Ind. Eng. Chem. Res. 2018, 57 (6), 2165–2177. (6) Verma, S.; Kim, B.; Jhong, H.-R. “molly”; Ma, S.; Kenis, P. J. A. A Gross-Margin Model for Defining Technoeconomic Benchmarks in the Electroreduction of CO2. ChemSusChem 2016, 9 (15), 1972–1979. (7) Kutz, R. B.; Chen, Q.; Yang, H.; Sajjad, S. D.; Liu, Z.; Masel, I. R. Sustainion Imidazolium-Functionalized Polymers for Carbon Dioxide Electrolysis. Energy Technology 2017, 5 (6), 929–936. (8) Yin, Z.; Peng, H.; Wei, X.; Zhou, H.; Gong, J.; Huai, M.; Xiao, L.; Wang, G.; Lu, J.; Zhuang, L. An Alkaline Polymer Electrolyte CO2 Electrolyzer Operated with Pure Water. Energy Environ. Sci. 2019. https://doi.org/10.1039/C9EE01204D. (9) Salvatore, D. A.; Weekes, D. M.; He, J.; Dettelbach, K. E.; Li, Y. C.; Mallouk, T. E.; Berlinguette, C. P. Electrolysis of Gaseous CO2 to CO in a Flow Cell with a Bipolar Membrane. ACS Energy Lett. 2018, 3 (1), 149–154. (10) Phillips, R.; Dunnill, C. W. Zero Gap Alkaline Electrolysis Cell Design for Renewable Energy Storage as Hydrogen Gas. RSC Adv. 2016, 6 (102), 100643–100651. (11) Li, Y. C.; Zhou, D.; Yan, Z.; Gonçalves, R. H.; Salvatore, D. A.; Berlinguette, C. P.; Mallouk, T. E. Electrolysis of CO2 to Syngas in Bipolar Membrane-Based Electrochemical Cells. ACS Energy Lett. 2016, 1 (6), 1149–1153. (12) Ren, S.; Joulié, D.; Salvatore, D.; Torbensen, K.; Wang, M.; Robert, M.; Berlinguette, C. P. Molecular Electrocatalysts Can Mediate Fast, Selective CO2 Reduction in a Flow Cell. Science 2019, 365 (6451), 367–369. (13) Kutz, R. B.; Chen, Q.; Yang, H.; Sajjad, S. D.; Liu, Z.; Masel, I. R. Sustainion Imidazolium-Functionalized Polymers for Carbon Dioxide Electrolysis. Energy Technol. 2017, 5 (6), 929–936. (14) Aeshala, L. M.; Uppaluri, R.; Verma, A. Electrochemical Conversion of CO 2 to Fuels: Tuning of the Reaction Zone Using Suitable Functional Groups in a Solid Polymer Electrolyte. Phys. Chem. Chem. Phys. 2014, 16 (33), 17588–17594. (15) Delacourt, C.; Ridgway, P. L.; Kerr, J. B.; Newman, J. Design of an Electrochemical Cell Making Syngas (CO + H2 ) from CO2 and H2O Reduction at Room Temperature. J. Electrochem. Soc. 2008, 155 (1), B42–B49. (16) Lee, W.; Kim, Y. E.; Youn, M. H.; Jeong, S. K.; Park, K. T. Catholyte-Free Electrocatalytic CO 2 Reduction to Formate. Angew. Chem. Int. Ed. 2018, 56, 11326. (17) Aeshala, L. M.; Rahman, S. U.; Verma, A. Effect of Solid Polymer Electrolyte on Electrochemical Reduction of CO2. Sep. Purif. Technol. 2012, 94, 131–137. (18) Merino-Garcia, I.; Albo, J.; Irabien, A. Productivity and Selectivity of Gas-Phase CO 2 Electroreduction to Methane at Copper Nanoparticle-Based Electrodes. Energy Technol. 2017, 5 (6), 922–928. 57 (19) Pătru, A.; Binninger, T.; Pribyl, B.; Schmidt, T. J. Design Principles of Bipolar Electrochemical Co-Electrolysis Cells for Efficient Reduction of Carbon Dioxide from Gas Phase at Low Temperature. J. Electrochem. Soc. 2019, 166 (2), F34–F43. (20) Hori, Y.; Ito, H.; Okano, K.; Nagasu, K.; Sato, S. Silver-Coated Ion Exchange Membrane Electrode Applied to Electrochemical Reduction of Carbon Dioxide. Electrochim. Acta 2003, 48 (18), 2651–2657. (21) Pan, J.; Han, J.; Zhu, L.; Hickner, M. A. Cationic Side-Chain Attachment to Poly(Phenylene Oxide) Backbones for Chemically Stable and Conductive Anion Exchange Membranes. Chem. Mater. 2017, 29 (12), 5321–5330. (22) Pan, J.; Chen, C.; Zhuang, L.; Lu, J. Designing Advanced Alkaline Polymer Electrolytes for Fuel Cell Applications. Acc. Chem. Res. 2012, 45 (3), 473–481. (23) Arges, C. G.; Kambe, Y.; Dolejsi, M.; Wu, G.-P.; Segal-Pertz, T.; Ren, J.; Cao, C.; Craig, G. S. W.; Nealey, P. F. Interconnected Ionic Domains Enhance Conductivity in Microphase Separated Block Copolymer Electrolytes. J. Mater. Chem. A Mater. Energy Sustain. 2017, 5 (11), 5619–5629. (24) Dischinger, S. M.; Gupta, S.; Carter, B. M.; Miller, D. J. Transport of Neutral and Charged Solutes in Imidazolium-Functionalized Poly (phenylene Oxide) Membranes for Artificial Photosynthesis. Ind. Eng. Chem. Res. 2019, 59 (12), 5257–5266. (25) Krödel, M.; Carter, B. M.; Rall, D.; Lohaus, J.; Wessling, M.; Miller, D. J. Rational Design of Ion Exchange Membrane Material Properties Limits the Crossover of CO2 Reduction Products in Artificial Photosynthesis Devices. ACS Appl. Mater. Interfaces 2020, 12 (10), 12030–12042. (26) Carter, B. M.; Keller, L.; Wessling, M.; Miller, D. J. Preparation and Characterization of Crosslinked Poly(vinylimidazolium) Anion Exchange Membranes for Artificial Photosynthesis. J. Mater. Chem. A Mater. Energy Sustain. 2019. https://doi.org/10.1039/C9TA00498J. (27) Ramírez, S. C.; Paz, R. R. Hydroxide Transport in Anion-Exchange Membranes for Alkaline Fuel Cells. New Trends in Ion Exchange Studies 2018, 51. (28) Grew, K. N.; Chiu, W. K. S. A Dusty Fluid Model for Predicting Hydroxyl Anion Conductivity in Alkaline Anion Exchange Membranes. J. Electrochem. Soc. 2010, 157 (3), B327–B337. (29) Merle, G.; Wessling, M.; Nijmeijer, K. Anion Exchange Membranes for Alkaline Fuel Cells: A Review. J. Memb. Sci. 2011, 377 (1), 1–35. (30) Dong, D.; Zhang, W.; van Duin, A. C. T.; Bedrov, D. Grotthuss versus Vehicular Transport of Hydroxide in Anion-Exchange Membranes: Insight from Combined Reactive and Nonreactive Molecular Simulations. J. Phys. Chem. Lett. 2018, 9 (4), 825–829. (31) Chen, C.; Tse, Y.-L. S.; Lindberg, G. E.; Knight, C.; Voth, G. A. Hydroxide Solvation and Transport in Anion Exchange Membranes. J. Am. Chem. Soc. 2016, 138 (3), 991–1000. (32) Helfferich, F. G. Ion Exchange; Courier Corporation, 1995. (33) Luo, T.; Abdu, S.; Wessling, M. Selectivity of Ion Exchange Membranes: A Review. J. Memb. Sci. 2018, 555, 429–454. (34) Geise, G. M.; Hickner, M. A.; Logan, B. E. Ionic Resistance and Permselectivity Tradeoffs in Anion Exchange Membranes. ACS Appl. Mater. Interfaces 2013, 5 (20), 10294–10301. (35) Le, X. T. Permselectivity and Microstructure of Anion Exchange Membranes. J. Colloid Interface Sci. 2008, 325 (1), 215–222. (36) Hosseini, S. M.; Madaeni, S. S.; Khodabakhshi, A. R. The Electrochemical Characterization of Ion Exchange Membranes in Different Electrolytic Environments: Investigation of Concentration and pH Effects. Sep. Sci. Technol. 2012, 47 (3), 455–462. (37) Długołęcki, P.; Nymeijer, K.; Metz, S.; Wessling, M. Current Status of Ion Exchange Membranes for Power Generation from Salinity Gradients. J. Memb. Sci. 2008, 319 (1), 214–222. (38) Guler, E.; Zhang, Y.; Saakes, M.; Nijmeijer, K. Tailor-Made Anion-Exchange Membranes for Salinity Gradient Power Generation Using Reverse Electrodialysis. ChemSusChem 2012, 5 (11), 2262–2270. (39) Strathmann, H.; Grabowski, A.; Eigenberger, G. Ion-Exchange Membranes in the Chemical Process Industry. Ind. Eng. Chem. Res. 2013, 52 (31), 10364–10379. 58 (40) Geise, G. M.; Falcon, L. P.; Freeman, B. D.; Paul, D. R. Sodium Chloride Sorption in Sulfonated Polymers for Membrane Applications. J. Memb. Sci. 2012, 423-424, 195–208. (41) Zhang, H.; Ohashi, H.; Tamaki, T.; Yamaguchi, T. Water Movement in a Solid-State Alkaline Fuel Cell Affected by the Anion-Exchange Pore-Filling Membrane Properties. J. Phys. Chem. C 2013, 117 (33), 16791–16801. (42) Oshiba, Y.; Hiura, J.; Suzuki, Y.; Yamaguchi, T. Improvement in the Solid-State Alkaline Fuel Cell Performance through Efficient Water Management Strategies. J. Power Sources 2017, 345, 221–226. (43) Omasta, T. J.; Park, A. M.; LaManna, J. M.; Zhang, Y.; Peng, X.; Wang, L.; Jacobson, D. L.; Varcoe, J. R.; Hussey, D. S.; Pivovar, B. S.; Mustain, W. E. Beyond Catalysis and Membranes: Visualizing and Solving the Challenge of Electrode Water Accumulation and Flooding in AEMFCs. Energy Environ. Sci. 2018, 11 (3), 551–558. (44) Li, Y. S.; Zhao, T. S.; Chen, R. Cathode Flooding Behaviour in Alkaline Direct Ethanol Fuel Cells. J. Power Sources 2011, 196 (1), 133–139. (45) Weng, L.-C.; Bell, A. T.; Weber, A. Z. Modeling Gas-Diffusion Electrodes for CO2 Reduction. Phys. Chem. Chem. Phys. 2018, 20 (25), 16973–16984. (46) Zawodzinski, T. A.; Davey, J.; Valerio, J.; Gottesfeld, S. The Water Content Dependence of Electro-Osmotic Drag in Proton-Conducting Polymer Electrolytes. Electrochim. Acta 1995, 40 (3), 297–302. (47) Xu, C.; Zhao, T. S. In Situ Measurements of Water Crossover through the Membrane for Direct Methanol Fuel Cells. J. Power Sources 2007, 168 (1), 143–153. (48) Cho, J. I. S.; Neville, T. P.; Trogadas, P.; Bailey, J.; Shearing, P.; Brett, D. J. L.; Coppens, M.-O. Capillaries for Water Management in Polymer Electrolyte Membrane Fuel Cells. Int. J. Hydrogen Energy 2018, 43 (48), 21949–21958. (49) Li, Y. S.; Zhao, T. S.; Yang, W. W. Measurements of Water Uptake and Transport Properties in Anion-Exchange Membranes. Int. J. Hydrogen Energy 2010, 35 (11), 5656–5665. (50) Hibbs, M. R.; Hickner, M. A.; Alam, T. M.; McIntyre, S. K.; Fujimoto, C. H.; Cornelius, C. J. Transport Properties of Hydroxide and Proton Conducting Membranes. Chem. Mater. 2008, 20 (7), 2566–2573. (51) Marino, M. G.; Melchior, J. P.; Wohlfarth, A.; Kreuer, K. D. Hydroxide, Halide and Water Transport in a Model Anion Exchange Membrane. J. Memb. Sci. 2014, 464, 61–71. (52) Liu, F.; Lu, G.; Wang, C.-Y. Low Crossover of Methanol and Water Through Thin Membranes in Direct Methanol Fuel Cells. J. Electrochem. Soc. 2006, 153 (3), A543–A553. (53) Xie, Z.; Song, C.; Andreaus, B.; Navessin, T.; Shi, Z.; Zhang, J.; Holdcroft, S. Discrepancies in the Measurement of Ionic Conductivity of PEMs Using Two- and Four-Probe AC Impedance Spectroscopy. J. Electrochem. Soc. 2006, 153 (10), E173–E178. (54) Soboleva, T.; Xie, Z.; Shi, Z.; Tsang, E.; Navessin, T.; Holdcroft, S. Investigation of the through-Plane Impedance Technique for Evaluation of Anisotropy of Proton Conducting Polymer Membranes. J. Electroanal. Chem. 2008, 622 (2), 145–152. (55) Jiang, R.; Mittelsteadt, C. K.; Gittleman, C. S. Through-Plane Proton Transport Resistance of Membrane and Ohmic Resistance Distribution in Fuel Cells. J. Electrochem. Soc. 2009, 156 (12), B1440–B1446. (56) Vielstich, W.; Lamm, A.; Gasteiger, H. A. Handbook of Fuel Cells: Fundamentals Technology and Applications; Wiley New York, 2003; Vol. 2. (57) Gardner, C. L.; Anantaraman, A. V. Measurement of Membrane Conductivities Using an Open-Ended Coaxial Probe. J. Electroanal. Chem. 1995, 395 (1), 67–73. (58) Cooper, K. Characterizing Through-Plane and In-Plane Ionic Conductivity of Polymer Electrolyte Membranes. ECS Trans. 2011, 41 (1), 1371–1380. (59) Karas, F.; Hnát, J.; Paidar, M.; Schauer, J.; Bouzek, K. Determination of the Ion-Exchange Capacity of Anion-Selective Membranes. Int. J. Hydrogen Energy 2014, 39 (10), 5054–5062. (60) Bouzek, K.; Moravcová, S.; Samec, Z.; Schauer, J. H+ and Na+ Ion Transport Properties of Sulfonated Poly (2, 6-Dimethyl-1, 4-Phenyleneoxide) Membranes. J. Electrochem. Soc. 2003, 150 (6), E329–E336. 59 (61) Pellerite, M.; Kaplun, M.; Hartmann-Thompson, C.; Lewinski, K. A.; Kunz, N.; Gregar, T.; Baetzold, J.; Lutz, D.; Quast, M.; Liu, Z.; Yang, H.; Sajjad, S. D.; Gao, Y.; Masel, R. Imidazolium-Functionalized Polymer Membranes for Fuel Cells and Electrolyzers. ECS Trans. 2017, 80 (8), 945–956. (62) 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. Anion-Exchange Membranes in Electrochemical Energy Systems. Energy Environ. Sci. 2014, 7 (10), 3135–3191. (63) Pan, J.; Lu, S.; Li, Y.; Huang, A.; Zhuang, L.; Lu, J. High-Performance Alkaline Polymer Electrolyte for Fuel Cell Applications. Adv. Funct. Mater. 2010, 20 (2), 312–319. (64) Thieu, L. M.; Zhu, L.; Korovich, A. G.; Hickner, M. A.; Madsen, L. A. Multiscale Tortuous Diffusion in Anion and Cation Exchange Membranes. Macromolecules 2019, 52 (1), 24–35. (65) Nishimura, Y.; Yoshida, D.; Mizuhata, M.; Asaka, K.; Oguro, K.; Takenaka, H. Solid Polymer Electrolyte CO2 Reduction. Energy Convers. Manage. 1995, 36 (6), 629–632. (66) Narayanan, S. R.; Haines, B.; Soler, J.; Valdez, T. I. Electrochemical Conversion of Carbon Dioxide to Formate in Alkaline Polymer Electrolyte Membrane Cells. J. Electrochem. Soc. 2011, 158 (2), A167–A173. (67) Larrazábal, G. O.; Strøm-Hansen, P.; Heli, J. P.; Zeiter, K.; Therkildsen, K. T.; Chorkendorff, I.; Seger, B. Analysis of Mass Flows and Membrane Cross-over in CO2 Reduction at High Current Densities in an MEA-Type Electrolyzer. ACS Appl. Mater. Interfaces 2019, 11 (44), 41281–41288. (68) Li, Y. C.; Yan, Z.; Hitt, J.; Wycisk, R.; Pintauro, P. N.; Mallouk, T. E. Bipolar Membranes Inhibit Product Crossover in CO2 Electrolysis Cells. Adv. Sustainable Syst. 2018, 2 (4), 1700187. (69) Leonard, M. E.; Clarke, L. E.; Forner‐Cuenca, A.; Brown, S. M.; Brushett, F. R. Investigating Electrode Flooding in a Flowing Electrolyte, Gas‐Fed Carbon Dioxide Electrolyzer. ChemSusChem 2019, 89. (70) Han, J.; Zhu, L.; Pan, J.; Zimudzi, T. J.; Wang, Y.; Peng, Y.; Hickner, M. A.; Zhuang, L. Elastic Long-Chain Multication Cross-Linked Anion Exchange Membranes. Macromolecules 2017, 50 (8), 3323–3332. (71) AMI-7001 Anion Exchange Membranes | Technical Specifications - Membranes International Inc https://ionexchangemembranes.com/ami-7001-anion-exchange-membranes-technical-specifications/ (accessed Dec 9, 2019). (72) Weng, L.-C.; Bell, A. T.; Weber, A. Z. Towards Membrane-Electrode Assembly Systems for CO 2 Reduction: A Modeling Study. Energy Environ. Sci. 2019, 12 (6), 1950–1968. (73) Lu, X.; Zhu, C.; Wu, Z.; Xuan, J.; Francisco, J. S.; Wang, H. In-Situ Observation of the pH Gradient near the Gas Diffusion Electrode of CO2 Reduction in Alkaline Electrolyte. J. Am. Chem. Soc. 2020. https://doi.org/10.1021/jacs.0c06779. (74) Salvatore, D.; Berlinguette, C. P. Voltage Matters When Reducing CO2 in an Electrochemical Flow Cell. ACS Energy Letters. 2019, pp 215–220. https://doi.org/10.1021/acsenergylett.9b02356. (75) Liu, Z.; Masel, R. I.; Chen, Q.; Kutz, R.; Yang, H.; Lewinski, K.; Kaplun, M.; Luopa, S.; Lutz, D. R. Electrochemical Generation of Syngas from Water and Carbon Dioxide at Industrially Important Rates. Journal of CO2 Utilization 2016, 15, 50–56. (76) Liu, Z.; Yang, H.; Kutz, R.; Masel, R. I. CO2 Electrolysis to CO and O2 at High Selectivity, Stability and Efficiency Using Sustainion Membranes. J. Electrochem. Soc. 2018, 165 (15), J3371–J3377. (77) Gabardo, C. M.; O’Brien, C. P.; Edwards, J. P.; McCallum, C.; Xu, Y.; Dinh, C.-T.; Li, J.; Sargent, E. H.; Sinton, D. Continuous Carbon Dioxide Electroreduction to Concentrated Multi-Carbon Products Using a Membrane Electrode Assembly. Joule 2019. https://doi.org/10.1016/j.joule.2019.07.021. (78) Jiang, K.; Siahrostami, S.; Zheng, T.; Hu, Y.; Hwang, S.; Stavitski, E.; Peng, Y.; Dynes, J.; Gangisetty, M.; Su, D.; Attenkofer, K.; Wang, H. Isolated Ni Single Atoms in Graphene Nanosheets for High-Performance CO2 Reduction. Energy Environ. Sci. 2018, 11 (4), 893–903. (79) Zheng, T.; Jiang, K.; Ta, N.; Hu, Y.; Zeng, J.; Liu, J.; Wang, H. Large-Scale and Highly Selective 60 CO2 Electrocatalytic Reduction on Nickel Single-Atom Catalyst. Joule 2018. https://doi.org/10.1016/j.joule.2018.10.015. (80) Peng, H.; Li, Q.; Hu, M.; Xiao, L.; Lu, J.; Zhuang, L. Alkaline Polymer Electrolyte Fuel Cells Stably Working at 80° C. J. Power Sources 2018, 390, 165–167. (81) Wang, M.; Preston, N.; Xu, N.; Wei, Y.; Liu, Y.; Qiao, J. Promoter Effects of Functional Groups of Hydroxide-Conductive Membranes on Advanced CO2 Electroreduction to Formate. ACS Appl. Mater. Interfaces 2019, 11 (7), 6881–6889. (82) Aeshala, L. M.; Verma, A. Amines as Reaction Environment Regulator for CO2 Electrochemical Reduction to CH4. In Macromolecular Symposia; Wiley Online Library, 2015; Vol. 357, pp 79–85. (83) Burdyny, T.; Smith, W. A. CO2 Reduction on Gas-Diffusion Electrodes and Why Catalytic Performance Must Be Assessed at Commercially-Relevant Conditions. Energy Environ. Sci. 2019. https://doi.org/10.1039/C8EE03134G. (84) Dettelbach, K. E.; Kolbeck, M.; Huang, A.; He, J.; Berlinguette, C. P. Rapid Quantification of Film Thickness and Metal Loading for Electrocatalytic Metal Oxide Films. Chem. Mater. 2017, 29 (17), 7272–7277. (85) Chen, D.; Hickner, M. A. Degradation of Imidazolium- and Quaternary Ammonium-Functionalized Poly(fluorenyl Ether Ketone Sulfone) Anion Exchange Membranes. ACS Appl. Mater. Interfaces 2012, 4 (11), 5775–5781. (86) Wang, B.; Sun, W.; Bu, F.; Li, X.; Na, H.; Zhao, C. Comparison of Alkaline Stability of Benzyltrimethylammonium, Benzylmethylimidazolium and Benzyldimethylimidazolium Functionalized Poly(arylene Ether Ketone) Anion Exchange Membranes. Int. J. Hydrogen Energy 2016, 41 (4), 3102–3112. (87) Deavin, O. I.; Murphy, S.; Ong, A. L.; Poynton, S. D.; Zeng, R.; Herman, H.; Varcoe, J. R. Anion-Exchange Membranes for Alkaline Polymer Electrolyte Fuel Cells: Comparison of Pendent Benzyltrimethylammonium- and Benzylmethylimidazolium-Head-Groups. Energy Environ. Sci. 2012, 5 (9), 8584–8597. (88) Qiu, B.; Lin, B.; Qiu, L.; Yan, F. Alkaline Imidazolium- and Quaternary Ammonium-Functionalized Anion Exchange Membranes for Alkaline Fuel Cell Applications. J. Mater. Chem. 2011, 22 (3), 1040–1045. (89) Salvatore, D. A.; Dettelbach, K. E.; Hudkins, J. R.; Berlinguette, C. P. Near-Infrared–driven Decomposition of Metal Precursors Yields Amorphous Electrocatalytic Films. Science Advances 2015, 1 (2), e1400215. (90) Daiyan, R.; Chen, R.; Kumar, P.; Bedford, N. M.; Qu, J.; Cairney, J. M.; Lu, X.; Amal, R. Tunable Syngas Production through CO2 Electroreduction on Cobalt–Carbon Composite Electrocatalyst. ACS Appl. Mater. Interfaces 2020, 12 (8), 9307–9315. (91) Xia, C.; Zhu, P.; Jiang, Q.; Pan, Y.; Liang, W.; Stavitsk, E.; Alshareef, H. N.; Wang, H. Continuous Production of Pure Liquid Fuel Solutions via Electrocatalytic CO2 Reduction Using Solid-Electrolyte Devices. Nature Energy 2019, 4 (9), 776–785. (92) Rosen, B. A.; Salehi-Khojin, A.; Thorson, M. R.; Zhu, W.; Whipple, D. T.; Kenis, P. J. A.; Masel, R. I. Ionic Liquid-Mediated Selective Conversion of CO₂ to CO at Low Overpotentials. Science 2011, 334 (6056), 643–644. (93) Wang, M.; Torbensen, K.; Salvatore, D.; Ren, S.; Joulié, D.; Dumoulin, F.; Mendoza, D.; Lassalle-Kaiser, B.; Işci, U.; Berlinguette, C. P.; Robert, M. CO2 Electrochemical Catalytic Reduction with a Highly Active Cobalt Phthalocyanine. Nat. Commun. 2019, 10 (1), 1–8. (94) Spurgeon, J. M.; Kumar, B. A Comparative Technoeconomic Analysis of Pathways for Commercial Electrochemical CO2 Reduction to Liquid Products. Energy Environ. Sci. 2018, 11 (6), 1536–1551. (95) Verma, S.; Kim, B.; Jhong, H.-R. “molly”; Ma, S.; Kenis, P. J. A. A Gross-Margin Model for Defining Technoeconomic Benchmarks in the Electroreduction of CO2. ChemSusChem 2016, 9 (15), 1972–1979. (96) Verma, S.; Hamasaki, Y.; Kim, C.; Huang, W.; Lu, S.; Jhong, H.-R. M.; Gewirth, A. A.; Fujigaya, T.; Nakashima, N.; Kenis, P. J. A. Insights into the Low Overpotential Electroreduction of CO2 to 61 CO on a Supported Gold Catalyst in an Alkaline Flow Electrolyzer. ACS Energy Lett. 2018, 3 (1), 193–198. (97) Kim, B.; Hillman, F.; Ariyoshi, M.; Fujikawa, S.; Kenis, P. J. A. Effects of Composition of the Micro Porous Layer and the Substrate on Performance in the Electrochemical Reduction of CO2 to CO. J. Power Sources 2016, 312, 192–198. (98) Weng, L.-C.; Bell, A. T.; Weber, A. Z. Towards Membrane-Electrode Assembly Systems for CO2 Reduction: A Modeling Study. Energy Environ. Sci. 2019, 12, 1950–1968. (99) Yang, H.; Kaczur, J. J.; Sajjad, S. D.; Masel, R. I. Electrochemical Conversion of CO2 to Formic Acid Utilizing SustainionTM Membranes. Journal of CO2 Utilization 2017, 20 (Supplement C), 208–217. (100) Pasaogullari, U.; Wang, C. Y. Liquid Water Transport in Gas Diffusion Layer of Polymer Electrolyte Fuel Cells. J. Electrochem. Soc. 2004, 151 (3), A399–A406. (101) An, L.; Zhao, T. S. Transport Phenomena in Alkaline Direct Ethanol Fuel Cells for Sustainable Energy Production. J. Power Sources 2017, 341, 199–211. (102) Mohanty, A. D.; Bae, C. Mechanistic Analysis of Ammonium Cation Stability for Alkaline Exchange Membrane Fuel Cells. J. Mater. Chem. A Mater. Energy Sustain. 2014, 2 (41), 17314–17320. (103) Wang, J.; Gu, S.; Kaspar, R. B.; Zhang, B.; Yan, Y. Stabilizing the Imidazolium Cation in Hydroxide-Exchange Membranes for Fuel Cells. ChemSusChem 2013, 6 (11), 2079–2082. (104) Lin, B.; Dong, H.; Li, Y.; Si, Z.; Gu, F.; Yan, F. Alkaline Stable C2-Substituted Imidazolium-Based Anion-Exchange Membranes. Chem. Mater. 2013, 25 (9), 1858–1867. (105) Wu, Q. X.; Zhao, T. S.; Chen, R.; Yang, W. W. Effects of Anode Microporous Layers Made of Carbon Powder and Nanotubes on Water Transport in Direct Methanol Fuel Cells. J. Power Sources 2009, 191 (2), 304–311. (106) Park, S.; Lee, J.-W.; Popov, B. N. Effect of PTFE Content in Microporous Layer on Water Management in PEM Fuel Cells. J. Power Sources 2008, 177 (2), 457–463. (107) Avcioglu, G. S.; Ficicilar, B.; Eroglu, I. Effect of PTFE Nanoparticles in Catalyst Layer with High Pt Loading on PEM Fuel Cell Performance. Int. J. Hydrogen Energy 2016, 41 (23), 10010–10020. (108) Uchida, M.; Aoyama, Y.; Eda, N.; Ohta, A. Investigation of the Microstructure in the Catalyst Layer and Effects of Both Perfluorosulfonate Ionomer and PTFE-Loaded Carbon on the Catalyst Layer of Polymer Electrolyte Fuel Cells. J. Electrochem. Soc. 1995, 142 (12). (109) Yin, C.; Gao, Y.; Li, T.; Xie, G.; Li, K.; Tang, H. Study of Internal Multi-Parameter Distributions of Proton Exchange Membrane Fuel Cell with Segmented Cell Device and Coupled Three-Dimensional Model. Renewable Energy 2020, 147, 650–662. (110) Weng, L.-C.; Bell, A. T.; Weber, A. Z. Towards Membrane-Electrode Assembly Systems for CO2 Reduction: A Modeling Study. Energy & Environmental Science 2019, 12 (6), 1950–1968. (111) Chaudhary, S.; Sachan, V. K.; Bhattacharya, P. K. Two Dimensional Modelling of Water Uptake in Proton Exchange Membrane Fuel Cell. Int. J. Hydrogen Energy 2014, 39 (31), 17802–17818. (112) Santamaria, A. D.; Das, P. K.; MacDonald, J. C.; Weber, A. Z. Liquid-Water Interactions with Gas-Diffusion-Layer Surfaces. J. Electrochem. Soc. 2014, 161 (12), F1184–F1193. 62 Appendices Appendix A: Chapter 3 supplementary material Table A3.1. X-ray fluorescence (XRF) measurements and corresponding Ag cathode catalyst loadings measured by gravimetric analysis. XRF count Ag catalyst loading Cathode sample # cps/μA mg/cm2 1 62.84 1.1 2 64.31 1.0 3 63.70 1.0 4 64.91 1.0 Figure A3.1. Salt precipitates formed on cathode flow field plates of CO2 electrolyzers containing (a) PSMIM and (b) PSTMA after applied current density of 100 mA/cm2. 63 Appendix B: Chapter 4 supplementary materials Figure B4.1 Water uptake (WU) and ion exchange capacity (IEC) values for membranes made from 1:1 and 3:1 (m:n) ionomers, defined as high water uptake membranes (HWUM) and low water uptake membranes (LWUM), respectively. The m:n corresponds to the ratio of styrene (m) and the imidazolium-containing 4-vinylbenzyl (n). 64 Figure B4.2 X-ray fluorescence (XRF) measurements of Ag cathodes with PTFE contents of 6, 17, and 29 wt% PTFE taken before electrolysis. Figure B4.3 Sample chronopotentiometry data at applied currents of 0.1, 0.2, 0.4, and 0.8 A, (i.e. 25, 50, 100, and 200 mA/cm2) held for 25 min each. Cell potentials (Ecell) in this study were taken as the average over a 200 s window within initiation of gas chromatography (GC) analysis, indicated by the light grey region. The data shown correspond to a reactor containing Ag cathodes with 17 wt% and 20 μm LWUM. 65 Figure B4.4 Vertical cell configuration used to measure JH2O,outlet and JH2O,inlet for the comparison of HWUM and LWUM at applied current densities (j) of 25, 50, 100 and 200 mA/cm2. Each electrolysis experiment held j for 25 min and used the same cathode (1 mg/cm2 of Ag with 17 wt% of PTFE on carbon paper). Figure B4.5 Horizontal cell configuration used to measure JH2O,outlet and JH2O,inlet for the comparison of different LWUM thicknesses (20, 40, and 55 μm) and cathode hydrophobicities (6, 17, 29 wt% PTFE). Each electrolysis experiment applied a current density of 200 mA/cm2 for 25 min. The horizontal configuration enabled detectable amounts of liquid to be collected at the cathode in a shorter period of time.66 Determination of the JH2O,reduction The Faradaic efficiencies reported in this study were used to determine JH2O, reduction (in mg/cm2⋅h), which is a function of the molar ratio between the consumed water and the number of electrons transferred (n in molH2O/mol e–), the total current density (I in A/cm2), Faradaic efficiency for product k (FEk in %), molar mass of water (MH2O in g/mol) and Faraday’s constant (96485 C/mol). 𝐽𝐻2𝑂,𝑟𝑒𝑑𝑢𝑐𝑡𝑖𝑜𝑛 = ∑3600⋅1000⋅𝑛⋅𝐼⋅𝐹𝐸𝐾⋅𝑀𝐻2𝑂𝐹𝑘 (Eq. A4.1) Table B4.1 Reaction variables and water fluxes for HWUM and LWUM collected after electrolysis experiments at applied current densities (j) of 25, 50, 100, and 200 mA/cm in a CO2 electrolyzer in the vertical configuration. Each electrolysis experiment held j for 25 min. Reaction variables Flux of water Membrane Current density Membrane thickness PTFE cathode content JH2O outlet JH2O,inlet JH2O, reduction mA/cm2 μm % mg/cm2⋅h mg/cm2⋅h mg/cm2⋅h HWUM 25 55 17% 29.2 28.3 8.0 HWUM 50 55 17% 30.7 28.3 16.2 HWUM 100 55 17% 31.9 28.3 31.5 HWUM 200 55 17% 33.4 28.3 56.9 LWUM 25 55 17% 22.7 28.3 8.0 LWUM 50 55 17% 20.9 28.3 16.3 LWUM 100 55 17% 23.8 28.3 31.1 LWUM 200 55 17% 40.0 28.3 56.2 67 Table B4.2 Reaction variables and water fluxes for LWUM at various thicknesses and PTFE cathode contents collected at an applied current density (j) of 200 mA/cm2 in a CO2 electrolyzer in the horizontal configuration. Each electrolysis experiment held j for 25 min. Reaction variables Flux of water Membrane Current density Membrane thickness PTFE cathode content JH2O outlet JH2O,inlet JH2O, reduction mA/cm2 μm % mg/cm2⋅h mg/cm2⋅h mg/cm2⋅h LWUM 200 20 17% 41.0 39.5 57.5 LWUM 200 40 17% 50.8 39.5 49.4 LWUM 200 55 17% 77.3 39.5 56.2 LWUM 200 40 6% 66.7 39.5 55.0 LWUM 200 40 29% 41.6 39.5 55.6 Impedance spectroscopy We performed in situ electrochemical impedance spectroscopy (EIS) measurements to rule out differences in ionic conductivity between HWUM and LWUM as the reason for CO2RR performance differences between these membranes (Figure B4.6). At 200 mA/cm2, the measured 0.5 Ω⋅cm2 ohmic resistance difference should account for only a ~100 mV gain in the Ecell (of the 350 mV observed) for LWUM. We thereby point to cathode flooding as the primary reason for differences in CO2RR performance rather than this modest change in membrane conductivity. 68 Figure B4.6 In situ electrochemical impedance spectroscopy measurements at an applied cell voltage of -1.75 V over a frequency range of 0.03 Hz to 10 kHz, showing the reduced semicircle diameter for HWUM compared to LWUM. Inset shows a magnified AC impedance spectra at the high frequency range (HFR), showing the lower ionic transport resistance at the HFR for HWUM compared to LWUM. Gas diffusion electrode characterization We confirmed that the differences in CO2RR performance between cathodes were not related to differences in surface area or morphology by collecting double-layer capacitance (Cdl) measurements (Figures B4.7 and B4.8), scanning electron microscopy (SEM) micrographs (Figures B4.9 and B4.10), and energy-dispersive X-ray (EDX) maps of each cathode. The Cdl values (directly proportional to electrochemically active surface area) for all three GDE designs were effectively identical at ~2 μF/cm2 (Figure B4.8). SEM images of the cathodes revealed no meaningful differences in morphology, but samples with higher PTFE content showed a higher distribution of PTFE than Ag at the surface (Figure B4.10). This trend is consistent with energy-dispersive X-ray (EDX) elemental maps of the GDEs that confirm an increase in fluorine and a 69 decrease in Ag present at the surface for more hydrophobic cathodes (Figure B4.10). Focused ion beam-SEM (FIB-SEM) micrographs of the GDEs also showed that a higher PTFE content at the cathode led to thicker catalyst layers (Figure B4.9). Higher PTFE content at the cathode surface is known to decrease porosity and increase hydrophobicity of the GDE microporous layer, thereby increasing the resistance for water transport from the anode to the cathode.106,107 An increased catalyst layer thickness is purported to elongate hydrophobic channels for improved transport of reactant gases to the catalytic sites.108 These results therefore support our hypothesis that increasing the PTFE content in the cathode catalyst layer leads to an increase in CO2RR performance because flooding is mitigated and reactant CO2 transport is improved. Figure B4.7 Electrochemical surface area measurements for a 29 wt% PTFE cathode yielded (a) cyclic voltammograms at different scan rates in a non-Faradaic potential window collected in 1M H2SO4 [dashed line indicates the potential (0.25116 V vs Ag/AgCl) at which the capacitive current was measured for all scan rates] and (a) a linear plot of capacitive current versus scan rate. 70 Figure B4.8 Double layer capacitance (Cdl) of cathodes with PTFE loadings of 6 wt%, 17 wt% and 29 wt%, showing no significant change in the Cdl of cathodes used in this study. 71 Figure B4.9 Scanning electron microscopy (SEM) images of the surface (left) and cross-section (right) for cathodic GDEs comprising a microporous layer (MPL) coated with Ag nanoparticles (AgNPs) and a) 6 wt% b) 17 wt% or c) 29 wt% PTFE. 72 Figure B4.10 SEM images (left) and energy-dispersive X-ray (EDX) elemental maps (right) of the cathode surface for GDEs coated with AgNPs and a) 6 wt% b) 17 wt% or c) 29 wt% PTFE. 73 Analysis of fluxes using 1D model We constructed a 1D model in COMSOL Multiphysics 5.3a to determine the effect of changing AEM thickness and PTFE loading on the diffusive (JDIFF), electro-osmotic (JEO), and back convective (JBC) fluxes. The physical scope of the 1D model extends from the anode/AEM interface to the cathode gas gas channel (Figure B4.11). A boundary condition of 100% water (nH2O,AN = 1) was assumed for the anode channel AEM interface. The boundary conditions at the cathode gas diffusion layer (GDL)/ gas channel interface were set such that the mass fraction of water in the humidified CO2 stream equals 85% relative humidity for the flow conditions measured experimentally by a digital flow meter on the cell outlet (nH2O,CAT = 0.22, T = 25 °C, and Pch = 107.9). Figure B4.11 Summary of model domains and dimensions. The system was assumed to be isothermal. Diffusion was estimated using Fick’s law with effective diffusivity through porous media calculated using Bruggeman’s correction. Ion migration within an electric field aside from the water flux arising from electro-osmotic drag were not considered. Water in the AEM is in the liquid phase, while water, CO2 and CO in the catalyst layer 74 and GDL are assumed to be gas phase. Darcy’s law was used to describe transport through porous media. Transport parameters of the membrane, catalyst layer, and gas diffusion layer were sourced from literature (Table S4). The CO2 reduction reaction was assumed to occur solely in the porous catalyst layer with 100% FECO used to calculate the stoichiometric rates of electrochemical CO2 and H2O consumption for a set current density. CO2 + H2O + 2e- → CO + OH– (Eq. A4.2) A case study was completed that changed the membrane thickness from 20 µm, 40 µm, and 55 µm with the corresponding changes in JDIFF, JBC, and JEO compared in Figure B4.12. Figure B4.12 The effect of membrane thickness on (a) each of the water flux (JH2O) to the cathode (diffusion, back convection, and electro-osmotic drag) and (b) the mol fraction of water (nH2O) from the anode/AEM interface to the cathode GDE at j = 200 mA/cm2. Increasing PTFE loading from 6 wt% to 17 wt% and 29 wt% was found to increase the thickness of the catalyst layer from 2.7 µm, to 3.4 µm, and 4.5 µm, and contact angle of 140°, 75 145°, and 145°, respectively (Figure B4.13). Changing these parameters in the model was found to have little effect on water flux. As the model contradicts the experimental results in this case, it highlights the shortcomings that most existing models cannot yet accurately describe the processes that occur inside a CO2 electrolyzer and therefore the need for empirical studies into complex phenomena such as flooding. Figure B4.13 The effect of cathodic PTFE content on (a) each of the water flux (JH2O) to the cathode (diffusion, back convection, and electro-osmotic drag) and (b) the mol fraction of water (nH2O) from the anode/AEM interface to the cathode GDE at j = 200 mA/cm2. Fluxes for JEOD and JBC at the AEM/cathode catalyst layer interface were calculated using the equations below. To calculate the back convection flux (Eq. B4.3) a liquid water saturation value (s = 0.01) was assumed in order to calculate the Leverett function (Eq. B4.4). (Eq. B4.3) (Eq. B4.4) 76 (Eq. B4.5) The diffusive flux for species i in the catalyst layer or GDL is calculated using Eq. B4.6. Density of gaseous mixtures is calculated using the ideal gas law for the calculated mass fractions. (Eq. B4.6) (Eq. B4.7) (Eq. B4.8) (Eq. B4.9) (Eq. B4.10) Diffusive mass flux of species i in the membrane is calculated by Eq. A4.11 with a density equivalent to liquid water: (Eq. B4.11) 77 Table B4.3 Model parameters and variable definitions Parameter/ Variable Symbol Value [units] Reference Effective diffusivity of water in membrane Deff 1.7E-10 [m2/s] 49 Diffusion coeff of water in gas DH2O 2.56E-5 [m2/s] 109 Porosity, CL εCL 0.5 110 Porosity, GDL εGDL 0.5 Sig 38BC datasheet Faraday Constant F 96485 [C/mol] Effective transport factor ƒe Current density j 0, 100, 200 [mA/ cm^2] Flux J Leverett function J(s) Eq 2 [-] 109 Permeability, CL KCL 8E-16 [m2] 110 Permeability, Membrane KMEM 1.8E-18 [m2] 111 Permeability, GDL KGDL 12E-12 [m2] 110 Electro-osmotic drag coefficient κ 2 49 Thickness, membrane LMEM 20, 40, 55 [μm] This work Thickness, catalyst layer LCL 2.7, 3.4, 4.5 [μm] This work Molar mass of species i Mi 18.02 [g/mol] Dynamic viscosity, CO2 μCO2 14.94E-6 [Pa∙s] Dynamic viscosity, water μH2O 8.9E-4 [Pa∙s] Mass fraction of water (AEM, anode side) nAN 1.0 [-] This work Mass fraction of water (channel, cathode side) nCAT 0.22 [-] This work Pressure, channel PCH 107.9 [kPa] This work Density of water ⍴H2O 1 [kg/L] Liquid water saturation in porous media s 0.01 [-] Assumption Surface tension σ 0.0625 [N/m] 109 Temperature T 25 [°C] This work Tortuosity 𝜏F Contact angle 𝛳 140, 145, 145 [°] 112 Mol fraction x