Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Investigation of capacity fade in flat-plate rechargeable alkaline MnO₂/Zn cells Mehta, Sean 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2016_february_mehta_sean.pdf [ 7.6MB ]
Metadata
JSON: 24-1.0223560.json
JSON-LD: 24-1.0223560-ld.json
RDF/XML (Pretty): 24-1.0223560-rdf.xml
RDF/JSON: 24-1.0223560-rdf.json
Turtle: 24-1.0223560-turtle.txt
N-Triples: 24-1.0223560-rdf-ntriples.txt
Original Record: 24-1.0223560-source.json
Full Text
24-1.0223560-fulltext.txt
Citation
24-1.0223560.ris

Full Text

     INVESTIGATION OF CAPACITY FADE IN FLAT-PLATE RECHARGEABLE ALKALINE MnO2/Zn CELLS  by  Sean Mehta   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF   MASTER OF SCIENCE   in  The Faculty of Graduate and Postdoctoral Studies  (Chemical and Biological Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   January 2016  ©Sean Mehta, 2016  ii   Abstract  The rechargeable alkaline manganese dioxide-zinc (RAMTM) battery system has been difficult to commercially develop in the past due to irreversible phase formation and progressive and cumulative capacity fade.  This system has many advantages however, such as low cost and environmentally sustainable materials, long shelf life, moderate energy density, and safety.  A flat-plate architecture was developed and investigated in half and full-cell apparatuses with the goal of understanding and improving cumulative capacity fade in the electrolytic manganese dioxide (EMD) cathode.   Two types of cathode current collectors (CCs) were developed, a thin film foil CC and an expanded metal mesh CC and used to assess the effect of various additives over 30+ cycles under various operating conditions.  Conductive carbon black (Super C65) and graphite (KS44) additives were shown to improve cell performance at 15 wt. % KS44 graphite providing an electrically conductive network between adjacent EMD particles.   In addition, other chemical additives (BaSO4, Sr(OH)2·8H2O, Ca(OH)2, and Bi2O3) were investigated at 5 wt. % with Bi2O3 providing a reproducible improvement over a control recipe.  Mechanical stability of the cathode electrode and pressure application were significant causes of cell failure.  Slow rates of discharge, and shallow depth of discharge (DOD) charge/discharge protocols reduced capacity fade by limiting electrochemically irreversible phase formation such as Mn2O3, Mn3O4, Zn2MnO4, and Mn(OH)2.  Analytical characterization techniques including Scanning Electron Microscopy/ Energy Dispersive X-Ray Spectroscopy (SEM/EDS), X-Ray Photoelectron Spectroscopy (XPS), Powder X-Ray Diffraction (XRD), and Potentiostatic Electrochemical Impedance Spectroscopy (PEIS) were used to provide supporting evidence indicating that the main causes of capacity fade are linked to the cathode electrode’s mechanical properties, increased cell resistance, and progressive and irreversible phase formation.   iii   Preface  The work presented in this thesis was conducted in the David P. Wilkinson laboratory at UBC chemical and biological engineering (CHBE) and Clean Energy Research Centre (CERC).  Under the supervision of and regular discussion with Dr. David P. Wilkinson and Dr. Arman Bonakdarpour, I performed all laboratory work (besides that which is stated here), data analysis, and writing.   A Co-op student, Faye Cuadra helped to prepare cells which are presented in section 3.1.3 with time cut-off and shallow depth of discharge protocols.  Several Co-op students (Kimia Yeganeh, Faye Cuadra, Beichen Zhang, and William Xi) helped to prepare electrolyte and electrode samples (control and 5 wt. % BaSO4 in Figure 3.30) which are in part presented in section 3.3.2.  Greg Afonso was responsible for the design and manufacturing of the full and half-cell hardware (and multiple iterations thereof) presented in Figures 2.1 - 2.3. A subset of this work investigating cathode additives and rate capability studies has been submitted for publication, titled “Impact of Cathode Additives on the Cycling Performance of Rechargeable Alkaline Manganese Dioxide-Zinc Batteries”.  Mehta, S.;Bonakdarpour, A.; Wilkinson, D. P.  A second manuscript is in progress at the submission date of this thesis, discussing the development of flat-plate rechargeable alkaline manganese dioxide-zinc cells with and without BaSO4 as a chemical additive in the cathode.       iv   Table of Contents  Abstract ........................................................................................................................................................ ii Preface ......................................................................................................................................................... iii Table of Contents ........................................................................................................................................ iv List of Tables ................................................................................................................................................ vi List of Figures ............................................................................................................................................. viii Acknowledgements ....................................................................................................................................xiii 1 Introduction ............................................................................................................................................... 1  1.1 Electrochemical Battery Energy Storage Technologies ............................................................. 1  1.1.1 Electrochemical Battery Energy Storage – Basics, Demand and Growth ................................ 1  1.1.2 Alkaline Cell Raw Materials Economics .................................................................................... 3  1.1.3 Rechargeable Alkaline Electrochemical Energy Storage Technologies .................................... 6  1.1.4 Basic Operation of the MnO2/Zn Alkaline Cell ......................................................................... 9  1.2  The Alkaline Cell ...................................................................................................................... 11  1.2.1 History of the Alkaline Cell ..................................................................................................... 11  1.2.2 Alkaline Cell Electrochemistry ................................................................................................ 13  1.2.3 Cathode Capacity Failure ....................................................................................................... 15  1.3  Manganese Dioxide Properties ............................................................................................... 17  1.3.1 Manganese Dioxide Solid State Chemistry ............................................................................ 17  1.3.2 Synthesis of Manganese Dioxide ........................................................................................... 18  1.4 Recent Advancements in Rechargeable Alkaline Cells ........................................................... 19  1.4.1 Cathode Additives .................................................................................................................. 19  1.4.2 Flat-Plate Architecture ........................................................................................................... 21  1.5 Objectives ................................................................................................................................. 23 2 Experimental Methods ............................................................................................................................ 24  2.1 Electrochemical Measurements .............................................................................................. 24  2.1.1 Electrochemical Measurements: Experimental Full-Cell Setup ............................................. 24  2.1.2 Electrochemical Measurements: Half-Cell Setup .................................................................. 26  2.1.3 Electrochemical Measurements: Cell Cycling Regimes .......................................................... 29  2.2  Electrode Current Collectors ................................................................................................... 31  v    2.3  Analytical Characterization ..................................................................................................... 34  2.3.1  X-Ray Diffraction (XRD) Studies ............................................................................................. 34  2.3.2 Scanning Electron Microscopy (SEM) and Energy Dispersive X-Ray Spectroscopy (EDS) ...... 35  2.3.3 Potentiostatic Electrochemical Impedance Spectroscopy (PEIS)........................................... 37  2.3.4  X-Ray Photoelectron Spectroscopy (XPS) ............................................................................. 38  2.3.5  Galvanostatic Intermittent Titration Technique (GITT) ........................................................ 39 3  Results and Discussion ........................................................................................................................... 40  3.1  Effect of Different Operating Conditions ................................................................................ 40  3.1.1 In-Situ Pressure Application ................................................................................................... 40  3.1.2 Depth of Discharge (DOD) ...................................................................................................... 43  3.1.3 Charging and Discharging Protocols ...................................................................................... 46  3.1.4 Galvanostatic Intermittent Titration Technique (GITT) ......................................................... 53  3.1.5 Differential Capacity Analysis ................................................................................................. 55  3.1.6 Coulombic and Energy Efficiency ........................................................................................... 57  3.1.7 Cyclic Voltammetry ................................................................................................................ 59  3.2 Electrode Thickness Effects ...................................................................................................... 61  3.2.1 Scalability and Mass Production Potential ............................................................................. 61  3.2.2  Effect of Current Collector on Identical Cathode Mix ........................................................... 64  3.2.3 Current Collector Discharge Profile Comparison ................................................................... 66  3.3 Effect of Cathode Additives ..................................................................................................... 69  3.3.1 Conductive Additives (KS44 graphite, Super C65 Carbon Black) ........................................... 69  3.3.2 Effect of Cathode Additives on Cycling Performance ............................................................ 75  3.3.3 Scanning Electron Microscopy (SEM) and Energy Dispersive X-Ray Spectroscopy (EDS) ...... 84  3.3.4 Powder X-Ray Diffraction (XRD) Analysis of the Cathode Electrode ..................................... 90  3.4 Analytical Characterization ...................................................................................................... 99  3.4.1 X-Ray Photoelectron Spectroscopy (XPS) Measurements ..................................................... 99  3.4.2 Potentiostatic Electrochemical Impedance Spectroscopy (PEIS)......................................... 100 4 Future Work and Recommendations.................................................................................................... 104  4.1 Improvements in Cell Fabrication and Assembly .................................................................. 104  4.1.1 Effect of Binder .................................................................................................................... 104  4.1.2 Effect of Gelling Agent ......................................................................................................... 105  4.1.3 Effect of Sonication vs. Ball Milling ...................................................................................... 105  vi    4.1.4 Current Collector Active Material Loading Density ............................................................. 106  4.1.5 Effect of Mixing/Electrode Homogeneity ............................................................................ 107  4.1.6 Automated Electrode Processing ......................................................................................... 107  4.1.7 Electrolyte Preparation, Dispensation, and Composition .................................................... 108  4.1.8 Anode Performance ............................................................................................................. 110 4.2 Improvement in Cell Performance through Materials Synthesis, Cathode Additives, and Characterization ........................................................................................................................... 111  4.2.1 Conductive Additives ........................................................................................................... 111  4.2.2 Cathode Additive Studies ..................................................................................................... 112  4.2.3 Alternative Synthesis Conditions of EMD ............................................................................ 114  4.2.4 Cell Setup to Focus on EMD Materials Testing .................................................................... 115  4.2.5 Analytical Characterization .................................................................................................. 115  4.3 Industry Recommendations ................................................................................................... 116 5 Conclusions ............................................................................................................................................ 119 Bibliography.............................................................................................................................................. 123 Appendices ............................................................................................................................................... 128  Appendix A – Materials Preparation ............................................................................................. 128  Appendix B – Materials Information ............................................................................................. 130  Appendix C - Instrumentation ....................................................................................................... 132     vii   List of Tables  Table 1.1: Common (Non-Exhaustive) Electrochemical Storage Technologies ........................................ 1 Table 1.2: Electrochemical Energy Storage Key Feature Comparison  ..................................................... 8 Table 1.3: Alkaline Cell History ................................................................................................................. 11 Table 1.4: Rechargeable Alkaline Cell Cathode Additives ......................................................................... 21 Table 1.5: Advantages of Rechargeable Alkaline MnO2/Zn Battery Chemistry (left) and Flat-Plate Cell Architectures (right) .................................................................................................................................. 22 Table 2.1: 10 A/m2 H2 Evolution Reaction (HER) Overpotentials for Common Metals ............................ 27 Table 2.2: Comparison of Cathode Electrode Current Collectors ............................................................. 33 Table 3.1: Comparison of estimated voltage losses dependent on rate of discharge ............................. 47  Table 3.2: Comparison of key parameters for thin film foil and expanded metal mesh CCs ................... 62  Table 3.3: Important parameters of Super C65 carbon black and KS44 graphite .................................... 69  Table 3.4: Published cathode compositions and rationale for testing ..................................................... 76 Table 3.5: Cathode additive chemical formula, molar mass, and solubility in water at 20oC .................. 77  Table 4.1: EMD to cathode additive molar ratio at 5 wt. % ...................................................................... 114           viii   List of Figures  Figure 1.1 - Battery Types Market Distribution (2009)1 ............................................................................ 2 Figure 1.2 - Manganese Ore Production by Country (2007-2011 average)2 ............................................ 3 Figure 1.3 - Relative cost per Watt-hour of common electrochemical battery energy storage technologies (2009)3 ................................................................................................................................. 5 Figure 1.4 - Volumetric and Gravimetric Energy Density of Common Electrochemical Battery             Storage Technologies4 ............................................................................................................................... 7 Figure 1.5 - Alkaline Electrochemical Cell: a) MnO2 cathode, b) Zinc anode, c) Separator, d) Potassium Hydroxide (KOH) electrolyte, e) External Load. ........................................................................................ 10 Figure 1.6 - Typical Voltage Capacity Plot of MnO2/Zn Cell in Flat-Plate Configuration with C/10 Rate of discharge to 0.9 V.  Capacity vs. cycle data for this cell is presented in Figure 1.2.3 ........................... 15 Figure 1.7 - Capacity vs. cycle for AA cylindrical alkaline cell (continuous 10 Ω discharge to 0.9 V at room temperature5) vs. flat-plate cell (C/10 rate of discharge to 0.9 V at room temperature). ............. 16 Figure 1.8 - Electrolytic Manganese Dioxide Structure6 ........................................................................... 17 Figure 1.9 - Proposed Electrolytic Manganese Dioxide Structure including Ruetschi Defects7................ 18 Figure 1.10 - Flat-Plate (left)8 and Cylindrical Bobbin (right)9 Rechargeable Alkaline MnO2/Zn Cell Architectures ............................................................................................................................................. 22 Figure 2.1 - Flat-Plate Rechargeable Alkaline Full-Cell Setup ................................................................... 24 Figure 2.2 - Schematic of Half-Cell Setup of EMD-based Cathode ........................................................... 28 Figure 2.3 - 3D rendering of half-cell hardware. Designed and manufactured by Greg Afonso .............. 29 Figure 2.4 - 3 step charge/discharge protocol.  1) Galvanostatic discharge, 2) Galvanostatic Charge, 3) Potentiostatic Charge. .............................................................................................................................. 30 Figure 2.5 - Cathode Electrode Current Collectors (CCs): Nickel Foil (top) and Expanded Nickel Mesh (bottom) .................................................................................................................................................... 32 Figure 2.6 - Electrode fabrication procedures for expanded nickel mesh (left) and thin film nickel foil (right) CC showing flexibility of both types of electrodes ......................................................................... 34 Figure 2.7 - Scanning Electron Microscopy Interaction Volume10 ............................................................ 36 Figure 2.8 - Idealized Nyquist plot noting bulk resistance (Rs) and charge transfer resistance (Rct) with equivalent circuit model ........................................................................................................................... 37   ix   Figure 3.1 – Current and Voltage vs. Time profile for no stack pressure applied (a) and 47 psi (3.3 kg/cm2) applied pressure (b). Cell discharged at a C/10 rate to -0.4 V vs.Hg/HgO in the flood half-cell setup.......................................................................................................................................................  40  Figure 3.2 - Voltage capacity profiles for 47 psi (30kg/cm2) stack pressure (a) and no applied pressure (b) for cycles 1, 3, and 5. Arrows indicate the directions of charge and discharge................................. 41  Figure 3.3 - Capacity vs. cycle number for 47 psi (3.3 kg/cm2) electrode stack pressure and no stack pressure .................................................................................................................................................... 42  Figure 3.4 - Uniform (left) and non-uniform (right) pressure distribution observed by the use of pressure sensitive paper ........................................................................................................................... 43  Figure 3.5 - Delamination of an electrode after cycling in a flooded half-cell setup.  Electrode cut in half with scissors after being removed from flooded half-cell setup ....................................................... 43  Figure 3.6 - Voltage vs. Capacity for 1.1, 0.9, 0.5 V potential cut-off for first and fifth cycles ................. 44  Figure 3.7 - Differential capacity profile for 0.5 V (a), 0.9 V (b), and 1.1 V (c) cut-off potentials for the first and fifth cycles.  Charge/discharge voltage vs. capacity profiles for 0.5 V, 0.9 V, and 1.1 V cut-off potentials for the first (d) and fifth (e) cycle ............................................................................................. 45  Figure 3.8 - Voltage capacity profile of baseline cells at C/1, C/10, and C/20 rates for first cycle with a 1.1 V cut-off. ............................................................................................................................................. 47  Figure 3.9 - Capacity and end of discharge potential for C/10 rate time cut-off of full MnO2/Zn cells at 10 and 25% time cut-off depth of discharge ............................................................................................ 48  Figure 3.10 - Shallow discharge (1.35 V C/2 rate and 1.4 V C/10 rate) with intermittent deep discharge (1.1 V) capacity vs. cycle ........................................................................................................................... 50  Figure 3.11 - Power and voltage versus capacity with 20 mA galvanostatic discharge at C/10 rate of discharge ................................................................................................................................................... 51  Figure 3.12 - C/40 Rate discharge, 12 hour discharge (30% depth of discharge):  a) voltage capacity profile, b) cut-off potential and specific capacity vs. cycle, and c) projected cell lifetime in pressurized half-cell setup ............................................................................................................................................ 52  Figure 3.13 - Voltage vs. fraction of total discharge for C/1, C/5, C/10, C/20, and C/40 rates of discharge to 1.1 V cut-off potential. GITT measurements at C/2 rate of discharge with 20 minutes discharge, 30 minutes rest period ................................................................................................................................... 54   Figure 3.14 - Voltage capacity profiles of cells at C/1, C/10, and C/20 rates showing polarization potential losses (a) and polarization potential losses vs. relative current for C/1, C/2, C/5, C/10, C/20, and C/40 rates of discharge (b) ................................................................................................................. 55  x   Figure 3.15 - Differential capacity plot for control recipe cycles #1, #10, and #20 showing capacity fade.  Figure is annotated to show discharge peak 1 (DP1), discharge peak 2 (DP2), and charge peak 1 (CP1) ....................................................................................................................................................... 56  Figure 3.16 - Differential capacity plots for control (a) and all additives at 5 wt. % (BaSO4 (b), Ca(OH)2 (c), Sr(OH)2·8H2O (d), Bi2O3 (e)) cycles 1, 3, and 5 with absolute value of differential capacity discharge/charge peak height vs. cycle number (inset) and cycle 1 differential capacity for control and all additives (f) ........................................................................................................................................... 57 Figure 3.17 - Coulombic (C.E.) and energy efficiency (E.E.) vs. cycle number for all additives at 5 wt. % with thin film foil CC electrodes ................................................................................................................ 59  Figure 3.18 - 0.05 mv/s CV of EMD-based cathode in half-cell setup (blue) with voltage vs. time (inset) and differential capacity (red) showing unique redox processes. ............................................................ 60 Figure 3.19 - Top view (left) showing electrode surface and side view (right) showing electrode width after processing ........................................................................................................................................ 61  Figure 3.20 - Optical micrographs of nickel foam (left), nickel foil (centre), and expanded nickel mesh (right) at 4x (top) and 10x (bottom) magnification ................................................................................... 64  Figure 3.21 - Normalized (left) and specific capacity (right) for control electrodes cycled at C/10 discharge rate to 1.1 V using the expanded metal mesh and thin film foil CCs ....................................... 65  Figure 3.22 - Voltage vs. capacity profiles for thin film foil CC and expanded metal mesh CC cathodes.  Cathode mix powder identical in each case (80% EMD, 15% KS44 graphite, 5% BaSO4). ........................ 67  Figure 3.23 - Midpoint Voltage vs. Cycle Number for thin film foil CC cells expanded metal mesh CC cells ........................................................................................................................................................... 68  Figure 3.24 - Powder X-Ray diffraction pattern of SuperC65 carbon black and KS44 graphite ............... 70  Figure 3.25 - 3 idealized representations of EMD-graphite connectivity. Large graphite particle size (left), intermediate graphite particle size (centre), and small graphite particle size (right). Central spherical particle represents EMD, with graphite particles surrounding it .............................................. 71  Figure 3.26 - Voltage vs. capacity profiles of 15% KS44 graphite (left), 15% Super C65 carbon black (right), and 7.5% KS44 Graphite/7.5% Super C65 Carbon black (bottom) ............................................... 72  Figure 3.27 - Capacity vs. cycle of electrodes prepared with 15% KS44 graphite, 15% Super C65 carbon black and 7.5% of both KS44 graphite and Super C65 carbon black using electrodes prepared with expanded metal mesh CCs ........................................................................................................................ 73   Figure 3.28 - Capacity vs. Cycle Number for addition of 15, 20, and 25% KS44 graphite at 1.2 and 0.9 V cut-off potentials ...................................................................................................................................... 75  Figure 3.29 - Normalized (left) and Specific Capacity (right) for control cathode composition (84% EMD, 16% KS44 graphite) with expanded metal mesh and foil CCs ......................................................... 78  xi    Figure 3.30 - Normalized (left) and specific (right) capacity vs. cycle number for baseline cathode composition (80% EMD, 15% KS44 graphite, 5% BaSO4) compared to control cathode composition (84% EMD, 16% graphite) ......................................................................................................................... 79  Figure 3.31 - Normalized(left) and specific (right) capacity vs. cycle for strontium hydroxide octahydrate cathode composition (80% EMD, 15% KS44 graphite, 5% Sr(OH)2.8H2O). Thin film foil vs. expanded metal mesh CC for C/2 and C/10 rates of discharge to 1.1V ................................................... 80  Figure 3.32 - Normalized and specific capacity vs. Cycle for Bismuth (III) oxide additive (80% EMD, 15% KS44 graphite, 5% Bi2O3). Thin film foil vs. Mesh CC with C/2 and C/10 rate of discharge to 1.1V.. ....... 81  Figure 3.33 - Normalized (left) and specific capacity vs. cycle number for calcium hydroxide cathode additive (80% EMD, 15% KS44 graphite, 5% Ca(OH)2). Thin film foil and expanded metal mesh CCs with C/2 rate of discharge to 1.1V .................................................................................................................... 82  Figure 3.34 - Thin film electrode full-cell capacities from galvanostatic discharge, galvanostatic charge, potentiostatic charge and total charge capacity vs. cycle number for control (a), 5 wt. % BaSO4 (b), 5 wt. % Ca(OH)2 (c), 5 wt. % Sr(OH)2·8H2O (d), 5 wt. % Bi2O3 (f), and specific discharge capacity vs. cycle number for all additives (f) ....................................................................................................................... 83  Figure 3.35 - Thin film electrode full-cell energies from galvanostatic discharge, galvanostatic charge, potentiostatic charge and total charge energy vs. cycle number for control (a), 5 wt. % BaSO4 (b), 5 wt. % Ca(OH)2 (c), 5 wt. % Sr(OH)2·8H2O (d), 5 wt. % Bi2O3 (f), and specific discharge energy vs. cycle number for all additives (f) ....................................................................................................................... 85  Figure 3.36 - Expanded metal mesh CC electrode surface analysis (5 wt. % BaSO4).  Backscattered electron micrograph (a) with energy dispersive x-ray spectroscopy mapping of potassium (b), carbon (c), barium (d), manganese (e) and oxygen (f).  Scale bar representing 80 μm ....................................... 86  Figure 3.37 - Secondary electron micrograph of an expanded metal mesh CC electrode cross section.  Image taken at the UBC CHTP ................................................................................................................... 87  Figure 3.38 - EMD particle size determination by SE (a, c, e) and BSE (b, d, f) imaging.  EMD particles outlined in solid red line (a-f), and graphite particle outlined in dashed blue line (c-f) ........................... 88  Figure 3.39 - Surface of 5 wt. % Sr(OH)2·8H2O electrode imaged in BSE mode (a) and EDS mapping for elemental Sr (b), Mn (c), and overlaid Sr and Mn map (d) ....................................................................... 89 Figure 3.40 - EDS potassium spot analysis through electrode cross section. ........................................... 90  Figure 3.41 - a) Powder x-ray diffraction patterns of a control cathode  (84% EMD, 16% Graphite) using expanded metal mesh CC cell cycled to a charged and discharged state; b) Magnified and overlaid pattern showing peak shifting .................................................................................................... 91  Figure 3.42 - a) Powder x-ray diffraction patterns of a baseline cathode (5 wt. % BaSO4) cell cycled to a charged and discharged state compared to fresh cathode mix; b) Magnified and overlaid pattern showing peak shifting. .............................................................................................................................. 92  xii    Figure 3.43 - Powder  x-ray diffraction pattern of baseline cathode mix (5% BaSO4) in 9M KOH with N2 atmosphere (top), 9M KOH with air (middle), and no electrolyte (bottom) for 30 days ......................... 93  Figure 3.44 - Powder X-ray diffraction pattern of barium sulfate (BaSO4, barite) in 9M KOH ................. 94 Figure 3.45 - Powder x-ray diffraction analysis of baseline cathode electrodes (5% BaSO4) from bottom to top: fresh cathode, discharged half-cell cathode, charged half-cell cathode, charged full cell cathode. .................................................................................................................................................... 95 Figure 3.46 - Powder x-ray diffraction analysis of cathode electrodes prepared with 5 wt. % Sr(OH)2·8H2O. ............................................................................................................................................ 97 Figure 3.47 - Powder x-ray diffraction analysis of fresh and cycled electrodes containing 5 wt. % Ca(OH)2 indicating zinc contamination. .................................................................................................... 98 Figure 3.48 - XPS measurements for a) fresh and b) cycled electrodes. .................................................. 100 Figure 3.49 - Typical control recipe full-cell Nyquist plot showing determination of charge transfer (Rct) and bulk resistance (Rs).  Measurements performed on BioLogic VMP3 potentiostat with frequency range 10 mHz to 200 kHz and 10 mV signal amplitude. ........................................................................... 101  Figure 3.50 - Bulk resistance (Rs) vs. cycle number for baseline (5 wt. % BaSO4) cathode full cell discharged to 1.1 V at a C/2 rate of discharge.  PEIS measurements were taken after galvanostatic discharge and after potentiostatic charge. ............................................................................................... 102  Figure 3.51 - Charge transfer resistance (Rct) vs. cycle number for baseline (5 wt. % BaSO4) cathode full cell discharged to 1.1 V at a C/2 rate of discharge.  PEIS measurements were taken after galvanostatic discharge and after potentiostatic charge. ............................................................................................... 103  Figure 3.52 – Annotated pourbaix diagram showing various discharge potentials11 ............................... 104 Figure 4.1 - 3D rendering of planetary mixer under construction for cathode electrode paste mixing and production.......................................................................................................................................... 108 Figure 4.2 - Flat-plate battery architecture using expanded metal mesh CCs.......................................... 117 Figure 4.3 - Flat-plate battery architecture volumetric and gravimetric energy density vs. number of cells with and without casing .................................................................................................................... 118 Figure 5.1 - Summary of Conclusions in 4 areas of investigation: 1) Operating Conditions, 2) Cathode Additives, 3) Electrode Processing, and 4) Characterization. ................................................................... 122       xiii   Acknowledgements  I would like to thank my supervisors Dr. David P. Wilkinson and Dr. Arman Bonakdarpour for the opportunity to work on an exciting and challenging project and for introducing me to the field of energy storage.  I am extremely grateful for their expertise, patience, and optimism.  They provided a stimulating environment which I am glad to be a part of.  I also wish to thank my Greg Afonso for cell hardware design, and the Co-op students Kimia Yeganeh, Faye Cuadra, Beichen Zhang, and William Xi for help with electrode and electrolyte preparation. To my parents who taught me that I can pursue any goals I set my sights on, and that the pursuit of knowledge is a lifelong goal.  To my siblings who always urged me to continue working hard and set a wonderful example for me to follow. I would also like to acknowledge NSERC and OTI Inc. for the generous funding without which this work would not have been possible.        1  1 Introduction 1.1 Electrochemical Battery Energy Storage Technologies 1.1.1 Electrochemical Battery Energy Storage – Basics, Demand and Growth Electrochemical battery energy storage has been a wide and growing area of research since the late 18th century with applications spanning everything from portable electronics to electric vehicles.12  The staggering diversity generated in electrochemical storage technologies is driven primarily by demand and future expected growth.  Electrochemical battery energy storage refers to the storage and release of electrical energy from chemical bonds and substrates, encompassing various technologies as summarized in Table 1.1 which have had a major impact worldwide.   Table 1.1 - Common (Non-Exhaustive) Electrochemical Storage Technologies.12 Electrochemical Storage Technology Primary/Secondary Inventor (Date) Common Applications Voltaic Pile Primary Allesandro Volta (1782) N/A Daniell Cell Primary John Daniell (1836) Early Railway Signaling Systems Lead-Acid Cell Secondary Gaston Planté (1859) Automotive Ignition Leclanché Cell Primary Georges Leclanché (1866) Early Telephones, Flashlights Zinc-Carbon Cell Primary Carl Gassner (1886) Early Flashlights Alkaline Cell Primary/Secondary Lewis Frederick Urry (1949) Portable Electronics Nickel Metal Hydride Cell Secondary Stanford R. Ovshinsky (Late 1960s) Electric Vehicles, Backup Telecommunications Lithium Ion Cell Secondary (1980s) High Power Electronics, Electric Vehicles  These technologies have been developed over the past several centuries and have produced a wide range of diversity with respect to chemistry and applications.  Battery research has received more attention recently in accordance with the demand for portable electronic applications such as cellular phones and   2  laptops as well as for energy storage and automotive applications.   The secondary or rechargeable market is expected to continue to grow and is expected in 2015 to increase to 82.6% of the market.1  The primary market is significant also, with alkaline cells currently leading the market.13  Figure 1.1 outlines the market distribution for the year 2009.   Figure 1.1 - Battery Types Market Distribution (2009).1  The prominence of Lithium-Ion cells is correlated to the massive demand and wide distribution of high power portable electronic devices such as smartphones, laptops, and gaming devices.   It is expected that this market will continue to grow and expand in the near future.14  Energy production must be coupled to energy storage in order to be of use for on-demand applications.  The price of grid electricity per kWh varies based on the method of generation, however is typically several orders of magnitude cheaper than battery technologies.  As a point of comparison, the Ontario Energy Board prices for electricity as of April 20th, 2015 vary between $0.08 (off-peak) and $0.161 (on-peak) per   3  kWh while the cost of primary alkaline batteries can range between $80 (D cell) and $406 (AAA cell) per kWh.15  The cost of battery technologies is highly dependent on economies of scale, with larger size batteries offering a significant cost reduction.   Currently, there is a significant drive to improve the known battery chemistries and to develop new types of battery technologies.   Among the various chemistries, alkaline MnO2/Zn chemistry is one which has been successfully implemented as a primary cell energy storage technology, however it is receiving new attention as a potential rechargeable chemistry.   1.1.2 Alkaline Cell Raw Materials Economics World manganese ore production was estimated by the United States Geological Survey (USGS) to be 16 Mt in 2011, increasing from 2010 by 6%.2  The distribution of manganese ore by country is detailed in Figure 1.2.    Figure 1.2 - Manganese Ore Production by Country (2007-2011 average).2    4  Although manganese ore production is dominated by South Africa, Australia, China and Gabon, the ore needs to be refined to produce the active cathode material, electrolytic manganese dioxide (EMD).  This material is often mined as various manganese compounds such as manganese carbonate, manganese sulfate, or even manganese dioxide.2   Manganese is the 12th most abundant element in the earth’s crust with significant and widely distributed deposits in locations such as in China, Australia and Africa.7 The production costs of manganese dioxide are significantly cheaper than electrolytic manganese metal (EMM) because the electro-winning process requires approximately 20% of the electrical energy required for EMM production.16  The major producers of EMD are located in China currently, with 98% of EMM production occurring in China and exporting approximately 30% internationally.16  The bulk import cost of EMD from 2009-2012 has averaged at $1.14 USD/lb.16  As of 2008, alkaline battery grade EMD was being produced in large scale by 8 companies in Greece, USA, Japan, South Africa, and China totalling over 160 000 tonnes per annum.17  While the majority of EMD production is destined for primary alkaline batteries, a growing demand for use in both primary and rechargeable Li-ion batteries is estimated at over 40 000 tonnes per annum and is increasing to meet demand for electric vehicles (EV) and hybrid electric vehicles (HEV).17 The active anode material in the MnO2/Zn cell is zinc and zinc oxide.  This material is mined around the world with China, Peru, Australia, and North America being the major producers according to the USGS.18  Once zinc ore is mined it is sent to a zinc smelter to be purified.  Approximately 30% of zinc is from recycled sources.19  Over the past 6 years, the price of zinc has risen from as low as $0.50 USD/lb to $0.99 USD/lb currently, averaging $0.90 USD/lb over this period.20  Zinc is used for many other purposes such as galvanizing and alloying metals, in addition to alkaline batteries.19  Comparatively, the need for zinc is significantly less than that of EMD for battery applications for two reasons.  First, the available reserves of zinc are distributed more evenly worldwide.20 Second, the stoichiometric ratio of EMD to zinc in alkaline batteries is 2:1, further reducing the need for zinc in comparison to EMD.  A cost analysis of the active   5  materials is a necessary step towards large scale commercialization of MnO2/Zn flat-plate RAM cells.  Figure 1.3 shows the relative cost of common electrochemical battery energy storage technologies.  Figure 1.3 – Relative cost per Watt-hour of common electrochemical battery energy storage technologies (2009).3  Evidently, the alkaline MnO2/Zn system clearly provides a significant cost advantage.  It should be noted that this cost does not take rechargeable capacity into consideration and is only comparing the single discharge energy cost.  This comparison provides a strong motivation to develop the rechargeable alkaline MnO2/Zn system.    6  1.1.3 Rechargeable Alkaline Electrochemical Energy Storage Technologies Rechargeable alkaline manganese dioxide-zinc (RAMTM) cells are an attractive battery technology for several reasons such as its low cost, moderate energy density, and environmentally friendly active materials.  However, they are not suitable for a number of uses which are currently being heavily researched, including high power and repeated discharge applications such as portable computers, electric motors, or multi-use electronics.  Two major factors currently limiting the expansion of RAM cells into commercial markets include cumulative and progressive capacity fade and the active material’s inherent energy density.  Cumulative capacity fade is a significant problem for any type of rechargeable alkaline battery and failure occurring after dozens to hundreds of cycles is often not appropriate for repeat discharge applications.8  The gravimetric and volumetric energy density of the MnO2/Zn electrochemical couple can be compared with various other currently used battery technologies such as lead-acid, lithium ion, metal/air, nickel metal hydride, and nickel cadmium cells as shown in Figure 1.4.   7   Figure 1.4 - Volumetric and Gravimetric Energy Density of Common Electrochemical Battery             Storage Technologies.4  Figure 1.4 compares volumetric and gravimetric energy density of various battery technologies.  Lead-acid batteries, which are commonly used to power automobile electronics, provide a low volumetric and gravimetric energy density compared to other battery chemistries.  Cylindrical AA alkaline provides a significant boost in comparison to lead-acid, nickel metal hydride, and nickel-cadmium cells in both volumetric and gravimetric energy density.  The energy density ranges that are exhibited by these different technologies are representations of variations in the form factors and pack architectures.  One can optimize the energy density of these technologies to a point limited by the properties of the active materials present.  The energy density of AA (cylindrical) alkaline cells is expected to be higher than that   8  of flat-plate (prismatic) alkaline cells due to the added mass and volume of the current collectors on the individual cell level.  However, when combining cells into the pack level, the flat-plate architecture is expected to improve the overall pack energy density as a result of improved stacking compared to a cylindrical format. Another important point of comparison between battery technologies is working cell potential.  Alkaline cells have a theoretical open circuit voltage of 1.608 V; however a practical working voltage is often between 1.1-1.4 V depending on the state of charge.  Cylindrical alkaline cells are typically advertised as having a working voltage of 1.5 V, although this voltage is only available near 100% state of charge (SOC).  Table 1.2 compares several important parameters for common battery technologies. Table 1.2: Electrochemical Energy Storage Key Feature Comparison.4 Battery Chemistry Nominal Cell Voltage (V) Gravimetric Energy Density (Wh/kg) Load Current (C-rate, best performance) Fast Charge Time (h) Operating Temperature Range (oC) Cycle Life (cycles at 80%) Self-Discharge Rate (%/month) NiMH 1.2 60-120 C/2 or lower 2-4 -20 to 60 300-500 30 NiCd 1.2 45-80 C/1 or lower 1 -40 to 60 1500 20 Pb-Acid 2.0 30-50 C/5 or lower 8-16 -20 to 60 500-800 5 Li Ion 3.7 100-160 C/1 or lower 2-4 -20 to 60 500-1000 10 Rechargeable Alkaline 1.5 80 C/5 or lower 2-3 0 to 65 50 0.3  The commercialization of any battery technology will be heavily influenced by the factors described in the above table.  In the case of rechargeable alkaline batteries, commercialization is primarily limited by the cycle life, which is a direct manifestation of cumulative capacity fade.   Other important parameters including energy density, load current, and operating temperature limit the penetration of alkaline cells to specific applications.  Based on these characteristics, alkaline cells are not ideal for low temperature (< 0oC), high power density, and high energy density applications.  However, alkaline cells are particularly well suited for low to moderate power and energy density as well as long shelf life applications.  It is necessary to work optimistically and realistically within the confines of particular battery chemistries and   9  engineering considerations when attempting to design a commercially viable battery energy storage product. 1.1.4 Basic Operation of the MnO2/Zn Alkaline Cell Electrochemical cells are composed of four main components called the anode, cathode, electrolyte, and separator.  In the case of the alkaline MnO2/Zn cell, the anode electrode is primarily zinc (Zn) and the cathode electrode is primarily manganese dioxide (MnO2).  Upon discharge of the electrochemical cell, the anode is oxidized (loses electrons) while the cathode is reduced (gains electrons) and a current is driven through an external load or circuit.  Upon charging, the reverse process occurs ideally restoring the anode and cathode to a charged state.  The potassium hydroxide (KOH or KOH/ZnO) electrolyte provides a medium for ions to flow balancing the charge distribution and it takes part in the electrochemical reaction.  The separator serves to prevent immediate discharge or short-circuiting while maintaining ionic contact between the anode and cathode electrodes.  Figure 1.5 shows a schematic of the MnO2/Zn cell with its associated discharge reactions.   10   Figure 1.5 - Alkaline Electrochemical Cell: a) MnO2 cathode, b) Zinc anode, c) Separator,                                d) Potassium Hydroxide (KOH) electrolyte, e) External Load.  2 MnO2 + 2 H2O + 2 e-  2 MnOOH + 2 OH- (Cathode, Eo = 0.36 V vs. SHE)  [Eq. 1.1] Zn + 2OH-  ZnO + H2O + 2e-  (Anodic Oxidation, Eo = -1.248 V vs. SHE)  [Eq. 1.2] Zn + 2MnO2 + H2O  ZnO + 2MnOOH  (Overall, Eo= 1.608 V)    [Eq. 1.3]     11  1.2 The Alkaline Cell 1.2.1 History of the Alkaline Cell Current alkaline battery technologies have a long history which stems back over 150 years to the invention of the Leclanché wet cell in 1866 by the French scientist Georges Leclanché.12  This cell was constructed of a zinc anode, a carbon cathode, and a manganese dioxide depolarizer in order to suppress hydrogen gas formation.12  A depolarizer accepts electrons to prevent the reduction of water to evolve hydrogen gas.  Table 1.3 shows a timeline of key advancements towards the current state of the art alkaline cell. Table 1.3: Alkaline Cell History.21,5,8 Year Technological Improvement Description 1866 Leclanché Cell Georges Leclanché wet cell with Zn anode, MnO2/C cathode, NH4Cl electrolyte 1886 Zinc-Carbon Cell Carl Gassner adapted Leclanché wet cell into a dry cell 1957 Primary Alkaline Cell US patent No. 2,960,558, Lewis Frederick Urry and Karl Kordesch detailing dry alkaline MnO2/Zn cell. 1970’s Commercial Rechargeable Alkaline Cell Removed from the market due to poor performance. 1980’s Commercial Rechargeable Alkaline Cell Re-Introduction Karl Kordesch. Rechargeable alkaline re-introduced with improved designs to reduce cathode expansion. 2004 Significant Improvement in AA Rechargeable Alkaline Cell Pure Energy Visions Inc. (Nova Scotia, Canada).  Current commercial standard for rechargeable alkaline cells. 2006 Flat-Plate Rechargeable Alkaline Cell Stani et. al, Development of flat plate rechargeable alkaline manganese dioxide–zinc cells at Graz University of Technology, Austria.   Zinc carbon cells are closely related to the Leclanché wet cell in that they are constructed of a zinc anode and manganese dioxide depolarizer surrounding a carbon rod.  Zinc carbon cells use a paste of zinc chloride and ammonium chloride as electrolyte.22  These cells are environmentally hazardous as the zinc anode was commonly amalgamated with heavy metals such as mercury or cadmium. Zinc carbon cells are   12  primary cells, meaning they would be discarded after a single use, and result in leeching of heavy metals into the surrounding environment.22 The invention of the primary alkaline cell was made public in 1957 with Lewis Frederick Urry (29 January 1927 – 19 October 2004, Canadian) and Karl Kordesch (18 March 1922 – 12 January 2011, Austrian) filing for US patent No. 2,960,558 which detailed a dry alkaline cell with a gelled zinc anode.21 The most significant advancement with these batteries was the inclusion of potassium hydroxide electrolyte.  These cells quickly replaced zinc carbon cells in the 1960s as they provided higher energy density, and eliminated the need and cost for heavy metal amalgamation in the anode.21 Primary alkaline cells are fundamentally similar to rechargeable alkaline cells in that the cells are commonly constructed in the same fashion.  Rechargeable alkaline cells, however, limit the depth of discharge in order to improve rechargeability.5  The electrochemistry of the manganese dioxide cathode will be described in further detail in later sections. Rechargeable alkaline cells were introduced commercially in the 1970s, but did not perform well economically and were discontinued.  This technology was pioneered by Dr. Karl Kordesch at the Graz University of Technology in Austria.5  Several new cell designs involving metal cages to constrain the expansion of the positive electrode were introduced in the 1980s.5  The licenses to commercialize batteries based on this technology were sold by a company named Battery Technologies Inc. which was founded in part by Dr. Karl Kordesch in the 1980s.   The licences were obtained by several companies worldwide.5  In 2004, a Nova Scotia based company called Pure Energy Visions introduced a significantly improved version of the AA rechargeable alkaline cell, which has now become the industry standard.5 In 2006, a flat-plate rechargeable alkaline cell was demonstrated by Stani et. al at the University of Graz which used a sandwich cell approach to apply pressure to the electrodes between plexiglass plates.8  This   13  work presented the cycling performance of a small scale flat-plate configuration over 25 discharge cycles with significant capacity fade.8 1.2.2  Alkaline Cell Electrochemistry The MnO2/Zn electrochemical couple provides a theoretical open circuit voltage (OCV) of 1.608 V.8 The theoretical one electron capacity of manganese dioxide can be calculated by the following equation: Theoretical Capacity = F/3.6M, where M = 86.96 g, F = 96,485 C/mol  [Eq. 1.4] Although the theoretical capacity is 308 mAh/gEMD, in practice it is uncommon to obtain a capacity above 250 mAh/gEMD. It should be noted that the first discharge cycle capacity is almost always higher than that of subsequent discharge cycle.  During the rechargeable discharge of a RAMTM cell, the first electron reduction of manganese dioxide occurs according to the following electrochemical reaction5: MnO2 + H2O + e-  MnOOH + OH- Eo = 0.36 V vs. SHE   [Eq. 1.5] The following reaction is considered to be reversible, and for this reason voltage limits are placed on RAMTM cells in order to prevent the formation of several “electrically irreversible” species such as the following manganese oxide and hydroxide species:5 MnOOH + H2O + 3OH-  [Mn(OH)6]3-      [Eq. 1.6] Primary use alkaline cells will typically discharge the second electron and have significantly higher specific capacities, however these reactions are considered to be non-rechargeable:5 [Mn(OH)6]3- + e-  Mn(OH)2 + 4OH-   Eo = -0.25 V vs. SHE  [Eq. 1.7]   14  Further reduction of manganese oxyhydroxide can form an electrochemically inactive phase called Hausmannite (Mn3O4), and is thought to be one of several reasons for cumulative capacity fade in RAMTM cells due to a significant increase in cell resistance:5,23 2 MnOOH + Mn(OH)2 + e-  Mn3O4 + 2 H2O      [Eq. 1.8] The electrochemical processes occurring at the manganese dioxide cathode have been characterized in more detail, and will be the focus of this work.  The anodic reaction consists of zinc oxidation and can be represented by Equation 9:5 Zn + 2 OH-  ZnO + H2O + 2 e-   Eo = -1.248 V vs. SHE   [Eq. 1.9] Combining this with the single electron discharge of manganese dioxide to manganese oxyhydroxide, the overall cell reaction and thermodynamic cell voltage is as follows:5 Zn + 2 MnO2 + H2O  ZnO + 2 MnOOH   Eo= 1.608 V   [Eq. 1.10] The discharge reduction of EMD to MnOOH occurs through a homogeneous process of proton and electron insertion into the EMD lattice, thus causing expansion of the cathode during discharge24,25.  This process can cause the cathode to expand by as much as 10%, increasing the porosity and influencing mechanical properties of the electrode.25  Upon charging (oxidation of MnOOH to EMD), the lattice relaxes back to its initial state.  The repeated expansion and relaxation occurring over many discharge/charge cycles of an EMD electrode may have implications towards the cycling performance and capacity fade of MnO2/Zn cells.   As previously stated, the thermodynamic cell voltage for a MnO2/Zn cell is 1.608 V.  As the cell is discharged the voltage drops until a defined voltage cut-off limit is reached.  The voltage plateau for a MnO2/Zn cell is between 1.2 to 1.3 V vs. Zn/ZnO, a reasonable estimate of the working voltage for a MnO2/Zn cell, and is shown in Figure 1.6.   15   Figure 1.6 - Typical Voltage Capacity Plot of MnO2/Zn Cell in Flat-Plate Configuration with C/10 Rate of discharge to 0.9 V.  Capacity vs. cycle data for this cell is presented in Figure 1.7. 1.2.3 Cathode Capacity Failure The performance of rechargeable MnO2/Zn cells in commercial markets has been limited primarily by the cumulative capacity fade present in deep discharge cycling applications.5  Under these conditions the capacity can drop as much as 75% over 25 cycles8, severely limiting the potential for long-term repeat discharge applications.  The deep discharge performance of AA rechargeable MnO2/Zn cells is similar to that of flat-plate rechargeable cells with respect to capacity fade, with dramatic capacity loss occurring within the first 5-10 cycles, and stabilizing to a linear rate of fade of 1-2%/cycle as shown in Figure 1.7.   16   Figure 1.7 – Capacity vs. cycle for AA cylindrical alkaline cell (continuous 10 Ω discharge to 0.9 V at room temperature5) and flat-plate RAM cell (C/10 rate of discharge to 0.9 V at room temperature).           17  1.3  Manganese Dioxide Properties 1.3.1 Manganese Dioxide Solid State Chemistry  Manganese dioxide is a black/brown solid which composes approximately 80% of the weight of cathode material in RAMTM cells.  In nature, manganese dioxide is commonly formed through mineral deposits of pyrolusite and ramsdellite.  In alkaline MnO2/Zn batteries a specific grade of manganese dioxide called electrolytic manganese dioxide (EMD or alternatively γ-MnO2) is used as a result of its improved performance.  EMD is composed of an intergrowth (De Wolff defect) of ramsdellite and pyrolusite.6    Figure 1.8 - Electrolytic Manganese Dioxide Structure.6  The structure of γ-MnO2 has been historically very difficult to characterize due to the complex structure including the intergrowth of various MnO2 phases in varying fractions, and manganese vacancies.  Some EMD samples are representative of mostly pyrolusite, while others may be representative of primarily ramsdellite or some mixture of both.7  Paul Ruetschi has attempted to further characterize the structure   18  of EMD including manganese vacancies.  Assuming that Mn4+ cations are generally in repeating lattices, some fraction of these ions will be vacant and coordinated with 4 hydroxide ions replacing O2- ions.7  In this way, protons coordinate a manganese vacancy and allow the lattice to locally retain its neutral charge.  In addition to these vacancies, some Mn4+ ions are replaced with Mn3+ ions with an additional hydroxide anion replacing a single O2- anion.  This combination of De Wolff and Ruetschi defects attempts to explain the wide variability of electronic and proton conductivity from sample to sample as well as other important parameters such as theoretical capacity and chemisorbed water content.7   The general formula Ruetschi proposed for γ-MnO2 is Mn1-x-y4+Mny3+O22-OH4x+y , where y is the number of Mn3+ replacements, and x is the number of Mn4+ vacancies.  This proposed structure is detailed in the following Figure 1.9 with Mn4+ vacancy highlighted with a dashed line and Mn3+ replacements with a solid line:7  Figure 1.9 - Proposed Electrolytic Manganese Dioxide Structure including Ruetschi Defects.7  1.3.2 Synthesis of Manganese Dioxide The electrolytic manganese dioxide composition is commonly above 90% MnO2 with approximately 5% free and bound water molecules, and 3-5% comprising insoluble HCl and sulfates.  The remainder (<1%) is made up of metal impurities from the electrolytic deposition process.16  Commonly, EMD is produced   19  through anodic oxidation of manganese sulfate in sulfuric acid.26  Other metals such as iron, calcium, potassium and sodium are soluble in acidic media and can be sources of contamination.26   Electrolytic manganese dioxide is the form of MnO2 most commonly used for alkaline battery applications. However, another form of MnO2 called chemical manganese dioxide (CMD) is produced by several methods using chemical processes.27  One method of CMD production involves using manganese sulfate as a starting material which is first dissolved in acidic media.  From there, various metals (e.g., nickel, bismuth, and titanium) can be included into the MnO2 lattice replacing Mn4+ by including other metal oxides.   The solution is titrated with sodium hydroxide to which a suitable oxidizing agent is added to produce manganese dioxide.  The role of nickel and bismuth in replacing Mn4+ ions is not well understood, however, the doping of nickel improves the  cycle life of CMD materials and bismuth improves the total discharge capacity.27  Commercially available samples of EMD are generally of higher purity and structural water content compared to CMD.  Higher water content has been described by Ruetschi28 and has been shown to influence the performance of EMD by improving proton diffusion into the EMD lattice structure.    1.4 Recent Advancements in Rechargeable Alkaline Cells 1.4.1 Cathode Additives The use of specific additives in the manganese dioxide cathode has been shown to have a considerable positive effect on the rechargeable capacity of Zn RAMTM cells.  The inclusion of barium sulphate (5 wt. %) was shown to improve the cumulative capacity fade of flat-plate Zn RAMTM cells by up to 24%.8  In addition, barium manganate (BaMnO4) has shown improved capacity, and it is hypothesized that the manganate (MnO42-) anion might be reduced to manganese dioxide, providing more active cathode material which can be further reduced to manganese oxyhydroxide as described previously.8   20  Other additives such as barium hydroxide ( Ba(OH)2 ) have been shown to suppress the dissolution of Mn(III) ions during cycling and prevent the formation of electrochemically inactive species.  The effect of this result is apparently most pronounced at 2-3 weight % Ba(OH)2.29   Titanium dioxide as an additive has been investigated by electrochemical impedance spectroscopy (EIS) in a half-cell setup to cycle the manganese dioxide cathode. It has been shown to decrease the charge transfer resistance (Rct) of the manganese dioxide cathode by up to 75%.30  These types of studies provide supporting evidence for the mechanism of capacity fade in RAMTM cells. The use of bismuth-doped manganese dioxide (BMD), by the inclusion of Bismuth Oxide (Bi2O3) at 10% by weight has been shown to improve the single cycle capacity of RAMTM cells.31  Similarly, a number of bismuth based compounds have been investigated for rechargeable alkaline cells including BaBiO3, and Ba0.6K0.4BiO3 in a thin film form factor tested over 30 recharge cycles.32  While these additives commonly had initial drops in capacity, the cycling performance and cumulative capacity fade has been improved over 30 cycles.32  It is theorized that the inclusion of these bismuth compounds prevents the formation of related manganese oxide compounds birnessite and hausmannite, which are contributors to cumulative capacity fade of the EMD-based cathode.32 The inclusion of TiB2 and TiS2 has been tested in small amounts to provide an improvement in cycling performance of the EMD-based cathode.  In addition, the combination of Bi2O3 and a small amount of TiS2 and TiB2 appears to have a synergistic effect on capacity fade.33   Evidently, a significant number of additives have been tested in EMD-based cathodes with limited success.  The rational investigation of these additives ultimately has been inconsistent, specifically with respect to a complete understanding of the electrochemical role these additives play in cycling performance.  The main goal of this thesis is to examine the role of a number of additives on the performance of flat plate alkaline MnO2/Zn batteries. Table 1.4 - Rechargeable Alkaline Cell Cathode Additives.   21  Cathode Additive Source Year Weight % Result(s) BaSO4 8 2006 5% Improved cycling performance over 25 cycles compared to control (no additive) BaMnO4 8 2006 5% Slightly improved performance over BaSO4 additive Ba(OH)2 29 2012 0-5% Mn3+ dissolution was prevented and cumulative capacity improved by 2.5% with 2% Ba(OH)2 over 10 cycles TiO2 30 2012 10% TiO2 incorporated into EMD bulk structure and reduced charge transfer resistance (Rct) by up to 75% Bi2O3 31 2010 10% Improved single cycle capacity by up to 25% BaBiO3 32 2005 1-5% Supressed formation of unwanted electrochemically inactive birnessite and hausmannite manganese oxide phases. TiB2/TiS2 33 2006 <5% TiB2 and TiS2 improve cycling performance and have a synergistic effect with Bi2O3 1.4.2 Flat-Plate Architecture The development of flat-plate rechargeable alkaline cells provides many advantages beyond battery technologies in use currently.  The rechargeable cylindrical alkaline cell is currently an attractive technology due to its low cost and environmentally superior properties.  The active materials are abundant, low cost, and environmentally superior to many other alternatives.  On top of these advantages, a flat-plate configuration would offer other improvements beyond the cylindrical alkaline geometry.     22   Figure 1.10 - Flat-Plate (left)8 and Cylindrical Bobbin (right)9 Rechargeable Alkaline MnO2/Zn Cell Architectures. Flat-plate geometries are expected to improve upon cylindrical cells specifically in terms of achievable current densities and utilisation of active materials.  Transitioning to very thin electrodes promises even better use of active materials.  These properties of flat-plate electrodes become significant when scaling up towards larger battery packs as well as when scaling down towards high energy density electrodes for portable electronics. Table 1.5: Advantages of Rechargeable Alkaline MnO2/Zn Battery Chemistry (left) and Flat-Plate Cell Architectures (right) Advantages of Rechargeable Alkaline MnO2/Zn (RAM) Cell Chemistry Advantages of Flat-Plate vs. Cylindrical Bobbin Cell Architectures Abundant, inexpensive active materials Improved mass transfer characteristics Moderate volumetric and gravimetric energy density compared to other battery chemistries Ability to achieve higher current densities due to thinner electrodes Non-toxic, Zero CO2 emissions (during operation) Improved utilisation of the active materials No memory effect Excellent charge retention  Although the targeted applications for flat-plate electrodes may be very different, many of the design principles are transferrable between small scale cells and larger scale battery packs.  Large scale flat-plate battery packs will benefit from the stacking geometry of flat-plate electrodes.  Small scale flat-plate cells   23  will benefit from the ability to achieve higher current densities which will be necessary for more demanding portable electronic applications. 1.5  Objectives  The main objectives of this thesis are to improve the long term cycling performance of flat-plate rechargeable alkaline MnO2/Zn cells and determine the mechanism(s) of capacity fade.  The specific objectives of this work are as follows: 1) To develop a lab-scale hardware to test and validate the performance of flat-plate rechargeable alkaline MnO2/Zn cells. 2) To characterize and understand the causes of cumulative capacity fade in flat-plate rechargeable alkaline MnO2/Zn cells. 3) To improve the rechargeability of flat-plate rechargeable alkaline MnO2/Zn cells through the investigation of cathode additives. 4) To improve the rechargeability of flat-plate rechargeable alkaline MnO2/Zn cells through the investigation of various operating conditions including charging protocols and electrode fabrication procedures.     24  2 Experimental Methods 2.1 Electrochemical Measurements 2.1.1 Electrochemical Measurements: Experimental Full-Cell Setup The full-cell setup is important step in the development of flat-plate rechargeable alkaline cells.  The full-cell performance is a close representation of that expected in a commercial cell, but scaled down to smaller electrodes (approximately 2 cm x 4 cm) in order to minimize raw materials use and to expedite electrode processing time.  Figure 2.1 - Flat-Plate Rechargeable Alkaline Full-Cell Setup.  Figure 2.1 shows the main components of the full cell setup developed in the Wilkinson research group.  This setup contains a zinc anode, an EMD-based cathode, separator components, and current collectors (CCs) all required for full cell operation. The cathode is constructed from EMD, graphite, and appropriate binders and chemical additives.   The cathode CC is a nickel mesh (Product ID: 5Ni7-077) from Dexmet Corporation chosen for its good stability in highly alkaline solutions and high electrical conductivity.  The   25  cathode has a 2 cm x 4 cm area and is approximately 0.5 mm thick after being processed with a metal sheet roller and pressed in a two-step process initially with 4 MPa (2 minutes) and then with 10 MPa (2 minutes) between two sheets of wax paper.    The anode is constructed from a 5:1 ratio of Zn/ZnO provided by Horsehead Corporation and Zochem Inc. along with specific additives to prevent hydrogen evolution (Indium Sulfate, Polyethylene Glycol) and improve mechanical stability (PTFE suspension).  The anode CC is comprised of a tin plated brass mesh (Product ID: 4Br7.5-080TP) provided by Dexmet Corporation which was chosen for its stability in alkaline solution and high overpotential for hydrogen evolution.   The details of electrode preparation are given in Appendix A.  There are a few important distinctions between the full and half-cell setups which may influence performance and ultimately many aspects of the eventual goal of this project, larger scale battery packs.  First, there is a zinc anode present in a full-cell instead of an inert nickel mesh counter electrode in a half-cell.  The electrochemical properties of the zinc anode may influence many factors of cell performance such as the potential formation of zinc dendrites or haeterolyte (Zn2MnO4) formation.34  Second, a minimum amount of electrolyte can be used in the full-cell setup, as ionic contact with a reference electrode is not required.   The added advantage of using minimal amounts of electrolyte is that hydrogen and oxygen evolution are a less prominent concern, as less water is available for this gas evolution to occur.  However, optimization of these parameters is an important factor in the performance and operation of a full-cell because electrolyte loss could result in cell failure.  In the full-cell setup a reference electrode is not used as the potential of the cell is measured as the difference in potential between the EMD-based cathode and the zinc based anode.  However, a reference electrode could be used in order to independently measure the potentials of the anode and cathode electrodes.   26  Another important factor in the full cell setup is the relative amounts of anode and cathode materials.  In this setup, an excess of anode active materials (approximately twice the stoichiometric ratio required) was deliberately used to ensure the anode capacity was not limited.  Additional factors present in the full-cell which may influence performance as a result of two electrochemically active electrodes such as possible zinc contamination in the cathode, or separator degradation due to the formation of zinc dendrites during charge. 2.1.2 Electrochemical Measurements: Half-Cell Setup An electrochemical cell is generally constructed with two electrodes referred to as the anode (electrochemical oxidation) and cathode (electrochemical reduction) in order to do electrical work through two separate reactions.  In order to isolate and study the fundamental properties of a single electrode a half-cell setup can be used.  In this work, a half-cell setup was constructed in order to study the EMD-based cathode’s cycling performance.   This half-cell setup was constructed from Delrin plastic to form the housing.  The choice of Delrin was obvious due to its affordable cost, ease of machining, and stability to 9M potassium hydroxide solutions, which is the most common electrolyte for alkaline cells.  The electrode assembly is comprised of an inert counter electrode of expanded nickel from Dexmet Corporation (Product ID: 5Ni7-077) with length and width dimensions 2 cm x 4 cm , a bilayer (woven and cellophane) separator provided by Neptco Inc. (dimensions 2.5 cm x 4.5 cm) and an EMD-based working electrode developed in house (dimensions 2x4 cm).  Two nickel plated stainless steel bolts serve as current collectors for the working and counter electrodes. On top of the electrode stack is a tin-plated brass plate used to compress the electrode stack.  Two springs each with a spring constant 30 kg/cm (168 lb/in) are placed on top of the brass plate and apply a constant   27  force of approximately 3.3 kg/cm2 during cycling.  The brass plates are plated with a two part electroless tin plating solution (provided by Transene Company, Inc.35) used because of the high overpotential for hydrogen evolution of tin coupled with stability in basic solutions. Table 2.1 - 10 mA/cm2 H2 Evolution Reaction (HER) Overpotentials for Common Metals.36   The choice of reference electrode is important for half-cell studies.  Many reference electrodes are not stable in highly basic (pH ~15) solutions. Hg/HgO reference electrodes from Koslow and Radiometer Analytical were used for these half-cell experiments37. Hg/HgO reference electrodes are the common choice as they are stable in basic solution and use the same electrolyte as their filling solution (albeit with lower concentration).  The Hg/HgO reference electrodes were calibrated regularly in order to ensure the accuracy of potential measurements, and subsequently the correct depth of discharge.  The procedure for reference electrode calibration can be found in Appendix C. Electrode Type Overpotential (V) Platinized Pt 0.015 Cu 0.479 Ag 0.475 Fe 0.403 Ni 0.563 Graphite 0.599 Sn 0.856   28   Figure 2.2 - Schematic of Half-Cell Setup of EMD-based Cathode.  Figure 2.2 details the half-cell setup used to cycle the EMD-based cathode.  This setup has the advantage of separating out extraneous factors present in the full-cell setup and helps to diagnose and isolate performance constraints of the cathode only.  This half-cell setup removes the zinc based anode and other hardware components necessary for full-cell operation, and allows experiments to focus on a single electrode and its performance.  The cause of cumulative capacity fade can now be attributed to the performance of the EMD-based cathode only.  This is an important tool in the development of flat-plate rechargeable alkaline cells as it allows one to study the specific properties of the cathode and to be confident that these properties are not a result of extraneous factors such as the anode or separator components.  In addition, one can be confident that there is no contamination from the zinc anode.  Due to the fact that a reference electrode is required for electrochemical measurements in the half-cell setup, an excess amount of electrolyte is required in order to ensure ionic contact with the reference electrode.  Figure 2.3 shows a 3D rendering of the half-cell hardware with the cell body marked   29  transparent to show the components including the Hg/HgO reference electrode, brass pressure plate, compression springs, and SHCS working and auxiliary electrode contacts.  Figure 2.3 – 3D rendering of half-cell hardware. Designed and manufactured by Greg Afonso. 2.1.3 Electrochemical Measurements: Cell Cycling Regimes Computer controlled battery chargers were used to charge/discharge full and half cells using various charging techniques.  Full cells were controlled by an 8 channel Maccor 4300 series38 or MTI battery analyzer.39  Half-cells were controlled using a Solartron 1470 E Potentiostat40 with a separate additional impedance analyzer.  The typical cycling regime employed is a 3 step protocol:   30   Figure 2.4 - 3 step charge/discharge protocol.  1) Galvanostatic discharge, 2) Galvanostatic Charge, 3) Potentiostatic Charge.  1. Galvanostatic Discharge:  Constant current is drawn from a cell in the charged state to a predefined cut-off voltage.  Figure 2.4 shows a typical discharge/charge cycle for a flat-plate rechargeable alkaline full and half-cell.  The cut-off voltage is the termination condition for this step and can range anywhere between the typical open circuit voltage (~1.55 V) and 0.90 V vs. Zn/ZnO or -0.40 V vs. Hg/HgO depending on the setup used.  The current drawn from the cell is a measure of the rate of discharge and depends on the mass of active cathode material present. It is assumed that EMD provides a nominal capacity of 200 mAh/g.  For example, an electrode comprised of 1 g EMD discharged at a rate of C/10 (10 hour discharge) will require a 20 mA current for discharge.    31  2. Galvanostatic charge:  A constant current is applied to charge a cell from the discharged state to the charged state until the desired cell potential is reached (1.75V vs. Zn/ZnO or 0.35V vs Hg/HgO).   Typically this step will use the same magnitude of current as the galvanostatic discharge step however any rate of charge can be employed.  3. Potentiostatic charge:  The cell is held at a constant potential (1.75 V vs. Zn/ZnO or 0.35 V vs. Hg/HgO) for a period of time (2-10 hours) until the current profile decays to approximately zero.  The amount of charge obtained during this step can be calculated by integrating the current vs. time curve.  The potential values indicated in Figure 2.4 are either measured against the zinc anode (full-cell setup) or against a Hg/HgO reference electrode (half-cell setup).  2.2  Electrode Current Collectors  Two types of flat-plate CCs for this project were developed in order to test the performance of additives and cathode recipes in both the half and full-cell setups.  Figure 2.5 and Table 2.2 compare the properties of the nickel mesh and nickel foil form CCs and the different aspects of their construction and development.     32   Figure 2.5 - Cathode Electrode Current Collectors (CCs): Nickel Foil (top) and Expanded Nickel Mesh (bottom).  There are significant advantages to both types of CCs which motivated the development of each.  The expanded nickel mesh CC can provide higher volumetric and gravimetric energy density compared to the nickel foil CC. This conclusion follows naturally from a comparison of EMD and CC mass in both types of electrodes.  Electrodes prepared with the expanded nickel mesh CC have an EMD mass percentage of 50%, while electrodes using the nickel foil CCs are 30% EMD.  The energy density and thickness of thin film form factors could be optimized to increase this ratio with further work.  The relative amounts of EMD in the same geometric area electrodes vary by a factor of 5.           33   Table 2.2: Comparison of Cathode Electrode Current Collectors.  Nickel Foil Expanded Nickel Mesh Electrode Mass (g) 0.3g 1g Thickness (mm) 0.15mm 0.6mm Dimensions (cm x cm) 2x4 cm 2x4 cm Composition EMD, C, Additives, CMC gelling agent EMD, C, Additives, CMC gelling agent, SBR binder Ease of preparation Easy, bulk production Difficult, handmade Wet Consistency Ink/Slurry Paste Dry Consistency Dried black powder Dried black powder Reproducibility More consistent and reliable Consistent EMD Mass (g) 0.1g 0.5g Current Collector Mass (g) 0.20g 0.30g EMD mass percentage (%) ~30% ~60%     The methods of preparation are also significantly different between the two types of CCs. The Ni foil CC electrode material is mixed with a probe tip sonicator ensuring a homogeneous mixture.  The expanded Ni CC electrodes are hand mixed extensively (procedure in Appendix A); however, this could be a source of inconsistency between electrode batches.  Once an ink has been prepared for the Ni foil CC, it is applied evenly with a block spreader to prepare a large number of electrodes with essentially uniform thickness and composition.  The expanded Ni mesh electrode materials are individually hand-mixed and with such a manual process electrodes are inevitably prone to some inconsistencies.  Figure 2.2.2 shows the process involved in the fabrication of both types of electrodes and their differences in flexibility.    34   Figure 2.6 – Electrode fabrication procedures for expanded nickel mesh (left) and thin film nickel foil (right) CC showing flexibility of both types of electrodes.  2.3 Analytical Characterization 2.3.1 X-Ray Diffraction (XRD) Studies X-Ray radiation with a wavelength of about 1 Å (similar to interatomic or intermolecular distances) can be used to probe the crystalline structure of various compounds on the atomic level.41  In order to generate X-rays of a known wavelength, a metal target (often copper) is bombarded by an electron beam which excites electrons in the target metal and in turn ejects X-rays.41  Generated X-rays will have a frequency and energy proportional (or wavelength inversely proportional) to the difference in energy level of the ejected (K-shell, 1s) electrons and higher energy (K or M-shell) electrons.41   35  X-rays will diffract with repeating planes within a crystal lattice, and produce a diffraction pattern which can be used as a fingerprint to identify the crystal phases present.  This diffraction occurs according to Bragg’s Law:41 nλ = 2dsinθ      [Eq. 2.1] where n is the order of diffraction, λ is the incident x-ray wavelength, d is the spacing between adjacent repeating crystal lattice planes, and θ is the X-ray incident angle.  Diffraction patterns (intensity of X-rays vs. various angles 2θ) can be compared against those of known samples in a database and identified based on structural information of the crystal lattice.  XRD analysis allows one to determine crystalline phase changes which may occur during cell charge/discharge experiments.  Electrodes were cut into 1 cm2 squares or carefully removed manually by scraping with a metal spatula and ground up with a mortar and pestle.  Samples were analyzed with a Bruker AXS D2 Phaser X-ray diffractometer (see Appendix C for details).  Samples were tested upon preparation, as well as after cycling in various states of charge.   2.3.2 Scanning Electron Microscopy (SEM) and Energy Dispersive X-Ray Spectroscopy (EDS) Scanning electron microscopy uses an electron beam to image the morphology of a solid surface.  These electrons interact with the sample producing either secondary electrons (SE) or backscattered electrons (BSE).10  The volume within the sample which electrons interact is often referred to as the interaction volume (see Figure 2.7).  The interaction volume depends on various factors including sample composition, density, morphology, and electron beam accelerating voltage.10  Two main types of electron emission result from the interaction of an electron beam and sample.  Secondary electrons result from inelastic scattering (interaction with atom’s electric field) and are defined   36  by energies lower than 50 eV and can give surface and near surface information.10  Backscattered electrons result from the interaction of incident electrons and the electric field of the nucleus of atoms in a sample.  This can provide information about elemental composition and can be used to deduce an elemental contrast map.10   Figure 2.7 - Scanning Electron Microscopy Interaction Volume.10  The spectrum of x-rays emitted from deeper within the sample can be used to identify the elements present in the sample, providing a signal for Energy Dispersive X-Ray Spectroscopy (EDS).  EDS detectors are normally coupled to SEMs, thus allowing the user to simultaneously investigate morphology and elemental composition.  The analysis of cross sections and surfaces of electrodes was conducted in order to provide supporting evidence and characterization of capacity fade mechanisms in the EMD-based cathode.  Samples were analyzed using a Helios NanoLab 650 Focused Ion Beam SEM at the UBC Centre for High Throughput Phenogenomics (CHTP), described in Appendix C.   37  2.3.3 Potentiostatic Electrochemical Impedance Spectroscopy (PEIS) The ideal resistor behaves according to Ohm’s law, where the resistance is equal to the  voltage divided by the current under all voltage and current values.  However, more complex systems such as a rechargeable alkaline cell must be considered as a combination of various circuit elements representing the capacitive behaviour of the electrodes and the electrode/electrolyte interface as well.  For systems such as these, a more general resistance parameter called impedance is more applicable.  The impedance of a cell is measured by applying a small amplitude (usually 1-10 mV) sinusoidal voltage and measuring the current response.  Typically, within a small range of voltages the current will respond linearly.  A small amplitude signal is used so as to not perturb the system from linearity, and so that its electrochemical properties will not be influenced by the excitation signal.    The response of the system (as measured by current) is generally phase shifted by some amount (Φ) in time compared to the excitation signal.  This phase shift can be used to express the impedance as a complex function with real and imaginary components using Euler’s relationship.  A typical method of graphically representing the real and imaginary impedance is accomplished with a Nyquist plot.  Figure 2.8 – Idealized Nyquist plot noting bulk resistance (Rs) and charge transfer resistance (Rct) with equivalent circuit model.   38   Figure 2.8 shows the idealized Nyquist plot and key parameters with the associated equivalent circuit.  The bulk resistance (Rs) is a measure of the ohmic resistance of the cell including all components of electrodes, electrolyte, CCs and is a measure of the ohmic resistance of the cell.  The charge transfer resistance is a measure of the propensity of a charge transfer reaction which occurs on the surface of a solid material in contact with the electrolyte.  With respect to the cathode of alkaline MnO2/Zn cells, this refers to the surface reduction of MnO2 and the associated process of proton intercalation.   Using PEIS measurements, it is possible to model the equivalent circuit which represents the impedance behaviour of the system under various voltage frequencies.  However, in practice many different equivalent circuits can model the same system thus requiring a careful and judicious choice of circuit model.  This work focuses on changes and trends observed in these key parameters and linking these changes to the performance of the cell.  It is also possible to measure impedance by using a sinusoidal current excitation signal (GEIS - as opposed to voltage excitation signal) and measuring the resulting voltage response.  GEIS measurements were not performed in this work, however. The PEIS measurements reported were collected using a BioLogic VMP3 potentiostat with a frequency range of 10 mHz to 200 kHz, a voltage amplitude of 10 mV, and with 6 points per frequency decade.  All the impedance measurements were performed at open circuit voltage after galvanostatic discharge and potentiostatic charge. 2.3.4 X-Ray Photoelectron Spectroscopy (XPS) XPS is a surface sensitive technique which allows detection of elemental composition, and oxidation state of the elements present within a sample.   XPS measurements are conducted under ultra-high vacuum (UHV) and are generally sensitive to a depth of 5-7 nm in a material. XPS spectra are obtained by shining X-rays at a sample and measuring the binding energy of electrons ejected from a sample.   39  XPS analyses were performed with a Kratos Axis Nova spectrometer using a monochromatic Al Kα source (15 mA, 14 kV).  The instrument work function was properly calibrated to give a binding energy of 86.93 eV for the 4f7/2 line and 932.62 eV for the Cu 2p3/2. The analysis was performed over an area of 300 x 700 microns on the electrode surface.  These measurements were conducted at the surface science lab at the University of Western Ontario.42 2.3.5 Galvanostatic Intermittent Titration Technique (GITT) GITT measurements were performed using the Maccor 4300 series battery charger by programming a galvanostatic charge/discharge protocol and intermittently resting the cell so that its potential can equilibrate.  GITT measurements can determine the deviations from the equilibrium state at a given state of charge.  GITT measurements can be contrasted from potentiostatic intermittent titration technique (PITT) measurements which alternate potentiostatic and rest periods, while measuring current.  The GITT measurements presented in section 3.1.4 were performed at a C/2 rate of discharge for 20 minute intervals with 30 minute rest periods until a lower cut-off potential of 1.1 V was reached.             40  3 Results and Discussion 3.1 Effect of Different Operating Conditions 3.1.1 In-Situ Pressure Application  Stack pressure, defined as the pressure applied to the anode and cathode electrodes during cell operation, has been shown to improve the cycling performance of flat-plate RAMTM cells.  This issue is most pronounced with no pressure applied resulting in progressive degradation until cell failure after 5 to 10 cycles.  In cells operating without stack pressure (electrodes suspended in a solution half-cell setup), the beginning of the galvanostatic discharge step is marked by a sudden voltage drop.  Figure 3.1 details the current and voltage vs. time profiles for two half cells with a) no in-situ pressure application and b) in-situ pressure application of 47 psi (3.3 kg/cm2).           Figure 3.1 – Current and Voltage vs. Time for a) no pressure applied, and b) 47 psi (3.3 kg/cm2) stack pressure. Half-cell setup, discharged at a C/10 rate to -0.4 V vs. Hg/HgO.  Large potential drops at the onset of the galvanostatic discharge step (Figure 3.1 a)) are indicative of poor electrical contacts within the electrode in cells without applied stack pressure.     41   Figure 3.2 presents the first, third and fifth closed loop voltage vs. capacity profiles of these two half cells.  Although a progressive voltage drop is present in the cell without applied stack pressure, the specific capacity of both cells is quite similar.  This suggests that the progressive failure of cells with no stack pressure is likely due to a physical phenomenon such as poor electrical contact between the active material and the CC.    Figure 3.2 – Voltage capacity profiles for 47 psi (30kg/cm2) stack pressure (a) and no applied pressure (b) for cycles 1, 3, and 5. Arrows indicate the directions of charge and discharge.  As the voltage drop occurring without stack pressure is progressive, the initial (first cycle) specific capacity is almost identical to cells with stack pressure.  Although the voltage drops significantly in cells without stack pressure, it will often recover before complete cell failure occurs.  The effect of stack pressure on the cycling performance of cells can be assessed by plotting the specific capacity per cycle.      42   Figure 3.3 – Capacity vs. cycle number for 47 psi (3.3 kg/cm2) stack pressure and no stack pressure.   Figure 3.3 shows the cycling performance of half-cells with and without stack pressure.  With no stack pressure, the cell’s ability to cycle vanishes after several cycles.   This is due to poor contact with the CC and delamination of the active layer.  Pressure sensitive paper (Fujifilm Prescale LLLW R270 5M 1-E Film) spanning a range of 29-87 psi (2-6.1 kg/cm2)43 was used to test the applied stack pressure inside the cell.  This paper consists of two sheets which undergo colour changes with applied pressure.  The intensity of this colour change quantitatively represents the pressure on any given area of the electrode surface.  Figure 3.4 shows two examples of good and poor pressure distributions.     43   Figure 3.4 – Uniform (left) and non-uniform (right) pressure distribution observed by the use of pressure sensitive paper.  Figure 3.5 shows delamination of active materials from the CC after cycling in a flooded half-cell with no stack pressure.   The electrode was removed from the cell after failure and cut in half down its length to indicate delamination from CC.  Figure 3.5 – Delamination of an electrode after cycling in a flooded half-cell setup.  Electrode cut in half lengthwise with scissors after being removed from flooded half-cell setup.  3.1.2 Depth of Discharge (DOD)  Depth of discharge (DOD) refers to the extent that a cell is discharged. DOD can be defined by the potential to which the cell is discharged.  The depth of discharge is an important feature of the cell operation   44  parameters.  Figure 3.6 shows the cell discharge profiles of 1.1 V, 0.9 V, and 0.5 V cut-off potentials for the first and fifth cycle.  The discharge profiles of all of the fifth cycles show reduced capacities.   Figure 3.6 – Voltage vs. Capacity for 1.1, 0.9, 0.5 V potential cut-off for first and fifth cycles.    The solid line (0.5 V cut-off) shows a second voltage plateau commencing at 1.0 V during the first discharge. This plateau corresponds to the 2nd electron discharge and is not present in subsequent cycles indicating that this process is irreversible. Capacity fade is more significant in cells when a lower potential cut-off is used.  A comparison of the 0.5 V and 1.1 V cut-offs show a drop of 120 mAh/g and 50 mAh/g, respectively. The relative capacity losses are significant as the 0.5 V and 1.1 V cut-offs show a 50% and 25% decrease in capacity over 5 cycles, respectively.  Deeper depth of discharge (0.5 V) leads to a larger capacity loss, in agreement with previous reports.5      45  Figure 3.7 shows the charge/discharge profiles of cells discharged to three different cut-off potentials.  Discharge capacity between 1.1 and 0.9 V is relatively small (approximately 5%) compared to the discharge capacity present between open circuit potential and 1.1 V, thus for most experiments a lower cut-off potential of 1.1 V was used.  The voltage plateau at 0.9 V corresponds to the second electron discharge of EMD, which proceeds according to equations 3-5 in the Introduction Chapter (p. 12) forming several manganese oxide and hydroxide species.5  A corresponding plateau during the charge step is observed as well.  This lower potential charge/discharge plateau is observed only during the first cycle and appears to be irreversible upon further cycling, which indicates one cause of capacity fade, the conversion of EMD to other electrochemically irreversible manganese phases.5     Figure 3.7 – Differential capacity profile for 0.5 V (a), 0.9 V (b), and 1.1 V (c) cut-off potentials for the first and fifth cycles.  Charge/discharge voltage vs. capacity profiles for 0.5 V, 0.9 V, and 1.1 V cut-off potentials for the first (d) and fifth (e) cycle.    46  Differential capacity plots in Figure 3.7 confirm the existence of these plateaus, as each plateau is manifested as a peak in the differential capacity profiles.  Regardless of the cut-off potential used (0.5, 0.9 or 1.1 V vs. Zn/ZnO) a minor plateau is present at 1.4 to 1.5 V during the discharge process.  A second major plateau is present at 1.25 V, and a final plateau at 0.9 V.  The plateau between 1.4 and 1.5 V has been noted previously in literature and has been attributed to the intercalation of protons into a manganese cation vacancy, referred to as Ruetschi defects and to the reduction of surface Mn4+ cations.44    The plateau between 1.2 and 1.3 V can be attributed to the reduction of the ramsdellite portion of EMD to groutite.44  As previously described, the irreversible second electron discharge of EMD occurs at potentials below 1.0 V.  The proposed chemical reactions for the reduction of ramsdellite and the irreversible second electron discharge are shown below. MnO2 + H2O + e-  MnOOH + OH- Eo = 0.36 V vs. SHE  [Eq. 3.1] MnOOH + H2O + 3OH-  [Mn(OH)6]3-     [Eq. 3.2] [Mn(OH)6]3- + e-  Mn(OH)2 + 4OH-   Eo = -0.25 V vs. SHE  [Eq. 3.3] 3.1.3 Charging and Discharging Protocols  Full-cells were discharged and charged according to a 3 step protocol previously described in the Experimental Methods chapter (section 2.1.3) and is briefly explained here: 1) galvanostatic discharge to 1.1 V, 2) galvanostatic charge to 1.75 V, and 3) potentiostatic charge at 1.75 V.  Figure 3.1.3.1 shows the cycling performance of full cells at different rates of discharge and charge. A C/20 rate (20 hour discharge) shows the best initial capacity of 215 mAh/gEMD.  A C/10 rate (10 hour discharge) shows similar performance with a capacity of 185 mAh/gEMD.  A C/1 rate (1 hour discharge) shows a much lower initial capacity of 65 mAh/gEMD.  This indicates that the performance is highly dependent on rate of discharge.  In addition, there does not appear to be a plateau present in the C/1 rate which is present at C/10 and C/20 rates.  This can also be characterized by comparing the midpoint voltage of each C-rate used.  The   47  midpoint voltage of the C/1, C/10, and C/20 rates are 1.2, 1.27, and 1.31 V, respectively.  This drop in voltage is consistent with the higher IR drop expected with the use of higher currents at faster discharge rates.  The IR drop associated with each C-rate assuming an EMD mass of 0.5 g, and cell resistance of 0.2 Ω is summarized in table 3.1. Table 3.1: Comparison of estimated voltage losses dependent on rate of discharge. Rate of Discharge Discharge Current (A) Voltage Loss (V) C/20 0.01 A 0.002 V C/10 0.02 A 0.004 V C/1 0.2 A 0.04 V    Figure 3.8 – Voltage capacity profile of baseline cells at C/1, C/10, and C/20 rates for first cycle with a 1.1 V cut-off.  Besides using a cut-off potential, discharge (and charge) steps can also be terminated using set periods of time.  For a nominal capacity of 200 mAh/gEMD, a percentage of full discharge capacity can be defined as   48  the termination condition for discharge.  For example, a cell discharging at a C/10 rate for 25% discharge would be set to complete discharge after 2.5 hours (150 minutes, 25% of 10 hour discharge).  Figure 3.9 shows the capacity vs. cycle performance of two time cut-off discharge regimes defined at 1 hour (60 minutes, 10% depth of discharge) and 2.5 hours (150 minutes, 25% depth of discharge) along with the end of discharge voltage.   Figure 3.9 – Capacity and end of discharge potential for C/10 rate time cut-off of full MnO2/Zn cells at 10 and 25% time cut-off depth of discharge.    The capacity is constant from cycle to cycle because the termination condition is based on time and is set to complete well before the full capacity of the cell is discharged.  Although capacity is constant, the voltage at the end of discharge drops with each cycle indicating cumulative cell failure.  These cells are failing in a similar manner to those discharged to a cut-off potential, however the manifestation of cell failure in this case is represented by a voltage fade rather than capacity fade.     49  A time cut-off discharge can provide hundreds of cycles if the discharge time is limited to a fraction (e.g. 5-25%) of the total one electron discharge time.  For example, the discharge of a cell at a C/10 rate could be terminated after 2 hours to utilise 20% of total discharge capacity and provide several hundred cycles.   In comparison, a cell fully discharged with a potential cut-off of 0.9 V vs. Zn/ZnO could provide approximately 100 cycles before complete failure with each cycle providing less capacity.  Although this cell would initially provide significantly more capacity per cycle, it would not last nearly as many cycles as a time limited cut-off cell.  Therefore, if a specific application requires consistent capacity per cycle then a time cut-off discharge protocol should be employed.    Employing a time or potential cut-off termination condition for cell discharge does not prevent cell failure; however it does use the cell’s capacity in two different operation modes.  Using a time limited cut-off (as in Figure 3.9) will provide a lower capacity at higher cell potentials, while a potential cut-off will provide a higher capacity although at lower voltages.  In both cases, cell failure is occurring but it is manifested in either a voltage or capacity drop per cycle.  Specific applications of MnO2/Zn cells may require intermittent deep discharges, such as portable devices requiring on demand use such as flashlights or portable charging devices.  In order to simulate intermittent use, Figure 3.10 shows two cells discharged every 20th cycle to 1.1 V while all other cycles are discharged to a higher potential (1.35 or 1.4 V).   50    Figure 3.10 – Shallow discharge (1.35 V C/2 rate and 1.4 V C/10 rate) with intermittent deep discharge (1.1 V) capacity vs. cycle.  Figure 3.10 shows two intermittent deep discharge regimes.  The first is a C/10 rate discharged to 1.4 V, while the second is a C/2 rate discharged to 1.35 V.  In both cases, the intermittent deep discharges are discharged to 1.1 V at their respective rates.  As expected, a faster discharge rate provides a lower capacity which is in agreement with previous experiments comparing various rates of discharge.  If the same C-rates were compared, it would be expected that lower depth of discharge (i.e., 1.35 V compared to 1.4 V) would provide a larger capacity.  In this case, the effect of a slower rate of discharge (C/10) provides a slightly larger boost in capacity (approximately 10 mAh/gEMD) compared to a 50 mV lower depth of discharge at a faster rate of discharge (C/2).  This effect is more pronounced comparing the intermittent deep discharge capacities which provide an increase in capacity of more than 70 mAh/gEMD for a deep discharge at C/10 versus C/2 rates.      51  Shallow depth of discharge voltage cut-off charging regimes may be suitable for high power, low capacity applications.  At higher voltages (i.e., OCV ~1.6 V) almost twice the power is supplied but it is only available for approximately 10% of cell discharge time as voltage drops as the galvanostatic discharge proceeds.  Figure 3.11 shows the available power throughout the galvanostatic discharge process.  Figure 3.11 – Power and voltage versus capacity with 20 mA galvanostatic discharge at C/10 rate of discharge.  The power versus capacity curve shown in Figure 3.11 shows that at higher potentials more power is supplied.  For example, although a higher potential of 1.6 V provides approximately 70% more power than that of a lower potential (0.9 V) with a 20 mA discharge, as discharge proceeds cell voltage and power drops as shown in Figure 3.11, and thus the high power is only available for a small percentage (10-15%) of the total cell capacity.   For this reason, higher power (shallow DoD) applications will suffer from lower energy density because more active material is needed in order to sustain a high cell potential for longer   52  periods of time.  Comparatively, lower power applications (deep DoD) can use the full discharge capacity of the cell.  Time cut-off discharge performance has also been demonstrated in a half-cell setup with performance detailed in Figure 3.12.  The rate of discharge has a significant effect on cumulative capacity fade as well as initial capacity, which agrees with the performance shown in a full-cell setup as in Figure 3.8.        Figure 3.12 – C/40 Rate discharge, 12 hour discharge (30% depth of discharge):  a) voltage capacity profile, b) cut-off potential and specific capacity vs. cycle, and c) projected cell lifetime in pressurized half-cell setup.  Typical modes of failure in the MnO2 cathode include deep discharge, fast rates of discharge, insufficient electrolyte, and zinc contamination from the anode.  In order to circumvent these modes of failure, a   53  MnO2 cathode has been cycled in a flooded half-cell with a time cut-off charge/discharge protocol with a C/40 rate of discharge.  Figure 3.12 a) shows the voltage vs. capacity profile for 20 cycles.  The first discharge voltage capacity profile in Figure 3.12 a) begins at approximately 0.2 V vs. Hg/HgO and completes discharge at just above 0 V vs Hg/HgO, with the highest cut-off potential compared to subsequent cycles.  It is evident that the end of discharge potential drops slightly as the cell cycles, however as shown in Figure 3.12 b) it appears to be stabilized after 10 cycles.    As with previous time cut-off discharge regimes, cut-off voltage is representative of capacity fade.   Similarly, capacity per cycle is constant as the time cut-off condition is met in order to maintain cell capacity.  Figure 3.12 c) shows a linearly projected cell lifetime before an end of discharge potential of -0.2 V vs. Hg/HgO (1.1 V vs. Zn/ZnO) is reached.  While this is speculative as cell data has only been obtained for 20 cycles, with this charge/discharge regime, 60 mAh/gEMD could be provided consistently for 140 or more cycles. 3.1.4 Galvanostatic Intermittent Titration Technique (GITT)   GITT measurements provide a measure of quasi-equilibrium potentials at various SOC.  For example, if discharge of the cathode can be simplistically viewed solely as a conversion of MnO2 to MnOOH, then at any given SOC some fraction of the cathode will be MnO2 and MnOOH.  The equilibrium potential of the cell will be a function of SOC or fraction MnOOH, dropping with increased fraction of MnOOH as discharge proceeds.  The GITT measurements performed also can provide kinetic information of the cell chemistry, as measured by the time required to equilibrate during a rest period. Gaining insight into the kinetics of the EMD-based cathode may help to explain why poor performance is observed at fast rates of discharge. Figure 3.13 shows the first cycle of control recipe cathode (84% EMD, 16% graphite) cells at various rates of discharge and the GITT measurements.    54   Figure 3.13 – Voltage vs. fraction of total discharge for C/1, C/5, C/10, C/20, and C/40 rates of discharge to 1.1 V cut-off potential. GITT measurements at C/2 rate of discharge with 20 minutes discharge, 30 minutes rest period.  The polarization losses associated with faster rates of discharge are very significant as seen in Figure 3.13.  At any given fraction of discharge, a parameter called polarization potential can be defined as the difference between charge and discharge potential (when plotted as in Figure 3.13).  Figure 3.14 shows the polarization potentials observed with each rate of discharge.   55    Figure 3.14 – Voltage capacity profiles of cells at C/1, C/10, and C/20 rates showing polarization potential losses (a) and polarization potential losses vs. relative current for C/1, C/2, C/5, C/10, C/20, and C/40 rates of discharge (b).  The polarization potential losses in Figure 3.14 are an indication of the efficiency of the cell and vary as much as 100 mV between a C/1 and C/40 rates of discharge.  Figure 3.14 b) shows that the polarization potential losses correlate to increased rate of discharge.  This further indicates that slow rates of discharge improve cell performance.   3.1.5 Differential Capacity Analysis Proton insertion in γ-MnO2  is generally attributed to two distinct processes occurring during discharge which can be easily identified in differential capacity plots.44 Figure 3.15 shows the differential capacity plots of EMD/Zn cells for a number of cycles. The first reduction peak (DP1) can be attributed to the reduction of Mn4+ on the surface of γ-MnO2 particles, while the second (DP2) can be attributed to the reduction of ramsdellite.44  Both DP1 and DP2 reduce with cycling indicating the loss of capacity which occurs during the cell cycling. Furthermore, we can observe a shift in the potential of the peaks. We note that potential of DP1, corresponding the reduction of surface Mn4+, does not significantly change with respect to cycling, however, DP2 shifts to lower potentials with cycling, while the corresponding charge   56  peak shows a shift to higher potentials. Overall the growing separation of peaks CP1 and DP2 as a function of cycling is an indication of increases in the cell polarization as a function of cycling. The source of this polarization increase is attributed to the formation of irreversible phases which occur during the cycling and we discuss these effects in a separate publication.11   Figure 3.15 - Differential capacity plot for control recipe cycles #1, #10, and #20 showing capacity fade.  Figure is annotated to show discharge peak 1 (DP1), discharge peak 2 (DP2), and charge peak 1 (CP1).  Figure 3.16 (a-f) shows the differential capacity of the first, third, and fifth cycles for each additive.  In each panel inset, the absolute value of the differential capacity peak (indicated as DP2 and CP1 in Figure 3.15) is also plotted vs. cycle number.  Decrease of peak heights corresponds to the capacity loss. Figure 3.16 (f) compares the first cycle data for all the additives. These sensitive differential plots show very similar charge/discharge indicating that the reaction mechanism of EMD has not changed. The best   57  performing additive, bismuth (III) oxide, shows a consistent gap between discharge and charge peak over 50 cycles.  The gap between discharge and charge differential capacity peak height is a result of approximately 25% of total charge capacity being produced by the potentiostatic charging step.   Figure 3.16 - Differential capacity plots for control (a) and all additives at 5 wt. % (BaSO4 (b), Ca(OH)2 (c), Sr(OH)2·8H2O (d), Bi2O3 (e)) cycles 1, 3, and 5 with absolute value of differential capacity discharge/charge peak height vs. cycle number (inset) and cycle 1 differential capacity for control and all additives (f). 3.1.6 Coulombic and Energy Efficiency Figure 3.17 shows the coulombic and energy efficiency of EMD/Zn cells with a number of additives cycled with a C/10 rate. The coulombic efficiency (CE) of the nth cycle is calculated by: Coulombic Efficiencyn =  CapacitynDischargeCapacityn−1Charge  ×  100%,          n > 1   [Eq. 3.4]   58  The CE calculated this way is about 94% on average. This value underestimates somewhat the true coulombic efficiency, because it is difficult to correctly account for the OER contribution of the charge regimes. A more accurate way of determining the coulombic efficiency is use the capacity loss during cycling data. The specific capacity (for C/10 rate) decreases to about 60% of its original value after 30 cycles. Using the following expression,  Capacityn =  CapacityinitialCE      [Eq. 3.5] we then obtain a value of CE = 98% which is the true measure of the coulombic efficiency. The difference of 3% provides a measure of the parasitic reaction (particularly OER) that occurs during the charge step.  The energy efficiency for the nth cycle is calculated by integrating the area under the Voltage – Capacity (c) curves: Energy Efficiencyn =  ∫ VoltagenDischarge dcn∫ Voltagen−1Charge dcn−1 ×  100%,    n > 1    [Eq. 3.6] The energy efficiency of the cells is on average about 75%. The losses in the efficiency are a result of rather large polarizations between discharge and charge step and can be reduced with the use of lower discharge/charge rates.    59   Figure 3.17 - Coulombic (C.E.) and energy efficiency (E.E.) vs. cycle number for all additives at 5 wt. % with thin film foil CC electrodes. 3.1.7 Cyclic Voltammetry  Cyclic voltammetry has been used as a method of investigating the EMD-based cathode in previous works45 to identify unique redox processes occurring which may not be evident by a traditional  galvanostatic discharge method as described in section 2.1. Figure 3.18 shows a cyclic voltammagram (blue) and an overlaid differential capacity plot (red).  The blue curve shows a 0.05 mV/s cyclic voltammagram which starts at the open circuit potential of the EMD-based half-cell (~0.1 V vs. Hg/HgO) with the voltage profile indicated in the inset plot.  Red and blue arrows indicate the temporal direction of the experiments, both of which start at the open circuit potential.   60   Figure 3.18 – 0.05 mv/s CV of EMD-based cathode in half-cell setup (blue) with voltage vs. time (inset) and differential capacity (red) showing unique redox processes. The total time required to perform a CV at 0.05 mv/s between the potentials indicated on Figure 3.18 is approximately 17 hours.  This slow scan rate allows for the sensitive detection of electrochemical processes occurring, much like differential capacity plots.  In this case, the fresh EMD electrode was charged by linearly increasing the voltage from OCV up until 0.4 V vs. Hg/HgO.  As the voltage was linearly decreased, a peak appears at 0.2 V vs. Hg/HgO which is above the OCV.  This peak is not evident in previous works, and could the result of charging the fresh electrode which may remove protons and/or structural water present in the cathode materials before discharge.  Otherwise, all the peaks present in the differential capacity plot are also present in the cyclic voltammagram and are discussed in section 3.1.2.   61  3.2 Electrode Thickness Effects 3.2.1 Scalability and Mass Production Potential  Two types of cathodes have been developed for use in flat-plate RAM cells.  These have been previously described in the Experimental Methods chapter with respect to electrode dimensions, production, and loading.    Figure 3.19 – Top view (left) showing electrode surface and side view (right) showing electrode width after processing.  The key differences between electrodes employing the thin film foil CC and the expanded metal mesh CC are the geometric loading density, mass production potential, and adhesion of active material.  The thin film foil CC has one fifth the geometric loading density (approx. 0.0125 gEMD/cm2) compared to the expanded metal mesh CC (approx. 0.0626 g EMD/cm2).  This has significant implications towards determining the energy density of the cell.  This parameter, however, could be optimized by further increasing the loading density of the thin film electrode, providing an estimated geometric loading density of 0.04 gEMD/cm2.  In addition to further increasing the loading density of the cell, a thinner CC could be   62  used.  Currently, the thin film foil CCs have a geometric density of 0.0253 g/cm2 compared to 0.037 g/cm2 for the expanded metal mesh. Table 3.2: Comparison of key parameters for thin film foil and expanded metal mesh CCs. Parameter Thin Film Foil Current Collector Expanded Metal Mesh Current Collector Geometric Loading Density (gEMD/cm2) 0.0125 0.0625 Current Collector Geometric Mass (g/cm2) 0.0253 0.037 Active Mass % (of total electrode mass) 30% 60% Electrode Thickness (mm) 0.15 0.6 Active Material Mass (gEMD) 0.1 0.5 Wet Consistency (before application) Ink Paste Dimensions (width and length) 2 cm x 4 cm 2 cm x 4 cm Binder Composition No SBR Binder SBR Binder Mass Production Potential Easy Difficult Electrode Flexibility Can bend/roll Cannot bend/roll   Another important factor is the ratio of mass of active material to mass of CC.  Currently, the thin film CC active material comprises roughly 30% of total electrode mass while the expanded metal mesh active material CC comprises roughly 60% of total electrode mass.  These values will differ slightly between electrode batches; however, it is clear that the current manufacturing method of thin film CCs needs to be optimized to improve energy density.  A significant improvement could be made by increasing the loading density of the thin film electrodes and by decreasing the thickness of the CC, as well as coating both sides of the thin film foil CC.      63  While the thin film foil CC is currently not optimized for energy density, it has other significant advantages over the expanded metal mesh CC.  The method of preparation of thin film electrodes involves preparation of an EMD-based ink which is sonicated using a probe tip sonicator.  Comparatively, the mesh CC is mixed by hand and is less likely to provide a homogeneous cathode mix.  This process could be automated and made more reproducible by employing a planetary type mixer and is discussed in section 4.1.5.  The choice of current collector has proven to be an important factor to investigate regarding parameters such as energy density, geometric loading density, mechanical stability, and preparatory considerations.  Another important factor is the average distance between any given particle of EMD and the current collector.  Reducing this distance can improve the mass transfer characteristics by minimizing diffusion distances, and therefore achieve higher current densities.  Another type of current collector, a 3-dimensional nickel foam CC, which has not been investigated, and will most likely be promising in this regard.  This type of electrode would significantly improve the surface area of the current collector.  Figure 3.20 shows optical micrographs of the three types of current collectors at 4 and 10 x magnifications.   64   Figure 3.20 – Optical micrographs of nickel foam (left), nickel foil (centre), and expanded nickel mesh (right) at 4x (top) and 10x (bottom) magnification.  3.2.2  Effect of Current Collector on Identical Cathode Mix  The CC employed has important considerations towards cell performance as well as production potential.  Figure 3.21 plots the % initial capacity (i.e., % of first discharge capacity) vs. cycle number for identical cathode mixes using thin film foil and expanded metal mesh CCs.    65   Figure 3.21 – Normalized (left) and specific capacity (right) for control electrodes cycled at C/10 discharge rate to 1.1 V using the expanded metal mesh and thin film foil CCs.  The thin film electrode fades in much more linear fashion compared to the expanded metal mesh CC which appears to fade at an exponential rate with a significant loss of capacity in the first 10 to 15 cycles.  Although the rate of fade is significantly improved using the thin film foil CC, the initial capacity is also approximately 100 mAh/gEMD lower than when using the expanded metal mesh CC.  The linearly projected capacity in Figure 3.21 may be misleading as cells using the expanded metal mesh CC typically show the most prominent capacity fade over the first 10 to 15 cycles.  Regardless, the rate of capacity fade is faster for the expanded metal mesh CC over the first 25 cycles, which is evident by both the normalized and specific capacity plots in Figure 3.21.  This difference in performance provides further motivation to optimize the preparation of thin film electrodes in order to improve long term cycling performance as well as initial capacity.  Recommendations for optimizing the fabrication of cathode electrodes using the thin film foil CC will be discussed in section 4.1.  This difference in performance could be a result of one or many factors including homogeneity of the electrode mixture and thus active material usage, binder content, and/or electrode contraction/expansion during cycling.  In addition, the electrode thickness may also influence active material usage.    66  Another property of the cathode material is that it undergoes expansion during discharge and contraction during charge.29,34,46,47  The repeated expansion and contraction of the cathode may influence the electrode structure over time, potentially reducing EMD to graphite connectivity.  Electrode expansion is a function of active material volume, ranging up to 18% during the discharge process25.  The thin film CC active material layer is thinner than the expanded metal mesh CC, and therefore will expand less relative to total electrode thickness.  In addition, the ink used to prepare the thin film foil electrodes is sonicated which may also produce smaller graphite and/or EMD particle sizes.  All of these factors may play a role in the performance difference between the thin film foil and expanded metal mesh CCs. 3.2.3 Current Collector Discharge Profile Comparison Figure 3.22 is a voltage vs. capacity plot showing the first cycle of discharge and charge for identical cathode mix powders (although differ with respect to the use of SBR binder. See Table 3.2) using the thin film foil and expanded metal mesh CCs.  The thin film foil CC electrode does not contain SBR binder.  The discharge profiles are the lower portion of the curve down to the voltage cut-off of 1.1 V, while the upper portions refer to the charge profile.    67   Figure 3.22 – Voltage vs. capacity profiles for thin film foil CC and expanded metal mesh CC cathodes.  Cathode mix powder identical in each case (80% EMD, 15% KS44 graphite, 5% BaSO4).  Figure 3.22 shows a key difference between the thin film and expanded metal mesh CCs.  The area inside the closed loop voltage capacity profile is indicative of the efficiency of the cell.  For example, a perfectly efficient cell charge/discharge would have the charge profile overlap directly with the discharge profile.  The thin film electrode has a thinner closed loop profile compared to the mesh electrode as measured by the difference in potential between charge and discharge profiles, although provides lower capacity.  It is currently unclear why the thin film electrodes provide lower specific capacity (approximately 100 mAh/gEMD lower), and this should be further investigated with respect to the effect of binder and electrode preparation.  Another metric of assessing the performance between the thin film and mesh CCs is midpoint voltage.   Midpoint voltage is defined as the voltage at half of total discharge capacity.  For example, if a cell   68  discharges to a capacity of 200 mAh/gEMD then the midpoint voltage is the voltage corresponding to a capacity of 100 mAh/gEMD on the voltage capacity profile.  This can be plotted vs. cycle number as in Figure 3.23.    Figure 3.23 – Midpoint Voltage vs. Cycle Number for thin film foil CC cells expanded metal mesh CC cells.   Figure 3.23 presents the midpoint voltage of two cells employing each type of cathode CC.  The midpoint voltage for the mesh CC is on average approximately 30 to 40 mV lower than that of the thin film foil CC.  This result of this difference is that the thin film foil CC will provide a slightly higher working voltage, and therefore provide a slight boost in power under similar conditions.   69  3.3 Effect of Cathode Additives 3.3.1 Conductive Additives (KS44 graphite, Super C65 Carbon Black)  The cathode is comprised of primarily manganese dioxide.  However, 15% graphite by weight is added to improve electrical conductivity.  The properties of the graphite additives can influence the performance of the electrodes in a number of ways.  Electrode porosity is an important parameter as it will influence the rate of electrolyte penetration.48  In addition, the electrode porosity will change as a function of % reduction of EMD.   This lattice expansion can increase the unit cell volume by as much as 18% as reduction of EMD proceeds from MnO1.97 to MnO1.58.25,49 Electrode porosity thus changes as different types of graphite additives are introduced into the cathode mix.  Two different carbon additives have been tested for their performance in flat-plate RAM cells, Super C65 carbon black and KS44 graphite.  These additives were chosen to assess the performance of two carbon additives with significantly different properties, which are summarized in Table 3.3. Table 3.3: Important parameters of Super C65 carbon black and KS44 graphite. Property Super C65 Carbon Black50 KS44 Graphite51 Scott Density (g/cm3) 0.16 0.19 Moisture Content (%) 0.10 0.10 Ash (%) 0.025 0.060 Particle Size 150 nm (aggregate)52 18.6 μm (D50), 45.4 μm (D90) BET Nitrogen Surface Area (m2/g) 62.0 9.0m   Super C65 carbon black and KS44 graphite differ from each other primarily with regards to their particle size, crystallinity, and surface area.  In particular, Super C65 carbon black has a significantly smaller aggregate particle size and a larger surface area.  In addition, the powder X-ray diffraction pattern of both additives presented in Figure 3.24 shows significant differences in crystallinity between these additives.      70   Figure 3.24 – Powder X-Ray diffraction pattern of SuperC65 carbon black and KS44 graphite.  The peaks at angles higher than 40o 2θ present in the KS44 graphite pattern are indicative of a higher degree of crystallinity and their absence is also indicative of the amorphous nature of the Super C65 carbon black additive.  The powder XRD patterns of these additives showing broader peaks for the Super C65 carbon black additive are in agreement with the properties presented in Table 3.3 indicating a smaller particle size and confirming the amorphous nature of the Super C65 carbon black additive.  The performance of the different carbon additives might be explained in terms of a model of graphite-EMD connectivity.6  In order for EMD to discharge effectively, each EMD particle requires contact with   71  electrolyte and graphite, in a so-called three-phase boundary.6  This three-phase boundary requires an electronic and ionic path to EMD allowing proton/electron insertion into the lattice during discharge.   The ratio of carbon black or graphite to EMD particle size may be indicative of the nature of this triple-phase boundary.  For example, if the graphite particle size is significantly larger than the EMD particle size, it may provide good ionic contact due to lower density packing of graphite and EMD.  Conversely, if the graphite particle size is significantly smaller than the EMD particle, electrolyte penetration may be limited.  In addition, contact between graphite and EMD particles is equally important to provide an electron path to the CC. Figure 3.25 illustrates how the ratio of graphite to EMD particle size may influence electronic and ionic conductivity.    Figure 3.25 – 3 idealized representations of EMD-graphite connectivity. Large graphite particle size (left), intermediate graphite particle size (centre), and small graphite particle size (right). Central spherical particle represents EMD, with graphite particles surrounding it.  These idealized cases show the trend in electrolyte penetration and electronic contact, both of which are required for homogeneous reduction of EMD.  If graphite particles are too tightly packed around EMD particles, poor electrolyte penetration limits the reduction of EMD, while if graphite particles are too large, poor electronic contact limits the reduction of EMD.  Applying this model to the discharge process, the   72  effective ratio of EMD particle size to graphite particle size may be a significant factor influencing utilisation of the active material.   Cathode electrodes were prepared in order to test the performance of both of these carbon additives and their effect on cycling performance.  Three compositions were tested in order to ascertain the effect of each individual carbon additive as well as a mixture of both.     Figure 3.26 – Voltage vs. capacity profiles of 15% KS44 graphite (left), 15% Super C65 carbon black (right), and 7.5% KS44 Graphite/7.5% Super C65 Carbon black (bottom).  The voltage vs. capacity profiles of three compositions of carbon additives in Figure 3.26 show a significant difference in performance between each composition.  15% KS44 graphite provided the best cycling performance as indicated in Figure 3.27.  15% Super C65 carbon black provided a higher initial capacity yet poor cycling performance.  The combination of graphite and carbon black, however, performed poorly   73  with cycling as compared to either graphite or carbon black on their own.  The results presented in Figures 3.26 and 3.27 was performed with two replicates of each cathode composition both from the same cathode mix batch.  More replicates may be necessary in order to confirm this result.   Figure 3.27 - Capacity vs. cycle of electrodes prepared with 15% KS44 graphite, 15% Super C65 carbon black and 7.5% of both KS44 graphite and Super C65 carbon black using electrodes prepared with expanded metal mesh CCs.  The performance of different carbon additives is a complex phenomenon which needs to be understood in the context of the physical and chemical environment.  In both cases when Super C65 carbon black was used, a higher initial capacity was achieved yet the cycling performance was poor.  In the context of improving electrical conductivity of an electrode material which expands (during discharge) and contracts (during charge), it is possible that smaller graphite particles may be displaced from the surface of EMD.  This hypothesis would explain a higher initial capacity, as well as poor cycling performance using a carbon black additive.       74  The poor performance obtained by using both carbon additives is not well understood, but may be a result of a physical interaction between larger graphite sheets (and/or EMD particles) and smaller carbon black particles.   Carbon black particles may be more likely to be displaced as a result of the expansion/contraction process during cycling, reducing electronic contact with the active material.  In addition, the carbon black particles may block the surface of EMD particles reducing electrolyte penetration, effectively isolating the EMD particles and reducing capacity as described in Figure 3.25 (right).  Further investigations into different graphite and carbon black additives may lead to a better understanding of the role which graphite plays in the cathode electrode.  In light of these limited results, KS44 graphite was subsequently used for electrode preparation.  To further investigate the effect of different graphite content, electrodes with three different compositions of KS44 graphite were prepared.  These electrodes were cycled in a full-cell setup in order to ascertain the effect of increased graphite content as well as cut-off voltage.     75  Figure 3.28 – Capacity vs. cycle number for addition of 15, 20, and 25% KS44 graphite at 1.2 and 0.9 V cut-off potentials.  Figure 3.28 can be divided into two bands representing the 0.9 V (higher band) and 1.2 V (lower band) cut-off potentials.   The gap between these two bands represents the capacity available between 1.2 and 0.9 V during discharge.  Using a lower cut-off voltage provides a significant boost (20-50% increase) in capacity as these cells will be drawing current for a longer period of time until the lower cut-off potential is reached.  By varying the graphite content between 15 and 25%, no significant change in capacity fade was obtained. 3.3.2 Effect of Cathode Additives on Cycling Performance Table 3.4 describes the cathode mix compositions along with a rationale for testing each of the specific additives.  All additives were tested at 5 wt. % composition and compared to a control which did not include any additives but kept the ratio of EMD to graphite constant.  These additives were chosen to compare a wide variety of compounds under the same discharge conditions.  In particular, alkaline earth metal based compounds (Ca, Sr, Ba) were chosen for our investigation because it has been reported that they have positive effects on the cycling performance of EMD.  Table 3.4: Published cathode compositions and rationale for testing. Cathode Mix Composition Year Publication Rationale Control (84% EMD, 16% KS44 Graphite) N/A N/A Used as a control with which to compare performance with other additives. EMD:Graphite ratio is kept constant across all tests. Baseline (80% EMD, 15% KS44 Graphite, 5% BaSO4) 2006 8 Stani et al. used 5% BaSO4 which showed minimal  improvement of cumulative capacity over 25 cycles 80% EMD, 15% KS44 Graphite, 5% Sr(OH)2·8H2O 2009 5 J Daniel-Ivad suggested that alkaline earth metals help to stabilize the reversible manganese phases present in EMD. 80% EMD, 15% KS44 Graphite, 2012, 2014 29,47 Barium Hydroxide was shown to suppress dissolution of Mn3+ into electrolyte.  It is also used   76  5% Ba(OH)2·8H2O as a point of comparison for Barium vs. Barium Sulfate 80% EMD, 15% KS44 Graphite, 5% Bi2O3 1995, 2003, 2008, 2009 5,53–55 Various Bi-based compounds (Bi2O3, BaBiO3) have been investigated and show an improvement in cycling performance compared to EMD. 80% EMD, 15% KS44 Graphite, 5% Ca(OH)2 1994, 2009 5,46 Ca(OH)2 has been used to precipitate Zinc discharge products.  Calcium compounds are also expected to stabilize the structure of electrochemically reversible manganese phases.  Since the electrolyte is 9M potassium hydroxide, the environment is already basic and therefore if these compounds are soluble the effect of 5 wt. % alkaline earth metal hydroxides on electrolyte alkalinity should be negligible. Thus it is our assumption that the effect of these additives will be a result of the cation present (i.e., calcium, strontium, barium etc.)   Table 3.5 details the solubility of each of the additives in water at 20oC.    Table 3.5: Cathode additive chemical formula, molar mass, and solubility in water at 20oC.56 Compound Name Chemical Formula Molar Mass (g/mol) Solubility in water at 20o C (g/100mL) Barium Sulfate BaSO4 233.43 0.0002448 Barium Hydroxide Octahydrate Ba(OH)2.8H2O 315.46 3.89 Strontium Hydroxide Octahydrate Sr(OH)2.8H2O 265.76 1.77 Calcium Hydroxide Ca(OH)2 74.093 0.173 Bismuth (III) Oxide Bi2O3 465.96 Insoluble  It can be seen from Table 3.5 that the solubility of the cathode additives tested vary widely from insoluble (Bi2O3) to almost completely soluble (Ba(OH)2.8H2O).     It should be noted that the solubility of the additives presented in Table 3.5 are in water, and not in 9M potassium hydroxide so the actual solubility can be expected to be lower (especially for alkaline earth metal hydroxides such as Ba(OH)2.8H2O which would produce hydroxide ions if soluble).  In addition, 5 wt. % of these additives was tested which would   77  mean that the stoichiometric ratio of additive to EMD would vary depending on the additive used.  This amount was used in order to maintain the EMD to graphite ratio constant between each additive tested.  The performances of all additives were compared against a control.  The control cathode composition was chosen in order to keep the same ratio of graphite to EMD as in all the other electrode compositions.    Figure 3.29 – Normalized (left) and Specific Capacity (right) for control cathode composition (84% EMD, 16% KS44 graphite) with expanded metal mesh and foil CCs.  Figure 3.29 shows the normalized and specific capacity vs cycle for control cathode compositions with both types of CCs.  Electrodes prepared with the expanded metal mesh CC provide a significant increase in specific capacity over the foil CC, however when comparing normalized capacity it becomes evident that the rate of fade is also more prominent.  In addition, capacity of electrodes prepared with the thin film foil CC appears to fade linearly, while the capacity of electrodes prepared with the expanded metal mesh CC fades more drastically over the first 15 cycles before fade becomes linear.     78   Figure 3.30 – Normalized (left) and specific (right) capacity vs. cycle number for baseline cathode composition (80% EMD, 15% KS44 graphite, 5% BaSO4) compared to control cathode composition (84% EMD, 16% graphite).  The effect of Barium Sulfate has been previously shown to improve cumulative capacity fade compared to a control in a flat-plate configuration by Stani et al8 and is in agreement with this data.  Although the control cathode composition provides higher initial capacity, it also fades more quickly which is also evident from the normalized capacity.  The performance of a number of cells averaged shows the statistical significance of Barium Sulfate as an additive compared to control in flat-plate rechargeable alkaline cells.  The use of alkaline earth metal additives such as strontium hydroxide octahydrate was investigated using both the thin film foil and expanded metal mesh CCs.   79   Figure 3.31 – Normalized (left) and specific (right) capacity vs. cycle for strontium hydroxide octahydrate cathode composition (80% EMD, 15% KS44 graphite, 5% Sr(OH)2·8H2O). Thin film foil vs. expanded metal mesh CC for C/2 and C/10 rates of discharge to 1.1V.  Similarly to other electrode compositions a C/10 rate of discharge provided the best performance.   The rate of capacity fade appears to stabilize after 20 cycles at a C/2 rate.  Compared to a control cathode composition, strontium hydroxide octahydrate provides a similar rate of capacity fade and higher specific capacity; however more data is needed to confirm this result.  Bismuth (III) oxide was tested as a cathode additive showing similar results as other additives at 5 wt. % and is presented in Figure 3.32.    80   Figure 3.32 –Normalized and specific capacity vs. Cycle for Bismuth (III) oxide additive (80% EMD, 15% KS44 graphite, and 5% Bi2O3). Thin film foil vs. Mesh CC with C/2 and C/10 rate of discharge to 1.1V.  Similarly, a slower C/10 rate of discharge improved performance and the stabilization of fade occurred at lower capacities with a C/2 rate of discharge.  In addition to the improved normalized and specific capacity, bismuth (III) oxide also shows quite repeatable performance as indicated by the width of the error bars in Figure 3.32.  Bismuth (III) oxide was chosen as an additive as a result of previous work indicating that bismuth modified manganese dioxide shows significantly improved cycling performance and different electrochemical properties compared to bismuth free electrodes.54   These previous results were obtained in a 3-electrode system with 45% graphite content, and thus may not be obtained at 15% graphite loading as presented in Figure 3.32.       81    Figure 3.33 –Normalized (left) and specific capacity vs. cycle number for calcium hydroxide cathode additive (80% EMD, 15% KS44 graphite, 5% Ca(OH)2). Thin film foil and expanded metal mesh CCs with C/2 rate of discharge to 1.1V.  Calcium hydroxide electrodes were prepared and tested in a full cell setup at 5 wt. %.  The results are similar to other additives which show a stabilization of capacity fade at approximately 40 mAh/gEMD.  One significant difference between calcium hydroxide and the other additives tested is that the gelling agent in the cathode material (1.7 wt. % CMC-carboxymethylcellulose) forms a precipitate with calcium hydroxide during electrode preparation.  Therefore, the inclusion of calcium hydroxide using CMC was not possible, and a different gelling agent was used called Carbopol 940.57  Carbopol 940 is a cross-linked polyacrylate polymer with high viscosity, and is also used in the zinc/zinc oxide anode electrodes.  The effect that gelling agent may have on electrode performance was not investigated, and thus testing other additives with the Carbopol 940 gelling agent should be done to verify the performance of calcium hydroxide as a cathode additive.    From a practical standpoint, calcium hydroxide may be an important cathode additive as it has been shown to precipitate soluble zinc species present in the anode.58  While this may seem unimportant as a cathode additive, throughout long term cycling soluble zinc species can migrate through and/or past the   82  separator to form zinc species in the cathode such as haeterolyte (Zn2MnO4).  This haeterolyte phase has been previously identified in MnO2/Zn cells as an irreversibly formed phase, and one cause of irreversible capacity fade.59  If these zinc ions are precipitated selectively by calcium hydroxide, these species may not form and thus the active cathode material may provide a longer cycle life.  The implications of this will be discussed further in section 3.3.4.  Figure 3.34 (a-f) shows the cycling performance of EMD/Zn for a number of additives tested at 5 wt% with thin film electrodes. Data presented show the i) discharge capacity, ii) total charge capacity and also the breakdown of the charge capacity regimes, i.e., the corresponding galvanostatic and potentiostatic charge steps.  Figure 3.34 – Thin film electrode full-cell capacities from galvanostatic discharge, galvanostatic charge, potentiostatic charge and total charge capacity vs. cycle number for control (a), 5 wt. % BaSO4 (b), 5 wt. % Ca(OH)2 (c), 5 wt. % Sr(OH)2·8H2O (d), 5 wt. % Bi2O3 (f), and specific discharge capacity vs. cycle number for all additives (f).   83   The overall charge capacity is the addition of both charging steps and appears to be slightly above the discharge capacities. All these experiments were performed with the thin film-type electrode (Ni foil CC). The discharge capacities of all the additives are compared in Figure 3.34 f).  It is evident from the data that both galvanostatic discharge and charge capacities over 50 cycles decrease during cycling for all the additives with a similar rate of  between 1.2 and 2 mAh g-1 cycle-1. The potentiostatic charge also decreases during the cycling, however, at a slower rate. It should be mentioned that decoupling of the OER contribution in general is not quite straightforward; hence, could result in an overestimation of the capacity delivered during the potentiostatic step. Figure 3.34 (f) compares the discharge capacities of the various additives for the first 50 cycles. No significant difference among these various additives can be observed, although the cathode with Ca(OH)2 additive appears to lose capacity faster than other additives. Ca(OH)2 electrodes were prepared with the Carbopol gelling agent instead of the typical cathode gelling agent carboxymethylcellulose (CMC) because Ca(OH)2 agglomerates with CMC and it is therefore not possible to disperse into the cathode paste during preparation.  It should be noted that all other additives were prepared with CMC and the poor performance of Ca(OH)2 could be a result of the different gelling agent.    Figure 3.35 (a-f) shows the breakdown of i) discharge energy, ii) total charge energy, and the breakdown of the charge energy for both the galvanostatic and potentiostatic charging steps from the cells presented in Figure 3.34.  The difference between discharge energy and total charge energy is significant, with the discharge energy accounting for approximately 75% of the total charge energy.  The energy efficiency (calculated with Eq. 3.6) of these cells is discussed in section 3.1.6, and is always lower than the coulombic efficiency.  As seen in Figure 2.4, during the galvanostatic charge step the potential quickly rises from the cut-off potential of the galvanostatic discharge step before a significant amount of charge is put into the cell.  This means that while the galvanostatic charging step is using the same current as the galvanostatic   84  discharge step, it is doing so at higher potentials and thus delivers more energy overall than if this step traced the discharge step voltage profile exactly in reverse.  This is a result of the ohmic polarization potential losses which are discussed in section 3.1.4.  Figure 3.35 – Thin film electrode full-cell energies from galvanostatic discharge, galvanostatic charge, potentiostatic charge and total charge energy vs. cycle number for control (a), 5 wt. % BaSO4 (b), 5 wt. % Ca(OH)2 (c), 5 wt. % Sr(OH)2·8H2O (d), 5 wt. % Bi2O3 (f), and specific discharge energy vs. cycle number for all additives (f). 3.3.3 Scanning Electron Microscopy (SEM) and Energy Dispersive X-Ray Spectroscopy (EDS)  Figure 3.36 details the surface composition of a cycled electrode with 5 wt. % BaSO4 additive.  The backscattered electron micrograph details the elemental density on the surface of the electrode.  Lighter portions of the image refer to elements present with higher atomic number.  By mapping the individual elements present in the electrode and comparing the oxygen and manganese EDS maps, it can be seen that a large manganese dioxide particle (approximately 80 μm) is present in the bottom right corner of   85  the image.  Smaller manganese dioxide particles are present all over the surface of the electrode, thus indicating a wider range in particle sizes than expected.  This may indicate that that the electrode mix needs to be mixed further in order to get better utilization of the active material.  The large black feature along the right side of each image corresponds to a crack in the surface of the electrode.  Cracks such as these may form more readily along the boundaries of larger particles such as the manganese dioxide particle in this image.    Figure 3.36 – Expanded metal mesh CC electrode surface analysis (5 wt. % BaSO4).  Backscattered electron micrograph (a) with energy dispersive x-ray spectroscopy mapping of potassium (b), carbon (c), barium (d), manganese (e) and oxygen (f).  Scale bar representing 80 μm.   The potassium concentration appears to be relatively constant across the electrode surface, yet shows a possible increase on the manganese dioxide particle.  Carbon concentration across the electrode surface is more homogeneous, with a few locations showing high a concentration of carbon due to the presence of 15% graphite in the electrode mix.  Barium is minimally present (i.e., in 5 wt. %) and is present in a few   86  locations in high concentration, which can also be seen as very bright spots in the backscattered electron micrograph.  Figure 3.37 – Secondary electron micrograph of an expanded metal mesh CC electrode cross section.  Image taken at the UBC CHTP.  Figure 3.37 presents a cross section of an expanded metal CC electrode at low magnification.  This cross section was prepared by freezing the electrode in liquid N2 for one minute, and pulling it apart with pliers by hand in order to ensure the electrode was not pinched or squeezed along the cross section, giving a more realistic indication of electrode composition.  In order to estimate the range of EMD particle size, the cathode electrode cross section was imaged with SE and BSE detectors at various magnifications.  Figure 3.38 a-b shows the low range  (148 x) magnification SE and BSE images noting a large EMD particle approximately 100 μm in diameter, and Figure 3.38 c-f   87  shows  the mid (2500 x) and higher (9995 x) range SE and BSE images indicating smaller size EMD particles between 0.1 and 10 μm.   Figure 3.38 – EMD particle size determination by SE (a, c, e) and BSE (b, d, f) imaging.  EMD particles outlined in solid red line (a-f), and graphite particle outlined in dashed blue line (c-f).  The particle size of EMD has been observed to range between 0.1 and 100 μm, thus indicating a wide range of particle sizes.  The two small EMD particles (circled in c-f) are representative of many others in both the mid and higher range SE and BSE images, and can be seen in the BSE image d) as small white particles with distinct elemental contrast against larger graphite flakes.  The EMD to graphite connectivity can also be seen, which has been argued to be a significant factor in cell performance.   88  Figure 3.39 – Surface of 5 wt. % Sr(OH)2·8H2O electrode imaged in BSE mode (a) and EDS mapping for elemental Sr (b), Mn (c), and overlaid Sr and Mn map (d).  The electrode surface EDS mapping in Figure 3.39 shows the distribution of Sr(OH)2·8H2O additive.  It is evident that the additive is not homogeneously distributed by ball milling of the electrode mix.  The effect of additives is still unclear, however considering the range of EMD particle sizes (0.1 – 100 μm) and the distribution of additive shown in Figure 3.39, it is proposed that the effect these additives have may not be a bulk effect.   In order to determine if potassium from the electrolyte is intercalating into the EMD lattice during cycling as suggested in literature60, EDS spot analysis was performed at four points along the cross section of an electrode.  The spots chosen correspond to flat areas of the electrode to ensure that electrode   89  morphology would not impact the quantification of potassium.  The electrode surface was rinsed with deionized water and dried in air prior to analysis in order to determine if potassium is being intercalated or deposited.  These results are presented in Figure 3.40.  Figure 3.40 – EDS potassium spot analysis through electrode cross section.  The results of the EDS spot analysis in Figure 3.40 indicate no significant potassium gradient is present through the cross section of the electrode.  The low Mn:K atomic ratio present at EDS Spot 3 is likely a result of EDS Spot 3 being located on a graphite particle.  The presence of potassium may indicate intercalation is occurring even though no discernable gradient was observed.  It is not conclusive as to whether potassium is being intercalated, deposited, or perhaps both.    This will need to be investigated further and in conjunction with testing other alkaline electrolytes (e.g., NaOH, LiOH) and is discussed further in section 4.1.7.      90  3.3.4 Powder X-Ray Diffraction (XRD) Analysis of the Cathode Electrode   Figure 3.41 – a) Powder x-ray diffraction patterns of a control cathode  (84% EMD, 16% Graphite) using expanded metal mesh CC cell cycled to a charged and discharged state; b) Magnified and overlaid pattern showing peak shifting.  Figure 3.41 b) shows the peak shifting occurring for control electrodes discharged and charged in a single and 20 cycles.  In the discharged state, the peaks corresponding to EMD shift to lower angles indicating a lattice expansion.  This is in agreement with previous literature which states that the homogeneous reduction of EMD occurs with proton insertion into the EMD lattice.44  Using the DIFFRAC.EVA software “tune cell” feature,  the lattice constants can be changed (from a = b = 2.786 Å, c = 4.412 Å to a = b = 2.854 Å, c = 4.557 Å) to correspond with the peaks observed in the powder x-ray diffraction pattern of the cathode discharged for a single cycle.  Although this is a rough approximation of the unit cell structure, it gives an insight into the process occurring during discharge.  A simple calculation determines that this peak shift corresponds to a volume expansion of 8.4%.  The electrode which is discharged and then again charged for a single cycle shows a slight peak shift to lower angles which indicates that the process occurring during charge (proton de-insertion) is not completely reversible.  Scherrer analysis was also   91  performed on these peaks showing no significant change in crystallite size with charge state or number of cycles.      Figure 3.42 – a) Powder x-ray diffraction patterns of a baseline cathode (5 wt. % BaSO4) cell cycled to a charged and discharged state compared to fresh cathode mix; b) Magnified and overlaid pattern showing peak shifting.  Figure 3.42 shows the same trend occurring in baseline cathode electrodes during cycling, with peak shifts occurring during the discharge process.  This peak shift to lower angles occurs only during the first discharge, and is not present in the discharge or charge state after 20 cycles, which indicates an irreversible process and is in agreement with the data presented in Figure 3.7.  In addition, two crystal phases are formed during cycling, witherite (BaCO3) and arcanite (K2SO4).  The experimental results presented in Figure 3.43 indicate that these crystal phases formed during cycling are not a result of an electrochemical phenomenon.   92   Figure 3.43 – Powder  X-ray diffraction pattern of baseline cathode mix (5% BaSO4) in 9M KOH with N2 atmosphere (top), 9M KOH with air (middle), and no electrolyte (bottom) for 30 days.  Based on the previous powder X-ray diffraction patterns in Figure 3.42 showing the formation of witherite (BaCO3), and arcanite (K2SO4) during cycling of baseline cathodes, two samples of baseline cathode mix were placed in 9M potassium hydroxide solution for 30 days, one in a N2 atmosphere and one open to air, in order to determine if these phases are forming through an electrochemical process or not.  The powder x-ray diffraction pattern of the baseline cathode mix (with no 9M KOH electrolyte) shows that it is composed of EMD, barite (BaSO4), and graphite (C) as expected.  The baseline cathode mix in 9M KOH and N2 atmosphere shows the formation of witherite alone, while the baseline cathode in 9M KOH and air shows the formation of both witherite and arcanite.  From these results, it follows that the formation of   93  both of these phases is not an electrochemical process, however, may still play a significant role in the electrochemistry of the EMD based cathode and should be investigated further.  In order to exclude the oxidation of graphite as the cause for carbonate formation, barite was placed in 9M potassium hydroxide electrolyte.  The powder X-ray diffraction pattern of the sample after 30 days indicated the formation of witherite and arcanite, thus indicating carbonate formation is not a result of graphite oxidation in the cathode.  Figure 3.44 – Powder X-ray diffraction pattern of barium sulfate (BaSO4, barite) in 9M KOH.               94  Figure 3.45 shows four powder x-ray diffraction patterns of 5 wt. % BaSO4 electrodes under various test conditions.    Figure 3.45 – Powder x-ray diffraction analysis of baseline cathode electrodes (5% BaSO4) from bottom to top: fresh cathode, discharged half-cell cathode, charged half-cell cathode, charged full cell cathode. The fresh electrode has been identified as expected to contain solely EMD, graphite, and barium sulfate.  Both cycled half-cell electrodes in the discharged and charged states indicate the formation of barium carbonate.  The full cell electrode shows the formation of haeterolyte (Zn2MnO4) which would indicate zinc contamination from the anode.  Previous reports indicate that the formation of haeterolyte is more   95  likely to occur at lower depths of discharge beyond 0.5 protons per manganese formula unit.61  Beyond this, competition between protons and zinc ions occurs resulting in the formation of groutite and haeterolyte.61  The mechanism with which the haeterolyte species forms is not clear; however it likely is produced through soluble zinc species crossing through or around the separator.   It is hypothesized that barium carbonate is formed through the following set of reactions, with the potassium hydroxide electrolyte acting as a carbon dioxide sink.  The formation of K2SO4 and BaCO3 which have been noted in Figures 3.42-3.44 are consistent with this proposed mechanism. 2KOH(aq) + CO2 (g)  K2CO3 (aq) + H2O (l) (step 1-carbon dioxide dissolution) [Eq. 3.7] K2CO3 (aq) + BaSO4 (s)  K2SO4 (aq)  + BaCO3 (s) (step 2-carbonate formation)                [Eq. 3.8]  Figure 3.46 describes a similar type of carbonate formation occurring in electrodes with 5 wt. % strontium hydroxide octahydrate.   96   Figure 3.46 – Powder x-ray diffraction analysis of cathode electrodes prepared with 5 wt. % Sr(OH)2·8H2O. The formation of strontium carbonate may be analogous to the formation of barium carbonate in electrodes comprised of 5 wt. % barium sulfate.  The fresh electrode has been characterized to be composed of strontium hydroxide octahydrate, graphite, and EMD as expected. The cycled electrode however indicates that strontium carbonate has formed during cycling.  This may provide some indication   97  of a trend between alkaline earth metal additives and their role in rechargeable alkaline cell chemistry, and subsequent formation of carbonates.  Figure 3.47 – Powder x-ray diffraction analysis of fresh and cycled electrodes containing 5 wt. % Ca(OH)2 indicating zinc contamination. As previously described, calcium hydroxide has been used as an additive to supress shape change in the zinc anode.46  This additive precipitates zinc discharge products in the anode and forms calcium zincate.62    98  Alternatively, using calcium hydroxide as an additive in the cathode may provide a similar effect.  The cycled electrode shows that calcium zincate (Ca(OH)2Zn(OH)3 ·2H2O) has formed during cycling.  If zinc species which are mobile across the separator are precipitated by an additive in the cathode, this could prevent the formation of haeterolyte (Zn2MnO4) and improve the performance of the EMD cathode.  Rather than being used as an anode additive, in the event that soluble zinc species migrate to the cathode they may be precipitated preferentially by calcium hydroxide.                99  3.4 Analytical Characterization 3.4.1 X-Ray Photoelectron Spectroscopy (XPS) Measurements XPS measurements were performed on a fresh and cycled electrode in order to identify amorphous manganese phases which may not be identified with other techniques such as powder x-ray diffraction.  Samples were sent to the surface science lab Western University for analysis.  Figure 3.48 – XPS measurements for a) fresh and b) cycled electrodes. The fresh electrode has some amount of Mn2O3 (30%) present even before cycled, which could be a surface phenomenon as XPS is sensitive up to a depth of 10 nm and is not likely a representation of the bulk properties prior to discharge.  These results indicate the formation of Mn2O3 during cycling of the   100  electrode which has been reported by several independent sources.23,58,63,64  This manganese oxide phase is formed irreversibly and cannot be efficiently oxidized to MnO2.65   3.4.2 Potentiostatic Electrochemical Impedance Spectroscopy (PEIS) As described in section 2.3.3, PEIS measurements were performed on full-cells in the charged and discharged state at OCV and the bulk and charge transfer resistances were extracted as in Figure 3.49 with the bulk resistance (Rs) from the real impedance axis intercept, and the charge transfer resistance (Rct) from the semi-circle horizontal width.    Figure 3.49 – Typical control recipe full-cell Nyquist plot showing determination of charge transfer (Rct) and bulk resistance (Rs).  Measurements performed on BioLogic VMP3 potentiostat with frequency range 10 mHz to 200 kHz and 10 mV signal amplitude.  The key parameters, bulk (Rs) and charge transfer resistance (Rct) were extracted as indicated in Figure 3.49 and are presented in Figures 3.50 and 3.51.   101   Figure 3.50 – Bulk resistance (Rs) vs. cycle number for baseline (5 wt. % BaSO4) cathode full cell discharged to 1.1 V at a C/2 rate of discharge.  PEIS measurements were taken after galvanostatic discharge and after potentiostatic charge.  It is known that the resistivity of haeterolyte (Zn2MnO4) and hausmannite (Mn3O4) phases are six orders of magnitude higher than EMD61, and thus it seems reasonable that a moderate increase in bulk resistance with cycling could be attributed at least in part due to the formation of one or more of these phases.  In addition, there is a slight increase in Rs (~0.05 Ω) after discharge which is consistent with a decrease in conductivity of the discharge product in the cathode.  Capacity fade in MnO2/Zn cells is progressive, and is in part due to the formation of electrochemically irreversible phases.  The gradual increase in Rs over a number of cycles as well as the discrepancy between the charged and discharged states is in agreement with this hypothesis.      102  The charge transfer resistance (Rct) in the context of the EMD cathode is indicative of the tendency of proton intercalation into the EMD lattice structure during discharge.  Figure 3.51 – Charge transfer resistance (Rct) vs. cycle number for baseline (5 wt. % BaSO4) cathode full cell discharged to 1.1 V at a C/2 rate of discharge.  PEIS measurements were taken after galvanostatic discharge and after potentiostatic charge.  Figure 3.51 shows the charger transfer resistance measured before and after discharge vs. cycle number.  A similar trend as seen in Rs also occurs with Rct, increasing with cycle number.  The charge transfer resistance is a measure of how readily charge is transferred at the interface of electrode and electrolyte. In this case, charge transfer is that of MnO2 reduction and intercalation of the associated proton.  The Rct can be thought of as an energy barrier which must be reached in order for electron/proton pairs to move in and out of the EMD lattice.47  Previous reports have shown that the Rct decreases below a potential of 0.05 V vs Hg/HgO (1.35 V vs. Zn/ZnO) in a half-cell setup during discharge.47  These measurements were taken in the charged and discharged state, and thus the change in Rct as a function of cell potential is not observed.  The increase in Rct observed with long term cycling shows that the charge and discharge processes are being constrained.  We can speculate that an increase in Rct during cycling is a result of less reactive films forming on the EMD grains such as Mn(OH)2.49  The progressive formation of these phases   103  may block the surface of the active material, limiting proton diffusion.  The formation of Mn(OH)2 is indicated Eq. 6 in section 1.2.2.  Figure 3.52 is an annotated Pourbaix diagram showing the range of possible manganese phases present at different potentials and pH values.  Figure 3.52 – Annotated manganese pourbaix diagram showing various discharge potentials.11 The pourbaix diagram in Figure 3.52 shows the various manganese phases present at different potential and pH values.  The red annotation denotes the various cell potentials at which MnO2/Zn cells operate between the charged and discharged state.  The majority of cell capacity is delivered between 1.4 V and 1.1 V vs. Zn/ZnO, which thermodynamically favours the formation of Mn2O3 and MnOOH.  The 1.1 V potential cut-off typically used lies on the edge of forming Mn3O4 (hausmannite) and Mn(OH)3- phases which are irreversibly formed.  The formation of either of these phases would be in agreement with the observed trends in Rs and Rct, and is consistent with the capacity fade observed with cells discharged to lower potential cut-offs as presented in section 3.1.2 (Depth of Discharge).    104  4 Future Work and Recommendations 4.1 Improvements in Cell Fabrication and Assembly 4.1.1 Effect of Binder While this body of work has developed two types flat-plate of cathodes based on two different current collectors, further systematic work is required to elucidate the impact of various components on the performance of these two types of electrodes. In preparation of the cathode ink, a significant amount of water and gelling agent is added to the cathode mix in order to allow for easy application to the thin film foil current collector.  The binder employed is a copolymer suspension of styrene-butadiene (SBR, see appendix B for details) which is soluble in water, and thus is not added to the cathode ink.  In addition, with the thin film foil current collector, mechanical stability is not as necessary as the ink dries and adheres to the current collector.    SBR binder added to the cathode electrode powder provides mechanical stability in electrodes prepared with the expanded metal mesh current collector.  It allows for the electrode mix to be pasted into and adhere to the current collector and maintain its shape.  Although this is required for electrodes prepared with the expanded metal mesh current collectors, it is not needed with the thin film foil current collectors and when added may increase the resistance of the cathode electrode as it is a non-conductive material.   In order to assess the effect of binder on cathode electrode performance, different binder contents should be tested in both the thin film foil and expanded metal mesh current collector electrodes.  By varying the binder content and assessing electrode cycling performance and resistance, an optimized cathode formulation may provide the proper mechanical stability while simultaneously minimizing electrode resistance.  Currently SBR binder comprises 2.5% of dry electrode weight, and examining a range of binder contents between 1 and 10% is recommended.   105  4.1.2 Effect of Gelling Agent The gelling agent used in the cathode electrode (1.7 wt. % carboxymethylcellulose, CMC) is commonly used in many applications as a viscosity modifier or thickening agent.  Its main purpose is to modify the electrode paste consistency so that it can be applied to the current collector more effectively.  However, when mixed with calcium hydroxide (Ca(OH)2) it forms a precipitate which prevents calcium hydroxide from mixing properly and limits the possibility of its use as a cathode additive if the CMC gelling agent is used. In order to test this additive, a different gelling agent (Carbopol 940) was used in the cathode paste.  The results of the calcium hydroxide additive at 5 wt. % showed a negative effect on cycling performance (as measured by capacity per cycle) and it is not clear if this effect is a result of the calcium hydroxide additive or a result of the gelling agent used.  To identify whether this effect is a result of the gelling agent, it is recommended that control electrodes be prepared with the Carbopol 940 gelling agent and tested to evaluate the impact of the Carbopol 940 gelling agent on the cycling performance. 4.1.3 Effect of Sonication vs. Ball Milling The cathode mixture for both types of electrodes are ball milled prior to the preparation of either the cathode ink or paste, however in both cases the ink and paste are mixed in different ways to provide a homogenous electrode composition.  The cathode paste is mixed by hand, however the cathode ink is mixed with a probe tip sonicator.  Sonication may provide more homogenous mixing of the cathode ink, and may impact the EMD and/or graphite particle and/or crystallite size.  The effect of sonication on the cathode ink has not been optimized and may influence utilisation of the active material as well as EMD-graphite connectivity.     106  In order to characterize the effect of sonication on the cathode ink, it is recommended that the cathode ink is sonicated for various amounts of time and with various power settings, dried, and characterized by powder X-ray diffraction and scherrer analysis.  The impact of sonication parameters (power, time) should be investigated by electrochemical and physical (e.g., SEM imaging) for optimum performance. 4.1.4 Current Collector Active Material Loading Density As described in Table 3.2, the thin film foil and expanded metal mesh current collectors comprise 60 and 30% of total cathode electrode weight, respectively.  Specifically with the thin film foil current collector, this amount is significantly higher than desired.  As the electrode ink is in contact with only one surface of the thin film foil current collector, the thickness of this could be reduced drastically in order to improve energy density.  In order to optimize the energy density of the cathode electrode, it would be useful to define a parameter such as the ratio of current collector to cathode ink film thickness.   The limiting factor for the thickness of the current collector would be the mechanical stability of the electrode.   By increasing the thickness of the cathode ink film, and minimizing the thickness of the nickel foil current collector, a significant increase in energy density could be obtained.  It is recommended to test a range of current collector thickness between 0.02 and 0.2 mm, and a range of cathode ink thickness between 0.05 and 0.2 mm. In addition, other types of current collectors could be investigated such as nickel foam (see Figure 3.20).  This type of current collector should have a similar geometric density loading as the other types of current collectors but is significantly more porous possibly providing better electrical contact than either the thin film foil or expanded metal mesh current collectors.     107  4.1.5 Effect of Mixing/Electrode Homogeneity A significant source of irreproducibility in cell performance has been attributed to inconsistencies in electrode preparation.   Currently, the cathode electrodes are prepared by hand which may influence several key factors during preparation such as electrode mix homogeneity, uniform pressure application, and even possibly contamination. In order to improve the consistency and homogeneity of the electrode paste, it is recommended that a planetary mixer be used to mix the electrode paste.  In addition to providing more consistent mixing, this will also allow for higher throughput of electrode preparation.  A schematic of a prototype planetary mixer currently under construction is shown in Figure 4.1.  Figure 4.1 – 3D rendering of planetary mixer under construction for cathode electrode paste mixing and production. 4.1.6 Automated Electrode Processing Currently, the electrodes used in cells are cut using a standard paper cutter and are not necessarily the exact same size.  The stack pressure calculated (47 psi, 3.3 kg/cm3) assumes a 2 cm x 4 cm electrode   108  dimensions and therefore if the electrodes between cells vary slightly in size they will also be under different amounts of pressure during cycling.  This could be improved by using a stencil type cutter to cut electrodes to uniform dimensions. Precision dimensions of the anode and cathode are very important for providing uniform pressure across the cathode electrode.  As seen in Figure 2.1, the springs, pressure plate, anode, separator are situated above the cathode and provide compression.  In addition, if the anode is not completely covering the area of the cathode, non-uniform pressure would be applied to the cathode and likely result in unpredictable performance. 4.1.7 Electrolyte Preparation, Dispensation, and Composition It is recommended that the electrolyte is prepared and stored in an inert atmosphere (i.e., N2) to prevent the dissolution of carbon dioxide and subsequent carbonate formation as described in section 3.3.5.  Although cells in preparation are open to air, they are sealed with a Viton O-ring ideally limiting the amount of atmospheric CO2 present in the cell.   The anode and cathode half-cell reactions both require alkaline electrolyte to proceed, and therefore the amount of electrolyte used in the cell will be an important factor to control.  The cathode half-cell reaction is as follows: MnO2 + H2O + e-  MnOOH + OH-     [Eq. 4.1] For example, if a typical cathode electrode contains 0.5 g of EMD, then theoretically 0.115 mL of 9M KOH electrolyte is required for full reduction during discharge.  Typical cells are assembled using 20 drops (12 mg/drop on average) of 9M KOH electrolyte and contain 0.5 g EMD.  Assuming a density of 1.38 g/mL, this amount corresponds to 0.17 mL of 9M KOH electrolyte which is approximately 50 % more than required at the cathode.    109  A similar comparison can be made at the anode, which oxidizes Zn in the following reaction: Zn + 2OH-  ZnO + H2O + 2e-      [Eq. 4.2]  In this case, 0.5 g of Zn in the typical anode will require 1.7 mL of 9M KOH electrolyte for full oxidation of Zn, although may not be required as these cells are cathode limited. Typically around 0.5mL of electrolyte is used which is approximately 70 % less than required at the anode.  During cell operation, hydroxide ions are produced at the cathode which may be available for oxidation at the anode.  In this scenario, the electrolyte retention capability of the anode vs. the cathode will determine which, if any, electrode is limited.  This provides a clear motivation to reduce the anode mass so that electrolyte is not limited in the cell.  The amount of electrolyte used in cells will be an important factor to optimize, as limited electrolyte will prevent utilisation of the active material while excess electrolyte will increase gassing.  Other electrolyte parameters can also be optimized such as the electrolyte concentration, and composition.  It is recommended that concentrations of 1M to 9M electrolyte are tested, along with various types of alkaline electrolytes.  There have been literature reports  suggesting that potassium ions (as well as lithium, titanium when used as electrolyte or cathode additives) can intercalate into the tunnel structure of EMD materials.30,60,66  If potassium intercalation occurs, this could also be responsible for capacity failure. Using alkaline and alkaline earth salts or a mixture (e.g., LiOH, NaOH, Ca(OH)2, Sr(OH)2) may have an impact on the cycling performance and is worth investigating further.     110  4.1.8 Anode Performance Very little work has been done towards developing the anode electrode as the cathode electrode has been identified as the source of capacity fade in rechargeable alkaline cells, as described by a number of independent reports.5,8,33  This result has also been confirmed by our results of half-cell cycling of the EMD-based cathode.  The anode recipe (see Appendix A) has not been modified, and is used in excess so that the full-cell performance is attributed to the performance of the cathode.  The full cell reaction is below, with a 2:1 ratio of MnO2 to Zn required. Zn + 2 MnO2 + H2O  ZnO + 2MnOOH  EO = 1.608 V   [Eq. 4.3] From this reaction stoichiometry, it can be deduced that the mass ratio of MnO2 to Zn required in a full cell is approximately 2.65.  Since the cathode is the main cause of capacity fade, it follows that if the anode mass is enough to fully utilise the cathode on the first discharge, subsequent cycles will also be cathode limited. During testing of the thin film current collector, the anode electrode was not modified.  In this case, the average mass of Zn in the anode is approximately 1 g, and the mass of EMD in the cathode is approximately 0.1 g.  The amount of Zn required to fully utilise 0.1 g of EMD is 0.038 g (as a result of a 2 electron oxidation of zinc to zinc oxide), and therefore the amount of zinc used is 26.5 times in excess by mass.  In order to optimize the energy density of the cell, it would be necessary to develop a thin film anode which has a smaller amount of Zn.  Thin film anodes could be developed using a copper foil current collector and an ink of similar consistency to that used in the cathode.  Thin film anode electrodes would have the added benefit of improving energy density, electrode fabrication time, and decreasing the amount of Zn required.    111  4.2 Improvement in Cell Performance through Materials Synthesis, Cathode Additives, and Characterization 4.2.1 Conductive Additives While the results presented in section 3.3.1 (Graphite Additives) preliminarily conclude that the best performance is achieved using 15 wt. % KS44 graphite, there are many aspects of conductive additives which have not yet been explored. The results presented in section 3.3.1 show that while super C65 carbon black as an additive can provide more utilisation of the active material, it does not improve the long-term cycling performance compared to KS44 graphite.  This result can be explained by the relative particle size of EMD (~40 μm), super C65 carbon black (0.15 μm), and KS44 graphite (18.6 μm D50, 45.4 μm D90).  The super C65 carbon black is able to fit better into the spaces between EMD particles and provide better electrical contact initially, however after repeated cycling (and thus expansion and contraction of the electrode) are not able to connect larger EMD particles in a way which the larger KS44 graphite particles are capable of providing.   The combination of equal amounts (7.5 wt. % super C65/KS44) of each additive reduced the performance (see Figures 3.26, 3.27), however, it may be useful to investigate a different combination of these additives (i.e., 12 wt. % KS44 and 3 wt. % super C65).  It is postulated that KS44 graphite can provide a conductive network to connect larger EMD particles while super C65 carbon black can fit in between smaller EMD particles to improve utilisation of the active material.  Ideally a mixture of both large graphitic carbon (KS44) and small carbon black (Super C65) particles for optimum conductivity.  The total amount of carbon and the ratio of these two additives need to be optimized. Carbon fiber is another type of conductive additive which could be used in the cathode.  Carbon fiber can be contrasted to graphite in several ways based on its atomic structure, size, and strength.  Graphite is typically stacked sheets of graphene (which interact weakly via Van der Waals forces) and is quite brittle.    112  Carbon fibers are generally microcrystalline fibers which are oriented to be parallel with the length of the fiber which improves tensile strength compared to graphite.  Carbon fibers are generally longer and could provide larger scale electrical conductivity.  In addition, the improved tensile strength may be less susceptible breaking due to the strain induced during electrode expansion and contraction.   Another factor to consider when choosing a conductive cathode additive is how it may influence the porosity of the electrode, and thus electrolyte penetration.  For example, super C65 carbon black may reduce electrolyte penetration as described in Figure 3.25 (right).  A simple test to roughly estimate electrolyte penetration and retention could be performed by placing electrodes prepared with various conductive additives in electrolyte and removing after set periods of time and comparing the weights before and after.  Each type of electrode would likely reach a saturation point at which no more electrolyte can be absorbed, and perhaps the amount of electrolyte and time required to reach saturation will vary with conductive additive.  In addition, contact angle analysis may provide information regarding the hydrophobicity of the electrode and correlate to electrolyte penetration.  The ideal electrode should hold a large amount of electrolyte and reach saturation quickly. 4.2.2 Cathode Additive Studies All of the cathode additives examined were tested at 5 wt. % in order to maintain a constant ratio of EMD to graphite mass.  It may be more prudent to compare the molar ratio of EMD to additive as each additive will have a different molar mass.  Table 4.1 summarizes the molar and weight percentages of each additive tested.      113  Table 4.1 – EMD to cathode additive molar ratio at 5 wt. % Compound Name Chemical Formula Molar Mass (g/mol) Molar Ratio of Additive to EMD at 5 wt. % Barium Sulfate BaSO4 233.43 0.0234 Barium Hydroxide Octahydrate Ba(OH)2.8H2O 315.46 0.0172 Strontium Hydroxide Octahydrate Sr(OH)2.8H2O 265.76 0.0204 Calcium Hydroxide Ca(OH)2 74.093 0.0733 Bismuth (III) Oxide Bi2O3 465.96 0.0117  If these additives have a stoichiometric effect on cell performance, this should be reflected by varying the amount of cathode additive.  Perhaps not coincidentally, calcium hydroxide which has the largest additive to EMD molar ratio (0.0733), showed a negative result on cycling performance compared to the control.  At the same time, bismuth oxide has the smallest additive to EMD molar ratio (0.0177) and showed the best performance.  This information alone does not indicate an ideal EMD to additive molar ratio, but it is reasonable to suggest that each additive will show different performance based on this ratio and it therefore can be optimized to provide the best obtainable performance.  It is recommended that each additive is tested at additive to EMD molar ratios between 0.01 and 0.1.  Prior to this type of analysis, it is recommended that electrode fabrication process is optimized in order to ensure reproducible cell performance as discussed in section 4.1.  As described in section 3.3.5, when barium sulfate was used as a cathode additive it was common for barium carbonate to form when in contact with potassium hydroxide electrolyte.  It is hypothesized that barium carbonate forms from the reaction between dissolved carbon dioxide and barium sulfate.  As a result of this, it is recommended that barium carbonate is tested as an additive in place of barium sulfate.  Figure 3.30 shows that barium sulfate has a statistically relevant improvement on cell performance, and a comparison between barium sulfate and barium carbonate as a cathode additive may indicate a cause   114  for the improvement in performance obtained by using barium sulfate.  In addition, it is recommended that the potassium hydroxide electrolyte is prepared in an inert atmosphere. The role of each additive is not completely understood, however, previous literature has shown that some additives can have a synergistic effect on cell performance such as TiB2/TiS2 with Bi2O333.  For example, in section 3.3.5 calcium hydroxide has been shown to selectively precipitate soluble zinc ions which are able to cross the separator during cell operation, thus preventing the formation of Zn2MnO4 (haeterolyte).  In addition, bismuth oxide has shown an improvement in cell performance (reduced capacity fade and improved reproducibility) although the precise mechanism is not yet understood.  For these reasons, it may be useful to use both these additives simultaneously.  As the roles of specific additives become known, many such additives could be added to the cathode for various purposes towards improving cycling performance. 4.2.3 Alternative Synthesis Conditions of EMD Higher water content and purity present in EMD materials and the resultant effect on primary cell performance has been described by Ruetschi,67 however, a systematic investigation of these factors with respect to rechargeable performance is needed.  The majority of EMD materials synthesized for alkaline batteries are destined for primary cells, and it is therefore recommended that new types of EMD materials are investigated.  By improving the process of EMD synthesis, new types of materials may be developed which are targeted to improve key features such as structural water content, proton conductivity, crystallinity, replacement of cation vacancies with other metals (i.e., titanium, nickel, cobalt) which may lead to better performance.      115  4.2.4 Cell Setup to Focus on EMD Materials Testing In order to test the performance of cathode materials, a different cell setup is recommended which uses a cathode pellet prepared by pressing EMD powder and graphite under approximately 20 tons of pressure.  This approach eliminates the possible impacts of binder and/or gelling agent.  Furthermore, the cell would ideally be operated in a half-cell configuration to eliminate the effects of the zinc anode.  The type of cell necessary for such measurements has already been constructed and tested. 4.2.5 Analytical Characterization The results presented in Figures 3.41 and 3.42 show that a phase transition in EMD occurs during the discharge process which is reflected by a shift in the peak occurring during the first discharge to lower angles 2θ.  This change in peak angle could be explained by the expansion of the lattice due to proton intercalation inducing an expansion in volume up to 18% (determined by a comparison of unit cell dimensions of starting material ramsdellite and major discharge product groutite).49  In-situ x-ray diffraction analysis has been reported over three cycles,68   no investigation has been reported over many (10 to 100+) cycles.  If this type of apparatus is constructed, it could be used to identify phases which are present in cells which have experienced dramatic capacity failure. Scanning electron microscopy (SEM)/energy dispersive X-ray spectroscopy (EDS) analysis can be performed on the additives tested in the cathode electrode (Bi2O3, Ca(OH)2, etc.).  These specific additives have shown to have a significant effect on cycling performance at 5 wt. %.  High resolution SEM/EDS analysis of cathode electrodes before and after cycling may elucidate the mechanism(s) of capacity failure.  The effect of each specific additive may differ based on the type of current collector used (thin film foil vs. expanded metal mesh), and should be investigated with this in mind.    116  4.3 Industry Recommendations  The work presented in this thesis represents lab scale performance of flat-plate rechargeable alkaline manganese dioxide-zinc cells in order to study the fundamental electrochemistry.  As this project is industry funded, this section will provide a framework and suggestions to scale this battery chemistry and architecture towards a commercial product which might be used in applications such as backup power systems.  Figure 4.2 shows a schematic of how this architecture could be scaled into a battery pack.    Figure 4.2 – Flat-plate battery architecture using expanded metal mesh CCs. This battery pack could scale based on the number of cells for a particular application.  Assuming a Delrin plastic casing, springs, and the dimensions specified in Figure 4.2, the energy density of the   117  battery pack will vary based on the number of cells (n) in the battery pack.  As a result of this geometry, the key factor in improving energy density is the percentage of the electrode weight and volume which the current collector constitutes.  When considering the battery pack application, the volumetric and gravimetric energy densities are important factors.  Varying the electrode area does little to influence the energy density of the battery pack as this geometry is two dimensional.  As such, a 1 cm2 electrode area will provide the same energy density as a 100 cm2 electrode area.  Figure 4.3 shows the volumetric and gravimetric energy density of this type of battery system and how it changes as a function of number of cells.    Figure 4.3 – Flat-plate battery architecture volumetric and gravimetric energy density vs. number of cells with and without casing. A battery pack consisting of 300 or more cells will provide a volumetric energy density of greater than 250 Wh/L, and a gravimetric energy density of near 100 Wh/kg.  This battery pack would be approximately 25 cm x 25 cm x 35 cm and weigh approximately 30 kg and provide 2.7 kWh (enough to   118  power a 270 W desktop computer for 10 hours).  The cells could be wired in series, parallel, or a combination of both in distinct modules to provide the necessary capacity and voltage. The construction of a battery pack would require extensive testing and validation as these are potentially high voltage systems with caustic electrolyte.  Rather than simply putting cells into a single compartment and linearly scaling, it is recommended that construction proceeds via smaller size (e.g., 12 V, 6 cells) modules which can be individually tested before being placed in a larger battery pack.       119  5 Conclusions   A lab scale flat-plate rechargeable alkaline MnO2/Zn system has been developed and investigated with the goal of improving cumulative capacity fade.  The effect of various operating conditions, cathode additives, cathode current collectors, charge/discharge protocols were investigated with respect to long term cycling performance.   The following conclusions could be drawn: Effect of Different Operating Conditions  1. Uniform pressure during operation of cells is a key factor in cell performance with approximately 50 psi providing good performance.  2. Lower depth of discharge (cut-off potential) increases the rate of capacity fade, and is associated with the formation of electrochemically irreversible phases.  3. Slower rates of discharge improve initial capacity and rate of fade.  4. Intermittent deep discharge and time cut-off charge/discharge protocols have been investigated for different applications.  Combined deep and shallow depths of discharge prolong the cycle life.  Electrode Thickness Effects 1. Thin film foil and expanded metal mesh cathode current collectors have been investigated, each providing significant benefits.  2. Thin film foil current collectors provide a higher plateau voltage, easier fabrication, lower energy density (active material loading), a more linear rate of capacity fade, and flexible electrodes.  3. Expanded metal mesh current collectors provide higher energy density, are more difficult to fabricate (at present), and fade more rapidly within the first 10 to 15 cycles compared to the thin film foil current collectors.      120  Effect of Cathode Additives 1. Graphite and carbon black additives have been tested in the cathode, showing best performance with KS44 graphite.  2. Amount of graphite (above 15 wt. %) does not have a significant effect on capacity fade.  3. Super C65 carbon black additives show greater utilisation of the active material but a higher rate of capacity fade.  4. Various cathode additives have been tested at 5 wt. %, with Bi2O3 showing best performance on capacity fade.  Analytical Characterization 1. Powder X-ray diffraction analysis shows the formation of carbonates (BaCO3) when BaSO4 and Sr(OH)2.8H2O (SrCO3) additives were used in the cathode.  2. Powder X-ray diffraction shows the formation of electrochemically irreversible haeterolyte phase (Zn2MnO4), and calcium zinc hydroxide hydrate (Ca(Zn(OH)3.2H2O) when calcium hydroxide was used as a cathode additive.  This additive could be used in order to prevent haeterolyte phase formation during cell operation.  3. X-ray photoelectron spectroscopy results indicate Mn2O3 forms during cycling.  Summary of Conclusions Figure 5.1 shows a high level summary of the conclusions of this research, organized into four main categories of investigation: 1) Operating Conditions, 2) Cathode Additives, 3) Electrode Processing, and 4) Characterization.   121   Figure 5.1 – Summary of Conclusions in 4 areas of investigation: 1) Operating Conditions, 2) Cathode Additives, 3) Electrode Processing, and 4) Characterization.       122  Final Comments 1. Flat-plate rechargeable alkaline MnO2/Zn batteries are suited to slow drain rates and intermittent use type applications.  2. Cathode additives can improve performance only marginally; however the effect of operating conditions (rate of discharge, cut-off potential, and stack pressure) is more significant.   3. Flat-plate rechargeable alkaline MnO2/Zn batteries provide significant benefits such as low cost, environmentally friendly and abundant materials, safety, and moderate energy density.                    123  Bibliography (1)  Buchmann, I. BU-103: Global Battery Markets http://batteryuniversity.com/learn/article/global_battery_markets. (2)  Corathers, L. A. 2011 Minerals Yearbook Manganese [ advance Release ]; 2013. (3)  Massachusetts Institute of Technology. Short Primer on Batteries http://web.mit.edu/2.009/www/resources/mediaAndArticles/batteriesPrimer.pdf (accessed Oct 2, 2015). (4)  Buchmann, I. What’s the Best Battery? http://batteryuniversity.com/learn/article/whats_the_best_battery (accessed Jan 14, 2015). (5)  Daniel-Ivad, J. Encycl. Electrochem. Power Sources 2009, 5, 497–512. (6)  Arnott, J. B.; Donne, S. W. J. Electrochem. Soc. 2007, 154, A776. (7)  Othmer, K. Encycl. Chem. Technol. 2007, 48. (8)  Stani, A.; Taucher-Mautner, W.; Kordesch, K.; Daniel-Ivad, J. J. Power Sources 2006, 153, 405–412. (9)  Engineers. Energizer Batter. Manuf. Inc. 2012, 7323, 1–14. (10)  Hafner, B. Scanning Electron Microscopy Primer http://www.charfac.umn.edu/sem_primer.pdf. (11)  Bonakdarpour, A.; Mehta, S.; Xi, W.; Afonso, G.; Wilkinson, D. P. . (12)  Lower, S.; Larsen, D. Electrochemistry 6: Electrochemical Energy Storage and Conversion http://chemwiki.ucdavis.edu/Analytical_Chemistry/Electrochemistry/Electrochemistry_6%3A_Electrochemical_Energy_Storage_and_Conversion. (13)  Buchmann, I. Battery Statistics http://batteryuniversity.com/learn/article/battery_statistics (accessed Jan 19, 2015). (14)  Roland Berger Strategy Consultants. Technology & Market Drivers for Stationary and Automotive Battery Systems; 2012. (15)  Buchmann, I. The Cost of Portable Power http://batteryuniversity.com/learn/article/the_cost_of_portable_power (accessed Sep 29, 2015).   124  (16)  Inc., A. M. Why Electrolytic Manganese Dioxide (EMD) and Chemical Manganese Dioxide (CMD) http://www.americanmanganeseinc.com/why-electrolytic-manganese-dioxide-emd-and-chemical-manganese-dioxide-cmd/ (accessed Jan 14, 2015). (17)  Energy, L.; First, P.; Liquids, G. T. L.; Chinchilla, a T. 2008, 1–5. (18)  (USGS), U. S. G. S. 2009 Minerals Yearbook - Zinc; 2011. (19)  American Galvanizers Association. Zinc Recycling http://www.galvanizeit.org/hot-dip-galvanizing/what-is-zinc/zinc-recycling (accessed Jan 20, 2015). (20)  Metals, M. V. & M. A. S. Metal Commodities Pricing Chart http://www.vincentmetals.com/1Pricing/Price_Data/. (21)  Kordesch, K.; Urry, L. F. US Patent No 2,960,558 - Alkaline Cell.pdf. 2,960,558, 1957. (22)  Davidson, M.; Florida State University. Zinc-Carbon Batteries http://micro.magnet.fsu.edu/electromag/electricity/batteries/zinccarbon.html (accessed Jan 14, 2015). (23)  Boden, D.; Venuto, C. J.; Wisler, D.; Wylie, R. B. J. Electrochem. Soc. 1967, 114, 415. (24)  Donne, S. W.; Lawrance, G. a; Swinkels, D. a J. 1997, 144, 2954–2961. (25)  Malloy, A. P.; Donne, S. W. J. Electrochem. Soc. 2008, 155, A817. (26)  Mesa Minerals Limited. Electrolytic Manganese Dioxide http://mesaminerals.com.au/products/electrolytic-manganese-dioxide-emd/ (accessed Jan 14, 2015). (27)  Li, X.; Li, Z.; Xia, T.; Dong, H.; Song, Y.; Wang, L. J. Phys. Chem. Solids 2012, 73, 1229–1234. (28)  Ruetschi, P. J. Electrochem. Soc 1984, 2737. (29)  Bailey, M. R.; Donne, S. W. J. Electrochem. Soc. 2012, 159, A999–A1004. (30)  Bailey, M. R.; Denman, J. A.; King, B. V; Donne, S. W. J. Electrochem. Soc. 2012, 159, A158. (31)  Reddy, T. Linden’s Handbook of Batteries; Reddy, T. B., Ed.; 4th Editio.; McGraw-Hill Professional, 2010. (32)  Raghuveer, V.; Manthiram, A. Electrochem. commun. 2005, 7, 1329–1332.   125  (33)  Raghuveer, V.; Manthiram, A. J. Power Sources 2006, 159, 1468–1473. (34)  Shen, Y.; Kordesch, K. J. Power Sources 2000, 87, 162–166. (35)  Transene Company, I. Bright Electroless Tin http://transene.com/sn/ (accessed Sep 22, 2015). (36)  Gerken, D. M. Chemistry 2810 Lecture Notes - University of Lethbridge http://classes.uleth.ca/200501/chem2810a/lecture_21.pdf (accessed Feb 1, 2015). (37)  Koslow Scientific Company. Mercury Oxide Reference Electrode - 5088 http://www.koslow.com/reference_electrodes/5088/mercury_oxide_reference_electrode_5088?cPath=62&zenid=9a7f3a328bd036cd631b275d65d9249b. (38)  Maccor Inc. Maccor 4300 Series http://www.maccor.com/Products/Model4300.aspx (accessed Feb 2, 2015). (39)  MTI Corporation. MTI Battery Analyzer http://www.mtixtl.com/8ChannelsBatteryAnalyzer-BST8-MA.aspx (accessed Feb 2, 2015). (40)  Ametek, I. Model 1470 E - Overview http://www.solartronanalytical.com/our-products/potentiostats/Model-1470e.aspx (accessed Sep 16, 2015). (41)  Hanno zur Loye. X-ray diffraction. How it works. What it can and cannot tell us. http://www.chem.sc.edu/faculty/zurloye/xrdtutorial_2013.pdf (accessed Jan 14, 2015). (42)  Western University. Surface Science Western | About About Surface Science Western http://www.surfacesciencewestern.com (accessed Sep 27, 2015). (43)  Fujifilm. Prescale Film http://www.fujifilm.ca/products/measurement-films/prescale/film/ (accessed Jun 29, 2015). (44)  Balachandran, D.; Morgan, D.; Ceder, G. J. Solid State Chem. 2002, 166, 91–103. (45)  Donne, S. W. J. Electrochem. Soc. 1997, 144, 2961. (46)  Roberge, P. R.; Farahani, M.; Tomantschger, K.; Oran, E. J. Power Sources 1994, 47, 13–26. (47)  Bailey, M. R.; Denman, J. A.; Donne, S. W. J. Electrochem. Soc. 2014, 161, A403–A409. (48)  IMERYS Graphite and Carbon. Specialty Carbons for Alkaline Batteries https://www.google.ca/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8  126  &ved=0CB0QFjAA&url=http://www.imerys-graphite-and-carbon.com/wordpress/wp-app/uploads/2014/04/Alkaline_batteries.pdf&ei=zzpVVZOWG4-qogS4iYCIBA&usg=AFQjCNGiv6ApH (accessed May 14, 2015). (49)  Malloy, A. P.; Donne, S. W. J. Power Sources 2008, 179, 371–380. (50)  MTI Korea. TIMICAL SUPER C65 Conductive Carbon Black as Conductive Additive for Lithium Ion Batteries http://www.mtikorea.co.kr/product/detail/3185.html?product_no=3185 (accessed May 12, 2015). (51)  MatWeb. TIMCAL TIMREX® KS44 Primary Synthetic Graphite http://www.matweb.com/search/datasheet.aspx?matguid=fde407c9c3044d12acaad2ef397d44e4&ckck=1 (accessed May 12, 2015). (52)  SCM New Energy Department. Lithium Ion Conductive Agents (IMERYS) http://www.scmbattery.com/en/product_show.php?id=101 (accessed May 14, 2015). (53)  Root, M. J. J. Appl. Electrochem. 1995, 25, 1057–1060. (54)  Im, D.; Manthiram, A. J. Electrochem. Soc. 2003, 150, A68. (55)  Minakshi, M.; Singh, P.; Carter, M.; Prince, K. Electrochem. Solid-State Lett. 2008, 11, A145. (56)  Patnaik, P. Handbook of Inorganic Chemicals; McGraw-Hill, 2002. (57)  The Lubrizol Corporation. Carbopol 940 Polymer https://www.lubrizol.com/PersonalCare/Products/Carbopol/Carbopol940.html (accessed May 16, 2015). (58)  Kozawa, a.; Powers, R. a. J. Electrochem. Soc. 1966, 113, 870. (59)  Ingale, N. D.; Gallaway, J. W.; Nyce, M.; Couzis, A.; Banerjee, S. J. Power Sources 2015, 276, 7–18. (60)  Biswal, A.; Tripathy, B. C.; Subbaiah, T.; Meyrick, D.; Ionescu, M.; Minakshi, M. Metall. Mater. Trans. E 2014, 1, 226–238. (61)  Gallaway, J. W.; Menard, M.; Hertzberg, B.; Zhong, Z.; Croft, M.; Sviridov, L. a.; Turney, D. E.; Banerjee, S.; Steingart, D. a.; Erdonmez, C. K. J. Electrochem. Soc. 2014, 162, A162–A168.   127  (62)  Zheng, Y.; Wang, J. M.; Chen, H.; Zhang, J. Q.; Cao, C. N. Mater. Chem. Phys. 2004, 84, 99–106. (63)  Ferrell, D. T.; Vosburgh, W. C. J. Electrochem. Soc. 1951, 98, 334. (64)  Scherson, D. a; Division, C. 1997, 144, 1598–1603. (65)  Kang, H. Y.; Liang, C. C. J. Electrochem. Soc. 1968, 115, 6. (66)  Dose, W. M.; Lehr, J.; Donne, S. W. Mater. Res. Bull. 2012, 47, 1827–1834. (67)  Ruetschi, P. J. Electrochem. Soc. 1988, 135, 2663. (68)  Mondoloni, C. J. Electrochem. Soc. 1992, 139, 954. (69)  Nippon Paper Industries Co. Ltd. [Chemical name: Carboxymethyl Cellulose (CMC)] http://www.npchem.co.jp/english/product/sunrose/index.html. (70)  MTI. Styrene-Butadiene Rubber (SBR) Binder for Li-ion Battery Anode 260g/bottle - EQ-Lib-SBR http://www.mtixtl.com/Styrene-ButadieneRubberSBRbinderforLi-ionBatteryAnode260g/bottle.aspx (accessed Sep 16, 2015). (71)  MTI. Polytetrafluoroethylene (PTFE) Condensed Liquid Binder for Li-ion Battery 360g/bottle EQ-Lib-PTFE http://www.mtixtl.com/PolytetrafluoroethylenePTFECondensedLiquidBinderforLi-ionBattery.aspx (accessed Sep 16, 2015). (72)  Bruker Corporation. X-ray Powder Diffraction on a desk https://www.bruker.com/products/x-ray-diffraction-and-elemental-analysis/x-ray-diffraction/d2-phaser/overview.html?amp%253Bprint=1 (accessed Sep 16, 2015). (73)  UBC Centre for High Throughput Phenogenomics. Helios NanoLab 650 Focussed Ion Beam SEM http://www.phenogenomics.dentistry.ubc.ca/equipment/HeliosFIB-SEM/ (accessed Sep 16, 2015). (74)  BioLogic Corporation. BioLogic VMP3 Potentiostat http://www.bio-logic.info/instruments/multi-potentiostat-vmp3/ (accessed Sep 16, 2015).        128  Appendices Appendix A – Materials Preparation  Anode Preparation Table A1: Anode Recipe. Component Weight (g) for ~16 g batch Wet Weight % Horsehead Zinc Powder 11 67.6 ZoCHEM Zinc Oxide Powder 2.75 16.9 Indium Sulfate Solution (sat’d) 0.22 1.4 Carbopol Gel 2 12.3 Polyethylene Glycol (PEG) Solution 0.1 0.6 60% PTFE suspension 0.2 1.2 Total 16.27 100  Anode Preparation Procedure: 1) Clean a 50 mL glass beaker thoroughly and dry. 2) Clean a glass rod and 3 spatulas thoroughly and dry. 3) Add ~2 g Carbopol gel to the empty beaker and record weight used. 4) Add 5 g zinc powder to beaker and mix for 5 minutes with glass rod. 5) Add 2 g zinc oxide powder to beaker and mix for 5 minutes with glass rod. 6) Add 6 g zinc powder to beaker and mix until homogeneous. 7)    Add the following additives and ensure they are thoroughly mixed. -0.22 g Indium Sulfate solution  -0.1 g Polyethylene Glycol (PEG) solution -0.75 g Zinc oxide powder -0.2 g PTFE solution        8)    Paste into tin plated brass mesh 2 cm x 4 cm (Product ID: 4Br7.5-080TP) before mixture dries.  Cathode Preparation Table A2: Control Cathode Recipe. Distributor/Component Weight (g) Weight % EAB EMD Lot # 21006 16.8 84% Timrex KS44 Graphite 3.2 16% Total 20 100%     129    Table A3: Baseline Cathode Recipe: Distributor/Component Weight (g) Weight % EAB EMD Lot # 21006 16 80% Timrex KS44 Graphite 3 15% Acros Barium Sulfate 1 5% Total 20 100%  Cathode Preparation Procedure: 1) Weigh out the above component materials in the correct ratios as indicated above. 2) High Power Ball-Mill for 10 minutes (BMR 5:1). 3) Weigh out 2 g of 1.7 wt. % Carboxymethyl Cellulose (CMC) into clean beaker. 4) Add 2 g cathode mix to beaker. 5) Mix 5 minutes with glass rod. 6) Add 2 g cathode mix to beaker. 7) Mix 5 minutes with glass rod. 8) Add 0.3 g Styrene Butadiene (SBR) binder with disposable pipet. 9) Add 3 g cathode mix to beaker. 10) Mix 5 minutes.  The mix should be peanut butter consistency. 11) Apply paste to Nickel Mesh (2 cm x 4 cm), leaving 1 cm x 2 cm on one end without paste. 12) Roll to 0.5 mm thickness using metal roller. 13) Press 4 MPa for 2 minutes between two sheets of wax paper on MTI hydraulic press. 14) Press 10 MPa for 2 minutes between two sheets of wax paper on MTI hydraulic press.  Electrolyte Preparation The electrolyte solutions of 9M potassium hydroxide (KOH) are prepared 1 L at a time.   The dissolution of KOH in deionized water is highly exothermic, and should only be attempted in glass flasks by dissolving small portions at a time in order to prevent cracking.  It is preferable to use a 1L plastic volumetric flask. Below is the procedure for electrolyte preparation. 1) Weigh out 505 g KOH pellets 2) Place approximately 500 mL deionized water into 1 L plastic volumetric flask 3) Place volumetric flask into cold water bath 4) Add 505 g KOH pellets in small portions, constantly stirring to improve head dissipation into cold water bath. 5) Add deionized water to 1 L total volume.   130  Appendix B – Materials Information  Binder/Gelling Agents Information (CMC, Carbopol 940, SBR, PTFE) Carboxymethyl Cellulose (CMC) - Cathode Electrode Gelling Agent69 CMC is a cellulose derivative commonly used as a viscosity modifier in a number of applications ranging from paints to food products.  CMC provides a macroviscous, long flow (drip) gel.  1.7 wt. % MAC 350 CMC is used in the cathode electrode paste as a gelling agent and has the following chemical structure.   It is prepared by the following method: 1) Add 1 L deionized water to glass container. 2) Start stirring with automatic propeller mixer at 1100 RPM. 3) Add 1.7 g MAC 350 CMC to mixture in small portions ensuring mixture does not stick to walls of glass container or to propeller. 4) Mix 8 hours until transparent viscous gel is produced. Carbopol 940 - Anode Electrode Gelling Agent57 Carbopol 940 is a cross linked polyacrylate polymer capable of producing a macroviscous, short flow (non-drip) gels typically used in applications towards personal care products and gels.  Carbopol 940 is used in the anode electrode paste as a gelling agent.  The precise details of its chemical structure are not provided by the supplier. Styrene Butadiene Rubber (SBR) Suspension- Cathode Electrode Binder70 SBR binder suspension (49-51%) is a water based co-polymer binder used in the cathode electrode at 2.5 dry weight % to provide mechanical stability and adherence to the current collector.  It is a slightly acidic (pH 6-7), water soluble, milky white liquid with a light blue gloss.  It has the following chemical structure.     131   Polytetrafluoroethylene (PTFE) Suspension – Anode Electrode Binder71 The 60% Teflon binder is a milky white fine dispersion liquid stabilized by a non-ionic surfactant (3-4 wt. %).   It is resistant to high temperatures and provides excellent electric insulation.  It has the following chemical structure.                 132  Appendix C - Instrumentation  Reference Electrode Calibration Procedure 1) Connect the following 3 electrode setup to Solartron Potentiostat: -Working Electrode: Platinum foil/wire. -Counter Electrode: Platinum foil/wire. -Reference Electrode: Reference you wish to calibrate.  2) Bubble N2 to degas electrolyte for 15 minutes  3) Perform a CV to clean the Pt W.E. surface: -200 cycles at 500 mV/s between -0.75 to 0.4 V vs. Hg/HgO (0.05 to 1.2 V vs RHE) -Choose range of CV based on potential of reference electrode you wish to calibrate. -Do this with a N2 blanket.          4)  Bubble H2 gas through solution for 15 minutes and measure OCV: -Potential of reference electrode will likely shoot down to reference potential vs. RHE very quickly. After 15 minutes it should be stabilized. -Read off stabilized potential and if fluctuating note the voltage range which it is fluctuating within.  Bruker D2 Phaser (Powder X-Ray Diffraction) PXRD analysis was performed with a Bruker D2 Phaser benchtop diffractometer.  The D2 phaser contains a LINXEYETM detector, scintillation counter, and standard size (51.5 mm) sample holder, and the DIFFRAC.SUITE software package for sample identification and analysis.  Table C1 shows the range of experimental run parameters available with the D2 Phaser. See reference for specifications.72 Table C1: Bruker D2 Phaser Experimental Run Parameters. Parameter Type Min Value Max Value Typical Value Purpose Scan Type N/A N/A Coupled Two Theta/Theta N/A Scan Mode N/A N/A Continuous PSD fast N/A 2θ [o] Start 4 (<End) 10 Start Angle 2θ 2θ [o] End (>Start) 90 80 End Angle 2θ Increment [o] 0.01 0.08 0.02 Goniometer 2θ increment. Time [s] 0.1 0.8 0.2-0.4 Time spent collecting at each increment. Variable Rotation 0 30 10 Sample rotation speed.   133   Helios NanoLab 650 (Scanning Electron Microscopy/Energy Dispersive X-Ray Spectroscopy – SEM/EDS)  SEM/EDS analysis was performed on a Helios NanoLab 650 at the UBC Centre for High Throughput Phenogenomics (CHTP).  The Elstar In-Lens and Everhart-Thornley detectors were used for secondary and backscattered electron imaging and an integrated EDAX Team Pegasus system with EDS (SDD) detector for elemental analysis.  See reference for specific details.73 Electrochemical Measurements The following instruments were used to perform and analyze electrochemical measurements of the full and half-cell apparatuses including various types of charge/discharge protocols and EIS measurements. BioLogic VMP3 Potentiostat The BioLogic VMP3 is a multi-channel Potentiostat capable of various electrochemical measurements including simultaneous multi-channel EIS measurements.  The EC-Lab® software allows for simultaneous, remote multi-user input measurements via Ethernet-LAN connection.  The system currently has 8 channels (up to 16 maximum).  See reference for specifications.74 Solartron 1470 E Potentiostat The Solartron 1470 E Potentiostat is an 8 channel test system used for half-cell and EIS measurements providing a wide range of experimental techniques including potentiostatic/potentiodynamic, galvanostatic/galvanodynamic, cyclic voltammetry, and EIS measurements.  See reference for specifications.40 Maccor 4300 Series Battery Analyzer The Maccor Model 4300 is an 8 channel battery analyzer used for full-cell testing protocols.  It is integrated with a comprehensive software package used for designing and programming different charge/discharge protocols, measurement acquisition, and data analysis.  The Maccor system has a lower current cut-off of 6 μA, allowing for slow discharge protocols used with the thin film cathode materials. See reference for specifications.38 MTI BTS8-WA Battery Analyzer The MTI BTS8-WA battery analyzer is an 8 channel system used for full cell testing protocols.  Three of these systems have been combined for a total of 18 channels.  It is integrated with a software package for programming of basic charge/discharge protocols and plotting cell data.  The MTI system has a lower current cut-off of 10 mA and is therefore used for full-cell measurements of the expanded metal mesh cathode materials which have larger mass.  See reference for specifications.39   

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0223560/manifest

Comment

Related Items