Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Bifunctional oxygen reduction/evolution catalysts for rechargeable metal-air batteries and regenerative… Hosseini-Benhangi, Pooya 2016

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

Item Metadata

Download

Media
24-ubc_2017_february_hosseinibenhangi_pooya.pdf [ 14.22MB ]
Metadata
JSON: 24-1.0340612.json
JSON-LD: 24-1.0340612-ld.json
RDF/XML (Pretty): 24-1.0340612-rdf.xml
RDF/JSON: 24-1.0340612-rdf.json
Turtle: 24-1.0340612-turtle.txt
N-Triples: 24-1.0340612-rdf-ntriples.txt
Original Record: 24-1.0340612-source.json
Full Text
24-1.0340612-fulltext.txt
Citation
24-1.0340612.ris

Full Text

Bifunctional Oxygen Reduction/Evolution Catalysts for Rechargeable Metal-Air Batteries and Regenerative Alkaline Fuel Cells by  Pooya Hosseini-Benhangi  M.Sc., Ferdowsi University of Mashhad, 2011 B.Sc., Ferdowsi University of Mashhad, 2009  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Chemical and Biological Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  December 2016  © Pooya Hosseini-Benhangi, 2016 ii  Abstract The electrocatalysis of oxygen reduction and evolution reactions (ORR and OER, respectively) on the same catalyst surface is among the long-standing challenges in electrochemistry with paramount significance for a variety of electrochemical systems including regenerative fuel cells and rechargeable metal-air batteries. Non-precious group metals (non-PGMs) and their oxides, such as manganese oxides, are the alternative cost-effective solutions for the next generation of high-performance bifunctional oxygen catalyst materials. Here, initial stage electrocatalytic activity and long-term durability of four non-PGM oxides and their combinations, i.e. MnO2, perovskites (LaCoO3 and LaNiO3) and fluorite-type oxide (Nd3IrO7), were investigated for ORR and OER in alkaline media. The combination of structurally diverse oxides revealed synergistic catalytic effect by improved bifunctional activity compared to the individual oxide components.  Next, the novel role of alkali-metal ion insertion and the mechanism involved for performance promotion of oxide catalysts were investigated. Potassium insertion in the oxide structures enhanced both ORR and OER performances, e.g. 110 and 75 mV decrease in the OER (5 mAcm-2) and ORR (-2 mAcm-2) overpotentials (in absolute values) of MnO2-LaCoO3, respectively, during galvanostatic polarization tests. In addition, the stability of K+ activated catalysts was improved compared to unactivated samples.  Further, a factorial design study has been performed to find an active nanostructured manganese oxide for both ORR and OER, synthesized via a surfactant-assisted anodic electrodeposition method. Two-hour-long galvanostatic polarization at 5 mAcm-2 showed the lowest OER degradation rate of 5 mVh-1 for the electrodeposited MnOx with 270 mV lower OER overpotential compared to the commercial γ-MnO2 electrode.   iii  Lastly, the effect of carbon addition to the catalyst layer, e.g. Vulcan XC-72, carbon nanotubes and graphene-based materials, was examined on the ORR/OER bifunctional activity and durability of MnO2-LaCoO3. The highest ORR and OER mass activities of -6.7 and 15.5 Ag-1 at 850 and 1650 mVRHE, respectively, were achieved for MnO2-LaCoO3-multi_walled_carbon_nanotube-graphene, outperforming a commercial Pt electrode. The factors affecting the durability of mixed-oxide catalysts were discussed, mainly attributing the performance degradation to Mn valence changes during ORR/OER. A wide range of surface analyses were employed to support the presented electrochemical results as well as the proposed mechanisms. iv  Preface The materials presented in the following dissertation are confidential and subject to United States and Canadian patent applications. The research work presented in this thesis including literature review, research proposal, experimental work, data interpretation, proposed hypothesis and preparation of this dissertation, four research manuscripts, five conference presentations along with one provisional and one U.S./Canadian patent applications are completed by Pooya Hosseini-Benhangi under the direct supervision of Professors Előd Gyenge and Akram Alfantazi at the Department of Chemical & Biological Engineering Department, the University of British Columbia. The following publications, manuscripts, presentations and patent applications are developed form the work presented in this dissertation: A version of chapter 3 and 4 was published and presented in: 1) P. Hosseini-Benhangi, A. Alfantazi, E. Gyenge, “Manganese dioxide-based bifunctional oxygen reduction/evolution electrocatalysts: Effect of perovskite doping and potassium ion insertion”, Journal of Electrochimica Acta, 123 (2014) 42-50. 2) P. Hosseini-Benhangi, M. A. Garcia-Contreras, A. Alfantazi, E. Gyenge, “Method for enhancing the bifunctional activity and durability of oxygen electrodes with mixed oxide electrocatalysts: Potential driven intercalation of potassium”, Journal of The Electrochemical Society, 162 (2015) F1356-F1366. 3) P. Hosseini-Benhangi, A. Alfantazi, E. Gyenge, “Investigation of non-precious metal bifunctional oxygen cathodes for alkaline fuel cells and batteries”, Hydrogen Fuel Cell conference (HFC 2013), Vancouver, Canada, June 2013. v  4) P. Hosseini-Benhangi, A. Alfantazi, E. Gyenge, “Doped MnO2-based oxygen reduction and evolution catalysts for bifunctional cathodes in alkaline electrochemical power sources”, 224th Electrochemical Society Meeting, San Francisco, U.S., October 2013. 5) P. Hosseini-Benhangi, A. Alfantazi, E. Gyenge, “MnO2-based bifunctional oxygen catalyst for rechargeable metal/air batteries: The effect of K+ intercalation”, 226th Electrochemical Society Meeting, Cancun, Mexico, October 2014 & AIChE 2014 Annual Meeting, Atlanta, U.S., November 2014. A version of chapter 5 is in preparation for publication and was presented at a conference: 6) P. Hosseini-Benhangi, A. Alfantazi, E. Gyenge, “Surfactant-assisted electrodeposition of Mn oxides as promising ORR/OER bifunctional non-PGM electrocatalysts: Factorial design study of the electrodeposition factors”, to be submitted. 7) E. Gyenge, P. Hosseini-Benhangi, C. H. Kung, A. Alfantazi, “Toward active and durable oxygen reduction/oxygen evolution (ORR/OER) bifunctional non-PGM electrocatalysts for rechargeable metal-air batteries and regenerative fuel cells”, International Workshop on Advanced Materials and Nanotechnology (IWAMN 2016), Hanoi, Vietnam, November 2016. A version of chapter 6 is in preparation for publication and was presented at a conference: 8) P. Hosseini-Benhangi, M. A. Garcia-Contreras, A. Alfantazi, E. Gyenge, “Carbon support effect on ORR/OER bifunctional activity and durability of non-PGM mixed-oxide catalyst: Graphene vs. commercial carbon materials”, to be submitted. vi  9) M. Garcia, P. Hosseini-Benhangi, A. Taheri, E. Gyenge, “The effect of graphene as a support for non-PGM bifunctional oxygen catalyst in rechargeable metal/air batteries”, AIChE 2014 Annual Meeting, Atlanta, U.S., November 2014. For the items 1 to 9, all of the experiments were performed by Pooya Hosseini-Benhangi and the manuscripts were co-authored by Előd Gyenge with the support from Akram Alfantazi. For items 2, 8 and 9, the experiments were performed by Pooya Hosseini-Benhangi with laboratory recommendations from Dr. Miguel Angel Garcia-Contreras. The manuscripts and presentations mentioned in items 2, 8 and 9 were also co-authored by Dr. Miguel Angel Garcia-Contreras. Parts of chapters 2, 3, 4, 5 and 6 from this dissertation were filed as United States and Canadian patent applications: 10)  E. Gyenge, P. Hosseini-Benhangi, “An oxygen electrode and a method of manufacturing the same”, U.S. (15/251,267) and Canadian (2,940,921) patent applications, filed on August 30th, 2016. where all of the experiments for this intellectual property were performed by Pooya Hosseini-Benhangi. The U.S. and Canadian patent applications were co-authored by Előd Gyenge and Brian Yat-Ming Lee (Patent Attorney).  vii  Table of Contents Abstract .......................................................................................................................................... ii Preface ........................................................................................................................................... iv Table of Contents ........................................................................................................................ vii List of Tables ............................................................................................................................... xii List of Figures ............................................................................................................................. xiv Nomenclature .......................................................................................................................... xxxii List of Abbreviations ...............................................................................................................xxxv Acknowledgements .............................................................................................................. xxxviii Dedication ................................................................................................................................... xlii Chapter 1: Introduction ................................................................................................................1 1.1 Summary ..................................................................................................................... 1 1.2 Bifunctional catalysts for ORR and OER in alkaline batteries and fuel cells ............ 9 1.2.1 Nobel metals and their alloys ............................................................................ 10 1.2.2 Manganese dioxide ........................................................................................... 11 1.2.2.1 Electro-reduction of MnO2 .......................................................................... 15 1.2.2.2 Oxygen reduction reaction on manganese oxides ....................................... 20 1.2.2.3 Oxygen evolution reaction on manganese oxides ....................................... 29 1.2.2.4 Density Functional Theory (DFT) studies on manganese oxides catalyzing both ORR and OER .......................................................................................................... 35 1.2.2.5 Dopants for manganese dioxide as cathode material in alkaline fuel cells and metal-air batteries.............................................................................................................. 37 1.2.2.6 Nanostructured manganese oxides .............................................................. 38 viii  1.2.3 Perovskite-type oxides ...................................................................................... 56 1.2.4 Fluorite-type oxides .......................................................................................... 58 1.2.5 Carbon support for ORR/OER bifunctional catalysts ....................................... 59 1.3 Knowledge gap and research objectives ................................................................... 61 1.3.1 Knowledge gap ................................................................................................. 61 1.3.2 Research objectives ........................................................................................... 62 Chapter 2: Experimental methods, apparatus and materials .................................................65 2.1 Material preparation .................................................................................................. 65 2.1.1 Catalyst powders ............................................................................................... 65 2.1.1.1 Perovskites .................................................................................................. 65 2.1.1.2 Nd3IrO7 ........................................................................................................ 65 2.1.1.3 Manganese dioxide ...................................................................................... 66 2.1.1.4 Carbonaceous materials .............................................................................. 66 2.1.1.5 Platinum ...................................................................................................... 66 2.1.2 Catalyst layer preparation ................................................................................. 66 2.1.3 Gas diffusion electrode preparation .................................................................. 67 2.1.4 Anodic electrodeposition of manganese oxides ................................................ 69 2.1.5 Surface modification: K+ intercalation ............................................................. 73 2.2 Surface and structural characterization ..................................................................... 74 2.3 Electrochemical measurements ................................................................................. 75 Chapter 3: Comprehensive studies on the ORR/OER electrocatalytic activity and durability of individual (MnO2, LaCoO3, LaNiO3 and Nd3IrO7) and mixed-oxide (MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7) catalysts  ..........................................................................80 ix  3.1 Introduction ............................................................................................................... 80 3.2 Results and discussions ............................................................................................. 81 3.2.1 Characterization of oxide catalysts ................................................................... 81 3.2.2 Initial stage electrocatalytic activity of oxide catalysts .................................... 86 3.2.3 Accelerated degradation testing of oxide catalysts in flooded test setup .......... 96 3.2.4 Long-term ORR durability of oxide catalysts in a flow-by test setup .............. 97 3.3 Conclusion .............................................................................................................. 103 Chapter 4: Oxide catalyst activation by alkali-metal ion intercalation  ...............................106 4.1 Introduction ............................................................................................................. 106 4.2 Results and Discussion ........................................................................................... 107 4.2.1 Oxide catalyst layer activation by open-circuit potential (OCP) K+ intercalation  ......................................................................................................................... 107 4.2.1.1 Healing effect ............................................................................................ 107 4.2.1.2 Activation of fresh catalysts ...................................................................... 116 4.2.2 Oxide catalyst layer activation by potential driven intercalation (PDI) of K+ 129 4.2.2.1 Initial stage electrocatalytic activity of PDI activated mixed-oxide catalysts   ................................................................................................................... 129 4.2.2.2 Initial stage electrocatalytic activity of PDI activated individual oxides .. 133 4.2.2.3 Galvanostatic long-term durability testing of unactivated and PDI activated mixed-oxide catalysts...................................................................................................... 138 4.2.3 Comparison of Bifunctional ORR/OER Activities: present work vs. literature ...  ......................................................................................................................... 141 4.3 Conclusion .............................................................................................................. 144 x  Chapter 5: Surfactant-assisted electrodeposition of Mn oxides as promising ORR/OER bifunctional non-PGM electrocatalysts: Factorial design study of the electrodeposition factors  .........................................................................................................................................146 5.1 Introduction ............................................................................................................. 146 5.2 Results and discussion ............................................................................................ 147 5.2.1 Anodic electrodeposition behavior of Mn oxides with and without surfactants ..  ......................................................................................................................... 147 5.2.2 Characterization of the electrodeposited samples ........................................... 150 5.2.3 Factorial design experiments .......................................................................... 154 5.2.3.1 Triton X-100 surfactant-assisted electrodeposition .................................. 156 5.2.3.2 SDS surfactant-assisted electrodeposition ................................................ 160 5.2.3.3 CTAB surfactant-assisted electrodeposition ............................................. 164 5.2.4 ORR/OER performance comparison .............................................................. 168 5.2.5 Galvanostatic long-term durability testing of deposited MnOx ...................... 173 5.3 Conclusion .............................................................................................................. 174 Chapter 6: Carbon support effect on ORR/OER bifunctional activity and durability of non-PGM mixed-oxide catalyst: Graphene vs. commercial carbon materials  ...........................177 6.1 Introduction ............................................................................................................. 177 6.2 Results and discussion ............................................................................................ 178 6.2.1 Microstructural studies of catalyst and support materials ............................... 178 6.2.2 RRDE studies of carbonaceous materials for ORR ........................................ 181 6.2.3 ORR/OER bifunctional electrocatalytic activity measurements: Carbonaceous materials as oxygen catalysts .............................................................................................. 184 xi  6.2.4 ORR/OER bifunctional electrocatalytic activity measurements of non-PGM oxide oxygen catalyst: MnO2-LaCoO3 supported on carbonaceous materials ................... 190 6.2.5 Bifunctional accelerated degradation testing: MnO2-LaCoO3 supported on MWCNT-Graphene or Vulcan XC-72................................................................................ 197 6.3 Conclusion .............................................................................................................. 200 Chapter 7: Conclusions and recommendations for future work ...........................................203 7.1 Conclusions ............................................................................................................. 203 7.1.1 ORR/OER electrocatalytic activity and durability of individual and mixed-oxide catalysts ......................................................................................................................... 203 7.1.2 Oxide catalyst activation by alkali-metal ion intercalation ............................. 204 7.1.3 Surfactant-assisted electrodeposition of Mn oxides: Factorial design study of the electrodeposition factors ..................................................................................................... 205 7.1.4 The effect of carbon supports: Graphene vs. commercial carbon materials ... 206 7.2 Contributions to knowledge .................................................................................... 207 7.3 Recommendation for future work ........................................................................... 209 Bibliography ...............................................................................................................................211 Appendices ..................................................................................................................................228 Appendix A : Electrode kinetics: The Butler-Erdey-Grúz-Volmer (BE-GV) equation .. 228 Appendix B : The rotating disk and ring disk electrode (RDE and RRDE) .................... 231 Appendix C : Break-in protocol test results for GDE flow-by cell ................................. 235 Appendix D : The effect of inter-stage OCP activation ................................................... 237 Appendix E : Factorial design study ................................................................................ 240 xii  List of Tables Table 1.1 MnO2 crystal structure evolution at several electrode potentials in fractional composition. Other conditions idem to Figure 1.10. Reprinted with permission from [45]. ........ 18 Table 1.2 Possible ORR pathways in alkaline media [2, 17, 45, 98-100]. ................................... 21 Table 2.1 Experimental design factors and their levels for 24-1+3 factorial design runs. ............. 70 Table 3.1 XRD structural analysis of: Commercial MnO2, synthesized LaCoO3, synthesized LaNiO3 and synthesized Nd3IrO7. The XRD spectrum of the powders are presented in Figure 3.2. The bold 2θ and Miller indices show overlapped peaks for different crystallographic plans. ..... 85 Table 3.2 BET surface area of single and mixed oxide catalyst layers. ....................................... 90 Table 3.3 The apparent exchange current densities and Tafel slopes for the initial stage ORR and OER activities of the investigated GDEs with fresh catalysts. O2 saturated 6 M KOH. 293 K. PO₂ of 1 atm. The exchange current densities are expressed per geometric area. The standard error of the mean calculated based on six replicates is indicated for each catalyst. .................................. 95 Table 5.1 XPS peak analysis of electrodeposited MnOx samples (T1, T9 and T10). The deconvoluted data for Mn 2p, Mn 3s and O 1s is presented. The error associated with binding energy of peak position is ±0.1 eV. Other conditions idem to Figure 5.2. ................................. 152 Table 5.2 Design matrix (in random order), Mn valence, calculated loadings and responses for the factorial design experiments in presence of Triton X-100. Other conditions idem to Table 2.1.157 Table 5.3 Design matrix (in random order), Mn valence, calculated loadings and responses for the factorial design experiments in presence of SDS. Other conditions idem to Table 2.1. ............ 161 Table 5.4 Design matrix (in random order), Mn valence, calculated loadings and responses for the factorial design experiments in presence of CTAB. Other conditions idem to Table 2.1. ......... 165 xiii  Table 6.1 The apparent exchange current densities and Tafel slopes for the initial stage ORR and OER activities of the investigated carbonaceous catalyst materials in Figure 6.3. The exchange current densities are expressed per geometric area. The apparent Tafel slope and exchange current density values are obtained over a potential range of min. 50 mV on six replicates. Other conditions idem to Figure 6.3. ...................................................................................................................... 189 Table 6.2 The apparent exchange current densities and Tafel slopes for the initial stage ORR and OER activities of the investigated catalyst materials in Figure 6.5. The exchange current densities are expressed per geometric area. The apparent Tafel slope and exchange current density values are obtained over a potential range of min. 50 mV on six replicates. Other conditions idem to Figure 6.5. ................................................................................................................................... 196  xiv  List of Figures Figure 1.1 Schematic illustration of electrically rechargeable metal-air batteries with its operating principle and bifunctional cathode. Reprinted with permission from [16]. .................................... 3 Figure 1.2 Schematic polarization curve of rechargeable Zn-air battery. The equilibrium potential of the Zn-air battery (black line) is 1.65 V, but the practical voltage (red line) in discharge is lower than 1.65 V due to the sluggish ORR. A large potential is needed to charge Zn-air battery, higher than the equilibrium potential (blue line). Reprinted with permission from [11]. .......................... 4 Figure 1.3 Challenges facing rechargeable aqueous Li-air batteries. Reprinted with permission from [21]. ........................................................................................................................................ 4 Figure 1.4 a) Discharge polarization curves of gas diffusion electrodes containing graphene nanosheets (GNSs), 20 wt% Pt/C and acetylene black at 0.5 mA cm-2, b) Galvanostatic discharge-charge cycle curves of heat-treated GNSs at 0.5 mA cm-2. 1 M LiClO4/ED/DEC was used as the organic electrolyte and 1 M LiNO3 + 0.5 M LiOH was used as the aqueous electrolyte. A solid-state electrolyte Li1+x+yAlx(Ti,Ge)2−xSiyP3−yO12 (LISICON) film was used as a separating membrane between the organic and aqueous electrolytes to prevent intermixing of the two solutions. The discharge-charge performance was carried out at a current density of 0.5 mA cm-2 between 2 and 4.8 Vvs. Li/Li+ for 2 hrs per each cycle. 298 K. Reprinted with permission from [24].......................................................................................................................................................... 5 Figure 1.5 Schematic representation of an anion exchange membrane-unitised regenerative fuel cell (AEM-URFC) as an energy storage unit. The AEM-URFC stores renewable energy as H2 while in electrolyzer mode and then uses that H2 to generate electric energy on-demand when in fuel cell mode. Reprinted with permission from [32]. .................................................................... 7 xv  Figure 1.6 Polarization curves for the AEM-URFC under eight fuel cell/electrolyzer cycles. MEA: Fumapem FAA-3, Fumatech as Anoion exchange memberane. Gas diffusion electrodes: 6 mgcat cm−2 Ni/C for the H2 electrode and 4 mgcat cm−2 MnOx/GC:Ni/C (5:1 weight ratio) for the O2 electrode. An alkaline ionomer (1.8 μL per 1 mg catalyst, Fumion FAA-3, Fumatech) was added to each of the electrodes. 20 min per each cycle. 335 K. H2/O2 gas flow rates of 300 Standard Cubic Centimeters per Minute (SCCM). All potentials are reported vs. reversible hydrogen electrode (RHE). Reprinted with permission from [32]. ................................................................ 8 Figure 1.7 A comparison between bifunctional oxygen performance of different catalyst materials in alkaline media. The ORR and OER overpotentials were calculated at superficial current densities specified on the graph. Reprinted with permission from [4]. ........................................ 11 Figure 1.8 Schematic representation of five different crystal structures of MnO2: A) Pyrolusite (β-MnO2):  Rutile structure with an infinite chain of MnO6 octahedra sharing opposite edges. Each chain is corner-linked with four similar chains, B) Ramsdellite (α-MnO2): Cross-linking of double or triple chains of the MnO6 octahedra resulting in two-dimensional tunnels within the lattice, C) Birnessite (δ-MnO2): Layered structure containing infinite two-dimensional sheets of edge-shared MnO6 octahedra, D) Spinel (λ-MnO2): A 3D spinel structure and E) Electrolytic manganese dioxide (γ-MnO2): An intergrowth of pyrolusite in ramsdellite matrix.  Each MnO6 octahedra composes of oxygen and manganese atoms in the corners and center, respectively. Reprinted with permission from [85, 88, 89]......................................................................................................... 14 Figure 1.9 Linear sweep voltammetry of 0.2 mg MnO2 on Ni mesh in N2 saturated 10.2 M KOH. 0.05 mV s-1. 298 K. ....................................................................................................................... 17 Figure 1.10 Cyclic voltammograms of different manganese oxides supported on carbon in N2 saturated 1 M KOH at 20 mV s-1 and 298 K. Samples labeled as A, B and C are MnO2/C, xvi  Mn2O3/C and Mn3O4/C, respectively. The potentials are reported vs. mercury-mercury oxide reference electrode (Hg/HgO). Reprinted with permission from [45]. ......................................... 18 Figure 1.11 Cyclic voltammograms of four different commercially available MnO2 powders supported on carbon in N2-purged 6 M KOH at 2 mV s-1. 295 K. Third cycle is presented. The potentials are reported vs. mercury-mercury oxide reference electrode (MOE). Reprinted with permission from [90]..................................................................................................................... 20 Figure 1.12 ORR polarization curves of various catalysts following 2e− or 4e− pathways in 0.1 M KOH. Theoretical limiting current densities for the 2e− or 4e− pathways are indicated by solid lines together with a ±10% margin (dashed lines). Currents are normalized to the geometrical area of the disk. 1600 rpm. 298 K. PO₂ equals to 1 atm. Reprinted with permission from [107]. ............ 23 Figure 1.13 Possible ORR pathways on manganese oxides in alkaline media: A) Four-electron pathway reducing O2 to hydroxide, B) Two-electron pathway reducing O2 to peroxide ion and C) Two-electron reduction of peroxide ions. Orange denotes species on the catalyst surface, and blue/purple denotes species in solution. Reprinted with permission from [107]. ......................... 25 Figure 1.14 Linear sweep voltammograms of four types of manganese dioxide: A) α-MnO2 (ramsdellite), B) β-MnO2, C) γ-MnO2 and D) ε-MnO2. N2 saturated 7 N KOH. 0.003 mV s-1. 298 K. Solid and dotted lines represent mass activity (I, currents normalized by the catalyst loading) and the fraction of one electron capacity expressed in percentage, respectively, vs. potential. The potentials are reported vs. mercury-mercury oxide reference electrode (Hg/HgO). Reprinted with permission from [41]..................................................................................................................... 27 Figure 1.15 Linear voltammograms of MnO2-catalyzed air electrode in: A) Argon and B) Air. The catalysts are labelled as follows: 1) α-MnO2, 2) β-MnO2, 3) γ-MnO2, 4) λ-MnO2 and 5) δ-MnO2. The reduction current is defined as positive (American current polarity convention). 6 M KOH. xvii  298 K. 1 mV s-1. Currents are normalized by the catalyst loading, presenting mass activity values. The potentials are reported vs. mercury-mercury oxide reference electrode (Hg/HgO). Reprinted with permission from [48]. ........................................................................................................... 28 Figure 1.16 Oxygen reduction polarization curves recorded in a flow cell for four types of commercially available MnO2 powders: Tronox (γ-MnO2), Riedel (β-MnO2), Merck (γ-MnO2) and Sigma (γ-MnO2 with high portion of α-MnO2). Gas diffusion electrodes are employed at a loading of 0.25 mg cm-2 for MnO2 mixed with Vulcan XC-72 (1:1 weight ratio). 6 M KOH. 293 K. Currents are normalized to the geometrical area of the electrode. The potentials are reported vs. mercury-mercury oxide reference electrode (MOE).  Reprinted with permission from [90]. .................... 29 Figure 1.17 OER mechanisms proposed for crystalline oxide surfaces in alkaline electrolytes: A) Four-step reaction mechanism proposed by Rossmeisl et al. for the OER on noble metal catalysts and oxide surfaces [124, 125], B) Four-step reaction mechanism proposed by Goodenough et al. for the OER on perovskite surfaces [126], C) Acid–base mechanism proposed for first-row transition-metal oxides [127], D) Reaction mechanism proposed by Faria et al. involving recombination of oxygen atoms to produce O2 [128] and E) Reaction mechanism proposed by Gerken et al. for electrodeposited cobalt oxides [129]. The orange and blue denote species on the catalyst surface and in solution, respectively.  Reprinted with permission from [123]. ............... 32 Figure 1.18 A comparison between the OER activity and durability of wide range of catalysts in 1 M NaOH: The x-axis is the overpotential required to achieve 10 mA cmgeo−2 at time equal to 0 s. The y-axis is the overpotential required to achieve 10 mA cmgeo−2 after two hrs of testing. The dashed diagonal line indicates where the stable catalysts would be. Catalyst loadings are 0.8 mg cm-2 for each case. 298 K. PO₂ of 1 atm. Reprinted with permission from [130]. ......................... 33 xviii  Figure 1.19 A comparison between the OER overpotentials (at 10 mA cm-2) for four different types of manganese oxides in alkaline media: α, β, δ-MnO2 and amorphous MnOx. 0.1 M KOH. 1600 rpm. 298 K. Reprinted with permission from [105]. .................................................................... 34 Figure 1.20 OER cyclic voltammograms of various catalysts in O2 saturated 0.1 M KOH: As-deposited 500 ALD MnO, annealed Mn2O3, glassy carbon, 20 wt% Ru/C and 20 wt% Pt/C. 20 mV s-1. 1600 rpm. 298 K. Reprinted with permission from [67].................................................. 35 Figure 1.21 A) Free-energy diagram for ORR (in reverse direction) and OER on an ideal catalyst. The vertical solid arrow shows ΔGHOO(ads)-ΔGHO(ads) on a perfect crystal which is 2.46 eV. B) Adsorption energy of HOO(ads) plotted against the adsorption energy of HO(ads) on the clean surfaces: Perovskites (○), rutiles (▵), MnxOy (□), TiO2 (◊) and Co3O4 (+). The best fit of all points is ΔEHOO(ads) = ΔEHO(ads) + 3.20 eV. The red star indicates where the binding energies need to be for an ideal catalyst. Reprinted with permissions from [121, 124]. .............................................. 37 Figure 1.22 Cyclic voltammetry on platinum working electrode in 2 M H2SO4, with varying Mn2+ concentrations of 0.018 M to 0.73 M at 308 K and 1 mV s-1 [156]. The potentials are reported vs. mercury-mercurous sulfate reference electrode (MSE). Reprinted with permission from [156]. 43 Figure 1.23 SEM plan-view and cross-sectional images of manganese oxide deposits synthesized from: A) 3 mM Mn(CH3COO)2 solution at 0.25 mA cm-2, B) 5 mM, C) 7 mM, D) 10 mM, E) 20 mM and F) 30 mM Mn(CH3COO)2 solutions at 5 mA cm-2. 10 min per each deposition. 373 K and pH of 7.5. Reprinted with permission from [155]. ....................................................................... 44 Figure 1.24 Schematic diagram showing the morphological evolution of electrodeposited manganese oxides (from left to right: thin sheets, rods, aggregated rods and non-uniform continuous coating) with increasing the Mn2+ concentration during anodic electrodeposition process. Reprinted with permission from [155]. ........................................................................... 45 xix  Figure 1.25 XRD patterns of MnOx electrodeposited at different anodic potentials (0.5 to 0.95 VSCE) in 0.25 M manganese acetate solution at 298 K. Arrowed peaks (at 2θ = 37.1° and 66.3°) correspond to the oxides formed on the carbon substrates. Potentials are versus saturated calomel reference electrode (SCE). Reprinted with permission from [158]. ............................................. 46 Figure 1.26 SEM micrographs showing the surface morphologies of manganese oxides electrodeposited at A) 0.5, B) 0.65, C) 0.8 and D) 0.95 VSCE characterized in Figure 1.25. Reprinted with permission from  [158]. ........................................................................................................ 47 Figure 1.27 Levich plots calculated from anodic electrodeposition of manganese dioxide in 0.1 M MnSO4+5 M H2SO4 at different temperatures. Reprinted with permission from [86]. ................ 48 Figure 1.28 SEM plan-view and cross-sectional images of manganese oxides deposited from 10 mM Mn(CH3COO)2 at 5 mA cm-2 for 10 min and pH of 7.5: A) 298 K, B) 333 K, C) 358 K. Reprinted with permission from [155]. ......................................................................................... 49 Figure 1.29 Surfactant classifications based on the charge of head group. Reprinted with permission from [169]................................................................................................................... 50 Figure 1.30 Molecular structure of: A) Cationic cetyltrimethylammonium bromide (CTAB), B) Anionic sodium n-dodecylbenzenesulfonate (SDBS) and C) Non-ionic t-octyl phenoxy polyethoxyethanol (Triton X-100). Reprinted with permission from [161]. ................................ 51 Figure 1.31 FESEM images of the manganese oxide (EMD) samples in the presence of various concentrations of SDS (in ppm) as anionic surfactants. Reprinted with permission from [162]. 52 Figure 1.32 SEM images of MnO2 thin film prepared in: A) Absence and B) presence of Triton X-100. Reprinted with permission from [172]. ............................................................................ 53 Figure 1.33 SEM images of electrodeposited manganese dioxide in presence of : a) 5.1 mM Triton X-100 and 9 mM CTAB. Reprinted with permission from [161]. ............................................... 54 xx  Figure 1.34 Surfactant aggregation to form micelle at critical micelle concentration (CMC). Reprinted with permission from [175]. ......................................................................................... 55 Figure 1.35 Geometrical shapes of surfactant micelles in aqueous solutions. Reprinted with permission from [176]................................................................................................................... 55 Figure 1.36 Schematic representation of perovskite-type oxides with the general formula of ABO3 where A sites includes rare-earth metal ions while B sites are transition-metal ions. Reprinted with permission from [184]................................................................................................................... 56 Figure 1.37 Volcano-type graphs showing the comparison between the electrocatalytic activity of various perovskite-type oxides for: A) ORR and B) OER. Figures reveal the ORR/OER overpotentials at 50 μA cm-2 in alkaline media as a function of eg electron occupancy at 298 K. Reprinted with permission from [123]. ......................................................................................... 58 Figure 1.38 Schematic representation of Nd3IrO7 with an orthorhombic crystal structure (space group Cmcm). Reprinted with permission from [195, 196]. ......................................................... 59 Figure 2.1 CNC controlled sprayer machine with IWATA air brusher (50 ml capacity). ........... 68 Figure 2.2 Gas diffusion electrodes consist of catalyst inks sprayed on: 1) 40 wt% PTFE treated carbon cloth from Fuel Cell Earth Co. and 2) PTFE treated carbon substrate supported on a Ni mesh (as current collector) from ZincNyx Energy Solution Inc.. ................................................. 69 Figure 2.3 Schematic diagram of three-electrode electrochemical half-cell setup used in this study........................................................................................................................................................ 72 Figure 2.4 Components of a quick-fit exchangeable sample holder (left) from radio Radiometer Analytical (#A35T450) with sample opening of 6 mm in diameter. The samples (right) were placed on a glassy carbon disk as a backing layer prior to be placed on the tip. ..................................... 72 xxi  Figure 2.5 Computer-controlled VoltaLab 80 potentiostat with its associated RDE setup and the electrochemical three-electrode cell. ............................................................................................ 73 Figure 2.6 A Pine jacketed electrochemical cell connected to a water bath in the three-electrode RDE half-cell test setup used for electrochemical measurements in this study. .......................... 76 Figure 2.7 Flow-by electrochemical test cell (HZ-PP01) and its components from Gaskatel GmbH used for ORR GDE tests: PTFE body, built-in platinum counter electrode, built-in Luggin-Haber capillary for the reference electrode. ............................................................................................ 79 Figure 3.1 SEM images of A) Commercial Sigma-Aldrich MnO2 powder (pillar and sphere-like particles are shown using dashed oval and circle shapes, respectively), B) Synthesized LaCoO3 powder, C) Synthesized LaNiO3 powder and D) Synthesized Nd3IrO7 powder. The black arrow in B and C points out the flakes in SEM morphology of studied oxides. ......................................... 82 Figure 3.2 XRD spectra of: A) Commercial MnO2, B) Synthesized LaCoO3, C) Synthesized LaNiO3 and D) Synthesized Nd3IrO7. (*), (▼), ( ▌) and (♦) present major peaks corresponding to MnO2, LaCoO3, LaNiO3 and Nd3IrO7, respectively. .................................................................... 84 Figure 3.3 EDX spectra of commercial and synthesized catalyst powders: A) MnO2, B) LaCoO3, C) LaNiO3 and D) Nd3IrO7. .......................................................................................................... 86 Figure 3.4 IR-corrected cyclic voltammograms of GDEs with MnO2, LaCoO3, LaNiO3, Nd3IrO7, MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7 catalysts. A) MnO2, LaCoO3, LaNiO3 and Nd3IrO7, B) MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7. Electrolyte: N2 saturated 6 M KOH at 293 K and PN₂ of 1 atm. The oxide loadings were 0.5 mg cm-2 each. Rotating electrode speed and potential scan rate were 400 rpm and 5 mV s-1, respectively. Cycle number five is reported in all cases. ...................................................................................................................... 88 xxii  Figure 3.5 SEM images of the GDE consisting of MnO2:LaCoO3:Vulcan XC-72:Nafion:PTFE (weight ratio of 1:1:1:0.6:0.6) sprayed on 40% PTFE treated carbon cloth. ................................ 89 Figure 3.6 Initial stage IR-corrected bifunctional ORR/OER Tafel-lines of GDEs with MnO2, LaCoO3, LaNiO3, Nd3IrO7, MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7 catalysts. A) ORR, B) OER. Electrolyte: O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm. Electrode potential scanning between 233 to 1683 mV. The oxide loadings were 0.5 mg cm-2 each. Rotating electrode speed and potential scan rate were 400 rpm and 5 mV s−1, respectively. Cycle number five is reported in all cases. The numbers associated with each line represent the respective apparent Tafel slopes............................................................................................................................................. 94 Figure 3.7 Electrocatalytic durability testing of GDEs with MnO2, MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7 catalysts: A) ORR at cycle one, B) OER at cycle one, C) ORR at cycle one hundred, D) OER at cycle one hundred. IR-corrected ORR/OER polarization curves obtained by potential scanning between 633 to 1633 mV for one hundred cycles (accelerated degradation testing) in O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm. Other conditions idem to Figure 3.6........................................................................................................................................................ 97 Figure 3.8 Long-term ORR durability testing of fresh GDEs containing Pt, MnO2, MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7 catalysts compared with commercial MnOx GDE from Gaskatel GmbH: A) With air at -40 mA cm-2 and B) With oxygen at -100 mA cm-2. IR-corrected galvanostatic polarization curves obtained in 11.7 M (45 wt%) KOH at 323 K for 24 hrs with either air (CO2 removed) or oxygen flowing through the gas chamber of a flow-by cell from Gaskatel GmbH after 24 hrs of break-in protocol explained in section 2.3 and Appendix C  . The absolute gas pressure and flow rate were fixed at 1 atm and 1.51×10-3 SLPM. The catalyst(s) loadings were 2 mg cm-2 each (except for Pt with 0.5 mg cm-2) with final weight ratio of 1:1:2:0.6:0.6 for MnO2 xxiii  or Pt:co-catalyst (if present):Vulcan XC-72:Nafion:PTFE in the catalyst layer. The catalyst loading for the commercial MnOx from Gaskatel was 20 mg cm-2. The standard error of the mean calculated based on min. two replicates is indicated for each data point. .................................. 102 Figure 3.9 Comparison between the XRD spectrum of the MnO2 catalyst: 1) Fresh electrode (black dotted line) and 2) After 48 hrs of galvanostatic tests (24 hrs at -67 mA cm-2 followed by 24 hrs at -100 mA cm-2) with O2 as feed gas in an unflooded flow-by test setup. ................................ 103 Figure 4.1 The effect of rest-time at open-circuit potential in 6 M KOH at 293 K following accelerated degradation testing in the flooded test setup (i.e. one hundred potential cycles between 633 to 1633 mV in O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm): A) ORR polarization curve for MnO2, B) OER polarization curve for MnO2, C) ORR polarization curve for MnO2-LaCoO3 and D) OER polarization curve for MnO2-LaCoO3. IR-corrected ORR/OER polarization curves obtained by potential scanning between 633 to 1683 mV in O2 saturated 6 M KOH at 400 rpm, 293 K and PO₂ of 1 atm. Other conditions idem to Figure 3.6. ................................................... 109 Figure 4.2 The effect of rest-time at open-circuit potential in 6 M KOH at 293 K following accelerated degradation testing in the flooded test setup (i.e. one hundred potential cycles between 633 to 1633 mV in O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm): A) ORR polarization curve for MnO2-LaNiO3, B) OER polarization curve for MnO2-LaNiO3, C) ORR polarization curve for MnO2-Nd3IrO7 and D) OER polarization curve for MnO2-Nd3IrO7. IR-corrected ORR/OER polarization curves obtained by potential scanning between 633 to 1683 mV in O2 saturated 6 M KOH at 400 rpm, 293 K and PO₂ of 1 atm. Other conditions idem to Figure 3.6. ...................... 110 Figure 4.3 Comparison between the XRD spectrum of the MnO2-LaCoO3 catalyst: 1) Fresh GDE (black dotted line) and 2) After accelerated degradation testing for one hundred cycles in the xxiv  potential range of 633 to 1633 mV followed by resting for six days at open-circuit in 6 M KOH at 293 K (red solid line) in a flooded test setup. ............................................................................. 113 Figure 4.4 XPS spectra of MnO2 GDE: A) Fresh, C) After accelerated degradation testing for one hundred cycles in the potential range of 633 to 1633 mV and rested for six days at open-circuit in 6 M KOH at 293 K...................................................................................................................... 115 Figure 4.5 Bifunctional activation effect of long-term (i.e. six days) exposure of MnO2-LaCoO3 to alkali-metal hydroxide solutions: LiOH, NaOH, KOH, CsOH. Initial stage IR-corrected polarization curves obtained by potential scanning between 633 to 1483 mV in O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm. Other conditions idem to Figure 3.6. ...................................... 118 Figure 4.6 Electrocatalytic durability testing of fresh activated MnO2-LaCoO3 and MnO2-Nd3IrO7 GDEs: A) ORR polarization curves and B) OER polarization curves. IR-corrected ORR/OER polarization curves obtained by potential scanning between 633 to 1483 mV for one hundred cycles (accelerated degradation testing) in O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm. Catalysts were activated by six-day-long exposure to 6 M KOH at 313 K and 400 rpm and open-circuit. Other conditions idem to Figure 3.6. .............................................................................. 121 Figure 4.7 XPS spectra of MnO2-LaCoO3 and MnO2-Nd3IrO7 before and after long-term exposure to 6 M KOH: A) MnO2-LaCoO3 fresh electrode, B) MnO2-LaCoO3 activated electrode, C) MnO2-Nd3IrO7 fresh electrode, D) MnO2-Nd3IrO7 electrode. Other conditions idem to Figure 4.6. ... 123 Figure 4.8 EDX spectra of the activated catalyst layers after six days of exposure to 6 M KOH: A) MnO2-LaCoO3, B) MnO2-Nd3IrO7. Conditions idem to Figure 4.6. .......................................... 124 Figure 4.9 EELS analysis of MnO2-LaCoO3: A) EELS spectrum showing Mn (L2,3) edges, B) EELS spectrum showing O (K) edges, C) L3:L2 Branching ratio versus valence state for Mn oxides. The sample was tested under three different conditions: 1) Initial stage, 2) After being activated in xxv  KOH for six days (idem to Figure 4.6) and cycled between the ORR and OER regions for ten cycles, 3) After being activated in KOH for six days and cycled between the ORR and OER regions for one hundred cycles (accelerated degradation test idem to Figure 4.6). The (■) and (▲) symbols represent reference data points obtained from literature [228] and the ones calculated directly from EELS spectrum for each sample, respectively. The standard error of the mean associated with Mn valence state, L3:L2 branching ratio and energy loss are ± 0.2, ± 0.001 and ± 0.1 eV, respectively...................................................................................................................................................... 128 Figure 4.10 Cell potential profile during potential driven K+ intercalation (PDI) on A) MnO2-LaCoO3 and B) MnO2-Nd3IrO7 as catalyst layers. 5.4 mA cm-2 was cathodically applied to each GDE in the RDE setup in a 0.036 M K2SO4 solution for 30 min up to seven times. The rotation speed and temperature were 400 rpm and 343 K, respectively. A platinum mesh was employed as both reference and counter electrode. The figure indicates the cell potential for the third round of PDI in each case. ......................................................................................................................... 130 Figure 4.11 XPS spectra of the potential driven K+ intercalation (PDI) activated catalyst layers: A) MnO2-LaCoO3 electrode after seven rounds of PDI activation, B) MnO2-Nd3IrO7 electrode after six rounds of PDI activation. ...................................................................................................... 131 Figure 4.12 The effect of potential driven K+ intercalation on the initial stage bifunctional polarization of mixed-oxide catalysts: A) ORR on MnO2-LaCoO3, B) OER on MnO2-LaCoO3, C) ORR on MnO2-Nd3IrO7 and D) OER on MnO2-Nd3IrO7. Other conditions idem to Figure 3.6...................................................................................................................................................... 133 Figure 4.13 The effect of potential driven K+ intercalation on the initial stage bifunctional polarization of individual oxides: MnO2, LaCoO3 and Nd3IrO7. A) ORR, B) OER. Other conditions idem to Figure 3.6. ...................................................................................................................... 136 xxvi  Figure 4.14 XPS spectra of the PDI activated catalyst layers: A) MnO2 after five rounds of K+ PDI activation, B) LaCoO3 after five rounds of K+ PDI activation and C) Nd3IrO7 after five rounds of K+ PDI activation. ....................................................................................................................... 137 Figure 4.15 Galvanostatic polarization of mixed-oxide catalysts without and with potential driven K+ intercalation activation: A) MnO2-LaCoO3 and B) MnO2-Nd3IrO7. Tests started with 5 mA cm-2 anodically applied to each GDE for 2 hrs followed by -2 mA cm-2 applied cathodically for 30 min in O2 saturated 6 M KOH using a flooded test setup. The rotation speed and temperature were 400 rpm and 293 K, respectively. PO₂ was 1 atm. The oxide loadings were 0.5 mg cm-2 each. . 140 Figure 4.16 Comparison between the ORR and OER overpotential values of the catalyst materials investigated here (shown as (▲)) with those reported in the literature for other bifunctional electrodes (shown as (♦)) [4, 64, 67, 69, 71, 72, 74, 76, 121, 259, 260]. For the catalyst investigated here: a) Fresh catalyst without activation. b) Activation by K+ insertion at open-circuit potential (OCP), c) Activation by K+ insertion using five rounds of potential driven intercalation (PDI), d) Activation by K+ insertion using six rounds of potential driven intercalation (PDI), e) Activation by K+ insertion using three rounds of PDI. The max error associated with overpotential values is ±5 mV. ........................................................................................................................................ 143 Figure 5.1 IR-corrected linear sweep voltammograms of nitric acid pre-treated 40 wt% PTFE treated carbon cloth starting from 0 to 2500 mVMOE in presence of: A) 5 vol% and B) 10 vol% of Triton X-100, SDS and CTAB. The solution was made of 0.2 M Mn(CH3COO)2 and 0.1 M Na2SO4 at 293 K. The scan rate and rotation speed were 5 mV s-1 and 400 rpm, respectively. .............. 149 Figure 5.2 XPS spectra of three representative electrodeposited MnOx samples at Mn 2p, Mn 3s and O 1s regions. The electrodeposition factors for each sample are as follows: T1 (C: 0.3 M, T: xxvii  295 K, S: Triton, 10 vol%, E: 800 mVMOE), T9 (C: 0.1 M, T: 343 K, S: Triton, 10 vol%, E: 800 mVMOE), T10 (C: 0.3 M, T: 295 K, S: Triton, 0 vol%, E: 1600 mVMOE). .................................. 151 Figure 5.3 FT-IR spectra of MnOx samples (after IPA washing) electrodeposited on the pre-treated carbon cloth as substrate in presence of Triton X-100, SDS and CTAB. The electrodeposition factors for each sample are as follows: Carbon cloth (no electrodeposited material), Triton (C: 0.1 M, T: 343 K, S: 10 vol%, E: 800 mVMOE), SDS (C: 0.3 M, T: 295 K, S: 10 vol%, E: 800 mVMOE), CTAB (C: 0.3 M, T: 343 K, S: 10 vol%, E: 1600 mVMOE). ....................................................... 154 Figure 5.4 Surface plots for 24-1+3 factorial design in presence of Triton X-100 including factorial responses with the most important factors and two-factor interactions based on the Pareto plots of estimates: A) ORR mass activity at 656 mV, B) OER mass activity at 1556 mV and C) ORR/OER potential window at -2 and 2 mA cm-2, respectively. Details of each run has been given in Table 5.2. Red and green colors in the surface plots correspond to highest and lowest values of each response, respectively. ................................................................................................................ 159 Figure 5.5 Surface plots for 24-1+3 factorial design in presence of SDS including factorial responses with the most important factors and two-factor interactions based on the Pareto plots of estimates: A) ORR mass activity at 656 mV, B) OER mass activity at 1556 mV and C) ORR/OER potential window at -2 and 2 mA cm-2, respectively. Details of each run has been given in Table 5.3. Red and green colors in the surface plots correspond to highest and lowest values of each response, respectively. ................................................................................................................................ 163 Figure 5.6 Surface plots for 24-1+3 factorial design in presence of CTAB including factorial responses with the most important factors and two-factor interactions based on the Pareto plots of estimates: A) ORR mass activity at 656 mV, B) OER mass activity at 1556 mV and C) ORR/OER potential window at -2 and 2 mA cm-2, respectively. Details of each run has been given in Table xxviii  5.4. Red and green colors in the surface plots correspond to highest and lowest values of each response, respectively. ................................................................................................................ 167 Figure 5.7 IR-corrected bifunctional performance comparison of electrodeposited MnOx in presence of different surfactants, i.e. Triton X-100, SDS and CTAB: A) ORR, B) OER. The electrodeposition factors for each sample are as follows: Carbon cloth substrate (no electrodeposited material), CTAB-Run no. 6 (C: 0.1 M, T: 343 K, S: 10 vol%, E: 800 mVMOE), SDS-Run no. 1 (C: 0.3 M, T: 295 K, S: 10 vol%, E: 800 mVMOE), Triton-Run no. 9 (C: 0.1 M, T: 343 K, S: 10 vol%, E: 800 mVMOE). Electrolyte: O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm. Rotating electrode speed and potential scan rate were 400 rpm and 5 mV s-1, respectively. Cycle number five is reported in all cases............................................................................................. 171 Figure 5.8 SEM micrographs of best performing electrodeposited MnOx oxides on nitric acid pre-treated carbon cloth: A and B) Triton run no. 9, C and D) SDS run no. 1, E and F) CTAB run no. 6. The electrodeposition factors are stated in Figure 5.7. ........................................................... 172 Figure 5.9 Galvanostatic polarization comparison of in-house made and commercial manganese oxide GDEs: I) Manganese oxide/C GDE from Gaskatel GmbH (20 mg cm-2), II) γ-MnO2/C from Sigma Aldrich (loading 5.6 mg cm-2) and III) Triton run no. 9 (C: 0.1 M, T: 343 K, S: 10 vol%, E: 800 mVMOE) (calculated loading 17.5 mg cm-2). The galvanostatic polarization started at 5 mA cm-2 anodically applied to each GDE for 2 hrs followed by -2 mA cm-2 applied cathodically for 30 min in O2 saturated 6 M KOH. The rotation speed and temperature were 400 rpm and 293 K, respectively. PO₂ was 1 atm. ........................................................................................................ 174 Figure 6.1 TEM images of materials investigated in this chapter as oxygen catalyst or catalyst support: A) Graphene, B) N-doped graphene, C) Vulcan XC-72, D) MWCNT, E) MnO2 and F) LaCoO3. ...................................................................................................................................... 180 xxix  Figure 6.2 Rotating ring disk electrode results for ORR on Graphene (I), N-doped Graphene (II), MWCNT (III) and Vulcan XC-72 (IV): A) O2 reduction current densities obtained from disk electrode (id) when polarized from 1170 to 500 mV (bottom) and the corresponding oxidation current densities on the ring at 1353 mV (ir) as a function of disc potentials (top). B) Calculated percentage of hydrogen peroxide ions produced during ORR (%HO2−). C) Calculated number of electrons transferred per molecule of oxygen during ORR (n). Carbon loadings of 0.5 mg cm-2 each. O2 saturated 6 M KOH. 5 mV s-1. 293 K. PO₂ of 1 atm. .................................................... 183 Figure 6.3 Bifunctional ORR/OER performance of carbonaceous materials, i.e. Graphene, N-doped graphene, MWCNT, Vulcan XC-72, MWCNT-Graphene and MWCNT-N-doped graphene: A) ORR, B) OER. Initial stage IR-corrected polarization curves obtained by potential scanning between 475 to 1673 mV, starting with anodic polarization. Carbon loadings are fixed at 0.5 mg cm-2, each. In cases where two materials were mixed, a weight ratio of 1:1 was used. Cycle number five is reported in all cases. O2 saturated 6 M KOH. 293 K. 400 rpm. 5 mV s-1. PO₂ of 1 atm. . 186 Figure 6.4 IR-corrected bifunctional ORR/OER Tafel-lines of investigated carbonaceous catalyst materials: A) ORR, B) OER. The numbers associated with each line represent the respective apparent Tafel slopes. Cycle number five is reported in all cases. The N2-baseline is subtracted from the ORR part for all samples. For the OER part, capacitive background current at open circuit potential is removed from each corresponding voltammogram. Other conditions idem to Figure 6.3................................................................................................................................................ 188 Figure 6.5 Bifunctional ORR/OER performance of MnO2-LaCoO3 catalyst supported on various carbonaceous materials: A) Graphene and N-doped graphene, B) MWCNT and Vulcan XC-72, C) MWCNT-Graphene and MWCNT-N-doped graphene. The 50 wt% Pt/Graphitized carbon is being shown as the baseline for comparison (Pt loading of 0.25 mg cm-2). The oxide and carbon loadings xxx  were 0.5 mg cm-2 each. A weight ratio of 1:1:1:1 for MnO2:LaCoO3:Carbon1:Carbon2 (if available) was used. Cycle number five is reported in all cases.  Other conditions idem to Figure 6.3................................................................................................................................................ 191 Figure 6.6 IR-corrected bifunctional ORR/OER Tafel-lines of investigated MnO2-LaCoO3 catalyst supported on various carbonaceous materials studied in Figure 6.5 (50 wt% Pt/Graphitized carbon as the baseline): A) ORR, B) OER. The numbers associated with each line represent the respective apparent Tafel slopes. Cycle number five is reported in all cases. The N2-baseline is subtracted from the ORR part for all samples. For the OER part, capacitive background current at open circuit potential is removed from each corresponding voltammogram. Other conditions idem to Figure 6.5................................................................................................................................................ 195 Figure 6.7 Electrocatalytic durability testing: ORR (bottom) and OER (top) current densities at 800 mV and 1750 mV, respectively, during one hundred successive potential cycling between 673 to 1873 mV on MnO2-LaCoO3-MWCNT-Graphene (weight ratio 1:1:1:1) and MnO2-LaCoO3-Vulcan XC-72 (weight ratio 1:1:1) GDEs in flooded test setup. The oxide and carbon loadings were kept at 0.5 mg cm-2, each, spayed on a 40 wt% PTFE treated carbon cloth as porous substrate. The error associated with ORR and OER current densities was found to be between 2% and 10%. Other conditions idem to Figure 6.3. .......................................................................................... 199 Figure A.1 Effect of the symmetry factor (β) on the symmetry of the current-overpotential curve described by the BE-GV equation (eq. 40). ................................................................................ 230 Figure B.1 Typical cathodic polarization curves obtained by a RDE as a function of angular velocity (ω). ω1 < ω2 < ω3. .......................................................................................................... 231 Figure C.1 Long-term ORR durability testing of fresh GDEs containing Pt, MnO2, MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7 catalysts compared with commercial MnOx GDE from xxxi  Gaskatel GmbH: A) With air at -33 mA cm-2 and B) With oxygen at -67 mA cm-2. IR-corrected galvanostatic polarization curves obtained in 11.7 M (45 wt%) KOH at 323 K for 24 hrs with either air (CO2 removed) or oxygen flowing through the gas chamber of a flow-by cell from Gaskatel GmbH as twenty-four-hour-long break-in protocol. The absolute gas pressure and flow rate were fixed at 1 atm and 1.51×10-3 SLPM. The catalyst(s) loadings were 2 mg cm-2 each (except for Pt with 0.5 mg cm-2) with final weight ratio of 1:1:2:0.6:0.6 for MnO2 or Pt:co-catalyst (if present):Vulcan XC-72:Nafion:PTFE in the catalyst layer. The catalyst loading for the commercial MnOx from Gaskatel was 20 mg cm-2. The standard error of the mean calculated based on min. two replicates is indicated for each data point. ........................................................................... 236 Figure D.1 The effect of inter-stage OCP activation: ORR and OER polarization curves of the activated MnO2-LaCoO3 electrodes tested for three hundred cycles with OCP activation in between, A) ORR polarization curves, B) OER polarization curves. The samples are kept at OCP in 6 M KOH solution for 12 hrs at 293 K and a rotation speed of 400 rpm after each one hundred cycles of durability testing. Other conditions are same as Figure 3.6. ....................................... 239 Figure E.1 Pareto plot of estimates for the effects of four main factors (i.e. surfactant concentration (S), temperature (T), Mn concentration (C) and applied anodic potential (E)) and three aliased two-factor interactions on the ORR mass activity in the 24-1+3 factorial design study on the anodic electrodeposition of MnOx in presence of Triton X-100. ..................................... 242  xxxii  Nomenclature Symbol Definition Unit 𝑎 Tafel parameter mV 𝑎𝑗 Activity of species j - 𝑏𝑂𝐸𝑅 OER apparent Tafel slope mV decade-1 𝑏𝑂𝑅𝑅 ORR apparent Tafel slope mV decade-1 𝐶 Mn2+ concentration mol L-1 𝐶𝑗 Bulk reactant concentration of species j mol L-1 𝐷𝑗  Diffusion coefficient of the electroactive species j m2 s-1 𝐸 Potential VMOE, VRHE 𝐸298𝐾0  Standard equilibrium potential at 298 K VSHE 𝐸𝑒 Equilibrium electrode potential VMOE, VRHE 𝐸0 Standard equilibrium potential  VSHE 𝐸𝑏 Binding energy eV 𝐹 Faraday constant (96500) C mol-1 𝑖 Measured current density mA cm-2 𝑖𝑑 Disk current density in a RRDE test setup mA cm-2 𝑖𝑘 Pure electrode kinetic current density mA cm-2 𝑖𝐿 Limiting current density mA cm-2 𝑖0,𝑂𝐸𝑅 Apparent exchange current density calculated from the OER part μA cm-2 xxxiii  Symbol Definition Unit 𝑖0,𝑂𝑅𝑅 Apparent exchange current density calculated from the ORR part μA cm-2 𝑖𝑟 Ring current density in a RRDE test setup mA cm-2 𝐼 (𝐿3) Normalized intensity of Mn(L3) edge in a Mn EELS spectrum - 𝐼 (𝐿2) Normalized intensity of Mn(L2) edge in a Mn EELS spectrum - 𝐾0 Standard heterogeneous rate constant  mol m-2 s-1 𝐾𝑚,𝑗 Mass transfer coefficient of species j m s-1 𝑀 Molarity mol L-1 𝑚𝑗 A power expressing the concentration/pressure dependence of the activity for species j - 𝑛 Electron stoichiometry coefficient - 𝑁 Collection efficiency of the ring in a RRDE test setup - 𝑂𝑂𝑥 An oxygen ion sitting on an oxygen lattice site with neutral charge in the Kröger-Vink notation - 𝑃𝑂2 Absolute partial pressure of oxygen atm 𝑃𝑔𝑎𝑠 Absolute gas pressure atm 𝑅 Gas constant (8.314) J mol-1 K-1 𝑅𝑒 Reynolds number - 𝑆 Surfactant concentration vol% xxxiv  Symbol Definition Unit 𝑠𝑗 Stoichiometric coefficient of species j - 𝑇 Temperature K 𝑉𝑥/// Cation vacancy in the Schottky notation - 𝑉𝑂∙∙ Lattice oxygen vacancy with double positive charge in the Kröger-Vink notation - 𝛼 Transfer coefficient - 𝛼𝑎 Anodic transfer coefficient - 𝛼𝑐 Cathodic transfer coefficient - 𝛽 Symmetry factor - 𝛿 Diffusion layer thickness m 𝜂 Overpotential mV 2𝜃 Diffraction angle degree (°) 𝜈 Kinematic viscosity of the electrolyte m2 s-1 𝜛 Angular velocity radian s-1 𝛱𝑗 Multiplication operator symbol -  xxxv  List of Abbreviations Abbreviation Definition ARFCs Alkaline Regenerative Fuel Cells  BET Brunauer-Emmet-Teller CMC Critical Micelle Concentration CNC Computer Numerical Control CV Cyclic Voltammetry ECSA Electrochemically Active Surface Area  EDX Energy-Dispersive X-ray Spectroscopy EELS Electron Energy Loss Spectroscopy FESEM Field Emission Scanning Electron Microscopy FT-IR ATR Fourier Transfer-Infrared Spectroscopy with Attenuated Total Reflectance FWHM Full Width at Half Maximum GDE Gas Diffusion Electrode GNS Graphene Nano Sheet IPA Iso-Propyl Alcohol IR-corrected Ohmic Drop Compensated LSV Linear Sweep Voltammetry MOE Mercury-Mercury Oxide Reference Electrode xxxvi  Abbreviation Definition MSE Mercury-Mercurous Sulfate Reference Electrode MWCNT Multi-Walled Carbon Nanotubes Non-PGM Non-Precious Group Metal OCP Open-Circuit Potential OER Oxygen Evolution Reaction ORR Oxygen Reduction Reaction PDI Potential Driven Intercalation PGM Precious Group Metal PTFE Poly-Tetra-Fluoro-Ethylene RDE Rotating Disk Electrode RHE Reversible Hydrogen Reference Electrode RRDE Rotating Ring Disk Electrode RTIL Room Temperature Ionic Liquid SCE Saturated Calomel Reference Electrode SEM Scanning Electron Microscopy SHE Standard Hydrogen Electrode SLPM Standard Liter Per Minute TEM Transmission Electron Microscopy URFC Unitised Regenerative Fuel Cell Wt% Weight Percentage xxxvii  Abbreviation Definition XANES X-ray Absorption Near Edge Structure XPS X-ray Photoelectron Spectroscopy XRD X-Ray Diffraction Vol% Volume Percentage VPSEM Variable Pressure Scanning Electron Microscopy  xxxviii  Acknowledgements Let me start with saying that I had an amazing experience at the University of British Columbia and will always be grateful for the opportunity to pursue my Ph.D. studies here in Vancouver.  This is going to be a quite different acknowledgment, neither formal nor informal. I have spent many hours, nights and days to come up with something that truly express my appreciation to all the people who have helped and empowered me to be who I am right now. But who am I kidding? I am neither a writer nor a poet. I am just a person who thinks everything in a scientific manner. It is such a strange feeling for me to write this section at this particular moment. One part of me wants to rush it while the other part tells me to put down every single word carefully from the bottom of my heart. I can say the nostalgic song that I am listening to right now will have a huge effect on the following paragraphs. I am digging deeper and deeper into my memory and soul, asking myself where to start; The moment my aunt called me “Mr. Doctor” for the first time, my indescribable feelings of joy and satisfaction which I have always experienced during my many years at school, my mother who I always adored for being a loving and patient physician, my father who is always proud of me, inspired me to pursue my passion and encouraged me to invent something useful, the laughs and support of my lovely sister, the moment I almost wanted to quit in the middle of the university entrance exam, the unsuccessful attempts to pursue my M.Sc. studies in Canada which almost crushed me, the moment I saw Maryam and something clicked in my heart, my first publication in the Journal of Power Sources, the very moment Akram replied to my email with an admission to UBC, the moment I had my study permit in one hand and Maryam’s rejected visa in the other, the awkward moment I left my hometown for Vancouver almost five years ago, my first meeting with xxxix  Elod at CHBE, shaking as I was excited and frightened at the same time or the relentless one-and-half-year-long wait till my precious wife could finally join me in Canada (Oh, I was relieved). Like others, I have gone through many ups and downs in my life. Those experiences have helped me grow so very much both mentally and physically, especially during the past five years. So, now, let me start expressing a small portion of my appreciation to the people who have helped me along my journey: First and foremost, I would like to extend my sincere gratitude and respect to my advisors Professors Előd Gyenge and Akram Alfantazi for giving me this opportunity to flourish and pursue my passion in science and engineering, their guidance throughout this work as well as everlasting support and patience. Előd, I have been always inspired by your honest and intellectual personality as well as in-depth knowledge and wisdom. I have enjoyed every minute of our many fruitful discussions and your productive suggestions during my Ph.D. program. You have gone well beyond being a supervisor and mentor. I will forever be indebted to you for your unconditional support, encouragement, nurturing mentorship and friendship. I am also grateful to my Ph.D. supervisory committee members, Professor Colin Oloman and Professor Fariborz Taghipour, for their invaluable help, sound advice and constructive comments throughout my Ph.D. studies. Moreover, I would like to specially thank all who have helped me with their knowledge and expertise during this work: Dr. Gianluigi Botton and Dr. Andreas Korinek from the Canadian Centre for Electron Microscopy at McMaster University for EELS analysis, Derrick Horne and Bradford Ross from the Bioimaging Facility at the University of British Columbia for FESEM and TEM trainings, Jacob Kabel from the Electron Microscopy Laboratory at the University of British Columbia for SEM analysis, Dr. Ken Wong from Interfacial Analysis & Reactivity Lab at the University of British Columbia for XPS analysis and Anita Lam xl  from the Department of Chemistry at the University of British Columbia for XRD analysis. I further extend my appreciation to CHBE community. My special thanks goes to Dr. Miguel Angel Garcia-Contreras, my dearest, amazing, knowledgeable friend and colleague from the National Institute of Nuclear Research at the United Mexican States who helped me through this journey and showed me the true meaning of passion to science, patience, honesty, hard-working and humbleness. My sincere appreciation is extended to all of my friends and colleagues for their support, joyful discussions and friendship: Amir, Brian, Amin, Andrew, Jeannette, Winton, Colin, Dustin, Lijun, Yan, Charanjit and Ivan. My true best friends from back home, Iran: Reza, Davood, Sadegh, Alireza and Masood; I cherish your irreplaceable friendship and lasting memories we had together. Davood and Reza: We are the “Three Musketeers”! I am so lucky to have such heartwarming and one-of-a-kind friends like you guys. You are like brothers to me. Reza, I wish you could be here in Canada with us right now. No proper words can describe my love and appreciation for my family. Maryam, my gorgeous wife, my soulmate, my love and my sweet partner in crime, thank you for being there through my ups and downs. You are the ultimate blessing in my life. I could have never done this without your support, patience, companionship and love. My parents, Zohreh and Hossein, whom I could have never been here without them. You enlighten my life. Far, far away yet I feel your warmth in my heart each and every day. My lovely sister, Parastoo, who always encouraged me to be a better person. Love you sis. My kind and precious in-laws, Robabeh, Reza, Marjan, Ata and last but never the least, my brother-in-law, Ehsan, who all brought joy to my life. And my childhood nanny, Maryam Khanoom, who loved and looked after me and my sister like her own children. I know you are watching us from the heaven. I am indebted to you for eternity, my lovely second mother. I am also grateful to my uncles, Mohammad and Mehdi, for being there in my times of needs. xli  Finally, I would like to acknowledge the University of British Columbia, Natural Sciences and Engineering Research Council of Canada (NSERC) and Electrochemical Society for giving me the opportunity to pursue my passion in science and engineering through their generous support, funding, travel grants and fellowships.   xlii  Dedication    To my beautiful wife, Maryam,  Your everlasting love and support sustains me. Your nurturing being completes me.   To my beloved parents and sister, Zohreh, Hossein and Parastoo,  Whose unceasing and unconditional love, compassion and thoughts empower me.  And, To my childhood nanny, Maryam Khanoom, who is in heaven now,  You are always in my heart. I miss you each day, every day, all the times.  1  Chapter 1: Introduction 1.1 Summary Oxygen electrochemistry, i.e. electrochemical oxygen reduction (ORR) and oxygen evolution reactions (OER), is known to be the heart of various electrochemical processes from renewable energy generation such as hydrogen-oxygen fuel cells, rechargeable metal-air and metal hydride-air batteries to chemical production technologies such as NaCl electrolysis along with hydrogen, hydrogen peroxide and chlorine production [1-6]. The development of a bifunctional catalyst for both ORR and OER reactions (cathodic and anodic directions, respectively, for eqs. 1 and 2) is of great interest due to their involvement in energy conversion systems such as regenerative fuel cells and rechargeable metal-air batteries, remaining an important challenge for electrochemists [2, 7, 8].   𝑂2 + 4𝐻+ + 4𝑒− ↔  2𝐻2𝑂   (𝐸298𝐾0 = 1.229 𝑉𝑆𝐻𝐸) (1) 𝑂2 + 2𝐻2𝑂 + 4𝑒− ↔ 4𝑂𝐻−   (𝐸298𝐾0 = 0.401 𝑉𝑆𝐻𝐸) (2)  Based on pH, two types of electrolytes, i.e. acid and alkaline, are currently used in fuel cells and batteries depending on the oxygen evolution/reduction reactions rates and the type of the metal catalyst. Alkaline electrolytes (such as 6-14 M KOH) are the primary choice for both ORR and OER in metal-air batteries such as Zn-air and Mg-air.  Metal-air batteries are being used in wide range of applications due to a combination of favorable properties such as low cost (Zn and Mg are the fourth and seventh most abundant metals in the earth’s crust, respectively [9, 10]), high safety due to non-flammable electrolyte and relatively high energy density as well as equilibrium cell voltage [10-13]. While the Zn-air battery 2  provides a standard equilibrium cell potential of 1.65 V (298 K) with theoretical energy density of 1330 Wh kg-1 *, Mg-air battery is characterized by a higher standard cell potential of 3.09 V (298 K) as well as high theoretical energy density of 6813 Wh kg-1 * [14]. However, the practical energy density of the existing metal-air batteries are low, e.g. 500 and 680 Wh kg-1 * for the Zn-air and Mg-air batteries, respectively [15, 16], limiting their commercial applications because of similar challenges, i.e. low utilization efficiency of the anode due to passivation and corrosion, sluggish kinetics of the cathode, sensitivity to contaminants from air (e.g. CO2) and electrolyte evaporation due to their open cathode structure [10, 13]. Electrically rechargeable metal-air batteries such as rechargeable Zn-air and Li-air operate based on the similar principle as of primary metal-air batteries except for the bifunctional cathode material which catalyzes both ORR and OER during discharging and charging, respectively, while the anode material is oxidized during discharge mode and reduced during the charge mode (Figure 1.1) [10, 11, 14, 16-18]. Figure 1.2 shows schematic polarization curve of a rechargeable Zn-air battery. Similar challenges to the primary metal-air batteries impede the commercialization of rechargeable Zn-air batteries such as low conversion rate of anode, i.e. Zn, dendrite formation on the anode and lack of active as well as durable bifunctional catalyst for both ORR and OER [11, 16, 19]. In 2007, Visco and Nimon introduced a protected anode for non-aqueous Li-air batteries which paved the way for development of rechargeable aqueous Li-air batteries [20, 21]. In theory, Li-air batteries could provide very high energy density of 11148 Wh kg-1 *, however, the practical energy density is much lower due to combination of challenges facing the aqueous Li-air batteries                                                  * Per kg of the metal. Calculated from the electrochemical equivalents of the metals and the cell electromotive forces based on the overall reaction in the battery. 3  such as low cycling efficiency, LiOH crystallization, CO2 contamination leading to Li2CO3 formation as well as lack of active and durable ORR/OER bifunctional catalyst (Figure 1.3) [11, 16, 17, 21-23].   Figure 1.1 Schematic illustration of electrically rechargeable metal-air batteries with its operating principle and bifunctional cathode. Reprinted with permission from [16].   4   Figure 1.2 Schematic polarization curve of rechargeable Zn-air battery. The equilibrium potential of the Zn-air battery (black line) is 1.65 V, but the practical voltage (red line) in discharge is lower than 1.65 V due to the sluggish ORR. A large potential is needed to charge Zn-air battery, higher than the equilibrium potential (blue line). Reprinted with permission from [11].   Figure 1.3 Challenges facing rechargeable aqueous Li-air batteries. Reprinted with permission from [21].  5  A typical discharge polarization and galvanostatic discharge-charge cycle curves are shown in Figure 1.4 for a rechargeable Li-air battery with a hybrid aqueous/organic electrolyte and various air electrodes. Yoo and Zhou introduced metal-free graphene nanosheets (GNSs) as air cathode with similar performances to Pt/C cathode as well as enhanced cycling ability, i.e. discharge and charge potential deviation of less than 0.2 V after 200 hrs of testing at 0.5 mA cm-2 [24].     Figure 1.4 a) Discharge polarization curves of gas diffusion electrodes containing graphene nanosheets (GNSs), 20 wt% Pt/C and acetylene black at 0.5 mA cm-2, b) Galvanostatic discharge-charge cycle curves of heat-treated GNSs at 0.5 mA cm-2. 1 M LiClO4/ED/DEC was used as the organic electrolyte and 1 M LiNO3 + 0.5 M LiOH was used as the aqueous electrolyte. A solid-state electrolyte Li1+x+yAlx(Ti,Ge)2−xSiyP3−yO12 (LISICON) film was used as a separating membrane between the organic and aqueous electrolytes to prevent intermixing of the two solutions. The discharge-charge performance was carried out at a current density of 0.5 mA cm-2 between 2 and 4.8 Vvs. Li/Li+ for 2 hrs per each cycle. 298 K. Reprinted with permission from [24]. 6  Alkaline fuel cells such as the alkaline hydrogen-oxygen fuel cell, direct borohydride fuel cells and others (e.g. methanol and formate fuel cells) have also received significant attention [25-28]. Particularly, alkaline regenerative fuel cells (ARFCs) are promising candidates as alternative energy storage technologies which in principle, include an electrolyzer and a fuel cell, employing hydrogen as energy carrier [29]. The electrolyzer is responsible for hydrogen production during the charging mode while the fuel cell part combines the hydrogen and oxygen (from air) to generate electricity during discharge mode [29-31]. A smaller and compact version of this unit is called unitised regenerative fuel cell (URFC) which both electrolyzer and fuel cell parts are combined into one cell, effectively reducing the costs and decrease the complexity of the system (Figure 1.5) [30-32]. Such a unit provides wide range of advantages including environmental friendliness of employed chemistries, i.e. hydrogen, oxygen and water, potentially high energy density due to use of H2 and customizable in size to appeal to variety of applications including grid-scale energy storage, centralized commercial/residential energy storage, transportation as well as space exploration [30-33]. However, issues such as low round-trip power efficiency† (less than 40%) and relatively high capital costs comparing to pumped hydroelectric and compressed air ($4000 USD per kW) hinder their wide-spread commercialization [30-32].  Figure 1.6 presents a typical polarization curve for an anion exchange membrane-unitised regenerative fuel cell (AEM-URFC) with Ni/C+MnOx/GC as O2 electrode and Ni/C as H2 electrode, providing round-trip power efficiencies of 34 to 40% at 10 mA cm-2 and a peak power density of 17 mW cm-2 in fuel cell mode [32].                                                   †  The round-trip power efficiency is defined as the power density in fuel cell mode over the power density in electrolyzer mode, both at certain current density. Here, the round-trip power efficiency was calculated at 10 mA cm-2. 7    Figure 1.5 Schematic representation of an anion exchange membrane-unitised regenerative fuel cell (AEM-URFC) as an energy storage unit. The AEM-URFC stores renewable energy as H2 while in electrolyzer mode and then uses that H2 to generate electric energy on-demand when in fuel cell mode. Reprinted with permission from [32].  8   Figure 1.6 Polarization curves for the AEM-URFC under eight fuel cell/electrolyzer cycles. MEA: Fumapem FAA-3, Fumatech as Anoion exchange memberane. Gas diffusion electrodes: 6 mgcat cm−2 Ni/C for the H2 electrode and 4 mgcat cm−2 MnOx/GC:Ni/C (5:1 weight ratio) for the O2 electrode. An alkaline ionomer (1.8 μL per 1 mg catalyst, Fumion FAA-3, Fumatech) was added to each of the electrodes. 20 min per each cycle. 335 K. H2/O2 gas flow rates of 300 Standard Cubic Centimeters per Minute (SCCM). All potentials are reported vs. reversible hydrogen electrode (RHE). Reprinted with permission from [32].  The alkaline electrolyte offers two significant advantages for the oxygen (or air) cathode: 1) Higher exchange current densities of oxygen reduction/evolution reactions compared to acidic media [2] and 2) The possibility of employing less expensive non-precious metal catalysts (e.g. transition metal oxides, perovskite or fluorite-type oxides) [34-37]. In acidic electrolytes, however, the use of noble metal catalysts seems inevitable at present for commercial applications of PEM fuel cells [38].  9  Manganese dioxide was found to be a potentially suitable catalyst for ORR in alkaline batteries and fuel cells due to a unique combination of properties, i.e. low cost, abundance, environmental friendliness, low self-discharge rate (open-circuit state) and fairly stable performance over a wide temperature range [39]. With regards to the oxygen electro-reduction mechanism, one of its major advantages is that it can facilitate the reduction of the peroxide ions (formed in ORR as an intermediate in neutral and alkaline media), reaching the theoretical four electrons per oxygen molecule exchanged during ORR [40-48]. However, some challenging issues such as low electrocatalytic activity for OER and poor rechargeability due to the conversion of MnO2 to non-rechargeable discharge products like Mn(OH)2, Mn2O3 and Mn3O4 limit the application of manganese dioxide as a lone bifunctional catalyst for both ORR and OER in rechargeable metal-air batteries and alkaline regenerative fuel cells [49, 50].  This work aims at developing active and durable MnOx-based catalysts for both ORR and OER in alkaline media, by modifying the microstructure and crystal structure of the manganese dioxide as well as introduction of co-catalysts to enhance its electrocatalytic activity and stability for both ORR and OER.  1.2 Bifunctional catalysts for ORR and OER in alkaline batteries and fuel cells In theory, it could be possible to develop an ideal, reversible, oxygen electrode with low ORR as well as OER overpotentials. In practice, a suitable catalyst should have certain characteristics, i.e. minimum changes in surface structure during the operation and the ability to fluctuate between the two reactions in a small potential window near the oxygen electrode equilibrium potential. The high activation overvoltage of the oxygen electrode in aqueous solutions, which implies a strong irreversible system, limits the choice of catalysts for the oxygen electrode to four major groups of 10  materials: 1) Pt, Ag, Ni, 2) Mixed valence oxides of Co, Ni and Mn with spinel‡ or perovskite§ crystal structures, 3) Metal sulfides, nitrides and carbides, and 4) Mixed metal oxides catalysts containing Pt, Ir, Ru, Os and/or Rh [7, 51-53].   1.2.1 Nobel metals and their alloys Noble metals and their alloys such as Pd [2], Ag [54], Pt [55], Pt-Au [56] and Pt-Co [57] have been extensively investigated for ORR in alkaline media but their lower electrocatalytic activity toward OER as well as high price compared to non-precious group metals (non-PGM) such as perovskite-type oxides (e.g. LaNiO3 and LaCoO3) [2, 4, 58-63] and Co oxides [64, 65], limit their widespread use as cost-effective bifunctional oxygen electrode catalysts. Moreover, other noble metals and their oxides such as Ru [66, 67], Ir [68, 69], RuO2 and IrO2 [2, 70], which are known as benchmarks for OER electrocatalysis, exhibit poor ORR electrocatalytic activity compared to MnOx [4, 67] and Pt [2, 55, 67], hindering their development as bifunctional oxygen electrocatalysts as well. Figure 1.7 summarizes the ORR and OER overpotentials of some candidates for bifunctional oxygen electrodes from the literature. For instance, high ORR overpotentials of -662, -502 and -444 mV (at -2 mA cm-2) have been reported for 83 wt% RuO2/C [71], 20 wt% Ru/C [67, 69] and 20 wt% Ir/C [69], respectively, while much lower values of -375, -365 and -284 mV (at -2 mA cm-2) were obtained for activated MnO2 [4], MnO2/Nitrogen doped carbon nano tube [72] and nano particulate Pt/C [73], respectively. In the OER region, 20 wt% Pt/C [67], Pt/TiO2 [74] and nano-sized Ag [75] have been reported to generate high OER overpotentials of 526, 473 and 464 mV (at 2 mA cm-2), respectively, comparing to lower values                                                  ‡ The spinels have the general formula of A2+B23+X42- where A and B are cations and X is an anion such as O, S, Se and Te. They crystallize with the same structure as the mineral spinel, i.e. MgAl2O4 [51] K.E. Sickafus, J.M. Wills, N.W. Grimes, Structure of Spinel, Journal of the American Ceramic Society, 82 (1999) 3279-3292..  § Please go to 1.2.3. 11  of 329, 252 and 206 mV (at 2 mA cm-2) for Core-Corona Structured Bifunctional Catalyst (CCBC) [76], nano-sized Co3O4 [65] and MnO2-LaCoO3 [4], respectively.    Figure 1.7 A comparison between bifunctional oxygen performance of different catalyst materials in alkaline media. The ORR and OER overpotentials were calculated at superficial current densities specified on the graph. Reprinted with permission from [4].  1.2.2 Manganese dioxide A considerable amount of work has been done on manganese dioxide since its introduction as depolarizing agent in the Zinc-Ammonium Chloride-Carbon battery by Leclanché in 1866 [39]. The use of chemically-synthesized manganese dioxide proposed by Glemser, instead of natural MnO2 first used by Leclanché, substantially enhanced its performance in the so called Leclanché 12  battery [39]. The next major step was in 1952 when Herbert introduced the application of electrochemically prepared MnO2, first proposed by Van Arsdale and Maier in 1918, in a concentrated KOH solution for commercial batteries [77]. In 1973, Zoltowski et al. showed that carbon supported air electrodes with manganese dioxide catalyst could improve the reversibility of the ORR, or in other words, it could enhance the performance of rechargeable metal-air batteries [44]. For 20 wt% MnO2/C, high ORR electrocatalytic activity was reported, close to the performance of 20 wt% Pt/C catalyst, i.e. a current density of about -3.5 mA cm-2 for both aforementioned catalysts at -300 mVMOE in 1 M KOH at 298 K and 1600 rpm [45]. Various synthesis methods of manganese oxides have been reported in the literature including hydrothermal synthesis, sol-gel synthesis, thermal decomposition, chemical co-precipitation and electrodeposition methods, leading to a number of different morphologies and crystal structures for synthesized Mn oxides [78-85]. Electrodeposition methods, in particular, have gained much attention due to several merits including ease of processing, low production cost, environmental compatibility, better control over properties of deposited materials such as morphology and crystallographic structure, high degree of reproducibility and high yield [84, 86, 87]. MnO2 exists in several crystallographic forms, which are known as α (Ramsdellite), β (Pyrolusite), γ (Electrolytic), δ (Birnessite), ε and λ (Spinel type) forms. The α, β, and γ forms possess 1D tunnels in their structures while the δ is a 2D layered compound and the λ is a 3D spinel structure (Figure 1.8) [85, 88, 89]. The physico-chemical and electrochemical properties of MnOx species depend very much on its crystallographic features [39]. Since the reduction potential of MnO2 is close to the oxygen reduction potential in alkaline media (between -110 and -126 mVSHE depending on the electrolyte’s composition and the crystal structure of manganese dioxide [90, 91]), the electro-reduction of manganese dioxide may occur 13  simultaneously with ORR (or at a more positive potentials due to the higher overpotential for the ORR). Cao et al. reported that the ORR happens in parallel with the reduction of MnO2 to MnOOH and the electrocatalytic activity of MnO2 for ORR is dependent on its electrochemical activity for reduction in alkaline media [48]. This further underlines the importance of Mn valence changes on the performance of MnOx-containing bifunctional electrodes which catalyze both ORR and OER.14      Figure 1.8 Schematic representation of five different crystal structures of MnO2: A) Pyrolusite (β-MnO2):  Rutile structure with an infinite chain of MnO6 octahedra sharing opposite edges. Each chain is corner-linked with four similar chains, B) Ramsdellite (α-MnO2): Cross-linking of double or triple chains of the MnO6 octahedra resulting in two-dimensional tunnels within the lattice, C) Birnessite (δ-MnO2): Layered structure containing infinite two-dimensional sheets of edge-shared MnO6 octahedra, D) Spinel (λ-MnO2): A 3D spinel structure and E) Electrolytic manganese dioxide (γ-MnO2): An intergrowth of pyrolusite in ramsdellite matrix.  Each MnO6 octahedra composes of oxygen and manganese atoms in the corners and center, respectively. Reprinted with permission from [85, 88, 89].15  1.2.2.1 Electro-reduction of MnO2 In 1966, Kozawa and Powers proposed a two-step mechanism for the reduction of electrolytic MnO2 [92]:  𝑀𝑛𝑂2 + 𝐻2𝑂 + 𝑒−⟶𝑀𝑛𝑂𝑂𝐻 + 𝑂𝐻−   (𝐸298𝐾0 = −0.022 𝑉𝑆𝐻𝐸) (3) 𝑀𝑛𝑂𝑂𝐻 + 𝐻2𝑂 + 𝑒−⟶𝑀𝑛(𝑂𝐻)2 + 𝑂𝐻−    (𝐸298𝐾0 = −0.072 𝑉𝑆𝐻𝐸) (4)  The first step begins with the formation of MnOOH in a solid-phase reaction without any changes in the basic structure of MnO2 [92]. During this step, protons are associated with the manganese dioxide lattice, while the reduction of Mn4+ into Mn3+ ions occurs on the surface [92]. It is reported that this discharge reaction is theoretically finished at an oxygen index of 1.5 in MnOx [92]. In addition to this reaction, parts of MnOOH are then converted to the Mn(OH)2, also known as deep-discharge process, which limits the rechargeability of MnO2 [92]. During the recharge process of the deep-discharged γ-MnO2, Mn(OH)2 is transformed to δ­MnO2 (layered birnessite structure) [92]. The reduction of birnessite during the subsequent cycles results in excessive formation of hausmannite (Mn3O4) which has been regarded as the main reason for the poor rechargeability of manganese dioxide [40, 41, 93].  For Leclanché-type Zn-MnO2 batteries in which MnO2 is being reduced as the cathode material, two solutions have been suggested in practice to avoid the deep discharge: 1) Limiting the discharge voltage of MnO2 and 2) Applying a limit to the capacity of the zinc anode electrode [42]. Moreover, other methods such as chemical or physical modifications of MnO2 cathodes have been reported to further improve the recharging ability of manganese dioxide [43, 94]. The feasibility of these methods for MnO2 acting as ORR catalyst have not been reported in the 16  literature thus far. This could arise from the fact that the ORR happens at more negative potentials compared to the potentials needed for Mn valence changes in alkaline media which makes it hard to limit the potential regime [48, 90, 91]. In 1976, Ruetschi investigated the discharge mechanism of γ-MnO2 deposited on fine Ni or Pt screens in 10.2 M KOH electrolyte. The following reaction mechanism was proposed [46]:  𝑀𝑛𝑂2 + 𝐻2𝑂 + 𝑒−𝑅𝑒𝑑𝑢𝑐𝑡𝑖𝑜𝑛→       𝑀𝑛𝑂𝑂𝐻 + 𝑂𝐻− (5) 𝑀𝑛𝑂𝑂𝐻 + 𝐻2𝑂 + 𝑂𝐻−𝐷𝑖𝑠𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛→        𝑀𝑛(𝑂𝐻)4− (6) 𝑀𝑛(𝑂𝐻)4− + 𝑒−𝑅𝑒𝑑𝑢𝑐𝑡𝑖𝑜𝑛→       𝑀𝑛(𝑂𝐻)42− (7) 𝑀𝑛(𝑂𝐻)42− 𝐶𝑟𝑦𝑠𝑡𝑎𝑙𝑖𝑧𝑎𝑡𝑖𝑜𝑛→          𝑀𝑛(𝑂𝐻)2 + 2𝑂𝐻− (8)  According to Ruetschi, Mn4+ is reduced to Mn3+ in MnOOH (eq. 5) and then, the manganese (III) oxy-hydroxide dissolves in water (eq. 6) [46]. Afterwards, the hydrated Mn3+ ions are further reduced to Mn2+ (eq. 7) and crystalized to produce stable manganese hydroxide (eq. 8) [46]. Therefore, two cathodic peaks were reported for the sequential reduction of Mn4+ to Mn2+ according to the aforementioned simplified mechanism (Figure 1.9) [46]. The reduction steps in reality are far more complex for manganese dioxide mainly due to the crystal structure changes during the discharge-charge processes [42, 45, 90, 94-98]. Lima et al. studied the redox performance of different manganese oxides supported on carbon in conjunction with in-situ X-ray Absorption Near Edge Structure (XANES) in alkaline media, showing high redox electrochemical activity for MnO2 and to lower extend for Mn2O3 and then Mn3O4 (Figure 1.10) [45]. In the cathodic direction, the reduction of MnO2 to MnOOH starts around 200 mVMOE, converting almost 17  40% of the MnO2 to MnOOH at -100 mVMOE (Figure 1.10 and Table 1.1) [45]. Further decrease in the potential leads to the formation of Mn3O4 and Mn(OH)2 starting at -300 and -700 mVMOE, respectively (Figure 1.10 and Table 1.1) [45]. In the anodic direction, the reduced MnO2, i.e. majorly Mn(OH)2, converts to Mn2O3 (slowly) and MnOOH at -300 and 0 mVMOE, respectively, followed by the oxidation of Mn3+ species, i.e. MnOOH, to MnO2 at 400 mVMOE while the generated Mn3O4 during cathodic polarization step stays intact in the anodic potential region investigated (Figure 1.10 and Table 1.1) [45].     Figure 1.9 Linear sweep voltammetry of 0.2 mg MnO2 on Ni mesh in N2 saturated 10.2 M KOH. 0.05 mV s-1. 298 K.  18   Figure 1.10 Cyclic voltammograms of different manganese oxides supported on carbon in N2 saturated 1 M KOH at 20 mV s-1 and 298 K. Samples labeled as A, B and C are MnO2/C, Mn2O3/C and Mn3O4/C, respectively. The potentials are reported vs. mercury-mercury oxide reference electrode (Hg/HgO). Reprinted with permission from [45].  Table 1.1 MnO2 crystal structure evolution at several electrode potentials in fractional composition. Other conditions idem to Figure 1.10. Reprinted with permission from [45].  19   Gyenge and Drillet further reported the existence of similar anodic and cathodic peaks for various commercially available MnO2 catalysts in alkaline media, depicting some deviations in the peak potentials of  these reduction and oxidation peaks compared to the literature, mainly due to differences in the experimental protocols such as MnO2 synthesis method, electrolyte composition and scan rate (Figure 1.11) [90]. Moreover, they have shown other small scan-rate-dependent peaks around the two major anodic and cathodic peaks in the cyclic voltammograms of all investigated MnO2 samples which they have contributed the existence of such peaks to the electrochemical, structural and physico-chemical phenomena (e.g. adsorption) together with the carbon support effect (Figure 1.11) [90].   20   Figure 1.11 Cyclic voltammograms of four different commercially available MnO2 powders supported on carbon in N2-purged 6 M KOH at 2 mV s-1. 295 K. Third cycle is presented. The potentials are reported vs. mercury-mercury oxide reference electrode (MOE). Reprinted with permission from [90].  1.2.2.2 Oxygen reduction reaction on manganese oxides The electro-reduction of oxygen in alkaline media is composed of series of complex electrochemical reactions. It is believed that the oxygen reduction reaction kinetics are governed by the binding energies between the catalyst and oxygen containing intermediates, i.e. O2−, HO2− and OH− [2, 17, 90, 99, 100]. Generally, two mechanisms have been proposed for ORR in alkaline 21  media [2, 17, 45, 98-100]: a) Four-electron pathway (direct or indirect) and b) Two-electron pathway. Table 1.2 summarizes the possible pathways for ORR on catalysts in alkaline media.   Table 1.2 Possible ORR pathways in alkaline media [2, 17, 45, 98-100]. Four-electron pathway (Direct) O2 + 2H2O + 4e− ↔ 4OH−   (E298K0 = 0.401 VSHE)   (2) Two-electron pathway 𝑂2 + 𝐻2𝑂 + 2𝑒− ↔ 𝐻𝑂2− + 𝑂𝐻−   (E298K0 = −0.076 𝑉𝑆𝐻𝐸)   (9) Four-electron pathway (Indirect) 1) 𝑂2 + 𝐻2𝑂 + 2𝑒− ↔ 𝐻𝑂2− + 𝑂𝐻−   (E298K0 = −0.076 𝑉𝑆𝐻𝐸)   (9)  2) 𝐻𝑂2− + 𝐻2𝑂 + 2𝑒− ↔ 3𝑂𝐻−   (E298K0 = 0.878 𝑉𝑆𝐻𝐸)   (10) Or 3) 2𝐻𝑂2− → 𝑂2 + 2𝑂𝐻−   (11)  The ORR mechanism is mainly affected by the configuration of adsorbed oxygen on the catalyst’s active sites, which itself depends on two factors: 1) Surface geometry of the catalyst and, 2) Oxygen binding energies on the catalyst (or the formation energy of lattice oxygen vacancies)  [101, 102]. In cases where a single atom of the adsorbed oxygen is perpendicularly coordinated to the surface, two-electron pathway mainly prevails, generating peroxide ions (eq. 9) [17]. This process can be further explained in details via eqs. 12, 13 and 14 [17, 90, 99]:  𝑂2 → 𝑂2,(𝑎𝑑𝑠)  (12) 𝑂2,(𝑎𝑑𝑠) + 𝐻2𝑂 + 𝑒− → 𝐻𝑂2,(𝑎𝑑𝑠) + 𝑂𝐻− (13) 𝐻𝑂2,(𝑎𝑑𝑠) + 𝑒− → 𝐻𝑂2− (14) 22   On some ORR catalyst surfaces the two-electron oxygen reduction reaction (eq. 9) could be followed by either the two-electron reduction of hydrogen peroxide ions (eq. 10) or a disproportionation reaction of HO2− (eq. 11), leading to indirect four-electron pathway, also known as 2 × 2e– [17, 45, 90, 98]. Different mechanism is proposed when both of the adsorbed oxygen atoms are coordinated parallel to the catalyst surface [17, 99, 101, 102]. This configuration of adsorbed oxygen favors O2 dissociation, leading to the direct four-electron pathway via eqs. 12, 15 and 16 (overall eq. 2):  𝑂2 → 𝑂2,(𝑎𝑑𝑠)  (12) 𝑂2,(𝑎𝑑𝑠) + 2𝐻2𝑂 + 2𝑒− → 2𝑂𝐻(𝑎𝑑𝑠) + 2𝑂𝐻− (15) 2𝑂𝐻(𝑎𝑑𝑠) + 2𝑒− → 2𝑂𝐻− (16)  Figure 1.12 shows typical ORR polarization curves for a variety of catalysts in alkaline media. The ORR limiting current densities (iL,j)**  in 0.1 M KOH solution were found to be around -5.8 mA cm-2 for catalysts following a 4e− pathway such as Pt/C, MnOx, LaMn0.5Cu0.5O3 and α-MnO2 while half of that, i.e. 2.9 mA cm-2, was reported as iL for ORR catalysts showing 2e− pathway, e.g. glassy carbon and δ-MnO2 [104-107]. ORR Tafel slopes and exchange current densities of between -80 to -120 mV dec-1 and 0.02 μA cm-2, respectively, were shown for carbon-based ORR catalysts which follow 2e− pathway, while Tafel slopes as low as -40 mV dec-1 and                                                  ** The limiting current density for a reactant is the current density under pure mass transfer control which is the max. current density that can be supported by that reactant [103] C. Oloman, Electrochemical processing for the pulp and paper industry, The Electrochemical Consultancy, Romsey, Hants, UK, 1996. 23  exchange current densities as high as 0.63 μA cm-2 were reported for catalysts following a 4e- pathway such as Pt, γ-MnO2 and α-MnO2 [5, 90, 104-109].   Figure 1.12 ORR polarization curves of various catalysts following 2e− or 4e− pathways in 0.1 M KOH. Theoretical limiting current densities for the 2e− or 4e− pathways are indicated by solid lines together with a ±10% margin (dashed lines). Currents are normalized to the geometrical area of the disk. 1600 rpm. 298 K. PO₂ equals to 1 atm. Reprinted with permission from [107].  Manganese dioxide has been intensely investigated as a catalyst for ORR in alkaline batteries and fuel cells due to a unique combination of properties, i.e., low cost, abundance, environmental friendliness and promising electrocatalytic activity for ORR [27, 39, 47, 90]. The study by Lima et al. reveals that the ORR on carbon supported manganese oxides (MnOx/C) in 1 M KOH undergoes a 2e− pathway [49]. According to this mechanism, Mn4+ ions rapidly reduce to Mn3+ while oxygen is promptly being adsorbed without splitting (eqs. 12 and 17) [49]. The rate determining step for overall ORR is reported to be the electron transfer of Mn3+ to the adsorbed 24  oxygen without molecular splitting (eq. 18) [49]. The speedy oxidation of water and formation of OH− and HO2− ions (eq. 19) are the final steps in the proposed 2e− pathway for ORR on MnOx/C [49, 110]:  𝑀𝑛4+ + 𝑒−⟷𝑀𝑛3+     (𝑓𝑎𝑠𝑡) (17) 𝑂2⟷ 𝑂2,𝑎𝑑𝑠     (𝑓𝑎𝑠𝑡) (12) 𝑀𝑛3+ + 𝑂2,𝑎𝑑𝑠⟷𝑀𝑛4+ + 𝑂2,𝑎𝑑𝑠−      (𝑠𝑙𝑜𝑤) (18) 𝑂2,𝑎𝑑𝑠− + 𝐻2𝑂 + 𝑒− → 𝐻𝑂2− + 𝑂𝐻−     (𝑓𝑎𝑠𝑡) (19)  Mao et al., however, proposed an overall four-electron mechanism for the reduction of O2 on MnOx catalysts in alkaline solutions [111]. It starts with the reduction of oxygen to hydrogen peroxide ions (eqs. 17 to 19) and continues with the disproportionation of HO2− on MnOx surfaces leading to O2 and OH− (eq. 11) when the regenerated oxygen gets to reduce again in the same first reduction steps (eqs. 17 to 19), resulting in a 2 × 2e− pathway for ORR [111]. Although Lima et al. proposed a two-electron mechanism for ORR on γ-MnO2, they have reported that under certain conditions of low rotation rates and/or low overpotentials, a complete disproportionation reaction of HO2− at the MnyOx sites could follow the main reaction which results in an overall 4e− process per O2 molecule, as suggested by Mao et al. and others [37, 45, 49, 50, 110, 111]. Figure 1.13 summarizes the ORR mechanisms on manganese oxides, revealing 4e− (Figure 1.13-A), 2e− (Figure 1.13-B) and 2 × 2e− pathways (Figure 1.13-B and C).  25   Figure 1.13 Possible ORR pathways on manganese oxides in alkaline media: A) Four-electron pathway reducing O2 to hydroxide, B) Two-electron pathway reducing O2 to peroxide ion and C) Two-electron reduction of peroxide ions. Orange denotes species on the catalyst surface, and blue/purple denotes species in solution. Reprinted with permission from [107].  It is worth mentioning that a competition between 2e− and 4e− mechanisms is reported for ORR on manganese dioxide/carbon electrodes in alkaline solutions which is unanimously believed to depend on the carbon content of the catalyst layer as well as the applied electrode potential [45, 48-50, 111]. Increasing the carbon support content of the catalyst layer enhances the two-electron mechanism for ORR [45, 111]. 26  The ORR electrocatalytic activity and reduction ability of different crystallographic types of manganese oxides have been intensively studied [39, 41, 48, 90]. Poinsignon et al. studied the electrochemical reduction of four allotropic forms of manganese dioxide, i.e. α, β, γ and ε-MnO2, in KOH solutions [41]. They have shown that the γ and ε types of manganese dioxide are reduced to MnOOH (eq. 5) at more positive potentials compared to either ramsdellite (α-MnO2) or pyrolusite (β-MnO2) in 7 N KOH solution (Figure 1.14) [41]. Further studies by Cao et al. revealed that not only the reduction currents of different crystallographic forms of MnO2 increases from β < λ < γ < α ≈ δ in argon saturated KOH solution but also the currents from ORR on manganese dioxide follow the same trend, showing a relationship between the ORR electrocatalytic activity of MnO2 and its electro-reduction (Figure 1.15) [48]. For instance, the γ-MnO2, an intergrowth of pyrolusite (β-MnO2) in the ramsdellite (α-MnO2) matrix [39, 112], is known as one of the most electrocatalytically active crystal structures of manganese dioxide for ORR in alkaline media, e.g. modest ORR Tafel slope of 40 mV dec-1 (Figure 1.16) and exceptionally low ORR overpotential of -375 mV (at -2 mA cm-2), while its facile reduction during ORR was reported in many studies [4, 39, 46, 48, 90, 113].  27   Figure 1.14 Linear sweep voltammograms of four types of manganese dioxide: A) α-MnO2 (ramsdellite), B) β-MnO2, C) γ-MnO2 and D) ε-MnO2. N2 saturated 7 N KOH. 0.003 mV s-1. 298 K. Solid and dotted lines represent mass activity (I, currents normalized by the catalyst loading) and the fraction of one electron capacity expressed in percentage, respectively, vs. potential. The potentials are reported vs. mercury-mercury oxide reference electrode (Hg/HgO). Reprinted with permission from [41]. 28   Figure 1.15 Linear voltammograms of MnO2-catalyzed air electrode in: A) Argon and B) Air. The catalysts are labelled as follows: 1) α-MnO2, 2) β-MnO2, 3) γ-MnO2, 4) λ-MnO2 and 5) δ-MnO2. The reduction current is defined as positive (American current polarity convention). 6 M KOH. 298 K. 1 mV s-1. Currents are normalized by the catalyst loading, presenting mass activity values. The potentials are reported vs. mercury-mercury oxide reference electrode (Hg/HgO). Reprinted with permission from [48].      A) B) 29   Figure 1.16 Oxygen reduction polarization curves recorded in a flow cell for four types of commercially available MnO2 powders: Tronox (γ-MnO2), Riedel (β-MnO2), Merck (γ-MnO2) and Sigma (γ-MnO2 with high portion of α-MnO2). Gas diffusion electrodes are employed at a loading of 0.25 mg cm-2 for MnO2 mixed with Vulcan XC-72 (1:1 weight ratio). 6 M KOH. 293 K. Currents are normalized to the geometrical area of the electrode. The potentials are reported vs. mercury-mercury oxide reference electrode (MOE).  Reprinted with permission from [90].  1.2.2.3 Oxygen evolution reaction on manganese oxides The mechanism originally proposed by Krasil’shchikov to describe the OER on nickel (or nickel oxide) and cobalt (or cobalt oxide) seems to inspire other researchers to introduce various complex pathways with distinctive rate determining steps and intermediates on variety of surfaces from perovskite-type oxides to MnOOH, NiOOH, NiCo2O4, Li-doped Co3O4, iron oxide and cobalt [52, 59, 114-121]. The complexity of these mechanisms arises from the several reaction intermediates with no available method to clearly identify them on different surfaces during OER 30  [2, 16, 122, 123]. Figure 1.17 summarizes the proposed OER mechanisms in the literature for variety of metal and oxide surfaces. The OER mechanism described in Figure 1.17-A is based on an extensive computational work on ORR/OER pathways for metals which was extended to oxide surfaces such as perovskites and rutile structures, (also referred to as acid-base mechanism) in which OH- (Lewis base) attacks a metal-bound electrophilic oxygen surface species (Lewis acid) [124, 125]. This mechanism will be discussed in details in the following sections. Goodenough et al. further proposed similar mechanism for OER on pyrochlores (Pb2(Ir or Ru)2-xPbxO7-y), perovskites (Sr1-xNbO3-δ) and rutile oxides (RuO2 and IrO2) (Figure 1.17-B) [126]. Computational works of Mavros et al. confirmed similar acid-base OER mechanism for first-row transition-metal oxides (Figure 1.17-C) [127]. The only difference between the first proposed OER mechanism (Figure 1.17-A) and the other two (Figure 1.17-B and C) is that the first one involves a step with a bare catalyst surface unlike the other two (Figure 1.17-A).  Nevertheless, oxide surfaces are more likely to adsorb negatively charged species such as OH*, OOH* and O* in alkaline media [126]. Figure 1.17-D describes a two-site reaction mechanism for OER on RuO2 surfaces, based on the work of Faria et al., which one involves the typical acid-base pathway while the recombination of surface oxygen species to produce oxygen molecules happens at the other site [128]. Moreover, Figure 1.17-E represents the OER mechanism proposed by Gerken et al. for electrodeposited cobalt oxides [129]. Jung et al. performed a systematic study on the OER electrocatalytic activity of wide range of catalysts in alkaline media, based on the primary work of McCory et al. in which they evaluated the performance of oxygen-evolving catalysts under standardized protocol, i.e. OER overpotential measurements at the beginning and after 2 hrs of testing at 10 mA cmgeo-2, 298 K and PO₂ of 1 atm [130, 131]. Figure 1.18 depicts a graphical representation of activity and stability comparison 31  between different OER catalysts.  Although IrO2, RuO2 and NiCoO2 provided the lowest OER overpotentials (at 10 mA cmgeo-2) of 380 to 390 mV at the beginning, the OER overpotential values of IrO2 and RuO2 increased by 50 mV  after two hrs of testing (Figure 1.18) [130].  Moreover, other catalysts such as LiNiO2, Mn3O4 and NiO provided stable OER overpotentials of 420 mV during the two-hour-long tests at 10 mA cmgeo-2 (Figure 1.18) [130].  32   Figure 1.17 OER mechanisms proposed for crystalline oxide surfaces in alkaline electrolytes: A) Four-step reaction mechanism proposed by Rossmeisl et al. for the OER on noble metal catalysts and oxide surfaces [124, 125], B) Four-step reaction mechanism proposed by Goodenough et al. for the OER on perovskite surfaces [126], C) Acid–base mechanism proposed for first-row transition-metal oxides [127], D) Reaction mechanism proposed by Faria et al. involving recombination of oxygen atoms to produce O2 [128] and E) Reaction mechanism proposed by Gerken et al. for electrodeposited cobalt oxides [129]. The orange and blue denote species on the catalyst surface and in solution, respectively.  Reprinted with permission from [123].33   Figure 1.18 A comparison between the OER activity and durability of wide range of catalysts in 1 M NaOH: The x-axis is the overpotential required to achieve 10 mA cmgeo−2 at time equal to 0 s. The y-axis is the overpotential required to achieve 10 mA cmgeo−2 after two hrs of testing. The dashed diagonal line indicates where the stable catalysts would be. Catalyst loadings are 0.8 mg cm-2 for each case. 298 K. PO₂ of 1 atm. Reprinted with permission from [130].  Further, Meng et al. investigated the bifunctional electrocatalytic activity of manganese oxides with various crystallographic structures (α, β, δ-MnO2 and amorphous) in alkaline media, proposing an OER mechanism similar to the acid-base mechanism based on Rossmeisl et al. work [105]. They have shown that the OER electrocatalytic activity of investigated manganese oxides increases in the following sequence: δ-MnO2 < β-MnO2 < Amorphous MnOx < α-MnO2 , with α-MnO2 providing OER overpotential (at 10 mA cm-2) and Tafel slope of 490 mV (about 100 mV 34  higher than 20 wt% Ir/C, 20 wt% Ru/C and RuO2) and 77.5 mV dec-1, respectively (Figure 1.19) [105]. Furthermore, other studies have reported modest OER electrocatalytic activities for different types of manganese oxides including MnO and Mn3O4 when being used as either bifunctional or oxygen-evolving catalysts in alkaline media (Figure 1.20) [32, 67, 69, 104, 113, 121, 130].    Figure 1.19 A comparison between the OER overpotentials (at 10 mA cm-2) for four different types of manganese oxides in alkaline media: α, β, δ-MnO2 and amorphous MnOx. 0.1 M KOH. 1600 rpm. 298 K. Reprinted with permission from [105].   35   Figure 1.20 OER cyclic voltammograms of various catalysts in O2 saturated 0.1 M KOH: As-deposited 500 ALD MnO, annealed Mn2O3, glassy carbon, 20 wt% Ru/C and 20 wt% Pt/C. 20 mV s-1. 1600 rpm. 298 K. Reprinted with permission from [67].    However, difficulties associated with the synthesis of pure crystalline manganese oxides and the structural changes during reduction/oxidation cycles, express the need for more fundamental studies on these oxides during both ORR and OER. 1.2.2.4 Density Functional Theory (DFT) studies on manganese oxides catalyzing both ORR and OER  Recent DFT studies have shown that the following simplified mechanism occurs for ORR (eqs. 20-23) and OER (eqs. 23-20) on oxide surfaces in alkaline media (Figure 1.21-A) [5, 16, 107, 121, 123, 132]:  𝑂2 + 𝐻2𝑂 + 𝑒− ↔ 𝐻𝑂𝑂(𝑎𝑑𝑠) + 𝑂𝐻− (20) 36  𝐻𝑂𝑂(𝑎𝑑𝑠) + 𝑒− ↔ 𝑂(𝑎𝑑𝑠) + 𝑂𝐻− (21) 𝑂(𝑎𝑑𝑠) + 𝐻2𝑂 + 𝑒− ↔ 𝐻𝑂(𝑎𝑑𝑠) + 𝑂𝐻− (22) 𝐻𝑂(𝑎𝑑𝑠) + 𝑒− ↔ 𝑂𝐻− (23)  where HOO(ads), HO(ads) and O(ads) are the oxygen containing intermediates binding to active sites through their oxygen atoms [121]. It is worth mentioning that this mechanism was developed with the assumptions of: 1) The reactions occur on a perfect single metal site without any defects, and 2) The possibility of oxygen recombination on any metal surfaces is excluded [121, 123]. Hence, it fails to address the challenging effects of crystal structure transformations as well as the existing and newly-formed defects, i.e. intertwined structures, twinnings and numerous types of vacancies, during ORR and OER which is the case in commercial scenarios [5, 39, 121]. The constant difference between the binding energy levels of the HOO(ads) and HO(ads) intermediates for many metals and their oxides, also known as universal scaling relationship, contributes largely to the overpotential of both ORR and OER (Figure 1.21-B) [121, 124, 133]. Breaking away from this linear scaling relationship (red star in Figure 1.21-B) via modification of catalyst surfaces, enhances its activity for both ORR and OER, significantly lowers the reaction overpotentials.  37   Figure 1.21 A) Free-energy diagram for ORR (in reverse direction) and OER on an ideal catalyst. The vertical solid arrow shows ΔGHOO(ads)-ΔGHO(ads) on a perfect crystal which is 2.46 eV. B) Adsorption energy of HOO(ads) plotted against the adsorption energy of HO(ads) on the clean surfaces: Perovskites (○), rutiles (▵), MnxOy (□), TiO2 (◊) and Co3O4 (+). The best fit of all points is ΔEHOO(ads) = ΔEHO(ads) + 3.20 eV. The red star indicates where the binding energies need to be for an ideal catalyst. Reprinted with permissions from [121, 124].   In a study by Su et al. on DFT calculations of MnOx single crystals catalyzing oxygen reduction and evolution reactions in alkaline media, it has been reported that HO(ads) covered α-Mn2O3 and O(ads) covered β-MnO2 sites are the most active surfaces for ORR and OER, respectively, among Mn3O4 (001), α-Mn2O3 (110) and β-MnO2 (110) single crystals [121].  1.2.2.5 Dopants for manganese dioxide as cathode material in alkaline fuel cells and metal-air batteries  There have not been many intensive studies on the effect of dopants and co-catalysts to enhance the ORR/OER bifunctional electrocatalytic activity of manganese oxides. Sun et al. showed that the core-shell like combination of β-MnO2 nano-rods (40-50 nm in diameter and 38  500-1000 nm in length) with 8.7 wt% Pd coating could shift positively the ORR onset and half-wave potentials by more than 250 mV compared to pure manganese dioxide in alkaline media [134]. In another study, Klápště et al. investigated the activity of MnOx/C composites doped with divalent and trivalent ions, such as Ca2+, Mg2+, Ni2+, Bi3+ and Cr3+, for ORR in alkaline solutions [135]. While alkali earth metals like Ca and Mg did not enhance the ORR electrocatalytic activity compared to the un-doped manganese oxide samples, transition metal ions such as Ni2+ or Cr3+ shifted positively the ORR half-wave potential by 50 to 60 mV [135]. The authors hypothesized that this could be due to the ability of transition metal ions to exist in several states of valence, mainly facilitating the charge transfer to oxygen [135].  1.2.2.6 Nanostructured manganese oxides When it comes to the particle size, nanostructured MnOx generally outperforms the micro-sized particles mainly due to its higher specific surface area. Hence, more active sites are available for nanostructured MnOx to facilitate the O2 reduction/oxidation reactions. The higher porosity for nanostructured manganese oxides also leads to extra space for oxygen bubbles to evolve and avoid entrapment [17, 67, 136, 137]. Recent studies showed promising ORR and OER electrocatalytic activities for electrochemically deposited nanostructured manganese oxides in alkaline media, i.e. ORR and OER overpotential of -311 mV (at -2 mA cm-2) and 405 mV (at 2 mA cm-2) [136, 137].  While several synthesis methods such as hydrothermal synthesis, sol-gel synthesis, thermal decomposition and chemical co-precipitation have been studied for production of nano-sized MnOx, the electrodeposition techniques provided huge opportunity in the synthesis of nanostructured manganese oxides by introducing an easy process with low production costs, environmental compatibility, enhanced control over properties of deposited material such as morphology and crystallographic structure, high degree of reproducibility and high yields [78-87]. 39  Using electrochemical deposition methods, MnO2 can be either synthesized via cathodic reduction of Mn7+ species or anodic oxidation of Mn2+ [87]. Cathodic electrodeposition of MnO2 from anionic MnO4− (Mn7+) species The overall reaction for cathodic electro-reduction of MnO4− ions to deposit MnO2 can be represented as follows in acidic (eq. 24) and neutral (eq. 25) media [87, 138, 139]:  𝑀𝑛𝑂4− + 4𝐻+ + 3𝑒− →  𝑀𝑛𝑂2 + 2𝐻2𝑂   (𝐸298𝐾0 = 1.679 𝑉𝑆𝐻𝐸) (24) 𝑀𝑛𝑂4− + 2𝐻2𝑂 + 3𝑒− →  𝑀𝑛𝑂2 + 4𝑂𝐻−   (𝐸298𝐾0 = 0.595 𝑉𝑆𝐻𝐸) (25)  It is noteworthy to mention that the kinetic pathway of Mn7+ reduction to Mn4+ depends on a number of factors such as electrode potential (as described in eqs. 26 and 27), pH, concentration of MnO4− and other species available in the solution [87, 138, 139]. Moreover, Nguyen et al. argued that the deposition factors such as applied potential, electrolyte concentration, applied charge and post-thermal treatments, can significantly affect the surface morphology and electrochemical response of the cathodically electrodeposited MnOx [140].  𝐸 = 𝐸0 +𝑅𝑇𝑛𝐹. 𝑙𝑛𝑎𝑀𝑛𝑂4− .𝑎𝐻+4𝑎𝑀𝑛𝑂2 .𝑎𝐻2𝑂2  (26) 𝐸 = 𝐸0 +𝑅𝑇𝑛𝐹. 𝑙𝑛𝑎𝑀𝑛𝑂4− .𝑎𝐻2𝑂2𝑎𝑀𝑛𝑂2 .𝑎𝑂𝐻−4  (27)  The cathodic electrodeposition process from Mn7+ draws some attention as the method of choice for the synthesis of nanostructured MnOx since it provides the opportunity to modify the composition of the catalyst layer by either chemical doping or co-deposition of other oxides for 40  fabrication of enhanced composite catalysts [87, 139]. Moreover, other advantages include avoiding anodic oxidation and dissolution of metallic substrates, no passivation of anode and low risk of product contamination [87, 139, 141].  Anodic electrodeposition of MnO2 from cationic Mn2+ species Anodic electrodeposition of MnO2 has recently gained more attention mainly due to its flexibility at scale and control on the morphology as well as crystallographic phases of the final deposited MnOx [86, 137, 142-144]. This electrochemical deposition of manganese dioxide involves oxidation of Mn2+ species on the anode while hydrogen evolution is happening on the cathode in aqueous media of manganese salt as described by eq. 28 [84]. A wide range of electrochemical techniques have been reported in the literature for deposition of MnO2 anodically, including galvanostatic, potentiostatic, potentiodynamic and pulse deposition methods [84, 87, 145-148].  𝑀𝑛2+ +  2𝐻2𝑂 → 𝑀𝑛𝑂2 + 4𝐻+ + 2𝑒−   (𝐸298𝐾0 = 1.23 𝑉𝑆𝐻𝐸) (28)  Recent study by Clark et al. revealed that the complex electrochemical oxidation of Mn2+ to MnO2 can proceed via two different multistep mechanisms: 1) Disproportionation and 2) Hydrolysis [86]. The first mechanism, i.e. disproportionation, consists of the following reactions (eqs. 29 to 31) [86, 149]:     𝑀𝑛2+ → 𝑀𝑛3+ + 𝑒− (electrochemical oxidation) (29) 2𝑀𝑛3+ → 𝑀𝑛2+ +𝑀𝑛4+ (disproportionation)  (30) 𝑀𝑛4+ + 2𝐻2𝑂 → 𝑀𝑛𝑂2 + 4𝐻+ (hydrolysis) (31) 41   This mechanism is believed to occur in concentrated acidic media where soluble Mn3+ ions are more stable. The second mechanism (i.e. hydrolysis), first introduced by Fleischmann et al. and further developed by Catwright and Paul, mainly starts in low acidic and neutral media [86, 150-154]. This helps un-stabilize the soluble Mn3+, allowing the MnOOH to precipitate on the electrode surface. Next, the MnOOH oxidizes to MnO2 as follows (eqs. 32 to 34) [86]:   𝑀𝑛2+ → 𝑀𝑛3+ + 𝑒− (electrochemical oxidation) (32) 𝑀𝑛3+ + 2𝐻2𝑂 → 𝑀𝑛𝑂𝑂𝐻 + 3𝐻+(hydrolysis) (33) 𝑀𝑛𝑂𝑂𝐻 → 𝑀𝑛𝑂2 + 𝐻+ + 𝑒− (electrochemical oxidation) (34)  The deposition of solid MnO2 from Mn2+ ions follows different pathways as mentioned in  each of the proposed mechanisms which is expected to be the reason behind distinctive morphology of any deposited product [86, 149].  Experimental factors affecting the electrosynthesis of nanostructured MnOx via anodic electrodeposition method While many factors affect the electrodeposition process for manganese oxides, some important ones are being reviewed here from the literature. Any changes in these factors could substantially alter the physical properties of deposited Mn oxides, i.e. crystal structure, morphology, pore density and Mn valence, leading to diverse ORR/OER electrocatalytic activity and durability for the deposited materials.  Many researchers tried to study the individual effect of these electrodeposition factors, i.e. Mn2+ concentration, applied potential, temperature and surfactant content, on the morphology and 42  later the electrochemical properties of electrodeposited manganese oxides. However, the important yet ignored piece in all of these studies is the complex interactions between all of the electrodeposition operating factors. These complicated interactions are crucial to more reliable predictions of the effect each factor has on electrochemical properties of the deposited manganese oxides in presence of other variable factors.  Manganese ion concentration Mn2+ concentration is one of the important factors that can affect morphology, crystal structure and the mechanism by which manganese oxides deposit on the substrate during anodic electrodeposition process [155, 156]. In a study of Mn2+ oxidation on Pt electrodes in 2 M H2SO4 solution with various Mn ion concentrations, Nijer et al. reported two anodic peaks representing Mn2+ oxidation to Mn3+ (around 0.8 mVMSE, eq. 32) followed by MnOOH oxidation to MnO2 (around 0.9 VMSE, eq. 34) (Figure 1.22) [156]. At high concentrations of Mn2+ ions in the solution (over 90 mM), the first sharp anodic peak (around 0.8 mVMSE) represented facile oxidation of Mn2+ to Mn3+ (Figure 1.22) [156]. The decrease in the peak current density of second anodic peak was attributed to the slow electrochemical oxidation of MnOOH to MnO2, leaving a combination of MnOOH and MnO2 as the final deposited product on Pt substrate (Figure 1.22) [156]. Moreover, they have proposed that the decrease in current density after the first sharp peak may be due to formation of intermediates such as MnOOH on the electrode which inhibit further oxidation of Mn2+ [156]. Nonetheless, the intermediate species may oxidize to MnO2 at higher potentials, justifying the existence of second anodic peak (Figure 1.22) [156]. At low Mn2+ concentrations (less than 90 mM), the two oxidation peaks seem to overlap, meaning similar Mn2+/Mn3+ oxidation rate to that of MnOOH/MnO2 step (the rate determining step). This leads to high contents of MnO2 in the final deposited layer (Figure 1.22) [156]. 43    Figure 1.22 Cyclic voltammetry on platinum working electrode in 2 M H2SO4, with varying Mn2+ concentrations of 0.018 M to 0.73 M at 308 K and 1 mV s-1 [156]. The potentials are reported vs. mercury-mercurous sulfate reference electrode (MSE). Reprinted with permission from [156].  Babakhani and Ivey further studied the significance of Mn2+ concentration as an electrodeposition factor on the morphology of anodically deposited manganese oxides (Figure 1.23) [155]. At extreme low manganese concentrations (3 to 5 mM for Mn2+ ions), discrete oxide clusters were reported to grow vertically on the substrate as thin sheets (Figure 1.23-A and B). Increasing the Mn2+ concentration to 10 mM leaded to free-standing MnOx rods with fibrous surfaces (Figure 1.23-C and D) [155]. Manganese ion concentrations of over 50 mM were shown to be unfavorable mainly due to the poor adhesion of deposited manganese oxides to the substrate [155]. Figure 1.24 shows schematically this morphological evolution of electrodeposited 44  manganese oxides from thin sheets to rod-like structure and continuous coating with increasing the Mn2+ concentration, as studied by Babakhani and Ivey [155].    Figure 1.23 SEM plan-view and cross-sectional images of manganese oxide deposits synthesized from: A) 3 mM Mn(CH3COO)2 solution at 0.25 mA cm-2, B) 5 mM, C) 7 mM, D) 10 mM, E) 20 mM and F) 30 mM Mn(CH3COO)2 solutions at 5 mA cm-2. 10 min per each deposition. 373 K and pH of 7.5. Reprinted with permission from [155].  45   Figure 1.24 Schematic diagram showing the morphological evolution of electrodeposited manganese oxides (from left to right: thin sheets, rods, aggregated rods and non-uniform continuous coating) with increasing the Mn2+ concentration during anodic electrodeposition process. Reprinted with permission from [155].  Applied anodic potential and current density The applied anodic charge can also alter the crystallinity, surface morphology and coverage, pore density and more importantly, Mn valence of the electrodeposited materials during anodic electrodeposition of MnOx [147, 157-159]. Chang and Tsai showed that the XRD peak intensity of electrodeposited manganese oxides (at 2θ equals to 37.1° and 66.3°) increases marginally when the deposition potential decreases, signifying formation of higher crystallinity at lower applied potentials (Figure 1.25) [158]. The higher crystallinity is attributed to the slower rate of oxide deposition at low anodic potentials, allowing more time for the formation of highly aligned and uniform MnOx nanostructures [158]. SEM images of electrodeposited MnOx in Figure 1.25 clearly demonstrate the changes in their surface morphologies with applied potentials (Figure 1.26). While high density for small-sized pores in the microstructure of electrodeposited MnOx is observed at low anodic potentials, the low pore density at high potentials results in formation of uniform, compact and dense manganese oxide layer (Figure 1.26) [158]. Moreover, Chang and Tsai  study revealed that based on the XPS analysis of the electrodeposited manganese oxides at various 46  applied anodic potentials, Mn4+ is present at higher potentials above 0.65 VSCE whereas Mn3+/Mn4+ are formed in the deposit at lower deposition potential [158].    Figure 1.25 XRD patterns of MnOx electrodeposited at different anodic potentials (0.5 to 0.95 VSCE) in 0.25 M manganese acetate solution at 298 K. Arrowed peaks (at 2θ = 37.1° and 66.3°) correspond to the oxides formed on the carbon substrates. Potentials are versus saturated calomel reference electrode (SCE). Reprinted with permission from [158].   47   Figure 1.26 SEM micrographs showing the surface morphologies of manganese oxides electrodeposited at A) 0.5, B) 0.65, C) 0.8 and D) 0.95 VSCE characterized in Figure 1.25. Reprinted with permission from  [158].  Deposition temperature Temperature is another factor that can play a paramount role on the nucleation and growth rates as well as morphology of the final electrodeposited MnOx [155, 160]. Clark et al. studied the effect of temperature on anodic electrodeposition behavior of manganese dioxide in acidic electrolyte, showing the deposition rate is charge transfer controlled at low temperatures (295 to 333 K), i.e. plateau line presenting the dependence of limiting current density and rotation speed 48  in Levich plot, while the process is mass transfer limited at high temperatures (i.e. 343 to 363 K), as demonstrated by Levich plots in Figure 1.27 [86]. The significant mass transport limitations at high temperatures might seem contradictory given the diffusion rate of species in electrolyte enhances with increased temperature, however, this can be interpreted in terms of an increase in electrocatalytic activity of electrode at elevated temperatures forcing more demands on mass transport for effective charge transfer, Clark et al. argued [86].    Figure 1.27 Levich plots calculated from anodic electrodeposition of manganese dioxide in 0.1 M MnSO4+5 M H2SO4 at different temperatures. Reprinted with permission from [86].   295 K 313 K 333 K 343 K 353 K 363 K 371 K 49  Babakhani et al. further investigated the effect of deposition temperature on anodically electrodeposited manganese oxides using SEM surface morphology analysis (Figure 1.28) [155]. They have shown that at room temperature, the nucleation rate for electrodeposition of MnOx is low, resulting in discrete oxide particles with cracks while well-ordered rod-like structure (1-3 µm in diameter) and aggregated rods with fibrous feature can form at 333 K and 358 K, respectively (Figure 1.28) [155].    Figure 1.28 SEM plan-view and cross-sectional images of manganese oxides deposited from 10 mM Mn(CH3COO)2 at 5 mA cm-2 for 10 min and pH of 7.5: A) 298 K, B) 333 K, C) 358 K. Reprinted with permission from [155].   50  Surfactants Surfactants can significantly change the surface coverage and morphology of electrodeposited materials by mainly adsorbing to the solid/liquid interface, acting as a deposition template, reducing the interfacial energy and controlling the nucleation/growth of the particles which result in distinctive electrochemical activities for the deposited particles [150, 161-166]. Surfactants are amphiphilic organic compounds consisting of a hydrophilic head group attached to a long aliphatic hydrocarbon chain, generally classified into four main classes according to the charge of head group: 1) Non-ionic surfactant (absence of charge on the head group), 2) Anionic surfactant (with negatively charged head group), 3) Cationic surfactant (with positively charged head group) and 4) Zwitter-ionic surfactant (the head group has either a positive or negative charge depending on the pH of solution), as shown in Figure 1.29 [167-169]. The molecular structure of some commonly used ionic and non-ionic surfactants in electrodeposition processes are displayed in Figure 1.30 as an example.   Figure 1.29 Surfactant classifications based on the charge of head group. Reprinted with permission from [169].  51   Figure 1.30 Molecular structure of: A) Cationic cetyltrimethylammonium bromide (CTAB), B) Anionic sodium n-dodecylbenzenesulfonate (SDBS) and C) Non-ionic t-octyl phenoxy polyethoxyethanol (Triton X-100). Reprinted with permission from [161].  Surfactant-assisted electrodeposition methods have gain much attention lately as the methods of choice for synthesis of nanostructured manganese oxides with diverse morphologies and electrochemical properties for wide range of applications such as batteries and electrochemical capacitors [150, 161, 162, 170, 171].          Biswal et al. studied the effect of anionic surfactants on the morphology and electrochemical performance of anodically electrodeposited electrolytic manganese dioxide (γ-MnO2 or EMD) [162]. They have reported that the addition of sodium dodecyl sulfate (SDS) at its optimum concentration (about 50 ppm) during electrodeposition of γ-MnO2 significantly enhances its surface area (BET surface area of 130 m2 g-1 comparing to 100 m2 g-1 in the absence of SDS) and galvanostatic charge-discharge cycle life [162]. Moreover, they have shown a variety of morphologies for the different concentrations of SDS in the solution during anodic electrodeposition of manganese oxide from a porous structure with narrow needle-like particles (SDS: 10 ppm) to smaller narrow needle-like particles (within a size range of 50-100 nm) with (A) (B) (C) 52  high surface area (SDS: 50 ppm) and platy morphology with randomly oriented particles (net-like appearance) (SDS: 100 ppm) (Figure 1.31) [162].   Figure 1.31 FESEM images of the manganese oxide (EMD) samples in the presence of various concentrations of SDS (in ppm) as anionic surfactants. Reprinted with permission from [162].  Further studies by Dubal et al. explored the effects of t-octyl phenoxy polyethoxyethanol (Triton X-100) as non-ionic surfactant on the morphological property and electrochemical performance of anodically electrodeposited manganese dioxide [172]. The sample deposited in presence of Triton X-100 (1 wt% in the final deposition solution) showed more uniform and porous 53  morphology with smaller particles compared to the MnO2 deposited without any surfactant, Dubal et al. argued (Figure 1.32) [172]. They have also reported enhanced supercapacitance of 278 F g−1 (at 100 mV s−1 and 298 K based on cyclic voltammetry tests in alkaline media) for the MnO2 deposited in presence of Triton X-100, about 70 F g-1 higher than the case without any surfactant [172].   Figure 1.32 SEM images of MnO2 thin film prepared in: A) Absence and B) presence of Triton X-100. Reprinted with permission from [172].  Similar studies targeting the effect of cationic cetyltrimethylammonium bromide (CTAB) and non-ionic Triton X-100 surfactants on the morphology and electrochemical behavior of electrodeposited γ-MnO2 revealed characteristic morphologies with enhanced activity in rechargeable alkaline batteries for these samples [161]. Better discharge performance and lower degradation rates were reported when optimum Triton X-100 concentration of 5.1 mM were employed to deposit manganese dioxide, leading to a morphology of small needle-like fibers roughly packed to each other with excellent orientation and high surface area (Figure 54  1.33-A) [161]. Moreover, CTAB at optimum concentration of 9 mM slightly promoted the charge/discharge cycle behavior of the deposited MnO2, Ghaemi et al reported [161]. The morphology of these CTAB assisted electrodeposited MnO2 showed rather rod-like crystals, smaller and more perpendicular to the electrode surface compared to Triton X-100 case (Figure 1.33).   Figure 1.33 SEM images of electrodeposited manganese dioxide in presence of : a) 5.1 mM Triton X-100 and 9 mM CTAB. Reprinted with permission from [161].  When the surfactant concentration reaches a certain level in the solution, known as the critical micelle concentration (CMC), they tend to aggregate to form micelles (Figure 1.34) [173-175]. The formation of micelles is governed by the molecular interactions such as the van der Waals and electrostatic forces, making different conformations for their arrangements [173, 175]. Some examples of these arrangements are: spherical, cylindrical, hexagonal, cubic and lamellar (bilayer) structures as demonstrated in Figure 1.35 [176]. At concentrations lower than CMC, surfactant molecules adsorb on the solid/liquid interface and effectively enhance the growth of deposits 55  whereas at concentrations higher than CMC, its molecules tend to form aggregates, lowering the surfactant concentration at the interface and hence, compromising the effectiveness of surfactant on directing growth of deposited materials such as nano-sized MnOx [171, 177].   Figure 1.34 Surfactant aggregation to form micelle at critical micelle concentration (CMC). Reprinted with permission from [175].    Figure 1.35 Geometrical shapes of surfactant micelles in aqueous solutions. Reprinted with permission from [176].  Many factors such as temperature, pressure, pH, ionic strength and the properties of surfactant species like hydrophobic volume, chain length, head group area, etc., can affect the micellization [178]. Among them, temperature has shown substantial influence on CMC where their relationship largely depends on the surfactant systems [178-182]. For most non-ionic surfactants, it has been 56  reported that the CMC decreases with an increase in temperature mainly due to enhanced hydrophobicity coming from the destruction of hydrogen bonds between the hydrophilic head groups of surfactant and water molecules [178]. On the other hand, for ionic surfactants, the CMC shows a U-shaped dependence on temperature where CMC decreases to a min. point with increasing temperature and then keeps ascending with any further increase in temperature [178, 179, 182]. However, experimental results for some non-ionic surfactants showed similar U-type temperature dependence of CMC to ionic surfactants [181].  1.2.3 Perovskite-type oxides Another important class of non-precious metal oxide electrocatalysts for oxygen cathodes is perovskites, with the general formula of ABO3, (where A and B correspond to rare-earth metal and transition-metal ions, respectively) (Figure 1.36) [13, 183].   Figure 1.36 Schematic representation of perovskite-type oxides with the general formula of ABO3 where A sites includes rare-earth metal ions while B sites are transition-metal ions. Reprinted with permission from [184]. 57   These are catalysts of great diversity because of the wide range of ions and valences that the structure can accommodate [17, 185]. Different types of perovskites have been synthesized and reported to possess improved physical properties as well as good electrocatalytic activity for ORR and OER in alkaline electrolytes [59, 60, 184, 186]. Lee et al. revealed that La0.6Ca0.4CoO3 prepared by the amorphous citrate precursor process, calcined at 923 K and then rapidly quenched, has a high surface area of 33 m2 g-1 [61]. They reported that 10 mg cm-2 of this catalyst can provide a significant bifunctional performance, i.e. -280 mA cm-2 for ORR and 300 mA cm-2 for OER at 600 and 1600 mVRHE in 30 wt% KOH at 298 K, respectively, when the gas diffusion electrodes (GDEs) are exposed to pure oxygen [61]. In another study, Li et al. reported fast ORR kinetics for La0.6Ca0.4CoO3-carbon composites with low Tafel slope of -60 mV dec-1 and relatively high exchange current density of 5.8×10-8 mA cm-2 in O2 saturated 6 M KOH [187]. Although 0.17 mg cm-2 of this perovskite-type catalyst reveals good electrocatalytic activity toward OER in both O2 and N2 saturated 4 M KOH solutions at 1600 rpm, i.e. 10 mA cm-2 at 1600 mVRHE, it provides poor ORR electrocatalytic activity of 1-2 mA cm-2 at 600 mVRHE comparing to γ-MnO2 [90, 187]. Promising electrocatalytic activity for OER have been also reported for other types of perovskites such as Sm0.5Sr0.5CoO3−δ [63], LaNiO3 [188, 189], LaCoO3 [58, 190] and layered LaSr3Fe3O10 [191] in alkaline media. Figure 1.37 summarizes the ORR and OER overpotentials as a function of eg electron occupancy on a variety of perovskite-type oxides in alkaline solutions. Even though perovskite-type oxides show high electrocatalytic activity for OER, their relatively low electrical conductivity (e.g. 1-10 S cm-1 for LaCoO3 [58] comparing to 1.7×104 S cm-1 for IrO2 [192] at 298 K) and surface area (e.g. 12.6 m2 g-1 for LaCoO3 [58] comparing to 200 m2 g-1 for Vulcan XC-72 [193]), as well as poor electrocatalytic activity toward ORR, compared to other 58  non-precious compounds like MnOx, limit their use as bifunctional electrodes in concentrated alkaline solutions [4, 59, 60, 90, 184, 187, 194].   Figure 1.37 Volcano-type graphs showing the comparison between the electrocatalytic activity of various perovskite-type oxides for: A) ORR and B) OER. Figures reveal the ORR/OER overpotentials at 50 μA cm-2 in alkaline media as a function of eg electron occupancy at 298 K. Reprinted with permission from [123].   1.2.4 Fluorite-type oxides Oxides with fluorite-related structure, such as Nd3IrO7 with an orthorhombic structure (space group Cmcm) (Figure 1.38), were also investigated as bifunctional oxygen electrode catalyst in the literature [108, 195]. According to the author’s knowledge, the study by Kortenaar et al. was the only investigation available in the literature on exploring these potentially active oxygen catalysts in alkaline media. Tafel slopes and exchange current densities for OER of 25 mV dec-1 and 1.5×10-15 μA cm-2, respectively, and 63 mV dec-1 and 8.5 μA cm-2 for ORR, were reported in 59  45 wt% KOH [108]. The very low exchange current density for OER compared to ORR renders unlikely the practical possibility of using Nd3IrO7 or other IrO6 or IrO7-containing compounds as a lone bifunctional catalyst [108].   Figure 1.38 Schematic representation of Nd3IrO7 with an orthorhombic crystal structure (space group Cmcm). Reprinted with permission from [195, 196].  1.2.5 Carbon support for ORR/OER bifunctional catalysts Non-precious group metals (non-PGMs) and their oxides are the alternative cost-effective solution for next-generation catalyst materials, showing similar, and in some cases, superior activity and long-term stability comparing to the noble metal catalysts [4-6, 17, 197]. While low 60  cost ($0.95 USD per kg for MnO2 ††) , ease of synthesis, high abundancy and environmental friendliness make these oxides favorable to be used as bifunctional oxygen catalysts, their inherent low electrical conductivity (10-5-10-6 S cm-1 for MnO2 [198]) is one of the major drawbacks toward their use for oxygen electrocatalysis [4, 5, 10, 39, 199]. To alleviate the low electrical conductivity of transition metal oxides and perovskites in ORR and OER applications, carbon-based materials are widely being used as catalyst supports [4, 5, 10, 17, 107, 122, 200]. Carbons provide favorable merits including low cost ($0.95 USD per kg for carbon black ‡‡), abundance, wettability, large active surface area (200 m2 g-1 for Vulcan XC-72 [193]), enhanced electrical conductivity (2.70×10-1 S cm-1 for Vulcan XC-72 [193]) and good stability in harsh concentrated acidic and alkaline media [17, 38]. It is believed that carbons with large meso-pores and thick crystalline walls provide more favorable properties such as high electrical conductivity and oxidation resistivity for ORR/OER applications [201].  However, carbonaceous materials alone are not the catalysts of choice to be used as ORR/OER bifunctional electrodes in aqueous solutions due to low electrocatalytic activity for both ORR and OER as well as durability issues mainly caused by carbon corrosion at the high anodic potentials [10, 17, 29, 137, 202-205]. Carbon structure modifications such as graphitization or hetero-atom (e.g. S, P and N) doping can boost the durability of carbon materials to enhance their role as either catalyst support or even ORR/OER bifunctional catalyst itself, mainly by increasing the defects and edge plan sites in graphitic matrix [10, 17, 203, 206-208]. Zhang et al. reported high ORR/OER electrocatalytic activity and durability for mesoporous carbon foams co-doped with N and P in                                                  †† From Alibaba.com on September 8th, 2016. ‡‡ From Alibaba.com on September 8th, 2016. 61  alkaline media, showing similar ORR electrocatalytic activity to commercial Pt/C catalyst as well as lower OER onset potential (up to 1700 mVRHE) comparing to Ru/C [206]. Using DFT calculations, they revealed that the most active sites for ORR and OER are N-dopant sites near the graphene edges and N/P co-doped graphene edges, respectively [206]. Although Vulcan XC-72 is known as the most conventional ORR catalyst support, it has been rarely used as either catalyst or support for OER due to the aforementioned carbon corrosion issues at high anodic potentials [2, 4, 5, 10, 17, 90, 202, 203]. Nanostructured carbons such as carbon nanotubes, graphene and especially N-doped graphene, however, showed promising performances as effective support or highly electrocatalytically active catalysts for ORR and OER, reviving the hope of finding durable, cheap, abundant and noble-metal-free oxygen catalysts [10, 17, 209-214]. In a study by Chen et al., it has been shown that the addition of nitrogen doped carbon nanotubes (NCNT) as support for MnO2 nanotubes could significantly boost both ORR and OER electrocatalytic activity of the catalyst, i.e. compare ORR overpotentials of -832 and -365 mV (at -2 mA cm-2) for MnO2 and MnO2-NCNT, respectively, as well as OER overpotentials of 482 and 395 mV (at 2 mA cm-2) for MnO2 and MnO2-NCNT, respectively [72]. They have also reported better durability for MnO2-NCNT in a homemade Zn-air battery [72].  1.3 Knowledge gap and research objectives 1.3.1 Knowledge gap The knowledge gap in the literature can be summarized as follows: 1) Talking about an ideal bifunctional catalyst for both ORR and OER, it is important to look at the initial stage electrocatalytic activity and long-term stability of the oxygen catalysts as inter-connected properties rather than two distinct un-related ones. This approach is not the case for most of the studies in the literature on non-PGM oxides (such as manganese oxides, 62  perovskites and fluorite-type oxides and their combinations) as ORR/OER bifunctional catalysts in alkaline media. Often times, the focus of these studies is only on the initial activity, neglecting the long-term durability of these oxides during ORR and OER. Moreover, it seems that there is a lack of sufficient experimental data on the structural evolution of the aforementioned non-PGM oxides to propose and support degradation mechanisms explaining the ORR/OER performance loss of these catalysts in operation. 2) The roles of surface treatments and possible elemental doping on the ORR/OER electrocatalytic activity and durability of aforementioned non-PGM oxides are other topics that have not been intensively investigated in the literature. 3) The studies on the electrosynthesis of manganese oxides, with diverse morphologies and crystal structures, as ORR/OER bifunctional catalysts neglect the complex interactions between different electrodeposition operating factors, e.g. temperature, applied current, Mn concentration, surfactant type and concentration, etcetera. The interactions between these factors need to be investigated comprehensively and systematically using factorial design studies to give better predictions for the effects of these tangled factors on the final electrochemical performance of deposited materials. 1.3.2 Research objectives The main goal of this study is to develop electrochemically active and durable MnOx-based bifunctional catalysts for both ORR and OER in alkaline media, by first, incorporating active OER co-catalysts and second, microstructural and surface modifications of manganese oxides using surfactant-assisted electrodeposition methods as well as alkali-metal ion intercalation techniques. The objectives can be further detailed as follows:  63  1) A comprehensive study on the ORR/OER electrocatalytic activity and durability of commercial MnOx/co-catalyst electrodes: a. To investigate the synergistic effects of synthesized active OER co-catalysts, e.g. LaCoO3, LaNiO3 and Nd3IrO7, when physically mixed with commercial MnO2 for the oxygen electrocatalysis. b. To give an insight on the structural changes of the prepared catalysts during ORR and OER using X-Ray Diffraction (XRD), X-Ray Photoelectron Spectroscopy (XPS), Electron Energy Loss Spectroscopy (EELS), Scanning Electron Microscopy (SEM) and Transmission Electron Microscopy (TEM) techniques.  c. To study the ORR/OER performances of individual and mixed-oxide catalysts using fundamental-study methods (flooded test setup) and commercial test protocols (flow-by Gas Diffusion Electrode (GDE) half-cell test setup). 2) A study on the effect of surface modifications on the ORR/OER electrocatalytic activity and durability of mixed oxides: a. To investigate the effect of alkali-metal ions on the ORR/OER performance of oxide catalysts. b. To develop time- and cost-effective alkali-metal ion intercalation methods for employing any beneficial effects provided by alkali-metal ion intercalation on the ORR and OER performance of oxide catalysts.  3) A systematic study on finding an electrochemically active nanostructured manganese oxide as a non-PGM binder-free ORR/OER bifunctional electrode, synthesized via anodic electrodeposition method on a pre-treated carbon cloth: 64  a. To investigate the main effects and important interaction effects of key operating factors, i.e. Mn2+ concentration, applied potential, temperature, surfactant type and concentration, that significantly influence the electrosynthesis of manganese oxides on the catalyst response for ORR and OER using a two-level half-fraction factorial design.  4) An investigation on the carbonaceous materials as catalyst support/additive for mixed-oxide non-PGM oxygen catalysts or as lone ORR/OER bifunctional catalyst: a. To study the electrochemical behavior of four carbonaceous materials, i.e. commercial Vulcan XC-72, commercial multi-walled carbon nanotubes (MWCNT), in-house made graphene and N-doped graphene, while catalyzing both ORR and OER or supporting highly active bifunctional non-PGM oxides, i.e. MnO2-LaCoO3. b. To find a cost-effective, active and durable catalyst/support combination for oxygen cathodes in alkaline metal-air batteries, regenerative fuel cells and electrolyzers. 65  Chapter 2: Experimental methods, apparatus and materials 2.1 Material preparation 2.1.1 Catalyst powders 2.1.1.1 Perovskites LaNiO3 and LaCoO3 were synthesized via a co-precipitation method [215]. The LaNiO3 powder was made by preparing a solution of 0.2 M lanthanum (III) nitrate hexahydrate (Sigma-Aldrich) and 0.2 M nickel(II) nitrate hexahydrate (Sigma-Aldrich) and adding ammonium hydroxide 30% (Fisher Scientific) as a precipitating agent until the pH reached 9.25. The solution was then heated for 2 hrs at 343 K followed by a heating sequence of: 3 hrs at 383 K, 1 hr at 573 K and another 2 hrs at 973 K in air using a box furnace (Pair equals to 1 atm). Afterwards, the sample was left to cool to room temperature in the furnace. The heating rate for all segments was set at 5 K min-1. For LaCoO3 synthesis, a similar procedure was carried out except that Ni(NO3)2·6 H2O was replaced with cobalt(II) nitrate hexahydrate (Sigma-Aldrich). 2.1.1.2 Nd3IrO7 Nd3IrO7 was made by a direct solid-state synthesis method [195]. Neodymium (III) oxide (Sigma-Aldrich) and iridium metal (Alfa Aesar) powders were mixed with a molar ratio of 1:1 in a glass mortar. The mixture was then heated for 12 hrs at 1323 K in an oxygen atmosphere (PO₂ equals to 1 atm) using a tube furnace and then left to cool down to room temperature in the furnace. Afterwards, the sample was grinded and heated again for 15 hrs at 1323 K under oxygen. The last step was cooling the sample in the furnace. The heating rate for all segments was kept at 5 K min-1. To avoid pyrochlore-type compound formation, i.e. Nd2Ir2O7, the excess oxygen from oxygen atmosphere was crucial during heat treatments [216]. 66  2.1.1.3 Manganese dioxide Manganese (IV) dioxide (reagent grade, ≥90%) was purchased from Sigma-Aldrich. This MnO2 is structurally a -MnO2, i.e. an intergrowth of pyrolusite (-MnO2) into a ramsdellite (-MnO2) matrix, and has higher ORR electrocatalytic activity in alkaline media compared to other commercially readily available MnO2 samples [90]. 2.1.1.4 Carbonaceous materials The in-house-made graphene and N-doped graphene sheets were acquired from Gyenge’s lab at the University of British Columbia. They were made using electrochemical exfoliation of graphite, assisted by ionic liquids (ILs), extensively explained elsewhere [217]. Vulcan XC-72 was acquired from Cabot. MWCNT (>95%) was purchased from Sigma-Aldrich. 2.1.1.5 Platinum 50 wt% Pt on Graphitized Carbon was acquired from Tanaka Kikinzoku Kogyo K.K. in Japan. Pt powder (assay 98%) was also purchased from Alfa Aesar.  2.1.2 Catalyst layer preparation To check the electrochemical performance of aforementioned carbonaceous materials (2.1.1.4) as support§§ for MnO2-LaCoO3 (weight ratio of 1:1) or oxygen catalyst alone, a mixture of oxides (if present), carbonaceous material(s), isopropyl alcohol (IPA) and 5 wt% Nafion solution was sonicated for 1 hr. Next, a specified volume of the catalyst ink (5-15 µL) was drop-casted on a polished glassy carbon (GC) electrode to reach a loading of 0.5 mg cm-2 for the carbonaceous material(s) or oxides, whichever was present, and left in air to dry for another hour                                                  §§ Note that throughout the entire experimental work performed in this thesis, the carbonaceous materials labeled as support for a non-PGM catalyst were added to the catalyst ink prior to its deposition on a substrate. 67  at 293 K. The final weight ratio of MnO2 (if present):LaCoO3 (if present):carbon(s):Nafion in the catalyst layer was 1:1:1:0.6. In all cases where more than one carbonaceous material was used, either as the catalyst or support, the loading of each carbon component was kept at 0.5 mg cm-2. 2.1.3 Gas diffusion electrode preparation Catalyst inks were prepared by 1 hr sonication of a mixture composed of catalyst material(s) (i.e. Pt, MnO2, LaCoO3, LaNiO3 and Nd3IrO7), carbon(s), IPA, water, 5 wt% Nafion solution and 60 wt% polytetrafluoroethylene (PTFE) suspension at 293 K. The carbon:isopropanol:water weight ratio was fixed at 1:50:16 in all catalyst inks based on previous studies aimed at finding the right catalyst ink composition for spraying using the CNC sprayer machine (Figure 2.1). The PTFE and dry Nafion content of the catalyst layer was the same for all samples, namely 0.3 mg cm-2 each. The catalyst inks were then sprayed on a 4×4 cm (16 cm2 geometric area) piece of 40 wt% PTFE treated carbon cloth from Fuel Cell Earth Co. using the CNC sprayer machine (Figure 2.1) to achieve main catalyst (i.e. MnO2 or Pt) and co-catalyst (i.e. LaCoO3, LaNiO3 and Nd3IrO7, if present) loadings of 0.5 mg cm-2 each. For GDE half-cell experiments in a flow-by test cell (Gaskatel Half Cell HZ-PP01), a PTFE treated carbon substrate supported on a Ni mesh (as current collector) from ZincNyx Energy Solution Inc. was employed to stand the vigorous GDE durability testing protocols during the course of 48 hrs at 323 K and Pgas of 1 atm (either O2 or purified air). In these cases, the final weight ratio of main catalyst (i.e. MnO2 or Pt):co-catalyst (i.e. LaCoO3, LaNiO3 and Nd3IrO7, if present):Vulcan XC-72:Nafion:PTFE in the catalyst layer was 1:1:2:0.6:0.6. Catalyst ink preparation and spraying procedures were similar to the aforementioned steps. The final loadings of main catalysts, i.e. MnO2 or Pt, were set at 2 and 0.5 mg cm-2, respectively.  Figure 2.2 shows 68  a representative image of the GDEs with catalyst inks sprayed on 40 wt% PTFE treated carbon cloth from Fuel Cell Earth Co. and ZincNyx’s PTFE treated carbon substrate. Further specifications of the employed GDEs are described in each corresponding chapter.   Figure 2.1 CNC controlled sprayer machine with IWATA air brusher (50 ml capacity).  69   Figure 2.2 Gas diffusion electrodes consist of catalyst inks sprayed on: 1) 40 wt% PTFE treated carbon cloth from Fuel Cell Earth Co. and 2) PTFE treated carbon substrate supported on a Ni mesh (as current collector) from ZincNyx Energy Solution Inc..  2.1.4 Anodic electrodeposition of manganese oxides Manganese oxides were electrodeposited onto a 6 mm in-diameter 40 wt% PTFE treated carbon cloth substrate (from Fuel Cell Earth Co.). The investigated electrodeposition factors and corresponding ranges are outlined in Table 2.1. Three different types of surfactants were used, i.e. Sodium dodecyl sulfate (SDS) as anionic, hexadecyl-trimethyl-ammonium bromide (CTAB, also known as cetrimonium bromide) as cationic and Triton X-100 as non-ionic surfactants. The critical micelle concentration (CMC) values of SDS, CTAB and Triton X-100 at 298 K are 7-10, 1-3 and 70  0.2-0.3 mM ***. The surfactant concentrations employed in this study, i.e. 5 or 10 vol%, were over the CMC in all cases. The electrolyte solution contained various concentrations of manganese (II) acetate tetrahydrate (Mn(CH3COO)2.4H2O) and 0.1 M sodium sulphate (Na2SO4) solution. A half-fraction 2n factorial design was constructed using the statistical software JMP 11. For a half-fraction factorial design of four factors (Table 2.1) with three center-points, the number of experimental runs required for each surfactant type was 11, compiling to a total of 33 random runs for the entire screening design experiments.   Table 2.1 Experimental design factors and their levels for 24-1+3 factorial design runs.  Factors (symbol, unit)* Levels  Low (-) Center (0) High (+) Mn(CH3COO)2.4H2O concentration (C, M) 0.1 0.2 0.3 Temperature (T, K) 295 319 343 Surfactant concentration (S, vol%) 0 5 10 Applied anodic potential (E, mVMOE) 800 1200 1600 * Pair is 1 atm. Prior to the electrodeposition process, the carbon substrate was pre-treated using nitric acid to reduce the hydrophobicity of the carbon cloth and remove any remaining impurity as well as surface oxides on the carbon fiber surfaces. The 40 wt% PTFE treated carbon cloth was dipped in acetone for 5 min and then washed thoroughly with DI water. Next, the substrate was soaked in 1 M nitric acid at 333 K for 30 min. Afterwards, the samples were washed thoroughly with 18 mΩ DI water and left to dry overnight at 343 K in an oven.                                                   *** All CMC values are obtained from Sigma-Aldrich website, the supplier of the three surfactants. 71  A conventional three-electrode electrochemical half-cell setup was used for the electrodeposition process (Figure 2.3). The working electrode was a punch-cut circular pre-treated carbon cloth piece with geometric surface area of 0.283 cm2 in a quick-fit exchangeable sample holder from Radiometer Analytical (#A35T450), attached to a rotating disk electrode (RDE) setup (Figure 2.4). The reference and counter electrodes were Hg/HgO/20 wt% KOH (MOE) and platinized titanium plate, respectively. The electrodes were connected to a computer-controlled VoltaLab 80 potentiostat in its associated RDE setup (Figure 2.5). The anodic electrodeposition was performed under different conditions as outlined in Table 2.1 using a potentiostatic method at rotation speed of 400 rpm and Pair of 1 atm for 30 minutes per each run. After the completion of electrodeposition process, the working electrode was washed thoroughly with DI water. In the case where a surfactant was used, the surfactant residue was removed by dipping the sample in IPA at 343 K for 15 minutes at 400 rpm. The catalyst-coated carbon cloth was then rinsed with DI water again.   72   Figure 2.3 Schematic diagram of three-electrode electrochemical half-cell setup used in this study.   Figure 2.4 Components of a quick-fit exchangeable sample holder (left) from radio Radiometer Analytical (#A35T450) with sample opening of 6 mm in diameter. The samples (right) were placed on a glassy carbon disk as a backing layer prior to be placed on the tip. 73   Figure 2.5 Computer-controlled VoltaLab 80 potentiostat with its associated RDE setup and the electrochemical three-electrode cell.  2.1.5 Surface modification: K+ intercalation  Two methods of K+ intercalation was investigated: open-circuit potential intercalation (OCP) and potential driven (electrophoretic) intercalation (PDI). In the open-circuit potential (OCP) method, each GDE was exposed to 6 M KOH solution for up to six days at 313 K under a rotation speed of 400 rpm. The samples were then thoroughly washed in 18 mΩ DI water for further electrochemical investigations. The same OCP method was also applied using LiOH, NaOH and CsOH to study comparatively the effect of exposure of the oxide catalysts to diverse alkali-metal ions. 74  In the potential driven intercalation (PDI) method, a constant cathodic current density of -5.4 mA cm-2 was applied for 30 min. to the electrodes under investigation in the RDE setup (at 400 rpm) in a 0.036 M K2SO4 solution at 343 K. The cathodic current density was selected such that to provide the necessary potential gradient for K+ migration toward the cathode while avoiding excessive H2 gas evolution. A platinum plate was used as a counter electrode. Next, the samples were thoroughly washed in DI water before further electrochemical investigations. The PDI procedure was repeated up to seven times to investigate the cumulative effect of the treatment method on the bifunctional performance. Each repeated PDI treatment was carried out using fresh K2SO4 solution. 2.2 Surface and structural characterization  The catalyst powders as well as GDEs were characterized by one or more of the following techniques: X-Ray Diffraction (XRD, D8 Advance Bruker diffractometer), X-ray Photoelectron Spectroscopy (XPS, Leybold Max 200 and Kratos AXIS Ultra), Energy Dispersive X-ray analysis (EDX, Hitachi S-2600N Variable Pressure Scanning Electron Microscope (VPSEM) equipped with an X-ray detector), Electron Energy Loss Spectroscopy (EELS, FEI Titan 80-300 LB equipped with a energy loss spectrometer Gatan 865 model), Field Emission Scanning Electron Microscopy (FESEM, Hitachi S-4700), Transmission Electron Microscope (TEM, FEI Tecnai G2 200kV), Fourier Transfer-Infrared spectroscopy with Attenuated Total Reflectance (FT-IR ATR, PerkinElmer Frontier) and Brunauer-Emmet-Teller analysis (BET, Micromeritics ASAP 2020 Accelerated Surface Area and Porosimetry Analyzer). The operating conditions for XRD were as follows: generator set at 40 kV and 40 mA; Cu as X-ray source; wave length of 1.54439 Å Kα1; step size of 0.04˚ (2θ); step time of 230.4 s; range: between 5˚ to 90˚ for 2θ. The XPS source was monochromatic Al Kα. The manganese oxidation state was determined from the multiplet splitting 75  of Mn 3s and the corresponding separation of peak energies at the XPS spectrum for the electrodeposited samples only. The EDX accelerating voltage was 10 kV.  FT-IR analysis was used to confirm the efficiency of the surfactant removal technique on the catalyst-coated carbon clothes. The FT-IR ATR operating conditions were as follows: an aperture setting of 4 cm-1; wavenumber range of 4000-600 cm-1; 64 scans per sample at a scan speed of 2.5 kHz. 2.3 Electrochemical measurements A conventional three-electrode RDE half-cell setup was used for the electrochemical analysis of the catalyst materials (Figure 2.6). Two different types of working electrodes were employed to test the electrocatalytic activity and durability of materials investigated here:  1) Catalyst ink drop-casted on a 5-mm-in-diameter polished GC. 2) A circular 8-mm-in-diameter punch-cut GDE fitted in a quick-fit exchangeable sample holder from Radiometer Analytical (#A35T450) with a geometric area of 0.283 cm2 exposed to the electrolyte. The reference and counter electrodes were Hg/HgO/0.1 M KOH (MOE) from Radiometer Analytical (XR400) and platinum mesh, respectively. The electrodes were connected to a computer-controlled VoltaLab 80 potentiostat in its associated RDE setup (Figure 2.5). The potential of MOE reference electrode was measured to be 977 and 1037 mV vs. RHE in 6 and 11.7 M (45 wt%) KOH solutions at 293 K, respectively, using the reversible hydrogen reference electrode (HydroFlex) from Gaskatel GmbH. All potentials are reported vs. RHE unless otherwise specified. All currents are normalized by the geometric surface area of the electrodes. All gas pressures specified in this study are absolute pressures. The equilibrium oxygen electrode potential in 6 and 11.7 M KOH solutions was calculated to be 1168 mVRHE (191 mVMOE) and 1153 mVRHE (116 mVMOE), respectively, at 293 K. 76    Figure 2.6 A Pine jacketed electrochemical cell connected to a water bath in the three-electrode RDE half-cell test setup used for electrochemical measurements in this study.  To study the electrode kinetics, rotating ring disc electrode (RRDE) measurements were performed by a Pine RRDE electrode (AFE6R1GCPK) with a GC disk (5.7 mm in diameter) and ring (6.4 mm inner and 7.8 mm outer diameter) using linear sweep voltammetry (LSV) tests in O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm, starting with cathodic polarization from 1173 to 398 mV at 5 mV s-1 and various rotation speeds (check Appendix A  and Appendix B  ). The collection efficiency of the RRDE was 0.34 and the potential of the ring electrode was held at 1353 mV during the RRDE measurements.  77  To investigate the electrocatalytic activity and durability of the catalysts, cyclic voltammetry (CV) tests up to one hundred successive potential cycles were employed on catalyst materials, either deposited on GC or 40 wt% PTFE treated carbon cloth, in 6 M KOH between 373 and 1823 mV, starting with anodic polarization, using a flooded test setup at 5 mV s-1, 293 K, 400 rpm and Pgas of 1 atm (either O2 or N2). In some cases, shorter potential range was adopted to avoid catalyst loss, specially for samples deposited on GC.  To further study the durability of the catalysts, galvanostatic polarization experiments (i.e. chronopotentiometry) were performed on GDEs as working electrodes in O2 saturated 6 M KOH at 293 K, 400 rpm and PO₂ of 1 atm by applying a constant current density (per geometric area) of 5 mA cm-2 for 2 hrs followed by -2 mA cm-2 for 30 min. The current densities were chosen to avoid mass transport limitations in the flooded electrode half-cell arrangement used in this study during galvanostatic longer-term experiments.  To test the ORR performance and stability of catalyst oxides developed here in commercial scenarios, a galvanostatic test protocol (i.e. chronopotentiometry) was employed in a flow-by test cell from Gaskatel GmbH (Half Cell HZ-PP01) as shown in Figure 2.7. Fresh GDEs composed of oxide containing catalyst layer on ZincNyx’s carbon substrate (see 2.1.3) were polarized at -100 mA (-34 mA cm-2) for 24 hrs followed by -120 mA (-40 mA cm-2) for another 24 hrs in 11.7 M (45 wt%) KOH at 323 K with air (CO2 removed) flowing through the gas chamber. In presence of pure oxygen, constant currents of -200 mA (-67 mA cm-2) for 24 hrs followed by -300 mA (-100 mA cm-2) for another 24 hrs were applied on the fresh electrodes with the same 78  conditions as air. The absolute gas pressure and flow rate were fixed at 1 atm and 1.51×10-3 SLPM†††.  Prior to the reported electrocatalytic performance tests, each electrode was subjected to a break-in polarization protocol composed of five potential cycles between 233 and 1683 mV at 5 mV s-1 and 400 rpm, starting with anodic polarization. For GDEs in the Gaskatel flow-by cell, the potential cycle for break-in polarization protocol was started at 373 mV to 1173 mV. All cyclic and linear sweep voltammograms have been repeated for at least five times to ensure the reproducibility of the presented results. For the galvanostatic measurements in the flooded test setup, min. three replicates were produced for each catalyst. In the case of ORR galvanostatic polarizations for 48 hrs in the flow-by cell, min. two replicates were employed to calculate the standard error of the mean for each sample.    In an electrochemical cell under operating conditions, the movement of ionic species in the electrolyte can bring about a number of sectional potential drops between the working and reference electrodes. When current is flowing between the working and counter electrodes, the movement of anions and cations at unequal rates in opposing directions under the imposed electric field sets up a total ohmic potential drop (IR-drop) in the electrolyte, mainly due to electrode/electrolyte contact resistance, ionic resistance of the electrolyte and diffusion limitations at the porous frit tips of the Luggin-Haber capillary and the reference electrode [218, 219].  Here, all cyclic and linear sweep voltammograms as well as galvanostatic polarization results are IR-drop corrected using “Static Manual” ohmic drop compensation feature of the VoltaLab 80 potentiostat.                                                   †††  Standard liter per minute (SLPM): Volumetric flow rate of a gas corrected to "standardized" conditions of temperature and pressure, i.e. temperature of 273.15 K and an absolute pressure of 100 kPa. 79    Figure 2.7 Flow-by electrochemical test cell (HZ-PP01) and its components from Gaskatel GmbH used for ORR GDE tests: PTFE body, built-in platinum counter electrode, built-in Luggin-Haber capillary for the reference electrode.   80  Chapter 3: Comprehensive studies on the ORR/OER electrocatalytic activity and durability of individual (MnO2, LaCoO3, LaNiO3 and Nd3IrO7) and mixed-oxide (MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7) catalysts ‡‡‡ 3.1 Introduction Non-precious metal bifunctional oxygen reduction and oxygen evolution reactions catalysts are of great interest for rechargeable metal-air batteries and regenerative alkaline fuel cells. In this chapter, both the initial stage electrocatalytic activities and the catalytic durability of novel bifunctional catalysts composed of MnO2 with perovskite (LaCoO3 or LaNiO3) or fluorite-type oxide (Nd3IrO7) as co-catalysts were studied. Gas diffusion electrodes (GDE) with a catalyst layer composed of MnO2:co-catalyst (LaCoO3, LaNiO3 or Nd3IrO7):Vulcan XC-72 were prepared and studied in alkaline media using two different cell configurations, i.e. flooded and flow-by test setups. The catalyst powders were carefully characterized before the electrochemical tests.  The initial stage bifunctional activities of mixed-oxide catalysts, i.e. MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7 are markedly superior compared to either MnO2 or individual co-catalysts, demonstrating a synergistic effect. However, the activity degradation during extensive potential cycling is more severe for all mixed-oxide formulations than MnO2 itself. MnO2-LaNiO3 revealed the best initial stage ORR/OER electrocatalytic activity whereas showing the worst long-                                                 ‡‡‡ Parts of this chapter have been published and filed as United States and Canadian patent applications: - P. Hosseini-Benhangi, A. Alfantazi, E. Gyenge, “Manganese dioxide-based bifunctional oxygen reduction/evolution electrocatalysts: Effect of perovskite doping and potassium ion insertion”, Journal of Electrochimica Acta, 123 (2014) 42-50. - P. Hosseini-Benhangi, M. A. Garcia-Contreras, A. Alfantazi, E. Gyenge, “Method for enhancing the bifunctional activity and durability of oxygen electrodes with mixed oxide electrocatalysts: Potential driven intercalation of potassium”, Journal of The Electrochemical Society, 162 (2015) F1356-F1366. - E. Gyenge, P. Hosseini-Benhangi, “An oxygen electrode and a method of manufacturing the same”, U.S. (15/251,267) and Canadian (2,940,921) patent applications, filed on August 30th, 2016.  81  term ORR/OER performance during accelerated degradation testing of one hundred cycles in flooded test setup.  Long-term ORR galvanostatic polarization curves of fresh MnO2 and mixed-oxide GDEs compared with commercial Pt and MnOx were obtained in a flow-by cell over the course of 24 hrs in 11.7 M KOH at 323 K and Pgas of 1 atm with either air (CO2 removed) or oxygen to resemble the commercial scenarios. MnO2-Nd3IrO7 possessed the lowest ORR overpotentials with both air and oxygen flowing through the gas chamber, followed by MnO2-LaCoO3 as the second best, both outperforming the commercial Pt and MnOx GDEs. An insight on the structural changes of GDEs during these long-term ORR durability tests was given. Further in-depth structural studies on the investigated catalysts during both ORR and OER are given in the next chapters. 3.2 Results and discussions 3.2.1 Characterization of oxide catalysts Figure 3.1 shows representative SEM images of the synthesized LaCoO3, LaNiO3, Nd3IrO7 and commercial MnO2 powders used in the catalyst layer preparation. The Sigma-Aldrich MnO2 powder clearly shows agglomerates of pillar and sphere-like particles (Figure 3.1-A). The size range of these agglomerates was found to be between 0.5 and 3 µm based on the SEM images (Figure 3.1-A). The synthesized perovskite powders, i.e. LaCoO3 and LaNiO3, have shown flaky and porous morphology with sharp edges (Figure 3.1-B and-C). Nd3IrO7, on the other hand, possesses particles with soft edges and round shapes comparing to perovskites’ morphology (Figure 3.1-D). All particles have shown great tendency to form agglomerates.    82     Figure 3.1 SEM images of A) Commercial Sigma-Aldrich MnO2 powder (pillar and sphere-like particles are shown using dashed oval and circle shapes, respectively), B) Synthesized LaCoO3 powder, C) Synthesized LaNiO3 powder and D) Synthesized Nd3IrO7 powder. The black arrow in B and C points out the flakes in SEM morphology of studied oxides.  The structural analysis of synthesized LaCoO3, LaNiO3, Nd3IrO7 and commercial MnO2 powders is presented in Figure 3.2 and Table 3.1, showing major peaks for the corresponding oxides. The diffraction pattern of Sigma-Aldrich MnO2 reveals that it can be structurally characterized as γ-MnO2, an intergrowth of pyrolusite in the ramsdellite matrix (Figure 3.2-A and Table 3.1) [39].  83  The EDX elemental analysis obtained from commercial and synthesized powders, previously shown in the SEM images of Figure 3.1, further confirms the existence of constituent elements for compounds reported by XRD results of each powder (Figure 3.2), i.e. Mn/O, La/Co/O, La/Ni/O  and Nd/Ir/O for commercial manganese oxide, synthesized lanthanum cobalt oxide, lanthanum nickel oxide and neodymium iridium oxide, respectively (Figure 3.3). Due to the presence of other compounds and impurities in the synthesized powders, the atomic percentage of each element calculated form EDX results cannot be linked to the chemical formula found for major compounds in each powder via XRD analysis.   84    Figure 3.2 XRD spectra of: A) Commercial MnO2, B) Synthesized LaCoO3, C) Synthesized LaNiO3 and D) Synthesized Nd3IrO7. (*), (▼), ( ▌) and (♦) present major peaks corresponding to MnO2, LaCoO3, LaNiO3 and Nd3IrO7, respectively. 85  Table 3.1 XRD structural analysis of: Commercial MnO2, synthesized LaCoO3, synthesized LaNiO3 and synthesized Nd3IrO7. The XRD spectrum of the powders are presented in Figure 3.2. The bold 2θ and Miller indices show overlapped peaks for different crystallographic plans. Catalyst powder 2θ (hkl) Crystal structure MnO2 1) 28.5˚ (110), 37.5˚ (101), 41˚ (200), 43˚ (111), 56.5˚ (211), 59.5˚ (220), 65˚ (002), 67.5˚ (310) and 72˚ (301) (marked as (*) in Figure 3.2-A) 2) 33˚ (marked as (►) in Figure 3.2-A) 3) 37.5˚ (201), 43˚ (211) and 56.5˚ (221)  1) Pyrolusite (β-MnO2) phase with a tetragonal Bravais lattice system (a=b=4.3999 Å, c=2.8740 Å) 2) Mn2O3 3) Ramsdellite (α-MnO2) LaCoO3 1) 23.3˚ (012), 33˚ (110), 33.5˚ (104), 40˚ (202), 47.5˚ (024) and 59˚ (214) (marked as (▼) in Figure 3.2-B) corresponding to (012), (110), (104), (202), (024) and (214) 2) 15.6˚, 27.3˚, 28˚, 39˚ and 48.5˚ (marked as (●) in Figure 3.2-B) 1) LaCoO3 with a rhombohedral Bravais lattice system (a=b=c=5.3778 Å, α=β=γ=60.798˚) 2) La(OH)3 with a hexagonal Bravais lattice system (a=b=6.5286 Å, c=3.8588 Å, α=β=90˚, γ=120˚) LaNiO3 1) 23.5˚ (012), 33˚ (110), 47.5˚ (024) and 59˚ (214) (marked as ( ▌) in Figure 3.2-C) 2) 15.6˚, 27.3˚, 28˚, 39˚ and 48.5˚ (marked as (●) in Figure 3.2-C) 1) LaNiO3 with rhombohedral Bravais lattice system (a=b=5.4510 Å, c=6.5640 Å, α=β=90˚, γ=120˚) 2) La(OH)3 with a hexagonal Bravais lattice system (a=b=6.5286 Å, c=3.8588 Å, α=β=90˚, γ=120˚) Nd3IrO7 1) 14.4˚ (110), 26.7˚ (021), 28.9˚ (220)/(202), 31.4˚ (221), 32.8˚ (400), 33.9˚ (022), 47.9˚ (422), 48.5˚ (004), 56.3˚ (620)/(602) and 57.6˚ (224)/(531) (marked as (♦) in Figure 3.2-D) 1) Nd3IrO7 with an orthorhombic Bravais lattice system (a=10.8903 Å, b=7.4400 Å, c=7.4893 Å, α=β=γ=90˚) 86    Figure 3.3 EDX spectra of commercial and synthesized catalyst powders: A) MnO2, B) LaCoO3, C) LaNiO3 and D) Nd3IrO7.  3.2.2 Initial stage electrocatalytic activity of oxide catalysts Figure 3.4 presents the cyclic voltammograms of the investigated oxide electrodes recorded in N2 saturated 6 M KOH. The upper potential limit in Figure 3.4 was selected such that to be lower than the oxygen equilibrium potential in order to reveal at this stage only the intrinsic responses of the oxides themselves and to avoid as much as possible interferences by dissolved oxygen. The reduction waves for MnO2 and Nd3IrO7 reach their respective peak currents at 300 and 500 mV, respectively (Figure 3.4-A). The features of MnO2 cyclic voltammograms and the 87  role of Mn4+, Mn3+ and Mn2+ species, in the 2×2 e− ORR electrocatalysis was extensively discussed in section 1.2.2.2.  For iridium oxide compounds with structures related to Nd3IrO7, similar voltammetry response to that shown by Figure 3.4-A at high pH was attributed to the Ir5+/Ir4+ couple [126]. The reduction onset potential for both MnO2 and Nd3IrO7 is the same, about 750 mV. Compared to MnO2 and Nd3IrO7, either oxidation or reduction waves associated with perovskites alone, i.e. LaCoO3 and LaNiO3, are virtually absent (Figure 3.4-A), corroborating previous reports of sluggish intrinsic electron transfer to or from LaCoO3 and LaNiO3 [123, 220].  For MnO2-LaCoO3 and MnO2-Nd3IrO7 (Figure 3.4-B), the reduction peak potential is observed at about 480 mV, which is characteristic mainly for Mn4+/Mn3+ reduction as opposed to Mn3+/Mn2+ reduction occurring at lower potentials (i.e. 300 mV in Figure 3.4-A). Furthermore, the reduction current densities for these two mixed oxides were larger than for each of the individual components, suggesting more extensive reduction in the catalyst layer (of mostly MnO2 and Nd3IrO7 where applicable). The MnO2-LaNiO3, however, shows a weak reduction wave starting at about 625 mV with no characteristic peak, mainly due to the scarce reduction of Mn4+ to Mn3+ and later Mn2+ (Figure 3.4-B). The low reduction current densities of MnO2-LaNiO3 comparing to MnO2 and the other two mixed-oxide formulations infer low reduction of its components, supposedly lower than MnO2 alone too (Figure 3.4-A and B).   88    Figure 3.4 IR-corrected cyclic voltammograms of GDEs with MnO2, LaCoO3, LaNiO3, Nd3IrO7, MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7 catalysts. A) MnO2, LaCoO3, LaNiO3 and Nd3IrO7, B) MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7. Electrolyte: N2 saturated 6 M KOH at 293 K and PN₂ of 1 atm. The oxide loadings were 0.5 mg cm-2 each. Rotating electrode speed and potential scan rate were 400 rpm and 5 mV s-1, respectively. Cycle number five is reported in all cases.  89  Figure 3.5 presents representative SEM micrographs of the mixed-oxide GDEs studied here with MnO2:LaCoO3:Vulcan:Nafion:PTFE (weight ratio of 1:1:1:0.6:0.6) as catalyst layer, sprayed on 40% PTFE treated carbon cloth. The SEM images reveal homogenized catalyst layer with an average thickness of about 25 µm, attached to carbon fibers as the substrates (Figure 3.5).     Figure 3.5 SEM images of the GDE consisting of MnO2:LaCoO3:Vulcan XC-72:Nafion:PTFE (weight ratio of 1:1:1:0.6:0.6) sprayed on 40% PTFE treated carbon cloth.  90  Next, the BET surface areas of catalyst layers investigated here are reported in Table 3.2. The oxide(s) and Vulcan XC-72 were physically mixed with a weight ratio of 1:1 or 1:1:1 (in case of mixed oxides) and sonicated for 1 hr in IPA. Each solution was then left in open air to dry over night before taking to the BET analyzer. The mixed oxide formulations have very similar BET surface areas between 53.7 and 55.1 m2 g-1.  With regard to Figure 3.4, it is noted that in a few reports, the electric double-layer portion of cyclic voltammograms for various oxides was used to estimate the electrochemically active surface area (ECSA) of these oxide catalysts [65, 131]. However, equating the total charged surface area obtained from electric double-layer capacitance measurements with the area of bifunctionally active sites for oxides with complex structures and involving various oxidation states with different activities, is unwarranted. In such cases, with respect to ECSA, the area obtained from electric capacitance measurements is hardly more accurate than the total BET area (Table 3.2). Hence, none of them are utilized here to represent the ECSA.    Table 3.2 BET surface area of single and mixed oxide catalyst layers. Catalyst layer (component weight ratios 1:1 or 1:1:1) BET Surface Area (m2 g-1oxide) MnO2-Vulcan 84.6 LaCoO3-Vulcan 107.5 LaNiO3-Vulcan 106.7 Nd3IrO7-Vulcan 104.6 MnO2-LaCoO3-Vulcan 55.2 MnO2-LaNiO3-Vulcan 54.3 MnO2-Nd3IrO7-Vulcan 53.7  91  The polarization curves for ORR and OER were recorded by potential scanning between 233 to 1683 mV in O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm, with a scan rate of 5 mV s-1. The results, representative for the initial stage electrocatalytic activity, are presented as Tafel plots in Figure 3.6, whereas Table 3.3 summarizes the calculated apparent exchange current densities and Tafel slopes (check Appendix A  for more details on the electrode kinetic theory). With respect to ORR, among the individual oxides, MnO2 has the highest activity, followed by LaCoO3, LaNiO3 and lastly Nd3IrO7 (Figure 3.6-A). Considering for comparison an overpotential of -350 mV, the ORR current density on MnO2 was 3, 9 and 20 times the ones on LaCoO3, LaNiO3 and Nd3IrO7, respectively. Furthermore, the combination of MnO2 with Nd3IrO7 (in a 1:1 weight ratio) increased the apparent exchange current density compared with either of the individual oxides, i.e. by 1.25 times vs. MnO2 alone and by over two-orders of magnitude vs. Nd3IrO7 alone (Table 3.3). The Tafel slope of the mixed oxide MnO2-Nd3IrO7 catalyst remained virtually the same as for MnO2 alone. The synergistic effect between MnO2 and Nd3IrO7 impacting the apparent exchange current density, brought about the highest initial stage ORR current densities among the investigated catalysts followed by MnO2-LaNiO3 for overpotentials greater than -500 mV (Figure 3.6-A).  At overpotentials lower than -500 mV, the MnO2-LaCoO3 combination prevailed due to lower Tafel slope, i.e. -84 mV dec-1 vs. -125 mV dec-1 for MnO2-Nd3IrO7 (Figure 3.6-A).  In the OER part of the polarization curve, the Tafel lines and the associated kinetic parameters are potential dependent (Figure 3.6-B and Table 3.3), a phenomenon well-known in the literature and usually attributed to changes in the rate determining step [221-223]. At overpotentials lower than 360 mV, the apparent Tafel slopes were between 69 mV dec-1 (LaNiO3) and 115 mV dec-1 (MnO2), whereas above 360 mV the Tafel slopes varied between 103 mV dec-1 (MnO2-LaCoO3) and 201 mV dec-1 (Nd3IrO7). In the high overpotential region, abnormal Tafel slopes such as 147 92  to 201 mV dec-1 in Figure 3.6-B, have been reported by others as well [224, 225], and it is due to hampered growth and detachment of O2 gas bubbles. In other words, at high overpotentials, there are many surface sites available for gas bubble nucleation but due to surface irregularities and other morphological features of the porous electrode causing entrapment, the bubble growth and break-off are inhibited, thereby, shielding the catalytic surface [226]. These effects are manifested as abnormally high apparent Tafel slopes in polarization experiments at high overpotentials. Considering as basis for comparison an OER current density of 10 mA cm-2 as per the benchmarking study of McCrory et al. [131], the corresponding overpotentials on the mixed-oxide catalysts were 425, 440 and 501 mV on MnO2-LaNiO3, MnO2-LaCoO3 and MnO2-Nd3IrO7, respectively (Figure 3.6-B). The latter overpotentials, representative for the initial stage activities, are significantly lower than for any of the individual oxides investigated, demonstrating clearly a strong beneficial synergistic electrocatalytic effect between the two components. Table 3.3 reveals that in the low OER overpotential region, the combination of MnO2 with LaCoO3 decreased the Tafel slope to 69 mV dec-1 from 115 and 98 mV dec-1, respectively, whereas the combination of MnO2 with Nd3IrO7 increased the exchange current density from 0.60 A cm-2 for MnO2 to 0.79 A cm-2 for MnO2-Nd3IrO7. However, MnO2-LaNiO3 possesses the best OER electrocatalytic activity with an OER Tafel slope and exchange current density of 93 mV dec-1 and 0.43 A cm-2, respectively, at overpotentials lower than 360 mV (Table 3.3). The prevailing modern theoretical concept regarding the oxygen electrode mechanism is based on the scaling relationships, indicating that the binding energies of intermediates, such as HOO(ads) and HO(ads), are linearly correlated regardless of the binding site. Both species adsorb on the same sites on the oxide surface with a single bond between O and the surface [124]. This so-called universal scaling relationship, i.e. approximately constant difference of 3.2 eV between the binding 93  energy levels of HOO(ads) and HO(ads) for various catalyst surfaces, leads to minimum “theoretical overpotential” of about 370 mV for the ORR and the OER on a variety of defect free materials. The practical translation of these results for “designing” the oxide catalytic surfaces is complicated by the fact that the synthesized oxides, such as MnO2, have a very complex crystallographic structure (ranging from  to ), with the possibility of intertwined structures, numerous types of vacancies, disorders and lattice defects and changes in the oxidation states during battery cycling [39]. In spite of the virtual impossibility of considering all these effects in a first principles model, a comparison between theoretically calculated and experimentally measured initial stage ORR and OER current densities for -Mn2O3 showed promising fit, specially in the ORR region, whereas some deviations are noted in case of OER [121]. The MnO2 used in the present work is of -type, which is a combination of  and  structures, and it was previously shown to provide good ORR activity compared to other commercial sources of MnO2 [90]. Theoretical studies suggest the need to break the scaling relationship between the HOO(ads) and HO(ads) binding energies in order to improve the bifunctional activity by favoring weaker HO(ads) binding. It is hypothesized that combining oxides with different structural features such as MnO2 and perovskites or MnO2 with fluorite-type structures, provides different binding sites and binding energies for HOO(ads) and HO(ads), that contribute to the observed synergistic electrocatalytic effect presented by Figure 3.6 and Table 3.3.    94   Figure 3.6 Initial stage IR-corrected bifunctional ORR/OER Tafel-lines of GDEs with MnO2, LaCoO3, LaNiO3, Nd3IrO7, MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7 catalysts. A) ORR, B) OER. Electrolyte: O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm. Electrode potential scanning between 233 to 1683 mV. The oxide loadings were 0.5 mg cm-2 each. Rotating electrode speed and potential scan rate were 400 rpm and 5 mV s−1, respectively. Cycle number five is reported in all cases. The numbers associated with each line represent the respective apparent Tafel slopes.95  Table 3.3 The apparent exchange current densities and Tafel slopes for the initial stage ORR and OER activities of the investigated GDEs with fresh catalysts. O2 saturated 6 M KOH. 293 K. PO₂ of 1 atm. The exchange current densities are expressed per geometric area. The standard error of the mean calculated based on six replicates is indicated for each catalyst. Catalyst bOER  (η< 360 mV) (mV dec-1) i0,OER  (η< 360 mV) (µA cm-2) bOER  (η> 360 mV) (mV dec-1) i0,OER  (η> 360 mV) (µA cm-2) bORR (mV dec-1) i0,ORR (µA cm-2) MnO2 115 ±2 0.63 ±0.01 187 ±4 9.0 ±0.2 -125 ±3 0.63 ±0.01 LaCoO3 98 ±2 0.20 ±4×10-3 105 ±2 0.39 ±8×10-3 -126 ±3 0.20 ±4×10-3 LaNiO3 69 ±1 1.5×10-2 ±3×10-4 147 ±3 9.4 ±0.2 -101 ±2 1.5×10-2 ±3×10-4 Nd3IrO7 70 ±1 2.5×10-3 ±0.05×10-3 201 ±4 5.0 ±0.1 -90 ±2 2.5×10-3 ±0.05×10-3 MnO2-LaCoO3 69 ±1 1.0×10-2 ±0.02×10-2 103 ±2 0.56 ±0.01 -84 ±2 1.0×10-2 ±0.02×10-2 MnO2- LaNiO3 93 ±2 0.43 ±9×10-3 129 ±3 5.2 ±0.1 117 ±2 0.43 ±9×10-3 MnO2-Nd3IrO7 108 ±2 0.79 ±0.02 182 ±4 18 ±0.4 -125 ±3 0.79 ±0.02 96  3.2.3 Accelerated degradation testing of oxide catalysts in flooded test setup Long-term durability investigations are of outmost importance for practical applicability of electrocatalysts. Unfortunately, many studies dealing with non-precious metal ORR or OER catalysts in alkaline media have focused only on the initial stage activities of fresh electrodes. The electrocatalytic long-term durability performances of MnO2 (as baseline) and mixed-oxide GDEs were investigated by performing one hundred continuous potential cycles between 633 and 1633 mV in O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm, using a flooded test setup. Figure 3.7 presents the ORR and OER polarization curves of the investigated GDEs at cycle one and one hundred, after continuous potential cycling.  Extensive cycling for one hundred continuous cycles caused significant drop in the ORR and OER electrocatalytic activities of all mixed-oxide catalysts (Figure 3.7). While the addition of perovskites and fluorite-type oxides to MnO2 enhances significantly both the ORR and OER initial stage electrocatalytic activities (as confirmed by Tafel plots in Figure 3.6), the activity degradation during one hundred cycles of accelerated degradation test is more severe than MnO2 alone for all mixed-oxide formulations. MnO2-LaNiO3 revealed the best initial stage ORR/OER electrocatalytic activity whereas showing the worst long-term ORR/OER performance during accelerated degradation testing of one hundred cycles in the flooded test setup (Figure 3.6 and Figure 3.7).  While it can be inferred that the perovskite and fluorite-type oxides as co-catalysts are most likely responsible for the degradation of the ORR/OER electrocatalytic activity of mixed-oxide catalysts during potential cyclying, an insight on the factors involving this complex degradation phenomenon is given in section 6.2.5 following further in-depth structural and electrochemical studies in the next chapters. 97    Figure 3.7 Electrocatalytic durability testing of GDEs with MnO2, MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7 catalysts: A) ORR at cycle one, B) OER at cycle one, C) ORR at cycle one hundred, D) OER at cycle one hundred. IR-corrected ORR/OER polarization curves obtained by potential scanning between 633 to 1633 mV for one hundred cycles (accelerated degradation testing) in O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm. Other conditions idem to Figure 3.6.  3.2.4 Long-term ORR durability of oxide catalysts in a flow-by test setup Figure 3.8 shows the long-term ORR galvanostatic polarization curves of fresh GDEs over the course of 24 hrs in 11.7 M (45 wt%) KOH at 323 K with either air (CO2 removed) or oxygen flowing through the gas chamber of the Gaskatel flow-by cell. The GDE consisting of Pt powder from Alfa Aesar sprayed on Zincnyx’s PTFE treated carbon substrate was used to compare the 98  performances of investigated catalysts to a commercial PGM catalyst norm. A MnOx GDE from Gaskatel was also used to compare the results to a non-PGM commercial electrode currently being used for industrial ORR applications. The first 24 hrs of applying low current densities in each case (i.e. -33 mA cm-2 in air or -67 mA cm-2 in oxygen) were adopted as a break-in protocol for the fresh GDEs (see section 2.3 and Appendix C  ). With air as feed gas at -40 mA cm-2, MnO2-Nd3IrO7 catalyst possessed the lowest average ORR overpotential of about -290 mV with a potential loss of -29 mV during 24 hrs of testing (Figure 3.8-A). MnO2-perovskite mixed oxides showed the next lowest ORR overpotentials, closely tailing the MnO2-Nd3IrO7 sample (Figure 3.8-A). With an average ORR overpotential and potential loss of -370 and -30 mV, respectively, for the in-house MnO2 sample, the synergistic effect of oxide co-catalysts on the ORR performance of MnO2 catalyst layer can be further confirmed at high cathodic current densities in an unflooded test setup (Figure 3.8-A). As mentioned in section 3.2.2, this synergistic effect could be as a result of breaking away from the universal scaling relationship between the HOO(ads) and HO(ads) binding energies with new binding sites available through the mixed-oxide catalysts. In terms of stability, however, it seems that perovskite-type oxides (specially LaNiO3) better enhance the long-term durability of MnO2 catalyst layer compared to Nd3IrO7 (Figure 3.8-A). Comparing the ORR performances of the in-house MnO2 with commercial PGM and non-PGM GDEs, Gaskatel MnOx and Pt samples showed more negative overpotentials (about -10 mV and -40 to -50 mV, repectively) comparing to the in-house MnO2 GDE during the course of 24 hrs (Figure 3.8-A). The overall potential losses of commercial Pt and MnOx GDEs were found to be similar to the in-house MnO2 sample, i.e. about -30 mV (Figure 3.8-A).  In the case of pure oxygen as feed gas, severe potential losses were observed during the one-day-long ORR durability tests except for MnO2-Nd3IrO7 sample (Figure 3.8-B). With an average 99  ORR overpotential of -248 mV and potential loss of -15 mV during 24 hrs of testing, MnO2-Nd3IrO7 is the best performing GDE at -100 mA cm-2 with oxygen feed, about 110 mV more positive than the commercial Pt after 23 hrs of testing (Figure 3.8-B). Contrary to the air feed case, the MnO2-LaNiO3 showed the highest ORR overpotentials as well as most drastic potential loss during 24 hrs of testing (Figure 3.8-B). Looking at the galvanostatic polarization curves in presence of oxygen, a considerable synergistic effect specially after 9 hrs of testing is observed for MnO2-Nd3IrO7 and MnO2-LaCoO3 samples, showing more positive ORR overpotentials compared to the in-house MnO2 GDE (Figure 3.8-B). Moreover, both commercial GDEs were unable to tolerate 24 hrs of durability tests with Gaskatel MnOx failing after 10 min. into the test while Pt sample failed after 23 hrs of testing (Figure 3.8-B). This could be related to the loss of active materials or the blockage in GDE pores, which are necessary for gas supply, during the previous testing conditions, i.e. break-in protocols. Comparing the results in air and oxygen, the trend in ORR electrocatalytic activity of samples is quite the same except for MnO2-LaNiO3 and commercial MnOx which could not tolerate high current density of -100 mA cm-2 (Figure 3.8-A and B). Looking at the stability performances of MnO2-based GDEs in both air and oxygen cases, it can be inferred that the decay in ORR electrocatalytic activity of mixed-oxide catalysts is mainly due to MnO2 degradation since performance losses for mixed oxides happen at about the same time MnO2 GDE starts to degrade, i.e. after 3 and 7 hrs of testing for air and oxygen tests, respectively (Figure 3.8-A and B). This can be backed by the literature studies which found less active Mn species, such as Mn(OH)2 and Mn3O4, toward ORR on cathodically polarized MnO2 catalysts for pro-longed times, as further explained in section 1.2.2.2 [4, 5, 39, 45, 46, 67, 90, 98, 113, 227]. 100  To further study the role of MnO2 in ORR performance loss of mixed-oxide catalysts, XRD characterization has been performed on MnO2 catalyst layer at two stages: 1) Fresh and 2) After 48 hrs of galvanostatic polarization tests (including the break-in protocol) (Figure 3.9). Due to the high contents of PTFE in both catalyst layer and carbon substrate leading to low signal-to-noise ratio, it is extremely hard to detect any small structural changes in the XRD spectrum of the investigated mixed-oxide catalyst during the long-term durability tests. Some even argue that other characterization methods such as EELS are more effective for Mn valence determination when it comes to complex catalyst layers with more than one component [228, 229]. Hence, the in-house made MnO2 GDE was chosen to avoid further complexity by other co-catalysts during the experiments. Section 4.2.1.2 will touch more on the EELS characterization method for the catalyst layer degradation of MnO2-based mixed oxides during ORR/OER accelerated degradation tests. Regarding the fresh MnO2 electrode, the peaks at 2θ = 18˚, 28.5˚, 41˚, 59.5˚, 65˚, 67.5˚ and 72˚ correspond to PTFE, β-MnO2 (110), β-MnO2 (200), β-MnO2 (220), β-MnO2 (002), β-MnO2 (310) and β-MnO2 (301), respectively (Figure 3.9). The broad peak at 25˚ is mainly due to an overlap of peaks corresponding to α-MnO2 (110) and graphite (Vulcan XC-72) (100). Other peaks around 37.5˚, 43˚ and 56.5˚ represent both β-MnO2 (101)/α-MnO2 (201), β-MnO2 (111)/α-MnO2 (211) and β-MnO2 (211)/α-MnO2 (221), respectively. After 48 hrs of galvanostatic durability testing, five new and relatively small peaks appear at 2θ = 17.6˚, 26.4˚, 30.4˚, 36.2˚ and 60.6˚ corresponding to MnOOH (010), MnOOH (111), Mn3O4 (112), Mn(OH)2 (311) and MnOOH (-123)/Mn3O4 (215). Hence, the XRD results further confirm the gradual reduction of MnO2 to Mn3+/Mn2+ species, attributing the ORR performance loss of mixed-oxide catalysts to formation of less active Mn species toward ORR during long-term galvanostatic measurements at cathodic currents. 101  The results here are contrary to the findings from the flooded experiments where the degradation of co-catalyst materials, i.e. perovskite and fluorite-type oxides, was likely the reason behind the performance loss of mixed-oxide GDEs for oxygen electrocatalysis during accelerated degradation tests (section 3.2.3). However, it should be noted that the test protocols and applied currents are totally different in each case, enabling different phenomena which can significantly affect the durability behavior of studied catalysts. In the flooded test setup, the GDEs were polarized between both ORR and OER regions where other phenomena such as MnO2 electro-corrosion, carbon corrosion and even Mn valence evolution at high anodic currents can influence the overall performance of mixed-oxide catalysts [48, 141, 203-205, 230]. In the unflooded experiments (flow-by cell), however, the currents were applied cathodically enabling ORR and MnO2 reduction/dissolution reactions only (eqs. 5 to 19).102    Figure 3.8 Long-term ORR durability testing of fresh GDEs containing Pt, MnO2, MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7 catalysts compared with commercial MnOx GDE from Gaskatel GmbH: A) With air at -40 mA cm-2 and B) With oxygen at -100 mA cm-2. IR-corrected galvanostatic polarization curves obtained in 11.7 M (45 wt%) KOH at 323 K for 24 hrs with either air (CO2 removed) or oxygen flowing through the gas chamber of a flow-by cell from Gaskatel GmbH after 24 hrs of break-in protocol explained in section 2.3 and Appendix C  . The absolute gas pressure and flow rate were fixed at 1 atm and 1.51×10-3 SLPM. The catalyst(s) loadings were 2 mg cm-2 each (except for Pt with 0.5 mg cm-2) with final weight ratio of 1:1:2:0.6:0.6 for MnO2 or Pt:co-catalyst (if present):Vulcan XC-72:Nafion:PTFE in the catalyst layer. The catalyst loading for the commercial MnOx from Gaskatel was 20 mg cm-2. The standard error of the mean calculated based on min. two replicates is indicated for each data point. 103   Figure 3.9 Comparison between the XRD spectrum of the MnO2 catalyst: 1) Fresh electrode (black dotted line) and 2) After 48 hrs of galvanostatic tests (24 hrs at -67 mA cm-2 followed by 24 hrs at -100 mA cm-2) with O2 as feed gas in an unflooded flow-by test setup.  In the next chapters, further discussions are made on the complex degradation phenomena of mixed-oxide catalysts using in-depth structural analysis coupled with electrochemical studies. 3.3 Conclusion The electrocatalytic activities for ORR and OER on individual and mixed-oxide catalysts, i.e.  MnO2, LaCoO3, LaNiO3, Nd3IrO7 and their combinations, were studied. A positive synergistic electrode kinetic effect between the oxide components was found as shown by either a decrease in the apparent Tafel slope or an increase in the apparent exchange current density for the mixed-oxide formulations compared to the respective individual oxide catalysts. The mechanism for the 104  mixed oxides’ synergistic electrocatalytic effect could be rationalized in terms of the linear scaling relationship between HOO(ads) and HO(ads) binding energies. The structurally diverse oxide combinations provide different binding energies for the key intermediates, thus, “breaking” away from the linear scaling relationship. However, further studies, both theoretical and experimental, are required to validate the proposed hypothesis. Furthermore, the bifunctional durability of catalysts was investigated by carrying out one hundred potential cycles between 633 and 1683 mV in a flooded test setup. The degradations in electrocatalytic activity of all mixed-oxide catalysts were found to be higher than MnO2 alone. MnO2-LaNiO3 revealed the best initial stage ORR/OER electrocatalytic activity whereas showing the worst long-term ORR/OER performance during accelerated degradation testing of one hundred cycles in the flooded test setup. The perovskite and fluorite-type oxides as co-catalysts are likely to be one of the reasons behind the degradation in the ORR/OER electrocatalytic activity of mixed-oxide catalysts. However, further in-depth structural and electrochemical studies are needed to investigate this complex phenomenon. The ORR long-term performances of fresh MnO2 and mixed-oxide GDEs in commercial scenarios were tested using twenty-four-hour-long galvanostatic polarizations in 11.7 M KOH at 323 K and Pgas of 1 atm with either air or oxygen flowing through a flow-by test cell. MnO2-Nd3IrO7 revealed the highest ORR activity and good stability followed by MnO2-LaCoO3 as the second best, both outperforming the commercial Pt and MnOx GDEs with up to 100 mV (air) and 150 mV (oxygen) more positive ORR overpotentials during 24 hrs of galvanostatic testing. The structural analysis on the MnO2-based GDEs during ORR galvanostatic polarization tests showed that the gradual transformation of MnO2 to less active forms of manganese species (i.e. 105  Mn3+/Mn2+) during ORR could attribute to the ORR performance degradation of the mixed-oxide catalysts. 106  Chapter 4: Oxide catalyst activation by alkali-metal ion intercalation §§§ 4.1 Introduction The present chapter discusses the effect of alkali-metal ions (Li+, Na+, K+ and Cs+), specially potassium ions, on the electrocatalytic activity and durability of oxide catalysts, i.e. MnO2, LaCoO3, Nd3IrO7, MnO2-LaCoO3 and MnO2-Nd3IrO7, for ORR and OER with wide range of in-depth structural and electrochemical characterizations.  The degradation of bifunctional electrocatalytic activity for the mixed-oxide catalysts, i.e. MnO2 with either perovskites or fluorite-type oxides, during extensive potential cycling is shown to be fully restored by long-term exposure to 6 M KOH at open-circuit. Insertion of potassium ion in the oxide structure either by longer-term exposure to 6 M KOH or by an accelerated potential driven intercalation method, has been found to be effective in enhancing the ORR and OER electrocatalytic activity of the investigated oxide catalysts. In addition, the stability of the potassium ion activated catalysts is reported to be improved. The electrode kinetic results presented here are supported by extensive surface analysis. Lastly, a thorough comparison of the results obtained in the present work with those reported in the literature for a variety of bifunctional catalysts is shown, demonstrating the effectiveness of                                                  §§§ Parts of this chapter have been published and filed as United States and Canadian patent applications: - P. Hosseini-Benhangi, A. Alfantazi, E. Gyenge, “Manganese dioxide-based bifunctional oxygen reduction/evolution electrocatalysts: Effect of perovskite doping and potassium ion insertion”, Journal of Electrochimica Acta, 123 (2014) 42-50. - P. Hosseini-Benhangi, M. A. Garcia-Contreras, A. Alfantazi, E. Gyenge, “Method for enhancing the bifunctional activity and durability of oxygen electrodes with mixed oxide electrocatalysts: Potential driven intercalation of potassium”, Journal of The Electrochemical Society, 162 (2015) F1356-F1366. - E. Gyenge, P. Hosseini-Benhangi, “An oxygen electrode and a method of manufacturing the same”, U.S. (15/251,267) and Canadian (2,940,921) patent applications, filed on August 30th, 2016.  107  potassium activation methods on enhancing the ORR and OER electrocatalytic activity of both individual and mixed-oxide catalysts. 4.2 Results and Discussion 4.2.1 Oxide catalyst layer activation by open-circuit potential (OCP) K+ intercalation 4.2.1.1 Healing effect After the oxide GDEs, i.e. MnO2, MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7, were subjected to one hundred potential cycles in a flooded test setup, the effect of GDE rest-time in 6 M KOH at open-circuit potential on the recovery of the ORR and OER electrocatalytic activities was investigated (Figure 4.1 and Figure 4.2). For all degraded samples, improvements in the bifunctional electrocatalytic activities can be observed already after 6 hrs of exposure to 6 M KOH at open-circuit. Moreover, a rest time of six days produced remarkable enhancements in the electrocatalytic activities of the degraded GDEs for both ORR and OER, exceeding even the initial performance of the fresh samples (compare  Figure 3.7, Figure 4.1 and Figure 4.2). For MnO2 alone, the ORR current density of fresh GDE (at 680 mV) started at -3 mA cm-2, decreasing to -0.9 mA cm-2 after one hundred cycles of accelerated degradation tests followed by a significant enhancement to -6 mA cm-2 after six-day rest in 6 M KOH (Figure 3.7 and Figure 4.1-A). In the OER region, the current density of fresh MnO2 catalyst at 1600 mV dropped from 1.7 to 1 mA cm-2 after extensive potential cycling between ORR and OER regions, reaching 31 mA cm-2 after being rested for six days in 6 M KOH at 293 K (Figure 3.7 and Figure 4.1-B). The degraded MnO2-Nd3IrO7 revealed highest ORR performance after the six-day-long rest in 6 M KOH, i.e. possessing about 10 times the ORR current density (at 680 mV) of the degraded sample prior to the treatment (Figure 3.7 and Figure 4.2-C), with MnO2-LaCoO3 showing the biggest enhancement in its ORR performance after the treatment, i.e. possessing 13 times the ORR current density (at 680 mV) of 108  the degraded sample prior to the treatment (Figure 3.7 and Figure 4.1-C). With regards to OER, the degraded MnO2-LaCoO3 catalyst possessed the highest OER current density of 61 mA cm-2 at 1600 mV after six-day-long rest in 6 M KOH, about 67 times that of the degraded sample right after the accelerated degradation test (Figure 3.7 and Figure 4.1-D).   109    Figure 4.1 The effect of rest-time at open-circuit potential in 6 M KOH at 293 K following accelerated degradation testing in the flooded test setup (i.e. one hundred potential cycles between 633 to 1633 mV in O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm): A) ORR polarization curve for MnO2, B) OER polarization curve for MnO2, C) ORR polarization curve for MnO2-LaCoO3 and D) OER polarization curve for MnO2-LaCoO3. IR-corrected ORR/OER polarization curves obtained by potential scanning between 633 to 1683 mV in O2 saturated 6 M KOH at 400 rpm, 293 K and PO₂ of 1 atm. Other conditions idem to Figure 3.6.  110   Figure 4.2 The effect of rest-time at open-circuit potential in 6 M KOH at 293 K following accelerated degradation testing in the flooded test setup (i.e. one hundred potential cycles between 633 to 1633 mV in O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm): A) ORR polarization curve for MnO2-LaNiO3, B) OER polarization curve for MnO2-LaNiO3, C) ORR polarization curve for MnO2-Nd3IrO7 and D) OER polarization curve for MnO2-Nd3IrO7. IR-corrected ORR/OER polarization curves obtained by potential scanning between 633 to 1683 mV in O2 saturated 6 M KOH at 400 rpm, 293 K and PO₂ of 1 atm. Other conditions idem to Figure 3.6.      111  Due to the high sensitivity of MnO2-LaCoO3 sample to the OCP activation method discussed here, this mixed-oxide catalyst has been chosen for further in-depth studies to unveil the mechanism behind the remarkable bifunctional activity recovery after the exposure to 6 M KOH at open-circuit. A preliminary structural investigation using XRD and XPS analysis is presented aiming at identifying possible structural changes in the catalyst during extensive potential cycling followed by the six-day-long treatment. Figure 4.3 shows the XRD spectra of MnO2-LaCoO3 GDEs in both initial stage and after being cycled for one hundred cycles (accelerated degradation testing) and rested for six days at open-circuit in 6 M KOH, i.e. black dotted and red solid lines, respectively. In case of the fresh GDE, the peaks at 2θ = 18˚, 28.5˚, 31.5˚, 37.5˚, 48.5˚, 56.5˚, 67.5˚, 68.5˚ and 72˚ correspond to the PTFE, β-MnO2 (110), LaCoO3 (110), β-MnO2 (101)/α-MnO2 (201), La(OH)3, β-MnO2 (211)/α-MnO2 (221), β-MnO2 (310), α-MnO2 (601) and β-MnO2 (301)/(112), respectively. The major broad peak at 2θ = 25˚ is mainly due to an overlap of peaks corresponding to LaCoO3 (012), α-MnO2 (110) and graphite (Vulcan XC-72) (100). Moreover, the minor peak at 2θ = 41˚ represents both LaCoO3 (202)/(006) and β-MnO2 (200). With regards to the MnO2-LaCoO3 catalyst layer after being cycled for one hundred cycles and rested for six days at open-circuit in 6 M KOH, six new peaks appear at 2θ = 16.5˚, 24˚, 29.5˚, 33.5˚, 43.5˚and 60.5˚ (labeled as “*” in Figure 4.3) which correspond to MnOOH (010), LaCoO3 (012), Mn3O4 (112), MnOOH (020), MnOOH (410) and both MnOOH (-123)/Mn3O4 (215), respectively. Looking back at section 1.2.2.1, the γ-MnO2 goes through different reduction and oxidation processes in the ORR and OER potential regions, starting with reduction to MnOOH and then Mn(OH)2 as well as hausmannite (Mn3O4) (unlikely to electrochemically oxidize in the potential 112  range studied here) in the ORR region followed by conversion to δ-MnO2 (layered birnessite structure) in the subsequent OER region and final formation of excessive Mn3O4 during the following ORR [40, 41, 45, 92, 93]. Poor ORR and OER electrocatalytic activity of Mn3O4 in alkaline media has been reported in the literature. 40 μg cm-2 of Mn3O4/C provides about -2.25 mA cm-2 at -400 mVMOE in O2 saturated 1 M KOH at 298 K while β-MnO2/C catalyst exhibits a much better ORR electrocatalytic activity, i.e. over -3.5 mA cm−2 under similar conditions, Lima et al. reported [45]. Moreover, Ramírez et al. showed that the amorphous MnOx (with some content of β-MnO2) outperforms Mn3O4 sample for OER in alkaline media, e.g. 2.3 times the OER current density of the Mn3O4 sample (at 1800 mVRHE and 298 K in 1 M KOH) for amorphous manganese oxide [136]. Hence, our XRD results in Figure 4.3 are consistent with the literature confirming MnO2 converts to MnOOH and inevitably hausmannite (Mn3O4) after being severely cycled [40, 41, 93, 228]. Thus, the loss of ORR/OER electrocatalytic activity can be attributed to defective regeneration of active MnO2 (i.e. γ-MnO2) from MnOOH and Mn3O4 during potential cycling. Regarding the possible degradation of LaCoO3 in the mixed-oxide catalyst, XRD analysis revealed the formation of LaCoO3 (012) after accelerated degradation testing, which could have a lower activity compared to the originally present LaCoO3 (110) (Figure 4.3). Further discussions on the role of perovskites in the performance degredation of mixed-oxide catalysts during long-term durability tests will be given in section 6.2.5.   113   Figure 4.3 Comparison between the XRD spectrum of the MnO2-LaCoO3 catalyst: 1) Fresh GDE (black dotted line) and 2) After accelerated degradation testing for one hundred cycles in the potential range of 633 to 1633 mV followed by resting for six days at open-circuit in 6 M KOH at 293 K (red solid line) in a flooded test setup.  In order to investigate the reason(s) behind the considerable activity recovery and enhancement of all cycled and rested electrodes for both ORR and OER, XPS analysis has been employed. Figure 4.4 shows the XPS spectra of the MnO2 GDE at its initial stage and after severe potential cycling and resting (for six days) in 6 M KOH at 293 K. The major peak of Mn overlaps with the one corresponding to F at about 690 eV. Fluorine atoms are present in both the catalyst layer (due to the Nafion ionomer and PTFE presence) and substrate (mainly because of 114  teflonation). Due to the low signal-to-noise ratio of the XPS spectra arising from the high F:Mn ratio, it was not possible to distinguish between the Mn 3s multiplet peaks and noises of the spectra for further Mn valence studies. Comparing the XPS spectra for the fresh MnO2 GDE and after accelerated degradation testing followed by six days of  resting at open-circuit in 6 M KOH, a new peak was observed in the latter sample at 380 eV (compare Figure 4.4-A and B). The peak at 380 eV can be specifically assigned to K 2s. Hence, it is proposed that the uptake and intercalation of K+ ions in the catalyst layer during extended exposure to 6 M KOH at open-circuit potential could induce a promoter effect for ORR and OER electrocatalysis, being responsible for the recovery (or “healing”) of the electrocatalytic activity of degraded samples. The reproducibility of “healing” effect by potassium ion intercalation into the structure of degraded mixed-oxide catalysts has been reported in Appendix D  . Further experiments investigating whether the initial stage activity (i.e. for the fresh GDEs) could be enhanced by exposure to 6 M KOH and insertion of K+ ions at open-circuit potential before use in polarization experiments will be presented in the next section. Moreover, the mechanisms involved with the potassium ion promotion effect on ORR/OER bifunctional electrocatalytic activity of investigated oxide catalysts will be discussed in the next sections.   115   Figure 4.4 XPS spectra of MnO2 GDE: A) Fresh, C) After accelerated degradation testing for one hundred cycles in the potential range of 633 to 1633 mV and rested for six days at open-circuit in 6 M KOH at 293 K. 116  4.2.1.2 Activation of fresh catalysts Here, the concept of oxide activation by K+ is further advanced by considering the following questions: 1) Is the effect specific to K+ or other alkali-metal ions produce similar effects? and 2) Can also the initial stage bifunctional activity be improved by activation and is this effect durable?  Bifunctional activation effect of alkali-metal ions Figure 4.5 shows the bifunctional polarization curves of MnO2-LaCoO3 recorded after six days of exposure to alkali-metal hydroxide solutions (at 313 K and 400 RPM) with concentrations near their respective ionic conductivity maximum [231]. With respect to ORR (Figure 4.5-A), clearly KOH induced the most significant activity improvement. At a potential of 730 mV, an ORR current density of -12.5 mA cm-2 was obtained, whereas in case of exposure to any of the other hydroxides and for the unactivated sample, the current density was at least three times lower, indicating a high level of K+ specificity. These results are corroborated by a different type of investigation, where the ORR on LaMnO3 was comparatively studied in either 0.1 M LiOH, or 0.1 M NaOH or 0.1 M KOH [232]. The ORR current density increased with increasing cation size in the electrolyte. It was proposed that the alkali metal ion may influence the rate determining step by interacting with the O22− species formed on the oxide surface. The smaller the cation size the stronger this interaction, inhibiting, therefore, the ORR rate determining step [232]. However, Figure 4.5 shows that in the present case, the performance with six-day exposure or without exposure to LiOH are virtually the same. Thereby, there is no evidence of ORR inhibition by Li+. Furthermore, exposure to Cs+ produces only a minor ORR improvement compared to K+, suggesting that no simple linear correlation can be established based on cation size.  In the OER section of the polarization curve (Figure 4.5-B), extended exposure to all the alkali-metal hydroxides increased the current density compared to the unactivated case. However, 117  the best results were obtained in the presence of K+ and Cs+ ions. At a potential of 1450 mV, the OER current density on the MnO2-LaCoO3 electrode was about an order of magnitude higher after the electrode was exposed to either KOH or CsOH. While some degree of non-specific contribution in Figure 4.5-B cannot be completely ruled out, where the longer-term exposure to any alkali-metal hydroxide solution could render the electrode more hydrophilic (e.g. partial PTFE wash-out), hence, a higher fraction of the pores are available for electrolyte penetration and oxygen evolution, Figure 4.5-A and Figure 4.5-B together point to a distinct bifunctional promotion effect mainly by K+ and to some extent by Cs+.  118    Figure 4.5 Bifunctional activation effect of long-term (i.e. six days) exposure of MnO2-LaCoO3 to alkali-metal hydroxide solutions: LiOH, NaOH, KOH, CsOH. Initial stage IR-corrected polarization curves obtained by potential scanning between 633 to 1483 mV in O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm. Other conditions idem to Figure 3.6.  119  Accelerated degradation tests on fresh activated mixed-oxide catalysts Next, the electrocatalytic durability of potassium activated (i.e. six days of exposure to 6 M KOH at 313 K and 400 rpm) MnO2-LaCoO3 and MnO2-Nd3IrO7 electrodes have been investigated under severe potential cycling between 633 to 1483 mV in 6 M KOH at 293 K (Figure 4.6). Extensive cycling drastically diminished the ORR electrocatalytic activity of fresh activated MnO2-LaCoO3 catalyst with ORR current density decreasing from about -21 to -7 mA cm-2 (at 680 mV) after one hundred cycles (Figure 4.6). Comparing to the fresh unactivated electrode, the activated MnO2-LaCoO3 catalyst after one hundred cycles possessed 3.5 times the ORR current density (at 680 mV) of the unactivated sample (compare Figure 3.7-A and Figure 4.6-A). With regards to the OER region, the activated MnO2-LaCoO3 showed excellent durability with current density loss of a mere 1 mA cm-2 at 1450 mV (i.e. about one tenth of the original OER current density for fresh activated sample) during one hundred cycles of accelerated degradation testing (Figure 4.6-B). For MnO2-Nd3IrO7, the activated sample revealed relatively high ORR and OER performance losses after one hundred cycles of accelerated degradation tests, i.e. about twice the ORR and OER current densities (at 680 and 1450 mV, respectively) of the activated sample at cycle 100 was observed for the fresh activated sample at cycle 1 (Figure 4.6).  Comparing the two K+ activated GDEs, i.e. MnO2-LaCoO3 and MnO2-Nd3IrO7, looking at the ORR region, the activated MnO2-LaCoO3 performs better in terms of both initial stage activity and durability, e.g. the current density at 680 mV is about -21 mA cm-2 compared to -4 mA cm-2 for the fresh activated MnO2-Nd3IrO7 (Figure 4.6-A). For OER, the same trend is maintained with the exception that the initial stage electrocatalytic activity of activated MnO2-Nd3IrO7 is superior at potentials higher than 1410 mV (Figure 4.6-A). Thus, one can conclude that the K+ activation 120  procedure is more effective when MnO2 is combined with perovskites as opposed to fluorite-type oxides.   121    Figure 4.6 Electrocatalytic durability testing of fresh activated MnO2-LaCoO3 and MnO2-Nd3IrO7 GDEs: A) ORR polarization curves and B) OER polarization curves. IR-corrected ORR/OER polarization curves obtained by potential scanning between 633 to 1483 mV for one hundred cycles (accelerated degradation testing) in O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm. Catalysts were activated by six-day-long exposure to 6 M KOH at 313 K and 400 rpm and open-circuit. Other conditions idem to Figure 3.6. 122  To gain further insights on the K+ promotion effect as revealed by electrode polarization experiments (Figure 4.5 and Figure 4.6), the XPS spectra of both fresh and K+ activated MnO2-LaCoO3 and MnO2-Nd3IrO7 catalysts are presented in Figure 4.7. The Mn and La major peaks overlap with the ones corresponding to F at about 690 and 835 eV, respectively (Figure 4.7-A and B). Fluorine is one of the main constituents of both the carbon cloth substrate (due to teflonation) and catalyst layer (due to Nafion ionomer and PTFE). Furthermore, the Nd major peaks around 980 eV overlap with O (Figure 4.7-C and D).  Comparing the XPS spectrum for fresh and K+ activated catalysts, the latter reveals peaks around 380 eV which correspond to K 2s. Two major spin-orbit splitting peaks appear for K 2p around 290 eV, but these peaks also overlap to a large extent with the ones from C(F) and C 1s due to the carbon material in both substrate and catalyst layer (Figure 4.7). The presence of potassium has been also confirmed by the EDX spectra of the activated catalyst layers (Figure 4.8).123    Figure 4.7 XPS spectra of MnO2-LaCoO3 and MnO2-Nd3IrO7 before and after long-term exposure to 6 M KOH: A) MnO2-LaCoO3 fresh electrode, B) MnO2-LaCoO3 activated electrode, C) MnO2-Nd3IrO7 fresh electrode, D) MnO2-Nd3IrO7 electrode. Other conditions idem to Figure 4.6.124   Figure 4.8 EDX spectra of the activated catalyst layers after six days of exposure to 6 M KOH: A) MnO2-LaCoO3, B) MnO2-Nd3IrO7. Conditions idem to Figure 4.6.  The intercalation of K+ in MnO2 structure could be understood in terms of the cation vacancy model [233]. During electrode potential cycling between the ORR and OER regions, the fraction of Mn4+ and Mn3+ ions is changing, as shown also by electron energy loss spectroscopy (EELS) in Figure 4.9. The first charge-transfer process associated with MnO2 can be represented as [90]:  𝑀𝑛(𝑠)4+ + 2𝑂(𝑠)2− + 𝑒− + 𝐻2𝑂 ⟶ 𝑀𝑛(𝑠)3+ + 𝑂(𝑠)2− + 𝑂𝐻(𝑠)− + 𝑂𝐻(𝑎𝑞)−  (35)  Thus, one electron and one proton is inserted per MnO2 leading to the formation of OH−(s) and Mn3+(s) with lattice expansions [233, 234]. Generally, the composition of partially reduced γ-MnO2 can be described as:  𝑀𝑛1−𝑥−𝑦4+  .𝑀𝑛𝑦3+ . 𝑂2−4𝑥−𝑦2−  . 𝑉𝑥∕∕∕ . 𝑂𝐻4𝑥+𝑦−  (36)  125  where x is fraction of vacancies, y is fraction of Mn3+ ions and V/// represents a cation vacancy in the Schottky notation [90].  In light of the cation vacancy model and eq. 36, it seems plausible that K+ could intercalate into the vacancies, also known as Schottky defects, surrounded by OH− ions. This intercalation may cause lattice distortion since the ionic radius of Mn4+ ions is much smaller than the one for K+, i.e. 53 and 138 pm, respectively [235]. It is then proposed that the lattice distortions induced by K+ affect the binding energies of intermediate species involved in ORR and OER such as HOO(ads) and HO(ads), contributing to the enhanced bifunctional activity by breaking away from the universal scaling relationship between their binding energies.  During potential cycling between the ORR and OER regions, diverse MnOx phases are forming with different activity and stability causing an overall complex behavior influencing the electrode durability  [4, 39, 45, 67, 90, 98, 113]. An effective way to find the Mn valence would be important to unveil the MnOx associated with different stages of ORR and OER. While XPS analysis can help determine the Mn valence using the Mn 3s peak separation method in the presence of pure MnOx [236-238], EELS is more effective for Mn valence determination when it comes to complex systems such as bifunctional catalyst layers with more than one component [228, 229]. EELS was performed on MnO2-LaCoO3 catalyst in three different conditions (Figure 4.9): 1) Fresh catalyst, 2) K+ activated catalyst after being cycled for ten cycles between 633 and 1483 mV in O2 saturated 6 M KOH at 293 K and 400 rpm, and 3) K+ activated catalyst after being cycled for one hundred cycles under the same conditions (accelerated degradation test).  Figure 4.9 shows the EELS spectra for manganese L edges and oxygen K edges as well as the calculated Mn valences for the MnO2-LaCoO3 catalysts during different stages of potential cycling. Figure 4.9-A indicates that not only the shape of Mn(L2,3) peaks changes during the 126  accelerated degradation testing but also its position shifts during potential cycling. The Mn(L3) main peak for MnO2-LaCoO3 catalyst shifted from 645.8 to 639.9 eV after one hundred cycles of durability testing (Figure 4.9-A). The O(K) main peak, however, fluctuates around 532 eV for both the fresh sample and the one subjected to one hundred potential cycling. According to the literature, the shifts in the Mn(L3) and O(K) main peaks could be related to the change of MnOx species from MnO2 to Mn2O3 and Mn3O4 after being cycled for one hundred cycles [239-241].  The shape of each Mn(L2,3) and O(K) edges can also represent the type of manganese oxide present in the catalyst layer. The Mn(L2,3) and O(K) edges of the fresh MnO2-LaCoO3 catalyst in EELS spectrum is similar to the ones shown in the literature for MnO2 [228, 240] (Figure 4.9-A and B). After the K+ activation and ten cycles, the EELS spectrum of the MnO2-LaCoO3 sample is similar to that of Mn2O3 reported in the literature [239-241], especially the Mn(L3) and O(K) peaks at 644.8 and 532.9 eV, respectively (Figure 4.9-A and B). The Mn (L2,3) edges of the activated MnO2-LaCoO3 after being cycled for ten cycles could be due to MnOOH.  Furthermore, the activated MnO2-LaCoO3 catalyst after one hundred cycles shows the typical EELS spectrum of Mn3O4, reported in the literature [239, 241], with two Mn(L3) and one O(K) peaks at 639.9, 641.3 and 531.7 eV, respectively (Figure 4.9-A and B). The Mn3O4 doublet peaks could be further deconvoluted to Mn(L3) edges from MnO to Mn2O3 or MnOOH (Figure 4.9-A) [239, 240, 242-244]. It has been reported that the ratio of Mn(L3) peaks corresponding to Mn3+ and Mn2+, in the present case 1.1 after one hundred cycles, could indicate the presence of vacancies in the tetrahedral sites of the Mn3O4 [245]. These vacancies could also act as sites for K+ intercalation. Figure 4.9-C shows the Mn valence vs. the L3:L2 branching ratio defined as: 127  𝐿3: 𝐿2 𝑏𝑟𝑎𝑛𝑐ℎ𝑖𝑛𝑔 𝑟𝑎𝑡𝑖𝑜 =𝐼(𝐿3)(𝐼(𝐿2) + 𝐼(𝐿3))⁄   (37)  where I(L3) and I(L2) are the intensities of Mn(L3) and Mn(L2) edges from the EELS spectrum of each sample.  In order to compare the calculated valences vs. reference values (Figure 4.9-C), the EELS spectra of MnO, Mn2O3 and MnO2 have been extracted from literature and the corresponding L3:L2 branching ratios have been used as reference points [228, 239]. This graph confirms that the Mn valence for the fresh catalyst is about 4, indicating the unreduced MnO2. The Mn valence decreases to 3.1 after ten cycles (i.e. MnOOH and Mn2O3), whereas extensive cycling up to one hundred cycles, lowers the Mn valence to 2.9 and 2.6 for the first and second Mn(L3) edge, respectively (Figure 4.9-C). The latter values indicate the increased formation of Mn2+ species such as MnO and Mn(OH)2 leading to hausmannite (Mn3O4) as the final composition. It has been reported that hausmannite, which is believed to appear at about 633 mV during MnO2 reduction, shows poor electrocatalytic activity for both ORR and OER compared to MnO2 alone [4, 40, 45, 246]. There is a debate whether hausmannite can be electrochemically oxidized or not. While some studies have claimed that Mn3O4 is electrochemically inactive and cannot be oxidized to Mn4+ oxide species [40, 45, 98, 247], others have identified hausmannite as an ideal candidate for supercapacitor applications and also proposed that it could be transformed to layered birnessite structure (δ-MnO2) after severe cycling [113, 248-251]. The present results in Figure 4.9-C show that MnO2 is not regenerated efficiently during potential cycling between 633 to 1483 mV. However, in spite of profound MnO2 structural changes during potential cycling, the electrodes activated by OCP exposure to K+ exhibited superior cycling durability (Figure 4.6). Furthermore, 128  the activity of degraded electrodes can be regenerated by OCP K+ treatment, called the “healing effect” (section 4.2.1.1 and Appendix D  ).    Figure 4.9 EELS analysis of MnO2-LaCoO3: A) EELS spectrum showing Mn (L2,3) edges, B) EELS spectrum showing O (K) edges, C) L3:L2 Branching ratio versus valence state for Mn oxides. The sample was tested under three different conditions: 1) Initial stage, 2) After being activated in KOH for six days (idem to Figure 4.6) and cycled between the ORR and OER regions for ten cycles, 3) After being activated in KOH for six days and cycled between the ORR and OER regions for one hundred cycles (accelerated degradation test idem to Figure 4.6). The (■) and (▲) symbols represent reference data points obtained from literature [228] and the ones calculated directly from EELS spectrum for each sample, respectively. The standard error of the mean associated with Mn valence state, L3:L2 branching ratio and energy loss are ± 0.2, ± 0.001 and ± 0.1 eV, respectively.  129  4.2.2 Oxide catalyst layer activation by potential driven intercalation (PDI) of K+ 4.2.2.1 Initial stage electrocatalytic activity of PDI activated mixed-oxide catalysts In order to accelerate the insertion of K+ into the fresh oxide catalyst layer, an electrophoretic method referred to as potential driven intercalation (PDI), has been developed and investigated. The PDI method includes a constant cathodic current density of -5.4 mA cm-2 applied for 30 min. (each round up to seven rounds) to the fresh electrodes in a 0.036 M K2SO4 solution at 343 K and 400 rpm (check section 2.1.5 for more details). Figure 4.10 shows two representative cell potential profiles while performing PDI method on the investigated mixed-oxide catalysts, i.e. MnO2-LaCoO3 and MnO2-Nd3IrO7. The presence of potassium in the PDI activated samples was confirmed by XPS analysis (Figure 4.11). The XPS spectra of the electrodes activated by the PDI and OCP methods are very similar (compare Figure 4.7 and Figure 4.11). One of the possible differences between the two activation methods could be the 3D distribution of potassium in the catalyst layers, which needs further experimentation in future works.   130   Figure 4.10 Cell potential profile during potential driven K+ intercalation (PDI) on A) MnO2-LaCoO3 and B) MnO2-Nd3IrO7 as catalyst layers. 5.4 mA cm-2 was cathodically applied to each GDE in the RDE setup in a 0.036 M K2SO4 solution for 30 min up to seven times. The rotation speed and temperature were 400 rpm and 343 K, respectively. A platinum mesh was employed as both reference and counter electrode. The figure indicates the cell potential for the third round of PDI in each case. 131   Figure 4.11 XPS spectra of the potential driven K+ intercalation (PDI) activated catalyst layers: A) MnO2-LaCoO3 electrode after seven rounds of PDI activation, B) MnO2-Nd3IrO7 electrode after six rounds of PDI activation. 132  Figure 4.12 shows the electrocatalytic activity of MnO2-LaCoO3 and MnO2-Nd3IrO7 catalysts after being activated using the PDI method for up to seven rounds, each activation round lasting 30 min. First it is noted that in the ORR region, for both catalyst layers, a peak current density is reached which is controlled by dissolved O2 mass transfer from the bulk solution to the reaction layer (Figure 4.12-A and C). Next, in case of MnO2-LaCoO3 (Figure 4.12-A), repeating the PDI activation procedure shifted positively the ORR peak potential from 465 mV (1st round) to 635 mV (6th round). Further repetition of the PDI activation beyond six rounds did not produce any additional benefits for ORR catalysis. Furthermore, the PDI method (6th round) increased about ten times the ORR current density at 730 mV compared to the unactivated case (Figure 4.5-A and Figure 4.12-A).  In the case of MnO2-Nd3IrO7 (Figure 4.12-C), the shift of the ORR peak potential was more limited, i.e. from 430 mV (1st round) to 500 mV (2nd round and beyond). The ORR current density at 730 mV increased for about two times after the PDI treatment (3rd round) compared to the unactivated case. Regarding the OER section of the MnO2-LaCoO3 GDE polarization curve (Figure 4.12-B) and considering 1450 mV as an arbitrary reference potential, PDI activation (after 6th rounds) generated a current density of 14 mA cm-2 while open-circuit K+ activation produced 9.5 mA cm-2 (Figure 4.5-B), whereas without any type of K+ activation the current density was only about 0.2 mA cm-2 (Figure 4.5-B). Similar improvements were observed in the case of PDI activated MnO2-Nd3IrO7 as well.  133   Figure 4.12 The effect of potential driven K+ intercalation on the initial stage bifunctional polarization of mixed-oxide catalysts: A) ORR on MnO2-LaCoO3, B) OER on MnO2-LaCoO3, C) ORR on MnO2-Nd3IrO7 and D) OER on MnO2-Nd3IrO7. Other conditions idem to Figure 3.6.   4.2.2.2 Initial stage electrocatalytic activity of PDI activated individual oxides To better understand the role of potassium intercalation on the bifunctional performance of the mixed-oxide catalysts, five rounds of PDI activation was also applied to individual oxide catalysts, i.e. MnO2, LaCoO3 and Nd3IrO7 (Figure 4.13). While K+ activation enhances the ORR electrocatalytic activity of all individual oxides (compare Figure 3.6-A and Figure 4.13-A), the method is most effective for MnO2. The OER performance of the individual oxides are also 134  improved by the PDI activation with MnO2 surpassing both LaCoO3 and Nd3IrO7 catalysts (Figure 4.13-B). XPS analysis confirmed the presence of potassium (K 2s peak at 380 eV) in all three oxides after PDI activation (Figure 4.14). The individual oxide polarization results presented by Figure 4.13 substantiate the hypothesis that the main mechanism for bifunctional activity enhancement is related to K+ intercalation into the vacancies or Schottky defects of MnO2 surrounded by OH− ions. This promotion effect of intercalated potassium ions on various oxide surfaces can be further explained considering the lattice oxygen vacancy. Simultaneous reduction of molecular O2 and its incorporation as solid-state oxygen ions are believed to occur during ORR on oxide surfaces [122, 252-254]. This oxygen incorporation reaction into the lattice vacancies in a solid, coupled with electron transfer, can be described using Kröger-Vink notation as follows [254, 255]:  𝑂2 + 4𝑒− + 2𝑉𝑂.. ↔ 2𝑂𝑂𝑥 (38)  where VO..  denotes a lattice oxygen vacancies with double positive charge and 𝑂𝑂𝑥 represents an oxygen ion sitting on an oxygen lattice site with neutral charge. These oxygen vacancies can behave as both acceptor and donors, facilitating the charge transfer between adsorbent and adsorbate [122, 253, 254, 256]. Hence, the continuous formation and annihilation of these lattice oxygen vacancies in oxygen-deficient oxides can significantly affect their ORR/OER electrocatalytic activity [122, 253, 254, 256]. Higher concentration of oxygen vacancies is believed to enhance the lattice oxygen mobility, facilitating the adsorption of reactants (such as OH−) as  well as enhancing the charge transfer, therefore, improving the electrocatalytic activity of deficient oxides for both ORR and OER [122, 256-258].  135  In consideration of lattice oxygen vacancies, it seems plausible that K+ could intercalate into these vacancies, further facilitating adsorption of OH− as well as inducing lattice distortions (compare atomic radius of 73 pm for O and ionic radius of 138 pm for K+) which could provide new binding sites for HOO(ads) and HO(ads) intermediates, help break away from the universal linear scaling relationship and  hence, enhance the ORR/OER bifunctional activity of oxides, specially for perovskite and fluorite-type oxides.  Comparing the mixed and individual oxides activated by PDI (Figure 4.12 and Figure 4.13), it is clear that due to the synergy between either MnO2 and LaCoO3 or MnO2 and Nd3IrO7, the mixed-oxide catalysts possess superior bifunctional electrocatalytic activity than any of the oxides individually. This could be due to the two beneficial phenomena occurring at the same time, first, extra binding sites such as additional lattice oxygen vacancies with different HOO(ads)/HO(ads) binding energies, obtained by mixing the individual oxides, and second, the extra induced potassium ions in the structure of activated oxides affecting the intermediates’ binding energies, both help breaking away from the universal linear scaling relationship and hence, improving the bifunctional activity of the catalysts.  136    Figure 4.13 The effect of potential driven K+ intercalation on the initial stage bifunctional polarization of individual oxides: MnO2, LaCoO3 and Nd3IrO7. A) ORR, B) OER. Other conditions idem to Figure 3.6.137    Figure 4.14 XPS spectra of the PDI activated catalyst layers: A) MnO2 after five rounds of K+ PDI activation, B) LaCoO3 after five rounds of K+ PDI activation and C) Nd3IrO7 after five rounds of K+ PDI activation.138  4.2.2.3 Galvanostatic long-term durability testing of unactivated and PDI activated mixed-oxide catalysts In addition to potential cycling experiments, galvanostatic polarization (i.e. chronopotentiometry) was also performed in order to assess the effect of PDI activation on electrocatalytic activity and stability. The oxide loading was the same as in all other experiments, namely, 0.5 mg cm-2 for each of the oxides. For OER, a constant current density (per geometric area) of 5 mA cm-2 (or 5 A g-1 per total catalyst mass) was applied for 2 hrs, whereas for ORR, -2 mA cm-2 (or 2 A g-1 per total catalyst mass) was applied for 30 min (Figure 4.15). The flooded electrode half-cell arrangement used in the present study imposes some limitations with respect to the current densities that can be applied during galvanostatic longer-term experiments. These conditions are different compared to the cell design that would be used in practice, for instance in a rechargeable zinc-air battery. Therefore, the experiments presented by Figure 4.15 provide only a preliminary insight into durability and further studies are required under conditions more relevant to the industrial practice.  The ORR current density was chosen to be sustainable by the availability of dissolved O2 in the O2 saturated 6 M KOH electrolyte for a more extended period of time (e.g. 30 min). In practice, a gas diffusion oxygen electrode would be used either air breathing or exposed to a convective air (or oxygen) flow (similar to section 3.2.4).  The OER current density of 5 mA cm-2 for 2 hrs, was selected to provide relevant longer-term electrocatalytic stability information, while avoiding the heavy O2 gas evolution expected at high current densities that could shield and/or damage the electrode surface in the present configuration. Two hours galvanostatic polarization was also proposed as an OER benchmarking criteria by McCrory et al. [131], albeit at a current density of 139  10 mA cm-2 but for an unspecified catalyst loading. Hence, it is difficult to employ identical conditions to the latter study.  Comparing first the unactivated fresh catalysts, the OER behavior of MnO2-LaCoO3 was superior over the 2 hr testing period compared to MnO2-Nd3IrO7 (Figure 4.15-A and B). For the latter catalyst (Figure 4.15-B), the potential increased from 1480 mV (at t = 1 min) to 1621 mV (at t = 2h), whereas in case of fresh MnO2-LaCoO3 the electrode potential was much more stable, i.e., 1549 mV (at t = 1 min) and 1568 mV (at t = 2h). PDI activation had a positive influence on both catalysts by reducing the O2 evolution potential by about 110 mV in case of MnO2-LaCoO3 (Figure 4.15-A) and up to 152 mV (at t = 2 h) on MnO2-Nd3IrO7 (Figure 4.15-B). Furthermore, the stability of the OER activity for the PDI activated MnO2-Nd3IrO7 catalyst is markedly superior compared to the unactivated case. Thus, for MnO2-Nd3IrO7 the rate of OER potential increase is lowered from 70.5 mV h-1 (fresh unactivated catalyst) to 10 mV h-1 (PDI activated).  Regarding the galvanostatic ORR response (at -2 mA cm-2), the electrode potential on fresh MnO2-Nd3IrO7 was about 43 mV (at t = 30 min) higher than on MnO2-LaCoO3. PDI activation increased the ORR electrode potential of the latter catalyst by about 75 mV (at t = 30 min) (Figure 4.15-A), whereas it had a lesser influence on MnO2-Nd3IrO7 (15 mV higher potential). These findings corroborate the cycling polarization experiments on the two catalyst formulations presented by Figure 4.12. The rate of ORR potential degradation was also dropped from -30 mV h-1 to -24 mV h-1 for MnO2-LaCoO3 and from -38 mV h-1 to -14 mV h-1 for MnO2-Nd3IrO7 after PDI activation.  140   Figure 4.15 Galvanostatic polarization of mixed-oxide catalysts without and with potential driven K+ intercalation activation: A) MnO2-LaCoO3 and B) MnO2-Nd3IrO7. Tests started with 5 mA cm-2 anodically applied to each GDE for 2 hrs followed by -2 mA cm-2 applied cathodically for 30 min in O2 saturated 6 M KOH using a flooded test setup. The rotation speed and temperature were 400 rpm and 293 K, respectively. PO₂ was 1 atm. The oxide loadings were 0.5 mg cm-2 each. OER ORR OER ORR 141  4.2.3 Comparison of Bifunctional ORR/OER Activities: present work vs. literature It is inherently difficult to compare catalysts from various literature sources because the apparent performance is dependent not only on the intrinsic electrocatalytic activity but also on other interacting factors such as the catalyst loading and dispersion, catalyst layer structure and composition (e.g. presence or absence of support and/or ionomer and/or PTFE) and electrode manufacturing conditions. In spite of the above-mentioned shortcomings, the author believes a comparison with literature results is warranted to place in a broader context the results obtained here with respect to representative precious and non-precious metal catalysts reported in the literature. Figure 4.16 presents a comprehensive comparison of the ORR and OER overpotentials (at -2 and 2 mA cm2, respectively) for the individual and mixed-oxide catalysts investigated here, and relevant catalyst examples from literature. The overpotentials at 2 and -2 mA cm-2 were chosen for comparison because of the available literature data in the latter current density range for diverse catalysts and catalyst loadings. Catalysts with the best and worst bifunctional activity are in the lower bottom-left and top-right corner of Figure 4.16, respectively.  The K+ activated oxide catalysts (i.e. MnO2, MnO2-LaCoO3 and MnO2-Nd3IrO7, with indices between 21 and 25, Figure 4.16) are all situated in the lower half of the diagram due to their low OER overpotential at 2 mA cm-2. The latter is between 100 to 150 mV lower than the reported OER overpotentials for catalysts such as: 20 wt% Ir/C (#13) [69], 20 wt% Ru/C (#12) [67, 69],  Pt/IrO2 (#6) [259], Pt/Ir-IrO2 (#7) [259] and Pt/Ir3(IrO2)7 (#8) [260]. Compared to the unactived MnO2, MnO2-LaCoO3 and MnO2-Nd3IrO7 catalysts (indices 17, 19 and 20, respectively), K+ activation lowered the OER overpotentials by up to 175 mV.    142  With respect to ORR, catalysts such as nano sized Ag (#16) [75], 20 wt% Pt/C (#11) [67], Pt/Ir-IrO2 (#7) [259] and Pt/Ir3(IrO2)7 (#8) [260] generated more positive overpotentials than those reported in the present work. However, other non-precious metal catalysts such as nanostructured Mn oxide thin film (#5) [69] and Core-Corona Structured Bifunctional Catalyst (CCBC) (#10) [76] showed lower ORR electrocatalytic activity with far more negative overpotentials compared to the oxide catalysts investigated here (Figure 4.16).  143   Figure 4.16 Comparison between the ORR and OER overpotential values of the catalyst materials investigated here (shown as (▲)) with those reported in the literature for other bifunctional electrodes (shown as (♦)) [4, 64, 67, 69, 71, 72, 74, 76, 121, 259, 260]. For the catalyst investigated here: a) Fresh catalyst without activation. b) Activation by K+ insertion at open-circuit potential (OCP), c) Activation by K+ insertion using five rounds of potential driven intercalation (PDI), d) Activation by K+ insertion using six rounds of potential driven intercalation (PDI), e) Activation by K+ insertion using three rounds of PDI. The max error associated with overpotential values is ±5 mV. 144  4.3 Conclusion The effect of alkali-metal ions insertion on the ORR/OER electrocatalytic activity and durability of fresh and degraded oxide catalysts was investigated using in-depth structural and electrochemical analysis. It has been shown that the degradation in the ORR/OER electrocatalytic activity of fresh oxide catalysts, i.e. MnO2, MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3irO7, after accelerated degradation tests can be fully restored by resting the electrodes in 6 M KOH at open-circuit potential, called the “healing” effect. It was observed that the performance recovery of the degraded electrodes improved with exposure to 6 M KOH from six hours to six days, with six-day-long activated electrodes exceeding the bifunctional activity of the fresh samples. In addition, the role of K+ insertion in the catalyst structure of fresh oxides was investigated by two methods: A) Longer-term exposure of the catalysts to 6 M KOH and B) Potential driven (electrophoretic) intercalation. Both methods are effective for enhancing the bifunctional activity and durability of the mixed-oxide catalysts. At a constant current density of 5 mA cm-2 (or 5 A gcatalyst-1) applied for 2 hrs, the OER overpotential is lowered by 110 mV and 152 mV due to potential driven potassium ion intercalation on MnO2-LaCoO3 and MnO2-Nd3IrO7, respectively. Furthermore, the rate of OER potential increase, a measure of electrocatalytic activity degradation, is diminished by the application of the potential driven potassium intercalation from 70.5 mV h-1 (for fresh unactivated MnO2-Nd3IrO7) to 10 mV h-1.   In the case of ORR, the potential driven intercalation of potassium was also effective lowering the electrode potential on MnO2-LaCoO3 by 75 mV at a constant current density of -2 mA cm-2 (or -2 A gcatalyst-1). The rate of ORR potential degradation was also improved from -30 mV h-1 to -24 mV h-1 for MnO2-LaCoO3 and from -38 mV h-1 to -14 mV h-1 for MnO2-Nd3IrO7 after K+ potential driven intercalation activation.  145  It is proposed that the reason for the enhanced ORR/OER performances of the activated catalysts is the uptake of K+ into the catalyst layer (mostly in the vacancies and defects of the oxide crystal structures such as Schottky and lattice oxygen vacancies), acting as a promoter for both ORR and OER. The K+ uptake was demonstrated by both XPS and EDX analysis. It is proposed that the insertion of potassium ions in the aforementioned vacancies provides new binding sites with distinct binding energies for HOO(ads) and HO(ads) intermediates, help breaking away from the universal scaling relationship and thus, enhancing the ORR/OER bifunctional activity of the investigated catalysts.    146  Chapter 5: Surfactant-assisted electrodeposition of Mn oxides as promising ORR/OER bifunctional non-PGM electrocatalysts: Factorial design study of the electrodeposition factors **** 5.1 Introduction A systematic study has been performed in this chapter to find an active nanostructured manganese oxide for both oxygen reduction and evolution reactions via surfactant-assisted anodic electrodeposition method. The main and interaction effects of key electrodeposition factors that significantly influence the electrosynthesis of manganese oxides, i.e. Mn2+ concentration (C), applied anodic potential (E), temperature (T), surfactant concentration (S), on the bifunctional activity of MnOx has been studied using a two-level half-fraction factorial design. Sodium dodecyl sulfate (SDS) as anionic, hexadecyl-trimethyl-ammonium bromide (CTAB) as cationic and Triton X-100 as non-ionic surfactants were used in this study to electro-synthesize the nanostructured MnOx. Triton X-100 samples provide the best performing nano-sized structures with promising ORR and OER performances comparing to both noble and non-precious metals and their oxides, i.e. between 50 to 150 mV more positive ORR overpotential (at -2 mA cm-2) comparing to CoMn2O4 and Core-Corona Structured Bifunctional Catalyst (CCBC) and min. 100 mV lower OER overpotential (at 2 mA cm-2) comparing to Ir, Ru and IrO2. Galvanostatic polarizations of Triton-run no. 9 sample at 5 mA cm-2 showed low OER potentials of 1446 mV (at t=2 hrs), about                                                  **** A version of this chapter is in preparation for publication. Parts of this chapter were filed as United States and Canadian patent applications: - P. Hosseini-Benhangi, A. Alfantazi, E. Gyenge, “Surfactant-assisted electrodeposition of Mn oxides as promising ORR/OER bifunctional non-PGM electrocatalysts: Factorial design study of the electrodeposition factors”, to be submitted. - E. Gyenge, P. Hosseini-Benhangi, “An oxygen electrode and a method of manufacturing the same”, U.S. (15/251,267) and Canadian (2,940,921) patent applications, filed on August 30th, 2016. 147  40 mV lower than the commercial MnOx, and degradation rate of 43 mV h-1, about 10 mV h-1 lower than its commercial counterpart. The surface modifications of MnOx via surfactant-assisted electrodeposition can help destabilizing the HOO(ads) and HO(ads) intermediates, breaking away from the linear scaling relationship between their binding energies as a major contributor to the ORR and OER overpotentials, enhancing the ORR and OER electrocatalytic activity of electrodeposited manganese oxides. The formation of hydrogen-bonded complexes, i.e. HO(ads)…H-OH, with specially configured water molecules called “activated water”, can further explain the enhancments in the ORR electrocatalytic activity of the catalysts. 5.2 Results and discussion 5.2.1 Anodic electrodeposition behavior of Mn oxides with and without surfactants  Linear sweep voltammetry (LSV) tests have been used to investigate the anodic electrodeposition behavior of manganese oxides on carbon cloth in presence of diverse types and different concentrations of surfactants, seeking to identify suitable potential range for MnOx electrodeposition while avoiding OER (Figure 5.1). As shown in Figure 5.1, the anodic peak, corresponding to the electrodeposition of MnOx on the pre-treated carbon cloth with no surfactant available, starts at around 450 mVMOE, reaches its max. current density of 1.2 mA cm-2 at 1365 mVMOE and is then joined by the OER at around 2000 mVMOE. The decrease in the anodic current density of the Mn2+ oxidation peak has been attributed to the formation of insulating MnOOH layer (eq. 33) (Figure 5.1). The addition of surfactants, i.e. Triton X-100, SDS and CTAB, with different concentrations, i.e. 5 and 10 vol%, to the 0.2 M Mn(CH3COO)2 and 0.1 M Na2SO4 solution seems to shift the anodic peak of MnOx deposition along with the OER onset potential to more negative potentials (Figure 5.1). While the Mn2+ oxidation peak stays between 1000 to 1200 mVMOE for all types of surfactants at the concentration levels studied here (Figure 5.1), the 148  OER onset potential, determined at the first increase in the current density after the Mn2+ oxidation peak, reaches its min. at around 1350 mVMOE for the solution with 5 vol% of Triton X-100 (Figure 5.1-A). This decrease in the OER onset potentials with introduction of surfactant has been reported in the literature for water electrolysis on carbon electrodes in acidic media and in the presence of a cationic surfactant [261]. In order to avoid the OER, the potential range was fixed to 800 to 1200 mVMOE with a center-point at 1000 mVMOE for all conditions.   149    Figure 5.1 IR-corrected linear sweep voltammograms of nitric acid pre-treated 40 wt% PTFE treated carbon cloth starting from 0 to 2500 mVMOE in presence of: A) 5 vol% and B) 10 vol% of Triton X-100, SDS and CTAB. The solution was made of 0.2 M Mn(CH3COO)2 and 0.1 M Na2SO4 at 293 K. The scan rate and rotation speed were 5 mV s-1 and 400 rpm, respectively.   150  5.2.2 Characterization of the electrodeposited samples XPS spectra were used to identify the Mn valence of electrodeposited manganese oxides. Three representative XPS spectra for T1, T9 and T10 samples are shown in Figure 5.2. The electrodeposition factors for each sample are as follows: T1 (C: 0.3 M, T: 295 K, S: Triton, 10 vol%, E: 800 mVMOE), T9 (C: 0.1 M, T: 343 K, S: Triton, 10 vol%, E: 800 mVMOE), T10 (C: 0.3 M, T: 295 K, S: Triton, 0 vol%, E: 1600 mVMOE). Table 5.1 summarizes the deconvoluted data for Mn 2p, O 1s and Mn 3s regions of these samples. The determination of Mn valence based on the location of Mn 2p peaks is usually associated with high uncertainties mainly due to the differential charging imposed by ejection of photoelectrons from inadequate conductivity of the material’s surface, leading to broadening or shifting of the peaks [237, 262]. However, a combined analysis of Mn 3s doublet peak splitting and O 1s constituents can provide a meaningful understanding of Mn valence in the manganese oxides [236-238, 262]. The Mn 3s peak separation is caused by the electron exchange interaction in the 3s-3d level of Mn upon photoelectron ejection [237]. Several Mn 3s doublet peak separation values have been reported in the literature including 4.5, 5.2, 5.4 and 5.8 for MnO2 (Mn4+), Mn2O3 (Mn3+), Mn3O4 (Mn2+,3+) and MnO (Mn2+), respectively [237, 262, 263]. As demonstrated in Figure 5.2 and Table 5.1, the Mn 3s doublet peak separation values are consistent with the literature showing Mn valence of 2, 4 and mixture of 3 and 4 for T1, T9 and T10 samples, respectively. Further, O 1s can be deconvoluted to three oxygen containing chemical bonds including Mn-O-Mn (oxide), Mn-O-H (hydroxide) and H-O-H (water molecule) [237, 262]. Relatively high content of hydroxide oxygen was detected on the surface of T1 and T9 samples, suggesting the co-existence of Mn3+/Mn2+ species for T1 and Mn4+/Mn3+ species for T9 (Figure 5.2 and Table 5.1).   151   Figure 5.2 XPS spectra of three representative electrodeposited MnOx samples at Mn 2p, Mn 3s and O 1s regions. The electrodeposition factors for each sample are as follows: T1 (C: 0.3 M, T: 295 K, S: Triton, 10 vol%, E: 800 mVMOE), T9 (C: 0.1 M, T: 343 K, S: Triton, 10 vol%, E: 800 mVMOE), T10 (C: 0.3 M, T: 295 K, S: Triton, 0 vol%, E: 1600 mVMOE). 152  Table 5.1 XPS peak analysis of electrodeposited MnOx samples (T1, T9 and T10). The deconvoluted data for Mn 2p, Mn 3s and O 1s is presented. The error associated with binding energy of peak position is ±0.1 eV. Other conditions idem to Figure 5.2.   Mn 2p Mn 3s O 1s  Sample name 3/2 (eV) 1/2 (eV) Eb1 (eV) Eb2 (eV) Eb2-Eb1 (eV) Mn-O-Mn (eV, area %, FWHM) Mn-O-H (eV, area %, FWHM) H-O-H (eV, area %, FWHM) Mn valence T1 642.6 654.4 84.6 90.4 5.8 529.6, 2.2,  1.4 532.1, 65.2, 2.6 533.3, 32.6, 2.5 2, 3 T9 642.3 654.1 84.5 89.1 4.6 529.7, 11.3, 1.4 532.1, 76.8, 2.9 533.4, 11.9, 2.9 3, 4 T10 642.9 654.3 84.5 89.5 5 529.9, 38.7, 1.9 531.8, 35.4, 2.8 533.4, 25.9, 2.7 3, 4 153  Several methods have been reported in the literature for surfactant removal from the electrodeposited samples including heat treatment, UV/ozone treatment and acetone/IPA washing methods [147, 264, 265]. IPA washing for 15 min at 343 K and 400 rpm rotation was chosen as a fast effective method without losing active material and damaging the crystal structure of electrodeposited MnOx. FT-IR analysis was utilized to examine the effectiveness of the IPA washing method for surfactant removal from the manganese oxides deposited in solutions with highest surfactant concentration, i.e. 10 vol%. Figure 5.3 shows FTIR spectra for the electrodeposited MnOx samples under various electrodeposition conditions. The black dotted line (graph I), which represents FTIR spectrum of 40 wt% PTFE treated carbon cloth after nitric acid pre-treatment, shows two characteristic peaks between 1100-1200 cm-1 that disappear after the completion of electrodeposition process (Figure 5.3). This could be an indication for deposition of materials on the substrate in all cases. In the case of electrodeposited MnOx in SDS containing solution, two major peaks for SDS at 1200 and 1460 cm-1 overlaps with peaks associated with MnO2 stretching and O-H bending vibrations of water, respectively [150, 266]. Similar interferences happen for both CTAB and Triton samples where same peaks overlap with major peaks for CTAB at 1486 cm-1 and Triton X-100 at 1113 and 1512 cm-1 [150, 266]. This makes it impossible to employ those peaks for evaluating the effectiveness of surfactant removal procedure. Since all of the surfactants used in this study contain a hydrocarbon chain composed of C and H, e.g. SDS (NaC12H25SO4), CTAB (C19H42BrN) and Triton X-100 (C14H22O(C2H4O)n(n=9-10)), traces of each surfactant on the electrodeposited MnOx can be detected using C-H stretching and C-H deformation vibrations of these hydrocarbon chains between 2700 to 3100 cm-1 and approximately 1490 cm-1, respectively [147]. As shown in Figure 5.3, the absence of such major peaks between 154  2700 and 3100 cm-1 for all of the samples indicates the efficiency of IPA washing method as a fast effective surfactant removal technique for the cases studied here.       Figure 5.3 FT-IR spectra of MnOx samples (after IPA washing) electrodeposited on the pre-treated carbon cloth as substrate in presence of Triton X-100, SDS and CTAB. The electrodeposition factors for each sample are as follows: Carbon cloth (no electrodeposited material), Triton (C: 0.1 M, T: 343 K, S: 10 vol%, E: 800 mVMOE), SDS (C: 0.3 M, T: 295 K, S: 10 vol%, E: 800 mVMOE), CTAB (C: 0.3 M, T: 343 K, S: 10 vol%, E: 1600 mVMOE).  5.2.3 Factorial design experiments A 2n half-fraction factorial design of four factors (Table 2.1) with three center-points (24-1+3) was constructed for each surfactant type, i.e. anionic, cationic and non-ionic, using JMP 11 155  statistical software compiling to a total of 33 random runs for the entire screening design experiments. With the defining relation of I=CTSE in 24-1 design, no main effect is aliased with any other main effect or any two-factor interaction. However, two-factor interactions are aliased with each other. The four main factors plus the three two-factor interaction alias pairs account for the seven degrees of freedom for the design. Pareto plots with estimates of factors and aliases were used to find the most important two-factor interaction in each alias pair based on the Ockham’s razor principle (check Appendix E  ). A single replicate factorial design is used to minimize the number of experimental runs. The single replicate strategy is a very common approach in variable screening experiments due to large number of factors under consideration [267]. In order to remedy for the random error, a wide range for factor levels as well as three center-points were introduced for each set of the experiments. Three different responses were introduced to better assess the ORR/OER bifunctional performances of the surfactant-assisted electrodeposited catalysts: 1) ORR mass activity at 656 mV, 2) OER mass activity at 1556 mV and 3) ORR/OER potential window (the absolute difference between the ORR onset potential at -2 mA cm-2 and OER onset potential at 2 mA cm-2). The catalyst loading was calculated using Fraday’s law via integration of the chronoamperometry graphs assuming the current efficiency for the main MnO2 anodic electrodeposition reaction is 100% (eq. 28). Since the catalyst loadings were different for each run due to various factors involved (Table 2.1), the mass activity values, defined as the ORR or OER current densities at 656 and 1556 mV, respectively, divided by calculated loadings, were employed as responses in the factorial design. The standard error for the mean value of each response, calculated based on the three center-point tests, has been assumed to be the max. error involved in all of the measurements for each surfactant type.   156  5.2.3.1 Triton X-100 surfactant-assisted electrodeposition Table 5.2 shows the design matrix and each response value for factorial runs in presence of Triton X-100 in a random order. Highest ORR and OER mass activities of -1359±202 and 20076±2098 mA g-1, respectively, are obtained at high Mn concentration, low temperature, high surfactant concentration and low applied potential, i.e. run no. 1 (Table 2.1). However, the lowest ORR/OER potential window of 600±20 mV is achieved under opposite conditions of low Mn concentration and high temperature for run no. 9 (Table 2.1). It is worth mentioning that the MnOx electrodeposited at run no. 9 possesses the second best ORR and OER mass activity of -334±50 and 8417±879 mA g-1, respectively.    157  Table 5.2 Design matrix (in random order), Mn valence, calculated loadings and responses for the factorial design experiments in presence of Triton X-100. Other conditions idem to Table 2.1.   Factor levels   Responses No. C T S E Mn valencea Calculated loadings (mg cm-2)b ORR mass activity (mA g-1)c OER mass activity (mA g-1)d ORR/OER potential window (mV)e 1 + - + - 2,3 3.5 -1359 20076 678 2 - - - - 3, 4 1.1 -197 143 1450 3 + + - - 3, 4 3.9 -152 350 1424 4 0 0 0 0 3, 4 19.3 -372 5547 585 5 - - + + 3, 4 7.2 -430 3088 744 6 0 0 0 0 3, 4 20.8 -226 3839 658 7 - + - + 3, 4 10.5 -104 730 914 8 + + + + 3, 4 27.7 -195 1807 664 9 - + + - 3, 4 17.5 -334 8417 600 10 + - - + 3, 4 0.6 -88 55 1450 11 0 0 0 0 3, 4 21.6 -367 4839 631 a The Mn valence calculated based on the XPS results. The bold value corresponds to higher content of that specific Mn valence in the electrodeposited manganese oxide. * Average factorial design response with standard error of the mean value calculated based on the center-point (0000) tests (The percentage of relative error is specified in parenthesis):  b 20.6±0.7 (3%) mg cm-2, c -322±48 (15%) mA g-1, d 4742±495 (10%) mA g-1, e 625±21 (3%) mV  Figure 5.4 shows the surface plots of three different responses studied here for the electrodeposited manganese oxides in presence of Triton X-100, correlating them to the most important factors and two-factor interactions based on the Pareto plots of estimates (Appendix E  ). The Pareto plot analysis of the estimates for factors has shown that the Mn concentration effect on the performance of electrodeposited MnOx is not significant (Appendix E  ). The highest ORR mass activity can be achieved at high surfactant concentration and low temperature (Figure 5.4-A). 158  Moreover, low applied anodic potential is found to further improve the ORR mass activity of the electrodeposited samples. Same trend is observed for the highest OER mass activity since it appears at high surfactant concentration, low temperature and low applied anodic potential (Figure 5.4-B). The lowest ORR/OER potential window can be obtained at high surfactant concentration, low applied anodic potential but high temperature (Figure 5.4-C). The temperature seems to be a defining factor for the optimum bifunctional characteristics of electrodeposited manganese oxides with high temperatures providing low ORR/OER potential window while low temperatures lead to high ORR/OER mass activities (Figure 5.4). Other than the temperature, high surfactant concentration together with low applied anodic potential bring best bifunctional performances for the electrodeposited oxides in presence of Triton X-100. The Rsquare for the responses, i.e. ORR mass activity, OER mass activity and ORR/OER potential window, is 0.987, 0.994 and 0.778, respectively, indicating negligible curvature effect for the first two responses (Rsquare close to 1) while a degree of non-linearity is observed in the behavior of the factors on the third response, i.e. ORR/OER potential window (check Appendix E  ).  159   Figure 5.4 Surface plots for 24-1+3 factorial design in presence of Triton X-100 including factorial responses with the most important factors and two-factor interactions based on the Pareto plots of estimates: A) ORR mass activity at 656 mV, B) OER mass activity at 1556 mV and C) ORR/OER potential window at -2 and 2 mA cm-2, respectively. Details of each run has been given in Table 5.2. Red and green colors in the surface plots correspond to highest and lowest values of each response, respectively.   C) B) A) 160  5.2.3.2 SDS surfactant-assisted electrodeposition In the presence of SDS, -988±45 and 31426±2481 mA g-1 are obtained as highest ORR and OER mass activities, respectively, for run no. 1 at high Mn concentration, low temperature, high surfactant concentration and low applied anodic potential (Table 5.3). While the lowest ORR/OER potential window of 620±7 mV is achieved for run no. 9 at the same conditions but low Mn concentration and low surfactant concentration (i.e. 0%), run no. 1 seems to be a better choice considering its high ORR and OER mass activities as well as second-lowest ORR/OER potential window, i.e. 658±8 mV (Table 5.3).   161  Table 5.3 Design matrix (in random order), Mn valence, calculated loadings and responses for the factorial design experiments in presence of SDS. Other conditions idem to Table 2.1.     Factor levels   Responses No. C T S E Mn valencea Calculated loadings (mg cm-2)b ORR mass activity (mA g-1)c OER mass activity (mA g-1)d ORR/OER potential window (mV)e 1 + - + - 3, 4 5.3 -988 31426 658 2 + - - + 3, 4 1.1 -298 886 1450 3 0 0 0 0 3, 4 12.7 -431 7329 682 4 - + + - 3 10.9 -125 6222 987 5 + + + + 3, 4 20.7 -218 2081 680 6 - + - + 3, 4 8.8 -80 534 985 7 - - + + 3 4.6 -543 39279 779 8 0 0 0 0 3, 4 11.5 -377 7693 699 9 - - - - 3, 4 7.9 -725 20431 620 10 0 0 0 0 3, 4 17.7 -378 5890 671 11 + + - - 3, 4 5.8 -140 405 1330 a The Mn valence calculated based on the XPS results. The bold value corresponds to higher content of that specific Mn valence in the electrodeposited manganese oxide. * Average factorial design response with standard error of the mean value calculated based on the center-point (0000) tests (The percentage of relative error is specified in parenthesis): b 13.9±1.9 (14%) mg cm-2, c -395±18 (5%) mA g-1, d 4742±953 (8%) mA g-1, e 684±8 (1%) mV  Figure 5.5 shows the surface plots representing the relationship between the most important two-factor interactions with different responses for the electrodeposited manganese oxides in presence of SDS as surfactant. The Pareto plot analysis of estimates for factors revealed that the applied anodic potential is the least significant factor contributing to each of the responses for electrodeposited MnOx in this case. LSV graphs in presence of SDS shown in Figure 5.1 further confirm the insignificancy of applied anodic potential in the investigated regime, i.e. 800 to 162  1600 mVMOE, since they all contribute about the same current density for deposition of manganese oxides on nitric acid pre-treated 40% PTFE treated carbon cloth.  As shown in Figure 5.5-A, the highest ORR mass activity of about 600 mA g-1 can be achieved at high surfactant concentration and high Mn concentration. Moreover, low temperature enhances the ORR mass activity of the electrodeposited MnOx. OER mass activity of about 35000 mA g-1 can be obtained at similar conditions of high surfactant concentration and low temperature but low Mn concentration (Figure 5.5-B). The lowest ORR/OER potential window of about 600 mV is achieved at high surfactant concentration, high Mn concentration and low temperature (Figure 5.5-C). One can conclude that the optimum bifunctional responses of deposited manganese oxides are sensitive to the Mn concentration with high Mn concentration leading to highest ORR mass activity and lowest ORR/OER potential window while low Mn concentration provides samples with highest OER mass activity. Overall, high surfactant concentration and low temperature lead to electrodeposited manganese oxides with preferable bifunctional properties.  Moreover, the Rsquare for the responses, i.e. ORR mass activity, OER mass activity and ORR/OER potential window, is 0.997, 0.924 and 0.833, respectively, showing negligible curvature effect for the first two responses (Rsquare close to 1) while a degree of non-linearity is observed in the behavior of the factors on the third response, i.e. ORR/OER potential window (check Appendix E  ).  163   Figure 5.5 Surface plots for 24-1+3 factorial design in presence of SDS including factorial responses with the most important factors and two-factor interactions based on the Pareto plots of estimates: A) ORR mass activity at 656 mV, B) OER mass activity at 1556 mV and C) ORR/OER potential window at -2 and 2 mA cm-2, respectively. Details of each run has been given in Table 5.3. Red and green colors in the surface plots correspond to highest and lowest values of each response, respectively.  C) B) A) 164  5.2.3.3 CTAB surfactant-assisted electrodeposition Looking at Table 5.4, CTAB-assisted electrodeposition of MnOx for run no. 8 generates -774±32 and 49237±2220 mA g-1 as highest ORR and OER mass activities, respectively, at high Mn concentration, low temperature, high surfactant concentration and low applied anodic potential. However, the lowest ORR/OER potential window of about 730 mV is obtained for run no. 5 and 6 at high surfactant concentration but different other electrodeposition factors such as high temperature (Table 5.4). Overall, run no. 6 provides the best compromise between the two types of responses for the bifunctional electrocatalytic activity of the electrodeposited MnOx with the second best ORR/OER mass activities and ORR/OER potential window among the other runs (Table 5.4).   165  Table 5.4 Design matrix (in random order), Mn valence, calculated loadings and responses for the factorial design experiments in presence of CTAB. Other conditions idem to Table 2.1.     Factor levels   Responses No. C T S E Mn valencea Calculated loadings (mg cm-2)b ORR mass activity (mA g-1)c OER mass activity (mA g-1)d ORR/OER potential window (mV)e 1 + - - + 3, 4 1.4 -182 140 1450 2 - - + + 3 3.9 -251 10816 978 3 - + - + 3, 4 10.2 -109 594 919 4 0 0 0 0 3, 4 9.2 -126 7760 859 5 + + + + 3 11.4 -242 8287 726 6 - + + - 3, 4 7.6 -310 13242 740 7 0 0 0 0 3, 4 7.6 -118 1480 1025 8 + - + - 3 1.0 -774 49237 1069 9 0 0 0 0 3, 4 8.0 -250 4826 787 10 + + - - 3, 4 5.2 -125 483 1312 11 - - - - 3, 4 1.2 -144 114 1450 a The Mn valence calculated based on the XPS results. The bold value corresponds to higher content of that specific Mn valence in the electrodeposited manganese oxide. * Average factorial design response with standard error of the mean value calculated based on the center-point (0000) tests (The percentage of relative error is specified in parenthesis): b 8.3±0.7 (6%) mg cm-2, c -231±10 (4%) mA g-1, d 4847±219 (5%) mA g-1, e 857±40 (5%) mV  Figure 5.6 represents surface plots depicting effect of most important two-factor interactions on the three responses during CTAB-assisted electrodeposition of MnOx. The estimate analysis of factors using Pareto plots revealed that applied anodic potential in the investigated range has lesser effect on the responses comparing to the other three electrodeposition factors. In presence of CTAB, LSV graphs have also shown almost a constant current density through the whole potential range investigated here, i.e. 800 to 1600 mVMOE, confirming that the applied anodic potential is 166  not a major factor affecting the sample responses during surfactant-assisted electrodeposition of MnOx (Figure 5.1). According to the factorial design analysis showcased as surface plots in Figure 5.6, the highest ORR mass activity of about 450 mA g-1 can be obtained at high Mn concentration and low temperature. High surfactant concentration can further enhance the ORR mass activity of the electrodeposited MnOx. The highest OER mass activity of almost 30000 mA g-1 is achieved at similar conditions of high surfactant concentration, low temperature and high Mn concentration (Figure 5.6-B). When it comes to the ORR/OER potential window, values as low as 800 mV can be reached at high surfactant concentration but low Mn concentration and high temperature, unlike the mass activity cases (Figure 5.6-C). The optimum level of electrodeposition factors for the best ORR/OER bifunctional performance of MnOx samples depends on the definition of each response. While the ORR/OER potential window reflects on the catalyst performance at low current densities neglecting the effect of loading, ORR and OER mass activities provide better insight on the high current density responses of electrodeposited MnOx, normalized based on the calculated catalyst loading. Both responses are valuable as the latter reflects on more practical version of the ORR/OER catalyst performance covering the implications of mass transport limitations due to the loading differences whereas the former looks at the intrinsic bifunctional activity of the electrodeposited MnOx with different morphologies and crystal structures.  In the case of CTAB as surfactant, the Rsquare is 0.906, 0.955 and 0.848 for the ORR mass activity, the OER mass activity and the ORR/OER potential window, respectively. While a degree of non-linearity is observed in the behavior of the factors, i.e. C, T, S and E, on the ORR mass activity as well as the ORR/OER potential window, negligible curvature effect (Rsquare close to 1) is shown for the OER mass activity (check Appendix E  ).   167   Figure 5.6 Surface plots for 24-1+3 factorial design in presence of CTAB including factorial responses with the most important factors and two-factor interactions based on the Pareto plots of estimates: A) ORR mass activity at 656 mV, B) OER mass activity at 1556 mV and C) ORR/OER potential window at -2 and 2 mA cm-2, respectively. Details of each run has been given in Table 5.4. Red and green colors in the surface plots correspond to highest and lowest values of each response, respectively.  C) B) A) 168  5.2.4 ORR/OER performance comparison Figure 5.7 shows a comparison between the ORR and OER performances of the most active electrodeposited MnOx from each surfactant category. XPS studies have shown a mixture of Mn3+ and Mn4+ with higher contents of Mn3+ species including MnOOH for these samples (Table 5.2, Table 5.3 and Table 5.4). However, it is not possible to determine the exact content of each Mn species in different samples or identify the crystal structure of MnOx by only employing XPS analysis. Further structural analysis is needed to find the crystallographic forms of these electrodeposited Mn oxides.  In the ORR region, the electrodeposited MnOx labeled as Triton run no. 9 reveals the highest ORR electrocatalytic activity with 1.4, 2.2 and 64 times the ORR current densities (at 406 mV) of the SDS run no. 1, CTAB run no. 6 and carbon cloth substrate samples, respectively. The characteristic cathodic peaks observed for Triton run no. 9 at about 706 and 406 mV resemble the performance of β-MnO2 (pyrolusite) in alkaline media with the first peak mainly due to the reduction of adsorbed oxygen on the unreduced active Mn4+ sites and the latter due to the reduction of dissolved oxygen on Mn3+/Mn4+ surfaces based on the mechanism discussed in eqs. 17-19 [45, 90, 98, 110, 268]. Pyrolusite has been known to possess largest oxygen sensitive specific area comparing to the other crystallographic forms of MnO2, adsorbing high contents of O2 and reducing it at more positive ORR potentials in alkaline media [90]. The reduction peak corresponding to adsorbed oxygen is reported to disappear at high rotation speeds when the mass transport limitations for the reduction of bulk dissolved oxygen (second cathodic peak) are lifted, or in the case of crystal structures with low ability of oxygen adsorption such as electrodeposited MnOx at SDS no. 1 and CTAB no. 6 [90]. One can also refer the enhanced ORR activity of electrodeposited oxide in Triton run no. 9 to its nano-sized petal-like microstructure of nano sheets 169  with high porosity, as shown in Figure 5.8-A and B. This unique microstructure can also provide high surface area, further improving the ORR activity of the sample (Figure 5.7-A, Figure 5.8-A and B). The lowest ORR performance obtained by the CTAB sample could be explained by its micron-sized petal-shape microstructure with needle-like fibers (having high length to thickness ratio) (Figure 5.7-A, Figure 5.8-C and D) comparing to SDS run no. 1 exhibiting nano-sized sphere-shape protrusions (between 200-400 nm in diameter) (Figure 5.8-E and F).  In the OER region, electrodeposited MnOx for Triton run no. 9 shows the lowest onset potential and highest OER current densities while being followed by the other two manganese oxides with virtually the same OER activity, e.g. OER onset potential of 1435 mV at 10 mA cm-2 for Triton run no. 9, about 30 mV lower than the SDS and CTAB samples and over 400 mV lower than the carbon substrate (Figure 5.7-B). Su et al. showed both experimentally and theoretically via DFT calculations that the β-MnO2 is one of the most active forms of MnOx for OER in alkaline media [121]. This, with the unique nano-sized porous petal-shape microstructure could be the main reasons behind enhanced OER electrocatalytic activity of electrodeposited manganese oxide at Triton run no. 9 sample (Figure 5.7-B, Figure 5.8-A and B). One can conclude that the surface modification of MnOx via surfactant-assisted electrodeposition can help destabilizing the HOO(ads) and HO(ads) intermediates by introducing more binding sites such as lattice oxygen vacancies, breaking away from the linear scaling relationship between their binding energies, enhancing the ORR and OER electrocatalytic activity of electrodeposited manganese oxides. While Su et al. proposed that low number of adsorbed water molecules enhances the ORR activity based on the DFT calculations, Staszak-Jirkovský et al. argue that it could be more complicated than that since sensitive interaction between covalently bonded OH(ads) and water molecules can form hydrogen-bonded complexes, i.e. HO(ads)…H-OH, with specially configured water molecules called 170  “activated water”, acting as a promoter for ORR [121, 269]. This can further enhance the ORR activity of the catalyst by facile transfer of protons to weakly adsorbed HOO(ads)/O(ads) intermediates and breaking away from the linear universal scaling relationship between binding energies of HOOads) and HO(ads) [121, 269]. The surface coverage of OH(ads) has a defining role on the electrocatalytic activity of the catalyst since an optimum coverage is needed to provide sites for formation of HO(ads)…H-OH complexes (promoter effect) as well as having bare catalyst sites necessary for formation of other reaction intermediates such as HOO(ads) and O(ads) (spectator effect) [269]. The comparison between ORR and OER overpotentials (at -2 and 2 mA cm2, respectively) for the electrodeposited manganese oxides investigated here with relevant catalyst materials from the literature, reveals modest ORR activity but superior electrocatalytic activity towards OER, i.e. ORR overpotential of -366 mV (at -2 mA cm-2) for the Triton run no. 9, between 50 to 150 mV more positive comparing to other non-precious metal catalysts such as Mn2O3 [67], CoMn2O4 [71], nanostructured Mn oxide thin film [69] and Core-Corona Structured Bifunctional Catalyst (CCBC) [76] as well as OER overpotential of 234 mV (at 2 mA cm-2) for the Triton run no. 9, min. 100 mV lower than 20 wt% Ir/C and 20 wt% Ru/C [67, 69] (Figure 4.16 and Figure 5.7).   171    Figure 5.7 IR-corrected bifunctional performance comparison of electrodeposited MnOx in presence of different surfactants, i.e. Triton X-100, SDS and CTAB: A) ORR, B) OER. The electrodeposition factors for each sample are as follows: Carbon cloth substrate (no electrodeposited material), CTAB-Run no. 6 (C: 0.1 M, T: 343 K, S: 10 vol%, E: 800 mVMOE), SDS-Run no. 1 (C: 0.3 M, T: 295 K, S: 10 vol%, E: 800 mVMOE), Triton-Run no. 9 (C: 0.1 M, T: 343 K, S: 10 vol%, E: 800 mVMOE). Electrolyte: O2 saturated 6 M KOH at 293 K and PO₂ of 1 atm. Rotating electrode speed and potential scan rate were 400 rpm and 5 mV s-1, respectively. Cycle number five is reported in all cases. 172     Figure 5.8 SEM micrographs of best performing electrodeposited MnOx oxides on nitric acid pre-treated carbon cloth: A and B) Triton run no. 9, C and D) SDS run no. 1, E and F) CTAB run no. 6. The electrodeposition factors are stated in Figure 5.7. A B C D E F 173  5.2.5 Galvanostatic long-term durability testing of deposited MnOx Figure 5.9 presents the galvanostatic polarization comparison of the best performing electrodeposited MnOx, i.e. Triton run no. 9, and commercial manganese oxides. These galvanostatic polarization tests provide only a preliminary insight into durability of catalysts since the applied current density and test duration were chosen to avoid limitations associated with the employed flooded electrode half-cell design such as limited dissolved oxygen, needed during ORR, and heavy oxygen gas evolution, shielding the electrode surface during OER. In practice, air/oxygen breathing gas diffusion electrodes and electrolyte circulation would be employed as solutions to avoid dissolved oxygen limitations during ORR and O2 gas evolution during OER, respectively.  In the OER region, electrodeposited MnOx sample in presence of Triton X-100 shows the highest OER electrocatalytic activity comparing to commercial manganese oxides, i.e. OER potentials of 1446, 1489 and 1694 mV (at 5 mA cm-2 and t=2 hrs) for the Triton run no. 9, commercial MnOx and γ-MnO2 samples, respectively (Figure 5.9). The rate of OER potential increase for the Triton run no. 9 sample is the second low, i.e. 43 mV h-1, for the total 2 hrs of testing comparing to 51 and 14 mV h-1 for commercial MnOx and γ-MnO2, respectively. However, the electrodeposited manganese oxide provides the lowest OER degradation rate of 5 mV h-1 for the second half of the OER testing compared to others tested here (Figure 5.9).  For the ORR part, commercial MnOx delivers the highest ORR electrocatalytic activity compared to others, i.e. electrode potentials of 856, 840 and 754 mV (at -2 mA cm-2 and t=2 hrs) for commercial MnOx, γ-MnO2 and the Triton run no. 9 samples, respectively (Figure 5.9). In terms of stability, however, γ-MnO2 sample possesses the lowest ORR potential degradation, i.e. 174  -2.4 mV h-1 comparing to -173 and -29 mV h-1 for the Triton run no. 9 and commercial MnOx samples, respectively (Figure 5.9).    Figure 5.9 Galvanostatic polarization comparison of in-house made and commercial manganese oxide GDEs: I) Manganese oxide/C GDE from Gaskatel GmbH (20 mg cm-2), II) γ-MnO2/C from Sigma Aldrich (loading 5.6 mg cm-2) and III) Triton run no. 9 (C: 0.1 M, T: 343 K, S: 10 vol%, E: 800 mVMOE) (calculated loading 17.5 mg cm-2). The galvanostatic polarization started at 5 mA cm-2 anodically applied to each GDE for 2 hrs followed by -2 mA cm-2 applied cathodically for 30 min in O2 saturated 6 M KOH. The rotation speed and temperature were 400 rpm and 293 K, respectively. PO₂ was 1 atm.  5.3 Conclusion A comprehensive study was performed via 2n half-fraction factorial design to investigate the effects of main factors, such as Mn2+ concentration, applied potential, temperature, surfactant 175  concentration, as well as their two-factor interactions on the ORR and OER electrocatalytic activity of anodically electrodeposited manganese oxides. Triton X-100 as non-ionic, Sodium dodecyl sulfate (SDS) as anionic and hexadecyl-trimethyl-ammonium bromide (CTAB) as cationic surfactants were employed in this chapter to electro-synthesize nanostructured MnOx while several surface characterization methods were used to analyze morphology and Mn valence of the synthesized catalysts. In the Triton X-100 cases, high surfactant concentration together with low applied anodic potential are believed to bring the best ORR/OER bifunctional performances for the electrodeposited Mn oxides. Mn concentration was found to be an insignificant player. Temperature, on the other hand, is believed to have different effect depending on its level with high temperature providing low ORR/OER potential window while low temperature leads to high ORR/OER mass activities. In the SDS category, the ORR/OER bifunctional responses of the deposited manganese oxides were sensitive to the Mn concentration with high Mn concentration leading to highest ORR mass activity and lowest ORR/OER potential window, while low Mn concentration provides samples with highest OER mass activity. Overall, high surfactant concentration and low temperature were found to lead to preferable bifunctional activities. For the CTAB samples, the highest ORR and OER mass activities were found to achieve at high surfactant concentration, low temperature and high Mn concentration. When it comes to the ORR/OER potential window, high temperature and low Mn concentration were more favorable. The effect of anodic applied potential on the ORR and OER activities of the electrodeposited samples was found to be negligible in the case of SDS and CTAB surfactants.  The surface modifications of MnOx via surfactant-assisted electrodeposition can significantly alter the morphology and Mn valence in the deposited material, provide new binding sites for the HOO(ads) and HO(ads) intermediates, help break away from the linear scaling relationship between 176  their binding energies as a major contributor to the ORR and OER overpotentials and hence, enhance the ORR and OER electrocatalytic activity of electrodeposited manganese oxides. The formation of hydrogen-bonded complexes, i.e. HO(ads)…H-OH, with specially configured water molecules called “activated water”, can further enhance the ORR activity of the catalyst by facile transfer of protons to weakly adsorbed HOO(ads)/O(ads) intermediates and break away from the linear universal scaling relationship. This, however, depends on the surface coverage of OH(ads) which provides binding sites for formation of HO(ads)…H-OH complexes (promoter effect). The electrodeposited MnOx for Triton run no. 9 was found to show the best ORR and OER electrocatalytic activities among other deposited manganese oxides investigated here, mainly due to a possible crystal structure of β-MnO2 and a nano-sized petal-like microstructure of nano sheets with high porosity. Comparing to wide range of noble metals and their oxides such as Ir, Ru and IrO2, the electrodeposited manganese oxide for Triton run no. 9 showed lower OER overpotential (min. 100 mV) at 2 mA cm-2. For the ORR, it provides between 50 to 150 mV more positive overpotential at -2 mA cm-2 compared to the other non-precious metal compounds such as CoMn2O4 and Core-Corona Structured Bifunctional Catalyst (CCBC). The galvanostatic polarization tests further confirmed the promising OER activity for the Triton run no. 9 sample with potentials as low as 1446 mV (at 5 mA cm-2 and t=2 hrs), about 40 mV lower than the commercial MnOx, and a degradation rate of 43 mV h-1, about 10 mV h-1 lower than its commercial counterparts.   177  Chapter 6: Carbon support effect on ORR/OER bifunctional activity and durability of non-PGM mixed-oxide catalyst: Graphene vs. commercial carbon materials †††† 6.1 Introduction This chapter aims at investigating the electrocatalytic activity and long-term durability of different carbonaceous materials, i.e. Vulcan XC-72, multi-walled carbon nanotubes (MWCNT), graphene and N-doped graphene, either as a support for a highly active ORR/OER bifunctional non-PGM catalyst, i.e. MnO2-LaCoO3, or an oxygen electrocatalyst itself in alkaline media. RRDE measurements were performed to give an insight on the mechanism involved on each carbonaceous material during ORR. Cyclic voltammetry tests on the carbons further revealed Vulcan XC-72 to be the most active bifunctional catalyst for both ORR and OER among other carbons investigated here. The carbonaceous materials employed as supports for MnO2-LaCoO3 were found to contribute differently to each part of oxygen electrocatalysis on the non-PGM catalyst. For ORR, the order of electrocatalytic activity for the samples containing MnO2-LaCoO3 vs. Pt is as follows: MWCNT-Graphene > Vulcan XC-72 > Pt > MWCNT-N-doped graphene > MWCNT > Graphene > N-doped graphene. However, the OER electrocatalytic activity of oxides containing carbons vs. Pt follows a different trend: MWCNT-Graphene > MWCNT-N-doped graphene > MWCNT >                                                  †††† A version of this chapter is in preparation for publication. Parts of this chapter were filed as United States and Canadian patent applications: - P. Hosseini-Benhangi, M. A. Garcia-Contreras, A. Alfantazi, E. Gyenge, “Carbon support effect on ORR/OER bifunctional activity and durability of non-PGM mixed-oxide catalyst: Graphene vs. commercial carbon materials”, to be submitted. - E. Gyenge, P. Hosseini-Benhangi, “An oxygen electrode and a method of manufacturing the same”, U.S. (15/251,267) and Canadian (2,940,921) patent applications, filed on August 30th, 2016.  178  Vulcan XC-72 > Pt > Graphene > N-doped graphene. The ORR/OER bifunctional durability performances of GDEs containing MnO2-LaCoO3 supported on MWCNT-Graphene and Vulcan XC-72 were tested using an accelerated degradation testing protocol. Overall, MnO2-LaCoO3-MWCNT-Graphene GDE showed better durability than the Vulcan XC-72 containing sample. The main factors affecting the performance degradation of mixed-oxide GDEs during ORR and OER were discussed. The role of lanthanum cobalt oxide in the performance loss of the mixed-oxide catalyst during accelerated degradation test was concluded to be insignificant. A wide range of characterization techniques including SEM and TEM studies were employed to investigate the microstructure-property relations as well. 6.2 Results and discussion 6.2.1 Microstructural studies of catalyst and support materials Figure 6.1 presents the TEM images of carbonaceous materials as well as non-PGM oxides, i.e. MnO2 and LaCoO3, used as support and catalyst material in this study. The microstructure of graphene consists of crumpled and randomly oriented large-area graphene sheets together with small number of flakes with an average length of 200 nm (Figure 6.1-A). Low-magnification TEM images show the tendency of graphene sheets to agglomerate even after hours of sonication in IPA. For N-doped graphene sample, the morphology is almost the same as graphene with less flakes and more ultrathin graphene sheets (Figure 6.1-B). The material’s tendency to agglomerate even after prolonged sonication in IPA is observed. Figure 6.1-C shows the nano-sized carbonaceous particles of Vulcan XC-72 that range from 20 to 30 nm in size. Excellent particle distribution is observed in low-magnification TEM images for Vulcan XC-72 dispersed in IPA. As for MWCNT, TEM micrographs show tangled tubes with outer diameter of 8-18 nm and length of 0.5-5 µm with modest distribution in IPA (Figure 6.1-D). The MnO2 particles vary in shape from pillars with 179  various length-to-width ratios to smooth spheres (Figure 6.1-E). For pillar-shaped particles, the length-to-width ratio changes from 3 to 7.5 with the lengths as small as 30 nm to max. 450 nm (Figure 6.1-E). The size of sphere-like MnO2 particles goes from 70 to 100 nm in diameter (Figure 6.1-E). LaCoO3 particle size extends between 50-100 nm (Figure 6.1-F).      180         Figure 6.1 TEM images of materials investigated in this chapter as oxygen catalyst or catalyst support: A) Graphene, B) N-doped graphene, C) Vulcan XC-72, D) MWCNT, E) MnO2 and F) LaCoO3.  181  6.2.2 RRDE studies of carbonaceous materials for ORR As mentioned in section 1.2.2.2, the ORR in alkaline media is composed of series of complex electrochemical reactions with two proposed pathways: a) Four-electron pathway or b) Two-electron pathway (Table 1.2). In order to better understand the ORR pathway on carbonaceous materials, RRDE experiments were performed (Appendix B  and Figure 6.2). A closer look at the shape of O2 electro-reduction voltammograms on different catalysts reveals a noticeable downward peak between 670 and 750 mV for all carbonaceous materials (Figure 6.2-A).  This reduction peak becomes less visible after increasing the rotation speed from 400 to 2500 rpm (Figure 6.2-A). The data for other rotation speeds are not presented in the Figure 6.2-A for the sake of clarity. The presence of this peak at low rotation speeds is attributed to the electro-reduction of adsorbed oxygen on the catalyst layer [270-272]. Increasing the rotation speed leads to higher diffusion-controlled ORR currents for the oxygen in bulk electrolyte, gradually masking the peak current from adsorbed O2. The absence of adsorbed O2 peak during the ORR on N-doped graphene could be related to its flat morphology of flakes and ultrathin graphene sheets (Figure 6.1-B) which limits the presence of adsorbed oxygen in different layers of catalyst material. As shown in Figure 6.2-A, Vulcan XC-72 possesses the highest electrocatalytic activity for ORR, i.e. lowest ORR onset potential of 781 mV at -0.2 mA cm-2 and highest ORR limiting current density of -0.248 mA cm-2 at 600 mV, followed by MWCNT, graphene and N-doped graphene, respectively. The carbonaceous materials that showed a peak in the early stages of ORR, possess distinguishable peaks in the same potential region for calculated hydrogen peroxide ion content (Figure 6.2-B) and the number of electrons involved in the ORR (Figure 6.2-C). This means that the adsorbed oxygen is going through a different ORR pathway comparing to the oxygen from bulk electrolyte. As an example, 182  Vulcan XC-72 shows HO2− content of 58% at 725 mV which corresponds to 2.8e− while it moves between 90% to 100% for HO2− ions below 660 mV, showing 2e− transfer during the ORR for oxygen from bulk electrolyte (Figure 6.2-B and C). This further reveals that the ORR is going through two competitive 2e− and 4e− pathways near the peak for the electro-reduction of adsorbed O2. It is most likely that the adsorbed oxygen coordinates parallel to the catalyst surface, showing direct 4e− pathway (eqs. 12, 15 and 16) while the ORR for the O2 in the bulk electrolyte follows a 2e− pathway leading to hydrogen peroxide ions (eqs. 12, 13 and 14). For both MWCNT and graphene, the competition between the 2e− and 4e− pathways in the potential region of adsorbed oxygen reduction continues with about 60% HO2− content and 2.8 e− (Figure 6.2-B and C). Out of this region, MWCNT favors 2e− pathway more while graphene shows lower HO2− content, moving toward 4e− pathway (Figure 6.2-B and C). N-doped graphene, however, depicts lowest content of hydrogen peroxide ions below 710 mV, meaning it favors 4e− pathway even though the 2e− pathway is dominant at some sites (Figure 6.2). While it is generally believed that the 4e− ORR pathway is dominant on metals and their oxides (both noble and transition metals) and the 2e− pathway primarily prevails on the surface of carbonaceous materials [2, 10, 17, 45, 90, 98, 110], many studies reported a mixture of 2e− and 4e− pathways for MWCNT, graphene and N-doped graphene with the nitrogen-doped graphene showing close-to-four values for the electrons involved in ORR per each mole of oxygen [10, 100, 199, 206, 273]. Lower ORR performances of graphene and N-doped graphene, despite the higher number of electrons involved comparing to Vulcan XC-72, could be related to the flat morphology of graphene-based materials, blocking the pores necessary for the electrolyte to reach the active sites and hence, reducing their ECSA for the same loadings of Vulcan XC-72 (Figure 6.2).  183    Figure 6.2 Rotating ring disk electrode results for ORR on Graphene (I), N-doped Graphene (II), MWCNT (III) and Vulcan XC-72 (IV): A) O2 reduction current densities obtained from disk electrode (id) when polarized from 1170 to 500 mV (bottom) and the corresponding oxidation current densities on the ring at 1353 mV (ir) as a function of disc potentials (top). B) Calculated percentage of hydrogen peroxide ions produced during ORR (%HO2−). C) Calculated number of electrons transferred per molecule of oxygen during ORR (n). Carbon loadings of 0.5 mg cm-2 each. O2 saturated 6 M KOH. 5 mV s-1. 293 K. PO₂ of 1 atm.184  6.2.3 ORR/OER bifunctional electrocatalytic activity measurements: Carbonaceous materials as oxygen catalysts  Figure 6.3 shows the ORR/OER electrocatalytic activity of the investigated carbonaceous materials at their initial stage. Higher ORR peak current densities for the carbonaceous materials, comparing to the ones from RRDE experiments on the same carbons, are mainly due to the employed testing protocol for assessing the ORR/OER bifunctional performance of these materials where the samples were anodically polarized first, forming both adsorbed and bulk oxygen at the active sites during OER. The resulting oxygen layer is easily accessible to be reduced during the subsequent cathodic polarization, enhancing the ORR current densities on the carbonaceous materials (compare Figure 6.2-A and Figure 6.3-A). While Vulcan XC-72 demonstrates the highest ORR peak current density of -0.9 mA cm-2 at around 760 mV, MWCNT containing catalyst layers, i.e. MWCNT, MWCNT-N-doped graphene and MWCNT-graphene, possess the lowest ORR overpotentials of about 305 mV at -0.2 mA cm-2 (Figure 6.3-A). Small synergistic effect after 593 mV is observed when MWCNT is mixed with N-doped graphene (Figure 6.3-A). Similar to the RRDE experiments, N-doped graphene shows the poorest ORR electrocatalytic activity when used as lone oxygen catalyst (Figure 6.2-A and Figure 6.3-A). As explained earlier, this is mainly due to the flat morphology of N-doped graphene with high content of ultra-thin sheets that tend to agglomerate and lie on top of each other (Figure 6.1-B), effectively reducing the ECSA of the N-doped graphene, leading to its poor ORR performance even though high ORR electrocatalytic activity for N-doped carbon-based materials is reported in the literature [10, 100, 199, 206, 273]. For the OER, Vulcan XC-72 reveals the highest electrocatalytic activity up to 1650 mV with the lowest OER overpotential of 211 mV at 0.2 mA cm-2 (Figure 6.3-B). A significant synergistic effect is observed when either graphene or N-doped graphene is mixed with MWCNT, showing 185  13 and 150 times, respectively, the OER current densities (at 1600 mV) of the individual graphene-based materials (Figure 6.3-B). Moreover, the MWCNT-graphene surpasses the OER electrocatalytic activity of Vulcan XC-72 after 1650 mV, further confirming the synergistic effect of mixed carbons (Figure 6.3-B). This synergistic effect can be attributed to the unflattened morphology of the carbonaceous mixture, i.e. MWCNT with either graphene or N-doped graphene. This morphology provides a porous structure with increased ECSA for ORR as well as necessary pore density for oxygen bubble evolution and release, which itself enhances OER performance of the mixture significantly [5, 226].   186    Figure 6.3 Bifunctional ORR/OER performance of carbonaceous materials, i.e. Graphene, N-doped graphene, MWCNT, Vulcan XC-72, MWCNT-Graphene and MWCNT-N-doped graphene: A) ORR, B) OER. Initial stage IR-corrected polarization curves obtained by potential scanning between 475 to 1673 mV, starting with anodic polarization. Carbon loadings are fixed at 0.5 mg cm-2, each. In cases where two materials were mixed, a weight ratio of 1:1 was used. Cycle number five is reported in all cases. O2 saturated 6 M KOH. 293 K. 400 rpm. 5 mV s-1. PO₂ of 1 atm.  187  The ORR Tafel plots further confirms the enhanced ORR electrocatalytic activity of MWCNT containing carbonaceous catalyst materials, i.e. MWCNT, MWCNT-graphene and MWCNT-N-doped graphene, comparing to each of the individual graphene-based materials as well as Vulcan XC-72 (Figure 6.4-A). The best performing ORR catalyst material, i.e. MWCNT-graphene reveals small synergistic effect comparing to each of its constituents at overpotentials more negative than -280 mV with Tafel slope and exchange current density values of -41mV dec-1 and 7.7×10-6 µA cm-2, respectively (Figure 6.4-A and Table 6.1). For the OER, the superior electrocatalytic activity of Vulcan XC-72 is easily observable as shown in Figure 6.4-B. However, almost all of the carbonaceous materials (excluding N-doped graphene) show abnormal Tafel slopes (i.e. 223 to 444 mV dec-1) and orders of magnitude higher OER exchange current densities comparing to the ORR part (Figure 6.4-B and Table 6.1). This could be related to the fact that the OER is not the main reaction occurring on the surface of carbonaceous materials at anodic potentials in alkaline media [204, 205]. Ross et al. reported that part of the overall anodic current obtained from acetylene black surface in the OER region is due to the corrosion of carbon, which is a function of both potential and temperature [204, 205]. According to their study, the carbon corrosion at room temperature mainly comes from two distinctive processes: 1) Dissolution to carbonate ion (CO32-) and 2) Carbon gasification to carbon monoxide [204, 205]. While the rate of these processes are different for carbons made under different treatments or catalyzed with diverse catalysts [203-205, 274], their occurrence decreases the current efficiency of OER which leads to unusual values for the calculated OER kinetic parameters such as Tafel slope and exchange current density (Figure 6.4-B and Table 6.1). It is important to note that the current efficiency of the oxygen evolution reaction on both catalyzed and un-catalyzed carbons is reported to increase at high anodic potentials, going over 75% at about 1600 mV in concentrated KOH solutions [204].  188    Figure 6.4 IR-corrected bifunctional ORR/OER Tafel-lines of investigated carbonaceous catalyst materials: A) ORR, B) OER. The numbers associated with each line represent the respective apparent Tafel slopes. Cycle number five is reported in all cases. The N2-baseline is subtracted from the ORR part for all samples. For the OER part, capacitive background current at open circuit potential is removed from each corresponding voltammogram. Other conditions idem to Figure 6.3.  189  Table 6.1 The apparent exchange current densities and Tafel slopes for the initial stage ORR and OER activities of the investigated carbonaceous catalyst materials in Figure 6.3. The exchange current densities are expressed per geometric area. The apparent Tafel slope and exchange current density values are obtained over a potential range of min. 50 mV on six replicates. Other conditions idem to Figure 6.3. Catalyst bORR  (mV dec-1) i0,ORR  (µA cm-2) bOER (mV dec-1) i0,OER (µA cm-2) Graphene -46 ±1 4.6×10-5  ±0.09×10-5 263 ±5 1.3 ±0.03 N-doped graphene -122 ±2 6.3×10-2 ±0.1×10-2 37 ±1 9.3×10-12  ±0.2×10-12 MWCNT -55 ±1 4.7×10-4 ±0.09×10-4 299 ±6 15 ±0.3 Vulcan XC-72 -83 ±2 2.0×10-2 ±0.04×10-2 444 ±9 101 ±2 MWCNT-Graphene -41 ±1 7.7×10-6 ±0.2×10-6 223 ±4 7.4 ±0.1 MWCNT-N-doped graphene -51 ±1 2.3×10-4 ±0.05×10-4 249 ±5 12 ±0.2   190  6.2.4 ORR/OER bifunctional electrocatalytic activity measurements of non-PGM oxide oxygen catalyst: MnO2-LaCoO3 supported on carbonaceous materials   Figure 6.5 shows the initial stage cyclic voltammograms of mixed-oxide catalyst, i.e. MnO2-LaCoO3 (1:1 weight ratio), supported on different investigated carbonaceous materials in O2 saturated 6 M KOH. The cyclic voltammogram of 50 wt% Pt/Graphitized carbon has been added as a baseline for comparison. The Pt voltammogram reveals two distinctive oxidation and reduction peaks at 861 and 756 mV which are associated with adsorption of hydroxyl ion (OH−), also known as formation of PtOH layer, and reduction of the oxide layer (PtOH), respectively (Figure 6.5) [275, 276]. The addition of a MnO2-LaCoO3 catalyst to the carbons significantly enhances the ORR/OER bifunctional performance of the carbonaceous materials (compare Figure 6.3 and Figure 6.5). Compared to Pt, the activity of MnO2-LaCoO3-N-doped graphene toward oxygen reduction and evolution reactions is virtually absent mainly due to its flattened morphology of ultra-thin sheets (Figure 6.5-A and Figure 6.1-B). A considerable synergistic effect is observed when MWCNT is being physically mixed with either N-doped graphene or graphene as a support for MnO2-LaCoO3 (Figure 6.5). The overall best performing catalyst layer was found to be the MnO2-LaCoO3-MWCNT-Graphene with visually enhanced ORR and OER waves, e.g. ORR and OER mass activities of -6.694 A g-1 at 850 mV and 15.536 A g-1 at 1650 mV (Figure 6.3 and Figure 6.5). This enhanced ORR/OER bifunctional electrocatalytic activity can be better presented in comparison to the benchmark used in this study, i.e. 50 wt% Pt/graphitized carbon performance. The Pt sample possesses ORR mass activity of -3.828 A g-1 (at 850 mV) and OER mass activity of 7.992 A g-1 (at 1650 mV), about 43% and 48%, respectively, lower than the ones associated with the MnO2-LaCoO3-MWCNT-Graphene catalyst layer (Figure 6.5-C). 191        Figure 6.5 Bifunctional ORR/OER performance of MnO2-LaCoO3 catalyst supported on various carbonaceous materials: A) Graphene and N-doped graphene, B) MWCNT and Vulcan XC-72, C) MWCNT-Graphene and MWCNT-N-doped graphene. The 50 wt% Pt/Graphitized carbon is being shown as the baseline for comparison (Pt loading of 0.25 mg cm-2). The oxide and carbon loadings were 0.5 mg cm-2 each. A weight ratio of 1:1:1:1 for MnO2:LaCoO3:Carbon1:Carbon2 (if available) was used. Cycle number five is reported in all cases.  Other conditions idem to Figure 6.3. 192  As discussed in section 1.2.2.4, a major contributor to ORR and OER overpotentials is found to be the difference in the binding energy levels of HOO(ads) and HO(ads) intermediates, known as the universal scaling relationship [121, 124, 133]. Breaking away from this confining linear scaling relationship is only attainable by modifications applied to catalyst surfaces and not blindly browsing other possible elements active for both ORR and OER. These modifications can include, but not limited to, manipulating the crystal structure of catalyst materials with introduction of dopants, defects or magnetic fields as well as changing the nature of catalyst/support interactions at their interfaces [122]. Now, this could explain the different ORR/OER electrocatalytic activity shown for mixed-oxide catalyst, i.e. MnO2-LaCoO3, supported on the wide range of carbonaceous materials studied here (Figure 6.5). To explain the synergistic effect of MWCNT mixed with either graphene or N-doped graphene, the combination could have provided near-the-ideal configuration/distribution of the materials for the electron transfer during both ORR and OER on mixed oxides, help destabilizing the HOO(ads) and HO(ads) intermediates with the new available binding sites and radically breaking away from the linear scaling relationship between their binding energies, thus enhancing the ORR and OER activity of the catalyst layer compared to other investigated oxygen catalysts here (Figure 6.5-C).  Turning the attention to the OER performance of studied MnO2-LaCoO3-Carbons catalyst layers, there are two other factors that can partially interfere with the anodic currents obtained in the OER region: 1) Carbon corrosion, 2) MnO2 electro-corrosion (eq. 39).   𝑀𝑛𝑂2 + 4𝑂𝐻− → 𝑀𝑛𝑂4− + 2𝐻2𝑂 + 3𝑒−   (𝐸298𝐾0 = 0.595 𝑉𝑆𝐻𝐸) (39)  193  As discussed in section 6.2.3, it has been shown that the rate of carbon corrosion processes on catalyzed or un-catalyzed carbons in concentrated alkaline solutions decreases dramatically at certain anodic potentials (about 1600 mV in 30 wt% KOH at 318 K) [204]. Different mechanisms have been proposed for this behavior with the most plausible theory based on limited wet area when gas bubble formation and entrapment in the microporous sublayer pushes out the electrolyte and dries the interior layer, effectively lowering the ECSA and hindering the electro-oxidation reactions to occur [205, 277]. This, however, fails to explain the experimentally-backed data showing an increase in the OER rate of catalyzed carbon at high anodic potentials where carbon corrosion rate drops dramatically, Staud and Ross argued [205]. With this in mind and the fact that the anodic current densities (at 1650 mV) for catalyzed MWCNT-N-doped graphene and MWCNT-Graphene samples are 7 and 17 times the ones for un-catalyzed samples, respectively, one can conclude that the carbon corrosion effect is likely to be negligible here, specially at high anodic potentials (Figure 6.3 and Figure 6.5-C).  While thermodynamics indicates the second possible corrosion factor in the OER region, i.e. MnO2 electro-corrosion to MnO4− (eq. 39), should take place at above 1690 mV, it kinetically depends on the nature of the electrolyte and the available ions [230]. Gao et al. studied the effects of alkali-metal cations on electro-corrosion of MnOx using atomic absorption spectroscopy, showing that small electro-corrosion currents diminish at a certain anodic potential (about 1900 mV in 0.5 M KOH) to almost zero [230]. They argue that it could be due to the formation of a passivating oxide layer without much effect on the catalyst’s electrocatalytic activity for OER [230]. Hence, it can be inferred that the major portion of anodic currents in the potential range studied here is associated with the OER catalyzed by MnO2-LaCoO3-Carbon(s) catalyst layers (Figure 6.5).  194  Next, the Tafel plots associated with the MnO2-LaCoO3 catalyst supported on wide range of carbons are shown in Figure 6.6, with the Tafel slopes and exchange current densities summarized in Table 6.2. With respect to ORR, the Tafel slope for different samples ranges from -52 to -86 mV dec-1, all within the kinetic control region limits (Figure 6.6). MWCNT-Graphene sample possesses the highest electrocatalytic activity with Tafel slope and exchange current density of -84 mV dec-1 and 2.1 µA cm-2, respectively, followed by Vulcan XC-72 and Pt (Figure 6.6-A and Table 6.2). For the OER part, Tafel slop values extend from 83 to 127 mV dec-1, still within the kinetic controlled region but near the max. limit, i.e. 120 mV dec-1 (Figure 6.6). This is expected since other parasitic reactions can influence the anodic currents in the OER region as discussed earlier. MWCNT-Graphene still shows the highest electrocatalytic activity followed by MWCNT-N-doped graphene and Pt, obtaining a Tafel slope and exchange current density of 123 mV dec-1 and 2.1 µA cm-2, respectively (Figure 6.6-B and Table 6.2). Similar exchange current densities calculated from the ORR and OER parts of the voltammograms in the presence of mixed oxides corroborate the discussion earlier on high current efficiencies for both ORR and OER in respective potential regions, especially in high anodic potentials where parasitic reactions are probable (Table 6.2). The synergistic effect of mixing MWCNT with either graphene or N-doped graphene is further confirmed with order of magnitude higher exchange current densities for the mixture comparing to each lone support, e.g. the exchange current densities for MWCNT:Graphene and MWCNT:N-doped graphene supported catalyst layers are 16 and 1000 times, respectively, the ones for the samples with the respective individual graphene-based material as catalyst support (Table 6.2).     195    Figure 6.6 IR-corrected bifunctional ORR/OER Tafel-lines of investigated MnO2-LaCoO3 catalyst supported on various carbonaceous materials studied in Figure 6.5 (50 wt% Pt/Graphitized carbon as the baseline): A) ORR, B) OER. The numbers associated with each line represent the respective apparent Tafel slopes. Cycle number five is reported in all cases. The N2-baseline is subtracted from the ORR part for all samples. For the OER part, capacitive background current at open circuit potential is removed from each corresponding voltammogram. Other conditions idem to Figure 6.5. 196  Table 6.2 The apparent exchange current densities and Tafel slopes for the initial stage ORR and OER activities of the investigated catalyst materials in Figure 6.5. The exchange current densities are expressed per geometric area. The apparent Tafel slope and exchange current density values are obtained over a potential range of min. 50 mV on six replicates. Other conditions idem to Figure 6.5. Catalyst bORR  (mV dec-1) bOER (mV dec-1) i0 (µA cm-2) MnO2-LaCoO3-Graphene -86 ±2 127 ±3 0.13 ±0.003 MnO2-LaCoO3-N-doped graphene -52 ±1 83 ±2 3.4×10-4  ±0.07×10-4 MnO2-LaCoO3-MWCNT -82 ±2 120 ±2 0.62 ±0.01 MnO2-LaCoO3-Vulcan XC-72 -72 ±1 125 ±3 0.61 ±0.01 MnO2-LaCoO3-MWCNT:Graphene -84 ±2 123 ±2 2.1 ±0.04 MnO2-LaCoO3-MWCNT-N-doped graphene -73 ±1 106 ±2 0.33 ±0.007 50 wt% Pt/Graphitized carbon -67 ±1 127 ±3 0.28 ±0.006 197  6.2.5 Bifunctional accelerated degradation testing: MnO2-LaCoO3 supported on MWCNT-Graphene or Vulcan XC-72   Figure 6.7 shows the ORR/OER bifunctional durability performance of GDEs with MnO2-LaCoO3 and two carbonaceous materials studied here, i.e. MWCNT-Graphene and Vulcan XC-72, during one hundred continuous potential cycling. MWCNT-Graphene containing catalyst layer was chosen as the best performing ORR/OER bifunctional catalyst investigated here while the Vulcan XC-72 was chosen as the baseline, mainly due to being the conventional support for oxygen catalysts. MWCNT-Graphene GDE provided superior ORR and OER electrocatalytic activity comparing to Vulcan XC-72 throughout the whole potential cycling tests. The ORR and OER performance losses of MWCNT-Graphene GDE was found to be about half of what Vulcan XC-72 GDE revealed during the first fifty cycles of accelerated degradation tests (Figure 6.7). Overall, MnO2-LaCoO3-MWCNT-Graphene GDE seems to provide better ORR/OER durability compared to Vulcan XC-72 supported GDE. It is well known in the literature that the MnO2 witnesses extensive changes in its Mn valence and crystallographic structure during potential cycling between ORR and OER regions, resulting in diverse Mn containing phases with different activity and durability for both ORR and OER [4, 5, 39, 45, 46, 67, 90, 98, 113, 227]. In section 4.2.1.2, we have shown using EELS analysis that the MnO2 in MnO2-LaCoO3-Vulcan XC-72 GDE was not regenerated efficiently in the same conditions as here during extensive potential cycling between ORR and OER regions, forming MnO, Mn(OH)2 and mostly Mn3O4 after one hundred cycles with low electrocatalytic activity for both ORR and OER comparing to the original manganese oxide [4, 5, 40, 45, 246]. Hence, one can argue that the loss in the ORR and OER performance of both GDEs is partly associated with the phase changes in the manganese oxide content of the catalyst layer to less active forms of 198  MnOx for both ORR and OER. Moreover, other phenomena such as carbon corrosion and MnO2 electro-corrosion (eq. 39) in high anodic potentials (as discussed in section 6.2.4) as well as catalyst layer detachment from the carbon substrate, no matter how small, could have a significant impact on the long-term bifunctional performance of these GDEs by affecting the surface interactions between the different components of the catalyst (specially catalyst and support), including lattice oxygen vacancy concentrations and HOO(ads)/HO(ads) binding energies as major contributors to ORR/OER overpotentials of oxide surfaces.  The role of lanthanum cobalt oxide in the performance loss of the mixed-oxide catalyst during accelerated degradation test is more complex than manganese oxide (Figure 6.7). The reason behind this is mainly due to the diversity of proposed reaction mechanisms for oxygen electrocatalysis on perovskite surfaces with no reliable method to experimentally detect the reaction intermediates and confirm what exactly is happening on these surfaces during ORR and OER [122, 123]. In the OER potential region, La-based perovskites, such as LaMnO3 and LaCoO3, are reported to be structurally stable during severe potential cycling according to a study using High-Resolution TEM and Fast Fourier Transform analysis by May et al. [257]. Moreover, the sluggish ORR performance of LaCoO3, comparing to manganese oxides, was previously reported in this work and the literature [4, 5, 123, 220]. Hence, one can conclude that the LaCoO3 itself has a low impact on the performance loss of mixed-oxide GDEs (e.g. MnO2-LaCoO3) during accelerated degradation tests. The real question here goes back to the surface interactions of LaCoO3 with other components of the catalyst layer, such as MnO2 and carbons, during severe potential cycling. Slight changes to these shared surfaces by either carbon corrosion or formation of passivating oxide layers during electro-corrosion of MnOx could have a drastic impact on the electrical conductivity of LaCoO3:Carbon or MnO2:LaCoO3 shared interfaces, effectively 199  insulating the electron paths of perovskite’s particles or manipulating the HOO(ads)/HO(ads) binding energies and therefore, lowering the overall ORR/OER electrocatalytic activity of the catalyzed GDEs. Further in-situ characterization is needed to examine the nature of each catalyst/support interfaces during ORR and OER to be able to fully understand the mechanism behind the performance losses of GDEs during the applied accelerated degradation tests.     Figure 6.7 Electrocatalytic durability testing: ORR (bottom) and OER (top) current densities at 800 mV and 1750 mV, respectively, during one hundred successive potential cycling between 673 to 1873 mV on MnO2-LaCoO3-MWCNT-Graphene (weight ratio 1:1:1:1) and MnO2-LaCoO3-Vulcan XC-72 (weight ratio 1:1:1) GDEs in flooded test setup. The oxide and carbon loadings were kept at 0.5 mg cm-2, each, spayed on a 40 wt% PTFE treated carbon cloth as porous substrate. The error associated with ORR and OER current densities was found to be between 2% and 10%. Other conditions idem to Figure 6.3.  200  6.3 Conclusion A systematic study has been performed to comparatively investigate the electrocatalytic activity and long-term durability of six different carbonaceous materials, i.e. Vulcan XC-72, multi-walled carbon nanotubes (MWCNT), graphene, N-doped graphene, MWCNT mixed with graphene or N-doped graphene, either as a support for a highly active ORR/OER bifunctional non-PGM oxide catalyst, i.e. MnO2-LaCoO3, or a lone oxygen electrocatalyst in alkaline media.  The RRDE experiments on the carbonaceous materials revealed a distinguished reduction peak at low rotation speeds (below 1000 rpm) and between 670-750 mV, mainly due to the electro-reduction of adsorbed oxygen on the catalyst surfaces, following a 4e− pathway. Vulcan XC-72 possessed the lowest ORR onset potential and highest ORR limiting current density followed by MWCNT, graphene and lastly, N-doped graphene. While Vulcan mostly catalyzed the bulk oxygen reduction through a 2e− pathway leading to hydrogen peroxide ions (except for the potential region around adsorbed oxygen reduction peak), MWCNT and graphene went through 2.8e− and 3e−, respectively, for the reduction of bulk oxygen indicating a competition between the 2e− and 4e− pathways. N-doped graphene, however, favors 4e− pathway with lowest ORR performance due to the flat morphology and small ECSA for the same loadings of Vulcan XC-72.  Next, the initial-stage bifunctional electrocatalytic activities of the carbons for both ORR and OER were investigated. Overall, Vulcan XC-72 showed highest current densities for both ORR and OER (up to 1650 mV). Significant synergistic effect was observed in the OER region when either graphene or N-doped graphene was mixed with MWCNT. This synergistic effect could be attributed to the unflattened morphology of the carbonaceous mixture by the introduction of MWCNT to it. Unusual values for kinetic parameters such as high OER Tafel slopes and orders 201  of magnitude higher OER exchange current densities compared to the ORR ones were attributed mainly to the carbon corrosion, interfering with the OER currents at high anodic potentials. The performance of carbonaceous materials as catalyst support for MnO2-LaCoO3 (weight ratio 1:1), was also investigated and compared to a 50 wt% Pt supported on graphitized carbon. The addition of MnO2-LaCoO3 to the carbons significantly increased their ORR and OER electrocatalytic activity. With regards to the ORR, the order of electrocatalytic activity for carbon(s) supported mixed-oxide catalysts vs. Pt is as follow: MWCNT-Graphene > Vulcan XC-72 > Pt > MWCNT-N-doped graphene > MWCNT > Graphene > N-doped graphene. The OER electrocatalytic activity of oxide containing samples decreases in a different sequence: MWCNT-Graphene > MWCNT-N-doped graphene > MWCNT > Vulcan XC-72 > Pt > Graphene > N-doped graphene. Highest ORR and OER mass activities of -6.694 A g-1 at 850 mV and 15.536 A g-1 at 1650 mV were achieved for MnO2-LaCoO3-MWCNT-Graphene compared to Pt sample.  Diverse ORR/OER electrocatalytic activities shown for mixed-oxide catalyst supported on wide range of carbonaceous materials studied here were explained by the difference in surface characteristics of each type of carbons. These surface characteristics such as morphology, active sites, defects, etc., can affect the nature of catalyst/support interactions at their interfaces. This provides various binding sites for the HOO(ads) and HO(ads) intermediates, leading to wide range of ORR and OER electrocatalytic activity for each combination. The effects of carbon corrosion and electro-reduction of MnOx on the OER electrocatalytic activity of MnO2-LaCoO3-Carbon(s) catalyst layers were also discussed. It was concluded that the currents associated with the two aforementioned processes were negligible comparing to OER in the potential range studied here.  202  The main factors affecting the ORR/OER durability of mixed-oxide GDEs were found to be: A) MnOx phase changes during ORR and OER, B) Carbon corrosion, C) MnOx electro-corrosion, and D) Catalyst layer detachment from the carbon substrate. The role of lanthanum cobalt oxide in the overall performance loss of the mixed-oxide catalyst during accelerated degradation test was concluded to be insignificant due to its structural stability in the OER region and lack of activity in the ORR region. 203  Chapter 7: Conclusions and recommendations for future work  7.1 Conclusions The primary goal of this study is to design electrochemically active and durable ORR/OER bifunctional non-PGM MnOx-based catalysts. To accomplish such a task, several steps including physical incorporation of active non-PGM co-catalysts, i.e. perovskite and fluorite-type oxides, microstructural modifications, i.e. surfactant-assisted electrodeposition of nanostructured manganese oxides, surface modification methods, i.e. alkali-metal ion insertion into the oxide structure and addition of carbon contents to the oxide catalyst layer were taken. In-depth structural characterizations and electrochemical measurements coupled with theoretical studies were further employed to carefully investigate the mechanisms by which the aforementioned methods affect the bifunctional electrocatalytic activity and durability of “designed” catalysts. 7.1.1 ORR/OER electrocatalytic activity and durability of individual and mixed-oxide catalysts A positive synergistic effect on the ORR/OER electrocatalytic activity of the physically-mixed oxide catalysts, i.e.  MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7, was found. This was shown by either a decrease in the apparent Tafel slope or an increase in the apparent exchange current density for the mixed-oxide formulations compared to the respective individual oxide catalysts. The mechanism for the mixed oxides’ synergistic electrocatalytic effect could be rationalized in terms of the scaling relationship between HOO(ads) and HO(ads) binding energies. The structurally diverse oxide combinations provide different binding energies for the key intermediates, thus, “breaking” away from the linear scaling relationship. 204  The MnO2-Nd3IrO7 GDE revealed the highest ORR activity and long-term stability followed by MnO2-LaCoO3 as the second best, both outperforming the commercial Pt and commercial MnOx GDEs with up to 100 mV (air) and 150 mV (oxygen) more positive ORR overpotentials during 24 hrs of galvanostatic polarizations in 11.7 M KOH at 323 K and Pgas of 1 atm. The structural analysis on the MnO2-based GDEs during ORR galvanostatic polarization tests showed that the gradual transformation of MnO2 to less active forms of manganese species (i.e. Mn3+/Mn2+) during ORR could attribute to the ORR performance degradation of mixed-oxide catalysts. 7.1.2 Oxide catalyst activation by alkali-metal ion intercalation  Potassium ion insertion in the catalyst structure of fresh oxides, i.e. MnO2, LaCoO3, Nd3IrO7, MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7, by A) Longer-term exposure to 6 M KOH and B) Potential driven (electrophoretic) intercalation of potassium, was found to be effective for enhancing the bifunctional activity and durability of the oxide catalysts, e.g. OER (at 5 mA cm-2 or 5 A gcatalyst-1) and ORR (at -2 mA cm-2 or -2 A gcatalyst-1) overpotentials are enhanced by 110 and 75 mV, respectively, for MnO2-LaCoO3 subjected to potential driven potassium ion intercalation or the rates of OER and ORR potential increase, a measure of electrocatalytic activity degradation, under similar conditions are diminished by 60.5 and 24 mV h-1, respectively, for MnO2-Nd3IrO7 after K+ potential driven intercalation activation. It is proposed that the reason for the enhanced ORR/OER performances of the activated catalysts is the uptake of K+ into the catalyst layer (mostly in the vacancies and defects of the oxide crystal structures such as Schottky and lattice oxygen vacancies), acting as a promoter for both ORR and OER and providing new binding sites with distinct binding energies for HOO(ads) and 205  HO(ads) intermediates, help breaking away from the universal scaling relationship and thus, enhancing the ORR/OER bifunctional activity of the investigated catalysts. 7.1.3 Surfactant-assisted electrodeposition of Mn oxides: Factorial design study of the electrodeposition factors For all surfactant cases studied here for the anodic electrodeposition of manganese oxides, high concentration of surfactants provided samples with the best ORR/OER bifunctional performances, based on all three specified responses of the factorial design study. Moreover, low temperature was reported to lead to high ORR/OER mass activities for all surfactant choices. While the effect of applied anodic potential on the ORR and OER activities of electrodeposited MnOx was found to be negligible in the case of SDS and CTAB surfactants, Mn concentration seemed to be an insignificant player for the Triton X-100 samples.  The electrodeposited MnOx for Triton run no. 9 was found to show the best ORR and OER electrocatalytic activities among other electrodeposited manganese oxides investigated here. Compared to Ir, Ru and IrO2, it showed lower OER overpotential (min. 100 mV) at 2 mA cm-2. For the ORR, the manganese oxide electrodeposited during Triton run no. 9 provided between 50 and 150 mV more positive overpotential (at -2 mA cm-2) compared to the other non-precious metal compounds such as CoMn2O4 and Core-Corona Structured Bifunctional Catalyst (CCBC). The galvanostatic polarization tests further confirmed the promising OER activity for the Triton run no. 9 sample with potentials as low as 1446 mV (at 5 mA cm-2 and t=2 hrs), about 40 mV lower than the commercial MnOx, and a degradation rate of 43 mV h-1, about 10 mV h-1 lower than its commercial counterparts. The surface modifications of MnOx via surfactant-assisted electrodeposition can significantly alter the morphology and Mn valence in the deposited materials, provide new binding sites for the 206  HOO(ads) and HO(ads) intermediates, help break away from the linear scaling relationship between their binding energies as a major contributor to the ORR and OER overpotentials and hence, enhance the ORR and OER electrocatalytic activity of electrodeposited manganese oxides. The formation of hydrogen-bonded complexes, i.e. HO(ads)…H-OH, with specially configured water molecules called “activated water”, can further explain the enhancement in the ORR activity of the catalysts, mainly by facile transfer of protons to weakly adsorbed HOO(ads)/O(ads) intermediates. This, however, depends on the surface coverage of OH(ads) which provides binding sites for formation of HO(ads)…H-OH complexes (promoter effect). 7.1.4 The effect of carbon supports: Graphene vs. commercial carbon materials The RRDE experiments on the carbonaceous materials revealed that Vulcan mostly catalyzes the bulk oxygen reduction through a 2e− pathway leading to hydrogen peroxide ions while MWCNT and graphene goes through 2.8e− and 3e−, respectively, indicating a competition between the 2e− and 4e− pathways. N-doped graphene, however, favors 4e− pathway. Unusual values for kinetic parameters such as high OER Tafel slopes and orders of magnitude higher OER exchange current densities compared to the ORR ones were attributed mainly to the carbon corrosion, interfering with the OER currents at high anodic potentials. Significant synergistic effect was observed in the ORR and OER region when either graphene or N-doped graphene was mixed with MWCNT. This synergistic effect could be attributed to the unflattened morphology of the carbonaceous mixture by the introduction of MWCNT to it, providing new binding sites for the HOO(ads)/HO(ads) intermediates. The orders of ORR and OER electrocatalytic activities for carbon(s) supported mixed-oxide catalysts vs. Pt are as follows:  207  ORR: MWCNT-Graphene > Vulcan XC-72 > Pt > MWCNT-N-doped graphene > MWCNT > Graphene > N-doped graphene.  OER: MWCNT-Graphene > MWCNT-N-doped graphene > MWCNT > Vulcan XC-72 > Pt > Graphene > N-doped graphene. Diverse ORR/OER electrocatalytic activities shown for the mixed-oxide catalyst supported on the carbonaceous materials studied were explained by the difference in surface characteristics of each type of carbon. These surface characteristics such as morphology, active sites, defects, etc., can affect the nature of catalyst/support interactions at their interfaces. This provides various binding sites for the HOO(ads) and HO(ads) intermediates, leading to wide range of ORR and OER electrocatalytic activity for each combination. The main factors affecting the ORR/OER durability of mixed-oxide GDEs were found to be: A) MnOx phase changes during ORR and OER, B) Carbon corrosion, C) MnOx electro-corrosion, and D) Catalyst layer detachment from the carbon substrate. The role of lanthanum cobalt oxide in the overall performance loss of the mixed-oxide catalyst during accelerated degradation test was concluded to be insignificant due to its structural stability in the OER region and lack of activity in the ORR region. 7.2 Contributions to knowledge Major novelties and contributions to science for this study are mentioned as follows: 1) The synergistic effect for incorporation of perovskite and fluorite-type oxides to manganese dioxide on the MnOx activity for oxygen electrocatalysis has been reported for the first time. The study paves the way toward fundamental understanding of oxide-oxide interactions in catalyst layers which could significantly help “design” an active and durable non-PGM bi-208  material catalyst for both ORR and OER to be used in wide range of energy generation and storage applications such as alkaline fuel cells and rechargeable metal-air batteries. 2) The effect of alkali-metal ions, i.e. Li+, Na+, K+ and Cs+, on the bifunctional performance of individual and mixed-oxide catalysts has been investigated and announced for the first time in this work. Novel potassium insertion methods such as open-circuit potential (OCP) and potential driven intercalation (PDI) were first introduced here which can “heal” the bifunctional performance of degraded catalysts or significantly enhance both initial stage activity and long-term durability of fresh oxide catalysts. The potassium activation methods, specially PDI, present a time-and-cost-effective way to enhance the activity and life-time durability of non-PGM oxide catalysts to replace the precious metal catalysts currently being used in industrial applications. 3) The comprehensive investigation on surfactant-assisted electrodeposition of MnOx using factorial design analysis provided a unique and comprehensive study on main electrodeposition factors affecting the final bifunctional activity of deposited manganese oxides.  The study helps understand the important role of two-factor interactions during the electrodeposition process, something that is usually neglected in the literature. The most active manganese oxide electrodeposited in presence of Triton X-100 provided one of the lowest OER overpotentials and highest OER stability ever reported for MnOx in the literature. This opens a new path for manganese oxides as active OER electrocatalysts in applications like water electrolyzers. 4) The systematic study on the carbon support/catalyst interactions for a non-PGM catalyst, i.e. MnO2-LaCoO3, provides better insights on the mechanisms by which the catalyst layer is degraded during extensive potential cycling and helped find optimum catalyst composition for better and more durable non-PGM ORR/OER bifunctional catalyst.  209  5) The novel utilization of EELS structural characterization method on the non-PGM catalyst layer during extensive potential cycling between ORR and OER regions provided an excellent experimental evidence for structural changes associated with manganese dioxide as well as corresponding Mn valence at each stage. This is specially helpful to understand the limits of developed MnOx-based catalyst and “design” them accordingly for commercial applications. 7.3 Recommendation for future work The following recommendations are proposed for future work based on the findings of this study: 1) Experimental evidence corroborated possibly by theoretical calculations of the binding energies of ORR and OER intermediate species on MnO2, LaCoO3 and Nd3irO7 with and without lattice distortions is required to validate the proposed hypothesis for enhanced bifunctional activity of oxides with induced potassium ions. 2) DFT calculations seems necessary to calculate the formation energy as well as concentration of lattice oxygen vacancies on the investigated catalysts. 3) Further experimentations and 3D distribution modeling is needed to better understand the possible differences between the potassium intercalation in each of two activation methods. 4) A study on the effect of alkali-ion insertion on surfactant-assisted deposited MnOx could translate to further enhancement in the ORR/OER electrocatalytic activity and durability of deposited manganese oxides. Simultaneous electrodeposition of Mn oxides and potassium activation should be considered to simplify the process. 5) Large-scale performance testing of catalyst materials in commercial scenarios for tandem ORR and OER should be performed. 210  6) The application of investigated catalysts and developed activation methods in aqueous solutions needs to be further studied in non-aqueous battery electrolytes such as metal induced room temperature ionic liquids (RTIL), e.g. [C2mim][dca]+Zn(dca)2 or gel-like [P6,6,6,14][Cl].2(MgCl2).37H2O, to seek the possibility of an active and durable rechargeable non-aqueous metal-air battery catalyzed by MnOx-based cathode catalyst. Room temperature ionic liquids have gained much attention lately to be used as the battery electrolyte due to their safety (non-volatile), high conductivity and support of some metals’ electrochemistry such as Zn and Mg [278].211  Bibliography [1] J. Lee, B. Jeong, J.D. Ocon, Oxygen electrocatalysis in chemical energy conversion and storage technologies, Current Applied Physics, 13 (2013) 309-321. [2] J. Ludwig, Bifunctional oxygen/air electrodes, J. Power Sources, 155 (2006) 23-32. [3] Y. Kiros, M. Pirjamali, M. Bursell, Oxygen reduction electrodes for electrolysis in chlor-alkali cells, Electrochimica Acta, 51 (2006) 3346-3350. [4] P.H. Benhangi, A. Alfantazi, E. Gyenge, Manganese Dioxide-based Bifunctional Oxygen Reduction/Evolution Electrocatalysts: Effect of Perovskite Doping and Potassium Ion Insertion, Electrochimica Acta, 123 (2014) 42-50. [5] P. Hosseini-Benhangi, M.A. Garcia-Contreras, A. Alfantazi, E.L. Gyenge, Method for Enhancing the Bifunctional Activity and Durability of Oxygen Electrodes with Mixed Oxide Electrocatalysts: Potential Driven Intercalation of Potassium, Journal of The Electrochemical Society, 162 (2015) F1356-F1366. [6] A. Serov, A. Aziznia, P.H. Benhangi, K. Artyushkova, P. Atanassov, E. Gyenge, Borohydride-tolerant oxygen electroreduction catalyst for mixed-reactant Swiss-roll direct borohydride fuel cells, Journal of Materials Chemistry A, 1 (2013) 14384-14391. [7] V. Nikolova, P. Iliev, K. Petrov, T. Vitanov, E. Zhecheva, R. Stoyanova, I. Valov, D. Stoychev, Electrocatalysts for bifunctional oxygen/air electrodes, J. Power Sources, 185 (2008) 727-733. [8] B. Sajdl, K. Micka, P. Krtil, Study of the rechargeable manganese dioxide electrode, Electrochim. Acta, 40 (1995) 2005-2011. [9] Y. Li, M. Gong, Y. Liang, J. Feng, J.-E. Kim, H. Wang, G. Hong, B. Zhang, H. Dai, Advanced zinc-air batteries based on high-performance hybrid electrocatalysts, Nat Commun, 4 (2013) 1805. [10] R. Cao, J.-S. Lee, M. Liu, J. Cho, Recent Progress in Non-Precious Catalysts for Metal-Air Batteries, Advanced Energy Materials, 2 (2012) 816-829. [11] J.-S. Lee, S. Tai Kim, R. Cao, N.-S. Choi, M. Liu, K.T. Lee, J. Cho, Metal–Air Batteries with High Energy Density: Li–Air versus Zn–Air, Advanced Energy Materials, 1 (2011) 34-50. [12] D. Higgins, Z. Chen, D.U. Lee, Z. Chen, Activated and nitrogen-doped exfoliated graphene as air electrodes for metal-air battery applications, Journal of Materials Chemistry A, 1 (2013) 2639-2645. [13] F. Cheng, J. Chen, Metal-air batteries: from oxygen reduction electrochemistry to cathode catalysts, Chemical Society Reviews, 41 (2012) 2172-2192. [14] K.F. Blurton, A.F. Sammells, Metal/air batteries: Their status and potential — a review, Journal of Power Sources, 4 (1979) 263-279. [15] T. Zhang, Z. Tao, J. Chen, Magnesium-air batteries: from principle to application, Materials Horizons, 1 (2014) 196-206. [16] D.U. Lee, P. Xu, Z.P. Cano, A.G. Kashkooli, M.G. Park, Z. Chen, Recent progress and perspectives on bi-functional oxygen electrocatalysts for advanced rechargeable metal-air batteries, Journal of Materials Chemistry A, 4 (2016) 7107-7134. [17] F. Cheng, J. Chen, Metal-air batteries: from oxygen reduction electrochemistry to cathode catalysts, Chem. Soc. Rev., 41 (2012) 2172-2192. [18] J. Desilvestro, O. Haas, Metal Oxide Cathode Materials for Electrochemical Energy Storage: A Review, Journal of The Electrochemical Society, 137 (1990) 5C-22C. 212  [19] A. R. Mainar, O. Leonet, M. Bengoechea, I. Boyano, I. de Meatza, A. Kvasha, A. Guerfi, J. Alberto Blázquez, Alkaline aqueous electrolytes for secondary zinc–air batteries: an overview, International Journal of Energy Research, 40 (2016) 1032-1049. [20] S. Visco, Y. Nimon, Li/air non-aqueous batteries, Google Patents, 2007. [21] P.G. Bruce, S.A. Freunberger, L.J. Hardwick, J.-M. Tarascon, Li-O2 and Li-S batteries with high energy storage, Nat Mater, 11 (2012) 19-29. [22] M.A. Rahman, X. Wang, C. Wen, High Energy Density Metal-Air Batteries: A Review, Journal of The Electrochemical Society, 160 (2013) A1759-A1771. [23] J. Lu, L. Li, J.-B. Park, Y.-K. Sun, F. Wu, K. Amine, Aprotic and Aqueous Li–O2 Batteries, Chemical Reviews, 114 (2014) 5611-5640. [24] E. Yoo, H. Zhou, Li−Air Rechargeable Battery Based on Metal-free Graphene Nanosheet Catalysts, ACS Nano, 5 (2011) 3020-3026. [25] E. Hollax, Advanced electrochemical energy sources for space power systems — a review, Journal of Power Sources, 4 (1979) 11-19. [26] S. Mekhilef, R. Saidur, A. Safari, Comparative study of different fuel cell technologies, Renewable and Sustainable Energy Reviews, 16 (2012) 981-989. [27] A. Aziznia, C.W. Oloman, E.L. Gyenge, Platinum- and Membrane-Free Swiss-Roll Mixed-Reactant Alkaline Fuel Cell, ChemSusChem, 6 (2013) 847-855. [28] A. Aziznia, C.W. Oloman, E.L. Gyenge, A Swiss-roll liquid–gas mixed-reactant fuel cell, Journal of Power Sources, 212 (2012) 154-160. [29] J. Pettersson, B. Ramsey, D. Harrison, A review of the latest developments in electrodes for unitised regenerative polymer electrolyte fuel cells, Journal of Power Sources, 157 (2006) 28-34. [30] W. Smith, The role of fuel cells in energy storage, Journal of Power Sources, 86 (2000) 74-83. [31] T. Ioroi, N. Kitazawa, K. Yasuda, Y. Yamamoto, H. Takenaka, IrO2-deposited Pt electrocatalysts for unitized regenerative polymer electrolyte fuel cells, J Appl Electrochem, 31 (2001) 1179-1183. [32] J.W. Desmond Ng, Y. Gorlin, T. Hatsukade, T.F. Jaramillo, A Precious-Metal-Free Regenerative Fuel Cell for Storing Renewable Electricity, Advanced Energy Materials, 3 (2013) 1545-1550. [33] F. Barbir, T. Molter, L. Dalton, Efficiency and weight trade-off analysis of regenerative fuel cells as energy storage for aerospace applications, International Journal of Hydrogen Energy, 30 (2005) 351-357. [34] L. Swette, N. Kackley, S.A. McCatty, Oxygen electrodes for rechargeable alkaline fuel cells. III, Journal of Power Sources, 36 (1991) 323-339. [35] S.K. Tiwari, J.F. Koenig, G. Poillerat, P. Chartier, R.N. Singh, Electrocatalysis of oxygen evolution/reductionon LaNiO3 prepared by a novel malic acid-aided method, J Appl Electrochem, 28 (1998) 114-119. [36] L. Swette, J. Giner, Oxygen electrodes for rechargeable alkaline fuel cells, Journal of Power Sources, 22 (1988) 399-408. [37] N. Ominde, N. Bartlett, X.-Q. Yang, D. Qu, Investigation of the oxygen reduction reaction on the carbon electrodes loaded with MnO2 catalyst, Journal of Power Sources, 195 (2010) 3984-3989. [38] Y. Shao, J. Liu, Y. Wang, Y. Lin, Novel catalyst support materials for PEM fuel cells: current status and future prospects, Journal of Materials Chemistry, 19 (2009) 46-59. 213  [39] Y. Chabre, J. Pannetier, Structural and electrochemical properties of the proton / γ-MnO2 system, Prog. Solid State Ch., 23 (1995) 1-130. [40] J. McBreen, The electrochemistry of β-MnO2 and γ-MnO2 in alkaline electrolyte, Electrochimica Acta, 20 (1975) 221-225. [41] C. Poinsignon, J.M. Amarilla, F. Tedjar, Electrochemical reduction of β-MnO2, ramsdellite, γ- and ε-MnO2, Solid State Ionics, 70–71, Part 1 (1994) 649-653. [42] K. Kordesch, J. Gsellmann, M. Peri, K. Tomantschger, R. Chemelli, The rechargeability of manganese dioxide in alkaline electrolyte, Electrochimica Acta, 26 (1981) 1495-1504. [43] M. Kloβ, D. Rahner, W. Plieth, The effect of alkaline earth titanates on the rechargeability of manganese dioxide in alkaline electrolyte, Journal of Power Sources, 69 (1997) 137-143. [44] P. Żółtowski, D.M. Dražić, L. Vorkapić, Carbon-air electrode with regenerative short time overload capacity: Part 1. Effect of manganese dioxide, J Appl Electrochem, 3 (1973) 271-283. [45] F.H.B. Lima, M.L. Calegaro, E.A. Ticianelli, Electrocatalytic activity of manganese oxides prepared by thermal decomposition for oxygen reduction, Electrochimica Acta, 52 (2007) 3732-3738. [46] P. Ruetschi, Discharge Mechanism of MnO2 Electrodes as Influenced by the Solubility of the Reaction Products, Journal of The Electrochemical Society, 123 (1976) 495-500. [47] A. Aziznia, C.W. Oloman, E.L. Gyenge, A Swiss-roll liquid–gas mixed-reactant fuel cell, J. Power Sources, 212 (2012) 154-160. [48] Y.L. Cao, H.X. Yang, X.P. Ai, L.F. Xiao, The mechanism of oxygen reduction on MnO2-catalyzed air cathode in alkaline solution, Journal of Electroanalytical Chemistry, 557 (2003) 127-134. [49] F.H.B. Lima, M.L. Calegaro, E.A. Ticianelli, Investigations of the catalytic properties of manganese oxides for the oxygen reduction reaction in alkaline media, Journal of Electroanalytical Chemistry, 590 (2006) 152-160. [50] I. Roche, K. Scott, Effect of pH and temperature on carbon-supported manganese oxide oxygen reduction electrocatalysts, Journal of Electroanalytical Chemistry, 638 (2010) 280-286. [51] K.E. Sickafus, J.M. Wills, N.W. Grimes, Structure of Spinel, Journal of the American Ceramic Society, 82 (1999) 3279-3292. [52] P. Rasiyah, A.C.C. Tseung, A Mechanistic Study of Oxygen Evolution on NiCo2 O 4 : II . Electrochemical Kinetics, Journal of The Electrochemical Society, 130 (1983) 2384-2386. [53] A. Daghetti, G. Lodi, S. Trasatti, Interfacial properties of oxides used as anodes in the electrochemical technology, Materials Chemistry and Physics, 8 (1983) 1-90. [54] B.B. Blizanac, P.N. Ross, N.M. Marković, Oxygen Reduction on Silver Low-Index Single-Crystal Surfaces in Alkaline Solution:  Rotating Ring DiskAg(hkl) Studies, The Journal of Physical Chemistry B, 110 (2006) 4735-4741. [55] J.N. Tiwari, K. Nath, S. Kumar, R.N. Tiwari, K.C. Kemp, N.H. Le, D.H. Youn, J.S. Lee, K.S. Kim, Stable platinum nanoclusters on genomic DNA–graphene oxide with a high oxygen reduction reaction activity, Nat Commun, 4 (2013) 3221-3221. [56] Y.-C. Lu, Z. Xu, H.A. Gasteiger, S. Chen, K. Hamad-Schifferli, Y. Shao-Horn, Platinum−Gold Nanoparticles: A Highly Active Bifunctional Electrocatalyst for Rechargeable Lithium−Air Batteries, Journal of the American Chemical Society, 132 (2010) 12170-12171. [57] B.C. Beard, P.N. Ross, The Structure and Activity of Pt‐Co Alloys as Oxygen Reduction Electrocatalysts, Journal of The Electrochemical Society, 137 (1990) 3368-3374. 214  [58] J.O.M. Bockris, T. Otagawa, The Electrocatalysis of Oxygen Evolution on Perovskites, Journal of The Electrochemical Society, 131 (1984) 290-302. [59] J.O.M. Bockris, T. Otagawa, V. Young, Solid state surface studies of the electrocatalysis of oxygen evolution on perovskites, J. Electroanal. Chem., 150 (1983) 633-643. [60] M. Bursell, M. Pirjamali, Y. Kiros, La0.6Ca0.4CoO3, La0.1Ca0.9MnO3 and LaNiO3 as bifunctional oxygen electrodes, Electrochim. Acta, 47 (2002) 1651-1660. [61] C.K. Lee, K.A. Striebel, F.R. McLarnon, E.J. Cairns, Thermal Treatment of La0.6Ca0.4CoO3 Perovskites for Bifunctional Air Electrodes, J. Electrochem. Soc., 144 (1997) 3801-3806. [62] J. Suntivich, K.J. May, H.A. Gasteiger, J.B. Goodenough, Y. Shao-Horn, A Perovskite Oxide Optimized for Oxygen Evolution Catalysis from Molecular Orbital Principles, Science, 334 (2011) 1383-1385. [63] S. Velraj, J.H. Zhu, Sm0.5Sr0.5CoO3−δ – A new bi-functional catalyst for rechargeable metal-air battery applications, Journal of Power Sources, 227 (2013) 48-52. [64] L. Wang, X. Zhao, Y. Lu, M. Xu, D. Zhang, R.S. Ruoff, K.J. Stevenson, J.B. Goodenough, CoMn2O4 Spinel Nanoparticles Grown on Graphene as Bifunctional Catalyst for Lithium-Air Batteries, Journal of The Electrochemical Society, 158 (2011) A1379-A1382. [65] N.H. Chou, P.N. Ross, A.T. Bell, T.D. Tilley, Comparison of Cobalt-based Nanoparticles as Electrocatalysts for Water Oxidation, ChemSusChem, 4 (2011) 1566-1569. [66] N.M. Marković, P.N. Ross, A Study of Bismuth Ruthenate as an Electrocatalyst for Bifunctional Air Electrodes, Journal of The Electrochemical Society, 141 (1994) 2590-2597. [67] K.L. Pickrahn, S.W. Park, Y. Gorlin, H.-B.-R. Lee, T.F. Jaramillo, S.F. Bent, Active MnOx Electrocatalysts Prepared by Atomic Layer Deposition for Oxygen Evolution and Oxygen Reduction Reactions, Adv. Energy Mater., 2 (2012) 1269-1277. [68] W.H. Lee, H. Kim, Oxidized iridium nanodendrites as catalysts for oxygen evolution reactions, Catalysis Communications, 12 (2011) 408-411. [69] Y. Gorlin, T.F. Jaramillo, A Bifunctional Nonprecious Metal Catalyst for Oxygen Reduction and Water Oxidation, Journal of the American Chemical Society, 132 (2010) 13612-13614. [70] Y. Lee, J. Suntivich, K.J. May, E.E. Perry, Y. Shao-Horn, Synthesis and Activities of Rutile IrO2 and RuO2 Nanoparticles for Oxygen Evolution in Acid and Alkaline Solutions, The Journal of Physical Chemistry Letters, 3 (2012) 399-404. [71] J. Sunarso, A.M. Glushenkov, A.A.J. Torriero, P.C. Howlett, Y. Chen, D.R. MacFarlane, M. Forsyth, Bi-Functional Water/Oxygen Electrocatalyst Based on PdO-RuO2 Composites, Journal of The Electrochemical Society, 160 (2013) H74-H79. [72] Z. Chen, A. Yu, R. Ahmed, H. Wang, H. Li, Z. Chen, Manganese dioxide nanotube and nitrogen-doped carbon nanotube based composite bifunctional catalyst for rechargeable zinc-air battery, Electrochimica Acta, 69 (2012) 295-300. [73] Y. Gorlin, C.-J. Chung, D. Nordlund, B.M. Clemens, T.F. Jaramillo, Mn3O4 Supported on Glassy Carbon: An Active Non-Precious Metal Catalyst for the Oxygen Reduction Reaction, ACS Catalysis, 2 (2012) 2687-2694. [74] S.-Y. Huang, P. Ganesan, H.-Y. Jung, B.N. Popov, Development of supported bifunctional oxygen electrocatalysts and corrosion-resistant gas diffusion layer for unitized regenerative fuel cell applications, Journal of Power Sources, 198 (2012) 23-29. [75] N. Sasikala, K. Ramya, K.S. Dhathathreyan, Bifunctional electrocatalyst for oxygen/air electrodes, Energy Conversion and Management, 77 (2014) 545-549. 215  [76] Z. Chen, A. Yu, D. Higgins, H. Li, H. Wang, Z. Chen, Highly Active and Durable Core–Corona Structured Bifunctional Catalyst for Rechargeable Metal–Air Battery Application, Nano Lett., 12 (2012) 1946-1952. [77] Y. Chabre, J. Pannetier, Structural and electrochemical properties of the proton / γ-MnO2 system, Progress in solid state chemistry, 23 (1995) 1-130. [78] X. Wang, Y. Li, Synthesis and Formation Mechanism of Manganese Dioxide Nanowires/Nanorods, Chemistry – A European Journal, 9 (2003) 300-306. [79] A. Perner, K. Holl, D. Ilic, M. Wohlfahrt-Mehrens, A New MnOx Cathode Material for Rechargeable Lithium Batteries, European Journal of Inorganic Chemistry, 2002 (2002) 1108-1114. [80] R. Chitrakar, H. Kanoh, Y.-S. Kim, Y. Miyai, K. Ooi, Synthesis of Layered-Type Hydrous Manganese Oxides from Monoclinic-Type LiMnO2, Journal of Solid State Chemistry, 160 (2001) 69-76. [81] B. Ammundsen, J. Paulsen, Novel Lithium-Ion Cathode Materials Based on Layered Manganese Oxides, Advanced Materials, 13 (2001) 943-956. [82] Y. Muraoka, H. Chiba, T. Atou, M. Kikuchi, K. Hiraga, Y. Syono, S. Sugiyama, S. Yamamoto, J.-C. Grenier, Preparation ofα-MnO2with an Open Tunnel, Journal of Solid State Chemistry, 144 (1999) 136-142. [83] S. Ching, D.J. Petrovay, M.L. Jorgensen, S.L. Suib, Sol−Gel Synthesis of Layered Birnessite-Type Manganese Oxides, Inorganic Chemistry, 36 (1997) 883-890. [84] D. Tench, L.F. Warren, Electrodeposition of Conducting Transition Metal Oxide/Hydroxide Films from Aqueous Solution, Journal of The Electrochemical Society, 130 (1983) 869-872. [85] S. Devaraj, N. Munichandraiah, Effect of Crystallographic Structure of MnO2 on its Electrochemical Capacitance Properties, The Journal of Physical Chemistry C, 112 (2008) 4406-4417. [86] C.J. Clarke, G.J. Browning, S.W. Donne, An RDE and RRDE study into the electrodeposition of manganese dioxide, Electrochimica Acta, 51 (2006) 5773-5784. [87] G.M. Jacob, I. Zhitomirsky, Microstructure and properties of manganese dioxide films prepared by electrodeposition, Applied Surface Science, 254 (2008) 6671-6676. [88] T. Brousse, M. Toupin, R. Dugas, L. Athouël, O. Crosnier, D. Bélanger, Crystalline MnO2 as Possible Alternatives to Amorphous Compounds in Electrochemical Supercapacitors, Journal of The Electrochemical Society, 153 (2006) A2171-A2180. [89] C. Julien, M. Massot, S. Rangan, M. Lemal, D. Guyomard, Study of structural defects in γ-MnO2 by Raman spectroscopy, Journal of Raman Spectroscopy, 33 (2002) 223-228. [90] E.L. Gyenge, J.-F. Drillet, The Electrochemical Behavior and Catalytic Activity for Oxygen Reduction of MnO2/C– Toray Gas Diffusion Electrodes, J. Electrochem. Soc., 159 (2012) F23-F34. [91] B. Šljukić, D.M.F. Santos, C.A.C. Sequeira, Manganese dioxide electrocatalysts for borohydride fuel cell cathodes?, Journal of Electroanalytical Chemistry, 694 (2013) 77-83. [92] A. Kozawa, R.A. Powers, The Manganese Dioxide Electrode in Alkaline Electrolyte; The Electron‐Proton Mechanism for the Discharge Process from MnO2   to MnO1.5, Journal of The Electrochemical Society, 113 (1966) 870-878. [93] D. Im, A. Manthiram, Role of Bismuth and Factors Influencing the Formation of Mn3O4 in Rechargeable Alkaline Batteries Based on Bismuth-Containing Manganese Oxides, Journal of The Electrochemical Society, 150 (2003) A68-A73. 216  [94] H.S. Wroblowa, N. Gupta, Rechargeable manganese oxide electrodes: Part II. physically modified materials, J. Electroanal. Chem., 238 (1987) 93-102. [95] M.A. Dzieciuch, N. Gupta, H.S. Wroblowa, Rechargeable Cells with Modified MnO2 Cathodes, Journal of The Electrochemical Society, 135 (1988) 2415-2418. [96] W. Jantscher, L. Binder, D.A. Fiedler, R. Andreaus, K. Kordesch, Synthesis, characterization and application of doped electrolytic manganese dioxides, Journal of Power Sources, 79 (1999) 9-18. [97] A. Stani, W. Taucher-Mautner, K. Kordesch, J. Daniel-Ivad, Development of flat plate rechargeable alkaline manganese dioxide–zinc cells, Journal of Power Sources, 153 (2006) 405-412. [98] F.H.B. Lima, M.L. Calegaro, E.A. Ticianelli, Investigations of the catalytic properties of manganese oxides for the oxygen reduction reaction in alkaline media, J. Electroanal. Chem., 590 (2006) 152-160. [99] W.L. Vielstich, Arnold; Gasteiger, Hubert A (Eds), Fuel Cell Electrocatalysis,  Handbook of fuel cells: fundamentals, technology, and applications, Wiley2003, pp. 443. [100] X. Ge, A. Sumboja, D. Wuu, T. An, B. Li, F.W.T. Goh, T.S.A. Hor, Y. Zong, Z. Liu, Oxygen Reduction in Alkaline Media: From Mechanisms to Recent Advances of Catalysts, ACS Catalysis, 5 (2015) 4643-4667. [101] P.A. Christensen, A. Hamnett, D. Linares-Moya, Oxygen reduction and fuel oxidation in alkaline solution, Physical Chemistry Chemical Physics, 13 (2011) 5206-5214. [102] C.F. Zinola, A.J. Arvia, G.L. Estiu, E.A. Castro, A Quantum Chemical Approach to the Influence of Platinum Surface Structure on the Oxygen Electroreduction Reaction, The Journal of Physical Chemistry, 98 (1994) 7566-7576. [103] C. Oloman, Electrochemical processing for the pulp and paper industry, The Electrochemical Consultancy, Romsey, Hants, UK, 1996. [104] Y. Gorlin, D. Nordlund, T.F. Jaramillo, The Role of Heat Treatment in Enhanced Activity of Manganese Oxides for the Oxygen Reduction and Evolution Reactions, ECS Transactions, 58 (2013) 735-750. [105] Y. Meng, W. Song, H. Huang, Z. Ren, S.-Y. Chen, S.L. Suib, Structure–Property Relationship of Bifunctional MnO2 Nanostructures: Highly Efficient, Ultra-Stable Electrochemical Water Oxidation and Oxygen Reduction Reaction Catalysts Identified in Alkaline Media, Journal of the American Chemical Society, 136 (2014) 11452-11464. [106] J. Suntivich, H.A. Gasteiger, N. Yabuuchi, Y. Shao-Horn, Electrocatalytic Measurement Methodology of Oxide Catalysts Using a Thin-Film Rotating Disk Electrode, Journal of The Electrochemical Society, 157 (2010) B1263-B1268. [107] K.A. Stoerzinger, M. Risch, B. Han, Y. Shao-Horn, Recent Insights into Manganese Oxides in Catalyzing Oxygen Reduction Kinetics, ACS Catalysis, 5 (2015) 6021-6031. [108] M.V. ten Kortenaar, J.F. Vente, D.J.W. Ijdo, S. Müller, R. Kötz, Oxygen evolution and reduction on iridium oxide compounds, Journal of Power Sources, 56 (1995) 51-60. [109] N.M. Marković, H.A. Gasteiger, P.N. Ross, Oxygen Reduction on Platinum Low-Index Single-Crystal Surfaces in Alkaline Solution:  Rotating Ring DiskPt(hkl) Studies, The Journal of Physical Chemistry, 100 (1996) 6715-6721. [110] I. Roche, E. Chaînet, M. Chatenet, J. Vondrák, Carbon-Supported Manganese Oxide Nanoparticles as Electrocatalysts for the Oxygen Reduction Reaction (ORR) in Alkaline Medium:  217  Physical Characterizations and ORR Mechanism, The Journal of Physical Chemistry C, 111 (2006) 1434-1443. [111] L. Mao, D. Zhang, T. Sotomura, K. Nakatsu, N. Koshiba, T. Ohsaka, Mechanistic study of the reduction of oxygen in air electrode with manganese oxides as electrocatalysts, Electrochimica Acta, 48 (2003) 1015-1021. [112] S.L. Brock, N. Duan, Z.R. Tian, O. Giraldo, H. Zhou, S.L. Suib, A Review of Porous Manganese Oxide Materials, Chemistry of Materials, 10 (1998) 2619-2628. [113] Y. Gorlin, B. Lassalle-Kaiser, J.D. Benck, S. Gul, S.M. Webb, V.K. Yachandra, J. Yano, T.F. Jaramillo, In Situ X-ray Absorption Spectroscopy Investigation of a Bifunctional Manganese Oxide Catalyst with High Activity for Electrochemical Water Oxidation and Oxygen Reduction, Journal of the American Chemical Society, 135 (2013) 8525-8534. [114] A.I. Krasil'shchikov, Intermediate stages in the anodic evolution of oxygen, 37 (1963). [115] R.L. Doyle, M.E.G. Lyons, An electrochemical impedance study of the oxygen evolution reaction at hydrous iron oxide in base, Physical Chemistry Chemical Physics, 15 (2013) 5224-5237. [116] H. Willems, A.G.C. Kobussen, J.H.W. De Wit, G.H.J. Broers, The oxygen evolution reaction on cobalt: Part I. Reaction order experiments and impedance measurements, Journal of Electroanalytical Chemistry and Interfacial Electrochemistry, 170 (1984) 227-242. [117] M.S. El-Deab, M.I. Awad, A.M. Mohammad, T. Ohsaka, Enhanced water electrolysis: Electrocatalytic generation of oxygen gas at manganese oxide nanorods modified electrodes, Electrochemistry Communications, 9 (2007) 2082-2087. [118] A.M. Fundo, L.M. Abrantes, Electrocatalytic behavior of electroless Ni-P alloys for the oxygen evolution reaction, Russ J Electrochem, 42 (2006) 1291-1297. [119] L.D. Burke, M.E. Lyons, O.J. Murphy, Formation of hydrous oxide films on cobalt under potential cycling conditions, Journal of Electroanalytical Chemistry and Interfacial Electrochemistry, 132 (1982) 247-261. [120] P. Rasiyah, A.C.C. Tseung, D.B. Hibbert, A Mechanistic Study of Oxygen Evolution on NiCo2 O 4 : I . Formation of Higher Oxides, Journal of The Electrochemical Society, 129 (1982) 1724-1727. [121] H.-Y. Su, Y. Gorlin, I.C. Man, F. Calle-Vallejo, J.K. Norskov, T.F. Jaramillo, J. Rossmeisl, Identifying active surface phases for metal oxide electrocatalysts: a study of manganese oxide bi-functional catalysts for oxygen reduction and water oxidation catalysis, Phys. Chem. Chem. Phys., 14 (2012) 14010-14022. [122] S. Gupta, W. Kellogg, H. Xu, X. Liu, J. Cho, G. Wu, Bifunctional Perovskite Oxide Catalysts for Oxygen Reduction and Evolution in Alkaline Media, Chemistry – An Asian Journal, 11 (2016) 10-21. [123] W.T. Hong, M. Risch, K.A. Stoerzinger, A. Grimaud, J. Suntivich, Y. Shao-Horn, Toward the rational design of non-precious transition metal oxides for oxygen electrocatalysis, Energy & Environmental Science, 8 (2015) 1404-1427. [124] I.C. Man, H.-Y. Su, F. Calle-Vallejo, H.A. Hansen, J.I. Martínez, N.G. Inoglu, J. Kitchin, T.F. Jaramillo, J.K. Nørskov, J. Rossmeisl, Universality in Oxygen Evolution Electrocatalysis on Oxide Surfaces, ChemCatChem, 3 (2011) 1159-1165. [125] J. Rossmeisl, A. Logadottir, J.K. Nørskov, Electrolysis of water on (oxidized) metal surfaces, Chemical Physics, 319 (2005) 178-184. 218  [126] J.B. Goodenough, R. Manoharan, M. Paranthaman, Surface protonation and electrochemical activity of oxides in aqueous solution, Journal of the American Chemical Society, 112 (1990) 2076-2082. [127] M.G. Mavros, T. Tsuchimochi, T. Kowalczyk, A. McIsaac, L.-P. Wang, T.V. Voorhis, What Can Density Functional Theory Tell Us about Artificial Catalytic Water Splitting?, Inorganic Chemistry, 53 (2014) 6386-6397. [128] L.A. De Faria, J.F.C. Boodts, S. Trasatti, Electrocatalytic properties of ternary oxide mixtures of composition Ru0.3Ti(0.7−x)CexO2: oxygen evolution from acidic solution, J Appl Electrochem, 26 (1996) 1195-1199. [129] J.B. Gerken, J.G. McAlpin, J.Y.C. Chen, M.L. Rigsby, W.H. Casey, R.D. Britt, S.S. Stahl, Electrochemical Water Oxidation with Cobalt-Based Electrocatalysts from pH 0–14: The Thermodynamic Basis for Catalyst Structure, Stability, and Activity, Journal of the American Chemical Society, 133 (2011) 14431-14442. [130] S. Jung, C.C.L. McCrory, I.M. Ferrer, J.C. Peters, T.F. Jaramillo, Benchmarking nanoparticulate metal oxide electrocatalysts for the alkaline water oxidation reaction, Journal of Materials Chemistry A, 4 (2016) 3068-3076. [131] C.C.L. McCrory, S. Jung, J.C. Peters, T.F. Jaramillo, Benchmarking Heterogeneous Electrocatalysts for the Oxygen Evolution Reaction, Journal of the American Chemical Society, 135 (2013) 16977-16987. [132] I. Katsounaros, S. Cherevko, A.R. Zeradjanin, K.J.J. Mayrhofer, Oxygen Electrochemistry as a Cornerstone for Sustainable Energy Conversion, Angewandte Chemie International Edition, 53 (2014) 102-121. [133] M.T.M. Koper, Thermodynamic theory of multi-electron transfer reactions: Implications for electrocatalysis, Journal of Electroanalytical Chemistry, 660 (2011) 254-260. [134] W. Sun, A. Hsu, R. Chen, Palladium-coated manganese dioxide catalysts for oxygen reduction reaction in alkaline media, Journal of Power Sources, 196 (2011) 4491-4498. [135] B. Klápště, J. Vondrák, J. Velická, MnOx/C composites as electrode materials II. Reduction of oxygen on bifunctional catalysts based on manganese oxides, Electrochim. Acta, 47 (2002) 2365-2369. [136] A. Ramírez, P. Hillebrand, D. Stellmach, M.M. May, P. Bogdanoff, S. Fiechter, Evaluation of MnOx, Mn2O3, and Mn3O4 Electrodeposited Films for the Oxygen Evolution Reaction of Water, The Journal of Physical Chemistry C, 118 (2014) 14073-14081. [137] J.W.D. Ng, M. Tang, T.F. Jaramillo, A carbon-free, precious-metal-free, high-performance O2 electrode for regenerative fuel cells and metal-air batteries, Energy & Environmental Science, 7 (2014) 2017-2024. [138] S.-B. Ma, K.-Y. Ahn, E.-S. Lee, K.-H. Oh, K.-B. Kim, Synthesis and characterization of manganese dioxide spontaneously coated on carbon nanotubes, Carbon, 45 (2007) 375-382. [139] J. Wei, I. Zhitomirsky, Electrosynthesis of manganese oxide films, Surface Engineering, 24 (2008) 40-46. [140] T. Nguyen, M. João Carmezim, M. Boudard, M. Fátima Montemor, Cathodic electrodeposition and electrochemical response of manganese oxide pseudocapacitor electrodes, International Journal of Hydrogen Energy, 40 (2015) 16355-16364. [141] E. Preisler, Material problems encountered in anodic MnO2 deposition, J Appl Electrochem, 19 (1989) 559-565. 219  [142] M. Nakayama, H. Tagashira, S. Konishi, K. Ogura, A Direct Electrochemical Route to Construct a Polymer/Manganese Oxide Layered Structure, Inorganic Chemistry, 43 (2004) 8215-8217. [143] B.A. Pinaud, Z. Chen, D.N. Abram, T.F. Jaramillo, Thin Films of Sodium Birnessite-Type MnO2: Optical Properties, Electronic Band Structure, and Solar Photoelectrochemistry, The Journal of Physical Chemistry C, 115 (2011) 11830-11838. [144] J. Tang, H.-M. Meng, L.L. Huang, Energy-saving and environmentally friendly electrodeposition of [gamma]-MnO2, RSC Advances, 4 (2014) 16512-16516. [145] W. Wei, X. Cui, X. Mao, W. Chen, D.G. Ivey, Morphology evolution in anodically electrodeposited manganese oxide nanostructures for electrochemical supercapacitor applications—Effect of supersaturation ratio, Electrochimica Acta, 56 (2011) 1619-1628. [146] C.-C. Hu, K.-H. Chang, Y.-T. Wu, C.-Y. Hung, C.-C. Lin, Y.-T. Tsai, Pulse deposition of large area, patterned manganese oxide nanowires in variable aspect ratios without templates, Electrochemistry Communications, 10 (2008) 1792-1796. [147] C.-W. Lee, K.-W. Nam, B.-W. Cho, K.-B. Kim, Electrochemical synthesis of meso-structured lamellar manganese oxide thin film, Microporous and Mesoporous Materials, 130 (2010) 208-214. [148] T. Shinomiya, V. Gupta, N. Miura, Effects of electrochemical-deposition method and microstructure on the capacitive characteristics of nano-sized manganese oxide, Electrochimica Acta, 51 (2006) 4412-4419. [149] A.J. Gibson, B. Johannessen, Y. Beyad, J. Allen, S.W. Donne, Dynamic Electrodeposition of Manganese Dioxide: Temporal Variation in the Electrodeposition Mechanism, Journal of The Electrochemical Society, 163 (2016) H305-H312. [150] A. Biswal, B. Chandra Tripathy, K. Sanjay, T. Subbaiah, M. Minakshi, Electrolytic manganese dioxide (EMD): a perspective on worldwide production, reserves and its role in electrochemistry, RSC Advances, 5 (2015) 58255-58283. [151] R.L. Paul, A. Cartwright, The mechanism of the deposition of manganese dioxide: Part II. Electrode impedance studies, Journal of Electroanalytical Chemistry and Interfacial Electrochemistry, 201 (1986) 113-122. [152] R.L. Paul, A. Cartwright, The mechanism of the deposition of manganese dioxide, Journal of Electroanalytical Chemistry and Interfacial Electrochemistry, 201 (1986) 123-131. [153] A. Cartwright, R.L. Paul, in: B. Schumm, H.M. Joseph, A. Kozawa (Eds.) Proceedings of the MnO2 SymposiumTokyo, 1980, pp. 290. [154] M. Fleischmann, H.R. Thirsk, I.M. Tordesillas, Kinetics of electrodeposition of [gamma]-manganese dioxide, Transactions of the Faraday Society, 58 (1962) 1865-1877. [155] B. Babakhani, D.G. Ivey, Effect of electrodeposition conditions on the electrochemical capacitive behavior of synthesized manganese oxide electrodes, Journal of Power Sources, 196 (2011) 10762-10774. [156] S. Nijjer, J. Thonstad, G.M. Haarberg, Oxidation of manganese(II) and reduction of manganese dioxide in sulphuric acid, Electrochimica Acta, 46 (2000) 395-399. [157] G. Qiu, H. Wen, Q. Guo, Y. Hu, D. Yang, F. Liu, Electrochemical preparation of nanosized manganese dioxides from manganese chloride solutions, Ionics, 17 (2011) 209-216. [158] J.-K. Chang, W.-T. Tsai, Material Characterization and Electrochemical Performance of Hydrous Manganese Oxide Electrodes for Use in Electrochemical Pseudocapacitors, Journal of The Electrochemical Society, 150 (2003) A1333-A1338. 220  [159] A. Cross, A. Morel, A. Cormie, T. Hollenkamp, S. Donne, Enhanced manganese dioxide supercapacitor electrodes produced by electrodeposition, Journal of Power Sources, 196 (2011) 7847-7853. [160] L. Chih-Hsiang, H. Chii-Shyang, Nanostructures and Capacitive Properties of Manganese Oxide Film Synthesized by Hydrothermal Electrochemical Process, Japanese Journal of Applied Physics, 47 (2008) 4682. [161] M. Ghaemi, L. Khosravi-Fard, J. Neshati, Improved performance of rechargeable alkaline batteries via surfactant-mediated electrosynthesis of MnO2, Journal of Power Sources, 141 (2005) 340-350. [162] A. Biswal, B.C. Tripathy, T. Subbaiah, D. Meyrick, M. Minakshi, Dual Effect of Anionic Surfactants in the Electrodeposited MnO2 Trafficking Redox Ions for Energy Storage, Journal of The Electrochemical Society, 162 (2015) A30-A38. [163] R. Vittal, H. Gomathi, Beneficial Effects of Cetyltrimethylammonium Bromide in the Modification of Electrodes with Cobalt Hexacyanoferrate Surface Films, The Journal of Physical Chemistry B, 106 (2002) 10135-10143. [164] J.F. Rusling, Controlling electrochemical catalysis with surfactant microstructures, Accounts of Chemical Research, 24 (1991) 75-81. [165] J.F. Rusling, Liquid crystal surfactant films for electrochemical catalysis, Microporous Materials, 3 (1994) 1-16. [166] J.O. Besenhard, J. Gürtler, P. Komenda, A. Paxinos, Corrosion protection of secondary lithium electrodes in organic electrolytes, Journal of Power Sources, 20 (1987) 253-258. [167] J.F. Rusling, D.J. Howe, Electron transfer in surfactant films on electrodes; copper phthalocyaninetetrasulfonate-didodecyldimethylammonium bromide, Inorganica Chimica Acta, 226 (1994) 159-169. [168] M.J. Rosen, Characteristic Features of Surfactants,  Surfactants and Interfacial Phenomena, John Wiley & Sons, Inc.2004, pp. 1-33. [169] J.A. Kent, SoapSoap, Fatty acidsFatty Acids, and Synthetic Detergents, in: J.A. Kent (Ed.) Riegel's Handbook of Industrial Chemistry, Springer US, Boston, MA, 2003, pp. 1098-1140. [170] A. Biswal, B.C. Tripathy, T. Subbaiah, D. Meyrick, M. Ionescu, M. Minakshi, Effect of Non-ionic Surfactants and Its Role in K Intercalation in Electrolytic Manganese Dioxide, Metallurgical and Materials Transactions E, 1 (2014) 226-238. [171] S. Devaraj, N. Munichandraiah, The Effect of Nonionic Surfactant Triton X-100 during Electrochemical Deposition of MnO2 on Its Capacitance Properties, Journal of The Electrochemical Society, 154 (2007) A901-A909. [172] D.P. Dubal, W.B. Kim, C.D. Lokhande, Surfactant assisted electrodeposition of MnO2 thin films: Improved supercapacitive properties, Journal of Alloys and Compounds, 509 (2011) 10050-10054. [173] M.J. Rosen, Micelle Formation by Surfactants,  Surfactants and Interfacial Phenomena, John Wiley & Sons, Inc.2004, pp. 105-177. [174] X. Cui, S. Mao, M. Liu, H. Yuan, Y. Du, Mechanism of Surfactant Micelle Formation, Langmuir, 24 (2008) 10771-10775. [175] A. Berthod, C.G. Coque, Micellar Liquid Chromatography, Marcel Dekker Inc., New York, 2000. [176] O.R. Fennema, M. Karel, G.W. Sanderson, S.R. Tannenbaum, P. Walstra, J.R. Whitaker, Food emulsions, Third ed., Marcel Dekker Inc., New York, 1997. 221  [177] M. Sawangphruk, S. Pinitsoontorn, J. Limtrakul, Surfactant-assisted electrodeposition and improved electrochemical capacitance of silver-doped manganese oxide pseudocapacitor electrodes, Journal of Solid State Electrochemistry, 16 (2012) 2623-2629. [178] H.-U. Kim, K.-H. Lim, A model on the temperature dependence of critical micelle concentration, Colloids and Surfaces A: Physicochemical and Engineering Aspects, 235 (2004) 121-128. [179] B.D. Flockhart, The effect of temperature on the critical micelle concentration of some paraffin-chain salts, Journal of Colloid Science, 16 (1961) 484-492. [180] K.-H. Kang, H.-U. Kim, K.-H. Lim, Effect of temperature on critical micelle concentration and thermodynamic potentials of micellization of anionic ammonium dodecyl sulfate and cationic octadecyl trimethyl ammonium chloride, Colloids and Surfaces A: Physicochemical and Engineering Aspects, 189 (2001) 113-121. [181] L.-J. Chen, S.-Y. Lin, C.-C. Huang, E.-M. Chen, Temperature dependence of critical micelle concentration of polyoxyethylenated non-ionic surfactants, Colloids and Surfaces A: Physicochemical and Engineering Aspects, 135 (1998) 175-181. [182] J.A. Stead, H. Taylor, Some solution properties of certain surface-active N-alkylpyridinium halides, Journal of Colloid and Interface Science, 30 (1969) 482-488. [183] C.O. Soares, M.D. Carvalho, M.E. Melo Jorge, A. Gomes, R.A. Silva, C.M. Rangel, M.I. da Silva Pereira, High surface area LaNiO3 electrodes for oxygen electrocatalysis in alkaline media, J Appl Electrochem, 42 (2012) 325-332. [184] J. Suntivich, H.A. Gasteiger, N. Yabuuchi, H. Nakanishi, J.B. Goodenough, Y. Shao-Horn, Design principles for oxygen-reduction activity on perovskite oxide catalysts for fuel cells and metal–air batteries, Nat. Chem., 3 (2011) 546-550. [185] C.O. Soares, M.D. Carvalho, M.E. Melo Jorge, A. Gomes, R.A. Silva, C.M. Rangel, M.I. da Silva Pereira, High surface area LaNiO3 electrodes for oxygen electrocatalysis in alkaline media, J Appl Electrochem, 42 (2012) 325-332. [186] X. Yang, Y. Liu, S. Li, X. Wei, L. Wang, Y. Chen, A direct borohydride fuel cell with a polymer fiber membrane and non-noble metal catalysts, Sci. Rep., 2 (2012). [187] X. Li, W. Qu, J. Zhang, H. Wang, Electrocatalytic Activities of La0.6Ca0.4CoO3 and La0.6Ca0.4CoO3-Carbon Composites Toward the Oxygen Reduction Reaction in Concentrated Alkaline Electrolytes, J. Electrochem. Soc., 158 (2011) A597-A604. [188] S.P. Singh, R.N. Singh, G. Poillearat, P. Chartier, Physicochemical and electrochemical characterization of active films of LaNiO3 for use as anode in alkaline water electrolysis, International Journal of Hydrogen Energy, 20 (1995) 203-210. [189] R.N. Singh, L. Bahadur, J.P. Pandey, S.P. Singh, P. Chartier, G. Poillerat, Preparation and characterization of thin films of LaNiO3 for anode application in alkaline water electrolysis, J Appl Electrochem, 24 (1994) 149-156. [190] M. Risch, A. Grimaud, K.J. May, K.A. Stoerzinger, T.J. Chen, A.N. Mansour, Y. Shao-Horn, Structural Changes of Cobalt-Based Perovskites upon Water Oxidation Investigated by EXAFS, The Journal of Physical Chemistry C, 117 (2013) 8628-8635. [191] T. Takeguchi, T. Yamanaka, H. Takahashi, H. Watanabe, T. Kuroki, H. Nakanishi, Y. Orikasa, Y. Uchimoto, H. Takano, N. Ohguri, M. Matsuda, T. Murota, K. Uosaki, W. Ueda, Layered Perovskite Oxide: A Reversible Air Electrode for Oxygen Evolution/Reduction in Rechargeable Metal-Air Batteries, Journal of the American Chemical Society, 135 (2013) 11125-11130. 222  [192] D.B. Rogers, R.D. Shannon, A.W. Sleight, J.L. Gillson, Crystal chemistry of metal dioxides with rutile-related structures, Inorganic Chemistry, 8 (1969) 841-849. [193] A.M. Dehkhoda, N. Ellis, E. Gyenge, Electrosorption on activated biochar: effect of thermo-chemical activation treatment on the electric double layer capacitance, J Appl Electrochem, 44 (2014) 141-157. [194] X. Yang, Y. Liu, S. Li, X. Wei, L. Wang, Y. Chen, A direct borohydride fuel cell with a polymer fiber membrane and non-noble metal catalysts, Sci. Rep., 2 (2012) 567. [195] J.F. Vente, D.J.W. Ijdo, The orthorhombic fluorite related compounds Ln3IrO7, Materials Research Bulletin, 26 (1991) 1255-1262. [196] H. Nishimine, M. Wakeshima, Y. Hinatsu, Crystal structures, magnetic and thermal properties of Ln3IrO7 (Ln=Pr, Nd, Sm, and Eu), Journal of Solid State Chemistry, 177 (2004) 739-744. [197] F. Jaouen, E. Proietti, M. Lefevre, R. Chenitz, J.-P. Dodelet, G. Wu, H.T. Chung, C.M. Johnston, P. Zelenay, Recent advances in non-precious metal catalysis for oxygen-reduction reaction in polymer electrolyte fuel cells, Energy & Environmental Science, 4 (2011) 114-130. [198] X. Lang, A. Hirata, T. Fujita, M. Chen, Nanoporous metal/oxide hybrid electrodes for electrochemical supercapacitors, Nat Nano, 6 (2011) 232-236. [199] Z.-L. Wang, D. Xu, J.-J. Xu, X.-B. Zhang, Oxygen electrocatalysts in metal-air batteries: from aqueous to nonaqueous electrolytes, Chemical Society Reviews, 43 (2014) 7746-7786. [200] W.G. Hardin, D.A. Slanac, X. Wang, S. Dai, K.P. Johnston, K.J. Stevenson, Highly Active, Nonprecious Metal Perovskite Electrocatalysts for Bifunctional Metal–Air Battery Electrodes, The Journal of Physical Chemistry Letters, 4 (2013) 1254-1259. [201] D. Banham, F. Feng, T. Fürstenhaupt, K. Pei, S. Ye, V. Birss, Effect of Pt-loaded carbon support nanostructure on oxygen reduction catalysis, Journal of Power Sources, 196 (2011) 5438-5445. [202] S. Müller, K. Striebel, O. Haas, La0.6Ca0.4CoO3: a stable and powerful catalyst for bifunctional air electrodes, Electrochimica Acta, 39 (1994) 1661-1668. [203] P.N. Ross, M. Sattler, The Corrosion of Carbon Black Anodes in Alkaline Electrolyte: III . The Effect of Graphitization on the Corrosion Resistance of Furnace Blacks, Journal of The Electrochemical Society, 135 (1988) 1464-1470. [204] P.N. Ross, H. Sokol, The Corrosion of Carbon Black Anodes in Alkaline Electrolyte: I . Acetylene Black and the Effect of Cobalt Catalyzation, Journal of The Electrochemical Society, 131 (1984) 1742-1750. [205] N. Staud, P.N. Ross, The Corrosion of Carbon Black Anodes in Alkaline Electrolyte: II . Acetylene Black and the Effect of Oxygen Evolution Catalysts on Corrosion, Journal of The Electrochemical Society, 133 (1986) 1079-1084. [206] J. Zhang, Z. Zhao, Z. Xia, L. Dai, A metal-free bifunctional electrocatalyst for oxygen reduction and oxygen evolution reactions, Nature Nanotechnology, 10 (2015) 444-452. [207] X. Bao, D.v. Deak, E.J. Biddinger, U.S. Ozkan, C.M. Hadad, A computational exploration of the oxygen reduction reaction over a carbon catalyst containing a phosphinate functional group, Chemical Communications, 46 (2010) 8621-8623. [208] C.H. Choi, S.H. Park, S.I. Woo, Heteroatom doped carbons prepared by the pyrolysis of bio-derived amino acids as highly active catalysts for oxygen electro-reduction reactions, Green Chemistry, 13 (2011) 406-412. 223  [209] S. Zhang, Y. Shao, G. Yin, Y. Lin, Carbon nanotubes decorated with Pt nanoparticles via electrostatic self-assembly: a highly active oxygen reduction electrocatalyst, Journal of Materials Chemistry, 20 (2010) 2826-2830. [210] R. Liu, S. Li, X. Yu, G. Zhang, Y. Ma, J. Yao, Facile synthesis of a Ag nanoparticle/polyoxometalate/carbon nanotube tri-component hybrid and its activity in the electrocatalysis of oxygen reduction, Journal of Materials Chemistry, 21 (2011) 14917-14924. [211] T. Fujigaya, T. Uchinoumi, K. Kaneko, N. Nakashima, Design and synthesis of nitrogen-containing calcined polymer/carbon nanotube hybrids that act as a platinum-free oxygen reduction fuel cell catalyst, Chemical Communications, 47 (2011) 6843-6845. [212] Y. Li, H. Dai, Recent advances in zinc-air batteries, Chemical Society Reviews, 43 (2014) 5257-5275. [213] D. Ye, T. Wu, H. Cao, Y. Wang, B. Liu, S. Zhang, J. Kong, Electrocatalysis of both oxygen reduction and water oxidation using a cost-effective three-dimensional MnO2/graphene/carbon nanotube, RSC Advances, 5 (2015) 26710-26715. [214] N.I. Andersen, A. Serov, P. Atanassov, Metal oxides/CNT nano-composite catalysts for oxygen reduction/oxygen evolution in alkaline media, Applied Catalysis B: Environmental, 163 (2015) 623-627. [215] M.M. Natile, E. Ugel, C. Maccato, A. Glisenti, LaCoO3: Effect of synthesis conditions on properties and reactivity, Appl. Catal. B., 72 (2007) 351-362. [216] H. Nishimine, Y. Doi, Y. Hinatsu, M. Sato, Phase Transition of Ln3IrO7 (Ln=Pr, Nd, Sm, Eu) and its Low-Temperature Structure, Journal of the Ceramic Society of Japan, 115 (2007) 577-581. [217] A.T. Najafabadi, E. Gyenge, High-yield graphene production by electrochemical exfoliation of graphite: Novel ionic liquid (IL)–acetonitrile electrolyte with low IL content, Carbon, 71 (2014) 58-69. [218] A.J.F. Bard, Larry R. , Electrochemical Methods: Fundamentals and Applications, 2nd ed., Wiley, New York, 2001. [219] E. Gyenge, CHBE 577: Fuel Cells and Electrochemical Engineering Course Notes, 2012. [220] T. Poux, F.S. Napolskiy, T. Dintzer, G. Kéranguéven, S.Y. Istomin, G.A. Tsirlina, E.V. Antipov, E.R. Savinova, Dual role of carbon in the catalytic layers of perovskite/carbon composites for the electrocatalytic oxygen reduction reaction, Catalysis Today, 189 (2012) 83-92. [221] J.O. Bockris, apos, M., Kinetics of Activation Controlled Consecutive Electrochemical Reactions: Anodic Evolution of Oxygen, The Journal of Chemical Physics, 24 (1956) 817-827. [222] P. Castelli, S. Trasatti, F.H. Pollak, W.E. O'Grady, Single crystals as model electrocatalysts, Journal of Electroanalytical Chemistry, 210 (1986) 189-194. [223] C. Bocca, A. Barbucci, G. Cerisola, The influence of surface finishing on the electrocatalytic properties of nickel for the oxygen evolution reaction (OER) in alkaline solution, International Journal of Hydrogen Energy, 23 (1998) 247-252. [224] J.O. Bockris, T. Otagawa, Mechanism of oxygen evolution on perovskites, The Journal of Physical Chemistry, 87 (1983) 2960-2971. [225] A. Kapałka, G. Fóti, C. Comninellis, Determination of the Tafel slope for oxygen evolution on boron-doped diamond electrodes, Electrochemistry Communications, 10 (2008) 607-610. [226] H. Vogt, Ö. Aras, R.J. Balzer, The limits of the analogy between boiling and gas evolution at electrodes, International Journal of Heat and Mass Transfer, 47 (2004) 787-795. 224  [227] A. Kozawa, T. Kalnoki‐Kis, J.F. Yeager, Solubilities of Mn(II) and Mn(III) Ions in Concentrated Alkaline Solutions, Journal of The Electrochemical Society, 113 (1966) 405-409. [228] M. Minakshi, Examining manganese dioxide electrode in KOH electrolyte using TEM technique, Journal of Electroanalytical Chemistry, 616 (2008) 99-106. [229] T. Riedl, T. Gemming, K. Wetzig, Extraction of EELS white-line intensities of manganese compounds: Methods, accuracy, and valence sensitivity, Ultramicroscopy, 106 (2006) 284-291. [230] Q. Gao, C. Ranjan, Z. Pavlovic, R. Blume, R. Schlögl, Enhancement of Stability and Activity of MnOx/Au Electrocatalysts for Oxygen Evolution through Adequate Electrolyte Composition, ACS Catalysis, 5 (2015) 7265-7275. [231] E.J. Rubin, R. Baboian, A Correlation of the Solution Properties and the Electrochemical Behavior of the Nickel Hydroxide Electrode in Binary Aqueous Alkali Hydroxides, Journal of The Electrochemical Society, 118 (1971) 428-433. [232] J. Suntivich, E. Perry, H. Gasteiger, Y. Shao-Horn, The Influence of the Cation on the Oxygen Reduction and Evolution Activities of Oxide Surfaces in Alkaline Electrolyte, Electrocatalysis, 4 (2013) 49-55. [233] P. Ruetschi, Cation‐Vacancy Model for MnO2, Journal of The Electrochemical Society, 131 (1984) 2737-2744. [234] P. Ruetschi, R. Giovanoli, Cation Vacancies in MnO2 and Their Influence on Electrochemical Reactivity, Journal of The Electrochemical Society, 135 (1988) 2663-2669. [235] R. Shannon, Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides, Acta Crystallographica Section A, 32 (1976) 751-767. [236] M. Toupin, T. Brousse, D. Bélanger, Influence of Microstucture on the Charge Storage Properties of Chemically Synthesized Manganese Dioxide, Chemistry of Materials, 14 (2002) 3946-3952. [237] M. Chigane, M. Ishikawa, Manganese Oxide Thin Film Preparation by Potentiostatic Electrolyses and Electrochromism, Journal of The Electrochemical Society, 147 (2000) 2246-2251. [238] N. Nagarajan, M. Cheong, I. Zhitomirsky, Electrochemical capacitance of MnOx films, Materials Chemistry and Physics, 103 (2007) 47-53. [239] L. Laffont, P. Gibot, High resolution electron energy loss spectroscopy of manganese oxides: Application to Mn3O4 nanoparticles, Materials Characterization, 61 (2010) 1268-1273. [240] H.K. Schmid, W. Mader, Oxidation states of Mn and Fe in various compound oxide systems, Micron (Oxford, England : 1993), 37 (2006) 426-432. [241] H. Kurata, C. Colliex, Electron-energy-loss core-edge structures in manganese oxides, Physical Review B, 48 (1993) 2102-2108. [242] G.H. Du, Z.Y. Yuan, G. Van Tendeloo, Transmission electron microscopy and electron energy-loss spectroscopy analysis of manganese oxide nanowires, Applied Physics Letters, 86 (2005) -. [243] Y. Li, H. Tan, O. Lebedev, J. Verbeeck, E. Biermans, G. Van Tendeloo, B.-L. Su, Insight into the Growth of Multiple Branched MnOOH Nanorods, Crystal Growth & Design, 10 (2010) 2969-2976. [244] H. Tan, J. Verbeeck, A. Abakumov, G. Van Tendeloo, Oxidation state and chemical shift investigation in transition metal oxides by EELS, Ultramicroscopy, 116 (2012) 24-33. [245] J. Kaczmarek, E. Wolska, Cation and Vacancy Distribution in Nonstoichiometric Hausmanite, Journal of Solid State Chemistry, 103 (1993) 387-393. 225  [246] M.-R. Gao, Y.-F. Xu, J. Jiang, Y.-R. Zheng, S.-H. Yu, Water Oxidation Electrocatalyzed by an Efficient Mn3O4/CoSe2 Nanocomposite, Journal of the American Chemical Society, 134 (2012) 2930-2933. [247] K.-W. Nam, M.G. Kim, K.-B. Kim, In Situ Mn K-edge X-ray Absorption Spectroscopy Studies of Electrodeposited Manganese Oxide Films for Electrochemical Capacitors, The Journal of Physical Chemistry C, 111 (2006) 749-758. [248] D.P. Dubal, A.D. Jagadale, C.D. Lokhande, Big as well as light weight portable, Mn3O4 based symmetric supercapacitive devices: Fabrication, performance evaluation and demonstration, Electrochimica Acta, 80 (2012) 160-170. [249] Y. Dai, K. Wang, J. Xie, From spinel Mn3O4 to layered nanoarchitectures using electrochemical cycling and the distinctive pseudocapacitive behavior, Applied Physics Letters, 90 (2007) 104102. [250] D.P. Dubal, D.S. Dhawale, R.R. Salunkhe, C.D. Lokhande, A novel chemical synthesis of Mn3O4 thin film and its stepwise conversion into birnessite MnO2 during super capacitive studies, Journal of Electroanalytical Chemistry, 647 (2010) 60-65. [251] B. Djurfors, J.N. Broughton, M.J. Brett, D.G. Ivey, Microstructural characterization of porous manganese thin films for electrochemical supercapacitor applications, Journal of Materials Science, 38 (2003) 4817-4830. [252] W.C. Chueh, S.M. Haile, Electrochemistry of Mixed Oxygen Ion and Electron Conducting Electrodes in Solid Electrolyte Cells, Annual Review of Chemical and Biomolecular Engineering, 3 (2012) 313-341. [253] J. Suntivich, H.A. Gasteiger, N. Yabuuchi, H. Nakanishi, J.B. Goodenough, Y. Shao-Horn, Design principles for oxygen-reduction activity on perovskite oxide catalysts for fuel cells and metal–air batteries, Nat Chem, 3 (2011) 546-550. [254] D.N. Mueller, M.L. Machala, H. Bluhm, W.C. Chueh, Redox activity of surface oxygen anions in oxygen-deficient perovskite oxides during electrochemical reactions, Nat Commun, 6 (2015). [255] C. Wagner, Über den Mechanismus der elektrischen Stromleitung im Nernststift, Naturwissenschaften, 31 (1943) 265-268. [256] C.-F. Chen, G. King, R.M. Dickerson, P.A. Papin, S. Gupta, W.R. Kellogg, G. Wu, Oxygen-deficient BaTiO3−x perovskite as an efficient bifunctional oxygen electrocatalyst, Nano Energy, 13 (2015) 423-432. [257] K.J. May, C.E. Carlton, K.A. Stoerzinger, M. Risch, J. Suntivich, Y.-L. Lee, A. Grimaud, Y. Shao-Horn, Influence of Oxygen Evolution during Water Oxidation on the Surface of Perovskite Oxide Catalysts, The Journal of Physical Chemistry Letters, 3 (2012) 3264-3270. [258] L. Li, X. Feng, Y. Nie, S. Chen, F. Shi, K. Xiong, W. Ding, X. Qi, J. Hu, Z. Wei, L.-J. Wan, M. Xia, Insight into the Effect of Oxygen Vacancy Concentration on the Catalytic Performance of MnO2, ACS Catalysis, 5 (2015) 4825-4832. [259] F.-D. Kong, S. Zhang, G.-P. Yin, Z.-B. Wang, C.-Y. Du, G.-Y. Chen, N. Zhang, Electrochemical studies of Pt/Ir–IrO2 electrocatalyst as a bifunctional oxygen electrode, International Journal of Hydrogen Energy, 37 (2012) 59-67. [260] F.-D. Kong, S. Zhang, G.-P. Yin, N. Zhang, Z.-B. Wang, C.-Y. Du, Preparation of Pt/Irx(IrO2)10−x bifunctional oxygen catalyst for unitized regenerative fuel cell, Journal of Power Sources, 210 (2012) 321-326. 226  [261] Z.D. Wei, M.B. Ji, S.G. Chen, Y. Liu, C.X. Sun, G.Z. Yin, P.K. Shen, S.H. Chan, Water electrolysis on carbon electrodes enhanced by surfactant, Electrochimica Acta, 52 (2007) 3323-3329. [262] M. Chigane, M. Ishikawa, M. Izaki, Preparation of Manganese Oxide Thin Films by Electrolysis/Chemical Deposition and Electrochromism, Journal of The Electrochemical Society, 148 (2001) D96-D101. [263] J.S. Foord, R.B. Jackman, G.C. Allen, An X-ray photoelectron spectroscopic investigation of the oxidation of manganese, Philosophical Magazine A, 49 (1984) 657-663. [264] F. Hahn, D. Floner, B. Beden, C. Lamy, In situ investigation of the behaviour of a nickel electrode in alkaline solution by uv-vis and ir reflectance spectroscopies, Electrochimica Acta, 32 (1987) 1631-1636. [265] M. Grdeń, M. Alsabet, G. Jerkiewicz, Surface Science and Electrochemical Analysis of Nickel Foams, ACS Applied Materials & Interfaces, 4 (2012) 3012-3021. [266] J. van Drunen, B. Kinkead, M.C.P. Wang, E. Sourty, B.D. Gates, G. Jerkiewicz, Comprehensive Structural, Surface-Chemical and Electrochemical Characterization of Nickel-Based Metallic Foams, ACS Applied Materials & Interfaces, 5 (2013) 6712-6722. [267] D.C. Montgomery, Design and Analysis of Experiments, 7th ed., Wiley, Hoboken, NJ, 2009. [268] M.L. Calegaro, F.H.B. Lima, E.A. Ticianelli, Oxygen reduction reaction on nanosized manganese oxide particles dispersed on carbon in alkaline solutions, Journal of Power Sources, 158 (2006) 735-739. [269] J. Staszak-Jirkovský, R. Subbaraman, D. Strmcnik, K.L. Harrison, C.E. Diesendruck, R. Assary, O. Frank, L. Kobr, G.K.H. Wiberg, B. Genorio, J.G. Connell, P.P. Lopes, V.R. Stamenkovic, L. Curtiss, J.S. Moore, K.R. Zavadil, N.M. Markovic, Water as a Promoter and Catalyst for Dioxygen Electrochemistry in Aqueous and Organic Media, ACS Catalysis, 5 (2015) 6600-6607. [270] X. Li, W. Qu, J. Zhang, H. Wang, Electrocatalytic Activities of La0.6Ca0.4CoO3 and La0.6Ca0.4CoO3-Carbon Composites Toward the Oxygen Reduction Reaction in Concentrated Alkaline Electrolytes, Journal of The Electrochemical Society, 158 (2011) A597-A604. [271] J. Sunarso, A.A.J. Torriero, W. Zhou, P.C. Howlett, M. Forsyth, Oxygen Reduction Reaction Activity of La-Based Perovskite Oxides in Alkaline Medium: A Thin-Film Rotating Ring-Disk Electrode Study, The Journal of Physical Chemistry C, 116 (2012) 5827-5834. [272] J. Jiang, B. Yi, Thickness effects of a carbon-supported platinum catalyst layer on the electrochemical reduction of oxygen in sulfuric acid solution, Journal of Electroanalytical Chemistry, 577 (2005) 107-115. [273] G.-L. Tian, M.-Q. Zhao, D. Yu, X.-Y. Kong, J.-Q. Huang, Q. Zhang, F. Wei, Nitrogen-Doped Graphene/Carbon Nanotube Hybrids: In Situ Formation on Bifunctional Catalysts and Their Superior Electrocatalytic Activity for Oxygen Evolution/Reduction Reaction, Small, 10 (2014) 2251-2259. [274] N. Staud, H. Sokol, P.N. Ross, The Corrosion of Carbon Black Anodes in Alkaline Electrolyte: IV . Current Efficiencies for Oxygen Evolution from Metal Oxide‐Impregnated Graphitized Furnace Blacks, Journal of The Electrochemical Society, 136 (1989) 3570-3576. [275] N.S. Marinković, N.M. Marković, R.R. Adz̆ić, An International Journal Devoted to all Aspects of Electrode Kinetics, Interfacial Structure, Properties of Electrolytes, Colloid and Biological ElectrochemistryHydrogen adsorption on single-crystal platinum electrodes in alkaline solutions, Journal of Electroanalytical Chemistry, 330 (1992) 433-452. 227  [276] F.T. Wagner, P.N. Ross, Long-range structural effects in the anomalous voltammetry on ultra-high vacuum prepared Pt (111), Journal of Electroanalytical Chemistry and Interfacial Electrochemistry, 250 (1988) 301-320. [277] D.E. Hall, Alkaline Water Electrolysis Anode Materials, Journal of The Electrochemical Society, 132 (1985) 41C-48C. [278] M. Kar, T.J. Simons, M. Forsyth, D.R. MacFarlane, Ionic liquid electrolytes as a platform for rechargeable metal-air batteries: a perspective, Physical Chemistry Chemical Physics, 16 (2014) 18658-18674. [279] J. Tafel, Polarization in cathodic hydrogen evolution, Zeitschrift fuer Physikalische Chemie, Stoechiometrie und Verwandtschaftslehre, 50 (1905) 641-712. [280] T. Erdey-Gruz, M. Volmer, The theory of hydrogen overvoltage, Z. Phys. Chem, 150 (1930) 203-213. [281] T. Erdey-Gruz, H. Wick, On the issue of hydrogen overvoltage, Z Physik Chem A, 162 (1932) 53-62. [282] J.A.V. Butler, Studies in heterogeneous equilibria. Part II.-The kinetic interpretation of the nernst theory of electromotive force, Transactions of the Faraday Society, 19 (1924) 729-733. [283] J.A.V. Butler, Studies in heterogeneous equilibria. Part III. A kinetic theory of reversible oxidation potentials at inert electrodes, Transactions of the Faraday Society, 19 (1924) 734-739. [284] J.A.V. Butler, The mechanism of overvoltage and its relation to the combination of hydrogen atoms at metal electrodes, Transactions of the Faraday Society, 28 (1932) 379-382. [285] U.A. Paulus, T.J. Schmidt, H.A. Gasteiger, R.J. Behm, Oxygen reduction on a high-surface area Pt/Vulcan carbon catalyst: a thin-film rotating ring-disk electrode study, Journal of Electroanalytical Chemistry, 495 (2001) 134-145. [286] K. Kinoshita, Electrochemical Oxygen Technology, 1992. [287] W. Xing, G. Yin, J. Zhang, Rotating Electrode Methods and Oxygen Reduction Electrocatalysts, Elsevier Science & Technology, Amsterdam, NLD, 2014. 228  Appendices Appendix A  : Introduction to the Butler-Erdey-Grúz-Volmer (BE-GV) equation In 1905, J. Tafel proposed, based on experimental studies, that the current density increases exponentially with the overpotential [279]. However, the fundamental equation of electrode kinetics was developed later in the late 1920s and 1930s mainly by T. Erdey-Grúz and M. Volmer with independent contribution from J. Butler [218, 219, 280-284]. The BE-GV equation expresses the net current density as a function of surface overpotential for a single reaction at a single electrode (eq.  40) (Figure A.1) [103, 218, 219]:  𝑖 = 𝑖0 [𝑒𝑥𝑝 (𝛼𝑎𝐹𝑅𝑇𝜂) − 𝑒𝑥𝑝 (−𝛼𝑐𝐹𝑅𝑇𝜂)] (40) 𝜂 = 𝐸 − 𝐸𝑒 (41) 𝛼𝑎 = (1 − 𝛽). 𝑛 (42) 𝛼𝑐 = 𝛽. 𝑛 (43)  where 𝑖0 is the exchange current density [A m-2], 𝛼𝑎 is the anodic transfer coefficient, F is the Faraday constant [C mol-1], 𝜂 is the surface overpotential [V], 𝑅 is the gas constant [J mol-1 K-1], 𝑇  is the temperature [K], 𝛼𝑐  is the cathodic transfer coefficient, 𝐸  is the operating electrode potential [VSHE], 𝐸𝑒 is the equilibrium electrode potential [VSHE], 𝛽 is the symmetry factor and 𝑛 is the number of electron exchanged in the rate determining step of the reaction.  The exchange current density is a measure of the electrocatalytic properties of the electrode surface. Considering multiple species that affect the rate of a reaction, the exchange current density can be written as [218, 219]: 229   𝑖0 = 𝑛𝐹𝑘0∏ 𝑎𝑗𝑚𝑗𝑗  (44)  where 𝑘0 is the standard heterogeneous rate constant of the electrochemical reaction [mol m-2 s-1], 𝑎𝑗 is the activities of species j and 𝑚𝑗 is a power expressing the concentration/pressure dependence of the activity for species j. The exchange current density (𝑖0 ) is a function of the catalytic properties of the electrode (via heterogeneous rate constant), temperature, activities of species and reaction mechanism. Typically, the exchange current density values over 1 A m-2 correspond to “fast” electrode kinetics [219]. For practical calculations, when the absolute value of overpotential exceeds 0.1 V (|η| > 0.1 V), either the anodic or cathodic partial current density is significant and the eq. 40 can be simplified to Tafel equation [103, 218, 219]:  |𝜂| = 𝑏. 𝑙𝑜𝑔|𝑖| + 𝑎 (45) 𝑏 =2.303𝑅𝑇𝛼𝐹 (46) 𝑎 = −2.303𝑅𝑇𝛼𝐹𝑙𝑜𝑔 𝑖0 = −𝑏 𝑙𝑜𝑔 𝑖0 (47)  where 𝑎 and 𝑏 [V decade-1] are Tafel parameters, Tafel intercept and Tafel slope, respectively, and 𝛼 is the transfer coefficient. Both Tafel parameters are temperature and composition dependent constants containing the electrode reaction parameters. The experimentally measured Tafel slopes are typically between 30 to 300 mV decade-1 with high apparent Tafel slope values (i.e. over 230  120 mV decade−1) indicating other polarization effects such as adsorption, mass transfer, etc. interfere with “pure” intrinsic kinetic measurement [218, 219].   Figure A.1 Effect of the symmetry factor (β) on the symmetry of the current-overpotential curve described by the BE-GV equation (eq. 40).     i (A m-2) η (V) η > 0 η < 0 i > 0 i < 0 β=0.75 β =0.5 β =0.25 231  Appendix B  : Introduction to the rotating disk and ring disk electrode (RDE and RRDE) The rotating disk and ring-disk electrodes provide a hydro-dynamically well-characterized, laminar-forced convective flow to the smooth electrode surface, up to rotation speeds of about 5000 rpm (rotation per minute). Generally, the polarization curve measured under these conditions has a sigmoid shape with three distinct regions (Figure B.1) [218]: 1) Pure electrode kinetic (or charge transfer) control region, where the convective mass transfer has no effect on the polarization curve. 2) Mixed control region, where both electrode kinetic and mass transfer influence the polarization curve. 3) Pure mass transfer control region, generating a limiting current density (𝑖𝐿).   Figure B.1 Typical cathodic polarization curves obtained by a RDE as a function of angular velocity (ω). ω1 < ω2 < ω3. 232  In the mass transfer control region, the limiting current density (iL) for the oxidation or reduction of an electroactive species j can be expressed using [218, 219]:  𝑖𝐿 =𝑛 ∙ 𝐹𝑠𝑗∙𝐷𝑗𝛿∙ 𝐶𝑗 =𝑛 ∙ 𝐹𝑠𝑗∙ 𝐾𝑚,𝑗 ∙ 𝐶𝑗 (48) 𝛿 = 1.61 ∙ 𝐷𝑗13 ∙ 𝜈16 ∙1√𝜛 (𝑙𝑎𝑚𝑖𝑛𝑎𝑟 𝑓𝑙𝑜𝑤, 102 ≤ 𝑅𝑒 ≤ 104) (49)  where sj  is the stoichiometric coefficient of species j, Dj  is the diffusion coefficient of the electroactive species j [m2 s-1], Cj is the bulk reactant concentration of species j [kmol m-3], Km,j is the mass transfer coefficient of species j [m s-1], δ is the diffusion layer thickness [m], ν is the kinematic viscosity of the electrolyte [m2 s-1] and ω is the angular velocity [rad s-1]. The RDE measurement is a commonly employed technique to determine number of exchanged electrons as well as kinetic parameters, i.e. Tafel slope and exchange current density, for an electrochemical reaction like ORR which uses Levich equation as follows [218, 219]:  𝑖𝐿,𝑗 = 0.62 ∙𝑛 ∙ 𝐹𝑠𝑗∙ 𝐷𝑗2/3∙ 𝜈−1/6 ∙ 𝜛1/2 ∙ 𝐶𝑗 = 𝐵𝐿,𝑗 ∙ 𝜛1/2 (50)  The Levich equation shows that 𝑖𝐿,𝑗  increases linearly with 𝜛1/2 . Hence, the number of electrons involved (𝑛) or the diffusion coefficient of species j (𝐷𝑗) can be calculated from the slope of experimentally observed limiting current density (iL) versus the square root of angular velocity (𝜛1/2), i.e. Levich slope (BL,j). 233  Kinetic parameters, e.g. Tafel slope and exchange current density, can be further extracted from the polarization curves by employing the reciprocal equation for the mixed control region (kinetic and mass transfer control) and substituting Levich equation in it [218, 219]:  1𝑖𝑗=1𝑖𝑘.𝑗+1𝑖𝐿,𝑗=1𝑖𝑘.𝑗+1𝐵𝐿,𝑗 ∙ 𝜛1/2 (51)  where ij is the measured current density of species j [A m-2] and ik.j is the pure electrode kinetic current density of species j [A m-2].  With the collection of a series of experimental data points from the polarization curves, i.e. 𝑖𝑗 at different potentials in the mixed control region for various rotation speeds, 1𝑖𝑗 vs. 𝜛−1/2 (Koutecky-Levich plot) can be plotted at different potentials in the mixed control region. The slope and intercept of the linear Koutecky-Levich plot allows the calculation of number of electrons (𝑛) and the purely electrode kinetic current density (𝑖𝑘.𝑗), respectively, at various potentials.  An alternative way for understanding the ORR pathway on different catalysts is the rotating ring disk electrode (RRDE) tests [218]. The relative formation rate of hydrogen peroxide ions (eq. 9) and hydroxyl ions (eq. 2) can lead to the number of electrons involved in the ORR on each catalyst surface [218, 285, 286]. This rate can be quantitatively determined by setting an oxidative potential (about 1353 mV) at the ring where the oxidation of the HO2−, formed by O2 reduction on the disk electrode in the first place, is diffusion limited [218, 285, 286]. The selectivity of catalyst toward hydrogen peroxide ion formation (%HO2−) is calculated by:  234  %𝐻𝑂2− =21 + (𝐼𝑑𝐼𝑟× 𝑁)× 100 (52)  where Id, Ir and N refer to the disk current (A), ring current (A) and collection efficiency of the ring,  respectively [285-287]. Then, the number of electrons transferred during ORR (n) can be calculated by:  𝑛 =41 + (𝐼𝑟𝐼𝑑 ×𝑁) (53)      235  Appendix C  : Break-in protocol test results for GDE flow-by cell  As mentioned in section 2.3, the ORR performance and stability of oxide catalysts developed here were tested in commercial scenarios using a galvanostatic test protocol (i.e. chronopotentiometry) in a flow-by test cell with air (CO2 removed) or oxygen flowing through the gas chamber. Prior to the reported electrocatalytic performance tests (Figure 3.8), each fresh electrode was subjected to a break-in polarization protocol composed of 24 hrs of galvanostatic polarization at constant cathodic current densities, i.e. -33 mA cm-2 in air and -67 mA cm-2 in oxygen. Figure C.1 shows the performance of each fresh GDE during the twenty-four-hour-long break-in protocol with air or oxygen flowing through the flow-by cell. 236   Figure C.1 Long-term ORR durability testing of fresh GDEs containing Pt, MnO2, MnO2-LaCoO3, MnO2-LaNiO3 and MnO2-Nd3IrO7 catalysts compared with commercial MnOx GDE from Gaskatel GmbH: A) With air at -33 mA cm-2 and B) With oxygen at -67 mA cm-2. IR-corrected galvanostatic polarization curves obtained in 11.7 M (45 wt%) KOH at 323 K for 24 hrs with either air (CO2 removed) or oxygen flowing through the gas chamber of a flow-by cell from Gaskatel GmbH as twenty-four-hour-long break-in protocol. The absolute gas pressure and flow rate were fixed at 1 atm and 1.51×10-3 SLPM. The catalyst(s) loadings were 2 mg cm-2 each (except for Pt with 0.5 mg cm-2) with final weight ratio of 1:1:2:0.6:0.6 for MnO2 or Pt:co-catalyst (if present):Vulcan XC-72:Nafion:PTFE in the catalyst layer. The catalyst loading for the commercial MnOx from Gaskatel was 20 mg cm-2. The standard error of the mean calculated based on min. two replicates is indicated for each data point.237  Appendix D  : The effect of inter-stage OCP activation In order to further investigate the “healing effect”, i.e. ORR and OER performance recovery of degraded mixed-oxide catalysts by potassium ion intercalation, a fresh MnO2-LaCoO3 exposed to 6 M KOH for six days at 313 K and 400 rpm, was subjected to three hundred potential cycles between 633 to 1483 mV with twelve-hour intervals after each one hundred cycles while resting at open-circuit in 6 M KOH solution at 293 K and 400 rpm. The ORR and OER electrocatalytic activities of K+ activated MnO2-LaCoO3 electrode after durability tests followed by inter-stage activation are shown in Figure D.1. As mentioned in 4.2.1.2, the ORR electrocatalytic activity of K+ activated MnO2-LaCoO3 drops drastically after being extensively cycled for one hundred cycles, i.e. almost 65% of loss in the ORR current density (at 683 mV) after one hundred potential cycles (Figure D.1-A). The so-called inter-stage OCP activation process, i.e. 12 hrs of rest time at open-circuit in 6 M KOH at 293 K and 400 rpm, significantly enhances the ORR performance of the degraded MnO2-LaCoO3 sample by increasing the ORR current density for 126% from -7.2 to -16.8 mA cm-2 at 683 mV (Figure D.1-A). This recovers the ORR electrocatalytic activity of degraded MnO2-LaCoO3 catalyst to some extend and bring its polarization curve close to that of K+ activated MnO2-LaCoO3 electrode at cycle one, i.e. almost identical current densities up to 783 mV (Figure D.1-A). Although the second round of durability test for one hundred cycles severely decreases the ORR electrocatalytic activity of the healed MnO2-LaCoO3 electrode, i.e. almost 60% drop in the ORR current density from -16.8 mA cm-2 (cycle 100) to -6.6 mA cm-2 (cycle 200) at 683 mV, the second twelve-hour-long inter-stage OCP activation, enhances the ORR current density of MnO2-LaCoO3 electrode at cycle one hundred by 120% from -6.6 to -14.6 mA cm-2 (at 683 mV) (Figure D.1-A). The same trend is also observed for the third round of durability test followed by inter-stage OCP activation process: First, decreasing the ORR current density of 238  MnO2-LaCoO3 electrode by almost 62% from -14.6 mA cm-2 (cycle two hundred) to -5.6 mA cm-2 (cycle three hundred) at 683 mV, and second, increasing the current density of MnO2-LaCoO3 electrode at cycle three hundred to -12.3 mA cm-2 at 683 mV, i.e. 120% increase in the ORR performance after third round of inter-stage OCP activation (Figure D.1-A). The total loss in the ORR current density of the K+ activated mixed-oxide catalyst after three hundred cycles of durability tests and subsequent inter-stage OCP activation process is found to be about 41% at 683 mV (Figure D.1-A)  In the OER region, unlike ORR, the K+ activated MnO2-LaCoO3 catalyst exhibits more stable OER performance during three hundred cycles of durability testing and three rounds of inter-stage OCP activation. The most drastic drop in OER electrocatalytic activity of K+ activated MnO2-LaCoO3 catalyst during the three durability stages corresponds to the third round of durability testing (Figure D.1-B), i.e. almost 20% drop in the OER current density from 16.5 mA cm-2 (cycle two hundred) to 13.1 mA cm-2 (cycle three hundred) at 1483 mV. The total OER performance drop for the K+ activated MnO2-LaCoO3 catalyst after three hundred cycles of durability tests and three rounds of inter-stage OCP activation is maximized at 27% (at 1483 mV).  The results here further confirm the reproducibility of “healing effect” by K+ activation of degraded mixed-oxide catalysts. This could be explained by constant formation of Schottky defects and lattice oxygen vacancies during ORR on the oxides’ surfaces which provides the ideal sites for potassium ions to intercalate. The intercalated potassium ions could introduce new binding sites for oxygen catalysis intermediates such as HOO(ads) and HO(ads) and hence, manipulating the ORR/OER electrocatalytic activities of degraded catalysts.  239    Figure D.1 The effect of inter-stage OCP activation: ORR and OER polarization curves of the activated MnO2-LaCoO3 electrodes tested for three hundred cycles with OCP activation in between, A) ORR polarization curves, B) OER polarization curves. The samples are kept at OCP in 6 M KOH solution for 12 hrs at 293 K and a rotation speed of 400 rpm after each one hundred cycles of durability testing. Other conditions are same as Figure 3.6. 240  Appendix E  : Factorial design study Pareto plots of estimates The JMP software employs Pareto plot of estimates to demonstrate the main and interaction effects of various factors on the responses that are being investigated in a factorial design study. In the factorial design investigations of the current work, i.e. 24-1+3 half-fraction 2n factorial design studies in presence of various surfactants, these plots are powerful tools for judging the significance of the effect that each factor or two-factor interaction can have on the final responses, hence help remove the insignificant two-factor interaction in each aliased pair as explained earlier in the section 5.2.3. To calculate the main (Xi) or two-factor interaction (XiXj) effects in each surfactant category, the following equations have been employed:   𝑀𝑎𝑖𝑛 𝑒𝑓𝑓𝑒𝑐𝑡 𝑜𝑓 𝑋𝑖 =∑(𝑅𝑒𝑠𝑝𝑜𝑛𝑠𝑒 𝑎𝑡 ℎ𝑖𝑔ℎ 𝑋𝑖) − ∑(𝑅𝑒𝑠𝑝𝑜𝑛𝑠𝑒 𝑎𝑡 𝑙𝑜𝑤 𝑋𝑖)(𝐻𝑎𝑙𝑓 𝑡ℎ𝑒 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑓𝑎𝑐𝑡𝑜𝑟𝑖𝑎𝑙 𝑟𝑢𝑛𝑠) (54) 𝐼𝑛𝑡𝑒𝑟𝑎𝑐𝑡𝑖𝑜𝑛 𝑒𝑓𝑓𝑒𝑐𝑡 𝑜𝑓 𝑋𝑖𝑋𝑗 =∑(𝑅𝑒𝑠𝑝𝑜𝑛𝑠𝑒 𝑎𝑡 ℎ𝑖𝑔ℎ 𝑋𝑖𝑋𝑗) − ∑(𝑅𝑒𝑠𝑝𝑜𝑛𝑠𝑒 𝑎𝑡 𝑙𝑜𝑤 𝑋𝑖𝑋𝑗)(𝐻𝑎𝑙𝑓 𝑡ℎ𝑒 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑓𝑎𝑐𝑡𝑜𝑟𝑖𝑎𝑙 𝑟𝑢𝑛𝑠) (55)  The levels of two-factor interactions (XiXj) for each factorial run in eq. 55 are determined based on the levels of each individual factor, i.e. Xi and Xj. For example, high level of factor Xi (“+”) and low level of factor Xj (“-”) lead to low level of their two-factor interaction XiXj (“-”). Figure E.1 shows the estimates of effects that each main factor and aliased two-factor interaction pair have on the ORR mass activity of the electrodeposited MnOx in presence of Triton X-100 as a response for the factorial design study. The graph shows clearly that between the four main factors (see Table 2.1), surfactant concentration (S), Temperature (T) and applied anodic 241  potential (E) have the most significant effect on the response while Mn concentration (C) is the most insignificant factor in this case for the defined response (Figure E.1). Employing the Ockham’s razor principle, one can assume the most significant two-factor interactions are S×T, S×E and E×T among the C×E+S×T, C×T+S×E and C×S+E×T aliased pairs, respectively, since Mn concentration is shown to have the least effect on the ORR mass activity of the electrodeposited manganese oxides. Hence, the surface plots can be easily constructed using the most significant three main factors and three two-factor interactions. This is similar to the case that Mn concentration is discarded from the beginning and a 23+3 full factorial design with three factors and three center-points leading to 11 runs in total is being constructed for the Triton X-100 samples.       242   Figure E.1 Pareto plot of estimates for the effects of four main factors (i.e. surfactant concentration (S), temperature (T), Mn concentration (C) and applied anodic potential (E)) and three aliased two-factor interactions on the ORR mass activity in the 24-1+3 factorial design study on the anodic electrodeposition of MnOx in presence of Triton X-100.  Curvature effect Traditionally, the curvature effect is estimated as the difference between the average of the center-point responses and the average of the factorial points. A strong curvature effect is then a reflection of a non-linear system behavior. JMP demonstrates the curvature effect in terms of “RSquare” when center-points are added to the factorial design study. RSquare estimates the proportion of variation in the response that can be attributed to the model rather than to random error. RSquare (also called the coefficient of multiple determination in the JMP software) is calculated as: 243   𝑅𝑆𝑞𝑢𝑎𝑟𝑒 =𝑆𝑢𝑚 𝑜𝑓 𝑠𝑞𝑢𝑎𝑟𝑒𝑠 (𝐶.  𝑡𝑜𝑡𝑎𝑙) − 𝑆𝑢𝑚 𝑜𝑓 𝑠𝑞𝑢𝑎𝑟𝑒𝑠 (𝐸𝑟𝑟𝑜𝑟)𝑆𝑢𝑚 𝑜𝑓 𝑠𝑞𝑢𝑎𝑟𝑒𝑠 (𝐶.  𝑡𝑜𝑡𝑎𝑙) (56)  where “sum of squares (C. total)” is the sum of the squared differences between the response values and the sample mean (representing the total variation in the response values) and “sum of squares (Error)” is the sum of the squared differences between the fitted values and the actual values (representing the variability that remains unexplained by the fitted model). A RSquare closer to 1 indicates a better fit to the data than does a RSquare closer to 0, meaning the curvature effect is negligible. A RSquare near 0 indicates that the model is not a much better predictor of the response than is the response mean, meaning a degree of non-linearity in the behavior of variables. 

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}]}"
                            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.24.1-0340612/manifest

Comment

Related Items