UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Electrochemical recycling of sodium borohydride for hydrogen storage : physicochemical properties of… Cloutier, Caroliine R. 2006

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

Item Metadata

Download

Media
831-ubc_2006-0411.pdf [ 13.13MB ]
Metadata
JSON: 831-1.0092686.json
JSON-LD: 831-1.0092686-ld.json
RDF/XML (Pretty): 831-1.0092686-rdf.xml
RDF/JSON: 831-1.0092686-rdf.json
Turtle: 831-1.0092686-turtle.txt
N-Triples: 831-1.0092686-rdf-ntriples.txt
Original Record: 831-1.0092686-source.json
Full Text
831-1.0092686-fulltext.txt
Citation
831-1.0092686.ris

Full Text

E L E C T R O C H E M I C A L R E C Y C L I N G OF SODIUM BOROHYDRTDE FOR H Y D R O G E N STORAGE: PHYSICOCHEMICAL PROPERTIES OF SODIUM M E T A B O R A T E SOLUTIONS. by Caroline R. Cloutier B. A . Sc., The University of Ottawa, 1999 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF M A S T E R OF APPLIED SCIENCE m THE F A C U L T Y OF G R A D U A T E STUDIES (Materials Engineering) THE UNIVERSITY OF BRITISH C O L U M B I A July 2006 © Caroline R. Cloutier, 2006 ABSTRACT The large-scale adoption of a "hydrogen economy" is hindered by the lack of a practical storage method and concerns associated with its safe handling. Chemical hydrides have the potential to address these concerns. Sodium borohydride (sodium tetrahydroborate, NaBH4), is the most attractive chemical hydride for H 2 generation and storage in automotive fuel cell applications but recycling from sodium metaborate (NaB02) is difficult and costly. An electrochemical regeneration process could represent an economically feasible and environmentally friendly solution. In this thesis, the properties of diluted NaB0 2 aqueous solutions and concentrated NaB0 2 alkaline aqueous solutions that are necessary for the development of electrochemical recycling methods have been studied. The conductivity and viscosity of dilute aqueous solutions of NaB0 2 were measured as a function of concentration at 25°C. Also, the solubility, pH, density, conductivity and viscosity of the filtrate of saturated aqueous NaB0 2 solutions containing varying weight percentages (1, 2, 3, 5, 7.5 and 10 wt%) of alkali hydroxides (NaOH, KOH and LiOH) were evaluated at 25°C. Selected experiments were repeated at 50 and 75°C to investigate the effect of temperature on the NaB0 2 alkaline aqueous solution solubility and physicochemical properties. Preliminary experiments to investigate the effect of glycine (C 2H 5N0 2), the smallest amino acid, on the solubility and physicochemical properties of NaB0 2 alkaline aqueous solutions were conducted at 25°C. Furthermore, the precipitates formed in the supersaturated 10 wt% alkaline aqueous NaB0 2 solutions at 25°C were characterized by X-Ray Diffraction and Scanning Electron Microscopy. The use of KOH as the electrolyte was found to be more advantageous for the H 2 storage and generation system based on NaB0 2 solubility and NaBH 4 half-life due to the pH effect. However, the addition of NaOH led to the highest ionic conductivity, and its use seems more ii suitable for the electroreduction of NaB0 2. Further investigations on the impact of KOH and NaOH on the electroreduction of NaB0 2 in aqueous media have the potential to enhance the commercial viability of this H 2 generation and storage system. ii i T A B L E O F C O N T E N T S Abstract i i Table of Contents iv List of Tables ix List of Figures xi List of Symbols xiii List of Acronyms xiv Acknowledgement xvi Dedication xvii C H A P T E R I Introduction 1 1.1 Motivation 1 1.2 Objectives 2 1.3 Approach 3 1.4 Overview 4 C H A P T E R II Background 5 2.1 N a B H 4 Production 5 2.1.1 Natural Resources 5 2.1.2 Current Production Process 5 A) Production Paths 5 B) The Schlesinger and Brown Process 6 C) The Bayer Process 8 2.1.3 Economics 8 A) Statistics 9 B) Borate Usage 9 C) NaBH4Cost •.. 11 2.2 Regeneration 12 2.2.1 Chemical Synthesis 12 A) Modified Schlesinger and Brown Process 13 B) Modified Schlesinger and Brown Process: Amendment 15 C) Three-Step Regeneration 16 D) Using Coke or Methane 17 E) Recycling Through Diborane 18 F) Reduction of NaB0 2 with H 2 19 G) Methods Using Metal Hydrides 19 a) Annealing 19 b) Ball Milling 20 H) Radiolysis Process 21 2.2.2 Electrochemical Synthesis 22 A) Direct Electroreduction 23 B) Kinetic Considerations 25 C) Patent Review 27 D) Reproducibility 31 E) Indirect Electroreduction 32 F) Electroreduction of Borate Ester 33 G) Electroreduction in Molten Hydroxide Media 34 iv H) Hydrogen Assisted Electrolysis 34 a) Molten NaOH Electrolysis 34 b) H 2 Assisted Molten Salt Electroreduction of Borate 36 c) Aqueous Alkaline Borate Solutions 38 I) Metathesis 38 J) Simulating Conventional Synthesis 39 2.2.3 Infrastructure 40 2.3 NaBH 4 H 2 Generation and Storage System Efficiency 41 2.3.1 Gravimetric Capacity 41 2.3.2 Volumetric Capacity 42 2.3.3 Fuel Comparison 43 2.4 Hydrolysis of NaBH 4 for H 2 Generation and Storage 44 2.4.1 Hydrolysis 45 2.4.2 Effect of pH 46 A) pH Equilibria of Borates 46 B) Effect of pH on the Solution Stability 48 C) Effect of Electrolyte Concentration on the NaBH4 Hydrolysis 49 2.4.3 Effect of Catalysts 50 A) Noble Metal Catalysts 50 a) Supported Platinum 50 b) Supported Ruthenium 51 B) Non-Precious Transition Metal Catalysts 52 a) Cobalt Boride 52 b) Nickel Boride 53 c) Filamentary Nickel Mixed Cobalt 54 d) Metal Hydride 54 e) Metal-Metal Oxide 55 C) Organic Catalysts 55 D) Catalysts Comparison 56 2.4.4 Effect of Temperature 58 2.4.5 Effect of N a B H 4 Concentration 59 2.4.6 Steam Hydrolysis of N a B H 4 59 2.5 NaBH 4 Applications 60 2.5.1 N a B H 4 Hydrolysis Reactors 60 A) Kipp Generator 60 B) Remote Fuel-Cell Power Source 62 C) Solid Hydride and Liquid Water or Water Vapour 62 D) Solid Hydride and Steam 63 E) Catalytic Reactors 64 2.4.2 B-PEMFC System 66 A) Hydrogen on Demand™ 66 2.4.3 Vehicle Prototype 67 A) B-PEMFC Vehicle 68 C H A P T E R III Physicochemical Properties of N a B 0 2 69 3.1 N a B 0 2 Solution Characteristics 69 3.1.1 Solubility 69 v A) NaB0 2 Solubility 69 B) NaBH, Solubility 71 C) Effect of Temperature on the Solubility of NaB0 2 71 3.1.2 Single 1:1 Electrolyte Theory 72 A) Conductivity 73 B) Dynamic Viscosity 76 a) Effect of Temperature on the Viscosity and Ionic Mobility 77 b) Effect of Temperature on the B Coefficients 78 C) Walden's Rule 79 3.1.3 Physicochemical Properties 80 A) Aqueous NaB0 2 Solutions 80 B) Alkaline Aqueous NaB0 2 Solutions 82 3.1.4 Organic additives 83 A) Amides 83 a) Urea '. 83 b) Thiourea 84 c) Acetamine 85 B) Ammonia 85 C) Diethylene glycol 86 D) Glycine 86 E) Organic Additive Comparison 87 3.2 Precipitate Characteristics... 88 CHAPTER IV Experimental Methods 90 4.1 Materials 90 4.1.1 Properties 90 4.1.2 Contamination 91 4.2 Experimental Plan 92 4.3 Solution Preparation 93 4.3.1 Dilute Aqueous N a B 0 2 Solutions 93 4.3.2 Concentrated Alkaline Aqueous N a B 0 2 Solutions 93 4.3.3 Organic Additives 94 4.4 Filtration 95 4.4.1 Filtration at 25°C 94 A) Vacuum Filtration 95 B) Syringe Filtration 96 4.4.2 Filtration at 50 and 75°C 97 4.5 Filtrate Characterization 98 4.5.1 Solubility 98 A) Boron Detection Methods 98 B) Induced Coupled Plasma Mass Spectrometry 99 4.5.2 pH 99 A) Glass Electrode :'. „ 100 B) pH Meter 100 4.5.3 Conductivity 100 4.5.4 Viscosity 101 4.5.5 Density 102 vi 4.6 Precipitate Characterization 102 4.6.1 X-Ray Diffraction 102 4.6.2 Scanning Electron Microscopy .. 103 CHAPTER V Results and Discussion. 104 5.1 Aqueous NaB0 2 Solutions Physicochemical Properties 104 5.1.1 Solubility 104 5.1.2 Transport Properties 105 A) Conductivity 105 B) Viscosity 107 5.2 Aqueous NaB0 2 Solutions Physicochemical Properties with Alkali Additive 109 5.2.1 Solubility 109 5.2.2 pH I l l A) NaBH4 Solution Half-Life 112 5.2.3 Transport Properties 113 A) Ionic Conductivity 113 a) Effect of Hydroxide Concentration on the Conductivity of Saturated NaB0 2 Solutions 113 b) Unsaturated NaB0 2 Alkaline Aqueous Solution 114 B) Molar Conductivity 115 a) Unsaturated NaB0 2 Alkaline Aqueous Solution 115 b) Effect of Hydroxide Concentration on the Molar Conductivity of Saturated NaB0 2 Solutions 115 C) Dynamic Viscosity 117 D) Walden Product 119 5.2.4 Specific Gravity 120 5.3 Effect of Temperature 121 5.3.1 Solubility 122 5.3.2 Transport Properties 123 A) Ionic Conductivity 123 a) Effect of Hydroxide Concentration on the Conductivity of Saturated NaB0 2 Solutions 123 b) Unsaturated NaB0 2 Alkaline Aqueous Solution 124 B) Molar Conductivity 126 a) Unsaturated NaB0 2 Alkaline Aqueous Solution 126 5.4 Effect of Glycine Additions 126 5.4.1 Solubility 127 5.4.2 pH 127 A) NaBH 4 Solution Half-Life 128 5.4.3 Transport Properties 129 A) Ionic Conductivity 129 B) Molar Conductivity 130 C) Dynamic Viscosity 131 5.4.4 Specific Gravity 131 5.5 Precipitate Characteristics 132 5.5.1 X-Ray Diffraction 132 vii 5.5.2 Scanning Electron Microscopy 134 CHAPTER VI Conclusions 136 6.1 Conclusions 136 6.1.1 System Considerations 137 6.1.2 Future Research and Recommendations 138 6.2 Summary of Contributions 140 References 141 APPENDIX 150 Appendix A: Comparison to DOE Range, Storage Cost, Energy Density and Specific Energy Targets 150 Appendix B: Filtrate Characterization Data 152 Appendix C: Precipitate XRD Data 157 Appendix D: Precipitate SEM Pictures 159 viii LIST OF TABLES Table 1.1: US Department of Energy Freedom Car Targets for On-Board H 2 Storage, (Modified from [Department of Energy, (2004)]) 3 Table 2.1: Partial List of Borate Uses (Modified from [Garret, D. E., (1998)]) 10 Table 2.2: List of Possible Chemical Regeneration Synthesis 13 Table 2.3: Chemical Synthesis Patent List 13 Table 2.4: List of Possible Electrochemical Regeneration Synthesis 22 Table 2.5: Electrochemical Borohydride Synthesis Patent List 23 Table 2.6: Exchange Current Densities (io) and Tafel Slopes for the H 2 Evolution Reaction at Various Materials in Basic Solutions at Room Temperature., *313K [Gouper, A.M. et al, (1990)]... 26 Table 2.7: H 2 Content of Various Chemical Hydrides 41 Table 2.8: Volume Required for the Storage of 1 kg of H 2 from Different Borohydrides 43 Table 2.9: Comparison of Some Catalyst Activity (Modified from [Kaufman, C. M. et al, (1985), Amendola, S. C. et al., (2000a)] 51 Table 2.10: Catalysts for NaBH4 Hydrolysis 57 Table 3.1: NaBH4 Solubility in Various Solvents 71 Table 3.2: Solubility of NaB0 2 in Water at Selected Temperatures 72 Table 3.3: Specific Gravity and Conductivity of 5.4 and 20.2 wt% NaB0 2 Aqueous Solutions (Modified from [Maksimova, I. N. etal, (1963)]) 81 Table 3.4: Conductivity at Infinite Dilution at Various Temperatures (Modified from [Maksimova, I. N. et al, (1963)]) 82 Table 3.5: Molar Conductivity Ratios at Varying NaB0 2 Concentrations (Modified from [Maksimova, I. N. et al, (1963)]) 82 Table 3.6: Apparent Mobility of the Borate Anion (Modified from [Frei, V. er al, (1963)]) 88 Table 4.1: Anion Properties 90 Table 4.2: Cation Properties 91 Table 4.3: Thermodynamic Data 91 Table 4.4: Deionized Distilled Water ICP Analysis 92 Table 4.5: Filtration Method Comparison 97 Table 4.6: List of Some Boron Detection Methods 99 Table 5.1: Comparison of NaB0 2 Solubility wt% in Water at 25°C : 104 Table 5.2: Comparison of Limiting Molar Conductivity Predictions at 25°C ± 3°C 106 Table 5.3: Infinite Dilution Transport Properties for NalB02" at 25°C ± 3°C 107 Table 5.4: Effect of Glycine on the Filtrate Solubility as a Function of Hydroxide Content at 25°C ± 3°C : . 127 Table 5.5: Effect of Glycine on the Filtrate pH as a Function of Hydroxide Content at 25°C ± 3°C... 128 Table 5.6: Effect of Glycine on the Filtrate Ionic Conductivity as a Function of Hydroxide Content at25°C±3°C 130 Table 5.7: Effect of Glycine on the Filtrate Molar Conductivity as a Function of Hydroxide Content at25°C±3°C 130 Table 5.8: Effect of Glycine on the Filtrate Dynamic Viscosity as a Function of Hydroxide Content at25°C±3°C 131 Table 5.9: Effect of Glycine on the Filtrate Specific Gravity as a Function of Hydroxide Content at 25°C±3°C 132 Table 6.1: Variable Effect on Various Parameters of the NaBH 4 H 2 Generation and Storage System (+ increase, - decrease, op=optimum point, n=none) 136 Table 6.2: Summary of Alkali Hydroxide Additive Effect on the H 2 Storage and Generation System and Electrochemical Recycling 137 Table A. 1: Fuel Cost for Target Range and Storage Cost Estimation 150 Table A.2: Volumetric and Gravimetric Energy Density Comparison 151 Table B.l: Physicochemical Properties Raw Data, Unsaturated Diluted Aqueous Solutions of NaB0 2 at 25°C ± 3°C, Trial 1 152 Table B.2: Physicochemical Properties Raw Data, Unsaturated Diluted Aqueous Solutions of NaBQ2 at 25°C ± 3°C, Trial 2 152 IX Table B.3: Physicochemical Properties Raw Data, Unsaturated Diluted Aqueous Solutions of NaB0 2 at 25°C ± 3°C, Trial 3 152 Table B.4: Summary of Averages, Conductivity of Unsaturated NaB0 2 Aqueous Alkaline Solutions Filtrate at 25, 50 and 75°C 153 Table B.5: Filtrate Physicochemical Properties Raw Data, Saturated Aqueous Solutions of NaB0 2 at25°C±3°C 154 Table B.6: Filtrate Physicochemical Properties Raw Data, Saturated Aqueous Solutions of NaB0 2 stabilized with 7.5 wt% KOH at 25°C ± 3°C 154 Table B.7: Summary of Averages, Effect of Alkali Hydroxide Addition on the Physicochemical Properties of Saturated NaB0 2 Aqueous Solutions Filtrate at 25°C ± 3°C 155 Table B.8: Summary of Averages, Effect of Temperature on the Physicochemical Properties of Saturated NaB0 2 Aqueous Alkaline Solutions Filtrate at 25, 50 and 75°C 156 Table B.9: Summary of Averages, Effect of Glycine on the Physicochemical Properties of Saturated NaB0 2 Aqueous Alkaline Solutions Filtrate at 25°C ± 3°C 156 Table C.l: Summary of Three Most Important XRD Peaks for Various Boron Compounds from Different Sources 157 Table C.2: Summary of Three Most Important XRD Peaks for the Precipitates of Saturates Alkaline Aqueous Solutions of NaB0 2 at 25°C 158 x LIST OF FIGURES Figure 2.1: Process Flow Diagram of the Schlesinger and Brown Process (Modified from [Wu, Y. etal, (2004b)]) 7 Figure 2.2: Schematic Diagram of the Three-Step NaBH4 Regeneration Process (Modified from [Mohring, R. M. et al, (2003)]) 16 Figure 2.3: Schematic of Electrolytic Cell for the Direct Reduction of Borates to Borohydrides: 1) Cathode, 2) Catholyte, 3) Membrane, 4) Anode, 5) Anolyte, 6) Electric Cell, 7) Wire, 8) Aqueous Acid Solution, 9) Borate Aqueous Solution, 10) Borohydride Withdrawal Line, (Modified from [Hale, C.H., (1990)]) 30 Figure 2.4: Hydrogen Electrolytic Synthesis of NaOH in Salt Melt 1) Cathode, 2) Catholyte, 3) Membrane, 4) Anode, 5) Anolyte, 6) Hydrogen sparger, 7) Molten sodium, (Modified from [Xu, J. etal, (2004)]) 35 Figure 2.5: Hydrogen Electrolytic Synthesis of NaBH4 in Salt Melt 1) Cathode, 2) Catholyte, 3) Membrane, 4) Anode, 5) Anolyte, 6) Hydrogen sparger (Modified from [Xu, J. et al, (2004)]) 37 Figure 2.6: NaBH4 Fuel Infrastructure Schematic Diagram (Modified from [MERIT, Ltd., (1986)] 40 Figure 2.7: Comparison of the Gravimetric and Volumetric Energy Densities of Various Fuels and H 2 carriers (Modified from [Chandra, D. et al., (2006)]) 44 Figure 2.8: Schematic Diagram of a Kipp Generator (Modified from [Mattson, B., (2005)]) 61 Figure 2.9: Schematic of Remote Hydride System (Modified from [Linden, D., (1984)]) 62 Figure 2.10: Schematic of Hydride Reactor (a) with Liquid Water, (b) with Water Vapor (Modified from [Kong, V. C. Y. et al, (1999)]) 63 Figure 2.11: Schematic of Steam Hydrolysis Reactor System (Modified from [Linden, D., (1984)]). 64 Figure 2.12: Reactor for Releasing H 2 from an Alkaline Aqueous Solution of NaBH4 (Modified from [Larminie, J. et al, (2003)]) 65 Figure 2.13: Schematic of a H 2 Generation Vessel with Catalytic Control of Evolution Rate (Modified from [Linkous, C. A. et al, (2004)]) 65 Figure 2.14: Schematic Diagram of the Hydrogen on Demand™ System (Modified from [Mohring, R. M. et al, (2003)]) 67 Figure 2.15: Daimler Chrysler Natrium Vehicle Power Train (Modified from [Wu, Y., (2003a)]).... 68 Figure 4.1: Molecular Chain Arrangement of Glycine 91 Figure 4.2: Experimental Plan Schematic Diagram 92 Figure 4.3: Vacuum Filtration Apparatus 95 Figure 4.4: Syringe Filtration Apparatus: (a) Schematic Diagram (b) Full View 96 Figure 4.5: pH and Conductivity Experimental Set-up 101 Figure 4.6: Viscometer Experimental Set-up 102 Figure 5.1: Molar Conductivity as a Function of the Square Root of the Concentration at 25°C ± 3°C 106 Figure 5.2: Relative Viscosity as a Function of the Square Root of the Concentration at 25°C ± 3°C 108 Figure 5.3: NaB0 2 Solubility as a Function of Hydroxide Content at 25°C ± 3°C 110 Figure 5.4: Filtrate pH as a Function of Hydroxide Content at 25°C ± 3°C I l l Figure 5.5: Estimated Half-Life of Alkaline Aqueous Solutions of NaBH4 as a function pH at 25°C ± 3°C using Eq. 2.68 112 Figure 5.6: NaB0 2 Ionic Conductivity in the Filtrate as a Function of Hydroxide Content at 25°C ± 3°C 113 Figure 5.7: Ionic Conductivity as a Function of Hydroxide Content of the Unsaturated 5 wt% NaB0 2 Aqueous Solution at 2 5°C ± 3°C 114 Figure 5.8: Molar Conductivity as a Function of Hydroxide Content of the Unsaturated 5 wt% NaB0 2 Aqueous Solution at 25°C ± 3°C 115 XI Figure 5.9: Molar Conductivity in the Filtrate as a Function of Hydroxide Content at 25°C ± 3°C.... 115 Figure 5.10: Filtrate Dynamic Viscosity as a Function of Hydroxide Content at 25°C ± 3°C 118 Figure 5.11: Filtrate Walden Product as a Function of Hydroxide Content at 25°C ± 3°C 120 Figure 5.12: Filtrate Specific Gravity as a Function of Hydroxide Content at 25°C ± 3°C 121 Figure 5.13: NaB0 2 Solubility as a Function of Hydroxide Content at 25, 50 and 75°C 123 Figure 5.14: NaB0 2 Ionic Conductivity in the Filtrate as a Function of Hydroxide Content at 25, 50 and75°C 124 Figure 5.15: Ionic Conductivity as a Function of NaOH of the Unsaturated 5 wt% NaB0 2 Aqueous Solution at 25, 50 and 75°C 125 Figure 5.16: Ionic Conductivity as a Function of KOH of the Unsaturated 5 wt% NaB0 2 Aqueous Solution at 25, 50 and 75°C 125 Figure 5.17: Molar Conductivity as a Function of Hydroxide Content of the Unsaturated 5 wt% NaB0 2 Aqueous Solution at 25, 50 and 75°C 126 Figure 5.18: Effect of Glycine on the Estimated Half-Life of Alkaline Aqueous Solutions as a function pH at 25°C ± 3°C using Eq. 2.68 - 129 Figure 5.19: Intensity as a Function of Diffraction Angle for Pure NaB0 2.2H 20 and Precipitates of 10 wt% Hydroxide NaB0 2.2H 20 Supersaturated Aqueous Solutions 133 Figure 5.20: Scanning Electron Microscope Precipitate Crystal Images, (a) Pure NaB0 2.2H 20 Flakes (SE, WDM. 1mm, x45, 1mm), (b) NaB02.2H20 Crystal (SE, WD15.8mm, xlO.Ok, 5um), (c) 10 wt% NaOH, Saturated NaB0 2.2H 20 Aqueous Solution Precipitate (BSE2, WD14mm, x2.0k, 20um), (d) 10 wt% ROH, Saturated NaB02.2H20 Aqueous Solution Precipitate (BSE2, WD 14mm, x500, lOOum), (e) 10 wt% LiOH, Saturated NaB02.2H20 Aqueous Solution Precipitate (BSE2, WD13.4mm, xl.Ok, 50um), (f) 10 wt% LiOH Saturated NaB02.2H20 Aqueous Solution Precipitate (BSE2, WD13.3mm, xl.Ok, 50um) 134 Figure D.l: Scanning Electron Microscope Precipitate Crystal Images of Pure NaB0 2.2H 20 (a) SE, WD27.6mm, x200, 200um, (b) SE, WD25.5mm, x700, 50um, (c) BSE2, WD14.3mm, x200, 200um, (d) BSE2, WD 14.1 mm, x200, 200um, (e) SE, WD26.0mm, x2.5k, 20um, (f) SE, WD26.0mm, x2.5k, 20um 159 Figure D.2: Scanning Electron Microscope Precipitate Crystal Images of Precipitates formed from Saturated Aqueous Solutions of NaB0 2 stabilized with 10 wt% NaOH (a) SE, WD26.2mm, x500, lOOum, (b) SE, WD25.5mm, x700, 50um, (c) SE, WD25.5mm, xl.Ok, 50um, (d) SE, WD26.1mm, x2.0k, 20um, (e) BSE2, WDH.Omm, x5.0k, lOum, (f) BSE2, WD14.0mm, xlOk, 5um 160 Figure D.3: Scanning Electron Microscope Precipitate Crystal Images of Precipitates formed from Saturated Aqueous Solutions of NaB0 2 stabilized with 10 wt% KOH (a) BSE2, WD13.9mm, xl.Ok, 50um, (b) BSE2, WD13.9mm, x3.0K, lOum, (c) SE, WD27.4mm, x500, lOOum, (d) SE, WD27.4mm, xl.Ok, 50um, (e) SE, WD27.1mm, x3.0k, lOum, (f) SE, WD28.1mm, x5.0k, lOum 161 Figure D.4: Scanning Electron Microscope Precipitate Crystal Images of Precipitates formed from Saturated Aqueous Solutions of NaB0 2 stabilized with 10 wt% LiOH (a) BSE2, WD14.1mm, xl.OK, 50um, (b) BSE2, WD14.1mm, x3.0k, lOum, (c) SE, WD29.3mm, x500, lOOum, (d) SE, WD29.4mm, xl.Ok, 50um, (e) BSE2, WD13.7mm, x500, lOOum, (f) BSE2, WD13.7mm, xl.Ok, 50um 162 Figure D.5: Scanning Electron Microscope Precipitate Crystal Images of Over dried Precipitates formed from Saturated Aqueous Solutions of NaB0 2 stabilized with 10 wt% LiOH (a) BSE2, WD 13.6mm, xlOO, 500um, (b) BSE2, WD 13.6mm, x500, lOOum, (c) BSE2, WD 13.6mm, xl.Ok, 50um, (d) BSE2, WD13.6mm, x3.0k, lOum, (e) BSE2, WD13.4mm, xl.Ok, 50um, (f) BSE2, WD13.5mm, x500, lOOum 163 X l l LIST O F S Y M B O L S Symbol Roman Symbols A b B B B E B s C D D D" dH° dG° dS° E° E°a E°c E c Ecell F ^e, cell F F H i i 0 K K a K B K.7 Mr P P R S tl/2 T t°i u Wh Ws y y Greek Symbols Description Viscosity Coefficient Tafel Slope Viscosity Coefficient Molar Conductivity Coefficient Extrapolated Viscosity Coefficient Viscosity Coefficient Molar Concentration Viscosity Coefficient Molar Conductivity Coefficient Limiting Diffusivity of Species i Standard Enthalpy of Formation Standard Gibbs Free Energy of Formation Standard Entropy of Formation Standard Cell Voltage Standard Anode Voltage Standard Cathode Voltage Actual Cathode Voltage Actual Cell Voltage Equilibrium Cathode Voltage Non-standard Equilibrium Cell Voltage Activation Energy of Viscous Flow Faraday Constant Constant Current Density Exchange Current Density Molar Conductivity Coefficient Ion Pair Association Constant Boltzmann Constant Conductivity at Temperature T Molecular Weight Ratio of NaBH4 to NaB0 2 Temperature Pre-exponential Factor Parameter Universal Gas Constant Solubility Half Life Temperature Limiting Transference Number of Species i Mobility Limiting Mobility of Species i Amount of H 2 Generated Amount of NaB0 2 Generated Organic Additive Concentration Molar Concentration Ratio Solute Solubility Units 1/M , / 2 V 1/M S.m2/mol.M 1/M 1/M M 1/M2 S.m2/mol.M3/2 cm2/s KJ/mol KJ/mol J/mol.K V V V V V V V J/mol 964853.34 C/mol Dimensionless A/m 2 A/m 2 S.m2/mol.M1/2 or S.m2/mol.M1/3 1/M J/K S/m Dimensionless Dimensionless Dimensionless J/mol.K wt% days C o r K Dimensionless cm2/s.V cm2/s.V g g wt% Dimensionless wt% a Temperature Coefficient of Conductivity 1/C Cathodic Transfer Coefficient Dimensionless AShyd Standard Entropy of Hydration J/mol.K A O o h m Ohmic Losses V e0 Dielectric Constant of Pure Solvent Dimensionless X l l l K Conductivity S/m Tl Dynamic Viscosity cP ^con Concentration Overpotential V Tls Surface Activation Overpotential V Tlr Relative Viscosity Dimensionless Tlo Viscosity of Pure Solvent cP PT Density of Solution at Temperature T kg/m3 PH20, T Density of Water at Temperature T kg/m3 A Molar Conductivity S.nrVmol A 0 Limiting Molar Conductivity S.m2/mol Limiting Molar Conductivity of Ionic Species i S.m2/mol xiv LIST OF ACRONYMS AA Atomic Absorption B-PEMFC Borohydride - Proton Exchange Membrane Fuel Cell CS Cellulose Ester DBFC Direct Borohydride Fuel Cell DOE Department of Energy EDA Efhylenediamine EDTA Ethylenediaminetetraacetic Acid ETV Electro-Thermal Vaporization FCE Fuel Cell Emulator FCV Fuel Cell Vehicle FCUPS Fuel Cell Uninterrupted Power Supply HMPA Hexamethylphosphoamide HOD Hydrogen On Demand ICP Induced Coupled Plasma MS Mass Spectrometry PEMFC Proton Exchange Membrane Fuel Cell PTFE Polytetrafluorethylene SHE Standard Hydrogen Electrode SEM Scanning Electron Microscopy SS Stainless Steel XRD X-Ray Diffraction XV ACKNOWLEDGEMENT Special acknowledgement is made to my supervisor, Dr. Akram Alfantazi, and co-supervisor, Dr. Elod Gyenge, for their stimulating suggestions, constructive criticism and valuable guidance. I would like to express my gratitude to Dr. David Dreisinger, for being an inspiring professor and to Dr. Mary Wells, for being an exceptional graduate advisor throughout the course of my degree. I wish to expand my sincere thanks to my lab colleagues, especially Edouard Asselin, James Vaughan, John Skrovan, Andy Jung, as well as Muhannad Al-Darbi, Alex Abughusa, Maslat Alwaranbi, Mansour Almansour and Vincent Lam for all the meaningful discussions, to Jennifer Cutting, for her valuable assistance with preliminary experiments, as well as to Dr. Boyd Davis and his research team, particularly Flora Lo, Daniel Calabretta, and Mohamed Abdul at Queen University, for their collaboration on this research project. Many thanks to the Science and Engineering librarians, Aleteia Greenwood and Anne Miele, for their assistance in locating hard to find articles and to Edith Czech and Libin Tong for their valuable help in translating relevant articles. I also want to thank Mary Mager for her effective instruction on the use of the SEM and XRD apparatus, Glenn Smith, at the department store, and the team at the shop, for promptly handling product orders and providing the necessary tools for pursuing my experimental work, and Serge Milaire for solving the odd computer problems. I am very grateful to David Chiu at International Plasma Labs Ltd. for performing ICP analysis on my samples and to Dr. Jonathan Owen at US Borax Inc. for agreeing to provide sodium metaborate for performing this research work. In addition, I am grateful to the Auto 21 Network of Centers of Excellence (NCE) for their funding contribution under project D06-DSB as well as to the Natural Sciences and Engineering Research Council of Canada (NSERC) for providing financial support through a PGS-M scholarship and for their interest in this applied research program. Last but not least, to my parents, family and close friends for their sustained and precious encouragement during the difficult times, a very special appreciation for your tremendous patience and continuous practical and emotional support as I struggled to fulfill the competing demands of my studies and personal life challenges. Your extreme generosity will be remembered always. To each of the above, I extend my deepest respect and appreciation. xvi DEDICATION lleutie^, Jm^ jbmwctm^ me mit/b unemi«dttumal l<we and Auffiowt, and Jtw (jfitwixjfs me tke Afaen^t/i, <mtwa^e and detemrimaM&ri t&fiwwkw ^eafo a/zidjodow m^ dreasmA. xvii CHAPTERI INTRODUCTION 1.1 Motivation It has been proposed that switching to hydrogen (H2) for energy production could alleviate many environmental concerns related to the combustion of fossil fuels [Hart, D., (2003)]. H 2 is an excellent energy carrier which can be produced from a variety of sources such as solar, wind, biomass, hydro, and geothermal. However, there is a long way to go before the cost of H 2 production and distribution to the users will be competitive with that of gasoline. The extensive use of H 2 as a fuel in commercial transportation is delayed by difficulties in developing a practical and efficient H 2 storage method and by concerns over its utilization hazards. The slow development of a suitable H 2 storage system to start the vehicle and rapidly respond to its fuel demand greatly restrains the commercialization of Fuel Cell Vehicles (FVCs). Among the different H 2 storage methods, chemical hydride and light metal complex hydride based H 2 generators have the potential to address these concerns. They possess the characteristics necessary to meet the design requirements of automotive fuel cell applications and do not require handling of gaseous H 2 . They currently constitute the most active area of research for H 2 storage for on-board applications [Dhanesh, C. et al, (2006)]. Various chemical and light metal complex hydrides can produce H 2 via a hydrolysis reaction. Sodium borohydride (sodium tetrahydroborate, NaBH4) is one of the most attractive material for this purpose, as it has a high H 2 content (10.58 wt%), based on the hydride itself and a high energy density (1.4 KWh/1), based on a 20 wt% NaBH4 solution feed to a Proton Exchange Membrane Fuel Cell (PEMFC). When in alkaline solution and in contact with a catalyst, NaBH 4 releases H 2 and produces sodium metaborate (NaB02). NaBH4 hydrogen generators are safe, clean and lightweight systems able to deliver pure H 2 to feed PEMFCs generating electricity in a FCV. The most important obstacles to the acceptance of this system as the H 2 carrier of choice for automotive fuel cells are the high production cost of NaBH4 and the difficult recyclability of the NaBH4 hydrolysis by-product, NaB02, back to NaBH4. 1 Depending on how NaB0 2 is recycled, the resulting overall H 2 generation and storage system energy cycle scheme could represent an emission free and economically viable solution. Hence, the development of an efficient NaB0 2 recycling path is crucial to make NaBH4 play a prevailing role in overcoming the growing energetic demand and environmental challenges of the future. 1.2 Objectives This thesis contributes to the development of an efficient H 2 storage method based on NaBH4 solutions. More particularly, the goal of this work is to address the current barriers hindering the use of NaBH4 for H 2 generation and storage, i.e. recyclability and cost. Currently, the knowledge of the fundamental physicochemical and transport properties of the NaB0 2 solution is very limited and this basic information is essential to the investigation of electrochemical recycling. In this study, the physicochemical properties of various NaB0 2 aqueous solutions with and without alkali additive were evaluated to determine which electrolyte media would be more favourable to the H 2 generation and storage system, as well as for the electrochemical reduction of NaB0 2 in water. Hence, the outcome of this study constitutes an incremental improvement in the knowledge of the properties of NaB0 2 aqueous solutions with and without alkali additive and also establishes a platform to address some of the key issues in the area of recyclability and cost reduction. Table 1.1 summarises the specific energy, energy density and storage cost goals for on-board H 2 generation and storage set by the US Department of Energy (DOE) for the next decade. Table 1.1: US Department of Energy Freedom Car Targets for On-Board H 2 Storage, (Modified from [Department of Energy, (2004)]). Year/Target 2005 2010 2015 [Units] Specific Energy 5.40 7.20 10.80 [MJ/kg] 1.50 2.00 3.00 [kWh/kg] Energy Density 4.30 5.40 9.70 [MJ/1] 1.19 1.50 2.69 [kWh/1] Storage Cost 1.70 1.10 0.60 [$/MJ] 6A2 3,96 2T6 [$/KwH] 2 An efficient way to enhance the by-product solubility and recyclability could result in cost savings and protection of the environment, and consequently contribute to overcome one of the main constraint to the adoption of NaBH 4 as the H 2 carrier of choice for automotive fuel cell applications. 1.3 Approach To demonstrate the uniqueness and significance of the contribution made by this thesis, pertinent background information and a thorough literature review on the relevant topics were provided. This will show the importance of the issue of NaB0 2 recycling, which remains to be addressed. It will also be clear the lack of knowledge of the physicochemical properties of aqueous NaB0 2 solutions. For this reason,.the main physicochemical transport properties of dilute aqueous NaB0 2 solutions (conductivity and viscosity) were determined in pure water at 25°C. The NaB0 2 limiting molar conductivity, important transport electrochemical parameter of electrolyte solutions, was extrapolated from experimental data at very low concentrations using the established Kohlrausch empirical correlations, while the NaB0 2 B coefficient was evaluated from the Jones-Dole viscosity correlations. It will be explained that increasing the solubility of NaB0 2 is attractive as it would improve not only the overall Ff2 storage system capacity by allowing the use of more concentrated reactant solutions but also the effectiveness of the electrochemical recycling of the spent solution to NaBH4. Some work has been reported on the solubility of NaB0 2 in aqueous solutions, but, as will be seen, very few researchers have reported solubility data for NaB0 2 in caustic solution. The potential replacement of sodium hydroxide (NaOH) with another stabilizing agent, such as potassium hydroxide (KOH) or lithium hydroxide (LiOH), was explored by determining the solubility, pH, conductivity, viscosity and density, of alkaline aqueous NaB0 2 solutions at different weight percentages for each of those hydroxides. The effect of temperature on the physicochemical properties of the solutions and the effect of the addition of an organic additive were also briefly investigated. 3 1.4 Overview This thesis is divided in six chapters. Chapter II contains background information on the provenance of boron mineral and the production of NaBH4. The hydrolysis reaction is described and details of the effect of various factors on the reaction are given. The storage efficiency is discussed along with the H 2 generation and storage system and its demonstrated application possibilities. A thorough literature review including relevant state of the art theory on the NaB0 2 solutions and precipitate characteristics is provided in Chapter III. The following Chapter indicates the materials, procedures and characterization methodologies used to carry the experimental work. In Chapter V, the results obtained are presented and thoroughly discussed. The aqueous NaB0 2 solution physicochemical properties are discussed prior to the alkaline aqueous NaB0 2 solution physicochemical properties. The effects of temperature and of the addition of an organic compound on the solution physicochemical parameters are described separately. The precipitates characteristics are also discussed in this Chapter. Finally, Chapter VI constitutes a summary of the major findings. The significance of the experimental data collected and its impact on the H 2 generation and storage system, as well as on the electrochemical recycling are presented and future work is recommended. The thesis is followed by a series of Appendices. Appendix A holds the NaBH4 solution cost calculations for the DOE target range, as well as the NaBH4 H 2 generation and storage system cost, gravimetric and volumetric density calculations. Appendix B contains the summary tables of the averages of the filtrate characterization measurements as well as sample of the raw data collected. Appendix C includes the most important XRD peaks for a variety of boron compounds as well as for the precipitates studied, whereas Appendix D shows additional SEM pictures of the precipitates. Parts of this thesis have been accepted for publication in two articles [Cloutier, C. R., et al., (2006), Cloutier, C. R. et al. (2007)]. 4 C H A P T E R II B A C K G R O U N D 2.1 N a B H 4 Production 2.1.1 Natural Resources The manufacturing of NaBH4 uses refined minerals containing boron-oxygen compounds as the raw material. Borates are mineral ores containing boric oxides (B(OH)3). There exist at least 230 types of borate minerals [Garret, D. E., (1998)]. The commercially common ones are sodium borates, borax (sodium tetraborate, Na2B4O7.10H2O) and kernite (Na2B406(OH)2.3H20), calcium borate colemanite (Ca2B 6O n.5H 20) and sodium-calcium borate ulexite (NaCaB509.8H20). As they are not easily recycled, most borates are obtained from fresh sources. Similarly to oil and gas natural resources, borate deposits are found in a limited number of geographic regions. Major deposits are located in the United States, Turkey, Chile and Argentina. Minor producers include Russia, China and other countries. This geographic dependence affects the viability and sustainability of a transportation economy entirely dependant on NaBH4. Nevertheless, NaBH4, as a H 2 carrier, could have an impact on the energy consumption patterns in locations where deposits are present. 2.1.2 Current Production Process A) Production Paths NaBH 4 is the only boron hydride used routinely on a commercially significant scale [Kirk, R. E. et al, (1992)]. Various raw materials can be used in the preparation of NaBH4: boron hydrides, alkoxyboron compounds, boric oxide, metal borates and boron halides [Roy M. A., (1964)]. Methods using components 5 generated by the processing of boron ores as its raw material are reported below (Modified from [Martin, D. R., (1961)]). (Eq. 2.1) Borax (Na2B4O7.10H2O) and other boron ores are converted to boric acid (H3BO3) by using acids (a). Boric oxide (B203) is obtained by the calcinations of boric acid (H3B03) (b). Trimethoxyborane ((CH30)3B) is prepared by the esterification of boric acid (H3B03) with methanol (CH3OH) (c). Boron trifluoride (BF 3) is prepared from Na2B4O7.10H2O or H 3 B0 3 by reaction with an acid in the presence of fluoride (e, f). Finally, NaBH 4 is obtained by a reaction of sodium and H 2 or sodium hydride with B 2 0 3 (g), (CH30)3B) (d) B F 3 (h). As it produces less greenhouse gas emissions, the production of H 2 from NaBH4 synthesized from the decomposition of boron-containing compounds, such as Na2B4O7.10H2O, is less harmful to the environment than the production H 2 from the processing of fossil fuel [Ay, M. et al, (2005)]. However, for that to happen, the H 2 required in the NaBH4 synthesis would have to originate from non-fossil fuel energy sources such as solar, wind, biomass or hydroelectricity. B) The Schlesinger and Brown Process Most NaBH 4 is produced by the Schlesinger and Brown process. It is based on the reduction of trimethoxyborane ((CH30)3B) obtained from borate ore refining. The chemical reactions involved and a general process flow diagram describing the six most important steps are provided below. Initial borate processing: NaB0 2 + 1/2H2S04 + H 2 0 ->l/2Na2S04 + H 3 B0 3 (Eq. 2.2) Preparation of boron intermediate: H 3 B0 3 + 3CH3OH -»B(OCH 3) 3 + 3H 20 (Eq. 2.3) 6 Preparation of sodium metal: Preparation of sodium hydride: Production of NaBH4: Recovery of organic solvent: NaCl - » N a + 1/2C12 (electrolysis) 4Na + 2H 2 — 4NaH B(OCH3)3 + 4NaH -> NaBH 4 + 3NaOCH3 3NaOCH3 + 3H 20 -» 3CH3OH + 3NaOH (Eq. 2.4) (Eq. 2.5) (Eq. 2.6) (Eq. 2.7) Overall Reaction: NaB0 2 + 4Na + 2H 2 + 1/2H2S04 -> NaBH 4 + 3NaOH + l/2Na2S04 (Eq. 2.8) CH 4 NaCl Borax Steam Reforming CO, H, Electrolysis Cl 2 Refining Na 4 I I H 2 S0 4 H 3 B0 3 T Na 2S0 4 NaH CH 3OH A B(OCH3)3 NaBH 4 3NaOCH3 3CH 3OH + 3NaOH Figure 2.1: Process Flow Diagram of the Schlesinger and Brown Process (Modified from [Wu, Y. et al, (2004b)]). This method involves the use of a mineral oil suspension of sodium hydride [Ullmann, F., (2000)]. The determining reaction step is the one producing NaBH4. Sodium dispersions are hydrogenated directly in the mineral oil and the oil provides a heat sink, which facilitates temperature control of the exothermic reaction with methyl borate [Amendola, S. C. et al, (1999)]. The reaction is carried out in solvents such as diethylene glycol (diglyme) or tetrahydrofuran at 250-275°C. The sodium methoxide, NaOCH3, can be separated from NaBH 4 by a series of extractions: first with amines, then with water, and finally by a counter-extraction with a solvent, such as 2-aminopropane [BCS Inc., (2002)]. The large amount of Gibbs free energy required to combine the three different elements (Na, B and H) as well as electrons and energy needed to produce NaBH4 makes it difficult to make. The process includes the formation of sodium (Na) by the electrolysis of sodium chloride (NaCl). This reaction step has the highest activation energy requirements and is the most energy consuming of the entire process [Mohring, R. M. et al, (2003)]. The electrolysis of Na requires 9.7 kWh per kg of Na produced [Wu, Y. et al, (2004a)] and only a quarter of the sodium produced is converted to NaBH4. As will be seen in Section 2.2.2 H), the energy consumption associated with the electrolysis of Na can be reduced by about 80% by 7 carrying out the electrolysis in a H2-assisted molten sodium hydroxide (NaOH) media [Wu, Y. et al, (2004a)]. This significant energetic improvement could reduce the production cost of NaBH4, but it would not attenuate the amount of by-products (Na2S04), wastes (Cl2) and greenhouse gases (C02) generated by the overall Schlesinger and Brown process. C) The Bayer Process Bayer developed a NaBH 4 production batch process which is commercially viable. It uses ground borosilicate glass, elemental sodium and H 2 as per the following reaction [Ullmann, F., (2000)]: (NaB407 + 7 Si02) + 16Na + 8H2 -»• 4NaBH4 + 7Na2Si03 (Eq. 2.9) The reaction takes place at temperatures between 500 to 700 °C and at pressures between 300 and 400 kPa. However, the risks of explosion increase when the reaction temperature exceeds the decomposition temperature of NaBH4 which is 400°C [Wu, Y. et al., (2004b)]. The Bayer process is operated in a batch mode. As in the case of the Schlesinger and Brown process, only one fourth of the sodium used is converted to NaBH4, and elemental sodium is needed as a raw material. Also, the production of large quantities of sodium silicate by-product needs to be dealt with. 2.1.3 Economics The intention of this section is not to do a detailed economic analysis, but to provide an overview of the economic principles of supply and demand. Some statistics regarding the. current world production of borates are provided. Then the use of borates is briefly reviewed and the production cost of NaBH4 is analyzed. 8 A) Statistics The world's reserve base was estimated to 410,000 thousand metric tons of boron in 2004 [Lyday, P. A., (2005)]. The 2000 world apparent consumption in terms of B 2 0 3 was of 461,000 metric tons while the gross weight world mine production, in various boron units (elemental boron content, gross weight boron minerals and compounds) was of 4,220,000 metric tons [Buckingham, D. A., et al. (2002)]. During the same year, the total borate production from the four California producers was reported to be 1,100,000 metric tons valued at S557US million. In 2004, the United States was one of the world's two largest producers of boron compounds and exported about half of its domestic production. All the production is concentrated in California. Exported materials competed with the ones from Turkey, the largest producer of boron ore in the world. Canada imported 42,100 metric tons of sodium borates from the US in 2000 [Lyday, P. A., (2000)]. The unit value is defined as the estimated value of boron apparent consumption in US dollars, of 1 metric ton of 100 % B 2 0 3 content. The unit value was S787US per ton in 2000 [Buckingham, D. A. et al, (2002)]. Boron production remains quite low despite its desirable properties. At the 2004 consumption levels, the world resources are sufficient for the next 100 years or the foreseeable future. The world demand for boron is expected to remain strong and to grow. The most recent annual world production capacity of B 2 0 3 has been cited as sufficient for forecasted requirements of borohydrides [Lyday, P. A., (2005)]. B) Borate Usage The borosilicate glass production is the largest borate use and includes fiberglass for insulation, fabrics and reinforcement and many specialty glass products. As the boride compounds are very hard, they are used as abrasives and refractories. One of the earliest uses of borates is in the production of glazes, frits and emanels to impart color, texture, heat, chemical or wear resistance to appliances, ceramics or tiles [Garret, D. E., (1998)]. Composites of boron or boride fibers in a matrix of plastics, ceramics or metals have great strength and a high modulus of elasticity. Those advanced composites found uses in air and spacecraft applications. Boric acid, borax and pentahydrates are employed in the fabrication of fire 9 retardants. The mild alkalinity of borax allows it to emulsify oil and greases and reduce the surface tension of water, which aids in loosening dirt particles. It also reacts with organics to form esters and has a moderate bacterial action [Garret, D. E., (1998)]. Boric acid and sodium borates are mild antiseptics used in eyewash and some heterocyclic boron compounds inhibit tumor growth, while borax and some of its ores are efficient parasite killers. Borohydrides are well-known reducing reagents. In organic and inorganic chemistry, they are used as the sources of H" or reductants. Table 2.1 summarizes the most important uses of the numerous borate ores and their related products. Table 2.1: Partial List of Borate Uses (Modified from [Garret, D. E., (1998)]). Abrasives Hair creams Adhesives Herbicides Alloys Hydraulic fluids Antiseptics Insecticides Bactericide Leather tanning Bleaches Lubrificating oil additives Boron filaments Magnets Buffering Medical applications Catalysts Metallurgical applications Cement Metal hardening Ceramics Nuclear applications Cleaning compounds Nylon Corrosion inhibitor Organic synthesis Cosmetics Paints and pigments Detergents Pharmaceuticals Disinfectants Photography Electrical insulation Plastics Electrolytic refining Plating solutions Electronic components Polymer stabilisers Electroplating Pulp bleaching Enamels Purifying specialty chemicals Enzyme stabilization Pyrotechnics Eye wash Refractories Fertilizer Shampoos Fiber optics Soil sterilant Fiber glass Swimming pool sanitizer Fire, flame retardants Taxidermy Frits Textile finishing Fuel additives Textile dyes Fuel Transformers Funficides Waste treatment Glass Wax emulsifier Glazes Wire drawing Goldsmifhing Wood preservative 10 C) NaBH 4 Cost For small quantities of specialty chemical with no contract in place, the listing price of NaBH 4 caplets ranges between $80-$100US/kg at Rohm and Haas Company. However, the listed price of NaBH4 is about $50US/kg for industrial quantities. The cost of the sodium required in the current production of NaBH4 represents about 60-65% of the NaBH4 cost [Wu, Y. et al, (2004a)]. The inefficiencies of the current NaBH4 production process render NaBH4 an expensive chemical, which price greatly depends on grade, form, packaging and expected annual purchasing volume. Based on NaBH 4 listed prices, it was shown in Appendix A that it would cost approximately $3000-3800US (specialty chemical price) or $1900US (industrial price) to meet the 2009 US Department of Energy (DOE) of 8 kg H 2 target for 400 km [Bonhoff, K. et al, (2005)]. The H 2 pre-tax cost target at the refuel pump station site has been set to $3.00US/kg for 2008 [Department of Energy, (2003)]. At the current prices of NaBH4, the cost of NaBH 4 needed to produce 1 kg of H 2 ranges within S375-470US (specialty chemical price) or S230US (industrial price), which is significantly above the desired cost. It is clear that major costs reductions are required to attain the DOE $3.00US/gallon of gasoline equivalent (gge) target for 2009, and the $2-3.00US/gge target for 2015 [Wipke, K. et al, (2006)]. While the high cost of NaBH4-based solutions is not as critical in small portable Direct Borohydride Fuel Cell (DBFC) applications, it is a major obstacle to its large scale acceptance as a H 2 carrier for H 2 storage and generation in FCVs. The energy required for the manufacturing of NaBH 4 is significantly more then the energy required for the production of methanol [Li, Z. P. et al, (2003b)] and the production cost of the NaBH4 solution is approximately twenty times that of the cost of gasoline, for the same energy density. In addition, the price of H 2 generated from NaBH 4 is significantly higher than H 2 generated from the reforming of natural gas, and even higher than H 2 produced by electrolysis using wind generated electricity [Wee, J.-H., (2006)]. As several US and European chemical companies becomes eager to sign up to mass produce NaBH4, it is possible that the price of NaBH4 drops significantly in the next few years. However, until the cost of NaBH4 drops by a factor of 10, technologies based on this H 2 carrier will only be economically viable for limited applications [Larminie, J. et al, (2003)]. Nevertheless, currently the NaBH4 manufacturing process is not only costly: it also creates a lot of wastes, which requires further 11 treatment, and carbon dioxide emissions. The extensive use of a NaBH 4 based solutions would require the disposal of large quantities of metaborate salts. A way to achieve important cost savings while reducing the quantity of waste generated would be to develop an effective method to recycle the returned spent solution of NaB02 to produce NaBFL,. Depending on how this is achieved, reductions in greenhouse gas emissions could be attained. 2.2 NaBH 4 Regeneration This Section provides an overview of some of the possible ways to regenerate the borate produced from the NaBH4 hydrolysis reaction. The syntheses were separated in two main categories: chemical regeneration processes and electrochemical regeneration processes. Regeneration methods falling in each category are revised in Sections 2.2.1 and 2.2.2. Based on the information available in the literature, it was not possible to determine which of the options would result in a lower NaBFL, cost. For this purpose the methods described need to be thoroughly evaluated and compared using many factors, such as energy requirements, product purity, reaction yield, overall efficiency, by-product generation and environmental impact. To a minimum, one major criterion must be met: to have an energy efficient cycle, the regeneration process should not require more energy than the amount of energy released or recovered in the fuel cell system. 2.2.1 Chemical Synthesis Aside from the Schlesinger and Brown and the Bayer processes described in the Section 2.1.2, there exist several other chemical processes to produce NaBH4. Chemical regeneration approaches comprise organic based reactions, inorganic synthesis and ball milling processes. The majority of the chemical regeneration processes involve multiple chemical reaction steps which add significantly to the total manufacturing cost. Hence, even though good yields are obtained, the cost reduction of NaBH4 would become increasingly difficult as the complexity and the number of chemical reaction steps increases. A list of the processes selected for discussion in this review is provided in Table 2.2. This list does not comprise 12 all of the possible ways to chemically produce NaBH4. Other possible syntheses have been covered in a separate study [Wu, Y. et al, (2004b)]. Table 2.2: List of Possible Chemical Regeneration Synthesis. Chemical Syntheses References Modified Schlesinger and Brown Process [Amendola, S. C. etal, (2003)] Process amendment [Snover, J. et al., (2004)] Three-step regeneration method [Mohring, R. M. et al, (2003)] Using coke or methane [Kojima, Y. et al, (2003)] Recycling through diborane [Filby, E. E., (1976)] Using a reducing agent [Knorre.H. et al, (1968)] Annealing methods [Kojima, Y. et al, (2003)] Ball milling processes [Suda, S., (2003), Li, Z. P. etal, (2003a), Seijiro, S., (2004)] Radiolysis Process [Bingham, D. N. et al, (2005), Wilding, B. et al, (2004), Bingham, D. et al, (2004)] A list of US patents pertaining to different NaBH4 chemical synthesis is provided in Table 2.3, however, in this review, these patents were not individually studied in details. Table 2.3: Chemical Synthesis Patent List. Patent Number Title References US 3379511 Production of sodium borohydride [Knorre, H. et al, (1968)] US 3993732 Method of recycling lithium borate to [Filby, E. E., (1976)] lithium borohydride through diborane WO 02062701 A process for synthesizing metal [Yu, Z., (2002)] borohydrides US 6524542 Processes for synthesizing borohydride [Amendola, S. C. etal, (2003)] compounds JP2004010446 Method for producing alkali metal boron [Seijiro, S., (2004)] hydride US 6706909 Bl Recycle of discharged sodium borate fuel [Snover, J. et al, (2004)] US 20050077170Al Method of forming a chemical composition [Bingham, D. N. et al, (2005)] A) Modified Schlesinger and Brown Process A modification of the Schlesinger and Brown process adapted for the recycling of metaborate has been patented [Amendola, S. C. et al, (2003)]. It allows for the recycling of excess reagents and by-products 13 generated within the process to provide a greater efficiency in the production of NaBH4 [Amendola, S. C. et al, (2003)]. The first reaction step is carried out in a stirred tank reactor where carbon dioxide, at 30-40 arm, reacts with water and NaB02 to form a bicarbonate compound and boron oxide, which is filtrated out. 4NaB02 + 4C0 2 + 2H 20 - • 4NaHC03 + 2B 20 3 (Eq. 2.10) The bicarbonate is converted in sodium oxide, carbon dioxide and water by thermal decomposition at high temperatures in a rotary calciner. 4NaHC03 -> 2Na20 + 4C0 2 + 2H 20 (Eq. 2.11) Bromide is obtained by converting HBr in hydrogen and bromide by passage over a metal catalyst at temperatures of 100-500°C in a fluidized bed reactor. 12HBr —* 6H2 + 6Br2 (Eq. 2.12) In another stirred tank reactor, boron oxide reacts with carbon and a halide compound to obtain a boron halide, boron bromide, and carbon monoxide: 2B 20 3 + 6C + 6Br2 -> 4BBr3 + 6CO (Eq. 2.13) H 2 is obtained along with carbon dioxide from the reaction of carbon monoxide and water in a shift reactor. 6CO + 6H20 6H 2 + 6C0 2 (Eq. 2.14) The boron halide reacts with H 2 to form diborane and a halide in an autoclave: 4BBr3 + 12H2 -> 2B 2H 6 + 12HBr (Eq. 2.15) 14 Finally, sodium oxide reacts with diborane to produce NaBH4, NaB0 2 and carbon dioxide: 2Na20 + 2B 2H 6 3NaBH4 + NaB0 2 (Eq. 2.16) The process can be carried in a closed loop cycle where borate, carbon and water need to be supplied, while NaBH 4 and carbon dioxide are produced. The process overall reaction is: As the overall equation is endothermic, so energy needs to be supplied to generate NaBH4. The overall energetic requirement for the process is evaluated to be about 17551 kJ and the overall efficiency is estimated to 65 % [Amendola, S.C. et al, (2003)]. It has to be noted that natural gas reforming is used to produce H 2 , which releases important amount of C0 2 . Furthermore, carbon needs to be obtained from coke or methane. Thus, the environmental impacts associated with this process need to be accounted for. B) Modified Schlesinger and Brown Process: Amendment To improve the recovery of boron from a mixture of alkali metal borate and alkali metal hydroxide discharged from a H 2 generation system, an amendment to the modified Schlesinger and Brown process was proposed [Snover, J. et al, (2004)]. In this process, Eq. 2.10 is replaced by a reaction of carbon dioxide and lower alcohol to form trialkyl borate (B(OR)3), alkali bicarbonate and water. The reactor is pressurized with carbon dioxide at 15-35 arm for 1-10 hours [Snover, J. et al, (2004)]. The sodium bicarbonate formed is heated to produce sodium carbonate as in Eq. 2.11. The trialkyl borate is isolated from the reaction by distillation and then reacts with water to form boric acid as per the equihbrium: 3NaB02 + 6C + 6H 20 -> 3NBH4 + 6 C 0 2 (Eq. 2.17) NaB02 + NaOH + 2C0 2 + 3CH3OH -> B(OCH3)3 + 2NaHC03 + H 2 0 (Eq. 2.18) 15 B(OH)3 + 3CH3OH <-> B(OCH3)3 + 3H20 (Eq. 2.19) In order to favour the equilibrium towards the formation of trialkyl borate, excess alcohol is required and water has to be removed. The non-reacted alcohol is recycled in the process. Due to their ability of absorbing water as it forms, the addition of solid porous water-adsorbing materials, such as alumina pellets and molecular sieves, to the reaction mixture improved the yield of trialkyl borate by 50 % [Snover, J. et al., (2004)]. The trialkyl borate is then converted to alkaliborohydride by reacting with sodium hydride as in the conventional Schlesinger and Brown process described earlier. Alternatively, it could possibly be used in the three-step regeneration process described below. C) Three-Step Regeneration A somewhat simpler three-step regeneration process was proposed [Mohring, R. M. et al., (2003)]. Figure 2.2 summarises the overall synthesis reactions. ROH k NaBO, B(OR)3 H, Na 2C0 3 CO, HB(OR) ' ' B T B2H<5 W H C r — • NaBH, Figure 2.2: Schematic Diagram of the Three-Step NaBH4 Regeneration Process (Modified from [Mohring, R. M. et al., (2003)]). The first step is a reaction (A) involving borate salts under reflux in alcohol under a C 0 2 atmosphere to produce a trialkyl borate, B(OR)3. The reaction by-product, Na 2C0 3, is reused in reaction (C) to form NaBH 4 from B 2 H 6 . The most important reaction step (B) is the one where HB(OR)2 is synthesized from 16 B(OR)3. This reaction is endothermic and requires temperatures above 500°C. It is known that dialkoxy borane, HB(OR)2, disproportionate to form borane, hence, difficulties in obtaining good reaction (b) yields were anticipated. No yields were reported for reaction (b) [Mohring, R. M. et al, (2003)]. Reaction (c) is described as follows: The yields for this reaction were reported to be as low as 6 to 25 %. Higher NaC0 3 concentrations and greater NaC0 3 to B 2 H 6 ratios somewhat improved the yields [Mohring, R. M. et al., (2003)]. The use of a ball milling technique, an increase of the B 2 H 6 pressure in the reactor and the addition of diglyme to the ball milling process, all contributed to improve the dissolution of NaBH4, which enhances the exposure of the reactant surface for more reaction. Even after using all those improvements, the maximum yield obtained was 75 % [ Mohring, R. M. et al, (2003)]. Based on the fact that the yields are low and that carbon dioxide is a reaction by-product, it is obvious that the process is not economically viable and environmentally advantageous. D) Using Coke or Methane The use of a classical chemical reaction, modified for the composition of the recovered NaB02, was proposed to regenerate NaBH4. It was demonstrated that NaB0 2 can be recycled back to NaBH4 using coke or methane, which are less expensive reducing agents. During synthesis, H 2 is produced from methane via a decomposition reaction at 850-950°C [Kojima, Y. et al, (2003)]. The overall regeneration reaction combined with the H 2 generation reaction gives the overall reaction for the system. Methane decomposition: 2CH 4 —• 2C + 4H 2 (Eq. 2.22) Regeneration: NaB0 2 + 2CH 4 + 2H20 NaBH4 + 2C0 2 + 4H 2 (Eq. 2.23) 2B 2H 6 + 2Na 2C0 3 3NaBH4 + NaB0 2 + 2C0 2 (Eq. 2.20) B 2 H 6 + 6H 20 2B(OH)3 + 6H2 (Eq. 2.21) 17 When using coke, the reaction is described as: Regeneration: NaB0 2 + 2C + 2H 20 NaBH4 + 2C0 2 (Eq. 2.24) Two moles of methane are required to produce eight moles of H 2 , whereas two moles of coke are needed to produce four moles of H 2 . The cheapest and most energy efficient reducing agent would be coke (coal) or petroleum product [Kojima, Y. et al, (2003)]. However, in both cases, the overall regeneration would produce two moles of C0 2 , which conflicts with the fundamental objective of clean energy and the main reason for using H 2 . E) Recycling Through Diborane A method for the recycling of lithium borate to lithium borohydride was patented [Filby, E.E., (1976)]. In this synthesis, the borate reacts with hydrogen chloride and water to produce boric acid and lithium chloride, which are then converted to borohydride through a diborane intermediate. The preferred diborane synthesis is: The diborane intermediate can be reacted with a lithium species, such as lithium hydride, to produce lithium borohydride. The multi-step process overall reaction would be as follows: This reaction is very similar to Eq. 2.24, but in non-aqueous media. The chemical reaction steps involved in the process are relatively well known and the chemical separations are not difficult [Filby, E. E., (1976)]. The process involves the electrolysis of lithium chloride to form chlorine and lithium. If transposed to the production of NaBH4, this reaction step would be equivalent to the electrolysis used in the Schlesinger and 2BC13 + 6H 2 -y B 2 H 6 + 6HC1 (Eq. 2.25) 2LiB0 2 + 3C + 7H 2 + 5/202 - * 2LiBH 4 + 3H20 + 3C0 2 (Eq. 2.26) 18 Brown process described earlier. This is highly energy consuming. As per the overall synthesis reaction, three molecules of C 0 2 are produced for every two moles of NaB0 2 recycled. Thus, the environmental impacts of this process are significant. F) Reduction of NaBO z with H 2 A process for the production of NaBH4 by treating NaB0 2 with a reducing agent in the presence of H 2 at 500°C and under atmospheric pressures up to 10 arm was patented [Knorre, H. et al, (1968)]. The reaction is described as follows: NaB0 2 + Na20 + 2H 2 + Si -» NaBH 4 + Na 2Si0 3 (Eq. 2.27) The reaction is exothermic and the desired temperature range between 420 to 500°C [Knorre, H. et al, (1968)]. A NaBH4 yield of 93 % with respect to NaB0 2 has been claimed. The reaction yield is increased when excess reducing agent is added. Adding sodium oxide to the mixture activates the exothermic reaction and allows it to be carried at milder pressures. Pure NaBH4 can be extracted from the reaction product with various organic solvents. The reaction can also be carried with reductants such as carbon black and calcined charcoal. G) Methods Using Metal Hydrides a) Annealing The Toyota research group has developed an annealing technique which uses alkaline-metal based hydride such as LiH, NaH, CaH 2 and MgH 2. NaBH4 can be synthesized at 350-750°C by annealing NaB0 2 and MgH 2 or Mg2Si under H 2 pressure ranging from 0.1 to 7 MPa [Kojima, Y. et al, (2003)]. The reaction of NaBQ2 with MgH 2 to produce NaBH4 is as follows: 19 NaB0 2 + 2MgH2 NaBH4 + 2MgO (Eq. 2.28) This reaction is spontaneous ( A G 0 = -270 KJ/mol NaBH4). NaBH 4 yields up to 97-98 % were obtained at 550°C and 7 MPa, independently of the annealing time. The melting point of NaBH 4 and the thermal decomposition of MgH 2 played an important role in the yield of NaBH4 [Kojima, Y. et al, (2003)]. This method could represent an effective amendment to the Bayer process described in Section 2.1.2 C), as the use of Mg as a reducing metal would be less expensive than using Si [Wu, Y. et al, (2004b)]. • A similar annealing generation method to convert NaB0 2 and produce NaBH 4 with Mg2Si under high H 2 pressure has been investigated by the same research group. The reactions are as follows: NaB0 2 + 2Mg + Si + 2H 2 -> NaBH4 + Mg2Si +2H2 (Eq. 2.29) NaB0 2 + Mg2Si +2H2 -> NaBH 4 + 2MgO + Si (Eq. 2.30) The reaction is spontaneous but yields less NaBH4 then the reaction with MgH 2 [Kojima, Y. et al, (2003)]. To render the annealing method practical, the process of regenerating the hydride used as the H 2 donor has to be cost-effective and the energy consumption has to be minimized. b) Ball Milling In a mechano-chemical process, collision energy is converted to chemical energy. A ball milling technique has been utilized in the reaction of NaB0 2 with MgH 2 to produce NaBH4, as in Eq. 2.28 described above [Suda, S., (2003)]. Using this method, very high yields of NaBH 4 can be obtained at ambient conditions in limited time. The yields can be further increased by increasing temperature and pressure, but is independent of the reaction time. The addition of NaOH to the NaB0 2 solution further improves the reaction yield [Suda, S., (2003)]. Another academic research group investigated the formation of NaBH4 through the reaction of hydrated borax and magnesium hydride with ball-milling at room temperature [Li, Z. P. et al, (2003a)]. It was found that sodium compounds have to be provided to compensate for the reactant Na insufficiency and 20 promote the formation of NaBH4. Na 2C0 3 was the most effective additive and lead to the following reaction: 8MgH2 + Na 2 B 4 0 7 + Na 2C0 3 -> 4NaBH4 + 8MgO + C 0 2 (Eq. 2.31) The maximum reaction yield was of 78 %. Excess addition of MgH 2 improved the borohydride yield. However, the yield decreased when more MgH 2 was added after the NaBH4 yield reached a maximum value [Li, Z. P. et al, (2003a)]. The same method for manufacturing NaBH4 using NaB0 2 and magnesium hydride as the raw materials, was patented to include the formation of MgH 2 [Seijiro, S., (2004)]. The process requires four reaction steps. First, a hydrogen halide, such as HCI, reacts with an alkaline earth metal oxide, such as MgO, to produce an alkaline earth metal halide, like MgCl 2. Secondly, the alkaline earth metal halide is electrolysed to an alkaline earth metal (Mg) and a halogen (Cl2). In the third step, Mg is hydrogenated by H 2 to obtain an alkaline earth metal hydride (MgH2). Here again, the hydrogenation can be carried at milder conditions when mechanical energy is induced by a ball-milling process where MgH 2 reacts with NaB02, to generate MgO and NaBH4. In this case, a NaBH 4 yield of 45 % was claimed based on NaB0 2 [Seijiro, S., (2004)]. H) Radiolysis Process Radiolysis is being investigated as a potential route for off-board borate regeneration [Wilding, B. et al, (2004)]. In this method, radioactive waste is used as the photolytic energy source. It provides the ionizing potential required to modify the chemical bonds of NaB0 2 and permit chemical reconfiguration to the desired compound [Bingham, D. N. et al, (2005)]. The photon reacts with water and/or borate causing a change in the electronic structure, which creates a molecular radical. At high concentration, the radicals collide with the neutral molecules and form a stable borohydride configuration [Wilding, B. et al, (2004)]. With efficient control in place, it was.possible to produce stable borohydrides. A H 2 conversion efficiency of 53 % was calculated based on the analytical conversion of energy adsorbed [Bingham, D. et al, (2004)]. 21 This process seems to be economical as it reuses radioactive waste materials, which would otherwise require processing and storage at great cost. However, the borate radiochemistry mechanisms and reactions are not well understood at this time and need to be further investigated. 2.2.2 Electrochemical Synthesis Most electrochemical syntheses are simple as they require one reaction step. The use of an electrolytic cell for reducing NaB0 2 would only require the input of electrical energy. Based on the review of the chemical synthesis in the previous Section, it is clear that the use of electrochemical techniques has the potential of significantly reducing the capital and energy costs compared to chemical processes. Hence, electrochemical regeneration methods are the most likely to result in NaBH4 cost reduction [Wu, Y. et al, (2004b)]. Table 2.4 gives a list of possible electrochemical NaBH4 regeneration synthesis from NaB02. Table 2.4: List of Possible Electrochemical Regeneration Synthesis. Electrochemical Syntheses References Electroreduction: Direct electroreduction Indirect electroreduction Electroreduction of borate ester Electroreduction in Molten NaOH H 2 assisted electrolysis: Molten NaOH Borate in salt melt Aqueous alkaline NaBH4 solutions Electrosyntheses: Methathesis Simulating conventional process [Cooper, H. B. H., (1973), Sharifian, H., (1990), Hale, C. H., (1990), Amendola, S. C , (2002)] [Cooper, H. B. H., (1973), Gyenge, E. L. et al, (1998)] [Gyenge, E. L. et al, (1998)] [Wu, Y., (2004)] [Xu, J. et al, (2004)] [Wu, Y. et al, (2004a), Xu, J. et al, (2004)] [Mikio, K., (2003)] [Huff,G. F. et al, (1958)] [Linkous, C. A., (2003-4)] Table 2.5 is a list of patents regarding various electrochemical synthesis of NaBH4. According to the first five patents listed in Table 2.5, electrochemical cells can be used to generate NaB0 2 by the electrochemical reduction of NaB0 2 in aqueous media in yields ranging from 20 to 80%. Considering the number of recent patents pertaining to the electrochemical synthesis of NaBH4, it can be said that there is 22 an increase interest towards the recycling of NaB0 2 and the development of electrochemical NaBH 4 production processes. Table 2.5: Electrochemical Borohydride Synthesis Patent List. Patent Number Title References US 3734842 Electrolytic process for the production of alkali metal [Cooper, H. B. H„ borohydrides (1973)] US 4904357 Production of quartenary ammonium and quartenary [Sharifian, H.,(1990)] phosphonium borohydride US 4931154 Production of metal borohydrides and organic onium [Hale, C. H., (1990)] borohydrides US 6497973 Electroconversion cell [Amendola, S. C , (2002)] CN 1396307A Process for preparing boron hydride by electrolytic [Sun, Y. etal, (2003)] method JP 2003247088 Method and apparatus for manufacturing boron [Mikio, K., (2003)] hydride compound US 20040011662 A1 Hydrogen-assisted electrolysis processes [Xu, J. et al, (2004)] US 2855353 Electrolytic method for the preparation of metal [Huff, G. F. etal, borohydrides (1958)1 The patents above can be divided in three main categories: direct electroreduction (US patents 3734842, 4904357, 4931154, 6497973 and CN 1396307A), H 2 assisted electrolysis (JP patent 2003247088, US patent 20040011662 Al) and electro-synthesis (US patent 2855353). Each type of electrochemical synthesis is discussed separately in the following Sections. Only electrochemical methods allow the direct reprocessing of NaB0 2 to NaBH4. Electrochemical regeneration processes require less energy, are emission free and are most likely to result in significant cost reduction over the Schlesinger and Brown process. A) Direct Electroreduction The most convenient way to regenerate NaBH4 would be to use an aqueous media. The direct borate electroreduction can be described as the electrolysis of an aqueous NaB0 2 solution where NaB0 2 is directly reduced to form NaBH4. The theoretical cathodic reduction of the NaB0 2 to NaBH 4 in alkaline aqueous media is as follows: 23 Cathode: B0 2- + 6H 20 + 8e" -> BH4" + 80H" E°c = -1.24 Vvs. SHE (Eq. 2.33) Since the electroreduction takes place in an alkaline media, oxygen (02) evolves at the anode from the hydroxide ion. The release of 0 2 from an aqueous basic solution at the anode is given by: Anode: 80H -->4H 20 + 202+8e-E°a = 0.4 V(pH=14) vs. SHE (Eq. 2.34) This leads to the following overall cell electrochemical reaction and standard equilibrium cell voltage for the electroreduction cell: Overall: B02" + 2H 20 BH4" + 20 2 Ken = K - K = -1.24 - 0A{pH = 14) = -1.64V (Eq. 2.35) The non-standard equilibrium potential, E e , cen, is given by the Nernst equation: p = p° _ RT/ . l n [BO~2\ (Eq. 2.36) The actual cell voltage takes into account the electrode surface activation overpotentials (TJS and T]s ), the concentration overpotentials (T]con and T]con ) and ohmic losses (AO o h m) into account. When assuming that the open circuit voltage is approximately equal to the equilibrium cell potential, the actual cell voltage is given by: Ecell = Ee,ceU ~ Vs. + VS, ~ 1'con. + Vcon, ~ A O ohm (Eq. 2.37) 24 To shift the reaction equilibrium to the right and favor the generation of NaBH4, it is desired to make Ecen more positive and closer to the equilibrium cell potential by minimizing the overpotentials and the ohmic drop. When pure kinetic control is assumed, the mass transfer effects and the ohmic drop can be neglected. B) Kinetic Considerations At the low cathodic voltage of Eq. 2.33, the release of H 2 from water is more favorable. Hence, the H 2 evolution reaction competes with the NaBQ2 reduction: As Eq. 2.38 is likely to proceed at the cathode instead of Eq. 2.33, the electrochemical cell is not likely to generate significant amounts of NaBH4. Energy is wasted as the total current is split between both reactions. Hence, a significant concern is that it takes less energy to reduce the solvent than to reduce NaB02. To minimize the electrocatalytic H 2 evolution and favor the direct electrochemical reduction of NaB0 2 to NaBH4, a large H 2 evolution overpotential is needed at the cathode. The H 2 overpotential (Tjs ) is the difference between the cathode potential produced by the H 2 evolution reaction (Ec) and the equilibrium potential for that same reaction (ECj e). Large overpotentials can be obtained by selecting a cathode material which has a low catalytic affinity towards the H 2 evolution reaction. For Eq. 2.38, the cathodic current corresponds to the following expression based on the Tafel equation: 2H 20 + 2e" - * H 2 + 20H - (Eq. 2.38) E° = -0.83V(pH = \4)vs.SHE ( - 2 . 1 = -lo,c e X P b (Eq. 2.39) V 2 5 This expression can not be used directly for electrochemical reactions involving the transfer of more then two electrons. The exchange current density, i 0 ) c , depends on the electrolyte, temperature and pressure and is linked to the electrode material properties. When a simplified electrode kinetic model is used, the cathodic transfer coefficient, (0 )^, is found in the denominator of the Tafel slope (b) expression. b = 2'3'RT/a • F (Eq.2.40) Electrode materials leading to small Tafel slope (b) values and large exchange current density (i0i c) values will result in a low surface activation overpotential (TJS ). Hence, in this case, a large H 2 overpotential can be obtained by selecting an electrode material which gives a large b value and a small i 0 value for Eq. 2.38. Table 2.6 lists the exchange current densities and Tafel slopes for the H 2 evolution reaction using various materials. Table 2.6: Exchange Current Densities (i0) and Tafel Slopes for the H 2 Evolution Reaction at Various Materials in Basic Solutions at Room Temperature., *313K [Gouper, A.M. et al, (1990)]. Material Concentration OH' [M] i 0 [A/cm ] Tafel slope [mV] Ag 1 3.2x10"' 120 C 40%* 2.9xl0"5 148 Cu 0.1 lxlO"7 120 Cd 6 4x0"7 160 Cr 6 lxlO"7 120 Fe 01 1.6xl0"6 120 Hg 0.1 3xl0"15 120 Ni 0.5 7.9xl0"7 96 . Pb 0.5 3.2xl0"7 130 P 0.1 6.7xl0"5 114 Ir 0.1 5.5xl0"4 125 Sn 6 3.2xl0"7 150 Ti 6 lxlO-6 140 Zn 6 4xl0"7 210 Materials with high H 2 overpotentials are Hg, Pb, Cd, Zn, Sn, C, Cu, as well as steels. In addition, Ag, Ni and Fe are also worth considering in alkaline solutions [Gouper, A.M. et al, (1990)]. Cost and availability often favours Ni or stainless steels (SS). Unfortunately, experiments have shown that materials with high H 2 overpotential are not effective to transfer electrons and H 2 to a boron center [Wu, Y. et al, (2004b)]. 26 Luckily there exist other measures to increase the H 2 evolution overpotential. For example, important overpotentials can be obtained in the presence of certain additives. When exceeding a certain concentration, the additive molecules cover the active surface of the electrode. The H atoms have no empty metal surface to adsorb on the cathode surface and the H 2 evolution reaction is minimized. The potential of the H 2 ion at the electrode increases, and thus, the H 2 evolution overpotential increases. A hydride layer forms on the cathode material surface, and the ability of H 2 to adsorb in the cathode metal increases. Some additives, such as thiourea (CH4N2S) and selenium dioxides (Se02), are catalytic poisons to the H 2 recombination reaction [Szklarska-Smialowska, S. et al, (1963)]. A systematic study of the direct electroreduction of NaB0 2 has been conducted in the presence of additives, such as quaternary ammonium compounds and thiourea, in order to further restrain the electrocatalytic H 2 evolution [Gyenge, E. L. et al, (1998)]. Even with those extra measures in place, the production of detectable amounts of NaBH4 by the direct electrochemical reduction of borates was not achieved. The counter electrode 0 2 evolution reaction can be maximized by using an anode material resulting in a low 0 2 overpotential. In other words, a material leading to small b value and a high i 0 value for Eq. 2.34 would be more favorable. Materials suitable for the 0 2 evolution are nickel or nickel-plated mild steel. Many coatings have been reported to have the ability of reducing the oxygen overpotential. They include metal oxides, spinels and perovskites [Gouper, A. M. et al, (1990)]. C ) Patent Review US patent 3734842 describes an electrolytic process for the reduction of alkali metal borohydrides [Cooper, H. B. H., (1973)]. An alkaline aqueous solution of NaB0 2 is supplied to the cathode compartment of an electrolytic cell. To ensure proper conductivity a catholyte solution of 0.1 to 5 wt% NaB0 2 concentration is recommended. The maximum concentration of the catholyte is limited by the solubility of NaB0 2 at the cell operating conditions. An aqueous solution of NaOH constitutes the electrolyte in the anode compartment [Cooper, H. B. H., (1973)]. A cationic selective membrane separates the anode and cathode compartments. This implies that the Na+ ions are responsible for the transport through the membrane. The cell electrochemical reactions, Eq 2.33 to 2.35, can be written as: 27 Cathode NaB0 2 + 8Na+ + 6H20 + 8e" NaBH 4 + 8NaOH (Eq. 2.41) Anode: 8NaOH -* 2Q2 + 4H 20 + 8Na+ + 8e_ (Eq. 2.42) 8NaOH —• 8Na+ + 80H" 80H - — 20 2 + 4H 20 + 8e" Overall: B02" + 2H 20 - • BH4" + 20 2 (Eq. 2.43) The overall cell electrochemical reaction remains the same as for Eq. 2.35. The reduction of borate ions to borohydride ions occurs in the cathode compartment to produce alkali metal borohydride solution, from which the borohydride may be later separated. The oxygen generated at the anode is vented from the anode compartment. It was suggested that the electrodes be made of graphite, carbon, nickel, cobalt, silver, stainless steel, iron, nickel-plated iron, platinized titanium or tantalum, platinum, palladium, iridium and others [Cooper, H. B. Ff., (1973)]. It was noted that the activation of the reaction between the borate ion and H 2 generated at the cathode surface could be advantageous. Hence, it was suggested to use hydrogenation catalytic materials as the cathode. Such materials include nickel boride, nickel, Raney nickel, cobalt, cobalt boride, platinum and palladium [Cooper, H. B. H., (1973)]. It is stated that the cell can be operated at current densities ranging from 620 to 1550 A/m2. The operating voltage of the cell had to be greater than 1.23V, which is the voltage required for the H 2 0 splitting. Conversions to NaBH4 ranging from 20 to 80 % were claimed [Cooper, H. B. H., (1973)]. US patent 4904357, extends the application of the electrolysis cell described in the earlier patent 3734842 to the production of quaternary ammonium and phosphonium borohydrides by the reduction of the corresponding ammonium or phosphonium boron oxide compounds [Sharifian, H., (1990)]. Various boron oxide compounds, such as perborates, tetraborates and metaborates, can be used as reactants in the process. Boron oxide concentrations ranging between 1 and 40 wt% are recommended in the aqueous catholyte solution, while concentration of salt ranging between 3 and 40 wt% are recommended in the anolyte [Sharifian, H., (1990)]. An electrolysis experiment was carried for two hours at a current density of 500 A/m2, with an anolyte composed of 25 wt% tetramethylammonium hydroxide, a catholyte composed of 7 wt% tetramethylammonium, platinum cathode and anode, and nickel boride particles in the catholyte 28 compartment. A current efficiency of 50 % was achieved for the synthesis of tetramethyl ammonium borohydride. In a similar experiment using nickel powder in the catholyte compartment this time, a conversion efficiency of 45 % was reached. When no catalysts were present in the catholyte compartment, the efficiency was as low as 8 % [Sharifian, H., (1990)]. US patent 4931154 describes an electrolysis cell for the production of metal borohydrides and organic onium borohydrides [Hale, C. H., (1990)]. The process is said to be particularly useful for reducing borate ions to borohydride ions. Like in the two previous patents described above, the aqueous catholyte solutions containing 1 to 40 wt% of borate compound and alkaline catholyte solutions are preferred to stabilize the borohydride formed. A cationic-selective membrane is used to favor the migration of cations from the anolyte to the catholyte and increase the alkalinity of the catholyte solution, and Nation™ membranes are suitable [Hale, C. H., (1990)]. Unlike the other patents reviewed so far, an aqueous acidic solution constitutes the anolyte. At the anode, water in the aqueous acidic solution is ionized. The H + generated migrates through the membrane to the catholyte solution where metaborate is reduced to borohydride. As the cation in the anolyte is a hydrogen ion, oxygen is evolved and escapes from the anolyte compartment [Hale, C. H., (1990)]. According to the patent, the direct electrochemical reduction of an alkaline aqueous solution of NaB0 2 to NaBH 4 in an electrical cell using an acidic anolyte solution is as follows: Cathode: B0 2 ' + 6H 20 + 8e" -> BH4" + 80H" (Eq. 2.44) Anode: 4H 20 8H+ + 20 2 + 8e" (Eq. 2.45) (8C4T + 8H+ 8H20) Overall: B02" + 2H 20 -> BH4" + 20 2 (Eq. 2.46) Again, the overall cell electrochemical reaction still remains the same as for Eq. 2.35. Figure 2.3 represents a schematic diagram of electrolytic cell for the direct electroreduction of NaB0 2 to formNaBH4. 29 A c i d / H 2 0 Figure 2.3: Schematic of Electrolytic Cell for the Direct Reduction of Borates to Borohydrides: 1) Cathode, 2) Catholyte, 3) Membrane, 4) Anode, 5) Anolyte, 6) Electric Cell, 7) Wire, 8) Aqueous Acid Solution, 9) Borate Aqueous Solution, 10) Borohydride Withdrawal Line, (Modified from [Hale, C. H., (1990)]). Passing a current through the cell reduces the borate ions to borohydride ions in the cathode compartment. The borohydride can be later separated from the alkali metal borohydride solution produced. During the electrolysis, the liquid in the cell is kept at temperatures of 50°C or lower [Hale, C. H., (1990)]. The electrolytic cell can be operated in a batch or continuous way. For this acidic anolyte electrolytic cell, anodes with non-passivable and catalytic films are recommended. As the oxygen reduction reaction is slow, good catalysts such as metallic noble metals such as platinum, iridium, rhodium or their alloys, as well as mixtures of electroconductive oxides with at least one oxide being a noble metal are recommended. Suggested cathode materials included nickel, iron, stainless steel, nickel plated titanium and platinum. The cathode material can be coated or dispersed on a metal or inert substrate [Hale, C. H., (1990)]. An electrolysis experiment was carried in a cell with a 10 wt% NaB0 2 solution in aqueous 1 M sodium hydroxide as the catholyte and an aqueous solution of 2 M sulfuric acid as the anolyte. By carrying the electrolysis at a current density of 50 mA/cm2 for two hours, a current efficiency of 20 % for the synthesis of NaB0 2 was achieved when using a platinum anode and a nickel cathode [Hale, C. H., (1990)]. US patent 6497973 claims an electroconversion cell, which has the ability to function in one or several modes [Amendola, S. C , (2002)]. This system is constituted of two electrochemical cells: one cell consuming borohydride ions to produce electricity and one cell which electrically generate borohydride ions from the borate. The cells may be separated or the same cell can be used for both processes. The 30 carrier may be an aqueous or non aqueous solution, but if a non-aqueous solution is used, a solubility or conductivity enhancer is needed [Amendola, S. C , (2002)]. Recharging the cell obviously assumes that it is possible to synthesize borohydride by the direct electroreduction of borates. In this patent the electrode material selection for the cathode is erroneously explained. It is said that a material with an overpotential equal to or greater than the difference in voltages between Eq. 2.33 and 2.38 are needed (1.24 - 0.8277 = 0.41 V) to minimizing Eq. 2.38 as much as possible [Amendola, S. C , (2002)]. This voltage difference was incorrectly attributed to the hydrogen overpotential for Eq. 2.38. As explained earlier, the overpotential desired is not equal to the difference of potential between those two reactions and is a function of the current exchange density and the transfer coefficient. According to the patent, the optimal cathode would be a small electrode with high current density and would be coated with a soft metal alloy [Amendola, S. C , (2002)]. It was said that soft materials have a tendency to have high H 2 overpotentials. For example, it is said that mercury, due to its E° = -2.7 V, would allow Eq. 2.38 to proceed at high efficiency. Other soft metals judged appropriate included bismuth, lead, tin, thallium, cadmium, gallium and indium. Tellurium restrains the ability of an electrode material to generate H 2 gas and can be included in the metal [Amendola, S. C , (2002)]. Also, according to this patent, the optimal anode would have a high surface area and low current density like an electrode coated with gold or iridium oxide. Recommended anode materials for low oxygen overpotentials are gold, iridium oxide, manganese dioxide and others [Amendola, S. C , (2002)]. The Chinese patent application CN1396307 describes an electrochemical process to generate NaBH4 from NaB0 2 [Sun, Y. et al, (2003)]. In this patent, the electroreduction of NaB0 2 in aqueous solution at more than 30 wt% efficiency was claimed [Wu, Y. et al, (2004b)]. D) Reproducibility The NaB0 2 direct electrolysis yields claimed in the patents described above remain to be ascertained. The laboratory reproduction the direct electrochemical reduction of NaB0 2 did not produce any detectable amounts of NaBH4 [Gyenge, E. L. et al, (1998), Suda, S., (2003)]. Hence, it is highly probable that the BH 4 ' ion was not produced in significant amounts because, as explained earlier, it can not avoid the 31 thermodynamically favorable conversion to B02~. If that is the case, the efficiency of the direct electrochemical production of NaBH4 in aqueous solution is too low to be practical in industrial applications [Yu, Z., (2002)]. Furthermore, it was reported that there might be erroneously high concentrations of BH4" in some solutions of other reductants formed during the electrolysis according to the commonly used iodate detection method [Mirkin, M. V. et al, (1991), Gyenge, E. L. et al, (1998)]. More accurate, reliable and selective detection methods are needed to validate the results obtained. Using more robust methods, such as a voltammetric detection method [Mirkin, V. M. et al, (1991)], and the novel phosphotungstic acid reduction analytical method [Gyenge, E. L. et al, (1998)], no NaBH4 was detected after the direct electroreduction of NaB0 2 alkaline aqueous solutions. Consequently, a square wave sweep voltammetric method having a low detection limit of 3 x 10"5 M has been developed to study the formation of borohydrides [Celikkan, H. et al, (2005)]. Logically, this leads to the conclusion that water is not a suitable solvent to carry the direct electroreduction of NaB02. E ) I n d i r e c t E l e c t r o r e d u c t i o n The cathode and anode reactions for the indirect electrochemical generation of NaB0 2 in aqueous alkaline media are as follows: (Eq. 2.47) (Eq. 2.48) (Eq. 2.49) In this case, the H 2 evolution reaction is used at the cathode. As it leads to a different overall reaction than in the case of the direct electroreduction of NaB02, this process can be described as an electrochemical hydrogenation. Based on the cell standard potential, it can be showed that this electroreduction process Cathode reaction: Anode reaction: Overall cell reaction: po _ pO _ po ^cell ~ fic 2H 20 + 2e" H 2 + 20FT B0 2- + 4H 2 BH4" + 2H 20 20H- -> H 2 0 + l/20 2 + 2e" H 2 0 -> l/20 2 + H 2 = -0 .83- (0.4) =-1.23V 32 requires a quantity of energy lower than that needed for the direct electroreduction process. This electroreduction process is said indirect as it aims at promoting the anode reaction to produce H 2 and make it react with the metaborate on the cathode material. The principle is similar to that of the H 2 assisted electrolysis process which will be described in a later Section. Hydrogenation catalysts, such as Ni, Raney-Ni, NiB, palladium or zinc, are recommended cathode materials [Cooper, H. B. H., (1973)]. In this case, the hydrogenation catalyst role is to promote the H 2 evolution reaction, and to affect the adsorption and dissociation of H 2 at the electrode surface. Experiments showed that the use of Raney-Ni was not successful as it catalyzes the H 2 evolution reaction but does not catalyse the reaction of B02" and H 2 [Gyenge, E. L. et al, (1998)]. This was not the case when using NiB as the cathode material. Nevertheless, experiments demonstrated that this process did not generate any detectable quantities of NaBH4 [Gyenge, E. L. et al., (1998)]. Another research group attempted to explain why the cathodic electroreduction of the borate ion of Eq. 2.33 follows an indirect reduction process [Wei, X.-Y. et al., (2003)]. The negatively charge borate ion has difficulties to approach the cathode. According to their findings, a possible scenario would be that the water molecules release an electron and produce H 2 at the cathode, which in turns reduces B02" to BH4". Hence, the mechanism for the reduction of NaB0 2 would be controlled by the absorption of H 2 on the cathode. If this was the case, Ni would be a good cathode material as it strongly absorbs H 2 . This also reemphasizes the importance of selecting a medium favourable to electron transfer. Another finding was that the imposition of a positive pulse on the cathode forced the anion to absorb directly on the cathode surface and rendered the electroreduction, as per Eq. 2.33, a direct process [Wei, X.-Y. et al, (2003)]. F) Electroreduction of Borate Ester The electroreduction of borate ester in organic media on graphite and aluminum cathodes has also been investigated [Gyenge, E. L. et al, (1998)]. This rigorous reduction based on solvated elections was performed using a hexamethylphosphoramide (HMPA)-ethanol mixture or ehtylenediamine (EDA) as the catholyte, and with lithium salts as supporting electrolyte. No NaBH4 was detected in these experiments. 33 G) Electroreduction in Molten Hydroxide Media As part of the US DOE Hydrogen Program, a research group discovered a stable molten hydroxide system in which NaB0 2 might be directly reduced to NaBH4. The apparent electrolytic activity of NaB0 2 in hydroxide melt was observed. The addition of NaB0 2 to a hydroxide melt caused significant changes in the cyclic voltammograms, but the reduction wave remains unidentified at this point [Wu, Y., (2004)]. H) Hydrogen Assisted Electrolysis Using H 2 gas at the anode lowers the overpotential on the anodic side of an electrolytic cell, and lowers the overall cell voltage required during the electroreduction process [Xu, J. et al, (2004)]. Hence, the use of H 2 gas lowers the cell voltage, resulting in an improved regeneration efficiency of borate or sodium electroreduction processes [Wu, Y., (2003b)]. Consequently, the H 2 assisted electrolysis for the regeneration of NaBH 4 represents a lower cost alternative to electrochemical processes generating 0 2 at the anode. The resulting energy efficient improvement could lower the production cost of NaBH4. Three different types of H 2 assisted electrolysis process are discussed in this Section: the electrolysis of molten NaOH, the electrolysis of borate in salt melt, and the electrolysis of aqueous alkaline borate solutions. a) Molten NaOH Electrolysis The electrolytic reduction of sodium hydroxide, NaOH, to sodium metal, Na, can be performed by the H 2 assisted molten NaOH electrolysis. The electrochemical reactions for the conventional sodium electrolysis and for the H 2 assisted molten NaOH electrolysis are described below [Xu, J. et al, (2004)]. Note that the cathode sodium reaction is the same in both cases. Conventional sodium electrolysis: Cathode: Na+ + e"->Na (Eq. 2.50) Anode: CI" + e" — Vi Cl 2 (Eq. 2.51) Overall: Na+ + !/ 2Cl 2 - * Na + CI" (Eq. 2.52) 2.71-1.36 =-4.07V 34 Hydrogen assisted molten NaOH electrolysis: Cathode: 2Na+ + 2e" — 2Na (Eq. 2.53) Anode: 20H - + H 2 -» 2H 20 + 2e" (Eq. 2.54) Overall: 2Na+ + 2H 20 2Na + 20H" + H 2 (Eq. 2.55) E, cell o = E°C-E° =-2.71-(-0.83) = -1.88F The standard cell potential of the H 2 assisted molten NaOH electrolysis is 53.76 % less then that for the conventional sodium electrolysis. A lower voltage results in cost savings. This process could be directly integrated in the currently used Schlesinger and Brown process described in Section 2.1.2 B) to improve the efficiency of Na production. Figure 2.4 represents a schematic diagram of electrolytic cell with H 2 at the anode for the electrolysis of sodium hydroxide in molten NaOH. Figure 2.4: Hydrogen Electrolytic Synthesis of NaOH in Salt Melt 1) Cathode, 2) Catholyte, 3) Membrane, 4) Anode, 5) Anolyte, 6) Hydrogen sparger, 7) Molten sodium, (Modified from [Xu, J. et al., (2004)]). H 2 is supplied at the anode compartment where it oxidizes to protons, which then react with hydroxide and produce water. In the cathodic compartment, the molten sodium hydroxide medium is reduced to produce metallic sodium. Molten NaOH is present in both cell compartments, which are separated by a membrane. Since the cathode compartment has to be free of water, the membrane has to be impermeable to water and water vapor produced during the reaction, but permeable to alkali metal cations [Xu, J. et al, (2004)]. Cation exchange ceramics, such as sodium beta alumina, are said to be appropriate membrane materials. The cathode material has to be inert at high temperatures for this reaction, such as nickel, copper and 1 I 4 7 -35 stainless steel. The anode is a H 2 diffusion electrode with a high specific surface area, like nickel or supported noble metals. The reaction is carried at 300°C and above to maintain the anolyte and catholyte in the molten state. The sodium produced by electrolysis in the cathode compartment floats on top of the catholyte as a molten layer. It can be continuously or intermittently removed, while the molten NaOH feed can be continuously or intermittently supplied to the cell [Xu, J. et al, (2004)]. Operating voltages ranging from 1.46 to 6 V are used to carry the electrochemical reaction [Xu, J. et al, (2004)]. The cell voltage necessary to convert the alkali metal oxide to reduced metal is about 1.46 V at 350°C while the cell voltage needed to convert the alkali metal hydroxide to alkali metal in the absence of H 2 gas at the same temperature is of 2.44 V. b) H2 Assisted Molten Salt Electroreduction of Borate The successes obtained in the sodium H 2 assisted electrolysis were transferred to the reduction of NaB0 2 for the production of NaBH4. It has been shown that borates are more active in a molten hydroxide melt than in water. Therefore, it might be preferable to carry the H 2 assisted borate electrolysis in a molten hydroxide melt. The alkaline solvent provides stability and solubility for the borate reactants and borohydride products. It was demonstrated that NaBH4 is stable in a mixed melt containing NaOH and KOH. At the eutectic point of NaOH and KOH, the melting point of the solids is below 190°C. At 200°C, NaBH4 hydrolysis was minimal [Wu, Y. et al, (2004a)]. The H 2 assisted electrochemical reaction, the H 2 used at the cathode and the electrolyte present in a molten state provides the H 2 required for forming NaB02, which would not readily be produced without H 2 . While H 2 is passed in the cathode compartment, the borate is electrochemically reduced to borohydride, according to the same electrochemical reactions as for the indirect electrochemical reduction process, Eq. 2.47 to 2.49. Figure 2.5 represents a schematic diagram of electrolytic cell with H 2 at the anode and cathode for the synthesis of NaBH 4 from hydroxide melt containing NaB0 2 [Xu, J. et al., (2004)]. 36 i i Figure 2.5: Hydrogen Electrolytic Synthesis of NaBH4 in Salt Melt 1) Cathode, 2) Catholyte, 3) Membrane, 4) Anode, 5) Anolyte, 6) Hydrogen sparger (Modified from [Xu, J. et al, (2004)]). As shown in Fig. 2.5, H 2 is passed in the anodic and cathodic compartments from an outside source. The cathodic compartment contains alkali metal borate dissolved in a molten ionic salt and the anodic compartment contains a molten solution of sodium hydroxide with or without ionic salt dissolved in it. Water has to be removed from the system to avoid its electrolysis and the back-electrolysis reaction of borohydrides to borates. Carrying this process in a molten anolyte and catholyte which contains no water is the best way to ensure the cathodic compartment is free of water [Xu, J. et al, (2004)]. The borohydride is formed in the cathodic compartment and water is formed in the anodic compartment. The electrolytic cell membrane's is only permeable to alkali metal ions and not permeable to other ions, water or water vapor. As for the sodium electrolysis process, a cation-exchange ceramic membrane made of sodium beta alumina is suitable. The reaction is carried at temperatures ranging from 100 to 500°C depending on the catholyte and anolyte melting points [Xu, J. et al, (2004)]. Here again, the catholyte can be continuously or intermittently processed to remove the alkali metal borohydride produced. The residual alkali metal hydroxide is recycled to the anode, while the alkali metal borate is fed to the cathode compartment. The water vapor forming at the anode is carried away by the non-reacted H 2 leaving the anode chamber. Alternatively, an alkali metal oxide, such as sodium oxide, can be used to convert the remaining water vapors to NaOH, and prevent them from entering the cathode chamber [Xu, J. et al, (2004)]. Theoretical calculations showed that from a 5 g of NaOH/NaB02 mixture containing 10 wt% NaB02, a current of 1000 mAh can generate 0.18 g of NaBH4 at 100 % efficiency [Xu, J. et al, (2004)]. This electrochemical 37 process has the potential of replacing the current NaBH4 production method. However, in an industrial setting, the external H 2 needed in this process would most likely be obtained from natural gas reforming. c) Aqueous Alkaline Borate Solutions A H 2 assisted direct electroreduction process to produce borohydride from metaborate in alkaline aqueous solution was patented [Mikio, K., (2003)]. In this process, an aqueous alkaline solution of metaborate is electroreduced in an electrolytic cell, supplied with H 2 gas only in the anode compartment. The ion exchange membrane is an anion conductive diaphragm. The cathodic oxidation reaction is the same as Eq. 2.33, while the anodic reaction can be either the oxygen evolution reaction, Eq. 2.34, or the H 2 reduction reaction, Eq. 2.38. The gas, the electrolyte and the electrode then form a three-phase interface [Mikio, K., (2003)]. When the anode reaction is the H 2 reduction reaction, an anode material having a small H 2 overvoltage is desired and a gas diffusion electrode is preferable for the anode. Recommended anode materials with low H 2 overvoltage include palladium, platinum, ruthenium, osmium, iridium, rhodium, gold, cobalt, silver, nickel, tungsten, iron, copper, titanium and carbon. Among those, materials having also a high 0 2 overvoltage, such as palladium, gold and silver, are preferable [Mikio, K., (2003)]. Cathode materials with a large H 2 overvoltage, such as tantalum, indium, zinc, lead or carbon, are recommended. On the other hand, when the anode reaction is the 0 2 reduction reaction, it is recommended to use cathode materials with large H 2 overpotentials such as indium, zinc, lead, and carbon, and anode materials with small 0 2 overpotentials, such as nickel, cobalt, platinum, and iron. With a gas diffusion anode, a tantalum cathode, an aqueous solution of NaB0 2 in 6 M NaOH as the catholyte and an aqueous solution of 6 M NaOH as the anolyte, BH4" was generated at 80 % current efficiency when a current of 1 V was imposed on the cell and supplying the anode with 2 MPa H 2 gas [Mikio, K., (2003)]. I) Methathesis This electrosynthesis process for the preparation of borohydrides utilizes a methathetic reaction between a metal halide and a borohydride [Huff, G.F. et al, (1958)]. It is an electrochemical displacement 38 of one metal by another metal. The anode metal selected has to correspond to the metal of the desired borohydride. For example, when NaBH4 is electrolysed using a magnesium anode and a mercury cathode in an ionizing non-aqueous solvent, the following reactions occur: Anode: Mg — Mg 2 + 2e" (Eq. 2.56) Cathode: Na+ + e + Hg — Na(Hg) (Eq. 2.57) Overall: Mg + 2Na+ + Hg - » M g 2 + + 2Na(Hg) (Eq. 2.58) The sodium cations are replaced with magnesium cations and magnesium borohydride precipitates. Reaction yields of 98 % were obtained [Huff, G. F. et al, (1958)]. J) Simulating Conventional Synthesis The Florida Solar Energy Center, Hydrogen R&D Division at the University of Florida has a project targeting stationary H 2 generation systems based on NaBH4 for NASA [Linkous, C. A., (2003-4)]. Their research is focused on the development of an electrochemical process, simulating the conventional Schlesinger and Brown synthesis in which NaBH4 is formed from the reaction between alkali metal hydride and trimethoxyborate, B(OCH3)3. The following reactions have been postulated: Anode 2H 2 -<• 4H + + 4e" (Eq. 2.59) 4H + + B O / + 3CH30" ->• B(OCH3)3 + 2H 20 (Eq. 2.60) Cathode 2H 2 + 4e~ —»4H"(s)B(OCH3)3 + 4H"(s) —» BH4" + 3CH30" (Eq. 2.61) Overall 4H 2 + B02" -> BH4" + 2H 20 (Eq. 2.62) The anode reaction would consist of H 2 oxidation to form H + , which would then be used to convert B02" to trimethoxyborate. At the cathode, H 2 would be reduced to hydride ion (H"), which then displaces the methoxide ion, generating BH 4'. Noble metal catalysts active towards the electrooxidation of H 2 , such as Pt and Pd, will be used to demonstrate the electrochemical hydriding of the trimethoxyborate intermediate 39 [Linkous, C. A., (2003-4)]. Preliminary experiments were conducted to identify proper electrode materials for the anodic and cathodic reactions, but no published papers regarding this work are available yet. 2.2.3 Infrastructure According to the life cycles proposed by various research groups, [MERIT, Ltd., (1986), Levy, A. et al, (1960), Jacques, S. et al, (2004)], a scheme analogous to the current gasoline distribution infrastructure could be used to dispense NaBH 4 solution. The electrochemical regeneration of NaB0 2 to NaBH4 would be carried out off-board. NaBH4 fuelling infrastructure is depicted in Fig. 2.6. Electrochemical Recycling Figure 2.6: NaBH4 Fuel Infrastructure Schematic Diagram (Modified from [MERIT, Ltd., (1986)]). Once H 2 is produced, the NaB0 2 slurry must be removed from the vehicle while fresh NaBH4 solution is added. Refuelling could be done within a few minutes. The recovered spent solution would then be transported to an industrial recycling facility for regeneration. Moreover, if an electrochemical regeneration process is used, it could be combined with load levelling operations with hydroelectric or nuclear power plants. Then, the electricity produced during off-peak electricity demand hours would be sufficient to supply the energy necessary to electrochemically regenerate NaBH4. However, the energy required for the large-scale electrochemical regeneration of borohydride is difficult to evaluate with accuracy at this time [Jacques, S. et al, (1970)]. The electricity provided could also come from other renewables such as wind, tide and solar energy sources. 40 2.3 NaBH4 H2 Generation and Storage System Efficiency Among the currently known chemical hydrides, NaBH4 seems to be the most suitable compound for H 2 storage as it releases more H 2 at lower reaction temperatures and has a higher H 2 density than other chemical hydrides. Even though NaBH4 alkaline aqueous solutions have a lower gravimetric capacity then gasoline, when it is used to feed a PEMFC, it has the potential to exceed the volumetric capacity of gasoline used in an engine. As it will be seen in this Section, the H 2 storage efficiency of NaBH 4 renders it competitive for H 2 generation. 2.3.1 Gravimetric Capacity The NaBH4 hydrolysis reaction is very efficient on a weight basis. NaBH4 has a theoretical H 2 content of 10.58 wt% based on the compound itself. Table 2.7 compares the theoretical H 2 content of various chemical hydrides. Table 2.7: H 2 Content of Various Chemical Hydrides. Name Formula Formula H 2 [wt%] Sodium borohydride NaBH4 10.6 Lithium borohydride LiBH 4 18.2 Lithium hydride LiH 13 Magnesium hydride MgH 2 7.6 Sodium Aluminum hydride NaAlH 4 7.4 During the hydrolysis of this compound, water becomes another source of H 2 . The extra water added has to be considered as a reactant and must be taken into account in the H 2 generation percentage. Out of the four moles of H 2 produced, half comes from NaBH4 and half comes from H 2 0 [Amendola, S. C. et al, 2000a)]. Thus, the hydrolysis using excess H 2 0 leads to superior storage capacity than the total amount of H 2 stored in the NaBH4 itself [Jacques, S. et al, (1970)]. If 1 mole of NaBH4 (37.8g) produces 4 moles of H 2 (8.00g), a theoretical storage density of about 21.2 wt% H 2 based on the weight of NaBH4 could be obtained. The theoretical specific gravimetric energy density of 30.55 kWh/kg was calculated in Appendix 41 A for 200 g of NaBH4, assuming 4 moles of H 2 are generated per mole of NaB02. The gravimetric density has been reported for various NaBH4 concentrations [Mohring, R. M. et al, (2003)]. The actual maximum H 2 storage capacity has been measured to be 6.7 wt% for a solution containing about 34.5 wt% NaBH4 and 3.3 wt% NaOH at 25°C [Hua, D. et al, (2003)]. The DOE system gravimetric energy targets for 2010 and 2015 are of 6 wt% (2 kWh/kg) H 2 and 9 wt% (3 kWh/kg) H 2 respectively [Department of Energy, (2004)]. It appears that the NaBH4 H 2 generation and storage system could reach 21.2 wt% and exceed the DOE FreedomCAR goals when water produced by the PEMFC is recycled back to the stored NaBH4 slurry. However, this decreases to 12 wt% when calculated based on the resulting NaB0 2 [Davis, B. et al, (2005)]. If the by-product hydration level and the quantity of alkaline additive would be taken into account, it is very likely that the system storage gravimetric energy density would be too low to meet the DOE targets. The balance of plant (BOP) also has a significant impact on the storage system capacity. For a 25 wt% NaBH4 solution, the gravimetric storage density of the system drops of about 53 % when a BOP of 40 wt% is assumed [Mohring, R. M. et al, (2003)]. Similarly, assuming a storage system BOP taking up 20 % of the size of the tank, the volumetric storage density of a 25 wt% NaBH4 solution decreases of about 16 %. 2.3.2 Volumetric Capacity In theory, a litre of 30 wt% NaBH4 solution yields about 67 g of H 2 . Assuming the ideal gas law, this is equivalent to 744.5 L of H 2 , which is an excellent system volumetric capacity. The volume of NaBH4 solution of different concentrations required to store varying amounts of H 2 was reported [Mohring, R. M. et al, (2003)]. As the NaBH4 solution concentration increases, more H 2 is stored in a given volume. When combined with a proton exchange membrane fuel cell (PEMFC, at 60 % efficiency), the NaBH4 solutions have the potential to exceed the volumetric energy density of gasoline being burned in an internal combustion engine (ICE, at 30 % efficiency) [Davis, B. et al, (2005)]. Furthermore, it was calculated in Appendix A that the theoretical volumetric energy density contained in a 20 wt% NaBH 4 0.2 M NaOH solution generating H 2 to feed a PEMFC is about 1.4 kWh/L of solution. This exceeds the energy density of a PEMFC feed with H 2 at 200 atm and 25°C, which is about 0.54 kWh/L of H 2 . The DOE FreedomCAR system energy density targets are of 1.5 kWh/L (0.045 kg/L) for 2010 and of 2.7 kWh/L (0.081 kg/L) for 42 2015 [Department of Energy, (2004)]. It is important to note that these are system targets and that the material energy density must be greater in order to meet the system requirements. Table 2.8 compares NaBH4 volumetric storage ability with that of other borohydrides. The values are representing the volume (L) required to store 1 kg of H 2 . The additional equipment required to contain the hydride and generate the H 2 was not taken into account in those volumetric calculations. Table 2.8: Volume Required for the Storage of 1 kg of H 2 from Different Borohydrides. Name Formula Volume to Store 1 kg H 2 [1] Sodium borohydride (30 wt%) NaBH4 + H 2 0 15 Sodium borohydride NaBH4 9.5 Lithium borohydride LiBH 4 8.1 Aluminum borohydride A1(BH4)3 11 The volumetric capacity of NaBH4 is only slightly lower then some other hydrides, but it is still very competitive has it possessed other advantageous characteristics that other hydrides do not possess. Of the available hydrides, NaBH4 has one of the lowest heat of reaction per mol of H 2 produced. It is stable compared to other hydrides so its hydrolysis is less violent and easier to control, while its handling is unproblematic. Furthermore, NaBH4 has a simple structure and is relatively easy to use. 2.3.3 Fuel Comparison Figure 2.7 gives a comparison of the theoretical volumetric and gravimetric energy densities of various fuels and H 2 carriers. 43 1 0 o Natural Gas Gasol ine it 7 Chemical Hydrides Compressed K, (69 MPa) Compressed H 2 (34 MPa) D O E 2 0 1 5 T a r g e t D O E 2 0 1 0 T a r g e t A Metal Hydrides 4 6 8 1 0 Gravimetric Storage Density (kWh/kg) 1 2 1 4 Figure 2.7: Comparison of the Gravimetric and Volumetric Energy Densities of Various Fuels and H 2 Carriers (Modified from [Chandra, D. et al, (2006)]). The theoretical storage system gravimetric and volumetric energy densities are very different than the actual values and the weight and volume of the equipment needed to generate the H 2 has a significant impact on the storage system gravimetric and volumetric capacities. For example, a storage system with a volume of 2 L and a weight of 2 kg, holding 1 L of solution composed of 30 wt% NaBH4, and 3 wt% NaOH, would have a storage efficiency of 3.35 wt% H 2 and a specific gravimetric energy density of 1.34 kWh/kg [Larminie, A. et al, (2003)]. Optimization of the solution composition and integration of the storage system water and heat utilisation is needed to maximize the actual system specific and volumetric energy densities. Hence, it is clear that significant improvements in actual storage efficiencies are still required to attain the DOE energy density targets. 2.4 Hydrolysis of NaBH* for H 2 Generation and Storage A description of the non-volatile and non-flammable NaBI-L, solution hydrolysis reaction is provided in this Section. The overall performance of the NaBH4 storage system is affected by numerous parameters, 44 including pH, catalyst material and catalyst loading, temperature, NaBH 4 concentration, electrolyte concentration, and steam. The individual impact of these parameters is reviewed separately in the following sections. 2.4.1 Hydrolysis The chemical reaction of a hydride with water to produce H 2 gas is defined as hydrolysis. The hydrolysis of NaBH4 at atmospheric pressure is described as follows. NaBH4(aq) + 4H 2 0(l)^4H 2 (g) + NaOH(aq) + H3B03(aq) (Eq. 2.63) NaOH(aq) + H3B03(aq)->NaB02.2H20(aq) (Eq. 2.64) Overall: NaBH4 (aq) + 4H 20 (1) —• 4H2(g) + NaB02.2H20 (aq) (Eq. 2.65) As per Eq. 2.63, the production of sodium hydroxide (NaOH) creates an increase in pH. In Eq. 2.64, boric acid (H3B03), which is a weak electrolyte, reacts further to generate a hydrous metaborate salt, NaB02.2H20, which is the predominant species. As Eq. 2.64 is slower, the solution becomes increasingly alkaline as the overall reaction progresses [Schlesinger, H. I. et al, (1953)]. Since NaBH4 exists as a complex ion in solution, the aqueous reactions are more realistic [Suda, S., (2003)]. In the early studies, the hydrolysis is reported to obey first order kinetics with respect to NaBH4 concentration at a constant pH [(Brown, J. B., (1957), Levy, A. et al, (1960), Davis, R. E. et al, (1962)]] and to obey first order kinetics with respect to H + ion [(Brown, J. B., (1957), Davis, R. E. et al, (1962)]. In a buffered media, the hydrolysis followed pseudo-first order kinetics [Kaufman, C. M. et al, (1985)]. Pseudo-first order kinetic models are typically used to find second order reaction by dividing the effective first order constant by the concentration of the reactant in excess. More recently, it is reported that the hydrolysis reaction follows zero-order kinetics with respect to NaBH4 concentration when using Ru or Ni based catalysts [Amendola, S. C. et al, (2000b), Hua D. et al, (2003), Pinto, A. M. F. R. et al, (2006)]. The NaBH4 hydrolysis reactions are not easily reversible, yet, recyclability is a DOE H 2 storage system requirement. After hydrolysis, the recovered solution contains crystalline NaB02, and sodium hydroxide 45 (NaOH) dissolved in water. The NaB0 2 by-product is a basic hydrated solid, which looks like a paste. It possesses a low solubility limit. Hence, its formation inhibits the heat and mass transfer properties of the reaction medium, the H 2 generation reaction rate, and eventually results in clogging the hydrolysis reactor. Hence, the solubility limit of NaB0 2 constitutes an important challenge that needs to be addressed to optimize the H 2 storage and generation system. The overall reaction is exothermic and liberates 75 kJ/mole of H 2 formed at 25°C [Amendola, S. C. et al, (2000)]. The enthalpy of reaction, A H , was reported to be about -218 kJ/mole [Larminie, J. et al, (2003), Wee, J.-H., (2006)] at unknown conditions. Using data reported in the literature [Dean, J. A., (1999), Li, J. et al, (2000)], at standard conditions, the enthalpy of formation of B02" (aq) (AH 0) and the Gibbs energy of formation of B0 2 ' (aq) (AG 0) were calculated to be -248.87 kJ/mole and -318.93 kJ/mole respectively. Hence, the reaction is very energetically favourable and no energy input is required to produce H 2 . The hydrolysis of NaBH4 liberates a determined amount of H 2 at ambient temperature and pressure with no side reactions or hazardous by-product. At ambient conditions and in alkaline media, the solution is stable. The H 2 generation rate is affected by many factors. As it is shown in Sections 2.2.2, 2.2.3, and 2.2.4, the hydrolysis of NaBH4 can be accelerated by the addition of an acid, by a catalyst and by an increase in temperature. 2.4.2 Effect of pH A) pH Equilibria of Borates In aqueous solution, boron is in the negatively charged borate ion or un-dissociated boric acid form, depending on the concentration and pH of the solution. For example, the pH of borax solution increases with increasing concentration and decreases with increasing temperature [Kirk, R. E. et al, (1992)]. Over the entire pH range, there are four borate compounds which can be formed in aqueous solution: boric acid, sodium pentaborate, sodium tetraborate, sodium metaborate. The addition of NaOH to a boric acid solution increases the pH and sodium pentaborate is formed. It has a comparably higher solubility than boric acid. At pH ranging from 7 to 10, sodium tetraborate, which has a lower solubility, is formed. At pH greater 46 then 10, sodium metaborate (NaB02), which has the lowest solubility, is formed. Thereby, it is the prevalent compound present in the highly alkaline solutions which are investigated in this study. However, traces of other polyborates, such as tetraborates and pentaborates, can also be present. The borate species generally stable in aqueous solutions are the othoboric acid (B(OH)3), the monoborate ion (B(OH)4), the triborate ion(B303(OH)4") and the polyborate ion(B405(OH)4"2). Boron compounds can exist in the form of mononuclear boric acid, a very weak acid, in water. In the free state, two boric acids exist: metaboric (HB02) and orthoboric (B(OH)3). At specific conditions, both can be hydrated or dehydrated to form the other. For example, in a dilute aqueous solution, boric acid exists as orthoboric acid H 3 B0 3 or B(OH)3 at pH lower then 7. Thus, the metaboric acid is rapidly converted to orthoboric by hydration under acidic conditions [Jigong, L. et al, (1999)]. Hence, depending on temperature, concentration and pH, boric acid weakly dissociates to form monoborates and polyborates in solution. For instance, boron occurs as a dissociated borate ion, B(OH)4", at pH greater than 9 [Jigong, L. et al, (1999)]. In aqueous solutions, boric acid reacts with strong bases and produce B(OH)4' according to the following equilibrium reaction: In concentrated solutions, boron condenses and polymeric ions are formed. The prevalent polyborate is the trimeric hydroxyborate [Jigong, L. et al, (1999)]: Numerous studies were carried to determine which ions are present in borate solutions, giving a sense of the system complexity. For example, the polyborates most likely to be found in concentrated solutions in the pH range from 7 to 9 were the focus of an entire study [Jigong, L. et al, (1999)]. It has been reported that various alkali polyborate species are likely to be formed in concentrated alkali borate aqueous solutions [Momii, R. K., (1967), Hampton, D. S. Jr. et al, (1972), Mesmer, R. E., (1972), Covington, A. K. et al, (1973), Salentine, C. G., (1983), Botello, J. C. et al, (2003)]. B(OH)3 + NaOH <-> B(OH)4" + Na+ . (Eq. 2.66) 2B(OH)3 + B(OHV <-+ B303(OH)4" + 3H20. (Eq. 2.67) 47 B) Effect of pH on the NaBH 4 Solution Stability Acid, such as hydrochloric acid (HC1), can be added to decrease the pH and, thus, improve the NaBH4 hydrolysis reaction rate. Generally, stronger acids produce greater H 2 generation. However, boric oxide, which is derived from a weak acid, was found to be as effective [Schlesinger, H. I. et al, (1953)]. For convenience, ideally, the acid accelerator should be able to be safely mixed and compressed with the solid NaBH4 to form pellets. Also, it was determined that the H 2 generation rate depends on the pH and not on the type of acid added. Thus, it is possible to accelerate the hydrolysis by adding an acid and terminate it by adding a base, such as NaOH. It was noted that, since at stoichiometric water addition efficient wetting is difficult to obtain, excess water is required to maximize the contact between NaBH 4 and the acid solution. Otherwise, some NaBH 4 remains unreacted and the reaction yield remained low [Schlesinger, H. I. etal, (1953)]. It was established that the rate of hydrolysis of the BH4" ion at a specific pH within the 12 to 14 range is independent of the nature of cation present. At lower pH values, the rate of hydrolysis dependency on the buffer system used was uncertain. However, it was confirmed that the alkali stabilization effect results from the suppression of the hydrolysis equilibrium of NaBH 4 [Weiren, R. et al, (1998)]. The rate of hydrolysis of NaBH4 solutions in water depends on the pH and temperature of the solution. The solution half-life (ti/2) was found to empirically correspond to the following expression [Kreevoy, M. M. et al, where t1/2 is in minutes and the temperature, T, is in K. Hence, a high alkali concentration extends the NaBH4 solution half-life at a given temperature. The stability of KBH 4 in alkaline solutions was previously investigated and compared to that reported for NaBH4 [Brown, B. B., (1957)]. (1979)]: (Eq. 2.68) 48 C) Effect of Electrolyte Concentration on the NaBH 4 Hydrolysis As seen in the last Section, the pH, has a direct impact on the H 2 generation, rate from the NaBH 4 hydrolysis. Decreasing the concentration of NaOH increased the H 2 generation rate when using a Ru-based catalyst [Amendola, S. C. et al, (2000a)]. As the OH" ion strongly complex the water molecules, high NaOH concentrations reduced the activity of water. Low concentrations of NaOH lead to higher H 2 generation rates and leave more by-product in solution, which directly improves the H 2 storage system efficiency. However, in another study using a Pt-LiCo02 catalyst, it was found that the H 2 generation rate was independent of the NaOH concentration [Kojima Y. et al, (2004a)]. Contradictorily, the H 2 generation rate increased with higher NaOH concentration when using a Ni based catalysts [Hua, D. et al, (2003), Pinto, A. M. F. R. et al, (2006)]. Those dissimilar results were attributed to different reaction mechanisms involved depending on the catalyst used. It was stipulated that, with Ni based catalyst, the OH' ions are involved in complex surface reactions. This effect compensated for the reduced water activity and lower by-product solubility at high NaOH concentrations. Whichever the catalyst is, one fact remains the same: as the reaction progresses, the quantity of NaB0 2 produced eventually exceeds its solubility limit in the alkaline aqueous solution, resulting in precipitation on the catalyst surface and prevention of further H 2 evolution [Hua, D. et al, (2003), Pinto, A. M. F. R. et al, (2006)]. To date, there seems to be no consensus on the optimum hydroxide concentration needed to carry out the catalytic hydrolysis reaction Eq. 2.65. A 0.01 M concentration of NaOH has been considered sufficient to prevent the NaBH4 decomposition, but dilute enough to cause negligible NaB0 2 and NaBH 4 solubility changes [Stockmayer, W. H. et al, (1955)]. Even though small amounts of alkali are sufficient to obtain a stable NaBH 4 solution at room temperature, alkali concentrations ranging from 4 to 25 wt% (about 1 to 6 M) were used to study this H 2 storage and generation system. NaOH is predominantly chosen to stabilize the solution as it is more readily available and-affordable than other hydroxides. 49 2.4.3 Effect of Catalysts As acid reagents are not desirable, catalysts are preferred to control the hydrolysis reaction. They are required in smaller quantities then acid accelerators and are as effective. Also, they do not directly alter the pH of the solution. With catalysts, the reaction proceeds at a more constant rate until completion. As in the case of acid accelerators, excess water improves the H 2 yield. This is because the reaction by-product is hydrated. Various catalysts were proposed for accelerating the NaBH4 hydrolysis. A summary table of the catalyst studied for the NaBH 4 hydrolysis since 2000 is presented in the literature [Wee, J.-H., (2006)]. Some of the most important catalysts are discussed in more details in this Section. A) Noble Metal Catalysts Although expensive materials, noble metals such as platinum (Pt) and ruthenium (Ru) are excellent catalysts for the NaBH4 hydrolysis. To limit the amount of noble metal used and improve the effective surface area, the noble metal is typically mounted on a bead support. Some of the supported Pt and Ru catalysts are discussed in this Section. a) Supported Platinum Carbon supported platinum catalysts (Pt/C), synthesised by an impregnation method, were adopted to the hydrolysis of NaBH4 solutions [Wu, C. et al, (2004)]. By comparing different carbon supports, it was found that the specific area of the support did not affect the H 2 generation performance. Increased Pt loading enhanced the H 2 generation rates and efficiencies. Unfortunately, the temperature under which the experiments were conducted was not reported. Higher Pt utilization was obtained with small particle size and uniform particle size distributions. The catalyst active area and adsorbability was enhanced with large quantities of micropores. The Pt/C catalyst showed good H 2 generation rates and the H 2 generation efficiencies were near 100 % after 20 minutes [Wu, C. et al, (2004)]. 50 b) Supported Ruthenium By the investigation of various metal salts, it was concluded that Ru and rhodium (Rh) salts rapidly liberate H 2 from NaBH4 solutions. However, Ru is less expensive and responds faster than metallic Rh. In comparison, studies showed that Co and Pt catalysts only have a modest H 2 generation rate [Amendola, S. C. et al, (2000a)]. High surface area Ru, supported on anion exchange resin beads, was investigated. Ion exchange beads have high stability in strong alkaline solutions [Amendola, S. C. et al, (2000a)]. Anionic resins gave better results than cationic resins as catalyst support [Amendola, S. C. et al, (2000b)]. The volume of H 2 generated by Ru catalyzed NaBH4 hydrolysis increased linearly with time. It was deduced that BH4" was adsorbed on the Ru catalyst surface. The activation energy for Ru catalyzed hydrolysis of NaBH4 was calculated to be 47 kJ/mol [Kojima, Y. et al, (2002)]. This value compares favourably with other reported activation energies found for NaBH4 hydrolysis catalyzed with other metals, which are summarized in Table 2.9. Table 2.9: Comparison of Some Catalyst Activity (Modified from [Kaufman, C. M. et al, (1985), Amendola, S. C. et al, (2000a)]. Metal Activation Energy [kJ/mol] Ru 47 Co 75 Ni 71 Raney Ni 63 From a systematic study using different loadings of a Ru catalyst and varied reactant concentrations, it was determined that the hydrolysis was limited by the reaction at the catalyst surface [Richardson, B. S. et al, (2005)]. Also, using a Ru catalyst supported on ion exchange resin IRA 400, it was found that the H 2 generation curve was divided in four distinct stages [Xia, Z. T. et al, (2005)]. First, the activity of the catalyst increased and the H 2 generation is improved. Secondly, as the NaBH4 concentration decreased, the H 2 generation went down. Thirdly, the catalyst particles were pulverized by the formation of NaB02. This increased the available catalyst surface and favoured the hydrolysis reaction. Finally, as the concentration of NaBH 4 decreased and that of NaB0 2 increased, the H 2 generation kept decreasing. Therefore, it seems 51 like NaB0 2 is beneficial for the catalyzed H 2 evolution. In addition, it was noted that Ru on alumina was unstable in aqueous alkaline NaBH4 solution, but that Ru on metal titanium (Ti) was stable [Gervasio, D. et al, (2005)]. B) Non-Precious Transition Metal Catalysts Among different transition metal salts, it was determined that, on a qualitative basis, the best hydrolysis catalyst accelerators were Ni salts, and that, on a quantitative molar basis, Ni and Co salts had higher catalytic efficiencies than monoprotic salts [Kaufman, C. M. et al, (1985)]. For example, after 10 minutes of reaction time at 21.4°C, 1 mol % of Ni generated 0.13 mol/dm3 of NaBH 4 while 1 mol % of Co generated 0.11 mol/dm3 of NaBH4. Under the same conditions, 4 mol % of Ni generated 0.17 mol/dm3 of NaBH 4 while 4 mol % of Co generated 0.14 mol/dm3 of NaBH4. Each of these salts is discussed separately. Other types of transition metal catalysts, including filamentary Ni mixed Co, metal hydrides and metal-metal oxides are also reviewed. a) Cobalt Boride Among the chloride salts of Mn, Fe, Co, Ni and Cu, Co, which were studied as catalysts for the hydrolysis of NaBH4 solutions, Co resulted in the highest H 2 generation rate [Schlesinger, H. I. et al., (1953)]. All of these chloride metal salts promptly formed dark suspensions of boride precipitates. These precipitates had noticeable catalytic action, especially in the case of Co. It was deduced that the active catalyst was the precipitate formed, a cobalt boride, Co2B [Schlesinger, H. I. et al, (1953)]. This was confirmed by some researchers [Levy, A. et al, (I960)] but contradicted by others, as the formation of metal borides in aqueous solution is not thermodynamically favourable [Kaufman, C. M. et al, (1985)]. Differences were found in the mechanisms and rate of the reduction of various metal salts. It seems that the catalysis is partly caused by the generation of protons during the metal salt reduction. Hence, in the presence of metal salts, the hydrolysis would be catalyzed by the metal salt and by the acid generated at the same time [Kaufman, C. M. et al, (1985)]. Although the chloride metal salt catalysts are active, the 52 dissolved metal ions tend to contaminate the aqueous spent NaB0 2 solution [Zhang, Q. et al, (2006)]. In a recent study, it was determined that Co-B catalysts have an activity for the hydrolysis of NaBH4 similar to that of Ru catalysts and that an increase in hydroxide ions concentration positively affected the activation of the Co-B catalyst [Jeong, S. U. et al, (2005)]. b) Nickel Boride Heat treated nickel boride catalyst, NixB (x = 4, 5), is easily prepared and inexpensive compared to noble metal catalysts. It is highly stable and enhances the H 2 generation from NaBH4 solutions. The heat treatment conditions were found to significantly affect the catalyst activity [Hua, D. et al, (2003)]. The H 2 generation rate of the NixB catalyzed hydrolysis of NaBH 4 in NaOH was compared with that of Ru/C catalyst. In the case of NixB rate of hydrolysis increased with increasing NaOH concentration, whereas in the case of Ru/C, the rate of hydrolysis decreased with increasing NaOH concentration [Hua, D. et al, (2003)]. Typically, it is assumed that the NaBH4 hydrolysis takes place by the reaction of borohydride ions with protons dissociated from water and that, in alkaline solution, the slow self-hydrolysis is due to the reduction of proton concentration. According to this hypothesis, the hydroxide ions would inhibit the catalyzed BH4" hydrolysis and greater NaOH concentration would reduce water activity and lower the solubility of the by-product NaB02. Interestingly, this explanation, which is valid in the case of Ru/C, does not hold for the NixB catalyzed NaBH4 hydrolysis. The accelerating effect of higher NaOH concentration on the NixB catalyzed NaBH4 hydrolysis was attributed to the fact that this reaction follows a different surface mechanism than Ru/C [Hua, D. et al, (2003)]. It was observed that, when using NixB, water is stoichiometrically consumed and gives rise to an equivalent NaOH concentration during the hydrolysis. The hydrolysis accelerating effect caused by the rise of NaOH concentration observed with NixB catalyst might be able to compensate for the H 2 0 concentration decrease and the accumulation of NaB0 2 over time. This would be very beneficial to the long term operational efficiency of the H 2 generation system. The activation energy using NixB was calculated to be 38 kJ/mole in 10 wt% NaOH and 1.5 wt% NaBH4, which is lower then the values reported for lower NaOH and greater NaBH4 concentrations [Hua, D. et al, 53 (2003)]. It is also quite low compared to the reported values for other metal catalysts listed in Table 2.6. This implies that it would require less energy to activate the NaBH4 hydrolysis reaction with NixB catalysts. Interestingly, it was recently discovered that a Ni powder based catalyst exhibited enhanced hydrolysis activity after-being recovered from previous use by washing, filtrating and drying [Pinto, A. M. F. R. et al, (2006)]. c) Filamentary Nickel Mixed Cobalt The filamentary Ni mixed Co catalyzed hydrolysis of NaBH 4 had a short response time and maintained the maximum H 2 generation rates for extended time periods due to its large surface area. Filamentary Ni is easy to handle and is inexpensive. It was found that the hydrophilic properties of stylene-butadiene-rubber make it an effective catalyst binder. The cyclic properties of the Ni mixed Co catalyst on the hydrolysis of NaBH4 in alkaline solution were also studied. The H 2 generation decreased overtime as a film was formed on the catalyst surface. About 76 % of the initial H 2 generation remained after 200 cycles [Kim, J.-H. et al, (2004b)]. It is suspected that the precipitation of NaB0 2 deteriorated the catalyst. With cycling, the hydrated NaB0 2 stacked upon the pasted catalyst and the catalyst surface oxidizes and eventually degrades. In addition, the catalyst agglomerated over cycling and its surface area decreased [Kim, J.-H. et al, (2004b)]. d) Metal Hydrides In general, metal hydrides catalysts result in inferior hydrogenation rates due to their mediocre kinetic properties. However, it was found that it is not the case for Mg2Ni, a high-temperature hydriding alloy, as the hydride phase of Mg2Ni is different from that of other metal hydrides. The hydride phase gradually forms from the surface towards the center of the particle [Suda, S. et al, (2001)]. A fluorination treatment, called the F-treatment, was investigated. It removes oxides to form fluoride and create hydride layers at the surface of the granular particles. The fluoride layer exhibits high affinity to proton uptakes. The F-treatment generates larger specific area then untreated particles. The hydride layer acted as the active site 54 for hydrolysis [Suda, S. et al, (2001)]. The kinetics of various fluorinated catalysts was compared. The experiments showed that the catalytic hydrolysis of F-treated Mg2Ni and F-treated Mg 2NiH 4 was superior to that of the untreated catalysts in both cases. e) Metal-Metal Oxide Compared to other metals and metal oxides, Pt-LiCo02 was found to be a good catalyst for the hydrolysis of NaBH4 solutions. The supercritical C 0 2 method generated superior catalysts then the conventional impregnation method [Kojima, Y. et al, (2002)]. The catalyst activity increased with the surface area of the Pt metal. The H 2 generation rate of Pt-LiCo02 increased with time, while that of a reaction with a Ru catalyst remained constant [Kojima, Y. et al, (2002)]. Pt-LiCo02 produced 100 % of the theoretical amount of H 2 using excess water [Kojima, Y. et al, (2004a)]. Nanosized Pt-LiCo02 also resulted in a superior catalytic activity than a mixture of Pt and LiCo0 2 when applied to the hydrolysis of LiBH 4 [Kojima, Y. et al, (2006)]. Furthermore, when using a PtRu-LiCo02 supported catalyst, instead of Pt-LiCo02 or Ru-LiCo02, the catalyst efficiency was doubled for concentration of NaBH 4 up to 10 wt% [Krishnan, P. etal, (2005)]. C) Organic Catalysts Organic pigments consisting of carbon and other non-metals were found to be effective catalysts to control the rate of H 2 evolution from NaBH4 [Linkous, C. A. et al, (2003-4)]. These compounds have been used as photocatalysts for water decomposition. The organic pigments investigated were fused hetero-aromatics, which possess distinctive molecular and electronic structures. Their structure is constituted of extended fused aromatic system with one or more hetero-atoms affiliated with the pi-electron cloud [Linkous, C. A. et al, (2003-4)]. The hetero-atoms lower the energy of the first excited state energy level, making it predisposed to reduction. The H 2 generation from NaBH4 solutions of the various pigments was evaluated using buffered NaBH4 solutions. Constant pigment catalyst loadings, 100 mg, were compared with an equivalent weight of cobalt powder. Pyranthrenedione, indanthrene Gold Orange and perylene 55 dime pigments were as good or better catalysts than the conventional cobalt powder. Pyranthrenedione increased the rate of H 2 generation from buffered NaBH4 solutions 6.5 times more then that without catalyst. D) Catalyst Comparison Different researchers used diverse solute concentrations to perform their experiment. Missing information regarding the experimental conditions, such as catalyst loading used, temperature, volume of solution, makes an overall catalyst comparison difficult based on the currently available data. Nevertheless, Table 2.10 attempts to compare the various catalysts reviewed in this section. Noble metals (Pt, Ru) are among the suitable catalysts, but other less expensive materials, such as Ni, are effective [Linkous, C. A. et al, (2003-4)]. Fabrication, treatment and ease of handling should be considered as they impact on the catalyst cost. Since the hydrolysis reaction rate is limited by catalyst surface area, more research needs to be done to enhance the catalyst specific surface area and optimize the catalyst loadings. Lower catalyst loadings would reduce the system cost. The effect of NaOH concentration on the NixB catalyst activity should be further examined. Research on catalyst durability has been limited. The catalysts long-term degradation behaviour due to by-product coating has to be studied in more detail. In addition, the catalyst performance at elevated pressures and temperatures, its chemical stability in hot caustic environment and its mechanical resistance to the formation of the bulky NaB0 2 hydrated molecules needs further investigation. 56 Table 2.10: Catalysts forNaBH4 Hydrolysis. Catalyst C h e m i c a l Catalyst /Support H 2 generated after 5 m i n T C o m p o s i t i o n Reference Formula ( A m o u n t [g], L o a d i n g [wt%)] [1] PC] [wt% N a B H 4 , w t% N a O H ] Supported P la t inum P t /C 0.1 g - 5 w t % 2 - 10 ,5 [ W u , C . et al, (2004)] Supported Ruthen ium R u / C 0.5 g - 2 wt% 0.5 - 20, 1 [Richardson, B . S. et al, (2005)] R u / I R A 4 0 0 0.02 g - 5 wt% 0.0075 25 1 0 , 1 0 [ X i a , Z . T. etal, (2005)] Coba l t Bor ide C o B 0.05 g 1.6 25 20,10 [Jeong, S . U . etal., (2005)] N i c k e l Bor ide N i x B 0.1 g 0.1 30 1.5, 10 [Hua , D . etal, (2003)] Filamentary N i m i x e d C o N i - C o 1 g - 20 wt% 0.425 30 10,0.01 M K O H . [ K i m , J . - H . etal, (2004a)] M e t a l - M e t a l O x i d e P t - L i C o 0 2 0.239 g , 1 0 w t % 0.1 25 • 5 , 5 [Kr i shnan , P. et al, (2005)] P t R u - L i C o 0 2 0.125 g , 10 w t % 0.25 25 5 , 5 Fluorinated M e t a l Hydr ides F - M g 2 N i R , 1.91 m 2 / g surface area 0.83 - 50 g / L N a B H 4 , 10 [Suda, S. etal, (2001)] Organic Pyranthrenedio ne 0.1 g 0.055 - p H buffer at 11 [L inkous , C . A . et al., (2003-4)] 2.4.4 Effect of Temperature To maximize the H 2 storage system energy density, it is important to minimize the size of the H 2 storage tank needed to instantaneously deliver H 2 to the fuel cell and the reactor start-up speed is directly affected by the hydrolysis temperature rise [Zhang, Q. et al., (2006)]. The temperature at which the hydrolysis is carried has a direct impact on the H 2 generation rate. The H 2 generation rate increases dramatically with temperature [Amendola, S. C. et al, (2000a), Pinto, A. M. F. R., et al, (2006)]. This is partly due to the increase in self-hydrolysis at high temperatures. Thus, from a reaction kinetics point of view, it is favourable to run the hydrolysis reactor at high temperatures. However, as explained in Section 2.3.2 B), high temperatures have a negative impact on the NaBH4 solution stability. When the vehicle is shutdown the NaBH4 solution will gradually cool down, but some self-hydrolysis will occur. Hence, a storage tank will be needed to store the small quantity of H 2 produced during the system shutdown. Since the hydrolysis of NaBH4 is exothermic, no preheating is needed and no heat sources are necessary to sustain the hydrolysis reaction. Thus, the hydrolysis reactors can be operated auto-fhermally [Zhang, Q. et al, (2006)]. For a peak H 2 generation rate of 2.5 g/s, derived for a vehicle with 68 kW net power and 24 % overall efficiency, the storage system must be designed to reject about 93 kW of heat. To maximize the cooling efficiency, it would be preferable to have a high reaction temperature to allow for a greater difference between ambient and rejection temperatures. Thus, for cooling efficiency purposes, it would be beneficial to carry the hydrolysis with steam rather than liquid H 2 0, as it will be discussed in Section 2.4.6. However, the H 2 may require cooling as the PEMFC have a limit on the allowable reactant inlet temperature, typically about 80°C. Cooling the H 2 before delivery to the fuel cell has the added benefit of increasing the relative humidity of the H 2 . The excess water provided for the hydrolysis cools the system and slows the reaction. If the reaction occurs too rapidly, the reaction vessel can become very warm due to the large amount of heat generated. 58 2.4.5 NaBFLt Concentration The range of a FCV depends on the amount of H 2 stored on-board and for a NaBH 4 H 2 storage system, the range is determined by the initial mass fraction of NaBH4 in the solution. The NaBH4 solution concentration and the mass of the solution carried on-board would then fix the FCV range. Another key design consideration for a FCV is weight which is directly affected by the NaBH 4 storage system mass. Increasing the initial mass fraction of NaBH4 will increase the mass of the solution, assuming that no fuel cell produced water is available for reuse in the hydrolysis of NaBH4. If the fuel cell produced water was reused, the mass of the storage system would be lower as less water would initially need to be stored on board. Thus, to minimize the mass of H 2 0 to be carried on board, the initial concentration of NaBH4 in the solution should be as large as practical. However, increasing the mass fraction of NaBH 4 can have a negative impact on the H 2 generation rate. Low concentrations of NaBH4 typically result in greater H 2 generation rates [Amendola, S. C. et al, (2000a), Pinto, A. M. F. R. et al, (2006)]. At 25°C and using a Ru catalyst, the H 2 generation rate was maximized in the range of 5-15 wt% NaBH4 [Amendola, S. C. et al, (2000a)]. This is partly because at low NaBH4 concentration, the low solution viscosity reduces the mass transport losses and allows more NaBH4 and water to contact the catalyst surface. Another reason is that at high NaBH 4 concentrations, while the reaction proceeds, the NaB0 2 by-product will eventually exceeds its solubility limit, precipitates on the catalyst surface and slow down the hydrolysis reaction [Amendola, S. C. et al, (2000a), Pinto, A. M. F. R. et al, (2006)]. It was reported that hydrolysis reactors were susceptible to clogging at NaBH4 concentrations above 25 wt% [Richardson, B. S. et al, (2005)]. Hence, decreasing the mass fraction of NaBH4 positively impacts the H 2 generation rate. High initial H 2 generation rates are crucial to quickly respond to the H 2 demand for FCV acceleration. 2.4.6 Steam Hydrolysis of N a B H 4 Instead of using liquid water or water vapour at ambient temperature, H 2 could be produced by hydrolysis with steam at high temperature. At these conditions, no acid addition or catalysts are necessary, 59 but excess steam is preferable to improve the reaction yield [Aiello, R. et al., (1999)]. It was found that the yield depended strongly on temperature and, to a lesser extent, on the steam flow rate. At 110°C, close to 100 % of the theoretical yield was obtained for the hydrolysis of NaBH4. Interestingly, the highest H 2 yields were obtained at the lowest temperature. At higher temperatures, the hydrated borate by-product released its hydration water and formed a solid layer on the unreacted hydride, inhibiting the reaction rate [Aiello, R., et al. (1999)]. The fastest rate of H 2 generation from 1 g NaBH4 was obtained at 110°C and a flow rate of 0.1 ml steam/min. However, only 88 % of the theoretical yield was reached. On the other hand, the hydrolysis of 1 g NaBH4 at 110°C and 0.035 ml steam/min took place at a slower rate but a yield of 99 % was reached. The differences in the rates of H 2 generation were attributed to mass transfer constraints [Aiello, R., et al. (1999)]. It was estimated that the exothermic hydrolysis reaction could provide the heat of vaporization necessary for steam generation, but a heat source would be required to start the reaction. It was proposed to use the excess heat to generate additional H 2 from the endothermic dehydrogenation of a metal hydride. 2.5 NaBH 4 Applications 2.5.1 NaBH 4 Hydrolysis Reactors The main engineering component for the NaBH4 H 2 generator use in combination with a PEMFC (B-PEMFC) system is a hydrolysis reactor. Generators using these solutions can take several forms. Some ionic hydrides generators were previously designed to generate H 2 for a large range of applications, while various small-scale reactors were developed for laboratory chemical hydride testing purposes. A selected variety of these reactors is reviewed in this Section. A) Kipp Generator The Kipp generator basically consists of a column of hydride inverted in water as shown in Fig. 2.8. Water enters the bottom of the column and reacts with the hydride. The H 2 generated during the reaction 60 rises to the top of the column, where it is removed through a control valve. The pressure exerted by the H 2 over the water column controls the reaction rate [Linden, D., (1984)]. \ / Figure 2.8: Schematic Diagram of a Kipp Generator (Modified from [Mattson, B., (2005)]). It was shown that the Kipp generator concept can be adapted to the catalyzed hydrolysis of NaBH 4 solution [Amendola, S. C. et al., (2000a)]. In that case, the column contained a catalyst. The NaBH4 solution, pushed by differential pressure, enters the column bottom where it reacts with the catalyst. H 2 is delivered to the fuel cell as needed. When no H 2 is required, the pressure builds up in the generator, forcing the solution away from the catalyst, ending the reaction. Only simple controls are required to ensure safe operation of this low cost and compact reactor. However, in order to keep control over the reaction, the column must remain in the upright position [Linden, D., (1984)]. Otherwise, a small pump has to be added to meter the NaBH4 solution flowing to the tubular catalyst bed and make the reactor functionality independent of its orientation [Amendola, S. C. et al., (2000a)]. This reactor design exhibits a very rapid response to the fuel cell H 2 demand. 61 B) Remote Fuel-Cell Power Source The US Army Mobility Equipment Research and Development Command developed a high energy density, refuelable metal hydride fuel-cell system for remote, equipment operating at the 1 to 50 W level for long periods of unattended operation [Linden, D., (1984)]. No auxiliary equipment is required. Figure 2.9 represents a schematic diagram of the generator. H 2 to Fuel Cell Figure 2.9: Schematic of Remote Hydride System (Modified from [Linden, D., (1984)]). In this system, the water flows from the reservoir to the water chamber adjacent to a porous hydrophobic membrane. The water vapour diffuses through the membrane, and reacts with the hydride to release H 2 . The H 2 flows out of the chamber and feeds a fuel cell [Linden, D., (1984)]. At no load, no H 2 is consumed: The pressure in the reaction chamber builds-up and the water is forced out of the chamber and to the reservoir, which hinders the generation of H 2 . As H 2 is consumed from the fuel cell, the water level will self-adjust and H 2 will be produced as required [Linden, D., (1984)]. C) Solid Hydride and Liquid Water or Water Vapour In the solid hydride and liquid water laboratory test scale reactor, the hydride is contained in a cylindrical basket made of fine nickel mesh. A schematic diagram of the liquid water reactor is shown in Figure 2.10 (a). Water is feed from the reactor bottom using a syringe. The reactor is simply a thick wall bottle sealed with a cap which is fixed with a clamp. The water flows up through the basket, where it contacts the Porous Hydrophobic Membrane _T Water : Reservoir Bed 62 hydride. The H 2 produced flows out to a water trap collector and measuring system [Kong, V. C. Y. et al, (1999)]. H, Out H,0 In (b) H 2Out Cap Outer N i M e s h Basket Inner N i M e s h Basket Con ta in ing H y d r i d e T T H 2 0 In Glass Bottle •Cy l ind r i ca l Paper Fi l ter W i c k Figure 2.10: Schematic of Hydride Reactor (a) with Liquid Water, (b) with Water Vapor (Modified from [Kong, V. C. Y. etal, (1999)]). A solid hydride and water vapour reactor can be seen in Figure 2.10 (b), this experimental reactor is very similar to the solid hydride and liquid water reactor described above, except that, for reaction with vapour, the hydride basket is elevated on a platform to ensure liquid water does not directly contact the hydride. Excess water is added to the reactor bottom. Also, an additional mesh cylinder is added around the inner cylinder to contain the hydroxide reaction product, which passes through the inner mesh basket [Kong, V. C. Y. et al, (1999)]. Wetted water filter can be placed around the inside wall of the bottle to act as a wick in order to facilitate water evaporation. Like in the previous reactor, the H 2 produced is collected and measured in a water trap. D) Solid Hydride and Steam The hydrolysis reaction with steam requires a more complex reactor design with additional control in place as shown in Figure 2.11. 63 Figure 2.11: Schematic of Steam Hydrolysis Reactor System (Modified from [Linden, D., (1984)]). A programmable syringe pump feeds a steady rate of water to the reactor. To achieve nearly isothermal operation, the reaction system needs to be enclosed in a mechanical convection oven. The solid hydride is added on a mesh, covered with glass wool, fitted to the bottom of the reactor. To ensure that no steam contacts the hydride prior to reaction, the reactor needs to be purged with nitrogen while the oven and the water are heated. The water from the preheating coil bypasses the reactor until constant steam production was obtained. The N 2 purge is stopped when the steam reaches the desired temperature. Then the syringe pump is reset and restarted to deliver the water rate needed. Finally, the reaction bypass valve is opened to allow steam to enter the reactor and contact the hydride [Linden, D., (1984)]. E) Catalytic Reactors The catalytic hydrolysis reactors are all based on the same fundamental principle: the solution is brought in contact with a suitable catalyst to release H 2 and the H 2 generation stops when the solution is removed from the catalyst. The catalyst can be either dipped in the NaBH4 solution or the NaBH 4 solution can be injected on the catalyst. An example of a catalytic reactor where the NaBH4 solution is pumped to the hydrolysis reactor is shown in Figure 2.12. 64 Catalyst Chamber H 2 to fuel cell • T Reactor NaBH. | solution Figure 2.12: Reactor for Releasing H 2 from an Alkaline Aqueous Solution of NaBH4 (Modified from [Larminie, J. et al., (2003)]). The rate of the H 2 generation is controlled via the stabilized aqueous NaBH 4 solution pumping rate. The pump motor is switched on/off by a controller, which senses the H 2 pressure. The solution flows through the reactor so that the catalyst is continuously in contact with fresh solution [Larminie, J. et al, (2003)]. A similar strategy developed by NASA is to add solid NaBH4 tablets to an alkaline aqueous solution. As shown in Figure 2.13, the catalyst's surface is exposed to the solution to accelerate the reaction as required [Linkous, C. A. et al, (2004)]. Figure 2.13: Schematic of a H 2 Generation Vessel with Catalytic Control of Evolution Rate (Modified from [Linkous, C. A. et al, (2004)]). A 10 kW scale H 2 generator composed of a storage tank for NaBH 4 solution, a pump, a by-product storage tank, a separator and a hydride reactor which contained a honeycomb monolith coated with catalyst Catalyst Lever Control Catalyst 65 has been built and tested [Kojima, Y. et al, (2004b)]. A controlled hydride solution rate was pumped to the bottom of the reactor and flowed upward through the channel. When contacting the honeycomb monolith, the solution generated H 2 gas and NaB0 2 and the temperature increased. The H 2 gas and the NaB0 2 solution were separated by a separator, which also acted as a small storage buffer for H 2 gas [Kojima, Y. et al, (2004b)]. 2.4.2 B-PEMFC Systems To date, the B-PEMFC system has been evaluated by few researchers, and the integrated system performance and technical feasibility has been scarcely documented. Possible B-PEMFC applications include portable fuel cells, fuel cell Uninterrupted Power Supply (FCUPS) and emergency power sources [Mohring, R. M. et al, (2003)], whereas the present study focuses on automotive related applications. A) Hydrogen on Demand™ Hydrogen on Demand™ (HOD) is a proprietary B-PEMFC system developed by Millennium Cell Inc. which relies on the catalytic hydrolysis of NaBH4. It can supply H 2 pure enough to feed directly the PEMFC without the need for purification, compression or liquefaction. A schematic diagram of the system is provided in Figure 2.14. In this system, heat and water management are of great importance. The H 2 generation and storage system comprises two storage tanks for NaBH4 and NaB0 2 respectively, a pump, a catalyst chamber reactor, a liquid/gas separator, and a heat exchanger. The NaBH 4 aqueous alkaline solution is kept in a light plastic tank at ambient conditions. The release of H 2 from the NaBH4 solution occurs by pumping the fluid mixture through a tubular reactor containing a proprietary catalyst. The heat supplied by the exothermic reaction partially humidifies the H 2 gas by evaporating water, which is beneficial to the PEMFC membrane. The H 2 produced contains less than 0.5 ppm of CO, and no NO x or SOx [Mohring, R. M. et al, (2003)]. 66 NaBH4 1s^ Fuel Tank. NaB0 2 kWaste Tank^ Catalytic Reactor Borate to Discharge Tank H 2 0 recycling PEMFC Heat Exchanger H2gas and Steam Figure 2.14: Schematic Diagram of the Hydrogen on Demand System (Modified from [Mohring, R. M. et al, (2003)]). The H 2 is cooled down in a heat exchanger before being directly fed to a PEMFC or to a H 2 internal combustion engine. The rate of H 2 generation is controlled by pumping a desired amount of NaBH 4 solution in the catalyst reaction chamber. Stopping the flow of NaBH 4 solution to the catalyst chamber brings the H 2 production to an end. Water generated from the PEMFC can be recycled back to the solution storage tank. The Hydrogen on Demand™ system is claimed to have a volumetric storage of about 63 g H 2/L using a 30 wt% NaBH4 solution. This value is comparable and equivalent to 71 g H 2 /L for liquid H 2 and about 39 g H 2 /L for H 2 gas pressurized at 68.93 MPa [Mohring, R. M. et al, (2003)]. 2.4.3 Vehicle Prototype There exist opportunities for the integration of a NaBH4 storage system in a FCV that are unachievable with either liquid or pressurized H 2 storage systems. Among other things, the NaBH 4 solution tank geometry is flexible, the H 2 pressure and flow rate can be easily regulated by feedback control, and refuelling with fresh NaBH4 solution is rapid, simple and can be done at any time. The NaBH4 H 2 generation system was evaluated with a fuel cell emulator (FCE) to test its performance under true vehicle operating conditions without the need of a real fuel cell [Mohring, R. M. et al, (2003)]. It was found that the system was capable of responding to the H 2 demand for the application. A vehicle prototype taking advantage of the B-PEMFC technology was developed and is reviewed in this Section. 67 A) B-PEMFC Vehicle In 2001, Daimler Chrysler built a Town and Country Natrium® minivan prototype which incorporated a H 2 storage system that produces H 2 from the hydrolysis of NaBH4. The manufacturer claimed that the vehicle has a range of 500 km and can accelerate from 0 to 100 km/h in 16 seconds. As shown in Figure 2.15, Daimler-Chrysler's prototype vehicle power train fits under the vehicle floor and is very compact. It is significantly more space-efficient way than using compressed H 2 as the H 2 storage system and do not take room away from the passenger and storage compartments [Daimler Chrysler, (2002)]. The acceleration and power demands require a maximum H 2 flow rate of about 1000 slpm. The storage system is designed to produce sufficient H 2 to satisfy the real time demand from the NaBH 4 stored on-board. Millennium Cell Inc. developed the NaBH4 catalyst system for the Daimler Chrysler Natrium® vehicle. The response curve of the vehicle's storage system was found to be closely identical to the response curve of a pressurized H 2 gas storage system. Electr ic Motor v N a B H 4 Fuel A n d N a B 0 2 Was te Tanks H 2 Generat ion & Storage Sys tem Lithium Battery Figure 2.15: Daimler Chrysler Natrium Vehicle Power Train (Modified from [Wu, Y., (2003a)]). 68 C H A P T E R III P H Y S I C O C H E M I C A L PROPERTIES O F N a B 0 2 Although the thermodynamic properties of NaBH4 and NaB0 2 have been investigated by a number of researchers through the years, the transport properties of NaB0 2 alkaline and non-alkaline aqueous solutions were not systematically determined. In this Chapter, the literature on the determination of the physicochemical properties of aqueous NaB0 2 solutions with and without alkali additive and on the characterization of the precipitates formed from those solutions is reviewed. 3.1 N a B 0 2 Solution Characteristics The alkaline waste solution (i.e. containing NaB02) solubility and physicochemical properties such as viscosity and conductivity, affect the H 2 generation and storage system as well as the electrochemical recyclability of the by-product. For this reason, it is important to determine and analyze the solubility and physicochemical transport properties of the borate ion in aqueous solutions (1:1 electrolyte). In this section, the relevant electrolyte theory, NaB0 2 solubility and physicochemical properties, as well as the expected effects of temperature and of organic additives on the solution properties are reviewed. 3.1.1 Solubility A) N a B O z Solubility No solubility studies were found specifically on the NaB0 2-H 20-MOH system, where M represents Na, K or Li. The solubility of mixed alkali borate systems, including combinations of sodium pentaborates and diborate, ammonium pentaborate, tetraborate and diborate, and potassium pentaborate and diborate in water at 20, 30 and 40°C, was studied. It was found that the maximum solubility was controlled by the 69 mixture's pH [Roathnaum, H. P. et al, (1936)]. The solubility of the binary salt-water NaB0 2-H 20 system was reported by Urusova and Valyashko at elevated temperatures up to 400°C [Urusova, M. A. et al, (1993)]. The salt solubility continuously increased with temperature up to the melting point. The solubility of NaB0 2 was determined to be 61.9 wt% at 245°C and exceeded 67 wt% at 350°C. The most relevant solubility isotherms for the system under investigation are found in reviews of the ternary system of sodium oxide - boric oxide - water (Na20-B203-H20). This system was studied over the temperature range of 0 and 100°C [Blasdale, W. C. et al. (1938)]. Previous work was reviewed in an attempt to correct discrepancies reported in the early literature [Nies, N. P. et al, (1967)]. Since the ensemble of the isotherms reported was not always homogeneous among the various publications, the entire system was once more re-investigated from -30 to 100°C [Kocher, J. et al, (1970)]. It was suspected that the ternary system solubility values determined so far were inflated due to the presence of silica and carbonate, which increase the apparent solubility [Mellor, J. W., (1980)]. Based on those studies, there exist three known NaB0 2 hydrates: a tetrahydrate (1:1:8, Na 2O.B 20 3.8H 20, NaB02.4H20), a dihydrate (1:1:4, Na 2O.B 20 3.4H 20, Na02.2H20) and hemihydrate (1:1:1, Na 2O.B 20 3.H 20, NaBO2.0.5H2O). The transition temperature to the dihydrate occurs around 54°C, while the transition temperature to the hemihydrate is around 105°C [Blasdale, W. C , et al. (1938), Nies, N. P. et al, (1967)]. Hence, at 25°C, the stable hydrate is NaB02.4H20 in saturated aqueous solutions. This was also confirmed in a crystallographic study of the NaB0 2-H 20 system [Menzel, H. et al, (1943)]. The solubility of saturated NaB0 2 in water at 25°C was reported at around 22 wt% in terms of anhydrous solids [Blasdale, W. C, et al. (1938), Nies, N. P. et al, (1967)]. It was visually determined that in water at 25°C, the solubility of NaB0 2 is 28g/100g of water, while that of NaBH4 is about 52 to 55g/100g of water [Kojima, Y., et al, (2002), Kojima, Y. et al, (2004b)]. For comparison, those values were converted to weight percentages by calculating the amount of solute per 100 g of saturated solution. These values are equivalent to about 22 wt% and 36 wt% respectively. The solubility of NaB0 2 was measured at various NaOH wt% in water at 25°C [Amendola, S. C. et al, (2000a)]. It was observed that the solubility of NaB0 2 decreased as the NaOH content increased, but no values were reported. The solubility of NaB0 2 in a 10 wt% NaOH solution was reported as 2.16 mol/1 (14.2 wt%) at 25°C [Suda, S., (2003)] while the solubility of NaBH 4 in the same solution was reported to be 550 g/1 (55 wt%) at 25°C [Suda, S. et al, 70 (2001)]. To keep NaB0 2 in solution, the maximum composition of NaBH 4 has to be less than 16g/100 g of water (13.8 wt%) at 25°C [Kojima, Y. et al, (2002)] or 26 wt% at 60°C [Kojima, Y. et al, (2004b)]. Even if there is limited information on the solubility of aqueous alkaline NaB0 2 available in the literature, the published solubility data compiled is sufficient to undertake a comparative study and form the basis to carry out further experiments. B) NaBH 4 Solubility As previously explained, when no alkali stabilizers are present in an oversaturated NaBH4 aqueous solution, the dissolved NaBH 4 will react with water at a certain reaction rate and this will trigger NaBH4 dissolution. No data was found to compare the NaBH4 rate of dissolution to its hydrolysis rate. However, it might be possible to measure the total boron content of the solution, i.e. that of NaBH 4 and the NaB0 2 resulting from the hydrolysis reaction. On the other hand, if an alkaline additive is used to stabilize the NaBH 4 aqueous solution and hinder the hydrolysis reaction, then, the NaBH 4 solubility measured may not be representative of the solubility of NaBH4 in water. As seen in Section 2.4.2, the addition of alkali stabilizers greatly affects the solubility of NaB0 2 in water. Table 3.1 shows the NaBH4 solubility values reported in the literature for various solvents. In water at 25°C, the solubility of NaBH4 is almost two times greater than that of NaB02. Table 3.1: NaBH4 solubility in Various Solvents. Solvent Temperature [°C] Solubility [wt%] Reference Water 25 55 [Lide, D. R., (2005-6)] D imethy 1 formamide 25 18 [Dean, J. A., (1999)] Methanol 20 16.4 [Dean, J. A., (1999)] C) Effect of Temperature on the Solubility of N a B 0 2 Table 3.2 shows the solubility of NaB0 2 in aqueous solution at 25, 50 and 75°C, as well as at the transition temperature. It is clear that the solubility of NaB0 2 increases as the temperature increase. 71 Table 3.2: Solubility of NaB0 2 in Water at Selected Temperatures. Reference Temperature [°C] NaB0 2 [wt%] NaB0 2.2H 20 [wt%] [Blasdale, W. C. etal. (1938)] 25.0 22.00 30.45 50.0 35.02 54.20 54.0 37.85 58.58 75.0 43.80 67.78 [Nies, N. P. et al, (1967)] 25.0 21.60 33.43 50.0 34.10 42.77 53.6 36.90 57.11 75.0 42.20 65.31 [US Borax] 25.0 21.58 33.40 50.0 34.12 52.80 53.6 36.90 57.10 75.0 42.20 65.30 It can be observed that a transition temperature exists at about 54°C. Below the transition temperature, NaB0 2.4H 20 is known to be the stable compound. At the transition temperature, NaB0 2.4H 20 coexists along with NaB02.2H20. At temperature above the transition temperature, NaB0 2.2H 20 becomes the stable compound. The hydration level of the solute impacts on the physico-chemical properties of the solution. Above the transition temperatures less water is needed to hydrate the hydrolysis by-product. The hydration level of the NaB0 2 by-product affects the mass and heat transport properties of the solution and influences the theoretical gravimetric energy density of the H 2 storage and generation system. Hence, it would be preferable to operate the H 2 generation system at a temperature above the transition temperature to favour the formation of a less hydrated form of NaB02. 3.1.2 Single 1:1 Electrolyte Theory Many publications pertaining to the theories of aqueous single 1:1 electrolyte solutions are available [Erdey-Gruz, T., (1974), Horvath, A. L., (1985)]. The equations resulting from those theories include a number of simplifying assumptions which more or less affect the properties of the solution as concentration changes. 72 A) Conductivity The solution molar conductivity (A) is related to the solution specific conductivity (K) and the molar concentration of solute (C) as per: A = § ' (Eq.3.1) where K is in S/m, C in mol/m3 and A in S.m2/mol. At infinite dilution, the ions are far apart and do not influence each other. The solute limiting molar conductivity can be determined by extrapolating A to infinite dilution according to the Kohlrausch's law for strong electrolytes given by: i\ = K°-KCX//l, (Eq.3.2) where A 0 is the limiting conductivity and K is a constant. Using this equation, it is possible to obtain the limiting value of conductivity with an accuracy of 0.1% from conductivity data in highly diluted solutions (<0.0001 M) [Erdey-Gruz, T., (1974)]. The validity of the square root law was brought to a wider concentration range by the addition of another term: A = A° -KC^2 -BC. (Eq.3.3) The addition of a third term further extended the equation's validity to concentration as high as 0.1 M: A = A° -KC% -BC-DC^2. (Eq.3.4) In more concentrated solution of strong electrolytes, the cube-root law is more appropriate: 73 A = A ° - K C l / / \ (Eq. 3.5) Kohlrausch's law of independent ion migration states that the solution limiting conductivity is the sum of the limiting molar conductivities (k°) of ionic species (i) in the solution. For the dissociation of the 1:1 electrolyte AB, z+ and z. are unity and the following relationships can be obtained: AVt Bv <r+v+-Az+ - v_ Bz~ A 0 =Sv, .A° . (Eq.3.6) A value of 50.11 S.cm2/mol was reported for limiting ionic molar conductivity of Na+ [Dean, J. A. (1999)]. Applying the Nernst-Einstein relation at infinite dilution to a simple 1:1 electrolyte system where zf is unity gives: F2D° i = F U ° i = ^ T ^ ' (Eq-3-7) where F is the Faraday constant, R the universal gas constant and T the temperature. Knowing the limiting molar conductivity for one ion renders it possible to calculate the limiting mobility (u°) and the limiting diffusivity (D° ) of the same ion. Also, in a binary electrolyte, the sum of the cation and anion transference numbers is equal to one. Therefore, it is possible to calculate the ionic limiting transference numbers (t°) based on the limiting value of the ionic molar conductivities as: A°=t°A°. (Eq.3.8) 74 The Debye-Huckel-Onsager theory provides an expression for evaluating the theoretical value of the constant K: K = A + BA" . (Eq. 3.9) For a binary salt electrolyte solution at 25°C, A = 60.21 1/M I / 2 and B = 2.289xl0"5 S.m2/mol.M l /2 [Horvath, A. L., (1985), Snyder, K. A., (2005)]. The Onsager expression gives a theoretical explanation for the coefficients in the Kohlrausch equations and somewhat extends the concentration range [Horvath, A. L., (1985), Helmy, F. M., et al, (1987)]: A = A 0 - (0A° + cr)c^ (Eq. 3.10) . 8.205 x lO 5 where 6 = ^—, (Eq. 3.11) 82.48 cr = - , (Eq.3.12) rj0(eoT)/2 and where e0 is the dielectric constant of pure water (78.56), n 0 is the viscosity of pure water (0.008948 P) and T is the temperature in K [Horvath, A. L., (1985)]. This equation is applicable to the alkali salts in a solvent of high dielectric constant, such as water, in which strong electrolytes are completely ionized [Helmy, F. M. et al, (1987)] and is valid over concentrations ranging from 0.001 to 2 M [Shedlovsky, T., (1938)]. Many additional equation extensions and other empirical correlations were developed to predict conductivity in low concentration binary electrolyte systems but were not applied in this investigation [Horvath, A. L., (1985), Helmy, F. M. et al, (1987)]. 75 B) Dynamic Viscosity For a Newtonian fluid such as water, the dynamic viscosity is the proportionality constant between the velocity gradient in the direction perpendicular to the plane on which a shear stress is applied. Hence, it represents the resistance of a fluid to deform under shear stress. Based on the study of several electrolytes at constant temperature, Jones and Dole found that the dynamic viscosity was empirically related to concentration up to 0.1-0.2 M as follows: rjr =-L- = \ + AC/2 + BC, (Eq.3.13) Vo where T)r is the relative dynamic viscosity, n is the solvent viscosity, T| 0 is the viscosity of the pure solvent and C is the solute molar concentration. The dynamic viscosity of the solution can be calculated from the product of the kinematic viscosity and the solution density. The coefficient A is always a positive value while coefficient B can have either signs depending on the solute and the solvent [Erdey-Gruz, T., (1974)]. Coefficient A is specific to the solvent properties and limiting conductivities of the ions. It represents the ion-ion interactions and can be calculated from the Debye-Huckel theory based on the limiting ionic molar conductivities according to [Breslau, B. R. et al, (1970)]: A = 1.45 Vo&oTY (Eq. 3.14) Coefficient B qualitatively represents the ion-solvent interaction and is specific to the individual ions. More precisely, the B coefficient is related to the effect of the solvated ion size, the orientation of the solvent molecules and the change in solvent structure on viscosity [Erdey-Gruz, T., (1974)]. This coefficient usually has the most important impact on the viscosity as the ion-ion interaction effect becomes significant only in very concentrated solutions [Erdey-Gruz, T., (1974)]. Coefficient B is the sum of the contributions of the individual ions constituting the solute. The Jones-Dole B + coefficient for Na+ was 76 reported as being 0.0863 1/M at 25°C [Horvath, A. L., (1985)]. Kaminsky extended the Jones-Dole equation by adding a quadratic term to fit viscosity measurements in concentrated aqueous systems, described as: T]r=^- = l + AC/2+BC + DC\ (Eq.3.15) and where the B and D parameters are interdependent [Martinus, N. et al, (1977)]. Hence, the coefficient B obtained from the original Jones-Dole equation (Eq. 3.13) is different from the one obtained from the extended Jones-Dole equation (Eq. 3.15). Several other viscosity correlations to interpret experimental viscosity data of single electrolytes exist [Horvath, A. L., (1985)], but were not considered in this work. a) Effect of Temperature on the Viscosity and Ionic Mobility Temperature has a significant impact on the solubility and transport properties of the ions in solution. The temperature dependence of viscosity is typically expressed by the following Arrhenius relationship [Uddin, F. et al, (2001)]: 77 = />-exp ^ vis RT (Eq. 3.16) where p is a pre-exponential factor, E v ; s is an activation energy of viscous flow, T is the temperature and R is the universal gas constant. A graph of ln(n) as a function of 1/T should illustrate a straight line relationship. Hence, as the temperature is increase, the viscosity of the solvent typically decreases while the ionic mobilities increase. The limiting value of the ion mobility is affected by the interaction changes between the solvent molecules and between the solvent molecules and the ions, by the mutual electrostatic interaction between 77 the ions, and by the changes in the degree of association or dissociation of ions [Erdey-Gruz, T., (1974)]. For a single ion in solution, the ionic mobility dependence on temperature follows a cubic equation: u° = u° + a(T - 298) + b(T - 298)2 + c(T - 298)3 (Eq.3.17) Where M°and u°2S are the limiting values of mobility at T and 298K respectively and where a, b, and c are temperature-independent coefficients. This expression leads to a cubic dependence of the conductivity on the temperature. Hence the limiting ionic mobility values increases with increasing temperature. b) Effect of Temperature on the B Coefficients The extended Jones-Dole B coefficients for a single electrolyte of Eq. 3.15 depend on temperature as per the following expression [Out, D. P. J. et al, (1980)]: A value of 0.023 can be attributed to parameter P. To evaluate the coefficient B of Eq. 3.18, the values of B and D have to be extrapolated from the rearranged extended Jones-Dole correlation using a least square analysis at zero concentration: The parameter B s depends on the standard entropy of hydration (AShyd) and is a measure of the structure breaking/making effects [Lencka, M. M. et al, (1998)]. For monovalent ions, the following relationship can be used: (Eq. 3.18) (Eq. 3.19) 78 =-0.00233 • A S . . -0.297 (Eq. 3.20) A decrease of B values with a rise in temperature indicates that the ions are structure promoters. This means that the ions ability to polarize water molecules decreases with increasing temperature due to increasing thermal motion [Afzal, M. et al., (1994)]. On the other hand, ions for which the B values increase with a rise in temperature are structure breaking and have the opposite effect. C ) Walden's Rule According to Walden's rule, the product of the molar conductivity of an electrolyte solution, A (S.m2/mol), times its dynamic viscosity, n (cP), is nearly constant for the same ions in different solvents at constant temperature as per Eq. 3.21. A -TJ» const (Eq. 3.21) The Walden product is generally constant for a given electrolyte in different solvent or solvent mixtures in which ion-solvent interactions are uniform [Chowdoji Rao, K. et al., (1999)]. The Walden product is generally not constant for highly concentrated salts and solutions of small ions. Nevertheless, it indicates that a change in molar conductivity is affected by a change in the dynamic viscosity. When generalized for the NaB0 2 aqueous alkaline systems under consideration, Eq. 3.21 becomes: ^ N a + B 0 2 H 2 0 ' ^ N a + B 0 2 H 2 0 ~ ^ N a + B 0 2 H 2 O M O H ' ^ N a + B 0 2 H 2 O M O H (Eq. 3.22) For a 1:1 electrolyte at a given temperature, the Walden product of the viscosity and the limiting molar conductivity is equal to: 79 A 0 -7 = 0.8201- — + — (Eq. 3.23) 's J where the hydrodynamic factor, (l/rs++ l/rs")" , quantify the ions hydrodynamic radii and represent the ion interactions with the solvent and A0.n is in Sm2.Cp/mol [Chowdoji Rao, K., et al., (1989)]. The ionic hydrodynamic radius is the apparent size of the dynamic hydrated or solvated ion based on its diffusion. 3.1.3 Physicochemical Properties This Section reviews the physicochemical properties of NaB0 2 aqueous solutions and alkaline aqueous solutions which have been reported in the literature. A) Aqueous NaB0 2 Solutions The electrical conductivity of 1 to 20 wt% aqueous NaB0 2 solutions has been reported from 20 to 300°C [Maksimova, I. N. et al., (1963)]. Over these aqueous NaB0 2 concentrations, conductivity was expressed by the equation: K2=K,(\ + a(T2-T,y), (Eq. 3.24) where K , is the specific conductivity at 40°C and T, = 40°C over temperature ranging from 20 to 200°C. The temperature coefficient of the conductivity, a -\ K \ J \dTj (Eq. 3.25) is also linked to the NaB0 2 concentration as per: 80 a = a + bC, (Eq. 3.26) where C is the Na 20 concentration in wt%, and where a and b are dimensionless constants equal to 2.045xl0"2 and 3.38xl0"4 respectively for temperatures within 20 to 200°C. A maximum, which shifts with temperature, is present in the solubility curve at temperatures lower than 80°C. For example, at 25°C and 40°C, the maximum conductivity is located at 15 wt% and 17 wt% NaB0 2 respectively [Maksimova, I. N. et al., (1963)]. The density of NaB0 2 solutions ( p x ) in the temperature range of 20 to 100°C, without taking account of the solution compressibility, can be calculated by [Maksimova, I. N. et al., (1963)]: PT = PH2O,T +1.14X10"2 C . (Eq. 3.27) Using Eq. 3.25 to 3.27, the molar conductivity and the specific gravity was calculated as a function of the temperature. The data reported for 5.4 wt% and 20.2 wt% NaB0 2 aqueous solutions is summarized in Table 3.3. Table 3.3: Specific Gravity and Conductivity of 5.4 and 20.2 wt% NaB0 2 Aqueous Solutions, (Modified from [Maksimova, I. N. et al., (1963)]). NaB0 2 (wt%) 5.4 20.2 5.4 20.2 Temperature [°C] Specific Gravity Molar Conductivity [S/m] 25 1.0585 1.2202 39 51 50 1.0484 1.1950 64 114 75 1.0364 1.1840 91 175 The Kohlrausch correlation, Eq. 3.24, was found to be applicable to NaB0 2 concentrations up to 1.4 M and was used to extrapolate the limiting value of the molar conductivity of NaB02. Thus, the limiting value of the molar conductivity of the B02" ion at infinite dilution was determined [Maksimova, I. N. et al., (1963)]. The values obtained are summarized in Table 3.4. 81 Table 3.4: Molar Conductivity at Infinite Dilution at Various Temperatures (Modified from [Maksimova, I. N. et al, (1963)]). i A 0 [S m /mole] Temperature [°C] NaB0 2 NaB 50 8 NaOH B0 2-25 78 100 250 28 75 196 200 416 80 The molar conductivity ratios calculated QJX°) were found to be independent of temperature, but decreased as the concentration increased. The ionic strength was used to calculate the corresponding activity coefficients. Table 3.5 show the reported activity coefficients and conductivity ratios obtained for different aqueous NaB0 2 solution concentration. Table 3.5: Molar Conductivity Ratios at Varying NaB0 2 Concentrations (Modified from [Maksimova, I. N. et al, (1963)]). Concentration [M] Activity Coefficient Conductivity Ratio 0.159 0.716 0.800-0.835 0.472 0.622 0.664-0.696 0.868 0.566 0.566-0.588 1.400 0.532 0.470-0.546 The viscosity B coefficients of B(OH)4" and of NaB(OH)4 were reported as 0.233 and 0.316 1/M at 25°C in dilute aqueous solutions, respectively [Corti, H. et al, (1980)]. In aqueous solutions of various metaborates, the equivalent mobility of the anion was reported to be 390000-400000 S.m2 [Kesans, A. et al, (1955)]. B) Alkaline Aqueous NaBO z Solutions To the author's knowledge, no experimental data on the conductivity, viscosity, pH or density of NaB0 2 in concentrated alkaline aqueous solutions has been reported in the literature at 25, 50 or 75°C. An investigation of the transport properties of the tetrahydroxoborate anion in alkaline aqueous media was found. The mobility of the tetrahydroxoborate anion, M „,_„.- , was calculated based on the dissociated constant of boric acid, K a H 3 B0 3 = 6.2xl0"10 [Frei, V. et al, (1963)] as per the following equation: 82 U—U * • C ^ + W M J + Na* B(OH)l (K C ^ BO, J + UH+ \H+}+UOH_\OH-\ (Eq.3.28) where w^,, and WQ//_ are the ionic mobilities of Na+ (45.4), H + (324.9) and OH' (181.4) respectively and C + and C f l 3 + are the concentrations of sodium hydroxide and boric acid, respectively. Using this expression, it was found that the apparent mobility of B(OH)4~ decreased with an increase of NaOH concentration as it forms undissociated complexes with Na+ [Frei, V. et al, (1963)]. 3.1.4 Organic Additives As mentioned in Section 2.2.2, water is not favorable to the direct electroreduction of NaB02. Other media such as molten salts, ionic liquids and anhydrous organic solvents might be more suitable. In a proper media, the electroreduction of NaB0 2 would occur before the electroreduction of the media itself. In aprotic organic solvents, NaBH4 is known to be stable and NaB0 2 is soluble to varying extents. In this Section, the few hydroborate solubility studies available in the literature are reviewed to determine which organic additive would be the most capable to enhance the solubility of NaB02. A) Amides The amides decrease the solubility of sodium tetraborate salts. The systems are simple eutonic as no complex salts or solid solution are formed. The effect of urea, thiourea and acetamide are reviewed. a) Urea The potassium borate - urea - water system, KB0 2 -N 2 H 4 CO-H 2 0, was studied at 15, 25 and 40°C [Druzhinin, G. et al, (1968)]. After the solution reached equilibrium, the liquid and solid phases were sampled and the solution's composition was determine using various methods. The solubility of KB0 2 in 83 pure water at 25°C is about 40 wt%. The addition of 12.8 wt% urea decreased slightly the solubility limit of KB0 2 to 37.5 wt% [Druzhinin, G. et al, (1968)]. This drop in solubility was attributed to the formation of a double compound, KB02.CO(NH2)2- A solubility diagram of the system sodium tetraborate-urea-water, Na2B407-N2H4CO-H20, was developed at 25°C [Sadetdinov, Sh. V., (1985)]. The solubility isotherm of the tetraborate was approximated by the linear equation: where y is the Na 2 B 4 0 7 solubility in wt% and x is the wt% of urea. Alkali metal metaborates are known to form different kind of solid solutions with urea in aqueous solution [Skvortsov, V. G., (1974)]. At 15°C, the sodium metaborate - urea - water system, NaB02-CO(NH2)2-H20, has a eutonic solution composition of 4.99 wt% NaB0 2 and 46.16 wt% urea. Furthermore, the salting out effect of urea on NaB0 2 was accentuated as the temperature was increased. The solubility of the sodium tetraborate-thiourea-water system, Na 2B 40 7-Na 2H 4CS-H 20, was investigated at 25°C [Sadetdinov, Sh. V., (1985)]. Thiourea additions lowered the salt solubility. The solubility isotherm of the borate was approximated by the linear equation: v =-0.0427 +3.12, (Eq. 3.29) b) Thiourea y = -0.0688 +3.07, (Eq. 3.30) where y is the sodium tetraborate solubility in wt% and x is the wt% of thiourea. 84 c) Acetamide The solubility of the sodium tetraborate - acetamide - water system, Na2B407-CH3CONH2-H20, was determined at 25°C [Sadetdinov, Sh. V., (1985)]. The presence of acetamide decreased the solubility of the salt. The solubility isotherm of the borate was approximated by the linear equation: where y is the sodium tetraborate solubility in wt% and x is the wt% of acetamide. B) Ammonia Ammonia (NH3) increases the solubility of boric acid in water as per the following relationship: where S represents the solubility in wt%, H is a dimensionless constant, and C is the salt concentration in wt%. Per linear interpolation, it was estimated that, at 25°C, K has a value of 1.1575 and log S is equal to 0.6292 [Constable, F. H. et al., (1953)]. The addition of the ammonium ion to the ternary system sodium oxide - boric oxide - water, Na20-B203-H20, enlarged the maximum solubility pH range of B 2 0 3 in the solution, thereby positively affecting the solubility of the total solids present in the solution [Rothbaum, H. P. et al, (1956)]. Hence, the peak B 2 0 3 solubility can be augmented by switching to the sodium oxide -ammonium oxide - boric oxide - water system, Na20-(NH4)20-B203-H20. In this system, the Na20:(NH4)20 ratio can be varied to change the pH without significantly affecting the B 2 0 3 solubility. Based on this knowledge, it would be interesting to investigate the effect of ammonia additions on the solubility of NaB0 2 and explore the hydrolysis of ammonium borohydride, (NH4BH4). y = -0.0423 -x + 3.11, (Eq.3.31) logS = log S 0 + # V C (Eq. 3.32) 85 C) Diethylene Glycol NaBH4 is only slightly soluble in ethylene glycol (monoglyme, C 4Hi 0O 2) however it is more soluble in diethylene glycol (diglyme, C 6Hi 40 3) or triethylene glycol (triglyme, C 8 H 1 8 0 4 ) [Brown, H. C. et al, (1955)]. Diethylene glycol is an organic solvent often used to remove remaining hydroborate reagents in solution after exchange reactions to improve product purity. The solubility of NaBH4 in diglyme at 20°C was measured to be 2.42 wt% [Konoplev, V. N., (1988)]. At 40°C, NaBH4 dissolves in diglyme to a concentration up to 3 M. The solubility decreases at higher or lower temperatures. The precipitate formed is a 1:1 solvate of diglyme and NaBH4 [Brown, H. C. et al, (1955)]. The solubility of lithium hydroborate (LiBH4) and hydroaluminate (LiAlH4) was studied in diglyme at temperatures ranging from 25 to 60°C [Mal'tseva, N. N. et al, (1991)]. At 25°C, the solubility of LiBH 4 in diglyme was reported as 9.92 wt% and the precipitate formed was a 1:1 LiBH4-diglyme compound. As temperature increased, the solubility of LiBH 4 increased and the solution viscosity significantly increased. D) Glycine The effect of glycine, C2H5NO2, on the solubility of lithium, sodium and potassium tetraborates [Skvortsov, V. G. et al, (1986a)], of lithium, sodium and potassium pentaborates [Skvortsov, V. G. et al, (1986b)], and of lithium, sodium and potassium metaborates [Skvortsov, V. G. et al, (1988)] was investigated in water at 25°C. The systems were all simple eutonic. The addition of glycine appears to enhance the solubility of sodium tetra- and pentaborates in water. The solubility improvement of the glycine addition was more pronounced on the tetraborates of lithium and the least on those of potassium while the solubility improvement of glycine addition was more pronounced on the pentaborates of lithium, and the least on those of potassium. The solubility of sodium pentaborate is linearly dependent on the glycine concentration as per the following expression: v = 0.152-JC 4-11.51 , (Eq.3.33) 86 where y is the sodium pentaborate solubility in wt% and x is the glycine concentration in wt% [Skvortsov, V. G. et al, (1986)]. Unexpectedly, although the medium remained alkaline, it was found that the solubility of sodium and potassium metaborates decreased in the presence of glycine, while that of lithium metaborates improved. It seems that the metaborates tend to react with glycine to produce polyborates, such as alkali metal tetraborates, and aminoacetates [Skvortsov, V. G. et al, (1988)]. The solubility of the simple eutonic ammonium tetraborate - glycine - water, (NH 4) 2B 40 7-NH 2CH 2COOH-H 20, and ammonium tetraborate - EDTANa 2 - water, (NH 4) 2B 4O 7-C 1 0H 1 4O 8N 2Na 2.2H 2O-H 2 0, systems was determined at 25°C [Skvortsov, V. G. et al, (1992)]. It was found that glycine did not dehydrate the borate and that its hydration properties were weaker than that of EDTANa 2. The solubility of the ammonium tetraborate and of the organic component both increased. At the eutonic point, the ammonium tetraborate concentration was 15.59 wt% while its solubility in water was only 8.76 wt% [Skvortsov, V. G. et al, (1992)]. The solubility isotherms of the glycine - boric acid - water system, glycine-H3B03-H20, were investigated at 0, 25, 50 and 70°C to determine possible interactions between the components [Eysseltova, J. et al, (1994)]. At 25°C, the solubility of boric acid in the glycine solution was correlated by the following linear equation: y = 0.034 +5.66, (Eq.3.34) where y is the boric acid solubility in wt% and x is the wt% of glycine in the solution. A weak interaction between glycine and boric acid was believed to exist in the low temperature range. £ ) Organic Additive Comparison Table 3.6 summarizes the effect of the addition of about 10 wt% organic compound has on the solubility of various borates. In all cases, the solubility of the borate compound is lower than in pure water. 87 Table 3.6: Effect of Organic Additive on the Solubility of Borate Compounds at 25°C. Solute Organic Additive Organic Additive Concentration [wt%] Solubility [wt%] Na 2 B 4 0 7 Urea 10.00 2.69 Na 2 B 4 0 7 Thiourea 10.00 2.38 Na 2B 40 7 Acetamine 10.00 2.69 H3BO3 Ammonia 11.00 3.20 NaB 50 8 Glycine 10.00 13.03 H3BO3 Glycine 10.00 6.00 NaB0 2 Glycine 9.68 2.47 K B 0 2 Glycine 10.66 25.10 LiB0 2 Glycine 10.36 2.11 3.2 Precipitate Characteristics Accurate determination of hydrated borates by X-Ray Diffraction (XRD) is difficult as the peaks of many compounds in the hydrated borate group almost overlap with each other [Kim, J.-H. et al, (2004b)]. Bouaziz investigated the radiocrystallography of lithium and sodium borates and found similarities between the two [Bouaziz, R., (1962)]. A compilation of data on crystal morphology, XRD and fusion of NaB0 2.2H 20 were made [John, K. R. C. Jr., (1951)], and a wide variety of XRD patterns and crystal pictures were reported [Menzel, H. et al, (1943)]. A large diversity in the XRD peak location and intensity exists depending on the conditions under which the crystals were prepared. The structure of the hydrolysis by-product was confirmed to be NaB02.2H20 by XRD, after drying the by-product solution at room temperature for 24 hours [Kojima, Y., et al, (2002)]. In another study, the intensity curve of the by-product obtained from a 25 wt% NaBH4 solution was compared to that of NaB0 2 and NaB02.2H20 [Kojima, Y. et al, (2004b)]. The same research group also evaluated the by-product formed from H 2 generation at high pressure of a solution with a H 20/NaBH 4 molar ratio of 2 by XRD. The by-product was allowed to evaporate at 50°C and vacuum dried at room temperature. The XRD intensity curve corresponded to amorphous NaB0 2 [Kojima Y. et al, (2004a)]. In the same study, the by-product generated at atmospheric pressure from a solution with a H 20/NaBH 4 molar ratio of 4 was analyzed. According to the XRD pattern, its by-product was found to be NaB0 2 as well. Since at high pressure, only half of the water was required to carry out the hydrolysis, it was deduced that NaB0 2.2H 20 transformed into amorphous NaB0 2 during the vacuum drying preparation step. On the other hand, the Kurcel-Merit 88 group mentioned that the hydrolysis by-product is NaB0 2.4H 20 [Suda, S., (2003)]. Even though many boron oxides could have formed, XRD data indicated that NaB0 2.2H 20 was the hydrolysis reaction product [Gervasio, D. et al., (2005)]. The most important peaks of XRD pattern data of boron compounds relevant to this study can be found in Appendix C. 89 CHAPTER IV EXPERIMENTAL METHODS 4.1 Materials The Fisher Scientific alkali hydroxides used were sodium hydroxide electrolytic pellets (NaOH, S318-500, 98.1% purity), potassium hydroxide pellets (KOH, P250-500, 87.7% purity) and lithium monohydrate crystals (LiOH.H20, L127-500, 99%+ purity). The three alkali hydroxides were used independently. Sodium metaborate 4 mol (NaB02.2H20, 99%+ purity) from U.S. Borax Inc. was used. The hydration level of NaB0 2.2H 20 was not constant over the entire study. Solutions were prepared with deionized distilled water having an electrical ionic conductivity of 1.5 uS/m at a pH of 6.1. Glycine, (C2H 5N0 2 , CH 2NH 2COOH), US Pharmacopeia grade, was obtained from Fisher Scientific (G48) in the white crystalline powder form. 4.1.1 Properties Tables 4.1 and 4.2 list the pertinent anion and cation properties, while Table 4.3 contains the thermodynamic data pertaining to the species under considerations in the system investigated. Figure 4.1 describes the molecular chain arrangement of glycine, which is a non-polar hydrophobic aliphatic. Table 4.1: Anion Properties. Property B(OH)4" BH4" OH" Reference Ionic radius [nm] 0.229 ±0.019 0.205 ±0.019 0.152 ±0.019 [Roobottom H. K. et Volume [nm3] 0.058 0.066 ±0.015 0.032 ±0.008 al, (1999)] [Jenkins, H. D. B. et al, (1999)] 90 Table 4.2: Cation Properties. Property (in H 2 0 at 25°C) Na+ K + L i + Reference Ionic radius [A0] 0.98 1.33 0.68 [Antropov, L. I., (1972)] Effective ionic radius [A0] 4 3 6 [Dean, J. A., (1999)] Limiting ionic conductivity [10 4.m2.S/mol] 50.11 73.5 38.69 [Dean, J. A., (1999)] Table 4.3: Thermodynamic Data. Species dG° [KJ/mol] dH° [KJ/mol] dS° [J/mol.K] B02" -678.94 -772.37 -37.24 OH" -157.28 -230.015 -10.90 BH4" 114.27 48.16 110.50 Na+ -261.88 -240.34 58.45 H 2 (g) 0 0 130.68 H 2 0 (1) -237.14 -285.83 69.95 NaOH (aq) -419.20 -469.15 48.10 KOH (aq) -440.53 -482.37 91.60 LiOH (aq) -451.9 -508.40 7.10 NaB0 2 (aq) -940.81 -1012.49 21.80 NaB0 2 (c) -920.70 -977.00 73.54 KB0 2 (aq) -962.19 -1024.75 65.30 KB0 2(c) -923.40 -981.60 79.98 LiB0 2 (c) -976.10 -1032.20 51.50 NaBH 4 (aq) -147.61 -199.60 169.50 B02" -678.94 -772.37 -37.24 [Dean, J. A., (1999)] NaB0 2.2H 20 -1415.20 -3228.07 LiB0 2.2H 20 -2873.80 B(OH)4" -1159.87 -1345.46 -622.47 L i + -278.48 -293.30 Na+ -240.12 -261.89 K + -252.38 -283.26 [Li, J. era/., (2000)1 o o-\ / H ~ C - H I N H 3 + Figure 4.1: Molecular Chain Arrangement of Glycine. 4.1.2 Contamination Table 4.4 summarizes the ICP-MS analysis of deionized distilled water used throughout this study. 91 Table 4.4: Deionized Distilled Water ICP-MS analysis. Element B Na K Li Concentration [mg/1] 3.5 8 <2 0.08 Other than the elements listed in Table 4.4, impurities present in the solutions are Ca, Cu, Fe and Zn. During the course of the experiment, a potential source of boron contamination can be attributed to the use of borosilicate glass. Hence, the use of glass was avoided whenever possible during the solution preparation and physicochemical properties determination. 4.2 Experimental Plan Figure 4.2 represents the experimental plan followed-in this study. Tests were conducted on aqueous NaB0 2 solutions, saturated NaB0 2 solutions, unsaturated NaB0 2 solutions and diluted NaB0 2 solutions containing no alkali or various wt% of alkali and organic additives at 25, 50 and 75°C. The solution preparation procedures are presented in Section 4.3. Filtration, as described in Section 4.4, was only performed when evaluating the properties of saturated solutions. NaBOj Aqueous Solution Saturated Unsaturated Diluted (0.0025-15M) (33-35 wtV.) (5wt%) 25,50,75'C No additive 25'C Figure 4.2: Experimental Plan Schematic Diagram. 92 The filtrate properties determined included solubility, conductivity, pH, viscosity and density. The precipitates were evaluated by X-Ray Diffraction and Scanning Electron Microscopy. The procedures followed for the filtrate and precipitate characterization are described in Sections 4.5 and 4.6, respectively. 4.3 Solution Preparation Three types of solutions were prepared in this study. Dilute solutions of NaB0 2 in water, saturated solutions of NaB0 2 in water, with or without alkali additives, and finally, saturated solutions of NaB0 2 with an organic additive with and without alkali additives. The solution preparation procedures are described below. 4.3.1 Dilute Aqueous NaB02 Solutions Solutions of different NaB0 2 concentrations were prepared by dilution. An appropriate weighted amount of NaB0 2 was placed in a 50 ml ±5 % graduated beaker and de-ionized water was added up to 30 or 40 ml depending on the desired concentration. The solutions were magnetically stirred for a few minutes to ensure homogeneity. An analog-controlled Cole-Parmer polystat heated circulating water bath (Model EW-12107-40) was used to keep the solutions at a constant temperature ±0.2°C. 4.3.2 Concentrated Alkaline Aqueous NaBC>2 Solutions An isothermal solution saturation methodology was followed to prepare the solutions. A predetermined quantity of alkali additive (1, 2, 3, 5, 7.5 or 10 wt%) was combined in a flask containing excess sodium metaborate (32-35 wt%). A fixed de-ionized distilled water volume (55 ml) was added to the mixture. The liquid to solid ratio was made sufficient to allow efficient stirring. The water to boron weight ratio before filtration ranged from 1.83 to 1.9 in most experiments. Polytetiafluorethylene (PTFE) flasks were used to avoid potential contamination with the boron contained in borosilicate glass. The flasks were sealed with rubber stoppers with a small vent hole open to the atmosphere. This prevented pressure 93 build up but did not prevent water evaporation to the ambient atmosphere during the high temperature experiments. The solutions were magnetically stirred for 5 hours to accelerate the homogeneous solid dissolution and allow equilibrium to be reached. At this point, temperature, pressure and composition were assumed to be constant. The stirring time required to establish equilibrium was estimated based on the time reported for similar experimental conditions and components with similar solubilities. After reaching equilibrium, stirring was stopped and the solutions were allowed to stand for one hour to enable excess solute settling and phase separation. The heated circulating water bath mentioned above was used to keep the solutions at the desired temperature ± 0.2°C throughout the sample preparation period. Saturation of the solution was confirmed by visual observation.' 4.3.3 Organic Additives Preliminary tests were conducted using glycine as an organic additive. The procedure followed to conduct the tests with glycine was very similar to the one used to prepare the concentrated alkaline aqueous solution described above. The tests were conducted using a mixture of 5 wt% NaOH or KOH, 10 wt% glycine, and excess sodium metaborate (35 wt%) in 50 ml of de-ionized distilled water. For saturated NaB0 2 alkaline aqueous solutions containing glycine, the time required to establish equilibrium was much longer than for the solutions containing no organic additives. The solutions were magnetically stirred for 36-39 hours, then stirring was stopped and the solutions were allowed to settle for one hour. 4.4 Filtration After the solution was prepared and equilibrium reached, a sample of the liquid was taken and filtered. Filtration was carried out differently, depending on the temperature at which the experiment was performed. 94 4.4.1 Filtration at 25°C Filtration of the solution at 25°C was conducted in two different ways. First, preliminary solubility experiments were conducted after vacuum filtration using different types of membranes, and then by syringe filtration. Syringe filtration was adopted to conduct the bulk of the experimental work at ambient temperature. A) Vacuum filtration In vacuum filtration, a flask was clamped to a ring stand and a funnel was connected to it with an adaptor. A piece of filter paper was placed in the funnel and the side of the flask was then connected to the vacuum source. The paper was wetted using a small amount of solution. The water circulation bath maintained the temperature constant throughout the filtration. The vacuum was turned on to start the filtration and the funnel was pressed on to engage seal with filter adapter. The filtrate solution was poured on the filter paper. After filtration, the rubber tubing was disconnected before turning off the vacuum. The filter paper was removed and the solids were collected and dried. Figure 4.3: Vacuum Filtration Apparatus. 95 Different paper filters with pore sizes of 0.45 um were tested: cellulose ester (CS) filters and hydrophobic Teflon (PTFE) filters. The CS filter tended to leak and break under high vacuum while the hydrophobic PTFE filter would not filter until made hydrophilic by pre-wetting with acetone. Hydrophillic PTFE filters were judged too expensive for the number of experiments to be conducted. B) Syringe Filtration In this case, sampling was conducted using a graduated plastic syringe and plastic tube. Syringe filters, fitted with PTFE membranes with 0.45 um diameter pores, were used to filter a clarified solution sample withdrawn from the flask. Difficulties were encountered in the withdrawal of saturated solution samples when the solution was highly viscous due to cavitation. Pushing the aqueous solution through the hydrophobic PTFE filter was very difficult. Hence, a filtration apparatus was built and designed to simplify the task. Filtration was conducted at room temperature using the stainless steel filtration apparatus assembly described in Figure 4.4, which ensured a steady filtration rate. Figure 4.4: Syringe Filtration Apparatus: (a) Schematic Diagram (b) Full View. Table 4.5 is a comparison of the NaB0 2 solubility results obtained in aqueous solution at 25°C using different filtration techniques described in Section 4.3.1. 96 Table 4.5: Filtration Method Comparison. Filtration Method Filter Type NaB0 2 Solubility in Water at 25°C rwt%l Average Standard Deviation [%] Vacuum CS 30.45 1.28 Vacuum PTFE 30.41 0.13 Syringe PTFE 29.21 0.79 It is important to note that temperature was not controlled or maintained to 25°C during syringe filtration. It was assumed that the solubility standard deviations were caused by temperature variations estimated to be of ± 3°C, during the filtration step. In all cases, during filtrate characterization, the filtrates were kept in sealed plastic bottles maintained at 25°C with the water circulation bath described earlier. 4.4.2 Filtration at 50 and 75°C Major difficulties were encountered when using the apparatus described above to filter the sample solutions in an oven set at the desired temperature. At higher temperatures, the syringe plastic became too soft, and the syringe shaft would fold under pressure rather than pushing the sample solution through the attached syringe filter. A SS plunger was built to use in the plastic syringe core. However, the sealing between the SS part and the soft plastic part was poor. Hence, the fluid was leaking on the side of the plunger and the liquid was not pushed through the filter. Another option was to use a simple filtering method. For the tests performed at 50 and 75°C, the equilibrium phases were decanted and filtrated by gravity using a Whatman paper filter No. 4, for course and gelatinous precipitates, placed on a preheated funnel. This process was conducted in an oven heated at a temperature slightly above the desired temperature in order to prevent recrystallization during sampling and filtration. It is important to note that the pore sizes of the paper filter were much larger than that of the syringe filters used to filter the solutions at ambient temperature. 97 4.5 Filtrate Characterization At a minimum, three solutions of the same composition, i.e. MOH and NaB0 2 concentration, were characterized. The physicochemical properties values reported represent the average of the measurements performed on at least three solutions. 4.5.1 Solubility There exist many methods to determine the content of boron in a liquid. The most relevant ones are reported in this Section. A) Boron Detection Methods In analytical methods, the solubility is.directly determined by a chemical analysis of the phases in equilibrium. Typically, this is done by carrying out a titration with an acid or base and a variety of reagents, such as polyols, and a color indicator to determine the end points. The volume of acid or base used to the end-point indicates how much boron is present in the solution. The most common titration method used for the determination of boron in alkaline aqueous solution is carried out using mannitol and methyl red. The results obtained tend too be too high as interferences with carbon dioxide and silica are present. Carbon dioxide interferences can be addressed by refluxing the solution prior to titration. Other possible boron determination methods include coulometry, chromatography, potentiometry, volumetry, voltammetry, atomic absorption and mass spectrophotometry. Some references for each method are listed in Table 4.6. 98 Table 4.6: List of Some Boron Detection Methods. Type References Analytic Volumetric Voltammetric Colorimetric Coulometric ICP Potentiometric Spectrophotometric Atomic Absorption Spectrophotometric [Brown, H. C , (1955), Rudie, C. N., (1979), Johnson, E. A. etal., (1954), Dodd, S., (1929), Blumenthal, H., (1951), Drescher, A. et al, (1998), Midgley, D., (1992), Belcher, B. et al, (1970)] [Kuleshova, O. D., (1985), Ross, W. J. et al, (I960)] [Wei, W. C. etal, (1995)] [Norkus, P. K , (1968), Drescher, A., (1989)] [Sagrado, S., (1998)] [Andrew, H. E., (1976)] [Staden, J. F., (2000), Van Staden, J. F. et al, (2000), Basson, W. D. et al, (1969), Chaurasia, S. C , (2004), Hayes, M. R. etal, (1962), Betty, K. R. etal, (1986), Lopez Garcia, I. et al, (1985), Sah, R. N., (1997)] [Chaikin, S. W., (1953), Lyttle, D. A., (1963)] [Mirkin, M. V., (1991), Celikkan, H. et al, (2005)] [Dawidowicz, A. L. et al, (1989)] B) Induced Coupled Plasma Mass Spectrometry In this study, induced couple plasma (ICP) mass spectrometry (MS) was chosen for the quantitative determination of the filtrate boron, sodium, potassium and lithium content. The ICP-MS analysis was carried at International Plasma Laboratories Ltd. Prior to ICP-MS analysis, samples were digested to ensure solubility of any precipitate which could have formed after filtration. Minimum detection ranged from 0.01 to 2 mg/1 and maximum detection was 999 or 9999 mg/1. It has been shown that the detection sensitivity for boron can be further improved with the use of mannitol as a chemical modifier in electrothermal vaporization inductively coupled plasma mass spectrometry (ETV-ICP-MS) [Wei, W. C. et A VWR Scientific digital pH meter (Model SP21), with a resolution of ±0.01 and a pH range from -2 to 19.99was used to determine the pH at the filtrate temperature. A Corning rugged bulb combination glass electrode (Model 476560) with Ag/AgCl internals was used. After 30 minutes of standing and after an hour of standing, the changes in pH were within ±0.4 pH units. At 25°C, the average standard deviation on the pH measurements of NaB0 2 saturated alkaline aqueous solutions was about ± 0.2 pH units. al, (1995)]. 4.5.2 pH 99 4.5.2.1 Glass Electrode General purpose glass electrodes are suitable in the pH range from 1 to 10, but at other pH values special "low sodium error" electrodes which can operate at high temperatures are preferable. Standard glass electrodes measure erroneously high pH values in highly alkaline solutions (pH > 10). In those solutions, the glass membrane is not selective to the H + ions. At low concentration of H + , alkali metal cations such as Li + , Na+, and K + , present in large excess, tend to combine with the membrane surface, resulting in an electrode response. The potential dependence on the pH becomes non-linear. This interference phenomenon is known as the alkaline error. A) pH Meter The pH meter was calibrated using standard buffers at pH 4, 7 and 10 between each series of measurements. The measured pH is affected by temperature. The chemical equilibrium and the electrode properties are both affected. E = E° pH (Eq.4.1) F The slope of the Nernst equation increased with temperature. The electrode did not reach thermal equilibrium spontaneously and pH fluctuations were likely to occur. Hence, temperature needed to be recorded along with the pH measurements. 4.5.3 Conductivity The filtrate electrical conductivity measurements were made with an Orion 105A portable conductivity meter with a conductivity measurement range from 0 to 199.9 S/m ±0.5%. Measurements were normalized with a temperature adjustment of 2.5%/°C at 25°C. The average standard deviation on the conductivity 100 measurements for saturated solutions at 25°C was about ±4.8 S/m. For the dilute aqueous NaB0 2 solutions, the conductivity measurement average standard deviation was approximately ±0.05 S/m in the higher concentration range and ±0.005 S/m in the lower concentration range. At higher temperatures, more time was allowed for the conductivity measurements to be stabilized. The filtrate temperature was recorded along with the conductivity measurements which were later adjusted to the desired temperature of 50 and 75°C using a compensation factor of 2.1°C per degree. The water circulation bath, pH meter and conductivity meter set-up are shown in Fig. 4.5. Figure 4.5: pH and Conductivity Experimental Set-up. 4.5.4 Viscosity Cannon-Fenske routine viscometers are accurately calibrated glass tubes used to measure the kinematic viscosity of a fluid. Viscometers universal size number 75 (1.6-8 cSt), 100 (3-15 cSt), 150 (7-35 cSt) and 200 (20-100 cSt) were used to determine the filtrate (i.e. NaB0 2 alkaline solution) viscosity within ±0.2 %. The viscometer was filled with the fluid to be characterized and immersed in a water bath maintained at 25°C ± 1°C. As the water bath temperature fluctuation was greater at higher temperatures (50 and 75°C), a silicone based oil of a viscosity of 350 cSt was used. Special care was taken to prevent crystallization from occurring in the viscometer for the high temperature measurements. Since crystallization was associated with an increase in volume, it could have strain the internal wall of the viscometer. Temperature was allowed to stabilize before the first efflux flow time measurement was taken. However, with the set-up used, as shown in Fig. 4.7, it was difficult to precisely reach the desired temperature and maintain it constant throughout the viscosity measurement period. 101 Figure 4.7: Viscometer Experimental Set-up. The kinematic viscosity was obtained by multiplying the viscometer constant, extrapolated at the measurement temperature, with the efflux time. The dynamic viscosity was calculated by multiplying the experimental kinematic viscosity by the sample density. Between each experiment, the viscometer was rinsed alternately with distilled water and acetone, and dried with Argon to ensure proper drainage and avoid particle accumulation. Calibration runs were performed with distilled water before and after all measurements were taken, and in all cases the viscometer constant did not change significantly (< 2 %) during the course of the experiments. In the case of the alkaline aqueous NaB0 2 filtrate at 25°C, three measurements were taken per filling sample while in the case of the dilute aqueous NaB0 2 solution and alkaline aqueous NaB0 2 filtrate at 50 and 75°C, two measurements were taken per filling samples. The averaged standard deviation on the viscosity measurements on a particular sample were ± 0.007 cP and ± 0.03 cP respectively. 4.5.5 Density The dilute aqueous NaB0 2 solution density was measured by weighting a pycnometer calibrated with de-ionized water and determined to have a volume of 25.034 ml at 25°C. Between measurements, the pycnometer was rinsed, dried in an oven at 70°C and cooled in a desiccator. A standard analytical balance with a precision of ±0.004 g was used for all weight measurements. The average standard deviation for the dilute NaB0 2 solution density measurements was ±0.001 g/ml. The average standard deviation for the 102 density measurements of saturated solutions at 25°C was about ±19 kg/m3. The concentrated alkaline aqueous solution filtrate density was determined by simple weight and volume measurements at room temperature. Volume was determined using graduated cylinders with a readability of ±0.5 ml. 4.6 Precipitate Characterization The 10 wt% saturated solution wet residue mixtures obtained at 25°C were characterized by X-Ray Diffraction and Scanning Electron Microscopy. The precipitates were decanted and dried in an air oven at 50°C overnight. The dried crystals were crushed and conserved in Petri dishes. Their compositions may quite differ from the pure crystals as they might still contain a significant amount of the mother solution. 4.6.1 X-Ray Diffraction X-Ray Diffraction (XRD) measurements were taken with a Rigaku Multiflex XRD 2 kW analyzer using Cu radiation. The diffraction patterns were recorded at 40 kV and 20 mA in the angular range of 9=10-65°, with a continuous scan speed of 2 deg./min and a sampling width of 0.05°. The intensities observed were corrected for background noise, and peaks were determined using the MDI Jade 7 software. 4.6.2 Scanning Electron Microscopy Scanning Electron Microscopy (SEM) images were taken using a Hitachi model S-3500N system. Most precipitate crystallographic morphologies were observed at low scanning speed under low vacuum mode. The vacuum pressure was 20 Pa and the accelerating voltage was 20 kV. 103 CHAPTER V RESULTS AND DISCUSSION 5.1 Aqueous N a B 0 2 Solutions Physicochemical Properties In this Section, the experimentally obtained solubility and transport properties of NaB02 aqueous solutions are analyzed. Tables B.l to B.4 in Appendix B contain samples of the physicochemical property raw data analysed in the following pages. 5.1.1 Solubility To test the validity of the experimental method, first the solubility of NaB0 2 was measured in water at 25°C by ICP-MS analysis of the filtrate and compared with values reported in the literature. The ICP-MS analysis boron concentration was converted to weight percentage in terms of NaB0 2.2H 20 or NaB0 2 for ease of comparison. Table 5.1 compares the experimental solubility value obtained with the literature data. Table 5.1: Comparison of NaB0 2 Solubility wt% in Water at 25°C. Solubility Solubility [wt% as NaBQ2.2H2Q] [wt% as NaBQ2] Source / Reference 30.45 19.68* Present work ICP-MS analysis ± 1.28 wt%, ± 3°C. 34.05* 22.00 [Blasdale, W. C. et al, (1938)] 33.43* 21.60 [Nies, N. P. et al, (1967)] 33.40 21.58* [US Borax] 33.83* 21.86** [Kojima, Y., et al, 2002 and 2004b] * Converted based on molecular weight ratio. ** Estimated from visual observation. An average solubility value of 30.45 ±1.28 wt% of NaB0 2.2H 20 was determined at 25°C ± 3°C. In terms of anhydrous NaB0 2 the mean solubility experimental value obtained was 2.32 and 1.92 wt% lower than 104 the solubility values reported in the literature [Blasdale, W. C. et al., (1938), Nies, N. P. et al., (1967)]. All of these early studies used the methyl red - mannitol titration method to measure solubility. As mentioned in Section 4.5.1, in this analytical method, the boric acid concentration represents the total amount of boron in the sample. It is unknown if measures were taken to repel C0 2 trapped in the solutions. If present, aqueous C 0 2 will react with water and form a bicarbonate which dissociates as a carbonic acid. This carbonate acid gets titrated as if it was boric acid, the resulting titration boron concentration values are falsified. Thus, if no special measures are taken to remove C0 2 , this analytical boron detection method is not as precise as ICP-MS analysis, which did not exist at that time. Based on these results, the boron determination methodology employed was judged to be adequate for the purposes of this study. 5.1.2 Transport Properties A) Conductivity The solution concentrations covered in the conductivity experiments ranged from 0.0033 to 3.3 M. The measured conductivities were corrected for the solvent impurities by subtracting the conductivity of the solvent. As explained in Section 3.1.2 A), the various conductivity models are only valid for a certain concentration range. Hence, the Kohlrausch square root second extension correlation (Eq. 3.4) and the Kohlrausch cubic root correlation (Eq. 3.5) were not valid over the concentration range studied and did not represent the experimental data. To extrapolate at zero concentration, the variation of the molar conductivity as a function of the square root of the concentration for the NaB02-water system was plotted according to the Kohlrausch correlation and its first extension (Eq. 3.2 and 3.3) in Fig. 5.1. The molar conductivity decreased as the concentration increased. The non-linear model fitting equations obtained from the Sigma Plot accurately represented the experimental data. As seen in Fig. 5.1, the molar conductivity measurements seem to obey the lower orders of the Kohlrausch square root correlation at very dilute concentrations. The empirical value of K in Eq. 3.2 (78.28) is only 0.51 % lower than the value predicted from the Debye-Huckel-Onsager theory as per Eq. 3.9 (78.68). Eq. 3.3 seems to graphically represent the best fit at concentrations lower than 0.65 M. A theoretical limiting molar conductivity of 105 76.33 S.cm2/mol was estimated by extrapolation using the Onsager expression (Eq. 3.10). Hence the limiting molar conductivity value obtained from Eq. 3.2 or 3.3 would have a relative error of 5.72 %. A = 80.70 - 78.28CA1/2 (Eq. 3.2) -A =80.70 - 38.64CA1/2 - 19.82C (Eq. 3.3)| Experimental Data - - . . • A X . X X 0 . 0 0 0 . 2 5 0 . 5 0 0 . 7 5 1 .00 1 .25 1 .50 Square Root of Concentration (M ) Figure 5.1: Molar Conductivity as a Function of the Square Root of the Concentration at 25°C ± 3°C. Table 5.2 summarizes and compares the limiting molar conductivity values predicted by fitting the different models. As can be seen, all the values are relatively close. The limiting molar conductivity obtained from Eq. 3.2 and 3.3 is within 3.34 % from the value reported in the literature, which was reported in Table 3.4 [Maksimova, I. N. et al, (1963)]. Table 5.2: Comparison of Limiting Molar Conductivity Predictions at 25°C ± 3°C. Equation Description Limiting Molar Conductivity _ = _ = = = = = = = = = _ _ = [S.cm2/mol] Kohlraush Square Root (Eq. 3.2) 80.70 Kohlraush Square Root Extention 1 (Eq. 3.3) 80.70 Kohlraush Square Root Extention 2 (Eq. 3.4) 84.92 Kohlraush Cubic Root (Eq. 3.5) 91.19 Onsager Limiting Law (Eq. 3.10) 76.33 106 The limiting ionic transport properties summarized in Table 5.3 were calculated with Eq. 3.7, based on the NaB0 2 limiting molar conductivity value obtained from the lower orders of the Kohlrausch square root correlations (Eq. 3.2 and 3.3), i.e. 80.70 S.cm2/mol. Table 5.3: Infinite Dilution Transport Properties for Na^CV at 25°C ± 3°C. Parameter Limiting Molar Ionic Diffusion Transport Number Conductivity Mobility Coefficient [Units] [S.cm2/mol] [10"4cm2/s.V] [105 cm2/s] [Dimensionless] Na+ 50.11 [Dean, J. A., (1999)] 5.19 1.33 0.621 B02" 30.59 3.17 0.81 0.379 The calculated B02" limiting molar conductivity is low compared to that of other monovalent ions, which is typically around 75 S.cmVmol [Dean, J. A., (1999)]. Yet, it is within 9 % of the value in Table 3.4 which was reported in the literature [Maksimova, I. N. et al, (1963)]. The ionic mobility calculated shows that Na+ moves faster when submitted to a potential gradient. However, the mobility value of B02" is outside of the reported mobility value range for various borate anions [Kesans, A. et al, (1955)]. The calculated B02" diffusion coefficient is more than two times lower than that reported for the N02" ion (1.912xl0"5 cm2/s) [Lide, D. R., (2005-6)]. The calculated ionic transport numbers are indicative of the fact that the Na+ ions carry most of the charge when current is applied to the NaB0 2 aqueous solution. B) Viscosity The upper limit of NaB0 2 concentration employed in the viscosity determination experiments was dictated by the solubility limit of NaB0 2.2H 20 in water at 25°C which ranged from 33.40 to 34.05 wt% as NaB0 2.2H 20 [Blasdale, W. C. et al, (1938), Nies, N. P. et al, (1967), Kojima, Y., et al, (2002 and 2004b)]. The solution concentrations covered in the viscosity experiments ranged from 0.0025 to 0.5 M. Using the limiting ionic molar conductivities of Table 5.3, the value of coefficient A was estimated to be equal to 0.0099 1/M1 / 2 using Eq. 3.14. This value is comparable to that reported for other electrolytes such as NaCl (0.0062 1/M1/2) [Horvath, A. L , (1985)]. The value of coefficient B was estimated by plotting (r)/r|0 -1)/C I /2 as a function of C l / 2 in accordance with the Jones-Dole and the extended Jones-Dole 107 correlations, corresponding to Eq. 3.13 and 3.15 respectively. The theoretical value of coefficient A at infinite dilution was used to fit the model. In Fig. 5.2, the best fitting model was plotted along with the experimental data. The NaB0 2 concentration increased as the viscosity increased and the equation was found more appropriate for very dilute concentrations. c in o u 10 > _ra a> or X Exper imenta l Data n/n.0 = 1 + 0 .00985C A 1/2 + 0.42C - 0 .06C A 2 (Eq. 3.15) X x-x-x- X X . -x 1.30 1.25 1.20 1.15 1.10 1.05 1.00 0.95 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Square Root of Concentration (M 1 ' 2) Figure 5.2: Relative Viscosity as a Function of the Square Root of the Concentration at 25°C ± 3°C. 0.8 Both model fittings were excellent, although Eq. 3.15 resulted in a slightly better fit than Eq. 3.13 as it is valid over a wider concentration range. Both correlations gave a small positive B coefficient value indicating that the association between the ions constituting the electrolyte are weak. The extended Jones-Dole correlation shown in Fig. 5.2 gave a NaB0 2 B coefficient of 0.4185 1/M. As the B coefficients are additive between the cation and anion, a B" coefficient of-0.4660 1/M was obtained for B 0 2 \ This value is one order of magnitude higher than that reported for the N03" ion (-0.0450 1/M) [Horvath, A. L., (1985] but much smaller than the B coefficients of B(OH)4" (0.233 1/M) in dilute aqueous solutions at 25°C [Corti, H. et al., (1980)]. It is expected that, in a solution of medium size ions (K+, CI"), the different viscous forces tend to cancel each other [Erdey-Gruz, T., (1974)]. It is possible that, due to its size and shape, the B02" ion slightly alters the solvent's viscosity and influences the orientation of the polar water molecules of the 108 solvent. However, the negative B coefficient values would indicate that, due to polarization effect, the ions have a stronger tendency to disorder the solvent's structure. 5.2 Aqueous NaBC>2 Solutions Physicochemical Properties with Alkali Additive The solubility and physicochemical properties of saturated NaB0 2 aqueous solutions are analyzed in this Section. Appendix B contains examples of the raw data gathered in Tables B.5 and B.6. The summary of the averages used to carry out the analysis can be found in Table B.7. Unless indicated otherwise, all properties reported in this Section were measured in the filtrate of NaB0 2 saturated alkaline aqueous solutions. The error bars shown on the Figures showing measured values represent the standard deviation values obtained from a minimum of three measurements performed on three different solutions of the same composition. The standard deviations were calculated using the conventional statistical formula with N> 3. 5.2.1 Solubility As seen in Chapter 4, in this study, NaB02.2H20 is the compound used to prepare the solutions. Hence, in this thesis all solubility values are reported in terms of NaB02.2H20. Figure 5.3 illustrates the solubility of NaB0 2.2H 20 as a function of the saturated solution hydroxide content at 25°C ± 3°C. 109 i 0 1 2 3 4 5 6 7 8 9 1 0 11 H y d r o x i d e C o n t e n t (wt%) Figure 5.3: NaB0 2 Solubility as a Function of Hydroxide Content at 25°C ± 3°C. The impact of alkali addition on NaB0 2 solubility differs greatly depending on the alkali hydroxide used. An addition of 1 wt% hydroxide resulted only in a minor solubility improvement in all cases over the solution containing no hydroxide. The solubility decreased in the cases of NaOH and LiOH, while it increased as the wt% KOH increased. The decrease in NaB0 2 solubility with increasing NaOH wt% agrees with the literature [Amendola S. C. et al, (2000a)]. More than 1 wt% of NaOH and LiOH resulted in solute precipitation. The trends observed are in accordance with the solubility of the alkali additive in pure water at 25°C, i.e. KOH is the most soluble, while LiOH is the least. However, the solubility of NaB02.2H20 in NaOH is slightly lower than in LiOH. The low solubility of NaOH can be attributed to the common ion effect, which further reduces the solubility of NaB02. At 10 wt% NaOH, the solubility was 10.89 wt% in terms of anhydrous NaB02, which is 4.11 wt% lower than the reported value [Suda, S., (2003b)]. Based on these results, it is not surprising that at room temperature, only a 2 wt% H 2 storage capacity can be achieved with concentrated NaOH solution before the precipitation of the reaction by-product [Hua, D. et al, (2003)]. Moreover, the solubility is linked to the hydration shell of the alkali cation. Small ions bind more solvent molecules and are more hydrated. The effective ionic radii of K + , Na+ and L i + in pure water are 3 A, 4 A and 6 A respectively [Dean, J. A., (1999)]. Thus, the effective available H 2 0 content is the lower in the case of L i + , yielding the lowest solubility for NaB02. Based on Fig. 5.4, KOH is a better alkali additive to maintain the hydrolysis by-product in solution. It is expected O 3 f & CO. re o 4 0 3 5 3 0 < 1 2 5 i i 2 0 1 5 10 • N a O H • K O H A L i O H • N o a d d i t i v e 110 that the non-monotonic behaviour observed about 7.5 wt% KOH is not due to experimental error but caused by a lost in the crystallization water of the NaB0 2.2H 20 powder used to prepare the solutions. Hence, this measurement should be verified. 5.2.2 pH Figure 5.4 shows that the filtrate pH increased as the quantity of alkali hydroxide added to the saturated NaB0 2 solution increased. As the pH measurements were taken after filtration of the saturated NaB0 2 alkaline aqueous solution, the concentration of NaB0 2 in the filtrate varied as shown in Fig. 5.3. 1 5 . 5 1 5 . 0 1 4 . 5 1 4 . 0 1 3 . 5 1 3 . 0 4-1 2 . 5 i { • N a O H • K O H • L i O H • N o a d d i t i v e 4 5 6 7 H y d r o x i d e C o n t e n t ( w t % ) 1 0 11 Figure 5.4: Filtrate pH as a Function of Hydroxide Content at 25°C ± 3°C. The filtrate of the KOH- and NaOH-stabilized solutions exhibited a higher pH than the filtrate of the LiOH-stabilized solutions. More precipitation resulted in a filtrate with a lower pH, while less precipitation resulted in a higher pH. The NaB0 2 precipitation decreased the total alkali content of the filtrate, which lead to lower pH values. The addition of LiOH resulted in the lowest NaB0 2 solubility, and had a lower pH at 10 wt%. In NaOH, increased pH suppressed the solubility of NaB0 2 due to the presence of the common Na+ ion. Hence, the KOH system was more alkaline than the NaOH system at 10 wt% due to the higher NaB0 2 solubility. I l l A) NaBR, Solution Half-Life Although the high pH of the stabilized NaBH4 solutions prompts safety concerns associated with the handling of caustic solutions, it is not a disadvantage as it significantly extends the life of the NaBH4 solution by preventing the uncontrolled release of H 2 . Employing Eq. 2.68, Fig. 5.5, demonstrates that the stability of NaBH 4 solutions depends on the solution's pH but is only indirectly independent of the cation present, through the resulting pH. 3500 -~ 3000 ] o 2500 o I m 2000 1500 1000 500 1 • NaOH • KOH A LiOH • No additive 10 wf% KOH — 12.5 1 Wt% NaOH 13 14.5 15 13.5 14 PH Figure 5.5: Estimated Half-Life of Alkaline Aqueous Solutions of NaBH 4 as a function pH at 25°C ± 3°C using Eq. 2.68. The addition of 1 wt% NaOH extended the half-life to about 200 days over the NaB0 2 solution containing no hydroxide stabilizer. A NaBH4 solution lifetime of nine years can be achieved with 10 wt% of KOH, providing hydrolysis catalysts are absent. 112 5.2.3 Transport Properties A) Ionic Conductivity a) Effect of Hydroxide Concentration on the Conductivity of Saturated NaB02 Solutions Figure 5.6 shows the ionic conductivity of the filtrated NaB0 2 saturated aqueous solution as a function of alkali concentration. 2 2 0 j -2 0 0 -1 8 0 -1 6 0 -> 1 4 0 -3 1 2 0 -•a c o o 1 0 0 -0) +J E 8 0 -i i 6 0 -O 4 0 • 2 0 -• 1 i A • N a O H • K O H A L i O H • N o a d d i t i v e 4 5 6 7 H y d r o x i d e C o n t e n t ( w t % ) 10 11 Figure 5.6: NaB0 2 Ionic Conductivity in the Filtrate as a Function of Hydroxide Content at 25°C ± 3°C. The NaB0 2 content in the saturated solution was about 32-35 wt%. As seen previously, the extent of precipitation varied depending on the type of hydroxide and its concentration. Hence, the quantity of NaB0 2 in the filtrate fluctuated as the saturated solution composition changed. The ionic conductivity was measured after filtration of the saturated solution. Although the filtrate conductivity was proportional to the ion concentration and depended on the type of ion present, the conductivity measurements were difficult to interpret as the weight percentage of solute contained in the filtrate varied slightly with the alkali hydroxide concentration and the type of alkali hydroxide present as previously shown in Fig. 5.3. 113 For example, in the case of the NaOH-stabilized solution, there was a difference of 12.43 wt% between the maximum and minimum NaB0 2 content depending on the NaOH concentration used over the concentration range studied. b) Unsaturated NaB02 Alkaline Aqueous Solution The ionic conductivity was measured at a constant concentration of NaB02. Unsaturated aqueous solutions containing 5 wt% NaB0 2.2H 20 and varying concentration of alkali additive were prepared. As the conductivity meter limit was reached in solutions containing more than 5 wt% hydroxide, no measurements could be made on solutions containing higher hydroxide concentrations. Figure 5.7 shows the ionic conductivity of the unsaturated solution as a function of the alkali hydroxide content. 2 0 0 1 7 5 1 5 0 ? 55. 1 2 5 If > 1 0 0 7 5 5 0 2 5 •o c o o • A • A • • N a O H • K O H A L i O H 0 1 2 3 4 5 6 Hydroxide Content (wt%) Figure 5.7: Ionic Conductivity as a Function of Hydroxide Content of the Unsaturated 5 wt% NaB0 2 Aqueous Solution at 25°C ± 3°C. As shown in Table 3.3, the reported conductivity of a 5.4 wt% NaB0 2 aqueous solution is 39 S/m [Maksimova, I. N. et al, (1963)]. The ionic conductivity of strong electrolyte solutions depended significantly on the hydroxide concentration and almost increased linearly. This confirms that the OH" ion is the most important contributor to the solution conductivity. This behaviour is also interlinked with the 114 increased in ionic mobility as the concentration increases. It would be expected that conductivity increase as the cation size decrease but this was not the case, when considering the effective ionic radius (Li+>Na+>K+). Similarly to the saturated solution case, the highest ionic conductivity was obtained for the NaOH-stabilized solution. However, in the unsaturated solutions, KOH as the alkali source yielded the lowest ionic conductivities rather than LiOH. B) Molar Conductivity a) Unsaturated NaB02 Alkaline Aqueous Solution Figure 5.8 shows the molar conductivity as a function of hydroxide content for the unsaturated NaB0 2 solutions. 2 3 4 5 6 Hydroxide Content (wt%) Figure 5.8: Molar Conductivity as a Function of Hydroxide Content of the Unsaturated 5 wt% NaB0 2 Aqueous Solution at 25°C ± 3°C. Unlike the ionic conductivity, the molar conductivity of NaB0 2 unsaturated solutions did not increase linearly with the hydroxide concentration. Furthermore, the molar conductivity did not follow the same relative hydroxide order as in Fig. 5.8. As it is inversely proportional to the molecular weight of the alkali hydroxide present, the total solute molar concentration in the solution is mainly impacted by the hydroxide 115 concentration. Hence, the relative molecular weight of the alkali hydroxide, KOH>NaOH>LiOH, will have a direct impact on the molar conductivity. This explains why, at all hydroxide concentrations, the molar conductivity was the lowest for the LiOH-stabilized solution. However, the conductivity of the NaOH-stabilized solutions did not agree with this trend, probably due to the common ion effect. The molar conductivity of the NaOH- and LiOH-stabilized solutions seems to even out as the hydroxide content increases pass 3 wt%, while the KOH-stabilized solution molar conductivity seams to get steady at hydroxide concentrations greater than 5 wt%. It would be interesting to extend the concentration range studied to see if a maximum is reached in the molar conductivity at a certain hydroxide concentration. It appears that the molar conductivity of the KOH-stabilized solution crossed the NaOH-stabilized solution molar conductivity around 4 wt% hydroxide addition. At low wt%," the molar conductivity was the highest for the NaOH-stabilized solution, while at 5 wt% hydroxide, it is the highest for the KOH-stabilized solution. The reasons behind this behaviour are not clear at this point and require further investigations. b) Effect of Hydroxide Concentration on the Molar Conductivity of Saturated NaB02 Solutions Figure 5.9 shows the molar conductivity as a function of hydroxide content for the saturated NaB0 2 solutions. 116 0.038 0.036 0.034 0.032 0.030 0.028 0.026 0.024 0.022 0.020 0.018 1 0.016 0.014 0.012 0.010 0.008 • N a O H • K O H • L i O H • N o a d d i t i v e 4 5 6 7 Hydroxide Content (wt%) 9 10 11 Figure 5.9: Molar Conductivity in the Filtrate as a Function of Hydroxide Content at 25°C ± 3°C. As can be seen, the filtrate molar conductivity value obtained at 5 wt% NaOH seems to be about 21 % lower than the value which would be expected from the linear data trend. This error might be because the ionic conductivity measure is too low. For this solution composition, the filtrate ionic conductivity measurement was verified, and, as can be seen in Fig. 5.7, although it appears to be slightly low, it is within the experimental data trend. Hence, the error was attributed to a too high concentration of NaB02, probably caused by NaB0 2.2H 20 hydration losses. C ) Dynamic Viscosity The deionized water viscosity was 0.86 cP, while the filtrated NaB0 2 saturated aqueous solution viscosity was 5.31 cP. Figure 5.10 shows dynamic viscosity as a function of hydroxide content in the NaB0 2 saturated solution. 117 • N a O H • K O H • L i O H • N o a d d i t i v e i I 0 1 2 3 4 5 6 7 8 9 H y d r o x i d e C o n t e n t (wt%) Figure 5.10: Filtrate Dynamic Viscosity as a Function of Hydroxide Content at 25°C ± 3°C. t f 1 0 11 At 25°C, the filtrate dynamic viscosity increased with increasing hydroxide content in the saturated solution, but not in a linear way. In pure water and at 25°C, the viscosity of a LiOH solution exceeds that of a NaOH solution, which in turns exceeds that of a KOH solution. However, in saturated NaB0 2 solutions, the filtrate dynamic viscosity of the saturated solutions containing LiOH or KOH exceeded the filtrate dynamic viscosity of the saturated solution containing NaOH. Furthermore, the addition of LiOH or KOH both resulted in higher dynamic viscosities than that of the NaB0 2 aqueous solution. Additions of NaOH resulted in a dynamic viscosity lower then that of the saturated NaB0 2 aqueous solution filtrate. As the addition of NaOH significantly decreased the solubility of NaB02, the resulting filtrate contained fewer Na+ ions. Hence, the filtrate resistance to flow was lower and its resulting dynamic viscosity was closer to that of water. It is possible that precipitation of NaB0 2 in NaOH solution resulted in structural changes in the solvent, due to the common ion effect, and that significantly decreased the filtrate dynamic viscosity. In general, the larger the effective ionic radii of the cation (Lf^Na^K*), the sharper was the extent of the dynamic viscosity increase. The filtrate dynamic viscosity increased more sharply with increasing concentration of LiOH than with increasing concentrations of NaOH and KOH. While additions of LiOH also resulted in low NaB0 2 solubility, the large hydration shell of L i + significantly restrained the movement 118 of ions in the filtrate. It seems that the ion hydration effect raised the dynamic viscosity of the filtrate. The large hydrated L i + ion created more and more friction as the LiOH concentration increased in the saturated solution filtrate and the ion mobility decreased sharply. This ionic electric field reduces the mobility of the molecules and increases the dynamic viscosity of the solution. Hence, on a dynamic viscosity basis, KOH or NaOH seem to be better electrolytes than LiOH for the system under consideration. D ) W a l d e n P r o d u c t From the experimentally determined viscosity and molar conductance values, it is possible to verify if the Walden's rule holds for the solutions studied. The filtrate Walden products for the saturated NaB0 2 aqueous solution as well as for the saturated NaB0 2 alkaline aqueous solutions were calculated and plotted as a function of the saturated NaB0 2 solution hydroxide content in Fig. 5.11. It was found that Eq. 3.21 was mostly applicable to the solutions at low hydroxide concentrations as the magnitude of the Walden product was significantly different at higher hydroxide concentrations. More specifically, the Wladen product remained constant for concentrations up to 3 wt% for NaOH and LiOH, and for concentrations up to 7.5 wt% for KOH. At 1 wt% hydroxide, the Walden product of the filtrate containing hydroxide matched that of the filtrate of the aqueous NaB0 2 saturated solution to within ±4 to 22 %, while at 3 wt% hydroxide, it matched the filtrate of the aqueous NaB0 2 saturated solution within ±0.8 to 37 %. The best match was obtained at 2 wt% hydroxide content, as the Walden product of the filtrate containing hydroxide matched that of the filtrate of the aqueous NaB0 2 saturated solution within ±1.8 %. At this hydroxide concentration, the Walden product was independent of the solvent as the hydroxide cations were not significantly hydrated and their concentration had a similar effect on the filtrate transport properties, i.e. molar conductivity and dynamic viscosity. 119 • • ! ' A • N a O H • K O H A L i O H • N o a d d i t i v e 0 1 2 3 4 5 6 7 8 H y d r o x i d e C o n t e n t ( w t % ) Figure 5.11: Filtrate Walden Product as a Function of Hydroxide Content at 25°C ± 3°C. 10 11 At high alkali hydroxide content, however, the right part of Eq. 3.22 yielded values over two orders of magnitude greater than the left part of Eq. 3.22. Therefore, it can be concluded that those solutions are outside of the applicability limit of the Walden rule as other ionic interactions were significant and had to be taken into account. At 10 wt% hydroxide, the structure of the filtrate of the saturated salt solution more closely resembles that of a molten salt than that of water. It is worth noting that the solute-solvent interactions in the filtrates were appreciably greater when the solution was stabilized with NaOH rather than KOH and LiOH. 5.2.4 Specific Gravity The filtrate specific gravity as a function of hydroxide concentration is shown in Fig. 5.12. The filtrate specific gravity corresponds to the ratio of the solution density and the density of the deionized distilled water used to make the solutions at the same temperature. The filtrate specific gravity increased proportionally with the hydroxide content in the saturated solution. Similarly, for a fixed NaBH4 concentration at 20°C, the specific gravity of alkaline aqueous NaBH4 solution increases as the wt% NaOH increases [Li, Z. P., (2004)]. A solution containing 25 wt% NaBH4 and 10 wt% NaOH had a specific 120 gravity of 1.9 at 20°C, this is higher than the specific gravity of the filtrate of a saturated NaB0 2 solution containing 10 wt% NaOH at 25°C which is about 1.34. In pure water, the density of KOH is greater than that of NaOH, which exceeds the density of LiOH, and parallels the order of magnitude of the molecular weight of the three alkali additive. Accordingly, the KOH-stabilized filtrate had the highest specific gravity, which exceeded the specific gravity of the LiOH-stabilized filtrate. However, the specific gravity of the NaOH-stabilized filtrate was the lowest, probably due to the low NaB0 2 solubility and the common ion effect. > E o o j£ 'o a> a. CO . 4 0 . 3 9 . 3 8 . 3 7 . 3 6 . 3 5 .34 ] . 33 . 32 .31 . 3 0 .291 . 2 8 • • A • • • • N a O H • K O H • L i O H • N o A d d i t i v e s 0 1 2 3 4 5 6 7 8 H y d r o x i d e C o n t e n t ( w t % ) Figure 5.12: Filtrate Specific Gravity as a Function of Hydroxide Content at 25°C ± 3°C. 1 0 11 5.3 Effect of Temperature The temperature effect was investigated by carrying out experiments on saturated NaB0 2 aqueous solutions containing 0, 1, 5 and 10 wt% NaOH and KOH at 50 and 75°C. Only two trials were conducted per solution. Due to experimental difficulties in retrieving a representative sample and in filtrating the gel-like solutions, some experimental results obtained did not represent the solution content and had to be discarded. At 50 and 75°C, the solutions were very viscous and sticky. It appears that equilibrium was not easily achieved within 5 hours of stirring as liquid and solid phase separation was not achieved, even after 121 an hour of settling. Hence, the retrieval of representative filtrate samples for solubility and physicochemical parameter measurements was nearly impossible. Table B.8 in Appendix 2 is a summary of the averaged physicochemical properties for the solution studied at 25, 50 and 75°C. 5.3.1 Solubility At 50 and 75°C, the filtrate samples sent for ICP-MS analysis were completely solidified since, as soon as the filtrate temperature decreased to room temperature, significant crystallization occurred in the super-saturated solutions. Even though digesting of the sample was carried out prior to ICP-MS analysis, it is suspected that the element concentration values obtained are not representative of the expected sample content. Figure 5.13 shows the experimentally measured NaB02.2H20 solubilities. As expected, the solubility increased as temperature increased. At 50 and 75°C, increasing hydroxide concentration reduced the solubility of NaB0 2 in all cases. However, it can be seen that NaB0 2.2H 20 solubility values in excess of 100 wt% were obtained, which is not mathematically possible. The NaB0 2.2H 20 solubility values obtained in aqueous solution at 50 and 75°C were much greater than the ones reported in the literature. At 50 and 75°C, average solubility values of 92.22 and 113.56 wt% NaB0 2.2H 20 were obtained, respectively. According to the values listed in Table 3.9, the reported average solubility for each temperature should be around 49.92 and 66.13 wt% NaB0 2.2H 20 respectively. Hence, the experimentally determined solubility values are almost twice as large as the ones reported in the literature when compared on the same NaB0 2.2H 20 solubility basis. This excess solubility can be partly attributed to NaB0 2.2H 20 hydration water lost during the solution preparation at 50 and 75°C. For example, based on NaB02, average solubility values of 59.59 and 73.38 wt% would be obtained at 50 and 75°C, respectively. However, even when no hydration water is taken into account, the solubility values are about 10 % higher than the ones reported in the literature. Similarly, at 75°C, solubility values in excess of 100 wt% were obtained for saturated NaB0 2 alkaline aqueous solutions stabilized with 1 and 5 wt% hydroxide. The experimental causes of the error in the measurement are not clear at this time and require further investigations. Hence, 122 the values shown in Fig. 5.14 should be used with caution. A suitable and reliable methodology to determine the solubility of NaB02.2H20 at 50 and 75°C needs to be developed. 1 2 5 4 5 6 7 Hydroxide Content (wt%) A N a O H a t 2 5 - C • K O H at 2 5 - C A N a O H at 5 0 " C • K O H at 5 0 " C A N a O H a t 7 5 - C • K O H at 7 5 " C o N o A d d i t i v e at 2 5 * C o N o A d d i t i v e a t 5 0 ' C • N o A d d i t i v e a t 7 5 ' C Figure 5.13: NaB0 2 Solubility as a Function of Hydroxide Content at 25, 50 and 75°C. 5.3.2 Transport Properties A) Ionic Conductivity a) Effect of Hydroxide Concentration on the Conductivity of Saturated NaB02 Solutions Figure 5.14 illustrates the ionic conductivity measurements as a function of NaOH and KOH content at 25, 50 and 75°C in the filtrate of saturated NaB0 2 solutions. As expected, conductivity increased with the hydroxide concentration and with temperature. However, it was not the case at 75°C at hydroxide content greater than 1 wt%. Hence, at this temperature, the composition of the filtrate sample characterized did not represent the content of the solution and the data were not included in Fig. 5.14. 123 4 5 6 7 Hydroxide Content (wt%) A N a O H at 2 5 * C • K O H at 2 5 ' C A N a O H at 5 0 " C • K O H at 5 C T C A N a O H a t 7 5 * C • K O H at 7 5 " C o N o A d d i t i v e at 2 5 " C o N o A d d i t i v e at 5 0 ' C • N o A d d i t i v e at 7 5 ' C Figure 5.14: NaB0 2 Ionic Conductivity in the Filtrate as a Function of Hydroxide Content at 25, 50 and 75°C. For a more meaningful analysis, conductivity measurements on saturated NaB0 2 alkaline aqueous solutions should be performed without filtration. To verify the conductivity trends, conductivity measurements were performed in unsaturated NaB0 2 alkaline aqueous solutions. b) Unsaturated NaB02 Alkaline Aqueous Solution Conductivity measurements were performed on unsaturated 5 wt% NaB0 2 solutions as a function of hydroxide content at 50 and 75°C. When no alkali additives were present, conductivity values of 36.25 and 59.40 S/m were obtained at 50 and 75°C respectively. When using Eq. 3.24 and 3.25 with the conductivity measured at 25°C, predicted conductivity values of 42.33 and 57.42 S/m were calculated at 50 and 75°C respectively. According to this, in solutions containing no alkali, the conductivity values measured would be about 14% too low at 50°C and about 3% too high at 75°C. The conductivity values measured were plotted along with the conductivity measurements at 25°C in Fig. 5.15 and 5.16, for NaOH and KOH, respectively. 124 200 175 B 150 i ? 125 ]> | 100 "O o 75 O 50 25 • o o o o 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 H y d r o x i d e C o n t e n t ( w t % ) 4.5 5.0 55 O N a O H a t 2 5 " C O N a O H a t 50*C • N a O H a t 75 "C o N o a d d i t i v e at 2 5 ' C o N o a d d t i v e at 50 "C • N o a d d i t i v e at 7 5 " C Figure 5.15: Ionic Conductivity as a Function of NaOH of the Unsaturated 5 wt% NaB0 2 Aqueous Solution at 25, 50 and 75°C. 200 175 E 150 55 £• 125 > | 100 •a c o 75 O 50 25 0.0 • • • • • • • • • 0.5 1.0 1.5 2.0 2.5 3.0 3.5 H y d r o x i d e C o n t e n t ( w t % ) 4.0 4 .5 5.0 • K O H a t 25*C o N o a d d i t i v e at 2 5 ' C • K O H a t 50 "C o N o a d d t i v e at 50 "C • K O H a t 7 5 " C • N o a d d i t i v e a t 75"C 5.5 6.0 6.0 Figure 5.16: Ionic Conductivity as a Function of KOH of the Unsaturated 5 wt% NaB0 2 Aqueous Solution at 25, 50 and 75°C. Lower hydroxide weight percentages were used as the conductivity meter limit of 200 S/m was easily reached at high temperatures. It seems that, at 75°C and 0.5 wt% hydroxide, the conductivity measurements are slightly lower than expected, based on the conductivity values obtained for the solution containing no alkali additive. These measurements should be verified. 125 B) Molar Conductivity a) Unsaturated NaB02 Alkaline Aqueous Solution The unsaturated conductivity values of Fig. 5.15 and 5.16 were used to construct the graph shown in Fig. 5.17. The molar conductivity increased as temperature increased. At 50°C, the molar conductivity of the solutions stabilized with NaOH exceeded the ones with KOH. However, at 75°C, the molar conductivity of solutions containing either stabilizer was the same. 1.5 2 2 . 5 Hydroxide Content (wt%) 3 . 5 • N a O H at 7 5 " C • K O H at 7 5 " C o N a O H at 5 0 " C • K O H a t 5 0 " C o N a O H at 2 5 ' C • K O H at 2 5 " C Figure 5.17: Molar Conductivity as a Function of Hydroxide Content of the Unsaturated 5 wt% NaB02 Aqueous Solution at 25, 50 and 75°C. 5.4 Effect of Glycine Additions In Section 3.1.4 D), it was reported that glycine improved the aqueous solubility of the sodium tetra-and pentaborates, as well as LiB0 2 , while it decreased the solubility of NaB0 2 and KB0 2 . To see if the addition of LiOH would have a positive impact on the NaB0 2 solubility results, preliminary experiments were conducted to investigate the effect of the addition of 10 wt% glycine, on the solubility and physicochemical properties of saturated aqueous solutions of NaB0 2 containing 0 and 5 wt% alkali hydroxide at 25°C ± 3°C. Two trials were conducted per solution composition. The summary of the averaged physicochemical parameters measured can be found in Table B.9 in Appendix B. 126 5.4.1 Solubility The addition of 10 wt% glycine significantly decreased the solubility of NaB0 2 in water. The solubility of NaB0 2 in 10 wt% glycine in water was found to be equal to 0.68 wt% at 25°C, which is 3.6 times lower than the value listed in Table 3.9. As shown in Table 5.4, the use of 5 wt% alkali stabilizers had little improvement effect on the resulting solubility. The observed solubility decrease in the presence of glycine confirms that, even in aqueous solutions stabilized with LiOH, NaB0 2 is the main solute formed. If LiB0 2 would be formed, the solubility would be expected to increase in the presence of glycine in aqueous solution as discussed in Section 3.1.4 D) [Skvortsov, V. G. et al, (1988)]. Table 5.4: Effect of Glycine on the Filtrate Solubility as a Function of Hydroxide Content at 25°C ± 3°C. Glycine Content Alkali Content NaB02.2H20 Solubility Standard Deviation [wt%] rwt%i [wt%] [± wt%] 0 0 30.45 1.28 0 5, NaOH 29.83 2.83 0 5, KOH 31.46 2.27 0 5, LiOH 31.51 0.00 10 0 0.68 0.01 10 5, NaOH 0.78 0.02 10 5, KOH 1.06 0.06 10 5, LiOH 1.17 0.02 Further tests should be performed to investigate the effect of lower glycine concentrations and higher alkali additive concentrations on the solubility of NaB0 2 in water. Also, solubility measurements should be made to investigate the solubility of glycine in alkaline aqueous NaB0 2 solutions. 5.4.2 pH The addition of glycine decreased the pH of the deionised distilled water. The addition of hydroxide increased the filtrate pH, but the resulting pH was still lower than that of the filtrate containing no glycine. When comparing Fig. 5.5 to Table 5.5, it is clear that the extent of the pH increase resulting from the 127 addition of 5 wt% alkali hydroxide is significantly lower when 10 wt% glycine is present in the alkaline aqueous saturated NaBC>2 solution filtrate. Table 5.5: Effect of Glycine on the Filtrate pH as a Function of Hydroxide Content at 25°C ± 3°C. Glycine Content Alkali Content Filtrate pH Standard Deviation [wt%] [wt%] [± pH unit] 0 0 12.79 0.22 0 5, NaOH 14.33 0.17 0 5, KOH 14.34 0.22 0 5, LiOH 13.85 0.19 10 0 11.8 0.01 10 5, NaOH 11.86 0.02 10 . 5, KOH 12.55 0.23 10 5, LiOH 12.03 0.25 A) NaBEL, Solution Half-Life Figure 5.18 presents the NaBH4 solution half-life estimates based on the filtrate pH values displayed in Table 5.5. As can be anticipated, the addition of glycine decreased the expected half-life of the solution. The decrease in half-life due to glycine addition is not as significant as that caused by a rise in temperature. 128 3 0 • A X X 1 1 . 6 0 1 1 . 8 0 1 2 . 0 0 1 2 . 2 0 1 2 . 4 0 1 2 . 6 0 PH 1 2 . 8 0 1 3 . 0 0 • 0 % a d d i t i v e • 1 0 % g l y c i n e , 0 % a lka l i A 1 0 % g l y c i n e , 5 % N a O H [ X 1 0 % g l y c i n e , 5 % K O H x 1 0 % g l y c i n e , 5 % L i O H Figure 5.18: Effect of Glycine on the Estimated Half-Life of Alkaline Aqueous Solutions as a function pH at 25°C ± 3°C using Eq. 2.68. 5.4.3 Transport Properties A) Ionic Conductivity Table 5.6 reveals that the 10 wt% glycine NaB0 2 alkaline aqueous solution conductivity decreased when NaOH and LiOH were used as the stabilizers, but increased when stabilized with KOH. However at 5 wt% KOH, the extent of the 10 wt% glycine conductivity improvement was not as significant as when no glycine was present in the solution. Nevertheless, more measurements should be taken to further investigate the conductivity improvement resulting from the combination of KOH and glycine at lower glycine concentrations. 129 Table 5.6: Effect of Glycine on the Filtrate Ionic Conductivity as a Function of Hydroxide Content at 25°C ±3°C. Glycine Content Alkali Content Ionic Conductivity Standard Deviation [wt%l rwt%i rs/mi [± S/m] 0 0 52.67 0.86 0 5, NaOH 99.77 1.40 0 5, KOH 68.18 3.50 0 5, LiOH 62.76 2.04 10 0 48.75 0.64 10 5, NaOH 50.75 1.34 10 5, KOH 56.55 0.21 10 5, LiOH 36.40 0.14 B) Molar Conductivity The molar conductivity was calculated based on the total solute concentration in the solution, i.e. glycine, hydroxide and NaB02. Table 5.7 shows that the molar conductivity was the greatest when only glycine was present in the saturated aqueous solution of NaB02. The addition of 5 wt% hydroxide decreased the molar conductivity of the saturated NaB0 2 aqueous solution filtrate containing 10 wt% glycine. The extent of the molar conductivity decrease was the greatest in the case of 5 wt% LiOH addition. Table 5.7: Effect of Glycine on the Filtrate Molar Conductivity as a Function of Hydroxide Content at 25°C ± 3°C. Glycine Content rwt%] Alkali Content [wt%] Molar Conductivity [S.m2/moll 0 0 0.0141 0 5, NaOH 0.0186 0 5, KOH 0.0130 0 5, LiOH 0.0092 10 0 0.0275 10 5, NaOH 0.0153 10 5, KOH 0.0189 10 5, LiOH 0.0080 130 C) Dynamic Viscosity As seen in Table 5.8, the dynamic viscosity of the filtrate of the saturated NaB0 2 aqueous solutions containing glycine was significantly lower than in the absence of glycine. Like for the filtrate containing no glycine, at 5 wt% hydroxide, the viscosity was the greatest when LiOH was used as the stabilizer. Table 5.8: Effect of Glycine on the Filtrate Dynamic Viscosity as a Function of Hydroxide Content at 25°C ±3°C. Glycine Content Alkali Content Dynamic Viscosity Standard Deviation [wt%] rwt%l [cP] [± cPl 0 0 5.31 0.49 0 5, NaOH 31.90 0.10 0 5, KOH 34.53 0.08 0 5, LiOH 39.40 0.17 10 0 2.18 0.09 10 5, NaOH 1.77 0.21 10 5, KOH 1.97 0.20 10 5, LiOH 3T4 0.00 5.4.4 Specific Gravity Glycine has a molecular weight of 75.07 g/mol, which is greater than that of the three alkali hydroxide additives used in this study. Table 5.9 proves that, as expected, the addition of glycine increases the specific gravity of the filtrate. Even though at 5 wt% alkali additive, the filtrate specific gravities increased, the extent of the increase was not as pronounced as for solutions containing no glycine as shown in Fig. 5.13. This is mainly because the glycine addition significantly decreased the solubility of NaB02. 131 Table 5.9: Effect of Glycine on the Filtrate Specific Gravity as a Function of Hydroxide Content at 25°C ± 3°C. Glycine Content Alkali Content Specific Gravity rwt%i rwt%i 0 0 1.2896 0 5, NaOH 1.3248 0 5, KOH 1.3644 0 5, LiOH 1.3600 10 0 1.3097 10 5, NaOH 1.2914 10 5, KOH 1.3304 10 5, LiOH 1.3272 5.5 Precipitate Characteristics X-Ray Diffraction and Scanning Electron Microscopic characterization was carried on the composition of the precipitates formed from the saturated aqueous solutions of NaB0 2 stabilized with 10 wt% hydroxide. At high pH, replacement of one solute by another could to take place. In the case of the solutions stabilized with KOH and LiOH, KBO2 and LiB02 could be formed as the precipitate. The hydration level of the precipitate could also be affected by the type of alkali additive present. Information on the precipitates formed is reviewed in this Section. 5.5.1 X-Ray Diffraction The precipitates formed from supersaturated NaB0 2 solutions containing 10 wt% alkali additive samples were examined by X-Ray Diffraction. Although there were analogies between most XRD patterns, there was a broad variety in the nature of the precipitates formed. Since this work did not intend to be a detailed crystallographic study, the drying process conditions were not strictly controlled and varied from one sample to another. Consequently, precipitate crystals formed from the same solution composition could vary. The molar ratio of Na20, B 2 0 3 and H 2 0, solution pH and chemical composition, heat treatment, hydration level, recrystallization or decomposition conditions and pressure are all factors affecting the precipitate crystal formation. Hence, depending on the hydroxide used, hydrated NaB02 might not be the only precipitate formed and hydrated borates of K B 0 2 and LiB0 2 could also be present. 132 Figure 5.20 shows selected XRD patterns obtained for each solution type and compares them to the peaks obtained from pure NaB02.2H20. The most important peaks of other precipitate XRD patterns obtained can be found in Appendix C. The XRD patterns shown in Fig. 5.19 are considered to be the most relevant as they were performed a shorter time period after the precipitate preparation. The XRD scans obtained for the precipitates formed from KOH and LiOH solutions seem to indicate that hydrated NaB0 2 is the predominant species in all cases. The most prominent diffraction peak of the NaOH-stabilized precipitate was recorded at a 29 of 16.786° and corresponded to that of pure NaB02.2H20. The most important peak of the KOH- and LiOH-stabilized precipitates occurred at a two-theta of 31.2. It was at a higher two-theta value than the most important NaB0 2.2H 20 peak. According to the MDI Jade 7 software database, this peak is a characteristic of NaB02.4H20. 180000 160000 140000 ^ 120000 to a C 80000 — 60000 40000 20000 Pure N a B 0 2 . 2 H 2 0 10 wt% LiOH 10wt% KOH 10 wt% NaOH 10 15 20 25 30 35 40 Two-Theta (Degree) 45 50 55 60 Figure 5.19: Intensity as a Function of Diffraction Angle for Pure NaB02.2H20 and Precipitates of 10 wt% Hydroxide NaB0 2.2H 20 Supersaturated Aqueous Solutions. It is likely that the NaOH-stabilized precipitate lost some of its hydration water while drying since the transition to the lower hydration level occurs at about 54°C [Blasdale, W. C. et al., (1938), Nies, N. P. et al., (1967)] and at 25°C, the stable hydrate is NaB02.4H20 [Blasdale, W. C. et al, (1938), Nies, N. P. et al, (1967), Kocher, J. et al, (1970), Menzel, H. et al, (1943)]. It is not easy to remove NaB02.2H20 crystals from the alkaline solution in which they are grown [Menzel, H. et al, (1943)]. Due to their 133 physical and chemical instability, they immediately transform to NaB0 2.4H 20 when in contact with water. Hydrolysis leads to the formation of a hydrated solid phase, but depending on the conditions under which the H 2 generation system operates, the hydration level of the by-product generated will vary. 5.5.2 Scanning Electron Microscopy Figure 5.20 shows selected SEM images of pure NaB02.2H20 and of the various precipitates obtained. Additional SEM precipitate images can be found in Appendix D. The pure NaB0 2.2H 20 flakes shown in Fig. 5.20 (a) are a few mm long and have a porous surface. Figure 5.20: Scanning Electron Microscope Precipitate Crystal Images, (a) Pure NaB0 2.2H 20 Flakes (SE, WD 14.1mm, x45, 1mm), (b) NaB0 2.2H 20 Crystal (SE, WD15.8mm, xlO.Ok, 5um), (c) 10 wr% NaOH, Saturated NaB02.2H20 Aqueous Solution Precipitate (BSE2, WD14mm, x2.0k, 20um), (d) 10 wt% KOH, Saturated NaB02.2H20 Aqueous Solution Precipitate (BSE2, WD 14mm, x500, lOOum), (e) 10 wt% LiOH, Saturated NaB02.2H20 Aqueous Solution Precipitate (BSE2, WD13.4mm, xl.Ok, 50um), (f) 10 wt% LiOH Saturated NaB0 2.2H 20 Aqueous Solution Precipitate (BSE2, WD13.3mm, xl.Ok, 50um). Freshly decanted crystals contained in alkaline solutions are clear, but as soon as the alkaline solution diffuses out of the crystals, they become opaque. The NaB0 2.2H 20 crystals have the distinctive columnar lozenge shape seen in Fig. 5.20 (b). The precipitate formed from saturated NaB0 2 solution stabilized with NaOH shown in Fig. 5.20 (c) resembles a dispersion of small individual eroded and crumbled crystals. The KOH-stabilized aqueous NaB0 2 solution precipitate of Fig. 5.201 (d) was similarly constituted. The precipitate of the LiOH-stabilized NaB0 2 saturated solutions was a thick suspension, which, once dried 134 overnight, had the aspect of slush. The scattered group of small individual crystals, shown in Fig. 5.20 (e), was formed. To investigate how heat treatment affects the crystal surface morphology, this precipitate was dried for a longer period of time. This gave the crystals the intricate microstructure revealed in Fig. 5.20 (f), likely caused by the evaporation of the alkaline solution it contained. 135 CHAPTER VI CONCLUSIONS 6.1 Conclusions In this thesis, the effect of three alkali additives on the solubility and physicochemical properties of diluted and saturated aqueous NaB0 2 solutions was investigated. The present study was designed to determine which of these alkali additives was the most suitable and which concentration of alkali additive was optimal for the electrochemical reduction of NaB0 2 in aqueous media. The most obvious finding emerging from the experimental measurements is that all solution characteristics cannot be optimized at once. In fact, from the background information reviewed, it is clear that various parameters have different impact on the NaBH4 H 2 generation and storage system properties. Table 6.1 summarizes the interactions between the variables affecting the system. Table 6.1: Variable Effect on Various Parameters of the NaBH4 H 2 Generation and Storage System (+ increase, - decrease, op=optimum point, n=none). Variable Effect V s Parameter H 2 gen. rate Solubi l i ty l l / 2 P H Viscos i ty Conduct iv i ty Weight Densi ty V o l u m e Densi ty Catalyst loading + + n n n n + n n T + + + - n - + + + [ N a B H 4 ] + op n n n - + + + [ M O H ] + - - + + + - - -Closed P + n n n n n - n The findings of this thesis further confirm that compromises are required to create either a viable H 2 generation and storage system or a favourable electrolyte media for the electrochemical recycling of the hydrolysis waste solutions. This illustrates the complexity of the system being investigated given the number of variables and the interrelationships. The physicochemical property data obtained enhanced the general understanding of the effect of alkali additives on the solution properties and their impact on the H 2 storage and generation system. It revealed that, based on conductivity and viscosity, NaOH is a more 136 suitable additive for the electrochemical recycling of the spent NaB0 2 solution, while KOH addition favours the H 2 generation and storage system, based solubility and pH. A number of recommendations were made to address the shortcomings of the present study and suggestions for future work directions are given. Finally, the significance of the knowledge acquire through this thesis is placed in perspective. 6.1.1 System Considerations Table 6.2 synthesizes the major findings of this study having a significant influence over future development and application. Table 6.2: Summary of Alkali Hydroxide Additive Effect on the H 2 Storage and Generation System and Electrochemical Recycling. Additive Property Consequence NaOH - Highest ionic conductivity - Lowers Ohmic drop - Lowest viscosity - Reduces mass transport losses KOH - Improved NaB0 2 solubility - Improves hydrolysis conversion efficiency - Permits the use of a more concentrated solution - Reduces spent tank emptying frequency - Highest pH - Extends NaBH4 solution half-life - Prevents unsafe H 2 release - Highest specific gravity - Lowers system gravimetric density LiOH - All properties - Not favourable Although all the filtrate ionic conductivities were high, the ionic conductivity of unsaturated NaB0 2 solutions containing NaOH was the highest, while the ionic conductivity of unsaturated NaB0 2 solutions containing KOH was the lowest. This high conductivity enhances the current carrying capacity of the solution and would stimulate the electrochemical reaction of the species it contains. The addition of NaOH resulted in the lowest dynamic viscosity and the addition of KOH resulted in a lower filtrate dynamic viscosity than the addition of LiOH. The low viscosity resulting from the addition of NaOH would reduce the mass transport losses and would allow more NaBH4 and H 2 0 to contact the catalyst surface, and thereby potentially increase the H 2 generation. It would also result in superior electrochemical cell performance as Ohmic losses would be reduced. 137 The results obtained reveal that the most important factors affecting the improvement of the solubility of NaB0 2 are the type of hydroxide present in the solution and its concentration. The optimal alkali additive concentration under which the NaB0 2 precipitate's interference with H 2 generation will be minimized is 10 wt% KOH. Only the addition of KOH was able to maintain and improve the solubility of NaB0 2 in alkaline aqueous solution and could possibly improve the hydrolysis reaction conversion efficiency. Raising the NaB0 2 solubility permits the use of more concentrated NaBH4 solution which increases the H 2 storage and generation capacity of the system, and reduces the frequency at which a spent solution tank would require emptying. The electrochemical recycling favors a reactant solution with a high concentration as it minimizes the volume of NaB0 2 solution required per unit of NaBH 4 produced. High H 2 evolution overpotentials are more likely to be achieved in highly concentrated alkaline aqueous solutions. Since the KOH-stabilized filtrate had the highest pH, KOH additions would extend the half-life of the NaBH4 solution and prevent the unsafe non-catalytic release of H 2 . Filtrate solutions containing KOH had the highest specific gravity, and the weight of the storage system would be higher with KOH as the stabilizer. Another practical implication is that the system weight increases as the hydroxide content increases. To some extent, a 10 wt% KOH addition would reduce the acceleration capabilities of the FCV. The addition of LiOH did not affect the solution properties in a way that was beneficial to either the NaBH4 H 2 generation or storage system or to the electroreduction of NaB02. From the XRD data and SEM pictures, it was not possible to confirm the hydration level of NaB0 2 precipitate from the various alkaline solutions. However, it is clear that the hydration level of the NaB0 2 by-product not only affects the mass and heat transport properties of the solution, but also lessens the theoretical gravimetric density of the H 2 storage and generation system. Hence, it is recommended to operate the H 2 generation system at a temperature favouring the formation of a less hydrated form of NaB02. 6.1.2 Future Research and Recommendations In the current work, it was assumed that NaB0 2 is the main NaBH 4 hydrolysis by-product in alkaline aqueous media. However, the exact composition of the NaBH4 hydrolysis alkaline aqueous waste solution as a function of pH and temperature should be verified to confirm that NaB0 2.2H 20 is the primary 138 hydrolysis borate formed and that the alkali additives do not react with NaBH 4 to form other complex polyborates. Among other things, the pH and temperature of NaBH4 alkaline aqueous solutions should be measured before and after hydrolysis. The difference between the NaBH 4 dissolution rate and the hydrolysis reaction rate should be compared as a function of pH and temperature. It would be useful to determine if other potassium-containing compounds than KOH would be suitable additives to enhance the solubility of NaB02. Solubility measurements should be carried out with NaB0 2 and NaBH4 combined in the alkaline aqueous solution to further investigate the common ion effect on the boron-containing anion solubility. Mixtures of KOH and NaOH should also be investigated to potentially optimize the NaB0 2 alkaline aqueous solution conductivity and solubility. The solubility measurements obtained at 5 and 7.5 wt% hydroxide require verification and special care should be taken to avoid NaB0 2.2H 20 dehydration during the course of the experiments. Other methodologies, such as atomic absorption (AA), should be used to verify the ICP-MS solubility measurements. The use of ammonia as a co-solvent should be investigated even though it is expected that low solubility ammonia borate compounds will be formed in the saturated aqueous NaB0 2 solutions. Conductivity and pH measurements should be taken prior to the solution filtration. To .ensure the validity of the pH measurements in concentrated sodium solutions, a low sodium error electrode should be used instead of a general purpose glass electrode. A suitable procedure has to be developed to determine the solubility of NaB0 2.2H 20 at 50-75°C as the solubility values measured were too high. A water bath capable of maintaining temperatures less than 5°C above ambient should be used. Experiments should be carried-out at temperatures close to PEMFC operating conditions and the freezing point of the various solutions should also be determined. Also, the effect of NaOH and KOH on the electroreduction of NaB0 2 to NaBH4 in water should be investigated. However, it has been reported that if the NaB0 2 electroreduction is not carried-out in an environment free of water, the production of NaBH 4 will be limited. Consequently, efforts should be focused on the use of dry or water free media such as organic solvents, ionic liquids, or molten salts, which are likely to be more favourable to successfully electroreduce NaB02. It is important to note that in an organic solvent, the conductivity is likely to be lower than in the aqueous solution, and that a suitable supporting electrolyte, like the ones studied in this thesis, would still be needed. 139 Future work should include electrochemical measurements in aqueous alkaline solutions of NaB0 2 to investigate the feasibility of the direct electroreduction of NaB0 2 to NaBH4 at various electrode materials. Organic additives should be used to increase the H 2 evolution reaction overpotential and thereby favor the reduction of NaB0 2 to NaBH4. In addition, the catalytic hydrolysis and electrooxidation of NaBH4 on a variety of electrode materials possessing catalytic affinities for the hydrolysis of NaBH4, such as Pt, Ru, Ni and Ag, should be looked at. The hydrolysis reaction rate should be studied by measuring the H 2 evolution rate using a specially designed set-up. Voltammetry at rotating disk electrodes and chronoamperometry at microdisk electrodes are among the techniques that could be used. Electrochemical quartz crystal microbalance and cyclic voltammetry measurements could be combined to identify potential unstable electro-oxidation reaction intermediates and rapid chemical reaction stages. 6.2 Summary of Contributions This thesis has gone some way towards enhancing the understanding of the effect of hydroxides on the solubility and physicochemical properties of NaB0 2 aqueous solutions. The physicochemical parameters determined in this study will form the basis upon which electrochemical recycling investigations NaB0 2 to NaBH4 in aqueous media will be pursued. The current findings add to a continuously growing body of literature on the use of NaBH4 as a H 2 carrier in emerging fuel cell power generation technologies, such as B-PEMFCs and DBFCs. It is believed that the low electrochemical activity of the NaB0 2 alkaline aqueous solutions could be alleviated by the development of an appropriate electrolyte media to carry the electroreduction. It was demonstrated that a practical electrochemical recycling procedure could open a wide range of commercial possibilities for NaBH4 and that a sustainable electrochemical method to recycle NaB0 2 to NaBH4 would remove one of the most important obstacles to the adoption of NaBH4 as the H 2 carrier of choice for automotive H 2 fuel cell applications. The success of this H 2 storage system clearly lies in the development of a sustainable and efficient electrochemical method to convert NaB0 2 to NaBH4, which would lead to major cost reductions. 140 R E F E R E N C E S A Afzal, M., Iqbal, M. J., and, Ahmad, H., (1994), Collect. Czech. Chem. Commun., 59, pp. 1296-1300. Aiello, R., Sharp, J. H , and Matthews, M. A , (1999), Int. J. Hydrogen Energy, 24, pp. 1123-1130. Amendola, S. C , (2002), US Patent 6497973. Amendola, S. C , Kelly, M.T., and Wu, Y., (2003), US Patent 6524542. Amendola, S. C , Onnerud, P., Kelly, M. T., Petillo, P. J.,.Sharp-Goldman, S. L and Binder, M. J., (1999), Journal of Power Sources, 84, pp. 130-133. Amendola, S. C , Sharp-Goldman, S. L., Saleem Janjua, M., Spencer, N. C , Kelly, M. T., Petillo, P. J., and Binder, M., (2000a), Int. J. Hydrogen Energy, 25, pp. 969-975. Amendola, S. C , Sharp-Goldman, S. L., Saleem Janjua, M., Spencer, N. C , Kelly, M. T., Petillo, P. J., and Binder, M., (2000b), Journal of Power Sources, 85, pp. 186-189. Andrew, H. E., (1976), Ceramic Bulletin, 55 (6), pp. 583-584. Antropov, L. I., (1972), Theoretical electrochemistry, Mir, Moscow, 567 p. Ay, M., Midilli, A., and Dincer, I., (2006), Journal of Power Sources, In Press. B Basson, W. D., Bohmer, R. G. and, Stanton, D. A., (1969), The Analyst, 94, pp. 1135-1141. BCS Inc., (2002), Energy and Environmental Profile of the US Mining Industry, Chapter 3, Potash, Soda, Ash and Borates, pp. 3-1. Belcher, B., Tully, S. W., and, Svehla, G., (1970), Anal. Chim. Acta, 50, pp. 261-267. Bettera, J., (2003), PSA Peugeot Citroen Fuel Cell Activities, Business Briefing: Global Automotive Manufacturing & Technology, pp. 1-5. Betty, K. R., and Day, G. T., (1986), The Analyst, 111 (3), pp. 455-457. Bingham, D., Wendt, K., and Wilding, B., (2004), Idaho National Engineering and Environmental Laboratory, US Department of Energy, Energy Efficiency and Renewable Energy, Presentation, http://wwwl.eere.energy.gov/hydrogenandfuelcells/pdfs/review04/st_4_wilding.pdf. Bingham, D. N., Wilding, B. M., Klinger, K. M., Zollinger, W. T., and Wendt, K. M., (2005), US Patent Application 20050077170 A l . Blasdale, W. C , and Slansky, C. M., (1938), Journal of the American Chemical Society, 61, pp. 917-920. Blumenthal, H., (1951), Analytical Chemistry, 23 (7), pp. 992-994. Bonhoff, K., Gronich, S., Garbak, J., and Hooker, D., (2005), DOE Hydrogen Program, FY 2005 Progress Report, VIII.A Vehicle Demonstrations, pp. 1143-1146. 141 Botello, J. C , Morales-Dominguez, E., Domingeuez, J. M., Gutierrez, A., Rojas-Hernandez, A., and Ramirez, M. T. (2003), Spectrochimica Acta Part A, 59, pp. 1477-1486. Bouaziz, R., (1962), Bulletin de la Societe Chimique de France, 5, pp. 1451-1459. Breslau, B. R., and Miller, I. F., (1970), Journal of Physical Chemistry, 74 (5), pp. 1056-1061. Brown, B. B., (1957), Journal of the American Chemical Society, 79, pp. 4241-4242. Brown, H. C , (1955), Analytical Chemistry, 27, pp. 156-158. Brown, H. C , Mead, E. J., and Subba Rao, B.C., (1955), Journal of the American Chemical Society, 77, pp. 6209-6213. Buckingham, D. A., and Lyday, P. A., (2002), Historical Statistics for Boron, US Geological Survey, Open File Report 01-006, pp. 1-6. C Celikkan, H., Aydin, H., and, Aksu, M. L., (2005), Turkish Journal of Chemistry, 29, pp. 519-524. Chaikin, S. W., (1953), Analytical Chemistry, 25 (5), pp. 831-832. Chandra, D., Reilley, J. J., and Chellappa, R., (2006), Journal of Materials, 58 (2), pp. 26-32. Chaurasia, S. C , (2004), National Center for Compositional Characterisation of Materials, 502 (2), pp. 265-270. Chowdoji Rao, K, Subha, M. C. S., and, Brahmaji Rao, S., (1999), Indian Journal of Chemistry, 28a, pp. 102-105. Cloutier, C. R., Alfantazi, A., and Gyenge, E., (2006), Materials Science Forum/Advanced Materials Research, Trans Tech Publications, In Press. Cloutier, C. R., Alfantazi, A., and Gyenge, E., (2007), Journal of Fuel Cell Science and Technology, In Press. Constable, F. Ft., and Isnel, A. FL, (1953), Reviews of Faculty of Science, University of Istanbul, Series C, 18a, pp. 166-177. Cooper, H. B. H., (1973), US Patent 3734842. Corti, H., Crovetto, R., and Fernandez-Prini, R., (1980), Journal of the Chemical Society, Faraday Transactions, 76 (10), pp. 2179-2186. Covington, A. K., and Newman, K. E., (1973), J.Tnorg.Nucl. Chem., 35, pp. 3257-3262. D Daimler Chrysler, Press Release, (2002), New York. Davis B , Lo, F., Calabretta, D , and Abdul, M , (2005), 44* Annual Conference of Metallurgists of CIM, Calgary, Alberta, Canada, pp. 595-605. Davis, E. R., Bromels, E., and Kibby, C. L., (1962), Journal of the American Chemical Society, 84 (6), pp. 885-892. 142 Dawidowicz, A. L., Matuzewitz, J., and Wysocka-Lisek, J., (1989), Talanta, 5, pp. 581-584. Dean, J. A., (1999), Lange's Handbook of Chemistry, 15th Edition, McGraw-Hill, USA. Department of Energy, (2003), Controlled H 2 Fleet and Infrastructure Demonstration and Validation Project, Draft Revision 3, Appendix A, pp. A1-A25. Department of Energy, (2004), Hydrogen Storage "Think Tank" Report, Office of Hydrogen, Fuel Cells and Infrastructure Technologies, Washington, D C , p. 6. Dhanesh, C , James, J. R., and Raja, C , (2006), Journal of Materials, 58(2), pp. 26-32. Drescher, A., (1989), Analytica Chemica Acta, 219, pp. 273-279. Drescher, A., and Kucharkowski, R., (1998), Wissenschaftliche Berichte, 38, pp. 8-19. Dodd, S., (1929), The Analyst, 54, pp. 282-285. Druzhinin, G. and Skvortsov, V.G., (1968), Russian Journal of Inorganic Chemistry, 13 (5), pp. 754-756. E Erdey-Gruz, T., (1974), Transport Phenomena in Aqueous Solutions, Adam Hilger Ltd., England. Eysseltova, J., and Dosoudilova, Jana, (1994), Collection Czech. Chem. Commun., 59, pp. 1337-1341. F Filby, E. E., (1976), US Patent 3993732. Frei, V., and Ustianovicova, A., (1963), Russian Journal of Physical Chemistry, 37 (5), pp. 612-613. G Garret, D. E., (1998), Borates, Handbook of Deposits, Processing, Properties and Use, Academic Press, London, UK, pp. 23-49. Gervasio, D., Xu, M. and Thomas, E., (2005), Properties of Aqueous Alkaline Sodium Borohydride Solutions and By-products Formed During Hydrolysis, Presentation, Fuel Cell Design, Fabrication and Materials Selection Workshop. Gouper, A. M., Pletcher, D., and Walsh, F. C , (1990), Chem. Rev., 90, pp. 837-865. Gyenge, E. L., and Oloman, C. W., (1998), Journal of Applied Electrochemistry, 28, pp. 1147-1151. H Hale, C. H. and Sharifian, H., (1990), US Patent 4931154. Hampton, D. S. Jr., and Wiersena, R. J., (1972), Inorganic Chemistry, 115, pp. 1152-1154. Hart, D., (2003), Frontiers in the Ecology and the Environment, 1(3), pp. 138-145. Hayes, M. R, and Metcalfe, J , (1962), The Analyst, 87, pp. 956-969. Helmy, F. M., Shebi, F. A., Sartori, P. and Solimam, A. D., (1987), Chemiker-Zeitung, 111 (12), p. 367. 143 Horvath, A. L., (1985), Handbook of Aqueous Electrolyte Solutions, Ellis Horwood Ltd., England. Hua, D., Hanxi, Y., Xinping, A., and Chuansin, C , (2003), Int. J. Hydrogen Energy, 28, pp. 1095-1100. Huff, G. F., Chapel, F., McElroy, A. D., City, E. and Adams, R. M., (1958), US Patent 2855353. J Jacques S., Berthet, M. P. and Bonnetot, B., (1970), Prog. Inorg. Chem., 11, pp. 99-231. . Jacques, S., Berthet, M. P., and Bonnetot, B., (2004), Chemical Engineering Transactions, 4, pp. 331-336. Jenkins, H. D. B., Roobottom, H. K., Passmore, J., and, Glasser, L., (1999), Inorganic Chemistry, 38, pp. 3609-3620. Jeong, S. U., Kim, R. K., Cho, E.A., Kim, H.-J., Nam, S.-W., Oh, I.-H., Hong, S.-A. and Kim, S.H., (2005), Journal of Power Sources, 144 (1), pp. 129-134. Jigong, L., Foutch, G. L., and Na, J. W., (1999), Separation Science and Technology, 34 (15), pp. 2923-2941. John, K. R. C , Jr., (1951), Analytical Chemistry, 24, p. 806. Johnson, E. A., and Toogood, M. J., (1954), The Analyst, 79, pp. 493-496. K Kaufman, C. M., and Sen, B., (1985), J. Chem. Soc, Dalton Trans., pp. 307-312. Kesans, A., Vimba, S., and Svarcs, E., (1955), Latvijas PSR Zinatnu Akad. Vestis, 8, pp. 127-134. Kim, J.-H., Lee, H., Han, S.-C, Kim, H.-S., Song, M.-S., and Lee, J.-Y., (2004a), Int. J. Hydrogen Energy, 29, pp. 263-267. Kim, J.-H., Kim, K.-T., Kang, Y.-M., Kim, H.-S., Song, M.-S., Lee, Y.-J. Lee, P. S., and Lee, J.-Y., (2004b), J. Alloys and Compounds, 379, pp. 222-227. Kirk, R. E., Othmer, D. F., Kroschwitz, J. I., and Howe-Grant, M., (1992), Kirk-Othmer Encyclopedia of Chemical Technology. Knorre, H., Main, H.A., Kloepfer, H, Main, F. A., and Bretschneider, G., (1968), US Patent 3379511. Kocher, J., and Lahlou, A., (1970), Bulletin de la societe chimique de France, 6, pp. 2083-2089. Kojima, Y., and Haga, T., (2003), Int. J. Hydrogen Energy, 28, pp. 989-993. Kojima, Y., Kawai, Y., Nakanishi, H., and Matsumoto, S., (2004a), J. Power Sources, 135, pp. 36-41. Kojima, Y., Suzuki, K.-L, and Kawai, Y., (2006), Journal of Power Sources, 155, pp. 325-328. Kojima, Y., Suzuki, K.-L, Fukumoto, K., Sakai, M., Yamamoto, T., Kawai, Y., and Hayashi, H., (2002), Int. J. Hydrogen Energy, 27, pp. 1029-1034. Kojima, Y., Suzuki, K.-L, Fukumoto, K., Kawai, Y., Kimbara, M., Nakanishi, H., and Matsumoto, S., (2004b), Journal of Power Sources, 125, pp. 22-26. 144 Kong, V. C. Y., Foulkes, F. R., Kirk, D. W., and Hinatsu, J.T., (1999), Int. J. Hydrogen Energy, 24, pp. 665-675. Konoplev, V. N., (1988), Russian Journal of Inorganic Chemistry, 33, pp. 433-434. Kreevoy, M. M., and Jacobson, R.W., (1979), Ventron Alembic, 15, pp. 2-3. Krishnan, P., Yang, T.-H., Lee, W.-Y., and Kim, C.-S., (2005), Journal of Power Sources, 143, pp. 17-23. Kuleshova, O. D., (1985), Journal of analytical chemistry of the USSR, 40 (1), pp. 151-153. L Larminie, J., and Dicks, A., 2003, Fuel Cell Systems Explained, Second Edition, John Wiley and Sons Ltd., West Sussex, England, pp. 297-301. Lencka, M. M., Anderko, A., Sanders, S. J., and Young, R. D., (1998), International Journal of Thermophysics, 19 (2), pp. 367-378. Levy, A., Brown J. B., and Lyons, C.J., (1960), Industrial and Engineering Chemistry, 52, pp. 211-214. Li, J., Li, B., and Gao, S., (2000), Xi-an Branch, Phys. Chem. Minerals, 27, pp. 342-346. Li, Z. P., Liu, B. H., Arai K., and Suda, S., J., (2003b), Electrochemical Soc, 150 (7), pp. 868-872. Li, Z. P., Morigazaki, N., Liu, B. H., and Suda, S., (2003a), J. Alloys and Compounds, 349, pp. 232-236. Li, Z. P., Liu, B. H., Arai, K., Asaba, K., and Suda, S., (2004), Journal of Power Sources, 126, pp. 28-33. Lide, D. R., (2005-6), CRC Handbook of Chemistry and Physics, 86th Edition, CRC Press, USA. Linden, D., (1984), Handbook of Batteries and Fuel Cells, McGraw Hill Inc., pp. 42-48. Linkous, C. A., (2003-4), Florida Solar Energy Center, Task Number III-C, http://www.fsec.ucf.edu/hydrogen/reports/taskiiic-linkous-fsec.pdf. Linkous, C. A., Bhuller, S. K. and Slattery, D. K., (2004), A H2 Storage Cycle Based on Alkali Metal Borohydrides, Conference Abstract, 206th Meeting of the Electrochemical Society, p. 550. Lopez Garcia, I., Hernandez Cordoba, M., and, Sanchez-Pedreno, C , (1985), The Analyst, 100 (10), pp. 1259-1262. Lyday, P. A., (2000), Boron, U.S. Geological Survey Minerals Yearbook, pp. 14.1-14.8. Lyday, P. A., (2005), Boron Statistics, US Geological Survey, Mineral Commodity Summaries, Annual Publication, pp. 36-37. Lyttle, D. A., (1963), Journal of analytical chemistry, 24 (11), pp. 1843-1844. M Maksimova, I. N., and Yushkevich, V. F., (1963), Russian Journal of Physical Chemistry, 37 (8), pp. 1005-1007. 145 Mal'tseva, N. N., Kedrova, N. S., Sizareva, A. S., and Konoplev, (1991), Russian Journal of Inorganic Chemistry, 36 (1), pp. 37-38. Martin, D. R., (1961), Borax to Boranes, Advances in Chemistry Series, American Chemical Society Applied Publications, pp. 1-4. Martinus, N., Crawford, D., Sinclair, D., and Vincent, C. A., (1977), Electrochimica Acta, 22 (10), pp. 1183-1187. Mattson, B., (2005), Educacion Quimica, 16 (4), pp. 114-127. Mellor, J. W., (1980), Supplement to Mellor's Comprehensive Treatise on Inorganic and Theoretical Chemistry, Longman Group Ltd., N.Y., USA. Menzel, H., and Schulz, H., (1943), Zeitschrift fur Anorganische und Allgemeine Chemie, 251, pp. 167-200. MERIT (Materials & Energy Research Institute Tokyo), Ltd., (1986), Nagano-Ken, Japan, http://merit.hvdrogen.co.ip/Whv BH4/WhvBH4.html. Mesmer, R. E., (1972), Inorganic Chemistry, 11 (3), pp. 537-544. Midgley, D., (1992), The Analyst, 117, pp. 199-202. Mikio, K., and Hitoshi, I., (2003), Japanese Patent Application JP 2003247088. Mirkin, M. V., (1991), Anal. Chem., 63, pp. 532-533. Mirkin, M. V., and Bard, A. J., (1991), Anal. Chem., 63, pp. 532-533. Mohring, R. M., and Wu, Y., (2003), American Institute of Physics Conference Proceedings, 671, pp. 90-100. Momii, R. K., and Nachtrieb, N. H., (1967), Inorganic Chemistry, 6 (6), pp. 1189-1192. N Nies, N. P., and Hulbert, R. W., (1967), Journal of Chemical and Engineering Data, 12 (3), pp. 303-313. Norkus, P. K., (1968), Journal of Analytical Chemistry of the URSS, 23 (4), pp. 558-562. O Out, D. J. P., and Los, J. M., (1980), J. Solut. Chem., 9 (1), pp. 19-35. P Pinto, A. M. F. R., Falcao, D. S., Silva, R. A , and Rangel, C. M., (2006), Int. J. of Hydrogen Energy, In Press. R Richardson, B. S., Birdwell, J. F., Pin, F. G., Janson, J. F., and Lind, R. F., (2005), Journal of Power Sources, 145, pp. 21-29. 146 Roobottom, H. K., Jenkins, H. D. B., Passmore, J., and, Glasser, L., (1999), Journal of Chemical Education, 76 (11), pp. 1570-1573. Ross, W. J., and, White, J. C , (1960), Talanta, 3, pp. 311-317. Rothbaum, H. P., Tood, H. J. and Walker, I. K , (1936), Chemical Engineering Data, 1 (1), pp. 95-99. Sadetdinov, Sh. V., (1985), Russian Journal of Inorganic Chemistry, 30, pp. 1913-1915. Roy, M. A., (1964), Boron, Metallo-Boron Compounds and Boranes, John Wiley & Sons, pp. 390. Rudie, C. N., (1979), Journal of the American Chemical Society, 56 (4), pp. 520-521. S Sadetdinov, Sh. V., (1985), Russian Journal of Inorganic Chemistry, 30, pp. 1913-1915. Salentine, C. G., (1983), Ignore. Chem., 32, pp. 3920-3924. Sagrado, S., (1998), Talanta, 45, pp. 835-842. Sah, R. N., (1997), Plant and Soil, 193, pp. 15-33. Schlesinger, H. I., Brown, H. C , Finholt, A. E., Gilbreath, J. R., Hoekstra, H. R. and Hyde, E. K., (1953), J. Am. Chem. Soc, 75, pp. 215-219. Seijiro, S., (2004), Patent JP2004-010446. Sharifian, H., and Dutcher, J. S., (1990), US Patent 4904357. Shedlovsky, T., (1938), J. Franklin Institute, 225, p. 739-743. Skvortsov, V. G., (1974), Russian Journal of Inorganic Chemistry, 19 (7), pp. 1088-1089. Skvortsov, V. G., Molodkin, A. K., Rodionov, N. S. and, Tsekhanskii, R. S., (1986a), Russian Journal of Inorganic Chemistry, 31 (6), pp. 934-935. Skvortsov, V. G., Molodkin, A. K., Rodionov, N. S. and Tsekhanskii, R. S., (1986b), Russian Journal of Inorganic Chemistry, 31, pp. 911-913. Skvortsov, V. G., Petrova, O. V., Belova, V. F., Sadetdinov, Sh. V., (1992), Russian Journal of Inorganic Chemistry, 37 (4), pp. 475-475. Skvortosov, V. G., Tsekhanskii, R. S., Molodkin, A. K., Petrova, O. V. and Akimov, V. M., (1988), Russian Journal of Inorganic Chemistry, 33 (6), pp. 921-923. Snover J., and Wu, Y., (2004), US Patent 6706909 Bl . Snyder, K. A., (2005), Cement and Concrete Research, 35, pp. 421-422. Staden, J. F., (2000), S. Afr. J. Chem., 53 (2), pp. 77-85. Stockmayer, W. H., Rice, D. W., and Stephenson, C. C , (1955), J. Am. Chem. Soc, 77, pp. 1980-1983. Suda, S., (2003), The Sodium Borohydride Hydrogen Energy Systems, R&D Policies, KUCEL-MERIT Group, pp. 1-11. 147 Suda, S., Sun, Y. M., Liu, B.-H., Zhou, Y., Morimitsu, S., Arai, K., Tsukamoto, N., Uchida, M., Candra, Y., and Li, Z.-P., (2001), Applied Physics A, Materials Science and Processing, 725, pp. 209-212. Sun, Y., and Liang, Z., (2003), Patent CN1396307. Szklarska-Smialowska, S., and Smialowski, M., (1963), J. of the Electrochemical Society, 110, pp. 444-448. U Uddin, F , and Saeed, R., (2001), Acta Cientifica Venezolana, 52, pp. 186-191. Ullmann, F., (2000), Ullmann's Encyclopedia of Industrial Chemistry. Urusova, M. A., and Valyashko, V. M., (1993), Institute of General and Inorganic Chemistry, Academy of Sciences of the URSS Journal of Inorganic Chemistry, 38 (4), pp. 662-664. US Borax, NaB0 2.2H 20 product data sheet. V Van Staden, J. F., and Van Der Merwe, T. A., (2000), The Analyst, 125, pp. 2094-2099. W Wee, J.-H., (2006), Journal of Power Sources, 155, pp. 329-339. Wei, W. C , Chen, C.-J., and, Yang, M.-H., (1995), Journal of Analytical Atomic Spectrometry, 10, pp. 955-961. Wei, X.-Y., Zhang, L.-J., Liang, J.-J., Wei, Z.-M., and Weng, D.-H., (2003), Huagong Jishu Yu Kaifa, 32 (3), pp. 1-3. Weiren, R., and Deren, Q., (1998), Journal of Fudan University, 37 (3), pp. 276-278. Wilding, B., Bingham, D., Wendt, K., and Klingler, K., (2004), Hydrogen Storage: Radiolysis for Borate Regeneration, DOE Hydrogen Program, Progress Report, pp. 200-204. Wipke, K., Welch, C , Thomas, H., Sprik, S., Gronich, S., Garbak, J., Hooker, D., (2006), Controlled Hydrogen Fleet and Infrastructure Demonstration and Validation Project, Progress Update, National Hydrogen Association Conference, Long Beach, California, pp. 1-21. Wu, C , Zhang, H., and Yi, B., (2004), Catalysis Today, 93-95, pp. 477-483. Wu, Y., (2003a), Hydrogen Storage via Sodium Borohydride, Presentation, GCEP Hydrogen Workshop, Stanford University, CA, USA, http://gcep.stanford.edu/pdfs/hydrogen_workshopAVu.pdf. Wu, Y., (2003b), Process for the Regeneration of Sodium Borate to Sodium Borohydride for Use as Hydrogen Storage Source, Fuel Cells and Infrastructure Technologies Program Review, DOE Hydrogen Program, Progress Report, pp. 1-2. Wu, Y., (2004), Low-Cost, Off-Board Regeneration of Sodium Borohydride, Fuel Cells and Infrastructure Technologies Program Review, DOE Hydrogen Program, Presentation, http://wwwl.eere.energy.gov/hydrogenandfuelcells/pdfs/review04/st_5p_wu.pdf. 148 Wu, Y., Kelly, M. T., and Ortega, J. V., (2004a), Low-Cost, Off-Board Regeneration of Sodium Borohydride, DOE Hydrogen Program, Progress Report, pp. 195-199. Wu, Y., Kelly, M. T., and Ortega, J. V., (2004b), Review of Chemical Processes for the Synthesis Sodium Borohydride, DOE Cooperative Agreement DE-FC36-04GO14008, pp. 1-24. X Xia, Z. T , and Chan, S. H., (2005), Journal of Power Sources, 152, pp. 46-49. Xu, J., Kelly, M., Pez, G., Wu, Y. and Sharp-Goldman, S., (2004), US Patent 20040011662. Y Yu, Z., (2002), Patent WO 02062701. Z Zhang Q., Smith, G., Wu, Y. and Mohring, R., (2006), Int. J. of Hydrogen Energy, In Press. APPENDIX Appendix A: Comparison to DOE Range, Storage Cost, Energy Density and Specific Energy Targets. Table A. 1: Fuel Cost for Target Range and Storage Cost Estimation. Fuel Cost for Range Target 1 [mol NaBH4] produces 4 [moles of H2] NaBH4 Cost (Specialty Chemical) = $80-100US/kg 37.83 [g NaBH4] produces 4*(2.0158) [g of H2] NaBH4 Cost (Industry) = $50US/kg [g NaBH4] needed to produce 8000 [g of H2]: [g NaBH4] needed to produce 1000 [g of H2]: 37533.49 [g] 4691.69 [g] 37.53 [kg] 4.69 [kg] [US$] for 8 kg H 2 : [US$] for 1 kg H 2 : Specialty Chemical Industry Specialty Chemical Industry $3,002.68 $3,753.35 $1,876.67 $375.33 $469.17 $234.58 (min.) (max.) (average) (min.) (max.) (average) DOE 2015 Target for Range is: 8 [kg H2] for 400 [km] Storage Cost Estimation [$/k\Vh] 0.07 [Wh/molH2] 0.03 [kWh/gH2] 261.65 [kWh] for 8000g H 2 565.15 [$] for 8 kg H 2 (Specialty Chemical) NaBH4 cost is: 5.31 times too high DOE 2015 Storage Cost Target is: 2.16 [$US/kWh] Complete hydrolysis of NaBH 4 produces 2.37 [1] H 2 at STP per gram of NaBH 4 37.83 [g NaBH4] produces 4*(2.0158) [g of H2] 1 [g NaBH4] needed to produce x[g of H2]: 0.21 [g H 2 produced] 1 [mol H2] = 2.0158 [g H 2 produced] ? [mol] =0.106571504 [g H2] 0.11 [mol H 2 produced] 22.4 [1H2 per mol] at STP 237 [1H2 per g NaBH4] at STP 150 Table A.2: Volumetric and Gravimetric Energy Density Comparison. Volumetric Energy Density [KWh/1] For PEMFC: H 2 2H + + 2c 2F per 1 mol Ff2 consumed = 2*96485C = [A.s] = 2*26.8 [A.h] -1.23 [V] Theoretical Energy Density 65.93 [Wh/molH2] 0.07 [kWh/molH2] PEMFC Feed with H 2 : Ideal Gas Law 200 [atm] PV = nRT 25 [°C] V = nRT/P 298.15 [°K] 0.12 [1] Theoretical Energy Density 0.54 [kWh/lH2l PEMFC Feed with NaBH 4 fuel: 1 [mol NaBH4] forms 4 [moles H2] Theoretical Energy Density 0.26 [kWh/molNaBH4] Assuming 20 wt % NaBH4 in 2M NaOH in water 200 [g NaBH4] per 1000 [g solution] 5.29 [mol NaBHVl or kg solution] Theoretical Energy Density 1.39 [kWh/1 solution] DOE 2015 Target is: 2.69 [kWh/1] Gravimetric Energy Density [kWh/kg] 4 [moles H2] = 8.63 [gH2] 200 [g NaBH4] = 5.29 [mol NaBH4] 1 [mol NaBH4] = 4 [moles H2] 5.29 [molNaBH4] = 21.15 [mol H2] 21.15 [mol H2] = 45.63 [gH2] 0.05 ' [kgH2] 0.26 [kWh/molNaBH4] 1.39 [kWh/0.05 kgH2] 27.5 [kWh/kg] DOE 2015 Target is: 3.00 [kWh/kg] Appendix B: Filtrate Characterization Data Table B. 1: Physicochemical Properties Raw Data, Unsaturated Diluted Aqueous Solutions of NaB0 2 at 25°C ±3°C, Trial 1. Concentration Density Conductivity Dynamic Viscosity fMl [g/mll TS/ml rcPi 0.4910 1.0329 2.35 1.0916 0.3683 1.0251 1.86 1.0590 0.2455 1.0145 1.40 1.0135 0.1841 1.0094 1.08 0.9701 0.1228 1.0087 0.98 0.9677 0.0614 1.0008 0.42 0.9403 0.0246 0.9983 0.19 0.9432 0.0184 0.9968 0.14 0.9252 0.0123 0.9977 0.10 0.9169 0.0061 0.9964 0.04 0.9161 0.0025 0.9971 0.02 0.9159 Table B.2: Physicochemical Properties Raw Data, Unsaturated Diluted Aqueous Solutions of NaB0 2 at 25°C ± 3°C, Trial 2. Concentration Density Conductivity Dynamic Viscosity [M] [g/ml] rs/mi [cP] 0.4910 1.0334 2.29 1.0861 0.3683 1.0235 1.87 1.0356 0.2455 1.0156 1.34 0.9972 0.1841 1.0111 1.09 0.9762 0.1228 1.0043 0.78 0.9506 0.0614 1.0005 0.42 0.9424 0.0246 0.9975 0.18 0.9201 0.0184 0.9986 0.15 0.9182 0.0123 0.9982 0.11 0.9162 0.0061 0.9972 0.05 0.9124 0.0025 0.9965 0.02 0.9043 Table B.3: Physicochemical Properties Raw Data, Unsaturated Diluted Aqueous Solutions of NaB0 2 at 25°C±3°C,Trial3. Concentration Density Conductivity Dynamic Viscosity [Ml [g/ml] [S/m] [cP] 0.4910 1.0352 2.35 1.0890 0.3683 1.0230 1.89 1.0233 0.2455 1.0159 1.38 1.0030 0.1841 1.0110 1.12 0.9657 0.1228 1.0063 0.75 0.9487 0.0614 ' 0.9987 0.42 0.9442 0.0246 0.9987 0.18 0.9293 0.0184 0.9975 0.14 0.9302 0.0123 0.9959 0.10 0.9189 0.0061 0.9967 0.05 0.9165 0.0025 0.9974 0.03 0.9082 152 Table B.4: Summary of Averages, Conductivity of Unsaturated NaBC>2 Aqueous Alkaline Solutions Filtrate at 25, 50 and 75°C. Solution Composition No additive NaOH KOH LiOH Alkali T Conductivity [wt%] r c i [S/m] 0 25.00 27.25 0 50.00 36.25 0 75.00 59.40 1 25.00 52.30 2 25.00 91.00 3 25.00 125.30 5 25.00 188.50 10 25.00 -1 50.00 82.55 2 50.00 136.30 3 50.00 174.27 5 50.00 -10 50.00 -0.5 75.00 65.85 1 75.00 132.10 1.5 75.00 169.50 2 75.00 -3 75.00 -5 75.00 -10 75.00 -1 25.00 41.20 2 25.00 70.33 3 25.00 98.15 5 25.00 156.95 10 25.00 -1 50.00 82.25 2 50.00 114.10 3 50.00 157.80 5 50.00 -10 50.00 -0.5 75.00 68.90 1 75.00 121.50 1.5 75.00 150.30 2 75.00 193.30 3 75.00 -5 75.00 -10 75.00 -1 25.00 47.20 2 25.00 78.07 3 25.00 108.60 5 25.00 171.90 10 25.00 _ 153 Table B.5: Filtrate Physicochemical Properties Raw Data, Saturated Aqueous Solutions of NaBOz at 25°C ±3°C. Solution pH Density Conductivity NaB02.2Ff20 Viscosity ID Code [kg/m3l rs/mi [wt %] [cPl 22 31.46 24 32.40 26 30.80 53 28.54 54 28.83 55 30.14 82 30.61 83 30.84 94 12.77 95 12.61 30.31 96 12.59 29.77 97 12.55 1207.34 30.50 98 12.53 1232.22 28.65 119 12.77 1249.87 24S$ 144 A 13.13 1219.90 144 B 13.11 1233.82 53.60 188 12.95 1286.92 51.90 5.66 190 12.87 1299.00 52.50 4.97 viscometer test A OA U . O 1 # of Observations 10 7 3 8 2 Arithmetic Mean 12.79 1247.01 52.67 30.45 5.31 Standard Deviation 0.22 34.17 0.86 1.28 0.49 Standard Error 4.26 509.25 37.24 11.52 5.32 Table B.6: Filtrate Physicochemical Properties Raw Data, Saturated Aqueous Solutions of NaB0 2 stabilized with 7.5 wt% KOH at 25°C ± 3°C. Solution pH Density Conductivity NaB0 2.2H 20 Viscosity ID Code rkg/m3l rs/mi [wt %1 • [cPl 268 UA9 1300.09 98.40 27.85 6.14 271 14.91 1359.02 96.90 28.20 6.52 274 U^S 1310.97 108.00 28.1271 293 14.70 1382.09 92.30 6.38 294 14.51 1325.61 91.90 38.1335 6.31 295 14.50 1345.96 91.60 6.45 # of Observations 4 6 6 3 5 Arithmetic Mean 14.66 1337.29 96.52 28.06 6.36 Standard Deviation 0.19 30.88 6.31 0.19 0.14 Standard Error 8.46 598.19 43.24 19.84 3.18 154 Table B.7: Surnmary of Averages, Effect of Alkali Hydroxide Addition on the Physicochemical Properties of Saturated NaB0 2 Aqueous Solutions Filtrate at 25°C ± 3°C. Solution Alkali pH Density Conductivity Viscosity Solubility Composition [wt%] [kg/m3] [S/m] [cP] [wt %] Water 0 6.14 967.00 1.5uS/m 0.86 -No alkali 0 12.79 1247.01 52.67 5.31 30.45 NaOH 1 13.73 1261.54 60.85 3.93 30.52 NaOH 2 14.14 1267.94 75.93 4.23 29.50 NaOH 3 14.23 1279.05 92.00 5.08 28.60 NaOH 5 14.33 1281.11 99.77 5.99 29.83 NaOH 7.5 14.49 1292.09 142.97 6.11 18.38 NaOH 10 14.62 1293.57 198.27 6.37 18.09 KOH 1 13.79 1281.60 • 52.30 5.58 31.67 KOH 2 14.07 1290.98 53.70 6.14 31.72 KOH 3 14.21 1298.71 62.85 6.30 32.11 KOH 5 14.34 1319.37 68.18 6.32 31.46 KOH 7.5 14.66 1337.29 96.52 6.36 28.06 KOH 10 14.87 1341.81 167.05 • 6.54 32.27 LiOH 1 13.28 1269.87 46.30 6.51 31.72 LiOH 2 13.68 1277.63 49.60 7.49 30.23 LiOH 3 13.76 1286.65 51.05 8.32 29.92 LiOH 5 13.85 1310.34 62.76 10.01 31.51 LiOH 7.5 13.97 1319.50 81.50 10.92 26.57 LiOH 10 14.10 1324.74 150.28 11.38 18.68 155 Table B.8: Summary of Averages, Effect of Temperature on the Physicochemical Properties of Saturated NaB0 2 Aqueous Alkaline Solutions Filtrate at 25, 50 and 75°C. Solution Alkali T pH Density Conductivity Viscosity Solubility Composition [wt%] p q [kg/m3l [S/m] [cP] [wt %] No alkali 0 25 12.79 1247.01 52.67 5.31 30.45 No alkali 0 50 13.74 1554.50 60.74 23.88 92.22 No alkali 0 75 10.52 1583.00 86.89 2.51 113.56 NaOH 1 25 13.73 1261.54 60.85 3.93 30.52 NaOH 5 25 14.33 1285.99 99.77 5.99 31.09 NaOH 10 25 14.62 1293.57 198.27 6.37 18.09 NaOH 1 50 12.33 1623.50 56.51 30.46 89.24 NaOH 5 50 12.98 1479.20 83.61 11.84 80.09 NaOH 10 50 12.10 1514.59 - - 62.01 NaOH 1 75 10.65 1568.86 - 2.17 100.32 NaOH 5 75 10.09 - - - -NaOH 10 75 - - - - -KOH 1 25 13.79 1281.60 52.30 5.58 31.67 KOH 5 25 14.34 1319.37 68.18 6.32 38.96 KOH 10 25 14.87 1341.81 167.05 6.54 32.27 KOH 1 50 12.17 1577.29 44.20 28.89 85.76 KOH 5 50 13.57 1549.26 69.85 83.16 76.14 KOH 10 50 11.66 1532.97 131.43 8.70 62.05 KOH 1 75 10.63 1593.64 88.55 2.85 103.68 KOH 5 75 - - - - 100.15 KOH 10 75 - 1380.25 - - 77.51 Table B.9: Summary of Averages, Effect of Glycine on the Physicochemical Properties of Saturated NaB0 2 Aqueous Alkaline Solutions Filtrate at 25°C ± 3°C. Solution Alkali Glycine pH Density Conductivity Viscosity NaB0 2.2H 20 Composition [wt%] [wt%]_ [kg/m3] [S/m] [cPl [wt %] No additive 0 0 12.79 1247.01 52.67 5.31 30.45 Glycine 0 10 11.80 1266.48 48.75 2.18 0.68 NaOH 5 0 14.33 1281.11 99.77 31.90 29.83 Glycine and NaOH 5 10 11.86 1248.81 50.75 1.77 0.78 KOH 5 0 14.34 1319.37 68.18 34.53 31.46 Glycine and KOH 5 10 12.55 1286.48 56.55 1.97 1.06 LiOH 5 0 13.85 1310.34 62.76 39.40 31.51 Glycine and LiOH 5 10 12.03 1283.41 36.40 3.14 1.17 156 Appendix C: Precipitate XRD Data Table C.l: Summary of Three Most Important XRD Peaks for Various Boron Compounds from Different Sources. Chemical Formula PDF Source Reference ID Code 1st Max Peak 2-theta A%/Int 2nd Max Peak 2-theta A%/Int 3rd Max Peak 2-theta A%/Int NaB02.2H20 JCPDS 06-0122 16.745 100 23.39 65 31.137 30 NaB02.2H20 Jade 081-1512 16.768 100 23.385 50.9 17.631 46.5 NaBH4.2H20 JCPDS 26-1233 15.491 100 33.18 30 - -NaB02(OH)2 JCPDS 27-1222 6.866 100 9.994 100 15.919 100 Na 2B 20 5.4H 20 JCPDS 24-1054 18.163 100 36.803 60 30.44 50 NaB0 2 JCPDS 32-1046 29.131 100 32.853 75 34.277 75 NaB0 2 JCPDS 37-0115 33.903 100 38.382 60 40.033 60 B 2 0 3 JCPDS 44-1085 32.054 100 26.002 51 43.08 40 B 2 0 3 JCPDS 24-0160 32.042 100 32.136 67 26.002 56 B 2 0 3 JCPDS 13-0570 27.768 100 14.556 35 23.389 25 B 2 0 3 JCPDS 06-0634 43.16 100 32.199 90 40.448 70 B 2 0 3 JCPDS 06-0297 27.768 100 14.556 35 39.854 14 Na 20 JCPDS 23-0528 46.133 100 32.172 41 27.769 33 Na 20 JCPDS 03-1074 46.534 100 32.411 40 27.946 40 Na 2B 20 4.H 20 JCPDS 20-1078 18.641 100 24.87 95 31.274 80 Na 2B 20 4 .H 20 JCPDS 16-0242 16.713 100 18.625 100 25.062 100 Na 2B 20 4.2H 20 JCPDS 14-0678 22.375 100 25.575 100 30.697 100 Na 2B 20 5.4H 20 Jade 076-0756 31.28 100 33.819 78.9 25.722 71.7 Na 2B 20 4.8H 20 JCPDS 14-0677 18.625 100 25.801 100 31.36 100 Na 2B 20 4.8H 20 JCPDS 09-0011 31,274 100 33.563 90 18.8 80 LiB0 2 JCPDS 18-0738 25.228 100 41.201 100 50.686 80 LiB0 2 JCPDS 20-0619 35.553 100 17.632 60 13.096 60 LiOH.H 20 Jade 025-0486 33.73 100 30.064 66 36.962 45 LiOH.H 20 Jade 076-1074 33.562 100 30.055 62.6 36.932 48.8 LiOH.H 20 Jade 075-0883 33.79 100 30.116 62.6 37.075 48.4 K B 0 2 JCPDS 19-0979 31.73 100 29.281 100 - -K B 0 2 JCPDS 03-0729 14.443 100 13.393 83 17.969 33 157 Table C.2: Summary of Three Most Important XRD Peaks for the Precipitates of Saturates Alkaline Aqueous Solutions of NaB0 2 at 25°C ± 3°C. Additive Additive Max Peak 2nd Max Peak 3rd Max Peak 4th Max Peak 5th Max Peak MDI Jade 7 [wt %] 2-theta A% 2-theta A% 2-theta A% 2-theta A% 2-theta A% Sample ID None - 41.444 100 30.698 70 26.922 63 20.403 62.2 16.734 61.4 unknown None - 16.762 100 17.633 18.1 30.952 12 23.188 10.1 30.49 9 NaB02.2H20 None - 16.753 100 23.352 18.5 31.053 8.3 45.753 8.2 46.666 7.9 NaB0 2.2H 20 None - 16.809 100 31.169 6.5 34.003 5.9 45.295 2.5 - - unknown NaOH 10.00 16.683 100 31.214 24.4 33.896 14.9 23.874 13.9 - - NaB0 2.2H 20 NaOH 10.00 16.786 100 23.355 14.4 38.503 14.4 31.053 11.8 - - NaB0 2.2H 20 NaOH 10.00 17.086 100 37.359 13.9 46.022 13.7 31.361 8.3 23.679 8.2 unknown KOH 10.00 31.26 100 33.755 63.6 16.736 58.1 25.701 34.1 - - NaB0 2.4H 20 KOH 10.00 17.128 100 18.944 14.3 32.56 11.1 31.337 10 41.511 8.5 NaB0 2.2H 20 KOH 10.00 16.889 100 38.608 14.3 34.046 6 23.483 5.7 31.156 5.4 NaB0 2.2H 20 KOH 10.00 16.952 100 31.228 9.9 23.502 4.3 34.048 3.1 - - NaB0 2.2H 20 KOH 10.00 16.793 100 16.948 15.8 16.665 15 31.143 11.1 23.406 11 NaB0 2.2H 20 KOH 10.00 54.97 100 23.424 77.6 31.255 68.1 16.511 56.2 54.935 50.1 unknown LiOH 10.00 31.154 100 16.462 35.8 17.585 25.6 25.684 21.2 - - NaB0 2.4H 20 LiOH 1.00 30.436 100 20.146 87 35.708 52.9 29.652 39 18.762 37.5 tincalconite LiOH 10.00 33.654 100 33.654 78.7 29.957 54.8 39.75 44.4 18.462 43.8 unknown LiOH 10.00 22.946 100 16.745 64 30.548 49.7 33.598 32.8 17.558 32.4 unknown 158 Appendix D: Precipitate SEM Pictures Fig. D.l: Scanning Electron Microscope Precipitate Crystal Images of Pure NaB0 2.2H 20 (a) SE, WD27.6mm, x200, 200um, (b) SE, WD25.5mm, x700, 50um, (c) BSE2, WD 14.3mm, x200,200um, (d) BSE2, WD14.1 mm, x200, 200um, (e) SE, WD26.0mm, x2.5k, 20um, (f) SE, WD26.0mm, x2.5k, 20um. 159 Fig. D.2: Scanning Electron Microscope Precipitate Crystal Images of Precipitates formed from Saturated Aqueous Solutions of NaB0 2 stabilized with 10 wt% NaOH (a) SE, WD26.2mm, x500, lOOum, (b) SE, WD25.5mm, x700, 50um, (c) SE, WD25.5mm, xl.Ok, 50um, (d) SE, WD26.1mm, x2.0k, 20um, (e) BSE2, WDH.Omm, x5.0k, lOum, (f) BSE2, WDH.Omm, xlOk, 5um. 160 Fig. D.3: Scanning Electron Microscope Precipitate Crystal Images of Precipitates formed from Saturated Aqueous Solutions of NaB0 2 stabilized with 10 wt% KOH (a) BSE2, WD13.9mm, xl.Ok, 50um, (b) BSE2, WD 13.9mm, x3.0K, lOum, (c) SE, WD27.4mm, x500, lOOum, (d) SE, WD27.4mm, xl.Ok, 50um, (e) SE, WD27.1mm, x3.0k, lOum, (f) SE, WD28.1mm, x5.0k, lOum. 161 Fig. D.4: Scanning Electron Microscope Precipitate Crystal Images of Precipitates formed from Saturated Aqueous Solutions of NaB0 2 stabilized with 10 wt% LiOH (a) BSE2, WD14.1mm, xl .OK, 50um, (b) BSE2, WD14.1mm, x3.0k, lOum, (c) SE, WD29.3mm, x500, lOOum, (d) SE, WD29.4mm, xl.Ok, 50um, (e) BSE2, WD13.7mm, x500, lOOum, (f) BSE2, WD13.7mm, xl.Ok, 50um. 162 Fig. D.5: Scanning Electron Microscope Precipitate Crystal Images of Over dried Precipitates formed from Saturated Aqueous Solutions of NaB0 2 stabilized with 10 wt% LiOH (a) BSE2, WD 13.6mm, xlOO, 500um, (b) BSE2, WD13.6mm, x500, lOOum, (c) BSE2, WD13.6mm, xl.Ok, 50um, (d) BSE2, WD13.6mm, x3.0k, lOum, (e) BSE2, WD13.4mm, xl.Ok, 50um, (f) BSE2, WD13.5mm, x500, lOOum. 163 

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:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092686/manifest

Comment

Related Items