hydrogen oxidation at solid oxide fuel cell anodes mechanistic, kinetic and structural studies

Abstract Solid oxide fuel cells SOFCs are environmentally clean, high efficiency devices that operate at >600°C.. 1.2 Project Objectives The goals of this project were as follows: to gai

NOTE TO USERS

This reproduction is the best copy available

®

UMI

IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE

DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF CHEMISTRY CALGARY, ALBERTA DECEMBER, 2005

© Peter George Keech 2005

ivi Archives Canada

Published Heritage Branch

395 Wellington Street Ottawa ON K1A 0N4

NOTICE:

The author has granted a non-

exclusive license allowing Library

and Archives Canada to reproduce,

publish, archive, preserve, conserve,

communicate to the public by

telecommunication or on the Internet,

loan, distribute and sell theses

worldwide, for commercial or non-

commercial purposes, in microform,

paper, electronic and/or any other

formats

The author retains copyright

ownership and moral rights in

this thesis Neither the thesis

nor substantial extracts from it

may be printed or otherwise

reproduced without the author's

permission

Direction du Patrimoine de l'édition

Archives Canada

395, rue Wellington Ottawa ON K1A ON4

Your file Votre référence

le monde, a des fins commerciales ou autres, sur support microforme, papier, électronique et/ou autres formats

L'auteur conserve la propriété du droit d'auteur

et des droits moraux qui protége cette these

Ni la thése ni des extraits substantiels de celle-ci ne doivent étre imprimés ou autrement reproduits sans son autorisation

In compliance with the Canadian

Privacy Act some supporting

forms may have been removed

from this thesis

While these forms may be included

in the document page count,

their removal does not represent

any loss of content from the

thesis

Canada

Conformément a la loi canadienne sur la protection de la vie privée, quelques formulaires secondaires ont été enlevés de cette these

Bien que ces formulaires aient inclus dans la pagination,

il n'y aura aucun contenu manquant.

Abstract Solid oxide fuel cells (SOFCs) are environmentally clean, high efficiency devices that operate at >600°C Present-day SOFCs consist of Ni-based anodes, 8% yttria stabilized zirconia (YSZ) electrolytes, and manganite-based cathodes The large-scale commercialization of SOFCs is hindered by several factors, including anode and cathode performance This work has focused on the kinetics and mechanism of hydrogen oxidation and water reduction at Ni-based anodes and at Pt, for comparative purposes

The electrochemical methods used to obtain this information were developed first

at single phase Pt and Ni, and then applied later to all other electrode materials Experiments were carried out using a three-electrode half-cell at 750-950°C (typically at 800°C) in 97:3 H»:H20 Using the low-field, high-field and Allen-Hickling approximations, as well impedance spectroscopy, the exchange current density (ip ), the activation energy (130 kJ/mol), and the anodic and cathodic transfer coefficients were determined In general, the rate determining step of H2 oxidation at Ni was found to be the second electron transfer step, while water reduction is significantly slower, with the first step being the slowest

While the anodic transfer coefficient at Ni is larger than at Pt, the ig values cannot

be compared as the reactive areas remain unknown Therefore, Ni point electrodes,

pressed against a YSZ disc, were examined to correlate the ig values with the known

electrode perimeter Unfortunately, the Ni/YSZ contact region was found to be porous,

so that the true area remained unknown Efforts to use coke deposition as a means of identifying the active anode area also proved to be unsuccessful

ili

derived (SD) materials were studied Compared to ceramic grade, micron-sized NiO and YSZ, deposited using the same methods, the SD materials are very promising, being 40

to 50 times more active

A new method to establish the NiO-YSZ electrode porosity, involving monitoring the Ni?** oxide redox response in aqueous alkaline solutions was developed As long as the sweep rate was sufficiently slow, a good correlation between the measured redox charge and the amount of NiO in the composite electrodes was seen

iv

Acknowledgements

I am grateful to my advisor Dr Viola Birss for the many hours she spent with me

on this project, and for the personal growth opportunities she has given me I have learned a lot under her advisement, within our field of research and about the cooperative nature of research

Mark Cassidy (formerly of Global Thermoelectric Inc.) and Tony Wood (Versa Power Systems) were instrumental in getting this project started, with their valuable experimental instructions, and by supplying materials in the early stages

My advisory committee has contributed in many ways to this project They are

Dr J Hill (core member), Dr G Shimizu (core member initially), Dr V Thangadurai (core member at the end), Dr W Shaw (during candidacy and final examinations), Dr K Thurbide (during my candidacy examination), and finally, Dr H White of the University

of Utah, the external examiner for my final defence

I would like to thank Drs Scott Paulson and Shen Jiang Xia for their help and advice pertaining to our experimental apparatus, and for assistance in the preparation of thin film electrodes Mark Toonen, Andy Read and Jose Lopez were invaluable for their parts in the construction of the electrochemical test cell Other department members, especially Bonnie King have made this project go smoothly as well Imaging was accomplished with the help of Rick Humphrey and John McGovern of the Microscopy and Imaging Facility and Rob Marr of Geology and Geophysics at the U of C

I am very grateful to the undergraduate students whom I supervised on projects related to this work: Danielle Trifan and Sherman Kung Their work on the nickel-based

flagging Karlie Haynes also gave significant experimental help on the aqueous work

Our diverse research group has contributed in many ways to my time at this university, whether through group meeting discussions, paper editing or taking time away from the lab Harry Tsaprailis arrived when I did, and has worked alongside me the entire time I’ve been here, and our mutual support has benefited us both as we’ve gone through the Ph.D process Rudolf Potucek taught me much about Linux and networking issues, and assisted in writing scripts for data analysis Anne Co helped me with sputtering and imaging, while within our research group, Dr Jingbo Liu and Tyler Smith contributed greatly to interpretation of imaging results Joseph Fournier assisted during the early stages of our work with sols Other contributing group members include Heather Andreas, Erfan Abu Irhayem, Aislinn Sirk, Jeff Soderburg, Eric McLeod, Amit Jhas, Peyman Khalafpour, Ana Mani, Jason Young, and Sun Li, as well as Melanie Paulson Undergraduate researchers not mentioned above, that have made the lab a fun place, include Kerry Holmes and Trinh Nguyen

Alberta Ingenuity Fund and the Natural Science and Engineering Research Council (NSERC) have generously supported this project with scholarship money, along with the University of Calgary, and additional thanks go to Dr Yeong Yoo at the National Research Council (NRC) in Ottawa for supplying us with Pt paste

Finally, there are people that have made my time in Calgary fun, and are not included above: Kim Samkoe, Jude Hannaford and Nicole Hildebrand My family and especially my wife, Andrea, have supported me throughout this process, so I am grateful

to these people as well

vi

transfer coefficient

anodic transfer coefficient cathodic transfer coefficient line broadening (XRD) symmetry coefficient anodic symmetry coefficient cathodic symmetry coefficient concentration of the reactant

capacitance

potential potential with respect to time activation energy

Potential amplitude frequency

adsorbed water species

imaginary unit current density exchange current density

current

exchange current time dependent current response peak current density

current amplitude potential drop

temperature Boltzmann constant

inductance wavelength total number of electrons passed CPE exponent

CPE exponent number of electrons passed prior to the rate

vũ

adsorbed hydroxide species perimeter

number of electrons transferred during the rate determining step

gas constant

resistance resistance element in EIS fit resistance element in EIS fit resistance element in EIS fit series resistance

polarization resistance charge transfer resistance

sweep rate

surface vacancy on the electrode surface vacancy on the electrolyte

stoichiometric coefficient, the number of

times the rds occurs for one occurrence of the full reaction

angular frequency

Viii

ceramic grade ceramic grade electrode synthesis “A” ceramic grade electrode synthesis “B” constant phase element/capacitance cyclic voltammetry

diameter direct current direct methanol fuel cell energy dispersive X-ray spectroscopy electrochemical impedance spectroscopy high field (region of Butler-Volmer) low field (region of Butler-Volmer)

Lad.) Sr,.Mn03

molten carbonate fuel cell nickel-yttria stabilized zirconia nickel oxide-yttria stabilized zirconia open circuit potential

phosphoric acid fuel cell proton exchange membrane fuel cell rate determining step

reference electrode root mean square sol-derived scanning electron microscopy sol-gel

solid oxide fuel cell transmission electron microscopy triple phase boundary

ultra-high vacuum working electrode X-ray diffraction

8 mol% Y203 in ZrO»

1X

Chapter 1: IntroductiOn càng ng TH HH TH TH Hi TH TH hm 1 1.1 ProJect BackgrOUunnd - - si TH Tờ 1

ĩ9) 5 ee 2

Chapter 2: BacKgTOUnid - ‹ s9 TH TH HH T370 5 2.1 Advantages and disadvantages of fueÌ ceÌÌs ác ghe 6 2.2 Fuel Cell TYp€S LH HT TH TH HH TH HH 9 2.3 Solid Oxide Fuel Cells (SOFCS) cee 11

2.3.1 S920 TT 13

2.3.2 }9)J002(-v J0 0 14

2.3.3 N90 nh 15

2.3.4 The SOFC triple phase bournd4ry sư ry 16 2.3.5 h9)206.1 1011 6 -“- 17

2.4 Literature review ofthe hydrogen oxidation mechanism under SOFC COTỞIEIOTNS - TH TH HH HT TT TT kí 18 2.5 Electrochemical techniques utilized in this theS1S Son ssssisey 19 2.5.1 0 /9JìiÀ/J12ii› 0077 19

2.5.2 Electrochemical mpedance SpeC{TOSCODY cay 20 2.6 Principles of sol-gel cher1SẨTV - Ăn TT HH ng ngư 25 2.6.1 S DTOC€SSITE Án TH TH TT Họ HH 26 2.6.2 SG synth€S1S TOU{€S HH HH TH HH Tu nh 27 2.6.3 Forms of SƠ DrOdUCS - -Q HH HH HT ng ngu 28 Chapter 3: General Experimental Methods c1 SH ng ng nhiệt 29 3.1 Maferlals Ă SH nh ng HT TT HT gu TH 29 3.2 Physical and chemical character1zation methods cv sssvsss 30 3.2.1 X-ray diffraction charaCf€T1Z2f1OT - HH HH ng HH 30 3.2.2 Transmission electron microscopy and energy dispersive x-ray spectroscopy Characterization .:.cccccescccssccssecsscessceeseeeeeecessesseseesnecessersesesaesessees 30 3.2.3 Scanning electron microscopy analysis and energy dispersive x-ray Spectroscopy charaC†€T1ZAE1OTN ác HT TH ng 31 3.2.4 Microprobe anaÏySIS - cuc ch HH HH ng Hè 32 3.3 Electrochernical measurern€niS - «siết 32 3.3.1 Three electrode half-cells for high temperature WOrẨK c «c- 32 3.3.2 Conditions used during high temperature half cell experiments 34

S6 oi ốc hố e 37 Chapter 4: Kinetic Analysis of Hydrogen Oxidation/Water Reduction at Single Phase Pt and Ni Electrodes in SOECS LH HH nh TH no TH TH HH nh 38 4.1 Literature review on kinetic analyses of SOECs - SĂcSs re 39 4.2 Chapter Spectfic Expertnental Methods - càng ieo 40 4.2.1 Preparation of single phase eleC†TOđ€S - Án He, 40 4.2.1.1 Simgle phase platinum eÌeCfrOđ€S á- Ăn ng ưkp 40 4.2.1.2 Single phase Ni elecfrOdeS HH ng ng ưu 41 4.2.2 Pt-YSZ composite electrode preparafIOn - ác HH ng 41 4.3 Results and DiscussIon SH HH TH Kiện 42

4.3.1 Electrochemical methods for determining the kinetic and mechanistic parameters of Hạ oxidatlon/H;O reduction at SOFC electrodes - 42 4.3.2 Electrochemical testing of single phase Pt on YSZ - cc<°⁄ 48 4.3.2.1 CV results obtained during H, oxidation/H2O reduction at single phase PUYSZ mnterfaces at 8Ö °C LH HT ng HH kg 48 4.3.2.2 High field (HF) analysis of Hạ oxidation/H;O reduction at single phase PƯYSZ interfaces at S0 C - HH ng HH nh no 50 4.3.2.3 Allen-Hickling (AH) analysis ofH; oxidation/H;O reduction at single phase PU/YSZ Interfaces at 800 °C - Án HH TH ni nh kưh 57 4.3.2.4 Low Field (LF) analysis of Hạ oxidation/HaO reduction at single phase Pt/YSZ interfaces at 800°, TH HH ngờ 59 4.3.2.5 EIS analysis ofH; oxidation/HaO reduction at single phase Pt/YSZ interfaces at 8ÖÚC - cv T TH TH ng HH 62 4.3.2.6 Summary ofelectrochemical results obtained for Hạ oxidation/HạO reduction at single phase Pt/Y5Z Interfaces at 800°C .- - co sec 67 4.3.2.7 SEM analysis of single phase P† electrodes - - 2c Ặ {se 68 4.3.2.8 Electrochemical performance of single phase Pt electrodes at other LEMMPETALULES 2.0 eee eeeeeeeeeeneceeneeeeteeceaeeeseessseeesneessecenaseeseeesseesssesessseneusenessessasessess 70 4.3.2.9 Hp» oxidation/H2O reduction at ceramic grade Pt-YSZ/YSZ composite S235 74

4.3.3 Ha oxidation/H¿O reduction at single phase N¡ electrodes 76

4.3.3.1 CV results obtained during Hạ oxidation/H;O reduction at single phase

Ni electrodes at 80 °C, LH HH TH HH TT TH HT 76

4.3.3.2 — EIS results for Hạ oxidation/H;O reduction at Ni/YSZ interfaces at 800°C 83

4.3.3.3 Summary ofH; oxidation/H2O reduction at single phase Ni/YSZ

1nterfaces at 800°C, and comparison with Pt/YSZ, Interfaces ‹ -«- 85 4.3.3.4 SEM analysis of single phase Ni electrodes co 87 4.4 — COMCIUSIONS nh 88 Chapter 5: Ni anodes with controlled triple phase boundaries -. -<c<< «<< 91 5.1 Literature Review on Controlled Triple Phase Boundary SOFC Anodes 92 5.2 Chapter Specific Experimental Methods Ăn ng 94 5.2.1 Preparation and characterization of Ni point anodes for SOFC applications

94 5.2.1.1 Fabrlcation ofNI poInf anod§ óc HH ng Hiệp 94 5.2.1.2 SEM analysis ofNi point anodes and YSZ electrolytes 96 5.2.1.3 Microprobe analysis of Ni point anodes and YSZ discs 97 5.2.1.4 Electrochemical Evaluation of Activity of Ni Point Anodes in 97:3

H;:HạO 97

5.2.1.5 Regeneration of Ni Point Surfaces via Electropolishing 99 5.2.1.6 Depositlon of Carbon on Ni point anodes - «xe, 100 5.3 Results and Discussion SH HT ng Tờ 100 5.3.1 Physical characterization of Ni Points and YSZ Electrolytes 100 5.3.1.1 SEM analysis ofNI point electrodes cư 100

XI

point 102

5.3.1.3 Microprobe analysis of YSZ electrolyte surfaces after Ni point removal

105

5.3.2 Electrochemical performance of Ni point anodes in H2/3% H20

ALMOSPHETE 00 eeseesessceseecssseesceesnecsseeecseeceeeeseeseseceeeenneceeecseecsseesseereteeeneeseneeenaeenses 109

5.3.2.1 CV analysis of Hz oxidation/H20O reduction at Ni point electrodes 109

5.3.2.2 EIS analysis oFNi point anodes in 97:3 H;:H;O -+.- 112

5.3.3 Correlation between SEM determined for Ni point/YSZ contact areas/ perimeter and measured electrochemical parameters at 97:3 H2:H2O and 800°C 116

5.3.4 Coke Deposition as a Means of Establishing Ni/YSZ Contact Area and Perimeter 123 5.3.5 Analysis of Ni point anode after exposure to methane ‹› 123

5.4 CONCIUSIONS 0 130

Chapter 6: The Development of High Performance Ni-YSZ Catalysts for SOFCs 132

6.1 Literature Review of Nanoparticulate NIO — YSZ Materials 132

6.2 Chapter Spectfic Experimental Methods - - Ăn ng 134 6.2.1 Sol and sol-derived powder synthe€SIS HH, 134 6.2.1.1 Synthesis o£N¡O sol and sol-derived powders - sec 134 6.2.1.2 Synthesis of 8% YSZ sol and sol-derived poWders «- 135

6.2.1.3 Synthesis of NiO-YSZ composite sol and sol-derived powders 136

6.2.2 Electrochemical Cells .- cv Tnhh gọn HH ng cư 136 6.2.2.1 Preparation of YSZ electroÏyf€S LH ng rey 136 6.2.2.2 Application of sol-derived NiO-YSZ electrodes and other electrodes onto YSZ electrolyte SUTÍAC€S - HH ng gu ng ng Hư 136 6.2.2.3 Preparation of ceramic grade NiO-YSZ electrodes and other electrodes 137 6.2.2.4 Preparation of screen printed Ni-YSZ electrodes -.- 138

6.2.2.5 Preparation of Reference and Counter electrodes - 138

6.2.3 Physical and Chemical Characterization Methods - 139

6.2.3.1 TEM character1ZzatiOn - ch HH ng nư 139 6.2.3.2 SEM analySIS LH HH HT HH TH th ng 139 6.2.3.3 Microprobe nh 139

6.2.4 Electrochemical perforrmance testing of Ni-YSZ electrodes 140

6.2.4.1 H> oxidation/H,O reduction measurements of Ni-YSZ in half-cell 9014016151000 140

6.3 Results and DisCuSSIOn - Sàn TT HT HH TH Hà HH 140 6.3.1 Physical and chemical characterization of sol-derived NiO, YSZ and NiO- YSZ POWETS 4 140

6.3.1.1 XRD characfer1zation - SH HH HH kg kg 140 6.3.1.2 TEM examination of NiO, YSZ and NiO-YSZ powders 142

6.3.2 Electrodes fabricated using sol-derived NIO-YSŠZ” cà 146 6.3.2.1 Improving NiO-YSZ/VYSZ, Interfacial properties « 146

6.3.2.2 SEM characterization of YSZ and NiO-YSZ electrode materials 147

6.3.2.3 Microprobe analysis of sol-derived NIO-YSZ electrodes 151

xH

6.3.3 Electrochemrical Evaluation ofNI-YSZ Electrodes - «+ 154 6.3.3.1 Electrochemical performance of ceramic grade (CG) Ni- YSZ electrodes

154 6.3.3.2 Electrochemical performance of sol-derived (SD) Ni-YSZ electrodes

164

6.3.4 Comparison of the H2/H2O electrochemistry at ceramic grade vs sol-

derived NI-YSZ elecfrO@S G G9 TH ng KT 168

6.3.4.1 Reaction Mechanism Considerations .cccccccccccccccsessssssssseseessseeseseens 169 6.3.4.2 Kinetic Considerations .cccccccccccssccsessssscenscescessesessessssssssesessceseeeens 171

6.3.5 Enhancing performance of Ni- YSZ composite electrodes through screen- printing 174

6.3.5.1 Electrochemical and mechanistic results for screen-printed Ni- YSZ

electrodes in H;/H;O at §00°C 0 TQ HS HH ng ng kg xế 174

6.3.5.2 Comparison ofscreen-printed CG vs spray-coated CG (CGB)

324002) 176 6.3.5.3 Comparison of screen-printed SD vs spray-coated SD electrodes 176 6.3.5.4 Electrochemical performance screen printed ceramic grade Ni-YSZ electrode in full cell config1UIfat1OT - c- sư niey 177 6.4 ConcÌuSIOTAS - HH TT 180 Chapter 7: Aqueous testing of sol-derived NIO/Y SZZ LH HH xe, 184 7.1 Literature review on aqueous electrochemistry of sol-related NiO-YSZ 186 7.2 Chapter Specific Experimental Methods 0 ceccesssscsssseseceesseeessessesseesseees 188 7.2.1 Deposition of sol-gel NiO and NiO-YSZ on Pt -.-c c2 188 7.2.1.1 Preparation ofPt subsfraf€S -L cQ SH HH 1 ng 188 7.2.1.2 Application of NiO on Pt substrates for electrochemical evaluation 188 7.2.1.3 Application of NiO-YSZ sols on Pt substrates for electrochemical 1011500100077 3 190 7.2.2 Electrochemical evaluation of Ni and NiO-YSZ on Pt in aqueous solution

190 7.2.2.1 Electrochemical cell setup 0 ce ceceeseseesseeececeseceeesteceaeceaetenesaeenaees 190 7.3 Results and disCusSion óc HH HH HH HH ghe 191 7.3.1 Transmission electron microscopy analysis of NiO and YSZ sols 191 7.3.1.1 TEM Of NIO SOI Gà nrkn 19] 7.3.1.2 Transmission electron microscopy analysis of YSZ sol 193 7.3.2 The Deposition of sols onto Pt substraf€S ó5 àc series 195

7.3.2.1 Selection of Pt substrates for electrochemical evaluation: adhesion issues 195

7.3.3 The electrochemical behaviour of NiO and YSZ sols on Pt substrates 197

7.3.3.1 Dip-coated NiO and NIO-YSZ composIte soÌS - 22s 197 7.3.3.2 Electrochemistry of NiO deposited by Aliquot on Pt substrates 199 7.3.3.3 Electrochemistry ofNIO-YSZ deposited by aliquot - 204

“x92 o5 209 Chapter 8: Conclusions and future WOTK Ăn SH ng HH ng vu 211 8.1 Conclusions pertaining to hydrogen oxidation/water reduction at single phase

Pt and Ni anOۤ - - kh HT HH TT TT TH TT HH ng 211

XI

XIV

List of Tables

Table 2-1 Percentage Total System Emissions for Gasoline Combustion (GC), Methanol Fuel Cells (MF), and Natural Gas Fuel Cells (NGFC) Driven Passenger Vehicles 7 Table 2-2 Percentage Total System Emissions for Combined Cycle Gas Turbine

(CCGT), Solid Oxide Fuel Cell (SOFC), and Solid Oxide Fuel Cell Gas Turbine (SOFCGT) Power Generation? 0 c.ccccssssesseessesssesssessessecsssessesssestesstssessesseesstsseeaeen 8 Table 2-3 Common Fuel Cells, Electrolyte Types, Transporting Ion, and Typical

Operating Tempe€rfUT€S - - - <9 4H TH nọ TH HH HH nu nà 10 Table 4-1 Calculated Tafel Slopes (mV/decade) for Typical Transfer Coefficients 45

Table 4-2 LF, HF, AH, and EIS determined values of œ and ¡¿ for Hạ oxidation/H;O

reduction at sputtered Pt/YSZ at 800” C ch HT ng nàn 54 Table 4-3 Proposed mechanism for H2 oxidation/H2O reduction at sputtered Pt electrodes Iin60)000.0.)0)0 90h 56

Table 4-4 Values of elements from best-fit 2CPE equivalent circuit for H2 oxidation/H,O reduction at sputtered Pt at §ÖOC .- HH ng HH ng HH rên 65 Table 4-5 ig and a values for Hz oxidation/H20 reduction at Pt/YSZ interfaces as a

10910000) 19812i1-x1101 N8 71 Table 4-6 Arrhenius parameters for Hz oxidation/H2O reduction at sputtered Pt at 750 to

950°C, all calculated from the ig ValUeS cccccccccssccscccesececesccescssescessssucasevsestneescs 73 Table 4-7 LF, HF and EIS ig and o values for Hz oxidation/H2O reduction at single

Ji 280 N3 cuivvv 68.00200107 80 Table 4-8 Values of elements from best-fit 2CPE equivalent circuit at Ni- YSZ in 97:3

H;:H;O at 800°C Q0 Họ ni Họ ng gu ng HH net 85

Table 5-1 Geometry and EIS data for four Ni point anodes in 97:3 H2:H20 at 800°C 119 Table 5-2 Area / perimeter corrected EIS for four Ni point electrodes in 97:3 H2:H20 at 1000001755 119

Table 5-3 Area and Perimeter Corrected ig Values Calculated from EIS, HF and LF 122 Table 6-1 Comparison of electrochemical parameters for H, oxidation/H2O reduction at

SD, CGA and CGB Ni-YSZ electrodes (at 800°C) ccc eeesecstscesneeesreseneessseesees 169 Table 7-1 Drying temperature dependence of initial and final cathodic CV charges (uC) arising from 4.6 pmol of NiO sol on Pt foil at 100 mÝV/§ cv 201 Table 7-2 Film composition dependence of 1nitial and final cathodic CV charges (nC) arising from 4.6 pmol of NiO sol on Pt foil at 100 mÝV/§ cccccc<sererers 205 Table 7-3 Sweep rate dependence of cathodic CV charges (nC) arising from 4.6 mol of NiO sol on Pt foil at 1 to 10Ô mV/S Q1 v ng HH HH nườn 207

XV

Fig 2-1 Schematic ofa single solid oxide fuel cell (SOFFC) .-c c2 5 Fig 2-2 Schematic of triple phase boundary for hydrogen oxidation at Ni in SOFC Elements are shown without charge for sITnpÏICIfV - .- 5-5 ssssssexseeseesssss 16 Fig 2-3 Cartesian and polar (|Z|, Ð) coordinates for impedance «s« «+ 22 Fig 2-4 Equivalent circuit used to model simple electrochemical systems 23 Fig 2-5 Simulated (a) Nyquist and (b) Bode plots, corresponding to equivalent circuit

shown as Fig 2-4, using values of R; = 10 Q, R2 = 40 Q, CPE = 10 pF, andn=1

(green squares) or n = 0.9 (blue CirCÏ€S) c1 v9 1111111 1 xe 24 Fig 3-1 Schematic of 3 electrode SOFC test cell, illustrating positions and dimensions of working, counter and reference electrodes, as well as of current collectors and

Clectrical leads eee csssssssssscesceseeeeeeseesecsecescesecsessecseceeessesessesseseeessecaseneseeseeeeasaes 33 Fig 3-2 Schematic of high temperature electrochemical half celÏ -«‹ 35 Fig 3-3 Assembled high temperature electrochemical half cell with shielding wire 36

Fig 4-1 CV of H» oxidation/H2O reduction at Pt paste at 800°C and 100 mV/s at 18, 48 and 110 ml H›/mI1 - S99 0 T550 5 SE g5 6k ki 48 Fig 4-2 CV (10, 50, and 100 mV/s) ofH; oxidation/H;O reduction at sputtered Pt at L000001ẼẺẼ.8® 49 Fig 4-3 (a) Anodic and (b) cathodic Tafel plot of Hạ oxidation/H20 reduction at

sputtered Pt at 800°C and 10 mV/s, with i in MA/CM? wo ese sssesssesteseessesstessenes 51 Fig 4-4.(a) Anodic and (b) cathodic Allen-Hickling plots of H2 oxidation/H2O reduction

at sputtered Pt at 800°C and 10 mV/s with ¡ in mA/CIỂ 5-5 56 ccccccczeri 58

Fig 4-5 LF region of CV for H2 oxidation/H2O reduction at sputtered Pt at 800°C and 10

¡0 1n 60 Fig 4-6 OCP EIS response of of Hạ oxidation/HaO reduction at 800°C showing: (a) decrease in arc diameter following periods of polarization and with 1CPE EC inset, with ECs and fit data shown as insets; (b) no change in response seen over period of days at 800°C in absence of polarization ::csccsscssscssscetsceesscessscesecesscnnecneeeneeees 63 Fig 4-7 Top down SEM views of as deposited (a) sputtered Pt Anode on YSZ, and

following heat treatment at 800°C in H2/3% HạO for 10h of (b) sputtered Pt on YSZ and (c) Pt paste on SZ so HH HH 69 Fig 4-8 Arrhenius plot for Hz oxidation/H2O reduction at sputtered Pt using ig values from (top down): Anodic high field Tafel Region #1, anodic Allen-Hickling Tafel Region #1, cathodic high field, cathodic Allen-Hickling, low field, electrochemical impedance spectroscopy, anodic high field Tafel Region #2, and anodic Allen- Hickling Tafel Region #/2 - cv HH g9 HH uc Hà ng nà 72 Fig 4-9 (a) CV and (b) anodic Tafel plot (IR corrected) for H, oxidation/H,O reduction

at Pt-YSZ composite electrode at 800°C with ¿ in mA/CTrỶ - - 5c 5cccscsc5¿ 75 Fig 4-9 (c) EIS data for Hạ oxidation/H;O reduction at Pt-YSZ composite electrode at

00001 76 Fig 4-10 CV (10 mV/s) of Hz oxidation/H2O reduction at vapour deposited Ni at 800°C (a) as acquired and (b) following IR compensation .:ccccccssessecssessrseseeeseesteeees 77 Fig 4-11 Hạ oxidation/H2O reduction at vapour deposited Ni at 800°C and 10 mV/s with i in mA/cm’ showing anodic: (a) Tafel plot (b) Allen-HIckling plot 78

Xvi

Fig 4-12 Cathodic branch ofa CV (10 mV/s) for H; oxidation/H;O reduction at vapour

deposited Ni at 800°C demonstrating infÏlexion r€g1on - «cv 81 Fig 4-13 Simulated Butler-Volmer CV for reactions at 800°C described by (a) & = 1.5 (maroon), O = 0.5 (blue dash), and sum of these (black), and by o, = 0.5 (red), O& = 0.5 (blue dash), and sum of these (solid dark blue) and (b) blow-up of cathodic region of these showing predicted inflexion region (red) -ccccsssscsee 82 Fig 4-14 OCP EIS data for Hạ oxidation/HO reduction at vapour deposited Ni at 800°C t1911 01111 TT HT TT TT TH TH HH HT TT TH TT Ti HT g1 84 Fig 4-15 Top down SEM views of vapour deposited Ni (a) as deposited and (b)

following heat treatment at 800°C in H2/3% H2O for 10 Beewee eee cceeetseneceneeeeeeeaes 88 Fig 5-1 (a) Side view of Ni point electrode pressed on YSZ electrolyte, and (b) top view

of Ni point electrode, showing contact area to YSZ, and two possible (simulated) textures of the Contact TEQION ee ceceescesccsseeceseceesceeseeseescseeeceseceesecesaserseeeesreneas 95 Fig 5-2 SEM end-on view of Ni points electrode that were (a) newly machined, and (b) following heat treatment 800°C for 3 days and electrochemical testing for several

Fig 5-3 Secondary SEM electron image of end-on view of electropolished Ni point is 102 Fig 5-4 (a) Secondary and (b) Backscattered (SEM) images of YSZ discs after removal

of Ni point electrode Dark region in (a) and (b) examined for NiO and found to be

Fig 5-5 (a) Backscattered image of YSZ disc surface following cell disassembly,

showing Ni traversing the YSZ surface, likely due to the Ni point being dragged across the disc during disassernbÌy - «kg HH ng ng ng ke 105 Fig 5-6 (a) Backscattered electron image and element map for (b) Ni and on YSZ disc following cell disassembly, (Ni is white region near centre of (b)) - 107 Fig 5-7 Protocol # 1 CV (10 mV/s) of H2 oxidation/H20 reduction at Ni point electrode

at 800°C showing (a) effect of DC polarization of 300 mV (with inset shows Tafel Plot) and (b) effect of several days at 800°C and testing sec 110 Fig 5-8 (a) Protocol # 2 CV (10mV/s) of H2 oxidation/ H2O reduction at Ni point

electrode at 800°C and 10 mV/s and (b) corresponding Tafel plot of same data with J 0Ú 111 Fig 5-9 Nyquist plots for two different EIS of Ni point electrodes in (97:3) H2:H20 at 800°C showing (a) one apparent time constant, and (b) two apparent time constants, with equivalent circuits Shown As insets eeescsecenecesetereeeeceneteeeteseaeessereeeeaaee 113 Fig 5-10 SEM images of Ni point electrodes surfaces, after cell disassembly that

produced electrochemistry of Fig 5-9 revealing (a) non-porous point and (b) porous

09511017 114 Fig 5-11 (a) and (c) SEM images of end-on view of two Ni points electrode with (b) and (d) showing estimated contact areas (whife T€BØ1OTIS) - án nghiệt 117 Fig 5-12 SEM images showing (a) higher perimeter and (b) lower perimeters of Ni on YSZ, but with approximately equivalent N/YSZ conftact areas - ‹«- 120 Fig 5-13 (a) Secondary image, (b) backscattered image, (c) carbon element map, and (d)

Ni map of end of Ni point electrode that was attached to YSZ and exposed to

methane at 800°C Warmer colours indicate higher element concentrations 125

XVil

was attached to YSZ and then exposed to methane at 800°C Images show an

4900101318081) 000001007877 126 Fig 5-15 (a) and (b) Two possible contact region after exposure to methane at 800°C, X28 ÄJ[ i10) 0ui n6 128 Fig 5-16 (a) and (b) Two possible contact regions for methane exposed Ni point at 800°C, same electrode as in Fig Š- Í4 Ăn HH HH Hiệp 129 Fig 6-1 XRD pattern for (a) NiO, (b) 8% YSZ, and (c) 1:1 NiO:8% YSZ sol-derived powders, with data standard (Jade, 6.5), shown below each set 142 Fig 6-2 (a) TEM image of NiO powder on C-coated Cu grid, and (b) Histogram of NiO particle sizes (diameters) cccsesssessccesessscesseesecssessecseeseesssasssesseseaseesecssesseseesans 143 Fig 6-3 TEM image of 8% YSZ SD powder on C-coated Cu grid - ‹ 144 Fig 6-4 (a) TEM image of 1:1 NiO: 8% YSZ sol-derived powder and EDX data for (b) Spot 1 (NiO) and (c) Spot 2, (8% YSZ) particles ccccscssscessecesteceseeceteeesseeeees 145 Fig 6-5 SEM Image of YSZ, anchoring layer on YSZ electrolyte disc 147 Fig 6-6 SEM analysis of ceramic grade NiO-YSZ produced using synthesis, method

“CGA” showing (a) YSZ-rich region with embedded NiO particles and (b) NiO-rich T€Đ]OHI HH HT TH HH TH TT HT TT HT TH H0 HH ch 148 Fig 6-7 SEM (backscattered) images of NiO-YSZ electrode on YSZ electrolyte disc (a) low resolution top-down view, (b) cross-section, (c) top-down view showing large agglomerates within electrode, and (d) low resolution top-down view, showing agglomerates on electrolyte and within electrOde - 5 St ssxrerres 150 Fig 6-8 Top-down microprobe analysis of sol-derived NiO-YSZ anode showing (a) backscattered image, and (b-e) element maps for (b) Ni, (c) Zr, (d) Y and (e) O Warmer colours indicate higher element concerifraf1OnS 5 55s ssssss 152 Fig 6-9 Cross-sectional microprobe analysis of sol-derived NiO-YSZ functional anode showing (a) backscattered image, and (b-c) element maps for (b) Ni, (c) Zr Warmer colours indicate higher element concentrations cccssesssseeseetesssctseereeseeesesees 153 Fig 6-10 (a) IR uncompensated CV of H2 oxidation at CG Ni-YSZ composite prepared

by synthesis method “CGA” showing high series resistance at sweep rate of 10mV/s and at 800°C, (b) Tafel plot for same data illustrating effect of high series resistance With 7 in MA/CI eeeeccsssecscsessesssssessesscsrcsscarssesssssesseseesussecasssesaesussucaucaesateaseueteeeees 155 Fig 6-11 (a) Anodic Tafel plot derived from IR compensated CV for H2 oxidation/H2O

90100 157 Fig 6-12 (a) IR uncompensated CV of H2 oxidation/H20 reduction at CG Ni-YSZ

composite prepared by synthesis method “CGB” showing Butler-Volmer kinetics at

10 mV/s and at 800°C eee ceeeseencssscteceeseseeseesecseseesescesesssesecsusesesssesessesesseaseaseneseees 160 Fig 6-13 (a) IR uncompensated CV of H2 oxidation/H20 reduction at SD Ni-YSZ composite prepared by spray-coating method showing Butler-Volmer kinetics at 800°C and at 10 MV/S 164 Fig 6-14 Anodic Allen-Hickling plot calculated from CV of H2 oxidation at CG Ni-YSZ composite using Value n/v = 1.25 instead of n/v = 2 showing distortion to Tafel region when compared to Flg Ó-]2C - c2 1v E111 81118111 11111111 xee 172

XVII

Fig 6-15 IR corrected CV of H2 oxidation at Ni-YSZ composite prepared by screen- printing on roughened YSZ discs for (a) ceramic grade Ni- YSZ anode and (b) sol- derived Ni-YSZ anode 0.0 ccccecsesccesesseeseeseeseeseesessececseessececsecesevaeenessecneesronseeesenees 175 Fig 6-16 (a) CV of H> oxidation at a screen printed Ni- YSZ composite anode, on

COMME CIAal YSZ oo eescscsseessesseceseseeseccsseeeseceseeseceeesseeceeeeesseeetaeseseecseessneeseaeeenatenas 178 Fig 7-1 Schematic of NiO-YSZ composite on Pt substrate, illustrating a non-active structure (on left), arising from lack of electrolyte or independent electrical

connections between NiO solution or Pt, and an active structure (on right) with good CONMUCHION PAthWAYS cccsccesescesccssscessetscssecsssesseessceseeeneseaseseeteesaesesseeeeeseaesnees 185 Fig 7-2 TEM image of 3 mM NiO sol, aliquoted onto a C-coated Cu grid, and air dried ẨOT S€V€FA] Ủ SG HH HT HT HH TT TH TH HH 192 Fig 7-3 TEM Images of 9 mM (a) YSZ gel and (b) YSZ sol, both aliquoted onto C- coated Cu grids and dried in air for several he oo eee eeeecesesscenecseeseeeeeneeeaeeneeaeens 193 Fig 7-4 CV of Pt foil substrate in 1 M NaOH before NiO sol application (squares) and after cleaning with H2SO, (circles), showing hydrogen adsorption /desorption peaks used for real area and roughness calculÏafIOTnS - -ó- 5 ngay 195 Fig 7-5 CV (20 mV/s) showing Ni(OH)2 S NIOOH oxidation/reduction for dip-coated NiO sol (with Triton), in 1 M NaOH, formed by withdrawal from sol 4 cm/s, and dried at 200°C for l5 mmim -.- - << 11 119v nh 197 Fig 7-6 CV (2mV/s) showing Ni(OH)2 S NiOOH oxidation/reduction reaction of dip- coated NiO-YSZ sol (with Mazawet), in 1 M NaOH for withdrawal rate at 4 cm/s, and dried at temperature of 350°C for 1Š TmIT - ó5 «cv cư 198 Fig 7-7 CV (100mV/s) showing an Ni(OH)2 S NiOOH oxidation/reduction for 15 pL aliquot of NiO sol , showing effect of drying temperature for 100°C, 200°C, 500°C ANd 800°C for lŠ mIm - 0 5 2s s13 99 1g TH TT Hà Hư 200 Fig 7-8 Dependence of the cathodic peak current on the sweep rate for a 15yL aliquot of

Ni Sol on Pt foil, dried at 100°C for 15 Min ccc cccccccssssceesssecessseceeseeeceeecessaes 203 Fig 7-9 CVs showing an NiOOH > Ni(OH), reaction for 20:80 NiO: YSZ dried at 200°C sols, for 15 min, showing effect o sweep rafe co seo 206 Fig 7-10 Dependence of the cathodic peak current on the sweep rate for a 30:70

NiO: YSZ film on Pt foil, dried at 200°C for 15 min c- 5 5S 55123 s++ 208

XIX

Solid oxide fuel cells (SOFCs) have a unique design that makes them attractive for stationary (e.g., residential) electricity generation Operating at temperatures of 600- 1000°C, SOFCs offer the opportunity to concurrently generate electricity and high- density heat; this co-generation permits the use of nearly all of the chemical energy, although conversion of conventional fuels to fuel cell compatible fuels can decrease efficiency In addition, the high temperature of SOFCs eliminates the need for external reforming of natural gas, which can be supplied to homes or other buildings as a fuel through existing infrastructure’.

2 This project was related to a collaboration between our group and Versa Power Systems (formerly Global Thermoelectric Inc.) in Calgary and the Alberta Energy Research Institute (AERI) through the Coordination of University Research for Synergy and Effectiveness (COURSE) program Goals at the project outset involved the development of reliable electrochemical methods for the evaluation of the performance of individual electrodes in SOFCs Subsequently, it has developed into a fundamental study

of anodes, at which hydrogen is oxidized to protons, and has also led to the development

of new anode materials for SOFCs

The performance of fuel cell systems is often measured using two-electrode electrochemical techniques, from which the overall cell performance is assessed The individual effects from the anode, cathode and electrolyte have to be subsequently

extracted from a combination of data In the present work, a three electrode, half-cell

was employed, which allowed the analysis of just the anode data, separate from the rest

of the cell

1.2 Project Objectives

The goals of this project were as follows: to gain an understanding of the mechanism and kinetics of the anode reaction under typical SOFC operating conditions, thereby contributing to the overall development of SOFCs for power sources; and to synthesize novel Ni catalyst structures using sol-gel chemistry and to evaluate them as SOFC anode materials For these purposes, the specific focus of the research presented

in this thesis has been:

1.3

To establish reliable methods for measuring the reaction kinetics and mechanism

of the hydrogen oxidation reaction, and the reverse reaction (water reduction), at primarily single phase Ni- and Pt-based anodes on yttria stabilized zirconia (YSZ) electrolytes

To utilize sol-gel chemistry to produce small particle NiO-YSZ composite materials as possible high performance Ni- YSZ anodes in SOFCs and to apply the electrochemical analysis methods established using single phase Ni and Pt to these anodes

To restrict and quantitatively establish the triple phase boundary length, where the electrochemistry occurs, and to correlate it with the observed electrochemistry

To establish the porosity and stability of sol-gel derived NiO-YSZ materials using aqueous electrochemistry approaches

Thesis Organisation

This thesis contains eight chapters Chapter 1 briefly introduces the thesis topic and overviews the research objective outline Chapter 2 contains background information, particularly literature pertaining to the fields relevant to this thesis The general experimental methods utilized for the generation of data presented in this thesis are described in Chapter 3, while experimental details specific to Chapters 4-7 can be found at the beginnings of each of these chapters The latter portion of Chapter 4 describes the results of the mechanistic analysis of hydrogen oxidation/water reduction at thin film, single phase Ni and Pt electrodes Chapter 5 describes efforts to restrict the active region of the anode, so that electrochemical activity could be correlated with the

4 triple phase boundary length The electrochemical analysis methods developed in Chapter 4 for single phase metallic anodes were then applied to H»2 oxidation at composite Ni-YSZ anodes in Chapter 6, including to new anode nanomaterials synthesized through sol-gel based chemistry Chapter 7 documents the continued exploration of the sol-gel based materials; in this work, the materials are tested in an aqueous environment to assess the accessibility of Ni within the Ni-YSZ structure to external reactants and to provide a measure of the internal porosity In Chapter 8, the conclusions reached in this thesis work are presented, and recommendations for future

work in this area are made

A fuel cell is a device that electrochemically converts chemical energy to

electrical energy It consists of three core components, an anode, where fuel is oxidized;

a cathode, where oxygen is reduced; and an electrolyte, which separates the two electrodes, allows charge balance through the passage of ions, and prevents the cell from electrically shorting Fig 2-1 shows these components for a solid oxide fuel cell (SOFC)

Fuel

H, /

Fig 2-1 Schematic ofa single solid oxide fuel cell (SOFC)

Fuel cells trace back to the 19" century, having been made first by Sir William Grove in 1839° to run on hydrogen, but their development since then has been plagued with problems They have much in common with batteries, as both are used to convert chemical to electrical energy, and the core components serve the same purpose in each device However, unlike batteries, fuel cells do not require recharging, as the reactants (fuel and oxygen) are continually replenished, and as the products (electricity, heat, unreacted fuels, and reactant products) are removed on an ongoing basis Consequently,

6 the time consuming process of recharging required for batteries is eliminated for fuel cells In addition, fuel cells are easily stacked, creating a system of high energy density, when compared to battery technologies

The work described in this thesis is directed towards solid oxide fuel cells (SOFCs) and the focus is primarily on the anode reaction As such, this chapter summarizes the advantages and disadvantages of fuel cells, specifically SOFCs Particular attention is paid to the state of the anodes in SOFCs, as well as the kinetics and mechanisms of the reactions relevant to SOFC anodes The electrochemical techniques used in this investigation, such as cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS), are also described in brief

2.1 Advantages and disadvantages of fuel cells

Two key driving factors for the development of fuel cells are their high energy conversion capability and their low environmental impact For fuel cells operating on hydrogen, water is the only chemical product, making them true zero emission energy devices Fuel cells that operate on carbon fuels produce the greenhouse gas, carbon dioxide, but their high efficiency means that this occurs to a lower extent than in combustion devices, where Carnot limitations’ significantly reduce efficiency In

addition, the emission of the pollutant NO,, which occurs through the combination of Na with O2 during conventional fuel combustion, is much lower in the case of fuel cells Particulate matter, SO, and CO are also much lower in the emissions from fuel cells

Table 2-1 compares the total system emissions (well to wheels) for equivalent mass cars

fuelled by gasoline combustion (GC), a methanol fuel cell (MFC); and a natural gas fuel cell (NGFC) The emissions are normalized to that from the GC vehicle

Table 2-1 Percentage Total System Emissions for Gasoline Combustion (GC), Methanol Fuel Cells (MF), and Natural Gas Fuel Cells (NGFC) Driven Passenger Vehicles°®

~50%° For these systems, the overall efficiency of energy conversion is enhanced through the use of “co-generation”, where usable heat energy is also harnessed and used

to drive a steam turbine as a by-product of the steam producing (and primary turbine driving) combustion reaction Using a similar strategy of usable heat recovery, fuel cell efficiency is projected to be as high as 90%’, although this level has yet to be achieved Table 2-2 shows a comparison between emission estimates for CCGT generation with

8 those for SOFC and SOFC-gas turbine (SOFCGT) generation; the data are normalized to

the emissions from the CCGT®

Table 2-2 Percentage Total System Emissions for Combined Cycle Gas Turbine (CCGT), Solid Oxide Fuel Cell (SOFC), and Solid Oxide Fuel Cell Gas Turbine

(SOFCGT) Power Generation®

However, fuel cells are not yet widely commercially available, for several important reasons There are some buses and cars which operate on fuel cells, but their cost is prohibitive, especially compared with diesel counterparts’ Only prototype cellular phones or laptop computers have been developed’®, and widespread commercialization has not been achieved Besides cost concerns, many other issues must

be overcome for the continued development of fuel cells, including: durability issues, such as thermal cycling of high temperature fuel cells or catalyst deactivation; catalyst poisoning by fuel impurities or reaction products; material incompatibility within fuel cells or fuel cell stacks; limitations in infrastructure for the delivery of fuels suitable for many fuel cells; as well as the acceptance of their integration into the existing marketplace as safe energy devices Some of the technological barriers remain poorly understood, relegating many fuel cells to the development stage

2.2 Fuel Cell Types

As anode and cathode reactions are often common between different types of fuel cells, their classification is generally done according to the nature of the electrolyte Some common fuel cells are the proton exchange membrane (PEMFCs) and the related direct methanol fuel cells (DMFCs); alkaline fuel cells (AFCs); phosphoric acid fuel cells (PAFCs); molten carbonate fuel cells (MCFCs); and solid oxide fuel cells (SOFCs) These fuel cells, as well as their electrolyte/charge carriers and typical operating temperatures, can be found in Table 2-3

The operating temperature of a fuel cell is also defined by the electrolyte type For example, for PEMFCs, the upper temperature is limited by the breakdown of the hydrated conducting polymer, Nafion™, while MCFCs and SOFCs require a high temperature to achieve a suitable level of conductivity of the (ionic) charge carriers through the respective electrolytes

10 Table 2-3 Common Fuel Cells, Electrolyte Types, Transporting Ion, and Typical

Operating Temperatures!

PEMEC / DMEC NafionTM H 60-80

2.3 Solid Oxide Fuel Cells (SOFCs)

SOFCs require elevated temperatures (600 to 1000°C) to ensure suitable conductivity of the charged species through the solid electrolyte, which typically is composed of 3-8% yttria stabilized zirconia (YSZ) This high temperature also promotes the reaction kinetics at both anode and cathode and can allow for the direct oxidation of carbonaceous fuel!* Owing to the large amounts of heat required for and generated by SOFCs, their primary use is likely to be for the generation of power at stationary locations, where heat can be managed most efficiently Two types of SOFC are typically seen, planar and tubular; both types are described below

Planar SOFCs can be classified according to the component material that gives the FC strength In general, the classifications includes electrolyte-, anode-, cathode- or interconnect-supported When the support is the anode, cathode or electrolyte, the supporting component is commonly tape-casted and the remaining materials are screen- printed onto the support For electrolyte-supported SOFCs, the ohmic resistance can be quite high, so a great deal of research is focused on anode-or cathode-supported cells

Some of the ongoing problems associated with planar SOFCs include the cost of the materials involved with cell construction, as well as the oxidation of the interconnect materials, especially on the cathode side of the cell Performance degradation occurs as a result of thermal cycling, high current operation, fuel impurities, coking by the fuel, and catalyst poisoning from system components In addition, gas sealants, which are required

to keep the fuel separate from air, have ongoing issues of stability, especially when SOFCs undergo thermal cycling

Tubular SOFCs, made by Siemens-Westinghouse and which are cathode- supported, are synthesized using a cathode tube support (ca 1.4 mm thick), upon which the YSZ electrolyte is deposited as a dense, thin film (40 pm) The Ni-YSZ anode is subsequently added as a slurry, prior to the cells being fired at high temperatures (> 1000°) The Acumentrics tubular SOFCs, on the other hand, are anode-supported In this case, the electrolyte is added as a slurry to the anode tube support prior to the initial firing, which densifies the YSZ The cathode layer is subsequently added and the cell undergoes a second firing

As this project was focussed on the performance of anode materials, the mechanisms used in practice to separate the gas streams are not relevant to the experimental parameters we employed in this work Here, the electrochemical experiments were performed using a three-electrode “half-cell’”’ configuration (Fig 3-1) This configuration negates the need for the passage of different gases over each electrode,

as each electrode is under a common atmosphere in this configuration Chapter 3

describes the full details of the half-cell, which differs from the conventional

two-electrode “full cell” methodology (Fig 2-1), where leads are attached to the anode and cathode and the overall fuel cell performance is measured

2.3.1 SOFC cathodes

One difficulty that arises with respect to SOFC cathode materials is that the nature

of the atmosphere, e.g., hot air/oxygen, can be very harsh on many electrochemically useful materials, particularly metals Virtually all non-noble metals oxidize rapidly in this atmosphere, eliminating most of them as oxygen reduction catalysts However, Ag has a relatively low melting point, 962°C'*, and Au is not well-known for its oxygen reduction electrochemistry Of the noble metals, Pt has been explored with the most success'*!’; however, Pt has proven not to be stable during O reduction, giving inductive features in electrochemical measurements In addition, the cost of Pt is prohibitive for use as commercial SOFC cathodes

Most SOFC cathodes are based on a conductive ceramic, which overcomes the

problem of the oxidizing atmosphere Substituted LaMnQ;3 is the most common cathode material, a compound that is electronically conductive and also ionically conductive to

O at high temperatures The LaMnO; thermal expansion coefficient (11.2 x 10° cm/cmK)!® matches that of YSZ (10.8 x 10° cm/cmK above 600°C)'” reasonably well, which minimizes the impact of internal forces during heating The conductivity of LaMn0O; can be enhanced by partial substitution of either divalent or monovalent cations

in place of La** The most common substitution is Sr°*, producing lanthanum strontium manganite (LSM)

14 LSM based cathodes containing various noble metal dopants are also of recent interest, and have been characterized by Haanappel, ef al*° While Pd, Pt and Ag additives seemed to show no positive influence on the cathode electrochemistry, Pd on activated carbon proved to enhance the oxygen reduction reaction rate significantly, particularly at lower operating temperatures

2.3.2 SOFC electrolytes

SOFC electrolytes must be ionically conductive, electrically insulating, non- porous to gases, and stable at high temperatures for long periods of time The most

common electrolyte material is ZrO, doped with 8 mole % Y203 (8YSZ, or YSZ in this

thesis) As mentioned previously, the charge carrier in SOFCs is O*; however, the ionic conductivity of YSZ is only 0.1 @''em'” at 1000°C and much lower at the typical SOFC operating temperatures of 700 to 800°C The relatively low ionic conductivity of YSZ is partially overcome by making it very thin, typically only a few pm, within an electrode- supported cell It is the other characteristics of YSZ, specifically its durability and very high electronic resistance, which make it the primary option for SOFCs at this time

Specifically, YSZ does not react with conventional anode or cathode materials, even after

very many hours at high temperatures

In order to operate SOFCs at intermediate temperatures (500 to 700°C), there has been some recent interest in the development of electrolytes with enhanced ionic conductivity Rare earth oxide electrolytes, such as doped ceria”', have received significant attention, but scandia-stabilized zirconia and lanthanum strontium magnesium gallate have also been investigated for this purpose’ Each material has a higher

conductivity than YSZ, but has not yet demonstrated enough stability to be considered a viable replacement for YSZ in SOFC technology

2.3.3 SOFC Anodes

Ni is typically the anode catalyst used for Hz oxidation in SOFCs It has a high electrochemical activity, is relatively inexpensive, and when mixed with YSZ in a 1:1 mole ratio, it has a thermal expansion coefficient which is comparable to that of YSZ

(12.5 x 10° cm/cmK vs 10.8 x 10° cm/emK above 600°C) Currently, it is the material

of choice for Siemens-Westinghouse, Acumentrics, and Versa Power Systems

However, the oxidation of Ni can cause volumetric increases of up to 30%, a

problem that may arise during a failure of a gas seal or the rapid shutdown of an SOFC stack Rapid volumetric expanses are likely to cause fractures within the SOFC cell or detachment of the Ni anode In addition, Ni has a low tolerance for sulphur within fuel feeds, so it must be removed prior to fuel introduction into SOFCs Another ongoing issue with Ni-based anodes is their inability to directly oxidize carbonaceous fuels, such

as natural gas, without pre-reforming them If these fuels are added directly to Ni-based anodes, they undergo cracking, leaving residual carbon deposits on the Ni-based catalysts’* and causing an irreversible drop in cell performance” Consequently, most fuel feed stocks for SOFC anodes undergo a pre-reforming step to convert the

carbonaceous fuels to H2, in addition to CO2 and CO, and minimize carbon deposition?

To avoid the difficulty of pre-reforming, substantial effort has been directed at developing catalysts that do not have these carbon deposition issues Some of the most

14,26 successful use Cu-doped CeO anodes”, at which methane undergoes direct

16 electrochemical oxidation and does not go through an internal reforming process In related work’’, the authors have demonstrated that Cu-CeO anodes can electrochemically

oxidize either H, or CHg, and that H oxidation does not diminish following exposure to

CHg, unlike Ni-YSZ, which rapidly undergoes irreversible deactivation upon exposure to

CHạ Within this study’’, it is notable that the Ni-YSZ anode began with a significantly

higher activity than the Cu-CeO anode; this may partially indicate why Cu-CeO anodes have not developed more rapidly

2.3.4 The SOFC triple phase boundary

The electrochemically active region of an SOFC is often referred to as a triple phase boundary (TPB), as fuel, electrolyte and electrode all must be present for the

electrochemical reaction to proceed The TPB is discussed extensively throughout

Chapters 4-7, while Fig 2-2 pictorially demonstrates the reaction site (TPB) for Hz

oxidation at Ni

eH

@0o

e @ Zr O?- from cathode reaction -——_+ @ — ° Y

@Ni

YSZ Ni Fig 2-2 Schematic of triple phase boundary for hydrogen oxidation at Ni in SOFC Elements are shown without charge for simplicity

As Fig 2-2 indicates, Hz adsorbs to the electrocatalyst, where it can be oxidized only if it

is in close proximity to the electrolyte, which supplies the counter ion, O°

2.3.5 SOFC fuels

As alluded to in Section 2.3.3, there are several fuels currently being used for

28,29 , carbon monoxide’, syngas”, ca 30 31 SOFCs These include hydrogen, natural gas”°, butane

and coal gas** Existing infrastructure for its delivery, as well as its relative abundance, make natural gas an attractive fuel for SOFCs However, as mentioned in Section 2.3.3, when carbonaceous fuels such as methane, butane or coal gas are supplied directly to Ni anodes, hydrocarbon cracking occurs, and carbon fibres are formed within the SOFC

anode Following this, Ni anodes deactivate, and in some cases, the SOFCs are damaged Cu-CeO anodes have shown a resistance to carbon formation, but the high activity of Ni for H> oxidation continues to attract most SOFC anode researchers

Consequently, significant efforts have been made to efficiently reform carbonaceous fuels to make them more suitable for Ni-based SOFC anodes One such method is steam reforming, combined with the water-gas shift reaction®®, which produces

H, and CO, from H20 and CH, Steam reforming is generally performed on supported metal catalysts, such as NẺ, or Ni-Cu”” and temperatures typically exceed 500°C However, this reaction is very endothermic, requires low pressure (i.e large reactor volumes and constant removal of products), and a high H,O content must be maintained

to prevent carbon deposition Partial oxidation of CHs produces CO and H), and minimizes carbon deposition, but uses chemical energy that could otherwise be used by the fuel cell, so this solution is not ideal either

Regardless of the specific method for making H2 from hydrocarbons, a mixture of gases is produced from any reforming process This mixture generally includes H2, CO:,

CO, H20; in many cases, the fuel contaminant HS, and unreacted CH, also emerge from

to test them within mixed gas streams However, within the scope of this thesis, H2/H2O

was selected as the fuel gas, to establish the activity of these new catalysts in the absence

of C or S problems H>/H2O was also used in the early portion of this work, to establish kinetic and mechanism parameters for Ni- and Pt- based electrodes

2.4 Literature review of the hydrogen oxidation mechanism under SOFC conditions

Much of the literature pertaining to the hydrogen oxidation mechanism at SOFC anodes is summarized in Chapter 4, but a brief overview is given here There have been several attempts to study hydrogen oxidation at Pt, and especially at Ni-based anodes, under SOFC conditions Using either electrochemical impedance spectroscopy (EIS)*° or high overpotential measurements*®, investigations have been performed for Ni-YSZ composite anodes, including for anode-supported cells’’ In many cases, the anode microstructure was found to affect the results obtained, so some researchers have focused

on pure Ni anodes (termed “single phase” in this thesis), which were press-contacted against YSZ electrolytes Once again, both high overpotential***? and EIs® measurements have been used to identify the reaction mechanism However, results have proven to be somewhat contradictory, as the electrochemical characteristics have tended

to vary between research groups In some cases, the reaction is presumed to be limited

by gas diffusion®®, while sometimes it has been determined that adsorption”® or electron transfer?’ determine anode kinetics It is partly because of these discrepancies in the literature that we directed our work towards establishing reliable electrochemical methods to identify the rate determining step in the H» oxidation reaction at Ni and Pt anodes, as described in Chapter 4

2.5 Electrochemical techniques utilized in this thesis

As mentioned in Section 2.3, the electrochemical experiments performed in this thesis utilized a three-electrode set-up, where the kinetics and performance of only the working electrode (WE) are established In these experiments, a potentiostat is required, which controls the potential of the WE against a reference electrode (RE) Current that flows at the WE is matched in magnitude at the counter electrode (CE), which balances the charge within the electrochemical cell In this work, the CE was placed directly opposite to the WE, on the YSZ electrolyte, while the RE was located on the same side as the CE, several mm away, as described in Chapter 3

2.5.1 Cyclic voltammetry

In cyclic voltammetry (CV), the potential between the WE and RE is linearly

increased with time and then decreased, at specific sweep rates, and the current which

flows through the WE is measured The current is then plotted as a function of the WE potential, which allows multiple cycles to overlay on the same plot This technique is a direct current (DC) analysis method