<%BANNER%>

Development of a High-Performance Composite Cathode for LT-SOFC

Permanent Link: http://ufdc.ufl.edu/UFE0042563/00001

Material Information

Title: Development of a High-Performance Composite Cathode for LT-SOFC
Physical Description: 1 online resource (119 p.)
Language: english
Creator: Lee, Byung
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: bismuthruthenate, cathode, glycine, infiltration, pyrochlore, sofc
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Solid Oxide Fuel Cell (SOFC) has drawn considerable attention for decades due to its high efficiency and low pollution, which is made possible since chemical energy is directly converted to electrical energy through the system without combustion. However, successful commercialization of SOFC has been delayed due to its high production cost mainly related with using high cost of interconnecting materials and the other structural components required for high temperature operation. This is the reason that intermediate (IT) or low temperature (LT)-SOFC operating at 600~800degreeC or 650degreeC and below, respectively, is of particular significance because it allows the wider selection of cheaper materials such as stainless steel for interconnects and the other structural components. Also, extended lifetime and system reliability are expected due to less thermal stress through the system with reduced temperature. More rapid start-up/shut-down procedure is another advantage of lowering the operating temperatures. As a result, commercialization of SOFC will be more viable. However, there exists performance drop with reduced operating temperature due to increased polarization resistances from the electrode electrochemical reactions and decreased electrolyte conductivity. Since ohmic polarization of the electrolyte can be significantly reduced with state-of-the art thin film technology and cathode polarization has more drastic effect on total SOFC electrochemical performance than anode polarization as temperature decreases, development of the cathode with high performance operating at IT or LT range is thus essential. On the other hand, chemical stability of the cathode and its chemical compatibility with the electrolyte should also be considered for cathode development since instability and incompatibility of the cathode will also cause substantial performance loss. Based on requirements of the cathode mentioned above, in this study, several chemico-physical approaches were carried out to develop a high-performance composite cathode, in particular, for LT-SOFC operating 650degreeC and below since stability and compatibility of the materials in interest are secured at low temperatures. First, a nano-sized pyrochlore bismuth ruthenate (Bi2Ru2O7 or BRO7 shortly), one of the promising cathode materials, was successfully synthesized using glycine-nitrate combustion (GNC) route. Stoichiometric Bi2Ru2O7 without any impurity phase was achieved with considerably improved processing condition, leading to the crystallite size of ~24nm in diameter. Even though the resulting powder tends to agglomerate, resulting in overall 200~400nm size range, it still showed better quality than the one prepared by solid state (SS) reaction route followed by extra milling steps such as vibro-milling and sonication for further particle size reduction. Glycine-to-nitrate (G/N) ratio was found to play a critical role in determining the reaction temperature and reaction duration, thus phase purity and particle morphology (particle size, shape, and agglomeration etc). Composite cathodes of such prepared BRO7 (GNC BRO7) combined with SS erbia-stabilized bismuth oxide, Bi1.6Er0.4O3 or ESB, showed better electrochemical performance than vibro-milled BRO7 (VM BRO7)-SS ESB. ASR values of 0.123?cm2 at 700degreeC and 4.59?cm2 at 500degreeC, respectively, were achieved, which follows well the trend of particle size effect on performance of composite cathodes. Additionally, the number of processing steps (thus time) was reduced by GNC route. Several issues in regard to synthesis process and characteristics of BRO7 material itself will be addressed in this dissertation. Secondly, a unique in-situ composite cathode synthesis was successfully developed and applied for BRO7-ESB composite cathodes to improve percolation and to reduce agglomeration of each phase inside the cathode so that the effective triple phase boundary (TPB) length was extended. To disperse and stabilize ESB powder in de-ionized (DI) water, zeta potential profile of ESB powder in DI water as a function of pH was first achieved. The effect of a dispersant (ammonium citrate dibasic) on the stability of ESB powder dispersed in DI water was also investigated. Knowledge of BRO7 wet chemical synthesis from previous study was utilized for final product of in-situ BRO7-ESB composite cathodes. Such prepared composite particles were characterized and the electrochemical performance of in-situ BRO7-ESB composite cathodes was examined as well. Performance enhancement was observed so that ASR values of 0.097?cm2 and 3.58?cm2 were achieved at 700degreeC and 500degreeC, respectively, which were 19% and 22% improvement, respectively compared to those of conventionally mixed composite cathodes of BRO7-ESB. Finally, a highly controlled nanostructured BRO7-ESB composite cathode was developed by infiltration of BRO7 onto ESB scaffolds to maximize the effective TPB length, to improve the connectivity of ESB phase inside the cathode for better oxygen-ion diffusion, and to minimize delamination between the electrolyte and cathode layers. ESB scaffolds were first established by adding a graphite pore-former and controlling heat treatment condition. Nano-sized BRO7 particles were successfully created on the surface of previously formed ESB scaffold by infiltration of concentrated (Bi, Ru) nitrate solution followed by the optimized heat treatment. Such prepared composite cathodes exhibited superior electrochemical performance to conventionally made BRO7-ESB composite cathodes and even better than GNC BRO7-SS ESB developed in this dissertation, e.g. 0.073?cm2 at 700degreeC and 1.82?cm2 at 500degreeC, respectively. This cathode system was revealed to be highly competitive among all the reported composite cathodes consisting of the same or different materials prepared by various processing techniques. It was demonstrated that the extended TPB length from continuous network of BRO7 nanoparticles and better connectivity of ESB scaffolds enabled the outstanding performance. Moreover, de-lamination of cathode from the electrolyte was prevented thanks to improved adhesion between ESB scaffolds and ESB electrolyte. Dissociative adsorption of oxygen gas were proposed to be the dominant rate-determining process for the overall oxygen reduction reaction at low temperatures (500-600degreeC) whereas all of the constituting sub-reactions such as oxygen gas dissociative adsorption, oxygen ion diffusion towards TPB region, and oxygen ion incorporation were found to play roles competitively in the overall reaction at relatively high operating temperature (650-700degreeC) based on analysis of impedance spectra.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Byung Lee.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Wachsman, Eric D.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042563:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042563/00001

Material Information

Title: Development of a High-Performance Composite Cathode for LT-SOFC
Physical Description: 1 online resource (119 p.)
Language: english
Creator: Lee, Byung
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: bismuthruthenate, cathode, glycine, infiltration, pyrochlore, sofc
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Solid Oxide Fuel Cell (SOFC) has drawn considerable attention for decades due to its high efficiency and low pollution, which is made possible since chemical energy is directly converted to electrical energy through the system without combustion. However, successful commercialization of SOFC has been delayed due to its high production cost mainly related with using high cost of interconnecting materials and the other structural components required for high temperature operation. This is the reason that intermediate (IT) or low temperature (LT)-SOFC operating at 600~800degreeC or 650degreeC and below, respectively, is of particular significance because it allows the wider selection of cheaper materials such as stainless steel for interconnects and the other structural components. Also, extended lifetime and system reliability are expected due to less thermal stress through the system with reduced temperature. More rapid start-up/shut-down procedure is another advantage of lowering the operating temperatures. As a result, commercialization of SOFC will be more viable. However, there exists performance drop with reduced operating temperature due to increased polarization resistances from the electrode electrochemical reactions and decreased electrolyte conductivity. Since ohmic polarization of the electrolyte can be significantly reduced with state-of-the art thin film technology and cathode polarization has more drastic effect on total SOFC electrochemical performance than anode polarization as temperature decreases, development of the cathode with high performance operating at IT or LT range is thus essential. On the other hand, chemical stability of the cathode and its chemical compatibility with the electrolyte should also be considered for cathode development since instability and incompatibility of the cathode will also cause substantial performance loss. Based on requirements of the cathode mentioned above, in this study, several chemico-physical approaches were carried out to develop a high-performance composite cathode, in particular, for LT-SOFC operating 650degreeC and below since stability and compatibility of the materials in interest are secured at low temperatures. First, a nano-sized pyrochlore bismuth ruthenate (Bi2Ru2O7 or BRO7 shortly), one of the promising cathode materials, was successfully synthesized using glycine-nitrate combustion (GNC) route. Stoichiometric Bi2Ru2O7 without any impurity phase was achieved with considerably improved processing condition, leading to the crystallite size of ~24nm in diameter. Even though the resulting powder tends to agglomerate, resulting in overall 200~400nm size range, it still showed better quality than the one prepared by solid state (SS) reaction route followed by extra milling steps such as vibro-milling and sonication for further particle size reduction. Glycine-to-nitrate (G/N) ratio was found to play a critical role in determining the reaction temperature and reaction duration, thus phase purity and particle morphology (particle size, shape, and agglomeration etc). Composite cathodes of such prepared BRO7 (GNC BRO7) combined with SS erbia-stabilized bismuth oxide, Bi1.6Er0.4O3 or ESB, showed better electrochemical performance than vibro-milled BRO7 (VM BRO7)-SS ESB. ASR values of 0.123?cm2 at 700degreeC and 4.59?cm2 at 500degreeC, respectively, were achieved, which follows well the trend of particle size effect on performance of composite cathodes. Additionally, the number of processing steps (thus time) was reduced by GNC route. Several issues in regard to synthesis process and characteristics of BRO7 material itself will be addressed in this dissertation. Secondly, a unique in-situ composite cathode synthesis was successfully developed and applied for BRO7-ESB composite cathodes to improve percolation and to reduce agglomeration of each phase inside the cathode so that the effective triple phase boundary (TPB) length was extended. To disperse and stabilize ESB powder in de-ionized (DI) water, zeta potential profile of ESB powder in DI water as a function of pH was first achieved. The effect of a dispersant (ammonium citrate dibasic) on the stability of ESB powder dispersed in DI water was also investigated. Knowledge of BRO7 wet chemical synthesis from previous study was utilized for final product of in-situ BRO7-ESB composite cathodes. Such prepared composite particles were characterized and the electrochemical performance of in-situ BRO7-ESB composite cathodes was examined as well. Performance enhancement was observed so that ASR values of 0.097?cm2 and 3.58?cm2 were achieved at 700degreeC and 500degreeC, respectively, which were 19% and 22% improvement, respectively compared to those of conventionally mixed composite cathodes of BRO7-ESB. Finally, a highly controlled nanostructured BRO7-ESB composite cathode was developed by infiltration of BRO7 onto ESB scaffolds to maximize the effective TPB length, to improve the connectivity of ESB phase inside the cathode for better oxygen-ion diffusion, and to minimize delamination between the electrolyte and cathode layers. ESB scaffolds were first established by adding a graphite pore-former and controlling heat treatment condition. Nano-sized BRO7 particles were successfully created on the surface of previously formed ESB scaffold by infiltration of concentrated (Bi, Ru) nitrate solution followed by the optimized heat treatment. Such prepared composite cathodes exhibited superior electrochemical performance to conventionally made BRO7-ESB composite cathodes and even better than GNC BRO7-SS ESB developed in this dissertation, e.g. 0.073?cm2 at 700degreeC and 1.82?cm2 at 500degreeC, respectively. This cathode system was revealed to be highly competitive among all the reported composite cathodes consisting of the same or different materials prepared by various processing techniques. It was demonstrated that the extended TPB length from continuous network of BRO7 nanoparticles and better connectivity of ESB scaffolds enabled the outstanding performance. Moreover, de-lamination of cathode from the electrolyte was prevented thanks to improved adhesion between ESB scaffolds and ESB electrolyte. Dissociative adsorption of oxygen gas were proposed to be the dominant rate-determining process for the overall oxygen reduction reaction at low temperatures (500-600degreeC) whereas all of the constituting sub-reactions such as oxygen gas dissociative adsorption, oxygen ion diffusion towards TPB region, and oxygen ion incorporation were found to play roles competitively in the overall reaction at relatively high operating temperature (650-700degreeC) based on analysis of impedance spectra.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Byung Lee.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Wachsman, Eric D.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042563:00001


This item has the following downloads:


Full Text

PAGE 1

1 DEVELOPMENT OF A HIGH PERFORMANCE COMPOSITE CATHODE FOR LT SOFC By BYUNG WOOK LEE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010

PAGE 2

2 2010 Byung Wook Lee

PAGE 3

3 To my lovely F amily

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my advisor, Dr. Eric D. Wachsman, for his support and guidance. His encouragement helped me to reach a high er level of success and expand my potential. I also would like to thank my committee members, Dr. David Norton, Dr. Juan C. Nino Dr. Wolfgang Sigmund and Dr. Mark Orazem for their invaluable advice, guidance and constructive comments. In particular, I w ould like to thank my colleague, Dr. Matthew Camaratta for his endless e ncouragement, collaboration and friendship Without him, it would have been much difficult for me to ge t to this memorable moment of be ing a PhD. I also wish to acknowledge my forme r and current group members; Dr. Eric Armstrong, Dr. Kang Taek Lee, N ick Vito and other members for providing me excellent research environment and helpful comments. Dr. Craciun and Kerry Sieben in Major Analytical Instrumentation Center (MAIC) are also the people that I would like to give special credit for their invaluable discussion and collaboration on materials characterization. Last but not least, I would like to give special thank s to my parents and my sister for their sincere support patience and t rust toward s me

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 15 2 LITERATURE REVIEW ................................ ................................ .......................... 18 2.1. Operating P rinciples of S olid O xide F uel C ell ................................ .................. 18 2.2. Bi L ayer E lectrolyte ................................ ................................ .......................... 20 2.3. Single P hase Bismuth Ruthenate ( B i 2 R u 2 O 7 ) C athode ................................ .... 21 2.3.1. Electronic C onductivity of Bismuth Ruthenate ( B i 2 R u 2 O 7 ) ...................... 21 2.3.2. Structure of Bismuth Ruthenate (Bi 2 Ru 2 O 7 ) ................................ ............ 22 2.4. Dual Phase Composite Cathode of Erbia Stabilized Bismuth Oxide (Bi 1.6 Er 0.4 O 3 ) Bismuth Ruthenate (Bi 2 Ru 2 O 7 ) ................................ ..................... 23 2.5. Factors Affecting Cathode Performance ................................ .......................... 24 2. 6 Infiltration or Impregnation Process ................................ ................................ 25 3 SYNTHESIS OF A NANO SIZED PYROCHLORE BISMUTH RUTHENATE, Bi 2 Ru 2 O 7 USING GLYCINE NITRATE COMBUSTION (GNC) FOR LT SOFC CATHODE APPLICATION ................................ ................................ ...................... 33 3.1. Introduction ................................ ................................ ................................ ...... 33 3.2. E xperimental ................................ ................................ ................................ .... 35 3.2.1. GNC Bi 2 Ru 2 O 7 S ynthesis ................................ ................................ ....... 35 3.2.2. C ell Fabrication ................................ ................................ ...................... 36 3.2.3. Characterization ................................ ................................ ..................... 37 3.3. R esults and Discussion ................................ ................................ .................... 38 3.3.1. The E ffect of G lycine to Nitrate ( G/N ) R atio on B i 2 R u 2 O 7 S ynthesis ....... 38 3.3.2. Particle S ize A nalysis ................................ ................................ ............. 39 3.3.3. Phase Stability ................................ ................................ ........................ 41 3.3.4. Electrochemical P erformance ................................ ................................ 41 3.4. Conclusion ................................ ................................ ................................ ....... 43 4 SYNTHESIS OF IN SITU COM POSITE CATHODES OF BISMUTH RUTHENATE AND STABILIZED BISMUTH OXIDE FOR LT SOFC ...................... 53 4.1. Introduction ................................ ................................ ................................ ...... 53

PAGE 6

6 4.2. E xperimental ................................ ................................ ................................ .... 56 4.2.1. Stabilization of ESB Suspension ................................ ............................ 56 4.2.2. Synthesis Process ................................ ................................ .................. 56 4.2. 3. Characterization ................................ ................................ ..................... 58 4.3. R esults and Discussion ................................ ................................ .................... 58 4.3.1. Zeta Potential of ESB Suspension ................................ ......................... 58 4.3.2. Phase Purity of BRO7 ESB Composite and Formation of BRO7 Nanoparticles ................................ ................................ ................................ 60 4.3.3. Electrochemical Performance ................................ ................................ 60 4.4. Conclusion ................................ ................................ ................................ ....... 62 5 PYROCHLORE BISMUTH RUTHENATE, B i 2 R u 2 O 7 INFILTRATED CATHODES FOR HIGH PERFORMANCE L T SOFCS ................................ .......... 72 5.1. Introduction ................................ ................................ ................................ ...... 72 5.2. E xperimental ................................ ................................ ................................ .... 74 5.2.1. Infiltration Process ................................ ................................ .................. 74 5.2.2. Characterization ................................ ................................ ..................... 75 5.3. R esults and Discussion ................................ ................................ .................... 76 5.3.1. Cathode Microstructure and the Effect of Heating Co ndit ion .................. 76 5.3.2. Compositional Analysis ................................ ................................ .......... 78 5.3.3. The Effect of Infiltration on Cathode Polarization ................................ ... 78 5.3.4. Performance Comparison ................................ ................................ ....... 81 5.4. Conclusion ................................ ................................ ................................ ....... 82 6 SUMMARY ................................ ................................ ................................ ............. 96 APPENDIX A ISSUE ON SYNTHESIS OF PURE Bi 2 Ru 2 O 7 ................................ ....................... 102 B IMPEDANCE SPECTROSCOPY ................................ ................................ .......... 107 LIST OF REFERENCES ................................ ................................ ............................. 113 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 119

PAGE 7

7 LIST OF TABLES Table page 3 1 Crystallite sizes of pure BR O7 depending on 2 and its average ................................ ................................ ................................ ... 52 3 2 ASR values of all BRO7 ESB composite cathodes compared at each temperature measured ................................ ................................ ....................... 52 4 1 ASR values of all BRO7 ESB composite cathodes compared at each temperature measured ................................ ................................ ....................... 71 5 1 Activation energies of GNC BRO7 SS ESB (chapter 3) and infiltrated BRO7 cathodes (sample # 1 and #2) calculated from Figure 5 12 using equation (5 2) ................................ ................................ ................................ ........................ 95 5 2 ASR values of all BRO7 ESB composite cathodes compared at each temperature measured ................................ ................................ ....................... 95

PAGE 8

8 LIST OF FIGURES Figure page 2 1 Schematic of a fuel cell, comprised of an anode, electrolyte, and cathode ........ 28 2 2 Schematic vi ew of polarizations (or overpotentials) in a SOFC .......................... 28 2 3 T emperature dependence of electrical conductivity for Pb 2 Ru 2 O 6.5 Bi 2 Ru 2 O 7 CaRuO 3 and SrRuO 3 ................................ ................................ ......................... 29 2 4 Successive two formular units out of a pyrochlore unit cell (A 2 B 2 O 6 ) .............. 30 2 5 Composite cathode showing two parallel paths for oxygen species and electrons respecti vely ................................ ................................ ......................... 31 2 6 (a) Rct eff as a function of electrode thickness (b) The effect of finer structure on Rct eff ................................ ................................ ................................ .............. 31 2 7 Schematic illustrat ion of the current constriction effects for different cathode microstructure and current collector geometry ................................ .................... 32 2 8 Schematic of the microstructure derived by two types of infiltration strategy. ..... 32 3 1 Flow chart of Glycine Nitrate Combustion (GNC) route for BRO7 synthesis ...... 45 3 2 Effect of Glycine to Nitrates ratio (G/N) o n the crystallinity of as synthesized BRO7 precursors ................................ ................................ ................................ 46 3 3 Evolution of high purity BRO7 as a function of calcining temperature ................ 46 3 4 TEM images of BRO7 nanoparticles and aggregates of those ........................... 47 3 5 (Left) SEM image of BRO7 aggregates calcined at 700 C for 2hrs and (Right) Energy Dispersive Spectroscopy (EDS) analysis o n the particle (red circle in the left figure) ................................ ................................ ................................ ...... 47 3 6 Particle size distribution of BRO7 particles measured by TSI PSD 3603 (Aerosizer) ................................ ................................ ................................ .......... 48 3 7 Thermogravimetric (TG) plot to show the stability of the final BRO7 calcined 700C 2hrs powder at temperature range of 25~1000C ................................ ... 48 3 8 a) Nyquist and b) Bode plots as a function o f temperature for GNC BRO7 SS ESB composite cathode tested in air ................................ ................................ .. 49 3 9 Overlapped Bode plots measured at 600C of BRO7 ESB composite cathodes ................................ ................................ ................................ ............. 50

PAGE 9

9 3 10 Arrhenius plot of ASR for comparison of GNC BRO7 SS ESB with the other BRO7 ESB composite cathodes reported ................................ .......................... 51 4 1 Molecular structure of ammonium citrate dibasic ................................ ................ 64 4 2 Flow chart of synthesis process for in situ ESB BRO7 composite cathodes ...... 64 4 3 Zeta potential of ESB powder suspension with or with out a dispersant (ammonium citrate) as a function of pH ................................ .............................. 65 4 4 XRD patterns of in situ ESB BRO7 composite cathodes depending on calcining temperatures ................................ ................................ ....................... 66 4 5 SEM image of in situ BRO7 ESB composites showing adsorbed BRO7 nanoparticles on ESB phase ................................ ................................ .............. 67 4 6 a) Nyquist and b) Bode plots of in situ BRO7 ESB composite cathode t ested at each temperature in air under open circuit potential ................................ ....... 68 4 7 Overlapped Bode plots measured at 600C of BRO7 ESB composite cathodes ................................ ................................ ................................ ............. 69 5 1 Schematic of infiltration process ................................ ................................ ......... 83 5 2 Temperature profiles of two different heat treatment conditions for sample #1 and sample #2 ................................ ................................ ................................ .... 83 5 3 Schematic of impedance testing set up ................................ .............................. 84 5 4 Cross sectional S EM images of a) sample #1 and b) sample #2 ....................... 84 5 5 SEM images of BRO7 nanoparticles: the same spot with (a) low (b) high (zoomed in) magnifications ................................ ................................ ................. 85 5 6 a ) Representative SEM image of infiltrated BRO7 ESB composite cathodes showing t he region (bright region in the middle of image) used for porosity calculation where 30 vol% of graphite was used as a pore former to construct ESB scaffolds and b) calculated average porosity and its standard deviation .... 86 5 7 TEM images and EDS spectra of a) ESB scaffold and b) BRO7 nanoparticles with red circle on Ru element energy spectrum indicating the presence of Ru element in BRO7 nanoparticles ................................ ................................ .......... 87 5 8 Nyquist plots of BRO7 infiltrated ESB composite cathode (sample #1) at different temperatures (550~700C) ................................ ................................ ... 88 5 9 Nyquist plots of BRO7 infiltrated ESB composite cathod e (sample #2) at different temperatures (550~700C) ................................ ................................ ... 89

PAGE 10

10 5 10 Overlapped Nyquist plots of GNC BRO7 SS ESB, infiltrated BRO7 ESB sample #1, and sample #2 at different temperatures (550~700C) to b etter show the improvement of performance ................................ .............................. 90 5 11 O verlapped Bode plots of BRO7 infiltrated ESB composite cathodes (sample #1 and #2) and GNC BRO7 SS ESB composite cathode at different temperat ures (550~700C) for comparison on dominating processes ................ 91 5 12 Arrhenius plot of ASR for performance comparison with several conventionally mixed composite cathodes and the othe r infiltrated c athodes reported ................................ ................................ ................................ .............. 92 5 13 Overlapped Bode plots of BRO7 infiltrated ESB cathodes (sample #1 and #2), conventionally mixed BRO7 ESB composite cathode, and YSB in filtrated LSM cathodes at 700C ................................ ................................ ...................... 93 5 14 Arrhenius plot of ASR for performance comparison with all conventionally mixed BRO7 ESB composite cathodes ................................ .............................. 94 A 1 Diffractio n patterns of Bi 3 Ru 3 O 11 (or Bi 2 Ru 2 O 7.3 ) synthesized with amorphous citrate reaction method ................................ ................................ ..................... 105 A 2 Diffraction patterns of BRO7 and impurity phases such as Bi 3 Ru 3 O 11 (or Bi 2 Ru 2 O 7.3 ) synthesized with GNC method ................................ ....................... 105 A 3 Diffraction patterns of BRO7 (bottom) before and (top) after leaching taken from the literature ................................ ................................ ............................. 106 B 1 T ypical impedance cell response on complex plane and its equivalent circuit 111 B 2 Typical Warburg impedance response on complex plane ................................ 111 B 3 Impedance response for as sintered (1500 C, 4 h r ) HS3Y samples (circles), annealed at 1200 C for 110 h r in both 10% H 2 balance N 2 atmosphere (triangles), and air (diamonds) showing the effect of increasing grain size ...... 112 B 4 a) E quivalent circuit and b) impedance response in complex plane for mixed kinetic and charge transfer control ................................ ................................ .... 112

PAGE 11

11 Abstract of Dissertation Presented to the Graduate Schoo l of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DEVELOPMENT OF A HIGH PERFORMANCE COMPOSITE CATHODE FOR LT SOFC By Byoung Wook Lee December 2010 Chair: Eric D. Wachsman M ajor: Materials Science and Engineering Solid Oxide Fuel Cell (SOFC) has drawn considerable attent ion for decades due to its high efficiency and low pollution which is made possible since chemical energy is directly converted to electrical energy through the system without combustion However, successful commercialization of SOFC has been delayed due to its high production cost mainly related with using high cost of interconnecting materials and the other structural components required for high temperatur e operation. This is the reason that intermediate (IT) or low temperature ( LT ) SOFC operating at 600~800 C or 65 0C and below, respectively, is of particular significance because it allows the wider selection of cheaper materials such as stainless steel fo r interconnects and the other st ructural components Also, extended lifetime and system reliability are expected due to less thermal stress through the system with reduced temperature. M ore rapid sta rt up/shut down procedure is another advantage of lowerin g the operating temperatures As a result, commercialization of SOFC will be more viable. However, there exists performance drop with reduced operating temperature due to increa sed polarization resistances from the electrode electrochemical reactions and

PAGE 12

12 d ecreased electrolyte conductivity. Since ohmic polarization of the electrolyte can be significantly reduced with state of the art thin film technology and cathode polarization has more drastic effect on total SOFC electrochemical performance than anode pol arization as temperature decreases, development of the cathode with high performance operating at IT or LT range is thus essential. On the other hand, chemical stability of the cathode and its chemical compatibility with the electrolyte should also be con sidered for cathode development since instability and incompatibility of the cathode will also cause substantial performance loss Based on requirements of the cathode mentioned above, in this study, several chemico physical approaches were carried out to develop a high performance composite cathode, in particular, for LT SOFC operating 650 C and below since stability and compatibility of the materials in interest are secured at low temperatures. First, a nano sized py ro chlore bismuth ruthenate ( Bi 2 Ru 2 O 7 or BRO7 shortly ), one of the promising cathode materials, was successfully synthesized using g lycine n itrate c ombustion (GNC) route. Stoichiometric Bi 2 Ru 2 O 7 without any impurity phase was achieved with considerably improved processing condition, leading to t he crystalli te size of ~ 24 nm in diameter Even though the resultin g powder tends to agglomerate resulting in overall 2 00~ 4 00nm size range, it still showed better quality than the one prepared by solid state ( SS ) reaction route followed by extra milling st eps such as vibro milling and sonication for further particle size reduction. Gly cine to n itrate (G/N) ratio was found to play a critical role in determining the reaction temperature and reaction duration, thus phase purity and particle morphology (particl e size, shape, and agglomeration etc) Composite cathodes of such prepared BRO7 (GNC BRO7) combined with SS erbia

PAGE 13

13 stabilized bismuth oxide, Bi 1.6 Er 0.4 O 3 or ESB showed better electrochemical performance than vibro milled BRO7 (VM BRO7) SS ESB. ASR values o f 0.12 3 2 at 700C and 4. 59 2 at 500C, respectively, were achieved, which follows well the trend of particle size effect on performance of composite cathodes. Additionally, the number of p rocessing steps (thus time) was reduced by GNC route. Several i ssues in regard to synthesis process and characteristics of BRO7 material itself will be addressed in this dissertation. Secondly, a unique in situ composite cathode synthesis was successfully developed and applied for BRO7 ESB composite cathodes to improv e percolation and to reduce agglomeration of each phase inside the cathode so that the effective t riple phase b oundary (TPB) length was extended. To disperse and stabilize ESB powder in de ionized (DI) water, zeta potential profile of ESB powder in DI wate r as a function of pH was first achieved. Th e effect of a dispersant ( ammonium citrate dibasic) on the stability of ESB powder dispersed in DI water was also investigated. Knowledge of BRO7 wet chemical synthesis from previous study was utilized for final product of in situ BRO7 ESB composite cathode s Such prepared composite particle s w ere characterized and the electrochemical performance of in situ BRO7 ESB composite cathode s was examined as well. Performance enhancement was observed so that ASR values of 0. 097 cm 2 and 3.58 2 were achieved at 7 00C and 5 00C respectively, which were 19% and 22% improvement respectively compared to those of conventionally mixed composite cathode s of BRO7 ESB. Finally a highly controlled nano structured BRO7 ESB composit e cathode was developed by infiltration of BRO7 on to ESB scaffold s to maximize the effective TPB

PAGE 14

14 length, to improve the connectivity of ESB phase insid e the cathode for better oxygen ion diffusion, and to minimize delamination between the electrolyte and c athode layers ESB scaffold s w ere first established by adding a graphite pore former and controlling heat treatment condition. Nano sized BRO7 particles were successfully created on the surface of previously formed ESB scaffold by infiltrat ion of concentra ted (Bi, Ru) nitrate solution followed by the optimized heat treatment. Such prepared composite cathode s exhibited superior electrochemical performance to conventionally made BRO7 ESB composite cathode s and even better than GNC BRO7 SS ESB developed in thi s dissertation, e.g. 0. 073 2 at 700C and 1.82 2 at 5 0 0C respectively This cathode system was revealed to be highly competitive among all the reported composite cathodes consisting of the same or different materials prepared by various processing techniques It was demonstra ted that the extended TPB length from continuous network of BRO7 nanoparticles and better connectivity of ESB scaffold s enable d the outstanding performance. Moreover, de lamination of cathode from the electrolyte was prevented thanks to improved adhesion b etween ESB scaffolds and ESB electrolyte. D issociative adsorption of oxygen gas were proposed to be the dominant rate determining process for the overall oxygen reduction reaction at low temperature s (500 60 0C) whereas all of the con stituting sub reaction s such as oxygen gas dissociative adsorption, oxygen ion diffusion towards TPB region, and oxygen ion incorporation were found to play roles competitively in the overall reaction at relativel y high operating temperature (6 5 0 700C) based on analysis of imp edance spectra

PAGE 15

15 CHAPTER 1 INTRODUCTION A fuel cell is a device that converts the chemical energy directly into electricity by electrochemically combining a fuel with an oxidant through an ion conducting electrolyte. Among various kinds of fuel cells depe nding on the t ype of electrolyte being used, s olid o xide f uel c ell (SOFC), in particular, has received great attention as an alternative electric power generation system due to its high electrical efficiency, fuel flexibility, and minimal envi ronmental imp act etc 1 2 Owing to the high installation costs and poor systems durability, however, conventional high temperature SOFC operating at around 1000 C has prevented its widespread commercialization. As a result, trem endous effort has been devoted to the development of low temperature SOFC ( LT SOFC) operating at 500 700 C since it can bring the advantages over the conventional high temperature S OFC as below 3 4 Use of low cost metallic materials such as fer r itic stainless steels for interconnect and construction materials, which makes both the stack and balance of plant cheaper Longer operational lifetime and system reliability More rapid start up and shut down procedures requi red for potential applications in transpo rtation and mobile applications From an electrical point of view, however, reduction of the operating temperature is detrimental because the electrochemical processes limiting performance of the cell (anodic reactio ns, ion transport in the electrolyte and cathodic reactions) are thermally activated and become considerably sl ower at reduced temperatures 5 In particular, the more temperature goes down, the bigger cathode polari zation plays a role in governing overall performance of the cell. Consequently, the development of a cathode with low

PAGE 16

16 cathode polarization is inevitable in order to achieve reasonable electrochemical performance of L T SOFC which is ~0.5 W/cm 2 at operating vo ltage of ~0.7 V 2 It is well known that a composite cathode, consisting of an electro catalyst with high electronic condu ctivity and an oxygen ion conducting phase, reduces the cathode polarization significantly by effectively extending the reaction zone from the electrode/electrolyte interface into the electrode and by adding the ionic conductivity to the electrode, which p rovides additional path for tra nsport of oxygen species to the cat hode/electrolyte interface 6 7 In addition, further decrease in the cathode polarization can be achieved by two distinct approaches. One is to cont rol the geometrical par ameters such as porosity particle siz e, volume fraction, and spatial distribution of each c omponent for composite cathodes based on pe rcolation theory for better gas diffusion and connectivity of constituent phases. The other is to improve the materials properties such as the electronic conductivity and/or catalytic activity towards the oxygen reduction by doping or replacing the electrode material by itself 5 Meanwhile, since the cathode polarization also depends on the electrolyte materials system upon which the cathode is deposited as well as the cathode system by itself, care should be taken to select the right combination between the cathode and electrolyte 8 Abhishek et al. from our group developed a composite cathode con sisting of Bismuth ruthenate (Bi 2 Ru 2 O 7 or BRO7) and 20 mol% erbia stabilized bismuth oxide ((Bi 2 O 3 ) 0.8 (Er 2 O 3 ) 0.2 or ESB) that shows area specific resistance (ASR) values as low as

PAGE 17

17 ~0.08 cm 2 and ~3. 47 2 at 700 C and 500 C respectively, on gadolinium doped ce ria (GDC) as the electrolyte 9 More recently, Camaratta showed in his study that the composite cathode of ESB BRO7 can be further optimized by introducing pur e BRO7 current collector and changing microstructure such as constituent particle si ze and electrode thickness 10 The best performance that h e obtained was ASR as low as ~0.03 2 2 at 700 C and 500 C respectively, on 20ESB electrolyte with micron size of ESB and nano size of BRO7, which is comparable to one of the best results reported by S hao and Haile for Ba 0.5 Sr 0.5 Co 0.8 Fe 0.2 O 3 (BSCF) 11 However, it should be pointed ou t that yield ratio to get f ine particles (<100nm) of BRO7 wa s prohibitively low. Therefore, the more effective processing route is essential to get reproducible and scalable BRO7 ESB composite cathode systems without compromising hi gh performance verified from p revious study. Consequently, if all the requirements mentioned above for the optimized composite cathode were satisfied it is expected to achiev e a high performance composite cathode with low production cost, enabling the re aliz ation of LT SOFC.

PAGE 18

18 CHAPTER 2 LITERATURE REVIEW 2.1. Operating P rinciples of S olid O xide F uel C ell The primary components of a fuel cell are an ion conducting electrolyte a cathod e, and an anode as shown in Fig ure 2 1. 4 In the simplest example, a fuel such as hydrogen is fed into the anode and an oxidant, typically oxygen, into the cathode. The electrolyte serves as a barrier to gas diffusion, but allows the transport of ions (e.g. O 2 ) across it. Oxygen gas, combining with electrons, is electro reduced at the cathode to produce O 2 which is driven across the electrolyte by the chemical potential (or oxygen partial pressure) difference between the two sides of the electrolyte due to the oxidation of fuel at the anode. Free electrons are released at the anode when the fuel is oxidized, and they travel to the cathode through an external circuit. Thus, t he oxygen partial pressure gradient ( Po 2 ) over th e electrolyte is a driving force for the electromotive force across the electrolyte, which sets up a terminal voltag e towards the external load 12 The theoretical open circuit potential for a cell with oxygen potential gradient is given by the well kno wn Nernst equation in a cell 4 (2 1) where Po 2 and Po 2 are the equilibrium partial pressure of oxygen at the anode and cathode respectively the absolute temperature. The coefficient (n) in the denominator rep resents the number of electrons transferred per mole of oxygen molecules during reduction in the cell

PAGE 19

19 However, the output potential of a SOFC in practical operation is lower than the ideal values due to the over potential ( ) or potential loss induced fr om polarization. The total cell polarization consists of three major contributions caused by different mechanisms ohmic polarization, activation polarization, and concentration polarization as shown in Fig ure 2 2 and it can be expressed as the equation as follows: (2 2 ) The drop of potential due to activation polarization is associated with the electrochemical reactions in the electrodes. The second drop of potential comes from the ohmic resistance when ions and electrons conduc t in the electrolyte and electrodes, respectively. The third drop, which could be significant at high current densities, attributes to the mass transport resistance, or concentration polarization, of gas species. The rate of electrode reaction can be relat ed to the activation polarization by the Butler Volmer equation ; 13 (2 3 ) where promot ing the cathodic reaction. R, T, and F are the gas constant, absolute temperature, the number of electron s involved in the reaction and Faraday constant respectively. i o is t he exchange current density and i is the current density (negative for the cathod e current). In the low current density regime, it can be simplified as when act << RT (2 4 )

PAGE 20

20 In large current density regime, on the ot her han d, Tafel equation is applicable : (2 5 ) where a and b are influenced by the electrode microstructure and thickness. 2.2. Bi L ayer E lectrolyte With regard to an electrolyte material for LT SOFC, doped ceria and stabilized bis muth oxide have been proposed and studied by many researchers due to their re latively high ionic conductivities at 500 700 C It has been shown that Gd doped ceria (G DC) with as low as ASR of ~0.15 cm 2 can be achieved with thickness of 15 m at 500 C 14 Stabilized cubic bismuth oxides on the other hand, are known to exhibit the highest ionic conductivities due to their inherent concentration of oxygen vacancies and weak metal oxide bonds 15 16 The highest ionic conductivity was obtained by the fcc phase of Er 2 O 3 stabilized Bi 2 O 3 (ESB) Verkerk et al. examined the oxygen ion conductivity of 20mol% Er 2 O 3 stabilized Bi 2 O 3 ( 20ESB ) and reported its value s as 0.37S/cm and 0.023 S/cm at 700 C and 500 C respectivel y 17 However, each of them has some critical issues to be used alone as an electrolyte. For GDC, when it is placed at low oxygen partial pressures (<10 14 atm), Ce 4+ is reduced into Ce 3+ leading n type electronic conduction with P(O 2 ) 1/4 depe ndence 18 This phenomenon lowers the ionic transference number ( t i ) and open circuit potential (OCP) of the cell using GDC as an electrolyte due to the electronic conduction through it, thus resulting in the lower cell efficiency. Meanwhile, bismut h oxide based electrolytes seem to show the decomposition of Bi 2 O 3 to Bi under a moderate reducing atmosphere, which seriously limits its application in fuel cell environments. Takahashi et al. showed that the

PAGE 21

21 decompositio n of Bi/Bi 2 O 3 happens under the oxygen partial pressure ( Po 2 ) of ~ 10 13.1 atm at 600 o C 19 As an alternative, Wachsman et al. developed a bi layer electrolyte concept to overcome the thermodynamic instability of both mat erials 20 21 In their work, they demonstrated that higher OCP was obtained using bi layer of doped ceria and s tabilized bismuth oxide than using each of single electrolyte layer respectively since the ceria layer p r events the bismuth oxide layer from decomposing by shielding it from low Po 2 in the anode side and in turn, the bismuth oxide layer serve s to block the electronic conduction through doped ceria into the cathode side. Therefore, the development of a catho de system applicable for GDC/ESB will be reasonable. 2.3. Single P hase Bismuth Ruthenate ( B i 2 R u 2 O 7 ) C athode 2.3.1. Electronic C onductivity of Bismuth Ruthenate ( B i 2 R u 2 O 7 ) For a single phase purely electronic conducting cathode, it has to meet the followin g requirements: high electronic conductivity at operating temperature, thermal and chemical compatibility with the electrolyte namely, similar thermal expansion coefficient s (TEC) and no reaction with the electrolyte and high catalyti c activity for oxyge n reduction 22 In the beginning of bismuth/ceria bi layer SOFC research, the cathode materials were mainly gold or platinum, a single phase purely electronic conducting cathode, since rather widely used perovskite cathodes with low cathode polarization have the high re activity with bismuth oxides 23 Meanwhile, pyrochlores based on bismuth ruthenate, lead ruthenate, and yttrium ruthenate have been studied as cathodes for application in SOFC since they exhibit relatively high electrical conductivity 24 26 In particular, Takeda et al. studie d pyrochlores

PAGE 22

22 such as Bi 2 Ru 2 O 7 and Pb 2 Ru 2 O 6.5 and perovskites such as CaRuO 3 SRRuO 3 as cathode materials for YSZ based SOF C. As can be seen in Fig ure 2 3 25 they observed metallic behavior for pyrochlores with almo st temperature independent conductivity (10 2 10 3 S/m from room temperature up to 900 ), w hich is comparable to the best conventional single phase cathode materials. However, since cathode reaction sites (TPBL) are limited to near the interface between the electrolyte and cathode for single phase purely electronic conducting, it is more de sirable to introduce an ionic conducting phase into the cathode layer so as to increase TPBL and ionic conductivity resulting in lower polarization. 2.3.2. Structure of Bismuth Ruthenate (Bi 2 Ru 2 O 7 ) Successive 2/8 formula units out of a pyrochlore A 2 B 2 O 7 un it cell (consisting of eight formula units) are shown in Fig ure 2 4. It belongs to the family of cubic pyrochlore type structure with space group and lattice parameter of 27 28 The cation sublattice consists of bigger A 3+ (Bi 3+ in Bi 2 Ru 2 O 7 ) and smaller B 4+ (Ru 4+ in Bi 2 Ru 2 O 7 ) which order into alternate (110) rows in every other (001) plane and into alternate ( ) rows in the other (001) places respectively. This cation ordering provides three distinguishable tetrahedral sites for the oxygen ions: 8a sites surrounded by 4 A 3+ (Bi 3+ in Bi 2 Ru 2 O 7 ) cations, 8b sites surrounded by 4 Ru 4+ cations, and 48f sites surrounded by 2 Bi 3+ and 2 B 4+ (Ru 4+ in Bi 2 Ru 2 O 7 ) cations. While 8a and 48f sites are occupied, 8b sites are vacant, which results in an ordered oxygen io n sublattice. Thus, the formula unit of the pyrochlore can also be written as A 2 B 2 O 6 O (Bi 2 Ru 2 O 6 O for Bi 2 Ru 2 O 7 ) to distinguish between the oxygen ions occupying the 48f site as O and those occupying the 8a sites as O

PAGE 23

23 An electronic band structure study was carried out for metallic Bi 2 Ru 2 O 7 which revealed that the metallic character originates from the R u t 2g block bandwidth where the more Ru t 2g block bandwidth increases resulting from an increase in Ru O Ru bond a ngle and an shortening of the Ru O bond, the more the material becomes metallic 29 30 2.4. Dual Pha se Composite Cathode of Erbia Stabilized Bismuth Oxide (Bi 1.6 Er 0.4 O 3 ) Bismuth Ruthenate (Bi 2 Ru 2 O 7 ) Since dual phase composite cathodes, consisting of an ionic conducting phase and an electronic conducting phase, have proved to have lower cathode polariza tion than a single phase pure ly electronic conducting cathod e 6, 31 32 the development of dual phase composite cathode s for bismuth/ceria bi layer electrolyte has been a focus of study in our group In general, com posite cathode s consisting of an electro catalyst with high electronic conductivity and an oxygen ion conducting phase, are expected to not only increase TPBL, which results in lower over potential toward oxygen reduction, but shorten surface diffusion pa th of dissociatively adsorbed oxygen to TPB sites upon addition of an ionic conducting phase as can be seen Fig ure 2 5 33 Moreover, there arises an additional path for transport of oxygen species such as bulk diffusion through the ionic conducting phase. T hermal expansion also can be matched well between the composite cathode and electrolyte due to use of the same ionic conducting phase in both layers. For instance, a composite cathode of ESB BRO7 exhibiting ASR as low as ~0.03 2 2 at 700 C an d 500 C respectively, on 20ESB electrolyte was developed thanks to the high electronic conductivity and catalytic activity towards oxygen reduction of BRO7 phase and the high ionic conductivity of ESB phase as well. As expected this composite cathode sys tem showed no chemical reacti vity with ESB electrolyte used 10

PAGE 24

24 2.5. Factors Affecting Cathode P erformance Virkar et al. developed theoretical model showing the effect of po rous composite electrodes on the overall charge transfer process by taking into account various parameters such as intrinsic charge transfer resistance, electrode thickness, ionic conductivity of the ionic conducting phase a nd porosity et c 32, 34 Low current regime is assumed so that the activation polarization can be approximated as being ohmic as follows ; (2 6 ) Diffusion of gaseous species in porous e lectrodes is assumed to be rapid so as not to be rate limiting. Also, the electronic conductivity in the electronic conducting phase is assumed to be high enough. The performance of the electrodes is judged by the effective charge transfer resistance (R ct e ff ), which is defined as ASR of the entire SOFC minus ASR of the dense electrolyte. It was demonstrated that R ct eff decreases as the thickness of the composite cathode increases, eventually approaching an asymptotic minimum as c an be seen in Fi gure 2 6 (a) On the contrary, as the electrode thickness approaches zero, R ct eff becomes close to the intrinsic charge transfer resistance R ct implying that the electrode behaves as thin single phase electronic conducting cathode. In addition, the model shows that fin er microstructures give better electrode performance Fig ure 2 6 (b) However, since concentration polarization increases with increasing electrode thickness, the optimum microstructure of the electrode should be made as fine as possible to permit the use of moderate thickness of

PAGE 25

25 cathodes to ensure that concentration polarization is not a limiting factor and simultaneously to obtain as low value of Rct eff as possible. On the other hand, an inhomogeneous current distribution is expected when the number of cont acts between the cathode and current collecting mesh and lead wire is insufficient, leading to higher in plane electrode resistance as can be seen in Fig ure 2 7. 35 Thus, a porous current collecting layer of coarse grains on top of a thin cathode layer of fine grain s was suggested for L T SOF C 36 This is the reas on that the pure BRO7 current collector layer substantially influences the overall performance. 2. 6 Infiltration or Impregnation P rocess A nano structure d cathode is theoretically an ideal microstructure possessing the advantages of high electrocatalytic activity and large TPB length. It has been reported that the nano sized oxides improved the catalytic properties due to an increase in surface vacancy concentration, ionic and electronic cond uctivities 37 38 Modeli ng studies show that larger TPB length is associated with smaller particles 39 40 However, since SOFC cathode s require the structur al integrity with adequate mechanical strength and stability at operating temperatu re s typically higher 500 C the cathode entirely comprised of nano structure may be not suita ble. Thus, the nano structure should be suppor ted on a more robust micro sized cathode backbone. The technique to construct this microstructure is called infiltra tion or impregnation. Although the infiltration has been used in relatively recent y ears to improve performance of the SOFC electrodes, it attract ed man y research activities both experimental ly 41 49 and theoretica lly through modeling s tudies 50 51 Several reviews with different emphas i s in this area are also available 52 54

PAGE 26

26 The infiltration/impregnation technique involves depositin g nanoparticles into a pre sintered backbone. The sintering of backbone is normally performed at high temperature Thus, mechanical bonding between the electrode backbone and the electrolyte is good enough to give the structural stability to the system. Mo reover, sufficient percolation of each constituting phases ensures high effective conduction of electron or oxygen ion On the other hand, t he firing process for the deposition of nanoparitcles and the formation of the desired phase can be conducted at tem perature s much lower than that needed for the traditional ceramic fabrication process. For example, while the firing temperature of the traditional LSM YSZ composite s is usually higher than 1100C 55 the infiltration process requires only 800 C for the deposition of LSM on the YSZ backbone 46 The low temperature fabrication brings about many advantages. Since the firing temperature is low, nanoscale cha racteristics of particles can be preserved which is beneficial to achieving high catalytic activity and large TPB length. A modeling study show ed that the TPB length in a cathode prepared by infiltration is much larger than that of the composite electrode prepared by the traditi onal ceramic mixing process 50 In addition, the reaction between the MIECs and YSZ can be avoided since the firing temperature is low, making it possible to us e these MIEC materials which have high electrochemical performance but high reactivity with YSZ electrolytes, as cathodes for SOFCs The infiltration may also alleviate the thermal expansion mismatch. For cathode s fabricated by the infiltration process, the coefficient of thermal expansion ( CTE ) is mainly dominated by the back bone materials. T hus infiltrating with a material of mismatched CTE has little effect on the overall CTE of the cathodes

PAGE 27

27 Fig ure 2 8 shows the schematic of the microstructures cons tructed by the infiltration method. Two types of nano structures are possible since the electrode is usually a composite consist ing of two phases, a n e lectron ic conducting phase and an ion ic conducting phase I nfiltrating an electron conduct ing nano siz ed electrocatalyst into a n ion conduct ing backbone is one option or vice versa For these two types of infiltration depositing nanoparticles into the backbone not only increases TPB reaction sites for oxyg en reduction, but also provid es the pathway for charg e transport. Another strategy is to infiltrate nanoparticles of the electrocatalyst and/or electrolyte into a mixed conducting backbone of composites co mprised of both ionic and electronic conductor s or of a single phase MIEC material. For this strategy, t he infiltrated nanoparticle s are not required to provide the function of charge transport though

PAGE 28

28 Figure 2 1 Schematic of a fuel cell, comprised of an ano de, electrolyte, and cathode 4 Fig ure 2 2 Schematic view of polarizations (or overpotentials) in a SOFC 12

PAGE 29

29 Figure 2 3. T emperature dependence of electrical conductivity for Pb 2 Ru 2 O 6.5 Bi 2 Ru 2 O 7 CaRuO 3 and SrRuO 3 25

PAGE 30

30 Figure 2 4. S uccessive two formula units out of a pyrochlore unit cell ( A 2 B 2 O 6 O )

PAGE 31

31 Figure 2 5. C omposite cathode showing two parallel paths for oxygen species and electrons respectively 33 Figure 2 6. (a) Rct eff as a function of electrode thickness (b) The effect of finer structure on Rct eff 32

PAGE 32

32 Fig ure 2 7 Schematic illustration of the current constriction effects for different cathode micr ostructure and current collector geometry 35 Figure 2 8. Schematic of the microstructure derived by two types of infiltration strategy. (l eft: ionically conducting nanoparticles with electronically conducting backbone right: electronically conducting nanoparticles with ionically conducting backbone ) 56

PAGE 33

33 CHAPTER 3 SYNTHESIS OF A NANO SIZED PYROCHLORE BISMUTH RUTHENATE Bi 2 Ru 2 O 7 USING GLYCINE NITRATE COMBUSTION (GNC) FOR LT SOFC CATHODE APPLICATION 3.1. Introduction Transition metal oxides with a pyrochlore structure described as A 2 [Ru 2 x A x ]O 7 y been actively investigated as potential electro catalyst s or electrode material s for batter ies sensor s and fuel cell applications over several decades due to their high catalytic activity toward s oxygen reduct ion and/or evolution, as well as relatively high electronic conductivity with metallic behavior (> 300Scm 1 ) at targeted temperatures 57 61 More recently, these ruthenium pyrochlores have drawn attention as a promi sing cathode material for intermediate or low temperature s olid o xide f uel c ell (SOFC) applications (600 8 00C and <6 00C respectively ) 23 26, 62 Among them, in particular, bismuth ruthenate is more of interest be cause of its better stability and less toxicity than lead ruthenate at temperature range of 500 700C 26 There exists two represent ative bismuth ruthenates, i.e. Bi 2 Ru 2 O 7 (BRO7) and Bi 2 Ru 2 O 7.3 (or Bi 3 Ru 3 O 11 ) in ambient atmosphere depending on the temperatures. Low temperature bismuth ruthenate Bi 2 Ru 2 O 7.3 with cubic structure transforms irreversibly to high temperature pyrochlore Bi 2 Ru 2 O 7 beyond 950C 63 64 Even though there is some discrepancy regarding the transition temperature, it is obvious that high tem perature pyrochlore BRO7 is difficult to form at temperature below 900C. In general, as a result, high temperature pyrochlore BRO7 is synthesized using solid state (SS) reaction where the appropriate amount s of constituting oxides (RuO 2 and Bi 2 O 3 ) for the final composition ( Bi 2 Ru 2 O 7 ) are mixed and heated to over 900 C for several tens of hou r s for long range diffusion among reactants Thus, t he resulting

PAGE 34

34 powder contains particles of several microns in size due to grain growth inherent from long period of high temperature heat treatment 25, 63 Several extra processing steps such as ball milling and v ibro milling etc are required to further break up the micron sized particles into the submicron range Accordingly this conventional SS reaction route is not suitable for producing a nano sized powder. O n the other hand it was recently reported from our group that ASR values as low as 0.03 1 1 at 700C and 500C, respec tively, can be achieved for an optimized size combination of BRO7 ESB composite cathodes where very fine particle size of BR O7 was prepared by SS reaction route with multiple extra processing steps such as vibro milling sonication and sedimentation 10 However, it should be pointed out that the yield ratio to get those f ine particles (<100nm) of BRO7 wa s prohibitively low. Therefore, the processing route attempted above is not an effective fabrication method for mass production. Contrary to th e conventional SS reaction route, wet chemical synthe sis methods such as Pechini type citrate route, sol gel route 65 66 co precipitation route 67 and g lycine n itrate c ombustion (GNC) route 68 69 etc require relatively lower processing temperature, thanks to more intimate (molecular level) mixing between reactants That is, the nucleation process takes place through the rearrangement and short distance diffusion of nearby atoms and mole cules, not long distance diffusion, which normally happens with conventional SS reaction route. As a result, more fine and homogeneous powder with large specific surface area can be achieved with wet chemical synthesis routes Among these, GNC route was sp ecifically chosen in this study to synthesize nano sized pyrochlore BRO7 since GNC route has to date never been reported for

PAGE 35

35 BRO7 synthesis while there have been several attempts to prepare submicron BRO7 through co precipitat ion 70 and sol gel route s 66 GNC allows for very short reaction time s which is a characteristic inherent from combustion and the gas by p roducts (CO 2 N 2 or H 2 O) produced during the reaction partially p revent the particles from agglomerat ing so that the development of nano sized powders is plausible even though the reaction temperature could be as high as ~1400C 69 In this work, g lycine to n itrate (G/N) ratio was carefully examined to determine the optimum ratio for better crystallinity and phase purity of BRO7. Also, calcining temperature and duration were controlled to get the smallest nano sized particle as possible. Such synthesized BRO7 was mixed with erbia stabilized bismuth oxide (Er 0.4 Bi 1.6 O 3 ESB) to make a composite cathode and its electrochemical performance was compared to that of the conventionally made BRO7 ESB c athodes reported previously from our group to examine the feasibility of GNC route for BRO7 nanoparticle synthesis The increased t riple p hase b oundary (TPB) reaction sites in BRO7 ESB composite cathode arising from more contacts between nano sized BRO7 an d micron sized ESB w ere expected to improve electrochemical performance which was evidenced already for composite cathodes consisting of vibratory milled BRO7 (VM BRO7) and SS route ESB (SS ESB) in the previous study from our group 10 3.2. E xperimental 3.2.1. GNC Bi 2 Ru 2 O 7 S ynthesis The synthesizing process is outlined in flowchart in Figure 3 1. Metal nitrates are employed as oxiding agents while the glycine plays roles as both a chelating agent for metal cations (Bi 3+ and Ru 4+ in this case) and a fuel, i.e. reducing agent. Stoichiometric amount of Ru((NO)(NO 3 ) 3 (Ru 31.3% min, Alfa Aesar) and Bi(NO 3 ) 3 2 O (Puratronic

PAGE 36

36 99.999%, Alfa Aesar) for the final BRO7 were d issolved s eparately in d e ionized (DI) wa ter to form an aqueous solution The solution w as then added to the glycine solution where varing amount s of glycine in the solid state were dissolved in DI water as determined by the required g lycine to n itrate (G/N) ratio. The whole solution was then stirred for 1 hr on a hot plate at 80 C for better mixing. Next the hot plate was heated to 120C for water evaporation. Once water is entirely evaporated, the solution beco mes viscous gel. Upon further heating of the hot plate to 250~300C, spontaneous, exothermic autoignition (combustion) begins and ends rapidly within 3s, yielding a porous and foamy black ash of BRO7 precursor. As prepared materials were then ball milled in ethanol for 24 hrs using YSZ grinding media and drie d on a hot plate with stirring to break up any aggregates insid e the precursor before calcin ation The dried precursors were further calcined at temperatures over 600C for 2hrs in the open air to remove any residual carbonates and unreacted phases. 3.2.2. C ell Fabrication ESB powder was obtained using conventional SS reaction synthesis and then crush ed by mortar and pestle and sieved (325 M esh). ESB electrolyte pellets were prepared with 3.0 g of these solid state powders by uniaxial pressing (approximatel y 12 Kpsi) in a half inch die followed by isostatic pressing (250 MPa). These green bodies were then fired at 890 C for 15 hrs. The sintered pellets had densities 94%2% of theoretical with 1.11 cm 0.02 cm in diameter and 0.3 cm 0.01 cm in thickness. C athode inks were prepared by combining organic vehicles with the mixture of BRO7 and ESB powders (50 50 wt%) in appropriate ratio for 35~40vol% of porosity in the final cathode after sintering. The inks were applied to both sides of the ESB electrolyte sub strates by paint brushing to make symmetrical cells. These cells were dried at

PAGE 37

37 120C for 1 h r and sintered at 800 C for 2 h rs Pure BRO7 powder was separately mixed with organic vehicles and applied on top of the cathodes as a current collector. The acti ve cathode area where th e cathode ink was applied ( the top or bottom surface area of the electrolyte), was 0.97cm 2 3.2.3. Characterization X Ray Diffraction (XRD) patterns of several precursors and powders were recorded with a Philips APD 3720 diffracto meter using (Cu) = 1.5406 to determine phase formation, purity, crystallinity, and approximate crystallite size. Step scans were taken over a range of 2 angles, 20~80, with 0 02 step scan and 0.5sec/step The particle size was estimated by a JEOL TEM 200CX tr ansmission electron microscope (TEM). T he evaluation of morphology and microstructure of cathode system was carried out with the aid of JEOL JSM 6400 scanning electron mi croscope (SEM) Energy Dispersive Spectrum (EDS) was also used to analyze the composit ion of the BRO7 powders. Particle size distribution was analyzed with TSI PSD 3603 (Aerosizer) from Ahmherst Process Instrument Co. Inc. to better understand behavior of particles such as agglomeration. Thermogravimetric and differential thermal analysis ( TG/DTA) was performed on the final BRO7 powder (calcined at 700 C 2hrs ) using Mettler Toledo TGA/ DTA instrument to examine the stability of the phase at temperature range of 25 ~1000 C In addition, two electrode Electrochemical Impedance Spectroscopy (EIS) was measured on GNC BRO7 SS ESB composite cathodes deposited on both sides of ESB electrolyte pellet using Solartron 1470E frequency response analyzer in standalone mode for unbiased testing with frequency range of 0.1MHz to 0.01Hz and AC voltage amplitud e of 100mV.

PAGE 38

38 3.3. R esults and Discussion 3.3.1. The E ffect of Glycine to Nitrate ( G/N ) R atio on B i 2 R u 2 O 7 S ynthesis According to the propellant chemistry 71 the maximum heat during combustion is obtained when the net oxidizing valency of metal nitrates is equal to the net reducing valency of fuel. In the case of GNC route, hydrogen and carbon can be considered to be r educing agents with valency of +1 and +4, respectively, whereas oxygen is regarded as an oxidizing agent with a valency of 2 and nitrogen is assumed to have a valency of zero 68 In addition, for BRO7, Bi and Ru can be assumed to have valency of +3 and +4, respectively. With these valency values, the stoichimetric redox reaction can be calculated as follows: ( 3 1) GNC route produces N 2 H 2 O, and CO 2 gases as by products in addition to BRO7 precu rsor. The optimum G/N ratio, which can generate the maximum reaction heat, was found to be about 0.6 based on the calculation a bove. Five different G/N ratios ( 0.4, 0.6, 0.8, 1.0, and 1.2 ) were chosen in this study for BRO7 synthesis to evaluate the effect of G/N ratio on the crystallinity and particle size of BRO7 precursor. As seen in Figure 3 2 B RO7 precursor s become more crystalline from amorphous as G/N ratio is varied from 0.4 to 1.2 even though the best crystallinity is expected to appear at G/N=0.6 based on the maximum reaction heat calculated above. This could be attributed to the presence of abundant oxygen gas from the environment so that more reducing agent, i.e. glycine, is required to generate the maximum heat, which is in turn evidenced as be tter crystallinity for G/N ratios above 0.6. However, the best peak pattern similar to pure

PAGE 39

39 BRO7 was achieved at G/N=1.0 with less impurity phases such as unreacted Bi 2 O 3 Ru and/or RuO 2 Therefore, G/ N ratio of 1.0 was selected for further study of phase evolution according to calcining conditions and microstructural analysis on the final powder. Figure 3 3 shows the effect of further heat treatment on the phase purity of y a critical role in phase evolution. BRO7 precursor calcined at 700C for 2hrs exhibited almost the same XRD peak pattern as the one calcined at 700C for 10hrs. The only difference was the peak sharpening of BRO7 calcined at 700C for 10hrs due to its gr ain growth. T hus, the duration was fixed at 2hrs for this study while the calci ning temperatures were varied from 600 900C at 100C interval As calcining temperature increased the crystallinity increased accordingly. Pure BRO7 phase was achieved even at 700 C for 2hrs, which is much lower compared to t he 900 C for tens of hours necessary for the formation of pure BRO7 with conventional SS route 10 Impurity phases such as unreacted Bi 2 O 3 was totally removed beyond 700 C 3.3.2. Particle S ize A nalysis The size of BRO7 crystallites obtained after 700 C 2hrs calcining was estimated using the Scherrer s equation 72 from X ray line broadening of several peaks as follows : ( 3 2) where D is the average particle size in nm, the wavelength of X ray used (Cu K radiation, 0.15406nm, in this case), B the width in radian at half maxim um intensity (called full width half maxima (FWHM) ), and the Bragg diffraction angle of the line. The true B value can be calculated from the following equation: ( 3 3 )

PAGE 40

40 W here B M is the measured full width at half maximum i ntensity and B S is the full width at half maximum intensity of standard silicon. The calculated values for several main peaks and average crystallite size are given in Table 3 1. The average crystallite size of ~24 nm in diameter was achieved for B RO7 calci ned at 700C for 2hrs. TEM images of BRO7 calcined at 700C for 2hrs (Figure 3 4) also confirmed that crystallites in size range of 20~40nm were actually formed. However, it should be pointed out that the resultin g particles tend to agg regate during calci n ation whi ch was evidenced with SEM image as shown in Figure 3 5 From SEM image, the agg rega ted par ticles seemed to be distributed in size of 2 00~ 4 00nm which was also confirmed by particle size distribution analysis (Figure 3 6). Peak position of partic le distribution was ~360nm in diameter thus most o f particles seemed to aggregate and behave as polycrystalline particles rather than single crystal. However, this size range is still much smaller than that of the conventionally made one. For example, the particle size of vibro milled (VM) SS BRO7 for 7 days was reduced from ~ 1.31 m of as prepared SS BRO7 to ~ 0.73 m which is still about two times larger than that of GNC BRO7 particles and contains a small number fraction but significant volume fraction of large, unbroken particles 10 On the other hand, EDS analysis on nano particle s confirmed the formation of Bi 2 Ru 2 O 7 ( Figure 3 5 ). Consequently, a nano sized BRO7 was synth esized with reduced processing time T his was made possible thanks to both high reaction temperature (> 9 00C) and shorter r eaction time (< 3s) a characteristic of GNC route, which is not pos sible with conventional SS synthesis and many other methods.

PAGE 41

41 3.3. 3. Phase Stability Figure 3 7 shows TG plot obtained on the final BRO7 powder calcined at 700 C for 2hrs to see if there s any weight loss involved at temperature range of 25 ~1000 C As can be seen, there was no weight loss detected at practical level up t o 100 0 C indicating that the stability of the final BRO7 should be preserved at testing condition (500~700 C ) 3.3.4 Electrochemical P erformance E lectrochemical impedance test ing was carried out on BRO7 ESB composite cathodes. The performanc e of this com posite cathode w as expected to be s uperior to composite cathodes comprised of SS BRO 7 SS ESB and VM BRO7 SS ESB d ue to the expanded TPB reaction sites resulting from the smaller particle size of BRO7 Figure 3 8 shows ty pical Nyquist plots of GNC BRO7 SS E SB cathodes as well as Bode plots ( a convenient graphical representation of impedance data for locating characteristic frequencies ) 73 obtained at different temperatures (700~550C ) in air under open circuit potential. The plot is R el corrected for easy comparison (Figure 3 8 a) That is, the high frequency real axis in tercept, R el is subtracted from the real component of each data point in spectr a In this way, the low frequency inter cept with the real axis directly corresponds to the cathode polarization R p T herefore area specific resistance ( ASR ) of the cathode ca n easily be calculated by multiplying R p by the electrode area (~0.96 cm 2 in this study) and divid ing by 2 to account for the symmetrical cell. The h igh frequency real axis intercept, R el is known to mainly consist of the bulk electrolyte resistanc e, l ead resistance and contact resistance between the cathode and mesh current collector 10 As expected, ASR increases as temperature goes down since the oxygen reduction reacti on is thermally activated process 74

PAGE 42

42 On the other hand, characteristic freque ncy is believed to be associated with the dominating electrochemical process. From the Bode plot in Figure 3 8 b it can be noticed that the characteristic frequency (hig hest point of each spectrum in Bode plot ) is changed from 10~1kHz at 700C to <10Hz at 500 C which is typically assumed to be related to oxygen adsorption (1~10Hz) and gas diffusion through the pores (<1Hz) respectively 55 To further examine whether the dominating process is affected by use of GNC BRO7 nanoparticles, the Bode plots of four d ifferent cathodes at 600 C were compared as shown in Figure 3 9 The characteristic frequencies of GNC BRO7 SS ESB, VM BRO7 SS ESB, and Sup BRO7 SS ESB were almost the same (5~6 Hz) indicating that the dominant electrochemical process was not changed amon g them Ho wever, SS BRO7 SS ESB showed a little bit lower characteristic frequency (~1Hz). Since ESB for all of these cathodes was prepared with the same SS synthesis (SS ESB), the difference in characteristic fr equency seemed to be caused by difference in BRO7 phase, that is, particle size depending on the synthesis routes. Although the difference in characteristic frequencies is not substantial, the extended TPT reaction sites and better percolation resulting from nano sized BRO7 (VM, Sup, and GNC BRO7) a re believed to influence on the dominating process. In addition, Figure 3 10 is the Arrhenius plot of ASR for direct comparison between GNC BRO7 SS ESB in this study and the reported cathode systems. 10 As expected, ASR values of GNC BRO7 SS ESB were lower than those of SS BRO7 SS ESB and VM BRO7 S S ESB at examined temperatures due to the extended TPB sites as the

PAGE 43

43 particle size of BRO7 is reduced. ASR values of all the BRO7 ESB composite cathodes compared are listed in Table 3 2 for better comparison among ASR values The activation energy of GNC BRO7 SS ESB cathode system was determined to be ~1.26 eV which is comparable to the previously reported values, 1.2~1.34eV 9 but higher than 1.02eV 10 H owever, it should be mentioned that since the impedance semicircle at 5 00C was incomplete due to delayed response at low temperature, the intercept difference (ASR) was estimated roughly with the instant fit (Zview software). Therefore, there should be so me errors involved in calculation of ASR at 500C The ASR can be further improved by optimizing the composition of composite cathodes based on the effective medium percolation theory 10, 75 Since BRO7 particles ar e a lot smaller than ESB particles, percolation and TPB length maximization can occur at non equal volume fraction of each phase, considering the fact that the densities of BRO7 and ESB20 are approximately the same (8.92 and 8.96 g/cm 3 respectively). Redu ced processing steps and high yield of GNC synthesis route are expected to enable such further optimization, which is not possible with the conventional SS reaction route followed by laborious breaking down steps. 3.4. Conclusion A nano sized Bi 2 Ru 2 O 7 (BRO 7) powder was successfully synthesized using g lycine n itrate c ombustion (GNC) route. Pure BRO7 phase was achieved upon calcin ing at 700 C for 2hrs with t he average crysta llite size of ~ 24 nm in diameter Even though the resulting powder tends to a ggregate resulting in overall 2 00~ 4 00nm size range it still showed much smaller particle size than the one prepared by SS reaction route followed by extra milling steps such as vibro milling and sonication for further

PAGE 44

44 particl e size reduction The g lycine to n itrat e (G/N) ratio turned out to play a critical role in determining the reaction temperature and reaction duration, thus phase purity and particle morphology (particle size, shape and agglomeration etc) Composite cathodes of such prepared BRO7 with SS ESB ( GNC BRO7 SS ESB) showed better electrochemical performance than SS BRO7 SS ESB and VM BRO7 SS ESB cathodes ASR values of 0.12 3 2 at 700C and 4. 59 2 at 500C were achieved respectively which follow s well the trend of particle size effect on performan ce of composite cathodes. Additionally, the number of p rocessing steps (thus time) was reduced by GNC route. In conclusion, i t wa s demonstrated that GNC can be an effective synthesis route to produce nanoparticles of pyrochlore BRO7 and its composite catho de would be a potential candidate as a cathode material for LT SOFC application s

PAGE 45

45 Figure 3 1. Flow chart of g lycine n itrate c ombustion (GNC) route for BRO7 synthesis

PAGE 46

46 Figure 3 2. Effect of g lycine to n itrates ratio (G/N) on the cry stallinity of as synthesized BRO7 precursors Figure 3 3 Evolution of high purity BRO7 as a function of calcining temperature

PAGE 47

47 Figure 3 4. TEM images of B RO7 nanoparticles and aggregate s of those taken from two different spots Figure 3 5 (Left) SE M image of BRO7 aggregates calcined at 700 C for 2hrs and (Right) Energy Dispersive Spectroscopy (EDS) analysis on the particle (red circle in the left figure )

PAGE 48

48 Figure 3 6 Particle size distribution of BRO7 particles measured by TSI PSD 3603 ( Aerosizer ) Figure 3 7 Thermogravimetric (TG) plot to show the stability of the final BRO7 calcined 700C 2hrs powder at temperature range of 25~1000C

PAGE 49

49 Figure 3 8 a) Nyquist and b) Bode plots as a function of temperature for GNC BRO7 SS ESB composite cathode tes ted in air

PAGE 50

50 Figure 3 9 Overlapped Bode plots measured at 600C of BRO7 ESB composite cathode s where the particle size of SS ESB is fixed whereas different processing routes were used for BRO7, resulting in different particle sizes ( VM: vibro milled, S up: sonication and sediment ation as described in ref. 10 ) The data other than the one for GNC BRO7 SS ESB was extracted from ref. 10

PAGE 51

51 Figure 3 10 Arrhenius plo t of ASR for comparison of GNC BRO7 SS ESB with the other BRO7 ESB composite cathodes reported (S S: prepared by solid state reaction route (~micron size), VM: SS+vibro milling (700~800nm), Sup:SS+VM+sonication+sedimentation ( <100nm) as described in ref.10 )

PAGE 52

52 Table 3 1. Crystallite sizes of pure BRO7 depending on 2 using Scherrer s equation and its average 2 FWHM hkl D (nm) Strain (%) 30.239 0.368 222 31 0.58 35.059 0.460 400 23 0.62 50.280 0.457 440 25 0.42 59.713 0.493 622 23 0.37 62.643 0.472 444 25 0.33 73.756 0.642 800 18 0.37 Average 24.13 95% co nfidence interval 20.01< 24.13 <28.26 Table 3 2. ASR values of all BRO7 ESB composite cathodes compared at each temperature measured ASR( cm 2 ) at each temperature Cathodes 700C 65 0C 625 C 6 00C 55 0C 50 0C SS BRO7 SS ESB 10 1.29 2.00 4.27 11.55 VM BRO7 SS ESB 10 0.52 0.72 1.57 4.28 GNC BRO7 SS ESB (in this study) 0.123 0.21 0. 46 1.32 4.59 Sup BRO7 SS ESB 10 0. 23 0. 35 0. 92 2 80

PAGE 53

53 CHAPTER 4 SYNTHESIS OF IN SITU COMPOSITE CATHO DES OF BISMUTH RUTHE NATE AND STABILIZED BISMUTH OXIDE FOR LT SOFC 4.1 Introduction Low temperature ( LT ) s olid o xide f uel c ell (SOFC) operating at 65 0C and below can reduce the cost of production by allowing cheaper and wider selection of materials such as stainless steel for interconnects an d other structural components. 76 However, since decrease of operating temperatures usually causes an overall degr adation of cell performance, high performance cell components are required 77 As t he substantial reduction in the electrolyte ohmic resistance has been achieved by employing either thin film technology or electrolyte materials with higher ionic conductivity such as gadolinia doped ceria (GDC), the main focus of research is now primarily geared to wards redu ction of the electrode polarization, in parti cular, the oxygen reductio n reaction in the cathode. 78 79 I t has proven that a composite cathode, consisting of an electro catalyst with high electronic conductivity and an oxygen ion conducting phase, can decrease the cathode polarization substantially by extending t riple phase boundary (TPB) reaction sites from the electrode/electrolyte interface into the electrode Moreover, adding ionic conductivity to the electrode provides additional path for transport of oxyg en species to the cathode/electrolyte interface 6 7 F or example it was recently reported from our group that ASR values as low as 0.03 1 1 at 700C and 500C, respe ctively, can be achieved for a optimized bismuth ruthenate ( Bi 2 Ru 2 O 7 or BRO7 ) erbia stabilized bismuth oxide ( Bi 1.6 Er 0.4 O 3 or ESB ) composite cathode where very fine particle s of BRO7 w ere prepared by convention al solid state (SS) reaction route followed by multiple extra processing steps such as vibro milling, sonica tion, and sedimentatio n 10

PAGE 54

54 However, it should be pointed out tha t the yield ratio to get those fine particles (<100nm) of BRO7 was prohibitively low. On the other hand, a gglomeration of each constituting phase is often observed with the conventionally made composite cathode s during mixing and sintering process, which i s deleterious to the overall cathode performance since the number of reaction sites will be reduced from the agglomeration of each phase. 10 In addition, percolation of each constituting phase is one of the most important issues to be considered when a composite cathode is made since only the effective oxygen ions transported to the interface between the electrolyte and electrode can contribute to the overall cathode performa nce. 32, 75 One way to resolve the issues of both low yield of BRO7 nanoparticles from extra processing steps and agglomeration and percolation of each BRO7 and ESB phase mentioned above is proposed in this study, c alled in situ composite synthesis, a new approach to synthesize a composite cathode through wet chemic al based in situ synthesis process The con cept is that nitrate solution of targeted cations (Bi, Ru) is employed to synthesize nanoparticles of one phase (BRO7 in this study) in situ on surface of the other phase (ESB in this study) while the other phase of powder (ESB) is dispersed and stabilized in aqueous solution through repulsive surface charge of the powder generated by adding a dispersant and contro lling pH of the solution. Dispersing an d stabilizing oxide materials in aqueous solution has been extensively studied in tape casting industry for various applications such as multil ayer ceramic capacitors (MLCCs), v aristors, inductors, resistors, and cera mic substrates etc. 80 85

PAGE 55

55 As a result, f ormation of BRO7 nanoparticles will be ensured because wet chemical route is employed (refer to chapter 3) Moreover, since nanoparticles of BRO7 are formed o n surface of ESB particles, agglomeration of BRO7 nanoparticles w ill be prevented. Similarly, agglomeration of ESB particles is also expected to be suppressed by addition of a dispersant due to electrostat ic repulsion between functional anions adsorbed on ESB particles in the solution state 86 87 and by formed BRO7 nanoparticles in the final dry state as well. Therefore, in this work, synthesis of in situ composite cathodes of ESB BRO7 was tried to get better percolation and less agg lomeration of each constituting phase in the cathodes. The processing cost was also expected to be reduced since separate process es for each phase w ere not required, which normally happens for conventional composite synthesis route. D ispersion behavior of ESB powder with/without a dispersant (ammonium citrate in this study) was carefully examined through zeta potential measurement to stabilize ESB powder in the aqu eous solution. Then, (Ru, Bi) nitrates were added to the solution for final in situ composite cathodes of BRO7 ESB using knowledge of wet chemical synthesis for BRO7 nanoparticle s (refer to chapter 3) E lectrochemical performance of such prepared composite cathodes was compared to t hose of the conventionally made BRO7 ESB cathodes reported previou sly from our group to examine the feasibility of in situ composite cathode synthesis route Better percolation and less agglomeration of each BRO7 and ESB phase were expected to improve electrochemical performance

PAGE 56

56 4.2. E xperimental 4.2.1. Stabilization of ESB Suspension A p arameter for characterizing degree of repulsive force thus the stability of a suspension, is the zeta potential. In colloid scie nce, a suspension having a zeta potential >I30mVI is considered to have suffici ent repulsive force between p arti cles to inhibit agglomeration. 88 Experimental zeta potential measurement of suspensions is most commonly carried out using electrophoretic light scattering (ELS) tec hnique. 89 In this study, s urface charge, thus zeta potential, of ESB was first carefully examined to find the optimum pH of the solution and the amount of a dispersant to be added for exceeding the zeta potential requirement of >I30mVI thus stabiliz ing ESB powder in the solution. To control pH of the solution, titration solution such as HNO 3 and NH 4 OH were employed. Also ammon ium citrate dibasic (C 6 H 14 N 2 O 7 Fisher Scientific) was used as an anionic dispersant Figure 4 1 shows the molecular structure of it. 4.2.2. Synthesis Process The overall process of in situ composite synthesis is outlined in Figure 4 2 for better understanding First, ESB powder was obtained by a conventional solid state (SS) reaction and then crushed by mortar and pestle and sieved (325 Mesh). Such prepared ESB powder was t hen stabilized in de ionized ( DI ) water based on zeta potential measurement by controlling pH an d adding ammonium citrate as a dispersant. Once stabilized ESB suspension was prepared stoichiometric amounts of Ru((NO)(NO 3 ) 3 (Ru 31.3% min Alfa Aesar) and Bi(NO 3 ) 3 2 O (Puratronic 99.999%, Alfa Aesar) for the final Bi 2 Ru 2 O 7 were separately dissolved in DI water and added into the ESB suspension. Cations such as Bi 3+ and Ru 4+ were expected to adsorb onto negatively charged functional anions of the dispersant (COO or O ) which are also

PAGE 57

57 adsorbed onto surface of ESB particles in suspension due to specifi c adsorption of anions to the surface of ESB particles. The whole so lution was then stirred for 30 min on a hot plate at 80C for better mixing. After then, the hot plate was heated to 120C for water evaporation. Once water is evaporated the solution bec omes viscous gel consisting of ESB particles surrounded by (Ru, Bi) oxides. Upon subsequent heating of the hot plate up to 250~300C weak self auto ignition occurs, producing precursor s of in situ synth esized BRO7 ESB composite s As prepared precursors we re then ball milled in ethanol for 24 hrs using YSZ grinding media and dried on a hot plate with stirring to break up any aggregates inside the precursor s before calcin ation The dried BRO7 ESB composite precursors were calcined at 700~900C for 2hr s to re move any residual carbonates and unreacted phases resulting in in situ BRO7 ESB powder On th e other hand, ESB pellets were fabricated from SS ESB powder by uniaxial pressing (approximately 12 Kpsi) in a half inch cylindrical die followed by isostatic pr essing (250 MPa). These green bodies were then fired at 890 C for 15 hrs producing the final pellets with 1.11 cm 0.02 cm in diameter and 0.3 cm 0.01 cm in thickness. Cathode inks were prepared by combining organic vehicles with in situ BRO7 ESB powde r in appropriate ratio for 35~40vol% of porosity in the final cathode after sintering. The inks were applied to both sides of the ESB electrolyte substrates by screen printing to make symmetrical cells. These cells were dried at 120C for 1 h r and sinter ed at 800C for 2 h rs Pure BRO7 powder was separately mixed with organic vehicles and applied on top of the cathodes as a current collector. The active cathode area where the cathode ink was applied (the top or bottom surface area of the electrolyte), wa s 0.97cm 2

PAGE 58

58 4.2.3. Characterization Zeta potential was measured using ZetaPlus from Brookhaven Instruments and analyzed with PALS Zeta potential analyzer software ver.3.16. Smoluchowski model was applied to get values of the zeta potential since ka>>1 for ESB powder in DI water 90 X r ay d iffra ction (XRD) patterns of calcined composite powders were recorded with a Philips APD 3720 diffractometer using (Cu) = 1.5406 to determine phase f ormation purity, and crystallinity Step scans were taken over a range of 2 angles, 20~80, with 0 02 step scan and 0.5sec/step Evaluation of particle size and microstructure of the cath ode systems was carried out with the aid of a JE OL JSM 6400 scanning electron microscope (SEM). I mpedance testing on symmetrical cells consisting of in situ BRO7 ESB composite cathode s on ESB electrolyte was performed using a Solartron 1470E frequency resp onse analyzer at open circuit potential over a frequency range of 0.01Hz to 0.1MHz with AC voltage amplitude of 50mV at te mperatures of 550~700C in air. 4.3. R esults and Discussion 4.3.1. Zeta Potential of ESB Suspension Figure 4 3 shows zeta potential pr ofiles of ESB powder suspension s as a function of pH with and without a dispersant ( ammonium citrate dibasic) For ESB suspension without ammonium citrate, t he isoelectric point (IEP) point of pH where zeta potential becomes zero, was determined to be ~ p H 3.1, which is very close to that obtained for Bi 2 O 3 from the literature. 87 Since Bi 2 O 3 is a base material for ESB and valence of Er is the same as that of Bi as +3, close values of IEP betwe en ESB and Bi 2 O 3 appear reasonable. Zeta potentials of ESB suspension without ammonium citrate show

PAGE 59

59 relative ly low (
PAGE 60

60 potential of ESB suspension at this pH showed the maximum negative value (about 50mV ), which is enough for effectively stabilizing ESB suspension Ho wever, s ince pH of the suspension during in situ composite synthesis is fluctuating due to the equilibrium reaction between H + /OH ions and metal oxides hydroxyl group, pH range of 5~7 was tried to be maintained by adding titration solution such as HNO 3 a nd NH 4 OH. 4.3.2. Phase Purity of BRO7 ESB Composite and Formation of BRO7 Nanoparticles As can be seen in Figure 4 4 BRO7 ESB composites prepared by in situ synthesis showed some extra peaks in XRD patterns other than peaks from each pure phase up to 800 C of calcining temperature However, impurity phases wer e totally removed with calcin ation of 900 C for 2hrs. When compared to overlap ped patterns separately obtained from each of pure ESB and BRO phas e (top of the figure), XRD pattern of in situ composite s calcined at 900 C for 2hrs w as matched well with those, indicating that BRO7 ESB composite phase was rea lly formed by in situ synthesis. No reaction between BRO7 and ESB phase was observed for in situ composites. In addition, Figure 4 5 is a SEM image of final ESB BRO7 composites calcined at 900 C for 2hrs. It clearly shows that BRO7 nanoparticles are formed on surface of ESB particles and each phase is well connected, which is beneficial for transport ing of conduct ive species such as electrons and oxygen ions in composite cathodes. 4.3.3. Electrochemical Performance E lectrochemical performance of in situ BRO7 ESB composite cathode s was expected to be superior to conventionally made composite cathodes using mechanical mixing due to better percolation and l es s agglomeration of each phase It should be emphasized that even though there could exist many of TPB reaction sites in the

PAGE 61

61 c athode, only conducti ng species which tr ansport to the interface between the electrolyte and cathode can contribute to the overal l performance. 75, 79 Figure 4 6 shows Nyquist and Bode plots of an in situ BRO7 ESB composite cathode ASR increases as temperature goes down since the oxygen reduction reaction i s thermally activated process. 74 The Nyquist plot (Figure 4 6 a ) was R el corrected for easy comparison. That is, the high frequency real axis intercept, R el was subtracted from the real component of each data point in spectra. In this way, the point of low frequency intersecting re al axis directly corresponds to the cathode polarization, R c From the Bode plot (Figure 4 6 b) t he dominant electrochemical process seemed to be changed from 10~1kHz at >65 0C to <10Hz at <65 0C Frequency range of 1~10Hz is typically assumed to be relat ed to oxygen adsorption (1~10Hz). 55 A rea specific resistance (ASR) of the cathodes can eas ily be calculated as follows: ( 4 2) where the cathode area is divided by 2 to account for the symmetrical cell. In Table 4 1 calculated ASR values of in situ BRO ESB composite cathodes and four other BRO7 ESB composite cathode s reported from our group, which were made by conventional mixing, 10 are shown to compare the performance of those measured at temperatures of 500~700C For detailed expla nation about preparation method s of compared BRO7 ESB composite cathodes, refer to chapter 3. Even though it w as not substantial as expected, performance improvement was observed for in situ composite cathode compared to all other BRO7 ESB composite cathod es except Sup BRO7 SS ESB In particular, since the particle size combination is very similar for GNC BRO7 SS ESB and in situ BRO7 ESB, p erf ormance improvement

PAGE 62

62 was believed to be caused by better percolation and less agglomeration of in situ composite cath odes. However, ASR of in situ BRO7 ESB cathodes w as still higher than that of S up BRO7 SS ESB To further examine whether the dominating process of in situ composite cathodes is different from those of conventionally made composite cathodes, the Bode plots of all the cathodes compared at 600C were overlapped as shown in Figure 4 7 The characteristic frequencies of VM BRO7 SS ESB, GNC BRO7 SS ESB In situ BRO7 ESB, and Sup BRO7 SS ESB were almost the same, indicating that the dominant electrochemical proce ss was not that changed Even though better percolation and less agglomeration were expected to expedite the transport of oxygen ions through the cathode, degree of improvement seemed not sufficient enough to affect the dominant process at practical level. On the other hand, Figure 4 8 is the Arrhenius plot of A SR for direct comparison among all the BRO7 ESB composite cathode s prepared in this study and reported in the literature 10 The activation energy from this plot was calculated using the equation as below: (4 3) The activation energy of the in situ composite cathode was determined to be ~1.24 eV, which is comparable to the previously reported values, 1.2~1.34eV 9 4.4. Conclusion I n situ BRO7 ESB composite cathode w as successfully developed by wet chemical based in situ synthesis of BRO7 nanoparticles inside the stabilized ESB suspension. Pure BRO7 ESB composites were obtained after calcin ation of 900 C for

PAGE 63

63 2hrs. Since ESB particles were dispersed and stabilized through electrost atic repulsion between functional anions adsorbed on ESB particles generated by adding a dispersant and controlling pH of t he solution BRO7 nanoparticle s were formed on surface of ESB particle s As a result, less agglomeration and better percolation of each BRO7 and ESB phase were achieved for the final product The processing cost was also reduced since separate process es fo r each phase were not required, which normally happens for conventional composite synthesis route. Electrochemical performance was improved for in situ BRO7 ESB composite cat hodes compared to conventionally mixed BRO7 ESB composite cathodes except Sup BRO7 SS ESB (refer to chapter 3) ASR values of 0.097 1 and 3.58 1 were obtained at 700C and 500C, respectively. Consequently, the feasibility of in situ composite synthesis route was successfully demonstrated Further optimization of dispersing and st abilizing strength of dispe rsants will make this sy nthesis route more attractive to the fabrication of LT SOFC cathodes.

PAGE 64

64 Figure 4 1 Molecular structure of ammonium citrate dibasic Figure 4 2 Flow chart of synthesis process for in situ ES B BRO7 composite cathodes

PAGE 65

65 Figure 4 3. Zeta potential of ESB powder suspension with or without a dispersant (ammonium citrate) as a function of pH

PAGE 66

66 Figure 4 4 XRD patterns of in situ ESB BRO7 composite cathodes depending on calcining temperature s and ov erlapped XRD patterns of each ESB and BRO7 phase synthesized by solid state (SS) reaction for comparison

PAGE 67

67 Figure 4 5 SEM image of in situ BRO7 ESB composites showing adsorbed BRO7 nano particles on ESB phase

PAGE 68

68 Figure 4 6 a) Nyquist and b) Bode plots o f in situ BRO7 ESB composite cathode tested at each temperature in air under open circuit potential

PAGE 69

69 Figure 4 7 Overlapped Bode plots measured at 600C of BRO7 ESB composite cathode s where the particle size of SS ESB is fixed whereas different processin g routes were used for BRO7, resulting in different particle sizes (VM: vibro milled, Sup: sonicated and sedimented as described in ref. 10. ) The data for SS BRO7 SS ESB, VM BRO7 SS ESB, and Sup BRO7 SS ESB was extracted from ref. 10 and the one for GNC BRO 7 SS ESB from chapter 3

PAGE 70

70 Figure 4 8 Arrhenius plot of ASR for comparison among five different cathode systems 10

PAGE 71

71 Table 4 1 ASR values of all BRO7 ESB composit e cathodes compared at each temperature measured ASR( cm 2 ) at each temperature Cathodes 700C 65 0C 625 C 6 00C 55 0C 50 0C SS BRO7 SS ESB 10 1.29 2.00 4.27 11.55 VM BRO7 SS ESB 10 0.52 0.72 1.57 4.28 GNC BRO7 SS ESB (Chapter 3) 0.123 0.21 0.46 1.32 4.59 In situ BRO7 ESB (in this study) 0.097 0.18 0.38 1.02 3.58 Sup BRO7 SS ESB 10 0. 23 0. 35 0. 92 2 80

PAGE 72

72 CHAPTER 5 PYROCHLORE BISMUTH RUTHENATE B i 2 R u 2 O 7 INFILTRATED CATHODES FOR HIGH PERFORMANCE L T SOFCS 5 .1. Introduction Low temperature ( L T) s olid o xide f uel c e ll (SOFC) operating at 650 C an d below can reduce the cost of production by allowing cheaper and wider selection of materials such as stainless steel for interconnect s and other structural component s 76 Also, improved system stability and dura bility are additional benefits. However, since decreasing the operating temperatures usually leads to an overall degradation of cell performance, development of cell components with higher performance is needed 77 Thanks to the substantial reduction in the electrolyte ohm ic resistance by employing either thin film technology or electrolyte materials with higher ionic conductivity such as gadolinia doped ceria (GDC) the focus of improving overall performance is now mainly on finding means to reduce the electrode polarizati on, in particular, the oxygen reduction reaction in the cathode 3, 78 79 The c oncept of a composite cathode consisting of an ionic conducting phase and an electronic conductin g phase is theoretically proposed to be a highly effective way of improving cathode performance by exten ding the active t riple p hase b oundary (TPB) region from the electrolyte/electrode interface into the electrode bulk and has been experimentally demonstrated by many researchers 6, 33, 40, 95 F or cathode s for example, a composite cathode consisting of strontium doped LaMnO 3 (LSM) and yttrium stabilized zirconia (YSZ) showed improved a rea s pecific r esistance (ASR) reducing the original ASR of 2 obtained with a pure LSM cathode on a YSZ electrolyte to 2 2 with GDC substituted for YSZ at the same temperature. It was concluded by authors that the higher ionic

PAGE 73

73 conductivity as well as the h igher oxygen surface exchange rate of GDC over YSZ are the main reason for the improved performance 6 In addition, there have been many studies on optimizing the micro structure of cathodes since electrochemical performance strongly depends on the microstructur al parameters such as the number of TPB reaction sites, percolation of each phase constituting the cathode, porosity for gas diffusion 32, 34 36, 96 Relatively recently, a more powerful processing approach to develop the novel cathode structure was established, called infiltration or impregnation 41 42, 45 46, 48 49, 97 98 The i nfiltration technique involves depositing nanoparticles of electronic, ionic conducting phase, or sometimes metal electrocatalyst into a preformed scaffold. Since the sintering of scaffold is separately carried out at high temperature for ensuring good adh esion between scaffold and electrolyte, good connection between particles to achieve the effective electronic and/or ionic conduction, and structural stability of the cathode, the infiltration of nanoparticles and the formation of desired phase can be perf ormed at relatively low temperature 56 Such prepared nanostructure allows for high catalytic activity and enlarged TPB length so that overall performance enhancement is expected. For example, Jiang et al. showed th at GDC infiltrated LSM cathode co uld reduce its ASR down to 0.21 2 at 700C when 5.8 mgcm 2 GDC was loaded, about f ive times lower than that (1.06 2 ) of the conventionally made GDC LSM composite cathode mentioned above 47 On the other hand, a composite cathode consisting of erbia stabilized bismuth oxide (ESB) as an ionic conducting phase and pyrochlore bismuth ruthenate (Bi 2 Ru 2 O 7 or BRO7 shortly) as an electronic conducting phase was demonstrated from our group to be o ne of the most promising cathode materials due to the high ionic conductivity

PAGE 74

74 (0.023 Scm 1 at 773K and 0.37 Scm 1 at 973K ) 99 of ESB and hig h electronic conductivity (~300 Scm 1 at room temperature up to 1173K ) 25 of BRO7 phase 9 10, 23 Moreover, the oxy gen surface exchange rate of stabilized bismuth oxide is the factor of 10 3 larger than that of YSZ 100 101 An ASR as low as 0.03 Scm 1 at 700C was achieved for conventionally made BRO7 ESB composite cathode with the optimized microstructure and particle size combination though the yield ratio to get those f ine particles (<100nm) of BRO7 wa s prohibitively low thus not suitable for mass production. 10 Thus, infiltrating BRO7 nanoparticles on ESB scaffold s can be an alternative effective way of fabricating BRO7 ESB composite cathode s The improved electrochemical performance is expected as long as the infilt rated BRO7 phase forms sufficient nano network s on the surface of ESB to provide enough electron conducting paths through the cathode layer. Furthermore, less agglomeration of each phase, often obser ved with the conventionally ma de composite cathode, and b etter adhesion between the electrolyte and c athode layers will be additional advantage for these infiltrated cathodes Based on knowledge of infiltration and wet chemical synthesis route BRO7 infiltrated ESB composite cathodes were developed in this study 5 .2. E xperimental 5.2.1. Infiltration Process Schematic of infiltration process is shown in Figure 5 1 for better understanding. ESB powder was obtained by a conventional solid state (SS) reaction route and then crushed by mortar and pestle and sieved (3 25 Mesh). ESB pellets were fabricated from these powders by uniaxial pressing (approximately 12 Kpsi) in a half inch cylindrical die Next, t hese gre en bodies were first weakly fired at 70 0 C for 4 hrs before depositing ESB scaffold s on top of those Then ESB scaffold s w ere constructed by screen printing

PAGE 75

75 ESB slurry mixed with 30 vol% graphite ( Catalog number G67 500, Fisher Scientific ) as a pore former on both sides of the weakly fired ESB pellets to give an appropriate porosity The pellets were then fire d at 890 C for 14 hrs Glycine added nitrate solution s for infiltration were prepared by the same method described in chapter 3 for GNC BRO7. Such prepared GNC BRO7 solution was stirred for 1 hr on a hot plate at 80C for better mixing and heated up to 120 C for water evaporation until a concentrat ed solution (about 0. 4 M) was obtained. The solution was then infiltrated onto both sides of ESB scaffold s using a vacuum apparatus for better infiltration by capillary force. The cells with the precursors were ove n dried for 20 min at 70 C before further heat treatment. The symmetric al cells were then heat treated with different temperature profiles (sample #1 and #2) as shown in Figure 5 2 to form BRO7 nanoparticles and also to remove any residual carbonates and u nreacted phases. The mass difference of symmetric cells before and after infiltration was measured to estimate the infiltrated BRO7 phase loading. Lastly, t hese cells were dried at 120C for 1 h r and sintered at 800C for 2 h rs Pure BRO7 powder was separ ately mixed with organic vehicles and applied on top of the cathodes as a current collector. The active cathode area where the cathode ink was applied (the top or bottom surface area of the electrolyte), was 0.97cm 2 5.2.2. Characterization Evaluation of particle size and microstructure of the cathode system s was carried out with the aid of a JEOL JSM 6400 scanning electron mi croscope (SEM) Further compositional analysis of infiltrated BRO7 particles was performed by a JEOL TEM 200CX transmission electro n microscope (TEM) combined with e nergy d ispersive X ray s pectrometry (EDS) on a sli ced cathode sample prepared by f ocused i on b eam (FIB). In

PAGE 76

76 addition, Amira image reconstruction software was utilized to calculate the average porosity of ESB scaffolds. Fin ally impedance test ing of BRO7 infiltrated cathodes on ESB pellet s was performed using a Solartron 1470E frequency response analyzer at open circuit potential over a frequency range of 0. 01 Hz to 0. 1M Hz with AC voltage amplitude of 50 mV at temperatures of 5 5 0~700C in air. Impedance set up is visualized in Figure 5 3 for better understanding. 5 .3. R esults and Discussion 5 .3.1. Cathode Microstructure and the Effect of Heating Condition Figure 5 4 shows typical cross sectional SEM images of BRO7 infiltrated ESB cathodes. ESB scaffold s and BRO7 nanoparticles are clear ly seen in the image s ESB scaffold s w ere well connected and good adhesion between scaffold and the elect rolyte was obtained for both sample #1 and #2 However the distribution of BRO7 particles turned out to be quite different for each sample For sample #1, m ostly the top half layer of ESB scaffold s was infiltrated and the deposited particles are scattered rather than continuous whereas more continuous and deeper infiltration through ESB scaffol d s was observed for sample #2. Gradual heating from room temperature with slow heating rate (sample #1) seemed to be an in effective way to penetrate the BRO7 nitrate solution into the scaffold because the solution returns to the external surface of the cat hode where it remains after firing 102 On the other hand, the auto ig nition for BRO7 phase formation takes place right away for sample #2 when the sample is inserted into the furnace already preheated to 300C. From our previous study on BRO 7 nanoparticle synthesis using GNC (chapter 3) it was found that if not enough heat (below 300C) is provided to viscous g lycine n itrate solution the auto ignition reaction is slugg ish wh en it starts self auto ignition thus resulting in formation of less pure and larger BRO7 particles. Also,

PAGE 77

77 low temperature heating affected degree of t he agglomeration of synthesized BRO7 powder because less gas byproduct was produced during the reaction for preventing the agglomeration of particles Consequently, the heating condition was concluded to affect degree of the infiltration (penetration depth distribution, and size of BRO7 nanoparticles etc). Magnified SEM image clearly showed BRO7 nanoparticles deposited on the surface of the ESB scaffold with size range of 100~200nm in diameter (Figure 5 5 ). The way of area selection for porosity calculati on and its average porosity value is shown in Figure 5 6 The average porosity was estimated to be ~ 23.7vol%, which is a little bit lower than expected because 30vol% of graphite was used as a pore former to construct ESB scaffolds. S i nce only bright regio n in the middle of image s was selected for porosity calculation there could be some errors involved depending on how to select the area. The loading of infiltrated BRO7 was determined by mass change before and after infiltration to be about 4.59 wt% (1.0 mg cm 2 ) and 5.46wt% (1.2 mgcm 2 ) relative to ESB scaffold s for sample #1 and #2, respectively, with 2 min of vacuum assisted nitrate solution infiltration. The infiltrated amo unt with vacuum assistance seem ed higher than that of drop coating by hand reported in the literature For example, it was rep orted that GDC loading was 0.72 mgcm 2 after one infiltration, increased to 1.68 mgcm 2 after twice, and to 5.8mgcm 2 after six times of repeated infiltration 47 Thus, it is believed that v ac uum assisted infiltration is a more powerful method to deposit materials than drop coating by hand. However, it should be pointed out that t h e BRO7 loading can be further optimized with repeated infiltration as long as the appropriate porosity for gas

PAGE 78

78 tr ansport is maintained. There should be the optimized loading at which the deposited nanoparticles provide both ionic and electronic conducting pathways effectively throughout the cathode. 5 .3.2. Compositional Analysis Compositional analysis was performed b y TEM with EDS on both ESB scaffold s and BRO7 nanoparticles to confirm the formation of BRO7 phase (Figure 5 7 ). While no Ru ele ment was detected on ESB scaffold s it was detected on BRO7 nanoparticles as shown in the bottom of the image ( Figure 5 7 b), cl early indicating that BRO7 formed on the surface of ESB scaffold. 5 .3.3. The Effect of Infiltration on Cathode Polarization Figure 5 8 shows the Nyq ui st plot of BRO7 infiltrated ESB cathode #1 at differen t temperatures (5 5 0~700C) At 700C it seems that multiple arcs are involved in the resulting shape of spectrum. In other words, the electrochemical processes are uniformly distributed over a wide range of frequenc ies (0.1Hz~1kHz). That is, all the sub reaction steps such as oxygen adsorption and dissocia tion, diffusion of oxygen ion, and ion incorporation are playing roles to certain degree in the overall oxygen reduction reaction. However, as temperature decreases, processes more related to low frequency range (0.1~10Hz) start to dominate th e overall pro cess as shown at 5 5 0 C which is typically assumed to be related to oxygen adsorption (1~10Hz) and gas diffusion through the pores (<1Hz) 55 ASR was calculated from impedance spectra at each temperature as follows: (5 1)

PAGE 79

79 For sample #1, an ASR value of 0.09 6 cm 2 was obtained at 700 C and 3.5 5 cm 2 at 500 C This is about 22% lower compared to that of conventionally mixed GNC BRO7 SS ESB composite cathode (ch apter 3). On the other hand, Figure 5 9 shows the Nyqu i st plot of BRO7 infiltrated ESB cathode #2 at different temperatures (5 5 0~700C) Even though t h e frequency distribution seemed to be similar to that of sample #1, the shape of spectrum looked differe nt The calculated ASR values were 0.073 cm 2 and 1.82 cm 2 at 700 C and 500 C respectively. This is the marked improvement compared to GNC BRO7 SS ESB. The overall performance improvement is better visualized in Figure 5 10 through overlapped Nyquist plots of all three composite cathodes compare d. About 41~68% of performance improvement was achieved depending on the temperatures However, since different degree of improvement was obtained at each temperature, it was inferred that different dominating process could be involved for each sample at e ach temperature. To further clarify this observation, three Bode plot s of GNC BRO7 SS ESB, sample #1, and sample#2 w ere overlapped to elucidate the dominating process for each cathode at each tem perature as shown in Figure 5 11 The most significant point from the figure is that the electrochemical process related to 10Hz~1kHz frequency range was dramatically reduced for sample #2 regardless of temperatures so that the dominating process is now changed to 1~10Hz frequency range for sample #2 Since the mai n difference between sample #1 and #2 is the degree of penetration and continuity of BRO7 nanoparticles through the cathode la yer, more TPB reaction sites

PAGE 80

80 are assumed to be formed and a more continuous electronic pathway is expected to be established deep down to the scaffold s /electrolyte interface for sample #2. Typically, the electrochemical process in the 10Hz~1kHz frequency range is attributed to diffusion of oxygen ion ( ) towards TPB reaction sites 55, 79 Since mainly the surface layer of ESB scaffold was infiltrated for sample #1, diffusion of available adsorbed oxygen and oxygen ion are limited, thus oxygen incorporation takes place primarily at the surface of the ESB scaffold. On the other hand, there exist more BRO7 nanoparticles distributed continuously throughout the scaffold s for sample #2. Therefore, the number of TPB sites is likely to dramatically increase throughout the cathode layer, resulting in much easi er diffusion of adsorbed oxygen and oxygen ion towards TPB sites, which is observed as the dramatic reduction of cathode polarization in 10Hz~1kHz frequency range. In other words, the effective TPB length is extended. The effect of infiltration on the ove rall performance becomes more substantial as temperature goes down, suggesting that different activation energies are involved. The activation energy for each sample was calculated using the equation as below: (5 2) The resulting values are shown in Table 5 1 It is clearly seen that the activation energy was indeed reduced for the infiltrated cathode with the optimized heating condition (sample #2). The more reduced cathode polarization at lower temp eratures demonstrates significance of both the effective TPB reaction sites and high ionic conductivity / oxygen surface exchange rate of the ionic conductive phase in the composite cathodes.

PAGE 81

81 5 .3.4. Performance Comparison To compare the performance of these cathodes with the others from literature s an Arrhenius plot of ASR for cathodes used in this study along with several other infiltrated cathodes in addition to conventionally mixed LSM YSZ and LSM GDC was drawn in Figure 5 12 As can be seen, the optimiz ed BRO7 infiltrated ESB cathode (sample #2) shows the best performance among all the cathodes compared. 103 Considering that YSB infiltrated LSM cathode also employs a stabilized bismuth ox ide (YS B) as an ionic conducting phase which has similar ionic conductivity and oxygen surface exchange rate to ESB, BRO7 phase is also assumed t o play an important role in the substantial reduction of cathode polarization. However, it should be noticed that the LSM phase rather than YSB was employed as the scaffold s for their system Thus, there is the possibility for further performance enhancement with infiltrated LSM YSB cathode if YSB is used as the scaffold s and LSM is infiltrat e d. This strategy is expected to work well since the electrochemical process in 10Hz~1kHz frequency range seems to be dominant for infiltrated LSM YSB as shown in Figure 5 13 indicating that diffusion of oxygen ion ( ) towards TPB reaction s ites is limited, which was considerably improved in ou r BRO7 infiltrated ESB cathodes (sample #1 and #2) In Figure 5 14 the performance of BRO7 infiltrated ESB cathodes is further compared t o the other conventionally m ade BRO7 ESB composite cathodes repo rted previously from our group 10 For detailed explanation about the preparation method of each composite cathode, refer to chapter 3. The optimized BRO7 infiltrated ESB ca thode (sample #2) exhibits the second best performance among all the BRO7 ESB composite cathodes shown However, since

PAGE 82

82 Sup BRO7 S up ESB composite cat hode showing the lowest ASR was prepared by conventional solid state reaction followed by multiple extra pr ocessing steps such as vibro milling sonication and sedimentation, the yield ratio to get the final product is prohibitively low yield thus not suitable for mass production Consequently from the realistic point of view, the processing route used to sy nthesize Sup BRO7 S up ESB is not appropriate for fabricat ion of high performance cathode s As a summary, ASR values of all BRO7 ESB composite cathodes are liested in Table 5 2 to show the exact values obtained. 5 .4. Conclusion A high performance composite cathode was developed by infiltrating a (Bi,Ru) nitrate solution with g lycine onto preformed ESB scaffold s It was found that care must be taken for infiltration process since different heating condition s can lead to different degree s of infiltration (pene tration depth, distribution, and size of BRO7 nanoparticles etc). The optimized heating condition enabled the formation of continuous BRO7 nanoparticles through the scaffold s resulting in substantial extension of the effective TPB reaction sites. Infiltra ted BRO7 nanoparticles substantially accelerated the cathode reaction, especially diffusion of adsorbed oxygen and oxygen ion s towards TPB reaction sites ASR values a s low as 0.073 cm 2 and 1.82 cm 2 were achieved with the optimized cathode at 700 C and 500 C respectively, which was the best performance among all the infiltrated cathodes compared and the second best among all BRO7 ESB composite cathodes compared. Furthermore, the activation energy was also reduced for the infiltrated cathode. In conclusion, it was demonstrated that the BRO7 infiltrated ESB cathode can be an excellent candidate as a cathode for LT SOFC.

PAGE 83

83 Figure 5 1 Schematic of infiltration process Figure 5 2 Temperature profiles of two different heat treatment conditions for sample # 1 and sample #2

PAGE 84

84 Figure 5 3 Schematic of impedance testing set up Figure 5 4 Cross sectional S EM images of a) sample #1 and b) sample #2

PAGE 85

85 Figure 5 5 SEM images of BRO7 nanoparti cles taken from sample #1 : the same spot with (a) low (b) high (zoomed in) magnifications

PAGE 86

86 Figure 5 6 a ) A SEM image of infiltrated BRO7 ESB composite cathodes taken from sample #1, showing the region (bright region in the middle of image) used for porosity calculation w here 30 vol% of graphite was used as a pore f ormer to construct ESB scaffolds and b) calculated average porosity and its standard deviation

PAGE 87

87 Figure 5 7 TEM images and EDS spectra of a) ESB scaffold and b) BRO7 nanoparticles with red circle on Ru element energy spectrum indicating the presence of R u element in BRO7 nanoparticles (taken from sample #1)

PAGE 88

88 Figure 5 8 Nyquist plots of BRO7 infiltrated ESB composite cathode (sample #1) at different temperatures (550~700 C)

PAGE 89

89 Figure 5 9 Nyquist plots of BRO7 infiltrated ESB composite cathode (samp le # 2) at different temperatures (550~700 C)

PAGE 90

90 Figure 5 10 Overlapped Nyquist plots of GNC BRO7 SS ESB, infiltrated BRO7 ESB sample #1, and sample #2 at different temperatures (550~700 C) to better show the improvement of performance

PAGE 91

91 Figure 5 11 O ver lapped Bode plots of BRO7 infiltrated ESB composite cathode s (sample #1 and # 2) and GNC BRO7 SS ESB composite cathode at different temperatures (550~700 C) for comparison on dominating processes

PAGE 92

92 Figure 5 12 Arrhenius plot of ASR for performance comp arison with several conventionally mixed composite cathodes and the other infiltrated cathodes reported. All the data other than BRO7 ESB composite cathode and infiltrated BRO7 cathodes (#1 and #2) in this study was extracted f rom the references cited in r ef 103

PAGE 93

93 Figure 5 13 O verlapped Bode plots of BRO7 infiltrated ESB cathode s (sample #1 and # 2 ), conventionally mixed BRO7 ESB composite cathode, and YSB infiltrated LSM cathodes at 700 C. Notice that imaginary impedance (Z ) is area normalized for direc t comparison. Data for YSB infiltra ted LSM was extracted from ref. 103

PAGE 94

94 Figure 5 14 Arrhenius plot of ASR for performance comparison among all conventionally mixed B RO7 ESB composite cathodes All the data other than GNC BRO7 ESB composite cathode and infiltrated BRO7 cathodes (#1 and #2) in this study was extracted from ref. 10

PAGE 95

95 Table 5 1 Activation energies of GNC BRO7 SS ESB (chapter 3) and infiltrated BRO7 cathodes (sample #1 and #2) calculated from Figure 5 12 using equation (5 2) Activation Energy (Ea) GNC BRO7 SS ESB (chapter 3) Infiltrated BRO7 #1 (in this study) Infiltrated BRO7 #2 (in this study) eV 1.26 1.24 1.09 kJmol 1 122 120 105 Table 5 2 ASR values of all BRO7 ESB composite cathodes compared at each temperature measu red ASR( cm 2 ) at each temperature Cathodes 700C 65 0C 625 C 6 00C 55 0C 50 0C SS BRO7 SS ESB 10 1.29 2.00 4.27 11.55 VM BRO7 VM ESB 10 0.61 0.90 2.09 5.89 VM BRO7 SS ESB 10 0.52 0.72 1.57 4.28 SS BRO7 VM ESB 10 0.49 0.63 1.34 4.35 GNC BRO7 SS ESB (chapter 3) 0.123 0.21 0.46 1.32 4.59 SS BRO7 Sup ESB 10 0.25 0.33 0.66 1.90 Sup BRO7 SS ESB 10 0.23 0.35 0.92 2.80 Sup BRO7 Sup ESB 10 0.10 0.14 0.31 1.22 Infiltrated BRO7 ESB #1 (in this study) 0.096 0.17 0.37 0.96 3.55 Infiltrated BRO7 ESB #2 (in this study) 0.07 3 0.12 0.20 0.42 1.82

PAGE 96

96 CHAPTER 6 SUMMARY S olid oxide fuel cell (SOFC) is a one of the most promising candidate s for futur e power generation due to its high efficie n cy and low pollution emission. However, for its successful commercialization the production cost should be reduced dramatically to compete with the other power generating systems. Reducing the operating temperatur es down to 500 ~650 C could be an option s o that the wider selection of cheaper materials such as stainless steel for interconnects and the other structural components will be realized. However, s ince reduction of operating temperatures generally causes performance degradation of the system due to its inherent characteristic of thermally activated process, development of high performance constituting components at low temperatures is necessary. With the inception of thin film electrolytes and electrolytes with ionic conductivities higher than conve ntional YSZ materials, much of research focus is now geared to wards the electrode (anode and cathode) development. Since th e oxygen reduction reaction becomes more of an issue than fuel oxidation reaction at low temperatures h ence it is essential to devel op a cathode with high catalytic activity towards oxygen reduct ion as well as a microstructure that maximizes the number of reaction sites ( T PBs) and faci litates oxygen transport toward s the electrolyte. It is also important that the cathode be chemically compatible with the electrolyte and mechanically stable with time so that performance does not degrade to a significant extent over the lifetime of the cell. In this dissertation, three different physicochemical approaches were attempted to develop a high performance composite cathode for LT SOFC applications. First, glycine nitrate c ombustion (GNC) synthesis route was thoroughly investigated to obtain

PAGE 97

97 pure high temperature pyrochlore Bi 2 Ru 2 O 7 (BRO7) E specially g lycine to n itrate ratio (G/N) was carefully examined to find the optimum ratio for better crystall inity and phase purity of BRO7. The optimum G/N ratio, which can generate the maximum reaction heat, was found to be about 0.6 based on the calculation above. Five different G/N ratios, 0.4, 0.6, 0.8, 1 .0, and 1.2, were chosen in this study for BRO7 synthesis to evaluate the effect of G/N ratio on the crystallinity and particle size of BRO7 precursor. BRO7 precursors become more crystalline from amorphous as G/N ratio is varied from 0.4 to 1.2 even thoug h the best crystallinity wa s expected to appear at G/N=0.6 resulting from the maximum reaction heat provided This could be attributed to the presence of abundant oxygen gas from the environment so that more reducing agent, i.e. glycine, is required to gen erate the maximum heat, which is in t urn evidenced as better crystal l inity for G/N ratios above 0.6. However, the best peak pattern similar to pure BRO7 was achieved at G/N=1.0 with less impurity phases such as unreacted Bi 2 O 3 Ru and/or RuO 2 G/ N ratio of 1.0 was selected for further study of phase evolution at different calcining conditions and microstructural analysis on the final powder. T he duration was fixed at 2hrs for this study while the calcining temperatures were varied at 600 900C with 100C interval. As calcining temperature arose, the crystallinity increased accordingly. Pure BRO7 phase was achieved even at 700C 2hrs, which is noticeable compared to 900C tens of hours heat treatment necessary for the formation of pure BRO7 with conventiona l SS route. Impurity phases such as unreacted Bi 2 O 3 was totally removed beyond 700C.

PAGE 98

98 For BRO7 powders calcined at 700C 2hrs t he average crystallite size of ~ 24 nm in diameter was achieved. However, it should be pointed out that the resulting particles t end to aggregate during calcining For better understanding on behavior of these aggregates, analysis on particle size distribution was performed. Most of agg regated particles exist in size range of 2 00~ 4 00nm in diameter However, this size range was still much smaller than t he conventionally made one. For example, the particle size of vibro milled (VM) SS BRO7 after 7 days of vibro milling was reduced from ~ 1.31 m of as prepared SS BRO7 to ~ 0.73 m which is still about two times bigger than nanoparticles o f GNC BRO7 and contains a small number fraction but significant volume fraction of large, unbroken particles On the other hand, energy dispersive s pectroscopy (EDS) analysis on nano particle s confirmed the composition of Bi 2 Ru 2 O 7 Composite cathodes of su ch prepared BRO7 with SS ESB showed better electrochemical performance than SS BRO7 SS ESB and VM BRO7 SS ESB as well ASR values of 0.12 3 2 at 700C and 4. 59 2 at 500C, respectively, w ere achieved, which follows well the trend of particle size effect on performance of composite cathodes. Additionally, the number of processing steps (thus time) was dramatically reduced by GNC route. In conclusion, i t wa s demonstrated that GNC can be an effective synthesis route to produce nanoparticles of pyrochlore BR O7 and its composite cathode would be a potential candidate as a cathode material for LT SOFC application. Secondly, a unique in situ composite synthesis route was developed and applied to synthesize a BRO7 ESB composite cathode with better percolation and less agglomeration of each phase through wet chemical based in situ synthesis process.

PAGE 99

99 The processing cost was also expected to be reduced since separate process es for each phase to ma ke a final composite cathode were not required, which normally happens for conventional composite synthesis route. Since ESB particles were dispersed and stabilized through electrosteric repulsion between functional anions adsorbed on ESB particles generated by adding a dispersant and controlling pH of the solution BRO7 nano particle s were formed on surface of ESB particle s As a result, less agglomeration and better percolation of each BRO7 and ESB phase were achieved for the final product The isoelectric point (IEP) of ESB suspension turned out to be about pH 3.1 However, as ammonium citrate s w ere added to the suspension, potential curves were shifted to the lower pH side and, more importantly, to the more negative values because the surface layer of ESB particles becomes more negative due to adsorption of functional anion s of the dispersant (COO or O ) Thus, suspensions can be stabilized from electrostatic repulsion between particles developed from overlap of the electric double layer in the solution. The IEP of ESB suspension s with ammonium citrate s was estimated to be ~ pH 1.9. Since the IEP of ESB suspension was changed before and after addition of ammonium citrate s there should be some interactions called specific adsorption, between ESB particles and dispersants 0. 5wt% (relative to ESB powder) ammonium citrate gave the suspension more negative zeta potential than 1 0wt% ammonium citrate due to its lower ionic strength of the solution with 0. 5wt%, thus extended As a result, stability of the particles in suspension w as enhanced with 0. 5wt% ammo nium citrate.

PAGE 100

100 pH range of 5 ~7 with 0. 5wt% of ammonium citrate was chosen as the optimized stabilizing condition for synthesis of in situ composite cathodes since the zeta potential of ESB suspension at this pH showed the maximum negative value (about 50mV ), which is enough for effectively stabilizing ESB suspension Final pure BRO7 ESB composites were obtained after calcin ation of 900 C for 2hrs. Electrochemical performance was improved for in situ BRO7 ESB composite cathodes compared to that of conv entionally made GNC BRO7 SS ESB composite cathodes. ASR values of 0.097 2 and 3.58 2 were obtained at 7 00C and 5 00C, respectively. T he activation energy of the in situ composite cathode was determined to be ~1.24 eV, which is comparable to the previou sly reported values, 1.2~1.34eV Consequently, the feasibility of in situ composite synthesis route was successfully demonstrated Further optimization of dispersing and stabilizing strength of dispersants will make this synthesis route more attractive to the fabrication of LT SOFC cathodes. Lastly the infiltration of BRO7 on ESB scaffold s was carried out to increase the electrochemical performance. Two different heating conditions were applied to deposit BR O7 nanoparticles on the surface of the ESB scaffo ld s. T he distribution of BRO7 nano particles turned out to be quite different for each sample For sample #1, mostly the top half layer of ESB scaffold s was infiltrated and the deposited particles are scattered rather than continuous whereas more continuous and deeper infiltration through ESB scaffold was observed for sample #2. Gradual heating from room temperature with slow heating rate (sample #1) seemed to be an in effective way to penetrate the BRO7 nitrate solution into the scaff old because the solution return ed to the external surface of the cathode where it remains after firing On the other hand, the

PAGE 101

101 auto ignition for BRO7 phase formation takes place right away for sample #2 when the sample is inserted into the furnace already pre heated to 300C. I t w as found that if not enough heat (below 300C) is provided to visc ous g lycine n itrate solution, the auto ignition reaction is slugg ish when it starts self auto ignition resulting in larger and less pure BRO7 particles. Also, low temperature heating affect ed degree of the agglomeration of synthesized BRO7 powder because less gas byproduct was produced during the reaction for preventing the agglomeration of particles Consequently, the heating condition was concluded to affect degree of the infiltration (pen etration depth, distribution, and size of BRO7 nanoparticles etc). Electrochemical performance of such prepared BRO7 ESB composite cathode was compared with the other infiltrated cathodes and also other conv entionally m ade BRO7 ESB composite cathodes repo rted in the literature s T he optimized BRO7 infiltrated ESB cathode (sample #2) show ed the second best performance among all BRO7 ESB composite cathodes compared It was believed that t he optimized heating condition enabled the formation of continuous BRO7 nanoparticles through ESB scaffold s resulting in substantial extension of the effective TPB reaction sites. The extended TPB substantially accelerated the cathode reaction, especially diffusion of adsorbed oxygen and oxygen ion. ASR values a s low as 0.07 2 and 1.82 2 were achieved with the optimized cathode at 700C and 500C, respectively. Furthermore, the activation energy was also reduced for the infiltrated cathode. In conclusion, it was demonstrated that the BRO7 infiltrated ESB cathode can be a n excellent candidate as a cathode for LT SOFC.

PAGE 102

102 APPENDIX A ISSUE ON SYNTHESIS OF PURE B i 2 R u 2 O 7 Throughout this dissertation pyrochlore bismuth ruthenate (Bi 2 Ru 2 O 7 or shortly BRO7) was used as one of the primary component s in composite cathodes for LT SO FC. Although satisfactory results were obtained in terms of electrochemical impedance test, phase purity of BRO7 was one of the main obstacles to be attacked many times because there were many impurity phases possible to be formed due t o incomplete reactio n among constituting elements i.e. Ru, Bi, and O. Sin ce BRO7 possesses a high temperature pyrochlore structure, the reaction temperature is one of the critical factors influencing degree of reaction, thus purity of BRO7 phase. For example, if the reaction temperature is not sufficient enough, Bi 3 Ru 3 O 11 (or Bi 2 Ru 2 O 7.3 ) phase appears instead. This is why many researcher s have chosen solid state reaction method for BRO7 synthesis which is high temperature heat treatment over 900 C for tens of hours. Differenc e betwe en phases is normally characterized from XRD patterns. As can be seen in Figure A 1, Bi 3 Ru 3 O 11 (or Bi 2 Ru 2 O 7.3 ) contains mor e number of diffraction peaks around 30 while BRO7 shows only two strong peaks around 30 (Figure A 2) the latter of w hich correspond to diffraction s from ( 311) and (222) planes Overall, the XRD pattern is matched well with pure BRO7 phase. This is a quite exciting result if it is considered that BRO7 was synthesized with a wet chemical synthesis i.e. glycine nitrate com bustion (GNC) rather than conventional solid state reaction, which is fast auto ignition reaction enabling formation of pure BRO7 with fine particle size resulting from high reaction temperature in short reaction time. Glycine plays a significant role as a chelating agent for cations (Ru, Bi) and a fuel for the reaction as well. Since only mild

PAGE 103

103 post heat treatment (700C for 2rhs) was applied to precursor to remove unreac ted phase such as Ru and Bi 2 O 3 particle size of the final product was kept small (20 ~40nm in diameter ). This is marked reduction in particle size compared to the one synthe sized with solid state reaction showing micron size. However, more careful examination on the XRD pattern confirmed that o ne extra peak other than two main BRO7 peaks e xists at about 28.5 for synthesized BRO7 with GNC It was realized that this peak comes from the different composition of bismuth ruthenate, i.e. Bi 12 RuO 20 (Figure A 2) It was assumed to form due to incomplete reaction between the constituting elements or insufficient of elements in the vicinity during reaction. However, t his is not surprising because t his phase was also observed sometimes even with c onventional solid state reaction Therefore, it should be emphasized that formation of small amount of Bi 12 RuO 20 phase do esn t compromise the superiority of GNC method to conventional solid state reaction for synthesis of fine particle size BRO7. Moreover, it was found that this extra phase could be removed with aid of leaching process. Figure A 3 shows XRD patterns for BRO7 before and after leaching taken from the literature. However, leaching process was not applied to GNC BRO7 in this dissertation for reasons below ; 1. I t is the extra processing step increasing the processing cost and time 2. M aterials loss from l eaching process was concerned 3. T he overall performance was not affected that much since it w as small amount compared to pure BRO7 so that sufficient electronic conductivity is provided from synthesized B RO7

PAGE 104

104 Consequently, pure BRO7 with small amount of Bi 12 RuO 20 was successfully synthesized by GNC and GNC is the strong synthesis technique for producing fine particle s of BRO7, which is preferred for SOFC cathode application since triple phase boundary (TPB) reaction sites for oxygen reduction can be dramatically increased from more contacts between constituting particles in the cathode. However, care must be taken depending on the application. If the electronic conductivity is the main concern of the application, it is desirable to use leaching process for remova l of the extra phase, Bi 12 RuO 20 since this extra phase doesn t contribute to the electronic conductivity.

PAGE 105

105 Figure A 1. Diffraction patterns of Bi 3 Ru 3 O 11 (or Bi 2 Ru 2 O 7.3 ) synthesized with amorphous citrate reaction method Figure A 2. Diff raction patterns of BRO7 and impurity phases such as Bi 3 Ru 3 O 11 (or Bi 2 Ru 2 O 7.3 ) synthesized with GNC method

PAGE 106

106 Figure A 3. Diffraction patterns of BRO7 (bottom) before and (top) after leaching taken from the literature 23

PAGE 107

107 APPENDIX B IMPEDANCE SPECTROSCOPY A short description of impedance spectroscopy is given in this section for better understanding the electrochem ical response since it was one of the main characterization technique s utilized in this dissertation. Electrochemical impedance spectroscopy (EIS) is a valuab le tool for characterization of electrochemical processes. A small AC potential (across a range of frequencies) is applied to the sample, and the current response (impedance using over a range of frequencies. This plot where the real part of the impe dance is the abscissa and the imaginary part of the impedance the ordinate (usually the negative and positive portions of the imaginary axis are reversed for simplicity). The response of the cell is usually modeled in ter ms of equivalent circuits, i.e. a g roup of electrical circuit elements (resistors, capacitors, inductors) that are connected in a way that would give the same response as the cell. A common cell response feature (a semi circle), and its equiv alent circuit representation (a resistor and capa citor in parallel) are shown in Figure B 1. Such behavior could be characteristic, for example, of a double layer capacitance ( resulting from charge separation between electrode and electrolyte) in parallel with a resistance to charge transfer or a polariz ation resistance. It should be noted that the magnitude of the impedance decreases as frequency increases. The semi circle is characteristic of a electrochemical cells contain more than one time constant ( semi circle) indicative of more than one electrochemical process, and often only a portion of one or more of the semi circles is seen.

PAGE 108

108 Often two time constants will overlap and the semi circles must be deconvoluted in order to determine each individual con tribution. Also it should be mentioned that many times in the study of solid samples, the center of the semi circle may be depressed below the x axis. The equivalent circui t is similar to that in Figure B 1, but the capacitor is replaced by a so called con stant phase element (CPE) A capacitor can be thought of as a constant phase element whose phase angle (the phase difference between voltag e and current responses) is 90 When this phase angle is somewhat less than this a depressed semi circle is observe d. This behavior has been explained in a number of ways. Surface roughness of the electrode is one explanation for example, it is common for electrochemical cells with solid electrodes (which typically have rough surfaces) to display this behavior while it is not observed on mercury electrodes (which are atomically smooth). Another common cell response feature on a Nyquist plot is a straight line with a 45 angle (Figure B 2). This feature is usually modeled by so called Warburg impedance and is characteris tic of semi infinite diffusion. As shown in Figure B 2, in many cases at low frequencies, the plot forms an arc. This is justified because at high frequencies, the time for a molecule to diffuse through for example, a porous cathode is much longer than th e period of the cathode is of finite thickness. The response of a cell can be perfectly modeled by a number of different equivalent circuits. Knowledge of the physical processes occurring in e ach cell can help identify the most appropriate model. The model can be justified by altering a single

PAGE 109

109 aspect of the cell ( e.g. grain size, Figure B 3) 104 and verifying that the impedance spectrum changes in such a way that is as predicted by the model. Another typical cell response and equivalent circuit mode l are shown in F igure B 4. This figure will be used as an example for calculation of various cell parameters. The simplest parameter to extract is the total ohmic resistance of the cell, given as R in the figure. This is also known as the solution or electrolyte resistance. It should be noted at this point that while in this case the electrolyte response behaves as a pure resistor, in many polycrystalline electrolytes the response may exhibit some capacitive behavior due to the grains (bulk) and the grain boundaries, hence up to two semi circles may appear in this region of the Nyquist plot (as in Figure B 4 ). R ct is the charge transfer resistance (which is controlled by the kinetics of the charge transfer reaction), and is measured as the difference between the extrapolated low frequency real axis intercept and the high frequency axis intercept. T he rate of the charge transfer reaction can be modeled by the Butler Volmer equation. Since in IS the applied signal is small, the overpotential (the electrode potential minus the equili brium potential for the reaction) should be small, and the Butler Volmer equation becomes (B 1) where R, T, n and F have their usual meanings, a nd i O is the exchange current density. Thus if R ct is known, the exchange current d ensity can be calculated. Diffusion of species toward and away from the reaction sites usually gives the linear response shown at the low frequency end of the figure. Naturally this is not the only form of spectrum observed, and R ct is not the only non ele ctrolyte resistance reported. Polarization resistance, R p and the more general electrode resistance, R el

PAGE 110

110 are frequently reported. However, in most all cases, the value for the resistance is measured as the difference between low and high frequency real a xis intercepts of the arc of interest. Since resistance is not a materials parameter, cell geometry is usually taken into account, and IS results are reported in terms of resistivities or conductivities. For a particular resistance, (B 2 ) (B 3 ) which a uniform current is carried, respectively. It is seen, for example that the ohmic contribution can be identified by performing a series of experiments, holding all experimental conditions the same while changing electrolyte thickness a plot of R versus electrolyte thickness should be a straight line with intercept zero. Area specific resistance is another parameter that is commonly reported and is simply the resistance of interest mul tiplied by the area of interest F or example, the electrode resistance multiplied by the electrode area gives the electrode area specific resistance ( 2 ) These properties will often show an Arrhenius relatio nship with temperature, and a log plot of these parameters versus reciprocal temperature will give a straight line, the slope of which is reported as the activation energy of the specific process. With this brief description of EIS, one can begin to imagin e how this technique can be utilized to help interpret cell behavior, as well as help determine rate controlling processes.

PAGE 111

111 Figure B 1. Typical impedance cell response on complex plane and its equivalent circuit Figure B 2. Typical Warburg impedance response on complex plane

PAGE 112

112 Figure B 3. Impedance response for as sintered (1500 C, 4 h r ) HS3Y samples (circles), annealed at 1200 C for 110 h r in both 10% H 2 balance N 2 atmosphere (triangles), and air (diamonds) showing the effect of increasing grain siz e 104 Figure B 4 a) E quivalent circuit and b) impedance response in complex plane for mixed kinetic and charge transfer control

PAGE 113

113 LIST OF REFERENCES 1. M. C. Williams, Fuel Cells 7 78 (2007). 2. O. Yamamoto, Electrochim Acta 45 2423 (2000). 3. J. M. Ralph, A. C. Schoeler and M. Krumpelt, J Mater Sci 36 1161 (2001). 4. S. M. Haile, Acta Mater 51 5981 (2003). 5. J. Fleig, Annu Rev Mater Res 33 361 (2003). 6. E. P. Murray and S. A. Barnett, Solid State Ion. 143 265 (2001). 7. V. Dusastre and J. A. Kilner, Solid State Ion. 126 163 (1999). 8. M. Liu and Z. Wu, Solid State Ion. 107 105 (1998). 9. A. Jaiswal, C. T. Hu and E. D. Wachsman, J Electrochem Soc 154 B1088 (2007). 10. M. Camaratta and E. Wachsman, J Electrochem Soc 155 B135 (2008). 11. Z. P. Shao and S. M. Haile, Nature 431 170 (2004). 12. P. W. Li and M. K. Chyu, Transactions of the ASME 127 1344 (2005). 13. J. Newman and K. E. Thomas Alyea, Electrochemical Systems Wiley Interscinece, Hoboken, New Jersey (2004). 14. B. C. H. Steele, Solid State Ion. 75 157 (1995). 15. E. D. Wachsman J Eur Ceram Soc 24 1281 (2004). 16. N. X. Jiang, E. D. Wachsman and S. H. Jung, Solid State Ion. 150 347 (2002). 17. M. J. Verkerk, K. Keizer and A. J. Burggraaf, J Appl Electrochem 10 81 (1980). 18. B. C. H. Steele, Solid State Ion. 129 95 (2000 ). 19. T. Takahashi, T. Esaka and H. Iwahara, J Appl Electrochem 7 299 (1977). 20. E. D. Wachsman, P. Jayaweera, N. Jiang, D. M. Lowe and B. G. Pound, J Electrochem Soc 144 233 (1997). 21. J. Y. Park, H. Yoon and E. D. Wachsman, Journal of the American Ceramic Society 88 2402 (2005).

PAGE 114

114 22. E. Girdauskaite, H. Ullmann, M. Al Daroukh, V. Vashook, M. Bulow and U. Guth, J Solid State Electr 11 469 (2007). 23. A. Jaiswal and E. D. Wachsman, J Electrochem Soc 152 A787 (2005). 24. J. M. Bae and B. C. H. St eele, J Electroceram 3 37 (1999). 25. T. Takeda, R. Kanno, Y. Kawamoto, Y. Takeda and O. Yamamoto, J Electrochem Soc 147 1730 (2000). 26. Z. M. Zhong, Electrochem Solid St 9 A215 (2006). 27. G. R. Facer, M. M. Elcombe and B. J. Kennedy, Aust J Chem 46 1897 (1993). 28. R. J. Bouchard and J. L. Gillson, Mater Res Bull 6 669 (1971). 29. K. S. Lee, D. K. Seo and M. H. Whangbo, J. Solid State Chem. 131 405 (1997). 30. W. Y. Hsu, R. V. Kasowski, T. Miller and T. C. Chiang, Appl. Phys. Lett. 52 792 ( 1988). 31. T. Tsai and S. A. Barnett, Solid State Ion. 93 207 (1997). 32. A. V. Virkar, J. Chen, C. W. Tanner and J. W. Kim, Solid State Ion. 131 189 (2000). 33. T. Kenjo, S. Osawa and K. Fujikawa, J Electrochem Soc 138 349 (1991). 34. C. W. Tanner, K. Z. Fung and A. V. Virkar, J Electrochem Soc 144 21 (1997). 35. K. Sasaki, J. P. Wurth, R. Gschwend, M. Godickemeier and L. J. Gauckler, J Electrochem Soc 143 530 (1996). 36. M. Kleitz and F. Petitbon, Solid State Ion. 92 65 (1996). 37. J. Maier, S olid State Ion. 148 367 (2002). 38. J. Maier, J Eur Ceram Soc 24 1251 (2004). 39. P. Costamagna, P. Costa and V. Antonucci, Electrochim Acta 43 375 (1998). 40. X. J. Chen, S. H. Chan and K. A. Khor, Electrochim Acta 49 1851 (2004). 41. S. P. Jiang, Y. J. Leng, S. H. Chan and K. A. Khor, Electrochem Solid St 6 A67 (2003). 42. S. P. Jiang and W. Wang, Solid State Ion. 176 1351 (2005). 43. Y. Huang, J. M. Vohs and R. J. Gorte, Electrochem Solid St 9 A237 (2006).

PAGE 115

115 44. Y. Y. Huang, K. Ahn, J. M. Voh s and R. J. Gorte, J Electrochem Soc 151 A1592 (2004). 45. Y. Y. Huang, J. M. Vohs and R. J. Gorte, J Electrochem Soc 151 A646 (2004). 46. T. Z. Sholklapper, C. Lu, C. P. Jacobson, S. J. Visco and L. C. De Jonghe, Electrochem Solid St 9 A376 (2006). 47. S. P. Jiang and W. Wang, J Electrochem Soc 152 A1398 (2005). 48. F. Zhao, Z. Wang, M. Liu, L. Zhang, C. Xia and F. Chen, J Power Sources 185 13 (2008). 49. J. Chen, F. L. Liang, B. Chi, J. Pu, S. P. Jiang and L. Jian, J Power Sources 194 275 (200 9). 50. W. Zhu, D. Ding and C. Xia, Electrochem Solid St 11 B83 (2008). 51. M. Shah, J. D. Nicholas and S. A. Barnett, Electrochem Commun 11 2 (2009). 52. S. P. Jiang, Mat Sci Eng a Struct 418 199 (2006). 53. T. Z. Sholklapper, C. P. Jacobson, S. J. Visco and L. C. De Jonghe, Fuel Cells 8 303 (2008). 54. J. M. Vohs and R. J. Gorte, Adv Mater 21 943 (2009). 55. M. J. Jorgensen and M. Mogensen, J Electrochem Soc 148 A433 (2001). 56. Z. Y. Jiang, C. R. Xia and F. L. Chen, Electrochim Acta 55 3595 (2010). 57. H. S. Horowitz, J. M. Longo and H. H. Horowitz, J Electrochem Soc 130 1851 (1983). 58. R. G. Egdell, J. B. Goodenough, A. Hamnett and C. C. Naish, J Chem Soc Farad T 1 79 893 (1983). 59. J. Prakash, D. Tryk and E. Yeager, J Power Sources 29 413 (1990). 60. M. Kleitz, T. Iharada, F. Abraham, G. Mairesse and J. Fouletier, Sensor Actuat B Chem 13 27 (1993). 61. N. M. Markovic and P. N. Ross, J Electrochem Soc 141 2590 (1994). 62. S. Linquette Mailley, A. Caneiro, E. Djurado, G. Mairesse and J. Fouletier, Solid State Ion. 107 191 (1998). 63. F. Abraham, Nowogroc.G and D. Thomas, Cr Acad Sci C Chim 278 421 (1974).

PAGE 116

116 64. M. Hrovat, S. Bernik and D. Kolar, J Mater Sci Lett 7 637 (1988). 65. M. M. Milanova, M. Kakihana, M. Arima, M. Yashim a and M. Yoshimura, J Alloy Compd 242 6 (1996). 66. A. Kahoul, P. Nkeng, A. Hammouche, F. Naamoune and G. Poillerat, J. Solid State Chem. 161 379 (2001). 67. C. Abate, V. Esposito, K. Duncan, J. C. Nino, D. M. Gattia, E. D. Wachsman and E. Traversa, Jo urnal of the American Ceramic Society 93 1970 (2010). 68. B. W. Liu and Y. Zhang, J Alloy Compd 453 418 (2008). 69. L. A. Chick, L. R. Pederson, G. D. Maupin, J. L. Bates, L. E. Thomas and G. J. Exarhos, Mater Lett 10 6 (1990). 70. K. Gurunathan, N. P. Vyawahare, G. J. Phatak and D. P. Amalnerkar, J Mater Sci Mater El 16 159 (2005). 71. S. R. Jain, K. C. Adiga and V. R. P. Verneker, Combust Flame 40 71 (1981). 72. B. D. C. A. Wesley, Elements of X Ray Diffraction, second ed., in, p. 284, Addison W esley, Reading, MA (1978). 73. M. E. Orazem, N. Pebere and B. Tribollet, J Electrochem Soc 153 B129 (2006). 74. M. Godickemeier, K. Sasaki, L. J. Gauckler and I. Riess, Solid State Ion. 86 8 691 (1996). 75. Z. L. Wu and M. L. Liu, Solid State Ion. 93 65 (1996). 76. B. C. H. Steele and A. Heinzel, Nature 414 345 (2001). 77. F. Tietz, Q. Fu, V. A. C. Haanappel, A. Mai, N. H. Menzler and S. Uhlenbruck, Int J Appl Ceram Tec 4 436 (2007). 78. C. R. Xia and M. L. Liu, Adv Mater 14 521 (2002). 79. S. B Adler, Chem Rev 104 4791 (2004). 80. N. Shohata, T. Matsumura, K. Utsumi and T. Ohno, Am Ceram Soc Bull 59 335 (1980). 81. D. Hotza and P. Greil, Mat Sci Eng a Struct 202 206 (1995). 82. R. E. Mistler, Am Ceram Soc Bull 77 82 (1998). 83. A. Navar ro, J. R. Alcock and R. W. Whatmore, J Eur Ceram Soc 24 1073 (2004).

PAGE 117

117 84. D. H. Yoon and B. I. Lee, J Eur Ceram Soc 24 739 (2004). 85. L. Y. Wang, G. Y. Tang and Z. K. Xu, Ceram Int 35 487 (2009). 86. M. J. Desroches, I. A. Castillo and R. J. Munz, Pa rt Part Syst Char 22 310 (2006). 87. C. H. Weng and W. C. J. Wei, Journal of the American Ceramic Society 93 3124 (2010). 88. R. J. Hunter, Introduction to Modern Colloid Science New York (1993). 89. V. A. Hackley, R. S. Premachandran, S. G. Malghan a nd S. B. Schiller, Colloid Surface A 98 209 (1995). 90. M. N. Rahaman, Ceramic Processing and Sintering Marcel Dekker, Inc., New York (1995). 91. P. C. Hidber, T. J. Graule and L. J. Gauckler, J Eur Ceram Soc 17 239 (1997). 92. J. Cesarano, I. A. Aksa y and A. Bleier, Journal of the American Ceramic Society 71 250 (1988). 93. B. J. Briscoe, A. U. Khan and P. F. Luckham, J Eur Ceram Soc 18 2169 (1998). 94. Y. Q. Liu, L. Gao and J. K. Guo, Colloid Surface A 193 187 (2001). 95. E. P. Murray, T. Tsai and S. A. Barnett, Solid State Ion. 110 235 (1998). 96. M. J. L. Ostergard, C. Clausen, C. Bagger and M. Mogensen, Electrochim Acta 40 1971 (1995). 97. Y. J. Leng, S. H. Chan, K. A. Khor and S. P. Jiang, J Solid State Electr 10 339 (2006). 98. L. Nie M. Liu, Y. Zhang and M. Liu, J Power Sources 195 4704 (2010). 99. N. M. Sammes, G. A. Tompsett, H. Nafe and F. Aldinger, J Eur Ceram Soc 19 1801 (1999). 100. J. C. Boivin, C. Pirovano, G. Nowogrocki, G. Mairesse, P. Labrune and G. Lagrange, Solid State Ion. 115 639 (1998). 101. B. A. Boukamp, Solid State Ion. 136 75 (2000). 102. S. W. Jung, C. Lu, H. P. He, K. Y. Ahn, R. J. Gorte and J. M. Vohs, J Power Sources 154 42 (2006). 103. Z. Jiang, L. Zhang, L. Cai and C. Xia, Electrochim Acta 54 3059 (2009).

PAGE 118

118 104. S. P. S. Badwal, Solid State Ion. 76 67 (1995).

PAGE 119

119 BIOGRAPHICAL SKETCH Byung Wook Lee was born and raised in Seoul, South Korea in 1979 He earned a Department of Mat erials Science and Engineering at Hany ang University in Seoul, South Korea, in August 20 05 H e enrolled in the Department of Mat erials Science and Engineering at Universit y of Florida in August 2006 to pursue Ph D with special interest in the electrochemical energy device such as solid oxide fu el cell (SOFC) under guidance of Professor Eric D. Wachsman.