<%BANNER%>

Microstructural Engineering of Composite Cathode Systems for Intermediate and Low-Temperature Solid Oxide Fuel Cells

Permanent Link: http://ufdc.ufl.edu/UFE0021667/00001

Material Information

Title: Microstructural Engineering of Composite Cathode Systems for Intermediate and Low-Temperature Solid Oxide Fuel Cells
Physical Description: 1 online resource (160 p.)
Language: english
Creator: Camaratta, Matthew Allan
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: asr, bismuth, bro7, cathode, cell, composite, engineering, esb, evolution, fuel, intermediate, it, itsofc, microstructural, microstructure, optimization, oxide, particle, polarization, ruthenate, silver, sofc, solid, stability, temperature
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 cells (SOFCs) are electrochemical devices with the potential to generate power at high efficiency with little environmental impact. However, in order to improve their commercial appeal, operating temperatures must be lowered from the 800-1000 degrees C temperature range to 500-700 degrees C and below. Due to the high bond strength of oxygen molecules, the kinetics of oxygen reduction are orders of magnitude slower than those of fuel oxidation. Consequently, much research in the reduced-temperature SOFC field is aimed at enhancing cathode performance. A composite cathode makes use of an electronic conducting phase as well as an ion conducting phase in order to spread the 3PB reaction zone beyond the cathode/electrolyte interface. Silver-stabilized bismuth oxide composite cathodes exhibit low resistance to oxygen reduction due to a combination of high catalytic activity for oxygen reduction of both phases, as well as high ionic conductivity of the bismuth oxide phase. Isothermal comparisons were made between pure silver cathodes, silver-yttrium stabilized bismuth oxide (YSB) cathodes, and silver-erbium stabilized bismuth oxides (ESB) at 650 degrees C. The performance of all cathodes was shown to degrade with time. Cathode area specific resistance (ASR) of both the Ag-YSB and Ag-ESB electrodes increased by around 70%, while the pure Ag system experienced a near fourfold increase during the same length of time under open circuit conditions. In light of the electrochemical, microstructural, and chemical evidence presented, it was concluded that electrode microstructural evolution due to growth, agglomeration, and coalescence of the silver phase, rather than chemical reactivity of the bismuth oxide phase, was responsible for the observed degradation in electrochemical performance. Attempts were made to reduce the microstructural evolution of the silver phase in Ag-ESB20 composites by introduction of small particles (nano-size 8YSZ or vibratory-milled ESB20 particles) into the electrode. The addition of 5 vol% 8YSZ nano powders significantly improved unbiased electrode stability by 97%, and reduced the initial, zero time ASR value by 31%. Similar results were obtained when YSZ-free electrodes were prepared from ESB20 powders composed of particles hundreds of nanometers in size as opposed to electrodes prepared from ESB20 powders composed of micron-sized particles?the zero time ASR value was reduced by 25%, and ASR vs. time slope during unbiased testing of the silver-ESB20 system at 650 degrees C was reduced by 95%. The ASR vs. time slopes during testing under a 250 mV external applied bias were lowered by 50% using the smaller ESB20 particles due to suppression silver phase electro-migration. Porous composite electrodes consisting of BRO7 and ESB20 were also synthesized and characterized using impedance spectroscopy on symmetrical cells. Using a combination of compositional and microstructural optimization, a minimum electrode ASR of 0.73 omegacm2 and 0.03 omegacm2 was achieved at 500 degrees C and 700 degrees C, respectively, making it one of the lowest resistance cathode materials reported to date at such low temperatures.
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 Matthew Allan Camaratta.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Wachsman, Eric D.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021667:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021667/00001

Material Information

Title: Microstructural Engineering of Composite Cathode Systems for Intermediate and Low-Temperature Solid Oxide Fuel Cells
Physical Description: 1 online resource (160 p.)
Language: english
Creator: Camaratta, Matthew Allan
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: asr, bismuth, bro7, cathode, cell, composite, engineering, esb, evolution, fuel, intermediate, it, itsofc, microstructural, microstructure, optimization, oxide, particle, polarization, ruthenate, silver, sofc, solid, stability, temperature
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 cells (SOFCs) are electrochemical devices with the potential to generate power at high efficiency with little environmental impact. However, in order to improve their commercial appeal, operating temperatures must be lowered from the 800-1000 degrees C temperature range to 500-700 degrees C and below. Due to the high bond strength of oxygen molecules, the kinetics of oxygen reduction are orders of magnitude slower than those of fuel oxidation. Consequently, much research in the reduced-temperature SOFC field is aimed at enhancing cathode performance. A composite cathode makes use of an electronic conducting phase as well as an ion conducting phase in order to spread the 3PB reaction zone beyond the cathode/electrolyte interface. Silver-stabilized bismuth oxide composite cathodes exhibit low resistance to oxygen reduction due to a combination of high catalytic activity for oxygen reduction of both phases, as well as high ionic conductivity of the bismuth oxide phase. Isothermal comparisons were made between pure silver cathodes, silver-yttrium stabilized bismuth oxide (YSB) cathodes, and silver-erbium stabilized bismuth oxides (ESB) at 650 degrees C. The performance of all cathodes was shown to degrade with time. Cathode area specific resistance (ASR) of both the Ag-YSB and Ag-ESB electrodes increased by around 70%, while the pure Ag system experienced a near fourfold increase during the same length of time under open circuit conditions. In light of the electrochemical, microstructural, and chemical evidence presented, it was concluded that electrode microstructural evolution due to growth, agglomeration, and coalescence of the silver phase, rather than chemical reactivity of the bismuth oxide phase, was responsible for the observed degradation in electrochemical performance. Attempts were made to reduce the microstructural evolution of the silver phase in Ag-ESB20 composites by introduction of small particles (nano-size 8YSZ or vibratory-milled ESB20 particles) into the electrode. The addition of 5 vol% 8YSZ nano powders significantly improved unbiased electrode stability by 97%, and reduced the initial, zero time ASR value by 31%. Similar results were obtained when YSZ-free electrodes were prepared from ESB20 powders composed of particles hundreds of nanometers in size as opposed to electrodes prepared from ESB20 powders composed of micron-sized particles?the zero time ASR value was reduced by 25%, and ASR vs. time slope during unbiased testing of the silver-ESB20 system at 650 degrees C was reduced by 95%. The ASR vs. time slopes during testing under a 250 mV external applied bias were lowered by 50% using the smaller ESB20 particles due to suppression silver phase electro-migration. Porous composite electrodes consisting of BRO7 and ESB20 were also synthesized and characterized using impedance spectroscopy on symmetrical cells. Using a combination of compositional and microstructural optimization, a minimum electrode ASR of 0.73 omegacm2 and 0.03 omegacm2 was achieved at 500 degrees C and 700 degrees C, respectively, making it one of the lowest resistance cathode materials reported to date at such low temperatures.
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 Matthew Allan Camaratta.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Wachsman, Eric D.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021667:00001


This item has the following downloads:


Full Text





MICROSTRUCTURAL ENGINEERING OF COMPOSITE CATHODE SYSTEMS FOR
INTERMEDIATE AND LOW-TEMPERATURE SOLID OXIDE FUEL CELLS





















By

MATTHEW CAMARATTA


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

2007





































O 2007 Matthew Camaratta



































To my mother Maria, father Frank, and inspiration Tonya









ACKNOWLEDGMENTS

A debt of gratitude is owed to my advisor Dr. Eric Wachsman for his support, guidance,

and oversight throughout this experience. Especially helpful were the meetings he led nearly

each and every week, providing myself and others a great degree of clarity and direction. I

would also like to thank Dr. Juan Nino, Dr. Wolfgang Sigmund, Dr. David Norton, and Dr. Mark

Orazem for their advice and participation as part of my committee. Funding was provided by the

US Department of Energy (contract DE-FC26-03NT41959) and US DOE, High Temperature

Electrochemistry Center (contract DE-AC05-76RL 1 83 0).

I also wish to acknowledge Dr. Keith Duncan for his valuable discussions on the various

theoretical aspects of the field, and transfer of experience with writing and presenting. Thanks to

Dr. He-Sung Yoon for his indispensable processing assistance, as well as Dr. Briggs White for

his collaboration with some experimental aspects of this work.

Many thanks are owed to Wayne Acree and Kerry Seibein of the Maj or Analytical

Research Center at the University of Florida for their patience and assistance with both

microstructural analysis and with instrument training sessions.

I also wish to acknowledge my lab mates Jin Soo Ahn, Dr. Abhishek Jaiswal, Doh Won

Jung, Su-Ho Jung, Dr. Sun-Ju Song, Dr. Jun Young Park, Sean Bishop, and everyone else for the

innumerous discussions and debates throughout the years. I want to thank you for your

friendship and for making this experience truly enj oyable.

I thank my mother for her strength, I know this has been a difficult period in her life. I

thank Baba for always being there. I thank my father for his wisdom. I thank Chris and Rob for

the computer help and sense of humor. Finally, I thank you Tonya, for your constant

encouragement throughout the ups and downs and helping me to persevere.












TABLE OF CONTENTS


page

ACKNOWLEDGMENTS .............. ...............4.....


LIST OF TABLES ............ ....._._. ...............7.....


LIST OF FIGURES .............. ...............8.....


AB S TRAC T ............._. .......... ..............._ 13...


CHAPTER


1 INTRODUCTION ................. ...............15.......... ......


2 LITERATURE REVIEW ................. ...............20................


2. 1. Basic Principles of SOFCs ................. ...............20........... ..
2.2. Dominant Component Losses ................. ...............24........... ..
2.2.1. The Electrolyte .............. ...............24....
2.2.2. The Electrodes ................. ...............27........... ..
2.3. Classes of Cathode Materials................ ...............2
2.3.1. Single-Phase Electronic Conductor ................. ...............29........... ..
2.3.2. Dual-Phase Composite Cathodes .............. ................ ...............3
2.3.3. Single-Phase Mixed Ionic and Electronic Conductors (MIEC s) ................... ........3 1
2.4. Cathode Materials Conventional and State-of-the-Art ................. .......................33
2.5 Strategies for Improving Cathodic Performance .............. ...............40....
2.6 Summary ................. ...............44.......... ......

3 STABILITY OF SILVER-BISMUTH OXIDE CATHODES .............. ....................6


3.1. Introduction. ................ ............. ......... ........ ......... ........ .......60
3.2. Experimental ................. ...............61................
3.3 Results and Discussion ............... ...............63....
3.3.1. Compositional Optimization .............. .. ... ........ ... ..... ...........6
3.3.2. Stability: Thermal Performance Decay due to Microstructural Evolution............64
3.4 Conclusions............... ..............6


4 IMPROVING THE STABILITY OF SILVER-BISMUTH OXIDE CATHODES
THROUGH MICRO STRUCTURAL CONTROL................ ...............81


4. 1 Introducti on ................. ...............8 1....._._ ...
4.2 Experimental ............._ ....._.. ...............8 1.....
4.3. Results and Discussion .............. ...............83....
4.3.1 Nano Y SZ Additions .............. ...............83....
4.3.2. ESB20 Particle Size Reduction .............. ...............85....
4.4. Conclusions and Future Work ........._.._.. ...._... ...............87...












5 HIGH PERFORMANCE COMPOSITE BI2RU207-BI1.6ERO.403 CATHODES FOR IT-
S OF CS .............. ...............99....


5 .1 Introducti on ................. ...............99........... ...
5.2. Experim ental ................. .... ........ ... ...............100....
5.2. 1. Electrolyte and Electrode Preparation ................ ...............100.............
5.2.2. Characterization............... ...........10
5.3. Results and Discussion ................ ...............104..............
5.3.1. Chemical Compatibility................... .... ........0
5 3.2. Reproducibility/C ompositi onal Optimization ............__.....___ ..............10 5
5.3.3. Optimization by Particle Size Ratio .............. ...............107....
5.3.4 Sonication and Sedimentation ............ .....___ ...............109
5.3.5. Effect of thickness and current collection ....___ ................ ..................11
5.3.6. Direct compari son with conventi onal cathode sy stems ................. ................. .11 3
5.3.7. Performance under operation ........._.. ....__. ......_.__...........16
5.4. Conclusions ........._.... ...............117.____......


6 CONCLUSIONS .............. ...............138....


APPENDIX EXPERIMENTAL TECHNIQUES .....__.__. ..... .._._. ....._ ._...........14


A. 1. Electrochemical Impedance Spectroscopy .............. .....................143
A.2. Current-Voltage Measurements ................. ...............146................
A.3. DC Electrical Conductivity............... .............14
A.4. Oxygen Exchange Measurements............... .............14
A.5. Thermogravimetric Analysis (TGA) .............. ...............150....
A. 6. Microstructure ................. ...............150................


LIST OF REFERENCES ................. ...............156................


BIOGRAPHICAL SKETCH ................. ...............160......... ......










LIST OF TABLES


Table page

5.1 Volume mean diameter (dv), as well as subscript and symbolic designations for
different sets of starting powders ................. ...............119.......... ...

5.2 Cell open circuit potentials (OCP), maximum power densities (MPD), and ASR for
selected SOFCs at 650 oC from current-density measurements...........__... ...............1 19










LIST OF FIGURES


Figure page

1-1 Energy production efficiency comparisons. ............. ...............19.....

2-1 Schematic representation of a typical SOFC button cell. ................. .................4

2-2 Theoretical values of oxygen partial pressure on the anode (calculated from Eq. 2-8)
and cell potential (calculated from Eq. 2-4) as a function of temperature. ................... .....45

2-3 Schematic representation of an SOFC stack. .............. ...............46....

2-4 Illustrative representation of typical current-voltage profile for an SOFC illustrating
different types of electrochemical losses that arise when current is drawn. ......................46

2-5 Cubic fluorite (AO2) CryStal structure with A-site cations occupying the cube corners
and face centers, and oxygen anions occupying the tetrahedral interstitials. ....................47

2-6 Reported ionic conductivity values for common SOFC electrolyte materials. ...............47

2-7 Calculated cell performance and polarization resistances. ............. .....................4

2-8 Schematic representation of three phase boundary lines between electronic
conducting particles, the ion conducting electrolyte............... ..............4

2-9 Schematic representation of three phase boundary lines between electronic
conducting particles, ion conducting electrolyte .............. ...............49....

2-10 ABO3 cubic perovskite crystal structure. (a) A-site cations at cube center, B-site
cations at corners, and oxygen anions at face centers. (b) Red spheres............................49

2-11 Schematic representation of various reaction pathways for the oxygen reduction
reaction in a MIEC ................. ...............50................

2-12 Effect of cathode thickness on effective charge transfer resistance .............. .................50

2-13 Effect of cathode thickness in I-V performance. ............. ...............5......1

2-14 Agreement between model (Eq. 2-15) and experiment for LSM-YSZ composite
cathodes............... ...............51

2-15 Effect of cathode thickness on polarization resistance for LSM-YSZ composite
cathodes............... ...............52

2-16 Effect of electrolyte phase composition on polarization resistance of various
conventional SOFC cathode materials. ....._ .....___ .......__ ...........5










2-17 Effect composition on (a) cathode overpotential, and (b) interfacial resistance for
Ag-Y SB composite cathodes. ............. ...............53.....

2-18 Theoretical ambipolar conductivity from EMPT. ......____ .... ... .__ ........__......54

2-19 Effect of composition on interfacial resistance of GDC-LSM composite cathodes. .........54

2-20 (a) Arrhenius conductivity plot of selected Fe loadings in LSCF, and (b) cathode
polarization performance comparison ................. ...............55................

2-21 IEDP for (a) LSM and (b) LCCF. ............. ...............56.....

2-22 Effect of A-site dopant on (a) electrical conductivity, (b) cathodic polarization, and
(c) fuel cell power density for various Ln0.6Sr0.4MnO 3 ................. .......... ...............57

2-23 Overpotential-current density curves for (a) diffusion, (b) charge transfer, and (c)
m ixed control. ............. ...............58.....

2-24 Effect of electron-conducting particle size on 3PB length. ............. .....................5

2-25 Arrhenius behavior of interface conductivity for LSC cathodes (a) varying Sr content
at E=0V, and varying voltage .............. ...............59....

3-1 (a) Impedance spectra and (b) electrode ASR vs silver content of Ag-ESB showing
compositional optimization at 600 OC .............. ...............71....

3-2 Arrhenius plot, adapted from Xia et. al.,39 COmparing the ASR of Ag-ESB (this
work) with the best cathodes reported in the literature at the start of this study. ........._....72

3-3 Impedance spectra obtained from a symmetrical Ag-ESB/ESB/Ag-ESB (a), Ag-
YSB/YSB/Ag-YSB (b), and Ag/ESB/Ag (c) cells tested at 650 oC for 100 h. .................73

3-4 Electrode ASR vs. time for Ag-ESB, Ag-YSB, and pure Ag at 650 oC. Linear
regression based on first 10 h of testing for each electrode system ................. ...............74

3-5 Change in electrolyte relative conductivity vs. time for ESB and YSB at 650 oC. ......._...75

3-6 EPMA cross-section linescan of a sample having pure silver electrodes, annealed at
750 oC for 48 h. .............. ...............76....

3-7 XRD spectra of Ag-ESB powder mixtures before and after co-firing at 750 oC for 48
h............... ...............77...

3-8 SEM micrographs comparing the morphology of Ag-ESB electrodes before (a) and
after (b) firing at 750 oC for 1 h and (c) after testing at 650 oC for 100 h. ........................78

3-9 SEM micrographs comparing the morphology of Ag-YSB electrodes before (a) and
after (b) firing at 750 oC for 1 h and (c) after testing at 650 oC for 100 h. ........................79










3-10 SEM micrographs comparing the morphology of pure Ag electrodes before (a) and
after (b) firing at 750 oC for 1 h and (c) after testing at 650 oC for 100 h .........................80

4-1 XRD spectra of Ag-ESB-YSZ powder mixtures before and after co-firing at 750 oC
for 10 h. .............. ...............89....

4-2 Nyquist plots for silver-ESB20 composite electrodes containing 0 vol% (a), 5 vol%
(b), 10 vol% (c), and 15 vol% (d) 8YSZ nanoparticles. .................. ................8

4-3 Effect of time on ASR of silver-ESB20 composite electrodes containing various
volumetric amounts of 8 mol% YSZ nanoparticles. Measurements taken at 650 oC.......90

4-4 Backscatter electron microstructural images of tested [650 oC, 100 h in air under no
applied bias] silver-ESB20 electrodes containing 8YSZ nanoparticles. ...........................91

4-5 Secondary electron microstructural images of an untested (a) and tested (b) silver-
ESB20 electrodes containing 15 vol% 8YSZ nanoparticles ................. ............. .......92

4-6 Results of particle size analysis (number average) for the sieved and vibratory-milled
ESB20 powders............... ...............92

4-7 Nyquist plots (a) and imaginary part of impedance plotted as a function of log-scale
frequency (b) for 50-50 vol% silver-ESB20 composite electrodes .............. ..................93

4-8 Electrode ASR vs. time for 50-50 vol% silver-ESB20 composite electrodes, where
the ESB20 phase was prepared from sieved (triangles) ........................... ...............94

4-9 Nyquist plots (a) and imaginary part of impedance plotted as a function of log-scale
frequency (b) for 50-50 vol% silver-ESB20 composite electrodes .............. ..................95

4-10 Electrode ASR vs. time for 50-50 vol% silver-ESB20 composite electrodes, where
the ESB20 phase was prepared from sieved (triangles) and vibratory milled...................96

4-11 Microstructural images of silver-ESB20 composite electrodes, where the ESB20
phase was prepared from sieved powders--surface before testing .............. ..................97

4-12 Cross-sectional microstructural images of silver-ESB20 composite electrodes, where
the ESB20 phase was prepared from vibratory milled powders ................. ................ ..98

5-1 XRD spectra for mixtures of BRO7-ESB20 before and after heat treatment at 800 OC
for 10 h ................. ...............120...............

5-2 Nyquist (a) and Bode (b) plots at 625 oC for different compositions of BRO7VM-
ESB20s electrodes tested in air ................. ...............121........... ...

5-3 Effect of electrode composition for the ESB20s-BRO7V M system ................. ...............122

5-4 SEM micrograph of as-prepared BRO7 powders before (a) and after (b) vibro-
milling as well as ESB20 powders before (c) and after (d) vibro-milling. ................... ...123










5-5 TEM micrograph of BRO7vM powders before (a) and after sonication and
sedimentation (b), as well as ESBVM powders before (c) and after (d) sonication..........124

5-6 SEM image of fully-fired BRO7-ESB20 cathode systems used in this study--
BRO7s-ESBs (a), BRO7VM-ESBs (b), BRO7s-ESBVM (c), BRO7VM-ESBVM (d)...........125

5-7 Nyquist (a) and Bode (b) plots at 625 oC for different 50-50 wt%/ BRO7-ESB20
electrode microstructures tested in air. ............. ...............126....

5-8 Arrhenius plot of ASR vs reciprocal temperature for the four different
micro strctures studied. ............. ...............127....

5-9 Nyquist (a) and Bode (b) plots at 625 oC for different 50-50 wt%/ BRO7-ESB20
electrode microstructures tested in air before (open symbols) and after ................... ......128

5-10 Arrhenius plot of ASR vs reciprocal temperature--a comparison between electrodes
prepared from as-prepared powders (open symbols)............... ...............12

5-11 Effect of electrode thickness on ASR at 625 OC for the four different electrode
microstructures prepared after ultrasonication and sedimentation ................. ...............130

5-12 Nyquist (a,c) and Bode (b,d) plots at 625 OC for 50-50 wt%/ BRO7-ESB20 at
different thicknesses without (open symbols) and with (closed symbols) pure BRO7...131

5-13 Arrhenius plot of ASR vs reciprocal temperature--a comparison between electrodes
without (open symbols) and with (closed symbols) current collectors............................132

5-14 Solid state adaptation of 3-point Luggin probe configuration (a) schematic
representation and (b) actual cell. ............. ...............133....

5-15 Impedance spectra for LSM-YSZ composite comparing and total cell impedance
measured using 2-point configuration .............. ...............133....

5-16 Current-voltage measurement for LSM-YSZ on Luggin probe cell at 650 oC using
hydrogen bubbled through water as the fuel gas and air as the oxidant gas. ................... 134

5-17 Cathode overpotential versus current density data for selected cathode materials on
Luggin probe cells at 650 oC using hydrogen bubbled through water. ...........................135

5-18 Current-voltage measurement for selected cells at 650 oC using hydrogen bubbled
through water as the fuel gas and air as the oxidant gas ................. .......................136

5-19 SEM image of optimized BRO7-ESB composite cathode on SOFC with Ni-GDC
anode support with ESB/GDC bilayer SOFC after current-voltage testing. ...................137

A-1 (a) Equivalent circuit model and (b) typical impedance spectroscopy cell response. .....152

A-2 Impedance response showing diffusion behavior at high frequencies.............................15










A-3 Impedance response for as-sintered (1500 oC, 4 h) HS3Y samples (circles), annealed
at 1200 oC for 110 h in both 10% H2 balance N2 atmosphere (triangles). ................... ....152

A-4 (a) Response and (b) equivalent circuit for mixed kinetic and charge transfer control...153

A-5 (a) Fuel cell testing schematic and (b) illustrative representation of typical response. ...153

A-6 (a) Typical current-interruption response, and (b) calculated experimental cathodic
overpotentials. ........... ...... ._ __ ...............153.....

A-7 Alternative electrochemical testing setup for cathodic overpotential. ................... ..........154

A-8 Illustrative representation of separation of electrode contribution from fuel cell test
response............... ...............15

A-9 Alternative determination of electrode polarization from symmetrical cell I-V
measurements ................. ...............155................

A-10 Electrode arrangement for DC conductivity measurements. ............. .....................5

A-11 Oxygen nonstoichiometry data for various Co and Fe B-site dopant concentrations. .....155









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

MICROSTRUCTURAL ENGINEERING OF COMPOSITE CATHODE SYSTEMS FOR
INTERMEDIATE AND LOW-TEMPERATURE SOLID OXIDE FUEL CELLS

By

Matthew Camaratta

December 2007

Chair: Dr. Eric Wachsman
Major: Materials Science and Engineering

Solid oxide fuel cells (SOFCs) are electrochemical devices with the potential to generate

power at high efficiency with little environmental impact. However, in order to improve their

commercial appeal, operating temperatures must be lowered from the 800-1000 oC temperature

range to 500-700 oC and below. Due to the high bond strength of oxygen molecules, the kinetics

of oxygen reduction are orders of magnitude slower than those of fuel oxidation. Consequently,

much research in the reduced-temperature SOFC field is aimed at enhancing cathode

performance .

A composite cathode makes use of an electronic conducting phase as well as an ion

conducting phase in order to spread the 3PB reaction zone beyond the cathode/electrolyte

interface. Silver-stabilized bismuth oxide composite cathodes exhibit low resistance to oxygen

reduction due to a combination of high catalytic activity for oxygen reduction of both phases, as

well as high ionic conductivity of the bismuth oxide phase.

Isothermal comparisons were made between pure silver cathodes, silver-yttrium stabilized

bismuth oxide (YSB) cathodes, and silver-erbium stabilized bismuth oxides (ESB) at 650 oC.

The performance of all cathodes was shown to degrade with time. Cathode area specific









resistance (ASR) of both the Ag-YSB and Ag-ESB electrodes increased by around 70%, while

the pure Ag system experienced a near fourfold increase during the same length of time under

open circuit conditions. In light of the electrochemical, microstructural, and chemical evidence

presented, it was concluded that electrode microstructural evolution due to growth,

agglomeration, and coalescence of the silver phase, rather than chemical reactivity of the

bismuth oxide phase, was responsible for the observed degradation in electrochemical

performance .

Attempts were made to reduce the microstructural evolution of the silver phase in Ag-

ESB20 composites by introduction of small particles (nano-size 8YSZ or vibratory-milled

ESB20 particles) into the electrode. The addition of 5 vol% 8YSZ nano powders significantly

improved unbiased electrode stability by 97 %, and reduced the initial, zero time ASR value by

31 %. Similar results were obtained when YSZ-free electrodes were prepared from ESB20

powders composed of particles hundreds of nanometers in size as opposed to electrodes prepared

from ESB20 powders composed of micron-sized particles--the zero time ASR value was

reduced by 25 %, and ASR vs. time slope during unbiased testing of the silver-ESB20 system at

650 oC was reduced by 95 %. The ASR vs. time slopes during testing under a 250 mV external

applied bias were lowered by 50 % using the smaller ESB20 particles due to suppression silver

phase electro-migration.

Porous composite electrodes consisting of BRO7 and ESB20 were also synthesized and

characterized using impedance spectroscopy on symmetrical cells. Using a combination of

compositional and microstructural optimization, a minimum electrode ASR of 0.73 OZcm2 and

0.03 OZcm2 was achieved at 500 oC and 700 oC, respectively, making it one of the lowest

resistance cathode materials reported to date at such low temperatures.









CHAPTER 1
INTTRODUCTION

There is an expanding list of reasons to explore alternatives to conventional energy

generation methods (namely coal-fired steam turbines and gasoline-powered combustion

engines). From an ecological standpoint, there has been a longstanding need to reduce the

polluting byproducts of conventional energy generation. Greenhouse emissions from

automobiles and power plants have long been suspected of contributing to global climate change.

Particulate matter exhausted from buses and larger vehicles can penetrate into the lungs, causing

respiratory and cardiac disease. From an economic standpoint, worldwide demand for energy

continues to rise as more and more nations, notably China, j oin the ranks of the industrialized

countries. Also, the cost of oil is subject to global political landscapes--the dramatic upswing in

gasoline and diesel fuel prices during winter and early spring 2000 were the direct result of an

intentional reduction in crude oil production by OPEC. This exemplifies how sensitive oil prices

are to a relatively small (about 4%)' reduction in supply, and given that oil is a finite resource,

this signifies what may happen as supplies eventually begin to dwindle. From a socio-political

standpoint, there has been a lot of tension regarding the foreign policies of oil-consuming

nations, as evidenced by frequent protests and even wars. Lowering the global dependence on

oil might reduce such tension. Yet despite all of this, changes to energy generation methods

throughout the world have been exceedingly slow, as evidenced by the internal combustion Otto

cycle engine, which has been used in automobiles for more than a century.

Obviously, the concept of alternative energy is not new. Many different sources of energy

exist (or are proposed) that can supplement, and have the potential to replace, conventional

generation methods. Popular examples include solar radiation, wind motion, and fusion energy.

Each of these technologies has its own set of problems that have plagued its commercial









inception, but much research is being conducted to overcome these problems. Presently, it is

unknown which of these sources, if any, will realize the greatest share of the commercial energy

market in the coming decades. What is certain, however, is that the market for alternative highly

efficient, eco-friendly energy production technologies is expanding.

There is a general push towards higher efficiencies. Current combustion-powered

automobiles have an overall efficiency of about 20%.2 That is, only 20% of the thermal energy

content of the gasoline is converted into useful mechanical work and the rest is wasted. Higher

efficiencies translate into reduced energy costs per unit of work done.

Fuel cells, another alternative energy technology, have received growing interest in recent

years. The fuel cell, an invention credited to Grove in 1939,3 is a device that converts the

chemical energy of fuels directly into electricity and heat by electrochemically combining the

fuel and an oxidant gas via an ion-conducting electrolyte. The need for direct combustion is

eliminated, giving fuel cells much higher conversion efficiencies than conventional thermo-

mechanical methods (Figure 1-1).4 As a result, fuel cells have the added advantage of lower CO2

emissions per unit of power output compared to conventional technologies.' Moreover, the

production of NOx and SOx, the main components of smog and acid rain, are expected to be 90%

lower than in conventional pulverized coal plants5--the NOx and SOx emissions from a 100 kW

Siemens Westinghouse SOFC system were both reported to be less than 1 ppmV.6 Fuel cell

power generation is virtually noise-free and can be tailored for standalone, off-grid applications,

eliminating the need for, and losses associated with power transmission lines. Additionally, fuel

cells have the potential to be used for combined heat and power (CHP or cogeneration)

applications. This further increases efficiency by utilizing the thermal energy that is produced










generating electricity. The current trend to deregulate the production of electric power will serve

to promote CHP, thus furthering the marketability of fuel cells.

There are several different kinds of fuel cells in development. The main difference

between each is the material used for the electrolyte, and consequently the operating temperature

ranges and thus plausible market applications. Solid oxide fuel cells (SOFC) utilize a fast

oxygen ion-conducting electrolyte, which is capable of supporting a high flux of oxide ions (Ol-)

while in the solid state. A great deal of heat is produced during the energy production process.

This heat can be utilized to partially reform common hydrocarbon fuels--gasoline, diesel, and

alcohol--internally without combustion, unlike the case of polymer electrolyte membrane fuel

cells (PEMFC) where external partial combustion reformers are required to produce hydrogen

gas. The CO gas created during the reforming process is consumed as an additional fuel (unlike

the case of PEMFC where CO gas is a catalyst poison). The generated heat can also be utilized

for CHP applications. These points add to the efficiency of SOFC. Also, SOFC are made from

inexpensive ceramic materials, and don't require precious metal catalysts like platinum (which is

required for lower temperature fuel cells). For these reasons, some experts predict that solid

oxide fuel cells will be the eventual winner in the automotive market.4

There is one maj or drawback of SOFC that has hurt their viability in this market. The high

operating temperatures (800-1000oC) of current SOFC technology would require extended

startup times--a period of fuel burning is needed to reach these operating temperatures. High

operating temperatures are also responsible for sealing problems and expensive interconnect and

balance-of-plant materials for SOFC stacks. Additionally, high temperature operation can

induce thermal stresses at electrolyte-electrode interfaces, as well as cause interdiffusion between

cell components.' Most of these problems will be solved if the operating temperature can be









lowered to around 500oC.8 An added bonus of lowered operating temperatures is the possibility

of direct oxidation of methane without carbon deposition, which would eliminate the need for

fuel reforming.9 For these and other reasons, the movement to lower the operating temperature

of SOFC has intensified. However, the barriers to low operating temperatures are significant.

To gain a better insight into lowering the operating temperature of SOFC, one must first

understand the construction of SOFC, the various reactions and losses that occur in the different

components of SOFC, as well as explore some of the progress already made by researchers.











A-X Eff
10 -1. 200 kw and larger use bottoming cycle -0
90 -1 2. Use either heat exchanging or combined cycle 90
3. Using methane fuel and pressurization
80 4. There is some potential for ceramic boilers and turbines -80


40 40
DieseldeFul al
20 -I 20
10 Stemc am Turbine4~~-~ -C 10



30 200~ 20000k





Figure 1-1. Energy production efficiency comparisons [Reprinted from B. Weins, The Future of
Fuel Cells, http://lwww.benwiens. com/energy4.html Copyright (2002), with
permission from Ben Wiens Energy Science, Inc.].









CHAPTER 2
LITERATURE REVIEW

2.1. Basic Principles of SOFCs

Solid oxide fuel cells are capable of operating over a wide range of temperatures.

Intermediate temperature SOFCs (IT-SOFCs) operate in the range from 500 oC to 700 OC.10 A

typical single SOFC [Figure 2-1] is composed of a dense ceramic oxide electrolyte separating

two porous electrode catalysts--a cathode on the oxidant (air) side and an anode on the fuel side.

Oxygen molecules are fed to the cathode where it combines with electrons (supplied from the

electrocatalyst in the cathode which is electrically connected to an external circuit) and is

reduced into oxygen ions, which are inj ected into the electrolyte (Eq. 2-la), where the subscripts

'c' and 'e' stand for cathode and electrolyte, respectively. Note that at times it is more

convenient to think of the process of oxygen ions moving through the electrolyte as the

equivalent process of oxygen vacancies (which can be considered simply as point defects in a

periodic oxide lattice where a normally occupied oxygen site is left vacant) migrating from the

anode side towards the cathode side, depicted in Eq. 2-1b using Kroiger-Vink notation. That is,

Oc + 4e = 202- (2-la)


02, + 2e + Vo = Oo (2-1b)


The electrolyte conducts the ions, but blocks electrons. The oxygen ions emerge on the anode

side of the electrolyte where they react with fuel (H2 and CO) to form H20, CO2 and electrons,

20,2- + Ha = H20 + 4e (2-2)

202- + CO, = CO2 + 4e (2-3)

where the subscript 'a' stands for the anode.









Electrons produced during the reaction are released to the electrocatalyst in the anode

(which is also electrically connected to the external circuit). Thus the circuit is complete and

useful work can be done in the external circuit. The gradient in oxygen concentration (or oxygen

partial pressure or oxygen chemical potential) across the electrolyte provides the driving force

for all of these electrochemical processes. Thus, SOFCs are considered oxygen concentration

cells with a theoretical reversible voltage (or electromotive force, emf), Eth, glVen by the Nemnst

equation

RT, /4pOp, 24
E,= In 2(24


where R is the gas constant, T is the absolute temperature, F is Faraday's constant, and pO,, and

pO,, are the partial pressures of oxygen at the anode and cathode, respectively. The coefficient

'4' in the denominator represents the number of electrons transferred per mole of oxygen

molecules reacted in the cell. As already stated, air is usually fed as the oxidant to the cathode

side, thus pO,, is 0.21 atm. The oxygen partial pressure on the anode side for any particular

fuel can be predicted using the Gibbs free energy change equation

AG, = AH, + TSD =RTln eqn) (2-5)

where AHO is'" then heat~ chnge and A is the entropy change of the fuel gas reaction at any


anode temperature, T. Keq is the equilibrium constant for the fuel gas reaction. For many

laboratory SOFC experiments, the fuel is hydrogen gas bubbled through water at room

temperature. The fuel gas reaction is


H,!, +-Ozi, = H,O(, (2-6)

and the mass action expression is










pH,O
K = (2-7)





Rearrangement and substitution of Eq. 2-7 into Eq. 2-6 yields

p0=/H2O 2 AHO + TASf O
pO2~I = ) e (2-8)
p ~HZ RT

The vapor pressure of water at any bubbler temperature, Tbub aS Well aS the heat and entropy

change values, AH/o and ASO ,,,,;, arereailyavalabe i tabulated form.11,12 Assuming no water


condensation occurs in the gas delivery system between the water bubbler and the SOFC anode,

the expression for pO2 in Eq. 2-8 can be substituted for pO,, in the Nernst Equation (Eq. 2-4) to

predict the theoretical Nernst potential at any cell temperature. Plots of pO,, and Nernst

potential versus cell temperature using hydrogen fuel bubbled through water at 25 OC are shown

in Figure 2-2. To increase theoretical cell voltage and power output, multiple cells are usually

combined in series into a 'stack' with adj oining anodes and cathodes separated by an

interconnect material (Figure 2-3).

No cell is perfect--that is the cell voltage during operation is always lower than the

theoretical Nernst value (Fig. 2-4). As current is drawn to the load, the cell voltage is reduced

due to various electrochemical losses. For an SOFC with an ionic transference number, ti=1

(discussed below in Section 2.2), cell polarization or overpotential, rl, is the difference between

theoretical and operating voltages. The total cell polarization can be considered the sum of three

individual contributions--resi stance ohmicc) polarization (Tla), charge transfer (activation)

polarization (T1A), and diffusion (concentration) polarization (rlD).

rltot = rln + T1A rlD (2-9)









Ohmic polarization is caused by resistance to the flow of ions in ionic conductors and

electrons in electronic conductors, and by contact resistances between cell components. As the

name implies, these potential losses are directly proportional to the amount of current passing

through the cell.

Activation polarization ensues whenever reacting chemical (including electrochemical)

species are involved. An activation energy barrier must be overcome in order for the reaction to

proceed. This results in a potential drop which may be regarded as the extra potential necessary

to overcome the energy barrier of the rate-determining step of the reaction, and is related to the

current passing through the cell by the phenomenological Butler-Volmer relation (discussed

below in Section 2.4). In the electrode reactions (Eq. 2-1 through 2-3), the activation

polarization is due to the transfer of charges between the electronic and ionic conductors. For

this reason this overpotential is often termed charge-transfer polarization.

Concentration polarization occurs when reacting species are supplied to reaction sites

slower than they are consumed, or when reaction products are not removed fast enough so that

they block the reaction sites. This effectively results in lower concentrations of reactant species

at the reaction sites (lower oxygen partial pressures at the cathode or lower partial pressures of

fuel at the anode) than in the bulk of the gas stream and leads to a concomitant reduction in

voltage. This type of loss becomes significant when large amounts of current are drawn from the

cell, and in some cases a limiting current is reached where the concentrations of gases are

unsustainable (approach zero at the reaction sites) causing fuel cell operation to cease.

While these losses cannot be completely eliminated, they can be minimized by proper

choice of materials and by microstructural engineering of each cell component.









2.2. Dominant Component Losses


2.2.1. The Electrolyte

The dominant polarization losses in SOFC electrolytes are ohmic losses caused by the

resistance to ionic conduction. In fact, the main contribution to the ohmic polarization of a

SOFC is due to the electrolyte.13 As stated earlier, the role of an SOFC electrolyte is to conduct

oxygen ions from the air side of the cell towards the fuel side. Good electrolyte materials have

high ionic conductivity (low resistivity) for oxygen ions (or equivalently oxygen vacancies), and

low electronic conductivity (to prevent internal short-circuiting and thus reduced cell

performance) under an applied field. That is, ideally, the electrolyte should have an ionic

ttrttrttrttrttr~ansference number equal to unity,


t = (2-10a)


where ti is the ionic transference number, oi is the ionic conductivity (S/cm), and oe is the

electronic conductivity (S/cm). This results in a deviation from theoretical Nernst potential,

allowing ti to be measured using

E,
t = (2-10b)
'EN

where EN is the theoretical Nernst potential, discussed above, and Eo is the actual cell open

circuit potential (OCP). A material that has a low migration enthalpy for oxygen vacancies (or

ions) in its lattice will have high ionic conductivity.14 The mobility of vacancies or ions within a

lattice is affected, for example, by the spacing between ions in the lattice (which is related to the

lattice parameter of the lattice), the binding energy between counterions and oxygen ions (which

is dependant on the type of cations in the lattice), and also what are known as defect associates,

which are caused by bonding between these defect species. The concentration of defects also









affects the ionic conductivity. Oxygen ions move by jumping into an oxygen vacancy. If only

one vacancy were available for such a jump, only one ion could move at a time. By increasing

the number of oxygen vacancies, the number of mobile ions also increases and thus the ionic

conduction is increased. However, at a certain concentration the defects can begin to interact,

and these interactions tend to lower ionic conduction.

One crystal structure that is seen repeatedly in SOFC electrolytes is the cubic fluorite

structure (Figure 2-5). A variety of oxides can exist in the fluorite structure (AO2), Such as

zirconia (ZrO2), ceria (CeO2), and bismuth oxide (Bi203). The configuration of the two

sublattices and its "open" or "spacious" nature allow for relatively facile defect migration in

fluorite oxides. For many of these materials however, the cubic fluorite structure is a high-

temperature polymorph. For example zirconia undergoes a cubic-tetragonal transition below

around 2400oC, and bismuth oxide undergoes a transition from cubic to monoclinic around

730oC.15 It has been shown that the cubic fluorite structures of these materials can be maintained

down to room temperature by doping with various trivalent (and divalent) transition metal and

rare-earth oxides. Typically, dopant cations with ionic radii larger than that of the host cation are

more effective in stabilizing the cubic fluorite phase--smaller dopant cations tend to distort the

lattice more than larger ones.16 In addition to imparting structural stability to these materials,

these oxides may also serve to improve ionic conduction by virtue of the requirement that the

lattice be electronically neutral. That is, the aliovalent dopant cations (Re+3), which substitute

the host cations (A+4), have a net negative charge compared to normal lattice positions. However

during this substitution process the lattice must maintain charge neutrality. Two of these

negative charges can be compensated for by the creation of a vacant oxygen site (or oxygen

vacancy), which has a +2 charge. In Kroiger-Vink notation the dopant incorporation reaction is









2RO o'> R +V +4Oo (2-11)

Thus, the stabilized material is also imparted with a certain amount of oxygen vacancy

defects, which (up to a point) increases ionic conduction. One can see from this that the

electrolytic behavior of a material can be improved by optimization of the type and concentration

of dopant cations in the lattice.

The conventional SOFC electrolyte, yttria-doped zirconia (YSZ), has an operating

temperature around 10000C. As the operating temperature is lowered, ohmic losses across the

solid electrolyte increase. "1 These losses can be minimized by reducing the distance that ions

and vacancies need to travel, i.e. by reducing electrolyte thickness, and by choosing a material

that has higher ionic conductivities than YSZ.8 Several physical and chemical techniques can be

used to deposit thin-fi1m electrolytes on to anode substrates including vapor deposition, slip

casting, and electrophoretic deposition.l Significant reductions in electrolyte resistances have

been achieved by use of anode-supported thin-film electrolytes. "1 Two alternative materials

which have shown improved performance over YSZ are doped ceria8,19 and stabilized bismuth

oxide.20,21

At intermediate temperatures (~500-750 oC), stabilized bismuth oxides show some of the

highest ionic conductivities among all studied fast ion conducting oxides (Figure 2-6).22 This is

attributable to the inherent nature of its lattice structure. Because bismuth is a trivalent cation,

and because it can be stabilized in the cubic fluorite form, bismuth oxides intrinsically have '/ of

the regular fluorite anion sites vacant. Consequently, there is a large excess of equipotential

lattice points among which the anions (oxygen ions) can be distributed, which in effect leads to

the observed high conductivities. Stabilized bismuth oxide thus has the potential to provide high

power densities at low temperatures. However use of this material in fuel cell electrolyte









applications has been limited because it becomes unstable in reducing conditions such as those

experienced at fuel cell anodes.23

While the ionic conductivity of stabilized ceria is less than that of stabilized bismuth oxide,

it has also shown great potential to replace YSZ in fuel cell electrolyte applications.8,24 However

ceria also has limitations at reducing conditions. The electrolytic domain of 10 mol% and 20

mol% GDC persists until ~10- atm at 800 OC and ~10-20 atm at 700 OC.25 Below these pO2

values, some of the Ce+4 ions take on an electron and are reduced to Ce+3, creating Cece defects

(also known as small polarons). These defects can be conducted through the electrolyte by a

hopping mechanism from Cece sites to Ceie sites. This results in n-type conduction, lowered

ionic transference number, short-circuiting of the cell and thus reduced cell efficiency. This

leads to further problems when reducing electrolyte thickness to lower ohmic losses. When

mixed conductivity is present, any electronic leakage current will increase as electrolyte

thickness is decreased.8

2.2.2. The Electrodes

The dominant losses in the electrodes of SOFC are activation polarization and

concentration polarization. Activation polarization is typically more of an issue in cathodes due

primarily to the kinetics of the oxygen reduction reaction (attributable to the high bond strength

present in oxygen molecules), which is several orders of magnitude slower than the reactions

involving fuel oxidation.26-32 The dissociated chemisorption of H2 in a typical SOFC anode, for

example, is a non-activated process.32 That is there is no barrier to the dissociation of a

hydrogen molecule into two separate adsorbed atoms on (open, atomically rough faces of) the

catalyst surface (typically Ni). At the cathode, the dissociative adsorption of oxygen is a

thermally activated process.33 As the operating temperature is lowered, the rates of chemical









reactions drop dramatically, and activation polarization at the cathode becomes an even larger

issue. Thus anode-supported cells have gained wider acceptance than cathode-supported cells in

low-temperature SOFC research. This use of thick anodes results in longer diffusion lengths,

and concentration polarization can become an issue, especially at high current densities. It

should be noted that an analysis of isothermal transport of gaseous species through porous

electrodes, based on commonly used anode and cathode gases, suggests that for comparable

porosities and electrode thickness, the concentration polarization effects should in general be

lower on the anode side.13 A complete polarization model of SOFC has been developed34 and

the various loss mechanisms for an anode-supported cell are shown against a power density

curve in Figure 2-7. These curves were generated using data extracted from literature that

focused on cells that operated at 8000C. It should be noted that the values used for exchange

current densities, electrolyte conductivity and thickness were among the highest values obtained

from the literature. Based on the fact that the cell voltage reaches zero before an anodic limiting

current is reached, the authors conclude that the anode limiting current density will not be

reached under normal operating conditions. In addition, experimental studies have shown that

the low-current interfacial resistance was 70-85% of the total cell resistance from 550oC to

800oC, and still played a major role in limiting cell performance at higher current densities. For

these reasons, development of highly active cathodes, based on improvements in materials and

microstructures that are thermally and chemically compatible with adj acent stack components

are critical to the successful operation of low temperature SOFC.

2.3. Classes of Cathode Materials

The oxygen reduction reaction (Eq. 2-1) is the net reaction of interest for all SOFC

cathodes. There are a series of steps involved in the reduction mechanism which depend on the

type of cathode used. The cathode resistance emerges from these processes. In order to improve










the performance of cathodes for low temperature SOFC, whether by optimizing microstructure

or composition, one must understand the various possible steps involved in this overall reaction.

A review of three basic categories of porous cathode materials--single phase electronic

conducting cathodes, dual phase electronic and ionically conducting cathodes, and single phase

mixed ionic and electronic conducting (MIEC) cathodes--is given below to help gain insight

into these processes. The review will include sections covering the mechanism for oxygen

reduction, experimental (and theoretical) results, and methods for the determination and

enhancement of the rate-limiting step of the reduction reaction. The experimental methods

mentioned in this section are discussed in some detail in Appendix A of this dissertation.

2.3.1. Single-Phase Electronic Conductor

The most conventional type of cathode is a porous single-phase electronic conductor.

Traditional cathodes for YSZ electrolytes such as platinum (and other noble metals), and the

electronically conducting ceramic material, Lal-xSrxMnO3+6 (LSM; 6 is the average number of

vacant oxygen sites per unit cell, or more simply, the oxygen nonstoichiometry, which ranges

from +0.01 to +0.06 at 1000 oC and 700 oC, respectively, in air, for the typical composition

x=0.2)35, fall into this category. This particular construction restricts the oxygen reduction

reaction (Eq. 2-1) to one-dimensional regions where the three phases--gas phase (pores),

electrode phase, and electrolyte phase--meet, or three-phase boundaries (3PB). In this case, one

can visualize that such 3PB are limited to the electrode/electrolyte interface (Figure 2-8).

The gas phase supplies the oxygen species, the electrode phase supplies the electrons and

catalyzes the reaction, the oxygen-deficient electrolyte phase provides the vacant sites into which

the ionic oxygen species are inj ected. Generally, the reaction mechanism for this type of cathode

involves gaseous oxygen molecules diffusing through the pores (or they may dissociate and

adsorb onto the surface of the cathode and diffuse along the pore walls) toward the reaction sites









(3PB), where charge transfer occurs and oxygen ions are injected into the electrolyte. Naturally,

the reaction cannot occur on a one-dimensional line. Before combination with an electrolyte

oxygen vacancy, it is expected that some diffusion of oxygen ad-atoms over the electrode and

electrolyte surfaces takes place, but this reaction zone is expected to be very narrow.16 LSM

cathodes have been particularly useful for these conventional SOFC due to their relatively low

cost (compared to noble metal catalysts) and their thermal and chemical compatibility with YSZ

electrolytes. In fact, the LSM cathode/YSZ electrolyte combination is the most widely studied

system in the SOFC field.36 Such cathode materials have proven effective for SOFC operating at

high temperatures (greater than 1000 oC) where reaction kinetics are fast. However, for cells

that are to operate at lower temperatures the polarization losses in these cathode materials

become so large that practical power densities cannot be achieved. A great deal of materials

engineering has thus been necessary in order to overcome these losses.

2.3.2. Dual-Phase Composite Cathodes

One of the means to this end has been the development of porous dual phase composite

cathodes where one phases is a purely electronic conductor and the other phase is a purely ionic

conductor. The rationale for the use of this type of cathode is that it allows the reaction zone to

spread from the cathode/electrolyte interface into the cathode (Figure 2-9). The length and

density of the 3PB is increased, i.e., more sites are available for the charge transfer reaction.13

The general reaction mechanism is essentially the same as described above for purely electronic

conducting cathodes except that the diffusion distances for the oxygen species to the 3PB

reaction sites are shorter for dual phase cathode materials. Also, there is an additional path for

transport of oxygen species--bulk diffusion through the ionically conducting phase.

The composition, including the materials used for each constituent phase and the relative

amounts of each phase, is a key factor in determining the performance of these dual phase









composite cathodes. Consideration must be given to ensure that the individual phases within the

composite electrode are continuous in order to create a path for the transport of the respective

species--one phase for electrons, one phase for ions, and pores for gases. According to the

effective medium percolation theory (EMPT),37 in Order for a randomly distributed solid phase to

be continuous, its volume fraction should exceed 1/3 of the total volume of the composite,

including porosity. In addition, the cathode must be chemically and thermally compatible with

the electrolyte. For this reason, the ionic conducting phase is usually chosen to be the same

material as the electrolyte. The choice of the electronically conducting phase has a large effect

on the inherent catalytic activity of the composite cathode towards the oxygen reduction reaction.

This phase must be compatible with the ionically conducting phase for structural stability and to

keep the intrinsic interfacial resistances between these two phases low.38 COmmon examples of

this type of cathode include LSM/YSZ, and Ag/YSB (yttria stabilized bismuth oxide).39

2.3.3. Single-Phase Mixed Ionic and Electronic Conductors (MIECs)

One class of SOFC cathodes that is growing in popularity includes single-phase oxides that

exhibit mixed ionic and electronic conductivity. The most frequently used and versatile crystal

structure in this category is the ABO3 perOvskite structure (Figure 2-10).

Typical host lattices of perovskites used for SOFC cathode applications are composed of a

trivalent lanthanide cation on the A-site and a transition metal cation on the B-site. Electronic

conduction is possible due to the multivalent nature of the B-site cation. If the cation has an

affinity towards increasing its valence state (say, from +3 to +4), p-type electronic conduction

can occur by an electron-hole hopping mechanism (also a small polaron mechanism). In

contrast, if the cation has a tendency to want to lower its valence state (from +3 to +2, for

example), n-type electronic conduction can also occur via a small polaron electron hopping

mechanism. Note that this latter example can also have the effect of creating oxygen vacancies









in order to charge-compensate the lattice, while the former example can have the opposite effect

of oxygen vacancy annihilation. Oxygen mobility through vacancies provides the basis for ionic

conductivity in these perovskite materials.40 Another way defects may be introduced is by

doping the A-site cation with a divalent cation, typically Sr+2 Or Ca+2. In air and at low cathodic

overpotentials, the perovskite material is nearly stoichiometric, hence charge compensation for

negative defects caused by substitution of divalent (e.g., Sr) cations for the trivalent (e.g., La)

cation is mostly obtained by formation of electron holes.41 Under conditions of reasonably high

cathodic overpotential, the pO2 lOcal to the reaction sites is lowered, and as a result of defect

equilibrium requirements, oxygen vacancies concentration is increased.

The presence of these electronic and ionic defects broadens the electrochemically active

region of the cathode to a much larger surface area--the oxygen reduction reaction is no longer

limited to regions close to the triple phase boundaries. In other words, since both electrons (or

electron holes) and oxygen vacancies are mobile defects in a MIEC cathode, it is possible for

oxygen to be reduced at the gas/MIEC interface according to (Eq. 2-12a) or (Eq. 2-12b).

O +2e(E) + VOCARUEC) = Oj(HC (2-12a)



02 O + yAh(EC) = O;(AhUEC) + 2h(Ah(EC) (2-12b)

Furthermore, an additional path through which oxygen species may be supplied to the

cathode/electrolyte interface--bulk transport through the electrode by vacancy jumping. At

reasonable cathodic overpotentials (noted above), this oxygen flux through the electrode is likely

to be increased. Consequently, MIEC cathode materials typically exhibit lower polarization and

have improved performance compared to the other classes of cathodes already discussed. The

reaction path for oxygen reduction in such MIEC cathodes has been described27,30,42 as including









(1) oxygen diffusion in the gas phase, (2) oxygen adsorption-desorption on the MIEC cathode

surface with charge transfer and oxygen incorporation at the MIEC cathode surface, (3) surface

diffusion of the adsorbed oxygen on the MIEC surface, (4) bulk diffusion of oxygen ions in the

MIEC cathode, (5) oxygen ion transfer at the cathode/electrolyte interface, and (6) charge

transfer and incorporation of the adsorbed oxygen at the 3PB (Figure 2-11i).

With some groundwork laid, some of the experimental and theoretical results and

comparisons between the performances of different SOFC cathode materials will now be

discussed.

2.4. Cathode Materials Conventional and State-of-the-Art

The rate at which charge transfer reactions occur at the electrocatalyst/electrolyte interface

is often expressed by the aforementioned Butler-Volmer equation, which relates this current

density, i, to the activation overpotential, rlact by


I = io exp~, expr )l~, (2-13)
RT RT




where io is the exchange current density, a is the transfer coefficient, n is the number of electrons

participating in the electrode reaction. F, R, and T were defined in Section 2.1i. At sufficiently

low current densities, Eq. 2-13 may be approximated by


'Gacn~ -i=R~ (2-14)

RT
Where Rc, =is an intrinsic charge transfer resistance which is restricted to the
nFio

electrolyte/electrocatalyst interface.13 Note that this resistance term (or any resistance term) is









often normalized into an area specific resistance (ASR). This is accomplished simply by

multiplying the resistance by an area term (such as electrode area).

Virkar et al.13 have developed an expression showing the profound effect of adding an

ionically conducting phase (such as YSZ) to a conventional electronically conducting cathode

(such as LSM). In composite electrodes, the reaction zone, or region over which the process of

charge transfer occurs, is spread out from the electrocatlayst/electrolyte interface into the

electrode and as a result the activation overpotential may be lower than the intrinsic one given in

Eq. 2-14. Thus an effective charge transfer resistance, Rf is defined and used in its place; i.e.

Tact = Rcfi The expression for R~f is rather lengthy and won't be reproduced here. However,

assuming the electrode is sufficiently thick, Rcf~ can be approximated by


Rf o (2-15)


where B is the grain size of the electrolyte (or ionically conducting phase) in the composite

electrode, Vv is the fractional porosity, and oi is the ionic conductivity of the electrode (or that of

the ionically conducting phase in the composite electrode). Using as an illustration, typical

values for the parameters-Re~t=1.2 OZcm2, oi=0.02 S/cm, B=2 Cpm, Vv=0.35--give an effective

charge transfer resistance for a porous dual-phase composite cathode, R~f =0.14 OZcm2. This is

nearly an order of magnitude reduction in the intrinsic charge transfer resistance value of 1.2

OZcm2 (Or equivalently an order of magnitude increase in the effective exchange current density

value) obtained for the porous single-phase electronic conducting cathode where the reaction is

limited to the cathode/electrolyte interface.









It should be noted that the non-simplified model for Rf takes into account the effect of

cathode thickness (Figure 2-12). In general, if oi is sufficiently high, the value of R~f will

decrease with thickness down to an asymptotic value (the plots where Rcf~ increase with

thickness simply illustrate the deleterious effect if oi is low enough and explains the peculiar

behavior of certain composite electrodes reported in the literature).43 This effect of improved

performance by increased thickness is validated by a series of current-voltage plots in Figure 2-

13.

The decrease in performance for the 85 Cpm thick sample is presumably due to

concentration polarization effects. A value for R~f can be extracted from these plots by

measuring the cell resistances from the near-linear regions and subtracting the electrolyte ASR

(which can be obtained using independent impedance spectroscopy measurements, see Appendix

A). These results agree well with the theoretical model (Figure 2-14). Mogensen and Skaarup32

observed the same behavior of lowered polarization resistance with increasing thickness of LSM-

YSZ composite cathodes (Figure 2-15). These authors also report that the addition of 50

weight% YSZ to Lao.ssSro.1SMnO3 decreased the polarization resistance from 0.77 OZcm2 to 0. 16

OZcm2 at 900oC.

Based on Eq. 2-15, increasing the ionic conductivity of the ion-conducting phase will

improve cathode performance, everything else being constant. This effect has been explained in

that higher ionic conductivities allow ions produced deeper within the electrode to reach the

electrolyte, thus effectively increasing the width of the reaction zone.38 Murray and Barnett

inspected this effect by substituting the YSZ phase with a material that is known (Figure 2-16) to

have higher ionic conductivity-gadolinium-doped ceria (GDC)-and compared the behavior of

LSM/YSZ and LSM/GDC composite cathodes, as well as that of pure LSM (Figure 2-16).38









The results show that replacing the YSZ phase with GDC significantly lowers (by

approximately 50%) the polarization resistance at each temperature observed. Note also the

large improvement in behavior of both composite cathodes over pure LSM. A comparison is

also made to show how the type of electrolyte onto which an electrode is applied can influence

performance--the polarization resistance of pure LSM on GDC is 1/3 that of pure LSM on YSZ.

This result indicates that the kinetics of LSM electrochemical reactions are faster on GDC

electrolyte surfaces than on YSZ surfaces. The effect of electrolyte is much smaller in

LSM/GDC composites, but the slight reduction in polarization resistance observed on GDC

electrolytes likely results from replacement of the LSM/YSZ 3PB at the cathode/electrolyte

interface with LSM/GDC 3PB. However, the LSM/GDC 3PB already present in this composite

cathode dominate the electrode performance, hence the small effect of the electrolyte on

performance .

A model incorporating the effects of concentration polarization and activation polarization at

higher current densities (using a Tafel approximation to the Butler-Volmer equation,

ract = a + b ln(i) (2-16)

where a and b are parameters influenced by electrode microstructure and thickness) also predicts

an effective exchange current density which is over an order of magnitude higher for LSM/YSZ

composite cathodes compared to single-phase LSM cathodes at 800C.13

Wu and Liu7 reported the performance of dual phase yttrium-doped bismuth oxide (YSB)-

silver (YSB/Ag) composite cathodes on BaCeo~sGd0.203 electrolytes (Figure 2-17). It is seen in

Figure 2-17a that Ag/YSB shows a much reduced overpotential than Ag-at 40mA/cm2, the

overpotential of YSB/Ag is only 52% that of pure Ag. It is worth mentioning that there are two

inherent problems that limit the performance of pure Ag cathodes. One is its relatively high









thermal coefficient of expansion, yielding poor adhesion to electrolyte surfaces. The other is that

it readily densifies at low temperatures (850-900oC), and results in dense electrodes and hence

poor performance. The addition of a ceramic phase to Ag, in addition to increasing the TBP

length, can prevent electrode densification and improve thermal compatibility and adhesion to

electrolyte surfaces.' Figure 2-17b shows the effect of the volume fraction of Ag in the

composite on cathode performance. A rapid drop (indicative of percolation behavior) is realized

around 55 vol% Ag. If porosity (20 vol%) is taken into account, the vol% Ag and YSB at which

minimum resistance is reached are 44% and 36%, respectively. This agrees well with the EMPT,

but is inconsistent with ambipolar diffusion theory, which is discussed below.

For a specific electrode, a number of conductivities may be defined such as electronic,

ionic and total. It is the simultaneous transport of electronic and ionic species that is crucial in

electrodes that have mixed conduction. This is measured as ambipolar conductivity. Note that

since electron mobilities are much larger than that of ions, ambipolar diffusion predictS37 that for

these dual-phase composites, higher ambipolar conductivities are expected if the volume fraction

of the ion-conducting phase is higher than that of the electron-conducting phase (Figure 2-18),

which is in contrast to the results in Figure 2-17b where the electron-conducting phase volume

fraction is larger. Yet in Figure 2-19, a minimum resistance is reached at higher GDC content in

LSM/GDC dual phase composites. The apparent contrast in behavior can be explained by the

simple fact that the overall performance of a porous electrode is not solely determined by mixed

ionic-electronic transport properties in the solid phase of the electrode, but also by the inherent

catalytic property of the 3PB, as well as by gas transport to or away from the 3PB. One might

conclude, therefore, that the inherent catalytic activity of Ag is higher than that of LSM.









Maguire et al.41 examined the performance of Co-rich (x=0.3) and Fe-rich (x=0.7)

La0.84Sr0.16001-xFexO3 (LSCFe3 and LSCFe7, respectively) as well as Pt cathodes on GDC

electrolytes. Figure 2-20a shows the results of DC electrical conductivity measurements

(Appendix A) for LSCFe3 and LSCFe7 versus temperature. As expected, the Co-rich

composition shows a higher conductivity than the Fe-rich composition--at 8000C, the

conductivities are 643 S/cm and 115 S/cm for LSCFe3 and LSCFe7, respectively. Figure 2-20b

shows the results from steady-state cathodic polarization measurements (Appendix A) for the

two LSCF compositions as well as for Pt. For simplicity the cathodic overpotential and current

densities are expressed as positive quantities. For a given overpotential, LSCFe3 will pass a

higher current than LSCFe7, and as expected both LSCF compositions will pass higher currents

than Pt electrodes, since LSCF is an MIEC.

An ion exchange depth profiling (IEDP) comparison between manganates and cobalt-

ferrites44 (Figure 2-21) shows that oxygen exchange coefficient, k (cm- ) (the rate constant

associated with transfer of oxygen species across a solid/gas interface), and the oxygen self-

diffusion coefficients, D* (cm2/S), are orders of magnitude higher in the cobalt-ferrites,

indicating the relative ease with which such chemical processes proceed in these materials

(Appendix A). These results give credence to the supposition that mixed conducting cathodes

extend the reaction zone beyond the 3PB.

Ishihara et al.31 Studied the effect of the A-site dopant in Ln0.6Sr0.4MnO3

(Ln=La,Pr,Nd, Sm,Gd,Yb, and Y) perovskite cathodes on YSZ electrolytes. Figure 2-22 shows

the results of electrical conductivity, cathodic polarization, and fuel cell power density

(Appendix A) measurements.









In all cases, the Pr substitution exhibits the highest performance. Note that the cathodic

overpotential of PrSM is only 25% that of LaSM at the current density and temperature used for

this measurement. At 1000oC, the maximum power density is nearly the same for PrSM and

LaSM, indicating that the electrolyte is the main contributor to the voltage losses. At lower

temperatures, the difference becomes more pronounced, illustrating the importance of the

activity for the cathodic reaction. The voltage drop due to cathodic overpotential was smaller in

the PrSM cell, allowing it to exhibit the same power density at 700oC as LaSM at 800oC.

Microstructural effects were negligible since BET surface area measurements and particle sizes

(Appendix A) were determined to be nearly the same in both these cases. Chemical effects were

thus predominant. This has been explained in that Pr oxides exhibit a large nonstoichiometry in

air since the stable valence number of Pr ions is intermediate between 3 and 4. In addition,

Steele et al. reported that the activity for the dissociation of oxygen molecules to oxygen ions can

be expressed by the exchange current density, and this strongly influences cathodic

overpotential.31 The slope of cathodic overpotential-current density plots provides a measure of

the exchange current densities (Appendix A). The exchange current density of PrSM was shown

to be nearly three times greater than that of LaSM.

Kharton et al.45 tested a number of different cathode materials on bismuth oxide-based

electrolytes. Bismuth oxide was detected on the Lnl-xSrxCoO3 cathode surface after sintering.

Other solid solution phases involving the diffusion of lanthanum and strontium ion diffusion

were also detected. The interaction of cobaltites with YSZ electrolytes occurs much more slowly

than with bismuth oxide-based electrolytes, even though YSZ electrolytes are sintered 200oC

higher than bismuth oxide, as evidenced by electrode resistances which are 2-4 times larger on

yttrium stabilized bismuth oxide (YSB) electrolytes than on YSZ electrolytes.









In a landmark paper, Shao and Haile46 TepOrted on Bao~sSro~sCoo.sFe0.203 (BSCF) as a new

cathode material for IT-SOFCs. Unlike typical MIEC cathodes, the A-site cation of BSCF is an

alkaline-earth element as opposed to a rare-earth element. Utilizing BSCF on a thin-film

samarium-doped ceria (SDC) electrolyte, high power densities (1010 mW/cm2 and 402 mW/cm2

at 600 oC and 500 oC, respectively) were obtained when operated with humidified hydrogen as

the fuel and air as the cathode gas. The cathode ASR, as determined symmetrical cell testing,

was 0.055-0.071 OZcm2 at 600 OC, and 0.51-0.60 OZcm2 at 500 OC. The excellent performance of

this material is reportedly due to its exceptionally high oxygen diffusivity, especially at lower

temperatures, allowing for high rates of oxygen electro-oxidation.

2.5 Strategies for Improving Cathodic Performance

In order to attack the problem of improving cathode performance, one must not only

understand the various mechanistic steps involved in the oxygen reduction reaction but must also

be able to identify the slowest (rate-determining) step. In addition, one must have a viable

strategy toward accelerating this step. These strategies may be broken down into two

categories-morphological improvements, which involve changes to cathode geometry or

microstructure and electrocatalytic improvements, which encompass all other improvement

methods. There are many approaches to these problems; a few examples are discussed below.

Tsai and Barnett9 illustrated several methods for identifying mass-transport limitations in

cathode materials. A voltage/power density versus current density plot is conducted first in air,

then in pure Oz. If the slope of the voltage-current density plot, i.e. the resistance, is larger in air

than it is in pure Oz, maSs-transport limitation can be assumed. Note that in the case of MIECs

the ionic conductivity of the cathode may be pO2-dependent, hence this behavior may not

necessarily be indicative of mass-transport limitation. Mass-transport limitations can arise due to









surface diffusion, gas diffusion, dissociative adsorption9 Or bulk transport.38 Both dissociative

adsorption and surface-diffusion are thermally activated processes that should have strong

temperature dependence, while gas diffusion will not. Thus, if any limiting current that is

present increases as temperature is increased, one of the first two steps may be the limiting

factor. Gas diffusion will depend on factors such as porosity and pore size. Thus, if any limiting

current that is present increases as porosity and pore size increase, gas diffusion may be the

limiting factor. These morphological factors can thus be utilized to accelerate the gas diffusion

process. Also, if a change in test apparatus flow rates changes the polarization properties of the

test cell, gas diffusion may be the limiting factor.20 For dual phase composite cathodes, if high

frequency arcs are observed in the impedance spectrum whose behavior is independent of pO2,

bulk ionic conduction in the ionic conducting phase may be the limiting factor. This can be

confirmed by a comparison of the activation energy (Appendix A) of the resistance of this high

frequency arc in the composite cathode to that for ionic conduction in the pure ceramic phase. If

these values are similar, bulk diffusion can be considered the limiting factor.38

Godickemeier et al.33 discussed how overpotential-current density relationships (Appendix

A) may behave under a pure diffusion-limited process, a pure charge-transfer controlled process

and combination of charge-transfer control at low current densities and diffusion control at high

current densities (Figure 2-23). It is seen that there is a Butler-Volmer-type behavior for pure

charge-transfer control, and a limiting current for pure diffusion control.

For dual phase mixed conducting cathodes, one morphological approach to enhance the

charge transfer rate is to reduce the particle size of the ionic conductor while keeping the

electronic conductor particles relatively large. The smaller particles form more contact points

with the larger particles, creating a greater density of reaction sites within the cathode compared









to the case when all particles are approximately the same size. This effect is illustrated in Figure

2-24 where a particle broken into four equal pieces will have twice the 3PB length per unit area

(8L) as the unbroken particle (4L).

For single phase mixed conductors, a wide particle size distribution is desired.13 Small

particles are advantageous since they tend to create large grain boundary areas. These grain

boundaries are high-energy sites, where the charge transfer reaction preferably occurs.13

However, small particles have the disadvantage of creating large tortuosity (harmful to gas phase

diffusion) and poor connectivity.13 One electrocatalytic approach to improving charge transfer

kinetics is to dope the electrode with redox cationS47 Such as Tb+4/Tb+3 Or Pr+4/ Pr+3

Arrhenius plots of log(R) versus 1/T, where R is the cathode resistance of interest, yield

straight lines with slopes that depend primarily on the activation energy of the rate limiting

process. A change in slope over the investigated temperature range thus indicates a change in

the rate controlling process. An investigation OfLal-xSrxCoO3-a cathodes (x=0.2, 0.3, and 0.4)

on Lal-xSrxGal.7MgO3-a (LSGM) electrolytes utilizes Arrhenius plots of the logarithm of

interface conductivity (defined as the reciprocal of area-specific polarization resistance) as a

function of temperature for different applied cathodic voltage (E) conditions (experimental

details are discussed in Appendix A) to help elucidate that the rate limiting step in these cells

was due to bulk diffusion of oxygen ions in the LSC cathode as opposed to surface diffusion of

adsorbed oxygen on the LSC surface. The general procedure and results are discussed, briefly,

below.

Figure 2-25 shows the Arrhenius behavior at E=0V. Since the activation energies for all

samples are similar, it can be assumed that similar reaction mechanisms and the same rate-

limiting step are at play. Figure 2-25b shows the Arrhenius behavior (for x=0.2 and x=0.4) for










applied cathodic voltage conditions. It is seen that the activation energies for the lower Sr

content sample are more sensitive to changes in applied voltage. As described in Section 2.3.3,

under cathodic polarization conditions, the pO2 at the cathode/electrolyte interface is lowered

and the oxygen vacancy concentration in the cathode increases due to defect equilibrium

requirements. The driving force for surface diffusion depends on the oxygen partial pressure

between the surface of the cathode grains and the reaction sites at the electrolyte surface. Since

this gradient is generated by the applied cathodic overpotential, it should be similar for all

samples (x=0.2, x=0.3, and x=0.4) at a given value of cathodic voltage. Thus, if this step were

rate limiting, one would expect to see the similar activation energies for x=0.2 and x=0.4 as at

each voltage. This behavior was not observed, thus surface diffusion is determined not to be the

rate-limiting step.

Bulk diffusion depends on oxygen vacancy concentration. As mentioned above, applied

cathodic voltages tend to increase the concentration of vacancies in the bulk of the cathode.

Changes in vacancy concentrations (on a percentage basis) for LSC cathodes with lower

concentrations of A-site dopants may be more sensitive to changes in applied voltage than for

those with high A-site dopant concentrations. As a result, the activation energies will be more

sensitive to changes in voltage as well. Since this is the experimentally observed behavior, this

is assumed to be the limiting step.

The surface oxygen exchange rate, k, determined by ion exchange depth profiling (IEDP)

can be used to evaluate the surface activity of a cathode (Appendix A). A comparison of the

activation energy of the k values with the activation energy of interface conductivity values at

E=0V (described above) can clarify whether or not oxygen adsorption/desorption on the cathode

surface with charge transfer and oxygen incorporation at the cathode surface is rate-limiting. If









these activation energies are similar, this step may be considered as rate-limiting. Oxygen

surface exchange rates can be increased by increasing oxygen vacancy concentrations,48 Or by

doping the cathode with a noble metal49-51 Since noble metals are known to be good catalysts for

the oxygen reduction reaction.

The exchange current density (which can be determined experimentally as described in

Appendix A) is related7,33 to pO2 at a given temperature by

io = k(pO2 /) (2-17)



where k is a constant which is independent of pO2 and y is an exponent that depends on the

reaction mechanism and rate-limiting step. A plot of log(io) versus log (pO2) Should thus give

insight into the rate-controlling mechanism at a given temperature of interest.

2.6 Summary

The above discussion detailed background information on the basic principles of SOFCs,

highlighted the importance of the cathode for reduced SOFC operating temperatures, and

presented the three basic classes of SOFC cathodes--single phase electronic conductors, dual

phase composite cathodes, and single phase MIEC conductors. In addition, conventional and

state-of-the-art cathodes were presented, as were several methods by which the dominant

mechanism for observed cathodic losses can be identified. Not all methods and materials

described above were used in the following studies, but a general awareness of these facts

provides an essential framework from which the studies could be built.























































Figure 2-2. Theoretical values of oxygen partial pressure on the anode (calculated from Eq. 2-8)
and cell potential (calculated from Eq. 2-4) as a function of temperature. Fuel is
hydrogen bubbled through water at room temperatures.


E~i3~~ Electr lyeg








Figure 2-1. Schematic representation of a typical SOFC button cell.


1.24

1.22
1.2

1.18
1.16

1.14
1.12
1.1

1.os
1.06

1.04


O nn




a o
n "O


n ~O


1000


0

-10

-20

-30

-40 O
-50

-60

-70

---80

-90
1200


0 200 400 600 800

Temperature (oq)














Cathode

~c Electrolyte

~ Anode
Interconnect


Figure 2-3. Schematic representation of an SOFC stack.


Nernst Potential
(t=1 & gas-thght sealing)
theoretical max efficiency


____________________ : _Cell overpotential



Cell output potential

Load current
density


'Activation
.Polarization

Ohmic
Polarization


'Concentration
.Polarization


1.0 -

0.8 -

0.6

0.4 -

0.2 -


Cell Current Density (A/cm2)
Figure 2-4. Illustrative representation of typical current-voltage profile for an SOFC illustrating
different types of electrochemical losses that arise when current is drawn from the
cell .


Open Circuit Potential





Figure 2-5. Cubic fluorite (AO2) CryStal structure with A-site cations occupying the cube corners
and face centers, and oxygen anions occupying the tetrahedral interstitial sites.


400 T ("C) 300


900 800 700


600 500


0.5
0
-0.5






U -2.5


S-3.5
-4

-4.5


0.8 0.9 1 .11.2 1.3 1.4 1.5 1.6 1.7 1.8
1000/T (K-l)
Figure 2-6. Reported ionic conductivity values for common SOFC electrolyte materials. [Figure
adapted from Solid State lonics Vol. 75, B.C.H. Steele, Interfacial reactions
associated with ceramic transport membranes, 157-165 (1995) with permission from
Elsevier.]















0.8 4000




0.2 1 1000


0 5f000 1000~0 15000 20000 25000 30000
Cuarent denalty trArm)
--*-CE0I voliage -r-s Ohmic pola~ristion
-Anode~ athallon polarisation -rc--athode actiation palarisation
-F-Anodcn~centration poladsation -s- athode cncentration polarisatior
-a-POrnr dBAsilV
Figure 2-7. Calculated cell performance and polarization resistances. [Reprinted from Journal of
Power Jources, Vol. 93, S.H. Chan, K.A. Khor, Z.T. Xia, A complete polarization
model of a solid oxide fuel cell and its sensitivity to the change of cell component
thickness, 130-140 (2001) with permission from Elsevier.]








e- conducting

particle


3PB line



electrolyte


Figure 2-8. Schematic representation of three phase boundary lines between electronic
conducting particles, the ion conducting electrolyte and the gas phase for single-phase
electronic conducting cathodes.











ion conducting
:::Mlli::ipa article

3PB line
e- conducting
particle


electrolyte

Figure 2-9. Schematic representation of three phase boundary lines between electronic
conducting particles, ion conducting electrolyte as well as ion conducting particles,
and the gas phase for two-phase composite cathodes.


A B
Figure 2-10. ABO3 cubic perovskite crystal structure. (a) A-site cations at cube center, B-site
cations at corners, and oxygen anions at face centers. (b) Black spheres--A-site
cations, grey tetrahedra--oxygen anions at corners, B-site cations at center.




























Figure 2-11i. Schematic representation of various reaction pathways for the oxygen reduction
reaction in a MIEC.


1oo 100.0 10KLD


P1

9

a
~s~


Figure 2-12. Effect of cathode thickness on R~f [Reprinted from Solid State lonics, Vol. 131,
A.V. Virkar, J. Chen, C. W. Tanner, J. W. Kim, The role of electrode microstructure
on activation and concentration polarizations in solid oxide fuel cells, 189-198 (2000)
with permission from Elsevier.]


6 0 nam)
0.01 01L 1.0





Of --g a- r i I
-3 -2 -1 0 1 2


0 '1 2 3
Current Density (A/cm2)


Figure 2-13. Effect of cathode thickness in I-V performance. [Reprinted from Solid State lonics,
Vol. 131, A.V. Virkar, J. Chen, C. W. Tanner, J. W. Kim, The role of electrode
microstructure on activation and concentration polarizations in solid oxide fuel cells,
189-198 (2000) with permission from Elsevier.]


1.4



IQ-

nrc"
QB


~ba: ,,,
6~


I -fu~aim~l


Figure 2-14. Agreement between model (Eq. 2-15) and experiment for LSM-YSZ composite
cathodes. [Reprinted from Solid State lonics, Vol. 131, A.V. Virkar, J. Chen, C. W.
Tanner, J. W. Kim, The role of electrode microstructure on activation and
concentration polarizations in solid oxide fuel cells, 189-198 (2000) with permission
from Elsevier.]





0 2 4 6 8 10 12
Cathode thichaess / pmn

Figure 2-15. Effect of cathode thickness on polarization resistance for LSM-YSZ composite
cathodes. [Reprinted from Solid State lonics, Vol. 86-88, M. Mogensen and S.
Skaarup, Kinetic and geometric aspects of solid oxide fuel cell electrodes, 1151-1160
(1996) with permission from Elsevier.]


Matenal Temaerature IoC) lonic Conducrlvlv ty cm,-1
YSZ 600 3.2X10-3
GDC 600 3.1X10-'2
YSZ 750 1.8X10-2
GDC 750 7.9X10-2


Eletroyte Cathode Polarization Resistance (Cem2)


"IZLSM-ilDCSO Og M gec

5 as DasIP





LSM 10.82S



Figure 2-16. Effect of electrolyte phase composition on polarization resistance of various
conventional SOFC cathode materials. [Reprinted from Solid State lonics, Vol. 143,
E. Perry Murray, S.A. Barnett, (La, Sr)MnO3-(Ce,Gd)O2-x COmposite cathodes for
solid oxide fuel cells, 265-273 (2001) with permission from Elsevier.]


3-


E























0 10 20 50 40 50 60 70
Call Curret (aA/ #








40 ~ ~ ~ a 50 60 7 80 100






Volume Practionl of Silver (%)

Figure 2-17. Effect composition on (a) cathode overpotential, and (b) interfacial resistance for
Ag-YSB composite cathodes. [Reprinted from Journal of the American Ceramic
Society, Vol. 81, Z. Wu, M. Liu, Ag-Bi1.5Yo.5O3 COmposite cathode materials for
BaCeo~sGd0.203-based solid oxide fuel cells, 1215-1220 (1998) with permission from
Blackwell Publishing.]




















rP"O









6 aca e Faume ar ton of s to a ap--an

Figure 2-18. Theoretical ambipolar conductivity from EMPT. [Reprinted from Solid State
lonics, Vol. 93, Z. Wu, M. Liu, Modelling of ambipolar transport properties of
composite mixed ionic-electronic conductors, 65-84 (1996) with permission from
Elsevier.]






-*--+- soot rate~8









l I a `' l e I I--




30 35 40 45 50 55 00
wneight% GDC

Figure 2-19. Effect of composition on interfacial resistance of GDC-LSM composite cathodes.
[Reprinted from Solid State lonics, Vol. 143, E. Perry Murray, S.A. Barnett,
(La, Sr)MnO3-(Ce,Gd)O2-x COmposite cathodes for solid oxide fuel cells, 265-273
(2001) with permission from Elsevier.]













e~___L____
CSEFe3


"-r__
~S~Fo7


O.0005

20000


0 001


0.00T5


0.002


0.0025


0 100 20 300


400 500


Figure 2-20. (a) Arrhenius conductivity plot of selected Fe loadings in LSCF, and (b) cathode
polarization performance comparison. [Reprinted from Solid State lonics, Vol. 127,
E. Maguire, B. Gharbage, F.M.B. Marques, J.A. Labrincha, Cathode materials for
intermediate temperature SOFCs, 329-335 (2000) with permission from Elsevier.]









I I


Depth (x1 0-6 cm)



D* = 1x10-7 CB72S-1
-k = 3x10-6 cm s-l


La0.6Ca0.4GO, gFe0.203-x


L


D* = 3x10
k = 9x10-8


5MnO3-x
-13 gg72S-1
Scm-l


T = 800 oC
t = 400 s


0.00


0.01


0.04


0.02 0.03
Depth (cm)


B
Figure 2-21. IEDP for (a) LSM and (b) LCCF. [Reprinted from Journal of Power Sources, Vol.
49, B.C.H. Steele, Oxygen transport and exchange in oxide ceramics, 1-14 (1994)
with permission from Elsevier.]












2 5 0.4 YL


20

315 era
vb o 01o p


0 7 0 8 0 9 1 00 1 10 1 20 1 30 0.0
1000/T (K-l) 0.0 0.5 1 .0 1 .5 2.0
A Current density (Acm-2) B
400 .
E Cathode
() LaneSro*MDO3
E 300 g Pro *04~MDOJ


-0 200
L

E 100



900 1000 1100 1200 1300
Temperature (K)
C
Figure 2-22. Effect of A-site dopant on (a) electrical conductivity, (b) cathodic polarization, and
(c) fuel cell power density for various Ln0.6Sr0.4MnO3 (Ln=La,Pr,Nd, Sm,Gd,Yb, and
Y) perovskite cathodes on YSZ electrolytes. [Reprinted from Journal of the
Electrochemical Society, Vol. 142, T. Ishihara, T. Kudo, H. Matsuda, Y. Takita,
Doped PrMnO3 perOvskite oxide as a new cathode of solid oxide fuel cells for low
temperature operation, 1519-1524 (1995) with permission from ECS.]










0.3 ~ -- 0,3






0.0 '- **0.0 '
0.0 0.1 0.2 0.3 0.4 0.5 0.0 0.1 0.2 0.3 0.4 0.5
Ji [A/c~mq Ji A Bm
0.4 CT anni~c alffusi


f 0.3 T nhrelecule~flus

control~
0.1 - -
contro
0.0 '
0.0 0.1 0.2 0.3 0.i4 0.i5
J= IA/cm 1

Figure 2-23. Overpotential-current density curves for (a) diffusion, (b) charge transfer, and (c)
mixed control. [Reprinted from Solid State lonics Vols. 86-88, M. Goidickemeier, K.
Sasaki, L. J. Gauckler and I. Riess, Perovskite cathodes for solid oxide fuel cells
based on ceria electrolytes, 691-701 (1996) with permission from Elsevier.]


Figure 2-24. Effect of electron-conducting particle size on 3PB length.












I I 1 I I I

La1-x~rx~oO3-d


I .I I. I I

Mi~g~g~r0.2Coo3 -d




E=Q,4V.

\tE=4.3V
bE=-0) 2V

E=-. 1V


0.80 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20
T-1/k (K-1)



"La0. 68 "0.4Coo3-d
0.8 -

0.4;


E--0 4V
-E=-0 3
-0.41 -E=-01 V
-0.81 E=0V

-1.2=

0.*30 0.95 1.00 1.05 1.10 1.15 1.20 1.;
T 1/k (K-1)


0.8

0.4

E


u-0.4

o1


E=0V


0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25
T-1/k (K-1)


Figure 2-25. Arrhenius behavior of interface conductivity for LSC cathodes (a) varying Sr
content at E=0V, and varying voltage for (b) Lao.sSr0.2CoOO3- and (c) La0.6Sr0.4oOO3-
a. [Reprinted from Electrochimica Acta, Vol. 46, T. Horita, K. Yamaji, N. Sakai, H.
Yokokawa, A. Weber and E. Ivers-Tiffee, Oxygen reduction mechanism at porous
Lal-xSrxCoO3-d cathodes/Lao~sSr0.2Gao.sMg0.202.8 electrolyte interface for solid oxide
fuel cells, 1837-1845 (2001) with permission from Elsevier.]











CHAPTER 3
STABILITY OF SILVER-BISMUTH OXIDE CATHODES

3.1. Introduction

The primary function of SOFC cathodes is to aid in the oxygen reduction reaction, Eq. 2-1.

It is evident from this equation that oxygen reduction must occur where the gas phase, the

cathode, and the electrolyte are in intimate contact. These boundaries form one-dimensional

lines called three phase boundary (3PB) lines. As shown in Fig. 2-8, the reaction zone is

restricted to a narrow region at the electrolyte/cathode interface. The reaction zone can be

extended to a three-dimensional volume above this interface by introducing a second, ion-

conducting phase, Fig. 2-9. The performance of these dual phase cathodes is significantly

improved due to the increase in available sites for oxygen reduction.

A well-behaved cathode must be a good catalyst for oxygen reduction, highly conductive,

and chemically and thermally compatible with the electrolyte. In addition, for mobile

applications, the cathode must be stable at operating temperature for over 40000 hours.52

As described in Section 2.4 (Eq 2. 15), increasing the ionic conductivity of the ion-

conducting phase should improve cathode performance, since higher ionic conductivities allow

ions produced deeper within the electrode to reach the electrolyte, thus effectively increasing the

width of the reaction zone. Due to the high concentration of oxygen defects present in its crystal

structure, stabilized bismuth oxide exhibits excellent ionic conductivity, ranking among the

highest of all fast-ion conductors at 600 oC.22 Thus, dual phase composite cathodes utilizing

some metal, Me, as the electronically conducting phase and a bismuth oxide-based material as

the ionically conducting phase are expected the exhibit low resistance compared to similar dual

phase cathodes containing the same metal Me and some other ceramic phase with lower ionic









conductivity. It should be noted that below 600 oC, stabilized bismuth oxide undergoes defect

ordering, where the conductivity rapidly decreases with time.53 The exact transition temperature

has not yet been fully resolved.

Silver is known to be an excellent catalyst for the oxygen reduction reaction, has high

electronic conductivity, and is relatively inexpensive. Composites of silver and YSB exhibit

some of the lowest area-specific resistance (ASR) values known to date for YSZ electrolyte

substrates, over the temperature range of interest.39 However, the long-term stability of these

materials remains unchecked.

In this study, the long-term stability of Ag-YSB and Ag-ESB composites were examined.

Ag-YSB composites were chosen to directly study the long-term stability of the system reported

in the literature.39 Ag-ESB composites were studied for verification purposes and to improve

performance (due to the higher ionic conductivity of ESB compared to YSB).54 YSB (for Ag-

YSB composites) and ESB (for Ag-ESB composites) were chosen as the electrolyte substrates,

since these should be chemically and thermally compatible with their respective composite

cathodes, and to help identify possible ordering phenomena or phase transformation of the

stabilized bismuth oxide phase. For comparative purposes, the composition used in this study

was based on that used by Xia et. al.39 The isothermal annealing temperature was chosen to be

650 oC to avoid defect ordering in the bismuth oxide phase. It should be noted that at this

temperature, YSB undergoes a cubic to rhombohedral phase transformation,55,56 but its onset is

somewhat sluggish and can be monitored by impedance measurements.

3.2. Experimental

YSB (Bio.75YO.2501.5) and ESB (Bio~sEr0.20t~s) powders were prepared by conventional

solid-state synthesis. Stoichiometric amounts of bismuth oxide (Alfa Stock #10658) and yttrium

oxide (Alfa Stock #11181) or erbium oxide (Alfa Stock #11309) were ball-milled in ethanol for









24 h. The solutions were then dried at 90 oC for 4 h, calcined at 800 oC for 15 h, crushed and

sieved (325 mesh size). These powders were uniaxially pressed at ~70 MPa using a cylindrical

V/2" die, isostatically pressed at 250 MPa, and sintered at 890 oC for 15 h to form electrolyte

support pellets. All pellets measured ~1.1 cm in diameter. ESB and YSB pellets measured ~3.1

mm and ~3.6 mm in thickness, respectively. The relative porosities of the ESB and YSB pellets

were 90% and 85% of theoretical, respectively, as measured using ImageJ software (1.37v,

Wayne Rasband).

Composite Ag-bismuth oxide cathode inks were prepared by adding Ag (Alfa Stock

#41598) to the sieved oxide powders (volume ratio of Ag:bismuth oxide = 60:40). Ethanol and

organic vehicles (a-terpineol as the solvent, di-n-butyl phthalate as the plasticizer, and

polyvinylbuteral as the binder) were added (generally in a 3:1:2 volume ratio) until a viscosity

appropriate for brushing was achieved. A similar recipe was followed for preparing the pure

silver electrode ink. Symmetric cells were prepared by paint brushing the cathode slurry to each

side of the bismuth oxide pellets, drying at 120 oC for 1 h, and firing at 750 oC for 1 h.

Electrochemical impedance analysis was performed using a Solartron 1260

Impedance/Gain-Phase Analyzer. Silver mesh current collectors and leads were pressed against

the samples in a quartz reactor by the use of a ceramic screw-and-bolt assembly. Cell

temperature was maintained at 650 oC for 100 h and monitored by a thermocouple flanking the

sample. Cell response was measured over a frequency range of 32 MHz to 0. 1 Hz, with an AC

voltage amplitude of 100 mV. Cell responses were not corrected for porosity effects.

Electron probe microanalysis (EPMA, JEOL Superprobe 733) was performed on a sample

with pure silver electrodes to determine if any diffusion of silver into the electrolyte had

occurred during testing. A sample was annealed at 750 oC for 48 h, encased in epoxy resin and









set for 24 h. The sample was then polished down to 1 Cpm, and a linescan was performed across

the electrolyte/cathode interface--the atomic percentage of each element present recorded at

each position in the scan.

Cathode morphology was characterized by the use of a JEOL model JSM-6335F field-

emission scanning electron microscopy (FESEM). Samples which had undergone long-term

stability testing were compared against virgin samples to detect any visible microstructural

changes that might have occurred during testing.

X-ray diffraction (XRD Philips model APD 3720) using Cu Kocl (h = 1.54056 A+) was

employed to determine inter-phase reactivity between silver and ESB. The diffraction patterns

of mixed raw powders were compared against mixed powders which were annealed at 750 oC for

48 h.

3.3 Results and Discussion

3.3.1. Compositional Optimization

The first goal of this study was to produce silver-bismuth oxide cathodes with ASR values

comparable to those reported by Xia et al.39 Figure 3-1 shows the impedance spectra and

corresponding trend in cathode ASR over a range of composition for the Ag-ESB system at 600

oC. A minimum value observed at 50 vol% ESB. This result is reasonable since, for a two-phase

composite whose particle sizes are roughly the same, and assuming the solid phases and pores

are randomly distributed and the porosity is open and sufficiently large, the 3PB length should

reach a maximum if the two solid phases are present in equal fractions of the overall electrode

volume. As discussed previously, 3PB length is important in two-phase composite cathode

systems since the oxygen reduction reaction that is occurring can only proceed at sites where all

three reactants are present--as the cathode 3PB length increases, so do the number of reaction

sites, and hence cathode activation polarization decreases.









Electrode area specific resistance (ASR) was calculated from


R,A
ASR = (3-1)


where Rp is the electrode polarization discussed above, A is the electrode area, and the factor '2'

is used to account for the fact that each cell contains two (approximately identical) electrodes.

Note that electrode area does not account for electrolyte surface variations due to surface

porosity.

Figure 3-2 shows the comparison of the ASR obtained in this study for Ag-ESB cathodes

(on an ESB electrolyte) with the results adapted from Xia et al.39 for Ag-YSB (on YSZ

honeycomb electrolytes) and other reported cathodes. The Ag-ESB composite cathode has an

ASR of 0. 18 OZcm2 at 600 oC, making it one of the lowest-resistance electrode systems reported

to date and a significant improvement over that produced by Xia et al.39 who obtained an ASR of

0.3 OZcm2 at 600 oC for their Ag-YSB composite. Note that two samples of the Ag-ESB

optimized composition were tested to confirm reproducibility. The second sample had a slightly

higher value of ASR (0.20 OZcm2 at 600 OC), but was still significantly lower than the Ag-YSZ

literature value.

3.3.2. Stability: Thermal Performance Decay due to Microstructural Evolution

Having met and exceeded the benchmark for cathode performance obtained in the

literature, the long-term stability of these composites could now be assessed. Impedance spectra

for the long-term testing of pure Ag, Ag-ESB and Ag-YSB systems at 650 oC are shown in

Figure 3-3, with the data Rs-corrected for ease in comparison. That is, the high-frequency real-

axis intercept (bulk resistance or Rs) of each spectrum has been subtracted from the real

component of each data point in the spectrum. Plotted this way, the low-frequency real-axis









intercept corresponds to the electrode polarization resistance, R,. It is clear from this figure that

the electrode polarization for each system increases with time.

Electrode ASR (calculated using Eq. 3-1) is plotted against time for all systems at 650 oC

in Figure 3-4. The initial zero time value of ASR for the Ag-YSB and Ag-ESB systems were

similar--0.04 OZcm2 and 0.06 OZcm2, TOSpectively. Surprisingly, the Ag-YSB electrode system

exhibited lower resistance than the Ag-ESB system, despite the higher ionic conductivity of the

ESB phase. This could be due to surface porosity effects discussed above, but other

microstructural factors will be addressed below. Also, the resistance of these electrodes is

significantly lower than that reported by Xia for their Ag-YSB system at the same temperature

(greater than 0. 1 Ocm2), again, making it one of the lowest-resistance electrode systems reported

to date. Differences in processing routines could lead to variations in electrode porosity,

thickness, relative particle sizes, and inter-particle necking, accounting for the observed

differences in performance between the two studies and between the two electrodes Ag-YSB and

Ag-ESB. In addition, since the present electrode study was performed on electrolytes having the

same composition as the ionic conducting phase in the electrodes, the interfacial (electrode-

electrolyte) resistance may be lower than similar electrodes placed on YSZ electrolytes.

All systems experience significant increases in ASR during the 100 h stability

experiment--both the Ag-YSB and Ag-ESB electrode ASR values increased by around 70%

(from 0.04 OZcm2 to 0.07 OZcm2 (75%) and from 0.06 OZcm2 to 0.10 OZcm2 (67%) for Ag-YSB and

Ag-ESB respectively), while the pure Ag system experiences a near fourfold increase (from 0.92

OZcm2 to 3.55 OZcm2) during the same length of time. Linear regression of the data (based on the

first 10 h of testing) yields a degradation rate of 4.4x10-4 GZCm2/h for Ag-ESB and 1.2x10-4

OZcm2/h for Ag-YSB. The pure Ag electrode degrades at a rate more than two orders of










magnitude faster--8.7x10-2 GZCm2/h. It should also be mentioned that the rate of change in ASR

with time is linear for Ag-ESB electrodes, while for the Ag-YSB system deviates slightly from

linearity beginning approximately 50 h after commencement of testing. For the pure Ag system,

this rate of change is strongly non-linear and the non-linear trend is in the opposite direction

compared to the Ag-YSB system. Various possible causes for these observations are addressed

below.

Figure 3-5 plots the relative conductivity (from Rs) versus time for both bismuth oxide

electrolyte systems. Relative conductivity is calculated as the ratio of the electrolyte

conductivity (determined from impedance spectra high-frequency intercepts) at time, t, to the

initial conductivity value at t=0. Both electrolytes seem to be relatively stable--both

maintaining over 90% of their initial conductivities-however, YSB experiences significant

decay in conductivity starting 50 h into testing, while ESB underwent a slight increase in

conductivity (+3.0x10-5 Scm- /h). The trend for YSB is consistent with results reported by Fung

and Virkar56 where YSB undergoes a cubic-rhombohedral phase transformation at this

temperature. Post-mortem XRD studies confirm formation of a rhombohedral phase. Jiang and

Wachsmanls showed that ESB undergoes no such phase transformation, which is consistent with

the trend for ESB in Figure 3-5.

Since the resistance of bulk-related phenomena of both systems remained relatively

constant during the first half of the stability test, it may be concluded that little if any defect

ordering has occurred in the YSB and ESB electrolytes up to 50 h. It follows that minimal defect

ordering has occurred in the bismuth oxide phase of each cathode over this period. Further,

during this initial 50 h regime, the trend in electrolyte conductivity is slightly upward for ESB,

and slightly downward for YSB. This trend is opposite to that observed for cathode ASR values,









where the cathode containing YSB decayed at a slower rate than that containing ESB, suggesting

the impact of the bismuth oxide phase on cathode degradation is minor at this temperature. In

addition, the ASR of the pure Ag electrode rises at a much faster rate than either composite

system. Thus, it is suspected that the silver phase is the maj or source of the observed increase in

electrode polarization. However, the more dramatic decay in YSB conductivity at longer times

may cause the observed non-linear increase in ASR with time for the Ag-YSB system (Fig. 3-4),

as the same phenomenon observed for the electrolyte conductivity may be occurring in the YSB

phase inside the composite electrode.

EPMA, XRD, and FESEM analysis were performed on several samples to elucidate the

compositional and microstructural changes that might transpire during testing. For EPMA

measurements, Y203 and Bi203 Standards were prepared by pressing and sintering pellets of the

commercial powders. Figure 3-6 shows the results of EPMA testing on a YSB sample prepared

with pure silver electrodes and annealed at 750 oC for 48 h. The atomic percent of all elements is

consistent with what is expected on both sides of the interface. Thus, it is concluded that no

silver has diffused across the interface into the electrolyte.

X-ray diffraction patterns of mixed silver and erbium-stabilized bismuth oxide powders

(ESB) before and after heat treatment at 750 oC for 48 h are shown in Figure 3-7. The patterns

reveal no evidence of inter-phase reactivity after heat treatment. Since silver neither diffused

into nor reacted with the electrolyte phase at 750 oC, it is reasonable to assume the same is true

for all samples tested at 650 oC.

Microstructural comparisons of the three different electrode systems before firing, after

firing, and after testing at 650 oC for 100 h are shown in Figures 3-8 through 3-10. Energy

dispersive spectroscopy (EDS) confirmed that for backscattered SEM images, the lighter phase









was ESB and the darker phase was Ag. All three systems exhibit a dramatic growth of the silver

phase after firing at 750 oC for 1 h. However, the silver phase in the composite electrodes show

significantly less growth and coalescence compared to the pure Ag system, as the ceramic phase

helps restrict the migration and ultimate agglomeration of the silver phase. After 100 h of testing

at 650 oC, the pure Ag electrode appears fully dense. The porosity of the Ag-ESB electrode

decreased significantly after testing. Domains of coalesced silver are evident, and ESB particles

between these domains have been forced closer together, further reducing porosity and 3PBs.

Growth and coalescence of the silver phase is also evident in the Ag-YSB electrode though to a

lesser extent. It is believed that the smaller starting size of the YSB particles (as evidenced in the

unfired electrode micrographs) helped to lower the mobility of the silver phase by providing the

electrode with a higher surface area through which the silver phase must migrate.

Although the primary function of the silver phase to conduct electrons toward the 3PBs is

unaffected by these microstructural changes, increased grain size correlates with reduced 3PB

lengths, and hence a smaller reaction zone for oxygen reduction. Further, increased density can

lead to concentration polarization effects, where the reaction species (oxygen admolecules) are

being supplied to the reaction sites more slowly than they are being consumed. The trends in

electrode ASR shown in Figure 3-4 can be rationalized in terms of these microstructural

observations. Pure Ag undergoes the most dramatic densification during testing, and also

exhibits the most dramatic rise in ASR with time. It is reasonable to expect that as the

densification process of Ag nears completion, changes in porosity and 3PB lines will slow, and

hence the electrode ASR will level off at longer times, as was observed. The composite

electrodes experience a much less severe change in microstructure after firing and during testing,

particularly the Ag-YSB system. This is reflected in the two-order slower rate of ASR increase









for these composite systems, and the slightly slower rate for Ag-YSB compared with Ag-ESB.

The finer silver phase microstructure of the fired, untested Ag-YSB electrode relative to the

untested Ag-ESB electrode may also explain why the electrode containing YSB exhibited the

lower initial zero time ASR value despite the higher ionic conductivity of ESB.

Transgranular platelet structures similar to those observed by Fung55,56 were found in the

YSB phase of the electrode and in the electrolyte near the electrode interface (not shown),

consistent with the assertion that a cubic-rhombohedral phase transformation occurred during

testing. This phase transformation likely led to the observed drop in electrolyte conductivity and

rise in electrode ASR at longer times, both of which seem to begin around 50 h.

From the above observations, it is concluded that the main source of instability of the

composite Ag-bismuth oxide electrodes is the migration of the silver phase, rather than reactivity

or defect ordering of the stabilized bismuth oxide phase. Moreover, silver migration was

demonstrated to be even more severe when operating under an applied bias 5. This issue as

well as microstructural approaches toward improving the stability of these electrodes is discussed

Chapter 4 of this dissertation.

3.4 Conclusions

Silver-bismuth oxide composite cathodes have been prepared which perform as well as

(and even better than) similar composites presented in the literature. The long-term isothermal

stability of pure silver, Ag-ESB, and Ag-YSB electrodes was examined. Each system exhibited

significant increases in electrode area specific resistance during 100 h of testing at 650 oC under

open-circuit conditions. On the basis of electrochemical impedance as well as chemical and

microstructural analysis, it is concluded that Ag-bismuth oxide composites have inadequate

microstructural stability for long-term, IT-SOFC cathode applications. Microstructural evolution

of the silver phase is deemed to be responsible for degradation in performance over time due to










grain growth and electrode densification. However, since the microstructural evolution of the

composite electrodes was not as severe as that of pure silver, and since the Ag-YSB composite,

at least initially, was more stable than the Ag-ESB composite, it is likely that further

improvements in stability for the composite system will result from replacing the relatively large

bismuth oxide particles with smaller particles, or by infusing inert nano-sized oxide particles into

the microstructure. This concept will be explored in Chapter 4 of this dissertation.





*


70%


N: -0.2


-0.1

2.7





0.6


**~,


2.9 3.1 3.3 3.5 3.7 3.9


"]0.4



S0.2


0
30 40 50 60 70

Ag content (vol%)


Figure 3-1. (a) Impedance spectra and (b) electrode ASR vs silver content of Ag-ESB showing
compositional optimization at 600 OC, between 40-70 vol% Ag using 5 vol%
composition steps.




















~iAg-ESB



"I 61- e- gdradd cathode. C..Ka
It" 1 d '-0- G~Dd~Ct..S, E.P.Munny

-A~- graded cathode. NT.Hart "
-- YS-LISM PMry
0rI 9 .0l 1.1 12a 1.3

1000tT, K'


Figure 3-2. Arrhenius plot, adapted from Xia et. al.,39 COmparing the ASR of Ag-ESB (this work)
with the best cathodes reported in the literature at the start of this study. [Adapted
from Applied Physics Letters, Vol.82, C. Xia, Y. Zhang and M. Liu, Composite
cathode based on yttria stabilized bismuth oxide for low-temperature solid oxide fuel
cells, 901-903 (2003) with permission from American Institute of Physics.]








-0.1


time
~
c=~- ;
,-. -
-~. ~" -


C:

i~N


0.3


-0.05

N:
0\


0.05


0.20


N:


1 2


3lsz


Figure 3-3. Impedance spectra obtained from a symmetrical Ag-ESB/ESB/Ag-ESB (a), Ag-
YSB/YSB/Ag-YSB (b), and Ag/ESB/Ag (c) cells tested at 650 oC over a period of
100 h. The data are Rs-corrected. Arrows indicate direction of increasing time.


0.1 Z'(R) 0.2


0.10
z'(a)>


0.15









0.4 4
ASR = 8.65x10-2 t + 0.95 Pure Ag


0.3 3




c: 0.2 -2 c:



ASR 4.44x10-4 t 0.062 Ag-ESB
0.1 --1


AiSR = 1.4lx10-4 t+ 0.040 A-S
0 10
0 50 100
Time (h)

Figure 3-4. Electrode ASR vs. time for Ag-ESB, Ag-YSB, and pure Ag at 650 oC. Linear
regression based on first 10 h of testing for each electrode system.








1.05

ESB

14e

b3 0.95



0.9


50


100


Time (h)
Figure 3-5. Change in electrolyte relative conductivity vs. time for ESB and YSB at 650 oC.














-A- Ag (L)
--Bi (M)
-* Y(M)
~--O (K)


100
90
80
70


-10


Distance from interface (Cim)
Figure 3-6. EPMA cross-section linescan of a sample having pure silver electrodes, annealed at
750 oC for 48 h.













.0 (200 (222
Ag F



I I u nfi red



20 30 40 50 60
26 (">
Figure 3-7. XRD spectra of Ag-ESB powder mixtures before and after co-firing at 750 oC for 48
h. "F" markers identify cubic fluorite peaks of the ESB phase, "Ag" markers identify
silver peaks.












































Figure 3-8. SEM micrographs comparing the morphology of Ag-ESB electrodes before (a) and
after (b) firing at 750 oC for 1 h and (c) after testing at 650 oC for 100 h.











































Figure 3-9. SEM micrographs comparing the morphology of Ag-YSB electrodes before (a) and
after (b) firing at 750 oC for 1 h and (c) after testing at 650 oC for 100 h.





Figure 3-10. SEM micrographs comparing the morphology of pure Ag electrodes before (a) and
after (b) firing at 750 oC for 1 h and (c) after testing at 650 oC for 100 h.
Delamination caused by fracturing for SEM analysis.











CHAPTER 4
IMPROVING THE STABILITY OF SILVER-BISMUTH OXIDE CATHODES THROUGH
MICROSTRUCTURAL CONTROL

4.1 Introduction

Recent literature results demonstrate that a composite cathode consisting of a silver phase

as the electronic conductor and a stabilized bismuth oxide phase (in this case yttrium stabilized

bismuth oxide, YSB) as the ionic conductor exhibits outstanding performance, compared with

other frequently studied systems (Fig 3.1).39 In Chapter 3, the stability of this system was

assessed and determined to be a problem due to the high mobility of the silver phase.59 It was

incidentally shown that composite electrodes which contained an oxide phase in addition to a

silver phase exhibited better stability compared to pure silver electrodes. This effect has also

been demonstrated in the literature.' Furthermore, microstructural evolution and cathode ASR

degradation was less severe when the oxide phase was composed of finer particles.

In addition, it was shown that operation under an applied DC bias causes an electro-

migration type effect on the silver phase, resulting in a build-up of silver phase at the electrode-

electrolyte interface at one side of a symmetrical cell, and a migration of silver to the surface of

the other side. '~5 Stable SOFC components should deliver relatively constant, non-changing

performance for thousands of hours of operation. The goal of this study was to improve the

stability of this electrode system under open-circuit conditions as well as under an applied DC

bias through microstructural means.

4.2 Experimental

In this work, silver-ESB20 composite electrodes were studied because of the high ionic

conductivity of ESB20. Also, based on the results presented in Chapter 3, ESB20 shows no

phase transformation or defect ordering at the temperature of interest. Erbium oxide and bismuth









oxide powders were weighed in proper amounts to yield Er0.4Bil.603. The powders were ball

milled in ethanol for 24 h using YSZ grinding media, dried on a hot plate with stirring, then

calcined at 800 oC for 10 h. The calcined powder was then crushed by mortar and pestle and

sieved (325 mesh). Green ESB20 pellets were prepared by uniaxial pressing (approximately 70

MPa) followed by isostatic pressing (250 MPa). These green bodies were then fired at 890 oC

for 15 h. The sintered pellets had densities 94% 2% of theoretical, were 1.12 cm 10.02 cm in

diameter and 0.29 cm 10.01 cm in thickness.

For the electrode slurry, the crushed and sieved ESB20 powders were either used as-

prepared or a vibratory mill was used to reduce particle size. For vibratory milling, 10 g of

powder plus 300 g of cylindrical zirconia grinding media and 200 ml isopropyl alcohol were

placed into 250 ml Nalgene bottles. These bottles were covered with duct tape and placed into a

Sweco model M18-5 vibratory mill for seven days using appropriate counter weights. Inks were

prepared by combining organic vehicles with mixtures of metallic (silver) and ceramic (ESB20

alone or ESB20 and Tosoh 8YSZ) powders. Once an appropriate viscosity was reached

(consistency of honey), the inks were applied to both sides of the electrolyte substrates by screen

printing (AMI model HC-53 screen printer) to give symmetrically-electroded cells. The hole

size of the screen printing mask was large enough to prevent preferential screening of larger

particles. The electrodes had thicknesses of ~3 0 Cpm and geometric surface areas of ~0.79 cm2

These cells were dried at room temperature for 1 h, followed by drying at 120 oC for 1 h. This

process was repeated for a second coat. The doubly-coated cells were then fired at 750 oC for 1



Electrochemical performance of the electrodes was assessed using impedance

spectroscopy. Silver mesh current collectors and platinum lead wires were pressed against the









samples in a quartz reactor by the use of a ceramic screw-and-bolt assembly. The quartz reactor

was placed into a temperature-controlled furnace, and air was fed through the reactor at 50

cc/min using a BOC mass flow controller. A Solartron 1260 frequency response analyzer was

used in standalone mode for unbiased testing using a frequency range of 32 MHz to 0.01 Hz and

an AC voltage amplitude of 100 mV. A Solartron 1287 interface was added for testing under an

applied bias of 250 mV.

X-ray diffraction (XRD) patterns of various powders were recorded with a Philips APD

3720 diffractometer using Cu Kocl (h = 1.54056 A+). Step scans were taken over a range of 20

angles from 20o to 100o with 0.04o steps.

Number-average particle size distributions were determined using a Brookhaven ZetaPlus

instrument using a light scattering technique dilute suspensions of powder in an ethanol medium.

Scanning electron microscopy (SEM) was used to analyze electrode microstructures. Both

JEOL JSM 6400 scanning electron microscope and JEOL JSM 6335F field emission scanning

electron microscopes were used in this study for generating secondary and backscattered

microstructural images.

4.3. Results and Discussion

4.3.1 Nano YSZ Additions

The addition of bismuth oxide grains acts to inhibit silver phase grain growth in silver-

YSB composites.' Further, a common method used to avoid grain growth in commercial alloys

is to add a dispersion of fine particles, which restricts grain size according to


RL (4-1)
6f

where RL is the limiting grain radius, r is the particle radius, and f is the volume fraction of

particles--thus grain size is reduced by decreasing the particle size radius, or increasing the









volume fraction of particles (both of which increase the surface area of the particles).60 It is

proposed that increasing the surface area of the ceramic phase in silver-containing cermets will

further inhibit silver phase mobility. The first attempt to improve the stability of this electrode

system involved the use of nano-sized ceramic particles. Nano-sized 8 mol% YSZ powder was

used in this study as it is readily available, relatively inexpensive (~$100 per kilogram, or 10%

the cost of GDC nano-sized particles), and non-reactive toward silver and ESB20 at the

temperature range of interest (Fig. 4-2). Various YSZ/ESB compositions were tested, keeping

the silver volume fraction constant. That is, cathodes with composition Agx-(YSZ,,-ESB1~,)1-x

were prepared, where x = 0.5, and y = 0, 0.05, 0.10, and 0.15 represent fractions of the total

cathode solids volume.

Impedance spectra and electrode ASR (calculated from Eq. 3-1) versus time under no

applied bias are shown in Figures 4-2 and 4-3, respectively. Two arcs are present in the spectra

for all electrodes suggesting the impedance is governed by two competing processes, however it

is not known at this time which electrochemical processes correspond to each arc. The arcs are

more distinct for electrodes containing nano-sized YSZ, and it is clear that all samples containing

the YSZ powder additions were significantly more stable compared to the YSZ-free sample. The

electrode ASR degraded at a rate of 4.Ix1 0-4 GZCm2/h, 1.3x10-5 Ocm2/h, 2.5x10-5 Ocm2/h and

3.1x10-5 Ocm2/h for 0%, 5%, 10%, and 15% YSZ, respectively. The largest stability

improvement (a 97% reduction in ASR vs. time slope) was achieved at the lowest YSZ loading,

5%. In addition, the initial zero-time ASR value of this composition was 31% less than that of

the YSZ-free composition (0.043 OZcm2 and 0.062 OZcm2 for 5% and 0% YSZ, respectively),

despite the substitution of 5 vol% YSZ for the high conductivity ESB20 phase. Samples with

higher concentrations of YSZ exhibited higher initial ASR values, as expected--0.057 OZcm2 and










0.077 OZcm2 for 10% and 15%, respectively--but showed electrode stability improvement over

the non-YSZ composite. However, the ASR vs. time slopes increased as YSZ content increased.

A backscatter electron microscopy comparison between tested samples (Fig. 4-4) shows

how the silver particles have been restrained from coalescence, allowing porosity to remain open

and three phase boundaries to remain high, compared to the case with no YSZ additions.

Further, the size of the silver grains decreases with increased YSZ content. Figure 4-5 is a

secondary electron image comparing the 15% YSZ sample before and after testing. Grain size

and porosity appear to be comparable before and after testing. However, the edges of the silver

particles in the as-fired sample are smooth, while in the tested sample, the edges are rough, due

to envelopment of YSZ into the silver particles. It is possible that this effect further reduces the

3PB length between metallic, gaseous, and ESB20 phases, contributing to the observed increase

in ASR with time for these electrodes. This could also explain the observed increased ASR vs.

time slopes with increasing YSZ content.

4.3.2. ESB20 Particle Size Reduction

The second attempt to improve electrode stability involved reduction of initial particle size

of the ESB20 powder. This strategy is similar to the addition of nano-YSZ particles--more

energy is required for the silver phase to migrate over a given distance in the electrode.

However, the overall ionic conductivity of the cathode as well as the reactive 3PB lines should

not be compromised, unlike YSZ additions. This is also a more cost-effective approach than

strategies involving alloying silver with precious metals to reduce Ag mobility, as suggested by

Jaiswal et. al.5

Figure 4-6 shows the reduction in particle size that was obtained from the use of vibratory

milling. Number average particle size was reduced from ~1 Clm to ~300 nm. Note that the size









distributions may be skewed to lower particle sizes since larger particles settle out of suspension

quickly, and consequently may be omitted from the measurement count. All ESB powders were

ground with YSZ ball media--both ball milled and vibratory milled powders. In addition, the

hardness of ESB20 is much lower than that of YSZ. Further, the mass of the ball media used

before and after ball milling and vibratory milling was nearly unchanged (< 0.01 wt%). Hence

YSZ contamination from milling should be minimal and is expected to have little influence on

the results.

The long term, unbiased impedance study comparing composites prepared with (ESBVM)

and without (ESBs) vibratory milling of the ESB20 phase, shown in Figures 4-7 and 4-8, reveal

the vast improvement in stability the reduction in ESB20 particle size imparts on this system. As

in Chapter 3, linear trends in ASR with time are observed, and the degradation rate of silver-

ESB20 composites at 650oC was reduced from 4. 10x10-4 GZCm2/h to 1.91x10-5 Ocm2/h -a 95%

drop with the reduced ESB20 particle size.

As can be seen in Figure 4-8, the initial, zero time ASR value was also lowered by 25%,

from an already-low 0.062 OZcm2 for the Ag-ESBs composite to 0.048 OZcm2 for the Ag-ESBvM

composite--this is most likely due enhanced 3PB and may be attributed to suppression of silver

migration during electrode sintering, leaving both porosity and 3PB length high.

This experiment was repeated with an external 250 mV DC bias applied across the cells to

simulate operating conditions, and the results are shown in Figures 4-9 and 4-10. The

improvement in performance and stability is evident, though not as pronounced as the unbiased

case. After about 15 hours of testing, the ASR exhibits a linear increase with time. The ASR vs.

time slope under 250 mV bias at 650 oC is reduced by 50% (from 1.6x10-3 GZCm2/h to 8.0x10-4

OZcm2/h) when smaller ESB20 particles are used to prepare the composite electrodes.









Both macroscopic and microscopic changes in the electrodes prepared with larger ESB20

particles before and after bias testing are immediately recognized (Fig. 4-11). Optically, after

testing, the electrode surface of the working electrode was silver in color. The counter electrode

of the tested cell appeared red in color, and a ring of silver color could be seen along its edge.

The microstructural changes are also dramatic. A cross-sectional view of the counter electrode

shows silver dendrite-like structures at the electrolyte interface, and nearly pure ESB20 at the

electrode surface. The silver phase clearly migrates in one direction--towards the

electrode/electrolyte interface in the counter electrode and towards the surface of the working

electrode. Oxygen is reported to have high solubility and mobility in silver.51,52 It is possible

that the application of a bias across the cell leads to an electro-migration effect where the silver

phase is dragged along in the direction of oxygen flux, as observed by Jaiswal et. al.5

In contrast, the microstructure of the counter and working electrodes prepared with

vibratory milled ESB20 particles after bias testing are quite comparable and no segregation of

silver phase at the counter electrode/electrolyte interface is detected (Fig. 4-12), indicating silver

phase migration was significantly suppressed by the smaller particle size. So, not only did the

small ESB20 particles reduce initial ASR, and enhance microstructural stability with no bias, but

also improved microstrural stability under bias testing.

4.4. Conclusions and Future Work

Addition of 5 vol% 8YSZ nano powders significantly improved unbiased electrode

stability by 97 %, and reduced the initial, zero time ASR value by 31 %. Similar results were

obtained when YSZ-free electrodes were prepared from ESB20 powders composed of particles

hundreds of nanometers in size as opposed to electrodes prepared from ESB20 powders

composed of micron-sized particles--the zero time ASR value was reduced by 25 %, and ASR

vs. time slope during unbiased testing of the silver-ESB20 system at 650 oC was reduced by 95









%. Finally, ASR vs. time slopes during testing under a 250 mV external applied bias were

lowered by 50 % using the smaller ESB20 particles due to suppression silver phase electro-

migration. The stability of composite silver-ESB20 electrodes under an applied bias still needs

some improvement. Improvements are likely with further reduction in ESB20 particle size down

to several tens of nanometers. Also, as the operating temperature of SOFCs is reduced, the

migration of the silver phase will be suppressed even further. These electrodes perform well

even in the 500-550 oC range, but defect ordering in the bismuth oxide phase becomes an issue at

these temperatures. Currently research is being done to overcome this issue as well.61



















h
12
L
~I
v
cl
cn
r
a,
r








20


30 40 50
26 (o)


Figure 4-1. XRD spectra of Ag-ESB-YSZ powder mixtures before and after co-firing at 750 oC
for 10 h.


-0.1






o 0.1 Z'(D)~0.2


5%


Z'(D)


0.1 Z'n0.2 0.3U
Z'(C)O 0 .1 Z'(D)0.2 0.3D

.Nyquist plots for silver-ESB20 composite electrodes containing 0 vol% (a), 5 vol%
(b), 10 vol% (c), and 15 vol% (d) 8YSZ nanoparticles. Samples were tested at 650 oC
in air for 100 h under no applied bias. Note that the electrolyte resistance has been
subtracted from all Nyquist plots.


Figure 4-2


















E
u
c: 0.07
pt
r


0.09


0.05
ASR = 1.25x10-ot + 4.03x10-02 5%



0.03

0 50 100
Time (h)
Figure 4-3. Effect of time on ASR of silver-ESB20 composite electrodes containing various
volumetric amounts of 8 mol% YSZ nanoparticles. Measurements taken at 650 oC in
air under no applied bias.










































~r~rr IrJrlrlul ~ru lll C~urlo I~ u y~l ~rur l D
Figure 4-4. Backscatter electron microstructural images of tested [650 oC, 100 h in air under no
applied bias] silver-ESB20 electrodes containing 0 vol% (a), 5 vol% (b), 10 vol% (c),
and 15 vol% (d) 8YSZ nanoparticles.



























Figure 4-5. Secondary electron microstructural images of an untested (a) and tested (b) silver-
ESB20 electrodes containing 15 vol% 8YSZ nanoparticles.


100 0

90-

80

70

60 -6

50 VM sieved

S40-

30-

20
0 300 600 900 1200 1500

Particle Size (nm)



Figure 4-6. Results of particle size analysis (number average) for the sieved and vibratory-
milled ESB20 powders.










t~-~SI


0.3


Figure 4-7. Nyquist plots (a) and imaginary part of impedance plotted as a function of log-scale
frequency (b) for 50-50 vol% silver-ESB20 composite electrodes, where the ESB20
phase was prepared from sieved (larger curves) and vibratory milled (smaller curves)
powders. Samples were tested at 650 oC in air for 100 h under no applied bias. Note
that the Nyquist plots have been shifted so that the high frequency intercept with the
real axis crosses at 0 02.


-0."1





NO



-0.075




-0.050




-).025


0.1l Z'(D;) 0.2


01 1*n**** I J hillid 1111111]| II1)C I ibi 1 1 ilti
10-10oQ 10" 1@2 103 10

Frequency (Hz)








0.1



x Ag-ESB,





0.06-



0.04
~ 50 10
C: Time (h)

Fiue48 lcrd S s iefr5-0vosle-S2 opst lcrds hr
th S2 hs a rprd rmsee tinls ndvbaoymle crls
podr.Sm lswretse t60o i i o 0 ndrn ple is








-0.4


-0.3




~-0.1 gE





0 0.1


0.2 0.3


0.4


-0.100

Ag-E~SB,
~t-0.075
C ~ I ~ ; t

-0.050


-0.025
Ag-ESB,


10-2 10-1 100 101 102 103 104

Frequency (Hz)


Figure 4-9. Nyquist plots (a) and imaginary part of impedance plotted as a function of log-scale
frequency (b) for 50-50 vol% silver-ESB20 composite electrodes, where the ESB20
phase was prepared from sieved (larger curves) and vibratory milled (smaller curves)
powders. Samples were tested at 650 oC in air for 40 h under a 250 mV bias. Note
that the electrolyte resistance has been subtracted from all Nyquist plots.








0.14




0.12 e





d Ag-ESBs




Ag-ESBvm se4

0.06
0 20 40
Time (h)
Figure 4-10. Electrode ASR vs. time for 50-50 vol% silver-ESB20 composite electrodes, where
the ESB20 phase was prepared from sieved (triangles) and vibratory milled (circles)
powders. Samples were tested at 650 oC in air for 40 h under a 250 mV bias.
































Figure 4-11. Microstructural images of silver-ESB20 composite electrodes, where the ESB20
phase was prepared from sieved powders--surface before testing (a) and cross-
section of the counter electrode (b), working electrode surface (c), and counter
electrode surface (d) of a cell after 48 h of testing at 650 oC under a 250 mV applied
bias.


1 top m ,;


e 20pmll~


































Figure 4-12. Cross-sectional microstructural images of silver-ESB20 composite electrodes,
where the ESB20 phase was prepared from vibratory milled powders, after 48 h of
testing at 650 oC under a 250 mV applied bias--working electrode (a) with close-up
view of the working electrode/electrolyte interface (b) and counter electrode (c) with
close-up view of the counter electrode/electrolyte interface (d).


IAM
I ~li-











CHAPTER 5
HIGH PERFORMANCE COMPOSITE BI2RU207-BI1.6ERO.403 CATHODES FOR IT-SOFCS

5.1 Introduction

The critical role microstructure plays in the electrochemical performance of composite

cathode systems has been demonstrated. In Chapter 3 the isothermal instability of a composite

cathode consisting of Ag as the electronic conducting phase and ESB20 as the ionic conducting

phase was demonstrated.59 The study showed a 70% rise in ASR for Ag-ESB20 from 0.06 OZcm2

to 0.10 OZcm2, after 100 h at 650 oC. Microstructural evidence revealed a correlation between

agglomeration of the constituent phases (arising from migration of the silver phase) and the

increased resistance. Chapter 4 showed that this microstructural evolution could be inhibited

using nano-sized ceramic particles, resulting in a 95% reduction in the ASR degradation rate, as

well as a 25% reduction in the initial ASR.62 From these studies, it is clear that relative particle

sizes, agglomeration, and phase segregation can have dramatic impact on 3PB lengths and the

surface area of catalytic active sites, and thus electrode resistance. In addition, the electrode

thickness and the addition of current collection layers can influence performance by altering in-

plane electronic conduction, as well as mass transport.13,39 In this chapter, these microstructural

considerations will be used in order to improve the performance of a new class of SOFC

cathodes-ruthenate oxide ESB20 composites.

Ruthenium oxides are known to be catalytically active for oxygen reduction.63,64 By

selection of a sufficiently large A-site dopant cation in (A2Ru207), such as Pb or Bi, its band

structure is altered in such a way so as to render its behavior metallic, with conductivity

increasing as temperature decreases.65 These properties make metallically-conductive ruthenate

oxides good candidates for SOFC composite cathodes.









It was recently demonstrated that composite cathodes consisting of Bi2Ru207 (BRO7) as

the electronic conducting phase and ESB20 as the ionic conducting phase on GDC electrolytes

exhibit outstanding performance compared with other frequently studied systems.66 The low

resistance (3.47 OZcm2 and 0.08 OZcm2 at 500 oC and 700 oC, respectively) of this composite may

be attributable in part to the high ionic conductivity of the ESB20 phase in combination with the

catalytic activity of the ruthenate phase. Note that bismuth was chosen as the A-site dopant, as it

is the same as the host cation of the ESB20 phase, and should improve chemical compatibility.

Additionally, bismuth ruthenates have been reported to have better stability than lead ruthenates

above 600 oC.67,68

The goal of this study is to reduce the ASR of the BRO7-ESB20 electrode system on

ESB20 electrolytes by microstructural optimization.

5.2. Experimental

5.2.1. Electrolyte and Electrode Preparation

A conventional solid-state synthesis route was used to prepare ESB20 and BRO7 powders.

Er203 (99.99 %, Alfa Aesar) and Bi203 (99.999 %, Alfa Aesar) powders were weighed in proper

proportions to yield Er0.4Bil.603. The same raw Bi203 pOwder and RuO2 H20 (99.99 %, Alfa

Aesar, where the moles of hydration, X, varies with batch used, normal X=2.8) were weighed in

stoichiometric amounts to yield Bi2Ru207. The respective powders were ball milled in ethanol

for 24 h using YSZ grinding media, dried on a hot plate with stirring, then calcined at 800 oC for

15 h (for ESB20) or at 900 oC for 10 h (for BRO7). The calcined BRO7 powder was leached in

HNO3 to remove an impurity sellinite phase as described by Jaiswall.66 The ESB20 and BRO7

powders were separately crushed by mortar and pestle and sieved (325 mesh).

Green ESB20 pellets for use as electrolyte supports were prepared by uniaxial pressing

(approximately 70 MPa) followed by isostatic pressing (250 MPa). These green bodies were









then fired at 890 oC for 15 h. The sintered pellets had densities 94% 2% of theoretical, were

1.12 cm 10.02 cm in diameter and 0.29 cm 10.01 cm in thickness.

For the electrode slurry, the crushed and sieved ESB20 powders were either used as-

prepared or a vibratory mill was used to reduce particle size. For vibratory milling, 10 g of

powder plus 300 g of cylindrical zirconia grinding media and 200 ml isopropyl alcohol were

placed into 250 ml Nalgene bottles. These bottles were covered with duct tape and placed into a

Sweco model M18-5 vibratory mill for seven days using appropriate counter weights.

Sedimentation was performed to further reduce particle size and size distribution. Powders

were mixed in a medium (1 g / 50 ml ratio) in Nalgene bottles, ultra-sonicated for 30 min to

break up soft agglomerates, and allowed to settle for ~24 48 h. The supernatant was carefully

collected with pipettes and dried to 80 oC. Note that isopropanol was used as the sedimentation

medium in preference to de-ionized water to avoid the formation of Bi(OH)3 platelettes which

convert to co-Bi203 upon firing. The subscript designation and volume mean diameter, dv, for

each starting powder is given in Table 5-1.

BRO7 and ESB20 powders were then weighed in appropriate ratios. For studies involving

the use of sonication and sedimentation, BRO7-ESB20 powders were combined in isopropanol

and ultra-sonicated to achieve more intimate mixing. Inks were prepared by combining organic

vehicles with these mixtures of BRO7 and ESB20 powders or pure BRO7 for the current

collector. Once an appropriate viscosity was reached, the inks were applied to both sides of the

electrolyte substrates by paint brushing to give symmetrically-electroded cells. These cells were

dried at room temperature for 1 h, followed by drying at 120 oC for 1 h, and firing at 800 oC for 2

h. Note that in order to vary electrode thickness, or for the addition of the current collector,

successive layers were added after the drying stage, but before the firing stage. This was done to









reduce densification and grain growth within the cathode, as well as minimize ruthenium loss in

the volatile BRO7 phase, both of which that may occur when firing each coat separately. Two

was the standard number of coats applied to each cell.

For testing under SOFC operating conditions, dense YSZ pellets with reference electrode

bores were prepared. YSZ pellets with 99% theoretical density were prepared by slip casting

into a porous mold. The slip suspension was prepared using 30 vol% 8YSZ powder (Tosoh) in

de-ionized water and an appropriate amount of citric acid pre-dissolved (~0.3 wt% citric acid to

YSZ). Ammonium hydroxide was added dropwise to the mixture, with vigorous shaking, until a

water-like consistence was reached. The slip was allowed to dry in the mold for 1 day and

polished into flat disks of ~2 mm thickness. Holes were carefully drilled halfway into the green

pellets with a 1/16 inch bit. The pellets were then fired at 1400 oC for 4 h. NiO-YSZ (50-50

wt%) pasts were applied to the hole-free side of the YSZ pellets, dried, and fired at 1300 oC for 1

h. Pt wire was beaded on one end using a torch, dipped into Pt paste, carefully inserted into the

hole, and fired at 1 100 OC for 2 h, along with a Pt current collector on the anode side of each cell.

The cathodes (LSM-YSZ, LSCF-GDC, and BRO7-ESB) were then applied and fired at

appropriate temperatures and times. LSM20 powders were obtained from Nextech Corporation,

8YSZ from Tosoh, LSCF from Praxair, and GDC10 from Rhodia, Inc. BRO7 and ESB powders

were prepared in-house as described above. Pt current collectors were applied to LSM-YSZ and

LSCF-GDC cathodes. A pure BRO7 current collector was applied to the BRO7-ESB cathode.

Current collectors were co-fired with each cathode. In order to improve the mechanical strength

of the reference electrode, the hole was backfilled with YSZ slip and fired at 700 OC.

Anode-supported cells were prepared by co-pressing ~0.35 g GDC powder onto a NiO-

GDC substrate (~4 g) in a 1 1/8" cylindrical die. The NiO-GDC substrate was first pressed at










~14 MPa. Next, GDC was carefully and uniformly spread across the surface, and pressed at~-42

MPa. The pellets were then pressed isostatically at 250 MPa, and fired at 1450 oC for 4 h using

a 3 OC/min heating rate and a 400 oC, 1 h binder burnout step. The cells were then electroded in

the usual fashion.

5.2.2. Characterization

Electrochemical performance of the electrodes was assessed using impedance

spectroscopy. Silver mesh current collectors and platinum lead wires were pressed against the

samples in a quartz reactor by the use of a ceramic screw-and-bolt assembly. The quartz reactor

was placed into a temperature-controlled furnace, and air was fed through the reactor at a rate of

50 cc/min using a BOC mass flow controller. A Solartron 1260 frequency response analyzer was

used in standalone mode for unbiased testing using a frequency range of 32 MHz to 0.01 Hz and

an AC voltage amplitude of 100 mV.

X-ray diffraction (XRD) patterns of various powders were recorded with a Philips APD

3720 diffractometer using hKa(Cu) = 1.5406 A+. Step scans were taken over a range of 26 angles

from 200 to 1000 with 0.040 steps.

A JEOL JSM 6400 scanning electron microscope (SEM) was used in this study for

microstructural characterization. In addition, a JEOL TEM 200CX transmission electron

microscope (TEM) was also used in this study for particle size characterization. Average

particle sizes and electrode porosities were characterized statistically from SEM and TEM

micrographs using ImageJ software.

For each powder, the volume mean diameter, dv, was calculated from Eqn. (5-1)














V (5.1)




where di is the equivalent sphere diameter of each particle and n is the total number of particles

analyzed by the ImageJ software. Note that particle sizes are reported in terms of a volume-

average as opposed to a number-average as this choice is more consistent with other commonly-

reported microstructural properties such as porosity and composition.

Statistical stereology on epoxy-resin embedded, polished samples revealed that all

electrodes had ~40 vol% porosity.

For current-voltage measurements, cells were sealed (anode side) to an alumina tube using

ceramabond (Aremco). The setup was then placed into a furnace, cured, and taken up to testing

temperature. Air and H2 H20 gas mixtures were used as the oxidant and fuel gases, respectively.

Flow rates were maintained at 30 ccm using mass flow controllers. Cell OCP was monitored

using a Solartron 1287 potentiostat until a stable value was reached, and current-voltage

measurements were taken with the same instrument. A Solartron 1260 frequency response

analyzer was used for impedance measurements.

5.3. Results and Discussion

5.3.1. Chemical Compatibility

Chemical stability testing results between BRO7 and ESB20 are shown in Fig. 5-1. XRD

patterns of BRO7 and ESB20 powder mixtures before and after heat treatment at 800 oC for 10 h

are comparable, and reveal no evidence of inter-phase reactivity, suggesting these materials are

chemically compatible at this temperature.









5.3.2. Reproducibility/Compositional Optimization

Before any microstructural optimization was conducted, BRO7-ESB20 composites were

first compositionally optimized on ESB20 electrolytes using readily available powders--ESB20

prepared from conventional solid-state synthesis and solid-state BRO7 powders which

underwent an additional vibratory milling step (ESB20s-BRO7Vas). Optimization in composition

was constrained between 25 and 75 wt% ESB with steps of 12.5 wt% ESB (and 6.25 wt% ESB

at critical intermediate compositions). ESB Figure 5-2 shows typical impedance and Bode

spectra obtained at 625 oC in air under no applied bias. Note that as in previous chapters, these

and all future impedance plots have been Rs-corrected for ease in comparison. That is, the high-

frequency real-axis intercept, Rs, of each spectrum, which is composed of the bulk electrolyte

resistance and possibly electrode sheet resistance and lead contact resistance, has been subtracted

from the real component of each data point in the spectrum. At 625 oC, each impedance plot

appears to be composed of a single arc (Fig. 5-2a), and the characteristic frequency increases

with ESB20 content (Fig. 5-2b). It is interesting to note that the width of the imaginary

impedance vs. log frequency spectrum is narrower toward the extremes of the concentration

range than at intermediate compositions, suggesting a mechanistic overlap at these intermediate

compositions, and in fact, a minimum value is observed at 56 wt% ESB20 over the range of

temperatures tested.

Figure 5-3 shows the trend in cathode ASR (calculated from Eq. 3-1) vs. composition for

the ESB20s-BRO7Vas system over a range of temperatures. Since the densities of BRO7 and

ESB20 are approximately the same (~8.9 g/cm3), the observation of an ASR minima near 50

wt% of each phase is consistent with the effective medium percolation theory and 3PB

maximization. For a two-phase composite whose particle sizes are roughly the same, and

assuming the solid phases and pores are randomly distributed and the porosity is open and









sufficiently large, the 3PB length should reach a maximum if the two solid phases are present in

equal fractions of the overall electrode volume. In the ESB20s-BRO7VM study, BRO7 particles

are a factor of ~2 smaller than ESB20 particles hence percolation and 3PB length maximization

can occur at non-equal volume fractions of the two phases. As discussed above, 3PB length is

important in two-phase cathode systems since the oxygen reduction reaction that is occurring can

only proceed at sites where all three reactants are present--as the cathode 3PB length increases,

so do the number of reaction sites, and hence cathode activation polarization decreases.

Note that this result differs from earlier results,66 where a minimum was observed between

31-43 wt% ESB. One possible explanation may be differences in powder synthesis techniques

between the two studies. ESB20 was obtained through conventional solid state synthesis in this

study (micron-sized particles), as opposed to a wet chemical route (yielding nano-sized particles)

used by Jaiswall, et. al.66 Further, in the present study the BRO7 powders were vibratory milled

for seven days versus three days (Jaiswall, et. al);66 hence, yielding, on average, smaller BRO7

particles in the present study. Consequently, the ESB20-to-BRO7 particle size ratio in the

present study should be appreciably larger. As the ESB20-to-BRO7 particle size ratio decreases,

percolation of the ESB20 phase--and hence compositional optimization of the cermet cathode--

becomes feasible at smaller ESB20 volume fractions.

Note that there also appears to be a difference in activation energies between the electrodes

obtained in this study (1.02 eV) compared with Jaiswall, et al. (~1.3 eV).66 The electrode ASR at

500 oC are comparable (3.11 OZcm2 for the present study compared to 3.37 OZcm2), but are quite

different at 700 oC (0.17 OZcm2 for the present study compared to 0.08 OZcm2). It is not clear at

this juncture what mechanism leads to this lower activation energy. It has frequently been

reported that the surface of bismuth oxide-based electrolytes is active for the adsorption of










oxygen species. It is believed that bismuth strongly enhances the surface oxygen exchange

rate.69-71 This is evidenced by the factor of 103 larger surface oxygen exchange rate for bismuth

oxide based solids compared to YSZ solids. In addition, bismuth-based electrolytes have been

reported to be catalytically active towards oxygen dissociation and charge transfer.72,73 It is thus

possible that the use of ESB20 (rather than GDC) as the electrolyte support, which results in

additional ESB20 sites at the cathode/electrolyte, may result in the observed drop in activation

energy .

5.3.3. Optimization by Particle Size Ratio

Having shown an acceptable level of reproducibility in performance, microstructural

optimization was next carried out. Given the difference in optimum concentration between the

two compositional studies discussed above, careful consideration was given with respect to the

choice of concentration to use for the microstrucural study. It was decided to fix the BRO7-

ESB20 composition at 50-50 wt% (approximately 50-50 vol%) so as to minimize any possible

bias that might arise from skewing the composition toward either end of the spectrum.

Electrode microstructure was first optimized according to starting particle size ratio,

constrained to the particle sizes obtained from crushing and sieving as well as vibratory milling.

Particle sizes of the constituent phases were altered using vibratory milling as well as a

combination of sonication and sedimentation, as explained in the experimental section. Figures

5-4 and 5-5 show representative samples of each particle system used in the present study.

Figure 5-4 compares BRO7 and ESB20 powders as prepared from solid state synthesis, and

directly after vibratory milling. Particle size analysis (from imaging software) reveals that the

volume mean diameter was reduced by a factor of approximately two after seven days of

vibratory milling--from ~1.31 Clm to ~0.73 Clm for BRO7 and from ~1.31 Clm to ~0.81 Clm for










ESB20. Note however that each set of vibro-milled powders still contains a small number

fraction (but significant volume fraction) of large, unbroken particles. Figure 5-5 shows the

further reduction in particle size and agglomeration that can be obtained by sonication and

sedimentation.

Four different electrode structures were first examined using only "as-prepared" powders

(BRO7s-ESB20s, BRO7s-ESB20VM, BRO7VM-ESB20s, and BRO7VM-ESB20vu), each

containing a 1:1 volume ratio of the constituent phases. An SEM image of the four fully-fired

electrodes is shown in Figure 5-6.

Figure 5-7 shows typical impedance spectra and Bode plots obtained at 625 oC for these

electrodes in air under no applied bias. Plots of ASR vs. reciprocal temperature are given in

Figure 5-8. These results show that the composites exhibiting the lowest resistance are those

composed of both large and small particles. All composites containing at least one vibro-milled

phase have significantly lower resistance than the composite composed entirely of large

particles, as expected, due to the increased 3PBs. The activation energies for all electrodes are

similar (1.03eV + 0.03eV), as are their characteristic frequencies (Figure 5-7b), suggesting a

common rate-determining mechanism for the different electrodes. It is not known at this time

which mechanism is dominant; more work is needed to clarify this issue via testing under

different oxygen partial pressures. However, it is evident that the magnitude of the impedance of

this mechanism has been markedly affected by these microstructural considerations.

Comparing the two systems where the different phases have more similar grain size

(BRO7VM-ESB20vy and BRO7s-ESB20s), the electrode comprised of vibro-milled particles

exhibits the lower ASR, as expected. The systems composed of dissimilarly-sized particles

(BRO7VM-ESB20s and BRO7s-ESB20VM) exhibit nearly the same performance, with the










BRO7VM-ESB20s electrode having the lowest ASR of 0.43 OZcm2 at 625 oC. Interestingly, the

electrodes containing dissimilar grain sizes exhibit lower ASR values than the electrodes

containing more similar grain sizes, despite being shifted towards non-optimal composition

ratios, as discussed above. Recall that, although the optimal composition of BRO7vM-ESB20vy

was 56 wt% ESB20, for the microstructural study a 1:1 ratio was used. One possible explanation

for this observation could be the formation of soft agglomerates of fine particles during

synthesis, as can be seen for the as-prepared ESB20vM and BRO7vM powder in Figure 5-5.

Another possibility, as discussed in Section 2.5, is that smaller particles create a larger degree of

microstructural tortuosity, which can restrict gas phase diffusion.13

Note also that initially, samples used in this study were prepared with a single coat of

cathode paste. However a large drop in the ASR of the BRO7s-ESBVM was observed when a

sample of greater thickness was tested, hence a two-coat standard for thickness was adopted.

The effects of cathode thickness on performance is discussed in Section 5.3.5.

5.3.4 Sonication and Sedimentation

As mentioned in Section 5.3.3, as-prepared vibratory-milled powders contain a significant

volume fraction of unbroken, micron-sized particles. Focused ion beam (FIB) analysis and

reconstruction (a video file which can not be reproduced here) also reveals a degree of phase

segregation in the fired electrodes. The next stage in the microstructural development involved

the use of sedimentation for further reductions in particle size and a narrower particle size

distribution. Also, ultra-sonication was used to break up soft agglomerates and improve phase

distribution in the cathode.

The results are shown in Figures 5-9 and 5-10. Impedance results show a large reduction

in resistance for all systems where the raw vibro milled phase was replaced with the supernatant

phase. It is clear from Figure 5-9b that the impedance spectrum of the fully-supernatant










composite is composed of two arcs. Comparing the spectrum of this electrode with the other

electrodes, it appears that the low frequency process is the most affected by these microstructural

changes, and its resistance has been lowered to the point where the low and high frequency

processes are in competition with each other.

Figure 5-10 overlays the Arrhenius behavior of composites which underwent sonication

and sedimentation with those composed of as-prepared powders. Both composites consisting of

a mixture of large and small particles underwent comparable ASR reductions. However, the

ASR reduction was the most dramatic for the composite comprised entirely of small particles.

Clearly this composite benefits the most from the combination of reduced particle size and the

more intimate sonicated mixing. Not only were the soft agglomerates of each phase broken up,

but also the reduced particle size distribution translates into a larger number fraction of sub-

micron sized particles in each vibratory milled phase. This in turn translates into a larger 3PB

length. As discussed in Section 5.3.3 reduction in particle size gives rise to increased tortuosity

which can inhibit gas phase diffusion. However, as mentioned above, the spectra presented in

Figure 5-9 indicate that a reduction in particle size, size distribution, and phase agglomeration

provided by sonication and sedimentation primarily reduces the impedance of a single low-

frequency process. Since a higher-frequency process is only discernable once the impedance of

the low-frequency process is dramatically reduced, it is reasonable to conclude that the

improvements provided by increased 3PB lengths vastly outweigh any degradation resulting

from increased tortuosity. The minimum ASR observed was 0.10 OZcm2 at 625 oC for the

ESB20SUP-BRO7SUP System.









5.3.5. Effect of thickness and current collection

The effect of thickness on ASR for the different composite microstructures is illustrated in

Figure 5-11. In each case, ASR is reduced as thickness is increased, over the range of

thicknesses studied. The ESB20SUP-BRO7s system exhibited a one order of magnitude drop in

ASR between the first and second coating. This observation was confirmed to be reproducible

(from repeated testing on once-coated samples), and is believed to be due to a current collection

issue where the electrode thickness is insufficient for BRO7 phase percolation. That the

ESB20s-BRO7s system did not show as significant an ASR reduction as the ESB20SUP-BRO7s

system can be attributed to a reduction in BRO7 phase connectivity caused by the fine ESB20SUP

particles percolating between adjacent BRO7s grains. The volume percent porosity (open

symbols) of each composite tested is overlaid on Figure 5-1 1, and is relatively constant (40+7

vol%). There is no apparent trend between thickness and porosity over this narrow range, and

hence the porosity variation between electrodes tested is expected to have minimal influence on

the thickness vs. ASR results. It is also interesting to note that electrode ASR continues to drop,

even at thicknesses beyond 100-200 Clm. As stated in Section 2.4, in general, if the conductivity

of the ionically conducting phase in a composite electrode is sufficiently high, the effective

charge transfer resistance will decrease with thickness down to an asymptotic value. This indeed

appears to be the case in Figure 5-1 1, and the exceedingly high ionic conductivity of the ESB20

phase may explain why dominant concentration polarization effects do not show up at these

electrode thickness values.

In order to improve connectivity of the electronic conducting phase, a new batch of twice-

coated samples was prepared, this time using a pure BRO7s current collector layer. This layer

did not adhere well to the two electrode systems which utilized fine particles of ESB20.









However, for the other systems, there is a dramatic drop in the resistance of the low-frequency

process, as shown in Figure 5-12. For the BRO7SUP-ESB20s system, the high-frequency process

now seems to dominate the performance. As shown in Figure 5-13, the ASR of these systems

was reduced significantly--by factor of four (from 0.58 OZcm2 to 0. 15 OZcm2) in the ESB20s-

BRO7s system, and by a factor of three (from 0.22 OZcm2 to 0.076 OZcm2) in the ESB20s-

BRO7SUP System. The former system composed of large BRO7 grains exhibited a larger overall

ASR reduction than the system composed of fine BRO7 grains, as expected from enhanced

current collection. That is, percolation and phase connectivity is more easily achieved when the

phase is comprised of small particles and more difficult when the phase is comprised of large

particles. Thus, it is expected that the ESB20s-BRO7SUP System will have inherently better

electronic phase connectivity than the ESB20s-BRO7s system, hence introduction of a current

collecting layer should have less of an influence in the former system. The ESB20s-BRO7SUP

system exhibited the lowest ASR of all systems studied (0.73 OZcm2 and 0.03 OZcm2 at 500 oC and

700 oC, respectively). This is a marked improvement over earlier results (3.47 OZcm2 and 0.08

OZcm2 at 500 oC and 700 oC, respectively) and is comparable to landmark results reported by

Zhao and Haile46 for Bao.sSro.sCoo.sFe0.203-a (BSCF), especially at lower temperatures due to the

lower activation energy of BRO7-ESB20 compared with BSCF.

Further improvements are expected by improving the adhesion of the current collection

layer to the system composed entirely of fine particles since this exhibited the best performance

of all systems without a current collector. In addition, further compositional optimization may

be performed on each specific system studied in this work. As mentioned previously, bismuth

ruthenate is known to be volatile, and bismuth oxides are known to undergo an ordering










phenomenon at temperatures below ~600 oC. Thus, the stability of this system above 600 oC

should be examined.

5.3.6. Direct comparison with conventional cathode systems

Rather than simply comparing results with literature data, in-house LSM-YSZ and LSCF-

GDC composites were synthesized. The similarity in processing routines of the different

composite systems provides a more direct comparison with conventional cathode systems. The

four different materials (LSM from Nextech, YSZ from Tosoh Corporation, LSCF from Praxair,

and GDC from Rhodia) were obtained and have very Eine initial particle sizes. The conventional

composites were first optimized according to composition (using 5 vol% steps between 40-60

vol% of each phase using 5 vol% composition steps), firing temperature (using 50 oC steps

between 1000 oC and 1300 oC), and firing time (using 1 h intervals between 1-4 h).

All three sets of composites (LSM-YSZ, LSCF-GDC, and BRO7VM-ESBs) were applied to

dense YSZ pellets with Ni-YSZ anodes. Using the same electrolyte substrate for each system

minimizes the effect the electrode/electrolyte interface has on cathode performance. YSZ was

selected as the substrate since it is more easily sintered than GDC, and unlike ESB, can

withstand the firing temperatures used for LSM-YSZ and LSCF-GDC composites (1100 oC and

1150 oC, respectively). Further, testing was done under actual SOFC operating conditions using

a solid state adaptation of the Luggin probe configuration (Figure 5-14). The embedded

reference electrode in this configuration more accurately samples a well-defined equipotential

surface, allowing for more accurate measurement of electrolyte and electrode impedances,

compared with other surface-configured' reference electrodes which sample a more averaged

effective potential, leading to greater likelihood of inaccurate electrolyte measurements and

distorted electrode arcs74. Further, 3-point testing under SOFC operating conditions allows









direct assessment of cathode polarization as a function of current density, giving insight into

real-world behavior of these cathode systems.

Cathodic polarization can be calculated using Eq. 5-2

cathode = Em Evsz,ref cat (5-2)

where Em is the potential measured between the cathode and the embedded reference electrode,

and Evsz, ref cat is the potential drop due to the YSZ electrolyte between the reference electrode

and the cathode. This potential drop can easily be measured in-situ using impedance

spectroscopy between the cathode and reference electrode. The high-frequency intercept of the

impedance spectrum with the real axis (Z'hf, pt ||cat) allOws calculation of the ASR of the YSZ

electrolyte between the reference electrode and the cathode (Eq. 5-3)

ASRvsz, ref cat = Z'hf, pt ||cat A (5-3)

where A is the active area of the cell (the cathode area).

The voltage drop due to the YSZ electrolyte between the reference electrode and the cathode is

then simply the ASR value obtained from Eq. 5-3 multiplied by the cell current density. Since

the electrolyte resistance is ohmic, its resistance is independent of current density, a single, open-

circuit impedance measurement of Z'hf, pt ||cat can be used to determine the cathode overpotential

at all current densities (i) measured (Eq. 5-4).

rcathode(i) = Em(i) i Z'hf, pt ||cat A (5-4)

A similar 3-point impedance measurement between the reference electrode and anode provides

the ohmic resistance between the anode and the reference electrode Z'hf, pt ||an, aS well as the

polarization resistance of the anode.

As one accuracy check, the embedded reference electrode effectively partitions the ohmic

contribution of the electrolyte and allows separation electrode polarization. Thus, summing up










the partitioned ohmic contributions from the 3-point impedance measurement (Z'hf, pt ||an and Z'hf,

pt || ct) Should give the same value of ohmic resistance as a simple 2-point measurement between

the anode and cathode (Eq. 5-5).

Z'hf, 2-point= Z'hf, pt ||anand Z'hf, pt ||cat (5-5)

Further, the sum of the real and imaginary parts of the cathodic and anodic impedances measured

using the 3-point configuration should be the same as the real and imaginary parts of the total

electrode impedance measured using the 2-point configuration (Eqs. 5.6).

Z')tot, 2-point = Z'cat, 3-point + Z'an, 3-point (5-6a)

Z))tot, 2-point = Z))cat, 3-point + Z))an, 3-point (5-6b)

Such an analysis is shown in Figure 5-15 for the LSM-YSZ system measured at 650 oC.

There is a slight (~2 %) deviation between the electrolyte resistance measured from 2-point and

3-point measurements. The agreement between the total electrode resistance measured from

both configurations is better (deviation ~1 %). The deviations arise most likely from human

error involved in fabrication of the Luggin probe cell, but overall, the agreement is acceptable.

The high value of total cell ASR (44.5 OZcm2) arises mainly from the fact that testing is done on

thick (~2 mm) YSZ pellets. The cathode ASR measured (6.6 OZcm2) agrees reasonably well with

data for LSM-YSZ reported in the literature (Figure 3-2).

Current-voltage data for the same cell under the same conditions is shown in Figure 5-16.

The slope of the line near open-circuit conditions gives a measure of the total cell ASR, which

under these conditions is 45.5 OZcm2. This shows good agreement with the total ASR determined

from impedance measurement discussed above. Cathodic polarization at 650 oC as a function of

current density for the 3 systems studied is shown in Figure 5-17. As expected, the polarization

drop across the BRO7VM-ESBs composite cathode was significantly lower than that of









conventional materials. The ASR of BRO7VM-ESBs, calculated from the slope of the cathodic

overpotential-current density plot, is 0.9 OZcm2. This value is significantly higher than that

obtained earlier for this system at 650 oC (~0.5 OZcm2). The discrepancy is most likely due to the

fact that in the current study, testing was done on a YSZ electrolyte while previous testing was

done on ESB electrolytes. YSZ not only has orders of magnitude lower ionic conductivity than

ESB at this temperature, but also does not have the catalytic activity of ESB, as discussed in

Section 5.3.2.

5.3.7. Performance under operation

To gauge the viability of using BRO7-ESB composites, good performance should be

demonstrated. In order to demonstrate good performance, three samples were prepared using Ni-

GDC anode-supported electrolytes with relatively thin (~50 Clm) GDC electrolytes. To evaluate

performance of this cathode on ESB, one sample was coated with a thin film (~25 Clm) of ESB,

yielding a bilayer electrolyte with GDC on the fuel side and ESB on the air side. Note that the

GDC layer is necessary to protect the ESB layer, which decomposes under reducing conditions.

A LSCF-GDC composite cathode with a Pt current collector was used on one of the cells as a

reference standard. The optimized BRO7-ESB composite with a pure BRO7 current collector

was used on the other two cells (one with an ESB/GDC bilayer, the other having a GDC single

layer electrolyte).

The results of current-voltage testing at 650 oC are shown in Figure 5-18 as well as in

Table 5.2. Results on the single-layer GDC show that the BRO7-ESB composite has slightly

better performance than LSCF-GDC (total cell ASR is 0.81 and 0.85 for BRO7-ESB and LSCF-

GDC, respectively). Since the anode supports and the electrolytes were all prepared at the same

time, resistances due to the anode and electrolyte should be the same for both cells, so the slight










improvement can be attributed mainly to the cathode. Since the electrolyte is ~50 Clm thick,

much of the total cell resistance should be due to the electrolyte. Thus the improved cathode

performance is masked to some extent due to the resistance of the electrolyte (as well as the

anode).

The cell utilizing a bilayer electrolyte had significantly better performance compared

with the other cells (362 mW/cm2 maximum power density and 0.53 OZcm2 ASR), despite the

added resistance caused by inserting an ESB layer between the GDC and the cathode. This is

partly due to a degree of catalytic activity present at the surface of the ESB electrolyte, and the

role the cathode/electrolyte interface plays in the cathode resistance, as discussed in Sections

5.2.3 and 5.2.6. The full picture of the improved performance is not fully understood at this

point. Further, there may be some stability issues at this temperature as post-mortem SEM

analysis reveals formation of an interlayer between the cathode and the ESB layer (Figure 5-19).

More work is required to resolve these issues. However, it is clear that these cathodes have

potential for use in lower temperature SOFCs.

5.4. Conclusions

BRO7 was shown to be chemically compatible with ESB20. Microstructural engineering

by a combination of mechanical crushing, ultrasonication, and sedimentation was shown to be an

effective way of lowering electrode ASR, and the results seem to be consistent with 3PB length

maximization theory. Application of a pure BRO7 current collector to the electrode surfaces

further improved electrode performance. The lowest value of ASR attained ranged from 0.73

OZcm2 at 500 oC to 0.03 OZcm2 at 700 oC is one of the lowest SOFC electrode ASR values

reported to date. Direct comparison of BRO7vM-ESBs with conventional LSM-YSZ and LSCF-

GDC composite cathodes were made using a solid-state adaptation of the Luggin probe









configuration to extract cathode overpotentials as a function of current density. Results show

BRO7VM-ESBs to have a significantly lower cathodic overpotential than these conventional

composites. Current-voltage testing of the optimized composite was done on anode-supported

cells. A maximum power density of 362 mW/cm2 was attained at 650 oC using a ~75 mm thick

ESB/GDC bilayer electrolyte. It is believed that this material would be a good candidate cathode

for low temperature SOFCs.












Table 5-1. Volume mean diameter (dv), as well as subscript and symbolic designations for
different sets of starting powders. Hashed and white circles represent ESB particles,
black and gray circles represent BRO particles.
Phase subscript Type of powder dv ESB20 (Clm) dv BRO7 (Cpm) Symbol
designation
S Sieved 1.31 1.31 ~ D

VM Vibro-milled 0.81 0.73 gD g
SUP Vibro-milled, 0.08 0.06
09
supernatant


Table 5-2. Cell open circuit potentials (OCP), maximum power densities (MPD), and ASR for
selected SOFCs at 650 oC from current-density measurements.
Cell Type OCP (V) MPD ASR (OZcm2)
(mW/cm2)
LSCF-GDC on 0.91 250 0.85
GDC
BRO7-ESB on 0.90 269 0.81
GDC
BRO7-ESB on 0.87 362 0.53
ESB/GDC










7000

., 600 T 800 oC, 10 h

4l 000 -
P Pu

L5 2000 -1 unfiredO c


1000 C
200 30 40 506



fo 100 h









-0.5


0 0.5 1.0 1.5 2.0
Z'(0) A
-0.5
25%b 37% 625 OC

-0.4-



-0.3 -4~dL 6-9%

N 6~ 3%

-0.2 tp 56%~~~ ~



-0.1




10-110 0 10 1 102 10" 10 4
Frequency (Hz) B
Figure 5-2. Nyquist (a) and Bode (b) plots at 625 oC for different compositions of BRO7VM-
ESB20s electrodes tested in air.










10-
500 oC




cj 550 oC
E
C:


4 600 oC



625 oC




20 30 40 50 60 70 80
W~t% ESB
Figure 5-3. Effect of electrode composition for the ESB20s-BRO7VM system.









































Figure 5-4. SEM micrograph of as-prepared BRO7 powders before (a) and after (b) vibro-
milling as well as ESB20 powders before (c) and after (d) vibro-milling.








BR 07,


-rit


~
C


L l~nl


ESB20SUP,


C1 Irm
Figure 5-5. TEM micrograph of BRO7val powders before (a) and after sonication and
sedimentation (b), as well as ESBral powders before (c) and after (d) sonication and
sedimentation. Powders were dispersed onto lacey carbon TEM grids.


rBRO7SUP r ,t


rTL r )r
1


'fibL


b;
1 Clm ~
Al '"t


17










































~~Jr y bar~s~E~R~s"-~C~5~. ~ .,~ F~ere~lagr~a-~s~ I-
Figure 5-6. SEM image of four fully-fired BRO7-ESB20 cathode systems used in this study-
BRO7s-ESBs (a), BRO7VM-ESBs (b), BRO7s-ESBVM (c), BRO7VM-ESBvu (d).









625T



01 2 34


625T







-0.25




10 100 10' 10 1
Frequency (Hz)B
Figure 5-7. Nyquist (a) and Bode (b) plots at 625 oC for different 50-50 wt% BRO7-ESB20
electrode microstructures tested in air.
















'


''

r


'''
''





I

.'IC
''
''
''
''
''
''


.. .
~ .
~ .

.r
'' 0'
.


N

E
U

C:

P1
V)


1.1 1.15 1.2 1.25 1.3


1000/T (K 1)

Figure 5-8. Arrhenius plot of ASR vs reciprocal temperature for the four different
microstructures studied.









-0.5


625oC


1.5


0~
0 0.5
z'(nz)
-0.4


625oC


-0.3


N -0.2


-0.1


0'
10


Id

B


Figure 5-9. Nyquist (a) and Bode (b) plots at 625 oC for different 50-50 wt% BRO7-ESB20
electrode microstructures tested in air before (open symbols) and after (closed
symbols) sonication and sedimentation of electrode powders.


14 10' 102
Frequency(Hz)












,U
r'l
rll
r'l
r'
r'll


1.15 1.2 1.25


1000/T (K l)

Figure 5-10. Arrhenius plot of ASR vs reciprocal temperature--a comparison between
electrodes prepared from as-prepared powders (open symbols) and powders which
underwent ultrasonication and sedimentation (closed symbols). Arrows indicate ASR
drop for comparable systems.









100


625 oC


1


E










0.1


60 0



40 0


I ,


It---------------


100


150


200


250


300


Thickness (Cpm)
Figure 5-11. Effect of electrode thickness on ASR at 625 OC for the four different electrode
microstructures prepared after ultrasonication and sedimentation of the as-prepared
powders. Hollow symbols represent electrode porosity.










-0.15


625oC


-0.10


2 coats + BRO cc


-0.25


-0.05


0
10 1


100 101 10
Frequency(Hz)


0 0.25
Z'(n)


0.50


0.75


-0.5


-0.4


C:-0.3

N


1 coat
- 2 coats
- 4 coats


-0.67


N -0.17


0


0, ,1 ,,


0 0.5 1.0 1.5 2.0 10' 10 10 10 ur I
Z'(n)C Frequency(Hz)D

Figure 5-12. Nyquist (a,c) and Bode (b,d) plots at 625 OC for 50-50 wt%/ BRO7-ESB20 at
different thicknesses without (open symbols) and with (closed symbols) pure BRO7
current collectors. Note: The BRO7SUP-ESBs system is shown at left and BRO7s-
ESBs is shown at right.


~m~625oC





~Do


0.1 .: .- .
.4 cat +BR c


0...." 0.1i 2coat + BO c


S1.05 1.1 1.15 1.2 1.25 1.3 i 1.05 1.1 1.15 1.2 1.25 1.3
1000/T (K-1) A 1000/T (K-1) B
Figure 5-13. Arrhenius plot of ASR vs reciprocal temperature--a comparison between
electrodes without (open symbols) and with (closed symbols) current collectors for
the BRO7SUP-ESBs (a) and BRO7s-ESBs (b) systems.
























Figure 5-14. Solid state adaptation of 3-point Luggin probe configuration (a) schematic
representation and (b) actual cell.



-5
E anode (3 pt) cathode (3 pt) total (2 pt) total (an ode+cathode)


1 0 20 z'(acm2)30 40 50
Figure 5-15. Impedance spectra for LSM-YSZ composite comparing and total cell impedance
measured using 2-point configuration with that calculated from the sum of the anode
and cathode impedance measured using 3-point Luggin reference probe
configuration. Data has been normalized according to cathode area.










1.5


Cathode: LSM-YSZ
Electrolyte: YSZ
Anode: Ni-YSz
T=650 OC
Fuel: H2/H20
Oxidant: Air


1.0


L..
0.5


ASR = 45.5 Ocm2


0.1 (Am ps/cm2 002


0. 03


Figure 5-16. Current-voltage measurement for LSM-YSZ on Luggin probe cell at 650 oC using
hydrogen bubbled through water as the fuel gas and air as the oxidant gas.








0.2


0.~8 -1Anode:JN'i-YSZ' SYY
T=650 oC
0.1 6 -1 Fuel: H2/H20
Oxidant: Air
0.14

0 12 -1LSCF-GDC


4 0.08
S0.06

0.0 &BRO7-ESB ~en
0.02


0 0.01 0.02
Current (mA/cm2)
Figure 5-17. Cathode overpotential versus current density data for selected cathode materials on
Luggin probe cells at 650 oC using hydrogen bubbled through water as the fuel gas
and air as the oxidant gas.










1.00


BRO-ESB
(on ESB/GDC)


0.75










0.25





0


0.3





0.2





0.1


BRO-ESB


Standard
( LSCF-GDC
on GDC)


0.5 1.0


I (Amp s/cm2)

Figure 5-18. Current-voltage measurement for selected cells at 650 oC using hydrogen bubbled
through water as the fuel gas and air as the oxidant gas.





































Figure 5-19. SEM image of optimized BRO7-ESB composite cathode on SOFC with Ni-GDC
anode support with ESB/GDC bilayer SOFC after current-voltage testing.









CHAPTER 6
CONCLUSIONS

The solid oxide fuel cell is a promising candidate for future generation power generation

technologies. Since they do not rely on combustion of fuels, they are more efficient, quieter, and

cleaner than conventional technologies. However, in order to be more practical and cost-

effective, cell operating temperatures must be lowered to 500 OC and below. With the inception

of thin-film electrolytes and electrolytes with ionic conductivities higher than conventional YSZ

materials, much of the temperature reduction focus has shifted to electrode (anode and cathode)

development. The oxygen reduction reaction (Eq. 2-la and 2-1b), being more of a thermally-

activated process than that of fuel oxidation (Eq. 2-2 and 2-3), becomes a severely limiting

process at lower temperatures. Hence, low and intermediate-temperature SOFCs require

cathodes with a high degree of catalytic activity towards oxygen reduction as well as a

microstructure which maximizes the number of reaction sites (3PBs) and facilitates oxygen

transport toward and incorporation into the electrolyte. It is also important that the cathode be

microstructurally stable with time so that performance does not degrade over the lifetime of the

cell.

In the first part of this dissertation, the isothermal stability of a low-resistance cermet

cathode, silver-stabilized bismuth oxide was examined. Prior to stability testing, a preliminary

study of compositional optimization (between 40 and 70 vol% Ag phase at intervals of 5 vol%

Ag) was carried out on Ag-ESB20 composites in order to achieve a degree of agreement with

data published in the literature. The minimum in cathode ASR (achieved at 50 vol% ESB20)

was 0.18 OZcm2 at 600 OC, a significant improvement over that produced by Xia et al.39 who

obtained an ASR of 0.3 OZcm2 at 600 OC for their Ag-YSB composite. For the stability study,

pure Ag, Ag-ESB20, and Ag-YSB electrodes were isothermally tested at 650 oC for 100 h. All









systems experienced significant degradation in electrochemical performance during the test--

both the Ag-YSB and Ag-ESB electrode ASR values increased by around 70% (from 0.04 OZcm2

to 0.07 OZcm2 (75%) and from 0.06 OZcm2 to 0.10 OZcm2 (67%) for Ag-YSB and Ag-ESB

respectively), while the pure Ag system experienced a near fourfold increase (from 0.92 OZcm2 to

3.55 OZcm2) during the same length of time. Linear regression of the data (based on the first 10 h

of testing) yielded a degradation rate of 4.4x10-4 GZCm2/h for Ag-ESB and 1.2x10-4 GZCm2/h for

Ag-YSB. The pure Ag electrode degraded at a rate more than two orders of magnitude faster--

8.7x10-2 GZCm2/h. Electrolyte conductivities during the first 50 h of testing were relatively stable.

SEM analysis revealed significant microstructural evolution during the 100 h of testing at 650

oC. The pure Ag electrode appears fully dense. The porosity of the Ag-ESB20 electrode

appeared lower after testing--domains of coalesced silver are evident, and ESB20 particles

between these domains have been forced closer together, further reducing porosity and 3PBs.

Growth and coalescence of the silver phase was also evident in the Ag-YSB electrode though to

a lesser extent. It is believed that the smaller starting size of the YSB particles (as evidenced in

the unfired electrode micrographs) helped to lower the mobility of the silver phase by providing

the electrode with a higher surface area through which the silver phase must migrate. XRD and

EPMA analysis revealed neither evidence of inter-phase reactivity between silver and stabilized

bismuth oxide nor diffusion of silver into the electrolyte. In light of the electrochemical,

microstructural, and chemical evidence presented, it was concluded that electrode

microstructural evolution due to growth, agglomeration, and coalescence of the silver phase,

rather than chemical reactivity of the bismuth oxide phase, was responsible for the observed

degradation in electrochemical performance.









Next, attempts were made to reduce the microstructural evolution of the silver phase in

Ag-ESB20 composites, and hence improve electrochemical performance stability. This was

done by infusing the electrode with small particles (nano-size 8YSZ or vibratory-milled ESB20

particles) in order to increase electrode surface area which in turn would increase the amount of

energy required for the silver phase to migrate over a given distance in the electrode. The

addition of 5 vol% 8YSZ nano powders significantly improved unbiased electrode stability by 97

%, and reduced the initial, zero time ASR value by 31 %. Similar results were obtained when

YSZ-free electrodes were prepared from ESB20 powders composed of particles hundreds of

nanometers in size as opposed to electrodes prepared from ESB20 powders composed of micron-

sized particles--the zero time ASR value was reduced by 25 %, and ASR vs. time slope during

unbiased testing of the silver-ESB20 system at 650 oC was reduced by 95 %. The ASR vs. time

slopes during testing under a 250 mV external applied bias were lowered by 50 % using the

smaller ESB20 particles due to suppression silver phase electro-migration. The stability of

composite silver-ESB20 electrodes under an applied bias still needs some improvement.

Improvements are likely with further reduction in ESB20 particle size down to several tens of

nanometers. Also, as the operating temperature of SOFCs is reduced, the migration of the silver

phase will be suppressed even further.

Finally, porous composite electrodes consisting of BRO7 and ESB20 were synthesized and

characterized using impedance spectroscopy on symmetric cells. Electrode performance was

first manipulated compositionally by varying the weight percent of each phase in the composite,

and a minimum ASR of 0. 17 OZcm2 at 700 oC was achieved at 56 wt% ESB20. Next,

microstructural influences on electrode resistance were examined by varying starting particle

sizes of BRO7 and ESB20 powders using combinations of as-prepared sieved powders and









vibro-milled powders. Comparing the two systems where the different phases have more similar

grain size (BRO7VM-ESB20vy and BRO7s-ESB20s), the electrode comprised of vibro-milled

particles exhibits the lower ASR, as expected. The systems composed of dissimilarly-sized

particles (BRO7VM-ESB20s and BRO7s-ESB20VM) exhibit nearly the same performance, with

the BRO7VM-ESB20s electrode having the lowest ASR of 0.43 OZcm2 at 625 oC. Further ASR

reductions were achieved using a combination of sedimentation to further reduce particle size

and size distributions as well as ultrasonication to break up soft agglomerates. It is clear from

Figure 5-9b that the impedance spectrum of the fully-supernatant composite is composed of two

arcs. Comparing the spectrum of this electrode with the other electrodes, it is the low frequency

process is the most affected by these microstructural changes, and its resistance has been lowered

to the point where the low and high frequency processes are in competition with each other.

Since the impedance reduction at one frequency range was conspicuous while there appeared to

be no corresponding impedance rise at any other frequency range, it was concluded that the 3PB

improvements provided by reduced particle size, size distribution, and phase agglomeration

outweighs any possible degredation provided by increased electrode tortuosity. The minimum

ASR observed was 0.10 OZcm2 at 625 oC for the ESB20SUP-BRO7SUP System. The effect of

electrode thickness was also studied by applying successive coats of the electrode inks to the

electrolyte substrates. For all electrodes tested, electrode ASR dropped as thickness increased,

even at thicknesses approaching 200 Clm. The exceedingly high ionic conductivity of the ESB20

phase minimizes concentration polarization effects, explaining why these effects are not

observed even at such high values of electrode thickness. Lastly, application of a pure BRO7

current collector was found to dramatically improve performance. Using these optimization

techniques, a minimum electrode ASR of 0.73 OZcm2 and 0.03 OZcm2 was achieved at 500 oC and









700 oC, respectively. This is a marked improvement over earlier results (3.47 OZcm2 and 0.08

OZcm2 at 500 oC and 700 oC, respectively) and is comparable to results reported BSCF, especially

at lower temperatures due to the lower activation energy of BRO7-ESB20 (~1.0 eV) compared

with that of BSCF (~1.2 eV), making it one of the lowest resistance cathode materials reported to

date at such low temperatures. Further optimization for these composites is likely with improved

current collector adhesion, since the ESB20SUP-BRO7SUP System performed better than the

ESB20s-BRO7SUP System without a current collector. In addition, compositional optimization

on the optimized microstructure is incomplete at this point, since only 1:1 wt% ratios were

explored, thus supplemental compositional studies are likely to yield lower ASR values still.

Direct comparison of BRO7VM-ESBs with conventional LSM-YSZ and LSCF-GDC

composite cathodes were made using a solid-state adaptation of the Luggin probe configuration

to extract cathode overpotentials as a function of current density. Results show BRO7VM-ESBs

to have a significantly lower cathodic overpotential than these conventional composites.

Current-voltage testing of the optimized composite was done on anode-supported cells. A

maximum power density of 362 mW/cm2 was attained at 650 oC using a ~75 mm thick

ESB/GDC bilayer electrolyte.

Low-temperature performance and long-term stability testing under cell operating

conditions should also be conducted to finalize the feasibility of utilizing these materials as

lower-temperature SOFC cathodes. Unfortunately since bismuth oxide-based materials tend to

undergo defect ordering phenomena and/or phase transformations over much of the temperature

range of interest (400 oC to 600 OC), care must be taken when analyzing these composite

materials within this temperature range.









APPENDIX A
EXPERIMENTAL TECHNIQUES

This section does not contain detailed information about the fundamentals of each

technique. Instead it focuses on experimental design, testing procedure, typical results, and

derivation of important parameters for SOFC systems. Some key issues will be discussed.

A.1. Electrochemical Impedance Spectroscopy

Electrochemical impedance spectroscopy (EIS) is a valuable 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 Ohm's Law) through the sample is measured

over a range of frequencies. This response is usually represented as a "Nyquist" or "Cole-Cole"

plot where the real part of the impedance 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 terms of equivalent circuits, i.e., a

group 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 equivalent circuit representation (a

resistor and capacitor in parallel) are shown in Figure A-1. Such behavior could be

characteristic, for example, of a double-layer capacitance (due to charge separation between

electrode and electrolyte) in parallel with a resistance to charge transfer or a polarization

resistance. Notice that the magnitude of the impedance decreases as frequency increases. The

semi-circle is characteristic of a single "time-constant". Typical impedance plots of

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.

Often two time constants will overlap and the semi-circles must be deconvoluted in order to









determine each individual contribution. 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 circuit is similar to that in Figure A-1, but the capacitor is replaced by a so-called

constant-phase element. A capacitor can be thought of as a constant phase element whose phase

angle (the phase difference between voltage and current responses) is 900. When this phase

angle is somewhat less than this, a depressed semi-circle is observed. 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 450 angle

(Figure A-2). This feature is usually modeled by a so-called Warburg impedance and is

characteristic of semi-infinite diffusion. As shown in Figure A-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 the period of the

applied AC stimulus, hence the molecule does not 'see' that the cathode is of Einite thickness.

The response of a cell can be perfectly modeled by a number of different equivalent

circuits. Knowledge of the physical processes occurring in each cell can help identify the most

appropriate model. The model can be justified by altering a single aspect of the cell (grain size,

for example--Figure A-3)75 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 model are shown in Figure A-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 Ra2 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 A-3). Rct 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. As mentioned in Chapter 2,

the speed 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 equilibrium

potential for the reaction) should be small, and the Butler-Volmer equation becomes

RT
Rc, (A-1)
nFio

where R, T, n, and F have their usual meanings, and io is the exchange current density. Thus if

Rct is known, the exchange current density 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 Rct is not the only non-electrolyte resistance reported. Polarization

resistance, R,, and the more general electrode resistance, Rel are frequently reported. However,

in most all cases, the value for the resistance is measured as the difference between low and high

frequency real-axis 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,









RA
p = (A-2)
1

cr (A-3)
p RA

where p is the resistivity, o is the conductivity, A and I are the area and length over 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 Ra2 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 multiplied by the area of interest--for example,

the electrode resistance multiplied by the electrode area gives the electrode area-specific

resistance [GZcm2.

These properties will often show an Arrhenius relationship 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 imagine how this technique can be utilized to help interpret cell behavior, as well as

help determine rate-controlling processes.

A.2. Current-Voltage Measurements

There are a number of different ways to perform current-voltage (IV) measurements.

Figure A-5 shows a common fuel cell testing apparatus and typical results.

A seal is used to isolate and expose the two electrodes to different oxygen potentials (high

pO2" On the cathode and low pO2' On the anode). This sets up a Nernst potential across the fuel

cell, as mentioned previously. Current is drawn from the cell, and the resulting cell voltage is

recorded as a function of the magnitude of current drawn per unit of electrolyte area. Power









density is often plotted on the same graph. Two samples can be compared by changing one

element of the cell, such as the cathode, while keeping all other aspects of the cell constant. The

performance can be considered to improve if the slope of the current-voltage plot is decreased or

if the power density maximum is increased.

To isolate the response of the electrodes, current-interruption is often used. The ohmic and

non-ohmic contributions of the voltage (rle) between the cathode and Ref C can be separated by

the use of a fast electronic switch since the ohmic drop (relaxation of ionic and electronic charge

carriers) is order of magnitudes faster than the non-ohmic processes (discharge of the double

layer at the cathode/electrolyte interface and diffusion processes). A plot of rlo versus time and a

deduction of the relative (between cathode and reference electrode) non-ohmic cathodic

overpotential are shown in Figure A-6a. This process is repeated over a range of currents and a

plot of cathodic overpotential versus current density is obtained (Figure A-6b).

Another technique to accomplish the same result, i.e., determination of cathodic

overpotential is shown in Figure A-7. Here, the sample is not sealed, instead the process

is driven by an applied voltage, E, between the working and reference electrodes. The cathodic

overpotential is estimated by

Te = E IR, (A-4)

where I is the current and Ra2 is the ohmic resistance, as described earlier. Note that an

impedance spectrometer (FRA in Figure A-7) is used to determine the value of Ra at each value

of E applied.

The same data can be obtained using a combination of a fuel cell test (Figure A-5) and EIS.

A horizontal line drawn at the voltage intercept (current density equals zero) represents the

theoretical open circuit cell voltage, Eo. As the current density is increased, the cell voltage









drops by an amount equal to Eo-E(i). The ohmic contribution to this drop is simply iRa2, where

Ra2 is determined, as before, by impedance spectroscopy. The electrode contribution to the

voltage drop is then given by

r = E" E(i) iR, (A-5)

For clarity, a typical analysis is shown in Figure A-8. Murray and Barnett38 have also

separated electrolyte in order to obtain cathode overpotentials using a simple symmetrical cell (a

cell composed of the two electrodes of the same [cathode] material separated by the electrolyte).

In their measurements, a voltage was applied and plotted against the current measured. Again,

EIS was used to determine the electrolyte ohmic contributions, and these were subtracted from

the applied voltage to yield cathode overpotential versus current results (Figure A-9). The

results can be verified by comparing the cathode polarization resistance (the slope of the cathode

overpotential at zero current) to the magnitude of the polarization resistance determined by EIS.

The authors report good agreement.

It should be mentioned that some authors use these cathodic overpotential versus current

density plots (in the range of low overpotential) as an alternate way to extract exchange current

density values. Rearrangement of the low current approximation to the Butler-Volmer equation

(Eq. 2-14) gives


RT .nF
an=" (A-6)


where i/ rl is the low current slope of the current density-overpotential plot.

A.3. DC Electrical Conductivity

Figure A-10 shows the current and voltage Pt reference electrode arrangement for DC

conductivity measurements. The sample, in this case, is a relatively (~90%) dense pellet of the









electrode material composition being surveyed. A constant current is passed between points a

and b, and then b and c and the resulting voltage drop is measured across d and c, or a and d,

respectively. Two resistance values are then estimated by the equations


Rab,cd = dc (A-7a)



Rad,cb lad (A-7b)


Taking into account the flat, disc-shaped geometry, the conductivity of a sample of thickness d

can be estimated by the equation

21n(2)
cr=7 (A-8)
mi(R\1 ab,cd +ad,cb/

A.4. Oxygen Exchange Measurements

The resistance of mixed conducting cathodes is caused by several different processes such

as the exchange of oxygen from the gas to the solid, the diffusion of oxygen through the

electrode, gas phase diffusion through the pores, and the transfer of oxygen from the electrode

into the electrolyte. Oxygen exchange measurements are a valuable tool in the determination of

which process most greatly influences the cathode resistance. One such experimental technique

that is gaining in popularity is the isotope exchange/depth profiling method (IEDP). In this

technique, dense pellets of the electrode material composition being surveyed are placed in a

secondary ion mass spectrometer (SIMS). The samples are first annealed in a high purity oxygen

atmosphere at the measurement temperature. After the initial anneal, the chamber is switched (at

a time labeled t=0) to an IO-rich atmosphere, and the sample is again annealed. At some time, t,

depth profiling is performed on the sample by the use of a 5-15 keV primary ion beam (typically

Xe) at normal incidence to the sample surface, and the normalized isotopic ratio [ "O/(160+ O0)]









is measured via the SIMS instrument and plotted as a function of depth below the sample

surface.

The data is then fitted to the solution to Fick' s second law of diffusion. The appropriate

surface boundary condition used to solve this differential equation has been described22 as


k C -C =-D(A-9)


where k is the surface exchange coefficient, D is the oxygen self-diffusion coefficient, C C ,

and C: are the isotopic ratios in the gas, surface, and in the solid at depth x, respectively. From

a fit of the data to the solution42 to Fick' s second law, one obtains values for the parameters k and

D. The magnitude of these parameters provides valuable insight into the interpretation of

oxygen reduction kinetics.24

A.5. Thermogravimetric Analysis (TGA)

A TGA is used to measure the change in mass of a sample as a function of temperature.

These weight changes can result from a number of different processes, such as chemical

reactions and decomposition. The sample is prepared and placed on a microbalance and heated

in an appropriate atmosphere. TGA is sensitive enough to detect weight changes on the order of

a fraction of a microgram. This technique is useful for quantitative determination of oxygen

nonstoichiometry, and the experimental procedure has been described in detail.76 Figure A-11

shows typical results of oxygen nonstoichiometry calculations using information obtained from

TGA experiments.76

A.6. Microstructure

Microstructural evaluation is of critical importance in the study of cathode materials.

Many models used to describe electrode performance involve microstructrual parameters such as









porosity, grain size, tortuosity, etc. In addition, it is important in many cases for a material to be

single phase--both before and after electrochemical testing. For example, minor impurities

introduced by interdiffusion between cathode and electrolyte can lead to resistive interlayer

phases. These phases can lead to an alteration of diffusion and surface exchange properties of

the system."

A variety of techniques can be utilized to study microstructures. Scanning electron

microscopy (SEM) is a powerful technique that can be used to help determine grain size,

porosity, and tortuosity.77 In addition SEM can easily measure other parameters that may be of

interest such as cathode thickness. X-ray diffractometry (XRD) is another commonly employed

technique. It can be used by application of Bragg' s law to determine the theoretical density of a

material. In addition, phase purity (both before and after electrochemical testing) can be judged

based on the peak spectrum it produces. A quantitative determination of surface area is often

desired, and this can be done on both powder samples and sintered, dense samples." Thermal

expansion coefficients can be determined by utilizing a dilatometer.

It should be mentioned that there are many complexities involved in each technique; this is

particularly true for electrochemical measurements. The possible sources of experimental error

are numerous and will not be discussed here. Care should be taken in the preparation of each test

in order to minimize experimental error.





0 55000i 110000 1Bsn000 210000 275000o
Z', ohmn em
Figure A-3 Impedance response for as-sintered (1500 oC, 4 h) HS3Y samples (circles), annealed
at 1200 oC for 110 h in both 10% H2 balance N2 atmosphere (triangles), and air
(diamonds) showing the effect of increasing grain size (decreasing grain boundary
length). [Reprinted from Solid State lonics, Vol. 76, S.P.S. Badwal, Grain boundary
resistivity in zirconia-based materials: effect of sintering temperatures and impurities,
67-80 (1995) with permission from Elsevier.]


(n)=OOw=0 (DC)



R, RB+R, RP+RB+R ra


Figure A-1. (a) Equivalent circuit model and (b) typical impedance spectroscopy cell response.



-Zi,


Z
real


Figure A-2. Impedance response showing diffusion behavior at high frequencies.


165000


5~aoc



Baa~o~
9 +
a o
a
B d
P aaaa 'a
9 d
a S
'3~4 b



~
.cg9aoqg,


B 'a











-Z.,

I Mass transfer
Kinetic control
Control





R, R +V2Rt R +Rct real


A B
Figure A-4. (a) Response and (b) equivalent circuit for mixed kinetic and charge transfer control.



1.2 250

1 0 **2



~i0.6 Co,
oE 0. -as~ 5100 '

Higl pO 0.2 5

02IEdlectroyt drw cl 0
e 0 200 400 600 800 1000
|Anode
Low pO2 A current density (A/cn9) B
Figure A-5. (a) Fuel cell testing schematic and (b) illustrative representation of typical response.



~700 *C0.24


T chr a droop ."
S120t + 0.1



40~ -- 0.00
0 100 200cjo'~~ x 300 400 00 R 0.2 .4 06 0.8 1.0 1.2
t [ps] A: [A] B
Figure A-6. (a) Typical current-interruption response, and (b) calculated experimental cathodic
overpotentials. [Reprinted from Solid State lonics Vols. 86-88, M. Goidickemeier, K.
Sasaki, L. J. Gauckler and I. Riess, Perovskite cathodes for solid oxide fuel cells
based on ceria electrolytes, 691-701 (1996) with permission from Elsevier.]


























Figure A-7. Alternative electrochemical testing setup for cathodic overpotential


1.2







S0.4

~~0.2o -

0.



0 200 400 600 800
current density (A/cm2)

Figure A-8. Illustrative representation of separation of electrode contribution from fuel cell test
response.






































































Figure A-11. Oxygen nonstoichiometry data for various Co and Fe B-site dopant concentrations.
[Reprinted from the Ph. D. Dissertation of J.E. ten Elshof, Dense inorganic
membranes: studies on transport properties, defect chemistry, and catalytic behavior,
(1997) with permission from the author and The Universite of Twente.]


-~~ telecrd
I I I & I I I


2- *- ~







I I 1
-0.12 -0.09 408 4.05 0.00 0.03 0.08 0 0.12
Cunent (A)


1.0






-1.5


0.- 12 -0.09 -0 06 -0.03 0.00 0 03 0 06 0 09 0 (2
Cunent (A)


__


Figure A-9. Alternative determination of electrode polarization from symmetrical cell I-V
measurements. [Reprinted from Solid State lonics, Vol. 143, E. Perry Murray, S.A.
Barnett, (La,Sr)MnO3-(Ce, Gd)O2-x COmposite cathodes for solid oxide fuel cells, 265-
273 (2001) with permission from Elsevier.]


V- d


b


d()/


b~


Figure A-10. Electrode arrangement for DC conductivity measurements.


0,15

0.12


0.09


0.06



0.031

200


400 600 800
T ("C)


1000










LIST OF REFERENCES

1. J.J. MacKenzie, Oil as a Finite Resource: When is Global Production Likely to Peak,
World Resource Inst., Washington, D.C., (2000).

2. Advanced Technologies & Fuel Efficiency, http://www.fueleconomy .gov/feg/atv.shtml.

3. W.R. Grove, Philosophical Ma'gazine and Journal of Science 14, 127 (1839).

4. B. Weins, The Future of Fuel Cells, http://www.benwiens. com/energy4.html, Ben Wiens
Energy Science Inc., (2002) and references therein.

5. S.S. Penner, A.J. Appleby, B.S. Baker, J.L. Bates, L.B. Buss, W.J. Dollard, P.J. Fanis, E.A.
Gillis, J.A. Gunsher, A. Khandkar, M. Krumpelt, J.B. O' Sullivan, G. Runte, R.F. Savinell,
J.R. Selman, D.A. Shores, P. Turman, Energy 20 (5), 331 (1995).

6. Business Wire news article "Siemens Westinghouse's 100 kW SOFC System Completes
Operating Period", Jan 5, 2001, Gale Group, Farmington Hills, MI (2001).

7. Z. Wu, M. Liu, J. Am. Ceramn. Soc. 81 (5), 1215 (1998).

8. R. Doshi, V.L. Richards, J.D. Carter, X. Wang, M. Krumpelt, J. Electrochem. Soc. 146 (4),
1273 (1999).

9. T. Tsai, S.A. Barnett, Solid State lonics 98, 191 (1997).

10. A. Taranc6n, S. J. Skinner, R. J. Chater, F. Hernandez-Ramirez, J.A. Kilner, J. Mater.
Chem., 17, 3175 (2007).

11. R.C. Weast, CRC Handbook of Chemistry and Physics, 48th Edition, The Chemical Rubber
Co., Cleveland, Ohio, D-110 (1967-1968).

12. D.V. Ragone, Thermodynamics of Materials, Volume I, John Wiley & Sons, Inc., New
York, 129 (1995).

13. A.V. Virkar, J. Chen, C. W. Tanner, J. W. Kim, Solid State lonics 131 (1-2), 189 (2000).

14. B. C. H. Steele, Solid State lonics 12, 391 (1984).

15. N. Jiang, E.D. Wachsman, J. Am. Ceramn. Soc. 82 (11), 3057 (1999).

16. M.W. den Otter, Ph. D Dissertation, "A study of oxygen transportrt~r~rtrt~r~trtt in mixed conducting
oxides using isotopic exchange and conductivity relaxation," University of Enschede
(2000).

17. S. de Souza, S.J. Visco, L.C. De Jonghe, Solid State lonics 98, 57 (1997).

18. T. Ishihara, K. Sato, Y. Takita, J. Am. Ceramn. Soc. 79 [4], 913 (1996).










19. J.M. Ralph, J.T. Vaughey, M. Krumpelt, Solid Oxide Fuel Cells yTI : Electrochentical
Society Series 2001 (16), 466 (2001).

20. H. Tanabe, S. Fukushima, Electrochinsica Acta. 31, 801 (1986).

21. A.A. Yaremchenko, V.V. Kharton, E.N. Naumovich, A.A. Tonoyan, Mat. Res. Bull. 35
(4), 515 (2000).

22. B.C.H. Steele, Solid State lonics 75, 157 (1995).

23. J.R. Jurado, C. Mourie, P. Duran, N. Valverde, Solid State lonics 28-30, 518 (1988).

24. B.C.H. Steele, Solid State lonics 134, 3 (2000).

25. S. Wang, T. Kobayashi, M. Dokiya, T. Hashimoto, J. Electrochent. Soc., 147 (10), 3606
(2000).

26. C. Xia, W. Rauch, F. Chen, M. Liu, Solid State lonics, 149 (1-2), 11 (2002).

27. T. Horita, K. Yamaji, N. Sakai, Y. Xiong, T. Kato, H. Yokokawa, T. Kawada, Journal of
Power Sources 106, 224 (2002).

28. K. Sasaki, J. Tamura, H. Hosoda, T.N. Lan, K. Yasumoto, M. Dokiya, Solid State lonics,
148 (3-4), 551 (2002).

29. S. H. Chan, C. F. Low and O. L. Ding, Journal of Power Sources, 103 (2), 188 (2002).

30. T. Horita, K. Yamaji, N. Sakai, H. Yokokawa, A. Weber and E. Ivers-Tiffee,
Electrochinsica Acta, 46 (12), 1837 (2001).

31. T. Ishihara, T. Kudo, H. Matsuda, Y. Takita, J. Electrochent. Soc., 142 (5), 1519 (1995).

32. M. Mogensen and S. Skaarup, SolidState lonics, 86-88 (2), 1151 (1996).

33. M. Goidickemeier, K. Sasaki, L. J. Gauckler and I. Riess, Solid State lonics, 86-88 (2), 691
(1996).

34. S.H. Chan, K.A. Khor, Z.T. Xia, Journal of Power Sources 93, 130 (2001).

35. L. Rormark, Kjell Wiik, S. Stolen, T. Grande, J. Mater. Chent., 12, 1058 (2002).

36. J. Van Herle, A.J. McEvoy, K. Ravindranathan Thampi, Electrochinsica Acta 39 (11-12),
1675 (1994).

37. Z. Wu, M. Liu, Solid State lonics 93, 65 (1997).

38. E. Perry Murray, S.A. Barnett, Solid State lonics 143, 265 (2001).

39. C. Xia, Y. Zhang and M. Liu, Appl. Phys. Lett. 82 (6), 901 (2003), and references therein.










40. H. Ullmann, N. Trofimenko, F. Tietz, D. Stoiver, A. Ahmad-Khanlou, Solid State lonics
138, 79 (2000).

41. E. Maguire, B. Gharbage, F.M.B. Marques, J.A. Labrincha, Solid State lonics 127, 329
(2000).

42. J.D. Sirman, J.A. Kilner, J. Electrochent. Soc. 143 (10), L229 (1996).

43. R.T. Baker, I.S. Metcalfe, Appl. Catal. A 126, 297 (1995).

44. B.C.H. Steele, Journal of Power Sources 49, 1 (1994).

45. V.V. Kharton, E.N. Naumovich, V.V. Samokhval, Solid State lonics 99, 269 (1997).

46. Z. Shao, S. M. Haile, Nature, 431, 170 (2004).

47. B.C.H. Steele, K.M. Hori, S. Uchino, SolidState lonics 135, 445 (2000).

48. W. Preis, E. Bucher, W. Sitte, Journal of Power Sources 106, 116 (2002).

49. H. Uchida, S. Arisaka, M. Watanabe, J. Electrochent. Soc. 149 (1), Al3 (2002).

50. H. Uchida, S. Arisaka, M. Watanabe, Solid State lonics 135, 347 (2000).

51. M. Sahibzada, S.J. Benson, R.A. Rudkin, J.A. Kilner, Solid State lonics 113-115, 285
(1998).

52. H. Tu, U. Stimming, J. Power Sources 127, 284 (2004).

53. E. Wachsman, J. European Cerant. Soc. 24, 1281 (2004).

54. P. Shuk, H.D. Wiemhofer, U. Guth, W. Gopel, M. Greenblatt, Solid State lonics 89, 179
(1996).

55. K.Z. Fung, A.V. Virkar, D.L. Drobeck, J. Am. Cerant. Soc. 77 (6), 1638 (1994).

56. K.Z. Fung, A.V. Virkar, J. Am. Cerant. Soc. 74 (8), 1970 (1991).

57. A. Jaiswal, E.D. Wachsman, Solid State lonics 177, 677 (2006).

58. M. Camaratta, E.D. Wachsman, Electrochentical Chemical Society Transactions 1 (7), 279
(2006).

59. M. Camaratta, E.D. Wachsman, SolidState lonics, 178 (19-20), 1242 (2007).

60. P. R. Rios and G. S Fonseca, Scripta2ateralia 50, 71 (2004).

61. D.W. Jung, K. Duncan, E.D. Wachsman, Electrochentical Chemical Society Transactions
1 (7), 63 (2006).










62. M. Camaratta, E.D. Wachsman, SolidState lonics, 178 (23-24), 1411 (2007).

63. M. Hrovat, J. Holc, D. Kolar, Solid State lonics, 68 (1-2), 99 (1994).

64. C. Abate, K. Duncan, V. Esposito, E. Traversa, E. D. Wachsman, Electrochem. Soc. Trans.
1 (7), 255 (2006).

65. K. S. Lee, D. K. Seo, M. H. Whangbo, J. Solid State Chem. 131 (2), 405 (1997).

66. A. Jaiswal, C.T. Hu, E.D. Wachsman, J. Electrochem. Soc. 154 (10), B1088 (2007).

67. S. Gallini, M. Hansel, T. Norby, M. T. Colomer, J. R. Jurado, Solid State lonics 162-163,
167 (2003).

68. Z. Zhong, Electrochemical and Solid-State Letters, 9 (4), A215 (2006).

69. B.A. Boukamp, K.J. de Vries, A.J. Burggraaf, Non-Stoichiometric Compounds Surfaces,
Grain Boundaries~~~~dddd~~~~ddd and Structural Defects, 299 (1989).

70. J.C. Boivin, C. Pirovano, G. Nowogrocki, G. Mairesse, Ph. Labrune, G. Lagrange, Solid
State lonics 113-115, 639 (1998).

71. B.A. Boukamp, Solid State lonics 136-137, 75 (2000).

72. M. Dumelie, G. Nowogrocki, J.C. Boivin, Solid State lonics 28-30, 524 (1988).

73. I.C. Vinke, K. Seshan, B.A. Boukamp, K.J. de Vries, A.J. Burggraaf, Solid State lonics 34,
235 (1989).

74. G. Hsieh, T.O. Mason, E.J. Garboczi, L.R. Pederson, Solid State lonics 96, 152 (1997).

75. S.P.S. Badwal, Solid State lonics 76, 67-80 (1995).

76. J.E. ten Elshof, Ph. D Dissertation, "Dense inorganic membranes: Studies on transport
properties, defect chemistry, and catalytic behavior," University of Enschede (1997).

77. S.B. Adler, Solid State lonics 111, 125 (1998).









BIOGRAPHICAL SKETCH

Matthew Camaratta was born in Lakewood, New Jersey on October 14th, 1975. He grew

up in Madison, Connecticut, where he gained an appreciation for outdoor activities including ice

skating, hiking, and climbing. In 1987 he moved to Huntsville, Alabama with his family, and

later to Pensacola, Florida, where he attended the International Baccalaureate program at

Pensacola High School. There he made several strong friendships including Chess Club friends

Eric Ray and Fletcher Thomas, as well as future best man, Robert Van Hoose. In 1994, Matthew

began attending the University of Florida. He was followed one year later by his friend Robert.

The two quickly formed a band and dubbed themselves 'Baker Act' after a particular Florida

law, and together with a talented horn section played over 50 live performances throughout

Florida. In 1997 Matthew began dating his future wife, Tonya Bervaldi, and in 1998 they

traveled the countryside. During this time, Matthew' s awareness of and sensitivity to his actions

and the ramifications they have on his surroundings blossomed and led him to j oin Dr. Eric

Wachsman' s solid oxide fuel cell research group. The wealth of knowledge, experience, and

friendships he formed during these years was immeasurable. On August 24th, 2002, Matthew

and Tonya were married. Five years later their bond continues to strengthen due to their many

shared ideologies and love for the outdoors. In all matters, both strive to keep a sense of urgency

and curiosity. He will receive his Ph. D. in materials science and engineering in December of

2007.





PAGE 1

1 MICROSTRUCTURAL ENGINEERING OF COMPOSITE CATHODE SYSTEMS FOR INTERMEDIATE AND LOW-TEMPERAT URE SOLID OXIDE FUEL CELLS By MATTHEW CAMARATTA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Matthew Camaratta

PAGE 3

3 To my mother Maria, father Frank, and inspiration Tonya

PAGE 4

4 ACKNOWLEDGMENTS A debt of gratitude is ow ed to my advisor Dr. Eric Wachsman for his support, guidance, and oversight throughout this experience. Especi ally helpful were the m eetings he led nearly each and every week, providing myself and others a great degree of clarity and direction. I would also like to thank Dr. Juan Nino, Dr. Wolfgang Sigmund, Dr. David Norton, and Dr. Mark Orazem for their advice and participation as part of my committee. Funding was provided by the US Department of Energy (contract DE-F C26-03NT41959) and US DOE, High Temperature Electrochemistry Center (c ontract DE-AC05-76RL01830). I also wish to acknowledge Dr. Keith Duncan for his valuable disc ussions on the various theoretical aspects of the field, and transfer of experience with writing and presenting. Thanks to Dr. He-Sung Yoon for his indispensable processing assistance, as well as Dr. Briggs White for his collaboration with some experimental aspects of this work. Many thanks are owed to Wayne Acree and Kerry Seibein of the Major Analytical Research Center at the University of Flor ida for their patience and assistance with both microstructural analysis and with instrument training sessions. I also wish to acknowledge my lab mates Ji n Soo Ahn, Dr. Abhishek Jaiswal, Doh Won Jung, Su-Ho Jung, Dr. Sun-Ju Song, Dr. Jun Young Pa rk, Sean Bishop, and everyone else for the innumerous discussions and de bates throughout the years. I want to thank you for your friendship and for making this experience truly enjoyable. I thank my mother for her strength, I know this has been a difficult peri od in her life. I thank Baba for always being there. I thank my father for his wisdom. I thank Chris and Rob for the computer help and sense of humor. Finally, I thank you Tonya, for your constant encouragement throughout the ups and downs and helping me to persevere.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES.........................................................................................................................8 ABSTRACT...................................................................................................................................13 CHAP TER 1 INTRODUCTION..................................................................................................................15 2 LITERATURE REVIEW.......................................................................................................20 2.1. Basic Principles of SOFCs...............................................................................................20 2.2. Dominant Component Losses.......................................................................................... 24 2.2.1. The Electrolyte...................................................................................................... 24 2.2.2. The Electrodes.......................................................................................................27 2.3. Classes of Cathode Materials........................................................................................... 28 2.3.1. Single-Phase Electronic Conductor.......................................................................29 2.3.2. Dual-Phase Composite Cathodes..........................................................................30 2.3.3. Single-Phase Mixed Ionic and Electronic C onductors (MIECs)........................... 31 2.4. Cathode Materials Conven tional and State-of-the-Art ................................................. 33 2.5 Strategies for Improvi ng Cathodic Perform ance.............................................................. 40 2.6 Summary...........................................................................................................................44 3 STABILITY OF SILVER-BIS MUTH OXIDE CATHODES ............................................... 60 3.1. Introduction.............................................................................................................. ........60 3.2. Experimental....................................................................................................................61 3.3 Results and Discussion..................................................................................................... 63 3.3.1. Compositional Optimization................................................................................. 63 3.3.2. Stability: Thermal Performance Decay due to Microstructural Evolution............ 64 3.4 Conclusions.......................................................................................................................69 4 IMPROVING THE STABILITY OF SI LVER-BI SMUTH OXIDE CATHODES THROUGH MICROSTRUCT URAL CONTROL................................................................. 81 4.1 Introduction............................................................................................................... ........81 4.2 Experimental............................................................................................................... ......81 4.3. Results and Discussion.................................................................................................... 83 4.3.1 Nano YSZ Additions..............................................................................................83 4.3.2. ESB20 Particle Size Reduction............................................................................. 85 4.4. Conclusions and Future Work......................................................................................... 87

PAGE 6

6 5 HIGH PERFORMANCE COMPOSITE BI2RU2O7-BI1.6ER0.4O3 CATHODES FOR ITSOFCS.......................................................................................................................... ..........99 5.1 Introduction............................................................................................................... ........99 5.2. Experimental..................................................................................................................100 5.2.1. Electrolyte and Electrode Preparation................................................................. 100 5.2.2. Characterization................................................................................................... 103 5.3. Results and Discussion.................................................................................................. 104 5.3.1. Chemical Compatibility....................................................................................... 104 5.3.2. Reproducibility/Compos itional Optim ization..................................................... 105 5.3.3. Optimization by Particle Size Ratio.................................................................... 107 5.3.4 Sonication and Sedimentation.............................................................................. 109 5.3.5. Effect of thickness and current collection ........................................................... 111 5.3.6. Direct comparison with conventional cathode systems....................................... 113 5.3.7. Performance under operation.............................................................................. 116 5.4. Conclusions....................................................................................................................117 6 CONCLUSIONS.................................................................................................................. 138 APPENDIX EXPERIMENTAL TECHNIQUES........................................................................ 143 A.1. Electrochemical Impedance Spectroscopy.................................................................. 143 A.2. Current-Voltage Measurements................................................................................... 146 A.3. DC Electrical Conductivity..........................................................................................148 A.4. Oxygen Exchange Measurements................................................................................ 149 A.5. Thermogravimetric Analysis (TGA)........................................................................... 150 A.6. Microstructure........................................................................................................... ...150 LIST OF REFERENCES.............................................................................................................156 BIOGRAPHICAL SKETCH.......................................................................................................160

PAGE 7

7 LIST OF TABLES Table page 5.1 Volume mean diameter (dV), as well as subscript and symbolic designations for different sets of starting powders..................................................................................... 119 5.2 Cell open circuit potentials (OCP), m aximum power densities (MPD), and ASR for selected SOFCs at 650 C from current-density measurements......................................119

PAGE 8

8 LIST OF FIGURES Figure page 1-1 Energy production efficiency comparisons....................................................................... 19 2-1 Schematic representation of a typical SOFC button cell................................................... 45 2-2 Theoretical values of oxygen partial pressure on the anode (calculated from Eq. 2-8) and cell potential (calculated from Eq. 2-4) as a function of temperature......................... 45 2-3 Schematic representation of an SOFC stack...................................................................... 46 2-4 Illustrative representation of typical current-voltage profile for an SOFC illustrating different types of electrochem ical losses that arise when current is drawn....................... 46 2-5 Cubic fluorite (AO2) crystal structure with A-site cations occupying the cube corners and face centers, and oxygen anions occ upying the tetrahedral interstitials..................... 47 2-6 Reported ionic conductivity values for comm on SOFC electrolyte materials. ............... 47 2-7 Calculated cell performance and polarization resistances. ................................................ 48 2-8 Schematic representation of three phase boundary lines betw een electronic conducting particles, the i on conducting electrolyte..........................................................48 2-9 Schematic representation of three phase boundary lines betw een electronic conducting particles, ion conducting electrolyte............................................................... 49 2-10 ABO3 cubic perovskite crystal structure. (a ) A-site cations at cube center, B-site cations at corners, and oxygen anions at face centers. (b) Red spheres............................ 49 2-11 Schematic representation of various reaction pathways for the oxygen reduction reaction in a MIEC. ............................................................................................................ 50 2-12 Effect of cathode thickness on effective charge transfer resistance .................................. 50 2-13 Effect of cathode thickness in I-V performance................................................................51 2-14 Agreement between model (Eq. 2-15) and experim ent for LSM-YSZ composite cathodes..............................................................................................................................51 2-15 Effect of cathode thickness on polari zation resistance for LSM-YSZ composite cathodes..............................................................................................................................52 2-16 Effect of electrolyte phase compos ition on polarization re sistance of various conventional SOFC cathode m aterials............................................................................... 52

PAGE 9

9 2-17 Effect composition on (a ) cathod e overpotential, and (b) interfacial resistance for Ag-YSB composite cathodes.............................................................................................53 2-18 Theoretical ambipolar conductivity from EMPT............................................................... 54 2-19 Effect of composition on interfacial resistance of GDC-LSM com posite cathodes.......... 54 2-20 (a) Arrhenius conductivity plot of selected Fe loadings in LS CF, and (b) cathode polarization performance comparison................................................................................ 55 2-21 IEDP for (a) LSM and (b) LCCF....................................................................................... 56 2-22 Effect of A-site dopant on (a) electr ical conductivity, (b) ca thodic polarization, and (c) fuel cell power density for various Ln0.6Sr0.4MnO3......................................................57 2-23 Overpotential-current density curves for (a) diffusion, (b) charge transfer, and (c) m ixed control.................................................................................................................. ...58 2-24 Effect of electron-conducti ng particle size on 3P B length................................................ 58 2-25 Arrhenius behavior of interface conductiv ity for LSC cathodes (a) v arying Sr content at E=0V, and varying voltage............................................................................................59 3-1 (a) Impedance spectra and (b) electrode ASR vs silver content of Ag-ESB showing com positional optimization at 600 C............................................................................... 71 3-2 Arrhenius plot, adapted from Xia et. al.,39 comparing the ASR of Ag-ESB (this work) with the best cathodes reported in th e literature at the st art of this study............... 72 3-3 Impedance spectra obtained from a sy mm etrical Ag-ESB/ESB/Ag-ESB (a), AgYSB/YSB/Ag-YSB (b), and Ag/ESB/Ag (c) cells tested at 650 C for 100 h..................73 3-4 Electrode ASR vs. time for Ag-ESB, Ag-YSB, and pure Ag at 650 C. L inear regression based on first 10 h of te sting for each electrode system................................... 74 3-5 Change in electrolyte relative conduc tivity vs. time for ESB a nd YSB at 650 C............ 75 3-6 EPMA cross-section linescan of a sample having pure silver elec trodes, annealed at 750 C for 48 h. ..................................................................................................................76 3-7 XRD spectra of Ag-ESB powder mixtur es before and after co-firing at 750 C for 48 h..........................................................................................................................................77 3-8 SEM micrographs comparing the morphol ogy of Ag-ESB el ectrodes before (a) and after (b) firing at 750 C for 1 h and (c) after testing at 650 C for 100 h.........................78 3-9 SEM micrographs comparing the morphol ogy of Ag-YSB electrodes before (a) and after (b) firing at 750 C for 1 h and (c) after testing at 650 C for 100 h.........................79

PAGE 10

10 3-10 SEM micrographs comparing the morphol ogy of pure Ag electrodes before (a) and after (b) firing at 750 C for 1 h and (c) after testing at 650 C for 100 h.........................80 4-1 XRD spectra of Ag-ESB-YSZ powder mixt ures before and af ter co-firing at 750 C for 10 h. ..............................................................................................................................89 4-2 Nyquist plots for silver-ESB20 composite electrodes containi ng 0 vol% (a), 5 vol% (b), 10 vol% (c), and 15 vol % (d) 8YSZ nanoparticles. .................................................... 89 4-3 Effect of time on ASR of silver-ESB 20 com posite electrodes containing various volumetric amounts of 8 mol% YSZ nanopart icles. Measurements taken at 650 C....... 90 4-4 Backscatter electron microstr uctural im ages of tested [650 C, 100 h in air under no applied bias] silver-ESB20 electrode s containing 8YSZ nanoparticles............................91 4-5 Secondary electron microstructural images of an untested (a) a nd tested (b) silverESB20 electrodes containing 15 vol% 8YSZ nanoparticles. .............................................92 4-6 Results of particle size analysis (number averag e) for the sieved and vibratory-milled ESB20 powders..................................................................................................................92 4-7 Nyquist plots (a) and imagin ary part of i mpedance plotte d as a function of log-scale frequency (b) for 50-50 vol% silv er-ESB20 composite electrodes................................... 93 4-8 Electrode ASR vs. time for 50-50 vol% si lver-ES B20 composite electrodes, where the ESB20 phase was prepared from sieved (triangles)..................................................... 94 4-9 Nyquist plots (a) and imagin ary part of i mpedance plotte d as a function of log-scale frequency (b) for 50-50 vol% silv er-ESB20 composite electrodes................................... 95 4-10 Electrode ASR vs. time for 50-50 vol% si lver-ES B20 composite electrodes, where the ESB20 phase was prepared from sieved (triangles) and vibratory milled................... 96 4-11 Microstructural images of silver-ESB 20 com posite electrodes, where the ESB20 phase was prepared from sieved powderssurface before testing................................... 97 4-12 Cross-sectional microstructural images of silver-ESB20 compos ite electrodes, where the ESB20 phase was prepared from vibratory milled powders........................................ 98 5-1 XRD spectra for mixtures of BRO7-ESB20 before and after heat treatm ent at 800 C for 10 h.............................................................................................................................120 5-2 Nyquist (a) and Bode (b) plots at 62 5 C for different com positions of BRO7VMESB20S electrodes tested in air........................................................................................ 121 5-3 Effect of electrode composition for the ESB20S-BRO7VM system.................................. 122 5-4 SEM micrograph of as-prepared BRO7 pow ders before (a) an d after (b) vibromilling as well as ESB20 powders before (c) and after (d) vibro-milling....................... 123

PAGE 11

11 5-5 TEM micrograph of BRO7VM powders before (a) a nd after sonication and sedimentation (b), as well as ESBVM powders before (c) and after (d) sonication.......... 124 5-6 SEM image of fully-fired BRO7-ESB 20 cathode system s used in this study BRO7S-ESBS (a), BRO7VM-ESBS (b), BRO7S-ESBVM (c), BRO7VM-ESBVM (d)........... 125 5-7 Nyquist (a) and Bode (b) plots at 62 5 C for different 50-50 wt% BRO7-ESB20 electrode m icrostructures tested in air............................................................................. 126 5-8 Arrhenius plot of ASR vs reciproc al tem perature for the four different microstructures studied.................................................................................................... 127 5-9 Nyquist (a) and Bode (b) plots at 62 5 C for different 50-50 wt% BRO7-ESB20 electrode m icrostructures tested in ai r before (open symbols) and after......................... 128 5-10 Arrhenius plot of ASR vs reciprocal tem peraturea comparison between electrodes prepared from as-prepared powders (open symbols)....................................................... 129 5-11 Effect of electrode thickness on ASR at 625 C for the four different electrode m icrostructures prepared after ul trasonication and sedimentation.................................. 130 5-12 Nyquist (a,c) and Bode (b,d) plots at 625 C for 50-50 wt% BRO7-ESB20 at different thicknesses without (open sym bols) and with (closed symbols) pure BRO7... 131 5-13 Arrhenius plot of ASR vs reciprocal tem peraturea comparison between electrodes without (open symbols) and with (c losed symbols) current collectors............................ 132 5-14 Solid state adaptation of 3-point Lu ggin probe configuration (a) schem atic representation and (b) actual cell..................................................................................... 133 5-15 Impedance spectra for LSM-YSZ compos ite co mparing and total cell impedance measured using 2-point configuration.............................................................................133 5-16 Current-voltage measurement for LSMYSZ on Luggin probe cell at 650 C using hydrogen bubbled through wate r as the fuel gas and ai r as the oxidant gas. ................... 134 5-17 Cathode overpotential versus current density data for selected cathode m aterials on Luggin probe cells at 650 C usi ng hydrogen bubbled through water............................135 5-18 Current-voltage measurement for sel ected cells at 650 C using hydrogen bubbled through water as the fuel gas and air as the oxidant gas. ................................................. 136 5-19 SEM image of optimized BRO7-ESB co m posite cathode on SOFC with Ni-GDC anode support with ESB/GDC bilayer SO FC after current-voltage testing.................... 137 A-1 (a) Equivalent circuit model and (b) t ypical im pedance spectroscopy cell response...... 152 A-2 Impedance response showing diffusi on behavior at high frequencies............................. 152

PAGE 12

12 A-3 Impedance response for as-sintered (1500 C 4 h) HS3Y sa mples (circles), annealed at 1200 C for 110 h in both 10% H2 balance N2 atmosphere (triangles)........................ 152 A-4 (a) Response and (b) equivale nt circuit for m ixed kinetic an d charge transfer control... 153 A-5 (a) Fuel cell testing schema tic and (b) illustrative repres entation of typical response. ... 153 A-6 (a) Typical current-interruption respon se, and (b) calculated experim ental cathodic overpotentials...................................................................................................................153 A-7 Alternative electrochemical tes ting setup for cathodic overpotential. ............................. 154 A-8 Illustrative representation of separation of electrode contribution from fuel cell test response............................................................................................................................154 A-9 Alternative determination of electrode polarization from symmetrical cell I-V m easurements...................................................................................................................155 A-10 Electrode arrangement for DC conductivity measurements............................................ 155 A-11 Oxygen nonstoichiometry data for various Co and Fe B-site dopant concentrations......155

PAGE 13

13 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MICROSTRUCTURAL ENGINEERING OF COMPOSITE CATHODE SYSTEMS FOR INTERMEDIATE AND LOW-TEMPERAT URE SOLID OXIDE FUEL CELLS By Matthew Camaratta December 2007 Chair: Dr. Eric Wachsman Major: Materials Science and Engineering Solid oxide fuel cells (SOFCs) are electrochem ical devices with the potential to generate power at high efficiency with li ttle environmental impact. Howe ver, in order to improve their commercial appeal, operating temperatures must be lowered from the 800-1000 C temperature range to 500-700 C and below. Due to the high bond strength of oxygen mo lecules, the kinetics of oxygen reduction are orders of magnitude slower than those of fuel oxidation. Consequently, much research in the reduced-temperature SOFC field is aimed at enhancing cathode performance. A composite cathode makes use of an electr onic conducting phase as well as an ion conducting phase in order to spread the 3PB reaction zone beyond the cathode/electrolyte interface. Silver-stabilized bismuth oxide comp osite cathodes exhibit low resistance to oxygen reduction due to a combination of high catalytic activity for oxygen reduc tion of both phases, as well as high ionic conductivity of the bismuth oxide phase. Isothermal comparisons were made between pur e silver cathodes, silver-yttrium stabilized bismuth oxide (YSB) cathodes, a nd silver-erbium stabilized bi smuth oxides (ESB) at 650 C. The performance of all cathodes was shown to degrade with time. Cathode area specific

PAGE 14

14 resistance (ASR) of both the Ag-YSB and Ag -ESB electrodes increased by around 70%, while the pure Ag system experienced a near fourfold increase during the same length of time under open circuit conditions. In light of the electrochemical, microstruc tural, and chemical evidence presented, it was concluded that electrode microstructural evolu tion due to growth, agglomeration, and coalescence of the silver pha se, rather than chemi cal reactivity of the bismuth oxide phase, was responsible for the observed degradation in electrochemical performance. Attempts were made to reduce the microstructural evolution of the silver phase in AgESB20 composites by introduction of small par ticles (nano-size 8YSZ or vibratory-milled ESB20 particles) into the elec trode. The addition of 5 vol% 8YSZ nano powders significantly improved unbiased electrode stabili ty by 97 %, and reduced the initial, zero time ASR value by 31 %. Similar results were obtained when YS Z-free electrodes were prepared from ESB20 powders composed of part icles hundreds of nanomete rs in size as opposed to electrodes prepared from ESB20 powders composed of micron-sized particlesth e zero time ASR value was reduced by 25 %, and ASR vs. time slope during unbi ased testing of the s ilver-ESB20 system at 650 C was reduced by 95 %. The ASR vs. time slopes during testing unde r a 250 mV external applied bias were lowered by 50 % using the sma ller ESB20 particles due to suppression silver phase electro-migration. Porous composite electrodes consisting of BRO7 and ESB20 were also synthesized and characterized using impedance spectroscopy on symmetrical cells. Us ing a combination of compositional and microstructural optimi zation, a minimum electrode ASR of 0.73 cm2 and 0.03 cm2 was achieved at 500 C and 700 C, resp ectively, making it on e of the lowest resistance cathode materials reported to date at such low temperatures.

PAGE 15

15 CHAPTER 1 INTRODUCTION There is an expanding list of reasons to e xplore alternatives to conventional energy generation m ethods (namely coal-fired stea m turbines and gasoline-powered combustion engines). From an ecological standpoint, th ere has been a longstanding need to reduce the polluting byproducts of conventional energy generation. Greenhouse emissions from automobiles and power plants have long been susp ected of contributing to global climate change. Particulate matter exhausted from buses and larger vehicles can penetrate into the lungs, causing respiratory and cardiac disease. From an economic standpoint, worldwide demand for energy continues to rise as more and more nations, notab ly China, join the ranks of the industrialized countries. Also, the cost of o il is subject to global political landscapesthe dramatic upswing in gasoline and diesel fuel prices during winter and early spring 2000 were the direct result of an intentional reduction in crude oil production by OPEC. This exemp lifies how sensitive oil prices are to a relatively small (about 4%)1 reduction in supply, and given th at oil is a finite resource, this signifies what may happen as supplies eventually begin to dwindle. From a socio-political standpoint, there has been a lot of tension re garding the foreign polic ies of oil-consuming nations, as evidenced by frequent protests and even wars. Lo wering the global dependence on oil might reduce such tension. Yet despite all of this, changes to en ergy generation methods throughout the world have been exceedingly slow, as evidenced by the inte rnal combustion Otto cycle engine, which has been used in automobiles for more than a century. Obviously, the concept of alternative energy is not new. Many different sources of energy exist (or are proposed) that can supplement, a nd have the potential to replace, conventional generation methods. Popular ex amples include solar radiati on, wind motion, and fusion energy. Each of these technologies has its own set of problems that have plagued its commercial

PAGE 16

16 inception, but much research is being conducted to overcome these problems. Presently, it is unknown which of these sources, if any, will reali ze the greatest share of the commercial energy market in the coming decades. What is certain, how ever, is that the market for alternative highly efficient, eco-friendly energy produ ction technologies is expanding. There is a general push towards higher e fficiencies. Current combustion-powered automobiles have an overall efficiency of about 20%.2 That is, only 20% of the thermal energy content of the gasoline is converted into useful mechanical work and the rest is wasted. Higher efficiencies translate into reduced energy costs per unit of work done. Fuel cells, another alte rnative energy technology, have received growing interest in recent years. The fuel cell, an invention credited to Grove in 1939,3 is a device that converts the chemical energy of fuels directly into electric ity and heat by electrochemically combining the fuel and an oxidant gas via an ion-conducting electrolyte. The need for direct combustion is eliminated, giving fuel cells much higher conve rsion efficiencies than conventional thermomechanical methods (Figure 1-1).4 As a result, fuel cells have the added advantage of lower CO2 emissions per unit of power output comp ared to conventional technologies.5 Moreover, the production of NOx and SOx, the main components of smog and aci d rain, are expected to be 90% lower than in conventional pulverized coal plants5the NOx and SOx emissions from a 100 kW Siemens Westinghouse SOFC system were both reported to be less than 1 ppmV.6 Fuel cell power generation is virtually noise-free and can be tailored for standalone, off-grid applications, eliminating the need for, and losses associated with power transmission lines. Additionally, fuel cells have the potential to be used for combined heat and power (CHP or cogeneration) applications. This further increases efficiency by utilizing the thermal energy that is produced

PAGE 17

17 generating electricity. The current trend to dere gulate the production of electric power will serve to promote CHP, thus furthering the marketability of fuel cells. There are several different kinds of fuel ce lls in development. The main difference between each is the material used for the electrolyte, and conse quently the operating temperature ranges and thus plausible market applications. Solid oxide fuel cells (SOFC) utilize a fast oxygen ion-conducting electrolyte, which is capable of supporting a high flux of oxide ions (O2-) while in the solid state. A great deal of heat is produced during the energy production process. This heat can be utilized to partially reform common hydrocar bon fuelsgasoline, diesel, and alcoholinternally without combustion, unlike the case of polymer electrolyte membrane fuel cells (PEMFC) where external partial combusti on reformers are required to produce hydrogen gas. The CO gas created during the reforming process is consumed as an additional fuel (unlike the case of PEMFC where CO gas is a catalyst pois on). The generated heat can also be utilized for CHP applications. These points add to the e fficiency of SOFC. Als o, SOFC are made from inexpensive ceramic materials, a nd dont require precious metal catalysts like platinum (which is required for lower temperature fuel cells). For these reasons, some experts predict that solid oxide fuel cells will be the eventu al winner in the automotive market.4 There is one major drawback of SOFC that has hurt their viability in this market. The high operating temperatures (800-1000C) of curre nt SOFC technology w ould require extended startup timesa period of fuel burning is needed to reach these operating temperatures. High operating temperatures are also responsible for sealing problems and expensive interconnect and balance-of-plant materials for SOFC stacks. Additionally, high temperature operation can induce thermal stresses at electroly te-electrode interfaces, as well as cause interdiffusion between cell components.7 Most of these problems will be solved if the operating temperature can be

PAGE 18

18 lowered to around 500C.8 An added bonus of lowered operati ng temperatures is the possibility of direct oxidation of methane without carbon deposition, which would eliminate the need for fuel reforming.9 For these and other reasons, the move ment to lower the operating temperature of SOFC has intensified. Howe ver, the barriers to low operati ng temperatures are significant. To gain a better insight into lowering the operating temperature of SOFC, one must first understand the construction of SOFC the various reactions and losse s that occur in the different components of SOFC, as well as explore some of the progress already made by researchers.

PAGE 19

19 Figure 1-1. Energy production efficiency comparisons [Rep rinted from B. Weins, The Future of Fuel Cells, http://www.benwiens.com/energy4.html Copyright (2002), with perm ission from Ben Wiens Energy Science, Inc.].

PAGE 20

20 CHAPTER 2 LITERATURE REVIEW 2.1. Basic Principles of SOFCs Solid oxide fuel cells are capable of opera ting over a wide range of tem peratures. Intermediate temperature SOFCs (IT-SOFCs) ope rate in the range from 500 C to 700 C.10 A typical single SOFC [Figure 2-1] is composed of a dense cera mic oxide electrolyte separating two porous electrode catalystsa ca thode on the oxidant (air) side a nd an anode on the fuel side. Oxygen molecules are fed to the cathode where it combines with electrons (supplied from the electrocatalyst in the cathode which is electrical ly connected to an external circuit) and is reduced into oxygen ions, which are injected into the el ectrolyte (Eq. 2-1a), where the subscripts c and e stand for cathode and electrolyte, re spectively. Note that at times it is more convenient to think of the pr ocess of oxygen ions moving th rough the electrolyte as the equivalent process of oxygen vacancies (which ca n be considered simply as point defects in a periodic oxide lattice where a normally occupied oxygen site is left vacant) migrating from the anode side towards the cathode si de, depicted in Eq. 2-1b using Krger-Vink notation. That is, 2 224e cOeO (2-1a) x OOOVeO .. 22 2 1 (2-1b) The electrolyte conducts the ions but blocks electrons. The o xygen ions emerge on the anode side of the electrolyte where they react with fuel (H2 and CO) to form H2O, CO2 and electrons, eOHHOa e4 22 2 2 (2-2) eCOCOOa e4 22 2 (2-3) where the subscript a stands for the anode.

PAGE 21

21 Electrons produced during the reaction are rele ased to the electrocatalyst in the anode (which is also electrically conn ected to the external circuit). Thus the circuit is complete and useful work can be done in the external circuit. The gradient in oxygen concentration (or oxygen partial pressure or oxygen chemical potential) acr oss the electrolyte provides the driving force for all of these electrochemical processes. Thus, SOFCs are considered oxygen concentration cells with a theoretical re versible voltage (or electromotive force, emf), Eth, given by the Nernst equation a c thpO pO F RT E2 2ln 4 (2-4) where R is the gas constant, T is the absolute temperature, F is Faradays constant, and apO2 and cpO2 are the partial pressures of oxygen at the a node and cathode, respectivel y. The coefficient in the denominator repres ents the number of electrons transferred per mole of oxygen molecules reacted in the cell. As already stated, air is usually fed as the oxidant to the cathode side, thus cpO2 is 0.21 atm. The oxygen partial pressure on the anode side for any particular fuel can be predicted using the Gibbs free energy change equation eq O T O T O TKRTSTHG ln (2-5) where O TH is the heat change and O TS is the entropy change of th e fuel gas reaction at any anode temperature, T. Keq is the equilibrium constant for the fuel gas reaction. For many laboratory SOFC experiments, the fuel is hydrogen gas bubbled th rough water at room temperature. The fuel gas reaction is )(2)(2 )(22 1g g gOHOH (2-6) and the mass action expression is

PAGE 22

22 2 1 22 2))(( pOpH OpH Keq (2-7) Rearrangement and substitution of Eq. 2-7 into Eq. 2-6 yields RT STH pH OpH pOO T O T2 exp 22 2 2 (2-8) The vapor pressure of water at any bubbler temperature, Tbub as well as the heat and entropy change values, O TH and O TS are readily available in tabulated form.11,12 Assuming no water condensation occurs in the gas delivery system between the water bubbler and the SOFC anode, the expression for pO2 in Eq. 2-8 can be substituted for apO2 in the Nernst Equation (Eq. 2-4) to predict the theoretical Nernst potentia l at any cell temperature. Plots of apO2 and Nernst potential versus cell temperature using hydrogen fuel bubbled through water at 25 C are shown in Figure 2-2. To increase theoretical cell voltage and power output, multiple cells are usually combined in series into a stack with adjoin ing anodes and cathodes separated by an interconnect material (Figure 2-3). No cell is perfectthat is the cell voltage during operation is always lower than the theoretical Nernst value (Fig. 2-4). As current is drawn to the load, the cell voltage is reduced due to various electrochemical losses. For an SOFC with an ionic transference number, ti=1 (discussed below in Section 2.2), cell polarization or overpotential, is the difference between theoretical and operating voltages. The total cell polarization can be considered the sum of three individual contributionsresis tance (ohmic) polarization ( ), charge transfer (activation) polarization ( A), and diffusion (concentration) polarization ( D). tot = + A + D (2-9)

PAGE 23

23 Ohmic polarization is caused by resistance to the flow of i ons in ionic conductors and electrons in electronic conductors and by contact resistances betw een cell components. As the name implies, these potential losses are directly proportional to the amount of current passing through the cell. Activation polarization ensues whenever re acting chemical (incl uding electrochemical) species are involved. An activa tion energy barrier must be overcom e in order for the reaction to proceed. This results in a potential drop which may be regarded as the extra potential necessary to overcome the energy barrier of the rate-determining step of the reaction, and is related to the current passing through the cell by the phenomenological Butler-Volmer relation (discussed below in Section 2.4). In the electrode reactions (Eq. 2-1 thr ough 2-3), the activation polarization is due to the transf er of charges between the elec tronic and ionic conductors. For this reason this overpotential is often termed charge-transfer polarization. Concentration polarization occurs when reac ting species are supplie d to reaction sites slower than they are consumed, or when reactio n products are not removed fast enough so that they block the reaction sites. This effectively re sults in lower concentrations of reactant species at the reaction sites (lower oxygen partial pressures at the cathode or lower partial pressures of fuel at the anode) than in the bulk of the gas stream and leads to a concomitant reduction in voltage. This type of loss becomes significant when large amounts of curr ent are drawn from the cell, and in some cases a limiting current is r eached where the concentrations of gases are unsustainable (approach zero at the reaction sites) causing fuel ce ll operation to cease. While these losses cannot be completely el iminated, they can be minimized by proper choice of materials and by microstructura l engineering of each cell component.

PAGE 24

24 2.2. Dominant Component Losses 2.2.1. The Electrolyte The dominant polarization losses in SOFC el ectrolytes are ohmic losses caused by the resistance to ionic conduction. In fact, the main contribution to the ohmic polarization of a SOFC is due to the electrolyte.13 As stated earlier, the role of an SOFC electrolyte is to conduct oxygen ions from the air side of the cell towards the fuel side. Good electrolyte materials have high ionic conductivity (low resist ivity) for oxygen ions (or equivalently oxygen vacancies), and low electronic conductivity (to prevent internal short-circuiting and thus reduced cell performance) under an applied fiel d. That is, ideally, the elec trolyte should have an ionic transference number equal to unity, ei i it (2-10a) where ti is the ionic transference number, i is the ionic conductivity (S/cm), and e is the electronic conductivity (S/cm). This results in a deviati on from theoretical Nernst potential, allowing ti to be measured using N O iE E t (2-10b) where EN is the theoretical Nernst po tential, discussed above, and EO is the actual cell open circuit potentia l (OCP). A material that has a low migration enthalpy for oxygen vacancies (or ions) in its lattice will have high ionic conductivity.14 The mobility of vacancies or ions within a lattice is affected, for example, by the spacing between ions in the lattice (w hich is related to the lattice parameter of the lattice), the binding energy between counterions and oxygen ions (which is dependant on the type of cations in the lattice), and also what are known as defect associates which are caused by bonding between these defect sp ecies. The concentration of defects also

PAGE 25

25 affects the ionic conductivity. Oxygen ions move by jumping into an oxygen vacancy. If only one vacancy were available for such a jump, only one ion could move at a time. By increasing the number of oxygen vacancies, the number of mob ile ions also increases and thus the ionic conduction is increased. However, at a certain concentration the defects can begin to interact, and these interactions tend to lower ionic conduction. One crystal structure that is seen repeatedly in SOFC electrolytes is the cubic fluorite structure (Figure 2-5). A va riety of oxides can exist in the fluorite structure (AO2), such as zirconia (ZrO2), ceria (CeO2), and bismuth oxide (Bi2O3). The configuration of the two sublattices and its open or spacious nature allow for relatively facile defect migration in fluorite oxides. For many of these materials however, the cubic fluorite structure is a hightemperature polymorph. For example zirconia u ndergoes a cubic-tetragonal transition below around 2400C, and bismuth oxide undergoes a transition from cubic to monoclinic around 730C.15 It has been shown that the cubic fluorite structures of these materials can be maintained down to room temperature by doping with variou s trivalent (and divalent ) transition metal and rare-earth oxides. Typically, dopant cations with i onic radii larger than that of the host cation are more effective in stabilizing th e cubic fluorite phasesmaller dopa nt cations tend to distort the lattice more than larger ones.16 In addition to imparting structur al stability to these materials, these oxides may also serve to improve ionic conduction by virtue of the requirement that the lattice be electronically neutral. Th at is, the aliovalent dopant cations (Re+3), which substitute the host cations (A+4), have a net negative charge compared to normal lattice positions. However during this substitution process the lattice must maintain charge neutrality. Two of these negative charges can be compensated for by the creation of a vacant oxygen site (or oxygen vacancy), which has a +2 charge. In Krger-Vink notation the dopant incorporation reaction is

PAGE 26

26 x O OA AOOV O 4 Re2 Re2.. 22 (2-11) Thus, the stabilized material is also impart ed with a certain amount of oxygen vacancy defects, which (up to a point) increases ionic conduction. One can see from this that the electrolytic behavior of a material can be improved by optimi zation of the type and concentration of dopant cations in the lattice. The conventional SOFC electr olyte, yttria-doped zircon ia (YSZ), has an operating temperature around 1000C. As the operating temp erature is lowered, ohmic losses across the solid electrolyte increase.8,17 These losses can be minimized by reducing the distance that ions and vacancies need to travel, i. e. by reducing electrolyte thickness, and by choosing a material that has higher ionic conductivities than YSZ.8 Several physical and chemical techniques can be used to deposit thin-film electrolytes on to a node substrates including vapor deposition, slip casting, and electrophoretic deposition.18 Significant reductions in electrolyte resistances have been achieved by use of anode-s upported thin-film electrolytes.8,17 Two alternative materials which have shown improved performance over YSZ are doped ceria8,19 and stabilized bismuth oxide.20,21 At intermediate temperatures (~500-750 C), stabilized bismuth oxides show some of the highest ionic conductivities among all studie d fast ion conducting oxides (Figure 2-6).22 This is attributable to the inherent nature of its lattice structure. Because bismuth is a trivalent cation, and because it can be stabilized in the cubic fluor ite form, bismuth oxides intrinsically have of the regular fluorite anion sites vacant. Conse quently, there is a larg e excess of equipotential lattice points among which the anio ns (oxygen ions) can be distributed, which in effect leads to the observed high conductivities. St abilized bismuth oxide thus has the potential to provide high power densities at low temperatur es. However use of this material in fuel cell electrolyte

PAGE 27

27 applications has been limited because it become s unstable in reducing conditions such as those experienced at fuel cell anodes.23 While the ionic conductivity of st abilized ceria is le ss than that of stabilized bismuth oxide, it has also shown great potenti al to replace YSZ in fuel cel l electrolyte applications.8,24 However ceria also has limitations at reducing conditions The electrolytic domain of 10 mol% and 20 mol% GDC persists until ~10-15 atm at 800 C and ~10-20 atm at 700 C.25 Below these pO2 values, some of the Ce+4 ions take on an electron and are reduced to Ce+3, creating CeCe defects (also known as small polarons). These defect s can be conducted through the electrolyte by a hopping mechanism from CeCe sites to x CeCe sites. This results in n-type conduction, lowered ionic transference number, shortcircuiting of the cell and thus reduced cell efficiency. This leads to further problems when reducing electrolyte thickness to lower ohmic losses. When mixed conductivity is present, any electronic leakage current will increase as electrolyte thickness is decreased.8 2.2.2. The Electrodes The dominant losses in the electrodes of SOFC are activatio n polarization and concentration polarization. Activ ation polarization is typically more of an issue in cathodes due primarily to the kinetic s of the oxygen reduction reaction (attr ibutable to the high bond strength present in oxygen molecules), whic h is several orders of magnit ude slower than the reactions involving fuel oxidation.26-32 The dissociated chemisorption of H2 in a typical SOFC anode, for example, is a non-activated process.32 That is there is no barr ier to the dissociation of a hydrogen molecule into two separate adsorbed atoms on (open, atomically rough faces of) the catalyst surface (typically Ni). At the cathode, the dissociative adsorption of oxygen is a thermally activated process.33 As the operating temperature is lowered, the rates of chemical

PAGE 28

28 reactions drop dramatically, and activation polarization at the cathode becomes an even larger issue. Thus anode-supported cells have gained wider acceptance than cathode-supported cells in low-temperature SOFC research. This use of thick anodes results in longer diffusion lengths, and concentration polarization can become an issu e, especially at high current densities. It should be noted that an analysis of isotherm al transport of gaseous species through porous electrodes, based on commonly used anode and ca thode gases, suggests that for comparable porosities and electrode thickness, the concentration polarization effects should in general be lower on the anode side.13 A complete polarization mode l of SOFC has been developed34 and the various loss mechanisms for an anode-suppor ted cell are shown agai nst a power density curve in Figure 2-7. These curves were generated using data extracted from literature that focused on cells that operated at 800C. It should be noted that the values used for exchange current densities, electrolyte c onductivity and thickness were am ong the highest values obtained from the literature. Based on the fact that the ce ll voltage reaches zero before an anodic limiting current is reached, the authors conclude that the anode limiti ng current density will not be reached under normal operating conditions. In addition, experimental studies have shown9 that the low-current interfacial resistance was 70-85% of the total cell resistance from 550 C to 800 C, and still played a major role in limiting ce ll performance at higher current densities. For these reasons, development of highly active catho des, based on improvements in materials and microstructures that are thermally and chemical ly compatible with adjacent stack components are critical to the successful operation of low temperature SOFC. 2.3. Classes of Cathode Materials The oxygen reduction reaction (Eq. 2-1) is th e net reaction of interest for all SOFC cathodes. There are a series of steps involved in the reduction mechanism which depend on the type of cathode used. The cathode resistance emerge s from these processes. In order to improve

PAGE 29

29 the performance of cathodes for low temperatur e SOFC, whether by opti mizing microstructure or composition, one must understand the various po ssible steps involved in th is overall reaction. A review of three basic cate gories of porous cathode mate rialssingle phase electronic conducting cathodes, dual phase el ectronic and ionically conducting cathodes, and single phase mixed ionic and electronic conduc ting (MIEC) cathodesis given below to help gain insight into these processes. The review will incl ude sections covering the mechanism for oxygen reduction, experimental (and theoretical) resu lts, and methods for the determination and enhancement of the rate-limiting step of the reduction reaction. The experimental methods mentioned in this section are discussed in some detail in Appendix A of this dissertation. 2.3.1. Single-Phase Electronic Conductor The most conventional type of cathode is a porous single-phase el ectronic conductor. Traditional cathodes for YSZ electro lytes such as platinum (and other noble metals), and the electronically conducting ceramic material, La1-xSrxMnO3 (LSM; is the average number of vacant oxygen sites per unit cell, or more si mply, the oxygen nonstoichiometry, which ranges from +0.01 to +0.06 at 1000 C and 700 C, resp ectively, in air, for the typical composition x=0.2)35, fall into this category. This particul ar construction restri cts the oxygen reduction reaction (Eq. 2-1) to one-dimensional regions where the three phasesgas phase (pores), electrode phase, and electrolyte phasemeet, or th ree-phase boundaries (3PB). In this case, one can visualize that such 3PB are limited to th e electrode/electrolyte interface (Figure 2-8). The gas phase supplies the oxygen species, the electrode phase supplies the electrons and catalyzes the reaction, the oxygendeficient electrolyte phase provi des the vacant sites into which the ionic oxygen species are injected. Generally, the reaction mech anism for this type of cathode involves gaseous oxygen molecule s diffusing through the pores (or they may dissociate and adsorb onto the surface of the cathode and diffuse along the pore walls) toward the reaction sites

PAGE 30

30 (3PB), where charge transfer occurs and oxygen ions are injected into the electrolyte. Naturally, the reaction cannot occur on a one-d imensional line. Before comb ination with an electrolyte oxygen vacancy, it is expected that some diffusion of oxygen ad-atoms over the electrode and electrolyte surfaces takes place, but this reac tion zone is expected to be very narrow.16 LSM cathodes have been particularly useful for these conventional SOFC due to their relatively low cost (compared to noble metal catalysts) and their thermal and chemical compatibility with YSZ electrolytes. In fact, the LSM cathode/YSZ electrolyte combination is the most widely studied system in the SOFC field.36 Such cathode materials have prov en effective for SOFC operating at high temperatures (greater than 1000 C) where reaction kinetics are fast However, for cells that are to operate at lower temperatures the polarization losses in these cathode materials become so large that practical power densities cannot be achieved. A great deal of materials engineering has thus been necessary in order to overcome these losses. 2.3.2. Dual-Phase Composite Cathodes One of the means to this end has been th e development of porous dual phase composite cathodes where one phases is a pure ly electronic conductor and the ot her phase is a purely ionic conductor. The rationale for the use of this type of cathode is that it allo ws the reaction zone to spread from the cathode/electrolyte interface into the cathode (Figure 2-9). The length and density of the 3PB is increased, i.e., more site s are available for the charge transfer reaction.13 The general reaction mechanism is essentially th e same as described a bove for purely electronic conducting cathodes except that th e diffusion distances for th e oxygen species to the 3PB reaction sites are shorter for dual phase cathode materials. Also, th ere is an additional path for transport of oxygen speciesbulk diffusi on through the ionically conducting phase. The composition, including the materials used for each constituent phase and the relative amounts of each phase, is a key factor in de termining the performance of these dual phase

PAGE 31

31 composite cathodes. Consideration must be given to ensure that the indi vidual phases within the composite electrode are continuous in order to create a path for the transport of the respective speciesone phase for electrons, one phase for io ns, and pores for gases. According to the effective medium percolation theory (EMPT),37 in order for a randomly di stributed solid phase to be continuous, its volume frac tion should exceed 1/3 of the total volume of the composite, including porosity. In addition, the cathode must be chemically and thermally compatible with the electrolyte. For this reason, the ionic conducting phase is usually chosen to be the same material as the electrolyte. The choice of the electronically conducting ph ase has a large effect on the inherent catalytic activity of the compos ite cathode towards the oxy gen reduction reaction. This phase must be compatible with the ionically conducting phase for structural stability and to keep the intrinsic interfacial resi stances between these two phases low.38 Common examples of this type of cathode include LSM/YSZ, and Ag/YSB (yttria stabi lized bismuth oxide).39 2.3.3. Single-Phase Mixed Ionic and Electronic Conductors (MIECs) One class of SOFC cathodes that is growing in popularity includes sing le-phase oxides that exhibit mixed ionic and electroni c conductivity. The most frequen tly used and versatile crystal structure in this category is the ABO3 perovskite structure (Figure 2-10). Typical host lattices of perovskites used for SOFC cathode applications are composed of a trivalent lanthanide cation on the A-site and a transition metal cation on the B-site. Electronic conduction is possible due to the multivalent nature of the B-site cation. If the cation has an affinity towards increasing its valence state (say from +3 to +4), p-type electronic conduction can occur by an electron-hol e hopping mechanism (also a small polaron mechanism). In contrast, if the cation has a tendency to want to lower its valence state (from +3 to +2, for example), n-type electronic c onduction can also occur via a small polaron electron hopping mechanism. Note that this la tter example can also have the e ffect of creating oxygen vacancies

PAGE 32

32 in order to charge-compensate the lattice, while the former example can have the opposite effect of oxygen vacancy annihilation. Oxygen mobility through vacancies provides the basis for ionic conductivity in these pe rovskite materials.40 Another way defects may be introduced is by doping the A-site cation with a divalent cation, typically Sr+2 or Ca+2. In air and at low cathodic overpotentials, the perovskite material is nearly stoichiometric, hence charge compensation for negative defects caused by substitu tion of divalent (e.g., Sr) catio ns for the trivalent (e.g., La) cation is mostly obtained by formation of electron holes.41 Under conditions of reasonably high cathodic overpoten tial, the pO2 local to the reaction sites is lowered, and as a result of defect equilibrium requirements, oxygen vacan cies concentration is increased. The presence of these electronic and ionic de fects broadens the electrochemically active region of the cathode to a much larger surface areathe oxygen reduction reaction is no longer limited to regions close to the triple phase boundari es. In other words, since both electrons (or electron holes) and oxygen vacancies are mobile defects in a MIEC cathode, it is possible for oxygen to be reduced at the gas/MIEC interface according to (Eq. 2-12a) or (Eq. 2-12b). x MIECO MIECO MIEC gOVeO)( .. )()()(22 2 1 (2-12a) )()( .. )()(22 2 1MIEC x MIECO MIEC gh OVO (2-12b) Furthermore, an additional path through which oxygen species may be supplied to the cathode/electrolyte interfacebulk transport thro ugh the electrode by vacancy jumping. At reasonable cathodic overpotentials (noted above), this oxygen flux th rough the electrode is likely to be increased. Consequently, MIEC cathode ma terials typically exhib it lower polarization and have improved performance compared to the othe r classes of cathodes already discussed. The reaction path for oxygen reduction in su ch MIEC cathodes has been described27,30,42 as including

PAGE 33

33 (1) oxygen diffusion in the gas phase, (2) oxygen adsorption-desorption on the MIEC cathode surface with charge transfer and oxygen incorpor ation at the MIEC catho de surface, (3) surface diffusion of the adsorbed oxygen on the MIEC surf ace, (4) bulk diffusion of oxygen ions in the MIEC cathode, (5) oxygen ion transfer at the cathode/electrolyte inte rface, and (6) charge transfer and incorporatio n of the adsorbed oxygen at the 3PB (Figure 2-11). With some groundwork laid, some of the e xperimental and theo retical results and comparisons between the performances of different SOFC cathode materials will now be discussed. 2.4. Cathode Materials Conven tional and State-of-the-Art The rate at which charge transfer reactions occur at the electrocatal yst/electrolyte interface is often expressed by the aforementioned Butler-V olmer equation, which relates this current density, i, to the activation overpotential, act by RT nF RT nF iiact act O 1 exp exp (2-13) where io is the exchange current density, is the transfer coefficient, n is the number of electrons participating in the electrode reaction. F, R, and T were defined in Section 2.1. At sufficiently low current densities, Eq. 2-13 may be approximated by iRi nFi RTct O act (2-14) Where O ctnFi RT R is an intrinsic charge transfer re sistance which is restricted to the electrolyte/electrocatalyst interface.13 Note that this resistance term (or any resistance term) is

PAGE 34

34 often normalized into an area specific resist ance (ASR). This is accomplished simply by multiplying the resistance by an area term (such as electrode area). Virkar et al.13 have developed an expression showing the profound effect of adding an ionically conducting phase (such as YSZ) to a conventional electronically conducting cathode (such as LSM). In composite electrodes, the re action zone, or region over which the process of charge transfer occurs, is spread out from th e electrocatlayst/electrolyte interface into the electrode and as a result the activ ation overpotential may be lower than the intrinsic one given in Eq. 2-14. Thus an effective charge transfer resistance, eff ctR, is defined and used in its place; i.e. iReff ct act. The expression for eff ctR is rather lengthy and wont be reproduced here. However, assuming the electrode is sufficiently thick, eff ctR can be approximated by Vi ct eff ctV BR R 1 (2-15) where B is the grain size of the electrolyte (or ionically conducting phase) in the composite electrode, VV is the fractional porosity, and i is the ionic conductivity of the electrode (or that of the ionically conducting phase in the composite electrode). Usi ng as an illustration, typical values for the parametersRct=1.2 cm2, i=0.02 S/cm, B=2 m, VV=0.35give an effective charge transfer resistance for a porous dual-phase composite cathode, eff ctR =0.14 cm2. This is nearly an order of magnitude re duction in the intrinsic charge transfer resistance value of 1.2 cm2 (or equivalently an order of magnitude incr ease in the effective exchange current density value) obtained for the porous single-phase electronic conducting cathode where the reaction is limited to the cathode/electrolyte interface.

PAGE 35

35 It should be noted that the non-simplified model for eff ctR takes into account the effect of cathode thickness (Figure 212). In general, if i is sufficiently high, the value of eff ctR will decrease with thickness down to an asymptotic value (the plots where eff ctR increase with thickness simply illustrate the deleterious effect if i is low enough and explains the peculiar behavior of certain composite elec trodes reported in the literature).43 This effect of improved performance by increased thickness is validated by a series of current-v oltage plots in Figure 213. The decrease in performance for the 85 m thick sample is presumably due to concentration polarization effects. A value for eff ctR can be extracted from these plots by measuring the cell resistances from the near-lin ear regions and subtracting the electrolyte ASR (which can be obtained using independent impedance spectroscopy measurements, see Appendix A). These results agree well with the theoreti cal model (Figure 2-14). Mogensen and Skaarup32 observed the same behavior of lo wered polarization resistance w ith increasing thickness of LSMYSZ composite cathodes (Figure 215). These authors also re port that the addition of 50 weight% YSZ to La0.85Sr0.15MnO3 decreased the polarization resistance from 0.77 cm2 to 0.16 cm2 at 900C. Based on Eq. 2-15, increasing the ionic conduc tivity of the ion-conducting phase will improve cathode performance, everything else being constant. This effect has been explained in that higher ionic conductivities al low ions produced deeper with in the electrode to reach the electrolyte, thus effectively increas ing the width of the reaction zone.38 Murray and Barnett inspected this effect by substituting the YSZ phase with a material that is known (Figure 2-16) to have higher ionic conductivityg adolinium-doped ceria (GDC)and compared the behavior of LSM/YSZ and LSM/GDC composite cathodes, as well as that of pure LSM (Figure 2-16).38

PAGE 36

36 The results show that replacing the YSZ phase with GDC significantly lowers (by approximately 50%) the polarization resistance at each temperature observed. Note also the large improvement in behavior of both com posite cathodes over pure LSM. A comparison is also made to show how the type of electrolyte onto which an electrode is applied can influence performancethe polarization resistance of pure LSM on GDC is 1/3 that of pure LSM on YSZ. This result indicates that the kinetics of LS M electrochemical reacti ons are faster on GDC electrolyte surfaces than on YSZ surfaces. The effect of electrolyte is much smaller in LSM/GDC composites, but the slight reduction in polarization resistance observed on GDC electrolytes likely results from replacement of the LSM/YSZ 3PB at the cathode/electrolyte interface with LSM/GDC 3PB. However, the LS M/GDC 3PB already present in this composite cathode dominate the electrode performance, he nce the small effect of the electrolyte on performance. A model incorporating the effects of concentra tion polarization and activ ation polarization at higher current densities (usi ng a Tafel approximation to the Butler-Volmer equation, ibaactln (2-16) where a and b are parameters influenced by electr ode microstructure and th ickness) also predicts an effective exchange current density which is over an order of magnitude higher for LSM/YSZ composite cathodes compared to single-phase LSM cathodes at 800C.13 Wu and Liu7 reported the performance of dual phase yttrium-doped bismuth oxide (YSB)silver (YSB/Ag) composite cathodes on BaCe0.8Gd0.2O3 electrolytes (Figure 217). It is seen in Figure 2-17a that Ag/YSB shows a much reduced overpotential than Agat 40mA/cm2, the overpotential of YSB/Ag is only 52% that of pure Ag. It is wort h mentioning that there are two inherent problems that limit the performance of pure Ag cathodes. One is its relatively high

PAGE 37

37 thermal coefficient of expansion, yielding poor adhe sion to electrolyte surfaces. The other is that it readily densifies at low temper atures (850-900C), and results in dense electrodes and hence poor performance. The addition of a ceramic pha se to Ag, in addition to increasing the TBP length, can prevent electrode de nsification and improve thermal compatibility and adhesion to electrolyte surfaces.7 Figure 2-17b shows the effect of the volume fraction of Ag in the composite on cathode performance. A rapid drop (i ndicative of percolation behavior) is realized around 55 vol% Ag. If porosity ( 20 vol%) is taken into account, the vol% Ag and YSB at which minimum resistance is reached are 44% and 36%, re spectively. This agrees well with the EMPT, but is inconsistent with ambipolar diff usion theory, which is discussed below. For a specific electrode, a numbe r of conductivities may be de fined such as electronic, ionic and total. It is the simultaneous transport of electronic and ionic species that is crucial in electrodes that have mixed conductio n. This is measured as am bipolar conductivity. Note that since electron mobilities are much larger th an that of ions, ambipolar diffusion predicts37 that for these dual-phase composites, higher ambipolar c onductivities are expected if the volume fraction of the ion-conducting phase is higher than that of the electron-conducti ng phase (Figure 2-18), which is in contrast to the results in Figure 2-17b where the electron-co nducting phase volume fraction is larger. Yet in Figur e 2-19, a minimum resistance is reached at higher GDC content in LSM/GDC dual phase composites. The apparent contrast in beha vior can be explained by the simple fact that the overall pe rformance of a porous electrode is not solely determined by mixed ionic-electronic transport propert ies in the solid phase of the electrode, but also by the inherent catalytic property of the 3PB, as well as by gas transport to or away from the 3PB. One might conclude, therefore, that the i nherent catalytic activity of Ag is higher than that of LSM.

PAGE 38

38 Maguire et al.41 examined the performance of Co -rich (x=0.3) and Fe-rich (x=0.7) La0.84Sr0.16Co1-xFexO3 (LSCFe3 and LSCFe7, respectively) as well as Pt cathodes on GDC electrolytes. Figure 2-20a s hows the results of DC electri cal conductivity measurements (Appendix A) for LSCFe3 and LSCFe7 versus temperature. As expected, the Co-rich composition shows a higher conductivity th an the Fe-rich compositionat 800C, the conductivities are 643 S/cm and 115 S/cm for LS CFe3 and LSCFe7, respectively. Figure 2-20b shows the results from steady-state cathodic po larization measurements (Appendix A) for the two LSCF compositions as well as for Pt. Fo r simplicity the cathodic ove rpotential and current densities are expressed as positive quantities. For a given overpotential, LSCFe3 will pass a higher current than LSCFe7, and as expected bo th LSCF compositions will pass higher currents than Pt electrodes, since LSCF is an MIEC. An ion exchange depth profiling (IEDP) co mparison between manganates and cobaltferrites44 (Figure 2-21) shows that oxygen exchange coefficient, k (cm-1) (the rate constant associated with transfer of oxygen species ac ross a solid/gas interface), and the oxygen selfdiffusion coefficients, D* (cm2/s), are orders of magnitude higher in the cobalt-ferrites, indicating the relative ease with which such ch emical processes proceed in these materials (Appendix A). These results give credence to the supposition that mixed conducting cathodes extend the reaction zone beyond the 3PB. Ishihara et al.31 studied the effect of the A-site dopant in Ln0.6Sr0.4MnO3 (Ln=La,Pr,Nd,Sm,Gd,Yb, and Y) perovskite ca thodes on YSZ electrolytes. Figure 2-22 shows the results of electrical c onductivity, cathodic polarization, and fuel cell power density (Appendix A) measurements.

PAGE 39

39 In all cases, the Pr substituti on exhibits the highest performa nce. Note that the cathodic overpotential of PrSM is only 25% that of LaSM at the current density and temperature used for this measurement. At 1000C, the maximum po wer density is nearly the same for PrSM and LaSM, indicating that the electrolyte is the main contributor to the volta ge losses. At lower temperatures, the difference becomes more pronounced, illustrating th e importance of the activity for the cathodic r eaction. The voltage drop due to ca thodic overpotential was smaller in the PrSM cell, allowing it to exhibit the same power density at 700C as LaSM at 800C. Microstructural effects were negligible since BET surface area measuremen ts and particle sizes (Appendix A) were determined to be nearly the sa me in both these cases. Chemical effects were thus predominant. This has been explained in that Pr oxides exhibit a large nonstoichiometry in air since the stable valence number of Pr ions is intermediate between 3 and 4. In addition, Steele et al. reported that the ac tivity for the dissociation of oxyge n molecules to oxygen ions can be expressed by the exchange current density, and this strongly influences cathodic overpotential.31 The slope of cathodic overpotential-curre nt density plots pr ovides a measure of the exchange current densities (Appendix A). Th e exchange current density of PrSM was shown to be nearly three times gr eater than that of LaSM. Kharton et al.45 tested a number of different cath ode materials on bismuth oxide-based electrolytes. Bismuth ox ide was detected on the Ln1-xSrxCoO3 cathode surface after sintering. Other solid solution phases involving the diffu sion of lanthanum and strontium ion diffusion were also detected. The interact ion of cobaltites with YSZ electrolytes occurs much more slowly than with bismuth oxide-based electrolytes, even though YSZ electrolyt es are sintered 200C higher than bismuth oxide, as ev idenced by electrode resistances which are 2-4 times larger on yttrium stabilized bismuth oxide (YSB) el ectrolytes than on YSZ electrolytes.

PAGE 40

40 In a landmark paper, Shao and Haile46 reported on Ba0.5Sr0.5Co0.8Fe0.2O3 (BSCF) as a new cathode material for IT-SOFCs. Un like typical MIEC catho des, the A-site cation of BSCF is an alkaline-earth element as opposed to a rare-e arth element. Utilizing BSCF on a thin-film samarium-doped ceria (SDC) electroly te, high power densities (1010 mW/cm2 and 402 mW/cm2 at 600 C and 500 C, respectivel y) were obtained when operate d with humidified hydrogen as the fuel and air as the cathode gas. The cathode ASR, as determined symmetrical cell testing, was 0.055.071 cm2 at 600 C, and 0.51.60 cm2 at 500 C. The excellent performance of this material is reportedly due to its exceptio nally high oxygen diffusivity especially at lower temperatures, allowing for high ra tes of oxygen electro-oxidation. 2.5 Strategies for Improving Cathodic Performance In order to attack the problem of improvi ng cathode performance, one must not only understand the various mechanistic steps involved in the oxygen reduction r eaction but must also be able to identify the slowest (rate-determining ) step. In addition, on e must have a viable strategy toward accelerating this step. Thes e strategies may be broken down into two categoriesmorphological improvements, which involve changes to cathode geometry or microstructure and electrocatal ytic improvements, which enco mpass all other improvement methods. There are many approaches to these pr oblems; a few examples are discussed below. Tsai and Barnett9 illustrated several met hods for identifying mass-transport limitations in cathode materials. A voltage/power density versus current density plot is conducted first in air, then in pure O2. If the slope of the voltage-c urrent density plot, i.e. the resistance, is larger in air than it is in pure O2, mass-transport limitation can be assumed. Note that in the case of MIECs the ionic conductivity of the cathode may be pO2-dependent, hence this behavior may not necessarily be indicative of mass-transport limitati on. Mass-transport limita tions can arise due to

PAGE 41

41 surface diffusion, gas diffusion, dissociative adsorption9 or bulk transport.38 Both dissociative adsorption and surface-diffusion are thermally activated processes that should have strong temperature dependences, while gas diffusion will not. Thus, if any limiting current that is present increases as temperature is increased, one of the first two steps may be the limiting factor. Gas diffusion will depend on factors such as porosity and pore size. Thus, if any limiting current that is present increases as porosity a nd pore size increase, gas diffusion may be the limiting factor. These morphological factors can thus be utilized to accelerate the gas diffusion process. Also, if a change in test apparatus flow rates changes the polarization properties of the test cell, gas diffusion ma y be the limiting factor.20 For dual phase comp osite cathodes, if high frequency arcs are observed in the impedance spectrum whose behavior is independent of pO2, bulk ionic conduction in the ionic conducting phas e may be the limiting factor. This can be confirmed by a comparison of the activation ener gy (Appendix A) of the resistance of this high frequency arc in the composite cat hode to that for ionic conducti on in the pure ceramic phase. If these values are similar, bulk diffusion can be considered the limiting factor.38 Godickemeier et al.33 discussed how overpotential-curren t density relations hips (Appendix A) may behave under a pure diffusion-limited process, a pure charge-transfe r controlled process and combination of charge-transfer control at lo w current densities and diffusion control at high current densities (Figure 2-23). It is seen that there is a Butler-Volmer-type behavior for pure charge-transfer control, and a limiting current for pure diffusion control. For dual phase mixed conducting cathodes, one morphological approach to enhance the charge transfer rate is to reduce the partic le size of the ionic c onductor while keeping the electronic conductor partic les relatively large. The smaller particles form more contact points with the larger par ticles, creating a greater de nsity of reaction sites wi thin the cathode compared

PAGE 42

42 to the case when all particles are approximately the same size. This effect is illustrated in Figure 2-24 where a particle broken into four equal pie ces will have twice the 3PB length per unit area (8L) as the unbroken particle (4L). For single phase mixed conductors, a wide particle size distribution is desired.13 Small particles are advantageous since they tend to create large grain boundary areas. These grain boundaries are high-energy sites, where the char ge transfer reaction preferably occurs.13 However, small particles have th e disadvantage of creating large to rtuosity (harmful to gas phase diffusion) and poor connectivity.13 One electrocatalytic approach to improving charge transfer kinetics is to dope the electrode with redox cations47 such as Tb+4/Tb+3 or Pr+4/ Pr+3. Arrhenius plots of log(R) versus 1/T, where R is the cathode resistan ce of interest, yield straight lines with slopes that depend primar ily on the activation energy of the rate limiting process. A change in slope over the investigated temperature range thus indicates a change in the rate controlling process. An investigation30 of La1-xSrxCoO3cathodes (x=0.2, 0.3, and 0.4) on La1-xSrxGa1-yMgyO3(LSGM) electrolytes ut ilizes Arrhenius plots of the logarithm of interface conductivity (defined as the reciprocal of area-specific polarizat ion resistance) as a function of temperature for different applied cathodic voltage (E) c onditions (experimental details are discussed in Appendix A) to help eluc idate that the rate limiting step in these cells was due to bulk diffusion of oxygen ions in th e LSC cathode as opposed to surface diffusion of adsorbed oxygen on the LSC surface. The general procedure and results ar e discussed, briefly, below. Figure 2-25 shows the Arrhenius behavior at E= 0V. Since the activat ion energies for all samples are similar, it can be assumed that similar reaction mechanisms and the same ratelimiting step are at play. Figure 2-25b shows the Arrhenius behavior (for x=0.2 and x=0.4) for

PAGE 43

43 applied cathodic voltage conditions. It is seen that the activat ion energies for the lower Sr content sample are more sensitive to changes in applied voltage. As described in Section 2.3.3, under cathodic polarization conditions, the pO2 at the cathode/electrolyte interface is lowered and the oxygen vacancy concentration in the cathode increases due to defect equilibrium requirements. The driving force for surface diffusion depends on the oxygen partial pressure between the surface of the cathode gr ains and the reaction sites at the electrolyte surface. Since this gradient is generated by the applied cathodic overpotential, it should be similar for all samples (x=0.2, x=0.3, and x=0.4) at a given valu e of cathodic voltage. T hus, if this step were rate limiting, one would expect to see the simila r activation energies for x=0.2 and x=0.4 as at each voltage. This behavior was not observed, t hus surface diffusion is determined not to be the rate-limiting step. Bulk diffusion depends on oxygen vacancy concen tration. As mentioned above, applied cathodic voltages tend to increase the concentrati on of vacancies in the bulk of the cathode. Changes in vacancy concentrations (on a per centage basis) for LSC cathodes with lower concentrations of A-site dopants may be more se nsitive to changes in applied voltage than for those with high A-site dopant concentrations. As a result, the activation energies will be more sensitive to changes in voltage as well. Since th is is the experimentally observed behavior, this is assumed to be the limiting step. The surface oxygen exchange rate, k, determin ed by ion exchange depth profiling (IEDP) can be used to evaluate the surface activity of a cathode (Appendix A). A comparison of the activation energy of the k values with the activation energy of interface conductivity values at E=0V (described above) can cl arify whether or not oxygen adso rption/desorption on the cathode surface with charge transfer and oxygen incorporation at the cathode surface is rate-limiting. If

PAGE 44

44 these activation energies are similar, this step may be considered as rate-limiting. Oxygen surface exchange rates can be increased by increasing oxygen vacancy concentrations,48 or by doping the cathode with a noble metal49-51 since noble metals are know n to be good catalysts for the oxygen reduction reaction. The exchange current density (which can be determined experimentally as described in Appendix A) is related7,33 to pO2 at a given temperature by 2pOkiO (2-17) where k is a constant which is independent of pO2 and is an exponent that depends on the reaction mechanism and rate-limiting step. A plot of log(iO) versus log (pO2) should thus give insight into the rate-controlling mechanis m at a given temperature of interest. 2.6 Summary The above discussion detailed background info rmation on the basic principles of SOFCs, highlighted the importance of the cathode fo r reduced SOFC operating temperatures, and presented the three basic classes of SOFC cathodessingle ph ase electronic conductors, dual phase composite cathodes, and single phase MIEC conductors. In addition, conventional and state-of-the-art cathodes were presented, as were several methods by which the dominant mechanism for observed cathodic losses can be identified. Not all methods and materials described above were used in the following st udies, but a general awar eness of these facts provides an essential framework from which the studies could be built.

PAGE 45

45 Cathode Electrolyte Anode Cathode Electrolyte Anode Figure 2-1. Schematic representation of a typical SOFC button cell. 1.04 1.06 1.08 1.1 1.12 1.14 1.16 1.18 1.2 1.22 1.24 020040060080010001200 -90 -80 -70 -60 -50 -40 -30 -20 -10 0Temperature ( C)log [pO2(atm)] ENernst(V) 1.04 1.06 1.08 1.1 1.12 1.14 1.16 1.18 1.2 1.22 1.24 020040060080010001200 -90 -80 -70 -60 -50 -40 -30 -20 -10 0Temperature ( C)log [pO2(atm)] ENernst(V) Figure 2-2. Theoretical values of oxygen partial pressure on the a node (calculated from Eq. 2-8) and cell potential (calculated from Eq. 2-4) as a function of temperature. Fuel is hydrogen bubbled through wate r at room temperatures.

PAGE 46

46 Cathode Electrolyte Anode Interconnect Cathode Electrolyte Anode Interconnect Figure 2-3. Schematic represen tation of an SOFC stack. Cell Potential (V)Cell Current Density (A/cm2) 0.2 0.4 0.6 0.8 1.0 0.5 1.0Open Circuit Potential NernstPotential (ti=1 & gas-tight sealing) theoretical max efficiency Activation Polarization Ohmic Polarization Concentration Polarization Cell overpotential Cell output potential (useful voltage) Load current density Cell Potential (V)Cell Current Density (A/cm2) 0.2 0.4 0.6 0.8 1.0 0.5 1.0Open Circuit Potential NernstPotential (ti=1 & gas-tight sealing) theoretical max efficiency Activation Polarization Ohmic Polarization Concentration Polarization Activation Polarization Ohmic Polarization Concentration Polarization Cell overpotential Cell output potential (useful voltage) Load current density Cell overpotential Cell output potential (useful voltage) Load current density Figure 2-4. Illustrative representation of typical current-voltage profile for an SOFC illustrating different types of electrochemical losses th at arise when current is drawn from the cell.

PAGE 47

47 Figure 2-5. Cubic fluorite (AO2) crystal structure with A-site cations occupying the cube corners and face centers, and oxygen anions occupyi ng the tetrahedral interstitial sites. Ce0.8Gd0.2O1.9 Zr2Gd2O7 BaCe0.9Ce0.1(H+) (ZrO2)0.9(Y2O3)0.1 (Bi2O2)0.75(Y2O3)0.25 L/ = 0.15 cm2 SELF SUPPORTED ELECTROLYTES FILM SUPPORTED ELECTROLYTES 1500 m 150 m 5 m .5 m .15 m 1000/T (K-1) T ( C)900800 700600 500400 300 0.5 0 -0.5 -1 -1.5 -2 -2.5 -3 -3.5 -4 -4.5 0.8 0.911.1 1.21.3 1.4 1.51.61.71.8log (S cm-1) (La2O3)0.95(SrO)0.05 -Bi2O3 Ce0.8Gd0.2O1.9 Zr2Gd2O7 BaCe0.9Ce0.1(H+) (ZrO2)0.9(Y2O3)0.1 (Bi2O2)0.75(Y2O3)0.25 L/ = 0.15 cm2 SELF SUPPORTED ELECTROLYTES FILM SUPPORTED ELECTROLYTES 1500 m 150 m 5 m .5 m .15 m 1000/T (K-1) T ( C)900800 700600 500400 300 0.5 0 -0.5 -1 -1.5 -2 -2.5 -3 -3.5 -4 -4.5 0.8 0.911.1 1.21.3 1.4 1.51.61.71.8log (S cm-1) (La2O3)0.95(SrO)0.05 -Bi2O3 Figure 2-6. Reported ionic conducti vity values for common SOFC electrolyte materials. [Figure adapted from Solid State Ionics Vol. 75, B.C.H. Steele, In terfacial reactions associated with ceramic transport memb ranes, 157-165 (1995) with permission from Elsevier.]

PAGE 48

48 Figure 2-7. Calculated cell performance and polarization resistances. [Reprinted from Journal of Power Jources, Vol. 93, S.H. Chan, K.A. Khor, Z.T. Xia, A complete polarization model of a solid oxide fuel cell and its sensitivity to th e change of cell component thickness, 130-140 (2001) with permission from Elsevier.] electrolyte 3PB line e-conducting particle electrolyte 3PB line e-conducting particle Figure 2-8. Schematic representation of three phase boundary lines between electronic conducting particles, the ion conducting electrolyte and the gas phase for single-phase electronic conducting cathodes.

PAGE 49

49 ion conducting particle 3PB line electrolyte e-conducting particle ion conducting particle 3PB line electrolyte e-conducting particle Figure 2-9. Schematic representation of three phase boundary lines between electronic conducting particles, ion conducting electrolyte as well as ion conducting particles, and the gas phase for two-phase composite cathodes. A B Figure 2-10. ABO3 cubic perovskite crystal structure. (a ) A-site cations at cube center, B-site cations at corners, and oxygen anions at face centers. (b) Black spheresA-site cations, grey tetrahedraoxygen anions at corners, B-site cations at center.

PAGE 50

50 Figure 2-11. Schematic representation of va rious reaction pathways for the oxygen reduction reaction in a MIEC. Figure 2-12. Effect of cathode thickness on eff ctR. [Reprinted from Solid State Ionics, Vol. 131, A.V. Virkar, J. Chen, C. W. Tanner, J. W. Kim, The role of electrode microstructure on activation and concentrati on polarizations in solid oxi de fuel cells, 189-198 (2000) with permission from Elsevier.]

PAGE 51

51 Figure 2-13. Effect of cathode thic kness in I-V performance. [Rep rinted from Solid State Ionics, Vol. 131, A.V. Virkar, J. Chen, C. W. Ta nner, J. W. Kim, The role of electrode microstructure on activation a nd concentration polarizations in solid oxide fuel cells, 189-198 (2000) with permission from Elsevier.] Figure 2-14. Agreement between model (Eq. 215) and experiment for LSM-YSZ composite cathodes. [Reprinted from Solid State Ionics, Vol. 131, A.V. Virkar, J. Chen, C. W. Tanner, J. W. Kim, The role of elec trode microstructure on activation and concentration polarizations in solid oxid e fuel cells, 189-198 (2000) with permission from Elsevier.]

PAGE 52

52 Figure 2-15. Effect of cathode thickness on polarization resi stance for LSM-YSZ composite cathodes. [Reprinted from Solid State Io nics, Vol. 86-88, M. Mogensen and S. Skaarup, Kinetic and geometric aspects of solid oxide fuel cell electrodes, 1151-1160 (1996) with permission from Elsevier.] Figure 2-16. Effect of electrolyte phase co mposition on polarization resistance of various conventional SOFC cathode materials. [Repri nted from Solid State Ionics, Vol. 143, E. Perry Murray, S.A. Barnett, (La,Sr)MnO3-(Ce,Gd)O2-x composite cathodes for solid oxide fuel cells, 265-273 (2001) with permission from Elsevier.]

PAGE 53

53 A B Figure 2-17. Effect composition on (a) cathode overpotential, and (b) interfacial resistance for Ag-YSB composite cathodes. [Reprinted from Journal of the American Ceramic Society, Vol. 81, Z. Wu, M. Liu, Ag-Bi1.5Y0.5O3 composite cathode materials for BaCe0.8Gd0.2O3-based solid oxide fuel cells, 1215-1220 (1998) with permission from Blackwell Publishing.]

PAGE 54

54 Figure 2-18. Theoretical ambipol ar conductivity from EMPT. [R eprinted from Solid State Ionics, Vol. 93, Z. Wu, M. Liu, Modelli ng of ambipolar transport properties of composite mixed ionic-electronic conducto rs, 65-84 (1996) with permission from Elsevier.] Figure 2-19. Effect of compositi on on interfacial resistance of GDC-LSM composite cathodes. [Reprinted from Solid State Ionics, Vol. 143, E. Perry Murray, S.A. Barnett, (La,Sr)MnO3-(Ce,Gd)O2-x composite cathodes for solid oxide fuel cells, 265-273 (2001) with permission from Elsevier.]

PAGE 55

55 A B Figure 2-20. (a) Arrhenius conducti vity plot of selected Fe loadings in LSCF, and (b) cathode polarization performance comparison. [Repri nted from Solid State Ionics, Vol. 127, E. Maguire, B. Gharbage, F.M.B. Marques, J.A. Labrincha, Cathode materials for intermediate temperature SOFCs, 329-335 (2000) with permission from Elsevier.]

PAGE 56

56 Depth (x10-5cm) 18O/(16O + 18O) La0.65Sr0.35MnO3-xD* = 3x10-13cm2s-1k = 9x10-8cm-1 Depth (x10-5cm) 18O/(16O + 18O) La0.65Sr0.35MnO3-xD* = 3x10-13cm2s-1k = 9x10-8cm-1A 18O/(16O + 18O) 0.00 0.01 0.02 0.03 0.04 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7Depth (cm) D* = 1x10-7cm2s-1k = 3x10-5cm s-1La0.6Ca0.4Co0.8Fe0.2O3-x T = 800 C t = 400 s 18O/(16O + 18O) 0.00 0.01 0.02 0.03 0.04 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7Depth (cm) 18O/(16O + 18O) 0.00 0.01 0.02 0.03 0.04 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7Depth (cm) D* = 1x10-7cm2s-1k = 3x10-5cm s-1La0.6Ca0.4Co0.8Fe0.2O3-x T = 800 C t = 400 sB Figure 2-21. IEDP for (a) LSM and (b) LCCF. [R eprinted from Journal of Power Sources, Vol. 49, B.C.H. Steele, Oxygen transport and exchange in oxide ceramics, 1-14 (1994) with permission from Elsevier.]

PAGE 57

57 1000/T (K-1) 1.0 1.5 2.0 2.5 3.0 0.70.80.91.001.101.201.30log ( / Scm-1) 1000/T (K-1) 1.0 1.5 2.0 2.5 3.0 0.70.80.91.001.101.201.30 0.70.80.91.001.101.201.30log ( / Scm-1)A Current densit y ( Acm-2 ) Cathodicoverpotential(V)0.0 0.1 0.2 0.3 0.4 0.5 0.00.51.5 1.0 2.0 Current densit y ( Acm-2 ) Cathodicoverpotential(V)0.0 0.1 0.2 0.3 0.4 0.5 0.00.51.5 1.0 2.0B Maximum power density (mWcm-2)Temperature (K)400 300 200 100 0 9001000110012001300 Maximum power density (mWcm-2)Temperature (K)400 300 200 100 0 9001000110012001300C Figure 2-22. Effect of A-site dopant on (a) elec trical conductivity, (b) ca thodic polarization, and (c) fuel cell power density for various Ln0.6Sr0.4MnO3 (Ln=La,Pr,Nd,Sm,Gd,Yb, and Y) perovskite cathodes on YSZ electrolyte s. [Reprinted fr om Journal of the Electrochemical Society, Vol. 142, T. Is hihara, T. Kudo, H. Matsuda, Y. Takita, Doped PrMnO3 perovskite oxide as a new cathode of solid oxide fuel cells for low temperature operation, 1519-1524 (1995) with permission from ECS.]

PAGE 58

58 A B C Figure 2-23. Overpotential-current density curves for (a) diffusion, (b) charge transfer, and (c) mixed control. [Reprinted from Solid Stat e Ionics Vols. 86-88, M. Gdickemeier, K. Sasaki, L. J. Gauckler and I. Riess, Per ovskite cathodes for solid oxide fuel cells based on ceria electrolytes, 691-701 (1996) with permission from Elsevier.] L L/2 L L/2 Figure 2-24. Effect of electron-conducting particle size on 3PB length.

PAGE 59

59 2 1 0 -1 -2 0.800.85 0.90 0.951.00 1.051.101.151.20T-1/k (K-1)log ( E/Scm-1) 2 1 0 -1 -2 0.800.85 0.90 0.951.00 1.051.101.151.20T-1/k (K-1)log ( E/Scm-1)A 1.2 0.8 0.4 0.0 -0.4 0.90 0.95 1.00 1.05 1.101.151.20T-1/k (K-1)log ( E/Scm-1)1.25 -0.8 -1.2 1.2 0.8 0.4 0.0 -0.4 0.90 0.95 1.00 1.05 1.101.151.20T-1/k (K-1)log ( E/Scm-1)1.25 -0.8 -1.2B 1.2 0.8 0.4 0.0 -0.4 0.90 0.95 1.00 1.05 1.101.151.20T-1/k (K-1)log ( E/Scm-1)1.25 -0.8 -1.2 1.2 0.8 0.4 0.0 -0.4 0.90 0.95 1.00 1.05 1.101.151.20T-1/k (K-1)log ( E/Scm-1)1.25 -0.8 -1.2C Figure 2-25. Arrhenius behavior of interface conductivity for LSC cathodes (a) varying Sr content at E=0V, and varying voltage for (b) La0.8Sr0.2CoO3and (c) La0.6Sr0.4CoO3[Reprinted from Electrochimica Acta, Vol. 46, T. Horita, K. Yamaji, N. Sakai, H. Yokokawa, A. Weber and E. Ivers-Tiffe, Oxygen reduction mechanism at porous La1-xSrxCoO3-d cathodes/La0.8Sr0.2Ga0.8Mg0.2O2.8 electrolyte interface for solid oxide fuel cells, 1837-1845 (2001) with permission from Elsevier.]

PAGE 60

60 CHAPTER 3 STABILITY OF SILVER-BISMUTH OXIDE CATHODES 3.1. Introduction The primary function of SOFC cathodes is to aid in the oxygen reduction reaction, Eq. 2-1. It is evident from this equa tion that oxygen reduction must occur where the gas phase, the cathode, and the electrolyte are in intimate contact. These boundaries form one-dimensional lines called three phase boundary (3PB) lines. As shown in Fig. 2-8, the reaction zone is restricted to a narrow region at the electrolyte /cathode interface. The reaction zone can be extended to a three-dimensional volume above this interface by introducing a second, ionconducting phase, Fig. 2-9. The performance of these dual phase cat hodes is significantly improved due to the increase in available sites for oxygen reduction. A well-behaved cathode must be a good catalyst for oxygen reduction, highly conductive, and chemically and thermally compatible with the electrolyte. In addition, for mobile applications, the cathode must be stable at operating temper ature for over 40000 hours.52 As described in Section 2.4 (Eq 2.15), in creasing the ionic conductivity of the ionconducting phase should improve cathode performance, since higher ionic conductivities allow ions produced deeper within the electrode to reach the electrolyte thus effectively increasing the width of the reaction zone. Due to the high conc entration of oxygen defects present in its crystal structure, stabilized bismuth oxide exhibits excellent ionic conductivity, ranking among the highest of all fast-ion conductors at 600 C.22 Thus, dual phase composite cathodes utilizing some metal, Me, as the electronically conducting phase and a bismuth oxide-based material as the ionically conducting phase are expected the exhibit low resistance compared to similar dual phase cathodes containing the same metal Me a nd some other ceramic phase with lower ionic

PAGE 61

61 conductivity. It should be noted that below 600 C, stabilized bismuth oxide undergoes defect ordering, where the conductivity rapidly decreases with time.53 The exact transition temperature has not yet been fully resolved. Silver is known to be an excellent cataly st for the oxygen reduc tion reaction, has high electronic conductivity, and is relatively inexpensive. Com posites of silver and YSB exhibit some of the lowest area-specifi c resistance (ASR) values known to date for YSZ electrolyte substrates, over the temperature range of interest.39 However, the long-term stability of these materials remains unchecked. In this study, the long-term stability of Ag-Y SB and Ag-ESB composites were examined. Ag-YSB composites were chosen to directly stud y the long-term stability of the system reported in the literature.39 Ag-ESB composites were studied for verification purposes and to improve performance (due to the higher ionic conductivity of ESB compared to YSB).54 YSB (for AgYSB composites) and ESB (for Ag-ESB composites) were chosen as the electrolyte substrates, since these should be chemically and thermally compatible with thei r respective composite cathodes, and to help identify possible orderi ng phenomena or phase transformation of the stabilized bismuth oxide phase. For compara tive purposes, the composition used in this study was based on that used by Xia et. al.39 The isothermal annealing temperature was chosen to be 650 C to avoid defect ordering in the bismuth oxide phase. It should be noted that at this temperature, YSB undergoes a cubic to rhombohedral phase transformation,55,56 but its onset is somewhat sluggish and can be monitored by impedance measurements. 3.2. Experimental YSB (Bi0.75Y0.25O1.5) and ESB (Bi0.8Er0.2O1.5) powders were prepared by conventional solid-state synthesis. Stoichiometric amounts of bismuth oxide (Alfa Stock #10658) and yttrium oxide (Alfa Stock #11181) or er bium oxide (Alfa Stock #11309) were ball-milled in ethanol for

PAGE 62

62 24 h. The solutions were then dried at 90 C for 4 h, calcined at 800 C for 15 h, crushed and sieved (325 mesh size). These powders were unia xially pressed at ~70 MPa using a cylindrical die, isostatically pressed at 250 MPa, and sintered at 890 C for 15 h to form electrolyte support pellets. All pellets measured ~1.1 cm in diameter. ESB and YSB pellets measured ~3.1 mm and ~3.6 mm in thickness, respectively. The relative porosities of the ESB and YSB pellets were 90% and 85% of theoretical, respectively, as measured using ImageJ software (1.37v, Wayne Rasband). Composite Ag-bismuth oxide cathode inks were prepared by adding Ag (Alfa Stock #41598) to the sieved oxide powders (volume rati o of Ag:bismuth oxide = 60:40). Ethanol and organic vehicles (a-terpineol as the solvent, di-n-butyl phthalate as the plasticizer, and polyvinylbuteral as the binder) were added (generally in a 3:1: 2 volume ratio) until a viscosity appropriate for brushing was ach ieved. A similar recipe was followed for preparing the pure silver electrode ink. Symmetric cells were prep ared by paint brushing the cathode slurry to each side of the bismuth oxide pellets, drying at 120 C for 1 h, and firing at 750 C for 1 h. Electrochemical impedance analysis was performed using a Solartron 1260 Impedance/Gain-Phase Analyzer. Silver mesh cu rrent collectors and leads were pressed against the samples in a quartz reactor by the use of a ceramic screw-and-bolt assembly. Cell temperature was maintained at 650 C for 100 h and monitored by a thermocouple flanking the sample. Cell response was measured over a frequency range of 32 MHz to 0.1 Hz, with an AC voltage amplitude of 100 mV. Cell responses were not corrected for porosity effects. Electron probe microanalysis (EPMA, JEOL Su perprobe 733) was performed on a sample with pure silver electrodes to determine if any diffusion of silver into the electrolyte had occurred during testing. A sample was annealed at 750 C for 48 h, encased in epoxy resin and

PAGE 63

63 set for 24 h. The sample was then polished down to 1 m, and a linescan was performed across the electrolyte/cathode interfaceth e atomic percentage of each element present recorded at each position in the scan. Cathode morphology was characterized by th e use of a JEOL m odel JSM-6335F fieldemission scanning electron microscopy (FESEM). Samples which had undergone long-term stability testing were compared against virgin samples to detect any visible microstructural changes that might have occurred during testing. X-ray diffraction (XRD Philips model APD 3720) using Cu K1 ( = 1.54056 ) was employed to determine inter-phase reactivity betw een silver and ESB. The diffraction patterns of mixed raw powders were compared agains t mixed powders which were annealed at 750 C for 48 h. 3.3 Results and Discussion 3.3.1. Compositional Optimization The first goal of this study wa s to produce silver-bismuth oxide cathodes with ASR values comparable to those reported by Xia et al.39 Figure 3-1 shows the impedance spectra and corresponding trend in cathode ASR over a range of composition for the Ag-ESB system at 600 C. A minimum value observed at 50 vol% ESB. This result is reas onable since, for a two-phase composite whose particle sizes are roughly the same, and assuming the solid phases and pores are randomly distributed and the porosity is ope n and sufficiently large, the 3PB length should reach a maximum if the two solid phases are pres ent in equal fractions of the overall electrode volume. As discussed previously, 3PB length is important in two-pha se composite cathode systems since the oxygen reduction re action that is occurring can onl y proceed at s ites where all three reactants are presentas the cathode 3PB length increases, so do the number of reaction sites, and hence cathode activa tion polarization decreases.

PAGE 64

64 Electrode area specific resist ance (ASR) was calculated from 2AR ASRp (3-1) where Rp is the electrode polarization discussed above, A is the el ectrode area, and the factor is used to account for the fact that each cell contains two (appr oximately identical) electrodes. Note that electrode area does not account for electrolyte surf ace variations due to surface porosity. Figure 3-2 shows the comparison of the ASR obtained in this study for Ag-ESB cathodes (on an ESB electrolyte) with the results adapted from Xia et al.39 for Ag-YSB (on YSZ honeycomb electrolytes) and othe r reported cathodes. The Ag-E SB composite cathode has an ASR of 0.18 cm2 at 600 C, making it one of the lowest -resistance electrode systems reported to date and a significant improvement over that produced by Xia et al.39 who obtained an ASR of 0.3 cm2 at 600 C for their Ag-YSB composite. Note that two samples of the Ag-ESB optimized composition were tested to confirm re producibility. The second sample had a slightly higher value of ASR (0.20 cm2 at 600 C), but was still signif icantly lower than the Ag-YSZ literature value. 3.3.2. Stability: Thermal Performance Dec ay due to Microstructural Evolution Having met and exceeded the benchmark for cathode performance obtained in the literature, the long-term stabilit y of these composites could now be assessed. Impedance spectra for the long-term testing of pure Ag Ag-ESB and Ag-YSB systems at 650 C are shown in Figure 3-3, with the data Rs-corrected for ease in comparison. That is, the high-frequency realaxis intercept (bulk resistance or Rs) of each spectrum has been subtracted from the real component of each data point in the spectrum. Plotted this wa y, the low-frequency real-axis

PAGE 65

65 intercept corresponds to the elec trode polarization resistance, Rp. It is clear from this figure that the electrode polarization for each system increases with time. Electrode ASR (calculated using Eq. 3-1) is plotted against time for all systems at 650 C in Figure 3-4. The initial zero time value of ASR for the Ag-YSB and Ag-ESB systems were similar.04 cm2 and 0.06 cm2, respectively. Surprisingly, the Ag-YSB electrode system exhibited lower resistance than the Ag-ESB system, despite the higher ionic conduc tivity of the ESB phase. This could be due to surface porosity effects discussed above, but other microstructural factors will be addressed below. Also, the resistance of these electrodes is significantly lower than that reported by Xia for their Ag-YSB system at the same temperature (greater than 0.1 cm2), again, making it one of the lowest-resistance electrode systems reported to date. Differences in processing routines c ould lead to variations in electrode porosity, thickness, relative particle sizes, and inter-p article necking, accounting for the observed differences in performance between the two studi es and between the two electrodes Ag-YSB and Ag-ESB. In addition, since the present electrode study was performed on electrolytes having the same composition as the ionic c onducting phase in the electrodes, the interfacial (electrodeelectrolyte) resistance may be lower than similar electrodes placed on YSZ electrolytes. All systems experience significant increases in ASR during the 100 h stability experimentboth the Ag-YSB and Ag-ESB electrode ASR values increased by around 70% (from 0.04 cm2 to 0.07 cm2 (75%) and from 0.06 cm2 to 0.10 cm2 (67%) for Ag-YSB and Ag-ESB respectively), while the pure Ag system experiences a near fourfold increase (from 0.92 cm2 to 3.55 cm2) during the same length of time. Linear regression of the data (based on the first 10 h of testing) yields a degradation rate of 4.4x10-4 cm2/h for Ag-ESB and 1.2x10-4 cm2/h for Ag-YSB. The pure Ag electrode degrad es at a rate more than two orders of

PAGE 66

66 magnitude faster.7x10-2 cm2/h. It should also be mentioned that the rate of change in ASR with time is linear for Ag-ESB electrodes, while for the Ag-YSB system deviates slightly from linearity beginning approximately 50 h after commen cement of testing. For the pure Ag system, this rate of change is strongly non-linear a nd the non-linear trend is in the opposite direction compared to the Ag-YSB system. Various possi ble causes for these observations are addressed below. Figure 3-5 plots the relati ve conductivity (from Rs) versus time for both bismuth oxide electrolyte systems. Relative conductivity is calculated as the rati o of the electrolyte conductivity (determined from impedance spectra hi gh-frequency intercepts) at time, t, to the initial conductivity value at t=0. Both electrolytes seem to be relatively stableboth maintaining over 90% of their initial conductivitieshowever, YSB experiences significant decay in conductivity starting 50 h into testi ng, while ESB underwent a slight increase in conductivity (+3.0x10-5 Scm-1/h). The trend for YSB is consis tent with results reported by Fung and Virkar56 where YSB undergoes a cubic-rhombohedral phase transformation at this temperature. Post-mortem XRD studies confirm formation of a rhombohedral phase. Jiang and Wachsman15 showed that ESB undergoes no such phase transformation, which is consistent with the trend for ESB in Figure 3-5. Since the resistance of bulk -related phenomena of both sy stems remained relatively constant during the first half of the stability test it may be concluded th at little if any defect ordering has occurred in the YSB and ESB electrolytes up to 50 h. It follows that minimal defect ordering has occurred in the bismuth oxide phase of each cathode over this period. Further, during this initial 50 h regime, the trend in elec trolyte conductivity is s lightly upward for ESB, and slightly downward for YSB. This trend is opposite to that observed for cathode ASR values,

PAGE 67

67 where the cathode containing YSB decayed at a slow er rate than that containing ESB, suggesting the impact of the bismuth oxide phase on cathode degradation is minor at this temperature. In addition, the ASR of the pure Ag electrode rises at a much faster rate than either composite system. Thus, it is suspected that the silver phase is the major source of the observed increase in electrode polarization. However, the more dram atic decay in YSB conductivity at longer times may cause the observed non-linear increase in ASR with time for the Ag-YSB system (Fig. 3-4), as the same phenomenon observed for the electrol yte conductivity may be occurring in the YSB phase inside the composite electrode. EPMA, XRD, and FESEM analysis were perf ormed on several samples to elucidate the compositional and microstructural changes that might transpire during testing. For EPMA measurements, Y2O3 and Bi2O3 standards were prepared by pre ssing and sintering pellets of the commercial powders. Figure 3-6 shows the result s of EPMA testing on a YSB sample prepared with pure silver electrodes and annealed at 750 C for 48 h. The atomic percent of all elements is consistent with what is expected on both sides of the interface. Thus, it is concluded that no silver has diffused across the interface into the electrolyte. X-ray diffraction pattern s of mixed silver and erbium-s tabilized bismuth oxide powders (ESB) before and after heat treatment at 750 C for 48 h are shown in Figure 3-7. The patterns reveal no evidence of inter-phase reactivity after heat treatment. Since silver neither diffused into nor reacted with th e electrolyte phase at 750 C, it is reasonable to assume the same is true for all samples tested at 650 C. Microstructural comparisons of the three diffe rent electrode systems before firing, after firing, and after testing at 650 C for 100 h are shown in Figures 3-8 through 3-10. Energy dispersive spectroscopy (EDS) confirmed that fo r backscattered SEM images, the lighter phase

PAGE 68

68 was ESB and the darker phase was Ag. All three sy stems exhibit a dramatic growth of the silver phase after firing at 750 C for 1 h. However, the silver pha se in the composite electrodes show significantly less growth and coalescence compared to the pure Ag system, as the ceramic phase helps restrict the migration and ultimate agglomer ation of the silver phase. After 100 h of testing at 650 C, the pure Ag electrode appears fully dens e. The porosity of the Ag-ESB electrode decreased significantly after testi ng. Domains of coalesced silver are evident, and ESB particles between these domains have been forced closer together, further reduci ng porosity and 3PBs. Growth and coalescence of the silver phase is al so evident in the Ag-Y SB electrode though to a lesser extent. It is believed that the smaller st arting size of the YSB particles (as evidenced in the unfired electrode micrographs) helped to lower the mobility of the silver phase by providing the electrode with a higher surface area through which the silver phase must migrate. Although the primary function of the silver phase to conduct electrons toward the 3PBs is unaffected by these microstructural changes, in creased grain size correla tes with reduced 3PB lengths, and hence a smaller reaction zone for oxyge n reduction. Further, increased density can lead to concentration polarizati on effects, where the reaction sp ecies (oxygen admolecules) are being supplied to the reaction sites more slowly than they are being consumed. The trends in electrode ASR shown in Figure 3-4 can be ratio nalized in terms of these microstructural observations. Pure Ag undergoes the most dramatic densifica tion during testing, and also exhibits the most dramatic rise in ASR with tim e. It is reasonable to expect that as the densification process of Ag near s completion, changes in porosity and 3PB lines will slow, and hence the electrode ASR will le vel off at longer times, as was observed. The composite electrodes experience a much less se vere change in microstructure after firing and during testing, particularly the Ag-YSB system. This is reflecte d in the two-order slower rate of ASR increase

PAGE 69

69 for these composite systems, and the slightly sl ower rate for Ag-YSB compared with Ag-ESB. The finer silver phase microstructure of the fi red, untested Ag-YSB electrode relative to the untested Ag-ESB electrode may also explain w hy the electrode containing YSB exhibited the lower initial zero time ASR value despite the higher ionic conductivity of ESB. Transgranular platelet structures similar to those observed by Fung55,56 were found in the YSB phase of the electrode and in the electrol yte near the electrode interface (not shown), consistent with the assertion that a cubic-r hombohedral phase transformation occurred during testing. This phase transformation likely led to the observed drop in electrolyte conductivity and rise in electrode ASR at longer times, both of which seem to begin around 50 h. From the above observations, it is concluded th at the main source of instability of the composite Ag-bismuth oxide electrodes is the migration of the silver phase, rather than reactivity or defect ordering of the stab ilized bismuth oxide phase. Mo reover, silver migration was demonstrated to be even more seve re when operating under an applied bias57,58. This issue as well as microstructural approaches toward improvi ng the stability of these electrodes is discussed Chapter 4 of this dissertation. 3.4 Conclusions Silver-bismuth oxide composite cathodes have been prepared which perform as well as (and even better than) sim ilar composites presented in the lite rature. The long-term isothermal stability of pure silver, Ag-ESB, and Ag-YSB electrodes was examined. Each system exhibited significant increases in electrode area specific resistance during 100 h of testing at 650 C under open-circuit conditions. On the basis of electrochemical impedance as well as chemical and microstructural analysis, it is concluded that Ag-bismuth oxide composites have inadequate microstructural stability for long-term, IT-SOFC cathode applications. Microstructural evolution of the silver phase is deemed to be responsible for degradation in perf ormance over time due to

PAGE 70

70 grain growth and electrode densification. Howe ver, since the microstructural evolution of the composite electrodes was not as severe as that of pure silver, and sin ce the Ag-YSB composite, at least initially, was more stable than th e Ag-ESB composite, it is likely that further improvements in stability for the composite system will result from replacing the relatively large bismuth oxide particles with smaller particles, or by infusing inert nano-sized oxide particles into the microstructure. This concept will be explored in Chapter 4 of this dissertation.

PAGE 71

71 -0.5 -0.4 -0.3 -0.2 -0.1 2.72.93.13.33.53.73.9 70% 65% 60% 40% 55% 45% 50%Z( )Z( ) -0.5 -0.4 -0.3 -0.2 -0.1 2.72.93.13.33.53.73.9 70% 65% 60% 40% 55% 45% 50%Z( )Z( ) A 0 0.1 0.2 0.3 0.4 0.5 0.6 304050607080ASR ( cm2)Ag content (vol%) 0 0.1 0.2 0.3 0.4 0.5 0.6 304050607080ASR ( cm2)Ag content (vol%) B Figure 3-1. (a) Impedance spectra and (b) electro de ASR vs silver content of Ag-ESB showing compositional optimization at 600 C, be tween 40-70 vol% Ag using 5 vol% composition steps.

PAGE 72

72 Ag-ESB Ag-ESB Figure 3-2. Arrhenius plot, adapted from Xia et. al.,39 comparing the ASR of Ag-ESB (this work) with the best cathodes reported in the literature at the start of this study. [Adapted from Applied Physics Letters, Vol.82, C. Xia, Y. Zhang and M. Liu, Composite cathode based on yttria stabilized bismuth oxi de for low-temperature solid oxide fuel cells, 901-903 (2003) with permission from American Institute of Physics.]

PAGE 73

73 005 00.10.20.3 -0.1 0 Z( )Z( )a) time 005 00.10.20.3 -0.1 0 Z( )Z( )a) time 005 00.10.20.3 -0.1 0 Z( )Z( )a) time 005 00.10.20.3 -0.1 0 Z( )Z( )a) time A 00.050.100.150.20 Z( ) -0.05 0 timeb) 2 ( ) Z( ) 00.050.100.150.20 00.050.100.150.20 Z( ) -0.05 0 timeb) 2 ( ) Z( ) 00.050.100.150.20 Z( ) -0.05 0 timeb) 2 ( ) Z( ) 00.050.100.150.20 00.050.100.150.20 Z( ) -0.05 0 timeb) 2 ( ) Z( ) BZ( )Z() ( ) 012345 -2 -1 0 c)timeZ( )Z() ( ) 012345 012345 -2 -1 0 c)time Z( )Z() ( ) 012345 -2 -1 0 c)timeZ( )Z() ( ) 012345 012345 -2 -1 0 c)time C Figure 3-3. Impedance spectra obtained from a symmetrical Ag-ESB/ESB/Ag-ESB (a), AgYSB/YSB/Ag-YSB (b), and Ag/ESB/Ag (c) cells tested at 650 C over a period of 100 h. The data are Rs-corrected. Arro ws indicate direction of increasing time.

PAGE 74

74 0 0.1 0.2 0.3 0.4 05 01 0 0 0 1 2 3 4ASR = 8.65x10-2 t + 0.95 ASR = 4.44x10-4 t + 0.062 ASR = 1.24x10-4 t + 0.040Pure Ag Ag-ESB Ag-YSBASR ( cm2)Time (h)ASR ( cm2) 0 0.1 0.2 0.3 0.4 05 01 0 0 0 1 2 3 4ASR = 8.65x10-2 t + 0.95 ASR = 4.44x10-4 t + 0.062 ASR = 1.24x10-4 t + 0.040Pure Ag Ag-ESB Ag-YSBASR ( cm2)Time (h)ASR ( cm2) Figure 3-4. Electrode ASR vs. time fo r Ag-ESB, Ag-YSB, and pure Ag at 650 C. Linear regression based on first 10 h of te sting for each electrode system.

PAGE 75

75 0.9 0.95 1 1.05 05 01 0 0 t/ Time (h) ESB YSB 0.9 0.95 1 1.05 05 01 0 0 t/ Time (h) ESB YSB Figure 3-5. Change in electrolyte relativ e conductivity vs. time for ESB and YSB at 650 C.

PAGE 76

76 0 10 20 30 40 50 60 70 80 90 100 -10-50510 Ag (L) Bi (M) Y (M) O (K)Distance from interface ( m)Atom % 0 10 20 30 40 50 60 70 80 90 100 -10-50510 Ag (L) Bi (M) Y (M) O (K)Distance from interface ( m)Atom % Figure 3-6. EPMA cross-section li nescan of a sample having pure silver electrodes, annealed at 750 C for 48 h.

PAGE 77

77 20304050602 )Intensity (arb.)unfired 750 C, 48 hF F F F F Ag Ag (111) (200) (220) (311) (222) (111) (200) 20304050602 )Intensity (arb.)unfired 750 C, 48 hF F F F F Ag Ag (111) (200) (220) (311) (222) (111) (200) Figure 3-7. XRD spectra of Ag-ESB powder mi xtures before and after co-firing at 750 C for 48 h. F markers identify cubic fluorite peak s of the ESB phase, A g markers identify silver peaks.

PAGE 78

78 ESB Ag ESB Ag ESB Ag ESB AgA ESB Ag ESB Ag ESB Ag ESB Ag B ESB Ag ESB Ag ESB Ag ESB Ag ESB Ag ESB AgC Figure 3-8. SEM micrographs comparing the morphology of Ag-ESB electrodes before (a) and after (b) firing at 750 C for 1 h and (c) after testing at 650 C for 100 h.

PAGE 79

79 YSB Ag YSB AgA YSB Ag YSB Ag B YSB Ag YSB AgC Figure 3-9. SEM micrographs comparing the morphology of Ag-Y SB electrodes before (a) and after (b) firing at 750 C for 1 h and (c) after testing at 650 C for 100 h.

PAGE 80

80 Ag ESB Ag ESBA ESB Ag ESB AgB ESB Ag ESB AgC Figure 3-10. SEM micrographs comparing the morphology of pure Ag electrodes before (a) and after (b) firing at 750 C for 1 h and (c) after testing at 650 C for 100 h. Delamination caused by fracturing for SEM analysis.

PAGE 81

81 CHAPTER 4 IMPROVING THE STABILITY OF SILVER -BI SMUTH OXIDE CATHODES THROUGH MICROSTRUCTURAL CONTROL 4.1 Introduction Recent literature results demons trate that a composite cathode consisting of a silver phase as the electronic conductor and a stabilized bismuth oxide phase (in this case yttrium stabilized bismuth oxide, YSB) as the ionic conductor exhi bits outstanding performance, compared with other frequently studied systems (Fig 3.1).39 In Chapter 3, the stability of this system was assessed and determined to be a problem due to the high mobility of the silver phase.59 It was incidentally shown that composite electrodes wh ich contained an oxide phase in addition to a silver phase exhibited better stability compared to pure silver electrodes. This effect has also been demonstrated in the literature.7 Furthermore, microstructural evolution and cathode ASR degradation was less severe when the oxide phase was composed of finer particles. In addition, it was shown that operation unde r an applied DC bias causes an electromigration type effect on the silver phase, resultin g in a build-up of silver phase at the electrodeelectrolyte interface at one side of a symmetrical cell, and a migra tion of silver to the surface of the other side.57,58 Stable SOFC components should de liver relatively constant, non-changing performance for thousands of hours of operation. The goal of this study was to improve the stability of this electrode system under open-ci rcuit conditions as well as under an applied DC bias through microstructural means. 4.2 Experimental In this work, silver-ESB20 composite electr odes were studied because of the high ionic conductivity of ESB20. Also, based on the results presented in Chapter 3, ESB20 shows no phase transformation or defect or dering at the temperature of interest. Erbium oxide and bismuth

PAGE 82

82 oxide powders were weighed in proper amounts to yield Er0.4Bi1.6O3. The powders were ball milled in ethanol for 24 h using YSZ grinding medi a, dried on a hot plate with stirring, then calcined at 800 C for 10 h. The calcined powder was th en crushed by mortar and pestle and sieved (325 mesh). Green ESB20 pellets were prepared by uniaxial pressing (approximately 70 MPa) followed by isostatic pressing (250 MPa). These green bodies we re then fired at 890 C for 15 h. The sintered pellets had densities 94% % of theoretical, were 1.12 cm .02 cm in diameter and 0.29 cm 0.01 cm in thickness. For the electrode slurry, the crushed and si eved ESB20 powders were either used asprepared or a vibratory mill was used to reduce particle size. For vibratory milling, 10 g of powder plus 300 g of cylindrical zirconia grinding media and 200 ml isopropyl alcohol were placed into 250 ml Nalgene bottles. These bottles were covered with duct tape and placed into a Sweco model M18-5 vibratory mill for seven days using appropriate counter weights. Inks were prepared by combining organic vehicles with mi xtures of metallic (s ilver) and ceramic (ESB20 alone or ESB20 and Tosoh 8YSZ) powders. Once an appropriate viscosity was reached (consistency of honey), the inks were applied to bo th sides of the electroly te substrates by screen printing (AMI model HC-53 screen printer) to give symmetrically-electroded cells. The hole size of the screen printing mask was large enoug h to prevent preferential screening of larger particles. The electrode s had thicknesses of ~30 m and geometric surface areas of ~0.79 cm2. These cells were dried at room temper ature for 1 h, followed by drying at 120 C for 1 h. This process was repeated for a second coat. Th e doubly-coated cells were then fired at 750 C for 1 h. Electrochemical performance of the el ectrodes was assessed using impedance spectroscopy. Silver mesh current collectors an d platinum lead wires were pressed against the

PAGE 83

83 samples in a quartz reactor by the use of a ceram ic screw-and-bolt assembly. The quartz reactor was placed into a temperature-controlled furnac e, and air was fed through the reactor at 50 cc/min using a BOC mass flow controller. A So lartron 1260 frequency response analyzer was used in standalone mode for unbiased testing us ing a frequency range of 32 MHz to 0.01 Hz and an AC voltage amplitude of 100 mV. A Solart ron 1287 interface was added for testing under an applied bias of 250 mV. X-ray diffraction (XRD) patterns of various powders were recorded with a Philips APD 3720 diffractometer using Cu K1 ( = 1.54056 ). Step scans were taken over a range of 2 angles from 20 to 100 with 0.04 steps. Number-average particle size distributions were determined using a Brookhaven ZetaPlus instrument using a light scatteri ng technique dilute suspensions of powder in an ethanol medium. Scanning electron microscopy (SEM) was used to analyze electrode microstructures. Both JEOL JSM 6400 scanning electron microscope and JEOL JSM 6335F field emission scanning electron microscopes were used in this st udy for generating seconda ry and backscattered microstructural images. 4.3. Results and Discussion 4.3.1 Nano YSZ Additions The addition of bismuth oxide grains acts to inhibit silver phase grain growth in silverYSB composites.7 Further, a common method used to a void grain growth in commercial alloys is to add a dispersion of fine particles, which restricts grain size according to f r RL6 (4-1) where RL is the limiting grain radius, r is the partic le radius, and f is the volume fraction of particlesthus grain size is re duced by decreasing the particle size radius, or increasing the

PAGE 84

84 volume fraction of particles (both of which in crease the surface area of the particles).60 It is proposed that increasing the surface area of the ceramic phase in silver-containing cermets will further inhibit silver phase mobility. The first a ttempt to improve the stability of this electrode system involved the use of nano-sized ceramic pa rticles. Nano-sized 8 mol% YSZ powder was used in this study as it is readily available, relatively inexpensive (~$100 per kilogram, or 10% the cost of GDC nano-sized particles), and non-reactive toward silver and ESB20 at the temperature range of interest (Fig. 4-2). Various YSZ/ESB co mpositions were tested, keeping the silver volume fraction constant. That is, cathodes with composition Agx-(YSZy-ESB1-y)1-x were prepared, where x = 0.5, and y = 0, 0.05, 0.10, and 0.15 represent fractions of the total cathode solids volume. Impedance spectra and electrode ASR (calcu lated from Eq. 3-1) versus time under no applied bias are shown in Figures 4-2 and 4-3, respectively. Two arcs are present in the spectra for all electrodes suggesting the impedance is governed by two competing processes, however it is not known at this time which electrochemical processes correspond to each arc. The arcs are more distinct for electrodes c ontaining nano-sized YSZ, and it is clear that all samples containing the YSZ powder additions were significantly more stable compared to the YSZ-free sample. The electrode ASR degraded at a rate of 4.1x10-4 cm2/h, 1.3x10-5 cm2/h, 2.5x10-5 cm2/h and 3.1x10-5 cm2/h for 0%, 5%, 10%, and 15% YSZ, respectively. The largest stability improvement (a 97% reduction in ASR vs. time sl ope) was achieved at the lowest YSZ loading, 5%. In addition, the initial zero-time ASR value of this composition was 31% less than that of the YSZ-free composition (0.043 cm2 and 0.062 cm2 for 5% and 0% YSZ, respectively), despite the substitution of 5 vol% YSZ for the high conductivity ESB20 phase. Samples with higher concentrations of YSZ exhibited highe r initial ASR values, as expected.057 cm2 and

PAGE 85

85 0.077 cm2 for 10% and 15%, respectivelybut showed electrode stability improvement over the non-YSZ composite. However, the ASR vs. tim e slopes increased as YSZ content increased. A backscatter electron microscopy comparison between tested sample s (Fig. 4-4) shows how the silver particles have been restrained fr om coalescence, allowing porosity to remain open and three phase boundaries to remain high, comp ared to the case with no YSZ additions. Further, the size of the silver grains decreases with increased YSZ content. Figure 4-5 is a secondary electron image compari ng the 15% YSZ sample before and after testing. Grain size and porosity appear to be comparable before and af ter testing. However, the edges of the silver particles in the as-fired sample are smooth, while in the tested sample, the edges are rough, due to envelopment of YSZ into the silver particles. It is possible that this effect further reduces the 3PB length between metallic, gaseous, and ESB20 phases, contribu ting to the observed increase in ASR with time for these electrodes. This c ould also explain the observed increased ASR vs. time slopes with increasing YSZ content. 4.3.2. ESB20 Particle Size Reduction The second attempt to improve electrode stability involved reduction of initial particle size of the ESB20 powder. This strategy is simila r to the addition of na no-YSZ particlesmore energy is required for th e silver phase to migrate over a given distance in the electrode. However, the overall ionic conductiv ity of the cathode as well as the reactive 3PB lines should not be compromised, unlike YSZ additions. This is also a more cost-effective approach than strategies involving alloying silver with precious metals to reduce Ag mobility, as suggested by Jaiswal et. al.57 Figure 4-6 shows the reduction in particle size that was obtaine d from the use of vibratory milling. Number average particle size was reduced from ~1 m to ~300 nm. Note that the size

PAGE 86

86 distributions may be skewed to lo wer particle sizes since larger particles settle out of suspension quickly, and consequently may be omitted from the measurement count. All ESB powders were ground with YSZ ball mediaboth ball milled and vi bratory milled powders. In addition, the hardness of ESB20 is much lower than that of YS Z. Further, the mass of the ball media used before and after ball milling and vibratory milling was nearly unchanged (< 0.01 wt%). Hence YSZ contamination from milling should be minimal and is expected to have little influence on the results. The long term, unbiased impedance study co mparing composites prepared with (ESBVM) and without (ESBS) vibratory milling of the ESB20 phase, shown in Figures 4-7 and 4-8, reveal the vast improvement in stability the reduction in ESB20 particle size imparts on this system. As in Chapter 3, linear trends in ASR with time are observed, and the degradation rate of silverESB20 composites at 650 C was reduced from 4.10x10-4 cm2/h to 1.91x10-5 cm2/h a 95% drop with the reduced ESB20 particle size. As can be seen in Figure 4-8, the initial, zero time ASR value was also lowered by 25%, from an already-low 0.062 cm2 for the Ag-ESBS composite to 0.048 cm2 for the Ag-ESBVM compositethis is most likely due enhanced 3PB a nd may be attributed to suppression of silver migration during electrode sintering, l eaving both porosity and 3PB length high. This experiment was repeated with an extern al 250 mV DC bias app lied across the cells to simulate operating conditions, and the results are shown in Figures 4-9 and 4-10. The improvement in performance and stability is ev ident, though not as prono unced as the unbiased case. After about 15 hours of test ing, the ASR exhibits a linear in crease with time. The ASR vs. time slope under 250 mV bias at 650 C is reduced by 50% (from 1.6x10-3 cm2/h to 8.0x10-4 cm2/h) when smaller ESB20 particles are used to prepare the composite electrodes.

PAGE 87

87 Both macroscopic and microscopic changes in the electrodes prepared with larger ESB20 particles before and after bias testing are immediately recognized (Fig. 4-11). Optically, after testing, the electrode surface of th e working electrode was silver in color. The counter electrode of the tested cell appeared red in color, and a ring of silver color could be seen along its edge. The microstructural changes are also dramatic. A cross-sectional view of the counter electrode shows silver dendrite-like structur es at the electrolyt e interface, and nearly pure ESB20 at the electrode surface. The silver phase clearly migrates in one directiontowards the electrode/electrolyte in terface in the counter electrode and towards the surface of the working electrode. Oxygen is reporte d to have high solubility and mobility in silver.51,52 It is possible that the application of a bias ac ross the cell leads to an electro-m igration effect where the silver phase is dragged along in the direction of oxygen flux, as observed by Jaiswal et. al.57 In contrast, the microstructure of the c ounter and working elec trodes prepared with vibratory milled ESB20 particles after bias testing are quite comparable and no segregation of silver phase at the counter electr ode/electrolyte interface is detect ed (Fig. 4-12), indicating silver phase migration was significantl y suppressed by the smaller partic le size. So, not only did the small ESB20 particles reduce initial ASR, and enha nce microstructural stability with no bias, but also improved microstrural st ability under bias testing. 4.4. Conclusions and Future Work Addition of 5 vol% 8YSZ nano powders sign ificantly improved unbiased electrode stability by 97 %, and reduced th e initial, zero time ASR value by 31 %. Similar results were obtained when YSZ-free electrodes were prepared from ESB20 pow ders composed of particles hundreds of nanometers in size as opposed to electrodes prepared from ESB20 powders composed of micron-sized particlesthe zero time ASR value was reduced by 25 %, and ASR vs. time slope during unbiased testing of the silver-ESB20 system at 650 C was reduced by 95

PAGE 88

88 %. Finally, ASR vs. time slopes during testi ng under a 250 mV external applied bias were lowered by 50 % using the smaller ESB20 partic les due to suppression silver phase electromigration. The stability of composite silver-E SB20 electrodes under an a pplied bias still needs some improvement. Improvements are likely wi th further reduction in ESB20 particle size down to several tens of nanometers. Also, as th e operating temperature of SOFCs is reduced, the migration of the silver phase will be suppresse d even further. These electrodes perform well even in the 500-550 C range, but defect ordering in the bi smuth oxide phase becomes an issue at these temperatures. Currently research is being done to overcome this issue as well.61

PAGE 89

89 2030405060ESB ESB ESB ESB ESB YSZ YSZ YSZ YSZ Ag Ag(111) (111) (200) (200) (111) (200) (220) (220) (311) (222) (311)2 Intensity (arb.)750 C, 10 h unfired 2030405060ESB ESB ESB ESB ESB YSZ YSZ YSZ YSZ Ag Ag(111) (111) (200) (200) (111) (200) (220) (220) (311) (222) (311)2 Intensity (arb.)750 C, 10 h unfired Figure 4-1. XRD spectra of Ag-ESB-YSZ powde r mixtures before and after co-firing at 750 C for 10 h. A B C D Figure 4-2. Nyquist plots for silver-ESB20 composite electrode s containing 0 vol% (a), 5 vol% (b), 10 vol% (c), and 15 vol% (d) 8YSZ nanopa rticles. Samples were tested at 650 C in air for 100 h under no applied bias. Note that the electrolyt e resistance has been subtracted from all Nyquist plots.

PAGE 90

90 0.03 0.05 0.07 0.09 0.11 0 50 100A S R = 4 1 0 x 1 00 4t + 6 3 0 x 1 00 2ASR = 3.06x10-05t + 7.67x10-02ASR = 2.51x10-05t + 5.65x10-02ASR = 1.25x10-05t + 4.03x10-020% 15% 10% 5%ASR ( cm2)Time (h) 0.03 0.05 0.07 0.09 0.11 0 50 100A S R = 4 1 0 x 1 00 4t + 6 3 0 x 1 00 2ASR = 3.06x10-05t + 7.67x10-02ASR = 2.51x10-05t + 5.65x10-02ASR = 1.25x10-05t + 4.03x10-020% 15% 10% 5%ASR ( cm2)Time (h) Figure 4-3. Effect of time on ASR of silv er-ESB20 composite electr odes containing various volumetric amounts of 8 mol% YSZ nanopart icles. Measurements taken at 650 C in air under no applied bias.

PAGE 91

91 ESB Ag ESB Ag A B C D Figure 4-4. Backscatter electron micr ostructural images of tested [650 C, 100 h in air under no applied bias] silver-ESB20 electrodes contai ning 0 vol% (a), 5 vol % (b), 10 vol% (c), and 15 vol% (d) 8YSZ nanoparticles.

PAGE 92

92 10 m 10 mA 10 m 10 m 10 m 10 m 10 m 10 m 10 m 10 mB Figure 4-5. Secondary electron mi crostructural images of an untes ted (a) and tested (b) silverESB20 electrodes containing 15 vol% 8YSZ nanoparticles. 20 30 40 50 60 70 80 90 100 0 300 600 900 1200 1500 Particle Size (nm)Count (% of highest)VM sieved 20 30 40 50 60 70 80 90 100 0 300 600 900 1200 1500 Particle Size (nm)Count (% of highest) 20 30 40 50 60 70 80 90 100 0 300 600 900 1200 1500 Particle Size (nm)Count (% of highest)VM sieved Figure 4-6. Results of particle size analysis (number average) for the sieved and vibratorymilled ESB20 powders.

PAGE 93

93 A B Figure 4-7. Nyquist plots (a) a nd imaginary part of impedance pl otted as a function of log-scale frequency (b) for 50-50 vol% silver-ESB20 composite electrodes, where the ESB20 phase was prepared from sieved (larger curves) and vibratory milled (smaller curves) powders. Samples were tested at 650 C in air for 100 h under no applied bias. Note that the Nyquist plots have been shifted so that the high frequency intercept with the real axis crosses at 0

PAGE 94

94 0.04 0.06 0.08 0.1 05 01 0 0Ag-ESBsAg-ESBvmA S R = 4 1 0 x 100 4t + 6 30 x 1 00 2ASR = 1.91x10-05t + 4.78x10-02ASR ( cm2)Time (h) 0.04 0.06 0.08 0.1 05 01 0 0Ag-ESBsAg-ESBvmA S R = 4 1 0 x 100 4t + 6 30 x 1 00 2ASR = 1.91x10-05t + 4.78x10-02ASR ( cm2)Time (h) Figure 4-8. Electrode ASR vs. time for 50-50 vol% silver-ESB20 composite electrodes, where the ESB20 phase was prepared from sieved (triangles) and vibratory milled (circles) powders. Samples were tested at 650 C in air for 100 h under no applied bias.

PAGE 95

95 A B Figure 4-9. Nyquist plots (a) a nd imaginary part of impedance pl otted as a function of log-scale frequency (b) for 50-50 vol% silver-ESB20 composite electrodes, where the ESB20 phase was prepared from sieved (larger curves) and vibratory milled (smaller curves) powders. Samples were tested at 650 C in air for 40 h under a 250 mV bias. Note that the electrolyte resistance has been subtracted from all Nyquist plots.

PAGE 96

96 0.06 0.08 0.1 0.12 0.14 02 04 0Ag-ESBsAg-ESBvmA S R = 1 6 x 1 03t + 7 7 x 1 02A S R = 8 0 x 1 04t + 5 5 8 x 1 02ASR ( cm2)Time (h) 0.06 0.08 0.1 0.12 0.14 02 04 0Ag-ESBsAg-ESBvmA S R = 1 6 x 1 03t + 7 7 x 1 02A S R = 8 0 x 1 04t + 5 5 8 x 1 02ASR ( cm2)Time (h) Figure 4-10. Electrode ASR vs time for 50-50 vol% silver-ESB 20 composite electrodes, where the ESB20 phase was prepared from sieved (triangles) and vibratory milled (circles) powders. Samples were tested at 650 C in air for 40 h under a 250 mV bias.

PAGE 97

97 Ag ESB10 m Ag ESB10 mA 20 m Ag ESB Electrolyte 20 m Ag ESB ElectrolyteB 10m ESB Ag 10m ESB Ag 10m ESB Ag C 10 m 10 mAg ESB 10 m 10 mAg ESB D Figure 4-11. Microstructural im ages of silver-ESB20 composite electrodes, where the ESB20 phase was prepared from sieved powderssurface before testing (a) and crosssection of the counter elec trode (b), working electrode surface (c), and counter electrode surface (d) of a cell after 48 h of testing at 650 C under a 250 mV applied bias.

PAGE 98

98 10 m 10 mA 10 m 10 mB 1 mElectrolyteElectrolyte 1 mElectrolyteElectrolyteC 1 m1 m 1 m1 mD Figure 4-12. Cross-sectional microstructural im ages of silver-ESB20 composite electrodes, where the ESB20 phase was prepared from vibratory milled powders, after 48 h of testing at 650 C under a 250 mV applied biasworki ng electrode (a) with close-up view of the working electrode/electrolyte interface (b) and counter electrode (c) with close-up view of the counter elec trode/electrolyte interface (d).

PAGE 99

99 CHAPTER 5 HIGH PERFORMANCE COMPOSITE BI2RU2O7-BI1.6ER0.4O3 CATHODES FOR IT-SOFCS 5.1 Introduction The critical role microstructure plays in the electrochemical performance of composite cathode systems has been demonstrated. In Chapte r 3 the isothermal instability of a composite cathode consisting of Ag as the electronic condu cting phase and ESB20 as the ionic conducting phase was demonstrated.59 The study showed a 70% rise in ASR for Ag-ESB20 from 0.06 cm2 to 0.10 cm2, after 100 h at 650 C. Microstructural evidence revealed a correlation between agglomeration of the constituent phases (arising from migration of the silver phase) and the increased resistance. Chapter 4 showed that this microstructural evolut ion could be inhibited using nano-sized ceramic particle s, resulting in a 95% reduction in the ASR degradation rate, as well as a 25% reduction in the initial ASR.62 From these studies, it is clear that relative particle sizes, agglomeration, and phase segregation can have dramatic impact on 3PB lengths and the surface area of catalytic ac tive sites, and thus electrode resist ance. In addition, the electrode thickness and the addition of current collection layers can in fluence performance by altering inplane electronic conduction, as well as mass transport.13,39 In this chapter, these microstructural considerations will be used in order to improve the performance of a new class of SOFC cathodesruthenate oxide ESB20 composites. Ruthenium oxides are known to be cat alytically active for oxygen reduction.63,64 By selection of a sufficiently la rge A-site dopant cation in (A2Ru2O7), such as Pb or Bi, its band structure is altered in such a way so as to render its behavior metallic, with conductivity increasing as temperature decreases.65 These properties make me tallically-conductive ruthenate oxides good candidates for SO FC composite cathodes.

PAGE 100

100 It was recently demonstrated that composite cathodes consisting of Bi2Ru2O7 (BRO7) as the electronic conducting phase and ESB20 as the ionic conducting pha se on GDC electrolytes exhibit outstanding performance compared with other frequently studied systems.66 The low resistance (3.47 cm2 and 0.08 cm2 at 500 C and 700 C, respectively) of this composite may be attributable in part to the high ionic conductivity of the ESB20 phase in combination with the catalytic activity of the ruthenate phase. Note th at bismuth was chosen as the A-site dopant, as it is the same as the host cation of the ESB20 pha se, and should improve chemical compatibility. Additionally, bismuth ruthenates have been reported to have better stability than lead ruthenates above 600 C.67,68 The goal of this study is to reduce the AS R of the BRO7-ESB20 electrode system on ESB20 electrolytes by microstructural optimization. 5.2. Experimental 5.2.1. Electrolyte and Electrode Preparation A conventional solid-state synthesis route was used to prep are ESB20 and BRO7 powders. Er2O3 (99.99 %, Alfa Aesar) and Bi2O3 (99.999 %, Alfa Aesar) powders were weighed in proper proportions to yield Er0.4Bi1.6O3. The same raw Bi2O3 powder and RuO2 XH2O (99.99 %, Alfa Aesar, where the moles of hydration, X, varies with batch used, normall X 2.8) were weighed in stoichiometric amounts to yield Bi2Ru2O7. The respective powders were ball milled in ethanol for 24 h using YSZ grinding media, dried on a hot pl ate with stirring, then calcined at 800 C for 15 h (for ESB20) or at 900 C for 10 h (for BRO7). The calcined BRO7 powder was leached in HNO3 to remove an impurity sellinit e phase as described by Jaiswall.66 The ESB20 and BRO7 powders were separately crushed by mort ar and pestle and sieved (325 mesh). Green ESB20 pellets for use as electrolyte supports were prepared by uniaxial pressing (approximately 70 MPa) followed by isostatic pr essing (250 MPa). These green bodies were

PAGE 101

101 then fired at 890 C for 15 h. The sintered pell ets had densities 94% % of theoretical, were 1.12 cm .02 cm in diameter and 0.29 cm .01 cm in thickness. For the electrode slurry, the crushed and si eved ESB20 powders were either used asprepared or a vibratory mill was used to reduce particle size. For vibratory milling, 10 g of powder plus 300 g of cylindrical zirconia grinding media and 200 ml isopropyl alcohol were placed into 250 ml Nalgene bottles. These bottles were covered with duct tape and placed into a Sweco model M18-5 vibratory mill for seven days using appropriate counter weights. Sedimentation was performed to further reduce particle size and size distribution. Powders were mixed in a medium (1 g / 50 ml ratio) in Nalgene bottles, ultra-sonicated for 30 min to break up soft agglomerates, and allowed to settle for ~24 48 h. The supernatant was carefully collected with pipettes and dried to 80 C. Note that isopropanol was used as the sedimentation medium in preference to de-ionized water to avoid the formation of Bi(OH)3 platelettes which convert to -Bi2O3 upon firing. The subscript desi gnation and volume mean diameter, dV, for each starting powder is given in Table 5-1. BRO7 and ESB20 powders were then weighed in appropriate ratios. For studies involving the use of sonication and sedimentation, BRO7-ESB20 powders were combined in isopropanol and ultra-sonicated to achieve more intimate mixing. Inks were prepared by combining organic vehicles with these mixtures of BRO7 and ESB20 powders or pure BRO7 for the current collector. Once an appropriate viscosity was reache d, the inks were applied to both sides of the electrolyte substrates by paint br ushing to give symmetrically-electroded cells. These cells were dried at room temperature for 1 h, followed by dryi ng at 120 C for 1 h, and firing at 800 C for 2 h. Note that in order to vary electrode thickness, or for the addition of the current collector, successive layers were added after the drying stage, but before the firing stage. This was done to

PAGE 102

102 reduce densification and grain growth within th e cathode, as well as minimize ruthenium loss in the volatile BRO7 phase, both of which that may occur when firing each coat separately. Two was the standard number of coats applied to each cell. For testing under SOFC operati ng conditions, dense YSZ pellets with reference electrode bores were prepared. YSZ pellets with 99% th eoretical density were prepared by slip casting into a porous mold. The slip suspension was prepared using 30 vol% 8YSZ powder (Tosoh) in de-ionized water and an appropria te amount of citric acid pre-dissolved (~0.3 wt% citric acid to YSZ). Ammonium hydroxide was added dropwise to the mixture, with vigorous shaking, until a water-like consistence was reached. The slip wa s allowed to dry in the mold for 1 day and polished into flat disks of ~2 mm thickness. Hole s were carefully drilled halfway into the green pellets with a 1/16 inch bit. The pellets were then fired at 1400 C for 4 h. NiO-YSZ (50-50 wt%) pasts were applied to the ho le-free side of the YSZ pellets, dried, and fired at 1300 C for 1 h. Pt wire was beaded on one end using a torch, dipped into Pt paste, caref ully inserted into the hole, and fired at 1100 C for 2 h, along with a Pt current collector on the anode side of each cell. The cathodes (LSM-YSZ, LSCF-GDC, and BRO7-E SB) were then applied and fired at appropriate temperatures and times. LSM20 powde rs were obtained from Nextech Corporation, 8YSZ from Tosoh, LSCF from Praxair, and GDC10 from Rhodia, Inc. BRO7 and ESB powders were prepared in-house as descri bed above. Pt current collectors were applied to LSM-YSZ and LSCF-GDC cathodes. A pure BRO7 current collector was applied to the BRO7-ESB cathode. Current collectors were co-fired w ith each cathode. In order to improve the mechanical strength of the reference electrode, the hole was back filled with YSZ slip and fired at 700 C. Anode-supported cells were prepared by co-p ressing ~0.35 g GDC powder onto a NiOGDC substrate (~4 g) in a 1 1/8 cylindrical die. The NiO-GDC substrat e was first pressed at

PAGE 103

103 ~14 MPa. Next, GDC was carefully and uniformly spread across the surface, and pressed at ~42 MPa. The pellets were then pr essed isostatically at 250 MPa, a nd fired at 1450 C for 4 h using a 3 C/min heating rate and a 400 C, 1 h binder bu rnout step. The cells we re then electroded in the usual fashion. 5.2.2. Characterization Electrochemical performance of the el ectrodes was assessed using impedance spectroscopy. Silver mesh current collectors an d platinum lead wires were pressed against the samples in a quartz reactor by the use of a ceram ic screw-and-bolt assembly. The quartz reactor was placed into a temperature-controlled furnace, and air was fed through the reactor at a rate of 50 cc/min using a BOC mass flow c ontroller. A Solartron 1260 fr equency response analyzer was used in standalone mode for unbiased testing us ing a frequency range of 32 MHz to 0.01 Hz and an AC voltage amplitude of 100 mV. X-ray diffraction (XRD) patterns of various powders were recorded with a Philips APD 3720 diffractometer using K (Cu) = 1.5406 Step scans were taken over a range of 2 angles from 20 to 100 with 0.04 steps. A JEOL JSM 6400 scanning electron microsco pe (SEM) was used in this study for microstructural characterization. In add ition, a JEOL TEM 200CX transmission electron microscope (TEM) was also used in this study for particle si ze characterization. Average particle sizes and electrode por osities were characterized statistically from SEM and TEM micrographs using ImageJ software. For each powder, the volume mean diameter, dV, was calculated from Eqn. (5-1)

PAGE 104

104 n d dn i i V 3 1 3 (5.1) where di is the equivalent sphere diameter of each particle and n is the to tal number of particles analyzed by the ImageJ software. Note that pa rticle sizes are reporte d in terms of a volumeaverage as opposed to a number-average as this choice is more consistent with other commonlyreported microstructural properties such as porosity and composition. Statistical stereology on epoxy-resin embedde d, polished samples revealed that all electrodes had ~40 vol% porosity. For current-voltage measurements, cells were sealed (anode side) to an alumina tube using ceramabond (Aremco). The setup was then placed in to a furnace, cured, and taken up to testing temperature. Air and H2/H2O gas mixtures were used as the oxida nt and fuel gases, respectively. Flow rates were maintained at 30 ccm using ma ss flow controllers. Cell OCP was monitored using a Solartron 1287 potentiostat until a stab le value was reached, and current-voltage measurements were taken with the same in strument. A Solartron 1260 frequency response analyzer was used for impedance measurements. 5.3. Results and Discussion 5.3.1. Chemical Compatibility Chemical stability testing results between BRO 7 and ESB20 are shown in Fig. 5-1. XRD patterns of BRO7 and ESB20 powder mixtures before and after heat treatment at 800 C for 10 h are comparable, and reveal no evidence of interphase reactivity, suggesting these materials are chemically compatible at this temperature.

PAGE 105

105 5.3.2. Reproducibility/Compositional Optimization Before any microstructural optimization was conducted, BRO7-ESB20 composites were first compositionally optimized on ESB20 electro lytes using readily available powdersESB20 prepared from conventional solid-state s ynthesis and solid-state BRO7 powders which underwent an additional vibr atory milling step (ESB20S-BRO7VM). Optimization in composition was constrained between 25 and 75 wt% ESB with steps of 12.5 wt% ESB (and 6.25 wt% ESB at critical intermediate compositions). ES B Figure 5-2 shows typical impedance and Bode spectra obtained at 625 C in air und er no applied bias. Note that as in previous chapters, these and all future impedance plots have been Rs-corrected for ease in comparison. That is, the highfrequency real-axis intercept, Rs, of each spectrum, which is composed of the bulk electrolyte resistance and possibly electrode sh eet resistance and lead contact resistance, has been subtracted from the real component of each data point in the spectrum. At 625 C, each impedance plot appears to be composed of a single arc (Fig. 52a), and the characteristic frequency increases with ESB20 content (Fig. 5-2b). It is intere sting to note that the width of the imaginary impedance vs. log frequency spectrum is narrowe r toward the extremes of the concentration range than at intermediate compositions, suggesting a mechanistic overlap at these intermediate compositions, and in fact, a minimum value is observed at 56 wt% ESB20 over the range of temperatures tested. Figure 5-3 shows the trend in cathode ASR (calculated from Eq. 3-1) vs. composition for the ESB20S-BRO7VM system over a range of temperatures. Since the densities of BRO7 and ESB20 are approximately the same (~8.9 g/cm3), the observation of an ASR minima near 50 wt% of each phase is consistent with the e ffective medium percolation theory and 3PB maximization. For a two-phase composite whos e particle sizes are roughly the same, and assuming the solid phases and pores are random ly distributed and the porosity is open and

PAGE 106

106 sufficiently large, the 3PB length should reach a maximum if the two solid phases are present in equal fractions of the overall el ectrode volume. In the ESB20S-BRO7VM study, BRO7 particles are a factor of ~2 smaller th an ESB20 particles hence percolation and 3PB length maximization can occur at non-equal volume fractions of the two phases. As discussed above, 3PB length is important in two-phase cathode systems since th e oxygen reduction reaction that is occurring can only proceed at sites where all three reactants are presentas the cathode 3PB length increases, so do the number of reaction sites, and hence cathode activ ation polarization decreases. Note that this result differs from earlier results,66 where a minimum was observed between 31-43 wt% ESB. One possible explanation may be differences in powder synthesis techniques between the two studies. ESB20 was obtained through conventional so lid state synthesis in this study (micron-sized particles), as opposed to a we t chemical route (yieldin g nano-sized particles) used by Jaiswall, et. al.66 Further, in the present study th e BRO7 powders were vibratory milled for seven days versus thr ee days (Jaiswall, et. al);66 hence, yielding, on average, smaller BRO7 particles in the present study. Consequently, the ESB20-to-BRO 7 particle size ratio in the present study should be appreciabl y larger. As the ESB20-to-BRO7 particle size ra tio decreases, percolation of the ESB20 phaseand hence co mpositional optimization of the cermet cathode becomes feasible at smaller ESB20 volume fractions. Note that there also appears to be a differen ce in activation energies between the electrodes obtained in this study (1.02 eV) compar ed with Jaiswall, et al. (~1.3 eV).66 The electrode ASR at 500 C are comparable (3.11 cm2 for the present study compared to 3.37 cm2), but are quite different at 700 C (0.17 cm2 for the present study compared to 0.08 cm2). It is not clear at this juncture what mechanism leads to this lo wer activation energy. It has frequently been reported that the surface of bismuth oxide-based electrolytes is active for the adsorption of

PAGE 107

107 oxygen species. It is believed that bismut h strongly enhances the surface oxygen exchange rate.69-71 This is evidenced by the factor of 103 larger surface oxygen exch ange rate for bismuth oxide based solids compared to YSZ solids. In addition, bismuth-based electrolytes have been reported to be catalytically active toward s oxygen dissociation and charge transfer.72,73 It is thus possible that the use of ESB20 (rather than GDC ) as the electrolyte support, which results in additional ESB20 sites at the cathode/electrolyte may result in the observed drop in activation energy. 5.3.3. Optimization by Particle Size Ratio Having shown an acceptable level of reproducib ility in performance, microstructural optimization was next carried out. Given the difference in optimum c oncentration between the two compositional studies discussed above, careful consideration was given with respect to the choice of concentration to use for the microstruc ural study. It was d ecided to fix the BRO7ESB20 composition at 50-50 wt% (approximately 50-50 vol%) so as to minimize any possible bias that might arise from skewing the com position toward either end of the spectrum. Electrode microstructure was first optimized according to starting particle size ratio, constrained to the particle sizes obtained from crushing and sieving as well as vibratory milling. Particle sizes of the constituent phases were altered using vibratory milling as well as a combination of sonication and sedimentation, as e xplained in the experimental section. Figures 5-4 and 5-5 show representative samples of each particle system used in the present study. Figure 5-4 compares BRO7 and ESB20 powders as prepared from solid state synthesis, and directly after vibratory milling. Particle size analysis (from imaging software) reveals that the volume mean diameter was reduced by a factor of approximately two after seven days of vibratory millingfrom ~1.31 m to ~0.73 m for BRO7 and from ~1.31 m to ~0.81 m for

PAGE 108

108 ESB20. Note however that each set of vibr o-milled powders still contains a small number fraction (but significant volume fraction) of large, unbroken particles. Figure 5-5 shows the further reduction in particle size and agglomeration that ca n be obtained by sonication and sedimentation. Four different electrode struct ures were first examined using only as-prepared powders (BRO7S-ESB20S, BRO7S-ESB20VM, BRO7VM-ESB20S, and BRO7VM-ESB20VM), each containing a 1:1 volume ratio of the constituent phases. An SEM image of the four fully-fired electrodes is shown in Figure 5-6. Figure 5-7 shows typical impedance spectra and Bode plots obtained at 625 C for these electrodes in air unde r no applied bias. Plots of ASR vs. reciprocal temperat ure are given in Figure 5-8. These results show that the compos ites exhibiting the lowest resistance are those composed of both large and small particles. All composites containing at least one vibro-milled phase have significantly lower resistance than the composite composed entirely of large particles, as expected, due to the increased 3PBs. The activation energies for all electrodes are similar (1.03eV 0.03eV), as ar e their characteristic frequenc ies (Figure 5-7b), suggesting a common rate-determining mechanism for the different electrodes. It is not known at this time which mechanism is dominant; more work is needed to clarify this issue via testing under different oxygen partial pressures. However, it is evident that the magnitude of the impedance of this mechanism has been markedly affected by these microstructu ral considerations. Comparing the two systems where the different phases have more similar grain size (BRO7VM-ESB20VM and BRO7S-ESB20S), the electrode comprised of vibro-milled particles exhibits the lower ASR, as expected. The sy stems composed of dissimilarly-sized particles (BRO7VM-ESB20S and BRO7S-ESB20VM) exhibit nearly the same performance, with the

PAGE 109

109 BRO7VM-ESB20S electrode having the lowest ASR of 0.43 cm2 at 625 C. Interestingly, the electrodes containing dissimilar grain sizes exhibit lower ASR values than the electrodes containing more similar grain sizes, despite being shifted towards non-optimal composition ratios, as discussed above. Recall th at, although the optimal composition of BRO7VM-ESB20VM was 56 wt% ESB20, for the micros tructural study a 1:1 ratio was used. One possible explanation for this observation could be the formation of soft agglomerates of fine particles during synthesis, as can be seen for the as-prepared ESB20VM and BRO7VM powder in Figure 5-5. Another possibility, as discussed in Section 2.5, is that smaller particles create a larger degree of microstructural tortuosity, which can restrict gas phase diffusion.13 Note also that initially, samples used in this study were prepared with a single coat of cathode paste. However a larg e drop in the ASR of the BRO7S-ESBVM was observed when a sample of greater thickness was tested, hence a two-coat standard for thickness was adopted. The effects of cathode thickness on performa nce is discussed in Section 5.3.5. 5.3.4 Sonication and Sedimentation As mentioned in Section 5.3.3, as-prepared vibratory-milled powders contain a significant volume fraction of unbroken, micron-sized particles. Focused ion beam (FIB) analysis and reconstruction (a video file which can not be re produced here) also reveals a degree of phase segregation in the fired electrodes. The next stage in the microstructural development involved the use of sedimentation for further reductions in particle size and a narrower particle size distribution. Also, ultra-sonication was used to break up soft agglomerates and improve phase distribution in the cathode. The results are shown in Figures 5-9 and 5-10. Impedance results show a large reduction in resistance for all systems where the raw vibr o milled phase was replaced with the supernatant phase. It is clear from Figur e 5-9b that the impedance spectr um of the fully-supernatant

PAGE 110

110 composite is composed of two arcs. Comparing the spectrum of this electrode with the other electrodes, it appears that the low frequency process is the most affected by these microstructural changes, and its resistance has been lowered to the point where the low and high frequency processes are in competition with each other. Figure 5-10 overlays the Arrhenius behavior of composites which underwent sonication and sedimentation with those composed of as-p repared powders. Both composites consisting of a mixture of large and small particles underwen t comparable ASR reductions. However, the ASR reduction was the most dramatic for the com posite comprised entirely of small particles. Clearly this composite benefits the most from the combination of reduced particle size and the more intimate sonicated mixing. Not only were the soft agglomerates of each phase broken up, but also the reduced particle si ze distribution translates into a larger number fraction of submicron sized particles in each vibratory milled phase This in turn translat es into a larger 3PB length. As discussed in Section 5.3.3 reduction in particle size gi ves rise to increased tortuosity which can inhibit gas phase diffusion. However, as mentioned above, the spectra presented in Figure 5-9 indicate that a reduction in particle size, size distri bution, and phase agglomeration provided by sonication and sedimentation primarily reduces the impedance of a single lowfrequency process. Since a higher-frequency pro cess is only discernabl e once the impedance of the low-frequency process is dramatically reduced, it is reasonable to conclude that the improvements provided by increa sed 3PB lengths vastly outw eigh any degradation resulting from increased tortuosity. The minimum ASR observed was 0.10 cm2 at 625 C for the ESB20SUP-BRO7SUP system.

PAGE 111

111 5.3.5. Effect of thickness and current collection The effect of thickness on ASR for the different composite microstructures is illustrated in Figure 5-11. In each case, ASR is reduced as thickness is increased, over the range of thicknesses studied. The ESB20SUP-BRO7S system exhibited a one order of magnitude drop in ASR between the first and second coating. This observation was confirme d to be reproducible (from repeated testing on once-coated samples), and is believed to be due to a current collection issue where the electrode thickness is insuffi cient for BRO7 phase percolation. That the ESB20S-BRO7S system did not show as significa nt an ASR reduction as the ESB20SUP-BRO7S system can be attributed to a reduction in BRO7 phase connectivity cau sed by the fine ESB20SUP particles percolating between adjacent BRO7S grains. The volume percent porosity (open symbols) of each composite tested is overlaid on Figure 5-11, and is relatively constant (40 vol%). There is no apparent trend between thic kness and porosity over this narrow range, and hence the porosity variation betw een electrodes tested is expected to have minimal influence on the thickness vs. ASR results. It is also interes ting to note that electrod e ASR continues to drop, even at thicknesses beyond 100-200 m. As stated in Section 2.4, in general, if the conductivity of the ionically conducting phase in a composite electrode is su fficiently high, the effective charge transfer resistance will decrease with thic kness down to an asymptotic value. This indeed appears to be the case in Figure 5-11, and the exceedingly high ionic c onductivity of the ESB20 phase may explain why dominant concentration polarization effects do not show up at these electrode thickness values. In order to improve connectivity of the electronic conducting phase, a new batch of twicecoated samples was prepared, this time using a pure BRO7S current collector layer. This layer did not adhere well to the two electrode systems which utilized fine particles of ESB20.

PAGE 112

112 However, for the other systems, there is a dram atic drop in the resistan ce of the low-frequency process, as shown in Figure 5-12. For the BRO7SUP-ESB20S system, the high-frequency process now seems to dominate the performance. As shown in Figure 5-13, the ASR of these systems was reduced significantlyby factor of four (from 0.58 cm2 to 0.15 cm2) in the ESB20SBRO7S system, and by a factor of three (from 0.22 cm2 to 0.076 cm2) in the ESB20SBRO7SUP system. The former system composed of large BRO7 grains exhibited a larger overall ASR reduction than the system composed of fi ne BRO7 grains, as expected from enhanced current collection. That is, percolation and pha se connectivity is more easily achieved when the phase is comprised of small particles and more difficult when the phase is comprised of large particles. Thus, it is expected that the ESB20S-BRO7SUP system will have inherently better electronic phase connec tivity than the ESB20S-BRO7S system, hence introduction of a current collecting layer should have less of an infl uence in the former system. The ESB20S-BRO7SUP system exhibited the lowest AS R of all systems studied (0.73 cm2 and 0.03 cm2 at 500 C and 700 C, respectively). This is a marked improvement over earlier results (3.47 cm2 and 0.08 cm2 at 500 C and 700 C, respectiv ely) and is comparable to landmark results reported by Zhao and Haile46 for Ba0.5Sr0.5Co0.8Fe0.2O3(BSCF), especially at lower temperatures due to the lower activation energy of BR O7-ESB20 compared with BSCF Further improvements are expected by improvi ng the adhesion of the current collection layer to the system composed enti rely of fine particles since this exhibited the best performance of all systems without a current collector. In addition, furt her compositional optimization may be performed on each specific system studied in this work. As mentioned previously, bismuth ruthenate is known to be volatile, and bism uth oxides are known to undergo an ordering

PAGE 113

113 phenomenon at temperatures below ~600 C. Thus, the stability of this system above 600 C should be examined. 5.3.6. Direct comparison with co nventional cathode systems Rather than simply comparing results with literature data, in-house LSM-YSZ and LSCFGDC composites were synthesized. The similari ty in processing routines of the different composite systems provides a more direct comp arison with conventional cathode systems. The four different materials (LSM fr om Nextech, YSZ from Tosoh Cor poration, LSCF from Praxair, and GDC from Rhodia) were obtaine d and have very fine initial pa rticle sizes. The conventional composites were first optimized according to composition (using 5 vol% steps between 40-60 vol% of each phase using 5 vol % composition steps), firing te mperature (using 50 C steps between 1000 C and 1300 C), and firing ti me (using 1 h intervals between 1-4 h). All three sets of composites (LSM-YSZ, LSCF-GDC, and BRO7VM-ESBS) were applied to dense YSZ pellets with Ni-YSZ anodes. Using th e same electrolyte subs trate for each system minimizes the effect the electrode/electrolyte interface has on cathode performance. YSZ was selected as the substrate since it is more easily sintered than GDC, and unlike ESB, can withstand the firing temperatures used for LSM-YSZ and LSCF-GDC composites (1100 C and 1150 C, respectively). Further, testing was done under actual SOFC operating conditions using a solid state adaptation of the Luggin probe configuration (Figure 5-14). The embedded reference electrode in this configuration more accurately samples a well-defined equipotential surface, allowing for more accurate measuremen t of electrolyte and electrode impedances, compared with other surface-configured refere nce electrodes which sample a more averaged effective potential, leading to greater likelihood of inaccurate electrolyte measurements and distorted electrode arcs74. Further, 3-point testing under SOFC operating conditions allows

PAGE 114

114 direct assessment of cathode pol arization as a function of curre nt density, giving insight into real-world behavior of these cathode systems. Cathodic polarization can be calculated using Eq. 5-2 cathode = Em EYSZ,ref || cat (5-2) where Em is the potential measured between the cathode and the embedded reference electrode, and EYSZ, ref || cat is the potential drop due to the YSZ elect rolyte between the reference electrode and the cathode. This potential drop can easily be measured in-situ using impedance spectroscopy between the cathode and reference electrode. Th e high-frequency intercept of the impedance spectrum with the real axis (Zhf, pt || cat) allows calculation of the ASR of the YSZ electrolyte between the reference electrode and the cat hode (Eq. 5-3) ASRYSZ, ref || cat = Zhf, pt || cat A (5-3) where A is the active area of the cell (the cathode area). The voltage drop due to the YSZ electrolyte between the referen ce electrode and the cathode is then simply the ASR value obtained from Eq. 53 multiplied by the cell current density. Since the electrolyte resistance is ohmic, its resistance is independent of curre nt density, a single, opencircuit impedance measurement of Zhf, pt || cat can be used to determ ine the cathode overpotential at all current densities (i) measured (Eq. 5-4). cathode(i) = Em(i) i Zhf, pt || cat A (5-4) A similar 3-point impedance measurement between the reference electrode and anode provides the ohmic resistance between the an ode and the reference electrode Zhf, pt || an, as well as the polarization resistance of the anode. As one accuracy check, the embedded referen ce electrode effectively partitions the ohmic contribution of the electrolyte and allows separation electrode polarization. Thus, summing up

PAGE 115

115 the partitioned ohmic contributions from the 3-point impedance measurement (Zhf, pt || an and Zhf, pt || cat) should give the same value of ohmic resist ance as a simple 2-point measurement between the anode and cathode (Eq. 5-5). Zhf, 2-point = Zhf, pt || an and Zhf, pt || cat (5-5) Further, the sum of the real a nd imaginary parts of the cathodic and anodic impedances measured using the 3-point configuration s hould be the same as the real and imaginary parts of the total electrode impedance measured using th e 2-point configur ation (Eqs. 5.6). Ztot, 2-point = Zcat, 3-point + Zan, 3-point (5-6a) Ztot, 2-point = Zcat, 3-point + Zan, 3-point (5-6b) Such an analysis is shown in Figure 5-15 fo r the LSM-YSZ system measured at 650 C. There is a slight (~2 %) deviat ion between the electrolyte resistance measured from 2-point and 3-point measurements. The agreement between the total electrode resistance measured from both configurations is better (d eviation ~1 %). The deviations arise most likely from human error involved in fabrication of the Luggin probe cell, but overall, the agreement is acceptable. The high value of total cell ASR (44.5 cm2) arises mainly from the f act that testing is done on thick (~2 mm) YSZ pellets. Th e cathode ASR measured (6.6 cm2) agrees reasonably well with data for LSM-YSZ reported in the literature (Figure 3-2). Current-voltage data for the same cell under th e same conditions is shown in Figure 5-16. The slope of the line near open-circuit conditions gives a measure of the total cell ASR, which under these conditions is 45.5 cm2. This shows good agreement with the total ASR determined from impedance measurement discussed above. Ca thodic polarization at 650 C as a function of current density for the 3 systems studied is show n in Figure 5-17. As expected, the polarization drop across the BRO7VM-ESBS composite cathode was signifi cantly lower than that of

PAGE 116

116 conventional materials. The ASR of BRO7VM-ESBS, calculated from the slope of the cathodic overpotential-current de nsity plot, is 0.9 cm2. This value is signifi cantly higher than that obtained earlier for this system at 650 C (~0.5 cm2). The discrepancy is most likely due to the fact that in the current study, testing was done on a YSZ electroly te while previous testing was done on ESB electrolytes. YSZ not only has orders of magnitude lower i onic conductivity than ESB at this temperature, but also does not have the catalytic ac tivity of ESB, as discussed in Section 5.3.2. 5.3.7. Performance under operation To gauge the viability of using BRO7-ES B composites, good performance should be demonstrated. In order to demonstrate good perfor mance, three samples were prepared using NiGDC anode-supported electrolytes with relatively thin (~50 m) GDC electrolytes. To evaluate performance of this cathode on ESB, one sa mple was coated with a thin film (~25m) of ESB, yielding a bilayer electrolyte with GDC on the fuel side and ESB on the air side. Note that the GDC layer is necessary to protect the ESB layer, which decomposes unde r reducing conditions. A LSCF-GDC composite cathode with a Pt current collector was used on one of the cells as a reference standard. The optimized BRO7-ESB co mposite with a pure BRO7 current collector was used on the other two cells (one with an ESB/GDC bilayer, the other having a GDC single layer electrolyte). The results of current-voltage testing at 650 C are shown in Figure 5-18 as well as in Table 5.2. Results on the single-layer GDC s how that the BRO7-ESB composite has slightly better performance than LSCF-G DC (total cell ASR is 0.81 and 0.85 for BRO7-ESB and LSCFGDC, respectively). Since the anod e supports and the electrolytes we re all prepared at the same time, resistances due to the anode and electrolyte should be the sa me for both cells, so the slight

PAGE 117

117 improvement can be attributed mainly to the cathode. Since the electrolyte is ~50 m thick, much of the total cell resistan ce should be due to the electro lyte. Thus the improved cathode performance is masked to some extent due to th e resistance of the electrolyte (as well as the anode). The cell utilizing a bilayer el ectrolyte had significantly better performance compared with the other cells (362 mW/cm2 maximum power density and 0.53 cm2 ASR), despite the added resistance caused by inserting an ESB laye r between the GDC and the cathode. This is partly due to a degree of catalytic activity present at the surface of the ESB electrolyte, and the role the cathode/electroly te interface plays in the cathode resistance, as discussed in Sections 5.2.3 and 5.2.6. The full picture of the impr oved performance is not fully understood at this point. Further, there may be some stability i ssues at this temperature as post-mortem SEM analysis reveals formation of an interlayer betw een the cathode and the ESB layer (Figure 5-19). More work is required to resolv e these issues. However, it is clear that these cathodes have potential for use in lower temperature SOFCs. 5.4. Conclusions BRO7 was shown to be chemically compatible with ESB20. Microstructural engineering by a combination of mechanical crushing, ultrason ication, and sedimentation was shown to be an effective way of lowering electrode ASR, and the results seem to be consistent with 3PB length maximization theory. Application of a pure BRO 7 current collector to the electrode surfaces further improved electrode performance. The lowest value of ASR attained ranged from 0.73 cm2 at 500 C to 0.03 cm2 at 700 C is one of the lowe st SOFC electrode ASR values reported to date. Dir ect comparison of BRO7VM-ESBS with conventional LSM-YSZ and LSCFGDC composite cathodes were made using a so lid-state adaptation of the Luggin probe

PAGE 118

118 configuration to extract cathode overpotentials as a f unction of current density. Results show BRO7VM-ESBS to have a significantly lower cathodi c overpotential than these conventional composites. Current-voltage testing of the optimized composite was done on anode-supported cells. A maximum power density of 362 mW/cm2 was attained at 650 C using a ~75 mm thick ESB/GDC bilayer electrolyte. It is believed that this material would be a good candidate cathode for low temperature SOFCs.

PAGE 119

119 Table 5-1. Volume mean diameter (dV), as well as subscript and symbolic designations for different sets of starting pow ders. Hashed and white circles represent ESB particles, black and gray circles represent BRO particles. Phase subscript designation Type of powder dV ESB20 (m)dV BRO7 (m) Symbol S Sieved 1.31 1.31 VM Vibro-milled 0.81 0.73 SUP Vibro-milled, supernatant 0.08 0.06 Table 5-2. Cell open circuit potential s (OCP), maximum power densities (MPD), and ASR for selected SOFCs at 650 C from current-density measurements. Cell Type OCP (V) MPD (mW/cm2) ASR (cm2) LSCF-GDC on GDC 0.91 250 0.85 BRO7-ESB on GDC 0.90 269 0.81 BRO7-ESB on ESB/GDC 0.87 362 0.53

PAGE 120

120 0 1000 2000 3000 4000 5000 6000 7000 20304050602 ()Intensity (Arbitrary Units ) F F F F F P P P P P(222) (400) (331) (440) (622) (111) (200) (220) (311) (222)P(531)P(311)800 C, 10 h unfired 0 1000 2000 3000 4000 5000 6000 7000 20304050602 ()Intensity (Arbitrary Units ) F F F F F P P P P P(222) (400) (331) (440) (622) (111) (200) (220) (311) (222)P(531)P(311) 0 1000 2000 3000 4000 5000 6000 7000 20304050602 ()Intensity (Arbitrary Units ) F F F F F P P P P P(222) (400) (331) (440) (622) (111) (200) (220) (311) (222)P(531)P(311)800 C, 10 h unfired Figure 5-1. XRD spectra for mixt ures of BRO7-ESB20 before and after heat treatment at 800 C for 10 h.

PAGE 121

121 A B Figure 5-2. Nyquist (a) and B ode (b) plots at 625 C for different compositions of BRO7VMESB20S electrodes tested in air.

PAGE 122

122 1 10 20304050607080ASR ( cm2)Wt% ESB500 C 550 C 600 C 625 C Figure 5-3. Effect of electrode composition for the ESB20S-BRO7VM system.

PAGE 123

123 BRO7S BRO7SA BRO7VM BRO7VMB ESB20S ESB20SC ESB20VM ESB20VMD Figure 5-4. SEM micrograph of as-prepared BRO7 powders befo re (a) and after (b) vibromilling as well as ESB20 powders before (c) and after (d) vibro-milling.

PAGE 124

124 BRO7VM BRO7VMA BRO7SUP BRO7SUPB ESB20VM ESB20VMC ESB20SUP ESB20SUP Figure 5-5. TEM micrograph of BRO7VM powders before (a) a nd after sonication and sedimentation (b), as well as ESBVM powders before (c) and after (d) sonication and sedimentation. Powders were di spersed onto lacey carbon TEM grids.

PAGE 125

125 10 mESB20SBRO7S 10 m 10 mESB20SBRO7S A 10 mESB20S BRO7VM 10 m 10 mESB20S BRO7VMB 10 mESB20VMBRO7S 10 m 10 mESB20VMBRO7S C 10 mBRO7VM+ESB20VM 10 m 10 mBRO7VM+ESB20VMD Figure 5-6. SEM image of four fully-fired BRO7-ESB20 cathode system s used in this study BRO7S-ESBS (a), BRO7VM-ESBS (b), BRO7S-ESBVM (c), BRO7VM-ESBVM (d).

PAGE 126

126 A B Figure 5-7. Nyquist (a) and B ode (b) plots at 625 C for different 50-50 wt% BRO7-ESB20 electrode microstructures tested in air.

PAGE 127

127 1 10 1.1 1.15 1.2 1.25 1.3ASR( cm2)1000/T (K-1) 1 10 1.1 1.15 1.2 1.25 1.3 1 10 1.1 1.15 1.2 1.25 1.3ASR( cm2)1000/T (K-1) Figure 5-8. Arrhenius plot of ASR vs reci procal temperature fo r the four different microstructures studied.

PAGE 128

128 00.51.01.5 -0.5 0 Z( ) ( ) 625 C 00.51.01.5 -0.5 0 Z( ) ( ) 625 C 00.51.01.5 -0.5 0 00.51.01.5 -0.5 0 Z( ) ( ) 625 C Z( ) 00.51.01.5 -0.5 0 Z( ) ( ) 625 C 00.51.01.5 -0.5 0 Z( ) ( ) 625 C 00.51.01.5 -0.5 0 00.51.01.5 -0.5 0 Z( ) ( ) 625 C Z( )A 10-1100101102103 -0.4 -0.3 -0.2 -0.1 0 Frequency(Hz)625 C 10-1100101102103 -0.4 -0.3 -0.2 -0.1 0 Frequency(Hz)625 C 10-1100101102103 -0.4 -0.3 -0.2 -0.1 0 10-1100101102103 -0.4 -0.3 -0.2 -0.1 0 Frequency(Hz)625 C Z( ) 10-1100101102103 -0.4 -0.3 -0.2 -0.1 0 Frequency(Hz)625 C 10-1100101102103 -0.4 -0.3 -0.2 -0.1 0 Frequency(Hz)625 C 10-1100101102103 -0.4 -0.3 -0.2 -0.1 0 10-1100101102103 -0.4 -0.3 -0.2 -0.1 0 Frequency(Hz)625 C Z( )B Figure 5-9. Nyquist (a) and B ode (b) plots at 625 C for different 50-50 wt% BRO7-ESB20 electrode microstructures tested in air before (open symbols) and after (closed symbols) sonication and sedime ntation of electrode powders.

PAGE 129

129 0.1 1 10 1.11.151.21.251.3 ASR( cm2)1000/T (K-1) 0.1 1 10 1.11.151.21.251.3 ASR( cm2)1000/T (K-1) Figure 5-10. Arrhenius plot of ASR vs reciprocal temperaturea comparison between electrodes prepared from as-prepared pow ders (open symbols) and powders which underwent ultrasonication and se dimentation (closed symbol s). Arrows indicate ASR drop for comparable systems.

PAGE 130

130 0.1 1 0 20 40 60 80 100 050100150200250300625 C ASR ( cm2)Thickness ( m)Porosity (vol %) 0.1 1 0 20 40 60 80 100 050100150200250300 0.1 1 0 20 40 60 80 100 050100150200250300625 C ASR ( cm2)Thickness ( m)Porosity (vol %) Figure 5-11. Effect of electrode thickness on AS R at 625 C for the four different electrode microstructures prepared after ultrasonicat ion and sedimentation of the as-prepared powders. Hollow symbols represent electrode porosity.

PAGE 131

131 00.250.500.75 -0.25 0 Z( )Z( )625 C a) b) 00.250.500.75 -0.25 0 Z( )Z( )625 C 00.250.500.75 -0.25 0 00.250.500.75 -0.25 0 Z( )Z( )625 C a) b) 00.250.500.75 -0.25 0 Z( )Z( )625 C a) b) 00.250.500.75 -0.25 0 Z( )Z( )625 C 00.250.500.75 -0.25 0 00.250.500.75 -0.25 0 Z( )Z( )625 C a) b) A 10-1100101102103 -0.15 -0.10 -0.05 0 Z( )Frequency(Hz)625 C 2 coats + BRO cc b) 10-1100101102103 -0.15 -0.10 -0.05 0 Z( )Frequency(Hz)625 C 2 coats + BRO cc 10-1100101102103 -0.15 -0.10 -0.05 0 10-1100101102103 -0.15 -0.10 -0.05 0 Z( )Frequency(Hz)625 C 2 coats + BRO cc b) 10-1100101102103 -0.15 -0.10 -0.05 0 Z( )Frequency(Hz)625 C 2 coats + BRO cc b) 10-1100101102103 -0.15 -0.10 -0.05 0 Z( )Frequency(Hz)625 C 2 coats + BRO cc 10-1100101102103 -0.15 -0.10 -0.05 0 10-1100101102103 -0.15 -0.10 -0.05 0 Z( )Frequency(Hz)625 C 2 coats + BRO cc b) B 00.51.01.52.0 -0.67 -0.17 Z( )Z( )625 C c ) 00.51.01.52.0 -0.67 -0.17 Z( )Z( )625 C 00.51.01.52.0 -0.67 -0.17 00.51.01.52.0 -0.67 -0.17 Z( )Z( )625 C c ) 00.51.01.52.0 -0.67 -0.17 Z( )Z( )625 C c ) 00.51.01.52.0 -0.67 -0.17 Z( )Z( )625 C 00.51.01.52.0 -0.67 -0.17 00.51.01.52.0 -0.67 -0.17 Z( )Z( )625 C c ) C 10-1100101102103 -0.5 -0.4 -0.3 -0.2 -0.1 0 Z( )Frequency(Hz)625 C 1 coat 2 coats 4 coats 2 coats + BRO cc d ) 10-1100101102103 -0.5 -0.4 -0.3 -0.2 -0.1 0 Z( )Frequency(Hz)625 C 1 coat 2 coats 4 coats 2 coats + BRO cc 10-1100101102103 -0.5 -0.4 -0.3 -0.2 -0.1 0 10-1100101102103 -0.5 -0.4 -0.3 -0.2 -0.1 0 Z( )Frequency(Hz)625 C 1 coat 2 coats 4 coats 2 coats + BRO cc d ) 10-1100101102103 -0.5 -0.4 -0.3 -0.2 -0.1 0 Z( )Frequency(Hz)625 C 1 coat 2 coats 4 coats 2 coats + BRO cc d ) 10-1100101102103 -0.5 -0.4 -0.3 -0.2 -0.1 0 Z( )Frequency(Hz)625 C 1 coat 2 coats 4 coats 2 coats + BRO cc 10-1100101102103 -0.5 -0.4 -0.3 -0.2 -0.1 0 10-1100101102103 -0.5 -0.4 -0.3 -0.2 -0.1 0 Z( )Frequency(Hz)625 C 1 coat 2 coats 4 coats 2 coats + BRO cc d ) D Figure 5-12. Nyquist (a,c) and Bode (b,d) plots at 625 C for 5050 wt% BRO7-ESB20 at different thicknesses without (open symbol s) and with (closed symbols) pure BRO7 current collectors. Note: The BRO7SUP-ESBS system is shown at left and BRO7SESBS is shown at right.

PAGE 132

132 0.1 1 11.051.11.151.21.251.3 ASR( cm2)1000/T (K-1) 2 coats + BRO cc 0.1 1 11.051.11.151.21.251.3 0.1 1 11.051.11.151.21.251.3 ASR( cm2)1000/T (K-1) 2 coats + BRO ccA 0.1 1 10 11.051.11.151.21.251.3 ASR( cm2)1000/T (K-1) 2 coats + BRO cc 0.1 1 10 11.051.11.151.21.251.3 0.1 1 10 11.051.11.151.21.251.3 ASR( cm2)1000/T (K-1) 2 coats + BRO ccB Figure 5-13. Arrhenius plot of ASR vs reciprocal temperaturea comparison between electrodes without (open symbol s) and with (closed symbols) current collectors for the BRO7SUP-ESBS (a) and BRO7S-ESBS (b) systems.

PAGE 133

133 Em i cathode anode Pt ref (embedded) YSZ electrolyte Em i cathode anode Pt ref (embedded) YSZ electrolyte A B Figure 5-14. Solid state adaptation of 3-poi nt Luggin probe configuration (a) schematic representation and (b) actual cell. 1020304050 -5 0 Z( cm2)Z( cm2)anode (3 pt) cathode (3 pt) total (2 pt) total (anode+cathode) 1020304050 -5 0 Z( cm2)Z( cm2)anode (3 pt) cathode (3 pt) total (2 pt) total (anode+cathode) Figure 5-15. Impedance spectra for LSM-YSZ composite comparing and total cell impedance measured using 2-point configuration with th at calculated from the sum of the anode and cathode impedance measured usi ng 3-point Luggin reference probe configuration. Data has been nor malized according to cathode area.

PAGE 134

134 00 0 10 0 20 0 3 0 0.5 1.0 1.5 I (Amps/cm2)E (Volts) ASR = 45.5 cm2Cathode: LSM-YSZ Electrolyte: YSZ Anode: Ni-YSZ T=650 C Fuel: H2/H2O Oxidant: Air 00 0 10 0 20 0 3 0 0.5 1.0 1.5 I (Amps/cm2)E (Volts) ASR = 45.5 cm2Cathode: LSM-YSZ Electrolyte: YSZ Anode: Ni-YSZ T=650 C Fuel: H2/H2O Oxidant: Air Figure 5-16. Current-voltage measurement for LSM-YSZ on Luggin probe cell at 650 C using hydrogen bubbled through wate r as the fuel gas and ai r as the oxidant gas.

PAGE 135

135 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 00.010.02cathode(V)Current (mA/cm2)LSM-YSZ LSCF-GDC BRO7-ESB Electrolyte: YSZ Anode: Ni-YSZ T=650 C Fuel: H2/H2O Oxidant: Air 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 00.010.02cathode(V)Current (mA/cm2)LSM-YSZ LSCF-GDC BRO7-ESB Electrolyte: YSZ Anode: Ni-YSZ T=650 C Fuel: H2/H2O Oxidant: Air Figure 5-17. Cathode overpotentia l versus current density data for selected cathode materials on Luggin probe cells at 650 C using hydrogen bubbled through water as the fuel gas and air as the oxidant gas.

PAGE 136

136 00 51 01 5 0 0.25 0.50 0.75 1.00 I (Amps/cm2)E (Volts) 0 0.1 0.2 0.3 0.4Power (Watts/cm2) Standard (LSCF-GDC on GDC) BRO-ESB (on GDC) BRO-ESB (on ESB/GDC) 00 51 01 5 0 0.25 0.50 0.75 1.00 I (Amps/cm2)E (Volts) 0 0.1 0.2 0.3 0.4Power (Watts/cm2) Standard (LSCF-GDC on GDC) BRO-ESB (on GDC) BRO-ESB (on ESB/GDC) Figure 5-18. Current-voltage measurement for selected cells at 650 C using hydrogen bubbled through water as the fuel gas and air as the oxidant gas.

PAGE 137

137 Figure 5-19. SEM image of optimized BRO7-ES B composite cathode on SOFC with Ni-GDC anode support with ESB/GDC bilayer SO FC after current-voltage testing.

PAGE 138

138 CHAPTER 6 CONCLUSIONS The solid ox ide fuel cell is a promising candi date for future generation power generation technologies. Since they do not re ly on combustion of fuels, they are more efficient, quieter, and cleaner than conventional technologies. However, in order to be mo re practical and costeffective, cell operating temperatures must be lo wered to 500 C and below. With the inception of thin-film electrolytes and el ectrolytes with ioni c conductivities higher than conventional YSZ materials, much of the temperature reduction focus has shifted to elec trode (anode and cathode) development. The oxygen reduction reaction (Eq. 2-1a and 2-1b), being more of a thermallyactivated process than that of fuel oxidation (Eq. 2-2 and 2-3), becomes a severely limiting process at lower temperatures. Hence, low and intermediate-temperature SOFCs require cathodes with a high degree of catalytic ac tivity towards oxygen re duction as well as a microstructure which maximizes the number of reaction sites (3PBs) and facilitates oxygen transport toward and incorporation into the electrolyte. It is al so important that the cathode be microstructurally stable with time so that perf ormance does not degrade over the lifetime of the cell. In the first part of this dissertation, the isothermal stability of a low-resistance cermet cathode, silver-stabilized bismuth oxide was examined Prior to stability testing, a preliminary study of compositional optimizatio n (between 40 and 70 vol% Ag phase at intervals of 5 vol% Ag) was carried out on Ag-ESB20 composites in order to achieve a degree of agreement with data published in the literature. The minimu m in cathode ASR (achieved at 50 vol% ESB20) was 0.18 cm2 at 600 C, a significant improvement over that produced by Xia et al.39 who obtained an ASR of 0.3 cm2 at 600 C for their Ag-YSB com posite. For the stability study, pure Ag, Ag-ESB20, and Ag-YSB electrodes were isot hermally tested at 650 C for 100 h. All

PAGE 139

139 systems experienced significant degradation in electrochemical performance during the test both the Ag-YSB and Ag-ESB electrode AS R values increased by around 70% (from 0.04 cm2 to 0.07 cm2 (75%) and from 0.06 cm2 to 0.10 cm2 (67%) for Ag-YSB and Ag-ESB respectively), while the pure Ag system expe rienced a near fourfold increase (from 0.92 cm2 to 3.55 cm2) during the same length of time. Linear regression of the data (based on the first 10 h of testing) yielded a de gradation rate of 4.4x10-4 cm2/h for Ag-ESB and 1.2x10-4 cm2/h for Ag-YSB. The pure Ag electrode degraded at a ra te more than two orders of magnitude faster 8.7x10-2 cm2/h. Electrolyte conductivities during the first 50 h of tes ting were relatively stable. SEM analysis revealed significan t microstructural evolution duri ng the 100 h of testing at 650 C. The pure Ag electrode appears fully dens e. The porosity of the Ag-ESB20 electrode appeared lower after testingdomains of coales ced silver are evident, and ESB20 particles between these domains have been forced closer together, further reduci ng porosity and 3PBs. Growth and coalescence of the silver phase was also evident in the Ag-YSB electrode though to a lesser extent. It is believed that the smaller starting size of the YSB pa rticles (as evidenced in the unfired electrode micrographs) helped to lo wer the mobility of the silver phase by providing the electrode with a higher su rface area through which the silver phase must migrate. XRD and EPMA analysis revealed neither evidence of inte r-phase reactivity between silver and stabilized bismuth oxide nor diffusion of s ilver into the electrolyte. In light of the electrochemical, microstructural, and chemical evidence pr esented, it was concluded that electrode microstructural evolution due to growth, agglomer ation, and coalescence of the silver phase, rather than chemical reactivity of the bismut h oxide phase, was responsible for the observed degradation in electrochemical performance.

PAGE 140

140 Next, attempts were made to reduce the micr ostructural evolution of the silver phase in Ag-ESB20 composites, and hence improve electroch emical performance stability. This was done by infusing the electrode with small partic les (nano-size 8YSZ or vibratory-milled ESB20 particles) in order to increase electrode surface area which in tu rn would increase the amount of energy required for the silver phase to migrat e over a given distance in the electrode. The addition of 5 vol% 8YSZ nano powders significan tly improved unbiased electrode stability by 97 %, and reduced the initial, zero time ASR valu e by 31 %. Similar results were obtained when YSZ-free electrodes were prepared from ESB20 powders composed of particles hundreds of nanometers in size as opposed to electrodes pr epared from ESB20 powde rs composed of micronsized particlesthe zero time ASR value was reduced by 25 %, and ASR vs. time slope during unbiased testing of the si lver-ESB20 system at 650 C was reduced by 95 %. The ASR vs. time slopes during testing under a 250 mV external applied bias were lowered by 50 % using the smaller ESB20 particles due to suppression silver phase electro-m igration. The stability of composite silver-ESB20 electrodes under an ap plied bias still needs some improvement. Improvements are likely with further reduction in ESB20 particle size down to several tens of nanometers. Also, as the operating temperature of SOFCs is reduced, the migration of the silver phase will be suppressed even further. Finally, porous composite electrodes consisti ng of BRO7 and ESB20 were synthesized and characterized using impedance spectroscopy on symmetric cells. Electrode performance was first manipulated compositionally by varying the weight percent of each phase in the composite, and a minimum ASR of 0.17 cm2 at 700 C was achieved at 56 wt% ESB20. Next, microstructural influences on el ectrode resistance were examined by varying starting particle sizes of BRO7 and ESB20 powders using combin ations of as-prepared sieved powders and

PAGE 141

141 vibro-milled powders. Comparing the two systems where the different phases have more similar grain size (BRO7VM-ESB20VM and BRO7S-ESB20S), the electrode comprised of vibro-milled particles exhibits the lower AS R, as expected. The systems composed of dissimilarly-sized particles (BRO7VM-ESB20S and BRO7S-ESB20VM) exhibit nearly the same performance, with the BRO7VM-ESB20S electrode having the lowest ASR of 0.43 cm2 at 625 C. Further ASR reductions were achieved using a combination of sedimentation to further reduce particle size and size distributions as well as ultrasonication to break up soft a gglomerates. It is clear from Figure 5-9b that the impedance spectrum of the fu lly-supernatant composite is composed of two arcs. Comparing the spectrum of this electrode w ith the other electrodes, it is the low frequency process is the most affected by these microstructu ral changes, and its resistance has been lowered to the point where the low and high frequency pr ocesses are in competition with each other. Since the impedance reduction at one frequency range was conspicuous wh ile there appeared to be no corresponding impedance rise at any other frequency range, it was concluded that the 3PB improvements provided by reduced particle size, size distribution, and phase agglomeration outweighs any possible degredati on provided by increased electrode tortuosity. The minimum ASR observed was 0.10 cm2 at 625 C for the ESB20SUP-BRO7SUP system. The effect of electrode thickness was also stud ied by applying successive coats of the electrode inks to the electrolyte substrates. For all electrodes test ed, electrode ASR dropped as thickness increased, even at thicknesses approaching 200 m. The exceedingly high ionic conductivity of the ESB20 phase minimizes concentration polarization e ffects, explaining why these effects are not observed even at such high values of electrode thickness. Lastly, ap plication of a pure BRO7 current collector was found to dramatically improve performance. Using these optimization techniques, a minimum electrode ASR of 0.73 cm2 and 0.03 cm2 was achieved at 500 C and

PAGE 142

142 700 C, respectively. This is a marked improvement over earlier results (3.47 cm2 and 0.08 cm2 at 500 C and 700 C, respectivel y) and is comparable to results reported BSCF, especially at lower temperatures due to the lower activ ation energy of BRO7-ESB20 (~1.0 eV) compared with that of BSCF (~1.2 eV), ma king it one of the lowest resist ance cathode materials reported to date at such low temperatures. Further optimi zation for these composites is likely with improved current collector adhesion, since the ESB20SUP-BRO7SUP system performed better than the ESB20S-BRO7SUP system without a current collector. In addition, compositional optimization on the optimized microstructure is incomplete at this point, since onl y 1:1 wt% ratios were explored, thus supplemental compositional studies are likely to yield lower ASR values still. Direct comparison of BRO7VM-ESBS with conventional LS M-YSZ and LSCF-GDC composite cathodes were made using a solid-state adaptation of the Luggin probe configuration to extract cathode overpotenti als as a function of current density. Results show BRO7VM-ESBS to have a significantly lower cathodic overpot ential than these conventional composites. Current-voltage testing of the optimized co mposite was done on anode-supported cells. A maximum power density of 362 mW/cm2 was attained at 650 C using a ~75 mm thick ESB/GDC bilayer electrolyte. Low-temperature performance and long-term stability testing under cell operating conditions should also be conducted to finalize th e feasibility of utilizing these materials as lower-temperature SOFC cathodes. Unfortunately since bismuth oxide-based materials tend to undergo defect ordering phenomena and/or phase transformations over much of the temperature range of interest (400 C to 600 C), care mu st be taken when analyzing these composite materials within this temperature range.

PAGE 143

143 APPENDIX A EXPERIMENTAL TECHNIQUES This section does not con tain detailed in formation about the fundamentals of each technique. Instead it focuses on experimental design, testing pr ocedure, typical results, and derivation of important parameters for SOFC systems. Some key issues will be discussed. A.1. Electrochemical Impedance Spectroscopy Electrochemical impedance spec troscopy (EIS) is a valuable tool for characterization of electrochemical processes. A small AC potential (a cross a range of frequenc ies) is applied to the sample, and the current response (impedance usi ng Ohms Law) through the sample is measured over a range of frequencies. This response is usually represented as a Nyquist or Cole-Cole plot where the real part of the impedance is the abscissa and the imaginary part of the impedance the ordinate (usually the negative and positive por tions of the imaginary axis are reversed for simplicity). The response of the cell is usually modeled in terms of equivalent circuits, i.e., a group of electrical circuit elemen ts (resistors, cap acitors, inductors) that are connected in a way that would give the same response as the cell. A common cell response feat ure (a semi-circle), and its equi valent circuit representation (a resistor and capacitor in para llel) are shown in Figure A-1. Such behavior could be characteristic, for example, of a double-layer capacitance (due to charge separation between electrode and electrolyte) in parallel with a resistance to char ge transfer or a polarization resistance. Notice that the magnitude of the im pedance decreases as frequency increases. The semi-circle is characteristic of a single time-constant. Typical impedance plots of 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. Often two time constants will overlap and the semi-circles must be deconvoluted in order to

PAGE 144

144 determine each individual cont ribution. Also it should be me ntioned that many times in the study of solid samples, the center of the semi-c ircle may be depressed below the x-axis. The equivalent circuit is similar to that in Figure A-1, but the capacitor is replaced by a so-called constant-phase element. A capacitor can be thoug ht of as a constant phase element whose phase angle (the phase difference between voltage and current responses) is 90 When this phase angle is somewhat less than this, a depressed semi -circle is observed. This behavior has been explained in a number of ways. Surface roughne ss of the electrode is one explanationfor 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 merc ury electrodes (which are atomically smooth). Another common cell response feat ure on a Nyquist plot is a st raight line with a 45 angle (Figure A-2). This feature is usually modeled by a so-ca lled Warburg impedance and is characteristic of semi-infinite diffusion. As shown in Figure A-2, in many cases at low frequencies, the plot forms an arc. This is ju stified because at high frequencies, the time for a molecule to diffuse through, for example, a porou s cathode, is much longer than the period of the applied AC stimulus, hence the molecule does not see that the cathode is of finite thickness. The response of a cell can be perfectly m odeled by a number of different equivalent circuits. Knowledge of the physic al processes occurring in each cell can help identify the most appropriate model. The model can be justified by altering a single aspect of the cell (grain size, for exampleFigure A-3)75 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 model are shown in Figure A-4. This figure will be used as an example for calculation of various cell parameters. The simplest

PAGE 145

145 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 electrol yte resistance. It should be not ed at this point that while in this case the electrolyte response behaves as a pu re resistor, in many polyc rystalline electrolytes the response may exhibit some capacitive behavi or due to the grains (bulk) and the grain boundaries, hence up to two semi-circles may appear in this region of th e Nyquist plot (as in Figure A-3). Rct is the charge transfer resistance (whi ch 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. As mentioned in Chapter 2, the speed of the charge transfer reaction can be modeled by the Butler-Volmer equation. Since in IS the applied signal is small, the overpotentia l (the electrode potential minus the equilibrium potential for the reaction) should be sma ll, and the Butler-Volmer equation becomes O ctnFi RT R (A-1) where R, T, n, and F have their usual meanings, and iO is the exchange current density. Thus if Rct is known, the exchange curren t density 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 Rct is not the only non-electrolyte re sistance reported. Polarization resistance, Rp, and the more general electrode resistance, Rel are frequently reported. However, in most all cases, the value for the resistance is measured as the difference between low and high frequency real-axis 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 resistivitie s or conductivities. For a particular resistance,

PAGE 146

146 l RA (A-2) RA l 1 (A-3) where is the resistivity, is the conductivity, A and l are the area and length over which a uniform current is carried, respectively. It is seen, for example that the ohmic contribution can be identified by performing a series of experime nts, holding all experimental conditions the same while changing electrolyte thicknessa plot of R versus electrolyte thickness should be a straight line with intercept zero. Area specific resistance is another para meter that is commonly reported, and is simply the resistance of interest multiplied by the area of interestfor example, the electrode resistance multiplied by the electrode area gives the electrode area-specific resistance [ cm2]. These properties will often show an Arrhenius relationship 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 pro cess. With this brief description of EIS, one can begin to imagine how this tec hnique can be utilized to help interpret cell behavior, as well as help determine rate-controlling processes. A.2. Current-Voltage Measurements There are a number of different ways to perform current-voltage (IV) measurements. Figure A-5 shows a common fuel cell te sting apparatus and typical results. A seal is used to isolate a nd expose the two electrodes to different oxygen potentials (high pO2 on the cathode and low pO2 on the anode). This sets up a Nernst potential across the fuel cell, as mentioned previously. Current is drawn from the cell, and the resulting cell voltage is recorded as a function of the ma gnitude of current drawn per unit of electrolyte area. Power

PAGE 147

147 density is often plotted on the same graph. Two samples can be compared by changing one element of the cell, such as the cathode, while ke eping all other aspects of the cell constant. The performance can be considered to improve if the slope of the current -voltage plot is decreased or if the power density maximum is increased. To isolate the response of the electrodes, curr ent-interruption is ofte n used. The ohmic and non-ohmic contributions of the voltage ( c) between the cathode and Ref C can be separated by the use of a fast electronic swit ch since the ohmic drop (relaxation of ionic and electronic charge carriers) is order of magnitudes faster than the non-ohmic processes (discharge of the double layer at the cathode/electrolyte interface and diffusion processes). A plot of c versus time and a deduction of the relative (between cathode and reference electrode) non-ohmic cathodic overpotential are shown in Figure A6a. This process is repeated over a range of currents and a plot of cathodic overpotential versus curre nt density is obta ined (Figure A-6b). Another technique to accomplish the same result, i.e., determination of cathodic overpotential is shown in Figure A-7. Here, the sample is not sealed, instead the process is driven by an applied voltage, E, between the working and reference electrodes. The cathodic overpotential is estimated by IREc (A-4) where I is the current and R is the ohmic resistance, as described earlier. Note that an impedance spectrometer (FRA in Figure A-7) is used to determine the value of R at each value of E applied. The same data can be obtained using a combinati on of a fuel cell test (Figure A-5) and EIS. A horizontal line drawn at the voltage intercept (current density equals zero) represents the theoretical open circuit cell voltage, E0. As the current density is increased, the cell voltage

PAGE 148

148 drops by an amount equal to E0-E(i). The ohmic contribution to this drop is simply iR, where R is determined, as before, by impedance spec troscopy. The electrode contribution to the voltage drop is then given by iRiEE)(0 (A-5) For clarity, a typical analysis is show n in Figure A-8. Murray and Barnett38 have also separated electrolyte in order to obtain cathode overpotentials using a simple symmetrical cell (a cell composed of the two electrode s of the same [cathode] material separated by the electrolyte). In their measurements, a voltage was applied and plotted against the current measured. Again, EIS was used to determine the electrolyte ohmic contributions, and these were subtracted from the applied voltage to yield ca thode overpotential versus curren t results (Figure A-9). The results can be verified by comparing the cathode polarization resistance (t he slope of the cathode overpotential at zero current) to the magnitude of the polarization resistance determined by EIS. The authors report good agreement. It should be mentioned that some authors us e these cathodic overpot ential versus current density plots (in the range of lo w overpotential) as an alternate way to extract exchange current density values. Rearrangement of the low current approximation to the Butler-Volmer equation (Eq. 2-14) gives i nF RT i 0 (A-6) where i/ is the low current slope of the current density-overpotential plot. A.3. DC Electrical Conductivity Figure A-10 shows the current and voltage Pt reference electrode arrangement for DC conductivity measurements. The samp le, in this case, is a relati vely (~90%) dense pellet of the

PAGE 149

149 electrode material composition being surveyed. A constant current is passed between points a and b, and then b and c and the resulting voltage drop is measured across d and c, or a and d, respectively. Two resistance values are then estimated by the equations ab dc cdabi V R (A-7a) bc ad cbadi V R (A-7b) Taking into account the flat, disc -shaped geometry, the conductivity of a sample of thickness d can be estimated by the equation cbad cdabRRd, ,)2ln(2 (A-8) A.4. Oxygen Exchange Measurements The resistance of mixed conducti ng cathodes is caused by severa l different processes such as the exchange of oxygen from the gas to the solid, the diffusi on of oxygen through the electrode, gas phase diffusion through the pores, and the transfer of oxygen from the electrode into the electrolyte. Oxygen exchange measuremen ts are a valuable tool in the determination of which process most greatly influences the cathod e resistance. One such experimental technique that is gaining in popularity is the isotope ex change/depth profiling method (IEDP). In this technique, dense pellets of the electrode materi al composition being surveyed are placed in a secondary ion mass spectrometer (SIMS). The samp les are first annealed in a high purity oxygen atmosphere at the measurement temperature. Afte r the initial anneal, the chamber is switched (at a time labeled t=0) to an 18O-rich atmosphere, and the sample is again annealed. At some time, t, depth profiling is performed on the sample by th e use of a 5-15 keV primary ion beam (typically Xe) at normal incidence to the sample surface, and the normalized isotopic ratio [18O/(16O+18O)]

PAGE 150

150 is measured via the SIMS instrument and plot ted as a function of depth below the sample surface. The data is then fitted to the solution to Fi cks second law of diffusion. The appropriate surface boundary condition used to solve this differential equation has been described22 as x C DCCki x i s i g (A-9) where k is the surface exchange coefficient, D is the oxygen self-diffusion coefficient, i gC, i sC, and i xC are the isotopic ratios in the gas, surface, and in the solid at depth x, respectively. From a fit of the data to the solution42 to Ficks second law, one obtains values for the parameters k and D. The magnitude of these parameters provide s valuable insight into the interpretation of oxygen reduction kinetics.24 A.5. Thermogravimetric Analysis (TGA) A TGA is used to measure the change in mass of a sample as a function of temperature. These weight changes can result from a number of different processes, such as chemical reactions and decomposition. The sample is prepared and placed on a microbalance and heated in an appropriate atmosphere. TGA is sensitiv e enough to detect weight changes on the order of a fraction of a microgram. This technique is useful for quantitative determination of oxygen nonstoichiometry, and the experimental pr ocedure has been described in detail.76 Figure A-11 shows typical results of oxygen nonstoichiometry calculations using information obtained from TGA experiments.76 A.6. Microstructure Microstructural evaluati on is of critical importance in the study of cathode materials. Many models used to describe electrode performa nce involve microstructrual parameters such as

PAGE 151

151 porosity, grain size, tortuosity, etc. In addition, it is important in many cases for a material to be single phaseboth before and afte r electrochemical testing. For example, minor impurities introduced by interdiffusion betw een cathode and electrolyte can l ead to resistive interlayer phases. These phases can lead to an alteration of diffusion and surface exchange properties of the system.77 A variety of techniques can be utilized to study microstr uctures. Scanning electron microscopy (SEM) is a powerful technique that can be used to help determine grain size, porosity, and tortuosity.77 In addition SEM can easily measure other parameters that may be of interest such as cathode thickness. X-ray diffractometry (XRD ) is another commonly employed technique. It can be used by application of Brag gs law to determine the theoretical density of a material. In addition, phase pur ity (both before and after electr ochemical testing) can be judged based on the peak spectrum it produces. A quant itative determination of surface area is often desired, and this can be done on both powder samples a nd sintered, dense samples.77 Thermal expansion coefficients can be determined by utilizing a dilatometer. It should be mentioned that ther e are many complexities involved in each technique; this is particularly true for electrochemical measuremen ts. The possible sources of experimental error are numerous and will not be discussed here. Care should be taken in the preparation of each test in order to minimize experimental error.

PAGE 152

152 R RBRPCBCdl R RBRPCBCdlA RRB+RRP+RB+R= =0 (DC) Zreal-Zim RRB+RRP+RB+R= =0 (DC) Zreal-ZimB Figure A-1. (a) Equivalent circuit model and (b ) typical impedance spectroscopy cell response. Zreal-Zim Zreal-Zim Figure A-2. Impedance response showing di ffusion behavior at high frequencies. Figure A-3 Impedance response for as-sintered (1500 C, 4 h) HS3Y samples (c ircles), annealed at 1200 C for 110 h in both 10% H2 balance N2 atmosphere (triangles), and air (diamonds) showing the effect of incr easing grain size (decr easing grain boundary length). [Reprinted from So lid State Ionics, Vol. 76, S.P.S. Badwal, Grain boundary resistivity in zirconia-based materials: effect of sintering temperatures and impurities, 67-80 (1995) with permission from Elsevier.]

PAGE 153

153 R RctCdl W R RctCdl WA Zreal-Zim RR+RctR+RctMass transfer control Kinetic control Zreal-Zim RR+RctR+RctMass transfer control Kinetic controlB Figure A-4. (a) Response and (b) eq uivalent circuit for mixed kineti c and charge transfer control. Anode Cathode Electrolyte Vcell idrawnO2-e-Low pO2High pO2 Anode Cathode Electrolyte Vcell idrawnO2-e-Low pO2High pO2A 0 0.2 0.4 0.6 0.8 1 1.2 02004006008001000 0 50 100 150 200 250 current density (A/cm2)power density (mW/cm2) cell potential (V) 0 0.2 0.4 0.6 0.8 1 1.2 02004006008001000 0 50 100 150 200 250 current density (A/cm2)power density (mW/cm2) cell potential (V)B Figure A-5. (a) Fuel cell testing schematic and (b) illustrative re presentation of typical response. A B Figure A-6. (a) Typical current-i nterruption response, and (b) calculated experimental cathodic overpotentials. [Reprinted from Solid Stat e Ionics Vols. 86-88, M. Gdickemeier, K. Sasaki, L. J. Gauckler and I. Riess, Per ovskite cathodes for solid oxide fuel cells based on ceria electrolytes, 691-701 (1996) with permission from Elsevier.]

PAGE 154

154 Anode Cathode O2-Low pO2High pO2Electrolyte Ereference electrode Potentiostat(1287) FRA (1260) Potentiostat(1287) FRA (1260) Anode Cathode O2-Low pO2High pO2Electrolyte Anode Cathode O2-Low pO2High pO2Electrolyte Ereference electrode Potentiostat(1287) FRA (1260) Potentiostat(1287) FRA (1260) Figure A-7. Alternative electrochemical testing setup for cathodic overpotential. 0 0.2 0.4 0.6 0.8 1 1.2 0200400600800current density (A/cm2)cell potential (V) iR 0 0.2 0.4 0.6 0.8 1 1.2 0200400600800current density (A/cm2)cell potential (V) iR Figure A-8. Illustrative representa tion of separation of electrode contribution from fuel cell test response.

PAGE 155

155 A B Figure A-9. Alternative determination of electrode polarization from symmetrical cell I-V measurements. [Reprinted from Solid Stat e Ionics, Vol. 143, E. Perry Murray, S.A. Barnett, (La,Sr)MnO3-(Ce,Gd)O2-x composite cathodes for so lid oxide fuel cells, 265273 (2001) with permission from Elsevier.] a b c d I(+) V(-) V(+) I(-) a b c d V(-) V(+) I(-) I(+) a b c d I(+) V(-) V(+) I(-) a b c d V(-) V(+) I(-) I(+) Figure A-10. Electrode arrangement for DC conductivity measurements. Figure A-11. Oxygen nonstoichiometry data for vari ous Co and Fe B-site dopant concentrations. [Reprinted from the Ph. D. Dissertati on of J.E. ten Elshof, Dense inorganic membranes: studies on transport properties, defect chemistry, and catalytic behavior, (1997) with permission from the author and The Universite of Twente.]

PAGE 156

156 LIST OF REFERENCES 1. J.J. MacKenzie, Oil as a Finite Resource: When is Global Production Likely to Peak, World Resource Inst., Wa shington, D.C., (2000). 2. Advanced Technologies & Fuel Efficiency, http://www.fueleconomy.gov/feg/atv.shtml 3. W.R. Grove, Philosophical Magazine and Journal of Science 14, 127 (1839). 4. B. Weins, The Future of Fuel Cells, http://www.benwiens.com/energy4.html, Ben Wiens Energy Science Inc., (2002) and references therein. 5. S.S. Penner, A.J. Appleby, B.S. Baker, J.L. Ba tes, L.B. Buss, W.J. Dollard, P.J. Fanis, E.A. Gillis, J.A. Gunsher, A. Khandkar, M. Krumpelt J.B. OSullivan, G. Runte, R.F. Savinell, J.R. Selman, D.A. Shores, P. Turman, Energy 20 (5), 331 (1995). 6. Business Wire news article Siemens Wes tinghouse's 100 kW SOFC System Completes Operating Period, Jan 5, 2001, Gale Group, Farmington Hills, MI (2001). 7. Z. Wu, M. Liu, J. Am. Ceram. Soc. 81 (5), 1215 (1998). 8. R. Doshi, V.L. Richards, J.D. Carter, X. Wang, M. Krumpelt, J. Electrochem. Soc. 146 (4), 1273 (1999). 9. T. Tsai, S.A. Barnett, Solid State Ionics 98, 191 (1997). 10. A. Tarancn, S. J. Skinner, R. J. Chater, F. Hernndez-Ramrez, J.A. Kilner, J. Mater. Chem., 17, 3175 (2007). 11. R.C. Weast, CRC Handbook of Chemistry and Physics, 48th Edition, The Chemical Rubber Co., Cleveland, Ohio, D-110 (1967-1968). 12. D.V. Ragone, Thermodynamics of Materials, Volume I, John Wiley & Sons, Inc., New York, 129 (1995). 13. A.V. Virkar, J. Chen, C. W. Tanner, J. W. Kim, Solid State Ionics 131 (1-2), 189 (2000). 14. B. C. H. Steele, Solid State Ionics 12, 391 (1984). 15. N. Jiang, E.D. Wachsman, J. Am. Ceram. Soc. 82 (11), 3057 (1999). 16. M.W. den Otter, Ph. D Dissertation, A study of oxygen transport in mixed conducting oxides using isotopic exchange and conductivity relaxation, University of Enschede (2000). 17. S. de Souza, S.J. Visco, L.C. De Jonghe, Solid State Ionics 98, 57 (1997). 18. T. Ishihara, K. Sato, Y. Takita, J. Am. Ceram. Soc. 79 [4], 913 (1996).

PAGE 157

157 19. J.M. Ralph, J.T. Vaughey, M. Krumpelt, Solid Oxide Fuel Cells VII : Electrochemical Society Series 2001 (16), 466 (2001). 20. H. Tanabe, S. Fukushima, Electrochimica Acta. 31, 801 (1986). 21. A.A. Yaremchenko, V.V. Kharton, E.N. Naumovich, A.A. Tonoyan, Mat. Res. Bull. 35 (4), 515 (2000). 22. B.C.H. Steele, Solid State Ionics 75, 157 (1995). 23. J.R. Jurado, C. Mourie, P. Duran, N. Valverde, Solid State Ionics 28-30, 518 (1988). 24. B.C.H. Steele, Solid State Ionics 134, 3 (2000). 25. S. Wang, T. Kobayashi, M. Dokiya, T. Hashimoto, J. Electrochem. Soc., 147 (10), 3606 (2000). 26. C. Xia, W. Rauch, F. Chen, M. Liu, Solid State Ionics, 149 (1-2), 11 (2002). 27. T. Horita, K. Yamaji, N. Sakai, Y. Xiong, T. Kato, H. Yokokawa, T. Kawada, Journal of Power Sources 106, 224 (2002). 28. K. Sasaki, J. Tamura, H. Hosoda, T. N. Lan, K. Yasumoto, M. Dokiya, Solid State Ionics, 148 (3-4), 551 (2002). 29. S. H. Chan, C. F. Low and O. L. Ding, Journal of Power Sources, 103 (2), 188 (2002). 30. T. Horita, K. Yamaji, N. Sakai, H. Yokokawa, A. Weber and E. Ivers-Tiffe, Electrochimica Acta, 46 (12), 1837 (2001). 31. T. Ishihara, T. Kudo, H. Matsuda, Y. Takita, J. Electrochem. Soc., 142 (5), 1519 (1995). 32. M. Mogensen and S. Skaarup, Solid State Ionics, 86-88 (2), 1151 (1996). 33. M. Gdickemeier, K. Sasaki, L. J. Gauckler and I. Riess, Solid State Ionics, 86-88 (2), 691 (1996). 34. S.H. Chan, K.A. Khor, Z.T. Xia, Journal of Power Sources 93, 130 (2001). 35. L. Rormark, Kjell Wiik, S. Stolen, T. Grande, J. Mater. Chem., 12, 1058 (2002). 36. J. Van Herle, A.J. McEvoy, K. Ravindranathan Thampi, Electrochimica Acta 39 (11-12), 1675 (1994). 37. Z. Wu, M. Liu, Solid State Ionics 93, 65 (1997). 38. E. Perry Murray, S.A. Barnett, Solid State Ionics 143, 265 (2001). 39. C. Xia, Y. Zhang and M. Liu, Appl. Phys. Lett. 82 (6), 901 (2003), and references therein.

PAGE 158

158 40. H. Ullmann, N. Trofimenko, F. Tietz, D. Stver, A. Ahmad-Khanlou, Solid State Ionics 138, 79 (2000). 41. E. Maguire, B. Gharbage, F.M.B. Marques, J.A. Labrincha, Solid State Ionics 127, 329 (2000). 42. J.D. Sirman, J.A. Kilner, J. Electrochem. Soc. 143 (10), L229 (1996). 43. R.T. Baker, I.S. Metcalfe, Appl. Catal. A 126, 297 (1995). 44. B.C.H. Steele, Journal of Power Sources 49, 1 (1994). 45. V.V. Kharton, E.N. Naumovich, V.V. Samokhval, Solid State Ionics 99, 269 (1997). 46. Z. Shao, S. M. Haile, Nature, 431, 170 (2004). 47. B.C.H. Steele, K.M. Hori, S. Uchino, Solid State Ionics 135, 445 (2000). 48. W. Preis, E. Bucher, W. Sitte, Journal of Power Sources 106, 116 (2002). 49. H. Uchida, S. Arisaka, M. Watanabe, J. Electrochem. Soc. 149 (1), A13 (2002). 50. H. Uchida, S. Arisaka, M. Watanabe, Solid State Ionics 135, 347 (2000). 51. M. Sahibzada, S.J. Benson, R.A. Rudkin, J.A. Kilner, Solid State Ionics 113-115, 285 (1998). 52. H. Tu, U. Stimming, J. Power Sources 127, 284 (2004). 53. E. Wachsman, J. European Ceram. Soc. 24, 1281 (2004). 54. P. Shuk, H.D. Wiemhofer, U. Gu th, W. Gopel, M. Greenblatt, Solid State Ionics 89, 179 (1996). 55. K.Z. Fung, A.V. Virkar, D.L. Drobeck, J. Am. Ceram. Soc. 77 (6), 1638 (1994). 56. K.Z. Fung, A.V. Virkar, J. Am. Ceram. Soc. 74 (8), 1970 (1991). 57. A. Jaiswal, E.D. Wachsman, Solid State Ionics 177, 677 (2006). 58. M. Camaratta, E.D. Wachsman, Electrochemical Chemical Society Transactions 1 (7), 279 (2006). 59. M. Camaratta, E.D. Wachsman, Solid State Ionics, 178 (19-20), 1242 (2007). 60. P. R. Rios and G. S Fonseca, Scripta Materialia 50, 71 (2004). 61. D.W. Jung, K. Duncan, E.D. Wachsman, Electrochemical Chemical Society Transactions 1 (7), 63 (2006).

PAGE 159

159 62. M. Camaratta, E.D. Wachsman, Solid State Ionics, 178 (23-24), 1411 (2007). 63. M. Hrovat, J. Holc, D. Kolar, Solid State Ionics, 68 (1-2), 99 (1994). 64. C. Abate, K. Duncan, V. Esposito, E. Traversa, E. D. Wachsman, Electrochem. Soc. Trans. 1 (7), 255 (2006). 65. K. S. Lee, D. K. Seo, M. H. Whangbo, J. Solid State Chem. 131 (2), 405 (1997). 66. A. Jaiswal, C.T. Hu, E.D. Wachsman, J. Electrochem. Soc. 154 (10), B1088 (2007). 67. S. Gallini, M. Hnsel, T. Norby, M. T. Colomer, J. R. Jurado, Solid State Ionics 162, 167 (2003). 68. Z. Zhong, Electrochemical and Solid-State Letters, 9 (4), A215 (2006). 69. B.A. Boukamp, K.J. de Vries, A.J. Burggraaf, Non-Stoichiometric Compounds Surfaces, Grain Boundaries and Structural Defects, 299 (1989). 70. J.C. Boivin, C. Pirovano, G. Nowogrocki, G. Mairesse, Ph. Labrune, G. Lagrange, Solid State Ionics 113-115, 639 (1998). 71. B.A. Boukamp, Solid State Ionics 136-137, 75 (2000). 72. M. Dumelie, G. Nowogrocki, J.C. Boivin, Solid State Ionics 28-30, 524 (1988). 73. I.C. Vinke, K. Seshan, B.A. Boukamp, K.J. de Vries, A.J. Burggraaf, Solid State Ionics 34, 235 (1989). 74. G. Hsieh, T.O. Mason, E.J. Garboczi, L.R. Pederson, Solid State Ionics 96, 152 (1997). 75. S.P.S. Badwal, Solid State Ionics 76, 67-80 (1995). 76. J.E. ten Elshof, Ph. D Dissertation, Dense inorganic membranes: Studies on transport properties, defect chemis try, and catalytic behavior, University of Enschede (1997). 77. S.B. Adler, Solid State Ionics 111, 125 (1998).

PAGE 160

160 BIOGRAPHICAL SKETCH Matthew Cam aratta was born in Lakewood, New Jersey on October 14th, 1975. He grew up in Madison, Connecticut, where he gained an appreciation for outdoor activities including ice skating, hiking, and climbing. In 1987 he move d to Huntsville, Alabama with his family, and later to Pensacola, Florida, where he attend ed the International Baccalaureate program at Pensacola High School. There he made several strong friendships including Chess Club friends Eric Ray and Fletcher Thomas, as well as future best man, Robert Van Hoose. In 1994, Matthew began attending the University of Florida. He was followed one y ear later by his friend Robert. The two quickly formed a band and dubbed themselv es Baker Act after a particular Florida law, and together with a talented horn sect ion played over 50 live pe rformances throughout Florida. In 1997 Matthew bega n dating his future wife, Tonya Bervaldi, and in 1998 they traveled the countryside. During this time, Matth ews awareness of and se nsitivity to his actions and the ramifications they have on his surroundin gs blossomed and led him to join Dr. Eric Wachsmans solid oxide fuel cell research gr oup. The wealth of know ledge, experience, and friendships he formed during these y ears was immeasurable. On August 24th, 2002, Matthew and Tonya were married. Five years later thei r bond continues to streng then due to their many shared ideologies and love for the outdoors. In all matters, both strive to keep a sense of urgency and curiosity. He will receive his Ph. D. in ma terials science and engineering in December of 2007.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101108_AAAAAN INGEST_TIME 2010-11-08T08:41:43Z PACKAGE UFE0021667_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 25271604 DFID F20101108_AAAKKF ORIGIN DEPOSITOR PATH camaratta_m_Page_153.tif GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
0f1c30b4bb2b8487eee22d0e3a0937eb
SHA-1
6455a11d99763a88418745d8003ef39bc68a6f1c
29494 F20101108_AAAKJQ camaratta_m_Page_065.QC.jpg
a7aff8c00e1d5beabad53c49ce4cf9d3
ceb98a1a8273ffcb926f6399e846e1ab0249a50a
40343 F20101108_AAAKKG camaratta_m_Page_151.pro
b2bbbfa6891a142161f7427fbe71531b
2ffb2e042f01b5250e6648c947e2dbc61a17f997
2154 F20101108_AAAKJR camaratta_m_Page_109.txt
1b39beceaf33a40c6a24b824df6d3a19
c4ef423ed1fdae4779062a6f7f6943118a10178a
55326 F20101108_AAAKKH camaratta_m_Page_069.pro
f3f917c154892f831cea646f2b39a1f5
7e8e304261eb59507ea1c50a958a7d243c77d5ae
6167 F20101108_AAAKJS camaratta_m_Page_142thm.jpg
104ef4e238369d573fb5119e9b480bfa
d226dcd5ac7e629dda06e38c5907aec4e444f1ae
28460 F20101108_AAAKKI camaratta_m_Page_039.QC.jpg
99cebb30e5c7ca841311ae152e312623
befb9b44996cab2efed21876d79ccf60d2bb5142
F20101108_AAAKJT camaratta_m_Page_041.tif
41787ccb093b246e4f736ce0aafc7b92
203b29709d769e36dd703530d4f0b3da5e95f89e
378 F20101108_AAAKKJ camaratta_m_Page_126.txt
3b9510066d9d71dd4afd654815b5261e
cc03b4799229fb78e37db52c0e731a69e2189cd8
455 F20101108_AAAKJU camaratta_m_Page_129.txt
6fd5db1d96feb302aa47b87f7bdbaf5c
bf535e01b62079ba089b9a26c3461cb4b9cbf671
4794 F20101108_AAAKKK camaratta_m_Page_092thm.jpg
e7721a4e2e0256772d3c15c06250bae6
50b0bba3ee3cbf5510956b252184ec9381799550
6490 F20101108_AAAKJV camaratta_m_Page_109thm.jpg
22a76c712a95cefeab070e95cc13d100
633aa6e88c65a512e9101b2c98a4b2f0a61f29aa
30793 F20101108_AAAKKL camaratta_m_Page_120.jp2
4be690e86b2cdb43526902f076ded690
941023017c47fbe3ca1fb70817fe4b698bedd00f
24905 F20101108_AAAKJW camaratta_m_Page_064.QC.jpg
94818ea44986d50d91de349acb93129d
1799c4d5d1dcad8c63ffb84036559d7088629819
F20101108_AAAKLA camaratta_m_Page_036.tif
9b2ddc7b5f575b8701087f52b54663fe
98734f3ba482d0f8b489a21095c3d3eb9e5ec60c
2121 F20101108_AAAKJX camaratta_m_Page_026.txt
02f43148786c0ff8e79e3e8bd0582409
10bdfc41fc9475430b7f44323c4e037b80e14909
4021 F20101108_AAAKLB camaratta_m_Page_075.pro
d9f5ef4051bcc0dd22572ddb31966f3a
48287ac189f5ddfcbb26d39dfc36f9ffb63ace45
27766 F20101108_AAAKKM camaratta_m_Page_036.QC.jpg
cd38a8cfc0c62f1e35101c620d558c0d
e564d74a257885d4ed5f14323b8fbab39bb3908f
29889 F20101108_AAAKJY camaratta_m_Page_029.QC.jpg
5b1d10eb30c75a9d9418d01d0d144897
12437a7bcca5e9234ef448419fc0a942ce519900
7087 F20101108_AAAKLC camaratta_m_Page_102thm.jpg
8a46d48d82370d11d070e9c76d5c048f
7c297f4a3b7551075b9bc9c3dde50dee52ebc248
14650 F20101108_AAAKKN camaratta_m_Page_135.QC.jpg
fcd0d0436b8b3e482cb0a8e602ce48ae
844b567d0387d484b9554bfeaec8e1057f4be729
784 F20101108_AAAKJZ camaratta_m_Page_131.txt
ae295c8309c524b9813d788792e6e549
ac5e8b5b45c0863e081a6ff5e4cfd36b9d3b9bac
8899 F20101108_AAAKLD camaratta_m_Page_129.pro
94d868b7de7ee8f894935e8a82e6ff4e
fd0bf084c47dd290fd2cd79bb4d032d6aa0586fc
F20101108_AAAKKO camaratta_m_Page_141.tif
1f5b52a5c3b72c0eb69fd884cb3c8c66
4ffa3bccdda3e18bb45d7f481d96a52941b857b3
1051803 F20101108_AAAKLE camaratta_m_Page_137.jp2
d6c817915f6d724ed792d20f216d0c5b
1c6573aa39151197864aa222d45217405e86cb68
F20101108_AAAKKP camaratta_m_Page_140.tif
6739bd4c90d9bba3d61a750b98d17082
885be492a92e7c395db4c5f1d995fb2bee3725e8
43736 F20101108_AAAKLF camaratta_m_Page_119.jpg
5c7d26d7661cdf8f74e565113f15e9c6
368cf3d6c0dfd050b150e2a075f8f7aebce72cd0
1051984 F20101108_AAAKKQ camaratta_m_Page_084.jp2
ca3740dd4d8862cffd176b3dfd54429f
5852b736c3718d2fba34fb4936ee291273af7e3b
F20101108_AAAKLG camaratta_m_Page_145.tif
8a49d977744661cb160eb3d4782e2650
04a0b979c51605cd8a6ea7c223c7c5bd26e3d5ad
16727 F20101108_AAAKKR camaratta_m_Page_095.pro
998894a18cf3debd74d8490f012e8192
9660cbec0237ab09e0cc7f3275387f8bf26b95f4
27698 F20101108_AAAKLH camaratta_m_Page_016.QC.jpg
3d49074d97b75581780a2a39a58a9a07
8a11ab9a247bd25eb343b77772004f7e08dbb9f0
316921 F20101108_AAAKKS camaratta_m_Page_134.jp2
3af45e11f7e63b2b05d656511fa0b640
0bc00d40615ba334f8d4208126cbee0049dfe897
90146 F20101108_AAAKLI camaratta_m_Page_109.jpg
6d8753460f4450f4abf476338b8cbea3
39c5c9b5b14c9eebac721c3df5d561d30c6f610a
F20101108_AAAKKT camaratta_m_Page_159.tif
5507133f41d601b704fc3aeb56e87888
9ea261a3203e88051da3e2bd56147a0553305d73
58888 F20101108_AAAKLJ camaratta_m_Page_031.pro
417dbdfcb1d18dc50fa2bd18f55eaf16
d536ef8ed795d5269cf215af6b42a3b3dee20ee8
55754 F20101108_AAAKKU camaratta_m_Page_089.jpg
3cbeefda187d5ce7839bbb7dccb36582
98434691817e9edca4c4fe557eb0573e5eccd7b9
52066 F20101108_AAAKLK camaratta_m_Page_085.pro
f83f5b72fb1d1efc261c6c67e4bb2187
5c54123ca9382677f1bb7f5354fcb0418782d2c3
6470 F20101108_AAAKKV camaratta_m_Page_061thm.jpg
491f25e70948898653100d39e419737a
ba89dca75d31a13c7570a3c639bad00aa052eb52
26820 F20101108_AAAKLL camaratta_m_Page_117.QC.jpg
09ccba67928ff7b4054a12512b53e0c2
c4edd19fd265b94b4ed472d7be07d963c803e3ec
4436 F20101108_AAAKKW camaratta_m_Page_123.pro
196ce2344437a4d8bf216374548cd493
aed9656a4c328407d3cf6f4ae0af55138c610c63
791630 F20101108_AAAKLM camaratta_m_Page_093.jp2
f04a85961b6aaa803493b3349134e7a9
d874f72e848157aab95dfc11bf2e7d2c29032764
24298 F20101108_AAAKKX camaratta_m_Page_146.QC.jpg
b09b4c08eece93ad7b78b9297b46347d
fa60048cae82c2d5623ec5c87dfea3c27e540f0b
6636 F20101108_AAAKMA camaratta_m_Page_116thm.jpg
a80aa9d4ba89aea6f98a1ce9e6816446
52b7060d3e6f03caf63e23f003e1cecd7ea7b007
27490 F20101108_AAAKKY camaratta_m_Page_042.QC.jpg
352735b8ccc10a9d637d0f230522ad5c
47d3380fc3a4ef06ca54f131b47b172daf8c9150
8719 F20101108_AAAKMB camaratta_m_Page_019.QC.jpg
c895c28833ef312e24f18fab3f227259
c2eea914c50f165687df590ecb75e4ec7b2e8f88
6427 F20101108_AAAKLN camaratta_m_Page_036thm.jpg
6593577e4c46a14e109eaea8231e2ce5
01d26fab970898bc00209ff2893ab6b361f8ebe8
29557 F20101108_AAAKKZ camaratta_m_Page_106.QC.jpg
1694446c15e5250c7b697f9276a572b8
87ce222f37597bd72a94bdf60e51466f915e3a2e
7840 F20101108_AAAKMC camaratta_m_Page_090.pro
7dc3f64b3b1950f9a6e1d61900011989
9af65e88692c7f39e08f5324ddf6658a444dfdc3
25751 F20101108_AAAKLO camaratta_m_Page_038.QC.jpg
866c654922751083935ba5f4520f105b
3ea555115338f2fc7cf31a0eafd59564fbf37de3
53221 F20101108_AAAKMD camaratta_m_Page_144.pro
f19f06bef7c007964f147be076cb18b8
a5529a0bacfc1687ee694aad7c9a12cb8fbce5a4
3790 F20101108_AAAKLP camaratta_m_Page_002.jpg
4da5b7b0efde631f684648f10058c74e
2cf15099babb56ce03e4362944bb98d227aa723f
1795 F20101108_AAAKME camaratta_m_Page_021.txt
e4b97852e6bbaa23d87cb38532822bba
eedbdb03e54561e8ef2d90c33718ce3a93a5831f
2011 F20101108_AAAKLQ camaratta_m_Page_116.txt
fad63ea87f00f3648f61a55e48158a69
a9fbd20bef01f60342c45f161ed9aeca985b353c
82966 F20101108_AAAKMF camaratta_m_Page_145.jpg
0831f0a62976ef8b530b206714f7ac5e
3f3ffc6238b649730f9ac07009d5d158f4b91d61
F20101108_AAAKLR camaratta_m_Page_017.tif
63b88845ce9d6b301a25fc1691c18927
4d8221f96195f32616be53e6b0a07342d0f4c182
1051925 F20101108_AAAKMG camaratta_m_Page_008.jp2
538e5670bd828e27fc7355eba146d2d0
0f0e7347803a295ee9bd55765bc91104891fc4f4
46913 F20101108_AAAKLS camaratta_m_Page_152.jpg
148f3fdad585349d236d75d94dd3f2f5
c0c3c9ec60a387d874a4efcb96d1fb8054ce7ed9
52564 F20101108_AAAKMH camaratta_m_Page_138.pro
0a9e309bdc6b7301451d98bcd0f24828
502140789f3c23522c906be33ea345d08f0f8d57
53268 F20101108_AAAKLT camaratta_m_Page_042.pro
f5fc40398c8373f5029ad5c096fbeb88
c469e85c7ce5dc668e462ef825bfc3c2eeaf348f
15814 F20101108_AAAKMI camaratta_m_Page_131.pro
cc0419a9b7735bc137a54f0fc4938ea7
06ee50601607e4379dc565444c92e1ff39d79069
47278 F20101108_AAAKLU camaratta_m_Page_024.pro
bbeb4d5f76e3dc836570acf34ae9fd49
cbcccc4a63d7a042988eb99677854031402d3386
898849 F20101108_AAAKMJ camaratta_m_Page_057.jp2
7251a736dd10bb72963c147ee697c5be
20eabf3bed9af33eeaf20ee1956d28570fdc4e0c
57480 F20101108_AAAKLV camaratta_m_Page_058.jpg
60a9a95791e27043a26df28222b16497
1aba3f115296be159ff2ac56acbfa2220e3e916f
1051972 F20101108_AAAKMK camaratta_m_Page_010.jp2
2664f095e2269281dc7bced5fc88b94f
d61b12b3931990c52f565df6b0c438d8be8c2214
2088 F20101108_AAAKLW camaratta_m_Page_107.txt
57c4d75451335c641dafc047611a8457
3d20c5fbd8ae40b3968525f72abeee428e7639ac
900 F20101108_AAAKML camaratta_m_Page_054.txt
8f6439b5fb240a6e8151d7cfcd0f3e9c
140daf22cb2995a78cbc46cbe3a8907425657c39
366505 F20101108_AAAKLX camaratta_m_Page_077.jp2
6b5fb9cac67ed138110c04987cd49049
242f349665197c270d561075c3a9f7cbe2ba80c8
21365 F20101108_AAAKNA camaratta_m_Page_020.QC.jpg
129ce332aaff3b6c930b36062a89dd46
53bd296339550afdc510b10a3953d70604705467
F20101108_AAAKMM camaratta_m_Page_098.tif
f7221109d80a1e23d807c7ec29cc69b3
71f755a3517790757204a5d75e81af6fb90eceb9
11677 F20101108_AAAKLY camaratta_m_Page_126.QC.jpg
9c6cf9002136d9576d4dc568c6a89087
3f546d1125f141defa9ecac22cee086ba46bab43
53143 F20101108_AAAKNB camaratta_m_Page_016.pro
361ece9f08b682075f3b9921cf17dd08
a2a7c6470fed2a2140269fbda2e5bfaf505cad8f
1687 F20101108_AAAKMN camaratta_m_Page_103.txt
27c592fd5fd11110b0aac279b3901fa3
68f5223e08b2e2cd75f2c14b6388d50508f94156
881664 F20101108_AAAKLZ camaratta_m_Page_044.jp2
2e588d7250bc4471964ba94bc60d0fcb
e4175eeeb0ad1a1314e3eafe857a37fe4e485484
12401 F20101108_AAAKNC camaratta_m_Page_129.QC.jpg
8567929924c902ce30dca0f42b80847d
f00d0772dc61e5540a1e202952f08dafff20b79d
26558 F20101108_AAAKND camaratta_m_Page_145.QC.jpg
062561071ab6c4ad8b9a69b971d6d054
00fd50de39335e6991433ac1b35de0ae397dcd61
1051978 F20101108_AAAKMO camaratta_m_Page_041.jp2
6d80e176b6d21848dc5dd5afe309552a
61ca10e639d99962f002f60e2db82e2886484fe4
F20101108_AAAKNE camaratta_m_Page_020.tif
6841ebee385814c91b9f997abdd4130b
a7a1cf118abf09fa4f40a4ae1f7d8da1f53289c6
16758 F20101108_AAAKMP camaratta_m_Page_057.QC.jpg
5357b2a489abb4c46ab66b3b712fc376
f80550999cdf99b47a459db8a72ce0a52d24aca3
F20101108_AAAKNF camaratta_m_Page_043.tif
982043b9dda686f234afd31dba5f74b4
50dbee80e0f712c4af437c344ae6561ca4e1043e
47606 F20101108_AAAKMQ camaratta_m_Page_038.pro
50af7a7735be7342845f2e2190c28c30
303a05e19da6af98a40d9c7d5af85e25dfb32292
1020983 F20101108_AAAKNG camaratta_m_Page_114.jp2
6b8f3b26e7704cc6d0c70803d7bb4e8d
cdc76f3c566cca68ce09611e37f7cd5753205f08
6558 F20101108_AAAKMR camaratta_m_Page_035thm.jpg
6b3883cf1ba09b27fb789f151248aad8
566f0dda82fc711069eb71288bb94f8193e6d590
26490 F20101108_AAAKNH camaratta_m_Page_023.QC.jpg
5023fd28f8be95c1a8054c12c65a1f4f
52684125468cc5fc0dfddb75843c9e4c87ab445b
6237 F20101108_AAAKMS camaratta_m_Page_110thm.jpg
8ba270c9be31e51c0ad5d8f2128b936c
dbf159bdecc904a301f84ac747d43e6893abbd0b
22597 F20101108_AAAKNI camaratta_m_Page_160.QC.jpg
fa478ba8fd9f1a0d5e86f2fb067ed87a
d5eab65a3b1254f856615feb7f679a74394b4d41
F20101108_AAAKMT camaratta_m_Page_093.tif
1dfd1a843fd0a676aa0c4397dadaed9d
b734924c43782fe754bd013e6a9aa4d249ed9e4e
2175 F20101108_AAAKNJ camaratta_m_Page_039.txt
40d0f0d25ccc01e0b8778c13878e8c54
96a31a37e2081c5a34fe1d65efb79f40a3c7c462
18727 F20101108_AAAKMU camaratta_m_Page_153.QC.jpg
ba015a596255a96d05e158b93cf22bc8
ca09bb03f456635520ade806cbe71e2722309235
28962 F20101108_AAAKNK camaratta_m_Page_069.QC.jpg
bf29c94ebaa7497c708d3ef1fe2b75c8
4493d30005a66ea537bad68cbcd0dad58894594f
57278 F20101108_AAAKMV camaratta_m_Page_041.pro
73635e3b03a20eb0332643bc26790aed
050cb38d5f73f2f9f06967e514c20be8dbff0a8c
4180 F20101108_AAAKNL camaratta_m_Page_125thm.jpg
f1f502b932b1b1ea16dbe01eb5e80b87
45d5d3d205f79eae8ac4bd3d62a0cf31341de429
20309 F20101108_AAAKMW camaratta_m_Page_098.QC.jpg
06a83f53c835fa91311c28ef4109bdc2
4a9e156eae1267ec8b7d73dbe5ccba6de2fe5070
6382 F20101108_AAAKOA camaratta_m_Page_014thm.jpg
8de4fe881cdcfa5dd9da3969b7d2f7d3
67b324770dd5fbd101d09f1ce427c50b74576773
27866 F20101108_AAAKNM camaratta_m_Page_082.QC.jpg
e8dd7119cc39cddde60355f63559ca14
263f86cfcdf1ee95fef1f2f3492baa3776fcccb6
643 F20101108_AAAKMX camaratta_m_Page_133.txt
59bfc5b7d3f8599073d242ed8f93270e
e54260765057ce90c66b23addd3fafcce5b285a3
87687 F20101108_AAAKOB camaratta_m_Page_144.jpg
ec4a52101f0a43df893131654b62d961
e9d736a61a07a22ffef30675ce35d92dafdb1804
12421 F20101108_AAAKNN camaratta_m_Page_090.QC.jpg
4a12efde45970432fe948afd7da64c00
cc7075d1c58143df077756af87a48593664dcb57
42786 F20101108_AAAKMY camaratta_m_Page_135.jpg
d944b01d81e504601ae8bb81b732587a
3ad2758ed9452fab3cfb7796ea3cfdc56ddeb7b7
6859 F20101108_AAAKOC camaratta_m_Page_112thm.jpg
0149e9227567eac6ed7893db666d0fbb
896111b4ad411c9176bf086e6d8d90afc6d7969f
60018 F20101108_AAAKNO camaratta_m_Page_059.jpg
7bb1b5a1ebca244a088492558b13303c
0fa8b8d479189136105320bf4a259cd8cef9b041
F20101108_AAAKMZ camaratta_m_Page_129.tif
f318d029f14c9e210fe23c2fe675f7ec
18351e47af60432fb5682e63c1f65b0eb37abc4d
F20101108_AAAKOD camaratta_m_Page_077.tif
8139816fb55fb2c3b4baf677b583911a
bd160dd39bcd977be1b0463d50cc6acbeb0bf765
5177 F20101108_AAAKOE camaratta_m_Page_104thm.jpg
d4d554e6d4154d89234c2d11a88fd53f
c61499ac9eea0bc073cb2304c861ee46bdcf412d
F20101108_AAAKNP camaratta_m_Page_024.tif
60ec5875897866482a10b2cf61951720
fde95e43f3ea635e8d0be6b206cc8201ac7db1d8
7567 F20101108_AAAKOF camaratta_m_Page_091.pro
c8ecb69ac2e00bc8db75ff4c2cbda7ae
6ae608fad71fea2d09300bbd68d0b4d6d2e35dca
62584 F20101108_AAAKNQ camaratta_m_Page_093.jpg
238dbfe42f7057ed0891c054ead8456c
7f125457a4eaddda10f7583131d056a20640ad9d
25840 F20101108_AAAKOG camaratta_m_Page_099.QC.jpg
d0295e293560c1d0d4694fbbeb95ac47
427afb8e06eed063965aadf1ff6ec0cbb8423c30
40594 F20101108_AAAKNR camaratta_m_Page_053.jpg
3c9f4b8b35bacef53316f070a3f491c2
e4785d65846b8f00ffe8def97b4794c2492c663e
6487 F20101108_AAAKOH camaratta_m_Page_032thm.jpg
f241124ae15eb1b5c25bd37742ecef51
8bc4a58b5348ac05406c0961a5209232bc71a351
8215 F20101108_AAAKNS camaratta_m_Page_118.QC.jpg
48835172a7983538bc46c875bdadc3dc
4d9fd3d9871e0e9fed21a4ddcee85121173adb27
60000 F20101108_AAAKOI camaratta_m_Page_054.jpg
adce0784ab6497c0ecab0387f0ed772e
bab257d35e849093d158f50add61c20c0018f835
2104 F20101108_AAAKNT camaratta_m_Page_144.txt
e3d2a992873a93abf901b10e9757c43c
6e5296cfaa8e7304c2cb5c17eb3282604bea6cd3
27908 F20101108_AAAKOJ camaratta_m_Page_002.jp2
372c782f36fbf24fdbe5ff002e83a451
c04e4e9ed00f83d15023ac9ad7a4dbfcd7b7a8c2
1051975 F20101108_AAAKNU camaratta_m_Page_027.jp2
abf7bd874c0fe68de06f7c30020ab318
b832fa89efcdc234783b7f7ea7e28ba98b8c38c9
15652 F20101108_AAAKOK camaratta_m_Page_050.QC.jpg
eb8506ada4709974074ce675fab0b662
42a56470028d92e80e680111b1376100aba662bb
9003 F20101108_AAAKNV camaratta_m_Page_070.QC.jpg
2d25b905c5f32e3cae856e9810c98422
03ef086b45a0e9ce6b76558d28e5971f9789c15f
F20101108_AAAKOL camaratta_m_Page_151.tif
7b75449cab5c749a3fe34290b0e4f6d8
7f5a95b6a18e8fad6cbdb73c29af37094273284a
50139 F20101108_AAAKNW camaratta_m_Page_099.pro
57eae3232b71540faffee05b5692bdad
4e81a53943768c5af1fd3bc7f55ef043300e9730
F20101108_AAAKOM camaratta_m_Page_051.tif
a6b8803e59fcb4fb95b8cd76313867b6
9b72efedc598a5cad8cb79c61b1500b1f502cade
1909 F20101108_AAAKNX camaratta_m_Page_013.txt
4e94f74fbdc9c9b09757b11ac6166904
be8d18bbdd575c8655321e03d05443b01994fc80
49988 F20101108_AAAKPA camaratta_m_Page_133.jpg
06c83435550bf477f9f66839f8577859
e22c36ff1861b2c22d7b24fb7b9a0ee08a0d54a2
638 F20101108_AAAKON camaratta_m_Page_003thm.jpg
2a05d591cb485d8c227871f9b2f32cfa
c16f0c3673fdff3de1b2fa0aa0e564fca7705469
27969 F20101108_AAAKNY camaratta_m_Page_062.QC.jpg
06e5cb0f188a81feea8b06b0674de998
a3392aaa997db20c12f16123e44fba4f10ac40a5
8423998 F20101108_AAAKPB camaratta_m_Page_094.tif
cbd22acf6113943f55ad4b45aeac7ca8
79f120a783dbe3d2831279cec213ba0ab8578915
6798 F20101108_AAAKNZ camaratta_m_Page_030thm.jpg
8474aa158bc614dbca4032736468e197
17ef2fcfb727a94e32533c8f4e43a72bef3c3e4f
53129 F20101108_AAAKPC camaratta_m_Page_067.pro
a9a44e6175a1447ca3b43266a56bc633
917d55298078257fc8326a8735d58f0abd76ec3c
1653 F20101108_AAAKOO camaratta_m_Page_007thm.jpg
c94304def38fdcc338c2da687e63cb04
9648ec6da489c7a2609ae7b1d399338ddf97a9b9
93317 F20101108_AAAKPD camaratta_m_Page_105.jpg
08cbbac502fecc880b41b8354347b4f2
0a6eab1a1075c2c360fff674a2481eb2285d5fe5
12369 F20101108_AAAKOP camaratta_m_Page_094.QC.jpg
6020548198f2442f1b4e214b0e4a5b83
4c858f71852e10a2897670b50d2674bb31ca6288
258 F20101108_AAAKPE camaratta_m_Page_075.txt
bb37f99318471b4e196888212405f9be
5a9e17437c89dee9718c3fddd5fab25833123849
21436 F20101108_AAAKPF camaratta_m_Page_159.QC.jpg
febccdfc547a5408e42780114635d316
748a9fe145531df0cd2a6d39b9e3657610a10f0a
56534 F20101108_AAAKOQ camaratta_m_Page_025.pro
db04dde7d46b24c0d46587045b654ff9
91529e90b0fde2595defadfccc8f22beedfcb778
F20101108_AAAKPG camaratta_m_Page_081.tif
e2344d72c256d22eb62774eaf8c9df07
01233c8358610e07d8e5f21639383672555e7673
46156 F20101108_AAAKOR camaratta_m_Page_158.pro
99a49b089a79c65cce17be6df8b54836
275e8bf80095d97e08b584d2dcf3187a08347f21
2158 F20101108_AAAKPH camaratta_m_Page_101.txt
e5f72dfb8b3cc7f4229e1ab894d9ded7
7f5106ad0a04899ba1908d36d81582981e35885c
54829 F20101108_AAAKOS camaratta_m_Page_101.pro
97a1221799158e92fa58e8d507ee7d05
6705551e7bb79dc91793fa3231138b90e8ce3fcb
90911 F20101108_AAAKPI camaratta_m_Page_069.jpg
f70cb1c51c483620d8ac632b3a0cbc8d
43ded48507a2bc8faee13c3ba16a37e60845b87a
6344 F20101108_AAAKOT camaratta_m_Page_138thm.jpg
e3531499f6f13a4ccb82823db6a2e6dd
a076d16a664ffcf50e5e68da940f391412e7bf53
F20101108_AAAKPJ camaratta_m_Page_055.tif
ec6185d79e038cbce5f550d68645b337
3b162b862c95535ed77c5f9a5cbb67c6e3043ada
27687 F20101108_AAAKOU camaratta_m_Page_086.QC.jpg
42bb4021a4d267e0cf11b654ccaba1f9
ea8ce4d91e76b3dd71b47f983462de8b7733c844
55975 F20101108_AAAKPK camaratta_m_Page_057.jpg
2d06bd76f20594466ba23302b5f13adb
75c77907ea8f636064aad82d1214fc41baa80f9c
79254 F20101108_AAAKOV camaratta_m_Page_158.jpg
f56e1d940f1c317db1d894325c068d37
e591a087e8c8b24aa2dd5c041f6aa56db25eab8a
1075 F20101108_AAAKPL camaratta_m_Page_057.txt
419aa4db6203b3b1c8f9a2ac25d4bda0
afd9408d716e36f0e90f33ae7561b1bd4ec6b862
3667 F20101108_AAAKOW camaratta_m_Page_090thm.jpg
472f0c8fb0e21bfc157e97368f5dc5b0
7a6664441422947616d8d183fbc73d1647f92a66
5524 F20101108_AAAKPM camaratta_m_Page_126.pro
b47d7774a4c9267a652fdd5940e645d0
0d28143353e0d40fd3fe18aed989caf054d0ef7f
523222 F20101108_AAAKOX camaratta_m_Page_090.jp2
ad6dbd575fb23229240b8e0d69350829
8a0a243a0baac6eb79d51a894e5a24ef3101cff1
5318 F20101108_AAAKQB camaratta_m_Page_003.jpg
8b5623b1cb438ee2ed265caeb3d32a22
c90a7431b518c0f32b4f9044a6c4dd85e5560fec
47080 F20101108_AAAKPN camaratta_m_Page_114.pro
0110b3147c541bbeee58ddd401fd6639
0239e47a1e09467d35e9c4ffda435f0ac49aae2c
38403 F20101108_AAAKOY camaratta_m_Page_044.pro
93f9bd7cc3fb847cc56392165aef0482
d4263eff77096f8056e7632fafae6abf2da1982a
84191 F20101108_AAAKQC camaratta_m_Page_004.jpg
55a59ab9bfd4b5deb7973348bac5d0a5
022b46b82e349203234c3414f153cff5434b15e7
1051967 F20101108_AAAKPO camaratta_m_Page_015.jp2
d75fa5e14619f361c8d8a4c282feaa25
d4c1f92ad520593c293dd4e1293bcfb9eaf84fa3
F20101108_AAAKOZ camaratta_m_Page_047.tif
4a936235ccb11fa2e4468176df694579
fe606c21638007a254eca35433abc7efc0eb987a
77197 F20101108_AAAKQD camaratta_m_Page_006.jpg
07a30c9fd30561eb2aff73b4e7075f41
022f178e129f2ba714fb3d1b6ad3eb5a08088cbc
399737 F20101108_AAAKPP camaratta_m_Page_018.jp2
602f0a5b927f4d50833675f0ef7ad4bf
42d915230d3f36d07870168701ba3922a71265d7
22412 F20101108_AAAKQE camaratta_m_Page_007.jpg
7e6885f2c2ad68c72022dc36e5819f57
25d8bee05b440f3f97f12df7a50f54316fc19766
5826 F20101108_AAAKPQ camaratta_m_Page_076.pro
80a5c5c8851efae710311147e74e19e6
7d512948e706c4e0e77c36ced393ab6dd3a109a2
105625 F20101108_AAAKQF camaratta_m_Page_008.jpg
8489dd5a90c3acae40c1a8388edbc89a
ad277f75020f15afa0bbb7a447ef48025675aaf4
128638 F20101108_AAAKQG camaratta_m_Page_010.jpg
116860e050748f35b793984cb05785bb
c1ea998580ffd54abc8b2b14ad799371f8349ecc
2216 F20101108_AAAKPR camaratta_m_Page_015.txt
c549289074ac4957a7dd8edb88180495
931f2ee1250266c0397afc198103171edc670ece
54668 F20101108_AAAKQH camaratta_m_Page_012.jpg
f4c6c0ea2a02d9f5f292e291b35abf04
6d0c156c9a6268a494cc43352f15d260e6de3e44
33745 F20101108_AAAKPS camaratta_m_Page_010.QC.jpg
92d2e515c2ab4300428dd84f88cb9950
9dd206e4ddd07243246c73b69745d7b0e54cf3ed
86464 F20101108_AAAKQI camaratta_m_Page_014.jpg
11cc95d014c2751d55a0efa28eb89816
007a7f630e613f38ece16b34fb588432ee9c9dfa
533 F20101108_AAAKPT camaratta_m_Page_098.txt
ec658f368f80c372c2827a4d8dd1d28a
677f20ec118fc7b4e495652a9cf6d012211a32fc
90564 F20101108_AAAKQJ camaratta_m_Page_015.jpg
8b871dbbb75b2b9f9165ead6368c979a
ba1b19ebeb7a82650c1b8f74e1583d0adf3e0f93
2404 F20101108_AAAKPU camaratta_m_Page_019thm.jpg
6eb8189502e8448f2438daa67b711ac0
b602f7bac7763d972f8b36276a4357388708d6bc
88171 F20101108_AAAKQK camaratta_m_Page_016.jpg
8ec602b7711384636c18beebdcf8a784
fd203d241ccf5b0202cad5d5dde3ffacdad8ad26
46201 F20101108_AAAKPV camaratta_m_Page_083.pro
697bfa80abcedeab94a214935b5e6159
7193bdd1c50c57365741d75250ba25722371abb9
69611 F20101108_AAAKQL camaratta_m_Page_020.jpg
9616b6258b4d8c48064f469ddd77f000
559eb28df6e427071ebdf42b3752a7d9821d23b0
82776 F20101108_AAAKPW camaratta_m_Page_063.jpg
9eb5e0def6a6ab9024378cfc8827a9de
f5572335f669cf87d066166832b1585d8f0d97ed
69337 F20101108_AAAKQM camaratta_m_Page_021.jpg
d25fa1f20f1241bbc9279f0188ea98af
668e1b2585ba1c4c61212db7002f9b35146b09f9
F20101108_AAAKPX camaratta_m_Page_012.tif
0be3623404c5503b8dbddac4b315a735
a500671bf88780839ed5f7c57aaddb9fbac46220
92569 F20101108_AAAKRA camaratta_m_Page_041.jpg
7729e0437c9a1c8921e9c6777463e3c1
ccf79d1588c508848bd3acc6e5a37ae0318f417b
70546 F20101108_AAAKQN camaratta_m_Page_022.jpg
c48a5e64e75d6bd4ea2d568ebfe63c0b
3dac03d44648bba49ab540671f8287d068cdc9ae
187130 F20101108_AAAKPY UFE0021667_00001.mets FULL
f234705711f361b044c2efcdf92d1f87
dde32b29a59c8a6a5af3c00eb3786d939d603b3d
64578 F20101108_AAAKRB camaratta_m_Page_044.jpg
7d862f7a770df51443b0d22c94158595
b368a72375910bdb633321df949b21c71f39c3e9
75190 F20101108_AAAKQO camaratta_m_Page_024.jpg
346e90ab5cc674a778ae1ab245c10c70
99efb4ad87732fd4032c6993f76d6c8169f84080
39318 F20101108_AAAKRC camaratta_m_Page_045.jpg
2140a4c2408c703a5c6e67b83a693168
7054943d68c0a4182a4030b327571bae91d5a453
91320 F20101108_AAAKQP camaratta_m_Page_025.jpg
83a46e03525a2e1709e0d1f551363b8b
d1b39cc1f113375f634c7ec6a57511611bda071d
55856 F20101108_AAAKRD camaratta_m_Page_048.jpg
de40edf75b27f36bb91214ff78ffc382
5c6ae05f3167dc27bd27c7c5c49bd9f22e151967
85431 F20101108_AAAKQQ camaratta_m_Page_026.jpg
10106d022477d906597fdb2f9db25351
5f8d080eed46c7da8ea2db79afc67df8ac4ffd83
64833 F20101108_AAAKRE camaratta_m_Page_049.jpg
06cf7fbf3b955bb703a65740964b65b3
7c1ff17dc2ce71785ca17ff1ca0247663f84919a
85602 F20101108_AAAKQR camaratta_m_Page_027.jpg
d86d200bbbf25e5991c0dd4d1b7c2751
eea8ddca979c9bcdca2cc1a87a8a6525e10b7cb2
48500 F20101108_AAAKRF camaratta_m_Page_050.jpg
a9de05d198bdc17737492d78fd0818d5
54880f7eb56f30431da752f85609eaa9c94fa867
55560 F20101108_AAAKRG camaratta_m_Page_052.jpg
fe416acdec32f5695ea1892fe7d6e6b6
02625fbb33082d3254d77436e25cb2c00bbe6f17
95039 F20101108_AAAKQS camaratta_m_Page_030.jpg
b405d83c61a997a162913fa8fbae9041
f3040c05386977b56c023cb35aa9d9c7c40add2e
83914 F20101108_AAAKRH camaratta_m_Page_060.jpg
50c48c9f298dd83770804aab12eb047a
4dda6d7788a3a14c12396391f2b72a1e7d8af3d6
85624 F20101108_AAAKQT camaratta_m_Page_032.jpg
8ca87cf985e15ef48ab8336b46a3948d
45eb29bb2dc2225c5079751762694a4161bf4a25
89267 F20101108_AAAKRI camaratta_m_Page_061.jpg
b946e312b0723da62d68b1fa59d508e8
6cb35fbe7be5e4fff4449bb60bcaadb16e9f5df5
66192 F20101108_AAAKQU camaratta_m_Page_033.jpg
187ffb86feaeb26b153c0d5f5c40fc85
dd607f537c9493521c9ea76c2f921dae97dedde9
89357 F20101108_AAAKRJ camaratta_m_Page_062.jpg
ff2bb4f963f0ee1942229a4253aa46eb
3d8a8093e99dad8f0cadab8d3bcf1b31e2f0dfed
75443 F20101108_AAAKQV camaratta_m_Page_034.jpg
7ac7d18defbe434ee768751ff3ec2f6c
166e7060cc4391de7a8b8c223f8e5c02d2ae8fa8
87891 F20101108_AAAKRK camaratta_m_Page_066.jpg
6ec4313e405a06c73cf3a92b2c97c60b
ada45a01789cc7551f9713ac9ea8a884612c56d1
87805 F20101108_AAAKQW camaratta_m_Page_036.jpg
2a50f2103a53b635b56d18e99514e6cb
922b26edbf326fd9564381e3cc889f63b8659a1a
87625 F20101108_AAAKRL camaratta_m_Page_067.jpg
83847d5d024299f54de07ac3a74d6eaa
aba2a932cd24ca21eaec821aba593f57f7b0f782
91091 F20101108_AAAKQX camaratta_m_Page_037.jpg
27911400f8c6916e72cdf1fa71916cda
eb73452c95e7f7f347f525ab9429c2234d6cbc88
51551 F20101108_AAAKSA camaratta_m_Page_095.jpg
930e94d05a225fc8f73de5fc39b5ade6
6b4859bd8d418131885d44877bdc8b6d7747e20f
94182 F20101108_AAAKRM camaratta_m_Page_068.jpg
f6ac6de516c1ddb5d4976207cdbde338
a41f12360de4edc422981b2bc11bb90442de99f9
80650 F20101108_AAAKQY camaratta_m_Page_038.jpg
c4dcff68207c24bb729d9690c70dbbba
83e2d0b18cce4c0f9e444138735aaaeb07990b90
34624 F20101108_AAAKSB camaratta_m_Page_096.jpg
d64da5921e64fcadcabbe3c4868f99e1
5e41adbbb59d847ee1a0408dea218698aca5b621
27376 F20101108_AAAKRN camaratta_m_Page_070.jpg
04ba06645df15c97afd24f3e26d34fc1
30fdd249c2dce529869500b8a22b58f5f8a8760b
91887 F20101108_AAAKQZ camaratta_m_Page_039.jpg
bbf8ff69e3ec1c5b123afaaabfab0df8
9358ff361d542075bcc7e32155d4072524c061b5
66464 F20101108_AAAKSC camaratta_m_Page_098.jpg
548a3b6e078830adf5434c155f2d1708
052f872d5b8e08afa15bca27b61c68eaea3038fe
44442 F20101108_AAAKRO camaratta_m_Page_072.jpg
9a61c88e2b0593089e50812566e2502f
220526e95cf8e94a62a5a4bdf0a04f35ef5d12af
83963 F20101108_AAAKSD camaratta_m_Page_099.jpg
eef5a08b5f83ecceb75242fd8606d55c
a76e62eca73503ba55d359044dd74199eb9da969
45616 F20101108_AAAKRP camaratta_m_Page_073.jpg
7280c95e6b25c5611211949e02a723d3
d0d59ceb011b19b615f23aee8a50707fc045311d
93861 F20101108_AAAKSE camaratta_m_Page_102.jpg
8ed91b05b6b7bf437b150fd0be9b1356
845e17cb9bb248da170583b47aa33194ea054ed6
36832 F20101108_AAAKRQ camaratta_m_Page_076.jpg
f72bb629bd428bbc0d7708ebf6f5da0a
9c0b1ec9eeea3fa9a5e7ac955dda733d7ce183f8
71327 F20101108_AAAKSF camaratta_m_Page_103.jpg
385c07d63435b7d16fddf39568cc58e2
6a3e700bb6037396e6b21eac989d6bce5340e817
29207 F20101108_AAAKRR camaratta_m_Page_077.jpg
d49c55dd473ff43c81ab11af170ff931
ba95ea7e9d575bc0b80508b3167a26993c69458f
93422 F20101108_AAAKSG camaratta_m_Page_106.jpg
47bb24b6f6f5cd8f1714f4124b0fc874
feaa35b01b621aa80fe86e8623bc99b69654370e
59242 F20101108_AAAKRS camaratta_m_Page_080.jpg
9febb5fb51e4ef6c0a1ccc55d66810c7
edd8313c965cb5cca01ae045e3fa6a82cbdf705f
87092 F20101108_AAAKSH camaratta_m_Page_107.jpg
bef5cadc439f10a1b26b946f97d6fc98
13327bc22899db9fb0f6d785c09a5225f00603b8
83854 F20101108_AAAKSI camaratta_m_Page_110.jpg
2116120115ee4e4661368716684736bd
6819be876630550ef6cc83da400a338060dc855c
75703 F20101108_AAAKRT camaratta_m_Page_083.jpg
c450a8278e7a06f368683e08b6877c4e
2e5da370e87f8ee86285d2e488862ff9599a45cc
92869 F20101108_AAAKSJ camaratta_m_Page_112.jpg
d6a084d9bb3d282bf6ab328b282bae9b
69418b8f2171d51848b96f1cc57182faf9c83a8d
86295 F20101108_AAAKRU camaratta_m_Page_085.jpg
6d204f29a8d9d0bc0462a14cd29ed309
f39f29249e88770b28f982e44a573a6a2d317451
76919 F20101108_AAAKSK camaratta_m_Page_114.jpg
6b9c6b7c71b89b8dfe5aebbeae38a4fc
d0ac4db2a10fbf8a0a1659eeabd51add7d47607a
89090 F20101108_AAAKRV camaratta_m_Page_086.jpg
a15864f4a086c06be00dad69819991a1
863284c52dc4fd07e9b9592aa28f8ad0f46a3305
83488 F20101108_AAAKSL camaratta_m_Page_115.jpg
4d1689fd15ad8ae8140a05e2fbf52561
07b8e4735d3c4d65b0046aa7ec289990f35d8a39
92757 F20101108_AAAKRW camaratta_m_Page_087.jpg
b2b6ab7cd82dc101c37ded5b1fc827d8
adf2069d10dc9dcfcd9237f8c4c9381d69c62d92
84805 F20101108_AAAKTA camaratta_m_Page_157.jpg
0195681ec795b85e336d1b1b220edf3d
7bb013928601a021850ea1af9aa6676e595c2f23
84878 F20101108_AAAKSM camaratta_m_Page_117.jpg
dd2ed2a7a46b65b9ddcb12fc1cf6a12c
8ecb2f7d85042cfdc3d974ee59d19bbd9805d9df
88896 F20101108_AAAKRX camaratta_m_Page_091.jpg
20c205b421c580c69a6ae812e8fcf8f3
71d2bb6032aa0e4f2dc8985e20a410915251ba53
73474 F20101108_AAAKTB camaratta_m_Page_160.jpg
7036f8599d813ee664d8b4fccfe29250
7baca4c147c73e4298f3853c9c723458723d9e90
63428 F20101108_AAAKSN camaratta_m_Page_125.jpg
7d7bb1efbed9ee31b38bc1772cdbd1be
27ff8b21379954bc2e3a2c1f33cdb0b6b6480127
53230 F20101108_AAAKRY camaratta_m_Page_092.jpg
e52ed7f6eba9bb0836609d69769a4d23
434665d7f1d52a33535955d339208fda1ca25c9d
44446 F20101108_AAAKTC camaratta_m_Page_003.jp2
dad01b12c0f9f07dd4abe8e39324db36
79454721a871613c8772569b21f10d5b009d1117
35894 F20101108_AAAKSO camaratta_m_Page_126.jpg
54a648506fc923d56f6556549de5ca77
bfa55c177e8c473ab9db2c5ba8888108bb7225bc
38249 F20101108_AAAKRZ camaratta_m_Page_094.jpg
6f59da4a7944629de762603389ad2d9e
329b5187236f21d2c0c5315dea01fac57041442e
1051971 F20101108_AAAKTD camaratta_m_Page_004.jp2
c471f410af5f992ca6a421c9c688bb50
f83ccece28b01bf949ef543de7aaab31be791339
38469 F20101108_AAAKSP camaratta_m_Page_129.jpg
8dd6423934e6fa38352c3ea96f654751
b64eccb7331144d00980a43c918587478211702b
F20101108_AAAKTE camaratta_m_Page_005.jp2
ebe09814b9da43525dfd353a259ce664
b12c3d8f2f1bb04b9528b4f6906b5ce5b9c5e299
34072 F20101108_AAAKSQ camaratta_m_Page_132.jpg
5ef39c44b4ce5d2ab84b79ca6abf987c
677160298146c05ce802a581cfb0a76b19d90d20
1051976 F20101108_AAAKTF camaratta_m_Page_006.jp2
cb0409f8d4740b823c1ecfacb4c22b20
f3789d3df4842f5da06ba4d549f580c8e3dfa68f
48026 F20101108_AAAKSR camaratta_m_Page_137.jpg
e3b0f0657df7f8a6bf01287a82f5fda4
6942fd602f3f3adaeb5492e10659b01c39c73160
438556 F20101108_AAAKTG camaratta_m_Page_007.jp2
cdfdb3689a7960debd0ec44519752458
487d62dbd3b5477cc1e32671437e3bc5d761f7db
88161 F20101108_AAAKSS camaratta_m_Page_138.jpg
22165c49c70efd2a4f109f52ab325f9b
cd5a7eb19d6f8d8a1d936be929af6a9ec1d2cc25
1051958 F20101108_AAAKTH camaratta_m_Page_009.jp2
2d51e548089d04332b939e2a8e4b93e4
91508a8def4ed6251966c75f872c048b94fd34bb
86356 F20101108_AAAKST camaratta_m_Page_143.jpg
44018ee2a5f1c1709c8ca6eb1df1fb00
9a987b7325c9a1bfa48242d88fbdd6bd9cc9d3c9
1051986 F20101108_AAAKTI camaratta_m_Page_011.jp2
3b5fe934002de758e6ae364f72ab48bb
b942e392cde8e319e2fccc41212e33650d24ae45
1051977 F20101108_AAAKTJ camaratta_m_Page_012.jp2
316b704673723180ecad3704b8fb2900
660c354d0a363f5778e45a5bfab5fc93917a6c3e
76191 F20101108_AAAKSU camaratta_m_Page_146.jpg
0cea2c3ccd0bdc8d222c22a91d75b693
f35205253959729867eb6cf01be414d781c4e3c0
1022815 F20101108_AAAKTK camaratta_m_Page_013.jp2
54d4f6100614c2db5c60a20a3f3c6ec0
0d14924db39886f38f22d8403ae95ae23878bcd6
73321 F20101108_AAAKSV camaratta_m_Page_148.jpg
f34517fbce0361edc7fb976ae0a83789
9d2e50feb1a0eaefa37bc357a9686a71fcd5be1a
73039 F20101108_AAAKSW camaratta_m_Page_149.jpg
976f714158c5dbf4c1ccee2cd6d47322
8d626cced8a7b8ece7d6a8a7173c276ac14fadad
1051985 F20101108_AAAKTL camaratta_m_Page_016.jp2
306d12cee9e771c61ac21249ae63b3b5
61982a5e4b3c783737d646c80f452dba9d498eba
58965 F20101108_AAAKSX camaratta_m_Page_153.jpg
8128cd821fee01bcf826d41ee4b4787e
de6425a4203c11b42bcf2099bfe305be9d95198b
771001 F20101108_AAAKUA camaratta_m_Page_055.jp2
cca9bd8685f6179aa9771218195b9476
113d7ede49cefd697dfbf4f63bbe21dca7b72814
329998 F20101108_AAAKTM camaratta_m_Page_019.jp2
c9e672074e5c0b49e80a34968fdda75d
183ef5279e36d937b12a85e4de54fd937f0e8191
30139 F20101108_AAAKSY camaratta_m_Page_154.jpg
791bf9746250ec67b4eefa1098dc1a04
2f42aeeba1c7563d00257285d66dfeaed11dbc04
931306 F20101108_AAAKUB camaratta_m_Page_058.jp2
735d5b12767471c4c9da1eff906432a6
a0388a539c29aa2b53cf9ef59d538c69575dcfbc
1051983 F20101108_AAAKTN camaratta_m_Page_023.jp2
e53cc0c63b7fd26ed4b57bdd39491c2a
53b5fa199b521dfeeac0b97a9dadf6cc4893654a
91705 F20101108_AAAKSZ camaratta_m_Page_156.jpg
2bdded9ed6a4fc2d7c2dc0f60725b893
08cb7631e73959559ba9e2c4b23600bafae4527d
1051922 F20101108_AAAKUC camaratta_m_Page_063.jp2
7c490b4025432467f8789822fd54d7fa
5eee46ed61d4c513692f78b42a50372011fef2e6
1051919 F20101108_AAAKTO camaratta_m_Page_025.jp2
3cff55e9cd90af8ba2ec6ffab81aa583
9732f207db6fcde9bb969c1f2a7a59ba0e29d98b
1051966 F20101108_AAAKUD camaratta_m_Page_064.jp2
01ae67464cc8e327bcab47fde172e918
7d54db889d66d8013812a4fefdf894dd731e726f
F20101108_AAAKTP camaratta_m_Page_026.jp2
f6a848d291a71de0a81b07f28b63272f
8aad0e75ae9117baddd41f1b7ee5ed598cb9994a
1051961 F20101108_AAAKUE camaratta_m_Page_065.jp2
e5c226fb0ba00e31b519534ad81a5407
67111a7eeaa407529607576572ff761ea9fc47c8
1051932 F20101108_AAAKTQ camaratta_m_Page_028.jp2
5f4b928428a15a4fa4b42118fdde9f82
04f42a53af8bdb958557398e4dc2a68a4a82c91f
55142 F20101108_AAALAA camaratta_m_Page_102.pro
5af18e126c01b597a6196657ec4b1c2f
c62ccb0acc0808b2936a91d38525ad8e97368465
1051973 F20101108_AAAKUF camaratta_m_Page_066.jp2
03b86c42f8d958719e01dad722ef4476
c5619b6ab9683e5939b2ebfbdfa5cce581e36b7f
1051949 F20101108_AAAKTR camaratta_m_Page_032.jp2
777214a2911ef843e569e8033683b9fe
815285462759dd17304b9d4fc39c50a00aa5b235
F20101108_AAAKUG camaratta_m_Page_067.jp2
108780db0125192a669dc825e7049fe6
161385c667a8e40d77fb6c54bc4d3c079a5900f6
880043 F20101108_AAAKTS camaratta_m_Page_033.jp2
9b03f1aac9ac6442042c031e6e18b618
1296a14e2041821793a6f8b35b33486d239d0b95
38852 F20101108_AAALAB camaratta_m_Page_104.pro
00f280701e8751099a11370a9d3a5781
5e0c295fa52642205f75a7bc8b4052e99ae91b78
F20101108_AAAKUH camaratta_m_Page_069.jp2
f6b93e8100d31f13aabe76522e28fb84
e96054fc9fb2c96775f2429e1a3042c7569c878f
1024143 F20101108_AAAKTT camaratta_m_Page_034.jp2
dafa7371a617e550f644f4a9270f4a4c
a98f88ccd8a677d5a3ce0fb646949e7a0721435c
55090 F20101108_AAALAC camaratta_m_Page_105.pro
8f1ad01918a0a31662611d9fb38bed8a
4e1a6d63e40e12a950e240aaf38deb05c0828d8d
335301 F20101108_AAAKUI camaratta_m_Page_070.jp2
a3b03b2d68a14c0bda23bd1d3b45c342
a6547a7f2e097deac5899501780a98a92d2928a6
F20101108_AAAKTU camaratta_m_Page_042.jp2
c29137751b263bda985b70dc8c840af2
96aa922cc20bbea9b28ec76029eed080841a563a
53011 F20101108_AAALAD camaratta_m_Page_107.pro
f249b9a6e5a06b6e0252b7390aeebee1
a6181c38b176e2fb1852048390d1cd9b3f99f38a
83271 F20101108_AAAKUJ camaratta_m_Page_073.jp2
4099634c673269f89c0f3fd516387778
75c3eca5a75f175b9cfa41b460fdc9d39558de0a
54586 F20101108_AAALAE camaratta_m_Page_109.pro
4384ba8ecb84dd4f6fab3c70860e5b06
d6b3d8cd9a4158b73e0049e0b65b3e55c7e20040
488834 F20101108_AAAKUK camaratta_m_Page_074.jp2
672fa87ad6e319ec9f2a53df3832e7f4
afa121aaa75e68764cbc579c9478112a72c97573
1051969 F20101108_AAAKTV camaratta_m_Page_043.jp2
bc08790554249c35180a8d9829a71d91
6c0f798b3977e0373c5b2a01bfd07a6a38340cce
54008 F20101108_AAALAF camaratta_m_Page_111.pro
ff76cca15f3a05f2bd69a80a4ea41b2b
386969eb4be7280c58ae547d0801390014aaea81
1051970 F20101108_AAAKUL camaratta_m_Page_078.jp2
eab24ebe4eabbfeb0c213171d569737e
4ddf05b3cccc1d437db6c8792f95a933550672c8
511974 F20101108_AAAKTW camaratta_m_Page_045.jp2
b39e591d3c3f177e2a5a97913057748f
82d9e3c974579a16864d413979a95a213c88a20c
52481 F20101108_AAALAG camaratta_m_Page_113.pro
1328ef6891655bab6badcd9950047f5d
94e973c342b0de54446934197568eab30a2ac84d
1051944 F20101108_AAAKVA camaratta_m_Page_112.jp2
db4acafd8a3bfac3365feaaa045653c4
58b2a612676b399909e43221f410182d2457bd39
1051894 F20101108_AAAKUM camaratta_m_Page_079.jp2
42798b7df56802487462afc2db647463
4947dff737e6fc64ba0917f55009f693c4b916ff
809053 F20101108_AAAKTX camaratta_m_Page_046.jp2
86ff1de9d6665d83aeb909e50fa40187
fbcc17331f50d9971e6a1202eedc705ddf296393
51608 F20101108_AAALAH camaratta_m_Page_115.pro
05a17737e628bb8b142cf466a5261ba9
da526abf0200ba1fbde98117216e4dd76e37f42f
1051911 F20101108_AAAKVB camaratta_m_Page_113.jp2
e38204826d0c29ca1356677301cac0f1
652184bacc891b35c210a0862f24a4c550f38173
F20101108_AAAKUN camaratta_m_Page_080.jp2
d45b6a6756bf0c97fd6398238063daa3
aee2c7ea751083f9e7ede709669c5eed6fead109
1051979 F20101108_AAAKTY camaratta_m_Page_051.jp2
6dbef78e4aa1af3ef8415ead2c5aec21
669ec66ed848c7e6164d2990ab240c282fab5713
51309 F20101108_AAALAI camaratta_m_Page_116.pro
043df5017937fd23fff2c0f3a7af3fcd
92eb5eaa9fef4b023dfe0797c4f4f01b93c8b9d7
308457 F20101108_AAAKVC camaratta_m_Page_118.jp2
dd5f05da360fb32129b14367e52d1730
2e6504266b385640d5284610bb22fc961358d031
F20101108_AAAKUO camaratta_m_Page_082.jp2
e0c6eadea14c030bd417641aca3a9909
9c091908a34b6139af8d016024b79544d5139eb2
790356 F20101108_AAAKTZ camaratta_m_Page_053.jp2
83b40b4cfa9cc2882a081aa25ae97c09
fcec17982943973d96a6e37019100473ecbf6f6d
50732 F20101108_AAALAJ camaratta_m_Page_117.pro
46e0a4e12762e92f3bc890e5c205c2c1
3f65cf3b970b1955e00a27a2f939d8542530f235
500497 F20101108_AAAKVD camaratta_m_Page_119.jp2
eb15b3505fe49a058e700f9c13e23545
24147dbc3cf9c29311bdb87680003ea32b15e258
1031633 F20101108_AAAKUP camaratta_m_Page_083.jp2
9d1269f8d08c743ee3aaaf73a984a2dc
61f56a65fb7e0d0e321a147fccca858ecec6a1c8
18964 F20101108_AAALAK camaratta_m_Page_119.pro
1550eecc0a83658c3d1da7a8e6f9d747
4524fd2a19f2f643d09d021dcaf9aecb080aa13d
713278 F20101108_AAAKVE camaratta_m_Page_121.jp2
ef8860f57e447f0268c3b13f8d4f8085
c4b93c8fca25950e6b875d7746b488e3a1315dc5
1051959 F20101108_AAAKUQ camaratta_m_Page_085.jp2
73c6d697f201e27cf131286474e43ffa
b5ec6487494f8e4c17351533a8da158afae29ea4
2394 F20101108_AAALBA camaratta_m_Page_006.txt
469149f50ba80243b48efa237f4e0e9d
aae2a7ec987e71e8f8af24908b3030859fa7d04b
3981 F20101108_AAALAL camaratta_m_Page_120.pro
5174fe99a5bb82984a9ad1d5d00b397d
7e26b5c1b4e0d64613e932bce726d21eb7b14dcd
603961 F20101108_AAAKVF camaratta_m_Page_128.jp2
5616a31593dcd69fd9bd2c22b17f5cad
42b632d0806a821844a5d9ec11819440f9a06fb9
F20101108_AAAKUR camaratta_m_Page_086.jp2
240ebb97559d94d6b62cdeffad6e65a1
fc0c74f9e30d473ed885348ad2e210eb3afada13
2093 F20101108_AAALBB camaratta_m_Page_016.txt
4d455ad6a9c5e0b86b31db85d45700c2
14b22bde68ac086241e382dfb7881b0c60145038
8122 F20101108_AAALAM camaratta_m_Page_121.pro
45eba48e28b7d47fa8caa56849b6a1a8
69fea4e8e08fa040de386a07aa1c0338958ee9ee
431656 F20101108_AAAKVG camaratta_m_Page_130.jp2
95dabe65fb2bab75ac3bdb1a349ee95d
7cad4a355876a69056f496cb23395df01f4e1ae3
447958 F20101108_AAAKUS camaratta_m_Page_088.jp2
5f33ac6aff211b0597ef25fad4fef1dc
fb3a776bceb22e04d1770908602a16ae31f6eb16
4300 F20101108_AAALAN camaratta_m_Page_122.pro
11cab977ff7ff38cd4af64caef8b96af
7e5dd7f0410b4f51385185cfcf1901f53dd91466
797185 F20101108_AAAKVH camaratta_m_Page_131.jp2
f0d69e3d7742ced4aa13895747e15461
7d4ef97648e212bdbad2d26bde14e90b3a14e883
1051897 F20101108_AAAKUT camaratta_m_Page_091.jp2
32a66bce2bda95368ff7bb979c51b327
84211e45c461c3ef88e210cfd0c0b8ba388d1780
788 F20101108_AAALBC camaratta_m_Page_019.txt
b1dd2e70e90a8924054071757ee10e20
c0a3632dfd0adb5107920808c8f80e71c9d06581
5816 F20101108_AAALAO camaratta_m_Page_125.pro
79697f405088c51d67ad783ea48d4406
f5dcf23b902432c83b0c9c2997e2fec954b71f94
524789 F20101108_AAAKVI camaratta_m_Page_132.jp2
d4b19d24785bb25fb24b41d9037a2522
c06935f520560bc699f054f207c62db1c1045f36
425562 F20101108_AAAKUU camaratta_m_Page_094.jp2
a447784213fa4f283d88ef716c504fba
1a72e1c51ede8b60000b3fff0e554df74f19786c
1966 F20101108_AAALBD camaratta_m_Page_024.txt
82d95e0d6eb2202ae217725ebfc47a1f
3c7b907cd6547a5950b59d44a7825724f9482313
9490 F20101108_AAALAP camaratta_m_Page_128.pro
80a7497280372a183cd72be8d9c2fe01
0e5a36637b9fcdee0ffde35a6af64a0d97c50bd9
F20101108_AAAJSH camaratta_m_Page_037.tif
7d92f1b99b17131d77dbd34b7742d748
0affe65d877c6fa3db74f7ba9be695e0cc78e6c7
918546 F20101108_AAAKVJ camaratta_m_Page_133.jp2
b8205a1ee185257d06379cbd8c89f7ac
2e550d564bd6a0c2819a7e56acd744b235e70295
1051968 F20101108_AAAKUV camaratta_m_Page_097.jp2
12aa7ad83ee308e22cabc6d1ec9b394c
4861499dbea6d61e8ba989d8c589e79316d2d44c
2215 F20101108_AAALBE camaratta_m_Page_025.txt
13500a04a49a0291d0b31ca31099b6f8
67b3359fb74cdbcd4975cdee7037d7b75763bc1c
9443 F20101108_AAALAQ camaratta_m_Page_130.pro
715bea71a59d85e5dcaf95f33c7fe77f
c005f23100dfc33a7bb71cac7ebee1de5799c5db
1883 F20101108_AAAJSI camaratta_m_Page_114.txt
026c7bd4ce8d9368f5eb158ef122c705
7d579396a2fc3364635ea29b1c118ccc23a5ec05
1051946 F20101108_AAAKVK camaratta_m_Page_139.jp2
6a15a72509fc49b41216e379698e78bf
53b7895d81192be39c1206806925fd3e46b22c39
2040 F20101108_AAALBF camaratta_m_Page_027.txt
ac5e312d9ccaf59e9b32d5fa5c1d1691
d5fec3c8c80ff58ebd963ede28af77eb87277b66
14004 F20101108_AAALAR camaratta_m_Page_133.pro
4f13f3c8e8483c9f3bca6e76df62fb52
7bf3d44257f6e2dfea061ebe76528975655022bc
83859 F20101108_AAAJSJ camaratta_m_Page_023.jpg
92182bfc559d205d1ee237a7dd4e4d7c
03fc4536b77b2029c541c00840ecddd11a7ff9cd
1051939 F20101108_AAAKVL camaratta_m_Page_140.jp2
433d6943f6c69464d941cb2162e43293
77c1bcf3e82482cd31a09f0d1e108b3e69ecc9ac
1051921 F20101108_AAAKUW camaratta_m_Page_099.jp2
1b6679cd532b178d352430378019ad3c
ea8bcaaa3db9189623c7f4d4238b0721d3ee2c92
2264 F20101108_AAALBG camaratta_m_Page_030.txt
5e3a2e26ef99b41da0baf82804965494
9148f51be95b155db785533a97359fa2870673e7
4280 F20101108_AAALAS camaratta_m_Page_137.pro
fe4066c4ffc78d0402c9ae414b50544c
4c6fc9857ae5f941210c0a422634a79d319929a9
501 F20101108_AAAJSK camaratta_m_Page_121.txt
d43e6f612318bfd574dc8eb4d708b80d
6fd9c3ed731fdb31d8e793cb175fbd01e4c4b34b
1051960 F20101108_AAAKVM camaratta_m_Page_141.jp2
159b74d35bafd0af2971480145d8aeef
a6ad87a77b93f385e0afa58087360739c4c1c871
1051963 F20101108_AAAKUX camaratta_m_Page_102.jp2
7afd4fba22189c42b3f057899be664f4
202ef6768f1ddab15ea6cfab849c24206322930b
F20101108_AAAKWA camaratta_m_Page_009.tif
0fef60fb7e003ca51649db07accb155d
5880ef8f5cac4b74f530f9c28ba7e38bf98b2cd8
2145 F20101108_AAALBH camaratta_m_Page_032.txt
3dccbbfefcf450d5c6d5b85a0ef05e1d
cac03738b02e4c54d257a9e85a402faa0ad05d19
54199 F20101108_AAALAT camaratta_m_Page_140.pro
6df2e319d572c355125c06f9a2e9e393
00f4f735889889a866ee3c765bf9421b7b89c213
939793 F20101108_AAAJSL camaratta_m_Page_047.jp2
f2541be8e86ecba6ed253ccbddc3af16
a88e31f23305d08185450c7b58e5159a3f11fe34
1051918 F20101108_AAAKVN camaratta_m_Page_142.jp2
f2ee970877f6309d90ea1a2f9689d025
ba21a77a8dd7a70a664c4801454dbf75aaae115e
F20101108_AAAKUY camaratta_m_Page_106.jp2
d5318978be63b75970c4d771ec47aad6
306c881193a1b73c4cb846a870414d4de54bbcc2
F20101108_AAAKWB camaratta_m_Page_010.tif
c1a24555f85c80cf77c4165d965ae96e
ba3e86483b2eaf95ed2e3f5119d91a7a429820dc
1718 F20101108_AAALBI camaratta_m_Page_033.txt
ac66900de326e9541d9b5c7a4573e285
7f6a3b0a75ecfcd23e0c9dd73a09733fa4e90033
46272 F20101108_AAALAU camaratta_m_Page_146.pro
a45325221072a72c4f390db192e715e1
13ad1d0ed35e448d03f275bd621d19f1ae0317d4
16511 F20101108_AAAJSM camaratta_m_Page_095.QC.jpg
050fc9af62fdc6d91c9867f23c32dbb2
96f55f3ab7dfd18c41d51230a578ac7d75ffef16
F20101108_AAAKVO camaratta_m_Page_144.jp2
7743318aea74524801e207fe03fab307
d05e3fa4338964622ebf6a168d642382a0c956e6
F20101108_AAAKUZ camaratta_m_Page_111.jp2
476ec8a3275678d384aa7e5a0eb67a00
33ae6a5ac8d16fca22142c0a4199285d1163866d
3039 F20101108_AAAJTA camaratta_m_Page_009.txt
2559efe0a1fc69ac743c66510a58e02a
c7be307ad3e0cf45df64037e395b332cc7afa962
F20101108_AAAKWC camaratta_m_Page_011.tif
1ac7748a17a6999df276b3742937d192
0fbecfae276d5d952f750c8f25a8974ff60adc8e
1808 F20101108_AAALBJ camaratta_m_Page_034.txt
fc56bcab817c71db8c5f0278676f3f06
b49df27fda33a44096bb1948e61a988ce55a82ba
49881 F20101108_AAALAV camaratta_m_Page_147.pro
e5b429eb2367e083f4bd026f2a37c046
c733030dee832cc1a7bcb12fab95c0a6c1749ac0
27955 F20101108_AAAJSN camaratta_m_Page_015.QC.jpg
a4fb4f654e923ec5ecdda26f48892c83
8316188b831644f7fe3b66f3cce260858ad665be
1051982 F20101108_AAAKVP camaratta_m_Page_145.jp2
73744f331cf6bb2c44132d0eb14a34d9
7e2c5d532691f9f600a23a91bab64eb4605d54ac
1051839 F20101108_AAAJTB camaratta_m_Page_098.jp2
42df75dcf0fba5533a227bf7c27ba221
42705ef71da63203f87131a102bdb04e5dea6166
F20101108_AAAKWD camaratta_m_Page_013.tif
e83885fc1688116039d6e5250e721b12
d0907512235a98253217f17cb3239b62a0073e1a
2051 F20101108_AAALBK camaratta_m_Page_035.txt
1ad6f5f634864c78c41384dd30e0727b
cbfc7f50021c3fe8ae744def63d4548958207dda
27800 F20101108_AAALAW camaratta_m_Page_153.pro
32444e15f233153d748827611ae74294
cce6801da6718096390677c981ef161b6d4e0502
F20101108_AAAJSO camaratta_m_Page_108.jp2
e7982ef6f4f1b8097611bf4fbc469ec2
56a5d54392d7d9e4a115783f6af5c3e71f8bacbb
F20101108_AAAKVQ camaratta_m_Page_147.jp2
ad9ef7817c80bbda8e5d47b79983ffff
de4b2e7e5d1c200ee959bb804cb8a1f83071c8e2
27147 F20101108_AAAJTC camaratta_m_Page_155.pro
dae292e5eddec40131e113ad7a829bba
84dad44fd69dfe0b4ebcb488f8ad9761d7af6f03
F20101108_AAAKWE camaratta_m_Page_014.tif
056b50514b221cc8d305b36f22edf942
0594a3f7cc1eb6fda0d19d375ef3ed6ed019ebbe
1981 F20101108_AAALCA camaratta_m_Page_063.txt
0e016076f7feee65d164642e7c137d14
d3366423a31b83e951dd77c8238ef4dd93374f1a
2013 F20101108_AAALBL camaratta_m_Page_036.txt
8fb5c68f5d3c2de54676202bf4f17e9e
3a46ea55cae2a807869da09bb7c2cdaacc6bb756
39981 F20101108_AAALAX camaratta_m_Page_159.pro
1d0f6f4cf55b5baacf4f083bbeb179fa
e06366089c824bbb594509de72edbb1198ccf8be
89559 F20101108_AAAJSP camaratta_m_Page_100.jpg
906976273456cc6ea350933f4c91b73f
3b08d4c48d6422ef60d19b81fc259a81afc7e183
977140 F20101108_AAAKVR camaratta_m_Page_148.jp2
ca9a89551944c1dd3fc8d4b8c838d537
89ad76ab11e8a6eb98d06ac36b3226143ca2a5ab
27387 F20101108_AAAJTD camaratta_m_Page_027.QC.jpg
f29df3ab060f42156c4f8457ea405046
36ca2fc1b896244ceb958d787bdd580c19cc3c79
F20101108_AAAKWF camaratta_m_Page_021.tif
f7571c11fe11ac00b7f802f783d6b516
196a45bc79bb8fa0640e193bcbde5fe2df7f7637
1854 F20101108_AAALCB camaratta_m_Page_064.txt
d3c1584cae4b09921b963c74fb0b801d
8d57f8b1ae1afc5973dc4bac4cf06e5076e7df55
2156 F20101108_AAALBM camaratta_m_Page_037.txt
d925be8c1476a0429699f578b4a464ac
a89939756d975ca8c2377b7f30637482b308ecbe
505 F20101108_AAALAY camaratta_m_Page_001.txt
8385624e2042ae040fae2e629f19364c
283eadfef67291c734678a6cf549642f278a0981
6390 F20101108_AAAJSQ camaratta_m_Page_027thm.jpg
2f8711e8e82b9ddcb9f7438283ed052c
79f96865412a5829aa7b2dda33d182b876072c78
984155 F20101108_AAAKVS camaratta_m_Page_149.jp2
bffaca7581841f1d1545598b3c24d43c
87237d542fb0cb1a26dffa007349833fdd02fb8e
1051923 F20101108_AAAJTE camaratta_m_Page_040.jp2
ac754f5b6f24baba9cf5b50c20de3430
3a558722e46873cdb91354db37b7354a598fc55f
F20101108_AAAKWG camaratta_m_Page_022.tif
f786236147d1abaaae9b34189a5c49ef
459eaeaef4f883b83d9f0ec35f3383ca54a2054c
2179 F20101108_AAALCC camaratta_m_Page_065.txt
60602a9b54091dc0603bac419a83c5a7
1c702b2271dc32d5e9fa8036075490d5a0424a54
2172 F20101108_AAALBN camaratta_m_Page_040.txt
b17550aee400ec442d2d1a0e4d47b0ff
c14cfc486bc54484102ec417a9a31d237c855331
1994 F20101108_AAALAZ camaratta_m_Page_004.txt
7db499e6054691bb7944d7f7082964ca
53ee4323930240628451edbfea038fe5f6cfc295
89740 F20101108_AAAJSR camaratta_m_Page_040.jpg
66efcdc3fafca3718a1389a11f4887f2
70ed3db557b4bd02956e3ea8b5f2525bf75697f7
901059 F20101108_AAAKVT camaratta_m_Page_151.jp2
89fadddd97493c521622e3bc09867da4
e3283ab44bf114139fb3012ca9d0172145157121
10570 F20101108_AAAJTF camaratta_m_Page_132.QC.jpg
a7e66b9a27deadd515357466233af838
cd893f20911c97475a404b31c6b6f98cdd9b14ef
F20101108_AAAKWH camaratta_m_Page_023.tif
d1206067ff6de928aaafa5a46c1736ca
fad2a7569914cef55ddf932b164daee639a750d8
2247 F20101108_AAALBO camaratta_m_Page_041.txt
1fb58ca149a58f692df45ea1117bc444
d3e738b6cd159d8ef14aa32e0ef8c16c5648c7fe
F20101108_AAAJSS camaratta_m_Page_057.tif
1b0fcaef6d6bdf7711288d8626822160
cbdf6abc54ff8e7958abb5d8927c509bc90e677f
1006109 F20101108_AAAKVU camaratta_m_Page_153.jp2
b5c439a4e944f5379cc92692f9b34d65
8b42f89e09af08b830464c25e4bd9c7fb9944d5b
95489 F20101108_AAAJTG camaratta_m_Page_141.jpg
50f5d1e843c6c96c9db98d6fe5d0b2dc
820391e309214240fe394e3102218bc9e70b2006
F20101108_AAAKWI camaratta_m_Page_025.tif
4944b13c03ccf5bdde99f52efc8ea224
f4f6bf6353ef95b945af82f243fc0bf1663c418a
2075 F20101108_AAALCD camaratta_m_Page_066.txt
dacae39734c018f90ba90100683487fc
029b0a6a5b713e1361e7b4e490b7747e9a49bc86
2123 F20101108_AAALBP camaratta_m_Page_043.txt
3190eebcef13f27061507f8026ec6132
413f8e05ed98a648b2b730b386221e097170f96f
525758 F20101108_AAAJST camaratta_m_Page_071.jp2
96b10b315424754b37bffb2e52e1373f
76284d7390f0e64b9f5511a504bdac2a14f8d875
1051974 F20101108_AAAKVV camaratta_m_Page_157.jp2
10b0bc30ee793276efbfe607fdc86dae
05ee9c8cf75418d68f5f95bac19629833d512d2a
3554 F20101108_AAAJTH camaratta_m_Page_096thm.jpg
40d5b7cd0a618b6d7011d58b867e3ec0
e12764f7eb4d076552dc8ec6e560ec29426553d0
F20101108_AAAKWJ camaratta_m_Page_026.tif
58034e8bb4564f76fa050e7ef5d388e9
8b331dfae23b871786543294c43d45cc9bd3f732
2098 F20101108_AAALCE camaratta_m_Page_067.txt
0f062a3f3c4edaea8fd37a19ef085f32
59f3986fa8c95c1fefe8147974a0a976d1a6d23c
1583 F20101108_AAALBQ camaratta_m_Page_044.txt
ad6378048f451411b1c423e54b4ec316
ae52b3d19860c1330558b5e24d3d02cc608b11eb
52731 F20101108_AAAJSU camaratta_m_Page_026.pro
7c87708b0f244dd574ab78d9fc8a8102
344b62f0970eb4cf134327a5bd18667d79f51e87
1047598 F20101108_AAAKVW camaratta_m_Page_158.jp2
06b8d262362be4083c9c7be9a9070ef9
1d57fc2e7c35360b5cfefe0609efb7210790f8aa
73668 F20101108_AAAJTI camaratta_m_Page_009.pro
4c4b8d497c50226e85eaf8e377100b46
99587f4afd452bc17d4e93c647058815f759e1ba
F20101108_AAAKWK camaratta_m_Page_027.tif
d90af58d8cca55ed0be0c8beac59e964
2576e1c4dfe9522b9b1da75705c94e1a60352dab
2213 F20101108_AAALCF camaratta_m_Page_069.txt
98b63a191955abeb0fa93a21d8a307a2
df0a7715d2401b468170da3a2b72fb57d0417ebd
842 F20101108_AAALBR camaratta_m_Page_046.txt
4f2dac344a00810cdaf4d4c6c7126128
98aea67e1c0965e11cdc9b3a6c250fc2254d3c03
F20101108_AAAJTJ camaratta_m_Page_082thm.jpg
52da6c7a9852e4d1a2cd07e5df9b59af
a416eff9d51ce094ae7dc2b866313a0fcaa71352
F20101108_AAAKWL camaratta_m_Page_029.tif
acfb1700aff66d33586dc383c86d82d7
5cb441101e0b79f95b5c751d3bd57aaa6eb76a4d
596 F20101108_AAALCG camaratta_m_Page_070.txt
11148813f025550fe98fe44e7dcc42c2
d5fabb454aab68546dc2e4ae7fc6edfbded7e855
769 F20101108_AAALBS camaratta_m_Page_047.txt
098b2b5bd99fb2abe49727de21680eff
bd953b5829cee40e6c69c38e596d3d50566b41a7
F20101108_AAAKXA camaratta_m_Page_064.tif
f429820055a3da88a3f95b260024feee
6fa5ae73264120b063117de18759a380e5df84d0
6473 F20101108_AAAJSV camaratta_m_Page_107thm.jpg
530141495ab20fe5ca1ae4322edb32cf
d9839f0b6cd03a948123d248bf5dcd9278cf10e3
912191 F20101108_AAAKVX camaratta_m_Page_159.jp2
6f871abe7c46da1782c5decde5945ec6
0bf95f54eff2ce1705311f5a9eeb28f6ccdb1ce3
5958 F20101108_AAAJTK camaratta_m_Page_146thm.jpg
24beba39545ea8b667e4fc4b63b18a93
90a5f3e557fab15fcac7e8246ff89b56892e5acd
F20101108_AAAKWM camaratta_m_Page_030.tif
abb27d897446f99c2a7d665a24a3672c
0c96ba0184f94ed34f9c08c23fd6d0a2ec83eb5f
933 F20101108_AAALCH camaratta_m_Page_072.txt
a8c926ac7e575e2110733a441956c9ed
32cb17fda0ec03f5133d553af431da159558ecfe
972 F20101108_AAALBT camaratta_m_Page_051.txt
ce82902ef512ce52fecd9abe5ffb5769
417c45ccfa955986c0578cee3a5b2335695dad35
F20101108_AAAKXB camaratta_m_Page_065.tif
e96e3cfc0d9d907cef8f4b56e98480bf
1fdf3d241821eb0d8c086d88f8720e46a126f72b
F20101108_AAAKWN camaratta_m_Page_033.tif
df07eaf1499e968a6c8b4e162e4ae3ef
a230d5c74b9673f7092bfaa424ef5c749a502b2d
F20101108_AAAJSW camaratta_m_Page_109.jp2
55891bdbde9c834c1d561d67c677b81b
3588b43d324772fb87cff5a726908ad5ff6ab1c7
F20101108_AAAKVY camaratta_m_Page_004.tif
09829535e5cd31a5444f088a0c0781cf
d57b0b3de6492b005c819b92b984c38f7c349447
F20101108_AAAJTL camaratta_m_Page_037.jp2
cc8312d10d96026053fd1b78e63fd064
a3d68a49f6217ef4b0e7c5d8d17ba96b39c41d8c
269 F20101108_AAALCI camaratta_m_Page_076.txt
618e43f03136798797c1a003c4e34e9c
009c20dd8a0f93dd3742b80e36efbfc1d57f6163
1114 F20101108_AAALBU camaratta_m_Page_052.txt
f20a579c7d4fa85db38db9ca476b9f2c
73637e0ac0cc4ff850055e77563d0cb75ba49457
F20101108_AAAKXC camaratta_m_Page_068.tif
d2d34bbbbd5ba33bf1476df361370cc0
b4919251f665dc5b5fbe7a6c4f89f3ed9562cd86
F20101108_AAAKWO camaratta_m_Page_035.tif
ca6c50c3374fe97e9d9f3b33b9b7b43f
4c9b8cbcf0b200c2426152af6fdae9a99d01dc46
42700 F20101108_AAAJSX camaratta_m_Page_021.pro
081990b9bccbd695c4a7b397835408b6
34bb450128de30f702c073f01bdac50b58435f78
F20101108_AAAKVZ camaratta_m_Page_008.tif
7912d035afa777c1c9d5e63b5d1f3342
f8993836956197c993b457cf88feb407df0d05f0
873 F20101108_AAAJUA camaratta_m_Page_127.txt
d27082efb9340e363a3a250664c82401
9251078b0c1fe4ed032641c6ffed6f7bee5a4e6c
2272 F20101108_AAAJTM camaratta_m_Page_028.txt
bac0d6067f9332c1f6f01d39385bec5b
634d25913fb629142c02662841b48a9351d07d83
420 F20101108_AAALCJ camaratta_m_Page_077.txt
bf290bbfe363f52e4497551b01d98820
061fa42775a9f83d423fbaf56fc1d61177fd3fe7
646 F20101108_AAALBV camaratta_m_Page_053.txt
8429d6f66d37308ddeb851b004c4f31b
01e268f12a0b55cb8e68aa9153157364d44ad02e
F20101108_AAAKXD camaratta_m_Page_071.tif
699cbe30604adc706d1635599a88b191
dae068985787c187d1f8f23a20f1b67f01350a30
F20101108_AAAKWP camaratta_m_Page_038.tif
3c1abad83fd266a2ded603c7a338cae7
d32150732f44063160cd28f7615f99c9a067e156
51894 F20101108_AAAJUB camaratta_m_Page_027.pro
417b382d678a01fbdf8b192d683ad23d
b357a8fae12300d216f2f6f407504681420372da
21053 F20101108_AAAJTN camaratta_m_Page_033.QC.jpg
5b2ed482a21081ee714686806015e969
a94d1e7e3f49331bc56c711018e77a2393f7f6cf
44879 F20101108_AAAJSY camaratta_m_Page_148.pro
9bc0c6cb53627e94d2939f3ec424ec18
45f2562d01f2c07dcd66b03a03dad0c369a47113
262 F20101108_AAALCK camaratta_m_Page_079.txt
fd5771bff351225dda8ab8b81ca9da81
26c29d195c3141c65e8f6bacefc808dac77299b9
585 F20101108_AAALBW camaratta_m_Page_055.txt
f6cff782f3ffaefde773ea3d86079a21
72bfe530d4bddbda4bf167e2e3cd1462d9f8ce9d
F20101108_AAAKXE camaratta_m_Page_072.tif
45ccc24e674ff06fd4cddd0df218ce6c
4bde1b76065b8a483c4f1c9719d9ebe59dd56f63
F20101108_AAAKWQ camaratta_m_Page_045.tif
0b682be3b046a9b72da8536386bbc1d6
a2236af57d537ef066660d878fd524d15cdfba54
5195 F20101108_AAAJUC camaratta_m_Page_020thm.jpg
366df3fbc815b3d1ad681b01732a0123
fc1736bf76e2ec3e33d71ad308e8222ab5647d62
50717 F20101108_AAAJTO camaratta_m_Page_014.pro
fee66840bbc62e0c7979a16cf67188a5
7c2f9e9a7c148580274703147b6d62dd7c5c0ab7
89630 F20101108_AAAJSZ camaratta_m_Page_139.jpg
9b50e692a7410611349f6a39683fdc9b
6eeb3d3ea2dc488fffc08bf33e24de1bd402b00b
1916 F20101108_AAALDA camaratta_m_Page_110.txt
785ac07efea5854f5f59dfa821cb1ba4
fc222386d896c37614618979031c8fa95cbcd55c
2085 F20101108_AAALCL camaratta_m_Page_081.txt
9501e5911f868702fd10ebbb9b8d487a
f8c377b4bbff66c58ea06f9e94b6b356ef028a2c
558 F20101108_AAALBX camaratta_m_Page_056.txt
2b7c5e1536d780b67541bf805058180b
01ca29d95b13cc7c2ad778faba0ec211375da6a8
1053954 F20101108_AAAKXF camaratta_m_Page_073.tif
18e07f7441e575718e56f8eb34d8af94
708f9e15cc9b6e720fe153068b3bc2cbc0a2c1fb
F20101108_AAAKWR camaratta_m_Page_048.tif
8a52408e52747b336e9de20ee2dbe3aa
c1a9cafc55017e7c51ccddf7b4543daf0ef839a8
2081 F20101108_AAAJUD camaratta_m_Page_060.txt
d56b1a8bda715d3cb0eaeb691d1bd166
b8fcf74475e7ab677c73f659156e4caa0d4c37f6
4476 F20101108_AAAJTP camaratta_m_Page_078thm.jpg
b43c11eea3455267665b5f9421f161f9
ae28a8ccedce10a5123b589c288351dc911ec7e2
2126 F20101108_AAALDB camaratta_m_Page_111.txt
cae5dcffcbe1e84a247c781a195dee5c
9889dbe0545e57709dd84a00b75f222368bf30e9
2159 F20101108_AAALCM camaratta_m_Page_084.txt
80289f63374590437685132f6e93cb23
7eba45e91edf659489e778652290af0e17477a8f
2127 F20101108_AAALBY camaratta_m_Page_061.txt
ae20bba119b0b5cefbc6a7c6ddc99902
7956789f384fafe5d7c16079726dbc75c4a40083
F20101108_AAAKXG camaratta_m_Page_076.tif
d420ea18e8c41f4a57d5d7b43f27a50e
fbfbfa2caee600b566027297d8b96e386ebd62b9
F20101108_AAAKWS camaratta_m_Page_050.tif
407d7312a33c603bcd47c22d55d8b2e8
70cb72fc6699366c8aea5f24ed03afd78c91ddad
6933 F20101108_AAAJUE camaratta_m_Page_065thm.jpg
442361dd2490b14b2ea55482f1e2f10d
5cd72d6b65faa1721cd61b3ff59e8c4f10d8a0df
F20101108_AAAJTQ camaratta_m_Page_075.tif
e02ec3c024646b30db23ab5fd2c4cb62
1f730af4b9369a3cf5d81600aa52930c09c66070
2072 F20101108_AAALDC camaratta_m_Page_115.txt
8f3896b7bac41758e6181e9127dc7577
228ec7f2b9c93b67d44ae5e99c372f73eed99beb
2060 F20101108_AAALCN camaratta_m_Page_085.txt
0d3890bf006aa46d54c592a09b976a79
369c7fd0a472c00a10e026193e44c7f44bd99951
F20101108_AAALBZ camaratta_m_Page_062.txt
ee63c0a1d253070397ca70b1e2861714
7fa0cd6bb8caaa32d7e50412f6a453ce0fc2777a
F20101108_AAAKXH camaratta_m_Page_078.tif
2abb2cf0ed68e8da0f5dd7d05742fe02
3de5563a067a7f13048d681e3c23f20f66f6b8d3
F20101108_AAAKWT camaratta_m_Page_056.tif
4190fb7508a56632f41e647f164c50e4
be514baf73f33ebefe6c2c17e2fe60d92783c8aa
F20101108_AAAJUF camaratta_m_Page_029.jp2
cdc3badf982a22ba119dd18fc84a444e
1663c1712888c1e8a5dcc4e3bfe799069adc77bc
506 F20101108_AAAJTR camaratta_m_Page_132.txt
ff16ec579f2966172b0d3e2432563eea
94db2c10c7ef6ad81966976ebfef264ef7086ab3
50358 F20101108_AAAKAA camaratta_m_Page_060.pro
3db364cb7b2278e642c48c1cdc829c47
3e06501602e0756e89ee5c7245c4c156f4d1fb0c
519 F20101108_AAALDD camaratta_m_Page_118.txt
39aa2be0dc8c9a210ee0cddc859057a4
0011fded08cee8f2ff7e8fd3f3c788d6f8439096
684 F20101108_AAALCO camaratta_m_Page_089.txt
1db7d086f1281ba87776a419f20d7dce
cb2ffa5c84ef41beb4f6ed10dbce87665bd04b40
F20101108_AAAKXI camaratta_m_Page_082.tif
5ce01dbbee71f02469138118cc8d6a26
28522cf21608015766d37b12629426f33178ee4f
F20101108_AAAKWU camaratta_m_Page_058.tif
48ce8669cdcf599d28ea337b71c831a5
a95bc1601880ab901177f4b8c9c07ccae64399b1
F20101108_AAAJUG camaratta_m_Page_087.jp2
3659895e47a05f6fd739af0f071c7d92
7891ad37b7176e06e422610026974f26214cdba0
288 F20101108_AAAJTS camaratta_m_Page_094.txt
48eaed2ce63ee415f5af34afa2b4bebd
93a74b49c4ab6ce1c8be5e6ec4629debbe36b114
F20101108_AAAKAB camaratta_m_Page_074.tif
4579352351c4dfbb76cd9c1540dd54f2
e67b922bfaeac152a5468e2792baa8c1c6565ce7
448 F20101108_AAALCP camaratta_m_Page_090.txt
f1ccd1f964fdbf900ce8e41d48c234c5
f1469156f390875a37969cc358c004b5ec310dde
F20101108_AAAKXJ camaratta_m_Page_085.tif
546c8150e8d8608731d26ccbb83ae9b5
c41e7d1b4f8157953717bd2567b7c57ad73b0510
F20101108_AAAKWV camaratta_m_Page_059.tif
1c2291455de9d710e3462364dd73ea21
d3f940c9573f6f347d32742e3a1fdf42f57b385c
15787 F20101108_AAAJUH camaratta_m_Page_049.pro
619c5d488af8e373e3dfd6b4dc55c8f6
0aba84c1f3dddd3c56deb30bf507140c53a568ec
F20101108_AAAJTT camaratta_m_Page_123.tif
88608c3b2de379ed86414df9e8a08bec
78618415fc1440163888d6393a78bd82df7223ee
198 F20101108_AAALDE camaratta_m_Page_120.txt
3dcd23b76b00828e05447f1d3afec8e4
272590524dbdad92cab6c6c2998cd60089954252
405 F20101108_AAALCQ camaratta_m_Page_091.txt
bc08397293fb4e06fd7b5c6cef8456fe
ae6210dcf3a900e996481572356bd51d3053cb55
F20101108_AAAKXK camaratta_m_Page_086.tif
bb042fc49ebb456046c5e4b2c62fbc57
c5c22752619e43e06c7b120c2cc54afcfe123a4e
F20101108_AAAKWW camaratta_m_Page_060.tif
27d5bb4bde133931f79a450019d19a25
621616033e5a6d4b7e71a3ef8ddbd4fc8cf44b00
691 F20101108_AAAJUI camaratta_m_Page_018.txt
e9a8ed7876238e25212c12ba34359917
2025bd098118d0c0c2daf8b1727b5ed5593c9c43
6614 F20101108_AAAJTU camaratta_m_Page_015thm.jpg
6adb43ab0983f5d23043ebd75cc91f8d
d59ed85bd16e6535c3b6116f56084000af3c27eb
F20101108_AAAKAC camaratta_m_Page_147.tif
9f004c749d58755b97d42ec3636f7ccf
d7cbae644cbdd754120e78a5180de0b85ec99a89
289 F20101108_AAALDF camaratta_m_Page_122.txt
07f977a8dac57eff81b2f8b78026feca
87977e951f290929d0e6d4e62f89135f79a6d786
770 F20101108_AAALCR camaratta_m_Page_093.txt
96845fbef318c3489177feaf2aaf1805
1264a815a4d83ef423bfc7f897e7d1baffc34777
F20101108_AAAKXL camaratta_m_Page_087.tif
3ef250550908d7c102071bdb1c8deee5
7194a86a198424579228b18602d63552d66ae64b
F20101108_AAAKWX camaratta_m_Page_061.tif
63ad8d51d6bd2f997560d20167e2e8f3
03f021f17843ac16a8eb1d3f0b01d8e0c31868fd
20886 F20101108_AAAJUJ camaratta_m_Page_075.jpg
87fce08c038a5e086daa2e5d1dd9c81e
c3d4e21782f92c946862d5c32d6111a25ab2337e
418 F20101108_AAAJTV camaratta_m_Page_007.txt
be793b8375a274ee7965f092586c5795
6b08a36a79a39af83832dfa0983a466808411975
2065 F20101108_AAAKAD camaratta_m_Page_082.txt
80926759f1901ce31d15a13aed95ab97
22c8ac3a4c294cb67506ac630543b2b86d9d05df
251 F20101108_AAALDG camaratta_m_Page_123.txt
fc6f93b6137dfe1fc6629af5dcdc6c70
7d3ad801d80c65a5a66c9019910b63c8fdda6020
821 F20101108_AAALCS camaratta_m_Page_095.txt
2441d688c6708872481de7e3dfdc7136
d558c02d797315351f9230ae6bd6179449c5cbfb
F20101108_AAAKYA camaratta_m_Page_117.tif
9be6c4debf090c693ade162f42f2bf45
447c460342e07b042f86d29bb3ec67c0b6882cfe
F20101108_AAAKXM camaratta_m_Page_089.tif
37a009c6c51c7e7d77a0d47bdade1f48
2f2b4fc0e4ac008257ad6780532fd06efcf9256e
77184 F20101108_AAAJUK camaratta_m_Page_097.jpg
d23639ab3ef69b28313036414b44f77c
7c2c65331c76a7232de134f9cbf67ce0a0302ac4
33837 F20101108_AAAKAE camaratta_m_Page_136.jpg
40adbb6d909ce9ff22086156ee2252ef
b4b92ffb264e1f0dd4637de330bb106d0749d5bc
493 F20101108_AAALDH camaratta_m_Page_124.txt
af3f1d61aa2e1f49763cf0ea59a22842
d18e2e68b88c21342bed074529960baabeaf91ed
482 F20101108_AAALCT camaratta_m_Page_097.txt
95ba5c5bd694c6a8f1c9d359912beb24
0b0ef78825612886115f783ae87c6d9c4d367899
F20101108_AAAKYB camaratta_m_Page_118.tif
7c62877a9bff526a0e7155dfad0468e8
ebb2f44f84d6aed77f5f59b3c4a41cbefe80da6b
F20101108_AAAKXN camaratta_m_Page_090.tif
7a6959bf6c895327ac1c5c9f9538e630
0f461d9bb6c3b3c917f289632fe0a0979af71059
F20101108_AAAKWY camaratta_m_Page_062.tif
e9b86296b7116a418821ca7159498d36
84487af8aa62ed78f300977eb06d3b443219a52a
F20101108_AAAJUL camaratta_m_Page_048.jp2
525911da928a8028ce914fc944434d56
2311e23a9caeab3f29834ac7ba4e35104c1306c6
14427 F20101108_AAAJTW camaratta_m_Page_073.QC.jpg
bce63de5c5ef16dfaafd8bf48053900c
3b0dd917046d1083ac1beb5e856a92f6736c4b0f
8404 F20101108_AAAKAF camaratta_m_Page_071.pro
3896f9ac67bed1bbee2c55ee8e59de43
961dbb81c0efe41816eab8ca271fde2ef4efa65f
458 F20101108_AAALDI camaratta_m_Page_128.txt
c6684d602acd17104ec85899beffaec3
ba893c7bbafa78a3fb290d9557936640b16f2dd1
2053 F20101108_AAALCU camaratta_m_Page_099.txt
f61f70ae21b5d51521b48aebd3a17def
cb40baa67b70dc520e7ea2cc1e1f8119f304a7e5
F20101108_AAAKYC camaratta_m_Page_120.tif
f1e424f0529170b84d74581fe25adbe8
35e8bc36b08cacff4e05fda9669827143e0ca7d9
F20101108_AAAKXO camaratta_m_Page_091.tif
42de77655fe2f942b783c9e43a827f2f
cfdeea3bfb92b99c8968fa6c4d0ac9c5e429129b
F20101108_AAAKWZ camaratta_m_Page_063.tif
b822f97c166734c1b627edc30e81059c
77ab48ec8579ab7890aecc3116424a0677e73e24
69830 F20101108_AAAJUM camaratta_m_Page_047.jpg
b3d892dfedb9226d7e5c6543c254bf02
6d1271ea2e7f9b9fb640af3b79060d1b8515c4ad
6924 F20101108_AAAJTX camaratta_m_Page_031thm.jpg
3ec529eaeacc332eaa83253b8cb09f46
a076e007040dd9688fa9184b7ded76ebd0fdb3d8
898581 F20101108_AAAKAG camaratta_m_Page_104.jp2
b2deee492b61156386fd10a5d9b628b3
f842e586798c60ff2f6ba064cfc998f76b7cf245
6222 F20101108_AAAJVA camaratta_m_Page_023thm.jpg
463c64a373f415efcc8bc0e6f1404e4f
423aa253ed31f8a944d21188580de6c5600ff58d
374 F20101108_AAALDJ camaratta_m_Page_134.txt
51ed7d95ef57d5656e39383d3b5637a8
9ea1d8361bf9c88ebd981d5bc6b38e7df3537ab2
2097 F20101108_AAALCV camaratta_m_Page_100.txt
b067cd7cd48e90bcb5afdeb71252ab47
dacc7b1fc106c762f51d2f57af99ac45fe45f909
F20101108_AAAKYD camaratta_m_Page_121.tif
f83022ff6fc89e994f23285fd66bdada
5a61d470081323b332efb8e0b2df3b747239a4f9
F20101108_AAAKXP camaratta_m_Page_095.tif
0a8bff519ef521becddbc8eb48d5d389
d1b7ebd332d78ae6965e1931f7207ee0e3610b7a
17175 F20101108_AAAJUN camaratta_m_Page_125.QC.jpg
438cc70ffe5c2ff02863d78c42747b86
8873db1f278cdbb3072d827e4c7da941bde946b3
6376 F20101108_AAAJTY camaratta_m_Page_026thm.jpg
ec1107037c10d9f4c82f014505c8ab8d
95785ecf4c62ba4c1be0f6c400228ceb203d1ad6
2270 F20101108_AAAKAH camaratta_m_Page_029.txt
0d309fd5bed69eab10618667b62576df
6172e9221cda2fd95e0ad8c088d1f9ac6f5a6194
F20101108_AAAJVB camaratta_m_Page_104.tif
41308fe37c8aaf53325377e3de8dc177
09954937ff679d89e57f8113710dfeae53f2a3f2
532 F20101108_AAALDK camaratta_m_Page_135.txt
fcd367799b2a04189e6e7acfc2410f06
7e64eaf90722c00bfa0dcfb10dd7f44d4b7c42a2
2165 F20101108_AAALCW camaratta_m_Page_102.txt
4280b7a7084f035b78b2064084d91d9e
4c0f3f7c3f214abb3e8d07a30253e047c3f1aeeb
F20101108_AAAKYE camaratta_m_Page_122.tif
a8424fde2d4b603facc041f494d0fce8
51e122d95a3bcc82796f79ced222df8e27113345
F20101108_AAAKXQ camaratta_m_Page_099.tif
989937845fc320772dbd4b239c6b01b1
17a9725a1b8767f195174ebad49f10a3bb7bdf53
56068 F20101108_AAAJUO camaratta_m_Page_051.jpg
919f3206ac941631a023cce0a49a3db7
b0a3de644066ec3f7f8938a0a39fc13c608d629e
13020 F20101108_AAAJTZ camaratta_m_Page_118.pro
2108818173b28425c6ffc3774fef1e04
01bccad4d3664f6c48e7e418f5bd08150ca035c5
3102 F20101108_AAAKAI camaratta_m_Page_077thm.jpg
7a9fcce07f4858361a83a339f088465e
ef1dc7523e20801f5258c9c8bc35a7aa7b9b58a7
2141 F20101108_AAAJVC camaratta_m_Page_145.txt
5c815452fa347d483eff1be9862bcd6a
bd214d800b24f748df51c678d31465d52cee408f
394 F20101108_AAALEA camaratta_m_Page_154.txt
09d46cc3b03eb6028e9ae95fa9ec66f1
18b4c0f0dd1eaaa9bc4f30e8cc1d12cfb3af4e61
F20101108_AAALDL camaratta_m_Page_136.txt
d95b77752a94bec6efa9ebb31563fec3
f5de949eae98d07fcf12c9b71cade90b5bf02132
1614 F20101108_AAALCX camaratta_m_Page_104.txt
4f2b41774247027867c027cfa777d410
025dff1ecc244d4231be94bbc6e3ba750ffb6e72
F20101108_AAAKYF camaratta_m_Page_126.tif
1b7389a2251e2c8dac5546c859d2ce1f
da53c7aed6f098e0f0efc6299c3e1c0ed06695b0
F20101108_AAAKXR camaratta_m_Page_100.tif
af0fa844b3bb0b9ec21b990c4df98ecd
503783dedcacd3a5decd08265c211453bace742a
11320 F20101108_AAAJUP camaratta_m_Page_088.QC.jpg
874183bfd0f0fd04144e592f2b98458b
de6fa7d73d0217cd3aaad985aa2535c0ca077dae
F20101108_AAAKAJ camaratta_m_Page_130.tif
d5f87e74000136459dff4bf566c22b61
d493d3585eac08bb54fd6d63f07ada6d2f03ff79
4327 F20101108_AAAJVD camaratta_m_Page_079thm.jpg
40b4030cfd2b0d431d62577a2bbb62d3
b8681155506d64f4ad79585dd0ba831fa79eb415
1356 F20101108_AAALEB camaratta_m_Page_155.txt
e3d8a4dbe110218efe2039c69c65fa43
8164a345573261eeb9798962ed38a23652010e3a
236 F20101108_AAALDM camaratta_m_Page_137.txt
60887cfbdeacba9fd1868d3bee7f23fe
37263bf220b43c8b700ba1d845fdcfa8c8e41c73
2224 F20101108_AAALCY camaratta_m_Page_106.txt
0663104be4252cd8b5ab6fd01dafe410
f351b76342e942b077a4f7f193c666b79d9ce0b7
F20101108_AAAKYG camaratta_m_Page_128.tif
5fe5c10f824ca25a4e8504ce22c8ba2c
dcfc2218212fb9aab4d3a47732b636d46f9ca49c
F20101108_AAAKXS camaratta_m_Page_101.tif
3656115ce255353bc398b505f7c0683e
3d441d7de17ab0bc11c39e506dada43b96f73ded
F20101108_AAAJUQ camaratta_m_Page_156.tif
dc5eb8a909e2126937763578c57458ff
47d93a435fc42a3fcc68f0c1a6c48f1d4758d950
F20101108_AAAKAK camaratta_m_Page_070.tif
52a8e0b363c6ad1dce1e4d070219412d
5b709b6e497a79fc89347a2573e3c859d054f826
961059 F20101108_AAAJVE camaratta_m_Page_103.jp2
7605e87fa4d5c45f08392847b5fbd371
b51a461573815800dbc9e874d9c84af3d471e27a
2188 F20101108_AAALEC camaratta_m_Page_156.txt
5abd964275ca2c4040d0ed93eb6967e7
d04deb1a4cbc6c9f4b67cfddb089769e5bbda447
2140 F20101108_AAALDN camaratta_m_Page_138.txt
1e642182908abbcf26acad4a3e88c6ad
ff12696311f357c280c06bbb9908bf5e62776b46
2029 F20101108_AAALCZ camaratta_m_Page_108.txt
dd7978f59bb20ca4b3d334250855bd0f
1871546e686347da3d2d0ce9a6008f8321545a95
F20101108_AAAKYH camaratta_m_Page_131.tif
9706e00db6980d5b317adb6aa9023a83
ea31492fa881179cc337bcf842cfded2e4d0ef7e
F20101108_AAAKXT camaratta_m_Page_105.tif
72feb1e793f1dcb85a0be99b3bc8884c
e3fab4fbecac3fed7932fd9c87c738e2a88b142b
29026 F20101108_AAAJUR camaratta_m_Page_061.QC.jpg
c8d06d9a9748272479ab11e121578961
4baa1af86b44305575dfa76b647a4144eef7ed5b
27911 F20101108_AAAKBA camaratta_m_Page_113.QC.jpg
d5bcdbc5e942f53ef8d3684ff9303826
e3fac83763f6643daac75f911a3e282b180e8b7d
91363 F20101108_AAAKAL camaratta_m_Page_101.jpg
6a63284c82e3fef23166cb8b4cb1009c
ccc5f43ef34bba2e4cd2e2c408f59b79bf8fb23a
1941 F20101108_AAAJVF camaratta_m_Page_118thm.jpg
a3141c54f24ea506f2e8e760daa2b300
ed6ae976f875acbd28dbf9c2954f44008f9af902
1704 F20101108_AAALED camaratta_m_Page_160.txt
1e796954427e484c917a72f362d3e0f4
15293169ab88706baa60000360b4ee459ad0200c
2079 F20101108_AAALDO camaratta_m_Page_139.txt
7c133b11fd489e08ec57ee70811c4379
83a0c0fca94e8622cf238e064e37b3d0e447443c
F20101108_AAAKYI camaratta_m_Page_133.tif
4667d064286e18d9ebcbfe9190f75121
629e7e3b47c221c934c545b1b92a6b2a26723348
F20101108_AAAKXU camaratta_m_Page_107.tif
3a7abaa5dbe6af84bb69a2502bf236b9
f7b238c84f4952310a7dd09ab653310dde843158
407997 F20101108_AAAJUS camaratta_m_Page_096.jp2
126514fd33c04fbde8e521efc195b24e
75493d3811b3678cf495106a4837fbd5b98cc693
29824 F20101108_AAAKBB camaratta_m_Page_102.QC.jpg
b714a7702ac071f3de8fcdb77f89c02f
1e5b0946cf9b8bdb0a8d9fe6f41e87adec6620e5
2266 F20101108_AAAKAM camaratta_m_Page_087.txt
c7c3775ff3ee469beee8b04409a3e1fc
3805020942fc994bf9e2c3ea62daf7ad3ce396e1
25834 F20101108_AAAJVG camaratta_m_Page_035.QC.jpg
0a169c545095b17daf7af9f419cb9639
7c7aef2667e2eb2a969e3ed0b0aa6618b9e14b11
1784 F20101108_AAALEE camaratta_m_Page_001thm.jpg
05f8083e5d6d827f4e451943f3daad78
4dac97963bc820aacaa53792dff4adc0dc6ae840
2129 F20101108_AAALDP camaratta_m_Page_140.txt
a25d7dc7cd044b1b67414d954f3cb63e
f2920907fcd3182f092ef440bd75f08aec83fc12
F20101108_AAAKYJ camaratta_m_Page_135.tif
5b1df6974d1404150da8316f5726703c
97586dbbc20305ae7226eb560a518e0295fae12d
F20101108_AAAKXV camaratta_m_Page_108.tif
fe865c10998fce55ca5d8cf428c0106b
2e2c3d7ded0c02db0442216235630ba2ae9b494e
F20101108_AAAJUT camaratta_m_Page_155.tif
a1e00312dbd13e3b486d81b3e50c9b09
338ccff43b4298315592856361c8c407f3636912
F20101108_AAAKBC camaratta_m_Page_112.tif
65c2870e9a9cf2849fde975b9c649427
6d90db92da7437837583f8ba2ef42fc98d8f3029
40193 F20101108_AAAKAN camaratta_m_Page_074.jpg
980b55da4eadacc381dbdeca98896ef4
dc61b41283bb4b9d78d17d16d212c2f2fa7e3bca
7637 F20101108_AAAJVH camaratta_m_Page_136.pro
6a9642769e652285fab52f1714269d2d
82ef3cc0b921c37c8bd629fe7b8a2627647294b0
2204 F20101108_AAALDQ camaratta_m_Page_141.txt
30c960142b459bbd590f77bf0788160c
fbfb52e74859f0458d27d37edfb26b3ec4126534
F20101108_AAAKYK camaratta_m_Page_139.tif
3d53152b84ebeb5d47cb1d602c626f77
6b019b9f1b673a22a2f66bc2fe50f2b169d3cb33
F20101108_AAAKXW camaratta_m_Page_111.tif
7c71292484c0b4d014a649ba74f07db7
e052188f557ae09a6f3b1cb86c50b942fff377d2
1051928 F20101108_AAAJUU camaratta_m_Page_068.jp2
17ba42fa1dbf64176c1d1319e804bcb1
fb32aeaa243254112f6cd5bdb6913431902ab5e0
1051956 F20101108_AAAKAO camaratta_m_Page_081.jp2
922c0bce4400c5eb978175a91bfdf1c6
e359f60aa63b30c746609c415a112c51f36c3784
20404 F20101108_AAAJVI camaratta_m_Page_093.QC.jpg
049ce084d3a724643166a069979dcedc
ccceb9a0b5f5a00bd5119b404e0fc654ac81df2d
4010 F20101108_AAALEF camaratta_m_Page_076thm.jpg
278ab002301a65d4c936a14aacebfbea
3fdba4ce0e6d9873e34c4d0cfdd0cf0b4fd19946
F20101108_AAALDR camaratta_m_Page_143.txt
f38cf705369071111e2a390eaf12e87a
36e6a9e075bbc05c84095028833180c2909116f6
F20101108_AAAKYL camaratta_m_Page_142.tif
896083b07ce9537f3118a9a336ae4ffe
2eb5d109e5e1fb688e3a829792998a541b04c7da
F20101108_AAAKXX camaratta_m_Page_113.tif
89637f8377e52a949dc5f243a1b17ca6
6ed23f1c93706a46662659a950aa7907781cd093
F20101108_AAAJUV camaratta_m_Page_114.tif
764d2b61ea2abbdb224fb90e6ade932b
e75417ac2b2bbb054792a4614b542a1facc26d88
282684 F20101108_AAAKBD camaratta_m_Page_001.jp2
890002c8d1858bb56185f252aae83558
4b02eaa90e60c121ae93fa8809cb9dd67ce54917
16076 F20101108_AAAKAP camaratta_m_Page_121.QC.jpg
6e15dac5619f504b30236fca08a21ea1
1ec515395c67b4613acbbc5918082802a91a595d
16681 F20101108_AAAJVJ camaratta_m_Page_046.pro
664f0b44889af0bccaee18ff484b9957
4cddf1450bb29dc1073f44f00d97c4d4116aa9ab
6849 F20101108_AAALEG camaratta_m_Page_041thm.jpg
7b23a28a0acb1aa67ed751e0b7b4309e
9f07ef6ac8f1ac35970896a52ada99e296a78880
1898 F20101108_AAALDS camaratta_m_Page_146.txt
b887a679228ed4146632d849ab511a2a
e7f395c4ddb681a08c70c032aa977004a8851025
43934 F20101108_AAAKZA camaratta_m_Page_022.pro
2d778ba9b177fc66e8c5dff1a117647f
44ae6803f190307c60352df1d45cb1faf70d8cc8
F20101108_AAAKYM camaratta_m_Page_143.tif
f434079f4611cc918efe8a214de10c31
e7ebdcc47a4c673138a63e14a4bf63607baf42d2
F20101108_AAAKXY camaratta_m_Page_115.tif
6624ac8d1cb8c2e408863cbeda53a244
1cc0f2cfd018dcec78445545366c2dd1f64ec45b
56474 F20101108_AAAJUW camaratta_m_Page_141.pro
fabfa1495cb86913f1568604815e375f
0a21893dae834355c6ecee9017321f90ecff7284
1717 F20101108_AAAKBE camaratta_m_Page_003.pro
d16339dde9ef5a76ecaeb55d3d18412d
db82475cac8669f439653b4431a5aa78906b8f23
11253 F20101108_AAAKAQ camaratta_m_Page_098.pro
06c9202f37a6f6402f7cd0d93bcce536
8cfec41accec90c02382a6df168aa58c75270a9d
7733 F20101108_AAAJVK camaratta_m_Page_010thm.jpg
830997ae4328e3991805d3514ece5c3e
7f4a6b703b3a109095856a91b83d2c37c2e7ffe8
4059 F20101108_AAALEH camaratta_m_Page_071thm.jpg
24a71a5defaa8f9718db8473a0b123f6
b193383e3de7d443301bb890f450af543558ec09
1979 F20101108_AAALDT camaratta_m_Page_147.txt
b54b49bd00d76e343728e07ec065e633
d328d141916c792bd385d780beb65ab66fbe157c
51522 F20101108_AAAKZB camaratta_m_Page_023.pro
e8f86100b6ff361cd3bdce4c438eac9e
a19e48c6362383f3adb9712ded7939dba7158307
F20101108_AAAKYN camaratta_m_Page_144.tif
55939555e4113d191e7febe1ab12b666
b7ae4141f6ed5d39969589da33e9217abe18b30f
44534 F20101108_AAAKBF camaratta_m_Page_056.jpg
23fd9611e70bfd038320427dd96df41b
0564aa862a32f0b6ed6481a01dd0f8c180657ccc
25580 F20101108_AAAKAR camaratta_m_Page_081.QC.jpg
b85c619dd053b8fe7327b322344c508c
d246e564dc2c70a80a50d0b2d8d93bf796b7c27b
F20101108_AAAJVL camaratta_m_Page_018.tif
05a58679dd27ecd74016333d18fb075e
736ccbe4b1629be471fe4476c86cdddc99ce1524
6685 F20101108_AAALEI camaratta_m_Page_105thm.jpg
c097c192ba78308bbf81d41231b2c08c
87f168a470706a0755dfda9810d1f54dd0be0f07
1874 F20101108_AAALDU camaratta_m_Page_148.txt
f4970735a8843cebf84f0a831178f470
3a76786ea54f3bcd91f500bdfee55ca36049016b
39541 F20101108_AAAKZC camaratta_m_Page_033.pro
44a2f89b857b42ca420f3735938eb7c2
d92856cfc0a404142a246a2db400528f6a3e5f8c
F20101108_AAAKYO camaratta_m_Page_146.tif
f1179f32e460c14409ee970748c60b61
64d52081fd04bbdc588bc906dbd3cdab0d356709
F20101108_AAAKXZ camaratta_m_Page_116.tif
6294b8402ccfb79b2ec3f59c3247f52f
1380bc47e91901cd404dfecb4fc4cecea10844d4
26172 F20101108_AAAJUX camaratta_m_Page_004.QC.jpg
ee38428be71c818177e29fb6e5ec7514
7d41ec6ceabd848929cb6c33d7eb15492c2c96d7
1194 F20101108_AAAKBG camaratta_m_Page_059.txt
869adf2b89b0d281b5b0164fb13a60b7
5769fd6b059569dbe63c597fe17700221effa269
85370 F20101108_AAAJWA camaratta_m_Page_108.jpg
9140911b8afd4e30417d50ac3dc7438a
acf73d58f47a4f1cd3d95a4e6fd13077a73d248f
9625 F20101108_AAAKAS camaratta_m_Page_124.pro
dc26313b4da0f2b1de744ab5761a715c
3bbfadf08ea69e891360c8f29d4ee5ad8e0420b1
F20101108_AAAJVM camaratta_m_Page_039.tif
9482484b3577ea578dc66800ca2bd138
1cec34fe6ecf4bc604f0779a7ece192cd26d62a0
6935 F20101108_AAALEJ camaratta_m_Page_029thm.jpg
5754f8c074661a9e19decbe9799366d6
21614b62ef4ed22c3bc1902b054dfe26ada37303
F20101108_AAALDV camaratta_m_Page_149.txt
3ca6a4b63c0ec8a54887c03eac725720
664f74e91b5436d9185295d5fd4253a16fb9bca9
13527 F20101108_AAAKZD camaratta_m_Page_045.pro
25430a15af6ef40344f247c189954c64
72376de55025329add6e43ed24864fd177625d41
F20101108_AAAKYP camaratta_m_Page_149.tif
b5738223dba8df19203d436919292821
e3b0933ed98892830b3cb36524f5603f2d0f8cd5
24714 F20101108_AAAJUY camaratta_m_Page_118.jpg
ff2f0c62601f89a12797a765092f059a
3ddd0e6eebe4acfeb3589ea34467c1879104ce98
4677 F20101108_AAAKBH camaratta_m_Page_078.pro
1473e8a5b317a1b01ee2e4f7fe407d53
978381409c409a7b243e943ab92f2f6c1e18808f
F20101108_AAAJWB camaratta_m_Page_006.tif
6b847de8ec253d2704baa38e2c29e4bd
9f9dbf0604b15065148055ab249aedbde4371a73
1056 F20101108_AAAKAT camaratta_m_Page_049.txt
a67debbbe3c50c6aafc4ae8715ac01af
266db9a876676d26f489158a536354aa79f959e9
49382 F20101108_AAAJVN camaratta_m_Page_063.pro
4f3ec866531e6c2a3e7ca41e340e48ad
0f64912ecfb61edcc16eff8ec85cd7159274a4ce
28179 F20101108_AAALEK camaratta_m_Page_112.QC.jpg
39051a65ae0587307609339e6bd39aa6
fcf5c33d58cc101891fe05bbcde5aada3beeed57
1786 F20101108_AAALDW camaratta_m_Page_150.txt
a302209654e24aae93d396ae5c0a08ac
50c7c267a990e0156412a520121ebbb8da6353a6
16469 F20101108_AAAKZE camaratta_m_Page_047.pro
1d0158db9f7e7f100ae7d3fc11c814a7
76b2ed2937c4ff2c14bff47d5dfdf0cc64abe682
F20101108_AAAKYQ camaratta_m_Page_150.tif
a222dee329d2170f0c6373ab09e80d74
983dd1cd05bccee77929559557a6750af22f325a
27103 F20101108_AAAJUZ camaratta_m_Page_026.QC.jpg
e81368fbbf2bf5515ddb3dc385c25d66
e83e59053cb13adff52a6fac46f5338f38c5bd52
25128 F20101108_AAAKBI camaratta_m_Page_127.jpg
e21ba04efb86451dcd639c8bf5989454
00ce69befeabab0954a40a4fa448ae4d0f815e1e
89627 F20101108_AAAJWC camaratta_m_Page_113.jpg
b2183a8c114c1b6d7c1873e98c2b2a6a
dd3f3da477ad1e15374570b712ce1844580eab6f
55519 F20101108_AAAKAU camaratta_m_Page_065.pro
ef0f68da2a09ae68e738df86a4cab286
4329cb2cc218d317532934ecd8a0e94cbc195eea
F20101108_AAAJVO camaratta_m_Page_157.tif
6590643ae4a57923f0e383cabce7b68e
10527c422dd832c8bbb2fc229deaf672a5eae83f
2311 F20101108_AAALFA camaratta_m_Page_075thm.jpg
3fe279fa8fe04db3fa12e5e33c0f18f6
6ef0f664a576fba49e188ee639496a705a6beaf2
27411 F20101108_AAALEL camaratta_m_Page_107.QC.jpg
7e0f7d72c03cdd5928d81e66fbe37383
307d39d7935f74127ab1ad739fbd355f4fd75a67
1611 F20101108_AAALDX camaratta_m_Page_151.txt
6849c3a31b354f9fd5ac2236e0eb3186
d95afdf80a3526862f613fb243b39e12b17c45b0
21599 F20101108_AAAKZF camaratta_m_Page_051.pro
cdf6ee104fbd52b4c9bf29ef288609e2
e061a0232065089f2da76dd6cce4dd33ac006425
F20101108_AAAKYR camaratta_m_Page_154.tif
9718a10b16c68faed1453c8e86a4add8
80d646274d1618828d52310197070360beaf5c04
87589 F20101108_AAAKBJ camaratta_m_Page_043.jpg
a24a15795f3aa1236f2c156ed25b1b93
005e3c5846493cf825b640497c473e92097cc750
F20101108_AAAJWD camaratta_m_Page_106.tif
1e01fb36e170dee63631e119e715b1cf
dea6a96c63cba8ecaa1f0534581d7d7b87cb8fe6
18499 F20101108_AAAKAV camaratta_m_Page_059.QC.jpg
396c9821357277beb3e14964632128bd
7f116bfe94ca58cbbce23c6fc78240f7473af2c0
2067 F20101108_AAAJVP camaratta_m_Page_113.txt
cb8fb45bbd1474fd8cbb0e28c50ffed0
8c303ec49305698530edaa54e4cc8ef33aa63f05
6604 F20101108_AAALFB camaratta_m_Page_111thm.jpg
5620fb80de0ac183f04106a70508528c
e31f24cc3643ec6055ebfb0299ee89645eb96425
20948 F20101108_AAALEM camaratta_m_Page_005.QC.jpg
b572bd15ee1742a404b45ca4b2e4ddef
ca709d170616f20ff62e39f355fdb65715fae43a
1125 F20101108_AAALDY camaratta_m_Page_152.txt
f20b75cd7405136e460faaea50abed54
47c61843e8307dff893bfae45269fcd3cc3ff3e5
24338 F20101108_AAAKZG camaratta_m_Page_052.pro
f6cdc1ec906908f289615e54a820b628
660a07ee6b1d43f548727887c9cecc71aa70fb5b
F20101108_AAAKYS camaratta_m_Page_158.tif
8200eefe137ec25e772a043cd1e10fcf
7a6b4a5db552a15d3a9116ab383fd2adf3239f11
26349 F20101108_AAAKBK camaratta_m_Page_060.QC.jpg
4f481785f9237c4f3f63c120a8a71b60
95c62505e3161f5a87aeb40bbb2acb3e20bf5f85
52755 F20101108_AAAJWE camaratta_m_Page_032.pro
f7069d3c5275666cbda0cae26d0e20a2
0a801fc21175b2a5687c61e8550f550e4fa06997
27384 F20101108_AAAKAW camaratta_m_Page_144.QC.jpg
3d5bde3b827d7115b6d053f024dbc8be
3e15330641ec2c6d2b98de27ee562aef7d942dc5
53810 F20101108_AAAJVQ camaratta_m_Page_156.pro
8d0c3b45fb8683ed52790b6d4c59100c
a15dc5ff96488c7ac085dc2474ece371233fb124
15400 F20101108_AAALFC camaratta_m_Page_133.QC.jpg
9efe426ee3635fefdba32f0bb85bcb3d
59e07e793d35137afa1b9006ec7f48659fc849af
10460 F20101108_AAALEN camaratta_m_Page_018.QC.jpg
43ed674ad7350a8848d26cf5e43dc4bf
ab89b215f9f79770aabfcd50256b64b91b8d437f
1788 F20101108_AAALDZ camaratta_m_Page_153.txt
09c11af8e31c3b0bc5690419b2654f9a
72bdec35a39e401f3b61aeee70395ec05ffa0e70
12604 F20101108_AAAKZH camaratta_m_Page_053.pro
ce3ae9941cb89190bd03480e8a7ba1c2
6fb6a65f258c4c2ee2f37382cf4bf104bd806de0
F20101108_AAAKYT camaratta_m_Page_160.tif
10152da3f00a8c7d317f52a6ff044073
e6975a259e60838df0edc3ac80f9ebaf929f66a2
F20101108_AAAKBL camaratta_m_Page_100.jp2
52cc35358a53862052fb9301fd81af1c
5b1efbf4f5c82da8de60018ba0adc84b26328716
21137 F20101108_AAAJWF camaratta_m_Page_151.QC.jpg
6de56354eeb7fb06d929abe2ba4ac758
380f9120f7a8bf4b52ba8231df96686adbb7ab2c
84322 F20101108_AAAKAX camaratta_m_Page_142.jpg
f62240e918d11fb6a31654fed072d2e1
7e4c252b4ee42041240ff7e8e67cddcfed627c1e
F20101108_AAAJVR camaratta_m_Page_134.tif
de6e7a525dc688fe5a25f73d63b29c08
6e8f5e2b646f0bd2d7769d9a3a374ac3e45d52f2
13984 F20101108_AAAKCA camaratta_m_Page_152.QC.jpg
83b76cc3e7717eb5f07cf4d97c8847bf
75499dba7193f5c734d7b4e87e92662a303acfc6
6436 F20101108_AAALFD camaratta_m_Page_091thm.jpg
682395b4586c9950629b3ca68faf31a3
33242ab35d32fdaf69e78e32315ef26816c804e4
6775 F20101108_AAALEO camaratta_m_Page_039thm.jpg
4f66046a8e8edb95da9a1878c7aea8f6
b68a9d3205bb41d39c9cc5b105eb1bd17b001036
18490 F20101108_AAAKZI camaratta_m_Page_054.pro
4153474a9df3d59c92c557c4895dd11c
3cd833c351b4b9e36a3ea6b41fd6252f218bda27
9107 F20101108_AAAKYU camaratta_m_Page_001.pro
f412717b96d5892d2b451ca25725e10c
3c45af4baea6d70f5d14247312e70bf8bfc80848
96925 F20101108_AAAKBM camaratta_m_Page_031.jpg
d3268e755537f451c9854c3002ac65aa
f08438b8550a8dc5581149431f2eebdf2d57676e
67381 F20101108_AAAJWG camaratta_m_Page_104.jpg
2b74e3f0bd976a44ef61afa303717449
8d5d79bbcf8537c8561093f010e0dbb6de984777
47291 F20101108_AAAKAY camaratta_m_Page_121.jpg
a7abda8789d99a59d3be887bbb979375
28c3d504149e04c0a7a71a598a7589852fd0c7ba
4242 F20101108_AAAJVS camaratta_m_Page_073thm.jpg
e17a0171fafe007dfc4016d7250b4b2d
a0fba6d49df0a3c8c6f0e79e55545f4fd1e31a24
20811 F20101108_AAAKCB camaratta_m_Page_057.pro
6e86cd1d48984994a28a4e603da645c9
0f8a91e251d8dafa5b878a3bda2b0cab675a601b
5040 F20101108_AAALFE camaratta_m_Page_123thm.jpg
3ef2df8913722deb306e9d89233959e8
bbbff43daefc31984835516c717856bb0c61ca09
20212 F20101108_AAALEP camaratta_m_Page_044.QC.jpg
3fd03280bc1f8020efd85a8fb4e2002b
2aa1795a4eda13d5f082f92afd104481961e4b57
12061 F20101108_AAAKZJ camaratta_m_Page_055.pro
95f28a01897b881651bc4b0eec113679
acdf4eba1717ff5d561aaf7c7b21ade8cf02786e
56524 F20101108_AAAKYV camaratta_m_Page_006.pro
994d617bd399a1bc3338c04e7bee29ed
890a1d0b10bf4e2cfecce3e193c22e29ca02239e
28665 F20101108_AAAKBN camaratta_m_Page_100.QC.jpg
98d6c274825760e8eb0c86bbe05c1ef6
cc7336f6d318fc8c220af573ee9e9e8dcc7e834c
942511 F20101108_AAAJWH camaratta_m_Page_022.jp2
f433c87c8d95ad33ff7334604edd438a
cce6d3079ff5a6fb6a85bb656e29409010f1d508
6705 F20101108_AAAKAZ camaratta_m_Page_047thm.jpg
b2134e0ed9a9755a641a0ac0ecaaddae
ad1e7f03cb5f41c20947ccc176b8b52af6e58211
51418 F20101108_AAAJVT camaratta_m_Page_108.pro
4d123cb6f79cdd1457171a3e4dc449af
8bab897ab5fce58bd04ff80b7af25d54f8cb2d8f
F20101108_AAAKCC camaratta_m_Page_067.tif
91bfbad092c85590df4fe65b5cf88ae9
08c1a5ed3f4d6b0ad051ec77f17e7306c16f4877
3517 F20101108_AAALFF camaratta_m_Page_094thm.jpg
2e10d1e22f4cce29706172e3dab62337
088b65a8aa342ec178b2ca8e52932ed0a7e1a1b9
5052 F20101108_AAALEQ camaratta_m_Page_095thm.jpg
e36124a61166cebfe589884118d01659
9c4882ef6c742ca0d9da7dc72ac81c2bad3edfef
11534 F20101108_AAAKZK camaratta_m_Page_056.pro
f12b6e34268ed4eaa51aa144be90243b
867cb8d942855d60339e704e7f671290036d2f25
10095 F20101108_AAAKYW camaratta_m_Page_007.pro
19f9fa2f3a863e955a367db6c15597ab
e901a6492d83f680cfae26f35e95f22caf8b0d19
F20101108_AAAKBO camaratta_m_Page_034.tif
3914b82dd0558bf491e457ad67b9146b
14e1c1276382b20d89fa6cf3aea18bf6fb66e786
10655 F20101108_AAAJWI camaratta_m_Page_132.pro
d53c83b9b99454d4782f7a81c70f42e5
8707ee6896cb616d820b4a9033c31528d4965d60
27495 F20101108_AAAJVU camaratta_m_Page_085.QC.jpg
734c2a5e982da827bb3233e80f13d7a2
7e30bdcf4a80cdaeb31ab70e4485f8bcccc3774a
980468 F20101108_AAAKCD camaratta_m_Page_150.jp2
202f00eaaebfc3d6d13f30f12a9c5429
a84817f8fabe550de28cf3c080b4db87a6950312
6494 F20101108_AAALER camaratta_m_Page_085thm.jpg
80103ccab56b0b9d34855c134b67eaa2
109c8d5a9a18df68d687da8094a5ea7daf5bded5
52960 F20101108_AAAKZL camaratta_m_Page_061.pro
828528bc2c0c6907ac9c5bd5351b059f
e99f59b7304e91787f3dc7b40f5b8bc0d48df559
63523 F20101108_AAAKYX camaratta_m_Page_008.pro
a5aa8c0bab270f04d3497f11a310d1df
1f5b3c997fa6df33fe17adc7b7c85e70271ecd83
691878 F20101108_AAAKBP camaratta_m_Page_089.jp2
e779c604a751762bb56144cef7cae4fe
1b6fb7197ed524e1cf90a631475251cb4583e5cd
22302 F20101108_AAAJWJ camaratta_m_Page_150.QC.jpg
427d9a250542e28737862b48b29ba7d1
8f296df1f06d5538cedcfcc2d5191791cb2ddd91
1088 F20101108_AAAJVV camaratta_m_Page_048.txt
21fc00ae32b16167971d8f9a476b0573
b54d0b55aa0ae105f19bff7aa5ab2e8ee704a174
26597 F20101108_AAALFG camaratta_m_Page_032.QC.jpg
8335cbacb8b7c1b48fbc431271e542aa
ea66658a466c0d2cbaaf8da488f0c5f474527308
33779 F20101108_AAALES camaratta_m_Page_011.QC.jpg
2577c06504db2c55a1ce593188fed19e
be1c1175d43b4091f6a367256ff7940b1f91d23d
46407 F20101108_AAAKZM camaratta_m_Page_064.pro
920aee7601b3a2352b87adeb93feddab
e8a5a1b1388713a82f638a12a503340734c349fb
43823 F20101108_AAAKYY camaratta_m_Page_013.pro
181f3687a8efc55b4aa2165b9e439837
077573665b218bc43c9bab4a3016cb2a2ab1646a
415547 F20101108_AAAKBQ camaratta_m_Page_076.jp2
d6c8ab72d2cae40d0c95ec552bcf3bab
e0206726fc686807ddefd39d98c396bf2f3dd60c
49105 F20101108_AAAJWK camaratta_m_Page_081.pro
e630e37b8fa4a994100f89a579005c8a
3607ccee70e4c4b4bd612fd2f50df6464fb120d7
500499 F20101108_AAAJVW camaratta_m_Page_135.jp2
2e94ab3e38279d31d49de8a23a8a5913
55b9a22fedc50fd80073ba2bc243047a85b860b4
121845 F20101108_AAAKCE camaratta_m_Page_009.jpg
953df75886afad1397ede44368a54103
e741fcaeff78004fb7086150a98fb3ce802e08fa
5920 F20101108_AAALFH camaratta_m_Page_034thm.jpg
d61156cc4682e1450689ecca656c0380
3935cddfa55a652e8210852074d136f1e5a5d45c
6773 F20101108_AAALET camaratta_m_Page_157thm.jpg
5263c5359c632011cf204cde32477016
c024a4a6eece098591f61a8683bd1cb6e187626e
14971 F20101108_AAAKZN camaratta_m_Page_070.pro
8275712b7b3b8bfcfe4701dfb15c328d
2157dde7550ad262036d403ae72289d9cf00f427
17273 F20101108_AAAKYZ camaratta_m_Page_018.pro
3674c11d7c138c8e5317db8490d7e3fe
1dc5807595e82993e09509df0d4ecce9125b4b93
F20101108_AAAKBR camaratta_m_Page_102.tif
e7c57d0dc9fb42de0d47806343dde9e9
70eef1f6079cde8e09e25c5740b389d0ad1a4e99
7414 F20101108_AAAJWL camaratta_m_Page_009thm.jpg
98da0131f6a17639d1915e993fc5e2e4
1915557bc043746db0de4500dbe365779bc0c8fe
436223 F20101108_AAAJVX camaratta_m_Page_122.jp2
ddc43c5c7935043aaa95c1bf6c608a7c
5ab7fe1c2f0c171f3a3748cf7c5fef3f6e1f8062
31649 F20101108_AAAKCF camaratta_m_Page_122.jpg
70c7fd2389806bee3f775dbcf32e94a5
a1246dae95738e9a0f5f5e1887681f0f31ab4caa
27753 F20101108_AAALFI camaratta_m_Page_111.QC.jpg
acceb695b9b2b4c233206543941f430e
14b3023824bdd678e272e733d6391e535d2c6a10
5557 F20101108_AAALEU camaratta_m_Page_013thm.jpg
b41ac967a7520254244552ce12a8266a
8662ff758f40d95d023d92fc8efc3042adc17cce
15798 F20101108_AAAKZO camaratta_m_Page_072.pro
8663c4007249b2d75ec6a0e061919fbc
85b667d5e6ab8f6cb97b3798fe8ef8ebe219364c
32353 F20101108_AAAJXA camaratta_m_Page_130.jpg
a6ffa65c7d4ec2d09c52ab10c985c718
a66ec0aa468e9e65b049163ee2aee7db50a25446
1724 F20101108_AAAKBS camaratta_m_Page_020.txt
0a516c9cac35026aae02cc815ecf4d1f
6985c0b468351b830fc217197c110115e3d76a0a
330642 F20101108_AAAJWM camaratta_m_Page_127.jp2
8e7b08f46958bbd6a73ee2ce15cbb8a8
75d8581bb1838aa5c19a62b0dd88c7582aef8bb8
5232 F20101108_AAAKCG camaratta_m_Page_033thm.jpg
e78c34edd2ad2c428f34ed205e1b29ae
67c68922c81f6724349f744cc71112ceb2c01fd9
28176 F20101108_AAALFJ camaratta_m_Page_139.QC.jpg
61172a54247e44a2d37a635ab4238481
cc8c8b306a73732511322637bb710a2dc865b366
28535 F20101108_AAALEV camaratta_m_Page_040.QC.jpg
67a191a088041eecfa34860a0d7aa70d
631756b45770c386b5f28992be86b8cdef433dfd
10853 F20101108_AAAKZP camaratta_m_Page_073.pro
0dd8f29616cf5f1836810b5bcf0d6af8
20a82792aee2f02e0dc093e9e932f77f1b8f16a0
45475 F20101108_AAAJXB camaratta_m_Page_034.pro
6627b41c816943ec08bfb31436e08033
d506705eb82342e0d93b962ac72a53e18bc185e7
49454 F20101108_AAAKBT camaratta_m_Page_004.pro
6498d150903b8e8b18a2d45118dbf865
3863937e5f58229a703bd6ff800a31d74892e9d3
F20101108_AAAJWN camaratta_m_Page_136.tif
ad4b1917fe595332bb2c603e6144960b
0c92ac82db683bb23c4aa5ce2fd9303fe1a973ee
1011 F20101108_AAAJVY camaratta_m_Page_058.txt
a5a6de431a5142fc73322a3eaa7c2bbe
1042d354437c79489f6e5d2c224e688324ef39fa
F20101108_AAAKCH camaratta_m_Page_017.jp2
688dda66689aca9af0688f545c6a893d
0ad3e80013b9eca9abdf50dc78f9ae20277a3f8b
6683 F20101108_AAALFK camaratta_m_Page_101thm.jpg
e4a9217deedef9516d462aa4960d1316
f4f88df0c909033d48854ae703aabd594cdd07ef
17041 F20101108_AAALEW camaratta_m_Page_080.QC.jpg
adfe445ace0139735a3760254c7c2f74
0ddf6d52b6f8a3620b149caacd340557fce906cb
9684 F20101108_AAAKZQ camaratta_m_Page_074.pro
71af7da384bec9938589a30830bed6a3
86cd386dee03eef4ce3354fe20f23a96623ad88d
29869 F20101108_AAAJXC camaratta_m_Page_141.QC.jpg
60da3bd767bc7cb7775f82ad2fe7ab76
a16e59303eca4fc0d40c46b21317e409ac016f96
F20101108_AAAKBU camaratta_m_Page_107.jp2
adb88d7e12f9c7e44591c5b81c901054
d247bc9f0fce787c779938742888b8052e0c05b1
4084 F20101108_AAAJWO camaratta_m_Page_056thm.jpg
54a0224d9ec57e35fbd0cd1bb2e1cdfb
c10461a12b2581b941e981145c865b66e06da355
23778 F20101108_AAAJVZ camaratta_m_Page_024.QC.jpg
8ca7579dd9e90acaa93d390393a51331
67cba75de3374513221e974791d4b23810cb3a2a
52237 F20101108_AAAKCI camaratta_m_Page_100.pro
6150c65bd582fbc831cb32a61c12b866
81cc62d3d6b6d2447208130bd80b44801de6fd67
5649 F20101108_AAALGA camaratta_m_Page_153thm.jpg
4b8e6c001761ef06ca385d185c50429e
a642dc16e50130d26f9575d21036cd65e089e19d
27518 F20101108_AAALFL camaratta_m_Page_108.QC.jpg
3ce4ee000861f881481e708ce80aefe6
b5dad72648357de3758580f57ac04d47597c82c5
27386 F20101108_AAALEX camaratta_m_Page_138.QC.jpg
db110a155253219175d43faed207ad9e
d2ffb5d981a9232488ade403e1ea915cad24c72e
7131 F20101108_AAAKZR camaratta_m_Page_077.pro
f37b3c457d346a347a4a832c4f09978b
438ee345ca5ac3a5352430d85a46dafd3f7f1b2b
36080 F20101108_AAAJXD camaratta_m_Page_055.jpg
c769001e240a2af32121b4fa82769a3d
8cf09be6bec5367ae4bf0b3526961515174b9214
83633 F20101108_AAAKBV camaratta_m_Page_081.jpg
b3931300052abe26346991cb05afa2da
0d7b6918d1ef550041de1396c9d7732b3aa435c8
92334 F20101108_AAAJWP camaratta_m_Page_065.jpg
36a06fb7b67095aa0299b0145789785a
90e619112e3eaef849cf94c3ca8c5e8f8a9ecc82
40048 F20101108_AAAKCJ camaratta_m_Page_090.jpg
bf3e4da51c246d0dd821fda1fbf5242f
77af04e7a7ecfedb5200e5067396f6228b9e83ed
23484 F20101108_AAALGB camaratta_m_Page_149.QC.jpg
1f407a17ef10b5c80d142398f3957241
df839b9d280afb3c013e6c924136952d08098460
4913 F20101108_AAALFM camaratta_m_Page_046thm.jpg
19a4f548838218b122ac7cae6f7cb3db
5f0785937718079ca0ee9c355f9602eec4a4099e
3498 F20101108_AAALEY camaratta_m_Page_055thm.jpg
f6a6a47dbc742792043521dcdf5ea5ed
0d579d8574497473b6bba80e339c559c23b7463d
F20101108_AAAKZS camaratta_m_Page_079.pro
c2454a13210e0e55870b34e93aa3b679
66d8dc33cd7197d6235005be266026f89f499a8b
F20101108_AAAJXE camaratta_m_Page_005.tif
a97cbe1d1ade9340bd47a02263813e15
1d7d91b312af1a140b11cda53e2fea79c2c66c22
989522 F20101108_AAAKBW camaratta_m_Page_024.jp2
fec6607a9750ebe1b0d84a285a4c316e
35c84e6dd1fdf7c947ab6ac32a961e4194db2257
48372 F20101108_AAAJWQ camaratta_m_Page_110.pro
4ace9fcd67835680c7a45dda23a99b9f
f87a680d10d73bafbab19564044862749b39d801
1927 F20101108_AAAKCK camaratta_m_Page_083.txt
06b045832e538009aefba8bdc36a20ad
cf44d298565d64d69d923c76da6b1616bf4def24
6188 F20101108_AAALGC camaratta_m_Page_099thm.jpg
b40893603213e98a8010226facd01112
79b6416fc311e96ef39e214b8522b4f70a537443
13551 F20101108_AAALFN camaratta_m_Page_137.QC.jpg
9a8576bf3f8a9b26d6eda5b8eef932c1
f93883656b0c3614315cc641362d5aa7c2e6a26d
10313 F20101108_AAALEZ camaratta_m_Page_130.QC.jpg
b427a633763cb8ba27d5702582f71589
5a3bbb206360822c3ecc0d2c5c70d40b1c9ddf8d
6096 F20101108_AAAKZT camaratta_m_Page_080.pro
60f263731f820a80b3e4405fea9ac4cb
c401874c15348986aa63cf85f2e162d3077414da
575676 F20101108_AAAJXF camaratta_m_Page_152.jp2
952c0211e798e80e88cfed349d7463e0
2df2da463136ba70545ff9c0596188abc5abb2ca
F20101108_AAAKBX camaratta_m_Page_117.jp2
281d457f52ad8f17a6853e7d283ff85c
3a1d195f15d87923af3020e6d449a79e39d23759
40068 F20101108_AAAJWR camaratta_m_Page_020.pro
cca617570ebccb1ff3bb75ca88007ebf
6cbb5ba9881ef73481567f320c6f2fc514242615
F20101108_AAAKDA camaratta_m_Page_030.jp2
ead9b47f501a82058aacee3b13e36b47
cdf69407c9fc6550610966d1a6de53243b56b8de
77582 F20101108_AAAKCL camaratta_m_Page_010.pro
81c11bc4b8484d3c44f164b1ca422654
821587c33611d3546318fe37549f59fc7943eb16
6247 F20101108_AAALGD camaratta_m_Page_093thm.jpg
4c0c3d2fa515f6b3d47a0487ee17e778
a6bcc295584bc3a05f2a286f64a7c6672dd05c7c
5569 F20101108_AAALFO camaratta_m_Page_021thm.jpg
af62d7c003446eacf50875928e9e0bbc
5d23aa3714fa198a00ac0ca877789a99da1ea41e
52283 F20101108_AAAKZU camaratta_m_Page_082.pro
b5d226dccec51171dc20d3834738a007
a757180ba963157362ef6f661f9ed8b7f1704aca
911 F20101108_AAAJXG camaratta_m_Page_002.pro
ac86ab48275de3d58c94c713a051f770
e67012d0df34eb742868e80b42df3d68e34babc6
626615 F20101108_AAAKBY camaratta_m_Page_095.jp2
bcffdd8d2fb06ba31925bdca282bb8fb
9bc283c98bcb833f1843dbf5e0a7b528846e966b
25360 F20101108_AAAJWS camaratta_m_Page_134.jpg
b4ddca816e455108710fa44b345bd225
f0cb8f8f335936300532c4b65a3931c64f0f24fb
35438 F20101108_AAAKDB camaratta_m_Page_088.jpg
cf83051ed0e906d95c5ccabefd8a7409
d572f37998d5d3edb69ffaf6b96be972543f7386
F20101108_AAAKCM camaratta_m_Page_060.jp2
0cb96559a1fdea053b6054d468e7855d
fe429623e2d91499248cd6a31c47049d74b0f324
18056 F20101108_AAALGE camaratta_m_Page_089.QC.jpg
cdad12f7814acc34d0844ea17eacf22c
dc148d3f8eba027dc29f36ef86dfc7e1baad7220
22570 F20101108_AAALFP camaratta_m_Page_013.QC.jpg
b3f20b0ad4f86225079e4d87cfe9f81f
980b43ac1d0a916ad3c7c95711d654a475cd18cd
55153 F20101108_AAAKZV camaratta_m_Page_084.pro
6c7ad23ace66c29f9da1264b42b33e0f
7e5bbffc58275ed698f7c11455256a81309a04ad
F20101108_AAAJXH camaratta_m_Page_103.tif
f632ece1a155d65a71d856fc34e1545c
a47d57602795d43e7ce2b3fa4842cc04e052f2ba
54954 F20101108_AAAKBZ camaratta_m_Page_037.pro
7a5651f89cc190219dad11b1d27fa0b6
6f465573e416f8c2f5e8603a774b6cb5e8aa19d6
1051906 F20101108_AAAJWT camaratta_m_Page_125.jp2
421f4dfcadb2fe46bc74f269911cbe17
2949e697210ad8adb94778daebb7ad665577f5b5
56099 F20101108_AAAKDC camaratta_m_Page_017.pro
c41749b4c9fc394474e997caebf46ce7
204446d212e69166d35b406823e5f8efcb25e052
27007 F20101108_AAAKCN camaratta_m_Page_157.QC.jpg
e160650a82588cba5217c7dfb238e53c
34324363db634340fee4ea4f2349228d66a3e07a
6440 F20101108_AAALGF camaratta_m_Page_143thm.jpg
877fe5f8573aa43c97994bb83b698999
37922d3c6ac9832d924088d852a3853725a551d7
5252 F20101108_AAALFQ camaratta_m_Page_048thm.jpg
28413e4ea80888e9108342e3d16558b4
eb674ac4344e528915d10d3c2a7d9edf15bd701e
10627 F20101108_AAAKZW camaratta_m_Page_092.pro
cfc515a0a5f17afb4d484ca0112d6ae8
f68fb0bcf90eaba5d3a1954fe38664473a0eb490
16719 F20101108_AAAJXI camaratta_m_Page_092.QC.jpg
4330b2981024e356c3557dc0d513ff9d
67bfceb9c2cc79d2db62ac606ef12933871aece4
94541 F20101108_AAAJWU camaratta_m_Page_028.jpg
51fdf9c0e5d32d49abc738346daba6f3
73777c8ded212a31d63f2d94b695e88fba216ec9
476 F20101108_AAAKDD camaratta_m_Page_096.txt
6fb7e7a61dff446ab5f7b2e52eef0c0b
b4743f86b5ef9332e95cd63f76ec5af97cfec0bb
25694 F20101108_AAAKCO camaratta_m_Page_147.QC.jpg
09743c8d0a421c62102e93a369caf758
e07b16d48391eb2f884d3307c9d08f0510071e8f
242526 F20101108_AAALGG UFE0021667_00001.xml
881ecb551fc7da24de2b2f282d82ff55
1454a122100592d8d9f2a0808ec0895b7f7c646f
5242 F20101108_AAALFR camaratta_m_Page_005thm.jpg
fd8a7b8fa8616c0d9cef4b6e40de7db9
27cc96245e7d25bf6770ec9770ca361cf003ccfc
17542 F20101108_AAAKZX camaratta_m_Page_093.pro
9a064677ce9da16f8cf350c443cc6e7c
99e365491dd9754791c56def8296555e0d2ded65
6964 F20101108_AAAJXJ camaratta_m_Page_106thm.jpg
5607e441cb5d4628c21a2875afd07bc7
a0c49ca28f7ec363ac18829ae1e4edc4db264cb6
23468 F20101108_AAAJWV camaratta_m_Page_047.QC.jpg
d4ee6831d42c16ec85b71322d87320e0
eb010beb6210ecf891b1484d513869a0772ea894
F20101108_AAAKDE camaratta_m_Page_069.tif
2e21cade6e1302e60dc5199804744362
69995bd4a3f6868b21cba2bfc9251a918abf72ff
F20101108_AAAKCP camaratta_m_Page_080.tif
c8d2ca76921f590e7ce32f3a867f0d7b
6beb084aa93198e4dfdda95004fb0c0049434f27
4221 F20101108_AAALFS camaratta_m_Page_152thm.jpg
60948a58dd17b77cb12a084f534ab624
5af966c7f933bd291cc913ec3bcbeb8d93643717
5328 F20101108_AAAKZY camaratta_m_Page_094.pro
214ba3e1d38d4a16deac5f289be9c5a8
6ee349c7c7668c76462344060f3a8273654d4f45
28332 F20101108_AAAJXK camaratta_m_Page_101.QC.jpg
43b83b37b59692cbbfeb15f3f5ddb102
1ab9a73c37b8cf4725c0ac50c30272a1bf60ab38
1051877 F20101108_AAAJWW camaratta_m_Page_124.jp2
7fccce5c3e8b7dcd56a829cfc0ed2636
996816efebc200637a23e9d63d93f23440438ee0
74036 F20101108_AAAKCQ camaratta_m_Page_011.pro
918ce1e61b258e5aab1a4d6cd5c2f776
558abec81b624ef847786cef3a421c759a7b5f2a
1304 F20101108_AAALGH camaratta_m_Page_002.QC.jpg
8db691e794c55ae692d6bb3c25bcf307
c6169e9f003d7cd42e449facee529c64cdcda6a6
6789 F20101108_AAALFT camaratta_m_Page_043thm.jpg
10cf4e5d46635ff170f60f2d27ae536b
8d7475cc0cb5dde740aa2649cdada816e7f42720
9936 F20101108_AAAKZZ camaratta_m_Page_097.pro
01021401b208734969c17388ec575e07
2aca2f19f074c1d274cbc39d933af8e2fc9ce57d
29195 F20101108_AAAJXL camaratta_m_Page_105.QC.jpg
684eaea146862d54541dd103c8161e41
464da461fc7ac8aa4e1090238867a5d5520bc4ea
F20101108_AAAJWX camaratta_m_Page_019.tif
c288b5705121610e2445c32b61b15b3f
cbcbd53d0dc28cd00eb1b3cfa90f9459b4750292
1051980 F20101108_AAAKDF camaratta_m_Page_143.jp2
a0d7c3410268ea31eda30a289a85f017
b1cffa1ee2785c4a1dc6120d81a4cea1dfa76042
22928 F20101108_AAAKCR camaratta_m_Page_103.QC.jpg
53bcf809d8c6ee758cac69dd483a1e36
a6eb9223f634391dd6ba9ec86e5be402511f62c0
F20101108_AAALGI camaratta_m_Page_008thm.jpg
4dd58ebbbf7ab7381f984d5774a0f4d5
25744a48569cc13d1d3324e0b2b3fdac4d06c459
5607 F20101108_AAALFU camaratta_m_Page_022thm.jpg
846b173f9ccae84c0bf2bcfef923eafb
58c2d1236378a3b0d6ca010e9173e81fcb63c664
F20101108_AAAJXM camaratta_m_Page_080.txt
83b8a0d402becb204795cd0eedb1a887
714ff238ce88f79642cc7ad734007092fea8cc12
22344 F20101108_AAAJWY camaratta_m_Page_049.QC.jpg
cf689883b9a2b4d0dbf1be42cb31d240
f6203dbce15c405b840a850b7f637186a3af4f82
88082 F20101108_AAAKDG camaratta_m_Page_042.jpg
8e74c9ac1d1c4b4104c505fb3e0bef93
d7e0b2089fdf41a38456dd7ebe6e107d05eabcb0
10498 F20101108_AAAJYA camaratta_m_Page_120.QC.jpg
58071110a686e8a470945faefc0d812a
334c21a2d05f750dff0c44274af0581c5846e9de
1051931 F20101108_AAAKCS camaratta_m_Page_061.jp2
be59d4f00c38913daa9add70dfd5cc4c
d4abd1276d7f72a262a42d55557875b94b9f1c49
7584 F20101108_AAALGJ camaratta_m_Page_011thm.jpg
7552a89b6111bcc3e6045b44f94d3fd8
a9a1c456c7c528599df6d0ec4a3250e1c6a015c4
25895 F20101108_AAALFV camaratta_m_Page_158.QC.jpg
914bebc1e3dd59cbc2af5f1bde33ac3c
169c2da617b18c38f7cab7a95f165163021bad75
5277 F20101108_AAAJXN camaratta_m_Page_159thm.jpg
d6c7842dfbaa0564b22990c1d0951bf9
718161c2a85cd7d019fbe220dec7e012a6e05979
F20101108_AAAKDH camaratta_m_Page_124.tif
2ad0f48b57b5f4fc7fdfe82a8d0df90b
a867a6681fcf0545e939e648fef158b36724bd51
F20101108_AAAJYB camaratta_m_Page_096.tif
85026c6c1b7352bad29968cce70b9864
fd40bfc1e110a3c5ef93de50653b803b6fa18c03
90325 F20101108_AAAKCT camaratta_m_Page_111.jpg
294d3927184ae607ebbffdc0ac5a540a
4cbbe4e2e65003c9bf60a6b3ef83ba9e6b4d6a83
15167 F20101108_AAALGK camaratta_m_Page_012.QC.jpg
148eef027e33d040d2e650b7b8f6aab0
dc6504820b1455dba453e4935c7c19e3c2453381
16217 F20101108_AAALFW camaratta_m_Page_052.QC.jpg
eb3955640531d2c7e067cb5776c788a7
8e98afe18689c4e6d79e9d97a177c0cc1f79ee70
50101 F20101108_AAAJXO camaratta_m_Page_157.pro
d4da73ed599d3be1066419395fffb306
13ca704f6a86eccb3452a8f432380dd8b69b7862
F20101108_AAAJWZ camaratta_m_Page_062.jp2
70f8a2fc6d02a09fce72052521d7e35c
18228c3cb9d2b8ddb3e16bc41ca5489c76d14dd1
F20101108_AAAKDI camaratta_m_Page_155.jp2
c82664dc062dd07ee84aba8fcae8700b
63e9a92f13e22a51f5ede2dd0851fd5edfe228ad
328 F20101108_AAAJYC camaratta_m_Page_125.txt
b41a47a07b3c2377c4e2b3dc979ece3b
e6dcee65166286db29c66cc60282abef74646876
57613 F20101108_AAAKCU camaratta_m_Page_079.jpg
6aae9feb7dfe770acc3b9427b903ef16
f99cee851a19bbd9533ae98661e342883cf9ba0a
4510 F20101108_AAALHA camaratta_m_Page_052thm.jpg
47d60c7a21640c08492f4217e8b235d7
953a09964f8e20fd8b72c5897e505a481a7992a4
3641 F20101108_AAALGL camaratta_m_Page_012thm.jpg
f8eed4a6eb5cfa078798c380f5185194
fa006bbdbd6846c21b50d3dc1e430b55e32d2063
2981 F20101108_AAALFX camaratta_m_Page_132thm.jpg
c527dcae086342f9615d8c91d7e4b92c
0f1f442adf2237eaf3147014fb19162ba6fbf159
15140 F20101108_AAAJXP camaratta_m_Page_089.pro
1a9104c52affdb2bd63fd7acc90bd4da
1cb5cde327eb323f4b0fe9a5befcd7c1db534e0a
97 F20101108_AAAKDJ camaratta_m_Page_002.txt
23a81b80649603202730661e02249d98
ca774cbe8de96299d0a8557c1ef40bc30b56f61e
3140 F20101108_AAAJYD camaratta_m_Page_010.txt
6d43521ef7a1774d6581159fadfea802
98de5ca232b6b88b59a3c554b6251a12de1e4214
5829 F20101108_AAAKCV camaratta_m_Page_148thm.jpg
c515a85d2b02d14fe8aa982421819eb8
39f150c42a0f8da575436d45b2b593ffaabcedb3
12641 F20101108_AAALHB camaratta_m_Page_053.QC.jpg
4d4eb8c0757155732d994c505252ca16
acfda026b09d5c43116c69ee808938ba9c5df120
26894 F20101108_AAALGM camaratta_m_Page_014.QC.jpg
e2990a5ceb56847fe805da8d65a438ba
830f33e53e6a8792d7129da8cfd22130df05b57d
17163 F20101108_AAALFY camaratta_m_Page_006.QC.jpg
cd66ed6d174ff27bf284b724c4cc79fb
d8959ccc24dc46d35d4b8b8565c8819fd34a8bd9
52985 F20101108_AAAJXQ camaratta_m_Page_112.pro
d1456cf889fe38d780b24e2160ba849b
e0a87950d61c5a0ffde6ec85c77bd22523d181f7
774 F20101108_AAAKDK camaratta_m_Page_088.txt
d163c1ccbfab50818890b22571a35c6c
d1698cc8a9be09493e73a58ec86fce31eae42513
1615 F20101108_AAAJYE camaratta_m_Page_159.txt
f41b72695ee971313f66682dc90bb30a
9fc3ec8c00ff2deb038e8a7f02d7ee7959b0c891
F20101108_AAAKCW camaratta_m_Page_042.tif
d67ebadaff80d4f5edfe514f7b121844
f7e5fe0766e653b7624895681af774c44a363f21
3767 F20101108_AAALHC camaratta_m_Page_053thm.jpg
f9747db8038a2bff89a417222e2abb93
c034741f25d862a06959c0a5b1d472e9c0871e76
6531 F20101108_AAALGN camaratta_m_Page_016thm.jpg
efa7092d56d848688654552f51bf075f
1201d6786ca1928d4014e0b6c34475e34aa4b713
30380 F20101108_AAALFZ camaratta_m_Page_030.QC.jpg
c392a72483c1e971fca6bb6688f60719
e46979a0e81a7023e0abd0e9d7f3476fd14e90d2
29946 F20101108_AAAJXR camaratta_m_Page_028.QC.jpg
5696435707bb22b5aef523478cb48d13
675e40263348d49b3f9e1acf019cf20c22e43251
909343 F20101108_AAAKEA camaratta_m_Page_020.jp2
4bcefc677a758118445b030f47c779f4
ebe06b883e9bf841469ba76ac702aec27c32779f
19523 F20101108_AAAKDL camaratta_m_Page_088.pro
96f713e61055a041ce72e4025aa4e6c4
b0d57bcfbcf456efc583cd01f54aaf71c9c1b931
F20101108_AAAJYF camaratta_m_Page_044.tif
61052b0535a14347b9997148ef648543
87699a2de66054b57bf1246e17d76f0460d26e15
1965 F20101108_AAAKCX camaratta_m_Page_142.txt
9b5f4f4830e21ae8a1532dd545bcaa96
967e1309d7d5896f6460f3c7ba644d26fb053c56
17800 F20101108_AAALHD camaratta_m_Page_054.QC.jpg
3b74412bd0256679fa66fb617e8bf72a
d0cec62993d61efa9085fb7603dd24613b7d889f
28698 F20101108_AAALGO camaratta_m_Page_017.QC.jpg
fa27c9bfffd5da45177a7f46fd53b508
0afa61ac774801c3838270fffa2199a4b28d3b99
6537 F20101108_AAAJXS camaratta_m_Page_108thm.jpg
548ee2806a83d5140e7e1a6d95c104a1
7f7ba011b26798b49e4ee424cd48724ea828197c
F20101108_AAAKEB camaratta_m_Page_040.tif
1e6a26fc72dff12c9f0f210c92a614da
90f35d940e40c8a08b7e2a26922364d33a5cac00
2002 F20101108_AAAKDM camaratta_m_Page_157.txt
6f5b33212a199b2528f369b7662f47e4
cc7840cb3b05911235ee96e7ea586026194c9218
19720 F20101108_AAAJYG camaratta_m_Page_123.QC.jpg
f7adbb9faf247bbfec35f61750b38199
0b8c79c4d096e6837ff0725ef00cece63068d95b
30228 F20101108_AAAKCY camaratta_m_Page_012.pro
63d2a4dfa486c179c8c650fd886cfeaf
99e30d25b6257f79b070148e4203668570f9f1a4
11104 F20101108_AAALHE camaratta_m_Page_055.QC.jpg
6273ded60f76dce3fc324d101839a13f
3256f94a503cc6654708c3884a2df6bbb79823a0
22358 F20101108_AAALGP camaratta_m_Page_021.QC.jpg
55a140411d3bf3803989e057b93784fd
38cde0a45491dfb3b5226a71690f3adc7cc851ac
92672 F20101108_AAAJXT camaratta_m_Page_017.jpg
c5880b0e0882d15288db999ce05d534a
d30bf00ea7694b2c77cfd6830fc3aa9a2d103c5a
29502 F20101108_AAAKEC camaratta_m_Page_068.QC.jpg
b3512d15eee437b6cf4090647789b7fc
494424f7c5f7f0bc252bddef43d0c2ca9ed4da55
677883 F20101108_AAAKDN camaratta_m_Page_136.jp2
6b8f5585f6da8cb3cd03763556107c3c
2dc9413618f68f32c13b3bef091db93109b28813
3588 F20101108_AAAJYH camaratta_m_Page_126thm.jpg
a263073e7e38eb67081a54a3fb134648
c5706d503bbb77bd69d21180e428b0dcd4e4f06d
23275 F20101108_AAAKCZ camaratta_m_Page_114.QC.jpg
291fd76f14f028c4fbd84d0e55557d5b
9bce7bd823d8f971d486a3f94cc1f0c53faefbf0
14184 F20101108_AAALHF camaratta_m_Page_056.QC.jpg
f3f186ce40202ba0f9361dd4ce5663e9
3e51c4dbba95f8129b322e46052a8dbfd84117ee
5794 F20101108_AAALGQ camaratta_m_Page_024thm.jpg
605b2d1ce41df25e0704321b496cb237
76323b23b33953809dfd33f70a980d65a118580c
495 F20101108_AAAJXU camaratta_m_Page_002thm.jpg
1c3387b949d00d66064eb902266c0c41
521a8a232bf7967eddb205308228af7fd7b469d3
960432 F20101108_AAAKED camaratta_m_Page_160.jp2
a8f32daeca4dd1db4cd8c70db22b6ddd
8715d73eff66cbe1419732b635c8579682144ac1
2551 F20101108_AAAKDO camaratta_m_Page_127thm.jpg
34f89a7a50813bd372cb4ead7b24855c
f6bfcb1a983c6732dbd5881ccab07b526565a165
56653 F20101108_AAAJYI camaratta_m_Page_106.pro
d60d93d9fb41c58edd6717f63e511255
92367f851a74e4c4a4f0dcc5c64a67c1a6541366
4612 F20101108_AAALHG camaratta_m_Page_057thm.jpg
e12296fbcc37386aae41396740ca593e
85c63c7e14054004ef3a3c0c9adb664fe2b1f853
28428 F20101108_AAALGR camaratta_m_Page_025.QC.jpg
516fd668314829f7bbf8493791649374
bb9123c8cd3223abd91232c99e07b909a7691d15
96120 F20101108_AAAJXV camaratta_m_Page_005.jpg
536d296d9ad49d1fd88c536979229fd1
13c6b7c51f053f1e998958dd6386ac676b81ef97
6626 F20101108_AAAKEE camaratta_m_Page_113thm.jpg
0c3344cba105ca7cfb5e1abd63e69d58
5fb3b5aee226d7393cf03251aace85fee351f213
31294 F20101108_AAAKDP camaratta_m_Page_018.jpg
9ba43362d6683e805303fb9ae49547c0
1407abce879639088664501f762e91d0914cbdec
F20101108_AAAJYJ camaratta_m_Page_132.tif
7f098f743f6c89696ae56991a044d2a9
9c57f8831e48ef53194d1b5f874582b07b0bdf15
19224 F20101108_AAALHH camaratta_m_Page_058.QC.jpg
8948fd1f43e1438a6cec4d91dada75c8
a5eee0753f6ca64dfa5c79263373f52ec737aaa3
6791 F20101108_AAALGS camaratta_m_Page_028thm.jpg
76f49db3631ea9e45f2b552a8193784f
7e24b3909748dabbc9bda972a8b68dc975a21cd2
681444 F20101108_AAAKEF camaratta_m_Page_056.jp2
a931334026ad936e40c1e379ee3f0bfb
475623ba22ba6e3fc415b4e49d87d531e1bd77ff
28297 F20101108_AAAKDQ camaratta_m_Page_120.jpg
ddca4b793e852a91313dd28f02dbb27e
cfa1dcca62a142cda0e0317971780ff8e1a43e6e
2771 F20101108_AAAJYK camaratta_m_Page_088thm.jpg
cf84ad879b4fdaa71e0a136299ba2dfc
5919ce50353b1765875789fbe6eb6f9e89dc4852
F20101108_AAAJXW camaratta_m_Page_110.jp2
9ffa5509adc4de39a4c0287c795f66b7
b0fb18a98e3d6e1d176ee21493a56353150e0d79
28348 F20101108_AAALGT camaratta_m_Page_037.QC.jpg
2a27158d8592df4b393577406cae89ce
44ab8a33a7db7ead46ce48d8ead9455924abeeba
31829 F20101108_AAAKDR camaratta_m_Page_009.QC.jpg
e5431aef9f3b4d6c0852b13d5215639d
c7e7bb619047a5aff10f6ddb8d35199b292feeab
483261 F20101108_AAAJYL camaratta_m_Page_126.jp2
728eabe3061845807722dc0d3942d361
1c984336361072807dc973d2a32ba15b13b5b9c2
52962 F20101108_AAAJXX camaratta_m_Page_143.pro
d97c8a1caec8f9907fbaad0c4d53854b
fa91f2d31d9cbc1b658b1cc506ccca38114f17e6
5647 F20101108_AAALHI camaratta_m_Page_058thm.jpg
c606a123dcd62f461bbcbd6febc974f5
d819828222e0f5cb686732fc42a033d077fc6ff3
6737 F20101108_AAALGU camaratta_m_Page_037thm.jpg
d8432949c9974731d75a5f7a1ef9759a
1a9815c8fd2fa9d50b56980845eebb6bd4a031fe
F20101108_AAAKEG camaratta_m_Page_014.jp2
140b3b1686ee29622227ae28c343ecf4
728244e8c0b5114b0ffa51d6c410c859f636ddf0
F20101108_AAAJZA camaratta_m_Page_079.tif
3e8e6f34e8ce06bc549c6a29c43c3e79
ab7ea69459eef4196e0e6bc607ba734f2938368d
7064 F20101108_AAAKDS camaratta_m_Page_049thm.jpg
dceac69b106e69e2b71340facb22768a
87181cb40fcb6317475b99974f13aee782abb1dc
57362 F20101108_AAAJYM camaratta_m_Page_068.pro
5f108ab22ef96c9fa3a230a3dfc92ebd
51df6569c37dbfd0e8e7790243ea8dfa9bda4c46
6306 F20101108_AAAJXY camaratta_m_Page_004thm.jpg
067b6046530e18a96b4db4798384e607
22a7d01672c5e09df15047cd0ecbbcfb92d6ec0b
6141 F20101108_AAALHJ camaratta_m_Page_060thm.jpg
34c9218398f0862a5d662eb59c53d8dc
f3c1915459f99499c9a02e745d91f34bb8e434ae
27292 F20101108_AAALGV camaratta_m_Page_043.QC.jpg
4d1d0203b5f4fcd21acda4d47521c81b
d78aabd55155b6ef865bfe8f669ec753ddbca849
19488 F20101108_AAAKEH camaratta_m_Page_058.pro
ef02f7bcff9f3e9ccccbe21e8618d643
7485507b2c8ed99bae0ac5d8b5e50171dd1603ff
F20101108_AAAJZB camaratta_m_Page_125.tif
83c52940c490b366d91927593548f91b
089c14a725024335b6083902a7d27066202abc77
1891 F20101108_AAAKDT camaratta_m_Page_038.txt
dae596a9b87fcfb6f56382767351afa9
aaf97119b1dd6bea02301713c33852b8935c5059
4680 F20101108_AAAJYN camaratta_m_Page_121thm.jpg
fdd23bd2e827081984b9b85aedaebd95
64e27488954219479e0178be69dd2769d52a1660
5045 F20101108_AAAJXZ camaratta_m_Page_059thm.jpg
4b54715e0d12012579435eeab9da2884
9a17c80512f8eb628c09c07601b5785c1d1f03ef
6721 F20101108_AAALHK camaratta_m_Page_062thm.jpg
f75a829e771fe0bd7975dc223e4d09c7
cf76a230a5a4e96c557f24e3d7b7784c36b4d4a1
3909 F20101108_AAALGW camaratta_m_Page_045thm.jpg
9578002a41008055405b9308d7055b78
4f68d25ddb8414a26f71c169eb05f42d8b7ae147
26199 F20101108_AAAKEI camaratta_m_Page_091.QC.jpg
455e2beafb3323d19a156081cd429af2
71567a1a5820d0c36ca54e69b43e6e074dd3a909
43058 F20101108_AAAJZC camaratta_m_Page_150.pro
aee5861307bf1404e20976295651a770
789aa4866516fa26eff99ec4bef83b22296f012a
30939 F20101108_AAAKDU camaratta_m_Page_031.QC.jpg
790ccf33f27265fd72314328654642b3
bb54da1754297b68c1bd1c078e9a68e6d3b663e1
2160 F20101108_AAAJYO camaratta_m_Page_105.txt
96081c37fe609efbffddc0fe2688eaae
62ede1f7d924b47692d1134d909b5247c03cf5dc
7001 F20101108_AAALIA camaratta_m_Page_084thm.jpg
abdb214f7bb21f86430509f31edbfd57
73b0e87c22773853523e1a4450354dcbe983fefe
26399 F20101108_AAALHL camaratta_m_Page_063.QC.jpg
296edc924816b4d177bdd1fbea765393
0725ec19d9be20cca0aac410cdc736256136dfd6
16105 F20101108_AAALGX camaratta_m_Page_046.QC.jpg
28d649aaa20b8768cabc2f8fbab04780
4727619a1b03c5d8ffe8c85d1632c09306768dfe
48609 F20101108_AAAKEJ camaratta_m_Page_131.jpg
3a2a3edfa91d7289e23925509c54e549
187606331820bfa48315ae46fcbb5f2742b63291
2107 F20101108_AAAJZD camaratta_m_Page_070thm.jpg
39e2d43bac6122c7427c24c4dae3bd45
cb6ccb5281f6ca83412b3e591d7076a7131713d8
1843 F20101108_AAAKDV camaratta_m_Page_158.txt
7bddb9551fa42e01dab727a5258e4882
47a9567947d6a78ff109d9732d542b6df849fba5
2493 F20101108_AAAJYP camaratta_m_Page_134thm.jpg
b2f6b6cf2fead75f6061e41671136e51
2cf942e225952a467f8fa4640396259f99bf842b
F20101108_AAALIB camaratta_m_Page_086thm.jpg
2650022cadf7255d5ca2f563ae143a81
60632b766d70096ac6a106343a99161ec834e801
6602 F20101108_AAALHM camaratta_m_Page_066thm.jpg
b96dc8387029b06ebb4f48e50e38c461
bfb3b54cc3265af97272c674d9b943c852087935
17509 F20101108_AAALGY camaratta_m_Page_048.QC.jpg
37fee9ef509c69f446c30b23a73c3d82
0b11d64161f28c0f37621501ab8a79548ba05c8d
66316 F20101108_AAAKEK camaratta_m_Page_123.jpg
23ec9698ced53270ec7813093702f2c2
abe99356c36db3720b00411dfb983e56e88f8a25
1051950 F20101108_AAAJZE camaratta_m_Page_138.jp2
07e51bf47e2fab42843a93ef3a57bfd0
d1d96166547959776723a92374dd96ff8ac79062
F20101108_AAAKDW camaratta_m_Page_052.tif
e7172a4bb799ee5cc166eb935114de77
2830db76414a0470ddf23094266629151c9a9e9e
2106 F20101108_AAAJYQ camaratta_m_Page_042.txt
bc61820106edf79a70d2e2ff1c0e07c5
fc917ab1c6988e6f6523b99badf7f84be1928970
29916 F20101108_AAALIC camaratta_m_Page_087.QC.jpg
17644e073e0a61a86215a559d47fa0a2
581bfd04da467db29137046d5e8306cf965b501a
F20101108_AAALHN camaratta_m_Page_067thm.jpg
1129cdcf12afb0c9adf9847dd5e0d2b7
30323145cfd23fc7475ed8bca897476657568129
4765 F20101108_AAALGZ camaratta_m_Page_050thm.jpg
6ee7fcaedf8cb11889a8283a5f2e5e59
d323ae6e954983eb30f93aaa94f6b2270ba259ed
F20101108_AAAKEL camaratta_m_Page_088.tif
9fa884a97c2d8be47a336a58dff2f2eb
3a6493e41d8dae14b56f1bc3b52f0a8f1c2f472f
52800 F20101108_AAAJZF camaratta_m_Page_066.pro
946d361767913f3b200fa5d94716c2e5
695b673f0221464ea8a15680d167d9fa36bfcd62
572797 F20101108_AAAKDX camaratta_m_Page_129.jp2
def1e101bad23d9a1fdcd0a00abafe15
b63501687759d98e7c6db7224627c564a3eee235
F20101108_AAAJYR camaratta_m_Page_123.jp2
7befe5340c1f391e97d0ccd5a851866d
38e76923aa8ede19c932ed6e4ed5a18666805cf2
F20101108_AAAKFA camaratta_m_Page_137.tif
240d03f5b2c18d51e339b7703694654c
240357120ee789468fa1477f91b94ef1f0444b1a
6891 F20101108_AAALID camaratta_m_Page_087thm.jpg
7442964d64f6cdc379db4c55be2b5e0e
7c20cf0476858d478638527dbe99ab82594c269d
7072 F20101108_AAALHO camaratta_m_Page_068thm.jpg
34baa0a0c3a299e6de949dfce953c20c
3cdeb4c549a73fe74395055f6dd3b9594b86c0f6
8538 F20101108_AAAKEM camaratta_m_Page_096.pro
c6b534b78794794629fe9ebb3085a9b3
5473d4fe402d5598dc16c7756565996c34817c66
77528 F20101108_AAAJZG camaratta_m_Page_013.jpg
85a72bfe81ca7834bcb65117c623646d
85bdc1127d68f2cf09781c34f68271aee6c53530
54690 F20101108_AAAKDY camaratta_m_Page_040.pro
405a66ceae070d48455cac8dfce51470
a6feb2cad76c764340969519d376e5f6153ac07d
218099 F20101108_AAAJYS camaratta_m_Page_075.jp2
3b6e887350b47e0687d692eac1929135
322ab48b064852491cd5b457e00c2ece16631f41
F20101108_AAAKFB camaratta_m_Page_002.tif
3137ac315edbb7fb7731a1100f58633b
885159ec5c64b40577ae7550cad4ce2bab310d06
5514 F20101108_AAALIE camaratta_m_Page_089thm.jpg
e10e3ad118bda8bf351ce213e5798fec
c21d978abae05d0566a23c6fb1db8f12e4b800a9
6694 F20101108_AAALHP camaratta_m_Page_069thm.jpg
ddee55c91153283c8adea668c5896145
90351efbe2d7257ee13c999f490ca136ad439ab6
F20101108_AAAKEN camaratta_m_Page_156.jp2
c955e5300d275f31aa263c8dea79f75d
bb8264a1b698903b6a31b7cba69e3a30ef087c81
481 F20101108_AAAJZH camaratta_m_Page_130.txt
5f95792efd43ddcab641a1cb027876a1
16585ac11983168b71e016d6600c2ac51369fee6
F20101108_AAAKDZ camaratta_m_Page_097.tif
21fc0883121c1ea63be4831857975ce5
7f32d3f292de5590c98dc73743c085551483e0eb
5839 F20101108_AAAJYT camaratta_m_Page_150thm.jpg
27cfbed491f20b509c070da7854e7379
fe98c4bd0038a4fbaff93a1f0cded96f48e5a720
22914 F20101108_AAAKFC camaratta_m_Page_152.pro
1c3c314cef8d114b214c3f6250ac949e
337d5bfdbc78501c4816716127ec3ced0fb6cc84
24271 F20101108_AAALIF camaratta_m_Page_097.QC.jpg
927165b56c449681b4b20d45b0474fd8
274929f38c32853dce3e46dae26b375b4da71d34
12510 F20101108_AAALHQ camaratta_m_Page_071.QC.jpg
ccd132190fadbcc2108de81caf2d9567
27617225cac36d62ee39de2d0def424189615123
6557 F20101108_AAAKEO camaratta_m_Page_139thm.jpg
87853b74ad201cbd90ed8706ea73892d
8d5e81629834cdad963203ef86170aba78518846
F20101108_AAAJZI camaratta_m_Page_007.tif
5e7e0f09db3798a690eafdbb681b8c66
3a260b7bcce504b4a66596d580f0031df8dea780
11282127 F20101108_AAAJYU camaratta_m.pdf
3e5e6aed0b08467c312c238816e95b63
8d195e3e9325ddf341c056a7590fa43926a5791d
52524 F20101108_AAAKFD camaratta_m_Page_086.pro
ab2de2aa34b020dc730b81926bbfc32a
04e515c072c9c5d02477e75e09d376487495c9cb
F20101108_AAALIG camaratta_m_Page_097thm.jpg
0c54a855a3044db01815bd0cb6423d4d
044d7cc3d40dfec8af478ca66370caaba6f92dee
14550 F20101108_AAALHR camaratta_m_Page_072.QC.jpg
e8d902fcbc668e5aa5a8040c6a00ae6d
9e41350a147580a6d7a40fe9c32d0adc701205b8
68963 F20101108_AAAKEP camaratta_m_Page_159.jpg
5c479b66887e7a50b245322076a04988
9c12aff490fc189d99482517f3f9ab9b9f738891
6912 F20101108_AAAJZJ camaratta_m_Page_141thm.jpg
a4f3a90bf0adaab9e3035b849c6d751f
8becd62a00b2e88d4049badf1299c0a3d44b5fc2
1051853 F20101108_AAAJYV camaratta_m_Page_054.jp2
b8b907b8b04f7d6fc57a8e1eb569f38d
e7526501ef3236f650d232be2230294f9d6d5d2d
41986 F20101108_AAAKFE camaratta_m_Page_103.pro
97f48a0163e9dc5cdf9f6c76750f8960
3dfdcd731878f7496a0c16f07b29206b3c9f6132
5189 F20101108_AAALIH camaratta_m_Page_098thm.jpg
06ccd03ba2c49fab9362d06794384752
8f90c4460ae689a23ce61678f359db34e47eeae1
4330 F20101108_AAALHS camaratta_m_Page_072thm.jpg
f97b53a10eb0a0821a5db4415290d6f4
8c022a9eee499510710fc636b7743c866b1e80e2
55357 F20101108_AAAKEQ camaratta_m_Page_039.pro
c72ebda7fced275426de87e8b96e6cde
4c72d6e4bb06f9f6a6da53f4985e4230210a1c73
F20101108_AAAJZK camaratta_m_Page_053.tif
71813ca519958d4e7e8d39ce65e8739e
a5297a3b7970bc03235ac2b2cf30e8222a2338e3
F20101108_AAAJYW camaratta_m_Page_152.tif
de81bb5656a35a5cc17102f272329884
ed73de47e422697429fe31531af25495a82ade7f
57912 F20101108_AAAKFF camaratta_m_Page_029.pro
e4df7f6bdc51c06ef57ddf79f49d5c0d
33c91eef9de0516b2f47da9837334f41e178dd31
6564 F20101108_AAALII camaratta_m_Page_100thm.jpg
4b188be4ae2069131f6424e53f7ca38f
e095814261e9acdd403392b4871a5df609f00333
3888 F20101108_AAALHT camaratta_m_Page_074thm.jpg
4f4bbe3f731e4f0da85b3d10b8b9fa41
cb62206ff613b92b85466e567d0de52748c2d8d6
39400 F20101108_AAAKER camaratta_m_Page_128.jpg
6f7d26e2fa33e73f7b9cdbe20c973de5
c08908965f16539eddc67028b96cfc28c46b88c7
F20101108_AAAJZL camaratta_m_Page_022.txt
3cdb13a7ba4eaf7c64f73d96de36e7ee
a09933011201cb3eff2ac44402685923e90273a6
926278 F20101108_AAAJYX camaratta_m_Page_021.jp2
5850b2b3eb8177e4c4059d75d7fcab4f
13a8bba1369d8b1ac4bfc0a37457db8e54a65d12
1997 F20101108_AAAKFG camaratta_m_Page_014.txt
756960a987e2596b32780f4a0700d608
f6ff288d108330a05d079f03b5c928d4e2b7e9c6
9506 F20101108_AAALHU camaratta_m_Page_077.QC.jpg
e6a716721602331dbc28713d76eb75ab
8bce010d3fd056a0c55cdffc7cc1e7eb89b5465f
62153 F20101108_AAAKES camaratta_m_Page_155.jpg
ee1b8e7a40aca358167d2c12dc2e8259
94afbc835df250c7dfefa1b940e08ad2c6931673
1012718 F20101108_AAAJZM camaratta_m_Page_146.jp2
7cbdb947581b44787331d832bbbdf30c
363fe652bec9240cb3a0368b42b4f4a2b78463c2
F20101108_AAAJYY camaratta_m_Page_084.tif
9819632617720b3f374872cd89750d50
920d4f64921298299b517b607da3f5a242f1c909
29121 F20101108_AAALIJ camaratta_m_Page_109.QC.jpg
a779449dd9beb94d565e6fb9068540ca
ba5ec75107615dd4843ee259c7e05663caf28bca
17182 F20101108_AAALHV camaratta_m_Page_078.QC.jpg
1080f25f77dbd38bf4dc610fae06e3e2
8a05f6cd9ef611f6ae297bdb735ba36edd381789
11532 F20101108_AAAJZN camaratta_m_Page_096.QC.jpg
91b4f260dda75eb6daf855a437e4ab54
ecbdc9325821eccf157af665abb043166fa974f0
4809 F20101108_AAAJYZ camaratta_m_Page_054thm.jpg
67ce85e5f4a60c090196d909e0cecaa7
9db2b6aa5ebcb4d2dd731b3aa867ac8761b4fdd7
2203 F20101108_AAAKFH camaratta_m_Page_017.txt
81b738ca7f88148134d941b30892e6b0
9eb97bff425bb8c57627c7994b8906f976d0769b
53014 F20101108_AAAKET camaratta_m_Page_139.pro
a3a17d88bae089c3fdccc3562cb9caf9
d978fd55db8ec97112d7536f5c3ff08f966cdb25
26670 F20101108_AAALIK camaratta_m_Page_110.QC.jpg
16b2a4634ed7d1194b509e3f6be9aa0d
e8f506bfbd6d2cd1a3f0181145517e21d3743589
17121 F20101108_AAALHW camaratta_m_Page_079.QC.jpg
6cbf9bf1c3703366f539e6564f055c14
f814f1900ce4904f4707dfcf2d57f88c8d7adee4
73390 F20101108_AAAJZO camaratta_m_Page_150.jpg
d58ff250e9b80978800cc08779cf96a5
4b91926b7f113e85a838c662d62cbe50336f70f5
F20101108_AAAKFI camaratta_m_Page_031.tif
7a812c194a4e1019e8a2a6cd72a60d34
65ea1ce8d4458e28c168449d7f7030eb7450b904
6850 F20101108_AAAKEU camaratta_m_Page_040thm.jpg
62d196c2cd1f7352cdb769d22f77ff81
2fc07f1991be241c2b1b8c76b04801dea462bcc8
4642 F20101108_AAALJA camaratta_m_Page_135thm.jpg
3b29942971e48f3564bd71cc59e25e46
83e172292ba913878056deafc0ba6c8378105d7c
6088 F20101108_AAALIL camaratta_m_Page_114thm.jpg
939b2580d86c7ff1f8a3967ab665760c
2264490546138b002006825d3208afcfe80f090e
4424 F20101108_AAALHX camaratta_m_Page_080thm.jpg
e656eb2380ec8a653dcb439eea430a3e
c728810ba739e9d51e1d1dd85fc06572b153139c
11997 F20101108_AAAJZP camaratta_m_Page_045.QC.jpg
0ebe5255989d91a096f879f0cb242fb3
8ea3b94bd0017453026ad7002cd5dff1531869fd
924232 F20101108_AAAKFJ camaratta_m_Page_092.jp2
9bc268ad28f7f3d9f84349a03f11dfd7
5f4da8ed3932dd8ed809d44b279badb195efec2a
2244 F20101108_AAAKEV camaratta_m_Page_068.txt
8156e7c770b4a9a4ab9b468c349e3760
05b95861394ac57e1cb9018ae187bc21fff88dad
10739 F20101108_AAALJB camaratta_m_Page_136.QC.jpg
9b1f79532b920aa6854ce93ab1f4657c
8a3c0ca9f341cffd77ae80ca5b64997b202726a0
27087 F20101108_AAALIM camaratta_m_Page_115.QC.jpg
0a5c2d3e1a76cc18b5458e523e69ef22
c7aff79da471a25ea2eda7c3078bf3e00b40f6cf
23937 F20101108_AAALHY camaratta_m_Page_083.QC.jpg
b8ce60f7d7e643bfe721ca749bba2cd6
6c03a8bc86391fb4c72429d796eb311dfe1c62f4
F20101108_AAAJZQ camaratta_m_Page_032.tif
be8880fc4011da7ecfd3dc61e90612cf
e3f2d6f59b11babca9e42e2f3deb53da2b600dfa
7402 F20101108_AAAKFK camaratta_m_Page_127.pro
016508eab1846466d8cc0153b8a5fadf
89ade3642280a4efae12ec3e96f1183899a0ed44
86833 F20101108_AAAKEW camaratta_m_Page_116.jpg
81d9cbb69586d667d9cd605764853a64
77bfd493f368f5416739304b675b4ced06e556e1
3524 F20101108_AAALJC camaratta_m_Page_136thm.jpg
591761a16d98459d3984954b92060a99
892c7848722cfc353ae0e639dca9c1da54ace581
6514 F20101108_AAALIN camaratta_m_Page_115thm.jpg
d32d3f684c7a406067b96941c6d0c931
f60a5c723cf9a3e957127272b995a6cc87f62e01
5943 F20101108_AAALHZ camaratta_m_Page_083thm.jpg
412f88a83590fb64f985b6fde4cde8fa
7443c3bac8fc49bc3fc244c45eb88296862aa670
F20101108_AAAJZR camaratta_m_Page_036.jp2
d713125b883dfbad90bbba064078ac37
11022650168243fc7fed14e401c92dc7619637e4
5551 F20101108_AAAKGA camaratta_m_Page_103thm.jpg
69326fc2628af385dad7b52ef78ad69a
9cf8fdddf0edb6cabc148329747564e7613dcbab
5126 F20101108_AAAKFL camaratta_m_Page_151thm.jpg
49a07023b79510f95dc58e08cf3a2674
18bbc3943eec311e3b880452f3b3118d46ea9dcb
967373 F20101108_AAAKEX camaratta_m_Page_052.jp2
dfbf2f6d92a72054a9d9b8da84d07855
4feb83b1a481894ed1f3f5533ce7680b6cca5455
3415 F20101108_AAALJD camaratta_m_Page_137thm.jpg
c2bbc4d1d9307c7564ede6ad07254e6b
8ed26851a77d4aa1798ac81cfe32148d15ed28b3
6407 F20101108_AAALIO camaratta_m_Page_117thm.jpg
bfa14fe1514cba945c774f16709f1001
89997dfdb6e7c6655e5cd7b69dd26f7ce8ee1c5d
1051948 F20101108_AAAJZS camaratta_m_Page_039.jp2
2741486c8afe203f6e8f51cc6b4e424d
72c2d3734b772dd4e9e6b444ea667672dfa44550
F20101108_AAAKGB camaratta_m_Page_109.tif
83f06b6324b5d4ea4085ab980f33eb64
56b673362724bba9f19bdf608e9c03cab711f8a1
56906 F20101108_AAAKFM camaratta_m_Page_087.pro
6a5885d8f5f15e435ab5706e8b77c08f
8a61e38e98e4f71bdeadbd78542f42cf882b6e3f
15382 F20101108_AAAKEY camaratta_m_Page_019.pro
9e4d8201970753669359acee63111650
07e1a8cda8475bee6f7a9ef7be9be2a36f34926c
6802 F20101108_AAALJE camaratta_m_Page_140thm.jpg
056db4488cbb8bfdf0af42463a87a5dd
44a49cac1eeff73d82e4e7fb98f1805f0a0228d5
3439 F20101108_AAALIP camaratta_m_Page_119thm.jpg
3bb6fd17015c9b748e85f627b12d163c
e042bbc3834c82aecba377c4de0e94f4e6621159
1244 F20101108_AAAJZT camaratta_m_Page_012.txt
421a929623cfe1ac8bb8a67bc1a4c141
8d6058d5f9402259e1e03c0d9d6b9fa6707044b5
35981 F20101108_AAAKGC camaratta_m_Page_071.jpg
633edca2986dfa1a2f7718877726814b
64f52b47a7157f58b12d81277e3ad3bb2a084342
53993 F20101108_AAAKFN camaratta_m_Page_043.pro
edf66f1a942e9f075379ce4bf1e937fb
4b7396ca1eb82cfb2505426d0def56c96fed5704
13200 F20101108_AAAKEZ camaratta_m_Page_050.pro
b8cdb6da819f4193fcbbfc943e7832de
ed3145a92dd718067c00cac5a512ce73742f776d
26798 F20101108_AAALJF camaratta_m_Page_142.QC.jpg
80d6f609e20fd1014925009e1a84da5c
5ba1b2e209e9c68f523074fe2c16f091506390fc
2978 F20101108_AAALIQ camaratta_m_Page_120thm.jpg
6888c85c4ae432aee8c63a5e278267ef
5b916453c80800d827fbaa493cb7053237c3b184
6654 F20101108_AAAJZU camaratta_m_Page_042thm.jpg
a83da69a401d6354f762ef8e3792ada7
0fed6ec06e7749d4889e9a358e9780fc14d1e4e4
8821 F20101108_AAAKGD camaratta_m_Page_134.pro
62e62018455bc39c280bbe9a9db48308
cf01b21e77c79cf21e4d44ff3e9091c7569623bb
832 F20101108_AAAKFO camaratta_m_Page_045.txt
484d9cd2581b5511e1e54a1e303a326b
c51212d2ac3065f8d3a624cfcbfd749cfa00f315
26901 F20101108_AAALJG camaratta_m_Page_143.QC.jpg
9e12bd5ecb343d3fc104722ce0ab6970
44579da7649a5feffe6a8b16872f6c95f6fa1c84
12161 F20101108_AAALIR camaratta_m_Page_122.QC.jpg
e138a871f03fcd5874d575f2c34b1054
b752a2cf22459cd8fecc8920c6d49555d8db88e9
F20101108_AAAJZV camaratta_m_Page_003.tif
3981eb0fb39541e9eea37582ee3f42d2
f3634b03a1206e87e4df48db2c5a955a834b2355
124 F20101108_AAAKGE camaratta_m_Page_003.txt
de3646ed241c651ea7f4e4a1cc89d2f1
52c9945e4b2d4aff61044cfc64a6f397dcc6f709
F20101108_AAAKFP camaratta_m_Page_028.tif
4d789898c5cae4af1c2ddb2e048c105f
4ac6a2b873525a37f17d78089b34532be147162a
6212 F20101108_AAALJH camaratta_m_Page_145thm.jpg
78365941a70d715104470984694a46c2
8db226c06e2729f1e6195b142ed3c8fba1828859
5399 F20101108_AAALIS camaratta_m_Page_124thm.jpg
fb0d51bcac1adbdc071298171e478220
ea42cb65b0effabf0a1e9cb42f440a57b5317416
5042 F20101108_AAAJZW camaratta_m_Page_044thm.jpg
d8efed85d193185dbcd4e47e1bf1f4f7
46e80f6b396eee5fc98bfa94483e94bc94d7ad5a
57763 F20101108_AAAKGF camaratta_m_Page_030.pro
42bb6c3086eee490aaee17f46cfbe179
147039c3d082bf48cde181f2b5d981b380f02629
6684 F20101108_AAAKFQ camaratta_m_Page_144thm.jpg
30c3e8d36ce7a4405b084ded842c2e8e
0fcfa084fc5a3b0b77c9880879898c236c123898
23802 F20101108_AAALJI camaratta_m_Page_148.QC.jpg
6a6c341770d536cde7e116fdafb272cf
c9d5876b1678c2578a4793f657769b5dc02d16ad
8384 F20101108_AAALIT camaratta_m_Page_127.QC.jpg
811daa9c47dc240de1394926b2010f74
077443fe7bf7942eeacb5187c7b5ae2dbb5acc89
16826 F20101108_AAAJZX camaratta_m_Page_051.QC.jpg
2a6000cbfa479ec7a05fef341c48e535
bb480023a69977850f9a4b549f3f43fdd4f6bafe
126333 F20101108_AAAKGG camaratta_m_Page_011.jpg
effe705e2f02d7e39db1683fd90465fc
4185410b21edf38aa9ef782424d0e8404fd13bc8
52130 F20101108_AAAKFR camaratta_m_Page_035.pro
550ed4665ff6666e602d4b500c1ff3b1
ef0b8df9ee185bdc82b1394f7f98ade39fa82c74
6016 F20101108_AAALJJ camaratta_m_Page_149thm.jpg
c473f7ae31ec21822009edd311e9cb57
a30853c720ff3640a2808be02e17ed3350c526ba
12068 F20101108_AAALIU camaratta_m_Page_128.QC.jpg
b338a4c33ae25aca9b1d30b66bc5f9fe
69e345d9ef13f2fd108f075c2f9d1cda8f50d59f
68076 F20101108_AAAJZY camaratta_m_Page_151.jpg
5cb8e28f70d6d1ad15e67a687bf98e95
b96bd6f63cb35c93a00b295ff583fa5bb591cc21
81827 F20101108_AAAKGH camaratta_m_Page_147.jpg
9e42c75286aff23e6220a21a46bea8c5
737e2afec693af13bc9941eab41cd338ed80fe23
1017523 F20101108_AAAKFS camaratta_m_Page_059.jp2
705df659c9eff5a4a6c81dc20277e9aa
023945bf584cc13acbd73692186e960f91ce5a66
3577 F20101108_AAALIV camaratta_m_Page_128thm.jpg
6a8a991e98850b51757195615d0fba70
ba5eeeda1210f6f35febd4f98138c1c1b83de6be
51007 F20101108_AAAJZZ camaratta_m_Page_036.pro
985f822ff95f2e379e50244103800e83
18fc6230a43016d342a22ea256112ba0b4923e5e
2307 F20101108_AAAKFT camaratta_m_Page_031.txt
3da07a2d7895b2376693c85b570429c5
026ccff2ffb95e9e19be626a869a2ab8fbcfa77d
9998 F20101108_AAALJK camaratta_m_Page_154.QC.jpg
200efcea4b5942cc01e0983d8b085e93
48ab13fa740957a12e17e43c5856b058b2b2a8f8
3468 F20101108_AAALIW camaratta_m_Page_130thm.jpg
5d2f0b15703e66249f0e701fe01ed7a2
6c7142b154a667dd4be743ebca8fd27397d5b340
24881 F20101108_AAAKGI camaratta_m_Page_001.jpg
b6c8e6b4d865f46b9dcdf8700736986e
4014c367fb7c98ecd46af2bd64c6660d0ebc69c0
18698 F20101108_AAAKFU camaratta_m_Page_124.QC.jpg
c23c583f061040f88dda857430a8e245
9c1b1eb3705d9ae37255dfef0726dd6339d014d6
3167 F20101108_AAALJL camaratta_m_Page_154thm.jpg
d55b2ebd58c4083aa731660b16d6a297
80a2e8b1773d6f5393b26501464eaba36b8ad88e
4112 F20101108_AAALIX camaratta_m_Page_131thm.jpg
cd27d4814ea532afc963c827f78e253a
4ce18bfa56f0f2199c28738f39d4c953ef8e7ccd
49720 F20101108_AAAKGJ camaratta_m_Page_142.pro
c9d804a6a0b6b1d58133f54e741a8cef
1471d8f474f54ee696e4007ae0749a35e9614623
54531 F20101108_AAAKFV camaratta_m_Page_015.pro
0db95c5e2ce7697cd4268b2b90f0329e
fa2b71e954a03d7da96e96b8e7ce76933d961bdf
F20101108_AAALJM camaratta_m_Page_156thm.jpg
d70672be4e40d76d60578f9cd3e44c7e
3378c4c05c670a460d07df4f8b5171a630e43237
4511 F20101108_AAALIY camaratta_m_Page_133thm.jpg
71546f17fb1943c52c3ae85871b579e5
d8382367c7769def8cbcf7b7d85941f266223060
F20101108_AAAKGK camaratta_m_Page_119.tif
c17b695c8d019c2787335a21cf3aa086
e4959a53b26a927c8a4d4180e1000160e9382149
27810 F20101108_AAAKFW camaratta_m_Page_008.QC.jpg
bdc63263994f6577d10078a70cccfc5d
bf98b56ca61fe396e9420fdca87a9563fb497dae
5206 F20101108_AAALJN camaratta_m_Page_160thm.jpg
636d613f61dbc19c51d57404cf9b307b
7519e001ef55983a1e059d179a2b637ac35ba012
8423 F20101108_AAALIZ camaratta_m_Page_134.QC.jpg
f36d95eb0f6dc65367cdc2c5fdbb62d6
dd73504f6cc21b92e8742bae3a7dc6c8686fabb2
F20101108_AAAKGL camaratta_m_Page_031.jp2
b7d87f98427ce57e5550a88ba8abb2b5
d658502af6d3f9bccfd8bed31e6ccc27accf2a64
F20101108_AAAKFX camaratta_m_Page_105.jp2
fd8741ef60d6a40d15463f2de43a4c70
b5c814e253a20f0728ec20a9a8e744aef649a5de
18278 F20101108_AAAKHA camaratta_m_Page_155.QC.jpg
7eea405cf8065f0edbde7de3d6ddb5ea
506435a9f35758df50fa51af38fe308473c5319d
F20101108_AAAKGM camaratta_m_Page_049.tif
dfc1946fd4cc8ee47b2808f3f2120fc7
b80506965e7dd5987e1d942818f24671a39bbbcc
F20101108_AAAKFY camaratta_m_Page_038.jp2
57b21449ae5059d34e926a1947f05db8
901575935725626c42721b229a1b6b8c2da75088
520751 F20101108_AAAKHB camaratta_m_Page_072.jp2
dfb4a1b196c63fe88ab4075d8bf90f90
f966f1a6e3a9bdf478d8b46b9c4080b2bf97f282
7290 F20101108_AAAKGN camaratta_m_Page_075.QC.jpg
e0b8eb93b8e93e6ac42df51788cd3a40
c65e982d3b2a60a822bdc3760d231d21358b6ce5
27546 F20101108_AAAKFZ camaratta_m_Page_066.QC.jpg
1371ca97fd12414e675ecd6b3526b8b8
ee99b2eebcf76c3f7d71958141397be964c65dda
3967 F20101108_AAAKHC camaratta_m_Page_122thm.jpg
8881ed16e912f5f73787af4bd213179b
f078490223ad715dbb004bc312e365bcc090dcff
87173 F20101108_AAAKGO camaratta_m_Page_035.jpg
3384ff4172adcbddd983fa235165ad70
9d0bc34a74a156917b9757cab0c04b90261b406d
12609 F20101108_AAAKHD camaratta_m_Page_074.QC.jpg
33b3bad912ce37735b35354945ed8b2a
51a78f1faf1c0d5aa63cb4052d28dca733123e78
F20101108_AAAKGP camaratta_m_Page_054.tif
09648d03f40a205fa2f934513e650205
c1f2b59321b6be9182efff6abebc5d97b0121879
53334 F20101108_AAAKHE camaratta_m_Page_046.jpg
b4d6c74150669705abb13c489d88fce6
0bd8264bbbc455e45f4e95a5020775e89d6c5232
F20101108_AAAKGQ camaratta_m_Page_035.jp2
d29cc808086b15956f0a012f38b0e3d5
93ce596cbd4a9642a5dbdf9cf92afe3b74098dae
6839 F20101108_AAAKHF camaratta_m_Page_025thm.jpg
537a3c643c03241336c08174f4358328
b905a40e233e2a379ac6a917b15708474c157ade
6265 F20101108_AAAKGR camaratta_m_Page_007.QC.jpg
3b452d7362c5b4d1a0d84d69a01f5da7
cd85050c7c9df8fafb96faf6dbdc9a9e8f025391
1051943 F20101108_AAAKHG camaratta_m_Page_116.jp2
ccd22b0196a48fcc9519a880bc24c2e5
089a71be931fe6ac585c56d1f00d6594e0d7bd69
530 F20101108_AAAKGS camaratta_m_Page_073.txt
07a43372ab86662722d451febcd592ab
7c102d9601d06ca10345aaf8d398a5cc6390a869
F20101108_AAAKHH camaratta_m_Page_046.tif
75578ba2dd71a72c5a744b77a0baa6f2
7d99f8434ff44fee64844b68b3cc6d7c66d78e1d
6151 F20101108_AAAKGT camaratta_m_Page_038thm.jpg
f2cccb12d260620cbf4f5abc4fca15f4
c72155a19c4d608374883da378fb2b1f89008c1b
5156 F20101108_AAAKHI camaratta_m_Page_155thm.jpg
f09e8a68258088f6f42162b73c8e4129
b2878284985645eae7c3d60b5bd0984580c8e2be
2591 F20101108_AAAKGU camaratta_m_Page_008.txt
e08b596cb9de4ea5f5b0c657f6fea335
28380bfeacb752d32bed3c36b172817cc5c4223d
6360 F20101108_AAAKGV camaratta_m_Page_063thm.jpg
d14f97c4e916b80476a387797e926a2f
1e941d28ff1e43011a224be85e29621f62d9d41b
58857 F20101108_AAAKHJ camaratta_m_Page_078.jpg
1af2835e739c7910cd191fb347666882
6d40a50771edab30d63acb95317993fa3c69b1ce
43768 F20101108_AAAKGW camaratta_m_Page_149.pro
e9baa4596ba7a5154d79d8a0bd5b53e3
989b95fbeda776f2e4962efb2743c9835207210b
7355 F20101108_AAAKHK camaratta_m_Page_001.QC.jpg
8cea631b73dbf374e996a74ccac52fcc
002630438d978c2755f4f49ba03795142526d43a
1051895 F20101108_AAAKGX camaratta_m_Page_101.jp2
df3a920405b87f60da7fc9a5ae9d457e
7c39eebf279f1f76fdd325f402fb9290bc3e5348
74211 F20101108_AAAKIA camaratta_m_Page_005.pro
18d381c3b706e93c597a21db2c7c0688
c31b41b44d1e12c770397a7285cffd852ea22262
941353 F20101108_AAAKHL camaratta_m_Page_050.jp2
3a00268e3428b010b3f2a5b23eeeddf2
4107ebcfb2f3736b08695522f49b70522d92f9fb
21514 F20101108_AAAKGY camaratta_m_Page_104.QC.jpg
0448397a13fb8aa856e3358a5daf8546
e2778e89e0e3d8486bdda672ea720bebe89afb1c
13361 F20101108_AAAKIB camaratta_m_Page_076.QC.jpg
902bad5210054138d8831ec138abb114
d0ff3a94d43f07b0b9b5d2b7d7539630700c8b63
92035 F20101108_AAAKHM camaratta_m_Page_140.jpg
e6caf91fade5102ce4bae978fcdbb5d7
48b44f2327c23af7913644997b4a98612e6fe1cc
F20101108_AAAKGZ camaratta_m_Page_066.tif
fa2d4eb9efd3f390b7f1974d3e2cc6f0
3f4a298b9164cc8544da2ade7591f11c483c9e4e
F20101108_AAAKIC camaratta_m_Page_001.tif
7269401f64b2a9091573091cb75eb86b
6ed8d8fa445bdc6ac2964e444834197f938eeb69
6904 F20101108_AAAKHN camaratta_m_Page_017thm.jpg
1a173b0ba9ba8b4222eb3340539bffa3
83f7fed2f2d6bebb6f7575b7439ea51afe5493f1
28729 F20101108_AAAKID camaratta_m_Page_041.QC.jpg
c12c2ed9dfd695d39a6d977f9ce33996
e7579220c2384a489f6df1cbb051bc8466f74df2
6092 F20101108_AAAKHO camaratta_m_Page_064thm.jpg
5fb5589a5115a37cd2924dade652ac8f
692b18364bb6ad38fdf9d36d61c2ec4c37bda2f1
607 F20101108_AAAKIE camaratta_m_Page_050.txt
9f034d24350595328f236b3aa6d45574
d665cd7abeae197e982c9df2acbbc8a9d1dabebb
3001 F20101108_AAAKHP camaratta_m_Page_011.txt
1e56c461f5c4f6a1b0d1d0db0a1f52f0
b8b88a57fbf0948a02f241898fe54951b2ccfbd4
F20101108_AAAKIF camaratta_m_Page_016.tif
e4b908232605a12bdbc2ef638c37f6b8
f2a7eb895a5686165db1adc1fdb35beeb0e02ebe
F20101108_AAAKHQ camaratta_m_Page_148.tif
c042da435be39d38f172eb80042954a5
8d728a2b8b99f883871c4786de4201d8f15a06c8
6145 F20101108_AAAKIG camaratta_m_Page_081thm.jpg
4e72f025239c5c5d1e3ddfbeaf02ea56
46f9262c81cbe2efc01d5ef4e0a352d471ebccab
57274 F20101108_AAAKHR camaratta_m_Page_028.pro
d51fa80226e2a5ab34cbf27e507f8bed
ccf036dd2a7b33aab0bc06fe1e79891f6d5ebf41
2464 F20101108_AAAKIH camaratta_m_Page_018thm.jpg
0b35fde329aadd4c9d8cf9847b8703ab
89660b45a6291509b16c929e98d204307584a528
F20101108_AAAKHS camaratta_m_Page_092.tif
ba54e90072c601b57d9595ea48e154be
fc9881281d1440b77a4603ded1fa780b53784037
95413 F20101108_AAAKII camaratta_m_Page_029.jpg
42522b392613535dc8cae6830da17289
54e3671ef85b539e7b0f02a889699ad946b8b29b
F20101108_AAAKHT camaratta_m_Page_115.jp2
1de97634b7ecf20a568b4f33eae9261e
a7afe0b6dd18e86961dc7e290775737eaba431cf
3851 F20101108_AAAKIJ camaratta_m_Page_006thm.jpg
d35c90fa4979f8a285aaddb2df2723a8
9330fd4f387fc0470a341c55ea299fead7250208
3226 F20101108_AAAKHU camaratta_m_Page_005.txt
523c6ab0e34dd19860aac52f605ed8d0
1357856f1359352179ddeb96d1b48d1a0d72d77e
F20101108_AAAKHV camaratta_m_Page_147thm.jpg
71adeaf0f3c3074a30f607ee6a97fbd5
e2cf03e04cbf43fea202622c62793e1fd517052d
21325 F20101108_AAAKIK camaratta_m_Page_022.QC.jpg
25e782891ad1e4a5f00adfe8b6773c71
558ccb95045c0fc6e4495b694849fd06f5e57605
F20101108_AAAKHW camaratta_m_Page_003.QC.jpg
b8445d504e83ebb869ad1bdd5e8a4517
b54b2a37a373c2654b388787db6bd493e120a732
2044 F20101108_AAAKJA camaratta_m_Page_023.txt
6bf118093ca7ddefde6096b93fb52172
99fd29b2080dac32433ba24dd157181ccb9fbbd1
28273 F20101108_AAAKIL camaratta_m_Page_156.QC.jpg
c0dfca6c59437b1d10364ee77fe8868c
90e38eb74d7a86545f0512857cb1f9a94c276532
59236 F20101108_AAAKHX camaratta_m_Page_124.jpg
58d6ad9914039a890cdb3d1c3b6405f2
63c6f428bdae4d60af06b2d271d4570e1b4d6784
27466 F20101108_AAAKJB camaratta_m_Page_067.QC.jpg
83fb2eaf5282793f41e5c200409da434
ce0c9f425f03e878c64aefe889562a3ca8446aeb
42074 F20101108_AAAKIM camaratta_m_Page_160.pro
43dc48ca1be89fcdad268aec30a18977
14f1faa81b1163f20684c7bef80959ea5bb71d02
23115 F20101108_AAAKHY camaratta_m_Page_048.pro
d352112074f8dcb79827120bca350342
7b794f11b747b842480473181d7a9b40f0697dba
400 F20101108_AAAKJC camaratta_m_Page_071.txt
2b3f9a8514abab3b411f038b41f37065
cc13ae161bbad307a2894f01c8187793a78d2256
264 F20101108_AAAKIN camaratta_m_Page_078.txt
4953c66dbbe90682b6303aea88b6ef04
23e34de8ba21c8c94fe6c1ef3c7deca24c6daaad
28717 F20101108_AAAKHZ camaratta_m_Page_140.QC.jpg
011d5b73d2acb2fb261eb093852d58c2
ae0b98d3f798d36d3cb60a3c7a02700b740b2f58
52520 F20101108_AAAKJD camaratta_m_Page_145.pro
09d9b75ee52e91666e90d8923330020c
fbb184ac08bc7f36f93690db2588d6a085fe5ca6
22869 F20101108_AAAKIO camaratta_m_Page_059.pro
d37b8d97485a26245671dd85788fb3f1
7b9dd512dc14c8d60d9a5ebdab73188b43b72de4
31072 F20101108_AAAKJE camaratta_m_Page_019.jpg
3c757b80ead6d84e0a68e5bea8067d82
d62f3195ad2321e9ef5305d3e76cc847265642af
F20101108_AAAKIP camaratta_m_Page_110.tif
bb59b6211a63554ed0bb3a34cf28c06b
62a5f3cbe23e9b4a97361490c478a291fbd37c26
53363 F20101108_AAAKJF camaratta_m_Page_062.pro
7f528a64c1f2ae655e0daacdb8d2ab6f
e44e28fa87eb43d67d875eaff721e3aa42fdbfd9
6377 F20101108_AAAKIQ camaratta_m_Page_158thm.jpg
e2826712252c3c5f77f79cac6e94d785
9faa56fca9f1061fe0443877e606f270d5d89fee
640 F20101108_AAAKJG camaratta_m_Page_092.txt
e6c36caaf4d668e13b5041974a1ed2f0
e4a89874cdea0625f423711dcbdeba5f5ae6909a
91623 F20101108_AAAKIR camaratta_m_Page_084.jpg
84b10b733308a97d52d5f9626f9c2719
a934ec6f5992929d2587aed9ce36e85a6bad7e51
13362 F20101108_AAAKJH camaratta_m_Page_119.QC.jpg
4a2314e8917280402354e2f18e6fb28a
ea64749d5cf9b796219af863939790bbc30ef9f9
899674 F20101108_AAAKIS camaratta_m_Page_049.jp2
cba8589706ff3ca4a31a2a530394ab95
9504f430a39bcc447c92c9634417004d5d88621c
984 F20101108_AAAKJI camaratta_m_Page_119.txt
597bfa87ade44214e0f5f991e1ddc442
43a5872cc6c4258bf767fd107a1f134ae972687f
88064 F20101108_AAAKIT camaratta_m_Page_082.jpg
0f6d0cd4c8e8b5c009fe6cd8585dff5a
ee8ee97531130e4493977f74ef8e598fbb4491d5
F20101108_AAAKJJ camaratta_m_Page_015.tif
68cac4ab8d050b305d32c80e1ba4a142
59bece992f4a81d03623877aa0b95d03df17ae03
24599 F20101108_AAAKIU camaratta_m_Page_034.QC.jpg
ceaa7d29a18ad027748d57c10caa6116
4fa9e908ef46ae21e3e9cdcc119ffc024a0a9578
F20101108_AAAKJK camaratta_m_Page_086.txt
d39b9c36232278645074266f38fce7c1
99e3d1e09ceb321f129bc6f93f9ba61915180d05
622 F20101108_AAAKIV camaratta_m_Page_074.txt
277eb24c6f2913df25d3e08418162efe
43f1d5fe3033dd1ef3293b652a10eb920d6844a3
14502 F20101108_AAAKIW camaratta_m_Page_131.QC.jpg
5541eec1398a97ed4782eade8bfdb2a2
d6cb1386dc4b5bcb616739c4f30be46b5d2cf148
28463 F20101108_AAAKJL camaratta_m_Page_116.QC.jpg
d3d15e64ba358626da8039a75c02b2d1
09e5a03acd6dd9fcdbba2cbdaf6595cecf73e7fb
F20101108_AAAKIX camaratta_m_Page_083.tif
3e103d79b5cebadb75cda4cc33d2ad88
24e5caf85b78f5f9bd94d44861f9806a467e2714
347346 F20101108_AAAKKA camaratta_m_Page_154.jp2
45bd7d931a58b95f8665cefbe4d48ad6
59bb2f35bb8a96f30f57dc713c781a35a28257c2
28705 F20101108_AAAKJM camaratta_m_Page_084.QC.jpg
ac38900dba21506e6d40c08983393805
61d8f4fc1f8a333a25cab8a76307bcbc1075a57d
2080 F20101108_AAAKIY camaratta_m_Page_112.txt
047fdca539d529a861a25624a73eb292
15a18cbaac58766b77aa5f84823106d55d42c2ad
2036 F20101108_AAAKKB camaratta_m_Page_117.txt
f3b13680bda211e32d01b060a83f800b
a492e57e80233b3538273b77ceec1df74250b291
10290 F20101108_AAAKJN camaratta_m_Page_135.pro
4ecb00c245cc823361635fc9d2291021
75b6371aece8a7ba73b80a0160016f1fc306ade4
F20101108_AAAKIZ camaratta_m_Page_138.tif
cbded092fba77d45ba7674bdcbb65f37
0a2ba314d0191aaef30c9cad3ac535b09e70d045
F20101108_AAAKKC camaratta_m_Page_127.tif
19eb261cd07c9096ab31a70b0e47aaa6
bc21e42679542e3e5d3497d3ee6f855638244427
4600 F20101108_AAAKJO camaratta_m_Page_051thm.jpg
68100db766ee32fe561d2ef9d3e779bf
7fcb157f27cd10b6047a0b3cb92845714962045e
3640 F20101108_AAAKKD camaratta_m_Page_129thm.jpg
c252707604239c57ec711d049cb55a0e
1b9c07f53d355864a7007a68c5de816cf6b63e46
79722 F20101108_AAAKJP camaratta_m_Page_064.jpg
555025e46ca67833783f941cdb343744
ba6012e3ed39cd355a963f46881e42a6a21f1197
7008 F20101108_AAAKKE camaratta_m_Page_154.pro
f1d9b7db3ca2d055c7f18ef43d4800ef
43e47438b70c660d38c4665cfa5951b4ce33ea4d