Interactions between SOFC Interconnect Spinel Coating Materials and Chromia 
 
by 
 
Yingjia Liu 
 
 
 
 
A dissertation submitted to the Graduate Faculty of 
Auburn University 
in partial fulfillment of the 
requirements for the Degree of 
Doctor of Philosophy 
 
Auburn, Alabama 
 December 8, 2012 
 
 
 
 
Keywords: SOFC interconnect, interaction, (Mn,Co)
3
O
4
 coating, Cr
2
O
3
 oxidation layer, doping 
 
 
Copyright 2012 by Yingjia Liu 
 
 
Approved by 
 
Jeffrey W. Fergus, Chair, Professor of Materials engineering  
Bart Prorok, Associate Professor of Materials engineering  
Dong-Joo Kim, Associate Professor of Materials engineering  
Curtis Shannon, Professor of Chemistry and biochemistry 
 
 
 
 
 
 
ii 
 
 
 
 
 
 
Abstract 
 
 
 The SOFC technology has been developed toward lower operation temperature in the 
range of 500
o
C - 800
o
C, which brings wider materials choice for the interconnect. Ferritic 
stainless steels have been widely used as SOFC interconnect candidates, but their high 
temperature oxidation and chromium volatilization can lead to cathode poisoning and cell 
degradation. Applying ceramic conductive coating is an effective technique to protect the 
metallic interconnects, and manganese cobalt spinel oxides Mn
1.5
Co
1.5
O
4 
have demonstrated 
promising performances as coating material. However, the reasons for their excellent properties 
have not been completely understood. 
This work aims to provide a fundamental understanding of (Mn,Co)
3
O
4
 coating in high-
temperature operation, and to improve the performance of protective coatings. The interactions 
between (Mn,Co)
3
O
4 
spinel coatings and chromia at high temperature were investigated, and 
Chapter 2 discusses the mass transport behavior. The interactions lead to the change in chemical 
composition and microstructure of the original coating. To evaluate the long-term stability, the 
properties of the reaction layer need further characterization. Chapter 3 focuses on the effect of
 
chromium doping on the electrical conductivity, cation distributions and thermal expansion of 
Mn
1.5
Co
1.5
O
4 
at SOFC operation condition. The relationship of cation distributions with transport 
properties in the reaction layer was also discussed. In addition, the effects of different transition 
metal dopants in the spinel coating material on the mass transport behavior were studied to 
identify the improved coatings. The effects of iron doping in spinel oxides are investigated on 
 iii 
transport behavior in the interaction, and compared with that of titanium doping in Chapter 4. 
The interaction of nickel or copper substituted (Mn,Co)
3
O
4 
with chromia at high temperature are 
evaluated in Chapter 5. The differences in the mass transport behavior of transition metal 
dopants studied in this work provide helpful information on selection in novel spinel coatings.  
 
 
 iv 
 
 
 
 
 
Acknowledgments 
 
 
 I would like to express my genuine gratitude to my advisor, Dr. Jeffrey Fergus, for his 
academic guidance and support. His constructive suggestions helped me a lot in my research 
work and dissertation writing. Besides the knowledge of materials science, the research attitude 
he conveyed is the most important spirit I learn from him during these years.  
          I also want to thank my committee member, Dr. Dong-Joo Kim, Dr. Bart Prorok and Dr. 
Curtis Shannon for their valuable advices and insight, which is beneficial to finish this 
dissertation. The appreciation will also be given to the outside reader Dr. Wei Zhan for his kind 
help in this work.  
 I would like to further thank Dr. Clarina Dela Cruz at Oak Ridge National Laboratory for 
providing me great help in analyzing the samples with neutron diffraction.  
           The special thanks are given to Dr. Kangli Wang, who helped me a lot on the chromium 
reaction experiment at the early stage of this work, and gave me useful discussions. Besides, I 
also thank Dr. Yu Zhao, Henglong Wang, Alex, et al., for helping me solve problems during my 
research. The friendship from my friends in Materials Engineering is highly appreciated. I enjoy 
the happy time in Auburn during my PhD study. 
           I also acknowledge the financial support from the Department of Energy through the 
Building EPSCoR-State/Nation Laboratory Partnerships Program.  
 v 
          Last but not least, I would like to express my great gratitude to my family for their 
consistent company, love and encouragement which enable me to discover life is always 
beautiful. 
 vi 
 
 
 
 
 
Table of Contents 
 
 
Abstract ......................................................................................................................................... ii 
Acknowledgments........................................................................................................................ iv  
List of Tables ............................................................................................................................... ix 
List of Figures ............................................................................................................................... x  
List of Abbreviations ................................................................................................................. xiv 
Chapter 1 Introduction .................................................................................................................. 1 
 1.1 Background of SOFC  ................................................................................................. 1 
 1.2 Interconnect of SOFC   ............................................................................................... 4 
                    1.2.1 Materials requirements for SOFC interconnect   ............................................. 4 
                    1.2.2 Lanthanum chromites interconnect   ................................................................ 5 
                    1.2.3 Metallic alloy interconnect   .......................................................................... 10 
1.2.3.1 Chromium-based alloys   ..................................................................... 12 
1.2.3.2 Ni-Cr-based alloys   ............................................................................. 13 
1.2.3.3 Fe-Cr-based alloys   ............................................................................. 15 
                    1.2.4 Problems with metallic interconnect materials   ............................................ 21 
1.2.4.1 Oxidation problem   ............................................................................. 21 
1.2.4.2 Chromium poisoning   ......................................................................... 22 
1.3 Interconnect coating materials for SOFC  ................................................................ 24 
1.3.1 Reactive elements coating   ............................................................................ 25 
 vii 
1.3.2 Perovskite coating   ........................................................................................ 26 
1.3.3 Spinel coating  ............................................................................................... 28 
1.4 Objectives of the dissertation  ................................................................................... 33 
Chapter 2 Interactions between (Mn,Co)
3
O
4
 coating materials and Cr
2
O
3
 scale ....................... 35 
2.1 Introduction  .............................................................................................................. 35 
2.2 Experimental procedure  ........................................................................................... 36 
2.3 Results and discussions  ............................................................................................ 37 
2.3.1 Surface morphology and structure ................................................................. 37 
2.3.2 Mass transport properties in Mn
1.5
Co
1.5
O
4
 ..................................................... 45 
2.4 Conclusions  .............................................................................................................. 47 
Chapter 3 Electrical properties, cation distributions and thermal expansion of manganese cobalt 
chromite Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x=0-2)
 
spinels ....................................................... 49 
3.1 Introduction  .............................................................................................................. 49 
3.2 Experimental ............................................................................................................. 51 
3.3 Results and discussions  ............................................................................................ 52 
3.3.1 X-ray diffraction ............................................................................................ 52 
3.3.2 Electrical conductivity ................................................................................... 54 
3.3.3 Cation distribution in Mn
1.5
Co
1.5
O
4 
and MnCoCrO
4 
at high temperature ...... 58 
3.3.4 Thermal expansion test .................................................................................. 64 
3.4 Conclusions  .............................................................................................................. 65 
Chapter 4 Electrical properties of transition metal-doped (Mn,Co)3O4 (TM = Fe, Ti) spinels and 
their interaction with chromia for SOFC interconnect coatings ................................. 66 
4.1 Introduction  .............................................................................................................. 66 
4.2 Experimental  ............................................................................................................ 67 
 viii 
4.3 Results and discussions  ............................................................................................ 68 
4.3.1 X-ray Diffraction  .......................................................................................... 68 
4.3.2 Neutron Diffraction  ....................................................................................... 69 
4.3.3 Interaction with Chromia ............................................................................... 75 
4.3.3.1 Surface Morphology ............................................................................ 75 
4.3.3.2 Concentration Gradients ...................................................................... 77 
4.3.4 Electrical conductivity ................................................................................... 81 
4.3.5 Comparison in the effect of Fe and Ti doping ............................................... 84 
4.4 Conclusions  .............................................................................................................. 86 
Chapter 5 Interaction of transition metal-doped (Mn,Co)
3
O
4
 (TM = Ni, Cu) spinels with chromia 
for SOFC interconnect coatings .................................................................................. 88 
5.1 Introduction  .............................................................................................................. 88 
5.2 Experimental  ............................................................................................................ 89 
5.3 Results and discussions  ............................................................................................ 90 
5.3.1 Mass transport in Mn
1.5-0.5x
Co
1.5-0.5x
TM
x
O
4
 (TM=Ni,Cu)  ............................. 90 
5.3.2 Effect of Co/Ni ratio .................................................................................... 100 
5.4 Conclusions  ............................................................................................................ 104 
Chapter 6 Conclusions and Perspectives .................................................................................. 106 
References   ............................................................................................................................... 111 
 
 ix 
 
 
 
List of Tables 
 
 
Table 1.1 CTEs of La
0.9
Sr
0.1
Cr
1-x
M
x
O
3
 (M= Mg, Al, Sc, Ti, Mn, Fe, Co, Ni, 0<x<0.10) 
perovskites at 50-1000 
o
C in air or H
2
 atmosphere. ...................................................... 9 
Table 1.2 Properties of different chromia forming alloys for SOFC interconnect. .................... 20 
Table 2.1 The chemical compositions and Co/Mn ratio in surfaces of Mn
1.5
Co
1.5
O
4
 in contact 
with or near Cr
2
O
3
 after reaction at 800
 o
C, 1200
 o
C for 144 h................................... 44 
Table 3.1 Lattice parameter and density of cubic Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x= 1, 1.5, and 2) .... 54 
Table 3.2 Crystal parameters of (a) Mn
1.5
Co
1.5
O
4 
and (b) MnCoCrO
4 
at high temperature ....... 60 
Table 4.1 Crystal parameters of MnCo
1.66
Fe
0.34
O
4 
and Mn
0.67
Co
1.22
Fe
0.22
CrO
4 
at 600
 o
C .......... 71 
Table 4.2 Chemical composition and Co/Mn ratio of MnCo
2-x
Fe
x
O
4
 (x= 0.15 - 0.7) after reaction 
at 800 
o
C in contact with Cr
2
O
3
 for more than two weeks ......................................... 77 
Table 4.3 CTEs of MnCo
1.66
TM
0.34
O
4
 (TM=Fe and Ti) and (Mn,Co)
3
O
4
 spinel oxides ............ 86 
Table 5.1 Chemical compositions of Mn
1.5-0.5x
Co
1.5-0.5x
TM
x
O
4 
(TM = Ni, x = 0.2-0.6, TM = Cu, x 
= 0.2, 0.4) before and after reaction at 800 
o
C in contact with Cr
2
O
3
 for 96 h ........... 93 
 
 x 
 
 
 
List of Figures 
 
 
Fig. 1.1 Schematic diagram of a planar SOFC structure. ............................................................. 2 
Fig. 1.2 Schematic diagram of SOFC principles. ......................................................................... 3 
Fig. 1.3 Electrical conductivity for undoped LaCrO
3
, doped LaCrO
3
, and 
(La
0.73
Sr
0.25
)(Cr
0.5
Mn
0.5
)O
3
 perovskite in (a) air and (b) 10% H
2
/N
2
. ............................ 8 
Fig. 1.4 Schematic of alloy design for SOFC applications.  ....................................................... 12 
Fig. 1.5 A dynamic-segregation model of the reactive element function during high temperature 
oxidation.  .................................................................................................................... 19 
Fig. 1.6 Vapor pressures of different volatile chromium species as a function  
of temperature ............................................................................................................... 23 
Fig. 1.7 Equilibrium vapor pressures of chromium?oxygen?hydrogen gas species at 800 
o
C with 
a water vapor pressure of 3 kPa as a function of oxygen partial pressure. ................... 23 
Fig. 1.8 Schematic diagram of crystal structure of AB
2
O
4
 spinel. ............................................. 30 
Fig. 1.9 SEM images and EDS analysis on Crofer 22 APU with Co
1.5
Mn
1.5
O
4
 coating after a 
period of 6 months under thermal cycling.  .................................................................. 33 
Fig. 2.1 Schematic diagrams of sample configurations  ............................................................. 36 
Fig. 2.2 SEM images of as-prepared three (Mn,Co)
3
O
4
 compositions surface, (a) Mn
1.5
Co
1.5
O
4
, 
(b) Mn
2
CoO
4
, and (c) MnCo
2
O
4
. .................................................................................. 38 
Fig. 2.3 SEM images of surfaces of (a-b) Mn
1.5
Co
1.5
O
4
, (c-d) Mn
2
CoO
4
, and (e-f) MnCo
2
O
4
 after 
reaction at 1200 ?C for 72 h in contact with (left) or near chromia (right). .................. 39 
Fig. 2.4 XRD patterns of surfaces of Mn
1.5
Co
1.5
O
4
 after reaction at 1200 ?C (a) in contact with or 
(b) near chromia for different hours................................................................................ 41 
Fig. 2.5 Surface morphologies (a - d) and cation concentration profiles (e) of Mn
1.5
Co
1.5
O
4
 at 
different distances from Cr
2
O
3
 (light regions on the right) after reaction at 1200 ?C for 
144 h.............................................................................................................................. 42 
Fig. 2.6 SEM images of surfaces of Mn
1.5
Co
1.5
O
4
 in contact with or near Cr
2
O
3
 after reaction at 
800
 o
C and 1200 
o
C for 144 h ....................................................................................... 43 
 xi 
Fig. 2.7 (a) SEM images and (b) cations profile of cross section of Mn
1.5
Co
1.5
O
4
 after reaction at 
1000 ?C for 36 h in contact with Cr
2
O
3
. ......................................................................... 46 
Fig. 2.8 Reaction layers formed during the reaction between Mn
1.5
Co
1.5
O
4
 and Cr
2
O
3
 at high 
temperature. .................................................................................................................... 47 
Fig. 3.1 XRD patterns of Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x= 0 - 2) spinel oxides at room  
temperature... ................................................................................................................ 53 
Fig. 3.2 Electrical Conductivity of Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x=0 - 2) spinel oxides at 800
o
C 
 in air ............................................................................................................................. 55 
Fig. 3.3 Electrical Conductivity of Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x=0 - 2) spinel oxides as a function 
of 1000/T in air ............................................................................................................... 56 
Fig. 3.4 Activation Energy as a function of Cr content in Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x=0-2) spinel 
oxides .............................................................................................................................. 56 
Fig. 3.5 Comparison between Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x=1 and 2) and Cr
2
O
3
 in electrical 
conductivity in air. .......................................................................................................... 57 
Fig. 3.6 Neutron diffraction patterns of Mn
1.5
Co
1.5
O
4
 in air at (a) 600
 o
C and (b) 800 
o
C, 
respectively. .................................................................................................................... 59 
Fig. 3.7 Neutron diffraction patterns of MnCoCrO
4 
at (a) 600
 o
C and (b) 800 
o
C,  
respectively. .................................................................................................................. 60 
Fig. 3.8 CTEs of Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x = 0 - 2) spinel oxides and Cr
2
O
3
 from room 
temperature up to 1000
o
C in air. ..................................................................................... 64 
Fig. 4.1 XRD patterns of MnCo
2-x
Fe
x
O
4
 (x= 0 - 0.7) spinel oxides at room temperature. ......... 69 
Fig. 4.2 Neutron diffraction patterns of MnCo
1.66
Fe
0.34
O
4
 in air at 600
 o
C................................. 70 
Fig. 4.3 Neutron diffraction patterns of Mn
0.67
Co
1.22
Fe
0.22
CrO
4 
at 600
 o
C in air. ....................... 70 
Fig. 4.4 SEM images of surface of MnCo
1.5
Fe
0.5
O
4
 before (a) and after (b) reaction at 800 
o
C in 
contact with Cr
2
O
3
 for more than two weeks.................................................................. 76 
Fig. 4.5 SEM images of surface morphologies of MnCo
2-x
Fe
x
O
4
 (x= 0.15, 0.34 and 0.7) after 
reaction at 800 
o
C in contact with Cr
2
O
3
 for more than two weeks. ............................... 77 
Fig. 4.6 Cr concentration profile in MnCo
2-x
Fe
x
O
4
 after reaction for 420 h in contact with Cr
2
O
3
 
at 800 
o
C. ......................................................................................................................... 79 
 xii 
Fig. 4.7 Fe concentration profile in MnCo
2-x
Fe
x
O
4
 after reaction for 420 h in contact with Cr
2
O
3
 
at 800 
o
C. ......................................................................................................................... 79 
Fig. 4.8 Co concentration profile in MnCo
2-x
Fe
x
O
4
 after reaction for 420 h in contact with Cr
2
O
3
 
at 800 
o
C. ......................................................................................................................... 80 
Fig. 4.9 Mn concentration profile in MnCo
2-x
Fe
x
O
4
 after reaction for 420 h in contact with Cr
2
O
3
 
at 800 
o
C.  ........................................................................................................................ 80 
Fig. 4.10 Electrical conductivities of MnCo
2-x
Fe
x
O
4 
at 800 
o
C in air.  ....................................... 82 
Fig. 4.11 Electrical conductivities of MnCo
2-x
Fe
x
O
4 
and
 
MnCoCrO
4 
spinels as a function of 
temperature. .................................................................................................................. 82 
Fig.4.12 Chromium concentration profile in MnCo
1.66
TM
0.34
O
4
 (TM=Fe and Ti) and 
Mn
1.5
Co
1.5
O
4
 after reaction in contact with Cr
2
O
3
 for 72 h at 1000 
o
C. ....................... 84 
Fig.4.13 Dopent (Fe and Ti) concentration profile in MnCo
1.66
TM
0.34
O
4
 (TM=Fe and Ti) after 
reaction in contact with Cr
2
O
3
 for 144 h at 900 
o
C.. ....................................................... 84 
Fig. 4.14 Comparison in electrical conductivities of MnCo
1.66
TM
0.34
O
4
 (TM=Fe and Ti) spinel 
oxides.  .......................................................................................................................... 86 
Fig. 5.1 SEM images of surface of Mn
1.3
Co
1.3
Ni
0.4
O
4
 before (a) and after (b) reaction at 800 
o
C in 
contact with Cr
2
O
3
 for 96 h. ............................................................................................ 91 
Fig. 5.2 SEM images of surface of Mn
1.5-0.5x
Co
1.5-0.5x
TM
x
O
4 
(TM=Ni, x=0.2, 0.6, TM=Cu, x=0.2, 
0.4) after reaction at 800 
o
C in contact with Cr
2
O
3
 for 96 h. .......................................... 92 
Fig. 5.3 (a) SEM image of cross section of Mn
1.4
Co
1.4
Ni
0.2
O
4
 in contact with Cr
2
O
3 
after reaction 
at 800 
o
C for 96 h and (b) its corresponding cations concentration profile. ................... 94 
Fig. 5.4 (a) SEM image of cross section of Mn
1.4
Co
1.4
Cu
0.2
O
4
 in contact with Cr
2
O
3 
after reaction 
at 800 
o
C for 96 h and (b) its corresponding cations concentration profile. ................... 95 
Fig. 5.5 (a) chromium and (b) nickel concentration profiles of Mn
1.2
Co
1.2
Ni
0.6
O
4
 in contact with 
Cr
2
O
3 
after reaction at 900 
o
C and 1000
 o
C for 72 h ....................................................... 97 
Fig. 5.6 Nickel concentration profiles of Mn
1.5-0.5x
Co
1.5-0.5x
Ni
x
O
4 
(x = 0.2 -0.6) in contact with 
Cr
2
O
3 
after reaction at 1000
 o
C for 72 h. ......................................................................... 98 
Fig. 5.7 Nickel concentration profiles of Mn
1.5-0.5x
Co
1.5-0.5x
Ni
x
O
4 
(x = 0.2 - 0.6) in contact with 
Cr
2
O
3 
after reaction at 1000
 o
C for 72 h. ......................................................................... 99 
 xiii 
Fig. 5.8 Nickel concentration profiles of Mn
1.5-0.5x
Co
1.5-0.5x
Ni
x
O
4 
(x = 0.2 - 0.6) in contact with 
Cr
2
O
3 
after reaction at 900
 o
C for 72 h.  .......................................................................... 99 
Fig. 5.9 Surface morphologies of (a) Mn
1.5
Co
0.9
Ni
0.6
O
4 
and (b) Mn
1.5
Co
0.6
Ni
0.9
O
4 
after reaction at 
900
o
C in contact with Cr
2
O
3
 for 144 h.......................................................................... 100 
Fig. 5.10 Ni concentration profiles of Mn
1.5
Co
0.9
Ni
0.6
O
4
 and Mn
1.5
Co
0.6
Ni
0.9
O
4 
after reaction at 
900
o
C in contact with Cr
2
O
3
 for 144 h.......................................................................... 101 
Fig. 5.11 Cr concentration profiles of Mn
1.5
Co
0.9
Ni
0.6
O
4,
 Mn
1.5
Co
0.6
Ni
0.9
O
4 
and Mn
1.5
Co
1.5
O
4
 
after 
reaction at 900
o
C in contact with Cr
2
O
3
 for 144 h. ..................................................... 102 
Fig. 5.12 Cr concentration profiles of Mn
1.5
CoNi
0.5
O
4
 and Mn
1.5
Co
1.5
O
4 
after reaction at 700
o
C in 
contact with Cr
2
O
3
 for 30 days. .................................................................................. 102 
Fig. 5.13 Electrical conductivities of (a) Mn
1.5
CoNi
0.5
O
4,
 and (b) Mn
1.5
Co
0.6
Ni
0.9
O
4 
spinels as a 
function of temperature. .............................................................................................. 103 
 
 xiv 
 
 
 
List of Abbreviations 
 
 
SOFC Solid Oxide Fuel Cell  
PEM Proton Exchange Membrane  
8-YSZ 8 mol% Yttria-stabilized Zirconia 
VOC          Volatile Organic Carbon  
TPB Three Phase Boundary 
APU Auxiliary Power Unit  
CHP Combined Heat and Power Generation 
OCV Open Circuit Potential 
CTE Coefficient of Thermal Expansion 
AE Alkaline Earth 
TM Transition Metal 
ODS Oxide Dispersion Strengthened 
DBTT Ductile-to-Brittle Transition Temperature 
BCC Body Centered Cubic  
FCC Face Centered Cubic  
ASR Area Specific Resistance 
FSS Ferritic Stainless Steel 
ASS Austenitic Stainless Steel 
LSM Lanthanum Strontium Manganite 
MOCVD Metal Organic Chemical Vapour Deposition 
 xv 
RF Radio-Frequency 
APS Atmospheric Plasma Spraying  
SEM Scanning Electron Microscope 
EDS Energy-Dispersive X-ray Spectroscopy 
XRD X-Ray Diffraction 
 
 
 
 1 
 
 
 
CHAPTER 1 
Introduction 
 
It is a severe challenge for people to achieve sustainable development, and energy is one 
of these key issues that have to be considered due to resource shortage.
[1] 
Therefore, it is 
necessary to develop new energy resources and materials. 
Fuel cells are promising energy conversion devices that can produce electrical energy 
directly from chemical energy through electrochemical processes.
[2]
 Due to their high energy 
conversion efficiency, low emissions and little pollution, they have become more and more 
attractive throughout the world. 
 
1.1 Background of SOFC 
Solid oxide fuel cells (SOFCs) are one kind of fuel cells. A SOFC stack usually consists 
of four key components, including the anode, cathode, electrolyte and interconnects. Unlike 
other fuel cells (except proton exchange membrane (PEM) fuel cell), SOFCs are 
all-solid-state configurations. The most widely used materials for SOFC components are: 8 
mol% yttria-stabilized zirconia (8-YSZ) electrolyte, nickel/yttria-zirconia cermet anode 
(Ni/ZrO
2
), strontium-doped lanthanum manganite (La
1-x
Sr
x
MnO
3
) cathode and doped 
lanthanum chromite (LaCrO
3
) interconnect.
[3]
 The solid-state characteristic of cell 
 2 
components gives little restriction that required on the cells? design.
 
Currently, there are four 
common stack configurations for SOFCs? fabrication: sealless tubular design, 
segmented-cell-in-series design, monolithic design, and flat-plate design.
 [4]
 Among all the 
configurations, it is reported by Mori et al that the planner cell design exhibits highest power 
density, shortest conductive path and lowest production cost.
 [5]
 The schematic diagram of a 
planar SOFC structure is shown in Fig. 1.1.
 [6]
 
 
 
Fig. 1.1 Schematic diagram of a planar SOFC structure. Adapted with permission from D. 
Bhattacharyya, and R. Rengaswamy 2009,
 [6]
 Courtesy American Chemical Society. 
 
A SOFC unit contains two porous electrodes separated by a gas-tight oxygen-conducting 
electrolyte. During operation, fuel gas, such as hydrogen, methane, and carbon monoxide is 
fed into the anode (fuel electrode). On the other side, the air containing oxygen is supplied to 
the cathode (air electrode), where the oxygen (O
2
) is reduced to the ions (O
2-
). The ions of 
oxygen pass through the electrolyte to the interface of electrolyte and anode, and react with 
the fuel which is then oxidized. Simultaneously the electrons (e) are released. The solid 
 3 
electrolyte is nonconductive to electrons, so if there exists an external electrical connection 
between the cathode and anode for the electrons to flow through, direct current will be 
produced. All the reactions happen at the three phase boundaries (TPBs), where gas, 
electrolyte, and porous electrode meet together. The total reactions of different fuels can be 
expressed by the following equations: 
H
2
+O
2
?2H
2
O                             (1.1) 
CH
4
+2O
2
?2H
2
O+CO
2
                       (1.2)
 
2CO+O
2
?2CO
2
                            (1.3) 
The principles of electrochemical processes with hydrogen as fuel in SOFC are illustrated 
in Fig. 1.2.
 [7]
 
 
Fig. 1.2 Schematic diagram of SOFC principles. Adapted from A. Boudghene Stambouli, and E. 
Traversa 2002,
 [7]
 Courtesy Elsevier. 
 
The internal electrochemical reactions make the SOFC quite unlike the conventional 
combustion engine. Because the SOFC converts the chemical energy into electrical energy 
without fuel combustion as an intermediate step, its conversion efficiency is not subject to the 
 4 
Carnot limitation.
[4]
 Similar to other fuel cells, SOFC has much low level of pollutants (e.g., 
SO
x
, NO
x
, volatile organic carbon (VOC)). However it shows more unique advantages. Its 
high working temperature, 500 
o
C - 1000 
o
C, promotes rapid reaction kinetics, realizes 
hydrocarbon fuels reforming within SOFC (internal reforming), and produces high-quality 
byproduct heat suitable for use in cogeneration or bottoming cycle.
 [2] 
It has higher energy 
conversion efficiency, 70-80%, compared with other kinds of fuel cells.
 [8]
 Besides, its 
solid-state construction gives SOFC more merits like modular construction, non-erosion, and 
reliability during operation. It can be used as auxiliary power unit (APU) in vehicles and 
trucks, combined heat and power generation (CHP) devices, or battery replacements on 
remote sites. Due to its superior performance and promising potential, SOFC technology is 
being extensively developed for military, residential and industrial applications.
 [9,10]
 
 
1.2 Interconnect of SOFC 
 
1.2.1 Materials requirements for SOFC interconnect 
High temperature operation makes the SOFC work higher efficient, but also limits its 
choice of material, so the major challenge of the SOFC is to develop suitable component 
materials.
 [3,4]
 The interconnect is a critical part in the SOFC. An open circuit potential (OCV) 
of a single hydrogen-air SOFC at 1000
o
C is about 1V. To obtain higher voltage output, 
SOFCs are operated in series or in form of a stack, in which interconnects connect the 
cathodes and anodes of the adjacent cells electrically. Besides, it also separates the air and 
fuel atmospheres of different units from contacting. Therefore, the requirements of the 
 5 
interconnect are stringent. To pursuit better performance, the interconnect must fulfill the 
following criteria 
[4,11]
: 
(1) Sufficiently high electrical conductivity and low ohmic losses. 
(2) Low permeability for fuel gas and air during cell operation. It is the leakage in 
interconnects that results in a noticeable decline in the cell efficiency. 
(3) The coefficient of thermal expansion (CTE) of the interconnect is supposed to be 
compatible with those of the other cell components. As the CTE of 8-YSZ 
electrolyte, around 10.5 x 10
-6
/
o
C, remains constant regardless of temperature and 
oxygen partial pressure, interconnects should have close CTE value to 8-YSZ 
electrolyte to minimize the thermal stresses. 
(4) The interconnect is exposed to both reducing and oxidizing atmospheres, so it must 
have excellent physical and chemical stability under different working conditions. 
(5) Adequate mechanical strength at high temperature. 
(6) Reasonable thermal conductivity, which is beneficial for the internal reforming 
within SOFC. 
(7) Easy to fabricate, and low manufactural cost for the industrial production and wide 
application of SOFC. 
The requirements above for interconnect eliminate all but a few materials from 
consideration. There are in general two types of materials - ceramics and metallic alloys. 
 
1.2.2 Lanthanum chromites interconnect 
In early times, CoCrO
4
 was used as ceramic interconnect, but soon replaced by 
 6 
lanthanum chromite (LaCrO
3
), which
 
is now widely used in high-temperature SOFCs 
(800-1000
o
C). LaCrO
3
 is a perovskite oxide with a melting point over 2400
o
C.
[12]  
But due to 
the high vapor pressure of chromium species, LaCrO
3
 has a poor sinterability even at high 
temperature.
[13]
 Many efforts have been made, and doping is found an effective way to 
improve its sinterability.
 
An acceptable relative density was reported greater than 94% by 
doping with calcium and strontium. 
[14] 
Dopants can help to reduce the volatility of chromium 
from the surface during sintering, and also form a transient liquid phase and enhance its 
densification. Several other approaches have been developed to decrease the sintering 
temperature of LaCrO
3
 below 1600
o
C, including synthesis of highly reactive LaCrO
3
 by 
sol-gel method, controlling sintering atmospheres, employing non-stoichiometric 
compositions, sintering with Cr
2
O
3
 plates in sandwich configuration, etc.
 [15-20]
 
To increase the output power, the internal resistance of the cell should be decreased. 
Undoped LaCrO
3
 is a p-type conductor, and its electrical conductivity is quite low, about 
0.6-1.0 S/cm at 1000
o
C in air.
 [5,21,22] 
Studies have shown its conductivity can be improved by 
doping with alkaline-earth elements (AE) on lanthanum site or transition metals (TM) on 
chromium sites.
 [2,5,25]
 In air atmosphere, for doped LaCrO
3
 charge neutrality is maintained by 
Cr
3+
?Cr
4+
 transition and release of an electron.
 
These released electrons are in semi-bound 
state and easy to be stimulated, which cause a considerable increase in the electrical 
conductivity. The charge neutrality can be described as
[2]
: 
[ ] [ ] [ ]
?
=
Cr
'
Cr
'
La
CrNorM                      (1.4) 
where, [ ] indicates the concentration, 
M is the AE cation, and N is the TM cation. 
 7 
Its electrical conductivity also shows dependence on temperature, dopant content and 
oxygen partial pressure.
 [23-26]
 Jiang et al investigated the electrical conductivity of different 
cations doped LaCrO
3
 in air and 10% H
2
/N
2 
atmospheres, respectively.
 [25]
 It is shown in Fig. 
1.3 that doping with calcium, strontium, or barium can increase the conductivity of LaCrO
3
, 
and conductivities of doped and undoped LaCrO
3 
in air were higher than that in 10% H
2
/N
2 
atmosphere. According to Equation 1.4, the concentration of Cr
4+
 increases with the dopant 
content, and then conductivity is improved. Under the reducing atmosphere (anode side), the 
conductivity decrease a lot due to the existence of oxygen vacancies from ionic compensation 
mechanism, which can be described by: 
[ ] [ ] [ ] [ ]
???
+=
OCr
'
Cr
'
La
V2CrNorM               (1.5) 
where, V
o
??
 is the oxygen vacancy. 
 
 
 8 
 
Fig. 1.3 Electrical conductivity for undoped LaCrO
3
, doped LaCrO
3
, and 
(La
0.73
Sr
0.25
)(Cr
0.5
Mn
0.5
)O
3
 perovskite in (a) air and (b) 10% H
2
/N
2
. Adapted from S. Jiang et al 
2008,
 [25]
 Courtesy Elsevier.
 
 
 
The interconnects tend to distort due to isothermal expansion in the reducing and 
oxidizing environment on both sides. Then thermal stresses could be generated, and may lead 
to a crack if the thermal expansion behavior between interconnect and other components are 
not matched. The CTE of undoped LaCrO
3
 is around 8.6?10
-6
/
o
C from room temperature to 
1000 
o
C, which is lower than that of 8-YSZ electrolyte.
 [27]
 Doping is also effective to adjust 
its CTE with calcium, strontium, magnesium.
 [27-30]
 Strontium doping is efficient to increase 
the CTE. Mori et al also studied the thermal expansion behavior of Cr-site doped 
La
0.9
Sr
0.1
CrO
3
.
 [27]
 Their results summarized in Table 1.1 showed doping with Co can increase 
the CTE of La
0.9
Sr
0.1
CrO
3
 close to that of 8-YSZ electrolyte in air. Similar to conductivity 
behavior, the thermal expansion of LaCrO
3
 also depends on oxygen partial pressure. This is 
related to the increase in oxygen vacancies concentration under the reducing atmosphere. At 
 9 
the same oxygen partial pressure, higher dopant content can generate more oxygen vacancies, 
and thus higher thermal expansion.
 [2,26]
 
 
Table 1.1 CTEs of La
0.9
Sr
0.1
Cr
1-x
M
x
O
3
 (M= Mg, Al, Sc, Ti, Mn, Fe, Co, Ni, 0<x<0.10) 
perovskites at 50-1000 
o
C in air or H
2
 atmosphere. Adapted from M. Mori, Y. Hiei, and T. 
Yamamoto 2008,
 [27]
 Courtesy John Wiley and Sons.
 
Sample 
CTE (?10
-6
/
o
C) 
In air 
In H
2
 
1st 2nd 
8-YSZ 10.3 10.3 10.3 
LaCrO
3
 8.6 8.6 8.6 
La
0.9
Sr
0.1
CrO
3
 9.5 10.5 10.6 
La
0.9
Sr
0.1
Cr
0.95
Mg
0.05
O
3
 9.2 10.3 10.1 
La
0.9
Sr
0.1
Cr
0.95
Al
0.05
O
3
 9.7 10.8 10.3 
La
0.9
Sr
0.1
Cr
0.95
Ti
0.05
O
3
 9.0 9.3 9.2 
La
0.9
Sr
0.1
Cr
0.95
Mn
0.05
O
3
 10.2 11.2 10.9 
La
0.9
Sr
0.1
Cr
0.95
Fe
0.05
O
3
 9.5 10.2 10.1 
La
0.9
Sr
0.1
Cr
0.95
Co
0.05
O
3
 10.4 11.5 10.9 
La
0.9
Sr
0.1
Cr
0.95
Ni
0.05
O
3
 9.4 10.1 9.7 
 
Although the properties of LaCrO
3
 have been much improved through cation doping, it 
still faces some challenges. With sintering aids, it still needs high temperature above 1400
o
C 
 10 
to achieve high relative density. Lanthanum and its rare-earth element dopants also are 
expensive. As a ceramic material, processing is difficult for LaCrO
3
, which limits the 
geometry design of interconnect. So significant efforts have been made to find alternative 
interconnect materials. 
 
1.2.3 Metallic alloy interconnect 
The SOFC technology has been developed toward lower operation temperature in the 
range of 500
o
C - 800
o
C, but their power density and performance still remain at the same 
level as high-temperature SOFCs. This progress is realized by decreasing the thickness of 
8-YSZ electrolyte for lower ohimc polarization in anode-supported SOFC or using new 
electrolyte materials like doped CeO
2
 and LaGaO
3
 with higher ionic conductivity.
[31-35]
 The 
reduced working temperature also brings wider materials choice for the interconnect, and 
therefore metallic alloys seem to be good candidates. Compared with LaCrO
3
-based ceramic 
counterparts, their fabrication is much easier for the production of large components and raw 
materials are inexpensive, both of which decrease the manufacturing cost. Besides, metallic 
alloys have higher electrical conductivity, and their excellent thermal conductivity can 
promote internal endothermal fuel reforming. Their ductility enables metallic alloys more 
durable and reliable from environmental attack under the dual atmospheres. Their 
thermal-grown oxide scale on the surface in the cathode side can provide sufficient oxidation 
resistance as long as 40 000 h for the service lifetime of SOFC.
[2]
 In addition, the interconnect 
provides mechanical support for other ceramic components and gas channel, and metallic 
alloys possess superior strength to LaCrO
3
-based ceramics.  
 11 
The metallic alloys also have some challenges, e.g. the oxidation when simultaneous 
exposure to both air and fuel atmospheres during high temperature.
 [36-38]
 According to 
Wagner?s oxidation theory, the growth of oxidation scale is related to the transport of oxygen 
ions or chromium cations through the scale by lattice diffusion.
[39,40]
 The growth of the oxide 
scale obeys the parabolic law:  
KT/E0
g
22
g
p
2
ox
ek
d
t
t
)(
k
tk
?
===
??
?
               (1.6) 
where, ?  is the oxide scale thickness at time t and t=0; 
      k
p
 and k
g
 are the rate constants in thickness and weight, respectively; 
      ?  is the weight fraction of oxygen in the oxide (e.g. for Cr
2
O
3
, ? = 0.316); 
      ?  is the oxide density (e.g. Cr
2
O
3
, ? = 5.225 g/cm
3
); 
     and E
ox
 is the activation energy for the oxide scale growth. 
The parabolic law can be applied to the thick and homogeneous oxide scale. The 
parabolic rate constant k
g
 is an important parameter to measure the oxidation resistance 
property of alloys. The candidate metallic interconnect materials should have a low k
p
. 
High-temperature oxidation-resistant alloys (namely, superalloys) are often considered 
as the choice for interconnect materials.
[37,41] 
Alumina forming alloys have good oxidation 
resistance during high temperature, because their active element Al is easy to oxidize at the 
alloy surface and form Al
2
O
3
 layer to prevent further oxidation.
[42,43]
 But Al
2
O
3
 is insulating, 
so alumina forming alloy is not as good as chromia forming alloy in terms of electrical 
conduction, whose oxidation scale Cr
2
O
3
 is a p-type semiconductor and has a higher 
conductivity, about 0.01 S/cm at 800
o
C in air.
 [44-48]
 Different kinds of chromia forming alloys 
for SOFC interconnect application can be represented in a Fe-Cr-Ni phase diagram illustrated 
 12 
in Fig. 1.4.
 [49]
 Their details are introduced in the following part. 
 
Fig. 1.4 Schematic of alloy design for SOFC applications.
 
Adapted from Z. Yang et al 2003,
 [49]
 
Courtesy The Electrochemical Society. 
 
1.2.3.1 Chromium-based alloys 
Chromium metal has been considered as an alloy base since the late 1940s due to its 
high melting point (1863
o
C), good oxidation resistance, lower density and higher thermal 
conductivity than most superalloys.
[49,50]
 Unalloyed chromium has quite low strength, about 
70 MPa at 1200?C under compression tests.
 [51]
 To reduce their embrittlement and increase 
mechanical strength at high temperature, trace mount of reactive elements like Y, La, Ce, etc 
are added as elements or oxides to form the oxide dispersion strengthened (ODS) alloys. 
[41,52,53] 
Strengthening is realized by increasing its ductile-to-brittle transition temperature 
(DBTT) or the alloy density. Their oxidation resistance to the thermal cycling is also 
increased by dispersion of the rare-earth oxides, because these oxides effectively improve the 
adherence of the scale to the metallic matrix and reduce the oxidation rate.
[54]
 A good 
example is Plansee Ducrolloy with a composition of 94 wt% Cr, 5 wt% Fe and 1 wt% Y
2
O
3
 
(Cr-5Fe-1Y
2
O
3
). Larring et al studied its high-temperature corrosion under different water 
 13 
vapor pressure.
 [55]
 The rates of both Cr
2
O
3
 formation and its evaporation was found to 
decrease after dip-coating in Ce(NO
3
)
3
 solution and heating in H
2
/Ar atmosphere at 900
o
C for 
3 h (Ce-treatment on the alloy surface). CeO
2
 have high oxygen ion conductivity, and if it 
segregated on the surface, the reason why the rate of Cr
2
O
3 
evaporation was reduced could be 
ascribed to the Cr
2
O
3
-free surface area. Chromium based alloys is body centered cubic (BCC) 
structure. Besides lower electrical resistance, their another advantage is their thermal 
expansion is close to that of 8-YSZ electrolyte. Quadakkers et al made a comparison of 
thermal expansion behavior between chromium based alloys and stainless steels, and stainless 
steels showed larger thermal expansion than Cr-0.4La
2
O
3
 and Cr-5Fe-1Y
2
O
3
.
 [56]
 The CTE 
value of Cr-5Fe-1Y
2
O
3 
is 11.8 x 10
-6
/
o
C from room temperature to 1000
o
C. 
Although ODS chromium based alloys possess excellent oxidation resistance, 
mechanical properties and compatible CTE, their fabrication is still difficult and costly 
especially compared with other commercial metallic alloys. In addition, some problems are 
inevitable. Their higher chromium concentration can lead to excessive Cr
2
O
3
 scale growth, 
and volatile chromium species cause cathode poisoning 
[56]
, which will be discussed in detail 
later. Therefore, increasing efforts have been focused on other Cr forming alloys. 
 
1.2.3.2 Ni-Cr-based alloys 
Ni-Cr-based alloys have a face centered cubic (FCC) structure. Compared with ferric 
stainless steels, they exhibit lower scale growth rate and much higher mechanical strength.
[57]
 
The chromium content is required to be around 15% in Ni-Cr-based alloys to form a 
continuous chromia layer, lower than that of Fe-Cr based alloys, which is a optimum content 
 14 
of 18-19%.
[38]
 England et al reported the oxidation kinetics of different Ni-based alloys in air 
and humidified hydrogen. 
[58,59]
 Their major composition in the oxidation scale is Cr
2
O
3
, and 
(Mn,Cr)
3
O
4
 was detected as a second phase. Their oxidation behavior obeys the parabolic law. 
Haynes 230 with the highest Cr content showed the slowest oxidation rates in air, 
corresponding with lowest ASR. However, the oxidation kinetics of these Ni-based alloys at 
800
o
C in wet hydrogen was faster than that in air. The ASR of the oxidation layer grown in 
wet hydrogen was also higher in comparison with that in air. Yang et al studied the oxidation 
behavior of different Ni-Cr-based alloys with different chromium content in air, hydrogen 
and moist air-moist hydrogen dual atmospheres.
 [60]
 A thin layer of Cr
2
O
3
 and (Mn,Cr,Ni)
3
O
4
 
spinels formed in air on Haynes 230 and Hastelloy with a relative high Cr content (22 wt% 
and 16 wt%, respectively). However, a thicker double layer more than 5 ?m with NiO scale 
on top of a chromia-rich substrate was observed in Haynes 242 with only 8 wt% Cr in the 
moist air and at the air side of dual-atmosphere conditions. A similar thin layer formed on the 
three alloys after oxidation in moist hydrogen, indicating high oxidation resistance. The 
layer?s composition are Cr
2
O
3
 and (Mn,Cr,Ni)
3
O
4
 spinels for Haynes 230 and Hastelloy S, 
and Cr
2
O
3
 for Haynes 242. NiO did not appear due to its thermodynamical instability in the 
reducing condition. At dual atmosphere hydrogen flux from the fuel side to the air side is 
likely to alleviate NiO growth and make the interface between the scale and metal dense. The 
FCC Ni?Cr-base alloys usually demonstrate a higher CTE than the BCC ferritic stainless 
steels, which typically have a CTE of 12 - 13 x10
-6
/
o
C.
 [49]
 Haynes 230 and Hastelloy S 
showed a higher CTE value about 15 x10
-6
/
o
C from room temperature to 800?C. Lowest CTE 
was obtained in Haynes 242, 13.1 x10
-6
/
o
C, which is ascribed to heavy additions of Mo, and 
 15 
the limited amount of Cr. Haynes 242 is the recently developed Ni-Cr-based alloy. It can be 
used as gas turbine components, because it has high mechanical strength and lower CTE than 
other Ni-based alloy like Haynes 230.
[61]
 Its thermal exapnsion is more compatible with other 
SOFC components. Geng et al also investigated Haynes 242 alloy in different 
environments.
[62, 63]
 Consistent with Yang et al?s work, a NiO scale was observed on the top 
of Cr
2
O
3
 and (Mn,Cr)
3
O
4
 scale in the air oxidation too. The NiO scale did not continuously 
distributed on the alloy surface, however there is no apparent scale spallation. Its mass gain 
after isothermal oxidation at 800 ?C for 500 h in air was less than that of Crofer, and its ASR 
of the oxide scale was lower that that of Ebrite and Crofer. Haynes 242 showed a quite 
different behavior in Ar + 4 vol% H
2 
+ 3vol% H
2
O atmosphere. The oxidation layer consisted 
of a top thin (Mn,Cr)
3
O
4
 spinel scale and an inner Cr
2
O
3
 scale. The calculated activation 
energy of oxidation was near that of chromium self-diffusion in Cr
2
O
3
, which suggested the 
chromia layer is formed by the outward diffusion of chromium. 
Ni?Cr-base alloys have shown excellent oxidation resistance and sufficient strength for 
SOFC interconnect. Its significant challenge is the CTE mismatch to other components. To 
take advantages of it potential, new designs are required for interconnects and stacks. The 
identification to optimize alloy compositions is also a good solution.
[64]
 
 
1.2.3.3 Fe-Cr-based alloys 
Fe-Cr-based alloys include a large number of alloys, and they are also termed as 
stainless steel (SS). Compared with Cr-based alloys and Ni-Cr-based alloys, they are more 
promising due to their easy fabrication, reasonable thermal expansion and low maunfactural 
 16 
cost. Stainless steels are often divided into four groups based on their crystal structure: (1) 
ferritic stainless steels (FSS); (2) austenitic stainless steels (ASS); (3) martensitic stainless 
steels; and (4) precipitation hardening steels.
[49]
 Among them, the ferritic stainless steels are 
the most attractive due to their BCC structure which enable them to have more matched CTE 
(11.5-14 x10
-6
/
o
C from room temperature to 800
o
C) with other SOFC components. Besides, 
they also have good oxidation resistance and low fabrication cost. 
In order to develop a continuous Cr
2
O
3
 scale, enough chromium content is required in 
the substrate alloy. According to the Fe-Cr phase diagram, at least 13 wt% Cr should be 
contained to stabilize their ferritic structure phase.
[65]
 In the commercial stainless steels, the 
Cr content is usually in the range of 11- 30 wt%.
[49] 
High-Cr ferritic alloys with Cr 
concentration over 16 wt% showed good oxidation resistances.
[66,67] 
The minimum chromium 
content to form a homogenous dense Cr
2
O
3
 scale is also affected by the oxidation 
environments. Othman et al?s work showed water vapor and high temperature (700 
o
C - 
950
o
C) both can accelerate isothermal scaling rates, and increaser the chromium level.
[68] 
The 
elements like C and N which lead to form austenite must be controlled in a low content. The 
impurities e.g. Si and S should be eliminated, because Si can easily produce continuous 
insulating layers at the interface, and trace amounts of S can cause spallation of oxide scale.
 
[69]
 
 
E-Brite is a highly pure ferritic stainless steel which was developed by Allegheny 
Ludlum Steel Corporation for SOFC. Its typical compositions are: Fe (balance), Cr (27.5 wt% 
max), Ni+Cu (0.5wt% max), Mn(0.4 wt% max), Si (0.4 wt% max), Mo (1.5 wt% max), N 
(0.0015 wt% max), and C (0.01 wt% max).
[70]
 Huang et al studied its oxidation kinetics in 
 17 
stagnant air.
 [37]
 Its scale contained primary Cr
2
O
3
, and its growth rate showed a parabolic 
dependence on time. Its CTE was 11.8 x10
-6
/
o
C from room temperature to 500
o
C, quite close 
to other cell materials. However, its ASR with oxide electrodes was also near the limit for 
SOFC. A similar results was also observed in Geng et al?s work.
[71]
 The weight gain of Ebrite 
was the smallest among the ferritic alloys they studied, which was related to the dense Cr
2
O
3
 
to prevent further oxidation. Besides, no oxide scale spallation was detected after the cyclic 
oxidation test.  
Crofer 22 APU is another widely-used high-temperature ferritic stainless steel for SOFC 
interterconnect, which was developed by Thyssenkrupp Steel AG. Its chemical compositions 
are: Fe (balance), Cr (24 wt% max), Mn (0.80 wt% max), Si (0.50 wt% max), Cu (0.5wt% 
max), Al (0.5 wt% max), S (0.0020 wt% max), P (0.050 wt% max), Ti (0.020 wt% max), and 
C (0.03 wt% max).
[72]
 This alloy contains high concentration of Mn, and its protective oxide 
scale consist of Cr
2
O
3
 and (Cr,Mn)
3
O
4
 spinel, which is different from E-brite. The diffusion 
rate of different alloy elements in Cr
2
O
3
 follows the increasing rank of Cr, Ni, Co, Fe, Mn. 
[73] 
So during oxidation, the rapid outward diffusion of Mn lead to the raction with Cr and 
oxygen and formation of a Mn-containing spinel layer on top of the continuous Cr
2
O
3 
matrix. 
Liu also reported MnO was observed with Mn-Cr spinel on the surface of Crofer 22 APU 
after oxidation at 900 ?C for 1000 h in the reducing atmosphere of 
Ar + 5 vol.% H
2
 + 3 vol.% H
2
O. 
[74] 
Fontana et al compared oxidation resistance of Crofer 22 
APU with Fe30Cr (30 wt% Cr but no Mn).
 [75]
 Fe30Cr (k
p
= 3.14 x10
-13
g
2
/cm
4
s in air, and 
1.54 x10
-12
g
2
/cm
4
s in H
2
/10%H
2
O) showed higher mass gain than Crofer 22 APU (k
p 
= 2.5 
x10
-14
g
2
/cm
4
s in air, and 3 x10
-14
g
2
/cm
4
s in H
2
/10%H
2
O) after heating at 800
o
C for 100 h. 
 18 
Crofer 22 APU also possessed low rate of chromium vaporization. Konysheva et al evaluated 
the chromium release of Crofer 22APU and ODS Cr-5Fe-1Y
2
O
3
 alloy.
[76]
 A (Cr,Mn)
3
O
4
 
spinel scale with a thickness of 0.5 micrometer effectively decreased the Cr evaporation rate 
of Crofer 22 APU by a factor of 3 compared with alloy only forming Cr
2
O
3
 scale at 800
o
C in 
humidified air for a time up to 1300 h. They also found change in the minor additives like Al 
and Si in Crofer 22 APU can influence the oxide scale thickness, morphology and the 
(Cr,Mn)
3
O
4
/Cr
2
O
3 
ratio. Yang et al evaluated oxidation behavior and the conductivity of 
Crofer 22 APU and other ferritic Fe-Cr-Mn steels.
[66]
 A chromia-rich scale beneath 
(Mn,Cr)
3
O
4
 spinel rich layer was formed both in air and moist hydrogen (97%H
2
/3%H
2
O) at 
800 
o
C for 1200 h. Its ASR was only 13.0 m?cm
2
 after 1800 h at 800 
o
C, much lower than 
that of other ferritic steels like E-Brite and AISI-SAE 446. The low value is ascribed to 
higher electrical conductivity of the (Mn,Cr)
3
O
4
 spinel phases than Cr
2
O
3
.
[77]
 Overall, Crofer 
22 APU appears to be one of the most promising candidate for SOFC interconnect alloys. 
It should be noted that the presence of reactive elements could affect the oxidation 
behavior of alloy. AISI-SAE 430 is a commercial ferritic grade without reactive elements. 
AISI-SAE 441 alloy is based on AISI-SAE 430?s composition, but with the addition of 
reactive elements Ti and Nb. Rufner et al? work showed their oxidation behavior at the dual 
atmosphere is quite different.
[78]
 A porous Fe
2
O
3
-rich surface layer was formed on the surface 
of AISI 441 at the air side, while AISI 430 had isolated Fe-rich surface nodules. This may be 
partly related to Nb promoting the hydrogen transport in 441 by segregating at grain 
boundaries as a short-cut pathway. Then hydrogen would react with the oxide layer to 
enhance the iron outward diffusion from its substrate. A study by Shaigan et al compared the 
 19 
metal-oxide scale performance between ZMG 232 and AISI-SAE 430. 
[79]
 ZMG 232 is 
Fe-22% Cr based alloy with small amounts of reactive elements like Zr and La, developed by 
Hitachi Metals especially for SOFC interconnect.
 [80,81]
 Both of them were oxidized at 800
o
C 
in static air for various periods. Scale spallation was observed in AISI-SAE 430 just after 
oxidation less than 100 h, while it did not appear in ZMG 232 ever after longer hours. This is 
attributed to the segregation of impurity elements like Si, S, Cl and so on at the interface 
between alloy and oxide scale in AISI-SAE 430, and therefore the scale adhesion and its 
contact area with alloy is reduced. However, the reactive elements Zr and La effectively 
prevent the segregation of impurity elements in ZMG 232. The reactive elements can 
suppress the oxide growth rate and impurities of segregation by influencing the diffusion of 
chromium and oxygen. 
[82-85]
 The reactive element oxide segregate at the interface between 
scale and substrate, and inhibit voids formation. A dynamic-segregation model of the reactive 
element outward diffusion during high temperature oxidation is illustrated in Fig. 1.5.
[69]
 
 
Fig. 1.5 A dynamic-segregation model of the reactive element function during high 
 20 
temperature oxidation. Adapted from B. A. Pint 1995,
 [69]
 Courtesy Springer.
 
 
 
A comparison between different chromia forming alloys is listed in Table 1.2.
 [49]
 
Ferritic stainless steel is considered as the most attractive candidate in terms of its good 
oxidation resistance, compatible thermal expansion with 8-YSZ electrolyte, and the low cost. 
 
Table 1.2 Properties of different chromia forming alloys for SOFC interconnect. Adapted from 
Z. Yang et al 2003,
 [49]
 Courtesy The Electrochemical Society. 
Alloys 
Matrix 
Structure 
CTE 
x10
-6
/
o
C 
(RT-800
o
C) 
Oxidation 
Resistance 
Mechanical 
Strengths 
Manufacturability Cost 
Cr base alloys BCC 11.0-12.5 Good High Difficult 
Very 
Expensive 
Ferritic 
stainless steels 
BCC 11.5-14 Good Low Fairly Readily 
Less 
Expensive 
Austenitic 
stainless steels 
FCC 18-20 Good Fairly High Readily 
Less 
Expensive 
Fe-Ni-Cr base 
alloys 
FCC 15-20 Good High Readily 
Fairly 
Expensive 
Ni(-Fe)-Cr 
base alloys 
FCC 14-19 Good High Readily Expensive 
 
 21 
In addition, some low-Cr Fe based alloys are also studied. A Fe-Co-Ni based steel with 
only 6wt% Cr was evaluated by Geng et al.
[86]
 After oxidation at 800 ?C in air for 12 weeks, a 
Cr-free (Fe,Co,Ni)
3
O
4
 spinel layer formed atop Cr
2
O
3
 scale, which effectively blocked Cr 
violation, decreased the scale growth rate and ASR. The low-Cr alloy also showed good scale 
adherence and compatibility with the La
0.8
Sr
0.2
MnO
3 
cathode material. 
 
1.2.4 Problems with metallic interconnect materials 
Although ferritic stainless steels are the promising metallic interconnects material, they 
still face some challenges at SOFC operation condition, like high growth rate of oxide scale, 
evaporation of Cr species, etc.  
 
1.2.4.1 Oxidation problem 
    The thermally grown Cr
2
O
3
 scale is inevitable during oxidation for all chromium 
forming alloys. For Fe-Cr-Mn stainless steels, their oxidation layer also contain (Mn,Cr)
3
O
4
 
spinel. The electrical conductivity of both Cr
2
O
3
 and (Mn,Cr)
3
O
4 
is much lower than that of 
alloy matrix. Intensive studies have shown the thickness of Cr
2
O
3
 scale on stainless steels can 
reach several micrometers or even tens of micrometers after oxidation for thousands of hours 
under SOFC working environment. The existence of water vapor can also promote the 
growth of Cr
2
O
3
 scale at high temperature.
 [68,87,88]
 Even there is no crack, as the scale grows 
further, the overall ASR will increase. In practical, attributed to the thermal stress between 
scale and alloy substrate, crack or spallation will take place, and results in the degradation of 
SOFC stack. Therefore, an improvement to alleviate oxidation of metallic alloys is required. 
 22 
 
1.2.4.2 Chromium poisoning 
Another inherent weakness of chromium forming alloys is Cr poisoning of electrodes, 
which can cause severe degradation in SOFC electrochemical performance. The negative 
effect of chromium on lanthanum strontium manganite (LSM) cathode poisoning has been 
confirmed by different groups when applying chromium forming alloys.
[89-93]
 In the cathode 
side, the Cr
2
O
3
 layer on the surface of metallic alloy reacts with water and oxygen molecules, 
and generates volatile CrO
2
(OH)
2
 and/or CrO
3
 with Cr in the oxidation state of +6. The 
reactions can be expressed by the following equations
[94-96]
: 
)(4)(3)(2
3232
gCrOgOsOCr ?+                    (1.6) 
)()(4)(4)(3)(2
222232
gOHCrOgOHgOsOCr ?++      (1.7) 
Then the volatile Cr species migrate through the cathode, and segregate at the 
three-phase boundary (TPB, cathode, electrolyte and oxygen) as Cr
2
O
3
. The reduction of 
CrO
2
(OH)
2
 to solid Cr
2
O
3
 can dramatically deteriorate the cathode electrochemical activity, 
described by Equation 1.9 and 1.10.  
??
++=+
2
2323
O3)g(OH2)s(OCre6)g(CrO2            (1.8) 
??
+=+
2
3222
O3)s(OCre6)g()OH(CrO2               (1.9)  
Yokokawa et al also analyzed the reactions of cathode with chromium vapors in 
thermodynamics, and found the Cr
2
O
3
 deposition or CrMn
2
O
4
 formation at the 
cathode/electrolyte interface is the direct reason for the chromium poisoning which results in 
slow oxygen ion diffusivity in LSM cathode.
 [97]
  
 23 
 
Fig. 1.6 Vapor pressures of different volatile chromium species as a function of temperature. 
Adapted from W. Z. Zhu, and S.C. Deevi 2003,
 [2]
 Courtesy Elsevier.
  
 
 
Fig. 1.7 Equilibrium vapor pressures of chromium?oxygen?hydrogen gas species at 800 
o
C 
with a water vapor pressure of 3 kPa as a function of oxygen partial pressure. Adapted from J. 
Fergus 2005,
 [54]
 Courtesy Elsevier. 
 
The vapor pressures of different volatile chromium species as a function of temperature 
is displayed in Fig. 1.6. 
[2]
 It is obvious that CrO
2
(OH)
2
 possesses the highest vapor pressure 
 24 
in the IT-SOFC working temperature range of 600-800
 o
C among all Cr species. Fig. 1.7 
shows the equilibrium vapor pressures for Cr-O-H species at 800
o
C with a water vapor 
pressure of 3 kPa as a function of oxygen partial pressure.
[54]
 Their equilibrium vapor 
pressure increases with oxygen partial pressure. So chromium poisoning favors to happen in 
the cathode side. Currently increasing studies has been focused on novel cathode materials 
with higher catalysis activity to resist chromium poisoning.
[98-104]
 
Chromium forming alloy is difficult to achieve satisfactory performance when using 
directly without any surface modification. To suppress its excessive Cr
2
O
3 
scale growth and 
chromium violation to cathode, many efforts have been made such as alloy composition 
optimization, surface treatment, etc. Among them, applying protective ceramic coating on the 
alloy surface is an effective method. 
 
1.3 Interconnect coating materials for SOFC 
The coating material is applied between alloy interconnect and cathode. It servers as a 
barrier to chromium outward diffusion and oxygen inward diffusion, and therefore decreases 
the oxidation rate of metallic interconnect and chromium evaporation rate. To fulfill its 
function, it should satisfy the following requirements:  
(1) Dense and low diffusion rate of chromium and oxygen to hinder their migration in 
the coating; 
(2) Matched thermal expansion with substrate and cathode material to avoid internal 
stress and local spallation after thermal cycling; 
(3) High electrical conductivity to decrease the cell resistance; 
 25 
(4) Non- volatile and chemical stability with the alloy interconnect, cathode, and contact 
layer at high temperature. 
 Various coatings have been developed to protect metallic interconnects, and they can 
be categorized mainly into three kinds: reactive element oxide, perovskite oxide, and spinel 
oxide coatings.
[43]
 
 
1.3.1 Reactive elements coating 
As discussed in the above section, the addition of reactive elements or its oxides in 
Cr-forming alloys to form dispersed phase effective improved the adherence of the oxide 
scale to the substrate and reduce the oxidation rate by changing the diffusion mechanisms of 
chromium and oxygen. They can also be used as protective coating on the alloy surface, and 
the often used coating elements are Y, La, Nd and Ce.
[105-107]
 Fontana et al studied three 
coatings including La
2
O
3
, Nd
2
O
3
 and Y
2
O
3
 on different metallic alloys (Crofer 22 APU, AL 
453 and Haynes 230) applied by metal organic chemical vapour deposition (MOCVD).
[108]
 
The oxidation behavior was compared between coated and uncoated alloys after aging at 
800
o
C for 100 h in air. The Cr
2
O
3
 scale thickness of Crofer 22 APU was reduced by a factor 
of 1.5 after applying coating. La
2
O
3
 was acted as barrier that limits the chromium diffusion 
during oxidation. It reacted with chromia and formed LaCrO
3
 perovskite. Similarly, Nd
2
O
3
 
was concentrated at the chromia scale/top spinel grains interface as NdCrO
3
 perovskite. 
However, Y
2
O
3
 has high oxygen permeability. The chromia scale formed between Y
2
O
3 
coating and alloy, and the generated stresses cause cracks in all Y
2
O
3
 coated samples. In their 
another study of Crofer 22 APU under H
2
/10%H
2
O atmosphere, Y
2
O
3
 coating also has a 
 26 
different oxidation behavior with other coatings.
[74]
 It did not react with chromia to form 
YCrO
3 
perovskite. The weight gain of Y
2
O
3
 coated Crofer 22 APU was higher than that of the 
other two coated samples. Kim et al compared thin coatings of Y, Co and Y/Co in the 
oxidation behavior of ferritic stainless steel AISI 444 in wet air up to 1000 h at 800 
o
C. 
[109]
 
Y(30 nm), Co(50 nm) and Y(30 nm)/Co (50 nm) coatings were applied using electron beam 
evaporation method. The scale growth was reduced after coating with Y and Y/Co. The 
average scale thicknesses of the uncoated, Y-coated, and Y/Co-coated samples were 1.9 ?m, 
0.9 ?m, and 1.1 ?m. An YCrO
4
 layer was formed between MnCrO
4
 region and Cr
2
O
3
 scale, 
which was responsible for the suppressing of Cr
2
O
3
 sub-layer growth. Y
2
O
3
 coated sample 
showed the lowest and stable ASR value. Reactive element oxide coatings increase the 
oxidation resistance of stainless steels and the adhesion of oxide scale, however, their coating 
layers are usually porous and quite thin which makes them not effective as barriers to 
chromium outward diffusion compared to other types of coatings.
[110]
 
 
1.3.2 Perovskite coating 
Perovskite oxides are widely used as cathode and interconnect of SOFC, due to their 
higher electrical conductivity and compatible thermal expansion. The formula of perovskites 
is expressed as ABO
3
, where A is rare earth cation (e.g. La, Pr, Ce, etc), and B is transition 
metal cation (e.g. Cr, Co, Mn, Fe, Cu, V, etc). 
[43] 
They are also applied as conductive coatings 
on alloy interconnect. The advantages of perovskite coating include their rare earth elements 
can retard the oxidation. Besides, as discussed in La
1-x
Sr
x
CrO
3
 interconnect,
 
the conductivity 
of perovskites can be improved by A-site or B-site doping.  
 27 
Lu et al deposited a dense LaCrO
3
-based coating on AISI 444 stainless steel by sol?gel 
process.
[111]
 A smooth and uniform coating can be formed after annealing at 800
o
C for 1 h in 
air. The ASR of the coated AISI 444 is much lower than that of uncoated sample after 
oxidation at 850 
o
C for 100 h. Uncoated AISI 444 showed notable weigh loss due to the 
several spallation of the scale, while the sol?gel coatings provided the protection during the 
cyclic oxidation and improved scale adhesion. Shaigan et al also developed the Co/LaCrO
3
 
coating on AISI 430 by electrodeposition.
[112]
 After thermal exposure at 800
o
C in air up to 
2040 h, the Co/LaCrO
3
-coated AISI 430 has a triple-layer scale including a Cr
2
O
3
 inner layer, 
(Co,Fe)
3
O
4
 mid-layer and Co
3
O
4
 outer layer. Compared with bare AISI 430, the 
Co/LaCrO
3
-coated AISI 430 had a lower ASR, not exceed 0.02 ? cm
2
 after 900 h at 800 ?C 
in air, which could be attributed to high conductivity of LaCrO
3
 perovskite and the Co 
containing spinel scale.  
The high temperature SOFC cathode LaMnO
3
 is also investigated as coating material of 
interconnect. Shong et al deposited La
0.8
Sr
0.2
MnO
3
 on Crofer 22 APU using spin coating 
method, and both the coated and uncoated samples were oxidized in ambient air at 800 ?C up 
to 1600 h.
[113]
 All samples had (Mn,Cr)
3
O
4
 spinel phase in the scale, but the Mn/Cr ratio of 
the coated sample was higher, in which La
0.8
Sr
0.2
MnO
3
 coating promotes the growth of 
(Mn,Cr)
3
O
4
 spinel and increases its Mn content. The ASR of coated Crofer 22 APU was 
reduced to almost one third of that of uncoated sample, which could be attributed to the 
higher Mn content in (Mn,Cr)
3
O
4
 spinel phase. Another work focused on the LaMn
0.9
Ti
0.1
O
3
 
coating, which was applied on the surface of Haynes 230 by radio-frequency (RF) 
sputtering.
[114]
 Ti was doped in Mn site to suppress the oxygen ion transport. After oxidation 
 28 
at 800
 o
C in air for 1080 h, the thickness of oxide scale of uncoated Haynes 230 was around 
2.2 ?m, while it was only 0.63 ?m in LaMn
0.9
Ti
0.1
O
3
-coated samples. Its ASR was also 
reduced after coating.  
La
1-x
Sr
x
CoO
3
, La
1-x
Sr
x
MnO
3
 and La
1-x
Sr
x
FeO
3
 have much high electrical conductivity 
than La
1-x
Sr
x
CrO
3
, but these three are mixed electrical and ion conductor which means they 
can promote the oxygen ion diffusion at the same time. Yang et al compared La
0.8
Sr
0.2
FeO
3
 
and La
0.8
Sr
0.2
CrO
3
 coating on different ferritic stainless steels (E-brite, Crofer 22 APU, and 
AL453).
[115]
 Both of the pervoskites coatings reduced the oxidation rate of Crofer 22 APU 
after oxidation at 800 
o
C in air up to 1200 h, however, La
0.8
Sr
0.2
CrO
3
 coating (k
g
= 3.2 x10
-14
 
g
2
cm
-4
s
-1
) was more effective than La
0.8
Sr
0.2
FeO
3
 (k
g
= 6.2 x10
-14
 g
2
cm
-4
s
-1
). The similar 
results were also obtained in the oxidation of AL453 and E-brite. The ionic conductivity of 
La
0.8
Sr
0.2
CrO
3
 is 2-3 orders of magnitude lower than that of La
0.8
Sr
0.2
FeO
3
. At the same 
condition, oxygen ions can diffuse inward more easily through La
1-x
Sr
x
FeO
3
 coating at high 
temperatures. In addition, chromium was detected to diffuse into the La
0.8
Sr
0.2
FeO
3
 coating. 
Their work indicates La
0.8
Sr
0.2
CrO
3
 is superior to La
0.8
Sr
0.2
FeO
3
 coating as barrier to prevent 
chromium migration. However, one potential problem for La
1-x
Sr
x
CrO
3
 is that it contains 
chromium element, and chromium is easy to evaporate at high temperature. Then the released 
chromium could still diffuse though the coating and cause cathode poisoning.
[42,43]
 
 
1.3.3 Spinel coating 
Recently, the research trend has been focused on spinel oxides as the coating material. 
Spinel oxides are always attractive due to their excellent electronic and magnetic properties 
 29 
over a wide range of temperatures. Especially, the spinel oxides with 3d transition metals 
occupied on the octahedral sites possess the unique physical properties like ferromagnetism, 
antiferromagnetism, charge ordering, superconductivity, which are depended on the valence 
of the cations.
[116]
 
The general formula of spinel oxide is AB
2
O
4
, where O
2-
 ion are in a close packing, and 
A and B are divalent, trivalent or quadrivalent cations occupying the octahedral and 
tetrahedral sites, respectively.
  
Its crystal structure is illustrated in Fig. 1.8.
 [38] 
Spinel oxides 
have normal and inverse structures. In a spinel unit cell, there are 32 octahedral sites, and 64 
tetrahedral sites. In the normal structure, B
3+
 cations occupy half of the octahedral holes, and 
A
2+
 cations occupy 1/8 of the tetrahedral holes. A normal spinel case is Co
3
O
4
 at room 
temperature, in which Co
2+
 on tetrahedral sites and Co
3+
 on octahedral sites.
[118]
 In the inverse 
structure, all A
2+ 
cations occupy the octahedral sites, and push half of the B
3+
 cations from the 
octahedral sites to the tetrahedral sites. An ideal case of inverse spinel is Fe
3
O
4
, which can be 
expressed crystallographically in (Fe
3+
)
tetra
[Fe
2+
Fe
3+
]
oct
O
2-
4
.
[119]
 However, for most spinel 
oxides, their cations distributions are between the above two structures.
[120]
 
 
 30 
 
Fig. 1.8 Schematic diagram of crystal structure of AB
2
O
4
 spinel.
 
Adapted from J. Wu, 
and X. Liu 2010,
 [38]
 Courtesy Elsevier. 
 
The high electrical conductivity and compatible thermal expansion can be obtained by 
adjusting the A and B cations? composition and their ratio. Spinel coatings have indicated 
promising performance. Petric?s group did an extensive study on the electrical conductivity of 
spinel oxides, and found Cu
x
Mn
3-x
O
4
 system had much higher conductivity than most other 
tested spinels.
[121,122]
 The conductivity of Cu
1.3
Mn
1.7
O
4
 can reach up to 225 S/cm at 750
o
C. 
They further applied Cu
0.4
Mn
1.6
O
4
 by electroplating technique on stainless steel UNS 430, 
and the Cu-Mn spinel coating restricted Cr
2
O
3
 growth at the interface between coating and 
UNS 430 and still had good adhesion to alloy matrix after cyclic oxidation at 750
o
C as long 
as 28 days due to its matched thermal expansion with the substrate.
 [123-125]
 
   Hua et al developed NiCo
2
O
4
 coating on the surface of SUS 430 by sol-gel process.
[126]
 
The oxidation resistance of SUS 430 was significantly improved with NiCo
2
O
4 
coating by 
limiting the access of oxygen to the cations via outward diffusion during thermal exposure at 
800
o
C for 200 h in air (k
g
: 8.3 x10
-14
 g
2
cm
-4
s
-1
 for uncoated sample, and k
g
: 8.1 x10
-15
 
 31 
g
2
cm
-4
s
-1 
for coated sample). A recent work evaluated CoFe
2
O
4
 spinel oxide as a potential 
coating on Crofer 22 APU by electroplating for cell testing as long as 180 h.
[127]
 The 
channeled Crofer 22 APU interconnect with CoFe
2
O
4
 coating showed higher cell 
performance and lower degradation rate (3?10
?5 
Wcm
?2
 h
?1
) than the bare interconnect (about 
3?10
?4
 Wcm
?2
 h
?1
).  
In addition, other spinel oxides were also studied like NiMn
2
O
4
, NiFe
2
O
4,
 etc.
[128,129]
 
They all exhibit to improve the oxidation resistance and block Cr diffusion effectively. 
Among these spinel coating, (Mn,Co)
3
O
4
 oxides are the most promising coating candidates 
due to their reasonable CTE (11.4?10
-6
/?C for Mn
1.5
Co
1.5
O
4
 and 12-13?10
-6
/?C for ferritic 
stainless steel) and high electronic conductivity(60 S/cm for Mn
1.5
Co
1.5
O
4
 at 800
o
C). 
[121, 
130,131]
  
As early as 2000, Larring et al first reported (Mn,Co)
3
O
4
 coating was more effective to 
prevent Cr outward diffusion.
 [132]
 Then much more extensive studies have been focused on 
(Mn,Co)
3
O
4
 system as coating candidate currently. Chen et al applied a dense MnCo
2
O
4
 layer 
by slurry coating on AISI 430. After oxidation at 800
o
C for 60 h and 850
o
C for 120 h, its 
scale thickness was only sub-micron, and the ASR was significantly reduced by one order of 
magnitude.
[133]
 Similar results for MnCo
2
O
4
 coating are confirmed by other researchers.
[134-136]
 
Yang et al investigated the effect of Mn
1.5
Co
1.5
O
4
 coating by screen printing on E-brite. 
[130] 
According to the Mn-Co binary phase diagram, Mn
1.5
Co
1.5
O
4
 is a single cubic phase at 
800
o
C.
[137]
 Besides, its electrical conductivity is much higher than that of Cr
2
O
3
 and MnCr
2
O
4
. 
A dense coating was formed by first heat treatment in a reducing atmosphere and then in air. 
Their result showed the contact resistance between cathode and coated E-brite was as low as 
 32 
7 m?cm
2
 at early stages, and only slightly increased after 400 h at 800
o
C. The thickness of 
Cr
2
O
3
 scale was just about 1.0 ?m, which indicated Mn
1.5
Co
1.5
O
4 
coating can effectively 
inhibit the Cr
2
O
3
 growth and prevent chromium outward migration through coating. They 
also tried Mn
1.5
Co
1.5
O
4
 on Crofer 22 APU. 
[138]
 After applying Mn
1.5
Co
1.5
O
4 
coating, the 
contact ASR between Crofer 22 APU and LSF cathode was reduced from 38.6 m?cm
2 
to 
12.8 m?cm
2
 after testing at 800
o
C for 400 h. A steep Cr decrease was still observed from the 
subscale/spinel coating interface after a half-a-year test of thermal cycling as shown in Fig. 
1.9.
[138] 
The matched CTE between Crofer 22 APU and Mn
1.5
Co
1.5
O
4
 and dense coating 
structure both give rise to a good long-term stability. Xin et al reported a Mn
0.9
Y
0.1
Co
2
O
4 
spinel oxide applied on Crofer 22 APU by slurry coating.
 [139]
 With the protection of 
Mn
0.9
Y
0.1
Co
2
O
4
 coating, the ASR was significantly lowered
 
to 4 m?cm
2
 from 17 m?cm
2 
after operation at 800
o
C for 538 h even including several thermal circles. MnCo
1.9
Fe
0.1
O
4
 is 
also (Mn,Co)
3
O
4
-based coating. X. Montero et al?s study showed it can improve the oxidation 
resistance of Crofer 22 APU and F18TNb after screen coating.
 [140]
 They also compared it 
with reactive oxide coating (CeO
2
, Y
2
O
3
) and pervoskite oxide coating (La
0.8
Sr
0.2
FeO
3
, and 
La
0.6
Sr
0.4
FeO
3
).
 [141]
 The MnCo
1.9
Fe
0.1
O
4
 coated sample still showed the lowest ASR, and it 
was more effective to limit the outward migration of chromium cations than the other 
coatings. The superior performance of (Mn,Co)
3
O
4
-based coating over pervoskite oxide 
coating was also confirmed in Larring et al?s work.
[132]
 
 33 
 
Fig. 1.9 SEM images and EDS analysis on Crofer 22 APU with Co
1.5
Mn
1.5
O
4
 coating after a 
period of 6 months under thermal cycling.
 
Adapted from Z. Yang et al 2005,
 [138]
 Courtesy The 
Electrochemical Society. 
 
(Mn,Co)
3
O
4
 system have attracted increasing focuses in recent years, and different 
coating techniques have been developed on them, e.g. electrodeposition of metals, 
thermal-growth or RF-sputtering, magnetron sputtering, atmospheric plasma spraying (APS), 
aerosol deposition, etc.
 [136,142-146]
 All these work indicate (Mn,Co)
3
O
4
 system are the 
promising coating candidate for SOFC metallic interconnects. However, more works are still 
needed to achieve satisfactory performance. 
 
1.4 Objectives of the dissertation 
The excellent performances of (Mn,Co)
3
O
4
 spinel oxides have been demonstrated as 
coating material for metallic interconnect by extensive works. However, the reasons for their 
superior properties have not been fully understood, so further study is considered necessary. 
 34 
This work focuses on their interfacial reaction and long-term stability, electrical conduction 
behavior and materials optimization for improvement in coating performance and 
effectiveness. The major objective of this dissertation is to investigate the growth mechanism 
of the interaction layer and the effect of its existence on the coating and interconnect 
performance. 
The interaction between (Mn,Co)
3
O
4
 spinel coatings and chromia at high temperature is 
first studied to simulate the spinel coatings? working condition. The mass transport behavior 
of (Mn,Co)
3
O
4
 spinel oxide will be proposed. Then the physical properties of reaction layer 
formed during the interaction between (Mn,Co)
3
O
4
 spinel coatings and chromia is 
investigated by analogous bulk analyses. The transport properties (e.g. electrical conductivity 
and interface reaction) are related to cation distribution. In spinel oxides, transition metals 
cations usually have multiple valences. Their site preferences, valence state and electrical 
conductivity are interrelated, so the cation distribution study can help to understand the 
degradation in electrical conduction and composition change after the interaction. In the work, 
the cystal structure and cation distribution in the (Mn,Co)
3
O
4
 based spinel solid solution at 
SOFC operation temperature is also studied. Finally, the effect of different transition metals 
doping in manganese cobalt spine coatings is presented on the interaction with chormia for 
the identification of potential coatings. 
In all, this work provide a fundamental understanding of the mass transport properties 
and phase stability of the (Mn,Co)
3
O
4 
coatings during high temperature operation, which will 
be beneficial for the development of protective spinel coatings.  
 
 35 
 
 
 
CHAPTER 2 
Interactions between (Mn,Co)
3
O
4
 coating materials and Cr
2
O
3
 scale 
 
2.1 Introduction 
As described in the first chapter, ceramic coating can provide long-time protection for 
metallic SOFC interconnects. As compared with perovskite oxides, spinel oxide coatings 
have shown better performance in reducing the contact resistance and suppressing Cr outward 
diffusion through the coating layer. However, a relatively thick scale of chromium-rich oxides 
grows between the spinel coating and chromium forming alloy after long-term operation in 
the SOFC environment. The interactions between the coating and chromia scale can result in 
the change in chemical composition and microstructure of the coating, which will then affect 
coating properties and cell performance. 
(Mn,Co)
3
O
4
 spinel oxide is currently considered to be the one of the most promising 
candidate coatings for SOFC interconnect.
[130, 138] 
The purpose of this part is to investigate the 
interaction between (Mn,Co)
3
O
4 
coating material and chromia scale at high temperature, and 
to understand the mass transport properties and phase stability of the (Mn,Co)
3
O
4 
coating. In 
this chapter, results on the mass transport behavior between manganese cobalt oxides and 
chromia are presented. 
 
 36 
2.2 Experimental procedure 
Manganese cobalt oxides (Mn,Co)
3
O
4 
were prepared by solid-state reaction method. 
MnO (99% pure, Alfa Aesar, Ward Hill, MA) and Co
3
O
4
 (99.83% pure, Fisher, Pittsburgh, PA) 
were weighed in appropriate ratio to form the desired stoichiometry. The powders were 
mixed by ball-milling for 48 h with zirconia milling medium in de-ionized water, and the 
slurry was dried in an oven at 80 
o
C overnight. The pre-mixed powders were die-pressed into 
pellets, and then sintered at 1200 
o
C in air to form spinel structure. 
The interaction between the prepared (Mn,Co)
3
O
4
 pellets and Cr
2
O
3
 was studied by 
placing them together, and heating in air at high temperature in the range of 800
 o
C -1200 
o
C. 
The sample configurations are shown in Fig. 2.1 (a). In each case, the side of (Mn,Co)
3
O
4
 
sample was in contact with Cr
2
O
3
 (labeled ?solid?), while the other side was exposed to air 
(labeled ?vapor?), so that vapor-phase transport from Cr
2
O
3
 to (Mn,Co)
3
O
4
 could occur. In 
the experiment, the location of the original interface was marked by applying platinum pastes 
on the (Mn,Co)
3
O
4
 sample surface prior to the reaction. After reaction, the microstructure and 
composition of the (Mn,Co)
3
O
4 
samples on the surface and cross sections were characterized 
using a JEOL JSM-7000F scanning electron microscope (SEM) equipped with 
energy-dispersive X-ray spectroscopy (EDS). The phase structures of the synthesized and 
reacted spinels were determined by x-ray powder diffraction (XRD). 
 
 
(a) 
Cr
2
O
3 
 
 
(Mn,Co)
3
O
4 
Solid 
Vapor 
 37 
 
(b) 
Fig. 2.1 Schematic diagrams of sample configurations. 
 
2.3 Results and discussions 
 
2.3.1 Surface morphology and structure 
The surface morphologies of as-prepared three different (Mn,Co)
3
O
4 
compositions, 
Mn
1.5
Co
1.5
O
4
, Mn
2
CoO
4
, and MnCo
2
O
4
 spinel oxides are shown in Fig. 2.2.
 [149]
 They were 
all sintered into a dense body with few isolated pores. The brighter stripe-like parts in Fig. 2.2 
(c) have high Co-rich content determined by EDS, which indicates MnCo
2
O
4
 contains two 
phases after sintered at 1200 
o
C. According to the Mn-Co-O phase diagram, at 1200 ?C 
Mn
2
CoO
4
 and Mn
1.5
Co
1.5
O
4
 are single-phase spinel solid solution, but MnCo
2
O
4
 have dual 
phases which consists of a cubic spinel phase (lower cobalt) and a rocksalt phase (higher 
cobalt).
 [138]
 
 
Mn
15
Co
15
O
4 
Cr
2
O
3 
 
(a) Mn
1.5
Co
1.5
O
4 
 38 
 
 
Fig. 2.2 SEM images of as-prepared three (Mn,Co)
3
O
4
 compositions surface: (a) 
Mn
1.5
Co
1.5
O
4
, (b) Mn
2
CoO
4
, and (c) MnCo
2
O
4
.
 [149]
 
 
   
(b) Mn
0.8
Co
1.3
Cr
0.9
O
4 
(b) Mn
2
CoO
4 
(c) MnCo
2
O
4 
(a) Mn
04
Co
0.6
Cr
2
O
4 
 39 
    
    
Fig. 2.3 SEM images of surfaces of (a-b) Mn
1.5
Co
1.5
O
4
, (c-d) Mn
2
CoO
4
, and (e-f) MnCo
2
O
4
 
after reaction at 1200 ?C for 72 h in contact with (left) or near chromia (right).
 [148]
 
 
The initial experiments were performed at temperatures higher than the SOFC operation 
temperature to overcome kinetic limitations. Fig. 2.3 show the SEM images of surface 
morphologies of Mn
1.5
Co
1.5
O,
 4
Mn
2
CoO
4
, and MnCo
2
O
4
 after reaction at 1200 ?C for 72 h in 
contact with or near chromia (namely, solid side and vapor side), respectively.
 [148]
 In spite of 
the differences in the surface morphology of as-prepared samples, their facetted surface 
morphologies for the regions in contact with Cr
2
O
3
 after reaction are similar. EDS analyses 
show that the reaction products are (Mn,Co)Cr
2
O
4
 spinels, in which the maximum chromium 
content reached to 2 per M
3
O
4
 unit. The upper side regions of the (Mn,Co)
3
O
4
 spinel oxides 
(a) Mn
0.65
Co
0.35
Cr
2
O
4 
(b) Mn
1.35
Co
0.85
Cr
0.85
O
4
 
(a) Mn
0.25
Co
0.75
Cr
2
O
4 
(b) Mn
1
Co
1.45
Cr
0.55
O
4 
 40 
that are not in direct contact with Cr
2
O
3 
also reacted via vapor phase transport, but their Cr 
content was relatively lower. It is shown that the faceted morphology did not form after 
reaction at 1200 
o
C and faster growth tends to take place at the grain boundary regions. 
The chemical compositions of the reaction products by both solid and vapor phase 
transport are also shown in Fig. 2.3. The Co/Mn ratio in the reaction surfaces was always 
higher than the original ratio in (Mn,Co)
3
O
4
 spinel oxides, and increased with the cobalt 
content in the as-prepared (Mn,Co)
3
O
4
 samples, which may be attributed to higher diffusion 
rate of Co within the reaction layer than that of Mn or the greater preference for tetrahedral 
site of Co
2+
 compared with that of Mn
3+
. 
[147]
 
 
28 32 36 40 44 48 52 56 60 64 68
2? (degree)
 0 hr
 72hr
 108hr
(a) Solid side
 
 41 
28 32 36 40 44 48 52 56 60 64 68
2? (degree)
 0 hr
(b) Vapor side
 72 hr
 108 hr
 
Fig. 2.4 XRD patterns of surfaces of Mn
1.5
Co
1.5
O
4
 after reaction at 1200 ?C (a) in contact 
with or (b) near chromia for different hours. 
 
   Fig. 2.4 plots XRD patterns of surfaces of Mn
1.5
Co
1.5
O
4
 after reaction at 1200 ?C (a) in 
contact with or (b) near chromia for 72 h and 108 h, respectively. The reaction layers 
produced by solid phase transport or vapor phase transport both have single cubic spinel 
structure, which is different from the dual-phase structure in Mn
1.5
Co
1.5
O
4
 at room 
temperature. XRD analyses also indicate chromium migration can stabilize the cubic 
structure. 
 
 42 
 
 
Fig. 2.5 Surface morphologies (a - d) and cation concentration profiles (e) of Mn
1.5
Co
1.5
O
4
 at 
different distances from Cr
2
O
3
 (light regions on the right) after reaction at 1200 ?C for  
144 h.
[148]
 
 
Fig. 2.5 shows the transition in surface microstructures (a-d) and corresponding cation 
concentration profiles (e) of Mn
1.5
Co
1.5
O
4
 at different distances from Cr
2
O
3
 after reaction at 
Boundary 
(b) (c) 
 
(d) 
(a) 
 
 43 
1200 ?C for 144 h using sample configurations in Fig. 2.1(b).
 [148]
 In this experiment, half of 
Mn
1.5
Co
1.5
O
4
 surface was covered with chromia shown as light regions on the right in Fig. 
2.7(a). The morphology becomes less faceted and the chromium concentration at the surface 
decreases as moving away from the boundary to left. The Cr
2
O
3
 activity in the vapor phase is 
different from Cr
2
O
3
 scale
 
boundary to the area far away, which would affect the morphology 
and chromium content in the reaction products. The transition in morphologies is consistent 
with the difference in the morphologies of interfaces that formed by solid phase transport and 
vapor phase transport as shown in Fig. 2.3. 
 
      
 
  
Fig. 2.6 SEM images of surfaces of Mn
1.5
Co
1.5
O
4
 in contact with or near Cr
2
O
3
 after reaction 
(a) 800
o
C solid 
(c) 1200
o
C solid 
(b) 800
o
C vapor 
(d) 1200
o
C vapor 
 44 
at 800
 o
C and 1200 
o
C for 144 h.
 [148]
 
 
Table 2.1 The chemical compositions and Co/Mn ratio in surfaces of Mn
1.5
Co
1.5
O
4
 in contact 
with or near Cr
2
O
3
 after reaction at 800
 o
C, 1200
 o
C for 144 h. 
Temperature 
chemical compositions Co/Mn ratio 
Powder side Vapor side Powder side Vapor side 
800 
o
C Mn
0.3
Co
0.8
Cr
1.9
O
4
 Mn
1.4
Co
1.4
Cr
0.2
O
4
 3.4 1.0 
1200 
o
C Mn
0.4
Co
0.6
Cr
2
O
4
 Mn
0.7
Co
1.3
Cr
1.0
O
4 
1.5 1.8 
 
Then the reaction temperature is lowered to the typical SOFC operation temperatures. 
Fig. 2.6 shows surface morphologies of Mn
1.5
Co
1.5
O
4
 in contact with or near Cr
2
O
3
 after 
reaction at 800
 o
C and 1200 
o
C for 144 h, and their chemical compositions and Co/Mn ratio 
of the reaction interface are summarized in Table 2.1.
 [148]  
For the solid side, a similar 
faceted microstructure still formed at 800
 o
C, but both the chromium concentration and grain 
size in the surface reaction layer decreased with decreasing the temperature. The Co/Mn 
ratios in the surface layer on both sides were always higher than the original ratio in 
Mn
1.5
Co
1.5
O
4
 spinel oxide even after reaction at low temperature 800
o
C. 
For the vapor side, chromium content was much lower than that in the side by solid 
phase transport. After reaction at 800 ?C, trace amount of chromium in the reaction layer 
indicated only a small amount of reaction took place via vapor phase transport. However 
unlike the morphology for the vapor side at 1200 ?C, a facet microstructure was generated as 
shown in Fig. 2.6 (b). The fact that the degree of faceting for lower Cr contents increases 
 45 
while reducing the reaction temperature suggests that this faceting mechanism is favored at 
lower temperatures (e.g. 800 ?C). 
 
2.3.2 Mass transport properties in Mn
1.5
Co
1.5
O
4
 
The cross sections of Mn
1.5
Co
1.5
O
4
 after reaction at 1000?C for 36 h in contact with 
Cr
2
O
3
 are shown in Fig. 2.7.
 [148]
 The lighter points in a line in Fig. 2.7 (a) are the platinum 
markers, which was used to locate the original surface of Mn
1.5
Co
1.5
O
4
 before reaction. A 
dense layer formed on its surfaces by solid phase transport. For the reaction layer formed by 
solid phase transport as shown in Fig. 2.7 (a), the Cr
2
O
3
 layer was on its right side. EDS 
analysis in Fig. 2.7 (b) showed it contains two scales between Mn
1.5
Co
1.5
O
4 
and Cr
2
O
3
: a 
dense (Mn,Co)Cr
2
O
4
 spinel outer scale and a Cr-containing (Mn,Co,Cr)
3
O
4 
intermediate scale. 
Throughout the outer reaction layer the chromium content remained relatively constant and 
its maximum value could be reached to 2 per M
3
O
4 
formula unit. However, in the 
(Mn,Co,Cr)
3
O
4
 intermediate layer the chromium content was observed to decrease 
dramatically. Although both reaction layers have the spinel structure, they grow by different 
mechanisms.
 
After reaction with Cr
2
O
3
, the fact that (Mn,Co)Cr
2
O
4
 spinel scale on the right 
of the platinum marker (which locate the original interface) indicates that this outer layer 
grows by the outward diffusion of manganese and cobalt from Mn
1.5
Co
1.5
O
4
 to the Cr
2
O
3
, 
while the (Mn,Co,Cr)
3
O
4
 intermediate layer grows by the inward diffusion of chromium 
through the outer layer.
[149] 
The growth mechanisms of reaction layer are illustrated 
schematically in Fig. 2.8. The thickness of Cr-rich reaction layer can also be confirmed by 
both the platinum markers and Cr concentration profile.  
 46 
 
 
 
40 35 30 25 20 15 10 5 0
0.0
0.5
1.0
1.5
2.0
 
 
Pt
Mn
1.5
Co
1.5
O
4
A
t
om
s
 per
 M
3
O
4
 uni
t
Distance from surface (mm)
 Cr
 Mn
 Co
(b)
 
 
Fig. 2.7 (a) SEM images and (b) cations profile of cross section of Mn
1.5
Co
1.5
O
4
 after reaction 
at 1000 ?C for 36 h in contact with Cr
2
O
3
.
 [148]
 
 
If the spinel coating is quite thin or not dense enough, the dissolved chromium would 
finally diffuse throughout the coating to the cathode and then result in chromium poisoning. 
Therefore, both of the reaction layers are helpful to retard chromium migration and 
volatilization and then to protect SOFC components. 
Pt 
(a) 
 47 
 
Fig. 2.8 Reaction layers formed during the reaction between Mn
1.5
Co
1.5
O
4
 and Cr
2
O
3
 at 
high temperature. 
 
2.4 Conclusions 
The interaction between manganese cobalt (Mn,Co)
3
O
4 
spinel oxide coating materials 
and Cr
2
O
3
 layer was studied in this chapter. The facetted surface morphologies for the regions 
in contact with Cr
2
O
3
 were observed after reaction at both 800 
o
C and 1200 
o
C. EDS and 
XRD analyses showed that the outer reaction layer are (Mn,Co)Cr
2
O
4
 spinel oxide with a 
cubic structure, in which the chromium content is reached to the maximum 2 per M
3
O
4
 unit at 
steady state (e.g. at 1200 
o
C).  
The reaction layer contains two scales, Cr-rich (Mn,Co)Cr
2
O
4
 outer scale and 
Cr-containing (Mn
,
Co
,
Cr)
3
O
4
 intermediate scale. They both have spinel structures, but their 
growth mechanisms are obviously different. In the outer layer in contact with Cr
2
O
3
, 
(Mn,Co)Cr
2
O
4
 grew by the outward diffusion of cobalt and manganese from (Mn
,
Co)
3
O
4
 
 
 
 
 
 
Cr
2
O
3 
scale
 
 
 
 
 
 
 
Mn
1.5
Co
1.5
O
4
 
coating 
Original surface
 
Mn
2+
,
 
Co
2+ 
 
Cr
3+ 
Mn
2+
, Mn
3+ 
Co
2+
, Co
3+ 
 
 
 
 
 
Cr-rich outer 
reaction layer 
(Mn,Co)Cr
2
O
4 
 
 
 
 
Cr-containing 
intermediate 
layer 
(Mn,Co,Cr)
3
O
4
 
 48 
coating material. At the same time, chromium dissolved into (Mn
,
Co)
3
O
4
 and formed the 
other (Mn
,
Co
,
Cr)
3
O
4
 intermediate layer. The growth of the Cr-rich reaction layer is helpful to 
minimize the migration of chromium species and protect the cathode from poisoning.  
 
. 
 49 
 
 
 
CHAPTER 3 
Electrical properties, cation distributions and thermal expansion of 
manganese cobalt chromite Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x = 0 - 2)
 
spinels 
 
3.1 Introduction 
    As discussed in the first chapter, chromia-forming alloys, especially ferritic stainless 
steels are widely used as SOFC metallic interconnect. However, their chromium composition 
also brings some problems. The volatile CrO
2
(OH)
2
 and CrO
3
 species deposit on the 
cathode/electrolyte interface, and cause degradation of cathode electrochemical 
performance.
[43,49]
 In addition, the electrical conductivity of Cr
2
O
3 
scale formed on the 
surface of stainless steels is lower that of alloys, which would increase the overall resistance 
of fuel cell.
 [150] 
Different conductive ceramic coatings have been applied on the surface of 
metallic interconnects to prevent chromium volatilization and increase oxidation resistance, 
and Mn
1.5
Co
1.5
O
4
 spinel oxide has been reported to be a promising coating candidate.  
To evaluate the long-term stability of Mn
1.5
Co
1.5
O
4
 coating, the transport properties of 
the reaction layer formed between Mn
1.5
Co
1.5
O
4
 and the Cr
2
O
3
 scale on the surface of 
chromium forming alloys need to be characterized, because the interaction can change the 
coating?s composition, and thus its properties, e.g. crystal structure, electrical conductivity 
and thermal expansion. The mass transport properties of the reaction layer have been 
 50 
discussed in the second chapter. The reaction layer between Mn
1.5
Co
1.5
O
4
 and Cr
2
O
3 
is 
detected to contain two regions in the second chapter, a chromium rich outer layer 
(Mn,Co)Cr
2
O
4
, and a chromium-containing intermediate layer (Mn,Co,Cr)
3
O
4
. Although both 
layers have the spinel structure, they form by different mechanisms. Another interesting 
phenomenon is that the Co/Mn ratio in the reaction layer was observed to be higher than 1, 
the original coating composition (i.e. Mn
1.5
Co
1.5
O
4
). Both the mass transport and ion 
transport (reaction layer growth and conductivity behavior) in the reaction layer are highly 
interrelated with how the cations are distributed on the octahedral and tetrahedral sites in 
spinel structure. Thus, determination of the atomic occupancies in manganese cobalt 
chromium spinel system, especially at the SOFC operation temperature 600 
o
C ? 800 
o
C, will 
be useful in understanding the growth and conduction mechanism of reaction layer. Although 
(Mn,Co,Cr)
3
O
4
 are the major compositions in the reaction layer, the information on their 
cation distributions is still lacking at high temperature under oxidizing atmosphere relevant to 
SOFC working environment. Since Mn, Co and Cr have similar numbers of electrons, the 
differences in x-ray scattering between them are small. However their neutron scattering 
lengths are quite different, 3.635 x10
-15
m for Cr, -3.73 x10
-15 
m for Mn, and 2.49 x10
-15
m for 
Co.
[151]
 So the in-situ neutron diffraction technique is employed here as an effective method 
to study cation distributions and other crystallographic parameters of (Mn,Co,Cr)
3
O
4
 in the 
temperature range of 600
 o
C - 800
o
C in air.  
To investigate the effect of Cr migration on the Mn
1.5
Co
1.5
O
4 
coating performance, 
manganese cobalt spinels with different chromium content, Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x = 0 - 
2), were synthesized. This work focuses on their electrical conductivity, cation distributions 
 51 
and thermal expansion at SOFC operation condition. The relationship of cation distributions 
with transport properties in the reaction layer was also discussed. 
 
3.2 Experimental 
Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x = 0 - 2) spinel oxides were synthesized by solid state 
reaction. MnO (99% pure, Alfa Aesar, Ward Hill, MA), Co
3
O
4
 (99.83% pure, Fisher, 
Pittsburgh, PA) and Cr
2
O
3
 (99% pure, Acros, Geel, Belgium) were weighed in appropriate 
ratio to form the desired stoichiometry. Then the powder mixtures were ball-milled for 48 h 
with zirconia milling medium in de-ionized water, and dried overnight in an oven. The 
pre-mixed powders were die-pressed, and then sintered at 1200 
o
C in air to form spinel 
structure. X-ray diffraction (XRD) was used for phase analysis by a Bruker D8 X-ray 
diffractometer. The samples were scanned over the ?2 range of 28
o
-70
o 
with the 
Cu
?
K radiation at room temperature. The in-situ high-temperature neutron diffraction 
measurements were performed on the high resolution neutron powder diffractometer (HB-2A) 
with a wavelength of 
o
A537.1=?  at Oak Ridge National Lab, Oak Ridge, USA. Since the 
vanadium could be easily oxidized in air at high temperature, a quartz sample holder was 
used due to its excellent chemical stability. Samples were placed in a quartz holder, and 
loaded into a tube furnace, which would be heated up to 800 
o
C under air atmosphere. The 
data were collected in the ?2  range of 8
o
 to 152
o
, and refined by the FULLPROF program 
using Rietveld method.
[152]
 The electrical conductivities of spinel samples were measured by 
four-probe d.c. method on rectangular bars with the dimension about 14 mm x 5 mm x 2 mm 
which were cut from sintered specimens. The measurements were conducted between 500
o
C 
 52 
and 900
o
C in random direction leaving enough time (10 h) to reach equilibration. Pt paste and 
wire were used as current collector and electrodes. A current chosen from 10-100 mA was 
applied during measurement. The coefficients of thermal expansion (CTEs) of the specimens 
were tested from a Unitherm
TM
 model 1161V high temperature vertical dilatometer 
(Lewisville, TX) from room temperature up to 1000 
o
C with a heating rate of 10 
o
C/min in air. 
The average CTE value was obtained from the difference of thermal expansion between 1000
 
o
C and room temperature divided by temperature difference. 
 
3.3 Results and Discussion 
 
3.3.1 X-ray diffraction 
The spinel structures include cubic phase (e.g. JCPDS No. 23-1237 for manganese cobalt 
oxide and No. 26-0474 for manganese cobalt chromium oxide) or tetragonal phase (e.g. 
JCPDS No. 18-0408 for manganese cobalt oxide).
 [153]
 Fig. 3.1 shows the XRD patterns of 
Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x= 0 - 2) spinel oxides at room temperature. Peaks positions 
indicated Mn
1.5
Co
1.5
O
4 
contains two phases, which is consistent with Yang et al?s 
results.
[130,138]
 The crystal structure of Mn
1.5
Co
1.5
O
4 
at room temperature is dual-phase 
including cubic Co-rich Mn
1+?
Co
2-?
O
4
 (Space group: m3Fd ) and tetragonal Mn-rich 
Mn
2-?
Co
1+?
O
4
 (Space group: I4
1
/amd), respectively.
 [130,154]
 As Cr content increased, the 
tetragonal phases become vanished. When x is greater than 1, XRD indicated 
Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 forms the single cubic phase. With increasing chromium doping, the 
cubic spinel structure can be stabilized at room temperature. 
 53 
 
Fig. 3.1 XRD patterns of Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x= 0 - 2) spinel oxides at room 
temperature. 
    
Table 3.1 lists the lattice parameter and different kinds of density of cubic 
Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x=1, 1.5, and 2). Their lattice parameter did not change much as 
chromium content increased. The lattice size of the cubic spinel structure is related to the 
ionic radius and site atomic occupancies. The different cations? radii are listed as following: 
R
 Cr
3+
(CN=VI) = 0.615
o
A , R
 Mn
3+
 (CN=VI) = 0.645
o
A , R
 Mn
2+
(CN=IV) = 0.66
o
A , R
 
Co
3+
(CN=VI) = 0.545
o
A , and R
 Co
2+
(CN=IV) = 0.58
o
A .
[155] 
It is shown that the size of Cr
3+
(VI) 
is intermediate between those of Mn and Co cations. The distribution occupancies of these 
three cations at the octahedral (M) and tetrahedral (T) sites would vary when Cr 
concentration increased. Both factors could affect the change in lattice parameters. Their 
measured density was obtained from the weight and geometric dimensions of samples, and 
the theoretical density was calculated from lattice parameters. A decrease in the relative 
density was shown as Cr content increased, which indicates that Cr-rich spinel oxides have a 
 54 
poor sinterability. 
 
Table 3.1 Lattice parameter and density of cubic Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x= 1, 1.5, and 2). 
 
Lattice 
parameter a(
o
A ) 
Theoretical 
density (g/cm
3
) 
Measured 
density (g/cm
3
) 
Relative 
density (%) 
MnCoCrO
4
 8.366 5.214 3.64 70 
Mn
0.75
Co
0.75
Cr
1.5
O
4
 8.379 5.130 3.18 62 
Mn
0.5
Co
0.5
Cr
2
O
4
 8.380 5.076 2.79 56 
 
3.3.2 Electrical conductivity  
Fig. 3.2 shows the conductivity of Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4 
(x=0 - 2) at 800 
o
C, the 
typical SOFC operation temperature. The conductivity decreases dramatically by several 
orders of magnitude as the Cr content increases. The conductivity at 800 
o
C drops from 60 
S/cm
 [130]
 to 0.007 S/cm by three orders of magnitude as x increases from 0 to 2 in 
Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
. The increase in porosity may contribute to the decrease in the 
conductivity. The temperature dependences of conductivity of Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4 
are 
shown in Fig. 3.3. The conductivity of all samples increased with the temperature. A linear 
dependence of )Tlog(?  with 1000/T is observed, which was also reported in other spinels, 
and suggests that the conduction in the spinel oxides occurs through the small polaron 
hopping mechanism on the octahedral site during high temperature as described by
[121, 156-158]
 
)exp(
0
Tk
E
T
b
a
?=
?
?                               (3.1) 
where, ?  is the electrical conductivity,  
 55 
T is the absolute temperature,  
0
?  is the pre-exponential factor,  
E
a
 is the activation energy,  
and k
b
 is the Boltzmann?s constant.  
 
The activation energies (E
a
) calculated from the slope of the fitting lines in Fig. 3.3 are 
shown in Fig. 3.4. The activation energy increases with the Cr content, which indicates Cr 
doping makes electrical conduction difficult in Mn
1.5
Co
1.5
O
4
. With small Cr content doping, 
the conductivity decreases with unchanged increase in activation energy (e.g. x lower than 
1.0), but the activation energy increases at higher Cr contents. This may indicate the 
mechanism does not change for low Cr content (i.e. a decrease number of available 
octahedral sites), but the change in activation energy at high Cr level doping may suggest 
other conduction process might be involved. 
 
Fig. 3.2 Electrical Conductivity of Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x = 0 - 2) spinel oxides at 800 
o
C 
in air. 
 56 
 
 
Fig. 3.3 Electrical Conductivity of Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x = 0 - 2) spinel oxides as a 
function of 1000/T in air. 
 
 
Fig. 3.4 Activation Energy as a function of Cr content in Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x = 0 - 2) 
spinel oxides. 
 
 57 
 
Fig. 3.5 Comparison between Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x =1 and 2) and Cr
2
O
3
 in electrical 
conductivity in air. 
 
Comparisons between Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x = 1 and 2) and Cr
2
O
3
 in electrical 
conductivity in air are shown in Fig. 3.4. The Cr
2
O
3
 scale formed on the stainless steel 
interconnect will also be part of the cell circuit. The conductivity of Cr
2
O
3
 at 800
o
C in air is 
in the range of 0.006 to 0.16 S/cm,
[44-48]
 which is comparable to that of Cr-rich spinel samples 
with x =1.5 and 2. The results indicate that the existence of both reaction layer and Cr
2
O
3
 
scale would increase the overall cell resistance, so it is important to suppress their further 
growth.  
As for the conductivity in the spinel oxides, Lu et al attributed the electrical conduction 
in the Mn
1+x
Cr
2-x
O
4
 system to the hopping of small polaron between Mn
3+
 and Mn
4+
 on the 
octahedral site.
 [77] 
So increasing Cr content will lead to a reduction in the concentrations of 
exchange pairs between Mn
3+
 and Mn
4+
 concentration and therefore the conductivity value. 
In the Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 ternary system, Co can also have multiple valence states, and 
 58 
all these three cations have quite different site preferences.
[155,159]
 The details about cation 
distributions and its relationship to conductivity will be discussed with the neutron diffraction 
data in the following section. 
 
3.3.3 Cation distribution in Mn
1.5
Co
1.5
O
4 
and MnCoCrO
4 
at high temperature 
The neutron diffraction patterns for Mn
1.5
Co
1.5
O
4 
and MnCoCrO
4 
at 600
 o
C and 800 
o
C 
under air atmosphere are given in Fig. 3.6 and Fig. 3.7, respectively. These two samples? 
detailed crystallographic information can help to understand the transport properties of 
interactions between Mn
1.5
Co
1.5
O
4
 coating and Cr
2
O
3
 scale during SOFC operation. 
 
 59 
 
Fig. 3.6 Neutron diffraction patterns of Mn
1.5
Co
1.5
O
4
 in air at (a) 600
 o
C and (b) 800 
o
C, 
respectively.  
 
 
 60 
 
Fig. 3.7 Neutron diffraction patterns of MnCoCrO
4 
at (a) 600
 o
C and (b) 800 
o
C, respectively.  
 
Table 3.2 Crystal parameters of (a) Mn
1.5
Co
1.5
O
4 
and (b) MnCoCrO
4 
at high temperature 
(a) Crystal parameters of Mn
1.5
Co
1.5
O
4
 (Space group: m3Fd  (227)) 
 600
o
C 800
o
C 
a (A) 8.4129(1) 8.4331(1) 
u oxygen 0.26164(9) 0.26176(9) 
[Mn]
T
 0.151(10) 0.209(9) 
[Co]
T
 0.849(10) 0.791(9) 
[Mn]
M
 0.689(7) 0.682(7) 
[Co]
M
 0.311(7) 0.318(7) 
R
wp 
(%) 5.27 5.07 
R
exp
 (%) 4.61 4.70 
 
 
 61 
 (b) Crystal parameters of MnCoCrO
4 
(Space group: m3Fd  (227)) 
 600
o
C 800
o
C 
a (A) 8.4140(1) 8.4301(1) 
u oxygen 0.26211(6) 0.26207(7) 
[Mn]
T
 0.252(9) 0.271(9) 
[Co]
T
 0.748(9) 0.729(9) 
[Mn]
M
 0.418(4) 0.398(5) 
[Co]
M
 0.082(4) 0.102(5) 
[Cr]
M
 0.500(0) 0.500(0) 
R
wp
 (%) 4.82 4.86 
R
exp
 (%) 4.35 4.35 
*
2/1
n,1i
2
ii
n,1i
2
ciiiwp
]yw/)yyw[(100R
??
==
?= , 
2/1
i
2
oiiexp
]yw/)pn[(100R
?
?= , where y
i
 and y
ci
 
are the observed and calculated numbers of counts at point i of the neutron diffraction pattern, 
w
i
 is the weigh of the observation = 1/variance(obs)
i
, R
exp
 is calculated supposing the best 
possible mode, and n-p is the number of degrees of freedom. 
 
The broad background spectrum was from the quartz sample holder, but a good contrast 
of diffraction peak intensity was still obtained. The refined crystal parameters of (a) 
Mn
1.5
Co
1.5
O
4 
and (b) MnCoCrO
4 
are listed in Table 3.2. Mn
1.5
Co
1.5
O
4
 is shown a single cubic 
phase (Space group: m3Fd ) from 600
 o
C to 800
o
C in Fig. 3.6, which is different from its 
dual-phase structure at room temperature. In the general spinel AB
2
O
4
, O anions form a cubic 
close-packing in which A and B cations occupy the tetrahedral (T) and octahedral (M) sites, 
 62 
and it can be expressed as (A
1-x
B
x
)[A
x/2
B
1-x/2
]
2
O
4
, where the parentheses represent cations on 
tetrahedral site, and square brackets represent cations on octahedral sites.
[117,119]
 Since 
Mn
1.5
Co
1.5
O
4
 contains only Mn and Co cations, the occupancies of these two at the same site 
are constrained to be one. Its refined crystallographic formula at 800
o
C is 
(Mn
0.209
Co
0.791
)[Mn
0.682
Co
0.318
]
2
O
4
. Its lattice size increased from 600
o
C to 800
o
C due to 
thermal expansion, but its site occupancies did not vary significantly. More Mn cations 
occupied the octahedral sites, and more Co cations occupied on the tetrahedral sites, although 
the Co/Mn chemical ratio is 1 in Mn
1.5
Co
1.5
O
4
. 
For MnCoCrO
4
, all cations? occupancies were first refined without any constraint, but 
the chemical formula obtained from the refinement is quite different from the nominal 
formula. For the ternary system, neutron diffraction is difficult to determine more than two 
cations? distributions in the same site (octahedral and tetrahedral) directly, so reasonable 
constraints have to be applied to the refinement. Navrotsky et al studied the thermodynamics 
of site preference for individual cations in the spinel structure, and found Cr
3+
 had the 
strongest octahedral site preference. 
[147]
 Cr
3+
 were also reported exclusively occupied the 
octahedral site in the Cr-containing spinels.
 [77,121,160]
 Based on these results, the octahedral 
site was constrained to be fully occupied with 50% occupation by Cr
3+
, and an agreement 
between the refined chemical formula with its nominal was achieved. The reliability of the 
refinement results were estimated from the agreement factors R
wp
 and R
exp
, which are listed 
in Table 3.2 (b). The crystal structure of MnCoCrO
4
 at high temperature is also a single cubic 
phase with space group m3Fd  (227), and its lattice parameter expanded as increasing the 
temperature. Similar to Mn
1.5
Co
1.5
O
4
, the occupancy of Co at tetrahedral site was much 
 63 
greater than that of Mn, which exhibited Co cations preferred the tetrahedral sites, while Mn 
cations like the octahedral sites. As mentioned above, the electrical conduction is related with 
cations on the octahedral sites. Both Mn and Co cations were found distributed in the 
octahedral sites in Mn
1.5
Co
1.5
O
4 
and MnCoCrO
4
, respectively. It was reported that Co could 
have multiple valences in spinel oxides, e.g. III (+3 valence state with low spin) and +2 
which was determined by thermal analysis.
[159]
 This indicates Co may also make 
contributions to the electrical conduction. In the unit cell of cubic spinel, there are four layers 
of octahedral chains, and each octahedral site is surrounded by six nearest octahedral 
neighbors. Using Mn on the octahedral sites as an example, the refined results at 800
o
C 
indicate that there are about 4 (6 x 0.682) nearest Mn cations on average for small polaron 
hopping on Mn cation at octahedral sites in Mn
1.5
Co
1.5
O
4
. In MnCoCrO
4
 the number of 
nearest Mn cations is decreased to around 2 (6 x 0.398). There is also a reduction in the 
available neighboring octahedral sites for hopping between Co cations after Cr doping. 
Therefore, as more Cr cations occupy the octahedral site with a fixed valence +3,
 [77,121,160]
 the 
contents of exchange pairs Mn
3+
/Mn
4+
 and Co
2+
/Co
III
 decrease, so the conductivity of 
Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 is reduced. In addition, the tetrahedral site preference of Co over Mn 
obtained from neutron diffraction data could somewhat explain the reason for higher Co/Mn 
ratio in the reaction layer between Mn
1.5
Co
1.5
O
4
 and Cr
2
O
3
 as investigated in the second 
chapter. Both Cr
3+
 and Mn
3+
 prefer octahedral sites during the interaction between 
Mn
1.5
Co
1.5
O
4
 coating and Cr
2
O
3
 scale, so Co
2+ 
tends to occupy the tetrahedral site. The 
neutron diffraction technique provides useful crystallographic information in understanding 
the transport mechanism in the reaction layer.  
 64 
 
3.3.4 Thermal expansion test 
   Thermal expansion compatibility is required to reduce the thermal stress and avoid 
component spallation for good long-term performance. Fig. 3.8 shows the CTEs of 
Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x = 0 - 2) spinel oxides from room temperature to 1000
o
C in air. The 
CTE of Cr
2
O
3
 from 25
o
C - 900
o
C, 9.6 x10
-6
/
o
C is shown for comparison.
[151] 
As Cr content 
increased, CTE of Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 spinel oxides decreased, which indicates Cr 
diffusion from Cr
2
O
3
 layer can reduce the thermal expansion of Mn
1.5
Co
1.5
O
4
 coating. Thus, 
the growth of the reaction layer formed between Mn
1.5
Co
1.5
O
4
 coating and Cr
2
O
3
 scale on the 
surface of metallic interconnects could lead to thermal stresses from the CTE mismatch and 
local spallation after long-time operation.  
 
Fig. 3.8 CTEs of Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x=0-2) spinel oxides and Cr
2
O
3
[151]
 from room 
temperature up to 1000
o
C in air. 
 
 
 65 
3.4 Conclusions 
Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x = 0 - 2) spinel oxides were synthesized by solid state reaction. 
The effects of Cr doping were studied on the crystal structure, electrical conductivity and 
thermal expansion at SOFC operation condition, and the relationship between cation 
distributions and transport properties in the reaction layer were discussed. With increasing Cr 
content, the cubic phase at room temperature is stabilized, the electrical conductivity 
decreases dramatically by several orders of magnitude, and CTE also decreases. The 
temperature dependence of conductivity suggested a small polaron hopping conduction 
mechanism in Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
. The refinements of neutron diffraction data for 
Mn
1.5
Co
1.5
O
4
 and MnCoCrO
4
 show they both were cubic structure at SOFC operation 
temperature, and indicate Mn prefers the octahedral sites, while Co prefers the tetrahedral 
sites, which is consistent with the Co/Mn ratio in the reaction layer keeping higher than that 
in the original coating. The formation of a Cr-containing reaction layer between the oxidation 
scale and the (Mn,Co)
3
O
4
 spinel coating could lead to an increase in electrical resistance and 
thermal stresses. Based on this work, reducing the thickness of Cr-containing reaction layer 
could improve the overall electrical conduction of ceramic coating.  
 
 66 
 
 
 
CHAPTER 4 
Electrical properties of transition metal (TM)-doped (Mn,Co)
3
O
4
 (TM = Fe, 
Ti) spinels and their interaction with chromia for SOFC interconnect 
coatings 
 
4.1 Introduction 
Manganese cobalt spinel oxides (Mn,Co)
3
O
4 
are promising coating candidates which 
can be applied on SOFC metallic interconnect ferritic stainless steels. However, the CTE 
value of MnCo
2
O
4 
is larger than that of ferritic stainless steels. In addition, as discussed in the 
third chapter, the formation of the Cr-containing reaction layer between the Cr
2
O
3
 scale and 
the (Mn,Co)
3
O
4
 spinel coating could lead to an increase in the electrical resistance and 
thermal stresses due to its low electrical conductivity and CTE mismatch. Therefore, 
adjusting CTE of coating material and depressing the growth of its reaction layer could 
reduce its effect on the overall cell performance, and cation doping in the coating material is 
considered. In this work, the high temperature electrical properties of transition metal 
(TM)-doped MnCo
2-x
TM
x
O
4 
(TM = Fe and Ti) and their interaction with chromia were 
investigated as potential SOFC interconnect coatings. 
 
 
 67 
4.2 Experimental 
MnCo
2-x
TM
x
O
4 
(TM = Fe, x = 0 - 0.7; TM = Ti, x = 0.34) spinel oxides spinel oxides 
were synthesized by solid state reaction. MnO (99% pure, Alfa Aesar, Ward Hill, MA), Co
3
O
4
 
(99.83% pure, Fisher, Pittsburgh, PA) and Fe
3
O
4
 (99.99% pure, Alfa Aesar, Ward Hill, MA) 
or TiO
2
 (99.17% pure, Fisher, Pittsburgh, PA) were weighed in appropriate ratio to form the 
desired stoichiometry. Then the powder mixtures were ball-milled for 48 h with zirconia 
milling medium in de-ionized water, and the slurry was dried overnight in an oven. The 
pre-mixed powders were die-pressed, and then sintered at 1200
o
C in air to form spinel 
structure.  
X-ray diffraction (XRD) was used for phase analysis by a Bruker D8 X-ray 
diffractometer. The samples were scanned over the ?2 range of 28-68
o 
with the 
Cu
?
K radiation at room temperature. The in-situ high-temperature neutron diffraction 
measurements were performed on the high resolution neutron powder diffractometer (HB-2A) 
with a wavelength of 
o
A539.1=?  at Oak Ridge National Lab, Oak Ridge, USA. A quartz 
sample holder was used due to its excellent chemical stability. Samples were placed in a 
quartz holder, and loaded into a tube furnace, which would be heated in air atmosphere. The 
data were collected in the ?2  range of 8
o
 to 152
o
, and refined by the FULLPROF program 
using Rietveld method.
[153]
 The electrical conductivity of spinel samples were measured by 
four-probe d.c. method on rectangular bars with the dimension about 14 mm x 5 mm x 2 mm 
which were cut from sintered specimens. The measurements were conducted between 500
o
C 
and 900
o
C in random order leaving 10 h to reach equilibration. Pt paste and wire were used as 
current collector and electrodes. A current chosen from 300mA -500mA was applied during 
 68 
measurement. The coefficients of thermal expansion (CTEs) of the specimens were tested 
from a Unitherm
TM
 model 1161V high temperature vertical dilatometer (Lewisville, TX) 
from room temperature up to 1000
o
C with a heating rate of 10 
o
C/min in air. The average 
CTE value was obtained from the difference of thermal expansion between 1000
 o
C and room 
temperature divided by temperature difference. 
The interaction between the prepared MnCo
2-x
TM
x
O
4
 pellets and Cr
2
O
3
 was performed 
by placing the pellets in contact, and heating in air at high temperature at 800 
o
C and 900
o
C 
using the sample configurations as shown in Fig. 2.1 (a). After reaction, the microstructure 
and composition of the MnCo
2-x
TM
x
O
4 
samples on the surface and cross sections were 
characterized using a JEOL JSM-7000F scanning electron microscope (SEM) equipped with 
energy-dispersive X-ray spectrum (EDS).  
 
4.3 Results and Discussions 
 
4.3.1 X-ray diffraction  
Fig. 4.1 shows the XRD patterns of MnCo
2-x
Fe
x
O
4
 (x= 0 - 0.7) spinel oxides at room 
temperature. Peaks positions indicated all the MnCo
2-x
Fe
x
O
4
 (x= 0.15-0.7) spinel samples 
have single cubic structure with space group m3Fd  after sintering. It is shown that as the 
iron content increases, the diffraction peaks shift left to the lower angle range. According to 
the Bragg?s Law, ?? =?sind2 , a smaller ?  will lead to a larger d-spacing, and therefore a 
larger lattice parameter. So the lattice size increases as increasing the iron concentration in 
MnCo
2-x
Fe
x
O
4
. The lattice parameter (a) of MnCo
2-x
Fe
x
O
4 
unit cell is listed as follows: 
 69 
8.285
o
A , 8.318
o
A , 8.341
o
A , 8.372
o
A , and 8.398
o
A as x increases from 0 to 0.7. The ionic 
radius could affect the lattice size of this cubic spinel structure. The radii of manganese, 
cobalt, and iron cations are in the following: R
 Mn
3+
 (CN=VI) = 0.645
o
A , R
 Mn
2+
(IV) = 0.66
o
A , 
R
 Co
3+
(VI) = 0.545
o
A , and R
 Co
2+
(IV) = 0.58
o
A , R
 Fe
3+
(CN=VI) = 0.645
o
A , R
 Fe
2+
(IV) = 
0.63
o
A .
[155] 
It is apparent that the size of Fe
3+
(VI) or Fe
2+
(IV) is bigger than those of Co
3+
(VI) 
or Co
2+
(IV) cations. In MnCo
2-x
Fe
x
O
4
 spinel oxides, the Mn content is fixed to 1. So 
increasing Fe content, the Co concentrate will be decreased, and thus the lattice size will 
expand.  
 
 
Fig. 4.1 XRD patterns of MnCo
2-x
Fe
x
O
4
 (x= 0 - 0.7) spinel oxides at room temperature. 
 
 
4.3.2 Neutron diffraction 
The neutron diffraction patterns for MnCo
1.66
Fe
0.34
O
4 
and Mn
0.67
Co
1.22
Fe
0.22
CrO
4 
at 
 70 
600
o
C under air atmosphere are shown in Fig. 4.2 and Fig. 4.3, respectively. The broad 
background spectra are due to the quartz sample holder.  
 
 
Fig. 4.2 Neutron diffraction patterns of MnCo
1.66
Fe
0.34
O
4
 in air at 600
 o
C.  
 
 
Fig. 4.3 Neutron diffraction patterns of Mn
0.67
Co
1.22
Fe
0.22
CrO
4 
at 600
 o
C in air.  
 
 71 
Table 4.1 Crystal parameters of MnCo
1.66
Fe
0.34
O
4 
and Mn
0.67
Co
1.22
Fe
0.22
CrO
4 
at 600
 o
C.  
(a) 50% Fe on tetrahedral and octahedral sites, respectively. 
 MnCo
1.66
Fe
0.34
O
4
 Mn
0.67
Co
1.22
Fe
0.22
CrO
4
 
Space group 
m3Fd  m3Fd  
a (A) 8.39329(12) 8.40503(9) 
[Mn]
T
 0.200(16) 0.206(9) 
[Co]
T
 0.630(16) 0.684(9) 
[Fe]
T
 0.17(0) 0.11(0) 
[Mn]
M
 0.445(5) 0.265(5) 
[Co]
M
 0.470(5) 0.180(5) 
[Fe]
M
 0.085(0) 0.055(0) 
[Cr]
M
 -- 0.5(0) 
R
wp 
(%) 3.71 3.57 
R
exp
 (%) 3.52 3.02 
 
 
 
 
 
 
 
 
 72 
(b) 70% Fe on tetrahedral and 30% Fe on octahedral sites, respectively. 
 MnCo
1.66
Fe
0.34
O
4
 Mn
0.67
Co
1.22
Fe
0.22
CrO
4
 
Space group 
m3Fd  m3Fd  
a (A) 8.39329(12) 8.40504(9) 
[Mn]
T
 0.276(16) 0.256(9) 
[Co]
T
 0.486(16) 0.590(9) 
[Fe]
T
 0.238(0) 0.154(0) 
[Mn]
M
 0.407(5) 0.241(5) 
[Co]
M
 0.542(5) 0.226(5) 
[Fe]
M
 0.051(0) 0.033(0) 
[Cr]
M
 -- 0.5(0) 
R
wp 
(%) 3.70 3.56 
R
exp
 (%) 3.52 3.25 
 
 
 
 
 
 
 
 
 
 73 
(c) 30% Fe on tetrahedral and 70% Fe on octahedral sites, respectively. 
 MnCo
1.66
Fe
0.34
O
4
 Mn
0.67
Co
1.22
Fe
0.22
CrO
4
 
Space group 
m3Fd  m3Fd  
a (A) 8.39328(12) 8.40503(9) 
[Mn]
T
 0.124(16) 0.157(9) 
[Co]
T
 0.774(16) 0.777(9) 
[Fe]
T
 0.102(0) 0.066(0) 
[Mn]
M
 0.483(5) 0.289(5) 
[Co]
M
 0.398(5) 0.134(5) 
[Fe]
M
 0.119(0) 0.077(0) 
[Cr]
M
 -- 0.5(0) 
R
wp 
(%) 3.70 3.57 
R
exp
 (%) 3.52 3.25 
 
For these two samples, all the cations? occupancies were first refined without any 
constraints, but the formula from the refinement is far away from the nominal formula. As 
discussed in the third chapter, for the ternary or even quaternary spinel system, neutron 
diffraction is difficult to determine more than two cations? distributions in the same site 
(octahedral and tetrahedral) directly, so constraints are required for the refinement. According 
to Navrotsky et al?s work, both Fe
3+
 and Fe
2+
 do not have strong preference to either 
tetrahedral or octahedral site.
 [148]
 So it is reasonable to assume the Fe occupancy on both 
tetrahedral and the octahedral sites. First, half of the Fe content is assumed on tetrahedral site 
 74 
and half on the octahedral site. The spectra indicate MnCo
1.66
Fe
0.34
O
4
 has a single cubic 
phase (Space group: m3Fd ) at 600
 o
C in Fig. 4.2, which is consistent with its crystal structure 
at room temperature. Its refined crystallographic formula is 
(Mn
0.200
Co
0.630
Fe
0.17
)[Mn
0.442
Co
0.473
Fe
0.085
]
2
O
4
, where the parentheses represent cations on 
tetrahedral site, and square brackets represent cations on octahedral sites. It is shown that 
more Mn cations prefer to occupy the octahedral sites. Then, different ratio of Fe on both 
sites are constrained (e.g., 70% of the Fe content on tetrahedral site and 30% on the 
octahedral site, 30% of the Fe content on tetrahedral site and 70% on the octahedral site). 
Their refined results are shown in Table 4.1 (b) and (c), respectively. In the three cases, they 
all have similar residuals, which may suggest the three cation distributions are possible. 
Because Fe
 
cations do not have preference to tetrahedral or octahedral site,
 [148]
 it is more 
safely tentative to constrain half of the Fe content on tetrahedral site and half on the 
octahedral site in Table 4.1 (a). 
 For Mn
0.67
Co
1.22
Fe
0.22
CrO
4
, besides the application of above constraint on iron?s 
distribution, the octahedral site was constrained to be fully occupied with 50% occupation by 
Cr
3+
 due to its strongest octahedral-site occupation.
[77,122,148,154,160]
 Mn
0.67
Co
1.22
Fe
0.22
CrO
4
 also 
shows a cubic phase at 600
 o
C with space group m3Fd (227). The refined crystal parameters 
of Mn
0.67
Co
1.22
Fe
0.22
CrO
4 
with different Fe occupancy ratio assumptions are listed in Table 
4.1. For the three conditions, the fitting results have similar residuals, and the occupancy of 
Mn in octahedral sites was higher than that in tetrahedral sites. Similar to MnCo
1.66
Fe
0.34
O
4
, it 
is also more safe and reasonable to constrain half of the Fe content on tetrahedral site and half 
on the octahedral site in Mn
0.67
Co
1.22
Fe
0.22
CrO
4
. 
 75 
 
4.3.3 Interaction with Chromia 
 
4.3.3.1 Surface Morphology  
The interactions with chromia were performed to simulate the SOFC working condition 
for 420 h. Fig. 4.4 (a) shows the surface of the as-prepared MnCo
1.5
Fe
0.5
O
4
 pellet. The grains 
were densely packing with little isolated pores. After reaction with chromia at 800
o
C, a dense 
faceted morphology was formed in the region in contact with Cr
2
O
3 
shown in Fig. 4.4 (b). 
EDS analysis indicate the chemical composition of this outer reaction scale is 
Mn
0.1
Co
0.9
Cr
2.0
O
4
 spinel oxide. The surface morphologies of other MnCo
2-x
Fe
x
O
4
 (x= 0.15, 
0.34 and 0.7) samples after reaction at 800
o
C in contact with Cr
2
O
3
 for 420 h are shown in 
Fig. 4.5. MnCo
2-x
Fe
x
O
4 
spinels (x= 0.15, 0.34 and 0.5) have similar faceted morphologies in 
the region in contact with Cr
2
O
3
, but the surface morphology of MnCo
1.3
Fe
0.7
O
4
 after reaction 
is a little different, which could be related to its growth of the reaction layer. 
Table 4.2 also lists the corresponding chemical composition and Co/Mn ratio of 
MnCo
2-x
Fe
x
O
4
 (x= 0.15 - 0.7) after reaction at 800
o
C. It is interesting to note that even Fe 
content is increased to 0.7 in MnCo
1.3
Fe
0.7
O
4
, no iron is observed in the interface in contact 
with Cr
2
O
3
, with the composition of Mn
0.2
Co
0.8
Cr
2.0
O
4
, which suggested iron may not prefer 
the tetrahedral site as compared with manganese and cobalt during the interaction with Cr
2
O
3
. 
In addition, Co/Mn ratio of the outer reaction layer is decreased as that decreased in the 
samples before reaction, but is still much higher than that in the original MnCo
2-x
Fe
x
O
4
 spinel 
oxides.  
 76 
 
   
Fig. 4.4 SEM images of surface of MnCo
1.5
Fe
0.5
O
4
 before (a) and after (b) reaction at 800 
o
C 
in contact with Cr
2
O
3
 for 420 h.  
 
 
 
 
 
(a) 0.15 
(b) 0.34 
(a) As-prepared (b)Contact with Cr
2
O
3
 
 77 
 
Fig. 4.5 SEM images of surface morphologies of MnCo
2-x
Fe
x
O
4
 (x= 0.15, 0.34 and 0.7) after 
reaction at 800
o
C in contact with Cr
2
O
3
 for 420 h.  
 
Table 4.2 Chemical composition and Co/Mn ratio of MnCo
2-x
Fe
x
O
4
 (x= 0.15 - 0.7) after 
reaction at 800 
o
C in contact with Cr
2
O
3
 for 420 h. 
Chemical Composition Co/Mn ratio 
As-prepared After reaction Before         After 
MnCo
1.85
Fe
0.15
O
4
 Mn
0.1
Co
0.9
Cr
2.0
O
4
 1.85 9 
MnCo
1.66
Fe
0.34
O
4
 Mn
0.1
Co
0.9
Cr
2.0
O
4
 1.66 9 
MnCo
1.50
Fe
0.50
O
4
 Mn
0.1
Co
0.9
Cr
2.0
O
4
 1.50 9 
MnCo
1.30
Fe
0.70
O
4
 Mn
0.2
Co
0.8
Cr
2.0
O
4
 1.30 4 
 
4.3.2.2 Concentration Gradients 
Fig. 4.6 shows the Cr concentration gradient of MnCo
2-x
Fe
x
O
4 
after reaction with Cr
2
O
3 
at 800
o
C for 420 h. The cation concentration gradients were measured at locations in different 
images of each sample. The Cr profile exhibited two layer forms in the reaction region, which 
is consistent with the earlier work in the second chapter. The outer Cr-rich spinel layer 
(c) 0.70 
 78 
contacted with chromia grew by diffusion of cobalt and manganese from MnCo
2-x
Fe
x
O
4
. At 
the same time, chromium dissolved into MnCo
2-x
Fe
x
O
4
 and formed the inner layer with Cr 
content decreased to 0. As for MnCo
2-x
Fe
x
O
4
 samples, when the Fe content is from 0.15 to 
0.5, the thickness of outer reaction layer did not change much. But for MnCo
1.3
Fe
0.7
O
4
, its 
thickness increased. Fig. 4.7 shows the dopant (Fe) concentration gradient
 
after 
high-temperature reaction. The thickness of outer reaction layer can also be reflected from Fe 
profile, because Fe did not appear throughout this scale in all MnCo
2-x
Fe
x
O
4
 spinel oxides. 
The composition differences after reaction are probably attributed to the cation occupancy 
preference. In the Cr-rich outer layer, Cr
3+
 has the strongest preference for the octahedral site 
in the spinel structure, so during reaction Cr pushed Co and Mn to the tetrahedral site. Fig. 
4.8 and Fig. 4.9 are the Co and Mn content profiles after reaction with Cr
2
O
3
, respectively. It 
was observed that the Co/Mn ratio in the outer reaction layer is much higher than that in the 
original MnCo
2-x
Fe
x
O
4
 spinels, which could be due to higher diffusion rate of cobalt in the 
reaction layer than that of manganese or greater preference of cobalt cation for tetrahedral site 
than that of manganese cation which is also confirmed in the third chapter. Compared with 
Co and Mn, Fe seems not prefer the tetrahedral site, which may suggest why Fe was not 
detected in this region.  
 
 79 
 
Fig. 4.6 Cr concentration profile in MnCo
2-x
Fe
x
O
4
 after reaction for 420 h in contact with 
Cr
2
O
3
 at 800
o
C.  
 
 
Fig. 4.7 Fe concentration profile in MnCo
2-x
Fe
x
O
4
 after reaction for 420 h in contact with 
Cr
2
O
3
 at 800
o
C. 
 
 80 
 
Fig. 4.8 Co concentration profile in MnCo
2-x
Fe
x
O
4
 after reaction for 420 h in contact with 
Cr
2
O
3
 at 800
o
C. 
 
 
Fig. 4.9 Mn concentration profile in MnCo
2-x
Fe
x
O
4
 after reaction for 420 h in contact with 
Cr
2
O
3
 at 800
o
C. 
 
 
 81 
4.3.4 Electrical conductivity 
The electrical conductivity is an important property for the interconnect coating. Fig. 
4.10 shows the conductivity data of MnCo
2-x
Fe
x
O
4
 at 800 
o
C in air. The conductivity 
increased when Fe was doped as 0.15 content. Then as Fe content was increased further, the 
conductivity at high temperature decreased. But still MnCo
1.66
Fe
0.34
O
4
 had comparable 
electrical conductivity with Mn
1.5
Co
1.5
O
4
, close to 60 S/cm at 800
o
C in air 
[131]
.
 
Similar to 
other spinel oxides, the conductivities of MnCo
2-x
Fe
x
O
4 
increased with temperature in a linear 
relationship between )log( T? and 1000/T as shown in Fig. 11, and suggests their conduction 
belongs to the small polaron hopping mechanism.
 [122, 156-158]
 Their activation energy (E
a
) was 
calculated from the slope of the fitting line in Fig. 4.11, and it does not vary significantly, in 
the range of 0.54-0.58 eV. Although the conductivity of MnCo
2-x
Fe
x
O
4
 decreased further, it 
was still much higher than that of (Mn,Co,Cr)
3
O
4
, shown in Fig. 4.8. So considering both the 
reactivity with chromia and conductivity properties, the reasonable Fe doping content would 
be controlled less than 0.5.  
 
 82 
 
Fig. 4.10 Electrical conductivities of MnCo
2-x
Fe
x
O
4 
at 800 
o
C in air. 
 
 
Fig. 4.11 Electrical conductivities of MnCo
2-x
Fe
x
O
4 
and
 
MnCoCrO
4 
spinels as a function of 
temperature. 
 
4.3.4 Comparison in the effect of Fe and Ti doping 
The effect of iron and titanium doping on the reaction between the manganese cobalt 
 83 
spinel oxide and Cr
2
O
3
 are illustrated in Fig. 4.12.
 [148]
 Similar to the results of Mn
1.5
Co
1.5
O
4
, 
the Cr concentration profile indicated a two-layer interaction region formed between 
MnCo
1.66
TM
0.34
O
4
 (TM=Fe and Ti) and Cr
2
O
3
, a Cr-rich outer layer and a Cr-containing 
intermediate layer. However, the thickness of the Cr-rich outer layer is reduced significantly 
after doping with iron and titanium. The Cr-rich layer formed in Mn
1.5
Co
1.5
O
4
 was around 2 
times as thick as that in MnCo
1.66
TM
0.34
O
4
. The reason for the decreased extent of the 
reaction in MnCo
1.66
TM
0.34
O
4
 is not fully understood yet, but would be related to the change 
in the transport properties of manganese cobalt spinel oxide doped with Fe and Ti or 
(octahedral/tetrahedral) site preferences. Even for the MnCo
1.66
Ti
0.34
O
4
 oxide, its Co/Mn ratio 
at the interface after reaction was still higher than that in the original spinel, which is 
consistent with Fe doped or un-doped manganese cobalt spinel oxides. Fig. 4.13 shows Fe 
and Ti concentration profile in MnCo
1.66
TM
0.34
O
4
 after reaction in contact with Cr
2
O
3
 for 72 
h at 1000 
o
C.
 [148]
 It should be noted that similar to Fe, no Ti was detected by EDS in the 
Cr-rich layer, which could be attributed to the site (octahedral/tetrahedral) preferences. Ti is 
reported to prefer the octahedral sites in the spinel oxides.
[147]
 So during the formation of the 
Cr-rich outer layer, when Cr
3+
 occupies the octahedral site, Ti cations would not be as 
competent as Mn
2+
 and Co
2+
 for the occupation in tetrahedral site. 
 
 84 
 
Fig. 4.12 Chromium concentration profile in MnCo
1.66
TM
0.34
O
4
 (TM=Fe and Ti) and 
Mn
1.5
Co
1.5
O
4
 after reaction in contact with Cr
2
O
3
 for 72 h at 1000
o
C.
[148]
 
 
 
Fig. 4.13 Dopant (Fe and Ti) concentration profile in MnCo
1.66
TM
0.34
O
4
 (TM=Fe and Ti) 
after reaction in contact with Cr
2
O
3
 for 72 h at 1000
o
C.
 [148]
 
 
 
 85 
 
Fig. 4.14 Comparison in electrical conductivities of MnCo
1.66
TM
0.34
O
4
 (TM=Fe and Ti) 
spinel oxides. 
 
The electrical conductivities of MnCo
1.66
TM
0.34
O
4
 (TM=Fe and Ti) spinel oxides as a 
function of temperature are plotted in Fig. 4.14. It is shown that the electrical conductivity of 
MnCo
1.66
Ti
0.34
O
4, 
about 25 S/cm at 800
o
C, is lower than that of MnCo
1.66
Fe
0.34
O
4
 in air. 
Table 4.3 lists the CTE values of MnCo
1.66
TM
0.34
O
4
 (TM=Fe and Ti) and (Mn,Co)
3
O
4
 
spinels.
[149]
 MnCo
2
O
4
 had a much higher CTE than most of the stainless steels interconnect 
(11.5-14 x10
-6
/
 o
C), and mismatch in thermal expansion may result in local spallation and 
crack. After substitution with iron and titanium, CTEs were reduced to compatible values, 
about 11.2?10
-6
/?C for MnCo
1.66
Ti
0.34
O
4
 and 12.0?10
-6
/?C for MnCo
1.66
Fe
0.34
O
4
. Higher 
cobalt can lead to a higher CTE, but both Fe and Ti can adjust CTE value of manganese 
cobalt spinel oxide to match that of stainless steels interconnect. 
 
 
 86 
Table 4.3 CTEs of MnCo
1.66
TM
0.34
O
4
 (TM=Fe and Ti) and (Mn,Co)
3
O
4
 spinel oxides.
[149]
 
Composition CTE (room temperature-1000
o
C, x10
-6
/
 o
C) in air 
Mn
1.5
Co
1.5
O
4
 10.8 
MnCo
2
O
4
 14.1 
MnCo
1.66
Fe
0.34
O
4
 12.0 
MnCo
1.66
Ti
0.34
O
4
 11.2 
 
4.4 Conclusions 
MnCo
2-x
TM
x
O
4 
(TM = Fe, x = 0-0.7) spinel oxides were synthesized to investigate the 
effects of Fe doping on the reaction with chromia and electrical property at high temperature. 
When Fe was doped up to 0.5, the thickness of the reaction layer did not increase. Although 
Fe dopant is increased to 0.7, it did not appear in the Cr-rich reaction layer, which would be 
related to that Fe may not prefer the tetrahedral site as Co and Mn cations during the reaction. 
As Fe content increased from x = 0.15, the conductivity decreased, but is still high enough. 
The electrical conductivity of MnCo
1.66
Fe
0.34
O
4 
at 800
o
C was quite close to that of 
Mn
1.5
Co
1.5
O
4
, the most promising coating candidate.  
The comparison between iron and titanium doping on the performance of 
MnCo
1.66
TM
0.34
O
4
 (TM=Fe and Ti) coatings was also performed. With the substitution of 
iron and titanium, the thicknesses of reaction layer formed between MnCo
1.66
TM
0.34
O
4 
and 
Cr
2
O
3
 were obviously reduced as compared with Mn
1.5
Co
1.5
O
4
,
 
which could improve the 
adherence of the coating on the alloy and the long-term stability of the SOFC interconnect. 
The thermal expansion was also well adjusted after doping with Fe or Ti. But neither Fe nor 
 87 
Ti were detected in the outer reaction layer. The reason for the decreased growth in the 
reaction layer after doping with Fe and Ti is not clear yet, but could be related to the change 
in the transport properties of MnCo
1.66
TM
0.34
O
4
 and/or site (octahedral/tetrahedral) 
preferences in the spinel. The electrical conductivity of MnCo
1.66
Ti
0.34
O
4
 is lower than that of 
MnCo
1.66
Fe
0.34
O
4,
 but still high enough for the spinel coating. Based on the high temperature 
properties (conductivity and reaction with Cr
2
O
3
), MnCo
2-x
Fe
x
O
4
 with iron content lower 
than 0.5 would be the potential composition as SOFC interconnect coating. 
 88 
 
 
 
CHAPTER 5 
Interaction of transition metal-doped (Mn,Co)
3
O
4
 (TM = Ni, Cu) spinels 
with chromia for SOFC interconnect coatings 
 
5.1 Introduction 
Manganese cobalt spinel oxides (Mn,Co)
3
O
4 
are promising coating candidates which 
can be applied on SOFC metallic interconnect ferritic stainless steels. As discussed in the 
forth chapter, doping iron and titanium in (Mn,Co)
3
O
4 
can depress the thickness and growth 
of the outer reaction layer in (Mn,Co)
3
O
4
, and the Cr concentrate gradient in the intermediate 
layer became steep. However no iron and titanium dopants were detected in the Cr-rich spinel 
area, which could be related to the cation site preference in spinel structure. It was reported 
that chromite spinel oxides (ACr
2
O
4
) are normal spinel structure, in which Cr
3+
 occupy all the 
octahedral holes, and A-site cations occupy the tetrahedral holes.
 [77,122,154,160]
 Since titanium 
has large octahedral site preference, but not as strong as chromium in the spinel oxide, it 
would explain why in the interaction with chroma, titanium cation is reluctant to occupy the 
octahedral sites in the Cr-rich reaction layer. Ni
2+
 also exhibits high preference in octahedral 
sites in many spinel oxides.
[148,161,162]
 Then the effect of Ni doping in Mn
1.5
Co
1.5
O
4
 on the 
reaction behavior with chromia is studied here. Wei et al reported copper has a much greater 
tendency to occupy the tetrahedral sites in Cu
x
Mn
3-x
O
4
 spinel oxides.
[95]
 So copper is also 
 89 
considered as a dopant as compared with nickel. In this work, the interaction with chromia of 
transition metal (TM)-doped (Mn,Co)
3
O
4 
(TM = Ni, Cu) at high temperature were 
investigated as potential SOFC interconnect coatings. 
 
5.2 Experimental 
(TM)-doped (Mn,Co)
3
O
4 
(TM = Ni, Cu) spinel oxides spinel oxides were synthesized 
by solid state reaction. MnO (99% pure, Alfa Aesar, Ward Hill, MA), Co
3
O
4
 (99.83% pure, 
Fisher, Pittsburgh, PA) and NiO (99% pure, Alfa Aesar, Ward Hill, MA) or CuO (99% pure, 
Alfa Aesar, Ward Hill, MA) were weighed in appropriate ratio to form the desired 
stoichiometry. Then the powder mixtures were ball-milled for 48 h with zirconia milling 
medium in de-ionized water, and the slurry was dried overnight in an oven. The pre-mixed 
powders were die-pressed, and then sintered at 1200
o
C in air. 
The interaction between the prepared TM-doped (Mn,Co)
3
O
4 
(TM = Ni, Cu) pellets 
and Cr
2
O
3
 was performed by placing the pellets in contact, and heating in air at high 
temperature using the sample configurations in Fig. 2.1 (a). After reaction, the microstructure 
and composition of samples on the surface and cross sections were characterized using a 
JEOL JSM-7000F scanning electron microscope (SEM) equipped with energy-dispersive 
X-ray (EDS). The cation concentration gradients were measured at locations in different 
images of each sample. The electrical conductivity was measured by four-probe d.c. method 
on rectangular bars with the dimension about 14 mm x 5 mm x 2 mm which were cut from 
sintered specimens. The measurements were conducted between 500
o
C and 900
o
C in random 
order leaving 10 h to reach equilibration. Pt paste and wire were used as current collector and 
 90 
electrodes. A current chosen from 300mA -500mA was applied during measurement. 
 
5.3 Results and discussion 
 
5.3.1 Mass transport in Mn
1.5-0.5x
Co
1.5-0.5x
TM
x
O
4
 (TM=Ni,Cu) 
The interactions with chromia were performed to simulate the SOFC working condition. 
Fig. 5.1(a) shows the surface of the as-prepared Mn
1.3
Co
1.3
Ni
0.4
O
4
 pellet. The pellet was 
sintered dense with close packing grains and little isolated pores. After reaction with Cr
2
O
3 
at 
800
o
C, similar to undoped (Mn,Co)
3
O
4
 samples investigated in the second chapter, a dense 
faceted morphology was formed in the region in contact with Cr
2
O
3 
as
 
shown in Fig. 5.2 (b). 
EDS analysis indicate the chemical composition of this outer reaction scale is 
Mn
0.25
Co
0.89
Cr
1.86
O
4
 spinel oxide. The surface morphologies of the other 
Mn
1.5-0.5x
Co
1.5-0.5x
TM
x
O
4 
(TM=Ni, x=0.2, 0.6, TM=Cu, x=0.2, 0.4) spinel samples after 
reaction in contact with Cr
2
O
3
 at 800
o
C for 96 h are shown in Fig. 5.2. They all have faceted 
morphologies in the region in contact with Cr
2
O
3 
after reaction, which indicated the growth 
mechanism of their reaction layer could be the same as the (Mn,Co)
3
O
4
 spinel oxides . 
 
 91 
   
Fig. 5.1 SEM images of surface of Mn
1.3
Co
1.3
Ni
0.4
O
4
 before (a) and after (b) reaction at 
800
o
C in contact with Cr
2
O
3
 for 96 h. 
 
 
 
(a) As-prepared (b)Contact with Cr
2
O
3
 
(a) Ni 0.2 
(b) Ni 0.6 
 92 
 
 
Fig. 5.2 SEM images of surface of Mn
1.5-0.5x
Co
1.5-0.5x
TM
x
O
4 
(TM=Ni, x=0.2, 0.6, TM=Cu, 
x=0.2, 0.4) after reaction at 800
o
C in contact with Cr
2
O
3
 for 96 h. 
 
Table 5.1 also lists the chemical compositions of the interface of 
Mn
1.5-0.5x
Co
1.5-0.5x
TM
x
O
4 
(TM=Ni, x=0.2-0.6, TM=Cu, x=0.2, 0.4) in contact with Cr
2
O
3
 after 
reaction at 800
o
C for 96 h. Similar to the iron and titanium, even nickel content is increased 
up to 0.6 in Mn
1.2
Co
1.2
Ni
0.6
O
4
, no nickel is observed in the interface in contact with Cr
2
O
3
, 
with the composition of Mn
0.21
Co
0.92
Cr
1.87
O
4
, which would also suggest nickel may not prefer 
to occupy the tetrahedral site as compared with manganese and cobalt during the interaction 
with Cr
2
O
3
. However, unlike previous dopants, copper appeared in the outer Cr-rich reaction 
(c) Cu 0.2 
(d) Cu 0.4 
 93 
layer, and its content increased as increasing in the Mn
1.5-0.5x
Co
1.5-0.5x
Cu
x
O
4
 samples
,
 which 
showed copper had a significant different behavior in the mass transport, and indicated it 
tends to occupy the tetrahedral site in the interaction with Cr
2
O
3
. 
 
Besides, the Co/Mn ratio 
in the outer reaction layer is still much higher than that in the original 
Mn
1.5-0.5x
Co
1.5-0.5x
TM
x
O
4
 spinel oxides. It did not change much when the nickel contact 
increased, but decreased a little in Mn
1.5-0.5x
Co
1.5-0.5x
Cu
x
O
4
 as copper concentration increased 
in the outer reaction layer. 
 
Table 5.1 Chemical compositions of Mn
1.5-0.5x
Co
1.5-0.5x
TM
x
O
4 
(TM=Ni, x=0.2-0.6, 
TM=Cu, x=0.2, 0.4) before and after reaction at 800 
o
C in contact with Cr
2
O
3
 for 96 h. 
Chemical Composition 
As-prepared After reaction 
Mn
1.4
Co
1.4
Ni
0.2
O
4
 Mn
0.2
Co
0.9
Cr
1.9
O
4
 
Mn
1.3
Co
1.3
Ni
0.4
O
4
 Mn
0.2
Co
0.9
Cr
1.9
O
4
 
Mn
1.2
Co
1.2
Ni
0.6
O
4
 Mn
0.2
Co
0.9
Cr
1.9
O
4
 
Mn
1.4
Co
1.4
Cu
0.2
O
4
 Mn
0.2
Co
0.7
Cu
0.2
Cr
1.9
O
4
 
Mn
1.3
Co
1.3
Cu
0.4
O
4
 Mn
0.1
Co
0.5
Cu
0.5
Cr
1.9
O
4
 
 
 94 
 
 
Fig. 5.3 (a) SEM image of cross section of Mn
1.4
Co
1.4
Ni
0.2
O
4
 in contact with Cr
2
O
3 
after 
reaction at 800
o
C for 96 h and (b) its corresponding cations concentration profile. 
 
 
(a)  
(a)  
 95 
 
Fig. 5.4 (a) SEM image of cross section of Mn
1.4
Co
1.4
Cu
0.2
O
4
 in contact with Cr
2
O
3 
after 
reaction at 800
o
C for 96 h and (b) its corresponding cations concentration profile. 
 
Fig. 5.3 shows (a) the SEM image of cross section of Mn
1.4
Co
1.4
Ni
0.2
O
4
 in contact with 
Cr
2
O
3 
after reaction at 800
o
C for 96 h and (b) its corresponding cations concentration profile. 
The lighter spots in a line in Fig. 5.3(a) are platinum makers which locates the original 
surface before reaction. Their current location indicates the growth mechanism of the reaction 
layer in Mn
1.4
Co
1.4
Ni
0.2
O
4
 is the same as that of (Mn,Co)
3
O
4
 spinel oxides investigated before. 
Chromium content profile in Fig. 5.3(b) also exhibited two layer contained in the reaction 
region. The outer Cr-rich spinel layer contacted with chromia grew by diffusion of cobalt and 
manganese from Mn
1.4
Co
1.4
Ni
0.2
O
4
. At the same time, chromium dissolved into 
Mn
1.4
Co
1.4
Ni
0.2
O
4
 and formed the inner layer with Cr content decreased to 0. Fig. 5.4 shows 
(a) the cross section of Mn
1.4
Co
1.4
Cu
0.2
O
4
 in contact with Cr
2
O
3 
(b) its corresponding cations 
concentration profile after reaction at 800
o
C for the same condition for comparison. Its 
morphology of the cross section is similar to that of Mn
1.4
Co
1.4
Ni
0.2
O
4
. Cr content profile in 
Fig. 5.4(b) also showed two-scale reaction layer, and Co/Mn ratio was much higher than 1, 
 96 
the original in the sample, which could be due to higher diffusion rate of cobalt in the 
reaction layer than that of manganese or higher preference of cobalt cation for tetrahedral site 
than that of manganese cation
 
which is also determined in the third chapter.
 
However, it 
should be noted that copper appeared in the reaction layer, and its content was comparable to 
that in the Mn
1.4
Co
1.4
Cu
0.2
O
4
 spinel oxide, which was observed in the sample 
Mn
1.3
Co
1.3
Cu
0.4
O
4
 after reaction too. Copper showed a quite different behavior in the mass 
transport from any transition metals dopant which had been studied in this work. As 
discussed before, the composition differences after reaction are probably related to the 
preference for the cation occupancy. In the Cr-rich outer layer, Cr
3+
 still has the strongest 
preference for the octahedral site in the spinel structure, so during reaction Cr pushed Co, Mn 
and Cu to the tetrahedral site. The observation of copper in the Cr-rich reaction layer would 
indicate that Cu prefer the tetrahedral site to the octahedral site, which is consistent with Wei 
et al?s results.
[95]
  
 
 
 97 
 
Fig. 5.5 (a) chromium and (b) nickel concentration profiles of Mn
1.2
Co
1.2
Ni
0.6
O
4
 in 
contact with Cr
2
O
3 
after reaction at 900 
o
C and 1000
 o
C for 72 h. 
 
Fig. 5.5 shows the (a) chromium and (b) nickel concentration profiles of 
Mn
1.2
Co
1.2
Ni
0.6
O
4
 in contact with Cr
2
O
3 
after reaction at 900 
o
C and 1000
 o
C for 72 h. The 
dash line in Fig. 5.5(a) indicates the location of the platinum makers, where the original 
surface is before reaction. The thickness of Cr-rich outer reaction layer in Mn
1.2
Co
1.2
Ni
0.6
O
4 
increases significantly as the reaction temperature. Different from the interaction at 800
o
C, Ni 
is observed to diffuse into the Cr-rich layer in Fig. 5.5(b), and its concentration in this region 
increases with the temperature, which may suggest enhancing reaction temperature can 
promote Ni diffusion into the reaction layer.  
 98 
 
 
 
Fig. 5.6 Surface morphologies and corresponding chemical compositions of (a) 
Mn
1.4
Co
1.4
Ni
0.2
O
4 
, (b) Mn
1.3
Co
1.3
Ni
0.4
O
4
, and (c)
 
Mn
1.2
Co
1.2
Ni
0.6
O
4
 after reaction at 1000
o
C 
in contact with Cr
2
O
3
 for 72 h. 
 
(a)Mn
0.3
Co
0.8
Cr
2.0
O
4
 
(b)Mn
0.25
Co
0.7 
Ni
0.05
Cr
2.0
O
4
 
(c)Mn
0.2
Co
0.70
Ni
0.1
Cr
2.0
O
4
 
 99 
 
 
Fig. 5.7 Nickel concentration profiles of Mn
1.5-0.5x
Co
1.5-0.5x
Ni
x
O
4 
(x = 0.2 - 0.6) in 
contact with Cr
2
O
3 
after reaction at 1000
 o
C for 72 h. 
 
 
Fig. 5.8 Nickel concentration profiles of Mn
1.5-0.5x
Co
1.5-0.5x
Ni
x
O
4 
(x = 0.2 - 0.6) in contact with 
Cr
2
O
3 
after reaction at 900
 o
C for 72 h. 
 
 100 
Fig. 5.6 are surface morphologies and corresponding chemical compositions of 
Mn
1.5-0.5x
Co
1.5-0.5x
Ni
x
O
4 
(x = 0.2 - 0.6) after reaction at 1000
o
C in contact with Cr
2
O
3
 for 72 h, 
and Fig. 5.7 shows their nickel concentration profiles in contact with Cr
2
O
3
. Ni is detected in 
the outer reaction layer, and its content is shown to increase with the original Ni content in 
Mn
1.5-0.5x
Co
1.5-0.5x
Ni
x
O
4
 after reaction at 1000
 o
C. Similar results are also observed after 
reaction at 900
 o
C for 72 h in Fig. 5.8, which is different from the reaction behavior at 800
o
C. 
 
5.3.2 Effect of Co/Ni ratio 
 
  
Fig. 5.9 Surface morphologies of (a) Mn
1.5
Co
0.9
Ni
0.6
O
4 
and (b) Mn
1.5
Co
0.6
Ni
0.9
O
4 
after 
reaction at 900
o
C in contact with Cr
2
O
3
 for 144 h. 
 
Fig. 5.9 are surface morphologies of (a) Mn
1.5
Co
0.9
Ni
0.6
O
4 
and (b) Mn
1.5
Co
0.6
Ni
0.9
O
4 
after reaction at 900 
o
C in contact with Cr
2
O
3
 for 144 h. The EDS analyses show their 
chemical compositions are Mn
0.38
Co
0.62
Ni
0.06
Cr
1.94
O
4
, and Mn
0.53
Co
0.36
Ni
0.16
Cr
1.95
O
4
, 
respectively. Fig. 5.10 shows their Ni concentration profiles in contact with Cr
2
O
3
 at the same 
reaction condition. The Co/Ni ratio is observed to have effect on the composition of both the 
(a) 
 
 
(b)  
 101 
reaction interface and outer layer. Similar to the mass transport behavior of 
Mn
1.5-0.5x
Co
1.5-0.5x
Ni
x
O
4 
(x = 0.2 - 0.6) at high temperature (e.g. 1000 
o
C), Ni cations diffuse 
outward into the Cr-rich reaction layer, and its content is increased with the Ni content in the 
original spinel oxide. Fig. 5.11 shows Cr concentration profiles of Mn
1.5
Co
0.9
Ni
0.6
O
4,
 
Mn
1.5
Co
0.6
Ni
0.9
O
4 
and Mn
1.5
Co
1.5
O
4
[148] 
after reaction at 900 
o
C in contact with Cr
2
O
3
 for 144 
h. As compared with Mn
1.5
Co
1.5
O
4
, their thicknesses of the Cr-rich layer are similar. For 
Mn
1.5
CoNi
0.5
O
4
, its thickness of the Cr-rich layer is almost the same as that of Mn
1.5
Co
1.5
O
4 
after reaction at 700 
o
C as long as 30 days, which is shown in Fig. 5.12. All these results 
could indicate the Co/Ni ratio in the spinel oxides may not affect the growth of the reaction 
layer. 
 
 
Fig. 5.10 Ni concentration profiles of Mn
1.5
Co
0.9
Ni
0.6
O
4
 and Mn
1.5
Co
0.6
Ni
0.9
O
4 
after reaction 
at 900
o
C in contact with Cr
2
O
3
 for 144 h. 
 
 102 
 
Fig. 5.11 Cr concentration profiles of Mn
1.5
Co
0.9
Ni
0.6
O
4,
 Mn
1.5
Co
0.6
Ni
0.9
O
4 
and 
Mn
1.5
Co
1.5
O
4
[148] 
after reaction at 900
o
C in contact with Cr
2
O
3
 for 144 h. 
 
 
Fig. 5.12 Cr concentration profiles of Mn
1.5
CoNi
0.5
O
4
 and Mn
1.5
Co
1.5
O
4 
after reaction at 
700
o
C in contact with Cr
2
O
3
 for 30 days. 
 
 103 
 
 
Fig. 5.13 Electrical conductivities of (a) Mn
1.5
CoNi
0.5
O
4,
 and (b) Mn
1.5
Co
0.6
Ni
0.9
O
4 
spinels as 
a function of temperature. 
 
The electrical conductivities of Mn
1.5
CoNi
0.5
O
4
 and Mn
1.5
Co
0.6
Ni
0.9
O
4 
spinel oxides as 
a function of temperature are shown in Fig. 5.12, respectively. Similar to other spinel oxides, 
the conductivities of Mn
1.5
CoNi
0.5
O
4
 and Mn
1.5
Co
0.6
Ni
0.9
O
4 
increase with temperature in a 
linear relationship between )log( T? and 1000/T, and suggest their conduction belongs to the 
 104 
small polaron hopping mechanism.
 [122, 156-158]
 Their activation energies are 0.53 eV and 0.55 
eV, respectively, which were calculated from the slope of the fitting line in Fig. 5.13. Their 
conductivities at 800 
o
C are close to each other, about half of that of Mn
1.5
Co
1.5
O
4
. The 
results may suggest the Co/Ni ratio might not influence the conductivity when Mn 
composition is kept fixed.  
 
5.4 Conclusions 
Ni and Cu-doped (Mn,Co)
3
O
4 
spinel oxides were synthesized by solid-state reaction 
method to investigate the effects of nickel and copper doping on the interaction with chromia 
at high temperature. Their Cr content profiles indicate a two-scale reaction layer, and Co/Mn 
ratio in the outer layer is much higher than 1 in the original Mn
1.5-0.5x
Co
1.5-0.5x
TM
x
O
4 
samples. 
As for the reaction at 800 
o
C, Ni is not detected in the Cr-rich reaction layer, while copper 
appears in this region where its content was comparable to the original Cu level in the spinel 
oxide. The different behavior of Cu in the mass transport from other dopants could be 
probably related to the preference for the cation occupancy in the spinel oxide. Further 
increasing reaction temperature (e.g. 1000
o
C), Ni was observed to diffuse outwards into the 
reaction layer in which its content also increases. The Co/Ni ratio was shown to have effect 
on the interface composition after reaction with chromia, but the thickness of reaction layer 
does not change. The electrical conductivities of Mn
1.5
CoNi
0.5
O
4 
and Mn
1.5
Co
0.6
Ni
0.9
O
4 
are 
close and high enough for the ceramic coating. As for Ni and Cu-doped (Mn,Co)
3
O
4 
samples, 
their thicknesses of the reaction layer are similar to that of Mn
1.5
Co
1.5
O
4
, therefore their 
potential as the coating to SOFC interconnect needs further investigation. These two 
 105 
transition metal dopants showed different mass transport behavior in the reaction with 
chromia, which is helpful for the dopant identification and the study on the interaction 
between spinel coating and oxides scale during long-term operation. 
 106 
 
 
 
CHAPTER 6 
Conclusions and Perspectives 
 
The oxidation of SOFC metallic interconnects and chromium volatilization can cause 
cathode poisoning and cell degradation. Ceramic conductive coating can provide long-term 
protection for the alloy interconnects. Manganese cobalt spinel oxides have demonstrated 
promising performances as coating material for metallic interconnect. However, the reasons 
for their excellent properties have not been completely understood. In this work, the 
interaction between (Mn,Co)
3
O
4 
spinel coatings and chromia and the transport properties and 
phase stability of the reaction layer are investigated to provide a fundamental understanding 
of (Mn,Co)
3
O
4
 coating in high-temperature operation. In addition, the effects of different 
transition metal dopants in the coating material on the mass transport behavior were studied 
for the development of improved coatings. 
After reaction, the interfaces of (Mn,Co)
3
O
4 
in contact with Cr
2
O
3
 are shown a facetted 
dense morphology. At the steady state (e.g. 1200
o
C), this outer reaction layer has a chemical 
composition of (Mn,Co)Cr
2
O
4
 with a cubic spinel structure, in which the chromium content is 
achieved to the maximum 2 Cr per M
3
O
4
 unit. Their cations concentration profiles of the 
cross section and the location of platinum markers indicate the reaction layer between 
(Mn,Co)
3
O
4 
and Cr
2
O
3
 contains two scales, Cr-rich (Mn,Co)Cr
2
O
4
 outer scale and 
 107 
Cr-containing (Mn
,
Co
,
Cr)
3
O
4
 intermediate scale. In the outer layer in contact with Cr
2
O
3
, 
(Mn,Co)Cr
2
O
4
 grows by the outward diffusion of cobalt and manganese from (Mn
,
Co)
3
O
4
 
coating material. At the same time, chromium diffuses into (Mn
,
Co)
3
O
4
 and formed the 
(Mn
,
Co
,
Cr)
3
O
4
 intermediate layer. At low reaction temperature (800
o
C and 900
o
C), the 
growth of the Cr-rich outer layer becomes dominant, which is beneficial to block the 
diffusion of chromium species. 
To evaluate the long-term stability, the properties of the reaction layer between the 
Mn
1.5
Co
1.5
O
4
 coating and Cr
2
O
3
 scale formed on the alloy surface needs further 
characterization. Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4
 (x = 0 - 2) were prepared to investigate their 
electrical properties, cation distributions and thermal expansion behavior at high temperature. 
With increasing Cr content the cubic crystal structure is stabilized, the electrical conductivity 
and thermal expansion coefficient both decrease. The temperature dependence of 
conductivities in Mn
1.5-0.5x
Co
1.5-0.5x
Cr
x
O
4 
indicates their conduction is through a small polaron 
hopping mechanism. Mn
1.5
Co
1.5
O
4
 and MnCoCrO
4
 were chosen for the neutron diffraction at 
SOFC operation temperature. The cation distributions determined from neutron data show Cr 
had stronger preference on octahedral sites compared with Mn and Co, which could suggest 
why Cr diffusion leads to reduction in conductivity of the spinel coating after operation. The 
tetrahedral site preference of Co over Mn may give the reason for higher Co/Mn ratio in the 
reaction layer than the original. Therefore, reducing the thickness of Cr-containing reaction 
layer between the oxide scale and the spinel coating could improve the overall electrical 
conduction and avoid local spallation of ceramic coating. 
Different transition metal substitutions were also studied in this work. The effects of Fe 
 108 
doping in MnCo
2-x
Fe
x
O
4 
(x = 0.15 - 0.7) spinel oxides were investigated on the reaction with 
chromia and electrical property at high temperature. Their lattice parameter expands as 
increasing iron content due to its larger cation size than that of cobalt. It is interesting to note 
that even iron content is increased to 0.7, it is not detected in the Cr-rich reaction layer, which 
would reflected that Fe may not prefer the tetrahedral site as Co and Mn during the reaction 
with Cr
2
O
3
. The electrical conductivity of MnCo
1.66
Fe
0.34
O
4 
at 800 
o
C is close to that of 
Mn
1.5
Co
1.5
O
4
, the current promising spinel coating material.  
MnCo
1.66
Ti
0.34
O
4
 was also prepared to compare the effect of doping with iron and 
titanium. Their thicknesses of reaction layer formed are depressed when compared with 
Mn
1.5
Co
1.5
O
4
. The CTE of MnCo
2
O
4
 can be also effectively decreased after iron and titanium 
doping. Similar to iron, no titanium is detected in the outer reaction layer. Ti cations prefer 
the octahedral site in spinel oxides, and the observed phenomenon may suggest Ti would not 
like the tetrahedral sites during the interaction. The conductivity of MnCo
1.66
Ti
0.34
O
4
 is lower 
than that of MnCo
1.66
Fe
0.34
O
4 
at high temperatures. 
In addition, the effects of nickel and copper doping on the interaction with chromia at 
high temperature were evaluated. Ni is not detected in the Cr-rich reaction layer after reaction 
at 800 
o
C. However, when the reaction temperature is further increased (e.g. 900 
o
C and 1000 
o
C), Ni is observed to diffuse outward into the reaction layer in which its content increases 
with the temperature. Improving Ni content in the spinel oxides and reaction temperature are 
found to promote Ni diffusion into the Cr-rich region. As for the Cu doped samples, copper 
appears in the reaction layer even with a low level at x = 0.2 after reaction at 800 
o
C. The 
different behavior in mass transport between nickel and copper could suggest they may 
 109 
possess different site preference. The differences in the mass transport behavior of transition 
metal dopants studied in our work provide useful information to identify new spinel coatings.  
The proposed future work to further understanding the interaction between the 
(Mn,Co)
3
O
4
 spinel coating materials and chromia scale and the compatiblilty of the reaction 
layer with coating material and oxidation scale related to SOFC interconnect application are 
as the following:  
When the interconnect alloy is coated with dense spinel coating, there will be oxygen 
partial pressure gradient through the interface of coating/scale to the coating layer during 
SOFC working condition. The interaction of (Mn,Co)
3
O
4
 with Cr
2
O
3
 during the reaction 
under low oxygen partial pressure will be investigated. The samples preparation will be the 
same as this work, but the reaction atmosphere will be controlled at low level of oxygen 
partial pressure (e.g. 10
-4 
- 10
-0.7
 atm, the pressure range on the cathode side 
[2]
). The reaction 
will be performed first higher than the SOFC operation temperature to overcome kinetic 
limitations, and then at 800
o
C. The phase structure of the reaction will be examined by XRD. 
Their reaction interface morphologies, compositions and the mass transport of the reaction 
layer are studied by SEM and EDS. The cation distribution in the reaction region under low 
oxygen pressure will be analyzed to learn the mass transport from neutron diffraction. The 
effect of low oxygen pressure on the growth rate of the reaction layer will be compared with 
that in air of this work for the same reaction temperature.  
The effect of low oxygen pressure will be tested on the properties of the reaction scales. 
(Mn,Co,Cr)
3
O
4
 spinel oxides will be prepared for bulk analyses of the reaction layer. The 
electrical conductivities and CTE will be measured at low oxygen partial pressure. The 
 110 
oxygen partial pressure dependence of the conduction behavior of (Mn,Co,Cr)
3
O
4
 at 800 
o
C 
will be investigated, which can provide better understanding of effect of the raction layer on 
the cell resistance for SOFC coating application. 
This work focus on the interaction behavior between the Cr-free spinel oxides and Cr
2
O
3
, 
and theses spinel oxides are expected to suppress the chromium migration problems. The 
MnCo
1.66
Fe
0.34
O
4
 and MnCo
1.66
Ti
0.34
O
4
 are observed to depress the growth of the reaction 
layer, so the long-term stability of MnCo
1.66
Fe
0.34
O
4
, MnCo
1.66
Ti
0.34
O
4
 and Mn
1.5
Co
1.5
O
4 
coating applied on the surface of alloy interconnect will be studied. The coated alloys will be 
oxidized at dual atmospheres at 800 
o
C (e.g. in air and H
2
/H
2
O, respectively) to simulate the 
SOFC working condition. After oxidation, the microstructures of the coating and will be 
characterized by SEM. The mass transport behavior (e.g. cations concentration gradient, 
thickness growth of the reaction layer formed between the coating and scale) will be analyzed 
by SEM and EDS. The mass transport behavior of dopant will be compared with results of 
bulk analysis in this work. The coated alloys will also be evaluated in the weight gain and 
ASRs. The comparison work between bulk and coating analyses can give useful information 
to further understand the transport property of the spinel coatings during operation.  
 
 111 
REFERENCES 
[1] O. Yamamoto, Solid oxide fuel cells: Fundamental Aspects and Prospects, Electrochimica 
Acta, 45(2000) 2423?2435. 
[2] W. Z. Zhu, and S.C. Deevi, Development of interconnect materials for solid oxide fuel 
cells, Mater. Sci. Eng. A, 348 (2003) 227-/243. 
[3] S. C. Singhal, K. Kendall, High temperature solid oxide fuel cells: Fundamentals, Design 
and Applications, Elsevier Advanced Technology, Oxford, UK, 2003. 
[4] N. Q. Minh, Ceramic fuel cells, J. Am. Ceram. Soc., 76 (1993) 563-588. 
[5] M. Mori, T. Yamamoto, H. Itoh, and T. Watanabe, Compatibility of alkaline earth metal 
(Mg, Ca, Sr) - doped lanthanum chromites as sparators in planar-type high-temperature 
solid oxide fuel cells, J. Mater. Sci., 32 (1997) 2423-2431. 
[6] D. Bhattacharyya, and R. Rengaswamy, A Review of Solid Oxide Fuel Cell (SOFC) 
Dynamic Models, Ind. Eng. Chem. Res., 48 (2009) 6068 - 6086. 
[7] A. Boudghene Stambouli, and E. Traversa, Solid oxide fuel cells (SOFCs): a review of an 
environmentally clean and efficient source of energy, Renewable and Sustainable Energy 
Reviews, 6 (2002) 433 -455. 
[8] M. C. Williams, J. P. Strakey, and W. A. Surdoval, The U.S. Department of Energy, 
Office of Fossil Energy Stationary Fuel Cell Program, J. Power Sources, 143 (2005) 
191-196. 
[9] P. Singh, and N. Q. Minh, Solid oxide fuel cells: Technology status, Int. J. Appl. Ceram. 
Technol, 1 (2004) 5-15. 
 112 
[10] L. Blum, W. A. Meulenberg, H. Nabielek, and R. S. Wilckens, Worldwide SOFC 
technology overview and benchmark, Int. J. Appl. Ceram. Technol, 2 (2005) 482-492. 
[11] N. Sakai, H. Yokokawa, T. Horita, and K. Yamaji, Lanthanum chromite-based 
internonnects as key materials for SOFC stack development, Int. J. Appl. Ceram. Technol, 
1 (2004) 23-30. 
[12] W. Vielstich, A. Hubert, M. Gasteiger and A. Lamm, Handbook of fuel cells ? 
Fundamentals, Technology and Applications, Willy, New York, 2003. 
[13] H. Yokokawa, N. Sakia, T. Kawada, and M. Dokiya, LaCrO
3
-Based perovskites, J. 
Electronchem. Soc., 138 (1991) 1018-1027. 
[14] M. Mori, Y. Hiei, and N. M. Sammes, Sintering behavior of Ca- or Sr-doped LaCrO
3
 
perovskites including second phase of AECrO
4
 (AE = Sr, Ca) in air. Solid State Ionics, 
135 (2000) 743-748. 
[15] M. Mori, and N. M. Sammes. Sintering and thermal expansion characterization of 
Al-doped and Co-doped lanthanum strontium chromites synthesized by the Pechini 
method, Solid State Ionics, 146 (2002) 301-312. 
[16] M. Mori, T. Yamamoto, T. Ichikawa, et al. Dense sintered conditions and sintering 
mechanisms for alkaline earth metal (Mg, Ca and Sr) - doped LaCrO
3
 perovskites under 
reducing atmosphere, Solid State Ionics, 148 (2002) 93-101. 
[17] S. Simmer, J. Hardy, J. Stevenson, et al, Sintering of non-stoichiometric strontium doped 
lanthanum chromite, J. Mater. Sci. Lett., 19 (2000) 863-865. 
 113 
[18] M. Mori, Y. Hiei, and N. M. Sammes, Sintering behavior and mechanism of Sr-doped 
lanthanum chromites with A site excess composition in air, Solid State Inoics, 123(1999) 
103-111. 
[19] N. M. Sammes, R. Ratnaraj, and M. G. Fe, The effect of sintering on the mechanical 
properties of SOFC ceramic interconnect materials, J. Mater. Sci., 29 (1994) 4319-4324. 
[20] L. W. Tai, and P. A. Lessing, Tape casting and sintering of strontium-doped lanthanum 
chromite for a planar solid oxide fuel cell bipolar plate, J. Am. Ceram. Soc., 74 (1991) 
155- 160. 
[21] J. W. Fergus, Lanthanum chromite-based materials for solid oxide fuel cell interconnects. 
Solid State Ionics, 171 (2004) 1-15. 
[22] B. Barnes, La
1-x
Sr
x
MnO
3
 based ceramics for use in solid oxide fuel cells, MS. Thesis. 
Queen?s university, Kingston, Ontario, CA, 2003. 
[23] K. P. Bansal, S. Kumari, B. K. Das, et al, On some transport properties of 
strontium-doped lanthanum chromite ceramics, J. Mater. Sci., 18 (1983) 2095-2100. 
[24] I. Yasuda, and T. Hikita, Electrical conductivity and defect structure of calcium - doped 
lanthanum chromites, J. Electrochem. Soc., 140 (1993) 1699-704. 
[25] S. Jiang, L. Liu, K. P. Ong, P. Wu, J. Li, and J. Pu, Electrical conductivity and 
performance of doped LaCrO
3
 perovskite oxides for solid oxide fuel cells, J. Power 
Sources, 176 (2008) 82?89. 
[26] K. P. Ong, P. Wu, L. Liu, and S. Jiang, Optimization of electrical conductivity of 
LaCrO
3
 through doping: A combined study of molecular modeling and experiment, Appl. 
Phys. Lett. 90 (2007) 1-3. 
 114 
[27] M. Mori, Y. Hiei, and T. Yamamoto, Control of the thermal expansion of 
stronitium-doped lanthanum chromite perovskites by B-site doping for high - 
temperature solid oxide fuel cell separators, J. Am. Ceram. Soc., 84 (2001) 781-786. 
[28] S. Srilomsak?D.  P. Schilling, and H. U. Anderson, Thermal expansion studies on 
cathode and interconnect oxide, Proceedings of the First International Symposium on 
Sold Oxide Fuel Cells, The Electrochemical Society, Pennington, NJ, 1989, 129-140. 
[29] M. Mori, H. Miyamoto, K. Takenobu, and T. Matsudaira, Controlling the chromite 
expansion under reductive atmosphere, in: Proceedings of 2nd European Solid Oxide 
Fuel Forum, Oslo, Norway, 1996, 541-546. 
[30] P. Duran, J. Tartaj, F. Capel, and C. Moure. Formation, sintering and thermal expansion 
behaviour of Sr- and Mg-doped LaCrO
3
 as SOFC interconnector prepared by the 
ethylene glycol polymerized complex solution synthesis method. J. Euro. Ceram. Soc., 
24 (2004) 2619?2629. 
[31] A. J. McEvoy, Thin SOFC electrolytes and their interfaces ? A near-term research 
strategy, Solid State Inoics, 132 (2000) 159-165.  
[32] H. Moon, S. D. Kim, S. H. Hyun, and H. S. Kim, Development of IT-SOFC unit cells 
with anode-supported thin electrolytes via tape casting and co-firing, Int. J. Hydrogen 
Energy, 33 (2008) 1758-1768. 
[33] B. Zhu, and M. D. Mat, Studies on dual phase ceria-based composites in 
electrochemistry, Int. J. Electrochem. Sci., 1 (2006) 383-402. 
 115 
[34] A. Moure, A. Castro, J. Tartaj, and C. Moure, Single-phase ceramics with 
La
1-x
Sr
x
Ga
1-y
Mg
y
O
3-t
 composition from precursors obtained by mechanosynthesis, J. 
Power Sources, 188 (2009) 489-497.  
[35] Y. J. Leng, S. H. Chan, K. A. Khor, and S. P. Jiang, Performance evaluation of anode - 
supported solid oxide fuel cells with thin film YSZ electrolyte, Int. J. Hydrogen Energy, 
29 (2004) 1025?1033 
[36] Z. Yang, Recent advances in metallic interconnects for solid oxide fuel cells, Intl Mater. 
Rev., 53 (2008) 39-16. 
[37] K. Huang, P. Y. Hou, J. B. Goodenough, Characterization of iron-based alloy 
interconnects for reduced temperature solid oxide fuel cells, Solid State Ionics, 129 
(2000) 237-250. 
[38] J. Wu, and X. Liu, Recent development of SOFC metallic interconnect, J. Mater. Sci. 
Technol., 26 (2010) 293-305. 
[39] C. Wagner, Atom Movements, Am. Soc. Metals, Cleveland, OH (1951). 
[40] K. Hauffe, Oxidation of Metals, Plenum Press, New York (1965). 
[41] W. J. Quadakkers, H. Greiner, M. Hansel, A. Pattanaik, A. S. Khanana, E. Mallener, 
Compatibility of perovskite contact layers between cathode and metallic interconnector 
plates of SOFCs, Solid State Ionics, 91 (1996) 55-67. 
[42] H. Hattendorf, R. Hojda, D. Naumenko, and A. Kolb-Telieps, A new austenitic alumina 
forming alloy: an aluminum-coated FeNi32Cr20, Mater. Corro., 59 (2008) 449-454. 
[43] N. Shaigan, W. Qu, D. G. Ivey, and W. Chen, A review of recent progress in coatings, 
surface modifications and alloy developments for solid oxide fuel cell ferritic stainless 
 116 
steel interconnects, J. Power Sources, 195 (2010) 1529-1542. 
[44] A. Holt, and P. Kofstad, Electrical conductivity and defect structure of Cr
2
O
3
. I. Higher 
temperatures (>~1000
o
C), Solid State Ionics, 69 (1994) 127-136. 
[45] A. Holt, and P. Kofstad, Electrical conductivity and defect structure of Cr
2
O
3
. II. 
Reduced temperatures (<~1000
o
C), Solid State Ionics, 69 (1994) 137-143.  
[46] W. C. Hagel, and A. U. Seybolt, Cation diffusion in Cr
2
O
3
, J. Electrochem. Soc., 108 
(1961) 1146-1152. 
[47] H. Nagai, T. Fujikawa, and K. I. Shoji, Electrical Conductivity of Cr
2
O
3
 Doped with 
La
2
O
3
, Y
2
O
3
 and NiO, Trans. Japan Inst. Met., 24 (1983) 581-588. 
[48] P. Kofstad, Nonstoichometry, diffusion and electrical conductivity in binary metal 
oxides, Robert E. Krieger Publishing Company, Malabar, FL, 1983. 
[49] Z. Yang, K. S. Weil, D. M. Paxton, and J. W. Stevenson, Selection and evaluation of 
heat-resistant alloys for SOFC interconnect applications, J. Electrochem. Soc., 150 
(2003) A1188-A1201. 
[50] Y. F. Gu, H. Harada, and Y. Ro, Chromium and chromium-based alloys: Problems and 
possibilities for high-temperature service, J. Minerals Metals Mater. Soc., 56 (2004) 
28-33. 
[51] A. H. Sully, Ductile Chromium and its Alloys, Am. Soc. Metals, Materials Park, OH, 
1957. 
[52] P. Kofstad, R. Bredesen, High temperature corrosion in SOFC environments, Solid State 
Ionics, 52 (1992) 69-75. 
[53] S. Linderoth, P. V. Hendriksen, M. Mogensen, Investigations of metallic alloys for use 
 117 
as interconnects in solid oxide fuel cell stacks, J. Mater. Sci., 31 (1996) 5077-5082. 
[54] J. Fergus, Metallic interconnects for solid oxide fuel cells, Mater. Sci. Eng. A, 397 (2005) 
271-283. 
[55] Y. Larring, R. Haugsrud, and T. Norby, HT corrosion of a Cr-5wt% Fe-1wt% Y
2
O
3
 
alloy and conductivity of the oxide scale: Effects of water vapor, J. Electrochem. Soc., 
150 (2003) B374-B379. 
[56] W. J. Quadakkers, H. Greiner, and W. Knock, Proceedings of the First European Solid 
Oxide Fuel Forum, Lucerne, Switzerland, (1994) 525. 
[57] N. Lahl, D. Bahadur, K. Singh, L. Singheiser, and K. Hilpert, Chemical interactions 
between aluminosilicate base sealants and the components on the anode side of solid 
oxide fuel cells. J. Electrochem. Soc., 149 (2002) A607-A614.  
[58] D. M. England, and A. V. Virkar, Oxidation kinetics of some nickel based supperalloy 
foils and electronic resistance of the oxide scale formed in air Part I, J. Electrochem. 
Soc., 146 (1999) 3196-3202. 
[59] D. M. England, and A. V. Virkar, Oxidation kinetics of some nickel-based supperalloy 
foils in humidified hydrogen and electronic resistance of the oxide scale formed Part II, 
J. Electrochem. Soc., 148 (2001) A330-A338. 
[60] Z. Yang, G. Xia, and J. W. Stevenson, Evaluation of Ni-Cr-base alloys for SOFC 
interconnect applications, J. Power Sources, 160 (2006) 1104-1110. 
[61] M. F. Rothman, and H. M. Tawancy, Low thermal expansion superalloy, U.S. Pat. 
4,818,486, (1988). 
 118 
[62] S. J. Geng, J. H. Zhu, and Z. G. Lu, Evaluation of Haynes 242 alloy as SOFC 
interconnect material, Solid State Inoics, 177 (2006) 559-568. 
[63] S. J. Geng, J. H. Zhu, and Z. G. Lu, Investigation on Haynes 242 Alloy as SOFC 
Interconnect in Simulated Anode Environment, Electrochem. Solid-State Lett., 9 (2006) 
A211-A214. 
[64] P. Jablonski, and D. Alman, Oxidation resistance and mechanical properties of 
experimental low coefficient of thermal expansion (CTE) Ni-base alloys, Int J. Hydro. 
Energy, 32 (2007) 3705-3712.  
[65] J. O. Anderson and B. Sundman, CALPHAD: Comput. Coupling Phase Diagrams 
Thermochem., 11 (1987). 
[66] Z. Yang, J. S. Hardy, M. S. Walker, G. Xia, S. P. Simner, and J. W. Stevenson, Structure 
and conductivity of thermally grown scales on ferritic Fe-Cr-Mn steel for SOFC 
interconnect applications, J. Electrochem. Soc., 151 (2004) A1825-A1831. 
[67] T. Brylewski, M. Nanko, T. Maruyama, K. Przybylski, Application of Fe-16Cr ferritic 
alloy to innterconnector for a solid oxide fuel cell, Solid State Ionics, 143 (2001) 
131?150. 
[68] N. K. Othman, J. Zang, D. J. Young, Temperature and water vapour effects on the cyclic 
oxidation behaviour of Fe?Cr alloys, Corr. Sci., 52 (2010) 2827-2836. 
[69] B. A. Pint, Experimental observations in support of the dynamic-segregation theory to 
explain the reactive-element effect, Oxida. Met., 45 (1996) 1-37. 
[70] E-Brite superferritic stainless steel (UNS S44627) product specifications, from website: 
http://www.atimetals.com/products/Pages/e-brite.aspx. 
 119 
[71] S. Geng, and J. Zhu, Promising alloys for intermediate-temperature solid oxide fuel cell 
interconnect application, J. Power Sources, 160 (2006) 1009-1016. 
[72] Crofer 22 APU material data sheet No. 4046, from website: http: 
//www.thyssenkrupp-vdm.com/. 
[73] J. Robertson, The mechanism of high temperature aqueous corrosion of stainless steels, 
Corro. Sci., 32 (1991) 443-465.  
[74] Y. Liu, Performance evaluation of several commercial alloys in a reducing environment, 
J. Power Sources, 179 (2008) 286-291. 
[75] S. Fontana, S. Chevalier, and G. Caboche, Metallic interconnects for solid oxide fuel cell: 
Effect of water vapor on oxidation resistance of differently coated alloys, J. Power 
Sources, 193 (2009) 136-145. 
[76] E. Konysheva, U. Seeling, A. Besmehn, L. Singheiser, an K. Hilpert, Chromium 
vaporization of the ferritic steel Crofer22APU and ODS Cr5Fe1Y
2
O
3
 alloy, J. Mater. 
Sci., 42 (2007) 5778-5784. 
[77] Z. Lu, J. Zhu, E. A. Payzant, M. P. Paranthaman, Electrical conductivity of the 
manganese chromite spinel solid solution, J. Am. Ceram. Soc., 88 (2005) 1050-1053. 
[78] J. Rufner, P. Gannon, P. White, M. Deibert, S. Teintze, R. Smith, and H. Chen, 
Oxidation behavior of stainless steel 430 and 441 at 800 degrees C in single (air/air) and 
dual atmosphere (air/hydrogen) exposures, Int. J. Hydro. Energy, 33 (2008) 1392-1398. 
[79] N. Shaigan, D. G. Ivey, and W. Chen, Metal-oxide scale interfacial imperfections and 
performance of stainless steels utilized as interconnects in solid oxide fuel cells, J. 
Elelctrochem. Soc., 156 (2009) 765-B770. 
 120 
[80] Fuel Cell Interconnector, from website: http://www.hitachimetals.com/. 
[81] A. Toji, and T. Uehara, Stability of oxidation resistance of ferritic Fe-Cr alloy for SOFC 
interconnects, ECS Trans., 7 (2007) 2117-2124.  
[82] A. Martinez-Villafane, J. G. Chacon-Nava, C. Gaona-Tiburcio, F. Almeraya-Calderon, 
G. Dominguez-Patino, and J. G. Gonzalez-Rodriguez, Oxidation performance of a 
Fe-13Cr alloy with additions of rare earth elements, Mater. Sci. Eng. A, 363(2003) 15-19. 
[83] X. Yu and Y. Sun, The oxidation improvement of Fe3Al based alloy with cerium 
addition at temperature above 1000
o
C, Mater. Sci. Eng. A, 363 (2003) 30-39.  
[84] P. Y. Hou, J. Stringer, The effect of reactive elements additions on the selective 
oxidation growth and adhesion of chromia scales, Mater. Sci. Eng. A, 202 (1995) 1-10. 
[85] W. J. Quadakkers, J. Piron-Abellan, V. Shemet, and L. Singheiser, Metallic 
interconnectors for solid oxide fuel cells - a review, Mater. High Temp., 20 (2003) 
115-127. 
[86] S. Geng, J. Zhu, M. P. Bardy, H. U. Anderson, X. Zhou, and Z. Yang, A low-Cr metallic 
interconnect for intermediate-temperature solid oxide fuel cells, J. Power Sources, 172 
(2007) 775-781. 
[87] M. Monteiro, S. Saunders, and F. Rizzo, The effect of water vapour on the oxidation of 
high speed steel, kinetics and scale adhesion, Oxida. Met., 75 (2011) 57-76. 
[88] G. Hultquist, B. Tveten, and E. H?rnlund, Hydrogen in chromium: Influence on the 
high-temperature oxidation kinetics in H
2
O, oxide-growth mechanisms, and scale 
adherence, Oxida. Met., 54 (2000) 1-10.  
[89] Y. Matsuzaki, and I. Yasuda, Electrochemical properties of a SOFC cathode in contact 
 121 
with a chromium-containing alloy separator, Solid State Ionics, 132 (2000) 271-278. 
[90] Y. Matsuzaki, and I. Yasuda, Dependence of SOFC cathode degradation by 
chromium-containing alloy on compositions of electrodes and electrolytes, J. 
Elelctrochem. Soc., 148 (2001) A126. 
[91] S. P. Jiang, J. P. Zhang, L. Apateanu, and K. Foger, Deposition of chromium species at 
Sr-doped LaMnO
3
 electrodes in solid oxide fuel cells. I. Mechanism and kinetics, J. 
Elelctrochem. Soc.,147 (2000) 4013-4022. 
[92] S. P. Jiang, J. P. Zhang, L. Apateanu, and K. Foger, Deposition of chromium species at 
Sr-doped LaMnO
3
 electrodes in solid oxide fuel cells. II. Effect on O
2
 reduction reaction, 
J. Elelctrochem. Soc.,147 (2000) 3195-3205. 
[93] S. P. Jiang, J. P. Zhang, L. Apateanu, and K. Foger, Deposition of chromium species at 
Sr-doped LaMnO
3
 electrodes in solid oxide fuel cells. III. Effect of Air Flow, J. 
Elelctrochem. Soc.,148 (2001) C447. 
[94] H. Asteman, J. E. Svensson, and L. G. Johansson, Evidence for chromium evaporation 
influencing the oxidation of 304L: The effect of temperature and flow rate, Oxida. Met., 
57 (2002) 193-216. 
[95] P. Wei, Spinel coatings for solid oxide fuel cell interconncts and crystal structure of 
Cu-Mn-O, Thesis, McMaster University, 2009. 
[96] D. Chatterjee, and S. Biswas, Development of chromium barrier coatings for solid oxide 
fuel cells, Int. J. Hydro. Energy, 36 (2011) 4530-4539.  
[97] H. Yokokawa, T. Horita, N. Sakai, K. Yamaji, M. E. Brito, Y. P. Xiong, and H. 
Kishimoto, Thermodynamic considerations on Cr poisoning in SOFC cathodes, Solid 
 122 
State Ionics, 177 (2006) 3193-3198. 
[98] C. Sun, R. Hui, and J. Roller, Cathode materials for solid oxide fuel cells: a review, J. 
Solid State Electrochem., 14 (2010) 1125-1144. 
[99] Z. Shao, and S. M. Haile, A high-performance cathode for the next generation of 
solid-oxide fuel cells. Nature, 431 (2004) 170-173. 
[100] C. Fu, K. Sun, X. Chen, N. Zhang, and D. Zhou, Electrochemical properties of A-site 
deficient SOFC cathodes under Cr poisoning conditions, Electrochim. Acta, 54 (2009) 
7305-7312. 
[101] B. Liu, Y. Gu, L. Kong, Y. Zhang, Evaluation of nano-structured Ir
0.5
Mn
0.5
O
2
 as a 
potential cathode for intermediate temperature solid oxide fuel cell, J. Power Sources, 
185 (2008) 946-951.  
[102] G. Y. Lau, M. C. Tucker, C. P. Jacobson, S. J. Visco, S. H. Gleixner, and L. C. 
DeJonghe, Chromium transport by solid state diffusion on solid oxide fuel cell cathode, 
J. Power Sources, 195 (2010) 7540-7547. 
[103] A. Montenegro-Hernandez, J. Vega-Castillo, L. Mogni, and A. Caneiro, Thermal 
stability of Ln
2
NiO
4+?
 (Ln: La, Pr, Nd) and their chemicalcompatibility with YSZ and 
CGO solid electrolytes, Int. J. Hydro. Energy, 36 (2011) 15704-15714. 
[104] Y. Li, M. W. Xu, and J. B. Goodenough, Electrochemical performance of Ba
2
Co
9
O
14
 + 
SDC composite cathode for intermediate-temperature solid oxide fuel cells, J. Power 
Sources, 209 (2012) 40-43. 
[105] W. J. Quadakkers, M. H?nsel, T. Rieck, Carburization of Cr-based ODS alloys in 
SOFC relevant environments, Mater. Corros., 49 (1998) 252?258. 
 123 
[106] W. Qua, J. Lia, and D. G. Iveyb, Sol-gel coatings to reduce oxide growth in 
interconnects used for solid oxide fuel cells, J. Power Sources, 138 (2004) 162-173. 
[107] S. Chevalier, C. Valot, G. Bonnet, J. C. Colson, and J. P. Larpin, The reactive element 
effect on thermally grown chromia scale residual stress, Mater. Sci. Eng. A, 343 (2003) 
257-264.  
[108] S. Fontana, R. Amendola, S. Chevalier, P. Piccarod, G. Caboche, M. Viviani, R. 
Molins, and M. Sennour, Metallic interconnects for SOFC : characterization of 
corrosion resistance and conductivity evaluation at operating temperature of differently 
coated alloys, J. Power Sources, 171 (2007) 652-662. 
[109] S.H. Kim, J.Y. Huh, J.H. Jun, J.H. Jun, J. Favergeon, Thin elemental coatings of 
yttrium, cobalt, and yttrium/cobalt on ferritic stainless steel for SOFC interconnect 
applications, Curr. App. Phys., 10 (2010) S86-S90. 
[110] R. Lacey, A. Pramanick, J. C. Lee, J. Jun, B. Jiang, D. D. Edwards, R. Naum, S. T. 
Misture, Evaluation of Co and perovskite Cr-blocking thin films on SOFC interconnects, 
Solid State Ionics, 181 (2010) 1294-1302. 
[111] J. H. Zhu, Y. Zhang, A. Basu, Z. G. Lu, M. Paranthaman, D. F. Lee, and E. A. Payzant, 
LaCrO
3
-based coatings on ferritic stainless steel for solid oxide fuel cell interconnect 
applications, Surf. Coat. Technol., 177-178 (2004) 65-72. 
[112] N. Shaigan, D. G. Ivey, and W. Chen, Co/LaCrO
3
 composite coatings for AISI 430 
stainless steel solid oxide fuel cell interconnects, J. Power Sources, 185 (2008) 
331-337. 
 124 
[113] W. J. Shong, C. K. Liu, C. Y. Chen, C. C. Peng, H. J. Tu, G. T. Fey, R. Y. Lee, and H. 
M. Kao, Effects of lanthanum-based perovskite coatings on the formation of oxide scale 
for ferritic SOFC interconnect, Mater. Chem. Phys., 127 (2011) 45-50. 
[114] G. V. Pattarkine, N. Dasgupta, and A. V. Virkar, Oxygen transport resistant of 
electrically conductive perovskite coatings for solid oxide fuel cell interconnects, J. 
Electrochem. Soc., 155 (2008), B1036-B 1046. 
[115] Z. Yang, G. Xia, G. D. Maupin, and J. W. Stevenson, Evaluation of perovskite overlay 
coatings on ferritic stainless steels for SOFC interconnect applications, J. Electrochem.  
Soc., 153 (2006) A1852-A1858. 
[116] R. V. Chopdekar, F. J. Wong, Y. Takamura, E. Arenholz, and Y. Suzuki, Growth and 
characterization of superconducting spinel oxide LiTi
2
O
4
 thin films, Physica C: 
Superconductivity, 469 (2009) 1885-1891. 
[117] C.B. Carter, N.M. Grant, Ceramic materials: Science and Engineering, Springer 
Science and Busines Media, LLC, New York, USA, 2007. 
[118] X. Liu and C. T. Prewitt, High-temperature X-ray diffraction study of Co
3
O
4
: 
Transition from normal to disordered spinel, Phys. Chem. Miner., 17 (1990) 168 -172. 
[119] S. Sasaki, Radial distribution of electron density in magnetite, Fe
3
O
4
, Acta. Cryst., B53 
(1997) 762-766. 
[120] R. J. Harrison, S. A. T. Redfern, and H. S. C. O?Neill, The temperature dependence of 
the cation distribution in synthetic hercynite from in-situ neutron structure refinements, 
Am. Miner., 83 (1998), 1092?1099. 
[121] A. Petric, H. Ling, Electrical conductivity and thermal expansion of spinels at elevated 
 125 
temperatures, J. Am. Ceram. Soc., 90 (2007) 1515-1520. 
[122] B. E. Martin, A. Petric, Electrical properties of copper?manganese spinel solutions and 
their cation valence and cation distribution, J. Phys. Chem. Solids, 68 (2007) 2262-2270. 
[123] M. R. Bateni, P. Wei, X. Deng, A. Petric, Spinel coatings for UNS 430 stainless steel 
interconnects, Surf. Coat. Technol., 201 (2007) 4677-4684. 
[124] P. Wei, X. Deng, M.R. Bateni, A. Petric, Oxidation and electrical conductivity behavior 
of spinel coatings for metallic interconnects of solid oxide fuel cells, Corrosion Science 
Section, 63 (2007) 529-536. 
[125] P. Wei, X. Deng, M. R. Bateni, and A. Petric, Oxidation behavior and conductivity of 
UNS 430 stainless steel and Crofer 22 APU with spinel coatings, ECS Trans., 7 (2007) 
2135-2143. 
[126] B. Hua, W. Zhang, J. Wu, J. Pu, B. Chi, L. Jian, A promising NiCo
2
O
4
 protective 
coating for metallic interconnects of solid oxide fuel cells, J. Power Sources, 195 (2010) 
7375-7379. 
[127] Z. H. Bi, J. H. Zhu, and J. L. Batey, CoFe
2
O
4
 spinel protection coating thermally 
converted from the electroplated Co?Fe alloy for solid oxide fuel cell interconnect 
application, J Power Sources, 195 (2010), 3605-3611. 
[128] Y, Liu, D. Y. Chen, Protective coatings for Cr
2
O
3
-forming interconnects of solid oxide 
fuel cells, Inter J. Hydrogen Energy, 34 (2009) 9220-9226. 
[129] W. Zhang, J. Pu, B. Chi, and L. Jian, NiMn
2
O
4
 spinel as an alternative coating material 
for metallic interconnects of intermediate temperature solid oxide fuel cells, J. Power 
Sources, 196 (2011) 5591-5594. 
 126 
[130] Z. Yang, G. Xia, X. Li, and J. W. Stevenson, (Mn,Co)
3
O
4
 spinel coatings on ferritic 
stainless steels for SOFC interconnect applications, Inter J. Hydrogen Energy, 32 (2007) 
3648-3654. 
[131] Z. Yang, G. Xia, C. Wang, Z. Nie, J. Templeton, J. W. Stevenson, and P. Singh, 
Investigation of iron-chromium-niobium-titanium ferritic stainless steel for solid oxide fuel 
cell interconnect applications, J. Power Sources, 183 (2008) 660-667. 
[132] Y. Larring, and T. Norby, Spinel and perovskite functional layers between plansee 
metallic interconnect (Cr-5 wt % Fe-1 wt % Y
2
O
3
) and ceramic (La
0.85
Sr
0.15
)
0.91
MnO
3
 
cathode materials for solid oxide fuel cells, J. Electrochem. Soc., 147 (2000) 
3251-3256. 
[133] X. Chen, P. Y. Hou, C. P. Jacobson, S. J. Visco, and L. C. De Jonghe, Prptective coating 
on stainless steel interconnect for SOFCs: oxidation kinetics and electrical properties, 
Solid State Ionics, 176 (2005) 425-433. 
[134] J. J. Choi, J. Ryu, B. D. Hahn, W. H. Yoon, B. K. Lee, and D. S. Park, Dense spinel 
MnCo
2
O
4
 film coating by aerosol deposition on ferritic steel alloy for protection of 
chromic evaporation and low-conductivity scale formation, J Mater Sci., 44 (2009): 
843-848. 
[135] B. Hua, Y. Kong, F. Lu, J. Zhang, J. Pu, and J. Li, The electrical property of MnCo
2
O
4
 
and its application for SUS 430 metallic interconnect, Chin. Sci. Bulletin, 55 (2010) 
3831-3837. 
[136] Y. Fang, C. Wu, X. Duan, S. Wang, and Y. Chen, High-temperature oxidation process 
analysis of MnCo
2
O
4
 coating on Fe-21Cr alloy, Inter. J. Hydro. Energy, 36 (2011) 
 127 
5611-5616. 
[137] R. S. Roth, Phase Equilibria Diagrams, Fig. 9570, Vol. 11, The American Ceramic 
Society, Westerville, OH, 1995. 
[138] Z. Yang, G. Xia, S. P. Simmer, J. W. Stevenson, Thermal growth and performance of 
manganese cobaltite spinel protection layers on ferritic stainless steel SOFC 
interconnects, J. Electrochem. Soc., 152 (2005) A1896-A1901.  
[139] X. Xin, S. Wang, Q. Zhu, Y. Xu, T. Wen, A high performance nano-structure conductive 
coating on a crofer22 APU alloy fabricated by a novel spinel powder reduction coating 
technique, Electrochem. Commun., (2009), doi: 10.1016 /j.elecom.2009.10.031. 
[140] X. Montero, F. Tietz, D. Sebold, H.P. Buchkremer, A. Ringuede, M. Cassir, A. 
Laresgoiti, I. Villarreal, MnCo
1.9
Fe
0.1
O
4
 spinel protection layer on commercial ferritic 
steels for interconnect applications in solid oxide fuel cells, J Power Sources, 184 
(2008) 172-179. 
[141] X. Montero, N. Jordan, J. Piron-Abellan, F. Tietz, D. Stover, M. Cassir, and I. 
Villarreal, Spinel and Perovskite Protection Layers Between Crofer22APU and 
La
0.8
Sr
0.2
FeO
3
 Cathode Materials for SOFC Interconnects, J. Electrochem. Soc.,156 
(2009) B188 - B196. 
[142] J. Wu, C. D. Johnson, R. S. Gemmen, X. Liu, The performance of solid oxide fuel cells 
with Mn?Co electroplated interconnect as cathode current collector. J. Power Sources, 
15 (2009) 1106 -1113. 
[143] W. Wei, W. Chen, and D. G. Ivey, Oxidation resistance and electrical properties of 
anodically electrodeposited Mn-Co oxide coatings for solid oxide fuel cell interconnect 
 128 
applications, J. Power Sources, 186 (2009) 428-434. 
[144] Z. Yang, G. Xia, G. D. Maupin, and J. W. Stevenson, Conductive protection layers on 
oxidation resistant alloys for SOFC interconnect applications, Surf. Coat. Tech., 201 
(2006) 4476-4483. 
[145] M. J. Lewis, and J. H. Zhu, A process of synthesis (Mn,Co)
3
O
4
 spinel coatings for 
proctecting SOFC Interconnect Alloys, Electrochem. Solid-State Lett., 14 (2011) 
B9-B12. 
[146] E. Saoutieff, G. Bertrand, M. Zahid, and L. Gautier, APS deposition of MnCo
2
O
4
 on 
commercial alloys K41X used as solid oxide fuel cell interconnect: The importance of 
post heat-treatment for densification of the protective layer, ECS Trans., 25 (2009) 
1397-1402. 
[147] A. Navrotsky, O.J. Kleppa, The thermodynamics of cation distributions in simple 
spinels, J. Inorg. nucl. Chem., 29 (1967) 2701-2714. 
[148] K. Wang, data report, Auburn, 2009. 
[149] K. Wang, Y. Liu, J. W. Fergus, Interactions between SOFC interconnect coating 
materials and chromia, J. Am. Ceram. Soc., 94 (2011) 4490-4495. 
[150] W. Qu, L. Jian, J. M. Hill, and D. G. Ivey, Electrical and Microstructural 
Characterization of Spinel Phases as Potential Coatings for SOFC Metallic 
Interconnects, J. Power Sources, 153 (2006) 114-124. 
[151] V. F. Sears, Neutron News, 3 (1992) 29-37. 
[152] J. Rodriguez-Carvajal, Proceedings of the Satellite Meeting on Powder Diffraction, 
Toulouse, France (1990). 
 129 
[153] JCPDS, International Center for Diffraction Data. 
[154] A. Purwanto, A. Fajar, H. Mugirahardjo, J. W. Fergus, K. Wang, Cation distribution in 
spinel (Mn,Co,Cr)
3
O
4
 at room temperature, J. Appl. Cryst, 43 (2010), 394-400.  
[155] R. D. Shannnon, Revised effective ionic radii and systematic studies of interatomic 
distance in Halides and Chalcogenides, Acta. Cryst., A32 (1976) 751. 
[156]
 
R. Schmidt, A. Basu, and A. W. Brinkman, Small Polaron Hopping in Spinel 
Manganates, Phys. Rev. B, 72 (2005) 1-9. 
[157]
 
N. M. Kovtun, V. K. Prokopenko, and A. A. Shamyakov, Electroconductivity and 
Electron Exchange in Spinel Structures, Solid State Comm., 26 (1978) 877-878. 
[158] R. Schmidt, A. Basu, and A. W. Brinkman, Small polaron hopping in spinel 
manganates, Phys. Rev. B, 72 (2005) 1-9. 
[159] H. Bordeneuve, A. Rousset, C. Tenailleau, and S. Guillement-Fritsch, Cation 
distribution in manganese cobaltite spinels Co
3?x
Mn
x
O
4
 (0?x?1) determined by 
thermal analysis, J. Therm. Anal. Calorim., 101 (2010) 137-142. 
[160] S. Wang, X. Liu, Y. Feng, Q. He, and H. Wang, In situ high-temperature powder X-ray 
diffraction study on the spinel solid solutions (Mg
1?x
Mn
x 
)Cr
2
O
4
, Phys. Chem. Mater., 
(2011) 10.1007/s00269-011-0474-8. 
[161] B. Gillot, J. L. Baudour, F. Bouree, R. Metz, R. Legros, and A. Rousset, Ionic 
configuration and cation distribution in cubic nickel manganite spinels Ni
x
Mn
3?x
O
4
 
(0.57<x<1) in relation with thermal histories, Solid State Ionics, 58 (1992) 155-161. 
[162] G. D. C. Csete de Gyorgyfalva, and I. M. Reaney, Decomposition of NiMn
2
O
4
 spinel: 
an NTC thermistor material, J. Euro. Ceram. Soc., 21 (2001) 2145-2148.