Fundamental Studies of Perovskite related Oxide Thin Films for Oxygen
Electrocatalysis at Intermediate Temperatures
By
Dongkyu Lee
Bachelor of Science, Metallurgy and Materials Engineering
Hanyang University, 2004
Master of Science, Materials Science and Engineering
Gwangju Institute of Science and Technology, 2006
SUBMITTED TO THE DEPARTMENT OF MECHANICAL ENGINEERING IN PARTIAL
FULLFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY IN MECHANICAL ENGINEERING
/Tq
g
MASSACHUSETS INSTITUTE
AT THE
OFTECHNOLOGY
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
AUG 15 2014
JUNE 2014
LIBRARIES
C Massachusetts Institute of Technology 2014. All rights reserved.
Signature redacted
Authored by
Dongkyu Lee
Department of Mechanical Engineering
May 19,2014
Certified
bySignature
redacted
Yang Shao-Horn
Gail E. Kendall Chair of Mechanical Engineering
Thesis Supervisor
Accepted by
Signatureredacted-
David E. Hardt
Cross Professor of Mechanical Engineering
Chairman, Department Committee of Graduate Students
2
-
This page is intentionallyleft blank
Fundamental Studies of Perovskite related Oxide Thin Films for Oxygen
Electrocatalysis at Intermediate Temperatures
by
Dongkyu Lee
Submitted to the Department of Mechanical Engineering on May 19, 2014, in partial fulfillment
of the requirements for the degree of Doctor of Philosophy in Mechanical Engineering
Abstract
Discovering highly active and stable catalysts for electrochemical energy conversion and
storage is essential to envision a new generation of renewable energy applications. Mixed ionic
and electronic conductors (MIECs) such as Lai.xSrxCoO3-8 (LSC 1 3 ) and La.xSrxCoj.yFeyO3-8
(LSCF,13 ) are currently utilized for applications including oxygen permeation membranes and
solid oxide fuel cells (SOFCs), but alternative materials with higher catalytic activity and
stability are required for intermediate temperature (500 - 700 'C) oxide electrocatalysts. In this
thesis, two promising strategies, 1) Ruddlesden-Popper (RP) oxides and 2) surface decoration on
the MIEC oxides are proposed to design highly active oxide materials and improve the
fundamental understanding of the oxygen electrocatalysis at intermediate temperature.
The oxygen surface exchange kinetics of a-axis-oriented La2NiO 4+8 (LNO) thin films
increases with decreasing film thickness. Increasing volumetric strains in the LNO films at
elevated temperatures are correlated with increasing surface exchange kinetics and decreasing
film thickness. Volumetric strains may alter the formation energy of interstitial oxygen and
influence on the surface oxygen exchange kinetics of the LNO films. The effect of strontium (Sr)
substitution on the oxygen electrocatalysis of RP oxides is also investigated using La2-.SrxNiO4 8
(LSNO, 0.0 5 xsr S 1.0) thin films. A structure reorientation occurs with increasing the Sr content,
which can result from different energies in each surface. The surface exchange kinetics of LSNO
is strongly dependent on the Sr content. This observed surface exchange kinetics can be
attributed to the different oxygen adsorption energies and crystallographic orientations.
The oxygen surface exchange kinetics of LSC11 3 is significantly enhanced by
La0 8Sro.2CoO3-8 (LSM1 1 3 ) surface decoration as shown in LSC 2 14 -decorated LSC11 3. In addition,
long-term stability of LSC113 is significantly improved by LSM11 3 coverage. The suppression of
Sr-enriched particles and substantial changes in the surface cationic ratios are associated with
LSM11 3 decoration, which can contribute the enhanced surface exchange kinetics and stability of
LSM113-decorated LSC113. In contrast to the LSC 214-decorated LSC113 , LSC214 decoration does
not lead to the enhancement of the surface exchange kinetics and the long-term stability of
LSCF113. The change in the surface electronic structure and the suppression of the formation of
secondary passive phases as a result of LSC 214 decoration can be responsible for observed
oxygen surface exchange kinetics.
Thesis Supervisor: Yang Shao-Horn
Title: Gail E. Kendall Chair of Mechanical Engineering
3
This page is intentionallyleft blank
4
List of Publications
.
1. Dongkyu Lee, Yueh-Lin Lee, Alexis Grimaud, Wesley T. Hong, Michael D. Biegalski, Dane
Morgan, and Yang Shao-Horn, "Strontium Influence on the Oxygen Electrocatalysis of La2
xSrxNiO 4 t 6 (0 X 1.0) Thin Films", JournalofMaterials ChemistryA, 2 (18), 6480-6487,
2014, DOI: 10.1039/C3TA14918H
2. Dongkyu Lee, Alexis Grimaud, Ethan J. Crumlin, Khaled Mezghani, Mohamed A. Habib,
Zhenxing Feng, Wesley T. Hong, Michael D. Biegalski, Hans M. Christen, and Yang ShaoHorn, "Strain Influence on Oxygen Electrocatalysis of the (100)-Oriented Epitaxial
La2NiO 4+8 Thin Films at Elevated Temperatures", Journalof Physical Chemistry C, feature
article, 117 (37), pp 18789-18795, 2013, DOI: 10.1021/jp404121p
3. Dongkyu Lee, Yueh-Lin Lee Wesley T. Hong, Alexis Grimaud, Michael D. Biegalski, Dane
Morge and Yang Shao-Horn, "Enhanced Surface Stability and Electrocatalytic Activity on
Epitaxial Lao.sSro.2CoO 3 -8 Thin Films by Lao.8 Sr0 .2 MnO 3 - Decoration", In revision, 2014
4. Dongkyu Lee, Yueh-Lin Lee, Michael D. Biegalski, Stuart Adler, Dane Morgan, and Yang
Shao-Horn, "The effect of (LaSr)2 CoO 4 &/Lao.sSr. 2 MO 3 .8 (M = Co, Fe) hetero-interface on
Oxygen Electrocatalsis and Durability at Elevated Temperatures", In preparation,2014
5. Dongkyu Lee, Jiang Lu, Ho Nyung Lee, and Yang Shao-Horn, "Crystallographic Orientation
Effect on the Oxygen Electrocatalysis of Lal.s5 Sro.lsCuO4-s Thin Films", In preparation,
2014
6. Lei Wang, Susumu Imashuku, Alexis Grimaud, Dongkyu Lee, Khaled Mezghani, Mohamed
A. Habib, and Yang Shao-Horn, "Enhancing oxygen permeation of electronically shortcircuited oxygen-ion conductors by decorating with mixed ionic-electronic conducting
oxides", ECS Electrochem. Letters, 2 (11), 2013, DOI: 10.1149/2.002311 eel
7. Zhenxing Feng, Ethan J. Crumlin, Wesley T. Hong, Dongkyu Lee, Eva Mutoro, Michael D.
Biegalski, Hua Zhou, Zhi Liu, Hans M. Christen, and Yang Shao-Horn, "In Situ Studies of
Temperature-Dependent Surface Structure and Chemistry of Single-Crystalline (001)Oriented Lao.8 r. 2 CoO 3.-8 Perovskite Thin Films", Journal of Physical Chemistry Letters, 4
(9) 1512-1518 May 2013, DOI: 10.1021/jz400250t
8. Ethan Crumlin, Sungjin Ahn, Dongkyu Lee, Eva Mutoro, Michael D. Biegalski, Hans M.
Christen, and Yang Shao-Horn, "Oxygen Electrocatalysis on Epitaxial Lao.6SrO. 4 CoO 3-8
Perovskite Thin Films for Solid Oxide Fuel Cells", Journalof The ElectrochemicalSociety,
159 (7) F219-F225 (2012), DOI: 10.1 149/2.018207jes
9. Deok-Hyung Lee, Sun-Ghil Lee, Jong Ryeol Yoo, Gyoung-Ho Buh, Guk Hyon Yon, DongWoon Shin, Dong Kyu Lee, Hyun-Sook Byun, In Soo Jung, Tai-su Park, Yu Gyun Shin,
Siyoung Choi, U-In Chung, Joo-Tae Moon, and Byung-Il Ryu, "Improved Cell Performance
5
for sub-50 nm DRAM with Manufacturable Bulk FinFET Structure", 2007 Symposium on
VLSI Technology Digest of Technical Papers, 9B-1, p 16 4 (2007)
10. Dongkyu Lee, Sungkweon Baek, Sungho Heo, Changhee Cho, Gyoungho Buh, Taisu Park,
Yugyun Shin, and Hyunsang Hwang, "Ultra-Shallow p+/n Junction Prepared by Low Energy
BF3 Plasma Doping and KrF Excimer Laser Annealing", Electrochem. Solid-State Lett., vol.
9 no. 1, G19 (2006), DOI: 10.1149/1.2138448
11. Dongkyu Lee, Sungho Heo, Changhee Cho, G. H. Buh, Tai-su Park, Jongryeol Yoo, Yugyun
Shin, and Hyunsang Hwang, "Electrical Characteristics of Ultra-Shallow p+/n Junction
Formed by BF3 Plasma Doping and Two Step Annealing Process", Electrochem. Solid-State
Lett. vol. 9 no. 4, G121 (2006), DOI: 10.1149/1.2170461
12. Sungho Heo, Sungkweon Baek, Dongkyu Lee, Musarrat Hasan, Hyungsuk Jung, JongHo
Lee, and Hyunsan Hwang, "Sub-15nm n+/p Germanium Shallow Junction Formed by PH3
Plasma doping and Excimer Laser Annealing", Electrochem. Solid-State Lett., vol. 9 no.4,
G136 (2006), DOI: 10.1149/1.2172470
13. Sungho Heo, Sungkweon Baek, Dongkyu Lee, Gyonho Buh, Yugyun Shin, and Hyunsang
Hwang, "Ultrashallow arsenic n+/p junction formed by AsH 3 plasma doping", Jpn. J. Appl.
Phys., vol. 45 no. 13, L373 (2006), DOI: 10.1 143/JJAP.45.L373
14. Hyundoek Yang, Sungho Heo, Dongkyu Lee, Sangmoo Choi, and Hyunsang Hwang,
"Effective Workfunction of Scandium Nitride Gate Electrodes on Si02 and HO2", Jpn. J
Appl. Phys., vol 45 no.3, L83 (2006), DOI: 10.1 143/JJAP.45.L83
6
Acknowledgements
I vividly remember the day when I got admission in the Massachusetts Institute of Technology in
winter 2009 and the day was one of the best moments of my life. Five years later, another best
day of my life lies ahead of me. I am grateful to my advisor, Professor Yang Shao-Horn for the
opportunities and support over the last five years. She has been a role model for not only the
scientist I want to become but also the person I want to follow. I would also like to thank my
committee members, Professor Ahmed F. Ghoniem, Alexie M. Kolpak, and Harry L. Tuller for
their valuable comments and time.
This thesis would not have been possible without the wonderful dedication and support of
various research collaborators I have had the pleasure of interacting with over my years
including Dr. Ethan Crumlin, Dr. Eva Mutoro, Dr. Lei Wang, Dr. Yueh-Lin Lee, Dr. Zhenxing
Feng, Wesley Hong, Libby Shaw, Kurt Broderick, Professor Dane Morgan (U of WisconsinMadison), Professor Stuart Adler (U of Washington-Seattle), Dr. Michael Biegalski (Oak ridge
national lab), Dr. Ho Nyung Lee (Oak ridge national lab), Lu Jiang (U of Tennessee -Knoxville).
I also thank Dongwook Lee for your assistance in EEL. My special thanks are extended to the
members of the Mechanical Engineering at Korean Graduate Student Association, MIT for their
help to reach my full potential.
I would like to dedicate this work to all my family. Especially, I could not have done my Ph.D.
without my wife, Soomi. I am truly thankful for that you have been through a hard time with me.
I have been able to overcome difficulties because of you. To my parents and parents-in-law, I
thank all of you for your endless love and unstinting support. Hopefully, I have shown you that I
not only have strength and wisdom, but the perseverance in achieving success despite the
challenges in front of me. To my great aunt and uncle, I would appreciate your confidence and
continued support. I will try harder to meet your expectations and please keep watching me. To
my sister, congratulations on your Ph.D. defense and hopefully you will have only happy days in
the future. My sincere thanks also go to my brother-in-law and his wife for their encouragement.
My love and gratitude extends to the end of the world for all of you so this work is dedicated to
you, my loving and devoted family.
7
This page is intentionallyleft blank
8
Table of Contents
A bstract...........................................................................................................................................3
List of Publications ........................................................................................................................
5
A cknow ledgm ents ..........................................................................................................................
7
Table of C ontents ...........................................................................................................................
9
List of Figures...............................................................................................................................13
List of Tables ................................................................................................................................
29
Chapter 1. Introduction & Background
1.1 M otivation..........................................................................................................................31
1.2 Solid O xide Fuel Cells...................................................................................................
33
1.3. Cathodes of Solid Oxide Fuel Cells.............................................................................
35
1.4 M ixed Ionic and Electronic Conductors ........................................................................
36
1.4.1 ABO 3 Perovskite Oxides......................................................................................
37
1.4.2 Ruddlesden-Popper (RP) Oxides .........................................................................
39
1.5 La2.xSrxN iO4+a (0.0 5 xsr 5 1.0) Oxide..........................................................................
40
1.5.1 Structure of La 2 .xSrxN iO4
8...........................................
........... ..................... ... . . .
1.5.2 Oxygen Nonstoichiometry (6) of La2-xSrxNiO 4 s .................................................
40
42
1.5.3 Oxygen Surface Exchange Kinetics of La2.xSrxNiO 4 8.........................................43
1.6 Surface D ecoration on the MIEC Oxides .....................................................................
44
1.7 Thin Film Cathodes for Solid Oxide Fuel Cells ................................................................
45
1.8 Technical Overview ......................................................................................................
46
1.9 References..........................................................................................................................48
Chapter 2. Experimental Approach
2.1 Synthesis & Fabrication.................................................................................................
9
55
2.1.1 Synthesis of Thin Film by Pulsed Laser Deposition (PLD) .................
55
2.1.2 Fabrication of Microelectrodes.............................................................................
56
2.2 Characterization.................................................................................................................57
2.2.1 Surface M orphology and Structural Analysis......................................................
57
2.2.2 Surface Chemistry Analysis..................................................................................
58
2.2.3 Electrochemical Impedance Spectroscopy (EIS).................................................
59
2.2.4 In situ High-resolution X-ray Diffraction (HRXRD) ..........................................
61
2.3 References..........................................................................................................................61
Chapter 3. Strain Influence on the Oxygen Electrocatalysis of the (100)-Oriented Epitaxial
La 2 NiO 4 +8 Thin Films at Elevated Temperatures
3.1 Introduction........................................................................................................................64
3.2 Experimental M ethods...................................................................................................
65
3.3 Results and Discussion .................................................................................................
67
3.4 Conclusion .........................................................................................................................
76
3.5 Supporting Information..................................................................................................
76
3.6 References..........................................................................................................................88
Chapter 4. Strontium Influence on the Oxygen Electro catalysis of La2.SrNiO4sa (0.0
< XSr
5 1.0) Thin Films
4.1 Introduction........................................................................................................................92
4.2 Experimental M ethods...................................................................................................
93
4.3 Results and Discussion .................................................................................................
97
4.3.1 Structural Reorientation of the LSNO Thin Films with increasing Sr content ........ 97
4.3.2 Oxygen Surface Exchange Kinetics of the LSNO Thin Films ...............................
101
4.3.3 Proposed Origin for the Oxygen Surface Exchange Kinetics of LSNO as a function
of the Sr content .....................................................................................................
4.4 Conclusion .......................................................................................................................
4.5 Supporting Information....................................................................................................107
10
103
106
4.6 References........................................................................................................................121
Chapter 5. Enhanced Surface Stability and Electrocatalytic Activity on Epitaxial
Lao.sSr. 2 CoO 3.. Thin Films by Lao.8Sro.2 MnO 3-&Decoration
5.1 Introduction......................................................................................................................128
5.2 Experim ental Methods.....................................................................................................130
5.3 Results and Discussion ....................................................................................................
133
5.4 Conclusion .......................................................................................................................
142
5.5 Supporting Inform ation....................................................................................................143
5.6 References........................................................................................................................159
Chapter 6. The effect of (La,Sr)2 CoO 45 /Lao.sSro. 2 MO 3- (M = Co, Fe) Hetero-interface on
Oxygen Electrocatalsis at Elevated Temperatures
6.1 Introduction......................................................................................................................166
6.2 Experim ental M ethods.....................................................................................................168
6.3 Results and Discussion ....................................................................................................
170
6.4 Conclusion .......................................................................................................................
185
6.5 Supporting Inform ation....................................................................................................186
6.6 References........................................................................................................................206
Chapter 7. Conclusions and Perspectives
7.1 Conclusions......................................................................................................................213
7.1.1 Layered RP Oxide Thin Films (LN O and LSN O)..................................................213
LSC 214 /LSC
3
3
,
7.1.2 ABO 3 Perovskite Oxide Thin Films with Surface Decoration (LSM,1 3 /LSC 1
, and LSC 2 , 4/LSCF,1 3).....................................................................214
7.2 Perspectives for Future Work ..........................................................................................
7.2.1 Higher order RP oxides (n=2, 3).............................................................................215
7.1.2 Double Perovskites Oxides.....................................................................................216
11
215
7.3 References-........................................................................................................216
12
List of Figures
Figure 1-1. A graph provided by U.S. Energy Information Administration (EIA), International
Energy Outlook 2013 depicting the increase in energy use of the world.................................31
Figure 1-2. A graph provided by the Association for the Study of Peak Oil and Gas (ASPO)
depicting the production of oil, coal, and gas natural resources in million tonne of oil equivalent
(Mtoe) since the 1900's and projected to 2100, suggesting a peak in production in the 2010
decade............................................................................................................................................32
Figure 1-3. A graph provided by The Green Box Systems Group 5 showing the comparison of
electrical efficiency between fuel cells and conventional combustion engines.........................32
Figure 1-4. Schematic of a fuel cell operating on oxygen at the cathode (in blue), and hydrogen
at the anode (in green) separated by the oxygen ion-conducting electrolyte (in orange)......33
Figure 1-5. Voltage losses as a function of SOFCs operating temperature representing the
highest voltage losses from the cathode due to the slow ORR kinetics....................................34
Figure 1-6. ABO 3 perovskite structure, where A is rare earth, B an transition metal. ............ 37
Figure 1-7. Schematic crystal structures of the n = 1, 2 and 3 members of the RuddlesdenPopper (RP) type A.+1B,03,+1 are shown. The denotation of n represents the number of stacked
octahedral layers separated by a rock salt LaO layer. NiO 6 octahedra and La atoms are shown
schem atically ..................................................................................................................................
39
Figure 2-1. A photo of pulsed laser deposition (PLD) for the growth of an epitaxial thin film on
YSZ with an interlayer of GDC. A graph shows the reflection high-energy electron diffraction
(RHEED), which can allow an in-situ monitoring of the film growth. ....................................
13
55
Figure 2-2. A photolithography process of thin film microelectrode.....................................
56
Figure 2-3. (a) Normal X-ray diffraction (XRD), and (b) off-normal XRD results of La 2
xSrxN iO 4s_ (0.0
5 XSr
<
1.0) thin films. 2 .............................................
.......... .............. ............. . . .
57
Figure 2-4. (a) AFM images of the -0.9 nm LSM1 13-decorated LSC 113 thin film with root
mean square (RMS) of 0.968 nm after annealing at 550
oC,3
and (b) SEM images of the as-
prepared LSNO with Sr = 0,4.2 ......................................................
58
Figure 2-5. AES and SEM analysis for bare LSC,1 3 thin film before and after annealing.
Annealing was performed at 550 *C in an oxygen partial pressure of 1 atm. (a) Sr Auger spectra
of bare LSC113 thin film probed for: as-deposited surface (gray), and particles (dark red) and film
surface (dark blue) after annealing. SEM image of (b) as-deposited and (c) annealed LSC113 .3 ..59
Figure 2-6. (a) Schematic of a Film/GDC/YSZ(001)/porous Pt sample and electrochemical
testing configuration (not drawn to scale) and (b) equivalent circuit (RI = YSZ electrolyte
resistance, R2 = electrode/electrolyte interface resistance9 , RORR = ORR resistance, CPE =
constant phase element) used to extract ORR kinetics, and (c) Typical nyquist plot of the thin
film electrode at 550 0C (HF: 104 ~ 10' Hz, MF: 10 3 ~ 104 Hz, and LF: 10-2 ~ 10 3 Hz).............60
Figure 2-7. (a) In situ HRXRD data of the unit cell volume of the (1 00)tetragonai-oriented epitaxial
LNO thin films with ~ 14, ~ 42, -178 and -390 nm as a function of temperature in a p(02) of 1
atm,4 and (b) a schematic of In situ HRXRD.............................................................................
61
Figure 3-1. High resolution X-ray diffraction (CuK a) analysis. (a) Normal XRD of the
(I00)tetragonap-oriented epitaxial LNO thin films (-14, -42, -178, and ~ 390 nm), and (b) Offnormal XRD of the (100)tetragnai-oriented epitaxial LNO thin films (-14 nm), GDC, and YSZ (c)
schematic of the crystallographic rotational relationships among the LNO ( 2 0 0 )tetragona, GDC
(00 2 )cubic and Y SZ ( 0 02 )cubic.....................................................................................................
14
68
Figure 3-2. Structural stability and unit cell volume of the (100)tetragona-oriented epitaxial LNO
thin films (a) In situ HRXRD data (CuKa) of normal scan in the 0-26 Bragg-Brentano geometry
as a function of temperature with ~ 178 nm LNO film, showing no phase change upon heating at
a p( 0 2) of 1 atm. The starred (*) peaks originated from the heater, and the peaks of the LNO
film, GDC buffer layer and YSZ substrate are indexed to the tetragonal, cubic and cubic
structure, respectively. (b) In situ HRXRD data of the unit cell volume of the (lOO)ttagonaloriented epitaxial LNO thin films with
-
14, ~ 42, -178
and -390 nm as a function of
..... 69
temperature in a p(02) of 1 atm . ..................................................................................
Figure 3-3. Ex situ AES data of the (100)tetragonai-oriented epitaxial LNO thin films with ~ 14, ~
42, -178 and -390 nm annealed at 550 *C in an oxygen partial pressure of 1 atm. (a) La cation
variation and (b) Ni cation variation at three different locations on the LNO film surface. (c) The
change of the surface Ni/La ratio as a function of film thickness. Normalized to the value
obtained at - 14 nm LN O film .................................................................................................
70
Figure 3-4. Electrochemical impedance spectroscopy (EIS) results of microelectrodes for the
(100)tetragona-oriented epitaxial LNO thin films with
-
14, ~ 42, -178 and -390 nm at 550 *C (a)
Nyquist plot of the (100)tetragonal-oriented epitaxial LNO thin films with
-
14,
-
42, -178 and
-390 nm in 1 atm p(O2). (b) Nyquist plot of the (100)tetragonai-oriented epitaxial LNO thin films
with
-
14 nm as a function of p(O2). (c) Oxygen partial pressure dependency of the surface
exchange coefficients, kq of the (100)tetragonai-oriented epitaxial LNO thin films with ~ 14, ~ 42,
-178
and -390 nm calculated from EIS spectra collected at 550 *C. Extrapolated bulk k*
(approximately equivalent to k ) 4 1 value at 550 *C obtained from previous data of (*-gray)
Skinner et al. 0 is plotted for comparison. (d) Oxygen partial pressure dependency of volume
specific capacitance (VSC) of the (100)tetragonai-oriented epitaxial LNO thin films with ~ 14, ~ 42,
-178 and --390 nm calculated from EIS spectra collected at 550 *C. .......................................
72
Figure 3-5. Thickness dependency of the kq of the (100)tetragonai-oriented epitaxial LNO thin films
with ~ 14, ~ 42, -178 and -390 nm calculated from EIS spectra collected at 550 *C in 1 atm
p(O2). The inset shows the schematic of the surface exchange in the (100)tetragonal-oriented
epitaxial LN O thin film s................................................................................................................73
15
Figure 3-6. (a) Constrained (M-black,
*-blue)
and Relaxed (E-black, 0-blue) lattice
parameters of the (l00)tetragona-oriented epitaxial LNO thin films as a function of the surface
oxygen exchange kinetics at 550 *C in 1 atm p(O2). Extrapolated bulk a (*-black), c (*-blue)
lattice parameters and k* value at 550 *C obtained from previous data of Skinner et al.30 39 are
plotted for comparison. (b) Thickness and volumetric strain of the (100)tetragona-oriented epitaxial
LNO thin films as a function of the surface oxygen exchange kinetics at 550 *C in 1 atm p(O2).
(c) Unit cell volume and 8 extrapolated from Nakamura et al.3 7 as a function of the (IOO)tetragonaloriented epitaxial LNO thin films thickness at 550 'C in 1 atm p(O2). Extrapolated bulk unit cell
volume and 6 at 550 'C obtained from previous data of (gray line) Nakamura et al.3 are plotted
for comparison. (d) k of the (100)tetragonai-oriented epitaxial LNO thin films as a function of 6
extrapolated from Nakamura et al.3 at 550 0 C in 1 atm p(O2). Extrapolated bulk k* value and 8
obtained from previous data of (*-gray) Skinner et al.3 0 and Nakamura et al.37 , respectively are
plotted for comparison...................................................................................................................75
Figure S3-1. Constrained lattice parameters of the (I00)tetragona-oriented epitaxial LNO thin films
with 14, 42, 178, and 390 nm extracted from normal and off-normal HRXRD data at room
temperature. The constrained normal and in-plane lattice parameters of the LNO films were
calculated from combining the interplanar distances of the ( 2 00 )tetragonali, (10 3 )tetragonal and
(1 1 4 )tetragonal peaks..........................................................................................................................81
Figure S3-2. (a) Relaxed a lattice parameter, and (b) ^ lattice parameter of the LNO films as a
function of temperature in an oxygen partial pressure of 1 atm. For determining the relaxed film
lattice parameter a and ^, we used the equation:
(c-c) = -2v (a-a)
=-2v
(--,
1-V a
C
tl
assuming /
3 =
3.284, 3.286, 3.286, 3.286, 3.283, and 3.282 A for 298, 423,
523, 623, 723, and 823 K, respectively and v = 0.3...................................................................
82
Figure S3-3. Normal and In-plan strain of the (I00)tetragonai-oriented epitaxial LNO thin films as a
function of the film thickness at room temperature. Normal and in-plan strains are calculated
using the equation: Normal strain, Eaa = (a-a) and in-plane strain, Ecc =
a
16
............
.... 83
Figure S3-4. AFM measurements of the as-prepared (a) LNO films ~ 14 nm with RMS of
0.216nm, (b) LNO films
-
42 nm with RMS of 0.341nm, (c) LNO films -178 nm with RMS of
0.353nm, and (d) LNO films ~ 390 nm with RMS of 0.407 nm. ..............................................
84
Figure S3-5. Scanning electron microscopy (SEM) images of (a) LNO films ~ 14 nm, (b) LNO
films ~ 42 nm, (c) LNO films ~ 178 nm, and (d) LNO films ~ 390 nm annealed at 550 *C in an
oxygen partial pressure of 1 atm for 6 hours. ...........................................................................
85
Figure S3-6. (a) Schematic of a LNO/GDC/YSZ(001)/porous Pt samples and electrochemical
testing configuration (not drawn to scale), and (b) equivalent circuit (Ri = YSZ electrolyte
resistance, R 2 = electrode/electrolyte interface resistance' 7 , RORR = ORR resistance, CPE =
constant phase element) used to extract ORR kinetics, and (c) Nyquist plot of the (l00)tetragonaPoriented epitaxial LNO thin films with ~ 14 nm at 550 0 C; inset shows a magnification (HF: 10 4
~ 105 Hz, M F: 10 3 ~ 10 4 Hz, and LF: 10- ~ 103 Hz).................................................................
86
Figure S3-7. Schematic of two different orientations of LNO on GDC (a) (100)tetragonai-oriented
epitaxial LNO thin film and (b) (00 1)tetragonai-oriented epitaxial LNO thin films......................87
Figure 4-1. Schematic of the crystallographic rotational relationships of LSNO (l00)tetm., GDC
(000cbic and YSZ (000cubic ((a) top) and LSNO (000tetra., GDC (000cbic and YSZ (000cubic ((b)
top). (100)ttra. top view of LSNO thin film ((a) bottom) and (00I)tetra. top view of LSNO thin film
((b) bottom )....................................................................................................................................97
Figure 4-2. HRXRD analysis (a) Normal XRD of the La 2-xSrxNiO4 z (LSNO) thin films with 0 5
xsr 5 1.0, (b) schematic of a structural reorientation from (100)teta. to (001)tetra. orientation with
increasing Sr contents from LNO (xsr = 0.0) to LSNO (xsr = 1.0), and (c) Off-normal XRD of
atetr.-axis-oriented LNO (10 3 )tetra. and ctetra.-axis-oriented LSNO (10 3 )tetragonal, GDC (2 0 2 )cubic,
and Y SZ (202)cub.ic...........................................................................................................................99
17
Figure 4-3. Sr content dependency of (a) strain energies (o-gray) of the La2 _-xSr.NiO4 z (LSNO)
thin films with 0.0 5 xsr 5 1.0 calculated using strain energy density equation, surface energies,
and (b) adsorption energies of an oxygen molecule on the (IOO)ttra. (o-blue) and (001)tetra. (o-red)
surfaces of the LSNO thin films calculated by density functional theory (DFT). The surface
oxygen adsorption sites of (c) the (100)tetra. slabs and (d) the (001)tetra. slabs. Details of the DFT
modeling approaches, the LSNO slab models, and strain energy calculation are provided in the
ESI. Both surface energies and adsorption energies represent the anisotropic feature of the LSNO
thin films......................................................................................................................................100
Figure 4-4. Electrochemical impedance spectroscopy (EIS) results of microelectrodes for La2
xSrxNiO4+- (LSNO) thin films with 0.0:5
xsr
5 1.0 at 550 0 C (a) Nyquist plot of the LSNO films
in 1 atm. (b) Nyquist plot of the LSNO thin film with xsr = 0.2 as a function ofp(O2). (c) Oxygen
partial pressure dependency of the surface exchange coefficients kq of LSNO thin films
calculated from EIS spectra collected at 550 0 C. Extrapolated bulk k* (approximately equivalent
to k )58 values obtained from previous data of (m-gray) Kilner et al., 79 (*-dark gray) and (*-light
gray) Boehm et al.,40 are plotted for comparison. (d) Oxygen partial pressure dependency of
volume specific capacitance (VSC) of LSNO thin films calculated from EIS spectra collected at
550 C ...........................................................................................................................................
102
Figure 4-5. Sr content dependency of the k of the La 2-xSrxNiO 4 8 (LSNO) thin films with 0.0 5
xsr 5 1.0 calculated from EIS spectra collected at 550 0 C in an oxygen partial pressure of 1 atm.
The schematic of ateta.-axis-oriented LNO thin film, both atetra.-axis and ctetra.-axis orientation
coexistence LSNO thin films, and ctetra.axis-oriented LSNO thin film are shown besides the
graph . ...........................................................................................................................................
103
Figure 4-6. AES data of the La2-xSrxNiO4+E(LSNO) thin films with 0.0 5 xsr 5 1.0 annealed at
550 *C in an oxygen partial pressure of 1 atm. (a) LamNN cation variation (RSF: 0.059), (b) SrLMM
cation variation (RSF: 0.027), and (c) NiLMM cation variation (RSF: 0.277) as a function of Sr
contents. The change of La and Sr cation spectra scales with the La:Sr concentration and
indicates that the surface chemistry is representative of the surface cationic ratio targeted. The
change in the Ni cation spectra represents the change in the thin film orientation according to
18
increase in NiO6 octahedra population (normalized by surface area) with increasing the Sr
content..........................................................................................................................................104
Figure S4-1. (a) Constrained (m-red, *-blue) and Relaxed (o-red, o-blue) lattice parameters of
the La2 _xSrxNiO4IZ (LSNO) thin films as a function of Sr content at room temperature.
Extrapolated bulk ateta. (V-gray) and ctetra. (A -gray) lattice parameters at room temperature
obtained from previous data of Gopalakrishanan et al.45 are plotted for comparison. The
constrained normal and in-plane lattice parameters of the LSNO films were calculated from
combining the interplanar distances of the (200)tetra., (1
0 3 )ttra.
and (006)tetra. peaks. (b) Out of
plane and in-plain strain as a function of the Sr content calculated using E..
(c-e)
=--
and
for in-plane strain and out of plane strain respectably. For determining the relaxed
Ezz =
film lattice parameter a and C, we used the equation:
(ce
_-2v
(a-fl)
=2v
1-V-va-,assuming
a/e 5 = 3.223, 3.305, 3.333, 3.339, 3.293, and 3.25 A for xsr = 0, xsr
'
eC
0.2, xsr = 0.4, xsr = 0.6, xsr = 0.8, and xsr = 1.0, respectably at 298 K, and v = 0.3. ................... 113
Figure S4-2. Schematic of two different orientations of La2 -x.SrxNiO4+6 (LSNO) on GDC (a)
(I00)tetra.-oriented epitaxial LSNO thin film and (b) (00 l)tetra.-oriented epitaxial LSNO thin films.
..................................
............................................................................
....................
...............
114
Figure S4-3. (a) Peak intensities and (b) d spacing of the La2.xSr.NiO 4 8 (LSNO) thin film
(00 6 )tetra. and (2 0 0 )tetra. as a function of Sr content obtained from HRXRD. The peak intensities of
(2 0 0 )tetra. significantly decreases with increasing the Sr content while those of (006 )tetra.
significantly increases, which suggests that once (00l)tetra. orientation growth begins, (l00)tetn.
orientation growth is suppressed..............................................................................................115
Figure S4-4. Electrochemical impedance spectroscopy (EIS) results of microelectrodes for the
La2-xSrxNiO4s (LSNO) thin films with 0 <
Xsr
5 1.0 at 550 0 C (a) Nyquist plot of the LNO thin
film with x=0, (b) Nyquist plot of the LSNO thin film with xsr = 0.4, and (c) Nyquist plot of the
LSNO thin film with xsr = 1.0 as a function of p( 0
19
2).
All films exhibited nearly perfect
predominant semicircle impedances, which indicates that the surface oxygen exchange kinetics
governs the oxygen electrocatalysis on the thin film surface.' ....................................................
116
Figure S4-5. AFM measurements of the as-prepared La 2-xSrxNiO 4 Es (LSNO) thin films with 0 <
xsr
5 1.0 deposited at 5,000 pulses (a) xsr = 0 with RMS of 0.341 nm, (b)
0.655 nm, (c) xsr = 0.4 with RMS of of 0.672 nm, and (d)
xsr =
0.8 with RMS of 0.611 un, and (f)
xsr
xsr =
XSr
= 0.2 with RMS of
0.6 with RMS of of 0.666 nm, (e)
= 1.0 with RMS of 0.374 nm. ............................... 117
Figure S4-6. Scanning electron microscopy (SEM) images of (a) La2NiO 4+8 (LNO) films, (b)
La2-xSrxNiO 4 k (LSNO) films with xsr = 0.2, (c) LSNO films with xsr = 0.4, (d) LSNO films with
xsr = 0.6, (e) LSNO films with xsr = 0.8, and (f) LSNO films with xsr = 1.0 annealed at 550 *C in
an oxygen partial pressure of 1 atm for 6 hours...........................................................................118
Figure S4-7. (a) Schematic of a LSNO/GDC/YSZ(001)/porous Pt samples and electrochemical
testing configuration (not drawn to scale), and (b) equivalent circuit (Ri = YSZ electrolyte
resistance, R 2 = electrode/electrolyte interface resistance1 9 , RORR = ORR resistance, CPE =
constant phase element) used to extract ORR kinetics, and (c) characteristic Nyquist plot
schematic (color key : orange = YSZ/bulk transport, green = GDC/interface, blue =
LSN O/O RR ). ...............................................................................................................................
119
Figure S4-8. Simulated La2-xSrxNiO4+8 (LSNO) slab models in the density functional theory
calculations in this work: (a) side view of the (001)tetra. slab, (b) side view of the (100)tetra. slab,
(c) top view of the top 3 layers of the (001)tetra. slab, and (d) top view of the top 3 layers of the
(100)tct. slab. The dotted circles in c and d represent the surface oxygen adsorption sites for the
(001)tefta. and (l00)tetra. surfaces, respectively. .............................................................................
120
Figure 5-1. AFM images of (a) as-deposited pristine LSC82 -85 nm, (b) LSC82 with -0. 1 nm
LSM82, (c) LSC82 with -0.3 nm LSM82, (d) LSC82 with -0.9 nm LSM82, (e) LSC82 with
-3.5 nm LSM82, and (f) LSC82 with -10 nm LSM82. RMS roughness values were in the range
of 0.74 - 1.12nm and comparable across all surfaces..................................................................131
20
Figure 5-2. X-ray diffraction (Cu Ka) analysis at room temperature. (a) Normal XRD of the
epitaxial LSC82 reference and the LSM82-decorated LSC82 films, (b) off-normal XRD of a
similarly prepared sample with a thicker (~10 nm) LSM82 coverage, and (c) schematic of the
crystallographic
rotational
relationships
among
the
LSM82(001),e,
LSC82(001)p,
G DC(001)cubic, and Y SZ(OOl)cubic................................................................................................134
Figure 5-3. Structural stability and strains of the epitaxial LSC82 with LSM82 coverage (-0.9
nm) thin film. (a) A full-range normal scan in the 6-26 Bragg-Brentano geometry from 150 *C to
550 *C, showing no phase change upon heating at a p(O2) of 1 atm. The starred (*) peaks
originated from the heater, and the peaks of the LSC82, GDC and YSZ are indexed to the pc (apc
~ 3.85 A5 2), cubic (ac ~ 5.42 A 65) and cubic (ac ~ 5.15 A 66) structure, respectively. (b) The inplane strains, Faa and (c) the out-of-plane strains, &,, of the epitaxial LSC82 and LSM82decorated LSC82 films as a function of temperature...................................................................135
Figure 5-4. Electrochemical impedance spectroscopy (EIS) results for the bare LSC82 films and
the LSC82 films with -0.1 (yellow), -0.3 (red), -0.9 (blue), -3.5 (green), and -10 nm (light
blue) LSM82 decorations at 550 *C. (a) Nyquist plot of the epitaxial LSC82 and the epitaxial
LSM82-decorated LSC82 films in 1 atm. Inset shows a magnification of the Nyquist plot of the
LSC82 films with partial LSM82 coverage (-0.1, -0.3, and -0.9 nm), (b) oxygen partial pressure
dependency of the surface exchange coefficients (kg) of the LSC82 and LSM82-decorated
LSC82 films calculated from EIS spectra collected at 550 *C, and (c) oxygen partial pressure
dependency of volume specific capacitance (VSC) of the epitaxial LSC82 and LSM82-decorated
LSC82 films calculated from EIS spectra collected at 550 *C. ...................................................
137
Figure 5-5. AES, SEM, and AFM analysis for bare LSC82 and LSM82-decorated LSC82 films
before and after annealing. Annealing was performed at 550 *C in an oxygen partial pressure of 1
atm. (a) Sr Auger spectra of bare LSC82 thin film probed for: as-deposited surface (gray), and
particles (dark red) and film surface (dark blue) after annealing. SEM image of (b) as-deposited
and (c) annealed LSC82. (d) Sr Auger spectra of LSC82 with -0.1 nm LSM82 probed for: asdeposited surface (gray), and particles (orange) and film surface (light blue) after annealing.
SEM image of (e) as-deposited and (f) annealed LSC82 with -0.1 nm LSM82. (g) Sr Auger
21
spectra of LSC82 with -0.3 nm LSM82 probed for: as-deposited surface (gray), and annealed
surface (blue). No particles were observed. SEM image of (h) as-deposited and (i) annealed
LSC82 with -0.3 nm LSM82. AFM images also showed particle formation on (j) annealed
LSC82, (k) annealed LSC82 with -0.1 nm LSM82, but no particles were observed on (1)
annealed LSC82 with -0.3 nm LSM82, (m) annealed LSC82 with -0.9 nm LSM82, (n) annealed
LSC82 with -3.5 nm LSM82, or (o) annealed LSC82 with -10 nm LSM82. ............................ 139
Figure 5-6. Surface exchange coefficients (k) of the LSM82-decorated LSC82 films calculated
from EIS spectra collected at 550 *C in a p( 0 2) of 1 atm and normalized cation intensity ratios
extracted from area mode using AES after annealing at 550 'C in a p(O2) of 1 atm. (a) kq (0gray) and normalized La and Sr intensity ratio (0-light blue) and (b) kq (0-gray) and
normalized Co and Mn intensity ratio (0-light red) as a function of LSM82 thickness............140
Figure 5-7. (a) Surface exchange coefficients (kq) of the LSC82 (*-black) and LSC82 with -0.9
nm LSM82 (*-blue) as a function of annealing time. (b) The energy of mixing for Lal-xSrxColyMnyO3 (LSCMO) based on the DFT calculations as a function of Mn concentration with Sr =
0.25 (*-light red), Sr = 0.5 (*-red), and Sr = 0.75 (*-dark red). LSCMO can be stabilized as
increasing M n concentration........................................................................................................142
Figure S5-1. X-ray diffraction (Cu K,,) analysis at room temperature. (a) Normal XRD and (b)
off-normal XRD of the epitaxial LSM82 reference film. Off-normal XRD shows the in-plane
crystallographic relationships between GDC and YSZ (a cube-on-cube alignment), and LSM82
and GDC (an in-plane 450 rotation with [IOO]pc LSM82 // [11 ]cubic GDC / [110 cubicYSZ). ... 151
Figure S5-2. In situ HRXRD data of (a) the normal scan of the LSC(002)p, peak and (b) the offnormal LSC(202),c as a function of temperature in a p(O2) of 1 atm. Here, we observe the peak
shifts towards lower angle in the 0-20 with increasing temperature from 25 *C to 550 'C.........152
Figure S5-3. A full-range normal scan in the 0-20 Bragg-Brentano geometry with the epitaxial
LSC82 with LSM82 coverage (-10 nm) from 150 *C to 550 *C, showing no phase change upon
heating at ap(02) of 1 atm. The starred (*) peaks originated from the heater, and the peaks of the
22
LSC82, GDC and YSZ are indexed to the pc (apc ~ 3.85 A52 ), cubic (ac ~ 5.42 A6 5) and cubic (a.
~5.15 A66) structure, respectively...............................................................................................153
Figure
S5-4.
(a)
Schematic
of a LSM/LSC/GDC/YSZ(001)/porous
Pt
sample
and
electrochemical testing configuration (not drawn to scale), and (b) equivalent circuit (R1 = YSZ
electrolyte resistance, R2 = electrode/electrolyte interface resistance4 3 , RoRR = ORR resistance,
CPE = constant phase element) used to extract ORR kinetics, and (c) Nyquist plot of the
epitaxial LSC82 with -0.3 nm LSM82 coverage at 550 0 C; inset shows a magnification (HF: 10 4
~ 10 5 Hz, M F: 10 3 ~ 10 4 Hz, and LF: 10-2 ~ 10 3 Hz)...................................................................154
Figure S5-5. Nyquist plot of the epitaxial LSM82 thin film as a function of oxygen partial
pressure at 550 0 C. EIS data of the LSM82 was found to show the p(0 2)-dependent impedance
responses, which suggest that the oxygen surface exchange kinetics governs the oxygen
electrocatalysis on the film surface. .....................................................
............. ......................... 155
Figure S5-6. AES data using area mode of the epitaxial LSC82 with and without LSM82
coverage annealed at 550 *C in an oxygen partial pressure of 1 atm. (a) LaMNN cation variation
(RSF: 0.059), (b) SrLMM cation variation (RSF: 0.027), (c) COLMM cation variation (RSF: 0.076),
and (d) MnLMM cation variation (RSF: 0.161) as a function of LSM82 coverage. ..................... 156
Figure S5-7. The bulk supercells used to calculate the Sr-La swapping energy of Lal.xSrxCoO3-8
and Lai.xSrxMnO 3 .8 (a) x=0.25 and (b) x=0.375. The dark blue octahedra, light blue, green and
red spheres represent Co/Mn centered octahedra, Sr, La, and 0 ions, respectively....................157
Figure S5-8. Time dependent surface exchange coefficient (k') of bare LSC82 (0-black),
(
LSC82 with LSM82 coverage -0.9 nm (*-blue), and LSC82 with LSC214 coverage ~2.5 nm
-yellow ).....................................................................................................................................158
Figure 6-1. X-ray diffraction (Cu K,) analysis at room temperature. (a) Normal XRD of the
epitaxial LSCF11 3 reference and the LSC 214-decorated LSCF113 films, (b) off-normal XRD of a
similarly prepared sample with a thicker (~5 nm) LSC2 14 coverage, and (c) schematic of the
23
crystallographic
rotational
relationships
among
the
LSC 2 14 (001)teta.,
LSCFl 3(001),,
GDC(001)cubic, and Y SZ(001)cubic................................................................................................172
Figure 6-2. Electrochemical impedance spectroscopy (EIS) results for the bare LSCF1 3 film (0
-black), LSC113 film (0-gray), LSCF 11 3 films with -0.3 (0-red), -0.8 (0-orange), -2.6 (0yellow), and -5 nm (0-green) LSC214 decorations, LSC113 film with -2.6 nm (0-blue) LSC 2 14
-
decoration at 550 'C. (a) Nyquist plot of the epitaxial LSCF,1 3 and the epitaxial LSC 2 14
decorated LSCF113 films in 1 atm. (b) Nyquist plot of the LSCF,1 3 thin film with -5 nm LSC 214
coverage as a function of p(O2). Inset shows a magnification of the Nyquist plot in 1 atm. (c)
,
Oxygen partial pressure dependency of the surface exchange coefficients (k) of the LSCF113
LSC1 1 3, LSC 214-decorated LSCF 1 3 , LSC 214-decorated LSC113 fihns calculated from EIS spectra
collected at 550 *C. Extrapolated bulk k* (approximately equivalent to kq) values at 550 *C
obstained from previous data of (*-light blue) Steele et al.56 and (*-light blue) De souza et al.6
are plotted for comparison. (d) Oxygen partial pressure dependency of volume specific
-
capacitance (VSC) of the epitaxial LSCF113 , LSC1 1 3, LSC 214-decorated LSCF11 3, and LSC2 14
decorated LSC11 3 films calculated from EIS spectra collected at 550 *C....................................174
Figure 6-3. Auger electron spectroscopy (AES) data from area mode for the LSCF11 3 (black),
LSC11 3 (gray), LSCF,1 3 film with -2.6 nm LSC214 decorations (yellow), and LSC 1 3 film with
-2.6 nm LSC 214 decoration (blue) after annealing. Annealing was performed at 550 *C for 8
hours in an oxygen partial pressure of 1 atm. (a) La Auger spectra and (b) Sr Auger spectra of the
LSC 1 3 and LSC214-decorated LSC11 3 films. (c) La Auger spectra and (d) Sr Auger spectra of the
LSCF 1 3 and LSC2 14-decorated LSCF,1 3 films. (f) Normalized La and Sr intensity ratio extracted
from AES of the LSCF,1 3 films with and without -2.6 nm LSC 214 decorations (yellow), and
LSC1 1 3 film with and without -2.6 nm LSC214 decoration (blue). Details of normalization
methods are provided in the Supporting Information..................................................................176
Figure 6-4. (a) The calculated SrLa substitution energies in bulk LSC2 14, LSC, 13 , and LSCFn3
(all relative to that of LSC113, which is set to 0). The error bars shown in the LSCF113 and
LSC 2 14
represent the upper bound and lower bound of the SrLa substitution energies from the
sampled A-site and B-site cation arrangements...........................................................................178
24
Figure 6-5. (a) The relative ratio of the surface exchange coefficient kq (with respect to the kq of
the base film) vs. the calculated bulk 0 2p band center shifts (relative to the Fermi level)
between the LSC2 14-decorated LSCF,1 3 (LSC 1 3 ) film surfaces and the undecorated LSCF113
(LSC1 1 3) film surfaces based on the surface Sr information from COBRA analysis 63 and DFT
modeling predictions. The inset shows the Sr content vs the calculated bulk 0 2p band centers of
LSC11 3 and LSC2 14 . (b) k of the LSC11 3 with and without LSC 214 coverage and (c) the LSCF113
with and without LSC 214 coverage as a function of annealing time. Annealing was performed at
550 'C in an oxygen partial pressure of 1 atm.............................................................................
182
Figure 6-6. Auger spectra and atomic force microscopy (AFM) images for bare LSC113,
LSCF113, LSC214-decorated LSC113, and LSC214-decorated LSCF113 thin films. (a) Sr Auger
spectra for: LSC113 (gray), LSC214-decorated LSC13 (blue), LSCF113 (green), and LSC214decorated LSCF13 (yellow) after annealing at 550 oC for 70 hours in an oxygen pressure of 1
atm. The dashed orange line is the Sr spectra of a pristine LSC214 reference sample. The peak-to
peak values in Auger spectra reflect the Sr concentrations. AFM images of as-deposited (b)
LSC113, (c) LSC214-decorated LSC113, (d) LSCF13, and (e) LSC214-decorated LSCF113. AFM
image showed particle formation on (f) 6 h annealed LSC13 but no particles were observed on
(g) 6 h annealed LSC214-decorated LSC113, (h) 6 h annealed LSC214-decorated LSC113, and (i) 6 h
annealed LSC214-decorated LSC113. After annealing for 70 h, particles were observed on all
surfaces; (j) annealed LSC113, (k) annealed LSC214-decorated LSC113 (1) annealed LSCF113 and
(m) annealed LSC214-decorated LSCF113.....................................................................................185
Figure S6-1. AFM images of (a) as-deposited pristine LSCF11 3 -63 nm, (b) LSCF11 3 with -0.3
nm LSC2 14 , (c) LSCF113 with -0.8 nm LSC2 14 , (d) LSCF11 3 with -2.6 nm LSC 214, and (e)
LSCFl 3 with -5 nm LSC 214. RMS roughness values were in the range of 0.24 - 0.32 nm and
com parable across all surfaces.....................................................................................................192
Figure S6-2. X-ray diffraction (Cu Ka) analysis at room temperature. (a) Normal XRD of the
epitaxial LSC11 3 reference and the LSC 214 -decorated LSC113 film, (b) off-normal XRD of a
similarly prepared sample with a -2.6 nm LSC 214 coverage, and (c) schematic of the
25
crystallographic
rotational
relationships
among
the
LSC 2 14 (001)tetra.,
LSCi 3 (001)pc,
GDC(001)ubic, and Y SZ(001)cuic................................................................................................193
Figure S6-3. (a) Relaxed lattice parameters of both LSCF11 3 and LSC11 3 as a function of LSC 2 14
thickness, calculated from HRXRD data. (b) In-plain and out-of-plane strains of both LSCFi1 3
and LSC113 as a function of LSC2 1 4 thickness, calculated from HRXRD data............................194
Figure S6-4. (a) Schematic of a LSC2 ,4/LSCFu 3 or LSC11 3/GDC/YSZ(001)/porous Pt sample
and electrochemical testing configuration (not drawn to scale), and (b) equivalent circuit (Ri =
YSZ electrolyte resistance, R2 = electrode/electrolyte interface resistance4 , RORR = ORR
resistance, CPE = constant phase element) used to extract ORR kinetics, and (c) Nyquist plot of
the epitaxial LSCF113 with -2.6 nm LSC2 1 4 coverage at 550
(HF:
104
0 C;
inset shows a magnification
~ 10' Hz, MF: 103 ~ 104 Hz, and LF: 102 ~ 103 Hz)....................................................195
Figure S6-5. Nyquist plot of the epitaxial LSC214-decorated LSC11 3 thin film as a function of
oxygen partial pressure at 550 0 C. EIS data of the LSC 214-decorated LSC113 was found to show
the p(0 2)-dependent impedance responses, which suggest that the oxygen surface exchange
kinetics governs the oxygen electrocatalysis on the film surface.
..........................
.................. 196
Figure S6-6. Sr Auger spectra for the undecorated LSC11 3 thin film after annealing at 550 OC
for 6 hours. Observed particles on the surface of LSC113 shows higher Sr peak intensity
compared to the rest of the film surface.......................................................................................197
Figure S6-7. Sr Auger spectra for the undecorated LSC13 thin film after annealing at 550 oC for
6 hours. Observed particles on the surface of LSC113 shows higher Sr peak intensity compared to
the rest of the film surface. ..........................................................................................................
198
Figure S6-8. Simulated LSCF13 and LSC214 configurations for calculating the energies of Sr
substitution
of La
(SrLa)
in
LSCFi13
and
LSC214 (a)
Lao.625Sro.375Feo.7sCoo.2503
Lao.sSro.sFeo.75Coo.2503 (with an additional Sr in the simulated 2x2x2
supercells),
(b)
(c)
(Lao.sSro.s)2CoO4, and (d) (Lao.4375Sro.5625)2CoO4. Elements are represented as: La (green), Sr
26
(light blue), Fe (brown, center of the octahedra), and Co3+(dark blue, center of the octahedra). 0
ions are located at the corners of all the octahedral.....................................................................199
Figure S6-9. (a) Schematics of the heterostructured interfaces with various A-site and B-site
arrangements in the DFT simulations. Lao.625Sro.375Feo.75Coo.2503 represents the LSCF3 phase
and (Lao.5Sro.5)2Co04 represents the LSC214 phase. Elements are represented as: La (green), Sr
(light blue), Fe (brown, center of the octahedra), and Co3+(dark blue, center of the octahedra). 0
ions are located at the corners of all the octahedra. The AO planes are numbered from Al
through A6 in the LSCF13 and BI through B4 in the LSC214 phase. The planes Al, B1 and B2
represent an interfacial region. The relative stability of SrLa substitution energy relative to
Lao.75Sro.25Co03, or E(SrLa)-E(Srta) of LSC13(25%Sr), with variation in the SrLa defect position
across the AO planes. Values are relative to a bulk LSCt13(25%Sr) reference (y=0). Also shown
is a dotted horizontal line representing the SrLa substitution energies for the bulk LSCF13(green
dotted line, E(Sra)LSCFi3- E(SrLa)LSC113(25%Sr)), and a black dash-dotted line for the bulk
LSC214(or E(SrLa)LSC214
-
E(SrLa)LSC13(25%Sr)). Note that the more negative values on the y
axis correspond to the easier substitution of the SrLa relative to bulk LSC113(25%Sr)................201
Figure S6-10. Bulk LaO. 7 5SrO. 2 5CoO 3 phase diagram at T=823 K and P = 1 atm. We note the
chemical potential of 0 is fixed by setting T and P, while the DFT total energy of
Lao. 75SrO. 25 Co0 3 provides a constraint for the three (effective) metal chemicals so that only two
(effective) metal chemical potentials are needed to construct the phase diagram. The two
independent effective metal chemical potentials are represented by AMC(LaUA
7 SrOMCoO3 ) and
APs (La0 75Sr 25CoO3 ) , where Ap
Ap
7(La
0 *Sr.25CoO
3)
=M
(La .,Sr.2sCoO3 )
= pg
4
(LaO 7 5Sr.CoO3 ) - ptj(Co3 O4 ) and
0(La
75Sr.2 5CoO3 ) - P(SrO). The shaded area in the phase diagram
represents the stable region for bulk Lao.7 5Sro.25Co0 3 vs. the LaCoO 3 , SrCoO 2.5, C0 304 , SrO, and
La2 0 3 oxides, based on the inequality equations - Equations (22)-(26) using the effective
chemical potentials of metals of Lao. 75SrO. 25Co0 3 , LaCoO 3, SrCoO 2.5, C030 4 , SrO, and La 20 3...
...........................
...............................................
..........................................................................
2 03
Figure S6-11. The LSC1 1 3 and LSCF1 1 3 (001) slab models used for the ab initio surface
27
thermodynamic analysis. Green and light blue spheres represent La and Sr, while brown and
deep blue polyhedral represent local Fe-O and Co-O environments, respectively. The top (and
bottom) two surface layers, where La/Sr and Co/Fe compositions (La.XSrCo.yFeyO3 with x=0,
0.25, 0.5, 0.75, 1 y=0, 0.25, 0.5, 0.75, 1) are varied, are specified by the rectangular frames. The
central part of the slabs, outside the frames, are fixed to a composition close to Lao.75 SrO. 2 5CoO
3
and Lao.62 5SrO.3 7 5Feo. 7 5Coo.2 5 0 3 . A total of 10 configurations (5 for the (001) AO surfaces and 5
for the (001) B0 2 surfaces) for LSC113, and a total of 50 configurations (25 for the (001) AO
surfaces and 25 for the (001) B02 surfaces) for LSCF113 , are calculated based on these 9-layer
2x2 symmetric slab models for the surface stability analysis. ....................................................
204
Figure S6-12. The predicted Lao.7 5 Sro.2 5CoO 3 surface stability diagram at T = 823 K and P = 1
atm based on the chemical potentials of bulk Lao. 75Sro.2 5CoO3 . The grid points represent the
sampled bulk effective chemical potentials of Sr (x-axis; x=0 represents the equilibrium between
Lao.75 Sro.2 5CoO 3
and
SrO) and
Co (y-axis;
y=0
represents
the
equilibrium
between
Lao.7 5SrO.2 5CoO 3 and C030 4 ) in Lao.7 5Sro. 2 5CoO 3 , and the contour plot beyond the grid is
constructed based on the
calculated lowest
surface energy among
the investigated
Lao.7 5SrO.2 5CoO 3 (001) surface configurations. ............................................................................
28
205
List of Tables
Table S4-1. Lattice mismatch between the film materials and the substrate materials. ............ 110
Table S5-1. Constrained and relaxed lattice parameters of LSC82 and LSM82-decorated LSC82
films extracted from normal and off-normal XRD data as a function of temperature. Constrained
normal and in-plane lattice parameters of all films were calculated from combining the interplanar distance of the LSC(002),c and LSC(202)pc peaks. ..........................................................
29
145
This page is intentionallyleft blank
30
Chapter 1.
Introduction & Background
1.1 Motivation
Over the last several decades, the energy generation of the world has mainly relied on
fossil fuels. According to the U.S. Energy Information Administration (EIA), International
Energy Outlook 2013, in 2010 fossil fuels provided 84% of the about 18 TeraWatts (TW) of
energy used worldwide. Due to the growth of population and average income, estimated energy
demand in 2040 will increase by up to 130 % compared to 2010 (Figure 1-1). Hence, it is
expected that the global energy generation will continue to depend on fossil fuels for the next
few decades. 1 However, there is presently a scientific consensus about the finite supply of fossil
)
fuels 2 (Figure 1-2), the environmental damage caused by emissions of carbon dioxide (C0 2
from fossil fuels,3 and the low energy efficiency generated from fossil fuels 4 (Figure 1-3). In
response to the demanding need for clean and reusable energy sources, some potential solutions
have emerged such as conserving energy through improving energy efficiency, reducing fossil
fuel consumption, and increasing the supply of environmentally friendly energy sources. One
potential solution to aid in improving the energy conversion efficiencies and generating the
renewable energy without environmental damage is fuel cells.
world energ consumption
quadrilion Btu
Projections
Histody
400
Other Non-OECD
uOECD
a Non-OECD Asia
300
200
100
0
1990
2000
2020
2010
2030
2040
Figure 1-1. A graph provided by U.S. Energy Information Administration (EIA), International
Energy Outlook 2013 depicting the increase in energy use of the world.
31
COrntinal Fossl Fuels
o I~wU
G
.alrN..20t)
9 CoM M!egWs~ft
T~ee,06
Ic
low0
Qu. 207)
am
'a
00W
4M0
is00
t40
1960
1940
2000
1110
202D
200
200
20W
2100
Figure 1-2. A graph provided by the Association for the Study of Peak Oil and Gas (ASPO)
depicting the production of oil, coal, and gas natural resources in million tonne of oil equivalent
(Mtoe) since the 1900's and projected to 2100, suggesting a peak in production in the 2010
decade.
Fuel cells are efficient and environmentally clean energy conversion devices, which can
convert the chemical energy of a fuel gas directly into electrical energy.6-8 Fuel cells are not
limited by the conventional Carnot efficiency governing thermal-mechanical conversion systems.
70
operation with natural gas
80compwe
50
PEMFC-
40-
wE
wer planIt
?team p~owr plant
-
Cwbturettoi e ngine
30-
>-gas turbine"
20
-
W
cyle
-
.- I
Il
fuel cells
10
0
combustion engines
I
10
i
I
I
I
I
100
1,000
10.000
100.000
1,000,000
Plant Power (kW)
Figure 1-3. A graph provided by The Green Box Systems Group 5 showing the comparison of
electrical efficiency between fuel cells and conventional combustion engines.
32
This is due to the fact that chemical energy of a fuel is directly converted into electricity, without
intermediate conversion into heat, as in conventional power schemes. Hence, theoretical fuel cell
efficiencies reach 83
%.7
In particular, solid oxide fuel cells (SOFCs), based on an oxide-ion
conducting electrolyte, have several advantages over other types of fuel cells, including
relatively inexpensive materials, low sensitivity to impurities in the fuel, and high efficiency. 9-12
1.2 Solid Oxide Fuel Cells
SOFC is characterized by having a solid ceramic electrolyte (hence the alternative name,
ceramic fuel cell), which is a metallic oxide. The basic components of the SOFC are the cathode,
at which oxygen is reduced to oxygen ions (Oxygen reduction reaction, ORR), which then pass
through the solid electrolyte under electrical load, to the anode, where they react with the fuel
excess of 80%.8 The primary components of a fuel cell are an ion conducting electrolyte, a
cathode, and an anode, as shown schematically in Figure 1-4.13
02/ Air
Cathode
Electrolyte
Anode
H 2 /fuel
Figure 1-4. Schematic of a fuel cell operating on oxygen at the cathode (in blue), and hydrogen
at the anode (in green) separated by the oxygen ion-conducting electrolyte (in orange).
33
There is an overall chemical driving force for the oxygen and the hydrogen to react to produce
)
water. Direct chemical combustion is prevented by the electrolyte that separates the fuel (H 2
from the oxidant (02). The electrolyte serves as a barrier to gas diffusion, but will let ions
migrate across it. Accordingly, half cell reactions occur at the anode and cathode, producing ions
which can traverse the electrolyte. For example, if the electrolyte conducts oxide ions, oxygen
will be electro-reduced at the cathode to produce
02
ions and consume electrons. Then, oxide
ions will react at the cathode with hydrogen and release electrons after migrating across the
electrolyte.
0.7
1 =A/cm
0.6
2
Anode loss
0. 0.5
0
O
0.4
Cathode loss
WU
t;-
0
0.3
62% 4 Due to t e slow reaction kinetic
0.2
0.1.
Electrolyte loss
04
550
600
650
700
TEMPERATURE
750
0
( C)
800
850
Figure 1-5. Voltage losses as a function of SOFCs operating temperature representing the
highest voltage losses from the cathode due to the slow ORR kinetics.
In solid electrolyte, there are relatively few solid materials that have ionic conductivities
comparable to that of liquid electrolytes. The oxide oxygen ion conductors all have thermally
activated conductivities, with values only approaching that of liquid electrolytes at very high
temperatures (>800 C). However, the elevated operating temperature of the SOFC has a number
of consequences which are reduce costs, particularly of interconnect, manifolding and sealing
materials. Therefore, there is considerable interest in lowering the operating temperature of
34
smaller SOFCs. Reducing the operating temperature of the SOFC to less than 600 *C would offer
many advantages including the use of low-cost metallic materials such as ferritic stainless-steels
for the interconnection and construction materials. This makes both the stack and balance-ofplant cheaper and more robust. Rapid start-up and shut-down operations are another benefit of
reduced operating temperature. Lastly it reduces the corrosion rates significantly."
On the other hands, this would accompany a serious problem, total voltage loss, and it
would cause a significant drop of the fuel cell efficiency because lower ionic and electronic
conductivity, slower reaction kinetics and changes in surface/interface properties or reaction
pathway will occur at lower temperature. 15 Figure 1-5 shows a resistance values as following
different operating temperatures. The total resistance includes a cathode resistance, an anode
resistance and an electrode resistance (ohmic resistance). The cathode resistance is almost half of
a total resistance and therefore the cathode performance should be improved.
1.3 Cathodes of Solid Oxide Fuel Cells
For an application as SOFC cathode, a material has to exhibit several general properties:9
high electronic conductivity (preferably more than 100 S cm 1 under oxidizing atmosphere), a
matched thermal expansion coefficient (TEC) and chemical compatibility with the electrolyte
and interconnect materials, adequate porosity to allow gaseous oxygen to readily diffuse through
the cathode to the cathode/electrolyte interface, stability under an oxidizing atmosphere during
fabrication and operation, high catalytic activity for the oxygen reduction reaction (ORR), and
low cost.
Transition metal oxides with perovskite structure such as Lai.xSrxMnO 3-s (LSM)16- 2 0 with
high electronic conductivity but low ionic conductivity are commonly utilized as a cathode
material for SOFCs at high temperature such as 1000 *C. Reducing the operating temperature is
of vital importance to reduce the degradation and improve the lifetime of these solid-state
devices. Therefore, many efforts at developing suitable cathodes for intermediate temperature
SOFCs have focused on developing new materials due to the difficulty of meeting the
requirements for fast surface exchange kinetics and stability with a LSM cathode material.
Mixed ionic and electronic conductors (MIECs) possess a high surface oxygen exchange
kinetics, which can allow the oxygen reduction/oxidation reaction to take place on the entire
35
oxide surface, increasing the active surface area and providing higher electrocatalytic
performances. 22
1.4 Mixed Ionic and Electronic Conductors (MIECs)
The concept of MIECs is to combine ionic
(02~)
and electronic conductivity. In general,
two reactions can occur at the cathode, depending on if an interstitial or vacancy ionic
conductivity mechanism is involved (Kriger-Vink notations) 23 :
02(g) + 2e ->O+
(1-1)
or
10 2 (g)
(1.2)
+ V0 + 2e- -+ 01
These reactions represent the reduction of ambient oxygen at the cathode/gaseous oxygen
interface, including the transfer to the electrolyte. As mentioned above, the MIECs can extend
the active surface areas in which the kinetics can be enhanced due to a double interface (DI)
/
gaseous oxygen / mixed conductor / electrolyte relative to the threefold contact cathode
electrolyte / gaseous oxygen. Hence, the increased kinetics should accompany a decrease in the
SOFC operating temperature. For the ORR in the MIECs electrodes, the characteristic active
width is limited by the oxygen surface exchange and soli state diffusion.6, 24 , 25 Therefore, the
cathode performance strongly relies on the oxygen self-diffusion coefficient (D*) and surface
exchange rate (k*).
MIECs such as La.SrxCoO3-S
(LSC)2 4 ,
26-28
and Lai-xSrxCol-yFy03.8
(LSCF)2 9 -3 2
perovskite oxides, where vacancy ionic conductivity mechanism is involved in the ORR reaction,
have been extensively studied to promote the oxygen surface exchange kinetics at intermediate
temperature (500 - 700 *C). However, strontium (Sr) segregation, which can hamper the surface
exchange kinetics, can occur on the surface of such MEICs, 33 -3 6 and therefore the surface
stability of MIECs perovskite oxides has been widely studied to prevent the activity fading
during operation. Recently, layered MIECs such as Ruddlesden-Popper (RP) oxides,3 7
42
where
the reaction is governed by interstitial oxygen, have attracted much attention as new cathode
36
materials for intermediate temperature SOFCs. In addition, many efforts at developing suitable
cathodes for intermediate temperature SOFCs have also focused on developing new approaches,
such as surface modification4 3 5~1 on the MIECs surface and composite cathode materials, 52-55 due
'
to the difficulty of meeting the requirements for fast surface exchange kinetics and stability2
with a single ABO3 perovskite oxide.
This work aims to enhance our understanding and develop highly active oxide cathode
materials for ORR at an intermediate temperature, using RP oxides40 ' 4 1and surface modification
on the MIECs surface.44 '56-58
1.4.1 ABO 3 Perovskite Oxides
ABO 3 perovskite oxides are widely used in electrochemical applications such as SOFCs
and oxygen permeation membranes due to their good reactivity and flexibility not only in terms
of oxygen stoichiometry but also with regards to A and B cation substitutions. For most of the
perovskite materials used as cathodes in SOFCs, the A-site cation is a mixture of rare and
alkaline earths (such as La and Sr, Ca or Ba), while the B-site cation is a reducible transition
metal such as Mn, Fe, Co, or Ni (or a mixture). The atomic structure of perovskites consists in a
3D network of B06 octahedra connected by apex, which are maintained by a cubic lattice of A
atoms of coordination 12. The ideal atomic arrangement, usually adopted at high temperatures, is
shown in Figure 1-6. The symmetries of ABO 3 perovskites can theoretically be directly
determined from the ionic radii of A and B cations, which govern strains inside the materials,
since the size mismatch is responsible for the distinct stacking.
0
B: Transition metal
C
b
A: Rare earth
Figure 1-6. ABO 3 perovskite structure, where A is rare earth, B an transition metal.
37
In the early stage, lanthanum cobaltite1 0 , 24 , 59 (LaCoO 3) and lanthanum manganiteO,
60
(LaMnO 3) based ABO 3 perovskite oxides were utilized as a cathode material for SOFCs. These
stoichiometric compositions, however, were found unsuitable in terms of electrochemical
performance for a cathode. Usually, the increase of ionic conductivities is more influenced by Sr
concentration at the A-site while the increase of the electronic conductivities is more influenced
by Fe and Co concentration at the B-site. The substitution of Sr2 + for La3+ can enhance the
oxygen ion transport kinetics due to an increase in the oxygen vacancy concentration in the
perovskite structure. 61 Therefore, high electrochemical performance at the cathode could be
achieved relative to undoped ABO 3 perovskite oxides. 62 , 63 The Sr substitution can also influence
the TEC of the cathode, where higher Sr substitution leads to higher TEC. Indeed, the Sr
segregation, which hampers the oxygen surface exchange kinetics and long term stability, also
arises from the Sr substitution. 33 -3 6 The effect of substitution of B-sties by Fe is to aid in
enhancing the TEC of the cathode materials whereas it leads to a reduction in the ionic
conductivity.3 1 ' 3 2
Lai.xSrxCoO 3-8 (LSC) 2 4 , 26-28 and Lai~xSrxCoi~yFyO3-8 (LSCF) 2 9- 3 2 perovskite oxides are
known to show oxygen deficiency, and the oxygen nonstoichiometry (8) affects the
electrochemical properties, conductivity and lattice expansion of the materials. In the case of
LSC, the 8 values increase with increasing temperature, increasing Sr content, and decreasing
oxygen partial pressure (p(02)). 2 6, 6 1, 64 , 65 However, it has been shown by Mizusaki et al.26 that
the log p(O2) dependence of 8 becomes weaker with increasing 8 for 8 > 0.01. In general, the
valence states of Co are in the range of 2+ to 4+, while those of La (3+), Sr (2+), and 0 (2-) are
fixed. Therefore, when a portion, x, of La3 +in LaCoO 3 is replaced by Sr2 + to form Lai..SrCo038, electroneutrality is maintained by both a decrease in oxygen content and an increase in mean
cobalt valence, which can be expressed by n (cobalt valence) = 3 + x - 28.26,64,65
The influence of B-site cations on the 8 of LSCF have also been reported by several
authors. 3 0, 66 Hashimoto et al. found that the 8 value of LSCF depends on the Co content and
significantly differed for Co-rich (Co = 0.2 and 0.4) LSCF and Fe-rich LSCF (Fe = 0.6 and
0.8).66 According to the authors, Co-rich LSCF can be decomposed at relatively high p(O2),
where oxygen content rapidly decreased independently of p( 0 2), and the corresponding 8 as a
function of p(0 2) can be approximated by a vertical line. In contrast, the oxygen vacancy
38
concentration increases steadily for the Fe-rich LSCF. More recently, Kuhn et al. have shown
that The 8 of the Co-rich LSCF can be reasonably well fitted with a metallic defect chemical
approach previously used for LSC whereas a semiconductor model previously established for
LSF best fits the data of the Fe-rich LSCF. The semiconducting character becomes dominant
with increasing Fe content.30
1.4.2 Ruddlesden-Popper (RP) Oxides
ABO
3
MO,
.4
perovskite layers
LaO.
rock salt layers
AO layer
n
A : Rare earth
4
<4
|b
a
.B: Transition metal
4
n=1
I
n=2
n=3
Figure 1-7. Schematic crystal structures of the n = 1, 2 and 3 members of the RuddlesdenPopper (RP) type An+1 Bn03n+1 are shown. The denotation of n represents the number of stacked
octahedral layers separated by a rock salt LaO layer. NiO 6 octahedra and La atoms are shown
schematically.
39
The general formula of Ruddlesden-Popper (RP) phases can be written as An+,Bn03n+l.67
The RP phases are comprised of n consecutive perovskite layers (ABO 3) alternating with rocksalt layers (AO) along the crystallographic c-axis direction. Their formula can be represented by
(AO)(ABO 3)n, where n represents the number of connected layers of vertex sharing B0 6
octahedra.3 , 38 Figure 1-7 presents the ideal tetragonal unit-cells for n = 1, 2, and 3, which
correspond to the stoichiometric compounds all of same space group, I4/mmm. For n > 1, the
additional ABO 3 blocks are introduced between two AO rock-salt layers. Commonly these
materials consist of rare or alkaline earth A site cations with transition metals on the B site,
forming an extensive series of compositions. Similar to the ABO 3 perovskite oxides, RP phases
show a rather high structural flexibility in the oxygen stoichiometry. Of particular interest in the
case of SOFCs are the n = 1 members of some RP series, notably those based on La 2NiO 4 +8
(LNO), which have been shown to accommodate oxygen interstitials in the AO layers, 68 giving
rise to fast ion conduction and hence potential application as cathodes at intermediate
temperatures. In this thesis, we have focused on La2-xSrxNiO4 8 (0.0 5 xsr 5 1.0) oxides, which
belong to the n = 1 members of the RP oxides.
1.5 La2 -xSr.NiO 4 8 (LSNO, 0.0 5
XSr
5 1.0) Oxide
1.5.1 Structure of La2-xSr 1NiO 4
The LNO (xsr = 0) structure can be described as the intergrowth of alternative NiO 2 and
La2 0 2 rock salt layers or layers of LNO separated by LaO layers, having in-plane lattice
parameters similar to that of perovskites in the pseudocubic notation. The oxygen stoichiometric
La2NiO 4 can be easily oxidized with additional interstitial oxygen incorporated into lanthanum
tetrahedra in the La2 0 2 rock salt layer, which results in increasing the La-O bond distance and in
decreasing the nickel ionic radius by the formation of Ni .69' 70 The alternating perovskite and
rock salt layers in the RP structure can lead to different oxygen diffusion and surface exchange
kinetics along different crystallographic directions. For example, the in-plane oxygen transport
kinetics via interstitial oxygen diffusion in the LaO planes is considerably higher by up to two
40
orders of magnitude compared to the out-of-plane kinetics 3 9 and the surface exchange kinetics
in-plane can be an order of magnitude greater than that out-of-plane for LNO single crystals. 37
LNO in the Ruddlesden-Popper structure is known to be orthorhombic with space group
Fmmm at room temperature for an oxygen overstoichiometry ()
~ 0.18 as reported by Jorgensen
et al. 68 As the Ruddlesden-Popper structure can be influenced greatly by the oxygen
stoichiometry, fully oxidized La2NiO 4.25 exhibits a monoclinic symmetry with space group C2 at
room temperature7 1 while fully stoichiometric La2NiO 4 is orthorhombic with space group
Bmab. 72 In our study, large experimental uncertainty exists in the inferred oxygen
nonstoichiometry of LNO thin films from the literature3 7 using the lattice parameters determined
from HRXRD and the XRD data of single-crystalline thin films do not allow a precise
refinement of the symmetry of the LNO structure. Therefore, we chose the tetragonal symmetry
with space group I4/mmm, which differs from the orthorhombic by the loss of the octahedral
tilting (5 1*) in this thesis.
With xsr in La2 -xSrxNiO4 s (LSNO) greater than 0 and equal to 1.0, LSNO oxides are well
known to have tetragonal symmetry with space group 14/mmm.73,
74
By substituting of a larger
cation such as Sr2 + (1.31 A) for La 3+ (1.22 A) in LNO, the structural stresses are released, which
leads to decrease in the amount of additional oxygen required to stabilize the structure and thus
results in the reduction of the oxygen diffusion coefficients.7 5'
76
Gopalakrishnan et al.7 3 have
explained the trends in the lattice parameters of LSNO in terms of a Jahn-Teller distortion caused
by orbital ordering in the low-spin Ni 3 + cation. The observation that atetragonal decreases while
Ctetragonal
increases up to xsr=0. 6 has been taken to indicate that the distortion of the low spin Ni 3+_
o octahedra produces four short (equatorial) and two long (axial) bonds and that the single eg
electron is ordered in the d,2 orbital. However, the extent of the distortion decreases quite
dramatically as Sr increases to 1.0. Further increase in the amount of low spin Ni3 + as Sr
increases from 0.6 to 1.0 would be expected to produce more distorted octahedra if the electrons
in the system remain localized. The substitution of Sr for La in LSNO and the subsequent Ni
oxidation have been shown to induce electron holes, which can reduce the negative net charge in
the NiO 2 planes. 77-79 According to the tilting of the NiO octahedra caused by the oxidation of
Ni 2 + into Ni 3 +, a lattice decreases while c lattice increases.
41
1.5.2 Oxygen Nonstoichiometry (8) of La 2 -xSr 1NiO 4
8
,
37 7 6 80
Several authors have reported that LNO shows large oxygen excess composition. , ,
81
Interstitial oxygen provides facile oxide ion conduction and it exists at the center of La
pseudo-tetrahedron. 68 Incorporated oxygen becomes a major ionic career, and holes are
simultaneously created to maintain charge neutrality. The amount of nonstoichiometric oxygen
content can significantly affect both ionic and electronic conductivity of LNO. Therefore, 6 can
play an important role in the electrochemical properties. The 8 of LSNO is strongly dependent on
the Sr content.42 , 82 Nakamura et al.42 have reported that the LSNO with 0.0 < xsr 5 0.1 showed
only the oxygen excess, while the LSNO with
xsr=O
and 0.2 shows both oxygen excess in the
higher p(O2) region and the oxygen deficiency in the lower p(O2) region. Moreover, LSNO with
xsr=0. 4 shows only the oxygen deficient composition but not the oxygen excess composition.
Vashook et al.8 1 have also shown that the oxygen excess region decrease and the oxygen
deficient region increase as the Sr content increase. When the Sr content is fixed, the amount of
the excess oxygen decrease as temperature increase and p(O2) decrease, while the amount of the
oxygen deficiency increase. More importantly, a thermodynamic driving force for the formation
of interstitial oxygen in LNO decreases with increasing 6, suggesting a decrease in the barrier of
interstitial oxygen formation and the larger thermodynamic driving force for interstitial oxygen
formation at lower 6 by Nakamura et al.42 However, LSNO with xsr>0.2 has both oxygen excess
and oxygen deficiency, and thus a thermodynamic driving force for the formation of interstitial
oxygen in LSNO with xsr>0.2 is different from that of LNO.
For the ABO 3 perovskite structure, decreasing p(O2) leads to the formation of oxygen
vacancies, charge-compensated by partial reduction of the B-site cations. The perovskite lattice
can be expanded by following two factors, which are the corresponding increase in the B-site
cation radius and rising coulombic repulsion between the cations. In the case of LNO, however,
these two factors produce opposite effects on the lattice in the a-b plane and along the c-axis.
Although dimensional changes in the a-b plane are still controlled by the size of B-sit ecations,
the chemically induced expansion/contraction of the c parameter is governed by the amount of
interstitial oxygen incorporated
into rock-salt La2O2 layers via anion-anion repulsion.
Consequently, oxygen losses from the LNO structure results in expansion along the perovskitelike layers due to decreasing Ni oxidation state, and in contraction along the perpendicular c-axis
due to decreasing concentration of oxygen interstitials. These effects compensate each other and,
42
thus, partly suppress the overall strain, which becomes almost independent of the oxygen
stoichiometry. 9
Nakamura et al.70 have also revealed the relationship between lattice parameters and the
amount of excess oxygen in LSNO using high temperature X-ray diffraction. According to their
study, As p(O2) increases, a decreases and c increases resulting in the slight change of the cell
volume for xsr=O and 0.2. On the other hand, the lattice parameters and the cell volume of
xsr=0.4 are almost independent of p(O2). In the p( 0 2) range between 10~4 and 1 bar, LNO and
LSNO with xsr=0.2 show significant oxygen content variation while LSNO with xsr=0.4 shows
stable stoichiometric oxygen content (6=0). As 8 increases, a linearly decreases and c linearly
increases. As a consequence, the variation of the cell volume with 8 is very small. To sum up, the
lattice parameter perpendicular to the perovskite and rock salt layers increases and that parallel to
the layers slightly decreases as the amount of excess oxygen increases. The amount of interstitial
oxygen decreases as temperature increases, and a slightly decreases and c increases with
increasing 6.
1.5.3 Oxygen Surface Exchange Kinetics of La2-xSrxNiO 4*8
A few studies have examined the anisotropic nature of the oxygen exchange kinetics in
LNO oxides, which requires the use of single-crystalline samples.4 4 4 5 4 9 83 ' 84 Bassat et al.3 have
shown that the surface exchange kinetics in the a-b plane is ~5 times greater than that along the
c-axis using LNO single crystals. Employing (001)tetragonaoriented epitaxial LNO thin films
grown on the (00 1)cubic-SrTiO3 (STO) and (1 10)bi-NbGaO3 (NGO) substrates, Burriel et al.39
have shown that the surface oxygen kinetics in the a-b plane is two orders of magnitude greater
than that along the c-axis from secondary ion mass spectroscopy (SIMS) measurements. This is
not clear if such a large anisotropy in the surface oxygen exchange kinetics found in these LNO
films is intrinsic to LNO as the surface oxygen kinetics along the c-axis in these thin films are
much lower than single crystals while the values in the a-b planes of LNO films are comparable.
It is interesting to note that decreasing compressive strains in the thin films on NGO substrate
with increasing film thickness increases oxygen transport kinetics but has no apparent influence
on the surface exchange kinetics. This is in contrast to recent studies 4 '4 ' 8, 5 revealing that strains
in the epitaxial thin films and different film stoichiometry can greatly influence surface exchange
43
kinetics. For example, Yamada et al.8 5 have shown that compression in the c-axis of the
(11 0)tetragona-oriented epitaxial Nd2 NiO 4 +, (NNO) film on the (1 00)cubic-Y 2
3
-stabilized ZrO 2
(YSZ) substrate reduces the surface exchange kinetics.
The influence of Sr substitution on the in-plane and out-of-plane oxygen surface
exchange kinetics of La2 -xSrxNiO4+8 (LSNO) is poorly understood due to difficulties in the
growth of single crystals of LSNO with high Sr substitution. 86 Few studies have reported the
oxygen surface exchange kinetics of polycrystalline LSNO with low Sr substitution (xsr = 0.1
and 0.2).38,76,87 Skinner et al. 76 have shown no clear effect of the Sr substitution (xsr = 0.1) on
the surface exchange kinetics whereas Boehm et al.38 have reported that the Sr substitution (xsr
=
0.1 and 0.2) can decrease the surface exchange kinetics.
1.6 Surface Decoration on the MIEC Oxides
To achieve enhanced surface exchange kinetics and stability, recent efforts have been
-
focused on developing advanced cathode materials based on MIECs with surface modification. 43
51, 53, 88-92 Oxide heterostructure interfaces, the combination of a Ruddlesden-Popper (RP)
(Lao.5Sro.5) 2 CoO4
8
(LSC 2 14 ) layer on top of the perovskite LSC, have shown remarkably high
oxygen surface exchange kinetics. 4 5, 90, 91,93-96 Yashiro et al. have reported -1 order of magnitude
enhancement in activity for the composite cathode screen-printed with the mixture of
Lao.6 Sro.4CoO 3. and LSC2 14.9 1 Sase et al. have also reported -3 orders of magnitude higher
oxygen surface exchange coefficient (k*) at the interfacial region between polycrystalline
Lao 6 Sr. 4 CoO 3. 8 and LSC 2 14 compared to their bulk values. 94 More recently, we have shown
using well-defined thin film systems that LSC 214 coverage on epitaxial (001)-oriented
Lao.Sro 2CoO 3- (LSCii 3 ) thin film surfaces can greatly enhance the surface exchange kinetics up
to -2 orders of magnitude.45
Our recent study using Coherent Brag Rod Analysis (COBRA) has revealed the atomic
structure and concentrations of the (001)-oriented LSC11 3 thin film on a SrTiO 3 (STO) substrate,
which shows strontium (Sr) segregation toward the LSC113 surface and Sr depletion near the
interface between LSC113 and STO.
More recently, COBRA has also revealed the markedly
,
enhanced Sr concentration at the interface of LSC113 and LSC 214 and near the surface of LSC2 14
proposing that the increased Sr content at the interface may contribute the enhanced catalytic
44
activity resulting in higher oxygen vacancy concentration.97 In addition, we have shown that
heating the (001)-oriented LSC113 surface leads to the formation of surface Sr-enriched particles
upon annealing while the LSC 2 1 4 -decorated LSC113 surface chemistry is stable upon heating.56
These observations have suggested that the surface decoration can modulate the surface Sr
segregation and the surface phase stability, which can greatly influence the oxygen surface
exchange kinetics and the surface stability in LSC 35 and LSCF. 8''9' Therefore, understanding the
surface decoration effect on the surface chemistry of perovskites is critical and other potential
surface modification materials need to be investigated to design highly active and stable cathodes
for SOFCs.
Similar to the LSC surface modification with LSC 2 14 , several studies have also reported
the enhanced surface electrocatalytic activity of porous LSCF cathodes with surface
decoration.4 3 ,
4648,
50,
88,
92
Depositing
thin
Lao.8 5 5ro.isMnO 3 -8
coatings
on
porous
Lao. 6 5rO.4 Coo.2Feo.8 03- (LSCF 1 3) electrodes using an infiltration process, Lynch et al.4 8 have
shown the enhanced surface electrocatalytic activity of decorated LSCF113 cathodes upon
polarization. A uniform coating of Smo. 5 5r0 .sCoO 3 -84 7 and Lao.48 75Cao.o 12 5Ceo. 50 2 .84
through
infiltration on porous LSCF113 cathodes have also shown the reduced polarization resistance of
the cathode. However, vast majority of research has been performed on porous LSCF11 3
electrodes, which lead to ambiguous structure and geometry, and therefore the physical origin
responsible for enhanced cathodic performance associated with surface decoration of perovskites
is not yet completely understood. In addition, further investigation of other potential surface
modification materials is required to design highly active and stable cathodes for SOFCs.
1.7 Thin Film Cathodes for Solid Oxide Fuel Cells
Epitaxial thin films can allow the understanding of the fundamental physicochemical
properties of these materials and have been used as model systems to develop design principles
for enhancing oxygen surface exchange. 4 ''
99
In addition, the structure of RP phase materials is
anisotropic, typically presenting different ionic and electronic transport properties along different
crystallographic directions as mentioned above. Therefore, in order to investigate the intrinsic
anisotropy feature of these compounds, it is essential to use a high degree of crystal orientation
45
and a low density of grain boundaries, which are either single crystals or highly textured
polycrystalline samples, ideally epitaxially grown thin films.1 00
The main advantages for transport properties control and characterization using thin films
are associated with the following 4 factors. 100 First, epitaxial thin films can be grown with
different orientations, which can allow evaluating anisotropic behaviors (appropriate selection of
the single crystal substrates and optimization of the deposition conditions). Second, the
termination plane of each surface structure can be controlled using epitaxial thin films, and thus
the different surface exchange kinetics can be elucidated. Third, dense thin films can allow
studying oxygen transport kinetics with no influence of non-kinetic issues, such as particle
morphology and connectivity, and porosity and tortuosity typical in porous electrodes. Lastly,
the influence of the strain induced by the substrate on the transport properties can be investigated.
We have therefore utilized epitaxial thin films for all our studies in this thesis. Using aaxis-oriented epitaxial LNO thin films, we have investigated the strain influence on the oxygen
electrocatalysis of RP phase oxides. 40 In addition, we have revealed how Sr can influence the
oxygen surface exchange kinetics and the structure of RP phase oxides using epitaxial LSNO
thin films.4 1 In order to develop a fundamental understanding of surface decoration effect on the
oxygen electrocatalysis, we have utilized epitaxial LSM11 3-decorated LSC13 thin films and
epitaxial LSC 2 14 -decorated LSCF,1 3 thin films. 5 7' 5 8
1.8 Technical Overview
Chapter 2. will provide a brief overview of the experimental procedures for the synthesis of
materials. Specific details for each study will be provided within their respective Chapters 3-6.
In Chapter 3, the strain influence on the oxygen electrocatalysis of Ruddlesden-Popper
materials such as La2NiO 4+8 (LNO) will be discussed.40 Pulsed laser deposition (PLD) is utilized
for the synthesis of the (l00)tetragonaoriented epitaxial LNO thin films. The surface oxygen
exchange kinetics determined from electrochemical impedance spectroscopy (EIS) were found to
increase with decreasing film thickness from 390 to 14 nm. No significant change of the surface
chemistry with different film thicknesses was observed using ex situ auger electron spectroscopy
46
(AES). Increasing volumetric strains in the LNO films at elevated temperatures determined from
in situ high-resolution X-ray diffraction (HRXRD) were correlated with increasing surface
exchange kinetics and decreasing film thickness. Volumetric strains may alter the formation
energy of interstitial oxygen and influence on the surface oxygen exchange kinetics of the LNO
films.
Chapter 4. is a continuation of Chapter 3, attempting to learn more about how Sr can influence
the oxygen electrocatalysis of LNO thin films. 4 1 The growth and oxygen surface exchange
kinetics of La2-xSrxNiO4 8 (LSNO, 0.0 5 xsr 5 1.0) thin films grown on the (00l)cubci-Y2O3stabilized ZrO 2 (YSZ) by pulsed laser deposition will be discussed. High-resolution X-ray
diffraction analysis revealed that the LSNO film orientation was changed gradually from the
(100)tetra. (in-plane) to the (001)tetra. (out-of-plane) orientation in the RP structure with increasing
Sr from La2NiO 4+s (xsr = 0) to LaSrNiO 4 Z (xsr = 1.0). Such a change in the LSNO film
orientation was accompanied with reduction in the oxygen surface exchange kinetics by two
orders of magnitude as shown from electrochemical impedance spectroscopy. Density function
theory (DFT) calculations showed that Sr substitution could stabilize the (001)tetra. surface
relative to the (100)tetra. surface and both Sr substitution and increasing (001)teta. surface could
greatly weaken adsorption of molecular oxygen on the La-La bridge sites in the RP structure,
which can reduce oxygen surface exchange kinetics.
Chapter 5. will expand the understanding of the surface decoration effect on the surface
chemistry of perovskites and other potential surface modification materials to design highly
active and stable cathodes for SOFCs. The influence of Lao.lSro. 2MnO 3 - (LSM82) surface
decoration on the surface exchange kinetics and the surface stability of (001)pseudocubic-oriented
epitaxial La. 8 Sro.2 CoO 3- (LSC82) thin films on (001)-oriented yttria-stabilized zirconia (YSZ)
will be discussed.5 8 In-plane and out-of-plane strains of the LSC82 films at elevated temperature
determined from in situ high resolution X-ray diffraction were not influenced by LSM82
decoration. The surface exchange coefficients (0), determined from electrochemical impedance
spectroscopy, increased significantly with partial LSM82 coverage while full LSM82 coverage
reduced k relative to undecorated LSC82 films. The formation of Sr-enriched particles observed
upon annealing on the undecorated LSC82 surface was eliminated by the LSM82 decoration as
47
revealed by auger electron spectroscopy, which was accompanied with increased Sr and Mn
concentration on the surface. This surface chemistry change is in agreement with density
functional theory calculations that showed a considerable energy gain for exchange of Sr in
LSC82 with La in LSM82. It is postulated that the enhanced surface oxygen exchange kinetics
can be attributed primarily to the increased Sr concentration in the perovskite structure near the
surface.
Chapter 6. leveraging the insight gained from the previous study,4 4 ~4 5' 48-49 they were applied to
further the understanding of the heterostructured interface using LSCF11 3 with LSC 214 coverage.
The influence of LSC 214 surface decoration on the oxygen surface exchange kinetics of epitaxial
LSCF11 3 thin films, comparing with an LSC214 -decroated LSC11 3 heterostructure thin film will be
discussed.5 7 Using high-resolution X-ray diffraction (HRXRD), it is shown that the LSC 214
decoration has no influence on the in-plane and out-of-plane strains of both LSCF11 3 and LSC11 3
films. The LSC 214 decoration can enhance the surface exchange coefficient (ku) only -2 times
relative to the pristine LSCF1 1 3 thin film while it can significantly enhance the k of the LSC11 3
thin film by -2 orders of magnitude. Auger electron spectroscopy (AES) reveals markedly
3
,
different surface Sr concentrations between LSC 2 14-decorated LSCF11 3 and LSC2 14 -LSC,1
which is supported by the distinct thermodynamic driving force for cation interdiffusion across
the heterointerfaces of LSC2 14-LSCF,1
3
vs. LSC 2 14 -LSCi1
3
in the DFT modeling.
Chapter 7. will provide my conclusion of RP oxides and surface decorated ABO 3 oxides for
intermediate temperature (500 - 700 *C) oxide electrocatalysts and perspectives of future work.
1.9 References
1.
U.S. Energy Information Administration (EA), InternationalEnergy Outlook 2013.
2.
Energy Investing with Energy and Capital, http://www.energyandcapital.com.
3.
L. Caspeta, N. A. A. Buijs and J. Nielsen, Energy Environ. Sci., 2013, 6, 1077-1082.
4.
US. Energy Use 2008, http://www.usehalf.com/peakoil/PeakOil.html.
5.
The Green Box Systems Group, http://www.esru.strath.ac.uk/EandE/Websites/9900/biofuelcells/groupproject/pagefraine.htm.
48
6.
S. B. Adler, Chem. Rev., 2004, 104, 4791-4843.
7.
J. Larminie and A. Dicks, Fuel Cell System Explained, John Wiley & Sons Inc, 2003.
8.
S. M. Haile, Acta Mater., 2003, 51, 5981-6000.
9.
A. J. Jacobson, Chem. Mat., 2010, 22, 660-674.
10.
N.
11.
N.
12.
Z. P. Shao and S. M. Haile, Nature, 2004, 431, 170-173.
13.
S. C. Singhal, Solid State Ion., 2000, 135, 305-3 13.
14.
R. Merkle and J. Maier, Phys. Chem. Chem. Phys., 2002, 4, 4140-4148.
15.
M. N.
Q. Minh, J. Am. Ceram. Soc., 1993, 76, 563-588.
Q. Minh, Solid State Ion., 2004, 174, 271-277.
Q.,
in SOFC Fuel Cell Technology, Materials Research Society, Boston, MA,
2002.
16.
L. da Conceicao, C. R. B. Silva, N. F. P. Ribeiro and M. Souza, Mater. Charact., 2009,
60, 1417-1423.
17.
R. A. De Souza and J. A. Kilner, Solid State Ion., 1998, 106, 175-187.
18.
R. A. De Souza and J. A. Kilner, Solid State Ion., 1999, 126, 153-161.
19.
A. Endo, M. Ihara, H. Komiyama and K. Yamada, Solid State Ion., 1996, 86-8, 11911195.
20.
T. Ioroi, T. Hara, Y. Uchimoto, Z. Ogumi and Z. Takehara, J. Electrochem. Soc., 1997,
144, 1362-1370.
21.
M. J. Zhi, G. W. Zhou, Z. L. Hong, J. Wang, R. Gemmen, K. Gerdes, A. Manivannan, D.
L. Ma and N.
22.
Q. Wu, Energy Environ. Sci., 2011,
4, 139-144.
W. C. Chueh and S. M. Haile, in Annual Review of Chemical and Biomolecular
Engineering, Vol 3, ed. J. M. Prausnitz, Annual Reviews, Palo Alto, 2012, pp. 313-341.
23.
F. Kroger, The Chemistry ofImperfect Crystals., North-Holland, Amsterdam, 1964.
24.
S. B. Adler, Solid State Ion., 1998, 111, 125-134.
25.
S. B. Adler, J. A. Lane and B. C. H. Steele, J. Electrochem. Soc., 1996, 143, 3554-3564.
26.
J. Mizusaki, Y. Mima, S. Yamauchi, K. Fueki and H. Tagawa, J. Solid State Chem.,
1989, 80, 102-111.
27.
V. G. Sathe, A. V. Pimpale, V. Siruguri and S. K. Paranjpe, Journal of PhysicsCondensedMatter, 1996, 8, 3889-3896.
49
28.
S. Y. Wang, J. Yoon, G. Kim, D. X. Huang, H. Y. Wang and A. J. Jacobson, Chem. Mat.,
2010, 22, 776-782.
29.
A. Esquirol, N. P. Brandon, J. A. Kilner and M. Mogensen, J. Electrochem. Soc., 2004,
151, A1847-A1855.
30.
M. Kuhn, Y. Fukuda, S. Hashimoto, K. Sato, K. Yashiro and J. Mizusaki, J. Electrochem.
Soc., 2013, 160, F34-F42.
31.
L. W. Tai, M. M. Nasrallah, H. U. Anderson, D. M. Sparlin and S. R. Sehlin, Solid State
Ion., 1995, 76, 259-271.
32.
L. W. Tai, M. M. Nasrallah, H. U. Anderson, D. M. Sparlin and S. R. Sehlin, Solid State
Ion., 1995, 76, 273-283.
33.
Z. Feng, Y. Yacoby, W. T. Hong, H. Zhou, M. D. Biegalski, H. M. Christen and Y. ShaoHorn, Energy Environ. Sci., 2014, 7, 1166-1174.
34.
J. S. Hardy, J. W. Templeton, D. J. Edwards, Z. G. Lu and J. W. Stevenson, J. Power
Sources, 2012, 198, 76-82.
35.
M. Kubicek, A. Limbeck, T. Fromling, H. Hutter and J. Fleig, J. Electrochem. Soc., 2011,
158, B727-B734.
36.
S. P. Simner, M. D. Anderson, M. H. Engelhard and J. W. Stevenson, Electrochem. Solid
State Lett., 2006, 9, A478-A48 1.
37.
J. M. Bassat, P. Odier, A. Villesuzanne, C. Marin and M. Pouchard, Solid State Ion.,
2004,167, 341-347.
38.
E. Boehm, J. M. Bassat, P. Dordor, F. Mauvy, J. C. Grenier and P. Stevens, Solid State
Ion., 2005, 176, 2717-2725.
39.
M. Burriel, G. Garcia, J. Santiso, J. A. Kilner, J. C. C. Richard and S. J. Skinner, J
Mater. Chem., 2008, 18, 416-422.
40.
D. Lee, A. Grimaud, E. J. Crumlin, K. Mezghani, M. A. Habib, Z. Feng, W. T. Hong, M.
D. Biegalski, H. M. Christen and Y. Shao-Horn, The Journal of Physical Chemistry C,
feature article, 2013, 117, 18789-18795.
41.
D. Lee, Y.-L. Lee, A. Grimaud, W. T. Hong, M. D. Biegalski, D. Morgan and Y. ShaoHorn, J. Mater. Chem. A, 2014, 2, 6480-6487.
42.
T. Nakamura, K. Yashiro, K. Sato and J. Mizusaki, Solid State Ion., 2009, 180, 368-376.
50
43.
J. Chen, F. L. Liang, B. Chi, J. Pu, S. P. Jiang and L. Jian, J. Power Sources, 2009, 194,
275-280.
44.
E. J. Crumlin, S. J. Ahn, D. Lee, E. Mutoro, M. D. Biegalski, H. M. Christen and Y.
Shao-Hom, J. Electrochem. Soc., 2012, 159, F219-F225.
45.
E. J. Crumlin, E. Mutoro, S. J. Ahn, G. J. la 0, D. N. Leonard, A. Borisevich, M. D.
Biegalski, H. M. Christen and Y. Shao-Hom, J. Phys. Chem. Lett., 2010, 1, 3149-3155.
46.
M. F. Liu, D. Ding, K. Blinn, X. X. Li, L. F. Nie and M. Liu, Int. J Hydrog. Energy,
2012,37, 8613-8620.
47.
X. Y. Lou, S. Z. Wang, Z. Liu, L. Yang and M. L. Liu, Solid State Ion., 2009, 180, 12851289.
48.
M. E. Lynch, L. Yang, W. T. Qin, J. J. Choi, M. F. Liu, K. Blinn and M. L. Liu, Energy
Environ. Sci., 2011,4,2249-2258.
49.
E. Mutoro, E. J. Crumlin, M. D. Biegalski, H. M. Christen and Y. Shao-Hom, Energy
Environ. Sci., 2011, 4, 3689-3696.
50.
L. F. Nie, M. F. Liu, Y. J. Zhang and M. L. Liu, J. Power Sources, 2010, 195,4704-4708.
51.
Y. Sakito, A. Hirano, N. Imanishi, Y. Takeda, 0. Yamamoto and Y. Liu, J. Power
Sources, 2008, 182, 476-48 1.
52.
H. J. Hwang, M. B. Ji-Woong, L. A. Seunghun and E. A. Lee, J Power Sources, 2005,
145,243-248.
53.
M. Muranaka, K. Sasaki, A. Suzuki and T. Terai, J Electrochem. Soc., 2008, 155, B860B863.
54.
M. Muranaka, K. Sasaki, A. Suzuki and T. Terai, J. Electrochem. Soc., 2009, 156, B743B747.
55.
E. P. Murray, M. J. Sever and S. A. Barnett, Solid State Ion., 2002, 148, 27-34.
56.
Z. Feng, E. J. Crumlin, W. T. Hong, D. Lee, E. Mutoro, M. D. Biegalski, H. Zhou, H.
Bluhm, H. M. Christen and Y. Shao-Hom, The Journal of Physical Chemistry Letters,
2013,4, 1512-1518.
57.
D. Lee, Y.-L. Lee, M. D. Biegalski, D. Morgan and Y. Shao-Horn, In Preparation,2014.
58.
D. Lee, Y.-L. Lee, A. Grimaud, W. T. Hong, M. D. Biegalski, D. Morgan and Y. ShaoHorn, The Journalof PhysicalChemistry C, 2014, Under Review.
59.
R. H. E. van Doom and A. J. Burggraaf, Solid State Ion., 2000, 128, 65-78.
51
60.
Y. Jiang, S. Z. Wang, Y. H. Zhang, J. W. Yan and W. Z. Li, Solid State Ion., 1998, 110,
111-119.
61.
J. Mizusaki, Solid State Ion., 1992, 52, 79-91.
62.
T. Horita, K. Yamaji, N. Sakai, H. Yokokawa, A. Weber and E. Ivers-Tiffee,
ElectrochimicaActa, 2001, 46, 1837-1845.
63.
Y. M. Kim, S. I. Pyun, J. S. Kim and G. J. Lee, J. Electrochem. Soc., 2007, 154, B802B809.
64.
T. Kawada, J. Suzuki, M. Sase, A. Kaimai, K. Yashiro, Y. Nigara, J. Mizusaki, K.
Kawamura and H. Yugami, J. Electrochem. Soc., 2002, 149, E252-E259.
65.
A. N. Petrov, V. A. Cherepanov, 0. F. Kononchuk and L. Y. Gavrilova, J. Solid State
Chem., 1990, 87, 69-76.
66.
S. Hashimoto, Y. Fukuda, M. Kuhn, K. Sato, K. Yashiro and J. Mizusaki, Solid State
Ion., 2010, 181, 1713-1719.
67.
S. N. Ruddlesden and P. Popper, Acta Crystallographica,1958, 11, 54-55.
68.
J. D. Jorgensen, B. Dabrowski, S. Pei, D. R. Richards and D. G. Hinks, Phys. Rev. B,
1989, 40, 2187-2199.
69.
V. V. Kharton, A. V. Kovalevsky, M. Avdeev, E. V. Tsipis, M. V. Patrakeev, A. A.
Yaremchenko, E. N. Naumovich and J. R. Frade, Chem. Mat., 2007, 19, 2027-2033.
70.
T. Nakamura, K. Yashiro, K. Sato and J. Mizusaki, Solid State Ion., 2010, 181, 292-299.
71.
A. Demourgues, F. Weill, B. Darriet, A. Wattiaux, J. C. Grenier, P. Gravereau and M.
Pouchard, J. Solid State Chem., 1993, 106, 317-329.
72.
J. Rodriguezcarvajal, M. T. Fernandezdiaz and J. L. Martinez, Journal of PhysicsCondensedMatter, 1991, 3, 3215-3234.
73.
J. Gopalakrishnan, G. Colsmann and B. Reuter, J. Solid State Chem., 1977, 22, 145-149.
74.
Y. Takeda, R. Kanno, M. Sakano, 0. Yamamoto, M. Takano, Y. Bando, H. Akinaga, K.
Takita and J. B. Goodenough, Mater. Res. Bull., 1990, 25, 293-306.
75.
S. J. Skinner and J. A. Kilner, Ionics, 1999, 5, 171-174.
76.
S. J. Skinner and J. A. Kilner, Solid State Ion., 2000, 135, 709-712.
77.
M. Hucker, K. Chung, M. Chand, T. Vogt, J. M. Tranquada and D. J. Buttrey, Phys. Rev.
B, 2004, 70, 10.
52
78.
T. Nakamura, K. Yashiro, K. Sato and J. Mizusaki, Phys. Chem. Chem. Phys., 2009, 11,
3055-3062.
79.
V. V. Vashook, N. E. Trofunenko, H. Ullmann and L. V. Makhnach, Solid State Ion.,
2000, 131, 329-336.
80.
V. V. Kharton, A. P. Viskup, A. V. Kovalevsky, E. N. Naumovich and F. M. B. Marques,
Solid State Ion., 2001, 143, 337-353.
81.
V. V. Vashook, S. P. Tolochko, L. Yushkevich, L. V. Makhnach, I. F. Kononyuk, H.
Altenburg, J. Hauck and H. Ullmann, Solid State Ion., 1998, 110, 245-253.
82.
V. V. Vashook, H. Ullmann, 0. P. Olshevskaya, V. P. Kulik, V. E. Lukashevich and L.
V. Kokhanovskij, Solid State Ion., 2000, 138, 99-104.
83.
E. J. Crumlin, E. Mutoro, Z. Liu, M. E. Grass, M. D. Biegalski, Y. L. Lee, D. Morgan, H.
M. Christen, H. Bluhm and Y. Shao-Horn, Energy Environ. Sci., 2012, 5, 6081-6088.
84.
G. J. la 0, S. J. Ahn, E. Crumlin, Y. Orikasa, M. D. Biegalski, H. M. Christen and Y.
Shao-Horn, Angew. Chem.-Int. Edit., 2010, 49, 5344-5347.
85.
A. Yamada, Y. Suzuki, K. Saka, M. Uehara, D. Mori, R. Kanno, T. Kiguchi, F. Mauvy
and J. C. Grenier, Adv. Mater., 2008, 20, 4124-+.
86.
S. Shinomori, M. Kawasaki and Y. Tokura, Appl. Phys. Lett., 2002, 80, 574-576.
87.
A. Aguadero, M. J. Escudero, M. Perez, J. A. Alonso and L. Daza, J Fuel Cell Sci.
Technol., 2007, 4, 294-298.
88.
H. P. Ding, A. V. Virkar, M. L. Liu and F. Liu, Phys. Chem. Chem. Phys., 2013, 15, 489496.
89.
Z. Liu, M. F. Liu, L. Yang and M. L. Liu, J. Energy Chem., 2013, 22, 555-559.
90.
M. Sase, F. Hermes, K. Yashiro, K. Sato, J. Mizusaki, T. Kawada, N. Sakai and H.
Yokokawa, J. Electrochem. Soc., 2008, 155, B793-B797.
91.
K. Yashiro, T. Nakamura, M. Sase, F. Hermes, K. Sato, T. Kawada and J. Mizusaki,
Electrochem. Solid State Lett., 2009, 12, B 135-B 137.
92.
X. B. Zhu, D. Ding, Y.
Q.
Li, Z. Lu, W. H. Su and L. Zhen, Int. J Hydrog. Energy, 2013,
38, 5375-5382.
93.
M. Sase, F. Hermes, T. Nakamura, K. Yashiro, K. Sato, J. Mizusaki, T. Kawada, N.
Sakai, K. Yamaji, T. Horita and H. Yokokawa, in Solid Oxide Fuel Cells 10, eds. K.
53
Eguchi, S. C. Singhai, H. Yokokawa and H. Mizusaki, Electrochemical Society Inc,
Pennington, 2007, pp. 1055-1060.
94.
M. Sase, K. Yashiro, K. Sato, J. Mizusaki, T. Kawada, N. Sakai, K. Yamaji, T. Horita
and H. Yokokawa, Solid State Ion., 2008, 178, 1843-1852.
95.
A. Unemoto, K. Nagao, T. Tairako, K. Amezawa and T. Kawada, in Ionic and Mixed
ConductingCeramics 7, eds. M. Mogensen, T. Armstrong, T. M. Gur and H. Yokokawa,
Electrochemical Soc Inc, Pennington, 2010, pp. 59-70.
96.
K. Yashiro, T. Nakamura, M. Sase, F. Hermes, K. Sato, T. Kawada and J. Mizusaki, in
Solid Oxide Fuel Cells 10, eds. K. Eguchi, S. C. Singhai, H. Yokokawa and H. Mizusaki,
Electrochemical Society Inc, Pennington, 2007, pp. 1287-1292.
97.
Z. Feng, Y. Yacoby, M. J. Gadre, Y. L. Lee, W. T. Hong, H. Zhou, M. D. Biegalski, H.
M. Christen, S. B. Adler, D. Morgan and Y. Shao-Horn, J Phys. Chem. Lett., 2014,
Submitted.
98.
D. Oh, D. Gostovic and E. D. Wachsman, J. Mater. Res., 2012, 27, 1992-1999.
99.
F. S. Baumann, J. Fleig, G. Cristiani, B. Stuhlhofer, H. U. Habermeier and J. Maier, J.
Electrochem. Soc., 2007, 154, B93 1 -B94 1.
100.
J. Santiso and M. Burriel, J. Solid State Electrochem., 2011, 15, 985-1006.
54
Chapter 2.
Experimental Approach
2.1 Synthesis & Fabrication
2.1.1 Synthesis of Thin Film by Pulsed Laser Deposition (PLD)
All powders will be synthesized by Pechini methods.' After synthesizing the powder,
pulsed laser deposition (PLD) target pellets with 25 mm diameter will be subsequently fabricated
by uniaxial pressing and then sintered at certain temperature.
-----------------
Rheed oscillation
-
-
Rheed Pattern
Figure 2-1. A photo of pulsed laser deposition (PLD) for the growth of an epitaxial thin film on
YSZ with an interlayer of GDC. A graph shows the reflection high-energy electron diffraction
(RHEED), which can allow an in-situ monitoring of the film growth.
55
For all ORR studies, the primary thin films investigated are LN0 2 14, LSN0
2 14
(0.0
< xsr <
1.0), LSC1 1 3, LSM1 1 3 , LSCF 1 3, LSM 1 3-decorated LSC 1 3 , LSC 214 -decorated LSC 1 3 , and
LSC 214-decorated LSCF 13 , which were epitaxially grown using pulsed laser deposition (PLD) on
a (001)-oriented yttria-stabilized zirconia (YSZ) substrate with an interlayer of epitaxially grown
gadolinium doped ceria (GDC), as shown in Figure 2-1. Single crystal 9.5 mol% YSZ wafers
with the (GO )cubic orientation and dimensions of 10
x
5 x 0.5 mm (MTI corporation, USA) were
used as the substrate. Prior to film and GDC deposition, platinum ink (Pt) (#6082, BASF, USA)
counter electrodes were painted on one side of the YSZ and dried at 900 *C in air for 1 hour. The
YSZ wafer was affixed to the PLD substrate holder using a small amount of silver paint
(Leitsilber 200, Ted Pella, USA) for thermal contact. PLD was performed using a KrF excimer
laser at A = 248 nm, 10 Hz pulse rate and
-
1.5 J/cm 2 pulse energy under an oxygen partial
pressure, p(O2) of 6.6 x 10- atm (50 mTorr). The utilization of reflection high-energy electron
diffraction (RHEED) enabled an in-situ monitoring of the film growth. After completing the film
deposition, the samples were cooled down to room temperature in the PLD chamber for -1 hour
under a p(O2) of 6.6 x 10-5 atm (50 mTorr).
2.1.2 Fabrication of Microelectrodes
Photoresist
20um
PR coating
PR
PR
Exposure
Gil
PR
P
PR strip
RP
FilmFimFm
Development
GD
Gfl
Figure 2-2. A photolithography process of thin film microelectrode.
56
Etching
To prepare the samples for electrochemical impedance spectroscopy (EIS), traditional
photolithography techniques were utilized to create microelectrodes ranging in size from -25
srm
to -200 pm, as depicted in Figure 2-2.% 4
2.2 Characterization
2.2.1 Surface Morphology and Structural Analysis
a)
2
N
0
0
0
Z
N
N
(0
Z
9
9
0
IS
goo
C*4
00
Z Z
a
0
0
Z
b)
LNO or LSNO (1 03)m.
0
.45
0
N
C,,
z_j
A
X=I
A
GqC
(
2 0 2
)cubi
V)
-W
A
0
'.j
-A-
j22)ub
X=0.2A&_
iY$h
rrVI_ fin
20
40
60
80
100
20/*
-180 -90
0
90 180
0/*
Figure 2-3. (a) Normal X-ray diffraction (XRD), and (b) off-normal XRD results of La2
xSrxNiO4 s (0.0 < XSr < 1.0) thin films. 2
Benefits for these dense thin-film electrodes are the comprehensive characterization that
can be applied. The oxide phase purity and the thin film orientations were investigated via highresolution X-ray diffraction (HRXRD) using a four-circle diffractometer (PANalytical, USA and
57
Bruker D8, Germany). Measurements were performed in normal and off-normal configurations,
as shown in Figure 2-3.2
The surface morphology was examined by scanning electron microscopy (SEM, JEOL
6320FV, Japan) and atomic force microscopy (AFM, Veeco, USA), as shown in Figure 2-4.2, 3
LSM 0.9 nm
LSNO (Sr=0.4)
RMS: 0.968 nm
[LM
10 P
Figure 2-4. (a) AFM images of the -0.9 nm LSM113-decorated LSC113 thin film with root
mean square (RMS) of 0.968 nm after annealing at 550 C,3 and (b) SEM images of the asprepared LSNO with Sr = 0.4.2
2.2.2 Surface Chemistry Analysis
Auger electron spectroscopy (AES) (Figure 2-5) was conducted using a Physical
Electronics 700 Scanning Auger Nanoprobe (PHI, USA) operating at an accelerating voltage of
10 kV to analyze the surface chemistry change of the films after heat treatment. The films were
annealed at 550 'C for 6 hours in an oxygen partial pressure of 1 atm before AES data were
collected. The AES data were collected using two different modes: area mode (three different 10
ptm x 10 pm regions selected across a sample) and point mode (two different ~0.45 tm diameter
spots selected on a sample) in an ultra-high vacuum chamber. Elemental quantification of AES
spectra utilized relative sensitivity factors (RSFs) of 0.059, 0.027, 0.076, 0.161, 0.227, and 0.212
for LaMNN, SrLMM, COLMM, MnLMM, NiLMM, and OKLL, respectively, as supplied by the AES
manufacturer (Physical Electronics). 3 ' 4
58
LSC Pristine
6C,
C
LSC Annealed
C
1612
1IM
16"4
2 iim
Kinetic Energy / eV
Figure 2-5. AES and SEM analysis for bare LSC11 3 thin film before and after annealing.
Annealing was performed at 550 *C in an oxygen partial pressure of 1 atm. (a) Sr Auger spectra
of bare LSC113 thin film probed for: as-deposited surface (gray), and particles (dark red) and film
surface (dark blue) after annealing. SEM image of (b) as-deposited and (c) annealed LSC1 13. 3
2.2.3 Electrochemical Impedance Spectroscopy (EIS)
Electrochemical impedance spectroscopy (EIS) is a common in situ high temperature and
high pressure electrochemical technique that can provide great insight into how an electrode is
performing. This technique has been utilized for many cathodic ORR studies. 2 8 EIS possesses
the ability to evaluate how sensitive each electrode is to temperature, pressure and when
combined with reaction models can be utilized to elicit various reaction mechanisms. For ORR
studies utilizing an oxygen conducting electrolyte such as YSZ, both working and counter
electrodes can be placed within a single chamber and be exposed to the same oxygen partial
pressure. Since both working and counter electrodes are active catalysts for ORR and oxygen
evolution reaction (OER), oxygen can essentially be oscillated between the working electrode
and counter electrode, thus allowing both electrodes to be in the same oxygen environment.
EIS measurements of microelectrodes -200 ptm in diameter were performed using a
microprobe station (Karl Suss, Germany) connected to a frequency response analyzer (Solartron
1260, USA) and dielectric interface (Solartron 1296, USA). Temperature was controlled at 550
59
'C using a heating stage (Linkam TS1500, UK) and data were collected between 1 MHz to 1
mHz using a voltage amplitude of 10 mV. EIS testing temperature was calibrated with a
thermocouple contacting the thin film surface and deviation of
5 'C was observed. EIS
experiments were performed as a function of p(O2) between 10-3 and 1 atm. Multiple electrodes
(at least three) of all films were measured by EIS at each temperature and p(O2) to ensure that the
EIS results were reproducible and representative. EIS data of all samples used in this study were
found to show nearly perfect semicircle impedances,5 and typical features in the Nyquist plots
are shown in the schematic in Figure 2-6. EIS data were analyzed using an equivalent circuit
(Figure 2-6b), where the high-frequency intercept corresponds to the oxygen ion conduction
resistance in YSZ and the mid-frequency feature is believed to result from the interface between
YSZ and GDC as reported previously.9
a)
b)
R2
ORR
02
(MF) Interface
(LF) Film
c)
-Z
ORR
z
02
Figure 2-6. (a) Schematic of a Film/GDC/YSZ(001)/porous Pt sample and electrochemical
testing configuration (not drawn to scale) and (b) equivalent circuit (RI = YSZ electrolyte
resistance, R 2 = electrode/electrolyte interface resistance9, RORR = ORR resistance, CPE =
constant phase element) used to extract ORR kinetics, and (c) Typical nyquist plot of the thin
film electrode at 550 0 C (HF: 104 ~ 105 Hz, MF: 103 ~ 104 Hz, and LF: 10-2 _ 103 Hz).
60
2.2.4 In situ High-resolution X-ray Diffraction (HRXRD)
In situ HRXRD was performed on a four-circle diffractometer (PANalytical) in an
oxygen partial pressure, p(O2) of 1 atm and a controlled temperature stage (DHS 900, Anton
Paar) as shown in Figure 2-7. Silver paste was used to adhere the thin film sample to the heating
plate. The heating rate was ~10 *C min-I and the temperature was held for 20 minutes at each
temperature (25 'C, 150 'C, 250 'C, 350 'C, 450 'C, and 550 'C) before XRD data were
collected. Sample realignment was conducted at each temperature to maximize the XRD
intensities. As the thermocouple for this experiment was placed inside the heating stage, a small
difference between set and actual temperatures on the sample surface cannot be excluded.
~194
.
["3
O
.
.
.
. I
V
.
V
.
.
b)
.
a)195
14 nm
42 nm
178 nm
E 193 V
390nm
0 192
191
.2190
Relaxed
189
300 400 500 600 700 800 900
Temperature / K
Figure 2-7. (a) In situ HRXRD data of the unit cell volume of the (1 00)tetragonaroriented epitaxial
LNO thin films with
-
14,
-
42, -178 and -390 nm as a function of temperature in a p(O2) of 1
atm, 4 and (b) a schematic of In situ HRXRD
2.3 References
1.
M. Pechini, US. PatentNo. 3,330,697, 1967.
61
2.
D. Lee, Y.-L. Lee, A. Grimaud, W. T. Hong, M. D. Biegalski, D. Morgan and Y. ShaoHorn, J Mater. Chem. A, 2014, 2, 6480-6487.
3.
D. Lee, Y.-L. Lee, A. Grimaud, W. T. Hong, M. D. Biegalski, D. Morgan and Y. ShaoHorn, The Journalof Physical Chemistry C, 2014, Under Review.
4.
D. Lee, A. Grimaud, E. J. Crumlin, K. Mezghani, M. A. Habib, Z. Feng, W. T. Hong, M.
D. Biegalski, H. M. Christen and Y. Shao-Horn, The Journalof Physical Chemistry C,
feature article, 2013, 117, 18789-18795.
5.
S. B. Adler, Chem. Rev., 2004, 104, 4791-4843.
6.
F. S. Baumann, J. Fleig, H. U. Habermeier and J. Maier, Solid State Ion., 2006, 177,
1071-1081.
7.
E. J. Crumlin, S. J. Ahn, D. Lee, E. Mutoro, M. D. Biegalski, H. M. Christen and Y.
Shao-Horn, J. Electrochem. Soc., 2012, 159, F219-F225.
8.
D. Lee, Y.-L. Lee, M. D. Biegalski, D. Morgan and Y. Shao-Horn, In Preparation,2014.
9.
G. J. la 0, S. J. Ahn, E. Crumlin, Y. Orikasa, M. D. Biegalski, H. M. Christen and Y.
Shao-Horn, Angew. Chem.-Int. Edit., 2010, 49, 5344-5347.
62
Chapter 3.
Strain Influence on the Oxygen Electrocatalysis
of the (100)-Oriented Epitaxial La2NiO 4+e Thin
Films at Elevated Temperatures
1
4
_____
I______
V_____-_______
550 OC
p(0 2 )mftm
10
-
0
1
2
3
4
5
6
7
%
Volumetric Strain /
Reproduced in part with permission from Dongkyu Lee, Alexis Grimaud, Ethan J. Crumlin,
Khaled Mezghani, Mohamed A. Habib, Zhenxing Feng, Wesley T. Hong, Michael D. Biegalski,
Hans M. Christen, and Yang Shao-Horn, Strain Influence on the Oxygen Electrocatalysis of the
(100)-Oriented Epitaxial La2NiO4 +6 Thin Films at Elevated Temperatures, The Journal of
Physical Chemistry C, 2013, 117 (37), 18789 - 18795, Copyright 2013 American Chemical
Society.
63
3.1 Introduction
Transition metal oxides such as Lal.,Sr.MnO 3.8(LSM) 1-5 are commonly used to promote
oxygen electrocatalysis for solid oxide fuel cells 6- 12 and oxygen permeation membrane
applications1 3 at high temperatures such as 1000 'C. Reducing the operating temperature is of
vital importance to reduce the degradation and improve the lifetime of these solid-state devices.
Mixed ionic and electronic conductors (MIECs) such as Lai.xSrxCoO3-8 (LSC)14 2 0 have been
studied intensively to promote oxygen electrocatalysis at intermediate temperatures such as 600
*C. These MIECs have high surface oxygen exchange kinetics, which can allow oxygen
electrocatalysis to take place on the entire oxide surface not just at the electrode/electrolyte
interface. Ruddlesden-Popper (RP) oxides such as La2NiO 4+8 (LNO) with high surface oxygen
exchange kinetics and oxygen transport properties 21-30 are interesting alternative materials to
LSC for intermediate temperature operation. 14-20 The RP structure can be described as an
intergrowth of alternative NiO 2 and La 2O 2 rock salt layers or layers of LNO separated by LaO
layers along the c axis, where oxygen transport kinetics by interstitial oxygen are considerably
higher in the a-b plane up to two orders of magnitude than along the c-axis.2 1 2 3
21 22 28
Although surface oxygen exchange kinetics of bulk RP oxides are well studied, , ,
30
few studies have examined the anisotropic nature of the oxygen exchange kinetics in these
oxides, which requires the use of single-crystalline samples. 15 7' 3'1 ,32 Bassat et al.2 ' have shown
that the surface exchange kinetics in the a-b plane is -5 times greater than that along the c-axis
using LNO single crystals. Employing (001)tetragona-oriented epitaxial LNO thin films grown on
the (001)cubic-SrTiO3 (STO) and (1 I0)cubic-NbGaO3 (NGO) substrates, Burriel et al. 3 have shown
that the surface oxygen kinetics in the a-b plane is two orders of magnitude greater than that
along the c-axis from secondary ion mass spectroscopy (SIMS) measurements. This is not clear
if such a large anisotropy in the surface oxygen exchange kinetics found in these LNO films is
intrinsic to LNO as the surface oxygen kinetics along the c-axis in these thin films are much
lower than single crystals while the values in the a-b planes of LNO films are comparable. It is
interesting to note that decreasing compressive strains in the thin films on NGO substrate with
increasing film thickness increases oxygen transport kinetics but has no apparent influence on
the surface exchange kinetics. This is in contrast to recent studies,15
17,33
revealing that strains in
the epitaxial thin films and different film stoichiometry can greatly influence surface exchange
kinetics. For example, Yamada et al.3 3 have shown that compression in the c-axis of the
64
(110)tetragona-oriented epitaxial Nd2NiO 4+8 (NNO) film on the (100)cubic-Y23-stabilized ZrO 2
(YSZ) substrate reduces the surface exchange kinetics.
In this study, we report strong film thickness dependence on the surface oxygen exchange
rates (k') of the (100)tetragonai-oriented epitaxial growth of LNO thin films grown on YSZ in
contrast to (001)tetragona.-oriented epitaxial growth of LNO films.2 3 Using in situ high-resolution
X-ray diffraction (HRXRD), we show that the unit cell volume of the films increases as a
function of increasing film thickness at elevated temperatures. As ex situ auger electron
spectroscopy (AES) reveals no apparent difference in the surface chemistry of the LNO thin
films upon heating, the thickness-dependent surface oxygen exchange rates of the LNO thin
films can be attributed to different stains in these films.
3.2 Experimental Methods
Pulsed laser deposition (PLD) was utilized to deposit the (100)tetragonai-oriented epitaxial
%
LNO thin films (-14, -42, -178, -390 nm) on YSZ with gadolinium-doped ceria (GDC, 20 mol
Gd) as the buffer layer with thickness ~ 5 nm. Single crystal 9.5 mol% Y 2 0 3-stabilized ZrO 2
(YSZ) wafers with (001) orientation and dimensions of 10
x
5
x
0.5 mm (MTI corporation,
USA), were used as substrate. Prior to LNO and GDC deposition, platinum ink (Pt) (#6082,
BASF, USA) counter electrodes were painted on one side of the YSZ and dried at 900 *C in air
for 1 hour. The YSZ wafer was affixed to the PLD substrate holder using a small amount of
silver paint (Leitsilber 200, Ted Pella, USA) for thermal contact. PLD was performed using a
KrF excimer laser at A= 248 nm, 10 Hz pulse rate and 45 mJ pulse energy under p(O2) of 6.6
105
x
atm (50 mTorr) with 500 pulses of GDC (-5 nm) at 550 'C, followed by 1,000 pulses, 5,000
pulses, 15,000 pulses and 35,000 pulses of LNO (-14, -42, -178, -390 nm, respectively) at 650
'C. The film thicknesses were determined by atomic force microscopy (AFM). The utilization of
reflection high-energy electron diffraction (RHEED) enabled diagnostic in-situ monitoring of the
LNO film growth. After completing the LNO deposition, the sample was cooled to room
temperature in the PLD chamber for -1 hour under an oxygen partial pressure of 6.6
x 10-
atm
(50 mTorr). The synthesis details of LNO and GDC PLD targets can be found in the Supporting
Information (SI). Oxide phase purity and orientation of the thin film systems were investigated
via high resolution X-ray diffraction (HRXRD) using a four-circle diffractometer (Panalytical,
65
USA and Bruker D8, Germany). Measurements were performed in normal and off-normal
configurations. LNO in the Ruddlesden-Popper structure is known to be orthorhombic with
space group Fmmm at room temperature for an oxygen overstoichiometry (6) ~ 0.18 as reported
by Jorgensen et al.34 As the Ruddlesden-Popper structure can be influenced greatly by the
oxygen stoichiometry, fully oxidized La2 NiO 4 .2 5 exhibits a monoclinic symmetry with space
group C2 at room temperature3 5 while fully stoichiometric La 2NiO 4 is orthorhombic with space
group Bmab.36 In this study, large experimental uncertainty exists in the inferred oxygen
nonstoichiometry of LNO thin films from the literature3 7 using the lattice parameters determined
from HRXRD (for details see the Supporting Information) and the XRD data of singlecrystalline thin films do not allow a precise refinement of the symmetry of the LNO structure.
Therefore, we chose the tetragonal symmetry with space group 14/mmm, which differs from the
orthorhombic by the loss of the octahedral tilting (<
10)
in this study. The LNO in-plane lattice
parameter (b and c lattice parameter) was determined from the off-normal (11 4 )tetragonal and
(10 3 )tetragonal peak position, respectively and the a lattice parameter of LNO normal to the film
surface was determined from the (2 0 0 )tetragonal peak positions. Surface morphology was examined
by optical microscopy (Carl Zeiss, Germany) and atomic force microscopy (AFM) (Veeco,
USA). AFM images of LNO films are shown in Figure S4.
In situ electrochemical impedance spectroscopy (EIS) measurements of microelectrodes
-200
sm
in diameter were performed using a microprobe station (Karl Sfss, Germany)
connected to a frequency response analyzer (Solartron 1260, USA) and dielectric interface
(Solartron 1296, USA). Temperature was controlled at 550 *C with heating stage (Linkam
TS1500, UK) and data were collected between 1 MHz to 1 mHz using a voltage amplitude of 10
mV. EIS testing temperature was calibrated with a thermocouple contacting the thin film surface
and deviation of
5 *C was observed. EIS experiments were completed between p(O2) of 10-3
atm and 1 atm. EIS data were analyzed using an equivalent circuit (Figure S3-6b), from which
the ORR resistance (RoRR) and surface oxygen exchange rate were obtained. EIS data of all
samples used in this study were found to be very similar in shape, and the detailed Nyquist plot
of the (100)teftagonai-oriented epitaxial LNO thin films with ~ 14 nm at 550 0 C is shown in Figure
S3-6c.
In situ HRXRD was performed on a four-circle diffractometer (Panalytical) in an oxygen
partial pressure of 1 atm and a controlled temperature stage (DHS 900, Anton Paar). Silver paste
66
was used to adhere the thin film sample to the heating plate. The heating rate was -10 *C min
1
and the temperature was held for 20 minutes at each temperature (25 *C, 150 'C, 250 *C, 350 *C,
450 'C, and 550 C) before XRD data were collected. Sample realignment was conducted at
each temperature to maximize the XRD intensities. A full range 9-29 normal scan was collected,
then high-resolution 9-29 normal scans of LNO (2 0 0 )teagonal and YSZ
(00 2 )cubic
were collected.
Finally, high-resolution off-normal scans of LNO (I 0 3 )teragonal, LNO (1l 4 )teagonal and YSZ
(2 0 2 )cubic
peaks were obtained. As the thermocouple for this experiment was placed inside the
heating stage, a small difference between set and actual temperatures on the sample surface
cannot be ruled out.
Ex situ auger electron spectroscopy (AES) was conducted with Physical Electronics 700
Scanning Auger Nanoprobe (PHI, USA) operating at an accelerating voltage of 10 kV to analyze
the surface chemistry change of the LNO films after heat treatment. The films were annealed at
550 'C for 6 hours in an oxygen partial pressure of 1 atm before AES data were collected. The
AES data were collected from three different locations (10 pm x 10 pm) selected on a sample in
an ultra high vacuum chamber. Details about AES measurement can be found in SI.
3.3 Results and Discussion
Normal X-ray diffraction (XRD) data (Figure 3-la) of LNO films collected at room
temperature clearly show the presence of the (00)teragonal (I is even) peaks of LNO and
(0)c.bi
(I is even) peaks of GDC and YSZ, which indicates the synthesis of (200)tgonai-oriented LNO
films having (200)teatgonaLNO//(002)eubicGDC//(002)cubicYSZ.
Off-normal phi-scan analysis
(Figure 3-1b) allowed us to identify the in-plane crystallographic relationships between LNO,
GDC and YSZ, LNO and GDC, as shown in Figure 3-1c. The data showed that
<2 0 0 >ttragonaiLNO was rotated by 45 0 with respect to the <002>ubicGDC, which is expected due
to lattice matching: c(LNO) ~ 3,F2 a(GDC). The measured lattice parameters a and b of the
LNO films were found to be nearly identical (a = 3.939 A and b = 3.932 A for 390 nm at RT,
details about all measured lattice parameters in Figure S3-1), which is in reasonably good
agreement with literature data (a = 3.864 A and b = 3.867 A for (001)tetagonai-oriented epitaxial
LNO thin film with ~ 330 nm 38 ). The LNO films have in-plane compressive strains and tensile
67
strains normal to the film surface at room temperature (Figure S3-3), where the magnitude of
strains decreases with increasing film thickness. The strains can be attributed to the large
differences in the atomic spacing values in the (1OO)LNO orientation between LNO and GDC,
where lattice mismatch in the a-b plane is 0.8% (aLNo = 3.862 A39 and aLNO
5.418
A 40) and that
in the c plane is 9.4% (CLNO
=
12.685 A 39 and CLNO
F-
2
3 -2aGDC =
aGDC, aGDC
11.497 A4 0),
as shown in Figure S3-7. The relaxed lattice parameters, a and , of LNO films at room
temperature were found to not change significantly with different film thicknesses, having values
in range of 3.866 - 3.869 A and 12.699 - 12.708 A (Figure S3-2), respectively.
0
(a)
(b)LN(1)
N0
(C)a
N
La
Ni
LNO
GDnnsC (202)CGc
Z~
C
j
C
20
C
4d
nm
4nm
40
60
80
100
40sz
S(202)
-180-900
90180
220
Figure 3-1. High resolution X-ray diffraction (CuK a) analysis. (a) Normal XRD of the
(l00)tetragonai-oriented epitaxial LNO thin films (~A4, ~42, ~178, and ~390 nm), and (b) Offnormal XRD of the (l00)tetragonaioriented epitaxial LNO thin films (~14 nm), GDC, and YSZ (c)
schematic of the crystallographic rotational relationships among the LNO (2 0 0 )tetragonal, GDC
(.
2 )cubic
and YSZ (0 0 2 )cubic.
In situ heating HRXRD was conducted to show that the LNO thin films were structurally
stable upon heating to 550 C0 in an oxygen partial pressure of 1 atm. Upon heating to 550 C, no
phase change was observed, and only peak shifts towards low diffraction angles associated to the
thermal expansion of LNO were noted, which resulted in increasing lattice parameters and
68
relaxed unit cell volume as a function of increasing temperature, as shown in Figure 3-2. The
volumetric thermal expansion coefficients (TECs) of the LNO thin films (11.1 x 10-6 ~ 11.9
10-6 K') were in good agreement with those reported in literature (11.0 x 10-6 ~ 11.6
x
x
10-6 K-1
for bulk LN0 3 7 ). Interestingly, the smaller relaxed unit cell volume of the LNO thin films at
smaller thicknesses became increasingly evident with increasing temperature. This observation
suggests that the oxygen overstoichiometry of these films decreases with decreasing the film
thickness at elevated temperatures, which will be discussed later.
0
CD
N
0-a
Zso
.
0a
y
0
Sz
-
(b)
G 194 0 14 nm
E
042 nm
%No
Co
N
-
.
195 U
193
'A171
j192 V39C
550c*.
*
AW
n
191
0
350 PC
0190
0
20
40
60
80
100
W 189
a
p(02) =1atm
300 400 500 600 700 800 900
Temperature / K
26/0
Figure 3-2. Structural stability and unit cell volume of the (1 00)tetragonai-oriented epitaxial LNO
thin films (a) In situ HRXRD data (CuK ,) of normal scan in the 6-26 Bragg-Brentano geometry
as a function of temperature with
-
178 nm LNO film, showing no phase change upon heating at
a p(O2) of 1 atm. The starred (*) peaks originated from the heater, and the peaks of the LNO
film, GDC buffer layer and YSZ substrate are indexed to the tetragonal, cubic and cubic
structure, respectively. (b) In situ HRXRD data of the unit cell volume of the (1 OO)tetagonaloriented epitaxial LNO thin films with
-
14, ~ 42, -178 and -390 nm as a function of
temperature in ap(02) of 1 atm.
To investigate the change of surface chemistry as a function of film thickness, ex situ
AES was conducted with the LNO thin films after annealing at 550 'C in an oxygen partial
69
pressure of 1 atm. Figure 3-3a and 3-3b show the changes in the surface La and Ni cations
spectral of each film. This clearly indicates that there is no significant change in the surface La
and Ni cations as a function of film thickness. The relative surface Ni/La ratios were found to be
(a)
La
(b)
NI
L2.0
.
.
.
-0-
.
.
.
.
independent on film thickness (Figure 3-3c).
W
c
Ni/La
S1.o
--
14 nm
14 nm
- 42 nm
178 nm
-
390nm .
580
600
I 0.5.
42 nm
-
178 nm
620
640
Kinetic Energy I eV
800
0
390nm
820
840
860
Kinetic Energy / eV
880
Z .0 1
0
100
200
300
400
Film Thickness / nm
-
Figure 3-3. Ex situ AES data of the (1 00)tetragonar-oriented epitaxial LNO thin films with ~ 14,
42, -178 and -390 nm annealed at 550 'C in an oxygen partial pressure of 1 atm. (a) La cation
variation and (b) Ni cation variation at three different locations on the LNO film surface. (c) The
change of the surface Ni/La ratio as a function of film thickness. Normalized to the value
obtained at ~ 14 nm LNO film.
EIS data collected from the LNO thin films with ~ 14, ~ 42, -178 and -390 nm at 550 'C
in an oxygen partial pressure of 1 atm are shown in Figure 3-4a. The real impedance of the
predominant semicircle decreased significantly with decreasing thickness of the LNO films. In
addition, the predominant semicircle was found to increase with decreasing oxygen partial
pressure, p(O2). Representative EIS data collected from the 14 nm film measured at 550 'C as a
function of p(O2) are shown in Figure 3-4b. Considering the fact that the film thicknesses are
much smaller than the critical thickness (350 Vm estimated for LNO single crystal 21 at 550 0C),
the p(02)-dependent impedance responses suggest that the surface oxygen exchange kinetics
governs the oxygen electrocatalysis on the LNO film surface. Figure 3-4c shows the kq of the
LNO thin films with different thicknesses (- 14,
-
42, -178 and -390 nm) at 550 'C as a
function of p(O2), which was obtained from the real impedance of the semicircle. The k values
70
of the LNO thin films were found to decrease with increasing thickness. Assuming that k' can be
approximated as k*l, it is noted that the k of the thinnest LNO film having a thickness of 14 nm
is higher than that of extrapolated for LNO bulk 30 by about two orders of magnitude at 1 atm.
Interestingly, the trend is in contrast to no thickness dependency of k* for the (0)tetmgonaloriented epitaxial LNO thin films by Burriel et al. , regardless of the surfaces perpendicular to
either the a-b plane or c-axis. The difference might be attributed to the fact that the LNO thin
3
have lower lattice mismatch with the substrate (1 % on STO and -0.19
%
films of previous work
on NGO at room temperature and no high-temperature data available). On the other hand, the
lattice mismatch between the b-c plane parameter in the LNO thin films and GDC in this work is
considerably large (Aa ~ Ab = 6.96 %, and Ac = 4.82 % on GDC at room temperature and Aa ~
Ab = 7.68 %, and Ac = 5.32 % at 550 'C for 14 nm film). Therefore, it is proposed that large
tensile strains on the (I00)tetragonai-oriented epitaxial LNO thin films can cause the thicknessdependent kq, as shown in Figure 3-5.
71
p(0 2 )=1atm
Z40
*
390 nm
80
-T
o
60
XV
-
20
E
120
150
0.1 atm
at
0
30
Real Impedance / kn
- 0
S 0.01 atm
40
-
42nm020 ,
&14 n
E
'
D
30
60
90
0
0.
E
178 nm
550 0 C
14 nm
atm
100
80
60
0.O
'
e 100
c
C: 120 -(b)
550 ;C
-
120 (a)
-
*
60
90
120
150
Real Impedance / k2
uuw
0-6
010
E
14 nm
390 nm
LL 500
10-8
42 nm
10.
Bulk
42 nm
400
178 n
14 nm-
390 nm
300
10.10
1 0"
600
-3
-2
-1
550 C
0
:UU Ommmmmmdhm-
-3
-2
-1
550 OC
0
log p(0 2) / atm
log p(0 2) / atm
Figure 3-4. Electrochemical impedance spectroscopy (EIS) results of microelectrodes for the
(100)tetragona-oriented epitaxial LNO thin films with ~ 14, ~ 42, -178 and ~390 nm at 550 0 C (a)
Nyquist plot of the (100)tetragona-oriented epitaxial LNO thin films with ~ 14, ~ 42, ~178 and
~390 nm in 1 atm p(O2). (b) Nyquist plot of the (100)tetragonai-oriented epitaxial LNO thin films
with ~ 14 nm as a function of p(O2). (c) Oxygen partial pressure dependency of the surface
exchange coefficients, kq of the (1 00)tetragonal-oriented epitaxial LNO thin films with ~ 14, ~ 42,
-178 and -390 nm calculated from EIS spectra collected at 550 *C. Extrapolated bulk k*
(approximately equivalent to k ) 4 1 value at 550 0C obtained from previous data of (*-gray)
Skinner et al.30 is plotted for comparison. (d) Oxygen partial pressure dependency of volume
72
specific capacitance (VSC) of the (100)tetragonai-oriented epitaxial LNO thin films with ~ 14, ~ 42,
~178 and -390 nm calculated from EIS spectra collected at 550 'C.
10
Mt4 film
0(100)-Orine
E
55
00
0
100
200
300
400
Thickness I nm
Figure 3-5. Thickness dependency of the k of the (10O)tetragonai-oriented epitaxial LNO thin films
with ~ 14,
-
42, -178 and -390 nm calculated from EIS spectra collected at 550 'C in 1 atm
p(O2). The inset shows the schematic of the surface exchange in the (100)tetragonal-oriented
epitaxial LNO thin films.
The greater difference between the relaxed and constrained lattice parameters of thinner
LNO films were found to accompany with greater surface oxygen exchange kinetics, as shown in
Figure 3-6a. The constrained a parameters of the LNO thin films decrease very slightly with
increasing film thickness whereas the constrained c parameters increase considerably with
increasing film thickness, which leads to increasing tensile volumetric strains with decreasing
film thickness (Figure 3-6b). Such strains can induce changes in the oxygen nonstochiometry, 6,
in LNO. To the first approximation, the nonstoichiometry, 6, of these LNO thin films can be
estimated from the unit cell volume based on the established correlation between the unit cell
volume of bulk LNO and
.37 As the unit cell volume of the LNO thin films increases as
increasing film thickness (Figure 3-2b), the extrapolated 6 value of the LNO films is larger at
greater thicknesses, as shown in Figure 3-6. The trend is further supported by the fact that
73
volume specific capacitances (VSCs) of LNO films extracted from EIS data (for details see the
Supporting Information), which correspond to changes in 8 induced by changes in the electrical
potential, were found to increase with increasing film thickness, as shown in Figure 3-4d.
Considering that Nakamura et al.4 have reported that thermodynamic driving force for the
formation of interstitial oxygen in bulk LNO decreases with increasing 8, it is proposed that the
increasing surface oxygen exchange kinetics of thinner LNO films can be attributed to the
decreasing barrier of interstitial oxygen formation and release associated with the larger
thermodynamic driving force for interstitial oxygen formation at lower 8, as shown in Figure 36d. Although the surface oxygen exchange kinetics are not correlated strongly with the electrical
conductivity in LNO thin films,2 3'
38
the influence of hole concentration and mobility of Ni
cations on the surface oxygen exchange kinetics cannot be excluded, which requires further
investigation.
74
(a)4.6
(b)
13.0
0
Bulk 390 nm 178 nm 42 nm
4.4
*0-0-0
0 4.2
O~
550 OC
14 nmn
.0 12.8 xCL
7
0 550
p(0 2)=latm
E
6 0
100
5
12.611
0,
0
12.4w
0 4.0
200
0
4
X-
C
.9
01
5.
.9
300
L.
12.2 CD
3.8
I')
-
_'% W1
k4 /
(.')(
3
0.
fl>
...
108
-
10
0.35
108
14 nm
03 14 nm
0.25
Bulk LNO
193.4
ssbo6
55606
p(0 2)=latm
'A
193.6
01
V
0.30
E 193.8
-
0
0.20 01
C0
E
E 10
42 nm 0
0.15
s
O 193.2 IV
P
: 193.0 - 0
178 nm
390 nm
Bulk LNO
550 0C
p(02 )=latm
-100
200
3 )0
J
0.10
W.Vu
400
18100.08
Thickness / nm
Figure 3-6. (a) Constrained (U-black,
I0
10"
k / cm sec-1
cm sec-1
o1<4.0
2
. ...
..MM
-
550 OC
P(C 2)=Iatm- 10 8
CD
-
0.12
0.16
-
0.20
-MMMMMMM
0.24
6
*-blue)
and Relaxed (Li-black, O-blue) lattice
parameters of the (100)tetragonal-oriented epitaxial LNO thin films as a function of the surface
oxygen exchange kinetics at 550 *C in 1 atm p(O2). Extrapolated bulk a (*-black), c (*-blue)
lattice parameters and k* value at 550 'C obtained from previous data of Skinner et al.3 0, 3 9 are
plotted for comparison. (b) Thickness and volumetric strain of the (1 00)tetragonal-oriented epitaxial
LNO thin films as a function of the surface oxygen exchange kinetics at 550 'C in 1 atm p(O2).
(c) Unit cell volume and 6 extrapolated from Nakamura et al.3 as a function of the (1
OO)tetagonal-
oriented epitaxial LNO thin films thickness at 550 *C in 1 atm p(O2). Extrapolated bulk unit cell
volume and 6 at 550 *C obtained from previous data of (gray line) Nakamura et al.3 are plotted
for comparison. (d) kq of the (100)tetragona-oriented epitaxial LNO thin films as a function of 6
extrapolated from Nakamura et al.3 at 550 0 C in 1 atm p(O2). Extrapolated bulk k* value and 6
75
obtained from previous data of (*-gray) Skinner et al. 3 0 and Nakamura et al.3 7 , respectively are
plotted for comparison.
3.4 Conclusions
We have successfully deposited the (l00)tetragonai-oriented epitaxial LNO thin films with
four different thicknesses. The k values of the LNO thin films decrease with increasing film
thickness, and such a trend is not observed in (00 l)tetragonai-oriented epitaxial LNO thin films. 2 3
Ex situ AES shows that there is no significant change of the surface chemistry as a function of
film thickness before and after exposure to elevated temperatures. In situ HRXRD reveals that
the unit cell volume of the LNO thin films increases with increasing the film thickness at 550 *C,
which indicates oxygen nonstiochiometry, 8, in La 2NiO 4+a decrease with decreasing LNO film
thickness. Our results demonstrate the key role of oxygen excess of RP phases on the oxygen
surface exchange process, where modifying the driving force to form interstitial oxygen by
strains is a new strategy to design highly active surface oxygen exchange materials for
applications such as SOFC cathodes or oxygen conducting membranes.
3.5 Supporting Information (SI)
Experimental Details
Target Synthesis.
La2NiO 4+8 (LNO) and Gdo 2 Ceo.80 2 (GDC) were prepared by the
Pechini methods 43. La(N0 3)3 *6H20 and Ni(N0 3)2 96H2 0, and separately
Gd(N0 3)3 and
Ce(N0 3)3 were dissolved in de-ionized water with ethylene glycol, and citric acid (SigmaAldrich, USA) mixture to synthesize LNO and GDC respectively. After esterification at 100 *C,
the resin was charred at 400 *C and finally calcined at 1000 *C in air for 12 hours. Pulsed laser
deposition (PLD) target pellets with 25 mm diameter were subsequently fabricated by uniaxial
76
pressing at 50 MPa. The LNO and GDC pellets were fully sintered at 1,300 "C in air for 10
hours and 1,100 "C in air for 14 hours, respectively.
The Relaxed
Relaxed lattice parameter determination by in situ IRXRD.
lattice
parameter 8 and ^ are derived from the following equation (where a and e are the relaxed lattice
Stt)16, 17,44 (c-c^) =
parameters for the film in an unstrained state),assuming
-2v
(a-ai)
tl '9= 3
3.286, 3.286, 3.286, 3.283, and 3.282 A for 298, 423, 523, 623, 723, and 823 K, respectively and
The in-plane strain is given by: Ecc =
and the out of plane strain by:
Unit cell volume calculation using in situ HRXRD data.
LNO is a tetragonal structure,
v
=
0.316 17, 4'
Eaa =
(a-a)
a
where the c-axis lattice parameter can be determined from the LNO (10 3 )tetagona1 reflection, and
assuming a- and b-axis lattice parameters are equivalent based on the measured HRXRD data,
they can be determined from the geometrical relationship between the c axis and the LNO (200)
.
2
tetragonal reflection. Once c and a = b are determined, the unit cell volume can be obtained by c - a
The volumetric strain are calculated using the following equation:
Volumetric strain (%) = (Monstrained -
Vrelaxed)
relaxed
Microelectrodes Fabrication.
In situ electrochemical
x 100
impedance spectroscopy (EIS)
measurements were conducted to probe ORR activity on geometrically well-defined LNO
microelectrodes fabricated by photolithography and acid etching, where sintered porous Pt
sintered onto the backside of the YSZ substrate served as the counter electrode. OCG positive
photoresist (Arch Chemical Co., USA) was applied on the LNO surface and patterned using a
mask aligner (Karl Stiss, Germany, A = 365 nm). The photoresist was developed using Developer
934 1:1 (Arch Chemical Co., USA) and the thin films were etched in hydrochloric acid (HCl) to
remove LNO film excess and create the circular microelectrodes (diameters -50 pm, -100 Pm,
-150
pm, and
-200 pm, exact diameter
determined by optical microscopy).
Before
electrochemical testing, microelectrode geometry and morphology was examined by optical
microscopy (Carl Zeiss, Germany) and atomic force microscopy (AFM) (Veeco, USA). AFM
77
measurements after acid-etching of the LNO film revealed thickness of -14, -42, -178, -390 nm
for 1,000 pulses, 5,000 pulses, 15,000 pulses and 35,000 pulses, respectively.
Electrochemical Characterization.
Figure S3-6b and Figure S3-6c detail the equivalent
circuit and corresponding Nyquist plot for this experimental system. ZView software (Scribner
Associates, USA) was used to construct the equivalent circuit and perform complex least squares
fitting. The EIS data were fitted using a standard resistor (RI) for HF and resistors (R2) in parallel
with a constant phase elements (CPE2) for MF and LF (Ri-(R2/CPE2)-(RoRR/CPEo0)). Based on
the p(O2) dependence of the three features, physical or chemical process with regard to each
frequency range can be determined.6'46-48 The HF feature (104
-
105 Hz) was found unchanged
with p(O2), and its magnitude and activation energy (~1.15 eV) were comparable to those of
oxygen ion conduction in YSZ reported previously 49 . The MF feature (10' - 104 Hz), which was
found to have a p(O2) independent feature, was attributed to interfacial transport of oxygen ions
between the LNO film and the GDC layer. In addition, the magnitude of its capacitance was
relatively small (~ 10-6 F) compared to the LF feature (-10-3 F). The LF feature (10-2
-
10' Hz)
was found to have a strong p(O2) dependence. The resistance of the LF feature drastically
increases as oxygen partial pressure decreases. In the case of thin film samples, the magnitude of
capacitance is due to the oxygen content change in the films. Therefore, the electrode oxygen
surface reaction corresponds with the LF feature. We obtained values for Ro. and knowing the
area of the microelectrode
(Aelectrode
specific resistance (ASRoR = Rop
=0.25
t delectrode 2 ).
Then, we can determine the ORR area
- Aelectrode). The electrical surface exchange coefficient (Ic),
which is comparable to k*, 4l was determined using the expression, 0 5
k =
RT /
4F2RO0RAe1.ctrodeCo
(1)
where R is the universal gas constant (8.314 J mol-1 K71 ), T is the absolute temperature, F is the
Faraday's constant (96,500 C mol-1), and c0 is the lattice oxygen concentration in LNO where
co = (4+6)/Vm,
(2)
V, is the molar volume of LNO at room temperature. In this study, c. was calculated with 6
extracted from previous reported values.42
VSC, indicative of changes in the oxygen nonstoichiometry induced by changes in the
electrical potential, can be obtained from EIS data via the expression5 2
VSC
=
[I/(Aelectrode x
thickness)]((RoRR) 1 ~"Q)"",
78
(3)
where
Q
is the non-ideal "capacitance", and n is the non-ideality factor of CPE. The fitted
values of n for semi-circle CPEoRR were found to range from
-
0.96 to 1.0 over the entire p02
range examined (n =1, ideal).
Calculation details of oxygen nonstoichiometry.
Lattice parameters of LNO depend
on oxygen nonstoichiometry and temperature. Nakamura et al.3 have formulated a relationship
between the lattice parameters, T, and 8 expressed by the equations below.
a(T, 6) = aref. + (a )6AT + ( )a
c(T, 6) = cref. +
aT
a ) = 3.92x0
, ac
and (0)T = -0.0626
92x10 and (a)T
0-616
By the first approximation, it was assumed that the partial differential terms are independent of T
and S. In the calculation, the reference conditions were T = 873 K and 6 = 0.08. The relationship
between bulk unit cell volume and 8 at 823 K was calculated using the same reference
conditions. Then, 6 of LNO films were extracted using the relaxed lattice parameter and unit cell
volume determined from in situ XRD.
Experimental details of ex situ AES.
In AES, the obtained energy spectrum for a
particular element is always situated on a large background (low signal-to-noise ratio), which
arises from the vast number of so-called secondary electrons generated by a multitude of
inelastic scattering processes. To obtain better sensitivity for detection of the elemental peak
positions, the AES spectra from this study are presented in the differentiated form. Elemental
quantification of AES spectra utilized relative sensitivity factors (RSFs) of 0.059, 0.227, and
0.212 for Lamm, NiLMM, and OKLL, respectively, as supplied by the AES manufacturer (Physical
Electronics). In addition, the Inelastic-Mean-Free-Path (IMFP) was calculated to correct signal
intensity for their different IMFPs (information depth). IMFPs were calculated using the NIST
Standard Reference Database 71 "NIST Electron Inelastic- Mean-Free-Path Database" version
1.2. The software program provides the ability to predict the IMFP for inorganic compounds
supplying the stoichiometric composition of La (2), Ni (1), 0 (4), the number of valence
electrons per molecule (assumed to be 40), the density (7.037 g/cm3)53 and a band gap energy
(for which we are assuming 0 eV as LNO is metallic like at high temperatures; additionally when
assuming a band gap of an insulator 5 eV, the IMFP increases by -0.03 nm). The IMFP for La
79
and Ni were determined to be 1.361 nm and 1.562 nm, respectively. A relative depth-scaling
factor (s) was determined as:
o-i=
I exp - xdx,
Ai
0 Ai
(4)
where A, is the IMFP, yielding aNi = 0.58, and ULa = 0.63. The intensities from different elements
were scaled using Iscaled - Imeasured*Si/Si.
80
-12.6
S12.4
E 12.2
12.0
~11.8
-
-
CD a lattice
0
blattice
c lattice
.E 4.0
03.900
0
100
200
300
400
Thickness I nm
Figure S3-1. Constrained lattice parameters of the (100)tetragonal-oriented epitaxial LNO thin films
with 14, 42, 178, and 390 nm extracted from normal and off-normal HRXRD data at room
temperature. The constrained normal and in-plane lattice parameters of the LNO films were
calculated from combining the interplanar distances of the
(11
4
)tetragonal peaks.
81
( 2 0 0 )tetragonal,
(10 3 )tetragonal and
p(0 2 ) = 1atm
(a)
0-
3.89
a,
E
(U
(U
0.
3.88
12.78
0314 nm
,
a,
E
2 12.74
cc
W
12.72
3.87
(U
Z 12.70
3.86
D.
12.76 . 042 nm
1
"relaxed
p(0 2 ) = L atm
0114 nm
V
042nm
178 nm
390 nm
-(b)
I
.
3.90
wO0
0
178 nm
390 nm
03
0)
4- "
300 400 500 600 700 800
relaxed
300 400 500 600 700 800
Temperature / K
Temperature / K
-
-
Figure S3-2. (a) Relaxed
a
,
lattice parameter, and (b) C lattice parameter of the LNO films as a
function of temperature in an oxygen partial pressure of 1 atm. For determining the relaxed film
lattice parameter a and C, we used the equation:
2
=) 1-v
2
(a)
a
assuming a/i
3 =
3.284, 3.286, 3.286, 3.286, 3.283, and 3.282 A for 298, 423,
523, 623, 723, and 823 K, respectively and v =0.3.
82
-
-
-
6
#OI
-4 S
z2 0
0
100
200
300
400
Film Thickness / nm
Figure S3-3. Normal and In-plan strain of the (1 OO)tetragonai-oriented epitaxial LNO thin films as a
function of the film thickness at room temperature. Normal and in-plan strains are calculated
using the equation: Normal strain, Eaa =
(a-d)
and in-plane strain, Ec-
83
(C__0
Figure S3-4. AFM measurements of the as-prepared (a) LNO films ~ 14 nm with RMS of
0.216nm, (b) LNO films ~ 42 nm with RMS of 0.341nm, (c) LNO films ~178 nm with RMS of
0.353nm, and (d) LNO films ~ 390 nm with RMS of 0.407 nm.
84
Figure S3-5. Scanning electron microscopy (SEM) images of (a) LNO films ~ 14 nm, (b) LNO
films ~ 42 nm, (c) LNO films ~ 178 nm, and (d) LNO films ~ 390 nm annealed at 550 *C in an
oxygen partial pressure of 1 atm for 6 hours.
85
(b)
(a)
R1
R2
RORR
02
"EV
(HF)
Substrate
200
(MF) interface
(LF) Film
C:
0/
0
0.01
..
0150
V5
0. 4
E
3 HF
1
*~10
r 2
S
0.100
IU2(
10
0
MF
to
-/
101 Hz)
O-0,
E
0
0
50
HHz)
LNO - 14 n
6
8 10 12 14 16
Real Impedance IkQ
SO OC
0.1 atm
.1tm
LF
(10-2 to
175
E
0.01 atm-
02
R
/
(c)
CPE
C-
LNO
14 nm
100 150 200 250 300
Real Impedance / k(
Figure S3-6. (a) Schematic of a LNO/GDC/YSZ(001)/porous Pt samples and electrochemical
testing configuration (not drawn to scale), and (b) equivalent circuit (RI = YSZ electrolyte
resistance, R 2 = electrode/electrolyte interface resistance'7, RORR = ORR resistance, CPE =
constant phase element) used to extract ORR kinetics, and (c) Nyquist plot of the (1
oriented epitaxial LNO thin films with ~ 14 nm at 550
~ 105 Hz, MF: 103 ~ 104 Hz, and LF: 10-2 _ 10 3 Hz).
86
0
OO)tetragonav-
C; inset shows a magnification (HF: 104
(a)(b)
LaNN)O4+6
1 4
GdO 2
4
7Ce
1 ..3GdO 24
.
c.
La 2NIO4*
Figure S3-7. Schematic of two different orientations of LNO on GDC (a) (100)tetragonai-oriented
epitaxial LNO thin film and (b) (001)tetragonal-oriented epitaxial LNO thin films.
87
3.6 References
1.
L. da Conceicao, C. R. B. Silva, N. F. P. Ribeiro and M. Souza, Mater. Charact., 2009,
60, 1417-1423.
2.
R. A. De Souza and J. A. Kilner, Solid State Ion., 1998, 106, 175-187.
3.
R. A. De Souza and J. A. Kilner, Solid State Ion., 1999, 126, 153-161.
4.
A. Endo, H. Fukunaga, C. Wen and K. Yamada, Solid State Ion., 2000, 135, 353-358.
5.
T. Ioroi, T. Hara, Y. Uchimoto, Z. Ogumi and Z. Takehara, J. Electrochem. Soc., 1997,
144,1362-1370.
6.
S. B. Adler, Chem. Rev., 2004, 104, 4791-4843.
7.
S. B. Adler, X. Y. Chen and J. R. Wilson, J. Catal., 2007, 245, 91-109.
8.
A. J. Jacobson, Chem. Mat., 2010, 22, 660-674.
9.
Y. L. Lee, J. Kleis, J. Rossmeisl and D. Morgan, Phys. Rev. B, 2009, 80.
10.
Z. P. Shao and S. M. Haile, Nature, 2004, 431, 170-173.
11.
B. C. H. Steele and A. Heinzel, Nature, 2001, 414, 345-352.
12.
C. W. Sun, R. Hui and J. Roller, J. Solid State Electrochem., 2010, 14, 1125-1144.
13.
S. M. Hashim, A. R. Mohamed and S. Bhatia, Adv. ColloidInterface Sci., 2010, 160, 88100.
14.
S. B. Adler, Solid State Ion., 1998, 111, 125-134.
15.
E. J. Crumlin, S. J. Ahn, D. Lee, E. Mutoro, M. D. Biegalski, H. M. Christen and Y.
Shao-Horn, J. Electrochem. Soc., 2012, 159, F219-F225.
16.
E. J. Crumlin, E. Mutoro, Z. Liu, M. E. Grass, M. D. Biegalski, Y. L. Lee, D. Morgan, H.
M. Christen, H. Bluhm and Y. Shao-Hom, Energy Environ. Sci., 2012, 5, 6081-6088.
17.
G. J. la 0, S. J. Ahn, E. Crumlin, Y. Orikasa, M. D. Biegalski, H. M. Christen and Y.
Shao-Horn, Angew. Chem.-Int. Edit., 2010, 49, 5344-5347.
18.
J. Mizusaki, Y. Mima, S. Yamauchi, K. Fueki and H. Tagawa, J Solid State Chem.,
1989, 80, 102-111.
19.
V. G. Sathe, A. V. Pimpale, V. Siruguri and S. K. Paranjpe, Journal of PhysicsCondensedMatter, 1996, 8, 3889-3896.
20.
S. Y. Wang, J. Yoon, G. Kim, D. X. Huang, H. Y. Wang and A. J. Jacobson, Chem. Mat.,
2010, 22, 776-782.
88
21.
J. M. Bassat, P. Odier, A. Villesuzanne, C. Marin and M. Pouchard, Solid State Ion.,
2004, 167, 341-347.
22.
E. Boehm, J. M. Bassat, P. Dordor, F. Mauvy, J. C. Grenier and P. Stevens, Solid State
Ion., 2005, 176, 2717-2725.
23.
M. Burriel, G. Garcia, J. Santiso, J. A. Kilner, J. C. C. Richard and S. J. Skinner, J
Mater. Chem., 2008, 18, 416-422.
24.
A. Chroneos, D. Parfitt, J. A. Kilner and R. W. Grimes, J. Mater. Chem., 2010, 20, 266270.
25.
A. R. Cleave, J. A. Kilner, S. J. Skinner, S. T. Murphy and R. W. Grimes, Solid State
Ion., 2008, 179, 823-826.
26.
G. Garcia, M. Burriel, N. Bonanos and J. Santiso, J. Electrochem. Soc., 2008, 155, P28P32.
27.
V. V. Kharton, A. P. Viskup, E. N. Naumovich and F. M. B. Marques, J. Mater. Chem.,
1999, 9, 2623-2629.
28.
J. A. Kilner and C. K. M. Shaw, Solid State Ion., 2002, 154, 523-527.
29.
G. Kim, S. Wang, A. J. Jacobson and C. L. Chen, Solid State Ion., 2006, 177, 1461-1467.
30.
S. J. Skinner and J. A. Kilner, Solid State Ion., 2000, 135, 709-712.
31.
E. J. Crumlin, E. Mutoro, S. J. Ahn, G. J. la 0, D. N. Leonard, A. Borisevich, M. D.
Biegalski, H. M. Christen and Y. Shao-Horn, J. Phys. Chem. Lett., 2010, 1, 3149-3155.
32.
E. Mutoro, E. J. Crumlin, M. D. Biegalski, H. M. Christen and Y. Shao-Horn, Energy
Environ. Sci., 2011, 4, 3689-3696.
33.
A. Yamada, Y. Suzuki, K. Saka, M. Uehara, D. Mori, R. Kanno, T. Kiguchi, F. Mauvy
and J. C. Grenier, Adv. Mater., 2008, 20, 4124-+.
34.
J. D. Jorgensen, B. Dabrowski, S. Pei, D. R. Richards and D. G. Hinks, Phys. Rev. B,
1989, 40, 2187-2199.
35.
A. Demourgues, F. Weill, B. Darriet, A. Wattiaux, J. C. Grenier, P. Gravereau and M.
Pouchard, J. Solid State Chem., 1993, 106, 317-329.
36.
J. Rodriguezcarvajal, M. T. Fernandezdiaz and J. L. Martinez, Journal of PhysicsCondensedMatter, 1991,3, 3215-3234.
37.
T. Nakamura, K. Yashiro, K. Sato and J. Mizusaki, Solid State Ion., 2010, 181, 292-299.
89
38.
M. Burriel, J. Santiso, M. D. Rossell, G. Van Tendeloo, A. Figueras and G. Garcia, J.
Phys. Chem. C, 2008, 112, 10982-10987.
39.
S. J. Skinner, Solid State Sci., 2003, 5, 419-426.
40.
B. Matovic, S. Boskovic, L. Zivkovic, M. Vlajic and V. Krstic, in CurrentResearch in
Advanced Materials and Processes, eds. D. P. Uskokovic, S. K. Milonjic and D. I.
Rakovic, Trans Tech Publications Ltd, Zurich-Uetikon, 2005, pp. 175-179.
41.
J. Maier, Solid State Ion., 1998, 112, 197-228.
42.
T. Nakamura, K. Yashiro, K. Sato and J. Mizusaki, Solid State Ion., 2009, 180, 368-3 76.
43.
M. Pechini, U.S. PatentNo. 3,330,697, 1967.
44.
H. M. Christen, E. D. Specht, S. S. Silliman and K. S. Harshavardhan, Phys. Rev. B,
2003,68,4.
45.
D. Y. Noh, Y. Hwu, J. H. Je, M. Hong and J. P. Mannaerts, Appl. Phys. Lett., 1996, 68,
1528-1530.
46.
S. B. Adler, J. A. Lane and B. C. H. Steele, J Electrochem. Soc., 1996, 143, 3554-3564.
47.
T. Kawada, J. Suzuki, M. Sase, A. Kaimai, K. Yashiro, Y. Nigara, J. Mizusaki, K.
Kawamura and H. Yugami, J. Electrochem. Soc., 2002, 149, E252-E259.
48.
Y. L. Yang, C. L. Chen, S. Y. Chen, C. W. Chu and A. J. Jacobson, J Electrochem. Soc.,
2000, 147, 4001-4007.
49.
P. S. Manning, J. D. Sirman, R. A. DeSouza and J. A. Kilner, Solid State Ion., 1997, 100,
1-10.
50.
J. Maier, Physical Chemistry of Ionic Materials:Ions and Electrons in Solids John Wiley,
Chichester, England; Hoboken, NJ, 2004.
51.
J. Fleig and J. Maier, J. Eur. Ceram. Soc., 2004,24, 1343-1347.
52.
J. Fleig, Solid State Ion., 2002, 150, 181-193.
53.
A. J. Ellett, Oxygen Permeation and Thermo-Chemical Stability of Oxygen Separation
Membrane Materialsforthe Oxyfuel Process, Forschungszentrum Jilich, 2009.
90
Chapter 4.
Strontium Influence
catalysis of La2 ..xSr
on the Oxygen
(0.0
NiO 4
5 XSr
<
Electro
1.0)
Thin
Films
106
.71
1
SU
10
La2-xSrxNio4 6
0
~I
SS
.10
10 1
0.0
0.2
0.4
0.6
0.8
1.0
xsr
Reproduced in part with permission from Dongkyu Lee, Yueh-Lin Lee, Alexis Grimaud, Wesley
T. Hong, Michael D. Biegalski, Dane Morgan and Yang Shao-Horn, Strontium Influence on the
Oxygen Electro catalysis of La 2-xSrxNiO41 (0.0 5 xsr : 1.0) Thin Films, Journal of Materials
Chemistry A, 2014, 2 (18) 6480 - 6487, Copyright 2014 The Royal Society of Chemistry.
91
4.1 Introduction
Perovskites (ABO 3) with high electronic and ionic conductivity and catalytic
characteristics 1-3 have been studied intensively for solid oxide fuel cells (SOFCs) '
4-14
and
oxygen permeation membranes' 5 at high temperatures. A major barrier that limits the efficiency
of SOFCs and oxygen permeation flux is the slow kinetics of oxygen surface exchange on the
22
oxide surface. Mixed electronic and ionic conductors (MEICs) such as Lai-xSrxCoO3-8 (LSC)16and La1.xSrxCoi-yFeyO3-, (LSCF) 23 -2 6 are used commonly to promote the oxygen surface
exchange kinetics, where the substitution of strontium (Sr) for lanthanum (La) in the A-site
increases the surface exchange rates.27' 28
Ruddlesden-Popper (RP) oxides such as La2NiO 4+8 (LNO) can lead to higher oxygen
ionic conductivity relative to ABO 3 perovskites 291 3 2 and are interesting alternative cathode
materials for SOFCs. The RP structure can be described as the intergrowth of alternative NiO 2
and La2 O 2 rock salt layers or layers of LNO separated by LaO layers (Fig. 1), having in-plane
lattice parameters similar to that of perovskites in the pseudocubic notation. The oxygen
stoichiometric La2 NiO 4 can be easily oxidized with additional interstitial oxygen incorporated
into lanthanum tetrahedra in the La2 O 2 rock salt layer, which results in increasing the La-O bond
distance and in decreasing the nickel ionic radius by the formation of Ni3 +.33 , 3 4 The alternating
perovskite and rock salt layers in the RP structure can lead to different oxygen diffusion and
surface exchange kinetics along different crystallographic directions. For example, the in-plane
oxygen transport kinetics via interstitial oxygen diffusion in the LaO planes is considerably
higher by up to two orders of magnitude compared to the out-of-plane kinetics3 1 and the surface
exchange kinetics in-plane can be an order of magnitude greater than that out-of-plane for LNO
single crystals.
By substituting of a larger cation such as Sr2+ (1.31 A) for La3 + (1.22
A)
in
LNO, the structural stresses are released, which leads to decrease in the amount of additional
oxygen required to stabilize the structure and thus results in the reduction of the oxygen diffusion
coefficients. 3 6, 37 However, the influence of Sr substitution on the in-plane and out-of-plane
oxygen surface exchange kinetics of La2-xSrNiO4=8 (LSNO) is poorly understood due to
difficulties in the growth of single crystals of LSNO with high Sr substitution. 38
Few studies have reported the oxygen surface exchange kinetics of polycrystalline LSNO
with low Sr substitution (xsr = 0.1 and 0.2).394 Skinner et al.3 have shown no clear effect of
the Sr substitution (xsr = 0.1) on the surface exchange kinetics whereas Boehm et al.40 have
92
reported that the Sr substitution (xsr = 0.1 and 0.2) can decrease the surface exchange kinetics.
Synthesis of oriented RP films4 1 grown on single-crystal substrates has enabled the studies of the
oxygen surface exchange kinetics. For instance, employing out-of-plane-oriented LNO thin films
grown on the (00l)cubi-SrTiO3 (STO) and (11 0)cubi-NbGaO3 (NGO) substrates, Burriel et al.
have shown using secondary ion mass spectroscopy (SIMS) measurements that the in-plane
oxygen surface exchange kinetics can be much greater than the out-of-plane kinetics, which is in
agreement with measurements on single crystal LNO.
More recently, we have shown that the
surface exchange coefficient, k', increases with increasing tensile volumetric strains and
decreasing oxygen nonstoichiometry (6) in LNO thin films. 42 Since the formation enthalpy of the
oxygen interstitial has been shown to increase (less negative) with greater 6, the more negative
oxygen interstitial formation enthalpy at lower 6 suggests increase of the driving force to form
interstitial oxygen in the RP structure. Therefore, the surface exchange kinetics of in-planeoriented LNO thin films grown on (00)cubic-Y 2 0 3-stabilized ZrO2 (YSZ) can be enhanced by
increasing tensile volumetric strains.
In this study, we investigate how Sr substitution can affect the oxygen surface exchange
kinetics of LSNO thin films with a wide range of Sr content (xsr = 0.0, 0.2, 0.4, 0.6, 0.8, and 1.0)
grown on YSZ. The surface orientation of these LSNO thin films (normal to the film surface)
can be modified from the (10)tetra. (in-plane) to the (001)tetr. (out-of-plane) in the RP structure
by increasing the Sr content from 0.0 to 1.0. Such a change in the film orientation can be
explained by the reduction in the surface energy of the (001)tetm. surface with increasing Sr as
revealed from Density function theory (DFT) calculations. Ex situ Auger electron spectroscopy
(AES) indicates no formation of secondary phases across the Sr contents. We show a strong Sr
dependence of the oxygen surface exchange rate (k') for the LSNO thin films, which can be
attributed to the surface reorientation and the adsorption energy changes of molecular oxygen on
the La-La bridge sites in the RP structure.
4.2 Experimental Methods
Film Deposition
Pulsed laser deposition (PLD) was used to deposit the LSNO (xsr = 0.0,
0.2, 0.4, 0.6, 0.8, and 1.0) thin films on (001)ubi-Y 2
93
3-stabilized
ZrO 2 (YSZ). A -5 nm
gadolinium-doped ceria (GDC, 20 mol % Gd) was also deposited between the LSNO and the
YSZ as a buffer layer. Single crystal 9.5 mol% YSZ wafers with the (001)cubic orientation and
dimensions of 10 x 5 x 0.5 mm (MTI corporation, USA) were used as the substrate. Prior to
LSNO and GDC deposition, platinum ink (Pt) (#6082, BASF, USA) counter electrodes were
painted on one side of the YSZ and dried at 900 *C in air for 1 hour. The YSZ wafer was affixed
to the PLD substrate holder using a small amount of silver paint (Leitsilber 200, Ted Pella, USA)
for thermal contact. PLD was performed using a KrF excimer laser at A = 248 nm, 10 Hz pulse
rate and ~ 1.5 J/cm2 pulse energy under an oxygen partial pressure, p(O2) of 6.6
x
10-5 atm (50
mTorr) with 500 pulses of GDC (-5 nm) at 550 *C, followed by 5,000 pulses of LSNO (xsr = 0.0,
0.2, 0.4, 0.6, 0.8, and 1.0) at 650 *C. The film thicknesses were determined by atomic force
microscopy (AFM). Details of the thickness information are provided in the Electronic
Supplementary Information (ESI). The utilization of reflection high-energy electron diffraction
(RHEED) enabled an in-situ monitoring of the LSNO film growth. After completing the LSNO
deposition, the samples were cooled down to room temperature in the PLD chamber for -1 hour
under a p(O2) of 6.6 x 10-5 atm (50 mTorr). The synthesis details of LSNO and GDC PLD
targets can be found in the ESI.
High Resolution X-ray Diffraction (HRXRD)
The oxide phase purity and the thin film
orientations were investigated via high-resolution X-ray diffraction (HRXRD) using a four-circle
diffractometer (PANalytical, USA and Bruker D8, Germany). Measurements were performed in
normal and off-normal configurations. The LSNO thin films showed a structural reorientation as
a function of Sr content and the determination of in-plane atera. and out-of-plane ctetra. lattice
parameters was described below, where the subscript "tetra." denotes the tetragonal notation.
The LNO structure can be influenced greatly by the oxygen stoichiometry and fully oxidized
La 2NiO 4 .2 5 exhibits a monoclinic symmetry with space group C2 at room temperature 4' while
stoichiometric La2NiO 4 is orthorhombic with space group Bmab.44 The LSNO oxides up to xsr =
1.0 are well known to have tetragonal symmetry with space group 14/mmm.45' 46 However, the
oxygen nonsotichiometry of the LNO thin film is experimentally challenging to obtain precisely,
and the XRD data of single-crystalline thin films do not readily allow a precise refinement of the
symmetry of the LNO structure. Therefore, the LNO film was considered to have tetragonal
symmetry with space group 14/mmm, which differs from the orthorhombic by the loss of the
94
octahedral tilting (s 10)44 and the tetragonal symmetry is generally used to describe the system in
thin film studies. 42 The in-plane atet. parameter of LSNO normal to the film surface was
determined from the (2 0 0)tetra. peak positions and the out-of-plane cetra. parameter of the LSNO
thin films with xsr = 0 and 0.2 was determined from the off-normal (10 3 )teta. peak position. For
the LSNO thin films with
xsr
= 0.4, 0.6, 0.8, and 1.0, the in-plane atetr. parameter and the out-of-
plane ctetr. parameter were determined from the off-normal (I0 3 )tetra., and (00 6 )ted. peak
positions, respectively.
Atomic Force Microscopy (AFM) The
surface
morphology
was examined
by optical
microscopy (Carl Zeiss, Germany) and atomic force microscopy (AFM) (Veeco, USA). The
RMS (root-mean-square) roughness value of LSNO with
about 0.3 nm while that of LSNO with
xSr
xsr
= 0.0 and 1.0 was found to show
= 0.2, 0.4, 0.6, and 0.8 was found to show about 0.6
nm as shown in Fig. S4-5.
Electrochemical Impedance Spectroscopy (EIS) In
situ
electrochemical
spectroscopy (EIS) measurements of LSNO microelectrodes -200
impedance
pm in diameter were
performed using a microprobe station (Karl Stiss, Germany) connected to a frequency response
analyzer (Solartron 1260, USA) and dielectric interface (Solartron 1296, USA). Temperature was
controlled at 550 *C using a heating stage (Linkam TS1500, UK) and data were collected
between 1 MHz to 1 mHz using a voltage amplitude of 10 mV. EIS testing temperature was
calibrated with a thermocouple contacting the thin film surface and deviation of
5 *C was
observed. EIS experiments were performed as a function of p(O2) between 104 and 1 atm.
Multiple electrodes (at least three) of all films were measured by EIS at each temperature and
p(O2) to ensure that the EIS results were reproducible and representative. EIS data of all samples
used in this study were found to show nearly perfect semicircle impedances,' and typical features
in the Nyquist plots are shown in the schematic in Fig. S4-7c. EIS data were analyzed using an
equivalent circuit (Fig. S4-7b), where the high-frequency intercept corresponds to the oxygen ion
conduction resistance in YSZ and the mid-frequency feature is believed to result from the
17 19 4 2
interface between YSZ and GDC as reported previously. , ,
Considering the fact that the film thicknesses are much smaller than the critical thickness, tert,
defined as D*Ik*, where D* is the tracer oxygen diffusivity and k* is the tracer surface exchange
95
coefficient4 7 (350 gm estimated for the LN0 3 5 and 888 pm estimated for LSNO having
xsr =
0.240 at 550 *C), the p(0 2 )-dependent impedance responses suggest that the oxygen surface
exchange kinetics governs the oxygen electrocatalysis on the LSNO thin film surface. The
surface exchange rate was calculated from the resistance of the low-frequency semicircle using19'
42,48,49
q
RT
where R is the universal gas constant (8.314 J mol-'-K-'), T is the
.
4F 2 RLFAlecrode
absolute temperature (823 K), F is the Faraday's constant (96,500 C-mol- 1), Aelectrode is the area
of the microelectrode, and co is the lattice oxygen concentration in LSNO. Details about the EIS
testing procedure, data analysis, and co estimation can be found in the ESI.
Auger Electron Spectroscopy (AES)
Auger electron spectroscopy (AES) was conducted
to analyze the surface chemistry change of LSNO films after annealing on a Physical Electronics
700 Scanning Auger Nanoprobe (PHI, USA) operating at an accelerating voltage of 10 kV. The
films were annealed at 550 *C for 6 hours in an oxygen partial pressure of 1 atm before AES data
were collected. The AES data were collected from three different areas (10 gm x 10 gm) selected
on a sample in an ultra high vacuum chamber. Elemental quantification of AES spectra utilized
relative sensitivity factors (RSFs) of 0.059, 0.027, 0.227, and 0.212 for LaM>N, SrLmm, NiLMM,
and OKLL, respectively, as supplied by the AES manufacturer (Physical Electronics).
Density Functional Theory (DFT)
Spin-polarized Density Functional Theory (DFT)
calculations were preformed with the Vienna Ab-initio Simulation Package5 0' 51 using the
Projector-Augmented plane-Wave method5 2 with a cutoff energy of 450 eV. Exchangecorrelation was treated in the Perdew-Wang-91. 53 Generalized Gradient Approximation (GGA)
using the soft Os oxygen pseudopotential. The GGA+U calculations5 4 were performed with the
simplified spherically averaged approach," where the Uff (Uff = Coulomb U - exchange J) is
applied to d electrons (Ueff(Ni) = 6.4 eV). All calculations were performed in the ferromagnetic
state in order to use a consistent and tractable set of magnetic structures, and the Ni cations are in
the high spin state for xsr = 0.0 with even reduction on spin moment of each Ni cation when
increasing Sr (hole) doping. Details about the calculation for bulk LSNO, surface energy and
oxygen adsorption energy can be found in the ESI.
96
La or Sr
,o
LSNO
LSNO
C
b
45
Ni
II
45'
-0
Azry
e zr[Y
YSZ
C
-0
btaO
YSZ
'0
ba
.a or Sr
Interstitial
oxygen site
0
Interstitial
I oxygen site
Nij
b
q
La or Sr
Ib
aW
c
(a) (100),,r..-oriented LSNO thin film
(b) (001),
.-oriented LSNO thin film
Figure 4-1. Schematic of the crystallographic rotational relationships of LSNO (l00)tetra., GDC
(000cubic and YSZ (000cubic ((a) top) and LSNO (00l)tetra., GDC (OOlcubic and YSZ (000cubic ((b)
top). (10O)teta. top view of LSNO thin film ((a) bottom) and (001 )tetra. top view of LSNO thin film
((b) bottom).
4.3 Results and discussion
4.3.1 Structural Reorientation of the LSNO Thin Films with increasing Sr
content
97
Normal HRXRD data in Figure 4-2a clearly showed a structural reorientation of LSNO
thin films as a function of Sr content. The LNO film (xsr = 0.0) was found to have the (l00)tetra. (1
is even) peaks only and the (OOl)cubic (1 is even) peaks of GDC and YSZ, which indicates that the
LNO film grew epitaxially with the atefra.-axis perpendicular to the film surface. With
xsr
in
LSNO equal to and greater than 0.2, the films were found to exhibit not only (l00)teta. (I is even)
but also (0)tetra. (I is even) peaks, indicating the presence of both domains with the ateta.-axis
and ctetra.-axis perpendicular to the film surface. The intensities of LSNO (00l)tetra. peaks were
found to increase with increasing Sr content in LSNO while those of the (l00)tetra. peaks
decreased, where the intensity ratio between the (2 00 )tetra. and the (00 6 )tetra. peaks of the LSNO
thin films decreasing with increasing Sr is shown as an example in Figure S4-3a. This
observation suggests that Sr substitution of La in LSNO thin films promotes the (OOl)tetra.
orientation growth but suppresses the growth of the (l00)tetra. orientation. Only the (000tetra. (1 is
even) peaks were found for LSNO with
xsr
= 1.0, indicating that the film was single crystalline
with the cetra.-axis perpendicular to the film surface, as shown in Figure 4-2a. Off-normal phiscan analysis (Figure 4-2c) allowed us to identify the in-plane crystallographic relationships
among LSNO, GDC and YSZ. The [100]teta. direction in the atetr.-axis-oriented LSNO and the
[001]tetra. direction in the cteta.-axis-oriented LSNO were rotated by 450 with respect to the
2
[001]c.ubi direction of GDC. Moreover, as expected from increasing Ni oxidation from Ni + in
LNO to Ni3 + in LSNO, the (l00)teta. and (00btetra. interplanar distances of LSNO were shifted
towards higher diffraction angles with increasing Sr (Fig. S4-3b).
98
(b)
)
(a)
.W
.W
6
0
'45
x=N00.N0
04
0
Gc C
(
Y Z
(
2
0 2 )cubi
2
0 2)ti
LSNO..
.1(001)t..-oriented
o
00
x=0.2
C0
x0.
-j
-j
20
LNO or LSNO ( 1 0 3 )ttra.
40
60
80
100
(100)tet..-oriented LNO
LIL
-180 -90
0
90 180
26I*#/
Figure 4-2. HRXRD analysis (a) Normal XRD of the La 2-xSrNiO 4 5 (LSNO) thin films with 0 <
xsr < 1.0, (b) schematic of a structural reorientation from (100)tetra. to (001)tetra. orientation with
increasing Sr contents from LNO (Xsr = 0.0) to LSNO (XSr = 1.0), and (c) Off-normal XRD of
2 2
3
atetra.-axis-oriented LNO (10 3 )tetra. and ctetra.-axis-oriented LSNO (10 )tetragonal, GDC ( 0 )cubic,
and YSZ(202)cubic.
The relaxed lattice parameters
atetra.
and ^tetra. of LSNO thin films at room temperature
were found to be in the range of 3.81 - 3.91 A and 12.44 - 12.72 A, respectively. Details of the
lattice parameters are provided in the ESI. Interestingly, the relaxed lattice atetra. parameter was
found to decrease with the Sr content for 0.0 5 xsr 5 0.6 while it slightly increased with
increasing Sr for XSr > 0.6. In contrast, the relaxed lattice ^tetra. parameter was found to exhibit the
opposite trend, which results in a maximum in the
ctetra./atetra.
ratio (tetragonality) at
XSr
= 0.6, as
shown in Figure S4-1. As described for bulk LSNO, 46 the decrease of the in-plane parameters
is due to the oxidation of nickel and the increase of the out-of-plane Ctetra.
parameter is correlated with the replacement of La3 by Sr 2+. For xsr > 0.6, the opposite trend is
(atetra.
and
btetra.)
observed, which has been explained by Jahn-Teller distortion.
99
La2 -xSrNiO4 6
0.08
0.5
O
x
,~
Vrft
'
,
,~
,
46
-
12,n
2-x
,
(b) 1.0
10~
(001)
0.07
..
100
-
(a) o.09
o 0.0
I-
0.06
10' =
0
w
-1.
10~ O =
0.03
00
-----0.0
0.2
0.4
-2.5
0.6
0.8
(001)-
0.0
1.0
XSr
0 .2
-
0.02
0.01
(100)
-2.
-
-
-20
.
0
-0
.
0.04
-0.5
W
(100)
5 0.05
0.4
0.6
0.8
1.0
XSr
(d)
(C)
iibi
Perspective view of (100) surface adsorption
Perspective view of (001) surface adsorption
Figure 4-3. Sr content dependency of (a) strain energies (o-gray) of the La 2-xSrxNiO 4 6 (LSNO)
thin films with 0.0 < xsr < 1.0 calculated using strain energy density equation, surface energies,
and (b) adsorption energies of an oxygen molecule on the (100)tetra. (o-blue) and (001 )teta. (o-red)
surfaces of the LSNO thin films calculated by density functional theory (DFT). The surface
oxygen adsorption sites of (c) the (100)tetra. slabs and (d) the (001)tetra. slabs. Details of the DFT
modeling approaches, the LSNO slab models, and strain energy calculation are provided in the
ESI. Both surface energies and adsorption energies represent the anisotropic feature of the LSNO
thin films.
The change in the orientation of LSNO films with increasing Sr can be explained by the
difference in the surface energy of in-plane and out-of-plane atomic terminations in the RP
structure.41, 5,57 As shown in Figure 4-3a, the calculated surface energies (details of the DFT
modeling are provided in the ESI) of LSNO (100)teta. and (00)tetra planes were found to be
strongly dependent on the Sr content, where the surface energy of (100)tetra. plane was found to
100
decrease while that of the (00l)tetra. plane increased with increasing Sr content in LSNO.1 7 It is
interesting to note that the strain energies of the LSNO thin films were found to be significantly
lower than the surface energy (Figure 4-3a, details of strain energy calculation are provided in
the ESI), which suggests that the strain energy in LSNO is not strongly correlated with the
structure reorientation of LSNO thin films. Furthermore, the lattice mismatch between LSNO
and GDC, YSZ was found to be independent of the Sr content (Table S4-1), which further
suggests that the structure reorientation is not a result of the film strain. Therefore, the film
reorientation can be attributed to the reduction of the surface energy of LSNO films.
4.3.2 Oxygen Surface Exchange Kinetics of the LSNO Thin Films
EIS data collected from the LSNO thin films as a function of Sr at an oxygen partial
pressure (p(O2)) of 1 atm and T = 550 *C are shown in Figure 4-4a. The real part of the
impedance of the predominant semicircle increased significantly with increasing the Sr content
of LSNO thin films, from which the k' values of LSNO were found to decrease with increasing
the Sr content. In addition, the predominant semicircle was found to increase with decreasing
p(O2) for all compositions, and representative EIS data collected from LSNO with Xsr = 0.2 at
550 *C as a function of p(O2) are shown in Figure 4-4b. Assuming that k" can be approximated
as k*," the k trend of the LSNO thin films is in good agreement with that of bulk LSNO with xsr
= 0.1 and 0.2.40 Representative k values of the LSNO thin films as a function of the Sr content
obtained at p(O2) = 1 atm and T = 550 'C are shown in Figure 4-5. The k" values of the ateta.axis-oriented LNO (Xsr = 0.0) are two orders of magnitude higher than that of the ctea.-axisoriented LSNO (Xsr = 1.0), where the k values of the LSNO thin films decreased with increasing
orientation having the ctetra.-axis perpendicular to the film surface with increasing Sr.
101
I.
.
.
.
.550 bC
.
(b)a 5000
'
(a) C 400
p0 2)=1atm,
0 X=1.0
x=0.8
0300 I
0 4000
0
3000
-0
E
0o 1000
E
I
100
D
200
300
400
Real Impedance / k
10-6
Bulk
10,
X=0
-I
x=0.2
9
10
0 10,10
X=0-4
1011
x=1-0
E
.
Ic
I 1000 2000 3000 4000 5000
Real Impedance / kQ
500
(d)
10
fa
0.1 atm
in
'
00.2
.
C
(c)
0.01 atm
L%
- 100
0
LSNO, x0.2
E 2000
x=0.4
-
550 OC
/
C 200
I
0.001 atm
x=0-1
U
Bulk
x=.
x=0.8
x=0.2
co, 3001
C,)
5500 C
-3
-2
x=0.2
4001
0
1012
10"13
E
Bulk
-1
200
0
X=0.4
550 OC
-3
-2
-1
0
log p(O 2 ) / atm
log p(O 2 ) / atm
-
Figure 4-4. Electrochemical impedance spectroscopy (EIS) results of microelectrodes for La 2
xSrxNiO 4
6
(LSNO) thin films with 0.0 < xsr < 1.0 at 550 0 C (a) Nyquist plot of the LSNO films
in 1 atm. (b) Nyquist plot of the LSNO thin film with XSr = 0.2 as a function of p(O2). (c) Oxygen
partial pressure dependency of the surface exchange coefficients kq of LSNO thin films
calculated from EIS spectra collected at 550 0 C. Extrapolated bulk k* (approximately equivalent
to k )5 8 values obtained from previous data of (m-gray) Kilner et al., 79 (*-dark gray) and (*-light
gray) Boehm et al., 40 are plotted for comparison. (d) Oxygen partial pressure dependency of
volume specific capacitance (VSC) of LSNO thin films calculated from EIS spectra collected at
550 0 C.
102
106
5500 C
p(0 2)= 1 atm
10 7
LSNO (002)
LSNO (200)
Eo
CE,
La 2 -. SrxNiO4
1010
0.0
0.2
6
0.4
0.6
0.8
1.0
Figure 4-5. Sr content dependency of the k of the La2-xSrNiO 4s (LSNO) thin films with 0.0 <
XSr
< 1.0 calculated from EIS spectra collected at 550 0 C in an oxygen partial pressure of 1 atm.
The schematic of atetra.-axis-oriented LNO thin film, both atetra.-axis and ctetra.-axis orientation
coexistence LSNO thin films, and cteta.-axis-oriented LSNO thin film are shown besides the
graph.
4.3.3 Proposed Origin for the Oxygen Surface Exchange Kinetics of LSNO as
a function of the Sr content
The volume specific capacitances (VSCs) extracted from EIS data of the LSNO thin films
(details are provided in the ESI), corresponding to the change in the oxygen nonstoichiometry (6)
induced by the change in the electrical potential, are in good agreement with the 8 trend of the
103
bulk LSNO,59 as shown in Figure 4-4d. However, the VSCs of the LSNO thin films with 0.4 < xsr
< 1.0 were found to be independent on the p(O2), whereas the k values of the LSNO thin films
were found to increase with increasing the p(O2) as expected (Figure 4-4c). Previously, we
reported that the 6 of the LSC thin films might not control the oxygen surface exchange
kinetics.' 7 Moreover, it has been reported that the diffusion coefficients (D*), which can be
influenced by the oxygen content in the LNO thin films have no correlation with the surface
exchange coefficients (k*). 3 1 Therefore, the change in the oxygen content induced by the Sr
content in the LSNO thin films cannot contribute to the modification in the k values observed.
4)
x=0
4..
Sr (c)
L al (b
(a)
x=0.2
x=0.4
x0
-x0.2
-x=0.4
x=0.6
x=0.8
x=1.0
x=0.6
x=0.8
-x1.0
580
-
a
600
-
S
620
-
M
640
Ni
-
x=0
-x=0.2
-x=0.4
x=0.6
x=0.8
2
1620
1665
-x=1.0
810
h
840
870
Kinetic Energy / eV
Figure 4-6. AES data of the La 2-xSrxNiO 4 , (LSNO) thin films with 0.0 < xsr < 1.0 annealed at
550 *C in an oxygen partial pressure of 1 atm. (a) LaMNN cation variation (RSF: 0.059), (b) SrLMM
cation variation (RSF: 0.027), and (c) NiLMM cation variation (RSF: 0.277) as a function of Sr
contents. The change of La and Sr cation spectra scales with the La:Sr concentration and
indicates that the surface chemistry is representative of the surface cationic ratio targeted. The
change in the Ni cation spectra represents the change in the thin film orientation according to
increase in NiO 6 octahedra population (normalized by surface area) with increasing the Sr
content.
104
To investigate the change in the surface chemistry of the LSNO thin films as a function
of the Sr content, AES was conducted on the LSNO thin films after annealing at 550 *C in an
oxygen partial pressure of 1 atm. Figure 4-6 shows the surface LamN, SrLMM, and NiLMM cations
spectra of each film. The intensity of La cation spectra was found to decrease with increasing the
Sr content (Figure 4-6a), whereas the intensity of Sr cation spectra was found to increase (Figure
4-6b), as expected from the nominal composition. This clearly indicates that the surface
chemistry is representative of the surface cationic ratio targeted. It should be noted that the
intensity of Ni cation spectra was found to increase with increasing the Sr content, as shown in
Figure 4-6c. In the case of ctet.-axis-oriented LSNO thin film with xsr = 1.0, the (001)tta. planes
have more NiO octahedra by surface area relative to the (IOO)teft. planes. The relative sensitivity
factor (RSF) of Ni (RSF: 0.277) is sensitively higher than that of La (RSF: 0.059) and Sr (RSF:
0.027). The intensity of Ni cation spectra can therefore be attributed to the thin film reorientation
from the (100)ter. to the (001)tet. orientation, as observed by HRXRD. Furthermore, no spectral
changes were observed for any of the cation Auger lines as a function of the Sr content,
suggesting that no formation of secondary phases occurred across the range of Sr content. This
was further confirmed by scanning electron microscopy (SEM) (Figure S4-6), where the surface
morphology of each film was found to be stable after annealing.
The surface exchange kinetics is highly anisotropic in the RP structure.31 ',5 6 Burriel et
al. 3 ' have shown from secondary ion mass spectroscopy (SIMS) measurements that the oxygen
surface exchange kinetics in the (100)tetr. plane is two orders of magnitude greater than that
along the (001)tet.. plane in LNO thin films. The difference can be attributed to the fact that the
exposed interstitial sites of the (100)teta. surfaces provide active sites for the oxygen surface
exchange and incorporation, whereas the (001)tet. planes have no interstitial oxygen sites due to
the perovskite layers, blocking interstitial site oxygen incorporation, as shown in Figure 4-1. Kim
et al.56 have also reported using electrical conductivity relaxation (ECR) measurement in LNO
thin films that the surface exchange kinetics in the (001)tct. plane is two orders of magnitude
slower than that in the (l00)tt.. plane, which also supports the influence of anisotropy on the
surface exchange kinetics in the RP structure. Therefore, the decrease of the observed kO values
of the LSNO thin films as a function of the Sr content can be partly attributed to the orientation
effect, as shown in Figure 4-5.
105
Surface exchange kinetics of the RP structure can also be influenced by the oxygen
adsorption energy. 60 The weaker adsorption (the less negative adsorption energy) was found to
accompany with higher Sr content for LSNO, regardless of the anisotropy of oxygen adsorption,
as shown in Figure 4-3b. The weakening in the oxygen adsorption upon Sr substitution can be
attributed to the increased oxidation state of the transition metal cations and the reduction of
surface polarity (surface layer charge) induced with the Sr substitution, both of which certainly
contribute to weaken the oxygen binding.61 Consequently, the reduced oxygen adsorption caused
by the Sr substitution may provide a reduced flux of oxygen for the oxygen incorporation
kinetics, and can be another reason for the strong Sr content dependence of the surface exchange
rate for the LSNO thin films, as shown in Figure 4-5.
Although the oxygen surface exchange kinetics are not strongly correlated with the
electrical conductivity in LNO thin films, 31 , 62 the influence of hole concentration and mobility of
Sr and Ni cations on the oxygen surface exchange kinetics cannot be excluded, and their role
requires further investigation.
4.4 Conclusion
In summary, we have successfully deposited epitaxial LSNO thin films with different Sr
content (0.0 5 xsr
1.0) by PLD. HRXRD reveals that the Sr substitution leads to a structural
reorientation from the (100)tetra. (in-plane) to (001)tetra. (out-of-plane) orientation as a function of
the Sr content, which is further supported by the DFT and AES analysis. The k value of the
LSNO thin films is found to strongly depend on the Sr content: the k of the LSNO thin films
decreases with increasing the Sr content. DFT modeling shows that the trend in adsorption
energies of the LSNO system decreases oxygen surface binding with increasing Sr contents. Ex
situ AES shows that there is no formation of second phases or phase segregation as a function of
the Sr content. Our results demonstrate the key role of Sr substitution of RP phases on the
oxygen surface exchange process, where modifying the orientation and the adsorption energy by
Sr substitution is a new strategy to design oxygen surface exchange materials with desired
anisotropic characteristics for applications such as SOFC cathodes or oxygen conducting
membranes.
106
4.5 Supporting Information (SI)
Experimental Details
La2-xSrxNiO4-+ (LSNO, 0 5
Target Synthesis.
5 1.0) and Gdo.2Ceo. 80 2 (GDC)
XSr
were prepared by the Pechini methods. 63 La(N0 3)3 06H2 0, Sr(N0 3)2 , Ni(N0 3)2 *6H20, and
separately Gd(N0 3)3 and Ce(N0 3)3 were dissolved in de-ionized water with ethylene glycol, and
citric acid (Sigma-Aldrich, USA) mixture to synthesize LSNO and GDC respectively. After
esterification at 100 *C, the resin was charred at 400 *C and then calcined at 1,000 *C and 1,200
*C for 12 hours in air for GDC and LSNO, respectably. Pulsed laser deposition (PLD) target
pellets with 25 mm diameter were subsequently fabricated by uniaxial pressing at 50 MPa. The
LSNO and GDC pellets were fully sintered at 1,400 *C in air for 10 hours and 1,100 *C in air for
14 hours, respectively.
The Relaxed
Relaxed lattice parameter determination by in situ HRXRD.
lattice
parameter d and C are derived from the following equation (where i and o6 are the relaxed lattice
c
1819&i(-c)
-
parameters
a/r" =3.223,for the film in an unstrained state),1964
, assuming
3.305, 3.333, 3.339, 3.293, and 3.25 A for XSr = 0,
= 1.0, respectably at 298 K, and v = 0.3.18,
ECC =
-
XSr
= 0.2,
(a-a)
_-2v
=
XSr
19, 42, 48, 64, 65
= 0.4,
XSr
l4
= 0.6,
Xsr =
0.8, and Xsr
The in-plane strain is given by:
and the out of plane strain by: Eaa = (a
In the LSNO thin films, the unit cell volume
Calculation details of strain energy density.
can be obtained by c -a2 . The volumetric strain are calculated using the following equation:
Volumetric strain
=
onstrained-Vrelaxed)
Vrelaxed
In a three dimensional linear elastic solid with loads supplied by external forces, the strain
66 67
energy density over entire volume can be expressed by the equation below. '
2
U (strain energy density, J m-3 )
I2
K
Vconstrained-Vrelaxed)
Vrelaxed
107
where
K (bulk modulus, GPa) =
E
3(1-2V)
Huang et al. 68 have reported that E (young's modulus) of LNO is ~ 155 GPa at room temperature
and v = 0.3 18,
19, 42, 48, 64, 65
was used in this study. Then, we can determine the strain energy
density for LSNO with 0.0 5 xs, < 1.0.
Microelectrodes Fabrication.
In situ electrochemical impedance spectroscopy (EIS)
measurements were conducted to probe ORR activity on geometrically well-defined LNO
microelectrodes fabricated by photolithography and acid etching, where sintered porous Pt
sintered onto the backside of the YSZ substrate served as the counter electrode. OCG positive
photoresist (Arch Chemical Co., USA) was applied on the LNO surface and patterned using a
mask aligner (Karl Suss, Germany, A = 365 nm). The photoresist was developed using Developer
934 1:1 (Arch Chemical Co., USA) and the thin films were etched in hydrochloric acid (HCl) to
remove LNO film excess and create the circular microelectrodes (diameters -50 p1m, -100 pm,
~150 pm, and
-200 pm, exact diameter determined by optical microscopy). Before
electrochemical testing, microelectrode geometry and morphology was examined by optical
microscopy (Carl Zeiss, Germany) and atomic force microscopy (AFM) (Veeco, USA). AFM
measurements after acid-etching of the LSNO film revealed thickness of ~ 42, ~ 64,
-
69, -73
nm, ~ 79 nm, and ~ 87 nm for Sr = 0, Sr = 0.2, Sr = 0.4, Sr = 0.6, Sr = 0.8, and Sr = 1.0,
respectively, at 5,000 pulses.
Electrochemical Characterization.
Fig.
S7.
details
the
equivalent
circuit
and
corresponding Nyquist plot for this experimental system. ZView software (Scribner Associates,
USA) was used to construct the equivalent circuit and perform complex least squares fitting. The
EIS data were fitted using a standard resistor (R1) for HF and resistors (R2) in parallel with a
constant phase elements (CPE2 ) for MF and LF (R1-(R2/CPE2)-(RoRR/CPEoR)). Based on the
p(O2) dependence of the three features, physical or chemical process with regard to each
frequency range can be determined." 69-71 The HF feature (104
-
10 5
Hz) was found unchanged
with p(O2), and its magnitude and activation energy (-1.15 eV) were comparable to those of
oxygen ion conduction in YSZ reported previously. 72 The MF feature (10' - 104 Hz), which was
found to have a p(O2) independent feature, was attributed to interfacial transport of oxygen ions
between the LNO film and the GDC layer. In addition, the magnitude of its capacitance was
108
relatively small (~10-6 F) compared to the LF feature (-10-3 F). The LF feature (10-2
-
103 Hz)
was found to have a strong p(O2) dependence. The resistance of the LF feature drastically
increases as oxygen partial pressure decreases. In the case of thin film samples, the magnitude of
capacitance is due to the oxygen content change in the films. Therefore, the electrode oxygen
surface reaction corresponds with the LF feature. We obtained values for RoRR; and knowing the
area of the microelectrode (Aelectrode = 0.25 7E delectrode 2 ) we can determine the ORR area specific
resistance (ASRoRR = RoRR . Aelectrode). The electrical surface exchange coefficient (ku), which is
comparable to k*,5 was determined using the expression ,,3
0 = R T / 4F2RORRAelectrodeco
(1)
where R is the universal gas constant (8.314 J mol-1 K71), T is the absolute temperature, F is the
Faraday's constant (96,500 C mol-1), and co is the lattice oxygen concentration in LSNO where
Co
= (4+6)/Vm,
(2)
Vm is the molar volume of LSNO at room temperature. In this study, co was calculated with c
74 75
extracted from previous reported values. '
VSC, indicative of changes in the oxygen nonstoichiometry induced by changes in the
electrical potential, can be obtained from EIS data via the expression 76
VSC = [1/(Aeectrode E thickness)]((RoRR)1-nQ)n,
where
Q
(3)
is the non-ideal "capacitance", and n is the non-ideality factor of CPE. The fitted
values of n for semi-circle CPEORR were found to range from
-
0.96 to 1.0 over the entire PO2
range examined (n =1, ideal).
Experimental details of ex situ AES.
In AES, the obtained energy spectrum for a
particular element is always situated on a large background (low signal-to-noise ratio), which
arises from the vast number of so-called secondary electrons generated by a multitude of
inelastic scattering processes. To obtain better sensitivity for detection of the elemental peak
positions, the AES spectra from this study are presented in the differentiated form. Elemental
quantification of AES spectra utilized relative sensitivity factors (RSFs) of 0.059, 0.027, 0.227,
and 0.212 for LamNN, SrLmm, NiLmm, and OKLL, respectively, as supplied by the AES
manufacturer (Physical Electronics). In addition, the Inelastic-Mean-Free-Path (IMFP) was
calculated to correct signal intensity for their different IMFPs (information depth). IMFPs were
calculated using the NIST Standard Reference Database 71 "NIST Electron Inelastic- Mean-
109
Free-Path Database" version 1.2. The software program provides the ability to predict the IMFP
for inorganic compounds supplying the stoichiometric composition of La2 -xSrNiO4,& (0
< x :
1.0), the number of valence electrons per molecule (assumed to be 40) and a band gap energy
(for which we are assuming 0 eV as LSNO is metallic like at high temperatures; additionally
when assuming a band gap of an insulator 5 eV, the IMFP increases by -0.03 nm). The IMFP for
La, Sr, and Ni were determined to be 1.361 ~ 1.395 nm, 2.611 ~ 2.667 nm, and 1.562
-
1.607
nm, respectively. A relative depth-scaling factor (s) was determined as:
or
=
(4)
f+exp (-- )dx,
0 Ai
Ai
where A, is the IMFP, yielding aNi = 0.58, asr = 0.41, and ULa = 0.63. The intensities from
different elements were scaled using Iscaied =
Imeasured*Si/Si.
Details of density functional theory (DFT) calculations.
Spin-polarized
Density
Functional Theory (DFT) calculations were preformed with the Vienna Ab-initio Simulation
Package5 0 ' 5 using the Projector-Augmented plane-Wave method5 2 with a cutoff energy of 450
eV. Exchange-correlation
was treated in the Perdew-Wang-9 177 Generalized
Gradient
Approximation (GGA) using the soft 0_s oxygen pseudopotential. The GGA+U calculations 54
are performed with the simplified spherically averaged approach55 , where the Ueff (Uff
=
Coulomb U - exchange J) is applied to d electrons (Uef(Ni) = 6.4 eV). All calculations are
performed in the ferromagnetic state in order to use a consistent and tractable set of magnetic
structures.
Fully relaxed bulk LSNO (Sr content : x=0, 0.5, and 1) calculations are performed using the
2 atetrax 2 aletra x
ctetra supercells with 3 x3 x2 kpoints. The Sr in the LSNO bulk is arranged to have
the farthest Sr-Sr pair distance in the simulated supercells. Based on the LSNO bulk
configurations, the LSNO (OOl)tetra. and (l00)tetra. surface energy and surface oxygen adsorption
energy are calculated using 2 atetrax2 atetra 9-layer (001) and 2 atetrax ctetra 6-layer (100)tlra. slabs in
periodic boundary conditions with 10
A
vacuum between the two surfaces, as illustrated in
Figure S4-9 (the kpoints setups are: 3x3xl kpoints for the 9-layer (001)tetra. slab and 1x3x2
kpoints for the 6-layer (100)tetra. slab).
110
The surface energies (Esr5f) are then obtained using the equation below:
Esurf =
2 (Esab
(5)
-N*EbuIk)Asurf
where Eslab is the calculated total energy of LSNO slabs,
Ebulk
is the calculated total energy of
LSNO bulk normalized as per formula unit, N is the number of LSNO units in the slab, and Asurf
is the surface area of the simulated slabs.
Oxygen adsorption energies (Ead) are calculated based on the following equation:
Eads
=
EO-adsorbed-slab
-Eslab -
(6)
2*EO 2
where EO-adsorbed-slab is the calculated total energy of LSNO slabs with a surface adsorbed oxygen
at the adsorption site illustrated in Figure S4-8 (c and d), and E0 2 is the calculated total energy of
isolated 02 molecule corrected with +0.33 eV/O, which is obtained by fitting to a series of binary
oxide experimental formation enthalpies at room temperature."
Finally, we distinguished that the (100)tetra. orientation in this work is equivalent to the defined
(1 O)tra. orientation in the previous theoretical study done by Read et al.5
Our calculated
surface energy based on the (IOO)tetra. surface configuration suggested by Read et al.5 7 (in Figure
1(a) and Figure 3 of Ref. 30) at xsr=0, 0.5, and 1 is found to be less stable (surface energy range
between 0.09-0.10 eV/A 2) than the two main surfaces investigated in this work (between
0.06-0.08 eV/A2 , see Fig. 3a of the main content), due to that the alternating 02 (layer charge of
-4) and A 2 B0 2 (layer charge of +4) layers give rise to polar instability.
relative surface stability of (11 0)tetra.
VS.
Nonetheless, the
(001 )tetra. and Sr solution energy reported by Read et al.",
which is equivalent to the (O00)teta. and (OO1)ttra. surfaces in this work, is consistent with the
calculated surface energetics in our DFT simulations.
111
Table S4-1. Lattice mismatch between the film materials and the substrate materials.
Lattice
ab
Al
c
/A
Substrate / A
(bulk)
(a-b plane)
41
,
La2 Sr NiO 4
x=0
,
La2 Sr NiO 4
x=0.24
5
3.854
3.881
3.842
-6.76%
NdGaO 3 (110), aNGO= 3.86331
~0.47
~7.95
YSZ (001), a sz= 5.1470
~6.9
~11.9%
GDC (001), aG= 5.41881
~1.3%
-8.8%
YSZ (001), aysz= 5.147
-6.7%
-14.6%
GDC (001), aG=
5.418
~0.3
-10.5%
YSZ (001), asz = 5.147
-5.6%
~16.3
-- 0.3%
-10.8%
-4.9%
-16.6%
-- 0.6%
-10.6%
YSZ (001), asz = 5.147
-4.6%
-16.4%
GDC (001), aGDC= 5.418
-- 0.4%
-9.3%
YSZ (001), asz = 5.147
-4.8%
~15
GDC (001), aGDC= 5.418
-- 0.2%
-8.1%
12.214
12.51
12.70
GDC (001), a
,
La2 Sr NiO 4
x=0.44
5
3.819
12.72
GDC (001), aG=
x=0.64
= 5. 4 18
YSZ (001), asz = 5.147
La2 Sr NiO4+,
5
5.418
12.722
,
La2 Sr NiO 4
3.807
%
12.51
%
Nd2NiO 4
3.881
3.906"
-- 0.6%
=
%
x0 45
SrTiO 3 (100), aSTO
(c plane)
%
NiO 4
,
La2 Sr
Lattice
mismatch
mismatch
pulk(bulk
%
Materials
La2_xSr NiO 4
x=1.04
3.824
12.559
%
3.814
12.429
YSZ (001), asz = 5.147
112
-5.1%
~13.8
%
5
,
x=0.84
<
12.7 (a
12.6
MA
12.5
o
La 2-.Sr.NiO4
8
2
-1 o
00
S12.4
2:;
0
(b)
d4 6
0
12.3
M
a ilattice
V
0 0 clattice
A bulk a, clattice
1
-2
0
*
w 3.9
m 3.7
M
-
0
--.
3
12.8
0.0
0.2
0.4
0.6
0.8
0
1.0
XSr
0.0
0.2
0.4
0.6
0.8
1.0
-3
XSr
Figure S4-1. (a) Constrained (u-red, *-blue) and Relaxed (o-red, o-blue) lattice parameters of
the La2..xSrxNiO 4 8 (LSNO) thin films as a function of Sr content at room temperature.
Extrapolated bulk atetra. (V-gray) and ctetra. (A -gray) lattice parameters at room temperature
obtained from previous data of Gopalakrishanan et al.45 are plotted for comparison. The
constrained normal and in-plane lattice parameters of the LSNO films were calculated from
combining the interplanar distances of the (200)tetra., (10 3 )tetra. and (006)tetra. peaks. (b) Out of
plane and in-plain strain as a function of the Sr content calculated using Exx =
Ezz=
-
,
C
and
for in-plane strain and out of plane strain respectably. For determining the relaxed
film lattice parameter a and C, we used the equation:
(c-e^)
__-2v
2
eC = 1-v
(a-g)
a
,assuming f/ile
5=
3.223, 3.305, 3.333, 3.339, 3.293, and 3.25 A for xsr
0.2, xsr = 0.4, xsr = 0.6, xsr = 0.8, and xs, = 1.0, respectably at 298 K, and v = 0.3.
113
0 ~xsr
0,
(a)
(b)
La2-xSrxNiO4 8
La 2-xSrNi0 4
NiO, octahedra
Ce14
. GdO 2.-
10,
6
octahedra
Cel-,Gd02.-o
b
b
45*
bL
45'
<
a
Figure S4-2. Schematic of two different orientations of La2.. SrxNiO
4 6
(LSNO) on GDC (a)
(100)tetra.-oriented epitaxial LSNO thin film and (b) (00 1)teta.-oriented epitaxial LSNO thin films.
114
(a)
La2-xSr.NiO
La 2.xSrNiO4 6
C 200
2.20
,(b)
,
,300,-,
250 C
---
2.15a
2.10
200 '
%Wtt150 -
1
-150
100
(200)
@0
50
50
50--%
0I
0.0
0.2
0.4
0.6
XSr
0.8
J
--
2.05
-
U
0.2.00
100
0
(006)
0<
0-
)
1.95
1.90
O~0 (
1.0
p
0.0
0.2
0.6
0.4
0.8
1.0
XSr
Figure S4-3. (a) Peak intensities and (b) d spacing of the La 2-,SrxNiO
4
6
(LSNO) thin film
(00 6 )tetra. and (2 0 0 )tetra. as a function of Sr content obtained from HRXRD. The peak intensities of
( 2 0 0 )tetra. significantly decreases with increasing the Sr content while those of (00 6 )tetra.
significantly increases, which suggests that once (00l)tetra. orientation growth begins, (l00)tetra.
orientation growth is suppressed.
115
550 0 C
LNO, x=0
0.001 atm
C
2000
C:
550 0C
550 0 C
LSNO, x=0.4
8000
-
C 3000
0.001 atm
C
040000U.001
atm
LSNO, x=1.0
*~6000
0.01 atm
E
'
A 0.01 atm
E
4000
.
2000
20000
1000
C
0
0.1 atm
1 atm
E
0
1000
(a)
2000
3000
Real Impedance / kfl
0.01 atm
0.
0.1 atm
7- 0.1 atm
1 atm
E
0
(b)
l at1
E
0
5000
10000
Real Impedance / ko
(c)
50000
Real Impedance I kQ
Figure S4-4. Electrochemical impedance spectroscopy (EIS) results of microelectrodes for the
La 2-xSrxNiO 4 6 (LSNO) thin films with 0 <
XSr
5 1.0 at 550 0 C (a) Nyquist plot of the LNO thin
film with x=0, (b) Nyquist plot of the LSNO thin film with
LSNO thin film with
XSr
XSr
= 0.4, and (c) Nyquist plot of the
= 1.0 as a function of p(O2). All films exhibited nearly perfect
predominant semicircle impedances, which indicates that the surface oxygen exchange kinetics
governs the oxygen electrocatalysis on the thin film surface.'
116
Figure S4-5. AFM measurements of the as-prepared La2..xSrNiO 4 8 (LSNO) thin films with 0 <
XSr
< 1.0 deposited at 5,000 pulses (a)
0.655 nm, (c)
XSr =
xsr
= 0 with RMS of 0.341 rm, (b) XSr = 0.2 with RMS of
0.4 with RMS of of 0.672 nm, and (d)
XSr = 0.8 with RMS of 0.611 nm, and (f)
XSr
XSr =
0.6 with RMS of of 0.666 nm, (e)
= 1.0 with RMS of 0.374 nm.
117
Figure S4-6. Scanning electron microscopy (SEM) images of (a) La 2NiO 4+6 (LNO) films, (b)
La2-xSrxNiO 4 6 (LSNO) films with xsr
Xsr
=
0.2, (c) LSNO films with xsr
= 0.6, (e) LSNO films with xsr = 0.8, and (f) LSNO films with
an oxygen partial pressure of 1 atm for 6 hours.
118
xsr
=
0.4, (d) LSNO films with
= 1.0 annealed at 550 C in
(b)
(a)
02
R1
R2
(HF)
CPE2
Substrate
(MF) Interface
RORR
CP ORR
(LF) Film
V
(c).
-Z
ORR
02
Figure S4-7. (a) Schematic of a LSNO/GDC/YSZ(001)/porous Pt samples and electrochemical
testing configuration (not drawn to scale), and (b) equivalent circuit (R 1 = YSZ electrolyte
resistance, R2 = electrode/electrolyte interface resistance19, ROR = ORR resistance, CPE =
constant phase element) used to extract ORR kinetics, and (c) characteristic Nyquist plot
schematic (color key : orange = YSZ/bulk transport, green = GDC/interface, blue
LSNO/ORR).
119
(a)
g
(b)
0
0
0
0!
*
0
0
0 00
00
gor
0
(10 0 )tetra. slab - side view
(001)tetra. slab - side view
(c)
(d)
0
Top 3 layers of the (0 0 1 )tetra. slab - top view
Figure S4-8. Simulated La 2-xSrxNiO 4
6
0-
-0
0
0
Top 3 layers of the (1 0 0 )tetra. slab - top view
(LSNO) slab models in the density functional theory
calculations in this work: (a) side view of the (001)tetra slab, (b) side view of the (10O)tetra. slab,
(c) top view of the top 3 layers of the (001)tetra. slab, and (d) top view of the top 3 layers of the
(100)tetra. slab. The dotted circles in c and d represent the surface oxygen adsorption sites for the
(001 )tetra. and (10 )tetra. surfaces, respectively.
120
4.6 References
1.
S. B. Adler, Chem. Rev., 2004, 104, 4791-4843.
2.
A. S. Bhalla, R. Y. Guo and R. Roy, Mater. Res. Innov., 2000, 4, 3-26.
3.
R. J. H. Voorhoeve, D. W. Johnson, J. P. Remeika and P. K. Gallagher, Science, 1977,
195, 827-833.
4.
S. B. Adler, X. Y. Chen and J. R. Wilson, J. Catal., 2007,245, 91-109.
5.
L. da Conceicao, C. R. B. Silva, N. F. P. Ribeiro and M. Souza, Mater. Charact., 2009,
60, 1417-1423.
6.
R. A. De Souza and J. A. Kilner, Solid State Ion., 1998, 106, 175-187.
7.
R. A. De Souza and J. A. Kilner, Solid State Ion., 1999, 126, 153-161.
8.
P. Decorse, G. Caboche and L. C. Dufour, Solid State Ion., 1999, 117, 161-169.
9.
T. Ioroi, T. Hara, Y. Uchimoto, Z. Ogumi and Z. Takehara, J Electrochem. Soc., 1997,
144, 1362-1370.
10.
A. J. Jacobson, Chem. Mat., 2010, 22, 660-674.
11.
Y. L. Lee, J. Kleis, J. Rossmeisl and D. Morgan, Phys. Rev. B, 2009, 80.
12.
Z. P. Shao and S. M. Haile, Nature, 2004, 431, 170-173.
13.
B. C. H. Steele and A. Heinzel, Nature, 2001, 414, 345-352.
14.
C. W. Sun, R. Hui and J. Roller, J. Solid State Electrochem., 2010, 14, 1125-1144.
15.
S. M. Hashim, A. R. Mohamed and S. Bhatia, Adv. ColloidInterface Sci., 2010, 160, 88100.
16.
S. B. Adler, Solid State Ion., 1998, 111, 125-134.
17.
E. J. Crumlin, S. J. Ahn, D. Lee, E. Mutoro, M. D. Biegalski, H. M. Christen and Y.
Shao-Horn, J Electrochem. Soc., 2012, 159, F219-F225.
18.
E. J. Crumlin, E. Mutoro, Z. Liu, M. E. Grass, M. D. Biegalski, Y. L. Lee, D. Morgan, H.
M. Christen, H. Bluhm and Y. Shao-Horn, Energy Environ. Sci., 2012, 5, 6081-6088.
19.
G. J. la 0, S. J. Ahn, E. Crumlin, Y. Orikasa, M. D. Biegalski, H. M. Christen and Y.
Shao-Horn, Angew. Chem.-Int. Edit., 2010, 49, 5344-5347.
20.
J. Mizusaki, Y. Mima, S. Yamauchi, K. Fueki and H. Tagawa, J Solid State Chem.,
1989, 80, 102-111.
121
21.
V. G. Sathe, A. V. Pimpale, V. Siruguri and S. K. Paranjpe, Journal of PhysicsCondensedMatter, 1996, 8, 3889-3896.
22.
S. Y. Wang, J. Yoon, G. Kim, D. X. Huang, H. Y. Wang and A. J. Jacobson, Chem. Mat.,
2010, 22, 776-782.
23.
F. S. Baumann, J. Fleig, H. U. Habermeier and J. Maier, Solid State Ion., 2006, 177,
1071-1081.
24.
A. Esquirol, N. P. Brandon, J. A. Kilner and M. Mogensen, J. Electrochem. Soc., 2004,
151, A1847-A1855.
25.
L. W. Tai, M. M. Nasrallah, H. U. Anderson, D. M. Sparlin and S. R. Sehlin, Solid State
Ion., 1995, 76, 259-271.
26.
L. W. Tai, M. M. Nasrallah, H. U. Anderson, D. M. Sparlin and S. R. Sehlin, Solid State
Ion., 1995, 76,273-283.
27.
A. V. Berenov, A. Atkinson, J. A. Kilner, E. Bucher and W. Sitte, Solid State Ion., 2010,
181, 819-826.
28.
Y. M. Kim, S. I. Pyun, J. S. Kim and G. J. Lee, J Electrochem. Soc., 2007, 154, B802B809.
29.
G. Amow, I. J. Davidson and S. J. Skinner, Solid State Ion., 2006, 177, 1205-12 10.
30.
G. Amow and S. J. Skinner, J. Solid State Electrochem., 2006, 10, 538-546.
31.
M. Burriel, G. Garcia, J. Santiso, J. A. Kilner, J. C. C. Richard and S. J. Skinner, J
Mater. Chem., 2008, 18, 416-422.
32.
R. Sayers and S. J. Skinner, J. Mater. Chem., 2010, 21, 414-419.
33.
V. V. Kharton, A. V. Kovalevsky, M. Avdeev, E. V. Tsipis, M. V. Patrakeev, A. A.
Yaremchenko, E. N. Naumovich and J. R. Frade, Chem. Mat., 2007, 19, 2027-2033.
34.
T. Nakamura, K. Yashiro, K. Sato and J. Mizusaki, Solid State Ion., 2010, 181, 292-299.
35.
J. M. Bassat, P. Odier, A. Villesuzanne, C. Marin and M. Pouchard, Solid State Ion.,
2004,167,341-347.
36.
S. J. Skinner and J. A. Kilner, Ionics, 1999, 5, 171-174.
37.
S. J. Skinner and J. A. Kilner, Solid State Ion., 2000, 135, 709-712.
38.
S. Shinomori, M. Kawasaki and Y. Tokura, Appl. Phys. Lett., 2002, 80, 574-576.
39.
A. Aguadero, M. J. Escudero, M. Perez, J. A. Alonso and L. Daza, J. Fuel Cell Sci.
Technol., 2007, 4, 294-298.
122
40.
E. Boehm, J. M. Bassat, P. Dordor, F. Mauvy, J. C. Grenier and P. Stevens, Solid State
Ion., 2005, 176, 2717-2725.
41.
A. Yamada, Y. Suzuki, K. Saka, M. Uehara, D. Mori, R. Kanno, T. Kiguchi, F. Mauvy
and J. C. Grenier, Adv. Mater., 2008, 20, 4124-+.
42.
D. Lee, A. Grimaud, E. J. Crumlin, K. Mezghani, M. A. Habib, Z. Feng, W. T. Hong, M.
D. Biegalski, H. M. Christen and Y. Shao-Horn, The Journal of Physical Chemistry C,
feature article, 2013, 117, 18789-18795.
43.
A. Demourgues, F. Weill, B. Darriet, A. Wattiaux, J. C. Grenier, P. Gravereau and M.
Pouchard, J Solid State Chem., 1993, 106, 317-329.
44.
J. Rodriguezcarvajal, M. T. Fernandezdiaz and J. L. Martinez, Journal of PhysicsCondensed Matter, 1991,3, 3215-3234.
45.
J. Gopalakrishnan, G. Colsmann and B. Reuter, J. Solid State Chem., 1977, 22, 145-149.
46.
Y. Takeda, R. Kanno, M. Sakano, 0. Yamamoto, M. Takano, Y. Bando, H. Akinaga, K.
Takita and J. B. Goodenough, Mater. Res. Bull., 1990, 25, 293-3 06.
47.
H. J. M. Bouwmeester, H. Kruidhof and A. J. Burggraaf, Solid State Ion., 1994, 72, 185194.
48.
E. J. Crumlin, E. Mutoro, S. J. Ahn, G. J. la 0, D. N. Leonard, A. Borisevich, M. D.
Biegalski, H. M. Christen and Y. Shao-Horn, J. Phys. Chem. Lett., 2010, 1, 3149-3155.
49.
J. Maier, Physical Chemistry of Ionic Materials: Ions and Electrons in Solids John Wiley,
Chichester, England ; Hoboken, NJ, 2004.
50.
G. Kresse and J. Furthmuller, Phys. Rev. B, 1996, 54, 11169-11186.
51.
G. Kresse and J. Hafner, Phys. Rev. B, 1993, 47, 558-561.
52.
P. E. Blochl, Phys. Rev. B, 1994, 50, 17953-17979.
53.
J. P. Perdew and Y. Wang, Phys. Rev. B, 1992, 45, 13244-13249.
54.
V. I. Anisimov, F. Aryasetiawan and A. I. Lichtenstein, Journalof Physics-Condensed
Matter, 1997, 9, 767-808.
55.
S. L. Dudarev, G. A. Botton, S. Y. Savrasov, C. J. Humphreys and A. P. Sutton, Phys.
Rev. B, 1998, 57, 1505-1509.
56.
G. Kim, S. Wang, A. J. Jacobson and C. L. Chen, Solid State Ion., 2006, 177, 1461-1467.
57.
M. S. D. Read, M. S. Islam, G. W. Watson and F. E. Hancock, J. Mater. Chem., 2001, 11,
2597-2602.
123
58.
J. Maier, Solid State Ionics, 1998, 112, 197-228.
59.
V. V. Vashook, Yushkevich, II, L. V. Kokhanovsky, L. V. Makhnach, S. P. Tolochko, I.
F. Kononyuk, H. Ullmann and H. Altenburg, Solid State Ion., 1999, 119, 23-30.
60.
J. W. Han and B. Yildiz, Energy Environ. Sci., 2012, 5, 8598-8607.
61.
F. Calle-Vallejo, N. G. Inoglu, H. Y. Su, J. I. Martinez, I. C. Man, M. T. M. Koper, J. R.
Kitchin and J. Rossmeisl, Chem. Sci., 2013, 4, 1245-1249.
62.
M. Burriel, J. Santiso, M. D. Rossell, G. Van Tendeloo, A. Figueras and G. Garcia, J.
Phys. Chem. C, 2008, 112, 10982-10987.
63.
M. Pechini, US. PatentNo. 3,330,697, 1967.
64.
H. M. Christen, E. D. Specht, S. S. Silliman and K. S. Harshavardhan, Phys. Rev. B,
2003, 68, 4.
65.
D. Y. Noh, Y. Hwu, J. H. Je, M. Hong and J. P. Mannaerts, Appl. Phys. Lett., 1996, 68,
1528-1530.
66.
U. C. Jindal, Strength ofMaterials,Pearson Education India, 2012.
67.
A. F. Bower, Applied mechanics ofsolids, CRC Press, 6000 Broken Sound Parkway NW,
suite 300, Boca Raton, FL 33487, 2009.
68.
B. X. Huang, J. Malzbender and R. W. Steinbrech, J. Mater. Sci., 2011, 46, 4937-4941.
69.
S. B. Adler, J. A. Lane and B. C. H. Steele, J. Electrochem. Soc., 1996, 143, 3554-3564.
70.
T. Kawada, J. Suzuki, M. Sase, A. Kaimai, K. Yashiro, Y. Nigara, J. Mizusaki, K.
Kawamura and H. Yugami, J. Electrochem. Soc., 2002, 149, E252-E259.
71.
Y. L. Yang, C. L. Chen, S. Y. Chen, C. W. Chu and A. J. Jacobson, J. Electrochem. Soc.,
2000, 147, 4001-4007.
72.
P. S. Manning, J. D. Sirman, R. A. DeSouza and J. A. Kilner, Solid State Ion., 1997, 100,
1-10.
73.
J. Fleig and J. Maier, J. Eur. Ceram. Soc., 2004, 24, 1343-1347.
74.
T. Nakamura, K. Yashiro, K. Sato and J. Mizusaki, Phys. Chem. Chem. Phys., 2009, 11,
3055-3062.
75.
V. V. Vashook, H. Ullmann, 0. P. Olshevskaya, V. P. Kulik, V. E. Lukashevich and L.
V. Kokhanovskij, Solid State Ion., 2000, 138, 99-104.
76.
J. Fleig, Solid State Ion., 2002, 150, 181-193.
77.
J. P. Perdew and Y. Wang, PhysicalReview B, 1992, 45, 13244.
124
78.
J. Goniakowski, F. Finocchi and C. Noguera, Rep. Prog. Phys., 2008, 71.
79.
J. A. Kilner and C. K. M. Shaw, Solid State Ion., 2002, 154, 523-527.
80.
M. Yashima, S. Sasaki, M. Kakihana, Y. Yamaguchi, H. Arashi and M. Yoshimura, Acta
Crystallogr. Sect. B-Struct. Commun., 1994, 50, 663-672.
81.
B. Matovic, S. Boskovic, L. Zivkovic, M. Vlajic and V. Krstic, in Current Research in
Advanced Materials and Processes, eds. D. P. Uskokovic, S. K. Milonjic and D. I.
Rakovic, Trans Tech Publications Ltd, Zurich-Uetikon, 2005, pp. 175-179.
125
This page is intentionally left blank
126
Chapter 5.
Enhanced Surface Stability and Electrocatalytic
Activity on Epitaxial LaOSSr. 2 CoO 3 .f Thin Films
by Lao.8 Sro.2 MnO 3 6. Decoration
10?
550 OC
LSM 0.1-1 nm
101
E
1041-2
-1
log P(02)
0
O tM
Reproduced in part with permission from Dongkyu Lee, Yueh-Lin Lee, Alexis Grimaud, Wesley
T. Hong, Michael D. Biegalski, Dane Morgan and Yang Shao-Horn, Enhanced Surface Stability
and Electrocatalytic Activity on Epitaxial Lao.sSro. 2 CoO 3 - Thin Films by Lao.8 Sr. 2MnO..
Decoration, The Journal of Physical Chemistry C, 2014, In Revision
127
5.1 Introduction
Lowering the operating temperature of solid oxide fuel cells (SOFCs) to the intermediate
temperature range (500 - 750 *C) in the development of advanced SOFCs is hampered by the
slow kinetics of the oxygen surface exchange (02 + 4e-
++
202-) at the cathode.'-3 Cathode
materials based on Lai.xSrxMnO3-8 (LSM) 4~8 with high electronic conductivity but low ionic
conductivity, 9 which are currently utilized in SOFCs operated; at high temperatures such as
1000 'C, have oxygen surface exchange kinetics too slow to achieve efficient SOFCs at
intermediate temperatures. Mixed ionic and electronic conductors (MIECs) such as LaixSrxCoO3..8 (LSC) 1-1 6 and LaixSrxCoiFyO- (LSCF)17
21
perovskite oxides have been intensely
studied to promote the oxygen surface exchange kinetics at intermediate temperatures. Recent
efforts have been focused on developing advanced cathode materials based on MIECs with
surface modification 22 -3 6 to achieve enhanced surface exchange kinetics and stability. For
example, Yashiro et al. have reported -1
order of magnitude enhancement in activity for the
composite cathode screen-printed with the mixture of Lao.6 Sro.4Co03- and (Lao. 5Sro.5 ) 2 Co0418.3 5
In addition, LSM has been widely used for surface modification of MIECs.2 5, 2 7 , 2 9 , 3 6 Depositing
thin Lao.8 5 Sro.15Mn03-8 coatings on porous Lao.6 Sro.4 Coo. 2 Feo.s0 3.- (LSCF6428) electrodes using
an infiltration process, Lynch et al.29 have shown the enhanced surface electrocatalytic activity of
decorated LSCF6428 cathodes upon polarization, resulting from a faster surface exchange
kinetics relative to uncoated LSCF6428. More recently, Zhu et al.36 have also shown that the
cathodic polarization resistance of LSCF6428 cathode can be reduced by surface modification
with a thin and continuous Lao 8 Sro.2Mn0 3. (LSM82) layer via a sol-gel process. Although
several studies have shown enhanced cathodic performance by LSM decorations, the influence
of LSM on the surface chemistry of perovskites that governs oxygen eletrocatalysis at elevated
temperatures is poorly understood.
Well-defined single-crystalline perovskite thin films have been used as a model system to
develop fundamental understanding into key parameters that govern oxygen surface exchange
kinetics.
23 24
- , 31,
34, 37-44
In particular, the surface oxygen exchange kinetics on epitaxial (001)-
oriented Lao.Sro 2Co03-8 (LSC82) thin film surfaces decorated with strontium (hydr)oxide3
or
(Lao. 5Sro.5)2 Co0418 (LSC 2 14 ) 2 3 - 2 4 particles have shown increased activities up to one and three
orders of magnitude, respectively. Our recent study using Coherent Brag Rod Analysis (COBRA)
has revealed the atomic structure and concentrations of the (001)-oriented LSC82 thin film on a
128
SrTiO 3 (STO) substrate, which shows strontium (Sr) segregation toward the LSC82 surface and
Sr depletion near the interface between LSC82 and STO. 4 ' More recently, COBRA has also
revealed the markedly enhanced Sr concentration at the interface of LSC82 and LSC214 and near
the surface of LSC 214 , proposing that the increased Sr content at the interface may contribute the
enhanced catalytic activity resulting in higher oxygen vacancy concentration.4 0 In addition, we
have shown that heating the (001)-oriented LSC82 surface leads to the formation of surface Srenriched particles upon annealing while the LSC 214 -decorated LSC82 surface chemistry is stable
upon heating. 39 These observations have suggested that the surface decoration can modulate the
surface Sr segregation and the surface phase stability, which can greatly influence the oxygen
surface exchange kinetics and the surface stability in LSC4 1 and LSCF.2 ,
4
Therefore,
understanding the surface decoration effect on the surface chemistry of perovskites is critical and
other potential surface modification materials need to be investigated to design highly active and
stable cathodes for SOFCs. In this study, we are particularly interested in addressing whether
LSM82 decoration on the LSC82 surface can lead to any enhancement of the surface exchange
kinetics as shown on porous LSCF cathodes and how LSM82 decoration can influence the
LSC82 surface chemistry using epitaxial thin films as a model system.
We here employ epitaxial LSC82 thin films with LSM82 surface decoration of varying
coverage to investigate how LSM82 decorations can alter the oxygen surface exchange kinetics.
Using in situ high-resolution X-ray diffraction (HRXRD), we show that the LSM82 decoration
has no influence on the in-plane and out-of-plane strains of the LSC82 films at elevated
temperatures. We find that partial LSM82 decorations (-0.1 - 0.9 nm) can enhance the surface
exchange coefficient (kg) up to two orders of magnitude relative to the pristine LSC82 thin film
while full LSM82 coverage (-3.5 - 10 nm) reduces the k values. Auger electron spectroscopy
(AES) reveals substantial changes in the surface cationic ratios associated with LSM82
decoration after annealing. DFT calculation shows A-site cation interdiffusion across the
interface between LSM82 and LSC82. The enhanced oxygen surface exchange kinetics of the
LSM82-decorated LSC82 thin films can thus be attributed to the stabilization of LSM82decorated LSC82 surface, where LSC82 surface cationic ratios are changed and the formation of
Sr-enriched particles on the surface is suppressed by LSM82 decoration.
129
5.2 Experimental Methods
Pulsed laser deposition (PLD) was utilized to deposit the epitaxial LSC82 thin films (-85
nm) on yttria-stabilized zirconia (YSZ(001)) with a
-
5 nm gadolinium-doped ceria (GDC)
buffer layer to prevent the formation of La2Zr 20 7.47 Varying thicknesses of LSM82, from partial
(-0.1, -0.3, -0.9 nm) to full coverage (-3.5, and -10 nm) were subsequently deposited on top of
the LSC82/GDC/YSZ. LSM82 decoration layer thickness was extrapolated from atomic force
microscopy (AFM) of the 500 pulses (-3.5 nm) and 1,500 pulses (-10 ntm) LSM82 coverage on
the LSC82 films after microelectrode patterning. The epitaxial LSM82 thin film (-10 nm) on
YSZ(001) with a GDC buffer layer was also prepared by PLD. Undecorated LSC82 and LSM82
films were used as reference samples. Details for the LSC, LSM, and GDC PLD target syntheses
and PLD deposition process can be found in the Supporting Information (SI).
Oxide phase purity and orientation of the thin film systems were investigated via high
resolution X-ray diffraction (HRXRD) using a four-circle diffractometer (PANalytical, USA and
Bruker D8, Germany). Measurements were performed in normal and off-normal configurations.
The LSC82 in-plane lattice parameter (a lattice parameter) was determined from the off-normal
(202),c peak position (where "pc" denotes the pseudocubic notation) and the c lattice parameter
of LSC82 normal to the film surface was determined from the (00 2 ),c peak position. Surface
morphology was examined by optical microscopy (Carl Zeiss, Germany) and atomic force
microscopy (AFM) (Veeco, USA). AFM images of as-deposited LSC82 and LSM82-decorated
LSC82 films revealed that the surfaces were smooth with the root-mean-square (RMS)
roughness values of 0.74 - 1.12 tm, as shown in Figure 5-1. The RMS roughness was
comparable across all surfaces.
In situ HRXRD was performed on a four-circle diffractometer (PANalytical) in an
oxygen partial pressure, p(O2) of 1 atm and a controlled temperature stage (DHS 900, Anton
Paar). Silver paste was used to adhere the thin film sample to the heating plate. The heating rate
was -10 *C min- and the temperature was held for 20 minutes at each temperature (25 *C,
150 *C, 250 *C, 350 0 C, 450 0 C, and 550 *C) before XRD data were collected. Sample
realignment was conducted at each temperature to maximize the XRD intensities. A full range 029 normal scan was collected, then high-resolution 0-26 normal scans of LSC82 (002)pc and YSZ
(002) were collected. Finally, high-resolution off-normal scans of LSC82 (202),c and YSZ (202)
peaks were obtained. As the thermocouple for this experiment was placed inside the heating
130
stage, a small difference between set and actual temperatures on the sample surface cannot be
excluded.
Figure 5-1. AFM images of (a) as-deposited pristine LSC82 -85 nm, (b) LSC82 with -0.1 nm
LSM82, (c) LSC82 with -0.3 nm LSM82, (d) LSC82 with -0.9 nm LSM82, (e) LSC82 with
-3.5 nm LSM82, and (f) LSC82 with -10 nm LSM82. RMS roughness values were in the range
of 0.74 - 1.12nm and comparable across all surfaces.
Electrochemical impedance spectroscopy (EIS) measurements were performed using a
microprobe station (Karl SUss, Germany) connected to a frequency response analyzer (Solartron
1260, USA) and dielectric interface (Solartron 1296, USA). Thin film microelectrodes -200 gm
in diameter were created by photolithography and acid-etching, and sintered porous Pt deposited
on the back-face of the YSZ prior to thin film deposition served as the counter electrode.
Temperature was controlled at 550 *C with heating stage (Linkam TS1500, UK) and data were
131
collected between 1 MHz to 1 mHz using a voltage amplitude of 10 mV. EIS testing temperature
was calibrated using a thermocouple contacting the thin film surface, and deviation of
5 'C
was observed. EIS experiments were completed in the p(O2) range of 10-3 atm to 1 atm. EIS data
were analyzed using an equivalent circuit shown in the SI (Figure S5-4b), from which the ORR
resistance (RoR) and oxygen surface exchange rate were obtained. EIS data of all samples used
in this study were found to predominantly exhibit a semicircle in the Nyquist plot. The electrical
oxygen surface exchange coefficient (k) was calculated from the resistance of the low-frequency
semicircle (RLF) using kq =
2 RT
4F2 &FAelectrodeCO
23,24,31,43,48-50
where R is the universal gas constant
(8.314 J molP'-K'), T is the absolute temperature (823 K), F is Faraday's constant (96,500
C'mol-1), Aelectode is the area of the microelectrode, and co is the lattice oxygen concentration in
LSC82. Details about the EIS testing procedure, data analysis, and co estimation can be found in
the SI.
Auger electron spectroscopy (AES) was conducted using a Physical Electronics 700
Scanning Auger Nanoprobe (PHI, USA) operating at an accelerating voltage of 10 kV to analyze
the surface chemistry change of the LSM82-decorated LSC82 films after heat treatment. The
films were annealed at 550 *C for 6 hours in an oxygen partial pressure of 1 atm before AES data
were collected. The AES data were collected using two different modes: area mode (three
different 10 pm x 10 pm regions selected across a sample) and point mode (two different -0.45
pm diameter spots selected on a sample) in an ultra-high vacuum chamber. Elemental
quantification of AES spectra utilized relative sensitivity factors (RSFs) of 0.059, 0.027, 0.076,
0.161, and 0.212 for LamN, SrLMM, COLMM, MnLMM, and
OKLL,
respectively, as supplied by the
AES manufacturer (Physical Electronics). Details about AES measurement and analysis can be
found in the SI.
Calculations for energy of La substituted with Sr (SrLa) in bulk Lao.&5r0 .2CoO 3 . and
Lao. 8Sro. 2MnO 3-5
were
performed
using
a 2x2x2
Lao. 7 5SrO. 2 5CoO 3 (with ap (Lao. 7 5Sro. 2 5CoO 3 ) = 3.868
A,
pseudocubic
supercell
structure
of
where ap is the GGA+U perovskite lattice
constant) and Lao. 75Sr0 .2 5MnO 3 (with a,(Lao. 75Sro.2 sMnO 3 ) = 3.945 A) with 2x2x2 k-point mesh
and 450 eV plane-wave energy cut-off. The SrLa substitution energy for Lao.7 5SrO. 2 5CoO
3
(Lao. 75 Sro.2 5MnO 3 ) bulk was taken as the difference in energies between a Lao.62 5SrO. 37 5CoO
3
132
(Lao. 62 5Sro.3 7 sMnO3 ) bulk and a Lao. 7 5Sro.2 5CoO 3 (Lao. 7 5 Sro.2 5 MnO 3) bulk. The supercell
configurations are illustrated in Figure S5-7. Details of DFT are provided in the SI.
Calculations for energy of mixing of La..ySryCol.xMnxO3 relative to the Lai.ySryCoO 3 and
Lai-ySryMnO 3 are performed using the same psudeocubic supercell configurations as shown in
Fig. S7, with varying A-site La and Sr concentration (y=0.25 0.5, and 0.75) and B-site Co and
Mn concentration (x=0, 0.125, 0.25, 0.375, 0.5, 0.625, 0.75, 0.875, and 1). Fully relaxed
calcuations are performed to obtain the relaxed volume (Vol) of the 2x2x2 supercell for
extracting the effective lattice constant ap of the pseudocubic perovskite supercell at various
La/Sr and Co/Mn concentrations (ap = (Vol) 1 3/2). Internal relaxation is then performed by fixing
the supercell lattice constant to 2ap (i.e. 2x2x2 pseudocubic perovskite supercell) to obtained the
DFT total energy. The energy of mixing (AEMIX , eV per formula unit) is calculated based on the
following equation:
AEMIX = EDFT(Lai.ySryCoi.xMnxO 3 ) - (1-x)xEDFT(Lai.ySryCoO 3) - xxEDFT(Lai.ySryMnO 3)
where
EDFT(Lai..ySryCoi.xMnxO 3),
EDFT(Lai-ySryCoO 3 ),
and
EDFT(Lai.ySryMnO 3)
are
.... ()
the
calculated DFT total energy of La1.ySryCoi..Mn.0 3 , Lai-ySryCoO 3, and Lai.ySryMnO 3 normalized
as per formula unit, respectively.
5.3 Results and Discussion
Normal XRD data (Figure 5-2a) of the undecorated LSC82 and LSM82-decorated LSC82
films clearly show the presence of the (00l), (1 is integer) peaks of LSC82 and (OOb)cubic (I is
even) peaks of GDC and YSZ, indicating that the LSC82 film grew epitaxially with the
following epitaxial relationships: (001),eLSC82 // (001)cubicGDC // (OOl)cub~iYSZ. With LSM82
coverage equal to -10 nm in thickness, the (00l)pc (1 is integer) peaks of LSM82 become visible,
which represents (001)pcLSM82 // (00 1),LSC82 // (001 )cubicGDC // (OOl)ubciYSZ. Off-normal
phi-scan analysis of the undecorated LSC82 and LSM82-decorated LSC82 films shows that
LSM82 {101},, LSC82 {101},, GDC {2 0 2 }cubic and YSZ {2 0 2 }cubic have strong peaks with 4fold cubic symmetry (Fig. 2b), which reveals the in-plane crystallographic relationships between
GDC and YSZ (a cube-on-cube alignment), LSC82 and GDC (an in-plane 45* rotation with
133
[100]pc LSC82 // [110]cubic GDC // [1 O]cubic YSZ), and LSC82 and LSM82 (no rotation with
[100]pc LSC82 // [100]pc LSM82), as shown in Figure 5-2c. The relaxed lattice parameters,
a of
the epitaxial LSC82 films with and without LSM82 surface decoration in this study at room
temperature did not change significantly with different LSM82 decoration thicknesses, ranging
from 3.838 - 3.843 A (Table S5-1). Details about lattice parameter calculation and HRXRD of
LSM82 reference film can be found in the SI.
b) LS (101)
a)
LSM - 10 nm
LSM -3.5
g SC (101)
45
nm
b
0.9 nm
LSM~0.1
La/Sr
LSC
Ca
bl)
45-
GDC (20
C
U)
0m
La/Sr
Mn
LSM
)
LSM -
C) C
nm
C
1
*-Ce/Gd
GDC
GC
YSZ (202)
a
P
Zr/Y
YSZ
20
40
60
80
100
iaLL
-180 -90
26/
0 90
(0/0
180
Figure 5-2. X-ray diffraction (Cu Ka) analysis at room temperature. (a) Normal XRD of the
epitaxial LSC82 reference and the LSM82-decorated LSC82 films, (b) off-normal XRD of a
similarly prepared sample with a thicker (-10 nm) LSM82 coverage, and (c) schematic of the
crystallographic
rotational
relationships
among
GDC(001)cuic, and YSZ(OOl)cubic.
134
the
LSM82(001)p,,
LSC82(001)pc,
a)
LSC,,o with
LSMmuo '0.9 nm
4
U
b)
40
0.7
0.6
100
80
60
26/0
c) 0.1
V
9-
-
20
p
HSI
0.0
0.5
-0.1
~0.4
0.3
.
h.tf
0
-0.2
-
.-
I..
U)
S0.2
-0- LSC ref.
0.1 .- 0- LSM - 0.3 nm
-0- LSM - 0.9 nm
0.0 -- 0- LSM - 3.5 nm
-0- LSM -10 nm
-0.1, wmwdWmmmMmmmmMmwmmMmmmmdhmmmdhm
100 200
300
-0.3
-
-0-
-0.4
400
500
600
.fa-5
Temperature / 0 C
LSC ref.
-0- LSM -~0.3 n n
-0LSM -~0.9n n,
-0- LSM - 3.5 n n
-0- LSM -10 n
0
100
200
300
400
500
600
Temperature I 0 C
Figure 5-3. Structural stability and strains of the epitaxial LSC82 with LSM82 coverage (-0.9
nm) thin film. (a) A full-range normal scan in the 0-20 Bragg-Brentano geometry from 150 *C to
550 'C, showing no phase change upon heating at a p(O2) of 1 atm. The starred (*) peaks
originated from the heater, and the peaks of the LSC82, GDC and YSZ are indexed to the pc (apc
~ 3.85 A52 ), cubic (ac ~ 5.42 A6 5) and cubic (ac ~ 5.15 A6 6) structure, respectively. (b) The inplane strains,
Eaa
and (c) the out-of-plane strains, E, of the epitaxial LSC82 and LSM82-
decorated LSC82 films as a function of temperature.
In situ HRXRD was conducted to show that the undecorated LSC82 and LSM82decorated LSC82 films were structurally stable upon heating to 550 *C in an oxygen partial
pressure of 1 atm. Upon heating to 550 *C, only peak shifts toward low diffraction angles
associated with the thermal expansion of LSC82 were observed, as shown in Figure 5-3a and
135
Figure S5-2. Both in-plane and out-of-plane strains of LSC82 films were not strongly influenced
by the LSM82 coverage, which is supported by the fact that the lattice constant of LSM82 (ap,
3.89
A
=
for LSM bulk") is very close to that of LSC82 (ape - 3.85 A for LSC bulk 2 ). The
volumetric thermal expansion coefficients (TECs) of the epitaxial LSC82 films with and without
LSM82 decoration (13.8 x 10-6 Kl
-
14.9 x 10-6 K~) were in good agreement with those
reported in the literature (14.9 x 10-6 K1 for LSC thin films 3 8 ). In-plane (s.) and out-of-plane
(S(e) strains of the epitaxial LSC82 films with and without LSM82 decoration were found to be
temperature dependent (Figure 5-3b and 3c), which can be attributed to the difference in the
thermal expansion coefficients between the LSC82 (14.9 x 10-6 K 1 ) and YSZ substrate (8.8 x 106 K 15 3 ). In-plane strains changed from tensile to compressive while the out-of-plane
strains
varied from compressive to tensile upon heating. Interestingly, in-plane and out-of-plane strains
of the LSC82 films with and without LSM82 coverage at 550 *C were found to be very small (0.05 - 0.06 %). Details about in situ HRXRD of LSC82 films with LSM82 decoration,
temperature dependent strains and lattice parameters are provided in the SI.
EIS data collected from the undecorated LSC82 and LSM82-decorated LSC82 films at
550 *C with an oxygen partial pressure, p(O2) of 1 atm are shown in Figure 5-4a. The real
impedance of the predominant semicircle decreased significantly with LSM82 coverage less than
or equal to -0.9 nm in thickness while it increased with larger LSM82 thicknesses. In addition,
the predominant semicircle was found to increase with decreasing p(O2), where EIS data of all
samples used in this study were found to show nearly perfect semicircle impedances.9
Representative EIS data collected from the LSC82 film with ~0.3 nm LSM82 coverage measured
at 550 'C as a function of p(O2) are shown in Figure S5-4c. Considering the fact that the film
thicknesses are much smaller than the critical thickness (estimated to 1 Pm for bulk LSC at 550
,C 6), the p(02)-dependent impedance responses suggest that the oxygen
surface exchange
kinetics governs the oxygen electrocatalysis on the film surface. The LSC82 films with partial
LSM82 coverage (-0.1, -0.3, and -0.9 nm) exhibit enhanced kq relative to the undecorated
LSC82 film by up to nearly two orders of magnitude while the LSC82 films with full LSM82
coverage (-3.5 and -10 nm) have similar or much lower kq relative to the undecorated LSC82
film, as shown in Figure 5-4b. Interestingly, the k value of the LSC82 film with -10 nm LSM82
coverage was found to be comparable with that of the LSM82 reference film. Details of EIS data
collected from the LSM82 reference film can be found in the SI.
136
Y/ so
2500
b) 104
0O.9 nm
0
0.3 nm
rA
2000
10l nm'
1150
10
9
U
- 0.9 nm
0.3 nm
-rb
0.
It
-
~so. lkf
I165
krOm
'S
10
0
0 mali
1000 2000 3000 4000
Real Impedance I ka
0
LM
450
0?
400
E0
E
0
base
film
S.
-I
100
-0.1'""n
- 3.5 nm
~10 nm
r
L&M-IUn
I
101 r
0 300250
1012 U
200 550
.1 At'I
-2
.1
I
LSM' film
U. 350:
-3
I
0'
L8M-3.Unm
400
S
-
In
S
M
HOC
-2
-3
-1
0
log p(0 2) / atm
0
log p(O2) / atm
Figure 5-4. Electrochemical impedance spectroscopy (EIS) results for the bare LSC82 films and
the LSC82 films with -0.1 (yellow), -0.3 (red), -0.9 (blue), -3.5 (green), and -10 nm (light
blue) LSM82 decorations at 550 *C. (a) Nyquist plot of the epitaxial LSC82 and the epitaxial
LSM82-decorated LSC82 films in 1 atm. Inset shows a magnification of the Nyquist plot of the
LSC82 films with partial LSM82 coverage (-0.1, -0.3, and -0.9 nm), (b) oxygen partial pressure
dependency of the surface exchange coefficients (kg) of the LSC82 and LSM82-decorated
LSC82 films calculated from EIS spectra collected at 550 *C, and (c) oxygen partial pressure
dependency of volume specific capacitance (VSC) of the epitaxial LSC82 and LSM82-decorated
LSC82 films calculated from EIS spectra collected at 550 *C.
The volume specific capacitances (VSCs) extracted from EIS data of the epitaxial LSC82
and
LSM82-decorated
LSC82
films,
corresponding
to
the
change
in the
oxygen
nonstoichiometry (6) induced by the change in the electrical potential, did not change
137
significantly with LSM82 decoration, as shown in Figure 5-4c. Indeed, the VSCs in this study
were comparable to those of epitaxial LSC82 thin films reported previously. 24 ,
43
This
observation indicates that the oxygen content in the LSC82 films with and without LSM82
coverage do not contribute to the modification of the k values observed. Details of VSCs are
provided in the SI.
To investigate the change of surface chemistry and surface morphology as a function of
LSM82 thickness, AES and AFM were conducted on the undecorated LSC82 and LSM82decorated LSC82 films before and after annealing at 550 *C for 6 hours in an oxygen partial
pressure of 1 atm, as shown in Figure 5-5. No particles were observed in the SEM micrographs
for undecorated LSC82 and LSM82-decorated LSC82 surfaces before annealing, as shown in
Figure 5-5b, 5-5e, and 5-5h. After annealing, discrete particles were noted on the undecorated
LSC82 and LSC82 with -0.1 nm LSM82 coverage surfaces (Figure 5-5c, 5-5f, 5-5j, and 5-5k),
whereas no particles were found on LSC82 surfaces with LSM82 decorations thicker than -0.3
nm (Fig. 5i, 51, 5m, 5n, and 5o). These discrete particles observed on the undecorated LSC82 and
-0.1-nm-LSM82-decorated surfaces after annealing were found to have higher Sr Auger signals
(Figure 5-5a and 5-5d) than the rest of the surface as well as the surfaces before annealing, which
indicates that Sr-enriched particles were formed after annealing in oxygen. This observation is
consistent with our recent in situ studies of surface structure and chemistry changes of LSC82
films 5 4 and the formation of Sr-enriched particles on annealed LSC64 film surfaces. 4 5 It is noted
that these Sr-enriched particles on the surface of LSC82 were suppressed by LSM decoration.
The number density of Sr-enriched particles on the -0.1-nm-LSM82-decorated LSC82 surface
(1.4 particles per gm 2, ~ 3 % area coverage) was considerably lower than that of uncoated
LSC82 surface (6.1 particles per llm2 , ~ 13 % area coverage, estimated by averaging particles in
three 5 x 5
1m2
areas in SEM images). LSM82 decoration thicker than -0.3 nm resulted in the
absence of Sr-enriched particles on the surface, which may result from the incorporation of Sr
from LSC82 into the LSM bulk lattice similar to the incorporation of SrO into LSM as suggested
previously.55 , 56 The chemistry of these Sr-enriched particles is not well understood. Recent
COBRA experiments suggest that these Sr-enriched particles on (001)-oriented LSC82 thin films
have a composition approaching to that of SrCoO 3.4 1 Unfortunately, the surface of SrCoO 3 is
most likely decomposed to secondary phases such as SrO/Sr(OH) 2 /SrCO 3 57 '
which can greatly impede the surface exchange kinetics. 60
138
58
and La 2 CoO 4 ,59
.
1612
1636
1M
Kinetc Energy / V
LSC Priwtie
LS
2
1012
1633
1664
Kinec Energy I eV
0
m
LSg
1612
1638
m
1664
Kinetic Energy / eV
Figure 5-5. AES, SEM, and AFM analysis for bare LSC82 and LSM82-decorated LSC82 films
before and after annealing. Annealing was performed at 550 *C in an oxygen partial pressure of 1
atm. (a) Sr Auger spectra of bare LSC82 thin film probed for: as-deposited surface (gray), and
particles (dark red) and film surface (dark blue) after annealing. SEM image of (b) as-deposited
and (c) annealed LSC82. (d) Sr Auger spectra of LSC82 with -0.1 nm LSM82 probed for: asdeposited surface (gray), and particles (orange) and film surface (light blue) after annealing.
SEM image of (e) as-deposited and (f) annealed LSC82 with -0.1 nm LSM82. (g) Sr Auger
spectra of LSC82 with -0.3 nm LSM82 probed for: as-deposited surface (gray), and annealed
surface (blue). No particles were observed. SEM image of (h) as-deposited and (i) annealed
LSC82 with -0.3 nm LSM82. AFM images also showed particle formation on (j) annealed
LSC82, (k) annealed LSC82 with -0.1 nm LSM82, but no particles were observed on (1)
annealed LSC82 with -0.3 nm LSM82, (m) annealed LSC82 with -0.9 nm LSM82, (n) annealed
LSC82 with -3.5 nm LSM82, or (o) annealed LSC82 with -10 nm LSM82.
The increase of the observed k values of LSC82 films with partial LSM82 coverage can
be attributed partly to the suppression of the Sr-enriched particles on the surface. Partial LSM82
coverage on LSC82 up to 0.9 nm (Figure 5-4b) enhanced the k values of LSC82 films, where
the highest k4 value was obtained with LSM82 coverage of -0.3 nm. This hypothesis is further
supported by the work of Kubicek et al.45 , which shows a detrimental influence of Sr-enriched
139
particles on the oxygen surface exchange activity of LSC64 films, having improved surface
activity by removal of these particles via chemical etching. In contrast, increasing LSM82
coverage on the LSC82 surface from 0.9 to 10 nm resulted in a reduction in the kq value and the
10-nm-LSM82-decorated LSC82 surface had the slowest surface oxygen exchange kinetics
approaching that of the LSM82 reference film, which will be discussed later.
5 ;0 0C
a )10
p(02 )=
,
b )10 -
0.40
,
;
P(
atm
100.35
2 )=
1 atm
*
-~
U
0.30
7
_
0.a
10
10
1.0
.
-
u-Mn/(Co+Mn)
0.60
10
\t
N
0.4
0.255.- ---
101
0.
LSCm
0.1
W 0.20
'
'
'
0.3
0.9
3.5
10
4 ).
0.15
10
.
10_
10
LSC
LSM Thickness /nm
0
0.3
0.1
0.3
0.0
0.9
3.5
10
LSM Thicknessl nm
Figure 5-6. Surface exchange coefficients (kg) of the LSM82-decorated LSC82 films calculated
from EIS spectra collected at 550 'C in a p(O2) of 1 atm and normalized cation intensity ratios
extracted from area mode using AES after annealing at 550 'C in a p(O2) of 1 atm. (a) kq (0gray) and normalized La and Sr intensity ratio (0-light blue) and (b) k' (0-gray)
and
normalized Co and Mn intensity ratio (0-light red) as a function of LSM82 thickness.
The increase of the observed kq values of LSC82 films with partial LSM82 coverage up
to 0.9 nm can be also attributed to increased Sr concentration in the perovskite structure near the
surface. As shown in Figure 5-6a, the Sr/(La+Sr) ratios extracted from the AES data using area
mode were found to increase with increasing LSM thickness up to 0.9 nm, and reach a plateau
from 0.9 nm to 10 nm considering experimental uncertainty. This hypothesis is in agreement
with previous findings that greater Sr concentration on the A-site of the perovskite structure
leads to increased oxygen surface exchange kinetics in LSC6 1 -63 and LSM. 5'6 The increased Sr
concentration near the surface of the LSC82 with LSM82 decoration may be associated with the
140
thermodynamic driving force of exchanging Sr between LSC82 and LSM82. It should be noted
that the energy gain for exchanging a Sr from LSC82 with a La from LSM82 was found to be
-0.6 eV, calculated from the same method as that used in previous work, 64 where -0.9 eV energy
gain for exchanging a Sr from Lao. 7 5Sro.2 5CoO 3 - with a La from LSC2 14 leads to a large driving
force for interdiffusion across the heterostructure interface. Although more detailed study of the
origin of the increased surface Sr concentration is needed, the thermodynamic driving force of
exchanging Sr between LSM82 and LSC82 can elucidate the change in the surface Sr
concentration. Details of DFT are provided in the SI.
With LSM82 coverage greater than -0.3 nm in thickness, the Mn/(Co+Mn) ratio of
LSM82-decorated LSC82 film surfaces significantly increased, as shown in Figure 5-6b.
Therefore, the reduction in the surface oxygen exchange kinetics of LSC82 with LSM82
decoration greater than 0.3 nm can be explained with increasing Mn concentration on the B-site
of the perovksite structure near the surface. This hypothesis is supported by the work of De
Souza and Kilner,'' 6 which shows that the k* values of Lai-xSr.Coi-yMnyO3+Z (LSCMO) decrease
with increasing Mn concentration. Details of AES analysis and AES spectra of each film can be
found in the SI.
It should be noted that LSM82 coverage can also prevent the time-dependent degradation
of the oxygen surface exchange kinetics of the LSC82 thin film, as shown in Figure 5-7a. The
LSC82 film with LSM82 coverage -0.9 nm was found to show stable k values for over 70 hours
annealing whereas the k values of the undecorated LSC82 were significantly reduced (- 2 orders)
after annealing for 20 hours. Moreover, LSM82 coverage was found to lead to higher timedependent stability of LSC82 relative to the LSC 214-decorated LSC82, as shown in Fig. S8. The
enhancement of the time-dependent surface exchange kinetics of the LSC82 with LSM82
coverage may be attributed to the enhancement of the surface stability by the Mn content. The
energy of mixing for LSCMO based on the DFT calculations (Figure 5-7b) suggests that the
formation of LSCMO is energetically favorable relative to the LSC and LSM at wide range of Sr
concentrations, where the greatest stabilization is around Mn concentration (xMn) equal to 0.5
with mixing energy of -0.10, -0.13, and -0.04 eV per formula unit at Sr = 0.25, 0.5 and 0.75,
respectively. The ab initio energy of mixing results indicate that the increase in Mn content can
stabilize the LSCMO with respect to the LSC at a wide range of Sr concentrations, which
suggests that alloying LSC with LSM can enhance the stability of LSC, preventing surface
141
secondary (non-perovskite) phase formation or phase decomposition, and thereby increases the
surface stability of the LSC82 at intermediate temperature. This hypothesis can also be supported
by the surface stabilization of the LSC82 with LSM82 coverage resulting from the suppression
of the surface Sr-enriched particles by LSM82 coverage, as shown in Figure 5-5. Our
experimental and theoretical results clearly elucidate that the time-dependent stability of LSC82
can be enhanced by the surface decoration with either LSM82 or LSC214, which can also lead to
the enhancement of the degradation of the surface exchange kinetics of LSC82 after long-term
annealing.
a) -7-
b)0-04
5;----------0
c
p(0) 2=1 atm
10
0.00 -----
--
Csell
LSM - 0.9 nm
010
U...0.08
-
% 10
ref.
40-@Sr
10
-p
C. -0.12
0
--
n0.5
.1
0.16
10
20
30
40
50
60
Sr a .75
0.0
70
Annealing Time / hrs
(LaSr)MnO3
M
0.2
0.4
0.8
0.6
X.
(LaSr)Co,.,Mn0
1.0
(LaSr)CoO 3
3
Figure 5-7. (a) Surface exchange coefficients (ku) of the LSC82 (0-black) and LSC82 with ~0.9
nm LSM82 (0-blue) as a function of annealing time. (b) The energy of mixing for Lai.xSrxCoi.
yMnyO 3 (LSCMO) based on the DFT calculations as a function of Mn concentration with Sr =
0.25 (0-light red), Sr
=
0.5 (0-red), and Sr
=
0.75 (0-dark red). LSCMO can be stabilized as
increasing Mn concentration.
5.4 Conclusion
We show that LSM82 surface decoration coverage can strongly affect the surface
exchange kinetics and the surface stability of epitaxial LSC82 thin films. The kq values of the
142
epitaxial LSC82 thin films with partial LSM coverage (-0.1, -0.3, -0.9 nm) are significantly
enhanced relative to the undecorated LSC82 film while those with full LSM coverage (-3.5 and
-10 nm) are similar or diminished. In situ HRXRD reveals that the in-plane and out-of-plane
strains of the LSC82 at elevated temperature are not influenced by LSM82 decoration. AFM and
SEM show the suppression of surface Sr-enriched particles by LSM82. AES analysis also shows
that the surface Sr and Mn contents increase with increasing LSM82 coverage. The change in the
surface cation concentration and the suppression of the formation of secondary passive phases as
a result of LSM82 decoration can be responsible for observed oxygen surface exchange kinetics.
In addition, the time-dependent degradation of the oxygen surface exchange kinetics of the
LSC82 thin film is markedly improved by LSM82 coverage, which may be attributed to the Mn
content resulting in the surface stabilization of LSC82 revealed by DFT calculation. Our results
demonstrate that small changes in the surface chemistry of perovskites can yield significant
increases in the oxygen surface exchange kinetics and the surface stability with control of surface
decoration thickness, which can be potentially utilized to develop highly active oxygen surface
exchange materials for applications in the field of solid-state electrochemistry such as
intermediate
temperature
SOFC cathodes,
solid-electrolyte-based
sensors,
and oxygen
conducting membranes.
5.5 Supporting Information (SI)
Experimental Details
Target Synthesis. Lao.8Sro.2CoO 3- (LSC82) was synthesized using solid state reaction from
stoichiometric mixtures of La20 3 , SrCO 3 , C030 4 (Alfa Aesar, USA) calcined at 1,000 *C in air
for 12 hours. The Lao.Sro.2MnO3-M (LSM82) and Gdo 2 Ceo. 80 2 (GDC) were prepared by the
Pechini methods 67 . La(N0 3)3*6H 20, Sr(N0 3)2 , Mn(NO 3)2 o6H2 0, and separately Gd(N0 3)3 and
Ce(N0 3)3 were dissolved in de-ionized water with ethylene glycol, and citric acid (SigmaAldrich, USA) mixture to synthesize LSM82 and GDC respectively. After esterification at 100
0C,
the resin was charred at 400 'C and finally calcined at 1000 "C in air for 12 hours. Pulsed
laser deposition (PLD) target pellets with 25 mm diameter were subsequently fabricated by
143
uniaxial pressing at 50 MPa. The LSC82, LSM82, and GDC pellets were fully sintered at 1,300
*C in air for 10 hours, 1,350 C in air for 12 hours, and 1,100 *C in air for 14 hours, respectively.
Sample preparation. Single crystal 9.5 mol% Y 20 3-stabilized ZrO 2 (YSZ) wafers with (001)
orientation and dimensions of 10 x 5 x 0.5 mm (MTI corporation, USA), were used as substrate.
Prior to LSM82, LSC82, and GDC deposition, platinum ink (Pt) (#6082, BASF, USA) counter
electrodes were painted on one side of the YSZ and dried at 900 *C in air for 1 hour. PLD was
performed using a KrF excimer laser at I = 248 nm, 10 Hz pulse rate and 45 mJ pulse energy
under p(O2) of 50mTorr with 500 pulses of GDC (-5 nm) at 550 *C, followed by 15,000 pulses
of LSC82 (-85 nm) at 650 "C. PLD was also performed using the same laser conditions under
p(O2) of 100mTorr with 500 pulses of GDC (-5 nm) at 550 'C, followed by 1,500 pulses of
LSM82 (-10 nm) at 750 *C. The film thicknesses were determined by atomic force microscopy
(AFM). The utilization of reflection high-energy electron diffraction (RHEED) enabled
diagnostic in-situ monitoring of the LSC82 film growth. Immediately after completing the
LSC82 base film deposition, LSM82 films were subsequently deposited; for the LSM82 surface
coverages consisting of 15 pulses (-0.1 nm, partial coverage), 30 pulses (-0.3 nm, partial
coverage), 100 pulses (-0.9 nm, partial coverage), 500 pulses (-3.5 nm, full coverage), and
1,500 pulses (-10 nm, full coverage). LSM82 decoration layer thickness is extrapolated from
AFM of the 500 pulses and 1,500 pulses LSM82 coverage on LSC82. After completing the final
deposition, the sample was cooled to room temperature in the PLD chamber for -1 hour under an
oxygen partial pressure of 50 mTorr.
HRXRD analysis of LSM82 thin film. Normal XRD data (Figure S5-la) of the undecorated
LSM82 film clearly show the presence of the (00l),c (I is integer) peaks of LSM82 and
(Obcubic
(I is even) peaks of GDC and YSZ, indicating that the LSM82 film grew epitaxially with the
following epitaxial relationships: (001),cLSM82 // (00l)cubicGDC // (001)cubicYSZ. Off-normal
phi-scan analysis of the undecorated LSM82 shows that LSM82 {101},, GDC {2 0 2 }cubic and
YSZ {2 02 }cubic have strong peaks with 4-fold cubic symmetry (Figure S5-lb), which reveals the
in-plane crystallographic relationships between GDC and YSZ (a cube-on-cube alignment), and
LSM82 and GDC (an in-plane 45" rotation with [100],c LSM82 // [110]cubic GDC / [1 10]cubic
144
YSZ). The relaxed lattice parameter,
a
of the epitaxial LSM82 film in this study at room
temperature were -3.889 A.
Relaxed lattice parameter determination by in situ HRXRD. The Relaxed lattice parameter 4
and o are derived from the following equation (where a and ^ are the relaxed lattice parameters
38, 436
for the film in an unstrained state),'e'
=
-2v (a-a!)
a-
, assuming
a=
,
and v
=
0.25.6.1 The
in-plane strain is given by: Eaa = (a-a) and the out of plane strain by: Ecc =
Temperature induced peak shift in in situ HRXRD. Based on Bragg's Law, a shift towards
smaller 26-values with increasing temperature is expected for materials having a positive thermal
expansion coefficients, such as LSC, GDC, and YSZ (with increasing temperature the spacing
between the planes increases ( dT= 25 *C < d = 550 *C) and thus OT= 25 *C > 6 r = 520 *C).
Unit cell volume calculation using in situ HRXRD data. Based on the Y-doping concentration
of the used ZrO 2 single crystals (YSZ) their crystal structure is cubic (lattice parameter a = b =
c), and thus the lattice parameter, a, can directly be determined from the YSZ (202) reflection.
Once a is known, the unit cell volume of YSZ can be calculated as a 3 . LSC is pseudo cubic,
where the c-axis lattice parameter can be determined from the LSC (002) reflection, and
assuming a- and b-axis lattice parameters are equivalent, they can be determined from the
geometrical relationship between the c axis and the LSC (101) reflection. Once c and a = b are
determined, the unit cell volume can be obtained by c-a2
Microelectrodes Fabrication.
In situ electrochemical
impedance spectroscopy (EIS)
measurements were conducted to probe ORR activity on geometrically well-defined LSM82decorated LSC82 microelectrodes fabricated by photolithography and acid etching, where
sintered porous Pt sintered onto the backside of the YSZ substrate served as the counter electrode.
OCG positive photoresist (Arch Chemical Co., USA) was applied on the LSM82-decorated
LSC82 surface and patterned using a mask aligner (Karl Stiss, Germany, . = 365 nm). The
photoresist was developed using Developer 934 1:1 (Arch Chemical Co., USA) and the thin
films were etched in hydrochloric acid (HCL) to remove LSM82-decorated LSC82 film excess
145
and create the circular microelectrodes (diameters -50 pm, -100 pm, ~150 gm, and -200 pm,
exact
diameter
determined
by
optical
microscopy).
Before
electrochemical
testing,
microelectrode geometry and morphology was examined by optical microscopy (Carl Zeiss,
Germany) and atomic force microscopy (AFM) (Veeco, USA).
Electrochemical Characterization.
Fig. S4b and S4c detail the equivalent circuit and
corresponding Nyquist plot for this experimental system. ZView software (Scribner Associates,
USA) was used to construct the equivalent circuit and perform complex least squares fitting. The
EIS data were fitted using a standard resistor (R1) for HF and resistors
(R2)
in parallel with a
constant phase elements (CPE2) for MF and LF (Ri-(R2/CPE2)-(Ro0R/CPEopj)). Based on the
p(O2) dependence of the three features, physical or chemical process with regard to each
frequency range can be determined. 9, 12 , 69, 70 The HF feature was found unchanged with p(O2),
and its magnitude and activation energy (-1.15 eV) were comparable to those of oxygen ion
conduction in YSZ reported previously 7 1 . The MF feature, which was found to have a p(O2)
independent feature, was attributed to interfacial transport of oxygen ions between the LSC82
film and the GDC layer. In addition, the magnitude of its capacitance was relatively small (~10-6
F) compared to the LF feature (~10-3 F). The LF feature was found to have a strong p(O2)
dependence. The resistance of the LF feature drastically increases as oxygen partial pressure
decreases. In the case of thin film samples, the magnitude of capacitance is due to the oxygen
content change in the films. Therefore, the electrode oxygen surface reaction corresponds with
= 0.25 x delectrode 2 ). Then, we can determine the ORR area specific resistance (ASRoRR = RoR
AeIectrode). The electrical surface exchange coefficient (kY), which is comparable to k*7
-
the LF feature. We obtained values for Roj and knowing the area of the microelectrode (Aeectrode
was
determined using the expression,50 ,7 3
0 = R T / 4F2RO1RAelectrodeCo
where R is the universal gas constant (8.314
(1)
J mol-1
K- 1),
T is the absolute temperature, F is the
Faraday's constant (96,500 C molP), and co is the lattice oxygen concentration in LSC82 where
co = (3-6)IV.,
(2)
V. is the molar volume of LSC82 at room temperature. In this study, c. was calculated with 6
.
extracted from previous reported values 14
146
The electrical surface exchange coefficient (/0) of the LSM82 thin film was also
determined using the same manner. EIS data collected from the undecorated LSM82 film at
550 *C as a function of p(02) are shown in Figure S5-5. The predominant semicircle was found
to increase with decreasing oxygen partial pressure, where EIS data of the LSM82 was found to
show nearly perfect semicircle impedances.9 Considering the fact that the film thickness is
smaller than the critical thickness (-65 nm for epitaxial LSM thin film at 750 *C74 ), the p(O2)dependent impedance responses suggest that the oxygen surface exchange kinetics governs the
oxygen electrocatalysis on the film surface.
The LSM82 surface coverage may change the c0 value of the system. For estimating this
influence we compared LSM82 co values with LSM82 c. values. However, calculated c, values
for LSM82 was only -1
- 2 % different from those for LSC82. We therefore decide to use co
values for LSC82 for all samples.
VSC, indicative of changes in the oxygen nonstoichiometry induced by changes in the
electrical potential, can be obtained from EIS data via the expression7 5
VSC = [lI/(Aelectrode e thickness)]((Ro) "Q)",
where
Q
(3)
is the non-ideal "capacitance", and n is the non-ideality factor of CPE. The fitted
values of n for semi-circle CPEORR were found to range from ~ 0.96 to 1.0 over the entire PO2
range examined (n =1, ideal).
Experimental details of auger electron spectroscopy (AES).
In AES, the obtained energy
spectrum for a particular element is always situated on a large background (low signal-to-noise
ratio), which arises from the vast number of so-called secondary electrons generated by a
multitude of inelastic scattering processes. To obtain better sensitivity for detection of the
elemental peak positions, the AES spectra from this study are presented in the differentiated
form. Elemental quantification of AES spectra utilized relative sensitivity factors (RSFs) of
0.059, 0.027, 0.076, 0.161, and 0.212 for LamNN, SrLM, COLMM, MnLMM, and OKLL, respectively,
as supplied by the AES manufacturer (Physical Electronics). In addition, the Inelastic-MeanFree-Path (IMFP) was calculated to correct signal intensity for their different IMFPs
(information depth). IMFPs were calculated using the NIST Standard Reference Database 71
"NIST Electron Inelastic- Mean-Free-Path Database" version 1.2. The software program
provides the ability to predict the IMFP for inorganic compounds supplying the stoichiometric
147
composition of La (0.8), Sr (0.2), Co (1), and 0 (3), the number of valence electrons per
molecule (assumed to be 29.8), the density (6.931 g/cm 3 ) and a band gap energy (for which we
are assuming 0 eV as LSC82 is metallic like at high temperatures; additionally when assuming a
band gap of an insulator 5 eV, the IMFP increases by -0.03 nm). The IMFP for La, Sr, and Co
were determined to be 1.337, 1.337 and 2.549 nm, respectively. A relative depth-scaling factor
(s) was determined as:
a
(4)
exp (-l xdx,
0 Ai
A
where A is the IMFP, yielding asr
=
elements were scaled using Iscaled
0.41, and
uLa
and
Imeasured*Si/Si.
Uco
= 0.63. The intensities from different
Similarly, the IMFP for La, Sr, and Mn of
LSM82 was determined to be 1.283, 1.283, and 2.640 nm, respectively, by using the the
stoichiometric composition of La (0.8), Sr (0.2), Mn (1), and 0 (3), the number of valence
electrons per molecule (assumed to be 27.8), the density (6.414 g/cm 3) and a band gap energy
(assumed 0 eV). The obtained values of the relative depth-scaling factor for LSM82 are thus
approximately equal to those of LSC82. The La and Sr concentration (CLa or Csr) was obtained by
normalizing to the their sum, Ci=Ii/(ILa+ISr). The Mn and Co concentration was also obtained by
using the same manner.
Details of density functional theory (DFT) calculations. Spin polarized Density Functional
76 77
Theory (DFT) calculations were preformed with the Vienna Ab-initio Simulation Package '
using the Projector-Augmented plane-Wave method78 with a cutoff of 450 eV. Exchangecorrelation was treated in the Perdew-Wang-91 79 Generalized Gradient Approximation (GGA).
The pseudopotential configurations for each atom are as follows: La: 5s2 5p65d'6s2 , Srsv:
0 are
4s24 65s2 , Mnpv: 3p63d 64s', Co: 3d4s' and Os: 2s2 2p . The GGA+U calculations
performed with the simplified spherically averaged approach 1 , where the Uff (Uff= Coulomb U
- exchange J) is applied to d electrons. (Uft(Mn)
=
4.0 eV and Uef(Co) = 3.3 eV) 64 , 82. All
calculations are performed in the ferromagnetic state in order to use a consistent and tractable set
of magnetic structures, and the spin states for the calculated La1.XSrXCoO3 and Lai..SrxMnO3
systems are: Mn: high spin and Co: intermediate spin
Calculations for energy of La substituted with Sr (SrLa) in bulk Lao.5r0 . 2 CoO 3-8 and
LaO. 8Sro.2 MnO3+8
were
performed
using
a 2x2x2
148
pseudocubic
supercell
structure
of
Lao.75 Sro.2 5CoO 3 (with a, (Lao. 7 5Sro.2 sCoO 3 )
=
3.868
A,
where ap is the GGA+U perovskite lattice
constant) and LaO.7 5 Sro.2 5MnO 3 (with a, (Lao.7 5 Sro.2 5MnO 3 )
=
3.945 A) with 2x2x2 k-point mesh
and 450 eV plane-wave energy cut-off. The SrLa substitution energy for Lao.7 5Sro.2 5CoO 3
(Lao.7 sSro.2 5MnO 3 ) bulk was taken as the difference in energies between a Lao.62 5 Sro.3 7 5CoO 3
(Lao. 62 5Sro.37 5MnO 3 ) bulk and a La0 .75Sro. 2 5CoO 3 (Lao.7 5Sro.2 5MnO 3) bulk. The supercell
configurations are illustrated in Figure S5-7.
149
Table S5-1. Constrained and relaxed lattice parameters of LSC82 and LSM82-decorated LSC82
films extracted from normal and off-normal XRD data as a function of temperature. Constrained
normal and in-plane lattice parameters of all films were calculated from combining the interplanar distance of the LSC(002),c and LSC(202)p, peaks.
In-plane
strain
Normal
strain
Constrained
in-plane
a/A
Constrained
normal
C/A
Relaxed lattice
parameter
aIA
RT
3.861
3.823
3.838
0.594
-0.396
150 'C
3.866
3.833
3.846
0.515
-0.343
250 0C
3.87
3.847
3.856
0.358
-0.239
3
C
3.872
3.857
3.863
0.233
-0.155
450 'C
3.874
3.866
3.869
0.124
-0.083
550 'C
3.879
3.877
3.878
0.031
-0.021
RT
3.865
3.825
3.841
0.617
-0.412
150 'C
3.869
3.837
3.850
0.497
-0.332
250 'C
3.872
3.848
3.857
0.370
-0.247
350 OC
3.874
3.859
3.865
0.236
-0.157
450 OC
3.876
3.869
3.872
0.100
-0.067
550 oC
3.879
3.880
3.879
-0.007
0.005
RT
3.867
3.825
3.842
0.655
-0.436
150 OC
3.871
3.835
3.849
0.563
-0.376
250 oC
3.873
3.845
3.856
0.431
-0.288
350 OC
3.875
3.856
3.864
0.288
-0.192
450 'C
3.878
3.867
3.871
0.160
-0.107
550 OC
3.880
3.877
3.879
0.047
-0.031
RT
3.867
3.826
3.842
0.635
-0.424
150 OC
3.869
3.836
3.849
0.510
-0.340
250 OC
3.871
3.847
3.857
0.373
-0.249
350 'C
3.873
3.858
3.864
0.221
-0.147
450 OC
3.877
3.869
3.872
0.117
-0.078
550 OC
3.880
3.880
3.880
0.004
-0.002
RT
3.866
3.827
3.843
0.616
-0.411
150 OC
3.870
3.836
3.850
0.525
-0.350
LSM
250 OC
3.872
3.847
3.857
0.399
-0.266
- 10nm
350 'C
3.874
3.859
3.865
0.237
-0.158
450 OC
3.878
3.869
3.873
0.135
-0.090
550 OC
3.881
3.877
3.879
0.062
-0.041
LSC ref.
LSM
~ 0.3 nm
LSM
~ 0.9 MnM
LSM
~ 3.5 nm
Temperature
150
(a
-
)/%
/
Samples
-
a
(c/)%
ai
(a)LSM
(202
040
26/
2LS0
20
40
60
80
*
-1
100
-180
-90
0
90
180
Figure S5-1. X-ray diffraction (Cu K,) analysis at room temperature. (a) Normal XRD and (b)
off-normal XRD of the epitaxial LSM82 reference film. Off-normal XRD shows the in-plane
crystallographic relationships between GDC and YSZ (a cube-on-cube alignment), and LSM82
and GDC (an in-plane 450 rotation with [IOO]pc LSM82 // [I
151
O]cubic
GDC // [1
I]cubic
YSZ).
(a)
LSC(002) with LSM ~ 0.9nm
(b)
LSC(202) with LS M
0.9 nm
550 0C
550 C
450 0 C
450 C
350 0
350 0 C
C)
RT
46
47
48
RT
67
49
20
68
69
70
71
20/ 0
Figure S5-2. In situ HRXRD data of (a) the normal scan of the LSC(002)p, peak and (b) the offnormal LSC(202)pc as a function of temperature in a P(O2) of 1 atm. Here, we observe the peak
shifts towards lower angle in the 0-20 with increasing temperature from 25 'C to 550 'C.
152
0.
00
-%
Q
C4
00
0
2 Q 0
U) 0) Lr)
C4
0
U
-
U
-
U
C4 C4
00
N a
0N
%...
L)
00
N
U)
-1 Cn
CD
~
C
C
%..wflU
_j _J
550 0
U
C
450 0 C
*
-
(ID
C)
350 0 C
U
-
20
40
60
a
a
80
100
NU
20/ 0
Figure S5-3. A full-range normal scan in the 6-26 Bragg-Brentano geometry with the epitaxial
LSC82 with LSM82 coverage (10
nm) from 150 'C to 550 'C, showing no phase change upon
heating at a p(02) of 1 atm. The starred (*) peaks originated from the heater, and the peaks of the
LSC82, GDC and YSZ are indexed to the pc (aPe ~ 3.85 A52 ), cubic (ac ~ 5.42 A65) and cubic (ac
~ 5.15 A 66) structure, respectively.
153
(a)
(b)
R1
R
(HF)
P
0
Substrate
PFLgR
(MF) Interface
800
V
RORR
(LF) Film
WF
EC
~14-
1.atm
0
I8-
010
~6
6 00
atm 4
2400
10.
r
0
5
10
15
20
2
25
9
Real Impedance I kn
2U
a200
0.1 atm
'009
Figure
S5-4.
(a)
Schematic
of a LSM/LSC/GDC/YSZ(OO1)/porous
Pt sample
and
electrochemical testing configuration (not drawn to scale), and (b) equivalent circuit (R 1 =YSZ
electrolyte resistance, R2
CPE
=
=
electrode/electrolyte interface reitne3
RORR
=
ORR resistance,
constant phase element) used to extract ORR kinetics, and (c) Nyquist plot of the
epitaxial LSC82 with ~O.3 nm LSM82 coverage at 550 0 C; inset shows a magnification (HF: i0 4
~i0 5 Hz, ME: i0 3
i 04 Hz, and LF: 10.2 ~i0 3 Hz).
154
6000
-
W
U
LSM ~10 nm
10
0
atm 0
-.4
M04000
*
0
oc
00
E 00
0.S2000
/
-..
O;- 1 atm
0
O
2000
0
4
m
1 atma
0U
A
4000
6000
8000
Real Impedance / kf
Figure S5-5. Nyquist plot of the epitaxial LSM82 thin film as a function of oxygen partial
pressure at 550 OC. EIS data of the LSM82 was found to show the p(0 2 )-dependent impedance
responses, which suggest that the oxygen surface exchange kinetics governs the oxygen
electrocatalysis on the film surface.9
155
Sr (c)
La (b)
(a)
6o (d)
Mn
6
/-LC.
C
-
560
600
0.
.3nm
--
nm
3.: nm
10 nm
620
-
640
1612
-LSC
--3.5
.3
-0.
3
n
nm
10 rm
1638
1664
740
0.3 nm
-3'
0.9 nm
nm
10 nm
760
0.3 nm
0. nm
nm
\
J
10 11m
760
600
560
66
616
Kinetic Energy / eV
Figure S5-6. AES data using area mode of the epitaxial LSC82 with and without LSM82
coverage annealed at 550 'C in an oxygen partial pressure of 1 atm. (a) LaMNN cation variation
(RSF: 0.059), (b) SrLMM cation variation (RSF: 0.027), (c) COLMM cation variation (RSF: 0.076),
and (d) MnLMM cation variation (RSF: 0.161) as a function of LSM82 coverage.
156
(a)
(b)
Figure S5-7. The bulk supercells used to calculate the Sr-La swapping energy of La1.xSrxCoO3-8
and Lai.xSrxMnO3-6 (a) x=0.25 and (b) x=0.375. The dark blue octahedra, light blue, green and
red spheres represent Co/Mn centered octahedra, Sr, La, and 0 ions, respectively.
157
550
*C
atm :
P()2=
LSM
0.9 nm
sell
I'e
010,
:
I
1
I
r]
I
LSC
S
10
-
0
a
10
-
rSf
I
a
a
a
20
30
40
$I
50
60
70
Annealing Time / hrs
Figure S5-8. Time dependent surface exchange coefficient (k) of bare LSC11 3 (S-black),
LSC11 3 with LSM113 coverage -0.9 nm (S-blue), and LSC113 with LSC2 14 coverage ~2.5 nm (0
-yellow).
158
5.6 References
1.
Z. P. Shao and S. M. Haile, Nature, 2004, 431, 170-173.
2.
B. C. H. Steele and A. Heinzel, Nature, 2001, 414, 345-352.
3.
L. Yang, C. D. Zuo, S. Z. Wang, Z. Cheng and M. L. Liu, Adv. Mater., 2008, 20, 32803283.
4.
L. da Conceicao, C. R. B. Silva, N. F. P. Ribeiro and M. Souza, Mater. Charact., 2009,
60, 1417-1423.
5.
R. A. De Souza and J. A. Kilner, Solid State Ion., 1998, 106, 175-187.
6.
R. A. De Souza and J. A. Kilner, Solid State Ion., 1999, 126, 153-161.
7.
A. Endo, H. Fukunaga, C. Wen and K. Yamada, Solid State Ion., 2000, 135, 353-358.
8.
T. Ioroi, T. Hara, Y. Uchimoto, Z. Ogumi and Z. Takehara, J Electrochem. Soc., 1997,
144, 1362-1370.
9.
S. B. Adler, Chem. Rev., 2004, 104, 4791-4843.
10.
S. B. Adler, Solid State Ion., 1998, 111, 125-134.
11.
A. V. Berenov, A. Atkinson, J. A. Kilner, E. Bucher and W. Sitte, Solid State Ion., 2010,
181,819-826.
12.
T. Kawada, J. Suzuki, M. Sase, A. Kaimai, K. Yashiro, Y. Nigara, J. Mizusaki, K.
Kawamura and H. Yugami, J. Electrochem. Soc., 2002, 149, E252-E259.
13.
J. A. Kilner, R. A. DeSouza and I. C. Fullarton, Solid State Ion., 1996, 86-8, 703-709.
14.
J. Mizusaki, Y. Mima, S. Yamauchi, K. Fueki and H. Tagawa, J. Solid State Chem.,
1989, 80, 102-111.
15.
V. G. Sathe, A. V. Pimpale, V. Siruguri and S. K. Paranjpe, Journal of PhysicsCondensedMatter, 1996, 8, 3889-3896.
16.
R. H. E. vanDoorn, I. C. Fullarton, R. A. deSouza, J. A. Kilner, H. J. M. Bouwmeester
and A. J. Burggraaf, Solid State Ion., 1997, 96, 1-7.
17.
J. M. Bae and B. C. H. Steele, Solid State Ion., 1998, 106, 247-253.
18.
A. Esquirol, N. P. Brandon, J. A. Kilner and M. Mogensen, J. Electrochem. Soc., 2004,
151, A1847-A1855.
19.
S. Hashimoto, Y. Fukuda, M. Kuhn, K. Sato, K. Yashiro and J. Mizusaki, Solid State
Ion., 2010, 181, 1713-1719.
159
20.
L. W. Tai, M. M. Nasrallah, H. U. Anderson, D. M. Sparlin and S. R. Sehlin, SolidState
Ion., 1995, 76, 259-271.
21.
L. W. Tai, M. M. Nasrallah, H. U. Anderson, D. M. Sparlin and S. R. Sehlin, Solid State
Ion., 1995, 76, 273-283.
22.
J. Chen, F. L. Liang, B. Chi, J. Pu, S. P. Jiang and L. Jian, J. Power Sources, 2009, 194,
275-280.
23.
E. J. Crumlin, S. J. Ahn, D. Lee, E. Mutoro, M. D. Biegalski, H. M. Christen and Y.
Shao-Horn, J Electrochem. Soc., 2012, 159, F219-F225.
24.
E. J. Crumlin, E. Mutoro, S. J. Ahn, G. J. la 0, D. N. Leonard, A. Borisevich, M. D.
Biegalski, H. M. Christen and Y. Shao-Horn, J. Phys. Chem. Lett., 2010, 1, 3149-3155.
25.
H. P. Ding, A. V. Virkar, M. L. Liu and F. Liu, Phys. Chem. Chem. Phys., 2013, 15, 489496.
26.
M. F. Liu, D. Ding, K. Blinn, X. X. Li, L. F. Nie and M. Liu, Int. J. Hydrog. Energy,
2012, 37, 8613-8620.
27.
Z. Liu, M. F. Liu, L. Yang and M. L. Liu, J. Energy Chem., 2013, 22, 555-559.
28.
X. Y. Lou, S. Z. Wang, Z. Liu, L. Yang and M. L. Liu, Solid State Ion., 2009, 180, 12851289.
29.
M. E. Lynch, L. Yang, W. T. Qin, J. J. Choi, M. F. Liu, K. Blinn and M. L. Liu, Energy
Environ. Sci., 2011, 4, 2249-2258.
30.
M. Muranaka, K. Sasaki, A. Suzuki and T. Terai, J. Electrochem. Soc., 2008, 155, B860B863.
31.
E. Mutoro, E. J. Crumlin, M. D. Biegalski, H. M. Christen and Y. Shao-Horn, Energy
Environ. Sci., 2011, 4, 3689-3696.
32.
L. F. Nie, M. F. Liu, Y. J. Zhang and M. L. Liu, J. Power Sources, 2010, 195, 4704-4708.
33.
Y. Sakito, A. Hirano, N. Imanishi, Y. Takeda, 0. Yamamoto and Y. Liu, J. Power
Sources, 2008, 182, 476-481.
34.
M. Sase, F. Hermes, K. Yashiro, K. Sato, J. Mizusaki, T. Kawada, N. Sakai and H.
Yokokawa, J. Electrochem. Soc., 2008, 155, B793-B797.
35.
K. Yashiro, T. Nakamura, M. Sase, F. Hermes, K. Sato, T. Kawada and J. Mizusaki,
Electrochem. Solid State Lett., 2009, 12, B135-B137.
160
36.
X. B. Zhu, D. Ding, Y.
Q. Li, Z. Lu, W.
H. Su and L. Zhen, Int. J. Hydrog. Energy, 2013,
38, 5375-5382.
37.
E. J. Crumlin, E. Mutoro, W. T. Hong, M. D. Biegalski, H. M. Christen, Z. Liu, H.
Bluhm and Y. Shao-Horn, The Journal of Physical Chemistry C, 2013, 117, 1608716094.
38.
E. J. Crumlin, E. Mutoro, Z. Liu, M. E. Grass, M. D. Biegalski, Y. L. Lee, D. Morgan, H.
M. Christen, H. Bluhm and Y. Shao-Horn, Energy Environ. Sci., 2012, 5, 6081-6088.
39.
Z. Feng, E. J. Crumlin, W. T. Hong, D. Lee, E. Mutoro, M. D. Biegalski, H. Zhou, H.
Bluhm, H. M. Christen and Y. Shao-Horn, The Journalof Physical Chemistry Letters,
2013, 4, 1512-1518.
40.
Z. Feng, Y. Yacoby, M. J. Gadre, Y. L. Lee, W. T. Hong, H. Zhou, M. D. Biegalski, H.
M. Christen, S. B. Adler, D. Morgan and Y. Shao-Horn, J Phys. Chem. Lett., 2014,
Submitted.
41.
Z. Feng, Y. Yacoby, W. T. Hong, H. Zhou, M. D. Biegalski, H. M. Christen and Y. ShaoHorn, Energy Environ. Sci., 2014, 7, 1166-1174.
42.
A. Kumar, D. Leonard, S. Jesse, F. Ciucci, E. A. Eliseev, A. N. Morozovska, M. D.
Biegalski, H. M. Christen, A. Tselev, E. Mutoro, E. J. Crumlin, D. Morgan, Y. ShaoHorn, A. Borisevich and S. V. Kalinin, ACS nano, 2013, 7, 3808-3814.
43.
G. J. la 0, S. J. Ahn, E. Crumlin, Y. Orikasa, M. D. Biegalski, H. M. Christen and Y.
Shao-Horn, Angew. Chem.-Int. Edit., 2010, 49, 5344-5347.
44.
D. N. Leonard, A. Kumar, S. Jesse, M. D. Biegalski, H. M. Christen, E. Mutoro, E. J.
Crumlin, Y. Shao-Horn, S. V. Kalinin and A. Y. Borisevich, Adv. Energy Mater., 2013,
3, 788-797.
45.
M. Kubicek, A. Limbeck, T. Fromling, H. Hutter and J. Fleig, J. Electrochem. Soc., 2011,
158, B727-B734.
46.
D. Oh, D. Gostovic and E. D. Wachsman, J. Mater. Res., 2012, 27, 1992-1999.
47.
A. Mitterdorfer and L. J. Gauckler, Solid State Ion., 1998, 111, 185-218.
48.
D. Lee, A. Grimaud, E. J. Crumlin, K. Mezghani, M. A. Habib, Z. Feng, W. T. Hong, M.
D. Biegalski, H. M. Christen and Y. Shao-Horn, The Journal of Physical Chemistry C,
feature article, 2013, 117, 18789-18795.
161
49.
D. Lee, Y.-L. Lee, A. J. L. Grimaud, W. Hong, M. Biegalski, D. Morgan and Y. ShaoHorn, J Mater. Chem. A, 2014.
50.
J. Maier, Physical Chemistry of Ionic Materials:Ions and Electrons in Solids John Wiley,
Chichester, England; Hoboken, NJ, 2004.
51.
M. Mori, Y. Hiei, N. M. Sammes and G. A. Tompsett, J. Electrochem. Soc., 2000, 147,
1295-1302.
52.
R. H. E. van Doom and A. J. Burggraaf, Solid State Ion., 2000, 128, 65-78.
53.
T. Ishihara, T. Kudo, H. Matsuda and Y. Takita, J. Electrochem. Soc., 1995, 142, 15191524.
54.
Z. Feng, E. Crumlin, W. Hong, D. Lee, E. Mutoro, M. Biegalski, H. Zhou, H. Bluhm, H.
Christen and Y. Shao-Hom, J. Phys. Chem. Lett., 2013, 4, 1512-1518.
55.
S. P. Jiang, J. Solid State Electrochem., 2007, 11, 93-102.
56.
W. Wang and S. P. Jiang, Solid State Ion., 2006, 177, 1361-1369.
57.
H. M. Reichenbach and P. J. McGinn, Appl. Catal. A-Gen., 2003, 244, 101-114.
58.
L. Karvonen, S. H. Yoon, P. Hug, H. Yamauchi, A. Weidenkaff and M. Karppinen,
Mater. Res. Bull., 2011, 46, 1340-1345.
59.
A. N. Petrov, V. A. Cherepanov, 0. F. Kononchuk and L. Y. Gavrilova, J. Solid State
Chem., 1990, 87, 69-76.
60.
E. Fischer, W. D. Shen and J. L. Hertz, J Electroceram.,2012, 29, 262-269.
61.
T. Horita, K. Yamaji, N. Sakai, H. Yokokawa, A. Weber and E. Ivers-Tiffee,
ElectrochimicaActa, 2001, 46, 1837-1845.
62.
J. Hwang, H. Lee, K. J. Yoon, H. W. Lee, B. K. Kim, J. H. Lee and J. W. Son, J.
Electrochem. Soc., 2012, 159, F639-F643.
63.
Y. L. Lee, J. Kleis, J. Rossmeisl, Y. Shao-Hom and D. Morgan, Energy Environ. Sci.,
2011, 4, 3966-3970.
64.
M. J. Gadre, Y. L. Lee and D. Morgan, Physical chemistry chemical physics : PCCP,
2012, 14, 2606-2616.
65.
B. Matovic, S. Boskovic, L. Zivkovic, M. Vlajic and V. Krstic, in CurrentResearch in
Advanced Materials and Processes, eds. D. P. Uskokovic, S. K. Milonjic and D. I.
Rakovic, Trans Tech Publications Ltd, Zurich-Uetikon, 2005, pp. 175-179.
162
66.
M. Yashima, S. Sasaki, M. Kakihana, Y. Yamaguchi, H. Arashi and M. Yoshimura, Acta
Crystallogr. Sect. B-Struct. Commun., 1994, 50, 663-672.
67.
M. Pechini, U.S. PatentNo. 3,330,697, 1967.
68.
H. M. Christen, E. D. Specht, S. S. Silliman and K. S. Harshavardhan, Phys. Rev. B,
2003, 68, 4.
69.
S. B. Adler, J. A. Lane and B. C. H. Steele, J. Electrochem. Soc., 1996, 143, 3554-3564.
70.
Y. L. Yang, C. L. Chen, S. Y. Chen, C. W. Chu and A. J. Jacobson, J Electrochem. Soc.,
2000, 147, 4001-4007.
71.
P. S. Manning, J. D. Sirman, R. A. DeSouza and J. A. Kilner, Solid State Ion., 1997, 100,
1-10.
72.
J. Maier, Solid State Ionics, 1998, 112, 197-228.
73.
J. Fleig and J. Maier, J. Eur. Ceram. Soc., 2004, 24, 1343-1347.
74.
J. Fleig, H. R. Kim, J. Jamnik and J. Maier, Fuel Cells, 2008, 8, 330-337.
75.
J. Fleig, Solid State Ion., 2002, 150, 181-193.
76.
G. Kresse and J. Hafner, PhysicalReview B, 1993, 47, 558.
77.
G. Kresse and J. Furthmuller, Phys. Rev. B, 1996, 54, 11169-11186.
78.
P. E. Blochl, Phys. Rev. B, 1994, 50, 17953-17979.
79.
J. P. Perdew and Y. Wang, Physical Review B, 1992, 45, 13244.
80.
V. I. Anisimov, F. Aryasetiawan and A. I. Lichtenstein, Journalof Physics-Condensed
Matter, 1997, 9, 767-808.
81.
S. L. Dudarev, G. A. Botton, S. Y. Savrasov, C. J. Humphreys and A. P. Sutton, Phys.
Rev. B, 1998, 57, 1505-1509.
82.
Y.-L. Lee, J. Kleis, J. Rossmeisl and D. Morgan, PhysicalReview B (CondensedMatter
andMaterialsPhysics), 2009, 80, 224101.
163
Thispage is intentionallyleft blank
164
Chapter 6.
The effect of (La,Sr)2CoO 4s/Lao. 8 Sro. 2 MO 3-8 (M =
Co, Fe) Hetero-interface on Oxygen
Electro-
catalsis at Elevated Temperatures
10
2
Vol 0LSC
24
0
1LSC 113
E 10
0101
10,
~LSIC 11 3
Sm
0.4
0.2
0.0
Relative bulk 0 2p band vs base film I eV
Reproduced in part with permission from Dongkyu Lee, Yueh-Lin Lee, Wesley T. Hong,
Michael D. Biegalski, Dane Morgan and Yang Shao-Horn, The effect of (LaSr)2CoO4 :k/
Lao.8Sro.2MO3 -, (M = Co, Fe) Hetero-interface on Oxygen Electrocatalsis at Elevated
Temperatures, In preparation
165
6.1 Introduction
The majority of efficiency loss in solid oxide fuel cells (SOFCs) results from slow
oxygen reduction reaction (ORR) kinetics at the cathode,1- 3 where lanthanum strontium
manganite (LSM) 4 -8 are currently utilized as a cathode material at high temperatures such as
1000 *C. Therefore, there is a need to search for electrode materials with high catalytic activity,
particularly for intermediate temperature (500 - 700 *C) SOFCs. Mixed ionic and electronic
20
9 15
conductors (MIECs) such as Lai.SrxCoO3-8 (LSC) - and La.xSrxCol.yFyO3-8 (LSCF)16perovskite oxides have been intensively studied to promote the oxygen surface exchange kinetics
at intermediate temperatures.
Recently, oxide heterostructure interfaces, the combination of a Ruddlesden-Popper (RP)
(Lao. 5 Sro.5 ) 2 CoO 4 +8 (LSC2 1 4 ) layer on top of the perovskite LSC, have shown remarkably high
oxygen surface exchange kinetics. 2 1-27 Yashiro et al. have reported -1
order of magnitude
enhancement in activity for the composite cathode screen-printed with the mixture of
Lao. 6Sro.4CoO 3- and LSC2 1 4.2 7 Sase et al. have also reported -3 orders of magnitude higher
oxygen surface exchange coefficient (k*) at the interfacial region between polycrystalline
Lao.6 Sro.4 CoO 3-8 and LSC21 4 compared to their bulk values. 2 4 More recently, we have shown
using well-defined thin film systems that LSC214 coverage on epitaxial (001)-oriented
Lao8 Sro.2 CoO 3-8 (LSCii 3 ) thin film surfaces can greatly enhance the surface exchange kinetics up
to -2 orders of magnitude.7
A number of studies have been focused on understanding the high catalytic activity of
oxide heterostructure interfaces, using well-defined thin film systems. 2 8-3 For example, Density
functional theory (DFT) calculations have shown that there is a large thermodynamic driving
force for exchanging a Sr from Lao. 7 5 Sro.25CoO3.-
with a La from LSC2 1 4 , where enhanced
oxygen kinetics result from increased vacancy concentration near the interface. 3 0 Using Coherent
Brag Rod Analysis (COBRA), we have shown the markedly enhanced Sr concentration at the
interface of LSC 1 3 and LSC21 4 and near the surface of LSC2 1 4 , proposing that the increased Sr
content at the interface may contribute the enhanced catalytic activity resulting in higher oxygen
vacancy concentration. 2 9
Similar to the LSC surface modification with LSC2 1 4 , several studies have also reported
the enhanced
surface electrocatalytic
activity of porous LSCF cathodes with surface
decoration.3-3 8 Depositing thin Lao. 85Sro. 15MnO 3 .3 coatings on porous Lao.6 Sro.4 Coo.2Feo.O3-8
166
(LSCF11 3) electrodes using an infiltration process, Lynch et al.36 have shown the enhanced
surface electrocatalytic activity of decorated LSCF113 cathodes upon polarization. A uniform
coating of Smo.5 Sro.5CoO 3 -2
and Lao.4 87 5Cao.oi2 5Ceo. 50 2 -
4
through infiltration on porous
LSCF113 cathodes have also shown the reduced polarization resistance of the cathode. However,
vast majority of research has been performed on porous LSCF11 3 electrodes, which lead to
ambiguous structure and geometry, and therefore the physical origin responsible for enhanced
cathodic performance associated with surface decoration of perovskites is not yet completely
understood. In addition, further investigation of other potential surface modification materials is
required to design highly active and stable cathodes for SOFCs. In this study, we are particularly
interested in addressing if and how LSC 214 decoration on the LSCF11 3 surface can lead to any
enhancement of the surface exchange kinetics using epitaxial thin films as a model system.
We here discuss the influence of LSC 2 14 surface decoration on the oxygen surface
exchange kinetics and the time-dependent surface stability of epitaxial LSCF113 thin films,
comparing with an LSC214 -decroated LSC11 3 heterostructure thin film. We find that the LSC214
decoration can enhance the surface exchange coefficient (k) of the LSCF11 3 thin film only -2
times while it can significantly enhance the k of the LSC113 thin film by -2 orders of magnitude.
Furthermore, the LSC 214-decorated LSC11 3 shows a dramatically improved the time-dependent
degradation of the surface activity in contrast to the LSC214-decorated LSCF11 3 thin film. Auger
electron spectroscopy (AES) reveals markedly different surface Sr concentrations between the
LSC 214-decorated LSCF 1 3 and LSC214 -decorated LSC113 thin films, which is supported by the
distinct thermodynamic driving force for cation interdiffusion across the heterointerfaces of
LSC 2 14 -LSCF1 1 3 vs. LSC 2 14 -LSCil 3 in the DFT modeling. In addition, ab initio surface stability
analysis is also performed to demonstrate different Sr concentrations near the top (001) surface
between LSCF11 3 and LSC 1 3. The different LSC 214 decoration effects on the surface exchange
kinetics of the LSCF11 3 and LSC11 3 thin films can thus be attributed to different surface Sr
concentrations resulted from different thermodynamic driving force and different capability of
suppressing formation of the surface secondary phases, which can also lead to the difference in
the time-dependent surface stability.
6.2 Experimental Methods
167
Film deposition
Pulsed laser deposition (PLD) was utilized to deposit the (001)-oriented
epitaxial LSCF11 3 thin films (-62.5 nm) on YSZ with a -5 nm gadolinium-doped ceria (GDC, 20
mol % Gd) as the buffer layer to prevent the formation of La 2 Zr2 07-.
LSC 2 14
(-0.26, -0.78,
-2.6,
and -5
39
Varying thicknesses of
nm) were subsequently deposited on top of the
LSCFil 3/GDC/YSZ. The epitaxial LSC11 3 thin films (-85 nm) with and without LSC 214
decoration (-2.6 nm) were also prepared on YSZ(001) with a GDC buffer layer using PLD.
Details for the LSCF11 3, LSC 2 14 , LSC11 3 , and GDC PLD target syntheses and PLD deposition
process can be found in the Supporting Information (SI).
Oxide phase purity and orientation of the
High resolution X-ray diffraction (HRXRD)
thin film systems were investigated via high resolution X-ray diffraction (HRXRD) using a fourcircle diffractometer (PANalytical, USA and Bruker D8, Germany). Measurements were
performed in normal and off-normal configurations. The in-plane lattice parameters (a lattice
parameter) of LSCF 113 and LSC11 3 was determined from the off-normal (202)p peak position
(where "pc" denotes the pseudocubic notation) and the c lattice parameter of LSCF11 3 and
LSC1 1 3 normal to the film surface was determined from the (002)p peak position. Surface
morphology was examined by optical microscopy (Carl Zeiss, Germany) and atomic force
microscopy (AFM) (Veeco, USA). AFM images of as-deposited LSCF11 3 and LSC214-decorated
LSCFj1 3 films revealed that the surfaces were smooth with the root-mean-square (RMS)
roughness values of 0.237 - 0.323 nm, as shown in Figure S6-1. Indeed, those of as-deposited
LSC1
3
and LSC 214-decorated LSC11 3 films also showed vey smooth surfaces with the RMS
roughness values of 0.769 - 1.133 nm, as shown in Figure 6-1. The RMS roughness was
comparable across all surfaces.
Electrochemical impedance spectroscopy (EIS)
Electrochemical impedance spectro-
scopy (EIS) measurements of microelectrodes -200 Pm in diameter were performed using a
microprobe station (Karl Snss, Germany) connected to a frequency response analyzer (Solartron
1260, USA) and dielectric interface (Solartron 1296, USA). Temperature was controlled at
550 *C with heating stage (Linkam TS1500, UK) and data were collected between 1 MHz to 1
mHz using a voltage amplitude of 10 mV. EIS testing temperature was calibrated with a
168
thermocouple contacting the thin film surface and deviation of
5 *C was observed.
EIS
experiments were completed between p(O2) of 10-3 atm and 1 atm. EIS data were analyzed using
an equivalent circuit (Figure S6-4b), from which the ORR resistance (RoRR) and surface oxygen
exchange rate were obtained. EIS data of all samples used in this study were found to
predominantly exhibit a semicircle in the Nyquist plot. The electrical oxygen surface exchange
coefficient (k) was calculated from the resistance of the low-frequency semicircle
k
RT
-
,21,40-45
(RLF)
using
where R is the universal gas constant (8.314 J molV'-K7'), T is the
4F2RF electrode
absolute temperature (823 K), Fis Faraday's constant (96,500 C-mol1'), Aelectrode is the area of the
microelectrode, and co is the lattice oxygen concentration in LSCF11 3 and LSC113. Details about
the EIS testing procedure, data analysis, and co estimation can be found in the SI.
Auger electron spectroscopy (AES)
Auger electron spectroscopy (AES) was conducted
using a Physical Electronics 700 Scanning Auger Nanoprobe (PHI, USA) operating at an
accelerating voltage of 10 kV to analyze the surface chemistry change of the LSC 214-decorated
LSCF 1 3 and the LSC214-derocated LSC113 films after heat treatment. The films were annealed at
550 *C for 6 hours in an oxygen partial pressure of 1 atm before AES data were collected. The
AES data were collected using two different modes: area mode (three different 10 pm x 10 pm
regions selected across a sample) and point mode (two different -0.45 pm diameter spots
selected on a sample) in an ultra-high vacuum chamber. Elemental quantification of AES spectra
utilized relative sensitivity factors (RSFs) of 0.059, 0.027, 0.076, 0.178, and 0.212 for LamN,
SrLMM, COLMM, FeLMM, and OKLL, respectively, as supplied by the AES manufacturer (Physical
Electronics). Details about AES measurement and analysis can be found in the SI.
Density functional theory (DFT)
Spin polarized Density Functional Theory (DFT)
calculations were preformed with the Vienna Ab-initio Simulation Package4 6 '
47
using the
Projector-Augmented plane-Wave method4" with a cutoff of 450 eV. Exchange-correlation was
treated in the Perdew-Wang-9 149 Generalized Gradient Approximation (GGA). The GGA+U
calculations
50
are performed with the simplified spherically averaged approach5 l, where the Uff
(Uff= Coulomb U - exchange J) is applied to d electrons. (Uef(Fe) = 4.0 eV and Ueff(Co) = 3.3
eV) 3 0, 5. All calculations are performed in the ferromagnetic state in order to use a consistent and
169
tractable set of magnetic structures, and the spin states for the calculated LSC 1 3 , LSC 2 14 , and
LSCF 113 systems are: Fe: high-spin and Co: intermediate-spin.
Calculations for SrLa substitution energies in bulk LSC113 and LSCF 113 are simulated
using
a
2ax2a,,,x2a,
pseudocubic
apv(Lao. 75Sro.2 5CoO 3 )= 3.88
A,
supercell
structure
of
LaO.7 5Sro.2 5CoO 3
(with
where a, is the relaxed GGA+U perovskite lattice constant) and
Lao. 62 5 Sro.37 5Feo. 75Coo.2 5 0 3 (ap (LSCF113) = 3.91 A) with 2x2x2 k-point mesh and 450 eV planewave energy cut-off. SrLa substitution energy in bulk LSC 214 is simulated using a 2 arpx2 apxcrp
supercell structure of LSC214 (where a,,p and crp are the relaxed GGA+U Ruddlesden-Popper
phase lattice constant: arp(LSC 2 14 )
=
3.86
A,
Crp(LSC 2 14)
=
12.50
A).
The supercell
configurations are illustrated in Figure S6-8 of the Supporting Information. The SrLa substitution
energy for Lao. 75SrO. 2 5CoO 3 (Lao. 62 5 Sro.3 7 5Feo. 75Coo.25 03 ) bulk was taken as the difference in
energies between a Lao.62 5Sro. 3 75CoO 3 (Lao.5 Sro. 5Feo.75Coo. 25 0 3 ) bulk and a Lao. 7 5Sro.2 5 CoO
3
(Lao.6 2 sSro. 3 7 5Feo.7 5Coo.2 5 0 3 ) bulk. Similarly, the SrLa substitution energy for LSC 2 14 was
calculated using the total energy difference between LSC 2 14 and (Lao.4 3 75Sro.562 5) 2 CoO4).
The LSC214 -LSCF11 3 heterointerface was simulated with a periodic 176-atom supercell
(2 an3x2 an3 supercell in the x-y plane (a113 = ap(LSCFii 3) =3.91
Lao.62 5Sro.3 7 5Feo.75Coo.2 50
A).
3
A)
with 12-layers of
and 6-layers of LSC 2 14 along z where the relaxed crp(LSC214)=12.42
The LSC214-LSCF1 1 3 heterointerface calculations and structural model were reported
previously and the simulated LSC 214 surface slab model. Three different Sr/La and Co/Fe
arrangements in the LSC214 -LSCF,1 3 interface model are investigated, as shown in Figure S6-8
of the Supporting Information.
Ab initio thermodynamic analysis for-LSC,1
3
and LSCF113 surface stability are simulated
using the 9-layer 2x2 symmetric (001) AO terminated and B0 2 terminated slabs with the central
5 layers fixed to a composition close to the bulk LSC11 3 and LSCF113 . The La/Sr (and Fe/Co for
LSCF113) content of the top two and bottom two layers are varied, as illustrated in Figure S6-11.
More details of the ab initio thermodynamic analysis approaches are provided in the Supporting
Information.
6.3 Results and Discussion
170
Structural relationship of epitaxial LSCF113 thin films with LSC21 decoration
Normal XRD data (Figure 6-1) of the undecorated LSCF 1 3 and LSC 214-decorated
LSCF 1 3 films clearly show the presence of the (00l)pc (I is integer) peaks of LSCF113 and
(00l)cubic
(1 is even) peaks of GDC and YSZ, indicating that the LSCF11 3 film grew epitaxially
with the following epitaxial relationships: (00l),cLSCF 1 3 // (00l)cubciGDC //(00l)cubicYSZ. With
LSC 2 14 coverage equal to ~5 nm in thickness, the (00t)tea. (I is integer) peaks of LSC214 become
visible, which represents (00l)tetra.LSC 2 14 // (00l)pcLSCF113 // (001)cubicGDC // (00l)cubciYSZ.
The subscript "tetra." denotes the tetragonal notation. Off-normal phi-scan analysis of the
undecorated LSCF11 3 and LSC 214-decorated LSCF11 3 films shows that LSC2 1 4 {10 3 }tetra., LSC82
{101 },, GDC {2 0 2 }cubic and YSZ {2 02 }cubic have strong peaks with 4-fold cubic symmetry
(Figure 6-1b), which reveals the in-plane crystallographic relationships between GDC and YSZ
(a cube-on-cube alignment), LSCF 11 3 and GDC (an in-plane 450 rotation with [100]pCLSCFiu I/
[1 10]cubicGDC // [l0]cubicYSZ), and LSCF 1 3 and LSC 214 (no rotation with [100]pLSCF11 3 //
[100]tetra.LSC 2 14 ), as shown in Figure 6-1c. The undecorated LSC11 3 and LSC 2 14-derocated
LSC1 1 3 thin films also had the same crystallographic relationships as shown in Figure S6-2.
Similar to our previous studies, 2 1,2, 4 0 the relaxed lattice parameters,
a of the
epitaxial LSCF113
films with and without LSC 214 surface decoration in this study at room temperature did not
change significantly with different LSC 214 decoration thicknesses, ranging from 3.902 - 3.906
A
(Figure S6-3a, Supporting Information). The a of the epitaxial LSC11 3 films with and without
LSC 2 14 surface decoration was also found to be no change significantly, having 3.838 - 3.839 A
(Figure S6-3a, Supporting Information). As shown in Figure S3b, Supporting Information, both
in-plane and out-of-plane strains of LSCF11 3 and LSC113 films were not strongly influenced by
the LSC 214 coverage, which is supported by the fact that the lattice constant of LSC 214 (atvra. ~
3.81 A for LSC 214 bulk53 ) is very close to that of LSCF113 (ateta. ~ 3.89 A for LSCF11 3 bulk") and
LSC11 3 (atetra. ~ 3.85
A for
LSC11 3 bulk 54). This can be further supported by our recent work, 28
where the LSC 2 14 decoration has no influence on the in-plane and out-of-plane strains of the
epitaxial LSC11 3 films at elevated temperatures. Details about lattice parameter calculation and
HRXRD of LSC 214 -decorated LSC11 3 film can be found in the SI.
171
a)
b)
-.
00U
1
0
N00
0
45
LSC214 (10 3 )tetra
C)
La/Sr
0
0U
0UU
Co
0
N
I
U
0
5 nm-
LSC214
LS
(202)p
00
LSCZ14 ~2.6 nm
0.78
LSC214
Co/Fe
GDC (202),
nm
0)
-
450
Ce/Gd
LC21
0
-j
-
0.26 nm
0)
0j
YSZ (202).
LSCF - ref.
20
40
60
80
100
-180
-90
0
90
180
0/I0
20/ 0
Figure 6-1. X-ray diffraction (Cu Ka) analysis at room temperature. (a) Normal XRD of the
epitaxial LSCF 1 3 reference and the LSC 214-decorated LSCF11 3 films, (b) off-normal XRD of a
similarly prepared sample with a thicker (-5 nm) LSC 2 14 coverage, and (c) schematic of the
rotational
relationships
among
the
LSC 2 14 (001)tetra.,
LSCF 1 3(001)p
,
crystallographic
GDC(001)cubic, and YSZ(001)cubic.
Oxygen surface exchange kinetics of epitaxial LSCF113 thin films with LSC 21 4 decoration
EIS data collected from the undecorated LSCF 1 3 and LSC 214-decorated LSCF113 films at
550 *C with an oxygen partial pressure of 1 atm are shown in Figure 6-2a. The real impedance of
the predominant semicircle decreased slightly with LSC 214 coverage. In addition, the
predominant semicircle was found to increase with decreasing oxygen partial pressure, where
EIS data of all samples used in this study were found to show nearly perfect semicircle
impedances. 55 Representative EIS data collected from the LSCF 1 3 film with -5 nm LSC 214
coverage measured at 550 'C as a function of p(O2) are shown in Figure 6-2b. Considering the
fact that the film thicknesses are much smaller than the critical thickness (estimated to 3.28 gm
for bulk LSCF 1 3 at 550 0C 56), the p(0 2)-dependent impedance responses suggest that the oxygen
172
surface exchange kinetics governs the oxygen electrocatalysis on the film surface. Similar to
LSCF11 3 thin films, LSC2 14 -decorated LSC11 3 thin film, where the film thicknesses are much
smaller than the critical thickness (estimated to 1
sm
for bulk LSC113 at 550 *C6 ), also showed
the p(02)-dependent impedance responses (Figure S6-5, Supporting Information), suggesting
that the oxygen surface exchange kinetics governs the oxygen electrocatalysis on the film
surface. The k of the undecorated LSCF113 was found to be higher than that of the undecorated
LSC1 1 3, which is in good agreement with our previous study" where the bulk oxygen p-band
center of LSCF113 is higher than that of LSC113 , resulting in higher surface exchange rate. It
should be noted that LSC21 4 coverage led to only 1 - 2 times enhancement of the k values of the
LSCF11 3 thin films while the kq of LSC11 3 with LSC21 4 coverage was found to be nearly 2 orders
of magnitude higher than that of the undecorated LSC11 3. Recently, Han et al. have proposed that
both the anisotropy of LSC21 4 and the lattice strain near the interface between LSC21 4 and LSCi1 3
are responsible for the significantly enhanced surface exchange kinetics of heterostructred oxide
interface.3 1 As shown in Figure S6-3, Supporting Information, however, LSC2
4
decoration had
no influence on the strains of either LSCF113 or LSC1 13. In addition, the same thickness (-2.6
nm) of LSC21 4 decoration resulted in different enhancement of the surface exchange kinetics of
LSCF 1 3 and LSC214 . Therefore, our observation cannot be explained by the anisotropy of LSC21 4
and the strains at the interface. The volume specific capacitances (VSCs), corresponding to the
change in the oxygen nonstoichiometry (6) induced by the change in the electrical potential, of
the entire LSCFl 3 and LSC 1 3 film did not change significantly with LSC21 4 surface coverage, as
shown in Figure 6-2d. This is in good agreement with our previous results2 ' where LSC21 4
surface coverage has no influence on the oxygen vacancy concentration of LSC113 thin films.
Moreover, the VSCs of LSC11 3 with and without LSC2 14 decoration in this study were
comparable to those of epitaxial LSC113 thin films reported previously.41 This indicates that the
oxygen content in the LSC113 and LSCF11 3 films with and without LSC214 coverage do not
contribute to the modification of the k" values observed. Details of VSCs are provided in the SI.
173
4000 U
0
LSCF - 62.4 nm
LSC21 4 - 0.26 nm
LSC214 - 0.78 nm
2.6 nm
--2.nm
LSCa
LS
150 -a,
.40.
'-0-LSC 214 - 5nm
10atm
0 2000
/ 0\
-
E
O-
M 50 I
0)
100
150
200
0
10
0am
1
.atm.
n Y Ll
0
1000 2000
0a) E
o
o
10o-7 I
o
o
\LSC214 5nm
C1000
0
50
Ria ha~wfdnce I k
10. tm
00
D
00-
~
E
0
25 0
.SCF - 62.4 n n
LSC214 - 0.26 nm
LSC214 - 0.78 nm
LSC 214 - 2.6 nm
5.0
0
0
C
o
ILSC,
LSC
21
0
113
C
0
LSC 214 -5nm
LSCF bulk
b)
3000
4000
5000
E
0
LSCF - 62.4 ni
LSC-214 0.26 nm
LSC - 0.78 nm
LSC 2 14 - 2.6 nm
550c
5m
LSC214
-
U) 10
0
I
Real Impedance / kf
Real Impedance / k
10
.
0
0 100 I
0-
0
at
3000
2 14
E
1
.
-0-
-I
550 OC
550 0C
p(0 2) =1 atm
0
E
0
LSC bulk
10
10
LSC 21/LSCI1
Ca
b
3
LSC.
LSC113
1010
.
a a
c)
10 I
-3
-2
-1
log p(O 2) / atm
1
102
0
-3
-2
pd)
-1
0
log p(O 2) / atm
Figure 6-2. Electrochemical impedance spectroscopy (EIS) results for the bare LSCF1 1 3 film (0
-black), LSC11 3 film (0-gray), LSCF 1 3 films with -0.3 (0-red), -0.8 (0-orange), -2.6 (0yellow), and -5 nm (0-green) LSC2 14 decorations, LSCi3 film with -2.6 nm (0-blue) LSC 2 14
-
decoration at 550 *C. (a) Nyquist plot of the epitaxial LSCF 1 3 and the epitaxial LSC2 14
decorated LSCF,1 3 films in 1 atm. (b) Nyquist plot of the LSCF113 thin film with -5 nm LSC214
coverage as a function of p(O2). Inset shows a magnification of the Nyquist plot in 1 atm. (c)
,
Oxygen partial pressure dependency of the surface exchange coefficients (k) of the LSCF113
LSC1 13 , LSC 214-decorated LSCF113, LSC 214 -decorated LSC 113 films calculated from EIS spectra
collected at 550 *C. Extrapolated bulk k* (approximately equivalent to kq) values at 550 0C
obstained from previous data of (*-light blue) Steele et al.56 and (*-light blue) De souza et al.6
174
are plotted for comparison. (d) Oxygen partial pressure dependency of volume specific
-
capacitance (VSC) of the epitaxial LSCF11 3, LSC11 3 , LSC 214-decorated LSCF113 , and LSC2 14
decorated LSC11 3 films calculated from EIS spectra collected at 550 *C.
Surface chemistry changes of the epitaxial heterostructure thin films
To investigate the change of surface chemistry by LSC 214 decoration, AES was
conducted on the LSCj1 3 and LSCF11 3 films with and without -2.6 nm LSC 214 coverage after
annealing at 550 'C for 6 hours in an oxygen partial pressure of 1 atm, as shown in Figure 6-3.
La Auger signals of the undecorated LSC113 and LSCF11 3 surfaces were found to decrease with
LSC 214 coverage (Figure 6-3a and c) while Sr Auger signals were found to increase (Figure 6-3b
and d). Interestingly, the change in La and Sr Auger signals of LSCF113 film surfaces with and
without LSC 214 coverage is relatively smaller than that of LSC11 3 film surfaces with and without
LSC 214 coverage, which leads to drastically different amount of surface Sr concentration between
LSC 2 1 4-decorated LSC11 3 and LSC 214-decorated LSCF11 3 as shown in Figure 6-3f. Our recent
study 2 9 using Coherent Brag Rod Analysis (COBRA) has revealed the atomic structure and
concentrations of the LSC 214-decorated LSC113 thin films on a SrTiO 3 (STO) substrate, where the
interface of LSC113 and LSC 214 and near the surface of LSC 214 show markedly enhanced
strontium (Sr) concentration and it has been proposed that the increased Sr content at the
interface may be responsible for the enhanced catalytic activity resulting in higher oxygen
vacancy concentration. Therefore, higher Sr concentration on the LSC214 -decorated LSC113 film
surface may lead to higher oxygen vacancy concentration, resulting in faster surface exchange
kinetics compared to the LSC 214-decorated LSCF1 13. This hypothesis is in agreement with
previous findings that greater Sr concentration on the A-site of the perovskite structure leads to
increased surface oxygen exchange kinetics in LSC 10 '" and LSCF. 8
175
Li by
a)
Sr
LSC
LSC
LSC113
LS
113
LSC214 LSC, 13
600
550
c .5
> 2.5
0 C,
I atm
LSC214
/LSC
a
620
640
162
164
1660l
Sr
La d
c)
LSCF
LSCF
a
I
LSC2iILSCI 3
Kinetic Energy / eV
C
U
U
0.0
C
S
f) 03.0
LSC113
0 1.5
Sr/La
0
N
cc0.5
I-
0
4
p
600
620
640
1520
1540
Kinetic Energy / eV
Undecorated
LSC or LSCF
1660
LSC214
Decoration
Figure 6-3. Auger electron spectroscopy (AES) data from area mode for the LSCF113 (black),
LSC11 3 (gray), LSCF,1 3 film with -2.6 nm LSC 2 14 decorations (yellow), and LSC113 film with
~2.6 nm LSC2 14 decoration (blue) after annealing. Annealing was performed at 550 C for 8
hours in an oxygen partial pressure of 1 atm. (a) La Auger spectra and (b) Sr Auger spectra of the
LSC11 3 and LSC 2 14-decorated LSC113 films. (c) La Auger spectra and (d) Sr Auger spectra of the
LSCF 1I 3 and LSC 2 14 -decorated LSCF 1
3
films. (f) Normalized La and Sr intensity ratio extracted
from AES of the LSCF11 3 films with and without ~2.6 nm LSC 2 14 decorations (yellow), and
LSC11 3 film with and without -2.6 nm LSC 2 14 decoration (blue). Details of normalization
methods are provided in the Supporting Information.
Results and Discussion for Density Functional Theory Modeling
In the following discussions, ab initio DFT modeling and thermodynamic analysis were
performed to investigate LSC11 3 and LSCF11 3 (001) surface stability with respect to the surface
,
layer Sr content, as well as relative stability for Sr substitution of La (SrLa) in bulk LSC11 3
LSC 2 14 ,
LSCF11 3, and the LSC 214-LSCF,1 3 heterointerfaces, to provide theoretical information
176
for rationalizing the observed surface chemistry changes and distinct surface exchange kinetics
between LSCF11 3 and LSC11 3 upon LSC21 4 decoration.
Ab initio thermodynamic assessment of Sr content for the LSCF113 and LSC11 3 (001)
surfaces
To probe the physical origin of experimentally observed surface chemistry changes, ab
initio thermodynamic analysis is performed to investigate stability of the LSC11 3 and LSCF11 3
unreconstructed (001) surfaces vs. surface layer Sr content at T=823 K and p(0 2) = 1 atm (details
of the ab initio methods and the thermodynamic analysis are provided in the SI). Within the
LSC1 1 3 bulk stability region set by chemical potential boundaries of thermodynamic equilibrium
,
between the LSC11 3 bulk and relevant lower order oxide compounds (such as SrO, La2 0 3 , C030 4
perovskite LaCoO 3, and brownmillerite SrCoO 2 .5), the ab initio thermodynamic analysis predicts
that the most stable LSC113 (001) surface among the investigated (001) surface configurations
(i.e., the AO and B0 2 surfaces with varying Sr content) is the AO surface containing -75% Sr
concentration (Figure S6-12). The predicted surface layer Sr content from the ab initio
thermodynamic analysis can be supported by our recent COBRA study 59, which shows Sr
segregation within the perovskite phase toward the LSCI
3
surface, and the top LSC11 3 surface
layer contains Sr concentration of -60%. It is noted that surface particles formed on the LSCi3
surfaces were also revealed by the COBRA measurements to be in the perovskite phase with the
A sites almost fully occupied by Sr, i.e., SrCoO3-6. However, the COBRA information only
reveals phases that are registered to the substrate while precipitated Sr-rich secondary phases
such as SrOx and Sr(OH) 2 may also coexist in these surface particles, which cannot be measured
by COBRA. Therefore, the surface particles, which may contain both SrCoO 3-8 and Sr-rich
secondary phases, should be distinguished from the perovskite base film surfaces.
A similar thermodynamic assessment is also performed for the LSCF11 3 (001) surfaces,
and the results predict that the most stable LSCF113 (001) surfaces among the investigated
surface configurations (i.e., the (001) AO and B0 2 surfaces with varying Sr/La amd Co/Fe
content) is the fully Sr-enriched (001) AO terminated surfaces, while the most stable surface
Co/Fe content is dependent on the activity of Co/Fe, as shown in Figure S6-13. The predicted
LSCF11 3 (001) surface stability results based on the ab initio thermodynamic analysis are also in
qualitative agreement with a recent Low Energy Ion Scattering study on LSCF 113 pellets6 0 , where
177
it was observed rapid disappearance of the transition metal upon annealing in air for 8 hours at
T= 600 'C and 800 *C with increase of Sr coverage at the LSCF11 3 surfaces.
-
0.0
-0.2
0
4,
a11
-0.4
-0.6
LU
-0.8
.
Cl)
I
P
-1.0
I
LSCF113
LSC113
I
LSC2
b)
4
I
-
+I
0.0
0
4,
-Il
lb.
Cl)
LU
*
4
414
I
I
I
~
I
I
4I
I
ty
II
-0.4
-0.8
SLSC
-1.2
SConfig. 1
214
Config. 2
SConfin. 3
Figure 6-4. (a) The calculated SrLa substitution energies in bulk LSC 2 14 , LSC11 3, and LSCF113
(all relative to that of LSC,1 3, which is set to 0). The error bars shown in the LSCF,1 3 and
LSC 2 14 represent the upper bound and lower bound of the SrLa substitution energies from the
sampled A-site and B-site cation arrangements.
178
-
Ab initio SrLa substitution energies in bulk LSC 214 , LSCu3 , LSCF,13 , and the LSC 2 14
decorated LSCF113 heterointerfaces
To understand the physical origin of different enhancement on the surface Sr content in
the LSCF1 3 and LSC 1 3 films upon LSC 214 decoration, ab initio DFT calculations were
,
performed to investigate energy for Sr substitution of La (SrLa) in the structures of LSC 214
LSC 1 3 , and LSCFi13 , as shown in Figure 6-4a (details of the DFT modeling approaches are
provided in the SI). The calculated SrLa substitution energies in relaxed bulk LSC 2 14 , LSC1 1 3, and
LSCF113 (all relative to that of LSCi3) suggest a weaker thermodynamic driving force (-0.12 eV)
for Sr interdiffusion from LSCF 1
3
to LSC2 14 than from LSC11 3 to LSC 2 14 (-0.7 eV). To further
understand the interfacial effect on the SrLa substitution energies, three LSC 214-LSCFil 3 interface
configurations with different Sr/La and Co/Fe arrangements are investigated (more detailed
information of the interface models is provided in the SI), as shown in Figure 6-4b. It is shown
that the most stable SrLa substitution is located in the first interface layer adjacent to the LSCF1 1 3
region (the interfacial region is labeled with the gray dotted line). Moving from this interface
layer with the most stable
SrLa
substitution toward the LSCF11 3 and LSC 214 regions, the SrLa
substitution energies become destabilized and gradually approach to the bulk LSCF113 and
,
LSC 2 14 values. In contrast to the results of the previous the LSC,1 3-LSC 2 1 4 interface study 30
where
SrLa
substitution becomes monotonically more stabilized moving from the LSC113 toward
the LSC 214 region, the closeness of SrLa substitution energies between the bulk LSCF113 and the
bulk LSC 214 further reveals an additional interfacial stabilization effect on SrLa substitution,
which can be attributed to the electrostatic effect by the interfacial charge introduced by LSCF113
bulk polarity6 1 . We note that such an interfacial charge effect may also occur in the previous
LSC 2 14 -LSCi1 3 interface model, but the interface charge effect contributes in the same manner as
the relative stability of the
SrLa
substitution in the bulk LSC 214 vs. bulk LSC113 , leading to an
overall monotonic stabilization of
across the LSC2 1 4 -LSC 1
3
SrLa
substitution moving from the LSC113 toward the LSC 214
interface. It is also noted that the energies in Figure 6-4b represent
thermodynamic driving forces of SrLa substitution (relative to the bulk LSCn 3) for the simulated
configurations close to the nominal bulk phase composition, rather than the equilibrated
,
interfacial configurations. The as-grown LSCF113 film surfaces may be initially Sr enriched59
and consequently the thermodynamic driving force for
179
SrLa
interdiffusion from the decorated
LSC 2 14
phase to the Sr-enriched LSC 214- LSCF113 interfaces is expected to be reduced as
compared to the energetic results shown in Figure 6-4b.
Combining the ab initio modeling results for the stable LSCF113 and LSC11 3
surfaces and the thermodynamic driving forces for SrLa substitution in bulk LSC113, LSC 2 14, and
LSCF11 3, as well as the LSC2 14 -LSC1 1 3 and LSC2 14 -LSCF11 3 heterointerfaces, here we discuss the
underlying origins that are responsible for the observed surface chemistry change upon the
LSC 214 decoration for the LSC
3
vs. the LSCF11 3 films.
In the case of the LSC113 (001) base film, the stable surface layer within the perovskite
phase may contain Sr concentration at 60% (COBRA59 )- 75% (DFT predictions, this work), and
additional surface particles, where SrCoO 3- 59 may coexist with other inactive Sr-rich passive
phases, are formed above the surfaces. Upon LSC 214 decoration, the decorated LSC214 phase acts
as a Sr sink (based on the thermodynamic driving forces of SrLa substitution from the LSC113 to
interface of LSC 1 3 and LSC 214, to the LSC 214 phase and surface 2 9, 30) and the perovskite phase
stabilizer 25 . Therefore, both the undecorated LSC113 film surface and the surface of the LSC 214
decorated phase can allow more Sr content to be placed within the perovskite phase and the
decorated LSC 214, as also revealed by the COBRA measurement 29 and supported by the DFT
2 9 30
LSC2 14 -on-LSCi1 3 interface models. ,
On the other hand, in the case of the LSCF11 3 base film, the top surface layer is predicted
to be fully Sr enriched based on the ab initio surface stability analysis (this work). Upon LSC 214
decoration, the already 100% Sr enriched LSCF113 surface layer has no room for introducing
more Sr, and the much smaller thermodynamic driving force for Sr interdiffusion from the bulk
LSCFl 3 to the bulk LSC214 (Figure 6-4a) suggests that the LSC 214 may contain smaller increase
in Sr content as compared to the LSC214 decorated on the LSC113, which leads to the observed
AES normalized cation intensity ratios shown in Figure 6-3f, where the LSC 214 decorated
LSCF113 film surface exhibits only a slight increase in the Sr content relative to the undecorated
LSCF11 3 base film, while the surfaces of the LSC 214-decorated LSC11 3 film contain significantly
more enhanced surface Sr content vs. the undecorated LSC11 3 base film.
Factors governing surface exchange kinetics of the LSCF113 and LSC1 13 with
and without LSC 214 decoration
180
To understand what is responsible for the observed surface exchange kinetics enhancement for
the LSC 214 decorated LSCF11 3 film (relative to the LSCF11 3 base film) vs. the LSC214 -decorated
LSC11 3 film (relative to the LSC113 base film), here we discuss on factors that govern the
observed oxygen surface exchange kinetics. We consider the surface electronic structure effect
upon Sr segregation within the perovskite phases and the formation of secondary passive phases,
2 5 57 62 63
both of which have been shown to strongly influence the surface exchange kinetics. , , ,
Surface Electronic Structure changed by Surface Decoration
The increased Sr content within the perovskite phases has been shown to lead to upshift
of the oxygen 2p band center relative to the Fermi level5 7,63, which further correlates with the
observed surface oxygen exchange kinetics. In addition, by utilizing the surface layer Sr content
within the perovskite phase from the COBRA measurement, it was also shown the upshift of the
oxygen 2p band centers with increasing Sr can correlate with the observed activity enhancement
for oxygen surface exchange of the LSC 2 14 decorated LSCn3 vs. LSCi1 32 9. In Figure 6-5a, we
compare the relative change of the surface exchange coefficient (kg) vs. the calculated 0 2p band
centers (relative to the Fermi level) between the LSC 214-decorated LSCF11 3 (LSC11 3) and the
LSCF11 3 (LSC1 1 3) base film surfaces by utilizing the surface Sr content of the stable LSC113 and
LSCF 1 3 (001) surfaces as well as the thermodynamic driving force of SrLa substitution between
LSC 2 14 vs. LSC 1 3
and LSC2 14 vs. LSCFil 3 . For LSC 1 3 and LSC21 4 decorated LSC113 films, the
Sr content in the surface layer of the LSC113 can be increased from 60%~75% for the
undecorated LSC11 3 film to 100% for the undecorated LSC113 film surface with the LSC214
decoration 29 . Furthermore, the fully Sr occupied top surface layer of the LSC214 islands on the
LSC11 3 film contains a higher 0 2p band center than the surface layer of the particles on the
LSCO1 1 3 film, which leads to higher activities (-2 orders magnitude enhancement in k values).
On the other hand, since the stable LSCF11 3 (001) surfaces may already contain -100%
Sr predicted by the DFT modeling, there is no change in Sr content for the LSCF1 13 film surfaces
upon LSC2 1 4 decoration. In addition, the much smaller thermodynamic driving force for SrLa
interdiffusion from the LSCF11 3 to the LSC2 1 4 shown in the DFT models (Figure 6-4) suggests
that much smaller amount of Sr may enter into LSC2 14 in comparison with the decorated LSC2 1 4
on the LSC11 3. Overall, both no increase in Sr content of the LSCF1 1 3 surface layer and minor
181
increase of Sr content in LSC 214 will lead to almost no change on the 0 2p band centers relative
.
to the Fermi level, resulting in the only 2 fold enhancement of the kq values of LSCF 1 3
Therefore, the unique Sr occupancy changes within the LSC 214 -on-LSCF11 3 and LSC 214-onLSC 1 3 heterostructure might be responsible for the observed enhanced surface exchange kinetics
relative to the undecorated LSCF113 and LSC1 13 base films.
E 100
5
a)
LSC214/LSC 1 (100% S
(A
S0
214 11
0
1000%0"r)
r Lr
o
0
.4
0 01
(Sr 62.5%)
L113
LC
-@-La
-0-
:.2
-2.50
*
*
@00
2
Sr CoO3
La Sr CoO 4
-1 .5
-2.00
-1.75
-2.25
0P.and c.nter.vs. E__, eV
.
1n- 7 0
-
0.4
0.2
0.0
Relative bulk 0 2p band center shifts vs base film / eV
LSC /LSC
113
0 214
10
E
E
-10
r
r
LSCF
LSC
r b)
0
10
20
30
40
1010
113
r c)
-
-10
io
10-a
I
S
50
60
70
0
10
20
30
40
50
60
70
Annealing Time / hrs
Annealing Time / hrs
Figure 6-5. (a) The relative ratio of the surface exchange coefficient kq (with respect to the kq of
the base film) vs. the calculated bulk 0 2p band center shifts (relative to the Fermi level)
between the LSC 214-decorated LSCF 13 (LSC113 ) film surfaces and the undecorated LSCF 113
182
(LSC 1 3) film surfaces based on the surface Sr information from COBRA analysis 6 3 and DFT
modeling predictions. The inset shows the Sr content vs the calculated bulk 0 2p band centers of
LSC1 1 3 and LSC2 14 . (b) k' of the LSC113 with and without LSC 214 coverage and (c) the LSCF113
with and without LSC 214 coverage as a function of annealing time. Annealing was performed at
550 *C in an oxygen partial pressure of 1 atm.
Surface Sr-enriched secondary phases
The surface decoration can modulate the surface Sr segregation and the surface phase
stability, which can greatly influence the oxygen surface exchange kinetics and the surface
stability in LSC64 and LSCF. 33 , 65 In the case of LSC11 3 with and without LSC214 coverage, no
particles were observed on the surfaces before annealing in the AFM images, as shown in Fig. 6b
and 6c. After annealing at 550 *C for 6 hours, discrete particles, which have higher Sr Auger
signals (Figure S6-7) than the rest of the surface as well as the surfaces before annealing, were
noted on the undecorated LSC,1 3 (Figure 6-6f). This observation is consistent with our recent in
situ studies of surface structure and chemistry changes of LSC,1 3 films 28 and the formation of Srenriched particles on annealed LSC11 3 (Sr=40%) film surfaces. 64 In contrast to the LSC113, no
particles were found on the LSC 214-decorated LSC11 3 surface after annealing for 6 hours, which
indicates that LSC 214 decoration can suppress the formation of Sr-enriched particles on the
LSC1 1 3 surface resulting in the enhanced surface activity of LSC11 3. This can be supported by
previous work25 showing that the porous LSC214 phase plays an important role in retaining the
perovskite phase of LSC11 3 film surface, leading to the surface activity enhancement. This can be
further supported by that a detrimental influence of Sr-enriched particles on the oxygen surface
exchange activity of LSC113 (Sr=40%) films, having improved surface activity by removal of
these particles via chemical etching.64 It should be noted that no particles were found on the
LSCF11 3 surface regardless of the LSC 214 coverage after at 550 'C for 6 hours (Figure 6-6h and
6i). As shown in Figure 6-4 and Ref. 18, LSCF113 is thermodynamically more stable than LSCu3,
which may lead to no particles on the LSCF11 3 surface after annealing for 6 hours, and therefore
LSC 2 14
decoration cannot further promote surface stabilization for the already more stable
LSCF11 3 surfaces, resulting in relatively small surface activity enhancement of the LSCF11 3, as
shown in Figure 6-2c.
183
As shown in Figure 6-5b, the LSC214 decoration can also prevent the time-dependent
degradation of the oxygen surface exchange kinetics of the LSC113 thin film. After annealing at
550 *C for 70 hours, the k values of the undecorated LSC113 was found to significantly decrease
as annealing time increases, while LSC214-decorated LSC113 showed a relatively small reduction
in the k" values after 70 hours annealing. In contrast, LSC214 decoration had no influence on
preventing the time-dependent degradation of the oxygen surface exchange kinetics of the
LSCF11 3 thin film. The VSCs of the LSC113 and LSCF11 3 with and without LSC2 1 4 coverage was
found to not change significantly with increasing the annealing time (Figure S6-6), which
indicates that the degradation of the oxygen surface exchange kinetics as a function of time does
not depend on the bulk oxygen content.
Interestingly, discrete particles, which also have higher Sr Auger signals (Figure 6-6a),
were observed on the surfaces of all samples after annealing for 70 hours, as shown in Figure 66j, 6k, 61, and 6m. The chemistry of these Sr-enriched particles is not well understood. In the
case of the undecorated LSC113, however, recent COBRA experiments suggest that these Srenriched particles on the LSC11 3 thin films have a composition approaching to that of SrCoO 3-4 66
'
which is likely to coexist with or decompose to secondary phases such as SrO/Sr(OH) 2 /SrCO 3 67
68
and (La,Sr) 2CoO 4 ,69 , leading to surface passivation for the oxygen exchange kinetics.
70
Similar
to the LSC 1 3 , LSCF 1 3 can decompose into A2B0 4 18 after annealing over 16 hours and it can
also lead to the formation of the surface Sr-enriched particles, 65 which contribute to degradation
of cell performance. 62 As the Sr spectra of the particles on the LSC214-decorated LSC11 3 was
found to resemble that of an as-deposited LSC214 sample prepared by PLD, it is postulated that
the discrete particles on the LSC214-decorated LSC113 result from the agglomeration of LSC21 4
71
2
coverage , which reduces active heterointerface region , and/or the decomposition of the
.
LSC 1
3
and LSC21 4 coverage 72, which can possess higher Sr concentration resulting from the
cation interdiffusion (Figure 6-4) for long annealing time, and thus the LSC2 14-decorated LSCM3
-
can lead to the reduction in the k values after 70 hours annealing. In the case of the LSC2 1 4
decorated LSCF113, the particles may be attributed to the decomposition of both LSCF113 18 and
LSC2 1 4 , which can be supported by the AES Sr spectra having a similar intensity and shape as
.
shown in the particles on the LSCF11 3
184
Annealed at 550 0 C
Particles
Pristine
LSC214
C
-
LSC
-
LSC21 LSC113
-
LSCF
11 3
--
1600
LSC
1620
a)
/LSCF
1660
1640
1680
Kinetic Energy / eV
,
Figure 6-6. Auger spectra and atomic force microscopy (AFM) images for bare LSC113
LSCFil 3 , LSC 214 -decorated LSC113 , and LSC21 4 -decorated LSCF,1 3 thin films. (a) Sr Auger
-
spectra for: LSC113 (gray), LSC 214-decorated LSC113 (blue), LSCF113 (green), and LSC2 1 4
decorated LSCF,1
3
(yellow) after annealing at 550 0 C for 70 hours in an oxygen pressure of 1
atm. The dashed orange line is the Sr spectra of a pristine LSC 214 reference sample. The peak-topeak values in Auger spectra reflect the Sr concentrations. AFM images of as-deposited (b)
LSC11 3 , (c) LSC214-decorated LSC113 , (d) LSCF 1 3 , and (e) LSC2 14 -decorated LSCF11 3. AFM
image showed particle formation on (f) 6 h annealed LSC,1 3 but no particles were observed on
(g) 6 h annealed LSC2 14-decorated LSC113 , (h) 6 h annealed LSC214 -decorated LSC113 , and (i) 6 h
annealed LSC214-decorated LSC113 . After annealing for 70 h, particles were observed on all
surfaces; (j) annealed LSC113 , (k) annealed LSC214-decorated LSC11 3 (1) annealed LSCF 1 3 and
(in) annealed LSC 214 -decorated LSCF1 13.
6.4 Conclusion
We show that LSC2 1 4 decoration can differently affect the oxygen surface exchange
kinetics and the surface stability of LSC,1 3 and LSCF,1
3
thin films. LSC2 1 4 decoration leads to
only -2 times enhancement of the kq values of the LSCF113 thin films while it enhances the kq
185
values of the LSC113 thin film up to -2 orders of magnitude. In addition, the time-dependent
degradation of the oxygen surface exchange kinetics of the LSC113 thin film is markedly
improved by LSC214 decoration in contrast to the LSCF11 3 with LSC 214 coverage. AES reveals
that the surface Sr concentration of the LSC11 3 significantly increases with LSC214 coverage
while LSC214 decoration does not increase the surface Sr concentration of the LSCF11 3, which is
good agreement with cation interdiffusion DFT modeling. This result can be further supported by
ab initio surface stability analysis, showing limited enhancement in Sr concentration is allowed
at the LSCF11 3 perovskite surfaces by LSC 214 decoration due to the already saturated surface Asite Sr content, while for the LSC11 3 (001) surfaces the unsaturated A-site Sr occupation has
available concentration space for increase of Sr content by promoted the LSC 214 decoration. We
show that the change in the surface electronic structure and the suppression of the formation of
secondary passive phases as a result of LSC 214 decoration can be responsible for observed
oxygen surface exchange kinetics and time-dependent surface stability. Our results represent that
the influence of heterostructured interface on the oxygen surface exchange kinetics is dependent
on the perovskite based film, which can be generally utilized to develop highly active and stable
oxygen surface exchange materials for oxygen electrocatalysis at intermediate temperature.
6.5 Supporting Information (SI)
Target Synthesis. Both Lao.6 Sro.4 Coo. 2 Feo.sO3-8 (LSCF 1 3) and LaSrCoO 4 8 (LSC 2 14 ) were
,
prepared by the Pechini methods. La(N0 3) 3*6H 20, Co(N0 3)3 06H2 0, Fe(NO 3)3 *9H2O, Sr(N0 3 ) 2
and separately La(N0 3) 3*6H20, Co(NO3) 306H 20, Sr(N0 3)2 were dissolved in de-ionized water
with ethylene glycol, and citric acid (Sigma-Aldrich, USA) mixture to synthesize LSCF11 3 and
LSC 2 14
respectively. After esterification at 100 *C, the resin was charred at 400 *C and finally
calcined at 1000 *C in air for 12 hours. The Lao. 8Sr. 2CoO 3..8 (LSC1 1 3 ) and Gdo.2Ceo.80 2 (GDC)
were also prepared by the Pechini methods 66 . La(N0 3)3 *6H2 0, Sr(N0 3 )2, Mn(NO 3)2 *6H2 0, and
separately Gd(N0 3)3 and Ce(N0 3)3 were dissolved in de-ionized water with ethylene glycol, and
citric acid (Sigma-Aldrich, USA) mixture to synthesize LSM82 and GDC respectively. After
esterification at 100 *C, the resin was charred at 400 *C and finally calcined at 1000 *C in air for
12 hours. Pulsed laser deposition (PLD) target pellets with 25 mm diameter were subsequently
186
fabricated by uniaxial pressing at 50 MPa. The LSCF113, LSCi 13 , LSC 2 14 , and GDC pellets were
fully sintered at 1,300 0C in air for 6 hours, 1,200 *C in air for 10 hours, 1,350 *C in air for 12
hours, and 1,100 *C in air for 14 hours, respectively.
Sample preparation. Single crystal 9.5 mol% Y 2 0 3-stabilized ZrO 2 (YSZ) wafers with (001)
orientation and dimensions of 10
x
5
x
0.5 mm (MTI corporation, USA), were used as substrate.
Prior to LSC214 , LSCF113, LSC1 1 3, and GDC deposition, platinum ink (Pt) (#6082, BASF, USA)
counter electrodes were painted on one side of the YSZ and dried at 900 *C in air for 1 hour.
PLD was performed using a KrF excimer laser at 1 = 248 nm, 10 Hz pulse rate and 45 mJ pulse
energy under p(O2) of 50mTorr with 500 pulses of GDC (-5 nm) at 550 *C, followed by 15,000
pulses of LSCF113 (-63 nm) at 650 *C. PLD was also performed using the same laser conditions
under p( 0
2)
of 1 O0mTorr with 500 pulses of GDC (-5 nm) at 550 *C, followed by 15,000 pulses
of LSC11 3 (-85 nm) at 650 *C. The film thicknesses were determined by atomic force
microscopy (AFM). The utilization of reflection high-energy electron diffraction (RHEED)
enabled diagnostic in-situ monitoring of the LSC82 film growth. Immediately after completing
the LSCF113 base film deposition, LSC 214 films were subsequently deposited; for the LSC 214
surface coverages consisting of 50 pulses (-0.3 nm), 150 pulses (-0.8 rum), 500 pulses (-2.6 nm),
and 1,000 pulses (-5 nm). The LSC 214 films (-2.6 rim) were also subsequently deposited on the
LSC1 1 3 base film. LSC 2 14 decoration layer thickness is extrapolated from AFM of the 500 pulses
and 1,000 pulses LSC 214 coverage on LSCF113 . After completing the final deposition, the sample
was cooled to room temperature in the PLD chamber for -1
hour under an oxygen partial
pressure of 50 mTorr.
HRXRD analysis of LSC 214 decorated LSC113 thin film. Normal XRD data (Figure S6-2a) of
the undecorated LSC113 and LSC 214-decorated LSC113 films clearly show the presence of the
(00l),e (1 is integer) peaks of LSC113 and (OOl)cubic (I is even) peaks of GDC and YSZ, indicating
/
that the LSC11 3 film grew epitaxially with the following epitaxial relationships: (001),eLSC11 3
(00 1)cubicGDC //(00l)cubicYSZ. With LSC 214 coverage equal to -2.6 nm in thickness, the (OOl)eNt.
/
(1 is integer) peaks of LSC 214 was found to show, representing (001)tetra.LSC 2 14 // (001)pcLSCii3
(001)cubicGDC / (001l)cubicYSZ. The subscript "tetra." denotes the tetragonal notation. Off-normal
phi-scan analysis of the undecorated LSC11 3 and LSC 2 14-decorated LSC113 films shows that
187
LSC2 14 {1
0 3}tetra.,
LSC113 {l0l}pc, GDC {2 0 2 }ubic and YSZ {2 0 2 }cubic have strong peaks with 4-
fold cubic symmetry (Figure S6-2b), which reveals the in-plane crystallographic relationships
between GDC and YSZ (a cube-on-cube alignment), LSC113 and GDC (an in-plane 450 rotation
with [100]pcLSC1 1 3 / [110]cUbicGDC // [1 l0]cubjcYSZ), and LSC113 and LSC 214 (no rotation with
[100]pcLSC11 3
/ [100]tet.LSC 214), as shown in Figure S6-2c.
-
Relaxed lattice parameter determination by HRXRD. The Relaxed lattice parameter a and
are derived from the following equation (where a and ^ are the relaxed lattice parameters for the
fimi nuntandstate),, 41' 67' 68 (CLC = -2v (a-a).7h i~ln
film in an unstrained
assuming a = c, and v = 0.25.67 The in-plane
strain is given by: Eaa = (a-a) and the out of plane strain by: Ecc =
a
Microelectrodes Fabrication.
In situ electrochemical impedance spectroscopy (EIS)
-
measurements were conducted to probe ORR activity on geometrically well-defined LSC2 14
decorated LSCF113 microelectrodes fabricated by photolithography and acid etching, where
sintered porous Pt sintered onto the backside of the YSZ substrate served as the counter electrode.
OCG positive photoresist (Arch Chemical Co., USA) was applied on the LSC 214-decorated
LSCF11 3 surface and patterned using a mask aligner (Karl Stss, Germany, 2 = 365 nm). The
photoresist was developed using Developer 934 1:1 (Arch Chemical Co., USA) and the thin
films were etched in hydrochloric acid (HCl) to remove LSC 214 -decorated LSCF113 film excess
and create the circular microelectrodes (diameters -50 gm, -100 gm, -150 pm, and -200 Pim,
exact
diameter
determined
by
optical
microscopy).
The
LSC 2 14-decorated
LSC11 3
microelectrodes were also fabricated by using the same manner. Before electrochemical testing,
microelectrode geometry and morphology was examined by optical microscopy (Carl Zeiss,
Germany) and atomic force microscopy (AFM) (Veeco, USA).
Electrochemical Characterization.
Figure S6-4b and 4c detail the equivalent circuit and
corresponding Nyquist plot for this experimental system. ZView software (Scribner Associates,
USA) was used to construct the equivalent circuit and perform complex least squares fitting. The
EIS data were fitted using a standard resistor (R 1) for HF and resistors (R2 ) in parallel with a
constant phase elements (CPE2 ) for MF and LF (R1-(R2/CPE2)-(RoRR/CPEoRR)). Based on the
188
p(O2) dependence of the three features, physical or chemical process with regard to each
frequency range can be determined. 1' 55 '6 9' 70 The HF feature was found unchanged with p(O2),
and its magnitude and activation energy (-1.15 eV) were comparable to those of oxygen ion
conduction in YSZ reported previously7".
The MF feature, which was found to have a p(O2)
independent feature, was attributed to interfacial transport of oxygen ions between the LSCF1
3
film and the GDC layer. In addition, the magnitude of its capacitance was relatively small (~ 10-6
F) compared to the LF feature (~ 10,3 F). The LF feature was found to have a strong p(O2)
dependence. The resistance of the LF feature drastically increases as oxygen partial pressure
decreases. In the case of thin film samples, the magnitude of capacitance is due to the oxygen
content change in the films. Therefore, the electrode oxygen surface reaction corresponds with
the LF feature. We obtained values for Roj and knowing the area of the microelectrode (A 1cn.trde
0
RoRR
-
= 0.25 7 deectrode 2 ). Then, we can determine the ORR area specific resistance (ASRoR =
Aelectrode). The electrical surface exchange coefficient (ku), which is comparable to k*,7 2 was
4 4 73
determined using the expression,
0
=
RT /
4F2ROR
(1)
electrodeCo
where R is the universal gas constant (8.314 J mol-1 K1), T is the absolute temperature, F is the
Faraday's constant (96,500 C mol-1), and co is the lattice oxygen concentration in LSCF113 where
(2)
co = (34)IV,
Vm is the molar volume of LSCF11 3 at room temperature. In this study, c. was calculated with 3
extracted from previous reported values. 74
The electrical surface exchange coefficient (kg) of the LSC11 3 and LSC 214 -decorated
LSC11 3 thin film was also determined using the same manner. EIS data collected from the
LSC 214-decorated LSC113 film at 550 *C as a function of p(O2) are shown in Figure S5. The
predominant semicircle was found to increase with decreasing oxygen partial pressure, where
EIS data of the LSC 214-decorated LSC113 was found to show nearly perfect semicircle
impedances." Considering the fact that the film thickness is smaller than the critical thickness
(- nm for bulk LSC,, 3 at 550 *C6 '75), the p(0 2 )-dependent impedance responses suggest that
the oxygen surface exchange kinetics governs the oxygen electrocatalysis on the film surface.
The LSC2 14 surface coverage may change the c. value of the system. For estimating this
influence we compared LSC 2 14 co values with LSCF1 3 co values. However, calculated co values
for LSC 2 14 were only -1 - 2 % different from those for LSCF1 1 3. We therefore decide to use c.
189
%
values for LSCF11 3 for all samples. Similarly, calculated c, values for LSC 214 were only -1 - 2
.
different from those for LSC1 1 3
VSC, indicative of changes in the oxygen nonstoichiometry induced by changes in the
electrical potential, can be obtained from EIS data via the expression 76
VSC = [ 1/(Aeectrode e thickness)]((RoRR)'"Q)"",
where
Q
(3)
is the non-ideal "capacitance", and n is the non-ideality factor of CPE. The fitted
values of n for semi-circle CPEORJ were found to range from ~ 0.96 to 1.0 over the entire PO2
range examined (n =1, ideal).
Experimental details of auger electron spectroscopy (AES).
In AES, the obtained energy
spectrum for a particular element is always situated on a large background (low signal-to-noise
ratio), which arises from the vast number of so-called secondary electrons generated by a
multitude of inelastic scattering processes. To obtain better sensitivity for detection of the
elemental peak positions, the AES spectra from this study are presented in the differentiated
form. Elemental quantification of AES spectra utilized relative sensitivity factors (RSFs) of
0.059, 0.027, 0.076, 0.178, and 0.212 for Lam, SrLMV, COLmm, FeLMM, and OKLL, respectively,
as supplied by the AES manufacturer (Physical Electronics). In addition, the Inelastic-MeanFree-Path (IMFP) was calculated to correct signal intensity for their different IMFPs
(information depth). IMFPs were calculated using the NIST Standard Reference Database 71
"NIST Electron Inelastic- Mean-Free-Path Database" version 1.2. The software program
provides the ability to predict the IMFP for inorganic compounds supplying the stoichiometric
composition of La (0.6), Sr (0.4), Co (0.2), Fe (0.8), and 0 (3), the number of valence electrons
per molecule (assumed to be 24.8), the density (6.36 g/cm 3 ) and a band gap energy (for which we
are assuming 0 eV as LSCF1 1 3 is metallic like at high temperatures; additionally when assuming
a band gap of an insulator 5 eV, the IMFP increases by ~0.03 nm). The IMFP for La, Sr, Co and
Fe were determined to be 1.395, 2.667, 1.607, and 1.404 nm, respectively. A relative depthscaling factor (s) was determined as:
o
Y
0 Ai
exp
(4)
x )dx ,
\Ai/
where A2is the IMFP, yielding OSr = 0.41,
Uco
= 0.58, and ULa and OrFe = 0.63. The intensities from
different elements were scaled using Iscaed = Imeasured*Si/Si. Similarly, the IMFP for La, Sr, and Co
190
of LSC11 3 was determined to be 1.1.337, 2.549, and 1.337 nm, respectively, by using the the
stoichiometric composition of La (0.8), Sr (0.2), Co (1), and 0 (3), the number of valence
electrons per molecule (assumed to be 29.8), the density (6.931 g/cm 3) and a band gap energy
(assumed 0 eV). The obtained values of the relative depth-scaling factor for LSCu3 are thus
approximately equal to those of LSCF113 . The La and Sr concentration (CLa or Csr) was obtained
by normalizing to the their sum, Ci=I/(ILa+ISr). The Co concentration was also obtained by using
the same manner.
191
Figure S6-1. AFM images of (a) as-deposited pristine LSCF 1 3 ~63 nm, (b) LSCF113 with -0.3
nm LSC 2 14 , (c) LSCF 1 3 with -0.8 nm LSC 2 14 , (d) LSCF11 3 with -2.6 nm LSC 2 14 , and (e)
LSCF,1
3
with ~5 tim LSC 2 14 . RMS roughness values were in the range of 0.24 - 0.32 nm and
comparable across all surfaces.
192
b)
a)
=0
LSC 214 (1 0 3 )etra.
45
C)La
a
Co
ONN
~20
LS (202)pc
CD Cc
-
CO
La/Sr
Co
O%
0
0jC
_0
GDC (202)c,
450S
Ce/Gd
'.
0
0
0)
20
40
60
80
100
I
I
I
I
I
I
I
I
-180
20/0
Z (2O2)~
Zr/Y
-90
0
0/ 0
90
180
Figure S6-2. X-ray diffraction (Cu Ka) analysis at room temperature. (a) Normal XRD of the
epitaxial LSC113 reference and the LSC 214 -decorated LSCIH film, (b) off-normal XRD of a
similarly prepared sample with a -2.6 nm LSC 2 14 coverage, and (c) schematic of the
crystallographic
rotational
relationships
among
GDC(001)cubic, and YSZ(O01)cubic.
193
the
LSC 2 14 (001
)tetra.,
LSC1 13(001)pc,
1.1.4 ~Iw
09
3.95 3.95
1.0
-0-
I..
E 3.901.
O
LSC 214/LSCF
e...---O
O-O
-0-
-0-- -C
LSC214/LSCF
LSC214/LSC, 1 3
0.5
0.5 Z
0
0.0
0.0 C
C-.
(U
a-a
_j
X
3.85 1
-0.5
0
-I
-0.5
LSC/214 LSC
13
3.80
2 a)
Ref.
-1.0
0.3
0.8
2.6
5.1
LSC2 14 Thickness / nm
b)
Ref.
0.3
0.8
2.6
5.1
LSC214 Thickness / nm
-1.0
Figure S6-3. (a) Relaxed lattice parameters of both LSCF1 13 and LSC113 as a function of LSC2 14
thickness, calculated from HRXRD data. (b) In-plain and out-of-plane strains of both LSCF, 1 3
and LSC 113 as a function of LSC214 thickness, calculated from HRXRD data.
194
(a)
(b)
V
RORR
R1
R2
Substrate
(MF) Interface
(c)
(IF) Film
60
550 C
LF
'
40
to
C 20
02
HF
02HF
1
0
MF
LSC 1 -2.6 nm
40
20
Real Impedance I kn
Figure S6-4. (a) Schematic of a LSC 214/LSCF1 13 or LSC,1 3/GDC/YSZ(001)/porous Pt sample
and electrochemical testing configuration (not drawn to scale), and (b) equivalent circuit (R1 =
YSZ electrolyte resistance, R2 = electrode/electrolyte interface resistance4, RORR
=
ORR
resistance, CPE = constant phase element) used to extract ORR kinetics, and (c) Nyquist plot of
the epitaxial LSCF 1 3 with ~2.6 nm LSC 214 coverage at 550
(HF: 104 ~ 105 Hz, MF: 103 ~ 104 Hz, and LF: 10-2 ~ 10 3 Hz).
195
0 C;
inset shows a magnification
50
I
*
I
*
I
*
I
*
I
"
"
,~
A ~
40LSC214 decorated on LSC113
30
20
-
)10latmn
0
10
20
30
40
50
60
Real Impedance (ku)
Figure S6-5. Nyquist plot of the epitaxial LSC 214-decorated LSC113 thin film as a function of
oxygen partial pressure at 550 0 C. EIS data of the LSC 214-decorated LSC 113 was found to show
the p(0 2)-dependent impedance responses, which suggest that the oxygen surface exchange
kinetics governs the oxygen electrocatalysis on the film surface.55
196
1000 I
.
W
-
V
.
.
V
800
-
4
E
LSCF
600 a
a
U
so1
00
14
400
a
S060"
200
0
10
LS
113
20
30
LSC
40
/LSC1
50
60
70
Annealing Time / hrs
Figure S6-6. Sr Auger spectra for the undecorated LSC13 thin film after annealing at 550 0 C for
6 hours. Observed particles on the surface of LSC11 3 shows higher Sr peak intensity compared to
the rest of the film surface.
197
Annealed at 550 0C
Particles
LM
CO)
-- LSC 11 particles
.
1600
base film
I 113
1620
I
-
1640
1660
Kinetic Energy /eV
-
LSC
---
1680
Figure S6-7. Sr Auger spectra for the undecorated LSC1 13 thin film after annealing at 550 0 C for
6 hours. Observed particles on the surface of LSC 1
the rest of the film surface.
198
3
shows higher Sr peak intensity compared to
Lao. 625Sro. 375Feo. 75 Coo.25 0 3
(a)
Uwi_)
0 meV/FU
5 meV/FU
11 meV/FU
(iv)
(v)
(vi)
21 meV/FU
23 meV/FU
45 meV/FU
Lao. 5Sro. 5Feo. 75Coo. 250 3
(b)
U,,i
(iv)
(vi)
(v)
199
(Lao. 5Sro. 5) 2CoO 4
(C)
(ii)
(i)
(iii)
(iv)
(ii) ~(iii)
(iv)
()
(d) (Lao. 437 5Sro.5625) 2CoO 4
(ii)
(i)
I
4
I
A6.
Figure S6-8. Simulated LSCF,1 3 and LSC 214 configurations for calculating the energies of Sr
substitution
of La
Lao.5 Sro.5 Feo.75 Coo. 2 5 0 3
(SrLa)
in
LSCF 1 i 3 and
LSC 214
(a)
Lao. 6 25 Sro. 3 75 Feo. 75 CO0.2 5 0
3
(b)
(with an additional Sr in the simulated 2x2x2 supercells), (c)
(Lao. 5Sro. 5) 2 CoO 4 , and (d) (Lao. 4 37 5Sro. 562 5) 2 CoO 4 . Elements are represented as: La (green), Sr
(light blue), Fe (brown, center of the octahedra), and Co3 (dark blue, center of the octahedra). 0
ions are located at the corners of all the octahedral.
200
(a)
Config. 1
Config. 2
Config. 3
Al
C
A4
A5
A6
B1
0.00
-0.40
0
B2
B3
B4
Al
+
,
A3
............
11
.
(b)
A2
S-0.80-
-1.20
U)
-1.60
0 Config 3
A Config 2
+Config 1
Figure S6-9. (a) Schematics of the heterostructured interfaces with various A-site and B-site
arrangements in the DFT simulations. LaO. 62 5 SrO. 3 7 5Feo. 75Coo.2 5 03 represents the LSCF 113 phase
and (Lao. 5Sro. 5) 2CoO 4 represents the LSC2 14 phase. Elements are represented as: La (green), Sr
(light blue), Fe (brown, center of the octahedra), and Co3 (dark blue, center of the octahedra). 0
ions are located at the corners of all the octahedra. The AO planes are numbered from Al
201
through A6 in the LSCF11 3 and BI through B4 in the LSC 214 phase. The planes Al, B1 and B2
represent an interfacial region. The relative stability of SrLa substitution energy relative to
Lao. 75Sro.2 5CoO 3 , or E(SrLa)-E(SrLa) of LSCj 1 3(25%Sr), with variation in the SrLa defect position
across the AO planes. Values are relative to a bulk LSC, 13(25%Sr) reference (y=O). Also shown
is a dotted horizontal line representing the SrLa substitution energies for the bulk LSCF11 3 (green
dotted line, E(SrLa)LSCFll
3
- E(SrLa)LSC11 3 (25%Sr)), and a black dash-dotted line for the bulk
LSC 2 14 (or E(SrLa)LSC 2 14 - E(SrLa)LSCl1 3(25%Sr)). Note that the more negative values on the y-
axis correspond to the easier substitution of the SrLa relative to bulk LSC11 3 (25%Sr).
202
n
SrCoO 2 . precip. line
C0304 precip. line
0
0.000
-0.014
-0.028
0
*
LaCoO 3
.
precip. line
000
0
-0.042
0
-0.056
c
-0.0708
0
La2 0 3 precip. lin
-0.14'
-0.126
T -823 K, P = I atm
-4
Apeffsr(Lao.7 5Sro.26CoO 3), eV
Figure S6-10. Bulk LaO. 75Sro.25CoO
3
phase diagram at T=823 K and P = 1 atm. We note the
chemical potential of 0 is fixed by setting T and P, while the DFT total energy of
Lao. 75Sro. 2 5Co0 3 provides a constraint for the three (effective) metal chemicals so that only two
(effective) metal chemical potentials are needed to construct the phase diagram. The two
independent effective metal chemical potentials are represented by Ay4f(La
A4f (LaO. 75Sr.25CoO 3 ) , where Auc (LaO. 75 SrO.25CoO 3 )
Ap
(La 0 75 Sr.2 5CoO 3 ) _ Y
0(La
7 5SrO.25CoO
3
)-
=
75
SrO.2 5 CoO3 ) and
pg (LaO 75SrO.2 5CoO 3 ) - pug (Co 3 04) and
(SrO). The shaded area in the phase diagram
represents the stable region for bulk Lao 75Sro. 2 5Co0 3 vs. the LaCoO 3 , SrCoO 2 .5 , Co 30 4 , SrO, and
La 2 0 3 oxides, based on the inequality equations - Equations (22)~(26) using the effective
3
.
chemical potentials of metals of Lao. 75Sro. 2 5Co0 3 , LaCoO 3 , SrCoO2. 5, C0 3 0 4 , SrO, and La 2 0
203
(001) AO-Slab
(001) B0 2-Slab
Figure S6-11. The LSC,1 3 and LSCF1 1 3 (001) slab models used for the ab initio surface
thermodynamic analysis. Green and light blue spheres represent La and Sr, while brown and
deep blue polyhedral represent local Fe-O and Co-O environments, respectively. The top (and
bottom) two surface layers, where La/Sr and Co/Fe compositions (Laj-xSrxCoj-yFeyO
3
with x=0,
0.25, 0.5, 0.75, 1 y=0, 0.25, 0.5, 0.75, 1) are varied, are specified by the rectangular frames. The
central part of the slabs, outside the frames, are fixed to a composition close to Lao.7 5Sro.2 5 CoO 3
and Lao.62 5Sro.37 5Feo. 7 5Co 0.2 50 3 . A total of 10 configurations (5 for the (001) AO surfaces and 5
for the (001) B0 2 surfaces) for LSC 1 3, and a total of 50 configurations (25 for the (001) AO
surfaces and 25 for the (001) B0 2 surfaces) for LSCF 1 3, are calculated based on these 9-layer
2x2 symmetric slab models for the surface stability analysis.
204
SrCoO 2 . precip. lin e
C0 30 4 precip. line
=/)
0.000
-0.014
LaCiO
-0.041
>|
-0.055
-
-0.027
precip. line
-0.082
VX
-0.110
r8
-0.123
oa
LaT23 precip line
T=823 K, P
= I atm
:L.
n.1?7
Ap'fSr (La 0 .7 SrO.2oO,),
eV
Figure S6-12. The predicted Lao 75Sro.2 5CoO 3 surface stability diagram at T = 823 K and P = 1
atm based on the chemical potentials of bulk Lao. 75Sro. 2 5CoO 3 . The grid points represent the
sampled bulk effective chemical potentials of Sr (x-axis; x=0 represents the equilibrium between
Lao. 75 Sro.25 CoO 3 and SrO) and Co (y-axis; y=O represents
Lao.75Sro.25CoO3 and C030 4) in La.
constructed
based
on the
75 Sro.25CoO 3 ,
calculated
and the contour plot beyond the grid is
lowest surface
Lao. 75Sro. 25CoO 3 (001) surface configurations.
205
the equilibrium between
energy
among
the investigated
6.5 References
1.
Z. P. Shao and S. M. Haile, Nature, 2004, 431, 170-173.
2.
B. C. H. Steele and A. Heinzel, Nature, 2001, 414, 345-352.
3.
L. Yang, C. D. Zuo, S. Z. Wang, Z. Cheng and M. L. Liu, Adv. Mater., 2008, 20, 3280-
3283.
4.
L. da Conceicao, C. R. B. Silva, N. F. P. Ribeiro and M. Souza, Mater. Charact., 2009,
60, 1417-1423.
5.
R. A. De Souza and J. A. Kilner, Solid State Ion., 1998, 106, 175-187.
6.
R. A. De Souza and J. A. Kilner, Solid State Ion., 1999, 126, 153-161.
7.
A. Endo, M. Ihara, H. Komiyama and K. Yamada, Solid State Ion., 1996, 86-8, 1191-
1195.
8.
T. Ioroi, T. Hara, Y. Uchimoto, Z. Ogumi and Z. Takehara, J. Electrochem. Soc., 1997,
144, 1362-1370.
9.
S. B. Adler, Solid State Ion., 1998, 111, 125-134.
10.
A. V. Berenov, A. Atkinson, J. A. Kilner, E. Bucher and W. Sitte, Solid State Ion., 2010,
181, 819-826.
11.
T. Kawada, J. Suzuki, M. Sase, A. Kaimai, K. Yashiro, Y. Nigara, J. Mizusaki, K.
Kawamura and H. Yugami, J. Electrochem. Soc., 2002, 149, E252-E259.
12.
J. A. Kilner, R. A. DeSouza and I. C. Fullarton, Solid State Ion., 1996, 86-8, 703-709.
13.
J. Mizusaki, Y. Mima, S. Yamauchi, K. Fueki and H. Tagawa, Journal of Solid State
Chemistry, 1989, 80, 102-111.
14.
V. G. Sathe, A. V. Pimpale, V. Siruguri and S. K. Paranjpe, Journal of Physics-
Condensed Matter, 1996, 8, 3889-3896.
15.
R. H. E. vanDoorn, I. C. Fullarton, R. A. deSouza, J. A. Kilner, H. J. M. Bouwmeester
and A. J. Burggraaf, SolidState Ion., 1997, 96, 1-7.
16.
J. M. Bae and B. C. H. Steele, Solid State Ion., 1998, 106, 247-253.
17.
A. Esquirol, N. P. Brandon, J. A. Kilner and M. Mogensen, J Electrochem. Soc., 2004,
151, A1847-A1855.
18.
S. Hashimoto, Y. Fukuda, M. Kuhn, K. Sato, K. Yashiro and J. Mizusaki, Solid State
Ion., 2010, 181, 1713-1719.
206
19.
L. W. Tai, M. M. Nasrallah, H. U. Anderson, D. M. Sparlin and S. R. Sehlin, Solid State
Ion., 1995, 76, 259-271.
20.
L. W. Tai, M. M. Nasrallah, H. U. Anderson, D. M. Sparlin and S. R. Sehlin, Solid State
Ion., 1995, 76, 273-283.
21.
E. J. Crumlin, E. Mutoro, S. J. Ahn, G. J. la 0, D. N. Leonard, A. Borisevich, M. D.
Biegalski, H. M. Christen and Y. Shao-Horn, J. Phys. Chem. Lett., 2010, 1, 3149-3155.
22.
M. Sase, F. Hermes, T. Nakamura, K. Yashiro, K. Sato, J. Mizusaki, T. Kawada, N.
Sakai, K. Yamaji, T. Horita and H. Yokokawa, in Solid Oxide Fuel Cells 10, eds. K. Eguchi, S.
C. Singhai, H. Yokokawa and H. Mizusaki, Electrochemical Society Inc, Pennington, 2007, vol.
7, pp. 1055-1060.
23.
M. Sase, F. Hermes, K. Yashiro, K. Sato, J. Mizusaki, T. Kawada, N. Sakai and H.
Yokokawa, J. Electrochem. Soc., 2008, 155, B793-B797.
24.
M. Sase, K. Yashiro, K. Sato, J. Mizusaki, T. Kawada, N. Sakai, K. Yamaji, T. Horita
and H. Yokokawa, Solid State Ion., 2008, 178, 1843-1852.
25.
A. Unemoto, K. Nagao, T. Tairako, K. Amezawa and T. Kawada, in Ionic and Mixed
Conducting Ceramics 7, eds. M. Mogensen, T. Armstrong, T. M. Gur and H. Yokokawa,
Electrochemical Soc Inc, Pennington, 2010, vol. 28, pp. 59-70.
26.
K. Yashiro, T. Nakamura, M. Sase, F. Hermes, K. Sato, T. Kawada and J. Mizusaki, in
Solid Oxide Fuel Cells 10, eds. K. Eguchi, S. C. Singhai, H. Yokokawa and H. Mizusaki,
Electrochemical Society Inc, Pennington, 2007, vol. 7, pp. 1287-1292.
27.
K. Yashiro, T. Nakamura, M. Sase, F. Hermes, K. Sato, T. Kawada and J. Mizusaki,
Electrochem. Solid State Lett., 2009, 12, B135-B137.
28.
Z. Feng, E. J. Crumlin, W. T. Hong, D. Lee, E. Mutoro, M. D. Biegalski, H. Zhou, H.
Bluhm, H. M. Christen and Y. Shao-Horn, J. Phys. Chem. Lett., 2013, 4, 1512-1518.
29.
Z. X. Feng, Y. Yacoby, M. J. Gadre, Y. L. Lee, W. T. Hong, H. Zhou, M. D. Biegalski,
H. M. Christen, S. B. Adler, D. Morgan and Y. Shao-Horn, J. Phys. Chem. Lett., 2014, 5, 10271034.
30.
M. J. Gadre, Y. L. Lee and D. Morgan, Physical chemistry chemical physics : PCCP,
2012, 14, 2606-2616.
31.
J. W. Han and B. Yildiz, Energy Environ. Sci., 2012, 5, 8598-8607.
207
32.
J. Chen, F. L. Liang, B. Chi, J. Pu, S. P. Jiang and L. Jian, J. Power Sources, 2009, 194,
275-280.
33.
H. P. Ding, A. V. Virkar, M. L. Liu and F. Liu, Phys. Chem. Chem. Phys., 2013, 15, 489-
496.
34.
M. F. Liu, D. Ding, K. Blinn, X. X. Li, L. F. Nie and M. Liu, Int. J Hydrog. Energy,
2012, 37, 8613-8620.
35.
X. Y. Lou, S. Z. Wang, Z. Liu, L. Yang and M. L. Liu, Solid State Ion., 2009, 180, 1285-
1289.
36.
M. E. Lynch, L. Yang, W. T. Qin, J. J. Choi, M. F. Liu, K. Blinn and M. L. Liu, Energy
Environ. Sci., 2011, 4, 2249-2258.
37.
L. F. Nie, M. F. Liu, Y. J. Zhang and M. L. Liu, J. Power Sources, 2010, 195,4704-4708.
38.
X. B. Zhu, D. Ding, Y.
Q.
Li, Z. Lu, W. H. Su and L. Zhen, Int. J. Hydrog. Energy, 2013,
38, 5375-5382.
39.
A. Mitterdorfer and L. J. Gauckler, Solid State Ion., 1998, 111, 185-218.
40.
E. J. Crumlin, S. J. Ahn, D. Lee, E. Mutoro, M. D. Biegalski, H. M. Christen and Y.
Shao-Horn, J. Electrochem. Soc., 2012, 159, F219-F225.
41.
G. J. la 0, S. J. Ahn, E. Crumlin, Y. Orikasa, M. D. Biegalski, H. M. Christen and Y.
Shao-Horn, Angew. Chem.-Int. Edit., 2010,49, 5344-5347.
42.
D. Lee, A. Grimaud, E. J. Crumlin, K. Mezghani, M. A. Habib, Z. X. Feng, W. T. Hong,
M. D. Biegalski, H. M. Christen and Y. Shao-Horn, J. Phys. Chem. C, 2013, 117, 18789-18795.
43.
D. Lee, Y.-L. Lee, A. J. L. Grimaud, W. Hong, M. Biegalski, D. Morgan and Y. Shao-
Horn, J. Mater. Chem. A, 2014.
44.
J. Maier, Physical Chemistry of Ionic Materials:Ions andElectrons in Solids John Wiley,
Chichester, England ; Hoboken, NJ, 2004.
45.
E. Mutoro, E. J. Crumlin, M. D. Biegalski, H. M. Christen and Y. Shao-Horn, Energy
Environ. Sci., 2011, 4, 3689-3696.
46.
G. Kresse and J. Hafner, PhysicalReview B, 1993, 47, 558.
47.
G. Kresse and J. Furthmuller, Phys. Rev. B, 1996, 54, 11169-11186.
48.
P. E. Blochl, Phys. Rev. B, 1994, 50, 17953-17979.
49.
J. P. Perdew and Y. Wang, PhysicalReview B, 1992, 45, 13244.
208
50.
V. I. Anisimov, F. Aryasetiawan and A. I. Lichtenstein, Journalof Physics-Condensed
Matter, 1997, 9, 767-808.
51.
S. L. Dudarev, G. A. Botton, S. Y. Savrasov, C. J. Humphreys and A. P. Sutton, Phys.
Rev. B, 1998, 57, 1505-1509.
52.
Y.-L. Lee, J. Kleis, J. Rossmeisl and D. Morgan, PhysicalReview B (CondensedMatter
and MaterialsPhysics), 2009, 80, 224101.
53.
M. James, A. Tedesco, D. Cassidy, M. Colella and P. J. Smythe, J Alloy. Compd, 2006,
419, 201-207.
54.
R. H. E. van Doom and A. J. Burggraaf, Solid State Ion., 2000, 128, 65-78.
55.
S. B. Adler, Chem. Rev., 2004, 104, 4791-4843.
56.
B. C. H. Steele and J. M. Bae, Solid State Ion., 1998, 106, 255-261.
57.
Y. L. Lee, J. Kleis, J. Rossmeisl, Y. Shao-Hom and D. Morgan, Energy Environ. Sci.,
2011, 4, 3966-3970.
58.
Y. M. Kim, S. I. Pyun, J. S. Kim and G. J. Lee, J. Electrochem. Soc., 2007, 154, B802-
B809.
59.
Z. Feng, Y. Yacoby, W. Hong, H. Zhou, M. Biegalski, H. Christen and Y. Shao-Hom,
Energy & EnvironmentalScience, 2014.
60.
J. Druce, T. Ishihara and J. Kilner, Solid State Ionics.
61.
N. Nakagawa, H. Y. Hwang and D. A. Muller, Nat Mater, 2006, 5, 204-209.
62.
Z. H. Cai, M. Kubicek, J. Fleig and B. Yildiz, Chem. Mat., 2012, 24, 1116-1127.
63.
A. Grimaud, K. J. May, C. E. Carlton, Y. L. Lee, M. Risch, W. T. Hong, J. Zhou and Y.
Shao-Hom, Nature Communication, 2013, 4, 2439.
64.
M. Kubicek, A. Limbeck, T. Fromling, H. Hutter and J. Fleig, J Electrochem. Soc., 2011,
158, B727-B734.
65.
D. Oh, D. Gostovic and E. D. Wachsman, J Mater. Res., 2012,27, 1992-1999.
66.
Z. Feng, Y. Yacoby, W. T. Hong, H. Zhou, M. D. Biegalski, H. M. Christen and Y. Shao-
Horn, Energy Environ. Sci., 2014, 7, 1166-1174.
67.
H. M. Reichenbach and P. J. McGinn, Appl. Catal. A-Gen., 2003, 244, 101-114.
68.
L. Karvonen, S. H. Yoon, P. Hug, H. Yamauchi, A. Weidenkaff and M. Karppinen,
Mater. Res. Bull., 2011, 46, 1340-1345.
209
69.
A. N. Petrov, V. A. Cherepanov, 0. F. Kononchuk and L. Y. Gavrilova, J. Solid State
Chem., 1990,87, 69-76.
70.
E. Fischer, W. D. Shen and J. L. Hertz, J Electroceram., 2012, 29, 262-269.
71.
H. Huhtinen, J. Jarvinen, R. Laiho, P. Paturi and J. Raittila, JournalofApplied Physics,
2001, 90, 1521-1528.
72.
T. Nitadori, M. Muramatsu and M. Misono, Chem. Mat., 1989, 1, 215-220.
73.
M. Pechini, U.S. PatentNo. 3,330,697, 1967.
74.
H. M. Christen, E. D. Specht, S. S. Silliman and K. S. Harshavardhan, Phys. Rev. B,
2003,68,4.
75.
E. J. Crumlin, E. Mutoro, Z. Liu, M. E. Grass, M. D. Biegalski, Y. L. Lee, D. Morgan, H.
M. Christen, H. Bluhm and Y. Shao-Horn, Energy Environ. Sci., 2012, 5, 6081-6088.
76.
S. B. Adler, J. A. Lane and B. C. H. Steele, J. Electrochem. Soc., 1996, 143, 3554-3564.
77.
Y. L. Yang, C. L. Chen, S. Y. Chen, C. W. Chu and A. J. Jacobson, J Electrochem. Soc.,
2000, 147,4001-4007.
78.
P. S. Manning, J. D. Sirman, R. A. DeSouza and J. A. Kilner, Solid State Ion., 1997, 100,
1-10.
79.
J. Maier, Solid State Ionics, 1998, 112, 197-228.
80.
J. Fleig and J. Maier, J Eur. Ceram. Soc., 2004, 24, 1343-1347.
81.
S. Hashimoto, Y. Fukuda, M. Kuhn, K. Sato, K. Yashiro and J. Mizusaki, Solid State
Ion., 2011, 186, 37-43.
82.
J. Fleig, H. R. Kim, J. Jamnik and J. Maier, Fuel Cells, 2008, 8, 330-337.
83.
J. Fleig, Solid State Ion., 2002, 150, 181-193.
84.
K. Johnston, M. R. Castell, A. T. Paxton and M. W. Finnis, PhysicalReview B, 2004, 70,
085415.
85.
S. Piskunov, E. Heifets, T. Jacob, E. A. Kotomin, D. E. Ellis and E. Spohr, Physical
Review B, 2008, 78, 121406(R).
86.
NIST, in NIST StandardReference DatabaseNo. 69, http://webbook.nist.gov/chemistry/,
eds. P. J. Linstrom and W. G. Mallard, National Institute of Standards and Technology,
Gaithersburg, MD, 2003.
87.
K. Reuter and M. Scheffler, PhysicalReview B, 2001, 65, 035406.
210
88.
J. Leitner, P. Chuchvalec, D. Sedmidubskf, A. Strejc and P. Abrman, Thermochimica
Acta, 2002, 395, 27-46.
89.
Y.-L. Lee and D. Morgan, Physical Chemistry Chemical Physics, 2012, 14, 290-302.
90.
L. Wang, T. Maxisch and G. Ceder, PhysicalReview B, 2006, 73, 195107.
91.
J. Cheng, A. Navrotsky, X.-D. Zhou and H. U. Anderson, Chemistry ofMaterials, 2005,
17, 2197-2207.
92.
J. H. Cheng, A. Navrotsky, X. D. Zhou and H. U. Anderson, J. Mater. Res., 2005, 20,
191-200.
93.
T. R. S. Prasanna and A. Navrotsky, Journalof Solid State Chemistry, 1994, 112, 192-
195.
211
This page is intentionallyleft blank
212
Chapter 7.
Conclusions & Perspectives
7.1 Conclusions
To develop highly active oxide materials and improve our fundamental understanding of
oxygen electrocatalysis at intermediate temperature for the cathode in SOFCs, I proposed two
promising strategies, which are layered RP oxide thin films'
2
and surface decorated ABO 3
perovskite oxide thin films3, 4 in this thesis.
7.1.1 Layered RP Oxide Thin Films (LNO and LSNO)
We revealed strong film thickness dependence on the surface oxygen exchange rates (kg)
of the (100)tetragonai-oriented epitaxial growth of LNO thin films grown on YSZ.1 The k values of
the LNO thin films decrease with increasing film thickness, and such a trend is not observed in
(00 1)tetragona-oriented epitaxial LNO thin films. 5 AES showed that there is no significant change
of the surface chemistry as a function of film thickness before and after exposure to elevated
temperatures. In situ HRXRD revealed that the unit cell volume of the LNO thin films increases
with increasing the film thickness at 550 *C, which indicates oxygen nonstiochiometry, 6, in
La2NiO 4+6 decrease with decreasing LNO film thickness. These results demonstrate the key role
of oxygen excess of RP phases on the oxygen surface exchange process, where modifying the
driving force to form interstitial oxygen by strains is a new strategy to design highly active
surface oxygen exchange materials for applications such as SOFC cathodes or oxygen
conducting membranes.
To further extend our understanding of the Sr substitution effect, which can lead to the
enhancement of the oxygen surface exchange kinetics for the ABO 3 perovskite oxides, on the
surface electrocatalysis of LNO were investigated.2 Interestingly, the Sr substitution led to a
structural reorientation from the (100)teta. (in-plane) to (OOl)tetra. (out-of-plane) orientation as a
function of the Sr content, confirmed by HRXRD. Using DFT modeling, the physical origin of a
structure reorientation was found to have a strong correlation with different surface energies as a
213
function of Sr content. In addition, the k of the LSNO thin films decreased with increasing the
Sr content, which may be attributed to both different adsorption energy and different
crystallographic orientation as a function of the Sr content. This work demonstrates that
modifying the orientation and the adsorption energy by Sr substitution is a new strategy to design
oxygen surface exchange materials with desired anisotropic characteristics for applications such
as SOFC cathodes or oxygen conducting membranes.
7.1.2
ABO 3 Perovskite
Oxide
Thin Films with
Surface
Decoration
)
(LSM1 1 3/LSC1 1 3, LSC 2 ,4 /LSC,1 3, and LSC 214/LSCF113
In order to improve our understanding of the surface decoration effect on the surface
chemistry of perovskites and develop other potential surface modification materials, the effect of
LSM1 1 3 surface decoration on the oxygen electrocatalysis of the LSC11 3 thin films was
investigated.4 LSM11 3 surface decoration coverage can strongly affect the surface exchange
kinetics and the surface stability of epitaxial LSC,1 3 thin films. The k' values of the epitaxial
LSC,1
3
thin films with partial LSM11 3 coverage (-0.1, -0.3, -0.9 nm) were significantly
enhanced relative to the undecorated LSC,1 3 film while those with full LSM113 coverage (-3.5
and -10 nm) were similar or diminished. In situ HRXRD showed that the in-plane and out-ofplane strains of the LSC113 at elevated temperature are not influenced by LSM11 3 decoration. We
showed the suppression of surface Sr-enriched particles by LSM113 using AFM and SEM. Indeed,
AES analysis also showed that the surface Sr and Mn contents increase with increasing LSM11 3
coverage, which can alter the surface exchange kinetics of the LSM,1 3-decorated LSC,1 3 films.
In addition, LSM11 3 decoration can prevent the degradation of the surface activity of LSC,1 3 for
over 70 h annealing. This work highlights that small changes in the surface chemistry of
perovskites can yield significant increases in the oxygen surface exchange kinetics with control
of surface decoration thickness, which can be potentially utilized to develop highly active
oxygen surface exchange materials.
The final study made in this thesis probes that surface decoration with a LSC214 can
differently affect the oxygen surface exchange kinetics and the surface stability of LSC,1 3 and
LSCF,1 3 thin films. 3 LSC214 decoration led to only -2 times enhancement of the k values of the
LSCF113 thin films while it enhanced the k values of the LSC113 thin film up to -2 orders of
214
magnitude. In addition, the long-term degradation of the oxygen surface exchange of the LSC11 3
thin film was markedly improved by LSC 214 decoration in contrast to the LSCF11 3 with LSC214
coverage. AES revealed that the surface Sr concentration of the LSC11 3 significantly increases
with LSC21 4 coverage while LSC21 4 decoration does not increase the surface Sr concentration of
the LSCF1 3,which is good agreement with cation interdiffusion DFT modeling. This result can
be further supported by ab initio surface stability analysis, showing the difference in the surface
Sr concentration between LSC,1 3 and LSCF113. We propose that the change in the surface
electronic structure and the suppression of the formation of secondary passive phases as a result
of LSC21 4 decoration can be responsible for observed oxygen surface exchange kinetics.
7.2 Perspectives for Future Work
Recently, higher order Ruddlesden-Popper (RP) oxides 6' 7 and double perovskite oxides8
10
have attracted much attention as new cathode materials for intermediate temperatures. The
anisotropy of oxygen diffusion and the presence of cationic ordering structures may lead to the
enhancement of the oxygen transport properties compared to LSC and LSCF, making these
materials suitable for intermediate temperature operation. The use of epitaxial thin films is a
direct way to study and exploit the intrinsic anisotropy of oxygen diffusion, lattice strain effects,
and surface properties; therefore, it is essential to utilize the epitaxial thin films of layered
MIECs for developing new cathode materials.
7.2.1 Higher order RP oxides (n=2, 3)
Higher order RP oxides (n=2, 3) have attracted attention as an alternate cathode material
due to significantly higher electronic conductivity than the n=1 RP oxide 6'
7.
However, the
anisotropic nature of the oxygen transport kinetics and oxygen surface exchange kinetics in
higher order RP oxides is not fully understood due to difficulties in the growth of epitaxial
layered perovskite thin films"'
12.
Therefore, epitaxial higher order RP thin films are essential to
investigate the influence of the crystal phase, orientation, and surface chemistry on oxygen
electrocatalysis at intermediate temperatures. The success of this work will enable a
breakthrough in utilizing layered perovskite thin films for electrochemical applications.
215
7.2.2 Double Perovskites Oxides
Double perovskites with the formula AA'B 205 +8, where A is a rare earth, A' an alkaline
earth, and B 3d transition metals, have been recently proposed as the next generation of cathode
materials for intermediate temperature operation due to their high electronic conductivity and
enhanced oxygen transport properties'
.13 This new class of material is characterized by an
ordering of lanthanide and alkali-earth cations which occupy alternative planes in the c
direction.14 This ordering gives rise to anisotropic oxygen diffusion channels in the a-b planes,
which can be utilized and enhanced to increase the cathode performances using epitaxial thin
films. So far, few studies have shown the oxygen transport kinetics of GdBaCo 205 +8 (GBCO)
and PrBaCo 205 +8 (PBCO) ceramic samples and few attempts at the growth of epitaxial GBCO 5
and PBCO1 3 thin films have shown the microstructural complexity of double perovskites in an
epitaxial thin film form. Despite the microstructural complexity, early thin films studies have
suggested that the electronic conductivity can be tuned using microstructural modifications. It is
therefore necessary to get a better understanding of the complex conduction processes occurring
in these phases to further develop new SOFCs cathodes (or MIECs oxides in general) with
enhanced conduction properties. The current state-of-the-art material, which is limited to a few
phases (e.g. GBCO and PBCO), must be extended to new phases in order to fully understand the
complex correlation existing between chemistry, anisotropic diffusion processes, and
electrocatalytic performances.
7.3 References
1.
D. Lee, A. Grimaud, E. J. Crumlin, K. Mezghani, M. A. Habib, Z. Feng, W. T. Hong, M.
D. Biegalski, H. M. Christen and Y. Shao-Horn, The Journal of Physical Chemistry C,
feature article, 2013, 117, 18789-18795.
2.
D. Lee, Y.-L. Lee, A. Grimaud, W. T. Hong, M. D. Biegalski, D. Morgan and Y. ShaoHorn, J. Mater. Chem. A, 2014, 2, 6480-6487.
3.
D. Lee, Y.-L. Lee, M. D. Biegalski, D. Morgan and Y. Shao-Horn, In Preparation,2014.
216
4.
D. Lee, Y.-L. Lee, A. Grimaud, W. T. Hong, M. D. Biegalski, D. Morgan and Y. ShaoHorn, The Journalof Physical Chemistry C, 2014, Under Review.
5.
M. Burriel, G. Garcia, J. Santiso, J. A. Kilner, J. C. C. Richard and S. J. Skinner, J
Mater. Chem., 2008, 18, 416-422.
6.
G. Amow, I. J. Davidson and S. J. Skinner, Solid State Ion., 2006, 177, 1205-1210.
7.
G. Amow and S. J. Skinner, J. Solid State Electrochem., 2006, 10, 53 8-546.
8.
M. Burriel, J. Pena-Martinez, R. J. Chater, S. Fearn, A. V. Berenov, S. J. Skinner and J.
A. Kilner, Chem. Mat., 2012, 24,613-621.
9.
S. Choi, S. Yoo, J. Kim, S. Park, A. Jun, S. Sengodan, J. Kim, J. Shin, H. Y. Jeong, Y.
Choi, G. Kim and M. Liu, Sci Rep, 2013,3, 6.
10.
D. Parfitt, A. Chroneos, A. Tarancon and J. A. Kilner, J Mater. Chem., 2011, 21, 21832186.
11.
M. Burriel, G. Garcia, M. D. Rossell, A. Figueras, G. Van Tendeloo and J. Santiso,
Chem. Mat., 2007, 19, 4056-4062.
12.
K. T. Wu, Y. A. Soh and S. J. Skinner, Mater. Res. Bull., 2013, 48, 3783-3789.
13.
G. Kim, S. Wang, A. J. Jacobson, Z. Yuan, W. Donner, C. L. Chen, L. Reimus, P.
Brodersen and C. A. Mims, Appl. Phys. Lett., 2006, 88, 3.
14.
A. A. Taskin, A. N. Lavrov and Y. Ando, Phys. Rev. B, 2005, 71, 28.
15.
M. Burriel, M. Casas-Cabanas, J. Zapata, H. Y. Tan, J. Verbeeck, C. Solis, J. Roqueta, S.
J. Skinner, J. A. Kilner, G. Van Tendeloo and J. Santiso, Chem. Mat., 2010, 22, 55125520.
217