d10.06.04.presentation
DESCRIPTION
TRANSCRIPT
11
Degradation Mechanisms in Degradation Mechanisms in Solid Oxide Electrolysis Anodes and Bond Layer:Solid Oxide Electrolysis Anodes and Bond Layer:
CrCr--Poisoning and Cation TransportPoisoning and Cation Transport
Vivek I. Sharma and Vivek I. Sharma and Bilge YildizBilge Yildiz
Department of Nuclear Science and EngineeringDepartment of Nuclear Science and EngineeringMassachusetts Institute of Technology, Cambridge, MAMassachusetts Institute of Technology, Cambridge, MA
International Workshop on High Temperature
Electrolysis Limiting
Factors, June
9-10, Karlsruhe, Germany.
Motivation:–
Enable durable, efficient, cost-competitive steam electrolysis technology.
–
Help identify degradation modes of SOECs
tested thus far.
Outline:Challenges, and research in:
degradation mechanisms in the
anode and bond layer materials in present SOECs
Post-mortem analysis of stack-cellsElectrochemical tests
Outstanding questions, and research needs
*Properietary
Information
SOECs
H2
O + 2e-
→ O2-
+ H2
CO2
+ 2e-
→ CO + O2-
O2-
→ ½ O2
+ 2e-
Cathode
Anode
Degradation rate ~40%/1000hrs is greater than in SOFCs
(<1.5%)
Air in
Air out
O2H2
H2
O
e- e-
Electrolyte
[ScSZ] Anode*Cathode
[Ni / Sc2
O3
-ZrO2 (ScSZ) Cermet]
O=
Bond Layer
[La0.8
Sr0.2
CoO3 (LSC)]
0
10
20
30
40
50
0.0 500.0 1000.0 1500.0 2000.0
Time (hours)
Cur
rent
, A
Time, hours
Stoots, O’Brien, Elangovan, Hartvigsen, Herring et al. 2008
0 200 400 600 800 1000
Time (hours)
10
15
20
Cur
rent
(A) 800oC 830oC
0 200 400 600 800 1000
Time (hours)
10
15
20
Cur
rent
(A)
0 200 400 600 800 1000
Time (hours)
10
15
20
Cur
rent
(A)
0 200 400 600 800 1000
Time (hours)
10
15
20
Cur
rent
(A) 800oC 830oC
O’Brien et al, Nuclear Technology, 2006
Components of the stack repeat unit
Interconnect (airside)–
Native scale (Cr2
O3
or AB2
O4
spinel)–
Bond (protective) layer
Flow field (not shown)Bond layerAnode: Oxygen Electrode ElectrolyteCathode: Steam/hydrogen electrodeBond layerFlow field Interconnect (H2O/H2 side)–
Native Scale –
Bond layerSeal
Interconnect
Bond (protective) layer
Bond layer
Oxygen electrode
Flow field here ↑↓
Electrolyte
Steam/H2
electrode
not shown
Composite SEM photo
Carter et al. ACS, 2008
Suspect causes of stack degradation
Oxygen bond layer–
Cr “contamination”
from interconnect–
Delamination
from the oxygen electrodeOxygen electrode–
Delamination from the electrolyteElectrolyte–
Possible aging due to Tetragonal → Monoclinic transitionSteam/hydrogen electrode–
Silica deposition from seal–
Nickel oxidation–
Mn diffusion from interconnect
Single cell tests suggest that SOEC mode has greater degradation rates than SOFC mode governing mechanisms differ.
Objective: Identify and control degradation mechanisms in SOEC anode / bond layer
E.g. Cation interdiffusion between cathode and electrolyte of SOFCs
E.g. Presence of Cr-containing species in SOFC cathode
1. Bond Layer Dissociation
Cation segregation Local variations in cation ratios at
the surface
2. Interdiffusion
of Cr-containing species from interconnects into bond layer and anode
Cr distribution across the bond layer and anode
Relationship between Cr and
local composition
(Cr,Mn)3
O4
Grosjean et al. SSI, 2006
Salvador et al. SECA report, 2006
Approach
Technique Objective
Raman SpectroscopyIdentification of secondary phases
formed
on the bond layer
Nanoprobe
Auger Electron Spectroscopy (NAES)
Electrode surface chemistry and microstructure
and its variation across the cross section at a small
scale (μm-nm)
Focused Ion Beam (FIB)Selectively choose the interface
of interest to prepare TEM samples
Energy Dispersive X-Ray Spectroscopy (EDX) / Transmission Electron
Microscopy (TEM)
High resolution
identification of the chemical composition and
secondary structures
formed
LSCAnode
LSC
ScSZ
Top View
C/S View
10cm
Oxygen electrode microstructural change
Microstructure evolved from round grains for as-prepared anode to faceted grains for used anode. Small changes evolved in the surface chemistry – not disclosed here.
Anode
LSC
Delamination and/or weak interface
3
20000 X4/18/2008
1
1.0 µm
As-prepared anode
Used anode
The weakly bound and delaminated interface could prevent Cr solid-state diffusion into the anode Cr not found on the anode surfaceThis can enable the anode stability, however, not desirable for electronically activating the anode.
Cr-contamination and dissociation of LSC bond layer
Regions or crystallites with high Cr content exist. Cr content varies locally (average 2-10%)
Drastic variations in La/Co at surface seen even at a local level and no Sr-presence was found on the surface of the bond layer.
–
The as-sintered LSC has A/B~6 consistently
SEM image of cross-section in the LSC bond layer of an
operated cell. AES spectra from the 3 points is on the right.
O Cr La Co La/CoArea 1 0.70 0.08 0.08 0.13 0.61Area 2 0.57 0.01 0.38 0.04 9.50Area 3 0.61 0.02 0.29 0.07 4.14
O Cr La Co La/CoArea 1 0.70 0.08 0.08 0.13 0.61Area 2 0.57 0.01 0.38 0.04 9.50Area 3 0.61 0.02 0.29 0.07 4.14
Cr association with LSC cations - surface
No trends in Cr distribution as a function of depth detected by the AES.
No Sr-presence was found on the surface of the bond layer across the entire
depth of LSC.
Interconnect/LSC interface
LSC Middle region
LSC/Anode interface
0
10
20
30
40
50
60
70
80
90
100
0 1 2 3
CrCoLa
Cat
ion
frac
tion,
at-%
Spatial scale for the LSC dissociation: Site-specific chemical and structural characterization
TEM sample B,Near the interconnect
interfaceTEM sample A,Near the anode
interfaceFIB-prepared TEM
membraneAnode Bond Layer, LSC
Focused Ion Beam (FIB) to selectively choose the interfaces of interest for high resolution structural and chemical studies :–
LSC-interconnect, LSC-Anode, Anode-SSZ Electrolyte
LSC Anod
e
SEM view of a TEM membrane being prepared using FIB
TEM membrane
High resolution image of a FIB-
prepared TEM
sample
Ga-ion damage was unavoidable
Dissociation of the LSC at the nanoscale
LSC dissociation evident even at the nanoscale.
Regions rich in Cr have higher La and lower Co.
Low Sr content, even in bulk.
1.2μm
CrCo
La Sr
600nm
CrCo
La Sr
Structural dissociation of LSC bond layer: Secondary phases
Raman spectra, showing secondary phases
with low conductivity
Phase Conductivity (S/cm)
La0.6
Sr0.4
CoO3
at 8000C 1.6x103
[9]Co3
O4
at 8000C 3.9x101
[4]Cr2
O3
at 10000C 1.0x10-3
[6]LaCrO3
at 8000C 3.4x10-1
[7]
UsedReference
Dissociation of the LSC at the nanoscale
Element
Bond Layer, Atomic % (TEM/EDX)
Near the anode Near the interconnect
Cr-rich region
Co-rich region
Cr-rich region
Co-rich region
La 54.92 22.45 50.00 23.08
Sr 2.81 2.85 4.16 3.84
Co 15.49 63.63 12.50 57.60
Cr 27.03 10.20 33.33 15.38
Co/Cr 0.57 6.24 0.38 3.74
La/Cr 2.03 2.20 1.50 1.50
? La2 CrO6, Co3 O4 ? ? La2 CrO6 , LaCrO3Cr2 O3, Co3 O4 ?
Transport of Sr and Co to LSC top
Sr-rich
Co-rich
Cr diffusing into LSCSr and Co transported to the LSC top –
due to Cr, or electronic/ionic current?
41%
>95%,17%.<5%,74%
Cr
O=
e-
SSZ
LSCCoSr
Anode
LSC dissociation via long-range transport of cations?
Mechanism for the LSC dissociation with local variations in La/Co ratio, and transport of Sr
and Co several tens of microns, to the top.
LSC bond
Anode 1
Anode 2
SSZElectrolyte
Cr distribution in SOFC
SOFC [Carter et al. ACS 2008]
Cr
O= e-
SSZ
LSC
CoSr
Anode
Cation distribution in SOEC
Hypotheses: 1) Electronic or ionic current drives the Sr
and Co cations
out of the LSC
structure to the interconnect interface.2) Cr-driven thermodynamics favors the Sr-
and Co-
rich phases near the
Cr-containing interconnects.
Reference half-cell electrochemistry w/o Cr
Degradation in the absence of Cr
Anode Electrolyte Reference Electrodes
CathodeA = 0.5cm2
0.2
0.25
0.3
0.35
0.4
0.45
0 100 200 300
Time (hours)
Pote
ntia
l (V
)
T = 830oCI = 400mA/cm2
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Z_real (ohm)
Z_i
mg
(ohm
)
0.10 Hz
10 Hz
100 Hz
1000 Hz
0.10 Hz5 Hz3500 Hz
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Z_real (ohm)
Z_i
mg
(ohm
)
100 Hz1000 Hz
10000 Hz
0.10 Hz10 Hz
750 Hz
30 Hz
O2
O2
Anode Cathode
Worse degradation in anode compared to cathode.
With Au mesh contacts
Chemical changes in LSC in the reference cells
No chemical differences detectable by the AES between the thermal reference cell, electrochemical reference cell, and the regions in each electrode.
20000 X 10.0 keV
23
1.0 µm
LSC Anode SSZ
20μm
No delaminationin the microstructure
Therm
Echem
Driving force for the Cr-reaction here?Electrochemical reduction of Cr-vapor phases, CrO3 or CrO2(OH) to Cr2O3, and block active sites at the cathode, in SOFC [Hilpert et al., 1996], in competition with O2 reduction under similar cathodic polarization potentials.A non-electrochemical process, kinetically limited by the nucleation reaction between the CrO3 or CrO2(OH) and nucleation agents [Jiang et
al., 2006].
62323
33
332
CrOLaOLaCrOLaCrOLaCrO
:ONCr
CrOOCr
x
(g)(oxide)
→+→+
−−
→
LSC
La2 O3
CrO3
LSC
La2 O3La2 CrO6
Here, LaO-
and SrO-species due to surface segregates?
Bond layer degradation mechanism differ from those proposed for SOFCs?
SOFC:
Formation of an
oxide scale at the bond layer/ interconnect interface is responsible [Yang et al. JPS 2006]. Oxide scale conductance depends on the bond layer composition.
SOEC:
Dissociation of LSC, favoring the Sr-
and Co-
rich phases near the
Cr-containing interconnects.Cr-driven chemical reaction thermodynamics and kinetics in electrolytic conditions.
LSC bond
Anode 1
Anode 2
SSZElectrolyte
Cr distribution in SOFC
SOFC [Carter et al. ACS 2008]
Cr
O= e-
SSZ
LSC
CoSr
Anode
Cation distribution in SOEC
Summary
LSC bond layer dissociation at the nanoscale, local variations in secondary phases, evidenced in the La/Co ratio.
Sr and Co have transported several tens of microns, to the top.
–
Likely driven by Cr-related thermodynamics and kinetics in electrolytic conditions.
Cr distributed throughout the bond layer, in various phases, not only as Cr2O3 blocking species.
Degradation rate and/or mechanism in SOEC, with and without Cr-interconnect, differ from that in SOFC.
Future needs
A combination of applied and fundamental studies to consistently quantify the causes and effects of degradation mechanisms.
–
In situ structural, chemical, and electrochemical characterization of interfaces and bulk, to correlate the evolving state to its reasons
X-ray scattering and spectroscopy, high resolution site-specific microscopy and chemical analyses, and surface chemistry.
–
Use of model material configurations, in addition to the conventional cells, to isolate the complicated degradation mechanisms.
–
Theoretical studies to understand the degradation mechanisms at a more fundamental level, to predict durable electrode compositions.
Phase stability, cation mobility, at the atomistic scale.
2323
AcknowledgementsAcknowledgements
Idaho National Laboratory, NHI and NGNP programs for Idaho National Laboratory, NHI and NGNP programs for financially supporting our work.financially supporting our work.
CeramatecCeramatec Inc. for providing the samples for Inc. for providing the samples for characterization in this work.characterization in this work.
Drs. S. Herring (INL), S. Elangovan, J. Hartvigsen Drs. S. Herring (INL), S. Elangovan, J. Hartvigsen ((CeramatecCeramatec Inc.), and D. Carter (ANL) for technical Inc.), and D. Carter (ANL) for technical discussions and feedback.discussions and feedback.
Center for Materials Science and Engineering at MIT, and Center for Materials Science and Engineering at MIT, and Center for Nanoscale Systems at Harvard University were Center for Nanoscale Systems at Harvard University were used in part for AES, FIB, and TEM/EDX.used in part for AES, FIB, and TEM/EDX.
References for SOEC degradation1.
O’
Brien et al, Nuclear Technology 2006, S. Elangovan et al, personal commn. 20082.
A. Grosjean, O. Sanseau, V. Radmilovic, A. Thorel, “Reactivity and diffusion between LSM and ZrO2 interfaces in SOFC cores by TEM analyses on FIB samples”, Solid State Ionics 177 (2006) 1977 –
1980 3.
P. Salvador, S. Wang, “Investigations of Cr contamination in SOFC cathodes using TEM”, FY 2007 Annual Report, Office of Fossil energy Fuel Cell Program 59 –
634.
Sakamato, Yoshinaka, Hirato
and Yamaguchi, “Fabrication, Mechanical Properties, and Electrical Conductivity of Co3O4 Ceramics”, Journal of American Ceramic Society, 80 [1] 267-68 (1997)
5.
Hu, Li, Huang and Chen, “Improve the electrochemical performances of Cr2O3 anode for lithium ion batteries”, Solid State Ionics 177 (2006) 2791 –
2799 6.
Atkinson, Levy, Roche and Rudkin, “Defect properties of Ti-doped Cr2O3”, Solid State Ionics 177 (2006) 1767 –
1770 7.
Ong, Wu, Liu and Jiang, “Optimization of electrical conductivity of LaCrO3 through doping: A combined study of molecular modeling and experiment”, Applied Physics Letters 90, 044109 (2007)
8.
Email communication with Ceramatec9.
P. Hjalmarsson, M. Soggard, A. Hagen and M. Mogensen, “Structural properties and electrochemical performance of strontium-
and nickel-substituted lanthanum cobaltite”, Solid State Ionics 179 (2008) 636-646
10.
Y. Zhen, A. Tok, F. Boey
et al., “Development of Cr-tolerant cathodes of solid oxide fuel cells”, Electrochemical and Solid State Letters 11 (3) B42-B46 (2008)
11.
J. Mayer, L. A. Gianuzzi, T. Kamino
and J, Michael, “TEM sample preparation and FIB-
induced damage”, MRS Bulletin 32 (May 2007) 400-407
12.
Z. Yang, G. Xia, P. Singh and J. W. Stevenson, “Electrical contacts between cathodes and metallic interconnects in solid oxide fuel cells”, J. Power Sources 155 (2006) 246-252
13.
D. Carter et al, Chemical Sciences and Engineering Division, Argonne National Laboratory