functionally graded cathodes for … library/events/2004/seca/georgia-tech...functionally graded...

Functionally Graded ElectrodesFunctionally Graded Electrodes

Functionally Graded Cathodes for SOFCsFunctionally Graded Cathodes for SOFCsFunctionally Graded Cathodes for SOFCs

Project Manager: Dr. Lane WilsonDOE National Energy Technology Laboratory

E. Koep, Q. Wu, H. Abernathy, J. Dong, Y. Liu & M. LiuCenter for Innovative Fuel Cell and Battery Technologies

School of Materials Science and EngineeringGeorgia Institute of Technology

Collaborators: T. Orlando and H. ChenChemistry and Biochemistry, GT

May 11 – 13, 2004


OutlineOutlineOutline• Technical Issues Addressed

• Objectives & Approach

• Phase I Accomplishments

• Recent Results (Phase II)– Further versification and interpretation of IR data– DFT Calculations and ESD– In-situ Raman and TERS– Patterned Electrodes with some sites selectively blocked– Measurement of Local (micro or nano-scale) properties – Fabrication of Graded Electrodes

• Applicability to SECA

• Activities for the next 6-12 Months


Critical Factor: Interfacial ResistanceCritical Factor: Interfacial ResistanceCritical Factor: Interfacial Resistance

400 450 500 550 6000

2

4

6

8

10

Res

ista

nce,

Ω

cm2

Temperature, °C

30 µm Electrolyte

Interfacial Resistance

Performance is determined by Rpat low temperatures

Cathode

30 µm Electrolyte

Anode


Origin of RP for a Porous MIEC ElectrodeOrigin of ROrigin of RPP for a Porous MIEC Electrodefor a Porous MIEC Electrode

Ions & e’ (current flow path)

Mass Transport

gas through pores

O2

Electrolyte

ReactionZones:

TPBs & MIEC Surfaces

The Concept of FGE

Macro-porous structureLarge pores for fast TransportHigh Electronic ConductivityCompatible with Interconnect

Inter-Mixed LayerProduce Turbulence Flow

Nano-porous structureHighly Catalytic ActiveCompatible with electrolyte


Critical IssuesCritical IssuesCritical Issues

• Intrinsic Properties of MIEC Cathodes– Fundamental processes at the surfaces?– Effect of surface defects/Nano-struture?– Effect of ionic and electronic transport?– In-situ characterization tools and predictive models?

• Effect of Microstructure/Architecture– Surface area/reaction sites– Rapid gas transport through pores– Predictive models for design of better electrodes

• Fabrication of FGE with desired microstructure and composition


ObjectivesObjectivesObjectives

• To develop tools for in-situ characterization of electrode reactions in SOFCs;

• To gain a profound understanding of the elemental processes occurring at cathode-electrolyte interfaces; and

• To rationally design and fabricate efficient cathodes for low temperature operation to make SOFC technology economically competitive.


Technical ApproachTechnical ApproachTechnical Approach

Modeling• Transport in Porous Media

• Active Reaction Sites• Reaction Pathways

• Mechanism

Patterned Electrodes• Reaction Pathway

• Active sites

In-situ Characterization• FTIR/Raman, TERS• Micro- or nano-IS

• GC/MS

Fabrication of FGE• Optimal Microstructure• Graded in Composition

• Cost-effective/Reproducible

SOFC Performance • High Performance

• Long-Term Stability


Major Phase I Accomplishments

• Successfully fabricated patterned electrodes (SSC and LSM) with well-defined geometries for determination of TPB width

• Studied SOFC cathode materials using in-situ FTIR emission spectroscopy under practical conditions, including Pt, SSC, LSC, LSF, and LSCF

• Characterized surface structures of SOFC materials using Raman spectroscopy

• Fabricated electrodes with vastly different microstructures and morphologies using combustion CVD and template synthesis

• Demonstrated functionally graded cathodes of low polarization resistances at low temperatures.


Recent Progress – Phase IIRecent Progress Recent Progress –– Phase IIPhase II

• Further versification and interpretation of in-situ vibrational spectra of cathode materials (thickness and time dependence)

• DFT calculation of vibration frequencies of surface oxygen species (adsorbed O2, O2

-, and O22-) – peak assignment

• Started study of electron-stimulated desorption (ESD) of oxygen• Dramatically increased sensitivity and spatial resolution in probing

surfaces using tip enhanced Raman scattering (TERS)• Successfully fabricated patterned electrodes for specific site

isolation• Set up for local (micro or nano-scale) electrochemical

measurements (IS) using SPM tips and patterned electrodes• Developed processes for fabrication of electrodes graded in both

composition and microstructure


Typical IR Spectra: Bulk and Kinetic PropertiesTypical IR Spectra: Bulk and Kinetic Typical IR Spectra: Bulk and Kinetic PropertiesProperties

C O 2

∆E

/Eo(%

)

4000 3500 3000 2500 2000 1500 1000

0

4

8

1 2

1 6

2 0

2 4

930

1236

A t o ve rp o te n tia l o f -0 .7 6 V

W ave N u m b er , cm -1

1124

A t O C V

B aselin e O ffse t (b u lk p ro p erties)

S u rface S p ec ies1 1 2 4 cm -1:

1 2 3 6 cm -1: p ertu rb ed 9 3 0 cm -1:

−2O

−22O

−2O

P ea k a ssig n m en ts:


FITR spectra of SSC film at different O2 partial pressureFITR spectra of SSC film at different O2 partial pressure

6 0 01 6 0 02 6 0 03 6 0 0

W a v e n u m b e r , c m -1

∆E/

E, %

0.5%

100%

Gas switching from N2 to a gas containing Po2 (0.5 to 100%)

Both negative-going baseline shift and positive-going IR peak increase with the oxygen partial pressure.


IR peak & Baseline shift ~ O2 partial pressuresIR peak & Baseline shift ~ OIR peak & Baseline shift ~ O22 partial pressurespartial pressures

The saturated oxygen partial pressure for oxygen adsorption is about 20%.

Corresponding to the IR peak height, the baseline shift at 825 cm-1 is also saturated when the O2 is about 20%.

0.2

0.22

0.24

0.26

0.28

0.3

0.32

0.34

0.36

0 20 40 60 80 100 120

Oxygen partial pressure (%)

Peak

hei

ght ∆

E/E

-3.3

-2.8

-2.3

-1.8

-1.3

-0.8

0 20 40 60 80 100 120

Oxygen partial pressure (%)

Base

line

shift

∆E/

E

Baseline shift and the IR peaks seem to go together!Lower O2 pressure means higher oxygen vacancy in the bulk. Negative-going baseline shift therefore means decrease in bulk oxygen vacancy concentration


Validity of Interpretation?Validity of Interpretation?Validity of Interpretation?

• How do we know that the peaks are corresponding to species adsorbed on surfaces, not changes in bulk properties?

• If so, can we prove that the peak assignments are correct?

• How do we know that the baseline shift is due really to changes in bulk properties, not in surface properties ?


Thickness DependenceThicknessThickness DependenceDependence

Active bulk < skin depthInactive bulk

Adsorbed moleculesEmitted IR

• If the peaks are corresponding to species adsorbed on surface, the heights of the peaks should be independent of film thickness.

• If the baseline shift is due really to changes in bulk properties, it should increase with thickness (until the thickness is greater than the skin depth).


Equilibrium FITR spectra for SSC thin film electrodesEquilibrium FITR spectra for SSC thin film electrodes

650165026503650

W avenumber, cm -1

∆E/

E, %

2.0µm

15 µm

2.5µm

3.0µm

3.5µm

5.0µm6.0µm

7.5µm

8.0µm

10 µm

• A broad oxide adsorbent peak centered at 1124 cm-1

• A negative-goingbaseline shift with increase thickness.


IR peak height and baseline shiftIR peak height and baseline shiftIR peak height and baseline shift

The O2- peak height is almost

independent of film thickness.

00.050.1

0.150.2

0.250.3

0.350.4

0.450.5

1 3 5 7 9 11 13 15Thickness (µ m)

∆E/

E

-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

1 3 5 7 9 11 13 15Thickness (µm)

∆E/

E

(a)

(c)

(b)

The baseline shifts at (a) 3020, (b) 2128, and (c) 825 cm-1 increase with film thickness.

Larger baseline shift is observed at lower wavenumber (or longer wavelength with larger skin depth).


Time Dependence of pd-FTIR ESTimeTime Dependence of pdDependence of pd--FTIR ESFTIR ES

OO-

OO-

PO2

O2−

Heating CartridgeDC

Voltage

• If the peaks are corresponding to species adsorbed on surface and the baseline shift is due really to changes in bulk properties, the rate of peak height change (rate of surface reaction) should be different from that of baseline shift (the rate of bulk transport) unless the rates of the two processes are coincidently identical.


Rapid scan: Time resolved pd-FTIR on LSCRapid scan: Time resolved pdRapid scan: Time resolved pd--FTIR on LSCFTIR on LSC

-0.35

-0.3

-0.25

-0.2

-0.15

-0.1

-0.05

0

0.05

1 6 11 16 21 26 31 36 41 46 51 56 61

Time (Sec.)

Peac

khe

ight

-1

0

1

2

3

4

5

6

Bas

elin

e

600 Peak height

650 Peak height

700 Peak height

600 Baseline

650 Baseline

700 Baseline

Peak height change is much faster than baseline shift, indicating again that the peak and the baseline shift correspond to different processes.

Further, it’s clear that surface reaction is faster than bulk diffusion.

Electrode kinetics

Bulk diffusivity


Conclusions for FTIR Studies Conclusions for FTIR Studies Conclusions for FTIR Studies

• Indeed, the IR peaks correspond to species adsorbed on surfaces or a change in surface properties;

• The IR baseline shift is due to a change in bulk properties of electrode material, most likely to the change in oxygen vacancy concentration in the bulk phase;

• However, we are still unable to conclude which peak corresponds to which species.- Theoretical Calculation- Isotope Exchange


DFT Calculation of O-O vibrationsDFT Calculation of ODFT Calculation of O--O vibrationsO vibrations

• Used DFT method (B3LYP) on Q-Chem software with 6-311+G(3df) basis set

• Calculated theoretical frequency of O-O bond for free O2, O2

-, and O22-

83511731592Frequency, cm-1

O22-O2

-O2Species


Effect of transition metal cationEffect of transition metal Effect of transition metal cationcation

• Began calculations of O-O vibrational frequency when bonded to cobalt cation in SSC or LSC

• Optimized geometry first, then calculated frequency

Co

O

On+

OptimizeCo

O

On+

orCo

O O

n+

Start Finish


Most Probable Surface Configuration

For (101) phase

Oxygen vacancy Adsorbed Oxygen


Effect of transition metal cationEffect of transition metal Effect of transition metal cationcation

DNCDNC70.5141.5∠Co-O-O, °

N/AN/A4671487O-O vibration, cm-1

DNCDNC1.20741.3100O-O bond, Å

CoO24+CoO2

3+CoO22+CoO2

+Species

*DNC denotes that geometry optimization calculation did not converge

• Calculations reveal superoxo-/peroxo- nature of oxygen species

• Frequencies unreliable – must include more of the lattice for better results


Electron Stimulated Desorption (ESD)

e-

Auger ElectronE-beam

1 2

X-ay

Ionization Auger relaxation


3-Step ESD Model

e-

Excitation~ 10-15 sec

Nuclear Separation~ 10-13 sec

Desorption~ 10-11 sec

Electron excites the target via an inelastic scatteringNuclear motion on the excited state potential surfaceOutgoing atom or molecule interacts with surface


What Can ESD Offer?What Can ESD Offer?What Can ESD Offer?

• Structures of adsorbate/adsorbent system

– Local bonding– Bonding geometry– Binding energy

• Dynamics of charge transfer

• Site specific desorption

– Defects (for example: oxygen vacancy)– Local disorder


QMS Sample Transfer Stage

XYZ-Rotation Stage

Sample Holder

Turbo

Loading Cell

E-Gun


XYZ-Rotation Stage

Sample Holder

Turbo

Loading Cell

E-Gun


XYZ-Rotation Stage

Sample Holder

Turbo

Loading Cell

E-Gun

ESD Experimental Set-Up

Prof. Thom OrlandoChemistry, GT


GDC Sample for ESD StudyGDC Sample for ESD Study

Compostition :Ce0.9Gd0.1O1.95 (gadolinia doped ceria, GDC )

Preparation: Combustion of metal nitrateCalcine in air at 600ºC for 2 hCold press into pelletsFired at 1450ºC for 5 h

Characterization:Relative density: 95% to 97%XRD: fluorite structureSEM: grain size about 5 µm


Electron Energy Dependence of ESD

0 20 40 60 80 100

O+ O+ threshold energy is 22 eVrelevant to the direct ionization of O2s level followed by an intra-atomic Auger cascade

The feature around 50 eV is related to the Auger process that involves the excitation of 5s level of cerium and gadolinium

Ready to study cathode materials (e.g, SSC)

Electron Energy (eV)


FTIR – Isotope DosingFTIR FTIR –– Isotope DosingIsotope Dosing

• Difficulties

— Precise control of Po2

— Fast oxygen exchange at high temperature

• Approach – FTIR in UHV

— Diffuse-reflectance IR (DRIFTS) set up for low temperature FTIR measurement

— Emission IR set up for high temperature FTIR measurement


High-Temp Emission IR SpectroscopyHighHigh--Temp Emission IR SpectroscopyTemp Emission IR Spectroscopy

1P

2F

6.02”6”3”

1F

Type of of mirrorsF: flat mirrorP: parabolic mirror

FTIR

Sequence of mirrors


Low-Temp (LN) DRIFTSLowLow--Temp (LN) DRIFTSTemp (LN) DRIFTS

FTIRA

B

B

C

A - plain mirrorB - parabolic mirror F=69, 100, 153, 180, 220, 250, 300, 400, 418 mmC - parabolic mirror F= 43 mm

DETECTOR

Vac. Chamber

3P

2P

1F

6.02”6”3”

6.02

”

4P

Detector

1.70”3”

FTIR

3P

2P

1F

6.02”6”3”

6.02

”

4P

Detector

1.70”3”

FTIR


Other Tools for in-Situ StudiesOther Tools for inOther Tools for in--Situ StudiesSitu Studies

• Probing and Mapping Surface Reactions using Raman Spectro-microscopy

• Tip Enhanced Raman Spectroscopy (TERS)

– Combination of Raman and SPM

• Patterned Electrodes for Isolation of Reaction Sites• Local (micro or nano-scale) measurement using SPM

tips and patterned electrodes; TERS, Raman mapping


Raman Spectra of LSC in a Controlled AtmosphereRaman Spectra of LSC in a Controlled AtmosphereRaman Spectra of LSC in a Controlled Atmosphere

200 700 1200 1700Wavenumber (cm-1)

Inte

nsity

(a.u

.)

O2 RT-H2H2 RT-H2H2 RTO2 RTAr RT


Raman Spectra of Cobalt Compounds in AirRaman Spectra of Cobalt Compounds in AirRaman Spectra of Cobalt Compounds in Air

200 700 1200 1700

Wavenumber (cm-1)

Inte

nsity

(a.u

.)

SC

SSC

LSCF

LSC

855 11401545

• 3 types of oxygen species are observed

• But the peaks are week


Tip-Enhanced Raman Scattering (TERS)

Nano-sized Au, Ag or Cu tip

Sample

Objective lens

e-e-

Illumination

SERS

K. SHIBAMOTO, et al., ANALYTICAL SCIENCES 2001, VOL.17 SUPPLEMENT

SERS Mechanism


Our Integrated Raman and SPMOur Integrated Raman and SPMOur Integrated Raman and SPM

Objective

Scanners


Schematic Arrangement for TERSSchematic Arrangement for TERSSchematic Arrangement for TERS

45o

•Tip with Ag coating and small diameter (50 nm) for TERS•Tuning fork technology removing the interruptions of other laser source

Unique Capabilities:Spatial Resolution: up to 50 nmVery Sensitive to the Chemical Nature of Surfaces


Capabilities of Our SystemCapabilities of Our SystemCapabilities of Our System

TERS

AFM ImageRaman Mapping

To be Obtained


TERS: Single Crystal SiliconTERS: Single Crystal SiliconTERS: Single Crystal Silicon

0 200 400 600 800 1000

0

500

1000

1500

2000

2500

3000

3500

Inte

nsity

Raman shift (cm-1)

with tip enhancement without tip enhancement (X10)


TERS: Oxygen species Adsorbed on LSMTERS: Oxygen species Adsorbed on LSMTERS: Oxygen species Adsorbed on LSM

0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200

0

500

1000

1500

2000

2500

3000

3500

Without enhancement

With enhancement

852

Inte

nsity

Raman shift (cm-1)

Peroxide species (O22-) on LSM surface was not observable with ordinary Raman or FTIR, but

observable with TERS. Mechanism on LSM is different from that on SSC or LSC


TERS: Carbon deposited on YSZTERS: Carbon deposited on YSZTERS: Carbon deposited on YSZ

1000 1200 1400 1600 1800 20000

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

YSZ

Without enhancement

With enhancement

D GIn

tens

ity

Raman shift (cm-1)

Preliminary results: The Raman intensities of the G and D mode are dramatically enhanced by the AFM tip.


TERS – A Sensitive Surface ProbeTERS TERS –– A Sensitive Surface ProbeA Sensitive Surface Probe

• Dramatically enhanced sensitivity to surface species (up to 108)

• Increased spatial resolution (up to 20 nm) for mapping of active sites for electrode reactions

• Unlike IR, TERS is not influenced by gas phase H2O and CO2 and especially sensitive to carbon and sulfur, making it a powerful tool for investigation of fuel reforming and anodes of SOFCs running on hydrocarbon fuels (e.g., gasified coal)


Patterned ElectrodesPatterned ElectrodesPatterned Electrodes

• Patterned electrodes with the length of TPB varying in 4 orders of magnitudes

• Patterned electrodes with some reaction sites selectively blocked

• Combination of patterned electrodes and SPM tip for performing local (micro- to nano-scale) electrochemical measurements


Possible Reaction SitesPossible Reaction SitesPossible Reaction Sites

O2

Path A.

Electrolyte

(a)

Metallic or MIECElectrode

e-

Electrolyte

Metallic or MIECElectrodee-

(b)

Os

Path B.

Metallic or MIECElectrode

e-

(c)Electrolyte

O2

Path D.

Electrolyte

MIECElectrodee-

O2

(d)


A Photolithographic Process for Isolation of Reaction SitesA Photolithographic Process for Isolation of Reaction SitesA Photolithographic Process for Isolation of Reaction Sites

YSZ electrolyte

PhotoresistPhotoresist Undercut

PMGI PMGI50µm

Photoresist

Ti-coating

YSZ electrolyte

LSM patterned electrode

Ti coating

PMGIPMGI

LSM electrode

0.26 µm LSMTi-coating

YSZ electrolyte

LSM patterned electrode 0.26µm

YSZ electrolyte


YSZ electrolyte



Typical MicrographsTypical MicrographsTypical Micrographs

YSZ electrolyte

LSM microelectrode

TiO2coating

TPB

Edge of TiO2

µm

YSZ electrolyteLSM Electrode

TiO2

0.26 µm x 50 µm

3.0 µmYSZ Electrolyte


Proof-of-Concept: Effect of Site BlockingProofProof--ofof--Concept: Effect of Site BlockingConcept: Effect of Site Blocking

0

0.005

0.01

0.015

0.02

0.025

0.03

500 600 700 800

Temperature (ºC)

1/R

p

Without Ti coating

With Ti Coating

- 1 0 0

- 8 0

- 6 0

- 4 0

- 2 0

00 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0

With TiO2

Without TiO2

• The surface of the patterned (0.26x50µm) LSM electrode contributes significantly to oxygen reduction.

• The micro-fabrication technique is effective in isolation of different reaction sites.


Local Electrochemical Measurements

Impedance Analyzer

SPMController

Patterned Electrode

Impedance Analyzer

SPMController

Patterned Electrode

To probe local properties using impedance spectroscopy performed on an SPM tip and patterned electrodes with micro- or nano-scale spatial resolution


Impedance of an Individual Grain/GB/TPB

1

2

Patterned Electrode

1

2

Patterned Electrode

Impedance Analyzer


Graded Electrodes Prepared by Combustion CVD

GDC

70 wt.% LSC+30 wt.% GDC

LSC

YSZYSZ-substrate

GDC-CCVD50wt%LSC-50wt%GDC-CCVD

LSC-powder spray


Applicability to SOFC CommercializationApplicability to SOFC CommercializationApplicability to SOFC Commercialization

• Generated some basic understanding of electrode reaction mechanisms

• Developed new tools for in-situdetermination of electrode properties under practical conditions

• Rational design of efficient electrodes


Activities for the Next 12 MonthsActivities for the Next 12 MonthsActivities for the Next 12 Months

1) Oxygen isotrope exchange experiments in UHV chamber for study of detailed reaction mechanisms using vibrational spectroscopy

• FTRI, pd-FTIR, Rapid Scan FTIR

• TERS and Raman Mapping

2) Refined calculations of vibrational spectra for different cathode materials to assist interpretation of FTIR and Raman data for different electrodes

• Oxygen reaction mechanism and kinetic parameters

• Bulk properties such as vacancy concentration and transport properties

3) Local measurements using SPM tips and patterned electrodes to probe local properties under in-situ conditions

4) Optimize Templated Synthesis and Combustion CVD for Fabrication of FGEs


AcknowledgementAcknowledgementAcknowledgement

Lane Wilson, NETL/DoE

SECA Core Technology ProgramDept of Energy/National Energy Tech Laboratory

Equipment Partially funded by DURIP/AROCenter for Innovative Fuel Cell and Battery Technologies,

Georgia Tech

functionally graded cathodes for … library/events/2004/seca/georgia-tech...functionally graded...

Documents