Solid-oxide fuel cells (SOFC) with

hydrocarbon and hydrocarbon-derived fuels

Colorado School of Mines

California Institute of Technology

Robert J. Kee and Huayang Zhu

Engineering Division, Colorado School of Mines
Golden, CO 80401, USA
David G. Goodwin
Engineering and Applied Science, California Institute of Technology
Pasadena, CA 91125, USA

[email protected]
(303) 273-3379

Presented:
International Symposium on Combustion
July 29, 2004

30_ICS_7/04.p1

Direct-oxidation fuel cells offer a number of

potentially important benefits
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p2

There is important fuel-cell physics and chemistry

spanning a great range of scales
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p3

There is an extensive and growing literature in

fuel cells generally and SOFC particularly
Colorado School of Mines

California Institute of Technology

General SOFC resources

The Fuel Cell Handbook, US Dept. of Energy, 6th Ed.,
DOE/NETL 2002/1179, available on the web, (2002).
Singhal and Kendall, High-temperature SOFC, Elsevier, (2003).
Minh and Takahashi, Science and Technology of Ceramic Fuel Cells,
Elsevier, (1995).
Seminal research on hydrocarbon-fueled SOFC
E. Perry Murray, T. Tsai, and S.A. Barnett, A direct-methane fuel cell
with a ceria-based anode, Nature, 400:651-659 (1999)
S. Park, J.M. Vohs, and R.J. Gorte, Direct oxidation of hydrocarbons
in a solid-oxide fuel cell, Nature, 404: 265-266 (2000)
Broad areas of research and literature
Materials (electrolyte, electrodes, catalysts, interconnects, seals,)
Membrane-electrode assemblies
!Fuel cell characterization and performance
Fundamental electrochemical processes
Systems and hybrid cycles

30_ICS_7/04.p4

Solid-Oxide Fuel Cells enable direct electrochemical oxidation (DECO) of hydrocarbons

Colorado School of Mines

California Institute of Technology

Planar architecture
Membrane-electrode assembly
Interconnect carries current
Stacked layers build voltage
Electrolyte (e.g., YSZ)
Polycrystalline ceramic
O2- ion conductor
Electrical insulator
Impervious to gas flow
Electrodes
Porous cermet composite
Anode supports MEA

Characteristic dimension: 1 mm
Channel diameter
Anode thickness

Interconnect
!High electrical conductivity
Maintains equal potential

Characteristic temperature: 700C

Flow channels
Formed in interconnect

30_ICS_7/04.p5

Fuel-cell operation depends on coupled

macroscopic and microscopic processes
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p6

There are similarities and differences

between PEM and SOFC technology
Colorado School of Mines

California Institute of Technology

Proton Exchange Membrane (PEM)

Polymer electrolyte
Proton-conducting electrolyte
Low temperature (~100C)
Requires H2 fuel
CO is a poison
Needs good reforming/separation
Precious-metal catalysts
Low thermal inertia
Fairly mature

PEM

Solid Oxide Fuel Cell (SOFC)

Ceramic electrolyte
Oxygen-ion conductor
High temperature (~700C)
Can use hydrocarbon fuel
CO is not a poison
May use reforming or CPOX
Inexpensive catalysts
Large thermal inertia
In development

SOFC

30_ICS_7/04.p7

Experiments to characterize and evaluate new material

systems are often done with small button cells
Colorado School of Mines

California Institute of Technology

From the laboratory of Prof. Sossina Haile, Caltech

30_ICS_7/04.p8

Direct electrochemical oxidation has been

demonstrated for a variety of fuels
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p9

A practical SOFC needs extensive three-phase

boundaries to facilitate charge transfer
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p10

There is a long and expanding set of materials that are

being developed for fuel-cell applications
Colorado School of Mines

California Institute of Technology

Interconnect
!Metals (Fe-Cr alloys)
Ceramics (e.g., LaCrO3)
Electrolyte
Yttria-stabilized zirconia, YSZ
Samarium doped ceria, SDC
Gadolinium doped ceria, GDC
Cathode
Mixed ionic-electronic conductors
- Doped (e.g., Sr) LaMnO3 , LSM
- La1-xSrxCo1-yFeyO 3-!, LSCF
Cermets
Anode
Cermets (e.g., Ni-YSZ, Ni-SDC)
Ceria-copper

30_ICS_7/04.p11

Membrane-electrode assemblies (MEA) can be

fabricated by tape-casting or extrusion technologies
Colorado School of Mines

California Institute of Technology

MEA fabricated by ITN energy Systems

30_ICS_7/04.p12

At equilibrium the electric potentials are related to the

chemical potentials of the fuel and oxidizer streams
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p13

Reversible cell potential is determined entirely

from thermodynamics of the gas streams
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p14

The electrical potential of the cell depends on

the fuel type and fuel dilution
Colorado School of Mines

California Institute of Technology

Assuming direct oxidation

C4H10 + (13/2) O2 -> 5 H2O + 4 CO2
CH4 + 2 O2 -> 2 H2O + CO2
H2 + (1/2) O2 -> H2O

30_ICS_7/04.p15

Irreversible losses, fuel depletion, and fuel utilization

serve to reduce fuel-cell efficiency
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p16

Cell performance depends on the chemical

potential of fuel and oxidizer and on internal losses
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p17

The ohmic overpotential through the electrolyte

depends greatly on temperature
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p18

Activation overpotentials represent the chemical

energy needed to overcome charge-transfer barriers
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p19

There is great uncertainty in the details of elementary

charge-transfer chemical reaction mechanisms
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p20

The electric-potential bias affects the

activation barriers to charge-transfer
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p21

The Butler-Volmer form of the charge-transfer rates

can be derived from mass-action concepts
Colorado School of Mines

California Institute of Technology

Butler-Volmer reaction orders are different from elementary orders

!Electric-potential difference versus Overpotential
30_ICS_7/04.p22

Estimates of adsorption, desorption, and diffusion

assist estimating the extent of the three-phase regions
Colorado School of Mines

California Institute of Technology

Most adsorbed H does not participate in charge transfer

Benefits to making nano-dispersed electrodes
30_ICS_7/04.p23

To what extent are parallel chargeexchange pathways coupled ?

Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p24

Pattern anodes can expose chemical processes

that are obscured from observation in a fuel cell
Colorado School of Mines

California Institute of Technology

Pattern anode SEM images courtesy

of Profs. G. Jackson, R. Walker, and
B. Eichhorn, University of Maryland

30_ICS_7/04.p25

Microfabrication technology is used to

make well defined three-phase boundaries
Colorado School of Mines

California Institute of Technology

From the laboratory of

Profs. G. Jackson, R. Walker, and B. Eichhorn (University of Maryland)
30_ICS_7/04.p26

Patterned-anode models assist interpreting

experimental electrochemical observations
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p27

Current-dependent losses (overpotentials) reduce

the actual cell voltage during operation
Colorado School of Mines

California Institute of Technology

Concentration overpotentials
Species transport through electrodes
Activation overpotentials
Electrochemistry at TPB
Ohmic overpotential
Resistive losses in electrolyte
Interface overpotentials
Resistance at materials interfaces
Leakage overpotential
Electrical conduction in electrolyte
30_ICS_7/04.p28

The Dusty-Gas model considers that the meanfree path length is comparable to the pore size
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p29

What is the role of the anode structure in

promoting reforming, shifting, and CPOX?
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p30

The models are capable of handling elementary

heterogeneous reforming/CPOX chemistry
Colorado School of Mines

California Institute of Technology

Reaction mechanism from Prof. Olaf Deutschmann, University of Karlsruhe

30_ICS_7/04.p31

MEA models do a good job of representing

measured electrochemical performance
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p32

Channel models couple flow, porous

transport, chemistry, and electrochemistry
Colorado School of Mines

California Institute of Technology

Channel flow
Plug flow is usually appropriate
Mass exchange with anode

Electrochemistry
Charge transfer rates
Ion transport

Anode transport and chemistry

Gas and surface species
Porous-media transport
Elementary surface chemistry

Finite-volume discretization
Computational solution

30_ICS_7/04.p33

With appropriate catalysts, there can be

significant reforming within the anodes
Colorado School of Mines

California Institute of Technology

Anode lnlet: 97% CH4, 3% H2O, 12.5 cm/s, 800C, 1 atm.

Cathode: Air
30_ICS_7/04.p34

The local voltage-current characteristics of the MEA

vary along the channel because of fuel dilution
Colorado School of Mines

California Institute of Technology

Anode lnlet: 97% CH4, 3% H2O, 12.5 cm/s, 800C, 1 atm.

Cathode: Air
30_ICS_7/04.p35

Homogeneous chemistry models predict

fuel conversion and deposits
Colorado School of Mines

California Institute of Technology

From the laboratory of Prof. A.M. Dean (Colorado School of Mines)

30_ICS_7/04.p36

Homogeneous fuel pyrolysis can lead to

the formation of polyaromatic deposits
Colorado School of Mines

California Institute of Technology

Initial steps lead to production of smaller unsaturated species

- Results from hydrogen-abstraction, "-scission sequences
Molecular weight growth (deposit formation) occurs at later times
- As unsaturates concentrations increase, radical-addition
reactions become more important
- As resonantly-stabilized radical concentrations increase,
recombination reactions become important

30_ICS_7/04.p37

Upstream fuel processing can help to

reduce coking and fouling
Colorado School of Mines

California Institute of Technology

CPOX and/or Reform

30_ICS_7/04.p38

Improved fuel-cell function

Increase H2 content
Reduce coking

Recently reported results provide

strong evidence for internal CPOX
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p39

Small SOFC systems can operate with

clean kerosene fuel
Colorado School of Mines

California Institute of Technology

20 Watt system designed and built by ITN Energy Systems, Inc.

and Mesoscopic Devices, Inc.
30_ICS_7/04.p40

There are several tubular configurations

that offer alternatives to planar systems
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p41

Hybrid cycles can offer overall system

efficiencies of over 70%
Colorado School of Mines

California Institute of Technology

30_ICS_7/04.p42

Biomass gasification and SOFC can be

integrated into a hybrid system
Colorado School of Mines

California Institute of Technology

System layout adapted from

Dr. Nikhil Patel, EERC,
Univ.of North Dakota

30_ICS_7/04.p43

There are a great many research opportunities

Colorado School of Mines

California Institute of Technology

Elementary electrochemical kinetics mechanisms and formalism

Charge-transfer processes and kinetics
Transport by surface diffusion, including field effects
Local three-phase boundary geometry
Selective heterogeneous and catalytic reactions
!Thermal and electrochemical competitions (e.g., H on surface)
Coking propensity and deposit formation
Influence of catalysts, including functionally graded systems
Influence of temperature, residence time, fuel mixes, etc.
In-situ fuel-altering processes in channels and porous electrodes
Partial oxidation and reforming
!Functionally graded materials
Hybrid systems
Heat engines or Combined heat and power
!Biofuels
Many materials issues

30_ICS_7/04.p44

Acknowledgements
Colorado School of Mines

California Institute of Technology

Office of Naval Research MURI

!CSM (Kee, Lusk, and Dean)
Caltech (Goodwin, Haile, and Goddard)
University of Maryland (Jackson, Walker, and Eichhorn)
DARPA (Palm Power)
ITN Energy Systems (Barker, Sullivan, and Thoen)
Review and Comments on Manuscript
Olaf Deutschmann (Karlsruhe)
!Greg Jackson (Maryland)
!Tony Dean (CSM)

30_ICS_7/04.p45