Production of Pure Hydrogen from a Source of Waste
sing Solid O
ide Membrane Electrolyzer
Electrol er
and Steam using
Oxide
Soobhankar Pati
Kyung Joong Yoon
Srikanth Gopalan
Uday B. Pal
Materials Science & Engineering
ECS 215
ECS 215th Conference, San Francisco
May 24 May 24 ‐‐ 30th, 2009
OUTLINE OF THE PRESENTATION
y Hydrogen as an energy carrier ?
y Conventional Solid Oxide Steam Electrolyzers (SOSE)
y Novel Solid Oxide Membrane (SOM) Electrolyzer
y Experiment
y Process model for the SOM Electrolyzer
y Evaluation of experimentally obtained data using the process model
d l
y Some ongoing experiments to improve the efficiency
ELECTROLYSIS: CONVENTIONAL SOLID OXIDE STEAM ELECTROLYZER (SOSE)
Air (O2 , N2) O2 (g) 2eO + O Æ O2(g)
Porous anode
O2‐ Æ O + 2e‐
V
Electrolyte
H2O(g) + 2e‐ Æ H2(g) + O2‐
Porous cathode
H2O (g) H (g) 2 2e-
Partial pressure of oxygen in air (pO2)air > Partial pressure of oxygen in H2-H2O (pO2)H2O/H2
60 – 70% of the energy = THERMODYNAMIC BARRIER
Natural gas*
THERMODYNAMIC BARRIER
REDUCTANT AT THE ANODE
Coal
Hydrocarbon
waste
t
PROBLEM: Conventional SOSE is not equipped to use reductants (waste, coal, etc.)
*J. Martinez‐Frias, Ai‐Quoc Pham and S. M. Aceves: Int. J. of Hydrogen Energy, 2003, vol.28, pp 483‐90
CONFIGURATION: SOLID OXIDE MEMBRANE (SOM)ELECTROLYZER Reductant (R)
(Coal, Hydrocarbon waste)
RO(g)
Syn gas Syn‐gas 2e-
Liquid metal anode
(Sn1, Ag2, Cu3)
[O]LMA + R Æ
R Æ RO(g)
RO( )
O2‐ Æ [O]LMA + 2e‐
V
O2‐
Electrolyte
(YSZ)
H2O(g) + 2e‐ Æ H2(g) + O2‐
Porous cathode
(Ni‐YSZ, Cu‐YSZ)
H2 (g) 2e-
H2O(g) 1 , 2
Charge transfer at the YSZ/Anode interface
LIQUID METAL ANODE
3
1
2
3
Reaction interface for oxidation of reductant by the [O]
y
[ ]LMA
T. Ramanarayanan and R.A. Rapp: Metall. Trans. B,3, 3239 (1972)
T. H. Etsell and S. N. Flengas, Metall. Trans. B, 2, 2829 (1971)
A. Krishnan, U. Pal and X. Lu: Metall. Trans. B, 36,463 (2005)
ADVANTAGES : SOLID OXIDE MEMBRANE (SOM) ELECTROLYZER
1.2
200
0.9
H2O (g)
( ) = H2(g)
( ) + 1/2 O2(g)
( )
100
06
0.6
o
ΔE (Voltss)
ΔGo (kJ / mol)
SOSE
0.3
Operating
Temperature
0
0.0
SOM
-0.3
H2O (g) + C = H2(g) + CO (g)
-100
400
600
800
1000
1200
o
Temperature ( C )
If h d
If hydrocarbon waste (HC) is used , b (HC) i d H2O(g) + (HC) Æ H2(g) + H2O(g) + CO(g)
‐ Electrochemical conversion of H2O(g) High purity H2
‐ Efficient way of converting energy value in waste
EXPERIMENTAL SET UP: ELECTROCHEMICAL PERFORMANCE Ni-YSZ
S cathode
~ 90 m
~ 10% porous
YSZ electrolyte
~ 2 mm
Operating Temperature : 1000
O
ti T
t 1000oC
ELECTROCHEMICAL CHARACTERIZATION AND PERFORMANCE
ELECTROCHEMICAL CHARACTERIZATION: OPEN CIRCUIT POTENTIAL
Open circuit potential (Eeq)*,**
Cathodic gas composition
:
poO2(a)
[C/CO(g) equilibrium]
3 % H 2O - H 2
10 % H2O - H2
20 % H2O - H2
40 % H2O - H2
Measured Values
(V)
-0.050
3.999 × 10-19
- 0.120
-0.154
- 0.207
‰ A negative open circuit potential (OCP) indicates the process is spontaneous
‰ OCP of SOSE (0.89 V) >> OCP SOM electrolyzer (-0.207 V) : 40%H2O – H2
* J. Martinez‐Frias, Ai‐Quoc Pham and S. M. Aceves: Int. J. of Hydrogen Energy, 2003, vol.28, pp 483‐90.
** P. Soral, U. Pal, H. Larson and B. Schroeder: Metall. Trans. B, 1999, vol.30, pp 307‐21.
ELECTROCHEMICAL CHARCTERIZATION: POTENTIODYNAMIC SCAN
5
40% H2O
2
4
Current (A))
C
10% H2O
0.3
3% H2O
3
0.2
2
1
0
-0.5
Mass transfer limited
0.0
0.5
1.0
1.5
Curren
nt Density (A/cm )
04
0.4
20% H2O
2.0
2.5
3.0
0.1
3.5
Potential (V)
‰ Current density increased with increase in steam content in the cathodic gas feed
‰ Mass transfer limitation was observed only at 3% H2O content
~ Unavailability of H2O(g) at the TPBs
ELECTROCHEMICAL PERFORMANCE: POLARIZATION MODEL DEVELOPMENT ƒ Open circuit potential (Eeq)
ƒ Ohmic Polarization (
:
= i Rohm
ohm) : ohm
ƒ Activation Polarization (
act)
Rohm : YSZ , Electrodes, Contact, Lead wire
(ne=2, α=1/2)
:
(Anode + Cathode)
ƒ Concentration Polarization (
conc)
:
Cathodic concentration polarization
Anodic concentration polarization
ELECTROCHEMICAL PERFORMANCE: CURVE FITTING
Open Ohmic loss
Circuit
Potential
Measured
Activation polarization
Cathodic conc. polarization
i0 : Fitting parameter
Impedance spectroscopy
Anodic conc. polarization
: Literature value* kc (Mass transfer coeff.
of dissolved oxygen yg
(Eff i bi
(Effective binary diffusivity)
diff i i )
in the liquid tin anode)
: Fitting parameter
: Fitting parameter
Maximum of 2 fitting parameters used for the curve fitting
* T. Ramanarayanan and R.A. Rapp: Metall. Trans. B, 1972, vol.3, pp 3239‐46
Appropriate assumption
ELECTROCHEMICAL PERFORMANCE: IMPEDANCE SPECTROSCOPY
2
Zreal(Ω-cm )
5
10
15
20
30
35
40
3% H2O
1.6
4 X 10-2 Hz
10% H2O
16
12
1 Hz
0.8
8
0.4
4
4000 Hz
0.0
0.5
1.0
1.5
2.0
Zreal(Ω)
2
40% H2O
- Zimag(Ω-cm )
20% H2O
1.2
- Zimag(Ω)
25
0
2.5
3.0
3.5
4.0
‰ High frequency intercept : Ohmic resistance* : Independent of steam content**
g
q
y
p
p
‰ Overlap of charge transfer resistance and diffusional (Warburg) impedance at higher
frequencies *** * J.R. Macdonald : Impedance Spectroscopy, John Wiley, New York,NY,1987
** M. A. Laguna‐Bercero, S. J. Skinner and J. A. Kilner, J. of Power Sources, doi:10.1016/j.jpowsour.2008.12.139,(2009)
*** S. Britten and U. Pal, Metall. and Mat. Transactions B, 31, 733 (2000)
ELECTROCHEMICAL PERFORMANCE: CURVE FITTING (40% H2O in cathodic gas)
Open O
Ohmic loss
Oh i l
Circuit
Potential
Activation polarization
A i i l i i
i0 : Fitting parameter
Measured
Cathodic conc. polarization
h d
l
Insignificant
Anodic conc. polarization
A di l i i
kc: Fitting parameter
Measured (Imp. Spectra)
3.5
A3.0
3SSUMPTION
0
: Cathodic concentration polarization (
conc, c
) is negligible*
) is negligible
Potentiial (V)
2.5
‐ Mass transfer limitation not observed ‐ SOM electrolyzer is electrolyte supported with thin Ni‐ YSZ cathode 2.0
1.5
1.0
0.5
0.0
-0.5
0
1
2
3
4
Fitting parameters
Curve fitting results
Exchange current (i0 )
1.00A
Mass transfer
f coefficient
ff
( kc)
( k
0.00056 cm/sec
5
Current (A)
* M. Ni, M.K.H. Leung and D.Y.C. Leung: Int. J of Hyd. Energy, 2007, vol. 32, pp 2305‐13
ELECTROCHEMICAL PERFORMANCE: CURVE FITTING (3% H2O in cathodic gas)
Open Ohmic loss
Circuit
Potential
Activation polarization
Measured
Cathodic conc. polarization
Cathodic conc polarization
kc : Independent 0f cathodic gas comp.*
i0 : Fitting parameter
: Fitting parameter
Measured (Imp. Spectra)
kc : 40 % curve fitting (0.00056 cm/sec
3% H2O
2.0
Anodic conc. polarization
Anodic conc polarization
Curve Fitted
Pote
ential (V)
1.5
1.0
0.5
0.0
Fitting parameters
Curve fitting results
Exchange current (i0 )
0.12 A
Effective Binary Diffusivity ( DeffH2‐H2O ) 0.025 cm2/sec
0.0
0.3
0.6
0.9
1.2
1.5
1.8
Current (A)
*S. Yuan, K.C. Chou, U. Pal: J. Elec. Soc.,1994, vol. 141, pp. 467‐74
ELECTROCHEMICAL PERFORMANCE: CURVE FITTING ( 10 % and 20 % H2O)
Open Ohmic loss
Circuit
Potential
p
Activation polarization
p
Cathodic conc. polarization
kc : 40 % curve fitting
(0.00056 cm/sec i0 : Fitting parameter
Measured
p
Anodic conc. polarization
:Independent of cathodic gas comp.*
Measured (Imp. Spectra)
3
10 % H2O
:3% curve fitting (0.025 cm2/sec)
20 % H2O
Pote
ential (V)
2
Curve Fitted
1
0
0
1
2
3
Cathode gas composition
Exchange current (i0)
10 % H2O – H2
0.26 A
20 % H2O – H2
0.68 A
4
Current (A)
R. B. Bird, W. E. Stewart and E. N. Lightfoot: Transport Phenomena, 2nd ed., John Wiley & Sons, New York, 2002
ELECTROCHEMICAL PERFORMANCE : ACTIVATION POLARIZATION
Cathode g
gas
Exchange current (i0)
composition
3 % H2O – H2
0.12 A
10 % H2O – H2
0.26 A
20 % H2O – H2
0.68 A
40 % H2O – H2
1.00 A
3% H2O
Ac
ctivation pola
arization (V)
0.4
10% H2O
20% H2O
0.3
H2O content
40% H2O
=> Surface coverage of H
S f f H2O (g) at TPB’s O ( ) TPB’ 02
0.2
0.1
0.0
0
1
2
3
Current (A)
4
5
Provides additional sites for the charge transfer reaction
ELECTRO. PERFORM.: CATHODIC CONC. POLARIZATION (40% H2O)
is composition independent 3.5
Total polarization
Cathodic conentration polarization
Polarrization (V)
3.0
25
2.5
2.0
1.5
Cathodic concentration polarization 1.0
0.5
1% of the total overpotential
0.0
-0.5
0
1
2
3
4
5
Current (A)
Cathodic concentration polarization is insignificant at 40% H2O
: Consistent with our initial assumption
ELECTRO. PERFORM.: IMPEDANCE SPECTROSCOPY IN SUPPORT OF POLARIZATION MODELING
Cathodic gas composition :40 % H2O
Equivalent circuit describing the SOM electrolyzer
2
Zreal (Ω-cm )
0
5
10
15
20
25
30
20
Cathodic gas composition 40 % H2O
O-H
H2
Curve fitted
1.5
16
-2
-Zimag (Ω)
1.0
0
12
Rohm = 0.585 Ω
Rct = 0.061 Ω
8
1 Hz
0.5
10 Hz
4000 Hz
4
0
0.0
0.5
1.0
1.5
2.0
2.5
Zreal (Ω)
At small activation polarization
At small activation polarization,
Polarization modeling
CNLS Fitting
Oh i Loss
Ohmic
L
(Rohm)
0 58Ω
0.58Ω
0 585Ω
0.585Ω
Charge transfer resistance (Rct,c)
0.054Ω
0.061 Ω
2
CNLS Fitting Results:
-Zimag(Ω-c
cm )
4.0 X 10 Hz
ELECTROCHEMICAL PERFORMANCE: VARIOUS POLARIZATION LOSSES
3% Steam content
40% Steam content
3.5
2.0
1.5
Conc. polarization(Cathode)
Activation polarization
1.0
Ohmic loss (Electrodes,
Contact, Current collector))
05
0.5
Total Polarizatio
T
on (V)
Conc. polarization(Anode)
Tottal Polarization
n (V)
Conc. polarization
( Anode)
Activation pol.
3.0
2.5
Ohmic loss
( Electrode,
Current collector)
2.0
1.5
1.0
Ohmic loss
( YSZ electrolyte)
0.5
Ohmic (YSZ electrolyte)
0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
0.0
0
1
2
3
4
Current (A)
Current (A)
™ At 3% H2O the ohmic part ≈ Overpotential due to electrode processes ™ At 40% H2O ohmic resistance is 80.5% of the total polarization p
Electrodes, Contacts, Current collector 52 % of the ohmic loss
YSZ electrolyte resistance 48 % of the ohmic loss
Ohmic loss due to the dominates the performance loss:
‐ Molybdenum current collectors on the anodic side
‐ SOM electrolyzer design ( Electrode supported) 5
SUMMARY
‰ The potential of hydrogen production from steam using a solid oxide membrane electrolyzer with a liquid anode was demonstrated.
l t l
ith li id d d
t t d
‰ Thermodynamic barrier was lowered using a reductant in the liquid metal anode.
‰ U
Using an electrolyte supported design and applying only 2.0
i l t l t t d d i d l i l 2 0 V , a current density of t d it f 2
0.5 A/cm was achieved.
‰ Polarization modeling results thus showed that the performance is rate‐controlled by
the ohmic loss. h h i l
‰ Experimental study and modeling will form the basis for redesigning the SOM
electrolyzer to improve its efficiency and for investigating various types of waste feed.