The PEM Fuel Cells
Nafion Thermal Behavior
Nafion 115
80oC
o
130 C
Cell Potential / V
1.0
Silicon Oxide/Aciplex 1004
130 oC
0.8
0.6
0.4
0.2
0.0
0
200
400
600
800
1000 1200 1400 1600 1800
Current Density / mA cm2
Frick Laboratory, Princeton University
Is the Metal Oxide Phase Water Retentive?
§The composite typically contains 3-6 wt% metal oxide.
§TGA indicates the same water content and dehydration
temperature for pure Nafion and the composite.
§The conductivity of the composite measured in a
mechanically unconstrained environment is the same or
slightly worse than the conductivity of pure Nafion.
 The metal oxide is not simply providing a water
retentive or hydrated interface.
Frick Laboratory, Princeton University
If it’s not a question of direct dehydration,
then what is occurring?
• First, we will seek a molecular picture.
• Then, we will attempt to make connections between our
understanding of the molecular structure and bulk materials
properties.
Frick Laboratory, Princeton University
Effect of Metal Oxide Identity on Membrane Performance
1.0
Cell Potential / V
130o C (Degussa-Huls)
TiO2; 21nm; 50 2 m /g (R - 0.18)
0.8
2
SiO2; 20nm; 90 m/g (R - 0.21)
2
Al2O3; 13 nm; 100 m /g (R - 0.76)
Recast Nafion Control (R - 0.5)
0.6
0.4
0.2
0
500
1000
1500
-2
Current Density / mA cm
Frick Laboratory, Princeton University
Interfacial Chemistry is Critical
1.0
TiO2 (AA)/Recast Nafion; 130˚C
Cell Potential / V
unmodified (R - 0.50)
silylated (R - 0.29)
H2SO4, HNO
, "degreased" (R - 0.25)
3
0.8
0.6
0.4
0.2
0
200
400
600
800
1000
1200
1400
1600
Current Density / mA-2cm
Frick Laboratory, Princeton University
The Effect of Relative Humidity on Recast Nafion
C ontrol m em brane(H 2 -O 2 )
0
1.0
130 C 30 psig
100% R H
88% R H
75% R H
0.9
cell potential(V )
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
200
400
600
800
1000
1200
1400
1600
2
current density (m A /cm )
Frick Laboratory, Princeton University
75% Relative Humidity
0
130 C 75% R H
C ontrol R ecast
D egussa H uls SiO 2
Alfa Aesar SiO 2
D egussa H uls TiO 2
Alfa Aesar TiO 2
1.0
0.9
cell potential(V)
0.8
0.7
0.6
125µ Film
0.5
3 atm pressure
0.4
0.3
40 ml/min
0.2
0.1
0
200
400
600
800
1000
1200
1400
2
current density (m A/cm )
Frick Laboratory, Princeton University
Potential Chemical Interactions
----CF2-CF2--OH OH
Metal
Oxide
-----CF----SO3-
HO
Metal Oxide
-----CF----O=S=O
O
OH
Ti
Metal Oxide
Frick Laboratory, Princeton University
Temperature Programmed Decomposition (TGMS) of Nafion 117
m/z 18 H2O
MS abundance / a.u.
Thermal decomposition of Nafion
m/z 131 C3F5
2nd step
C3F5+
m/z 64 SO2
m/z 47 COF
CFO
+
0
-
Weight loss/ %
-10
-20
H2O
-30
H2O
-40
SO2
SO2
C3F5+
CFO+
-50
1st step
H2O
-60
-70
0
100
200
300
400
500
Temperature / C
Frick Laboratory, Princeton University
TG-MS Profile of Nafion/TiO2 Composite Membranes
m/z 18 H2O
MS abundance / a.u.
Thermal decomposition of Nafion
3rd step
m/z 131 C3F5
C3F5+
m/z 64 SO2
2nd
step
CFO
+
m/z 47 COF
0
Weight loss/ %
-10
-
-20
HO
-30
H2O
-40
SO2
-50
C3F5+
CFO+
TiO2
1st step
SO2
H2O
-60
-70
0
100
200
300
400
500
Temperature / C
Frick Laboratory, Princeton University
TPD-MS profiles of Nafion/Inorganic
composite membranes
SO2 (m/z 64)
CFO (m/z 47)
Nafion + 3% SiO2
Nafion117
Nafion + 3% TiO2
Nafion + 3% Al2O3
Nafion + 3% SiO2
Intensity / a.u.
Intensity / a.u.
Nafion117
Nafion + 3% TiO2
Nafion + 3% ZrO2
200
250
300
350
400
Nafion + 3% Al2O3
Nafion + 3% ZrO2
200
250
Temperature / C
Nafion117
Nafion + 3% TiO2
Nafion + 3% Al2O3
Nafion + 3% SiO2
Temperature / C
350
400
Intensity / a.u.
Intensity / a.u.
Nafion + 3% SiO2
Nafion + 3% ZrO2
300
400
C3F5 (m/z 131)
Nafion117
Nafion + 3% TiO2
250
350
Temperature / C
H2O (m/z 18)
200
300
-
Nafion + 3% Al2O3
Nafion + 3% ZrO2
200
250
300
350
400
Frick
Laboratory, Princeton University
Temperature / C
Molecular Model
MOx
•Crosslinking controls the
mechanical properties of the
polymer
•Glass transition temperature
•Bulk rigidity – better water
retention under stress load
Frick Laboratory, Princeton University
Dependence of Nafion Glass Transition on Metal Oxide
Frick Laboratory, Princeton University
SAXS Studies
Frick Laboratory, Princeton University
Order-Disorder Transition
Crystalline
Heat
Self Assembled
Disordered
Frick Laboratory, Princeton University
Membrane Mechanical Properties
Affect Cell Response
Ionic inclusions swell with water uptake, requiring the
membrane to push the electrodes apart.
Frick Laboratory, Princeton University
Stress-Strain Response
5.5x106
Stress (N/m 2)
5.0x106
4.5x106
4.0x106
Metal Oxide Composite
6
3.5x10
Nafion 112
3.0x106
2.5x106
2.0x106
1.5x106
1.0x106
5.0x105
0.0
0
20
40
60
80
100
120
140
160
180
200
Strain(%)
Frick Laboratory, Princeton University
Too Much of a Good Thing is Bad
Applied Pressure by
Current Collector Plates
8
7
Swelling Pressure of
Polymer Membrane
5
4
3
2
60
35
34.5
current oscillations observed
1
34
0
0
0.5
1
1.5
2
2.5
33
55
32.5
32
31.5
Turns past finger tight
31
30.5
50
20000
22000
24000
26000
28000
30
30000
Time (s)
Frick Laboratory, Princeton University
Power (mW)
33.5
Current (mA)
MEA Resistance (ohm)
6
Carbon
support
membrane
Membrane Swelling
(a)The membrane is in contact
with the catalyst support
particles.
(b)Applied pressure enhances
the membrane/catalyst contact.
(c)Additional pressure further increases
the membrane/catalyst contact.
However, the larger pressure forces
water out of the membrane.
Frick Laboratory, Princeton University
Hydrogen Crossover
Crossover Current (mA/cm 2)
950
3.0
900
2.0
850
1.0
800
0.0
125µm
40µm
40µm Composite
Open Circuit Voltage (mV)
1000
4.0
Membrane
Frick Laboratory, Princeton University
What Role Does the Metal Oxide Play?
•Increased
Tg allows maintenance of hydrated proton
conduction paths at elevated temperatures.
•Improved mechanical rigidity allows for dimensional
stability under conditions where water content of the
membrane may be changing.
•Maintains good catalyst contact on deswelling
•Eliminates water loss on swelling.
Frick Laboratory, Princeton University
Carbon Monoxide Tolerance
o
Nafion 115 - 80 C Pt Anode
w/o CO
w/100 ppm CO
1.0
Cell Potential / V
0.9
o
TiO - 130 C Pt/Ru Anode
2
w/100 ppm CO
w/500 ppm CO
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0
200
400
600
800
1000
1200
1400
-2
1600
Current Density / mA cm
Frick Laboratory, Princeton University
Summary

High Temperature Nafion Based PEM Fuel Cells overcome
several limitations associated with current cell design
Addition of a metal oxide phase affects the mechanical
properties of the membrane:


Increased Tg

Improved gas barrier

Mitigation of swelling/deswelling effects
Frick Laboratory, Princeton University
Bonus Material
(It’s not electrochemistry, but it is
interesting)
So, How Does One Store
Hydrogen on the Run?
Storage Issues


Safety
For mobile applications range & power should be maintained.
5-10Kg of H2 needed for a 65-75kW engine.
H2 feed rate is ~1000 liters/minute

Weight

Effective Density of Hydrogen

Volume Requirements



Refill



Size
Geometry
Availability
Recharge rate.
Cost
Frick Laboratory, Princeton University
Hydrogen Storage Phases
Limiting Densities of Molecular Hydrogen
Solid Hydrogen (4.2K)
Moles Hydrogen
Atom/cm3
0.088
Liquid Hydrogen (20K)
0.070
100%
Hydrogen Gas 200bar (300K)
0.016
~1%
Hydrogen Phase
W eight Pe rcent
100%
Frick Laboratory, Princeton University
Storage Options

Standard steel tanks (2000-5000psi)


Known technology.
Good Safety Record




Tanks are challenging to fill because hydrogen heats upon expansion
Heavy


Subject to hydrogen imbrittlement
Forms projectiles if structure is breached
Storage capacity is only 0.5-1% by weight
Poor volumetric storage due to non-ideality of hydrogen:
 n 2 a 
P  2 V  nb  nRT
V 

a  0.2444
b  0.02661
~ 20% volumet ric expansion at 5000psi
Frick Laboratory, Princeton University
Storage Options

Composite Tanks (~10,000psi)





High storage capacity:
Light weight
Can store 7% H2 by weight!
Does not fragment upon failure
Cost
Frick Laboratory, Princeton University
Storage Options

Generation on the fly: in-situ or ex-situ reforming
of hydrocarbon fuels using an on-site reformer.





Energy density of gasoline
Easy access to fuel (gasoline stations)
Systems integration is poor
No carbon mitigation.
Solid-state storage by intercalation (metal hydrides, carbons)




Safe
Heavy
Expensive
Chemical thermodynamics and kinetics are difficult
Significant heating is required to release the hydrogen
∆H losses up to 30% are typical with operating temperatures of 200-300C.
 Tank filling is very exothermic Chemical kinetics are a difficult to handle

Frick Laboratory, Princeton University
Hydride Storage Capacity
Metal Hydride Storage Systems
Hydrogen Phase
LaNi5H6
Moles Hydrogen
Atom/cm3
0.091
W eigh t Pe rcen t
1.37%
TiFeH2
0.10
1.89%
Mg2NiH4
0.098
3.6%
MgH2
0.11
7.6%
Frick Laboratory, Princeton University
Storage Options

Chemical Hydrides

“Hydrogen on Demand” (Sodium Borohydride)







Not flammable
High Effective hydrogen pressure (~7000psi)
Low Volume
Simple system
Chemical Safety
Recyclable
Cost??
lyst
NaBH4  Aqueous Base Cata

H2  NaBO2  2H2O
Frick Laboratory, Princeton University