Diagnostics of PEM fuel cells
N. Wagner, A. Bauder, K.A. Friedrich,
Deutsches Zentrum für Luft und Raumfahrt (DLR), Institute of Technical Thermodynamics, Pfaffenwaldring 3848, D-70569 Stuttgart, Germany
2nd International Workshop on Degradation Issues in Fuel Cells
21-23 September, 2011
Thessaloniki, Greece
Test Stand Diagnostics versus Non-traditional in-situ
Diagnostics using Physical Methods
Test stand diagnostics (methods integrated into test stands):
U(i) characteristics, OCV…
EIS on single cells and stacks
Performance degradation with time U(t); i(t) …
Cell voltage distribution Ustack= U1 + U2 + U3….
Pressure loss / Gas tightness test
Gas utilization measurement
Temperature distribution and control
Current interrupt methods
Locally resolved measurements
Physical non-traditional in-situ methods:
Optical imaging methods / CCD Camera; Microscopy
Optical spectroscopies and x-ray tomography (Outlook)
Synchrotron and neutron tomography
In-situ x-ray diffraction
Test Stand Diagnostics versus Non-traditional in-situ
Diagnostics using Physical Methods
Test stand diagnostics (methods integrated into test stands):
U(i) characteristics, OCV…
EIS on single cells and stacks
Performance degradation with time U(t); i(t) …
Cell voltage distribution Ustack= U1 + U2 + U3….
Pressure loss / Gas tightness test
Gas utilization measurement
Temperature distribution and control
Current interrupt methods
Locally resolved measurements
Physical non-traditional in-situ methods:
Optical imaging methods / CCD Camera; Microscopy
Optical spectroscopies and x-ray tomography (Outlook)
Synchrotron and neutron tomography
In-situ x-ray diffraction
Test Stand Diagnostics versus Non-traditional in-situ
Diagnostics using Physical Methods
Test stand diagnostics (methods integrated into test stands):
U(i) characteristics, OCV…
EIS on single cells and stacks
Performance degradation with time U(t); i(t) …
Cell voltage distribution Ustack= U1 + U2 + U3….
Pressure loss / Gas tightness test
Gas utilization measurement
Temperature distribution and control
Current interrupt methods
Locally resolved measurements
Physical non-traditional in-situ methods:
Optical imaging methods / CCD Camera; Microscopy
Optical spectroscopies and x-ray tomography (Outlook)
Synchrotron and neutron tomography
In-situ x-ray diffraction
Motivation
Characterization of Fuel Cells by Electrochemical Impedance
Spectroscopy:
Determination of electrode structure and reactivity, separation
of electrode structure from electrocatalytical activity
Determination of reaction mechanism and separation of
different overvoltage contributions to the fuel cell
performance loss
Determination of degradation mechanism of electrodes,
electrolyte and other fuel cell components (bipolar plates, end
plates, sealings, etc.)
Determination of optimum operation condition (e.g. gas
composition, temperature, partial pressure), cell design (flow
field) and stack design
Electrochemical Impedance Spectroscopy:
Application to Fuel Cells
Electrochemical Impedance Spectroscopy:
Application to Fuel Cells
PEFC: Schematic Diagram (cross section)
Schematic representation of the different steps and their
location during the electrochemical reactions as a function
of distance from the electrode surface
N. Wagner, K.A. Friedrich, Dynamic Response of Polymer Electrolyte Fuel Cells in „Encyclopedia of Electrochemical Power Sources“
(Ed. J. Garche et al.), ISBN-978-0-444-52093-7, Elsevier Amsterdam, Vol.2, pp. 912-930, 2009
Overview of the wide range of dynamic processes in FC
Membrane
humidification
Electric double layer
charging
Liquid water
transport
Charge transfer fuel cell
reactions
Degradation and
ageing effects
Gas diffusion processes
Changes in catalytic
properties / poisoning
Temperature
effects
microseconds
10
-6
10
-5
10
-4
milliseconds
10
-3
10
-2
10
-1
seconds
10
0
10
1
Time / s
minutes
10
2
10
3
10
4
months
days
hours
10
5
10
6
10
7
10
8
Bode representation of EIS measured at different current
densities, PEFC operated at 80°C with H 2 and O2 at 2 bar
Impedance / mΩ
Ω
50
Diffusion
Charge transfer of
ORR
Phase
Charge
transfer of
HOR
o
RM
80
30
60
20
40
O V=597 mV; i=400 mAcm-2
V=497 mV; i=530 mAcm-2
∆ V=397 mV; i=660 mAcm-2
+ V=317 mV; i=760 mAcm-2
15
20
10
0
10m
100m
1
10
100
Frequency / Hz
1K
10K
100K
Common Equivalent Circuit for Fuel Cells
Rct,c
Rct,a
RM
Cdl,c
Cdl,a
Common Equivalent Circuit for Fuel Cells
Diffusion
of O2
Zdiff
Rct,c
Rct,a
RM
Cdl,c
Cdl,a
Common Equivalent Circuit for Fuel Cells
Zdiff
Rct,c
Rct,a
RM
Cdl,c
Cdl,a
Diffusion
of H2
Zdiff
EIS at Polymer Fuel Cells (PEFC):
Contributions to the cell impedance at different current densities
0.161
1
Cell impedance
/ Ohm
Cell impedance /Ohm
10
0.121
0,1
0,01
0,001
0
0.081
22
117
236
Current density / mAcm
656
-2
0.041
Rdiffusion
Rmembrane
R anode
R cathode
0.001
0
100
200
300
400
500
Current density /mAcm
600
-2
700
EIS at Polymer Fuel Cells (PEFC):
Contributions to the overal U-i characteristic determined by EIS
1100
E0
}∆E
}∆ E
Cell voltage / mV
1000
900
C
800
700
A
600
RN
500
RK
} ∆E
RA
M
RM
400
} ∆ EDiff.
300
CN
200
0
100
Cdl,c
200
Cdl,a
300
400
500
Current density / mAcm-2
600
700
800
Current density / mA cm-2 , Cell voltage / mV
Change of cell voltage during constant load at 500 mA cm-2
170 µV/h
12 35
4
8
9
10
11
6
7
12
13
Current density
Cell voltage
Elapsed time / h
Investigation of Degradation Mechanisms in Polymer
Electrolyte Fuel cells – Durability Measurements
Impedance / mΩ
|Phase|
30
Time
0
90
75
25
60
20
RN
Rct,C
Rct,A
RM
CN
Cdl,C
15
45
10
30
7
15
Cdl,A
100m
1
3
10
100
Frequency / Hz
1K
10K
0
100K
Variation of RC during PEFC operation at 500 mA cm-2
17
RN
16
RC
RA
RM
Impedance / mOhm
15
14
CN
Cdl,c
Cdl,a
13
New start
up
after 24 h
12
11
10
Cathode
9
0
200
400
600
Elapsed time / h
800
1000
1200
Variation of RA during PEFC operation at 500 mA cm-2
10
Anode
9
8
Impedance / mOhm
7
6
5
4
RN
RC
RA
3
RM
2
New start up
after 24 h
1
Cdl,c
CN
0
0
200
400
Cdl,a
600
Elapsed time / h
800
1000
1200
Analysis of Fuel Cell Losses by Impedance
Spectroscopy
RN
RC
RA
RM
16
140
6
4
2
Cathode
100
80
60
0
Anode
68.8 mV
400
Cathode
43.2 mV
600
120
100
Anode
68.8 mV
Irreversible
voltage loss
48.6 mV
Cathode
43.2 mV
20
Irreversible voltage loss
Irreversible
voltage loss
48.6voltage
mVloss
Reversible
1200
Time / h
Anode 5.4 mV
0
Irreversible voltage loss
Reversible
voltage
loss
88.8 mV
60
40
1000
Cdl,a
80
0
800
Cd l,c
Reversible
voltage
Cathode
20 mV
loss
88.8 mV
Anode 5.4 mV
200
20
CN
140
RCathode
RAnode
40
0
Cathode
20 mV
Voltage loss / mV
10
8
New start up
After 24 h
120
12
Voltage loss / mV
Impedance / mΩ
14
Reversible voltage loss
Overall cell voltage loss
Overall cell voltage loss
Evaluation of EIS with the porous electrode model
I
I
1
Rp,a ; Rct,a ; Rpor,a /Ohm
R p ,a =
0.1
( R por ,a ⋅ Rct ,a ) 2
2 
 R
 por ,a  
tanh 
 
R
 ct ,a  
1
0.08
0.06
0.04
0.02
0
0
200
400
600
800
-2
Current density /mAcm
I
I
Porous electrode resistance (Rp, a), charge transfer
resistance (Rct, a) and electrolyte resistance (Rpor, a)
in the pore of the anode at different current densities
Evaluation of EIS with the porous electrode model
i-V characteristic and current dependency of pore electrolyte resistance
of the anode and cathode
1200
35
1000
■ Rel,por,Anode
♦ Rel,por,Kathode
30
25
800
20
600
15
400
10
200
5
0
0
0
2
4
6
8
10
Current / A
12
14
16
18
Cell voltage / mV
Pore electrolyte resistance / mOhm
40
Improved Evaluation Techniques
Real time drift compensation
Time course interpolation
Z-HIT compensation
Reforming of Methane
CH4 + 2H2O => 4H2 + CO2 (CO)
Compres.
Methane
a) Reformer-heating
Reformer
Shiftreactor HT
9% CO
3% CO
b) Cat-Burner
Residual
Gas
Shiftreactor LT
COcleaning
0,5% CO
Air (O2)
H2, CO2
<
0,005%CO
PEMFuel Cell
E-Energy
Heat
Time resolved EIS - CO poisoning of Pt-anode
Ω
Imaginary part / mΩ
600
a
700
500
600
400
b
500
300
400
200
c
300
d
e
f
Overv. CO / mV
Cell voltage / mV
800
-200
f
e
-100
c
d
b
0
a
100
100
200
0
3000
6000
9000
0
12000
200
Time / s
0
200
400
Ω
Real part / mΩ
Time progression of cell voltage and overvoltage in
galvanostatic mode of PEFC operation (217 mAcm-2)
Pt-anode , H2 + 100 ppm CO at 80°C
Nyquist plot of EIS measured at
different times during poisoning of
the Pt-anode with CO
Improved evaluation techniques
Time course interpolation
drift affected data
interpolated data
time course of the
impedance at
single frequencies
impedance
Requirements
time
frequency
• Series measurement
• Time for each
• measured frequency
AND
for each spectrum
Mitigation of EIS Problems
Reduction of the high frequency disturbances by optimization of
the wiring of the electrical sensing of stacks.
Variation of the operating conditions (gases, temperature) in
order to determine the different impedances of the electrode
processes
Modeling of the spectra by an equivalent circuit.
Development of advanced EIS equipment for high currents / high
frequencies in corporation with instrument manufacturer (e.g.
Zahner Elektrik GmbH).
Lowering Mutual Induction Errors
I
Current
Magnetic
Coupling
Zf
Zm
Zs
Potential
Current Flow
Zf
Minimizing the
magnetic coupling
Zm
Zs
Current Flow
Circuit
E
Potential Sense
Circuit
Twisted pairs of current feeding and
potential sensing lines suppress the
mutual inductance errors
Impedance Spectroscopy at fuel cell stacks
U5
Cell 5
U4
Voltage
U3
probes
Cell 4
I
Cell 3
U2
Cell 2
„Zahner IM6“
(diff. Potentiostats//
diff. Electronic Loads)
U1
Cell 1
Current
probes
AC Voltage Input Signal
U(f)=10 mV
f = 30 mHz – 500 kHz
Imax = 40 A
Experimental EIS set-up with 16 parallel channels for
stack measurements
Investigation of oscillatory fluctuations in
PEFC with the segmented cell
Motivation for using locally resolved diagnostic tools
Variation of gas concentrations along the flow field as a result of
the electrochemical reaction
The reaction leads to temperature gradients in the cell
The electrical and chemical potential distribution is based on the
local product and reactant concentrations
The gas composition has a signifcant influence on the
degradation (e.g. corrosion) of the cell
HO
2
O2
O2
H2
H2
HO
2
O2
Segmented Cells
Enables locally resolved
measurements, important for the
qualification and determination of:
Lokale M essung en
• Functionality of MEAs
• Catalysts / Kinetic (poisoning)
• Water management and balance
• Mass transport (diffusion)
4 mm
Time dependency of current density distribution
measured at 540 mV, Tcell = 70°C
Rh Cathode = 0%
Rh Anode > 100%
Flow rate H2 (Anode): 260 ml/min
Flow rate Air (Cathode) : 830 ml/min
Film
Time dependency of cell voltage at 0% r.H.
Galvanostatic mode of operation: 20 A green; 15 A red; 10 A blue
Voltage / mV
H2H2-rate: 209 ml/min; Air: 665 ml/min; Tcell: 80 °C; Humidifier : 80°
80°C; Pressure: 1.5 bar
Time / s
Comparison between asymmetrical humidification Anode/Cathode
(0% r.F. Anode and 0% r.F. Cathode)
Voltage / mV
Anode 0% r.F.
Current: 25 A
H2-flow: 261 ml/min
Air: 830 ml/min
Tcell: 70 °C
Cathode 0% r.F.
Time / s
Flow Field and Ion Power MEA Performance
1.0
Cell Voltage / V
0.8
0.6
T = 70 °C
P = 1.5 bar
λH2 = 1.4
λAir = 2
0.4
0.2
0.0
0.0
0.2
0.4
0.6
0.8
-2
Current Density / A cm
Configuration A (co-flow):
Configuration B (counter flow):
Anode: input G1 output A7
Cathode: input G1 output A7
Anode: input A1 output G7
Cathode: input G1 output A7
1.0
1.2
Oscillations of PEM fuel cells at low cathode humidification
Cathode with r.h.=0%, Anode with nominal r.h.=152%
(condensing conditions)
Cell Voltage:
500mV;
H2=180ml/min
Air=560ml/min
Co-flow
Current Density Distribution of Two Operation
Points
Impedance spectra of
two operation points
High Current Density (H)
0.8
Z' / mΩ
0.6
-15
500 Hz
-10
599 mV - 3.199 A
0.4
Low Current Density (L)
0.2
-5
599 mV - 17.088 A
500 Hz
0
R=5.3 mOhm
R=28 mOhm
1.2 kHz
30 kHz
5
0.0
0
50
100
150
200
250
10
Time / s
5
10
15
20
25
30
Ucell: 500mV, Cathode 660ml min-1, Anode 200 ml min-1; Tcell: 70°C; counter
flow
35
Z'' / mΩ
Current Density / A cm
-2
-20
Interpretation of the Oscillation Cycle
Formation of liquid water
High ionic conductivity
Liquid water resevoir at anode High electro-osmotic drag
LVP
and water permeation
Rehydration Anode
~ 20 s
Low electro-osmotic drag
and low water permeation
Low ionic conductivity
Anode without liquid water
H2 + H2 O
H2 + H2O
Dehydration anode
αdrag≈2.7
H+
H+
αdrag< 1
Anode without liquid water
High ionic resistance
O2 + N2
O 2 + N2 + H2 O
In situ Radiography
Helmholtz-Zentrum Berlin
Bessy ll
Zentrum für Sonnenenergie
und Wasserstoff-Forschung
In situ investigation of water
management behavior of
homemade DLR-MPL at
cathode side (+SGL 25BA),
anode with SGL 25BC, CCM
from Gore
Zelle 699
Zelle 700
Current Density [A/cm²]
0,80
0,70
1
galvanostatic
operation
Voltage [V]
0,60
0,8
0,50
0,6
0,40
0,30
0,4
0,20
0,2
0,10
0
0
200
400
600
800
1000
1200
0,00
1400
Time [s]
Radiography image
MPL thickness
33µm
80µm
Current Density [A/cm²]
1,2
Zelle 699
Zelle 700
Current Density [A/cm²]
0,80
0,70
1
galvanostatic
operation
Voltage [V]
0,60
0,8
0,50
0,6
0,40
0,30
0,4
0,20
0,2
0,10
0
0
200
400
600
800
1000
1200
0,00
1400
Time [s]
Radiography image
MPL thickness
33µm
80µm
Current Density [A/cm²]
1,2
Zelle 699
Zelle 700
Current Density [A/cm²]
0,80
0,70
1
galvanostatic
operation
Voltage [V]
0,60
0,8
0,50
0,6
0,40
0,30
0,4
0,20
0,2
0,10
0
0
200
400
600
800
1000
1200
0,00
1400
Time [s]
Radiography image
MPL thickness
33µm
80µm
Current Density [A/cm²]
1,2
Conclusion
Diagnostic Methods are used
To understand the processes in fuel cells in great detail
- Influence of the operation conditions on the conditions inside the
fuel cell
- Understanding of the degradation processes
- Inhibition of degradation processes
To support the development and for improvement of fuel cells
- Detection of problems of components
- Targeted development
To control fuel cell systems
- Optimized operation conditions
- Increasing of life-time
- Early error detection, increased reliability
Thank You for the Attention!