Professional Documents
Culture Documents
Available at www.sciencedirect.com
Article history: After 500 h accelerated lifetime test on an automotive PEM (proton exchange membrane)
Received 10 June 2009 fuel cell stack, the performance of the cells at the stack’s foreside is the worst. The elec-
Received in revised form trochemical surface area, internal resistance, particle size and hydrophobic nature of
6 July 2009 catalyst are measured based on the electrochemical methods, electronic-lens and dropped
Accepted 22 September 2009 water test. The results show that the internal resistance increases to double and the
Available online 25 October 2009 average particle diameter increases to 3 times when electrochemically active area surface
coefficient decreases to one-fifth. The open circle voltage of the worst cell is 0.7 V and its
Keywords: maximum current density is 200 mA/cm2. Especially, the performance of part at air-outlet
PEM Fuel cell is the worst, followed by the part at air-inlet, hydrogen-outlet, and hydrogen-inlet. The
Performance degradation phenomenon mentioned above at the inlet of stack is the most serious among all the cells
Electrochemical in the stack.
Test ª 2009 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved.
Life
1. Introduction 2. Experimental
The automotive fuel cell’s lifetime is only around 2000 h at 2.1. Performance degradation in the stack
present. There is a wide gap to the minimum requirement
of 5000 hours for the automotive commercial purpose [1]. After 500 hours accelerated lifetime test on an automotive
Performance degradation of fuel cell is closely related to PEM fuel cell stack with conditions of idling cycle, load
catalyst agglomeration, loss on membrane electrode changing cycle, high power cycle, and start-stop cycle, the
assembly (MEA) [2–5], catalyst poisoning [6,7] and MEA performance is deteriorate obviously. As shown in Fig. 1, the
contaminated [6–8]. The start-stop process, idling and average cell voltage curve of every cell from No.1 to No.14 cell
frequent load changing are the most important reasons is lower than that of No.15 to No.100 cell.
affecting the performance of PEMFC. Starvation of gas and
water flooding occur frequently in PEMFC used in vehicle 2.2. Preparation of MEA for experiment
[9–15]. Considering the different location on flow field in
a cell and in different cells, electrochemical methods, TEM MEAs from No.1 to No.14 cell have been taken out of the 100
(transmission electron microscopy) and drop water method cells stack. Three representative cells, No.1 cell, No.4 cell and
are used to test and analyze performance degradation of No.8 cell, are chosen for destructive test. As shown in Fig. 2,
different fuel cells and different flow field positions. five small sections have been taken out from each MEA for
Fig. 3 – Polarization curve of small cells in different part of primary stack cells. (a) No.1 cell. (b) No.4 cell. (c) No.8 cell.
electrode at a cell temperature of 60 C and gas pressure of (Fig. 3 (c)) is better than the new cell. The open circuit potentials
1.05 MPa. Impedance of new MEA cell is about 120 mU cm2. is more than 0.85 V. The polarization curves of them drop gently
in the ohmic potential loss area. Their maximum current
density is more than 1000 mA/cm2, which is the highest of all.
Some open circuit potentials of small cells are lower than
3. Results and discussion 0.75 V, and membranes of which may have been damaged
after working more than 500 hours.
3.1. Polarization curve
The small cells tests are performed on the PEMFC testing 3.2. Electrochemically active surface area, particle size,
platform, the temperature of which is 60 C, and the absolute hydrophobic nature and impedance
pressure of fully hydrated hydrogen and air is 1.05 MPa with
the flow rate of 10 mL/min passed through anode and 200 mL/ Electrochemically active surface area coefficient is defined as:
min passed through cathode in cell. The Fig. 3 expresses the Z Z
1 SIV
polarization curve of small cells of different parts taken from Qh ¼ Idt ¼ IdV ¼ (1)
m m
different primary cells.
As shown in the Fig. 3, the small cells taken from air-outlet
Qh SIV
parts of primary cells have the worst performance of all parts. Ac ¼ ¼ (2)
Qm Ageometric mQm Ageometric
They also have lowest open circuit potentials of all. The polari-
zation curves of them drop fast in the ohmic potential loss area, Qh (/C) stands for total quantity of electricity in the adsorbed
and the maximum current density is less than 300 mA/cm2. process, SIV (/W) is given from Cyclic voltammograms curve,
The small cells taken from hydrogen-outlet parts of primary m (mV/s) stands for voltage scan rate, Qm (/C/cm2) stands for
cells are also performed poorly, sometimes even worse than the atomic electric charge adsorbed smooth catalyzer’s surface,
air-outlet part such as that in No.4 cell (Fig. 3 (b)). The small cells Ageometric (/cm2) stands for the geometric area of a MEA.
taken from air-inlet, middle parts of primary have the average Fig. 4 (a) shows the electrochemically active surface area
performance of all. coefficients of different parts in every cell. Electrochemically
The small cells taken from hydrogen-inlet parts of primary active surface area coefficients of hydrogen-inlet parts are the
cells have the best performance, especially which of No.8 cell biggest of all parts, which shows that the catalyzers in these
3150 international journal of hydrogen energy 35 (2010) 3147–3151
Fig. 4 – Results comparison of all parts in every cell. (a) Electrochemically active surface area coefficient. (b) Average particle
diameter values. (c) Contact angle values. (d) Impedance values.
areas are the most active of all after working long time. At the Fig. 4 (b) compares the particle diameter size among different
same time, electrochemically active surface area coefficients position of three representative foreside cells MEA in the 100
of hydrogen-outlet, especially and hydrogen-outlet parts are cells stack, which worked more than 500 hours. As shown in this
low, catalyzers on some parts have little activity. Primary figure, after working more than 500 hours, the cathode particles
characters are acquired by coefficient analysis and compare are highly agglomerated compared with that of new. Average
for this method has a commonly exactitude. Along air flowing particle diameter and standard deviation of particle diameter
from inlet to outlet in an actual single cell, the electrochemi- are defined to describe the degree of particle agglomeration and
cally active areas decrease gradually, which certainly much distribution extension on diameter values as:
less than that of a new cell. Along air flowing in primary Pn
Si X
m
distributive channel from No.1 cell to No.8 cell in an actual S¼ i¼1
¼ Pi Si (3)
stack, the electrochemically active areas increase gradually n i¼1
centralized. At the same time, the more board extension those references
of air-outlet part distribute, the more proportions assemble on
big diameter values. As shown in Table 1, standard deviations
of particle diameter of air-outlet part are bigger than air-inlet, [1] DOE and DOD of USA Freedom CAR-Fuel Cell Technologies
and of course are bigger than which of new MEA. Roadmap, August 10, 2005.
Contact angle is gained in the experiments referred [2] Michael W, et al. Incorporation of voltage degradation into
before. If the contact angle is bigger than 90 degree, MEA a generalised steady state electrochemical model for a PEM
fuel cell. J Power Sources 2002;106:274–83.
surface is hydrophobic, and in other case, MEA surface is
[3] Luo Z, Li D, et al. Degradation behavior of membrane–
hydrophilic. O. Chapuis, M. Prat’s studies show that, the electrode-assembly materials in 10-cell PEMFC stack. Int J
hydrophobic nature of MEA surface has no difference when Hydrogen Energy 2006;31:1831–7.
the Contact angle is from 120 to 140 degree [17]. As shown in [4] Collier Amanda, Wang Haijiang, Yuan Xiao Zi, Zhang Jiujun,
Fig. 4 (c), the contact angles in different parts of every cell Wilkinson David P. Degradation of polymer electrolyte
are from 124 to 130 degree, which makes clear that the membranes. Int J Hydrogen Energy 2006;31:1838–54.
[5] Gülzow E, et al. Dry layer preparation and characterisation of
hydrophobic nature of any part in MEA surface has no
polymer electrolyte fuel cell components. J Power Sources
obvious change after working for more than 500 hours. The
2000;86:352–62.
drainage nature of cell is steady, which shows that proba- [6] Ahn SY, Shin SJ, Ha HY, et al. Performance and lifetime
bility of flooding in the flow channel of old cell is almost the analysis of the kW-class PEMFC stack. J Power Sources 2002;
same compared with new cell. 106:295–303.
As shown in Fig. 4 (d), the impedances of air-outlet and [7] Mattsson B, Ericson H, Torell LM, et al. Degradation of a fuel
hydrogen-outlet part are biggest of all parts in every cell, cell membrane, as revealed by micro-Raman spectroscopy.
Electrochimica Acta 2000;45:1405–8.
which of hydrogen-inlet part is the smallest of all, but all of
[8] Pozio A, Silva RF, Francesco MD, et al. Nafion degradation in
them are bigger than that of the new MEA (120 mU cm2). The PEFCs from end plate iron contamination. Electrochimica
impedances of No.1 cell are biggest of all cells, and which of Acta 2003;48:1543–9.
No.8 are smallest of all. [9] Olga Polevaya, Kevin Beverage, Antonio Maggiore. Accelrated
testing methodology of fuel cell stacks. Fuel cell durability &
performance, 2nd Edition Proceedings, Miami Beach, Florida,
4. Conclusions 2006.
[10] Kev Adjemian. Realizing automotive stack needs through
After an accelerated lifetime test, catalyzes on the MEA MEA development. Fuel cell durability & performance, 2nd
agglomerate to some degree. The average particle diameter Edition Proceedings, Miami Beach, Florida, 2006.
[11] Weng Fang-Bor, Jou Bo-Shian, Li Chun-Wei, Su Ay,
grows bigger and internal resistance increases while hydro-
Chan Shih-Hung. The effect of low humidity on the
phobic nature remains the same. In addition, the catalyzes uniformity and stability of segmented PEM fuel cells. J Power
electrochemically active surface area coefficient increases, Sources 2008;181:251–8.
oppositely the maximum of loading current decreases. [12] Ruichun Jiang H, Russell Kunz, James M. Fenton, Influence of
There are enormous differences on performance degrada- temperature and relative humidity on performance and CO
tion among different flow field positions and among different tolerance of PEM fuel cells with Nafion – Teflon – Zr(HPO4)2
cells. Open circuit potential of the worst position is 0.7 V while higher temperature composite membranes. Electrochimica
Acta 2006;51:5596–605.
the best one is 0.96 V. Maximum of current density of the
[13] Najjaria Mustaph, Khemili Faycel, Ben Nasrallah Sassi. The
worst position is 200 mA/cm2, which is one-sixth of the best effects of the cathode flooding on the transient responses of
one. Electrochemically active surface area coefficient of the a PEM fuel cell. Renewable Energy 2008;33:1824–31.
worst position is 25, which is less than one-fifth of the best [14] Kima Hyoung-Juhn, et al. Development of shut-down
one. Average particle diameter of the worst position is 21 nm, process for a proton exchange membrane fuel cell. J Power
which is 2 times bigger than the best one. Impedance of the Sources 2008;180:814–20.
[15] Paquin Mathieu, Fre0 chette Luc G. Understanding cathode
worst position is 250 mU cm2, which is twice of the best one.
flooding and dry-out for water management in air breathing
But contact angle of the worst position and the best one does
PEM fuel cells. J Power Sources 2008;180:440–51.
not differ by more than 10%. [16] Vielstich W, Lamm A, Gasteiger H. Handbook of Fuel Cells:
The best performance of fuel cell is the hydrogen-inlet Fundamentals, Technology and Applications. John Wiley
part, while the worst part is air-outlet part. This may be &Sons 2003;3:17–8.
caused by frequent air starvation or flooding in the air-outlet. [17] Chapuis O, Prat M, Quintard M, Chane-Kane E, Guillot O,
The performance of first cell is the worst and that of eighth cell Mayer N. Two-phase flow and evaporation in model fibrous
media Application to the gas diffusion layer of PEM fuel cells.
is the best. The probable reason is Non-uniform gas intake and
J Power Sources 2008;178:258–68.
water distribution.