Journal of Power Sources 105 (2002) 256–260

Metallic bipolar plates for PEM fuel cells

J. Wind*, R. Späh, W. Kaiser, G. Böhm
Research and Development, European Aeronautic Defence and Space Company/Dornier, Elektrochemical Technology,
88039 Friedrichshafen, Germany

Abstract

Metallic bipolar plates for Polymer electrolyte membrane (PEM) fuel cells with and without coatings were tested in single cell tests.
Current–voltage curves, lifetime curves and the contamination with metal ions were measured. Additionally the surface of the plates was
analyzed by several methods. So far the investigations revealed that principally stainless steel covered with a thin coating is suitable as
material for bipolar plates in PEM fuel cells. Cell performance is the same as in PEM fuel cells with graphite bipolar plates. Concerning the
cost it has to be considered that not only the material itself but also the coating process has to be evaluated. # 2002 Elsevier Science B.V.
All rights reserved.

Keywords: Fuel cell; Bipolar plate; Stainless steel; Coatings

1. Introduction suitability of metals as materials for bipolar plates is inves-

tigated. It is well known that corrosion-resistant metals such
Polymer electrolyte membrane fuel cells (PEM-FC)are as stainless steel form passive surface layers which have a
the most promising power sources in the near future for high ohmic resistance under PEM fuel cell operating con-
residential and mobile applications. They offer the potential ditions. The direct use of these materials leads to a voltage
of compactness, lightweight, high power density and low drop in the fuel cell. The power output and efficiency then
temperature operation. In the past 5 years many research are too low for a commercial application of fuel cells.
organizations, huge and small companies have intensified In order to reduce the contact resistance of the metallic
their efforts towards commercialization of fuel cells extre- bipolar plates various types of coatings and surface treat-
mely. Among these Daimler Chrysler together with Ballard ments were investigated and applied to the metallic plates.
Power Systems and Ford have initiated a big boost for fuel The coatings were tested in fuel cell tests. So far, a number
cell technology. The applied research of European Aero- of promising candidates has been identified. The coatings
nautic Defence and Space Company/ Dornier as part of the investigated up to now are either metallic or based on
research of Daimler–Chrysler is working on future devel- metallic compounds, like oxides and nitrides. One of the
opments for the following generations of fuel cells. A first key issues is to investigate long time performance and to
market introduction of fuel cell vehicles will be seen in reduce the costs of such coatings. Up to 1000 h stable
the near future, however there are still issues to work on operation of PEM fuel cells with metallic bipolar plates
before fuel cell cars will have a big market share. One of the have been demonstrated in small single cells.
important issues in this context is the availability of low So far, the focus of the investigations was on stainless
cost bipolar plates. Bipolar plates based on carbon materials steel as a base material for bipolar plates. Previous work of
have been the main focus of the development activities so other groups [1,2] revealed that stainless steel 316L is the
far. These materials will fulfill all requirements in the near material of choice for bipolar plates in PEM fuel cells.
future. Nevertheless, further cost reduction and increase of During our experiments the base material was tested with
power density is beneficial for fuel cell technology. Bipolar and without coatings.
plates based on metals offer a high potential to reduce costs
and enhance power density.
As part of a public funded project (Development of the 2. Experimental
PEM Fuel Cell Technology for Mobile Applications) the
*
Corresponding author. Tel.: þ49-7545-82845; fax: þ49-7545-85775. The production of bipolar plates with gas and coolant flow
E-mail address: [email protected] (J. Wind). fields is possible with different methods, namely machining,

0378-7753/02/$ – see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 7 7 5 3 ( 0 1 ) 0 0 9 5 0 - 8
 J. Wind et al. / Journal of Power Sources 105 (2002) 256–260 257

etching or embossing. Out of these, embossing is the pre- steel was measured by chemical analysis. If the coatings
ferred process for mass production due to economical showed changes after the test the composition of the surface
reasons. The plates for the tests presented here were pro- was measured by surface analytical methods like Photoelec-
duced by etching the flow fields into the stainless steel. With tron spectroscopy (ESCA) and glow discharge optical emis-
this method, a variety of flow fields can be produced very sion spectroscopy (GDOES).
easily. After production of the flow fields the plates were
coated with different materials by various methods like
electroplating, evaporation, sputtering and chemical vapour 3. Results and discussion
deposition. The coating materials used for cell and stack
tests were chosen either by known conductivity and corro- 3.1. Stainless steel 316L without coating
sion protection properties or after a screening process with
small samples. As it is not easy to simulate the conditions in In our first experiments we tested bipolar plates made of
a fuel cell correctly, cell tests were the preferred measure to stainless steel 316L without coating in comparison with
verify the suitability of the tested plates. These cell tests bipolar plates made of graphite. The flow field geometry was
were conducted in single cells of 49 cm2 active area. The identical with a slight difference in the channel profile due to
membrane electrode assembly consisted of Nafion 113.5 different production methods of the flow fields (machining
membranes and Torray paper TGP90 with a catalyst loading and etching). Fig. 1 shows a current–voltage curve of a
of 4 mg/cm2 on each electrode. The flow field chosen for stainless steel bipolar plate compared with one made of
comparison was a serpentine flow field with two parallel graphite. Our experiments revealed a voltage drop up to
channels. Operation temperature was 75 8C, operation pres- 300 mV at a current of 700 mA/cm2 by using untreated
sure 2.6 bar on both sides. Air stoichiometry was 2.0, fuel stainless steel for bipolar plates. This is far too high for
stoichiometry was 1.2. The gases were humidified with a commercial applications of PEM fuel cells. Surface analysis
dew point of 70 8C. At the beginning of each test a I–V-curve on the air side showed that an oxide layer is formed on the
was measured to achieve the performance of the cell in the surface. Thus, the voltage drop is caused by a thin passivat-
complete operation range from 0 to 1 A/cm2. Voltage probes ing oxide layer on the surface of the bipolar plates. This
inside the cell allowed the measurement of the voltage oxide layer has a high resistivity and thus leads to a series
directly at the electrodes and also the resistance between resistance inside the fuel cell. Chemical analysis after 100 h
electrodes and bipolar plates. Thus, the resistance of pos- cell operation revealed an amount of 76 mg/cm2 Nickel and
sibly formed oxide layers could be measured in situ during almost no iron and chromium in the MEA (Fig. 2).
the cell test. Before and after testing the cells with metallic
plates the cells were tested with graphite bipolar plates to get 3.2. Stainless steel 316L with gold coating
a base value for the MEA performance. Most plates were
tested up to 500 h, whereas very promising candidates were In order to avoid the formation of the oxide layer and
investigated for longer times. After the test, the contamina- nickel dissolution the bipolar plates were coated with a thin
tion of the MEA due to metal ions coming out of the stainless gold layer. The current–voltage curve measured in single

Fig. 1. Comparison of current–voltage curves of single cells with metallic bipolar plates (SS 316L) and graphite bipolar plates (beginning of life).
258 J. Wind et al. / Journal of Power Sources 105 (2002) 256–260

Fig. 2. Contamination of the MEA of single cells with Fe, Cr and Ni, measured after the cell test with different coated metallic bipolar plates.

cells with gold coated bipolar plates (SS 316L) clearly that the gold coating is a good diffusion barrier for nickel
demonstrates that there is no difference between theses under the operation conditions of a PEM fuel cell. As almost
plates and graphite plates. After measuring the I–V-curves, no additional resistance is measured oxygen also does not
the fuel cell was operated under the standard conditions for diffuse under the gold coating. A clear explanation of the
1000 h. Fig. 3 shows that there is almost no deterioration of increase of iron cannot be given up to now, further inves-
the cell voltage. This means that a possibly growing oxide tigations have to be done.
layer is so thin that it has no effect on cell resistance and also
the amount of contaminants in the MEA is so low that it does 3.3. Stainless steel 316L with low cost coatings
not effect the cell performance. The chemical analysis of the
MEA after the cell test shows that the amount of nickel is For commercial applications of fuel cells it is necessary to
reduced considerably, whereas the iron content in the MEA reduce costs to a very low amount. Therefore, gold as a
is enhanced. The reduction of the amount of nickel shows coating material cannot be used due to the high material

Fig. 3. Lifetime curve of a fuel cell with a gold coated metallic bipolar plate (SS316L).
 J. Wind et al. / Journal of Power Sources 105 (2002) 256–260 259

Fig. 4. Lifetime curves of single cells with bipolar plates made of stainless steel 316L coated with different coating materials.

price of gold. In order to reduce the costs of the coatings, cell voltage and high contact resistance similar to uncoated
materials which have the potential for low cost coatings had SS316L and coating C shows also an intermediate perfor-
to be found. Additionally, the coating process also has to be a mance and a relatively high resistance. A detailed analysis
low cost process. In the first step, materials which have a low of additional resistance and corrosion is shown in Table 1. It
price were identified. A number of sheet steel samples were appears that not all coatings which show good cell results
coated and then the conductivity was measured in a through also do not suffer corrosion, but the trend is that a coating
plane measurement. The most promising coating materials which leads to good performance data also has good corro-
were used to coat bipolar plates made of stainless steel 316L sion preventing properties. As the necessary operation time
which were tested in single cell tests. The lifetime curves for for fuel cells is several thousand hours, some promising
500 h of a number of these plates are shown in Fig. 4. As can candidates of low cost coatings were also tested in single cell
be seen the performance of most of these cells is similar to tests for up to 1000 h. An example of such a test is shown
cells with graphite bipolar plates. Coating A exhibits a poor in Fig. 5. As can be seen almost no contact resistance is

Fig. 5. Lifetime curve of a single cell with metallic bipolar plate and selected coating.
260 J. Wind et al. / Journal of Power Sources 105 (2002) 256–260

Table 1 Both oxide formation and poisoning of the MEA are

Oxide resistance and corrosion stability of various coatings after test in a reduced. In this work the focus was on functionality of
PEM fuel cell
protecting layers. The materials used have the potential to be
Resistance (mO cm1) Chemical stability low cost coatings and thus lead to low cost bipolar plates.
Coating A 668.36 
Coating B 63.7 þ
Coating C 205.8  Acknowledgements
Coating D 63.7 þ
Coating E 24.5 þþ
We like to thank the federal ministry of research for
Coating F 19.6 þþ
financial support under contract No.327067A. We are also
indebted to A. Löffler and M. Kolb for technical support.
Special thanks go to F. Jäger, B. Kaltenmark, E. Wegmann
observed, although surface analysis revealed the formation and K. Ruedel for their excellent support with the fuel cell
of a thin oxide layer. test stations.

4. Conclusion
References
The results shown above indicate that metallic bipolar
[1] R. Mallant, F. Koene, C. Verhoeve, A. Ruiter, in: Proceedings of Fuel
plates can be used in PEM fuel cells when coated with Cell Seminar (1994) pp.503.
materials which prevent the formation of oxide layers with [2] C. Zawodzinski, S. Mahlon, S. Gottesfeld, in: Proceedings of Fuel
high resistivity. These layers are stable for at least 1000 h. Cell Seminar (1996) pp.659.