PEMFC Cathode Contamination Mechanisms for
several VOCs - Acetonitrile, Acetylene,
Bromomethane, Iso-propanol, Methyl Methacrylate,
Naphthalene and Propene
Jean St-Pierre, Junjie Ge, Yunfeng Zhai, Tatyana V.
Reshetenko, Michael Angelo
Hawaii Natural Energy Institute, University of Hawaii –
Manoa, Honolulu, Hawaii 96822, USA
Proton exchange membrane fuel cells (PEMFC) are
currently being demonstrated and marketed as
replacements for more established technologies such as
the energy inefficient and unsustainably fueled internal
combustion engine (1). However, a gap still exists
between PEMFC durability targets and current system
specifications. PEMFCs are fed with ambient air that
contains the oxygen oxidant. The fuel cell exposure risk
to a multitude of air contaminants is significant (2).
Contamination mechanisms were only established for a
few species but they are needed to devise robust
prevention and recovery procedures.
Contaminants were selected using a two tiered approach
that greatly decreased the number of species requiring an
experimental evaluation (3,4). Seven organic
contaminants with different functional groups were
chosen for extensive characterization tests: acetonitrile (a
nitrile, solvent and chemical intermediate), acetylene (an
alkyne, welding fuel and chemical intermediate),
bromomethane (an halocarbon and fumigant), iso-
propanol (an alcohol, solvent, chemical intermediate and
windshield de-icer), methyl methacrylate (an ester and
synthesis precursor for poly(methyl methacrylate), a
shatter-resistant alternative to glass), naphthalene (an
aromatic, chemical intermediate and fumigant) and
propene (an alkene and synthesis precursor for
polypropylene used for films, packaging, etc). Impedance
spectroscopy was first used to assess resistance loss types:
kinetic, ohmic, mass transfer. All seven species led to
changes in kinetic and mass transfer resistances. Only
acetonitrile affected the ohmic resistance. Subsequently,
additional tests were completed to resolve in more detail
each resistance loss type. A rotating ring/disc electrode, a
membrane conductivity cell, a segmented cell, a single
fuel cell coupled with a gas chromatograph and a tracer
based method able to measure the liquid water content in
flow field channels and gas diffusion electrodes (5,6)
were either used or planned to either localize more
precisely or gain additional insight into each resistance
loss type.
Acetylene experimental data were synthesized into a
consistent and cohesive contamination mechanism (Fig.
1). As the acetylene moves along the flow field channels,
it is not appreciably dissolved into liquid water droplets
because its solubility is too low. In other words, the
scavenging effect of liquid water is negligible (7). As the
acetylene penetrates the gas diffusion layer, it may adsorb
on the C fibers, sub-layer C particles and catalyst C
support affecting its hydrophobic and liquid water
management properties, and induce a change in mass
transfer resistance. This hypothesis remains to be tested as
the carbon surface area and acetylene adsorption energy
are expected to be smaller than on Pt. Subsequently, the
acetylene reaches the ionomer. However, acetylene does
not impact catalyst layer ionomer conductivity. Further
along its transport path, the acetylene reaches the catalyst
layer where it adsorbs on the Pt surface, reduces the
active area and induces an oxygen reaction mechanism
change which is detected by an increase in the fraction of
peroxide produced from 3 to 14% at 30C and 0.5 V vs
RHE. The acetylene deleterious impact persists within an
application relevant cell voltage range of 0.55 to 0.85 V
despite a large change in acetylene conversion of
respectively <1 to ~100% (mostly CO
2
with CO traces).
Mass transfer in the ionomer layer is also affected by the
presence of adsorbed acetylene on the catalyst surface
(longer transport path) and is expected to be the major
contributor in comparison to acetylene adsorption on C.
Planned residence time distribution measurements (tracer
based method) are expected to clarify the source of the
mass transfer loss (C versus Pt adsorption) by quantifying
the amount of liquid water in the gas diffusion electrode.
The change in liquid water content indirectly provides an
assessment of the C surface hydrophobic state. Finally,
the acetylene reaches the membrane. However, acetylene
does not impact the membrane conductivity.
Current distribution measurements indicate that a
contamination front moves from the cell inlet to the outlet
during the cell voltage transient, from the contaminant
injection to the steady state. After a temporary acetylene
exposure, the cell voltage recovers to almost its original
value (>90% recovery). The small remaining loss in
performance is ascribed to residues on the catalyst
surface. A revised version of the acetylene mechanism
will be presented as well as an experimental data
summary for all other selected contaminants.
REFERENCES
1. 2011 Fuel Cell Technologies Market Report, United
States Department of Energy, Energy Efficiency and
Renewable Energy (2012).
2. J. St-Pierre, in Polymer Electrolyte Fuel Cell
Durability, F. N. Büchi, M. Inaba, and T. J. Schmidt,
Editors, p. 289, Springer (2009).
3. J. St-Pierre, M. S. Angelo, and Y. Zhai, Electrochem.
Soc. Trans., 41(1), 279 (2011).
4. J. St-Pierre, Y. Zhai, and M. Angelo, Int. J. Hydrogen
Energy, 37, 6784 (2012).
5. J. Diep, D. Kiel, J. St-Pierre, and A. Wong, Chem. Eng.
Sci., 62, 846 (2007).
6. J. St-Pierre, A. Wong, J. Diep, and D. Kiel, J. Power
Sources, 164, 196 (2007).
7. B. Wetton and J. St-Pierre, Electrochem. Soc. Trans., 50
(2), 649 (2012).
Flow field channel Gas diffusion layer Catalyst layer Membrane
Air + C
2
H
2
Products
Pt
C
2
H
2
CO
2
,CO
Pt
C
2
H
2
CO
2
,CO
Pt
C
2
H
2
0.85
0.55
E vs H
2
In situ (80 C)
C
2
H
2
C
2
H
2 X
Pt
O
2
H
2
O (97 %)
H
2
O
2
(3 %)
Pt
O
2
H
2
O (86 %)
H
2
O
2
(14 %)
Ex situ (30 C,0.5 V vs RHE)
C
2
H
2
C
2
H
2
OOO
O H
H
OOO
O H
H H C C
X
C C
or ?
H
<
>
C
2
H
2
C
2
H
2 X
Pt Pt
C
2
H
2
O
2
O
2
?
Flow field channel Gas diffusion layer Catalyst layer Membrane
Air + C
2
H
2
Products
Pt
C
2
H
2
CO
2
,CO
Pt
C
2
H
2
CO
2
,CO
Pt
C
2
H
2
0.85
0.55
E vs H
2
In situ (80 C)
Pt
C
2
H
2
CO
2
,CO
Pt
C
2
H
2
CO
2
,CO
Pt
C
2
H
2
0.85
0.55
E vs H
2
In situ (80 C)
C
2
H
2
C
2
H
2 X
C
2
H
2 X
Pt
O
2
H
2
O (97 %)
H
2
O
2
(3 %)
Pt
O
2
H
2
O (86 %)
H
2
O
2
(14 %)
Ex situ (30 C,0.5 V vs RHE)
C
2
H
2
C
2
H
2
OOO
O H
H
OOO
O H
H H C C
X
C C
or ?
H
<
>
C
2
H
2
C
2
H
2
OOO
O H
H
OOO
O H
H H C C
X
C C
or ?
H
<
>
C
2
H
2
C
2
H
2 X
Pt Pt
C
2
H
2
O
2
O
2
C
2
H
2
C
2
H
2 X
C
2
H
2 X
Pt Pt
C
2
H
2
O
2
O
2
?
Fig. 1. Proposed acetylene contamination mechanism.
Abstract #1330, 224th ECS Meeting, © 2013 The Electrochemical Society