ECS Transactions, 58 (1) 867-880 (2013)
10.1149/05801.0867ecst ©The Electrochemical Society
Effect of GDL Material on Thermal Gradients along the Reactant Flow Channels in
PEMFCs
E. J. See and S. G. Kandlikar
Department of Mechanical Engineering, Rochester Institute of Technology, Rochester,
New York, 14623, USA
Water and thermal management techniques remain as key
roadblocks to widespread commercial implementation of PEMFCs.
Typically, within PEMFCs a temperature gradient is seen between
the inlet and the outlet of the flow channels. The thermal properties
and morphological structure of the gas diffusion layer (GDL) are
believed to play an important role in the temperature variation. In
this work, an experimental PEMFC with thin foil thermocouples
were used to measure the temperature just above the membrane
near the inlet and outlet locations. Four commercially available
GDLs (Toray TGP-H-060, MRC-105, SGL 25BC, and
Freudenberg H2315) were tested in order to investigate their role
in liquid water transport. The GDL morphology was analyzed
using confocal laser scanning microscopy. Thermal gradients were
measured for all four GDLs at various conditions. It is seen that
the GDL material properties, in particular the in-plane thermal
conductivity, plays an important role in the overall cell
performance due to changes temperature profile and water
distribution in the flow channels.
Introduction
Proton Exchange Membrane Fuel Cells (PEMFCs) are a strong area of interest for the
transportation sector. However, water and thermal management techniques remain as key
roadblocks to widespread commercial implementation. Zhang and Kandlikar (1)
reviewed various cooling techniques and identified the importance of the resulting
temperature gradient along the flow channels. Typically, within PEMFCs a temperature
gradient is seen between the inlet and the outlet of the flow channels. This can be induced
by several causes including uneven reaction and poor cooling. The thermal properties and
morphological structure of the gas diffusion layer (GDL) are believed to play an
important role in the temperature variation. Additionally, these thermal gradients can
result in condensation within the GDL consequently affecting the cell performance (2).
Thermal Gradients within PEMFCs
Coppo et al. (3) illustrated the importance of temperature on PEM fuel cell operation
through the use of a 3D implementation of previously developed models. It was noted
that at higher temperatures, higher membrane ionic conductivity caused improved
performance within the ohmic region. They also highlighted that higher temperatures
improved water removal through both the increase in gas velocities and the change in the
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
867 subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 867-880 (2013)
advection coefficient. Due to these factors, local temperature can play a primary role in
the performance of a PEMFC.
Under typical operating conditions, it has been observed that a temperature difference
in the through-plane direction can occur (4). These temperature differences have been
shown to be as high as 1 °C at the interface of the membrane. Through the overall
thickness of the GDL, these temperature difference can be significantly higher (5,6).
Kim and Mench (7,8) provided thorough reviews of experimental investigations of
temperature-driven water transport, thermo-osmosis. They found the water flux to
proportional to the temperature difference between side of the membrane, as well as
increase with average temperature. They highlighted that thermo-osmosis water flux was
on the same order of magnitude as diffusion flux, and should not be neglected in overall
water balance of PEMFCs. Fu et al. (9) also provided notable experimental data for the
water transport across membranes in the PEMFCs under thermal gradients.
Several methodologies have been used to quantify these temperature differences
within operating PEMFCs. Notably, Daino et al. (2) studied the thermal profile in the
through-plane direction of in situ PEMFCs using optical and infrared imaging. This
methodology allowed for the temperature distribution within the GDL at a 5 µm spatial
resolution. Temperature differences on the order of a few degrees were observed.
Similarly, most studies have focused on through-plane; however in-plane can also have
significant effects on local channel flow conditions (10).
Anisotropic GDL Properties in PEMFCs
The modeling and furthered understanding of the effects of the highly anisotropic
nature of GDLs has garnered much interest in recent years (11-16). Many groups have
investigated and reported values for effective thermal conductivity of GDLs, in both the
in-plane and through-plane directions (5,11-12). Numerical estimation of the in-plane and
through-plane thermal conductivity of GDLs was undertaken by Zamel et al. (13). The
thermal conductivity of the GDL depended on porosity, and it followed that thermal
conductivity increased with a decrease in porosity. It was also emphasized that the inplane thermal conductivity of the GDL was significantly higher than the through-plane
thermal conductivity.
Yablecki et al. (14) modeled the effective thermal conductivity of GDLs, taking into
account the inherent anisotropy of the GDL structure. Both the in-plane and throughplane effective thermal conductivity were modeled using 2-D and 3D methodologies. It
was found that using a two dimensional approach, results were almost an order of
magnitude smaller than that of three dimensional results. This revealed one of the innate
complexities of GDL modeling, as only in three dimensional modeling can the fibers be
properly connected providing a preferential path for heat transport.
He et al. (15) investigated the effect of anisotropic versus isotropic GDLs on
temperature distributions within PEMFCs. This investigation indicated that the
anisotropic GDL results in higher cell temperature gradients within the PEMFC.
However, it is important this has not yet been successfully experimentally validated.
Alhazmi et al. (16) performed an investigation of the effect of the anisotropy of
thermal conductivity of GDLs on the overall performance of fuel cells through 3D
modeling. The focus of the investigation was on the sensitivity of the PEMFC
performance to the thermal conductivity of the GDL. Through a three-dimensional
multiphase model, six cases varying both in plane and through plane thermal conductivity
of the GDL were evaluated. They noted that low in-plane thermal conductivity caused
regions of the PEMFC to not heat, thus facilitating the formation of water pockets in
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
868 subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 867-880 (2013)
these aforementioned regions. Over all of the operating temperatures investigated an
increase in either the in-plane and/or through-plane thermal conductivity augmented
performance. It was also noted that the performance was more sensitive to changes in the
in-plane thermal conductivity.
Despite significant evidence that thermal gradients within PEMFCs directly affect
performance, there has been minimal in situ investigation of membrane temperature.
Additionally, through-plane temperature gradients have been a primary area of interest
while in-plane thermal gradients have received less attention. However, in-plane thermal
gradients have been noted by many researchers (1,3,4,10) and can cause phase change
within the channel length, potentially significantly affecting the water management and
pressure drop, and consequently performance, of the system.
Experimental Set-up
Fuel Cell Testing Facility
The PEMFC is tested using a Greenlight Innovation G40 Fuel Cell Test Stand
operated under constant current control from a TDI Load Bank. The air is supplied via a
Parker Balston Zero Air Generator, while hydrogen and nitrogen are supplied from ultrahigh purity grade compressed gas cylinders. The G40 test stand includes an integrated
humidification system providing humidification via mist injection and optional dry gas
bypass system. The water for used for the humidification system is supplied on-demand
from a custom Siemens water de-ionization system. To provide back pressure, Fairchild
T6000 electro-pneumatic E/P Transducers were used in conjunction with Go Regulator
Inc. BP-Series Pneumatic Back Pressure Regulators. Control for test stand systems is
provided through HYWARE II testing software on a Dell Optiplex 790, while auxiliary
systems, visualization, and data acquisition systems are controlled and monitored through
an HP Z800 Workstation.
Reactant gases are supplied via ¼ inch 316 stainless steel tubing through an in-house
developed heated gas line system. The stainless steel lines are coated with Kapton
polyamide film to prevent electrical discharge to reactants and minimize static
accumulation. Each reactant line is heated via an Omega 120V rope heater through the
use of 30VDC switch solid state relays and insulated using braided fiberglass insulation.
Control for heating is provided through a Watlow EZ-Zone Integrated PID Limit
Controller with full auto-tuned PID output with K-type thermocouple feedback loop.
Limit control is monitored through the Watlow EZ-Zone system via K-type
thermocouples inside the insulation connected to a non-sparking mechanical relay which
disconnects all power in limit situations.
The experimental PEMFC is dual vibration isolated with a Newport SMART Table
UT2 and rubber isolation mounts with a 50A durometer. The manifold-to-manifold anode
and cathode pressure drop are measured with Honeywell FDW differential pressure
transducers with a range of 0-5 psi. Thirty three Honeywell 060-G763-07 pressure
sensors are used to acquire individual channel pressure drops in the entrance region to
obtain individual channel flow rates, as demonstrated by Kandlikar et al. (17).
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
869 subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 867-880 (2013)
Figure 1. Experimental PEMFC with integrated micro-foil thermocouples.
A 50 cm2 optically transparent fuel cell, previously developed for two-phase flow
observation, was utilized in this study. This experimental PEMFC is shown in Figure #.
The flow field geometry and overall dimensions of the active area were used from a study
by Owejan et al. (18) that is scalable to commercial hardware.
The flow fields were 400 µm gold-plated copper plates which were cut via wire EDM
(electrical discharge machining). The gold-plated copper flow fields form two of the
channel walls within the visualization cell, while the GDL forms the third wall. The
fourth wall is created by an optically transparent sheet of Lexan® which also provides
mechanical support. Two machined 6061 aluminum blocks provided compression as well
an inlets and outlets for reactants. The lack of coolant passages necessitated heaters to be
placed on the aluminum compression blocks to heat the fuel cell during warm-up
operation.
At higher current densities, the heaters are disconnected and heat output from the
reaction is used to maintain the cell temperature. In order to monitor the temperature of
the MEA, four thin film thermocouples were placed on the edge of the active area on
both anode and cathode at the inlet and outlet.
Fuel Cell Materials
In the experimental PEMFC, the membrane used was a W.L. Gore Inc. 18 μm
perfluorosulfonic acid (PFSA) membrane. The anode catalyst layer had a target loading
of 0.05 mg Pt/cm2. The cathode catalyst layer had a target loading 0.3 mg Pt/cm 2. Four
commercially available GDLs were tested in order to investigate its role in liquid water
transport in the channel. All samples had a MPL coating, nominally 5 wt. % PTFE
treatment, and approximate thickness of ~210 μm (with the exception of SGL 10BC).
The material properties for all four GDLs are summarized in Table I.
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
870 subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 867-880 (2013)
TABLE I. Summary of GDLs utilized for Experimental PEMFC testing
Property
Freudenberg
H2315
MRC-105
Toray
TGP-H-060
SGL 10BC
Type*
Air-laid hydroentangled CFP
CFP
Wet-laid CFP
2D CFP
MPL Coating
Yes
Yes**
Yes
Yes
Thickness* (μm)
210
245
190
420
Contact Angle (°)
153
148
150
-
PTFE wt. %
5
5**
5
5
* As reported by manufacturer
** In-house by General Motors
Despite their similar material properties, there are significant differences in the
structure of each GDL. The GDL samples used are shown in Figure 2 using confocal
laser scanning microscopy (CLSM). For the CLSM, SGL-25BC was used in place of
SGL-10BC due to availability at time of measurement. Both SGL GDLs have near
identical structure, however overall thickness is the primary difference between the GDLs,
which was negligible for the CLSM analysis.
(a)
(b)
(c)
(d)
Figure 2. CLSM Images of Fiber Structure, Binder, and PTFE of Tested GDLs
a) MRC-105 b) Freudenberg H2315 c) SGL 25 BC d) Toray 060.
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
871 subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 867-880 (2013)
The fiber structure of MRC-105, SGL-10BC, and Toray 060 are very similar, all 2-D
fiber orientation and typically straight fiber orientation. Freudenberg due to its air-laid
hydro-entangled manufacturing process has significantly different fiber orientation with a
truly 3-D orientation and curved fibers. Additionally, the binder which holds the fiber
together varies between each GDL. The binder used in the SGL 25 BC sample is
extremely coarse and does not span between fibers. On the other end of the spectrum, the
binder used in the Toray 060 sample is smooth and spans fibers restricting the open area.
(a)
(b)
Figure 3. CLSM of Fiber Interaction a) MRC-105 b) Freudenberg H2315.
Testing Procedure
In order to ensure steady state measurement, testing followed a standardized protocol
defined through previous studies. Fuel cell conditioning was performed for a minimum of
8 hours after first assembly of the PEM fuel cell to ensure the membrane has been
properly hydrated. At the beginning of each test run (typically a set of approximately 310 test conditions), the PEM fuel cell was conditioned for a minimum of 2 hours with
fully humidified gases at 40 °C and operation at ~0.60 V. Constant current control was
maintained and the load was adjusted every 15 minutes to maintain ~0.60 V. At the
beginning of each test run, the OCV was recorded and compared to previous test runs to
ensure no performance degradation had occurred. If the test run was to consist of tests
with a dry inlet stream, the G40 test stand was set to bypass the humidifiers and
conditioned at ~0.6V for an additional hour before commencing the test run.
Once the PEM fuel cell was properly conditioned, tests covering the full range of
operating conditions were performed. For the experimental PEMFC the following
procedure was followed during testing:
1. Current density was increased to next test condition.
2. Cell temperature, gas temperature, and humidifier dewpoint were set and
allowed to reach steady operation.
3. Prior to data acquisition, the fuel cell operated for up to 60 minutes to reach
steady state
4. Temperature and pressure drop data was recorded for a total of 120 minutes
5. After data acquisition, the load and voltage were recorded via the G40 test
stand.
After each test condition, the process was repeated until either the desired conditions
were completed or the PEM fuel cell is unable to increase load without becoming
unstable.
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
872 subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 867-880 (2013)
Results and Discussion
Fuel Cell Performance Validation
In order to ensure results were representative of commercial automotive PEMFCs, the
performance of the experimental PEMFC was evaluated using a Greenlight Innovation
G40 Fuel Cell Test Stand operated under constant current control from a TDI Load Bank
for the full operating range. The cell was conditioned by constant current operation at
approximately 0.6 V for 8 hours after first assembly to ensure membrane hydration.
Figure 4. Polarization Curve for PEMFC compared with Alhazmi et al. (16)
During testing, a polarization curve was generated for each assembly and testing
condition. As seen in Figure 4, performance did not show much variance due to
stoichiometry with a fully humidified inlet and the MRC-105 GDL. Additionally, as
compared in Figure 4, the performance of the experimental PEMFC exceed that of
similar investigation by Alhazmi et al. (16) which was validated against their numerical
model for performance with anisotropic thermal conductivity GDL. The results for
maximum power output within the tested range are summarized in Table II. As seen in
the table, in dry conditions produced very similar maximum power densities for the
MRC-105 and Freudenberg GDLs. However, in wet conditions the Freudenberg GDL
produced higher power densities by up to 20%. This was investigated by Sergi et al. (19)
and attributed to membrane hydration through high frequency resistance (HFR)
measurements.
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
873 subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 867-880 (2013)
TABLE II. Summary of Maximum Power Density over Tested Range
Stoichiometry
1.5 : 2.5
1.5 : 5
2
3:8
2
Baseline GDL
Inlet RH 0%
0.183 W/cm
at 0.35 A/cm2
0.143 W/cm
at 0.3 A/cm2
0.100 W/cm2
at 0.2 A/cm2
Baseline GDL
Inlet RH 95%
0.254 W/cm2
at 0.5 A/cm2
0.248 W/cm2 at 0.5
A/cm2
0.221 W/cm2
at 0.5 A/cm2
Freudenberg GDL
Inlet RH 0%
0.186 W/cm2
at 0.3 A/cm2
0.126 W/cm2
at 0.3 A/cm2
0.100 W/cm2
at 0.2 A/cm2
Freudenberg GDL
Inlet RH 95%
0.267 W/cm2
at 0.5 A/cm2
0.262 W/cm2
at 0.5 A/cm2
0.281 W/cm2
at 0.5 A/cm2
Temperature Variation Due to Operating Conditions
Once the performance of the experimental PEMFC was validated to ensure results
were representative of commercial automotive PEMFCs, various operating conditions
were tested for both in-plane and through-plane thermal gradients. The majority of testing
focused on in-plane (also known as down-the-channel) thermal gradients. For the results
presented, a positive temperature difference is representative of a temperature increase
from inlet to outlet. A negative temperature difference represents a decrease in
temperature from inlet to outlet. For the initial investigation MRC-105 and Freudenberg
GDLs were tested to show the variation between thermal gradients on the anode and
cathode of the experimental PEMFC.
(a)
(b)
Figure 5. Comparison of temperature difference between cathode and anode
a) Freudenberg b) MRC-105
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
874 subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 867-880 (2013)
For both GDLs, the temperature difference on the anode side was consistently higher
than on the cathode as shown in Figure 5. Generally, the temperature difference on the
anode was more than twice that of the cathode. This was attributed to the significantly
lower superficial gas velocity and specific heat capacity of reactants on the anode.
For the Freudenberg GDL, the maximum temperature difference on the cathode was
just over 2 °C, while on the anode it exceeded 4 °C. For the MRC-105 GDL, on the
cathode the maximum temperature difference was similarly just over 2 °C, while on the
anode the temperature difference surpassed 7 °C. On the cathode side, both GDLs
maintain very comparable temperature gradients at low stoichiometry. However, on the
anode only the highest stoichiometry showed an analogous trend between the two GDLs.
After the trends for temperature gradient within the fuel cell had been established, the
investigation focused on key operating parameters that have been shown to affect
performance. The first parameter investigated was the inlet relative humidity of reactant
gases. Inlet relative humidity was tested at both a fully humidified inlet, as well as dry
inlet gases. In Figure 6, the temperature gradients on the cathode side are compared for
two GDLs, MRC-105 and SGL 10BC, at both inlet relative humidity conditions.
(a)
(b)
Figure 6. Comparison of in-plane temperature difference between dry and fully
humidified inlets a) MRC-105 b) SGL 10 BC.
At low stoichiometry (1.5 : 2.5), very little change in the temperature difference is
seen for the MRC-105 GDL. Additionally, despite doubling the stoichiometry on the
cathode, both fully humidified tests with MRC-105 showed comparable temperature
gradients. However, at higher stoichiometry, the dry gas inlet showed a much smaller
temperature difference. This is most likely due to membrane dry-out seen during these
tests, as previously mentioned in the performance validation.
The SGL-10BC with a fully humidified inlet had near identical temperature
differences to that of the MRC-105, despite the dissimilar stoichiometry. The variation
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
875 subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 867-880 (2013)
between the two tests was typically less than ~0.2 °C. However, with a dry gas inlet, as
current density increased the temperature gradient dropped off significantly.
Temperature Variation Due to GDL
Once the effect that operating condition had on in-plane thermal gradients within the
PEMFC had been established, the focus of the investigation was on the effect of GDL
anisotropy and structure on the thermal gradients. Tests at two stoichiometries were
completed for Toray 060, Freudenberg H-2315, and MRC-105. SGL 10 BC was tested
for one of the stoichiometries for comparison. In Figure 7, the in-plane cathode
temperature differences are compared at two stoichiometries for all four GDLs under
investigation.
(a)
(b)
Figure 7. Comparison of in-plane temperature difference between GDLs on cathode
a) Stoich 3 : 8 b) Stoich 1.5 : 5
Figure 7a compares all four of the GDLs’ temperature gradient on the cathode side at
a stoichiometry of 3 : 8. Significant variation in the temperature differences along the
channel are seen between the four GDLs. The MRC-105 exhibited the lowest temperature
difference across the range of current densities, reaching a maximum difference of
1.03 °C. Conversely, Toray 060 exhibited the greatest magnitude temperature difference,
nearing 3.74 °C. When the stoichiometry is reduced, as shown in Figure 7b, Toray 060
showed near identical performance, with a maximum temperature difference of 3.96 °C.
Contrariwise, the temperature gradient of the MRC-105 doubled from that of the higher
stoichiometry, reaching a maximum of 2.12 °C. The behavior of the Freudenberg GDL
most closely mimicked that of the MRC-105.
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
876 subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 867-880 (2013)
(a)
(b)
Figure 8. Comparison of in-plane temperature difference between GDLs on anode
a) Stoich 3 : 8 b) Stoich 1.5 : 5
In Figure 8, the in-plane anode temperature differences are compared at two
stoichiometries for all four GDLs under investigation. Compared to the results for the
cathode, it should be noted that the change in temperature difference with respect to
current density is more linear. Figure 8a compares all four of the GDLs’ temperature
gradient on the anode side at a stoichiometry of 3 : 8. Minor variation in the temperature
differences along the channel are seen between three out of four GDLs. The SGL 10 BC
exhibited the maximum temperature difference across the range of current densities,
reaching a difference of nearly 8 °C. Conversely, the three other GDLs exhibited similar
performance with temperature differences on the order of 4 °C. When the stoichiometry
is reduced, as shown in Figure 8b, Toray 060 and Freudenberg showed near identical
performance, with a maximum temperature difference of ~4 °C. However, the
temperature gradient of the MRC-105 increased from that of the higher stoichiometry,
reaching a maximum of 7.08 °C.
Through-Plane Temperature Variation
In addition to the in-plane thermal gradients that were investigated, the experimental
PEMFC was utilized to measure the thermal gradient across the membrane. This metric
can be extremely valuable, as it determines the thermo-osmotic drag and consequently
the overall water balance within the PEMFC. This has been studied extensively in several
ex situ studies, relating thermal gradient to thermo-osmotic drag, however in situ data of
membrane surface temperatures is still relatively scarce.
In order to investigate through-plane temperature gradients, tests were run with MRC105 and Freudenberg GDLs at three stoichiometries. Figure 9 shows the temperature
difference between each side of the membrane at both the inlet and the outlet. A positive
value indicates that the cathode is at a higher temperature than the anode. Conversely, a
negative value indicates the anode is at a higher temperature than the cathode.
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
877 subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 867-880 (2013)
(a)
(b)
Figure 9. Comparison of through-plane temperature difference
a) Freudenberg b) MRC-105
As shown in Figure 9, for both GDLs tested at the inlet the cathode side was at a
higher temperature, while at the outlet the anode was at a higher temperature. Since the
experimental PEMFC was not actively cooled for the purposes of this investigation, this
can be attributed to the significantly lower superficial gas velocity and specific heat
capacity of reactants on the anode. These factors greatly reduce the amount of heat that
can be removed from the PEMFC via the reactants on the anode versus cathode side.
Figure 9a shows the through-plane temperature differences with the Freudenberg
GDL. Only minor variation in the temperature differences along the channel are seen
between stoichiometries. However, it should be noted that the stoichiometries that exhibit
the largest magnitude temperature difference at the inlet, also exhibited the largest
magnitude on the outlet.
Figure 9b shows the through-plane temperature differences with the MRC-105 GDL.
Only minor variation in the temperature differences along the channel are seen between
stoichiometries on the inlet. However, at the outlet significant variation in the
temperature differences along the channel are seen between stoichiometries. The increase
in temperature difference increased with stoichiometry. The magnitude of the maximum
temperature difference was significantly higher than that of the Freudenberg, 4.20 °C and
1.94 °C respectively.
Summary
In this study, thermal gradients and temperature differences in both the in-plane and
through plane direction are reviewed. A methodology to measure temperature differences
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
878 subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 867-880 (2013)
at the GDL interface was established and tested over operating conditions of interest for
PEMFCs.




In order to study the morphology of the tested GDLs, confocal laser scanning
microscopy (CLSM) was utilized and it was noted that Freudenberg H-2315
GDL has a significantly different fiber structure from that of Toray 060, SGL
10BC, and MRC-105.
The thermal gradients within PEMFCs were shown to be much greater on the
anode versus the cathode due to the significantly lower superficial gas
velocity and specific heat capacity of reactants on the anode.
The inlet relative humidity of reactant gases was shown to affect in-plane
thermal gradient under low-load conditions, with a fully humidified inlet
increasing the temperature difference.
The GDL utilized in the experimental PEMFC was shown to have a
significant effect on thermal gradients within the fuel cell in both in-plane and
through-plane direction.
The change in temperature along the flow direction was observed to vary based on GDL
structure, specifically due to a change in its in-plane thermal conductivity. This difference
in temperature profile was seen to affect the overall cell performance. Under conditions at
which membrane dehydration was likely, the change of in-plane thermal conductivity of
the GDL allowed improved performance. It is seen that the GDL material properties, in
particular the in-plane thermal conductivity, plays an important role in the overall cell
performance due to changes temperature profile and water distribution in the flow
channels.
Acknowledgments
Support for this work was provided by the US Department of Energy under the award
number DE-EE0000470. This work was conducted at the Thermal Analysis,
Microfluidics, and Fuel Cell Laboratory at the Rochester Institute of Technology. Gas
diffusion media and technical support was provided by General Motors Electrochemical
Energy Research Laboratory.
References
1. G. Zhang and S. Kandlikar, Int. J. of Hydrogen Energy, 37, 2412-2429 (2012).
2. M. Daino, Z. Lu, J. LaManna, J. Owejan, T. A. Trabold, and S. G. Kandlikar,
Electrochem. Solid-State Lett., 14, B51-B54 (2011).
3. M. Coppo, N. P. Siegel, and M. R. von Spakovsky, J. Power Sources, 159, 560569 (2006).
4. M. Mench, Fuel Cell Engines, Wiley, New Jersey (2008).
5. M. Khandelwal and M. Mench, J. Power Sources, 161, 1106-1115 (2006).
6. A. Z. Weber and J. Newman, J. Electrochem. Soc., 153, A2205-A2214 (2006).
7. S. Kim and M. Mench. ECS Trans., 13 (28), 89-105 (2008).
8. S. Kim and M. Mench, J. Membrane Sci., 328, 113-120 (2009).
9. R. Fu, J. Preston, U. Pasaogullari, T. Shiomi, S. Miyazaki, Y Tabuchi, D Hussey,
and D Jacobson. J. Electrochem. Soc., 158 (3), B303-B312 (2011).
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
879 subject to ECS license or copyright; see ecsdl.org/site/terms_use
ECS Transactions, 58 (1) 867-880 (2013)
10. E. See and S. Kandlikar. ECS Trans., 50 (2), 99-11 (2013).
11. J. Ramousse, O. Lottin, S. Didierjean, and D. Maillet, J. Power Sources, 192,
435-441 (2009).
12. G. Karimi, X. Li, J. Shen, and P. Teertstra, Electrochemica Acta , 55, 1619-1625
(2010).
13. N. Zamel, X. Li, J. Shen, A. Wiegmann, and J. Becker, Chem. Eng. Sci., 65,
3994-4006 (2010).
14. J. Yablecki, A Nabovati, and A. Bazylak, J. Electrochem Soc., 159 (6), B647B653 (2012).
15. G. He, Y. Yamazaki, and A. Abudula, J. Power Sources, 195, 1551-1560 (2010).
16. N. Alhazmi, D.B. Ingham, M.S. Ismail, K.J. Hughes, L. Ma, M. Pourkashanian.
Int. J. of Hydrogen Energy, 38, 603-611 (2013).
17. S. Kandlikar, Z. Lu, W. Domigan, A. White, M. Benedict, Int. J. of Heat and
Mass Transfer , 52 (7/8) 1741-1752 (2009).
18. J. Owejan, J. Gagliardo, J. Sergi, S. Kandlikar, and T. Trabold, Int. J. of
Hydrogen Energy, 34(8), 3436-3444 (2009).
19. J. Sergi and S. Kandlikar, Int. J. Hydrogen Energy, 36 (19), (2011).
Downloaded on 2013-12-06 to IP 129.21.102.159 address. Redistribution
880 subject to ECS license or copyright; see ecsdl.org/site/terms_use