- Email: [email protected]
Effect of a GDL based on carbon paper or carbon cloth on PEM fuel cell performance
Fuel 90 (2011) 436–440
Contents lists available at ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Short communication
Effect of a GDL based on carbon paper or carbon cloth on PEM fuel cell performance Sehkyu Park ⇑, Branko N. Popov Center for Electrochemical Engineering, Department of Chemical Engineering, University of South Carolina, Columbia, SC 29208, USA
a r t i c l e
i n f o
Article history: Received 28 September 2009 Received in revised form 12 August 2010 Accepted 8 September 2010 Available online 24 September 2010 Keywords: Proton exchange membrane fuel cells Gas diffusion layer Carbon paper Carbon cloth Microporous layer
a b s t r a c t A commercially available GDL based on carbon paper or carbon cloth as a macroporous substrate was characterized by various physical and electrochemical measurements: mercury porosimetry, surface morphology analysis, contact angle measurement, water permeation measurement, polarization techniques, and ac-impedance spectroscopy. SGL 10BB based on carbon paper demonstrated dual pore size distribution and high water flow resistance owing to less permeable macroporous substrate, and more hydrophobic and compact microporous layer, as compared to ELAT-LT-1400 W based on carbon cloth. The membrane-electrode-assembly fabricated using SGL 10BB showed an improved fuel cell performance when air was used as an oxidant. The ac-impedance response indicated that a microporous layer which has high volume of micropores and more hydrophobic property allows oxygen to readily diffuse towards the catalyst layer due to effective water removal from the catalyst layer to the gas flow channel. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction A gas diffusion layer (GDL) is embedded between the catalyst layer and the gas flow channel in proton exchange membrane (PEM) fuel cells. The GDL mainly functions as (i) a gas diffuser, (ii) a current collector, and (iii) a physical support, thus determining the catalyst utilization and the overall performance. It also permits gas-phase water to reach the membrane and liquid-phase water to come out of the catalyst layer. A GDL is wet-proofed to prevent water flooding and enhance reactants transport to the catalytic active sites [1–3]. A GDL typically consists of a macroporous substrate and a microporous layer (MPL) of carbon black. Woven carbon cloth or non-woven carbon paper is widely used as a macroporous substrate due to its high gas permeability and electronic conductivity [2]. An MPL reduces ohmic resistance between the catalyst layer and the macroporous substrate, provides non-permeable support during catalyst deposition, and manages liquid water flow [4,5]. The effect of a single-layer GDL (e.g., carbon paper and carbon cloth) on the fuel cell performance has been studied by several researchers, who demonstrated that carbon cloth led higher performance, primarily due to higher porosity and less water saturation [6–8]. Also, extensive work has been performed to examine how the MPL properties such as (i) carbon powder type [4,9], (ii) carbon loading (or MPL thickness) [4,10–13], and (iii) hydrophobic agent concentration [14–16] control water management in PEM
fuel cells. However, the effect of the macroporous substrate in a GDL on pore characteristics for the reactant and the product transport has not been addressed in literature extensively. The objective of this work is to characterize physical properties of a commercially available GDL prepared with carbon paper or carbon cloth and examine how the GDL properties affect water management and oxygen flow during PEM fuel cell operation.
2. Experimental 2.1. Physical characterization of gas diffusion layer Porous structures of the GDLs were analyzed by using a mercury porosimeter (Micromeritics Autopore 9500). In order to perform analysis, small pieces of a GDL were weighed and loaded onto a penetrometer which consists of a sample cup integrated with a metal-clad and glass capillary stem, followed by outgassing from a GDL in a vacuum. Then the penetrometer was automatically filled with mercury. Pore size distribution (PSD) curve was determined from the mercury intrusion data, i.e., the volume of mercury penetrating the pores versus the applied pressure p. Under the assumption that all pores are cylindrical, the pore diameter dp was calculated from the value of p using a well-known capillary law [17]:
dp ¼ ⇑ Corresponding author. Present address: Pacific Northwest National Laboratory, Richland, WA 99352, USA. Tel.: +1 509 371 6277; fax: +1 509 371 6498. E-mail addresses: [email protected], [email protected] (S. Park). 0016-2361/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.fuel.2010.09.003
4c cos h p
ð1Þ
where c and h denote the surface tension of mercury and the contact angle of mercury with the sample, respectively.
437
S. Park, B.N. Popov / Fuel 90 (2011) 436–440
Water transport characteristics through the GDL were determined using the laboratory-fabricated water permeation cell, as shown in Fig. 1. A GDL sample with a diameter of 5 cm was placed into the water permeation cell. Water was slowly added to the inner cylinder in the cell until the hydrostatic head (i.e., water height from the GDL) reaches 102 cm (10 kPa) with the valve closed. When the value was open and the water started to flow, the amount of water passing through the GDL was recorded with time. The water at the outer cylinder was circulated at 20 °C using the water bath. Surface morphology of a GDL was examined using scanning electron microscope (Hitachi). Hydrophobic nature of an MPL was characterized by surface contact angle measurement using a contact angle standard goniometer (Ramé–Hart Instrument). 2.2. Preparation of membrane-electrode assembly The cathode catalyst ink was prepared by ultrasonically blending Pt/C powder (45 wt.% Pt, Tanaka) with Nafion solution (5 wt.% Nafion, Alfa Aesar), deionized water and methyl alcohol for 2 h. The catalyst ink was sprayed onto one side of the Nafion 112 membrane until a total Pt loading of 0.4 mg cm2 was achieved. A commercially available catalyzed GDL (20 wt.% Pt/C, 0.5 mg cm2 Pt, E-TEK) was used as the anode for all fuel cell tests. The Nafioncoated anode (1.2 mg cm2) was hot-pressed to the uncatalyzed side of the membrane at 140 °C and at 15 atm for 90 s. Finally, the GDL of interest was placed on the cathode catalyst layer. 2.3. Electrochemical measurements Electrochemical experiments were carried out in a single cell with serpentine flow channels. Pure hydrogen gas humidified at 77 °C and air humidified at 75 °C were supplied to the anode and cathode compartments. All the measurements were performed at 75 °C and at ambient pressure. Polarization technique was conducted with a fully automated test station (Fuel Cell Technologies Inc.) using a 30 mV potential step and a 5 min dwell time. The stoichiometry of hydrogen and air was 2.0 and the geometric area of the MEA used was 25 cm2. The electrochemical impedance measurement was performed by applying an ac-amplitude of 10 mV over the frequency range from 10 mHz to 10 kHz. 3. Results and discussion Table 1 lists the physical characteristics for four commercially available GDLs: SGL 10CA (carbon paper loaded with 10 wt.% PTFE,
Water Jacket
GDL
Water Collection Water Bath
Fig. 1. Schematic diagram of the water flow measurement across a GDL.
Table 1 Physical characteristics for commercially available GDLs. Property
SGL 10CA
Carbon Cloth A
SGL 10BB
ELAT-LT1400 W
Total thickness dt (lm) Total pore volume Vt (cm3 g1) Median pore diameter dp,med (lm) Average pore diameter dp,ave (lm) Characteristic length lch (lm)
380 3.5 52.1 6.8 77.4
360 2.3 71.2 5.9 194.6
420 2.3 37.7 0.5 45.3
380 1.6 7.8 0.3 34.9
SGL Carbon), Carbon Cloth A (carbon cloth loaded with 10 wt.% PTFE, E-TEK), SGL 10BB (carbon paper loaded with 5 wt.% PTFE and the MPL, SGL Carbon) and ELAT-LT-1400 W (carbon cloth loaded with no PTFE and the MPL, E-TEK). All pore characteristics were estimated from the analyses of mercury intrusion data. The average pore diameter dp,ave was determined using the Carman–Kozeny theory [18]:
dp;ave ¼
4V t At
ð2Þ
where Vt and At denote the total pore volume and the total pore surface area in a GDL. As summarized in Table 1, the median pore diameter dp,med and the characteristics length lch which represents the largest drainable pore size for Carbon Cloth A is larger than those for SGL 10CA, while SGL 10CA demonstrates higher value of dp,ave. It is typically attributed to their different porous structures between non-woven carbon paper and woven carbon cloth shown in Fig. 2(a) and (b) [2]. For a dual-layer GDL, as presented in Fig. 2(c) and (d), MPLs are densely coated on the different substrates and surface morphology for both MPLs is quite similar. However, SGL 10BB is ca. 4.8 times higher than ELAT-LT-1400 W in terms of dp,med, although dp,ave for SGL 10BB is slightly higher. The results indicate there exists different pore geometry coupled with macroporous substrate between SGL 10BB and ELAT-LT-1400 W. Fig. 3 shows the PSD curves (dV/dlogdp) for SGL 10CA, Carbon Cloth A, SGL 10BB, and ELAT-LT-1400 W. As seen in Fig. 3, most of pores in SGL 10CA are observed between 20 and 100 lm, indicating that carbon fibers randomly arrayed result in single PSD. However, Carbon Cloth A exhibits dual PSD in the 2–50 lm and 100–300 lm ranges, which results from the void volume between individual carbon fibers and between carbon yarns (bundles of carbon fibers). It is also observed that there is no significant difference between two single-layer GDLs at smaller pores (dp < 2 lm). Comparing the PSD data for dual-layer GDLs, it is obvious that in the case of SGL 10BB, the pore size ranges between 0.01 and 0.1 lm in the MPL and from 6 to 300 lm in the carbon paper. In contrast, the PSD for ELAT-LT-1400 W is highly uniform over the whole pore sizes. Furthermore, higher volume of pores ranging from 0.1 and 10 lm is observed for ELAT-LT-1400 W. This indicates that the MPL in ELAT-LT-1400 W is significantly entrenched into carbon cloth, reducing large pores (dp > 6 lm) during the MPL deposition. Fig. 3 also illustrates differential pore volume (dV/ddp) against pore size for SGL 10BB and ELAT-LT-1400 W (see the inset). SGL 10BB contains more micropores ranging from 0.01 to 0.1 lm, when compared to ELAT-LT-1400 W. Hence, the mercury intrusion porosimetry in this study specifies that a single-layer GDL based on carbon cloth has larger characteristic length and pore volume at dp,ave > 150 lm due to its woven structure, as compared to that based on carbon paper, which makes carbon particles readily introduced into the pores between carbon yarns during the MPL deposition [19]. In PEM fuel cell, the liquid water produced and condensed at the cathode catalyst layer flows through the GDL, depending on pore geometry as well as hydrophobicity in the GDL [20,21]. For a dual-layer GDL, liquid water transport is not strongly affected
438
S. Park, B.N. Popov / Fuel 90 (2011) 436–440
Fig. 2. SEM micrographs for the GDLs. (a) SGL 10CA, (b) Carbon Cloth A, (c) SGL 10BB, and (d) ELAT-LT-1400 W.
by the macroporous substrate but the MPL adjacent to the catalyst layer because of smaller pores and more hydrophobic property which increase capillary pressure at the interface between the catalyst layer and the MPL [5,22]. In order to examine wettability of the MPL in a dual-layer GDL, the static contact angle hc on the MPL was measured. The value of hc was 157.0° for SGL 10BB and 142.5° for ELAT-LT-1400 W, respectively. This typifies that the MPL of SGL 10BB contain more hydrophobic pores occupied by non-wetting fluid (i.e., hydrophobic agent). Fig. 4 exhibits the superficial velocity v against pressure gradient rp for two commercially available dual-layer GDLs. The experiments were conducted using the laboratory-fabricated water permeation cell, as illustrated in Fig. 1. Volumetric flow rate of water was estimated by weighing water discharged from the water permeation cell with time. The value of v was determined from the volumetric flow rate divided by the geometric GDL area which is in contact with water. Water permeability kw can be expressed by Darcy’s law [17]:
-1
3
SGL 10BB ELAT-LT-1400W
3
4
-1
dV/ddp (cm g μm )
5
2
1
0 0.01
0.1
1
Pore Diameter, dp (μm)
kw ¼ lw
μ Fig. 3. PSD curves (dV/dlogdp) for SGL 10CA, Carbon Cloth A, SGL 10BB, and ELATLT-1400 W by mercury porosimetry. (Inset) PSD curves (dV/ddp) for SGL 10BB and ELAT-LT-1400 W.
v rp
ð3Þ
where lw denotes the water viscosity. As depicted in Fig. 4, when the water starts to flow through the GDL v linearly reduces at higher pressure gradients and then v decreases slowly with decreasing kw. Finally, no water flow is observed at smaller pressure gradients. It is known that at small pressure gradients, nonlinear relationship between superficial velocity and pressure gradient shown in
439
3
-1
Superficial Velocity X 10 (m s )
S. Park, B.N. Popov / Fuel 90 (2011) 436–440
1.2
Porosity of Dual-Layer GDL,ε
-2
Limiting Current Density, ilim (A cm )
1.0 0.8
εGDL
1.0
0.6
0.8
0.4
0.6 εGDL,lim
0.2 0
0.4 0.2
-1
Pressure Gradient (bar m )
Fig. 4 corresponds to non-Darcy behavior. According to Klausner and Kraft [23], nonlinear Darcy flow depends on two independent parameters: the wall force and the pore diameter. Particularly, the hydrophobic treatment of a porous medium is highly responsible for non-Darcy flow, since the hydrophobic agent (e.g., polytetrafluoroethylene (PTFE) and fluorinated ethylene propylene (FEP)) introduced into the pores reinforces wall force and simultaneously alters pore geometry. As seen in Fig. 4, the nonlinear region for SGL 10BB is greater than that for ELAT-LT-1400 W. Larger nonlinear region is attributed to not only higher fraction of hydrophobic pores but also higher volume of micropores, especially in the MPL. In addition, for ELAT-LT-1400 W the minimum pressure gradient which represents the pressure gradient at which water does not flow any longer is close to zero. i.e., water continues to flow across the GDL until liquid pressure is nearly equal to atmospheric pressure, which is related to highly homogeneous porous structure and high wettability. Similarly, Mathias et al. [2] demonstrated that water transport through bare carbon paper (no MPL and no PTFE, TGP-H-1.0T, Toray) was in agreement with Darcy’s flow, while water flow across the carbon paper loaded with 28 wt.% PTFE appreciably deviated from Darcy’s law at low pressure difference. Gostick et al. [22] showed that the carbon paper (SGL 10BA, SGL Carbon) with punctured mask and the carbon paper with the MPL (SGL 10BB, SGL Carbon) increase water flow resistance without rapid water saturation at low capillary pressure, in contrast to bare carbon paper (SGL 10BA) and different breakthrough capillary pressure was observed. Hence, the wettability and porous structure of an MPL which is dependent on macroporous substrate induce non-Darcy’s flow through a GDL, resulting in better water management in PEM fuel cells [22,24,25]. Fig. 5 portrays polarization curves of the PEMFCs measured using SGL 10BB and ELAT-LT-1400 W. The experiments were performed with fully-saturated H2/air under the constant stoichiometry mode at kH2 = 2.0 and kair = 2.0. As illustrated in Fig. 5, SGL 10BB results in better fuel cell performance, which can be attributed to higher total pore volume. Also, the improved performance may be explained in terms of the water management, which results in better oxygen counter flow through the GDL. As can be seen in Fig. 4, in case of SGL 10BB a higher liquid pressure is necessary
Fig. 5. Polarization curves of PEMFCs under H2/air for SGL 10BB and ELAT-LT1400 W. The experiments were performed under the constant stoichiometry mode at kH2 = 2.0 and kair = 2.0. (Inset) Bulk porosity, effective porosity, and limiting current density.
to force water away from the GDL and water ceases flow at a minimum pressure, whereas ELAT-LT-1400 W allows water to drain continuously because of lower water flow resistance, resulting in reduced pores available for oxygen transport within the GDL during fuel cell operation. i.e., lower effective porosity eeff as a result of higher water saturation (see the inset). The equation for effective porosity is fully described in [26]. Fig. 6 displays typical Nyquist plots of the ac-impedance spectra measured on the PEM fuel cell at different values of Ecell. The measured real impedance Z0 was subtracted by the ohmic resistance RX. Since the catalyst layer is made up of the carbon-supported
Imaginary Impedance, - Z " (Ω)
Fig. 4. Superficial velocity as a function of pressure gradient for SGL 10BB and ELATLT-1400 W.
0.8 V
0.6 V
Real Impedance, Z ' (Ω) Fig. 6. Nyquist plots ELAT-LT-1400 W.
of
the
ac-impedance
spectra
for
SGL
10BB
and
440
S. Park, B.N. Popov / Fuel 90 (2011) 436–440
Pt particles and the Nafion ionomer, interfacial oxygen reduction kinetics is strongly dependent on oxygen diffusion through the catalyst layer and oxygen concentration at the interface between the catalyst layer and the MPL. As illustrated in Fig. 6, a smaller acimpedance spectrum was obtained for SGL 10BB at 0.8 V. The reason is that SGL 10BB shows higher gas-phase pore volume in the absence of severe mass transport limitation, thereby enhancing oxygen diffusion kinetics in the catalyst layer. The magnitude of ac-impedance spectra at 0.6 V for SGL 10BB in Fig. 6 is also smaller than that for ELAT-LT-1400 W. This means that smaller charge transfer resistance is responsible for lower oxygen concentration gradient across the GDL mainly due to faster water removal without significant saturation in the macroporous substrate (or carbon paper), although higher water pressure acting on the MPL reduces catalytic active area [5]. i.e., SGL 10BB leads to effective water management, resulting in higher oxygen concentration at the interface between the catalyst layer and the MPL. 4. Conclusion SGL 10BB based on carbon paper and ELAT-LT-1400 W based on carbon cloth were characterized physically and electrochemically. The ex-situ water permeation experiments indicated that the hydrophobic MPL deposition onto carbon paper appreciably enhances the resistance to water flow through the GDL due to smaller pores and higher surface energy in the MPL. Electrochemical polarization demonstrated that SGL 10BB resulted in better fuel cell performance using H2/air. Consequently, the results showed that the hydrophobic and dual PSD GDL supported by carbon paper provides more pore volume available for oxygen transport as well as prevent water accumulation in the GDL, resulting in higher oxygen concentration in the catalyst layer. Acknowledgment Financial support provided by FUJIFILM Manufacturing U.S.A., Inc. is acknowledged gratefully. References [1] Barbir F. PEM fuel cells. Burlington: Elsevier Academic Press; 2005. [2] Mathias M, Roth J, Fleming J, Lehnert W. In: Vielstich W, Lamm A, Gasteiger HA, editors. Handbook of fuel cells, New York: John Wiley & Sons; 2003. [3] Williams MV, Begg E, Bonville L, Kunz HR, Fenton JM. Characterization of gas diffusion layers for PEMFC. J Electrochem Soc 2004;151:A1173–80.
[4] Jordan LR, Shukla AK, Behrsing T, Avery NR, Muddle BC, Forsyth M. Diffusion layer parameters influencing optimal fuel cell performance. J Power Sources 2000;86:250–4. [5] Weber AZ, Newman J. Effects of microporous layers in polymer electrolyte fuel cells. J Electrochem Soc 2005;152:A677–88. [6] Ralph TR, Hards GA, Keating JE, Campbell SA, Wilkinson DP, Davis M, et al. Low cost electrodes for proton exchange membrane fuel cells. J Electrochem Soc 1997;144:3845–57. [7] Frey T, Linardi M. Effect of membrane electrode assembly preparation on the polymer electrolyte membrane fuel cell performance. Electrochim Acta 2004;50:99–105. [8] Wang Y, Wang CY, Chen KS. Elucidating differences between carbon paper and carbon cloth in polymer electrolyte fuel cells. Electrochim Acta 2007;52:3965–75. [9] Jordan LR, Shukla AK, Behrsing T, Avery NR, Muddle BC, Forsyth M. Effect of diffusion-layer morphology on the performance of polymer electrolyte fuel cells operating at atmospheric pressure. J Appl Electrochem 2000;30:641–6. [10] Paganin VA, Ticianelli EA, Gonzalez ER. Development of electrochemical studies of gas diffusion electrodes for polymer electrolyte fuel cells. J Appl Electrochem 1996;26:297–304. [11] Qi Z, Kaufman A. Improvement of water management by a microporous sublayer for PEM fuel cells. J Power Sources 2002;109:38–46. [12] Lee HK, Park JH, Kim DY, Lee TH. A study on the characteristics of the diffusion layer thickness and porosity of the PEMFC. J Power Sources 2004;131:200–6. [13] Park S, Lee JW, Popov BN. Effect of carbon loading in microporous layer on PEM fuel cell performance. J Power Sources 2006;163:357–63. [14] Georgi L, Antolini E, Pozio A, Passalacqua. Influence of the PTFE content in the diffusion layer of low-Pt loading electrodes for polymer electrolyte fuel cells. Electrochim Acta 1998;43:3675–80. [15] Lufrano F, Passalacqua A, Squadrito G, Patti A, Giorgi LJ. Improvement in the diffusion characteristics of low-Pt loaded electrodes for PEFCs. J Appl Electrochem 2004;29:445–8. [16] Park S, Lee JW, Popov BN. Effect of PTFE content in microporous layer on water management in PEM fuel cells. J Power Sources 2008;177:457–63. [17] Bear J. Dynamics of fluids in porous media. New York: American Elsevier Publishing Company; 1972. [18] Kaviany M. Principles of heat transfer in porous media. New York: SpringerVerlag; 1991. [19] Benziger J, Nehlsen J, Blackwell D, Brennan T, Itescu J. Water flow in the gas diffusion layer of PEM fuel cells. J Membr Sci 2005;261:98–106. [20] Gostick JT, Ioannidis MA, Fowler MW, Pritzker MD. Direct measurement of the capillary pressure characteristics of water–air–gas diffusion layer systems for PEM fuel cells. Electrochem Commun 2008;10:1520–3. [21] Park S, Popov BN. Effect of hydrophobicity and pore geometry in cathode GDL on PEM fuel cell performance. Electrochim Acta 2009;54:3473–9. [22] Gostick JT, Ioannidis MA, Fowler MW, Pritzker MD. On the role of the microporous layer in PEMFC operation. Electrochem Commun 2009;11:576–9. [23] Klausner Y, Kraft R. A capillary model for non-Darcy flow through porous media. J Rheol 1966;10:603–19. [24] Nam JH, Lee KJ, Hwang GS, Kim CJ, Kaviany M. Microporous layer for water morphology control in PEMFC. Int J Heat Mass Transfer 2009;52:2779–91. [25] Kang JH, Lee KJ, Yu SH, Nam JH, Kim CJ. Demonstration of water management role of microporous layer by similarity model experiments. Int J Hydrogen Energy 2010;35:4264–9. [26] Springer TE, Wilson MS, Gottesfeld S. Modeling and experimental diagnostics in polymer electrolyte fuel cells. J Electrochem Soc 1993;140:3513–26.
Contents lists available at ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Short communication
Effect of a GDL based on carbon paper or carbon cloth on PEM fuel cell performance Sehkyu Park ⇑, Branko N. Popov Center for Electrochemical Engineering, Department of Chemical Engineering, University of South Carolina, Columbia, SC 29208, USA
a r t i c l e
i n f o
Article history: Received 28 September 2009 Received in revised form 12 August 2010 Accepted 8 September 2010 Available online 24 September 2010 Keywords: Proton exchange membrane fuel cells Gas diffusion layer Carbon paper Carbon cloth Microporous layer
a b s t r a c t A commercially available GDL based on carbon paper or carbon cloth as a macroporous substrate was characterized by various physical and electrochemical measurements: mercury porosimetry, surface morphology analysis, contact angle measurement, water permeation measurement, polarization techniques, and ac-impedance spectroscopy. SGL 10BB based on carbon paper demonstrated dual pore size distribution and high water flow resistance owing to less permeable macroporous substrate, and more hydrophobic and compact microporous layer, as compared to ELAT-LT-1400 W based on carbon cloth. The membrane-electrode-assembly fabricated using SGL 10BB showed an improved fuel cell performance when air was used as an oxidant. The ac-impedance response indicated that a microporous layer which has high volume of micropores and more hydrophobic property allows oxygen to readily diffuse towards the catalyst layer due to effective water removal from the catalyst layer to the gas flow channel. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction A gas diffusion layer (GDL) is embedded between the catalyst layer and the gas flow channel in proton exchange membrane (PEM) fuel cells. The GDL mainly functions as (i) a gas diffuser, (ii) a current collector, and (iii) a physical support, thus determining the catalyst utilization and the overall performance. It also permits gas-phase water to reach the membrane and liquid-phase water to come out of the catalyst layer. A GDL is wet-proofed to prevent water flooding and enhance reactants transport to the catalytic active sites [1–3]. A GDL typically consists of a macroporous substrate and a microporous layer (MPL) of carbon black. Woven carbon cloth or non-woven carbon paper is widely used as a macroporous substrate due to its high gas permeability and electronic conductivity [2]. An MPL reduces ohmic resistance between the catalyst layer and the macroporous substrate, provides non-permeable support during catalyst deposition, and manages liquid water flow [4,5]. The effect of a single-layer GDL (e.g., carbon paper and carbon cloth) on the fuel cell performance has been studied by several researchers, who demonstrated that carbon cloth led higher performance, primarily due to higher porosity and less water saturation [6–8]. Also, extensive work has been performed to examine how the MPL properties such as (i) carbon powder type [4,9], (ii) carbon loading (or MPL thickness) [4,10–13], and (iii) hydrophobic agent concentration [14–16] control water management in PEM
fuel cells. However, the effect of the macroporous substrate in a GDL on pore characteristics for the reactant and the product transport has not been addressed in literature extensively. The objective of this work is to characterize physical properties of a commercially available GDL prepared with carbon paper or carbon cloth and examine how the GDL properties affect water management and oxygen flow during PEM fuel cell operation.
2. Experimental 2.1. Physical characterization of gas diffusion layer Porous structures of the GDLs were analyzed by using a mercury porosimeter (Micromeritics Autopore 9500). In order to perform analysis, small pieces of a GDL were weighed and loaded onto a penetrometer which consists of a sample cup integrated with a metal-clad and glass capillary stem, followed by outgassing from a GDL in a vacuum. Then the penetrometer was automatically filled with mercury. Pore size distribution (PSD) curve was determined from the mercury intrusion data, i.e., the volume of mercury penetrating the pores versus the applied pressure p. Under the assumption that all pores are cylindrical, the pore diameter dp was calculated from the value of p using a well-known capillary law [17]:
dp ¼ ⇑ Corresponding author. Present address: Pacific Northwest National Laboratory, Richland, WA 99352, USA. Tel.: +1 509 371 6277; fax: +1 509 371 6498. E-mail addresses: [email protected], [email protected] (S. Park). 0016-2361/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.fuel.2010.09.003
4c cos h p
ð1Þ
where c and h denote the surface tension of mercury and the contact angle of mercury with the sample, respectively.
437
S. Park, B.N. Popov / Fuel 90 (2011) 436–440
Water transport characteristics through the GDL were determined using the laboratory-fabricated water permeation cell, as shown in Fig. 1. A GDL sample with a diameter of 5 cm was placed into the water permeation cell. Water was slowly added to the inner cylinder in the cell until the hydrostatic head (i.e., water height from the GDL) reaches 102 cm (10 kPa) with the valve closed. When the value was open and the water started to flow, the amount of water passing through the GDL was recorded with time. The water at the outer cylinder was circulated at 20 °C using the water bath. Surface morphology of a GDL was examined using scanning electron microscope (Hitachi). Hydrophobic nature of an MPL was characterized by surface contact angle measurement using a contact angle standard goniometer (Ramé–Hart Instrument). 2.2. Preparation of membrane-electrode assembly The cathode catalyst ink was prepared by ultrasonically blending Pt/C powder (45 wt.% Pt, Tanaka) with Nafion solution (5 wt.% Nafion, Alfa Aesar), deionized water and methyl alcohol for 2 h. The catalyst ink was sprayed onto one side of the Nafion 112 membrane until a total Pt loading of 0.4 mg cm2 was achieved. A commercially available catalyzed GDL (20 wt.% Pt/C, 0.5 mg cm2 Pt, E-TEK) was used as the anode for all fuel cell tests. The Nafioncoated anode (1.2 mg cm2) was hot-pressed to the uncatalyzed side of the membrane at 140 °C and at 15 atm for 90 s. Finally, the GDL of interest was placed on the cathode catalyst layer. 2.3. Electrochemical measurements Electrochemical experiments were carried out in a single cell with serpentine flow channels. Pure hydrogen gas humidified at 77 °C and air humidified at 75 °C were supplied to the anode and cathode compartments. All the measurements were performed at 75 °C and at ambient pressure. Polarization technique was conducted with a fully automated test station (Fuel Cell Technologies Inc.) using a 30 mV potential step and a 5 min dwell time. The stoichiometry of hydrogen and air was 2.0 and the geometric area of the MEA used was 25 cm2. The electrochemical impedance measurement was performed by applying an ac-amplitude of 10 mV over the frequency range from 10 mHz to 10 kHz. 3. Results and discussion Table 1 lists the physical characteristics for four commercially available GDLs: SGL 10CA (carbon paper loaded with 10 wt.% PTFE,
Water Jacket
GDL
Water Collection Water Bath
Fig. 1. Schematic diagram of the water flow measurement across a GDL.
Table 1 Physical characteristics for commercially available GDLs. Property
SGL 10CA
Carbon Cloth A
SGL 10BB
ELAT-LT1400 W
Total thickness dt (lm) Total pore volume Vt (cm3 g1) Median pore diameter dp,med (lm) Average pore diameter dp,ave (lm) Characteristic length lch (lm)
380 3.5 52.1 6.8 77.4
360 2.3 71.2 5.9 194.6
420 2.3 37.7 0.5 45.3
380 1.6 7.8 0.3 34.9
SGL Carbon), Carbon Cloth A (carbon cloth loaded with 10 wt.% PTFE, E-TEK), SGL 10BB (carbon paper loaded with 5 wt.% PTFE and the MPL, SGL Carbon) and ELAT-LT-1400 W (carbon cloth loaded with no PTFE and the MPL, E-TEK). All pore characteristics were estimated from the analyses of mercury intrusion data. The average pore diameter dp,ave was determined using the Carman–Kozeny theory [18]:
dp;ave ¼
4V t At
ð2Þ
where Vt and At denote the total pore volume and the total pore surface area in a GDL. As summarized in Table 1, the median pore diameter dp,med and the characteristics length lch which represents the largest drainable pore size for Carbon Cloth A is larger than those for SGL 10CA, while SGL 10CA demonstrates higher value of dp,ave. It is typically attributed to their different porous structures between non-woven carbon paper and woven carbon cloth shown in Fig. 2(a) and (b) [2]. For a dual-layer GDL, as presented in Fig. 2(c) and (d), MPLs are densely coated on the different substrates and surface morphology for both MPLs is quite similar. However, SGL 10BB is ca. 4.8 times higher than ELAT-LT-1400 W in terms of dp,med, although dp,ave for SGL 10BB is slightly higher. The results indicate there exists different pore geometry coupled with macroporous substrate between SGL 10BB and ELAT-LT-1400 W. Fig. 3 shows the PSD curves (dV/dlogdp) for SGL 10CA, Carbon Cloth A, SGL 10BB, and ELAT-LT-1400 W. As seen in Fig. 3, most of pores in SGL 10CA are observed between 20 and 100 lm, indicating that carbon fibers randomly arrayed result in single PSD. However, Carbon Cloth A exhibits dual PSD in the 2–50 lm and 100–300 lm ranges, which results from the void volume between individual carbon fibers and between carbon yarns (bundles of carbon fibers). It is also observed that there is no significant difference between two single-layer GDLs at smaller pores (dp < 2 lm). Comparing the PSD data for dual-layer GDLs, it is obvious that in the case of SGL 10BB, the pore size ranges between 0.01 and 0.1 lm in the MPL and from 6 to 300 lm in the carbon paper. In contrast, the PSD for ELAT-LT-1400 W is highly uniform over the whole pore sizes. Furthermore, higher volume of pores ranging from 0.1 and 10 lm is observed for ELAT-LT-1400 W. This indicates that the MPL in ELAT-LT-1400 W is significantly entrenched into carbon cloth, reducing large pores (dp > 6 lm) during the MPL deposition. Fig. 3 also illustrates differential pore volume (dV/ddp) against pore size for SGL 10BB and ELAT-LT-1400 W (see the inset). SGL 10BB contains more micropores ranging from 0.01 to 0.1 lm, when compared to ELAT-LT-1400 W. Hence, the mercury intrusion porosimetry in this study specifies that a single-layer GDL based on carbon cloth has larger characteristic length and pore volume at dp,ave > 150 lm due to its woven structure, as compared to that based on carbon paper, which makes carbon particles readily introduced into the pores between carbon yarns during the MPL deposition [19]. In PEM fuel cell, the liquid water produced and condensed at the cathode catalyst layer flows through the GDL, depending on pore geometry as well as hydrophobicity in the GDL [20,21]. For a dual-layer GDL, liquid water transport is not strongly affected
438
S. Park, B.N. Popov / Fuel 90 (2011) 436–440
Fig. 2. SEM micrographs for the GDLs. (a) SGL 10CA, (b) Carbon Cloth A, (c) SGL 10BB, and (d) ELAT-LT-1400 W.
by the macroporous substrate but the MPL adjacent to the catalyst layer because of smaller pores and more hydrophobic property which increase capillary pressure at the interface between the catalyst layer and the MPL [5,22]. In order to examine wettability of the MPL in a dual-layer GDL, the static contact angle hc on the MPL was measured. The value of hc was 157.0° for SGL 10BB and 142.5° for ELAT-LT-1400 W, respectively. This typifies that the MPL of SGL 10BB contain more hydrophobic pores occupied by non-wetting fluid (i.e., hydrophobic agent). Fig. 4 exhibits the superficial velocity v against pressure gradient rp for two commercially available dual-layer GDLs. The experiments were conducted using the laboratory-fabricated water permeation cell, as illustrated in Fig. 1. Volumetric flow rate of water was estimated by weighing water discharged from the water permeation cell with time. The value of v was determined from the volumetric flow rate divided by the geometric GDL area which is in contact with water. Water permeability kw can be expressed by Darcy’s law [17]:
-1
3
SGL 10BB ELAT-LT-1400W
3
4
-1
dV/ddp (cm g μm )
5
2
1
0 0.01
0.1
1
Pore Diameter, dp (μm)
kw ¼ lw
μ Fig. 3. PSD curves (dV/dlogdp) for SGL 10CA, Carbon Cloth A, SGL 10BB, and ELATLT-1400 W by mercury porosimetry. (Inset) PSD curves (dV/ddp) for SGL 10BB and ELAT-LT-1400 W.
v rp
ð3Þ
where lw denotes the water viscosity. As depicted in Fig. 4, when the water starts to flow through the GDL v linearly reduces at higher pressure gradients and then v decreases slowly with decreasing kw. Finally, no water flow is observed at smaller pressure gradients. It is known that at small pressure gradients, nonlinear relationship between superficial velocity and pressure gradient shown in
439
3
-1
Superficial Velocity X 10 (m s )
S. Park, B.N. Popov / Fuel 90 (2011) 436–440
1.2
Porosity of Dual-Layer GDL,ε
-2
Limiting Current Density, ilim (A cm )
1.0 0.8
εGDL
1.0
0.6
0.8
0.4
0.6 εGDL,lim
0.2 0
0.4 0.2
-1
Pressure Gradient (bar m )
Fig. 4 corresponds to non-Darcy behavior. According to Klausner and Kraft [23], nonlinear Darcy flow depends on two independent parameters: the wall force and the pore diameter. Particularly, the hydrophobic treatment of a porous medium is highly responsible for non-Darcy flow, since the hydrophobic agent (e.g., polytetrafluoroethylene (PTFE) and fluorinated ethylene propylene (FEP)) introduced into the pores reinforces wall force and simultaneously alters pore geometry. As seen in Fig. 4, the nonlinear region for SGL 10BB is greater than that for ELAT-LT-1400 W. Larger nonlinear region is attributed to not only higher fraction of hydrophobic pores but also higher volume of micropores, especially in the MPL. In addition, for ELAT-LT-1400 W the minimum pressure gradient which represents the pressure gradient at which water does not flow any longer is close to zero. i.e., water continues to flow across the GDL until liquid pressure is nearly equal to atmospheric pressure, which is related to highly homogeneous porous structure and high wettability. Similarly, Mathias et al. [2] demonstrated that water transport through bare carbon paper (no MPL and no PTFE, TGP-H-1.0T, Toray) was in agreement with Darcy’s flow, while water flow across the carbon paper loaded with 28 wt.% PTFE appreciably deviated from Darcy’s law at low pressure difference. Gostick et al. [22] showed that the carbon paper (SGL 10BA, SGL Carbon) with punctured mask and the carbon paper with the MPL (SGL 10BB, SGL Carbon) increase water flow resistance without rapid water saturation at low capillary pressure, in contrast to bare carbon paper (SGL 10BA) and different breakthrough capillary pressure was observed. Hence, the wettability and porous structure of an MPL which is dependent on macroporous substrate induce non-Darcy’s flow through a GDL, resulting in better water management in PEM fuel cells [22,24,25]. Fig. 5 portrays polarization curves of the PEMFCs measured using SGL 10BB and ELAT-LT-1400 W. The experiments were performed with fully-saturated H2/air under the constant stoichiometry mode at kH2 = 2.0 and kair = 2.0. As illustrated in Fig. 5, SGL 10BB results in better fuel cell performance, which can be attributed to higher total pore volume. Also, the improved performance may be explained in terms of the water management, which results in better oxygen counter flow through the GDL. As can be seen in Fig. 4, in case of SGL 10BB a higher liquid pressure is necessary
Fig. 5. Polarization curves of PEMFCs under H2/air for SGL 10BB and ELAT-LT1400 W. The experiments were performed under the constant stoichiometry mode at kH2 = 2.0 and kair = 2.0. (Inset) Bulk porosity, effective porosity, and limiting current density.
to force water away from the GDL and water ceases flow at a minimum pressure, whereas ELAT-LT-1400 W allows water to drain continuously because of lower water flow resistance, resulting in reduced pores available for oxygen transport within the GDL during fuel cell operation. i.e., lower effective porosity eeff as a result of higher water saturation (see the inset). The equation for effective porosity is fully described in [26]. Fig. 6 displays typical Nyquist plots of the ac-impedance spectra measured on the PEM fuel cell at different values of Ecell. The measured real impedance Z0 was subtracted by the ohmic resistance RX. Since the catalyst layer is made up of the carbon-supported
Imaginary Impedance, - Z " (Ω)
Fig. 4. Superficial velocity as a function of pressure gradient for SGL 10BB and ELATLT-1400 W.
0.8 V
0.6 V
Real Impedance, Z ' (Ω) Fig. 6. Nyquist plots ELAT-LT-1400 W.
of
the
ac-impedance
spectra
for
SGL
10BB
and
440
S. Park, B.N. Popov / Fuel 90 (2011) 436–440
Pt particles and the Nafion ionomer, interfacial oxygen reduction kinetics is strongly dependent on oxygen diffusion through the catalyst layer and oxygen concentration at the interface between the catalyst layer and the MPL. As illustrated in Fig. 6, a smaller acimpedance spectrum was obtained for SGL 10BB at 0.8 V. The reason is that SGL 10BB shows higher gas-phase pore volume in the absence of severe mass transport limitation, thereby enhancing oxygen diffusion kinetics in the catalyst layer. The magnitude of ac-impedance spectra at 0.6 V for SGL 10BB in Fig. 6 is also smaller than that for ELAT-LT-1400 W. This means that smaller charge transfer resistance is responsible for lower oxygen concentration gradient across the GDL mainly due to faster water removal without significant saturation in the macroporous substrate (or carbon paper), although higher water pressure acting on the MPL reduces catalytic active area [5]. i.e., SGL 10BB leads to effective water management, resulting in higher oxygen concentration at the interface between the catalyst layer and the MPL. 4. Conclusion SGL 10BB based on carbon paper and ELAT-LT-1400 W based on carbon cloth were characterized physically and electrochemically. The ex-situ water permeation experiments indicated that the hydrophobic MPL deposition onto carbon paper appreciably enhances the resistance to water flow through the GDL due to smaller pores and higher surface energy in the MPL. Electrochemical polarization demonstrated that SGL 10BB resulted in better fuel cell performance using H2/air. Consequently, the results showed that the hydrophobic and dual PSD GDL supported by carbon paper provides more pore volume available for oxygen transport as well as prevent water accumulation in the GDL, resulting in higher oxygen concentration in the catalyst layer. Acknowledgment Financial support provided by FUJIFILM Manufacturing U.S.A., Inc. is acknowledged gratefully. References [1] Barbir F. PEM fuel cells. Burlington: Elsevier Academic Press; 2005. [2] Mathias M, Roth J, Fleming J, Lehnert W. In: Vielstich W, Lamm A, Gasteiger HA, editors. Handbook of fuel cells, New York: John Wiley & Sons; 2003. [3] Williams MV, Begg E, Bonville L, Kunz HR, Fenton JM. Characterization of gas diffusion layers for PEMFC. J Electrochem Soc 2004;151:A1173–80.
[4] Jordan LR, Shukla AK, Behrsing T, Avery NR, Muddle BC, Forsyth M. Diffusion layer parameters influencing optimal fuel cell performance. J Power Sources 2000;86:250–4. [5] Weber AZ, Newman J. Effects of microporous layers in polymer electrolyte fuel cells. J Electrochem Soc 2005;152:A677–88. [6] Ralph TR, Hards GA, Keating JE, Campbell SA, Wilkinson DP, Davis M, et al. Low cost electrodes for proton exchange membrane fuel cells. J Electrochem Soc 1997;144:3845–57. [7] Frey T, Linardi M. Effect of membrane electrode assembly preparation on the polymer electrolyte membrane fuel cell performance. Electrochim Acta 2004;50:99–105. [8] Wang Y, Wang CY, Chen KS. Elucidating differences between carbon paper and carbon cloth in polymer electrolyte fuel cells. Electrochim Acta 2007;52:3965–75. [9] Jordan LR, Shukla AK, Behrsing T, Avery NR, Muddle BC, Forsyth M. Effect of diffusion-layer morphology on the performance of polymer electrolyte fuel cells operating at atmospheric pressure. J Appl Electrochem 2000;30:641–6. [10] Paganin VA, Ticianelli EA, Gonzalez ER. Development of electrochemical studies of gas diffusion electrodes for polymer electrolyte fuel cells. J Appl Electrochem 1996;26:297–304. [11] Qi Z, Kaufman A. Improvement of water management by a microporous sublayer for PEM fuel cells. J Power Sources 2002;109:38–46. [12] Lee HK, Park JH, Kim DY, Lee TH. A study on the characteristics of the diffusion layer thickness and porosity of the PEMFC. J Power Sources 2004;131:200–6. [13] Park S, Lee JW, Popov BN. Effect of carbon loading in microporous layer on PEM fuel cell performance. J Power Sources 2006;163:357–63. [14] Georgi L, Antolini E, Pozio A, Passalacqua. Influence of the PTFE content in the diffusion layer of low-Pt loading electrodes for polymer electrolyte fuel cells. Electrochim Acta 1998;43:3675–80. [15] Lufrano F, Passalacqua A, Squadrito G, Patti A, Giorgi LJ. Improvement in the diffusion characteristics of low-Pt loaded electrodes for PEFCs. J Appl Electrochem 2004;29:445–8. [16] Park S, Lee JW, Popov BN. Effect of PTFE content in microporous layer on water management in PEM fuel cells. J Power Sources 2008;177:457–63. [17] Bear J. Dynamics of fluids in porous media. New York: American Elsevier Publishing Company; 1972. [18] Kaviany M. Principles of heat transfer in porous media. New York: SpringerVerlag; 1991. [19] Benziger J, Nehlsen J, Blackwell D, Brennan T, Itescu J. Water flow in the gas diffusion layer of PEM fuel cells. J Membr Sci 2005;261:98–106. [20] Gostick JT, Ioannidis MA, Fowler MW, Pritzker MD. Direct measurement of the capillary pressure characteristics of water–air–gas diffusion layer systems for PEM fuel cells. Electrochem Commun 2008;10:1520–3. [21] Park S, Popov BN. Effect of hydrophobicity and pore geometry in cathode GDL on PEM fuel cell performance. Electrochim Acta 2009;54:3473–9. [22] Gostick JT, Ioannidis MA, Fowler MW, Pritzker MD. On the role of the microporous layer in PEMFC operation. Electrochem Commun 2009;11:576–9. [23] Klausner Y, Kraft R. A capillary model for non-Darcy flow through porous media. J Rheol 1966;10:603–19. [24] Nam JH, Lee KJ, Hwang GS, Kim CJ, Kaviany M. Microporous layer for water morphology control in PEMFC. Int J Heat Mass Transfer 2009;52:2779–91. [25] Kang JH, Lee KJ, Yu SH, Nam JH, Kim CJ. Demonstration of water management role of microporous layer by similarity model experiments. Int J Hydrogen Energy 2010;35:4264–9. [26] Springer TE, Wilson MS, Gottesfeld S. Modeling and experimental diagnostics in polymer electrolyte fuel cells. J Electrochem Soc 1993;140:3513–26.