Approved for public release; disfribution is unlimited.

r o / ~ y/ 8 Z.

Title:

Author@):

Submitted to.

POROSIMETRIC STUDY OF CATALYST LAYER OF POLYMER ELECTROLYTE FUEL CELLS

JIAN XIE KENNARD WILSON, JR THOMAS ZAWODZINSKI

THIRD INTERNATIONAL SYMPOSIUM ON PROTON CONDUCTING MEMBRANE FUEL CELLS

/ + A Los Alamos N A T I O N A L L A B O R A T O R Y

Los Alamos National Laboratory, an affirmative actionlequal opportunity employer, is operated by the University of California for the US. Department of Energy under contract W-7405-ENG-36. By acceptance of this article, the publisher recognizes that the US. Government retains a nonexclusive, royalty-free license to publish or reproduce the published form of this contribution, or to allow others to do so, for U.S. Government purposes. Los Alamos National Laboratory requests that the publisher identify thls article as work performed underthe auspices of the U.S. Department of Energy. Los Alamos National Laboratory strongly supports academic freedom and a researcher's right to publish; as an institution, however, the Laboratory does not endorse the viewpoint of a publication or guarantee its technical correctness.

Form 836 (8lOO)

3 CP

POROSIMETRIC STUDY OF CATALYST LAYER OF POLYMER ELECTROLYTE FUEL CELLS

Jian Xie’, Kennard V. Wilson Jr2, and Thomas Zawodzinski3 Los Alamos National Laboratory, Electronic and Electrochemical Materials and

Devices, Los Alamos, New Mexico 87545 Los Alamos National Laboratory, Polymers and Coatings, Los Alamos, New Mexico

87545 3Department of Chemical Engineering, Case Western Reserve University, Cleveland, OH

1

2

44 106-72 17

ABSTRACT

The porosimetry of the catalyst layer made by the “decal process” was studied using mercury porosirnetry. The comparison of the porosimetric profiles between the carbon powder and the catalyst supported on carbon suggests that the loading of the catalyst onto the Vulcan XC-72 carbon changes the porosimetry of the catalyst/carbon. The porosimetry of the catalyst layer depends on the catalyst used and the NafionB content. Boiling the catalyst layer effectively increases the pore area and the porosity of the catalyst layer. The correlation of porosimetric data with performance of catalyst layers suggests that the Nafiona content in catalyst layers plays a vital role on the structure of catalyst layer such as pore size and pore distribution and further influences the performance of PE:FC.

INTRODUCTION

The cathode in a polymer electrolyte fLlel cell (PEFC) contributes the largest energy loss due to the slow kinetics of oxygen reduction reaction (ORR). This issue can be resolved by either developing new noble catalysts for ORR or increasing the catalyst utilization_The recently developed Pt alloy catalysts significantly improved the rate of ORR, hence the performance of the PEFC [ 11. However, much more significantly improved performance has been achieved using the “thin film” catalyst layer with very low catalyst loading (0.12-0,16 mg Pt/cm2), which was developed by Wilson at Los Alamos National Lab (LANL) [2,3]. This “thin film” catalyst layer is made by forming a very thin catalyst layer ( - 4 0 pm thick) on the decals, and then hot pressing the formed thin catalyst layer onto the both sides of membrane to form a membrane electrode assembly (MEA). Tn this “thin film” catalyst layer, recast NafionB ionomer is used instead of PTFE to bind the catalyst clusters together and this makes the structure of the “thin film” catalyst layer different from the catalyst layers made by other methods. This suggests the importance of the microstructure of the catalyst layer on the determination of the performance.

I

A complete description of the microstructure of the catalyst layer will include the size and the size distribution of the catalyst clusters inside the catalyst layer, the orientation of the catalyst clusters and the recast NafionB ionomer network, the

continuity and the uniformity of the recast NafionB ionomer network, the porosimetry of the catalyst layer (pore size, and pore size distribution). To facilitate the oxygen reduction reaction inside the catalyst layer, the gas electrode (here is the catalyst layer) must provide the maximum number of the active catalyst sites, efficient electronic and protonic conductivities because the ORR, taking place at the catalyst sites, requires the gas reactants, the protons and the electrons at the same location simultaneously. The use of the NafionB ionomer as binder provides the efficient protonic conductivity, and the electron conductivity is sufficient using carbon support. The porosimetry of a catalyst layer is a very important property because of the special requirements for the structure of the catalyst layer-it has to have the efficient gas reactant access to the catalyst sites and it also has to provide as many active catalyst sites as possible. The utilization of the catalyst is usually low because a quite number of catalyst clusters are buried in the matrix of the catalyst layer and they are not accessible for the gas reactants, which makes these catalyst sites electrochemically inactive. Therefore, the increase of the number of electrochemically active catalyst sites and the improvement of the gas accessibility are the key to improve the performance of the catalyst layer. Understanding the porosimetry of catalyst layer and the effect of the starting materials and the fabrication process on the porosimetry of the catalyst layer will provide the guidance for designing the high performance catalyst layer.

A lot of studies on porosimetry of electrodes for solid oxide fuel cells (SOFC) have been carried out [4-81 because the porous ceramic electrodes of SOFC are easy to work with. Relatively, not very much work has been done on the porosimetry of the catalyst layer for PEFC due to the difficulties on sample preparation and the data interpretation. Although, several groups [9-113 reported their work on the porosimetry of the PEFC electrode, often, the results varies due to the different materials used and the electrode making methods. Here we present our study of the porosimetry of the catalyst layer, which was made by the “decal processgy, as the part of our comprehensive study on the microstructure of the catalyst layer.

EXPERIMENTAL

All MEAs were prepared by the “decal process” [2,3]. In this process, a catalyst ink was made by mixing a catalyst (carbon supported) with glycerol (Fisher, NJ), 5% NafionB solution (EWI 100, Solution Technology Inc. PA), and tetrabutylammonium hydroxide (TBAOH) (1 .O M solution in methanol, Aldrich, WI) to form a homogeneous ink and by stirring ovcrnight [2,3]. The formed catalyst ink then was hand painted onto a decal surface which was fiberglass coated with Teflon. After painting, the painted decal was heated in an oven at 140 “C for 30 minutes. The process of painting and heating was repeated until the desired catalyst loading was reached, which was equivalent to 0.2 mg Pt/crn2 in this study. The catalyst layer formed on the decal was hot pressed onto a NafionB N112 membrane (sodium form) (Du Pont, DE) at 454 kg/cm2 for 5 minutes. After hot pressing, the catalyst layer was peeled off and a membrane electrode assembly (MEA) was formed. The formed MEA was boiled for two hours first in 0.5M

,

sulfuric acid followed by a two hour boiling in deionized water. After boiling, the MEA was dried on a heated vacuum Table at 60 "C for 30 minutes.

1 Ox 1 Omm squares were cut from the middle of the MEAs to prepare samples for mercury porosimetric measurement. It is worth to point out that the special attention is needed for the sample preparation, because porosimetry of a catalyst layer depends on the type of catalyst used. The anode and cathode catalyst layers, which were hot pressed on two sides of Nafion membrane, must be the same catalyst for the mercury porosimetric measurement. Otherwise, the measured porosimetry data will be the average of the two different catalyst layers. Mercury porosimetry was used because of its wide range of pore size, which is from 0.5 to 1000 run and well-established theory for data interpretation. A Micromeritics mercury porosimeter was used to measure all samples.

The single-cell fuel cell hardware and the cell testing system have been previously described [12]. For the MEA with different Nafion contents used for the polarization measurement, carbon supported PtRu catalyst (20% PtRu/C, Tanaka Kiknzoku Kogyo, Japan) with nominal atomic ratio 2:3 was used for the anode, and carbon supported PtCr catalyst (20% PtCr/C, E-TEK, Natick, MA) with nominal atomic ratio 3: 1 was used for the cathode. For the catalyst loadings, the anode catalyst layers were 0.20 k 0.01 mg Pt/cm2 of PtRu catalyst (20%, PtRu/C, TKK) and the cathode catalyst layers were 0.20 -C 0.01 mg Pt/cm2 of Pt3Cr catalyst (20%, Pt3Cr /Cy E-TEK). A NafionB 112 membrane was used. The geometric active area of all MEAs prepared was 50 cm2. The NafionB ionomer content in all catalyst layers was controlled at wt. 28.8% except noted somewhere else. For all fuel cell performance experiments, the anode catalyst layers were kept as: 0.20 kO.01 mg Pt/cm2 of PtRu catalyst (20%, PtRu/C, TKK) and made with same procedure and parameters. Thus, the measured polarization curves of the MEAs with different cathode catalyst layers or different NafionB content represent the performance of the cathode catalyst layers only. In all single-cell experiments, the cathode was fed with air at a flow rate of 2100 standard cubic centimetedminute (sccm), and the anode was fed with pure hydrogen with flow rate of 1046 sccm. Both air and hydrogen were pre-humidified and the humidified temperatures were set at 80 "C and 105 "C for air and hydrogen respectively. Backpressure was imposed upon the anode and cathode outlet flows, which was 30 psi for both air and hydrogen, to ensure the cell was pressurized. The cell operating temperature was set at 80 "C. A carbon cloth backing (E- TEK, Natick, MA) was used as a gas diffusion layer.

RESULTS AND DISCUSSIONS

Porosimetrv of Vulcan XC-72 and Catalyst Supported On Carbon

The catalyst layer made by the "decal process" is the composite which consists of the recast NafionQ ionomer, and the nano-particle precious metal catalyst supported on micro-particle carbon. The structure of the catalyst layer is determined by the composition of the NafionB content, the catalyst clusters and the carbon support. Since

the catalyst clusters are loaded onto the surface of the carbon through wet chemistry and the size of the carbon cluster (Vulcan XC-72, 0.72 pm in diameter) is much bigger than the size of catalyst cluster (1 0-20 nm in diameter), the porosimetry of the carbon powder (Vulcan, XC-72) was studied first. It is also interested to know how the catalyst loading process changes the porosimetry of the Vulcan XC-72 carbon powder. Therefore, powders of Pt/C, PtRu/C, and PtCr/C catalysts were also studied.

Figure 1 shows the volume pore size distribution curves of four powder samples and the data is summarized in Table 1. For the Vulcan XC-72 powder, the volume pore size distribution curve is divided into three categories: the primary pores, which pore size is concentrated in the range of 24-84 nm and is 53% of total pore, the secondary pores, which pore size is in the range of 3.5-24 nm and is 34% of total pore, and the big pores, which are larger than 84 nm and is only 13% of total pores. The volume pore size distribution curves of PtRu/C and PtWC catalysts powders have the similar trends except slightly difference in the range of pore diameters. It is noticed that the volume pore size distribution curve of Pt/C is different from others; it only has the primary and big pores and do not have the secondary pores. Since Vulcan XC-72 carbon exists in the form of clusters (average size is 0.72 pm in diameter), each of which is made of the primary particles of Vulcan XC-72 (30 nm in diameter), the primary pores and secondary pores of the Vulcan XC-72 are believed to be the interparticle pores between the primary particles. For the PtCr catalyst powder, it seems that the number of the secondary pores become less (23% comparing 34% of Vulcan XC-72) after the PtCr catalyst has been loaded on the Vulcan XC-72 powder. This suggests that a quite of number of small pores was filled by the formed PtCr catalyst clusters and it sounds like that the surface of the Vulcan XC- 72 within the small pores are favorable for the catalyst cluster formation.

For the PtRu catalyst powder, the primary pore range shifts to smaller pores, 24- 55 nm for PtRu instead of 24-84 nm for Vulcan XC-72, and the number of secondary pores increases significantly (49% comparing with 34% of Vulcan XC-72). This suggests that the PtRu catalyst clusters may form along the walls of the primary pores and make these pores smaller. The increased number of secondary pores may attribute to the formed interparticle pores between the formed catalyst clusters. The Pt catalyst is a special case, it does not have the secondary pores and its pore range is relative smaller than others. This indicates that the Pt catalyst forms clusters within both secondary pores (small pores) and the primary pore. It seems that the formed Pt clusters completely fill the secondary pores. Since the Pt is easily to form crystal structure on the carbon surface, unlike the formed PtCr and PtRu catalyst clusters, which are mixture of crystal and amorphous, there may not be interparticle pores for Pt catalyst clusters in the range of 3.5-24 nm.

It seems that the pore size and pore size distribution of the Vulcan XC-72 carbon have been changed after the catalyst loading. But the trend of pore size and pore size distribution are same for all three catalysts and the carbon support. This suggests that the catalyst clusters may not only be formed on the surface but also are formed within these pores.

Boiling Effect on the Porosimetry of Catalyst Layers of MEA

The catalyst layer in a MEA made by the “Decal Process” is a very thin film (7-10 prn in thickness) and lightweight, and it is very difficult to make a catalyst layer, which can exist alone. Since the catalyst layer has to attach on a substrate, in order to study the porosimetry of the catalyst layer, the substrate either has not to contribute any porosimetry profile, in other word, the substrate must be a non-porous material, or has a very defined porosimetry (not interact with the catalyst layer). If the NafionB membrane was nonporous or had a well-defined porosimetry, it would be the ideal substrate. The use of NafionB membrane as substrate would make it possible to directly measure the porosimetry of MEA instead of measuring the isolated catalyst layer. Thus, NafionB 112 membrane was examined to see whether it is a non-porous membrane. For comparison, a piece of glass (cut from a optical glass slide) was also examined. The results are shown in Table 2. It is clear that the N112 membrane is non porous. This guarantees that the measured porosimetric profile of a MEA is only the porosimetry of the catalyst layers on both sides of the membrane without any contribution from the membrane. Interestingly, the glass is shown more porous than the NafionB membrane.

Following the standard MEA-making procedure [2,3], after hot pressing, a MEA has to be boiled in 0.5 M sulfuric acid and Deionized water respectively to transfer the recast NafionB ionomer from sodium form (Na’) to proton form (H’). After boiling, the MEA was dried on a heated vacuum Table at 60 “C for 30 minutes. It was found that the boiling process had significant impact on the porosimetry of the catalyst layer. From Table 2, it can be seen that the total pore area and porosity of all catalyst layers-PtRu/C and PtCr/C with either same or different Nafion contents, increase tremendously after the layers have been boiled. It is believed that the increased pore area and porosity is caused by the expansion and the contraction of the catalyst layer during the boiling followed by drying on vacuum Table. The membrane in the middle of a MEA swells during boiling, and this deforms the catalyst layer from inside to outside. The deformation of the catalyst layer creates a lot of void space inside the catalyst layer which expresses in the form of increase of pore area, After boiling, the MEA is dried on a vacuum Table and this make the membrane contract. Therefore, the deformation of the catalyst layer occurs again and more pores are created.

Effect of the T w e of the Catalyst

It clearly shows in Table 2. that the pore area and porosity of PtRu and PtCr catalyst layers with the same NafionB content are different. Thus, the porosimetry of the catalyst layer is dependent of the type of the catalyst used. It is needed to point out that to measure the porosimetry of a catalyst layer, the catalyst layers on both sides of a membrane must use the same catalyst; otherwise, the obtained porosimetry will be the average of the two different catalyst layers,

For comparison, the porosimetry of four different catalyst layers were measured and the results are shown in Figure 2 and Table 3. The NafionB content for all four catalyst layer were controlled at 28% except the Vulcan XC-72 carbon layer, The

NafionB content was controlled at 33% in the carbon layer so that the carbon to NafionB ratio was the same as in a catalyst layer with 28.8% NafionB. The reason for keeping the NafionB to carbon ratio the same is because (1) the majority of the carbon supported catalyst is carbon (80%) and (2) the duster size of carbon support (approximately 0.72 pm) is much bigger than the catalyst cluster size (1 0-20 nm).

From Figure 2 and Table 3, the four layers have the similar trend for the volume pore size distribution curves except for the PtRu catalyst layer. All three catalyst layers, Vulcan 32-72, Pt/C, E'tCr/C have two categories of pores, primary pores (diameter in nm; 3.5-6.6 for XC-72,3.5-5.6 for Pt/C, 3.8-6.7 for PtCrK) and the secondary pores (diameter in nm; 6.6-10 for XC-72, 5.6-10 for Pt/C, 6.7-34.1 for PtCr/C). The PtCr/C catalyst layer has the slightly large range for secondary pores 6.7-34.1 nm comparing those of XC-72 and Pt/C. The volume pore size distribution curve of PtRdC catalyst layer is different from others, it has three categories of pores (in diameter): primary 3.5- 4.2 nm, 4.2-5.6 nni, and 5.6-19 nm. This clearly shows that the porosimetry of a catalyst layer depends on the type of the catalyst used. This also suggests that the catalyst layers remain the same trend of'tlie isrosimetry as the powder of catalysts.

From the early discussion about the powder samples, clearly, the porosimetry of the starting materials is different from the final catalyst layers in terms of the pore area, pore ranges, volume pore size distribution curves and the porosity. This indicates that no matter how good the starting materials are or how small the catalyst clusters are to begin with, the porosimetry of the finally formed catalyst layer plays a critical role in the determination of the performance. On the other hand, the size and the distribution of the catalyst clusters still play a critical role on the performance, because within those pores and on the surface of the catalyst layer, the number of catalyst clusters per area and the distribution of the non overlapped clusters are the keys for high catalyst utilization and performance.

Effect of NafionB Content

As a major cornponent of the catalyst layer, the content of recast NafionB in a catalyst layer deteimine the structure of the catalyst layer and further the porosimetry of the catalyst layer. Figure 3 shows effect of the recast NafionB content. The range of the primary pore (in diameter) shifts from 3.8-6.7 nm of 28% NafionB to 4.3-8.2 nm for 40% NafionB: the range of secondary pore (in diameter) shifts from 6.7-68.2 nm of 28% NafionB to 8.1-84 nm of 40% NafionB. This clearly demonstrates that the Nafion content affects the pore size distribution. Both the primary pores and the secondary pores tend to become large with increased NafionB content in the catalyst layers. This may be attributed to the structure change of the catalyst layer with the increased NafionB content, As the NafionB content increases, the continuity and uniformity of the recast NafionB ionomer network inside the catalyst layer is improved, the network becomes more uniform, thus, the NafionB network can hold the clusters of catalyst/carbon much tighter than low NafionB content. Meanwhile, increased NafionB content also creates more interfaces between the clusters of catalyst/carbon and recast NafionB, and after boiling, more big pores are produced.

Porosimetry and Performance of the Catalyst Layer

Catalyst layers with different NafionO contents (ranging from 20 to 66% NafionB) were measured, and the pore area data and porosity data are shown in Figure 4. The pore size distribution data is quite complicated and the detailed analysis will be published later. Froin Figure 4, both the pore area and the porosity are highest for the catalyst layer with 28% NafionB. There is an increase in porosity for the catalyst layer of 50% NafionB in Figure 4. The cause of the abnormal increase in porosity for the catalyst layer with 50% Nafion is not clear at this time.

Figure 5 shows the performance of the catalyst layers as the function of NafionB content. The catalyst layer with 28% NafionB outperforms all other catalyst layers at all cell voltages except 0.9V. This matches the porosimetric data in Figure 4, in which the catalyst layer with 28% NafionB has the highest pore area and porosity. This demonstrates that the pore area and porosity are critical properties for the performance of the catalyst layer. It is noticed that the majority of pore size concentrates in the range of 3.8 to 6.7 nm in diameter for catalyst layers with 28 wt. % Nafion. This pore size is much smaller than the primary pore size, which is 20 to 40 nm, reported by Makoto Uchida et al. [ 101. This difference on pore size attributes to the different materials and techniques used in making catalyst layers. We speculate that such small pore size benefits the ORR because water may exist in the form of vapor phase instead of liquid phase in these tiny pores, which makes the diffusion of oxygen much faster than in liquid water. This may be one of explanations for the best performance achieved by the catalyst layer with 28 wt. % Nafion. It is noticed that the catalyst layer with 25% NafionB has the best performance at cell voltage 0.9V. It is not clearly what causes this.

CONCLUSIONS

The porosimetries of the powder of the Vulcan XC-72 and the catalyst supported on the Vulcan XC-72 carbon powder are different due to the catalyst loading process through wet chemistry. The catalyst layers and the catalyst powder, which are the starting materials for making catalyst layer, have different porosimetries, and this demonstrates that the porosimetry of the finally formed catalyst layer is the key for the determination of the performance. Thus, it is worth to address the importance of studying the effect of MEA fabrication process on the porosimetry of the catalyst layer rather than on the starting materials. The study of the starting materials should be focused on the size of the catalyst cluster and the distribution of these catalyst clusters. The porosimetry of the catalyst layer depends on the catalyst used and the NafionB content. The correlation of porosimetric data with performance of catalyst layers suggests that Nafion content in catalyst layers plays a vital role on the structure of catalyst layer such as pore size and pore distribution and further influences the performance of PEFC.

ACKNOWLEDGEMENTS

This work was supported by the U.S. Department of Energy, Office of Advanced Automotive Technology and the Office of Basic Energy Sciences-Chem. Sciences. Acknowledgement is also made to Dr. Mahlon Wilson for the very helpful discussion.

REFERIENCES

[ 13 S. Gottesfeld and T. A, Zawodzinski, Adv. in Electrochem Sci. and Eng., Yol. 5, Eds. Alkire, R.C., Gerischer, H., Kolb, D.M. & Tobias, C.W., p. 197 (1 998). [2] M.S. Wilson and S. Gottesfeld, J. Appl. Electrochem, 22, 1 (1992). [3] Mahlon Wilson and Shimshon Gottesfeld, J. Electrochem. SOC., 139, L28 (1992) [4] J. Divisek, R. Wilkenhoner, and Y. Volfkovich, J. Appl. EZectrochem., 29, 153(1999) [5] D. Simwonis, A. Naoumids, F.J. Dias, J. Linke, A. Moropoulou, J. Mat. Res. 26,1508 (1 997) [6] W. Liu, ZP. Zhang, CR Xia, JQ Xie, CS Chen, J. Inorg. Mat, 15, 849(2000) [7] R. Ohashi, F. Hayashi, 0. Yamamoto, Denki Kagaku, 62,797(1994) [8] H. Itoh, T Yamamoto, M Mori, T Watanabe, T Abe, Denki Kagaku, 64,549(1996) [9] Makoto Uchida, Yuko Fukuoka, Yasushi Sugawara, Nobuo Eda, and Akira Ohta, J. ElectrochemSoc, 143,2245 (1996) [ IO] Makoto Uchida, Yuko Aoyama, Nobuo Eda, and Akira Ohta, ibid, 142,4143(1995) [ 111 E. Gulzow, M, Schulze, N. Wanger, T. Kaz, R. Reissner, G. Steinhilber, A. Schneider, J. Power Sources, 86,352(2000) [12] T.E. Springer, M. S. Wilson, and S. Gottesfeld, J. ElectrochemSoc, 140,3513 (I 993)

0.08

0.07

-A- PtRulXC-72 0.06

0.05

I

0.04 c L I 8 Oo3 -

0 02

0 01

0

-0.01

Log Pore Radlur(nm)

Poros&(%) Pore Range(Dia.nm)

Figure 1, Volume pore size distribution curves of four powder samples: Vulcan XC-72 carbon; Pt/XC-72; PtCr/XC-72; and PtRdXC-72.

Dist.(%) I==-- r XC-72

Pt/XC-72

PtCr./XC-72

I ptRu/xc-72

Total Pore Area (m2/g)

13.35

27.77

86.97

135.23

1 1

1 3.5-24 134 1 24-84 84-3986 86.74

36-68 65.17 68-2196

3.5-24 23

81.04 24-84 84-2 184

77.2

3.5-24 24-55 55-2184

Sample Total Pore Area

1.85 Glass

__I Pt Ru( unboi led) 8.74 PtRu(boi1ed) - 386.21

Pt3Cr(boiled) 347.47 - Pt3Cr( - - 33.3%Nafion,unboiled) --- -__ _ _ _ _ _ _ ~ - - - 10.49 Pt3Cr( 33.3%Nafion, boiled) 31 9.46

--____---_____---I___.

(m21Q) __-_I ________. __

- Membrane(N112) ______ 0.00

Pt3Cr( unboiled) 10.451

_-__ ___ -_.-____̂ ___

__________-._ ___-______

--____ ___ -

4 1 I - 4 - Pt3CrlXC-72, 28% Nafion I

Med. Pore Dia. Porosity (nm) (%)

4.5 0.99 0.0 0.00 4.9 3.76 4.5 95.82 4.8 4.27

78.99 ""-I- 4.7, 3.68 4.8 I 84.12

----_____

06

05

- E 0 4 ;si" 1

0 3

02

0 1

0

-0 1

Log Pore Radius(nm)

Figure 2. Volume pore size distribution curves of four different catalyst layers: Vulcan XC-72 carbon; Pt/XC,-72; PtCr/XC-72; and PtRu/XC-72.

0.35

0 3

0.25

0.2

5-

F

P 8 015 *I

- 0

0.1

0.05

0

-0.05

Log Pore Radlus(nm)

Figure 3. The effect of NafionCD content in catalyst layer on the volume pore size distribution curves of the catalyst layers.

18 0.018

16 0.016

14 0.014

12

10

0.012 A

si 0.01 E

8 0.008 5 6 0.006 n

4 0.004

2 0.002

cn

Y

Q

E

- 0

60 70

0 - --I I 1 i I I ! 1 : 1 I 1 1 1 I 10 20 30 40 50

Nafion Content (Weight %)

Figure 4. Porosimetry of the PtCr catalyst layers with different NafionB content.

1.6

1.4

A 3 1.2

- E 8 N l

5 a 0.8

0.6 E al

5 0.4 0

0.2

.- v) E

.c1

0

0 10 20 30 40 50 60 Nafion Content in Catalyst Layer(Weight%)

Figure 5. Performance of the catalyst layers at various cell voltage as function of NafionB contents.