Introduction. Jeong-Hi Kim, Hyoung-Juhn Kim,Tae-HoonLim, and Ho-In Lee

Transcription

1 Jeong-Hi Kim, Hyoung-Juhn Kim,Tae-HoonLim, and Ho-In Lee School of Chemical and Biological Engineering & Research Center for Energy Conversion and Storage, Seoul National University, Seoul , Korea Fuel cell Research Center, Korea Institute of Science and Technology, Seoul , Korea Received April 16, 2007; Accepted May 3, 2007 Abstract: Pt/C catalyst layers with a phosphoric acid-doped polybenzimidazole ionomer binder were used as the electrode for a high temperature polymer electrolyte fuel cell. Catalyst layers were prepared with a transparent ionomer solution in trifluoroacetic acid solvent, an opaque ionomer dispersion in a trifluoroacetic acid/propanol-based solvent, and a mixture of the above two substances. The nano-sized ionomers in the transparent ionomer solution were located inside the primary pores, while the micro-sized ionomer agglomerates in the opaque ionomer dispersion were located outside the primary pores and produced proper proton conduction connections without blocking the primary pore openings. The optimum cell performance was obtained with an appropriate mixture of the transparent ionomer solution and opaque ionomer dispersion. Keywords: high temperature polymer electrolyte fuel cell, ionomer binder, membrane electrode assembly, phosphoric acid-doped polybenzimidazole Introduction 1) In recent years, polymer electrolyte fuel cells (PEFCs) have been spotlighted as promising electricity generators with high energy efficiency and ease of carriage. Mostly, perfluorosulfonic acid (PFSA) proton conducting polymers such as Nafion are used for the membrane and catalyst binder. However, the operation of the cells is limited to temperatures below the boiling point of water under atmospheric pressure, because the proton conductivity of the polymers is strongly dependent on the relative humidity [1]. High temperature cell operation has the advantage of allowing for simple water management and high tolerance to CO [2-9]. In the case of Nafion, the water drag generated by the proton transport causes the anode to become dry and the cathode to become flooded with liquid water, which decreases the cell performance. To make up for the water imbalance, very delicate operations are To whom all correspondence should be addressed. ( hilee@snu.ac.kr, thlim@kist.re.kr) needed. The operation of the cell at high temperature does not produce liquid water and simplifies the equipment required for water supply and recovery. Another difficulty associated with low temperature fuel cell operation under 100 o C is the poisoning of the Pt catalyst by CO. The interaction of CO with Pt is very strong and the CO adsorbed on Pt blocks the active sites needed for proton formation. Thus, the cell performance decreases rapidly with increasing CO content in the fuel. However, when the cell is operated at temperatures over 150 o C, the cell performance is not affected by the CO in the feed gas. Therefore, reformate gas with a high CO content can be used without the need for a multi-step fuel purification process. Much effort has been made to develop ionomers for high temperature PEFCs ( o C), and remarkable progress has been achieved with acid-doped polybenzimidazole (PBI) derivatives. Since PBI doped with phosphoric acid was first reported to have high thermal stability and high proton conductivity at high temperature [10,11], many different types of benzimidazole polymers have been synthesized and the modification of

2 Improvement of High Temperature Polymer Electrolyte Fuel Cell Performance with Phosphoric Acid-Doped Polybenzimidazole Ionomer Binder in Catalyst Layer 851 these polymers with additives has also been reported. Recently, phosphoric acid-doped poly(2,5-benzimidazole) (ABPBI), which is the simplest member of the polybenzimidazole family, was reported to have better proton conductivity than phosphoric acid-doped PBI by Asensio and coworkers [12-14]. ABPBI can be prepared by the condensation reaction of a single commercial monomer which is cheaper than the binary monomers used for the synthesis of PBI. The direct casting of the ABPBI film obtained from the homogeneous polymer solution developed by Kim and coworkers [15,16] made the preparation process very simple. There have been several reports on the synthesis of PBI derivatives and their application to PEFCs [17-27]. However, most of these studies were focused on the use of acid-doped polybenzimidazole ionomers as a membrane and very few results have been reported for fabrication methods using the ionomer as a binder polymer in a catalyst layer. The previously reported methods of fabricating catalyst layers with phosphoric acid-doped PBI ionomers can be classified as follows: (1) Post doping of phosphoric acid on the electrode impregnated with PBI/N,N-dimethylacetamide solution on the pre-cast Pt/C layer [17], (2) Post doping of phosphoric acid on the electrode pre-cast with Pt/C and PBI dispersion in acetone [27], and (3) Casting of the electrode using a catalyst slurry impregnated with phosphoric acid and PBI in trifluoroacetic acid [16]. This paper presents a new fabrication method designed to enhance the utilization of the Pt catalyst. The morphological changes were studied with different ionomer binder contents using a transparent ionomer solution and opaque ionomer dispersion. The size of the ionomer was measured and compared with the morphological changes of the catalyst layers. The combined impregnation with the transparent ionomer solution and opaque ionomer dispersion improved the cell performances. ABPBI Membrane and PBI Binder Polymer ABPBI film was prepared following the previously reported method [15]. The dried ABPBI film was immersed in aqueous 50 wt% phosphoric acid solution for 5 days to obtain the phosphoric acid-doped ABPBI membrane. PBI polymer powder was prepared by the method reported previously [16]. 3,3'-diaminobenzidine and isophthalic acid were used as the monomer, and the volume ratio of CF 3 SO 3 HtoCH 3 SO 3 H was 1:1. Ionomer Solution and Catalyst Slurry Two types of ionomer solution were made for the impregnation of Pt/C with the phosphoric acid/pbi ionomer. A transparent ionomer solution was prepared by the dissolution of PBI powder with phosphoric acid in trifluoroacetic acid (TFA), and an opaque ionomer dispersion was made by dropping excess iso-propanol into the transparent ionomer solution with stirring. The ratio of phosphoric acid to PBI was fixed at 6 H 3 PO 4 molecules per PBI repeating unit. The total weight percentage of H 3 PO 4 and PBI in the catalyst layer was defined as the ionomer content. Ultrasonic treatment was carried out with a commercial Pt/C catalyst (20 wt% Pt/C, E-tek) in the ionomer solution or dispersion to obtain a catalyst slurry with adequate viscosity for casting. The catalyst slurry impregnated with the phosphoric acid/pbi ionomer was cast on carbon paper (Toray, TGPH-090) using an adjustable doctor-blade, followed by evaporation of the solvent to obtain the ionomer-impregnated catalyst layers coated on the carbon paper. The catalyst layers impregnated with an ionomer content of x wt% using the transparent ionomer solution and opaque ionomer dispersion were denoted as xt and xc, respectively, and the catalyst layer with a mixture of both the transparent ionomer solution and opaque ionomer dispersion was denoted as TC (for example, 1T9C) and impregnated stepwise as follows; 1) impregnation of fresh catalyst with a small amount of transparent ionomer solution, 2) removal of TFA solvent, and 3) impregnation of pre-impregnated catalyst with opaque ionomer dispersion. The size of the ionomer in the transparent ionomer solution was measured by the dynamic light scattering method (Otsuka electronics, Photal ELS-8000), and the size of the ionomer agglomerate in the opaque ionomer dispersion was measured by the laser light scattering method (Malvern, Mastersizer-E). The pore size distribution of the catalyst layer and metallic surface area which was not covered by the ionomer were measured by nitrogen physisorption and CO chemisorption (Micromeritics, ASAP 2010), respectively. Single Cell Assembly The phosphoric acid-doped ABPBI membrane (thickness: 105 ± 5 µm) and two catalyst layers on carbon paper were assembled without hot pressing process. The anode was made of the 10T catalyst layer in each case, while the cathode was changed. The active electrode area for a single cell was 2 cm 2. The Pt loadings of the anode and cathode were 0.45 ± 0.05 and 0.50 ± 0.05 mg/cm 2, respectively.

3 852 Jeong-Hi Kim, Hyoung-Juhn Kim, Tae-Hoon Lim, and Ho-In Lee Figure 1. Pore size distributions of catalyst layer with different ionomer content using transparent ionomer solution. Figure 2. Particle size distribution in opaque ionomer dispersion. Hydrogen and oxygen were used as the fuel and oxidant, respectively. All gases were supplied without external humidification and the cell temperature was kept at 150 o C. The polarization curves were obtained using an electric loader (Daegil Electronics, EL500P). The impedance spectra were plotted using a potentiostat (Zahner, IM6) at different current densities in the frequency range of 100 khz 0.1 Hz, and the specific ohmic resistance was obtained from the impedance curve. Characterization of Ionomer and Catalyst Layer The size of the ionomer in the transparent ionomer solution measured by the dynamic light scattering method is between 8 and 20 nm. It shows a relatively narrow distribution and the minimum diameter is 8 nm. The pore size distributions of the catalyst layers with different ionomer binder contents in the electrode using the transparent ionomer solution are presented in Figure 1. Considering that the minimum size of the ionomer in the solution is 8 nm, it is supposed that the ionomer particles could not enter into those pores smaller than 8 nm. Accordingly, the Pt sites placed in the pores smaller than 8 nm could not be utilized for the electrochemical reaction, due to the lack of ionomer. The decrease of the pore volume by pores under 8 nm was due to the blocking of the pore openings by the ionomer particles, as shown in Figure 1. As the ionomer binder content in the catalyst layer increased to 1 and 2T, the pore volume of the pores between 4 30 nm increased compared with that of the Base (0 wt% ionomer content) and the maximum peak shifted from 50 to 35 nm. On the other hand, in the cases of the ionomer binder contents of 5, 10, and 20T, the pore volumes of all the pore sizes decreased with increasing ionomer binder content, and the maximum peak approached 50 nm which is similar to that of the Base. From these results, it is suggested that the primary pores with a diameter bigger than that of the Figure 3. Pore size distributions of catalyst layer with different ionomer content using opaque ionomer dispersion. ionomers were coated inside by the ionomer and became narrower, resulting in the shift of the maximum peak position to 35 nm without the blocking of the openings in the cases of the ionomer binder contents of 1 and 2T. However, the pores were gradually filled and their volume decreased in the case of the high ionomer contents from 5 to 20T without the shift of the maximum peak. The particle size distribution measured by the laser light scattering method in the opaque ionomer dispersion is presented in Figure 2. The ionomer agglomerates had a broad diameter range and the minimum size was 0.4 µm. The pore size distributions of the catalyst layers with different ionomer binder contents ranging from 1 to 10 wt% using the opaque ionomer dispersion are presented in Figure 3. Considering that the minimum size of the ionomer agglomerates is 0.4 µm, it is supposed that the ionomer agglomerates could not enter into the primary pores. As expected, the volumes of most of the primary pores with diameters of 5 30 nm did not decrease and only those of the secondary pores bigger than 30 nm decreased with increasing ionomer content. From these results, it is suggested that the ionomer agglomerates were located outside of the primary pores without blocking the pore openings and that the Pt sites inside the primary

4 Improvement of High Temperature Polymer Electrolyte Fuel Cell Performance with Phosphoric Acid-Doped Polybenzimidazole Ionomer Binder in Catalyst Layer 853 Figure 4. Metallic surface areas of catalysts with different ionomer content. pores did not come into contact with the ionomer agglomerates. The metallic surface areas measured by the CO chemisorption method with various ionomer contents are presented in Figure 4. Severe and mild reductions of the metallic surface area were observed with increasing ionomer content in the catalyst layers impregnated using the transparent ionomer solution and opaque ionomer dispersion, respectively. The different decay rates of the metallic surface area also supported the morphological differences, as mentioned above. Ionomers Inside the Catalyst Layers Schematic illustrations of the distribution of the ionomer inside the catalyst layers are shown in Figure 5. The Pt inside the primary pores is in contact with the ionomer and gas flow to make a 3-phase interface in the case of 1T (Figure 5(a)). However, the ionomer bridging to the membrane is so poor that the 3-phase interface inside the primary pores cannot be utilized for the electrochemical reaction. The ionomer bridging to the membrane becomes richer with increasing ionomer content in the case of 10T (Figure 5(b)). However, the filling of the primary pores with the ionomer is so much that the Pt inside the primary pores cannot make contact with the gas flow, resulting in a severe decrease of the 3-phase interface. The ionomer bridging to the membrane is sufficient and the primary pores are open in the case of 10C (Figure 5(c)). However, the Pt inside the primary pores Figure 6. Pore size distributions of catalyst layers using various impregnation methods. is not in contact with the ionomer and the 3-phase interface decreases severely. Eventually, the Pt inside the primary pores cannot be utilized for the electrochemical reaction in both the 10T and the 10C cases for different reasons. The fabrication method was modified in order to produce the 1T9C specimen, in an attempt to enhance the utilization of the Pt metals inside the primary pores, as shown in Figure 5(d). In this specimen, the thin ionomers inside the primary pores are connected with the thick ionomers outside the primary pores. At the same time, the primary pores are kept open to the gas flows. The transparent ionomer solution with an ionomer content of 1 wt% in the catalyst layer was first used to impregnate the inside of the primary pores without blocking the openings. Then, the opaque ionomer dispersion with an ionomer content of 9 wt% in the catalyst layer was used to impregnate the outside of the primary pores to keep the pores open. The pore size distribution of 1T9C is compared with those of 10T and 10C in Figure 6. In the case of 1T9C, most of the primary pores with diameters of 8 30 nm were successively impregnated on the outside without any severe plugging of the pore openings by the ionomer agglomerates. Considering that the inside of the primary pores was previously impregnated with a 1 wt% ionomer content, the 1T9C specimen was assumed to have optimum ionomer connections between the inside and outside of the primary pores. The SEM images of the cut sides are presented in Figure 7. The catalyst particles were covered with ionomer and well interconnected in Figure 5. Schematic illustrations for the distribution of ionomer in catalyst layer.

5 854 Jeong-Hi Kim, Hyoung-Juhn Kim, Tae-Hoon Lim, and Ho-In Lee Figure 7. SEM images of cutting side of catalyst layer impregnated by various impregnation methods. Acknowledgements This work was financially supported by the Fuel Cell Research Center of Korea Institute of Science and Technology and by the ERC Program of MOST/KOSEF (Grant No. R ). References Figure 8. ir free polarization curves of catalyst layers using various impregnation methods. the cases of 10C and 1T9C. The polarization curves are presented in Figure 8 with the ir-free voltage (cell voltage corrected for resistance). The cell performance with 10C was slightly better than that with 10T. However, the cell performance with 1T9C was much better than that with 10C. Conclusion The impregnation of the Pt/C catalyst with the phosphoric acid-doped polybenzimidazole ionomer for use in a high temperature polymer electrolyte fuel cell was performed using a transparent ionomer solution and opaque ionomer dispersion. The ionomers in the transparent ionomer solution and opaque ionomer dispersion were nano-sized and micro-sized, respectively. The informations about both the 3-phase interface inside the primary pores and the ionomer bridging to the membrane could be obtained by gradually increasing the ionomer content from an initially low content. The morphology changes of the catalyst layer were dependent on the size of the ionomer. An improvement of the cell performance was obtained using the combined impregnation of the transparent ionomer solution and opaque ionomer dispersion. 1. J. S. Wainright, M. H. Litt, and R. F. Savinell, in Handbook of Fuel Cells, W. Vielstichm, A. Lamm, and H. A. Gasteiger Ed., pp , John Wiley & Sons, Inc., 3 (2003). 2. D. Weng, J. S. Wainright, U. Landau, and R. F. Savinell, J. Electrochem. Soc., 143, 1260 (1996). 3. Y. M. Kim, S. H. Choi, H. C. Lee, M. Z. Hong, K. Kim, and H.-I. Lee, Electrochim. Acta, 49, 4787 (2004). 4. Y.-L. Ma, J. S. Wainright, M. H. Litt, and R. F. Savinell, J. Electrochem. Soc., 151, A8 (2004). 5. Q. Li, R. He, R. W. Berg, H. A. Hjuler, and N. J. Bjerrum, Solid State Ionics, 168, 177 (2004). 6. L. Xiao, H. Zhang, E. Scanlon, L. S. Ramanathan, E.-W. Choe, D. Rogers, T. Apple, and B. C. Benicewicz, Chem. Mater., 17, 5328 (2005). 7. C. Pan, R. He, Q. Li, J. O. Jensen, N. J. Bjerrum, H. A. Hjulmand, and A. B. Jensen, J. Power Sources, 145, 392 (2005). 8. J.-H. Kim, H. Y. Ha, I.-H. Oh, S.-A. Hong, H.-N. Kim, and H.-I. Lee, Electrochim. Acta, 50, 797 (2004). 9. O.-H. Kwon, H.-J. Ryu, and S.-Y. Jeoug, J. Ind. Eng. Chem., 12, 2, 306 (2006). 10. J. S. Wainright, J.-T. Wang, D. Weng, R. F. Savinell, and M. Litt, J. Electrochem. Soc., 142, L121 (1995). 11. H.-J. Kim and T.-H. Lim, J. Ind. Eng. Chem., 10, 1081 (2004). 12. J. A. Asensio, S. Borrós, and P. Gómez-Romero, J. Membr. Sci., 241, 89 (2004). 13. J. A. Asensio, S. Borrós, and P. Gómez-Romero, J. Electrochem. Soc., 151, A304 (2004). 14. J. A. Asensio and P. Gómez-Romero, Fuel Cells, 5,

6 Improvement of High Temperature Polymer Electrolyte Fuel Cell Performance with Phosphoric Acid-Doped Polybenzimidazole Ionomer Binder in Catalyst Layer (2005). 15. H.-J. Kim, S. Y. Cho, S. J. An, Y. C. Eun, J.-Y. Kim, H.-K. Yoon, H.-J. Kweon, and K. H. Yew, Macromol. Rapid Commun., 25, 894 (2004). 16. H.-J. Kim, S. J. An, J.-Y. Kim, J. K. Moon, S. Y. Cho, Y. C. Eun, H.-K. Yoon, Y. M. Park, H.-J. Kweon, and E.-M. Shin, Macromol. Rapid Commun., 25, 1410 (2004). 17. J.-T. Wang, R. F. Savinell, J. Wainright, M. Litt, and H. Yu., Electrochim. Acta, 41, 193 (1996). 18. J.-T. Wang, J. S. Wainright, R. F. Savinell, and M. Litt, J. Appl. Electrochem., 26, 751 (1996). 19. J. J. Fontanella, M. C. Wintersgill, J. S. Wainright, R. F. Savinell, and M. Litt, Electrochim. Acta, 43, 1289 (1998). 20. Q. Li, H. A. Hjuler, and N. J. Bjerrum, Electrochim. Acta, 45, 4219 (2000). 21. O. Savadogo and B. Xing, J. New Mater. Electrochem. Syst., 3, 345 (2000). 22. Q. Li, H. A. Hjuler, and N. J. Bjerrum, J. Appl. Electrochem., 31, 773 (2001). 23. Q.Li,R.He,J.Gao,J.O.Jensen,andN.J.Bjerrum, J. Electrochem. Soc., 150, A1599 (2003). 24. Q. Li, R. He, J. O. Jensen, and N. J. Bjerrum, Fuel Cells, 4, 147 (2004). 25. J. O. Jensen, Q. Li, R. He, C. Pan, and N. J. Bjerrum, J. Alloys. Compd., 404, 653 (2005). 26. F. Seland, T. Berning, B. Børresen, and R. Tunold, J. Power Sources, 160, 27 (2006). 27. J. Lobato, M. A. Rodrigo, J. J. Linares, and K. Scott, J. Power Sources, 157, 284 (2006).