Investigation of Solvent and Carbon Particles Behavior during Drying Process from Catalyst Ink to Catalyst Layer

Mass transport properties of catalyst layers (CLs) for proton exchange membrane fuel cells (PEMFCs) strongly depends on its structure. Structure control, therefore, is important to achieve higher cell performance. However, the CLs have fabricated by trial and error approach. Understanding mechanism and key parameters of the structure determination during the fabrication process is necessary. In this study, we especially focused on drying process from a catalytic slurry named catalyst ink to CL. Coating catalyst ink on a membrane or a substrate is one of the well-known fabrication methods of the CLs. Catalyst ink consists of components of the CLs, which is catalyst-supported carbon and ionomer, and solvents such as water and alcohol. During drying process from catalyst ink to CL, the solvent evaporates and then density of the components increase in the slurry. The components agglomerate or deposit, and finally the porous structure of the CL is formed. The porous structure affects cell performance [1]. Therefore, controlling behavior of components during the drying process has potential to form a further well-established porous structure. For the fundamental understanding, there are several attempt to characterize catalyst ink [2,3]. However, there are few studies about the drying process, although there is some simulation of the process [4].

To determine drying fabrication process and control the resultant structure of the CLs, revealing transport phenomena and dynamics of solvent, carbon and ionomer in the catalyst ink during drying fabrication process is necessary. For the fundamental understanding of the process, solvent evaporation and carbon agglomeration were detected in this study. The solvent evaporates during the drying process, thus the weight of the coated catalyst ink decreases as time goes along. Therefore, weight variation during the drying process was detected by using electric balance. According to evaporation, a density of carbon particles in the coated catalyst ink increases and the particles gradually form a network with each other. Carbon is prime electric conductive material in the catalyst ink. Therefore, electric conductivity was measured during the drying process by using microelectrode. Catalyst ink is coated on the silicon substrate with gold terminals. Four-terminal sensing was conducted by means of high-frequency resistance. For the simple evaluation, carbon black (ketjenblack, Lion Corp.) and Nafion^® ionomer (DE2020, Wako Pure Chemical Industries) were blended in the solvent which consist of water and propanol to form catalyst ink. Three types of the catalyst ink which differ in carbon density (1.8, 3.6 and 5.4wt.%) was synthesized. Ionomer to carbon ratio was 1.0. Although the weight measurement and resistance measurement were conducted independently, the same type of silicon substrate with SiO₂layer was used as a substrate of the coating. Drying condition was room temperature and humidity (ca. 20°C and 30%RH). Aimed thickness of the CLs was 10 μm in all kinds of catalyst ink according to the previous work [5].

Figure 1 shows weight variation during the drying process. Lower carbon content ink takes a longer time to dry completely. This is because the ink contains a larger amount of solvent to form the CL with aimed thickness. Figure 2 shows resistance variation during the drying process. The result indicates that lower carbon content ink takes a longer time to form carbon network. It is showed that the carbon content in the catalyst ink affects both drying and agglomeration behaviors. As a comparison of the two results of weight and resistance variation, catalyst ink with 1.8wt.% carbon shows much faster agglomeration time of less than 15 min than drying time of about 50 to 60 min. A low-carbon content ink can show independent agglomeration behavior from drying behavior although high-carbon content inks show the relatively correlated behavior of evaporation and carbon agglomeration.

References

[1] T. Suzuki, R. Hashizume, M. Hayase, J. Power Sources, 286 (2015) 109-117.

[2] Z. Xie, T. Navessin, X. Zhao, M. Adachi, S. Holdcroft, T. Mashio, A. Ohma, K. Shinohara, ECS Trans., 16 (2008) 1811-1816.

[3] F. Xu, H. Zhang, D. Ho, J. Ilavsky, M. Justics, H. Petrache, L. Stanciu, J. Xie, ECS Trans., 41 (2011) 637-645.

[4] T. Munekata, T. Suzuki, S. Yamakawa, R. Asahi, Physical Review E, 88 (2013).

[5] T. Suzuki, S. Tsushima, S. Hirai, J. Power Sources, 233 (2013) 269-276.