Investigation of Solvent Effects on the Dispersion of ...

ECS Transactions, 75 (14) 361-371 (2016) 10.1149/07514.0361ecst ©The Electrochemical Society

Investigation of Solvent Effects on the Dispersion of Carbon Agglomerates and Nafion Ionomer Particles in Catalyst Inks Using Ultra Small Angle X-Ray Scattering Method Fan Yanga,b, Le Xinb, Aytekin Uzunogluc, Lia Stanciuc,d, Jan Ilavskye, Steven Sona, and Jian Xieb a

School of Mechanical Engineering, Purdue University, West Lafayette, Indiana, 47907 b Department of Mechanical Engineering, Indiana University- Purdue University Indianapolis, Indianapolis, Indiana, 46202 c School of Materials Engineering, Purdue University, West Lafayette, Indiana 47097 d Weldon School of Biomedical Engineering, Purdue University, West Lafayette, Indiana, 47097 e X-Ray Science Division, Argonne National Laboratory, Argonne, Illinois, 60439

Ink formulation is important to improve the performance of membrane electrode assembly (MEA) in polymer electrode fuel cells (PEMFCs). In this study, ultra-small angle x-ray scattering method was applied to study the dispersion of carbon black in different solvent to get a better understanding of the influence of the solvent on the dispersion of carbon black. The dimension of carbon black aggregates before and after adding Nafion ionomer was compared in different solvent. Dielectric constant and viscosity were found to be the key factors that affect the dispersion of carbon black in solvents. Carbon black particles in a solvent with higher dielectric constant will tend to form larger aggregates after adding Nafion ionomer. The higher viscosity of a solvent will prevent the aggregation of carbon black particles after the addition of Nafion ionomer.

Introduction Polymer electrolyte membrane fuel cells (PEMFCs) have received increasing attention as the best candidates for realizing the hydrogen economy, not only for zero emissions, but also for higher energy efficiency (1-4). However, currently, PEFCs are still not commercially ready for the market because of their high cost and moderate performance. The membrane electrode assembly (MEA) is the core component of PEMFCs and directly determines the performance and cost of PEMFCs. Hence, recent research has been focused on developing a high-performance, low-cost, and durable MEA (5, 6). Several aspects such as the activity of Pt, the interaction between Pt and catalyst support, stability of catalyst support and the dispersion of catalyst

361

Downloaded on 2016-09-27 to IP 149.166.233.236 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

ECS Transactions, 75 (14) 361-371 (2016)

support will influence the performance of MEA (7-10). One of the most commonly used catalyst is carbon black, so the dispersion of the carbon black aggregates and the ionomer particles in a solvent is the first and very critical step to achieving a fine structure of the catalyst layer. In order to prepare an appropriate ink, the carbon black powder must be well dispersed in the solvent. Therefore, to verify whether a formulated catalyst ink disperses the carbon black aggregates and the ionomer particles well is of great importance. A typical MEA is like a sandwich structure with a composite of a polymer electrolyte membrane and two porous catalyst layers. The catalyst layer is made of recasting Nafion ionomer with carbon aggregates together. The platinum catalyst nanoparticles reside on the surface of carbon black aggregates. In order to get maximum property, the catalyst layer needs to have (i) the plentiful boundary of Nafion/ catalyst for gas reactions, (ii) the appropriate pore structure to allow reactant gas diffusion and water dissipation (8, 11), and (iii) optimized Nafion and the carbon aggregates network offers the structural integrity for the catalyst layer, and provides a proton conduction path (12). There are several new techniques for the fabrication of catalyst layers/MEAs being explored, such as doctor-blade spreading, electrophoretic deposition (EPD), sputtering, electrospraying, and rolling (13-15). A method developed at Los Alamos National Laboratory (LANL) using the “thin film decal” process (16) is still widely used today. The essential part of this method is dispersing the catalyst powder and Nafion ionomer particles in a liquid media first, and then forming a porous solid catalyst layer. There are two key points keeping the special structure of catalyst layers. One is the dispersion of the Pt/C aggregates and the ionomer particles in the catalyst ink and the other is the evaporation of the solvent in the catalyst ink to form the porous solid catalyst layer (the condensation of the catalyst ink) (17). In the well done system, the Pt/C catalyst powder is well dispersed and forms small aggregates (on a 100 nm scale) and the ionomer particles form small diameter strips or rods. The final porous, rigid solid catalyst layer/MEA promises the nanoscale catalyst particles be contacted uniformly with each other. From above description, in a catalyst ink, the solvent plays a critical role for the dispersion of both the Nafion ionomer and the carbon black particles. The polarity of the solvent is indicated by its dielectric constant. For instance, when the Nafion ionomer is mixed with various kinds of organic solvents, the geometry and the micro-structure of the Nafion ionomer particles tends to change (18). These changes are mainly due to the dielectric constant (ε) of the organic solvent. Therefore, it is of great interest to study the effects of different organic solvents on the dispersion of Nafion ionomer and Pt/C catalyst particles. It’s of great importance to certify whether a formulated catalyst ink disperses the Pt/C aggregates and the ionomer particles well. Instead of a “trial-and-error” approach--measuring the MEA fuel cell performance, designing “ideal” ink guided by

362



well known about the catalyst ink system is a more accurate and effective way. Even though some research has been conducted to investigate the Nafion solution microstructure via dynamic light scattering (DLS) (19), small-angle X-ray scattering (SAXA) (20), and electron spin resonance (ESR) (21). These are only from the curve fitting of the scattering data without direct observation. In 2010, our group first combined the ultra-small angle x-ray scattering and cryo-TEM technology to explore the size and geometry of both the Pt/C aggregates and the ionomer particles in catalyst inks and the effects on dispersion in catalyst ink (17). Ultra-small angle x-ray scattering can reflect the size and geometry of Nafion ionomer particles and Pt/C aggregates in liquid media with overcoming absorption problem. With direct observation from cryo-TEM, the USAXS data could be well validated. In this work, three different solvents, ethanol, glycerol and 20% isopropanol aqueous solution were used to study the solvent effects on the catalyst supports in the ink systems. The effects of the solvent on the dispersion of both Nafion ionomer particles and carbon black particles were studied using ultra small angle x-ray scattering (USAXS). Sizes change of carbon particles in different solvents before and after adding Nafion ionomer were studied to get a comprehensive understanding of the influence of the solvent on the dispersion of catalyst and Nafion ionomer. Experimental Materials and sample preparation Samples were prepared by mixing carbon black, solvent and 5% Nafion solution based on requirement. The inks were ultrasonicated for 30 mins in ice water and stirred vigorously overnight to achieve a uniform suspension. The sample information is listed in Table I. TABLE I. Sample Preparation for USAXS Measurements sample

Abbreviation

XC72 (mg)

Nafion (ul)

Solvent (ml)

XC72+Ethanol

CB_E

30

0.9333

0

CB_E_NF

30

0.9333

0

XC72+Ethanol+Nafion XC72+IPA solvent

CB_I

0

0.9333

0

XC72+IPA solvent+Nafion

CB_I_NF

0.9333

0.9333

0

CB+Glycerol

CB_G

0

0

0.9333

CB+Glycerol+Nafion

CB_G_NF

0.9333

0

0.9333

Characterization Ultra-Small Angle X-Ray Scattering

363



USAXS measurements were conducted by the versatile USAXS instrument at the Advanced Photon Source (APS), Argonne National Laboratory. The USAXS has angular and energy resolutions of the order of 10-4, accurate and repeatable x-ray energy tunability over its operational energy range from 8 to 18 keV, a dynamic intensity range of 108 to 109, and a scattering vector range from 0.0001 to 1 Å-1. The scattering data are well described by a global unified fit analysis. Porod and Guinier scattering regimes are employed in the fitting analysis. The deconvoluted USAXS curves were analyzed by comparing them with scattering model functions implemented in an Igor Pro based software package “Irena” (17, 22). Results and Discussion In order to investigate solvent effects on the dispersion of carbon agglomerates, three ink systems were measured by USAXS for both before and after adding Nafion solution. Using a systematic approach, each component in a catalyst ink was added into a solvent, and the size and microgeometry were studied using USAXS via curve fitting. Multilevel fitting was employed and the results shows each level have had a different particle size and geometry. The typical X-ray scattering result of carbon black in ethanol aggregate system is shown in figure 1. There are three different power slopes in the curve, which divide the total scattering pattern into three distinctive regions corresponding to three fitting levels. The global unified function fitting result in high q range (0.02 – 0.06 A-1) shows that the radius of gyration of the first level is 5.86 nm which is corresponding to the diameter D= 11.72 nm with power slope P = 2.62. This level is related to the scattering of the isolated primary particle of carbon black. In the middle q range (0.002 - 0.02 A-1), the fitting result shows the radius of gyration for the 2nd level is 103.6 nm with power slope P of 3.05. This level represents the dimension of the carbon black aggregates. The 3rd level fitting of this fractal system has an infinity radius of gyration which means that the size of the carbon agglomerates is too large which exceeds the limit of the instrument.

364



Figure 1. USAXS measured and fitting result of carbon black ethanol ink system before adding Nafion ionomer After adding Nafion ionomer, significant differences can be observed in the USAXS fitting result which is shown in figure. Two fitting levels instead of three can be observed in the result of CB_E_NF system. Compared to the result of CB_E system, the level in the high q range disappears due to the interaction between carbon black primary particles and Nafion ionomer which will be discussed later. The first level fitting in the high q range shows that the radius of gyration of carbon black particles is 130.9 nm, which represents the aggregate of the carbon black. The low q range fitting still shows an infinity radius of gyration corresponding to the large agglomerate of carbon black which is also similar to the result of CB_E system. To compare the influence of difference solvent on the dispersion of carbon black, glycerol which has a different dielectric constant is also used as solvent in the carbon black fractal system. As can be seen in figure 3, three different power slopes can be found in the fitting range, which leads a two-level fitting to be applied to fit the CB_G aggregate system. In the high q region, the fitting result shows that the radius of gyration of the 1st level fitting is 5.47 nm which is corresponding to the primary carbon black particles. In the medium q range, it can be found that the radius of gyration of the carbon black aggregates is 213.8 nm, which is larger than the one in the CB_E system. The 3rd level fitting here is still from the scattering of the large agglomerate of carbon black particles.

365



Figure 2. USAXS measured and fitting result of carbon black ethanol ink system after adding Nafion ionomer

Figure 3. USAXS measured and fitting result of carbon black glycerol ink system before adding Nafion ionomer Similarly, Nafion ionomer was also added into CB_G system to compare the influence of Nafion ionomer on the dispersion of carbon black in glycerol solvent. Different from CB_E system after adding Nafion ionomer, three different power

366



slopes instead of two can still observed, from which we can deconvolute three different fitting levels. The first level fitting in the high q region still results from the scattering of the carbon black primary particles which has the radius of gyration of 3.66 nm. The fitting of the 2nd level is carried out in the medium q range (between 0.002 A-1 to 0.02 A-1), in which we can observe that the radius of gyration of the carbon black aggregates in this level is 156.8 nm. Again, the third fitting level in low q range is still related to the scattering of the large carbon black agglomerates in glycerol solvent.

Figure 4. USAXS measured and fitting result of carbon black glycerol ink system after adding Nafion ionomer Isopropanol aqueous solution (20 vol%) was also used to compare the behavior of carbon black particles in different solvents. Similar to other ink systems without Nafion ionomer, three fitting levels can still be applied to fit the scattering curve according to three different power slopes which can be seen in figure 5. The first level fitting in high q range which represents carbon black primary particles with the radius of gyration of 11.87 nm. In the medium q range, the second level fitting result shows that the radius of gyration of carbon black aggregates is 122.6 nm, which is formed by carbon black primary particles. The fitting result of the third level in low q region with infinite radius of gyration is the result of the scattering of large carbon black agglomerates.

367



Figure 5. USAXS measured and fitting result of carbon black isopropanol solution ink system before adding Nafion ionomer The scattering curve and the fitting result of CB_I aggregate system after adding Nafion ionomer is shown in figure 6. CB_I_NF system is fitted by three levels. The radius of gyration from the fitting result of the first level is 170.0 nm which represents the dimension of the carbon black aggregates. The second fitting level in low q region with the power slope of 3.41 results from the scattering of the large carbon black agglomerates.

Figure 6. USAXS measured and fitting result of carbon black isopropanol solution ink system after adding Nafion ionomer

368



The size of the carbon black aggregates affects the performance and structure of MEA most, so we focus on the influence of different solvent on the radius of gyration of carbon black aggregates. The radius of gyration of carbon black aggregates in different solvent system with different dielectric constant is shown in table 2. By comparing the size change of carbon black aggregates in solvents with different dielectric constant, it can be found that the radius of gyration of carbon black aggregates increased 26.4% after adding Nafion ionomer in ethanol solvent and increased 38.7% in isopropanol aqueous solution. This result shows that the solvent with higher dielectric constant will facilitate the aggregating process of carbon black particles after Nafion ionomer. Following this trend, carbon black aggregates in glycerol, which has the dielectric constant between isopropanol aqueous solution and ethanol should have a size increase between 26.4% and 38.7% after Nafion ionomer. However, the USAXS fitting result shows that the Rg of carbon black aggregates in glycerol decreases by 26.7% after adding Nafion ionomer. It can be speculated that there are some other factors that influence the particles size of carbon black aggregates. By comparing other physical properties of glycerol, isopropanol aqueous solution and glycerol, it can be found that viscosity is a key factor that may influence the behavior of carbon black aggregates after adding Nafion ionomer. It can be found that isopropanol aqueous solution and ethanol have similar viscosity at room temperature, however, glycerol has a viscosity a thousand times larger than them. Higher viscosity will impose a strong friction that will prevent the aggregation of carbon black particles. It is the reason why the increase of the size of carbon black aggregates was not observed after the addition of Nafion ionomer from the USAXS fitting result. TABLE II. Dielectric Constant of Different Solvent and Rg of Carbon Black Aggregates Sample CB_E CB_E_NF CB_G CB_G_NF CB_I CB_I_NF

Dielectric Constant 23.8 42.5 68

Rg of Aggregates (nm) 103.6 130.9 213.8 156.8 122.6 170.0

Change after Adding Nafion +26.4% -26.7% +38.7%

Conclusion The USAXS measurement has been successfully applied to study the dispersion of carbon black particles in different solvent systems. Carbon black aggregates is the major component that will influence the performance and structure of MEA. Two key factors that will affect the behavior of carbon black after adding Nafion ionomer. In solvents with similar viscosity (isopropanol aqueous solution and ethanol), the radius of gyration of carbon black aggregates will increase more after adding Nafion

369



ionomer in solvent with higher dielectric constant. However, in the solvent with a much higher viscosity, the high friction from the solvent will prevent the carbon black particles from aggregating, which makes the radius of gyration of carbon black aggregates in glycerol decrease. From the discussion, USAXS is a useful tool to study the dispersion of carbon black in different solvent. With the help of it, we can discover a better ink formulation to improve the performance of PEFCs.

References 1. H. A. Gasteiger, S. S. Kocha, B. Sompalli and F. T. Wagner, Appl. Catal., B, 56, 9 (2005). 2. F. T. Wagner, B. Lakshmanan and M. F. Mathias, J. Phys. Chem. Lett., 1, 2204 (2010). 3. H. A. Gasteiger and N. M. Markovic, Science, 324, 48 (2009). 4. C. C. Chan, Proc. IEEE, 95, 704 (2007). 5. C. Houchins, G. J. Kleen, J. S. Spendelow, J. Kopasz, D. Peterson, N. L. Garland, D. L. Ho, J. Marcinkoski, K. E. Martin and R. Tyler, Membranes, 2, 855 (2012). 6. Y. Shao, G. Yin and Y. Gao, J. Power Sources, 171, 558 (2007). 7. M.-x. Wang, F. Xu, Q. Liu, H.-f. Sun, R.-h. Cheng, H. He, E. A. Stach and J. Xie, Carbon, 49, 256 (2011). 8. Z.-F. Li, L. Xin, F. Yang, Y. Liu, Y. Liu, H. Zhang, L. Stanciu and J. Xie, Nano Energy, 16, 281 (2015). 9. M.-x. Wang, F. Xu, H.-f. Sun, Q. Liu, K. Artyushkova, E. A. Stach and J. Xie, Electrochim. Acta, 56, 2566 (2011). 10. L. Xin, F. Yang, S. Rasouli, Y. Qiu, Z.-F. Li, A. Uzunoglu, C.-J. Sun, Y. Liu, P. Ferreira, W. Li, Y. Ren, L. A. Stanciu and J. Xie, ACS Catal., 6, 2642 (2016). 11. Z.-F. Li, H. Zhang, F. Yang, L. Stanciu and J. Xie, ECS Trans., 64, 131 (2014). 12. A. L. Dicks, J. Power Sources, 156, 128 (2006). 13. E. Taylor, E. Anderson and N. Vilambi, J. Electrochem. Soc., 139, L45 (1992). 14. S. Hirano, J. Kim and S. Srinivasan, Electrochim. Acta., 42, 1587 (1997). 15. J.-H. Wee, K.-Y. Lee and S. H. Kim, J. Power Sources, 165, 667 (2007). 16. M. S. Wilson and S. Gottesfeld, J. Electrochem. Soc., 139, L28 (1992). 17. F. Xu, H. Zhang, J. Ilavsky, L. Stanciu, D. Ho, M. J. Justice, H. I. Petrache and J. Xie, Langmuir, 26, 19199 (2010). 18. M. Uchida, Y. Aoyama, N. Eda and A. Ohta, J. Electrochem. Soc., 142, 463 (1995). 19. S. Jiang, K.-Q. Xia and G. Xu, Macromolecules, 34, 7783 (2001). 20. G. Gebel, Polymer, 41, 5829 (2000). 21. S.-J. Lee, T. L. Yu, H.-L. Lin, W.-H. Liu and C.-L. Lai, Polymer, 45, 2853 (2004).

370



22. J. Ilavsky and P. R. Jemian, J. Appl. Crystallogr., 42, 347 (2009).

371
