Understanding Ink Dispersion as an Essential Tool in Large-Scale CCM Manufacturing

EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476303 B1711 Page 1/7 B1711 Understanding Ink Dispersion as an Essential Tool in Large-Scale CCM Manufacturing Jessica Chamier*, Douriya Khedaioui, Chloé Pignol, Octavio Graniel Tamayo Symbio, High Valued Materials, 10 Rue Specia, 69190 Saint Fons/France; *Contact corresponding authors: www.EFCF.com/ContactRequest Abstract At the basic R&D scale (<50 ml), the primary focus is on preserving the properties of the catalyst and ionomer and their interactions within a given solvent system. These inks are applied using single-sheet techniques such as bar coating, doctor blade, and Mayer rod coating—methods that are more forgiving in terms of viscosity, particle size distribution, and ink rigidity. However, at prototype scale (2L) and during second-phase validation, the rheological properties of the ink, ensuring adaptability to roll-to-roll (R2R) manufacturing, become a priority alongside safety considerations. Achieving consistent, scalable, and safe prototype-scale ink dispersions is a complex process requiring careful evaluation of available technologies. This paper explores the influence of different dispersion technologies on ink performance, particularly focusing on how particle size distribution (PSD), viscosity, and ionomer-catalyst interactions impact coating quality and electrochemical performance. We demonstrate that both grinding (BM) and high-pressure dispersion (HPD) techniques achieve desirable PSDs and coating homogeneity. However, differences in viscosity significantly affect catalyst layer morphology and high-current-density performance. Our findings highlight that beyond achieving target PSD, the drying behaviour and ink flow—strongly influenced by viscosity and dispersion method—are critical to optimizing the final electrode structure and performance. EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476303 B1711 Page 2/7 Introduction The commercial viability of proton exchange membrane fuel cells (PEMFCs) hinges on balancing cost with scalable production. Increasing power density through improved electrode performance is a priority. While the effects of catalyst and ionomer selection are well-documented, less attention has been given to how dispersion technology influences ink properties and downstream manufacturability. Scaling production is essential not just for meeting demand but also for reducing system costs through economies of scale [1]. Most prior studies [2], [3], [4] focus on small-scale formulations where ink behavior is not necessarily representative of roll-to-roll (R2R) coating performance. Techniques such as sonication are not scalable, while shear mixing, grinding (BM), and high-pressure dispersion (HPD) offer more realistic industrial relevance. This study aims to bridge that gap by comparing these dispersion technologies at the prototype scale, providing insight into how ink structure influences both coating behavior and electrochemical performance. We focus on three critical ink parameters: particle size distribution (PSD), viscosity, and ionomer-catalyst interaction. These govern not only the ink's coatability but also the morphology and function of the resulting catalyst layer. Importantly, we show that solvent system and dispersion technique interact in complex ways to shape ink structure and film quality. Our goal is to provide a practical framework for selecting dispersion methods that align with industrial-scale coating and CCM production. 1. Experiments Catalyst inks were prepared with an alcohol content ranging between 60 to 20% (v/v) and a solid content of 10% (g). Pt/C was dispersed in PFSA ionomer at a fixed I/C ratio. The ink was dispersed using bead milling (BM), high pressure dispersion (HPD) and shear mixing respectively. Small batches of the ink were coated onto proton exchange membranes (PEMs) in the laboratory using bar coater and prototype coated substrates were coated with a R2R slot-die. The total membrane electrode assembly (MEA) loading was 0.4 mgPt/cm² sandwiched between gas diffusion electrodes. 2. Results 2.1 Ink dispersion Particle size distribution (PSD) is a key metric when evaluating dispersion performance. Smaller PSDs promote more uniform catalyst packing and smoother coatings. In cathode catalyst layers with loadings as low as 0.2 mgPt/cm², film thickness ranges from 6 to 11 μm. For such thin layers, PSDs exceeding 10 μm are undesirable and those above 100 μm may block slot-die coating heads. The various dispersion techniques studied here resulted in unique PSD profiles presented in Figure 1. Only the PSD larger then 0,1 µm should be considered based on the limitation of instrument detection. EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476303 B1711 Page 3/7 Figure 1: PSD of the ink prepared using the various dispersion technologies in A) water rich (80% water) and B) alcohol rich (40% water) formulations. Shear mixing is more effective in reducing particle size in water-rich systems (Dx90, 8.73µm) but is less effective with alcohol-rich solvents (Dx90,13.6). All round, HPD and BM result in smaller PSD compared to shear mixing, as expected, due to the ability of these technologies to break aggregate structures. BM or grinding technology is the most popular dispersion technology, due to its reliability in scalability and its effectiveness in achieving lower PSD. Even smaller PSD can be achieved by increasing either the time of agitation or the flowrate. BM applies high localized shear to break down agglomerates, while HPD leverages cavitation high velocity impact and shear effects for consistent size reduction. Alcohol-rich formulations naturally exhibit different PSD behaviour due to changes in ionomer aggregation. Alcohol reduces the dielectric constant of the solvent system, suppressing ionization of –SO₃H groups and weakening electrostatic repulsion between charged chains. Shear mixing is therefore significantly less successful in deagglomeration in alcohol rich inks as it strongly relies on the hydrodynamic forces to tear the flocculation structures. In contrast, BM and HPD disrupt these aggregates more thoroughly. The Dx90 for the alcohol rich and water rich formulations for HPD are remarkably similar at 1.71 versus 1.70 µm respectively. The same is observed for BM with 2.72 µm (water rich) versus 2.01 µm. The inks are well within the acceptable range for slot die coating. The PSD, however, is quite unique, illustrating the different deagglomeration approaches of the dispersion technologies. Figure 2A shows that HPD and BM significantly lower viscosity in water-rich inks, bringing them close to Newtonian behavior. In HPD, water has high surface tension and is therefore more forceful cavitation during high-pressure dispersion resulting in a more efficient particle size reduction. Smaller particle sizes may have higher surface area with a better distributed charge, less agglomeration, and therefore lower bulk density. In a similar manner, BM uses localized, high shear forces to break down particle size as well as weak gel-like networks, and overall lower the shear viscosity. Smaller particles, as is the case for BM and HPD, have higher surface charge density, stronger electrostatic repulsion and therefore reduced flocculation. EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476303 B1711 Page 4/7 Figure 2: A) The viscosity of the water rich ink in various dispersion technologies. B) The viscosity before and after BM and high-pressure dispersion of alcohol rich inks. As expected, the viscosities of the inks are higher with increased alcohol content. With the addition of less polar alcohol, sulfonic groups are not as well solvated and the ionomer experiences stronger inter-chain interactions and collapses into a more compact, aggregated form. However, Figure 2B reveals an opposite effect in alcohol-rich inks: BM and HPD increase viscosity. These technologies expose more ionic groups and surface area, promoting stronger inter-chain interactions in lower dielectric media. As a result, chain collapse and microdomain formation increase, forming a denser network that resists flow. This outcome underscores the dual influence of dispersion and solvent composition on ink rheology. The interaction of Nafion with polar solvents and its phase-separated morphology is critical to understanding dispersion behavior [5]. 2.2 Coating Slot-die coating is highly sensitive to ink rheology. Inks from BM and HPD show smoother, more homogeneous coatings compared to shear-mixed inks (Figure 3). The improved surface morphology is due to smaller PSD, lower surface tension, and better leveling in alcohol-rich systems. Higher viscosity from BM and HPD also helps suppress flow instabilities during drying [6], [7], [8]. Figure 3: Surface morphology of substrates coated with alcohol rich ink using R2R slotdie (68 x magnification). Figure 4 considers the crack density of water-rich coatings of the various dispersed inks. Water rich inks are more prone to cracking due to the high surface tension of water. The water pulls the particles together during evaporation, and as water limits chain entanglement, results in weaker adhesion during drying. The link between surface cracking and electrochemical performance is still to be established and will in most likelihood not be linear. Similar correlations between surface cracking and MEA durability have been noted in earlier studies [9]. EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476303 B1711 Page 5/7 Figure 4: Crack density for water-rich coated inks dispersed with shear, BM and HPD (100 x magnification). In water-rich coatings, crack density (Figure 4) was lowest for BM and HPD. This reduction is linked to smaller particle size (Figure 1) and better substrate wetting, which minimizes internal stress and allows for rearrangement during drying. In contrast, shear-mixed inks exhibited higher crack density due to larger particles and stronger capillary forces. Inks with near-Newtonian behaviour are less ideal for slot-die coating. They do not respond to shear, which hampers self-levelling and increases the risk of coating defects like ribbing. These challenges become more pronounced with low solids content in water-rich systems. 2.3 Electrochemical evaluation The performance curves were measured multiple times (>3) with the highest performance recorded in Figure 5. The electrochemical surface area of the MEAs varied between 51 and 56 cm²/g, showing no relation to the dispersion technology. The high frequency resistance (HFR) and hydrogen crossover was equivalent for all MEAs. Polarization curves (Figure 5A) show that for water-rich MEAs, electrochemical performance at 1 A/cm² is comparable across all dispersion methods. However, at higher current densities, BM-based electrodes outperform others. Variation in high current density performance is related to the catalyst layer structure. While HPD achieves smaller PSD than shear mixing, this does not directly translate into better performance, likely due to surface cracking and porosity affecting mass transport. Figure 5: Polarization curves for MEAs A) prepared from water rich inks and B) prepared from alcohol rich inks. EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476303 B1711 Page 6/7 Figure 5B illustrates that the MEA performance is more sensitive to PSD and surface structure [10], [12], [12] compared to water-rich coated substrates. BM and HPD result in nearly flawless coatings and similar performance up to 1.4 A/cm². Above this, BM outperforms HPD, possibly due to its lower ink viscosity, which enables better ionomercatalyst contact. Higher viscosity in HPD inks may cause local inhomogeneities as surface drying limits ionomer mobility. Further studies, including BET surface area and porosity analysis, are needed to confirm this hypothesis. These findings reinforce that high-current-density performance is governed not only by PSD but also by how ink viscosity and drying dynamics shape the catalyst layer microstructure. 3.Conclusions In the evolution of dispersion technologies, it has become evident that shear mixing and sonication are less suited for scaled catalyst ink production that meets industrial requirements. There is now a competitive landscape between grinding and high-pressure dispersion technologies. These methods offer different advantages, particularly in handling both waterand alcohol-rich solvent systems. Shear mixing technologies can be optimized for water-rich formulations and serve as a cost-effective entry point. However, BM and HPD are more effective across a broader solvent range. Both achieve smaller particle sizes with higher charge density, and their viscosities can be finely tuned by adjusting alcohol content. Once the target PSD is achieved, it is the ink's flow and drying behaviour—shaped by its rheology and dispersion quality—that dictate the resulting electrode microstructure and ionic pathways. Our study underscores that successful CCM production requires more than just achieving the right particle size. The ink's ability to coat evenly and dry into a cohesive, highperforming electrode layer hinges on managing viscosity and dispersion strategy. These parameters ultimately influence mass transport, triple-phase boundary formation, and electrochemical performance at high current densities. Future work should focus on optimizing drying conditions and porosity control to complement the benefits of advanced dispersion technologies. Previous findings have shown that non-uniform ionomer distribution can impair ionic pathways and triple-phase boundary formation [11]. References [1] Accelerazero, "How Economies of Scale Will Increase Hydrogen Technologies Accessibility and Adoption," Accelerazero, 2023. [Online]. Available: https://www.accelerazero.com/news/how-economies-scale-will-increase-hydrogentechnologies-accessibility-and-adoption [2] Yang, D., Guo, Y., Tang, H., Yang, D., Ming, P., Zhang, C., Li, B. and Zhu, S., 2022. Effect of rheological properties of catalyst slurry on the structure of catalyst layer in PEMFC. International Journal of Hydrogen Energy, 47(14), pp.8956-8964; [3] Li, B., Liu, Y., Guo, Y., Yang, D., Yang, D., Ming, P. and Zhang, C., 2021. Controlling the microscopic morphology and permeability of catalyst layers in proton exchange membrane fuel cells by adjusting catalyst ink agglomerates. International Journal of Hydrogen Energy, 46(63), pp.32215-32225; [4] Bapat, S., Fricke, S., Kohsakowski, S., Goessling, S., Peinecke, V. and Segets, D., 2020. Tailoring of electrocatalyst inks for performance enhancement in proton exchange membrane fuel cells. ECS Transactions, 97(7), p.651. [5] K. A. Mauritz and R. B. Moore, "State of understanding of Nafion," Chemical Reviews, vol. 104, no. 10, pp. 4535–4586, 2004. EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476303 B1711 Page 7/7 [6] T. K. Nguyen, K. Sivula, and T. Meyer, "Slot die coating of nanostructured materials for energy applications," Advanced Materials Interfaces, vol. 5, no. 1, p. 1700680, 2018. [7] C. N. Hoth, P. Schilinsky, S. A. Choulis, S. F. Tedde, and C. J. Brabec, "Long-term stability of P3HT:PCBM solar cells prepared by hot slot-die coating from a chlorinefree solvent system," Solar Energy Materials and Solar Cells, vol. 95, no. 5, pp. 1344– 1348, 2011. [8] F. C. Krebs, "Fabrication and processing of polymer solar cells: A review of printing and coating techniques," Solar Energy Materials and Solar Cells, vol. 93, no. 4, pp. 394– 412, 2009. [9] T. A. Greszler, J. D. Boncella, and S. A. Mauger, "Understanding membrane electrode assembly performance and durability impacts of catalyst layer cracking," Journal of The Electrochemical Society, vol. 159, no. 6, pp. F831–F840, 2012. [10] M. A. Yandrasits, D. M. Karl, and M. K. Debe, "Properties and characterization of Aquivion short side chain ionomer dispersions," Journal of The Electrochemical Society, vol. 162, no. 6, pp. F559–F565, 2015. [11] Z. Shi, J. Zhang, and R. Liu, "Ionomer distribution and its impact on PEM fuel cell catalyst layer structure and performance," Electrochimica Acta, vol. 258, pp. 1162– 1174, 2017. [12] Y. H. Cho, J. S. Park, S. J. Yoo, and H. Kim, "High current density operation of PEM fuel cells with ultra-thin catalyst layers," Journal of Power Sources, vol. 268, pp. 180– 185, 2014. [13] S. Takamuku, K. Ueno, and K. Dokko, "Colloidal behavior of perfluorosulfonic acid ionomers in mixed solvents," Langmuir, vol. 24, no. 22, pp. 12231–12239, 2008. [14] V. A. Sethuraman, L. J. O’Neill, C. P. Huang, and J. W. Weidner, "Role of ionomer content and structure on catalyst layer properties," Journal of Power Sources, vol. 195, no. 12, pp. 4035–4042, 2010. [15] S. Litster and G. McLean, "PEM fuel cell electrodes," Journal of Power Sources, vol. 130, no. 1–2, pp. 61–76, 2004. [16] U.S. Department of Energy, "Technical Targets for Fuel Cell Systems for Heavy-Duty Vehicles," 2023. [Online]. Available: https://www.energy.gov/eere/fuelcells/doetechnical-targets-fuel-cell-systems-heavy-duty-vehicles Keywords: EFCF2025, H2, LowTemp. Fuel Cells & Electrolysers, Dispersion, Industrial scale, ink Remark: This work is licensed under Creative Commons Attribution 4.0 International