From Stochastic to Deterministic Transport: Advanced Gas Diffusion Layers for Polymer Electrolyte Fuel Cells

Abstract

The upcoming energy transition from fossil to renewable, clean energy requires multiple energy conversion and storage solutions to increase efficiency and reduce emissions in all energy-consuming sectors. Hydrogen, when produced from green electricity, can play a crucial role in this transition. Polymer electrolyte fuel cells (PEFCs) are the most efficient energy converters to re-electrify hydrogen, combining it with oxygen from the air to produce only water, electrical energy, and heat. They, therefore, are applied for both mobile and stationary applications. However, the high system cost of PEFCs has limited their wider use in the energy and mobility markets. Decreasing the cost per installed power requires indispensably material developments of all essential fuel cell components. In a PEFC, hydrogen and oxygen are supplied through gas channels from both sides of the polymer electrolyte membrane at the center. The gas diffusion layers (GDLs) distribute the feed gases homogeneously across the catalyst layers located at the membrane interfaces. Water accumulation in the porous gas diffusion layers increases the oxygen transport resistance to the cathode catalyst and thus prevents current PEFC technology from reaching higher current and power densities. This thesis thoroughly investigates water and gas transport in conventional and advanced GDL materials. It highlights the benefits of deterministic structures, as compared to today’s stochastic materials, leading product water more direct to the gas channel and thereby reducing the liquid water saturation in the pore space. On the one hand, deterministic modifications of stochastic materials, and, on the other hand, intrinsically deterministic structures are investigated and enable for proposing design guidelines for future materials. After the fundamentals and experimental means used are described, the results of this thesis are split into two parts. The first part investigates the potentials and limits of modifying conventional carbon fiber GDLs by laser perforation. This technique introduces deterministic, straight channels that locally direct liquid water. A wide variety of patterns was studied to evaluate the influence of the pattern geometry on fluid transport through porous materials. With this approach, a suitable descriptor was found that accurately relates the optimum perforation density with fuel cell performance. The competing effects of improved water transport (and thus reduced oxygen transport resistance) and increased membrane drying (by increasing the effective diffusivity) limit the maximum power density increase to 20%. Furthermore, the water transport through the laser-structured GDL of one specific perforation pattern was investigated with sub-second operando X-ray tomographic microscopy (XTM). The comparison with a conventional GDL showed that the total water saturation is similar for both gas diffusion layers. However, the perforations distribute product water more evenly, reducing the water saturation close to the catalyst layer, leading to increased current densities for the perforated GDL. These studies show that the introduced deterministic structure can direct and improve water transport in conventional, stochastic GDLs and thereby increase fuel cell power density. The second part follows a different approach and investigates the potential of novel, fully deterministic GDLs. Woven and gold-coated PET fabrics offer unique tuning properties, enabling a rational, bottom-up design of transport optimized gas diffusion layers. In a first study, these woven materials are successfully used for the first time in a polymer electrolyte fuel cell (PEFC), and their superior performance under cold and saturated operating conditions was analyzed. The water transport, investigated again by operando XTM, is fundamentally different. Water moves much faster through the structure, as it needs to pass only one pore throat before reaching the gas channel. The last study focuses on the thermal conductivity of GDLs, as heat removal is another essential transport duty. Different thickness of gold coatings enable changing the thermal conductivity independently of the pore structure. The strongly reduced thermal conductivity is unexpectedly not a performance limiting factor. It leads to significant thermal gradients across the GDL and thus drives through water evaporation a heat pipe. The temperature gradient is reduced by using product water evaporation as cooling power on the hot side. Thus, more water is transported in the vapor phase, and the GDL structure stays dry for improved oxygen transport. In comparison with a conventional GDL it was shown that the structure is the key to the improved performance and that the thermal conductivity, different by orders of magnitude, causes just smaller differences between the materials. Both approaches emphasize the benefits of designed, deterministic GDL structures that locally reduce liquid water saturation and facilitate oxygen transport. Future GDL development can therefore focus on optimizing the structure and tuning it to the locally heterogeneous operating conditions present in a fuel cell to further increase PEFC power density.