Edge effects on the single cell level of polymer electrolyte fuel cells

Abstract

Polymer electrolyte fuel cells (PEFC) are about to gain an important role in an energy supply based on renewable energy sources. In order to facilitate the market entry of PEFCs, various targets regarding lifetime, functionality and costs have to be fulfilled. An aspect of PEFCs which has not gathered much attention so far is the outer perimeter of the active cell area. The design of this region is substantially determined by the sealing concept and sealant manufacturing process and can have a considerable influence on the overall cell design. This work aims to illustrate the impact of the configuration of this specific area on operating conditions and degradation effects of the whole cell. The results enable the appropriate design of the sealing solution in order to mitigate unfavorable local operating conditions and degradation effects in the outer cell perimeter. In Chapter 3 state-of-the-art cell concepts are presented with respect to the design of the cell edge region and the impact on cell design, local operating conditions and manufacturing processes. General impact of the edge region design on the water transport in a cell is discussed in Chapter 5. Five cells with different sealing concepts were operated and the in-plane water distribution was analyzed by means of neutron radiography. It was shown that void volumes in the outer perimeter of a cell favor accumulation of liquid water there, as long as they are not fed by a direct gas flow. As water transport between edge region and flowfield is slow with time constants of > 1 h the removal of these water clusters is not possible with the applied cell operation protocols. Cells with a gas feed to the outer cell perimeter are in turn subjected to bypass flows around the flowfield. Particularly for flowfields with high flow resistances - e.g. with serpentine-shaped flowfield channels - this can lead to a significantly reduced stoichiometry in the flowfield and hence to decreased liquid water discharge. As a result the measured water content in the flowfield reached a maximum of twice the water content compared to a cell without a flowfield bypass. In general, local operating conditions were strongly influenced by the cell setup in the outer perimeter. Startup of PEFCs under freezing conditions is a requirement for many mobile applications. It can be challenging for cell design and operation as electrochemically produced water as well as residual water from a previous operation can freeze and block gas transport pathways, leading to cell failure. In Chapter 6 it is shown by a recently developed dual spectrum neutron radiography method that freezing of water over a limited fraction of a cell can occur while water in the rest of the cell remains liquid. At moderate temperatures of ≥ −5°C this partial freezing occurred simultaneously with the beginning of a cell voltage decline while the final cell failure could be assigned to the freezing of water over the entire cell area. Furthermore it was shown that residual water in the edge region and flowfield of a cell can have a negative influence on cold start capability. Residual water freezes as soon as the cell is cooled below 0°C and poses a nucleus for a fast phase transition of liquid product water to ice. Chapter 7 focuses on specific degradation mechanisms occurring in the outer cell perimeter. It is shown that if a catalyst coated membrane (CCM) is sandwiched between gaskets in its outer perimeter and between gas diffusion layers (GDL) in the flowfield, a gap between gasket and GDL can lead to accelerated mechanical deterioration of the membrane. Particularly under oscillating humidification conditions shrinking and swelling of the membrane can induce high local stresses in the membrane at the edges of GDL and gasket. As a result cell failure occurred after 10 000 cycles as cracks or pinholes in the membrane led to strong leakages. The experiment showed that a mechanically favorable integration of the CCM with the sealing setup is essential in order to mitigate membrane stress, especially for applications with lifetime requirements of more than 10 000 h. Manufacturing and assembling tolerances can lead to a lateral offset of the gaskets or single layers of a sub-gasket on both sides of a CCM. As a result the gas supply to the anode and cathode catalyst layer (CL) can be asymmetric in the outer perimeter of the active area. In Chapter 7 it is shown experimentally that particularly a local interruption of the anode gas supply can cause massive carbon corrosion of the cathode catalyst support. From 1 mm onwards under the covered area strong thinning of the cathode CL was seen, while the thickness of membrane and anode CL remained unchanged. The results were confirmed by numerical simulation. A specific characteristic of the discussed case was found to be the local electrical isolation between GDL and CL on the anode side by the introduced sub-gasket layer. Thereby a strong negative electrical potential gradient in the anode CL can emerge towards the outer cell perimeter, favoring a low local electrolyte potential since anode overpotentials remain small. As the electrical potential of the cathode CL does not exhibit pronounced potential gradients, the low electrolyte potential leads to high cathode electrode potentials and therefore to significant carbon corrosion rates. It is concluded that cells should intentionally exhibit a lateral offset of gaskets or sub-gasket layers on both sides of the CCM, so that with respect to the assembling tolerances local oxygen starvation occurs on the cathode side rather than hydrogen starvation on the anode side in every case.