Hydrogen is expected to play an important role as a crosslinking technology between power generation on one hand and transport and industry on the other hand. It can directly replace fossil fuels in transport and industry when produced by water electrolysis renewable energies such as solar or wind, which are converted with low efficiencies. The relevant technologies are either the mature alkaline electrolysis or the newer proton exchange membrane (PEM) water electrolysis. The main advantage of PEM electrolysis technology is that it provides high efficiency, high power density with low footprint, and flexibility without the need of corrosive chemicals such as KOH (alkaline electrolysis). On the other hand, questions about degradation mechanisms, long-term stability and the high capital costs compared to the alkaline technology hinder the large-scale industrial penetration of this technology in the market. Moreover, these challenges are mutually dependent. Cost reductions of materials and stack components, for example, can have a negative impact on long-term stability. Thus, more information about the degradation mechanisms is necessary to determine an optimum between cost reduction and long-term stability. Furthermore, degradation measurements are expensive and evidently require a lot of time. Consequently, the development of accelerated stress tests (AST) protocols is in the focus of the current research by the academy and industry in particular.

This presentation will discuss degradation processes of MEAs under high current density which has been found to accelerate degradation in our laboratory. Degradation was investigated in a HYLYZER™ Hydrogen Generator (HHG) from Hydrogenics under operation at 4 A cm^-2 for almost 3000 h. The degradation of the bipolar plates and gas diffusion layers are neglected since these components are well protected against corrosion. The cells were analyzed by electrochemical impedance spectroscopy. Furthermore samples of ion exchange resin, which keeps the water resistivity above 10 MΩ, were collected and analyzed by XPS to quantify the accumulated ions. In addition, post-test investigations of the aged MEAs by conductive and material sensitive AFM, SEM with EDX and XPS were performed to identify degradation mechanisms. An important observation is a decrease of the ohmic resistance during operation at high current densities which is associated to fluoride release into the water.

To analyze the time dependence of the cells in the hydrogen generator system a zero-dimensional system model with effective parameters was developed. This model enables quantifying the degradation measurements and the discussion in the context of system behavior. Preventing metal oxidation by corrosion protective layers leads to increased efficiency mainly due to decreasing ohmic resistances. This decrease is assumed to be depending on current density and time. The observed F release rate supports this conclusion. The degradation mechanism is assumed to be an increased permeability for water, oxygen and hydrogen with corresponds to a membrane thinning and ionomer loss in the electrodes. Such an interpretation was backed by the ex-situ post-test analysis. At the anodes a general increase of conductive area fraction evaluated by AFM indicates a loss of ionomer. At the cathodes, a general decrease of conductive area fraction indicates a loss of catalysts. EDX confirmed the loss of Pt and Ir from cathode and anode respectively, and in addition a loss of Ti from the anode was found. Membrane thinning occurred in all cells. It is explained by ionomer decomposition catalyzed by the well distributed Pt content in the membranes detected by XPS.

Acknowledgements

The research leading to these results has received funding from the European Union’s Seventh Framework Program (FP7/ 2007-2013) for the Fuel Cells and Hydrogen Joint Technology Initiative under grant n° 621237, “ INSIDE- In -situ Diagnostics in Water Electrolyzers“ and from national projects “LastElSys” and “WESpe” by Federal Ministry for Economic Affairs and Energy.