Polymer electrolyte membrane water electrolysis (PEL) cells are studied in-operando by synchrotron X-ray radiography. Two-phase flow phenomena associated with the evolution of oxygen and hydrogen in the surrounding water are investigated on a running electrolyzer cell. We examine the gas bubble discharge from the porous transport layer (PTL) into the flow channel and discuss the transport of bubbles in the flow channel. The transport of gas inside the PTL and the number of gas bubble discharge sites is examined and correlated with current density.

The future of energy conversion and storage is expected to rely on the production and storage of hydrogen via water electrolysis 1 . In this scenario, water electrolyzers will play a key role to the establishment of an energy matrix based on renewable but intermittent power sources (e.g. wind turbines and photovoltaics). Other technologies could also be used to store the surplus of energy, such as pumped hydro, geo thermal, batteries, air compression, and the like 2 . However, none of those have the same value proposition as hydrogen. Hydrogen has the advantage of being able to drive multiple revenue streams like transportation, chemicals, green production of fertilizers, regeneration of electricity through fuel cells, and also initially supplement the energy gap through methanation, when coupling with CO 2 sequestration 1 . Moreover, the production, storage, or distribution can be chosen to be centralized or decentralized and it is the only option with a multi-GWh storage capacity 1 . To date, only alkaline and polymer electrolyte membrane (PEM) water electrolyzers are commercially available 1,3 . In order to meet the future demand for water electrolyzers, investment and operational costs still have to be reduced. Moreover, it is fundamental to develop electrolyzers that are able to operate at high current densities, variable partial load, overload, and on/off conditions. These requirements usually place PEM water electrolysis as the best alternative to couple with intermittent power sources. In any case the high costs of PEM water electrolysis components (based on Pt, Ir, and Ti materials) still hamper its large-scale commercial application 3 . Though consistent R&D one can pursue to considerably reduce the costs of PEM water electrolyzers, so that the loading of the expensive based materials can be reduced or even to the point of being completely substituted 3 . On the other side, alkaline water electrolysis stays as a long-established, well matured, and comparatively low cost available technology and another approach could be pursued by improving the performance and operational characteristics of this KOH based system. The hydroxide transport across the diaphragm inside alkaline electrolyzers promoted by the KOH responds very slowly to the power input, limiting the efficiency of the electrochemical reaction, and consequently resulting in low current densities 1,3 . Moreover, the porous structure of the diaphragm allows the diffusion and extensive mixture of the produced hydrogen and oxygen gases when operating at low current densities, limiting the safety range for its operation. Conventional electrodes used in alkaline electrolyzer also tend to possess low active surface area, poor catalyst utilization, and many associated voltage losses 1,3 . When conventional diaphragms or separators are replaced by thin polymer based separators or anion exchange membranes (AEM), the performance of the alkaline electrolyzers can be substantially enhanced. These are the so-called advanced alkaline electrolyzer units and are already expected to reach performance levels close to that of PEM water electrolyzers. The direct comparison between those three presented alternatives is absolutely not trivial and yet cannot be avoided. It is therefore essential to establish a fair performance range comparison when using standard and available materials for classic alkaline, alkaline PEM, and PEM water electrolysis. As an example for the performance behavior in PEM water electrolysis, by using thin PFSA based membranes (< 50 µm), performances reaching up to 10 Acm -2 are obtained. Due to the low ohmic losses when using thin membranes, lower cell voltages are also found, mitigating the voltage inducing corrosion, allowing the use of less expensive material. Nonetheless, the thinner membrane shall increase the hydrogen permeation to the oxygen side limiting its partial load and differential pressure operation 4 . Another important point is the loading of noble metals used in the catalyst layer. To date, Ir loadings range between 2 and 3 mg Ir cm -2 , Pt loadings range between 0.8 and 1.5 mg Pt cm -2 . 1,3 By using advanced methods for the membrane electrode assembly (MEA) fabrication, loadings were dramatically reduced without statistically affecting the performance. In conclusion, a new, robust and efficient benchmark study is presented showing the up-to-date performance behavior of classic alkaline, alkaline PEM, and PEM water electrolysis. This study shall be able to contribute to validate the R&D potential for each technology and its future incorporation into our energy matrix for energy storage and conversion. [1] J. Mergel, M. Carmo and D. Fritz. in Transition to Renewable Energy , D. Stolten, V. Scherer, Editors, p. 423-450, Wiley-VCH (2013) [2] W.F. Pickard et al Energy Reviews ; 13 (8), 1934 (2009). [3] M. Carmo, D. Fritz, J. Mergel and D. Stolten, International Journal of Hydrogen Energy , 38 , 4901 (2013) [4] M. Schalenbach, M. Carmo, D. Fritz, J. Mergel and D. Stolten, International Journal of Hydrogen Energy , 38 , 14921 (2013) Figure 1

When coupled to renewable but nonetheless intermittent power sources such as wind or solar, polymer electrolyte membrane (PEM) electrolyzers are intended to be in operation for tens of thousands of hours 1 . Hence, the long term behavior of membrane electrode assemblies (MEAs) under these power operating characteristics is of great importance. At moderate conditions (25 cell stack, around 50°C and 1.3 A/cm²) 2 , commercial PEM electrolyzers show little to no significant degradation (< 4 µVh) after 55,000 h 2 . However the anode catalyst loading in commercial systems is generally high (> 2 mg/cm²) 3 and the state-of-the-art catalyst material, iridium, is expensive and rare. A reduced catalyst loading will contribute to reduce the costs of electrolyzer stacks, but to this date the effect of anode catalyst loading on the durability of PEM electrolysis has not yet been thoroughly investigated in literature. This study aims to show the durability of PEM electrolysis single cells under reduced anodic catalyst loadings. As shown in Fig. 1 (top), with IrO 2 loadings as low as 0.6 mg/cm² we were able to reach identical performance when compared to our standard catalyst loading of 2.25 mg/cm² IrO 2 . Moreover, stable cell performances over 680 h in an initial life time test were obtained (Fig. 1 bottom). Further reductions in catalyst loading show deteriorating cell performances in 1000 h long term testing. In this study the performance degradation data of MEAs with lower anode loadings were shown and the degradation effects differentiated. The degradation characteristics were also partially localized using electrochemical impedance spectroscopy and reference electrode measurements 4 in order to separate cathodic and anodic overvoltages. We aim to contribute to the understanding of the degradation mechanisms and to possible cost reductions of MEAs for PEM electrolysis by showing the relation between anode loading and long term stability. This information shall assist in making PEM water electrolysis both cost effective and durable. Fig. 1 : Top: single cell performance under a heavily reduced anode loading of 0,60 mg/cm² IrO 2 . We were able to obtain similar cell performances as for the regular loadings of 2,25 mg/cm² IrO 2 and stable cell performances over time (bottom). References: 1. M. Carmo, D. Fritz, J. Mergel and D. Stolten, International Journal of Hydrogen Energy, 38, 4901-4934 (2013) 2. K. Ayers, PEM Electrolysis R&D Webinar, presented at the DOE Fuel Cell Technologies Program Webinar, May 23 (2011) 3. Debe, M. K.; Hendricks, S. M.; Vernstrom, G. D.; Meyers, M.; Brostrom, M.; Stephens, M. et al.:. In: J. Electrochem. Soc. 159 (6) (2012) 4. Jung, Myunghee; van Zee, John W., 216th ECS Meeting, 1585–1593 (2009) Figure 1

The energy excess of intermittent power sources can be stored as hydrogen. Polymer electrolyte membrane (PEM) electrolysis is an efficient way to produce electrolytic hydrogen.[1] Due to the acidic environment only scarce and expensive noble metals like platinum and iridium are used as cathode and anode catalysts respectively. To cut down the investment cost of a PEM electrolyzer, one of the main tasks is the reduction of the noble metal amount while retaining or increasing the performance of the electrolyser.[2] Thereby the influence of the decrease of the noble metal loading and possible degradation of the active material on the durability of the HER catalyst are both essential. In this study we investigated the degradation of catalyst coated membranes (CCM) under constant and intermittent PEM electrolysis conditions. For this purpose we conducted long term test (1000h) with 4 different commercial CCMs each having different cathode loadings. To date, the durability of CCMs is examined under moderate current densities (0.9 to 1.4 A cm-2).[3] However, in this study the long-term test were performed at a current density of 2 A cm-2 to simulate real electrolysis conditions. As a next step we manufactured CCMs with a low platinum loading and different Nafion amounts at the cathode to examine the influence of the ionomer content on the long-term durability. TEM, XRD and EIS were applied to monitor physical and electrochemical changes during the durability test. The cathode and anode potentials were measured with the help of a reference electrode to determine the cause of degradation.[4] CCMs with a platinum loading of 0.8 mgPt cm-2 showed a small degradation rate at a constant and dynamic operation mode. We observed that the reduction of the cathode loading to 0.05 mgPt cm-2 lead to a 5 fold increase of the degradation rate at the constant and the dynamic operation mode. The increase of the cathode potential during the durability test indicates a degradation of the cathode catalyst, which was verified by TEM and XRD analyses: Agglomeration of the platinum particles on the carbon support was observed. Our study clearly identifies the mechanisms responsible for the degradation of the cathode. This information opens up new possibilities to implement counter measures against the deterioration of the Pt/C catalyst, and to produce cost effective CCMs with low loadings and high durability. 1) M. Carmo, D. L. Fritz, J. Mergel, D. Stolten, Int. J. Hyd. Energ., 2013, 38(12),4901 - 34. 2) K. Ayers, Annual Merit Review DOE Hydrogen and Fuel Cells and Vehicle Technologies Programs, Washington, DC, 2014. Http://www.hydrogen.energy.gov/pdfs/review14/pd098_ayers_2014_o.pdf 3) M. Debe, J. Electrochem. Soc., 2012, 159(6), K165-K176. 4) S.A. Grigoriev, P. Millet, V.N. Fateev, J. Power Sources, 2008, 177(2), 281 - 285. Figure 1

Water electrolysis is an established method to produce hydrogen from the surplus power of renewable energy sources 1 . There are two technologies considered to be the most suitable when connected to intermittent power sources: polymer electrolyte membrane (PEM) water electrolysis and alkaline electrolysis with a liquid electrolyte 2 . Both technologies are linked to certain advantages and limitations. High current densities for example are achieved with PEM electrolysis but the acidic regime requires scarce and expensive platinum-group-metal (PGM) catalysts and Ti-based stack components 3 . On the other hand alkaline electrolysis is regarded as a mature and less expensive technology but is limited to low current densities 4 . In order to combine the benefits of both technologies while simultaneously overcoming their drawbacks, a thin anion exchange membrane (AEM) is used as the electrolyte, allowing the use of low-cost transition-metal catalysts and stack material 5 . In this study we developed membrane electrode assemblies (MEAs) for the AEM water electrolysis that closely match the performance of state-of-the-art PEM water electrolysis. Furthermore we have assessed commercial AEMs to determine the ionic conductivity and ion-exchange capacity. To precisely measure the hydrogen permeability of the membranes at different temperatures, an electrochemical monitoring technique was developed. Best results were achieved with Tokuyama´s A201 AEM as the ionic conductivity reached 0.4 S/cm and a hydrogen permeability of only 10 -16 mol/(cm · s · Pa) at 70 °C. A screening method was also developed to identify and characterize the most active set of electrocatalysts. Nickel nanoparticles were found to have an overpotential of about 0.4 V towards HER and OER at 10 mA/cm 2 and was thus chosen for the development of MEAs. Single cells exhibited a current density of 220 mA/cm 2 at 2.2 V and 50 °C in deionized-water. We suppose that the low current density originates from the poor ionic conductivity of the AS4 ionomer used in the electrode layer. Nevertheless the use of 1 M KOH solution as an additional electrolyte led to an improved ionic conductivity. This optimization resulted in single cells reaching current densities of 1.5 A/cm 2 at 2.2 V and 50 °C. Long-term experiments were additionally conducted to gain information on the impact of both operational modes on the ionomer stability. Our results clearly show the potential of AEM water electrolysis to meet the performance of state-of-the-art PEM electrolyzers without using scarce and expensive electrocatalysts and stack components. Figure 1: Polarization curves of a MEA containing Nickel nanoparticles on anode and cathode electrode. Tokuyama AS4 was used as ionomer and A201 as AEM. A 25 cm 2 single cell was measured at 50 °C in deionized water and 1M KOH solution. Current density ranges after [1]. References: 1. D. Stolten, Ed., Transition to Renewable Energy Systems , John Wiley & Sons, Weinheim (2013) 2. D. Stolten, Hydrogen and Fuel Cells: Fundamentals, Technologies and Applications , Wiley-VCH, Weinheim (2010) 3. M. Carmo, D. L. Fritz, J. Mergel, and D. Stolten, Int. J. Hydrogen Energy , 38 , 4901–4934 (2013) 4. K. Zeng and D. Zhang, Prog. Energy Combust. Sci. , 36 , 307–326 (2010) 5. J. R. Varcoe et al., Energy Environ. Sci. (2014) DOI: 10.1039/C4EE01303D Figure 1

The electrodes in fuel cells simultaneously realize electric and ionic conductivity. In the case of acidic polymer electrolytes, the electrodes are typically made of composites of carbon-supported catalyst and Nafion polymer electrolyte binder. In this study, the interaction of the proton conduction, the electron conduction, and the electrochemical hydrogen conversion in such composite electrode materials was examined. Exposed to a hydrogen atmosphere, these composites displayed up to 10-fold smaller resistivities for the proton conduction than that of Nafion membranes. This effect was ascribed to the simultaneously occurring electrochemical hydrogen oxidation and evolution inside the composite samples, which are driven by different proton and electron resistivities. The parasitic electrochemical currents resulting were postulated to occur in the anode of fuel cells with polymer, solid oxide, or liquid alkaline electrolytes, when the ohmic drop of the ion conduction in the anode is higher with the anodic kinetic overvoltage (as illustrated in the graphical abstract). In this case, the parasitic electrochemical currents increase the anodic kinetic overpotential and the ohmic drop in the anode. Thinner fuel cell anodes with smaller ohmic drops for the ion conduction may reduce the parasitic electrochemical currents.

One of the main goals for improvement of high-temperature polymer electrolyte fuel cells (HT-PEFCs) is the increase of the fuel cell performance under different operating conditions. We investigated the correlation between operating conditions and structural changes in the electrodes by means of in-situ through-plane synchrotron X-ray radiography. From the radiographs it is possible to clearly distinguish between the electrode crack structure beneath the ribs and beneath the channels of the flow field. We present a statistical method to analyze these crack structures. For this purpose a 'radar' method was developed in order to obtain the width of the cracks at many different locations and the distribution of crack widths. We found a different behavior of cracks located beneath the ribs and beneath the channels and an influence of the operating conditions on local regions of the crack structure.