Design of an In-Operando Cell for X-Ray and Neutron Imaging of Oxygen-Depolarized Cathodes in Chlor-Alkali Electrolysis

materials

Article Design of an In-Operando Cell for X-Ray and Neutron Imaging of Oxygen-Depolarized Cathodes in Chlor-Alkali Electrolysis

Marcus Gebhard 1,* , Melanie Paulisch 2, André Hilger 2, David Franzen 3, Barbara Ellendorﬀ 3, Thomas Turek 3, Ingo Manke 2 and Christina Roth 1 1 Institute of Chemistry and Biochemistry, Freie Universität Berlin, Takustraße 3, 14195 Berlin, Germany; [email protected] 2 Institute of Applied Materials, Helmholtz-Zentrum Berlin für Materialien und Energie GmbH, Hahn-Meitner-Platz 1, 14109 Berlin, Germany; [email protected] (M.P.); [email protected] (A.H.); [email protected] (I.M.) 3 Institute of Chemical and Electrochemical Process Engineering, Clausthal University of Technology, Leibnizstraße 17, 38678 Clausthal-Zellerfeld, Germany; [email protected] (D.F.); barbara.ellendorﬀ@tu-clausthal.de (B.E.); [email protected] (T.T.) * Correspondence: [email protected]

Received: 28 March 2019; Accepted: 16 April 2019; Published: 18 April 2019

Abstract: Oxygen-depolarized cathodes are a novel concept to be used in chlor-alkali electrolysis in order to generate significant energy savings. In these porous gas diffusion electrodes, hydrophilic and catalytically active microsized silver grains and a hydrophobic polytetrafluoroethylene cobweb structure are combined to obtain the optimum amount of three-phase boundaries between the highly alkaline electrolyte and the oxygen gas phase to achieve high current densities. However, the direct correlation between specific electrode structure and electrochemical performance is difficult. In this work, we report on the successful design and adaptation of an in-operando cell for X-ray (micro-computed tomography, synchrotron) and neutron imaging of an operating oxygen-depolarized cathode under realistic operation conditions, enabling the investigation of the electrolyte invasion into, and distribution inside, the porous electrode for the first time.

Keywords: in-operando and in-situ; X-ray and neutron imaging; cell design; chlor-alkali electrolysis; gas diﬀusion electrode; oxygen-depolarized cathode; oxygen reduction reaction

1. Introduction Chlorine is a commodity chemical, needed in large quantities for the manufacture of more than 60% of all chemical products [1]. Until recently, it has been mainly synthesized by the amalgam and the diaphragm process [2,3]. These two versions of the chlor-alkali electrolysis (CAE) have been subject to development and research to minimize the harmful and controversial use of mercury and asbestos [4]. The membrane process developed in the 1970s [5] is environmentally friendly in regards to the used materials and more energy efficient than the older amalgam and diaphragm processes, but still huge amounts of electrical energy are required, resulting in high CO2 emissions [6]. In context to the efforts to reduce global greenhouse gas emissions, a novel approach has been proposed, by which the CO2 emissions can be reduced significantly [7]. Herein, an oxygen-depolarized cathode (ODC) replaces the hydrogen-evolving cathode in the CAE system. Electrical energy savings of up to 30% can be obtained due to the lower Gibbs free energy of the oxygen reduction reaction (ORR) compared to the hydrogen evolution reaction (HER). In this manner, the cell potential can be theoretically lowered by 1.23 V under standard conditions, resulting in savings up to 1 V under conditions relevant to industrial

Materials 2019, 12, 1275; doi:10.3390/ma12081275 www.mdpi.com/journal/materials Materials 2019, 12, 1275 2 of 14

application, where the ODC is operated at 80 ◦C in 30 wt.% NaOH electrolyte. Considering the total energy demand for the chlor-alkali process, this technology could cut up to 1% of the total electrical energy demand in Germany [7,8]. The life cycle assessment of ODCs shows that the total energy and cost savings achieved will be an essential step forward in order to meet future emission goals [9–11]. The ODC is a special version of widely used gas diffusion electrodes (GDE) [12–14] with the oxygen reduction reaction taking place as the main reaction. Platinum is a commonly utilized, very efficient, and stable ORR catalyst, but for industrial CAE application a porous silver electrode has been proposed in the past because of its comparable activity and stability [15] in the highly alkaline electrolyte (30 wt.% NaOH). Even though silver is much less costly than platinum, a significant quantity of the catalytically active material is still required [16]. Consequently, an optimal catalyst utilization is of utmost importance. Since the catalyst utilization is directly linked to the number of particles, which are electrochemically addressed, and this is achieved by the distribution of three-phase boundaries, the porous electrode structure plays an important role. Consequently, it is essential to correlate the electrode’s 3D architecture with the electrochemical performance. In the ODC, the highly alkaline electrolyte and the gas phase get into contact at the three-phase boundary regions; ORR-active, electron-conducting, and hydrophilic µm-sized silver particles will form agglomerates that are partly filled with the electrolyte. The gas phase is introduced in the porous electrode by a hydrophobic polytetrafluoroethylene (PTFE) phase encompassing the silver particles in a cobweb-like fashion [16]. Routinely, such electrodes are prepared by airbrushing techniques, followed by a hot-pressing step to compact the structure and then a heat treatment procedure to obtain porous structures and to sinter the material. Since many parameters can be varied in the process, it is difficult to completely control the final electrode structure. Moreover, to our knowledge a direct and in-depth correlation between structure characteristics and electrochemical performance has not been made yet. Different imaging techniques have been used to characterize the 3D structure of porous electrodes (e.g. focused ion beam (FIB)/scanning electron microscopy (SEM) [17,18], X-ray as well as neutron tomography and radiography), each of them addressing a different length-scale with different resolution and specific advantages [19]. In particular, significant advances have been made in the area of fuel cells. Thus, it was possible to show, explain, and quantify the water transport within polymer electrolyte membrane fuel cells in-situ and in-operando with synchrotron X-ray [20–24] and neutron radiography [25–32]. However, the invasion and distribution of electrolyte within metallic GDEs and ODC in a realistic testing set-up have not yet been imaged during operation. Furthermore, due to the different phases, silver, and PTFE in the GDEs, it is necessary to ensure in-operando measurements with X-rays and neutrons. While X-ray measurements provide excellent spatial and time resolution [33], neutron radiography leads to a high contrast of the electrolyte to the metallic GDE, when NaOD is used as electrolyte and thus provides additional insight. To understand how ODCs work, it is of high importance to perform in-operando tests to image, analyze, and quantify the electrolyte flow within the GDE. In this paper, we report on the successful design and adaptation of an in-operando cell for X-ray and neutron imaging of an operating ODC under realistic operation conditions. We will describe the considerations based on which materials were chosen and how the cells were modified. The commercial Gaskatel cell served as a starting point and was optimized for the imaging process, leading to the first generation (1st Gen.) cell. The further development to the second generation (2nd Gen.) cell improved the electrochemistry features of the cell, so that the 2nd Gen. shows an optimal compromise between the electrochemical and the imaging requirements.

2. Materials and Methods

2.1. Electrode Preparation The ODCs investigated in this study were fabricated by an airbrushing process similar to [16]. In brief, an aqueous suspension of the microsized silver catalyst (SF9ED, Ferro, Frankfurt a. M., Germany) Materials 2019, 12, 1275 3 of 14 and PTFE (PTFE Dispersion TF 5060GZ, 3M Dyneon, Burgkirchen, Germany) as a hydrophobic agent were mixed with a methylcellulose solution (WALOCEL® MKX 70000 PP 01, Dow Wolff Cellulosics GmbH, Bomlitz, Germany) as a pore building agent and thickener. The suspension was mixed with an Ultra-Turrax (IKA®-Werke GmbH & CO. KG, Staufen im Breisgau, Germany) and homogenized with sonication, then consecutively air-brushed onto a conductive supporting material (Ni-mesh 106 118 × µm mesh size, 63 µm thickness, Haver & Boecker OHG, Oelde, Germany). The resulting electrodes were hot-pressed with a pressing load of 15 MPa at 130 ◦C for 5 min and sintered at 330 ◦C for 15 min to burn off the methylcellulose and improve the mechanical stability through PTFE sintering. The resulting ODC was subsequently cut into a rectangular shape (20 20 mm, 350 µm thickness) to fit × into the half-cell set-up.

2.2. Electrochemical Characterization Electrochemical characterization was performed in a three-electrode set-up using either the commercial GDE testing cell (FlexCell® HZ-PP03, Gaskatel GmbH, Kassel, Germany) or the newly developed in-operando cells (1st Gen. cell and 2nd Gen. cell). A potentiostat (Gamry Ref 3000, Gamry Instruments, Warminster, PA, USA) was used for electrochemical measurements. As the working electrode (WE), silver ODCs as mentioned above, with a geometric surface area of 3.14 cm2 were utilized. A platinum wire electrode (99.95% purity, 1 mm diameter, 25 cm length) was used as a counter electrode (CE) and a HydroFlex® (Gaskatel GmbH, Kassel, Germany) reversible hydrogen electrode (RHE) represented the reference electrode (RE). The industrially used electrolyte (30 wt.% NaOH solution) was prepared by dissolution of caustic flakes (> 99% NaOH, Carl Roth, Germany) in deionized water (Merck Millipore, Burlington, MA, USA) and provided to the electrolyte side of the ODC. On the backside of the cathode, oxygen ( 99.5%; Air Liquide, Paris, France) was provided via ≥ a pressure controller (Everwand Druckgastechnik GmbH, Solingen, Germany) by constant flow of 1 20–50 mL min− (Influx Measurements Ltd., Alresford, UK) and minimal backpressure using a 5 mm water column. For electrode characterization the following electrochemical methods were applied, as single procedure or periodically scripted procedures for long-term measurements. The operating temperature of the cell was set to 20, 50, or up to 80 ◦C by regulation of cell and electrolyte tank temperature. Linear sweep voltammetry (LSV) measurements were performed either without correction for inner resistance (iR) or in current interrupt (CI) mode. Starting at open circuit (1.0–1.1 V vs. RHE), the 1 potential was changed with a constant sweep rate of 1 or 2 mV s− to a lower limit of approximately 0.2 V vs. RHE. Depending on the iR correction, or current limitation of the potentiostat, the measurements were terminated before reaching the lower potential limit. Long-term operation chronoamperometric measurements (CA) were performed at different potentials (0.2, 0.5, and 0.9 V vs. RHE, not iR corrected) up to 1 h per potential. For post iR compensation of the electrochemical measurements, potentiostatic electrochemical impedance spectroscopy (EIS) was performed in a range from 1 MHz to 100 mHz with an AC voltage of 5 mV at 0.2–0.9 V vs. RHE respectively.

2.3. Micro-CT Measurements The laboratory X-ray micro-computed tomography (micro-CT) radiography set-up consisted of a Hamamatsu X-ray tube (L8121-03, tungsten anode, Hamamatsu Photonics K.K., Hamamatsu, Japan) and a Hamamatsu flat panel detector (C7942SK-05, Hamamatsu Photonics K.K., Hamamatsu, Japan). The in-operando cell was placed between the source and detector as seen in Figure1a. The half-cell was furthermore shielded by a modified plastic housing, providing sufficient protection of the measurement environment for possible electrolyte contamination. The separate electrolyte tank (PTFE container, 250 mL) was placed in a heating bath outside the micro-CT housing, while the peristaltic pump, oxygen flow controller, temperature controller unit and the potentiostat were connected and placed in the micro-CT housing. The magnification achieved in this way was the largest that projected the entire field of view onto the detector, resulting in a pixel size of 12.25 µm. To achieve maximum Materials 2019, 12, x FOR PEER REVIEW 4 of 14

2.4. Synchrotron Radiography Preliminary synchrotron experiments without operating the ODC were conducted at the BESSYII BAMLine to demonstrate suitable resolution and the experimental feasibility. To simulate the electrolyte behavior in these preliminary experiments, the 1st Gen. cell was flooded with deionized water. For the experiments, a monochromatic photon energy of 25 keV was used. The optical set-up, consisting of a (4008 × 2672 pixel) CCD camera (PCO 4000, PCO AG, Kelheim, Germany) and a cadmium tungstate (CdWO4) scintillator screen were used to capture images with a pixel size of 985 nm. Resulting images were calculated using the imaging software Fiji (v1.5.2) [34].

2.5. Neutron Radiography Neutron experiments were carried out at the CONRAD/V7 Beamline facility of the Helmholtz- Zentrum Berlin, Institute of Applied Materials (BER II research reactor). The in-operando cell was operated at 20 °C. The half-cell was connected to the oxygen and electrolyte supply (30 wt.% NaOD in D2O, 99 at.% D, Sigma Aldrich, St. Louis, MI, USA), as well as the potentiostat. A specially prepared Materialsshielding,2019 consisting, 12, 1275 of a modified thin plastic housing was provided and connected to the sample4 of 14 holder to secure the measurement environment for possible electrolyte contaminations (see Figure 1b). Images of the ODC were taken with white beam at 2 to 6 Å (peak at approximately 3.5 Å). The contrastneutrons and were the detected best signal-to-noise at a gadolinium ratio, oxide the X-ray (Gd2O tube3) coated was operated scintillator at 130in combination kV and at 230 withµA a withNeo a 0.55.5 mmsCMOS copper camera ﬁlter. (Andor, The normalization Oxford Instruments, of the Abingdon, captured images UK) with was 5.5 performed megapixel usingproviding the imaginga pixel softwaresize of 6.5 Fiji µm. (v1.5.2) [34].

FigureFigure 1.1. Photographs of the experimental in-operando in-operando set-up set-up (a (a) )micro-computed micro-computed tomography tomography at at Helmholtz-ZentrumHelmholtz-Zentrum Berlin; (b) neutron radiography at at CONRAD2 CONRAD2 Berlin. Berlin. 2.4. Synchrotron Radiography 3. Results and Discussion Preliminary synchrotron experiments without operating the ODC were conducted at the BESSYII BAMLine3.1. General to Considerations demonstrate suitable resolution and the experimental feasibility. To simulate the electrolyte st behaviorFor incharacterization these preliminary and experiments,imaging of all the processes 1 Gen. taking cell was place ﬂooded at an with ODC, deionized a rather water.complex For set the experiments,up is required. a monochromaticIn CAE where ODCs photon are energy utilized, of 25 the keV appropriate was used. supply The optical of gaseous set-up, oxygen consisting to one of a (4008 2672 pixel) CCD camera (PCO 4000, PCO AG, Kelheim, Germany) and a cadmium tungstate side of× the ODC and liquid electrolyte to the other side are critical operating factors. Furthermore, (CdWOthe cell4 set-up) scintillator needs screento be stable were against used to corrosion capture images caused with by the a pixel highly size al ofkaline 985 nm.electrolyte Resulting (30 imageswt.% wereNaOH) calculated at temperatures using the up imaging to 80 °C software[11,16]. In Fiji addition, (v1.5.2) a [sophisticated34]. periphery is necessary to keep the process conditions constant for electrochemical characterization and during imaging of the 2.5. Neutron Radiography temperature sensitive process. Consequently, the reliable characterization of the electrochemical performanceNeutron of experiments ODCs for ORR were can only carried be achieved out at if the the CONRADhalf-cell set-up/V7 is Beamline constructed facility in a way of to the Helmholtz-Zentrumguarantee constant measurement Berlin, Institute cond ofitions. Applied Additionally, Materials(BER a three-electr II researchode reactor). set-up has The to in-operando be used to cellensure was the operated accurate at measurement 20 ◦C. The half-cell of the wasODC connected characteristics. to the Therefore, oxygen and the electrolyte set-up has supply to include (30 wt.%the NaODworking in electrode D2O, 99 at.% (ODC), D, Sigmaa reference Aldrich, electrode St. Louis, compat MI,ible USA), to the as operating well as the conditions, potentiostat. and Aa suitable specially preparedand stable shielding, counter electrode. consisting The of apositioning modiﬁed thinof the plastic electrodes housing should was be provided done carefully and connected to minimize to the sample holder to secure the measurement environment for possible electrolyte contaminations (see Figure1b). Images of the ODC were taken with white beam at 2 to 6 Å (peak at approximately 3.5 Å). The neutrons were detected at a gadolinium oxide (Gd2O3) coated scintillator in combination with a Neo 5.5 sCMOS camera (Andor, Oxford Instruments, Abingdon, UK) with 5.5 megapixel providing a pixel size of 6.5 µm.

3. Results and Discussion

3.1. General Considerations For characterization and imaging of all processes taking place at an ODC, a rather complex set up is required. In CAE where ODCs are utilized, the appropriate supply of gaseous oxygen to one side of the ODC and liquid electrolyte to the other side are critical operating factors. Furthermore, the cell set-up needs to be stable against corrosion caused by the highly alkaline electrolyte (30 wt.% NaOH) at temperatures up to 80 ◦C[11,16]. In addition, a sophisticated periphery is necessary to keep the process conditions constant for electrochemical characterization and during imaging of the temperature sensitive process. Consequently, the reliable characterization of the electrochemical performance of Materials 2019, 12, 1275 5 of 14

ODCs for ORR can only be achieved if the half-cell set-up is constructed in a way to guarantee constant measurement conditions. Additionally, a three-electrode set-up has to be used to ensure the accurate measurement of the ODC characteristics. Therefore, the set-up has to include the working electrode (ODC), a reference electrode compatible to the operating conditions, and a suitable and stable counter electrode. The positioning of the electrodes should be done carefully to minimize their inﬂuence on the electrochemical measurement (i.e., not causing a large ohmic drop and disturbance of the electrical ﬁeld). The cell geometry and position of all components play an important role when imaging the physical processes taking place at the ODC. In general, the set-up has to be constructed in a way that the incoming beam of either X-rays or neutrons is absorbed in a minimal way. The absorption of the incident beam can be caused by the cell material, cell components, the electrolyte, or the ODC itself. For radiography investigation with X-ray radiation, in-plane measurements were excluded, due to the strong absorption of the electrode material (in-plane 2 cm of silver and nickel). If radiography imaging is used to investigate the electrolyte behavior in through-plane geometry, the ODC structure exerts less impact (electrode thickness 350 µm). Therefore, the experiment has been designed in a way that the beam direction is perpendicular to the electrode with a minimized impairment by components in beam direction. If all of these requirements regarding the operating conditions, electrochemical measurements and the imaging are met, an investigation of the electrolyte behavior during the operation of an ODC Materials 2019, 12, x FOR PEER REVIEW 6 of 14 at technically relevant conditions is possible.

3.2.the GaskatelODC, the Cell gas supply compartment is situated and designed to tighten the top of the electrode mesh with the current collector and to keep the ODC in position with a screwed holder plate. Gaskets betweenA well-established the block and half-cellholder plate set-up seal for the the ODC electrochemical surface and characterization restrict it to 3.14 of cm². GDEs The is provided working ® byelectrode Gaskatel. area The is the commercially same as in the available Gaskatel half-cell cell and (FlexCell was not changed) allows toto measureobtain a wider under field industrially of view relevanton the electrode conditions surface. (30 wt.% After NaOH, interaction 80 ◦ C)of withthe be aam three-electrode with the ODC, set-up the beam in a chemicallyexits at the resistantbackside cellof the environment. ODC to a cylindrical The cell consists oxygen ofreservoir two compartments, with connections one forto the the gas supply supply. of oxygen A sealed to theradiation ODC andwindow, an electrolyte milled in the compartment direction of on the the beam opposite to a th sideickness of theof 1 ODCmm and for therefore, the electrolyte absorbing supply as little (see Figureof the 2beama). However, as possible, the referencecloses the electrode, reservoir. counter From Figure electrode, 2b it electrolyte can be seen reservoir, that the and incident other partsbeam ofpasses the Gaskatel only 2 mm cell of are block placed material, in positions, the 5 mm where thick theelectrolyte beam would chamber have in tofront pass of for the through-plane ODC, and the radiographyODC itself. In imaging this way, (see the Figure absorption2a). Therefore, of the inci thedent cell set-upbeam by has the to behalf-cell modiﬁed was signiﬁcantly kept to a minimum, to allow imagingresulting of in the promising electrode results without obtained constraints from duepreliminary to the cell X-ray components. radiography measurements.

FigureFigure 2. 2.SchematicSchematic diagram diagram of of ( (aa)) Gaskatel Gaskatel cell andand ((bb)) ﬁrst first generation generation (1 (1st stGen.) Gen.) cell cell with with relevant relevant parts. parts.

st 3.3. DevelopmentHowever, not of 1includedGen. Cell in the schematic figures, but important for the operation of the cell, are seals,The heating position cartridges, of key components and tubing inthe connections, commercial as Gaskatel well as cell the does temperature not allow for sensors. the radiography Heating imagingcartridges of were operating placed ODCs. in the The electrolyte focus for block developing near the the electrolyte 1st Gen. cellreservoir (see Figure and inlet2b) was to allow on adapting heating theup existingto 80 °C, design regulated in a wayby a thatseparate it attenuates temperature the radiation controller. as little In this as possible, way, the while developed keeping set-up the e ﬀwasect onoptimized the electrochemical for radiography measurement imaging asand low met as possible.all requirements Therefore, for the operating novel design the cell was at divided technically into relevant electrochemical conditions.

3.4. Preliminary Synchrotron Measurements

The 1st Gen. cell was investigated at the synchrotron BESSYII to test the achievable resolution and the experimental feasibility. Figure 3a presents an image of the GDE with its microstructure clearly visible. The nickel mesh (which is utilized as conducting and mechanically supporting structure) appears clearly, also some artifacts due to the manufacturing of the radiation windows. In particular, some parts of the radiation windows are not fully planar, which is caused by the milling process. However, this in-situ measurement demonstrated the suitability of our cell concept. Additionally, in Figure 3b–d the cell is flooded with deionized water to demonstrate that the electrolyte shows enough contrast, so that its interaction with the ODC can be imaged. Already in this experiment, different structural features of the ODC can be observed. For example, relatively large pores appear in between the mesh structure of the Ni-substrate.

Materials 2019, 12, 1275 6 of 14 two compartments, like the Gaskatel cell, to keep its functionality; but with the electrolyte, CE and RE were positioned differently to allow for successful imaging. Polyether ether ketone (PEEK) was chosen as the material for the half-cell, because of its high resistivity against alkaline solutions, good thermal stability, and good machinability. PEEK also provides good durability under X-ray radiation [35] and was chosen as material for components remaining in the path of the beam. The strongly absorbing electrolyte reservoir with the counter electrode was placed above the beam channel to minimize the electrolyte volume in the beam path. However, a thin layer of electrolyte in front of the electrode could not be completely eliminated, as it is required for the ODC to operate properly. Consequently, a conically-shaped electrolyte chamber with 5 mm thickness was placed in front of the ODC. Due to the small volume, the electrolyte had to be pumped to maintain stable measurement conditions. The electrolyte inlet was positioned below the beam with the electrolyte flowing onto the electrolyte holder plate, not directly onto the ODC surface. This has another beneficial side effect that gas bubbles passing through the ODC will collect at the top of the electrolyte chamber and will be pumped in the flow direction with evolving gas from the CE into the outlet. In addition, the reference electrode was positioned beside the electrolyte reservoir and connected to the electrolyte chamber in front of the ODC by a capillary. In this way, important components could be eliminated from the beam path, while the three-electrode set-up remained operational. As a result, the incident beam can pass the electrolyte block with minimal attenuation. The radiation window of the electrolyte block is milled to a minimum thickness of 1 mm, which the beam has to pass, before entering the 5 mm thick electrolyte chamber and the ODC. Behind the ODC, the gas supply compartment is situated and designed to tighten the top of the electrode mesh with the current collector and to keep the ODC in position with a screwed holder plate. Gaskets between the block and holder plate seal the ODC surface and restrict it to 3.14 cm2. The working electrode area is the same as in the Gaskatel cell and was not changed to obtain a wider field of view on the electrode surface. After interaction of the beam with the ODC, the beam exits at the backside of the ODC to a cylindrical oxygen reservoir with connections to the gas supply. A sealed radiation window, milled in the direction of the beam to a thickness of 1 mm and therefore, absorbing as little of the beam as possible, closes the reservoir. From Figure2b it can be seen that the incident beam passes only 2 mm of block material, the 5 mm thick electrolyte chamber in front of the ODC, and the ODC itself. In this way, the absorption of the incident beam by the half-cell was kept to a minimum, resulting in promising results obtained from preliminary X-ray radiography measurements. However, not included in the schematic figures, but important for the operation of the cell, are seals, heating cartridges, and tubing connections, as well as the temperature sensors. Heating cartridges were placed in the electrolyte block near the electrolyte reservoir and inlet to allow heating up to 80 ◦C, regulated by a separate temperature controller. In this way, the developed set-up was optimized for radiography imaging and met all requirements for operating the cell at technically relevant electrochemical conditions.

3.4. Preliminary Synchrotron Measurements The 1st Gen. cell was investigated at the synchrotron BESSYII to test the achievable resolution and the experimental feasibility. Figure3a presents an image of the GDE with its microstructure clearly visible. The nickel mesh (which is utilized as conducting and mechanically supporting structure) appears clearly, also some artifacts due to the manufacturing of the radiation windows. In particular, some parts of the radiation windows are not fully planar, which is caused by the milling process. However, this in-situ measurement demonstrated the suitability of our cell concept. Additionally, in Figure3b–d the cell is ﬂooded with deionized water to demonstrate that the electrolyte shows enough contrast, so that its interaction with the ODC can be imaged. Already in this experiment, diﬀerent structural features of the ODC can be observed. For example, relatively large pores appear in between the mesh structure of the Ni-substrate. Materials 2019, 12, 1275 7 of 14 Materials 2019, 12, x FOR PEER REVIEW 7 of 14

FigureFigure 3. 3.Images Images of of oxygen-depolarized oxygen-depolarized cathode cathode (ODC) (ODC) obtained obtained by by synchrotron synchrotron radiation radiation in in 1 1ststGen. Gen. cell,cell, ( a()a) microstructure microstructure of of ODC ODC as as obtained; obtained; (b –(bd–)d normalized) normalized (by (by dry dry cell) cell) images images of ﬁllingof filling process process of 1ofst Gen.1st Gen. cell cell with with deionized deionized water, water, the th liquide liquid front front is clearly is clearly visible. visible.

3.5. Electrochemical Testing of 1st Gen. Cell 3.5. Electrochemical Testing of 1st Gen. Cell While the 1st Gen. cell turned out to be a suitable set-up concerning material and imaging While the 1st Gen. cell turned out to be a suitable set-up concerning material and imaging properties, our next aim was to apply in-operando imaging during the operation of the ODC at properties, our next aim was to apply in-operando imaging during the operation of the ODC at industrially relevant conditions. For operation and electrochemical characterization of ODCs different industrially relevant conditions. For operation and electrochemical characterization of ODCs electrochemical methods, such as LSV and CA, can be applied. LSVs reveal the performance of the ODC different electrochemical methods, such as LSV and CA, can be applied. LSVs reveal the performance for the ORR. Each measurement is affected by the iR of the set-up, reducing the effective potential that of the ODC for the ORR. Each measurement is affected by the iR of the set-up, reducing the effective is applied to the ODC. Therefore, the LSVs have to be corrected in order to compare the performance of potential that is applied to the ODC. Therefore, the LSVs have to be corrected in order to compare electrodes in different set-ups. Commonly, the measurements are either post-corrected by a separately the performance of electrodes in different set-ups. Commonly, the measurements are either post- determined iR value or corrected by the operating program of the potentiostat (CI-mode) directly corrected by a separately determined iR value or corrected by the operating program of the during the measurement. For post-correction, the iR can be evaluated by additional EIS measurements potentiostat (CI-mode) directly during the measurement. For post-correction, the iR can be evaluated andby additional be used for EIS correction measurements of the measured and be used LSVs for in correction the analysis of program.the measured In contrast, LSVs in the the CI-mode analysis interruptsprogram. theIn contrast, LSV measurement the CI-mode for interrupts a short time the LSV and measurement compensates thefor a LSV short instantaneously time and compensates for the determinedthe LSV instantaneously iR-value. for the determined iR-value. st ForFor comparison comparison of of the the di differentfferent cell cell set-ups, set-ups, ODCs ODCs were were characterized characterized in thein the Gaskatel Gaskatel cell cell and and 1 Gen. cell by LSVs under the same conditions (20, 50, and 80 C, 30 wt.% NaOH). The correction of 1st Gen. cell by LSVs under the same conditions (20, 50, and 80◦ °C, 30 wt.% NaOH). The correction of thethe potential potential is is either either done done by by post-correction post-correction for for one one determineddetermined iRiR valuevalue oror measuredmeasured inin CI-mode.CI-mode. LSVsLSVs obtained obtained withwith thethe Gaskatel Gaskatel cellcell of of a a technical technical ODC ODC are are shown shown in in Figure Figure4 a.4a. The The results results reveal reveal distinctdistinct didifferencesfferences between between the the two two iR iR correction correction modes. modes. During During characterization, characterization, ODCsODCs achieve achieve highhigh currentcurrent densitiesdensities andand asas aa result result of of the the current current flow flow and and the the iR, iR, the the electrolyte electrolyte heats heats up. up. Thus,Thus, thethe iR iR value value of of the the set-up set-up changes changes during during the the measurement measurement to to a a lower lower resistance. resistance. Consequently,Consequently, the the post-correctionpost-correction leads leads to to erroneous erroneous results, results, as the as LSVthe curvesLSV curves are corrected are corrected for one for false one iR false value iR (dotted value lines).(dotted However, lines). However, the CI-mode, the CI-mode, designed todesigned avoid this to avoid problem, this faces problem, another faces issue. another Theelectrolyte issue. The inelectrolyte the Gaskatel in the cell Gaskatel (see Figure cell 4(seeb) isFigure not pumped, 4b) is not but pumped, remains but stationary remains stationary so that evolving so that oxygenevolving fromoxygen the from counter the electrodecounter electrode will stay will inside stay the inside electrolyte the electrolyte and hinder and the hinder correction the correction for the right for the iR value.right iR The value. determination The determination of the iR is of disturbed the iR is bydisturbed the oxygen by bubblesthe oxygen and bubbles as a result and the as potentiostat a result the correctspotentiostat the LSVs corrects for changing the LSVs andfor changing false iR values, and false leading iR values, to noisy leading measurements to noisy measurements (full lines). Hence, (full itlines). can be Hence, shown it that can thebe shown characterization that the characterization of ODCs using theof ODCs Gaskatel using cell the is problematicGaskatel cell at is high problematic current 2 densitiesat high current above 5 densities kA m− . above The heating 5 kA upm−². of The the heating electrode up and of electrolytethe electrode during and operationelectrolyte and during the evolvingoperation gas and bubbles the areevolving a big disadvantagegas bubbles ofare the a Gaskatel big disadvantage cell, when measurements of the Gaskatel at high cell, current when densitiesmeasurements are required. at high current densities are required.

Materials 2019, 12, 1275 8 of 14 Materials 2019, 12, x FOR PEER REVIEW 8 of 14

FigureFigure 4. 4.( a(a)) Linear Linear sweep sweep voltammograms voltammograms (LSV) (LSV) of of technical technical ODC ODC in in Gaskatel Gaskatel cell cell (20, (20, 50, 50, and and 80 80◦ °C,C, 3030 wt.% wt.% NaOH, NaOH, correction correction of appliedof applied potential potential by post-correction by post-correction (dotted (dotted lines) orlines) CI-mode or CI-mode (full lines), (full 1 −1 sweeplines), rate sweep 2 mV rate s− 2 );mV (b) s photography); (b) photography of operating of operating Gaskatel Gaskatel cell. cell.

EvenEven though though heating heating up up of of the the electrolyte electrolyte was was remedied remedied by pumpingby pumping electrolyte electrolyte through through the cell, the andcell, evolving and evolving gas on gas the on counter the counter electrode electrode was remedied was remedied by purging by purging of the electrolyteof the electrolyte reservoir reservoir in the 1stinGen. the 1st cell Gen. design, cell design, electrochemical electrochemical testing testing with an with ODC an atODC technical at technical current current densities densities revealed revealed that thethat ﬂow the areaflow chosenarea chosen between between the working the working and and counter counter electrodes electrodes was was still still too too small small in scale. in scale. As As a result,a result, the cellthe resistancecell resistance between between the working the working and counter and counter electrodes electrodes was high, was limiting high, the limiting operation the ofoperation the ODC of to currentthe ODC densities to current below densities 4 kA m- belo2. Pumpingw 4 kA them-². electrolyte Pumping couldthe electrolyte not compensate could fornot thecompensate small exchange for the area small between exchange the working area between and counter the working electrode, and making counter measurements electrode, atmaking high currentmeasurements densities at impossible high current with densities the chosen impossible potentiostat. with Thesethe chosen results potentiostat. show that operationThese results of ODCs show atthat high operation current densitiesof ODCs isat highlyhigh current problematic, densities leading is highly to the problematic, 2nd Gen. cell leading with to further the 2nd improved Gen. cell designwith further features. improved design features.

nd 3.6.3.6. Development Development of of 2 2nd Gen.Gen. Cell

ToTo enable enable the the operation operation and and electrochemical electrochemical characterization characterization of ODCs of ODCs at temperatures at temperatures up to 80up◦ Cto and80 °C 30 wt.%and 30 NaOH, wt.% theNaOH, design the of design the 1st ofGen. the cell 1st wasGen. revised cell was (Figure revised5a). (Figure We focused 5a). We on increasingfocused on theincreasing exchange the area exchange between area the between working the and working counter and electrode counter toelectrode allow forto allow measurements for measurements at high currentat high densities. current densities. The suggested The suggested new design new is showndesign in is 3Dshown in Figure in 3D5b. in The Figure position 5b. The ofthe position electrolyte of the reservoirelectrolyte and reservoir counter and electrode, counter as wellelectrode, as the as shape well ofas the the radiation shape of window, the radiation were modified.window, were The reservoirmodified. was The moved reservoir into was the moved center ofinto the the block, center surrounding of the block, the surrounding radiation window. the radiation The existingwindow. conicalThe existing electrolyte conical chamber electrolyte was chamber extended was by extend a tubulared by reservoir a tubular providing reservoir an providing increased an exchange increased areaexchange between area WE between and CE. WE The and ring-shaped CE. The ring-shaped counter electrode counter around electrode the around radiation the window radiation provides window aprovides more homogeneous a more homogeneous electrical field, electrical which is favorablefield, which for theis electrochemicalfavorable for the measurement. electrochemical The pumpmeasurement. inlet and The outlet pump are placedinlet and perpendicular outlet are placed to the perpendicular direction of the to beam.the direction The top of position the beam. of theThe outlettop position allows forof the the outlet removal allows of evolving for the removal gas from of the evolving CE. The gas components from the CE. of theThe cell components were modified of the ascell outlined, were modified but the materialas outlined, in the but direction the material of the in beam the direction (2 mm radiation of the beam window (2 mm material radiation and window 5 mm electrolytematerial and volume) 5 mm was electrolyte kept identical volume) to thewas 1 keptst Gen. identical cell design. to the The 1st imagingGen. cellproperties design. The of imaging the cell should;properties therefore, of the remaincell should; the same. therefore, remain the same.

Materials 2019, 12, 1275 9 of 14 MaterialsMaterials 2019 2019, 12, 12, x, x FOR FOR PEER PEER REVIEW REVIEW 9 9of of 14 14

nd FigureFigureFigure 5. 5.5. ( (a(aa)) )Schematic SchematicSchematic diagram diagramdiagram of ofof second secondsecond generation generationgeneration (2 (2(2ndnd Gen.) Gen.) cell cell with with relevant relevant parts; parts; ( (b(bb)) )3D-cell 3D-cell3D-cell nd designdesigndesign of ofof 2 22ndnd Gen. Gen. cell. cell.

ElectrochemicalElectrochemicalElectrochemical investigation investigationinvestigation using using the thethe newly newlynewly designed designeddesigned cell cellcell showed showedshowed significant significantsignificant improvements improvementsimprovements withwithwith respect respectrespect to toto the thethe 1 11ststst Gen. Gen.Gen. cell. cell. The The cell cell set-up set-up could could be be us useduseded at at every every temperature temperature without without showing showingshowing anyanyany corrosion corrosioncorrosion effects. eeffects.ffects. Furthermore, Furthermore, the thethe modified modifiedmodified ce cellcellll design design demonstrates demonstrates a a significant significantsignificant improvement improvement whenwhenwhen measuring measuringmeasuring LSVs LSVsLSVs up upup to toto high highhigh current currentcurrent densities. densities.densities. In In Figure Figure 6a,6 6a,a, the the LSVsLSVs ofof technicaltechnical ODCODC areare comparedcomparedcompared for forfor three threethree different didifferentfferent oper operatingoperatingating temperatures. temperatures.temperatures. The The ORR ORR activity activity increased increased with with temperature, temperature, − − 2 reachingreachingreaching current currentcurrent densities densities up up to to 9 9 kA kA m m ².−². The .The The LSVs LSVs LSVs were were were recorded recorded recorded using using using the the the CI-mode. CI-mode. CI-mode. Contrarily Contrarily Contrarily to to 1to1stst Gen. 1Gen.st Gen. cell, cell, cell,we we did wedid not didnot observe notobserve observe any any anydrawbacks, drawbacks, drawbacks, such such suchas as a a aslimited limited a limited current current current density, density, density, due due to dueto the the to cell thecell design.celldesign. design. To To demonstrate demonstrate To demonstrate that that the thatthe ce ce thellll can can cell be be can used used be reliably usedreliably reliably to to perform perform to perform long-term long-term long-term measurements measurements measurements while while applyingwhileapplying applying a a constant constant a constant potent potential, potential,ial, CA CA measurements measurements CA measurements were were carried carried were carriedout out at at different different out at di temperatures.ff temperatures.erent temperatures. In In Figure Figure In 6bFigure6b itit isis6 b shownshown it is shown thatthat at thatat 8080 at °C°C 80 andand◦C and threethree three differentdifferent different appliedapplied applied potentials,potentials, potentials, stablestable stable measurementsmeasurements measurements cancan be bebe achievedachievedachieved for forfor several several hours. hours. The The applied applied potentialspotentials potentials inin inthe the the CA CA CA measurements measurements measurements were were were not not not corrected corrected corrected for for for iR, iR,butiR, but but were were were chosen chosen chosen to representto to represent represent diff different erentdifferent operation operation operation conditions. conditions. conditions. In brief,In In brief, brief, the the 2thend 2 Gen.2ndnd Gen. Gen. cell cell cell set-up set-up set-up fulfills fulfills fulfills all allrequirementsall requirementsrequirements for the forfor detailed thethe detaileddetailed and systematic andand systematsystemat imagingicic imaging ofimaging processes ofof processes occurringprocesses duringoccurringoccurring ODC duringduring operation ODCODC in operationtechnicallyoperation in in relevanttechnically technically conditions. relevant relevant conditions. conditions.

Figure 6. Electrochemical characterization in 2nd Gen. cell (a) LSV of ODC recorded up to high current FigureFigure 6. 6. Electrochemical Electrochemical characterization characterization in in 2 2ndnd Gen. Gen. cell cell ( a(a) )LSV LSV of of ODC ODC recorded recorded up up to to high high current current densities (20, 50, and 80 ◦C, 30 wt.% NaOH, correction of applied potential by CI-mode, sweep rate densities1 (20, 50, and 80 °C, 30 wt.% NaOH, correction of applied potential by CI-mode, sweep rate densities2 mV s− (20,); (b 50,) Chronoamperometry and 80 °C, 30 wt.% NaOH, (CA) at correction 0.2, 0.5, and of applied 0.9 V vs. potential RHE (80 by◦ C,CI-mode, 30 wt.% sweep NaOH, rate no -1 2iR-correction)2 mV mV s -1s);); ( b(b) )Chronoamperometry Chronoamperometry showing very stable (CA) features. (CA) at at 0.2, 0.2, 0.5, 0.5, and and 0.9 0.9 V V vs. vs. RHE RHE (80 (80 °C, °C, 30 30 wt.% wt.% NaOH, NaOH, no no iR- iR- correction)correction) showing showing very very stable stable features. features. 3.7. Micro-CT testing of 2nd Gen. Cell 3.7. Micro-CT testing of 2ndnd Gen. Cell 3.7. Micro-CTThe cell testing set-up of was 2 furtherGen. Cell tested while performing radiographies in a lab-scale micro-CT. In FigureTheThe7a–c cell cell the set-up set-up ﬂooding was was offurther further the in-operando tested tested while while 2 nd perfor perforGen.ming cellming with radiographies radiographies 30 wt.% NaOH in in a a lab-scale electrolytelab-scale micro-CT. micro-CT. is displayed. In In FigureForFigure this 7a–c 7a–c purpose, the the flooding flooding a technical of of the the ODC in-operando in-operando was mounted 2 2ndnd Gen. Gen. in the cell cell 2nd with withGen. 30 30 cellwt.% wt.% and NaOH NaOH the electrolyte electrolyte electrolyte was is is displayed. pumpeddisplayed. at ForFor this this purpose, purpose, a a technical technical ODC ODC was was mounted mounted in in the the 2 2ndnd Gen. Gen. cell cell and and the the electrolyte electrolyte was was pumped pumped atat a a reduced reduced pumping pumping rate rate to to obtain obtain a a higher higher time time resolution. resolution. The The flooding flooding process process within within the the field field

Materials 2019, 12, 1275 10 of 14

Materials 2019, 12, x FOR PEER REVIEW 10 of 14 a reduced pumping rate to obtain a higher time resolution. The flooding process within the field of viewof view took took 23 min23 min using using an exposurean exposure time time of 2.5 of s2. per5 s frame.per frame. The The ODC ODC structure structure is visible is visible and theand fine the meshfine mesh structure structure of the of conductingthe conducti Ni-substrateng Ni-substrate appears appears to be to regular.be regular. The The rising rising electrolyte electrolyte front front is recognizableis recognizable from from the the higher higher absorption absorption of of the the X-ray X-ray by by the the electrolyte electrolyte and and represented represented byby aa darkerdarker color.color. TheThe in-operandoin-operando cellcell waswas operated with didifffferenterent technicaltechnical electrodeselectrodes andand operatingoperating conditionsconditions forfor severalseveral days days without without any any problems. problems. During During this this process, process, images images of the of operating the operating ODCs ODCs were takenwere (Figuretaken (Figure7d–f) at 7d–f) di fferent at different time stages. time Thesestages. results These represent results represent the interaction the interaction of the electrode of the structureelectrode withstructure the electrolyte. with the Theelectrolyte. appearing The droplets appearing could droplets be formed could by evaporationbe formed andby evaporation condensation and of electrolytecondensation on theof electrolyte colder backside on the ofcolder the electrode. backside of A the more electrode. detailed A investigation more detailed and investigation discussion willand bediscussion subject of will a forthcoming be subject of publication a forthcom bying Paulisch publication et al. by Paulisch et al.

FigureFigure 7.7. ((aa––cc)) FloodingFlooding processprocess ofof thethe 22ndnd Gen.Gen. cell, recorded usingusing aa lab-scalelab-scale micro-CT.micro-CT. (technical(technical ODC,ODC, 3030 wt.%wt.% NaOHNaOH solution);solution); ((dd––ff)) in-operandoin-operando radiographyradiography measurementsmeasurements at at 75 75◦ °C.C. 3.8. Neutron Testing 3.8. Neutron Testing In addition to the radiography measurements with X-rays, the cell should also enable imaging In addition to the radiography measurements with X-rays, the cell should also enable imaging using neutrons, as complementary information can be obtained. However, testing the cell with neutron using neutrons, as complementary information can be obtained. However, testing the cell with radiation is only possible with a modified cell set-up, adapted to the respective requirements. Since neutron radiation is only possible with a modified cell set-up, adapted to the respective requirements. neutron radiation is strongly absorbed by hydrogen- and carbon-containing materials, the cell materials Since neutron radiation is strongly absorbed by hydrogen- and carbon-containing materials, the cell had to be replaced, in parts, with more suitable polymers. PEEK consists of a significant amount of materials had to be replaced, in parts, with more suitable polymers. PEEK consists of a significant carbon and hydrogen in its monomer structure. Thus, it had to be exchanged. A possible alternative amount of carbon and hydrogen in its monomer structure. Thus, it had to be exchanged. A possible material is polychlorotrifluoroethylene (PCTFE), since it provides a sufficient resistivity to the highly alternative material is polychlorotrifluoroethylene (PCTFE), since it provides a sufficient resistivity alkaline electrolyte (up to 80 C) and a high material hardness, making it suitable to be processed. to the highly alkaline electrolyte◦ (up to 80 °C) and a high material hardness, making it suitable to be Additionally, its low carbon content and absence of hydrogen in the polymer will result in a low processed. Additionally, its low carbon content and absence of hydrogen in the polymer will result absorption of the neutron beam. The standard electrolyte, an aqueous solution of 30 wt.% NaOH, in a low absorption of the neutron beam. The standard electrolyte, an aqueous solution of 30 wt.% would absorb a large fraction of the beam. Therefore, the in-operando tests with neutron radiation NaOH, would absorb a large fraction of the beam. Therefore, the in-operando tests with neutron have been performed using an electrolyte of 30 wt.% NaOD in D2O. With the replacement of the radiation have been performed using an electrolyte of 30 wt.% NaOD in D2O. With the replacement electrolyte, the in-operando set up is also suitable for neutron imaging and was tested at CONRAD2 of the electrolyte, the in-operando set up is also suitable for neutron imaging and was tested at beamline (see experimental set-up Figure1b). In Figure8a–c the process of emptying the 2 nd Gen. CONRAD2 beamline (see experimental set-up Figure 1b). In Figure 8a–c the process of emptying the cell from deuterated electrolyte is demonstrated and shows the suitability of the cell for neutron 2nd Gen. cell from deuterated electrolyte is demonstrated and shows the suitability of the cell for radiography. A clearly visible electrolyte contrast and interaction of electrolyte with the ODC structure neutron radiography. A clearly visible electrolyte contrast and interaction of electrolyte with the ODC can be demonstrated and detected. Furthermore, in Figure8d an inverted in-operando radiography structure can be demonstrated and detected. Furthermore, in Figure 8d an inverted in-operando radiography image is presented to show the detectable electrolyte contrast. Bright spots indicate a higher amount of absorbing electrolyte in the beam path.

Materials 2019, 12, 1275 11 of 14 image is presented to show the detectable electrolyte contrast. Bright spots indicate a higher amount of Materialsabsorbing 2019 electrolyte, 12, x FOR PEER in the REVIEW beam path. 11 of 14

FigureFigure 8. ((aa––cc)) Neutron radiography of pumping process inin 22ndnd Gen.Gen. cell; cell; ( (dd)) in-operando in-operando neutron neutron radiographyradiography image image of of ODC. ODC.

InIn Table Table 1,1, capabilities capabilities of of each each previously previously described described cell cell design design and and set-up set-up are summarized. are summarized. The st advantagesThe advantages of the of the1st Gen. 1 Gen. cell, cell, over over the thecommercial commercial Gaskatel Gaskatel design, design, were were the the changes changes to to the the electrolyteelectrolyte and and gas gas supply. supply. By By careful careful design, design, it it wa wass possible possible to to have have the the lowest lowest amount amount of of absorbing absorbing st materialmaterial in in the the beam beam path. path. Most Most requirements requirements fo forr successful successful imaging imaging were being met, and the 1 st Gen.Gen. cellcell enabled enabled in-situ in-situ radiography radiography imaging imaging of of ODC ODC for for the the first first time. time. However, However, it it was was not not possible possible to to meetmeet the the necessary electrochemical requirements, requirements, in in particular high current densities. The The further further nd improvementimprovement to to 2 2nd Gen.Gen. cell cell allowed allowed in-operando in-operando imaging imaging with with both both X-ray X-ray and and neutron neutron radiation radiation at at nd industriallyindustrially relevant relevant conditions forfor the firstfirst time. Therefore, thethe 2 nd Gen.Gen. cell cell enabled enabled us us to to investigate investigate thethe electrolyte electrolyte invasion invasion and and flooding flooding taking place during the operation of an ODC.

Table 1. Development of cell capabilities. Table 1. Development of cell capabilities. st nd CellCell Requirements requirements GaskatelGaskatel Cell Cell 1 Gen. 1st Gen. Cell Cell 2nd Gen.Gen. Cell ResistivityResistivity to alkaline to alkaline media media    ElectrolyteElectrolyte and and gas supplygas supply ( ) ()   Electrochemical ()()  Electrochemicalcharacterization characterization () () Imaging radiation: Imaging radiation: X-ray (PEEK windows)  ()  X-ray (PEEK windows) () Neutron (PCTFE windows)  ()  Neutron (PCTFE windows) ()  Suitable, Suitable, ( () suitable) suitable with with restrictions,restrictions, not not suitable. suitable.

4. Conclusions and Outlook 4. Conclusions and Outlook In-operando cells could be devised and optimized to allow for the collection of X-ray and neutronIn-operando imaging data cells during could be operation devised andof ODCs optimized in highly to allow alkaline for theconditions. collection While of X-ray the and1st Gen. neutron cell st stillimaging suffered data from during a high operation cell resistance, of ODCs in ina seco highlynd approach, alkaline conditions. the electrolyt Whilee reservoir the 1 wasGen. modified cell still to allow for the beam to pass unhindered through the cell and the cell performance was significantly improved. The 2nd Gen. cell demonstrated reliable, stable and reproducible LSV curves in the X-ray imaging sample environment. For allowing neutron imaging, the beam windows were manufactured in a fashion that they can be easily exchanged to less neutron-absorbing polymer types. First X-ray

Materials 2019, 12, 1275 12 of 14 suffered from a high cell resistance, in a second approach, the electrolyte reservoir was modified to allow for the beam to pass unhindered through the cell and the cell performance was significantly improved. The 2nd Gen. cell demonstrated reliable, stable and reproducible LSV curves in the X-ray imaging sample environment. For allowing neutron imaging, the beam windows were manufactured in a fashion that they can be easily exchanged to less neutron-absorbing polymer types. First X-ray and neutron imaging results of an operating ODC in realistic working conditions were obtained and demonstrated the filling of the porous electrode structure with the liquid electrolyte. In future experiments we will focus on the investigation of the electrolyte behavior during electrochemical measurements, applying different electrode structures and compare pristine and aged electrodes.

Author Contributions: Conceptualization, T.T., I.M., and C.R.; data curation, M.G., M.P. and A.H.; formal analysis, M.G., M.P. and A.H.; funding acquisition, T.T., I.M. and C.R.; investigation, M.G., M.P., D.F. and B.E.; methodology, M.G. and A.H.; project administration, M.G. and C.R.; resources, T.T., I.M. and C.R.; supervision, C.R.; validation, M.G. and M.P.; visualization, M.G. and M.P.; writing—original draft, M.G. and C.R.; writing—review and editing, M.G. and C.R. Funding: This research was funded by [Deutsche Forschungsgemeinschaft] research grants [FOR 2397; RO 2454/16-1, MA 5039/3-1 and TU 89/13-1). The APC was funded by [Freie Universität Berlin]. Acknowledgments: We also want to thank the HZB for the allocation of synchrotron and neutron radiation beamtime at BAMline/ V7Beamline and Nikolay Kardjilov and Henning Markötter for the support. In addition, the help of the workshop at Freie Universität Berlin is gratefully acknowledged. Conﬂicts of Interest: The authors declare that they have no conﬂicts of interest.

References

1. IHS Chlor-Alkali World Analysis. 2017. Available online: https://www.ihs.com/products/world-petro- chemical-analysis-chlor-alkali.html (accessed on 10 November 2017). 2. Grotheer, M.; Alkire, R.; Varjian, R. Industrial Electrolysis and Electrochemical Engineering. Electrochem. Soc. Interface 2006, 15, 52–54. 3. O’Brien, T.F.; Bommaraju, T.V.; Hine, F. Handbook of Chlor-Alkali Technology; Springer Nature: New York, NY, USA, 2005. 4. Bolaños-Álvarez, Y.; Alonso-Hernandez, C.M.; Morabito, R.; Díaz-Asencio, M.; Pinto, V.; Gómez-Batista, M. Mercury contamination of riverine sediments in the vicinity of a mercury cell chlor-alkali plant in Sagua River, Cuba. Chemosphere 2016, 152, 376–382. [CrossRef][PubMed] 5. Millet, P. Chlor-Alkali Technology: Fundamentals, Processes and Materials for Diaphragms and Membranes; Elsevier BV: Amsterdam, The Netherlands, 2013; pp. 384–415. 6. Recent trends in the OECD: Energy and CO2 Emissions. Available online: https://www.iea.org/media/ statistics/Recent_Trends_in_the_OECD.pdf (accessed on 28 February 2017). 7. Moussallem, I.; Jörissen, J.; Kunz, U.; Pinnow, S.; Turek, T. Chlor-alkali electrolysis with oxygen depolarized cathodes: History, present status and future prospects. J. Appl. Electrochem. 2008, 38, 1177–1194. [CrossRef] 8. Neues Verfahren senkt Energieverbrauch bei Chlor-Produktion um 30%. Available online: https://www. process.vogel.de/neues-verfahren-senkt-energieverbrauch-bei-chlor-produktion-um-30-a-408117/ (accessed on 2 June 2016). 9. Jung, J.; Postels, S.; Bardow, A. Cleaner chlorine production using oxygen depolarized cathodes? A life cycle assessment. J. Clean. Prod. 2014, 80, 46–56. [CrossRef] 10. Garcia-Herrero, I.; Margallo, M.; Onandía, R.; Aldaco, R.; Irabien, A. Environmental challenges of the chlor-alkali production: Seeking answers from a life cycle approach. Sci. Total. Environ. 2017, 580, 147–157. [CrossRef] 11. Kintrup, J.; Millaruelo, M.; Trieu, V.; Bulan, A.; Mojica, E.S. Gas Diffusion Electrodes for Efficient Manufacturing of Chlorine and Other Chemicals. Electrochem. Soc. Interface 2017, 26, 73–76. [CrossRef] 12. Danner, T.; Eswara, S.; Schulz, V.P.; Latz, A. Characterization of gas diffusion electrodes for metal-air batteries. J. Power Sources 2016, 324, 646–656. [CrossRef] 13. Park, J.; Oh, H.; Ha, T.; Lee, Y.I.; Min, K. A review of the gas diffusion layer in proton exchange membrane fuel cells: Durability and degradation. Appl. Energy 2015, 155, 866–880. [CrossRef] Materials 2019, 12, 1275 13 of 14

14. Park, S.; Lee, J.-W.; Popov, B.N. A review of gas diffusion layer in PEM fuel cells: Materials and designs. Int. J. Hydrogen Energy 2012, 37, 5850–5865. [CrossRef] 15. Chatenet, M.; Genies-Bultel, L.; Aurousseau, M.; Durand, R.; Andolfatto, F. Oxygen reduction on silver catalysts in solutions containing various concentrations of sodium hydroxide—Comparison with platinum. J. Appl. Electrochem. 2002, 32, 1131–1140. [CrossRef] 16. Moussallem, I.; Pinnow, S.; Wagner, N.; Turek, T. Development of high-performance silver-based gas-diffusion electrodes for chlor-alkali electrolysis with oxygen depolarized cathodes. Chem. Eng. Process. Intensif. 2012, 52, 125–131. [CrossRef] 17. Neumann, M.; Osenberg, M.; Hilger, A.; Franzen, D.; Turek, T.; Manke, I.; Schmidt, V. On a pluri-Gaussian model for three-phase microstructures, with applications to 3D image data of gas-diffusion electrodes. Comput. Mater. Sci. 2019, 156, 325–331. [CrossRef] 18. Zils, S.; Timpel, M.; Arlt, T.; Wolz, A.; Manke, I.; Roth, C. 3D Visualisation of PEMFC Electrode Structures Using FIB Nanotomography. Fuel Cells 2010, 10, 966–972. [CrossRef] 19. Banhart, J.; Borbély, A.; Dzieciol, K.; Garcia-Moreno, F.; Manke, I.; Kardjilov, N.; Kaysser-Pyzalla, A.R.; Strobl, M.; Treimer, W. X-ray and neutron imaging—Complementary techniques for materials science and engineering. Int. J. Mater. 2010, 101, 1069–1079. [CrossRef] 20. AlRwashdeh, S.S.; Manke, I.; Markötter, H.; Klages, M.; Göbel, M.; Haußmann, J.; Scholta, J.; Banhart, J. In Operando Quantification of Three-Dimensional Water Distribution in Nanoporous Carbon-Based Layers in Polymer Electrolyte Membrane Fuel Cells. ACS Nano 2017, 11, 5944–5949. [CrossRef] 21. Markötter, H.; Manke, I.; Krüger, P.; Arlt, T.; Haussmann, J.; Klages, M.; Riesemeier, H.; Hartnig, C.; Scholta, J.; Banhart, J. Investigation of 3D water transport paths in gas diffusion layers by combined in-situ synchrotron X-ray radiography and tomography. Electrochem. Commun. 2011, 13, 1001–1004. [CrossRef] 22. Fazeli, M.; Hinebaugh, J.; Fishman, Z.; Tötzke, C.; Lehnert, W.; Manke, I.; Bazylak, A. Pore network modeling to explore the effects of compression on multiphase transport in polymer electrolyte membrane fuel cell gas diffusion layers. J. Power Sources 2016, 335, 162–171. [CrossRef] 23. Sasabe, T.; Deevanhxay, P.; Tsushima, S.; Hirai, S. Investigation on the effect of microstructure of proton exchange membrane fuel cell porous layers on liquid water behavior by soft X-ray radiography. J. Power Sources 2011, 196, 8197–8206. [CrossRef] 24. Manke, I.; Hartnig, C.; Grünerbel, M.; Lehnert, W.; Kardjilov, N.; Haibel, A.; Hilger, A.; Banhart, J.; Riesemeier, H. Investigation of water evolution and transport in fuel cells with high resolution synchrotron x-ray radiography. Appl. Phys. Lett. 2007, 90, 174105. [CrossRef] 25. Kardjilov, N.; Hilger, A.; Manke, I.; Griesche, A.; Banhart, J. Imaging with Cold Neutrons at the CONRAD-2 Facility. Phys. Procedia 2015, 69, 60–66. [CrossRef] 26. Kardjilov, N.; Manke, I.; Woracek, R.; Hilger, A.; Banhart, J. Advances in neutron imaging. Mater. Today 2018, 21, 652–672. [CrossRef] 27. Lehmann, E.; Boillat, P.; Scherrer, G.; Frei, G.; Lehmann, E. Fuel cell studies with neutrons at the PSI’s neutron imaging facilities. Instrum. Methods Phys. Sect. A Accel. Spectrometers Detect. Assoc. Equip. 2009, 605, 123–126. [CrossRef] 28. Boillat, P.; Scherer, G.; Wokaun, A.; Frei, G.; Lehmann, E.; Lehmann, E. Transient observation of 2H labeled species in an operating PEFC using neutron radiography. Electrochem. Commun. 2008, 10, 1311–1314. [CrossRef] 29. Bellows, R.J.; Lin, M.Y.; Arif, M.; Thompson, A.K.; Jacobson, D. Neutron Imaging Technique for In Situ Measurement of Water Transport Gradients within Nafion in Polymer Electrolyte Fuel Cells. J. Electrochem. Soc. 1999, 146, 1099. [CrossRef] 30. Schröder, A.; Wippermann, K.; Lehnert, W.; Stolten, D.; Sanders, T.; Baumhöfer, T.; Kardjilov, N.; Hilger, A.; Banhart, J.; Manke, I. The influence of gas diffusion layer wettability on direct methanol fuel cell performance: A combined local current distribution and high resolution neutron radiography study. J. Power Sources 2010, 195, 4765–4771. [CrossRef] 31. Manke, I.; Hartnig, C.; Kardjilov, N.; Messerschmidt, M.; Hilger, A.; Strobl, M.; Lehnert, W.; Banhart, J. Characterization of water exchange and two-phase flow in porous gas diffusion materials by hydrogen-deuterium contrast neutron radiography. Appl. Phys. Lett. 2008, 92, 244101. [CrossRef] Materials 2019, 12, 1275 14 of 14

32. Manke, I.; Hartnig, C.; Grünerbel, M.; Kaczerowski, J.; Lehnert, W.; Kardjilov, N.; Hilger, A.; Banhart, J.; Treimer, W.; Strobl, M. Quasi–in situ neutron tomography on polymer electrolyte membrane fuel cell stacks. Appl. Phys. Lett. 2007, 90, 184101. [CrossRef] 33. Banhart, J. Advanced Tomographic Methods in Materials Research and Engineering; Oxford University Press: Oxford, NY, USA, 2008. 34. Schindelin, J.; Arganda-Carreras, I.; Frise, E.; Kaynig, V.; Longair, M.; Pietzsch, T.; Preibisch, S.; Rueden, C.; Saalfeld, S.; Schmid, B.; et al. Fiji: An open-source platform for biological-image analysis. Nat. Methods 2012, 9, 676–682. [CrossRef] 35. Heiland, K.; Hill, D.J.T.; O’Donnell, J.H.; Pomery, P.J. Radiation degradation of poly(arylene ether ketone)s. Polym. Adv. Technol. 1994, 5, 116–121. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).