Understanding Mass Transport at High Current Densities in an Industrial Scale PEM Electrolysis Test Cell - a Segmented Along-the-Channel Approach

Abstract

Industrial PEM water electrolysis stack designs can suffer from an unevenly distributed water amount over the cell area. This can lead to performance differences and thermal hot spots due to the lack of reactant supply and poor thermal management. Undersupplied spots could also degrade more quickly [1,2]. Especially in the water flow direction, gradients are expected due to the water consumption and gas evolution and accumulation along the supply channels.To face these issues, we built up a segmented along the channel (AtC) PEM electrolysis test cell in industrial scale for locally-resolved investigations. The cell has a length of 30 cm at 2 cm cell width with a straight parallel flow field. To minimize transverse currents the cell is divided into 10 equal segments along the channels. The main focus of our investigation is locally-resolved electrochemical impedance spectroscopy (EIS). With a shunt resistor approach, it is feasible to measure the impedance of each segment and the mean cell in parallel. Additionally, it is possible to measure the current density and temperature distribution of the anodic bipolar plate highly resolved using 120 sensors over the cell length.With this test cell we want to understand mass transport processes along the channel in industrial scale. Diffusion and mass transport overpotentials are usually dominant at high current densities. To be able to face extreme conditions and investigate possible future operation points the cell was designed up to 10 A∙cm-2 (600 A absolute) and successfully tested. Furthermore, an operation of 10 bar differential and equal pressure is possible.Figure 1 shows the cell design (a), the mean cell polarization curve of the AtC cell (60 cm²) in comparison with measurements of the same setup in our 4 cm² test cell (b) and locally-resolved high frequency resistance (HFR) free EIS at 7 A∙cm-2 (c). It is noticeable, that the here used commercially available materials do not show high mass transport limitations, neither in the 4 cm² ISE-reference cell nor in the 60 cm² AtC cell. An increase of the slope of the polarization curve towards higher current densities remains out, see Figure 1 (b). Instead of this the slope of the polarization curve is constantly decreasing. This can not only be explained by the temperature increase of the catalyst-coated membrane (CCM) due to higher heat dissipation. Using EIS, we here can detect an additional electrochemical process at low frequencies of inductive type which is increasing with increasing current density. In the locally-resolved EIS in Figure 1 (c) the inductive process (positive imaginary values) is detectable as the so called “inductive loop”. This phenomenon has a negative resistance magnitude and therefore a beneficial impact on the cell performance. Inductive Loops are discussed in several electrochemical applications, like batteries and fuel cells [3,4]. Despite this, in PEM electrolysis this process has not been discussed yet. We could show that this phenomenon is reproducible and has an important impact on the cell performance and impedance-based performance analyses when operating at current densities > 1 A∙cm-2 [5].For diffusion analyses, the inductive loop is essential to be considered since both processes happen at similar frequencies. In Figure 1 (c) an increasing of the number of time constants at frequencies < 100 Hz is detectable towards the cell outlet (e.g. segment 9 and 10). These slow processes are most likely diffusion related.Our upcoming investigations will face locally-resolved measurements with varying structural parameters of porous transport layers (PTL) and CCMs with different anodic catalyst loadings to gain a better understanding of mass transport processes for future PEM electrolysis cells along the channel.Reference List:[1] IMMERZ, C., et al. Experimental characterization of inhomogeneity in current density and temperature distribution along a single-channel PEM water electrolysis cell. Electrochimica acta, 2018, 260. Jg., S. 582-588.[2] VERDIN, B., et al. Operando current mapping on PEM water electrolysis cells. Influence of mechanical stress. International Journal of Hydrogen Energy, 2017, 42. Jg., Nr. 41, S. 25848-25859.[3] KLOTZ, Dino. Negative capacitance or inductive loop?–A general assessment of a common low frequency impedance feature. Electrochemistry Communications, 2019, 98. Jg., S. 58-62.[4] GERLING, Christophe, et al. Experimental and Numerical Investigation of the Low-Frequency Inductive Features in Differential PEMFCs: Ionomer Humidification and Platinum Oxide Effects. Journal of The Electrochemical Society, 2023, 170. Jg., Nr. 1, S. 014504.[5] HENSLE, Niklas, et al. On the role of inductive loops at low frequencies in PEM electrolysis. Electrochemistry Communications, 2023, 155. Jg., S. 107585. Figure 1