Optimization of Flow-By CO2 Electrolyzers: Impact of Differential Pressure and Gas Diffusion Layer Composition

Abstract

The electrochemical reduction of CO2 (eCO2R) into valuable chemicals offers a great ecological solution to lower anthropogenic CO2 in the atmosphere, since it can close the carbon cycle by utilizing renewable energy, resulting in a sustainable carbon recycling system. Consequently, it has gathered significant scientific interest over the past decade [1,2]. However, to make significant progress towards making the process industrially feasible, it would be beneficial to replace the typical anodic oxygen evolution reaction at the counter electrode with an economically more interesting one, like alkane dehydrogenation, while at the same time lowering cell potential and increasing energy efficiency. However, this requires the CO2 reduction to operate efficiently at elevated temperatures (up to 100°C) [3]. Unfortunately, little is known on the impact of elevated temperatures on the overall performance of CO2 electrolyzers and its components.In this research, we studied the effect of increasing the temperature on CO2 electrolyzers to ultimately enable selective and stable eCO2R to formic acid not only at room temperature. Heating the system leads to many changes (e.g. decreased electrolyte surface tension, decreased hydrophobicity of the gas diffusion layer (GDL), lower dissolved CO2 content and, increased CO2 diffusion coefficient), all affecting the performance of a conventional CO2 electrolyzer, consequently raising the need for re-evaluating many of its components and their configuration, considering they have been optimized for operation at ambient conditions. Specifically, we focused on the three-phase boundary, where the eCO2R reaction actually takes place. Logically, it is important that this boundary is at the exact location of the catalyst layer (CL), as the catalyst is the active species towards the desired reaction. To this end, GDLs are specifically designed to align this boundary to the right location through variations in hydrophobic additives, thickness, porosity, etc. However, we have found that all these efforts to perfect the GDL's properties can be forfeited if the differential pressure across it were to change as it shifts the three-phase boundary. A shift inwards the GDL will result in a flooded CL, lengthening the diffusion path of the gaseous CO2 in the electrolyte to the active sites of the CL resulting in increased hydrogen evolution. A shift outward the GDL will result in a CL that is not fully used, as such lowering the activity. Since the CL is a thin layer, it can be easily understood that the margin within which the boundary can shift without becoming catastrophic for the overall performance is very small, making it extremely critical to have precise control over its position. Changing the temperature of the system will affect the location of the boundary layer and by optimizing the differential pressure we can shift its location back to its optimal position and increase the performance of the CO2 electrolyzer.We investigated the influence of deliberately altering the differential pressure across a GDE in a flow-by electrolyzer with Bi2O3 nanoparticles for the eCO2R towards formate, at different temperatures. At higher temperatures (i.e. 85°C), flooding was a pronounced problem, which resulted in a rapid decrease of performance, and after 24 hours of electrolysis only a 40% faradaic efficiency towards formate (FEHCOOH) was maintained from the initial 70%. We discovered that by increasing the differential pressure, by elevating the backpressure at the gas side, we could limit flooding of the GDE. By mitigating the flooding, the system was able to maintain a better performance, i.e. FEHCOOH of 65% after 24 hours of electrolysis was maintained, which is a 1.6 factor increase (Figure). The present findings confirm that, without altering any of the more significant electrochemical aspects of the electrolyzer, it is possible to increase performance by solely varying the differential pressure over the GDE, showing the importance of also optimizing the operational conditions.Once the optimal differential pressure of the system was established, it was still necessary to investigate the GDE structure itself. Indeed, the GDE should be designed to appropriately mediate all transport processes necessary to achieve high eCO2R performance and further reduce flooding phenomena, by favourably affecting CO2 transport, local pH and water transport [4]. Therefore, the next step was to investigate and compare other commercially available GDLs. For this research we selected commercially available GDLs that are primarily used in high-temperature proton exchange fuel cells, since their required characteristics (hydrophobicity, thickness and pore structure) are similar to the needs of our system.By re-evaluating and optimizing these (operational) conditions, we are getting closer to making the electrochemical CO2 reduction efficient, also at elevated temperatures.[1] https://doi.org/10.1016/j.petlm.2016.11.003[2] https://doi.org/10.1016/j.joule.2017.09.003[3] https://doi.org/10.1002.celc.201900872[4] https://doi.org/10.1038/s41578-021-00356-2 Figure 1