Modelling and Numerical Investigation of Taylor Flow on the Performance of Electrochemical Flow Cells

Abstract

Electrochemical reduction of CO2 using renewable energy sources is one of the promising avenues to pursue towards mitigating the emissions of the notorious CO2. However, the CO2 electrolysis in aqueous systems, due to the low solubility of CO2, are severely limited by mass transfer. State of the art review shows that a significant amount of the research is done to improve mass transfer, where a variety of electrolyser designs were studied. Despite the effort, the challenge to enhance mass transfer remains and is the focus of the present work. To improve mass transfer, an innovative concept is proposed - Taylor flow in an electrochemical flow cell. Taylor flow has been extensively studied in the literature, especially in micro channels and monolith reactors. Therefore the characteristics of the Taylor flow are known to a certain extent. But they were never tested in an electrochemical system. For that reason, a numerical investigation is carried out to assess the performance of Taylor flow on the flow cell. A simplified 2D model was formulated using a unit cell strategy and is verified based on experimental data and theoretical concepts. The effect of dissolving bubbles is also modelled using a quasi-steady-state analysis The results of the 2D model show a significant improvement in the performance, i.e. in the current densities of the electrochemical cell compared to a typical flow cell. The maximum calculated current densities increase by an order of magnitude under certain flow conditions. The dissolution studies showed that the current densities deteriorate with time. Nevertheless, the overall performance is still higher than the typical flow cell. Finally, based on the insights from the 2D model, a 1D model is suggested to estimate the current densities and dissolution rates. The present study showed promising results for using Taylor flow in an electrochemical cell. The proposed 2D model can help in aiding future modelling studies while the 1D model can give simple estimates for the experimental work.