In alkaline and neutral media CO₂ to multicarbon electrolyzers, CO₂ rapidly converts to (bi)carbonate, crosses over the membrane and mixes with the O₂-rich anodic gas stream. This so-called CO₂ crossover phenomenon limits the carbon efficiency in those systems to ~25%, unless an energy-intensive, chemical separation step is performed to recover CO₂ from the anodic gas stream.

Bipolar membranes (BPMs), under reverse bias, are known to eliminate CO₂ crossover and convert (bi)carbonate back to CO­₂ – by generating a proton flow towards the cathode. However, implementing BPMs in CO₂ reduction electrolyzers is challenging: one must keep the cathode locally alkaline to enable the reaction and ensure the regenerated CO₂ can participate in the cathodic reactions.

Assisted by a numerical simulation model (Fig. 1a, 1c and 1d), we sought to develop a BPM-based electrolyzer design (Fig. 1b) that meets the criteria mentioned above. We discovered that the cathodic alkalinity and mass transfer efficiency of the regenerated CO₂ are greatly affected by the gap distance between cathode and BPM, and by the buffering capacity of the electrolyte filling in this gap.

The simulation results suggest the following design principles for the BPM-based CO₂ reduction electrolyzers: the local cathode pH and the diffusion layer thickness of the regenerated CO₂ increase as the electrolyte thickness increases; the buffering capacity of the catholyte increases the diffusion layer thickness and reduces transport. Precise control of the thickness of a non-buffering catholyte should thus offer a route to high carbon efficiency, CO₂ reduction selectivity and reaction rate.

The experimental results verified the designed functions of the BPM-based CO₂ reduction electrolyzer. We report the CO₂ crossover < 0.5% with a single-pass carbon efficiency of up to 78%, with other key performance metrics close to the anion membrane-based counterparts.

Fig. 1 | The configuration and simulation of the electrolyzer design. (a) The CO₂ (solid lines) and pH distributions (dashed lines) in the 65 μm-thick electrolyte layer. The positions where the (bi)carbonates revert to CO₂ are marked (red for non-buffering and black for buffering electrolyte). (b) The schemes and the mass transfer in the BPM-based CO₂ reduction electrolyzer. (c) The pH distribution inside the electrolyte layer. (d) The dissolved CO₂ concentration profile inside the electrolyte layer.