Operando cathode activation with alkali metal cations for high current density operation of water-fed zero-gap carbon dioxide electrolyzers

Abstract

Continuous-flow electrolyzers allow CO₂ reduction at industrially relevant rates, but long-term operation is still challenging. One reason for this is the formation of precipitates in the porous cathode from the alkaline electrolyte and the CO₂ feed. Here we show that while precipitate formation is detrimental for the long-term stability, the presence of alkali metal cations at the cathode improves performance. To overcome this contradiction, we develop an operando activation and regeneration process, where the cathode of a zero-gap electrolyzer cell is periodically infused with alkali cation-containing solutions. This enables deionized water-fed electrolyzers to operate at a CO₂ reduction rate matching that of those using alkaline electrolytes (CO partial current density of 420 ± 50 mA cm^-2 for over 200 hours). We deconvolute the complex effects of activation and validate the concept with five different electrolytes and three different commercial membranes. Finally, we demonstrate the scalability of this approach on a multi-cell electrolyzer stack, with a 100 cm² / cell active area.

Fig. 1

Unintended cation crossover and precipitate formation in alkaline anolyte fed zero-gap CO₂ electrolyzers. (a) Schematic illustration of the operation of an AEM separated zero-gap CO₂ electrolyzer cell with alkaline anolyte. (b) Ion chromatographic quantification of K⁺ and Cs⁺ crossover through different commercially available AEMs during CO₂ electrolysis in a zero-gap electrolyzer cell. (c) Cross-section SEM-EDX and (d) micro-CT images of a GDE after continuous CO₂ electrolysis in a zero-gap cell (T = 50 °C 1 M KOH anolyte, ΔU = 3.0 V). The red and green colors in the SEM-EDX and Micro-CT images represent Ag and K atoms, respectively. Such experiments were repeated on separate cell assemblies independently 3 times, with similar results.

Fig. 2

Schematic piping and instrumentation diagram of the test framework employed. In the inset “1” shows the default positions of the manual valves, forming a continuous gas path to the cell, bypassing the activation loop. Turning the valves into position “2” the gas is driven through the activation loop, carrying the activation fluid into the cell.

Fig. 3

Cathode activation using different commercially available AEMs. (a) Contact angles of different water/isopropanol solvent mixtures on the microporous side of a Sigracet 39BC GDL. (b-d) Chronoamperometric curves and CO-formation partial current densities (T_cathode = 60 °C, 12.5 cm cm⁻² min⁻¹ CO₂ feed rate, pure water anolyte)measured using (b) Sustainion X37-50 (ΔU = 3.1 V), (c) PTFE-reinforced 15 μm thick PiperION TP-85 (ΔU = 3.2 V), (s) Fumasep FAB-PK-130 (ΔU = 3.1 V)AEMs. 10 cm 0.5 M KOH (for (b) and (d)) or 0.5 M CsOH (c) solutions in 1:3 isopropanol/water mixture were used to activate the cathode at the times marked with asterisks in the figures. Such experiments were repeated on separate cell assemblies independently at least 3 times, with similar results. The values in (a) are the mean value of 3 independent measurements, together with the calculated standard deviations.

Fig. 4

Mechanism and reversibility of cathode activation. (a) Chronoamperometric curves and (b) EIS spectra recorded before and after activating the cathode GDE with 10 cm 0.5 M KOH solution in 1:3 isopropanol/water mixture (Sustainion membrane. ΔU = 3.1 V, T_cathode = 60 °C, 12.5 cm cm⁻² min⁻¹ CO₂ feed rate,). (c) Chronoamperometric curve and CO formation partial current density measured under identical conditions as in (a). The cathode GDE was activated at the marked times with 10 cm 0.5 M KOH solution in 1:3 isopropanol/water mixture, while it was rinsed with 10 cm 1:3 isopropanol/water mixture to de-activate it. (d) Time-resolved current density and product stream composition change during, and immediately after activating the electrolyzer cell (with 3 cm 1 M CsOH solution in 1:3 isopropanol/water mixture, ΔU = 3.2 V, T_cathode = 60 °C, 12.5 cm cm⁻² min⁻¹ CO₂, PiperION membrane). In the upper panel, the pressure in the CO₂ inlet pipe during activation is indicated. Such experiments were repeated on separate cell assemblies independently at least 3 times, with similar results.

Fig. 5

Deconvolution of the complex effect of the activating electrolyte. Partial current densities for CO and H₂ production during constant voltage electrolysis with water anolyte, after cathode activation using 10 cm solution (in 1:3 isopropanol/water mixture) of (a) different alkali metal hydroxides (c = 0.5 M) (b) different potassium salts (c(K⁺) = 0.5 M). The cell was operated at ΔU = 3.1 V, T_cathode = 60 °C with 12.5 cm cm⁻² min⁻¹ CO₂ feed rate, using a Sustainion membrane in the cell. The plotted values are the mean from 3 gas composition measurements (GC analysis), together with the calculated standard deviations. All experiments were repeated independently on separate cell assemblies 3 times, with similar results.

Fig. 6

Long-term operation of a CO₂ electrolyzer with water anolyte and periodic activation. (a) Total current density and (b) CO and H₂ partial current densities during constant voltage electrolysis, using a PTFE-reinforced 15 μm thick PiperION TP-85 membrane separated cell (ΔU = 3.2 V, T = 60 °C water anolyte, 12.5 cm cm⁻² min⁻¹ CO₂ feed rate). The cathode was activated with 5 cm 1 M CsOH solution in 1:3 isopropanol/water mixture after every 12 hours of the electrolysis. Long term experiments (over 100 h) were repeated on separate cell assemblies independently 5 times, with similar results.

Fig. 7

Cathode activation experiments in larger electrolyzer cells and stack. Chronoamperometric measurements on an A = 100 cm² single layer electrolyzer cell (T = 60 °C, water anolyte, 12.5 cm cm⁻² min⁻¹ CO₂ feed) at (a) ΔU = 3.3 V and (b) ΔU = 3.5 V. (c) CO partial current densities 10 minutes after performing the cathode activation (20 cm 1 M CsOH solution in 1:3 isopropanol/water mixture) and the ratio of CO formation partial current densities 90 and 10 minutes after performing the cathode activation, recorded at different cell voltages. (d) Chronoamperometric measurement on a A = 100 cm² 3-layer electrolyzer cell stack (T = 60 °C, water anolyte, 12.5 cm cm⁻² min⁻¹ CO₂ feed, cathode activation with 60 cm 1 M CsOH solution in 1:3 isopropanol/water mixture) at ΔU = 9.9 V. Such experiments were repeated on separate cell assemblies independently 3 times, with similar results.