The future of low-temperature carbon dioxide electrolysis depends on solving one basic problem

Carbonate formation is the primary source of energy and carbon losses in low-temperature carbon dioxide electrolysis. Realigning research priorities to address the carbonate problem is essential if this technology is to become a viable option for renewable chemical and fuel production.

The plummeting cost and daily curtailment of renewable electricity have spurred growing interest in using CO₂ electrolysis to produce chemicals and fuels. High-temperature solid oxide cells that convert CO₂ to CO and O₂ have reached nascent commercialization. Low-temperature CO₂ electrolysis is an attractive alternative that offers more convenient and flexible operation and the ability to generate multicarbon products such as ethylene, ethanol, and propanol. Over the past 10 years, a dramatic expansion of research in this area has yielded substantial progress in fundamental understanding and prototype devices. Leveraging insights from fuel cells and membrane water electrolyzers, researchers have developed gas diffusion electrode (GDE) cells demonstrating synthetically relevant CO₂ electrolysis current densities (>100 mA cm⁻²) and promising stability. Despite these advances, the energy efficiency (power-to-product) and carbon efficiency (CO₂-to-product) of low-temperature CO₂ electrolysis remain too low to support large-scale applications^1,2. While much current research is focused on CO₂ reduction catalyst design, the biggest obstacle to improving performance is an often overlooked basic chemistry problem: the rapid and thermodynamically favorable reaction of CO₂ with hydroxide (OH^–) to form carbonate (CO₃^2–) imposes steady state electrolysis conditions that result in large voltage and CO₂ losses. Although recent work has brought attention to some aspects of the CO₃^2– problem^2,3,4, it is far more pernicious than what is widely appreciated. Here we explain how CO₃^2– formation compromises efficiency to highlight the need for new research directions that address this problem.

Hydroxide consumption makes alkaline CO₂ electrolyzers fuel-wasting devices

The state-of-the-art for low-temperature CO₂ electrolysis has been obscured by studies that utilize a reservoir of flowing alkaline electrolyte to maintain a high pH in the cell^5,6,7,8. High pH minimizes the cell voltage, which makes these systems appear to have high energy efficiency, but consumption of OH^– in the reservoir by CO₃^2– formation results in a net negative energy balance. Understanding why the cell voltage is minimized at high pH helps to clarify the CO₃^2– problem (Fig. 1a). For most known catalyst materials, including Au and Cu, the CO₂ reduction rate depends on the electron transfer driving force but not explicitly on pH^{9,10,11,12,13}. As a result, synthetically relevant current densities require rather negative potentials versus an absolute reference such as the standard hydrogen electrode (SHE). Even with high surface area electrodes, CO₂ reduction at a geometric current density of >200 mA cm^–2 has generally required potentials <–1.3 V versus SHE. Because the thermodynamic electrode potentials become more negative versus SHE as the pH is increased, the cathode overpotential is minimized by increasing pH at a fixed cathode potential versus SHE. The overpotential for oxygen evolution at the anode is also generally lowest in base. These contributions significantly reduce the cell voltage at high pH (Fig. 1a).

Fig. 1: The CO₃^2– problem.

a Visualization of various contributors to the cell voltage for a CO₂ electrolysis cell operating under alkaline conditions, carbonated conditions, or with a bipolar membrane (BPM). The contributions shown are the thermodynamic cell potential (E°_cell), and cathode, anode, and BPM overpotentials (η). Electrode potentials are referenced to the SHE scale. For simplicity, cell resistance was not included, which would add to the cell voltage. The green dotted line represents the least negative cathode potential reported for low-temperature CO₂ reduction at >200 mA cm^–2 7. The red dotted lines provide a visual reference for cell voltages. The thermodynamic potentials for the cathode (E°_cathode) and anode (E°_anode) shift positive versus SHE as the pH is decreased. Alkaline conditions minimize cell voltage but cannot be maintained at steady state because of CO₃^2– formation. b Schematic of anion transport in an alkaline flow cell showing OH^– consumption by CO₂. c Schematic of a cell at steady state after carbonation showing the carbon loss due to CO₂ released at the anode. d Schematic of a BPM cell at steady state.

While minimizing cell voltage is obviously desirable, cells operating under flowing alkaline conditions are not at steady state. The OH^– in the electrolyte reservoir is continuously consumed by exergonic CO₃^2– formation as it contacts CO₂ at the cathode (Eq. 1¹⁴ and Fig. 1b):

$$2{{\mathrm{OH}}^-}_{({\mathrm{aq}})} + {\mathrm{CO}}_{2({\mathrm{g}})} \to {{\mathrm{CO}_3}^{2 - }}_{({\mathrm{aq}})} + {\mathrm{H}}_2{\mathrm{O}}_{({\mathrm{l}})}\,{\Delta}G^\circ = -56\,{\mathrm{kJ}}\,{\mathrm{mol}}^{-1}$$

(1)

In practice, the energy required to regenerate CO₂ and 2 OH^– from aqueous CO₃^2– is much larger than |ΔG°|, for example >230 kJ mol^–1 with an optimized system using a calcination cycle¹⁵. Depending on the product, the energy stored by CO₂ electrolysis is ~100–130 kJ mol^–1 of electrons. Thus, the energy balance with flowing alkaline electrolyte will be negative if carbonation consumes ~1 OH^– equivalent for each electron of CO₂ electrolysis current. Though it is rarely quantified, one recent study reported that ~3 OH^– were consumed per electron of CO₂-to-CO electrolysis current with a flowing 1 M KOH electrolyte⁸. Unfortunately, it is common practice to ignore OH^– consumption and calculate an “energy efficiency” based solely on the applied voltage, thermodynamic voltage, and faradaic efficiency. These values are qualitatively incorrect. In reality, flowing alkaline CO₂ electrolysis cells are fuel-wasting devices because it would require more fuel to regenerate the spent electrolyte than the amount of fuel produced by the electrolysis. Using data from flowing alkaline conditions in technoeconomic analyses will grossly overestimate the state-of-the-art for CO₂ electrolysis.

Consequences of operating in carbonated electrolyte

Conversion of power to fuel with a positive energy balance requires operating under steady state conditions where there is no net electrolyte consumption. In a cell operating at steady state, CO₃^2– is still formed continuously at the cathode by reaction of CO₂ with the electrogenerated OH^–, but it is protonated elsewhere in the cell to release CO₂. In an anion-transporting cell such as an anion exchange membrane (AEM) cell, CO₃^2– is transported to the anode (Fig. 1c). In order to protonate CO₃^2–, the anode pH equilibrates to near-neutral (pH ~ 8)², which increases the cell voltage compared to high pH because the anode thermodynamic potential moves in the positive direction (Fig. 1a). In addition, there is a much greater oxygen evolution overpotential at near-neutral pH with available catalysts. Thus, most of the energy penalty from the CO₃^2– problem is a consequence of what happens at the anode. In fact, the same problem is seen in AEM water electrolysis, where the cathode catalyzes H₂ evolution. Contamination of an AEM water electrolyzer with trace amounts of air (400 ppm CO₂) results in the formation of CO₃^2–, which causes a >1 V increase to the cell voltage¹⁶.

The steady state carbon flux imposed by CO₃^2– formation also sets an upper limit on carbon efficiency in anion transporting cells. For CO₂ reduction to CO, one CO₂ is released at the anode for every CO that is produced at the cathode, resulting in a maximum carbon efficiency of 50%. For ethylene, the maximum is 25% (Fig. 1c). Many cells operate at much lower carbon efficiencies than these maxima because a large excess of CO₂ is supplied to the cathode to support high current densities. Low carbon efficiency is a major barrier to large-scale chemical and fuel electrosynthesis because there is a substantial energy and financial cost to obtain a CO₂ feedstock of suitable purity.

An alternative to an anion transporting cell is to operate with a bipolar membrane (BPM) configured such that CO₂ electrolysis is coupled with water dissociation at the BPM. At steady state, CO₃^2– generated at the cathode is protonated at the BPM interface, releasing CO₂ on the cathode side (Fig. 1d). While this design avoids releasing CO₂ at the anode¹⁷, it has a lower energy efficiency than an AEM cell because the BPM imposes an additional overpotential in order to drive water dissociation (Fig. 1a). It is important to note that CO₂ crossover through a BPM is low but non-zero¹⁸. As such, BPM cells that start with high pH on the anode side will transform to a CO₃^2– electrolyte at steady state.

Current state-of-the-art and outlook

The consequences of the CO₃^2– problem are evident in the performance that has been demonstrated at steady state. For the production of CO, the best reported performance is for lab-scale (5 cm²) devices that have been operated at steady state for >4000 h at ~200 mA cm^–2 CO₂-to-CO current density, nicely demonstrating the viability of durable CO electrosynthesis at a reasonable rate. However, the carbon efficiency is 50% and the cell voltage is 3.0 V, corresponding to an energy efficiency of only 43%^1,19. Both of these values are consistent with the analysis in Fig. 1. To make a more reduced product such as ethylene, the CO₃^2– problem is more pronounced. The best reported performance at steady state is 60 h of operation at ~500 mA cm^–2 CO₂-to-ethylene current density, ~2% carbon efficiency, and a cell voltage of 3.9 V, corresponding to an energy efficiency of ~15%²⁰. The evaluation of full cell metrics under steady state conditions remains uncommon in CO₂ electrolysis research, but such experiments are essential for assessing progress and should become standard to demonstrate the impact of a material or design advance.

Low-temperature CO₂ electrolysis will not be competitive with other electrical energy storage or CO₂ conversion technologies without major gains in energy and carbon efficiency. Reducing the overpotential of CO₂ reduction catalysis remains an important objective provided it can be realized under steady state conditions at high current density. Avoiding the losses imposed by CO₃^2– formation demands a much broader research effort that includes strategies to control the formation and clearance of CO₃^2–, creative cell designs, and far greater attention to the anode. The progress on the CO₃^2– problem will determine the trajectory of CO₂ electrolysis in the next 10 years and its impact beyond the laboratory.