Journal Archive

Johnson Matthey Technol. Rev., 2023, 67, (1), 110
doi: 10.1595/205651323X16703459968311

Cathodes for Electrochemical Carbon Dioxide Reduction to Multi-Carbon Products: Part II

Rapid improvement in cathode performance

  • Harry Macpherson$, Toby Hodges, Moyahabo Hellen Chuma
  • Johnson Matthey, Blounts Court, Sonning Common, Reading, RG4 9NH, UK
  • Connor Sherwin
  • University of Southampton, B29 Building, East Highfield Campus, University Road, SO17 1BJ, UK
  • Urša Podbevšek, Katie Rigg
  • Johnson Matthey, Blounts Court, Sonning Common, Reading, RG4 9NH, UK
  • Veronica Celorrio
  • Diamond Light Source, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, UK
  • Andrea Russell
  • University of Southampton, B29 Building, East Highfield Campus, University Road, SO17 1BJ, UK
  • Elena C. Corbos*
  • Johnson Matthey, Blounts Court, Sonning Common, Reading, RG4 9NH, UK
  • Email: crina.corbos@matthey.com
  • $Present address: Deep Science Ventures, 51 Eastcheap, 3rd Floor, London, EC3M 1JP, UK

 

PEER REVIEWED
Submitted 7th July 2022; Revised 26th October 2022; Accepted 27th October 2022; Online 31st October 2022

Article Synopsis

This is Part II of a focused review of recent highlights in the literature in cathode development for low temperature electrochemical carbon dioxide and carbon monoxide reduction to multi-carbon (C2+) products. Part I (1) introduced the role of CO2 reduction in decarbonising the chemical industry and described the catalysts and modelling approaches. Part II describes in situ characterisation to improve the understanding and development of catalysts, the catalyst layer and the gas diffusion layer.

1. In Situ Characterisation

1.1 Characterising Reaction Intermediates

Copper electrodes have been characterised through a range of different in situ and operando techniques to study the CO2 reduction mechanism in real time. Prominent characterisation techniques include surface enhanced infrared adsorption spectroscopy (SEIRAS) and Raman spectroscopy which can both probe the surface intermediates during the reaction. Such characterisation can help improve the understanding and development of catalysts for improved C–C coupling.

The first step in the electrochemical CO2 reduction reaction is the formation of the CO2 radical anion (*), proposed by Hori et al. (2). This is the common intermediate between all reduction paths, beyond which formate or CO form. Chernyshova et al. (3) identified this intermediate through surface enhanced Raman spectroscopy (SERS) of a roughened copper electrode (Table I). They identified the first intermediate with the structure η2(C,O)- (Figure 1) which is stabilised onto the copper surface through the covalency of its two bonds to the copper surface, the charge polarisation of the system, the electrostatic interaction with the hydrated electrolyte cation and the positive charge of the coordinating copper atoms.

Table I

Table of Infrared and Raman Bands and their Assignments

Electrode Electrolyte Band Position, cm–1 Infrared/Raman Assignment Reference
Rough Cu 0.1 M NaHCO3 2500–2070 Raman C≡O Stretch of *CO (3)
280 Cu-CO frustrated rotation
360 Cu-CO translation
1065 v1 (symmetric C-O)
1540 vas (η2(C,O)-CO2–)
350 vCu-C (η2(C,O)-)
Polycrystalline Cu CO sat. 0.05 M Li2CO3 2000–2100 IR C≡O Stretch (COatop) (4)
1800–1900 C≡O Stretch (CObridge)
Cu(100) CO sat. 0.1 M LiOH 1191 IR C-O-H Stretch (OCCOH) (5)
1584 C-O Stretch (OCCOH)
OD-Cu CO sat. 0.05 M KOH 2058 IR C≡O Stretch of *CO (6)

The adsorbed η1(C)- or η2(C,O)- can be converted to CO through protonation to form an adsorbed carboxyl and then reductively dissociates to CO and H2O (2, 3), Equation (i) and (ii):

(i)
(ii)
Fig. 1.

Possible first intermediates of the CO2 reduction reaction. Reprinted from (3)

CO is the important intermediate in the formation of C2+ products. Favourable adsorption of CO onto the copper surface promotes further reduction into C2+ products. The adsorption mode of CO onto the copper is also important, with a top-bound CO promoting further reduction whereas bridge-bound CO is less active (Figure 2) (4). The adsorption mode can be clearly distinguished from the infrared (IR) adsorption spectrum, with the bridge bound adsorption having an adsorption between 1800–1900 cm–1 and linearly bound between 2000–2100 cm–1 (Table I) (4).

Fig. 2.

Bonding modes of carbon monoxide onto copper surfaces. Reprinted with permission from (4), Copyright (2018) American Chemical Society

Through in situ Fourier transform infrared of Cu(100) electrodes coupled with density functional theory calculations, stretching frequencies associated with C–O–H and C–O have been identified during CO reduction. These were attributed to formation of the hydrogenated dimer OC–COH (Table I) (5). This provides evidence for C–C coupling taking place through a reductive dimerisation process early in the reaction. This process was however not found on Cu(111) electrodes under identical conditions, demonstrating the structure sensitivity of C–C coupling and its preference for square symmetry (6).

In contrast to metallic copper, electrodes prepared using copper oxides and derivates of copper oxide are favourable for producing C2+ products. Copper derived from Cu2O has been shown to be highly selective towards C2+ products during CO reduction, with a 57% Faradaic efficiency (FE) at a modest potential of –0.3 V vs. reversible hydrogen electrode (RHE) in CO sat. 0.1 M KOH. Operando investigations of oxide-derived copper (OD-Cu) using attenuated total reflection SEIRAS (ATR-SEIRAS) have shown an adsorption band at 2058 cm–1 (7) similar to that of the linearly bound CO on Cu(100) (4). Thus, the ability of the oxide derived copper to facilitate C–C coupling at low overpotentials has been attributed to preferential exposure of the Cu(100) facet (7).

1.2 Probing Catalyst Changes

In situ techniques have also been used to determine the chemical and structural changes occurring at oxidised copper electrodes during the CO2 reduction reaction. X-ray absorption spectroscopy (XAS) can determine bulk oxidation state changes, X-ray diffraction (XRD) spectroscopy provides information regarding changes to the crystal structure of the catalyst during the reaction, whilst Raman spectroscopy and X-ray photoelectron spectroscopy (XPS) can provide surface specific information.

According to the Pourbaix diagrams (8), in neutral electrolyte at CO2 reduction reactions (CO2RR) potentials copper should exist as copper metal. Therefore, over the course of CO2RR a copper electrode typically reduces to metallic copper, a worse C2+ coupling catalyst than the oxides (9). Time dependent in situ Raman has been employed to look at the surface degradation of Cu2O films during the CO2RR. After holding the electrode at –0.99 V vs. RHE in CO2 saturated KHCO3, after 30 s the vibrations associated with Cu2O were attenuated and two bands associated with intermediately reduced copper oxides appeared. After 200 s no peaks could be seen in the Raman spectrum suggesting it had been fully reduced to copper. Once the cathodic potential was removed, the surface re-oxidised to Cu2O (10).

The Kauffman group (11) used in situ XAS, Raman spectroscopy and XRD to investigate the bulk and surface changes of porous copper oxides during CO2 reduction (Figure 3). At open circuit potential the copper remains in the 2+ oxidation state however when a potential of –0.2 V vs. RHE is applied the presence of metallic copper is detected. At –0.6 V vs. RHE the material had almost completely reduced to Cu0. This was also confirmed through in situ XRD, with the evolution of face-centred cubic copper when the potential was held at –0.6 V vs. RHE. However, once the cathodic potential was removed, the metallic copper was also found to re-oxidise. To investigate the surface changes and possible presence of surface oxides during the reaction (12–14), in situ Raman was also performed at –0.6 V vs. RHE. The Ag feature of CuO was monitored during the reaction, after 30 min this feature had completely gone, indicating the formation of Raman inactive metallic copper.

Fig. 3.

(a) Cyclic voltammetry of CuO-IO electrode in CO2 saturated 0.1 M KHCO3; (b) potential-dependent k2-weighted R-space extended X-ray absorption fine structure analysis (no phase correction) from –0.2 V to –1.2 V vs. RHE (collected at 30 min at each potential); (c) in situ Raman spectra for tracking surface structure of CuO-IO during EC-CO2RR at –0.6 V (using 785 nm laser source); (d) single-crystal X-ray diffraction patterns of CuO-IO electrode under open circuit and steady state at –0.6 V vs. RHE. Used with permission of Royal Society of Chemistry from (11), Copyright (2019); permission conveyed through Copyright Clearance Center, Inc

Copper oxide catalysts have been shown to reduce during the CO2RR, but subsurface oxygen has been suggested as the reason for the improvement of oxide derived catalysts over metallic copper for C–C couplings (15). In situ ambient pressure XPS results has shown the absence of any residual oxide phases during CO2RR, but did show the presence of subsurface oxygen (15, 16). Subsurface oxygen is believed to increase the CO binding energy through reduced sigma repulsion, favouring C–C bond formation through increased CO coverage (15).

Different methods of stabilising oxygenated copper species have been investigated. One promising example introduced nanocavities into the Cu2O catalyst which confines the carbon intermediates formed in situ to remain in the cavity and cover the catalyst surface, stabilising the Cu+ species. Through in situ Raman the characteristic Raman peaks for Cu2O remained after 20 min at –0.61 V vs. RHE (Figure 4). Additional XAS of the catalysts during the 20 min cathodic hold shows 32.1% of the Cu+ species remained after the potential hold (18).

Fig. 4.

(a) Operando Raman spectroscopy of Cu2O during a potential hold of –0.61 V vs. RHE; (b) copper K-edge XAS spectra of Cu2O at different times during a potential hold at –0.61 V vs. RHE. Reprinted with permission from (17), Copyright (2017) American Chemical Society

De Luna et al. (19) have suggested ‘electro re-deposited’ copper catalyst made through partial dissolution of a copper precursor and redeposition of the copper ions as another catalyst capable of staying active for C2+ products for extended periods. They found this material has a C2+ FE of 52% at –1.2 V vs. RHE. In situ XAS during a potential hold at –1.2 V vs. RHE showed a reduction of the Cu+ to metallic copper, from 84% Cu+ at the start of the reaction to 23% after 1 h. However, when a potential of –1.87 V vs. RHE was applied, the Cu+ species was completely reduced to metallic copper.

In situ and operando techniques have been fundamental in understanding the electrochemical CO2 reduction reaction mechanisms. SEIRAS and Raman spectroscopy have proven essential in characterising the initial intermediates forming during the reaction aiding in catalyst design to improve the selectivity and activity towards C2+ products. XAS, XRD, Raman spectroscopy and XPS have all been used to effectively show the changes occurring in the bulk and surface of the catalyst. This has helped understand the stability of copper oxides and how oxide derived copper catalysts can give higher selectivity towards carbon coupled products.

2. Catalyst Layer

The cathode catalyst layer (CL) must balance many competing demands and is one of the most crucial electrolyser components (20). Typically, it is a layer several hundred nanometres thick that at minimum comprises the catalyst, which is often in the form of nanoparticles. Polymers, especially ionomers, are commonly incorporated into the CL, primarily to provide mechanical stability and coherence by coating catalyst particles and adhering them to the substrate and each other. Polymers in the CL are therefore often called ‘binders’. Polymers can, however, also be deposited in layers on top of the catalyst, and carbon nanoparticles are sometimes incorporated in the CL to improve electrical conductivity (21). The substrate onto which the CL is deposited depends on the electrolyser configuration. In a membrane-electrode assembly configuration akin to a polymer electrolyte water electrolyser, the CL may be directly deposited onto an ion exchange membrane. However, in the more commonly used configurations, in which electrolyte flows through a channel between the ion exchange membrane and the cathode, the CL is deposited onto a porous gas diffusion layer (GDL) and this configuration is the focus of this review. CL deposition is commonly achieved by producing an ink, containing binder and catalyst, which is deposited onto the GDL by spray coating, calendaring, drop casting, hot pressing or bar coating. The method of coating affects reproducibility and product selectivity, with more controllable methods such as automatic spray coating producing more consistent and homogeneous layers (20). A different possible approach involves directly coating the catalyst to the GDL by sputtering or physical or chemical vapour deposition.

The number one priority in CL design in CO2 electrolysis to C2+ products is high FE for such products. Achieving this not only requires high activities (i.e. high partial current densities) for C2+ products but also the suppression of the parasitic hydrogen evolution reaction (HER) and pathways to C1 products. Looking at Table II, CO2RR require CO2 and protons, in contrast to HER. Therefore, sufficient mass transport of CO2 to the electrocatalyst surface is required to kinetically favour CO2RR. To maintain sufficient CO2 transport, it is essential the CL does not become fully wetted or ‘flooded’ with electrolyte, and this is a leading cause of cell failure.

Table II

The Electrochemical Half Equations of CO2 to Commonly Reported Products, with their Associated Reduction Potentials vs. Standard Hydrogen Electrode (22)

Half-cell reaction Potential vs. standard hydrogen electrode
CO2(g) + 2H+ + 2e– ↔ CO(g) + H2O(l) –0.106
2CO2(g) + 12H+ + 12e– ↔ C2H4(g) + 4H2O(l) 0.064
2H+ + 2e– ↔ H2(g) 0.000

A high local pH at the catalyst surface also promotes C2+ products by suppressing pathways to C1 products for which proton transfer is the rate limiting step.

Therefore, CLs should ideally be engineered for maximum transport of CO2 but relatively slow transport of OH– produced in the cathode reactions and buffering ions such as HCO3– from the bulk electrolyte.

An important function of the CL is to maintain triple phase boundaries where CO2, protons and electrons can meet. This is challenging because liquid transport is enhanced by hydrophilicity, yet this hinders gas transport, and furthermore hydrophobic materials tend not to be electrically conductive. Therefore, the design of active and selective CLs involves an optimisation of competing material properties.

Maintaining sufficient hydrophobicity in long term operation is made challenging by several phenomena. First, electrowetting of copper CL’s can occur merely as a result of the application of potential (23). Second, liquid products such as formic acid and alcohols decrease surface tension of the electrolyte and can cause flooding at high concentrations (24). Third, the precipitation of hydrophilic solid carbonates and bicarbonates in the CL can also cause flooding and block gas transport (25). A common strategy to mitigate flooding and reduce selectivity for HER is to increase the hydrophobicity of the CL and this is often achieved by incorporating binders such as polytetrafluoroethylene (PTFE). However, the addition of too much PTFE can significantly reduce the conductivity of the CL (26). A 2022 study (27) compared coatings onto copper nanoneedles of polyacrylic acid (PAA), which is hydrophilic; Nafion®, which contains hydrophilic and hydrophobic domains; and fluorinated ethylene propylene (FEP), which is totally hydrophobic and is similar to PTFE in structure.

As shown in Figure 5(a), the contact angle of the electrolyte with the CL was highest for the electrode incorporating FEP, followed by Nafion® and PAA, confirming that the more hydrophobic the binder the more hydrophobic the CL. Using the captive bubble contact angle method, the CO2-phillicity was determined to be highest for the FEP electrode, followed by Nafion® and PAA, confirming that gas accessibility increases with hydrophobicity. Potentiostatic electrolysis was then conducted in a flow cell using GDEs with CLs incorporating the three polymers. As detailed in Figure 5(b), the more hydrophobic the CL, the greater the FE for CO2RR products and the lower the FE for HER, and furthermore the higher the FE for C2+ products. The improved selectivity for CO2RR products is explained by increased concentration of CO2 relative to H2O at the catalyst surface, due to the microhydrophobic environment in the vicinity of the binder, which increases CO2 gas access while limiting H2O access, as illustrated in Figure 5(c) and 5(d). It was further posited that the improved selectivity to C2+ products was a result of improved accessibility of CO at the catalyst surface which further reacts to C2+ compounds.

Fig. 5.

(a) Water and CO2 contact angles for CLs prepared with PAA, Nafion® and FEP; (b) total current density and FEs for H2 and C2+ products as a function of the hydrophilicity, acquired in flow cell at –0.71 V vs. RHE, with aqueous 1.0 m KOH as electrolyte; (c) hydrophilic and (d) hydrophobic schematic illustrations of the local reaction environment near the copper surface in a flow cell. Reprinted from (27) under a Creative Commons Attribution-NonCommercial License

Other recent studies have improved CO2RR selectivity by employing hydrophobic binders in the CL (28–30). Kim et al. (31) measured the water concentration on thin ionomer films deposited onto bare metallic copper, finding that both Nafion® and Sustainion®, a hydrocarbon anion exchange ionomer, reduced water concentration compared to bare copper. Nafion®, with its very hydrophobic perfluorinated backbone, had a lower water concentration than Sustainion®, and Nafion® with an equivalent weight of 1100, denoted ‘Naf1100’ in Figure 6, had a lower concentration than Naf850, which has a higher density of hydrophilic sulfonate groups. This highlights the prospect of varying the equivalent weight of ionomers to modulate hydrophobicity.

Fig. 6.

Measurement of water concentration for ionomer-coated copper at 100% relative humidity; each sample was ion-exchanged using 0.1 M CsHCO3 solution before the measurement. Reprinted from (31), Copyright (2021) Springer Nature

An area which has so far received little attention is the influence of ink composition on hydrophobicity and more generally CL performance. Although the ink solvent is evaporated following, or even during, CL deposition and so does not remain in the CL, it can still have a profound effect on the CL structure. A 2021 study (32) produced inks of a silver catalyst with Aemion CNN-8 anion exchange ionomer using methanol, ethanol and different ratios of water and isopropanol, finding the ink solvent to have a large influence on CL properties and that the CLs exhibiting the highest hydrophobicity and capillarity had the highest FEs for CO2RR. It was also found that an intermediate level of ionomer aggregation was beneficial for CO2RR as this facilitated the formation of ionomer-catalyst aggregates. Too little aggregation resulted in the formation of a less porous CL, whereas too much resulted in regions of concentrated ionomer and exposed catalyst.

Several parameters in CL design affect local pH, one of the simplest being the thickness of the CL, which can be controlled by varying the catalyst loading. A 2021 study (33) found that high catalyst loadings of around 1.0 mg cm–2 increased ethylene selectivity because the thicker CLs act as a barrier for mass transport of buffering HCO3– ions that were present in the electrolyte, thereby increasing local pH in the region of the CL closest to the microporous layer (MPL) of the GDL. Similarly, thick CLs can also increase local pH by inhibiting transport of OH– produced at the cathode (34). Contrariwise, in concentrated alkaline electrolytes, such as 10 M KOH, C2+ selectivity is enhanced by employing CLs as thin as 25 nm (21). This is because in such alkaline electrolytes, CO2 is unable to penetrate far into the CL before being consumed by reaction with OH–, meaning regions of the CL furthest from the MPL receive insufficient mass transport of CO2. However, the rapid consumption of the electrolyte in highly alkaline cells has been recognised as a serious problem for overall energy consumption in recent years, making cells using near-neutral electrolyte more relevant to study (35). In near-neutral buffering electrolytes, high Nafion® loading can also enhance C2+ selectivity by further slowing transport of HCO3– ions, however increasing the loading to 50 wt% reduces selectivity by limiting transport of CO2. Cation exchange ionomers such as Nafion® are proposed to slow transport of OH– and HCO3– via Donnan exclusion (31, 36).

Taking a different approach to increasing local pH, a 2021 paper (37) reported the production of GDEs by co-electrodeposition of copper and polyamines, as shown schematically in Figure 7. Polymers with various degrees of amine functionalisation were deposited from solution along with copper, such that the amines were adhered to the surface of the deposited copper particles. The polymers with the highest density of amine groups were found to give the highest current density and FE for ethylene when tested in an alkaline flow cell, and this was rationalised by their ability to increase pH at the catalyst surface by setting up the following acid-base equilibrium, Equation (iii):

(iii)
Fig. 7.

Schematic illustration of co-electrodeposition of copper and polyamine to a GDL

The presence of amine groups shifts the equilibrium to the right towards greater concentration and therefore higher surface pH. The increase in ethylene selectivity may however also have been due to an increase in surface CO concentration facilitated by the polymer aiding C–C coupling.

A less discussed parameter affecting local CO2 concentration at the catalyst surface is the solubility of CO2 in the binder. For example, the solubility of CO2 in Sustainion® is 20 times higher than in Nafion® (31) due to its imidazolium functional groups (38). As shown in Figure 8, increasing the CO2 concentration in the CL and, in particular, increasing the ratio of CO2 to water, increases the partial current density for CO2 reduction.

Fig. 8.

Estimation of local CO2:H2O on the basis of measured water concentration and corresponding partial current densities at –1.15 V vs. RHE in 0.1 M CsHCO3. Reprinted from (31), Copyright (2021) Springer Nature

For significant progress to be made in sustaining high selectivity, an innovation in the CL will be required which allows it to break out of the current compromise between gas transport, electrolyte species transport and conductivity. One proposal (39) for how to do this is to de-couple hydroxide, CO2 and water transport by making use of binders with separate hydrophilic and hydrophobic domains and potentially by bringing polyfluorosulfonic acid and anion exchange ionomers into close contact with the catalyst.

3. Gas Diffusion Layer

The GDL serves as a mechanical support to the CL to form a GDE. Crucially, the GDL provides a pathway for CO2 gas to reach the CL and thus enables an order of magnitude higher CO2RR current densities (hundreds of mA cm–2) than when gas is fed to the electrode through dissolution and diffusion in bulk electrolyte (tens of mA cm–2) (40, 41). At the same time, the catalyst must be in direct contact with a liquid or solid electrolyte that provides protons and sustains ionic current conduction. The GDE must also sustain electrical contact without significant ohmic loses, either by using an inherently conductive GDL, such as carbon paper, or by containing an additional conductive layer (21).

Since it defines the rate of mass transport to and from the CL, a GDL has a defining role to play in tuning the selectivity of the conversion by defining the concentration of reactants and products in the local catalyst environment. Intentional design of GDL structures and compositions that would provide favourable conditions for selectively producing a desired product has been identified as an important challenge in the development of economically feasible electrolysers yet is not well understood (39–41).

In addition to selectivity tuning, the GDL must also provide sufficient resistance to flooding by simultaneously allowing the presence of water as a proton source in the CL while preventing detrimental permeation of water and liquid products into the layer.

3.1 Selectivity Studies by Model Electrodes

To guide GDE design towards improving selectivity for C2+ products, well-defined structures made of copper have been designed and studied (42, 43). Well-defined geometries of such model electrodes enable insight into the complex interplay between the mass transport and local chemical environment. They offer a rationalised context to the many experimental observations reporting sensitivity of the CO2RR selectivity to the structure of the GDE (41). Pérez-Ramírez et al. designed a microstructure electrode using laser ablation to define uniform and evenly distributed conical cavities in a copper foil (Figure 9(a)) (43). The well-defined and tuneable micrometre-scale geometry enabled quantitative description of the local chemical environment established inside the cavities during the CO2RR reaction. By combining experimental current densities and FEs with the descriptors of the chemical environment inside the pores (average local CO2 concentration and pH), they were able to map selectivity towards selected products versus the local conditions. These descriptors are well-known to strongly influence the selectivity of CO2 reduction (44, 45), however accurately quantifying their contribution has remained elusive.

Fig. 9.

(a) Scanning electron micrographs of the surface and cross-section of the laser-etched copper foil with conical cavities; (b) overlaid selectivity maps plotting the FE towards selected CO2RR in relation to average pH and CO2 concentration as descriptors. Reprinted from (43), Copyright (2020), with permission from Elsevier

Figure 9(b) shows a visualisation of the sensitivity of reaction selectivity on the local conditions. The overlap of ethylene and ethanol production peaks suggests it may not be possible to discriminate between these two products by tuning the local conditions. Another insight can be gained by analysing the two FE peaks for propanol, which suggest two different reaction pathways may be favoured at different local conditions. The intermediate pH and high CO2 concentration are hypothesised to offer the optimal balance between C1 and C2 production, leading to C3 products.

As well as model electrodes, practical GDLs can also be manufactured with well-defined and controlled structural parameters such as pore size and shape. A three-dimensional (3D) printable GDL has been developed with tuneable microporous structure of a fluoropolymer-based GDL, which allowed CO2 permeance through the GDL to be modulated over several orders of magnitude (46). By tuning the structural parameters, an optimised GDL structure was obtained that allowed the FE towards ethylene to be maximised at high current densities. The authors suggested prolonging the residence time of the CO intermediate near the catalyst, achieved with the optimised GDL design, as the main reason for improved performance.

3.2 Composition for Flooding Prevention

While using GDL enables high current densities and allows the conversion selectivity to be modulated, the most pertinent challenge that still requires a breakthrough remains water management (47, 48). Flooding is the key issue preventing electrolyser operation for relevant times (24). Even the currently best-performing systems, which sustain stable operation for over 100 h (21) are still far away from the durability that will eventually be required in practical CO2 electrolysers (>5000 h) (49).

Flooding is governed by surface properties of the GDE components. Most commonly the wettability is managed by the GDL by either using hydrophobic materials such as fluorinated polymers as coatings for the surface of otherwise hydrophilic carbon (50), or as the main GDL component (21, 46). Preventing flooding failure through micro-structuring (optimising pore geometry, size distribution and connectivity) has also been suggested (24, 51).

Polymer-coated carbon-based GDLs have been reported to enable C2+ products with comparatively high FEs (21). The initial hydrophobicity, however, is lost within a few hours under high-current density operation conditions, which leads to flooding. It has been suggested that this is due to the applied potential (50) and degradation of the coating, accelerated under the locally alkaline conditions (47, 50). Another downside of using carbon as the GDL material is that it can catalyse the HER, which decreases the FE towards carbon products (50).

Significantly longer operational times (>100 h) at high current density have been achieved when the carbon GDL was replaced by an expanded PTFE membrane (21). What is more, the selectivity towards C2+ product at relevant current densities (>100 mA cm–2) was improved (21, 52). To introduce sufficient electron conductivity, a layer of sputtered copper catalyst has been additionally coated with a layer of carbon black particles and graphite (21).

The desired improvement in production rates, FEs and durability can, however, make flooding prevention even more challenging by accumulating liquid C2+ products that lower the surface tension of the electrolyte, such as ethanol and 1-propanol. Although concentrating the products in the electrolyser outlet is highly desired from a technoeconomic point of view, producing such mixtures can be detrimental even to the currently top-performing hydrophobic fluoropolymer-based GDLs, as demonstrated by Leonard et al. (24) for water-alcohol mixtures on PTFE surface. They suggested introducing additional surface treatments to the hydrophobic fluoropolymer-based GDLs in order to make them oleophobic and resistant to wetting of product-rich outlets.

In a recent report, Lapkin et al. present a case for deposition of a hydrophilic residue on the electrode surface while producing ethylene on copper (53). They proposed that acrolein, an unsaturated C3 aldehyde not previously reported as a CO2RR product on copper, was formed during the reaction in low concentrations. Under the locally alkaline conditions at the catalyst surface, acrolein was hypothesised to polymerise into hydrophilic oligomers and polymers that were found deposited on the electrode surface in post-mortem analysis. Acrolein thus avoided direct operando detection, but nevertheless promoted flooding by making the catalyst surface hydrophilic.

The pressing challenge for CO2RR to C2+ products is designing a GDL that will provide wetting-resistant functionalities for extended run times in the presence of concentrated products. Additionally, the mass-controlling function of the GDL can in combination with the design of the CL modulate the selectivity of the overall conversion by careful tuning of the local conditions that facilitate formation of the selected C2+ product(s). There is growing need to expand our understanding of this complex but relevant effects (46).

4. Conclusions

While the rate of improvement in cathode performance has been rapid in recent years, achieving prolonged product selectivity will require both greater understanding and ingenuity. Not only must the cathode components be optimised in their own right, but their interactions must be understood and harnessed. While this complexity and interdependence makes establishing solid facts challenging, the potent combination of standard electrochemical testing with in situ characterisation and modelling will likely continue to be instrumental in advancing the field.

References

  1. 1.
    H. Macpherson, T. Hodges, M. H. Chuma, C. Sherwin, U. Podbevšek, K. Rigg, V. Celorrio, A. Russell and E. C. Corbos, Johnson Matthey Technol. Rev., 2023, 67, (1), 97 LINK https://technology.matthey.com/article/67/1/97-109/
  2. 2.
    Y. Hori, ‘Electrochemical CO2 Reduction on Metal Electrodes’, in “Modern Aspects of Electrochemistry 42”, eds. C. G. Vayenas, R. E. White and M. E. Gamboa-Aldeco, Ch. 3, Springer-Verlag, New York, USA, 2008, pp. 89–189 LINK https://doi.org/10.1007/978-0-387-49489-0_3
  3. 3.
    I. V. Chernyshova, P. Somasundaran and S. Ponnurangam, Proc. Natl. Acad. Sci. USA, 2018, 115, (40), E 9261 LINK https://doi.org/10.1073/pnas.1802256115
  4. 4.
    C. M. Gunathunge, V. J. Ovalle, Y. Li, M. J. Janik and M. M. Waegele, ACS Catal., 2018, 8, (8), 7507 LINK https://doi.org/10.1021/acscatal.8b01552
  5. 5.
    E. Pérez-Gallent, M. C. Figueiredo, F. Calle-Vallejo and M. T. M. Koper, Angew. Chem. Int. Ed., 2017, 56, (13), 3621 LINK https://doi.org/10.1002/anie.201700580
  6. 6.
    K. J. P. Schouten, E. Pérez Gallent and M. T. M. Koper, ACS Catal., 2013, 3, (6), 1292 LINK https://doi.org/10.1021/cs4002404
  7. 7.
    A. S. Malkani, M. Dunwell and B. Xu, ACS Catal., 2019, 9, (1), 474 LINK https://doi.org/10.1021/acscatal.8b04269
  8. 8.
    B. Beverskog and I. Puigdomenech, J. Electrochem. Soc., 1997, 144, (10), 3476 LINK https://doi.org/10.1149/1.1838036
  9. 9.
    S. Nitopi, E. Bertheussen, S. B. Scott, X. Liu, A. K. Engstfeld, S. Horch, B. Seger, I. E. L. Stephens, K. Chan, C. Hahn, J. K. Nørskov, T. F. Jaramillo and I. Chorkendorff, Chem. Rev., 2019, 119, (12), 7610 LINK https://doi.org/10.1021/acs.chemrev.8b00705
  10. 10.
    D. Ren, Y. Deng, A. D. Handoko, C. S. Chen, S. Malkhandi and B. S. Yeo, ACS Catal., 2015, 5, (5), 2814 LINK https://doi.org/10.1021/cs502128q
  11. 11.
    T.-D. Nguyen-Phan, C. Wang, C. M. Marin, Y. Zhou, E. Stavitski, E. J. Popczun, Y. Yu, W. Xu, B. H. Howard, M. Y. Stuckman, I. Waluyo, P. R. Ohodnicki and D. R. Kauffman, J. Mater. Chem. A, 2019, 7, (48), 27576 LINK https://doi.org/10.1039/c9ta10135g
  12. 12.
    L. Mandal, K. R. Yang, M. R. Motapothula, D. Ren, P. Lobaccaro, A. Patra, M. Sherburne, V. S. Batista, B. S. Yeo, J. W. Ager, J. Martin and T. Venkatesan, ACS Appl. Mater. Interfaces, 2018, 10, (10), 8574 LINK https://doi.org/10.1021/acsami.7b15418
  13. 13.
    H. Mistry, A. S. Varela, C. S. Bonifacio, I. Zegkinoglou, I. Sinev, Y.-W. Choi, K. Kisslinger, E. A. Stach, J. C. Yang, P. Strasser and B. R. Cuenya, Nat. Commun., 2016, 7, 12123 LINK https://doi.org/10.1038/ncomms12123
  14. 14.
    A. Eilert, F. S. Roberts, D. Friebel and A. Nilsson, J. Phys. Chem. Lett., 2016, 7, (8), 1466 LINK https://doi.org/10.1021/acs.jpclett.6b00367
  15. 15.
    A. Eilert, F. Cavalca, F. S. Roberts, J. Osterwalder, C. Liu, M. Favaro, E. J. Crumlin, H. Ogasawara, D. Friebel, L. G. M. Pettersson and A. Nilsson, J. Phys. Chem. Lett., 2017, 8, (1), 285, LINK https://doi.org/10.1021/acs.jpclett.6b02273
  16. 16.
    D. Gao, I. Zegkinoglou, N. J. Divins, F. Scholten, I. Sinev, P. Grosse and B. Roldan Cuenya, ACS Nano, 2017, 11, (5), 4825 LINK https://doi.org/10.1021/acsnano.7b01257
  17. 17.
    S. Ma, M. Sadakiyo, M. Heima, R. Luo, R. T. Haasch, J. I. Gold, M. Yamauchi and P. J. A. Kenis, J. Am. Chem. Soc., 2017, 139, (1), 47 LINK https://doi.org/10.1021/jacs.6b10740
  18. 18.
    P.-P. Yang, X.-L. Zhang, F.-Y. Gao, Y.-R. Zheng, Z.-Z. Niu, X. Yu, R. Liu, Z.-Z. Wu, S. Qin, L.-P. Chi, Y. Duan, T. Ma, X.-S. Zheng, J.-F. Zhu, H.-J. Wang, M.-R. Gao and S.-H. Yu, J. Am. Chem. Soc., 2020, 142, (13), 6400 LINK https://doi.org/10.1021/jacs.0c01699
  19. 19.
    P. De Luna, R. Quintero-Bermudez, C.-T. Dinh, M. B. Ross, O. S. Bushuyev, P. Todorović, T. Regier, S. O. Kelley, P. Yang and E. H. Sargent, Nat. Catal., 2018, 1, (2), 103 LINK https://doi.org/10.1038/s41929-017-0018-9
  20. 20.
    H.-R. “M.” Jhong, F. R. Brushett and P. J. A. Kenis, Adv. Energy Mater., 2013, 3, (5), 589 LINK https://doi.org/10.1002/aenm.201200759
  21. 21.
    C.-T. Dinh, T. Burdyny, M. G. Kibria, A. Seifitokaldani, C. M. Gabardo, F. P. García de Arquer, A. Kiani, J. P. Edwards, P. De Luna, O. S. Bushuyev, C. Zou, R. Quintero-Bermudez, Y. Pang, D. Sinton and E. H. Sargent, Science, 2018, 360, (6390), 783 LINK https://doi.org/10.1126/science.aas9100
  22. 22.
    M. Jouny, W. Luc and F. Jiao, Ind. Eng. Chem. Res., 2018, 57, (6), 2165 LINK https://doi.org/10.1021/acs.iecr.7b03514
  23. 23.
    K. Liu, W. A. Smith and T. Burdyny, ACS Energy Lett., 2019, 4, (3), 639 LINK https://doi.org/10.1021/acsenergylett.9b00137
  24. 24.
    M. E. Leonard, M. J. Orella, N. Aiello, Y. Román-Leshkov, A. Forner-Cuenca and F. R. Brushett, J. Electrochem. Soc., 2020, 167, (12), 124521 LINK https://doi.org/10.1149/1945-7111/abaa1a
  25. 25.
    D. G. Wheeler, B. A. W. Mowbray, A. Reyes, F. Habibzadeh, J. He and C. P. Berlinguette, Energy Environ. Sci., 2020, 13, (12), 5126 LINK https://doi.org/10.1039/d0ee02219e
  26. 26.
    K. J. Puring, D. Siegmund, J. Timm, F. Möllenbruck, S. Schemme, R. Marschall and U.-P. Apfel, Adv. Sustain. Syst., 2021, 5, (1), 2000088 LINK https://doi.org/10.1002/adsu.202000088
  27. 27.
    T. H. M. Pham, J. Zhang, M. Li, T.-H. Shen, Y. Ko, V. Tileli, W. Luo and A. Züttel, Adv. Energy Mater., 2022, 12, (9), 2103663 LINK https://doi.org/10.1002/aenm.202103663
  28. 28.
    P. An, L. Wei, H. Li, B. Yang, K. Liu, J. Fu, H. Li, H. Liu, J. Hu, Y.-R. Lu, H. Pan, T.-S. Chan, N. Zhang and M. Liu, J. Mater. Chem. A, 2020, 8, (31), 15936 LINK https://doi.org/10.1039/d0ta03645e
  29. 29.
    M. Wang, L. Wan and J. Luo, Nanoscale, 2021, 13, (6), 3588 LINK https://doi.org/10.1039/d0nr08369k
  30. 30.
    Q. Chang, J. H. Lee, Y. Liu, Z. Xie, S. Hwang, N. S. Marinkovic, A.-H. A. Park, S. Kattel and J. G. Chen, JACS Au, 2021, 2, (1), 214 LINK https://doi.org/10.1021/jacsau.1c00487
  31. 31.
    C. Kim, J. C. Bui, X. Luo, J. K. Cooper, A. Kusoglu, A. Z. Weber and A. T. Bell, Nat. Energy, 2021, 6, (11), 1026 LINK https://doi.org/10.1038/s41560-021-00920-8
  32. 32.
    B. A. W. Mowbray, D. J. Dvorak, N. Taherimakhsousi and C. P. Berlinguette, Energy Fuels, 2021, 35, (23), 19178 LINK https://doi.org/10.1021/acs.energyfuels.1c01731
  33. 33.
    T. Möller, T. N. Thanh, X. Wang, W. Ju, Z. Jovanov and P. Strasser, Energy Environ. Sci., 2021, 14, (11), 5995 LINK https://doi.org/10.1039/d1ee01696b
  34. 34.
    S. Suter and S. Haussener, Energy Environ. Sci., 2019, 12, (5), 1668 LINK https://doi.org/10.1039/c9ee00656g
  35. 35.
    J. Sisler, S. Khan, A. H. Ip, M. W. Schreiber, S. A. Jaffer, E. R. Bobicki, C.-T. Dinh and E. H. Sargent, ACS Energy Lett., 2021, 6, (3), 997 LINK https://doi.org/10.1021/acsenergylett.0c02633
  36. 36.
    D. A. Vermaas, S. Wiegman, T. Nagaki and W. A. Smith, Sustain. Energy Fuels, 2018, 2, (9), 2006 LINK https://doi.org/10.1039/c8se00118a
  37. 37.
    X. Chen, J. Chen, N. M. Alghoraibi, D. A. Henckel, R. Zhang, U. O. Nwabara, K. E. Madsen, P. J. A. Kenis, S. C. Zimmerman and A. A. Gewirth, Nat. Catal., 2021, 4, (1), 20 LINK https://doi.org/10.1038/s41929-020-00547-0
  38. 38.
    S. Zulfiqar, M. I. Sarwar and D. Mecerreyes, Polym. Chem., 2015, 6, (36), 6435 LINK https://doi.org/10.1039/c5py00842e
  39. 39.
    E. W. Lees, B. A. W. Mowbray, F. G. L. Parlane and C. P. Berlinguette, Nat. Rev. Mater., 2022, 7, (1), 55 LINK https://doi.org/10.1038/s41578-021-00356-2
  40. 40.
    D. Wakerley, S. Lamaison, J. Wicks, A. Clemens, J. Feaster, D. Corral, S. A. Jaffer, A. Sarkar, M. Fontecave, E. B. Duoss, S. Baker, E. H. Sargent, T. F. Jaramillo and C. Hahn, Nat. Energy, 2022, 7, (2), 130 LINK https://doi.org/10.1038/s41560-021-00973-9
  41. 41.
    D. Higgins, C. Hahn, C. Xiang, T. F. Jaramillo and A. Z. Weber, ACS Energy Lett., 2019, 4, (1), 317 LINK https://doi.org/10.1021/acsenergylett.8b02035
  42. 42.
    K. D. Yang, W. R. Ko, J. H. Lee, S. J. Kim, H. Lee, M. H. Lee and K. T. Nam, Angew. Chem. Int. Ed., 2017, 56, (3), 796 LINK https://doi.org/10.1002/anie.201610432
  43. 43.
    F. L. P. Veenstra, N. Ackerl, A. J. Martín and J. Pérez-Ramírez, Chem, 2020, 6, (7), 1707 LINK https://doi.org/10.1016/j.chempr.2020.04.001
  44. 44.
    Y. C. Tan, K. B. Lee, H. Song and J. Oh, Joule, 2020, 4, (5), 1104 LINK https://doi.org/10.1016/j.joule.2020.03.013
  45. 45.
    A. S. Varela, Curr. Opin. Green Sustain. Chem., 2020, 26, 100371 LINK https://doi.org/10.1016/j.cogsc.2020.100371
  46. 46.
    J. Wicks, M. L. Jue, V. A. Beck, J. S. Oakdale, N. A. Dudukovic, A. L. Clemens, S. Liang, M. E. Ellis, G. Lee, S. E. Baker, E. B. Duoss and E. H. Sargent, Adv. Mater., 2021, 33, (7), 2003855 LINK https://doi.org/10.1002/adma.202003855
  47. 47.
    M. E. Leonard, L. E. Clarke, A. Forner-Cuenca, S. M. Brown and F. R. Brushett, ChemSusChem, 2020, 13, (2), 400 LINK https://doi.org/10.1002/cssc.201902547
  48. 48.
    M. Li, M. N. Idros, Y. Wu, T. Burdyny, S. Garg, X. S. Zhao, G. Wang and T. E. Rufford, J. Mater. Chem. A, 2021, 9, (35), 19369 LINK https://doi.org/10.1039/d1ta03636j
  49. 49.
    A. J. Martín, G. O. Larrazábal and J. Pérez-Ramírez, Green Chem., 2015, 17, (12), 5114 LINK https://doi.org/10.1039/c5gc01893e
  50. 50.
    K. Yang, R. Kas, W. A. Smith and T. Burdyny, ACS Energy Lett., 2021, 6, (1), 33 LINK https://doi.org/10.1021/acsenergylett.0c02184
  51. 51.
    Z.-Z. Niu, F.-Y. Gao, X.-L. Zhang, P.-P. Yang, R. Liu, L.-P. Chi, Z.-Z. Wu, S. Qin, X. Yu and M.-R. Gao, J. Am. Chem. Soc., 2021, 143, (21), 8011 LINK https://doi.org/10.1021/jacs.1c01190
  52. 52.
    C. M. Gabardo, C. P. O’Brien, J. P. Edwards, C. McCallum, Y. Xu, C.-T. Dinh, J. Li, E. H. Sargent and D. Sinton, Joule, 2019, 3, (11), 2777 LINK https://doi.org/10.1016/j.joule.2019.07.021
  53. 53.
    M. K. Kovalev, H. Ren, M. Z. Muhamad, J. W. Ager and A. A. Lapkin, ACS Energy Lett., 2022, 7, (2), 599 LINK https://doi.org/10.1021/acsenergylett.1c02450

Author Contributions

Harry Macpherson wrote the ‘Catalyst Layer’ section and created the layout of the review; Toby Hodges wrote the ‘Catalysts’ section; Moyahabo H. Chuma wrote the ‘Modelling’ section; Connor Sherwin wrote the ‘In Situ Characterisation’ section; Urša Podbevšek wrote the ‘Gas Diffusion Layer’ section; Katie Rigg, Veronica Celorrio and Andrea Russell contributed to the ‘In Situ Characterisation’ section; Elena C. Corbos proofread the manuscript and gave overall advice.

Acknowledgements

We acknowledge European Commission Horizon 2020 research and innovation program for funding Renewable Electricity-based, Cyclic and Economic Production of Fuel (EcoFuel) project under grant agreement 101006701 and Heterogenous Photo(electro)catalysis in Flow using Concentrated Light: modular integrated designs for the production of useful chemicals (FlowPhotoChem) under grant agreement No 862453. The material presented and views expressed here are the responsibility of the authors only. The EU Commission takes no responsibility for any use made of the information set out.

Connor Sherwin would like to acknowledge Chris Zalitis (Johnson Matthey, UK) for his support and guidance, and fruitful discussions.

The Authors

Harry Macpherson completed his Master’s in the Carbon Nanomaterials group at the University of Oxford, UK, before developing technology in noble metal alloy processing, lithium-ion battery recycling and CO2 electrolysis at Johnson Matthey, UK. He is now researching novel approaches to the direct air capture of CO2 with Deep Science Ventures, UK.

Toby Hodges graduated from the University of York, UK, in 2019 with an MChem in Chemistry, having completed a research project in photochemical CO2 reduction at the Tokyo Institute of Technology, Japan. His current work as a research scientist at Johnson Matthey focuses on catalyst synthesis for CO2 electrochemical reduction and electrolyser technologies.

Moyahabo H. Chuma is a research scientist in core capabilities group at Johnson Matthey. Her expertise is in the computational simulation of materials. By making use of modelling techniques, she develops atomic-scale models to investigate physical properties and study chemical reactions. She is interested in heterogeneous catalysis, materials and surface science.

Connor Sherwin is a PhD student in Professor Andrea Russell’s group at the University of Southampton, UK. He is currently researching the operando X-ray characterisation of gas evolving and consuming electrocatalysts in an industrial collaboration from Johnson Matthey and Diamond Light Source, UK.

Urša Podbevšek is an electrochemist working on development of low-temperature CO2 reduction systems. She obtained her PhD in Chemical Sciences from University of Ljubljana, Slovenia, focusing on the development of carbon-based electrocatalysts. Her interests include high temperature CO2 electrolysis, advanced coupled electrochemical techniques and catalyst stability studies.

Katie Rigg graduated from the University of Leeds, UK, in 2018 with a MNatSci in Maths and Chemistry. She is currently working as a researcher at Johnson Matthey specialising in catalyst coated membrane testing for proton exchange membrane water electrolysis, and has also contributed to research in oxygen evolution reaction and CO2RR catalysis.

Veronica Celorrio is a Beamline Scientist on B18, the core X-ray absorption spectroscopy beamline at Diamond Light Source. Her research interests are the characterisation and understanding of materials with application in electrochemical sciences. She is highly enthusiastic about the opportunities that this research field provides, especially at the interface between materials chemistry and X-ray techniques.

Andrea E. Russell is Professor of Physical Electrochemistry in the School of Chemistry at the University of Southampton. Her research focuses on the discovery and characterisation of electrocatalysts for sustainable energy and chemical production, with emphasis on the use of spectroscopic tools to characterise the catalysts under operando conditions. She has collaborated with Johnson Matthey for over 25 years and was a Royal Society Industry Fellow from 2004–2006 at Johnson Matthey.

Elena C. Corbos is leading the Electrochemical transformations team at Johnson Matthey. She joined Johnson Matthey in 2010 as a Marie Curie fellow, and has since held several research positions. Her research interests are in materials synthesis with application in sustainable chemicals and fuels production and the development of sustainable technologies to achieve net zero emissions.

Read more of this issue

Related articles

Guest Editorial: Industrial Biotechnology

View
Download

Microbial Production of Hydrogen

View
Download
MORE ARTICLES

Find an article

ArticleSearch