1. Introduction
The UN Climate Conference on COP26 is taking place in Glasgow, Scotland, from 31 October to 12 November 2021, at which countries must agree on action to curb global warming [
1]. The leaders of almost 200 countries approved agreement, committing to fight climate change by cutting CO
2 emissions to slow down the increase in its concentrations in the atmosphere, which has now reached 415 ppm [
2].
Since CO
2 is the final product of combustion of fossil fuels, its capture at the source (i.e., in exhaust sources), which is more implementable than directly capturing from the atmosphere and subsequent conversion using green electricity into useful products, have become an area of interest for researchers worldwide [
3]. There are numerous catalyst-activated routes (thermal, photo, electrochemical and biologic) to convert anthropogenic CO
2 into fuels and other chemicals employing renewable energy selected in number of projects [
4,
5].
Electrocatalytic CO
2 reduction in electrochemical electrolysis cell is highly promising since the reduction process can be carried out at ambient conditions with green electricity and without the need of external hydrogen (since hydrogen is released at the cathode during the water electrolysis) [
6]. Catalysts are needed to improve not only the selectivity but also the yield of all of these reactions, which involve transfer of two or more electrons and protons in multiple steps on three phase contact points—a catalyst surface, gas (CO
2), and electrolyte. There are many catalyst materials screened up to now: Au, Ag, Zn primarily catalyse the formation of CO; Bi, Sn, In, Pb catalyse formic acid, and only Cu, copper oxides, and hydroxides drive the formation of C-C coupling and formation of multi-carbon molecules [
6,
7,
8,
9].
Copper (Cu) is identified as catalyst material that can produce different hydrocarbons and oxygenates carbon monoxide (CO), formic acid (HCOOH), methanol (CH
3OH), methane (CH
4), ethanol (C
2H
5OH), and ethylene (C
2H
4) [
6,
7,
8,
9]. Typical cathode reactions start with hydrogen evolution reaction (HER) and follows through CO
2 reduction reaction (CO2RR) [
7]:
These half-reaction standard potential values are very close, and thus slight changes in the environment (such as local pH, morphology, crystalline facets, doping and etc.) will determine the selectivity of reaction products. To ensure the desired product, the charge transfer must be led. However, this has proven to be a challenging problem.
In case of ethylene, a multiple-step 12 electron and proton transfer reaction (7) on copper-based electrodes proceed at fairly high Faradaic efficiencies (FE) but with low currents. FE in the presence of two or more concurrent reactions indicates the percentage of electrons (charge) that contributes to the formation of a particular product.
More efforts have to be devoted to developing an electrode with a 3D structure and mixed hydrophobic/hydrophilic surfaces for powerful catalysts for efficient CO
2 reduction to higher hydrocarbons. Single crystal metal surfaces or foil electrodes are incapable of this, due to the mass transport limitation of CO
2, due to the poor solubility of CO
2 in aqueous electrolytes (33 mM at 298 K and 1 atm.). The mass transport constraint can be managed by using gas diffusion electrodes (GDE). GDEs in electrocatalytic electrolysis cell provide environment in which a solid catalyst promotes the electrochemical reaction between the liquid and gaseous phases [
7].
Another challenge is to keep the system stable over a long period of time—the falling-off and corrosion of the catalyst coating must be prevented. To this end, researchers propose different solutions, for example B-doped Cu and integrated sacrificial zinc anode [
10]. Despite some more efforts on oxygen plasma-activated Cu [
11], electro-redeposited Cu [
12] and a copper catalyst modified with a polymer [
13] the stabilisation issue of Cu under CO2RR condition remains challenging. In cases where the required CO
2 reforming products are gaseous (CO, C
2H
4), the membrane-electrode system helps to hold the catalyst in place on the electrode surface—the catalyst (GSS/Cu) would be enclosed at the interface between the gas diffusion electrode and the proton conducting membrane [
14]. No catholyte is needed; to power carbon dioxide electroreduction process, moist CO
2 gas has to flow onto the cathode side in the proton exchange membrane cell. As concluded by the authors of [
15], much effort has been devoted to achieving a stable CO
2 electroreduction performance by catalyst electrode design. The biggest issue is low productivity and difficulty in product separation in low current regime and catalyst lose at higher currents. Through the introduction of zero-gap electrolyser, the amount of electrolyte reduces and accelerates the CO
2 conversion reaction rate.
The graphene/nanoparticle hybrid structures exhibit additional advantageous and often synergistic properties that greatly augment their potential for catalysis applications. There are number of studies about modification of the graphene oxide surface by copper nanoparticles to develop efficient electrocatalysts for CO
2 reduction to liquid multi-carbon products (see as example [
16,
17,
18]). N-doped graphene supported Cu nanoparticle catalysts [
16] prepared by simple wet chemical method shows high selectivity for ethanol with FE of 25.72% at −1.0 V (vs. RHE). The defective vertical graphene and stabilized Cu/Cu
xO enhances CO
2 adsorption and promotes electron transfer to the adsorbed CO
2 and intermediates on the catalyst surface, thus improving the overall CO
2 reduction performance [
17]. The Cu-terminated armchair graphene nanoribbons were more efficient catalysts for producing methanol from CO
2, offering the advantages of a lower over-potential and higher selectivity than bulk Cu and other graphene-supported Cu structures [
18]. Only a few studies can be found on the use of graphene/copper heterostructure for the synthesis of ethylene in the process of CO
2 electro-reduction. For example, the authors of [
19] claimed that the manifestation of several parameters including oxide contents in an electrocatalyst, Raman peaks I
D/I
G ratio due to GO support and mode of fabrication of working electrodes along with porous microstructures of the nanocomposite contributes crucially towards CO
2 electro-reduction to ethylene.
Carbon-based materials are potentially interesting catalyst carriers for the CO
2 reduction reaction due to their low cost and especially due to their ability to form a wide range of hybrid nanostructures [
20,
21]. They are chemically inactive at negative potential ranges and present high overpotentials for the hydrogen evolution reaction compared to metal surfaces [
21]. Pristine graphene does not exhibit any activity. However, introducing dopants and defects during the synthesis tailors the electronic structure and catalytic properties of nanostructured carbon materials. In particular, N-doping has been shown to significantly enhance the CO
2 reduction activity [
21].
These preceding studies have led us to elaborate a theoretical model for stable C
2H
4-selective electrocatalyst based on copper nanocluster decorated graphene and understand how this selectivity came to fruition [
22]. In this paper, we report the design of electrocatalytic electrode from graphene sheet stacks (GSS) coated with copper nanocrystals to follow our theoretical predictions [
22,
23]. The novelty of this study is in obtaining stacks of graphene sheets (GSS) using high-frequency modulated DC pulses for electrochemical exfoliation. The exfoliation process is also useful to dope the resulting material with selected atoms, such as nitrogen, though, a suitable electrolyte must be found. Industrial graphite waste is taken as the starting material. Two methods have been tested for the application of the copper catalyst-chemical and electrochemical deposition. It has been found that the size of the copper crystals and the uniformity of the coating can be controlled by the length and amplitude of the current pulses, which prefers the electrochemical deposition method to obtain good catalyst electrodes. Carbon paper and a gas diffusion layer (GDL) were used as substrates to design the electrocatalyst cathode. A gas diffusion layer electrode is shown as best to perform production of ethylene with satisfying FE at high currents.
3. Results and Discussion
As it is shown from theoretical model developed by one from authors of this article Sergei Piskunov [
22,
23], the Cu7 cluster is quite strongly physically adsorbed to the graphene layer with binding energy of −1.54 eV/Cu atom. The negative binding energy means energy is released after the substrate-adsorbate coupling. Single Cu atoms tend to adsorb at the hollow sites of graphene with the binding energy of −2.65 eV/Cu atom. Thus, Cu atoms deposited at graphene could reproduce the facets of the most stable Cu (111) surface [
22]. Nevertheless, Cu deposited at graphene forms quite weak Cu-C graphene bonds, with a bond population of 80 millielectrons. The presence of the nitrogen atoms at graphene support allows for strong chemisorption of Cu atom with the bond population of Cu-N = 303 millielectrons and N-C graphene = 344 millielectrons [
23].
The SEM micrographs of exfoliated GSS powder is shown below in
Figure 1. Clear GSS edges are visible. As it is seen in SEM image (
Figure 2), due to chemical deposition process the Cu coating can be rare, with chaotically scattered copper nanocrystals of various sizes (range from 20 to 200 nm).
The concentration and temperature of the reaction solutions should be optimized. Perhaps exfoliated graphene is too reduced and has insufficient amount of defect sites with oxygen/nitrogen atoms attached to its surface which acts as seed forming centres. The following experiments are performed to optimize the material, and mainly Electrochemical deposition pulse method can be noted as a solution to the concerns raised above.
Integration of rGSS with carbon paper/gas diffusion electrode is performed by electrophoretic deposition of rGSS–SEM micrographs clearly shows GSS flakes on the substrate (
Figure 2a).
Electrochemical copper catalyst deposition on carbon paper and GDL electrodes with GSS flakes gives coatings with different structures (needles, wires, flowers, cubes). From SEM and XRD results it has been concluded that copper creates dendrite and needle such as Cu/Cu
2O structures when high current density (up to 100 mA/cm
2) is applied, on the other hand micron dimension grains with developed crystalline Cu/Cu
2O structures (
Figure 3 and
Figure 4) appear when low current densities (0.1–0.01 mA/cm
2) with shorter current pulses and longer growth times are applied. Upon long exposition, the crystals grow in size until dense coverage is reached (as shown in
Figure 3). It is important to note that in this process the copper coating is formed as the monovalent copper oxide Cu
2O, which at negative potential is reduced to copper without changing morphology (SEM (
Figure 3) and XRD (
Figure 4) images before and after 60 min of electrolysis).
The best electrodes with higher stability in 1 h tests and higher FE for ethylene production in our experiments were Freudenberg GDL/N-GSS with Cu/Cu
2O coating (2–3 microns thick) obtained in pulse electrodeposition process (0.3 mA/cm
2 current pulses for 7 ms in length and separated by 400 ms). As it is seen in
Figure 4, after prolonged electrolysis at higher currents XRD diffractograms of electrochemically deposited Cu coatings on GDL/GSS is with smaller coper oxide peaks. Some etching of copper crystals can be seen after prolonged CO
2 reduction reaction, SEM pictures before and after process are shown in
Figure 3 (up) and (below) accordingly.
Electrocatalytic tests were carried out for three different electrodes:
(a) with 2D electrodes (carbon paper/GSS/Cu). Before starting experiment an electrolyte 0.5 M KHCO
3 was saturated with CO
2 by purging it for 20 min. The time of electrolysis experiments (2–3 V, 6–60 mA/cm
2) was 1 h, and gas samples were collected in syringe (volume 12 cm
3) from catholyte chamber. Low concentrations of ethylene were detected (<0.5 vol%). As it was found in the mass spectra, about 60% was carbon dioxide, about 15% -air. The rest 25% of the gases selected for analysis were H
2, CO, C
2H
4, HCOOH. (b) with 3D electrodes (GDL/N-GSS/Cu). Experiment was performed at fixed current of 150 mA/cm
2 in 1.0 M KHCO
3 electrolyte and CO
2 was introduced to cathode through the non-electrolyte side of cathode at flow rate of 20 cm
3/min. Gas collection was carried out using Tedlar gas bag. MS analyse showed the presence of ethylene in concentration 2.8 vol% (FE~27% for both C
2H
4 and H
2—see
Figure S4).
(c) with 2D electrodes (N2H4 assisted chemical Cu deposition on carbon paper/GSS). Before starting experiments an electrolyte 0.5 M KHCO3 was saturated with CO2 by purging it for 20 min. The time of electrolysis experiments (2–3 V, 6–60 mA/cm2) was 1 h, and gas samples were collected in syringe (volume 12 cm3) from catholyte chamber.
Faraday efficiency describes the efficiency with which charge (electrons) is transferred in a system facilitating electrochemical reactions on both-cathode and anode. In case of CO
2 reduction in electrocatalysis on cathode, several gaseous and liquid products may be formed (gaseous: H
2, CO, CH
4, C
2H
4, C
2H
6; liquid: formic acid, methyl and ethyl alcohols and more). Dissolved products, especially lower hydrocarbons, are usually volatile and may be present in the collected gas sample. It is more difficult to estimate their amount, but if their presence in the gas is not noticeable, then it is assumed that they do not occur. The analysis of the mass spectrometer shows that the gas samples collected at our cathode for our synthesized electrodes contain 30–60% of overflowed CO
2, 30–50% of H
2 generated by electrolysis and the rest of the volume is occupied by CO, HCOOH and C
2H
4. By determining the partial volumes of the sample for the gases produced by the electrocatalysis process and assuming it as 100%, the Faraday efficiencies (FE) for the particular reactions can be calculated (see
Table 1).
It can be seen from
Table 1 that the highest FE of C
2H
4 in the tested samples is for the GDL/N-GSS/Cu sample, where on the 3D gas diffusion layer/nitrogen-doped GSS substrate a fine-grained copper electrode was obtained in the electrochemical deposition process. Compared to the two other samples tested which are on carbon paper, it can be concluded that both the higher currents and the higher FE are given to the substrate of the gas diffusion layer for ethylene and the lowest FE for hydrogen. It also plays a role of N-doped GSS, since as our calculations show, the composite graphene/Cu structure functions as a CO
2 and proton absorber, facilitating hydrogenation and carbon–carbon coupling reactions through carbene mechanism (CH
2 + CH
2 [
23]) on graphene/Cu-nanocluster for the formation of C
2H
4. The authors are convinced that carbon, which is in close contact with copper catalyst, plays an important role in the electrocatalysis of CO
2 for higher hydrocarbons, serving as proton absorber.
Research work on GDL/N-GSS/Cu electrode (b) will be continued. In our experiments of the CO2 electrocatalysis process, N-GSS flakes are the first to delaminate from the electrode surface. One should think of strengthening N-GSS flakes in the carbon fabric with additional binder. Therefore, further research should focus on the development of an ink-based N-GSS slurry on an electrically conductive carbon fabric to form 3D electrode substrate for electrochemical Cu deposition to be applied for electrocatalytic CO2 reforming.
4. Conclusions
In our design electrocatalytic electrode was made from nitrogen-doped graphene sheet stacks coated with copper nano- micro structures. The novelty of this study is to obtain nitrogen-doped stacks of graphene sheets (N-GSS) using high-frequency modulated DC pulses for electrochemical exfoliation. The exfoliation process is grateful to dope the resulting material with selected atoms, such as nitrogen—a suitable electrolyte must be used. Industrial graphite waste is taken as the starting material.
Two methods have been tested for the application of the copper catalyst—chemical and electrochemical deposition. Although chemical deposition makes it possible to obtain a suspension of particulate GSS/Cu material, which can be used to form an electrode, the deposition of copper on the surface of GSS sheets is very fragmentary, covering the edges of the sheets more. Further research and improvement of this method is needed.
It has been found that the size of the copper crystals and the uniformity of the coating can be controlled by the length and amplitude of the current pulses, which prefers the electrochemical deposition method to obtain good catalyst electrodes.
Theoretical calculations show that the composite N-doped graphene/Cu structure functions as a CO2 and proton absorber, facilitating hydrogenation and carbon–carbon coupling reactions on Cu-nanocluster/graphene for the formation of C2H4.
In this research, electrochemical deposition method was recognized as successful and most promising to grow Cu crystals on conducting carbon substrates. Further research can focus on the two tasks need to be solved—how to strengthen the bonding of graphene sheet stacks with the carbon substrate and how to increase adhesion of copper nanocrystals to graphene plates.