Electrochemical CO₂ Reduction to CO Catalyzed by 2D Nanostructures

Abstract

Electrochemical CO₂ reduction towards value-added chemical feedstocks has been extensively studied in recent years to resolve the energy and environmental problems. The practical application of electrochemical CO₂ reduction technology requires a cost-effective, highly efficient, and robust catalyst. To date, vigorous research have been carried out to increase the proficiency of electrocatalysts. In recent years, two-dimensional (2D) graphene and transition metal chalcogenides (TMCs) have displayed excellent activity towards CO₂ reduction. This review focuses on the recent progress of 2D graphene and TMCs for selective electrochemical CO₂ reduction into CO.

1. Introduction

The increasing utilization of fossil fuels leads to huge amounts of CO₂ emissions which causes inevitable damage to the planet’s ecosystem [1,2,3,4,5]. One possible way to counter this is to capture and convert CO₂ to industrially important organic compounds [6,7,8,9,10]. To date, several strategies have been employed for the reduction of CO₂; for example, chemical [11], thermochemical [12,13], photocatalytic [14,15,16,17,18,19,20,21], electrocatalytic [22,23,24], biological [25], and inorganic transformation [26]. However, among those, a great deal of research has focused on electrochemical CO₂ reduction technology aimed towards large scale applications due to its environmental compatibility and cost-effectiveness [27,28,29].

However, the low reactivity of CO₂ requires high energy input for its activation [30,31,32,33,34]. CO₂ can be reduced to various useful products by employing suitable cell potential in combination with appropriate catalysts (as shown in

Table 1. The standard redox potentials for the main products of CO₂ reduction in aqueous solution.

Half−Electrochemical Thermodynamic Reactions	Product	E° Redox	Equation
CO2 + e− → CO2•	CO2 anion radical	−1.90 V	(1)
CO2 + 2H+ + 2e− → HCOOH	Formic acid	−0.61 V	(2)
CO2 + 2H+ + 2e− → CO + H2O	Carbon monoxide	−0.53 V	(3)
CO2 + 4H+ + 4e− → HCHO + H2O	Formaldehyde	−0.48 V	(4)
CO2 + 6H+ + 6e− → CH3OH + H2O	Methanol	−0.38 V	(5)
CO2 + 8H+ + 8e− → CH4 + 2H2O	Methane	−0.24 V	(6)
2CO2 + 12H+ + 12e− → C2H4 + 4H2O	Ethylene	−0.41 V	(7)

(Equations (1)–(7)) [35,36]. For example, CO₂ reduction to C₁ products such as CO is thermodynamically more favorable because it requires only two electrons and two protons for the activation, and such C₁ products are useful as building blocks for the production of higher hydrocarbons [37]. In contrast, long-chain hydrocarbons are industrially significant but CO₂ reduction to higher hydrocarbons is a complex process, since C–C bond formation demands high energy input [38]. Hence, rather than CO₂ reduction to higher hydrocarbons, the product selectivity and efficiency are crucial factors to be addressed.

However, developing a low cost, highly efficient, and stable catalyst remains an obstacle to be overcome. To date a number of catalysts including metals [39], alloys [40], metal oxides [41,42], metal complexes [43], carbides [44,45], and metal-organic frameworks [46,47,48] have been studied for electrocatalytic CO₂ reduction. Of these, transition metal-based catalysts (e.g., Au, Pt, Ni, and Ag) displayed notable activity, but their scarcity in nature and high-cost limit them from large-scale employment [49]. Metal oxides, namely, TiO₂, RuO₂, and IrO₂, have been widely investigated [50,51,52], but they also possess notable disadvantages—low efficiency, high cost, poor stability, and low current exchange density. Recently 2D materials, in particular, graphene derivatives and transition metal chalcogenides (TMCs), have emerged as active and inexpensive alternatives to traditional electrocatalysts [52]. They possess unique physicochemical properties and their large surface areas facilitate the redox reaction process to boost the catalytic activity [53,54]. Moreover, exposed active sites of 2D nanostructures are favorable for interfacial charge transfer [55]. Enhancement in mass transport and electron conductivity of 2D electrocatalysts can be achieved by co-catalyst doping [56]. For instance, doping of Nb into MoS₂ increases the overall catalytic performance at the minimal potential of 50–150 mV [57]. Therefore, all these collective advantages of 2D nanostructures make them suitable for electrochemical CO₂ reduction.

This short review will describe the recent progress and advances in emerging 2D materials in electrochemical CO₂ reduction towards CO production. We will be discussing the fundamentals and pathways of electrochemical CO₂ reduction. Various advanced synthesis methodologies of 2D catalysts and the effect of metal/and non-metal doping on catalytic activity will also be discussed along with the theoretical studies.

2. Fundamentals of Electrochemical CO₂ Reduction

CO₂ is a linear molecule and a large input of energy is needed to break the carbon and oxygen bonding (i.e., C=O) [58,59]. H₂O reduction to H₂ and ½O₂ demands 237.2 kJ mol⁻¹ of energy under standard conditions. Whereas, CO₂ reduction to hydrocarbon fuels CH₄ and CH₃OH requires much higher free energy towards positive change, equivalent to 818.3 kJ mol⁻¹ and 702 kJ mol⁻¹, respectively [59]. Therefore, CO₂ reduction is an uphill reaction and considered as a grand challenge [60]. The electrochemical cell used for CO₂ reduction reaction consists of a cathode, an anode, and an ion-exchange membrane placed in-between them, as demonstrated in Figure 1. The process of electrochemical CO₂ reduction begins with diffusion of CO₂ molecule from the solution to cathode surface which is coated with a suitable catalyst. One of the postulated mechanisms for activating CO₂ molecule is a single electron transfer to CO₂, thereby generating a CO₂^•− radical anion [61]. A single-electron addition bends the linear molecule because of the repulsion between acquired electron over electrophilic carbon atom and free electron pairs on the oxygen atom. However, this mechanism has a setback, because for such a process to take place, a very high redox potential is required (i.e., −1.9 V versus NHE) [62]. But then, as seen in Table 1, which depicts the standard redox potential for CO₂ reduction reactions, the possibility of single-electron transfer to a CO₂ molecule becomes inconceivable [63]. In terms of thermodynamics, the edges of conduction band and valence band of a semiconductor must be above and below the standard reduction and oxidation potentials, respectively. Therefore, this makes it quite clear why a single electron transfer to the CO₂ molecule is highly improbable. Hence, a better efficiency can be achieved with a multiple proton-coupled electron transfer (PCET) process. This process involves the transfer of protons with electrons, thereby bypassing the formation of high energy CO₂^• radical [64]. Even though the PCET phenomenon makes CO₂ reduction feasible, the process is dependent upon the concentration of available electrons and protons. For example, as manifested from Equations (2)–(7), which show the main products obtained during CO₂ reduction in the presence of H₂O, by the PCET phenomenon (pH = 7) [65], generation of formic acid and carbon monoxide is more favorable than generating methane. The reason behind such a phenomenon is that the carbon monoxide requires two electrons and protons, whereas eight electrons and protons are necessary for producing methane. Another possible pathway involved in CO₂ reduction to CO is decoupled electron-proton transfer, where the CO production can be carried out by the formation of a CO₂²⁻ radical [55]. However, theoretical studies and quantum chemical calculations provide a better understanding of the mechanism and reaction pathway of electrochemical CO₂ reduction. For example, the computational hydrogen electrode (CHE) model is one of the most convenient and widely employed methods for the PCET process [66]. Paterson et al. reported a CHE model for electrochemical CO₂ reduction on the metallic copper electrode and it has been found that metal electrodes can easily generate hydrocarbon products upon CO₂ reduction [67,68]. On the other hand, the natures of the catalysts used for catalytic reaction play important roles in the reaction pathway.

Another challenge in electrochemical CO₂ reduction is competition with proton reduction, i.e., hydrogen evolution reaction (HER), because it is a two-electron and proton process, requiring far less energy [69]. This leads to poor selectivity to carbonaceous products. Last but not the least, CO₂ has very low solubility in an aqueous medium (≈0.034 M), which negatively affects the overall CO₂ reduction activity to chemical fuels. There are efforts to overcome this solubility problem by using gas diffusion layers to use the polymer electrolytes. In conclusion, both thermodynamic and kinetic barriers for the transformation to value-added fuels make it a challenging reaction. In the anode half of the electrochemical cell, oxygen evolution takes place. Lastly, the ion exchange membrane functions through avoiding oxidation of carbonaceous products by acting as a barrier to prevent the flow of electrons and allowing passage of protons only. A schematic of the electrochemical cell and reactions going on is shown in Figure 1, with the following key points:

• CO₂ adsorption on cathode surface;
• PCET process;
• Desorption of products formed from the cathode surface.

2.1. Parameters of Electrochemical CO₂ Reduction

To examine the utility of electrochemical CO₂ reaction there have been key parameters set that include current density (j), overpotential (η), faradaic efficiency (FE), turnover frequency (TOF), effect of electrolyte, and stability. Those parameters will be discussed in this section, which will assist the readers with understanding the concepts and applying them while evaluating their experiments pertaining to electrochemical CO₂ reduction in the laboratory.

2.1.1. Current Density (j)

Current density is an important parameter for evaluating the activity of electrochemical CO₂ reduction. Application of a constant cell voltage gives rise to electrolysis current. Therefore, current density (j) is described as the ratio between current (i) and the geometric surface area of working electrode (A), as shown in Equation (8).

$$\mathit{j}=\frac{\mathit{i}}{\mathit{A}}$$

(8)

2.1.2. Faradaic Efficiency (FE)

Faradaic efficiency is yet another important guideline with which to evaluate the amount of product cross-over and electrochemical selectivity of CO₂ reduction reaction [70]. It is given as the ratio of the charge from total product generated (αNF, whereas α is the number of electrons transferred, N is the amount of desired product’s moles, and F is the faraday constant 96,485 C mol⁻¹) to the total amount of charges passed during the electrolysis (Q), as shown in Equation (9) [55]. It also correlates with the selectivity of products during CO₂ reduction.

$${\mathit{E}}_{\mathit{f}\mathit{a}\mathit{r}\mathit{a}\mathit{d}\mathit{a}\mathit{i}\mathit{c}}= \frac{\mathit{\alpha }\mathit{N}\mathit{F}}{\mathit{Q}}$$

(9)

2.1.3. Overpotential (η)

The difference between the actual required potential and the equilibrium potential is expressed as overpotential. An electrocatalyst showcasing low overpotential is considered to show a better and efficient catalytic activity. Overpotential is calculated by the difference between the standard potential of the reaction of CO₂ and the reduced product (E_eq), and actual electrode potential (E), as shown in Equation (10) [71].

$$\mathit{\eta }=\mathit{E}-{\mathit{E}}_{\mathit{e}\mathit{q}}$$

(10)

2.1.4. Turnover Frequency (TOF)

It is determined as the number of CO₂ conversion products obtained over number of active sites per unit time. The catalyst with a high turnover frequency means it possess many innate active sites, which in turn, provide better activity. TOF is calculated by the ratio of the number of moles of the target product produced per unit time (N_p) to the maximum number of moles of the active site of the catalyst (N_c), as shown in Equation (11) [71].

$$\mathit{T}\mathit{O}\mathit{F}= \frac{ {\mathit{N}}_{\mathit{p}}}{{\mathit{N}}_{\mathit{c}}}$$

(11)

2.1.5. Effect of Electrolyte

The product selectivity can be altered/controlled by varying the electrolyte solution in electrochemical CO₂ reduction reaction [72]. To date, various types of electrolytes have been used, such as non-aqueous (organic) solvents, aqueous and ionic liquids. Among them, aqueous electrolytes are cheap, environmentally friendly, and possess high sustainability. Moreover, in aqueous electrolytes, the concentration, pH value, and buffer capacity are of equal importance for eventually affecting the efficiency and product selectivity [73]. CO₂ is highly soluble in organic electrolytes, but they are toxic, expensive and require large overpotentials compared to aqueous electrolytes. The ionic liquids are polar in nature which further helps to improve the conductivity and solubility of CO₂.

2.1.6. Stability

Stability is one of the most important criteria in order to commercialize a catalyst. Therefore, even if a catalyst has high efficiency, it is equally as important to have long term stability.

3. 2D Electrocatalysts

The 2D materials possess excellent structural and morphological features, as shown in Figure 2. In the following sections, we will briefly describe the various 2D electrocatalysts for CO₂ reduction into CO.

3.1. 2D Molybdenum (Mo) Transition Metal Dichalcogenides (TMDCs)

It is commonly known that the defects, edge sites, and dopants present with a catalyst greatly influence the activity of electrochemical CO₂ reduction. In recent years, 2D MoS₂ has shown great potential because of its unique physicochemical and structural properties. Asadi et al. reported bulk MoS₂ with a layer-stacked structure for electrocatalytic CO₂ reduction to produce CO [74]. STEM (Scanning transmission electron microscopy) investigation confirmed the presence of vertically aligned multi-layered sheets of MoS₂ flakes. Remarkably, as-obtained metal terminated edges of MoS₂ were considered reactive sites for electrocatalytic CO₂ reduction. Excellent faradic efficiency of 98% was achieved with a high current density operated at a minimal overpotential of 54 mV. Therein, the presence of H⁺ from the water molecule (in electrolyte solution) and electron generated on the catalyst’s surface contribute to the reduction of CO₂ to form CO. Density functional theory (DFT) studies revealed that the presence of MoS₂ edge sites and metallic character of Mo atoms were collectively responsible for high catalytic activity and stability over 10 h. That theoretical study, significantly, provides a new idea for the experimental fabrication of a competent electrocatalyst. In another example, Nørskov et al. improved catalytic performance by breaking the linear scaling relationship among *CO, *CHO, and *COOH intermediates at the edge sites of MoS₂ and MoSe₂ [56]. Furthermore, in another study NÒrskov and coworkers demonstrated that transition metal-doped MoS₂ significantly improved the catalytic activity [75]. Product selectivity mainly depends upon binding energies of edge sites present over the catalyst. For example, S edges of Ni-doped MoS₂ and Mo edges of MoSe₂ selectively produce CO. However, S edges of undoped MoS₂ are selective towards H₂ formation.

In 2016, Abbasi and coworkers reported Nb-doped vertically aligned MoS₂ electrocatalyst for CO₂ reduction into CO [57]. A small shift in Raman and XPS spectra confirmed Nb doping in MoS₂ (Figure 3A,B). The CO₂ reduction experiment was performed in an electrochemical cell by employing cyclic voltammetry at a −0.8 ± 1.0 V potential range. All tests were performed in an electrolyte solution composed of a CO₂ saturated, 50:50 vol% mixture of 1-ethyl-3-methylimidazolium tetrafluoroborate (EMIM-BF₄) and deionized (DI) water. This optimized composition of electrolyte solution provides the maximum number of protons for CO₂ reduction reaction. Here, 5% Nb-doped MoS₂ nanosheets showed an improved current density compared to pristine MoS₂ and Ta-doped MoS₂ (Figure 3C). The catalyst displayed faradaic efficiencies oscillating from 12% to 82% towards CO formation in the potential range of −0.8 ± 0.16 V (Figure 3D). The formation rates for CO and H₂ at 0.8 V were 2.23 × 10⁻⁹ mol min⁻¹ and 1.98 × 10⁻⁹ mol min⁻¹, respectively. As a result, Nb-MoS₂ nanosheets exhibited an excellent catalytic activity in terms of turnover frequency (TOF) at an overpotential of 50–150 mV. In-depth DFT calculations were also conducted to understand the CO₂ reduction pathway. The results showed that Mo edges of pure MoS₂ preferably form an *COOH intermediate at lower overpotential. Moreover, Nb atoms help to reduce the binding strength among Mo and CO, thereby minimizing onset potential to produce CO via formation of *CO and *COOH intermediates. These findings revealed that Nb-doped MoS₂ promotes faster desorption of CO compared to pristine MoS₂. Furthermore, DFT calculations revealed that a low concentration of Nb-doping selectively reduced CO₂ to CO. However, an increased Nb concentration over 5% showed a negative effect on CO₂ reduction performance because of the enlarged work function of Mo_1−xNb_xS₂ composite. Therefore, an appropriate amount of metal doping to a catalyst can significantly influence electronic properties that may further intensify the overall catalytic performance. MoS₂ is a well-known catalyst for hydrogen evolution due to its unique structural characteristics where S atoms can make a strong bond with H⁺ in the electrolyte solution [76,77,78]. Therefore, a smaller number of S atoms are available for the CO₂ reduction reaction. Addressing the issue, Yu and coworkers employed a structure-directing approach and fabricated 3D TiO₂@MoS₂ to make bonding between the unsaturated S atom and Ti metal, consequently making a large number of Mo atoms available for CO₂ reduction reaction [76]. The porous nanosheets of TiO₂ acted as a CO₂ adsorbent due to high surface area and abundant active sites (Figure 4A). Bonding between Ti and S atoms improved the electronic and adsorption properties of MoS₂ (Figure 4B showing Raman spectra of Ti-S band). Here, the electrolyte solution played an important role in product selectivity. The N₂-saturated solution led to hydrogen evolution, whereas current density was increased towards CO formation in the CO₂-saturated solution. Faradic efficiency of ≈2.8% towards CO formation was observed at the potential of −0.21 V, while at a more negative potential it reached up to 85% (Figure 4C). The catalyst showed excellent stability over 1000 cycles which revealed its superiority (Figure 4D). Interestingly, long-term stability over 14 h was observed at a steady potential of −0.70 V.

3.2. 2D Graphene

Graphene has been playing a significant role in many advanced applications because of its high chemical stability, superior electronic and thermal properties, and high specific surface area [79]. Typically, pristine graphene is inactive towards electrocatalysis because of its ineffectiveness towards CO₂ activation. This allows a greater extent of freedom for surface modifications to tune the catalytic properties. The graphene templates, graphene oxide (GO), and reduced graphene oxide (RGO) have been largely employed for catalytic applications [80,81,82,83]. The metal and non-metal doping and composite formation with other semiconductors have been exercised to improve the catalytic performance of graphene. To date, metals like Cu, Ni, Au, and Co have been successfully used as dopants/co-catalyst [84]. Organometallic complexes and metal-organic frameworks (MOFs) have also been tried with graphene for CO₂ reduction [85,86]. Heteroatom doping (such as by nitrogen, boron, and sulfur) to graphene can effectively improve the CO₂ adsorption [87]. These heteroatoms act as reaction centers for CO₂ reduction and improve the catalytic activity [88,89]. For example, pyridinic N in the N-doped graphene possesses a lower free energy threshold to produce the COOH intermediate and selectively produce CO [90,91,92,93]. Additionally, the electrocatalyst comprising both metal and non-metal is significant for heterogeneous catalysis aiming towards product selectivity [94]. Su et al. synthesized Ni-nitrogenated-graphene (Ni-N-Gr) sheets for generation of CO from CO₂ [95]. The CO₂ reduction capability was examined by using cyclic voltammetry in the buffer solution and products were detected by GC-MS technique. At an optimized condition, Ni-N-Gr displayed the highest faradaic efficiency over 90% for CO production at −0.7 to −0.9 V. The Ni-N-Gr showed excellent catalytic performance with much higher activity than pristine Ni and N-Gr samples. To explore the significance of Ni in Ni-N-Gr composite, Cu-N-Gr was also tested as a control sample and results showed positive feedback towards the presence of Ni in composite. Further, the turnover frequencies of CO per electrochemically active Ni atom was up to 2700 h⁻¹ at −0.7 V and 4600 h⁻¹ at −0.8 V versus RHE. In another similar study, Ni²⁺ on nitrogen-doped graphene (Ni²⁺@NG) was investigated for CO formation through electrocatalysis [88]. Previously, such types of catalysts were synthesized by utilizing the conservative pyrolysis technique (800–1000 °C) and acid leaching process, which are not so ecofriendly. Therefore, in this work the synthesis was carried out by employing a simple solution-chemistry approach which is convenient for large scale applicability. In this method, first N-doped graphene with an ultrathin nanosheet structure was synthesized by pyrolyzing a mixture of g-C₃N₄ and glucose. Then, immobilization of Ni²⁺ in N-doped graphene was carried out by reacting Ni(NO₃)₂ with N-doped graphene in the aqueous solution. The presence of Ni²⁺ ions and Ni valence state on N-doped graphene were confirmed by XPS and XANES spectra, respectively. The electrocatalyst produced CO with the faradaic efficiency of 92% at −0.68 V versus RHE. In contrast, the efficiency of pristine N-doped graphene was below 60% at all potential capacities. Therefore, Ni²⁺ sites significantly facilitated the kinetics of CO₂ reduction. Furthermore, to understand the effects of other transition metals ion, metals such as Co²⁺ and Fe³⁺ were likewise restrained on N-doped graphene. All the catalysts showed almost 4–5 times higher activity than pristine N-doped graphene. The composite with Ni²⁺ showed the highest faradaic efficiency of 92% compared to that of Fe³⁺ (85%) and Co²⁺ (65%); see Figure 5A. The formation of CO occurred via the *COOH (* is surface sites) intermediate (Figure 5B). The quadruple-pyridinic configuration of N-doped graphene exhibited the maximum free energy barrier which protonates *COOH to form *CO.

In certain composites, stabilization of the single metal atom within the 2D nanostructure is crucial to prevent aggregation. To address the issue, Jeong et al. reported single-atom Ni activated on RGO sheets for CO formation which showed a faradaic efficiency of 97% at 0.8 V versus RHE [96]. The linkage between Ni atoms and GO was accomplished by using the tris(2-benzimidazolylmethyl)amine (NTB) ligand (Figure 6A). Firstly, Ni(NTB) complex was prepared. The NTB ligand makes a strong bond with GO via π–π interaction and Ni ions via ligation by forming Ni(NTB)-GO complex. A uniform distribution of Ni atoms on 2D-RGO sheets was confirmed by TEM analysis. XANES analysis was used to confirm that nickel was in Ni²⁺ state rather than Ni⁰. In an inert condition, the catalysts displayed moderate activity towards the H₂ at −0.71 V versus RHE. Significantly the current density was improved in CO₂-saturated electrolyte, reaching the onset potential of −5 mA cm⁻² at −0.60 V versus RHE. The observed current density was almost double at −1.1 V versus RHE, which implies its superior activity compared to that of control samples (Figure 6B). Ni-N-RGO showed the maximum faradaic efficiency (97%) for CO formation at −0.8 V versus RHE (Figure 6C). The reactivity of Ni-N₄ active sites towards the intermediate formed in the reaction are responsible for such high catalytic performance. DFT study revealed that the proton interactions with active sites of Ni-N-RGO are unfavorable towards HER. Therefore, selectivity was observed towards CO₂ reduction rather than H₂ formation.

Jiang and coworkers also demonstrated that graphene shells with embedded Ni atoms were active electrocatalysts for CO₂ reduction [97]. From DFT calculations it was observed that Ni atoms in the graphene layer can significantly lower the activation barrier of CO₂ molecule, which further weakens the binding with CO to form a facile product by lowering the proton reduction reaction. In this way, Ni-graphene electrocatalyst showed an excellent catalytic activity towards CO formation with a faradaic efficiency of 93.2% at current of 20 mA/mg. The turnover frequency (TOF) of 8 s⁻¹ was observed for the active Ni center and a 454,000 cumulative turnover number (TON) was recorded below −0.7 V overpotential during 20 h of continuous electrolysis.

Jiang and coworkers reported more results with single Ni atoms emended in a graphene sheet [98]. Resulting catalysts endow CO selectivity over 90% of in aqueous solution, and the highest faradaic efficiency of CO (95%) was obtained at an overpotential of 550 mV. Specifically, Ni single atomic sites present in graphene were responsible for high selectivity of CO with 90% of faradaic efficiency over the stability of 20 h. Further, CO formation current, more than 200 mA (50 mA cm⁻²) under a faradaic efficiency of 97%, was established by using a 4 cm² anion membrane electrode setup. Other transition metals, including Co, Fe, and Mn, were also dispersed on a graphene layer and it was observed that the reaction pathway changes with change in transition metals, specifically hydrogen was produced in the presence of Co. DFT studies revealed that for the Ni single atoms, compared to Co atoms, the weaker binding of CO and higher barrier for hydrogen evolution reaction accounted for the superior CO₂ reduction towards selective formation of CO.

The stability of the oxidation state among metals is very crucial for catalytic reactions. Copper is proven as one of the best and inexpensive catalysts for CO₂ reduction reaction. Karapinar et al. reported highly active, single-site Cu-N-C for electroreduction of CO₂ [99]. They found that during the catalytic process, isolated copper sites converted into metallic copper, which eventually act as active sites for the CO₂ reduction. Apart from nickel and copper, the iron-based catalysts are also considered as the superior and most active materials due to their chemical stability towards the CO₂ reduction under negative potential. The key parameter of this material is that Fe³⁺ can easily coordinate with N heteroatom supported on carbon and maintains its 3+ oxidation state throughout the electrocatalytic reaction [100]. Also, such Fe³⁺ sites can adsorb a greater number of CO₂ for electrocatalysis. Zhang and co-workers reported a hybrid composite of Fe/N-Graphene [101]. The electrocatalyst was fabricated by annealing the mixture of GO with iron chloride (FeCl₃) at 700–800 °C in an inert atmosphere, as shown in Figure 7A. The oxygen functional groups in GO adsorbed metal cations through the electrostatic interaction. Whereas, Fe atoms in N-graphene matrices created Fe-N₄ sites, which further acted as CO₂ activators. N dopant acted as an immobilizer among Fe atoms and the graphene surface through metal–nitrogen interaction. As a result, catalysts undergo selective CO₂ reduction towards CO formation with a faradaic efficiency of 80% at low potential of −0.60 V versus RHE. The catalyst was operated over 10 h at −0.60 V versus RHE to examine its long-term stability. The mechanism of CO₂ to CO conversion was studied by theoretical simulations. The formation of CO took place by sharing of two protons and two electrons via COOH* intermediate, as revealed in Figure 7. Here, substitution of nitrogen to graphene lowered the energy barrier of COOH* intermediate and facilitated the CO* formation.

Zhu and coworkers reported the superior catalytic activity of 94.2%, the faradaic efficiency for CO formation by employing 2D N-doped carbon nanorod arrays/ultrathin carbon nanosheet catalysts [102]. In the synthesis, aniline polymerization was carried out in the presence of carbon nanosheets and acid treatment, and resulting catalysts were obtained by pyrolysis. It was observed that the catalyst was proficient of retaining its activity for 30 h at a minimal onset overpotential of 97 mV. In the Tafel slope, a low value of 65 mV dec⁻¹ specified a fast pre-equilibration process of transfer of one electron followed by a rate-determining step. DFT calculation suggests that the N pyridinic sites and 2D structure were responsible for lowering the free energy barriers of intermediates (*COOH and *CO) and boosting CO desorption to form CO. Therefore, these results showed an excellent catalytic activity among other reported doped carbon materials.

Apart from single metal atom insertion, the transition metal sulfides have also been tried with graphene. For example, silver sulfide (Ag₂S) possesses excellent physical, chemical, and thermal properties that are suitable for catalytic applications. Along those lines, Zeng et al. reported Ag₂S/N-S doped rGO hybrid for CO₂ reduction [103]. A hydrothermally synthesized, Ag₂S/N-S-doped rGO electrocatalyst was employed for CO₂ reduction toward CO. Here, CO, H₂, and CH₄ were obtained as products, while selectivity for CO was observed in the wide potential range of −1.08 to −0.58 V versus RHE. The catalyst manifested a current density of 470 µAcm⁻² at −0.759 V versus RHE. An excellent catalytic activity owing to a faradaic efficiency of 87%, and long-term stability over 40 h, were achieved by the catalyst. The faradaic efficiency of this catalyst was enormously higher than pristine N-S-rGO (3.2%) and Ag₂S nanowires (41.7%) [104]. Significantly, an increase in current density was observed, as the potential was swept cathodically which specified the increase in CO₂ reduction kinetics in the system. In the CO formation mechanism, Ag ions bind with CO₂ molecule and generate a COOH intermediate by combining one proton and one electron. Afterwards, the intermediate (COOH) further reacts with a second proton and electron pair to form CO. The high surface area of rGO maximized the reactive sites for CO₂ molecule and N-S doping eventually enhanced the electrical conductivity of rGO.

Pitch derived 2D N-doped porous carbon nanosheets, which are similar to graphene, were also found to be a superior catalyst for selective CO₂ reduction with high stability [105]. For synthesis, the authors used a NH₃ etching strategy using NaCl template. NH₃ etching increases the catalyst’s surface area by creating nanopores, and further, helping to insert nitrogen atoms into a carbon framework. Resulting catalysts showed high catalytic performance with 84% of faradaic efficiency at an overpotential of −0.49 V and a negligible onset overpotential of −0.9 V for CO production. Such excellent catalytic behavior was due to presence of nanoporous 2D architecture and the high content of pyridinic N active sites which favored fast charge transfer, and high CO₂ adsorption. DFT studies confirmed that the pyridinic N sites favorably bind with COOH* intermediates and further facilitate catalytic reduction.

3.3. 2D Tungsten TMDCs

Tungsten is an emerging candidate for sustainable energy applications due to its low price, non-toxicity, and large atomic size compared with other TMDCs [106,107,108,109,110,111]. Due to its exciting electronic properties, tungsten dichalcogenides act as a catalytic site for electrochemical CO₂ reduction reactions. Also, tungsten dichalcogenides have the anisotropic properties of its basal and edge plane; therefore, most of the catalytic reactions occur on the edge sites [112]. These limited catalytic sites can be increased by transition metal doping [113]. In this regard, Asadi et al. fabricated the WSe₂ nanoflakes (NFs) by a chemical vapor transport growth and exfoliation technique [114]. They found that the CO₂ reduction activity of TMDCs in 50% EMIM-BF₄ with water was much higher compared to other TMDCs and Ag NPs. According to work functions estimation, the trend of monolayer TMDCs showed activity in the order WSe₂ < MoSe₂ < WS₂ < MoS₂ (Figure 8A). WSe₂ NFs had the lowest work function of those comprised of monolayered TMDCs. Also, electrochemical impedance spectroscopy (EIS) outcomes were matched with the work function trend (Figure 8B).

Charge transfer resistance (R_ct) is an important kinetic factor for the electrochemical rate constant. Charge transmission resistance of WSe₂ NFs (≈180 Ω) was lower than MoS₂ and Ag nanoparticles (Figure 8B) [114]. Therefore, WSe₂ possesses a fast electron transfer property which eventually affects the CO₂ reduction reaction. Experiments were performed in an electrolyte solution of 50% EMIM-BF₄, 50% water. It was found that WSe₂ NFs displayed a 18.95 mA cm⁻² current density for CO at a minimal potential of 54 mV (Figure 8C,D). Here, EMIM⁺ ion in the ionic liquid acted as a transporter of CO₂ to the catalyst surface. Also, the CO intermediate was more stable on the metallic edges of WSe₂ NFs, suggesting the possibility of CO formation after CO₂ reduction. Therefore, such systems owing to high catalytic performance operating at low overpotentials have the disruptive potential to solve the current obstacles of CO₂ reduction technology.

3.4. 2D InSe

Among various 2D materials, InSe is also considered as an emerging electrocatalyst. Although InSe is not a transition metal chalcogenide, we intentionally added this section to our study. Recently, the theoretical study about InSe was carried out by Jiang et al., which demonstrated its potential for CO₂ reduction [115]. The single metal (e.g., Rh, Mn, Cu, Zn, and Cr) was embedded in a 2D InSe substrate and catalytic reactivity towards CO, HCOOH, H₂, and CH₄ were studied. They showed that InSe systems with Rh, Mn, and Cu metals obeyed the linear scaling relationships among adsorption energies of intermediates, such as *CO and *COOH. However, the relationship broke down for system containing Zn and Cr metals. Such an interesting feature of 2D InSe offers a reliable method for the experimental usage, and in turn, the product selectivity. The Zn@2D InSe system effectively produces CO with limiting the potential of −0.78 V because it is situated on the scaling line among *COOH and *CO. Its lowered desorption energy (0.14 eV) executes selectivity towards CO. Therefore, such theoretical studies enrich the understanding of 2D systems for electrocatalytic applications.

The summary of various 2D nanomaterials for electrochemical CO₂ reduction is displayed in

Table 2. The summary of various 2D nanomaterials for electrochemical CO₂ reduction. Details included reactor type, example of catalysts, current density, onset potential, over potential, and the main product of CO₂ reduction with faradaic efficiency and stability.

No	Catalyst	Current Density (versus Applied Potential)	Onset Potential (V) or Overpotential (versus RHE)	CO Faradaic Efficacy (FE% versus RHE)	Stability (versus RHE)	Reactor (Type, Volume), Electrolyte	Ref, Year
1	MoS2	65 mA cm−2	−0.764 V	98% @ −0.764 V	10 h @ −0.764 V	Two-compartment three-electrode electrochemical cell in Ionic liquid, 96:4 mol% water:EMI-BF4	[74], 2014
2	Metal doped MoS2	Theoretical study					[56], 2016
3	MoS2 and MoSe2	Theoretical study					[75], 2014
4	Nb-MoS2	237 mA cm−2	31 mV @ −0.8 V	82% @ −0.8 V	10 h @−0.8 V	Two-compartment three-electrode electrochemical cell, Ionic liquid, CO2 saturated50:50 vol% EMIM-BF3:water	[57], 2016
5	3D TiO2@MoS2	68 mA cm−2	100 mV @ −0.7 V	85% @ −0.70 V	14 h @ −0.70 V	Three-electrode system, CO2-saturated 0.1 M Potassium bicarbonate	[76], 2018
6	Nickel-nitrogen-modified graphene (Ni-N-Gr)	≈1.45 mA cm−2	−0.5 V	90% @ −0.7 to −0.9 V	5 h @ −0.65 V	CO2-saturated 0.1 M Potassium bicarbonate	[95], 2016
7	Ni2+ on N-doped graphene (Ni2+@NG)	10.2 mA cm−2	≈−0.40 V	92% @ −0.68 V	20 h (≈2.85 → ≈−2.44 mA cm2) @ −0.58 V	Three-electrode system, CO2-saturated 0.5 M Potassium bicarbonate	[88], 2019
8	Atomic iron dispersed on N-doped graphene (Fe/NG)	2.6 mA cm−2	− 0.28 V	80% @ −0.60 V	10 h @ −0.60 V	CO2-saturated 0.1 M Potassium bicarbonate solution	[101], 2018
9	Nickel- and nitrogen-doped reduced graphene oxide (Ni-N-rGO)	5 mA cm−2	−0.60 V	97%@ −0.8V	-	Two-compartment electrochemical cell (Nafion 117 membrane) CO2-saturated 0.5 M Potassium bicarbonate	[96], 2018
10	Nickel-N-doped graphene shell (Ni-N-GS)	4 mA cm−2	0.7 V	93.2% @ 0.7 V	20 h @ 0.7 V	A homemade Teflon H-cell, CO2-saturated 0.5 M Potassium bicarbonate	[97], 2017
11	Ni single atoms in graphene nanosheets (Ni-G)	50 mA cm−2	480 mV	90% @ −0.87 V	20 h @ 0.64 V	A homemade Teflon H-cell, CO2-saturated 0.5 M Potassium bicarbonate	[98], 2018
12	1D/2D N-doped carbon nanorod arrays/ultrathin carbon nanosheets	3.78 mA cm−2	97 mV, @ −0.45 V	94.2% @ −0.45 V	30 h, @ −0.45 V	A homemade three-electrode system, CO2-saturated 0.5 M Potassium bicarbonate	[102], 2019
13	Ag2S/N-S-doped rGO	70 μA cm−2	−0.34 V	87.4% @ −0.76V	40 h @ −0.759 V	Double-chamber electrochemical cell with a Nafion 117 membrane 0.1 M Potassium bicarbonate	[103], 2018
14	2D N-doped porous carbon nanosheets	1.15 mA cm−2	−0.19 V	84% @ −0.49 V	8 h @ −0.60 V	Ar or CO2-saturated 0.1 M Potassium bicarbonate	[105], 2018
15	WSe2 nanoflakes (NFs)	18.95 mA cm−2	54 mV	24% @ 54 mV	27 h @ −0.364 V	Three-electrode two- compartment cell ionic liquid set-up (50% EMIM-water)	[114], 2016
16	Metal@2D InSe	Theoretical study					[115], 2019

.

4. Conclusions

In the field of electrocatalytic CO₂ reduction, recent studies were mainly concentrated on the design of various types of efficient catalysts. A variety of tailored 2D nanostructures were fabricated by researchers with the rational design of experimental procedures. In particular, metal and non-metal doped 2D templates could be highly efficient and stable catalysts. In this study, we have briefly reviewed the recent developments in selective electrochemical CO₂ reduction to CO using various 2D nanostructures. The foremost goal of this study was to introduce the fundamentals and different synthetic strategies of 2D materials. The high surface area of 2D nanostructure encourages CO₂ adsorption, and the presence of edge sites/dopants helps to generate reactive sites for CO₂ reduction. Besides, single atom doped 2D structures hold great potential for electrocatalytic CO₂ reduction, since they have been shown to have high faradaic efficiency along with CO selectivity. While highlighting the different fabrication methods, we have also paid considerable attention to the theoretical studies because theoretical simulations are of equal importance to understand the reactivity, product selectivity, and mechanism of CO₂ reduction pathway. Although several noteworthy successes in CO₂ reduction have been achieved by 2D based nanostructures, further efforts are required for large scale applications. Additionally, more focus should be given toward improving upon the green, scalable, and sustainable electrocatalyst fabrications, which is becoming a crucial factor for realistic industrial applications. Apart from that, CO₂ transformation into higher hydrocarbon products such as ethane, ethanol, or ethylene is also of similar importance, and has a great challenge ahead due to the low efficiency of the existing catalysts. Copper-based catalysts have been proven to be superior materials for generating higher hydrocarbons, including C₂ and C₃ products. However, we believe that the interesting surface, electronic, and structural properties of 2D materials could also be significant in higher hydrocarbon production in the near future.