Photocatalytic CO2 Conversion into Solar Fuels Using Carbon-Based Materials—A Review

Carbon materials with elusive 0D, 1D, 2D, and 3D nanostructures and high surface area provide certain emerging applications in electrocatalytic and photocatalytic CO2 utilization. Since carbon possesses high electrical conductivity, it expels the photogenerated electrons from the catalytic surface and can tune the photocatalytic activity in the visible-light region. However, the photocatalytic efficiency of pristine carbon is comparatively low due to the high recombination of photogenerated carriers. Thus, supporting carbon materials, such as graphene, CNTs (Carbon nanotubes), g-C3N4, MWCNs (Multiwall carbon nanotubes), conducting polymers, and its other simpler forms like activated carbon, nanofibers, nanosheets, and nanoparticles, are usually combined with other metal and non-metal nanocomposites to increase the CO2 absorption and conversion. In addition, carbon-based materials with transition metals and organometallic complexes are also commonly used as photocatalysts for CO2 reduction. This review focuses on developing efficient carbon-based nanomaterials for the photoconversion of CO2 into solar fuels. It is concluded that MWCNs are one of the most used materials as supporting materials for CO2 reduction. Due to the multi-layered morphology, multiple reflections will occur within the layers, thus enhancing light harvesting. In particular, stacked nanostructured hollow sphere morphologies can also help the metal doping from corroding.


Introduction
Global climate change due to the rise in global warming and the energy crisis due to the depletion of fossil fuels have become significant threats to the Earth in recent years. The rising human population, excessive transportation use, and rapid industrialization have all contributed to the excessive usage of fossil fuels, which has led to increased CO 2 emissions into the atmosphere. The combustion of fossil fuels, such as petrol, oil, diesel, natural gas, and biogas, is the major source of CO 2 emission. Though several gases contribute to environmental pollution, CO 2 emission is the utmost among them. The dependence on fossil fuels has grown significantly, particularly since the industrial revolution, which will prevent the next few decades' worth of coal, oil, and natural gas reserves from being fully utilized.
Reports on greenhouse gas emissions and their detrimental effects on the atmosphere's natural balance are frequently released. Figure 1a depicts a new assessment by the International Energy Agency (IEA) on the global extreme CO 2 emissions from industrial processes and energy combustion from 1900 to 2021. The report states that the amount of gas being released into the environment has increased annually and will reach 36.3 billion tons in 2021. As was already mentioned, CO 2 addresses both the question of global warming and the issue of energy. As a result, it is constantly imperative and global warming and the issue of energy. As a result, it is constantly imperative and incumbent upon researchers to address the problems related to CO2 emission. The capture, storage, and conversion of carbonic acid gas to its derivatives are the three fundamental steps for the research community to reduce the rate of CO2 release into the environment and utilize its derivatives for further energy needs. There are numerous processes available for the reduction of CO2 into its derivatives, such as photochemical, electrochemical, thermochemical, and photo-biochemical approaches with certain both benefits and shortcomings. Thermochemical processes demand high temperatures, while electrochemical processes have high voltaic requirements. Biological activities additionally necessitate ambient temperature and pressure. Among them, photochemical or photocatalytic techniques may become the most suitable technique for photocatalytic redox reactions [1]. It avails natural resources, such as water and sunlight for CO2 reduction. Semiconductor materials with appropriate bandgap and efficient catalytic properties are frequently used as stand-alone catalysts and composites with variously doped metal oxides, especially TiO2 metal sulfides, perovskites oxides/halides, and graphitic carbon nitrides (g-C3N4), as shown in Figure 1b   Carbon is one of the Earth's most abundant, widely distributed elements. Carbonbased materials have been largely used in photocatalytic CO2 reduction. Considering their vast availability, low cost, and eco-friendliness, carbon-based materials with elusive 0D, 1D, 2D, and 3D nanostructures and high surface area provide certain emerging applications in electrocatalytic and photocatalytic CO2 utilization, solar cells, batteries, supercapacitors, biosensors, and energy storage and conversions [4]. Graphene [5], graphene oxides [6], carbon nanotubes (CNTs) [7], g-C3N4 [8], conducting polymers [9], and carbon aerogels [10] are the most used carbonaceous materials for photocatalysis. These are the most prevailing catalysts, exhibiting a better catalytic efficiency both as a stand-alone and supporting material [11]. Carbon materials are excellent supporters as they could highly help in improving the adsorption and activation of CO2. In addition, they also play a major role in a wide range of light absorption and help in enhancing the stability of the catalysts with which it has been composited.
This review elucidates the advantages of utilizing carbon materials for photocatalytic CO2 reduction. In the front part of the review, we make a brief note on the fundamental ideas underlying photocatalytic CO2 reduction and the basic requirements that should be satisfied by a photocatalyst to efficiently participate in CO2 reduction. In the later part, we make a detailed discussion about the contribution of carbon materials in enhancing the efficiency of the catalysts using some of the recently published works. Carbon is one of the Earth's most abundant, widely distributed elements. Carbonbased materials have been largely used in photocatalytic CO 2 reduction. Considering their vast availability, low cost, and eco-friendliness, carbon-based materials with elusive 0D, 1D, 2D, and 3D nanostructures and high surface area provide certain emerging applications in electrocatalytic and photocatalytic CO 2 utilization, solar cells, batteries, supercapacitors, biosensors, and energy storage and conversions [4]. Graphene [5], graphene oxides [6], carbon nanotubes (CNTs) [7], g-C 3 N 4 [8], conducting polymers [9], and carbon aerogels [10] are the most used carbonaceous materials for photocatalysis. These are the most prevailing catalysts, exhibiting a better catalytic efficiency both as a stand-alone and supporting material [11]. Carbon materials are excellent supporters as they could highly help in improving the adsorption and activation of CO 2 . In addition, they also play a major role in a wide range of light absorption and help in enhancing the stability of the catalysts with which it has been composited.
This review elucidates the advantages of utilizing carbon materials for photocatalytic CO 2 reduction. In the front part of the review, we make a brief note on the fundamental ideas underlying photocatalytic CO 2 reduction and the basic requirements that should be satisfied by a photocatalyst to efficiently participate in CO 2 reduction. In the later part, we make a detailed discussion about the contribution of carbon materials in enhancing the efficiency of the catalysts using some of the recently published works.

Principles of Photocatalytic CO 2 Reduction
The most promising method for reducing the high levels of greenhouse gases in the atmosphere involves photocatalytic processes since they can provide the required thermal energy for their effective reduction into their fuel derivatives [12]. However, there are still challenges associated with each step of its photocatalytic reduction. Here are the four principal steps that have been involved in CO 2 reduction into its derivatives: adsorption of light energy by the photocatalyst, generation of electron-hole pairs, transportation of the charge carriers, and the chemical reaction between the charge carriers and surface species [13].
The first and foremost task in photocatalysis is to activate a semiconductor by absorbing a maximum amount of natural sunlight or from artificial light energy sources. The photons of light energy that are being absorbed should be equal to or more than those of the bandgap energy of the semiconductor photocatalysts [14]. After absorption of the required light energy, electrons will be generated. The photogenerated electrons become excited and transfer from the completely filled valence band (VB) to the empty conduction band of the semiconductor. The received charge carriers in the conduction band (CB) undergo a redox reaction with the adsorbed surface species on the semiconductor to obtain the aiming reduced products. To be effective, an ideal photocatalyst must have a conduction band edge potential above the reduction potential for the given product, while the valence band edge potential should be well below the oxidation level of the universal solvent, as shown in Figure 2a [15]. Depending on the number of electrons involved, a wide range of C 1 and C 2 products, such as methane (CH 4 ), methanol, formic acid, and carbon monoxide (CO), can be generated from the proton-assisted multi-electron reduction of CO 2 . CO and CH 4 are the most obtained products so far [16]. Along with the useful redox reactions of CO 2 , several other backward reactions can also occur. At times, photocatalytic reduction processes become more complex when intermediates and products are converted back into CO 2 . Despite extensive research in this area, the fast electron-hole recombination rate prevents this photocatalytic reduction of CO 2 from achieving the desired photocatalyst activity and stability [17]. Incorporating carbon materials as supporting materials could help in diminishing these shortcomings and increase photocatalytic efficiency by domineering the forward CO 2 reduction reaction to solar fuels.

Principles of Photocatalytic CO2 Reduction
The most promising method for reducing the high levels of greenhouse gases in the atmosphere involves photocatalytic processes since they can provide the required thermal energy for their effective reduction into their fuel derivatives [12]. However, there are still challenges associated with each step of its photocatalytic reduction. Here are the four principal steps that have been involved in CO2 reduction into its derivatives: adsorption of light energy by the photocatalyst, generation of electron-hole pairs, transportation of the charge carriers, and the chemical reaction between the charge carriers and surface species [13].
The first and foremost task in photocatalysis is to activate a semiconductor by absorbing a maximum amount of natural sunlight or from artificial light energy sources. The photons of light energy that are being absorbed should be equal to or more than those of the bandgap energy of the semiconductor photocatalysts [14]. After absorption of the required light energy, electrons will be generated. The photogenerated electrons become excited and transfer from the completely filled valence band (VB) to the empty conduction band of the semiconductor. The received charge carriers in the conduction band (CB) undergo a redox reaction with the adsorbed surface species on the semiconductor to obtain the aiming reduced products. To be effective, an ideal photocatalyst must have a conduction band edge potential above the reduction potential for the given product, while the valence band edge potential should be well below the oxidation level of the universal solvent, as shown in Figure 2a [15]. Depending on the number of electrons involved, a wide range of C1 and C2 products, such as methane (CH4), methanol, formic acid, and carbon monoxide (CO), can be generated from the proton-assisted multi-electron reduction of CO2. CO and CH4 are the most obtained products so far [16]. Along with the useful redox reactions of CO2, several other backward reactions can also occur. At times, photocatalytic reduction processes become more complex when intermediates and products are converted back into CO2. Despite extensive research in this area, the fast electron-hole recombination rate prevents this photocatalytic reduction of CO2 from achieving the desired photocatalyst activity and stability [17]. Incorporating carbon materials as supporting materials could help in diminishing these shortcomings and increase photocatalytic efficiency by domineering the forward CO2 reduction reaction to solar fuels.

Multi-Dimensional Carbon-Based Materials for Photocatalytic CO2 Reduction
Carbon, a tetravalent non-metallic element that is abundant in the atmosphere, can be classified into three categories based on their dimensionality, such as zero-dimensional (0D) fullerenes, one-dimensional (1D) carbon nanotubes or carbon nano-fibers, two-dimensional (2D) graphite and graphene nanosheets, and three-dimensional (3D) hybrid carbon networks and mesoporous carbons, as shown in Figure 2b. Carbon can be chemically embellished or functionalized using metallic nanoparticles (NPs). The large surface area with defective sites, high porosity, excellent electron conductivity, and chemical

Multi-Dimensional Carbon-Based Materials for Photocatalytic CO 2 Reduction
Carbon, a tetravalent non-metallic element that is abundant in the atmosphere, can be classified into three categories based on their dimensionality, such as zero-dimensional (0D) fullerenes, one-dimensional (1D) carbon nanotubes or carbon nano-fibers, two-dimensional (2D) graphite and graphene nanosheets, and three-dimensional (3D) hybrid carbon networks and mesoporous carbons, as shown in Figure 2b. Carbon can be chemically embellished or functionalized using metallic nanoparticles (NPs). The large surface area with defective sites, high porosity, excellent electron conductivity, and chemical inertness makes carbon both a standalone catalyst or catalyst support with semiconductors to obtain an effective reduction of CO 2 . While considering utilizing a standalone catalyst with exceptional physical and chemical stability, low cost, and relatively simple synthesis processes, g-C 3 N 4 with a layered structure and a moderate bandgap of 2.7 eV is the most popular visible-light active photocatalyst. Most crucially, g-C 3 N 4 is capable of conducting photocatalytic CO 2 reduction, which is thanks to its strongly negative conduction band potential of −1.1 V against the standard hydrogen electrode (NHE). As opposed to photocatalysts that incorporate with precious metal and metal salts to prepare, g-C 3 N 4 is independent of metals and so can be easily made by thermally polycondensing cost-effective N-rich precursors, such melamine, cyanamide, dicyanamide, and urea [18,19]. The carbon-supported catalysts use high-surface-area activated carbon and carbon black as their primary ingredients since they are inexpensive and widely available [20,21]. One-dimensional fullerene (C 60 ) is an allotrope that has a highly conjugated three-dimensional structure. The delocalized conjugated structure of such a singular kind gives C 60 an exceptional electronic property. Particularly, its ability to accept electrons is crucial for electron-transfer processes, as they can render a quick charge separation and a slow carrier recombination [22]. CuO-C 60 [23] and C 60 polymer [24] are some of the few examples by which we could observe these heterostructures exhibiting improved charge separation and photocatalytic activity. Nanoscale carbon compounds with at least one dimension smaller than 10 nm are known as carbon dots (CDs), which consist of hybridized carbon atoms with sp 2 and sp 3 and have different surface functional groups. The inherent quantity of edges and flaws in carbon dots allows for quantum-confinement processes that boost the exposure of such pyridinic N atoms. As reported in recent work on CDs/CdS [25] and in VCQDs/C 3 N 4 heterojunctions with carbon vacancy [26], due to the quantum confinement effect, the bandgap is widened in graphene quantum dots in a size-dependent manner. In addition to optical sensing, light-emitting diodes, these 0D graphene carbon dots are also promising catalysts for photocatalysis and energy storage [27].
Carbon nanotubes (CNTs) are hollow and one-dimensional materials with nanocapillaries having the capacity to absorb gas molecules, a significant specific surface area of about 1000 m 2 /g, resilient electrical conductivity, and thermal conductivity [28]. These are ideal substrates for heterogeneous catalysis since they are distinctive templates for the immobilization or deposition of metallic nanoparticles [29]. Single-walled carbon nanotubes (SWCNTs) and multi-walled carbon nanotubes (MWCNTs) are the two classes that broadly fall under CNTs. In theory, rolling single-layered graphene in a particular direction to form a cylindrical shape produces a single-walled carbon nanotube [30]. Noble metals, such as Au, Ag, and Pd, are the most used metal catalysts [31], because they could give CO and formate as a reduction product of CO 2 and also shows a higher selectivity toward it. Considering the availability and cost, non-noble metals, like Cu, Sn, Bi, Pb, In, and their compounds, are usually used as an alternative to the high-cost noble metals [32]. Due to high availability, low toxicity, and less cost, Sn metal is believed to satisfy the needs, and some of the reports on SnO 2 nanowires and nanosheets show an increase in catalytic active sites and therefore an improved catalytic activity [33]. Therefore, a composition of Sn with carbon supports, like MWCNTs, will create a better conduit for CO 2 gas diffusion. Cu-MWCNTs/pCN [34] and g-C 3 N 4 -MWCNTs [35] are a few examples that show how MWCNTs act as support for the photocatalysts to obtain better results.
A semiconductor with the finest catalytic activity must have properties including (1) a larger specific surface area with extensive active sites for the potential absorption of absorbate on the catalytic surface; (2) a nanoscale-range thickness or indeed even a reduced range to sub-nanometer scale for an effective charge-transfer process; and (3) a 2D flexible planar structure with the possibility of making modifications, like heterojunction formation and vacancy introduction. [36]. Carbon-based materials with π-π conjugation can encourage the flow of charge carriers and provide a large surface area for efficient dispersion. The three types of 2D materials that are most frequently utilized in photocatalysis are transition metal dichalcogenides, graphene, and g-C 3 N 4 [37]. Hexagonally packed 2D graphene nanosheets with π-π have exceptional thermal and electrical conductivity and abundant CO 2 adsorption. With a zero-band-gap energy, it could absorb almost the full arc light spectrum. Due to its one-atom-thick structure, it is usually regarded as having the highest specific surface area of any material [38]. Graphene materials with 0D, 2D, 3D, and a combination of all of the three-dimensional graphene-based catalysts have been widely used for photocatalytic CO 2 reduction [6,39]. g-C 3 N 4 is an n-type semiconductor and a member of the wider group of materials known as covalent organic polymers, which also include conjugated porous polymers [40,41], covalent organic frameworks [42,43], and covalent triazine frameworks [44,45]. g-C 3 N 4 has a moderate band gap between 2.7 and 2.8 eV, where it allows for direct photocatalytic CO 2 conversion because of its sufficiently negative conduction band (CB) edge that may convert CO 2 into valuable hydrocarbon [46]. Other than graphene and g-C 3 N 4 , in recent days, graphdiyne, a novel 2D carbon material, has begun to make progress. It is made up of aromatic rings comprising sp-and sp 2 -carbon as well as carbon-carbon triple bonds connected by acetylenic linkages. In order to reach the most stable state, electrons in the high-energy acetylenic links are prone to delocalization, and as a result, they are likely to end up acting as the active centers for adsorption and potential catalysis [47]. The photocatalytic performance of porous 3D porous carbon could possess a greater photocatalytic activity due to its wide surface area, multiple reactive sites, and porous shape. Carbon-carbon networks with a larger surface area along with the porous structure on a nanometer scale highly help CO 2 reduction to obtain C 1 and C 2 products.

Advantages of Utilizing Carbon Materials for CO 2 Reduction
The advantages of using carbon materials for photocatalytic reactions include (1) enlarging the specific surface area, (2) improving CO 2 adsorption and activation, (3) separating photogenerated charges, and (4) increasing light absorption.

Enlarging the Specific Surface Area
Chemical and structural modifications on the surfaces of solid materials possessing high surface areas have been investigated to improve CO 2 adsorption and selectivity [48]. Carbon-based materials, such as graphene, graphene oxides, carbon nanotubes (CNTs), and g-C 3 N 4 , provide greater surface area to the catalysts where it is composited [49]. In the case of an increase in surface area, there will be an increase in the adsorption of CO 2 molecules. Carbon compounds also function as capping agents to control the growth of nanoparticles, increasing the specific surface area of the catalysts. [50]. g-C 3 N 4 has a high proportion of surface defects since it contains a number of hydrogen atoms and possesses electron-rich properties. It has proven helpful in catalysis due to its high value in encouraging electron relocalization on the catalytic surface. Yang et al. designed NiAllayered double hydroxide nanosheets (NALDH) and then combined them with g-C 3 N 4 nanosheets, and witnessed an ultra-tight nanosheet-nanosheet heterojunction. They also added graphene aerogels, which further helped in expanding the structure to a network framework. In this work, they utilized both graphene nanosheets and aerogels to increase the efficiency of the photocatalyst. The formation of ultra-tight sheet-sheet heterojunctions helped in shortening the distance for charge transfer, and it also provided abundant active sites. Graphene aerogel with a 3D structure can act as excellent catalytic support and provide a larger surface area for CO 2 adsorption. The exceptional absorptivity and stability of the catalyst resulted in the production of 4 and 16 times more pure NALDH and bare carbon nitride. A wide range of increased specific surface areas and pore volume was calculated using N 2 adsorption-desorption isotherms, as shown in Figure 3a [51]. CNTs with peculiar structures and a nanometric diameter range have a larger specific surface area [52]. Nitrogen-doped porous carbon nanofibers were recently coupled with a dispersion of nickel and molybdenum phosphide (Mo/Ni-PS@PAN) by Chen et al. They followed a phosphatizing strategy to increase the size of the catalyst and achieved an increased diameter of 20 nm to 50 nm. Carbonization was performed to further increase Molecules 2023, 28, 5383 6 of 15 the size. The combination of nanoparticles of MoP with highly dispersed Ni atoms resulted in a NiMoP@NC PF material. As a result of carbonization, the size of the catalyst increased from 50 nm to 100 nm, as shown in Figure 3b,c. The specific surface area of Ni-MoP@NC PF was 60.5 m 2 g −1 , which was larger than that of Ni-MoP@NC SF (30.7 m 2 g −1 ), showing that NiMoP@NC PF has a much higher amount of mesopores than NiMoP@NC SF . The obtained porous structure of the material helped in promoting the CO 2 adsorption for the photoreduction reaction as shown in Figure 3d [53].
Molecules 2023, 28, x FOR PEER REVIEW 6 of 14 increased diameter of 20 nm to 50 nm. Carbonization was performed to further increase the size. The combination of nanoparticles of MoP with highly dispersed Ni atoms resulted in a NiMoP@NCPF material. As a result of carbonization, the size of the catalyst increased from 50 nm to 100 nm, as shown in Figure 3b,c. The specific surface area of Ni-MoP@NCPF was 60.5 m 2 g − 1 , which was larger than that of Ni-MoP@NCSF (30.7 m 2 g − 1 ), showing that NiMoP@NCPF has a much higher amount of mesopores than NiMoP@NCSF. The obtained porous structure of the material helped in promoting the CO2 adsorption for the photoreduction reaction as shown in Figure 3d [53].

Increasing the CO2 Adsorption and Activation
In addition to the pore structures, the nature of the gas-surface interaction is also crucial for CO2 adsorption. Thus, surface-modified carbons could polarize CO2 molecules and thereby improve their adsorption by introducing basic functional groups into the carbon framework [54]. There have been several porous adsorbents investigated for their potential to capture CO2, such as zeolites, metal-organic frameworks (MOFs), porous carbons, and organic-inorganic hybrid sorbents [55]. Among them, carbon materials with several forms have been used as supporting materials for CO2 photoreduction. Materials based on carbon have great potential as highly efficient materials. The tunable morphology and high surface area make it an efficient adsorbent for CO2 [56]. As we have already seen, g-C3N4 itself could possess a greater surface area. In some cases, graphene materials can be protonated using protonating agents to increase their specific surface area. In recent work, Wu et al. made a novel work in which they have g-C3N4 as a template and fabricated layers of g-C3N4 along with a co-doping of Ni/Co metals in the cavities present in g-C3N4. g-C3N4 material was protonated using phosphoric acid. With the addition of a varied proportion of metal dopings, there was a change in microstructure and differently sized g-C3N4-Co1.6Ni0.4 were obtained, as shown in Figure 4a-c. The addition of bimetallic doping created numerous holes in g-C3N4 and thus increased the nitrogen vacancies. Individual g-C3N4 itself showed a CO production rate of about (3.49 mmol g − 1 h − 1 ), but the addition of metal doping enhanced the CO2 adsorption by creating nitrogen vacancies. Figure 4d depicts the catalytic activity of the catalyst at different proportions. The photocatalytic CO2 reduction of g-C3N4 and g-CN-CoxNiy resulted in CO as the primary product, demonstrating the strong selectivity of CO. This illustrates the advantage of the extended p-p conjugated system and the red shift in the visible edge in the bimetal-doped g-C3N4. Most

Increasing the CO 2 Adsorption and Activation
In addition to the pore structures, the nature of the gas-surface interaction is also crucial for CO 2 adsorption. Thus, surface-modified carbons could polarize CO 2 molecules and thereby improve their adsorption by introducing basic functional groups into the carbon framework [54]. There have been several porous adsorbents investigated for their potential to capture CO 2 , such as zeolites, metal-organic frameworks (MOFs), porous carbons, and organic-inorganic hybrid sorbents [55]. Among them, carbon materials with several forms have been used as supporting materials for CO 2 photoreduction. Materials based on carbon have great potential as highly efficient materials. The tunable morphology and high surface area make it an efficient adsorbent for CO 2 [56]. As we have already seen, g-C 3 N 4 itself could possess a greater surface area. In some cases, graphene materials can be protonated using protonating agents to increase their specific surface area. In recent work, Wu et al. made a novel work in which they have g-C 3 N 4 as a template and fabricated layers of g-C 3 N 4 along with a co-doping of Ni/Co metals in the cavities present in g-C 3 N 4 . g-C 3 N 4 material was protonated using phosphoric acid. With the addition of a varied proportion of metal dopings, there was a change in microstructure and differently sized g-C 3 N 4 -Co 1.6 Ni 0.4 were obtained, as shown in Figure 4a-c. The addition of bimetallic doping created numerous holes in g-C 3 N 4 and thus increased the nitrogen vacancies. Individual g-C 3 N 4 itself showed a CO production rate of about (3.49 mmol g − 1 h − 1 ), but the addition of metal doping enhanced the CO 2 adsorption by creating nitrogen vacancies. Figure 4d depicts the catalytic activity of the catalyst at different proportions. The photocatalytic CO 2 reduction of g-C 3 N 4 and g-CN-Co x Ni y resulted in CO as the primary product, demonstrating the strong selectivity of CO. This illustrates the advantage of the extended p-p conjugated system and the red shift in the visible edge in the bimetaldoped g-C 3 N 4 . Most importantly, the porous structure can offer a lot of sites and a wide surface area for photocatalytic activity, which accelerates the reaction. Thus, this helped with the enhanced CO production rate of 13.5l mmol g − 1 h − 1 , which was 3.9 times higher than that of g-C 3 N 4 , as shown in Figure 4e [57].
importantly, the porous structure can offer a lot of sites and a wide surface area for photocatalytic activity, which accelerates the reaction. Thus, this helped with the enhanced CO production rate of 13.5l mmol g − 1 h − 1 , which was 3.9 times higher than that of g-C3N4, as shown in Figure 4e [57].

Separation of Photogenerated Electron-Hole Pairs
Carbon materials have good conductive properties. The photogenerated electronhole transfer is the second step in photocatalytic CO2 reduction, which takes place after the absorption of light by the catalyst. When light of a variety of wavelengths strikes the catalysts, the electrons in the valence band are excited and migrate to the conduction band. Along the path of the transfer, charge recombination may occur during the electron transfer process, which reduces the number of electrons accessible to participate in the forward redox reaction. To prevent the recombination of charges, hole scavengers are therefore required. The highest capacity for capturing and retaining electrons is found in carbonbased materials, which makes them interesting to be considered [58].
Being the most widely used supporting semiconductor material, g-C3N4 with a larger specific surface area has drawn a lot of attention due to its exceptional merits in expanding the range of visible-light absorption, improving photoexcited charge separation, and offering more active sites for CO2 activation, but it suffers from a major disadvantage of fast charge recombination of charges. A viable photocatalyst should have a larger bandgap that correlates with a higher oxidation/reduction ability. At the same time, a reduced bandgap boosts the light-harvesting capacity and will raise the efficiency of solar energy utilization simultaneously. In order to enhance the efficiency of the photocatalytic system, it is crucial to design a heterojunction between these kinds of single catalysts to help in overcoming these challenges. The creation of heterojunction devices has the potential to be successful due to the separation of photogenerated electron-hole pairs and the combination of the distinctive efficiencies of both catalysts [59]. The charge transfer pathway in heterostructure photocatalysts typically follows either a type II heterojunction or a Zscheme diagram depending on the arrangement of the band-positions and Fermi-levels of

Separation of Photogenerated Electron-Hole Pairs
Carbon materials have good conductive properties. The photogenerated electron-hole transfer is the second step in photocatalytic CO 2 reduction, which takes place after the absorption of light by the catalyst. When light of a variety of wavelengths strikes the catalysts, the electrons in the valence band are excited and migrate to the conduction band. Along the path of the transfer, charge recombination may occur during the electron transfer process, which reduces the number of electrons accessible to participate in the forward redox reaction. To prevent the recombination of charges, hole scavengers are therefore required. The highest capacity for capturing and retaining electrons is found in carbon-based materials, which makes them interesting to be considered [58].
Being the most widely used supporting semiconductor material, g-C 3 N 4 with a larger specific surface area has drawn a lot of attention due to its exceptional merits in expanding the range of visible-light absorption, improving photoexcited charge separation, and offering more active sites for CO 2 activation, but it suffers from a major disadvantage of fast charge recombination of charges. A viable photocatalyst should have a larger bandgap that correlates with a higher oxidation/reduction ability. At the same time, a reduced bandgap boosts the light-harvesting capacity and will raise the efficiency of solar energy utilization simultaneously. In order to enhance the efficiency of the photocatalytic system, it is crucial to design a heterojunction between these kinds of single catalysts to help in overcoming these challenges. The creation of heterojunction devices has the potential to be successful due to the separation of photogenerated electron-hole pairs and the combination of the distinctive efficiencies of both catalysts [59]. The charge transfer pathway in heterostructure photocatalysts typically follows either a type II heterojunction or a Z-scheme diagram depending on the arrangement of the band-positions and Fermi-levels of the heterojunction components [60]. Liu et al. used a morphology-inherited strategy to de-sign a hollow-scheme SnS 2 /g-C 3 N 4 /C (SCC) photocatalyst. The formation of a Z-scheme heterojunction highly helped in increasing the CO 2 adsorption sites and thus increased the production rate. It also increased the reduction ability of electrons in the CB edge of g-C 3 N 4 . The introduction of hollow carbon spheres improved the quantity of light energy utilized, increased the dispersion of SnS 2 /g-C 3 N 4 for more reactive active sites, and also the capacity of photogenerated electrons transfer. Figure 5a shows the electron transport mechanism of SCC, in which the obtained Z-scheme photocatalysts follows a typical type II heterojunction mechanism and the electrons formed on g-C 3 N 4 will be transferred to the CB of SnS 2 . Since the calculated conduction band potential of SnS 2 (−0.46 eV) in this work is higher than the theoretical CO 2 -CO reduction band potential (−0.53 eV), additional energy is needed to push the redox reaction of CO 2 to CO. The Z-scheme formation greatly aids electron-hole separation. Electrons on the CB of SnS 2 are speedily transmitted to the VB of g-C 3 N 4 through the conductive carbon sphere, where they are combined with the photogenerated holes. The internal electric field holds back the hole transfer on the VB of SnS 2 . Therefore, the Z-scheme system greatly helps to separate the electron and holes and enhance the charge flow for CO 2 reduction to CO [61].
Molecules 2023, 28, x FOR PEER REVIEW 8 of 14 the heterojunction components [60]. Liu et al. used a morphology-inherited strategy to design a hollow-scheme SnS2/g-C3N4/C (SCC) photocatalyst. The formation of a Z-scheme heterojunction highly helped in increasing the CO2 adsorption sites and thus increased the production rate. It also increased the reduction ability of electrons in the CB edge of g-C3N4. The introduction of hollow carbon spheres improved the quantity of light energy utilized, increased the dispersion of SnS2/g-C3N4 for more reactive active sites, and also the capacity of photogenerated electrons transfer. Figure 5a shows the electron transport mechanism of SCC, in which the obtained Z-scheme photocatalysts follows a typical type II heterojunction mechanism and the electrons formed on g-C3N4 will be transferred to the CB of SnS2. Since the calculated conduction band potential of SnS2 (−0.46 eV) in this work is higher than the theoretical CO2-CO reduction band potential (−0.53 eV), additional energy is needed to push the redox reaction of CO2 to CO. The Z-scheme formation greatly aids electron-hole separation. Electrons on the CB of SnS2 are speedily transmitted to the VB of g-C3N4 through the conductive carbon sphere, where they are combined with the photogenerated holes. The internal electric field holds back the hole transfer on the VB of SnS2. Therefore, the Z-scheme system greatly helps to separate the electron and holes and enhance the charge flow for CO2 reduction to CO [61]. The increase in the separation of charges decides the lifetime and stability of the material. Wang et al. fabricated a layered g-C3N4/rGO/NiAl-LDHs heterojunction. Reduced graphene oxide (rGO) was used as an electron transfer bridge between the g-C3N4 and NiAl-LDHs and facilitated a rapid charge migration. During the electron transfer mechanism, electrons became excited to both CB edges of Ni3Al-LDHs and g-C3N4. The interfacial internal electric field at the interface and the type II heterojunction facilitated the photogenerated electron transfer from the VB of the g-C3N4 to the VB of the Ni3Al-LDHs. The photogenerated holes accelerated in the CB of g-C3N4, which advanced the electron-hole charge separation, as shown in Figure 5b. The usage of an excellent conductive rGO material resulted in an increased CO production of 14.2 times that of pure g-C3N4 [62].

Increasing Light Absorption
Increased visible-light photocatalytic activity is brought about by the inclusion of carbon materials like MWCNT, graphene, fullerenes, carbon nanoparticles, and g-C3N4 [63]. As we know already, TiO2 is the most used metal oxide catalyst for photocatalysis of its appropriate bandwidth to absorb visible-light photocatalysis [64]. Camarillo et al. used two supercritical methods to prepare CNT/TiO2 and CNT/TiO2/Cu nanocomposites. The addition of CNT greatly aided the synthesized composites to exhibit a greater catalytic The increase in the separation of charges decides the lifetime and stability of the material. Wang et al. fabricated a layered g-C 3 N 4 /rGO/NiAl-LDHs heterojunction. Reduced graphene oxide (rGO) was used as an electron transfer bridge between the g-C 3 N 4 and NiAl-LDHs and facilitated a rapid charge migration. During the electron transfer mechanism, electrons became excited to both CB edges of Ni 3 Al-LDHs and g-C 3 N 4 . The interfacial internal electric field at the interface and the type II heterojunction facilitated the photogenerated electron transfer from the VB of the g-C 3 N 4 to the VB of the Ni 3 Al-LDHs. The photogenerated holes accelerated in the CB of g-C 3 N 4 , which advanced the electronhole charge separation, as shown in Figure 5b. The usage of an excellent conductive rGO material resulted in an increased CO production of 14.2 times that of pure g-C 3 N 4 [62].

Increasing Light Absorption
Increased visible-light photocatalytic activity is brought about by the inclusion of carbon materials like MWCNT, graphene, fullerenes, carbon nanoparticles, and g-C 3 N 4 [63]. As we know already, TiO 2 is the most used metal oxide catalyst for photocatalysis of its appropriate bandwidth to absorb visible-light photocatalysis [64]. Camarillo et al. used two supercritical methods to prepare CNT/TiO 2 and CNT/TiO 2 /Cu nanocomposites. The addition of CNT greatly aided the synthesized composites to exhibit a greater catalytic activity than bare TiO 2 . The majority of CNT/TiO 2 composites have been created through either mechanically combining CNT and TiO 2 or through the use of a coating, sol-gel, and solvothermal techniques. Here, in this work, they used superfluids for preparing the composite. In addition to facilitating the elimination of ligand breakdown products by desorption during decompression, this encourages the transfer of the precursor to the fluid/solid interface at high concentrations. Mere TiO 2 was mentioned as P25, different ratios of CNT and TiOs were taken, and in addition to that, they doped Cu metal over CNT/TiO 2 composite to find out more about the catalytic efficiency with the addition of metal doping.
As depicted in Figure 6, the introduction of CNT made a great impact on narrowing the bandgap of the prepared composites and enhanced the absorption in the visible-light region. The charge transfer mechanism to the nanotubes took place effectively since there was less radiative recombination of photoinduced electrons trapped on the TiO 2 particle surface. Even without the presence of metal doping, CNT50/TiO 2 50 exhibited the best photocatalytic performance among the prepared composites by showing a wide range of absorbance in the visible-light region. While considering CNT/TiO 2 /Cu ternary composites, the effect of metal doping was shielded by CNT. Therefore, an optimal ratio of carbon supporters should be used to obtain maximum productivity [65].
activity than bare TiO2. The majority of CNT/TiO2 composites have been created t either mechanically combining CNT and TiO2 or through the use of a coating, sol-g solvothermal techniques. Here, in this work, they used superfluids for preparing th posite. In addition to facilitating the elimination of ligand breakdown products by tion during decompression, this encourages the transfer of the precursor to the flu interface at high concentrations. Mere TiO2 was mentioned as P25, different ratios and TiOs were taken, and in addition to that, they doped Cu metal over CNT/TiO posite to find out more about the catalytic efficiency with the addition of metal do As depicted in Figure 6, the introduction of CNT made a great impact on nar the bandgap of the prepared composites and enhanced the absorption in the visib region. The charge transfer mechanism to the nanotubes took place effectively sin was less radiative recombination of photoinduced electrons trapped on the TiO2 surface. Even without the presence of metal doping, CNT50/TiO250 exhibited t photocatalytic performance among the prepared composites by showing a wide r absorbance in the visible-light region. While considering CNT/TiO2/Cu ternary sites, the effect of metal doping was shielded by CNT. Therefore, an optimal ratio bon supporters should be used to obtain maximum productivity [65]. Surface modification by defect engineering and nanostructuring are some of th egies that have been followed in recent years to construct a defect-rich photocatal face. Yang et al. created an accordion-like structure in a nitrogen-vacancy defect, in C-doped graphitic carbon-nitride nanosheets (g-CN-X) for the first time utiliz steam technique and ethanol. The enhanced g-CN-10 demonstrated an improved light response over the whole visible-light spectrum. The designing of a defect-ri lytic surface leads to a more localized charge density distribution and promotes catalytic active sites, which increase the light absorption efficiency and impro transport of charge carriers involved in photocatalysis. A graphical model of the l of nitrogen-vacancy defects over g-CN-X with C-doping is shown in Figure 7a. Du gaps between the ultrathin nanosheets in the accordion structure, there was the for of more mesopores and macropores of g-CN-X, which were higher than those of CN, as depicted in Figure 7b. Hence, the accordion-like morphology with a dist Figure 6. UV-vis DRS spectra of P25 and TiO 2 catalysts produced in supercritical media, both in their pure form and as composites. Reproduced with permission from reference [65]. Copyright 2020 Elsevier.
Surface modification by defect engineering and nanostructuring are some of the strategies that have been followed in recent years to construct a defect-rich photocatalyst surface. Yang et al. created an accordion-like structure in a nitrogen-vacancy defect, including C-doped graphitic carbon-nitride nanosheets (g-CN-X) for the first time utilizing the steam technique and ethanol. The enhanced g-CN-10 demonstrated an improved visible-light response over the whole visible-light spectrum. The designing of a defectrich catalytic surface leads to a more localized charge density distribution and promotes photocatalytic active sites, which increase the light absorption efficiency and improve the transport of charge carriers involved in photocatalysis. A graphical model of the location of nitrogen-vacancy defects over g-CN-X with C-doping is shown in Figure 7a. Due to the gaps between the ultrathin nanosheets in the accordion structure, there was the formation of more mesopores and macropores of g-CN-X, which were higher than those of bulk g-CN, as depicted in Figure 7b. Hence, the accordion-like morphology with a distributed hierarchy surface effectively promotes photoconversion efficiency. The energy gap of the g-CN-X (1.82-2.48 eV) was precisely controlled, resulting in an improved light absorption throughout the visible spectrum region, as shown in Figure 7c,d [66]. The efficiency of photo-energy conversion in a non-metal carbon-based nanomaterial can be enhanced by designing a defect-rich catalytic surface, which helps in increasing visible-light absorption for the photocatalytic reduction reaction.
Molecules 2023, 28, x FOR PEER REVIEW 10 of 14 g-CN-X (1.82-2.48 eV) was precisely controlled, resulting in an improved light absorption throughout the visible spectrum region, as shown in Figure 7c,d [66]. The efficiency of photo-energy conversion in a non-metal carbon-based nanomaterial can be enhanced by designing a defect-rich catalytic surface, which helps in increasing visible-light absorption for the photocatalytic reduction reaction.

Conclusions
As seen in Table 1, a wide range of typical carbon materials, like CNT, g-C3N4, conducting polymers, and its other simpler forms like activated carbon, nanofibers, nanosheets, and nanoparticles, have been used with other metal and non-metal nanocomposites for photocatalytic CO2 reduction. These materials with varied dimensions and morphology were used as supporters with photocatalysts for photocatalytic CO2 reduction to increase the selectivity, stability, and reducing ability. Carbonaceous materials could highly help in improving the photocatalytic CO2 reduction mechanism in each of the stages. These are good electron acceptors for aiding in rapid electron transfers. Graphene and rGO materials possess higher surface area, which helps in the adsorption of an adequate quantity of CO2. Since it has a larger surface area to wrap around the photocatalyst, it could avail a uniform distribution of nanoparticles and helps in reducing the size of the nanoparticles. In addition, it is concluded that MWCNs are one of the most used materials as supporting materials for CO2 reduction due to their multi-layered morphology and stacked nanostructured hollow sphere morphologies, which can help the metal doping from corroding.

Conclusions
As seen in Table 1, a wide range of typical carbon materials, like CNT, g-C 3 N 4 , conducting polymers, and its other simpler forms like activated carbon, nanofibers, nanosheets, and nanoparticles, have been used with other metal and non-metal nanocomposites for photocatalytic CO 2 reduction. These materials with varied dimensions and morphology were used as supporters with photocatalysts for photocatalytic CO 2 reduction to increase the selectivity, stability, and reducing ability. Carbonaceous materials could highly help in improving the photocatalytic CO 2 reduction mechanism in each of the stages. These are good electron acceptors for aiding in rapid electron transfers. Graphene and rGO materials possess higher surface area, which helps in the adsorption of an adequate quantity of CO 2. Since it has a larger surface area to wrap around the photocatalyst, it could avail a uniform distribution of nanoparticles and helps in reducing the size of the nanoparticles. In addition, it is concluded that MWCNs are one of the most used materials as supporting materials for CO 2 reduction due to their multi-layered morphology and stacked nanostructured hollow sphere morphologies, which can help the metal doping from corroding. 9 g-C 3 N 4 /TiO 2 /Ti 3 AlC 2 2D/0D/2D composite Enhanced selectivity The 2D Ti 3 AlC 2 MAX acted as a conductive substrate for the 2D/0D S-scheme heterojunction of g-C 3 N 4, g-C 3 N 4 increased the selectivity toward CO 2 to CH 4 [75] 10 Z-scheme FeV 2 O 4 /g-C 3 N 4 Increased charge separation and active sites The ion exchange liquid chemical.The method combined with the post-annealing technique helped in activity and selectivity [76] Carbon materials, like graphene, can absorb almost the whole spectrum of full arc light. If the amount of light absorbed is high, then there will be more production and supply of photogenerated electrons to the redox mechanism. Another advantage is that carbon materials could be used as photosensitizers to increase visible-light absorption. Despite the numerous advantages that carbon materials can offer for photocatalysts to increase efficiency, there are still some disadvantages that exist for practical application, where the photocatalyst's reusability is a crucial parameter that needs to be evaluated. At times, the addition of rGOs substantially decreases the stability of the photocatalysts. To improve the stability, heterostructured CNTs along with nitrogen vacancies could be used to further increase the stability of the catalysts. Since carbon materials are black in color, it absorbs the maximum wavelength of light, which in turn creates a negative photothermal effect, called the shielding effect. An excessive addition of carbon materials could shield the efficiency of metal dopants on the catalytic surface. These fascinating features of these carbon-based materials can be used and evaluated to construct highly selective and more effective photocatalysts for CO 2 reduction application.