Photocatalytic CO2 Reduction Using TiO2-Based Photocatalysts and TiO2 Z-Scheme Heterojunction Composites: A Review

Photocatalytic CO2 reduction is a most promising technique to capture CO2 and reduce it to non-fossil fuel and other valuable compounds. Today, we are facing serious environmental issues due to the usage of excessive amounts of non-renewable energy resources. In this aspect, photocatalytic CO2 reduction will provide us with energy-enriched compounds and help to keep our environment clean and healthy. For this purpose, various photocatalysts have been designed to obtain selective products and improve efficiency of the system. Semiconductor materials have received great attention and have showed good performances for CO2 reduction. Titanium dioxide has been widely explored as a photocatalyst for CO2 reduction among the semiconductors due to its suitable electronic/optical properties, availability at low cost, thermal stability, low toxicity, and high photoactivity. Inspired by natural photosynthesis, the artificial Z-scheme of photocatalyst is constructed to provide an easy method to enhance efficiency of CO2 reduction. This review covers literature in this field, particularly the studies about the photocatalytic system, TiO2 Z-scheme heterojunction composites, and use of transition metals for CO2 photoreduction. Lastly, challenges and opportunities are described to open a new era in engineering and attain good performances with semiconductor materials for photocatalytic CO2 reduction.


Introduction
From the beginning of the 20th century, energy resources have been a major concern of human beings due to the Industrial Revolution and advancement in technology. Fossil fuels were the mainstream source of energy in that era. Consumption of fossil fuels on a large scale causes environmental pollution and increases the amount of CO 2 in the environment. The average global temperature has increased by 1 • C in recent years due to global warming which has dangerous effects on our world [1]. CO 2 is one of the major greenhouse gases which is produced from the burning of fossil fuel [2]. Therefore, it is necessary to reduce the emission of CO 2 by decreasing fossil fuel usage and utilizing renewable energy (solar, water, biomass etc.) [3][4][5]. It is very important to develop systems to capture CO 2 from the atmosphere and reduce it to useful hydrocarbon which we use as a fuel instead of fossil fuel. This technique will help to solve the global warming issue and energy shortage simultaneously [6][7][8][9][10]. The energy demand in the world is increasing day by day. A survey is done about the utilization of energy resources in 2008 [2]. The survey is described in pictorial form which is given below (Figure 1) [11]). The sun provides us with an enormous amount of energy; the solar energy we receive in 1 h is 9200 (4.3 × 10 20 J h −1 ) times the energy that was consumed on the earth in 1 h in 2001 (4.7 × 10 16 J h −1 ) [12]. Researchers have conducted a lot of work to propose a mechanism to utilize solar energy. Giacomo Ciamician was an Italian photochemist; in 1912, he proposed the design of solar fuel by implementing the science of photosynthesis [13].
The reduction of CO 2 in the presence of water by photosynthesis produces hydrocarbons and oxygen, which is solar to chemical energy conversion. The reduction of CO 2 gives us many valuable chemicals such as methanol (CH 3 OH), methane (CH 4 ), and methanol (C 2 H 5 OH) which can be used as a non-fossil fuel energy source instead of gasoline and diesel [14][15][16]. Methanol (CH 3 OH) is produced from the reduction of CO 2 and can directly be used as fuel instead of gasoline and diesel [17]. It is not required to change the current energy bass system of the combustion engine or to make major alterations in the engine system of the motor vehicle [18]. It can be directly used as an energy source as an alternative to fossil fuel without making a change in the energy distribution system of the internal combustion engine of the motor vehicle.
In 1979, Fujishima and coworker first reported the reduction of CO 2 by using semiconductor photocatalysts such as TiO 2 and CdS [19]. Willner and his team purposed the photocatalyst reduction of CO 2 to CO by using fac-Re (bpy) (CO) 3 Cl; this is selective reduction of CO 2 to CO which has the highest quantum yield [20,21]. The researcher developed numerous semiconductor photocatalysts which have been used for this purpose such as Ti 2 O, CdS, ZnO, and Bi 2 [22][23][24][25][26][27]. It is a difficult mechanism because it consists of multi-electron steps which lead to various useful products such as CO, CH 4 , higher hydrocarbons, alcohol, aldehydes, and carboxylic acids [28]. So, it is the main concern of researchers to study the selective photoreduction of CO 2 to obtain the selective product.
Many efforts have been made to synthesize and modify different photocatalysts for the photocatalyst reduction of CO 2 but their efficiency is still not enough for practical use on a commercial level. The main reasons are deficiency in the charge transfer channel and the active sites not being sufficient to fulfill the requirement for CO 2 reduction. It is a big task to synthesize such efficient materials for CO 2 photocatalytic reduction which have earned more light to perform better results. To resolve this problem, many techniques have been used to further activate the electronic deposition such as atomic doping and metal deposition. Some emerging concepts and platforms are facilitating the design of cocatalyst-free solar activity of TiO 2 in a cost-effective way, with good quantum efficiency for expanding the TiO 2 's photocatalytic applications [29].
Rutile TiO 2 has a promising application for its effective usage of indoor illumination in photocatalytic environmental purification [30]. Single atom catalysts have gained more attention by achieving outstanding properties for CO 2 photocatalytic reaction. These single atom catalysts can attain good utilization of prepared samples, heighten catalytic activity, and are at an economical price. Many of the single atom catalysts have been synthesized for CO 2 reduction and display excellent results. These catalysts can also offer active sites to increase the CO 2 photocatalytic activity. In this aspect, TiO 2 has great potential due to its unique properties and this can play a vital role in photocatalytic CO 2 reduction [31].
Titanium dioxide has been widely studied in material science and technology due to its functional properties and application. TiO 2 and its composite with other materials such as graphene, g-C 3 N 4, and transition metal oxide have been reported for photocatalytic activity. Historically, TiO 2 has been widely used as pigment for paints [32][33][34] and for other products such as toothpastes [35] and sunscreens [36][37][38]. A historical breakthrough took place with the discovery of photocatalytic water splitting on a TiO 2 electrode illuminated by UV light [39], which brought about extensive interest and efforts toward the scientific and technological research on titania. In the present day, the most important applications of TiO 2 can be roughly classified into two categories, namely environmental and energy applications. For the reasons discussed in the next sections, these applications often rely on modified forms of TiO 2 . MXene, carbon quantum dots, CdS and ZnS, and non-oxide semiconductors have exhibited excellent photocatalytic properties. The most commonly used photocatalysts have been selected from metal oxide semiconductors; some examples of them are represented by TiO 2 , ZnO, SnO 2 , CeO 2 , ZrO 2 , WO 3 , MoO 3 , Fe 2 O 3 , and Fe 2 O 3 . Among all of them, TiO 2 has some advantages when used as a photocatalyst.
It is not expensive, while production on a large scale is cost effective. It has excellent ability to resist corrosion as well as having good photo stability. Many physical and chemical techniques have been well developed already to synthesize the porous film and nanoparticulate powder of TiO 2 . It has exhibited excellent photocatalytic efficiency. It can be activated in visible sunlight and can start a chemical reaction.
Lei Ji et al. report the direct utilization of FeP nanoarray on Ti mesh (FeP NA/TM) as an efficient 3D catalyst electrode to electrochemically convert CO 2 to alcohols [40]. In 0.5 MKHCO 3 , the FeP NA/TM offers a high FECH 3 OH up to 80.2%, with a total FECH 3 OH+C2H 5 OH of 94.3% at −0.20 V vs. reversible hydrogen electrode (RHE). It has shown an extraordinarily high stability during 36 h of continuous electrocatalysis. Anthony Vasileff et al. report the electrocatalyst using some derived model; namely, by using iodide-derived copper (ID-Cu) and oxide-derived copper (OD-Cu) foams, they study the CO 2 to ethane pathway [41].
After significant work has been conducted on the photoreduction of CO 2 , selectivity of the product and conversion efficiency have not improved much. To address these issues, extensive study of CO 2 reduction processes is required, because it depends on different factors [42]. It depends on the composition of the photocatalyst and the reductant and solvent that are used in the reaction system [43].
Electron (e − ) and hole (h + ) pairs are generated during the photocatalytic reaction in the conduction and valance band, respectively, by absorption of photon. These species then move on the surface of material to take part in the oxidation and reduction reaction. Absorption of light, generation, and separation of e − /h + pairs and their performance in catalyzing a reaction are major factors that will define the efficiency. These three steps are explained in Figure 2.
The electronic energy band gap of the semiconductor plays the most important role in the overall process because it controls the extent of light absorption. To induce absorption of light in visible regions, different kinds of techniques have been used such as doping and usage of light sensitizers. To enhance the charge separation efficiency, the cocatalyst and Schottky junction have been used as well as doping of the correctly aligned band structure with the semiconductor employed for this purpose. There are many uses of carbon dioxide on the industrial level. We can directly use CO 2 in food and carbonated drinks and also to increase the recovery of oil. It can also be used in the biological and chemical transformation for manufacturing synthetic fuels and also for the preparation of chemicals and materials. These materials and chemicals are helpful in the manufacturing of polymers, plastics, minerals, and in organic chemistry. Thus, we can say that CO 2 will have more scope on an industrial level in the future. The diagram which details the uses of CO 2 is given below (Figure 3).

Basic Principle of Photocatalysis
Usually, SC materials have been used as photocatalysts; when light with suitable energy falls on the catalyst, the electrons move from the valance band (VB) to the conduction band (CB). There is an energy difference between the valance band and the CB of the photocatalyst which is called the band gap [46]. The charge carriers (e − and h + ) are generated when the light strikes on the surface of photocatalysts with suitable energy greater than the band gap energy of SCs. These electrons and holes play a vital role in the oxidation and reduction reactions to obtain the product [24,47]. In this process, if photogenerated charge carriers (e − and h + ) are not rapidly exhausted, then there is the possibility to recombine and lose their energy. During this process, e − and h + could undergo modification on the surface of photocatalyst material and become recombined. There are two possibilities for this recombination; one is small E BG of catalyst material where electrons and holes may be easily recombined and the other is where these photogenerated particles cannot find the material to react and are recombined due to a very small lifetime [48]. Some basic steps such as: (i) absorbance of light; (ii) production of e − and h + pairs; (iii) transportation of these photogenerated charges to the surface of the material; (iv) adsorption on the catalyst surface; (v) oxidation and reduction reactions; and (vi) formation of desired product occur during this reaction [49]. The schematic diagram of the charge recombination process across the photocatalyst material is given in Figure 4. When light falls on material then it adsorbs according to the band gap energy of SC material and the photocatalytic activity started. This absorbed energy is used in the reductive process for the designed purpose, such as reduction of CO 2 or removal of organic dyes from water. After the absorbance of light energy on the surface of the semiconductor, the electrons move from the valance band to the conduction and due to this, +ve charged particles (h + s) are produced in the valance band of the semiconductor material. This movement of charge carriers can only be possible if the band gap energy is less than the energy of the incoming light falling on the photocatalyst. The catalyst material performance and efficiency are not restricted to the band gap of material with appropriate irradiated light and inhibition (e − and h + ) recombination with different material modification methods. It also depends on many factors such as size of particle, SA, deposition process, and NHE potential of the final product. In future, to obtain better results, effectiveness of these factors and restriction of electron and hole recombination can play a significant role in photocatalytic activity [51]. The diagram of the typical photoreaction process for the production of different renewable fuels is given below in Figure 5. In the photocatalytic process, when light strikes with the material highly active, charge carriers are produced for reaction. In the valance band, the uppermost energy band is filled with electrons, whereas in the conduction band there are no or very few electrons at its ground state. For the reduction reaction it is needed for CB to have more negative than the required potential [53]. In the opposite site the oxidation process occurs due to the conduction band electrons reducing the absorbed species; however, valence band electrons oxidize these species through surface-bound hydroxyl radicals [54]. Normally, these electrons move on the surface of material and combine at the trapping areas then react with species such as CO 2 [55]. If the CR is slower than the transition reaction, then this process can only occur. When the electrons flop to search the trapped places or use those materials which have small band gap energy, then the recombination process happens in the photocatalytic reaction. If recombination occurs, then small amounts of energy are released at the surface of material [49]. A schematic diagram of photo-excitation and electron transformation is given below in Figure 6.  The photocatalytic CO 2 reduction has to face two challenges: one is thermodynamics and the other is kinetics. The CO 2 molecule has large bonding energy which is 750 kJ mol −1 . The CO 2 reduction is required to absorb an appropriate amount of light which is necessary to break the double carbon oxygen bond (C=O). The breaking of the C=O bond is essential to form the C-H bond and this bond is the reason for developing the hydrocarbon chemical. The oxidation and reduction reactions with expectations of CO 2 and water are difficult reactions which require more environmental energy compared to the water splitting reaction which is sometimes the reason for low photocatalytic activity. When we use semiconductor photocatalysts, some problems are faced in the energy band structure which affects the absorption of photons or photocatalytic activity. So there is a need to improve these factors and obtain better efficiency for photocatalytic CO 2 reduction [56]. To reduce these problems, we have synthesized many 2D materials which may be useful for CO 2 reduction. In this aspect, ultrathin 2D materials have become an attractive candidate due to their exclusive properties for photocatalytic CO 2 reduction. Figure 8 reveals the advantages of ultrathin 2D nanomaterials.

Thermodynamics of Photocatalytic Reduction of CO 2
The thermodynamics in photocatalytic activity for CO 2 reduction can be described with some factors such as temperature, light which falls on the photocatalyst, and CB and VB of semiconductor. Commonly, the SCs we use in the photocatalytic reaction have fully filled VB and incompletely filled or empty CB. The energy levels present in the atom have different quantities of photogenerated charge carriers [50]. Some factors which are mostly effective on CO 2 photoreduction are given below.

Effect of Temperature
When light falls on the SC material with energy >E g , then it varies the quantity of electrons and holes in the CB and valance band, respectively. Electrons attain equilibrium within a few seconds in the energy level of CB as compared to band gap energy because relaxation time is small in CB [58,59]. The thermodynamic force is used to start the photocatalytic activity and is in direct relation to the difference of the e − and h + quantity in the CB and VB, respectively [60]. Through experimental work, it is proven that heat cannot induce the photocatalytic phenomenon but the output of CO 2 reduction via the photocatalytic reaction rises with temperature [61]. In this process, temperature helps in the desorption of product from the photocatalyst surface that increases the carbon dioxide reduction yield [62]. Thus, the escalation in the temperature is helpful in the photocatalytic process to increase the rate of reaction.

Effect of Light
In the photocatalytic process, light radiation plays an important role in starting the process because when light falls on the SC material, excitation occurs and electron hole pairs are generated. The chemical potential also helps to elucidate the process because chemical potential of e − is greater in quasi Fermi level than h + potential and is consequently ∆G positive. So, the light radiation gives necessary ∆G and starts the photocatalytic reduction of CO 2 at normal temperature. Thus, it is proven that thermodynamics of this photocatalytic reaction may be explained by the free energy system with the help of light radiation.
When we discuss the energy aspect, then photocatalytic reduction of CO 2 primarily depends on the intensity and L of light. The wavelength of light is important for the excited e − and h + energy because it depends on L of light. The quantity of e − and h + in the conduction band is due to the intensity of falling light [63]. If we consider UV light with high-energy photons, then it can generate e − and h + pairs in the large energy band gap of SC [64]. The band gap and L of light should correspond with one another. The highenergy photon such as for UV light is necessary to initiate the stimulation of large band gap SCs. As compared to large band gap materials, the smaller band gap photocatalysts are activated with the help of visible light (>380 nm) [65]. Similarly, the light with high power intensity can produce a large number of e − and h + pairs when it falls on the surface of the photocatalyst. Thus, it is proven that the efficiency of CO 2 reduction via photocatalytic activity rises with intensity of light and decreases with escalation in the L of light [66]. However, sometimes in the presence of high power intensity light, results are not so good due to rapid recombination of e − and h + in photocatalytic reduction of CO 2 . The band gap of SC materials and the incoming light wavelength are important to each other to obtain better performance [65]. The diagram which details the relation between L of light and the band gap of semiconductor materials is given below in Figure 9.

Effect of CB and VB Potential
In the photoreduction of CO 2 , two types of reactions are involved: one is CO 2 reduction and the other is oxidation of H 2 O. Usually, we consider the thermodynamics of CO 2 reduction; then we also explain it in respect to CB and VB of the photocatalyst. The possibility of thermodynamics for the production of renewable and green fuels can be determined with the help of the position of CB in the photocatalyst material. The position of CB, the VB of various photocatalysts, their band gap energy and reduction potential related with promising green and renewable fuels are described in Figure 10. We take an example of TiO 2 with a valence band and conduction band potential of 2.7 eV and −0.50 eV, respectively, as the photocatalyst for CO 2 reduction. If we discuss this thermodynamically, then some environmentally friendly fuel may be formed which has less (-ve) reduction potential than the conduction band of TiO 2 from carbon dioxide photoreduction such as CH 4 (−0.24 eV) and CH 3 OH (−0.38 eV). Similarly, TiO 2 photocatalyst can oxidize those materials which have less redox potential than its valence band potential [67]. Through results, it is proven that CB, VB, and energy obtained from the light play a significant role in thermodynamics of photocatalytic CO 2 reduction.

Recombination of Charges and Effect of Metal-Modified Surface
The recombination of e − and h + is a big problem in the efficiency of photocatalytic activity of CO 2 reduction. Many research articles have been written on the photocatalysts with metal loading in the photocatalytic CO 2 reduction process. These photocatalysts have been used as charge carrier traps to increase the efficiency of the photocatalytic reaction. Those materials which have this property have the ability to overcome the recombination of e − and h + pairs and can raise the lifetime of these photogenerated charge carriers. This phenomenon is described in Figure 11 which is given below. Many scientists have published papers on the performance of metal-modifying photocatalysts. When Cu/TiO 2 is used as a photocatalyst in the CO 2 reduction process for CH 4 , then it is observed that the layer of copper (Cu) on the surface of the photocatalyst helps in the higher efficiency of photocatalytic activity [68]. Similarly, the composites of Hg and Pt with TiO 2 also showed good performance for formaldehyde [69]. Some other composites of metals such as Pd, Rh, Au, and Ru with TiO 2 have given better performance for the production of methane and acidic acid [70]. In the above given metals, the combination of Pd/TiO 2 was the promising material, and researchers have usually selected this for the production of methane through the CO 2 reduction process [71]. In 1987, Thampi et al. described how rhodium/TiO 2 showed a good performance for photocatalytic CO 2 reduction in the existence of hydrogen [72]. From the above given metal composites, it is observed that when more electrons are available, the efficiency of CO 2 photoreduction for the production of CH 4 , methanol, and formaldehyde remarkably increased. The reason is that when metal atoms make contact with the SC surface, the e − easily comes from the semiconductor (SC) and smoothly spreads on the surface of the material. Moreover, the oxidation process can happen because in metal/TiO 2 composites, holes are easily diffused to the SC surface. In these composites, two things are important for the photocatalytic CO 2 reduction process: one is that metals are uniform and equally distributed over the surface of the photocatalyst. The other is that large amounts of metal loading create problems in the illumination of the catalyst surface due to the reflection and large amounts of photons cannot be absorbed [73].

Synthesis Methods of TiO 2 -Based Photocatalyst
In this section, we discuss various synthesis methods such as hydrothermal, sol-gel, impregnation, one-pot, co-precipitation, and other methods. Figure 12 illustrates some possible synthesis techniques.

Sol-Gel Synthesis
Xia et al. [74] developed the MWCNT and TiO 2 composites by utilizing the sol-gel technique for the reduction of CO 2 with H 2 O. They synthesized the nanocomposites by coating the anatase TiO 2 on the MWCNT. The existence of MWCNT in the synthesized composite can reduce the accumulation of the particles of TiO 2 and move the electron hole pairs developed due to the irradiation of UV light and can mitigate the electron hole pair recombination and consequently enhance the photocatalytic performance of TiO 2 . The synthesized nanocomposite materials lead to the major development of C 2 H 5 OH. It was observed that the MWCNT shows excellent performance for TiO 2 as compared to the activated carbon. The synthesized nanocomposite showed higher performance as compared to the pure TiO 2 in the reusing cycles.
IH Tseng and JCS Wu [75] fabricated the Cu/TiO 2 and Ag/TiO 2 catalysts by using the modified sol-gel approach as shown in Figure 13. They evaluated the reduction of CO 2 and observed that the division of Cu in the particles of TiO 2 is crucial to enhance the production of methanol. A maximum value of methanol production was~1000 µmol g catalyst −1 with the 25 molar% of complete loading of Cu on the layer of the particles of TiO 2 . The photocatalytic performance of the Ag/TiO 2 was lower as compared to the Cu/TiO 2 owing to the vigorous affinity among the clusters of Ag and photoelectrons [75]. Nishimura et al. [76] also fabricated the films of Cr-doped TiO 2 by using the sol-gel approach to improve the CO 2 reduction.

Co-Precipitation Method
Ong et al., [77] successfully fabricated the CNT@Ni/TiO 2 photocatalyst for the reduction of CO 2 into CH 4 under the irradiation of visible light. They synthesized the nanocomposites by utilizing the co-precipitation technique. The results showed that the synthesized nanocomposites have band gap of 2.22 eV, due to which they can absorb a large amount of energy from visible light. The photocatalytic performance revealed that the fabricated composite showed maximum CH 4 production (~0.145 µmol g catalyst −1 h −1 ) as compared with pristine TiO 2 and Ni/TiO 2 . This improvement was due to the collaborative interaction among the CNTs and TiO 2 . Figure 14 illustrates the charge transfer procedure for the reduction of CO 2 .

Impregnation Method
Gui et al. [78] successfully developed the MWCNT@TiO 2 by doping them with metal oxides by utilizing the early wetness impregnation approach. The nanocomposites doped with CuO, CoO, and Fe 2 O 3 played a pivotal role in the expansion of the absorption band into the visible light portion and also facilitated the charge separation. Among the developed nanocomposites, CuO-MWCNT@TiO 2 showed an excellent photocatalytic performance and production of CH 4 was recorded as 0.93 µmol g catalyst −1 at the eighth hour of irradiation time. They also proposed the mechanism of charge transfer for these synthesized metal-oxide-doped MWCNT@TiO 2 nanocomposites for the reduction of CO 2 [78].
Kohno et al. [79] fabricated the nanocomposites of Rh and TiO 2 by using the facile impregnation technique for the reduction of CO 2 with hydrogen. The reaction started in the absence of light but it was improved under irradiation, and was considered as a "photo-enhanced reaction". When Rh in the nanocomposites of Rh/TiO 2 was turned down into a complete metallic condition, the performance was reduced and the major product was converted from CO to CH 4 . The same behavior was also observed by raising the loading amount of Rh.

Hydrothermal Synthesis
Xia et al. [74] fabricated the MWCNT and TiO 2 composites by employing the hydrothermal technique for the reduction of CO 2 with H 2 O. They developed the nanocomposites by the uniform deposition of rutile TiO 2 nanorods on the MWCNT. The existence of MWCNT in the synthesized composite can reduce the accumulation of the particles of TiO 2 and move the electron hole pairs developed due to the irradiation of UV light and can mitigate the electron hole pair recombination and consequently enhance the photocatalytic performance of TiO 2 . The synthesized nanocomposite materials lead to the major development of HCOOH. It was observed that the MWCNT shows excellent performance for TiO 2 as compared to the activated carbon.

One-Pot Synthesis
Gui et al. [80] successfully fabricated the Ag-MWCNT@TiO 2 nanocomposite by using the simple one-pot synthesis approach for the reduction of CO 2 . Different concentrations of Ag were used to observe the effect on the reduction of CO 2 . Among all the samples, the sample having 2% concentration of Ag was considered the ideal sample that showed the maximum reduction of CO 2 into C 2 H 4 and CH 4 with the values of ca. 0.68 µmol g catalyst −1 and 6.34 µmol g catalyst −1 , respectively, as shown in Figure 15.

Other Methods
Gui et al. [81] fabricated the core-shell nanocomposites of MWCNT and TiO 2 for the reduction of CO 2 under the irradiation of visible light. The nanocomposites were synthesized by utilizing the new coating technique. UV-Vis evaluation showed that the synthesized nanocomposites revealed improved photocatalytic performance under visible light. The synthesized nanocomposites showed continuous conversion of CO 2 into CH 4 under the irradiation of low visible light at atmospheric pressure. The maximum production of CH 4 was recorded as ca. 0.17 µmol g catalyst −1 h −1 at the sixth hour of irradiation time [81].

TiO 2 -Based Photocatalysts for CO 2 Reduction
Semiconductors are useful for CO 2 reduction. Titanium dioxide is widely explored for photocatalysts for CO 2 reduction among the semiconductors due to its suitable electronic/optical qualities, availability at low cost, thermal stability, low toxicity, and high photoactivity. Anatase, rutile, and brookite are three different polymorph structures of TiO 2 . Anatase phase has the highest photoconductivity and has a band gap of 3.2 eV [82,83]. The photoreactivity of the rutile phase is lower than the other two phases due to its greater proficiency in recombination of the electron-hole pairs and smaller surface area. However, it has limited efficiency due to its wider band gap of 3.2 eV; after absorption of UV photons, it produces low oxygen, and the charge transport property is relatively poor. As it is active in the UV region, it is thus limited due to UV photosensitivity because 3 to 5 percent of UV light reaches the earth's surface. So, that is the reason that it should be photosensitive under visible light. For TiO 2 , high energy is required for activation but in UV light only 4% of photons have high energy. This is the main reason TiO 2 has limitations for photocatalytic activity. To eliminate these limitations, researchers are working on various modification techniques for the TiO 2 surface, such as dye sensitization, doping, synthesis of composite semiconductors, quantum dots, modification using metal oxides, the formation of nanomaterials, modification by surface photosensitization, and formation of heterostructures [84]. A detailed survey of different TiO 2 -based photocatalysts for CO 2 reduction is summarized in Table 1, where major products produced, reactants, temperature, and pressure for TiO 2 photocatalysts are mentioned.

Doping
The most common method to modify the catalytic properties to tune the band gap of TiO 2 is doping. As metals are electron-rich, they can behave as electron-trapping agents. Efficiency of TiO 2 and charge separation properties can be enhanced and improved by metal doping. At the interface of the metal-semiconductor, a strong electric field is created when TiO 2 is doped. Numbers of charge carriers are increased at the interface by the produced electric field. So, these charge carriers move toward the surface where they can be used in a reduction reaction. Therefore, doping for the reduction of CO 2 is a successful and easy method and improvement of the photocatalytic efficiency of TiO 2 can be achieved. So, we can dope TiO 2 with metals and non-metals. The process of doping can be carried out by using various methods such as the sol-gel, metal-ion implantation [85], co-precipitation method [86], impregnation, and microwave-assisted method [84,87].

Metal-Doped TiO 2 Photocatalysts
When a TiO 2 is doped with metals, then electrical properties can be adjusted or modified by the impurities of metal in the pure structure of TiO 2 . The band gap of pure TiO 2 is 3.2 eV, which is large enough that it is incapable of absorbing high-energy photons. So, the band gap is decreased with the doping of metals. Doping of TiO 2 has been conducted by different researchers with metals, noble metals, rare-earth metals, and transition metals.
Metals such as Ag, Cr, Cu, Rh, Fe, and Pt are commonly used to make metal-doped TiO 2 -based photocatalysts for CO 2 reduction. Enzhou et al. successfully prepared Ag/TiO 2 nanocomposites for reduction of CO 2 via the microwave-assisted method by doping different percentages of Ag. The composite with 2.5% Ag/TiO 2 showed the best performance as a photocatalyst under visible light and UV that was 9.4 times higher and better than pure TiO 2 . It produced methanol that yielded 405.2 µmol/g-cat ( Figure 16) [87]. The pure TiO 2 structure is doped with transition metals to modify its band gap due to which the heteroatom-doped photocatalyst is formed and on the doped metal atom, the photoconversion of CO 2 takes place. This photoconversion occurs when electrons are transferred from the valance band (VB) to the conduction band (CB) and then to the heteroatom of the metal.
At the metal-semiconductor interface, the Schottky barrier is formed, when the n-type semiconductor (TiO 2 ) is doped with the metal, due to which holes cannot migrate from semiconductor to metal. This is because of work function; the work function of metal W m is greater than the work function of the semiconductor (W s ). To overcome this problem, the work function of the semiconductor (W s ) should be greater than the work function of metal [43]. Charge carriers can easily migrate in the case of p-type semiconductors (e.g., Cu 2 O). The choice of dopant metal is crucial to know because it determines the pathway of the photocatalyst reaction and type of the product. When TiO 2 is doped with Pt, it produces methane from CO 2 which has higher yield than methanol. The reason for this is that the work function of Pt is greater than the work function of TiO 2 , due to which photogenerated electrons can easily migrate or be transferred. Among all the metals that are used for CO 2 reduction, copper Cu and silver Ag are the most frequently used metals that showed higher TiO 2 photocatalytic activity and higher selectivity of methanol production in CO 2 reduction.

Non-Metal-Doped TiO 2 Photocatalysts
Doping with non-metals like C, B, S, N, and P, etc. is a good method to modify the TiO 2 photocatalytic system. Nitrogen is the most famous and frequently used candidate by the researchers for doping with TiO 2 . Because of its atomic size that is similar to oxygen, low ionization energy, and high stability, it is used to narrow the band gap. Under visible light, Kumar et al. used the nitrogen doping for CO 2 reduction into methanol. They used nitrogen-doped graphene supported on the copper complex. They reported that during the photocatalytic process, nitrogen helped to reduce recombination. Different methods have been reported by the researchers to synthesis the nitrogen-doped TiO 2 . Figure 17 shows the different methods for the TiO 2 doped with nitrogen. On the other hand, phosphorus is the least used non-metal for the photoreduction of CO 2 with TiO 2 . Wang et al. described the P-doped TiO 2 nanotubes for photoreduction of CO 2 . They reported that high methanol yield was achieved that was up to 860.4 µmolg/cat.

Defect Chemistry of Doped TiO 2
When defects occur in semiconductor-based photocatalysts, their performance can be greatly impacted. There are always defects in semiconductors, which significantly affect the charge separation and transport processes. In semiconductors, localized states that are due to defects can selectively capture approaching charge carriers, resulting in improved photocatalytic activity. Alternatively, these defect states can work as centers of recombination for charge carriers. In addition, the defect states also influence charge transport, altering free carrier acceleration vectors and generating potential barriers. To improve the efficiency and effectiveness of charge separation and photocatalysis, a clear understanding of the effects of defects is required [88]. The photoactivity of the composite catalyst is determined by the crystallinity of the composite catalyst, the crystal phase, and defects. There have been several methods for tuning TiO 2 surfaces by introducing defects into the forbidden gap. Adding cations to either the interstitial or substitutional sites of TiO 2 causes a change in density of states near Fermi levels [89].
Dileep Maarisetty et al. studied the defect chemistry of Ce-doped TiO 2 . In order to enhance the interfacial contact between the oxides of Ce and Ti, they introduced defects on the surface of the composite catalyst by thermal treatment. The rate constant of Ce-doped TiO 2 increased by 40% under solar light by thermal treatment in an inert atmosphere at 900 • C compared to catalysts of the same composition annealed at an alternative heating condition ( Figure 18) [89].

Modification Using Metal Oxides (MO)
Pure TiO 2 is also doped with metal oxides because metal oxides improve the light absorption, surface chemistry, structural properties, and charge separation of the photocatalysts that give the higher photocatalytic activity. Charge separation is achieved at the heterojunction of MO and TiO 2 , due to which the redox reaction is separated which is helpful to obtain more product yield from the CO 2 reduction. Due to the smaller band gap of MO, the light absorption capacity of TiO 2 is improved. Slamet et al. reported methanol product in the photoreduction process of CuO that gave the highest yield. Cu 2 O is very promising and is the most frequently used dopant candidate for the photoreduction of CO 2 . It has a band gap ranging from 2 to 2.2 eV that absorbs visible light effectively.
Nasution et al. also reported the methanol production when TiO 2 is doped with CuO. The synthesis method they used was the improved impregnation method. The light source they used was a black light lamp with UV of 10 W with 415 nm to 700 nm wavelength and 2.45 mW cm 2 . The production of methanol was 442.5 and 19.23 3 µmolh −1 per g-cat.

Formation of Nanomaterials
Nanostructured nanomaterials have various advantages in the photocatalytic process due to their faster diffusion rate and high surface area. The recombination between electrons and holes is inhibited. The surface to volume ratios is high for nanomaterials, so they perform good catalytic activity. Nanostructured nanomaterials have been widely researched due to their advantages. Scientists successfully synthesized the materials that have nanostructure such as nanofibers, nanorods, nanostrings, core-shell, nanowires, and ribbons. Yolk-shell, multi-shell, and hollow shells are also nanostructured nanomaterials that are known as heterostructured nanomaterials. These nanostructures improve the photocatalytic response of the photocatalyst. The abovementioned properties and morphologies of nanomaterials and nanocomposites result in well-defined pathways for electrons and these electrons are interconnected so that their transport rate becomes rapid and wider. These nanostructures help to inhibit recombination and allow separation of electron-hole pairs effectively [84,91].
The spatial electronic configuration of carbon nanotubes and 1D nanowires shows brilliant results. Wang et al. investigated the nanowires of Pt-TiO 2 and stated that enhanced photocatalytic activity and these nanowires have properties like graphene that were more effective for separation of electron hole pairs [92]. Limited works exist that describe the importance of heterostructure nanomaterials such as core-shell structure as reported by

Crystal Phases
TiO 2 has three kinds of crystal phases: anatase, rutile, and brookite. For photocatalytic activity, the first two phases are usually reported. By creating the mixture of its different phases, its photocatalytic performance can be improved [98][99][100]. For this purpose, Chai et al. reported the controlled composition of the anatase and rutile phase in TiO 2 to improve its photoactivity by simply varying the annealing temperature. They found that TiO 2 nanoparticles were composed of both anatase and rutile crystal phase, the heterojunction of which leads to an increase in electron-hole separation. Morphology is another factor that can affect the photoactive ability of TiO 2 . To make sure these hypotheses are clear, 1D hierarchical mesoporous TiO 2 nanofibers were synthesized using a combination of electrospinning and sol-gel methods ( Figure 21) [101].

Modification by Surface Photosensitization
Surface photosensitization expands the useable light wavelength and the efficiency of the excitation process is increased. The absorption rate of visible light can be enhanced by using sensitizers. Sensitizers such as dye sensitizers are ones that are used widely. Figure 22 illustrates the photoreduction scheme of CO 2 by using dye-sensitized TiO 2 photocatalyst into methanol. In this system, dye served as the light harvester and the TiO 2 photocatalysts as electron accepter. The process begins like other photocatalytic systems and photoexcitation occurs. The photoexcitation of electrons occurs from the highest occupied molecular orbital (HOMO) of the dye to its lowest unoccupied molecular orbital (LUMO), which is followed by the photogenerated electron transference from the LUMO of the dye to CB of TiO 2 and then it is used up by CO 2 reactant. Then it is converted into hydrocarbons. The main advantage of this method is the possibility of recycling the oxidized dye through oxidation of H 2 O. The photosensitive dye provides photoelectrons that are in contact with the semiconductor material. At the surface between an electrolyte, semiconductor and dye charge separation phenomenon occurs. The excitation process is increased by the surface photosensitization through chemisorbed or physiosorbed dye that expands the wavelength of useable light through excitation. Numerous research works have used this technique for reduction of CO 2 . However, the efficiency of CO 2 conversion into other hydrocarbons and methane is minimal [102][103][104]. A comparison of the different TiO 2 -based photocatalysts for CO 2 reduction is given in Table 1, where reactants and temperature and pressure, major products produced, and reactor type are presented for TiO 2 photocatalysts.

TiO 2 Z-Scheme Heterojunction Composites for CO 2 Photoreduction
Heterojunction engineering is observed to be a very encouraging technique for suppression of charge recombination as well as for obtaining applicable conduction and valance band edge positions for enhanced photocatalytic activity [124,125]. Particularly, the Zscheme approach, encouraged with the natural photosynthesis process, promotes excellent redox facilities that would be advantageous for holes and radicals generation and propagation [124,126,127]. Fabricating Z-scheme between TiO 2 and other semiconductors having higher potential of the conduction band is considered to be a challenging task for CO 2 . Such that, g-C3N4 CB edge is about −1.23 eV greater than TiO 2 (−0.50 eV) potential energy; however, g-C3N4/TiO 2 Z-scheme composite is synthesized under UV-visible to facilitate charge carrier separation [128][129][130]. For visible light irradiation, WO 3 /TiO 2 Z-scheme heterojunction composites with suppressed charge recombination are considered to be an encouraging combination [131,132]. Raza. A et al. successfully synthesized WO 3 -TiO 2 /Cu 2 ZnSnS 4 Z-scheme heterojunction with CO/CH 4 yield rates of 15.37/1.69 µmol h −1 g −1 for CO 2 production [133].
Moreover, surface engineering of photogenerated electrons and absorbing reactants is very important to enhance the yield rate of products; the fact of the matter is that approximately eight electrons are needed for CO 2 photoreduction to CH 4 [134]. It is well understood that the noble metal Ag supports the surface plasmonic effects, which could be triggered by nanoparticles, and is effective in increasing the catalytic activity [135,136]. In the same way, CO 2 photoreduction conversion into CH 4 and CO, with the yield rate 37.4 and 21.7 µmol g −1 h −1 , respectively, was achieved using (Au/A-TiO 2 )@g-C3N4 Z-scheme heterojunction [137]. The proposed mechanism is shown in Figure 23. The CO 2 photoreduction signifies a promising route for transformation into hydrocarbon fuels using light irradiation. Currently, the productivity of CO 2 photoreduction is significantly poor when it comes to fulfilling practical demands. Furthermore, the production of particular products of CO 2 photoreduction is still a challenge, which will require advanced investigation in upcoming studies.

Challenges and Recommendations
Photocatalytic CO 2 reduction has become an attractive area of research because it provides a green and environmentally friendly process to produce renewable fuels. By using these renewable fuels, we can reduce the use of fossil fuels and increase the potential to utilize low-carbon fuels for sustainable development in the future. The factors which can affect the performance of CO 2 photoreduction are surface affinity, separation of charge carriers, and the ability of the materials to absorb light. Surface affinity suppresses the side reaction, which is important for selectivity and output of the process. If we raise the basic sites on the surface, then we can improve the surface affinity. The capacity of materials for the absorption of light and charge separation capability are related to the quantum yields and product yields [138].
The main challenges in the photocatalytic CO 2 reduction are improvement in e − and h + pair separation, light absorbance efficiency, and increment in the productive yields. So, in the future perspective, many effective and supportive methods are projected to remarkably increase in the photocatalytic activity of CO 2 reduction [139].
The greenhouse gases and the increasing requirement of energy are also large challenges from the photochemical point of view. In this aspect, many appropriate SCs with narrow band gap have been investigated to reduce these problems. The process of photocatalytic CO 2 reduction is not an easy and simple method. There are some complications in the kinetic and thermodynamic process; due to this, CO 2 photoreduction is not yet suitable on an industrial level. During the past few years, many research articles have been written and useful results have been obtained with different suitable photocatalyst materials. There is another problem while using present materials because they have some limitations in photocatalytic CO 2 reduction.
In this aspect, TiO 2 has been researched many times, but its large band gap and its low light absorption capability are hurdles in good performance. To reduce these problems, many methods have been used but have not yet obtained desirable results. Some techniques such as doping and making of heterostuctures have given better results. Many researchers have also investigated and used different methods to increase the efficiency of TiO 2 catalyst. TiO 2 has many properties such as the following: it has good stability, is environmentally friendly, has easy availability, and its reuse ability is better as compared to the other SC materials, but it is less photoactive. So there is a need to find those materials which are more active for light, stable, and suitable for the photocatalytic CO 2 reduction. These materials should be active in the visible light region, and should decrease the recombination rate, improve the charge separation ability, and be suitable for CO 2 photoreduction. In this aspect, some materials such as g-C 3 N 4 , graphene, and GO, etc. have been used for CO 2 reduction but there is need to improve and some modifications are needed for making suitable candidates for photocatalytic CO 2 reduction. Now, many researchers are working on the heterostuctures and these structures may play a significant role in increasing the productivity of CO 2 reduction. Thus, there is a need to find some novel materials and progressive methods to increase the photocatalytic activity of CO 2 reduction [84].
Those materials which are used in CO 2 reduction define rate of reactant (CO 2 ) adsorption, recombination rate, flow of e − and h + , rate of desorption, and in reality the product of this process. Therefore, it is required to improve the present catalyst materials or find some new materials to obtain a better performance. Moreover, the information about reaction mechanism, reaction paths, and kinetic and thermodynamic studies for CO 2 reduction is still not sufficient. The intensity of falling light and absorbance rate are important for the photocatalytic activity of CO 2 reduction. The only way to increase the efficiency of photocatalytic reduction of CO 2 is to understand the mechanism of CO 2 reduction, how recombination occurs, and how to adjust these materials for a better product. Furthermore, more investigations in the field of photochemistry and surface interaction will give a better understanding about actions on the surface of materials and also give better guidelines about the selection of materials. For this purpose, simulations and innovative characterization methods can extend our knowledge to obtain better results.
Many research articles have been written on CO 2 photoreduction and semiconductors such as TiO 2 have become an attractive candidate to obtain creditable results. In spite of recorded attainments many SCs for photoreduction still face some problems such as low thermal and chemical stability, small absorbance of light, and low efficiency under specific circumstances. On the other hand, polymeric materials show attractive properties such as the following: being economical in price, low toxicity, easy availability, light in weight, easy to use and synthesize, and having good flexibility [140]. g-C 3 N 4 is one of the attractive polymeric photocatalysts which showed better results even without composite of noble metals. The polymer materials for photoreduction have many properties such as good absorbance of light in VLR, porosity in structure, and appropriate band gap materials. Many of them have good mobility of charge carriers and long lifetime of excited charges [141]. In polymer materials, the generation of e − and h + pairs is faster than the SC photocatalysts such as in g-C 3 N 4 due to stacked π bonds [142]. These collected π bonds in the polymeric materials are helpful in the fast charge transformation. Thus, due to the unique and distinct properties of polymeric materials, they are suitable candidates for the photocatalytic activity of CO 2 reduction [142,143]. These stacked π bonds can play a vital role in the photoreduction of CO 2 but the study is in the beginning stage for π-conjugated materials [140].
Increasing research into the C-based polymeric materials may be more helpful for the photocatalytic CO 2 reduction as compared to the present SC photocatalysts. Although these promising materials have revealed good results, there is need for some modifications and suitable band gaps to obtain the best results. The design of materials and reaction methods is important but the performance and sustainability of the process for energy production and conversion also have great importance. When we discuss energy production and conversion, the life cycle of the device will also have more importance in future perspectives. So, there is need for more detailed and extensive study about photocatalytic CO 2 reduction reactions using different and advanced tools. These strategies will give a complete and sustainable assessment about photocatalytic activity of CO 2 reduction for the fabrication of renewable fuels and will solve the greenhouse gas effect in future perspectives [144].

Conclusions
The basic cause of global warming is the emission of CO 2 which is escalating day by day due to the burning of fossil fuels in excessive amounts to fulfill the energy demands. The most environmentally friendly and economical technology for sustainable development is photocatalysis. So, the reduction of CO 2 into various environmentally friendly products is a gratifying solution to reduce and utilize the emitted CO 2 . It is a good approach to sustain societies through generation of renewable energy. Titanium dioxide (TiO 2 ) is widely explored for photocatalysts for CO 2 reduction among the semiconductors due to its suitable electronic/optical qualities, availability at low cost, thermal stability, low toxicity, and high photoactivity. In this review article, photoreduction of CO 2 has been discussed by using the TiO 2 -based photocatalysts. A detailed survey of TiO 2 -based photocatalysts, composites with TiO 2 , and metal-and non-metal-loaded TiO 2 -based photocatalysts has been presented. Moreover, the Z-scheme heterojunction composites for CO 2 photoreduction have been studied and it was concluded that they provide an easy way to enhance efficiency of CO 2 reduction. Various suitable and environmentally friendly synthesis methods to prepare photocatalysts for CO 2 reduction into green products have also been studied and discussed in detail. In future, several effective and helpful methodologies are expected to be developed which will increase the efficiency of CO 2 photoreduction and production rate.

Conflicts of Interest:
The authors declare that they have no competing financial interests or personal relationships that could have appeared to influence the work reported in this article.