A Comprehensive Review on Advances in TiO₂ Nanotube (TNT)-Based Photocatalytic CO₂ Reduction to Value-Added Products

Abstract

The photocatalytic reduction of CO₂ into solar fuels by using semiconductor photocatalysts is one of the most promising approaches in terms of pollution control as well as renewable energy sources. One of the crucial challenges for the 21st century is the development of potential photocatalysts and techniques to improve CO₂ photoreduction efficiency. TiO₂ nanotubes (TNTs) have recently attracted a great deal of research attention for their potential to convert CO₂ into useful compounds. Researchers are concentrating more on CO₂ reduction due to the rising trend in CO₂ emissions and are striving to improve the rate of CO₂ photoreduction by modifying TNTs with the appropriate configuration. In order to portray the potential applications of TNTs, it is imperative to critically evaluate recent developments in synthesis and modification methodologies and their capability to transform CO₂ into value-added chemicals. The current review provides an insightful understanding of TNT production methods, surface modification strategies used to enhance CO₂ photoreduction, and major findings from previous research, thereby revealing research gaps and upcoming challenges. Stability, reusability, and the improved performance of TNT photocatalysts under visible light as well as the selection of optimized modification methods are the identified barriers for CO₂ photoreduction into valuable products. Higher rates of efficacy and product yield can be attained by synthesizing suitable photocatalysts with addressing the limitations of TNTs and designing an optimized photoreactor in terms of the proper utilization of photocatalysts, incident lights, and the partial pressure of reactants.

1. Introduction

Development is required; however, it has to be conducted in a sustainable manner otherwise it will eventually pose a threat to humanity. In recent times, there has been a lot of focus on the imbalanced consequences of carbon dioxide (CO₂) on the ecosystem. In the overall atmospheric system, when the amount of CO₂ increases it creates a notable imbalance in the energy input into the planet, thereby leading to global temperature rise [1,2]. It is widely known that rapid industrialization, exponential growth in the use of fossil fuels, and extensive deforestation are all contributing to a sharp rise in CO₂ emissions [3,4,5]. If current trends continue, CO₂ concentrations will reach 550 ppm by 2050, then a global temperature rise of +2 °C will be reached [3,6]. To mitigate the consequences of global warming, some strategies such as shifting to renewable energy sources, the effective use of energy as well as CO₂ reduction and utilization have to be considered.

Recently, various methods are being used for the conversion and utilization of CO₂. Artificial photosynthesis is a promising and trustworthy method for CO₂ reduction, and it is also a potential technique to achieve value-added compounds by using solar energy [7,8]. This is one of the most appealing alternatives, since it can both benefit from recycling atmospheric CO₂ with direct solar energy usage and meet the demand for renewable fuels [9,10]. TiO₂ has been extensively utilized as a photocatalyst in the photosynthesis process for CO₂ reduction due to its high photoactivity, excellent durability, affordability, and low toxicity [11,12]. Figure 1 depicts the general mechanism of the CO₂ photoreduction process when TiO₂ is used as a photocatalyst. However, practical applications of TiO₂ are greatly hindered by its wide inherent band gap (E_g = 3.2 eV for anatase), quick recombination of photogenerated charges, and low solar light utilization (about 5%) [13,14]. Therefore, the modification of TiO₂ is crucial for practical application.

The photocatalytic performance of TiO₂ can be improved by modifying the surface of TiO₂. Due to their cost-effective construction, promising optical properties, and increased surface-to-volume ratio, TiO₂ nanotubes (TNTs) synthesized by electrochemical anodization have received special attention among the one-dimensional (1D) TiO₂ nanostructured materials [15,16]. In recent years, the use of TNTs for photocatalytic CO₂ reduction has grown tremendously, which can be visualized in Figure 2. Beginning in 2009, TNT photocatalysts have been used to reduce CO₂; since then, there has been a definite upward trend. Recently, remarkable studies are also going on to modify TNTs with different techniques such as doping, sensitization, and heterojunction for improving photocatalytic CO₂ reduction performance under visible light [17,18,19,20,21]. Therefore, a comprehensive overview of TNTs for applicability in reducing CO₂ is required along with new perspectives to facilitate potential directions for future research.

2. Scope and Overview of This Review

CO₂ is regarded as the most powerful and major contributor to global warming among other greenhouse gases. One of the most concerning environmental issues currently facing the world is CO₂ emissions [22,23]. To save the environment, it is indispensable to prevent the source of CO₂ emissions and at the same time convert CO₂ into valuable resources. In this connection, photocatalytic CO₂ reduction to value-added products is a promising approach. With the presence of substantial input energy to break down the C=O bond and suitable photocatalysts, CO₂ can be converted to value-added products such as hydrocarbons (CH₄, C₂H₄, C₂H₆) and oxygenated hydrocarbons (CH₃OH, C₂H₅OH, HCHO, HCOOH, CH₃COOH). During the photoreduction of CO₂, CO is generally observed as the most important intermediate product. Product selectivity is a very challenging task in the overall CO₂ photoreduction process [24,25]. Presently, plenty of research is going on using TiO₂-based photocatalysts to find suitable methods for enhancing the photocatalytic CO₂ reduction rate. Along with original research, some review works are also ongoing to portray the recent scenario to scientific bodies. Although there have been some reviews of photocatalytic CO₂ reduction using TiO₂ to produce CH₄ or other value-added chemicals, none of these reviews was entirely concerned with TNTs [6,26,27]. Consequently, in this review, we focused on photocatalytic CO₂ reduction to value-added products by employing TNTs. A comparative summary of previously reviewed articles on different structures and composites of TiO₂ for applications in diverse fields is presented in

Table 1. Comparative summary of reviewed articles on different structures and composites of TiO₂ for variety of applications.

Review Type	Focused on	Application	Critically Reviewed	Reviewed Literature	Reference
Comprehensive	TiO2 nanotubes	Photocatalytic CO2 reduction to value-added products	Fabrication of TNTs, photocatalytic performance of TNTs, modification methods for improving CO2 photoreduction activity under visible light and kinetics of CO2 photoreduction	2010–2022	This study
Critical	TiO2-based materials	Photocatalytic CO2 reduction	Fundamentals and recent developments of TiO2 photocatalyst, different modification techniques, future challenges for CO2 photoreduction	2005–2017	[28]
Narrative	TiO2-based nanostructures	Photocatalytic CO2 conversion to valuable chemicals	Different structures of TiO2, photocatalytic performance and their favorable reaction conditions	2008–2019	[29]
Narrative	TiO2-based photocatalysts	Conversion of CO2	Preparation and surface modification of TiO2	1997–2018	[6]
Narrative	TiO2 nanoparticles	H2 production	Synthesis and characterization methods for doped-TiO2; influence of dopants for improving photocatalytic performance of TiO2	2005–2019	[30]
Narrative	TiO2-based materials	Solar fuel production	Fabrication strategies and performance of visible light-responsive TiO2-based materials for solar fuel production	2010–2020	[31]
Comprehensive review	TiO2-based photocatalysts	H2 fuel production	Key operating parameters affecting the photocatalytic performance of TiO2-based photocatalysts	2002–2020	[26]
Narrative	TNTs	Dye-sensitized solar cells	Fabrication methods of TNT photoelectrode; modification of TNTs for enhancing power conversion efficiency	2001–2019	[32]
Narrative	TiO2-based composites	Degradation of organic pollutants	Modification methods and performance of doped TiO2-based photocatalysts	2008–2020	[33]
Narrative	TiO2 nanorods	Applications as a photocatalyst and electrode	Synthesis and characterization of TiO2 nanorods; mechanism and photocatalytic activity of TiO2 nanorods	2005–2021	[34]
Narrative	Graphene coupled TiO2 photocatalysts	Environmental applications	Fundamental mechanism, functionalization, and dynamics in TiO2-graphene Nanocomposites	2000–2020	[35]
Perspective review	TiO2-x-based materials	Photocatalytic CO2 reduction	Recent progress in reduced TiO2 catalysts for photocatalytic CO2 reduction performance	2012–2022	[27]
Mini review	TiO2-based Photocatalysts	Contaminant degradation	Different modification techniques; photocatalytic performance of TiO2 for removal of emerging contaminants	2000–2021	[36]

. Recent articles focusing on CO₂ photoreduction based on TNTs or modified TNTs are critically described. The authors believe that this comprehensive overview of the recent advances of TNT-based photocatalytic CO₂ reduction to value-added products would be useful to concerned researchers focusing on nanotechnology, materials science, chemical engineering, chemistry, and environmental science.

3. Fundamentals of TNT

Titanium dioxide (TiO₂) or titania, has become a very recognized and highly studied photocatalyst in recent times due to its relatively long-term photostability, non-toxicity, availability, and photocatalytic activity [26,36]. In natural conditions, TiO₂ exists in four major crystal forms: anatase, rutile, brookite, and TiO₂(B). It is well documented in the literature that TiO₂ in the form of anatase has better photocatalytic performance than the rutile and brookite forms [35,37]. Among these forms, anatase is more advantageous for CO₂ photoreduction at the nanoscale due to its faster charge carrier separation, higher kinetic stability, lower surface energy, and acceptable bandgap energy [38,39]. The crystal structures of TiO₂ nanoparticles can be controlled by utilizing sugar alcohol such as D-sorbitol [40]. To improve the photocatalytic performance of pristine TiO₂, it is tuned to several shapes including 0D (nanoparticles), 1D (nanofibers, nanowires, nanorods, nanotubes), 2D (nanosheets, nanoflakes), and 3D (nanoflowers) (Figure 3). The nanotube shape has a larger surface area than other 1D forms of TiO₂ due to its additional inner walls. Both the inner and outer surfaces of TNTs are susceptible to modification. In addition, photons may have multiple reflections within the TNT hollow structure to facilitate its photocatalytic applications [41,42]. A longer diffusion length of materials leads to improved photoconversion efficiency, and this fact is widely acknowledged [43,44]. Lynch et al. [45] revealed that the electron diffusion length of nanotubes was roughly 30 times longer than that of other nanoparticles. Along with the morphology of the TNT photocatalyst, the composites’ bonding strength also facilitates charge transfer [28]. TiO₂ nanotubes, having the advantages of relatively low toxicity, high chemical and thermal stability, high surface area, and reduced recombination rate, are known to be a desired photocatalyst for CO₂ photoreduction. Even though TNT has been utilized for CO₂ conversion to CH₄ and other value-added compounds, for improving photoreduction performance under visible light it is required to modify TNT.

4. Synthesis of TNT

The one-dimensional (1D) tubular-structured TiO₂ nanotube (TNT) has received pervasive interest due to its unique structural morphology and wide range of applications [16]. TiO₂ nanotubes have a stronger catalytic performance than other 1D nanostructures in the activation- or diffusion-governed photocatalytic process due to their superior characteristics [39]. Sol-gel, template-assisted synthesis, hydrothermal treatment, and electrochemical anodization are frequently used methods for the synthesis of TNTs. The anodization approach, among the other methods, allows for the development of self-organized TiO₂ nanotube arrays with the simplicity of geometry control (length, diameter, and wall thickness) by the use of appropriate anodization parameters [46]. The parameters that have a direct influence on the morphology of TNTs are electrolyte types, concentration, pH, temperature, anodization voltage, time, and type of electrodes [47,48]. The typical preparation of TNTs via the electrochemical anodization process is shown in Figure 4. The following steps are involved in TNTs’ growing process: The initial oxide layer forms on a Ti substrate when a potential voltage is applied; the as-formed oxide layer is attacked by the Fluoride ions to produce TiF₆^2–, which will diffuse into the electrolyte; at the same time, oxidation continues to spread into the substrate; a relatively porous layer with a short anodization time is formed, and the interior of the layer dissolves as the anodization time increases, resulting in well-organized tubes with sufficient length and diameter.

The formation of TNTs by using sol-gel and hydrothermal methods is oriented randomly [49]. However, well-organized and vertically oriented TNTs with an average length can be achieved by electrochemical anodization, which enhances the charge transfer efficiency [50]. The anodization method is also preferable to the template assistant approach, in which the formation of nanotube morphology is limited by the used template geometry [43]. Hence, this method is widely utilized because of its controllable, strong adherent strength, achievability in tailoring the size and form of the NT arrays to the necessary dimensions, and capability to fulfill the requirements of specific applications.

Figure 4. Illustration of electrochemical anodization process used to form TNTs, adopted from [51].

Figure 4. Illustration of electrochemical anodization process used to form TNTs, adopted from [51].

5. Basics about Photocatalytic CO₂ Reduction

Recently, a growing interest has been directed to conducting research in the field of photocatalytic CO₂ reduction, a novel technology for reducing CO₂ emissions while also producing renewable and sustainable fuels. However, CO₂ is a highly stable molecule (∆G° = –400 kJ mol⁻¹) with linear symmetrical configuration, entirely oxidized carbon, and gross carbon-oxygen energy of up to 804.4 kJ mol⁻¹ (at 298 K) [52]. As a result, it is not possible to convert it into value-added chemicals without the need for a photocatalyst or energy input. When exposed to photons with energy over the photocatalyst’s band gap, the excited electrons (e⁻) move from the valence band (VB) to the conduction band (CB), leaving empty holes (h⁺) in the valence band. After that, some of the e⁻/h⁺ pairs go through rapid recombination and do not take part in the subsequent redox processes, which results in the loss of irradiation energy and conversion to some heat. The remaining electrons and holes move to the photocatalyst’s surface to participate in reduction and oxidation reactions. The photocatalytic reduction of CO₂ includes complex photocatalytic reaction mechanisms comprising proton-assisted multiple electron reduction steps with elevated energy barriers, low efficiency, activation and adsorption of CO₂ molecules, and product selectivity [53].

According to the literature, CO₂ photoreduction follows distinct chemical pathways depending on the reductant [54,55]. The formaldehyde and carbene pathways are the most widely acknowledged and confirmed CO₂ to CH₄ conversion processes, as shown in Figure 5. The first step in a reduction pathway is the formation of a ^•CO₂⁻ anion radical by acquiring an electron from the CB of the photocatalyst. In the formaldehyde pathway, there are a number of possible intermediates in the conversion path of CO₂ to CH₄, which is indicated by blue arrows in Figure 5. The carbene pathway, which commences with deoxygenation followed by hydrogenation, is used to convert CO₂ to CH₄. The black arrows in Figure 5 illustrate the carbene pathway for CH₃OH/CH₄ production. There is still a lack of understanding regarding what happens during CO₂ photocatalysis. A thorough investigation of product selectivity is required to fully understand the mechanism of CO₂ reduction pathways [56,57].

6. Photocatalytic CO₂ Reduction to Value-Added Products Solely Using TNT

TiO₂ in the configuration of nanotube arrays is highly desirable for practical application due to the morphological advantages that allow the most appealing diffusion of photocatalytic reduction products and a relatively low recombination rate [58]. The recent studies regarding photocatalytic performance for reducing CO₂ to hydrocarbon fuels using TNTs as a photocatalyst are summarized in

Table 2. Photocatalytic performance of TNTs for CO₂ reduction.

Photocatalyst	Synthesis Method	Reactor	Reactant	Light Source	Product	Yield (µmol g−1 h−1)	Reference
Black TNTA	Electrochemical anodization	Gas-closed circulation system	CO2 + H2O	300 W Xe lamp	CO	185.4	[62]
Flame-annealed TNT	Electrochemical anodization	Stainless steel reactor with transparent quartz window	CO2 + H2O vapor	50 W LED lamp	CH4	156.5	[63]
TNTA	Electrochemical anodization	Batch reactor with Fresnel lens and auto-tracking system	CO2 + H2O	Sunlight	CH4 C2H4 C2H6	861.1 100.8 53.3	[64]
Amorphous TNTA	Electrochemical anodization	Three-way quartz reactor	CO2 + H2O	UV-light bulbs	CH4	14.0	[65]
Self-doped TNT	Electrochemical anodization	Electric-assisted quartz reactor	CO2 + H2O vapor	UV LED lamp	CH4 C2H6	682.3 52.6	[66]

. Depending on the length, diameter, wall thickness, and crystal structure of the nanotubes the photocatalytic activity of TNT photocatalysts is greatly influenced. For instance, Savchuk et al. [59] found that the photocatalytic CO₂ reduction rate of multi-walled TNTs was 2-fold faster for CH₄ and 4-fold faster for CH₃OH production than single-walled TNTs. Oxygen vacancies and defects in the surface play a crucial role in enhancing the photoactivity of photocatalysts [60,61]. The photoactivity of synthesized black TiO₂ nanotube arrays (TNTA) was investigated by Gao et al. [62]. Black TNTA displayed excellent CO₂ photoreduction to CO, yielding 185.39 mol g¹ h⁻¹ under irradiation of visible light. This was made possible by the oxygen vacancy self-doping, which significantly improved three important aspects, namely the development of photoinduced charge, transport of interfacial charge, and interfacial reaction.

Light intensity, reaction temperature, and the partial pressure of the reactant in the photocatalytic reactor have a significant impact on CO₂ photoreduction [67]. The photocatalytic performance of TiO₂ nanotube arrays (TNTA) under concentrated sunlight and different H₂O and CO₂ partial pressure was investigated by Zhang et al. [64]. The results exhibited that the photocatalytic performance of TNTA was 100 times greater than that of pure TiO₂ at 400 sunlight concentrated ratio and 0.05 MPa CO₂ pressure. For improving the performance of TNTs as a stand-alone photocatalyst different techniques have been used by researchers. When evaluating the photocatalytic CO₂ reduction, amorphous TNTA displayed promising photoactivity, providing a CH₄ generation rate of 14.0 without any modification methods [65]. Under visible light irradiation, flame-annealed TNTs synthesized in an aqueous electrolyte medium demonstrated superior photocatalytic performance and produced 156.5 µmol g⁻¹ h⁻¹ of CH₄ [63]. Pan et al. [66] established a unique electric-assisted photocatalytic method to lower the rate of electron and hole recombination. Using this technique, self-doped TNT photocatalyst was produced without using any electrolyte and counter electrode. Self-doped TNTs were able to convert CO₂ at maximum rates of 682.3 µmol g⁻¹ h⁻¹ for CH₄ and 52.6 µmol g⁻¹ h⁻¹ for C₂H₆, respectively. This indicates a 3.5-fold improvement in photocatalytic activity with the electric-assisted technique over conventional photocatalysis.

7. Modified TNT for Improving Photocatalytic Performance

To get over its well-known shortcomings of wide band gap and fast recombination rate, pure TiO₂ material typically undergoes various morphological and surface modifications. However, for improving the photoreduction activity of TNTs under visible light some modification methods are commonly used, as shown in Figure 6.

Table 3. Comparison of various modification methods utilized for improving photocatalytic activity of TNTs.

Modification Methods	Criteria of Utilized Metals or Materials	Benefits	Limitations
Metal Doping	Transition metals	Effective method to expand the optical response of TiO2-based nanocomposites.	Transition metals are expensive and rare as well as there is a possibility of metal leaching into the environment.
Metal-oxide Doping	Alkaline-earth metal oxides	Utilized to improve charge separation, light absorption, and structural properties.	Leads to structural defects.
Non-metal Doping	Mostly N and C	Narrow down the bandgap of TiO2 by introducing new energy states.	Act as recombination centers.
Heterojunction	Semiconductor materials with narrow bandgap	Extends light absorption into visible range and separates reduction and oxidation sites.	This is complex approach with relatively lower stability.
Sensitization	Highly light-sensitive materials	Applied to improve the light response.	Development of photosensitized composite is challenging task.
Carbon-based Materials Doping	Chemically inert, nontoxic, and economically feasible	Extends light absorption into visible range, enhances electron–hole separation and improves CO2 adsorption on catalytic surface.	Inhibits light absorption by TiO2 materials.

provides a summary of the benefits and limitations of the most adopted approaches for enhancing the CO₂ photoreduction of the TNT-based photocatalysts. Details of different modification methods and the performance of modified TNTs are described in the following sections.

7.1. Metal Doping

Metal doping is the most used technique to improve the photoreduction performance of TiO₂-based photocatalysts by introducing different energy levels under the CB and slowing down the fast recombination rate [68,69]. Transition metals, rare earth, and noble metals are the most reported metal dopants utilized to increase the photoactivity of TiO₂ [36,70,71,72]. For instance, Qian et al. [73] employed the transition metal niobium (Nb) with TNTA for enhancing CO₂ photoreduction under simulated solar irradiation. The results showed the production of CH₃CHO was over ∼500 μmol g⁻¹ h⁻¹ with superior reusability and product selectivity. Nguyen et al. [74] examined molybdenum (Mo)-doped TNTs for enhancing CO₂ photoreduction. The incorporation of Mo substantially decreased the oxidation capability of TNTs, while improving photoreduction performance. Higher yields of CH₄ and CO have been reported for Mo-modified TNTs compared to bare TNTs. In another study, Ni and Cu metal was doped in TNT for exploring CO₂ photoreduction activity by Celaya et al. [18]. Doping Cu and Ni with TNT structures increased the density of the carriers, which enhances photoactivity, especially for Ni-doped TNT. In order to increase photoactivity, TiO₂ structures can also be altered using single-atom (SAs) catalysts, where all of the metals remain as isolated single atoms [75,76]. Single atom (SAs–Pt and Au) catalysts were deposited on TNTs by Pan et al. [77] for enhancing CO₂ photoreduction. The Pt-Au/R-TNTs with the addition of 0.33% of SA metals showed a 149-fold higher photocatalytic activity than untreated R-TNT, yielding CH₄ and C₂H₆ of 360.0 and 28.8 μmol g⁻¹ h⁻¹, respectively. The quick shifting of photo-induced electrons from the defective sites to the SAs was made possible by the strong metal support interactions between Pt and Au. This also improved the separation of the electron holes and the transmission of the charge carriers.

With the aid of recent advances in plasmonic photocatalysis, the performance of CO₂ photoreduction under visible light exposure has rapidly increased [78]. It uses precious metal nanoparticles (NPs) spread over photocatalysts and has exceptional qualities including localized plasmonic surface resonance (LSPR), which helps to strengthen the absorption of visible light and excite active charge carriers [79]. The plasmonic effect is the consequence of the interaction between the incident beam of light and free electrons in metal nanoparticles. Additionally, plasmonic metal nanoparticles exhibit distinct optical, electrical, and thermal characteristics [80,81]. The most used plasmonic metals in metal doping with TiO₂ are gold (Au), silver (Ag), platinum (Pt), rhodium (Rh), and ruthenium (Ru) [82]. To enhance the yield of CH₄ through CO₂ photoreduction, Kar et al. [83] synthesized TNT arrays with Au and Ru nanoparticles. Under identical testing settings, the results showed that modified TNTAs produced CH₄ at a rate that was almost ten times greater than that of unmodified TNTAs. Low et al. [15], and Khatun et al. [84] investigated the photocatalytic performance of TNT arrays loaded with Ag and Au, respectively. In both cases, they found improved CO₂ photoreduction performance due to the LSPR behavior of the plasmonic metals.

7.2. Metal-Oxide Doping

A metal-oxide doping technique is also used to increase the photoactivity of pure TiO₂ due to their enhanced light absorption, charge separation, and surface chemistry [85,86]. Separation of charge is achieved at the junction of the metal oxide and TiO₂ material, resulting in a separation of redox processes that inhibits reactions from back and side and increases the production of the product from CO₂ photoreduction [11]. For instance, MgO/TNT films have been constructed by Li et al. [87] for investigating photocatalytic performance under visible light. They reported MgO/TNTs had excellent efficiency for the conversion of CO₂ to CH₄ compared with the TNTs alone due to MgO having a strong adsorption ability of CO₂.

Oxides of copper are the most active dopant compared with other metals used for improving the photoreduction performance of TiO₂ [11]. Shi et al. [88] used TNT-supported CuO-ZnO-CeO₂ (CZC) catalysts for enhancing the CO₂ hydrogenation to CH₃OH. Figure 7a shows that among all catalysts, the catalyst (CZC) without TNT support exhibits the lowest CO₂ conversion. They asserted that adding TNT support to CZC catalysts enhanced CO₂ adsorption and increased the conversion of CO₂ to CH₃OH in addition to promoting CuO reducibility, improving metallic Cu dispersion, and increasing specific surface area. They also suggested a potential bifunctional CO₂ hydrogenation pathway over TNT-based CZC catalysts, as illustrated in Figure 7b. In another study, TNTs were modified with different ratios of Cu₂O/CuO for enhancing the CH₃OH yield from CO₂ photo-reduction [89]. When lactic acid was used as the electrolyte, TNT adjusted with an 80:20 CuO/Cu₂O ratio demonstrated the maximum CO₂ reduction performance. Recently, Goto et al. [90] examined the gas phase complex CO₂ photoreduction mechanism in high vacuum over Cu₂O-doped TNTAs. The positive synergistic effects of Cu₂O nanoparticles, and the appearance of TNA in the strong electron spin resonance signals, indicated the improvement of photoreduction. It is worth mentioning that although CO₂ deoxygenation to CH₄ is often the primary gas phase reaction, they found that CO₂ photoreduction happens by hydrogenation in the gas phase over Cu₂O-TNA photocatalysts even at high vacuum.

7.3. Non-Metal Doping

The photocatalytic activity of TiO₂ nanoparticles under the irradiation of visible light can be accelerated by non-metal doping. The deposition of dopants in the TiO₂ lattice may produce impurity states that impact the band gap, which controls important factors that influence the activity of photocatalysts, such as the charge transfer and recombination rate [91]. Although its usefulness is still controversial, doping with non-metallic anions, particularly nitrogen (N), has received a lot of attention in recent years [92]. For instance, Delavari et al. [93] synthesized N-doped TNT arrays for CO₂ photoreduction in the presence of CH₄. The results showed that the produced TNT arrays were highly organized, vertically aligned, and maintained their structural integrity after doping nitrogen. The photo-reduced products mostly consisted of H₂ with a selectivity of 80.5%, along with a number of additional by-products such as CO, C₂H₂, C₂H₄, and C₃H₈. Parayil et al. [94] examined nitrogen and carbon (C) doping with sodium TNT to enhance CO₂ photoconversion to hydrocarbon fuels. The N, C-TNT sample with a moderate doping concentration exhibited the highest CH₄ yield of 9.75 μmol g⁻¹ h⁻¹. Light absorption, specific surface area, and the concentration of Na⁺ ions in TNT, which act as CO₂ adsorption sites and center the photogenerated electrons, were the main parameters that improved the performance of the modified photocatalyst.

7.4. Heterojunction

The heterojunction technique is becoming a very promising approach for the control of the charge recombination rate as well as for achieving optimal conduction band (CB) and valance band (VB) edge positions for improved photocatalytic activity [95,96]. In particular, the Z-scheme method encourages strong redox facilities that are helpful for the production and propagation of electron holes [97]. The challenge in CO₂ reduction is the appropriate fabrication of Z-scheme between TiO₂ and other semiconductors with higher CB potential. Lai et al. [98] successfully synthesized ZnO–Au-TNT Z-scheme heterojunction with a CH₃OH yield of 7.78 mmol g⁻¹ h⁻¹ for CO₂ reduction. In another study, Wu et al. [99] investigated the CO₂ photoreduction performance of P–O-linked g-C₃N₄-TNT Z-scheme composites. The synergistic effect between g-C₃N₄ and P–O links plays a crucial role in the modified Z-scheme heterostructure and showed a more than 3-fold higher CO₂ photoreduction than bare TNTs. The schematic diagram of the charge transfer and separation mechanism of the structured Z-scheme P–O-connected g-C₃N₄-TNTs is shown in Figure 8. When both TNTs and g-C₃N₄ were excited appropriately, the photogenerated electrons on the CB of the TNTs interact with the holes on the VB of g-C₃N₄. As a result, the energy-rich holes left on the VB of the TNTs and the electrons left on the CB of g-C₃N₄ participate in redox processes, promoting charge separation and enhancing CO₂ photoreduction. The photoreduction performance of the g-C₃N₄-rGO-TNT heterostructure was also examined by Ikreedeegh and Tahir [100]. Using an optimized g-C₃N₄-rGO-TNT combination after 4 h irradiation, they found elevated CH₄ and CO production of 3.32 and 47.12 mmol m⁻², respectively, which represents a significant improvement compared to pristine TNTs.

A TNT photocatalyst synthesized by the hydrothermal method was modified to a CdS(or Bi₂S₃)/TNT heterostructure by Li et al. [101] for enhancing photocatalytic CO₂ reduction activity under the visible spectrum of light. The maximum production of CH₃OH, which was almost 2.2 times more than that of bare TNTs, was achieved on a modified Bi₂S₃/TNT photocatalyst. Kim et al. [19] constructed a TNT-ZnIn₂S₄ type II heterojunction by a simple hydrothermal approach for highly selective CO₂ photoreduction. The results showed a significant improvement with a 4.41 mmol g⁻¹ h⁻¹ yield of CO, which was 1.5 times higher than bare TNTs. The TNT-ZnIn₂S₄ composite extends the responsive spectral range and accelerates the separation of photoexcited electrons during CO₂ conversion.

7.5. Sensitization

The wavelength of light can be expanded by surface photosensitization, which also improves the performance of the excitation process [96]. By utilizing sensitizers, the rate of visible light absorption can also be increased. This approach has been utilized in various studies for CO₂ photoreduction. For instance, Cheng et al. [102] incorporated CdS/ZnS quantum dots as a sensitizer with TNT arrays for enhancing the CO₂ conversion rate under the illumination of visible light. Before the incorporation of sensitizer, the photocatalytic conversion efficiency of TNTA under visible light was too low. Results showed a 2.73 times improvement in CH₃OH yield after sensitizing TNTA with CdS/ZnS quantum dots. Xiao et al. [20,21] successfully synthesized Ni(II) tetra(4-carboxylphenyl)porphyrin (NiTCP) and cobalt tetra(4-carboxyphenyl)porphyrin (CoTCP) with TNT arrays for efficient CO₂ photoreduction to CH₄. Under visible light sources, CoTCP-TNTAs produced 5.5 times more CH₄ than bare TNT arrays, while NiTCP-TNTAs produced 5 times higher CH₄. The photoexcitation of electrons occurs from the highest occupied molecular orbital (HOMO) of the sensitizer to its lowest unoccupied molecular orbital (LUMO). As depicted in Figure 9, photogenerated electrons are transferred from the LUMO of the sensitizer to the CB of the TiO₂, where they are then consumed by the CO₂ reactant.

7.6. Carbon-Based Material Doping

Carbon-based nanostructures such as carbon quantum dots (CQDs), graphene quantum dots (GQDs), carbon nanotubes (CNT), graphene oxide (GO), and reduced graphene oxide (rGO) are often used to enhance the photoactivity of TiO₂-based photocatalysts [103]. Due to strong photoluminescence, high electrical conductivity, wide solar absorption spectrum, and affordable cost, carbon-based materials have drawn the greatest attention to be used as a co-catalyst in enhancing photoactivity [17,104]. To increase CO₂ photoreduction performance under visible light, GO and rGO were used with TNTs as co-catalyst as presented in Table 2. Under visible light irradiation, the graphene oxide (GO) layer links the NTs on their surface, which helps to enhance the mobility of the electrons [105]. The combination of the carbon-based materials along with TNTs acts as a strong photocatalyst to improve CO₂ photoreduction [17,106]. A possible synergetic mechanism to enhance the CO₂ photoreduction over the CQD-modified TNTs is shown in Figure 10. The following are some major ways that CQDs contribute to the improvement of CO₂ photoreduction. The electrons in the CB of the TNTs are transferred to CQDs during the photocatalytic process and quickly react with CO₂, which prevents charge recombination; in the composite long-wavelength visible light is absorbed and converted to short-wavelength by CQDs, which then excites TNTs to increase charge separation; and in the third case, the CQDs serve as a reservoir to trap electrons emitted from TNTs, which prevents charge recombination [17].

As shown in

Table 4. Summary of carbon-based material-doped TNT photocatalysts used for CO₂ photoreduction.

Composite Photocatalysts	TNT Fabrication Method	Photoreactor	Light Source	Product	% of Yield Increased than TNT	Reference
CQDs-TNTs	Hydrothermal method	Electrochemical workstation	300 W Xe lamp	CO CH4	2.4 2.5	[17]
TNT-rGO-Pt	Electrochemical anodization	Circular stainless steel batch reactor	400 W metal-halide lamp	CH4	2.8	[105]
GQD-TNTA	Anodization	Stainless steel reactor	100 W Xe lamp	CH4	5.6	[107]
rGO-Pt-TNTA	Anodization	Gas–solid phase photoreactor	500 W tungsten–halogen lamp	CH4	1.9	[108]
rGO-TNTA	Electrochemical anodization	Stainless steel cell	100 W Xe light	CH4	4.4	[109]
GO-TNT	Electrochemical anodization	Cylindrical quartz reactor	200 W UV-A lamp	CO	2.3	[110]

, substantial improvements have been noticed converting CO₂ into solar fuels when utilizing carbon-based materials as a co-catalyst with TNT. The maximum rate of CO₂ conversion to produce CH₄ was reported by Zubair et al. [107] with the use of GQDs. It has been documented that the mobility of photogenerated charge carriers between the GQDs and TNT crystals plays a crucial role in the photocatalytic performance of the attained composites. A considerable improvement in the production of CH₄ has also been noticed with TNT arrays modified by rGO.

8. Kinetic Modeling to Compute Rate of CO₂ Photoreduction

There is a very limited number of studies focusing on the kinetic model of CO₂ photoreduction. An intrinsic model is highly useful, since it avoids scale dependency and it explains the kinetics of the entire process where all active sites on the photocatalyst surface have an equal possibility to participate in the reaction [111]. Empirical modeling, in situ analysis, and microkinetic approaches are generally used for computing the rate of CO₂ photoreduction [112]. In heterogeneous photocatalysis, the kinetic models are utilized for the well-stable reactants and products in terms of the concentrations of reactant and product on the surface of photocatalysts. In general, the constants of reaction rate and adsorption equilibrium are chosen as kinetic parameters.

There are five key steps that generally occur within the photocatalytic reactor as shown in Figure 11. Step 1 is the initiation of reactants to the reactor, i.e., the diffusion of H₂O and CO₂ to the photocatalyst’s surface; step 2 consists of diffusion along the surface-active sites; step 3 is the provision of the energy for the reaction; step 4 consists of the photoreaction between the two adjacent active sites; while step 5 belongs to the desorption of products from the surface of the photocatalyst [111]. The most crucial factors influencing photocatalytic reduction efficiency are the interactions between the CO₂ molecules and the photocatalyst surfaces. Computational modeling employing density function theory (DFT) simulations has been applied to investigate the CO₂ photoreduction processes occurring at the surface of TiO₂ photocatalysts by Liu et al. [113], and Núñez et al. [114]. Computational fluid dynamics and the Sips model were used by Khalilzadeh and Shariati [115], and Lu et al. [116] to investigate the CO₂ photoreduction rate over TiO₂ photocatalysts. So far, no study regarding computational modeling has been conducted on TNT photocatalysts. When evaluating the assumptions established for empirical kinetic studies, it is also vital to understand the molecular processes appearing at the surface of the photocatalyst. The use of in situ analysis methods is a more promising way to study the CO₂ photoreduction activities appearing on the surface of photocatalysts [111]. Photoluminescence (PL), Fourier Transform Raman Infrared Spectroscopy (FTIR), and X-ray photoelectron spectroscopy (XPS) are examples of in situ spectroscopy that provide vital information on the CO₂ photoreduction process. A kinetic study of CO₂ photoreduction over TNTs using in situ FTIR, UV–vis, and X-ray absorption spectroscopies, has been reported by Liu et al. [113], and Chang et al. [42]. Their findings provided evidence that carboxylate species adsorbed on TNT might carry out surface reactions that are accelerated by UV-vis light to produce low-carbon compounds.

Microkinetics is the study of the fundamental processes taking place at a molecular level on the surface of photocatalysts without taking into account the effects of heat and mass transfer [117]. Microkinetic analysis is a viable technique for propelling CO₂ photoreduction forward. A benefit of employing this technique is that it saves time and money in the lab by producing models that can be used to screen and direct photocatalyst design [114]. The development of scientifically justified Langmuir–Hinshelwood (LH)-based kinetic models provides an alternative approach to microkinetic modeling. This type of model was previously used by Bjelajac et al. [118], and Tan et al. [119] to illustrate the kinetics of TiO₂-based CO₂ photoreduction. The benefit of this method is its simplicity, and it takes into account the effects of partial pressure as well as irradiation on the kinetic data. The time and effort necessary to acquire CO₂ photoreduction kinetic data are a drawback. Reaction rates in non-homogeneous photocatalytic approaches are typically related to the effective desorption of the end products on the photocatalyst surface after reactant adsorption [119,120]. The reaction rate is an indicator of photocatalytic activity represented as the number of moles that have reacted in a given amount of time and space, as evaluated by this approach. The typical assumption is that the surface reactions are the slowest and ultimately the rate-determining step. The L–H mechanism can be utilized for such reactions if indeed the reactants are effectively adsorbed within the same adsorbent surface with distinct rate constants. When CO₂ and H₂O are employed as reactants, and it is assumed that they are adsorbed on the identical active sites just on the catalyst surface, the L–H model can be used to compute the rate of reaction as illustrated in Equation (1).

$$\mathrm{Rate}=({\mathrm{kI}}^{\mathsf{\alpha }})[\frac{K_{C{O}_2}{P}_{C{O}_2{K}_{H_{2}O}{P}_{H_{2}O}}}{{\left(1+K_{C{O}_2}{P}_{C{O}_2}+K_{H_{2}O}{P}_{H_{2}O}+K_{C{H}_4}{P}_{C{H}_4}+K_{CO}{P}_{CO}+ K_{H_2}{P}_{H_2}\right)}^2 }]$$

(1)

Here, k stands for the rate constant of any specific product, and I is the amount of incident light that is used to calculate the kinetic constants. Generally, the rate of a photocatalytic reaction is proportional to I^α, where α denotes the reaction order of light intensity and depending on the intensity of light its value never exceeds one. $K_{C{O}_2}$, $K_{H_{2}O}$, $K_{C{H}_4}$, $K_{CO}$, and $K_{H_2}$ are the ratios of rate constants for the adsorption and desorption of CO₂, H₂O, CH₄, CO, and H₂, respectively. If only reactants are adsorbed on the surface of the photocatalysts and all products are instantly desorbed following chemical reaction, Equation (1) can be simplified. The simplified form of the L–H model will be as Equation (2).

$$\mathrm{Rate}=({\mathrm{kI}}^{\mathsf{\alpha }})[\frac{K_{C{O}_2}{P}_{C{O}_2{K}_{H_{2}O}{P}_{H_{2}O}}}{{\left(1+K_{C{O}_2}{P}_{C{O}_2}+K_{H_{2}O}{P}_{H_{2}O}\right)}^2 }]$$

(2)

By comparing it to the experimental data of $P_{C{O}_2}$, $P_{H_{2}O}$, light intensity, and the rate of hydrocarbon production, the constants of the L–H model can be resolved. Delavari et al. [84] used the L–H model for investigating the CO₂ photoreduction rate over N-doped TNTA. It has been observed that when only a small portion of the sites is utilized the reaction rate is entirely proportional to $P_{C{O}_2}$ at a lower partial pressure. However, with the increase in $P_{C{O}_2}$, the rate eventually slows down and becomes less reliant on it, probably because of CO₂ adsorption on the whole surface of the photocatalyst. Moreover, the rate of reaction is slowed down since CH₄ molecules must compete with CO₂ molecules to commence chemical reactions.

The catalyst surface is frequently employed as a reaction space in heterogeneous catalytic processes [121]. The participation of light in the initiation of certain catalytic processes adds some complexity to photocatalysis. The reaction space cannot be described by only one single parameter because both the quantity of the photocatalyst and the light are essential to measure the rate of reaction at a specific position in the reaction space. The Kinetic Monte Carlo (KMC) simulation model is rapidly becoming the benchmark for bridging the gap between the large diversity of length scales and time scales across which heterogeneous catalysis unfolds [122]. In a nutshell, the kinetics of CO₂ photoreduction are greatly influenced by the source of light, temperature, and pressure in the photoreactor [111]. In both gas and liquid phase systems, partial pressure influences the possibility of adsorption over the photocatalyst, increasing the yield of the photocatalytic CO₂ reduction [28].

9. Concluding Remarks and Future Perspective

TiO₂ nanotubes are promising photocatalysts that have been utilized for CO₂ photoreduction to harvest solar fuels. TiO₂ nanotubes can be produced in a wide range of morphologies utilizing a simple electrochemical anodization process. This review critically assesses the advancement of TNT application on CO₂ photoreduction over the last ten years. The detailed description of kinetic approaches to compute the rate of CO₂ photoreduction has also been summarized in this study. Overall, notable progress has been made in the reduction of CO₂ using bare TNTs or modified TNTs. TiO₂ nanotube photocatalysts utilized by raising the incident light intensity and reaction temperature by using concentrated sunlight have been shown to achieve excellent CO₂ photoreduction to CH₄. In the case of modified TNTs, ZnO–Au–TNT Z-scheme heterojunction showed the maximum CO₂ reduction with a CH₃OH yield of 7.78 mmol g⁻¹ h⁻¹. Among the carbon-based materials, GQD-modified TNTs showed 5.6 times higher CO₂ photoreduction performance than pristine TNTs. Langmuir–Hinshelwood (LH) is the most utilized model to compute the rate of reaction during photocatalytic CO₂ reduction to value-added products.

Despite extensive research in this area, the photocatalytic reduction of CO₂ is still far from practical application due to the low yield of products. Studies have been undertaken so far on a lab scale, and there is still a long way to go before the technologies are taken into consideration for large-scale applications. The control of charge-carrier recombination, performance under visible light, quick charge separation, peeling off TNTs from substrate along with the utilization of the whole surface of the photocatalyst in a photoreactor, and cost-effective and environmentally friendly surface modification methods are all issues that need to be addressed for improving the performance of TNT photocatalysts. The stability and reusability of photocatalysts as well as cost-efficiency and more comprehensive kinetic modeling are crucial for scaling up the CO₂ photoreduction process. However, these have been rarely addressed in previous studies. More research is required on the cost-effective and reusable photocatalyst synthesis method, photoreactor design, kinetic modeling, and pilot-scale implementation.