Recent Advances in the Strategies for Developing and Modifying Photocatalytic Materials for Wastewater Treatment

Abstract

In recent decades, the rising wastewater output from industrial pollution has inflicted severe harm on both surface and groundwater, leading to substantial environmental damage. The elimination of harmful, toxic materials and wastewater remediation are pressing global concerns and pose a formidable challenge for scientists worldwide. Heterogeneous photocatalysis has been recognized as a promising, effective, energy-free, and eco-friendly process capable of completely degrading various organic pollutants. Finding a material that simultaneously satisfies various thermodynamic and kinetic criteria, coupled with good thermal- and photo-stability, is a challenging task necessitating the modification of existing materials or the synthesis of new ones to meet the required standards. This present study comprehensibly elaborates on different approaches to the modification of various photocatalytic systems, both organic and inorganic, in order to obtain more efficient and feasible catalysts for practical applications. In addition, the current status of the application of photocatalysts in dye wastewater treatment is summarized, projecting the future direction for wastewater management by photocatalytic processes.

1. Introduction

Due to rapid industrialization, urbanization, economic development, and unlimited anthropogenic activities in the recent decade, the world is witnessing continuous destruction of natural resources [1]. With this irrepressible development of human society, considerable industrial pollution has been triggered, leading to continuous and careless discharging of hazardous materials into the aquatic environment without adequate treatment, causing severe environmental damage and contamination of both surface and groundwater [2,3]. In the scientific community, red flags were raised when the World Health Organization issued a statement concluding that half of the world’s population will suffer due to the water crisis by 2025 [1,4]. Considering the paramount scale of the ongoing environmental damage, protecting water resources is a pressing concern worldwide, and it is intrinsically linked to climate change. Not only that, the released hazardous materials have an effect on aquatic ecology; these materials also cause air and soil pollution and present many hidden dangers to human health, considering their carcinogenicity, teratogenicity, mutagenicity, and interference with the endocrine system [2,3]. Various industries are direct sources of water pollution, which include the pharmaceutical industries, textile industries, agricultural industries, petrochemical industries, iron and steel industries, fertilizers, metal plating industries, leather and sugar industries, plastic industries, etc. [5]. A challenging problem that arises in this domain is the emission of toxic substances into the environment as a consequence of a variety of industrial products that are released. Most of these pollutants have recently been classified as potentially toxic elements (PTEs) and contaminants of emerging concern (CECs). According to literature, CECs, that are defined as a wide range of organic molecules that are the most challenging pollutants concerning removal techniques due to their different physical and chemical properties [3]. For example, pharmaceutical products, such as antibodies, antiseptics, anti-inflammatory medicines, steroids, and hormones, are directly discharged into the sewage system, and due to their stability, their removal with the use of conventional techniques is impossible. Given these points, antibiotics are usually not completely removed during the standard wastewater treatment process and have the ability to interact with other organic pollutants or heavy metals, forming complex mixtures and contributing to antibiotic-resistant infections [6,7]. Moreover, agricultural products, such as pesticides, insecticides, herbicides, and fertilizers, are adsorbed in the soil and are directly integrated with groundwater, impacting the water quality and causing severe acute and chronic diseases to both human and aquatic lives [8]. According to a World Bank report, approximately 17–20% of industrial pollution originates from the textile industry, leading to extensive environmental contamination [9]. Due to the significant increase in the use of clothing and goods by the modern society, dyes in wastewater are growing explosively worldwide [10]. There are more than 10,000 commercially available dyes that have been widely used in numerous fields, such as dyestuff, textiles, paper, plastics, rubber, cosmetics, leather, etc. [11]. Organic dyes from wastewater released by the industry are detrimental to the environment, as even a small amount of color can be carcinogenic, mutagenic, and disruptive to blood formation. Moreover, dyes are water-soluble organic molecules and stable in the presence of light, heat, and oxidizers. When discharged into an aquatic environment, they alter the habitat’s aesthetic value, obstruct sunlight transmission through water, and increase chemical and biochemical oxygen demand, resulting in reduced photosynthesis. The toxic nature of dyes can be directly attributed to the presence of stable aromatic rings in majority of dyes. Considering their chemical structure, color source, and fiber type, organic dyes can be classified as anionic, cationic, acidic, or basic, whereas based on their functional groups, they can be classified as azo, indigo, phthalocyanine, anthraquinone, sulfur, etc. [2,12,13,14]. Most common textile dyes that also represent the largest group of artificial organic dyes are azo dyes, and they have been investigated over the past decades. Cationic dyes are a class of dyes that contain cationic functional groups, which enable them to dissociate into positively charged ions when dissolved in an aqueous solution, such as methylene blue (MB), rhodamine B (RhB), crystal violet (CV), etc., and as anionic dyes with anionic functional groups, such as sulfonic or carboxylic acid groups such as methyl orange (MO), acid orange 7 (AO7), acid red 14 (AR14), etc. [9]. Specifically, rhodamine B, a class of xanthene, is a water-soluble, reddish violet powder mostly used in the printing, textile, and photographic industries, and it can affect the eyes, skin, and respiratory tract of animals and humans. Methylene blue, associated with thiazine dyes, is a water-soluble green crystalline solid, which contains tertiary amine and is known to be poisonous to water. As the representative of anionic dyes, methyl orange is an acidic and a water-soluble orange-yellow powder with a benzenesulfonate structure, and it is known to be extremely hazardous to the environment when released [12,15]. A challenging problem that arises in this domain is the removal of these versatile dyes. A wide range of different approaches and methods have been available in order to tackle this problem, such as ozonation, membrane filtration, bioadsorption, ion exchange removal, adsorption, photocatalytic degradation, catalytic reduction, biological/aerobic treatment, and coagulation [8,9,10]. Even though various treatments and techniques have been developed over time, each method still remains to be improved, considering their flaws. Most of the methods mentioned have various disadvantages in different application aspects and mostly have the ability of only transferring harmful contaminants from one phase to another without decomposing toxic pollutants to nontoxic substances [16]. Additionally, some of these methods are energy-intensive, have low efficiency, tend to be foul, and they are only effective in removing specific types of pollutants.

Table 1. Conventional approaches to the removal of dyes.

	Advantages	Disadvantages	References
Physical-mechanical
Adsorption Activated carbon	Good efficiency	High cost, expensive regeneration, loss of adsorbents, non-destructive process, Disposal of adsorbents	[17]
Low-cost adsorbents	Low cost	Poor capacity, disposal of adsorbents	[18]
Coagulation/flocculation	Low cost, easy operation, high efficiency	High sludge production	[19]
Sedimentation	Relatively inexpensive, works by gravity and does not require energy	Lack of process control, poor environmental condition for treatment, not effective for removing colloidal solids or dissolved solids	[20]
Ultrafiltration-microfiltration	Low pressure	Unsatisfactory quality of the treated wastewater	[21]
Ion-exchange	Low efficiency for dispersive dyes, regeneration: no adsorbent loss	Concentrated sludge production	[22]
Reverse osmosis	High efficiency in removing hydrolized reactive dyes and salts	High pressure	[23]
Biological	Low cost, slow process	Low biodegradability of dyes	[24]
Chemical
Chlorine	Effectiveness	Dangerous gas	[25]
Ozone	Highest redox potential	Highly unstable, must be generated onsite	[26]
Fenton process	High efficiency, lack of high equipment requirements, availability of reagents, no harmful by-products	Excessive sludge generation, narrow range of operational pH, chemical consumption	[27]
Photocatalysis	Low cost, reusable, complete degradation, eco-friendly	Low photocatalytic efficiency and poor stability, long-scale applicability, light absorption and utilization	[28]

summarizes various approaches to wastewater treatment, highlighting their benefits and drawbacks.

Recently, several advanced oxidation processes have gained wide popularity due to their high removal efficiency of chemical pollutants from wastewater, the most promising being non-toxic and non-polluting heterogeneous photocatalysis, which is well known as a cost-effective and environmentally friendly method [16,29]. Photocatalysis has been proven to be highly suitable for the treatment of dye wastewater and has advantages over the traditional removal techniques, such as the following: (i) prompt time for the complete degradation process (usually a few hours); (ii) mild conditions required and consequent energy conservation with the use of solar light; (iii) complete mineralization of organic pollutants without the formation of toxic secondary products, etc. [2,9,16,30,31]. The process is also effective for the removal of a wide range of contaminants, as it is capable of completely mineralizing complex and degradation-resistant compounds, such as landfill leachate and CECs, into simple compounds such as CO₂ and non-harmful inorganic ions. Additional benefits of the photocatalysis process include its applications across various fields, such as environmental remediation and energy utilization, while requiring minimal energy consumption. Despite the semiconductor photocatalysis technology being a promising self-sustaining technique that harnesses inexhaustible solar energy, the efficiency of photocatalysts in converting solar energy is significantly lower than anticipated. Therefore, the modification of present photocatalytic systems and the development of new ones would be beneficial for achieving sustainable development goals.

A wide range of semiconductors have been employed as photocatalysts for water treatment. Various groups of inorganic compounds, including metal oxides (such as TiO₂, ZnO, ZrO₂, SnO₂) and metal chalcogenides (such as CdS, CdSe, CdTe, PbS, InP), have been used to remove harmful organic compounds. Metal oxides like ZnO and TiO₂ have suitable conduction band (CB) and valence band (VB) edges for reduction and oxidation reactions. However, their wide energy band gap (Eg > 3.2 eV) and high electron-hole (e⁻/h⁺) recombination rates limit their efficiency in visible light harvesting. In contrast, most metal sulfides possess narrow band gaps and adequately negative conduction band potentials, making them suitable for reduction processes, such as solar-to-fuel conversion. However, the less positive VB position of these materials presents a significant barrier to producing hydroxyl (^•OH) radicals. Despite the high visible light harvesting efficiency (Eg = 0.41 eV), the CB and VB potentials of PbS are insufficient for hydrogen evolution and ^•OH formation from water reduction and oxidation, respectively. Photocorrosion and a low quantum efficiency are additional, significant drawbacks that limit their effectiveness as photocatalysts.

Figure 1 illustrates the band positions of commonly used semiconductors and the redox potentials versus the Normal Hydrogen Electrode (NHE) for hydroxyl radical formation, hydrogen evolution, and CO₂ reduction at a pH of 7. The alignment of band gap energy with the redox potentials of photo-induced reactions determines the types of reactive species that can be generated.

Finding a material that meets all thermodynamic and kinetic criteria simultaneously is challenging. Therefore, it is often necessary to modify existing materials or develop new ones to achieve the desired properties.

In general, effective strategies designed to tackle the current challenges and develop an optimized photocatalyst can be categorized into three main areas: structural engineering, compositional enhancement, and heterojunction design. Precisely designing the morphology, employing facet engineering, and optimizing the textural properties can significantly boost photocatalytic efficiency. For instance, increasing the specific surface area and creating interconnected porous networks can improve the diffusion of reactants and products, improve the adsorption of organic pollutants, and enhance photon flux densities. Enhancing the material composition through metal and non-metal doping has proven to be an effective strategy for altering the optical and electrical properties of semiconductors by adjusting the band gap structure of photocatalysts. This approach can improve light absorption, facilitate electron transfer, and ultimately increase photo-efficiency. Moreover, the optimal concentration of doped metals can help reduce charge separation. Coupling two semiconductors with different energy band alignments is another effective, yet challenging, method to suppress charge recombination and extend absorption into the visible light region, thereby further enhancing photo-efficiency. The formation of a thin layer of photocatalyst on a support appears to be a practical approach to reduce the agglomeration of nanoparticles and facilitate their recovery and separation from reaction mixtures. Thin films can be employed in various setups, from simple batch reactors to more complex continuous flow systems, depending on the application and the desired outcome.

Based on the above content and inspired by the current research trend, this review article summarizes the current state-of-the-art strategies for designing and optimizing photocatalytic materials to improve their stability and enhance their application in dye wastewater treatment. Particular attention is devoted to unraveling the effects of structural and morphological engineering, band gap energy tuning, facet engineering, and heterojunction design, aiming to modulate charge separation and enhance light harvesting efficiency. Finally, the review provides a comprehensive analysis of photocatalytic performance in the treatment of organic dyes from textile wastewater.

2. Principles of Photocatalytic Oxidation

Typical heterogeneous photocatalytic reactions consist of various oxidation and reduction reactions occurring on the photocatalyst surface. The process of photocatalytic pollutant removal can be described in several steps [30,33,34,35]:

• Adsorption of the pollutant from the surrounding environment onto the photocatalyst surface;
• Absorption of light with energy greater than the band gap of the photocatalyst, generating e⁻/h⁺ pairs in the bulk phase;
• Separation of e⁻/h⁺ pairs and their diffusion to the photocatalyst surface. Some photogenerated carriers simultaneously recombine, both on the surface and within the photocatalyst;
• Photo-oxidation and photoreduction reactions between the trapped electrons and holes with adsorbents on catalytic active sites. Specifically, holes in the valence band (VB) oxidize H₂O molecules on the photocatalyst surface to ^•OH, while electrons in the conduction band (CB) reduce O₂ adsorbed on the surface to superoxide radicals (^•O²⁻). This process simultaneously degrades pollutants into smaller molecules such as H₂O and CO₂;
• Desorption of reaction products from the interface into the bulk solution, allowing for a new photoreaction cycle to begin.

The photocatalytic processes are governed by both kinetic and thermodynamic criteria [36]. Thermodynamically, the redox potential of the redox couple must be higher (less negative) than the conduction band potential and lower (less positive) than the valence band potential of the photocatalyst to enable the reactions [30,34]. Apart from thermodynamic criteria, kinetically, several barriers must be overcome, including light absorption, charge carriers’ separation, and the adsorption and activation of pollutants and water molecules [37]. While solar light harvesting and charge separation/migration are key characteristics of a photocatalyst, the surface interactions of charge carriers with pollutants primarily differentiate among various photocatalytic systems. Since most of the photogenerated e⁻/h⁺ pairs (approximately 90%) recombine rapidly after excitation [34], this has been identified as a hinderance to the application of this promising technology. Consequently, its improvement has been the subject of much research [38]. Additionally, properties such as non-toxicity, stability, a low production cost, and high recyclability are crucial for real-time applications. However, finding materials that satisfy all these criteria simultaneously is challenging, and compromises often need to be made to balance these properties effectively.

Photocatalytic Mechanism

A photocatalytic reaction occurs when photocatalysts absorb light and convert the absorbed electromagnetic radiation into chemical energy, generating negative electron and positive hole pairs that accelerate the chemical reaction rate [2,39]. Upon irradiation with photons, electrons from the valence band are excited and move to the conduction band, e_CB⁻, while positive holes remain in the valence band, h_VB⁺, as shown in Equation (1) and Figure 2 [2,8,27,40,41].

Photocatalyst + hν (UV or VIS) → (e_CB⁻ + h_VB⁺)

(1)

Figure 2. Mechanisms of photocatalytic degradation of organic pollutants (reprinted with permission from Ref. [28]).

Figure 2. Mechanisms of photocatalytic degradation of organic pollutants (reprinted with permission from Ref. [28]).

The photogenerated charge carriers can recombine and easily release heat in 10 and 100 ns [2,40]. The goal of many studies addressing this challenging aspect of photocatalysis is to impede the recombination of reactive species. A significant reduction of recombination and an enhanced separation of charge carriers can be achieved by adding scavengers or incorporating trap sites on the photocatalyst surface (surface defects, surface adsorbents, or creation of other surface sites) [40]. The remaining photogenerated electrons and holes migrate to the photocatalyst surface and react with other species adsorbed on the surface [41]. Positive holes readily react with surface water, producing highly reactive hydroxyl radicals, ^•OH, which non-selectively oxidize various organic molecules, and hydrogen ions, H⁺ (Equation (2)) [33,39,40,41].

h_VB⁺ + H₂O → H⁺ + ^•OH

(2)

At the same time, due to their strong reduction capacity, photogenerated electrons react with oxygen, O₂, or hydrogen ions, H⁺, adsorbed on the photocatalyst surface, creating a superoxide radical anion, ^•O₂⁻ (Equation (3)), or hydrogen peroxide, H₂O₂ (Equation (4)) [33,40,41].

e_CB⁻ + O₂ → ^•O₂⁻

(3)

e_CB⁻ + (O₂, H⁺) → H₂O₂

(4)

These reactions prevent the recombination of e⁻/h⁺ pairs, creating highly reactive radicals, ^•OH and ^•O₂⁻, that react with pollutants, forming other reactive species [29]. Additionally, the resulting reactive species can then further react, forming different radicals (Equations (5)–(9)) [39,40]. Additionally, the electron, e_CB⁻, can react with hydrogen peroxide, H₂O₂, producing hydroxyl radicals (Equation (5)).

H₂O₂ + e_CB⁻ → ^•OH

(5)

A superoxide radical, ^•O₂⁻, can react with hydrogen ion, H⁺, or water, generating a hydroperoxide radical, ^•OH₂, and hydroxyl ion (Equations (6) and (7)):

^•O₂⁻ + H⁺ → ^•OH₂

(6)

^•O₂⁻ + H₂O → ^•OH₂ + OH⁻

(7)

Intermediates, such as a hydroperoxide radical, ^•OH₂, and hydrogen peroxide, H₂O₂, subsequently produce highly reactive hydroxyl radicals (Equations (8) and (9))

^•OH₂ + H₂O → H₂O₂ + ^•OH

(8)

H₂O₂ → 2 ^•OH

(9)

Lastly, either positive holes oxidize the organic pollutants directly or other highly reactive oxygen species, such as H₂O₂, ^•OH, or ^•O₂⁻, decompose organic pollutants and totally mineralize them, converting organic molecules to CO₂, H₂O, and other products, as in Equation (10) [27,28].

(^•OH, ^•O₂⁻, H₂O₂, h_VB⁺) + Organic pollutants → CO₂ + H₂O + other products

(10)

Besides the presented reactions, photocatalytic treatment can also include antimicrobial processes and disinfection [35,40]. Highly reactive oxygen species (^•OH, ^•O₂⁻, H₂O₂, h_VB⁺) are able to apply different oxidative stresses, causing bacterial cell damage; disturbing ribosomes, proteins, and genomic materials (DNA, RNA) in cells; and impairing the electron transport chain system and the peptidoglycan layer (Equation (11)). These highly reactive oxygen species damage cell membranes by modifying cell permeability, causing the leakage of cell cytoplasmic content. Moreover, these reactive oxygen species can hinder some cell protein activities, damaging normal cell function. Bacterial damage can also be achieved by the termination of organic covalent bonds (H-O, C-O, C-H, and C-C present in biomolecules like DNA, amino acids, proteins, nucleic acids, and carbohydrates) by hydroxyl radicals [35,40].

ROS (^•OH, ^•O₂⁻, H₂O₂, e_CB⁻, h_VB⁺) + Bacterial cell → Cell damage

(11)

At the end of the photocatalytic process, highly unstable reactive e⁻/h⁺ pairs recombine, releasing heat (Equation (12)) [39].

Photocatalyst + hν → Photocatalyst + e_CB⁻+ h_VB⁺→ Photocatalyst + heat

(12)

3. Photocatalytic Materials

Various types of materials have been studied and used as photocatalysts, with semiconductors being the most commonly utilized due to their unique electronic structures [6,16,33,40,42]. Semiconductors have a valence band occupied by electrons and a conduction band unoccupied by electrons, separated by a band gap. They also exhibit other favorable properties, such as chemical and biological stability, non-toxicity, and an effective response to solar radiation [33,40]. Nano-sized semiconductors (TiO₂, ZnO, SnO₂, WO₃, CeO₂, ZrO₂, MoO₃, ZnS, Fe₂O₃, SiO₂) have been successfully employed for the degradation of organic pollutants, microbial inactivation, water splitting, sensors, and solar cells [33,40]. In addition to conventional semiconductors, recent studies have explored materials such as carbon nitride (g-C₃N₄) metal–organic framework compounds, graphene-based photocatalysts, and BiOCl and black phosphorus as photocatalysts [6]. The following section provides a concise overview of the most commonly used photocatalysts, highlighting their advantageous properties, limitations, and applications.

3.1. Metal Oxides

3.1.1. TiO₂

Among various semiconductor metal oxide photocatalysts, titanium dioxide (TiO₂), especially in its anatase phase, has received significant interest and has been the most extensively investigated. Its outstanding properties can be assigned to its large surface area and shape-dependent electronic, optical, and catalytic properties, durability, low cost, low toxicity, chemical and photochemical stability, and superhydrophilicity [9,11,33,34,40,42]. TiO₂ has demonstrated significant activity in the degradation of organic pollutants (such as dyes, phenolic compounds, pesticides, herbicides, benzenes, furans, chlorinated alkanes, and dioxins) as well as inorganic contaminants (including toxic metal ions like Pt²⁺, Rh³⁺, and Cr⁶⁺). Additionally, TiO₂ exhibits excellent antibacterial and antiviral activities against various bacteria and viruses [34,40,42]. However, TiO₂ has drawbacks such as a photocatalyst, including limited absorption of visible light due to its wide band gap (3.2 eV for anatase); rapid recombination of photogenerated e⁻/h⁺ pairs; reduced adsorption of hydrophobic organic pollutants on photocatalysts; aggregation of nanoparticles during photocatalytic degradation; and difficult recovery of nanosized particles from the treated wastewater [11,34,40,42]. To improve its photocatalytic properties, researchers have focused on structural and morphological engineering, as well as crystal-facet engineering. For example, optimizing the {101} (reducing sites) and {001} (oxidizing sites) facets of TiO₂ through careful adjustments of synthesis conditions can increase photocatalytic activity [43]. This optimization enables faster electron transfer, reduces inter-crystalline contacts, and shifts the band gap toward 3.1 eV.

3.1.2. ZnO

ZnO is another widely used photocatalyst due to its non-toxicity and insolubility in water, excellent piezoelectric effect, high electron mobility, high photosensitivity, environmentally friendly nature, and relatively low cost [9,33,34,40,42]. However, the main drawbacks of ZnO include a wide band gap (3.37 eV) and susceptibility to photocorrosion. Photocorrosion occurs in aqueous solutions under UV irradiation due to dissolution caused by the reaction between valence band holes and surface oxygen. Additionally, ZnO operates within a limited pH range since it dissolves in strong acids and alkalis [34]. To improve ZnO’s photocatalytic performance and address its drawbacks, strategies such as doping with metal or non-metal ions can be employed to narrow the band gap and enhance visible-light absorption. Additionally, surface modifications and the creation of heterostructures with other semiconductors can help reduce photocorrosion and extend ZnO’s operational pH range, thereby increasing its overall stability and efficiency [35,44].

3.1.3. WO_x

WO_x is an n-type semiconductor with a narrow band gap (2.8 eV) that absorbs electromagnetic irradiation effectively in the blue region of the visible solar spectrum [33,40,42]. WO₃ is known for its inertness to chemical action, fine metal interaction, and mechanical strength [21,30]. Despite its potential for photodegrading organic pollutants and its chemical stability at various pH levels, WO_x has been less studied than TiO₂ and ZnO [33,42]. The lack of published studies on WO_x can be attributed to several proposed key reasons. One reason could be the historical focus and extensive research already carried out on TiO₂ and ZnO, and their already recognized effectiveness and versatility as photocatalysts. Additionally, WO_x may present challenges such as synthesis complexity, cost, and scalability issues when compared to the mentioned TiO₂ and ZnO. Furthermore, its photocatalytic efficiency under visible light, although promising, may not be as well-documented or understood as that of TiO₂ and ZnO, leading researchers to prioritize more established materials.

3.1.4. Cu-Based Oxides

CuO exhibits unique catalytic, mechanical, electrical, and optical properties and can be used as an electrode material, for field emission, as gas sensors, as a catalyst support, as a semiconductor, as batteries, and as a catalyst [33]. Cu₂O, however, is thermodynamically unstable in aqueous solutions and tends to reduce to Cu. Therefore, a protective layer on a Cu₂O electrode is necessary [34]. By varying transition metals and the Cu(I) coordination environment in ternary Cu-based metal oxides, the band gap can be adjusted across a wide range within the sunlight spectrum, from approximately 1.2 eV to greater than 3.0 eV [34].

CuO exhibits unique catalytic, mechanical, electrical, and optical properties and can be used as an electrode material, for field emission, as gas sensors, as a catalyst support, as a semiconductor, as batteries, and as a catalyst [33]. Cu₂O, however, is thermodynamically unstable in aqueous solutions and tends to reduce to Cu. Therefore, a protective layer on a Cu₂O electrode is necessary [34]. By varying transition metals and the Cu(I) coordination environment in ternary Cu-based metal oxides, the band gap can be adjusted across a wide range within the sunlight spectrum, from approximately 1.2 eV to greater than 3.0 eV [34].

3.1.5. SnO₂

SnO₂ is a semiconductor with a wide energy band gap (3.6 eV) and cannot be used independently as a photocatalyst since it cannot reduce oxygen molecules due to its conduction band position [33,40,42]. However, more positive conduction band edge makes SnO₂ a better electron acceptor, allowing for it to be coupled with other semiconductors to form heterostructures with efficient charge separation and improved photocatalytic properties [28,30]. SnO₂ also exhibits high stability and tunable optoelectronic and optical properties [21].

3.1.6. Bi₂O₃

Bi₂O₃ is a semiconductor with the bandgap varying from 2.09 eV to 2.9 eV, representing a visible-light responsive photocatalyst suitable for the degradation of various organic contaminants [45,46]. Bismuth trioxide has many polymorphs, such as α-Bi₂O₃ (monoclinic), β-Bi₂O₃ (tetragonal), γ-Bi₂O₃ (body-centered cubic), δ-Bi₂O₃ (face-centered cubic), ω-Bi₂O₃ (triclinic), and ε-Bi₂O₃ (tetragonal) [45,46]. The polymorphs have different photocatalytic activity; for instance, β-Bi₂O₃ has higher photocatalytic activity for the organic dye degradation than α-Bi₂O₃, whereas γ-Bi₂O₃ showed higher photocatalytic activity in Rhodamine B degradation than α-Bi₂O₃ [46]. Nevertheless, the recombination of photogenerated electrons and holes in single-component Bi₂O₃ photocatalysts lowers the photocatalytic performance, and these drawbacks can be removed by the synthesis of a photocatalyst with a heterojunction composite construction with components that combine the advantages of individual components [45,47].

3.1.7. Bismuth-Based Multi-Component Oxides

Bismuth-based multi-component oxides exhibit enhanced photocatalytic performances when compared to pristine Bi₂O₃ and Bi₂S₃ photocatalysts owing to their developed surface area and porosity, as well as due to the photocatalytic activity in visible light and NIR region [45]. Bi₂WO₆ has a perovskite layered structure with a bandgap of ~2.8 eV and photocatalytic activity in the visible light region with a disadvantage of charge recombination that could be decreased by modifying the morphology, crystal facets, and defects [45,46,48,49].

BiVO₄ has a bandgap of 2.4 eV and outstanding visible light adsorption capacity [45,46,49]. BiVO₄ has three polymorphs: monoclinic, orthorhombic, and tetragonal, with the monoclinic BiVO₄ being the most photocatalytic-active [45]. However, the photocatalytic performance of BiVO₄ is still inadequate due to the tendency of charge carriers’ rapid decline through recombination [46,49]. Bi₂MoO₆ has a bandgap of 2.7 eV and photocatalytic activity under visible light, but it also has a serious charge recombination in the bulk that could be decreased by developing nanostructures, introducing oxygen vacancies and modifying crystal facets [46]. BiFeO₃ has a bandgap that ranges from 2.1 eV to 2.7 eV and a perovskite structure where Bi sites have ferroelectricity and Fe sites magnetism. These photocatalysts are applied for organic pollutants’ photocatalytic degradation, and their photocatalytic performances could be enhanced by engineering nanostructures’ heterojunctions and doping [46]. The following bismuth titanates show photocatalytic activity: Bi₄Ti₃O₁₂ (bandgap: 3.0–3.5 eV), Bi₂Ti₂O₇ (bandgap: 2.95 eV), Bi₁₂TiO₂₀ (bandgap: 2.4 eV), and Bi₂₀TiO₃₂ (bandgap: 2.38 eV). Their narrow bandgap facilitates photocatalytic activity under visible light. The photocatalytic activity of Bi₄Ti₃O₁₂ for azo-dyes degradation depends on exposed facets, particle size, and morphology [45,46]. Bismuth oxyhalides, BiOX (X = F, Cl, Br, I), have a layered structure with [Bi₂O₂] slabs interleaved by double slabs of halogen atoms [46]. The layered structure enables the formation of an internal static electric field perpendicular to layers, facilitating efficient separation of photoinduced charge carriers [50]. Their bandgaps are 3.64 eV (BiOF), 3.22 eV (BiOCl), 2.64 eV (BiOBr), and 1.77 eV (BiOI); thus, the variation of the Bi and X contents can tailor the bandgap and band edge positions of the photocatalyst [46]. Bismuth oxyhalides appear to be non-hazardous and inert, having good carrier transport properties [48]. Nevertheless, a fast recombination of photo-generated electrons and holes limits the photocatalytic performance of BiOX [50].

3.2. Metal Sulfides

Transitional metal sulfide compounds are intriguing photocatalytic materials due to their electronic structures, which make them more suitable for visible light photocatalytic degradation processes and energy conversion. Additionally, they offer tunable electronic, optical, physical, and chemical properties [33,40].

3.2.1. ZnS

ZnS has a high conduction band position and a wide band gap (3.6 eV), making it effective as a photocatalyst only in the UV light region. The advantage of ZnS are exceptional: transport properties for reducing the scattering and recombination of charge carriers, as well as thermal stability, high electron mobility, non-toxicity, insolubility in water, and a lower cost [51]. The photocatalytic properties of ZnS can be improved by doping with metal ions (Mn, Ni, Pb, Bi, Cu, etc.) to expand visible-light absorption and maintain strong reducibility by forming a new donor energy level within the ZnS band gap [9,40]

3.2.2. CdS

CdS has a narrow band gap (2.3 eV), which enables its excitation under solar light, and facilitates CdS activity in various organic pollutant photodegradation reactions [33,40,42]. The photocatalytic activity of CdS depends on its crystalline structure, morphology, and particle size [40]. Nevertheless, CdS has significant drawbacks, such as poor e⁻/h⁺ separation, susceptibility to photocorrosion due to its self-oxidative nature, poor stability, and high toxicity [33,40]. These shortcomings can be eliminated by surface modification, doping with noble metals (RuO₂, Pd, etc.), or coupling with semiconductors of different band gaps, such as CdS-HgS, CdS-ZnO, CdS-ZnS, CdS-TiO₂, CdS-PbS, ZnOCdSTiO₂, CdSZnO, and ZnSCdS [9,33,40]. Due to the presence of toxic elements like cadmium, CdS is unlikely to be used on a larger scale despite its beneficial properties in photocatalysis [6].

3.2.3. PbS

PbS has a narrow and small band gap (0.41–0.78 eV) with a large excitation Bohr radius (20 nm) because it also has a wide quantum size effect, high carrier mobility, and a high dielectric constant. Its efficiency in photodegrading organic compounds can be improved by doping (Ag-doped PbS) or coupling with other materials (PbS/BiOBr) [33].

3.2.4. Bi₂S₃

Bi₂S₃ has remarkable visible-light absorption properties owing to its bandgap varying from 1.28 eV to 1.84 eV [45,46]. Nevertheless, the limitation of Bi₂S₃ is its susceptibility to photocorrosion, wherefore Bi₂S₃ is mainly combined with other semiconductors in order to broaden their absorption range [46]. However, Bi₂S₃ nanoparticles that have a layered structure and occur in an orthorhombic phase forming different morphological structures (nanorods, nanowires, and nanoplates) showed better photocatalytic performance due to enhanced permeability and exceptional light-harvesting in their hierarchical architectures [45].

3.3. Non-Metal-Based Semiconductors: SiC

SiC is a third-generation semiconductor widely utilized in photocatalytic water purification due to its indirect and wide band gap (2.3–3.3 eV), high mechanical strength, high surface area, high melting point, excellent resistance to chemical oxidation, and high thermal stability. SiC showed photocatalytic activity for degradation of organic pollutants under UV light irradiation, but the bandgap of SiC can be tuned to absorb light in the visible range, depending on the crystalline structure, carbon content, and presence of doping atoms [52].

Moreover, SiC has also been investigated in photocatalytic reactions at higher wavelengths, taking advantage of its bandgap capability to absorb irradiation in the visible light range. In fact, depending on its crystalline structure, carbon content, and presence of doping atoms, the SiC bandgap can be slightly shifted to lower values (typically 2.4 and 3.0 eV for β-SiC and α-SiC, respectively), thus reaching the domain of visible light.

Its photocatalytic activity can be enhanced by coupling with other semiconductors, such as Pt/SiC, Ag₃PO₄/Ag/SiC, Co₃O₄/SiC, SiC/Cd, g-C₃N₄/SiC, and SiC/graphene [33].

3.4. Organic Materials

When compared to inorganic photocatalysts, organic photocatalysts exhibit superior responsiveness to visible and even infrared light. Their band structure can be readily modified at the molecular and atomic levels through the incorporation of specific molecular functional groups [40]. The most studied organic photocatalysts are graphene, GO, graphitic carbon nitride, g-C₃N₄, and metalorganic framework materials, MOFs [6,40,42].

3.4.1. Graphene

Graphene is a nanomaterial with a large surface area (~2630 m²g⁻¹), high electrical conductivity (108 Acm⁻²), high thermal conductivity (~2000–5000 W mK⁻¹), high mechanical stiffness (2.4 ± 0.4 TPa), high charge carrier mobility in room temperature (~100,000 cm² V⁻¹ s⁻¹), and optical transparency [40,42]. Moreover, the advantages of graphene-based nanocomposites as photocatalysts are also chemical functionalization, biocompatibility, and efficient water dispersion; these have been used for water purification under UV–visible-light irradiation [8]. To enhance photocatalytic performance, graphene is often integrated with other inorganic or organic semiconductors, owing to its extensive surface area, capability to prevent the agglomeration of immobilized metal oxides, and improvement in charge carrier separation [40,42].

3.4.2. Graphitic Carbon Nitride

Graphitic carbon nitride (g-C₃N₄) has high thermal and chemical stability and a molecular structure resembling graphite with a two-dimensional morphology [10,42]. Its basic tectonic units consist of tri-s-triazine or s-triazine rings, with a carbon-to-nitrogen molar ratio of approximately 0.75. It features a tunable medium band gap (around 2.7 eV) and is responsive to visible light, with well-defined lowest unoccupied molecular orbital (LUMO) and highest occupied molecular orbital (HOMO), along with a favorable conduction band edge position. Moreover, it possesses numerous surface groups that facilitate coupling with other photocatalysts [6,10,40,42]. The g-C₃N₄ photocatalyst showed activity for hydrogen evolution, carbon dioxide reduction, organic pollutant removal, and organic synthesis [10]. Despite advantages such as high stability, high chemical resistance, non-toxicity, high reduction ability, and cost effectiveness, g-C₃N₄ has several drawbacks including insufficient visible-light absorption, low oxidation ability, high recombination rate, low solvent-accessible surface area, and susceptibility to degradation by hydroxyl radicals [6,15,40]. In order to increase g-C₃N₄ surface area, various morphologies have been developed (hollow structures, nanochips, flowers, nanotubes, nanorods, etc.) [40].

The discussed findings of numerous studies on photocatalytic materials are summarized in

Table 2. Summarized comparison of numerous photocatalytic materials: advantages and drawbacks.

Photocatalytic Materials	Advantages	Drawbacks	Ref.
Metal oxides
TiO2 (anatase)	Significant activity in the degradation of organic pollutants (dyes, phenolic compounds, pesticides, herbicides, benzenes, furans, chlorinated alkanes, and dioxins), significant activity in the degradation of inorganic contaminants (including toxic metal ions like Pt2+, Rh3+, and Cr6+), excellent antibacterial and antiviral activity against various bacteria and viruses, durability, low cost, low toxicity, chemical and photochemical stability, superhydrophilicity.	Limited absorption, only in UV solar light region due to wide band gap (3.2 eV), rapid recombination of photogenerated e−/h+ pairs, reduced adsorption of hydrophobic organic pollutants on photocatalysts, aggregation of nanoparticles during photocatalytic degradation, difficult recovery of nanosized particles from the treated wastewater.	[9,11,33,34,40,42,43]
ZnO	Non-toxicity, insolubility in water, excellent piezoelectric effect, high electron mobility, high photosensitivity, environmentally friendly nature, relatively low cost.	Limited absorption, only in UV solar light region due to wide band gap (3.37 eV), susceptibility to photocorrosion, limited pH range of operation (ZnO dissolves in strong acids and alkalis).	[9,33,34,35,40,42,44]
WOx	Narrow band gap (2.8 eV), effective in the blue region of the visible solar spectrum, chemical stability at various pH levels.	Synthesis complexity, cost, and scalability issues when compared to TiO2 and ZnO.	[33,40,42]
Cu-based oxides	CuO can be used as catalyst support, semiconductor, and catalyst; the band gap can be adjusted across a wide range within the sunlight spectrum, from approximately 1.2 eV to greater than 3.0 eV by varying transition metals and the Cu(I) coordination environment in ternary Cu-based metal oxides.	Cu2O is thermodynamically unstable in aqueous solutions and tends to reduce to Cu. A protective layer on a Cu2O electrode is necessary,	[33,34]
SnO2	When coupled with other semiconductors, SnO as good electron acceptor, forms heterostructures with efficient charge separation and improved photocatalytic properties.	Limited absorption, only in UV solar light region due to wide band gap (3.6 eV), cannot be used independently as photocatalyst since it cannot reduce oxygen molecules due to its conduction band position.	[33,40,42]
Bi2O3	Visible light responsive photocatalyst.	Recombination of photogenerated electrons and holes.	[45,46,47]
Bismuth-based multi-component oxides (Bi2WO6, BiVO4, Bi2MoO6, BiFeO3, Bismuth titanates)	Photocatalytic activity in visible light and NIR regions.	Charge recombination.	[45,46,47,48,49]
Bismuth oxyhalides, BiOX (X = F, Cl, Br, I)	Efficient separation of photoinduced charge carriers, the variation of the Bi and X contents can tailor the bandgap and band edge positions of the photocatalyst, non-hazardous and inert.	Fast recombination of photo-generated electrons and holes.	[46,48,50]
Metal sulfides
ZnS	Exceptional transport properties for reducing the scattering and recombination of carriers, thermal stability, high electron mobility, non-toxicity, insolubility in water, lower cost.	Limited absorption, only in UV solar light region due to wide band gap (3.6 eV).	[9,30,40]
CdS	Photocatalytic activity in visible light region	Poor e−/h+ separation, susceptibility to photocorrosion due to its self-oxidative nature, poor stability, high toxicity.	[6,9,33,40,42]
PbS	Narrow and small band gaps (0.41–0.78 eV), high carrier mobility, high dielectric constant.	Moderate photocatalytic activity.	[33]
Bi2S3	Remarkable visible-light absorption properties.	Photocorrosion.	[45]
Non-metal-based semiconductors
SiC	Indirect and wide band gaps (2.3–3.3 eV) tunable to adsorb visible light, high mechanical strength, high melting point, excellent resistance to chemical oxidation, high thermal stability.	Very low surface area of SiC powder, moderate photocatalytic activity.	[33,52,53]
Organic materials
Graphene	Large surface area, high electrical conductivity, high thermal conductivity, high mechanical stiffness, high charge carrier mobility in room temperature, optical transparency, biocompatibility, efficient water dispersion, photocatalytic activity in UV-VIS region.	Pristine graphene is rarely used, graphene-based nanocomposites have enhanced photocatalytic performance.	[8,40,42]
g-C3N4	tunable medium band gap (~2.7 eV), photocatalytic activity in UV-VIS region. Numerous surface groups that facilitate coupling with other photocatalysts, high stability, high chemical resistance, non-toxicity, high reduction ability, cost effectiveness.	Insufficient visible-light absorption, low oxidation ability, high recombination rate, low solvent-accessible surface area, susceptibility to degradation by hydroxyl radicals.	[6,10,15,40,42]

.

4. Strategies for Improving Photocatalytic Efficiency

Enhancing the photoactivity of TiO₂ is a critical area of the ongoing research on photocatalysis, driven by the need to optimize both kinetic and thermodynamic parameters to achieve higher photocatalytic efficiency. Beyond the inherent thermodynamic properties such as band gap energy and the alignment of the conduction band (CB) and valence band (VB), the efficiency of TiO₂ photocatalysis is significantly influenced by kinetic factors. Key kinetic challenges in photocatalytic processes include efficient solar light harvesting, effective charge separation and migration, and the facilitation of surface reactions. Addressing these challenges involves multiple strategies, including structural engineering through metal and non-metal doping, morphology design, facet engineering, photosensitization, and the creation of semiconductor heterojunctions. These approaches aim to optimize the interactions at the microscopic level, thereby enhancing the overall photocatalytic performance of TiO₂.

4.1. Structural Engineering (Metal and Non-Metal Doping)

Doping of metal/non-metal ions/atoms has emerged as a promising strategy to modulate the band gap structure of photocatalysts, shifting the adsorption edge from UV to visible photon absorption, thereby enhancing visible-light harvesting efficiency [54]. Numerous studies have focused on designing optimal metal-doped semiconductors to improve their activity in the photodegradation of organic pollutants in wastewater systems. Commonly used metals for doping include platinum (Pt) [55], palladium (Pd) [56], gold (Au) [57], silver (Ag) [58], copper (Cu) [59], and nickel (Ni) [60]. Doped metals act as electron acceptors, capturing photo-excited electrons from the valence band. The formation of a Schottky barrier at the metal–semiconductor junction lowers the Fermi levels of the metal below the conduction band of TiO₂, enhancing charge transfer and preventing charge recombination (Figure 3). A higher work function of the metal improves its ability to accept electrons, thereby suppressing charge recombination. Notably, lowering the conduction band minimum (i.e., making it less negative) relative to the redox potential of water can limit the scope of potential reduction processes.

Pintar et al. [55] conducted a study on the synthesis of TiO₂ + Pt plasmonic solids, with Pt particles supported on a different TiO₂ support, and investigated their photo, thermal, and photo-thermal activity in gas–solid and gas–liquid–solid environmental applications. The study concluded that the nature of the TiO₂ support influenced the size, distribution, and interaction degree of Pt particles, which, along with Schottky barrier height, were critical factors determining catalyst performance. Sakthivel [62] compared the photoactivity of noble metal-deposited TiO₂ in the photocatalytic degradation of leather dye Acid Green 16. The results revealed that Pt and Au exhibited higher photocatalytic efficiency compared to Pd, attributed to their suitable work function and high electron affinity properties. Zheng and co-workers [63] advanced this research by synthesizing edge-site-deposited Au/TiO₂ nanosheets on active {001} facets, rather than randomly depositing metal-doped nanosheets. This approach significantly enhanced photocatalytic activity in the degradation of rhodamine B. The strong interfacial contact between edge-deposited Au nanocrystals and TiO₂ nanosheets was considered to be the decisive factor for improved charge separation efficiency and enhanced visible-light harvesting efficiency.

Additionally, noble metals improve the charge carrier migration through provoking a surface plasmon resonance (SPR) effect, which can result in a substantial increase in charge carrier formation. In this case, the electrons are being directly transferred to the CB of the semiconductor, resulting in a thousand-times-larger charge carrier formation [64]. The visible-light activity of BiVO₄ materials in phenol decomposition has been reported to be improved by Au and Pt nanoparticles’ deposition due to the effects of localized surface plasmon resonance (LSPR). Moreover, the selective modification of Pt-Au cocatalyst on the {010} facets of BiVO₄ exhibited higher photocatalytic performance compared to a random deposited Pt-Au/BiVO₄ catalyst [65].

However, the ability of metals to suppress charge recombination depends on their preparation method and the concentration of the used metal dopant. Excessive amounts of deposited metals can act as a recombination center, having detrimental photoactivity effects [66,67,68].

Many studies have reported that co-doping showed better results than single doping. Chippada and co-authors [69] investigated the photocatalytic degradation of carcinogenic food/textile dye Amaranth under visible light irradiation using structurally modified nano-titania doped with barium and copper. The optimal amount of Ba and Cu was found to be 0.22 and 0.73 wt%, respectively, which significantly increased photocatalytic efficiency. However, a further increase in metal dopant concentration decreased the photo-efficiency due to metals acting as recombination centers and blocking photon absorption. In another study performed by Khan and co-workers [70], the photocatalytic performance of Cu- and Ni-doped ZnO was examined in methylene orange dye degradation. The similar ionic radius of Cu²⁺ (73 pm) and Ni²⁺ (69 pm) with Zn²⁺ resulted in an easy incorporation of these metals in crystals, forming acceptor sites in conjunction with a neighboring oxygen vacancy. Co-doped catalysts displayed superior activity compared to single-doped systems as a result of the formed Zn–O–Ni and Zn–O–Cu functional groups and more generated oxygen vacancies.

The major drawbacks of using metal-doped photocatalysts lies in their photocorrosion, poor thermal stability, and probability to act as recombination centers. In contrast, non-metal doping has been used to improve the spectrum sensitivity of semiconductors, similarly to metal doping, but overcoming the above-mentioned weaknesses. As illustrated in Figure 4, non-metal dopants create a new intermediate energy level above VB, positively shifting the band edges. Since the groundbreaking work of Asahi et al. [71] on incorporating nitrogen into the TiO₂ lattice to enhance light absorption into the visible spectrum, various non-metal-doped nanomaterials have been extensively researched. Visible-light-active TiO₂ structures doped with boron (B) [72], carbon (C) [73], sulfur (S) [74], nitrogen (N) [75], or halogens [76] generally show good performance toward the photo-oxidation of different organic pollutants. The non-metal doping can be classified as substitution (N^3–, replacing oxygen) and interstitial doping (N0, N…O groups with lattice oxygen). Substitution doping replaces oxygen with a foreign atom, while interstitial doping places the foreign non-metals in the interstitial sites. Both states can broaden the spectral range for light absorption to longer wavelengths. Substitutional doping results in an upward shift of the TiO₂ valence band (VB) or creates discrete interband states above the VB, rather than forming a continuous VB [77,78]. Other reports indicate that N doping might be interstitial [79]. However, it is essential to manage the elevation of the valence band maximum because raising the VBM above the H₂O/OH redox potential can prevent the creation of strong oxidative hydroxyl radicals [80].

The type of doping depends on the type of non-metal (size and charge-state), synthesis environments, dopant precursor [82], etc. As a rule, oxygen-rich conditions favor the formation of interstitial N and NO_x species, while oxygen-deficient conditions favor the substitutional N doping [83]. Generally, X-ray photoelectron spectroscopy (XPS) is considered to be a useful tool for determining whether nitrogen doping in the TiO₂ structure is interstitial or substitutional. However, some researchers question the reliability of fitting the nitrogen peak from XPS characterization [84].

The debate over whether substitutional or interstitial doping displays higher photocatalytic activity is ongoing. While Song and co-authors [85] found that substitutional nitrogen samples displayed higher activity in the degradation of gaseous benzene compared to samples containing interstitial nitrogen, other studies reported enhanced activity of TiO₂ structures with interstitial doped nitrogen [86]. However, it is important to recognize that comparing the activity of different TiO₂ systems can be irrational. This is not only due to variations in their intrinsic properties caused by different synthesis methods, but also because of diverse reaction conditions such as reactor type, nature of light irradiation, light spectrum and intensity, type of pollutant, and so forth. Therefore, a significant challenge lies in studying the relationship between nitrogen doping type and photocatalytic activity of N-doped systems.

DFT calculations have been recognized as a powerful tool to analyze the formation and position of N-localized states. Di Valentin and co-authors [87] concluded that substitutional N states are positioned just above the valence band, while interstitial nitrogen states are positioned higher in the gap. In the study of Fadlallah [88], the DFT method was applied to study the electronic, optical, and photocatalytic properties of different non-metal-doped (B, C, N, P, and S) and halogen-doped (F, Cl, Br, and I) anatase TiO₂ nanotubes (TNTs). The reduction in the band gap of non-metal-doped titanium nanotubes (TNTs) was found to be greater compared to the halogen-doped TNTs.

The simultaneous co-doping of metals and non-metals into TiO₂ nanomaterials has garnered considerable interest due to its reported enhancement of photocatalytic activity. Recent studies have focused on investigating co-doped [89,90] and multi-element-doped TiO₂ systems [91,92].

4.2. Morphology Design

Advances in the scientific understanding of catalyst preparation processes have accelerated the design and development of new catalyst groups. Improving photocatalytic performance can be achieved by designing catalysts with desired morphologies. Nakata and Fujishima proposed a classification of TiO₂ photocatalysts in four categories based on their dimensionality (Figure 5) [93]. Given the vast scope of morphology design, the following section focuses exclusively on the formation and characteristics of 1-D and 2-D architectures, including their associated advantages, drawbacks, and applications in dye degradation systems.

The use of 1-D nanostructures offers benefits such as reducing particle size, leading to a larger surface-to-volume ratio. Additionally, the reduced charge carrier diffusion distance in 1-D nanostructures decreases the possibility of electron-hole recombination. So far, different methods have been proposed for the preparation of 1-D nanostructured materials.

Table 3. Preparation methods for the synthesis of 1-D nanostructured photocatalysts.

Composition	Morphology	Preparation Techniques	Ref.
RuTe2	nanorods	Chemical precipitation and coprecipitation	[94]
vanadium-doped mesoporous silica	nanospheres	Sol–gel approach	[95]
TiO2, CeO2, Y2O3-ZrO2 and SrTiO3	nanofibers	Electrospinning	[96]
ZnO	nanorods	Solid-state reaction	[97]
Bi4Ti3O12	nanorods and nanotubes		[98]
Titanates	nanotubes	Hydrothermal/solvothermal	[99]
Titanates	nanorods	Vapor-phase hydrothermal	[100]
Carbon	nanotubes	Aerosol flame synthesis	[101]

summarizes the most commonly used preparation methods for synthesizing bulk 1-D nanostructured photocatalysts.

Following the pioneer work on titanate nanotube synthesis by Kasuga et al. [102,103], the fabrication of 1-D titanate nanomaterials with various morphologies, such as nanotubes [104,105], nanowires [106], nanofibers [107], and nanoribbons [108], has been widely studied. The advantage of hydrothermal synthesis lies in its ability to tune the structure and morphological features of the final titanates by changing the TiO₂ precursor type [109], the type and concentration of the alkaline solution [110], hydrothermal treatment conditions (temperature and pressure) [111], post-acid washing [112], etc. Figure 6 presents distinct TiO₂ nanomaterials’ morphologies.

Various case studies and review studies documenting the use of the 1-D nanomaterials for the removal of the different dye pollutants have been reported in the academic literature [117,118,119].

The review of Wang and co-workers provides comparisons of nanomaterials based on their dimensionality [120]. A variety of morphologies and dimensionalities offer distinct advantages, yet no single dimensionality dominates by encompassing all advantages over the others [121].

Two-dimensional (2D) nanomaterials refer to materials with a thickness of just one or a few atomic layers and virtually infinite dimensions in the other two directions. The exceptional properties of 2D nanomaterials offer significant advantages for enhancing photocatalytic processes, making them highly promising for tackling environmental and energy-related issues. These advantages include the following:

−

High surface area, which leads to an increased number of active sites available for photocatalytic reactions, resulting in improved overall efficiency;

−

The planar structure, which improves light absorption, resulting in more effective excitation of electrons and holes, and enhanced charge separation and electron transfer, resulting in reduced charge recombination;

−

The unique electronic properties and high surface energy, as well as the tunability of their electronic, optical, and chemical properties;

−

The reduced structural defects compared to their bulk counterparts;

−

The easier development of 2D-based composite photocatalysts facilitated by their unique characteristics, which introduce new functionalities;

−

The easier development of large-scale photocatalytic systems for industrial applications.

Although numerous techniques exist for creating 2D nanoscale structures, these methods are typically categorized into top–down and bottom–up approaches [122]. The bottom–up method is more effective at generating 2D nanosheets with a consistent chemical composition and fewer structural imperfections. Also, 2D titania nanosheets can be obtained as an intermediate product in the hydrothermal synthesis of 1D titania. A precise regulation of reaction conditions during the hydrothermal process is crucial for achieving the desired 2D nanosheet morphology. Yang et al. [123] successfully synthesized anatase nanosheets with predominant {001} facets using the alcohol solvothermal technique. According to the SEM images obtained, the average dimension of a single crystal anatase nanosheet is five times greater than its thickness.

Additionally, the combination of different dimensionalities can enhance certain material properties. The most common method for constructing 1D/2D heterogeneous photocatalysts is by growing 1D nanostructures on 2D nanostructures [124]. Depending on the conduction band potential, reduction and oxidation processes can occur selectively on the surface of either 1D or 2D nanostructures. When 1D nanowires/nanorods possess a narrower bandgap and a more negative conduction band (CB), photogenerated electrons from the 1D nanowires/nanorods can be readily transferred to the CB of the 2D-based material, thus facilitating charge separation. Figure 7 presents cross-sectional SEM image of GaN nanowires grown on a graphene monolayer.

4.3. Facets Engineering

In recent years, surface structure engineering has emerged as a promising research direction for improving material properties [125]. The main scientific challenge in nanomaterials design is the exposure of high-energy facets on the surface of photocatalytic crystals. The main low-index facets of anatase TiO₂ phase are {001}, {010}, and {101}, with surface energies of 0.90 Jm⁻², 0.53 Jm⁻², and 0.44 Jm⁻², respectively [126,127,128]. The low surface energy {101} facet is thermodynamically the most stable, accounting for more than 90% of the total surface [129]. Theoretically, the facet surface energy determines their reactivity. Due to the lower surface energy, higher stability, and fewer active sites, the {101} facets display a much poorer performance compared to the {001} facets [130]. However, one must not neglect the importance and the effect of additional factors, such as the degree of oxygenation, the existence and density of defects on the facet surfaces, flatband potentials, surface atomic structures, facet–pollutant interactions, and the selectivity of facet surfaces to specific functional groups of adsorbed molecules [130,131], which can exert a crucial impact on the photocatalytic performance of certain facet surface. Thus, many contradictory results regarding the photoactivity order of different low-index facets have been reported. For example, Dudziak et al. [132] reported the highest activity of anatase {101} facets, while {001} revealed the lowest photoactivity in degradation of all four studied pollutants. Contrarily, Han et al. [133] found higher activity of {001} facets than {101} facets in the degradation of methyl orange dye. Lei and co-authors [134] found that the photoactivity order of {001} and {101} facets depended on the treatment method. Thus, the {001} facet was found to be the most active in the degradation of methylene blue dye when samples were calcined. Conversely, when samples were treated with NaOH, the {001} facets were less efficient than the {101} facets. Ye and co-workers [135] found opposite orders of photocatalytic activity of anatase’s low-index facets for rhodamine B photodegradation and photoreduction of CO₂ to CH₄. Specifically, the {001} facet was the most active for rhodamine B degradation, while the {010} facet was the most active for the CO₂ photoreaction. The authors concluded that differences in CO₂ molecules’ adsorptive properties on various exposed facets, the conduction band level position, and the efficiency of separating photo-generated carriers are the three primary reasons explaining the differences in photocatalytic activity order between liquid-phase and gas-phase reactions.

Numerous efforts have been directed towards preparing TiO₂ anatase crystals with exposed facets of high surface energy and high activity. Significant success has been achieved in preparing {001} facets, and many studies have been focused on the formation of anatase with these facets and their application in the photocatalytic degradation of pollutants [136], hydrogen production [137], photo-enhanced thermal CO₂ methanation [138], etc. Domen and his group [139] succeeded in obtaining an internal quantum efficiency for overall water splitting of almost unity, using a modified aluminum-doped strontium titanate (SrTiO₃:Al) photocatalyst by photodepositing Rh/Cr₂O₃ and CoOOH on different SrTiO₃ crystal facets.

Recently, attention has increasingly been directed towards the controllable synthesis of other high-energy facets, such as the {100}, {110}, {111}, and {10l} (l > 1) facets. Additionally, it has been reported that nanocrystals with different exposed facets exhibit higher activity compared to those synthesized with only one exposed facet. Zuohua Liu and co-workers [140] synthesized tetragonal anatase TiO₂ nanorods with designed {100}, {001}, and {101} facets. They reported that an optimal composition of 25% low-energy {101} facets and 75% high-energy {100} and {001} facets, with a {001}/{101} ratio of 1:9, led to enhanced H₂ production. The high-magnification FESEM images in Figure 8 distinctly reveal TiO₂ nanorods synthesized at 150 °C with various exposed facets. The discrepancy in results between this study and others underscores the importance of additional factors beyond surface energy, such as molecule adsorption properties and synergistic effects between coexisting facets, which have a decisive impact on the performance of specific surfaces of anatase nanocrystals.

4.4. Photosensitization

In the photosensitization process, a molecular entity (sensitizer) absorbs light and transfers energy to the desired reactant (semiconductor). The photosensitization of semiconductors with organic and organometallic molecules has been considered as a simple and effective approach to developing visible-light responsive photocatalysts. The advantage of using organic molecules lies in their tunable HOMO and LUMO levels through the anchoring of different ligands, low cost, and great variety. The concept of semiconductor photosensitization was inspired by Brian O’Regan and Michael Grätzel who firstly invented dye-sensitized solar cells, commonly known as Grätzel cells [141,142].

The photosensitization of semiconductors with appropriate organic molecules results in a narrowing band gap, thus moving the light absorption threshold from the UV to the visible region or near-IR (NIR) region, improving the light harvesting efficiency of photocatalytic and photovoltaic devices. Regarding charge transfer from the sensitizer molecule to the semiconductor, two distinct mechanisms can be classified: Type I (straddling) and Type II (staggered) [143,144]. In Type I, the dye sensitizer is firstly photoexcited to a dye’s excited state, and then an electron is transferred from the excited dye level (LUMO) to the semiconductor conduction band. Type II is regarded as a direct mechanism and involves the transfer of electrons from a dye’s ground state level (HOMO) to the semiconductor conduction band.

In the last two decades, different sensitizers have been used for the surface modification of semiconductors, such as benzene derivatives (catechol [145,146], salicylic acid [147], thiosalicylic acid [148]) and different dyes, such as organic dyes (porphyrins and metallic porphyrins [149,150], phthalocyanines [151], erythrosin B [152], rhodamine B [153], eosin Y [154], rose bengal [155]), metal-organic dyes [156,157], and natural dyes (chlorophylls [158], flavonoids [159], tannins [160]).

Photosensitizers generally have different absorption abilities, causing diversity in their efficiencies and costs. For example, natural dyes are abundant, low cost, and stable, while the synthetic dyes are more tunable and display higher efficiency. In particular, ruthenium dyes demonstrate high conversion efficiency but have the drawback of being expensive to synthesize, environmentally harmful, and unstable in water [161]. Thus, intense studies have been directed toward the design of metal-free dyes as more environmentally friendly solutions.

The photocatalytic performance of a surface sensitized catalyst is strongly correlated with the sensitizer’s characteristics, such as absorption wavelength, electron transfer facility between the semiconductor and organic sensitizer, organic sensitizer structure, etc.

The photosensitizing approach can lead to equal or even larger activity compared to other methods (non-metal or metal doping or heterojunction formation). Moreover, it was revealed that the photosensitization method can be used to transform even the wide band gap isolators to visible-light-responsive hybrid materials. The group of Nedeljkovic [162] used 5-aminosalicylic acid (5-ASA) to modify non-absorbing Al₂O₃, moving the absorption threshold to the visible region. Photosensitizing Al₂O₃ facilitated the development of photocatalytically active hybrid materials demonstrating comparable performance under simulated solar and visible light.

However, the main challenge regarding the photosensitizing approach is to obtain a sensitized photocatalytic system with a good stability, high solar light harvesting efficiency, and high photoconversion efficiency. Therefore, a number of studies have been directed toward the optimization of the parameters affecting the photocatalytic efficiency of a sensitized semiconductor system, such as the type and content of the photosensitizer [163], the type of the electronic injection mechanism (direct or indirect) [164], and photocatalytic reaction conditions (the type of light irradiation, nature of the pollutant and the solvent, solution pH) [162]. In this regard, it is likely that a sensitizer optimal for one system might display lower activity compared to another sensitizer in a different photocatalytic system, depending on the specific reaction conditions applied.

4.5. Design of Semiconductor Heterojunction

Coupling two semiconductors with different energy band alignments is an efficient and challenging method to suppress charge recombination and shift absorption to the visible light region, thereby enhancing photo-efficiency. Appropriately matching their CB and VB levels leads to vectorial charge transfer from one semiconductor to another. Depending on energy levels and charge transfer direction, three types of heterojunctions have been suggested: type I (straddling gap), type II (staggered gap), and type III (broken gap) [165]. A schematic diagram of possible heterojunction types is presented in Figure 9.

In the described heterojunction systems, the electrons move from a more negative CB of one semiconductor to a lower CB of another semiconductor, while holes move from a more positive VB to a less positive VB, which results in the lower redox potential of the CB and VB [167]. The difference in the described systems lies only in the different directions of charge transfer and relative positions of the CB and VB levels. Although the formation of these heterojunctions enables narrowing the band gap, it also results in redox sites that are less active compared to single-component photocatalyst. Therefore, a new class of heterostructured photocatalytic systems with enhanced reaction efficiency have been developed, i.e., Z-scheme. Unlike conventional type II heterojunctions, in the Z scheme, the electrons move from the CB of one (PS II) semiconductor to the VB of (PS I) another semiconductor, while holes remain in the VB of PS II. This results not only in a lower energy band gap, but also in the highest reduction and oxidation potentials within the heterojunction system [168,169]. With their favorable properties, Z-scheme heterojunctions have found application in different photocatalytic fields, such as water splitting [170], CO₂ photoreduction [171], and waste-water treatment [172].

Different methods have been proposed to determine (identify) the type of semiconductor heterojunction. Thus, the photoluminescence (PL) method and radical trapping experiments have been successfully applied to elucidate the type of heterojunction photocatalyst system. Yu and co-workers [173] investigated the photocatalytic formaldehyde degradation by TiO₂/g-C₃N₄, analyzing the formation of ^•OH using the photoluminescence (PL) method with terephthalic acid as a probe molecule. They observed a progressive increase in PL signal with irradiation time, which indicated the increase in ^•OH formation. This finding further confirmed that the TiO₂/g-C₃N₄ heterojunction belonged to a direct Z-scheme system. If the TiO₂/g-C₃N₄ heterojunction were a type-II heterojunction, there would be no increase in PL intensity over time, as the VB of g-C₃N₄ has an unfavorable energy level for the formation of ^•OH. Another noteworthy finding regarding the determination of the TiO₂/g-C₃N₄ heterojunction type was detailed in a recent study [172]. Through scavenger experiments, the authors concluded that the g-C₃N₄/TiO₂ system is a type II heterojunction rather than a direct Z-scheme, as ^•OH was identified as a byproduct species. A PL analysis further confirmed the heterojunction type. The absence of an additional PL peak with energy below the lower band gap of the individual components served as an indirect indication of a type II heterojunction.

A significant breakthrough of the 1990s that captivated the scientific community was seminal papers by O’Regan and Graetzel in 1991 [141,142]. The authors demonstrated that TiO₂ could be photosensitized with a dye in the presence of a redox couple to produce electrical power, leading to the development of what is now known as the Graetzel solar cell. Since then, hundreds, if not thousands, of papers have been published in this area.

5. Summary of Photocatalysis Application

Numerous case studies and reviews have been devoted to examining the application of photocatalysis for the removal of organic dyes from wastewater.

Table 4. Comparison of the photocatalytic performance of TiO₂-based nanocomposites in degradation of various dye pollutants.

Type of Materials	Dye Pollutant	Source of Light	Removal Efficiency (%); Reaction Time	Stability vs. Number of Cycles	Ref.
Composites
TiO2/Fe2O3	Reactive blue dye	Sunlight irradiation	100%; 120 min	9.7% after 4th run	[174]
Bi2WO6/TiO2	Rhodamine B (RhB), methylene blue (MB), and methyl orange (MO)	Simulated solar irradiation using a 500 W xenon lamp (AM1.5, 100 mW/cm2)	95.2%; 30 min (RhB) 89.3% within 60 min (MB) 15.1% MO; 60 min (MO)		[175]
Metal doped
W-TiO2	Procion Red MX-5B	Visible light irradiation (35 W Xe-arc)	92.05%; 60 min	90% after 5th run	[176]
Fe-TiO2	Methyl orange	Visible light irradiation (300 W xenon lamp with a UV cut off filter (λ > 400 nm))	98%; 60 min	n.a.	[177]
Nonmetal doped
C-TiO2	Methylene blue	Sunlight irradiation	98.86%; 140 min.	85.94%, after 4th run	[178]
N-TiO2	Methylene blue (MB) and Methyl orange (MO)	Visible light irradiation (300 W xenon lamp with a UV cut off filter (λ > 420 nm))	92%; 60 (MB); 95%; 180 min (MO)	100% after 6th run	[179]
Metal/nonmetal co-doped
P/Zr co-doped TiO2	Rose Bengal	Visible light irradiation	95%; 40 min	78% in 4 th run	[180]
Co/S co-doped TiO2	Methylene Blue	Visible light irradiation (200 W halogen lamp)	93%; 120 min	n.a.	[181]

presents the comparative analysis of the photocatalytic performance of TiO₂-based photocatalytic systems in the degradation of various dyes.

However, comparing the performance of different photocatalytic systems might be unreasonable due to several factors:

−

The intrinsic properties, including morphology, textural, and structural characteristics, vary significantly among different photocatalysts, making them challenging to standardize and compare;

−

Reaction conditions such as the type of reactor and the nature of light irradiation—including the light spectrum and intensity—vary among photocatalytic systems. Light absorption and scattering play an important and often underrated role in the overall photocatalytic process, significantly affecting the reaction rate and being difficult to standardize;

−

The type of pollutant, its concentration, the catalyst mass, and the pollutant-to-catalyst ratio differ between photocatalytic systems.

Although some attempts have been made to develop standardization methods using the Degussa P25 photocatalyst and chlorophenol, there are still other parameters that need to be addressed. For example, identifying the relationship between the structure of the pollutant and its degradation rate is crucial for understanding the efficiency of photocatalytic processes, yet it remains insufficiently researched. The pollutant composition and the interaction between the catalyst surface and the pollutant molecule are critical factors affecting both photocatalytic kinetics and mechanisms. Although several studies have investigated the relationship between the pollutant structure and its decolorization efficiency in photocatalytic systems [182,183], a quantitative understanding of this relationship is still lacking. The Linear Free Energy Relationship (LFER) model, utilizing the Hammett equation [184], has been effectively employed to establish a quantitative connection between the structure of organic pollutant molecules and their chemical reactivity within photocatalytic systems.

Generally, photocatalytic processes involve the formation of various reactive species, including ^•OH radicals, which can react with aromatic compounds. Since the reaction of ^•OH radicals with aromatic compounds is electrophilic in nature, it is expected that the reaction should be favored by a high electron density at the reaction site. The study by Parra et al. [185] demonstrated that the reactivity of phenolic compounds was influenced by the nature and position of substituents. The results indicate that electron-donating groups accelerated the reaction, while electron-withdrawing groups inhibited it, which is in alignment with the electrophilic nature of OH-induced oxidation (Figure 10, left).

In contrast, the study by Dostanic et al. [186] reported different findings. A number of arylazo pyridone dyes were synthesized by varying the substituent type in the diazo moiety, ranging from strong electron-donating to strong electron-accepting groups. The reactivity of the synthesized dyes was tested through their decolorization efficiency in a Degussa P-25 TiO₂ photocatalytic system. The results showed that the reaction was facilitated by electron-withdrawing groups and retarded by electron-donating ones. This outcome is likely due to the transfer of negative charge from the pyridone oxygen to the azo nitrogen at the reaction site (Figure 10, right).

Understanding reaction pathways and degradation mechanisms is crucial, especially for evaluating the safety and environmental impact of the process. High-Performance Liquid Chromatography coupled with Mass Spectrometry (HPLC-MS) is a widely used technique for identifying and analyzing intermediates formed during photocatalytic degradation processes. In the work of Meetani et al. [187], the LC-MS/MS technique was used to study the degradation mechanism and to track numerous intermediate products formed during the course of orange G degradation. Degradation typically involves breaking the azo bond or other types of reactions where the azo group undergoes an electrophilic attack [188]. Depending on the reactive species produced, variations in dye structure and size, and degradation pathways, different intermediates may be formed, which can further influence the adsorption and photodegradation processes. The primary intermediates found during the degradation of orange G dye were substituted phenols, aromatic hydroxylamines, nitroso compounds, and dicarboxylic aromatic compounds, which are formed through the desulfonation and hydroxylation of the aromatic ring, along with oxidative cleavage of the azo bond, destruction of aromaticity, and ring cleavage [181]. According to the study of Rauf et al. [189], π bonds in the azo group are the preferred site for ^•OH radical attack, which can occur on either nitrogen atom of the azo group, leading to a diverse array of degradation products.

The importance of identifying the degradation mechanism lies in the fact that the resulting aromatic amines are often highly toxic and carcinogenic, posing greater risks than the original dye [190]. Detecting the nature of intermediates is, therefore, of primary concern to ensure the environmental safety of the photocatalytic process. In the work of Dostanic et al. [186], quantum mechanical calculations were employed to elucidate the mechanism of the photocatalytic oxidation reactions of arylazo pyridone dyes and interpret the dyes’ reactivities within the framework of the Density Functional Theory (DFT). The results indicated that both the azo nitrogen atom linked to the benzene ring and the pyridone carbon atom connected to the azo bond served as active sites for a ^•OH radical attack. The ^•OH attack preferentially occurred on the nitrogen atom for all dyes, except for the dye with a NO₂ substituent, where the attack rather occurs on the carbon atom. The study demonstrates that altering the type of substituent in the dye structure can direct the degradation pathways of dyes toward non-toxic or less toxic by-products, thereby significantly reducing the toxicity of wastewater.

Both thermodynamic and kinetic factors must align to achieve an efficient photocatalytic process and complete mineralization. Thermodynamically, for the generation of hydroxyl radicals through the oxidation of water or hydroxide ions, the redox potential of the H₂O/^•OH couple must be less positive than the valence band potential of the photocatalyst. However, if the kinetic parameters are not optimal, the concentration of reactive species like OH radicals might be insufficient to effectively degrade the intermediates formed during the reaction. Ensuring that kinetic parameters such as the concentration of electron and holes, charge separation, charge transfer, and kinetics of surface reaction are within an optimal range is essential for maintaining a high concentration of reactive species and, consequently, for the effective degradation of initial pollutants and intermediates. Materials capable of fully mineralizing dyes into CO₂ and H₂O often consist of advanced photocatalysts, which generally demonstrate strong oxidative abilities and favorable kinetic properties.

6. Conclusions and Future Perspectives

In conclusion, the continuous advancements in the design and development of photocatalytic materials have significantly contributed to enhancing their efficiency and application range. Various methods of doping, photosensitization, and the formation of heterojunctions have shown promising results in improving the photocatalytic performance of semiconductors, particularly TiO₂.

Looking forward, several key areas require further research and development in order to maximize the potential of photocatalytic materials. Future research should focus on optimizing the combination and concentration of metal and non-metal dopants to achieve the best possible photocatalytic performance while maintaining stability and reducing the negative effects of recombination centers. Furthermore, research should aim to improve the solar light harvesting efficiency and photoconversion efficiency of sensitized photocatalytic systems. Considering the ongoing dispute on high-energy facets, additional systematic studies are needed to understand the relationship between facet exposure, molecule adsorption properties, and photocatalytic performance. Developing methods to synthesize nanocrystals with optimal facet compositions could lead to significant improvements in activity. Lastly, understanding the interaction between photocatalysts and various pollutants would contribute to the precise mechanisms of charge transfer and separation and enable the formation of new pathways for enhancing photocatalytic efficiency.

By addressing these challenges and exploring new horizons in photocatalytic material design, the field can make significant progress towards developing highly efficient, stable, and versatile photocatalysts for a wide range of applications.