TiO₂ Photocatalysis for the Transformation of Aromatic Water Pollutants into Fuels

Abstract

The growing world energy consumption, with reliance on conventional energy sources and the associated environmental pollution, are considered the most serious threats faced by mankind. Heterogeneous photocatalysis has become one of the most frequently investigated technologies, due to its dual functionality, i.e., environmental remediation and converting solar energy into chemical energy, especially molecular hydrogen. H₂ burns cleanly and has the highest gravimetric gross calorific value among all fuels. However, the use of a suitable electron donor, in what so-called “photocatalytic reforming”, is required to achieve acceptable efficiency. This oxidation half-reaction can be exploited to oxidize the dissolved organic pollutants, thus, simultaneously improving the water quality. Such pollutants would replace other potentially costly electron donors, achieving the dual-functionality purpose. Since the aromatic compounds are widely spread in the environment, they are considered attractive targets to apply this technology. In this review, different aspects are highlighted, including the employing of different polymorphs of pristine titanium dioxide as photocatalysts in the photocatalytic processes, also improving the photocatalytic activity of TiO₂ by loading different types of metal co-catalysts, especially platinum nanoparticles, and comparing the effect of various loading methods of such metal co-catalysts. Finally, the photocatalytic reforming of aromatic compounds employing TiO₂-based semiconductors is presented.

1. Introduction

Water is essential for the existence of all living beings. However, its pollution with organic and inorganic compounds remains a threat and poses great risks to the environment and human health. The water quality is merely a concept reflecting the kind and quantity of contaminants contained in it. Mining and petrochemical industries are instrumental in the economic growth of many countries and their products are regarded as privileges to modern communities [1]. However, the wastes generated from the activities of these industries are toxic and carcinogenic [2]. Thus, these wastes have been classified as “hazardous” [3], and there is a constant increase in the pollution concerns associated with various petrochemical compounds and their by-products in the form of water, air, and soil pollution. Many of these by-products are still extensively employed, especially in the chemical, medical, and other industrial fields, as irreplaceable and inevitable raw materials [4,5,6,7]. Aromatic compounds, such as benzene, phenol, and chlorobenzene, are some of the most encountered volatile organic compounds (VOCs). The primary sources of VOCs are originated from a large number of anthropogenic activities, such as refinery streams, especially from catalytic reforming and cracking, and petroleum refining, petrochemical processing, and solvent use [8,9]. Other VOCs, such as methane and chlorofluorocarbons, are classified as “greenhouse gases”, which cause global warming.

The aromatic ring is the basic constituent of many organic pollutants, such as polyaromatic hydrocarbons (PAHs), dyes, pesticides, and pharmaceuticals. Aromatic compounds, such as benzene, phenols, and benzoic acid, are the most frequently used model substrates to investigate the photocatalytic mechanism and to test the activity of the photocatalysts [10,11,12,13]. Detailed studies have been made on the harm caused by the aromatic compounds, for example, the potential relationship between the benzene-related compounds and the risk of hematologic cancers, such as lymphoid malignancies [14]. Moreover, long-term exposure to a low concentration of such compounds could predispose to the development of type 2 diabetes (T2D) and affect human metabolism [15,16]. Aromatic organic compounds also contribute to serious environmental problems, such as water pollution, which may result in the demise of scarce species, and biological genetic variation, which in many cases is an irreversible problem [17,18].

Semiconductor photocatalysis has been extensively studied in the past 30 years as a promising method of environmental cleanup and sterilization. However, an earlier description of the photocatalytic properties of some metal oxides was given in 1955 by Markham [19], who dealt with ZnO, Sb₂O₃, and TiO₂, and the various types of photochemical changes that these oxides could undergo, including the catalyzed oxidation of organic compounds under UV irradiation. Later, and in 1972, a short note published in Nature by Fujishima and Honda [20] demonstrated that water could be photolyzed electrochemically at an illuminated TiO₂ and Pt electrode combination to yield stoichiometric quantities of H₂ and O₂. What followed soon thereafter was a frenzied series of studies in search of the photocatalytic materials to produce H₂ fuel as part of the beginnings of the hydrogen economy. Many semiconductors showed photocatalytic properties like MoS₂ [21], WO₃ [22], BiFeO₃ [23], Fe₂O₃ [24], and CdTe [25,26], but only a few of them fulfill the thermodynamic requirements for overall water splitting, such as KTaO₃, SrTiO₃, TiO₂, ZnS, and SiC. What seems a simple basic function like the excitation of the semiconductor by absorption of light results in the formation of the charge carrier, i.e., the valence band holes, $h_{vb}^{+}$, and the conduction band electrons, $e_{cb}^{-}$. However, because of the strong oxidation ability of $h_{vb}^{+}$ and reactive oxygen species like ^•OH, ^•OOH, and H₂O₂, which are formed from the $h_{vb}^{+}$ oxidation of H₂O and $e_{cb}^{-}$ reduction of O₂, most organic compounds can be oxidized, even mineralized to CO₂ and H₂O in the photocatalytic systems. Due to the challenges that remain in achieving overall pure water splitting, the alternative approach is to combine light-induced splitting of water and photooxidation of organic substrates into a single process in so-called photocatalytic reforming [27].

TiO₂ is one of the most studied photocatalysts, it was greatly debated in many aspects like the nature of the oxidative agent (${}^{•}OH$ radicals vs. $h^{+}$), the site at which the reaction takes place (surface versus bulk solution), how to improve performance, and efficiency of TiO₂ photocatalytic properties. Unfortunately, TiO₂ has a relatively large bandgap (3.2 eV) so that only UV radiation can activate it, its conduction band is somewhat positive in relation to the redox potential for H₂ evolution. Clearly, new semiconductor-based nanostructured materials are needed that would use lower-energy photons available in the visible spectral region. A strategy developed mostly in the past decade was to push the absorption onset of TiO₂ toward longer wavelengths (≥387 nm) by doping TiO₂ with anions and/or cations (N, C, S, F,..., and metal ions) [28]. Despite the many studies carried out with TiO₂, no other metal oxide has yet been found that might act as an efficient photoanode with conduction and valence band edges that straddle the redox potentials of water and many water organic pollutants.

In the first part of this review, we discuss the different classes of aromatic organic hydrocarbons pollutants with an emphasis on their chemical, structural, and physical properties. Their environmental hazards, health threats, and their ability to leak into the different components of the environment especially the aquatic environment are also discussed. In the following parts of this review, we provide the reader with an overview of the conventional methods adopted for the treatment of organic pollutants against the recent methods used and the new trends to be evaluated and debated. The fundamental and electronic structure of semiconductor–photocatalyst, the significant parameters affecting its performance, and photocatalytic water splitting versus photocatalytic reforming are reviewed in detail. The next part of this review is dedicated to TiO₂, which is one of the most widely studied semiconductors, with emphasis on the photocatalytic properties of its different phases and how to enhance the performance of pristine TiO₂ by the loading of noble metals. Finally, and based on the recent research investigations, the current perspective for photocatalytic reforming of aromatic-based pollutants towards H₂ production and water decontamination is reviewed and highlighted.

2. Aromatic Hydrocarbons as Water Pollutants

Many pollutants discarded into the environment contain non-degradable substances like heavy metals and organic pollutants [29,30,31]. The persistent organic pollutants, such as pesticides [32], aromatic organic compounds (OCs) [33,34], semi-volatile organic compounds (SVOCs) [35,36,37], and organic dyes [38,39] are gaining great environmental concerns due to their impacts on health and environment. These compounds have grasped much attention due to their carcinogenic potential and ubiquitous presence in the environment, which pose a major threat to water reservoirs and the surrounding ecosystem.

The aromatic organic compounds, such as benzene, toluene, ethylbenzene, and xylenes (BTEX), polyaromatic hydrocarbons (PAHs), phenols, and their derivatives are frequently detected in different wastewater resources [40]. The removal of these organic pollutants is a must to reuse this water since such pollutants cannot be eliminated efficiently during conventional treatment processes. The reuse of this treated water, especially in irrigation of crops, contains great risks due to the transfer of these pollutants to plants, thus, through the food chain to living organisms [32,35]. The following subsections provide a brief description of the main types of these pollutants.

2.1. Phenols

Phenolic compounds are a class of organic compounds that consists of a hydroxyl group(s) directly bonded to one or more aromatic rings. The phenolic compounds are classified as priority pollutants due to their carcinogenic, mutagenic properties, and high toxicity even at low concentrations [41]. These compounds represent serious threats to human health, e.g., skin and eye irritations, anemia, respiratory, and vertigo [42]. The Environmental Protection Agency (EPA) sets the level standard of phenols in the surface water to less than 1 µg·L⁻¹, while, the toxicity levels for both humans and aquatic life are usually in the range 9–25 mg·L⁻¹ [41]. Phenols are one of the main intermediates for household and industrial productions of cleaners, dyes, pesticides, herbicides, paint, pharmaceuticals, petrochemicals, cooking operations, resin manufacturing, plastics, pulp, paper, and wood products [33,42]. They are usually detected in the wastewater effluents in very high concentrations up to thousands of mg·L⁻¹ [43].

The first member of this category of organic compounds is phenol with the chemical formula of C₆H₅OH. All other members of the group are derivatives of phenol [44]. Chlorophenols are the largest group and most spread group of phenols. They are formed in the environment by chlorination of mono and polyaromatic compounds present in soil and water [42]. Moreover, the reaction of hypochlorite with phenolic acids during the treatments and disinfection processes leads to the formation of such chlorinated compounds. Chlorinated phenols, e.g., 2-chlorophenol, 2,4-dichlorophenol, 2,4,6-trichlorophenol, and pentachlorophenol (Chart 1), are listed by the U.S. EPA as priority organic pollutants [45,46].

2.2. Polyaromatic Hydrocarbons

Another kind of organic pollutants that causes water contamination is the PAHs, which are classified as hazardous persistent environmental pollutants [47]. They are a group of over 100 different organic compounds containing two or more fused aromatic benzene rings connected linearly, angularly, or in a cluster arrangement [48,49]. PAHs are found naturally and released into the environment by anthropogenic sources. The incomplete combustion of wood, coal, oil, gas, garbage, and other organic substances, pyrosynthesis or pyrolysis of hydrocarbons (petrogenesis), and the leakage of crude oil and refined petroleum products are considered the main sources of the PAHs [47,50,51]. The surface runoff from roads is another major source of the PAHs in the aquatic system [52,53]. Surface-active compounds and humic substances increase the solubility of PAHs several times. Huang and Buekens [54] reported the formation of the PAHs under insufficient combustion conditions of the aliphatic fuels. Under these conditions, carbon containing-compounds are not oxidized completely to carbon dioxide, rather, hydrocarbon fragments that are generated during incomplete combustion interact with each other to yield complex polycyclic structures. Many other resources for the PAHs [49,53] are shown in Figure 1.

PAHs have low aqueous solubility and are considered as lipophilic organic compounds that are widely distributed in the environment and characterized by their high toxicity, genotoxicity, and carcinogenicity. [49,55].

Table 1. Physiochemical properties of the 16 U.S. Environmental Protection Agency (EPA) PAHs [53,72,73].

Compound Name	Chemical Structure	Chemical Formula	Number of Rings	Molecular Weight (g\mol)	Melting Point (°C)	Boiling Point (°C)	Aqueous Solubility (mg/L)	Vapor Pressure (Pa)	Log Kow
Naphthalene		C10H8	2	128.17	80.26	218	31	1.0 × 102	3.37
Acenaphthene		C12H10	3	154.21	93.4	279	3.8	3.0 × 10−1	3.92
Acenaphthylene		C12H8	3	152.19	92–93	265–275	16	9.0 × 10−1	4.00
Fluorene		C13H10	3	166.22	116–117	295	1.9	9.0 × 10−2	4.18
Anthracene		C14H10	3	178.23	218	340–342	0.045	1.0 × 10−3	4.54
Phenanthrene		C14H10	3	178.23	100	340	1.1	2.0 × 10−2	4.57
Fluoranthene		C16H10	4	202.25	110.8	375	0.26	1.2 × 10−3	5.22
Pyrene		C16H10	4	202.25	156	393–404	0.13	6.0 × 10−4	5.18
Benzo[a]anthracene		C20H12	4	228.29	158	438	0.011	2.8 × 10−5	5.91
Chrysene		C18H12	4	228.29	254	448	0.006	5.7 × 10−7	5.91
Benzo[b]fluoranthene		C20H12	5	252.31	168.3	No data	0.0015	-	5.80
Benzo[k]fluoranthene		C20H12	5	252.31	215.7	480	0.0008	5.2 × 10−8	6.00
Benzo[a]pyrene		C20H12	5	252.31	179–179.3	495	0.0038	7.0 × 10−7	5.91
Dibenzo[a,h]anthracene		C22H14	6	278.35	262	No data	0.0006	3.7 × 10−10	6.75
Benzo[ghi]perylene		C22H12	6	276.33	273	550	0.00026	1.4 × 10−8	6.50
Indeno[1,2,3-cd]pyrene		C22H12	6	276.33	163.6	530	0.00019	-	6.50

shows the physicochemical properties of 16 compounds of the PAHs that have been listed as priority pollutants by the United States Environmental Protection Agency [50,56]. PAHs of two and three aromatic rings, e.g., naphthalene and anthracene, are known as low molecular weight (LMW). Those compounds possess higher solubility in water and higher volatility than that of the high molecular weight (HMW) PAHs [57]. In fact, higher concentrations of the LMW PAHs have been reported in wastewater influent and effluent comparing to the HMW PAHs, which can be related to their higher water solubility [35,40,58,59,60,61].

Naphthalene (C₁₀H₈) is the simplest form of PAHs and possesses higher volatility besides its higher solubility in water (31.7 mg·L⁻¹ at 25 °C) compared to other PAH compounds. Naphthalene is widely used in industry as an intermediate in the production of pesticides, phthalic anhydride, dyes, resins, and surfactants [62,63]. Moreover, it is found in many consumer products like mothballs and some insect repellent products that are used to kill moths in airtight spaces, and to repel vertebrate pests in attics and wall voids spaces [64]. In general, naphthalene was found the most ubiquitous and abundant PAH in wastewater with concentrations ranged between ng·L⁻¹ to µg·L⁻¹ [47,59,60,61,65,66].

2.3. Organic Dyes

Dyes are colored substances that have an affinity for the substrate to which they are being applied. They have colors due to their absorption of light at a certain wavelength in the visible range. Due to their high molar extinction coefficients, a small amount of dye in an aqueous solution can produce a vivid color [67,68]. Synthetic dyes possess very different chemical and physical properties. Azo, anthraquinone, xanthene, indigoid, triphenylmethane, and phthalocyanine derivatives are the most frequent chemical classes of dyes employed in the industry (Chart 2) [67,68,69,70,71].

Synthetic organic dyes are introduced in the aquatic environment [74,75] because of their extensive usage in printing, paint, and textile industries. These compounds are characterized especially by their non-reactivity, long-lasting coloring, and highly stable structures [74]. Besides their carcinogenic effect, many dyes affect human life, such as dysfunction of the central nervous system (CNS), kidney, reproductive system, brain, and liver [74,76,77]. Wastewaters from textile and other dyes industrial processes contain large quantities of these organic pollutants, which are difficult to degrade during the standard biological methods and resist aerobic degradation. Moreover, Due to their high solubility in water, the removal of the dyes from wastewater through conventional methods is very difficult and ineffective [71,78]. Degradation of certain types of dye produces more hazardous pollutants than the dye itself. For example, under anaerobic conditions, organic dyes, such as azo dye, can be reduced to potentially carcinogenic aromatic amine [68,79].

3. Methods of Treatment

As outlined beforehand, toxic organic pollutants are widespread in the environment, thus, it is highly recommended to eliminate or reduce the concentration of such pollutants in the aquatic environment to safe levels [53,80,81,82,83]. Numerous conventional treatment processes have been applied and tested for wastewater treatment, such as adsorption, coagulation, precipitation, biodegradation, ozonation, electrochemical oxidation, and advanced oxidation processes [43,47,83,84,85,86,87,88]. Besides that, combining some of these processes, such as the biological–physical, or chemical processes, has been successfully applied in many wastewater treatment plants [89,90,91,92]. Although many of these processes have been considered effective and efficient for removing a wide spectrum of organic pollutants from the wastewater; however, each process has disadvantages that limit the large-scale application, e.g., small capacity, high costs, pH-dependency, limited recyclability, high-energy requirements, incomplete pollutant removal, and generation of toxic secondary materials (

Table 2. Different removal techniques used for wastewater treatment and their advantage(s) and disadvantage(s).

Removal Techniques	Advantage(s)	Disadvantage(s)
a)  Coagulation	▪ The additive coagulants easily settled with the suspended particle. ▪ Rapid and efficient for insoluble contaminants. ▪ Low-cost operation.	▪ pH monitoring of the effluent. ▪ The dissolved organic pollutants are not completely removed. ▪ Formation of sludge and secondary pollutants.
b)  Electrochemical oxidation	▪ Recycling of valuable metals. ▪ Increases biodegradability. ▪ Not require auxiliary chemicals or high temperatures.	▪ Required pre-filtration; formation of sludge. ▪ High initial cost of the equipment ▪ Low selectivity and low reaction rates.
c)  Biological process	▪ Simple, economically attractive. ▪ Ecologically favorable process.	▪ Poor decolorization. ▪ Formed uncontrolled degradation products. ▪ High capital and operational cost. ▪ The secondary sludge problems.
d)  Adsorption	▪ Cost-effective and simple method. ▪ The most profitable process and more efficient than the conventional methods (i.e., precipitation, solvent extraction, membrane filtration, etc.).	▪ Removes the pollutants from one phase (aqueous) to another (solid matrix). ▪ Expensive regeneration process especially if the pollutants are strongly bound to the adsorbents.
e)  Chemical precipitation	▪ Adapted to high pollutant loads. ▪ Simple equipment and processes.	▪ Chemical consumption. ▪ High sludge production.
f)  Advanced Oxidation Processes (AOP)
I. Ozonation	▪ Powerful oxidation technique for a large number of pollutants.	▪ Complex technology. ▪ High capital/operational cost. ▪ High electric consumption.
II. UV	▪ An effective method that typically does not produce harmful by-products.	▪ Low efficiency when the wastewater contains a high number of particulates that absorb UV light.
III. UV/H2O2	▪ An effective technique in the oxidation and mineralization of most organic pollutants. ▪ Ease formation of OH• radicals.	▪ Less effective, when the wastewater has high absorbance. ▪ High operational cost.
IV. O3/UV/H2O2	▪ Most effective process due to the fast generation of OH• radicals. ▪ Can treat a wide variety of contaminants.	▪ Needs to compete with high turbidity, solid particles, and heavy metal ions in the aqueous stream. ▪ High operational cost.
V. Fenton reaction	▪ Simple process. ▪ Easy availability of chemicals.	▪ Production of iron sludge waste, bringing logistical problems with handling.
VI. Photo-Fenton reaction	▪ Reduction of sludge iron waste compared to the original Fenton reaction. ▪ Effective and fast degradation.	▪ Requires a controlled pH medium for better performance.
VII. Heterogeneous photocatalysis	▪ Long-term stability at high temperature. ▪ Resistance to attrition. ▪ Low-cost and environmentally benign treatment technology.	▪ Formation a harmful byproduct to the environment. ▪ Requires efficient catalysts that can absorb a wide range of light.

) [87,93,94,95,96].

Among many developed and examined methods for eliminating the persistent organic pollutants from the environment (especially aquatic environment), advanced oxidation processes (AOPs) are the most promising techniques. They are also the most studied and the best environmental-friendly techniques for removing these pollutants. These processes are based on the formation of in-situ highly reductive or oxidative free radicals, e.g., hydroxyl radicals (OH^•), at sufficient concentration to effectively mineralize the hazardous organic compounds and decontaminate water under ambient conditions [95,97]. Several AOP techniques have been explored to decompose the organic pollutants in the water resources by chemical oxidation or reduction such as ozonation, H₂O₂ photolysis, Fenton process, photo-Fenton process, and heterogeneous photocatalysis [97,98,99,100]. Scheme 1 shows the types and the general classification of the AOP [100].

By far, heterogeneous photocatalysis has gained the most attention as one of the most realistic and viable solutions due to its ability to clean-up a wide range of environmental pollutants besides the use of low-cost and chemical stable photocatalysts [101,102,103,104,105,106,107,108]. This promising approach relies on the excitation of a semiconductor with suitable light, e.g., the sunlight, to drive different redox reactions. Heterogeneous photocatalysis is a process that includes a large variety of reactions, such as oxidation, dehydrogenation, water splitting (reduction, H₂ production; oxidation, O₂ production), organic synthesis, photoreduction, metal deposition, hydrogen production, gaseous pollutant removal, and water purification [109]. Photocatalysis is a sustainable and economical technology that can exploit the inexhaustibly abundant clean energy of the sun [104,109,110]. The use of an efficient nanoparticulate semiconductor is required for the detoxification of the wastewater via photocatalysis, which has the potential to degrade the toxic substances in the water, such as contaminants and microorganisms [111,112,113]. Due to their narrow bandgap and distinct electronic structure (unoccupied conduction band and occupied valence band) [20,114,115], various kinds of photocatalysts, including TiO₂, Gr-TiO₂, CdS, SnO₂, WO₃, SiO₂, ZnO, Nb₂O₃, Fe₂O₃, have been studied to degrade a variety of organic and inorganic pollutants [102,115,116,117].

Nowadays, the photooxidation of organic pollutants based on TiO₂ nanomaterials is still gain huge attention. Several recent studies have shown the effective role of TiO₂ in oxidizing and mineralizing a wide range of hazardous organic contaminants [118,119], such as alcohol [120,121], organic acids [122], aromatic hydrocarbons [104], phenols [123], dyes [124], pharmaceuticals [125,126], and pesticides [126]. The photocatalytic oxidation of organic pollutants proceeds either by the direct attack of the photogenerated holes or via the attack of the highly reactive hydroxyl radicals generated at the surface of the photocatalyst [110]. The photocatalytically generated OH^• radicals can abstract hydrogen atoms from the organic molecules, causing a chain of reactions toward lower molecular intermediates and may end up in the complete mineralization of the pollutants [127].

Another major field in photocatalysis includes light-driven water splitting into H₂ and O₂ [128]. H₂ is regarded as the most recommended replacement for fossil fuels since its energy cycle is free of pollutants and greenhouse gases [105,129,130]. Achieving dual-functional photocatalysis, i.e., the photocatalytic degradation of organic pollutants and the simultaneous production of hydrogen gas is an added value of this technique [131]. Unfortunately, as will be discussed in the following sections, different operational conditions should be applied for each process to achieve its optimal reaction yield [11].

4. Semiconductor-Based Heterogeneous Photocatalysis

A photocatalytic system is thermodynamically defined as a system, in which a reaction with ΔG < 0 is driven through the photon absorption by a suitable material, i.e., the light energy is exploited to drive a reaction having extremely low kinetics outside this system [132]. The photons are absorbed by such a system to generate accordingly charge carriers, i.e., electrons and holes, which induce a redox reaction. The semiconductors could be the light-absorbing materials in heterogeneous systems and they are then known as photocatalysts [133]. Thus, heterogeneous photocatalysis depend on the distinctive properties of powdered semiconductor materials in harvesting incident light, generating charge carriers, and subsequently initiating surface reactions. This may provide a simple means for environmental remediation and photochemical energy conversion into fuels [134,135].

4.1. The Electronic Structure of a Semiconductor–Photocatalyst

The band model based on the concept of molecular orbitals can be used to explain the electronic structure of a semiconductor. The electronic orbitals merge and split into two bands, i.e., the valence band (VB) and the conduction band (CB). VB and CB of a semiconductor are formed from the highest occupied molecular orbitals (HOMO) and the lowest unoccupied molecular orbitals (LUMO), respectively [136]. At temperature 0 K, the VB is the lower band that is completely filled with electrons, while the CB is the higher band that is empty [137]. The difference in energy between the highest energy level in the VB and the lowest energy level in the CB creates a region known as the energy bandgap (E_g) [138]. The interaction between the electronic orbitals, forming the band structure of a semiconductor is shown in Figure 2. Interestingly, the E_g values for semiconductors are sufficiently small that the electrons promotion from the VB to the CB can be initiated through an energy transfer to those materials [132]. The light energy that is higher or equal to E_g induces the excitation of electrons from the VB to occupy partially filled states in the CB generating an electron vacancy in the VB, which is known as the positively charged hole [139]. This hole is considered as a mobile entity since it can be filled by another electron creating a vacancy in the space where it has been transferred from [140]. The electrons in the CB are, likewise, mobile entities having often higher mobility than those of the holes (e.g., for Si, µ_n(electrons) = 1500 cm²·V⁻¹·s⁻¹ > µ_p(holes) = 450 cm²·V⁻¹·s⁻¹ [141]). Electrons have consequently a higher diffusion coefficient than holes; however, the trapping of the electrons leads to a decrease in their mobilities [142]. The e⁻/h⁺ species migrate then to the surface of the semiconductor, where they can react with the adsorbed molecules. The photogenerated holes act as oxidants (+1.0 to +3.5 V vs. NHE), while the photogenerated electrons are potential reductants (+0.5 to −1.5 V vs. NHE) [143].

Semiconductor photocatalysis is considered, from this point of view, as a multi-step process, which is illustrated in Figure 3. Such a process is initiated by the photoexcitation with electromagnetic radiation equal to or exceeding E_g (1), the separation of the charge carrier pairs (2), the diffusion of e⁻/h⁺ species within the material towards the surface, and the surface charge transfer for the reduction of adsorbed electron acceptors (3), and the oxidation of adsorbed electron donors (4), respectively [145,146]. Accordingly, the photo-induced electrons and holes should migrate to reach the surface of the material and react with adsorbed chemical species via surface charge transfer [144]. Therefore, the E_g of a semiconductor is the minimum thermodynamic requirement for photocatalysis [147,148,149].

One of the main limitations of semiconductor photocatalysis is the recombination of the photogenerated charge carriers, dissipating the absorbed energy as heat [138] and affecting negatively the lifetime of the electrons and holes [144]. This undesired recombination occurred either indirectly, i.e., via surface defects (5), or directly, i.e., by band-to-band recombination (6). Such phenomena are highly reliant on the crystal structure of the semiconductor. To enhance effectively the redox reactions while minimizing recombination, the photogenerated charge carriers must migrate to the liquid junction through the solid and should react with adsorbed species directly at the semiconductor surface [148].

4.2. Photocatalytic Water Splitting vs. Photocatalytic Reforming

Photocatalytic H₂ production from water splitting is accomplished under ambient operating conditions and consists of two half-reactions as shown in Equations (1) and (2), i.e., the reduction of proton and the 4-electron oxidation of water, respectively [150]. A change in free energy of ΔG⁰ = 237.2 kJ mol⁻¹ is associated with the splitting of one H₂O molecule to H₂ and ½ O₂, which equals to ΔE° = 1.23 V according to the Nernst equation [151]. Thus, the semiconductor should absorb photon energy of more than 1.23 eV (wavelengths shorter than 1000 nm) to drive the water splitting photoreaction. The semiconductor can use their photogenerated electrons/holes to convert the photon energy into H₂ and O₂ when the energy of the conduction band-edge and the valence band-edge straddle the electrochemical potentials E° (H⁺/H₂) and E° (O₂/H₂O), respectively [150,152] (Figure 4). Accordingly, the thermodynamic requirement for the water splitting is more cathodic and more anodic energy levels of the CB bottom and VB top of a photocatalyst compared to the standard electrode potential of (H⁺/H₂) and (O₂/H₂O), respectively [152,153]. Therefore, from a thermodynamic point of view, only a few photocatalysts, e.g., TiO₂, are proficient to drive the water splitting reaction. However, the efficiencies of heterogeneous photocatalytic water splitting remain relatively low due to many reasons outlined in the next sections.

2H⁺ + 2e⁻ → H₂ (↑)    E⁰ = 0 V vs. NHE

(1)

H₂O → ½O₂ (↑) + 2H⁺ + 2e⁻  E⁰ = 1.23 V vs. NHE

(2)

As a hybrid field, dual-functional photocatalysis is a combination of different photocatalytic fields for 2-fold purposes achieved in a single step [130]. The coupling of H₂ evolution and photocatalytic degradation of organic pollutants yielding CO₂ can be achieved in the so-called photoreforming process [38,132,153,154,155,156]. Such a technique has a great advantage as it can benefit from solar light to treat wastewater, meanwhile, the evolved CO₂ can be consumed by natural photosynthesis [157]. In the photocatalytic reforming process, the photogenerated holes in the valence band can oxidize adsorbed organic substrates (electron donors or sacrificial reagents), whereas the photogenerated electrons in the conduction band can reduce the protons (electron acceptor) to molecular hydrogen [104,105,130,158,159]. Such an adsorbed organic substrate can react irreversibly with the photogenerated holes, minimizing the undesired electron/hole recombination [150].

Even though H₂ can be formed simultaneously with other processes, e.g., the organic synthesis of organic compounds [160,161], however, such processes should not be considered as a dual function process [130]. The reforming process can be considered as a dual function photocatalysis process only when some requirements have been met: (i) H₂ is mainly derived from the reduction of water and (ii) the targeted organic molecules are pollutants, or their oxidation aimed to synthesize value-added products, such as aldehyde, organic acid, and imine [130,162]. Consequently, photocatalytic reforming is an intermediate process between photocatalytic water splitting and the photocatalytic oxidation of organic pollutants as shown in Figure 5.

Organic substrates are generally stronger reducing agents than water, hence, a less positive potential is necessary to oxidize these compounds. Accordingly, the energetic separation of the redox half-reactions in photoreforming is narrower compared to that of the overall water splitting [163]. As O₂ is not produced in these systems, the back reaction to produce water is suppressed, avoiding a subsequent gas separation stage [164]. A wide range of organic compounds such as alcohols, organic acids, and hydrocarbons had proven activity as electron donors for photocatalytic H₂ production [101,104,105,131,150,165,166,167]. The evolution of H₂ and its kinetic reactions pathway is dependent on the concentration and the nature of the organic substrate [168,169].

4.3. Titanium Dioxide (TiO₂) as a Photocatalyst

Titanium dioxide has been one of the most widely studied semiconductors in the last decade for various photocatalytic applications [170]. This is related to its high reactivity, hydrophilicity, low cost, and availability, physical and chemical stability, resistance to photocorrosion, and optimal electronic and optical capacity [97,111,139]. TiO₂ is a transition-metal oxide semiconductor composed of Ti⁴⁺ atoms and six O²⁻ coordinated together to form a TiO₆ octahedron [171]. Like other transition metal oxides, TiO₂ is often nonstoichiometric with oxygen vacancies (O_v) as predominant defects at the near-atmospheric oxygen pressure, granting it the properties of an intrinsic n-type semiconductor [172]. The oxygen vacancies (O_v) at the surface of n-type TiO₂ appear as extra unpaired electrons in the CB [164,173], which act as donor-like states. This creates an accumulation layer in the surface, resulting in a downward band bending [174].

The photocatalytic activity of TiO₂ is highly related to its charge carrier dynamics. The e⁻/h⁺ pairs are generated within a few femtoseconds upon irradiation and they can recombine easily either in the bulk or at the surface. However, other charge carriers escape recombination and migrate to the surface, where they might be trapped before the interfacial charge transfer in redox reactions [175]. Figure 6a and Equations (3)–(6) demonstrate the potential fates of charge carriers upon the irradiation of TiO₂, while Figure 6b reports the time scale of each process [143,176].

Photogeneration of the charge carriers:

TiO₂ + hv → TiO₂ (e⁻_CB/h⁺_VB)

(3)

Trapping of the charge carriers:

e⁻_CB + ≡Ti^IVOH → ≡Ti^IIIOH (shallow trapping)

(4a)

e⁻_CB + ≡Ti^IV → ≡Ti^III (deep trapping)

(4b)

h_VB⁺ + ≡Ti^IVOH → {Ti^IVOH^•}⁺ + (h⁺_tr)

(4c)

Recombination of the charge carriers:

e⁻_CB + {Ti^IVOH^•}⁺ → Ti^IVOH

(5a)

h+_tr + ≡Ti^IIIOH → Ti^IVOH

(5b)

h⁺_tr + Ti^III → Ti^IV

(5c)

Interfacial charge transfer to the acceptor (A) or donor (D) adsorbed on the surface:

≡Ti^IIIOH + A → ≡Ti^IVOH + A^•−

(6a)

{Ti^IVOH^•}⁺ +D → ≡Ti^IVOH + D^•+

(6b)

Serpone et al. found that in the absence of scavengers, more than 90% of the initially formed charge carriers recombine rapidly within 10 ns upon the irradiation of TiO₂ in aqueous media. Such high recombination results in less than 10% quantum yields of photooxidation [177].

On the other hand, the photogenerated charge carriers can be trapped either in the bulk or at the surface as trapped holes and trapped electrons, with the surface trapping being preferred for the subsequent interfacial charge transfer reactions [178]. Yoshihara et al. showed in their Transient Absorption Spectroscopy study that both trapped holes and electrons are found to be localized at the surface of photoexcited TiO₂ particles, while free electrons are distributed in the bulk [179]. Howe and Gratzel demonstrated in their Electron Paramagnetic Resonance spectroscopy (EPR) studies on irradiated TiO₂ that the photogenerated electrons are localized in the d orbitals of Ti⁴⁺ while the photogenerated holes are trapped at the lattice oxygen atoms, forming EPR-active paramagnetic centers, i.e., Ti³⁺ and O^•−, respectively [180,181]. Simultaneously, upon the generation, separation, and transport of charge carriers in TiO₂, e^–/h⁺ pairs might participate in redox reactions via interfacial charge carrier transfer. In aqueous media, water layers adsorb, physically and chemically, on the TiO₂ surface creating a TiO₂/H₂O interface [182]. The photogenerated holes can react on the surface either with hydroxyl groups or with H₂O resulting in the formation of hydroxyl radicals, OH^•. Therefore, not only h⁺ is produced by the photoexcitation of TiO₂ but also hydroxyl radicals can be formed on hydrated TiO₂ surfaces.

TiO₂ has three main crystal phases: anatase, rutile, and brookite. While anatase and rutile exhibit the same tetragonal crystal structures, brookite has an orthorhombic crystal structure. These three polymorphs have also different E_g values, i.e., 3.2 eV, 3.0 eV, and 3.3 eV for anatase, rutile, and brookite, respectively [30]. Anatase has been generally considered as the most active phase of the three TiO₂ polymorphs for photocatalytic applications [183,184].

Anatase and rutile TiO₂ have shown differences in their respective charge carrier recombination kinetics [185,186]. Using transient absorption spectroscopy, Sachs et al. [185] compare the ultrafast charge carrier kinetics for anatase and rutile in dense and nanostructured TiO₂ films. They found that bulk rather than surface recombination was the key determinant of charge carrier lifetime. They also monitored that recombination was dependent on the crystal phase. Rutile shows faster recombination than anatase, which is consistent with the doping density (n-type doping due to oxygen vacancies) in rutile being higher than in anatase. Besides, Wang et al. [187] investigated anatase and rutile TiO₂ with photoluminescence spectroscopy under weak excitation conditions. Anatase showed a visible emission, while a NIR emission was reported in rutile; however, both emission spectra exhibited long lifetimes up to milliseconds. They explained that the NIR luminescence band in rutile TiO₂ was due to the recombination of trapped electrons with free holes. Hence, trap states in TiO₂ may play a very important role in the photocatalysis processes. The depth of trap states in rutile TiO₂ is much deeper than that in anatase TiO₂, which has shallowed-trapped electrons in addition to a higher number of free electrons as shown in Figure 7a,b. On the other hand, Durrant et al. [188] employed transient absorption spectroscopy (TAS) to investigate the kinetic of photocatalysis in anatase and rutile TiO₂ films. Although rutile exhibited 10 times slower recombination kinetics than anatase, mesoporous anatase film was around 30 times more efficient than mesoporous rutile film in the photocatalysis of the “intelligent ink” model system. They found also that in the presence of alcohols, faster and irreversible hole scavenging was achieved on anatase than in the case of rutile, resulting in the creation of long-lived electrons (τ ≈ 0.7 s). The authors explained the lower activity of rutile to the deficiency of rutile holes to drive efficient and irreversible alcohol oxidation rather than to the differences in recombination kinetics.

Choi et al. [189] compared the recombination kinetics in anatase and rutile using time-resolved diffuse reflectance (TDR) spectroscopy. They demonstrated that during the 355 nm laser excitation, the time-resolved decay at 550 nm was slower in anatase than rutile as shown in Figure 8a. This results in a longer lifetime of photogenerated charge carriers with subsequent higher reactive oxygen species generation in anatase. The authors observed also the generation and the diffusion of OH^• from the illuminated TiO₂ surface to the solution bulk using a single-molecule detection method. They found that only anatase generates mobile OH^• radicals, therefore, the photocatalytic oxidation on rutile is limited to adsorbed species. Schindler and Kunst [190] studied the excess charge carrier kinetics in anatase and rutile TiO₂ powders using the time-resolved microwave conductivity (TRMC) method. Figure 8b shows the transient change of the reflected microwave power after excitation by a 20-ns laser pulse at 266 nm. The photoconductivity in anatase decays very slowly compared to rutile powder. They demonstrated that this signal can be attributed to excess electrons in the CB because of the n-doping characteristics and the larger electron mobility compared to the hole mobility. Therefore, the short electron lifetime in rutile could be due to a higher recombination rate, while in anatase fast trapping of the minority charge carriers (holes) may take place. This would decrease the availability of holes for recombination and reduce the recombination probability, leading to a longer lifetime in anatase.

The development of TiO₂ materials has led to mixed-phase titania photocatalysts. One example is P25-TiO₂, which is a mixture of anatase and rutile (75:25). Due to its higher activity, anatase is conventionally considered to be the active component in P25, with rutile serving as an electron sink. Some reports showed that such mixed-phase titania has slower rates of charge carrier recombination, higher photo-efficiencies, and lower energy light activation [191]. Knorr et al. [186] studied the room-temperature photoluminescence spectra of nanocrystalline TiO₂ in the anatase and rutile phases and mixed-phase films. They showed that the photoluminescence of anatase results from at least two spatially isolated trap-state distributions, i.e., trapped electrons and trapped holes, which are, respectively, about 0.7−1.6 eV and 1.8−2.5 eV below the conduction band edge. The signal of trapped electrons was largely quenched in P25 and the presence of hole scavengers. The authors, hence, concluded a bidirectional electron transport between anatase and rutile phases in P25, with solvents having a strong impact on the competition for electrons between the two phases.

Additional recombination or trapping in the rutile part decreases the lifetime of electrons compared to pure anatase, but it would be much longer than in pure rutile due to the deep trapping of the holes in the anatase part. On the other hand, Hurum et al. [191] studied the charge separation characteristics of P25 by EPR spectroscopy. They showed, as presented in Figure 9, that the visible light irradiation of rutile produced charge carriers, which are stabilized through electron transfer to lower energetic trapping sites in the lattice of anatase. The authors suggested that the morphology of nanoclusters P25 consists of small rutile crystallites interwoven with anatase crystallites. The transition points between these two phases permit a rapid electron transfer from rutile to anatase. Hence, rutile in P25 acts as an antenna to extend the photoactivity into visible wavelengths and the structural arrangement creates catalytic “hot spots” at the rutile−anatase interface.

Using the time-resolved microwave conductivity (TRMC) method, Schindler and Kunst [190] found that the transient photoconductivity in P25 was rather more like the decay behavior observed in anatase than that in rutile shown in Figure 8b. They expected that in the mixed powder, fast recombination, such as in rutile for the electron–hole pairs created in the rutile part. Nevertheless, the deep trapping of holes in the anatase part would prevent the transfer of holes to the rutile part for the electron–hole pairs created in the anatase part.

4.4. Enhancing the Performance of Pristine TiO₂

As discussed above, although pristine TiO₂ exhibits advantages, some limitations are also presented. The main drawbacks to using pristine TiO₂ as an active photocatalyst are the lack of visible light activation, the fast recombination of the photogenerated electrons and holes, the relatively low charge carrier mobility.

Various attempts have been made to improve the capability to exploit visible photons for the TiO₂ photocatalytic process. Doping with transition metal ions is one approach that has been extensively employed, especially the incorporation of Fe³⁺ into the TiO₂ matrix [164]. This has been proven as a promising method to create additional states in the bandgap and, consequently, to an increase in the absorption of the visible light [102]. Moreover, it can introduce electron capture centers, resulting in a decrease in electron/hole recombination centers [164]. Compared to pristine TiO₂, Fe-doped TiO₂ has enhanced light-harvesting; however, controversial results on its photocatalytic activity have been reported [102]. Choi et al. [192] studied the photocatalytic oxidation of chloroform using TiO₂ doped with 21 transition metal ions and discovered that the doping with Fe³⁺, Mo⁵⁺, Ru³⁺, Os³⁺, Re⁵⁺, V⁴⁺, and Rn³⁺ cations is beneficial. Moreover, nonmetal doping has been widely studied, especially with N, C, F, B, and other elements having an atomic radius similar to that of the O atom. Among them, nitrogen has attracted much attention. Asahi et al. [193], for example, showed that nitrogen-doped TiO₂ exhibits enhanced visible light absorption and photocatalytic activity. Other strategies are the use of noble metals (e.g., Pt, Au, Pd, Rh, Ni, Cu, and Ag) as a co-catalyst to decrease the recombination of the charge carriers and provide additional active sites for H₂ evolution [102]. We will focus in the next sections on the modification of pristine TiO₂ with noble-metal co-catalysts, particularly platinum nanoparticles, due to their higher catalytic performance driving the reduction reaction of protons [104]; hence, increasing the photocatalytic reforming of organic compounds.

The energy of the photogenerated electrons in the conduction band for both rutile (E_CB = −0.11 V at pH 0) and anatase (E_CB = −0.32 V at pH 0) [145] is sufficient to form H₂ by reducing water. However, pristine TiO₂ has been reported as an inactive photocatalyst for H₂ production because of the fast recombination of charge carriers and the inability to reduce protons to H₂ due to the higher overpotential for hydrogen evolution reaction (0.05 V) [194]. Hence, even in the presence of an electron donor, pristine TiO₂ has shown an inability to catalyze the hydrogen evolution reaction [131]. TAS data revealed that the generated electrons are trapped as blue Ti³⁺ ions instead of reducing H⁺ upon the consumption of holes by the electron donor [105,195]. Consequently, it is highly recommended to modify pristine TiO₂ with an appropriate co-catalyst, which can effectively catalyze the cathodic H₂ evolution reaction. One successful strategy is the surface modification with noble metal nanoparticles, e.g., Pt and Au NPs. Noble-metal-modified TiO₂ photocatalysts have been widely studied in the literature, in which the noble-metal NPs act as Hydrogen Evolution Reaction (HER) catalysts.

HER on metallic platinum, as an example, induce via the Volmer reaction, in which H^•_ads atoms are produced when the accumulated electrons in Pt are transferred to the proton adsorbed H⁺_ads and H₂O_ads, respectively, as described in Equations (7a) and (7b) [196]. The reaction proceeds afterward through two possible pathways, either the Heyrovsky reaction (Equation (7c)) or the Tafel reaction (Equation (7d)), in which H^•_ads react with H⁺_ads or/and the direct recombination of two H^•_ads with each other, respectively [155]. Figure 10 illustrates the two-electron transfer reaction that occurs on the metal surface in acidic solutions. HER on Pt has been shown to exhibit pseudo-first-order kinetics, which indicates that the rate-determining step of the HER is the Volmer reaction [197]. Rabani et al. [197] have presented a linear increase in e⁻_TiO2 decay rate while increasing H⁺ concentration at a given Pt concentration, suggesting that H₂ is most likely generated by reduction of H⁺ rather than the reduction of H₂O. They have also highlighted that the presence of Pt is vital for reactions 7a and 7b to occur.

H+ads (Pt) + e−(Pt) → H•ads (Pt)

(7a)

H2Oads(Pt) + e−(Pt) → H•ads (Pt) + OH-

(7b)

H•ads (Pt) + e−(Pt) + H+ads (Pt) + e−(Pt) → H2(g)

(7c)

H•ads (Pt) + H•ads (Pt) → H2(g)

(7d)

It has been widely accepted that enhancement of the activity through the modification of TiO₂ with noble-metal NP_S is due to a better charge separation according to the Schottky barrier model. Noble-metal NPs have higher Fermi level energy, i.e., 5.65 eV and 5.10 eV for Pt and Au, respectively [199] compared to that of TiO₂, i.e., 4.2 eV [200]. Therefore, photogenerated electrons can transfer from TiO₂ to the metal NPs through the interface until a thermodynamic equilibrium is achieved [110] as shown in Figure 11a–d. Schottky barrier ΦB is defined as the barrier against the flow of electrons from the metal to the n-type semiconductor, i.e., TiO₂ [201]. During the irradiation, this thermodynamic equilibrium will be unsettled, permitting the photogenerated electrons to continuously flow from the CB of TiO₂ to the metal NPs [110,202]. It has been generally recognized that such a Schottky barrier smooths electron trapping by the metal, providing better charge separation. The trapped electrons have, therefore, a longer lifetime to promote the reduction reactions [203,204]. Correlations between photocatalytic H₂ evolution rates and metal work functions have been thoroughly established [105,205,206]. However, EPR experiments for irradiated Pt/TiO₂ revealed simultaneously signals for the Ti³⁺ centers, which confirms that the photogenerated electrons are not transferred completely to the Pt NPs, rather a certain number of them are trapped as Ti³⁺ ions in TiO₂ [105,110,181].

Scavenging the photogenerated electrons from TiO₂ by the noble-metal NPs is essential but is not the only factor that enhances the HER. According to the Sabatier principle [207], an ideal catalyst for (HER) is characterized by its optimal binding energy with adsorbed atomic hydrogen (H^•_ads). This binding energy should be neither too strong nor too weak. On the one hand, the active sites for the HER reaction can be blocked and the desorption of H₂ becomes the rate-limiting reaction in the case of a strong binding. On the other hand, proton reduction is rate-limiting in the case of weak binding energy with H^•_ads [208]. Consequently, a volcano-type dependence between HER rates and metal-H^•_ads bond strength has been proposed [209], in which platinum provides the best activity to drive the HER as shown in Figure 12. In conclusion, Pt/TiO₂ has been demonstrated to exhibit the highest photocatalytic activity towards H₂ production compared to other metal-loaded TiO₂ [105,155,210], such as Au/TiO₂. This has been explained by the highest work function of Pt that enhances electrons “sinking” properties, the lowest overpotential for H₂ formation, and the optimal binding energy adsorbing atomic hydrogen.

4.5. Effect of the Loading Method on H₂ Production

As discussed in Section 4.4, surface decoration of metal (e.g., Au, Ag, Cu, and especially Pt) on TiO₂ nanoparticles is an outstanding technique to revamp the electronic properties of TiO₂ without affecting its original crystallinity, thus, enhance the photocatalytic activity, and enrich the H₂ production efficiency [212,213]. Different co-catalyst loading methods “techniques” have been successfully applied [38,214,215]; however, the structure and the properties of the co-catalyst were found to play a critical role in achieving superior photocatalytic activity [216]. It has been reported that many structural factors affect the activity of the platinized TiO₂, such as the size of Pt NPs [213,217,218], their dispersion of Pt NPs [219,220], the interaction between the metal and the support [221,222,223], and the chemical state of Pt deposits [220,224]. Nevertheless, all of these factors can be optimized by using proper preparation methods [210,214,220,225,226].

The most commonly adopted techniques for the loading of Pt nanoparticles on the surface of TiO₂ include photodeposition [104,105,131,167,210,227], deposition–precipitation [210], chemical reduction [212,228], impregnation [227], electrodeposition [229], and physical mixing [210]. Some of these methods require adding a reducing agent, such as NaBH₄, to reduce the metal ions to metal particles. However, the weak adhering of the metal nanoparticles to the semiconductor surface, the larger size of the metal nanoparticles, and the nucleation of isolated metal nanoparticles in the electrolyte are the main problems associated with such methods. Such a poor interaction between the metal nanoparticles and the semiconductor surface negatively affects the electron transfer to the metal, increasing the electron/hole recombination rate [212,230,231]. On the other hand, some techniques need elevated temperatures or an applied bias, and a longer preparation period [128,210,214,232,233].

Photocatalytic hydrogen production over Eosin Y-sensitized Pt-loaded TiO₂–ZrO₂ mixed oxide photocatalysts was investigated under visible light irradiation by Sreethawong and Yoshikawab [234]. The authors prepared the platinized material by using two different methods, i.e., single-step sol-gel (SSSG) and photochemical deposition (PCD). At the optimum loading ratio (0.5 wt.%) of Pt, the authors found that the platinized photocatalyst prepared by the PCD method exhibited a higher H₂ production rate of 2.37 mL/h g comparing to 1.42 mL/h g to that prepared by the SSSG method. They attributed the difference in the photocatalytic activity to the different oxidation states of Pt in both samples. The loaded Pt nanoparticles synthesized by the SSSG method were partly in the oxide form, whereas those prepared via the PCD method consisted of particles in their metallic form having better dispersion on the surface of the semiconductor. Accordingly, the latter provided an efficient charge carrier separation at the interfacial contact between the photochemical-deposited Pt nanoparticles and the TiO₂–ZrO₂.

Alternatively, the photodeposition method is the most adopted and recommended technique among other loading methods to prepare Pt/TiO₂ [128,214]. The interest of the scientific community with the photodeposition method has been greatly expanded since 1978 when Kraeutler and Bernhard employed this technique to synthesize well-dispersed Pt nanoparticles on TiO₂ to use this composite in the photocatalytic decomposition of acetic acid to methane [214,235]. Many beneficial features can be controlled during the photodeposition method such as well-defining of co-catalyst nanoparticles, preparing facet-engineered nanoparticles, geometrical distributing of nanoparticles, controlling the size and the oxidation state of the deposited nanoparticles. During the photodeposition process, the metal ions are reduced by the conduction band photogenerated electrons, which leads to a uniform dispersion of the metal nanoparticles on the photocatalyst surface and avoids the self-nucleation of metal particles in the solution [236].

Several structural properties contribute to the photoactivity of the loaded photocatalyst, such as aggregation, Pt-assisted network formation, and Pt dispersion. Wang et al. [222] studied a 1 wt.% platinization of colloidal TiO₂ by two methods, i.e., the photodeposition and the mixing with colloidal Pt prepared by chemical reduction of Pt⁴⁺. The authors found that during the photocatalytic oxidation of methanol, the quantum yields of HCHO formation increased by 70% and 50%, for the photocatalyst prepared by photodeposition and mixing, respectively. Additionally, they showed that, in a deoxygenated system, the platinized-TiO₂ prepared by photodeposition method was more efficient for photocatalytic H₂ and HCHO formation than the other platinized sample prepared by physical mixing during the reforming of CH₃OH. The authors explained that Pt clusters on the TiO₂ surface were formed via the photodeposition process, while, Pt particles were surrounded by TiO₂ particles by mixing colloidal Pt with colloidal TiO₂, as shown in Figure 13. Therefore, the better activity of the former is attributed to the better dispersity and the stronger contact between the Pt particles with the TiO₂ surface.

Moreover, the deposition of Pt on the surface of TiO₂ enhances the optical property of the Pt-TiO₂. Chen et al. [237] reported that the deposition of Pt on TiO₂ surface via the photodeposition method promoted the optical absorption property of the prepared material to the visible region of light. The authors attributed this enhancement to the formation of Ti⁺³ due to the reduction of the Ti⁺⁴ during the photodeposition of the Pt. Similarly, F. Li and X. Li [224] found that the deposition of Pt nanoparticles on the TiO₂ surface enhanced the photocatalytic activity due to the formation of a defect energy level near the valence band of TiO_2, as a result of the Ti^III formation in the lattice. The authors attributed the formation of Ti^III to the interaction between Pt and TiO₂ during the photoreduction process.

Although many reports have shown that the photodeposition technique produces a high active photocatalyst system [210,212], several reports of the metal/semiconductor prepared with other techniques have claimed the contrary [213,225,238]; thus, no general conclusion could be deduced. Apparently, the shape and the nature of Pt NPs besides their interaction with the support are expected to be different by the various platinization methods, which results in diverse photocatalytic behaviors.

5. Photocatalytic Reforming of Aromatic Compounds

Aromatic compounds such as phenols, dyes, and PAHs are important industrial chemicals due to their wide usage. Therefore, the development of novel and simple processes is desired to remove these compounds from the environment, according to the viewpoint of “green chemistry” [239]. Several investigations of photooxidation of such pollutants have been carried out by using TiO₂ and Pt/TiO₂ photocatalysts in the presence of molecular oxygen [113,117,124,240,241,242,243,244,245,246].

Due to the low efficiency of overall photocatalytic water splitting, the photoreforming of the organic compounds has shown significantly higher rates and longer-term stability of H₂ production. Therefore, a huge number of photocatalytic reforming studies have been reported. However, simple organic compounds like methanol (the most studied), ethanol, formaldehyde, formic acid, etc., have been mostly used as model pollutants. In this section, we will focus on the reported investigations that using aromatic compounds like benzene, phenols, dyes, and PAHs as electron donors (hole scavengers), especially over modified TiO₂ materials for the same goal.

As mentioned in Section 4.2, surface-modified TiO₂ and other semiconductors, such as Cu₂O, WO₃, have been widely used as photocatalysts for the photooxidation and the reforming of the hazardous organic pollutants found in wastewater [38,105,130,247]. Upon the total mineralization, the photoreforming process is demonstrated by the following stoichiometric reaction (Equation (8)) [248].

$${\mathrm{C}}_{\mathrm{x}}{\mathrm{H}}_{\mathrm{y}}{\mathrm{O}}_{\mathrm{z} }+\left(2\mathrm{x}-\mathrm{z}\right){\mathrm{H}}_2\mathrm{O} \to {\mathrm{xCO}}_2+\left(2\mathrm{x}-\mathrm{z}+\mathrm{y}/2\right){\mathrm{H}}_2$$

(8)

5.1. Monoaromatic and Phenolic-Based Compounds

Benzene is considered a toxic and carcinogenic pollutant. It naturally exists in the environment and is artificially manmade through a wide range of products, such as plastics, paints, mucilage, rubber, and gasoline. It was confirmed that exposure to benzene for a high level or long-time results in several ailments, such as drowsiness, nausea, headache, lightheadedness, dizziness, and cancers. Therefore, it was classified in Category A as a carcinogenic compound by the Environmental Protection Agency.

Hashimoto et al. [247] investigated the photocatalytic H₂ production from different solutions of aliphatic and aromatic compounds using Pt-TiO₂ photocatalyst. Assisted by light energy and in the presence of the photocatalyst, both types of hydrocarbons produced H₂ by reacting with water at room temperature. The maximum H₂ formation was obtained at a 30:1 ratio between the water/benzene mixtures while increasing the benzene ratio decreased the H₂ formation. The authors have claimed that water is the main source of H₂ since no H₂ was detected upon the use of pure benzene in the presence of the Pt-TiO₂ under irradiation. Comparing to the water–alcohol mixture, the authors observed that H₂ and CO₂ are produced at an early stage of irradiation, and the aromatic hydrocarbons produced a higher CO₂ amount than their corresponding derivatives, such as phenol, hydroquinone, and catechol. Therefore, they suggested a higher reactivity of the aromatic hydrocarbons comparing to the hydroxylated aromatic compounds. The authors, hence, proposed that the direct oxidation of benzene by photogenerated holes is the main reaction pathway, followed by the ring-opening producing the corresponding organic acid that decomposes via photo–Kolb reaction (Path C in Scheme 2). The authors have excluded phenol and catechol as the main intermediates in this path since benzene swiftly oxidized to muconic acid, whose reactivity is larger than that of benzene itself.

Many other investigations have documented the H₂ formation during the photocatalytic transformation of benzene; however, they have mainly discussed the H₂ formation as a secondary product. These reports focused on other purposes, such as the mechanistic studies of the photocatalytic reaction [12,13,249,250,251] and the chemical synthesis [162,250,252], rather than the transformation of the aromatic water pollutants into fuels.

On the other hand, phenolic compounds—as we mentioned previously—are one of the most abundant aromatic pollutants in wastewater. Few research groups reported the transformation of these kinds of pollutants into fuels. In 2008, Choi et al. [253] reported the photocatalytic degradation of 4-chlorophenol and bisphenol A on the surface of bare TiO₂ (P25), F-TiO₂, Pt/TiO₂, and F-TiO₂/Pt under anoxic conditions. The authors found that F-TiO₂/Pt exhibited the highest photocatalytic activity towards the conversion of these compounds compared to the other materials. They attributed this activity to the unique synergic effect of two different surface species, i.e., fluoride and platinum, on the photo-induced charge transfer process. Such an effect inhibits the charge recombination on F-TiO₂/Pt as shown in Scheme 3. However, the mineralization of these aromatic compounds could not be achieved for all photocatalysts, since the total organic carbon content in the suspensions remained unaltered during the irradiation. Nevertheless, the authors did not discuss the possibility of molecular hydrogen formation in anoxic conditions and they ignored it in their reaction mechanism.

Two years after, the same research group presented similar results, taking into account the simultaneous production of H₂ during the photooxidation of such organic compounds [248]. Interestingly, the authors found that the % photonic efficiencies for H₂ formation during the photooxidation of the simple organic compounds, i.e., dichloroacetic acid and N-nitrosodimethylamine were found 0.122 and 0.139, respectively over Pt/TiO₂. These values were higher than those reported for F-TiO₂/Pt, i.e., 0.046 and 0.1, respectively. In contrast, using other aromatic compounds, i.e., hydroquinone, 4-chlorophenol, 4-chlorobenzoic acid, and bisphenol, higher % photonic efficiencies for H₂ were achieved over F-TiO₂/Pt (0.094, 0.116, 0.249, 0.334, respectively) than those over Pt/TiO₂ (0.052, 0.003, 0.002, 0.044, respectively). The authors reported, additionally, a complete total organic carbons (TOC) removal in the 4-chlorophenol/F-TiO₂/Pt suspension after 8-h irradiation.

It is well known that the adsorption of the photodegraded organic intermediates to the photocatalyst surface and the insufficient management of the photo-generated charge carrier inhibits the H₂ evolution reaction and/or the photocatalytic degradation of the organic pollutants in the dual-functional photocatalysis process. [38,104,254]. To this end, Kim et al. [11] studied the enhancement of the dual-functional photocatalysis process by modifying titania photocatalysts with fluoride or phosphate beside the deposition of different metals, i.e., Pt, Pd, Au, Ag, Cu, or Ni. The authors found that the dual-function photocatalysis worked only when both the anion and the metal coexisted on the surface of TiO₂, whereas TiO₂ modified with a single surface component such as F-TiO₂, P-TiO₂, or M/TiO₂ was inactive under the same experimental condition (Figure 14a). Almost similar dual-functional photocatalysis activities were reported for F-TiO₂/Pt and P-TiO₂/Pt; however, the synergistic effect greatly depended on the kind of deposited metal and the pH (Figure 14a,b). F-TiO₂/Pt was found to be active in the acidic pH region since its activity gradually decreased with increasing pH. In contrast, P-TiO₂/Pt exhibited a consistent activity over a wide range of pH, due to the strong chemical bonding of phosphates on TiO₂. Therefore, they suggested that P-TiO₂/Pt could be more appropriate for practical dual-functional applications (Figure 14c). The authors claimed that the modification of the TiO₂ surface with fluorides or phosphates with the deposition of metal acts synergistically to reduce the charge recombination and enhance the interfacial electron transfer, which enhances the photocatalytic activity.

Furthermore, the enhancement of the dual-functional photocatalytic process toward the simultaneous H₂ formation and 4-chlorophenol degradation was achieved by designing a ternary components photocatalyst [156]. Cr₂O₃/Rh/SrTiO₃ was prepared by covering the Rh nanoparticles on the surface of SrTiO₃ with a thin barrier layer of Cr₂O₃ to selectively control and maximize the dual-functional photocatalytic activity. Under the same experimental condition, the as-prepared Cr₂O₃/Rh/SrTiO₃ photocatalyst exhibited a higher activity towards H₂ production and 4-chlorophenol degradation than that of F-TiO₂/Pt and was unaffected by the pH change from the acidic medium to neutral medium (Figure 15).

According to the authors, the better photocatalytic behavior of Cr₂O₃/Rh/SrTiO₃ can be related to two features. Firstly, the Cr₂O₃ barrier layer selectively allows the conduction band electrons to be consumed by protons, hindering their transfer to O₂ or other electron acceptors. Secondly, the valence band holes are utilized to oxidize both the 4-chlorophenol and H₂O (to O₂), since the in-situ generated O₂ simultaneously and immediately consumed in the oxidation reaction to help in the mineralization of the organic pollutants, as shown in Scheme 4.

Cho et al. [255] modified the TiO₂ surface by adding the graphene oxide (GO) as a ternary component besides the modification with F and Pt to enhance the dual-functional photocatalytic activity. Pt/GO/TiO₂-F showed 1.7 and 3.8 times higher H₂ production than Pt/TiO₂-F and Pt/GO/TiO₂, respectively, during the photocatalytic degradation of 4-chlorophenol. Since the GO attracts electrons, the interfacial electron transfer was facilitated by the direct contact between GO and the TiO₂ surface, while holes are kept in TiO₂. Such an electron transfer to GO reduces the possibility of recombination of photogenerated charge carriers and extends the lifetime of charge carriers. Moreover, as the work function of Pt is higher than that of GO, i.e., 5.64 and 4.42 eV, respectively, the transfer of photogenerated electrons from GO to Pt is energetically favorable, which enhances the H₂ production. On the other hand, F ions replace the surface hydroxyl groups on the TiO₂ surface, which act as the main hole trap sites. This in turn reduces hole-trapping efficiency and hinders the chemisorption of organic substrates, thus, prevents the direct attack of the organic molecules by the trapped hole. Since the electrons are trapped by Pt, the preferred path of holes is to react with H₂O to generate unbound OH^• radicals that can diffuse out from the surface and react with the organic molecules in the medium. The authors explained that such a ternary hybrid system retards the recombination of the charge carrier and enhances both the H₂ production and 4-chlorophenol degradation, as shown in Scheme 5.

Recently, many efforts have been made to use visible-light active dual-functional photocatalysts. For example, the two-dimensional (2D) black phosphorous/2D carbon nitride (2D BP/2D C₃N₄) was synthesized and employed for efficient H₂ evolution with the simultaneous photodegradation of bisphenol A pollutant (BPA) [256]. The H₂ evolution rate and the BPA removal over 2D C₃N₄ nanosheets were found to be ~45 μmol·h⁻¹·g⁻¹ and 43%, respectively. Upon the introduction of 2D BP, both the H₂ production rate and the simultaneous BPA removal were improved. The optimum ratio of 5% 2D BP exhibits an H₂ evolution rate of 259.04 μmol·h⁻¹·g⁻¹ and BPA removal rate of 88% with an external quantum efficiency of 0.56% at 420 nm (Figure 16a). The authors attributed the high efficiency of this material to the intimate electronic interaction between 2D BP and 2D C₃N₄, besides the excellent charge mobility between the two composites.

Another effort was exerted towards the development of new materials possessing optical properties in the visible light region. Jiang et al. [257] prepared a photocatalyst consists of carbon quantum dots/CdS quantum dots/g-C₃N₄ (CDs/CdS/GCN) photocatalyst composite. The photocatalytic activity of this material under visible-light illumination was evaluated for concurrent H₂ production and the decomposition of typical wastewater pollutants like p-chlorophenol, bisphenol A, and tetracycline. The 3%CDs/10%CdS/GCN photocatalyst exhibited the best photocatalytic efficiency under the visible-light irradiation for H₂ evolution from water splitting in an aqueous solution containing organic pollutants (Figure 16b). The addition of p-chlorophenol decreased the photocatalytic H₂ evolution rate compared with the pure water system, due to the consumption of some photogenerated electrons in the degradation of p-chlorophenol. Although the photocatalytic degradation rate of p-chlorophenol was higher than those of bisphenol A and tetracycline, the H₂ evolution rate increased with the addition of bisphenol A or tetracycline. The authors have explained such a result by the consumption of all photogenerated electrons to split water for H₂ production.

On the other hand, some reports have shown the inability to use phenol as a sacrificial reagent in the dual functional photocatalytic processes. Mogyorósi et al. [258] investigated the photocatalytic H₂ production and the decomposition of various organics using 1% Pt-, Au-, and Ag-deposited on the surface of Degussa P25 photocatalysts. The photocatalytic decomposition of oxalic acid and formic acid was increased upon the deposition of noble metals compared to that of the bare photocatalyst. However, in phenol containing system, the authors reported a decrease in the decomposition activity, indicating that the noble metals block the active sites on the surface of the photocatalyst. On the other hand, they practically reported no H₂ production over the bare and the modified P25 in the presence of phenol as a sacrificial reagent. While a very high quantum yield for H₂ production over Pt-TiO₂ photocatalyst was reported in the presence of oxalic and formic acids. They have concluded that O₂ was a requirement for the photooxidation of phenol in presence of any photocatalysts since no decomposition was detected in its absence. Hence, the inhibition of phenol photooxidation in anoxic conditions negatively affects the ability of H₂ production.

5.2. Dyes and Polyaromatic-Based Pollutants

A wide variety of photocatalysts was designed to achieve the goal of the dual-functional photocatalysis technology; simultaneous H₂ production and wastewater purification by the degradation of the persistent dyes [38,39,189,259,260,261,262]. The production of hydrogen with a simultaneous degradation of azo dye solution (commercial name Acid Orange 7; AO7) using the well-known photocatalyst Pt/TiO₂ suspensions was examined by Patsoura [261] under UV-vis light. The authors have investigated the effect of the dye concentration, the pH, and the temperature on the H₂ production rate. Besides, the effect of the Pt loading ratio on the H₂ formation rate was thermodynamically investigated through the dynamic of the charge carrier during the reaction. In the absence of the Azo-dye and after the deposition of Pt (0.5 wt.%) on the TiO₂ surface, the H₂ production rate increased to a maximum during the irradiation before dropping to a very low steady-state rate value comparable to those obtained over bare TiO₂. Although, it is well known that bare TiO₂ is inactive for the H₂ production due to the driving force for this reaction is small, and the presence of a large overpotential for the H₂ evolution [263], the authors attributed such activity to the presence of (i) metal or organic impurities in the semiconductor; (ii) partially reduced titania species; (iii) small size semiconductor particles that exhibit a higher efficiency in photocatalytic reactions. The authors also reported an improvement in the H₂ formation during the photo-induced water splitting reaction over Pt/TiO₂ by increasing the pH and the temperature. Interestingly, the presence of a small quantity of Azo dye in the reaction medium significantly enhanced the H₂ formation rate, which depends on dye concentration, solution pH, and to a lesser extent to the solution temperature. They found that using a higher dye concentration resulted in increasing H₂ formation over a longer reaction period. However, afterward, the formation rate was decreased to a steady-state value comparable to that obtained in the absence of the azo dye. The authors have attributed this decrease to the complete mineralization of the AO7 by-products in the reaction solution, due to their oxidation by consuming the photogenerated oxygen from the surface of the photocatalyst.

In the same study, another two azo dyes, namely Basic Blue 41 and Basic Red 46 have been tested at a neutral pH solution to ascertain that the use of a dye is generally beneficial for the rate of H₂ production. Similar behavior to the addition of AO7 was observed for the other azo dyes. The H₂ formation rate increased during the first few hours of irradiation and then progressively dropped to steady-state values similar to those obtained for pure water.

The authors have also highlighted that the adsorption of the reaction intermediates on Pt cannot be effectively removed under the experiment conditions, which leads to retarding the H₂ evolution. This behavior had been observed in many similar photocatalytic systems dealing with the TiO₂ and aromatic compounds [104,167,247,250].

On the other hand, increasing pH from 4 to 10 resulted in a significant increase in the maximum formation rate from 0.28 to 0.67 μmol/min, which was related to the enhanced kinetics of dye degradation with increasing the solution pH. According to these authors, this enhancement indicating that the rate of H₂ production is limited by the rate of consumption of photogenerated O₂. Therefore, they conclude that the azo dye acts as a sacrificial agent that rapidly remove the photogenerated holes and consume the photogenerated oxygen. This suppresses electron-hole and O₂-H₂ recombination, enhancing the H₂ production until complete degradation of the dye to CO₂ and inorganic ions.

The modification of the TiO₂ in a way that increases the adsorption of the organic molecules on its surface is considered one of the methods that enhance the photocatalytic activity since the direct hole transfer to the organic molecules is dominant [264,265]. Bifunctional TiO₂ photocatalysts have been developed by Kim et al. [266] through the modification of the surface of TiO₂ with two different components, platinum, and Nafion (Pt/TiO₂/Nf). The simultaneous H₂ production and rhodamine B (RhB) degradation was successfully achieved using Pt/TiO₂/Nf under visible light (𝜆 > 420 nm). Pt/TiO₂/Nf exhibited high activity for H₂ production in the presence of RhB as a photosensitizer and organic dye pollutant, besides EDTA as an electron donor. However, the modification with only one component, i.e., Pt or Nf, resulted in a negligible activity for H₂ production under the same experimental conditions. According to the authors, the negative charge of the Nafion layer improves the adsorption of cationic RhB and pulls protons to the surface of TiO₂ through electrostatic attraction, enhancing the RhB photooxidation. Simultaneously, these protons are reduced to H₂ on the deposited Pt that acts as an electron sink and a temporary electron reservoir for the reduction half-reaction. The authors found that RhB was not degraded in the absence of EDTA, which is involved in the reaction mechanism by converting the RhB to 𝑁-deethylation. In this dual-functional photocatalytic system, a 20 µM (0.6 µmol) of RhB approximately produced 70 µmol of H₂, while the RhB and its intermediates were completely removed over 12 h period.

Polycyclic aromatic hydrocarbons (PAHs) are a kind of semi-volatile persistent aromatic pollutants [267]. These compounds are frequently detected in different types of wastewater [58,267]. As for many other pollutants, advanced oxidative processes based on photocatalysis have often been reported for the removal of PAHs [268,269]. Although several studies have explored their photocatalytic degradation in anoxic conditions [268,270,271]; however, very limited reports on the remediation of PAHs with simultaneous H₂ production had been documented in the literature [104,167]. On the other hand, several reports on the simplest aromatic compound benzene have proved its ability to act as a sacrificial electron donor (hole scavenger) to photo-catalyze molecular hydrogen [247,250,251]. Bahnemann’s research group has considered this shortage in the literature and spotted the light on employing these compounds as sacrificial electron donors in the dual-functional photocatalysis system [104,167,272].

The hydrogen production with the simultaneous degradation of the simplest PAH compounds naphthalene based on Pt/TiO₂ has been investigated by the Bahnemann research group [167]. In this study, two different commercial TiO₂ photocatalysts, Aeroxide P25 (ATiO₂) and Sachtleben Hombikat UV100 (HTiO₂) were loaded with different fractional ratios of Pt nanoparticles using the photodeposition method. The aim was to evaluate the role of the loaded Pt on hydrogen production and the simultaneous degradation of naphthalene. The 0.5 wt.% Pt was found to be the optimum loading ratio on the surface of HTiO₂, which increased the conversion of naphthalene from 71% for bare HTiO₂ to 82% and produces 6 µmol of H₂ (Figure 17). However, the authors found that using a higher Pt content than the optimal platinization ratio inhibited both processes, the H₂ formation, and naphthalene photooxidation. On the other hand, they claimed that loading ATiO₂ with the Pt nanoparticles regardless of the platinization ratio decreased naphthalene conversion, while no dependency between the Pt ratio and the H₂ formation rate was found since all OF the platinized ATiO₂ materials showed a similar H₂ formation of around 3 µmol.

Based on the EPR technique, the authors concluded that the Pt NPs on ATiO₂ acted as recombination centers for the photogenerated charge carrier. They have additionally related the decreases of H₂ formation rate and naphthalene conversion during the photocatalytic process to the deactivation of the photocatalyst due to adsorption of the formed intermediates on the surface of the photocatalyst. Interestingly, the authors demonstrated that the reforming of PAHs over the Pt-HTiO₂ exhibits higher photonic efficiencies than that of their corresponding hydroxylated compounds, such as 1 and 2-naphthols.

In another study, the effect of the co-catalyst loading methods on the physicochemical properties of the dual-functional photocatalyst was studied by the same research group [272]. Anatase TiO₂ (Sachtleben Hombikat UV100) was loaded with Pt nanoparticles using two alternative methods: photodeposition by reduction of PtCl₆²⁻ (Pt_PD-TiO₂) and physical mixing of TiO₂ with Pt nanoparticles synthesized by laser ablation (Pt_LA-TiO₂). Both as-prepared materials were fully characterized, and their photocatalytic activities were evaluated for the photoreforming of naphthalene and methanol. Over both photocatalysts, the authors reported a huge difference in H₂ formation between the two-electron donors, which can be related to the different nature of the organic compounds. Methanol reacts swiftly with the photogenerated holes, while the reaction of naphthalene involved multi-complicated steps. On the other hand, Pt_PD-TiO₂ exhibited better photocatalytic activity toward naphthalene oxidation and H₂ formation compared to Pt_LA-TiO₂. Based on the transient absorption spectroscopy and the electron paramagnetic spectroscopy techniques, the higher activity of Pt_PD-TiO₂ was related to the better charge carrier transfer between the TiO₂ and the loaded Pt nanoparticles. The authors explained these results by the better dispersion of Pt nanoparticles and their strong interaction with the surface of TiO₂.

The mechanism of the dual-functional photocatalysis process for molecular hydrogen formation concurrent with naphthalene degradation over Pt-TiO₂ (Hombikat UV100) has been investigated [104]. The authors reported photonic efficiencies of 0.33% and 0.970% for naphthalene conversion and H₂ formation, respectively, under simulated sunlight. After 4 h irradiation, the authors were able to determine the formed organic by-products in the system by the mean of gas chromatography - mass spectrometry, high-performance liquid chromatography, and high pressure ion chromatography techniques. Moreover, through the spin-trapping experiments, they proved that only the photogenerated holes play the main role in the photooxidation of naphthalene, while, the isotopic labeling analyses showed that the evolved H₂ originated mainly from water. According to these results, the authors suggested the following mechanism for hydroxylation of naphthalene (Equations (9)–(15)), while the total mineralization mechanism was shown in Scheme 6.

$${\mathrm{TiO}}_2 \to \mathrm{n}\left({\mathrm{e}}^{-}+{ \mathrm{h}}^{+}\right){\mathrm{TiO}}_2$$

(9)

(10)

(11)

(12)

(13)

$$\mathrm{Pt}-{\mathrm{TiO}}_2\left[{\mathrm{e}}^{-}\right]+{\mathrm{H}}_{\left(\mathrm{ads}\right)}^{+} \to \mathrm{Pt}-{\mathrm{TiO}}_2+{ \mathrm{H}}_{\left(\mathrm{ads}\right)}^{•}$$

(14)

$$2{\mathrm{H}}_{\left(\mathrm{ads}\right)}^{•} \to { \mathrm{H}}_2(\uparrow )$$

(15)

6. Conclusions

The increase in the quantity and quality of pollutants associated with industrial progress and population growth makes it necessary to match this increase with efficient and sustainable ways to treat it; hence, the urge to develop new materials or to modify and/or enhance the performance of some existing materials. There is no doubt that abundance and low cost are advantages that every semiconductor must meet for their application in large-scale photocatalytic systems. These two properties turned TiO₂ into an attractive material in this field—attractive enough to devote large scientific efforts to overcome its main limitations: fast charge carrier recombination rates and a relatively large bandgap (3.2 eV), so that only UV radiation can activate it. However, the adopted strategies to improve TiO₂ performance and make it more appealing for large-scale applications—some of these strategies are discussed in this review, and are, to a large extent working, and seem promising. However, miniaturizing or synthesizing in the nanoscale is not the only way to achieve high efficiency. The fast recombination of the photogenerated charges could be significantly reduced by the loading of co-catalyst, which is normally noble metal nanoparticles. The high cost of the noble metals and their limited availability make them a not idealistic choice. One more time, cost-effectiveness comes into play but this time as a limiting factor. The search for co-catalysts that demonstrate high efficiency combined with cost efficiency is a challenging issue in photocatalysis. The wide scale of chemical and physical properties of both pollutants and semiconductors could anticipate the use of oxide-oxide or metal-oxide hetero nanostructures to create new properties that achieve higher performance and enhanced ability to remove, reform, or degrade pollutants. Not to mention that heterostructures could demonstrate the same function as a catalyst and co-catalyst without the resolve to high-cost noble metals, which could be a working strategy with enormous numbers of materials. Finally, despite a large amount of photocatalytic reforming studies, there is a huge deficiency in the investigation of the photocatalytic reforming of aromatic-based pollutants, especially the PAHs. Such pollutants have shown the ability for photocatalytic oxidation in the oxygen atmosphere; however, few reports have been published that deal with the H₂ production based on their photoreforming. Hence, this research line can be a rich area for further future investigation.