g-C₃N₄-Based Direct Z-Scheme Photocatalysts for Environmental Applications

Abstract

Photocatalysis represents a promising technology that might alleviate the current environmental crisis. One of the most representative photocatalysts is graphitic carbon nitride (g-C₃N₄) due to its stability, cost-effectiveness, facile synthesis procedure, and absorption properties in visible light. Nevertheless, pristine g-C₃N₄ still exhibits low photoactivity due to the rapid recombination of photo-induced electron-hole (e⁻-h⁺) pairs. To solve this drawback, Z-scheme photocatalysts based on g-C₃N₄ are superior alternatives since these systems present the same band configuration but follow a different charge carrier recombination mechanism. To contextualize the topic, the main drawbacks of using g-C₃N₄ as a photocatalyst in environmental applications are mentioned in this review. Then, the basic concepts of the Z-scheme and the synthesis and characterization of the Z-scheme based on g-C₃N₄ are addressed to obtain novel systems with suitable photocatalytic activity in environmental applications (pollutant abatement, H₂ production, and CO₂ reduction). Focusing on the applications of the Z-scheme based on g-C₃N₄, the most representative examples of these systems are referred to, analyzed, and commented on in the main text. To conclude this review, an outlook of the future challenges and prospects of g-C₃N₄-based Z-scheme photocatalysts is addressed.

1. Introduction

Environmental protection and energy generation are two of the great challenges that mankind is presently facing, since they have a large negative impact on our society, affecting, among others, air pollution, climate change, water pollution, thermal pollution, and solid waste disposal [1,2,3]. Such problems have been caused by rapid industrialization, uncontrolled environmental pollution, and current energy scenario based on fossil fuels [4]. Aiming to solve these issues, current policymakers are passing regulations to reach full environmental sustainability and avoid a prolonged energy crisis and environmental deterioration. One interesting approach to solve this issue is to replace the use of carbon-based energy with solar energy [5,6]. Indeed, solar energy is a green, inexhaustible, and clean resource [6,7]. Another essential approach to achieve the sustainability of our planet is the degradation of pollutants to less toxic products and the reduction of CO₂ to obtain value-added products, which can be achieved under the irradiation of solar light [8,9].

Photocatalysis represents a promising technology that might alleviate the current environmental crisis [10]. Semiconductor photocatalysts may generate H₂ by water splitting, degrade organic pollutants, reduce CO₂ into high value-added products, etc. using solar energy. Moreover, heterogeneous photocatalysis can be performed under very mild conditions (room temperature and atmospheric pressure) [11,12]. Although photocatalytic technology presents relevant characteristics, such as being clean, safe, and renewable, it is still far from commercial implementation, especially in solar-to-fuel conversion [13,14]. Titanium dioxide (TiO₂) has been widely studied in the field of solar energy conversion after the pioneering work reported by Fujishima and Honda in 1972 based on photoelectrochemical water splitting on a TiO₂ electrode [15]. Furthermore, this semiconductor has received great interest from the scientific community for its use in pollutant abatement due to its advantages, e.g., low price, high durability, and abundance [16]. To understand the relevance of photocatalysis, it is important to briefly describe the reaction mechanism [17]. Initially, a photon with energy equal to or greater than the energy band gap (Eg) of the semiconductor is absorbed by the semiconductor. Then, a photoexcited electron (e⁻) is promoted from the valence band (VB) to the conduction band (CB). This effect leads to the formation of a hole (h⁺) in the VB and, consequentially, to the generation of an electron-hole pair (e⁻-h⁺). The produced e⁻ and h⁺ can recombine on the surface or within the photocatalyst rapidly and thus decrease the photocatalytic activity or they can also migrate to the surface of the semiconductor and initiate the redox reaction(s) with the species adsorbed on the surface of the photocatalyst. The generated holes can induce oxidation processes, which occurs in organic molecules abatement, and the electrons can promote reduction processes such as hydrogen evolution in water splitting and CO₂ reduction. To develop efficient photocatalysts, light absorption is an essential step in the manufacture of the material [18]. Pristine TiO₂ is active only under UV light with a wavelength below 387 nm due to its wide bandgap (3.2 eV). This fact is the main drawback of TiO₂ since solar energy is mainly concentrated (between 95 and 97%) in the IR and visible light regions [16]. Consequently, the exploration of visible-light-responsive photocatalysts with high efficiency is a highly interesting topic for the scientific community [19,20].

Graphitic carbon nitride (g-C₃N₄) is a very promising material due to its potential application in photocatalytic pollutant degradation, photocatalytic H₂ production, and CO₂ reduction. Its bandgap of 2.7 eV makes it a good candidate as a visible-light-responsive photocatalyst. In addition, g-C₃N₄ presents a unique electronic band structure, low cost, and easy preparation, allowing it to be a possible alternative to TiO₂ in solar energy applications [21,22]. However, pristine g-C₃N₄ still exhibits low photoactivity in several applications such as water splitting and CO₂ reduction due to the rapid recombination of photo-induced e⁻-h⁺ pairs [23]. In recent years, several strategies have been developed by the scientific community to improve the photocatalytic efficiency of g-C₃N₄, i.e., morphological control, doping elements, deposition of noble metals, and construction of heterojunctions [24,25].

The latest advancements on g-C₃N₄-related photocatalytic systems are based on g-C₃N₄-based heterostructures due to the effective separation of photogenerated e⁻-h⁺ pairs [26,27]. Traditional heterojunctions are categorized into type-I, type-II, and type-III, as shown in Figure 1. A different pathway of the traditional type-II is the Z-scheme family (Z-schemes and S-schemes) [28,29], as shown in Figure 1 [30,31]. Since the present review is focused on direct Z-scheme heterojunctions, only type II heterojunctions will be described, as shown in Figure 1b. Type II heterojunction formation using g-C₃N₄ is a noteworthy method used to increase the photocatalytic efficiency by suppressing the recombination of charge carriers [24]. However, type II heterostructures limit the oxidation and reduction efficiency of the heterojunction material from the parent materials [27]. To solve this drawback, the synthesis and study of Z-scheme photocatalysts may be a good compromise since Z-scheme-based photocatalysts present the same band configuration but follow a different charge carrier recombination mechanism. Thus, in the Z-scheme, the strongest reduction and oxidation potentials of each semiconductor are preserved similarly to what happens in a natural process such as photosynthesis [32,33].

There are two types of Z-scheme semiconductor heterojunctions [34,35]. The first one is the indirect Z-scheme (redox-mediated and all solid state), where the transport route of photogenerated charge carriers is not achieved directly but by the addition of an electron mediator. The second type is the direct Z-scheme, which consists of two semiconductors in close contact, eliminating the demand for an electron mediator. These novel direct Z-scheme photocatalysts have attracted the attention of the scientific community, increasing the use of these photocatalysts. Indeed, the absence of mediators eliminates the backward reaction and light-shielding effects [36,37]. The layered structure of g-C₃N₄ is a suitable building block for Z-scheme photocatalyst construction due to the possible high surface area, provided that a layered material is obtained, and its ability to perform photocatalytic reduction reactions (water splitting and CO₂ reduction), making it possible to couple g-C₃N₄ with a wide range of oxidation-type photocatalysts to fabricate Z-scheme photocatalysts [21,38,39]. This fact has boosted the study of Z-scheme photocatalysts based on g-C₃N₄ for their use in environmental applications, as evidenced by the increase in the number of scientific publications on this topic during the last decade (Figure 2). Therefore, it becomes timely to carry out a bibliographic review in which the most relevant works of direct Z-scheme photocatalysts based on g-C₃N₄ and their environmental applications are compiled and summarized.

In this review, we will briefly describe pristine g-C₃N₄, although it still exhibits low photoactivity in several applications such as water splitting and CO₂ reduction due to the rapid recombination of photo-induced e⁻-h⁺ pair. Then, a new alternative to improve the photocatalytic activity of g-C₃N₄ based on the design of novel Z-schemes will be widely described. Although Z-schemes are applied in relevant photocatalytic reactions, such as environmental applications [34,40], energetic applications [40], transformation of organic compounds to added-value products [41,42,43], and photocatalytic biorefineries [44], among others, this review will cover most relevant works on g-C₃N₄-based Z-scheme photocatalysts and their use in environmental applications such as pollutant abatement and CO₂ reduction due to the high impact in our society, as was mentioned previously. In addition, energy aspects, e.g., water splitting, are also briefly reviewed.

2. Pristine g-C₃N₄

Graphitic carbon nitride is a stable polymeric and layered semiconductor formed by the polymerization of abundant nitrogen-containing precursors, such as melamine or urea [45,46]. The first report about g-C₃N₄ was written in the 1830s by Berzelius and Liebig [47]. Since then, many research studies have been carried out to investigate the structure of this material and to develop different synthetic routes [45]. The structure of g-C₃N₄ is based on triazine or heptazine units polymerized in a layered structure (Figure 3a) [48,49]. In the last decades, g-C₃N₄ has been presented as a promising material for its use in several applications such as sensors, energy storage, and photovoltaic cells, among others, since it presents facile, low-cost, and environmentally friendly preparation methods with promising stability and good physicochemical properties [45,50]. Focusing on the application of this review, g-C₃N₄ is a good candidate as a photocatalyst because it is a visible-light-active photocatalyst with a bandgap of ~2.7 eV (~460 nm) [51,52]. In contrast to other semiconductors, g-C₃N₄ does not contain metal ions, which may be leached into the ecosystem and cause unfavorable impacts therein. Instead, it is composed of Earth-abundant carbon and nitrogen elements and can be easily synthesized using nitrogen-rich organic precursors by various methods. The obtained g-C₃N₄ is endowed with desirable electronic structures and unique morphologies, and high thermal stability up to 600 °C in air [46].

Nowadays, there are several innovative and relevant synthesis methods of g-C₃N₄, that may be classified as solvothermal, chemical vapor deposition (CVD), plasma sputtering reaction deposition, and polycondensation [21,53]. Among them, thermal polycondensation is the most attractive method due to its simplicity and low cost. Nitrogen-rich small molecules can polymerize into g-C₃N₄ following a calcination process at 450–650 °C (Figure 3b) [54]. The choice of the precursor exhibits effects on the electronic band structures and the textural properties of the obtained pristine bulk g-C₃N₄ [55]. In this sense, Wang et al. demonstrated in 2009 that g-C₃N₄ was a promising visible-light photocatalyst for H₂ evolution. The as-prepared g-C₃N₄ achieved steady H₂ production from water containing triethanolamine as a sacrificial electron donor on light illumination (λ > 420 nm) even in the absence of noble metal co-catalysts such as Pt (4 μmol/h), opening a new approach for the search of materials for H₂ production [56]. After this pioneering work, the scientific community has carried out a lot of work to synthesize g-C₃N₄ by thermal treatment of nitrogen-rich precursors such as urea, thiourea, melamine, cyanamide, and dicyandiamide, among others, to obtain g-C₃N₄ in different morphologies such as nanosheets or nanotubes [57,58]. However, pristine g-C₃N₄ presents several drawbacks for its use in photocatalysis, including a low specific surface area, insufficient visible light utilization, and, the most relevant, rapid recombination of photogenerated charge carriers [23].

3. Direct Z-Scheme Photocatalysts Based on g-C₃N₄

3.1. Direct Z-Scheme Photocatalysts

To compensate for the drawbacks of pristine g-C₃N₄ for its use in photocatalysis, as described in Section 2, a novel alternative is the design of heterojunctions, defined as the interface between two different semiconductors, which can result in suitable band alignments [59,60]. The design of type II heterojunctions using g-C₃N₄ is an interesting method to increase the photocatalytic efficiency by suppressing the recombination possibilities of the generated e⁻-h⁺ pairs. In type-II heterojunction photocatalysts, the CB and the VB levels of semiconductor 1 are higher than the corresponding levels of semiconductor 2. Electron and hole pairs are separately generated in 1 and 2 subjected to light irradiation. The photogenerated electrons will transfer to semiconductor 2 while the photogenerated holes will migrate to semiconductor 1. Consequently, type II heterostructures diminish the oxidation and reduction efficiency of the isolated materials, which is the main drawback of this heterostructure (Figure 4c) [21].

Inspired by natural photosynthesis (Figure 4a), the artificial Z-scheme heterojunction design is a great alternative for the development and manufacture of novel photocatalysts, with outstanding results in numerous applications [61,62]. The Z-scheme photocatalytic system concept was proposed by Bard in 1979 for the first time [63]. Similar to what happens in photosynthesis, under light excitation, the electrons generated in a semiconductor with low reduction potential are recombined with the holes with low oxidation potential, i.e., those with the highest potential. This phenomenon generates electrons and holes isolated in the Z-scheme system with maximum redox abilities, making it the greatest advantage of these materials (Figure 4b) [34]. The traditional Z-scheme introduced by Bard [63] needs a shuttle redox mediator (electron acceptor/donor pair) to form a liquid-phase Z-scheme. Thus, Z-scheme photocatalysts suffer from the limitations of redox mediator reversibility and their specific applications, e.g., CO₂ reduction, can only be applied in the liquid phase [34]. The second generation of Z-scheme (all solid-state Z-scheme photocatalysts) was discovered in 2006 by Tada et al. [64]. These are composed of two different semiconductors with a solid-phase electron mediator as a noble metal nanoparticle (NP) or carbon material (graphene and carbon nanotubes) [65]. To solve the inconvenience of using electron mediation, Yu et al. [66] constructed a direct Z-scheme photocatalyst by combining g-C₃N₄ and TiO₂ in 2013. The interfacial contact between the semiconductors facilitates the direct electron transfer without the help of an electron mediator. This novel direct Z-scheme system presented the advantage of significantly reducing the construction cost [37].

In fact, g-C₃N₄ has received significant attention due to its potential use as a reductant semiconductor in direct Z-scheme photocatalysts given its superior chemical and physical features [26,33]. In addition, the great advantage of its use as part of Z-scheme photocatalysts is that the VB and CB positions are located at approximately +1.6 and −1.1 eV, respectively [28]. This redox potential converts g-C₃N₄ into an interesting reduction semiconductor for its use in several applications, such as H₂ production and CO₂ reduction (Figure 5) [67]. Nevertheless, the formation of the g-C₃N₄/semiconductor interface is a challenge for a Z-scheme heterostructure [33,68]. Therefore, it is necessary to design and modify novel routes of synthesis to tackle the drawbacks of these systems.

3.2. Synthesis and Characterization of Z-Scheme Photocatalysts Based on g-C₃N₄

Generally, the synthesis of Z-scheme photocatalysts is crucial to obtain systems with high efficiency in photocatalysis. In the literature, there are several methods to synthesize catalysts or composites such as deposition, solid-state synthesis, and hydrothermal synthesis, among others [22,37,69]. In the synthesis of Z-scheme photocatalysts, two semiconductors are combined to optimize the oxidation and reduction potential by the recombination of e⁻-h⁺ pairs, which makes intimate contact between both semiconductors crucial [63]. To increase the activity of photocatalysts based on g-C₃N₄, there are several methods [58,64], such as varying the texture of g-C₃N₄ by means of template synthesis or deposition of metal cocatalysts. However, this section of the manuscript will focus on the synthesis of the direct Z-scheme without any mediator.

The most relevant synthetic methodologies used by the scientific community to design direct Z-scheme systems based on g-C₃N₄ are (Figure 6):

Solid-state synthesis. This methodology is broadly used in the synthesis of pristine g-C₃N₄ materials [47,70] and it is most employed to synthesize Z-scheme photocatalysts based on g-C₃N₄ [71]. This methodology is based on the calcination of one or a mixture of precursors in air or an inert gas atmosphere at high temperatures [72]. There are some crucial experimental parameters in the synthesis of materials using solid-state methodologies such as the heating rate, calcination temperature, and calcination time to control the crystallinity, morphology, surface properties, and phase structure of the composite [73,74]. This methodology is very interesting for the design of Z-scheme materials, as reported by W. Yu et al. [75], who synthesized g-C₃N₄ in the presence of pre-synthesized WO₃ to obtain a direct Z-scheme g-C₃N₄/WO₃ photocatalyst. Another noticeable work that uses this methodology was reported by L. Lu et al. [74]. In this work, the authors studied the effects of the calcination temperature on the photocatalytic activity of direct Z-scheme TiO₂/g-C₃N₄. Although this methodology is widely used in the synthesis of the g-C₃N₄-based direct Z-scheme, the high temperatures used and the low control of the composition are the main drawbacks.

Deposition precipitation method. This methodology is commonly used for the synthesis of photocatalysts when one of the precursors is cationic and the other is anionic. A uniform precipitated composite is formed [76,77]. The deposition precipitation technique is based on the formation of a precipitate on the surface of another component by slow addition or in situ growth of a substance, following the addition of a precipitating agent at a low temperature [78,79]. Although this methodology is not widely used in direct Z-schemes, it is an interesting methodology due to the easy fabrication of these systems [80].

Impregnation. Impregnation is another popular method for the synthesis of catalysts [81,82] and it has been used for the fabrication of Z-scheme photocatalysts [83,84]. In this approach, a solid precursor or material is in contact with a solution containing the precursor to be deposited on the solid surface. There are two methods of impregnation: (1) the wet impregnation method, in which the solid precursor is introduced with an excess volume of the second precursor solution, and (2) incipient wetness impregnation, in which the volume of the second precursor solution used is equal or less than the pore volume of the solid [82]. Due to the advantages of this methodology, Feng et al. [85] reported a composite synthesized using a simple impregnation-heating method in which MoO₃ nanoparticles were in situ supported on g-C₃N₄ sheets. This Z-scheme photocatalyst had photocatalytic activity in CO₂ reduction to fuels under simulated sunlight radiation. Zhou et al. [86] reported a simple impregnation method to synthesize Z-scheme g-C₃N₄ decorated with TiO₂ nanotubes with improved visible-light photocatalytic activity in pollutant abatement. Another example is the work of Jin et al. [84], where the authors reported the use of a one-step impregnation method to prepare direct Z-scheme LaCoO₃/g-C₃N₄ photocatalysts.

Hydrothermal synthesis. In the 21st century, hydrothermal technology is one of the preferred methods for the synthesis of materials in various interdisciplinary fields such as advanced materials technology, nanotechnology, biotechnology, etc. due to the ease of processing particles with high purity, high crystallinity, controlled stoichiometry, and controlled chemical and physical characteristics, and to the environmental friendliness [87,88]. Hydrothermal processing is defined as a heterogeneous reaction performed under high temperatures and pressure in the presence of an aqueous solvent to dissolve and recrystallize substances that are relatively insoluble under ambient conditions [89]. It became one of the mostly used synthetic methodologies for the synthesis of g-C₃N₄-based Z-scheme photocatalysts. Jo et al. reported the synthesis of Z-scheme g-C₃N₄/TiO₂ photocatalysts for isoniazid degradation. Moreover, they studied the effect of the TiO₂ morphology in the synthesis of the Z-scheme photocatalysts and their catalytic activity [90]. Di et al. [91] reported the synthesis of a direct Z-scheme based on g-C₃N₄, synthesizing a C₃N₄/SnS₂ photocatalyst with an in situ hydrothermal method at 140 °C, and the photocatalysts had a superior visible-light CO₂ reduction performance. Recently, Lu et al. [92] reported a 2D/2D g-C₃N₄/BiVO₄ Z-scheme heterojunction using the hydrothermal methodology with remarkable photocatalytic activity enhancement of CO₂ conversion promoted by efficient interfacial charge transfer. In this sense, Wu et al. [93] synthesized 2D g-C₃N₄-supported nanoflower-like NaBiO₃ using a facile hydrothermal synthesis.

Photo-deposition methodology. Photo-deposition is a common technique for loading a cocatalyst (such as Pt NPs) onto a photocatalyst via photoreduction [94]. Photo-deposition is the phenomenon through which a cocatalyst is deposited on the surface of a semiconductor, upon illumination of a solution containing the cocatalyst precursor and the support [95]. In the last years, the scientific community has developed a great interest in the photo-deposition method to obtain Z-schemes [17,95]. One representative example is the work reported by Jiang et al. [96], where two routes for constructing the Fe₂O₃/g-C₃N₄ direct Z-scheme through photo-deposition were demonstrated.

The characterization of direct Z-scheme photocatalysts is crucial to identify Z-scheme heterojunctions because type II heterojunctions and Z-scheme photocatalysts have similar structures [37]. The main difference between both heterojunctions is the charge carrier mechanism, as described in Section 3.1. In the last years, researchers have studied and developed several experimental and theoretical simulation methods to characterize these novel heterostructures (Figure 6) [33,37]. The most interesting and widely used methodology to characterize and understand the mechanism of Z-scheme systems is the radical species trapping methodology [97,98]. Indeed, this methodology can be applicable because Z-scheme semiconductor 1 with a high oxidizing capacity can produce ^•OH while semiconductor 2 with a sufficient reduction potential is capable of generating O₂^•− species. However, type-II heterojunction photocatalysts with a low redox (reduction or oxidation) ability can only generate one type of radicals (either O₂^•− or ^•OH). To elucidate this effect, the radical scavenging methodology is applied. In radical scavenging experiments, a chemical agent is introduced in the photocatalytic medium system to quench a radical, consequently decreasing the activity of the studied reaction [99,100]. Common scavengers used in the literature are tert-butyl alcohol (TBA) and isopropanol (IPA) for ^•OH, and N₂ gas and p-benzoquinone (BQ) for O₂^•− while ammonium oxalate (AO), triethanolamine (TEA), and disodium EDTA are used for holes (h⁺) [100]. Other characterization techniques used to study Z-scheme systems are photoluminescence (PL) spectroscopy and electron paramagnetic resonance (EPR) spectroscopy [101,102,103]. During the last years, there has been an increase in the application of other spectroscopic characterization techniques such as ultraviolet photoelectron spectroscopy (UPS), transient absorption spectroscopy (TAS), surface photovoltage spectroscopy (SPS), and in situ irradiated X-ray photoelectron spectroscopy (ISI-XPS) to verify the Z-scheme charge transfer mechanism [26,104]. Another methodology used to verify Z-schemes in heterojunctions is photocatalytic reduction testing because the photogenerated electrons with adequate reduction potential in a semiconductor can be used to produce some selective products that a semiconductor with a lower reduction potential is unable to generate [33,105]. To experimentally verify the proposed system, this methodology usually needs the aid of computational calculations based on density functional theory [106,107,108]. In the last years, this methodology has been used to calculate the Fermi level and to interpret the charge transfer mechanism.

Finally, it is relevant to indicate that there are several methodologies used to synthesize direct Z-scheme photocatalysts, as described in this section. However, it is essential to choose and develop easy and sustainable methodologies focusing on the intimate contact between both semiconductors to obtain direct Z-scheme photocatalysts. It is also necessary to use and develop powerful characterization tools to investigate the charge-transfer mechanism to elucidate whether the heterojunctions synthesized and studied are Z-scheme photocatalyst or another heterostructure, such as type II.

4. Environmental Applications of Direct Z-Schemes Based on g-C₃N₄

As mentioned in previous sections, the use of semiconductors capable of utilizing sunlight has received great interest in environmental remediation and in contributing to the solution of the energy crisis [10,109]. g-C₃N₄ is a semiconductor with optimal properties (absorption in visible light) for its use in this field, although it has several drawbacks [54,110], as shown in Section 2. Therefore, an alternative for g-C₃N₄ to be a key material in the current scenario is the use of the direct Z-scheme based on g-C₃N₄ [33,67]. Hence, g-C₃N₄-based direct Z-schemes have been widely studied by scientists for their application in photocatalytic pollutant remediation [111,112,113], photocatalytic H₂ production [114,115], and photocatalytic CO₂ reduction [85,116,117]. In this section, the most representative direct Z-schemes based on g-C₃N₄ for their use in environmental applications focusing on pollutants remediation, H₂ production, and CO₂ photoreduction will be presented. Additionally, in the main text, the most relevant aspects of the developed direct Z-schemes based on g-C₃N₄ for these three relevant environmental applications are described.

4.1. Pollutant Remediation

During the last decades, photocatalytic pollutant remediation has been intensively studied, leading to the synthesis of a plethora of photocatalysts active in the degradation of different pollutants. Many reports were also focused on the study of the so-called advanced oxidation processes (AOPs), which are fundamental for pollutant remediation. Indeed, AOPs include by definition the formation of highly reactive oxygen species (ROS), e.g., O₂^•−, OH^•, and more recently SO₄^2•− radicals, which can generate cascade processes during pollutant degradation. Indeed, these radicals have the capability of rapidly reacting with most common pollutants. This will generate radicals on the organic skeleton of the molecule and bring about defragmentation, which might lead to a complete mineralization of the contaminant in the form of CO₂ [118]. The use of heterogeneous photocatalysts can positively influence the formation of these species. As shown in Equation (1), the promotion of electrons from the VB to the CB origins holes, which might react with water or SO₂O₈²⁻ for the generation of OH^• (see Equations (2) and (3)) or SO₄^2•− radicals (Equation (4)), respectively, which participate in the subsequent radical reaction pathway for pollutant degradation [119,120]. On the other hand, the electrons accumulated at the CB might react with O₂ to generate O₂^•− radicals (Equation (5)), fundamental for the following cascade reactions [121]:

$$SC+h\nu \to e_{CB}^{-}+h_{VB}^{+}$$

(1)

$$h_{VB}^{+}+O{H}_{surface}^{-}\to O{H}^{•}$$

(2)

E⁰ = 2.8 V (V vs. NHE, pH 7, 25 °C, 1 atm)

$$h_{VB}^{+}+H_2{O}_{absorbed}\to O{H}^{•}{}^{}+H^{+}$$

(3)

$$h_{VB}^{+}+S_2{O}_{8 adsorbed}^{2-}\to S{O}_4^{•-}$$

(4)

E⁰ = 2.6 V (V vs. NHE, pH 7, 25 °C, 1 atm)

$$e_{CB}^{-}+O_{2 absorbed}\to O_2^{•-}$$

(5)

E⁰ = −0.33 V (V vs. NHE, pH 7, 25 °C, 1 atm)

As mentioned in the text, the use of heterojunctions plays a key role in overcoming the main drawbacks of photocatalysts. Thus, the oxidation processes involved in the formation of ROS might be more favored by the composite than its isolated components [122]. The coupling of g-C₃N₄ with different semi-conductors gives birth to an extended family of direct Z-scheme heterojunctions, active in AOPs (Figure 7) and, subsequently, applied in pollutant remediation [67].

In the following pages, we will analyze some of the most important examples of pollutant photodegradation by direct Z-schemes based on g-C₃N₄. Mainly, we will focus on some of the most representative pollutants used in photodegradation studies and some of the most illustrative results are summarized in Table 1.

Among the different pollutants, rhodamine B (RhB) is one of the most widely studied in recent reports related to this topic. Indeed, this water-soluble molecule is used as a red dye in the textile industry, thus representing one of the most concentrated wastes in water. Being classified as carcinogenic, RhB must be removed from the residual water, and this can be achieved by the use of photocatalysis [123]. Metal-free g-C₃N₄-based direct Z-scheme photocatalysts have been extensively applied in the degradation of RhB with highly remarkable results. M. Tang et al. [124] developed a ternary system composed of AgI/Ag₃PO₄/g-C₃N₄ with high activity in the degradation of RhB and a neonicotinoid pesticide due to the low recombination rate assured by the presence of the three compounds, as shown in Figure 8a. Indeed, the use of isolated components or even binary systems containing these materials did not lead to the same high yield in degradation. This finding was also demonstrated by Y. He et al. [116] using SnO_2-x/g-C₃N₄ composite in RhB degradation. Figure 8b depicts another advantage of this heterostructure, i.e., an increase in the surface area, which is known to be one of the main drawbacks of the application of g-C₃N₄ in catalysis [116]. Another noteworthy example is the 2D/2D SnS₂/g-C₃N₄ heterostructure developed by H. Che et al. [125], where the efficiency of the charge carrier separation is enhanced by the high interfacial contact area generated from the contact between the 2D materials. In addition, the photocatalytic activity of this catalyst is boosted because both OH^• and O₂^•− radicals might be generated, as presented in Figure 8c [126]. Considering lower dimensionality systems, 0D materials have found applicability in the generation of novel composites. For example, Y. Fu et al. [111] developed the MoS₂ QD/g-C₃N₄ heterostructure and demonstrated that the interaction between the 0D QD and 2D moieties can stabilize the QDs. Photocurrent measurements (Figure 8d) highlight that in the composites, the interfacial charge transfer is faster and the recombination of photoexcited charges is inhibited, thus boosting the photocatalytic activity in RhB degradation.

Tetracycline (TC) is a common antibiotic but with many side effects to live bodies. Its wide usage in husbandry (cattle, swine, poultry, and fishery) leads to an increase in its concentration in water and poses a threat to the natural balance. Among other degradation techniques, photocatalysis has found application in the degradation of this molecule [127,128,129], and g-C₃N₄-based direct Z-scheme photocatalysts have been extensively applied for this purpose. One recent report developed by J. Kang et al. [130] shows how the construction of the heterostructure g-C₃N₄/Na-BiVO₄ can boost the photodegradation of TC, promoted in this case by the formation of peroxymonosulfate radical. Indeed, the calculated CB and VB presented in Figure 9a demonstrate that the use of g-C₃N₄ can overcome one of the main drawbacks of the promising photocatalyst BiVO₄, i.e., the low CB potential. Similarly, peroxymonosulfate-mediated TC photodegradation was demonstrated by C. Jin et al. [131] to be successful when using the 2D/2D Co₃O₄/g-C₃N₄ heterostructure. In this case, the two-dimensionality allows a better separation and transfer of the charges, as demonstrated by the photoluminescence measurements in Figure 9b. In addition, S. Wang [132] et al. showed that the use of α-Fe₂O₃/g-C₃N₄ in TC degradation could improve the photocatalytic activity of the isolated components due to the formation of interfacial -NH/-NH₂ groups on the g-C₃N₄ surface. The synergy between the two components is determined by the fact that the defect-rich g-C₃N₄ improves the absorption of visible light on the material and α-Fe₂O₃ diminishes the recombination rate.

RhB and TC represent two of the most common pollutants that have been studied in photodegradation processes, but, certainly, according to the latest publications, there are other dangerous pollutants, which are present worldwide in high concentrations in water [133]. Consequently, photocatalysts based on g-C₃N₄ have also been used for the degradation of different pharmaceutical products or persistent organic pollutants (POPs), e.g., fluoroquinolone antibiotics, benzodiazepines, or parabens. Y. Zhou et al. [134] successfully used a 3D/2D MOF-derived CoCeO_x/g-C₃N₄ Z-scheme heterojunction in the photocatalytic degradation of carbamazepine. The presence of the MOF-derived CoCeO_x increased the surface area, overcoming one of the main drawbacks of g-C₃N₄. Additionally, J. Huang et al. [135] showed that the combination of g-C₃N₄ and TiO₂ originated an active photocatalyst for the visible-light elimination of enrofloxacin in water, as indicated in Figure 9c. In this case, the presence of g-C₃N₄ favors the absorbance of TiO₂ under visible light, giving more applicability to the well-known photocatalyst TiO₂. Furthermore, J. Meng et al. [136] synthesized the WO₃/g-C₃N₄ heterojunction for the degradation of methylparaben. In this case, the heterostructure avoids the fast recombination typical in both materials and the presence of g-C₃N₄ enhances the formation of O₂^•− radicals, which are involved in the degradation process (Figure 9d).

In summary, direct Z-schemes are a great alternative to improve the catalytic activity of bare photocatalysts in pollutant abatement, obtaining total conversion in a few hours. However, it is necessary to test the photocatalysts under more realistic conditions since most of the catalytic tests were performed with isolated pollutants in pure water.

Figure 7. ROS formation catalyzed by a heterostructure. Reproduced from [137]. Copyright 2022, Elsevier.

Figure 7. ROS formation catalyzed by a heterostructure. Reproduced from [137]. Copyright 2022, Elsevier.

Figure 8. (a) Photocatalytic mechanism in the presence of the AgI/Ag₃PO₄/g-C₃N₄ heterostructure. Reproduced from [124]. Copyright 2020, Elsevier. (b) TG profiles of g-C₃N₄, its composite with SnO_2-x (SC), and their physical mixture (SC-PM). Reproduced from [116]. Copyright 2015, Elsevier. (c) Proposed photocatalytic mechanism for the degradation of RhB by SnS₂/g-C₃N₄. Reproduced from [126]. Copyright 2019, Elsevier. (d) Transient photocurrent response of g-C₃N₄ and MoS₂ QD/g-C₃N₄ composites. Reproduced from [111]. Copyright 2017, Elsevier.

Figure 8. (a) Photocatalytic mechanism in the presence of the AgI/Ag₃PO₄/g-C₃N₄ heterostructure. Reproduced from [124]. Copyright 2020, Elsevier. (b) TG profiles of g-C₃N₄, its composite with SnO_2-x (SC), and their physical mixture (SC-PM). Reproduced from [116]. Copyright 2015, Elsevier. (c) Proposed photocatalytic mechanism for the degradation of RhB by SnS₂/g-C₃N₄. Reproduced from [126]. Copyright 2019, Elsevier. (d) Transient photocurrent response of g-C₃N₄ and MoS₂ QD/g-C₃N₄ composites. Reproduced from [111]. Copyright 2017, Elsevier.

Figure 9. (a) UV–vis DRS spectra (i) and Mott–Schottky plots (ii) of catalysts, XPS valence spectra (iii) of g-C₃N₄, BiVO₄, and Na-BiVO₄. Reproduced from [130]. Copyright 2021, Elsevier. (b) Photoluminescence spectra: (i) g-C₃N₄, (ii) 1%Co₃O₄/g-C₃N₄, (iii) 3%Co₃O₄/g-C₃N₄, (iv) 5%Co₃O₄/g-C₃N₄, (v) 10%Co₃O₄/g-C₃N₄, and (vi) 15%Co₃O₄/g-C₃N₄. Reproduced from [131]. Copyright 2020, Elsevier. (c) Proposed visible-light photocatalytic enrofloxacin degradation mechanism by g-C₃N₄/TiO₂. Reproduced from [135]. Copyright 2020, Elsevier. (d) Charge carrier transfer mechanism in the presence of WO₃/g-C₃N₄ direct Z-scheme photocatalyst. Reproduced from [136]. Copyright 2021, Elsevier.

Figure 9. (a) UV–vis DRS spectra (i) and Mott–Schottky plots (ii) of catalysts, XPS valence spectra (iii) of g-C₃N₄, BiVO₄, and Na-BiVO₄. Reproduced from [130]. Copyright 2021, Elsevier. (b) Photoluminescence spectra: (i) g-C₃N₄, (ii) 1%Co₃O₄/g-C₃N₄, (iii) 3%Co₃O₄/g-C₃N₄, (iv) 5%Co₃O₄/g-C₃N₄, (v) 10%Co₃O₄/g-C₃N₄, and (vi) 15%Co₃O₄/g-C₃N₄. Reproduced from [131]. Copyright 2020, Elsevier. (c) Proposed visible-light photocatalytic enrofloxacin degradation mechanism by g-C₃N₄/TiO₂. Reproduced from [135]. Copyright 2020, Elsevier. (d) Charge carrier transfer mechanism in the presence of WO₃/g-C₃N₄ direct Z-scheme photocatalyst. Reproduced from [136]. Copyright 2021, Elsevier.

Table 1. g-C₃N₄-based direct Z-scheme photocatalysts for pollutant remediation.

Z-Scheme Photocatalyst	Fabrication	Irradiation Source	Pollutant/Photodegradation Performance/Reaction Time (min)	Reference
SnO2−x/g-C3N4	solid state	350W Xenon lamp (420 nm < λ < 800 nm)	RhB/100%/60 min	[116]
MoS2 QD/g-C3N4	microemulsion method	350W Xenon lamp (λ > 420 nm)	RhB/100%/9 min	[111]
Bi3O4Cl/g-C3N4	solid phase calcination method	250W Xenon lamp (λ > 420 nm)	RhB/98.3%/90 min	[125]
β-Bi2O3/g-C3N4	sonication	350W Xenon lamp (λ > 420 nm)	RhB/98%/80 min	[138]
g-C3N4/ZnO	thermal atomic layer deposition	300W Xenon lamp	Cephalexin/92.7%/60 min	[139]
SnS2/g-C3N4	solvothermal method	300W Xenon lamp (λ > 400 nm)	RhB/94.8%/60 min	[126]
Fe3O4-OQs/Bi2O4/g-C3N4	hydrothermal method	250W Xenon lamp (λ > 420 nm)	RhB/92.7%/160 min	[140]
g-C3N4/TiO2	solvent evaporation method	500 W Xenon lamp	RhB/100%/20 min	[141]
α-Fe2O3/g-C3N4	calcination	100W LED lamp (λ = 420 nm)	Tetracycline/95.0%/60 min	[142]
ZnO/g-C3N4	hydrothermal method	Visible light	Atrazine/90%/180 min	[143]
Co3O4/g-C3N4	solid state	300W Xenon lamp (λ > 420 nm)	Tetracycline/90.0%/60 min	[131]
AgI/Ag3PO4/g-C3N4	in situ ion exchange method	300W Xenon lamp (λ > 420 nm)	Nitenpyram/100%/4 min	[124]
g-C3N4/anatase TiO2	calcination	350W Xenon lamp (λ > 420 nm)	Enrofloxacin/98.5%/60 min	[135]
g-C3N4/Na-BiVO4	hydrothermal method	300W Xenon lamp (λ > 420 nm)	Tetracycline/98.2%/40 min	[130]
WO3/g-C3N4	hydrothermal method	300W Xenon lamp (λ > 400 nm)	MPB/98.2%/60 min	[136]
ZnO/g-C3N4	hydrothermal method	Solar light	Rh B/98%/100 min	[144]
α-Fe2O3/g-C3N4	sonication	500 W Xenon lamp	Tetracycline/97.1%/80 min	[132]
CoCeOx/g-C3N4	solid state	300W Xenon lamp (λ > 420 nm)	Carbamazepine/90.1%/60 min	[134]

Table 1. g-C₃N₄-based direct Z-scheme photocatalysts for pollutant remediation.

Z-Scheme Photocatalyst	Fabrication	Irradiation Source	Pollutant/Photodegradation Performance/Reaction Time (min)	Reference
SnO2−x/g-C3N4	solid state	350W Xenon lamp (420 nm < λ < 800 nm)	RhB/100%/60 min	[116]
MoS2 QD/g-C3N4	microemulsion method	350W Xenon lamp (λ > 420 nm)	RhB/100%/9 min	[111]
Bi3O4Cl/g-C3N4	solid phase calcination method	250W Xenon lamp (λ > 420 nm)	RhB/98.3%/90 min	[125]
β-Bi2O3/g-C3N4	sonication	350W Xenon lamp (λ > 420 nm)	RhB/98%/80 min	[138]
g-C3N4/ZnO	thermal atomic layer deposition	300W Xenon lamp	Cephalexin/92.7%/60 min	[139]
SnS2/g-C3N4	solvothermal method	300W Xenon lamp (λ > 400 nm)	RhB/94.8%/60 min	[126]
Fe3O4-OQs/Bi2O4/g-C3N4	hydrothermal method	250W Xenon lamp (λ > 420 nm)	RhB/92.7%/160 min	[140]
g-C3N4/TiO2	solvent evaporation method	500 W Xenon lamp	RhB/100%/20 min	[141]
α-Fe2O3/g-C3N4	calcination	100W LED lamp (λ = 420 nm)	Tetracycline/95.0%/60 min	[142]
ZnO/g-C3N4	hydrothermal method	Visible light	Atrazine/90%/180 min	[143]
Co3O4/g-C3N4	solid state	300W Xenon lamp (λ > 420 nm)	Tetracycline/90.0%/60 min	[131]
AgI/Ag3PO4/g-C3N4	in situ ion exchange method	300W Xenon lamp (λ > 420 nm)	Nitenpyram/100%/4 min	[124]
g-C3N4/anatase TiO2	calcination	350W Xenon lamp (λ > 420 nm)	Enrofloxacin/98.5%/60 min	[135]
g-C3N4/Na-BiVO4	hydrothermal method	300W Xenon lamp (λ > 420 nm)	Tetracycline/98.2%/40 min	[130]
WO3/g-C3N4	hydrothermal method	300W Xenon lamp (λ > 400 nm)	MPB/98.2%/60 min	[136]
ZnO/g-C3N4	hydrothermal method	Solar light	Rh B/98%/100 min	[144]
α-Fe2O3/g-C3N4	sonication	500 W Xenon lamp	Tetracycline/97.1%/80 min	[132]
CoCeOx/g-C3N4	solid state	300W Xenon lamp (λ > 420 nm)	Carbamazepine/90.1%/60 min	[134]

4.2. H₂ Production

One of the greatest current challenges of our society from an environmental and energy point of view is the production and use of clean fuel from renewable energies [2,145]. During the last decades, H₂ has become a highly relevant alternative to fossil fuels because it has clean combustion and high energy density [145]. Therefore, it is essential to develop technologies to allow the production of H₂ without using fossil fuels [146]. This subsection is placed here, as the water splitting uses similar starting radicals to pollutant abatement in aqueous ambience. The generation of H₂ by water splitting using sunlight has received great interest from the scientific community and in society [147,148]. The water splitting reaction that generates H₂ and O₂ (Equation (6)) is a non-spontaneous endergonic reaction (ΔG = +237 KJmol⁻¹), which means that energy is needed for the reaction to occur [148]:

$$H_{2}O\to \raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.O_2+H_2 \left(\mathsf{\Delta }G=237{ \mathrm{KJmol}}^{-1}\right)$$

(6)

One alternative to the generation of H₂ from water splitting is the use of photocatalysis, which would require the use of sunlight and a semiconductor (photocatalysis). For the reaction to occur, the CB potential of the semiconductor should be more negative than the reduction potential of H₂O (0 V vs. NHE) for H₂ generation, and the VB potential of the semiconductor should be more positive than the oxidation potential of H₂O (1.23 V vs. NHE) for O₂ generation [148,149]. An interesting semiconductor for this reaction is g-C₃N₄ since the VB and CB positions are located at approximately +1.6 and −1.1 eV, respectively, with a band gap of ~2.7 eV, making the reaction of water splitting possible using sunlight [56,150]. However, the g-C₃N₄ photocatalyst presents low activity in water splitting [23], as mentioned in Section 2. To solve this low activity of g-C₃N₄ in water splitting, the scientific community has focused on studying heterojunctions based on g-C₃N₄ [26,60]. A heterojunction that has aroused the interest of the scientific community is the direct Z-scheme because the latter is present in the systems responsible for the water splitting reaction in photosynthesis due to the decrease in the e⁻-h⁺ pair recombination rate [38,98]. This makes the direct Z-schemes promising systems, which may lead to adequate quantum efficiency values for their commercialization [38]. In this subsection, several direct Z-scheme photocatalysts based on g-C₃N₄ in different conditions used in water splitting are summarized in Table 2. In addition, a detailed discussion of the most relevant Z-scheme photocatalysts from the point of view of the catalytic activity and reaction mechanism will be described in the main text.

R. Ye et al. [151] described a CoTiO₃/g-C₃N₄ novel Z-scheme heterostructure photocatalyst for H₂ evolution (858 µmol g⁻¹ h⁻¹) under visible light with effectively intimate interfaces linked through the Co–O–N or Ti–O–N bonds, as shown in Figure 10a. The authors showed that this system presents a direct Z-scheme mechanism by PL, PEC, and ESR analyses. The work described by R. Ye et al. focusing on Z-scheme CoTiO₃/g-C₃N₄ was useful for further development of novel heterojunction photocatalytic systems based on perovskite, polymer, and composite photocatalysts for photocatalytic H₂ production. X. She et al. [152] showed a novel strategy for the synthesis of an α-Fe₂O₃/2D g-C₃N₄ Z-scheme hybrid structure containing ultrathin g-C₃N₄ platelets. In the synthesis methodology used in this study, α-Fe₂O₃ promotes the formation of g-C₃N₄ nanosheets. Moreover, the authors claimed that the Fe₂O₃/2D g-C₃N₄ direct Z-scheme photocatalyst exhibits unprecedented photocatalytic activity toward H₂ evolution (31,400 µmol g⁻¹ h⁻¹), with a remarkably high quantum efficiency of 44.35% at a wavelength of 420 nm, as shown in Figure 10b. The enhanced performance of H₂ evolution in α-Fe₂O₃/2D g-C₃N₄ was attributed to the 2D structure of g-C₃N₄, allowing an efficient transfer of the photoexcited electron to the reactant. Another factor relevant to improving the photocatalytic activity of this system is the Z-scheme structure due to the low electron-hole recombination in both α-Fe₂O₃ and 2D g-C₃N₄ and the close interface contact between α-Fe₂O₃ and 2D g-C₃N₄, enabling fast electron transfer from the conduction band of α-Fe₂O₃ to the valence band of 2D g-C₃N₄. This work shows that the dimensionality (2D) and the porous hierarchy of the systems present a great effect on the photocatalytic activity of Z-schemes. Moreover, Fe₂O₃/2D g-C₃N₄ direct Z-scheme photocatalysts exhibit great values in H₂ production, indicating the effectivity and interest of these artificial Z-schemes for their use in this application. Considering the effect of the 2D materials, Y. Yang, et al. [153] designed and constructed a direct Z-scheme van der Waals heterojunction composed of ultrathin WO₃·H₂O and g-C₃N₄ nanosheets using simple hydrothermal synthesis to achieve efficient water splitting without the addition of any sacrificial agents (482 µmol g⁻¹ h⁻¹). This WO₃·H₂O/g-C₃N₄ system could efficiently transport the electrons, decreasing the e⁻-h⁺ recombination rates, as shown in Figure 10c, which led to a considerable improvement in the photocatalytic performance. Another relevant parameter in the study of Z-schemes is the presence of interfacial defects since they might be beneficial to the recombination of interfacial photogenerated e⁻/h⁺ and could serve as Z-scheme pathways for charge transfer. L. Kong et al. [154] fabricated a direct Z-scheme g-C₃N₄/Ti³⁺-doped TiO₂ to show this effect. The authors observed that the Ti³⁺ defects were introduced under CB of TiO₂, which improved the interfacial charge transfer channels for Z-scheme g-C₃N₄/Ti³⁺-doped TiO₂ composite and improved the visible light absorption. These properties were studied using a radical species trapping methodology, UV-Vis spectroscopy, and EPR analysis. Due to the enhanced properties, this system has high activity (1938 μmol∙h⁻¹∙g⁻¹) under simulated solar light and good stability in photocatalytic H₂ evolution. S. Qin et al. [155] showed a promising way to construct Z-scheme heterostructures based on metal-organic cages (MOCs) and semiconductors for their use in photocatalytic H₂ generation. The authors synthesized an MOC, integrating four organic photosensitized ligands M⁴⁻ and two Pd²⁺ catalytic centers. Then, the MOC was successfully immobilized by hydrogen bonds to obtain a robust heterogeneous direct Z-scheme g-C₃N₄/MOC-Q1 (Figure 10e). This innovative system presents high H₂ evolution activity (4495 μmol g^–1 h^–1). In this manuscript, it was described that Z-scheme systems improved the photocatalytic H₂ production rate, but, on the other hand, the photoconversion efficiency is still far from the requirement for practical applications.

One novel methodology based on direct Z-scheme system photocatalysts is the design of dual-Z-scheme systems that ensure the efficient transmission of photogenerated carriers and further optimize the structure of the Z-scheme system. Y. Yang et al. [114] successfully constructed an efficient dual-Z-scheme heterojunction, CdS/W₁₈O₄₉/g-C₃N₄ (CWOCN), by introducing CdS nanoparticles in the W₁₈O₄₉/g-C₃N₄ (WOCN) composite. This dual-Z-scheme system exhibits extraordinary photocatalytic activity and stability. The TRPL measurement results demonstrated that the formation of a dual-Z-scheme system facilitates the transmission of photogenerated electrons and prolonged the electrons’ lifetime, thus boosting the photocatalytic activity. It led to a prominent photocatalytic H₂ evolution rate (11,658 μmol h⁻¹ g⁻¹). This study offers guidance for exploiting and designing efficient and practical photocatalysts.

To conclude this section, there are several methodologies based on direct Z-scheme system photocatalysts used to obtain photocatalysts that are active in H₂ production. However, this technology is far from commercialization due to the low reproducibility of the materials and low quantum efficiency.

Figure 10. (a) Schematic illustration of the mechanisms for charge carrier separation of CoTiO₃/g-C₃N₄ direct Z-schemes and the molecular interaction between CoTiO₃ and g-C₃N₄ through the formation of chemical bonds. Reproduced from [151]. Copyright 2016, ACS Publications. (b) Catalytic activity (Turnover frequency) of ML g-C₃N₄, α-Fe₂O₃/ML g-C₃N₄, 2D g-C₃N₄, and α-Fe₂O₃/2D g-C₃N₄ for water splitting. Reproduced from [152]. Copyright 2017, Willey. (c) PL spectra of g-C₃N₄, WO₃·H₂O, and WO₃·H₂O/g-C₃N₄ and visible-light-driven photocatalytic H₂ and O₂ generation for WO₃·H₂O/g-C₃N₄. Reproduced from [153]. Copyright 2018, Willey. (d) Schematic diagram of the mechanism for the separation and transfer of photoinduced charge carrier in CN-T-H composite. Reproduced from [154]. Copyright 2018, Elsevier. (e) Schematic illustration of a heterostructure of g-C₃N₄/MOC-Q1. Reproduced from [155]. Copyright 2021, ACS Publications. (f) The mechanism of dual-Z-scheme heterojunction (CWOCN) for H₂ production. Reproduced from [114]. Copyright 2021, Willey.

Figure 10. (a) Schematic illustration of the mechanisms for charge carrier separation of CoTiO₃/g-C₃N₄ direct Z-schemes and the molecular interaction between CoTiO₃ and g-C₃N₄ through the formation of chemical bonds. Reproduced from [151]. Copyright 2016, ACS Publications. (b) Catalytic activity (Turnover frequency) of ML g-C₃N₄, α-Fe₂O₃/ML g-C₃N₄, 2D g-C₃N₄, and α-Fe₂O₃/2D g-C₃N₄ for water splitting. Reproduced from [152]. Copyright 2017, Willey. (c) PL spectra of g-C₃N₄, WO₃·H₂O, and WO₃·H₂O/g-C₃N₄ and visible-light-driven photocatalytic H₂ and O₂ generation for WO₃·H₂O/g-C₃N₄. Reproduced from [153]. Copyright 2018, Willey. (d) Schematic diagram of the mechanism for the separation and transfer of photoinduced charge carrier in CN-T-H composite. Reproduced from [154]. Copyright 2018, Elsevier. (e) Schematic illustration of a heterostructure of g-C₃N₄/MOC-Q1. Reproduced from [155]. Copyright 2021, ACS Publications. (f) The mechanism of dual-Z-scheme heterojunction (CWOCN) for H₂ production. Reproduced from [114]. Copyright 2021, Willey.

Table 2. Summary of direct Z-scheme photocatalysts based on g-C₃N₄ for water splitting.

Z-Scheme Photocatalyst	Fabrication Methodology	Irradiation Source	H2 Production Activity (µmol g−1 h−1) and AQE	Reference
CoTiO3/g-C3N4	Solid-State	Xenon lamp (300 W, λ ≥ 420 nm)	858 AQE: 38.4% (365 nm)	[151]
g-C3N4/ZnO	Deposition	Xenon lamp (300 W, λ ≥ 420 nm)	322	[156]
g-C3N4/PSi	Polycondensation reaction	Xenon lamp (300 W, λ ≥ 400 nm)	870	[157]
2D α-Fe2O3/g-C3N4	Solid-State	Xenon lamp (300 W, λ ≥ 420 nm)	31400 AQE: 44.35% (420 nm)	[152]
WO3.H2O/g-C3N4	Hydrothermal method	Xenon lamp (300 W, λ > 400 nm)	482 AQE: 6.2% (420 nm)	[153]
g-C3N4/Ti3+-TiO2	Solid-State	Xenon lamp (300 W, λ > 400 nm)	1938	[154]
Nb2O5/g-C3N4	Hydrothermal	Xenon lamp (1000 W, 1.5G)	110,000	[115]
Bi2O2CO3/g-C3N4	Heat treatment method	Xenon lamp (300 W, λ ≥ 400 nm)	965 AQE: 7.14% (420 nm)	[158]
2D/2D g-C3N4/Sn3O4	Calcined in N2	Xenon lamp (300 W, λ ≥ 400 nm)	1960	[159]
g-C3N4/MOC-Q1	Deposition	Xenon lamp (300 W, λ ≥ 420 nm)	4495 AQE: 0.50% (425 nm)	[155]
CdS/W18O49/g-C3N4 (CWOCN)	Chemical bath deposition	Xenon lamp (300 W, λ ≥ 420 nm)	11,658 AQE: 26.73% (420 nm)	[116]
Cu2O/g-C3N4	Solid-State	Xenon lamp (300 W, λ ≥ 420 nm)	266.3 AQE: 13.40% (420 nm)	[160]

Table 2. Summary of direct Z-scheme photocatalysts based on g-C₃N₄ for water splitting.

Z-Scheme Photocatalyst	Fabrication Methodology	Irradiation Source	H2 Production Activity (µmol g−1 h−1) and AQE	Reference
CoTiO3/g-C3N4	Solid-State	Xenon lamp (300 W, λ ≥ 420 nm)	858 AQE: 38.4% (365 nm)	[151]
g-C3N4/ZnO	Deposition	Xenon lamp (300 W, λ ≥ 420 nm)	322	[156]
g-C3N4/PSi	Polycondensation reaction	Xenon lamp (300 W, λ ≥ 400 nm)	870	[157]
2D α-Fe2O3/g-C3N4	Solid-State	Xenon lamp (300 W, λ ≥ 420 nm)	31400 AQE: 44.35% (420 nm)	[152]
WO3.H2O/g-C3N4	Hydrothermal method	Xenon lamp (300 W, λ > 400 nm)	482 AQE: 6.2% (420 nm)	[153]
g-C3N4/Ti3+-TiO2	Solid-State	Xenon lamp (300 W, λ > 400 nm)	1938	[154]
Nb2O5/g-C3N4	Hydrothermal	Xenon lamp (1000 W, 1.5G)	110,000	[115]
Bi2O2CO3/g-C3N4	Heat treatment method	Xenon lamp (300 W, λ ≥ 400 nm)	965 AQE: 7.14% (420 nm)	[158]
2D/2D g-C3N4/Sn3O4	Calcined in N2	Xenon lamp (300 W, λ ≥ 400 nm)	1960	[159]
g-C3N4/MOC-Q1	Deposition	Xenon lamp (300 W, λ ≥ 420 nm)	4495 AQE: 0.50% (425 nm)	[155]
CdS/W18O49/g-C3N4 (CWOCN)	Chemical bath deposition	Xenon lamp (300 W, λ ≥ 420 nm)	11,658 AQE: 26.73% (420 nm)	[116]
Cu2O/g-C3N4	Solid-State	Xenon lamp (300 W, λ ≥ 420 nm)	266.3 AQE: 13.40% (420 nm)	[160]

4.3. CO₂ Photoreduction

The great use of fossil fuels in the last decades has led to significant CO₂ emissions into the atmosphere, resulting in global warming responsible for climate change, which represents one of the greatest challenges of the current society [161,162]. An innovative alternative to reduce the emissions of CO₂ into the atmosphere would be to capture CO₂ [163] and transform it into high value-added products using sunlight, mimicking natural photosynthesis [164]. However, CO₂ is a stable covalent molecule due to the C=O bond, resulting in high dissociation energy (531 kJ mol⁻¹) [165]. CO₂ photoreduction is an alternative method to break the C=O bond using solar energy. Nevertheless, this process is a multielectron transfer process, which may result in a great variety of products with different carbon oxidation states, including CH₄, CH₃OH, HCHO, HCOOH, and CO [166] (see Equations (7)–(13)):

$$C{O}_2+2e^{-}\to C{O}_2^{·-}$$

(7)

E⁰ = −1.90 V (V vs. NHE, pH 7, 25 °C, 1 atm)

$$C{O}_2+2H^{+}+2e^{-}\to HCOOH$$

(8)

E⁰ = −0.61 V (V vs. NHE, pH 7, 25 °C, 1 atm)

$$C{O}_2+2H^{+}+2e^{-}\to CO+H_{2}O$$

(9)

E⁰ = −0.53 V (V vs. NHE, pH 7, 25 °C, 1 atm)

$$C{O}_2+4H^{+}+4e^{-}\to HCHO+H_{2}O$$

(10)

E⁰ = −0.48 V (V vs. NHE, pH 7, 25 °C, 1 atm)

$$C{O}_2+6H^{+}+6e^{-}\to C{H}_{3}OH+H_{2}O$$

(11)

E⁰ = −0.38 V (V vs. NHE, pH 7, 25 °C, 1 atm)

$$C{O}_2+8H^{+}+8e^{-}\to C{H}_4+2H_{2}O$$

(12)

E⁰ = −0.24 V (V vs. NHE, pH 7, 25 °C, 1 atm)

$$2H_{2}O+2e^{-}\to 2O{H}^{-}+H_2$$

(13)

E⁰ = −0.41 V (V vs. NHE, pH 7, 25 °C, 1 atm)

Another challenge in the photocatalytic reduction of CO₂ is that the reduction of H₂O to H₂ (see Equation (1)) has a similar reduction potential to the reduction of CO₂. This fact means that in the aqueous ambience, there is competition between both reactions [167]. To overcome the challenges presented above in the reduction of CO₂, the scientific community has focused on designing selective photocatalysts for the reduction of CO₂ to various high value-added products [168]. Among the different photocatalysts developed, g-C₃N₄ should be highlighted. G. Dong et al. reported that pristine g-C₃N₄ without any cocatalyst presented selectivity in the photocatalytic reduction of CO₂ into CO under visible-light irradiation in the presence of water vapor [169]. However, the efficiency of this photocatalyst was low due to the high e⁻-h⁺ pair recombination rate [47]. As mentioned throughout this review, an alternative used to reduce the e⁻-h⁺ recombination ratio, which is one of the challenges of g-C₃N₄, is the design and synthesis of heterojunctions [26]. Within the use of heterojunctions for the photoreduction of CO₂, artificial direct Z-scheme photocatalysts have received great interest from the scientific community since these systems mimic the electronic processes that occur in photosynthesis to reduce CO₂ into hydrocarbons [33,34]. Therefore, the study of these artificial systems for the reduction of CO₂ has grown noticeably in the last years, being a key approach to understanding the photocatalytic mechanism and to design efficient systems capable of being marketed [39,170,171]. In this subsection, several direct Z-scheme photocatalysts based on g-C₃N₄ used in CO₂ photoreduction are summarized in Table 3. In addition, a detailed discussion of the most relevant photocatalyst Z-scheme from a synthesis, catalytic activity, and reaction mechanism perspective will be described in the main text.

W. Yu et al. [172] described the coupling effect of ZnO for the improved photoactivity of g-C₃N₄ and the authors also proposed a direct Z-scheme mechanism to explain the better performances of the g-C₃N₄/ZnO binary composite photocatalytic system. The g-C₃N₄/ZnO photocatalytic system exhibited enhanced photocatalytic activity for CO₂ reduction compared with pure g-C₃N₄, with selectivity towards CH₃OH. This enhancement of photocatalytic CO₂ reduction activity is attributed to the highly efficient ZnO-to-g-C₃N₄ electron transfer occurring at the intimate contact interface between the g-C₃N₄ and ZnO phases. This work provided new insights into the rational construction of a g-C₃N₄-based photocatalytic system and the design of a direct Z-scheme system without an electron mediator for the photocatalytic CO₂ reduction reaction. To enhance the photocatalytic activity of the materials, Z. Jiang et al. [170] described that the Z-scheme hybrid material should present several characteristics to accelerate the reduction process, including a 3D hierarchical structure and preferably basic sites, which promote CO₂ adsorption. In addition, these systems should present an efficient separation of the e⁻-h⁺ pairs and an enhancement in the reduction character of electrons in the conduction band of g-C₃N₄. In this work, a hierarchical direct Z-scheme hybrid was developed by combining 3D urchin-like α-Fe₂O₃ and g-C₃N₄ photocatalysts (Figure 11a) for the photocatalytic reduction of CO₂ without the use of any sacrificial agent or cocatalyst. The hierarchical system improved the photocatalytic reduction of CO₂ to form CO (27.2 µmol g⁻¹ h⁻¹), and the system was stable for four cycles (Figure 11a). N.T. Thanh Truc et al. [173] synthesized a Nb-TiO₂/g-C₃N₄ direct Z-scheme system, in which the photo-excited e⁻ in the CB of the Nb-TiO₂ combines with the photo-excited h⁺ in the VB of the g-C₃N₄, preserving the existence of e⁻ in the CB of the g-C₃N₄ and h⁺ in the VB of Nb-TiO₂. Thus, the established Nb-TiO₂/g-C₃N₄ direct Z-scheme system produced a huge amount of available e⁻/h⁺ pairs for the reduction of CO₂ into various valuable fuels. Moreover, the authors observed that the photo-current response of the prepared Nb-TiO₂/g-C₃N₄ (~57.55 µA/cm²) was much higher than that of g-C₃N₄ (~3.92 µA/cm²) and Nb-TiO₂ (~12.41 µA/cm²) under the full visible light spectrum (Figure 11b). This result showed that the photo-excited e⁻ in the CB of the Nb-TiO₂ tends to combine with photo-excited h⁺ in the VB of the g-C₃N₄, preserving the existence of e⁻ in the CB of the g-C₃N₄ and h⁺ in the VB of the Nb-TiO₂, indicating the power of this characterization technique to address the Z-scheme mechanism. M. Lu et al. [92] synthesized 2D/2D ultrathin nanosheets of g-C₃N₄/BiVO₄ Z-scheme heterojunction via thermal-polymerization and a subsequent hydrothermal method. The 2D/2D g-C₃N₄/BiVO₄ ultrathin nanosheets exhibited a great photocatalytic evolution rate for CO (145 µmol g⁻¹ h⁻¹) and CH₄ (133 µmol g⁻¹ h⁻¹) (Figure 11c). The study presented by Lu et. al. evidenced the relevance of using 2D materials. Indeed, the direct Z-scheme heterojunction and face-to-face interfacial contact between 2D g-C₃N₄ and 2D BiVO₄ accelerated the charge transfer and separation, improving the catalytic activity of these systems in CO₂ reduction. B. Tahir et al. [117] designed 3D/2D WO₃/g-C₃N₄ microspheres with an effective interfacial contact using a facile single-step hydrothermal method. The direct growth of WO₃ microspheres with g-C₃N₄ enables good interaction among both semiconductors, enabling proficient charge carrier separation (Figure 11d). The synergistic effect with the larger interfacial contact area and proficient charge carrier separation of this system improves the CO and CH₄ production over WO₃/g-C₃N₄ (145 and 133 μmol g⁻¹ h⁻¹, respectively), which are 2- and 4-fold higher than the use of pristine g-C₃N₄. This work reveals that effective interfacial contact between both semiconductors present on the Z-scheme photocatalyst is a key factor to promoting the photocatalytic CO₂ conversion to solar fuels under visible light irradiation. J. Zhao et al. [174] showed the development of a novel direct Z-scheme Bi₁₉S₂₇Br₃/g-C₃N₄ composite using the ionic liquid-assisted solvent-thermal method. The Bi₁₉S₂₇Br₃/g-C₃N₄ composites showed enhanced CO₂ photoreduction activity for the production of CO (12.87 μmol g⁻¹ h⁻¹). The authors observed in this work that the C-S bond boosts the transfer of photogenerated charge between Bi₁₉S₂₇Br₃ and g-C₃N_4, as shown in Figure 11e. Therefore, the construction of a chemical bond-bridged direct Z-scheme is a promising strategy for precisely tailoring the photogenerated charge separation direction of the photocatalyst.

In conclusion, although the synthesis of Z-schemes has been successfully achieved and they are active in CO₂ photoreduction, nowadays, the photocatalytic efficiencies of Z-schemes are still too low for their commercialization.

Table 3. Summary of direct Z-scheme photocatalysts based on g-C₃N₄ for CO₂ photoreduction.

Z-Scheme Photocatalyst	Fabrication	Irradiation Source	Products/Production (µmol g−1 h−1)/AQE	Reference
ZnO/g-C3N4	Solid-state	300 W simulated solar Xe arc lamp	CH3OH: 0.6	[172]
SnO2-x/g-C3N4	Solid-state	500 W Xe lamp	CO: ~19 CH3OH: ~4 CH4: ~2	[116]
g-C3N4/SnS2	Hydrothermal method	300 W Xenon lamp (λ ≥ 420 nm)	CH3OH: 2.24 CO: 0.64	[91]
MoO3/g-C3N4	impregnation method	350 W Xenon lamp (800 nm > λ > 420 nm)	CO: ~18 CH3OH: ~7 CH4: ~1	[85]
α-Fe2O3/g-C3N4	Impregnation–hydrothermal method	Xenon lamp 0.21 Wcm−2	CO: 27.2 AQE: 0.963% (420 nm)	[170]
AgCl/g-C3N4	Deposition-precipitation method	11 W fluorescent lamp	CH4: ~2 CH3COOH: ~0.75 HCOOH: ~0.31 AQE: 0.211% (475 nm)	[175]
Cu2V2O7/g-C3N4	Calcination methodology	20 W white bulbs (700 nm > λ > 400 nm)	CH4: 305 CO: 166 O2: 706	[176]
g-C3N4/FeWO4	Sonochemical method	300 W Xenon lamp (100 mW/cm2)	CO: 6 AQE: ~0.3% (420 nm)	[83]
(Nb)TiO2/g-C3N4	Calcination methodology	Two 30 W white bulbs	CH4: 562 CO:420 HCOOH:698	[173]
2D/2D g-C3N4/BiVO4	Hydrothermal method	300 W Xenon lamp (λ ≥ 420 nm)	CO: ~5.2 CH4: ~4.6	[92]
NiMoO4/g-C3N4	Calcined methodology	30 W LED, (700 nm > λ > 400 nm)	CH4: 635 CO: 432 O2: 1853 HCOOH: 647	[177]
α-Fe2O3/g-C3N4	Hydrothermal method	300 W xenon lamp	CO: 17.8 AQE: 0.31% (420 nm)	[178]
3D/2D WO3/g-C3N4	Hydrothermal method	300 W Xenon lamp (100 mW cm−2)	CO: 145 CH4: 133	[117]
La2Ti2O7/g-C3N4	Ultrasonic-deposition method	Four blue LED (4 × 3 W, 420 nm)	CH3OH: ~4 CO: ~2.5 AQE: 3.61% (420 nm)	[179]
Bi2S3/g-C3N4	Hydrothermal method	300 W xenon lamp	CO: 6.84 CH4: 1.57 H2: 1.38 AQE: 2.31% (420 nm)	[180]
ZnO/ZnWO4/g-C3N4	Calcination method	Xenon lamp (300 W, 0.95 mW/cm2)	CH4: 6.2 CH3OH: 3.8 CH3CH2OH: 2.1 CO: 1.3	[181]
NiTiO3/g-C3N4	Ultrasonic-calcination method	300 W Xenon lamp (λ ≥ 420 nm)	CH3OH: 13.74	[182]
Bi19S27Br3/g-C3N4	Physical mixture (Strong grinding)	300 W xenon lamp	CO: 12.87	[174]

Table 3. Summary of direct Z-scheme photocatalysts based on g-C₃N₄ for CO₂ photoreduction.

Z-Scheme Photocatalyst	Fabrication	Irradiation Source	Products/Production (µmol g−1 h−1)/AQE	Reference
ZnO/g-C3N4	Solid-state	300 W simulated solar Xe arc lamp	CH3OH: 0.6	[172]
SnO2-x/g-C3N4	Solid-state	500 W Xe lamp	CO: ~19 CH3OH: ~4 CH4: ~2	[116]
g-C3N4/SnS2	Hydrothermal method	300 W Xenon lamp (λ ≥ 420 nm)	CH3OH: 2.24 CO: 0.64	[91]
MoO3/g-C3N4	impregnation method	350 W Xenon lamp (800 nm > λ > 420 nm)	CO: ~18 CH3OH: ~7 CH4: ~1	[85]
α-Fe2O3/g-C3N4	Impregnation–hydrothermal method	Xenon lamp 0.21 Wcm−2	CO: 27.2 AQE: 0.963% (420 nm)	[170]
AgCl/g-C3N4	Deposition-precipitation method	11 W fluorescent lamp	CH4: ~2 CH3COOH: ~0.75 HCOOH: ~0.31 AQE: 0.211% (475 nm)	[175]
Cu2V2O7/g-C3N4	Calcination methodology	20 W white bulbs (700 nm > λ > 400 nm)	CH4: 305 CO: 166 O2: 706	[176]
g-C3N4/FeWO4	Sonochemical method	300 W Xenon lamp (100 mW/cm2)	CO: 6 AQE: ~0.3% (420 nm)	[83]
(Nb)TiO2/g-C3N4	Calcination methodology	Two 30 W white bulbs	CH4: 562 CO:420 HCOOH:698	[173]
2D/2D g-C3N4/BiVO4	Hydrothermal method	300 W Xenon lamp (λ ≥ 420 nm)	CO: ~5.2 CH4: ~4.6	[92]
NiMoO4/g-C3N4	Calcined methodology	30 W LED, (700 nm > λ > 400 nm)	CH4: 635 CO: 432 O2: 1853 HCOOH: 647	[177]
α-Fe2O3/g-C3N4	Hydrothermal method	300 W xenon lamp	CO: 17.8 AQE: 0.31% (420 nm)	[178]
3D/2D WO3/g-C3N4	Hydrothermal method	300 W Xenon lamp (100 mW cm−2)	CO: 145 CH4: 133	[117]
La2Ti2O7/g-C3N4	Ultrasonic-deposition method	Four blue LED (4 × 3 W, 420 nm)	CH3OH: ~4 CO: ~2.5 AQE: 3.61% (420 nm)	[179]
Bi2S3/g-C3N4	Hydrothermal method	300 W xenon lamp	CO: 6.84 CH4: 1.57 H2: 1.38 AQE: 2.31% (420 nm)	[180]
ZnO/ZnWO4/g-C3N4	Calcination method	Xenon lamp (300 W, 0.95 mW/cm2)	CH4: 6.2 CH3OH: 3.8 CH3CH2OH: 2.1 CO: 1.3	[181]
NiTiO3/g-C3N4	Ultrasonic-calcination method	300 W Xenon lamp (λ ≥ 420 nm)	CH3OH: 13.74	[182]
Bi19S27Br3/g-C3N4	Physical mixture (Strong grinding)	300 W xenon lamp	CO: 12.87	[174]

5. Conclusions and Outlook

In conclusion, g-C₃N₄-based direct Z-scheme photocatalysts are innovative alternatives for overcoming the main drawbacks of the parent material, i.e., high e⁻-h⁺ recombination rate and inadequate redox potential. In this review, a comprehensive overview of the recent research on g-C₃N₄-based Z-scheme photocatalysts is described, focusing on the utilization of these systems in three environmental applications, such as pollutant abatement, H₂ evolution, and CO₂ reduction. There is undoubtedly a substantial amount of ground to cover by the scientific community since there are few reports where the synthesis and catalytic activity of novel g-C₃N₄-based direct Z-scheme photocatalysts have been studied. Especially, several synthetic methodologies (hydrothermal, solid-state, impregnation, etc.), characterization techniques to study the properties of Z-schemes (EPR, PL …), and the application of these materials in environmental reactions have been investigated, focusing on the effect of the catalysts and their catalytic activity to obtain efficient solar photocatalysts. Nevertheless, the photocatalytic efficiencies are still low, and many catalytic mechanisms remain unclear. These characteristics represent the main drawbacks hindering the commercialization of these systems. The challenges for new researchers in the design and applicability of g-C₃N₄-based direct Z-scheme photocatalysts might be in the synthetic methodology of this photocatalyst, focusing on the interaction between g-C₃N₄ and another SC, since the intimate contact between both systems has a strong influence on the photocatalytic activity.

The effect of the dimensionality and surface chemistry of g-C₃N₄ and the hetero- counterpart present in the Z-scheme photocatalyst is a paradigm in heterostructure synthesis, and thus, in the improvement of the photocatalytic activity. Although the transfer directions of photogenerated charge carriers in Z-scheme systems are studied by EPR, photoluminescence spectroscopy, reactive species scavenging experiments, and theoretical calculations, it is necessary to develop more powerful characterization tools to investigate the charge-transfer mechanism. Subsequently, it is also mandatory to understand the electron photoexcitation, trapping, and migration at the interface through a series of operando characterization techniques and computational studies.

Regarding the applicability of Z-schemes in environmental reactions, it is essential to develop new techniques to elucidate the mechanisms of catalytic reactions in all the stages of these catalytic processes (adsorption, reaction, and desorption), considering that the monitoring of the reaction is sometimes complex due to the formation of by-products. This scenario opens the door for the design and engineering of new Z-schemes based on g-C₃N₄, and the development of characterization techniques and studies of new mechanisms to obtain active photocatalysts in environmental applications under visible light.