Modification Strategies of g-C₃N₄-Based Materials for Enhanced Photoelectrocatalytic Degradation of Pollutants: A Review

Abstract

Graphite carbon nitride (g-C₃N₄) is a low band gap non-metallic polymer semiconductor that has broad application prospects and is an ideal material for absorbing visible light, as g-C₃N₄ materials have strong oxidation properties and are easy to modify. The structure formation of g-C₃N₄-based materials makes a series of photocatalytic synthesis reactions possible and improves photocatalytic reaction activity. In this paper, the development history, structures, and performance of g-C₃N₄ are briefly introduced, and the modification strategies of g-C₃N₄ are summarized to improve its photocatalytic and photoelectric catalytic properties via doping, heterojunction construction, etc. The light absorption and utilization of the catalysts are also analyzed in terms of light source conditions, and the application of g-C₃N₄ and its modified materials in photocatalysis and photocatalytic degradation is reviewed.

1. Introduction

At present, pollutants caused by rising energy use and human activity are currently one of the most serious environmental issues. The world is trying to use solar energy for energy conversion, environmental protection, and water purification [1]. Various environmental pollutants (dyes [2], pesticides [3], drugs [4], phenols [5], metal ions [6], etc.) in water are toxic and recalcitrant in the environment, and their removal is of great concern [7].

Semiconductor photocatalytic technology is precisely based on the chemical conversion and storage of solar energy as the core. Removing toxic and harmful pollutants through photocatalysis will provide people with a green and environmentally friendly living space. In contrast, the conversion of solar energy into hydrogen as well as other chemical energy sources, such as photolysis of water and carbon dioxide conversion, is expected to completely solve the energy crisis brought about by the increasing depletion of energy. However, the low utilization efficiency of absorbed light and the high carrier recombination seriously restrict the practical application of photocatalytic technology. Photocatalysis technology is the foundation of photoelectrocatalysis (PEC), which can effectively promote the separation of photogenerated electron-hole pairs by utilizing photoelectric synergy, which has the advantages of easy separation and recovery. According to studies, applying a certain bias voltage to semiconductor materials can segregate the photogenerated charge while significantly extending the life of the photogenerated carrier, and the photocatalytic efficiency will thereafter be significantly increased [8,9,10]. Currently, photoelectrocatalytic technology is widely used in the fields of multi-pollutant degradation, carbon dioxide reduction, water oxidation, and hydrogen production [11,12,13].

Carbon materials (graphene, g-C₃N₄, carbon dots, etc.) have excellent electrical conductivity [14,15,16]. Through doping elements, efect engineering, grafting functional groups, and other methods, they can easily design their surface chemical properties and regulate their structure. The interface interaction between the inorganic carbon material and semiconductor can enhance the semiconductor’s photo-response performance, photogenerated carrier migration characteristics, adsorption efficiency, and catalytic active site, all of which contribute to the semiconductor’s photoelectrocatalytic performance [17]. Because of this, carbon materials have a promising future in photoelectrocatalytic degradation of pollutants and other areas [18]. As g-C₃N₄ is a unique 2D layered nonmetallic material [19,20,21], it offers the benefits of a straightforward synthesis process and strong thermal stability, and its energy band structure is highly suitable for the two key steps of hydrogen and oxygen production in the photocatalytic decomposition of water. As a photocatalytic material, g-C₃N₄ has broad application prospects [22,23,24]. As shown in Figure 1, g-C₃N₄ exhibits significant application potential in the degradation of both water and gas pollutants. In terms of water pollutant degradation, it demonstrates good treatment effects for both organic and inorganic contaminants. For organic pollutants, g-C₃N₄ can oxidatively decompose dye molecules such as methylene blue (MB), rhodamine B (RhB), and methyl orange (MO) in dye wastewater under light irradiation, while for antibiotic and pharmaceutical residues like tetracycline (TC), ciprofloxacin (CIP), and ibuprofen, it can mineralize them into CO₂ and H₂O, avoiding secondary pollution in traditional water treatment. For phenolic compounds and persistent organic pollutants (POPs) such as phenol, bisphenol A (BPA), and p-nitrophenol (4-NP), some intermediate products remain toxic, necessitating catalyst optimization to achieve complete mineralization. For inorganic pollutants, g-C₃N₄ can utilize photogenerated electrons to reduce heavy metal ions, such as by converting highly toxic Cr(VI) to Cr(III). In gas pollutant degradation, nitrogen oxides (NO_x) and sulfides (H₂S) such as NO, SO₂, and H₂S can be degraded through photocatalytic oxidation (e.g., NO→NO₃⁻) and surface adsorption activation to promote gas molecular decomposition. For volatile organic compounds (VOCs) in indoor and outdoor air purification, constructing porous g-C₃N₄ to increase gas adsorption sites and compositing with metal oxides to enhance oxidation capacity are effective improvement strategies.

2. Development History and Structures of Graphite Carbon Nitride

C₃N₄, a polymer derivative containing carbon and nitrogen (C₆N₉H₃), is considered to be one of the first chemicals synthesized, and it was reported in 1830 [25]. Jöns Jakob Berzelius heated the “mystery” mixture (potassium thiocyanate with a large amount of AlCl₃), and when the temperature was about 350 °C, a molten product was produced which could be successfully isolated with KOH [26,27]. Justus von Liebig then suggested naming it “melon”, which objectively made it one of the oldest artificial polymerized derivatives [28]. Franklin first proposed the concept of “carbon nitride” in 1922 when he obtained an amorphous carbon-nitrogen compound by pyrolyzing precursors such as Hg(CN)₂ and Hg(SCN)₂ [29]. In 1937, L. Pauling and J. H. Sturdivant first proposed that C₃N₄ is a cluster compound with coplanar triazine as the basic structural unit, and this view was proven through X-ray crystallography [30]. In the 1990s, C₃N₄ came back into the focus of developers after a long period of neglect, but back then, it was all about the breakthrough of the superhard material β-C₃N₄ [31,32]. In 1985, A.Y. Liu and M.L. Cohen [33] had theoretically predicted the structure of β-C₃N₄ by replacing Si with C atoms, based on the existing crystal structure of β-Si₃N₄. They deduced that the material had a hardness comparable to diamond. In 1996, D. M. Teter and R. Hemley recalculated the structure of C₃N₄ using first principles and proposed that C₃N₄ has five structures [34], namely α phase (P3₁c), β phase (P3), c phase (I$\overline{4}$3d), p phase (P$\overline{4}$2m), and g phase (P$\overline{6}$m2). Among these, the first four are superhard materials, while g-C₃N₄ is not. Compared with traditional metal catalysts, g-C₃N₄ has the advantages of high stability, acid and alkali resistance, and convenient modification, making it extremely suitable for inorganic metal catalyst design and synthesis [35,36]. In 2006, g-C₃N₄ began to be used in the field of heterogeneous catalysis. In 2009, the research group of Professor X.C. Wang [37] from Fuzhou University proved that g-C₃N₄ non-metallic semiconductors can achieve photocatalytic hydrogen production. In 2013, J. Xu et al. [38] prepared g-C₃N₄ nanosheets with a single atomic layer structure via the chemical exfoliation method, which exhibit stronger intrinsic and novel properties and thus shows great potential for application in many fields. The photocatalytic H₂ production activity and pollutant decomposition activity of monolayer g-C₃N₄ nanosheets are much higher than that of bulk g-C₃N₄, which has great potential for photocatalysis and photosynthesis [39,40].

g-C₃N₄ is in the form of powder with a yellowish surface. This material has a low density and a thermal decomposition temperature of 600 °C, and it is incompatible with water. Under light conditions, g-C₃N₄ nanomaterials are mixed with the water phase to form hydroxyl radicals, which have strong oxidizing properties and can degrade organic pollutants at the same time. In recent years, thanks to the efforts of researchers, the practical application of g-C₃N₄ in the field of photocatalysis has been expanded, and even mass production has been successfully achieved [41].

Among the five structural compositions of carbon nitride, g-C₃N₄ is the easiest material for achieving practical applications due to its band gap width and adjustable band structure. Its band gap is 2.7 eV, which is much smaller than those of other types. g-C₃N₄ has a two-dimensional layered structure similar to graphite stacks, and it was shown that two different polymer structures can be used as the basic structural units of each lamella, being connected by terminal N atoms to form an infinitely extended plane. Currently, there are two types of g-C₃N₄ nanomaterials, S-triazine units and 3-s-triazine units, in which the g-C₃N₄ nanomaterials based on 3-s-triazine units are much more stable, and this result has been confirmed by Jacob et al. [42], who investigated the calculation of the electronic band gap for both g-C₃N₄ and the triple-s-triazine-based g-C₃N₄ by using the multibody Green’s function method.

But it has been shown both experimentally and theoretically that melon, i.e., the polymorph with partial condensation of heptazine monomers, is the structure most frequently synthesized through thermal treatments [43,44]. B.V. Lotsch et al. [45], on the other hand, suggested that this model is inaccurate, and they argued that melons are planar layers and compounds consisting of carbon (IV) nitride cores (heptazine units) are regarded as “defective g-C₃N₄” materials. Since the preparation of g-C₃N₄ from hydrogen-containing precursors essentially yields materials with similar structural features and hydrogen contents, the graphite carbon nitride synthesis products should actually be polymers such as melons or related compounds. The discussion on g-C₃N₄ has a focus of many researchers. g-C₃N₄ is one of the representative nonmetallic semiconductors [46] which have received much attention this decade for its application in photocatalytic environmental remediation.

3. Graphitic Carbon Nitride Modification

Although g-C₃N₄ has the advantages of a narrow band gap, low cost, and non-toxicity, pure g-C₃N₄ still has the disadvantages of a narrow absorption spectral range and a high photogenerated electron-hole recombination rate [47], which impede its application in engineering practice. Compared with other means, the modification process of g-C₃N₄ via elemental doping greatly simplifies and streamlines the experimental steps and is more effective in broadening its light absorbance efficiency [48]. The formation of heterojunctions between g-C₃N₄ and one or more different substances does not only inhibit the complexation of photogenerated carriers; the synergistic effect between g-C₃N₄ and other components in the heterojunctions can give the photocatalysts some new characteristics [49]. Therefore, this section will focus on outlining three modifications of g-C₃N₄: doping and constructing heterojunctions, as shown in Figure 2.

Table 1. Comparison of modification methods.

Strategy		Principle	Advantages	Disadvantages
Doping	Metal doping	The doped metal ions are positively charged and have strong interactions with the negatively charged C and N, which can form coordination bonds. This changes the lattice structure of g-C3N4.	Alteration of the g-C3N4 lattice structure to reduce the band gap and expand the absorption range of visible light.	Limited resources and high prices.
	Non-metal doping	The non-metallic elements themselves have high electro-negativity and ionization energies and can form covalent bonds with other compounds during the reaction process.	Non-toxic and harmless, abundant sources, simple preparation process, good thermal and chemical stability, excellent light absorption performance, and adjustable band gap structure.	Performance is not yet up to the performance requirements of noble metal-based catalysts.
Heterojunction		The interface between two regions of different semiconductors with unequal band structures create interfacial band alignments.	The heterojunction structure of g-C3N4 was constructed to be able to effectively inhibit carrier composite.	The method for determining the type of heterojunction is complex and needs to be analyzed on an experimental basis.

shows a comparison of them.

3.1. Dopant Modification

3.1.1. Metal Doping

Metal ion-doped g-C₃N₄ inhibits electron-hole recombination, improves its optical and physicochemical properties, extends the absorption wavelength range to the visible range, and also increases its specific surface area [50]. N atoms with a lone pair of electrons in the g-C₃N₄ triazine structure can easily form coordination bonds with metal cations and improve the performance of photocatalysts by lowering the band gap of the material [51]. Doped metals include alkali metals, transition metals, and rare metals. Metal ions are mainly located in the vacant region of g-C₃N₄ because g-C₃N₄ has a larger atomic radius than the atomic radii of nitrogen and carbon in the structure [52,53]. Cavity doping occurs when metal ions enter the triangular pores between g-C₃N₄-linked triazine structures, leading to strong coordinative interactions between metal ions and negatively charged nitrogen atoms. Metal ion-doped g-C₃N₄ nanocomposites can be used for the rapid degradation of organic pollutants in wastewater [54]. Figure 3 presents the optimized structures of intrinsic g-C₃N₄ and doped alkali metal g-C₃N₄, with some metal ions as examples (Li, Na, K, and Rb).

H. Zhang et al. [55] prepared metal-doped g-C₃N₄ through the solvothermal method, and S. Ye et al. [56] successfully synthesized photocatalytic Fe₂O₃/g-C₃N₄ composites. They demonstrated that K-doped g-C₃N₄, Rb-doped g-C₃N₄, and Fe₂O₃/g-C₃N₄ have lower forbidden bandwidths and strong light-absorbing properties compared with pure g-C₃N₄. Y.Z. Zhang et al. [57] prepared sodium-doped graphite nitride. It was found that sodium doping significantly improved the photocatalytic ability of H₂, which was 9.2 times higher than that of graphite nitride. The experiments of P.Q. Deng et al. [58] realized doped Ni²⁺ in the cavities of the g-C₃N₄ framework. The increase in Ni content affects the possible gap sites of the doped Ni and changes the positions of its valence band (VB) and conduction band (CB). The Cu⁺- g-C₃N₄ complex prepared by J.Q. Ma et al. [59] has high Fenton activity, and its mechanism is shown in Figure 4. Cu²⁺ is reduced to Cu⁺ by a hydroxyl radical mechanism and finally embedded in the lamellar g-C₃N₄ via coordination with pyridine N. The bonded Cu⁺ participates in the generation of ¹O₂ and HO·, and then, due to its structural stability, it can be rapidly reduced to the cuprous state and thus has a high catalytic ability in Fenton-like reactions.

L.W. Ruan [60] computationally investigated the optical properties of lithium-doped g-C₃N₄ monolayers. There is no LUMO or HOMO in lithium atoms. Therefore, the active sites will not change in the doped systems compared to the pristine g-C₃N₄. However, by calculating LUMO and HOMO, it was found that doped Li atoms can enhance the hybridization of LUMO and HOMO values, indicating that the light-generated electron and hole pairs can be effectively separated. Ta et al. [61] synthesized metal-doped ZnO/g-C₃N₄ composites with a sponge-like porous structure via a facile one-pot pyrolysis method, among which the magnesium (Mg)-doped composite exhibited the highest photocatalytic performance. A study by Li et al. [62] demonstrated that Ag-doped g-C₃N₄-TiO₂ exhibits superior photocatalytic degradation performance due to its narrower band gap compared with pure g-C₃N₄-TiO₂, as well as faster separation and transfer of photogenerated charge carriers. Ji et al. [63] experimentally derived the photo-Fenton reaction mechanism of Fe-doped g-C₃N₄ as follows. Firstly, Fe-g-C₃N₄ generates photo-induced electron-hole pairs under light irradiation. The photo-induced holes of Fe-g-C₃N₄ can oxidize H₂O/OH⁻ to generate ·OH. Fe doping remarkably improved the redox properties. Owing to the interfacial charge transfer effect, the Fe³⁺ is reduced into Fe²⁺ by photo-excited electrons, followed by the reaction of Fe²⁺ with H₂O₂, which forms the ·OH radical and effectively oxidizes RhB. Doping Fe decreases the photoelectron-hole complex efficiency, and the transfer of photoelectrons to Fe improves the photocatalytic activity of g-C₃N₄. As shown in

Table 2. Examples of the performance of metal-doped g-C₃N₄.

	Doped Metal	Pollutants	Degradation Efficiency	References
Alkali metal-doped g-C3N4	Pure g-C3N4	CO2	3.6 μmol g−1	[55]
	Li		5.6 μmol g−1
	Na		7.4 μmol g−1
	K		9.8 μmol g−1
	Rb		12.1 μmol g−1
Ni-doping of g-C3N4	Pure g-C3N4	MO	47.8 (140 min)	[58]
	Ni		97.3 (90 min)
Cu-graphitic carbon nitride	Cu	RhB	99.2 (60 min)	[59]
ZnO/g-C3N4	ZnO	MB	91 (20 min)	[61]
Fe-g-C3N4	Pure g-C3N4	RhB	69 (45 min)	[63]
	5% Fe		92.9 (45 min)
	10% Fe		95.5 (45 min)
	15% Fe		90.5 (45 min)

, metal doping modification can significantly enhance the photocatalytic degradation performance of g-C₃N₄, and multiple metal-doped systems exhibit higher degradation rates in pollutant degradation.

Modification of g-C₃N₄ using metal doping can enhance the catalyst degradation rate. However, it has been found that neither mono nor bimetallic metals can change the g-C₃N₄ structure, and relying only on gap trapping of electrons and inhibition of some of the photo complex reactions cannot completely and drastically enhance the catalytic efficiency. Therefore, an attempt can be made to fully merge the respective advantages of metal and g-C₃N₄ into one to explore the doping and application studies of multi-metals.

3.1.2. Non-Metal Doping

Non-metallic atoms have high ionization energies and different electronegativities. When introduced, non-metallic atoms accept electrons from other substances to form covalent bonds, thus changing the electron distribution around the dopant sites [64,65]. In general, highly electronegative dopant atoms promote the migration of electrons from neighboring C atoms to the dopant sites; the original chemical inertia is destroyed, and the C atoms will have induced polarization to form new reaction sites. When the electronegativity of the dopant atoms is low, it can increase the asymmetric spin density of the neighboring C atoms, thus increasing the electron-donating ability of g-C₃N₄ [66,67]. Thus, non-metallic doping (O, N, P, C, S, B, etc.) enables reactions with other species to form covalent bonds, thereby altering electron distribution around dopant sites via charge polarization [68]. The photogenerated electron mobility of the material is increased, and the electron-hole complexation rate is decreased, which effectively adjusts the electronic and molecular structure of g-C₃N₄, expands its light responsive range [69], and ultimately improves the photocatalytic properties.

In M. Arumugamm’s research [70], g-C₃N₄ doped with nonmetals (B, O, P, and S) was prepared, and various physicochemical analyses were performed, which confirmed that the doped nonmetallic elements could improve the performance of photocatalysts by improving charge separation. Its efficiency in reducing carbon dioxide and water to CH₄ was 1~2 times that of the original. The S-g-C₃N₄ prepared by G. Liu et al. [71] was also effective in catalyzed phenol degradation at a light source with a wavelength > 400 nm. J. Marie and D.W. Dikdim et al. [72] investigated activated carbon/g-C₃N₄ and found that the addition of activated carbon effectively inhibited the electron-hole pair complexation. A non-metallic photocatalyst was prepared by J.J. Zhang et al. [73]. The experimental results showed that PI-g-C₃N₄ (perylene imide-modified g-C₃N₄) could realize the activation of persulfate to degrade BPA under visible light. The above studies have shown that doping of most of the nonmetals (e.g., S, P, O, C, I, and B) can effectively narrow the band gap of g-C₃N₄ and thus broaden the optical absorption spectral range of the material, making it easier to be excited by visible light [74]. However, when the band gap of the material is too small, the redox potential of the valence or conductance band of the catalytic material decreases [75], which is also detrimental to the photocatalytic property of the material. Therefore, the ideal outcome of g-C₃N₄ band gap regulation should be that the regulated material has an appropriate band gap (~2.0 eV) [76] which can fully absorb sunlight while having appropriate conduction and valence band positions such that the photogenerated carrier has a sufficient redox capacity and the synergy of different properties can produce better effects. Figure 5 shows the band gap variation in g-C₃N₄ and the energy band structure of g-C₃N₄ with different nonmetal doping.

Non-metal doping strategies are generally favored due to their low cost and ease of preparation. Although some results have been achieved in the research of non-metallic doping modification of g-C₃N₄, the current attention on the non-metallic doping modification strategy focuses on single non-metallic doping, and there is less exploration of multi-non-metallic co-doping modification of g-C₃N₄. Studies on the synergistic promotion of g-C₃N₄ photocatalysis via non-metallic doping and other modification methods are even less common.

3.2. Constructing Heterojunction Structures

The heterojunction structure of g-C₃N₄ was constructed to effectively inhibit the carrier composite. A heterojunction is generally defined as the interface between two regions of different semiconductors with unequal band structures and creating interfacial band alignments [78,79,80,81]. The formation of g-C₃N_4-based heterojunctions often includes the following categories: (1) Type I heterojunction, (2) Type II heterojunction, (3) p-n junction, (4) Schottky junction, and (5) Z-scheme heterojunction. These categories’ energy band structures are shown below in Figure 6 [82]. Based on the understanding of a traditional heterojunction, a new concept of S-type heterojunction has been proposed. The concept was first proposed by Prof. Jianguo Yu’s group in 2019, and its full name is the step-scheme heterojunction, shown in Figure 7 [83,84]. Table 3 presents examples of constructing different heterojunctions and their applications, providing a practical reference for understanding heterojunction - based design strategies.

Table 3. Examples of constructing different heterojunctions and their applications.

Classification	Typical Example		Application
Type I heterojunction	ZnIn2S4/ultrathin-g-C3N4 [85]		Photocatalytic H2 production reaction under visible light
	CdIn2S4/g-C3N4 [86]		Photodegradation of RB19 under visible-light irradiation
Type II heterojunction	g-C3N4/g-C3N4 [87] (with different raw materials)		Enhancement of photocatalytic H2 production and CO2 reduction activity
	g-C3N4@Cs2AgBiBr6 (CABB) [88]		CO2 photoreduction process
p-n junction	Co3O4/n-type g-C3N4 [89]		CO2 reduction
	P-group intercalated g-C3N4 (NP–CN)/Bi2WO6 (BWO), NPB [90]		Photodegradation of p-nitrophenol into harmless products
Schottky junction	Ag-decorated P-doped g-C3N4 nanosheets (Ag-(P/CNNS)) [91]		Water splitting and degradation of rhodamine B (RhB)
	g-C3N4/Ti3C2 [92]		Efficient visible-light photocatalytic hydrogen
Z-scheme heterojunction	Direct Z-scheme	BiOI/g-C3N4 [93]	Photocatalytic degradation of phenol
	Indirect Z-scheme	TiO2/BC/g-C3N4 [94]	Photocatalytic reduction of Cr(VI) in aqueous solution
S-scheme heterojunction	C–O-bridged CeO2/g-C3N4 (cCN) [95]		Photofixation of N2, H2 energy generation, and methyl orange photodegradation

Table 3. Examples of constructing different heterojunctions and their applications.

Classification	Typical Example		Application
Type I heterojunction	ZnIn2S4/ultrathin-g-C3N4 [85]		Photocatalytic H2 production reaction under visible light
	CdIn2S4/g-C3N4 [86]		Photodegradation of RB19 under visible-light irradiation
Type II heterojunction	g-C3N4/g-C3N4 [87] (with different raw materials)		Enhancement of photocatalytic H2 production and CO2 reduction activity
	g-C3N4@Cs2AgBiBr6 (CABB) [88]		CO2 photoreduction process
p-n junction	Co3O4/n-type g-C3N4 [89]		CO2 reduction
	P-group intercalated g-C3N4 (NP–CN)/Bi2WO6 (BWO), NPB [90]		Photodegradation of p-nitrophenol into harmless products
Schottky junction	Ag-decorated P-doped g-C3N4 nanosheets (Ag-(P/CNNS)) [91]		Water splitting and degradation of rhodamine B (RhB)
	g-C3N4/Ti3C2 [92]		Efficient visible-light photocatalytic hydrogen
Z-scheme heterojunction	Direct Z-scheme	BiOI/g-C3N4 [93]	Photocatalytic degradation of phenol
	Indirect Z-scheme	TiO2/BC/g-C3N4 [94]	Photocatalytic reduction of Cr(VI) in aqueous solution
S-scheme heterojunction	C–O-bridged CeO2/g-C3N4 (cCN) [95]		Photofixation of N2, H2 energy generation, and methyl orange photodegradation

3.2.1. Type I Heterojunction

In the Type I heterojunction structure (Figure 6a), the energy bands of two semiconductors exhibit a straddling alignment. Electrons migrate to the semiconductor with the lower conduction band edge, while holes migrate to the semiconductor with the higher valence band edge. As an example, K.F. Bai [96] investigated the properties of GaS/g-C₃N₄ van der Waals (vdWs) heterostructures using density functional theory. The formation energies of GaS/g-C₃N₄ vdWs heterostructures with different structural configurations were compared, and it was found that the stabilized GaS/g-C₃N₄ vdWs heterostructures have an I-type band alignment. It was found that the H₂ yield of the ternary catalyst NiS/ZnIn₂S₄/ultrathin-g-C₃N₄ (NiS/ZIS/UCN) was more than five times higher than that of ZnIn₂S₄ and that NiS/ZIS/UCN constructed channels with dual high-speed charge transfer to promote photocatalytic H₂ generation [95]. As shown in Figure 8, the photocatalytic mechanism, the formation of a Type I heterojunction between ZIS and UCN, and the formation of high-speed charge transfer nano-channels greatly accelerated charge transfer and significantly improved the separation efficiency.

3.2.2. Type II Heterojunction

g-C₃N₄ and another semiconductor construct a conventional g-C₃N₄-based Type II heterojunction structure (Figure 6b), in which both the valence and conduction bands of g-C₃N₄ are above or below the position of the other half of the conductor [97]. The potential difference between the two semiconductors can drive the migration of photogenerated electrons and holes and promote effective space charge redistribution at the interface, which greatly facilitates charge separation and thus greatly improves photocatalytic performance. F. Dong et al. [98] constructed a g-C₃N₄/g-C₃N₄ heterojunction composite with different raw materials, a typical type II heterojunction composite structure, and experimentally demonstrated that its visible-light properties were significantly enhanced.

3.2.3. The p-n Junction

The construction of p-n junction structures (Figure 6c) is an effective strategy for the development of noble metal-free hybridized photocatalysts. In the past decades, many novel p-n heterojunctions have been proposed and have shown excellent photocatalytic performance in the oxidative degradation of organic pollutants [99]. According to the theory of semiconductor physics, when p-type and n-type semiconductors with different Fermi energy levels are in contact, an internal electric field is generated at the interface of the p-n junction formed by the two, which has been widely demonstrated to be favorable for the separation of photogenerated carriers [100]. For example, 2D/2D BiOBr/g-C₃N₄ heterojunctions prepared using self-assembly methods were characterized by p-n junctions [101]. Compared with g-C₃N₄ alone, the kinetic constants for the degradation of bisphenol A and norfloxacin through the optimized BiOBr/g-C₃N₄ were improved by factors of 14.74 and 4.01, respectively. Compared with BiOBr alone, the kinetic constants for the degradation of bisphenol A and norfloxacin by the optimized BiOBr/g-C₃N₄ were improved by factors of 2.20 and 1.36, respectively. The efficient spatial separation and transport of photoexcited electron-hole pairs are important reasons for the good photocatalytic activity and stability of BiOBr/g-C₃N₄. Hydrothermal pretreatment and the calcination process developed a heterojunction of Co₃O₄ nanoparticles supported by g-C₃N₄ microtubules [89]. The p-type Co₃O₄ and n-type g-C₃N₄ constructed a p-n junction. The generated internal electric field accelerates the charge transfer. The charge transfer path on the TCN/Co₃O₄ NP is shown in Figure 9.

3.2.4. Schottky Junction

The Schottky junction catalyst is a heterogeneous structure formed by metal and semiconductors (Figure 6d). Its more obvious characteristic is its rectification properties, which can be adjusted according to the current pattern. Schottky conducted a more systematic study of metal-semiconductor contacts in 1874. He pointed out the existence of a potential barrier on the semiconductor side close to the interface and analyzed it [102,103,104]. The Schottky barrier is able to control the migration of free charge, effectively separating the photogenerated carriers and favoring the participation of holes in the reaction, thus improving the photocatalytic activity. Taking g-C₃N₄/Ti₃C₂ as an example, J. Li et al. [92] constructed a kind of unique 2D/3D structure of g-C₃N₄/Ti₃C₂, and its photocatalytic hydrogen precipitation mechanism is shown in Figure 10. Due to the presence of a Schottky junction and a wide contact interface in the composite, the excited electrons in the conduction band of g-C₃N₄ can easily migrate to Ti₃C₂ under the visible light condition under the action of an internal electric field, while the corresponding holes remain in the valence band of g-C₃N₄. Subsequently, the higher conductivity of Ti₃C₂ accelerates the consumption of the trapped electrons.

3.2.5. Z-Scheme Heterojunction

The Z-scheme heterojunction was proposed based on artificial photosynthesis that mimics natural plant photosynthesis. The Z-type system comprises an oxidized photocatalyst and a reduced photocatalyst [105]. When exposed to visible light, electrons on the oxidized photocatalyst CB are transferred and then recombined with holes on the reduced photocatalyst VB, leaving more electrons in the reduced photocatalyst and more holes in the oxidized photocatalyst to participate in the reduction and oxidation reactions [106]. It has the advantages of effectively separating electron-hole pairs, reducing the chance of compounding, retaining more active sites, and expanding the photo-response range. The photocatalytic activity is improved relative to the single photocatalyst of traditional heterojunction photocatalysts.

The Z-scheme heterojunction is usually divided into direct Z-scheme and indirect Z-scheme heterojunctions, as shown in Figure 6e,f. The fundamental differences between direct Z-scheme and indirect Z-scheme heterojunctions primarily lie in the charge transfer pathways, semiconductor configurations, and the separation efficiency of photogenerated carriers. In the direct Z-scheme, two semiconductor materials achieve direct electron-hole pair recombination through intimate interfacial contact (without any additional mediators), forming an internal electric field-driven Z-scheme pathway. This mechanism offers high charge transfer efficiency (minimizing energy loss caused by mediators) and relies on the intrinsic band alignment of the semiconductors. Moreover, the direct Z-scheme features a simple structure and lower fabrication complexity. In contrast, the indirect Z-scheme requires electron mediators (e.g., noble metal nanoparticles or carbon materials) acting as “electron bridges” to facilitate indirect charge recombination. This process depends on the redox cycling of the mediators and often necessitates additional electrolyte components. While the mediators can enhance charge transfer efficiency, they may also introduce additional complexity to the system.

Direct Z-scheme heterojunctions are formed by direct contact between two semiconductors, where the two semiconductor materials are in close contact to form stronger interactions, which helps reduce the resistance to carrier transfer [107]. For example, BiOI quantum dots and g-C₃N₄ were fabricated at room temperature via the precursor in situ transformation method. By reducing the size of BiOI to the quantum scale, a complete intimate contact between BiOI and g-C₃N₄ was achieved, which in turn improved the catalytic activity [93]. Compared with the direct Z-type heterojunction of TiO₂/g-C₃N₄, the indirect Z-scheme heterojunction of TiO₂/BC/g-C₃N₄ uses BC as an electron mediator, which facilitates the separation of electron-hole pairs and reduces the loss in the electron transfer process. During the experiments, the efficient removal of Cr(VI) was realized by enhancing the interfacial mass transfer and accelerating the electron transfer in TiO₂/BC/g-C₃N₄. The presence of a fully fixed Z-scheme connection structure can effectively solve the charge separation problem of the traditional II connection structure. Such a composite structure allows the use of narrow band gap semiconductor pairs without a loss in photoelectron and hole-strong redox capacity [108,109,110], preventing self-corrosion of certain unstable semiconductors by transferring the associated reduced electron/oxide holes. D.M. Chen et al. [111] prepared a AgBr/g-C₃N₄ Z photocatalyst for Ag/AgBr/g-C₃N₄ reduction, in which Ag nanoparticles acted as electron mediators to facilitate electron transfer from AgBr to g-C₃N₄. S.S. Lu et al. [112] prepared Bi/γ-Bi₂O₃/EtCN (O-doped g-C₃N₄), and an all-solid Z-type heterojunction was formed between γ-Bi₂O₃ and EtCN. The enhancement of the composite activity is mainly due to the modification of Bi after the formation of Z-type heterojunction, which broadens the light absorption range of the catalyst and promotes the effective separation of electron-hole pairs.

The composite materials with Z-scheme heterojunction, such as g-C₃N₄/Bi₂WO₃ [113], g-C₃N₄/Ag₃PO₄ [114], and g-C₃N₄/Ag₃VO₄ [115], usually have suitable electronic structures and good photochemical stability. The composite structure formed by mixing and calcining g-C₃N₄ with semiconductors has a good enhancement effect on photocatalytic performance [116,117]. The matched electronic structure of the two semiconductors and the combination of g-C₃N₄ with BiOX at low temperatures can improve the visible-light photocatalytic performance [118]. Under visible light, both types of semiconductors produce photogenerated charge carriers that can quickly separate photogenerated electrons from holes. Photogenerated charge carriers in a novel ternary heterojunction structure of g-C₃N₄/Bi₂MoO₆/Bi are transferred via an indirect Z-order structure. The synergistic effect of the Z-type heterojunction structure and the SPR effect of the doped Bi resulted in stronger light absorption and lower photogenerated charge carrier binding of the material (Figure 11) [119]. Z.L. Huang [120] fabricated Z-scheme two-dimensional/one-dimensional g-C₃N₄/MoO^3-_x composites, which exhibited strong visible and near-infrared absorption.

3.2.6. S-Scheme Heterojunction

In the field of photocatalysis, direct Z-scheme and S-scheme heterojunctions exhibit significant differences in their electron transfer and energy band mechanisms. The direct Z-scheme heterojunction forms a “Z”-shaped pathway relying on the band potential difference between semiconductors, requiring no mediator and featuring a simple structure, though it has high requirements for interfacial contact and is suitable for scenarios with clear band matching, such as CO₂ reduction. In contrast, the S-scheme heterojunction is more distinctive; it drives the selective recombination of electrons and holes through a built-in electric field, preferentially consuming low-energy carriers to form an “S”-shaped transfer pathway. By relying on regulation of the Fermi levels of n-type and p-type semiconductors, it not only significantly improves charge separation efficiency and broadens the photoresponse range but also retains strong redox capabilities while efficiently adapting to complex catalytic needs such as water splitting and deep mineralization of pollutants. Despite higher requirements for preparation processes, the S-scheme heterojunction is becoming a new hotspot in the research and application of photocatalysis by virtue of its unique carrier regulation mechanism.

As shown in Figure 8, this heterojunction is mainly constructed from a reduced-type semiconductor photocatalyst (RP) with smaller work functions and higher Fermi energy levels and an oxidized-type semiconductor photocatalyst (OP) with larger work functions and lower Fermi energy levels through a staggered approach. When observing the structural schematic, it is similar to the Type Ⅱ heterojunction. However, in a typical Type Ⅱ heterojunction, photogenerated electrons and holes are accumulated in the conductive and valence bands of the RP and OP, respectively, leading to a weakened redox capacity. In contrast, in S-scheme heterojunctions, the effective electrons and holes are preserved, and the meaningless photogenerated carriers are recombined. This is named for the electron transfer schematic, which resembles a step. G. Huang et al. [121] synthesized stable ZnO/g-C₃N₄ S-scheme heterojunctions through simple solvent evaporation and high-temperature thermal polymerization reactions. Under the same experimental conditions, the photocatalytic degradation of pyridine via pure g-C₃N₄ and a ZnO/g-C₃N₄ heterojunction was 65.8% and 98.9%, respectively. A. Lsulmi et al. prepared an S-scheme AgI/g-C₃N₄ heterojunction with 96% photocatalytic degradation of RhB dye. The AgI nanoparticles degraded 40% of the RhB under the same conditions, and the pristine g-C₃N₄ was even less efficient [122].

Generally, in combination with a suitable semiconductor, a lower energy level semiconductor can pull the excited electrons of the graphite carbon nitride into its own conduction band, which can use this energy not only to excite the photo drive of the semiconductor but also reduce the electron-hole recombination rate of graphite carbon nitride and improve the absorption and utilization of visible light, thereby enhancing the photocatalytic ability. For Type I heterojunctions, electrons and holes in two semiconductors migrate in the same direction. In contrast, Type II heterojunctions allow electrons and holes to migrate to different semiconductors, resulting in space charge separation [123]. Combining p-type semiconductors with n-type semiconductors leads to the redistribution of electrons within the catalytic system, which reduces the compounding rate of photogenerated charges and enhances the photocatalytic performance of the catalysts. A Schottky junction is a close contact between the interface of a metal and a semiconductor, where the semiconductor bends upward for the sake of the Fermi energy levels between them, and the electrons in the conduction band move toward the metal. The Schottky junction can effectively separate the photogenerated charges, extend the lifetime of the photogenerated charges, and improve the photocatalytic performance of the photocatalytic material [124]. Conventional heterojunction has the advantages of expanding the range of photo-response and promoting carrier separation, but it has the problem of an insufficient oxidation-reduction capacity. The Z-scheme heterojunction has many advantages over it, including effective separation of electron-hole pairs, reduction of compounding chances, retention of strong oxidation-reduction active sites, broadening of the photo-response range, and an improvement in photocatalytic activity. In conventional Z-scheme heterojunctions, charge transfer between the two semiconductors is achieved using redox electron mediator pairs, and these ions can only achieve sufficient migration rates in a solution. This limits the system to the reaction of the solution phase. S-scheme heterojunctions achieve spatial separation of semiconductor photogenerated electrons and holes with a guaranteed strong redox capacity through three factors: a built-in electric field, energy band bending, and electrostatic interactions. Some applications of S-scheme heterojunctions have been described in this paper, including photocatalytic hydrogen production, carbon dioxide reduction, and pollutant degradation [125,126]. However, as a new heterojunction system, S-scheme heterojunctions have been studied less.

4. Application of Graphitic Carbon Nitride in Photocatalytic and Photoelectrocatalytic Degradation

Since the beginning of this century, photocatalytic research results have emerged and shown broad development [127,128]. Scientists have researched and developed various photocatalysts [129,130,131]. In the last 12 years, two-dimensional g-C₃N₄ has attracted much attention for applications in photocatalytic environmental remediation and solar energy conversion to fuel [132]. When irradiated with light, the g-C₃N₄ material mixes with the aqueous solution to form hydroxyl radicals, which have a strong oxidizing ability and decompose organic pollutants, realizing a reduction in the number of carbon atoms in the molecules of organic compounds, reducing the molecular weight, and even decomposing them into carbon dioxide and water.

4.1. Methodological Principles of Photocatalysis and Photoelectrocatalysis

Photocatalysis is an efficient oxidation technology that has emerged in recent years. In short, photocatalytic oxidation technology is an advanced redox technology that uses light energy. The reaction equation is simply described as follows:

Semiconductor (SC) + hv → h⁺ + e⁻

(1)

Different types of semiconductors have different optical properties and colors, and these properties depend on the size of the band gap. The band gap is usually between 1.5 eV and 3.0 eV. Different band gap energies also reveal the difference in the absorption wavelengths, which directly leads to a difference in photocatalytic capacity [133,134].

PEC is an effective strategy for improving oxidation efficiency by utilizing photoelectric collaborative processes instead of traditional photocatalysis and electrocatalysis. As a new advanced oxidation technology, it combines the advantages of photocatalysis and electrocatalysis. By applying low auxiliary pressure to the electrode, photogenerated electrons are promoted, which are captured by the anode, transferred through the external circuit, and quickly reacted at the cathode, improving the key problem of easy recombination of e⁻-h⁺ in photocatalysis [135]. Studies have confirmed that PEC technology exhibits far superior catalytic performance to single catalytic methods under identical conditions [136,137]. As shown in Figure 12, the PEC system consists of a photocatalytic anode and an electrode cathode. Electrode cathodes can be composed of various photocatalysts. Wires then connect the cathode and anode to form a closed-loop system. The core of the system is a photocatalytic semiconductor material. The photocatalytic anode generates electrons and holes when exposed to light. The electrons generated by light are transferred to the cathode through an external circuit. Under anaerobic conditions, they can react with hydrogen ions transferred from the anode to form H₂. In the presence of oxygen, oxygen can be reduced to water. Therefore, the photo hole on the photo anode surface will oxidize and decompose the pollutants in the anode chamber, realizing the closed circuit and recovering the electric energy while eliminating the pollutants.

4.2. Relationship Between g-C₃N₄ Modification and Photo Response

The photoanode has always been a hot research topic as a key element of the photocatalytic and photoelectrocatalytic reaction system. Researchers are paying more attention to developing and improving photocatalysts for photoanodes [138]. g-C₃N₄ is attracting attention as a CN-based material. However, the photocatalytic stability of g-C₃N₄ is greatly limited due to its short life and the weak redox ability of the photo-generating carriers [139]. The g-C₃N₄ composite structure can transfer photogenerated electrons or holes to semiconductors, metals, oxides, and other materials [140,141], which enhances the photoelectrocatalytic reaction and efficiently prevents the generation of composite photogenerated carriers. g-C₃N₄-modified material catalysts have four advantages:

• Being combined with a co-catalyst

An appropriate co-catalyst combined with g-C₃N₄ can increase the active site, reduce the overpotential, and improve the photocatalytic performance [142].

2.

Enhanced light absorption

g-C₃N₄ can act as a photocatalyst under visible light, and its modified materials can further enhance its light absorption. For example, assisted by nitrogen defects, g-C₃N₄ undergoes a pressure-driven structural transition (PIA) from a graphitic to an amorphous phase. Under high pressures, the narrowing of the band gap enhances light absorption in the visible range [143].

3.

Improvement in transmission efficiency

The separation of the photogenerated carrier can effectively separate the electron hole to improve transport efficiency.

4.

Good stability

Stable, non-toxic, and non-polluting g-C₃N₄ has good thermal and chemical stability. g-C₃N₄ and its modified materials are capable of stable performance at high temperatures. It is only at over 600 °C that the thermal stability begins to decrease.

Elemental doping and composite modification can significantly improve the visible-light reactivity of g-C₃N₄ by disrupting the intra-layer hydrogen bonding and enhancing the visible-light photocatalytic performance [144]. For example, the introduction of the S element replaces the edge N atom of sp² hybridization, breaks the intra-layer hydrogen bond, and greatly improves the carbon atom density of g-C₃N₄, which not only promotes the n→π* electronic transition but also broadens the visible-light response range and lowers the potential barrier, which is conducive to photogenerated charge-carrier migration and separation and photogenerated electron trapping [145], with absorbance in the visible region for lithium-doped and silver-doped g-C₃N₄-modified composites, inhibition of photogenerated electron-hole pair recombination, and improved generation of reactive oxide species [146]. Constructing a heterojunction expands the range of light absorption and realizes different effects depending on the material, expanding the application of g-C₃N₄ materials. For example, by constructing heterojunctions using bismuth-based (BiOX, X = F, Cl, Br, and I) halogen oxides, it is possible to convert UV-active BiOF into visible-active BiOF, resulting in improved efficiency of the applied bias photon current and significant chemical stability over long periods [147].

4.3. Photocatalytic Performance of g-C₃N₄ and Its Modified Material Under Different Light Sources

g-C₃N₄ and its modified materials have shown great potential in the application of photocatalysis degradation under different light sources [148]. As we all know, the full spectrum contains the spectral curves of ultraviolet rays (UV), visible light (Vis), and near-infrared light (NIR). Compared with traditional TiO₂ photocatalysts, g-C₃N₄ has a wider absorption spectrum. It has the natural advantage of being able to act as a photocatalyst under normal visible light without the need for UV light. However, the application of g-C₃N₄ under more light source conditions is still worth investigating.

4.3.1. UV Photocatalytic Degradation

UV is a general term for radiation in the electromagnetic spectrum with wavelengths from 0.01 to 0.40 μm (between the violet end of visible light and X-rays). UV light has a high energy level and can trigger reactions in a wide range of molecules or ions. Upon UV irradiation, atmospheric oxygen molecules decompose to generate free oxygen radicals. As free oxygen radicals carry unpaired electrons (in an unstable state), they readily combine with oxygen molecules to form ozone.Ozone has a strong oxidizing effect on exhaust gas molecules and is, therefore, highly effective in treating volatile organic gases (VOCs) [149]. Some researchers have also found that the UV 3D electrode’s electro-Fenton removal rate is consistently higher for dye wastewater than the 3D electrode’s electro-Fenton removal rate. UV-catalyzed H₂O₂ degradation of dyes is more effective under acidic conditions. The UV/H₂O₂ system promotes the oxidation of organic matter through hydrogen extraction, electron transfer, and radical addition [150,151]. Under UV irradiation, the photocatalytic degradation of RhB by TiO₂/ZnO-g-C₃N₄ prepared by H.X. Jing [152] was superior to that of other light sources up to 99.6%. J.G. Kim et al. [153] prepared iron-modified g-C₃N₄ (FG materials) for the removal of arsenate (As(V)) and arsenite (As(III)) from water. The cerium fluoride (CeF₃) semiconductor CeF₃ has upconversion properties for converting Vis and NIR light into UV light. The CeF₃/g-C₃N₄ heterojunction is capable of photocatalytic degradation of dibenzothiophene (DBT) [154]. The synergistic action of CeF₃ and g-C₃N₄ achieves up to 84.2% desulfurization of DBT in 3 h. This composite has potential application in deep desulfurization.

4.3.2. Visible-Light Photocatalytic Degradation

Visible-light photocatalysis is a robust emerging area for organic transformation and is of great value [155]. Since the pioneering study of photocatalysis of g-C₃N₄ under visible-light irradiation by X.C. Wang et al. in 2009 [156], g-C₃N₄ has been recognized as a semiconductor with a narrow band gap (~2.7 eV) and visible light activity (~460 nm) [157,158]. The photocatalytic activity of Bi₂MoO₆/g-C₃N₄ composites under visible light significantly improved after introducing Bi₂MoO₆. After a certain time of visible-light irradiation, the degradation efficiency of organic pollutants was 93.88%, which was 3.9 times higher than that of g-C₃N₄ [159]. M. Sheydaei et al. [160] prepared visible-light active g-C₃N₄/Ce-ZnO/Ti nanocomposites and found that the visible-light photoelectrocatalytic process was able to degrade cefixime in aqueous solution with the highest efficiency. The WO₃/g-C₃N₄ prepared by D.M. Li et al. had a porous and wrinkled structure. Its light absorption coefficient was larger than that of g-C₃N₄, and its band gap was 0.3 eV, which is smaller than that of g-C₃N₄. Therefore, the recombination rate of photogenerated electron-hole pairs of WO₃/g-C₃N₄ was greatly reduced, and the visible light-induced catalytic activity was enhanced. The removal rate of RhB can reach 97.3% [161].

4.3.3. Near-Infrared Photocatalytic Degradation

Infrared light makes up more than 50% of the Sun’s spectrum. At present, the application research of photocatalysts is mainly focused on the ultraviolet and visible regions, while reports on the near-infrared region are relatively few in number. For efficient solar fuel conversion, researchers need to scrutinize the extension of light absorbance into the near-infrared (NIR) region. Theoretically, NIR-responsive photocatalysts for diffuse solar energy can be utilized to alleviate energy crises and environmental pollution [162].

A series of studies has shown that extending the photo response into the infrared optical region is possible by preparing heterojunctions with noble metals. The p-n heterojunction constructed using g-C₃N₄ and Ag₂O not only extends the absorption spectrum into the near-infrared region but also inhibits the recombination of photogenerated carriers, thus further improving the catalytic degradation performance of Ag₂O in the near infrared [163]. F. Wu et al. found that p doping can extend the response spectrum of g-C₃N₄ from 450 nm to 1000 nm, which produces a significant near-infrared response during PEC [164]. Furthermore, the Ag₂O/Ag₂S₂O₇ heterojunction catalyst with a full spectral response, prepared by H. Li et al. [165], can effectively improve the activity and stability of Ag₂O in the UV, visible, and near-infrared regions. UCNPs can convert near-infrared photons into ultraviolet (UV) and visible light, enabling the utilization of near-infrared light. S.Y. Lee combined g-C₃N₄ with upconversion nanoparticles (UCNPs). The use of an NIR light source enhanced the photocatalytic activity of UCNP and g-C₃N₄ nanocomposites for efficient degradation of methyl orange (Figure 13) [166].

4.4. Application of g-C₃N₄ and Its Modified Materials in Photoelectrocatalysis Under Different Light Sources

Photoelectrocatalysis is an effective strategy for improving oxidation efficiency. In contrast to photocatalysis, PEC can promote electrochemical oxidation of pollutants at the anode and reduction reactions at the cathode by generating a bias voltage. In PEC technology, the fixation of g-C₃N₄ as a matrix material can minimize potential secondary contamination through material leaching and avoid additional post-processing steps. At present, in the photoelectrocatalytic system, the most commonly used light sources are xenon lamps (simulating sunlight, emitting light as a continuous spectrum from 200 nm to 2000 nm, and having energy distribution) and UV-visible light. Meanwhile, g-C₃N₄ can activate molecular oxygen and generate superoxide radicals for the photocatalytic conversion of organic functional groups and photocatalytic degradation of organic pollutants, which is more suitable for the treatment of indoor air pollution and the degradation of organic matter rather than TiO₂ [167].

The PEC system prepared by impregnating g-C₃N₄ film onto ITO glass and irradiating it with a 500 W xenon lamp with a bias voltage of 2.5 V was able to increase the mineralization rate of phenol from 37% to 89% in a much shorter period of time. W. Zhen et al. [168] prepared a g-C₃N₄/TiO₂ film photoanode, and the mineralization of phenol was only increased to 100% after 1.5 h. A g-C₃N₄/TiO₂-NTs electrode was synthesized by Y. Zhang et al. [169], and the g-C₃N₄/TiO₂-NTs electrodes showed higher TOC mineralization of phenol and benzyl alcohol than the TiO₂-NTs electrodes within 2 h. The g-C₃N₄/TiO₂-NTs electrodes can be used for the simultaneous removal of Cr(VI) from wastewater under visible-light irradiation. As shown in Figure 14, hexavalent chromium, phenol, or other organic pollutants in wastewater can be removed simultaneously under visible-light irradiation.

It is worth noting that the performance of the PEC system is highly dependent on the properties and type of the conductive substrate as well as the photocatalytic activity, especially under visible-light irradiation. It is an excellent method for constructing a composite photoanode by coupling g-C₃N₄ with a visible-light active semiconductor with a matching band structure, which can enhance light absorption and charge transfer through a highly interactive interface. The PEC system investigated using g-C₃N₄/Ag/AgCl/BiVO₄ photoanode materials showed high effectiveness in treating wastewater of the pollutants diphenyl ketone and Escherichia coli [170]. Ag/AgCl media not only increase the light absorption and electron-hole transfer pathways of the g-C₃N₄/BiVO₄ heterojunction but also promote high photoanode recovery through their polarization and plasma effects. In a study by C. Murugan [171], g-C₃N₄-BiVO₄ direct Z-type heterojunctions were synthesized using a hydrothermal method, and Figure 15 shows a schematic diagram of their photoelectrocatalytic overall water separation. The results confirmed that the CSI response of BiVO₄ photoanodes containing g-C₃N₄ was improved under AM 1.5 G (100 mWcm⁻²) light. The formation of g-C₃N₄-BiVO₄ dielectric-free direct Z-type heterojunctions enhances interfacial charge transfer and inhibits charge recombination.

The g-C₃N₄/CNTs/Al₂O₃ photoanode was successfully integrated into the membrane filtration of the PEC system [172]. The conductive carbon nanotube layer and the visible-light responsive g-C₃N₄ layer were sequentially coated on the Al₂O₃ membrane carrier, and the composite membrane with a photo adsorption side length of 500 nm and high PEC stability was fabricated. In continuous flow mode, the phenol removal rate was up to 94% from 10 to about 60 min. L. Zhang et al. [173] used MnOx/g-C₃N₄ nanocomposites as a photoanode to decompose chloric acid (CA). Superoxide radicals and positive holes play a leading role in the PEC degradation of CA. The photoanode MnOx/g-C₃N₄-10 showed the highest photocatalytic performance under acidic conditions. Therefore, in a PEC system containing a g-C₃N₄-based anode, pollutant removal occurs at both electrodes. The photogenerated electrons separated from the g-C₃N₄-based photoanode, which are moved to the counter electrode through an external circuit, can reduce O₂ to form reactive oxygen species and further oxidize pollutants. That aside, the holes in the photoanode also can directly oxidize pollutants.

5. Conclusions and Prospective

This article reviewed the modification of g-C₃N₄ and summarized its role in photocatalysis and photoelectrocatalysis and its catalytic degradation applications under different light sources. The modification and optimization of g-C₃N₄ can start from syntheszing a single material or a series of modifications to the material. The photodegradation performance of g-C₃N₄ photocatalysts can be improved by enlarging the effective reaction sites for photocatalysis, prolonging the lifecycle of photogenerated electron holes, and enhancing the visible-light responsiveness of the catalysts. In recent years, the modification of g-C₃N₄-based photo- or photoelectrocatalytic materials has broadened the range of light absorption and increased the absorption and utilization of light by the catalysts. Attempts have been made to apply them to the treatment of various organic pollutants, realizing the efficient degradation of organic pollutants in water. Despite the extensive academic research on and achievements with g-C₃N₄-based materials, there are still many challenges to be faced, such as the following:

• Although the photoelectrocatalytic activity of g-C₃N₄-based photoelectrocatalytic materials has been significantly improved, it still cannot meet the production demand. For the final application of g-C₃N₄ photocatalysts in real pollution treatment, research needs to consider the simplicity and environmental safety of the synthesis method, as well as the recovery and recycling of catalysts.
• There is a lack of systematic understanding of how modifying g-C₃N₄-based photocatalysts affects catalytic activity. Therefore, it is necessary to explore the modification mechanism of g-C₃N₄, elucidate the electron transfer pathways inside the catalysts, and improve the controllability of the modification of g-C₃N₄-based materials.
• More kinds of pollutants with different properties were selected for photo- and photoelectrocatalytic experiments, and the effect of the solution injection volume on degradation efficiency should not be neglected so as to provide ideas and references for the development of more efficient g-C₃N₄-based photo- and photoelectrocatalysts.

Currently, the research on g-C₃N₄ is gradually deepening and has moved from the theoretical development stage to the practical application stage.