Graphitic Carbon Nitride as Visible-Light Photocatalyst Boosting Ozonation in Wastewater Treatment

Light can boost ozone efficiency in advanced oxidation processes (AOPs), either by direct ozone photolysis with UV light or by using a photocatalyst that can be excited with UV-Vis or solar light. The present review summarizes literature data on the combination of ozone and the g-C3N4 photocatalyst for the degradation of probe molecules in water, including oxalic, p-hydroxybenzoic and oxamic acids as well as ciprofloxacin and parabens. g-C3N4 is a metal-free visible-light photocatalyst based on abundant elements that establishes a synergistic effect with ozone, the efficiency of the combination of the photocatalysis and ozonation being higher than the sum of the two treatments independently. Available data indicate that this synergy derives from the higher efficiency in the generation of hydroxyl radicals due to the efficient electron quenching by O3 of photogenerated conduction band electrons in the g-C3N4 photocatalyst. Given the wide use of ozonizers in water treatment, it is proposed that their implementation with g-C3N4 photocatalysis could also boost ozone efficiency in the AOPs of real waste waters.

Although g-C 3 N 4 possesses similar structure to graphite, g-C 3 N 4 exhibits a stacked 2D structure of sheets ideally consisting of s-tris triazine units condensed by tertiary nitrogen atoms. These layers interact by van der Waals forces, thus imparting this material with high thermal and chemical stability. The electronic structure of this material has a band gap of about E g~2 .7 eV, thus finding applications in various fields, including the heterogeneous catalysis and photodegradation of organic pollutants [24][25][26][27][28][29]. The energy values of the conduction band and valence band are −1.1 and +1.6 eV, respectively, which are suitable to perform various redox reactions (Scheme 1). Some of the interesting features of g-C 3 N 4 compared to other conventional photocatalysts are that g-C 3 N 4 is a metal-free solid possessing solely carbon and nitrogen, which are highly abundant in earth, costeffective, environmentally benign and highly safe, visible-light photoresponse and suitable band energy alignment. These interesting factors encouraged researchers to develop photocatalysts based on g-C 3 N 4 for the degradation of organic pollutants without the use of transition metals.
Preparation of g-C 3 N 4 is always a challenging process and many synthetic methods have been reported in the literature [31][32][33]. A variety of g-C 3 N 4 structures have been prepared by employing liquid-based approaches through shaping and casting to obtain large surface-area solid; however, these approaches employ toxic chemicals [34,35]. One of the common strategies to prepare g-C 3 N 4 is through the solid-state reaction from cyanuric chloride and/or, calcium cyanide, lithium azide, or melamine [36][37][38]. In recent years, the thermal decomposition of single precursors such as cyanamide [39], dicyanamide [40], melamine [41], thiourea [42], or urea [43] has also resulted in high-quality and low-defect density g-C 3 N 4 materials (Scheme 2). Nanomaterials 2022, 12, x 2 of 16 photocatalysts based on g-C3N4 for the degradation of organic pollutants without the use of transition metals.

Scheme 1.
Conduction and valence band energy values of different types of g-C3N4 samples. Reproduced with permission from ref. [30]. Copyright 2017 Elsevier.
Preparation of g-C3N4 is always a challenging process and many synthetic methods have been reported in the literature [31][32][33]. A variety of g-C3N4 structures have been prepared by employing liquid-based approaches through shaping and casting to obtain large surface-area solid; however, these approaches employ toxic chemicals [34,35]. One of the common strategies to prepare g-C3N4 is through the solid-state reaction from cyanuric chloride and/or, calcium cyanide, lithium azide, or melamine [36][37][38]. In recent years, the thermal decomposition of single precursors such as cyanamide [39], dicyanamide [40], melamine [41], thiourea [42], or urea [43] has also resulted in high-quality and low-defect density g-C3N4 materials (Scheme 2).
The morphology of pristine g-C 3 N 4 can significantly be altered due to the presence of acidic and C−N bond in its structure. Thinner nanosheets of g-C 3 N 4 can also be prepared retaining their characteristic structural features by exfoliation and treatment with inorganic acids. Transition metal carbides can also be used to prepare thin g-C 3 N 4 nanosheets, but these metal carbides can also be an active site in the photocatalytic reaction. Porosity has been tried to introduce in g-C 3 N 4 by making use of the bubble effect, but it is still a challenge to regulate pore size. Overall, the morphology of g-C 3 N 4 can easily be designed and controlled for a required application with the advantages of easy recovery, high mechanical resistance, and good photocatalytic performance.
As it will be discussed in different examples, the generation of • OH is highly prominent upon visible-light irradiation of g-C 3 N 4 in the presence of O 3 due to the synergistic effect summarized in Equations (1)-(4). In the Vis/O 3 /g-C 3 N 4 photocatalyst, g-C 3 N 4 can absorb the photons under the visible-light irritation to generate eupon CB and holes (h + ) on the VB (Equation (1)). The CB potential of g-C 3 N 4 can be as low as −0.78 V (versus SCE at pH 7), which significantly facilitates CB electron capture by O 3 . As a result, an ozonide radical (2)), and it quickly protonated in the medium to generate a HO 3 • radical (Equation (3)). This trioxide radical easily decomposes into • OH (Equation (4)). The operation of these equations requires the combination of g-C 3 N 4 and O 3 and would not take place in the absence of one of these reagents.
Considering the recent progress made on wastewater treatment through the AOP [45,46], the present review focuses on the AOP combining O 3 as the oxidizing reagent and graphitic carbon nitride (g-C 3 N 4 ) without any metal as the photocatalyst. However, readers are directed to refer the catalytic activity of metal-doped TiO 2 or Ag-ZnO photocatalysts for the degradation of pollutants [47,48]. The purpose of this review is to show that g-C 3 N 4 , in the absence of any precious metal or even any other transition metal, is a very efficient visible-light photocatalyst to activate O 3 generating ROS. Among the various possible ROS, hydroxyl radicals ( • OH) are the most powerful and aggressive species [49][50][51][52], since they have a high oxidation potential. They are also a strong electrophilic species and are able to abstract a hydrogen from virtually any C-H bond, generating carbon-centered radicals [53]. It will be shown in this review that the combination of O 3 and a photocatalyst such as g-C 3 N 4 is a general method to produce high fluxes of • OH and, therefore, is a very powerful AOP treatment.
In many cases, model molecules have been used to evaluate and demonstrate the advantage of combining the g-C 3 N 4 photocatalyst with visible light and O 3 . The present review is organized according to the probe used to demonstrate the efficiency of combining the g-C 3 N 4 photocatalyst and O 3 . Oxalic acid (OA) as well as other reluctant organic pollutants that are selectively degraded by • OH are the favorite probes to show the synergy between the g-C 3 N 4 photocatalyst and O 3 . The last section summarizes the main achievement reached so far with the g-C 3 N 4 photocatalyst and ozonation and provides our prospects for future developments in this field.

Oxalic Acid (OA)
In one of the studies showing the activity to generate • OH radicals, bulk g-C 3 N 4 was synthesized either from thiourea (GCN-T) or dicyandiamide (GCN-D), respectively. Figure 1 provides SEM images of the two GCN samples. The photocatalytic performance of these solids was tested in the mineralization of OA and p-hydroxybenzoic acid (PHBA, Figure 2) under UV and visible-light irradiation [54]. A synergy between photocatalysis by g-C 3 N 4 under visible light and ozonation was found. Under the optimized reaction conditions, the rate constant observed for OA removal using Vis/O 3 /GCN-D was 20.6 times higher than the sum of that in Vis/GCN-D and ozonation. On the other hand, TOC removal of PHBA with Vis/O 3 /GCN-D was 98%, which is about 39.3% higher compared to the sum of the value observed with Vis/GCN-D and ozonation. Interestingly, the Vis/O 3 /GCN-D photocatalytic system showed stronger oxidizing capacity than UV/O 3 /GCN-D for OA degradation with the same light intensity. The inferior activity of UV/O 3 /GCN-D was proposed to be due to the partial direct irradiation of ozone that competes with ozone quenching of the photoinduced electrons on GCN that is the only operating process under visible-light irradiation. Consequently, the amount of produced • OH decreased with UV in comparison to visible light. by g-C3N4 under visible light and ozonation was found. Under the optimized reaction conditions, the rate constant observed for OA removal using Vis/O3/GCN-D was 20.6 times higher than the sum of that in Vis/GCN-D and ozonation. On the other hand, TOC removal of PHBA with Vis/O3/GCN-D was 98%, which is about 39.3% higher compared to the sum of the value observed with Vis/GCN-D and ozonation. Interestingly, the Vis/O3/GCN-D photocatalytic system showed stronger oxidizing capacity than UV/O3/GCN-D for OA degradation with the same light intensity. The inferior activity of UV/O3/GCN-D was proposed to be due to the partial direct irradiation of ozone that competes with ozone quenching of the photoinduced electrons on GCN that is the only operating process under visible-light irradiation. Consequently, the amount of produced • OH decreased with UV in comparison to visible light.  Recently, the photocatalytic generation of charge separation with the appearance of electrons and holes in C3N4 and their trapping by dissolved O2 and O3, as well as ROS evolution, was experimentally determined by spin-trapping upon feeding O3 into the Vis/O2/C3N4 photocatalytic system in aqueous media [55]. EPR measurements showed that a gas mixture of O3 (2.1 mol %)−O2 (97.9 mol %) can facilely trap about double to triple the amount of conduction band (CB)-e − in an aqueous C3N4 suspension irradiated by by g-C3N4 under visible light and ozonation was found. Under the optimized reaction conditions, the rate constant observed for OA removal using Vis/O3/GCN-D was 20.6 times higher than the sum of that in Vis/GCN-D and ozonation. On the other hand, TOC removal of PHBA with Vis/O3/GCN-D was 98%, which is about 39.3% higher compared to the sum of the value observed with Vis/GCN-D and ozonation. Interestingly, the Vis/O3/GCN-D photocatalytic system showed stronger oxidizing capacity than UV/O3/GCN-D for OA degradation with the same light intensity. The inferior activity of UV/O3/GCN-D was proposed to be due to the partial direct irradiation of ozone that competes with ozone quenching of the photoinduced electrons on GCN that is the only operating process under visible-light irradiation. Consequently, the amount of produced • OH decreased with UV in comparison to visible light.  Recently, the photocatalytic generation of charge separation with the appearance of electrons and holes in C3N4 and their trapping by dissolved O2 and O3, as well as ROS evolution, was experimentally determined by spin-trapping upon feeding O3 into the Vis/O2/C3N4 photocatalytic system in aqueous media [55]. EPR measurements showed that a gas mixture of O3 (2.1 mol %)−O2 (97.9 mol %) can facilely trap about double to triple the amount of conduction band (CB)-e − in an aqueous C3N4 suspension irradiated by Recently, the photocatalytic generation of charge separation with the appearance of electrons and holes in C 3 N 4 and their trapping by dissolved O 2 and O 3 , as well as ROS evolution, was experimentally determined by spin-trapping upon feeding O 3 into the Vis/O 2 /C 3 N 4 photocatalytic system in aqueous media [55]. EPR measurements showed that a gas mixture of O 3 (2.1 mol %)−O 2 (97.9 mol %) can facilely trap about double to triple the amount of conduction band (CB)-e − in an aqueous C 3 N 4 suspension irradiated by visible light compared to pure O 2 . This is mainly due to the much higher redox potential and water solubility of O 3 in comparison to O 2 . The capture of CB-e − by O 2 forms • O 2 − , which later converted to • OH through the H 2 O 2 -mediated consecutive three-electronreduction pathway. Quantification of the EPR signal has shown a 17-fold enhancement in the formation of • OH (characterized by the DMPO-OH adduct) and an 84-fold increase in the rate of OA mineralization when 2.1 mol % O 3 is introduced in the Vis/O 2 (O 3 )/bulk C 3 N 4 photocatalytic system ( Figure 3). Interestingly, the rate constant was further increased by a factor of 41 when bulk C 3 N 4 is exfoliated to the nanosheet (NS) C 3 N 4 form ( Figure 4). These results indicate that the use of NS C 3 N 4 exhibits superior performance compared to bulk C 3 N 4 due to its high surface area and upshifted CB edge, which favors more CB-e − to be trapped by dissolved O 3 and O 2 and the formation of higher • OH yield in photocatalytic ozonation compared to bulk C 3 N 4 .
visible light compared to pure O2. This is mainly due to the much higher redox potential and water solubility of O3 in comparison to O2. The capture of CB-e − by O2 forms • O2 − , which later converted to • OH through the H2O2-mediated consecutive three-electron-reduction pathway. Quantification of the EPR signal has shown a 17-fold enhancement in the formation of • OH (characterized by the DMPO-OH adduct) and an 84-fold increase in the rate of OA mineralization when 2.1 mol % O3 is introduced in the Vis/O2(O3)/bulk C3N4 photocatalytic system ( Figure 3). Interestingly, the rate constant was further increased by a factor of 41 when bulk C3N4 is exfoliated to the nanosheet (NS) C3N4 form ( Figure 4). These results indicate that the use of NS C3N4 exhibits superior performance compared to bulk C3N4 due to its high surface area and upshifted CB edge, which favors more CB-e − to be trapped by dissolved O3 and O2 and the formation of higher • OH yield in photocatalytic ozonation compared to bulk C3N4.  In another report, systematic studies were performed with a series of dimensionstructured nanocarbons as the visible-light photocatalysts in the presence of O3 as the metal-free AOP catalyst for water disinfection [56]. Single-walled carbon nanotube (SWCNT), multi-walled carbon nanotube (MWCNT), reduced graphene oxide (rGO) and fullerene (C60) exhibited superior catalytic performance in catalytic ozonation, while g-C3N4 and C60 outperformed in the visible light-O3 coupled photocatalytic process. The results are presented in Figure 5. The coupling coefficient of visible light with ozone on g-C3N4 measuring the synergy arising from the combination of ozonation and photocatalysis in comparison with the sum of the separate treatments was found to be as 95.8. Both visible light compared to pure O2. This is mainly due to the much higher redox potential and water solubility of O3 in comparison to O2. The capture of CB-e − by O2 forms • O2 − , which later converted to • OH through the H2O2-mediated consecutive three-electron-reduction pathway. Quantification of the EPR signal has shown a 17-fold enhancement in the formation of • OH (characterized by the DMPO-OH adduct) and an 84-fold increase in the rate of OA mineralization when 2.1 mol % O3 is introduced in the Vis/O2(O3)/bulk C3N4 photocatalytic system ( Figure 3). Interestingly, the rate constant was further increased by a factor of 41 when bulk C3N4 is exfoliated to the nanosheet (NS) C3N4 form ( Figure 4). These results indicate that the use of NS C3N4 exhibits superior performance compared to bulk C3N4 due to its high surface area and upshifted CB edge, which favors more CB-e − to be trapped by dissolved O3 and O2 and the formation of higher • OH yield in photocatalytic ozonation compared to bulk C3N4.  In another report, systematic studies were performed with a series of dimensionstructured nanocarbons as the visible-light photocatalysts in the presence of O3 as the metal-free AOP catalyst for water disinfection [56]. Single-walled carbon nanotube (SWCNT), multi-walled carbon nanotube (MWCNT), reduced graphene oxide (rGO) and fullerene (C60) exhibited superior catalytic performance in catalytic ozonation, while g-C3N4 and C60 outperformed in the visible light-O3 coupled photocatalytic process. The results are presented in Figure 5. The coupling coefficient of visible light with ozone on g-C3N4 measuring the synergy arising from the combination of ozonation and photocatalysis in comparison with the sum of the separate treatments was found to be as 95.8. Both In another report, systematic studies were performed with a series of dimensionstructured nanocarbons as the visible-light photocatalysts in the presence of O 3 as the metalfree AOP catalyst for water disinfection [56]. Single-walled carbon nanotube (SWCNT), multi-walled carbon nanotube (MWCNT), reduced graphene oxide (rGO) and fullerene (C60) exhibited superior catalytic performance in catalytic ozonation, while g-C 3 N 4 and C60 outperformed in the visible light-O 3 coupled photocatalytic process. The results are presented in Figure 5. The coupling coefficient of visible light with ozone on g-C 3 N 4 measuring the synergy arising from the combination of ozonation and photocatalysis in comparison with the sum of the separate treatments was found to be as 95.8. Both g-C 3 N 4 and C60 promoted the synergism between the visible-light photocatalysis and O 3 in the generation of • OH radicals for the efficient removal of OA compared to O 3 /nanocarbon (SWCNT, MWCNT and rGO) or even benchmark photocatalysts such as Vis/O 3 /metal oxides (WO 3 and TiO 2 ). Among these carbon materials, the superior activity of g-C 3 N 4 is due to the narrow bandgap and upshifted CB minimum of g-C 3 N 4 for the visible-light photocatalytic ozonation.
g-C3N4 and C60 promoted the synergism between the visible-light photocatalysis and O3 in the generation of • OH radicals for the efficient removal of OA compared to O3/nanocarbon (SWCNT, MWCNT and rGO) or even benchmark photocatalysts such as Vis/O3/metal oxides (WO3 and TiO2). Among these carbon materials, the superior activity of g-C3N4 is due to the narrow bandgap and upshifted CB minimum of g-C3N4 for the visible-light photocatalytic ozonation. In another study, the coupling of g-C3N4 or chlorine modified g-C3N4 (Cl/g-C3N4) photocatalysts with ozonation was employed as an effective strategy for the mineralization of OA under a visible-light irradiation condition [57]. The use of g-C3N4 and Cl/g-C3N4 was able to trigger a synergy between photocatalysis and ozonation with a coupling coefficient of 17.8 and 9.9, respectively, compared to the sum of the OA degradation by the two treatments separately ( Figure 6). Further, the combination of CB electrons and ozone effectively promoted surface-charge separation of g-C3N4 and self-decomposition of ozone to • OH in a much higher selectivity. The experimental results indicated that • OH is primarily responsible for the mineralization of OA in Vis/O3/g-C3N4, while in comparison • O2 − and other ROS are more important than • OH radicals in the g-C3N4 photocatalytic mineralization of OA. The removal efficiency of OA was 28% with Cl/g-C3N4 after 120 min under visible-light irradiation, which is 10% higher compared to g-C3N4. Further, the photocatalytic efficiency of Cl/g-C3N4 was 1.55 times higher than pure g-C3N4, a fact that has been attributed to the higher active surface area upon chlorination of g-C3N4. In another study, the coupling of g-C 3 N 4 or chlorine modified g-C 3 N 4 (Cl/g-C 3 N 4 ) photocatalysts with ozonation was employed as an effective strategy for the mineralization of OA under a visible-light irradiation condition [57]. The use of g-C 3 N 4 and Cl/g-C 3 N 4 was able to trigger a synergy between photocatalysis and ozonation with a coupling coefficient of 17.8 and 9.9, respectively, compared to the sum of the OA degradation by the two treatments separately ( Figure 6). Further, the combination of CB electrons and ozone effectively promoted surface-charge separation of g-C 3 N 4 and self-decomposition of ozone to • OH in a much higher selectivity. The experimental results indicated that • OH is primarily responsible for the mineralization of OA in Vis/O 3 /g-C 3 N 4 , while in comparison • O 2 − and other ROS are more important than • OH radicals in the g-C 3 N 4 photocatalytic mineralization of OA. The removal efficiency of OA was 28% with Cl/g-C 3 N 4 after 120 min under visible-light irradiation, which is 10% higher compared to g-C 3 N 4 . Further, the photocatalytic efficiency of Cl/g-C 3 N 4 was 1.55 times higher than pure g-C 3 N 4 , a fact that has been attributed to the higher active surface area upon chlorination of g-C 3 N 4 . Reproduced with permission from ref. [57]. Copyright 2015 Elsevier.
In one of the earliest reports, g-C3N4 prepared from thiourea in air at 550 °C was reported as an active photocatalyst for the removal of OA (Figure 7) and bisphenol A (Figure 8) coupling photocatalysis and ozonation. [58] Experimental data showed that OA degradation using g-C3N4/Vis/O3 was 65.2%, which is higher than the sum of the OA degradation values reached by g-C3N4/Vis and O3 separately. The C/C0 of g-C3N4+O3 is higher In one of the earliest reports, g-C 3 N 4 prepared from thiourea in air at 550 • C was reported as an active photocatalyst for the removal of OA (Figure 7) and bisphenol A (Figure 8) coupling photocatalysis and ozonation. [58] Experimental data showed that OA degradation using g-C 3 N 4 /Vis/O 3 was 65.2%, which is higher than the sum of the OA degradation values reached by g-C 3 N 4 /Vis and O 3 separately. The C/C 0 of g-C 3 N 4 +O 3 is higher than with g-C 3 N 4 +Vis due to the generation of a high flux of • OH radicals. On the other hand, the TOC removal of bisphenol A with g-C 3 N 4 /Vis/O 3 was 2.17 times higher than the sum of the ratio with g-C 3 N 4 /Vis and O 3 . This superior performance of the g-C 3 N 4 /Vis/O 3 system was attributed to the synergistic effect between photocatalysis and ozonation by g-C 3 N 4 . This synergistic effect results in the generation of higher • OH yields, which are the species responsible for the enhanced degradation of organic pollutants. Figure 6. Degradation of OA by various oxidation processes upon g-C3N4 (A) and Cl/g-C3N4 (B). Reproduced with permission from ref. [57]. Copyright 2015 Elsevier.
In one of the earliest reports, g-C3N4 prepared from thiourea in air at 550 °C was reported as an active photocatalyst for the removal of OA (Figure 7) and bisphenol A (Figure 8) coupling photocatalysis and ozonation. [58] Experimental data showed that OA degradation using g-C3N4/Vis/O3 was 65.2%, which is higher than the sum of the OA degradation values reached by g-C3N4/Vis and O3 separately. The C/C0 of g-C3N4+O3 is higher than with g-C3N4+Vis due to the generation of a high flux of • OH radicals. On the other hand, the TOC removal of bisphenol A with g-C3N4/Vis/O3 was 2.17 times higher than the sum of the ratio with g-C3N4/Vis and O3. This superior performance of the g-C3N4/Vis/O3 system was attributed to the synergistic effect between photocatalysis and ozonation by g-C3N4. This synergistic effect results in the generation of higher • OH yields, which are the species responsible for the enhanced degradation of organic pollutants.   As commented earlier, C3N4 has been effectively employed as heterogeneous photocatalysts for water disinfection under visible-light irradiation [59]. However, one of the issues that needs to be addressed is the stability of C3N4 under photocatalytic conditions since the reaction process generates ROS and other reactive radical species. In this aspect, Cao and coworkers have reported the chemical stability of C3N4 under exposure to ROS during photocatalytic water treatment (Figure 9). The experimental results indicated that • OH can attack the photocatalyst removing the heptazine unit from the C3N4 sheets, generating secondary pollutants in the aqueous environment. In contrast, C3N4 is chemically stable toward • O2 − and O3. Interestingly, the decomposition of C3N4 was fully or partially inhibited in the presence of organic pollutants due to their competition for • OH. Hence, As commented earlier, C 3 N 4 has been effectively employed as heterogeneous photocatalysts for water disinfection under visible-light irradiation [59]. However, one of the issues that needs to be addressed is the stability of C 3 N 4 under photocatalytic conditions since the reaction process generates ROS and other reactive radical species. In this aspect, Cao and coworkers have reported the chemical stability of C 3 N 4 under exposure to ROS during photocatalytic water treatment (Figure 9). The experimental results indicated that • OH can attack the photocatalyst removing the heptazine unit from the C 3 N 4 sheets, generating secondary pollutants in the aqueous environment. In contrast, C 3 N 4 is chemically stable toward • O 2 − and O 3 . Interestingly, the decomposition of C 3 N 4 was fully or partially inhibited in the presence of organic pollutants due to their competition for • OH. Hence, the photocatalyst exhibited high activity and stability under these conditions. This work provides useful information about the chemical instability of C 3 N 4 -based materials in those processes where • OH is the major involved ROS in various applications such as water treatment and organic synthesis. mission from ref. [58]. Copyright 2014 Elsevier.
As commented earlier, C3N4 has been effectively employed as heterogeneous photocatalysts for water disinfection under visible-light irradiation [59]. However, one of the issues that needs to be addressed is the stability of C3N4 under photocatalytic conditions since the reaction process generates ROS and other reactive radical species. In this aspect, Cao and coworkers have reported the chemical stability of C3N4 under exposure to ROS during photocatalytic water treatment (Figure 9). The experimental results indicated that • OH can attack the photocatalyst removing the heptazine unit from the C3N4 sheets, generating secondary pollutants in the aqueous environment. In contrast, C3N4 is chemically stable toward • O2 − and O3. Interestingly, the decomposition of C3N4 was fully or partially inhibited in the presence of organic pollutants due to their competition for • OH. Hence, the photocatalyst exhibited high activity and stability under these conditions. This work provides useful information about the chemical instability of C3N4-based materials in those processes where • OH is the major involved ROS in various applications such as water treatment and organic synthesis.

p-Hydroxybenzoic Acid (PHBA)
Porous g-C3N4 (PGCN) has recently received wide attention due to the easy access to the interior of the nanoporous framework [60,61]. In this aspect, a one-pot template-free approach was employed to obtain honeycomb-like PGCN by the reaction between 3. p-Hydroxybenzoic Acid (PHBA) Porous g-C 3 N 4 (PGCN) has recently received wide attention due to the easy access to the interior of the nanoporous framework [60,61]. In this aspect, a one-pot templatefree approach was employed to obtain honeycomb-like PGCN by the reaction between ammonium chloride and the precursor of g-C 3 N 4 , followed by calcination. The photocatalytic activity of PGCN was examined in the photocatalytic activity for PHBA degradation upon visible-light irradiation ( Figure 10) [62]. However, PHBA was difficult to mineralize by PGCN as a photocatalyst; an unfavorable factor was the larger band gap of PGCN compared to g-C 3 N 4 . To overcome these difficulties, the photocatalytic activity of PGCN was coupled with ozonation in a Vis/PGCN/O 3 AOP. The photoactivity data show that photocatalysis by PGCN and O 3 establish a synergistic effect. Under optimized AOP conditions, Vis/O 3 /PGCN promotes quantitative PHBA mineralization with the dosage of O 3 as 1.5 mg/min. The process is further accelerated by increasing the O 3 dosage. This synergism derives from the enhanced generation of • OH. These generated • OH radicals spontaneously react with PHBA and its O 3 -recalcitrant intermediates, such as carboxylic acids, leading to complete mineralization to CO 2 and H 2 O (Figure 11). This is a nice example illustrating the possible integration of sunlight/PGCN with O 3 as a metal-free photocatalyst for the AOP in water disinfection.
ditions, Vis/O3/PGCN promotes quantitative PHBA mineralization with the dosage of O3 as 1.5 mg/min. The process is further accelerated by increasing the O3 dosage. This synergism derives from the enhanced generation of • OH. These generated • OH radicals spontaneously react with PHBA and its O3-recalcitrant intermediates, such as carboxylic acids, leading to complete mineralization to CO2 and H2O (Figure 11). This is a nice example illustrating the possible integration of sunlight/PGCN with O3 as a metal-free photocatalyst for the AOP in water disinfection.   ditions, Vis/O3/PGCN promotes quantitative PHBA mineralization with the dosage of O3 as 1.5 mg/min. The process is further accelerated by increasing the O3 dosage. This synergism derives from the enhanced generation of • OH. These generated • OH radicals spontaneously react with PHBA and its O3-recalcitrant intermediates, such as carboxylic acids, leading to complete mineralization to CO2 and H2O (Figure 11). This is a nice example illustrating the possible integration of sunlight/PGCN with O3 as a metal-free photocatalyst for the AOP in water disinfection.

Ciprofloxacin
In another report, the photocatalytic degradation of a ciprofloxacin (CIP) antibiotic in water was performed using nanosheets of g-C 3 N 4 as catalysts under visible-light irradiation using white light LEDs. [63] The degradation of CIP was around 90% in 60 min using g-C 3 N 4 under ozonation conditions using visible-light irradiation. Further, the other objective of this work was the identification of the intermediate byproducts formed upon degradation and to establish the sequential pathway for CIP degradation with possible experimental evidence from liquid chromatography coupled to high-resolution mass spectrometry using a Q-TOF instrument. Seven intermediates were proposed, three of them reported for the first time. Kinetic studies showed that CIP degradation proceeds through a pseudofirst order kinetics with a rate constant of 0.035 min −1 . The addition of triethanolamine significantly decreased the rate constant to 0.00072 min −1 , suggesting that CIP degradation is initiated by the holes generated in the catalyst. In addition, the main pathway for CIP degradation was the attack to the piperazine ring by • OH radicals, followed by the rupture of the heterocyclic ring and a suite of consecutive reactions, including the loss of two carbon atoms as CO 2 , defluorination, oxidation and cleavage of the cycles of this intermediate. Figure 12 summarizes the proposed CIP degradation sequence.

Oxamic Acid (OMA)
Recently, g-C3N4 has been reported as a heterogeneous photocatalyst for the photocatalytic ozonation of OMA in aqueous solution. The bulk g-C3N4 material was thermally post-treated at 500 °C to obtain g-C3N4-500 that exhibits an increased surface area respect to the bulk material [64]. Experimental data show that the photocatalytic ozonation by C3N4 was highly effective in the removal of OMA, reaching complete OMA degradation with C3N4-500 after 120 min of irradiation ( Figure 13). The high activity of C3N4-500 is due to the combination of photoinduced charge separation along with ozonation to produce a higher number of • OH radicals. On the other hand, the decrease in the rate of OMA removal in the presence of scavengers is compatible with photogenerated holes on the catalyst surface, playing a dominant role in OMA degradation in comparison to • OH radicals. Although the solid was reused for three cycles without much change in its physicochemical properties, a slight decay in the degradation performance was observed and

Oxamic Acid (OMA)
Recently, g-C 3 N 4 has been reported as a heterogeneous photocatalyst for the photocatalytic ozonation of OMA in aqueous solution. The bulk g-C 3 N 4 material was thermally post-treated at 500 • C to obtain g-C 3 N 4 -500 that exhibits an increased surface area respect to the bulk material [64]. Experimental data show that the photocatalytic ozonation by C 3 N 4 was highly effective in the removal of OMA, reaching complete OMA degradation with C 3 N 4 -500 after 120 min of irradiation ( Figure 13). The high activity of C 3 N 4 -500 is due to the combination of photoinduced charge separation along with ozonation to produce a higher number of • OH radicals. On the other hand, the decrease in the rate of OMA removal in the presence of scavengers is compatible with photogenerated holes on the catalyst surface, playing a dominant role in OMA degradation in comparison to • OH radicals. Although the solid was reused for three cycles without much change in its physicochemical properties, a slight decay in the degradation performance was observed and attributed to modifications in the C 3 N 4 -500 exposed structure occurring in the course of the photocatalytic reaction. attributed to modifications in the C3N4-500 exposed structure occurring in the course of the photocatalytic reaction. Figure 13. OMA removal by ozonation (◼), photolysis (•), photochemical ozonation (♦), and photocatalytic ozonation using g-C3N4 (▼), g-C3N4-500 (▲) and TiO2 (⬟). Reproduced with permission from ref. [64]. Copyright 2020 Elsevier

Parabens
Besides the above discussed examples with visible-light irradiation, g-C3N4 was reported as a cost-effective and efficient photocatalyst for the degradation of a mixture of parabens through photo-assisted processes [65]. Control experiments indicated that the use of UV-A radiation exhibited higher activation of g-C3N4 compared to visible light. The photocatalytic ozonation process showed higher degradation rates of parabens with a ozone dosage lower than the corresponding dark ozonation process. Optimization studies revealed that the medium with basic and neutral conditions (pH = 7-11) provides a better interaction between catalysts and contaminants as well as the highest generation of radicals. Under the optimized reaction conditions of a 500 mg L −1 catalyst concentration and a paraben concentration of 1 mg L −1 , >95% removal was achieved for the three parabens (methyl-, ethyl-and propylparaben) in less than 15 min ( Figure 14). Further, these conditions were also effective for the degradation of Allivibrio fischeri bacteria by a significant decrease in its luminescence inhibition, providing a non-toxic, disinfected solution. attributed to modifications in the C3N4-500 exposed structure occurring in the course of the photocatalytic reaction. Figure 13. OMA removal by ozonation (◼), photolysis (•), photochemical ozonation (♦), and photocatalytic ozonation using g-C3N4 (▼), g-C3N4-500 (▲) and TiO2 (⬟). Reproduced with permission from ref. [64]. Copyright 2020 Elsevier

Parabens
Besides the above discussed examples with visible-light irradiation, g-C3N4 was reported as a cost-effective and efficient photocatalyst for the degradation of a mixture of parabens through photo-assisted processes [65]. Control experiments indicated that the use of UV-A radiation exhibited higher activation of g-C3N4 compared to visible light. The photocatalytic ozonation process showed higher degradation rates of parabens with a ozone dosage lower than the corresponding dark ozonation process. Optimization studies revealed that the medium with basic and neutral conditions (pH = 7-11) provides a better interaction between catalysts and contaminants as well as the highest generation of radicals. Under the optimized reaction conditions of a 500 mg L −1 catalyst concentration and a paraben concentration of 1 mg L −1 , >95% removal was achieved for the three parabens (methyl-, ethyl-and propylparaben) in less than 15 min ( Figure 14). Further, these conditions were also effective for the degradation of Allivibrio fischeri bacteria by a significant decrease in its luminescence inhibition, providing a non-toxic, disinfected solution.

Parabens
Besides the above discussed examples with visible-light irradiation, g-C 3 N 4 was reported as a cost-effective and efficient photocatalyst for the degradation of a mixture of parabens through photo-assisted processes [65]. Control experiments indicated that the use of UV-A radiation exhibited higher activation of g-C 3 N 4 compared to visible light. The photocatalytic ozonation process showed higher degradation rates of parabens with a ozone dosage lower than the corresponding dark ozonation process. Optimization studies revealed that the medium with basic and neutral conditions (pH = 7-11) provides a better interaction between catalysts and contaminants as well as the highest generation of radicals. Under the optimized reaction conditions of a 500 mg L −1 catalyst concentration and a paraben concentration of 1 mg L −1 , >95% removal was achieved for the three parabens (methyl-, ethyl-and propylparaben) in less than 15 min ( Figure 14). Further, these conditions were also effective for the degradation of Allivibrio fischeri bacteria by a significant decrease in its luminescence inhibition, providing a non-toxic, disinfected solution.
The Table 1 summarizes the evidence for the generation of • OH formation and their quantification methods for the various catalysts that have been discussed in this review.  Figure 14. Degradation of methyl-, ethyl-and propylparaben over photocatalytic oxidation and photocatalytic, catalytic, and single ozonation process. Reproduced with permission from ref. [65]. Copyright 2022 Elsevier.
The Table 1 summarizes the evidence for the generation of • OH formation and their quantification methods for the various catalysts that have been discussed in this review.  Figure 14. Degradation of methyl-, ethyl-and propylparaben over photocatalytic oxidation and photocatalytic, catalytic, and single ozonation process. Reproduced with permission from Ref. [65]. Copyright 2022 Elsevier.

Conclusions and Prospects
The examples discussed above refer to the degradation of probe molecules by combining photocatalysis by g-C 3 N 4 and O 3 . Activity data have shown the involvement of synergistic effects by this combination as the AOP, resulting in a degradation level that is much higher than the sum of the degradation degree reached independently by any of the two components. The available mechanistic data indicate that this synergy derives from the higher efficiency of • OH formation with the combined g-C 3 N 4 photocatalysis/ozonation process due to the capture of the photogenerated electrons in the g-C 3 N 4 semiconductor by O 3 as electron acceptor. It has also been commented that, although g-C 3 N 4 undergoes self-attack by photogenerated • OH, releasing some additional pollutant in water, selfdegradation is a minor process when there are some organic molecules present competing for • OH attack. In that way, together with the absence of any transition metal, the combination of the g-C 3 N 4 photocatalyst and ozonation appears as a practical method, easy to implement for wastewater treatment. Feasibility of implementation also derives from the commercially available large-scale ozonizers that have been already deployed in many plants for wastewater treatment and, thus, a burst in the efficiency can be easily anticipated just by complementing these ozonizers with natural sunlight photocatalysis. The target in this field will be just to confirm laboratory data with probe molecules in real wastewater treatment plants.

Conflicts of Interest:
The authors declare no conflict of interest.