Salt-Mediated Structural Transformation in Carbon Nitride: From Regulated Atomic Configurations to Enhanced Photocatalysis

Abstract

Molten salts-assisted synthesis is widely used in the construction of high efficiency graphitic carbon nitride (g-C₃N₄) photocatalysts, and two isotypes of g-C₃N₄ have been synthesized by such method, namely poly (heptazine imide) (PHI) and poly (triazine imide) (PTI). However, the understanding of the structural changes taking place during the molten salt process and the structure–activity relationship of g-C₃N₄ polymorphs remain blurred. Herein, by regulating the treatment duration of g-C₃N₄ nanosheets (melon) in molten salts, we successfully synthesized g-C₃N₄ with phases of PHI, PHI/PTI and PTI. A continuous structural transformation induced by ions, in which melon transforms to a stable phase PTI via PHI, an intermediate state, was unveiled for the first time. In addition, results reveal that atomic configurations play a vital role in photo absorption, and charge carrier transfer and surface reaction, leading to significant differences in photocatalytic degradation. Among them, PHI with K⁺ and cyan groups modification, as well as high crystallinity, exhibits remarkable degradation efficiency, with 90% removal of tetracycline in 10 min and 80% removal of phenol in 60 min. This study sheds light on a deeper understanding for the molten salt-assisted synthesis and provides new ideas for preparing efficient organic semiconductor photocatalysts.

1. Introduction

Solar-driven photocatalysis has attracted much attention in the field of water treatment due to its utilization of renewable energy in recent years [1,2]. During the photocatalytic reactions, various reactive oxygen species, such as H₂O₂, can be generated to remove pollutants from water bodies [1,3]. However, the bottleneck limiting the practical application of this advanced technology is to improve the efficiency of photocatalysts. In recent years, organic polymer semiconductors, especially graphitic carbon nitride (g-C₃N₄), has been widely studied in photocatalysis owing to a low-cost processibility, a finely tunable structure and environmental friendliness characteristics [4,5,6]. The commonly studied g-C₃N₄ was first proposed by Wang et al. in 2009 [7], and it can be synthesized from nitrogen-rich precursors such as cyanamide, dicyandiamide or urea by simple calcination [8]. During synthesis, the C-N precursors undergo a solid-state condensation reaction by releasing ammonia to form in-plane aromatic C-N rings, resulting in a 2D graphite-like structure named “melon” [9]. In contrast to the ideal 2D network proposed by Teter and Hemley [10], the melon structure is composed of incompletely polymerized chains made up of heptazine rings, with adjacent chains linked by hydrogen bonds [9]. The presence of incomplete polymerization and hydrogen bonding tends to reduce the crystallinity of g-C₃N₄ and leads to poor photogenerated charge transfer and photocatalytic efficiency [11].

A number of methods have been introduced to improve the crystallinity of g-C₃N₄, including solvothermal treatment [12], rapid heating [13] and molten salts synthesis [14], of which molten salts synthesis is considered to be the most promising one. In the molten salt process, molten salts (e.g., potassium chloride, lithium chloride or a mixture of them) are able to dissolve monomers and intermediates (e.g., melem or melon) and provide a special liquid environment for the polycondensation, thus improving local order [14,15]. More importantly, a molten salt-assisted synthesis offers opportunities for the formation of new isomers of g-C₃N₄. For example, Wirnhier et al. reported that s-triazine-based crystalline g-C₃N₄, that is, poly (triazine imide) (PTI), could be obtained by the polycondensation of dicyandiamide in mixed molten salts of KCl and LiCl [16]. This material exhibited superior photocatalytic activity compared to the polymeric melon [17,18]. Under similar molten salt conditions, Lin et al. synthesized tri-s-triazine-based crystalline g-C₃N₄, that is, poly (heptazine imide) (PHI), using melon type g-C₃N₄ as a starting material [11]. In comparison with PTI, PHI presented a higher photocatalytic efficiency [11]. However, substantial efforts have been focused on enhancing the photocatalytic activity of g-C₃N₄ by molten salts treatment or on resolving the atomic structures of PTI and PHI, seldom aimed at exploiting the evolution of structure and property of g-C₃N₄ throughout the molten salts process.

In this work, we explore in depth the structural transformation during the recrystallization process of melon in a molten salt environment at the molecular level and report a PHI catalyst with excellent photocatalytic degradation activity. Urea-based g-C₃N₄ nanosheets (melon type) were selected as the starting material and calcined in the molten salts of KCl and LiCl with different treatment times. Upon increasing the duration in molten salts, PHI, PHI/PTI and PTI phases appeared successively in the prepared materials. Therefore, we propose a new route for the polycondensation of g-C₃N₄ in the molten salts, where the melon transformed into a stable and highly crystalline PTI phase via an intermediate state, PHI, as schematically illustrated in Figure 1a. Due to the variance in atomic configurations, the PHI sample exhibited better crystallinity and featured a special microstructure and electronic structure, which enhances the photo absorption performance, accelerates the photogenerated charge separation and transfer process, and provides more surface reaction sites, leading to excellent photocatalytic degradation activity.

2. Results and Discussion

2.1. Structure and Composition

In the context of the molten salt-assisted preparation, we first carried out a detailed analysis of the structural composition of the carbon nitride, including the melon, PHI and PTI structures, as seen in Figure 1a. It is clearly seen that from the pristine melon-type g-C₃N₄, the formation of PHI structure undergoes mainly two processes: the breaking of inter-chain hydrogen bonds and the bridging of adjacent melon chains. In addition to the above processes, the formation of the PTI structure also requires the rupture of the heptazine ring. Furthermore, we used density functional theory (DFT) calculations to investigate the correlation between the energy and atomic configurations of g-C₃N₄. As shown in Figure S1 (Supporting information), the internal energy decreases from melon to PTI, indicating the order of thermodynamic stability is PTI > PHI > melon. As proof to the concept, in this study, the target g-C₃N₄ materials were prepared by a two-stage calcination method [17], as illustrated in Figure 1b.

The structural changes were investigated by X-ray diffraction (XRD), Fourier transform-infrared (FT-IR) spectra, X-ray photoelectron spectra (XPS) and solid-state nuclear magnetic resonance (NMR). Figure 2a shows the XRD patterns of the g-C₃N₄ samples.

The pristine melon presented two characteristic peaks at 13.0° and 27.5°, corresponding to the (100)_hex or (210)_ortho and (002) lattice planes, which can be attributed to the in-plane periodic repeating units and graphite-like stacking [7,9,19]. As for PHI, the (100) peak shifts from 13.0° (in melon) to 8.0°, indicating an extension of the in-plane repeating units, while the (002) peak shifts from 27.5° to 28.0°, demonstrating that the adjacent layers of PHI are closer than that of melon [11,20]. Both of the peaks could be indexed to the PHI structure and the expanded in-plane space/narrowed interlayer spacing could be attributed to the embedded K⁺ ions in PHI framework [11]. In contrast, multiple sharp peaks were observed in the XRD pattern of PTI, and all peaks can be indexed to those of PTI structure [21]. In addition, characteristic peaks for both PHI and PTI were observed in the PHI/PTI sample, indicating the presence of both phases in this sample. The above results clearly show that the structure of g-C₃N₄ undergoes a transition from melon to PHI to PTI in the molten salts as the treatment time increases. Furthermore, we also performed the simulation of XRD peaks of the melon, PHI and PTI structures. Figure 2b–d shows the comparison of the calculated data and the experimental data. It is clearly seen that the simulated XRD peaks of melon and PTI match well with the experimental XRD patterns. However, the simulated XRD data of PHI presented a much stronger intensity of the (100) peak than the (002) peak, different from the experimental data, which may be caused by the incomplete bonding of in-plane units.

FT-IR was used to further investigate the structural evolution of the samples. As shown in Figure 3a, all the g-C₃N₄ samples show similar vibrational patterns. Two characteristic bands are found at 810 and 1100~1700 cm⁻¹, which can be attributed to the breathing and stretching modes of the triazine or heptazine C-N rings, respectively [11]. The broad band located between 3000~3500 cm⁻¹ can be assigned to the stretching vibration mode of the surface unpolymerized amino or hydroxyl groups [16,22]. Notably, this band weakens significantly from melon to PTI, reflecting the improvement of condensation. In addition, a weak band at 2180 cm⁻¹ was observed in PHI and PHI/PTI, and it can be attributed to surface unpolymerized cyan groups [11,20]. In addition, PTI exhibited a weak band at 650 cm⁻¹. However, this band becomes weaker in PHI/PTI and has even disappeared in melon and PHI, suggesting that heptazine units are present in melon, PHI, and PHI/PTI while absent in PTI, which is consistent with their structural characteristics [20]. This phenomenon will be further verified in XPS and NMR analyses.

XPS was used to investigate the chemical compositions and local atomic environments of the g-C₃N₄ samples. In Figure 3b, all samples show strong C-1s and N-1s signals, along with a weak O-1s signal. A K-2p signal can be observed in PHI and PHI/PTI, indicating that PHI structure is present in both samples [11,23]. In contrast, a Cl-2p signal existed in PHI/PTI and PTI; in addition, a weak Li-1s signal can be found in the PTI sample. The presence of Li and Cl signals is consistent with the previously reported PTI/Li⁺Cl⁻ structure [24]. As shown in Figure S2a (Supporting information), the C-1s spectra contains three components with binding energy of 284.8, 286.5 and 288.2 eV. These peaks are associated with the adventitious C, the C adjacent to the amino groups and the sp²-hybridized C in the N=C-N [11]. Notably, the N=C-N peak of PTI is weaker compared to the other three samples, which may be caused by its longest duration in the salt treatment. As shown in Figure 3c, the N-1s spectra can be decomposed into three or four peaks. The main peak at 398.6 eV can be attributed to the sp²-hybridised N in C=N-C, and the peaks at 400.0 and 401.0 eV are associated with the tertiary N in N-(C)₃ and the N in -NH_x groups [11,20], the former of which may be the characteristic N atoms of heptazine ring units. By the combined analysis of C-1s and N-1s spectra, the basic structural units of melon, PHI, and PHI/PTI are determined to be heptazine rings. Interestingly, the PTI sample exhibited only one peak between 399 and 402 eV, representing the bridging N in the PTI frameworks [24], without the contribution of N-(C)₃, suggesting that PTI contained little or no heptazine rings. Therefore, the basic units of PTI are composed of triazine rings. Moreover, a weak peak was observed at 404.5 eV in N-1s spectra, representing the charge effect or positive charge localization effect in the C-N heterocyclic rings [24]. For the K-2p, Cl-2p and Li-1s spectra, two peaks were observed in K and Cl spectra, which can be attributed to the 2p1 and 2p3 contributions, respectively (Figure S2b,c, supporting information) [11]; whereas the latter exhibited only one peak representing the Li 1s structure (Figure S2d, supporting information) [24].

To further investigate the atomic environment and structural changes, we measured the solid-state ¹³C NMR spectra of the g-C₃N₄ samples. As shown in Figure 3d, the three major peaks marked 1, 2, 3 and 4 can be recognized. The peak 1, at 156~158 ppm, is attributed to core C of C-(N)₃ in heptazine units [23,25]. The predominant peak 2, at 163–164 ppm, is associated with the C next to amino groups [25]. The weak peak 3, at 168~171 ppm, is assigned to the C nucleus adjacent to non-protonated ring nitrogen [16,23]. Peak 1 was present in the melon, PHI and PHI/PTI samples, the structures of which contained heptazine rings, as shown in Figure 1a. In contrast, peak 1 was barely observed in PTI, which may be due to the absence of heptazine rings in PTI, in agreement with the XPS analysis finding seldom or no N-(C)₃ units (Figure 3c). The weak peak 3 and 4 may be caused by the negatively charged groups, such as CN₂(N⁻) and cyan groups [23]. When the individual units responsible for these peaks are marked 1, 2, 3 and 4 in the insets of Figure 3d, it becomes clear that those major NMR peaks agree with the crystal structures of all samples as determined by the XRD.

2.2. Microscopic Morphology

The morphologies of the samples were studied by scanning electron microscope (SEM) and transmission electron microscope (TEM). Figure 4a–d and Figure 4e–h show SEM and bright-field TEM images of melon, PHI, PHI/PTI and PTI, respectively, while the inset in Figure 4h is a high-resolution TEM image of PTI. The pristine melon exhibited the morphology of ultrathin nanosheets, with most of which curled into nanotubes due to the high surface energy. Rod-like nanostructures were recognized in PHI, and the bent nanorods were constructed from the ultrathin particles with size of approximately 20 nm. As for PTI, highly crystalline hexagonal prisms were observed, which is the typical morphology of PTI samples as reported by previous articles [15,18,26]. In addition, both morphologies of PHI and PTI can be recognized in the PHI/PTI sample, in accordance with the coexistence of PHI and PTI characteristic peaks observed in the XRD pattern.

The microstructural changes in the g-C₃N₄ samples were investigated by N₂ adsorption-desorption at 77.15 K (Figure S3, supporting information). According to the IUPAC classification, all samples exhibit type IV isotherms with H1 hysteresis loops, indicating the typical mesoporous microstructures. It is noteworthy that the relative surface area of PHI was substantially increased compared to pristine melon, which may be caused by the smaller size of particles. In addition, because of the formation of PTI hexagonal prisms, which exhibited a larger particle size than PHI, the specific surface area was reduced from PHI to PTI.

2.3. Structural Transformation Pathway

Encouraged by the above structural characterizations, we propose a hypothesis that the pristine melon-type g-C₃N₄ undergoes a two-stage sequential structural transformation in the molten salts, that is, melon transforms to the intermediate state PHI and then to the stable state PTI. The evolution at the molecular level is demonstrated in Scheme 1 [27]. In detail, in the molten salts, the interchain hydrogen bonds in melon were first broken under the attack of metal ions. The PHI structure is then formed by the bridging of terminated amino groups and the emission of NH₃. In the meantime, K⁺ were anchored in the cavities of the PHI frameworks due to the coulombic force between it and negatively charged groups such as cyan groups and N⁻. With the prolongation of treatment in molten salts, the PHI structure would be broken and the heptazine units would be decomposed into triazine rings under the attack of Li⁺ and Cl⁻. Thereafter, the PTI structure was formed by a hex-merization reaction of the residual triazine-containing species. Notably, K⁺ is replaced by Li⁺ and Cl⁻ due to the reduction of negatively charged groups and shrinkage of central space in frameworks in PTI compared to PHI.

Inspired by the above findings, we could not help but wonder whether the same structural transformation would be observed if melamine was used as the precursor under the same condition. Fortunately, in the melamine-based system, the same structural transformation was observed upon increasing the duration of salt treatment, but with more severe synthesis conditions. As shown in Figure S4 (Supporting information), the aforementioned PHI nanorods can be observed after 4 h of calcination (Figure S4a,c, supporting information), whereas the transformation from the PHI phase to the PTI phase took as long as 60 h (Figure S4a,d, supporting information). The difference in duration can be explained that the urea-based melon exhibited porous flake-like morphology, thereby facilitating the reaction with salt melt.

2.4. Band Structures and Charge Carrier Behaviors

In order to clarify the structure–activity relationship, the electronic properties of the g-C₃N₄ materials were investigated through a combination of experiments and theoretical calculations. UV-visible diffuse reflectance spectra (DRS) were used to study the photo absorption properties of the samples. As shown in Figure 5a, melon and PHI exhibited visible light response, while the PTI sample barely absorbed visible light. Compared to melon and PTI, the absorption edge of PHI showed a distinct redshift, indicating a narrow band gap. It is also worth noting that the absorption in the UV region was progressively enhanced from melon to PTI to PHI, suggesting that the π-conjugation and electron delocalization increased gradually from melon to PTI to PHI [28], which is consistent with the structural changes illustrated in Figure 1a. Conduction band potentials of the samples were determined from Mott-Schottky (MS) plots, as shown in Figure S5 (Supporting Information) and Figure 5b. All MS plots exhibited positive slopes, indicating that they were all n-type semiconductors. It is generally accepted that the flat band potential of an n-type semiconductor is approximate to its conduction band (CB) potential. The flat band potentials of melon, PHI and PTI relative to saturated calomel electrode (SCE) can be obtained from the cross-intercepts of MS curves, so that their conduction band potentials are determined to be −1.49, −1.52 and −1.52 V, respectively. The valence band (VB) potentials of the samples can be obtained from the sum of the conduction band potential and the band gap, which are determined to be +1.41, +1.15 and +1.59 V (compared to SCE), respectively. For comparison with the electrode potentials, the SCE potentials has also been converted to potentials relative to normal hydrogen electrode (NHE), the values of which are shown in Figure 5c. Thereafter, the energy band structures can be obtained, which is shown schematically in Figure 5c. It is clear that the three samples exhibit similar conduction band potentials, indicating that the photogenerated electrons generated by them may have similar reduction activity. However, their valence band potentials differ somewhat, with the PTI sample exhibiting the most positive valence band potential, indicating that the photogenerated holes generated by it may have the strongest oxidation activity.

Based on the atomic arrangements of melon, PHI and PTI, we further analyzed the electron density of states (DOS) by means of theoretical DFT calculations. For comparison, the top of the VB was artificially positioned at the zero point. As shown in Figure 5d, the theoretical band gaps for melon, PHI and PTI were determined to be 2.42, 2.02, and 3.04 eV, respectively, which are smaller than those measured from the DRS experiments, possibly due to a well-known shortcoming of the DFT [29]. Whereas, the order of the theoretical bandgaps is consistent with the experiment, suggesting that the calculations are reasonable. In addition, the bottom of the CB of PHI was dominated by doped K, compared to melon and PTI, which may narrow its band gap and improve its photo absorption, thus enhancing its photocatalytic performance.

The differences in structure and composition of the three phases not only affect their electronic properties, but also influence the photogenerated charge separation and transfer efficiency. Steady-state and time-resolved photoluminescence (PL) spectra were collected to study the recombination behaviors of the photogenerated charge. As depicted in Figure 6a, a considerable emission intensity decrease was observed for PHI and PTI compared to melon, which denotes the radiative recombination of electron-hole pairs was efficiently suppressed [30]. Similarly, PHI and PTI exhibited fast decay in time-resolved PL (Figure 6b). The average lifetime of the pristine melon sample was calculated to be 6.74 ns, while that of PHI and PTI was reduced to 3.41 and 4.82 ns, respectively (Table S1). This could be associated with the decrease of exciton and the enhancement of charge transfer over the interface [31,32]. To gain a deeper understanding of the charge transportation behaviors of the g-C₃N₄ samples, measurements of transient photocurrent response and the electrochemical impedance spectra (EIS) were carried out. As displayed in Figure 6c, all electrodes exhibited sensitive and stable responses to the light irradiation, and the photo-response phenomenon could be entirely repeated in the ten cycles. PHI exhibited the highest photocurrent density, almost 2.5 times the PTI and 10 times the pristine melon, indicating the best charge separation and transfer performance, in accordance with PL analysis. As for EIS, similar trend was observed. The equivalent circuit is schematically illustrated in the inset of Figure 6d [33]. In this model, a smaller arc radius stands for a smaller charge transfer impedance [34]. PHI exhibited a much smaller arc radius than melon and PTI, especially under the light irradiation (Figure 6d), suggesting that charge transfer impedance is effectively reduced in PHI.

2.5. Photocatalytic Degradation

Tetracycline (TC), a widely used antibiotic to treat human infections, the abuse of which leads to water pollution, which further causes antibiotic resistance problems [35,36]. Phenol is an important intermediate for the production of phenolic resin, pharmaceuticals, pesticide, etc. [37,38]. Wastewater containing phenol can be found everywhere in industrial areas, and the drinking water contaminated by phenol with very low concentration could bring about serious diseases [37]. Dyes such as rhodamine b (RhB) are broadly used in textile and printing industries, and they are rated as one of the biggest polluters of water resources [39,40]. Because of the chemical steady of these organic pollutants, they are hard to be removed from water [35,37,39]. Therefore, such three organic compounds (TC, phenol and RhB), were selected as the model pollutants to study the degradation activity of the g-C₃N₄ samples. The calibration curves are shown in Figure S6 (Supporting information), which illustrate good linearity (R² > 0.999) over the test range (0.1~50 mg∙L⁻¹ for TC; 0.1~10 mg∙L⁻¹ for phenol and RhB). The photocatalytic efficiency was evaluated by the residual concentration of pollutants C/C₀, where C₀ and C represented the initial concentration and the concentration after t min degradation, respectively (Figure 7). Before the photocatalytic experiments, photocatalyst-pollutants solutions were continuously stirred in dark to reach the equilibrium of adsorption-desorption. It is clear that the residual concentrations barely changed after 30 min stirring, demonstrating that the equilibrium was reached at 30 min, which is shown in Figure S7 (Supporting information). The reaction rate constants were fitted by pseudo primary reaction kinetics, the results of which are presented in Figure S8 (Supporting information). As shown in Figure 7a, all samples were able to completely remove 50 mg·L⁻¹ TC within 60 min. Among them, PHI exhibited the best degradation efficiency with 90% removal within 10 min. The reaction rate constant was estimated to be 0.22 min⁻¹, which was approximately 4–6 times higher than that of the other samples (Figure S8a, Supporting Information). In the case of phenol, PHI demonstrated excellent photocatalytic degradation performance as well, with 80% of the contaminants removed within 60 min (Figure 7b). The reaction rate constant for PHI was calculated to be 0.024 min⁻¹, which was approximately 5–6 times higher than that of the other samples (Figure S8b, supporting information). As for RhB, each photocatalyst was able to completely remove it within 60 min, with PHI again exhibiting the best degradation activity (Figure 7c and Figure S8c, Supporting Information).

In addition, we examined the stability of PHI before and after the photocatalytic degradation of TC, which is shown in Figure S9 (Supporting Information). The XRD pattern and the FT-IR profile showed that the crystal structure of PHI had no distinct change after the photocatalytic reaction (Figure S9a,b, Supporting Information). However, the K-2p spectrum indicated that the atomic percentage of K decreased from 4.72% to 1.84% after the reaction (Figure S9c, supporting information), and the photocatalytic activity exhibited a decrease in the cyclic stability tests for the degradation of TC (Figure S9d, supporting information). The reduction of K content may be attributed to the proton exchange on the surface of PHI photocatalyst because of the long-time immersion in the water [11,23,41], thereby leading to the reduction of photocatalytic activity in cyclic tests. Benefiting from the synergistic effects of high crystallinity and surface modification by K⁺ and cyan groups, PHI exhibited the best photo absorption performance, efficient charge separation and transfer, and numerous reaction sites, improving the photocatalytic degradation efficiency. However, the reduction in photocatalytic activity in the cyclic tests indicates the obtained PHI sample is highly active but only an intermediate state.

To investigate the degradation mechanism of TC, we added multiple sacrificial agents to capture the reactive species (photogenerated electron e⁻, photogenerated hole h⁺, superoxide radical •O₂⁻ and H₂O₂) in photocatalytic experiments [34,42]. The •OH capture experiments were not performed because the VB potentials of all photocatalysts did not reach the OH⁻/•OH reaction potential [43]. As shown in Figure 7d, •O₂⁻ is considered to be the dominant working species for both melon and PTI samples. In addition, h⁺ also plays a crucial role in PTI, which may be due to the most positive VB potential. Different from melon and PTI, the main working species of PHI was not •O₂⁻, but H₂O₂ and e⁻, suggesting that the e⁻ of PHI tend to produce H₂O₂ rather than •O₂⁻ [3].

3. Materials and Methods

3.1. Materials

Urea (99%), potassium chloride (KCl, 99.8%), N-Methylpyrrolidone (NMP, 99%), sodium sulfate (Na₂SO₄, 99%), ammonium chloride (NH₄Cl, 99.5%), ammonia water (NH₃·H₂O, 25~28%), potassium ferricyanide (K₃[Fe(CN)₆], 99.5%) and 4-aminoantipyrine (99%) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Lithium chloride (LiCl, 99.0%) was purchased from Macklin Biochemical Technology Co., Ltd. (Shanghai, China). Poly (vinylidene fluoride) (PVDF, average Mw ~400,000), tetracycline (98%), phenol (99%) and rhodamine b (RhB, 99%) were purchased from Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). All chemicals were used without further purification.

3.2. Preparation of Photocatalysts

To synthesize melon, typically, 10 g urea was loaded in a capped alumina crucible and heated at 500 °C for 2 h with a ramping rate of 3 °C·min⁻¹ in muffle furnace. The resulting material was grounded in agate mortar, and the luminous yellow powder was marked as melon. The yield of melon was about 0.5 g.

The condensed g-C₃N₄ materials were synthesized by a molten salt-assisted procedure reported in the previous literature [11]. In detail, 0.2 g melon powder was grounded in agate mortar with 3.96 g KCl and 3.24 g LiCl to a uniform mixture. The mixture was then transferred into a capped alumina boat placed in the center of a quartz tube and heated at 550 °C with different treatment times (0.5 h, 1 h and 12 h) with a ramping rate of 30 °C·min⁻¹ under flow N₂ atmosphere (100 mL·min⁻¹) in tube furnace. The as-obtained products were rinsed with hot deionized water and centrifugation for 3 times to remove residual salts. Finally, the powders were dried overnight at 60 °C. The samples were marked as PHI (0.5 h), PHI/PTI (1 h) and PTI (12 h). The yield of condensed g-C₃N₄ was about 120 mg.

3.3. Characterization

XRD measurements were performed on D8-Advance diffractometer (Bruker, Germany) with Cu-Kα radiation (λ = 1.54059 Å). FT-IR spectra were measured on Nicolet 6700 spectrometer (Thermo, Waltham, MA, U.S.). Solid state ¹³C CP/MAS NMR spectra were recorded on 400M spectrometer (Bruker, Karlsruher, Germany). XPS profiles were measured on ESCALAB 250 X-ray photoelectron spectrometer (Thermo, America) equipped with an Al-Kα source (1486.6 eV photon energy, 300 W). The binding energy of all spectra were calibrated by the peak of adventitious C at 284.8 eV. DRS were collected on UV-2600i spectrophotometer (Shimadzu, Kyoto, Japan). UV-visible absorption spectra were collected on a UV-2550 spectrophotometer (Shimadzu, Japan). High performance liquid chromatography measurements were conducted on LC-20A (Shimadzu, Japan). PL spectra were recorded on FLSP-920 fluorescence spectrophotometer (Edinburgh, Livingston, UK). The morphologies and elemental distribution of as-prepared materials were observed by SUPRA55 SEM (Zeiss, Oberkochen, Germany). TEM observations were performed on Tecnai G² F20 microscope (FEI, Hillsboro, FL, USA). The BET measurements were characterized on ASAP 2460 (Micromeritics, Norcross, GA, USA).

3.4. Electrochemical Analysis

MS plots, transient photocurrent response (at 0 V) and EIS (at 0 V) measurements of the as-prepared samples were performed in a three-electrode system on the CHI 660D electrochemical workstation (Chenhua Limited, Shanghai, China). A 300 W Xe lamp was used as the light source in the illumination tests and the electrolyte was 0.2 M Na₂SO₄ solution. We chose platinum wire electrode as the counter electrode and SCE as the reference electrode. To make working electrodes, we first prepare Poly vinylidene fluoride (PVDF) solution through dissolving PVDF in NMP. Afterwards the as-prepared materials were grounded in agate mortar with PVDF solution. Finally, the suspension was coated on a F-doped tin oxide (FTO) glass and dried at 80 °C overnight.

3.5. Photocatalytic Degradation

The light source used in the experiments was a 300 W Xe lamp (PLS-SXE 300D, Perfectlight Sci&Tech, Beijing, China) equipped with a AM1.5 G filter, the emission spectrum of which is shown in Figure S10. The irradiation intensity was adjusted to 55 mW·cm⁻². In each photocatalytic experiment, 20 mg photocatalyst was dispersed in 40 mL TC (50 mg·L⁻¹), phenol (10 mg·L⁻¹) or RhB (10 mg·L⁻¹) solution and stirred in dark for 30 min to reach the balance of absorption and desorption. The photocatalysis lasted for 1 h, in which 3 mL suspension was taken at different time. The liquid was filtered by 0.22 μm filter (Nylon 66, 13 mm diameter). The performance of the filter that examined by the measurements of UV-vis absorption spectra is shown in Figure S11.

For the preparation of standard solutions, quantification of pollutants and capture of reactive species, see supporting information.

3.6. Theoretical Calculations

Model building and XRD simulations were performed on a Visualization for Electronic and STructural Analysis (VESTA) package [44]. All periodic density functional theory (DFT) calculations were performed on a Quantum Espresso (QE) 6.4.1 package [45] with PAW pseudopotentials [46] and PBE XC functional [47]. The wave function was expanded using plane waves with a cut-off energy of 60 Ry (approximates to 816 eV). Crystal information were drawn from the earlier literature [16,19,48] and fully optimized until the total energy and force converged to 10⁻⁴ Ry and 10⁻⁴ Ry·Bohr⁻¹ individually. DOS were calculated after a self-consistent calculation.

4. Conclusions

In summary, we presented new insights into the structural transformation mechanism of g-C₃N₄ in the molten salts and synthesized a highly efficient PHI photocatalyst for water treatment in this work. Melon type g-C₃N₄ undergoes a two-stage transformation in molten salts, in which melon transforms to PHI, an intermediate state, and then to a stable phase PTI. In addition, remarkable improvement has been observed for PHI over pristine melon and crystalline PTI in the photocatalytic degradation of organic pollutants. Through the combined study, we update the understanding of the structure–activity relationship of the three types of g-C₃N₄: melon, PHI and PTI. That is, high crystallinity as well as K⁺ and cyan groups modification were observed in PHI, which improves the photo absorption performance, charge separation and transfer efficiency and serves as reaction sites, thereby enhancing the photocatalytic activity. This work provides a deeper understanding for the molten salt-assisted synthesis as well as the ionic modification of g-C₃N₄, therefore providing new ideas for the design and synthesis of high-performance photocatalysts in the field of environment remediation and water treatment.

Author Contributions

Conceptualization, D.Z. and K.Z.; methodology, D.Z. and K.Z.; formal analysis, K.Z.; investigation, K.Z., C.L. and Q.L.; writing—original draft preparation, K.Z.; writing—review and editing, D.Z. and Z.M.; visualization, K.Z.; supervision, D.Z.; project administration, D.Z. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

The data presented in this study are available on request from the corresponding author.

Acknowledgments

The authors are grateful to Analysis and Measurement Center (Institute of metal research), Technical Support Division (Shenyang National Laboratory for Materials Science) and Shiyanjia Lab (www.shiyanjia.com, accessed on 19 March 2023) for their help in physicochemical studies of photocatalysts.

Conflicts of Interest

The authors declare no conflict of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Scheme

Figure 1. (a) Structures of melon, PHI and PTI and the proposed structural transformation pathway of g-C3N4 in molten salts. (b) Synthesis route of crystalline g-C3N4.

Figure 2. Crystal structure characterization of the g-C3N4 samples. (a): Powder XRD patterns of melon, PHI, PHI/PTI and PTI samples. (b–d): Comparison of the calculated and experimental XRD data of melon, PHI and PTI.

Figure 3. Chemical structures of the g-C3N4 samples. (a): FTIR spectra. (b): XPS survey profile. (c): N-1s High-Resolution XPS profile. (d): 13C CP/MAS NMR spectra.

Figure 4. Morphologies of the g-C3N4 samples. (a–d): SEM images of melon, PHI, PHI/PTI and PTI. The inset in (d) is the SEM image of PTI while at larger magnification. (e–h): Bright-field TEM images of melon, PHI, PHI/PTI and PTI samples. The inset in (h) is the HRTEM image of PTI.

Scheme 1. The proposed structural transformation pathway in molecular level.

Figure 5. Electronic properties of g-C3N4. (a): UV-vis diffuse reflectance spectra, and the inset is the corresponding bandgaps. (b): Mott-Schottky plots. (c): Band structure alignments. (d): Electron density of states calculated by density functional theory.

Figure 6. Photoelectric properties of the g-C3N4 samples. (a): Steady-state PL spectra. (b): Time-resolved PL spectra. (c): Transient photocurrent response. (d): EIS profile in dark and under light irradiation.

Figure 7. Photocatalytic efficiency of the g-C3N4 samples (a): Degradation of phenol (10 mg·L−1). (b): Degradation of TC (50 mg·L−1). (c): Degradation of RhB (10 mg·L−1). (d): Photocatalytic activity for the degradation of TC with the introduction of multiple scavengers.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/catal13040717/s1, Figure S1: Internal energy of melon, PHI and PTI calculated by DFT; Figure S2: High resolution XPS profiles of the g-C₃N₄ samples. (a): C-1s spectra. (b): K-2p spectra. (c): Cl-2p spectra. (d): Li-1s spectra; Figure S3: Microstructures of the g-C₃N₄ samples. (a): N₂ adsorption-desorption isotherms. (b): Pore distribution; Figure S4: Structures and morphologies of the MBCN-based samples. (a): XRD patterns. (b–d): SEM images of MBCN, MBCN-4h and MBCN-60h; Figure S5: MS plots of (a): melon, (b): PHI and (c): PTI; Table S1: Fitting parameters for the time-resolved PL of melon, PHI and PTI; Figure S6: (a): Chromatograph of TC (50 mg∙L⁻¹) standard solution. (b,c): UV-vis absorption spectra of phenol (10 mg∙L⁻¹, After colorimetric reaction) and RhB (10 mg∙L⁻¹) standard solution. (d–f): Calibration curves of TC, phenol, and RhB; Figure S7: Adsorption-desorption tests of photocatalyst-pollutants solution in dark; Figure S8: Rate constants for the photocatalytic degradation of (a): TC, (b): phenol and (c): RhB; Figure S9: Stability tests of PHI photocatalyst for the degradation of TC. (a–c): XRD patterns, FTIR spectra and K-2p spectra before and after the reaction. (d): Cyclic tests; Figure S10: Emission spectrum of PLS-SXE 300D Xe lamp equipped with AM1.5 G filter. (Obtained from Perfectlight Sci&Tech, China); Figure S11 UV-vis absorption spectra of pure water, RhB solution (10 mg∙L⁻¹), phenol solution (10 mg∙L⁻¹, after colorimetric reaction) as well as photocatalyst solution before and after filtration [49].

References

1. Chong, M.N.; Jin, B.; Chow, C.W.; Saint, C. Recent developments in photocatalytic water treatment technology: A review. Water Res.; 2010; 44, pp. 2997-3027. [DOI: https://dx.doi.org/10.1016/j.watres.2010.02.039] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20378145]

2. Das, R.; Vecitis, C.D.; Schulze, A.; Cao, B.; Ismail, A.F.; Lu, X.; Chen, J.; Ramakrishna, S. Recent advances in nanomaterials for water protection and monitoring. Chem. Soc. Rev.; 2017; 46, pp. 6946-7020. [DOI: https://dx.doi.org/10.1039/C6CS00921B] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28959815]

3. Nosaka, Y.; Nosaka, A.Y. Generation and Detection of Reactive Oxygen Species in Photocatalysis. Chem. Rev.; 2017; 117, pp. 11302-11336. [DOI: https://dx.doi.org/10.1021/acs.chemrev.7b00161] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28777548]

4. Dai, C.; Liu, B. Conjugated polymers for visible-light-driven photocatalysis. Energy Environ. Sci.; 2020; 13, pp. 24-52. [DOI: https://dx.doi.org/10.1039/C9EE01935A]

5. Wang, X.; Blechert, S.; Antonietti, M. Polymeric Graphitic Carbon Nitride for Heterogeneous Photocatalysis. ACS Catal.; 2012; 2, pp. 1596-1606. [DOI: https://dx.doi.org/10.1021/cs300240x]

6. Zhang, G.; Lan, Z.A.; Wang, X. Conjugated Polymers: Catalysts for Photocatalytic Hydrogen Evolution. Angew. Chem. Int. Ed. Engl.; 2016; 55, pp. 15712-15727. [DOI: https://dx.doi.org/10.1002/anie.201607375]

7. Wang, X.; Maeda, K.; Thomas, A.; Takanabe, K.; Xin, G.; Carlsson, J.M.; Domen, K.; Antonietti, M. A metal-free polymeric photocatalyst for hydrogen production from water under visible light. Nat. Mater.; 2009; 8, pp. 76-80. [DOI: https://dx.doi.org/10.1038/nmat2317] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/18997776]

8. Ong, W.J.; Tan, L.L.; Ng, Y.H.; Yong, S.T.; Chai, S.P. Graphitic Carbon Nitride (g-C3N4)-Based Photocatalysts for Artificial Photosynthesis and Environmental Remediation: Are We a Step Closer to Achieving Sustainability?. Chem. Rev.; 2016; 116, pp. 7159-7329. [DOI: https://dx.doi.org/10.1021/acs.chemrev.6b00075]

9. Lotsch, B.V.; Doblinger, M.; Sehnert, J.; Seyfarth, L.; Senker, J.; Oeckler, O.; Schnick, W. Unmasking melon by a complementary approach employing electron diffraction, solid-state NMR spectroscopy, and theoretical calculations-structural characterization of a carbon nitride polymer. Chem. Eur. J.; 2007; 13, pp. 4969-4980. [DOI: https://dx.doi.org/10.1002/chem.200601759]

10. Teter, D.M.; Hemley, R.J. Low-compressibility carbon nitrides. Science; 1996; 271, pp. 53-55. [DOI: https://dx.doi.org/10.1126/science.271.5245.53]

11. Lin, L.; Ou, H.; Zhang, Y.; Wang, X. Tri-s-triazine-Based Crystalline Graphitic Carbon Nitrides for Highly Efficient Hydrogen Evolution Photocatalysis. ACS Catal.; 2016; 6, pp. 3921-3931. [DOI: https://dx.doi.org/10.1021/acscatal.6b00922]

12. Xia, P.; Antonietti, M.; Zhu, B.; Heil, T.; Yu, J.; Cao, S. Designing Defective Crystalline Carbon Nitride to Enable Selective CO2 Photoreduction in the Gas Phase. Adv. Funct. Mater.; 2019; 29, 1900093. [DOI: https://dx.doi.org/10.1002/adfm.201900093]

13. Wang, L.; Hong, Y.; Liu, E.; Wang, Z.; Chen, J.; Yang, S.; Wang, J.; Lin, X.; Shi, J. Rapid polymerization synthesizing high-crystalline g-C₃N₄ towards boosting solar photocatalytic H2 generation. Int. J. Hydrogen Energy; 2020; 45, pp. 6425-6436. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2019.12.168]

14. Huang, L.; Hu, Z.; Jin, H.; Wu, J.; Liu, K.; Xu, Z.; Wan, J.; Zhou, H.; Duan, J.; Hu, B. et al. Salt-Assisted Synthesis of 2D Materials. Adv. Funct. Mater.; 2020; 30, 1908486. [DOI: https://dx.doi.org/10.1002/adfm.201908486]

15. Liu, M.; Wei, C.; Zhuzhang, H.; Zhou, J.; Pan, Z.; Lin, W.; Yu, Z.; Zhang, G.; Wang, X. Fully Condensed Poly (Triazine Imide) Crystals: Extended pi-Conjugation and Structural Defects for Overall Water Splitting. Angew. Chem. Int. Ed. Engl.; 2022; 61, e202113389. [DOI: https://dx.doi.org/10.1002/anie.202113389] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34750939]

16. Wirnhier, E.; Doblinger, M.; Gunzelmann, D.; Senker, J.; Lotsch, B.V.; Schnick, W. Poly(triazine imide) with intercalation of lithium and chloride ions [(C3N3)2(NH(x)Li(1-x))3LiCl]: A crystalline 2D carbon nitride network. Chem. Eur. J.; 2011; 17, pp. 3213-3221. [DOI: https://dx.doi.org/10.1002/chem.201002462]

17. Liao, C.Z.; Lau, V.W.; Su, M.; Ma, S.; Liu, C.; Chang, C.K.; Sheu, H.S.; Zhang, J.; Shih, K. Unraveling the Structure of the Poly(triazine imide)/LiCl Photocatalyst: Cooperation of Facile Syntheses and a Low-Temperature Synchrotron Approach. Inorg. Chem.; 2019; 58, pp. 15880-15888. [DOI: https://dx.doi.org/10.1021/acs.inorgchem.9b02287]

18. Lin, L.; Wang, C.; Ren, W.; Ou, H.; Zhang, Y.; Wang, X. Photocatalytic overall water splitting by conjugated semiconductors with crystalline poly(triazine imide) frameworks. Chem. Sci.; 2017; 8, pp. 5506-5511. [DOI: https://dx.doi.org/10.1039/C7SC00900C]

19. Fina, F.; Callear, S.K.; Carins, G.M.; Irvine, J.T.S. Structural Investigation of Graphitic Carbon Nitride via XRD and Neutron Diffraction. Chem. Mater.; 2015; 27, pp. 2612-2618. [DOI: https://dx.doi.org/10.1021/acs.chemmater.5b00411]

20. Chen, Z.; Savateev, A.; Pronkin, S.; Papaefthimiou, V.; Wolff, C.; Willinger, M.G.; Willinger, E.; Neher, D.; Antonietti, M.; Dontsova, D. “The Easier the Better” Preparation of Efficient Photocatalysts-Metastable Poly(heptazine imide) Salts. Adv. Mater.; 2017; 29, 1700555. [DOI: https://dx.doi.org/10.1002/adma.201700555]

21. Bojdys, M.J.; Muller, J.O.; Antonietti, M.; Thomas, A. Ionothermal synthesis of crystalline, condensed, graphitic carbon nitride. Chem. Eur. J.; 2008; 14, pp. 8177-8182. [DOI: https://dx.doi.org/10.1002/chem.200800190] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/18663712]

22. Li, Y.-Y.; Zhou, B.-X.; Zhang, H.-W.; Ma, S.-F.; Huang, W.-Q.; Peng, W.; Hu, W.; Huang, G.-F. Doping-induced enhancement of crystallinity in polymeric carbon nitride nanosheets to improve their visible-light photocatalytic activity. Nanoscale; 2019; 11, pp. 6876-6885. [DOI: https://dx.doi.org/10.1039/C9NR00229D]

23. Savateev, A.; Pronkin, S.; Epping, J.D.; Willinger, M.G.; Wolff, C.; Neher, D.; Antonietti, M.; Dontsova, D. Potassium Poly(heptazine imides) from Aminotetrazoles: Shifting Band Gaps of Carbon Nitride-like Materials for More Efficient Solar Hydrogen and Oxygen Evolution. ChemCatChem; 2017; 9, pp. 167-174. [DOI: https://dx.doi.org/10.1002/cctc.201601165]

24. Schwinghammer, K.; Mesch, M.B.; Duppel, V.; Ziegler, C.; Senker, J.; Lotsch, B.V. Crystalline carbon nitride nanosheets for improved visible-light hydrogen evolution. J. Am. Chem. Soc.; 2014; 136, pp. 1730-1733. [DOI: https://dx.doi.org/10.1021/ja411321s] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24432762]

25. Jürgens, B.; Irran, E.; Senker, J.; Kroll, P.; Müller, H.; Schnick, W. Melem (2,5,8-Triamino-tri-s-triazine), an Important Intermediate during Condensation of Melamine Rings to Graphitic Carbon Nitride:  Synthesis, Structure Determination by X-ray Powder Diffractometry, Solid-State NMR, and Theoretical Studies. J. Am. Chem. Soc.; 2003; 125, pp. 10288-10300. [DOI: https://dx.doi.org/10.1021/ja0357689]

26. Lin, L.; Lin, Z.; Zhang, J.; Cai, X.; Lin, W.; Yu, Z.; Wang, X. Molecular-level insights on the reactive facet of carbon nitride single crystals photocatalysing overall water splitting. Nat. Catal.; 2020; 3, pp. 649-655. [DOI: https://dx.doi.org/10.1038/s41929-020-0476-3]

27. Lee, T.-G.; Kang, H.-J.; Kim, J.-H.; Suzuki, N.; Fujishima, A.; Choi, M.; Jun, Y.-S. Eutectic iodide-based salt as a melem-to-PTI conversion stopping agent for all-in-one graphitic carbon nitride. Appl. Catal. B; 2021; 294, 120222. [DOI: https://dx.doi.org/10.1016/j.apcatb.2021.120222]

28. Zhang, G.; Savateev, A.; Zhao, Y.; Li, L.; Antonietti, M. Advancing the n → π* electron transition of carbon nitride nanotubes for H2 photosynthesis. J. Mater. Chem. A; 2017; 5, pp. 12723-12728. [DOI: https://dx.doi.org/10.1039/C7TA03777E]

29. Masoumi, S.; Nadimi, E.; Hossein-Babaei, F. Electronic properties of Ag-doped ZnO: DFT hybrid functional study. Phys. Chem. Chem. Phys.; 2018; 20, pp. 14688-14693. [DOI: https://dx.doi.org/10.1039/C8CP01578C]

30. Bai, S.; Li, H.; Guan, Y.; Jiang, S. The enhanced photocatalytic activity of CdS/TiO₂ nanocomposites by controlling CdS dispersion on TiO₂ nanotubes. Appl. Surf. Sci.; 2011; 257, pp. 6406-6409. [DOI: https://dx.doi.org/10.1016/j.apsusc.2011.02.007]

31. Qiu, C.; Xu, Y.; Fan, X.; Xu, D.; Tandiana, R.; Ling, X.; Jiang, Y.; Liu, C.; Yu, L.; Chen, W. et al. Highly Crystalline K-Intercalated Polymeric Carbon Nitride for Visible-Light Photocatalytic Alkenes and Alkynes Deuterations. Adv. Sci.; 2019; 6, 1801403. [DOI: https://dx.doi.org/10.1002/advs.201801403] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30643725]

32. Zhang, G.; Liu, M.; Heil, T.; Zafeiratos, S.; Savateev, A.; Antonietti, M.; Wang, X. Electron Deficient Monomers that Optimize Nucleation and Enhance the Photocatalytic Redox Activity of Carbon Nitrides. Angew. Chem. Int. Ed. Engl.; 2019; 58, pp. 14950-14954. [DOI: https://dx.doi.org/10.1002/anie.201908322] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31424624]

33. Pan, Z.; Zhao, M.; Zhuzhang, H.; Zhang, G.; Anpo, M.; Wang, X. Gradient Zn-Doped Poly Heptazine Imides Integrated with a van der Waals Homojunction Boosting Visible Light-Driven Water Oxidation Activities. ACS Catal.; 2021; 11, pp. 13463-13471. [DOI: https://dx.doi.org/10.1021/acscatal.1c03687]

34. Xiao, J.; Liu, Q.; Song, M.; Li, X.; Li, Q.; Shang, J.K. Directing photocatalytic pathway to exceedingly high antibacterial activity in water by functionalizing holey ultrathin nanosheets of graphitic carbon nitride. Water Res.; 2021; 198, 117125. [DOI: https://dx.doi.org/10.1016/j.watres.2021.117125] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33878660]

35. Xie, Z.; Feng, Y.; Wang, F.; Chen, D.; Zhang, Q.; Zeng, Y.; Lv, W.; Liu, G. Construction of carbon dots modified MoO₃/g-C₃N₄ Z-scheme photocatalyst with enhanced visible-light photocatalytic activity for the degradation of tetracycline. Appl. Catal. B; 2018; 229, pp. 96-104. [DOI: https://dx.doi.org/10.1016/j.apcatb.2018.02.011]

36. Watkinson, A.J.; Murby, E.J.; Costanzo, S.D. Removal of antibiotics in conventional and advanced wastewater treatment: Implications for environmental discharge and wastewater recycling. Water Res.; 2007; 41, pp. 4164-4176. [DOI: https://dx.doi.org/10.1016/j.watres.2007.04.005] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17524445]

37. Aksu, Z.; Gönen, F. Biosorption of phenol by immobilized activated sludge in a continuous packed bed: Prediction of breakthrough curves. Process Biochem.; 2004; 39, pp. 599-613. [DOI: https://dx.doi.org/10.1016/S0032-9592(03)00132-8]

38. Busca, G.; Berardinelli, S.; Resini, C.; Arrighi, L. Technologies for the removal of phenol from fluid streams: A short review of recent developments. J. Hazard. Mater.; 2008; 160, pp. 265-288. [DOI: https://dx.doi.org/10.1016/j.jhazmat.2008.03.045]

39. Bizani, E.; Fytianos, K.; Poulios, I.; Tsiridis, V. Photocatalytic decolorization and degradation of dye solutions and wastewaters in the presence of titanium dioxide. J. Hazard. Mater.; 2006; 136, pp. 85-94. [DOI: https://dx.doi.org/10.1016/j.jhazmat.2005.11.017]

40. Akpan, U.G.; Hameed, B.H. Parameters affecting the photocatalytic degradation of dyes using TiO₂-based photocatalysts: A review. J. Hazard. Mater.; 2009; 170, pp. 520-529. [DOI: https://dx.doi.org/10.1016/j.jhazmat.2009.05.039]

41. Kröger, J.; Jiménez-Solano, A.; Savasci, G.; Lau, V.W.h.; Duppel, V.; Moudrakovski, I.; Küster, K.; Scholz, T.; Gouder, A.; Schreiber, M.L. et al. Morphology Control in 2D Carbon Nitrides: Impact of Particle Size on Optoelectronic Properties and Photocatalysis. Adv. Funct. Mater.; 2021; 31, 2102468. [DOI: https://dx.doi.org/10.1002/adfm.202102468]

42. Wang, W.; Fang, J.; Shao, S.; Lai, M.; Lu, C. Compact and uniform TiO₂@g-C₃N₄ core-shell quantum heterojunction for photocatalytic degradation of tetracycline antibiotics. Appl. Catal. B; 2017; 217, pp. 57-64. [DOI: https://dx.doi.org/10.1016/j.apcatb.2017.05.037]

43. Burns, J.M.; Cooper, W.J.; Ferry, J.L.; King, D.W.; DiMento, B.P.; McNeill, K.; Miller, C.J.; Miller, W.L.; Peake, B.M.; Rusak, S.A. et al. Methods for reactive oxygen species (ROS) detection in aqueous environments. Aquat. Sci.; 2012; 74, pp. 683-734. [DOI: https://dx.doi.org/10.1007/s00027-012-0251-x]

44. Momma, K.; Izumi, F. VESTA 3 for three-dimensional visualization of crystal, volumetric and morphology data. J. Appl. Crystallogr.; 2011; 44, pp. 1272-1276. [DOI: https://dx.doi.org/10.1107/S0021889811038970]

45. Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.; Cavazzoni, C.; Ceresoli, D.; Chiarotti, G.L.; Cococcioni, M.; Dabo, I. et al. QUANTUM ESPRESSO: A modular and open-source software project for quantum simulations of materials. J. Phys. Condens. Matter; 2009; 21, 395502. [DOI: https://dx.doi.org/10.1088/0953-8984/21/39/395502] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21832390]

46. Blöchl, P.E. Projector augmented-wave method. Phys. Rev. B; 1994; 50, pp. 17953-17979. [DOI: https://dx.doi.org/10.1103/PhysRevB.50.17953]

47. Perdew, J.P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett.; 1996; 77, pp. 3865-3868. [DOI: https://dx.doi.org/10.1103/PhysRevLett.77.3865] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/10062328]

48. Doblinger, M.; Lotsch, B.V.; Wack, J.; Thun, J.; Senker, J.; Schnick, W. Structure elucidation of polyheptazine imide by electron diffraction—A templated 2D carbon nitride network. Chem. Commun.; 2009; 12, pp. 1541-1543. [DOI: https://dx.doi.org/10.1039/b820032g]

49. Lacoste, R.; Venable, S.; Stone, J. Modified 4-aminoantipyrine colorimetric method for phenols. Application to acrylic monomer. Anal. Chem.; 1959; 31, pp. 1246-1249.