Synthesis of ternary g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube composite photocatalysts for the decolorization of dyes under visible light and direct sunlight irradiation

2.1 Synthesis of Bi₂MoO₆ nanoplate photocatalyst

Bi₂MoO₆ nanoplates were synthesized using a hydrothermal method [38]. A mixture of Bi(NO₃)₃·5H₂O and Na₂MoO₄·2H₂O was dissolved in deionized water, and then, the pH was adjusted to 6.0 by adding HCl. After that, it was stirred for 1 h at room temperature. The mixture was transferred to a Teflon-lined stainless steel autoclave and heated at 453 K for 20 h in an electric oven. After 24 h, the autoclave was cooled to room temperature. The obtained yellowish precipitate was centrifuged, washed with distilled water, and dried at 353 K for 24 h.

2.2 Synthesis of g-C₃N₄ photocatalyst

g-C₃N₄ was prepared through the direct calcination of urea [39]. Urea was placed in a crucible, which was covered with a lid and heated at 773 K for 3.5 h. Subsequently, the crucible was cooled to room temperature, and the obtained yellow g-C₃N₄ product was collected and stored for further composite synthesis.

2.3 Synthesis of TiO₂ nanotube photocatalyst

The TiO₂ nanotube was synthesized using TiO₂ P25 nanoparticles through a hydrothermal method [40]. First, a commercial TiO₂ P25 was added to a 10-N NaOH solution, and it was stirred vigorously at room temperature and transferred to a Teflon-lined stainless steel autoclave and heated at 403 K for 48 h. Afterward, the autoclave was allowed to cool to room temperature, and the obtained nanotube was centrifuged and washed with distilled water until the pH of the filtrate was approximately 6–7. Then, the nanotube was treated with 0.1 M HCl solution, which was later stirred overnight at room temperature. Thereafter, the nanotube was centrifuged, washed with distilled water, and dried in an oven at 343 K for 12 h. The dried nanotube was calcined in a tubular furnace at 523 K for 2 h and used for further composite synthesis.

2.4 Synthesis of binary (g-C₃N₄/Bi₂MoO₆) and ternary (g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube) composite photocatalysts

g-C₃N₄/Bi₂MoO₆ and g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube composites were synthesized using a wet impregnation method. Binary composites (g-C₃N₄/Bi₂MoO₆) with different weight percentages of g-C₃N₄ were synthesized. Initially, an appropriate amount of Bi₂MoO₆ was dispersed in 100 ml of methanol under sonication. After that, the required amount of g-C₃N₄ was added and stirred for 24 h at ambient temperature. After the complete evaporation of methanol, the obtained g-C₃N₄/Bi₂MoO₆ composites were dried at 353 K for 12 h and, then, calcined at 573 K for 1 h. The prepared binary composites (g-C₃N₄/Bi₂MoO₆) with different weight percentages of g-C₃N₄ were abbreviated as x%g-C₃N₄/Bi₂MoO₆(x=5%, 10%, 30%, 50% and 70%). In the preparation of ternary nanocomposites (g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube), different amounts of TiO₂ nanotubes were added to the binary composites, and the same aforementioned procedure was followed.

2.5 Characterization

X-ray diffraction (XRD) analysis was performed using a Bruker D8 Advance Eco diffractometer (Bruker Corporation, San Jose, CA, USA) to identify the crystal structure of the photocatalysts. The purity of the synthesized photocatalysts and their elemental constituents were determined using energy-dispersive X-ray (EDX) spectroscopy (Hitachi Co., Tokyo, Japan). The morphologies of the photocatalysts were examined using a Hitachi S-4800 scanning electron microscope (SEM) (Hitachi Co., Tokyo, Japan). The Fourier transform infrared (FT-IR) spectra of the samples were recorded using a Jasco FT-IR-6500 Fourier transform infrared spectrometer (JASCO International Co. Ltd., Tokyo, Japan) over the wavenumber range 400–4000 cm⁻¹ (KBr pellets were used).

2.6 Photocatalytic activity

The photocatalytic activity of the synthesized unitary, binary, and ternary composites was evaluated by studying the decolorization behavior of an aqueous dye solution in the presence of visible light (Figure 1A) and direct sunlight irradiation (Figure 1B). Reactive black 5 and methylene blue dyes were chosen as target pollutants for decolorization studies. In the decolorization experiment, 0.5 g/l of photocatalyst and 100 ml of dye solution (10 mg/l) were placed in a beaker, and the photocatalyst was dispersed under sonication for 2 min. Prior to light irradiation, the reaction mixture was stirred under dark atmospheric conditions to obtain an adsorption-desorption equilibrium between the photocatalyst and the dye solution. Afterward, the reaction mixture was exposed to light irradiation, and the dye solution was withdrawn at predetermined time intervals for absorbance measurements by using a UV-vis spectrophotometer (Genesys 10s UV-Vis spectrophotometer). Before the UV analysis, the catalyst particles were separated through centrifugation; then, the absorbance for reactive black 5 and methylene blue dyes was measured at λ_max=595 nm and λ_max=665 nm, respectively. The concentration of the decolored dye solutions was determined using their calibration curves, and the percentage decolorization was calculated.

Figure 1:

Experimental setup for the photocatalytic decolorization of dyes using synthesized photocatalysts under (A) visible light and (B) direct sunlight irradiation.

3.1 X-ray diffraction and Fourier transform infrared analyses

The phase purity and crystallinity of the synthesized composites were determined using XRD, and the results are shown in Figure 2. Figure 2A presents the XRD patterns of the following: (1) synthesized g-C₃N₄, Bi₂MoO₆, and TiO₂ nanotube photocatalysts; (2) inorganic crystal structure database (ICSD) for Bi₂MoO₆ and TiO₂ nanotube; and (3) commercial TiO₂ P25. The synthesized g-C₃N₄ reveals two distinct diffraction peaks at 13.04° and 27.40°, which correspond to (100) and (002) diffraction planes of the graphite-like structure of g-C₃N₄, respectively. Moreover, the peak at 13.04° is attributed to the (100) plane of the tri-s-triazine unit of g-C₃N₄; the peak at 27.40° corresponds to the characteristic inter-planar stacking of the conjugated aromatic systems and the interlayer structural packing of the (002) plane of graphite-like materials [29]. For Bi₂MoO₆, strong diffraction peaks at 2θ=11.00°, 28.55°, 32.66°, 33.24°, 36.17°, 47.42°, 55.61°, 56.43° and 58.54° can be assigned to (0 2 0), (1 3 1), (2 0 0), (2 1 0), (1 5 1), (2 6 0), (3 3 1), (1 9 1) and (2 6 2) crystal planes, respectively. For the TiO₂ nanotube, the diffraction peaks at 2θ=25.4°, 39°, 48.05°, 55° and 63° are indexed to the (1 0 1), (1 1 2), (2 0 0), (2 1 1) and (2 0 4) planes of the anatase phase of TiO₂, respectively. The XRD patterns of the synthesized Bi₂MoO₆ and TiO₂ nanotube match their corresponding ICSD data well. Furthermore, no peaks correspond to layered titanates, and a change in the anatase phase was observed after the hydrothermal treatment of TiO₂ nanocrystalline powder. Therefore, the XRD pattern validates the complete transformation of TiO₂ nanoparticles to nanotube, and the TiO₂ nanotube comprised the anatase phase only [40]. Figure 2B presents the XRD patterns of the following photocatalysts: (1) g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube, (2) Bi₂MoO₆/TiO₂ nanotube, (3) g-C₃N₄/Bi₂MoO₆ and (4) g-C₃N₄/TiO₂ nanotube. Peaks corresponding to g-C₃N₄ in binary (g-C₃N₄/Bi₂MoO₆) and ternary (g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube) composites were not observed. This may be due to the diffraction peak of g-C₃N₄ at 27.40° being superimposed with the Bi₂MoO₆ diffraction peak at 28.55°. Similarly, the TiO₂ nanotube peak was not observed in the ternary composite because of the low concentration of TiO₂ nanotube in the composite. However, the g-C₃N₄ presence was clearly observed at 2θ=13° and 27° in the g-C₃N₄/TiO₂ nanotube composite. From the XRD patterns, we conclude that unitary, binary, and ternary composite photocatalysts were successfully synthesized, and the introduction of g-C₃N₄ and TiO₂ nanotube did not change the crystal structure of Bi₂MoO₆ in the ternary composite.

Figure 2:

XRD patterns of (A) g-C₃N₄, Bi₂MoO₆, and TiO₂ nanotube and (B) binary and ternary composite photocatalysts.

Figure 3 shows the FT-IR spectra of TiO₂ P25 and the TiO₂ nanotube. The bands at 3420 and 1629 cm⁻¹ in the TiO₂ nanotube are attributed to the stretching and bending vibrations of the water molecule hydroxyl groups present on the surface of the TiO₂ nanotube. The higher intensity of the hydroxyl group peaks in the TiO₂ nanotube, relative to that in TiO₂ P25, and indicates that the surface of the TiO₂ nanotube contains a higher concentration of hydroxyl groups. This enhancement in the surface hydroxyl groups favors the adsorption capability of the TiO₂ nanotube and enhances its photocatalytic decolorization efficiency. Figure 4 displays the FT-IR spectra of g-C₃N₄, g-C₃N₄/Bi₂MoO₆, g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube and Bi₂MoO₆. In the spectrum of g-C₃N₄, the strong peaks at 3420 and 3200 cm⁻¹ can be ascribed to the absorbed water molecules and stretching mode of N–H, respectively. Four peaks at 1250, 1325, 1420 and 1572 cm⁻¹ observed in g-C₃N₄ are related to the typical stretching modes of the CN heterocycle. For Bi₂MoO₆, the peaks at 400–900 cm⁻¹ originate from Bi–O, Mo–O stretching and Mo–O–Mo bridging stretching modes. The distinct absorption peaks of g-C₃N₄ and Bi₂MoO₆ were also observed in the composites of g-C₃N₄/Bi₂MoO₆ and g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube.

Figure 3:

FT-IR spectra of TiO₂ P25 and TiO₂ nanotube photocatalysts.

Figure 4:

FT-IR spectra of g-C₃N₄, Bi₂MoO₆, g-C₃N₄/Bi₂MoO₆, and g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube photocatalysts.

3.2 Morphology

The morphological properties of the synthesized photocatalysts were characterized using SEM at different magnifications, and the results are shown in Figure 5. The SEM image of the Bi₂MoO₆ photocatalyst (Figure 5A) shows a plate-like structure in which the length was in the range of nanometer scale. The SEM image of g-C₃N₄ (Figure 5B) illustrates that it exhibits a thin sheet-like structure with wrinkled surface and irregular shapes. Figure 5C describes that the commercial TiO₂ P25 particles were spherical in shape. The SEM image of the synthesizedTiO₂ nanotube (Figure 5D) indicates that several long fibers were assembled together, with the diameter in the range of 10–30 nm. A binary composite photocatalyst (g-C₃N₄/Bi₂MoO₆) reveals a structure comprising two layers of g-C₃N₄ nanosheets and Bi₂MoO₆ nanoplates (Figure 5E). For the ternary composite photocatalyst (g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube), a three-layered structure is indicated in Figure 5F. The intimate contact between these semiconductors could enhance the charge carrier separation and transfer rate and photocatalytic decolorization efficiency.

Figure 5:

SEM images of (A) Bi₂MoO₆, (B) g-C₃N₄, (C) TiO₂ P25, (D) TiO₂ nanotube, (E) g-C₃N₄/Bi₂MoO₆, and (F) g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube photocatalysts.

3.3 Energy-dispersive X-ray spectroscopic analysis

The presence of Bi₂MoO₆ and g-C₃N₄ in the ternary composite was confirmed by EDX analysis. Figure 6 illustrates the EDX spectrum of a ternary composite photocatalyst (g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube). It indicates only the presence of Bi, Mo, O, C, N and Ti peaks in the composite, without the presence of any other characteristic peaks. This result confirms the successful synthesis of the ternary g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube photocatalyst; it contains no impurities, which supports the powder XRD and FT-IR results.

Figure 6:

EDX spectra of a ternary composite photocatalyst (g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube).

3.4 Photocatalytic performance

Initially, the photocatalytic performance of unitary photocatalysts (g-C₃N₄, TiO₂ nanotube, Bi₂MoO₆, and P25 TiO₂) was investigated by decolorizing reactive black 5 with 0.5 g/l of catalyst under visible light irradiation, and the results are shown in Figure 7A. The results reveal that no decolorization was observed under blank conditions (without a catalyst). For Bi₂MoO₆ and P25 TiO₂, 18.9% and 51.68% of reactive black 5 was decolorized after 5 h of reaction. For g-C₃N₄ and the TiO₂ nanotube, the percentage decolorization of reactive black 5 was enhanced to 88.31% and 96.14%, respectively. This significant enhancement in the photocatalytic activity is due to the layered and tubular structure of g-C₃N₄ and TiO₂, which enhances the adsorption quantity of reactive black 5 on these photocatalysts. Among the photocatalysts, the TiO₂ nanotube exhibited the highest adsorption ability for reactive black 5, which leads to enriched photocatalytic decolorization efficiency.

Figure 7:

Photocatalytic decolorization of reactive black 5 caused by (A) unitary, (B) binary composite, and (C) ternary composite photocatalysts under visible light irradiation.

The photoactivity of the binary photocatalyst (g-C₃N₄/Bi₂MoO₆) with different amounts of g-C₃N₄ (5–70%) was evaluated in terms of the decolorization of reactive black 5, and the results are shown in Figure 7B. The results reveal that the percentage decolorization of reactive black 5 increased from 18.9% (Bi₂MoO₆) to 27.54%, 91.20% and 94.09% for 5%, 10% and 30% loading of g-C₃N₄, respectively; increasing further the g-C₃N₄ loading to 50% and 70% decreased the percentage decolorization of reactive black 5 to 93.81% and 88.27%, respectively. The decreased photocatalytic decolorization efficiency at higher g-C₃N₄ loadings was due to the shielding effect of the added g-C₃N₄, which led to a decrease in the light intensity reaching the catalyst surface and a lowering of the generation of photogenerated charge carriers. Therefore, 30% g-C₃N₄ loading is the optimum amount, which was used for further ternary composite synthesis. The photocatalytic activity of the ternary composite with different loadings of the TiO₂ nanotube (5–70%) was determined, and the results are shown in Figure 7C. The results demonstrate that the ternary composite (5% TiO₂ nanotube loaded with 30% g-C₃N₄/Bi₂MoO₆) exhibits a higher percentage decolorization (99.02%) of reactive black 5; increasing further the TiO₂ nanotube loading to 10%, 30%, 50% and 70% resulted in a decrease in the dye percentage decolorization. From the results, it is observed that 30% g-C₃N₄/Bi₂MoO₆/5%TiO₂ nanotube is the optimum catalyst for the highest percentage decolorization of reactive black 5. We evaluated the photocatalytic decolorization efficiency of this optimum catalyst on the cationic methylene blue; the results reveal that 73.61% of methylene blue was decolorized.

In addition to the aforementioned initial decolorization studies, we also studied the effect of different parameters, such as the catalyst amount, initial dye concentration, pH, and temperature, on the decolorization of reactive black 5 and methylene blue dyes. To study the effect of the photocatalyst loading on the decolorization of reactive black 5 and methylene blue dyes, we varied the photocatalyst amount from 0.1 to 0.5 g/l while keeping the dye concentration constant (10 ppm), and the results are shown in Figure 8A and B. The results demonstrate that the percentage decolorization of reactive black 5 and methylene blue dyes increased with the catalyst amount. The highest percentage decolorization was observed for 0.5 g/l of catalyst, suggesting that 0.5 g/l is the optimum amount; 99.02% and 73.61% of reactive black 5 and methylene blue were decolorized, respectively.

Figure 8:

Effect of photocatalyst on the decolorization of (A) reactive black 5 and (B) methylene blue dyes.

The effect of the initial dye concentration on the decolorization of dyes was determined by varying the dye concentration from 10 to 50 ppm while keeping the catalyst amount (0.5 g/l) constant, and the results are shown in Figure 9A and B. The results reveal that the percentage decolorization of dyes decreased with increasing initial dye concentration. The percentage decolorization of reactive black 5 was decreased from 99.02% to 18.93% by increasing the initial dye concentration from 10 to 50 ppm (Figure 9A). Similarly, the percentage decolorization of methylene blue was decreased from 73.61% to 1.59% (Figure 9B). This is due to the increased amount of the dye adsorption on the catalyst surface, which led to only fewer photons reaching the photocatalyst surface and a decrease in the concentration of reactive ˙OH and O₂˙⁻ radicals being produced, thereby, declining the photocatalytic decolorization efficiency.

Figure 9:

Effect of the initial dye concentration on the decolorization of (A) reactive black 5 and (B) methylene blue dyes.

The effect of pH on the decolorization of dyes (reactive black 5 and methylene blue) was examined by adjusting the initial pH of the dye solution from 3 to 12 while keeping the catalyst amount (0.5 g/l) and dye concentration (10 ppm) constant. The results reveal that 11 was the optimum pH for the highest percentage decolorization of dye molecules (Figure 10A and B). The photocatalyst surface charge changed with a change in pH of the initial dye solution because of the amphoteric behavior of a semiconducting photocatalyst. In acidic solutions (pH<5), the dye photodegradation is retarded by the high concentration of protons, resulting in lower degradation efficiency. However, in alkaline solutions (pH>10), the presence of hydroxyl ions neutralizes the acidic end products of the photodegradation reaction.

Figure 10:

Effect of pH on the decolorization of (A) reactive black 5 and (B) methylene blue dyes.

The effect of temperature on the decolorization of reactive black 5 and methylene blue dyes was investigated at 35°C, 45°C and 55°C, with the photocatalyst amount and dye concentration maintained at 0.5 g/l and 10 ppm, respectively. The results are shown in Figure 11A and B. The results reveal that a higher percentage decolorization of reactive black 5 and methylene blue dyes was observed at 45°C. This may be due to the increased adsorption of pollutants on the catalytic surface at a higher reaction temperature, which led to a higher percentage decolorization of dyes. Furthermore, the increase in temperature improved the decolorization reaction and overcame the electron-hole (e⁻–h⁺) recombination. However, further increasing the temperature to 55°C resulted in an increase in the adsorption quantity of pollutants, which decreased the light intensity reaching the catalyst surface and the subsequent production of reactive radicals. In addition, lowering the temperature favored the adsorption of the final reaction product, whose desorption tended to inhibit the reaction. At a low reaction temperature, the desorption of the formed products limited the reaction because it was slower than the degradation on the catalyst surface and the adsorption of the pollutants [41].

Figure 11:

Effect of temperature on the decolorization of (A) reactive black 5 and (B) methylene blue dyes.

From the results, we observed that 30% g-C₃N₄/Bi₂MoO₆/5%TiO₂ nanotube was the best visible light active photocatalyst for the decolorization of reactive black 5 and methylene blue dyes. Therefore, to validate its visible light activity, we investigated the photocatalytic decolorization efficiency of the ternary composite under sustainable direct sunlight irradiation, and the results are shown in Figure 12A and B. The results reveal that the photocatalytic decolorization efficiency under direct sunlight was faster than under artificial visible light irradiation.

Figure 12:

Photocatalytic decolorization of reactive black 5 and methylene blue dyes under (A) visible light (B) direct sunlight irradiation.

3.5 Stability and recyclability of photocatalysts

The photocatalyst stability was evaluated through a recycling test on the decolorization of reactive black 5 and methylene blue dyes using a ternary composite photocatalyst (g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube) under visible light irradiation. Each cycle was for 5 h, which consisted of 1-h adsorption followed by 4-h reaction under light irradiation, and the results are shown in Figure 13A and B. The results demonstrate that the photocatalytic decolorization efficiency of the ternary composite was not reduced after five cycles of decolorization experiments; however, the percentage decolorization of reactive black 5 and methylene blue dyes was decreased to 70% and 50%, respectively. This may be due to the adsorption of dye molecules on the surface of the ternary composite photocatalyst (g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube), which decreases the light intensity that reaches the photocatalyst surface and declines the percentage decolorization of dyes. From the recyclability test, it was further confirmed that the ternary composite photocatalyst (g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube) was highly stable and reusable, and it could be promoted for practical applications.

Figure 13:

Recyclability of a ternary composite photocatalyst for the decolorization of (A) reactive black 5 and (B) methylene blue dyes under visible light irradiation.

3.6 Kinetic study

The enhanced photocatalytic decolorization efficiency of the ternary composite photocatalyst (g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube) was confirmed by a kinetic study. It was observed that the photocatalytic decolorization of dyes followed pseudo first-order kinetics in association with the Langmuir-Hinshelwood mechanism, as shown in Eq. (1) (C₀ is the initial dye concentration, C_t is the concentration at a given time (t), and k is the first order rate constant) [42].

The kinetic study on the decolorization of reactive black 5 dye was conducted. The rate constant was calculated from the slope of ln(C_o/C_t) vs. time plot, and the results are tabulated in Table 1. The results demonstrate that the ternary composite photocatalyst (g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube) shows a higher rate constant (0.0159 min⁻¹) compared with the TiO₂ nanotube (0.0122 min⁻¹), g-C₃N₄ (0.0084 min⁻¹), Bi₂MoO₆ (0.0012 min⁻¹), g-C₃N₄/Bi₂MoO₆ (0.0101 min⁻¹), and P25 TiO₂ (0.0028 min⁻¹); the correlation coefficient R² was 0.9719 (Table 1). From the results, we conclude that the use of a ternary composite photocatalyst (g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube) yields a higher visible light photocatalytic decolorization efficiency compared with the use of the other photocatalysts listed in Table 1.

3.7 Mechanism of photocatalytic decolorization

Under visible light irradiation, Bi₂MoO₆, g-C₃N₄, and TiO₂ nanotube photocatalysts are excited; then, the excited electrons are transferred to the conduction band (CB) and leave behind a hole in the valence band (VB), generating e⁻–h⁺ pairs in each photocatalytic material. From the literature, it is understood that VB and CB positions of g-C₃N₄, Bi₂MoO₆, and TiO₂ are −1.13 and +1.57 eV, −0.32 and +2.44 eV, and −0.2 and +3.0 eV, respectively [29], [43]. Therefore, photogenerated e⁻–h⁺ pairs are efficient, transferring electrons from a higher CB of g-C₃N₄ to the CB of Bi₂MoO₆ and then to the CB of TiO₂ nanotube, which react with the adsorbed O₂ to yield a reactive superoxide radical ion (O₂˙⁻) and enhance its decolorization efficiency. Similarly, the holes in VB of the TiO₂ nanotube are transferred to the VB of Bi₂MoO₆ and then to the VB of g-C₃N₄. Figure 14 depicts the probable charge separation processes in a ternary composite photocatalyst (g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube) during the photocatalytic decolorization of dye molecules. However, the accumulated holes in the VB of g-C₃N₄ cannot oxidize the H₂O or surface-adsorbed –OH groups into hydroxyl radicals (˙OH) because the VB potential of CN (+1.57 eV) is much lower than the required potential for ˙OH generation (H₂O/˙OH=+2.4 eV). However, the VB positions of Bi₂MoO₆ and TiO₂ nanotube are sufficient to oxidize the water molecules and produce reactive ˙OH radicals. Therefore, as discussed above, the photogenerated e⁻–h⁺ pairs are separated, producing reactive radicals (˙OH and O₂˙⁻) in the ternary composite. On the other hand, holes in the VB of Bi₂MoO₆ and the TiO₂ nanotube may directly oxidize the dye molecules and enhance the decolorization efficiency. Furthermore, the electrons in the CB of g-C₃N₄, Bi₂MoO₆ and TiO₂ nanotube may also directly react with the adsorbed O₂ and produce O₂˙⁻. The reactive radical species are reacted with the dye molecule, producing a range of intermediates, which are further mineralized into carbon dioxide, water, and inorganic nitrogen with ammonium and nitrate ions. These photogenerated electrons and hole transfer processes can be described as follows:

Figure 14:

Probable mechanism of the separation of photogenerated e⁻–h⁺ pairs during the decolorization of dye caused by a ternary composite photocatalyst (g-C₃N₄/Bi₂MoO₆/TiO₂ nanotube).

g-C3N4+hv(visible light)→g-C3N4(eCB−+hVB+)

Bi2MoO6+hv(visible light)→Bi2MoO6(eCB−+hVB+)

TiO2nanotube+hv(visible light)→ TiO2nanotube(eCB−+hVB+)

TiO2 nanotube (eCB−+hVB+)+g-C3N4(eCB−+hVB+) +Bi2MoO6(eCB−+hVB+)→TiO2 nanotube (eCB−) +g-C3N4(hVB+)

g-C3N4(hVB+)+H2O or OH→No formation of ⋅OH

_{TiO2 nanotube (eCB−) or g-C3N4(eCB−) or Bi2MoO6(eCB−) +O2→O2⋅−+TiO2 nanotube or g-C3N4 or Bi2MoO6}

Bi2MoO6(hVB+)+H2O or OH→Bi2MoO6+⋅OH

TiO2 nanotube (hVB+)+H2O or OH→TiO2 nanotube+⋅OH

Bi2MoO6(hVB+)+TiO2 nanotube (hVB+) +reactive black 5 dye or methylene blue dye →oxidized products

^{O2⋅−+⋅OH+reactive black 5 dye or methylene  blue dye→intermediate products→ CO2+H2O+NH4++NO3−}