Synthesis and Investigation of TiO₂/g-C₃N₄ Performance for Photocatalytic Degradation of Bromophenol Blue and Eriochrome Black T: Experimental Design Optimization and Reactive Oxygen Species Contribution

Abstract

Graphitic carbon nitride (g-C₃N₄) based photocatalyst was synthesized and the photocatalytic performance was investigated for the removal of Eriochrome Black T (EBT) and Bromophenol Blue (BPB) under UV irradiation. The prepared materials were characterized by SEM-EDX, XRD, Raman, FTIR and DRS. Higher degradation efficiency for the same initial concentrations of EBT and BPB in presence of TiO₂/g-C₃N₄ have been achieved within 160 min of irradiation. The kinetic study showed that the photodegradation of BPB by TiO₂/g-C₃N₄ follows pseudo-first-order kinetics with an R2 value of 0.98. The addition of persulfate (PS) in BPB solution improved the degradation yield from 8.81% to 80.14% within 20 min of UV light irradiation. A Box-Behnken model was developed from three factors and Response surface methodology (RSM) was employed to identify the optimum conditions for the treatment of BPB solution by TiO₂/g-C₃N₄. The experimental values of degradation of BPB agreed with predicted values obtained from central composite design (CCD) analysis with an R² value of 0.9999. The scavenger study revealed that superoxide radical anion (O₂^•−) plays a key role (68.89% of contribution) followed by OH^• and h⁺ with 22.40% and 15.55% of contribution, respectively. This study has obviously exhibited the potential of TiO₂/g-C₃N₄ composite as a promising catalyst for photocatalytic purposes.

1. Introduction

Water is a natural resource essential to human life. However, the scarcity of this precious resource has become a major problem faced by most countries in the world. This is usually due to the contamination of its sources by wastewater from human activities, mainly industry and agriculture [1]. Several pollutants found in water are organic in nature, such as dyes which are used in the textile and printing industries [2]. About 15–20% of dyes are lost during the dyeing process and therefore end up in the environment and can become a source of pollution [3]. Pharmaceutical pollutants such as drugs are also found in water when they are unmetabolized, and can be considered pollutants if their by-products are discharged into the environment [4].

The presence of pollutants in the water can be very toxic for living beings, causing health problems such as cancerogenic diseases, hepatitis, cholera and other diarrheal disease [5]. Therefore, the removal of these toxic organic pollutants from water is essential to protect public health. Reverse osmosis, ion exchange, electro-dialysis, electrolysis, coagulation, flocculation and adsorption are examples of some conventional methods commonly used for the removal of pollutants from water [6,7]. Adsorption is an effective technology for the removal of organic and pharmaceutical pollutants from water [8,9,10,11]. In adsorption, pollutants from water adhere to the surface of an adsorbent to purify the water, the pollutants are not eliminated but transferred to the surface the adsorbent which may be considered as another possible future source of pollution [12].

Advanced Oxidation Processes (AOPs) are also promising methods for removal of contaminants from wastewater effluents due to their fast reaction rate and strong oxidation capability [13]. There exists a variety of AOPs, such as Fenton [14], ozonation or catalytic ozonation [15], photocatalysis [16], electrochemical oxidation [17], discharge plasma [18], Sonolysis [19] and ionizing radiation [20]. Among AOPs, photocatalysis is widely studied for wastewater treatment. Thus, our interest in this present work to focus on the photocatalysis which has is an emerging technology for the degradation of organic pollutants in wastewater [21,22].

Part of photocatalysis is the use of catalysts to enhance removal of organic pollutants in wastewater. Due to the increase in demand for catalysts with high performance in photocatalytic degradation of organic compounds, researchers have been looking for means to synthesize new catalysts or modify the existing ones to increase their photocatalytic activities [23]. Various oxidants were used in order to intensify the degradation performance of photocatalyst as reported in different works such as such as H₂O₂, O₂ and K₂S₂O₈ [24,25]. Photocatalytic activated persulfates system was reported to have the merits of environmental protection, a great chemical stability and a sustainability [26]. For example, Sun et al. reported an increase of 75% in quinoline degradation efficiency with increasing persulfate concentration in WO₃-CuFe₂O₄ Z-scheme system within 150 min [27].

Over these years, researchers have developed an interest in graphitic carbon nitride (g-C₃N₄) catalyst in the photocatalytic treatment of pollutants in water [28]. g-C₃N₄ is a low-cost metal-free polymer with properties such as facile synthesis, high stability, unique optical and electronic characteristics which makes g-C₃N₄-based catalysts suitable candidates for photocatalytic applications [29]. Semiconducting metal oxides such as TiO₂ are known to be efficient in the degradation of organic pollutants and combining it with other compounds may increase their photocatalytic activities [30,31]. To promote the photocatalytic performance of TiO₂, different strategies were used, such as surface modification, doping, morphology control, facet engineering, coupling different semiconductors and adding oxidants (i.e., H₂O₂, O₃, and persulfate) [32,33]. Herein, the last two strategies were adopted to achieve these objectives. Indeed, coupling g-C₃N₄ with TiO₂ can increase the reaction site on the surface of g-C₃N₄, and can also act as an active site to promote the transfer of radical species [34,35]. Heterojunction composites like TiO₂/g-C₃N₄ help to reduce electron-hole recombination rate which leads to increase in photocatalytic reactions [36]. Such catalysts have been widely used in the photocatalytic hydrogen production [37], to enhance visible photocatalytic activity [38] and in the degradation of organic pollutants [39]. Eriochrome Black T (EBT) and Bromophenol Blue (BPB) used in printings, pharmaceutical and textiles industries as dyes, They are commonly used to evaluate the photocatalytic performance of semiconductor materials in aqueous media [16,40,41,42].

In this study, a simple synthesis route of g-C₃N₄ based catalysts (g-C₃N₄, 30% TiO₂/g-C₃N₄) were tested and the degradation performances of obtained photocatalysts were studied on the removal of EBT and BPB solutions. The behavior of TiO₂/g-C₃N₄ toward BPB degradation under in influence of a simultaneous different variable parameters was examined via Response Surface methodology (RSM). Furthermore, the potential of TiO₂/g-C₃N₄ in the degradation of dye solutions under various conditions (catalyst dosage, dye concentration, pH values, presence of salt and some oxidants) were also studied. Kinetics of photocatalytic degradation was studied and the contribution of active species in the degradation process was investigated.

2. Materials and Methods

2.1. Materials

Chemicals used for this study were: Bromophenol blue was purchased from REACTIF RAL (Paris, France), Eriochrome Black T and sodium persulfate (Na₂S₄O₈ ≥ 98.0%) were obtained from PROLABO (Paris, French), Hydrogen peroxide (H₂O₂, 30%, from Millipore, Darmstadt, Germany), sodium chloride (NaCl ≥ 99.0%, purchased from SIGMA-ALDRICH, Missouri, USA), hydrochloric acid (HCl ≥ 37%, from Honeywell, Seelze, Germany), sodium hydroxide (NaOH, 98% purity) were purchased from Sigma–Aldrich (Grenoble, French). Isopropanol (C₃H₈O ≥ 99.7%) and methanol (CH₄O ≥ 99.8%) obtained from MERCK company (Darmstadt, Germany) and potassium dichromate (K₂Cr₂O₇ ≥ 99.5%) purchased from Panreac (Barcelone, Spain). All the solutions were prepared using reverse osmosis water. Eriochrome black t and Bromophenol blue properties are shown in the

Table 1. Physico-chemical properties of Eriochrome black t and Bromophenol blue.

Dye	Molecular Formula	Structure	λ Max
Eriochrome black t	C20H12N3O7SNa		535 nm
Bromophenol blue	C19H10Br4O5S		592 nm

:

2.2. Synthesis of g-C₃N₄ and TiO₂/g-C₃N₄ Composites

The synthesis of bulk g-C₃N₄ was done by thermal polymerization based on the previous work [43,44]. Briefly, g-C₃N₄ was synthesized by thermal polymerization where 10 g of dicyandiamide was placed into an alumina crucible, and then heated at 550 °C in a muffle furnace during 4 h (20 °C/min). Afterward cooled to room temperature, the obtained solid was grinded into powders for further usage.

The synthesis of TiO₂ was done by the sol-gel method inspired by the work of Askari et al. [45]. Briefly, titanium isopropoxide was dissolved in acetic acid and the mixture was maintained under stirring. The reverse osmosis water was added drop by drop for the hydrolysis. The resultant gel was dried at 200 °C for 2 h. The TiO₂ nanoparticles were obtained after calcination of the sample at 400 °C during 4 h in order to improve the crystallinity of the nanoparticles.

However the TiO₂/g-C₃N₄ composite was prepared by wet- impregnation according to the literature [46,47]. In which, a suitable quantity of g-C₃N₄ and TiO₂ were dissolved separately in methanol and then, sonificated in an ultrasonic bath during 30 min. The two solutions were mixed and stirred at room temperature for 24 h. The resultant powder was preserved under room temperature for 12 h eliminate solvent by evaporation.

The schematic illustration of the materials preparation was regrouped in Figure 1.

2.3. Photocatalytic Treatment

EBT and BBP solutions used in the experiments were prepared at different concentrations (2, 5, 10 and 12 ppm). Photodegradation experiments were performed in the batch system of 1000 mL beaker and stirred at 400 rpm. The mass of the catalyst was dispersed into a volume of 500 mL solutions. The mixture was left in the dark and continuously stirred for 30 min to achieve the adsorption-desorption equilibrium before UV irradiation (UVA lamp of 24 W with wavelength of 365 nm as reported in Figure S1). After reaching the adsorption-desorption equilibrium, the dye solution was then exposed to UV-light for 180 min. At 20 min time intervals, 3 mL samples of the solution were collected from the photoreactor and filtered to remove the photocatalysts by a syringe filter (pore size 0.45 µm).

A dual-beam spectrophotometer (SHIMADZU UV-1800) was used to measure the absorbance at maximum wavelengths of 535 nm and 592 nm for EBT and BPB, respectively (example of BPB degradation in Figure S2). The photodegradation rate was estimated using the following equation:

$$R (\%)=\frac{C_0-C_t}{C_0}\times 100$$

(1)

where C₀ is the initial concentration of pollutant in mg/L and C_t is the concentration of dyes in mg/L at time t. For all photocatalytic experiments the pH value was adjusted only at the beginning of the experiment. The pH of the solution was adjusted by HCl and NaOH (0.01 M).

All figures (at the exception of figures generated from RSM study) was drawn using OriginPro 2022, from OriginLab Corporation, Northampton, MA, USA.

2.4. Photocatalyst Characterization

-

The X-ray diffraction (XRD) patterns of as-prepared catalysts were made using a D8 Bruker spectrometer (Cu Ka radiation with wavelength λ = 0.15418 nm as a wavelength) the incident angle of 2θ, 5–130° using 0.017° each step, and the acceleration tension is 40 kV and current emission equals 30 mA.

-

Raman spectra were acquired by using Raman spectrometer of JobinYvon company model T64000. The wavelength of laser was 514.5 nm (2.41 eV) and the power was set at 100 mW. The measurement was carried out in solid state by dispersing the sample powder upon glass slide under air at room temperature.

-

The scanning electron microscopy (SEM) images of the photocatalysts were obtained using an (JEOL 5910 LV) apparatus with EDS elementary analysis using SDD detector (Bruker)

-

The UV–vis diffuse reflectance spectra (UV–vis DRS) of the photocatalysts were recorded by Cary 300 instrument with scan Rate of 600 nm/min in shifting range of 80 to 500 cm⁻¹.

3. Results and Discussion

3.1. Characterization of Photocatalyst

The SEM images and EDX spectrum of the photocatalytic materials were shown in Figure 2. It seems that g-C₃N₄ and TiO₂/g-C₃N₄ were composed of many stacked particles, which presented an irregular spherical shapes morphology with more homogeneity for g-C₃N₄. Moreover, g-C₃N₄ particles appear as platelets, which is in good agreement with the 2D structure however in TiO₂/g-C₃N₄ images in addition to the observed platelets there are small particles which may be attributed to TiO2. Images presented also a porous surface with various sizes. EDX spectrum confirmed the presences of C, N and O elements in g-C3N4 sample, and the presence of C, N, O and Ti for TiO₂/g-C₃N₄ composite sample. This may demonstrate that both photocatalyst composites were well synthesized.

The crystallinity structure of the synthesized materials g-C₃N₄ and g-C₃N₄ modified TiO₂ nanocomposites were characterized and the results were shown in Figure 3a. The XRD pattern of TiO₂/g-C₃N₄ revealed numerous peaks at 25.3°, 37.8°, 48.1°, 54.0°, 55.1°, 62.8°, 68.9°, 70.1°, 75.3° and 83.2°, corresponding to the diffraction planes of (101), (004), (200), (105), (211), (204), (116), (220) and (215) which corresponds to the anatase phase JCPDS 21-1272 [44,48]. The g-C₃N₄ shows the diffraction peaks at 12.8° and 27.57°, corresponds to the inter-plane structural packing motif (100) and the interlayer diffraction plane (002) of the hexagonal graphitic carbon nitride [49]. However, while adding the g-C₃N₄ to the TiO₂ the intensity of g-C₃N₄ peaks was relatively narrowed compared to TiO₂ peaks in the composite. The result confirmed that the crystallinity structure of TiO₂ is highly reduced in TiO₂/g-C₃N₄ composites [50].

Raman spectra of photocatalytic materials are shown in Figure 3b. Theses spectra were recorded to obtain more structural information on synthesized composite. Three characteristic Raman active modes of anatase crystalline form of TiO2 with symmetries B1g, A1g and Eg were detected at 396.7, 514.6 and 638.5 cm⁻¹, respectively [51]. The presence of these characteristic vibrational frequencies and their relative intensity confirmed the phase pure anatase TiO₂. These results are in an agreement with XRD analyses and confirm the formation of phase pure anatase TiO₂ [52]. The spectrum of g-C₃N₄ did not present exploitable peaks in this zone.

FTIR spectra of g-C₃N₄ and TiO₂/g-C₃N₄ composites with different g-C₃N₄ contents are presented in Figure 3c. Pure g-C₃N₄ shows characteristic IR peaks similar to those of the previous results [53,54]. The peak at 1640 cm⁻¹ is assigned to C=N stretching vibration mode, while those at 1247, 1325 and 1408 cm⁻¹ are associated with C−N heterocycle stretching of g-C₃N₄, fully condensed C–N moieties/partially condensed C–NH moieties and aromatic C−N heterocycles, respectively [55]. The peak at 808 cm⁻¹ can be attributed to the characteristic bending mode of s-triazine ring system [54]. A broad band in the range of 3150–3300 cm⁻¹ corresponds to the stretching vibration modes of terminal −NH₂ and −NH− groups [56]. Pure TiO₂ shows characteristic broad absorption band at 500–700 cm⁻¹ [57]. It can be clearly seen that the main characteristic peaks of g-C₃N₄ and TiO₂ appeared in TiO₂/g-C₃N₄ sample suggesting the formation of a composite between g-C₃N₄ and TiO₂ [58].

The UV–Vis DRS spectra of g-C₃N₄ and TiO₂/g-C₃N₄ were presented in Figure 4. It can be seen that a combination of TiO₂ and g-C₃N₄ could significantly change the absorption band of composites compared to that of g-C₃N₄ or TiO₂ components. Accordingly, the presence of g-C₃N₄ in the composite gives to TiO₂ the capacity for absorbing light in visible region rather than UV region. Based on these data, the band gap values for g-C₃N₄ and TiO₂/g-C₃N₄ were calculated as 2.65 eV and 2.48 eV, respectively.

3.2. Catalyst Activity

The photocatalytic performances of the catalysts were evaluated by the photodegradation of 10 ppm solutions of EBT and BPB as pollutants using UV light and 200 mg/L of catalyst. For EBT, under UV light irradiation, we obtained a higher degradation efficiency after 180 min with TiO₂/g-C₃N₄ followed by g-C₃N₄, for both dyes. From Figure 5, we observed that degradation efficiencies of EBT are 87.94% and 100% for g-C₃N₄ and TiO₂/g-C₃N₄, respectively.

For BPB, we observe a degradation efficiency of 38.53% and 100% for g-C₃N₄ and TiO₂/g-C₃N₄, respectively. There is a significant difference in the degradation efficiency of the dyes when using g-C₃N₄ with 87.94% and 38.53% for EBT and BPB, respectively. These results may be attributed to the fact that under UV light irradiation and in the absence of catalyst, we observe 36% degradation of EBT, while no degradation of BPB. From another hand, EBT and BPB belong to the Azo and sulphonephthalein families respectively and therefore differ in their physico-chemical properties. It is easier to break the unsaturated bonds of BPB than those of EBT (azo bonds).

A comparative table of photocatalytic degradation of dyes with g-C₃N₄ based photocatalyst with different light sources is shown in

Table 2. Comparison of photocatalytic activity of g-C₃N₄ based photocatalysts.

Photocatalyst	Pollutant	Degradation Efficiency (%)	Light Source	Reference
40 wt% g-C3N4/Ag3VO	Basic Fuchsin	95.0	Visible	[59]
g-C3N4/TiO2(NT)	Rhodamine B	96.7	Visible	[60]
g-C3N4-TiO2	Rhodamine B	99.0	Visible	[61]
9 wt% Bi2O3/porous g-C3N4	Reactive Black 5	84.0	UV-vis	[62]
NP-GQDs-90/g-C3N4	Methyl Orange	96.0	UV	[63]
g-C3N4/TiO2-1.5	Methylene Blue	95.3	UV	[64]
Nanosheet g-C3N4/CNMBGt	Methylene Blue	98.2	UV	[65]
g-C3N4	EBT	87.9	UV	Present work
TiO2/g-C3N4	EBT	100.0	UV	Present work
g-C3N4	BPB	38.5	UV	Present work
TiO2/g-C3N4	BPB	100.0	UV	Present work

below:

3.3. Effect of Photocatalyst Dosage

In views of the results obtained during photolysis were BPB showed no sensitivity to UV-light compared to EBT, it was preferable to continue our studies for more precise results with BPB. The effect of photocatalyst dosage was studied using 10 ppm of BPB solution. An increase in photocatalyst dosage from 50 mg/L to 200 mg/L increases BPB degradation efficiency from 58.17 to 100% during 180 min irradiation time as shown in Figure 6.The addition of catalyst results in an increase in the photocatalytic performance due to an increase in active sites which leads to an increase in radical species formation [66]. Excess dosage of catalyst into the reaction mixture may enhance light reflectance due to the increase of the solution turbidity, which in turn leads to preventing light absorption by the catalyst reducing the degradation rate [30,67,68]. The degradation efficiency increases rapidly with an increase in photocatalyst from 50 to 100 mg/L and increases slightly with an increase in photocatalyst from 100 to 200 mg/L after 180 min irradiation. In order to obtain high degradation efficiencies and avoid the use of excess catalyst, the quantity of photocatalyst chosen to continue our studies was 100 mg/L.

3.4. Effect of Initial Dye Concentration

The effect of initial BPB concentration on degradation was explored under concentrations from 2 to 20 ppm with 100 mg/L of catalyst. Figure 7 shows that an increase in dye concentration from 2 to 20 ppm decreases overall degradation, from 100% to 27.33%.

This can be attributed to the quantity of active sites available when using the same quantity of catalyst for different dye concentration [69]. With low initial concentration, TiO₂/g-C₃N₄ provides a sufficient number of active sites for BPB molecules adsorption. With higher initial concentrations, the number of active sites provided by the TiO₂/g-C₃N₄ reduces, thus decreasing the degradation efficiency of the dye [70].

The opacity of the solution at high concentration may also reduce the intensity of the light travelling through the system which leads to a decrease in the photocatalytic activity of the photocatalyst. Increasing dye concentration facilitates light adsorption by the dye molecules and reduces the amount of light reaching the surface of the catalyst, thus decreasing generation of active sites [71,72].

3.5. Kinetic Study

To determine BPB degradation kinetics, pseudo-first-order kinetic and Langmuir-Hinshelwood models were investigated. The kinetic parameters of the reaction were obtained using the pseudo-first-order kinetic model expressed as follows:

$$\ln (\frac{C_0}{C})=k_{1 }t$$

(2)

The terms k₁, C and C₀ are respectively the rate of pseudo-first-order reaction, expressed in min⁻¹, BPB concentration at the time t, expressed in mg/L, and BPB initial concentration, also in mg/L.

The pseudo-first-order (k₁) reaction rate was obtained from the linear plot of ln(C₀/C) versus time as seen in Figure 8. From

Table 3. Variation of the pseudo-first-order rate constant (k₁) at various initial concentration of BPB (C₀).

C0 (mg/L)	k1 (min−1)	R2
2	0.0561	0.9976
5	0.0200	0.9756
10	0.0072	0.9881
15	0.0024	0.9837
20	0.0011	0.9516

, it can be seen that k₁ decreases with increase in initial concentration which is quite common in photocatalytic degradation of organic compounds in solution [73]. The correlation coefficients close to 1 (R² > 0.95) confirmed that photodegradation of BPB by TiO₂/g-C₃N₄ effectively follows a pseudo first-order kinetics model [74].

Most heterogeneous photocatalytic degradation reactions follow Langmuir–Hinshelwood (L-H) kinetics to describe the relationship between initial degradation rate and initial concentration [75,76].The L-H model used in this study can be expressed as follow:

$$\frac{1}{r_0}=\frac{1}{k}+\frac{1}{kK}\frac{1}{C_0}$$

(3)

where k denotes the reaction rate constant for the process (mg L⁻¹ min⁻¹), K is the adsorption coefficient of reactants (L mg⁻¹), C₀ is the initial concentration of dye (mg/L) and r₀ is the initial rate of disappearance of dye (mg L⁻¹ min⁻¹).

A linear expression can be conveniently obtained by plotting the reciprocal initial rate against the reciprocal initial concentration. The slope is 1/(kK) and the intercept is 1/K (Figure 9). The linear transform of this expression yielded k = 0.108 mg L⁻¹ min⁻¹ and K = 0.0585 L mg⁻¹.

3.6. Effect of Persulfate

The effect of persulfate (PS) on the performance of TiO₂/g-C₃N₄ towards BPB degradation was evaluated by introducing 2.4 mmol/L of sodium persulfate (Na₂S₂O₈) in the suspension. In the first 20 min we observed a removal efficiency of 80.14% and 98.83% after 180 min irradiation (Figure 10). This improvement in removal efficiency comes from the presence of PS which inhibits the recombination of electron/hole pairs in the catalyst by capturing photo induced electrons [77]. Addition of PS accelerates the generation of radicals, thus boosting decontamination rate [67]. This can be explained by recent studies of persulfate activation of g-C₃N₄ and TiO₂ based catalysts.

A review on the activation of persulfate on g-C₃N₄-based catalyst for environmental remediation states that PS-captures photoinduced electrons of g-C₃N₄ and reduces electron-hole pairs recombination thereby increasing the formation of radicals which increases the photodegradation of the pollutant (bisphenol A; BPA) [78]. This was further explained through the following reactions:

g-C₃N₄ + hν → e⁻ + h⁺

(4)

e⁻ + O₂ → O₂^•−

(5)

e⁻ + S₂O₈⁻² → [ S₂O₈⁻²] ⁻

(6)

[S₂O₈^{− 2}] ⁻ + O₂ → S₂O₈^{− 2} + O₂^•−

(7)

O₂^•−/h⁺ + organic pollutants → organic pollutants

(8)

From the recent advances in persulfate-assisted TiO₂-based photocatalysis for wastewater treatment, TiO₂ can be excited to produce electron-hole pair and persulfate is then activated by photo-induced e⁻ to generate SO₄^•− radicals [33]. The h⁺ can react with H₂O to produce hydroxyl radicals (^•OH). Under UV light, persulfate can also be activated to form SO₄^•− radicals [79] which can also react with OH⁻ ions to form ^•OH radicals [80]. SO₄^•− and ^•OH radicals with strong oxidation ability can degrade organic pollutants in water, as explained in these reactions [33]:

TiO₂ + hν → e⁻ + h⁺

(9)

h⁺ + H₂O → ^•OH + H⁺

(10)

S₂O₈²⁻ + hν → 2SO₄^•−

(11)

SO₄^•− + OH⁻ → SO₄²⁻ + ^•OH

(12)

SO₄^•−/^•OH + organic pollutants → degradation products

(13)

3.7. Effect of pH, H₂O₂ and NaCl

Three-level CCD (low, central and high) with three factors was used for create a model for the photodegradation of BPB in the presence of TiO₂/g-C₃N₄ under UV-light. Barzegar et al., applied this design to express mathematical relationships among variables such as process time, catalyst mass and initial concentrations of methylene blue (MB) and rhodamine B (RhB) in the degradation process with TiO₂/g-C₃N₄ [35]. The model examined was made of 13 experiments to determine the contribution of the factors (A: pH, B: C_H2O2, C: C_NaCl) on photocatalytic degradation of BPB and

Table 4. Experimental design for the mutual effects of pH, NaCl, and H₂O₂ on photocatalytic degradation BPB by TiO₂/g-C₃N₄.

Run	Factor 1 A: pH	Factor 2 B: CH2O2 (mmol/L)	Factor 3 C: CNaCl (mg/L)	Response Degradation %
1	7	0.4	5	28.69
2	11	0.4	10	18.44
3	7	2.4	5	14.26
4	11	2.4	10	67.52
5	3	0.4	10	80.60
6	7	2.4	15	6.06
7	3	1.4	15	77.59
8	11	1.4	15	65.87
9	7	1.4	10	10.00
10	7	0.4	15	21.56
11	3	2.4	10	80.70
12	3	1.4	5	80.88
13	11	1.4	5	57.54

represents the obtained BPB removal percentage for each experimental run.

The experimental results obtained were represented also as a spider or radar graph visualization to better highlight the effect of different parameters on the degradation yield of BPB as shown in Figure 11. To get a good presentation the three studied parameters (pH, NaCl, and H₂O₂) was reported with theirs coded values (−1, 0 and 1) however the degradation yield was reported from zero to one. I t seems that a basic pH and hydrogen peroxide are suitable for a high BPB molecules degradation, nevertheless, the NaCl presence in some case have a negative effect and, in another case, have no effect in the dye elimination.

Response surface methodology (RMS) was applied to determine the best conditions for BPB degradation. A Box-Behnken design model was used to investigate the effects of 3 independent variables at three pH levels (3, 7, and 11), concentration of H₂O₂ (0.4, 1.4 and 2.4 mmol L⁻¹) and concentration of NaCl (5,10 and 15 mg L⁻¹). The modified quadratic model equation was as follows:

%R degradation = +108.56630

−13.20569 * pH + 31.65992 * C_H2O2 +8.74214 * C_NaCl

−14.24516 * pH * C_H2O2 − 2.76016 * pH * C_NaCl

+ 0.916335 * pH² − 0.370286 * C_NaCl²

(14)

1.23617 * pH² * C_H2O2 + 0.063672 * pH² * C_NaCl

+ 0.100700 * pH * C_NaCl²

The ANOVA results of BPB removal (%) by TiO₂/g-C₃N₄ composite is presented in the

Table 5. ANOVA Table and goodness-of-fit coefficients for response variables.

Source	F-Value	p-Value
Model	2545.44	0.0004
A-pH	3239.21	0.0003
B-CH2O2	511.38	0.0019
C-CNaCl	134.20	0.0074
AB	1369.07	0.0007
AC	77.05	0.00127
A2	17642.96	< 0.0001
C2	447.28	0.0022
A2B	1785.98	0.006
A2C	118.45	0.0083
AC2	462.95	0.0022

. As seen, the p-value of <0.0001 for the model signifies that the model is statistically significant and indicate that the applied mathematical model better fitted to the experimental data as confirmed by lack of fit [81]. To be determined as significant, p-value for a model parameter (individual or interaction) must be < 0.05. In this case, all the model terms are significant [81].

The Predicted R² of 0.9877 is in reasonable agreement with the Adjusted R² of 0.9995 (

Table 6. Fit Statistics.

Parameter	Value
R2	0.9999
Adjusted R2	0.9995
Predicted R2	0.9877
Adequate Precision	122.1613

). Adequate Precision measures the signal to noise ratio. A ratio greater than 4 is desirable. The obtained ratio of 122,161 indicates an adequate signal, which confirm that this model can be used to navigate the design.

A comparison of the experimental data to the predicted values by the refitted model was made through a predicted vs. actual plot, their closeness can be better observed in Figure 12. Data points were fragmented consistently on the diagonal line of 45° proving the goodness of fit. These results are similar to those obtained when using optimization approach in phosphoric acid-treated spent tea residue biochar for wastewater decoloring [82] and the sorption and removal of crude oil spills from seawater using peat-derived biochar [83].

The 3D surface response graph was plotted against two parameters to illustrate the combined effects of parameters on simultaneous photocatalytic degradation of BPB by TiO₂/g-C₃N₄ as shown in Figure 13.

In Figure 13a,b, the degradation efficiency reveals that lower pH leads to more photodegradation efficiency with approximately 75–80% of BPB removal at pH 3 and C_H2O2 (0.4–2.4 mmol L⁻¹). At pH > 3 we observe a decrease in degradation and a gradual increase from pH 9 up to pH 11 where we observed a degradation efficiency of approximately 70% at C_H2O2 ranging between 1.9 to 2.4 mmol/L.

Figure 13c,d showed similar results as the previous plot with photodegradation efficiency of approximately 75–80% of BPB removal at pH 3 and with C_NaCl (5–15 mgL⁻¹). At pH 11 there is a degradation efficiency of approximately 60% at C_NaCl (5–15 mgL⁻¹).

3.8. Reactive Oxygen Species Contribution

To determine the contribution of Reactive Oxygen Species (ROS) in the degradation of BPB by TiO₂/g-C₃N₄ under UV light illumination, different ROS trappers were used at sufficient concentrations to suppress the corresponding ROS. Isopropanol, methanol and potassium dichromate were used as specific trappers of hydroxyl radicals (HO^•), holes (h⁺) and superoxide ions (O₂^•−) respectively in bulk solutions [24,84].

The experiments were carried out by addition of 43.5 mM isopropanol (IsoPro), 50 mM methanol (MeOH), and 1.1 mM potassium dichromate (K₂Cr₂O₇) into a 500 mL BPB (10 ppm) to examine their impact on photocatalytic degradation of BPB with 100 mg/L of catalyst.

From Figure 14, the presence of these trappers in the TiO₂/g-C₃N₄ suspensions affected the BPB degradation, which suggested that all of the h⁺, HO^•−and O₂^•− contributed to the BPB degradation. Using IsoPro, MeOH and K₂Cr₂O₇ we obtained a degradation percentage of 71.87, 78.22 and 27.89%, respectively. From this result, Figure 14 confirmed that O₂^•− was the primary ROS involved in the photocatalytic degradation of BPB with 68.89% contribution followed by HO^• and h⁺ with 22.40% and 15.55% contribution, respectively.

The amount of radicals in the system involved in the dye degradation can differ significantly due to electron-hole recombination and recombination reactions of radical species [24].

Considering the scavenging tests, the photodegradation of BPB with the TiO₂/g-C₃N₄ catalyst could be described by following steps: TiO₂/g-C₃N₄ was excited by UV-light to generate electron-hole pairs with photogenerated electrons flowing from g-C₃N₄ to the surface of TiO₂. Part of the electrons produced by g-C₃N₄ are captured by the oxygen adsorbed on the surface, generating superoxide radicals (O₂^•−) with strong oxidizing properties, then decomposing the pollutant in the wastewater, as shown in reactions below [85]:

TiO₂/g-C₃N₄ + hν → e⁻ + h⁺

(15)

O₂ + e⁻ → O₂^•−

(16)

H₂O + h⁺ → OH⁻+ H⁺

(17)

h⁺ + OH⁻ → HO^•

(18)

HO^• + O₂^•− + pollution → degradation products

(19)

Researches on photocatalytic degradation of organic pollutants generally align with the fact that h⁺ and O₂^•− are the main radical species using g-C₃N₄-based catalyst [78] which is slightly different from the results obtained.This implies that the presence of g-C₃N₄ with TiO₂ and UV irradiation can affect the physico-chemical properties of the photocatalyst and influence main radical species in the photocatalytic organic treatment process.

4. Conclusions

In this work, TiO_2/g-C₃N₄ was synthesized by wet- impregnation and applied in the photodegradation of EBT and BPB dye solutions. The heterojunction phenomenon in TiO₂/g-C₃N₄ facilitates the photoexcited charges transfer and simultaneously reduces photogenerated electron-hole recombination, which in turn markedly improves its photocatalytic performance which is higher than that of g-C₃N₄ alone. The degradation efficiency indicated that a decrease in initial dye concentration and increase in catalyst dosage positively affects photodegradation. The study of the degradation kinetics showed that the degradation of BPB follows both the pseudo-first-order kinetic and Langmuir-Hinshelwood model with an average value of R² equal to 0.979 and 0.975 respectively. The addition of Na₂S₂O₈ to the suspension increased the degradation rate in 20 min from 8.8% to 80% yield, an amplitude of 9.1. The optimization of the photodegradation of BPB was investigated using CCD and RSM methodologies where R% reached maximum values at pH 3 in the presence of NaCl and H₂O₂. Isopropanol, methanol and potassium dichromate were used as scavenger radicals and O₂^•− were identified to be the predominant radical species in the photodegradation using TiO₂/g-C₃N₄ catalyst. The present work provides the efficient photocatalysts as a promising material for environmental remediation purposes.