Kinetics of photocatalytic decolourization of cationic dye using porous TiO₂ film

Peer review under responsibility of Taibah University.

Abstract

In this work, the kinetics of the photocatalytic decolourization of methylene blue (MB) is investigated using different surface morphologies of multilayer TiO₂ coating onto a glass plate under irradiation from a 55-W household florescent lamp. A simple direct dip-coating technique was used, and the coating properties of TiO₂ powder were improved by adding epoxidized natural rubber (ENR) as an organic binder in the coating formulation. The effects of the fundamental parameters that govern the kinetics of the photocatalytic decolourization of MB, such as the mass of TiO₂ coated onto the glass plate, the pH and the TiO₂ surface morphology, were also studied. The kinetics of the MB decolourization in all cases was found to be pseudo-first-order kinetics and was fitted to the Langmuir–Hinshelwood model. The degraded part of the ENR binder led to generating pores within the surface of the TiO₂/ENR film and converting it into porous form, as confirmed by SEM analysis. Furthermore, TGA, FTIR and leachability analyses were conducted to further confirm the depletion of the ENR from the TiO₂/ENR film. The kinetics of the MB decolourization and the efficiency of the MB colour removal indicated that the porous TiO₂/ENR film becomes approximately twice as fast as the non-porous TiO₂/ENR film.

Keywords:

• Kinetics
• Decolourization
• Methylene blue
• Immobilization
• TiO₂

1 Introduction

Among all of the semiconductor photocatalysts, titanium dioxide (Anatase) currently stands as an ideal benchmark photocatalyst in environmental photocatalysis applications because of its many desirable properties, such as being inexpensive and readily available, biologically and chemically inert, and having good photoactivity compared to other semiconductors [1,2]. However, there are two major technical challenges that have remarkably limited its potential field applications. First is the relatively wide band gap of TiO₂, which absorbs only 3–4% of the energy of the solar spectrum and has restricted its applications to UV excitation sources [3]. Second, the effective slurry or suspension mode application of TiO₂ needs a post-treatment catalyst-recovering step, which is normally difficult, energy consumptive and not cost-effective [4]. Moreover, TiO₂ powder has a strong tendency to aggregate, especially at high concentrations, which provides a real limitation to applications to continuous flow systems and suffers from the scattering of incident UV light by the suspended particles [5]. Another important factor that must be considered in photocatalysis applications is the pH of the solution; the pH of the aqueous solution significantly affects TiO₂, including the charge on the particles, the size of the aggregates it forms, and the positions of the conductance and valence bands [6,7].

Immobilization of TiO₂ powder on solid supports is an alternative convenient method for solving the problem of the post-treatment catalyst powder recovery and for facilitating the photocatalyst usage for long-term applications [8]. However, the photocatalytic efficiency of the immobilized TiO₂ system could be less than that of the slurry system due to having less surface area that is accessible for photocatalytic reactions as well as low porosity of the supported catalyst layer [9,10]. In addition, cracking and peeling off of the catalyst layer can also be expected because of the poor adherence of the photocatalyst/support. Thus, it was proven that adding an organic binder, such as epoxidized natural rubber (ENR) combined with phenol-formaldehyde resin (PF) in an immobilizing solution can produce a good quality coating formulation of TiO₂ powder onto a solid support [11]. However, mixing more than one organic binder for TiO₂ powder immobilization is responsible for generating different types of transformation products and undesirable intermediates during the photocatalytic reaction [12].

In recent years, a large amount of attention has been paid to using a compact household fluorescent lamp as an alternative light source to replace a conventional UV light source. It was found that a very low UV content in a compact fluorescent lamp is sufficient to induce TiO₂ particles for a wide range of applications, such as photocatalytic oxidation of biopolymers [13], crosslinked biopolymers [14,15], colourless organic water pollutants [11,12] and cationic dye [16]. Therefore, the aim of this work is to study the decolourization kinetics of methylene blue on the surface of an immobilized TiO₂ layer under most significant experimental parameters, such as the mass of TiO₂ coated onto a glass plate, pH and surface morphology of the immobilized TiO₂ film. Among various types of cationic dyes, MB was chosen as a probe in this study due to its complicated chemical structure and because it is not easy to be removed from wastewater by conventionally used techniques such as biological treatments and chemical precipitation.

2 Materials and methods

2.1 Materials

Titanium (IV) oxide (99% anatase) powder with a particle size of ≤150 μm was obtained from Sigma–Aldrich. Epoxidized natural rubber (ENR50) was obtained from Kumpulan Guthrie Sdn. Bhd., Malaysia. Toluene and acetone were purchased from BDH Analar. Hydrochloric acid (HCl) obtained from R & M Chemicals and sodium hydroxide (NaOH) pellets purchased from Mallinkoot were used to adjust the pH. Methylene Blue dye (c.a. 98%, Colour Index Number: 52015) was purchased from R & M, and Table 1 summarizes the chemical structure, molecular weight and λ_max for MB. All of the materials were used as received without further purification. Ultra-pure water (18.2 MΩ cm⁻¹) was used to prepare all solutions in this study. The chemicals used in this experiment were all of the analytical grade and were used without further purification.

Table 1 Characteristics of Methylene blue dye.

Chemical structure of MB	Chemical formula	Molecular weight (g mol^−1)	λmax (nm)
	C16H18ClN3S·2H2O	319.85	661 nm

2.2 Instruments and methods

A Daihan Ultrasonic cleaner model WUC-D06H (50–60 KHz) from Scientific Co., Inc. Company was used to homogenize TiO₂ dip-coating solutions. A spectrophotometer, model DR 2800 from HACH, was used to determine the concentration of MB. The chemical oxygen demand (COD) was determined by using a high-range digestion solution according to an analytical procedure [17]. In this analysis, the samples were placed in vials that contained a high range of COD digestion solution and a small amount of mercury sulfate. The vials were inverted many times to ensure complete mixing before being placed in a COD reactor (HACH 2000), which was preheated to 150 °C. The samples were kept in the COD reactor for 2 h. The samples were then withdrawn from the reactor and cooled down to room temperature, followed by COD measurement at 620 nm by a DR 2800 HACH spectrophotometer. Digested ultrapure water was used as the blank reference solution, and the difference between the absorbance of the digested blank sample and the digested sample was the COD measurement of the sample. The peel-off rate of the immobilized TiO₂/ENR film was observed using sonication according to a published procedure [12]. The leaching ratio (LR) of TiO₂/ENR film was calculated according to a published procedure [13]. A scanning electron microscope (SEM) analyser model LEO SUPRA 50 VP Field Emission SEM was used to investigate the surface morphology of the TiO₂ layer. The Fourier transform infrared spectroscopy analyser model Nicolet, from Thermo Electron Corporation, was used to study the changes in the functional groups in the TiO₂ formulation. A thermogravimetric analyser (TGA) from TA Instruments, model STDQ600, was used to determine the composition of the polymeric binders that were used in TiO₂ dip-coating formulation at a heating rate of 10 °C min⁻¹, and scanning was performed from 30 °C to 800 °C under an inert N₂ gas flow.

2.3 Preparation of the TiO₂ formulation

The TiO₂ photocatalyst formulation was obtained by adding a fixed amount of 5 g ENR solution (11.32% solution of ENR in toluene) into an amber bottle that contained 12 g of TiO₂ powder. Afterward, 60 mL of acetone was poured into the bottle before being homogenized by a sonication process for 5 h. A simple dip-coating method was used to immobilize the TiO₂ formulation onto a glass plate. Glass plates with smooth surfaces on one side with dimensions of 45 mm (width) × 85 mm (height) were used as inert solid support materials for immobilizing the photocatalyst formulation. Varying amounts of TiO₂ photocatalyst were produced by dipping and drying a glass plate for a varying number of times in the TiO₂ formulation. In the dip-coating method, it was found that the catalyst loading was directly proportional to the number of dips into the TiO₂ formulation.

2.4 Photo-reactor and light source

A custom-made glass photo-reactor cell with a length of 58 mm, width of 10 mm and height of 90 mm was used for all photocatalytic experiments. A compact 55-W fluorescent lamp (Qusun) with a UV content of 7.32 Wm⁻² and visible 441 Wm⁻² was used as the light source. The UV and visible content in a compact 55-W fluorescent lamp was measured using a radiometer (Solar light co. PMA 2100) connected to a UV-A and UV-B broadband detector (PMA 2130) and a visible light detector (PMA 2132). This lamp was placed vertically in contact with the outer surface of the photo reactor cell. Then, 20 mL of 20 mg L⁻¹ MB solution was poured inside the photo-reactor cell. The initial pH of the MB solution was adjusted by adding either 0.1 M HCl or NaOH until the desired pH value was obtained by monitoring with a Metrohm, 827 pH lab. Afterward, an immobilized TiO₂/ENR film was placed upright inside the photo-reactor cell and exposed directly to the light source. The aeration flow rate was maintained at 50 mL min⁻¹ throughout the photocatalytic degradation process. The schematic diagram of the complete photo-reactor and light source setup was presented in our previous work [14]. To generate open pores on the surface of an immobilized TiO₂/ENR film, the TiO₂/ENR film was irradiated in 20 mL ultra-pure water with a 55-W compact fluorescent lamp, and an aeration of 50 mL min⁻¹ was performed for 5 h continuously. The degraded part of the ENR binder during the photocatalytic process acts as a pore-forming agent and converts the surface of the immobilized TiO₂/ENR film to be more porous. To monitor the ENR by-product mineralization, the TiO₂/ENR film was immersed in ultra-pure water under irradiation with a 55-W compact fluorescent lamp. After 1 h of irradiation, the water sample was collected and replaced with another fresh batch of ultra-pure water using the same TiO₂/ENR film until 5 h of irradiation was reached. The collected water samples were subjected to COD testing.

2.5 Photocatalytic efficiency and kinetic study

The change in the concentration of MB during the photocatalytic decolourization process was measured for its respective absorbance using a DR 2800 HACH spectrophotometer, and the results were converted into the corresponding concentrations (C). The photocatalytic efficiency was calculated using the following (Eq. (1)(1) $\text{MB}\text{colour}\text{removal}(\%)=(C_o-C_t)/C_o\times 100$(1) ):(1) $\text{MB}\text{colour}\text{removal}(\%)=(C_o-C_t)/C_o\times 100$(1) where C_o and C_t (both in mg L⁻¹) are the initial and remaining concentration of MB in solution, respectively, and C_t is the absorbance at any irradiation time t (min).

The kinetics of the photocatalytic decolourization rate of MB was determined using the Langmuir–Hinshelwood kinetics model, as given in the following (Eq. (2)(2) $\text{ln}\left(\frac{C_o}{C_t}\right)=k_{\text{app}}t$(2) ):(2) $\text{ln}\left(\frac{C_o}{C_t}\right)=k_{\text{app}}t$(2)

The pseudo-first-order rate constant, k_app (min⁻¹), was calculated from the slope of ln (C_o/C_t) versus irradiation time t.

3 Results and discussion

3.1 Adhesion strength of the TiO₂ layer

Fig. 1 shows the percentage of the TiO₂ film that remains attached on the glass plate at different sonication times. As observed, the gradual decreases in the catalyst amount occurred with increasing sonication time. At 30 s of sonication, the percentage of TiO₂ that remains is approximately 20%. This relative strength can be attributed to the role of ENR organic binder, which offers good elasticity and adhesion. Thus, ENR acted as an adhesive to strengthen the coating conditions of the immobilized catalyst.

Fig. 1 The percentage of TiO₂/ENR film remains attached on the glass plate at different sonication times.

3.2 Effect of TiO₂ loading on the MB decolourization kinetics

The optimization of TiO₂ loading is an important issue that is significant for avoiding the use of excessive TiO₂ and ensuring the maximum absorption of photon light. A series of TiO₂ loadings that ranged from 0.65 to 3.93 mg cm⁻² were obtained by varying the number of dip-coatings on the glass plate. In the dip-coating process, the TiO₂ loading and thickness increase almost linearly with an increase in the number of dip coatings [18]. Fig. 2 shows that the photocatalytic decolourization of MB was fitted to the Langmuir–Hinshelwood model by plotting ln (C_o/C_t) against the irradiation time (t) to obtain straight lines with linear regression coefficients (R²). The pseudo-first-order rate constants, k_app (min⁻¹), were calculated from the slopes of the plots and are presented in Table 2. Furthermore, Fig. 3 shows the pseudo-first-order rate constants for the photocatalytic decolourization of MB by TiO₂/ENR film at different loadings onto the glass plate. From Fig. 3, it can be concluded that the pseudo-first-order rate constant becomes 2.8 times faster by increasing the TiO₂ loading from 0.65 to the optimum loading of 2.62 mg cm⁻². The SEM analysis shows that the thickness of the TiO₂ film at the optimum loading of 2.62 mg cm⁻² was 34.31 μm (image not shown). After this loading of 2.62 mg cm⁻², no remarkable change in the pseudo-first-order rate constants can be observed. This observation can be attributed to the limited surface area of the solid support (glass plate); in addition, loading more TiO₂ leads to increasing the thickness of the immobilized catalyst on the glass plate without significantly increasing the number of photo-induced catalyst particles. A similar observation was also reported by Jawad and Nawi [12] of the photocatalytic performances of the glass/chitosan/TiO₂ interface composite film for the degradation of phenol. Below the optimal TiO₂ loading, there are insufficient TiO₂ particles immobilized on the glass surface, and therefore, less light absorption will occur and there will be less production of oxidative radicals due to the thin TiO₂ layer.

Table 2 Linearity of ln C_o/C_t versus time for photocatalytic decolourization of 20 mg L⁻¹ MB at different TiO₂/ENR loading on glass plate.

Amount of TiO2/ENR (mg cm^−2)	Rate constant (min^−1)	Linear regression (R^2)
0.65	0.0180	0.960
1.31	0.030	0.984
1.96	0.042	0.981
2.62	0.051	0.971
3.27	0.053	0.969
3.93	0.054	0.957

Fig. 2 Plot of ln(C_o/C_t) versus time for the photocatalytic decolourization of MB by TiO₂/ENR film at different loading of TiO₂ onto glass plate (experimental condition: [MB]_o: 20 mg L⁻¹, aeration rate: 50 mL min⁻¹, initial pH MB: 6.6).

Fig. 3 Effect of TiO₂ loading on MB colour removal at ambient pH (pH: 6.6) with initial MB concentration of 20 mg L⁻¹ and aeration rate of 50 mL min⁻¹.

3.3 Effect of pH on the MB decolourization kinetics

It is well known that pH will influence both the surface state of titanium and the ionization state of the ionizable dye molecules [4]. Thus, Fig. 4 shows that the photocatalytic decolourization of MB under a wide range of pH values was fitted to the Langmuir–Hinshelwood model by plotting ln (C_o/C_t) versus irradiation time (t). The pseudo-first-order rate constants, k_app (min⁻¹), were calculated from the slopes of the plots and are presented in Table 3. The influence of the initial pH of the solution on the calculated pseudo-first-order rate constant for the photocatalytic decolourization of MB is depicted in Fig. 5. As observed, the pseudo-first-order rate constants (k_app) remarkably increased from 0.0102 min⁻¹ to 0.0485 min⁻¹. In other words, the pseudo-first-order rate constant becomes almost five times faster by increasing the initial pH of the MB solution from 2.0 to 8.0, and no more noticeable changes can be observed with further increasing of the pH up to 10.0. Thus, the optimal pH that was obtained was found to be pH ∼8.0. In fact, the point of zero charge (pHpzc) of TiO₂ is 6.8 [19]. Therefore, the TiO₂ surface becomes positively charged with the species TiOH₂⁺ at pH < pHpzc, negatively charged with the species TiO⁻ at pH > pHpzc, and neutral with the species TiOH at pH = pHpzc [20]. Furthermore, the protonation and deprotonation condition of the photocatalyst surface by taking TiO₂ as a model can be shown in Eqs. (3)(3) $\text{TiOH}+{\text{H}}^{+}\to \text{TiO}{\text{H}}_2^{+}\text{in}\text{acidic}\text{solution}$(3) and (4)(4) $\text{TiOH}+{\text{OH}}^{-}\to {\text{TiO}}^{-}+{\text{H}}_2\text{O}\text{in}\text{alkaline}\text{solution}$(4) :(3) $\text{TiOH}+{\text{H}}^{+}\to \text{TiO}{\text{H}}_2^{+}\text{in}\text{acidic}\text{solution}$(3) (4) $\text{TiOH}+{\text{OH}}^{-}\to {\text{TiO}}^{-}+{\text{H}}_2\text{O}\text{in}\text{alkaline}\text{solution}$(4)

Table 3 Linearity of ln C_o/C_t versus pH solution of 20 mg L⁻¹ MB at fixed TiO₂/ENR loading on glass Plate 2.62 mg cm⁻².

pH	Rate constant (min^−1)	Linear regression (R^2)
2	0.0102	0.9531
3	0.0203	0.9478
4	0.0337	0.9332
5	0.0383	0.9242
6	0.0432	0.9962
7	0.0456	0.9455
8	0.0485	0.9168
9	0.0487	0.9276
10	0.0482	0.9441

In alkaline medium, the TiO₂ surface is negatively charged (TiO⁻) and MB is basically a cationic dye; therefore, a strong electrostatic interaction can be achieved. Another reason is that increasing the pH of the solution increases the amount of hydroxyl radicals generated (because more hydroxyl ions are available on the surface of the TiO₂). For both of the above-mentioned reasons, the rate constant of the photocatalytic activity for MB decolourization is expected to increase at high pH values.

Fig. 4 Plot of ln(C_o/C_t) versus time for the photocatalytic decolourization of MB by TiO₂ film at different pH values (experimental condition: [MB]_o: 20 mg L⁻¹, aeration rate: 50 mL min⁻¹ and 2.62 mg.cm⁻² TiO₂ loading onto glass plate).

Fig. 5 A plot of pseudo-first-order rate constant of the photocatalytic decolourization of 20 mL 20 mg L⁻¹ MB as a function of pH changes at fixed TiO₂ loading 2.62 mg cm⁻².

3.4 Preparation of porous TiO₂/ENR film

The main advantage of using an organic binder for immobilization of the photocatalyst is to generate pores and enhance the surface area of the immobilized photocatalyst by systematic degradation of the binder during the photocatalytic reaction. Moreover, the transformation by-products that are generated from the degraded binder in the treated water can be effectively and totally mineralized by the photocatalytic reaction [21]. Thus, to investigate the effect of the porosity and change in the surface morphology of an immobilized TiO₂/ENR film on the photocatalytic decolourization of MB, an experiment was conducted by irradiating TiO₂/ENR film in 20 mL ultra-pure water for 5 h. In fact, during the 5-h irradiation process in the presence of ultra-pure water, a partial degradation of ENR catalyst binder will take place [11]. The degraded or depleted part of the ENR will act as a pore-forming agent on the TiO₂ surface. To confirm the formation of porous structure on the surface of the TiO₂ film and the depletion of the ENR binder, analysis of the surface morphology of the TiO₂ in addition to quantitative and qualitative analyses were conducted as described below.

3.4.1 Leaching ratio (LR) measurements

Fig. 6 shows the weight loss of the irradiated TiO₂/ENR film in ultra-pure water in the form of the leaching ratio (LR) at different times of irradiation. As observed, the maximum LR of the irradiated TiO₂/ENR film was 7.3% after 4 h of irradiation, and no remarkable changes can be observed by extending the irradiation time. The weight loss of the TiO₂/ENR film was assumed to be due to the leaching or degradation of some of the ENR binder within the TiO₂/ENR film. The transformation of the ENR by-products into minerals during the photocatalytic-oxidation process will be discussed separately in detail in the upcoming mineralization study section.

Fig. 6 Measuring the weight loss as a function of the leaching ratio (LR) of the irradiated TiO₂ film at different times of irradiation in ultra-pure water.

3.4.2 SEM analysis

Fig. 7 shows representative SEM images of TiO₂/ENR film before and after 5 h of the irradiation process in ultra-pure water. As observed, the surface morphology of the irradiated TiO₂/ENR film has remarkably changed to be more porous after the irradiation process. The high proportion of pores generated on the immobilized TiO₂ surface was attributed to the degraded part of the ENR binder.

Fig. 7 SEM micrographs of the TiO₂/ENR film at fixed TiO₂ loading 2.62 mg cm⁻². (a) Before irradiation, (b) and (c) after 5 h of irradiation in ultra-pure water at 1000 and 5000 magnification power respectively.

3.4.3 FTIR and TGA analysis

The FTIR spectra of the TiO₂ powder, non-irradiated immobilized TiO₂/ENR film (non-porous TiO₂/ENR film) and irradiated immobilized TiO₂/ENR (porous TiO₂/ENR film) for 5 h in ultra-pure water are shown in Fig. 8(a)–(c), respectively. In the case of non-porous TiO₂/ENR film (spectrum b), the broad band characteristic at 3459.39 cm⁻¹ is attributed to the O–H stretch in the TiO₂/ENR formulation. The peak at 2937.65 cm⁻¹ is assigned to asymmetric and symmetric stretching vibration of the methyl (–CH₃) group. The band at 1715.56 cm⁻¹ is assigned to the C=O in ketones and saturated aliphatic. The band at 1646.82 cm⁻¹ is attributed to C=C stretching. The bands at 1453.32 cm⁻¹ and 1381.39 cm⁻¹ are assigned to CH₂ and CH deformations of the hydrocarbon backbone [22,23]. After 5 h of light irradiation, the intensity of the bands contributed by the ENR polymer was significantly reduced (see spectrum c). This obvious reduction in the band intensities of spectrum (b) indicated clearly the depletion of the ENR binder in the TiO₂ formulation during the photocatalytic process. For further confirmation, TGA analysis was conducted for the same samples, and the TGA curves are given in Fig. 9. As observed, the stage degradation pattern of the ENR in the TiO₂ formulation occurred at a peak temperature of approximately 380–600 °C. The weight loss was 5.49% and the residual was 94.51% for the non-irradiated TiO₂ formulation. On the other hand, the weight loss was 2.22% and the residual 97.78% in the case of TiO₂/ENR that was irradiated for 5 h with ultra-pure water. This difference in the weight loss of the TiO₂/ENR formulation before and after irradiation indicates clearly that a considerable amount of ENR binder was degraded during the photocatalytic reaction. Thus, the degraded part of the ENR binder acted as a pore-forming agent with the immobilized TiO₂ surface.

Fig. 8 FTIR spectra of (a) TiO₂ powder, (b) non-irradiated TiO₂/ENR (fresh TiO₂/ENR film) and (c) irradiated TiO₂/ENR (irradiated TiO₂/ENR film for 5 h in ultra-pure water).

Fig. 9 TGA curves of (a) TiO₂ powder, (b) non-irradiated TiO₂/ENR (fresh TiO₂/ENR film) and (c) irradiated TiO₂/ENR (irradiated TiO₂/ENR film for 5 h in ultra-pure water).

3.5 Photocatalytic efficiency and control testing

The effects of the key parameters on the decolourization of MB, such as photolysis (light-without TiO₂), adsorption (dark-with TiO₂), and photocatalysis (light-with TiO₂) at different surface morphologies of TiO₂/ENR film, were examined. The initial MB concentration was fixed at 20 mg L⁻¹ at the optimal conditions with TiO₂ loading at 2.62 mg cm⁻² and pH 8.0, as shown in Fig. 10. As observed, the highest removal of MB colour can be achieved by porous TiO₂/ENR film in both the processes of photocatalysis and adsorption compared to non-porous TiO₂/ENR film. It was found that almost 99.07% of MB decolourization was achieved after 45 min of irradiation using porous TiO₂/ENR film. Almost the same percentage (99.01%) of MB decolourization was also reached but after 90 min of irradiation using non-porous TiO₂/ENR film. Regarding the adsorption process, the maximum MB colour removal was recorded as 35.42% and 24.08% after 45 min of contact time by using porous TiO₂/ENR film and non-porous TiO₂/ENR film, respectively. Moreover, the photolysis process was extremely poor, with only 13.25% removed by irradiation with light only and without using TiO₂/ENR film. The kinetics study shows the straight lines with linear regression coefficients (R²) that are obtained by plotting ln (C_o/C_t) versus irradiation time (t), which indicates that the kinetics of the photocatalytic decolourization of MB by both non-porous and porous TiO₂/ENR films was fitted to the Langmuir–Hinshelwood model, as given in Fig. 11. The calculated value of the pseudo-first-order rate constant shows that the photocatalytic decolourization of MB by porous TiO₂/ENR film becomes approximately twice as fast as the non-porous TiO₂/ENR film, and the pseudo-first-order rate constant value was increased from 0.048 min⁻¹ to 0.094 min⁻¹. The kinetic result was in good agreement with the percentage of MB colour removal. This enhancement of the photocatalytic decolourization and the percentage of decolourization of MB by porous TiO₂/ENR film is mainly attributed to the generation of the macro-pores on the surface of the TiO₂/glass system. This generation of the macro-pores on the TiO₂ surface offers better diffusion of the MB molecules and hydroxyl radicals across the inner layer of the TiO₂ [11].

Fig. 10 Percentage of MB colour removal during the adsorption, direct photolysis and photocatalytic processes (experimental condition: [MB]_o: 20 mg L⁻¹, aeration rate: 50 mL min⁻¹, TiO₂ loading 2.62 mg cm⁻² and pH 8).

Fig. 11 A plot of pseudo-first-order rate constant of the photocatalytic decolourization of 20 mL 20 mg L⁻¹ MB by non-porous and porous TiO₂/ENR film (experimental condition: [MB]_o: 20 mg L⁻¹, aeration rate: 50 mL min⁻¹, TiO₂ loading 2.62 mg.cm⁻² and pH 8).

3.6 Mineralization study

The heterogeneous photocatalytic process usually involves partial or complete mineralization of the organic substances by the oxidizing species that exist on the surface of titanium oxide. In this study, the main advantage of using ENR as an organic binder for immobilization of the TiO₂ photocatalyst is that it can be regularly decomposed by the photocatalytic reaction to generate pores and enhance the surface area of the immobilized TiO₂ film. Therefore, further verification of the photocatalytic degradation of this ENR binder is required. However, the photocatalytic decolourization of MB by the porous TiO₂/ENR film does not necessarily indicate a full mineralization of the MB into CO₂ and H₂O and some types of mineral acids. In some circumstances, the intermediates produced during the photocatalytic reaction could be even more toxic than the starting materials. Therefore, the chemical oxygen demand (COD) test was conducted as a function of the presence of dissolved organic matter that is attributed to the degradation of ENR within the TiO₂ film as well as MB and its by-products during the photocatalytic reaction. In the first step, the immobilized TiO₂ film was immersed in ultra-pure water under the irradiation of a 55-W compact fluorescent lamp. After 1 h of irradiation, the water sample was withdrawn and replaced with fresh ultra-pure water using the same immobilized TiO₂ film until 6 h of irradiation. The withdrawn water samples were subjected to the COD test. Fig. 12(a) presents the detected COD values of the treated water due to the degraded part of the ENR polymer inside the immobilized TiO₂ film system. In the second step, the irradiated TiO₂ film that was obtained from step 1 was taken to evaluate the ability of the porous TiO₂ in the mineralization of 20 mg/L MB and its intermediates, as shown in Fig. 10(b). It can be observed that MB and its main by-products were mostly mineralized after 5 h of irradiation using an immobilized porous TiO₂ film. This finding can be attributed to the porous structure of the immobilized TiO₂ film, which led to enhanced MB and hydroxyl radical diffusion into the inner layers of the TiO₂ film [11]. In fact, once the porous TiO₂ film is illuminated with a 55-W fluorescent lamp with a UV leakage of 7.32 Wm⁻², the electrons are excited from the lower valence band (VB) towards the higher conduction band (CB), which leads to the formation of a paired positive hole and negative electron on the surface of the TiO₂ film. The absorbed oxygen and molecular water on the surface of the TiO₂ film can directly react with the photogenerated electron-hole pair to give O₂•⁻, HO₂• or •OH as products (Eqs. (5)–(8)) [24,25]:(5) ${\text{TiO}}_2+h\upsilon \to {\text{TiO}}_2({\text{e}}_{\text{cb}}^{-}+h_{\upsilon \text{b}}^{+})$(5) (6) ${\text{TiO}}_2(h_{\upsilon \text{b}}^{+})+{\text{H}}_2{\text{O}}_{\text{ads}}\to {\text{TiO}}_2+{\text{H}}^{+}+{\text{OH}}^{-}$(6) (7) ${\text{TiO}}_2({\text{e}}_{\text{cb}}^{-})+{\text{O}}_{2\text{ads}}\to {\text{TiO}}_2+{\text{O}}_{2\text{ads}}^{-}$(7) (8) ${\text{TiO}}_2(h_{\upsilon \text{b}}^{+})+{\text{OH}}^{-}\to {\text{TiO}}_2+{\text{OH}}^{\text{•}}$(8)

Fig. 12 COD values of (a) treated ultra-pure water sample of irradiated TiO₂/ENR film, and (b) of the 20 mg/L irradiated in the presence of TiO₂/ENR film.

The ^.OH radicals that are derived from an irradiated TiO₂ film (Eq. (8)) would directly oxidize MB dye in the bulk solution, as in Eq. (9)(9) $\text{MB}+{\text{OH}}^{\text{•}}\to \text{oxidation products}$(9) :(9) $\text{MB}+{\text{OH}}^{\text{•}}\to \text{oxidation products}$(9)

4 Conclusions

The kinetic characteristics of the photocatalytic decolourization of MB using an immobilized TiO₂/glass were experimentally determined with respect to the different TiO₂ loading, pH conditions and surface morphology of the TiO₂ film. A kinetic model was proposed according to the operational parameters. The apparent pseudo-first-order rate constant increased when there was an increase in the loaded amount of TiO₂ formulation on the glass surface and remained at almost a plateau after reaching the optimal amount. The same observation was obtained by increasing the pH value towards basic media. It was also found that depletion of the ENR binder from the TiO₂ formulation during the photocatalytic process plays a significant role in changing the surface morphology of TiO₂ to be more porous and enhancing the diffusion of the MB molecules and increasing the hydroxyl radical penetration to reach the inner layer of the immobilized TiO₂. The apparent pseudo-first-order rate constant and decolourization of MB using the porous structure of the TiO₂ film became almost two times faster than for the non-porous TiO₂.

Acknowledgments

The authors would like to thank the Faculty of Applied Sciences, Universiti Teknologi MARA, Arau Campus, for facilitating this work. Special thanks go to Miss Nur Lia Ali for facilitating the TGA analysis.

Notes

Peer review under responsibility of Taibah University.