Study on the lifetime of photocatalyst by photocatalytic membrane reactors (PMR)

The superior oxidizing ability and the utilization of solar energy are the main advantages of the photocatalytic degradation process (Sun et al. 2003; Wang et al. 2019). Photocatalyst deactivation is an important reason to hinder industrial applications of the photocatalytic degradation process. The researches on photocatalyst deactivation problems mainly include: photocatalyst lifetime, deactivation mechanism and regeneration method. In fact, the mechanisms of photocatalyst deactivation and regeneration have been widely studied, organic pollutants occupying the adsorption sites of the photocatalyst are considered as the deactivation mechanism (Jing et al. 2004), and the regeneration methods to solve the deactivation problem include: washing, oxidation, and thermal treatment (Gandhi et al. 2012). However, researches on photocatalyst lifetime were rarely studied, and no researches on the factors of photocatalyst lifetime have been found.

To study photocatalyst lifetime and its factors, it is necessary to use a continuous photocatalytic reactor. Photocatalytic membrane reactor (PMR) is a new type of continuous photocatalytic reactor. PMRs are hybrid reactors in which photocatalysis is coupled with a membrane process (Mozia 2010; Yan et al. 2018). Membrane filtration can resolve the water/catalyst separation problem and maintain the high quality of the TiO₂ constant in the photocatalytic reactor (Shon et al. 2008; Boonya-atichart et al. 2018; Gao et al. 2018; Rani & Karthikeyan 2018). Photocatalyst deactivation rep­resents a major drawback that hinders many industrial applications. Therefore, this study intends to change the photocatalytic reaction conditions in PMR to determine the effect of different reaction conditions on the lifetime of the photocatalyst. The study of the factors affecting the lifetime of the photocatalyst can optimize the photocatalytic reaction conditions and extend the lifetime of the photocatalyst, which is significant to the photocatalytic degradation process.

In this paper, the methyl orange (MO) dye was treated by PMR until the photocatalyst deactivated. The deactivation mechanism was understood by Fourier transform infrared spectroscopy (FTIR) and Brunauer-Emmett-Teller (BET) surface area analysis. The regeneration effects of methanol cleaning, hydrogen peroxide oxidation, and calcination regeneration were investigated. In addition, the effects of reaction conditions on the lifetime of the photocatalyst were investigated by changing the concentration of TiO₂ and MO, hydraulic retention time (HRT) and UV light intensity in PMR. The effects of changing the TiO₂ concentration, MO concentration, HRT and irradiance of the UV lamp in the PMR on the lifetime of the photocatalyst were investigated.

The cylindrical PMR had a volume of 2.16 L and a 33 μW/cm² UV lamp with a wavelength of 254 nm. The schematic diagram of the PMR setup is shown in Figure 1. A hollow microfiltration (MF) membrane module was immersed in the PMR. Polyethylene (PE) MF membranes (Mitsubishi, Japan) had a pore size of 0.1 μm, an inner diameter of 0.27 mm, an outer diameter of 0.41 mm and an effective filtering area of 0.06 m². Titanium dioxide (P25 TiO₂, Degussa, Germany) was the photocatalyst, and it was suspended in the PMR. The influent was mixed with the photocatalyst in the PMR and continuously filtered by a suction pump. The water level of the PMR was maintained with a level sensor that controlled the influent pump and valve. Air was pumped through a gas diffuser placed at the bottom of the reactor to provide dissolved oxygen, prevent the accumulation of TiO₂ particles on the membrane surface, and mix the dye and photocatalyst. In order to obtain the MO concentration in the effluent of the PMR, we measured the absorbance value of the effluent with the UV spectrophotometer at the wavelength of 465 nm, using distilled water as the background.

The spent TiO₂ was collected from the PMR after the photocatalyst was completely deactivated by sedimentation and filtration and then dried at 50 °C. The spent TiO₂ catalysts were regenerated via three methods. The spent catalyst was washed with methanol (0.1 g of catalyst per 10 ml of solution) and then dried at 50 °C in a vacuum drying oven for 2 h; this sample was designated as ST_W. After oxidation using 30% H₂O₂ solution (0.1 g of catalyst per 1 ml of solution), the spent catalyst was dried at 100 °C in a vacuum drying oven for 20 min; this sample was designated as ST_O. Lastly, the spent catalyst was calcinated in air at 420 °C for 3 h in a muffle furnace, this sample was designated as ST_C.

The regenerated photocatalysts were transferred into the photocatalytic batch reactor for the regeneration performance test. A 33 μW/cm² UV lamp with a wavelength of 254 nm was placed inside the reactor. The photocatalytic batch reactor was equipped with 2.16 L of 10 mg/L MO solution and an air diffuser, which provided oxygen for the photocatalytic reaction. The concentration of the TiO₂ catalyst was 0.5 g/L.

All samples were characterized by FTIR, and BET surface area analysis. To identify any changes in the surface functional groups, FTIR was performed using a spectrometer (BRUKER, TENSOR27). The samples were mixed with KBr powder at a ratio of 1:100, evenly ground in an agate mortar with a pestle, and then pressed for measurement. The infrared band used for determination was 400–4,000 cm⁻¹. The surface area of the sample was measured using a BET specific surface area tester (Micromeritics Instrument Corp., Tristar II 3020, USA).

The continuous photocatalytic degradation of MO was carried out in a PMR. The concentration of MO in the effluent of the PMR increased with the increase in the volume of treated MO. The photocatalyst was deactivated when the concentration of MO in the effluent of the PMR was stable (Yan et al. 2016). To characterize the lifetime of the photocatalyst, we applied g MO/g TiO₂. Approximately 1.12 g of MO was destroyed in the presence of 2.16 g of photocatalyst before complete catalyst deactivation. The lifetime of the photocatalyst was 0.52 g MO/g TiO₂. The concentrations of TiO₂, and MO were 1 g/L and 10 mg/L, respectively. The HRT was 0.75 h.

The spent TiO₂ catalyst was recovered from the PMR by sedimentation and filtration. Subsequently, it was regenerated by washing with methanol and treating with H₂O₂ and heat and reused for the photocatalytic degradation of MO as shown in Figure 2. The spent TiO₂, which was regenerated by washing with methanol (ST_W) and treating with heat (ST_C), degraded slowly, resulting in an MO degradation of 12.44% and 17.50% after 120 min, respectively. Compared to the washed spent catalyst (ST_W) and the heat-treated catalyst (ST_C), the regenerated photocatalyst treated with H₂O₂ (ST_O) showed slightly faster removal of MO, resulting in an MO degradation of 57.12% after 120 min. These results indicate that the catalyst could be regenerated to some extent by H₂O₂ treatment, which may have been due to the fact that H₂O₂, a strong oxidizing agent, provided plenty of ·OH radicals. The organic species adsorbed in the spent photocatalyst were oxidized by ·OH radicals into CO₂ and H₂O, thereby freeing all the surface active sites (Gandhi et al. 2012).

The regeneration recovery rate is defined as the ratio of the MO removal rate after 120 min by the regenerated photocatalyst and the MO removal rate after 120 min by the fresh photocatalyst. The regeneration recovery rates were 13.16%, 18.52%, and 60.43% for the ST_W catalyst, ST_C catalyst, and ST_O catalyst, respectively, indicating that the photocatalyst could not be completely regenerated by washing with solvents or treating with H₂O₂ or heat.

Fresh TiO₂ is a white powder, whereas spent TiO₂ is a light yellow powder. Furthermore, the ST_O catalyst turned white, and the ST_W and ST_C catalysts remained light yellow. The fresh, spent and regenerated TiO₂ catalysts were characterized by BET surface area analysis and FTIR.

The BET surface areas of the fresh TiO₂, spent and regenerated TiO₂ by washing, heat treatment, and H₂O₂ treatment were 59.86 m²/g, 58.87 m²/g, 59.66 m²/g, 60.19 m²/g, 58.40 m²/g, respectively. Compared to the fresh catalyst, no significant change was observed in the BET surface area of the spent catalyst, revealing that photocatalyst deactivation was not due to pore blocking by the reactant (MO) or intermediate products. Gandhi et al. (2012) obtained similar results when the spent photocatalyst for the degradation of phthalic acid was regenerated by washing and treating with H₂O₂ and heat.

The FTIR spectra of the fresh, spent, and regenerated TiO₂ are shown in Figure 3. In the spent TiO₂ spectrum, a new peak with strong intensity appeared at 1,262 cm⁻¹, which assigned the C–O stretch in –C–OH. These results were consistent with those of Kaewgun & Lee (2010), who evaluated the deactivation of polymorphic brookite titania by observing the photocatalytic degradation of MO. FTIR peaks, which can be assigned for the C–O stretch in –C–O–C– and –C–OH, were identified, and their relative intensities increased with the increase in the reaction cycle number. In a study by Li et al. (2012a), product formation of MO was measured by liquid chromatography mass/spectrometry (LC/MS); the results indicate –C–OH existing in the products. These results indicate that the deactivation of TiO₂ was due to the deposition of the decomposed MO products on the TiO₂.

From the FTIR spectra of the regenerated catalyst, the intensity of the FTIR peak at 1,262 cm⁻¹ decreased, and it completely disappeared in the ST_O photocatalyst. The decrease in the intensity of the peak was consistent with the regeneration recovery rate by different regeneration methods. ST_W and ST_C photocatalysts exhibited a lower regeneration rate, which was due to deposited species that could not be removed by washing or treating with heat. However, H₂O₂ oxidized the species on the surface of the photocatalyst, so the ST_O photocatalyst exhibited a higher regeneration rate.

When the initial concentration of organic pollutants is small, the Langmuir-Hinshelwood equation can be simplified as a quasi first-order kinetic equation. The effects of TiO₂, MO concentration and light intensity on the photocatalytic degradation process were analyzed based on quasi first-order reaction kinetics. At the same time, the effect of the above factors on lifetime was investigated through the dynamic reaction in the PMR.

The concentration of TiO₂ is important in treatment. The TiO₂ concentration is directly proportional to the reaction rate (Feilizadeh et al. 2015; Hosseini et al. 2016). The effect of the photocatalyst concentration on the photocatalytic degradation of MO was investigated as shown in Figure 4(a). Before the suspension concentration of TiO₂ increased to 3 g/L, the photocatalytic degradation rate increased with the increase in the concentration of TiO₂. The reason was that the number of active sites increased with the increase in the dosages of TiO₂, while the photocatalytic degradation slightly decreased when the suspension concentration of TiO₂ was 4 g/L. It is well known that the rate of photon uptake by TiO₂ should depend on the amount of light transmitted to the suspension, which can be influenced by the concentration of TiO₂ (Huang et al. 2008). This attributed that the high TiO₂ concentration led to the increased light shielding and scattering as well as the loss in surface area occasioned, which was caused by TiO₂ particle agglomeration (Kaneco et al. 2009).

The effect of the TiO₂ concentration on the photocatalyst lifetime was investigated using the PMR, as shown in Figure 4(b). The MO concentration in the effluent of the PMR was stable after treating with 57.6 L of a 10 mg/L MO solution with a TiO₂ concentration of 0.5 g/L. The lifetime of the photocatalyst was 0.53 g MO/g TiO₂. The MO concentration in the effluent of the PMR was stable after treating with 112.3 L of a 10 mg/L MO solution with a TiO₂ concentration of 1 g/L. The lifetime of the photocatalyst was 0.52 g MO/g TiO₂. These results show that the lifetime of the photocatalyst was similar under different photocatalyst concentrations, indicating that the concentration of the photocatalyst had few effects on the lifetime of the photocatalyst.

The effect of the initial MO concentration on the photocatalytic degradation rate was investigated over a concentration range of 5–20 mg/L of MO. Figure 5(a) shows that the degradation rate decreased with the increase in the initial MO concentration. There may be two reasons: firstly, when the dosage of photocatalyst is fixed, its specific surface area is also fixed. The increased occupation of TiO₂ active sites by contaminants hindered generation of the oxidants, such as ·OH, ·O₂⁻, which led to the reduction of the degradation rate. Secondly, more photons were absorbed, resulting from a higher concentration of MO, which caused the number of effective photons reaching the surface of TiO₂ to be reduced, and then the reaction rate is reduced (Niu & Hao 2013).

We used the PMR to investigate the effect of initial MO concentration on the lifetime of the photocatalyst as shown in Figure 5(b). The MO concentration in the effluent of the PMR was stable after treating with 60.5 L of a 5 mg/L MO solution, and the lifetime of the photocatalyst was 0.14 g MO/g TiO₂. The MO concentration in the effluent of the PMR was stable after treating with 112.3 L of a 10 mg/L MO solution, and the lifetime of the photocatalyst was 0.52 g MO/g TiO₂. The lifetime of the photocatalyst after treatment with 10 mg/L MO was 3.71 times that of 5 mg/L MO, indicating that increasing concentrations of MO can markedly increase the lifetime of the photocatalyst. It maybe attributed that the efficiency of the photocatalyst per unit mass was enhanced under high concentrations of MO.

The irradiance has a direct impact on the photocatalytic degradation rate. Figure 6(a) shows that the degradation rate increased linearly within a light intensity of 10–72 μW/cm², which was consistent with the view that the rate of degradation typically changes linearly with the intensity up to a certain value beyond which the reaction shows a square root dependency on the intensity (Gogate & Pandit 2004). Since the TiO₂ particles were suspended in a stirred solution, the degree of absorption of light by the TiO₂ particle surface increased with the increase of light intensity (Kaneco et al. 2009). This is because the number and rate of photons supplied to the semiconductor surface will be higher when the irradiation intensity of light was higher (Julson & Ollis 2006).

As shown in Figure 6(b), we investigated the effect of light intensity on the lifetime of the photocatalyst in the PMR. The MO concentration in the effluent of the PMR was stable after treating with 25.9 L MO solution under the 10 μW/cm² light intensity, and the lifetime of the photocatalyst was 0.12 g MO/g TiO₂. The MO concentration in the effluent of the PMR was stable after treating with 112.3 L MO solution under the 33 μW/cm² light intensity, and the lifetime of the photocatalyst was 0.52 g MO/g TiO₂. The MO concentration in the effluent of the PMR was stable after treating with 164.2 L MO solution under the 72 μW/cm² light intensity, and the lifetime of the photocatalyst was 0.76 g MO/g TiO₂. The lifetime of the photocatalyst after treatment under 72 μW/cm² light intensity was 1.46 and 6.33 times that under 33 and 10 μW/cm² light intensity. It is thus clear that the effect of light intensity on the lifetime of the photocatalyst is very significant.

The HRT, an important parameter of the PMR (Moradi & Nasrollahpour 2017), was changed by altering the membrane flux. Figure 7 shows the effect of the HRT on the lifetime of the photocatalyst.

At an HRT of 0.75 h, the photocatalyst was deactivated after treating with 119.5 L of an MO solution, and the MO concentration in effluent was 8.30 mg/L. At an HRT of 3 h, the MO concentration in effluent was less than 2 mg/L; after treating with 119.5 L of the MO solution, it increased slowly. The experiment was terminated after treating with 220 L of MO solution due to the slow increase, and the lifetime of the photocatalyst was not calculated. Furthermore, changing the HRT from 0.75 to 3 h prolonged the lifetime of the photocatalyst, which may have been due to the complete photodegradation of MO and the intermediate products, which could prevent photocatalyst deactivation. Meng et al. (2005) reported that organic carbon could be totally mineralized as long as a relatively longer HRT was applied in the PMR.

The deactivation of TiO₂ was caused by the deposition of MO products but not by the blocking of pores by the reactant (MO) or intermediate products. H₂O₂ treatment yielded the highest regeneration rate over washing with methanol and treating with heat, because H₂O₂ oxidized the deposited species on the surface of the photocatalyst. The MO concentration, the irradiance of the UV lamp and the HRT of the PMR were found to affect the lifetime of the photocatalyst more significantly than the TiO₂ concentration. This may have been due to the fact that when the HRT was long, the MO and intermediate products adsorbed on the surface of the photocatalyst were thoroughly photodegraded, which prevented photocatalyst deactivation.

This work was supported by the Fundamental Research Funds for the Central Universities (2020B1041X), National Science Foundation of China (No. 51508383, 51741807 and 51878357), and the Natural Science Foundation of Tianjin Province (18JCQNJC09000).