Effects of Hydrothermal Time on Structure and Photocatalytic Property of Titanium Dioxide for Degradation of Rhodamine B and Tetracycline Hydrochloride

Using butyl titanate and absolute ethanol as raw materials, TiO2 was prepared by a hydrothermal method with different hydrothermal times, and the influences of hydrothermal time on the structure and photocatalytic performance of TiO2 were investigated. The obtained samples were characterized by XRD, SEM, TEM, BET, PL and DRS, separately. The results show that TiO2 forms anatase when the hydrothermal time is 12 h, forms a mixed crystal composed of anatase and rutile when the hydrothermal time is 24 h, and forms rutile when the hydrothermal time is 36 h. With the extension of hydrothermal time, anatase gradually transforms into rutile and the surface area decreases. Although TiO2-24 h and TiO2-36 h show lower photoinduced charge recombination and higher light source utilization, TiO2-12 h exhibits the highest photocatalytic activity owing to its largest surface area (145.3 m2/g). The degradation degree of rhodamine B and tetracycline hydrochloride reach 99.6% and 90.0% after 45 min.


Introduction
Employing photocatalytic technology to degrade pollutants is an effective route for environmental governance. TiO 2 has received extensive attention due to its advantages such as chemical stability, low cost, mild reaction conditions and high photocatalytic activity [1][2][3][4][5]. The crystal structure, crystallinity, surface morphology, specific surface area and optical property of TiO 2 are closely related to the preparation method. The most commonly used methods are mainly sol-gel [6][7][8] and hydrothermal methods [9][10][11][12]. Zhu et al. [6] used sol-gel methods to synthesize TiO 2 under calcination at 540 • C. The particles are spherical with a certain extent agglomeration and the surface area is 40 m 2 /g. It is convenient to control the morphology of photocatalyst by a hydrothermal method, which does not require high-temperature calcination and is conducive to obtaining a large surface area and high photocatalytic activity [13][14][15][16][17][18]. Esparza et al. [16] prepared nanostructured TiO 2 by a hydrothermal method. The crystal grain size was 13 nm, and the particles were composed of nanotubes and nanosheets with a large surface area (269 m 2 /g). Methylene blue (MB) was completely degraded under UV light after 120 min. Zhu et al. [13] prepared Cu-doped TiO 2 under the conditions of 200 • C for 12 h by hydrothermal method. It was found that the photocatalyst was anatase/rutile mixed crystal structure and the surface area was 73.9 m 2 /g. The decolorization degree of rhodamine (RhB) was 99.4% after 60 min. Nesic et al. [18]. prepared lanthanum and vanadium co-doped titanium dioxide by a microwave-assisted hydrothermal method. The samples showed high crystallinity and were all anatase structures. The specific surface area of 0.02V-2La/TiO 2 was 125 m 2 /g, and the decolorization degree of RhB was more than 90% after 90 min.
Hydrothermal conditions will affect the crystal structure, surface morphology and specific surface area, thereby affecting the photocatalytic performance [19,20]. Lee et al. [19] studied the effects of hydrothermal temperature from 120 to 210 • C on the morphology and photocatalytic performance of TiO 2 . The results show that the particles gradually change from granular to nanotube with the increasing temperature and TiO 2 prepared at 180 • C displays the best photocatalytic performance. Changing the ratio of reactants can also affect the structure and photocatalytic performance of the products [21,22]. Li et al. [21] prepared TiO 2 by using different molar ratios of tartaric acid to TiCl 3 . When the tartaric acid: TiCl 3 is 0.1, the photocatalytic activity is the highest.
In the present study, at the fixed hydrothermal temperature 200 • C, TiO 2 photocatalysts were prepared with a hydrothermal time of 12, 24 and 36 h, respectively. The obtained samples were characterized by XRD, SEM, TEM, BET, PL, DRS and the degradation of RhB and tetracycline hydrochloride (TC) to study the effects of hydrothermal time on the crystal structure, morphology, specific surface area, optical property and photocatalytic performance of TiO 2 .

Characterization
The crystal structure was characterized by DX-2700 X-ray diffractometer (XRD). Morphology was observed using Hitachi SU8220 scanning electron microscope (SEM) and FEI-Tecnai G2 F20 transmission electron microscope (TEM). Surface area was measured by an ASAP2460 surface area analyzer (BET). The optical property was studied using UV-3600 ultraviolet-visible spectrophotometer (DRS) and F-4600 fluorescence spectrometer (PL).

Photocatalysis Experiment
To achieve adsorption and desorption equilibrium, 0.1 g TiO 2 powder and 100 mL (10 mg/L) RhB solution or 100 mL (30 mg/L) TC solution were mixed and then stirred 30 min in the dark. Using a 250 W xenon lamp as the light source, the mixture was taken every 15 min to measure the absorbance and the degradation degree was calculated by the formula (A 0 − A t )/A 0 × 100%. Figure 1 exhibits the XRD patterns of samples. All peaks in TiO 2 -12 h correspond to anatase structure, indicating TiO 2 forms anatase when the hydrothermal time is 12 h. The peak intensity of anatase (101) plane in TiO 2 -24 h drops sharply, and the peaks of anatase (004), (200), (105), (204) planes disappear. Only the peaks around 25.3 and 48.1 • ascribing to the (101) and (200) crystal planes of anatase structure can be detected. Meanwhile, the rutile diffraction peaks appear and the intensity is much higher than anatase, implying that TiO 2 -24 h forms anatase/rutile mixed crystal structure. The mass fraction of anatase (X A ) can be calculated by the following formula [2,21]:

Crystal Structure
where I A and I R represent the intensities of anatase (101) plane and rutile (110) plane, respectively. The mass fraction of anatase phase is 7.1% and the mass fraction of rutile phase is 92.9% in TiO 2 -24 sample. The anatase diffraction peaks in TiO 2 -36 h disappear wholly and all the peaks can be attributed to rutile, which indicates that the transformation where IA and IR represent the intensities of anatase (101) plane and rutile (110) plane, respectively. The mass fraction of anatase phase is 7.1% and the mass fraction of rutile phase is 92.9% in TiO2-24 sample. The anatase diffraction peaks in TiO2-36 h disappear wholly and all the peaks can be attributed to rutile, which indicates that the transformation from anatase to rutile has been completed when hydrothermal time is 36 h [23,24]. The grain sizes (D) of samples were calculated by the Scherrer formula [2]: where λ represents the wavelength of Cu Ka, β represents the full width at half maximum of the XRD peak ((101) plane for anatase and (110) plane for rutile), and 2θ represents the Bragg diffraction angle. The grain sizes (D) of TiO2-12, TiO2-24 and TiO2-36 are 9.9 nm, 13.2 nm (anatase)/35.0 nm (rutile) and 27.5 nm.      Figure 3a, the particles are relatively dispersed and the size of a single particle is around 10 nm. The interplanar spacing in Figure 3b is 0.351 nm, corresponding to the (101) crystal plane of anatase [5]. It is observed from Figure  3c that the size of a single particle is 15-30 nm, which is larger than that of TiO2-12 h. In addition to the granular shape, several particles exhibit rod and block shapes. The length of the nanorods is about 50 nm and the width is 15 nm. The size of the blocks is 50-100   Figure 3a, the particles are relatively dispersed and the size of a single particle is around 10 nm. The interplanar spacing in Figure 3b is 0.351 nm, corresponding to the (101) crystal plane of anatase [5]. It is observed from Figure  3c that the size of a single particle is 15-30 nm, which is larger than that of TiO 2 -12 h. In addition to the granular shape, several particles exhibit rod and block shapes. The length of the nanorods is about 50 nm and the width is 15 nm. The size of the blocks is 50-100 nm. In Figure 3d, the marked interplanar spacing 0.348 nm corresponds to the (101) crystal plane of anatase and 0.320 nm corresponds to the (110) crystal plane of rutile [13,20], indicating that TiO 2 -24 h is a mixed crystal composed of anatase and rutile, which is in line with XRD results. Nanoparticle almost disappears in Figure 3e and the particles are completely made of rods and blocks. The length of rods is 100 nm and the width is approximately 20 nm. The size of the blocks is around 120 nm. The interplanar spacing marked in Figure 3f is 0.322 nm, corresponding to the (110) crystal plane of rutile.   Figure 3a, the particles are relatively dispersed and the size of a single particle is around 10 nm. The interplanar spacing in Figure 3b is 0.351 nm, corresponding to the (101) crystal plane of anatase [5]. It is observed from Figure  3c that the size of a single particle is 15-30 nm, which is larger than that of TiO2-12 h. In addition to the granular shape, several particles exhibit rod and block shapes. The length of the nanorods is about 50 nm and the width is 15 nm. The size of the blocks is 50-100 nm. In Figure 3d, the marked interplanar spacing 0.348 nm corresponds to the (101) crystal plane of anatase and 0.320 nm corresponds to the (110) crystal plane of rutile [13,20], indicating that TiO2-24 h is a mixed crystal composed of anatase and rutile, which is in line with XRD results. Nanoparticle almost disappears in Figure 3e and the particles are completely made of rods and blocks. The length of rods is 100 nm and the width is approximately 20 nm. The size of the blocks is around 120 nm. The interplanar spacing marked in Figure 3f is 0.322 nm, corresponding to the (110) crystal plane of rutile.  The morphology of TiO2 has a great impact on surface area and adsorption performance. It is found in Figure 3 that with the extension of hydrothermal time, the single particle size increases and the morphology changes significantly, which may lead to the surface area difference. To clarify the influences of hydrothermal time on surface area and the porosity of samples, the textural properties of samples have been implemented and the results are shown in Figure 4 and Table 1. Both TiO2-12 h and TiO2-24 h are mesoporous materials. The pore size distribution curve of TiO2-12 h shows a narrow peak, and its pore size distribution is uniform, and the pore size is between 5-15 nm. The pore size distribution curve of TiO2-24 h shows a broad peak shape, and the pore size distribution is uneven with a size of 5-50 nm. There is no peak in the pore size distribution curve of TiO2-36 h, indicating that no obvious mesopores can be detected in TiO2-36 h. The BET surface area, pore volume and average pore size of samples are summarized in Table 1. As the hydrothermal time increases, the BET surface area decreases from 145.3 m 2 /g to 43.0 m 2 /g and 13.3 m 2 /g, and the pore volume is reduced from 0.264 cm³/g to 0.107 cm³/g and 0.029 cm³/g. TEM images show that as hydrothermal time increases from 12 to 24 h, part of fine nanoparticles aggregate to form nanorods and nanoblocks. When the reaction The morphology of TiO 2 has a great impact on surface area and adsorption performance. It is found in Figure 3 that with the extension of hydrothermal time, the single particle size increases and the morphology changes significantly, which may lead to the surface area difference. To clarify the influences of hydrothermal time on surface area and the porosity of samples, the textural properties of samples have been implemented and the results are shown in Figure 4 and Table 1. Both TiO 2 -12 h and TiO 2 -24 h are mesoporous materials. The pore size distribution curve of TiO 2 -12 h shows a narrow peak, and its pore size distribution is uniform, and the pore size is between 5-15 nm. The pore size distribution curve of TiO 2 -24 h shows a broad peak shape, and the pore size distribution is uneven with a size of 5-50 nm. There is no peak in the pore size distribution curve of TiO 2 -36 h, indicating that no obvious mesopores can be detected in TiO 2 -36 h. The BET surface area, pore volume and average pore size of samples are summarized in Table 1. As the hydrothermal time increases, the BET surface area decreases from 145.3 m 2 /g to 43.0 m 2 /g and 13.3 m 2 /g, and the pore volume is reduced from 0.264 cm 3 /g to 0.107 cm 3 /g and 0.029 cm 3 /g. TEM images show that as hydrothermal time increases from 12 to 24 h, part of fine nanoparticles aggregate to form nanorods and nanoblocks. When the reaction time is 36 h, all the nanoparticles aggregate to form nanorods and nanoblocks. The agglomeration phenomenon is further intensified, and the BET surface area and pore volume are reduced. rous materials. The pore size distribution curve of TiO2-12 h shows a narrow peak, and its pore size distribution is uniform, and the pore size is between 5-15 nm. The pore size distribution curve of TiO2-24 h shows a broad peak shape, and the pore size distribution is uneven with a size of 5-50 nm. There is no peak in the pore size distribution curve of TiO2-36 h, indicating that no obvious mesopores can be detected in TiO2-36 h. The BET surface area, pore volume and average pore size of samples are summarized in Table 1. As the hydrothermal time increases, the BET surface area decreases from 145.3 m 2 /g to 43.0 m 2 /g and 13.3 m 2 /g, and the pore volume is reduced from 0.264 cm³/g to 0.107 cm³/g and 0.029 cm³/g. TEM images show that as hydrothermal time increases from 12 to 24 h, part of fine nanoparticles aggregate to form nanorods and nanoblocks. When the reaction time is 36 h, all the nanoparticles aggregate to form nanorods and nanoblocks. The agglomeration phenomenon is further intensified, and the BET surface area and pore volume are reduced.   Controlling the morphology of TiO 2 and increasing its specific surface area is a research hotspot [25][26][27]. Du et al. [25] prepared porous Sn-doped TiO 2 using polystyrene microspheres as a template. After calcination to remove the template, the porous structure was fabricated and the surface area reached 71.1 m 2 /g. Huang et al. [26] used a sol-gel method combined with a hydrothermal method to prepare TiO 2 microspheres by a two-step reaction. The diameter of the microspheres was about 200-500 nm and the surface area was 91.1 m 2 /g. TiO 2 nanospheres with diameter of 50-100 nm and surface area of 70.0 m 2 /g were fabricated by hydrothermal method in Mohamed et al.'s work [27]. In the present study, TiO 2 -12 h exhibits a relatively large surface area (145.3 m 2 /g), which may result in high photocatalytic activity. Figure 5 shows the UV-visible absorption spectra of samples. The absorption edge of TiO 2 -12 h is 387 nm [28,29], which corresponds to anatase structure. The absorption edges are 403 and 402 nm for TiO 2 -24 h and TiO 2 -36 h, respectively. The band gap of rutile is smaller than anatase, thus TiO 2 -24 h and TiO 2 -36 h, which mainly consist of rutile, show red shift compared to TiO 2 -12 h.

Optical Property
The PL peaks are derived from the recombination of photogenerated electrons and holes, thus the lower peak intensity and the lower recombination rate [30,31]. The PL spectra of samples are shown in Figure 6. It is generally believed that the recombination rate of anatase is lower than rutile [32,33], however, the PL peak intensity of TiO 2 -12 h is the highest in the present work. XRD results show that the peak intensity of TiO 2 -12 h is low and the half-height width of peak is large, indicating that TiO 2 -12 h displays poor crystallinity with plentiful defects and oxygen vacancies. Photoinduced charges will be captured by defects and oxygen vacancies, which is in favor of retarding the recombination. However, excess defects or oxygen vacancies will introduce new recombination centers, enhancing the PL peak intensity [34]. With the increase in hydrothermal time, the crys-tallinity of the sample improves and the defects and oxygen vacancies reduce. Moderate defects and oxygen vacancies are beneficial to the separation of photogenerated electrons and holes [35,36]. Therefore, the PL peak intensity of TiO 2 -24 h and TiO 2 -36 h is lower than TiO 2 -12 h. step reaction. The diameter of the microspheres was about 200-500 nm and the surface area was 91.1 m 2 /g. TiO2 nanospheres with diameter of 50-100 nm and surface area of 70.0 m 2 /g were fabricated by hydrothermal method in Mohamed et al.'s work [27]. In the present study, TiO2-12 h exhibits a relatively large surface area (145.3 m 2 /g), which may result in high photocatalytic activity. Figure 5 shows the UV-visible absorption spectra of samples. The absorption edge of TiO2-12 h is 387 nm [28,29], which corresponds to anatase structure. The absorption edges are 403 and 402 nm for TiO2-24 h and TiO2-36 h, respectively. The band gap of rutile is smaller than anatase, thus TiO2-24 h and TiO2-36 h, which mainly consist of rutile, show red shift compared to TiO2-12 h. The PL peaks are derived from the recombination of photogenerated electrons and holes, thus the lower peak intensity and the lower recombination rate [30,31]. The PL spectra of samples are shown in Figure 6. It is generally believed that the recombination rate of anatase is lower than rutile [32,33], however, the PL peak intensity of TiO2-12 h is the highest in the present work. XRD results show that the peak intensity of TiO2-12 h is low and the half-height width of peak is large, indicating that TiO2-12 h displays poor crystallinity with plentiful defects and oxygen vacancies. Photoinduced charges will be captured by defects and oxygen vacancies, which is in favor of retarding the recombination. However, excess defects or oxygen vacancies will introduce new recombination centers, enhancing the PL peak intensity [34]. With the increase in hydrothermal time, the crystallinity of the sample improves and the defects and oxygen vacancies reduce. Moderate defects and oxygen vacancies are beneficial to the separation of photogenerated electrons and holes [35,36]. Therefore, the PL peak intensity of TiO2-24 h and TiO2-36 h is lower than TiO2-12 h. The PL main peak originates from photogenerated electrons in the conduction band returning directly to the valence band and recombining with holes, therefore, the wavelength corresponding to main peak in PL spectra is related to the band gap [37,38]. The main peak wavelengths of TiO2-12 h, TiO2-24 h and TiO2-36 h are 400, 415 and 415 nm, which are right shifted about 13 nm compared to their absorption edges due to the Stokes shift [39,40].  The PL main peak originates from photogenerated electrons in the conduction band returning directly to the valence band and recombining with holes, therefore, the wavelength corresponding to main peak in PL spectra is related to the band gap [37,38]. The main peak wavelengths of TiO 2 -12 h, TiO 2 -24 h and TiO 2 -36 h are 400, 415 and 415 nm, which are right shifted about 13 nm compared to their absorption edges due to the Stokes shift [39,40]. Figure 7a shows the RhB decolorization curves of samples. Without catalyst, the decolorization degree of RhB is 2.6%, which indicates that the decolorization of RhB is mainly due to the degradation of photocatalysts. The decolorization degrees of TiO 2 -12 h, TiO 2 -24 h and TiO 2 -36 h are 99.6, 46.3 and 81.8% after 45 min. Figure 7b displays the kinetics fitting curves of samples. The apparent first-order rate constants k of TiO 2 -12 h, TiO 2 -24 h and TiO 2 -36 h are 0.125, 0.013 and 0.024 min −1 , respectively. TiO 2 -12 h shows the highest photocatalytic activity. Although PL spectra and DRS spectra show that TiO 2 -24 h and TiO 2 -36 h exhibit a lower photogenerated charge recombination rate and higher visible light absorption, their photocatalytic activity is lower than that of TiO 2 -12 h yet. Morphology and BET results show that TiO 2 -12 h is composed of fine particles and possesses a relatively high surface area (145.3 m 2 /g), which is much higher than TiO 2 -24 h (43.0 m 2 /g) and TiO 2 -36 h (13.3 m 2 /g). A high surface area provides more reactive sites, thus TiO 2 -12 h shows the highest photocatalytic activity. Several RhB decolorization data reported by literatures via hydrothermal method are summarized in Table 2.   [41] hydrothermal method Ag-TiO2 Xenon lamp (800 W, >420 nm) 96.0% in 270 min 0.011 (RhB) [42] hydrothermal method TiO2 Mercury lamp (300 W) 58.0% in 15 min 0.104 (RhB) [43] hydrothermal method SDBS-TiO2 Xenon lamp (500 W) 90.0% in 120 min 0.0185 (RhB) [44] hydrothermal method C-TiO2 Xe lamp (500 W, >400 nm) 94.3% in 120 min 0.022 (RhB) [45] hydrothermal method Ag-TiO2 Xenon lamp (500 W) 80.0% in 240 min - [46] hydrothermal method Ag-TiO2 Xenon lamp (350 W) 100% in 45 min - [47] hydrothermal To study the degradation effect of the prepared photocatalyst on pharmaceutical waste, tetracycline hydrochloride (TC) was selected as the target pollutant. The results are shown in Figure 8. The degradation degree of TiO2-12 h, TiO2-24 h and TiO2-36 h are 90.0, 39.1 and 62.4%. The apparent first-order rate constants k of TiO2-12 h, TiO2-24 h and TiO2-36 h are 0.050, 0.011 and 0.022 min −1 , respectively. The photodegradation results of RhB and TC confirm that TiO2-12 h is an efficient photocatalyst, which shows potential application prospects in the field of dye wastewater and pharmaceutical wastewater.  To study the degradation effect of the prepared photocatalyst on pharmaceutical waste, tetracycline hydrochloride (TC) was selected as the target pollutant. The results are shown in Figure 8 To study the degradation effect of the prepared photocatalyst on pharmaceutic waste, tetracycline hydrochloride (TC) was selected as the target pollutant. The results a shown in Figure 8. The degradation degree of TiO2-12 h, TiO2-24 h and TiO2-36 h are 90 39.1 and 62.4%. The apparent first-order rate constants k of TiO2-12 h, TiO2-24 h and TiO 36 h are 0.050, 0.011 and 0.022 min −1 , respectively. The photodegradation results of Rh and TC confirm that TiO2-12 h is an efficient photocatalyst, which shows potential app cation prospects in the field of dye wastewater and pharmaceutical wastewater.

Photocatalytic Mechanism
To verify the free radicals generated during the photocatalytic reaction, nitro-blue tetrazolium (NBT) and salicylic acid (SA) tests were carried out on TiO 2 -12 h sample. The detailed processes of NBT and SA experiments are as follows: Add 0.1 g TiO 2 -12 h powder into 100 mL NBT solution (0.05 mmol/L) and 100 mL SA solution (0.02 mol/L), respectively, keep stirring, and test their absorbance every 15 min after light irradiation. The results are shown in Figure 9. The photoinduced electrons are excited to conduction band and react with O 2 to produce O 2 − radicals, which further react with NBT. SA react with ·OH radicals, forming 2,3-HBA. Therefore, the decrease in NBT absorbance and the increased 2,3-HBA absorbance suggest that O 2 − and ·OH radicals are generated under irradiation [48,49].
Materials 2021, 14, x FOR PEER REVIEW

Photocatalytic Mechanism
To verify the free radicals generated during the photocatalytic reaction, nitro tetrazolium (NBT) and salicylic acid (SA) tests were carried out on TiO2-12 h sample detailed processes of NBT and SA experiments are as follows: Add 0.1 g TiO2-12 h po into 100 mL NBT solution (0.05 mmol/L) and 100 mL SA solution (0.02 mol/L), re tively, keep stirring, and test their absorbance every 15 min after light irradiation results are shown in Figure 9. The photoinduced electrons are excited to conduction and react with O2 to produce O2 − radicals, which further react with NBT. SA react ·OH radicals, forming 2,3-HBA. Therefore, the decrease in NBT absorbance and th creased 2,3-HBA absorbance suggest that O2 − and ·OH radicals are generated under diation [48,49]. The active species in photocatalytic reaction process were investigated through ing benzoquinone (BQ), ammonium oxalate (AO) and isopropanol (IPA) as scaven The detailed processes of active species experiments are as follows: in the photocat experiment, 2 mL (0.1 mol/L) BQ, AO and IPA solutions were added, respectively, ke other test conditions unchanged. The results are shown in Figure 10. The decoloriz degree of RhB for TiO2-12 h declines from 99.6 to 58.8, 93.6 and 94.0% in the presen BQ, AO and IPA, respectively. Meanwhile, The degradation degree of TC for TiO declines from 90.0 to 46.5, 83.4 and 80.3% in the presence of BQ, AO and IPA. Sinc AO and IPA capture ·O2 − , h + and OH species, it can be concluded that O2 − radicals a main active groups in the degradation process, and h + and OH radicals play a secon role. The active species in photocatalytic reaction process were investigated through adding benzoquinone (BQ), ammonium oxalate (AO) and isopropanol (IPA) as scavengers. The detailed processes of active species experiments are as follows: in the photocatalytic experiment, 2 mL (0.1 mol/L) BQ, AO and IPA solutions were added, respectively, keeping other test conditions unchanged. The results are shown in Figure 10. The decolorization degree of RhB for TiO 2 -12 h declines from 99. 6  other test conditions unchanged. The results are shown in Figure 10. The decolorization degree of RhB for TiO2-12 h declines from 99.6 to 58.8, 93.6 and 94.0% in the presence of BQ, AO and IPA, respectively. Meanwhile, The degradation degree of TC for TiO2-12 h declines from 90.0 to 46.5, 83.4 and 80.3% in the presence of BQ, AO and IPA. Since BQ, AO and IPA capture ·O2 − , h + and OH species, it can be concluded that O2 − radicals are the main active groups in the degradation process, and h + and OH radicals play a secondary role.

Conclusions
In summary, the influences of hydrothermal time on the structure and photocatalytic performance of TiO 2 were studied systematically. TiO 2 -12 h forms anatase, TiO 2 -24 h forms anatase/rutile mixed crystal, and TiO 2 -36 h forms rutile. As the hydrothermal time increases, TiO 2 gradually transforms from fine particles to lumps and the surface area decreases. TiO 2 -24 h and TiO 2 -36 h show a lower photogenerated charge recombination rate and higher visible light absorption, however, their photocatalytic activities are lower than that of TiO 2 -12 h, which can be attributed to the relatively high surface area (145.3 m 2 /g) of TiO 2 -12 h. Active species tests confirm that·O 2 − radicals are the main active groups in the degradation process.