Signiﬁcantly Enhanced Self-Cleaning Capability in Anatase TiO 2 for the Bleaching of Organic Dyes and Glazes

: In this study, the Mg 2+ -doped anatase TiO 2 phase was synthesized via the solvothermal method by changing the ratio of deionized water and absolute ethanol V water /V ethanol ). This enhances the bleaching efﬁciency under visible light. The crystal structure, morphology, and photocatalytic properties of Mg-doped TiO 2 were characterized by X-ray diffraction, scanning electron microscopy, high-resolution transmission electron microscopy, N 2 adsorption-desorption, UV-Vis spectroscopy analysis, etc. Results showed that the photocatalytic activity of the Mg 2+ -doped TiO 2 sample was effectively improved, and the morphology, speciﬁc surface area, and porosity of TiO 2 could be controlled by V water /V ethanol . Compared with the Mg-undoped TiO 2 sample, Mg-doped TiO 2 samples have higher photocatalytic properties due to pure anatase phase formation. The Mg-doped TiO 2 sample was synthesized at V water /V ethanol of 12.5:2.5, which has the highest bleaching rate of 99.5% for the rhodamine B dye during 80 min under visible light. Adding Mg 2+ -doped TiO 2 into the phase-separated glaze is an essential factor for enhancing the self-cleaning capability. The glaze samples ﬁred at 1180 ◦ C achieved a water contact angle of 5.623 ◦ at room temperature and had high stain resistance (the blot ﬂoats as a whole after meeting the water).


Introduction
With the deterioration of environmental pollution, low-consumption and high-efficiency pollution technologies have received more attention [1,2]. As the durative utilizes clean energy, solar energy has vast potential for exploitation and application. Titanium dioxide is an important photocatalyst that has been widely studied because of its high activity, non-toxic characteristics, environmental friendliness, and good chemical stability [3][4][5][6]. As the energy barrier of the metastable phase was less than that of the stability phase, it was more likely to excite electrons and holes for the metastable phase [7,8]. Hence, anatase TiO 2 is considered to be the best photocatalyst of all of the structures of TiO 2 [9,10]. It can fully effectively utilize UV light from sunlight [11][12][13]. Several factors affect anatase TiO 2 photoactivity, such as crystal size, specific surface area, and crystallinity [14][15][16]. The performance of the TiO 2 was optimized by doping [17][18][19][20], loading [21,22], and thin-film preparation [23,24]. Available studies indicated that some ions could enter the lattice as substitutional or interstitial; the titanium ions are substituted by metal ions in the crystal lattices. Some studies illustrate that rare-metal-ion-doped titania nanoparticles were prepared by the hydrothermal method, and their photocatalytic performance was greatly improved under UV irradiation [25,26]. At present, there exist a few studies concerning magnesium-ion-doped TiO 2 obtained by the sol-gel reaction synthesis route and the solvothermal method [27,28], but its processing is complex and needs HF as a capping agent to form the anatase phase. It would therefore be interesting to investigate how a simple method can be used for preparing a glaze containing Mg(II)-doped anatase that is stable in a medium-/high-temperature (>1000 • C) ceramic glaze [29] and has self-cleaning properties, as anatase TiO 2 has a nanometer size. how a simple method can be used for preparing a glaze containing Mg(II)-doped anatase that is stable in a medium-/high-temperature (>1000 °C) ceramic glaze [29] and has selfcleaning properties, as anatase TiO2 has a nanometer size.
This study presents the simple synthetic procedure of producing Mg-doped TiO2 anatase samples without surfactants or templates and evaluates the influence of the structure and Vwater/Vethanol on their photocatalytic activity in decomposing rhodamine B (RhB). The self-cleaning activities of Mg-doped and undoped TiO2 anatase glaze samples are evaluated by comparing their anti-pollution ability.

Preparation of the Samples
The samples, with various deionized water and absolute ethanol contents, were prepared from tetrabutyl titanate (TBOT), MgCl2•6H2O, and NaOH using the hydrothermal method. In a typical synthesis, firstly, solution A was made, which included MgCl2•6H2O, deionized water, and absolute ethanol. Subsequently, solution B was made, which included TOBT and ethanol. Finally, suspension C was prepared by dripping solution B into system A. The molar ratio of MgCl2•6H2O: TBOT: ethanol: water was 0.03: 1: 10: 50. After 15 min, after adding suspension C into the reactor, it was heated at 180 °C for 36 h and then naturally cooled to room temperature. The final sample obtained was centrifuged and washed with deionized water and absolute ethanol. The photocatalytic properties of the samples were investigated by changing the molar ratio of water/ethanol (Vwater: Vethanol), keeping other experimental parameters unchanged. Figure 1 is the schematic diagram of Mg-doped TiO2 sample preparation. The Mg-doped TiO2 in the glaze sample was fabricated by sintering at 1180~1200 °C using raw powders, i.e., 95% of the as-prepared Kaolin clay was subjected to phase separation melting at 1500 °C for 4 h and 5% by adding 5% Mg-doped TiO2 (Vwater/Vethnol of 12.5:2.5) photocatalysts, and the self-cleaning and hyper-hydrophilic properties of the fired glaze samples were characterized and tested, respectively. Figure 2 is the schematic diagram of the glaze firing processes. The Mg-doped TiO 2 in the glaze sample was fabricated by sintering at 1180~1200 • C using raw powders, i.e., 95% of the as-prepared Kaolin clay was subjected to phase separation melting at 1500 • C for 4 h and 5% by adding 5% Mg-doped TiO 2 (V water /V ethnol of 12.5:2.5) photocatalysts, and the self-cleaning and hyper-hydrophilic properties of the fired glaze samples were characterized and tested, respectively. Figure 2 is the schematic diagram of the glaze firing processes.

Characterization of the Samples
The crystalline phase was identified by X-ray diffractometer (XRD, D8 Advance Bruker AXS, Germany) using Cu Kα radiation. Compared with the standard pattern in the XRD standard database, including JCPDS (i.e., PDF cards), the phase composition of the sample was analyzed using Jade 6.0 software. Photocatalyst morphology was investigated by scanning electron microscopy (SEM, JSM-6700F, Japan) using a device equipped with an EDS system operating at an accelerating voltage of 5.0 kV or 15 kV (15 kV for EDS). The crystal surface of nanocrystals was evaluated by high-resolution microscopy. The microstructures of the samples were studied by transmission electron microscopy (TEM, FEI Tecnai G2 F-30, Holland) and high-resolution transmission electron microscopy (HRTEM, FEI Tecnai G2 F-30, Holland) at accelerating voltages of 160 kV and 200 kV, respectively. The valence states of the samples were characterized by X-ray photoelectron spectroscopy (XPS, ESCALAB Xi+, United States) using Al Kα radiation. The specific surface areas were determined by the Brunauer-Emmett-Teller method, and the pore size was determined by the Barrett-Joyner-Hallenda method. Nitrogen adsorption-desorption isotherms were collected on a Micromeritics TriStar ii 3020 analyzer at 77 K. The analysis of samples by UV-Vis diffuse reflectance spectroscopy was carried out. The hydrophilicity of the samples was tested by a contact angle meter (JGW-360D, China).

Photocatalytic Activity of the Samples
The photocatalytic activity of the TiO2 was evaluated by bleaching the RhB with a concentration of 10 −4 mol/L. The total volume of RhB was 50 mL, irradiated with 0.05 g of the photocatalyst and a 500 WXeon light with a cut-off filter of 420 nm. This was to prove that the RhB was exhibiting bleaching rather than adsorption after the dark experiment was carried out. Samples were taken out at 20 min intervals and analyzed with a spectrophotometer. The photocatalytic activity was characterized by the apparent first-order rate constant k, as in equation k = ln(A0/A), where A was the absorbance of RhB at 553 nm after bleaching and A0 was the absorbance of the initial RhB solution at 553 nm.

Structural and Morphology
The crystal phase of the samples was studied as shown in Figure 3. The obtained diffraction peak of the doped TiO2 matched very well with the standard values (PDF-#21-1272) and the diffraction peaks at 2θ = 25.281(101), 37.800(004), 48.049(200), 53.890(105), and 62.688(204), illustrating that the samples were in the anatase phase. However, the obtained undoped TiO2 was in a mixed phase of anatase and brookite. The cell volume was calculated by Fourier synthesis with the program SHELXS−97 [30]. When the solvent was water, the sample consisted of nanoparticles 10~20 nm in mean size, as determined by

Characterization of the Samples
The crystalline phase was identified by X-ray diffractometer (XRD, D8 Advance Bruker AXS, Germany) using Cu Kα radiation. Compared with the standard pattern in the XRD standard database, including JCPDS (i.e., PDF cards), the phase composition of the sample was analyzed using Jade 6.0 software. Photocatalyst morphology was investigated by scanning electron microscopy (SEM, JSM-6700F, Japan) using a device equipped with an EDS system operating at an accelerating voltage of 5.0 kV or 15 kV (15 kV for EDS). The crystal surface of nanocrystals was evaluated by high-resolution microscopy. The microstructures of the samples were studied by transmission electron microscopy (TEM, FEI Tecnai G2 F-30, Holland) and high-resolution transmission electron microscopy (HRTEM, FEI Tecnai G2 F-30, Holland) at accelerating voltages of 160 kV and 200 kV, respectively. The valence states of the samples were characterized by X-ray photoelectron spectroscopy (XPS, ESCALAB Xi+, United States) using Al Kα radiation. The specific surface areas were determined by the Brunauer-Emmett-Teller method, and the pore size was determined by the Barrett-Joyner-Hallenda method. Nitrogen adsorption-desorption isotherms were collected on a Micromeritics TriStar ii 3020 analyzer at 77 K. The analysis of samples by UV-Vis diffuse reflectance spectroscopy was carried out. The hydrophilicity of the samples was tested by a contact angle meter (JGW-360D, China).

Photocatalytic Activity of the Samples
The photocatalytic activity of the TiO 2 was evaluated by bleaching the RhB with a concentration of 10 −4 mol/L. The total volume of RhB was 50 mL, irradiated with 0.05 g of the photocatalyst and a 500 WXeon light with a cut-off filter of 420 nm. This was to prove that the RhB was exhibiting bleaching rather than adsorption after the dark experiment was carried out. Samples were taken out at 20 min intervals and analyzed with a spectrophotometer. The photocatalytic activity was characterized by the apparent first-order rate constant k, as in equation k = ln(A 0 /A), where A was the absorbance of RhB at 553 nm after bleaching and A 0 was the absorbance of the initial RhB solution at 553 nm.

Structural and Morphology
The crystal phase of the samples was studied as shown in Figure 3. The obtained diffraction peak of the doped TiO 2 matched very well with the standard values (PDF-#21-1272) and the diffraction peaks at 2θ = 25.281(101), 37.800(004), 48.049(200), 53.890(105), and 62.688(204), illustrating that the samples were in the anatase phase. However, the obtained undoped TiO 2 was in a mixed phase of anatase and brookite. The cell volume was calculated by Fourier synthesis with the program SHELXS−97 [30]. When the solvent was water, the sample consisted of nanoparticles 10~20 nm in mean size, as determined by Nano Measurer 1.2 software using 10 nanoparticles. The average crystallite size of TiO 2 samples with different Mg-doped ions was calculated by XRD-Scherrer formula: d = 0.91 λ/βcos θ, where d is the mean crystallite size, k is 0.9, λ is the wavelength of Cu Kα (i.e., λ = 0.15420 nm), β is the full width at half maximum intensity of the peak (FWHM) in radian, and θ is Bragg's diffraction angle [31]. The crystallite size and cell volume were calculated as shown Table 1. When increasing V water /V ethanol , there are differences in the diffraction peak intensity and minor shifts in the peak occur, which indicates a reduction in crystalline size and an increase in the volume of unit cells (Table 1). Since the ionic radius of Mg 2+ (0.072 nm) is close to that of Ti 4+ (0.061 nm), Mg 2+ easily enters the TiO 2 lattice [32] and the lattice volume increases ( Table 1), indicating that the formation of a crystal defect. Based on the experimental results, the formation of the crystal defect promotes the formation of the anatase phase, which is accordance with the reported literature [27,29]. Hence, after the addition of the magnesium source, a pure-anatase TiO 2 phase appears. The intensity of the (004) direction is significantly enhanced compared to undoped TiO 2 . In addition, the FWHM of the (101) peak was calculated by using Lorentz fitting. According to the Scherrer formula, d = 0.91 λ/βcos θ, the crystallite size was calculated; it is shown in Table 1. Nano Measurer software using 10 nanoparticles. The average crystallite size of TiO2 samples with different Mg-doped ions was calculated by XRD-Scherrer formula: d = 0.91 λ/βcosθ, where d is the mean crystallite size, k is 0.9, λ is the wavelength of Cu Kα (i.e., λ = 0.15420 nm), β is the full width at half maximum intensity of the peak (FWHM) in radian, and θ is Bragg's diffraction angle [31]. The crystallite size and cell volume were calculated as shown Table 1. When increasing Vwater/Vethanol, there are differences in the diffraction peak intensity and minor shifts in the peak occur, which indicates a reduction in crystalline size and an increase in the volume of unit cells (Table 1). Since the ionic radius of Mg 2+ (0.072 nm) is close to that of Ti 4+ (0.061 nm), Mg 2+ easily enters the TiO2 lattice [32] and the lattice volume increases ( Table 1), indicating that the formation of a crystal defect. Based on the experimental results, the formation of the crystal defect promotes the formation of the anatase phase, which is accordance with the reported literature [27,29]. Hence, after the addition of the magnesium source, a pure-anatase TiO2 phase appears. The intensity of the (004) direction is significantly enhanced compared to undoped TiO2. In addition, the FWHM of the (101) peak was calculated by using Lorentz fitting. According to the Scherrer formula, d = 0.91 λ/βcosθ, the crystallite size was calculated; it is shown in Table  1.   Figure 4 shows SEM images of the as-synthesized samples. When the solvent was water, the sample consisted of nanoparticles 5-10 nm in size. When the Vwater/Vethnol ratio was 12.5:2.5, agglomerated nanoparticles had a grape-like morphology (Figure 4b). With the increase in ethanol dosage, nanoparticles increased (Figure 4c,d). The experimental results show that the morphology of the samples was greatly affected by Vwater/Vethnol. Their morphology is determined by the relationship between crystal formation and growth.   Figure 4 shows SEM images of the as-synthesized samples. When the solvent was water, the sample consisted of nanoparticles 5-10 nm in size. When the V water /V ethnol ratio was 12.5:2.5, agglomerated nanoparticles had a grape-like morphology (Figure 4b). With the increase in ethanol dosage, nanoparticles increased (Figure 4c,d). The experimental results show that the morphology of the samples was greatly affected by V water /V ethnol . Their morphology is determined by the relationship between crystal formation and growth. Moreover, crystal growth is influenced by the adsorption of certain crystalline facets into OH − . This adsorption hinders the growth of these facets, resulting in different rates of crystalline growth. Ethanol is a typical polar solvent and amphiphilic molecule. It was vertically adsorbed on the hydrophilic surface of the TiO 2 particles, forming a twoamphiphilic bilayer, which limited the immersion of the water molecule in the hydrophilic side surface and the TiO 2 particles [33]. The rapid hydrolysis of TBOT promoted the rapid generation of TiO 2 , which led to TiO 2 particle agglomeration with an increase in V water /V ethanol . Figure 5a,b show TEM and the corresponding SAED pattern (inset) and HRTEM images of the sample prepared at V water /V ethnol = 12.5:2.5. From Figure 5a, it is observed that the aggregated particles in Figure 4b consist of nanoparticles. The major diffraction rings for the crystal surface at (101), (004), and (105) match well with XRD analysis. The d spacing is 0.325 nm (Figure 5b), and it matches well with the lattice spacing of anatase TiO 2 (101). Furthermore, the corresponding EDX spectrum shown in Figures 5c and S1 verifies the existence of Mg, Ti, and O ions. Other impurities were not detected in the EDX spectra. Moreover, crystal growth is influenced by the adsorption of certain crystalline facets into OH − . This adsorption hinders the growth of these facets, resulting in different rates of crystalline growth. Ethanol is a typical polar solvent and amphiphilic molecule. It was vertically adsorbed on the hydrophilic surface of the TiO2 particles, forming a two-amphiphilic bilayer, which limited the immersion of the water molecule in the hydrophilic side surface and the TiO2 particles [33]. The rapid hydrolysis of TBOT promoted the rapid generation of TiO2, which led to TiO2 particle agglomeration with an increase in Vwater/Vethanol. Figure 5a,b show TEM and the corresponding SAED pattern (inset) and HRTEM images of the sample prepared at Vwater/Vethnol = 12.5:2.5. From Figure 5a, it is observed that the aggregated particles in Figure 4b consist of nanoparticles. The major diffraction rings for the crystal surface at (101)  As can be seen from Table 1 and Figure 3, the morphologies of the samples strongly depend on Mg-doped ions and Vwater/Vethanol. Because the current system contains ethanol, water, Mg-doped ions, and TBOT, we can reasonably assume that the formation of anatase TiO2 is due to the dehydrating condensation between Ti(OH)6 2− and Mg-doped ions under solvothermal conditions [34]. Thus, due to the formation of a lower number of active OH − ions and a lower number of soluble species, Ti(OH)6 2− and TiO6 octahedrons in one cluster may construct a chain via the corner-sharing of Ti(OH)6 2− growth units. Due to doped Mg ions entering the TiO2 lattice, resulting in TiO6 octahedron lattice distortion (Table 1) and an increase in the charge density of Ti and reduction in the electron density of oxygen, the preferred TiO6 octahedron chain-shaped clusters further adsorb OH − soluble species into the (101) plane ( Figure 5b) and anatase TiO2 monomers form through a dehydrating condensation process. Therefore, these planes could be freely bonded by interactions between OH-and nuclei to obtain aggregated nanoparticles (Figure 4). The solubility of salt increases with the dielectric constant of the solvent [35], and the dielectric constant of water is bigger than that of ethanol. When Vwater/Vethnol decreases, that is, ethanol content increases, this could decrease the solubility of the precursor and increase the viscosity of the As can be seen from Table 1 and Figure 3, the morphologies of the samples strongly depend on Mg-doped ions and V water /V ethanol . Because the current system contains ethanol, water, Mg-doped ions, and TBOT, we can reasonably assume that the formation of anatase TiO 2 is due to the dehydrating condensation between Ti(OH) 6 2− and Mg-doped ions under solvothermal conditions [34]. Thus, due to the formation of a lower number of active OH − ions and a lower number of soluble species, Ti(OH) 6 2− and TiO 6 octahedrons in one cluster may construct a chain via the corner-sharing of Ti(OH) 6 2− growth units. Due to doped Mg ions entering the TiO 2 lattice, resulting in TiO 6 octahedron lattice distortion (Table 1) and an increase in the charge density of Ti and reduction in the electron density of oxygen, the preferred TiO 6 octahedron chain-shaped clusters further adsorb OH − soluble species into the (101) plane ( Figure 5b) and anatase TiO 2 monomers form through a dehydrating condensation process. Therefore, these planes could be freely bonded by interactions between OH − and nuclei to obtain aggregated nanoparticles (Figure 4). The solubility of salt increases with the dielectric constant of the solvent [35], and the dielectric constant of water is bigger than that of ethanol. When V water /V ethnol decreases, that is, ethanol content increases, this could decrease the solubility of the precursor and increase the viscosity of the solution, thereby decreasing the diffusion ability of Ti(OH) 6 2− ions and causing the crystal size of the TiO 2 sample to decrease (Table 1).    Figure 6b). This may be due to the addition of small amounts of Mg atoms, causing new oxygen vacancies [36]. Oxygen vacancies in TiO2 are usually created in doped TiO2 to maintain charge neutrality and improve the service life of the photocatalyst [37]. When oxygen vacancies are generated, a higher energy peak can be seen due to the decrease in the electron density of oxygen [37]. A peak at 49.93 eV was associated with Mg 2p, which is further verified by the incorporation of Mg 2+ into the titanium dioxide lattice.   Figure 6b). This may be due to the addition of small amounts of Mg atoms, causing new oxygen vacancies [36]. Oxygen vacancies in TiO 2 are usually created in doped TiO 2 to maintain charge neutrality and improve the service life of the photocatalyst [37]. When oxygen vacancies are generated, a higher energy peak can be seen due to the decrease in the electron density of oxygen [37]. A peak at 49.93 eV was associated with Mg 2p, which is further verified by the incorporation of Mg 2+ into the titanium dioxide lattice. Figure 7 shows the typical FT-IR spectrum of undoped TiO 2 and Mg-doped TiO 2 samples with different V water /V ethnol ratios. All samples have absorption peaks at 3380 cm −1 and 1640 cm −1 , corresponding to O-H stretching vibration and bending vibration, respectively [38]. For the undoped TiO 2 sample, the bands at 1450 cm −1 and 1538 cm −1 are attributed to the H-O-H bending of the lattice water [39]. The band centered at 510 cm −1 is due to isolated tetrahedral TiO 4 stretching vibrations and only occurs in the pure TiO 2 sample [40]. As a result of Mg-doping, the bands at 1065 cm −1 and 458 cm −1 show the vibration of Ti-O-Mg [41]. With the increase in ethanol content, the intensities of the absorption peaks at 3380 cm −1 and 458 cm −1 increase, respectively. This indicates that Mg ions are doped into the lattice of TiO 2 , and the HRTEM, TEM, and XRD results further confirmed this point.  Figure 7 shows the typical FT-IR spectrum of undoped TiO2 and Mg-doped TiO2 sa ples with different Vwater/Vethnol ratios. All samples have absorption peaks at 3380 cm −1 a 1640 cm −1 , corresponding to O-H stretching vibration and bending vibration, respective [38]. For the undoped TiO2 sample, the bands at 1450 cm −1 and 1538 cm −1 are attribut to the H-O-H bending of the lattice water [39]. The band centered at 510 cm −1 is due isolated tetrahedral TiO4 stretching vibrations and only occurs in the pure TiO2 samp [40]. As a result of Mg-doping, the bands at 1065 cm −1 and 458 cm −1 show the vibration Ti-O-Mg [41]. With the increase in ethanol content, the intensities of the absorption pea at 3380 cm −1 and 458 cm −1 increase, respectively. This indicates that Mg ions are doped in the lattice of TiO2, and the HRTEM, TEM, and XRD results further confirmed this poin  Figure 8 shows the BET analysis of the samples using nitrogen adsorption-desorption. For all samples, the isotherms are type IV, and clear hysteresis loops can be identified.   Figure 8 shows the BET analysis of the samples using nitrogen adsorption-desorption. For all samples, the isotherms are type IV, and clear hysteresis loops can be identified. With the increase in V water /V ethnol , the BET surface area of the Mg-doped TiO 2 samples decreases. However, the pore volume and porosity of the samples exhibit a prominent enhancement compared with the undoped TiO 2 sample, as shown in Table 1 and Figure 8. The BJH average pore diameters, calculated from the adsorption branch of the isotherms, are 11.205 nm, 12.560 nm, 12.365 nm, and 12.807 nm for pure TiO 2 and Mg-doped TiO 2 samples prepared with different V water /V ethnol ratios of 12.5:2.5, 10:5, and 7.5:7.5, respectively. The mesoporous structure is mainly due to the porous accumulation of nanoparticles [42]. The porosity increase is due to the crystal size reducing with the decrease in V water /V ethnol .  Figure 8 shows the BET analysis of the samples using nitrogen adsorption-desorption. For all samples, the isotherms are type IV, and clear hysteresis loops can be identified. With the increase in Vwater/Vethnol, the BET surface area of the Mg-doped TiO2 samples decreases. However, the pore volume and porosity of the samples exhibit a prominent enhancement compared with the undoped TiO2 sample, as shown in Table 1 and Figure 8. The BJH average pore diameters, calculated from the adsorption branch of the isotherms, are 11.205 nm, 12.560 nm, 12.365 nm, and 12.807 nm for pure TiO2 and Mg-doped TiO2 samples prepared with different Vwater/Vethnol ratios of 12.5:2.5, 10:5, and 7.5:7.5, respectively. The mesoporous structure is mainly due to the porous accumulation of nanoparticles [42]. The porosity increase is due to the crystal size reducing with the decrease in Vwater/Vethnol.  Figure 9 shows the UV-Visible diffuse reflectance spectra of TiO2. The absorption edge of doped TiO2 had more of a blue shift than the undoped TiO2. The Kulbeka-Munk formula, (E(ev) = hC/λ, h = 6.626 × 10 −34 Js, C = 3.0 × 10 8 ms −1 ), was used to acquire the exact  Figure 9 shows the UV-Visible diffuse reflectance spectra of TiO 2 . The absorption edge of doped TiO 2 had more of a blue shift than the undoped TiO 2 . The Kulbeka-Munk formula, (E(ev) = hC/λ, h = 6.626 × 10 −34 Js, C = 3.0 × 10 8 ms −1 ), was used to acquire the exact band gap of TiO 2 from 3.26 eV to 3.13 eV, which can be attributed to the Mg 2+ -doped TiO 2 in the framework. Since Mg 2+ ions generated from oxygen vacancies are known to cause the photoexcitation of long-wavelength light, the UV-Vis absorption spectrum was inferred to verify the presence of Mg 2+ in the TiO 2 -doped sample.

Optical Properties
Inorganics 2023, 11, x FOR PEER REVIEW 9 of 15 band gap of TiO2 from 3.26 eV to 3.13 eV, which can be attributed to the Mg 2+ -doped TiO2 in the framework. Since Mg 2+ ions generated from oxygen vacancies are known to cause the photoexcitation of long-wavelength light, the UV-Vis absorption spectrum was inferred to verify the presence of Mg 2+ in the TiO2-doped sample. Moreover, from the spectrum, the energy gap of the semiconductor nanoparticles is related to the particle size. The band gap increases as the particle size decreases, resulting in a phenomenon known as a "blue shift" in light absorption at a specific wavelength due to the quantum size effect [43]. With the increase in ethanol content, the absorption edge Moreover, from the spectrum, the energy gap of the semiconductor nanoparticles is related to the particle size. The band gap increases as the particle size decreases, resulting in a phenomenon known as a "blue shift" in light absorption at a specific wavelength due to the quantum size effect [43]. With the increase in ethanol content, the absorption edge of the doped TiO 2 is blue-shifted, illustrating the particle size reduction. The results obtained are well-matched with the sizes of the crystals that were measured. The band gap energies of the prepared TiO 2 doped by adding 0 to 7.5 mL ethanol were found to be 3.17 ev, 3.03 ev, 3.13 ev, and 3.25 ev, respectively. From Figure 4, it is clear that the size of anatase nanoparticles increases with the increase in ethanol content. Optical absorption is highly dependent on the internal structure of the material [44]. Compared with pure TiO 2 , the longer-wavelength region of Mg-doped TiO 2 samples implies that the only possible transition is from the oxygen vacancies causing a red shift of the absorption edge (Figure 6), which also implies that Mg 2+ has been incorporated into the lattice of TiO 2 ( Table 1). From Figure 6, it can be observed that compared with the pure TiO 2 sample, the Ti and O binding energy in Mg-doped TiO 2 samples has been shifted to a lower energy and a higher energy peak, because some Ti 4+ ions are replaced by Mg 2+ ions in order to increase the charge density of Ti and reduce the electron density of oxygen [45]. The new oxygen vacancies are created through the doping of small amounts of Mg atoms [46]. For the Mg-doped TiO 2 sample, the peak of 49.9 eV is ascribed to Mg 2p (Figure 6c), which is consistent with the value of Mg 2+ [27,41]. These observations further verify the existence of Mg 2+ in the Mg-doped TiO 2 sample, which is consistent with XRD ( Figure 3), increased cell volume (Table 1), and FT-IR spectrum (Figure 7). Figure 10 shows the photocatalytic bleaching of RhB through the as-prepared sample under visible light. As shown in Figure 10, RhB concentration is unchanged, illustrating that RhB adsorbed on the TiO 2 surface had reached equilibrium in 30 min. Figure 10b shows kinetic curves of ln(C 0 /C) versus irradiation time during RhB bleaching under visible light irradiation. It has been found that the apparent rate constants [47] for the reaction of RhB with Mg-doped TiO 2 samples (V water /V ethanol = 15:0, 12.5:2.5, 10:5, 7.5:7.5) and Mg-undoped TiO 2 (V water /V ethanol = 12.5:2.5) were 0.01704, 0.06335, 0.04153, 0.01668, and 0.00203 min −1 , respectively, which illustrates that the photocatalytic activity of the samples was effectively improved by Mg 2+ -doping (due to pure anatase phase formation (Figure 3)). Moreover, the photocatalytic properties of Mg-doped TiO 2 can be further improved by changing the ratio of water to ethanol. The photocatalytic properties of the samples increased first and then decreased gradually with the increase in V water /V ethanol . When the V water /V ethanol ratio was 12.5:2.5, Mg-doped TiO 2 had the maximum photocatalytic activity. In addition, by combining Table 1 with Figures 4 and 9, we can observe that the aggregated nanoparticles increase in size and thus E g increases, which leads to the easy recombination of the electron and hole in the migration process, and therefore, the photocatalytic activity of the samples decreases with the increase in ethanol volume (i.e., V water /V ethanol decreases). Although TiO 2 (V water /V ethanol = 15:0) has a larger specific surface area and smaller crystal size (Table 1) compared with the Mg-doped samples, the sample had lower porosity and pore size, which caused the decrease in the sample of RhB adsorption. This clearly indicates that the adsorption of samples was determined by the surface area and characteristics of the pore. Obviously, Mg-doped TiO 2 samples exhibited better photocatalytic activities than pure TiO 2 samples. The narrowing of the band gap is a result of Mg doping into the TiO 2 lattice, which enables the trapping of the photo-induced electron and facilitates the separation of electron-hole pairs (Figure 11a). pore size, which caused the decrease in the sample of RhB adsorption. This clearly in cates that the adsorption of samples was determined by the surface area and characte tics of the pore. Obviously, Mg-doped TiO2 samples exhibited better photocatalytic ac ities than pure TiO2 samples. The narrowing of the band gap is a result of Mg doping i the TiO2 lattice, which enables the trapping of the photo-induced electron and facilita the separation of electron-hole pairs (Figure 11a).   cates that the adsorption of samples was determined by the surface area and characteristics of the pore. Obviously, Mg-doped TiO2 samples exhibited better photocatalytic activities than pure TiO2 samples. The narrowing of the band gap is a result of Mg doping into the TiO2 lattice, which enables the trapping of the photo-induced electron and facilitates the separation of electron-hole pairs (Figure 11a).

Self-Cleaning Properties of Mg-Doped TiO 2 in Glaze Sample
It can be seen the wet angle of pure TiO 2 glaze samples is obviously higher than those of Mg-doped TiO 2 glaze samples ( Figure 12). The super-hydrophilicity of Mg-doped TiO 2 glaze samples is attributed to several comprehensive factors. Based on the experimental results, Mg ions are helpful for the growth of the TiO 2 crystal grain, and thus separates the phase size in Mg-doped TiO 2 glaze more than pure TiO 2 . This makes the Mg-doped TiO 2 glaze surface rougher than that of the pure TiO 2 glaze (Figure 12). A large surface roughness could improve the hydrophilicity, according to the Wenzel equation (1): cos θ r = rcos θ, where r denotes the surface roughness of the glaze, cos θ is the classical contact angle depicted by the Young equation, and θ r is the measured real contact angle. Moreover, the partial substitution of Mg 2+ ions for Ti sites increases the slight TiO 2 lattice distortion, which is available for a low initial contact angle and hydrophilicity [48]. From Figure 12, it can be seen that the contact angles of Mg-doped TiO 2 samples are smaller than that of the pure TiO 2 glaze sample in the dark condition, indicating that the greater roughness and lattice distortion are helpful for decreasing the contact angle. This could be because the incorporation of Mg makes the band gap of TiO 2 narrow, thus the visible light can excite pairs of electrons and holes (Figure 11a), just as in the case of ultraviolet irradiation for the pure TiO 2 glaze. Ti 4+ ions could be united with the photo-induced electron and thus Ti 3+ ions could be obtained. Ti 3+ sites can be substituted by Mg 2+ ions, which produces one excess positive charge. Those excess positive charges could capture the photo-induced electrons quickly, and thus photo-generated holes are available for combining more H 2 O adsorbed on the glaze surface and react with water, producing hydroxyl radicals that are also available for maintaining the hydrophilicity of Mg-doped TiO 2 glaze samples [29]. Therefore, the super-hydrophilicity of Mg-doped TiO 2 glaze samples could be attributed to the visible-light-exciting photo-induced pairs of electrons and holes. For the sample with a V water /V ethanol ratio of 10:5 and 7.5:7.5, the contact angles of water droplets on Mg-doped TiO 2 glaze samples increase slightly, which could be attributed to the decrease in the V water /V ethanol ratio. However, when V water /V ethanol is 10:5 and 7.5:7.5, the hydrophilicity of Mg-doped TiO 2 glaze samples decreases slightly, though it still has super-hydrophilicity. The hydroxy groups anchoring on the Mg-doped TiO 2 glaze surface have a significant impact on the hydrophilicity. The formation of hydroxy groups results in the dissociative adsorption of water molecules at oxygen vacancy sites on the Mg-doped TiO 2 glaze surface. The extra hydroxy groups and oxygen vacancies on the surface are produced by electron-hole pairs, which lead to the hydrophilicity of the Mg-doped TiO 2 glaze surface [39]. Because oxygen vacancy is produced by the doping of Mg in the TiO 2 crystal and the separation of electron-hole pairs is facilitated (Figure 11a), the Mg-doped TiO 2 glaze surface has more photo-induced wettability than the pure TiO 2 glaze surface. The self-cleaning performance was tested using a Japan Marker pen. The glaze surface was drawn on after drying for 1 h. After that, after placing a few drops of water on the glaze, we could observe whether the ink blots were floating. Table 2 shows that after firing at 1180~1200 °C, the water contact angle (5.623° vs. 15.23°) and stain resistance The self-cleaning performance was tested using a Japan Marker pen. The glaze surface was drawn on after drying for 1 h. After that, after placing a few drops of water on the glaze, we could observe whether the ink blots were floating. Table 2 shows that after firing at 1180~1200 • C, the water contact angle (5.623 • vs. 15.23 • ) and stain resistance (the blot floats as a whole vs. not floating, as shown in Figure 13) of the sample fabricated were improved compared to commercial self-cleaning ceramic glazes [49]. The above results indicate the great potential application for enhancing the self-cleaning properties of glazes by introducing Mg-doped TiO 2 .

Conclusions
In this paper, Mg-doped TiO2 samples with various Vwater/Vethanol ratios were successfully prepared through the solvothermal method at 180 °C for 36 h. The Mg-doped (Vwater/Vethanol = 12.5:2.5) sample had higher surface area, porosity, optical performance, and photocatalytic activity than other samples. Undoped and Mg-doped TiO2 glaze ceramic samples were prepared using a medium-/high-temperature solid-firing process. Mg-doped TiO2 samples (Vwater/Vethanol = 12.5:2.5) illustrated superior hydrophilicity properties, photocatalytic activity in terms of bleaching organic dye, and self-cleaning capability in ceramic glaze samples than other samples after visible light exposure. This study provides a preparation approach for the synthesis of TiO2 while controlling crystal size and morphology, which can be utilized with solar energy for bleaching the contaminants in water and enhancing the self-cleaning properties of medium-/high-temperature glazes.

Conclusions
In this paper, Mg-doped TiO 2 samples with various V water /V ethanol ratios were successfully prepared through the solvothermal method at 180 • C for 36 h. The Mg-doped (V water /V ethanol = 12.5:2.5) sample had higher surface area, porosity, optical performance, and photocatalytic activity than other samples. Undoped and Mg-doped TiO 2 glaze ceramic samples were prepared using a medium-/high-temperature solid-firing process. Mg-doped TiO 2 samples (V water /V ethanol = 12.5:2.5) illustrated superior hydrophilicity properties, photocatalytic activity in terms of bleaching organic dye, and self-cleaning capability in ceramic glaze samples than other samples after visible light exposure. This study provides a preparation approach for the synthesis of TiO 2 while controlling crystal size and morphology, which can be utilized with solar energy for bleaching the contaminants in water and enhancing the self-cleaning properties of medium-/high-temperature glazes.