One-step synthesis of OH-TiO2/TiOF2 nanohybrids and their enhanced solar light photocatalytic performance

TiO2/TiOF2 nanohybrids were quickly synthesized through a hydrothermal process using titanium n-butoxide (TBOT), ethanol (C2H5OH) and hydrofluoric acid as precursors. The prepared nanohybrids underwent additional NaOH treatment (OH-TiO2/TiOF2) to enhance their photocatalytic performance. In this paper, the mechanism of NaOH affecting the pathway of transformation from TBOT (Ti precursor) to TiO2 nanosheets was discussed. The synthesized TiO2/TiOF2 and OH-TiO2/TiOF2 were characterized by field emission scanning electron microscopy (FE-SEM), high-resolution transmission electron microscopy (HRTEM), X-ray diffraction pattern (XRD), Fourier infrared spectroscopic analysis (FT-IR), Photoluminescence (PL) emission spectra and UV–visible diffuse reflection spectra (UV–vis DRS). The photocatalytic activity and stability of synthesized samples were evaluated by degradation of methylene blue (MB) under the simulated solar light. The results showed that a larger ratio of TiO2 to TiOF2 in TiO2/TiOF2 and OH-TiO2/TiOF2 nanohybrids could allow for even higher MB conversion compared with only TiO2 nanosheets. NaOH treatment can wash off the F ions from TiOF2 and induce this larger ratio. The highest efficiency of MB removal was just above 90% in 1 h. Lower electron–hole pairs recombination rate is the dominant factor that induces the photocatalytic performance enhancement of TiO2/TiOF2 nanohybrids. The synthesized OH-TiO2/TiOF2 nanohybrids exhibit great potential in the abatement of organic pollutants.


Introduction
As one of the most important materials, TiO 2 has been widely used as a promising catalyst due to its lack of toxicity, high stability and easy preparation. However, it has intrinsic faults of a wide energy band gap (3.1-3.2 eV, meaning it only responds to UV light) and high electronhole recombination, which hinders its use under solar or visible light [1][2][3][4]. Many studies examining TiO 2 has been devoted to reducing its energy band gap or photoelectron-hole separation [5][6][7][8][9][10][11].
Recently, Wen et al. [12] proposed a new visible light-driven TiOF 2 photocatalyst for H 2 evolution. Furthermore, Wang et al. [13] discovered a type of TiOF 2 photocatalyst that possesses proper activity and strong durability in photocatalytic degradation of rhodamine B and 4-chlorophenol under visible light, although its photocatalytic performance is not ideal. Heterostructured photocatalysts have attracted increasing attention during the past few years. The electronic assembling of different nanomaterials possessing dissimilar crystal structure and band edge positions allows the complete utilization of incident photons, while in-built electric fields at the interface assist effective charge carrier separation and induce excellent performance in terms of photocatalytic activity [5][6][7][8][9][10][11]. However, only a few investigations on the TiO 2 /TiOF 2 nanocomposites with heterostructures have been performed so far. Zhao et al. [14] reported on a Pd@TiO 2 /TiOF 2 photocatalyst made of TiO 2 shell and TiOF 2 core (labelled as TiO 2 /TiOF 2 ) and further improved its performance by loading Pd nanoparticles onto the surfaces of TiO 2 /TiOF 2 heterostructure, although its synthesis process is complex.
Among the typical synthesis methods, titanium (IV) butoxide and hydrofluoric acid (HF) are the most common precursors to provide anatase TiO 2 nanosheets with exposed facets [15][16][17][18][19][20][21][22]. In these papers, TiOF 2 was sometimes characterized by X-ray diffraction with a peak at 2θ = 23.9°, which occurs independently of the main (101) peak of anatase TiO 2 [17][18][19][20][21][22]. Almost all of these studies noted that TiOF 2 is the intermediate compound during the transformation of Ti 4+ to titanium nanosheets (TiO 2 ) and aimed to suppress its appearance to obtain pure TiO 2 nanosheets [17][18][19][20][21][22]. A few of these studies paid attention to the photocatalytic activity of TiOF 2 or TiO 2 /TiOF 2 systems, which performed more poorly when compared with (001)TiO 2 under UV or solar light [20][21][22]. For example, Lv et al. found that when TiOF 2 was calcined at 300°C, anatase TiO 2 and TiOF 2 were both observed, although the TiOF 2 and TiO 2 /TiOF 2 samples showed poor photocatalytic activity compared with (001) TiO 2 under UV light for reactive brilliant red X3B [20]. Huang et al. also found that TiOF 2 has poor photocatalytic performance, although the performances of TiO 2 /TiOF 2 systems were enhanced with an increased proportion of TiO 2 [21]. Yua et al. discussed the appearance of TiOF 2 and TiO 2 under different F/Ti atomic molar ratios, although they aimed to find the optimal ratios of exposed (101) and (001) facets of TiO 2 . However, they did not mention the role of TiOF 2 in the TiO 2 /TiOF 2 system [22]. Zhang et al. found that the mixture of (001)TiO 2 and TiOF 2 showed better photocatalytic activity for Rhodamine B (RhB) under simulated sunlight when they studied the reaction products of titanium butoxide and hydrogen fluoride. However, this did not show such good performance on MB for a reaction time as long as 10 h [23].
Alkali modification of catalysts has also been proven to be an effective way to enhance the photoactivity with methane dehydroaromatization, cumene cracking and CO oxidation [23][24][25][26][27]. For example, Han et al. found that alkali modification could form more hydroxyl groups on Au catalysts to enhance their catalytic activity [24]. Alkali-modified ZSM-5 zeolite also showed enhanced catalytic performance due to the formation of additional mesopores and the improvement in mass transfer and reaction kinetics [25,26]. NaOH-modified Pt/TiO 2 also showed enhanced performance in terms of the oxidation of formaldehyde under room temperature [27]. These studies have encouraged the modification of TiO 2 /TiOF 2 by NaOH to enhance its catalytic performance.
In this study, TiO 2 /TiOF 2 nanohybrids were synthesized and (001) TiO 2 was obtained through NaOH washing to remove the surface fluorine ions. NaOH was also used to modify the TiO 2 /TiOF 2 nanohybrids to enhance their performance. This has not been investigated before. TiO 2 /TiOF 2 nanohybrids even showed superior catalytic photoactivity towards methylene blue (MB) degradation under simulated sunlight for the first time. The possible mechanism of the TiO 2 /TiOF 2 system in enhancing photocatalytic performance was also discussed.

Material
Shanghai, China. All reagents were A.R. grade and used without further purification. Ultra-pure water (18.2 MΩ•cm) was used as the water in all experiments.

Synthesis of TiO 2 /TiOF 2 nanohybrids
A total of 15.2 ml of ethanol was added into 17.6 ml of TBOT, which was named solution A. Another 15.2 ml of ethanol and 5 ml of HF were added into 90 ml of ultra-pure water, which was named solution B. Solution A was dropped into solution B under medium-speed magnetic stirring at room temperature for 1.5 h to obtain a faint yellow sol. After this, the sol was transferred into a 200-ml Teflon-lined stainless steel autoclave. The autoclave was placed into an oven, which was maintained at 100°C for 0.5, 1, 1.5 and 2 h before being cooled to room temperature naturally to obtain a white precipitate. Ultra-pure water and C 2 H 5 OH were used to wash the precipitates several times to reach a pH of 7, before the precipitates were dried at a temperature of 100°C. The prepared samples were denoted as S0.5, S1, S1.5 and S2. The samples were then dispersed in 80 ml of 5 M NaOH solution for 30 min, before being washed with ultrapure water and C 2 H 5 OH to reach a pH of 7. Finally, these samples were dried at 100°C for 12 h. The prepared samples were denoted as OH-S0.5, OH-S1, OH-S1.5 and OH-S2.

Characterization
The crystal structure was analysed by a XD-2 X-ray diffractometer (Beijing Purkinje, China) with Cu-Kα radiation. The morphology was examined by FE-SEM (JEOL JSM6700, Japan) equipped with EDS to probe elemental analysis and high-resolution transmission electron microscopy (HRTEM; Tecnai G2 F20, FEI, USA) using an accelerating voltage of 200 kV. Specific surface area and porosimetry were measured using Micromeritic TriStar II 3020 micrometrics (Micromeritics, USA), and the Brunauer-Emmett-Teller (BET) method was used to calculate the surface area (S BET ). Fourier transform infrared (FT-IR) spectra were recorded using a TENSOR27 (Bruker, Germany). The optical properties were determined by UV-vis diffuse reflectance spectroscopy (UV-vis DRS; Shimadzu 2600, Japan). Photoluminescence (PL) emission spectra were measured at room temperature with a fluorescence spectrophotometer (Hitachi F-2700, Japan) using a 325-nm line with an Xe lamp. X-ray photoelectron spectroscopy (XPS) took place under an ultra-high vacuum (10 Pa) at a pass energy of 100 eV on a Escalab 250 Xi system (ThermoFisher, USA) equipped with a dual X-ray source by using a Al K Alpha anode and a hemispherical energy analyser. All binding energies were calibrated with contaminant carbon (C1 s = 284.6 eV) as a reference.

Photocatalytic experiments
Photocatalytic activity was measured by degradation of MB. A total of 0.015 g of the catalyst was dispersed in a 150-ml double-layered quartz reactor containing 100 ml of a 10.0 mg l −1 MB solution. Cooling water was introduced into the interlayer of the quartz reactor to maintain the solution at room temperature. A Jiguang-300 W Xe lamp (simulating solar light) was located 30 cm away from the MB solution. A JB-420 cutoff filter was chosen to filter off light less than 420 nm to simulate visible light. The solution was magnetically stirred for 0.5 h in the dark to obtain the adsorption-desorption equilibrium, before the Xe lamp was turned on to start the degradation. At time intervals of 0.5 h, about 4.0 ml of the solution was extracted and centrifuged at a speed of 10 000 r.p.m. to remove catalysts. After this, the MB concentration was analysed with a Purkinje UV1901 UV-vis spectrophotometer at 665 nm. The photocatalyst was separated from the MB solution, before another run was started to investigate the durability of catalysts.

The formation mechanism of TiOF 2 , TiO 2 /TiOF 2 and TiO 2
The XRD patterns of the prepared samples were obtained under different experimental conditions (figure 1). All S0.5, S1, S1.5 and S2 samples showed diffraction peaks at 23.42°for TiOF 2 (100) facet and at 25.3°for TiO 2 (101) facet, indicating that the TiOF 2 (JCPDS No. 01-0049) and anatase TiO 2 (JCPDS No. 21-1272) phases coexist. Therefore, the S0.5, S1 and S1.5 samples are TiO 2 /TiOF 2 nanohybrids. It also can be seen that with a longer reaction time, the ratio of peak height of (100) for TiOF 2 and TiO 2 for (101) changed from 0.44 to 4.4-4.5 and then to 0.55. This indicates that TiO 2 existed at the beginning, before the level decreased and was dominated by other compounds, then appeared and dominated again. This result is consistent with previous research [22]. After being washed with NaOH, the peak at 23.42°for (100) facet of TiOF 2 decreased and the peak at 25.3°for TiO 2 (101) facet increased, indicating an increase in the TiO 2 /TiOF 2 ratio with a longer reaction time. For OH-S1.5 and OH-S2, the peak at 25.3°for TiO 2 (101) facet increased, indicating an increase in TiO 2 /TiOF 2 ratio with a longer reaction time. For OH-S1.5 and OH-S2, the TiOF 2 complex disappeared. The (004) facet of TiO 2 even emerged, which was named the (001)-faceted TiO 2 existence. It seems that OH in NaOH can exchange with F ions in the crystal lattice of TiOF 2 , converting it into (001)-faceted TiO 2 nanosheets [14,18,22]. In summary, the chemical reactions of the formation of titanium nanosheets can be proposed. The first step is the hydrolysis reaction. The reaction of TBOT to form Ti(OH) 4 is shown in equation (3.1). Secondly, in equation (3.2), Ti(OH) 4 can react with HF to produce TiOF 2 through water condensation [28,29]. Finally, TiOF 2 can react with NaOH to form TiO 2 , as shown in equation (3.3) [30]. and This study provides a simple conversion method of TiOF 2 to TiO 2 in addition to the calcination of TiOF 2 [20,21].

Morphology analysis
The morphology of the prepared samples was characterized by FE-SEM and high-resolution transmission electron microscope (HRTEM) in figures 2 and 3. The SEM and HRTEM images shown in figure 2 and figure 3a,b, respectively, were used to obtain the morphology of photocatalysts. The S0.5, S1 and S1.5 samples exhibit mixtures of two types of crystals connected together, which correspond to TiOF 2 and TiO 2 according to the XRD results in figure 1. By contrast, S2 shows a nano-network look,  Figure 2. SEM of (a) S0.5, (b) OH-S0.5, (c) S1, (d) OH-S1, (e) S1.5, (f ) OH-S1.5, (g) S2 and (h) OH-S2.
corresponding to TiO 2 . The OH-S0.5, OH-S1 and OH-S1.5 samples exhibited a nano-network look after reaction with NaOH, permitting light scattering inside the catalyst and enhancing its absorption in light.
table S1, the S BET and pore volume of S2 can reach as high as 59.3 m 2 ·g −1 with a pore size of 62.5 nm, while the S BET of S0.5 and OH-S0.5 are significantly lower with values of 23.6 and 27.21 m 2 ·g −1 , respectively, and pore sizes of 122.6 and 14.49 nm, respectively. After NaOH washing, the S BET of OH-S0.5 was a bit lower than that of S0.5, but the pore size was dramatically decreased.

FT-IR analysis
As shown in figure 5, broad absorptions centered around 3418 and 2920 cm −1 as well as the weak sharp absorption band centred around 1624 cm −1 were attributed to the hydroxyl free radicals, associated hydrogen bonds and absorption water. The OH and water in all NaOH-reacted samples are all stronger than those modified before, with the exception of OH-S2. For the OH-S2 sample, the OH from the broad absorptions centred around 3418, 2920 and 1624 cm −1 sharply decreased. According to XRD results, the S2 and NaOH-reacted samples are all TiO 2 /TiOF 2 nanohybrids, except OH-S2. It was concluded that TiO 2 /TiOF 2 nanohybrids can adsorb more OH to enhance photocatalytic performance [31][32][33][34][35][36][37].

XPS analysis
In figure 6, XPS survey spectra of the prepared samples of S0.5 and OH-S0. 5   to the contamination of the XPS instrument itself [23,[38][39][40][41]. The F 1s BE of 684.8 eV in this spectrum (figure 6b) corresponds to that of F adsorbed on TiO 2 , while there was no sign of F ions in the lattice (BE = 688.5 eV) [38,39]. The measured binding energies of Ti 2p3/2 and Ti 2p1/2 (figure 6c) for S0.5 were 459.3 and 465.2 eV, respectively. The main sharp peaks of Ti 2p3/2 were assigned to Ti 4+ in TiO 2 [23,[39][40][41]. As compared to S0.5, the binding energies of OH-S0.5 shifted to 459.16 and 464.9 eV, respectively. One possible explanation is that TiO 2 is partially reduced into Ti 3+ [23,41]. As Ti 3+ is more hydrophilic than Ti 4+ , oxygen and water molecules are easily adsorbed on the surface of TiO 2 , which consequently facilitates the formation of surface OH groups. The measured BE of O 1s (figure 6d) was 530.42 eV, while this shifted to 530.36 eV in OH-S0.5. This can be explained by OH-S0.5 possibly having many oxygen vacancies that are generated at the interface of TiO 2 /TiOF 2 . Oxygen from the gas phase could dissociate and become adsorbed on such defects, thus resulting in a decrease in binding energies of O 1s to TiO 2 lattice oxygen (Ti−O−Ti) [23]. So it can be concluded that NaOH treatment can induce more surface OH groups and oxygen vacancies at the interface of TiO 2 /TiOF 2 , which would enhance the photocatalytic performance of TiO 2 /TiOF 2 nanohybrids. Figure 7 shows the UV-vis absorption spectroscopy and band gap of the prepared TiO 2 /TiOF 2 nanohybrids and TiO 2 samples. It can be seen that S2 and OH-S0. 5   1.4 S2 S0.5 S1 S1.5 S2 OH-S0.5 OH-S1 OH-S1.5 OH-S2

PL analysis
PL emission spectra were used to investigate the efficiency of charge carrier trapping, immigration and transfer as well as to understand the fate of electron hole pairs in catalysts (figure 8). Six peaks were observed in the spectra. The broad emission bands centred at 385.3 nm (peak 1), 399.0 nm (peak 2) and 412.4 nm (peak 3) were ascribed to form the boundaries of exciton emission due to the trapping of free excitons by titanite groups near defects [42]. The long wavelength range of 452.5-468.7 nm (peaks 4 and 5) is attributed to the oxygen vacancy with two trapped electrons. Oxygen vacancy sites are important for the formation of superoxide (O 2 •−) and hydroxyl (•OH) radicals for photocatalytic degradation. A lower PL intensity also indicates a lower recombination rate of electron-hole pairs and higher separation efficiency, thus representing higher photocatalytic activity [43].  OH-S1.5
The reaction rate of all of the samples is shown in figure 9b. The data were fitted with the first-order reaction equation as follows: where t is the reaction time, C 0 is the concentration of RhB at time 0, C is the concentration of RhB at time t, and k is the reaction rate constant. Figure 9b shows that P25 had a rate constant of only 0.07 h −1 , indicating poor photocatalytic performance. The calculated rate constants are 1.4 and 2.3 h −1 for S2 and OH-S0.5, respectively. The OH-S0.5 sample shows the best performance among all the photocatalysts, with a degradation rate that is much higher than that of the P25, TiO 2 /TiOF 2 and TiO 2 samples. This excellent performance could be mainly attributed to its stronger light adsorption and TiO 2 /TiOF 2 combination. The catalytic abilities of 0.006 g/100 ml, 0.030 g/100 ml and 0.050 g/100 ml (OH-S0.5) are also shown in figure 9c. For the catalytic amount of 0.015 g/100 ml, the photocatalytic degradation rate of MB was the strongest and had the best degradation effect (figure 9c). When the content of OH-S0.5 was increased, its adsorptive capacity increased but photocatalytic degradation performance decreased (figure 9c). Figure 9d shows the UV-vis absorption spectral changes of the MB solution. According to previous research, the discoloration of MB can be caused in two ways: the oxidative degradation and the twoelectron reduction to leuco-MB, which can be detected by the UV-vis absorption at 256 nm [44,45]. Figure 9d shows that there is a blue-shift from 665 to 625 nm with an absorbent peak emerging at 256 nm after the spectral change of MB with the irradiation time by OH-S0. 5. This means that a reductive conversion to leuco-MB exists in the degradation path of MB.

Radical-scavenging experiments
Radical-scavenging experiments were performed to complete an in-depth study of the photocatalytic degradation mechanism. The reactive species were detected through trapping experiments of hydroxyl radicals (•OH), holes (h + ) and superoxide radical anions (O 2 •−) by introducing terephthalic acid [46], potassium iodide [46] and p-benzoquinone [47]. Terephthalic acid can combine with the hydroxyl radicals (•OH), potassium iodide can combine with holes (h + ), and p-benzoquinone can combine with the superoxide radical (O 2 •−), to decrease the activity of the catalyst [46,47]. The effects of a series of scavengers on the photocatalytic oxidation towards the MB dye over the photocatalysts are shown in figure 10. Under simulated sunlight illumination, the percentage of MB loss decreased most rapidly after the addition of potassium iodide, indicating h + was the main active species in the photocatalytic process. When terephthalic acid and p-benzoquinone were added, this also reduced activity, which implied that •OH and O 2 •− radicals also played a role in the photooxidation of MB. The order of importance of the active species is h + , •OH and O 2 •−.

Discussion
TiOF 2 photoactivity was quite low, nearly 150 times lower than P25 in the X3B dye degradation experiment [36]. Lv et al. reported that surface fluorination can greatly enhance the photocatalytic activity of TiO 2 due to the formation of free •OH radicals, which are highly reactive [37]. In this research, a combination of TiO 2 and TiOF 2 demonstrated even higher photoactivity performance in MB degradation compared with the pure TiO 2 nanosheets. In order to further understand the reason for the enhanced photocatalytic performance, a possible mechanism of charge separation and transfer on the surface of TiO 2 /TiOF 2 nanohybrids is proposed. The conduction band (E c ) and valence band (E v ) potentials of TiO 2 and TiOF 2 at the point of zero charge can be calculated by the following empirical equation [11,14,44]: where E c is the energy of the conduction band, X is the bulk Mulliken electronegativity of the compound, E 0 is the energy of free electrons on the hydrogen scale (about 4.5 eV), and E g is the band gap energy of the semiconductor. The x-values for TiO 2 and TiOF 2 are approximately 5.8 and 7.3 eV, respectively. The energy gap (E g ) was estimated from the intercept of the tangent in the plots of (αhν) 1/2 versus photon energy (hν), which is shown in figure 7b. The E g of TiO 2 and TiOF 2 were evaluated to be 3.2 eV and 2.94 eV, respectively.
The position of the valence band edge (E v ) is determined by the following equation: The calculated E c and E v of TiO 2 and TiOF 2 are shown in the electronic supplementary material,  recombination is reduced, which is consistent with PL analysis. Thus, under simulated solar light irradiation, more h + radicals can react with H 2 O or OH − to produce •OH, while more electrons can react with the dissolved oxygen molecules to yield superoxide radical anions (O 2 •−) [48,49]. MB can also be decomposed by the holes directly [50]. Because the photocatalytic activity of TiO 2 is better than that of TiOF 2 , a larger ratio of TiO 2 /TiOF 2 (S2 and OH-S0.5) results in better photocatalytic activity. NaOH treatment can induce more F ions being disconnected from the surface of TiOF 2 , which means a larger ratio of TiO 2 /TiOF 2 (XRD results). Furthermore, the presence of OH bonds on the surface of the photocatalyst (FT-IR analysis) and more oxygen vacancies (XPS and PL analysis) also induce the production of more •OH with enhancement of the photocatalytic performance. This can explain why NaOH-treated nanohybrids showed greater performance than untreated ones. OH-S0.5 showed the best photocatalytic activity, even better than S2; this may be attributed to its lower recombination rate of electron-hole pairs introducing more h + , and more Ti 4+ being reduced into Ti 3+ , because its OH, S BET and absorption of light are not the largest according to FT-IR and UV-vis DRS analysis. So a lower electron-hole pairs recombination rate is the dominant factor that induces the photocatalytic performance enhancement of TiO 2 /TiOF 2 nanohybrids. This was in accord with the radical-scavenging experiments.

Conclusion
In summary, an easy one-step hydrothermal route to synthesize TiO 2 /TiOF 2 and OH-TiO 2 /TiOF 2 hybrids has been demonstrated. The introduction of NaOH facilitated the conversion from TiOF 2 /TiO 2 and induced a network structure. The prepared TiO 2 /TiOF 2 , especially the OH-TiO 2 /TiOF 2 nanocomposite, exhibited excellent activity towards the degradation of MB under simulated sunlight irradiation. A larger ratio of TiO 2 /TiOF 2 in TiO 2 /TiOF 2 and OH-TiO 2 /TiOF 2 nanohybrids could enable better performance. NaOH treatment can wash off the F ions from TiOF 2 and induce this larger ratio. The highest efficiency of MB removal was just above 90% in 1 h. A lower electron-hole pairs recombination rate is the dominant factor that induces the photocatalytic performance enhancement of TiO 2 /TiOF 2 nanohybrids.
Therefore, this work opens an avenue to efficiently synthesize TiO 2 /TiOF 2 through a one-step nanocomposite for the removal of organic pollutants.