Quasi-Spherical Brookite TiO2 Nanostructures Synthesized Using Solvothermal Method in the Presence of Oxalic Acid

Brookite TiO2, the latest TiO2 photocatalyst, promises to be an interesting candidate for photocatalytic applications because of its unique physical and chemical properties. In this study, pure-phase brookite TiO2 with a quasi-spherical nanostructure was successfully synthesized via a solvothermal method using tetrabutyl titanate (Ti(OC4H9)4, TBOT) as the Ti source in the presence of oxalic acid. NaOH was used to regulate the pH of solution. The structure and morphology of the samples were then analyzed using multiple methods, such as X-ray diffraction (XRD), Raman spectroscopy, scanning electron microscopy (SEM), transmission electron microscopy (TEM), Brunauer–Emmett–Teller (BET) measurements and ultraviolet–visible diffuse spectroscopy (UV–Vis). Photocatalytic activities of the as-synthesized brookite TiO2 were evaluated by degrading aqueous methylene blue solution under UV light irradiation. The effect of thermal treatment temperature on photocatalytic activity of the samples was also investigated. The produced brookite TiO2 nanopowders calcined at 500 °C for 2 h showed the highest photocatalytic activity, and the corresponding degradation rate of methylene blue (10 mg/L) reached 96.7% after 90 min of illumination. In addition, the formation mechanism of pure-phase brookite TiO2 was investigated. It was found that the formation of pure-phase brookite TiO2 in this study was ascribed to the combined action of oxalic acid and sodium hydroxide.


Introduction
It is known that TiO 2 shows three types of crystalline polymorphs: rutile (tetragonal, space group: P42/mnm), anatase (tetragonal, space group: I41/amd), and brookite (orthorhombic, space group: Pbca) [1]. Of these polymorphs, anatase and rutile TiO 2 have been widely studied over the past decades. However, few reports have focused on the preparation and properties of brookite TiO 2 , due to the difficulty involved in preparing its pure form [2][3][4]. Nevertheless, theoretical studies have demonstrated that brookite TiO 2 exhibits superior photocatalytic performance compared to anatase and rutile TiO 2 because of its unique orthorhombic crystalline structure [5][6][7]. Thus, brookite TiO 2 has generated significant interest due to its physicochemical properties and possible applications.
It is worth mentioning that oxalic acid is a rigid bidentate ligand that can link two metallic centers, and is a good chelating agent that effectively chelates with metal ions, including a titanium ion [17,22,23]. The decomposition of oxalate groups is primarily bi-coordinated or monocoordinated to titanium atoms via a reaction with Ti-OH groups, which results in the evolution of CO, CO 2 and a minor amount of H 2 O [14]. In addition, the titanium oxalate complex, which has a specific octahedral arrangement that is similar to the titanium glycolate complex, facilitates the growth of brookite TiO 2 polymorph with a high pH value [12,19,21]. Dambournet et al. [15] reported that mesoporous brookite TiO 2 with a high specific surface area and controllable morphology could be synthesized using titanium oxalate hydrate as a precursor in the presence of alkali ions (Li + , Na + ). It is thus known that oxalic acid can be used to prepare brookite TiO 2 nanostructures with excellent properties. However, due to the strong dependence of brookite TiO 2 formation on the synthesis method and experimental conditions involved, more research is required to understand brookite TiO 2 formation under different conditions.
In this study, brookite TiO 2 was synthesized via the solvothermal method using tetrabutyl titanate (TBOT) as a Ti source and oxalic acid as a complexing agent. N,Ndimethylformamide (DMF) was used as a solvent, and the pH value of the solution was maintained at 10 by adding 2 mol/L NaOH. X-ray diffraction (XRD), Raman spectroscopy, field emission scanning electron microscopy (FESEM), high-resolution transmission electron microscopy (HRTEM), Fourier transform infrared spectroscopy (FTIR), and ultraviolet-visible diffuse spectroscopy (UV-Vis) were employed to characterize the structure and morphology of the products. The photocatalytic activity of the produced brookite TiO 2 was evaluated by photodegrading methylene blue (MB) solution under UV light irradiation. In addition, this work investigated the formation mechanism of brookite TiO 2 in the presence of oxalic acid during the solvothermal process.

Synthesis of Brookite TiO 2 Nanostructures
In total, 1.27 g of oxalic acid was first dissolved in 50 mL of DMF solution via vigorous stirring, and 1.0 mL of TBOT was then slowly dropped into the mixture. Vigorous stirring was applied until a transparent yellow solution was formed. The pH value of the resulting solution was adjusted to 10 using 2.0 mol/L NaOH during this process. The solution was then stirred for another 1 h, and the slurry formed was transferred to a 100 mL Teflon-lined autoclave and placed in an oven at 180 °C for 8 h. It was then slowly cooled to room temperature, and the white precipitate was subsequently collected using centrifugation. The precipitate was washed several times with distilled water and absolute ethanol and then dried at 80 °C overnight in a drying cabinet.

Characterization
The crystalline phase, phase composition, and crystallite size of the product were characterized using powder XRD via a DX-2000 X-ray diffractometer with CuKα radiation at a scan rate of 0.06°/s. The accelerating voltage and applied current were 35 kV and 25 mA, respectively. The average TiO 2 sample crystallite size (D) was calculated from X-ray line broadening of reflections of the three main peaks using the Scherrer equation.
The morphologies and particle sizes of prepared TiO 2 powders were examined using an FESEM (Hitachi, S-4800), and the detailed microscopic nanostructure was observed using a HRTEM (JEOL, JEM 2100F operated at 200 kV). Raman spectra were recorded using a DXR Microscope Raman spectrometer system with an excitation wavelength of 532 nm at room temperature. Nitrogen adsorption-desorption measurements were taken at 77 K using a Micromeritics TriStar nitrogen adsorption apparatus (Micromeritics Instrument Corporation, USA). The samples were degassed at 300 °C for 4 h prior to Brunauer-Emmett-Teller (BET) measurements, and the BET surface areas, S BET , were calculated within the relative pressure range of 0.05 to 0.20. Desorption isotherms were used to determine the pore-size distribution via the Barrett-Joyner-Halenda (BJH) method. UV-Vis diffuse reflectance spectra were recorded using a UV-visible spectrophotometer (UV-2600, Shimadzu, Japan) and then converted into absorption spectra via Kubelka-Munk transformation.

Photoelectrochemical Evaluation
The electrochemical performance of brookite TiO 2 was evaluated at a CHI660E electrochemical workstation (Shanghai Chenhua Instrument Co., China). Electrochemical measurements were taken using a conventional three-electrode configuration with an as-prepared brookite TiO 2 /FTO employed as the working electrode (1 cm × 1 cm), a Pt wire as the counter electrode, and a Ag/AgCl electrode as the reference. The working electrode of TiO 2 /FTO was prepared according to the method reported in our previous work [24]. The electrodes were placed 3 cm apart, and the supporting electrolyte was 0.1 mol/L Na 2 SO 4 solution. Electrochemical impedance spectroscopy (EIS) measurements were taken in the frequency range of 100-0.01 Hz using amplitude of 5 mV and an open-circuit potential without illumination. The flat band potential (E FB ) of the samples was determined from Mott-Schottky (M-S) analysis at an alternating-current potential frequency of 1 kHz under dark conditions.

Photocatalytic Performance Test
The photocatalytic performance of brookite TiO 2 powders was evaluated using the photodegradation of an aqueous solution of MB. A Xenon lamp (HXS-F/UV 300, Beijing NBET Technology Co., Ltd) with a 365-nm band filter served as the light source and was placed 10 cm away from the liquid surface, where the light density was approximately 10 MW/cm. An appropriate amount of photocatalyst (15 mg) was added into 75 mL of MB solution (10 mg/L). To equilibrate the adsorption and desorption, the suspension was stirred in the dark for 30 min prior to UV illumination. Samples were periodically collected and centrifuged, and the concentration of the MB solution was analyzed using a UV-Vis spectrophotometer (T6, China).
To determine the effect of active species generated by the photocatalytic system on the photocatalytic activity of samples, various scavengers, including 2-propanol (IPA, 20 mmol/L) and EDTA-2Na (6 mmol/L), were introduced into the MB solution.

XRD, Raman, and FTIR Analyses
The crystalline structures of brookite TiO 2 synthesized via the solvothermal method in the presence of oxalic acid at 180 °C for 8 h were confirmed using XRD patterns and indexed using the Joint Committee on Powder Diffraction Standards (JCPDS), as shown in Fig. 1a.
All peaks of the sample synthesized in the presence of oxalic acid were confirmed to show the orthorhombic crystal structure of brookite TiO 2 (PDF# 29-1360). There were no other characteristic peaks observed belonging to impurities or other TiO 2 polymorphs (anatase and rutile), which indicated that the sample was pure-phase brookite TiO 2 . To investigate the effect of oxalic acid on brookite phase formation, another sample was synthesized under the same solvothermal conditions in the absence of oxalic acid and was found to be titanic acid (H 2 Ti 2 O 5 ·H 2 O; PDF# 47-0124) instead of brookite TiO 2 (as shown in Fig. 1a) [9]. This result indicates that oxalic acid plays an important role in the crystallization and phase formation of brookite TiO 2 . However, oxalic acid is not the sole factor influencing the phase formation of pure brookite; a different crystal phase (pure anatase, pure rutile, or anatase-brookite biphase) TiO 2 was synthesized at a different pH value when oxalic acid was used as a chelating agent [17,19]. It is thus concluded that brookite formation in this study is primarily attributed to the synergistic effect between oxalic acid and sodium hydroxide. 10   In addition, the pure brookite phase is only obtained under mild alkaline conditions, and titanate is formed when the pH value of the solution is higher than 10 [25]. The thermal phase stability of synthesized brookite TiO 2 was investigated by annealing the sample at different temperatures for 2 h, as shown in Fig. 1b. No phase transition was observed in the XRD patterns, even at a calcination temperature of 700 °C, which indicates that brookite TiO 2 synthesized under the present conditions possesses good thermal stability. The average crystallite sizes of samples without calcination and those calcined at 300, 500, 600 and 700 °C were calculated using Scherrer equation, and the (121) peaks were found to be 27.65, 28.32, 31.86, 33.02, and 35.93 nm, respectively. This result reveals that the thermal treatment temperature has no influence on the phase structure or composition of the samples and it only affects crystal size.
Raman spectroscopy is a sensitive technique used to confirm the existence of pure brookite in a sample, where the lower symmetry and larger unit cell of brookite result in a larger number of Raman and infrared-active phonons compared to anatase and rutile structures [26]. The relatively complex vibrational spectrum includes a stretching vibration (Ti-O) and a flexural vibration (Ti-O-Ti) [8], and the two types of O atoms that exist in brookite make all the bond lengths between the titanium and oxygen atoms different [1]. Figure 2 shows the Raman spectra of the assynthesized brookite TiO 2 powders without calcination and those calcined at 500 °C, respectively. The detected Raman vibration peaks are indexed as A 1g (127, 152, 194, 246, 324, 412, 545 and 640 cm −1 ), B 1g (212 and 283 cm −1 ), B 2g (366, 460 and 542 cm −1 ), and B 3g (500 cm −1 ) [26,27]. Raman peaks are thus in good agreement with the Raman-active modes of brookite TiO 2 . As no characteristic peak belonging to anatase TiO 2 exists at 516 cm −1 , it is suggested that the samples are pure brookite, and that chelation of C 2 O 4 2− with different types of O atoms results in the formation of distorted TiO 6 octahedron that facilitates the growth of brookite crystallites. The Raman spectra of brookite TiO 2 calcined at 500 °C are also presented in Fig. 2, and no phase transition is observed. This result is in agreement with results of XRD. One possible explanation for the thermal stability is the correlation between surface enthalpies and the brookite TiO 2 crystalline size. The energies of anatase, brookite, and rutile are sufficiently close to be reversed by small differences in surface energies. Zhang and Banfield [28] demonstrated that brookite is more stable than anatase for crystal sizes greater than 11 nm, while rutile is the most stable phase for sizes greater than 35 nm. All of the crystal sizes synthesized are within this reference range.

SEM and TEM Analyses
To confirm the morphological features of the as-obtained TiO 2 powders calcined at different temperatures, the samples were further characterized using FESEM and HRTEM. Figure 3a-e shows SEM images of samples without calcination and those calcined at 300, 500, 600 and 700 °C. It can be observed that brookite TiO 2 without calcination exhibits quasi-spherical structures (Fig. 3a), which consist of many smaller particles measuring approximately several nanometers in diameter. After calcination at 300 °C, the quasispherical structures with a rough surface are improved, as shown in Fig. 3b; this is attributed to an improvement in crystallinity. The surface of the spherical structures changes from rough to smooth, and the grain size increases with an increase in the calcination temperature, as shown in Fig. 3c-e. This indicates that the product is sintered when the calcination temperature exceeds a certain value.
To further investigate the crystal morphology, TEM and HRTEM analyses were conducted to confirm the quasispherical TiO 2 structures, as shown in Fig. 3f-i. Figure 3f shows the TEM image of brookite TiO 2 without calcination, which confirms the existence of abundant quasi-spherical nanoparticles with an average particle size of approximately 20-40 nm. It is in good agreement with results from the SEM image (Fig. 3a) and the crystallite size calculated using Scherrer equation. With an increase in the calcination temperature, the diameters of the quasi-spherical structures increase (Fig. 3g, h); this is attributed to agglomeration of the nanoparticles during the thermal treatment process. Moreover, a large number of white dots are observed on the surface of the quasi-spherical structures, as shown in Fig. 3h, which indicates that the sample calcined at 500 °C exhibits a porous structure resulting from an improvement in crystallinity and the accumulation of small particles, and it is consistent with results from the SEM image shown in Fig. 3c. The HRTEM image of a typical quasi-spherical structure is shown in Fig. 3i, and the distinct lattice spacing  Figure 4 shows the nitrogen adsorption-desorption isotherms and pore-size distribution curves of prepared TiO 2 powders calcined at various thermal treatment temperatures. It is observed from Fig. 4a that all synthesized samples show a typical Brunauer-Deming-Deming-Teller (BDDT) type-IV curve. At a high relative pressure range between 0.8 and 1.0, all curves exhibit clear hysteresis loops, which indicates the presence of a mesoporous structure (2-50 nm).

BET Analysis
The N 2 adsorption-desorption isotherms at a high relative pressure, as shown in Fig. 4a, show that the hysteresis loop area increases with an increase in the calcination temperature. It indicates that the calcination temperature has an important influence both on the BET specific surface area and on the average pore size. The average pore sizes were calculated using the Barrett-Joyner-Halenda (BJH) model derived from the desorption branches of isotherms. BET surface area values and average pore sizes are presented in Table 1; it is evident that the BET surface areas of samples change as the calcination temperature increases, which is attributed to the porous structure of the samples resulting from decomposition of the oxalate precursor. The average pore size of samples increases with an increase in the   calcination temperature; this is related to the compacted structure and the improved crystallinity due to the increase in calcination temperature. In addition, more than one peak in the pore-size distribution is observed, as shown in the insets of Fig. 4b-f, which indicates that the as-prepared TiO 2 has a hierarchically porous structure.

Possible Mechanism for Growth of Brookite TiO 2 in the Presence of Oxalic Acid
In recent years, brookite TiO 2 nanostructures (synthesized via thermal treatment of titanium complexes derived from different Ti sources and complexing agents) have been widely reported, and their corresponding growth mechanisms have been investigated [15][16][17][18][19][20]. Some researchers have suggested that a ligand type is crucial to TiO 2 phase formation during the hydrothermal treatment process, because ligand may coordinate to titanium in an octahedral environment and be assembled into the initial structure of anatase, brookite, or rutile crystallites [9-12, 21, 29-32]. In general, in an aqueous medium, titanium cations are solvated and form 6-coordinate charged complexes [TiL 6 ] z+ , in which the nature of the ligand (L) is indicated by the solution pH and the presence of species that act as complexing agents [29]. Oxalic acid is a rigid bidentate ligand; it is a good chelating agent that has been used to synthesize TiO 2 . However, no consensus currently exists about the effect of oxalic acid on TiO 2 phase formation. For example, Dambournet et al. [22] reported that the [C 2 O 4 2− ]/ [Ti 4+ ] molar ratio (denoted as R) was the key factor in TiO 2 phase formation and that the oxalate anion either acts as a ligand to stabilize the oxalate hydrate phase Ti 2 O 3 (H 2 O) 2 (C 2 O 4 )·H 2 O (R = 1) before the formation of brookite TiO 2 after thermal treatment, or acts as a chelating agent (R = 2) to stabilize the rutile phase. However, Kanna and Wongnawa [30] suggested that oxalic acid inhibited the formation of rutile because the bi-bonds of C 2 O 4 2− on the TiO 6 octahedra occupy one full octahedral face and inhibit the growing chain along opposite edges. Truong et al. [19] proposed that the presence of oxalic acid in solution prevented anatase formation through its chelation to the TiO 6 octahedron at the bridging oxo group, which benefited the growth of rutile crystallites. However, the results obtained in this study differed from the reported results when using oxalic acid as a complexing agent. Specifically, pure-phase brookite TiO 2 instead of rutile was obtained under solvothermal conditions with R > 2, which indicated that brookite TiO 2 derived from the titanium oxalate phase could be obtained with higher oxalate content by tuning other synthesis conditions. It can thus be inferred that the formation of brookite TiO 2 is related to the existence of NaOH, which is introduced to maintain the pH value at approximately 10. Dambournet et al. [15] reported that the titanium oxalate phase played an important role in the morphology and crystallization of TiO 2 due to the type of preferentially adsorbed alkali ions (Li + , Na + ) on its crystal facets. The three crystal forms of TiO 2 , including anatase, rutile, and brookite, are composed of a [TiO 6 ] octahedron that has oxygen ions at its vertices and titanium atoms at its center; these have different spatial arrangements that share edges and corners in different manners. The [TiO 6 ] octahedral structure prefers corner sharing to edge sharing to maintain its thermodynamic stability, because the repulsive Coulomb forces that result from the shared edges lead to a decrease in stability [33]. Therefore, it is here considered that the addition of oxalic acid and NaOH changes the spatial arrangement of the [TiO 6 ] octahedron, which results in the formation of brookite TiO 2 . Na + ions have been reported to favor the brookite transition because they retard the collapse of layers due to stronger electrostatic interactions with titanate layers and induce the phase transition of titanate into brookite TiO 2 [34]. In addition, it has been reported that the pH value of solution significantly influences the TiO 2 crystalline structure [3]; a high pH supports edge-shared bonding (anatase crystallization), while a low pH favors corner-shared bonding (rutile crystallization). Brookite has both shared edges and corners and is midway between anatase and rutile in terms of edge-shared bonding, and at a low NaOH concentration, OH − deficiency leads to the formation of [TiO 6 ] octahedra with brookite rather than anatase.
To further investigate the effects of oxalic acid and NaOH on brookite TiO 2 formation, a sample was synthesized using the same solvothermal conditions in the absence of oxalic acid. The XRD result in Fig. 1a demonstrates that titanic acid instead of brookite TiO 2 was obtained. Therefore, we conclude that the formation of brookite TiO 2 is a result of the combined action of oxalic acid and NaOH, and the possible mechanism involved in this process can be deduced. Specifically, the bond between ligands and the Ti precursor is fragile in a high pH solution and is readily cleaved because of the frequent nucleophilic attack on Ti atoms from abundant hydroxide ions. Thus, hydrolysis and condensation occur primarily along the edge-sharing bonds, which results in the formation of anatase and brookite crystallites. Therefore, the action of oxalic acid and the adjustment of pH value with sodium hydroxide are the main factors involved in forming pure brookite TiO 2 .
We thus propose that the formation mechanism of brookite TiO 2 under the current experimental conditions is related to four processes: (1) the formation of Ti-OH groups derived from the tetrabutyl titanate hydrolysis reaction, (2) the complexation reaction between the oxalate ligands and the Ti-OH groups via the bi-coordination process, (3) the transformation of titanium oxalate-based compound with the addition of NaOH, and (4) the formation of brookite TiO 2 under solvothermal conditions, including a high pH value (pH 10), which is described in Fig. 5.

UV-Vis Spectra
The diffuse absorbance spectra of as-synthesized brookite TiO 2 were analyzed using a UV-Vis spectrophotometer, as shown in Fig. 6a. The absorption edges were estimated to be 367.3, 364.4, 364.0, 363.6 and 363.9 nm for samples without calcination and those calcined at 300, 500, 600 and 700 °C, respectively. A slight blueshift occurs with an increase in calcination temperature. The optical band gap (E g ) of assynthesized brookite TiO 2 was determined using Eq. (1): where α is the linear portion of the absorption coefficient; hν is photon energy; and A is a constant [35]. As brookite TiO 2 is an indirect band gap semiconductor, the value of n is determined to be 2. The plots of (αhv) 1/2 versus hv from the spectral data in Fig. 6a are presented in Fig. 6b-f. Extrapolation of the linear parts of the curves yields indirect band gaps of 2.939, 3.313, 3.330, 3.335 and 3.338 eV for samples without calcination and those calcined at 300, 500, 600 and 700 °C, respectively. Di Paola et al. [36] reported that an increased band gap corresponds to a more powerful redox ability, which contributes to enhanced photocatalytic activity. It is thus concluded that the photocatalytic activity of the samples is improved after calcination.

Photoelectrochemical Analysis
It is known that the photocatalytic activity of TiO 2 depends on its structure and physical properties [37]. Examination of electrochemical impedance spectroscopy (EIS) and flat band potential (E FB ) of the brookite TiO 2 /FTO electrodes may be helpful in understanding the relationship between the photocatalytic activity of brookite TiO 2 and its physical properties. EIS is an effective approach for investigating the capacitance and resistance of electrode materials and has been used to investigate the electrochemical properties of brookite TiO 2 . EIS measurements were taken in a 0.1 mol/L Na 2 SO 4 solution at an AC frequency from 100 kHz to 0.1 Hz without irradiation; results of EIS are presented in Fig. 7a in the form of Nyquist plots. It was found that the impedance arc diameter of brookite TiO 2 first sharply decreased as the calcination temperature increased below 500 °C and then increased when the calcination temperature reached higher than 500 °C. This indicates that the electron-transfer rate from the electrolyte to the electrode surface is enhanced by the increasing calcination temperature below 500 °C, which is attributed to the improved crystallinity and particle size of brookite TiO 2 . However, the sample was sintered when the calcination temperature was too high (over 500 °C), which resulted in an increased number of crystalline defects. The crystal defects correspondingly increased, thereby promoting the recombination of electrons and holes.
The semiconductor flat band potential (E FB ) is a fundamental thermodynamic property during interfacial electrontransfer steps. In this study, E FB at the electrolyte interface of the brookite TiO 2 /FTO electrode was obtained using Mott-Schottky analysis (measured in the dark); results were then used to further investigate the intrinsic electronic properties of brookite TiO 2 . The semiconductor capacitance is described using the Mott-Schottky equation: where e 0 is the fundamental charge constant; ε is the dielectric constant of TiO 2 ; ε 0 is the permittivity of vacuum; N D is donor density; E is the electrode applied potential, E FB is the flat band potential; and kT/e 0 is the temperature-dependent correction term [38]. The Mott-Schottky plots of the brookite TiO 2 /FTO electrode are presented in Fig. 7b- Table 2, the N D value for brookite TiO 2 calcined at 500 °C is the highest (2.83 × 10 22 cm −3 ).
To further investigate photoelectrochemical properties, the transient photocurrent of brookite TiO 2 /FTO electrodes in a 0.1 mol/L Na 2 SO 4 electrolyte without any sacrificial reagents or catalysts was assessed. The photocurrent response was measured at a constant bias of 0.2 V, as shown in Fig. 8. All samples were observed to have good photoresponses in the chopped light cycles. Without illumination, the current values were approximately zero. However, the photocurrent rapidly rose to a steady-state value upon illumination, which was reproducible for several on/off cycles using an approximately identical photocurrent and dark current. As the calcination temperature increased, the steady-state photocurrent first increased below 500 °C and then decreased. This is attributed to the improved charge-transfer power due to improved crystallinity.

Photocatalytic Performance Evaluation
The photocatalytic performances of brookite TiO 2 nanopowders were evaluated by photodegrading an MB solution. For comparison, a blank experiment for MB  ). In addition, the photodegradation kinetics of MB over different samples was investigated to quantitatively characterize the photocatalytic activity of the samples, from which the apparent pseudo-first-order reaction rate constants (k) were calculated (as presented in Fig. 9b).
The k values of samples without thermal treatment and those calcined at 300, 500, 600 and 700 °C were 0.01172, The same results were observed for photocatalytic activity. The photocatalytic activity of TiO 2 catalysts strongly depends on two factors: the adsorption behavior and the separation efficiency of electron-hole pairs [40][41][42][43]. Active oxygen species (such as hydroxyl radicals (OH), superoxide radicals (O 2 − ), and H 2 O 2 derived from the photogenerated electrons and holes) are considered to be the major active species responsible for the photocatalytic activity of TiO 2 [40][41][42]. To further understand the photocatalytic mechanism of the as-synthesized brookite TiO 2 , hydroxyl radicals and hole trapping experiments were designed using 2-propanol as the radical scavenger and EDTA-2Na as the holes scavenger. Results are shown in Fig. 9c, where it can be observed that the decolorization efficiency of MB with addition of EDTA-2Na is slightly smaller than that of MB with no additive. However, the decolorization efficiency of MB with addition of 2-propanol sharply decreases (65% after illumination for 90 min). This result demonstrates that ·OH is one of the major active species in this experiment and that it possesses a strong oxidation ability (its oxidation-reduction potential is + 2.27 V, vs. NHE, pH 7). It is known that ·OH radicals are generated via two pathways: (1) photoelectrons are captured by adsorbed O 2 on the surface of TiO 2 to form intermediate species, such as superoxide radicals (O 2 − ) and H 2 O 2 , and ·OH radicals are then generated after the series reaction; (2) the holes oxidize adsorbed water or the surface hydroxyl groups to form OH radicals [41]. Therefore, based on the result of our hole trapping experiment, we infer that superoxide radicals (O 2 − ) and H 2 O 2 play important roles.
The absorption behavior is also another important factor contributing to the photocatalytic activity of TiO 2 catalysts [42][43][44]; the absorption behavior of organic species on the catalyst surface influences the photoinduced electron transfer, which reduces the rate of the compound cavity electronic structure and increases the photocatalytic activity. However, absorption behavior is significantly influenced by crystallinity, specific surface area, particle size, and morphology [29][30][31][32][33][45][46][47][48][49][50]. Samples calcined at different temperatures had similar particle sizes and morphologies according to the XRD and SEM analyses. However, the specific surface areas and pore sizes were significantly different. Brookite TiO 2 calcined at 500 °C for 2 h had the largest specific surface area (shown in Fig. 4 and Table 1), which explains the highest photocatalytic activity. It is evident that the crystallinity of samples, which is improved after calcination at a relatively high temperature (below 500 °C), is beneficial in reducing the recombination rate of photogenerated electrons and holes (due to a decrease in the number of defects) [51].

Conclusions
In the current study, we have synthesized quasi-spherical brookite TiO 2 nanostructures via a solvothermal method in the presence of oxalic acid; a NaOH solution was introduced to maintain the pH value at approximately 10. Experimental results indicated that pure-phase brookite TiO 2 with a quasi-spherical structure was successfully synthesized under solvothermal conditions of 180 °C for 8 h. Oxalic acid played an important role in the morphology of products, and the formation of pure-phase brookite TiO 2 was attributed to the combined action of oxalic acid and the pH value, which was adjusted with sodium hydroxide. Photocatalytic activity and photoelectrochemical performance tests indicated that the resulting brookite TiO 2 nanopowders calcined at 500 °C for 2 h exhibited high photocatalytic activity for methylene blue (MB) solution, and the degradation rate of methylene blue (10 mg/L) reached 96.7% after UV light illumination for 90 min at 10 mW/cm 2 .