Effect of Annealing and Etching Times on Anatase TiO 2 Hollow Sphere

. The development of high-efficiency photocatalysts plays an important role in the application of solar energy conversion. Titanium dioxide (TiO 2 ) with an anatase crystalline phase is well-known as semiconductor thin layers for solar cells. This work has constructed a novel TiO 2 hollow sphere (HST) in a 2-step synthesis. The first step is coating the SiO 2 template with TiO 2 to build a core-shell of SiO 2 @TiO 2 (CSST). The second step is etching via sonication to elute the SiO 2 from CSST and construct the HST. The annealing of CSST for 1 to 6 hours and etching for 1 to 7 hours has resulted in the HST with different crystallite sizes and microstrains. The HST of ~90 nm has been fabricated with crystallite size of 9.53 to 20.54 nm and microstrain from 0.34 to 3.42. It was found that the optimum crystallite size and microstrain of HST obtained via annealing the CSST for 2 hours and etching for 5 hours has the best photooxidation of I - under UV irradiation. The optimum crystallite size and microstrain of HST via annealing and etching times can be recommended for the future of solar cell fabrication and applications.


INTRODUCTION
In recent years, the architecture of hollow sphere TiO2 (HST) has attracted particular attention in photocatalyst and solar cell applications because of the advantages of HST which have a wide band gap and incident light scattering (Fang et al., 2022). These characteristics are essential to enhancing the efficiency of the photocatalytic activity (Shakeel et al., 2017). Furthermore, the size of particles and cavities in HST also impacts the scattering of light of the photocatalysts, which can be advantageous in improving performance. The small size of HST makes a large surface area into a sizeable light-contact space (Liao et al., 2012). While the cavities of HST as this morphological uniqueness provide an extension of the wavelength to enhance the scattering effect (Jafarzadeh et al., 2016;Tsai et al., 2014). Those properties can be designed based on the HST fabrication.
HST fabrication has two synthesis pathways, namely hydrothermal and sol-gel. The hydrothermal synthesis of HST has a single step while sol-gel involves two steps. Hydrothermal was implemented in an autoclave at high temperatures (Wang et al., 2014), producing more than 100 nm HST of particle size (Zhu et al., 2013). Meanwhile, the sol-gel synthesis of HST began with the templating through core@shell formation and continued with removing the core to form HST (Choi et al., 2011). The sol-gel by carbon-based templates have fabricated HST with particle sizes between 60 and 365 nm (Choi et al., 2011) while using the SiO2-based templates could produce 25-100 nm HST particle size (Chen et al., 2012;Oh et al., 2010).
Photocatalysts with smaller particles have large surface areas which enhance the contact area and scatter the incident light, thus increasing the photocatalytic activity. Consequently, the sol-gel method using the SiO2 template is preferred for fabricating the HST because it can produce smaller particles compared to the hydrothermal method (Jeng et al., 2013). The HST fabrication begins with the formation of CSST by coating the SiO2 template synthesized by the Stöber method with a TiO2 precursor (Choi et al., 2011;Kim et al., 2010 ). The particle size, cavity, and shell thickness of HST can be designed at this step, while the anatase phase has been constructed after annealing of CSST at more than 500 °C (Kanjana et al., 2020;Oh et al., 2010;Yao et al., 2020). Hence, an annealing process is an essential step to construct the anatase TiO2. Zheng et al. (2018) have reported the optimum annealing temperature of CSST to construct anatase HST at 900 °C. The anatase HST constructed at 500 C has the low peak of the XRD diffractogram and collapsed at 1100 °C. Yun et al. (2015) synthesized anatase HST at 6 hours of annealing at 900 °C. The appropriate annealing time in synthesizing TiO2 nanoparticles enhances the crystallite size, dislocation density, microstrain, and oxygen vacancy responsible for the increased magnetic moment in TiO2 nanoparticles (Doubi et al., 2019;Thejas et al., 2022). The annealing time also affects the optimum optical properties in the preparation of the TiO2 nanoparticles (Tighineanu et al., 2010). The optimum microstrain and dislocation density showed a better adsorption performance (John et al., 2021). The optimum microstrain of anatase TiO2 has also a high efficiency of photocatalytic activity. Therefore, to get the optimum microstrain, the optimization of annealing time has been studied to enhance the photocatalysis activity of HST.
Next, removing the SiO2 core from CSST using an alkaline solution (etching process) is the second step to produce the HST. Sodium hydroxide (NaOH) solution was chosen because of its high selectivity to soluble SiO2 (Choi et al., 2011;S. Li et al., 2013;Lyu et al., 2020). The NaOH solution in this process pushes high energy intensity into the voids between the silica-titania and causes partial etching of the SiO2 core (Chen et al., 2012). The etching of SiO2 by NaOH solution was reported at different times, 3 to 6 hours (Oh et al., 2010;Yun et al., 2015;Zheng et al., 2018). Meanwhile, the etching time is thought to affect the anatase HST microstrain after losing the SiO2 core.
To avoid concerns about changes in HST morphology because of microstrain changes from the impact of etching time, the etching time optimization is studied in this article to obtain the optimum etching time.
Time is our concern for the fabrication of anatase of HST, while the annealing and etching are the main steps to constructing the ones. Although the optimal annealing temperature of HST was reported, the annealing and etching times have not been studied previously. Generally, studies of HST as a photocatalyst have widely reported has superior morphology with unique pore characteristics and the ability to scatter light to enhance the efficiency of photocatalytic activity. Thus, in this work, we studied the effect of annealing and etching times to obtain the optimum crystallite size and microstrain of anatase HST. The photocatalytic activity of the anatase HST was determined using iodide ion (I -) photooxidation. According to , photooxidation Imimics the process of electron regeneration in the redox electrolytes (I -/I3 -) of a DSSC system. Therefore, the photocatalytic activity of HST on this system can serve as an indicator of its performance when applied to DSSCs, much like the redox performance in the DSSC system.

EXPERIMENTAL SECTION Synthesis of HST
All chemicals are of analytical grade quality without any further purification. The synthesis of HST began with the construction of the SiO2 sphere by the Stöber method (Meier et al., 2018;Stober, 1968). Firstly, 3.6 mL of titanium(IV) isopropoxide (TTIP) (Merck) precursor was dissolved into 8 mL of ethanol (Merck). The solution was mixed with a solution of 6 mL of acetonitrile (Merck) and 10 mL of ethanol, then added to 30 mL of colloidal SiO2 nanoparticles. The TTIP solution and SiO2 were reacted by stirring via a sol-gel coating process for 6 hours above an ice bath. The colloid CSST formed was obtained by centrifuging and sequentially washing with ethanol and deionized water. The CSST was annealed at 900 °C for altered times: 0, 1, 2, 4, and 6 hours. The HST was etched by eluting the CSST in 2.5 M NaOH (Merck) solution with a weight-to-volume ratio of 1:4, then sonicating for time variations of 0, 1, 3, 5, and 7 hours.

Photocatalytic Activity
The photocatalytic activity of HST 1-6 (different of CSST annealing times 1-6 hours and after etching 5 hours) was analyzed based on the I3formed from the photooxidation of Iin 10 mL KI (Merck) 0.2 M and 50 mg of HST by stirring and irradiation at 25 cm under UV light (40 W, with a maximum emission at about 254 nm) at room temperature for 15, 30, 45 and 60 minutes. The suspension was centrifuged, then the supernatant was diluted 10 times. The absorbance of the solution was monitored by a UV/Vis Spectrophotometer (Thermo Scientific, Orion AquaMate 8100) at 288 nm (Pavasupree et al., 2006). The concentration of I3was determined by Equation 1. In the dark and without the HST samples were observed no I3formed. A = ɛ.b.c (1) Where A: absorbance, a: the molar extinction coefficient 4.010 4 (cm mol/L) -1 , b: cuvette thickness, and c: concentration.

Characterization
The HST sample was characterized using transmission electron microscopy (TEM, JEOL: JEM-1400) and a field emission scanning electron microscope (FESEM, Quanta FEG-650). While the diffractograms of the crystallite phase of the sample were determined by X-ray diffraction (Shimadzu, XRD-6000) with a Cu Kα radiation source (λ = 1.5406) at 40 kV and 300 mA (12 W). Fourier transform infrared (FTIR, Shimadzu prestige 21) was used to identify the functional groups of the sample, and a UV/Vis Spectrophotometer (Shimadzu, UV-2401PC) determined the UV/Vis diffuse reflectance spectra. The SiO2 loss via the etching process was monitored using an energy-dispersive X-ray spectrometer (EDS, JEOL JSM-6510LA).

RESULTS AND DISCUSSION Synthesis of Hollow Sphere TiO2
The fabrication of HST begins by preparing the surface of the SiO2 template in a positive charge with an ammonia solution. The positive charge on the surface of SiO2 by NH4 + drives the interaction with Ti-OH and constructs the CSST via the condensation process at low temperatures ; the SiO2 template was then eluted by an etching process using sodium hydroxide solution to produce HST (S. Lyu et al., 2020). An illustration of these reactions can be seen in Figure 1.a.
The morphology and structure of synthesized HST are presented in Figure 1.b to 1.e. FESEM image (Figure 1.b) showed a SiO2 template with a particle size of ~65 nm generated from the Stöber method. (d) CSST has an inner diameter of ~45 nm, with a shell thickness of ~25 nm. The SiO2 template has shrunk after CSST construction and annealing at 900 C because of the arrangement of the anatase TiO2 (Yun et al., 2015). While Figure 1.e shows HST with a cavity of ~45 nm and a shell thickness of ~26 nm. The shell thickness of CSST and HST is supposed to be constant because there is no TiO2 crystal etch in the etching process.
The functional groups in CSST and HST can be identified through the FTIR spectra (Figure 2.a). The Si-O-Ti peak was identified at 949 cm -1 in the SiO2 template and SiO2/TiO2, as reported by Chang    On the other hand, Figure 2.c shows the UV-Vis diffuse reflectance spectra. The edge reflectance at 400 nm supports the morphological features of CSST and HST, suggesting the presence of the anatase profile of TiO2. This observation is consistent with previous findings on a core-shell structure of TiO2(rod)-SiO2 (Wahyuni et al., 2018;Wahyuni et al., 2017). The presence of a cavity can extend the wavelength path (Lei et al., 2014), which makes the HST has a lower reflectance than P25 and CSST at a wavelength of 200-400 nm, which is 3.04 eV related to the band gap of HST.

Effect of Annealing and Etching Time
The crystal structure of HST anatase with the annealing process was slowed down by Si-O-Ti in CSST, thus requiring more high temperatures than the pristine (Zheng et al., 2018). The construction of the anatase crystal structure began at one hour of annealing and was steady after six hours (Figure 3). The peak intensity of 101 at 25 degrees of 2 was the highest at 2 hours and decreased after 4 hours of annealing. The annealing time affects the crystal size and microstrain accessed by Williamson-Hall (W-H) plotting (Table 1). The W-H analysis used the XRD profile by linear plotting of βcos vs sin, as following Equation 2 (Kibasomba et al., 2018). βcos = Kλ/D + 4ɛsin (2) Where β is the full-width-at-half-maximum (FWHM) of the i-th peak,  is the angle from the i-th peak radian found, K is the instrument alignment constant with a magnitude of 0.94, and λ is the wavelength of the Xray source of Cu K at 1.54 Å, and ɛ is the microstrain (Rehani et al., 2006).
The W-H analysis involved creating a graph where βcos was plotted on the y-axis and 4sin on the x-axis. From this graph, a linear regression was determined, and the resulting intercept and slope were utilized to calculate the crystallite size (D) and microstrain (ɛ). Table 1 shows the optimum crystal size and microstrain of HST anatase at 20.54 nm and 3.42, respectively, after 2 hours of annealing. Crystal size and microstrain decreased after a longer annealing process. This affects the reduced surface area and oxygen content (John et al., 2021). From Table 1, etching time affected the crystallite size and microstrain of HST. The crystallite size of HST was decreased in a longer time of etching, except in 7 hours of etching were enhance the microstrain of HST. The investigation of the SiO2 etching process by an energy-dispersive X-ray spectrometer is shown in Figure 4.a.
The optimum time for removing the SiO2 core using 2.5 M NaOH solution was 5 hours (Figure 4.b). The etching process via sonication after 5 hours was steady at less than 2% mass percent of Si. The etching time has improved the crystal size and microstrain HST ( Table 1). The XRD profile shows the crystallite size of TiO2 anatase CSST shrinks after the etching process. Etching for 5 hours constructed a crystallite size of 12.66 nm, with the lowest microstrain of 1.84. The etching via sonication for 5 hours is the optimum time to construct anatase TiO2 of HST via sonication.

Photocatalytic Activity
The photocatalytic of iodide ion (I -) from KI solution by HST has changed the clear solution to bright yellow which is indicative of the formation of I3 -. The photocatalytic activity of HST was evaluated by the I3formation from the oxidation of Ito I2 in the excess iodide, following the reaction as shown in Equation 3-7 (Adachi et al., 2003;Watanabe et al., 2019). While the involvement of HST refers to the theory and principle of the nanostructure-TiO2 universal photocatalyst, namely the indirect photocatalytic mechanism (Long et al., 2017). The concentration of I3from photooxidation with HST 2 catalyst at 60 min of 10.210 -6 M is higher than HST 4, 6, and 1 ( Figure 6). At the same irradiation time, HST 4, 6, and 1 produced I3concentrations of 9.0810 -6 , 8.8510 -6 , and 6.7310 -6 M, respectively. Crystallite size and microstrain have affected the photooxidation of KI. The photocatalytic activity showed that the large crystallite size and microstrain of HST enhance the concentrations of I3 -. The I3was produced after the KI Figure 6. Photocatalytic activity (I3¯ concentration) of the HST annealed at altered times solution and the catalyst were stirred under UV light. Overall, we have shown that morphology engineering through the HST formation has resulted in improved photocatalytic performance. A similar finding of improved performance of the photocatalyst due to morphological changes has also been observed for silver phosphate photocatalysts (Sulaeman et al., 2022). As a result, there is a growing interest in conducting further research on shaping the morphology of photocatalysts.

CONCLUSIONS
Hollow sphere TiO2 (HST) has been successfully synthesized by the sol-gel method through the formation of a core-shell silica@titania (CSST). The anatase crystal of HST was constructed at 1 hour of annealing at 900 C. Meanwhile, the removal of the SiO2 core to form HST was obtained after 3 hours of the etching process with 2.5 M NaOH solution. Synthesis from this method fabricates HST with a particle size of ~90 nm, shell thickness of ~26 nm, and cavity size of ~45 nm. The synthesis of HST with different annealing and etching times has improved the crystallite size and microstrain. The large crystallite size and microstrain can increase the photocatalytic activity of photooxidation of I¯.