The precursor was titanium(IV) isopropoxide (TTIP, 97%, Sigma-Aldrich). A 0.2 M TTIP solution was made by adding TTIP to 64 mL of deionised water while stirring at room temperature. Hydrochloric acid (HCl, 37%; 0.4 mL) was subsequently introduced dropwise under continued stirring. After an additional 3 h of stirring, the sol was aged at room temperature for 48 h before use.

Glass substrates measuring 25.4 mm × 10.0 mm × 1.0 mm were cleaned successively with acetone, ethanol, and distilled water in an ultrasonic bath for 10 min. The substrates were then dried at 110˚C for 2 h. For coating, about 90 μL of aged sol was placed on each substrate and spin-coated under vacuum at 1500 rpm for 30 s, using a ramp rate of 500 rpm∙s⁻¹. Each substrate received two coating cycles, followed by drying at 110˚C for 1 h.

The coated substrates were subsequently annealed for 3 h at four temperatures: 200˚C, 300˚C, 400˚C, and 500˚C. This temperature series was selected to track the effect of progressively stronger thermal treatment on phase constitution, microstructure, optical response, and MB-removal performance.

GAXRD measurements were obtained with a PANalytical X’Pert PRO MPD (PW 3060/60) instrument using Cu Kα radiation (λ = 1.5406 Å). The scans covered 2θ = 10˚ - 80˚ with a 4˚ incidence angle; the operating conditions were 40 kV and 30 mA. Phase identification was performed using X’Pert HighScore. The crystallite size was calculated from the principal brookite diffraction peak using the Scherrer equation:

Here, D is the mean coherent-domain dimension, k is the shape factor (0.94), λ is the Cu Kα wavelength (0.15406 nm), β is the FWHM expressed in radians, and θ is the Bragg angle. The principal diagnostic reflections for anatase, rutile, and brookite occur near 25˚, 27˚, and 31˚, respectively. Because only brookite reflections were detected from the 200˚C - 400˚C films, the brookite (111) peak near 31˚ was used for the size determination.

TEM examination was carried out with an HT-7700 transmission electron microscope at 120 kV. Micrographs were collected at magnifications of 600,000× to examine the film microstructure.

Optical absorption was measured from 200 to 1100 nm using a PerkinElmer Lambda 35 UV-Vis spectrophotometer. The optical band gap was estimated from the Tauc representation, using the linear region of (αhν)² versus hν and extrapolating that region to the energy intercept. The resulting values should therefore be regarded as optical-edge estimates obtained from the measured spectra.

For photocatalytic evaluation, a 150 W germicidal UV-C lamp was positioned 16 cm from the solution. A coated film was placed in 35 mL of 1 × 10⁻⁶ M MB solution for each experiment. Before illumination, the suspension was kept in darkness for 30 min so that adsorption and desorption could approach equilibrium. Irradiation was then performed for 4 h, and samples were taken at 60 min intervals using separate beakers for absorbance measurements. The visible-light experiment reported in this work used a 200 W OSRAM model.

MB concentration was followed from the absorbance measured at 665 nm with a SHIMADZU UV-1700 UV-Vis spectrometer. Under the Beer-Lambert proportionality assumption, the percentage removal was calculated from either concentration or absorbance according to:

C₀ is the MB concentration after the dark-equilibration period, and C is the concentration at the selected irradiation time. A₀ and A denote the corresponding absorbances before and after irradiation, respectively.

The diffraction response changed markedly with heat-treatment temperature (Figure 1). At 200˚C, reflections assigned to brookite (111) and (023) occurred at approximately 31˚ and 66˚. The ~31˚ reflection gives a d-spacing of about 2.8 Å and is more consistent with JCPDS 84-1750 (31.3˚, 2.81 Å) than with JCPDS 29-1360 (30.8˚, 2.9 Å). After treatment at 300˚C and 400˚C, the brookite (111) reflection remained the only detectable crystalline TiO₂ peak and became weaker with increasing temperature. No distinct crystalline TiO₂ peak was resolved after treatment at 500˚C, so that specimen was classified as predominantly amorphous on the basis of XRD.

The progressive weakening of the brookite reflection is compatible with the metastability of brookite and its sensitivity to the chemical and thermal history of the precursor-derived coating. Earlier work indicates that brookite formation depends on nucleation conditions, crystallite dimensions, and possible deviations from ideal stoichiometry at both surfaces and within the material [15][16]. Allen et al. [17] reported a comparable reduction in brookite fraction, from 39.7% at 110˚C to 15.1% at 600˚C, followed by loss of detectable brookite at still higher temperatures [18]. The present data therefore suggest that the selected thermal schedule can progressively reduce the stability or detectability of brookite in this particular film system.

Figure 1. GAXRD patterns of spin-coated TiO₂ films after heat treatment at the indicated temperatures.

The amorphous appearance at 500˚C may also reflect the chemistry established during sol preparation. In the present procedure, TTIP was introduced at approximately one drop per minute, and HCl was used to influence hydrolysis and nucleation [19]. Solvent composition can modify precursor interactions and the hydrolysis/condensation sequence; differences between water- and methanol-containing systems have been reported [20]. Accordingly, the 500˚C observation should be linked to the complete processing route used here, rather than interpreted as a general rule for every TTIP-derived brookite coating.

Using the brookite (111) peak, the Scherrer analysis produced 47.9 nm ± 19.9 nm at 200˚C, 58.4 nm ± 12.0 nm at 300˚C, and 7.8 nm ± 1.0 nm at 400˚C (Figure 2). A size value was not assigned to the 500˚C specimen because its XRD pattern did not contain a sufficiently resolved brookite reflection.

Figure 2. Crystallite size estimated from the brookite (111) reflection versus heat-treatment temperature.

The size increase between 200˚C and 300˚C suggests development of larger coherent diffracting regions during the initial stage of thermal treatment. This behaviour changed at 400˚C: the calculated value fell to 7.8 nm ± 1.0 nm while the brookite (111) intensity was strongly reduced. Such a reversal can be associated with destabilisation of brookite and a reduction in the size of coherently diffracting regions [21]. The measured sequence should therefore be described as non-monotonic rather than as continuous thermal grain growth.

Table 1 compiles the crystalline-phase assignments from GAXRD. Brookite was detected at 200˚C, 300˚C, and 400˚C, with no resolvable anatase or rutile reflections. At 500˚C, the pattern was predominantly amorphous and did not provide a reliable crystalline-phase assignment. Within this experimental window, the principal change at the highest treatment temperature was thus the disappearance of detectable brookite crystallinity rather than a demonstrable conversion to anatase or rutile.

Table 1.Detectable TiO₂ crystalline phases in the spin-coated films after heat treatment.

TEM and SAED results for the 200˚C and 300˚C films are presented in Figure 3. At 200˚C, the film contained mainly near-spherical features and lattice fringes of approximately 0.28 nm, matching the brookite (111) assignment obtained from GAXRD. The corresponding SAED pattern was also compatible with brookite. After treatment at 300˚C, the observed morphology was more sheet-like; nevertheless, the ~0.28 nm spacing persisted, and the SAED pattern continued to support the brookite assignment.

Figure 4 shows the UV-Vis response of the films between 250 and 700 nm. A strong absorption feature centred near 280 nm was present for all specimens, in agreement with previously reported brookite TiO₂ behaviour [22][23]. Tauc plots were used to estimate the optical band gap from the linear portion of the (αhν)² versus hν relationship. The resulting intercepts were compared to determine how the optical edge varied with thermal treatment.

The calculated optical gaps were 3.42 eV (200˚C), 3.37 eV (300˚C), 3.55 eV (400˚C), and 3.90 eV (500˚C). Thus, the band gap reached a minimum at 300˚C and subsequently widened. The 200˚C - 400˚C values fall within the range reported for brookite TiO₂ [5]. The changes may reflect temperature-dependent microstructural disorder and defect-related electronic states, which have previously been associated with shifts in TiO₂ optical-gap behaviour [24] (Figure 5).

Figure 3. TEM images of the TiO₂ films treated at (a) 200˚C and (b) 300˚C; insets show corresponding SAED patterns.

Figure 4. UV-Vis absorption spectra of the brookite TiO₂ coatings after heat treatment.

Figure 5. Optical band-gap estimates for the brookite TiO₂ coatings at different heat-treatment temperatures.

Under UV-C irradiation, the blank solution exhibited 18.5% apparent MB removal after 1 h and 38.1% after 4 h (Figure 6). The presence of a TiO₂ coating increased removal relative to the blank for every heat-treatment condition. After 4 h, the efficiencies were 63.0% at 200˚C, 67.7% at 300˚C, 62.0% at 400˚C, and 57.1% at 500˚C. At the 1 h point, the corresponding values were 38.4%, 44.9%, 38.8%, and 31.7%. The maximum four-hour removal was therefore obtained from the 300˚C film.

The superior UV-C result for the 300˚C film coincided with its largest calculated crystallite size among the crystalline specimens, 58.4 nm ± 12.0 nm, and with the strongest observed MB adsorption. Literature reports have connected photocatalytic response with crystallite dimensions over a broad range, approximately 1.67 - 67 nm [25]-[27], and crystallite size can influence both adsorption and charge transport [28]. The present data consequently support a relationship between structural development and activity at 300˚C, but they do not demonstrate that crystallite size alone determines the degradation rate.

The 300˚C sample also possessed the narrowest measured band gap, 3.37 eV. For comparison, Komaraiah et al. [27] reported 92.03% MB degradation from a

Figure 6. Methylene blue (MB) degradation under UV-C irradiation for TiO₂ films treated at different temperatures.

brookite TiO₂ thin film with a 3.30 eV band gap. The present 400˚C and 500˚C films gave lower four-hour removal values of 62.0% and 57.1%, respectively. Their structural states also differed: the 400˚C film retained a measurable 7.8 nm ± 1.0 nm crystallite size, whereas the 500˚C film was predominantly amorphous. The activity decrease at higher treatment temperatures therefore occurs together with changes in both crystallinity and the optical gap.

Under the reported visible-light condition, the blank solution reached 19.2% removal after 1 h and 28.9% after 4 h (Figure 7). All coated specimens performed better than the blank. Removal by the 300˚C film rose from 58.1% at 1 h to 97.8% at 4 h, while the 200˚C film increased from 34.3% to 95.4%. At 4 h, the 400˚C and 500˚C films reached 53.1% and 51.8%, respectively; their 1 h values were 35.0% and 32.6%. Thus, the four-hour order remained 300˚C > 200˚C > 400˚C > 500˚C.

Figure 7. MB degradation under the reported visible-light condition for TiO₂ films treated at different temperatures.

For the 300˚C film, the measured crystallite size and optical gap were 58.4 nm ± 12.0 nm and 3.37 eV, respectively. The corresponding 200˚C values were 47.9 nm ± 19.9 nm and 3.42 eV, while the 400˚C specimen showed 7.8 nm ± 1.0 nm and 3.55 eV. The 500˚C coating, for which no crystalline domain size could be obtained from XRD, had the widest measured optical gap at 3.90 eV. The weaker visible-light response at 400˚C and 500˚C therefore coincided with both diminished crystalline development and a wider optical gap.

A common trend emerges when the photocatalytic measurements are considered with the structural data: the 300˚C film was the most active under both irradiation conditions and also possessed the largest calculated crystallite size, 58.4 nm ± 12.0 nm. Its diffraction indicated the strongest crystalline development among the tested films. Since crystallinity and crystallite dimensions have previously been linked with TiO₂ photocatalysis [29]-[31], these observations provide a useful framework for discussing the temperature dependence measured in this work.

Earlier TiO₂ studies have demonstrated that photocatalytic behaviour can vary with crystal development. Tanaka et al. [29] reported a dependence of photocatalytic action on TiO₂ crystallinity, while Ohtani et al. [30] examined photocatalysis in amorphous-anatase systems. Armaković et al. [31] also found temperature-dependent photocatalytic performance in TiO₂-based materials. These reports support considering crystallinity as one contributor to the present activity trend, although the simultaneous changes in optical and surface properties mean that the individual contribution of crystallite size cannot be isolated from the current dataset.

The band-gap results provide another correlation with photocatalytic performance. The 300˚C film showed the smallest measured gap among the crystalline samples, 3.37 eV, and simultaneously delivered 67.7% UV-C and 97.8% visible-light MB removal after 4 h. In principle, a smaller optical gap can lower the photon-energy threshold associated with electronic excitation under suitable illumination. Nevertheless, this study did not independently determine carrier generation, recombination kinetics, or wavelength-specific absorption. The observed association should therefore be treated as evidence of correlation rather than proof that the band gap controls the degradation rate.

The conventional interpretation of TiO₂ photocatalysis involves photoexcitation followed by formation of electron-hole pairs. As described by Wang et al. [32], photogenerated electrons can participate in oxygen-reduction reactions that produce superoxide species, whereas photogenerated holes can drive oxidation processes involving water or surface hydroxyl groups. These species can subsequently contribute to organic-dye degradation. For the present films, however, no direct measurements of radical concentration, carrier lifetime, or interfacial charge-transfer kinetics were performed. The pathway shown in Figure 8 should therefore be regarded as a literature-supported mechanistic interpretation of the MB-removal results, not as direct experimental proof of each elementary step.

Figure 8. Schematic interpretation of the proposed photocatalytic pathway for the brookite TiO₂ film.