In-Situ Functionalization of Cotton Fabric by TiO 2 : The Inﬂuence of Application Routes

: The desirable chemical, physical, electronic, and optical properties of TiO 2 , as well as its high availability, non-toxicity, and low price, make it very popular in the modern functional textile industry. Here, TiO 2 from titanium tetraisopropoxide (TTIP) precursors at concentrations of 2, 4, and 6% and commercial TiO 2 nanoparticles (NPs) in dispersion form were applied to cotton textiles using low-temperature application methods (i

In the functionalization of textiles, TiO 2 synthesis can be carried out using various precursors, the most common of which include titanium tetraisopropoxide [15,[18][19][20][21][22][23], titanium butoxide [24][25][26], titanium chloride [27][28][29], titanium oxysulfate hydrate [30], ammonium hexafluorotitanate [31], potassium titanium oxalate [32], and tetrabutyl titanate [33]. Both ex-situ and in-situ methods can be used for TiO 2 synthesis; however, in-situ synthesis has several advantages. For example, during in-situ synthesis, textile fibers in the TiO 2 precursor solution act as stabilizers that prevent the agglomeration of particles, resulting in the uniform distribution of smaller particles. Various synthesis routes have been used; among them, the in-situ sol-gel application technique is frequently utilized for the functionalization of textiles with TiO 2 [19,22,24,29]. In addition, in-situ hydrothermal synthesis is also commonly used to produce TiO 2 and results in a coating with optimal thickness and the high adhesion of TiO 2 to the textile surface [18,20,25,27,32,33]. In addition to surface deposition, TiO 2 absorption inside the fibers has also been reported [23]. Moreover, the phase composition, TiO 2 particle size, and crystalline homogeneity of the coated cotton fabric can be more precisely controlled with the hydrothermal process. During the hydrothermal process, the nucleation and in-situ growth of TiO 2 has been attributed to the strong attraction between the hydroxyl group of cotton and titanium ions, resulting in less particle agglomeration [33]. Aside from TiO 2 precursors, commercial TiO 2 products in the form of powders or dispersions have also been used for the functionalization of textiles [34][35][36][37]. Commercial TiO 2 products have been shown to have some challenges with respect to preparing a stable dispersion and achieving the uniform and thin deposition of TiO 2 particles on the fiber surface, as the particles tend to form agglomerates.
The most commonly used technique for obtaining functional textiles involves the immersion of the fabric for a certain time followed by drying, while the pad-dry-cure method has also been investigated due to its simplicity and the ease of particle uptake on the surface [21,22,24,35]. Since TiO 2 requires longer contact times due to its rather poor adhesion to the fibers, the exhaustion method is another suitable option for application, despite some disadvantages compared to the pad-dry-cure method, i.e., higher energy consumption, the discontinuity of the process, and the long reaction time of the chemicals with the textile substrate [19,34,35].
Undoubtedly, much research has been conducted on TiO 2 -functionalized textiles over the last decade. However, to date, this research has mainly focused on improving the adhesion of TiO 2 to the textile substrate and enhancing its photocatalytic activity toward visible light through various strategies and approaches, while little attention has been paid to the contribution of TiO 2 application routes to the functional properties. Indeed, the morphology, distribution, crystallinity, and Ti concentration on the coated fibers can greatly influence the UV-protective and photocatalytic performance of the functionalized textiles. To the best of our knowledge there is only one research project that investigated the three different process routes, focusing on the influence of precursor concentration and duration of the pad-dry-hydrothermal route on UV protection and self-cleaning properties [38]. However, studied application routes included some limitation of their feasibility on industrial level, i.e., sonication of the samples in the studied sols, which requires the use of specific application equipment and prolonged hydrothermal treatment, which is energy and time consuming. Focusing on the establishment of an in-situ TiO 2 application route that would be easily feasible in industry, in this work, the influence of titanium isopropoxide concentration applied via the most common low-temperature application methods, i.e., sol-gel pad-dry-cure, pad-hydrothermal, and exhaustion-hydrothermal, using three different concentrations of the TiO 2 precursor, was systematically investigated considering the morphological, chemical, and functional properties of the functionalized cotton fabric in relation to the crystalline and optical properties of the resulting TiO 2 . For comparison, a commercial TiO 2 product applied via a simple pad-dry-cure method was also included.

Morphological and Chemical Properties
The influence of the studied TiO 2 nanosols and the methods of their application on the morphological properties of the cotton fibers was studied using SEM and compared to the fibers obtained using commercial TiO 2 ( Figure 1). While the UN_CO sample exhibited a pristine cotton fabric with a relatively smooth surface, sample NP, which was treated with commercial TiO 2 , showed a thick film formation with an abundance of agglomerates on the fiber surface. In contrast, the sample treated with studied TiO 2 nanosols exhibited a thinner TiO 2 deposition. A detailed examination of the corresponding SEM images showed that the TTIP concentration and application method strongly influenced TiO 2 uptake and distribution on the fiber surface. Accordingly, the sol-gel pad-dry-cure application route was associated with the formation of a continuous film, while the formation of TiO 2 particles was detected on the surface of the fibers of the 2H and 4H samples treated according to the pad-hydrothermal route. In contrast, TiO 2 could hardly be perceived on the surface of the xE samples, suggesting that the exhaustion-hydrothermal application route either did not ensure the sufficient uptake of TiO 2 on the cotton fibers or that TiO 2 penetrated the inner part of the fibers rather than being deposited on their surface.
Catalysts 2022, 12, x FOR PEER REVIEW to the fibers obtained using commercial TiO2 (Figure 1). While the UN_CO sample e ited a pristine cotton fabric with a relatively smooth surface, sample NP, which treated with commercial TiO2, showed a thick film formation with an abundance glomerates on the fiber surface. In contrast, the sample treated with studied TiO2 nan exhibited a thinner TiO2 deposition. A detailed examination of the corresponding images showed that the TTIP concentration and application method strongly influ TiO2 uptake and distribution on the fiber surface. Accordingly, the sol-gel pad-dry application route was associated with the formation of a continuous film, while th mation of TiO2 particles was detected on the surface of the fibers of the 2H and 4H sam treated according to the pad-hydrothermal route. In contrast, TiO2 could hardly b ceived on the surface of the xE samples, suggesting that the exhaustion-hydroth application route either did not ensure the sufficient uptake of TiO2 on the cotton fib that TiO2 penetrated the inner part of the fibers rather than being deposited on thei face. Figure 1. SEM images of the untreated cotton sample (UN_CO), the sample treated with a com cial TiO2 product (NP), and samples treated via in-situ synthesis routes, i.e., sol-gel pad-dry (xS), pad-hydrothermal (xH), and exhaustion hydrothermal (xE), using different concentrati titanium isopropoxide (x = 2, 4, and 6%) in the nanosol. The results of EDS, shown in Figure 2, confirmed the presence of TiO 2 , as several peaks for titanium (Ti) were detected in the EDS spectra of the representative samples 4S, 4H, 4E, and NP. The comparatively low peak intensity of the Ti particles was caused by low TiO 2 concentration and was thus blurred by the signal of the fabric surface, which was also observed in another study [22]. On the other hand, the high peak intensity of O, which originated from the association of TiO 2 nanoparticles, was due to the inherent chemical structure of cellulose; thus, its concentration was comparatively higher on the TiO 2 -treated samples. EDS elemental mapping also showed that unlike commercial TiO 2 , which ensures a homogeneous but thick deposition of TiO 2 , the sol-gel pad-dry-cure and pad-hydrothermal application methods resulted in TiO 2 deposition that was mainly concentrated on the perimeter of the fibers (sample 4S and 4H), while on the surface of the fibers of sample 4E, the presence of TiO 2 was barely detectable in the EDS analysis.
The results of EDS, shown in Figure 2, confirmed the presence of TiO2, as several peaks for titanium (Ti) were detected in the EDS spectra of the representative samples 4S, 4H, 4E, and NP. The comparatively low peak intensity of the Ti particles was caused by low TiO2 concentration and was thus blurred by the signal of the fabric surface, which was also observed in another study [22]. On the other hand, the high peak intensity of O, which originated from the association of TiO2 nanoparticles, was due to the inherent chemical structure of cellulose; thus, its concentration was comparatively higher on the TiO2-treated samples. EDS elemental mapping also showed that unlike commercial TiO2, which ensures a homogeneous but thick deposition of TiO2, the sol-gel pad-dry-cure and padhydrothermal application methods resulted in TiO2 deposition that was mainly concentrated on the perimeter of the fibers (sample 4S and 4H), while on the surface of the fibers of sample 4E, the presence of TiO2 was barely detectable in the EDS analysis.  The presence of TiO 2 was further verified by ICP MS analysis which revealed that, of the application routes studied, sample 4E had the highest Ti content, i.e., 7400 mg per kg of fabric, followed by samples 4S and 4H with 6600 and 5100 mg Ti per kg of fabric, respectively. Sample NP contained the highest amount of TiO 2 , i.e., 15,000 mg Ti per kg of fabric, which is consistent with the SEM images. The high Ti content of sample 4E undoubtedly confirms that exhaustion-hydrothermal application route induces the absorption of TiO 2 inside the fibers, since the presence of TiO 2 could hardly be detected with EDS, while small particles must have been formed. The lower Ti content of sample 4H compared to sample 4S can be explained by the partial removal of TiO 2 during the hydrothermal treatment, as the boiling water became slightly white and turbid after the sample was immersed.
The chemical properties of the untreated and treated samples were studied using FTIR analysis (Figure 3). The IR ATR spectrum of sample UN_CO showed the absorption bands of pure cellulose, i.e., absorption bands in the 3700-3000 cm −1 spectral region ascribed to the -OH vibration of adsorbed water, an absorption band at 2900 cm −1 associated with the CH 2 stretching vibrations, and absorption bands in the 1500-800 cm −1 cellulose fingerprint region, indicating C-O-C, C-H, C-O, and O-H vibrations [39]. From the literature, TiO 2 -related vibrations appear at 700-600 cm −1 , 525-460 cm −1 , 489 cm −1 , and 360-320 cm −1 [26,39]. The IR ATR spectra of the representative samples 4S, 4H, and 4E were not influenced by the TiO 2 coating obtained using either the sol-gel curing, hydrothermal, or exhaustion method, as the direct comparison of these spectra with the spectrum of the UN_CO sample did not reveal any significant changes. The reason for this could be small concentration of TiO 2 , which was blurred by the strong absorption bands of the cellulose. To the contrary, in the IR ATR spectrum of the NP sample, a new absorption band appeared at 430 cm −1 , ascribed to Ti-O-Ti stretching. Its occurrence was related to the thick deposition of the TiO 2 coating and the highest Ti concentration among the samples studied.
The presence of TiO2 was further verified by ICP MS analysis which revealed that, of the application routes studied, sample 4E had the highest Ti content, i.e., 7400 mg per kg of fabric, followed by samples 4S and 4H with 6600 and 5100 mg Ti per kg of fabric, respectively. Sample NP contained the highest amount of TiO2, i.e., 15,000 mg Ti per kg of fabric, which is consistent with the SEM images. The high Ti content of sample 4E undoubtedly confirms that exhaustion-hydrothermal application route induces the absorption of TiO2 inside the fibers, since the presence of TiO2 could hardly be detected with EDS, while small particles must have been formed. The lower Ti content of sample 4H compared to sample 4S can be explained by the partial removal of TiO2 during the hydrothermal treatment, as the boiling water became slightly white and turbid after the sample was immersed.
The chemical properties of the untreated and treated samples were studied using FTIR analysis ( Figure 3). The IR ATR spectrum of sample UN_CO showed the absorption bands of pure cellulose, i.e., absorption bands in the 3700-3000 cm −1 spectral region ascribed to the -OH vibration of adsorbed water, an absorption band at 2900 cm −1 associated with the CH2 stretching vibrations, and absorption bands in the 1500-800 cm −1 cellulose fingerprint region, indicating C-O-C, C-H, C-O, and O-H vibrations [39]. From the literature, TiO2-related vibrations appear at 700-600 cm −1 , 525-460 cm −1 , 489 cm −1 , and 360-320 cm −1 [26,39]. The IR ATR spectra of the representative samples 4S, 4H, and 4E were not influenced by the TiO2 coating obtained using either the sol-gel curing, hydrothermal, or exhaustion method, as the direct comparison of these spectra with the spectrum of the UN_CO sample did not reveal any significant changes. The reason for this could be small concentration of TiO2, which was blurred by the strong absorption bands of the cellulose. To the contrary, in the IR ATR spectrum of the NP sample, a new absorption band appeared at 430 cm −1 , ascribed to Ti-O-Ti stretching. Its occurrence was related to the thick deposition of the TiO2 coating and the highest Ti concentration among the samples studied.   [25,40]. No differences were observed between the XRD spectrum of UN_CO and that of samples 4S, 4H, and 4E, suggesting that TiO 2 on the surface of these samples existed in the amorphous form. These results were further supported by the analysis of the Raman spectra ( Figure 4b). Again, while the anatase-phase TiO 2 on the surface of the NP sample was confirmed by the absorption bands at 360, 397, and 518 cm −1 [29,30], the latter was not observed for samples 4S, 4H, and 4E. Accordingly, it was concluded that the commercial TiO 2 -treated sample (NP) exhibited a crystalline anatase structure, while the synthesis procedures applied for samples 4S, 4H, and 4E resulted in the formation of amorphous TiO 2 . When conducting a pad-dry-cure application route, the formation of amorphous TiO 2 was also confirmed by other researchers [38]. In contrast, the absence of the characteristic anatase TiO 2 peaks in the case of sample 4H was a surprise, since in a previous study [23], a few minutes of hydrothermal treatment caused the formation of crystalline TiO 2 . However, in this case, the XRD analysis was performed on a TiO 2 powder synthesized in the same way as for the TiO 2 in-situ functionalization of fabric. In many studies, the formation of anatase TiO 2 by hydrothermal treatment was confirmed, but a prolonged hydrothermal treatment of at least 1.5 h was required to achieve high crystallinity of TiO 2 deposited on the fabric [18,33,38].
the Raman spectra ( Figure 4b). Again, while the anatase-phase TiO2 on the surface of the NP sample was confirmed by the absorption bands at 360, 397, and 518 cm −1 [29,30], the latter was not observed for samples 4S, 4H, and 4E. Accordingly, it was concluded that the commercial TiO2-treated sample (NP) exhibited a crystalline anatase structure, while the synthesis procedures applied for samples 4S, 4H, and 4E resulted in the formation of amorphous TiO2. When conducting a pad-dry-cure application route, the formation of amorphous TiO2 was also confirmed by other researchers [38]. In contrast, the absence of the characteristic anatase TiO2 peaks in the case of sample 4H was a surprise, since in a previous study [23], a few minutes of hydrothermal treatment caused the formation of crystalline TiO2. However, in this case, the XRD analysis was performed on a TiO2 powder synthesized in the same way as for the TiO2 in-situ functionalization of fabric. In many studies, the formation of anatase TiO2 by hydrothermal treatment was confirmed, but a prolonged hydrothermal treatment of at least 1.5 h was required to achieve high crystallinity of TiO2 deposited on the fabric [18,33,38].

Optical Properties
The optical properties were studied by analyzing the absorption spectra of the studied samples-NP, 4S, 4H, and 4E-in the spectral range of 200-400 nm ( Figure 5). All the studied samples strongly absorbed light in the UV spectral region and showed a broad absorption band with a sharp absorption edge (λg) that varied depending on the TiO2 application route. While the NP sample showed a λg value at 390 nm, samples 4S, 4H, and 4E were blue-shifted to 381, 365, and 376 nm, respectively.

Optical Properties
The optical properties were studied by analyzing the absorption spectra of the studied samples-NP, 4S, 4H, and 4E-in the spectral range of 200-400 nm ( Figure 5). All the studied samples strongly absorbed light in the UV spectral region and showed a broad absorption band with a sharp absorption edge (λ g ) that varied depending on the TiO 2 application route. While the NP sample showed a λ g value at 390 nm, samples 4S, 4H, and 4E were blue-shifted to 381, 365, and 376 nm, respectively.
Based on the absorbance values, the energy band gap (E g ) of TiO 2 was then determined. According to the literature, E g can be determined through two methods, either by using the absorption method [41] or by Tauc plotting [42]. In the absorption method, E g can be acquired using the following relationship: where λ g is determined from the intercept between the tangent of the absorption curve and the abscissa coordinate. Using Tauc plotting, E g can be calculated according to the Tauc and Davis-Mott relationship: where B is a constant, α is the absorption coefficient, and hν is the energy of incident photons. The value of n depends on the nature of the transition, where n = 1 2 and n = 2 are used for indirect allowed and direct allowed transitions, respectively. E g can be determined by plotting (αhν) n against (hν), where the linear portion of the curve is used to extrapolate the value of the E g . Based on the absorbance values, the energy band gap (Eg) of TiO2 was then determined. According to the literature, Eg can be determined through two methods, either by using the absorption method [41] or by Tauc plotting [42]. In the absorption method, Eg can be acquired using the following relationship: , where λg is determined from the intercept between the tangent of the absorption curve and the abscissa coordinate. Using Tauc plotting, Eg can be calculated according to the Tauc and Davis-Mott relationship: where B is a constant, α is the absorption coefficient, and hν is the energy of incident photons. The value of n depends on the nature of the transition, where n = ½ and n = 2 are used for indirect allowed and direct allowed transitions, respectively. Eg can be determined by plotting (αhν) n against (hν), where the linear portion of the curve is used to extrapolate the value of the Eg.
In the present study, both methods of Eg determination were used; for Tauc plotting, the absorption data were fitted for both indirect (n = ½) and direct (n = 2) transitions. As can be seen from the results in Figure 5 and in Table 1, using the absorption edge method, the commercial TiO2 showed a perfect fit for Eg of 3.2, which corresponds to the band gap of bulk TiO2 with an anatase structure [43]. Due to the blue shift of λg, the synthesized In the present study, both methods of E g determination were used; for Tauc plotting, the absorption data were fitted for both indirect (n = 1 2 ) and direct (n = 2) transitions. As can be seen from the results in Figure 5 and in Table 1, using the absorption edge method, the commercial TiO 2 showed a perfect fit for E g of 3.2, which corresponds to the band gap of bulk TiO 2 with an anatase structure [43]. Due to the blue shift of λ g , the synthesized TiO 2 samples showed an increasingly wider E g in the following order: 4S < 4E < 4H. Indirect fitting using a Tauc plot yielded E g values of 2.91-3.07 eV. Indeed, the E g of the commercial sample (NP) was comparable to that of P25 TiO 2 determined by Yu et al. [44]. The lower E g value of 2.95 eV for P25, compared to 3.2 eV for the bulk anatase form, can be explained by the presence of very fine particles with about 25% rutile content in P25. In our case, the narrowest E g of 2.91 eV was determined for sample NP, followed by samples 4E, 4S, and 4H. However, given that the E g values of samples 4S and 4E were determined to be approximately the same while no crystal structure was detected by XRD, these values are too low and do not seem realistic. Better agreement with the E g values determined by the absorption edge method was obtained with Tauc plotting using direct band gap fitting, which is also consistent with previous reports [45][46][47]. In this case, a slightly higher E g value of 3.28 eV was obtained for sample NP, while sample 4S again had the most comparable E g to that of commercial TiO 2 (i.e., 3.45 eV). Samples 4H and 4E exhibited wider E g values of 3.55 eV and 3.51 eV, respectively. The increase in the E g of the synthesized TiO 2 suggests differences in the structures of the synthesized TiO 2 [48]. While differences in E g values between the NP sample and those of synthesized TiO 2 are most likely due to anatase vs. amorphous TiO 2 , the differences in E g between samples treated with synthesized TiO 2 imply a certain common structure of TiO 2 formed by the pad-hydrothermal and exhaustion-hydrothermal methods, as 4H and 4E demonstrated an almost identical E g .

UV Protection
To evaluate the UV protection properties of the untreated and treated cotton textiles, the ultraviolet protection factor (UPF) values were determined based on the transmittance measurements ( Figure 6a) and are presented in Table 2. The untreated cotton sample UN_CO had a UPF of only 4.2. Despite its high reflectance (Figure 6b), due to the bleaching process of the raw cotton, and associated high light emission, a high UV transmittance of 27.7% in the UVA range and 22.0% in the UVB range makes it unsuitable for protection against solar UV light. TiO 2 absorbs UV light and thus increased the UPF value of the studied samples. Accordingly, the UPF of the treated samples increased and provided "good" to "very good" UV protection. For the xS and xH samples, a TTIP concentration of 4% in the sol provided the highest UV protection, mainly due to the strong reduction in light transmission and reflection, especially in the UVB range. Accordingly, a high UVB-blocking ability of 97.7% was observed for the 4S sample and 98.3% for the 4H sample, corresponding to UPFs of 29.6 and 30.3, respectively, thus providing "very good" UV protection. Interestingly, further increasing the TTIP concentration in the sol to 6% (samples 6S and 6H) did not further increase the UPF but resulted in a slight deterioration of the UV protection properties. For the xE samples, the UPF gradually increased with increasing TTIP concentration in the sol. However, even at the highest TTIP concentration (6%) in the sol, a UPF of 24.4 was obtained for sample 6E, providing "good" UV protection despite the highest Ti content compared to samples xS and xH. With a UPF value of 55.8, sample NP provided "excellent" UV protection, with a sharp decrease in light transmission from UVA and UVB (Figure 6a). Accordingly, 91.2% UVA-and 98.6% UVB-blocking were determined for this sample. This is a reflection of the crystalline anatase structure of commercial TiO 2 , which absorbs UV radiation more actively [49] than synthesized amorphous TiO 2 . Nevertheless, compared to samples 4S, 4H, and 4E, sample NP had 2.3, 2.9 and 2.0 higher Ti content, respectively, and thus a much thicker deposition, which also contributed to its excellent UV protective effect. Nonetheless, "good" to "very good" UV protection provided by the xS and xH samples functionalized according to the sol-gel pad-dry-cure and pad-hydrothermal application routes is in good agreement with results obtained by others [22,23,32,38]. Therefore, despite the low deposition and amorphous structure of TiO 2 , synthesized TiO 2 provided satisfactory UV protection, which exceeded our expectations considering that the TiO 2 synthesis was performed at a low temperature compared to the energy-consuming calcination process of commercial TiO 2 at high temperatures, which assures the formation of anatase TiO 2 with a high UV-absorbing capacity

Self-Cleaning
The photocatalytic self-cleaning efficiency of the studied cotton samples was investigated based on the degradation of MB dye under illumination at different time intervals

Self-Cleaning
The photocatalytic self-cleaning efficiency of the studied cotton samples was investigated based on the degradation of MB dye under illumination at different time intervals of 15, 30, 45, 60, and 90 min. The results are shown in Figure 7. In general, the degradation of MB on the surface of the studied samples increased with increasing illumination time. This phenomenon was also observed in the CO_UN sample but could be attributed to the lower light-fastness of the MB dye, since all samples treated with TiO 2 showed a much higher intensity of degradation of MB, regardless of the application method and illumination time. Indeed, TiO 2 is a strong oxidant for organic contaminants, and reactive oxygen species (ROS, i.e., superoxide, excited oxygen, and hydroxyl radicals) can be formed on the surface of TiO 2 particles in the presence of UV light. The generated ROS trigger the oxidation of the MB dye via various intermediates to the final products CO 2 , H 2 O, SO 4 2− , NH 4 + , and NO 3 − , which causes the self-cleaning of the stained surface. As can be seen from the results in Figure 7a, the TTIP concentration and application route affected the generation of ROS and thus self-cleaning activity. The examination of the results after the initial time interval for visible light exposure of 15 min revealed the highest self-cleaning activity for sample NP followed by samples 4S, 4H, and 2E, which degraded 52.3, 51.0, 44.3, and 39.7% of the MB dye, respectively. This agrees well with the results of the band gap energy determination, where the narrowest values, indicating higher photocatalytic activity, were determined for samples NP and 4S. The decreased degradation of the MB dye on the surface of sample xE at the beginning of visible illumination again implied the delayed photocatalytic degradation of the dye as ROS needed to pass from the inner parts to the surface of the fibers. At the end of the illumination period (Figure 7b), samples 2E and 4S showed the highest photocatalytic degradation of the MB dye, at 80.0 and 79.9%, respectively. Such a high photocatalytic activity of sample 2E, despite two times lower concentration of the TTIP precursor than that for sample 4S, can confirm the formation of very small sized TiO 2 particles during exhaustion-hydrothermal application route. Namely, it is well known that the photocatalytic activity of TiO 2 particles is size-dependent, whereas smaller TiO 2 particles reflect greater ROS formation due to a high specific surface area. As for sample 4S, a TiO 2 film was formed on the surface of the fibers with a large surface area but a lower surface-to-volume ratio, and thus a higher Ti concentration was required to achieve comparable photocatalytic activity. Due to the highest Ti content and the crystalline anatase form of commercial TiO 2 , it was expected that sample NP would demonstrate the highest photocatalytic self-cleaning activity among the samples studied. However, this was not the case, as a comparable MB dye degradation activity, i.e., 77.1%, was obtained at the end of the experiment, which was due to the formation of large agglomerates on the surface of the sample NP, as confirmed by SEM. Accordingly, due to the amorphous structure of the synthesized TiO 2 , the formation of ROS was most likely delayed, which would explain the lower self-cleaning activity of samples xS, xH, and xE at the beginning of illumination. With increasing illumination time, the levels of ROS continued to increase and finally exceeded the photocatalytic activity of the commercial TiO 2 . Namely, it has been suggested that amorphous TiO 2 is an excellent mediator for electron transfer from excited dyes to oxygen, which is attributed to the enhanced light harvesting and high specific surface area, due to a certain disordered structure and isotropic physical and chemical properties [50]. Accordingly, higher visible light induced photosensitized dye degradation of amorphous TiO 2 compared to crystallized TiO 2 was also obtained by others [51]. Undoubtedly, the obtained results confirm that a simple low-temperature synthesis can result in the formation of TiO 2 with better photocatalytic self-cleaning activity than commercial TiO 2 , despite the amorphous nature of TiO 2 and two-to three-times lower Ti content. Among the application routes studied, the simple sol-gel pad-dry-cure and exhaustion methods using TTIP concentrations of 4 or 2% proved to be the best.

Materials
Woven 100% cotton fabric was purchased from Tekstina tekstilna industrija d.o.o. (Ajdovščina, Slovenia). The warp thread, weft thread, and fabric density were 51 threads/cm, 31 threads/cm, and 120 g/m 2 , respectively. Titanium isopropoxide (concentration ≥ 97.0%) with a molecular weight of 284.22 and density of 0.96 g/mL (at 20 °C) was purchased from Sigma Aldrich (St. Louis, MO, USA. Commercial TiO2 NPs in dispersion form with a crystal size of 5 nm and density of 1.2 g/cm 3 were purchased from Cinkarna metalurško kemična industrija Celje, dd (Celje, Slovenia). The pH of the dispersion solution was maintained at 7-9. Ethanol (concentration ≥ 99.8 %) and glacial acetic acid were supplied from Honeywell research chemical (Seetze, Germany) and CARLO ERBA reagents S.A.S (Barcelona, Spain), respectively. All chemicals were used without further modification.

Preparation of TiO2 Nanosols and Their Application to Cotton Fabric
TiO2 nanosols were prepared by mixing TTIP in ethanol to obtain concentrations of 2, 4, and 6%. Immediately, the corresponding amount of glacial acetic acid was added dropwise into the TiO2 nanosols under constant stirring, where the mass ratio between

Materials
Woven 100% cotton fabric was purchased from Tekstina tekstilna industrija d.o.o. (Ajdovščina, Slovenia). The warp thread, weft thread, and fabric density were 51 threads/cm, 31 threads/cm, and 120 g/m 2 , respectively. Titanium isopropoxide (concentration ≥ 97.0%) with a molecular weight of 284.22 and density of 0.96 g/mL (at 20 • C) was purchased from Sigma Aldrich (St. Louis, MO, USA. Commercial TiO 2 NPs in dispersion form with a crystal size of 5 nm and density of 1.2 g/cm 3 were purchased from Cinkarna metalurško kemična industrija Celje, dd (Celje, Slovenia). The pH of the dispersion solution was maintained at 7-9. Ethanol (concentration ≥ 99.8 %) and glacial acetic acid were supplied from Honeywell research chemical (Seetze, Germany) and CARLO ERBA reagents S.A.S (Barcelona, Spain), respectively. All chemicals were used without further modification.

Preparation of TiO 2 Nanosols and Their Application to Cotton Fabric
TiO 2 nanosols were prepared by mixing TTIP in ethanol to obtain concentrations of 2, 4, and 6%. Immediately, the corresponding amount of glacial acetic acid was added dropwise into the TiO 2 nanosols under constant stirring, where the mass ratio between TTIP and glacial acid was set at 7:1. Afterwards, the TiO 2 nanosols were vigorously stirred for 120 min.
The TiO 2 nanosols were applied to the cotton fabric using different application routes, i.e., sol-gel pad-dry-cure, pad-hydrothermal, and exhaustion-hydrothermal methods. In the sol-gel pad-dry-cure and pad-hydrothermal routes, impregnation was first performed by immersing the cotton samples in the corresponding TiO 2 nanosols for 3 min, followed padding on a two-roller padder (Mathis, Switzerland), with a pressure of 0.8 bars to obtain 80 ± 2% wet pick-up. For the pad-dry-cure application route, the samples were dried at 100 • C for 1 min and cured at 140 • C for 5 min in the laboratory dryer (Mathis, Switzerland). These samples were named 2S, 4S, and 6S, corresponding to the concentration of TIIP in the TiO 2 nanosol. For the pad-dry-hydrothermal application route, the impregnated samples were firstly dried in the same conditions previously mentioned followed by immersion in boiling water for 30 min using a liquid-to-good ratio of 1:50. Afterwards, the samples were rinsed with deionized water and left to air dry. These samples were named 2H, 4H, and 6H. For the exhaustion-hydrothermal application route, the samples were immersed in the corresponding TiO 2 nanosols for 1 h under constant stirring in a Gyrowash (James Heal, Great Britain). The liquid-to-good ratio was set as 1:50. Afterwards, the samples were squeezed on a two-roller padder using the same settings described above; thus, the same wet pick-up of 80 ± 2% was achieved. Finally, the samples were dried and hydrothermally treated, similar to the pad-hydrothermal application. These samples were named 2E, 4E, and 6E.
For the purpose of comparison, commercial TiO 2 was also applied to the cotton fabric using the pad-dry-cure method. A commercial TiO 2 NP dispersion was diluted to 2.5% with double-distilled water and stirred vigorously for 2h, followed by the impregnation, padding, drying, and curing of the samples in the same conditions as previously described. The sample was designated as NP. The details of the application process, along with the sample codes, are presented in Table 3. The schematic presentation of the studied in-situ low-temperature application routes is shown in Figure 8. The surface morphology of the functionalized cotton samples was studied by a scanning electron microscope SEM-6060 LV (JEOL, Tokyo, Japan) operating at a 10 kV accel- Figure 8. Schematic presentation of the studied in-situ low-temperature application routes. The surface morphology of the functionalized cotton samples was studied by a scanning electron microscope SEM-6060 LV (JEOL, Tokyo, Japan) operating at a 10 kV acceleration voltage. A thin Au/Pt coating layer was applied for the sufficient electrical conductivity during measurement procedures.

Energy Dispersive Field Emission Scanning Electron Microscopy (EDS)
To investigate the chemical properties of the studied cotton samples, a Thermo Fisher Scientific Quattro S (Thermo Fisher Scientific, Waltham, MA, USA) energy dispersive field emission scanning electron microscope (FEG-SEM) working at an accelerating voltage of 1 kV in high vacuum using a concentric backscattered electron detector (CBS) was used. An energy dispersive X-ray detector (EDS; Oxford Instruments, Santa Barbara, CA, USA) was used to analyze the chemical composition of the observed particles using Aztec software. EDS spectra and elemental mappings of O and Ti were obtained. Prior to analysis, a thin conductive carbon layer was applied to the cotton samples.

Inductively Coupled Plasma-Mass Spectroscopy (ICP MS)
The total Ti content in the studied fabric samples was determined by inductively coupled plasma-mass spectroscopy (ICP MS) using a PerkinElmer SCIED Elan DRC spectrophotometer. A sample of 0.5 g was prepared in a Milestone microwave system by acid decomposition with 65% HNO 3 and 30% H 2 O 2 . Titanium concentrations values are given as the mean of two measurements made for each sample.

X-ray Diffraction (XRD)
The XRD characterization of the untreated cotton sample and Ag-containing functionalized cotton samples was performed using a PANalytical X'Pert PRO X-ray diffractometer (XRD, Malvern PANalytical Ltd., Almelo, The Netherlands) (CuK∝1 = 1.5406 Å) with a fully open X'Celerator detector (2.122 • 2θ). The XRD pattern was measured from 10 to 70 • 2θ with a step size of 0.034 • 2θ and an integration time of 100 s.

Raman Spectroscopy
The Raman spectra of TiO 2 functionalized cotton textile were recorded using a WITec alpha300 RA spectrometer (Ulm, Germany) equipped with a microscope. A laser of 532 nm excitation wavelength, power of 12 mW, and the 50 × objective was considered for the spectral region of 3600-200 cm −1 , and the final data were normalized.

Fourier Transform-Infrared (FT-IR) Spectroscopy
The infrared spectra of the studied cotton samples was performed with a Vertex 70V (Bruker, Bremen, Germany), integrated with an attenuated total reflection (ATR) accessory (Bruker, Germany) with a diamond crystal (n = 2.0). The spectra were recorded over the range of 4000-300 cm −1 with a resolution of 4 cm −1 and an average set of 128 spectra per sample.

UV Protection Property
The UV protective properties of the studied TiO 2 treated cotton fabric samples were determined according to the according to the AATCC TM 183 standard by measuring the transmission and reflection properties of the samples using a Lambda 800 UV/Vis spectrophotometer (Perkin Elmer, Buckinghamshire, UK). Three measurements of transmittance, T, and reflectance, R, were obtained in the wavelength range of 200-800 nm at different angles of the sample warp alignment. For the determination of the UV protection factor (UPF), average transmittances were calculated at wavelengths of 315-400 nm (UV-A) and 280-315 (UV-B). The UPF was calculated using the following equation: where E λ is the relative erythermal spectral effectiveness, S λ is the solar spectral irradiance, T λ is the spectral transmittance of the specimen, and ∆ λ is the measured wavelength interval in nm. Based on the obtained UPF values, the UV protection categories were determined according to the Australian/New Zealand Standard Sun protective clothing Evaluation and Classification (AS/NZS, 2017). Accordingly, UPF values of 15-24 correspond to a protection category of "good"; UPF values of 25-39 correspond a protection category of "very good"; UPF values of 40-50 and 50 and above correspond to a protection category of "excellent".
3.3.8. Self-Cleaning Activity TiO 2 -treated cotton fabric samples were stained with 0.05% methylene blue (MB) (Sigma Aldrich, St. Louis, MO, USA) water solution and illuminated by a Xenotest Alpha instrument (Atlas, Mount Prospect, IL, USA) equipped with a visible xenon arc lamp (radiation attitude of 0.8-2.5 kVA and extended radiation range of 300-400 nm). The studied samples were illuminated for 30, 60, 90, 120, and 180 min. After illumination, the degradation of the dye was assessed based on the reflectance, R, of the studied samples, measured with a Datacolor Spectraflash SF 600 spectrophotometer using D 65/10 light. For each sample, ten measurements of the R-value were obtained, and the corresponding K/S values were calculated according to the Kubelka-Munk equation: where K/S is the ratio of the coefficient of light absorption (K) to the coefficient of light scattering (S) and R is the reflectance at the maximum absorption wavelength determined at 610 nm. Based on the K/S values, MB dye degradation, D, due to the photocatalytic activity of TiO 2 was determined using the following equation: where (K/S) 0 is the K/S value of the sample before illumination and (K/S) t is the K/S value of the sample after a certain illumination time.

Conclusions
The in-situ low-temperature application routes affected the morphology, optical, and functional properties of the cotton fibers, which were significantly different from those obtained with commercial TiO 2 . Important conclusions from the results are the following: -Commercial anatase TiO 2 with a narrow band gap formed a rather thick layer with a high add-on level and the presence of agglomerates on the surface of the cotton fibers, which granted excellent UV protection with a UPF of 50+ but resulted in a slight impairment of the self-cleaning activity. - The low-temperature in-situ application routes studied resulted in the formation of amorphous TiO 2 with blue-shifted bandgap energies and greatly reduced Ti content, but with different morphology. While the sol-gel pad-dry-cure method resulted in the formation of a TiO 2 film on the fiber surface, the pad-hydrothermal method induced a combination of TiO 2 particles and continuous film formation, while the TiO 2 particles were mainly absorbed inside the fibers in the exhaustion-hydrothermal method. -Very good UV protection was obtained with 4% TTIP sol applied via the sol-gel pad-dry-cure or pad-hydrothermal routes, while the samples treated with 2 and 4% TTIP sol applied via the exhaustion-hydrothermal and sol-gel pad-dry-cure methods, respectively, exhibited the highest self-cleaning activity during visible light illumination time, exceeding that of commercial TiO 2 .
All the investigated in-situ low-temperature application routes showed high practical potential with lower energy and chemical consumption, which brings them in line with sustainable development guidelines.