Photocatalytic Degradation of Di ﬀ erent VOCs in the Gas-Phase over TiO 2 Thin Films Prepared by Ultrasonic Spray Pyrolysis

: In this study, we deposited TiO 2 thin ﬁlms onto borosilicate glass by ultrasonic spray pyrolysis at 350 and 450 ◦ C. The aim of study is to determine the e ﬀ ect of deposition temperature on photocatalytic activity of TiO 2 thin ﬁlms and to investigate the performance of TiO 2 thin ﬁlms on photocatalytic degradation of methyl tert-butyl ether (MTBE), acetone, acetaldehyde, and heptane as functions of di ﬀ erent operating parameters. TiO 2 thin ﬁlms deposited at 350 and 450 ◦ C have a thickness value of 190 and 330 nm, respectively. All as-prepared TiO 2 ﬁlms possess an anatase crystalline structure. According to the X-ray photon spectroscopy (XPS) study, the TiO 2 thin ﬁlm deposited at 350 ◦ C showed a higher amount of oxygen vacancies and hydroxyl groups on the ﬁlm surface after UV treatment. The aged-TiO 2 thin ﬁlm deposited at 350 ◦ C showed a water contact angle (WCA) value of 0 ◦ after 10 min UV irradiation, showing superhydrophilic surface behavior. The TiO 2 ﬁlm deposited at 350 ◦ C exhibited the highest amount of conversion of MTBE (100%). The results also showed that TiO 2 ﬁlms are capable of photocatalytic degradation of MTBE (100%) and acetaldehyde (approx. 80%) in humid air conditions and high airﬂow rate. The visible-light-activity of TiO 2 thin ﬁlms was tested with 5 ppm MTBE and acetone. TiO 2 thin ﬁlms deposited at 350 ◦ C with a surface area of 600 cm 2 showed 60% of MTBE and 33% of acetone degradation under VIS light.


Introduction
Volatile organic compounds (VOCs) are the most common air pollutants, which are present in both indoor and outdoor air. It was reported that the VOC concentration indoors is often 2 to 5 times higher than that of the outdoors [1]. Among VOCs, acetone and acetaldehyde are common chemicals used extensively in a variety of industrial and domestic applications and are found in appreciable concentrations in indoor air [2,3]. Methyl tert-butyl ether (MTBE) and heptane are fuel components released to the atmosphere from gasoline and motor vehicle exhaust [4,5]. They are considered as outdoor generated VOCs, which also affect indoor air quality, especially in buildings close to parking lots or streets. Therefore, it is essential to obtain healthy living environments in modern buildings, integrated with air cleaning solutions [6]. According to the international standard, acetone and acetaldehyde together with heptane, toluene, and formaldehyde are the model air pollutants for the potential international standard testing of air cleaners [7]. Figure 1 shows the surface morphology of as-prepared TiO 2 thin films on borosilicate glass. TiO 2 film deposited at 350 • C (Figure 1a) showed a plane surface structure, whereas the film deposited at 450 • C (Figure 1c) showed a grain-like structure. The thickness of TiO 2 films sprayed at 350 and 450 • C were 190 and 330 nm (Table 1), respectively, as estimated from the cross-sectional scanning electronic microscope (SEM) images (inset in Figure 1a,c).

Surface Morphology
Catalysts 2019, 9, x FOR PEER REVIEW 3 of 18 2.1.1. Surface Morphology Figure 1 shows the surface morphology of as-prepared TiO2 thin films on borosilicate glass. TiO2 film deposited at 350 °C ( Figure 1a) showed a plane surface structure, whereas the film deposited at 450 °C ( Figure 1c) showed a grain-like structure. The thickness of TiO2 films sprayed at 350 and 450 °C were 190 and 330 nm (Table 1), respectively, as estimated from the cross-sectional scanning electronic microscope (SEM) images (inset in Figure 1a  TiO2 films deposited at 450 °C showed grains with a size of ca 50 nm and agglomerated grains with a size of ca 200 nm. Different surface topography of the films deposited at 350 to 450 °C resulted also in a slight increase of the root mean square (RMS) roughness from 0.6 to 1.0 nm, respectively. The SEM and AFM images revealed the formation of agglomeration as elliptical clusters for the TiO2 thin film deposited at 450 °C. The formation of agglomerated grains was observed in TiO2 thin films deposited by spray pyrolysis at high temperatures [36].  [37]. No other crystalline phases of TiO2 were detected. The Scherrer formula was applied to calculate the mean crystallite size of TiO2 films using the (101) peak of the anatase phase. Similar mean crystallite sizes of 26 and 22 nm were found for the TiO2 films deposited at 350 and 450 o C, respectively. This confirms that deposition temperature did not have any size of ca 200 nm. Different surface topography of the films deposited at 350 to 450 • C resulted also in a slight increase of the root mean square (RMS) roughness from 0.6 to 1.0 nm, respectively. The SEM and AFM images revealed the formation of agglomeration as elliptical clusters for the TiO 2 thin film deposited at 450 • C. The formation of agglomerated grains was observed in TiO 2 thin films deposited by spray pyrolysis at high temperatures [36].  [37]. No other crystalline phases of TiO 2 were detected. The Scherrer formula was applied to calculate the mean crystallite size of TiO 2 films using the (101) peak of the anatase phase. Similar mean crystallite sizes of 26 and 22 nm were found for the TiO 2 films deposited at 350 and 450 • C, respectively. This confirms that deposition temperature did not have any remarkable effect on the structural properties of sprayed TiO 2 thin films.  Figure 3 shows the total transmittance spectra of TiO2 thin films deposited at 350 and 450 °C and annealed at 500 °C for 1 h. Both TiO2 films showed total transmittance of ca. 80% in the visible spectral region. The indirect band gap of the TiO2 film was determined according to the Tauc plot [5]. The optical band gap of both samples was found to be 3.38 eV (inset of Figure 3).   Figure 3 shows the total transmittance spectra of TiO 2 thin films deposited at 350 and 450 • C and annealed at 500 • C for 1 h. Both TiO 2 films showed total transmittance of ca. 80% in the visible spectral region. The indirect band gap of the TiO 2 film was determined according to the Tauc plot [5]. The optical band gap of both samples was found to be 3.38 eV (inset of Figure 3).

XPS Study
XPS is a useful and sophisticated measurement technique for investigating the chemical constituents of a material, the ionic states of the constituent elements, and the ratio of the amounts of the different ionic states of a single constituent element. Estimation of oxygen vacancies and carbon species was confirmed from the XPS data in many studies [38][39][40][41][42] with the deconvoluted O1s and C1s XPS spectrum of the TiO 2 .
The surface chemical composition and bonding structure of the aged and UV-treated TiO 2 thin films were investigated with the analysis of XPS data. Figures 4 and 5 show the O1s and C1s core level spectra of the surface of aged-TiO 2 thin films deposited at 350 and 450 • C before and after UV-treatment, respectively. The asymmetric O1s and C1s core level peaks were deconvoluted by using Lorentzian-Gaussian (function pseudo-Voigt) fitting analysis. The background subtraction was applied by using Shirley type fitting [5]. Figure 3 shows the total transmittance spectra of TiO2 thin films deposited at 350 and 450 °C and annealed at 500 °C for 1 h. Both TiO2 films showed total transmittance of ca. 80% in the visible spectral region. The indirect band gap of the TiO2 film was determined according to the Tauc plot [5]. The optical band gap of both samples was found to be 3.38 eV (inset of Figure 3).

XPS Study
XPS is a useful and sophisticated measurement technique for investigating the chemical constituents of a material, the ionic states of the constituent elements, and the ratio of the amounts of the different ionic states of a single constituent element. Estimation of oxygen vacancies and carbon species was confirmed from the XPS data in many studies [38][39][40][41][42] with the deconvoluted O1s and C1s XPS spectrum of the TiO2.
The surface chemical composition and bonding structure of the aged and UV-treated TiO2 thin films were investigated with the analysis of XPS data. Figure 4,5 show the O1s and C1s core level spectra of the surface of aged-TiO2 thin films deposited at 350 and 450 °C before and after UVtreatment, respectively. The asymmetric O1s and C1s core level peaks were deconvoluted by using Lorentzian-Gaussian (function pseudo-Voigt) fitting analysis. The background subtraction was applied by using Shirley type fitting [5].

XPS Data Analysis of Aged-TiO2 Thin Films
The O1s core level peak of the aged-TiO2 thin films was deconvoluted into three peaks ( Figure  4a,b). The peak observed in binding energy (BE) value at 530.0 ± 0.3 eV was assigned to the Ti-O bond and the peak at BE value at 530.7 ± 0.4 eV was ascribed to the presence of oxygen vacancy defects (Vo). In addition, the peak at 532.1 ± 0.1 eV was attributed to surface hydroxyl groups (OH -), revealing that the film surface contained adsorbed surface OHgroups [43][44][45].
Figures 4c and d show the C1s core level spectra of aged-TiO2 thin films deposited at 350 and 450 °C, respectively. The spectra could be resolved into four peaks at BE of 285.0, 285.7, and 286.7, ±0.1, and 288.8 ± 0.3 eV. In both TiO2 samples, the C1s peaks located at 285.0 and 285.7 ± 0.1 eV corresponded to C=C and C-C/C-H, respectively [43]. It is highly probable that hydrocarbons and increasing deposition temperature. In the literature, it was observed that there is an inconsistency trend on the amount of OH − groups on the film surface concerning the increase in the film thickness [42] and heat treatment [48]. Simonsen et al. [48] reported that the surface of the TiO2 films grown by microwave assisted sol-gel at 550 °C had the lowest amount of surface OHgroups, while the hydroxyl groups on the film surface increased gradually with an increase in calcination temperature from 25 to 450 °C. A similar fluctuation in the amount of OH − groups was observed also in our previous study [5]. The O1s core level spectra of the surface of the UV-treated aged-TiO2 thin films could be deconvoluted into four peaks (Figure 5a,b). The peaks at 530.3 ± 0.3, 531.1 ± 0.3, and 532.0 eV were assigned as Ti-O, Vo, and OH -, respectively. In the literature, the peak observed in the BE value at 532.9 ± 0.1 was associated with different species, such as C-O, C=O, or chemisorbed-H2O molecules [41,50]. As seen in Table 1, after UV-treatment, the TiO2 thin film deposited at 350 °C possessed a higher level of oxygen vacancies on the top of the film surface compared to the 450 °C deposited sample. The amount of hydroxyl groups ((OH)/(Ti-O), Table 1), however, decreased considerably after UV treatment in the case of the TiO2 thin film sprayed at 450 °C.

XPS Data Analysis of Aged-TiO 2 Thin Films
The O1s core level peak of the aged-TiO 2 thin films was deconvoluted into three peaks (Figure 4a,b). The peak observed in binding energy (BE) value at 530.0 ± 0.3 eV was assigned to the Ti-O bond and the peak at BE value at 530.7 ± 0.4 eV was ascribed to the presence of oxygen vacancy defects (Vo). In addition, the peak at 532.1 ± 0.1 eV was attributed to surface hydroxyl groups (OH − ), revealing that the film surface contained adsorbed surface OH − groups [43][44][45]. Figure 4c,d show the C1s core level spectra of aged-TiO 2 thin films deposited at 350 and 450 • C, respectively. The spectra could be resolved into four peaks at BE of 285.0, 285.7, and 286.7, ±0.1, and 288.8 ± 0.3 eV. In both TiO 2 samples, the C1s peaks located at 285.0 and 285.7 ± 0.1 eV corresponded to C=C and C-C/C-H, respectively [43]. It is highly probable that hydrocarbons and other carbon-like species could be formed on the surface of TiO 2 layers as a result of preparation procedure or due to carbon contamination arising from aging of samples in ambient atmosphere [43,46]. Further, there were other peaks observed at 286.7 ± 0.1, 288.8 (Figure 4c), and 289.6 eV (Figure 4d) that belonged to oxygen species containing carbonyl moieties, such as C-O, C=O, and O=C-O, respectively [41][42][43]47,48].Moreover, the peak positioned at 290.07 ± 0.4 eV could have resulted from the π-π shake-up process (Figure 4c) [42]. The Scofield' cross-sections were applied to investigate the integrated areas of O1s core level spectra for the calculation of atomic concentrations of the components, such as Ti-O, Vo, and OH − presented in the O1s and C1s core level spectrums. The ratios of the components (OH)/(Ti-O) and (Vo)/(Ti-O) are shown in Table 1. As seen in Table 1, the amount of oxygen vacancies ((Vo)/(Ti-O) ratios) on the surface of TiO 2 films decreased from 0.17 up to 0.09 with the increase in the deposition temperature from 350 to 450 • C. This result agrees with our previous study, showing that the amount of oxygen vacancies decreases with increasing deposition temperature [5]. The lower number of oxygen vacancy defects on TiO 2 thin films deposited at 450 • C can be attributed to the effect of the high heat deposition temperature that repairs the surface oxygen defects [5]. It should be noted that the deposition temperature has a large effect on the properties of TiO 2 films deposited by spray pyrolysis. Even though the TiO 2 thin films were exposed to annealing treatment at high temperatures, it was shown in our previous studies that the structural, morphological, and chemical properties of TiO 2 thin films were found to depend on the film deposition temperature [5,32,49]. As seen in Table 1, the aged-TiO 2 films showed an increase in the amount of OH − groups on the film surface with increasing deposition temperature. In the literature, it was observed that there is an inconsistency trend on the amount of OH − groups on the film surface concerning the increase in the film thickness [42] and heat treatment [48]. Simonsen et al. [48] reported that the surface of the TiO 2 films grown by microwave assisted sol-gel at 550 • C had the lowest amount of surface OH − groups, while the hydroxyl groups on the film surface increased gradually with an increase in calcination temperature from 25 to 450 • C. A similar fluctuation in the amount of OH − groups was observed also in our previous study [5].
XPS Data Analysis of Aged-TiO 2 Thin Films after UV-Treatment Figure 5 represents the O1s and C1s core level spectra of UV-treated aged-TiO 2 thin films fabricated at 350 and 450 • C.
The O1s core level spectra of the surface of the UV-treated aged-TiO 2 thin films could be deconvoluted into four peaks (Figure 5a,b). The peaks at 530.3 ± 0.3, 531.1 ± 0.3, and 532.0 eV were assigned as Ti-O, Vo, and OH − , respectively. In the literature, the peak observed in the BE value at 532.9 ± 0.1 was associated with different species, such as C-O, C=O, or chemisorbed-H 2 O molecules [41,50]. As seen in Table 1, after UV-treatment, the TiO 2 thin film deposited at 350 • C possessed a higher level of oxygen vacancies on the top of the film surface compared to the 450 • C deposited sample. The amount of hydroxyl groups ((OH)/(Ti-O), Table 1), however, decreased considerably after UV treatment in the case of the TiO 2 thin film sprayed at 450 • C.
A smaller amount of carbon species was observed on the surface of TiO 2 thin films after UV-treatment (Figure 5c [45][46][47][48]. The peak at 287.0 eV in the carbon spectra of the TiO 2 sample grown at 450 • C ( Figure 5d) could be assigned as C-O-C [45]. In the case of the TiO 2 film grown at 350 • C, the peak at BE value at 289.0 eV ( Figure 5) could be assigned to C=O [41,45]. In addition, it should be noted that the peak located in BE energy at 285.7 attributed to C-H (

Wettability
The water contact angles (WCA) for as-prepared TiO 2 films deposited at 350 and 450 • C were 12 and 17 • , respectively. The films were aged in air over a month, and the increase was observed up to 46 and 55 • ( Figure 6). The increase in the WCA during aging could be due to the formation of a hydrocarbon layer on the film surface in air (Table 1) [52].
As seen in Figure 6, short UV irradiation of 5 min decreased the WCA of TiO 2 film deposited at 350 • C from 46.8 • to 10 • . Further UV irradiation (in total 10 min) decreased the WCA even up to 0 • . Thus, this revealed that a relatively short UV irradiation time of 5 min was necessary to modify the surface behavior of TiO 2 film deposited at 350 • C from hydrophilic to superhydrophilic. Opposite to that, the WCA of TiO 2 films deposited at 450 • C decreased gradually from 55 • to 0 • within 35 min ( Figure 6).

Wettability
The water contact angles (WCA) for as-prepared TiO2 films deposited at 350 and 450 °C were 12 and 17°, respectively. The films were aged in air over a month, and the increase was observed up to 46 and 55° ( Figure 6). The increase in the WCA during aging could be due to the formation of a hydrocarbon layer on the film surface in air (Table 1) [52]. As seen in Figure 6, short UV irradiation of 5 min decreased the WCA of TiO2 film deposited at 350 °C from 46.8° to 10°. Further UV irradiation (in total 10 min) decreased the WCA even up to 0°. Thus, this revealed that a relatively short UV irradiation time of 5 min was necessary to modify the surface behavior of TiO2 film deposited at 350 °C from hydrophilic to superhydrophilic. Opposite to that, the WCA of TiO2 films deposited at 450 °C decreased gradually from 55° to 0° within 35 min ( Figure 6).
Different behaviors of WCA under UV-irradiation of TiO2 films deposited at 350 and 450 °C can be explained by different surface chemical compositions of TiO2 thin films. The changes in the degree of hydroxyl groups, oxygen defects, and hydrocarbons present on the TiO2 surface could be related In comparison to TiO 2 films, the amount of hydroxyl groups was found to be considerably higher in the TiO 2 thin film deposited at 350 • C after UV irradiation (Table 1), compared to films deposited at 450 • C. This can be explained by the high amount of inherent oxygen vacancy defects on the surface of aged-TiO 2 films (Table 1), leading to the trapping of hydroxyl groups that enhanced the hydrophilic surface [41,42,53]. Moreover, the rapid wetting ability of the TiO 2 sample deposited at 350 • C could be due to the difference in the ability of photodecomposition of the hydrocarbon layer on the TiO 2 surface [51,54]. to the rapid wettability conversion (in total 10 min) of the TiO2 thin film deposited at 350 °C, correlated with the results obtained by XPS.

Photocatalytic Activity
In comparison to TiO2 films, the amount of hydroxyl groups was found to be considerably higher in the TiO2 thin film deposited at 350 °C after UV irradiation (Table 1), compared to films deposited at 450 °C. This can be explained by the high amount of inherent oxygen vacancy defects on the surface of aged-TiO2 films (Table 1), leading to the trapping of hydroxyl groups that enhanced the hydrophilic surface [41,42,53]. Moreover, the rapid wetting ability of the TiO2 sample deposited at 350 °C could be due to the difference in the ability of photodecomposition of the hydrocarbon layer on the TiO2 surface [51,54].  As seen in Figure 7, the TiO2 thin films deposited at 350 °C showed higher photocatalytic activity than TiO2 thin films deposited at 450 °C. The MTBE at the initial concentration of ca.10 ppm was completely converted to CO2 and H2O at a residence time of 62.4 s on the TiO2 thin films deposited at 350 °C, while TiO2 thin films deposited at 450 °C exhibited MTBE conversion of ca. 75% after 78 s.  As seen in Figure 7, the TiO 2 thin films deposited at 350 • C showed higher photocatalytic activity than TiO 2 thin films deposited at 450 • C. The MTBE at the initial concentration of ca.10 ppm was completely converted to CO 2 and H 2 O at a residence time of 62.4 s on the TiO 2 thin films deposited at 350 • C, while TiO 2 thin films deposited at 450 • C exhibited MTBE conversion of ca. 75% after 78 s. The higher photocatalytic activity of TiO 2 thin film deposited at 350 • C can be ascribed to several factors including the surface morphology, higher level of oxygen vacancy defects, and OH − groups on the film surface compared to the samples deposited at 450 • C (Table 1). It is a well-known phenomenon that the presence of the surface-active OH − groups benefit photocatalytic reactions. It was reported that OH − groups can participate directly in the interface reactions by trapping photo-generated holes that diffuse to TiO 2 surfaces and producing reactive surface hydroxyl radicals. They can also act as active sites for adsorption/reaction of reactant molecules on TiO 2 [55]. Further, there is a mutual effect between the photo-induced electrons bound by oxygen vacancies and the adsorbed O 2 , and thus oxygen vacancies can support O 2 adsorbing [56]. This indicates that oxygen vacancies can favor the adsorbed O 2 to capture photo-induced electrons, simultaneously producing reactive oxygen species (ROS) and, thus, promoting the oxidation of organic substances. Therefore, it can be suggested that oxygen vacancies and other surface defects are in favor of photocatalytic reactions [57].
In the present study, the formation of large grains was observed on the TiO 2 thin film deposited at 450 • C compared to that of the films deposited at 350 • C (Figure 1a,c) due to the agglomeration of grains during the deposition process at higher temperature. This could be one additional reason for the reduction of the photocatalytic activity of TiO 2 films deposited at 450 • C. It was reported that the surface area decreases with the increase in the grain size, which is unfavorable for the enhancement of photocatalytic efficiency [58,59]. Furthermore, the agglomeration decreases the interface between the grains, and thus leads to the reduction of surface active sites available for photocatalytic reactions [59,60].
Further study of the characterization of gas-phase photocatalytic activity of synthesized thin films was carried out with the TiO 2 films deposited at 350 • C due to their higher photocatalytic activity ascertained with MTBE.

The Effect of Air Humidity on Photocatalytic Activity of TiO 2 Thin Films
VOCs, such as MTBE, acetone, acetaldehyde, and heptane were chosen as single model air pollutants for this study. The effect of different operating conditions, i.e., humidity, inlet concentration of pollutant, residence time, and source of irradiation, on the photocatalytic process performance was studied. The effect of increase in air humidity from ca. 6 to 40% RH on photocatalytic oxidation of heptane, acetaldehyde, acetone, and MTBE is presented in Figure 8 (residence time in reactor section was 15.6 s).
The results showed that under lower humidity the conversions of heptane, acetaldehyde, and acetone were 46, 74, and 93% at residence time of 78 s, respectively, and 100% conversion of MTBE was achieved at 62.4 s, as previously discussed (Figure 8a). During the photocatalytic oxidation of those VOCs, tert-butyl formate (TBF) and formic acid were detected as intermediate gas-phase products of MTBE and heptane, respectively. Under a relatively dry atmosphere, the formation of intermediate products followed by their degradation was observed. There was an increase in the concentration of formic acid achieving its maximum at residence time of 46.8 s and subsequently it decreased with the further increase in residence time (at 78.0 s, 1% of degraded heptane emerged as gaseous formic acid), whereas no TBF was observed at residence time of 78 s as this intermediate product was completely degraded.
Air humidity usually dually affects the photodegradation processes. It can contribute to surface re-hydration of photocatalysts, which could lead to the increase in concentration of •OH radicals and hence to an increase in reaction rate. Competitive adsorption between water vapor and the organic compounds, which inhibits the binding of reactant molecules, is also common at higher water vapor content in reactors [61]. Therefore, the effect of air humidity on the performance of photocatalytic oxidation of MTBE, acetone acetaldehyde, and heptane were investigated on TiO 2 thin films deposited at 350 • C. As could be observed from Figure 8a,b, the conversion efficiency of most VOCs diminished with the increase in the relative humidity from 6 to 40%.
(residence time in reactor section was 15.6 s).
The results showed that under lower humidity the conversions of heptane, acetaldehyde, and acetone were 46, 74, and 93% at residence time of 78 s, respectively, and 100% conversion of MTBE was achieved at 62.4 s, as previously discussed (Figure 8a). During the photocatalytic oxidation of those VOCs, tert-butyl formate (TBF) and formic acid were detected as intermediate gas-phase products of MTBE and heptane, respectively. Under a relatively dry atmosphere, the formation of intermediate products followed by their degradation was observed. There was an increase in the concentration of formic acid achieving its maximum at residence time of 46.8 s and subsequently it decreased with the further increase in residence time (at 78.0 s, 1% of degraded heptane emerged as gaseous formic acid), whereas no TBF was observed at residence time of 78 s as this intermediate product was completely degraded. Air humidity usually dually affects the photodegradation processes. It can contribute to surface re-hydration of photocatalysts, which could lead to the increase in concentration of •OH radicals and hence to an increase in reaction rate. Competitive adsorption between water vapor and the organic compounds, which inhibits the binding of reactant molecules, is also common at higher water vapor content in reactors [61]. Therefore, the effect of air humidity on the performance of photocatalytic oxidation of MTBE, acetone acetaldehyde, and heptane were investigated on TiO2 thin films deposited at 350 °C. As could be observed from Figure 8a and b, the conversion efficiency of most VOCs diminished with the increase in the relative humidity from 6 to 40%. The conversion of heptane decreased almost twice with the increase in the relative humidity. In contrast to dry air, in humid air no gaseous organic intermediate products of heptane were monitored, as formic acid was expected to stay adsorbed on the surface of hydrated photocatalyst film.
Humid atmosphere influenced also the formation of the MTBE intermediate product TBF and retarded slightly the overall degradation of MTBE. Complete conversion of MTBE on a coated glass with the surface area of TiO 2 thin film deposited at 350 • C of 600 cm 2 (78 s) was achieved also under humid air conditions. Only TBF was found as an intermediate product from the degradation of MTBE in dry and humid conditions. In dry air, TBF was converted to CO 2 and H 2 O at a residence time of 62.4 s, while the detected amount of TBF was higher at a higher humidity (at 78.0 s, 5% of degraded MTBE appeared as gaseous TBF). Thus, the increase in air humidity was slightly hindering the photocatalytic mineralization of MTBE on thin films. Acetone degradation was affected by relative humidity, showing a sharp drop in the conversion of acetone from 93 to 30%. No gas-phase intermediate products were detected in either dry or humid air during photodegradation of acetone over TiO 2 films. It is interesting to note that acetone is a polar molecule, which is supposed to penetrate through water to the surface of TiO 2 . However, in the present study, it was seen that higher humidity (40% RH) showed a detrimental effect on the adsorption and photocatalytic decomposition of acetone. Similar results were observed during the photodecomposition of acetone [28] and chlorobenzene [62], which are soluble in water, on the TiO 2 surface under high humidity conditions. The only compound of which oxidation was uninfluenced by variations in air humidity was acetaldehyde. There were no significant changes in the conversion of acetaldehyde (approx. 75% conversion at 78 s) with the increase in relative humidity from 6 to 40%. No gas-phase products other than CO 2 and H 2 O were observed during the photocatalytic degradation of acetaldehyde in both dry and humid air conditions (Figure 8a,b). It is known that transport of the reactant molecules through water molecules depends on the solubility of the molecules of interest in water. Acetaldehyde is a polar molecule, which is water-soluble. This explains the "immunity" of photocatalytic oxidation of acetaldehyde molecules to the difference in air humidity. Generalizing, it could be assumed that the potential influence of air humidity on the performance of thin films depends not only on the character of the initial compounds to be oxidized, but also on the character (adsorption affinity, ease of degradation) of the intermediate products. The difference in the trend of acetone and acetaldehyde degradation under humid conditions is assumed to be attributed to the formation of intermediate products, which in case of acetone presumably deteriorate the overall performance of the photocatalytic process.

The Effect of Residence Time and Inlet Concentration of VOCs on Photocatalytic Activity of TiO 2 Thin Films
The effect of shorter residence time and lower inlet concentration of VOCs was examined on the surface of TiO 2 thin film with initial pollutant concentration of 5 ppm and relative decrease in residence time in the reactor section from 15.6 to 7.8 s (Figure 9). The residence time in the reactor section shifted from 15.6 to 7.8 s with the increase in the airflow rate from 0.5 L min −1 to 1.0 L min −1 . Generally, when the airflow rate is increased, two opposing processes can be envisaged: the decrement in residence time and the enhancement in mass transfer rate, which are, respectively, inhibiting and promoting factors in photocatalytic oxidation processes [63]. By reducing the residence time, the pollutants a have shorter time for adsorption on the surface and participating in oxidation reactions. As seen in Figure 9b, a general decrease in the conversion of VOCs was observed within five sections of the reactor, i.e., on the same photocatalytic surface area (600 cm 2 ), but with a twice shorter residence time if compared to the results in Figure 9b. If comparing the same residence time in the reactor at different airflow regimes, then at shorter residence time of 15.6 s (one section of reactor in Figure 9a and two sections in Figure 9b) there is a decrease in the conversions of VOCs despite the higher photocatalytic area in two sections of the reactor in Figure  9b. However, at residence time of 31.2 s (two sections of the reactor in Figure 9a and four sections in Figure 9b), the two times higher airflow rate is favoring the photocatalytic oxidation of VOCs at higher photocatalytic surface: the conversions of all the compounds increased from 25, 38, 47, and 86% to 34, 47, 64, and 100% for heptane, acetone, acetaldehyde, and MTBE, respectively. The process of photocatalytic oxidation of acetaldehyde and MTBE was revealed as especially benefitting from the intensification of mass transfer in the reactor at the same residence time. Thus, again, the effect of the different airflow regimes was expected to be different for each VOC.
The order in conversion of VOCs at residence time of 39 s in five sections was MTBE > acetaldehyde > acetone > heptane (Figure 9b), while it was MTBE > acetone > acetaldehyde > heptane at longer residence time (Figure 9a). For instance, the conversion of acetaldehyde and MTBE was affected slightly by shorter residence time, while the impact of shorter residence time was remarkable for photocatalytic decomposition of acetone. A sharp decrease in the conversion of acetone from 100% at residence time of 62.4 s (Figure 9a) to 47% at residence time of 31.2 s (Figure 9b) was observed on the same photocatalytic surface area of 480 cm 2 (four sections), whereas high conversions of acetaldehyde (82% conversion at 39.0 s) and MTBE (100% conversion at 31.2 s) were easily achievable at shorter residence time (Figure 9b). This could be attributed to the different reaction rates of each VOCs. Ibrahim et al. [64] studied the kinetics model of degradation of acetone and acetaldehyde on Degussa P25. It was reported that initial reaction rate of acetaldehyde exceeded that of acetone by more than six times (833 and 125 µmol m −3 min −1 , respectively). The decrease in inlet concentration of VOCs from 10 to 5 ppm (Figures 8a and 9a) resulted in higher conversions of all model compounds. With the decrease in concentration from 10 to 5 ppm in 78 s the conversion of heptane increased from 46 to 65% (4.6 and 3.3 ppm of heptane degraded, respectively) and conversion of acetaldehyde increased from 74 to 85% (7.4 and 4.2 ppm of acetaldehyde degraded, respectively), whereas complete degradation of 5 ppm of acetone and MTBE was obtained within ca. 62 and 47 s, respectively, while 93% of acetone with initial concentration of 10 ppm was degraded (9.3 ppm) in 78 s. Thus, despite the lower conversions of VOCs at higher initial concentrations, the increase in inlet concentration leads to the larger absolute amounts of the degraded compounds.
The residence time in the reactor section shifted from 15.6 to 7.8 s with the increase in the airflow rate from 0.5 L min −1 to 1.0 L min −1 . Generally, when the airflow rate is increased, two opposing processes can be envisaged: the decrement in residence time and the enhancement in mass transfer rate, which are, respectively, inhibiting and promoting factors in photocatalytic oxidation processes [63]. By reducing the residence time, the pollutants a have shorter time for adsorption on the surface and participating in oxidation reactions. As seen in Figure 9b, a general decrease in the conversion of VOCs was observed within five sections of the reactor, i.e., on the same photocatalytic surface area (600 cm 2 ), but with a twice shorter residence time if compared to the results in Figure 9b. If comparing the same residence time in the reactor at different airflow regimes, then at shorter residence time of 15.6 s (one section of reactor in Figure 9a and two sections in Figure 9b) there is a decrease in the conversions of VOCs despite the higher photocatalytic area in two sections of the reactor in Figure 9b. However, at residence time of 31.2 s (two sections of the reactor in Figure 9a and four sections in Figure 9b), the two times higher airflow rate is favoring the photocatalytic oxidation of VOCs at higher photocatalytic surface: the conversions of all the compounds increased from 25, 38, 47, and 86% to 34, 47, 64, and 100% for heptane, acetone, acetaldehyde, and MTBE, respectively. The process of photocatalytic oxidation of acetaldehyde and MTBE was revealed as especially benefitting from the intensification of mass transfer in the reactor at the same residence time. Thus, again, the effect of the different airflow regimes was expected to be different for each VOC.
The order in conversion of VOCs at residence time of 39 s in five sections was MTBE > acetaldehyde > acetone > heptane (Figure 9b), while it was MTBE > acetone > acetaldehyde > heptane at longer residence time (Figure 9a). For instance, the conversion of acetaldehyde and MTBE was affected slightly by shorter residence time, while the impact of shorter residence time was remarkable for photocatalytic decomposition of acetone. A sharp decrease in the conversion of acetone from 100% at residence time of 62.4 s (Figure 9a) to 47% at residence time of 31.2 s (Figure 9b) was observed on the same photocatalytic surface area of 480 cm 2 (four sections), whereas high conversions of acetaldehyde (82% conversion at 39.0 s) and MTBE (100% conversion at 31.2 s) were easily achievable at shorter residence time (Figure 9b). This could be attributed to the different reaction rates of each VOCs. Ibrahim et al. [64] studied the kinetics model of degradation of acetone and acetaldehyde on Degussa P25. It was reported that initial reaction rate of acetaldehyde exceeded that of acetone by more than six times (833 and 125 µmol m −3 min −1 , respectively).

The Effect of Irradiation Source on Photocatalytic Activity of TiO 2 Thin Films
The photocatalytic activity of TiO 2 thin films sprayed at 350 • C was evaluated with 5 ppm MTBE and acetone in gas-phase under visible light in dry air ( Figure 10). TiO 2 thin film showed MTBE conversion of 60% and acetone conversion of 33% at residence time of 78 s. Under the same operating conditions except light source (UV-A light), the sample showed 100% conversion of MTBE at residence 46.8 (360 cm 2 ) and 100% conversion of acetone at residence time 62.4 s (480 cm 2 ) (Figure 9a). The remarkable reduction in photocatalytic activity of TiO 2 film was due to the optical gap width of stoichiometric TiO 2 (>3.1 eV) determining its photocatalytic activity under the action of near UV radiation, which constitutes only a small percent of the visible light spectrum [65,66].  (Figure 9a). The remarkable reduction in photocatalytic activity of TiO2 film was due to the optical gap width of stoichiometric TiO2 (>3.1 eV) determining its photocatalytic activity under the action of near UV radiation, which constitutes only a small percent of the visible light spectrum [65,66]. TBF was detected as gaseous intermediate of MTBE, while no gaseous intermediate products were monitored during the decomposition of acetone. The highest amount of TBF was observed at residence time of 46.8 s, which was 0.8 ppm of TBF from 1.65 ppm of degraded MTBE (48% of MTBE transformed to TBF). At higher residence times, the amount of TBF on TiO2 thin film was steadily decreasing, achieving 16% of degraded MTBE oxidized to TBF at 78 sec. In comparison, the TiO2 thin film was capable of overcoming by-product accumulation at a surface area residence time of 46.8 s (360 cm 2 ) under UV-A light ( Figure 9a); moreover, under UV-A light with the same other parameters, no other gaseous products than water and carbon dioxide were observed. The visible-light-activity of TiO2 could be related to the defect disorder and the consequent electronic structure. Firstly, it could be ascribed to the formation of oxygen defects on the film surface (Table 1). It was proposed that the oxygen vacancy defects act as shallow trapping sites, which is related to the formation of some sub-bands in the electronic band structure of TiO2, considered as the main dominating factor of the visible emission [32]. Additionally, it was observed that TiO2 thin films still have oxidized carbon species after UV irradiation (C-O, C=O, O=C-O, etc.) on the surface of TiO2 film, (Figure 5a), known as defect-containing sp 2 -hybridized carbons. It was reported that these defect-associated functional groups (carbon species) provide surface active sites, which influence the effective charge transfer .

Thin Film Synthesis and Materials Characterization
TiO2 thin films were fabricated by USP onto borosilicate glass at 350 and 450 °C. The TiO2 thin films were annealed at 500 °C for 1 h in air in a furnace and are labelled as as-prepared samples TBF was detected as gaseous intermediate of MTBE, while no gaseous intermediate products were monitored during the decomposition of acetone. The highest amount of TBF was observed at residence time of 46.8 s, which was 0.8 ppm of TBF from 1.65 ppm of degraded MTBE (48% of MTBE transformed to TBF). At higher residence times, the amount of TBF on TiO 2 thin film was steadily decreasing, achieving 16% of degraded MTBE oxidized to TBF at 78 sec. In comparison, the TiO 2 thin film was capable of overcoming by-product accumulation at a surface area residence time of 46.8 s (360 cm 2 ) under UV-A light ( Figure 9a); moreover, under UV-A light with the same other parameters, no other gaseous products than water and carbon dioxide were observed.
The visible-light-activity of TiO 2 could be related to the defect disorder and the consequent electronic structure. Firstly, it could be ascribed to the formation of oxygen defects on the film surface (Table 1). It was proposed that the oxygen vacancy defects act as shallow trapping sites, which is related to the formation of some sub-bands in the electronic band structure of TiO 2 , considered as the main dominating factor of the visible emission [32]. Additionally, it was observed that TiO 2 thin films still have oxidized carbon species after UV irradiation (C-O, C=O, O=C-O, etc.) on the surface of TiO 2 film, (Figure 5a), known as defect-containing sp 2 -hybridized carbons. It was reported that these defect-associated functional groups (carbon species) provide surface active sites, which influence the effective charge transfer .

Thin Film Synthesis and Materials Characterization
TiO 2 thin films were fabricated by USP onto borosilicate glass at 350 and 450 • C. The TiO 2 thin films were annealed at 500 • C for 1 h in air in a furnace and are labelled as as-prepared samples throughout the article. The details of the solution preparation and the operating parameters of ultrasonic spray pyrolysis were reported in our previous study [5]. Borosilicate glass was selected as a substrate for TiO 2 thin films since it contains few Na + ions, unlike soda-lime glass substrates [68]; it is known that Na + ions have a detrimental effect on photocatalytic activity [68].
Zeiss HR FESEM Ultra 55 SEM (Jena, Germany) with an acceleration voltage of 4.0 kV was performed to obtained surface morphology and the film thickness. NT-MDT Solver 47 PRO AFM system was used to study the surface morphology of the films. The AFM measurement was carried out in the non-contact mode (resolution in the range of 3 nm) and the samples were studied in 1 µm × 1 µm area per scans. The RMS roughness analysis was performed through the 3-D AFM scan according to the ISO 4287/1 standards.
X-ray diffraction (XRD) measurements were performed with a Rigaku Ultima IV diffractometer (Tokyo, Japan) with CuKα radiation ( Catalysts 2019, 9, x FOR PEER REVIEW 13 of 18 × 1 µm area per scans. The RMS roughness analysis was performed through the 3-D AFM scan according to the ISO 4287/1 standards. X-ray diffraction (XRD) measurements were performed with a Rigaku Ultima IV diffractometer (Tokyo, Japan) with CuKα radiation (ʎ = 1.5406 A˚, 40 kV at 40 mA). The thin films were studied in grazing-incidence diffraction geometry using PBA (parallel beam analyzer 0.112 deg.) soller slit. The angle between the film surface and the incident beam was fixed at 0.5°. The scintillation detector was used during the measurements. The measurements were carried out in 2 theta configurations with the 2 theta range of 20-50°, with a step of 0.02°. The MCS of thin films was calculated by the Scherrer method using the FWHM (full width at half maximum) of the (101) reflection of the TiO2 anatase phase.
The total transmittance spectra of the TiO2 films on borosilicate glass were measured in the wavelength range of 300-800 nm on a Jasco V-670 UV-VIS-NIR spectrophotometer equipped with a 40 nm integrating sphere.
XPS studies were performed on a Kratos Axis Ultra DLD (delayline detector) spectrometer in conjunction with a 165 mm hemispherical electron energy analyzer. The analysis was carried out with monochromatic Al Kα X-rays (1486.6 eV) operating at 15 kV and 225 W. All XPS spectra were recorded using an aperture slot of 300-700 mm and pass energy of 20 eV. The spectra were calibrated using C1s core level peak centered data at a binding energy of 285.0 eV. XPS measurements were carried out for the aged-TiO2 thin films before and after UV treatment. UV treatment was applied for 12 hours to the aged-TiO2 thin films in ambient air with an Actinic BL 15 W fluorescent lamp for removing contaminants from the TiO2 thin film surface.
Wettability of the films was determined by water contact angle (WCA) measurements. DSA 25 (KRUSS Instrument) was carried out at room temperature. A sessile drop fitting method was applied for WCA measurements. Actinic BL 15 W fluorescent lamp (Philips), with max emission at 365 nm, was used to investigate the dependence of WCA of aged-TiO2 films on UV irradiation under ambient conditions.

Photocatalytic Measurements
The photocatalytic activity of TiO2 thin films was studied in a five-section plug-flow photocatalytic reactor with a section volume of 130 mL. Details of the reactor setup were reported in a previous study [5]. MTBE (C2H12O), acetone (C3H6O), acetaldehyde (C2H4O), and heptane (C7H16) were used as model air pollutants (all reagents ≥ 99% purity). The inlet concentrations of the gaseous pollutant were 5 and 10 ppm. The relative humidity was maintained at 6 ± 1% and 40 ± 5% for dry and humid air conditions, respectively. In the photocatalytic experimental setup, air flows of 0.5 and −1 = 1.5406 A • , 40 kV at 40 mA). The thin films were studied in grazing-incidence diffraction geometry using PBA (parallel beam analyzer 0.112 deg.) soller slit. The angle between the film surface and the incident beam was fixed at 0.5 • . The scintillation detector was used during the measurements. The measurements were carried out in 2 theta configurations with the 2 theta range of 20-50 • , with a step of 0.02 • . The MCS of thin films was calculated by the Scherrer method using the FWHM (full width at half maximum) of the (101) reflection of the TiO 2 anatase phase.
The total transmittance spectra of the TiO 2 films on borosilicate glass were measured in the wavelength range of 300-800 nm on a Jasco V-670 UV-VIS-NIR spectrophotometer equipped with a 40 nm integrating sphere.
XPS studies were performed on a Kratos Axis Ultra DLD (delayline detector) spectrometer in conjunction with a 165 mm hemispherical electron energy analyzer. The analysis was carried out with monochromatic Al Kα X-rays (1486.6 eV) operating at 15 kV and 225 W. All XPS spectra were recorded using an aperture slot of 300-700 mm and pass energy of 20 eV. The spectra were calibrated using C1s core level peak centered data at a binding energy of 285.0 eV. XPS measurements were carried out for the aged-TiO 2 thin films before and after UV treatment. UV treatment was applied for 12 hours to the aged-TiO 2 thin films in ambient air with an Actinic BL 15 W fluorescent lamp for removing contaminants from the TiO 2 thin film surface.
Wettability of the films was determined by water contact angle (WCA) measurements. DSA 25 (KRUSS Instrument) was carried out at room temperature. A sessile drop fitting method was applied for WCA measurements. Actinic BL 15 W fluorescent lamp (Philips), with max emission at 365 nm, was used to investigate the dependence of WCA of aged-TiO 2 films on UV irradiation under ambient conditions.

Photocatalytic Measurements
The photocatalytic activity of TiO 2 thin films was studied in a five-section plug-flow photocatalytic reactor with a section volume of 130 mL. Details of the reactor setup were reported in a previous study [5]. MTBE (C 2 H 12 O), acetone (C 3 H 6 O), acetaldehyde (C 2 H 4 O), and heptane (C 7 H 16 ) were used as model air pollutants (all reagents ≥ 99% purity). The inlet concentrations of the gaseous pollutant were 5 and 10 ppm. The relative humidity was maintained at 6 ± 1% and 40 ± 5% for dry and humid air conditions, respectively. In the photocatalytic experimental setup, air flows of 0.5 and 1 L min −1 were used, which gave the residence time of 15.6 or 7.8 s per reactor section, respectively. The temperature in the reactor was not additionally regulated; equilibrium was reached between the surrounding temperature, heating by the UV-A lamps, and application of reflectors resulting in 33 ± 2 • C. The following lamps such as UV Philips Actinic BL 15 W with an average irradiance of 3.5 mW cm −2 with reflector (integrated in the range of 180-400 nm, with max emission at 365 nm, UV-B/UV-A ratio < 0.2%), or VIS Philips TL-D 15 W, irradiance 3.3 mW cm −2 with reflector (integrated in the range of 180-700 nm, UV/UV VIS ratio < 5%) were used as the UV and VIS sources. Irradiance from UV and VIS sources was measured by a fiber optic spectrometer (USB-2000 + UV-VIS, Ocean Optics).
MTBE, acetone, acetaldehyde, and heptane peaks were quantified by INTERSPEC 200-X FTIR spectrometer with a Specac Tornado 8-m gas cell at the IR bands of 1063-1124, 1172-1245, 2630-2910, and 2825-3010 cm −1 , respectively. The gas-phase intermediate products of MTBE and heptane, TBF and formic acid, respectively, were also quantitatively monitored by means of FTIR at the IR bands from 1138 to 1190 and from 1103 to 1107 cm −1 , respectively.

Conclusions
TiO 2 thin films were deposited by ultrasonic spray pyrolysis on borosilicate glass substrates at 350 and 450 • C, followed by annealing at 500 • C for 1 h in air. Photocatalytic activity of TiO 2 thin films were evaluated as a function of their morphological, structural, and optical properties. According to the SEM cross-sectional images, the thickness of TiO 2 films deposited at 350 and 450 • C were 190 and 330 nm, respectively. TiO 2 films possess smooth surface morphology, showing the RMS roughness in the range of 0.6-1 nm. The films consist of an anatase crystalline structure with the mean crystallite size in the range of 22-26 nm and indicate total transmittance of 80% in the visible spectral region. XPS study confirmed that UV-treatment was effective to reduce the amount of hydrocarbons on the film surface and to modify the surface chemical composition. It was confirmed that after UV-treatment, the surface of TiO 2 films deposited at 350 • C possessed higher amounts of oxygen vacancy defects and hydroxyl groups ( Photocatalytic activity tests of TiO 2 thin films revealed that the films deposited at 350 • C with surface area of 480 cm 2 effectually degraded 100% of the MTBE, whereas TiO 2 thin film deposited at 450 • C showed the MTBE conversion of ca. 75% with a surface area of 600 cm 2 . TiO 2 thin film deposited at 350 • C with surface area of 600 cm 2 effectually degraded, 90% of acetone and 75% of acetaldehyde vapor (inlet 10 ppm) under dry air (6% RH) conditions without the formation of gas-phase intermediate products. Complete degradation of MTBE as well as 80% conversion of acetaldehyde were also achieved under humid air (40% RH) conditions at residence time of 78 s. With the decrease in inlet concentration from 10 to 5 ppm the increase in conversion efficiency of heptane and acetaldehyde was obtained at residence time of 78 s, whereas complete degradation of 5 ppm of acetone and MTBE was obtained within ca. 62 and 47 s, respectively. TiO 2 thin films deposited at 350 • C were photocatalytically active under visible light degrading 60% of MTBE and 33% of acetone (inlet 5 ppm).
Based on the conducted analysis, it could be concluded that the amount of oxygen vacancy and hydroxyl groups on the TiO 2 surface highly influences the photocatalytic activity and wettability of sprayed TiO 2 thin films. This study demonstrated that TiO 2 thin films prepared by ultrasonic spray pyrolysis are sufficient to degrade different VOCs under different air conditions and irradiation sources and thereby are prospective coatings for self-cleaning and air purification applications.
Author Contributions: I.D. carried out the thin film deposition and material characterizations (XRD, optical properties, and wettability test), attended data collection for the photocatalytic activity test, took part in the design of the all figures and tables, participated in the writing of the manuscript, and drafted the manuscript. M.K. (Marina Krichevskaya) carried out the photocatalytic activity test and interpretation of data, and participated in the writing of the manuscript. A.K. made the XPS data analysis. I.O.A. carried out the design of the study, participated in data analysis, and was associated with preparing of the manuscript. M.K. (Malle Krunks) participated in data analysis. All authors gave final approval for publication.