Annealing Temperature-Dependent E ﬀ ects of Fe-Loading on the Visible Light-Driven Photocatalytic Activity of Rutile TiO 2 Nanoparticles and Their Applicability for Air Puriﬁcation

: Commercial rutile TiO 2 particles (200–300 nm) were modiﬁed by the temperature-regulated chemical vapor deposition (tr-CVD) of Fe-oxide and subsequent annealing at various temperatures (300~750 ◦ C). As a result of the modiﬁcation, the photocatalytic activity of the TiO 2 regarding acetaldehyde removal under visible light was enhanced, and the enhancement e ﬀ ects were dependent on the annealing temperature. Speciﬁcally, the enhancement e ﬀ ects of the modiﬁcation were most pronounced when Fe-TiO 2 was annealed at 375 ◦ C, whereas the e ﬀ ects were signiﬁcantly reduced by annealing at higher temperatures (525 and 750 ◦ C). The analytical results with various techniques, including two surface-sensitive methods (XPS (X-ray photoelectron spectroscopy) and TOF-SIMS (time of ﬁght-secondary ion mass spectrometry)), revealed that the stronger metal support interaction between TiO 2 and the loaded Fe-oxide at high temperature ( > 375 ◦ C) resulted in the decreased charge separation e ﬃ ciency and photocatalytic activity of the Fe-TiO 2 under light irradiation. The production scale for the Fe-TiO 2 photocatalysts can be easily increased (from 200 g to 8 kg per the unit process) by upsizing the reactor volume. The mass-produced samples exhibited similar activity to the samples produced at small scale, and they were photocatalytically active after being spread on a cement block (stainless steel plate) using a surface hardening agent (paint), showing the high applicability in real applications.


Introduction
The photocatalytic removal of harmful organic compounds in either the aqueous or gas phase has been studied extensively [1][2][3][4][5][6]. When the surface of a semiconductor is exposed to light whose energy exceeds the band gap of the semiconductor, electron-hole pairs can be generated upon the absorption of the light energy by the semiconductor. The light-induced electron-hole pairs can interact with O 2 and H 2 O molecules at the surface of the semiconductor, generating strong oxidizing agents such as O 2 − and OH radicals, respectively. These oxidizing agents can interact with other organic compounds on the surface, oxidizing them into less harmful species, e.g., CO 2 and H 2 O, by total oxidation. Acetaldehyde is in the class of volatile organic compounds (VOCs) that have been blamed for adverse effects on human health [7][8][9][10][11]. The photocatalytic removal of acetaldehyde using semiconducting materials has been studied by many research groups [12][13][14][15]. Among semiconducting

Characterization
The amount of Fe-loading in TiO 2 particles was determined to be 0.4 wt% by ICP-OES analysis. Figure 1a-c show TEM images of three samples (TiO 2 , and Fe-TiO 2 -375 and -750). In all three cases, the TiO 2 particles were hexagonal in shape, with a size range of 200-300 nm. Any changes in the structure of the TiO 2 particles upon Fe-loading and annealing were not noticeable in the TEM images, Catalysts 2020, 10, 739 3 of 18 which can be attributed to the very small amount of Fe-loading (0.4 wt%) on the TiO 2 particles. The crystallinities of all the synthesized samples (Fe-TiO 2 -300, -375, -450, -525, and -750) were examined by obtaining their XRD patterns. Figure 1d shows the XRD patterns of three samples (TiO 2 , and Fe-TiO 2 -375 and -750) and pristine TiO 2 (a comparison of the XRD patterns of all the synthesized samples and pristine TiO 2 can be found in Supporting information). All the XRD patterns were almost identical to those of pristine TiO 2 and were well matched with the standard rutile phase TiO 2 spectrum (JCPDS No. 88-1174), and other peaks related to anatase or brookite TiO 2 were not observed. This indicated that no phase transformations of TiO 2 took place, and the loading amount of Fe was too small to be detected in the XRD analysis.
structure of the TiO2 particles upon Fe-loading and annealing were not noticeable in the TEM images, which can be attributed to the very small amount of Fe-loading (0.4 wt%) on the TiO2 particles. The crystallinities of all the synthesized samples (Fe-TiO2-300, -375, -450, -525, and -750) were examined by obtaining their XRD patterns. Figure 1d shows the XRD patterns of three samples (TiO2, and Fe-TiO2-375 and -750) and pristine TiO2 (a comparison of the XRD patterns of all the synthesized samples and pristine TiO2 can be found in Supporting information). All the XRD patterns were almost identical to those of pristine TiO2 and were well matched with the standard rutile phase TiO2 spectrum (JCPDS No. 88-1174), and other peaks related to anatase or brookite TiO2 were not observed. This indicated that no phase transformations of TiO2 took place, and the loading amount of Fe was too small to be detected in the XRD analysis.
The specific Brunauer-Emmett-Teller (BET) surface areas of all the synthesized samples and pristine TiO2 were examined (Supporting information), and Table 1 shows the specific BET surface areas of three samples (TiO2, and Fe-TiO2-375 and -750). The surface area of TiO2 gradually decreased as the annealing temperature was increased from 375 to 750 °C after Fe-loading. However, only a marginal decrease in the surface area of the TiO2 was observed (from 15.1 to 11.8 m 2 /g), even after the 750 °C annealing following the Fe-loading.   The specific Brunauer-Emmett-Teller (BET) surface areas of all the synthesized samples and pristine TiO 2 were examined (Supporting information), and Table 1 shows the specific BET surface areas of three samples (TiO 2 , and Fe-TiO 2 -375 and -750). The surface area of TiO 2 gradually decreased as the annealing temperature was increased from 375 to 750 • C after Fe-loading. However, only a marginal decrease in the surface area of the TiO 2 was observed (from 15.1 to 11.8 m 2 /g), even after the 750 • C annealing following the Fe-loading.

Photocatalytic Removal of Acetaldehyde
The photocatalytic degradation of gaseous acetaldehyde under visible light irradiation was carried out with TiO 2 and Fe-TiO 2 -300, -375, -450, -525, and -750, and the results are shown in Figure 2. The experiments were carried out with a drop-casted film consisting of pristine TiO 2 and the synthesized Fe-TiO 2 samples prepared by a drop-casting method (the experimental details can be found in Section 3.3.). Scanning tunneling microscopic (SEM) analysis revealed that each drop-casted film consisted of individual TiO 2 particles without noticeable agglomeration among the particles, indicating that the individual TiO 2 particles were almost intact after the drop-casting process (Supporting information).

Photocatalytic Removal of Acetaldehyde
The photocatalytic degradation of gaseous acetaldehyde under visible light irradiation was  carried out with TiO2 and Fe-TiO2-300, -375, -450, -525, and -750, and the results are shown in Figure  2. The experiments were carried out with a drop-casted film consisting of pristine TiO2 and the synthesized Fe-TiO2 samples prepared by a drop-casting method (the experimental details can be found in Section 3.3.). Scanning tunneling microscopic (SEM) analysis revealed that each drop-casted film consisted of individual TiO2 particles without noticeable agglomeration among the particles, indicating that the individual TiO2 particles were almost intact after the drop-casting process (Supporting information). Figure 2a,b show the concentration changes of acetaldehyde and CO2 during the photocatalysis experiments under visible light irradiation. The negative time scales correspond to the pre-adsorption stage in the dark, which was performed prior to the photocatalysis experiments. No notable changes in the concentration of acetaldehyde were observed for any of the TiO2 samples (TiO2 and Fe-TiO2-300, -375, -450, -525, and -750) at the pre-adsorption stage, indicating that the adsorption of acetaldehyde in the dark was negligible for all the cases of the samples.    concentration changes of acetaldehyde and CO 2 during the photocatalysis  experiments under visible light irradiation. The negative time scales correspond to the pre-adsorption  stage in the dark, which was performed prior to the photocatalysis experiments. No notable changes  in the concentration of acetaldehyde were observed for any of the TiO 2 samples (TiO 2 and Fe-TiO 2 -300,  -375, -450, -525, and -750) at the pre-adsorption stage, indicating that the adsorption of acetaldehyde in the dark was negligible for all the cases of the samples.
Upon the visible light irradiation of the surface of all the samples, the concentration of acetaldehyde started to decrease, and that of gaseous CO 2 increased without an induction time, showing that the photocatalytic removal of acetaldehyde took place (Figure 2a,b). It is clear that the photocatalytic activity of TiO 2 in terms of acetaldehyde removal and CO 2 evolution under visible light irradiation was enhanced upon Fe-loading, irrespective of the annealing temperature within the range of 300 to 750 • C (Figure 2a,b). However, the enhancement effects of Fe-loading on the photocatalytic activity of TiO 2 under visible light depended on the annealing temperature. Almost all of the gaseous acetaldehyde molecules in the reactor were removed by three Fe-TiO 2 samples (Fe-TiO 2 -300, -375, and -400) after 590 min of blue light irradiation (Figure 2a). By contrast, only roughly half of the acetaldehyde molecules were removed by Fe-TiO 2 samples annealed at higher temperatures (Fe-TiO 2 -525 and -750) under the same experimental conditions (Figure 2a).
One might note that the evolution of CO 2 continued even after all the gaseous acetaldehyde molecules in the reactor were removed. For instance, in the presence of Fe-TiO 2 -375 samples, the concentration of gaseous CO 2 continued to increase even after 410 min when the acetaldehyde concentration in the reactor reached zero (Figure 2a,b). This phenomenon can occur when gaseous acetaldehyde is first removed by partial oxidation on the surface of a photocatalyst, and the partially oxidized species is later oxidized into CO 2 by a longer exposure to light [28][29][30]. The formation of acetic acid as an intermediate species on the surface of photocatalysts during the photocatalytic oxidation of acetaldehyde into CO 2 was previously reported [28][29][30]. The previous results indicated that the reaction rate of acetic acid oxidation into CO 2 was slower than that of acetic acid formation from acetaldehyde, which can result in the CO 2 selectivity being higher than 100% at longer reaction times [28][29][30].
The concentration change of acetaldehyde during each photocatalysis experiment was further examined by obtaining the C/C o value (%) for acetaldehyde and the acetaldehyde removal efficiency (%) for each duration of light irradiation. The C/C o value (%) for acetaldehyde was calculated as follows in Equation (1): Acetaldehyde concentration measured with each sample at each time Acetaldehyde concentration measured without sample The acetaldehyde removal efficiency (%) for each irradiation duration was determined as follows in Equation (2): For this calculation, the C/C o of each sample at each time was used. Table 2 summarizes the calculated values of the removal efficiency (%) of each sample at 410 and 590 min, and the removal efficiency (%) of each sample at 410 min is displayed in Figure 2c. The acetaldehyde removal efficiency of Fe-TiO 2 at 410 min increased significantly from that of TiO 2 (16.90%) when the Fe-TiO 2 sample was annealed at 300 • C (92.88%), and the efficiency was further increased upon the annealing of the Fe-TiO 2 sample at 375 • C (100%). On the other hand, as the annealing temperature increased above 375 • C, the enhancement effect of Fe-loading on the photocatalytic activity of TiO 2 for acetaldehyde removal under visible light irradiation became less pronounced. In particular, a significant reduction of the enhancement effect of Fe-loading was observed when the annealing temperature was higher than 450 • C. The selectivity of CO 2 was calculated for each sample as follows in Equation (3): The number of gaseous CO 2 molecules evolved with each sample for each time (The number of removed acetaldehyde molecules with sample for each time) × 2 × 100 (3) The denominator corresponds to the number of CO 2 molecules generated by the total oxidation of the removed acetaldehyde; the total oxidation of 1 mol of acetaldehyde generates 2 mol of CO 2 . Figure 2d shows that the CO 2 selectivity of the photocatalytic removal of acetaldehyde on the TiO 2 surface was also increased by Fe-loading and subsequent annealing. In addition, the enhancement effect of Fe-loading on the CO 2 selectivity also depended on the annealing temperature, similarly to the case of the acetaldehyde removal efficiency. The enhancement effects of Fe-loading on the CO 2 selectivity were most pronounced when the Fe-TiO 2 sample was annealed at 375 • C, whereas the effects became significantly less pronounced when the Fe-TiO 2 particles were annealed at higher temperatures (525 and 750 • C).

Further Analyses (UV Diffuse Reflectance Spectrophotometry (UV-DRS), Photoluminescence (PL), XPS, and TOF-SIMS)
The Fe-TiO 2 -375 samples exhibited the highest photocatalytic activity in terms of both acetaldehyde removal and CO 2 evolution under visible light irradiation among the five Fe-TiO 2 samples (Fe-TiO 2 -300, -375, -450, -525, and -750). The enhancement effects of Fe-loading on the photocatalytic activity of TiO 2 were reduced with increasing annealing temperature, and in particular, annealing at temperatures higher than 450 • C led to a significant reduction of the photocatalytic activity of Fe-TiO 2 . In order to understand the annealing temperature dependency of the enhancement effects of Fe-loading, three representative samples (Fe-TiO 2 -375, -525, and -750) were further analyzed by UV-DRS, PL, XPS, and TOF-SIMS.
The UV-vis reflectance of each sample was measured, and the obtained reflectance values were converted into values corresponding to the absorption values based on the Kubelka-Munk function [31][32][33] (Figure 3a). The UV-vis reflectance of the TiO 2 samples was also measured, and the values obtained using the Kubelka-Munk function are displayed in Figure 3a for comparison. The optical band gap of each sample can be determined by obtaining the X-axis intercept values of the extrapolated straight line of each Kubelka-Munk plot, and the optical band gap of TiO 2 was determined to be~3.01 eV. The optical band gap of TiO 2 was reduced upon the Fe-loading and subsequent annealing (Figure 3a), and higher temperature annealing at 750 • C reduced the optical band gap a little further than in the cases of annealing at 375 and 525 • C. However, the modification effects on the optical band gap of TiO 2 did not seem to be significant; the band gap value of the Fe-TiO 2 -750 sample was determined to be~2.96 eV. In addition, the optical band gap changes upon Fe-loading, and the annealing did not correlate to the annealing temperature dependency of the modification effects on the photocatalytic activity of the Fe-TiO 2 samples; Fe-TiO 2 -375 exhibited significantly higher photocatalytic activity than Fe-TiO 2 -525 and -750, but the band gap reduction was most pronounced for the case of the Fe-TiO 2 -750 sample. We would also like to mention that the optical band gaps of other synthesized Fe-TiO 2 samples (Fe-TiO 2 -300, -450, and -525) were also determined, and the variation among all the synthesized samples and pristine TiO 2 were only marginal (Supporting information). These results indicate that the modification effects on the photocatalytic activity and its annealing temperature dependency cannot be simply explained by the optical band gap reduction due to Fe-loading and thermal annealing.   shows photoluminescence (PL) spectra of four samples (TiO2 and Fe-TiO2-375, -525, and -750). The PL intensities of three Fe-TiO2 samples were significantly lower than that of TiO2, and Fe-TiO2-375 showed the lowest PL intensity among the three Fe-TiO2 samples. The PL signal from each sample is related to the separation efficiency of the electron-hole pairs under light excitation. The energy of the irradiating light was absorbed by each sample, and electron-hole pairs were generated (electrons in the conduction band and holes in the valence band). When the light-induced electron-hole pairs are recombined, light can be emitted and detected as a PL signal. Thus, the weaker the PL signal, the higher the efficiency of the electron-hole separation (less recombination of electronhole pairs took place). The PL analysis results indicated that Fe-loading can significantly improve the electron-hole separation efficiency, and the positive effects of Fe-loading on the charge separation efficiency depended on the annealing temperature; the effects were most pronounced upon the 375 °C -annealing of Fe-TiO2, whereas increased annealing temperatures (>375 °C, e.g., 525 and 750 °C) resulted in the reduction of the enhancement of the charge separation efficiency by Fe-loading.
Later, three Fe-TiO2 samples (Fe-TiO2-375, -525, and -750) were analyzed by two surface sensitive analysis techniques (XPS and TOF-SIMS), and the results are shown in Figures 4 and 5. Figure 4a-c show the Ti 2p, Fe 2p, and Al 2p core-level XPS spectra, respectively, of the three Fe-TiO2 samples. The intensity of each XPS spectrum was normalized to the respective Ti 2p peak intensity. The binding energy of the Ti 2p3/2 peaks of all three samples were ~ 458.7 eV, corresponding to Ti 4+ states [34][35][36], and the peak shapes and positions for the three samples were almost identical to each other ( Figure 4a). An Al 2p core-level XPS peak centered at ~75 eV was also observed for all three samples (Figure 4b), attributed to the existence of the alumina shell on the as-received TiO2 particles. Note  . The PL intensities of three Fe-TiO 2 samples were significantly lower than that of TiO 2 , and Fe-TiO 2 -375 showed the lowest PL intensity among the three Fe-TiO 2 samples. The PL signal from each sample is related to the separation efficiency of the electron-hole pairs under light excitation. The energy of the irradiating light was absorbed by each sample, and electron-hole pairs were generated (electrons in the conduction band and holes in the valence band). When the light-induced electron-hole pairs are recombined, light can be emitted and detected as a PL signal. Thus, the weaker the PL signal, the higher the efficiency of the electron-hole separation (less recombination of electron-hole pairs took place). The PL analysis results indicated that Fe-loading can significantly improve the electron-hole separation efficiency, and the positive effects of Fe-loading on the charge separation efficiency depended on the annealing temperature; the effects were most pronounced upon the 375 • C-annealing of Fe-TiO 2 , whereas increased annealing temperatures (>375 • C, e.g., 525 and 750 • C) resulted in the reduction of the enhancement of the charge separation efficiency by Fe-loading.
Later, three Fe-TiO 2 samples (Fe-TiO 2 -375, -525, and -750) were analyzed by two surface sensitive analysis techniques (XPS and TOF-SIMS), and the results are shown in Figures 4 and 5. Figure 4a-c show the Ti 2p, Fe 2p, and Al 2p core-level XPS spectra, respectively, of the three Fe-TiO 2 samples. The intensity of each XPS spectrum was normalized to the respective Ti 2p peak intensity. The binding energy of the Ti 2p 3/2 peaks of all three samples were~458.7 eV, corresponding to Ti 4+ states [34][35][36], and the peak shapes and positions for the three samples were almost identical to each other (Figure 4a). An Al 2p core-level XPS peak centered at~75 eV was also observed for all three samples (Figure 4b), Catalysts 2020, 10, 739 8 of 18 attributed to the existence of the alumina shell on the as-received TiO 2 particles. Note that alumina shells are generally considered to be required to increase the compatibility of TiO 2 with plastics or binders forming more uniform composite functional structures. In Supporting information, the Ti 2p and Al 2p core-level XPS spectra of the as-received TiO 2 are displayed together with those of the three Fe-TiO 2 samples (Fe-TiO 2 -375, -525-, and -750). It is shown that the Ti 2p and Al 2p XPS spectra did not undergo significant changes upon the modification, excepting for the slight decrease in the Al 2p signal with the increase in the annealing temperature after Fe deposition from 375 to 750 • C (Supporting information). It is also worth noting that a significant Ti signal was detected together with the Al signal in the XPS analysis, although the surface of as-received TiO 2 was covered with alumina, indicating that either the alumina shell was very thin (~2 nm) or the surface of the TiO 2 was not completely covered by the alumina shell.  Three Fe-TiO2 samples (Fe-TiO2-375, -525, and -750) were further analyzed by TOF-SIMS using a Bi3 + ion beam as a primary ion source. The intensities of the secondary ion signals measured by TOF-SIMS for each type of sample were normalized to the respective signal intensity of the Ti + ion. All the measured secondary ion signals are summarized in Supporting information. The secondary mass signals other than Al + and Fe + were categorized into two groups, FexOy + and FexTiy + , as shown of the loaded Fe-oxide by TiO2 (forming a compound such as FeTi2O3-x, in which Fe has an oxidation state of II or I). The SIMS analysis results were in line with the results of XPS analysis; both analysis results indicated the stronger SMSI effects of TiO2 on the loaded Fe-oxide upon Fe-TiO2 annealing at higher temperature (>375 °C). It is worth noting that both the SIMS and XPS analysis methods are surface-sensitive techniques, of which the typical analysis depths are ~2 nm and ~5 nm from the topmost surface, respectively. We suggested the following scenario, which can reconcile all our observations of UV/vis, PL, XPS, SIMS, and the temperature dependency of the enhancement effects of Fe-loading on the photocatalytic activity of Fe-TiO2 samples (Figure 6a). The electron-hole pairs were generated upon photon absorption. Electrons in the conduction band (or holes in the valence band) can transfer into Fe(III) atoms, reducing Fe(III) atoms into Fe(II) states (or oxidizing Fe(III) atoms into Fe(IV) states). If the Fe atoms in Fe(II) states (or Fe(IV) states) were available to the oxygen molecules (or water) adsorbed on the surface, electrons (or holes) could further transfer into oxygen molecules (or water), generating O2 -ions (or OH radicals), and then the Fe atoms would return to the original Fe(III) states (Figure 6a). When these processes take place, the electron-hole separation efficiency of the Fe-TiO2 samples under light irradiation can be improved, which can lead to the increased photocatalytic activity of Fe-TiO2.
The stronger SMSI effect of TiO2 at higher temperatures (525 and 750 °C) not only facilitated the scrambling up of loaded Fe-oxide by the underlying TiO2 but also resulted in the reduction of the loaded Fe atoms by the inter-diffusion of metal atoms between the TiO2 and oxidized Fe structure (Figure 6b) [43][44][45][46]. The former effect made Fe atoms unavailable to the adsorbed oxygen or water molecules, while the latter made the charge transfer from TiO2 to Fe atoms difficult [47,48]. That is to say, the effects of the stronger SMSI between TiO2 and the loaded Fe-oxide at higher temperature resulted in the decrease in the charge separation efficiency of Fe-TiO2, although, in our case study, the reduction of the oxidation states of the Fe atoms seemed to be more critical for decreasing the charge separation efficiency, and photocatalytic activity of Fe-TiO2 upon higher temperature annealing (>375 °C). Upon Fe-loading and annealing, Fe 2p 3/2 core-level XPS peaks appeared in the binding energy region of 713-703 eV (Figure 4c). The intensities of the Fe 2p core-level XPS spectra of the Fe-TiO 2 samples decreased with increasing annealing temperature, especially when the temperature was increased from 375 to 525 • C. This is similar to the case of the Al 2p core-level XPS spectra, where an intensity decrease was observed upon annealing at higher temperature, but the decrease in Fe 2p signal was more pronounced than that in the case of the Al 2p core-level. Along with the decrease in the Fe 2p signal intensity, a shift of the Fe 2p XPS peaks to the lower binding energy region was observed upon the higher temperature annealing; the Fe 2p 3/2 peak of Fe-TiO 2 -375 was centered at~710 eV, whereas the center position of the Fe 2p 3/2 peak of Fe-TiO 2 -525 was at~707 eV. The Fe 2p 3/2 peak position of Fe-TiO 2 -375 corresponded to oxidic Fe either in Fe(II) or Fe(III); by contrast, the peak positions of the two Fe-TiO 2 samples annealed at higher temperatures (Fe-TiO 2 -525 and -750) corresponded to Fe atoms in the metallic state or Fe(I) state. Both the intensity decreases and the lower binding energy shift of the Fe 2p XPS peak with increasing annealing temperatures (>375 • C) seemed to be caused by the strong metal-support interaction (SMSI) of TiO 2 with the loaded Fe-oxide. The SMSI between the group VIII metals (e.g., Fe and Ni) and TiO 2 substrate had been reported, which varied the catalytic activities of the metals (metal oxides) of the group VIII atoms regarding Fischer-Tropsch (F-T) synthesis [37][38][39]. The SMSI between Fe-oxide and TiO 2 can result in (1) the encapsulation of the Fe-oxide by the TiO 2 (the scrambling up of the underlying TiO 2 [40][41][42] or interdiffusion of the Fe atoms into the TiO 2 substrate [43][44][45][46]) and (2) the reduction of the Fe-oxide by the charge transfer [43][44][45][46]. The surface structure changes of the Fe-TiO 2 samples upon higher temperature annealing due to SMSI effects were further evidenced by the TOF-SIMS analysis, the details of which are discussed below.
Three Fe-TiO 2 samples (Fe-TiO 2 -375, -525, and -750) were further analyzed by TOF-SIMS using a Bi 3 + ion beam as a primary ion source. The intensities of the secondary ion signals measured by TOF-SIMS for each type of sample were normalized to the respective signal intensity of the Ti + ion. All the measured secondary ion signals are summarized in Supporting information. The secondary mass signals other than Al + and Fe + were categorized into two groups, Fe x O y + and Fe x Ti y + , as shown in Figure 5a. The intensity of FexOy+ corresponds to the sum of the secondary ion signals containing Fe and O atoms, e.g., FeO + , FeO 2 , Fe 2 O + , etc., whereas the Fe x Ti y + intensity shown in Figure 5a corresponds to the sum of the secondary ion signals containing Fe and Ti atoms. As the annealing temperature for the Fe-TiO 2 increased above 375 • C, the intensities of Fe x O y + and Fe x Ti y + with respect to Ti + gradually decreased, indicating the pronounced diffusion of Ti atoms to the top surface of the Fe-TiO 2 sample at higher temperatures (525 and 750 • C), which resulted in the higher coverage of the loaded Fe-oxide surface by TiO 2 . It is important to note that the decrease in the Fe x O y + /Ti + signal with increasing annealing temperatures above 375 • C was more pronounced than in the Fe x Ti y + /Ti + case. Figure 5b compares the Fe x Ti y + signal to FexO y + signal ratios of three samples (Fe-TiO 2 -375, -525, and -750). An increase in the Fe x Ti y + /Fe x O y + signal ratio was notable as the annealing temperature was increased above 375 • C. The SIMS data only monitor the topmost surface structure, and therefore, intermixing between Fe and Ti in deeper layers is not fully reflected here. Nevertheless, these results indicate that the interaction between Fe and TiO 2 became stronger at higher temperatures (especially >375 • C), which can lead to more interdiffusion of Ti atoms to the loaded Fe-oxide and the reduction of the loaded Fe-oxide by TiO 2 (forming a compound such as FeTi 2 O 3-x , in which Fe has an oxidation state of II or I). The SIMS analysis results were in line with the results of XPS analysis; both analysis results indicated the stronger SMSI effects of TiO 2 on the loaded Fe-oxide upon Fe-TiO 2 annealing at higher temperature (>375 • C). It is worth noting that both the SIMS and XPS analysis methods are surface-sensitive techniques, of which the typical analysis depths are~2 nm and~5 nm from the topmost surface, respectively. We suggested the following scenario, which can reconcile all our observations of UV/vis, PL, XPS, SIMS, and the temperature dependency of the enhancement effects of Fe-loading on the photocatalytic activity of Fe-TiO 2 samples (Figure 6a). The electron-hole pairs were generated upon photon absorption. Electrons in the conduction band (or holes in the valence band) can transfer into Fe(III) atoms, reducing Fe(III) atoms into Fe(II) states (or oxidizing Fe(III) atoms into Fe(IV) states). If the Fe atoms in Fe(II) states (or Fe(IV) states) were available to the oxygen molecules (or water) adsorbed on the surface, electrons (or holes) could further transfer into oxygen molecules (or water), generating O 2 − ions (or OH radicals), and then the Fe atoms would return to the original Fe(III) states (Figure 6a). When these processes take place, the electron-hole separation efficiency of the Fe-TiO 2 samples under light irradiation can be improved, which can lead to the increased photocatalytic activity of Fe-TiO 2 .
Catalysts 2020, 10, x FOR PEER REVIEW 11 of 18 Figure 6. The suggested scenario. Schematic depictions of (a) the suggested scenario that explains the

Large-Scale Production of Catalysts and Their Potential Applications
Finally, we explored the potential applications of the Fe-TiO2-375 particles. Firstly, we compared the photocatalytic activity of the Fe-TiO2-375 sample with that of the Fe-loaded (0.1 wt% of Fe) commercial TiO2 particles (P25, Evonik) annealed at 750 °C, hereafter denoted as Fe-P25-750. Previously, the Fe-P25-750 sample has been proven to exhibit the highest photocatalytic activity regarding acetaldehyde degradation under visible light irradiation among the P25 samples modified by the tr-CVD of Fe-oxide and thermal annealing [28]. The Fe-P25-750 sample containing 0.1 wt% of Fe was prepared via the tr-CVD of Fe-oxide on commercially available TiO2 (P25, Evonik) and subsequent annealing at 750 °C, as described in the literature [28]. The photocatalytic removal of acetaldehyde on the surface of Fe-P25-750 was performed under the same conditions as in the Fe-TiO2-375 case (Figure 7a,b). The Fe-P25-750 sample exhibited slightly better photocatalytic performance in terms of acetaldehyde removal and CO2 evolution compared to the Fe-TiO2-375 sample under the same conditions; a 100% removal efficiency for acetaldehyde was achieved by 370 min of visible light irradiation on the surface of Fe-P25-750, whereas a 100% removal efficiency was achieved on the surface of Fe-TiO2-375 by 410 min of visible light irradiation. However, one should note that the TiO2 used as a base material for Fe-TiO2-375 in this work was produced via a sulfuric acid method, and its price (3 USD/g) is much lower than that of the P25 (50-70 USD/g). In addition, the modification process for the Fe-TiO2-375 consumed much less energy than that for Fe-P25-750, Figure 6. The suggested scenario. Schematic depictions of (a) the suggested scenario that explains the surface structure changes associated with higher temperature annealing and (b) the annealing temperature dependency of the photocatalytic activity of the Fe-TiO 2 sample.
The stronger SMSI effect of TiO 2 at higher temperatures (525 and 750 • C) not only facilitated the scrambling up of loaded Fe-oxide by the underlying TiO 2 but also resulted in the reduction of the loaded Fe atoms by the inter-diffusion of metal atoms between the TiO 2 and oxidized Fe structure (Figure 6b) [43][44][45][46]. The former effect made Fe atoms unavailable to the adsorbed oxygen or water molecules, while the latter made the charge transfer from TiO 2 to Fe atoms difficult [47,48]. That is to say, the effects of the stronger SMSI between TiO 2 and the loaded Fe-oxide at higher temperature resulted in the decrease in the charge separation efficiency of Fe-TiO 2 , although, in our case study, the reduction of the oxidation states of the Fe atoms seemed to be more critical for decreasing the charge separation efficiency, and photocatalytic activity of Fe-TiO 2 upon higher temperature annealing (>375 • C).

Large-Scale Production of Catalysts and Their Potential Applications
Finally, we explored the potential applications of the Fe-TiO 2 -375 particles. Firstly, we compared the photocatalytic activity of the Fe-TiO 2 -375 sample with that of the Fe-loaded (0.1 wt% of Fe) commercial TiO 2 particles (P25, Evonik) annealed at 750 • C, hereafter denoted as Fe-P25-750. Previously, the Fe-P25-750 sample has been proven to exhibit the highest photocatalytic activity regarding acetaldehyde degradation under visible light irradiation among the P25 samples modified by the tr-CVD of Fe-oxide and thermal annealing [28]. The Fe-P25-750 sample containing 0.1 wt% of Fe was prepared via the tr-CVD of Fe-oxide on commercially available TiO 2 (P25, Evonik) and subsequent annealing at 750 • C, as described in the literature [28]. The photocatalytic removal of acetaldehyde on the surface of Fe-P25-750 was performed under the same conditions as in the Fe-TiO 2 -375 case (Figure 7a,b). The Fe-P25-750 sample exhibited slightly better photocatalytic performance in terms of acetaldehyde removal and CO 2 evolution compared to the Fe-TiO 2 -375 sample under the same conditions; a 100% removal efficiency for acetaldehyde was achieved by 370 min of visible light irradiation on the surface of Fe-P25-750, whereas a 100% removal efficiency was achieved on the surface of Fe-TiO 2 -375 by 410 min of visible light irradiation. However, one should note that the TiO 2 used as a base material for Fe-TiO 2 -375 in this work was produced via a sulfuric acid method, and its price (3 USD/g) is much lower than that of the P25 (50-70 USD/g). In addition, the modification process for the Fe-TiO 2 -375 consumed much less energy than that for Fe-P25-750, owing to the lower temperature of the thermal annealing process (375 • C vs. 750 • C). Thus, considering the lower production costs of the Fe-TiO 2 -375 relative to Fe-P25-750 and their comparable photocatalytic activities under visible light irradiation, the Fe-TiO 2 -375 sample seemed to be a more favorable choice for industrial applications than Fe-P25-750.
Catalysts 2020, 10, x FOR PEER REVIEW 12 of 18 owing to the lower temperature of the thermal annealing process (375 °C vs. 750 °C ). Thus, considering the lower production costs of the Fe-TiO2-375 relative to Fe-P25-750 and their comparable photocatalytic activities under visible light irradiation, the Fe-TiO2-375 sample seemed to be a more favorable choice for industrial applications than Fe-P25-750. We also tested the possibility of the mass production of Fe-TiO2-375 photocatalysts. The Fe-TiO2-375 samples were produced on a much larger scale; the amount of Fe-TiO2-375 particles (8 kg) produced in one sequential tr-CVD and annealing process was 40 times greater than that in the aforementioned case (200 g). The photocatalytic activity of the mass-produced Fe-TiO2-375 regarding acetaldehyde removal under the visible light irradiation was examined under the same conditions as for the case of the Fe-TiO2-375 produced at a small scale (200 g per sequential tr-CVD and annealing process) (Figure 7c,d). The mass-produced Fe-TiO2-375 exhibited a slightly lower efficiency of acetaldehyde removal (100% acetaldehyde removal at 600 min) than the Fe-TiO2-375 sample produced at the small scale, which was most likely due to the existence of an induction period for the activity of the mass-produced samples due to the higher level of impurities, but it exhibited slightly higher efficiency in terms of CO2 evolution. The results demonstrated the feasibility of the mass production of Fe-TiO2-375 and its usability as a photocatalyst for air purification.
One of the potential applications of photocatalyst particles for improving outdoor or indoor air quality is putting those particles on the outside or inside walls of buildings. In order to explore this We also tested the possibility of the mass production of Fe-TiO 2 -375 photocatalysts. The Fe-TiO 2 -375 samples were produced on a much larger scale; the amount of Fe-TiO 2 -375 particles (8 kg) produced in one sequential tr-CVD and annealing process was 40 times greater than that in the aforementioned case (200 g). The photocatalytic activity of the mass-produced Fe-TiO 2 -375 regarding acetaldehyde removal under the visible light irradiation was examined under the same conditions as for the case of the Fe-TiO 2 -375 produced at a small scale (200 g per sequential tr-CVD and annealing process) (Figure 7c,d). The mass-produced Fe-TiO 2 -375 exhibited a slightly lower efficiency of acetaldehyde removal (100% acetaldehyde removal at 600 min) than the Fe-TiO 2 -375 sample produced at the small scale, which was most likely due to the existence of an induction period for the activity of the mass-produced samples due to the higher level of impurities, but it exhibited slightly higher efficiency in terms of CO 2 evolution. The results demonstrated the feasibility of the mass production of Fe-TiO 2 -375 and its usability as a photocatalyst for air purification.
One of the potential applications of photocatalyst particles for improving outdoor or indoor air quality is putting those particles on the outside or inside walls of buildings. In order to explore this possibility, samples of Fe-TiO 2 -375 particles were mixed with a commercially available paint and a surface hardening agent, and the mixtures were spread on a stainless use steel (SUS) plate and a cement block, respectively. Then, the photocatalytic removal of acetaldehyde under visible light in the presence of each sample was examined under the same experimental conditions as in the aforementioned case (Fe-TiO 2 -375 drop-casted on a SUS plate) (Figure 8a,b). For both cases, the initiation of the photocatalytic removal of acetaldehyde and CO 2 evolution was observed upon visible light irradiation (Figure 8a,b). It is interesting that both samples exhibited similar acetaldehyde removal efficiency, but the Fe-TiO 2 -375 on the cement block showed a much higher efficiency of CO 2 evolution than the Fe-TiO 2 on the paint. This might be caused by the different natures of the adhesive materials between the two samples, but further systematic investigations are needed to reveal the effects of the adhesive materials on the photocatalytic activity of TiO 2 -based particles. Nevertheless, it is worth noting that the Fe-TiO 2 -375 particles can exhibit their photocatalytic activity under visible light irradiation after being mixed with other adhesive materials, which makes them applicable for air purification in outdoor or indoor environments.
Catalysts 2020, 10, x FOR PEER REVIEW 13 of 18 possibility, samples of Fe-TiO2-375 particles were mixed with a commercially available paint and a surface hardening agent, and the mixtures were spread on a stainless use steel (SUS) plate and a cement block, respectively. Then, the photocatalytic removal of acetaldehyde under visible light in the presence of each sample was examined under the same experimental conditions as in the aforementioned case (Fe-TiO2-375 drop-casted on a SUS plate) (Figure 8a,b). For both cases, the initiation of the photocatalytic removal of acetaldehyde and CO2 evolution was observed upon visible light irradiation (Figure 8a,b). It is interesting that both samples exhibited similar acetaldehyde removal efficiency, but the Fe-TiO2-375 on the cement block showed a much higher efficiency of CO2 evolution than the Fe-TiO2 on the paint. This might be caused by the different natures of the adhesive materials between the two samples, but further systematic investigations are needed to reveal the effects of the adhesive materials on the photocatalytic activity of TiO2-based particles. Nevertheless, it is worth noting that the Fe-TiO2-375 particles can exhibit their photocatalytic activity under visible light irradiation after being mixed with other adhesive materials, which makes them applicable for air purification in outdoor or indoor environments.

Small-Scale Production
FeOx/TiO2 photocatalysts were prepared by a temperature-regulated chemical vapor deposition (tr-CVD) method. Commercial titanium dioxide (TiO2), bought from Lomo Billions, China, was used as a substrate. Bis(cyclopentadienyl)iron (Ferrocene, Fe(Cp)2, Sigma Aldrich, USA) was used as a metal precursor, whereas oxygen and water in ambient air were used as the oxidizing agent. Two hundred grams of TiO2 was mixed with 0.667 g of Fe(Cp)2 (weight ratio of 300:1), and the mixture was placed in a quartz beaker (250 mL). Our previous studies showed that Fe-loading of less than 2% can have a significant enhancement effect on photocatalytic activity [28]. We used a different Feloading here, and the Fe-loading of 0.4 wt% prepared by a 1:300 ratio of TiO2 and Fe(Cp)2 resulted in better CO2 selectivity, even though the acetaldehyde removal rate appeared to be less sensitive to the Fe-loading in this range (Supporting information). Thus, we concentrated on the Fe-loading of 0.4 wt% in the present work. The top part of the quartz beaker was covered with aluminum foil, and the gap between the beaker and aluminum foil was sealed with polyimide (PI) tape. The beaker was located in a box furnace (DMF-3T, U1TECH, Hwaseong, Korea) with an internal volume of 3 L. The temperature was increased to 60 °C and kept at 60 °C for 1 h for the evaporation of the Fe(CP)2. After

Small-Scale Production
FeO x /TiO 2 photocatalysts were prepared by a temperature-regulated chemical vapor deposition (tr-CVD) method. Commercial titanium dioxide (TiO 2 ), bought from Lomo Billions, China, was used as a substrate. Bis(cyclopentadienyl)iron (Ferrocene, Fe(Cp) 2 , Sigma Aldrich, USA) was used as a metal precursor, whereas oxygen and water in ambient air were used as the oxidizing agent. Two hundred grams of TiO 2 was mixed with 0.667 g of Fe(Cp) 2 (weight ratio of 300:1), and the mixture was placed in a quartz beaker (250 mL). Our previous studies showed that Fe-loading of less than 2% can have a significant enhancement effect on photocatalytic activity [28]. We used a different Fe-loading here, and the Fe-loading of 0.4 wt% prepared by a 1:300 ratio of TiO 2 and Fe(Cp) 2 resulted in better CO 2 selectivity, even though the acetaldehyde removal rate appeared to be less sensitive to the Fe-loading in this range (Supporting information). Thus, we concentrated on the Fe-loading of 0.4 wt% in the present work. The top part of the quartz beaker was covered with aluminum foil, and the gap between the beaker and aluminum foil was sealed with polyimide (PI) tape. The beaker was located in a box furnace (DMF-3T, U1TECH, Hwaseong, Korea) with an internal volume of 3 L. The temperature was increased to 60 • C and kept at 60 • C for 1 h for the evaporation of the Fe(CP) 2 . After 1 h, the temperature was increased to 200 • C, and this temperature was maintained for 1 h to oxidize the Fe(Cp) 2 on the surface of TiO 2 . The aluminum foil and PI tape were removed after the beaker was cooled down to room temperature. Then, the tr-CVD-prepared samples were further annealed at various temperatures (300, 375, 450, 525, and 750 • C) for 2 h under ambient conditions. Hereafter, the Fe-oxide-loaded TiO 2 samples prepared via the tr-CVD of Fe-compounds and subsequent annealing at various temperatures are referred as Fe-TiO 2 -annealing temperature, e.g., Fe-TiO 2 -375 for 375 • C-annealed Fe-deposited TiO 2 sample, whereas TiO 2 is referred to as the as-received TiO 2 .

Large-Scale Production
For the mass production of the Fe-deposited TiO 2 photocatalysts, a box furnace (PMF27, LAB HOUSE, Seoul, Korea) with a larger internal volume (27 L) was used, and the quartz beaker was replaced by an alumina crucible (size, 135 mm × 135 mm × 60 mm). Eight kilograms of TiO 2 was mixed with 27 g of Fe(Cp) 2 for 30 min using a cement mixer (DCM-300R, DCM corporation, Yangsan, Korea). The mixture of TiO 2 and Fe(Cp) 2 was distributed in the alumina crucibles (15 EA). The top of each alumina crucible was covered with an alumina lid, and the gap between the crucible and the lid was sealed with PI tape. Then, the crucibles were placed in the box furnace (internal volume of 27 L), and the tr-CVD of Fe-oxide on TiO 2 particles was initiated by increasing the temperature in the same manner as in the aforementioned case (small-scale production, 1 h of 60 • C-heating followed by 1 h of 200 • C-heating). After the tr-CVD of the Fe-compound on the TiO 2 particles, the crucibles were cooled down until the temperature reached room temperature, and then the lids and PI tape were removed from the tops of all the crucibles. Then, the Fe-TiO 2 was further annealed at 375 • C for 2 h under ambient conditions using the same furnace (27 L-box furnace).

Characterization of Photocatalysts
The amounts of the Fe loading of Fe-TiO 2 were determined by inductively coupled plasma optical emission spectroscopy (ICP-OES, Varian, Agilent, Santa Clara, California, USA). Three samples (TiO 2 , and Fe-TiO 2 -375 and -750) were selected for transmission electronic microscopy (TEM), X-ray diffraction (XRD), and surface area analyses. Structural images of three samples were obtained using TEM (JEM-3010, Jeol, Japan). The crystallinity of each sample was studied by obtaining X-ray diffraction (XRD) patterns using a diffractometer (Ultima IV, Rigaku, Tokyo, Japan) and Cu Kα radiation. The N 2 adsorption/desorption isotherm (3Flex, Micromeritics, Norcross, Georgia, USA) of each sample was obtained, and the surface area of each sample was determined using the Brunauer-Emmett-Teller (BET) method.
Time-of-flight SIMS (TOF-SIMS) analyses were performed for the three selected Fe-TiO 2 samples (Fe-TiO 2 -375, -525, and -750) using a TOF-SIMS 5 (ION-TOF GmbH, Muenster, Germany). The instrument was equipped with a Bi liquid metal ion gun aligned along the axes at 45 • to the surface of the sample. For SIMS analysis, Bi 3 + was used as the primary ion with an acceleration voltage of 30 keV, and the secondary positive ions emitted from the surface of the sample were analyzed with the TOF mass spectrometer. The pressed pellet of each sample was mounted on the sample stage of the chamber using carbon tape. A 500 µm × 500 µm area of the sample was analyzed, and each SIMS spectrum was obtained by accumulating signals from 30 scans. The mass scale of the SIMS experiment was calibrated using H + , CH 3 + , C 2 H 5 + , C 3 H 7 + , and C 4 H 9 + .

Photocatalytic Decomposition of Acetaldehyde
The photocatalytic removal of acetaldehyde on the surface of TiO 2 samples (TiO 2 and Fe-TiO 2 -300, -375, -450, -525, and -750) was studied under visible light irradiation using a batch-type reactor (Supporting information). A high vacuum chamber (base pressure of~2.0 × 10 −5 Torr, reactor volume of~5.3 L) was used as a reactor, and it was equipped with a convection vacuum gauge (SUPER BEE, InstruTech, Longmont, Colorado, USA), oil rotary (W2V40, WSA CO. LTD., Hwaseong, Korea), and turbomolecular pump (TMG 260P, PFEIFFER VACUUM Technology AG, Germany). The chamber was connected to the gas chromatograph (GC, HP 6890 series, Hewlett Packard) via 1 4 " SUS lines equipped with an HP-PLOT/Q column (19095P-Q04, 30 m × 0.530 mm × 40 µm, Agilent, Santa Clara, California, USA) and flame ionization detector (FID). Along the 1 4 " SUS lines, a diaphragm pump (DA70EEAC, flow rate of 7 L/min, YLKTECH) was installed between the GC and chamber to circulate the gas through the reactor. Polytetrafluoroethylene (PTFE) tubes ( 1 4 ") were used for the inlet and outlet lines of the diaphragm pump to accommodate the vibration of the diaphragm pump. The top side of the chamber was composed of a quartz view port, which allowed the efficient transmission of light from the light source located on the outside of the chamber.
The thin film of each photocatalyst was prepared by drop-casting 0.15 g of each sample on a SUS (stainless steel) plate (6 × 6 cm 2 ) [28], and it was mounted on the sample stage of the chamber. After the sample mounting, the reactor was pumped down until the reactor pressure reached~2.0 × 10 −5 Torr, and then the reactor was filled with a gas mixture (total pressure of 760 Torr) consisting of H 2 O, O 2 , dry air, and acetaldehyde diluted with N 2 (201 mol ppm). Each gas was injected into the reactor using a leak valve while monitoring its partial pressure with the vacuum gauge. The composition of the total gas mixture corresponded to ambient air containing 30 ppm of acetaldehyde with a relative humidity (RH) of 33% at 25 • C. After the reactor was filled with the gas mixture, the diaphragm pump was turned on to circulate the gas mixture through the reactor. Every 10 min, 50 µL of the gas mixture inside the reactor was injected into the GC to monitor the acetaldehyde and CO 2 concentration. Prior to the initiation of the GC analysis, the gas mixture in the reactor was pre-circulated by the diaphragm pump for 30 min in the dark to achieve a homogeneous gas composition throughout the reactor. The adsorption of acetaldehyde on each sample was monitored for 3 h in the dark. The blue LED lamp emitted visible light with a maximum intensity at 460 nm. The wavelength of the emitted light was in the range of 420-490 nm (Supporting information), and the distance between the blue LED and sample surface was 10 cm. The reactor temperature was kept at~24 • C throughout the experiments by controlling the temperature of the laboratory using an air conditioning system. The standard deviation of the reactor temperature over 10 h was 0.32 • C (24.3 • C ± 0.32 • C). We would like to mention that our testing method for the photocatalytic removal of acetaldehyde is different from the ISO standard method (ISO 22197-2:2011, details are provided in Supporting information). The main difference between our method and the ISO standard is that a batch-type reactor was used in this study, which is beneficial for examining the reaction rates compared to the flow-type reactor (ISO standard).
Additional photocatalytic experiments were performed with a selected Fe-TiO 2 sample (Fe-TiO 2 -375) to demonstrate the potential applications of the Fe-loaded TiO 2 sample. For this purpose, samples of Fe-TiO 2 -375 were mixed with a surface hardening agent (BUFFHARD, STEL corporation, Korea) and a paint (NOROO, Korea), and the resulting mixtures were loaded on a cement block (6 × 6 cm 2 ) and SUS plate (6 × 6 cm 2 ), respectively. The Fe-TiO 2 -375 content in each mixture (Fe-TiO 2 -375 mixed with surface hardening agents and Fe-TiO 2 -375 mixed with paint) was 10 wt%, and 1 g of each mixture was loaded. The photocatalytic experiment with these samples was carried out in the same manner, and under the same conditions, as those performed with the TiO 2 thin films on SUS plates.

Conclusions
Commercial rutile TiO 2 particles (200~300 nm) whose surfaces were covered with a thin alumina shell were modified by the tr-CVD of Fe-oxide (0.4 wt%) and subsequent thermal annealing in the temperature range of 300~750 • C. The particle size and crystallinity of the rutile TiO 2 particles were not changed by the modification. The photocatalytic performance of the TiO 2 in terms of acetaldehyde removal and CO 2 evolution under visible light irradiation was significantly enhanced by the modification. The enhancement effect of the Fe-loading was dependent on the annealing temperature (300~750 • C), and the effect was most pronounced at 375 • C. The enhancement effect decreased gradually as the annealing temperature increased above 375 • C, and in particular, the Fe-TiO 2 -525 and Fe-TiO 2 -750 samples showed significantly lower activity than the Fe-TiO 2 -375 sample.
Three representative Fe-TiO 2 samples (Fe-TiO 2 -375, -525, and -750) were further analyzed using various techniques, including two surface-sensitive methods (XPS and TOF-SIMS). The highertemperature annealing (525 and 750 • C) led to the stronger metal support interactions of TiO 2 with the structure of oxidized Fe. It resulted in the reduction of the oxidation states of the loaded Fe-oxide as well as a decrease in the surface Fe-oxide available to the adsorbed molecules at the higher annealing temperature (>375 • C). These effects resulted in the decreased charge separation efficiency of the Fe-TiO 2 sample under light irradiation, which led to the reduction of the enhancement effect of Fe-loading on the photocatalytic activity of the Fe-TiO 2 sample.
Furthermore, we demonstrated that the amounts of Fe-TiO 2 -375 sample produced in one sequential tr-CVD and annealing process can be easily increased up to 8 kg from 200 g. The mass-produced sample also exhibited photocatalytic activity regarding acetaldehyde removal under visible light that was comparable to that of the small-scale-produced Fe-TiO 2 -375 sample. In addition, the Fe-TiO 2 -375 particles exhibited noticeable photocatalytic activity regarding acetaldehyde removal under visible light after they were mixed with either paint or a surface-hardening agent and spread on the SUS or on a cement block. These results showed the high applicability of the Fe-modified rutile TiO 2 for applications for air purification in either outdoor or indoor environments.