Fabrication and Photocatalytic Properties of Electrospun Fe-Doped TiO2 Nanofibers Using Polyvinyl Pyrrolidone Precursors

For the removal of pollutants, a modified TiO2 photocatalyst is attracting attention. Fe-doped TiO2 nanofibers were prepared through a combination of electrospinning and calcination. Morphological characterization of the sample was conducted using field-emission scanning electron and transmission electron microscopy. The crystal structure of each sample was analyzed using high-resolution transmission electron microscopy, selected area electron diffraction, and Fast Fourier Transform imaging. The average diameter of the Fe-doped TiO2 nanofibers was measured to be 161.5 nm and that of the pure TiO2 nanofibers was 181.5 nm. The crystal phase when heat treated at 350 °C was anatase for TiO2 nanofibers and rutile for Fe-doped TiO2 nanofibers. The crystal phase of the TiO2 matrix was easily transitioned to rutile by Fe-doping. The photocatalytic performance of each sample was compared via the photodegradation of methylene blue and acid orange 7 under ultraviolet and visible light irradiation. In the Fe-doped TiO2 nanofibers, photodegradation rates of 38.3% and 27.9% were measured under UV irradiation and visible light, respectively. Although other catalysts were not activated, the photodegradation rate in the Fe-doped TiO2 nanofibers was 9.6% using acid orange 7 and visible light. For improved photocatalytic activity, it is necessary to study the concentration control of the Fe dopant.


Introduction
Since Honda and Fujishima reported photoelectrolysis photoelectrodes without an external power source in 1972 [1], TiO 2 has been drawing substantial attention, and many studies have been conducted to apply TiO 2 to various industrial fields, such as sensors [2][3][4][5], drug delivery systems [6][7][8], photocatalysts [9][10][11][12], and photoelectrodes [13][14][15]. TiO 2 has many attractive properties; among them, non-toxicity, strong durability, and excellent chemical stability are regarded as suitable photocatalyst properties for water purification. Despite these excellent properties of TiO 2 , some issues must be addressed to ensure its applicability in industry: electron-hole pair recombination and electromagnetic wave absorption bands that are too wide. In order to accomplish the reduction-oxidation cycle that is targeted for applications as a photocatalyst, electrons in the valence band are excited by absorbing electromagnetic waves that are incident from the outside, and the excited electrons and holes must then be transferred to the surface of the semiconductor to react

Electrospinning Process
First, 10.0 g of PVP was added to 60.0 g of EtOH and was stirred for 24 h using a magnetic stirrer to prepare a polymer solution. In another beaker, an aqueous solution of 7.5 g of TTIP, 10.0 g of ACAC, 2.5 g of Fe(NO 3 ) 2 ·9H 2 O, and 10.0 g of deionized water was stirred for 4 h. The polymer solution and the aqueous solution were then mixed and stirred for 2 h to prepare a reddish-brown precursor solution. In order to compare the photocatalytic activity caused by the Fe-doping, a precursor solution for the pure TiO 2 nanofibers was prepared by only removing the Fe(NO 3 ) 2 ·9H 2 O from the same composition.
The two prepared precursor solutions were loaded into a polypropylene syringe with a diameter of 15.56 mm and a volume of 12 mL and were then mounted on a syringe pump. The syringes were connected to a stainless steel adapter and a 23-gauge needle using polypropylene tubing. We subsequently applied high voltage to the nozzle adapter using a power supplier and a constant flow rate at the same time to induce jet spinning from the droplets of the solution to the grounded aluminum foil collector. The applied electrospinning conditions were as follows: distance between electrodes of 20 cm, an applied voltage of 20 kV, a flow rate of 1.0 mL/h, room temperature, and humidity less than 40%. The composite nanofibers collected on the aluminum foil during the electrospinning process were collected every 1 h and were dried at 60 • C in a dry oven. The calcination process was then conducted at 350 • C using a box furnace. The temperature ramping speed was 5 • C/min.

Characterization
The crystal structure and morphology of the calcined nanofibers were analyzed by a field emission scanning electron microscope (FE-SEM, Inspect F, FEI Korea Co., Ltd., Gyeonggi, Korea), a transmission electron microscope (TEM, JEM-2200FS, Jeol Co., Ltd., Tokyo, Japan), and energy dispersive X-ray spectroscopy (EDS). In addition, a photodegradation test was conducted using a photocatalyst/methylene blue aqueous solution and an ultraviolet light source, and the degradation rate was compared using an ultravioletvisible spectrophotometer (UV-Vis, G1103A, Agilent Co., Ltd., Santa Clara, CA, USA). To determine the calcination temperature, Thermogravimetric analysis (TGA) was conducted using a thermogravimetric analyzer (TGA, Q500, TA instruments, New Castle, DE, USA).

Photocatalytic Degradation Test
All steps of the photocatalytic degradation tests were conducted in a condition where the natural sunlight was blocked. Methylene blue, AO7, and distilled water were used to simulate contaminated water. Since it was necessary to apply the same contamination concentration of all test samples, 5 mg/L and 20 mg/L of methylene blue and AO7 aqueous solutions, respectively, were prepared in large-capacity bottles. As catalysts for photodegradation, P25, TiO 2 nanofibers, and Fe-doped TiO 2 nanofibers were used. To fabricate the mixture samples, 0.2 g of each catalyst and 20 mL of deionized water were added to a Pyrex beaker and were stirred for 30 min. Subsequently, 200 mL of Polymers 2021, 13, 2634 4 of 14 methylene blue and AO7 aqueous solution were added. The beaker was then wrapped with aluminum foil to completely block the incident light, and the mixture was stirred for 2 h. A photocatalytic degradation test of each mixture was conducted in a darkroom using UV (6W, 365 nm) and Vis (500 W, Xenon lamp) light sources for 3 h. During degradation, the catalyst filtered solution was sampled by using a syringe and a polyvinylidene fluoride syringe filter (PVDF filter, 0.2 µm, Whatman, Marlborough, MA, USA) every 30 min and was stored in a cuvette wrapped in aluminum foil. The irradiation distance between the light source and the sample was fixed at 10 cm, and the rotation speed of the stirrer was 240 rpm.

Results and Discussion
To determine the calcination temperature, TGA was conducted. Figure 1 shows the TGA curves of the TiO 2 nanofibers and the Fe-doped TiO 2 nanofibers. N 2 was used as a purge gas for analysis, and the ramping speed was 5 • C/min. Both samples showed similar thermal decomposition behavior. Until the temperature in the chamber reached 120 • C, the moisture and the solvents adsorbed on the samples evaporated, and the weight decreased. Crystallization started at 210 • C, and the carbonization and the thermal decomposition of the PVP started near 350 • C, causing rapid weight loss. Since N 2 was used as the purge gas, it showed thermal decomposition behavior and not combustion. The final weights of the TiO 2 and the Fe-doped TiO 2 nanofibers were 20.3 wt % and 22.3 wt %, respectively. The difference was due to the non-volatile components that increased with the addition of Fe nitrate in the same precursor solution.
Polymers 2021, 11, x FOR PEER REVIEW 4 of 13 mixture samples, 0.2 g of each catalyst and 20 mL of deionized water were added to a Pyrex beaker and were stirred for 30 min. Subsequently, 200 mL of methylene blue and AO7 aqueous solution were added. The beaker was then wrapped with aluminum foil to completely block the incident light, and the mixture was stirred for 2 h. A photocatalytic degradation test of each mixture was conducted in a darkroom using UV (6W, 365 nm) and Vis (500 W, Xenon lamp) light sources for 3 h. During degradation, the catalyst filtered solution was sampled by using a syringe and a polyvinylidene fluoride syringe filter (PVDF filter, 0.2 μm, Whatman, Marlborough, MA, USA) every 30 min and was stored in a cuvette wrapped in aluminum foil. The irradiation distance between the light source and the sample was fixed at 10 cm, and the rotation speed of the stirrer was 240 rpm.

Results and Discussion
To determine the calcination temperature, TGA was conducted. Figure 1 shows the TGA curves of the TiO2 nanofibers and the Fe-doped TiO2 nanofibers. N2 was used as a purge gas for analysis, and the ramping speed was 5 °C/min. Both samples showed similar thermal decomposition behavior. Until the temperature in the chamber reached 120°C, the moisture and the solvents adsorbed on the samples evaporated, and the weight decreased. Crystallization started at 210 °C, and the carbonization and the thermal decomposition of the PVP started near 350 °C, causing rapid weight loss. Since N2 was used as the purge gas, it showed thermal decomposition behavior and not combustion. The final weights of the TiO2 and the Fe-doped TiO2 nanofibers were 20.3 wt % and 22.3 wt %, respectively. The difference was due to the non-volatile components that increased with the addition of Fe nitrate in the same precursor solution.   Figure 2a,b are low and high magnification images of the TiO2 nanofibers, and c and d are images of the Fe-doped TiO2 nanofibers, respectively. It was confirmed that uniform fibrous structures were obtained, thus reflecting that the electrospinning process and calcination were conducted using appropriate conditions in which defects such as cracks or beads did not form.   Figure 2a,b are low and high magnification images of the TiO 2 nanofibers, and c and d are images of the Fe-doped TiO 2 nanofibers, respectively. It was confirmed that uniform fibrous structures were obtained, thus reflecting that the electrospinning process and calcination were conducted using appropriate conditions in which defects such as cracks or beads did not form. The average diameters of the nanofibers were measured in the obtained FE-SEM images. Figure 3 presents the diameter histogram of the nanofibers. The average diameters of the TiO 2 and Fe-doped TiO 2 nanofibers were measured to be 181.5 nm and 162.5 nm, respectively. The content of non-volatile components remaining after the calcination was higher in the Fe-doped TiO 2 nanofibers, but the average diameter was lower than that of the TiO 2 nanofibers. We considered that the stretching process was more vigorous because the dielectric constant of the solution was increased due to the added Fe(NO 3 ) 2 ·9H 2 O. The average diameters of the nanofibers were measured in the obtained FE-SEM images. Figure 3 presents the diameter histogram of the nanofibers. The average diameters of the TiO2 and Fe-doped TiO2 nanofibers were measured to be 181.5 nm and 162.5 nm, respectively. The content of non-volatile components remaining after the calcination was higher in the Fe-doped TiO2 nanofibers, but the average diameter was lower than that of the TiO2 nanofibers. We considered that the stretching process was more vigorous because the dielectric constant of the solution was increased due to the added Fe(NO3)2·9H2O. Images of the morphology of a single nanofiber were obtained using TEM. Figure  4a,b show bright-field images (BF) and high-angle annular dark-field images (HADDF) of the TiO2 nanofibers, respectively. Figure 4c,d are TEM images of Fe-doped TiO2 nanofibers. It can be seen that the inside and surface of the TiO2 nanofiber have a uniform single phase, but there are many particles of another phase on the Fe-doped TiO2 nanofibers. The diameters of the particles formed on the Fe-doped TiO2 nanofibers are distributed in the range of 23-29 nm. The reason for this morphological difference was confirmed in other Images of the morphology of a single nanofiber were obtained using TEM. Figure 4a,b show bright-field images (BF) and high-angle annular dark-field images (HADDF) of the TiO 2 nanofibers, respectively. Figure 4c,d are TEM images of Fe-doped TiO 2 nanofibers.
It can be seen that the inside and surface of the TiO 2 nanofiber have a uniform single phase, but there are many particles of another phase on the Fe-doped TiO 2 nanofibers. The diameters of the particles formed on the Fe-doped TiO 2 nanofibers are distributed in the range of 23-29 nm. The reason for this morphological difference was confirmed in other analysis results using TEM accessories.   Figure 5 shows the EDS mapping results of a single nanofiber sample. In both samples, titanium and oxygen are uniformly distributed throughout the whole nanofiber, but it can be seen that Fe was mainly located in another phase site that was discontinuously distributed on the Fe-doped TiO2 nanofiber. This means that the Fe dopant did not diffuse into the TiO2 lattice as an interstitial or substitutional atom during crystal growth and did not form a secondary phase. In order to quantify the ratio of each element, an additional EDS spectrum analysis using FE-SEM was conducted for the Fe-doped TiO2 nanofibers. The atomic ratios of Ti, O, and Fe were 29.18 at%, 63.45 at%, and 7.37 at%, respectively, which means that the TiO2 and gamma-ferrite phases were well separated and heterogeneous, as confirmed in the TEM image.  Figure 6a,b. Both Figure 6a,b show a ring-type diffraction pattern rather than a circle. This means that both samples were well crystallized. In the case of the amorphous phase, the pattern is diffused and shows a circular pattern. Figure 6a shows the diffraction pattern of the TiO2 nanofiber, and the (101) plane, which has a significant peak of the anatase phase, was identified. On  Figure 5 shows the EDS mapping results of a single nanofiber sample. In both samples, titanium and oxygen are uniformly distributed throughout the whole nanofiber, but it can be seen that Fe was mainly located in another phase site that was discontinuously distributed on the Fe-doped TiO 2 nanofiber. This means that the Fe dopant did not diffuse into the TiO 2 lattice as an interstitial or substitutional atom during crystal growth and did not form a secondary phase. In order to quantify the ratio of each element, an additional EDS spectrum analysis using FE-SEM was conducted for the Fe-doped TiO 2 nanofibers. The atomic ratios of Ti, O, and Fe were 29.18 at%, 63.45 at%, and 7.37 at%, respectively, which means that the TiO 2 and gamma-ferrite phases were well separated and heterogeneous, as confirmed in the TEM image.  Figure 5 shows the EDS mapping results of a single nanofiber sample. In both samples, titanium and oxygen are uniformly distributed throughout the whole nanofiber, but it can be seen that Fe was mainly located in another phase site that was discontinuously distributed on the Fe-doped TiO2 nanofiber. This means that the Fe dopant did not diffuse into the TiO2 lattice as an interstitial or substitutional atom during crystal growth and did not form a secondary phase. In order to quantify the ratio of each element, an additional EDS spectrum analysis using FE-SEM was conducted for the Fedoped TiO2 nanofibers. The atomic ratios of Ti, O, and Fe were 29.18 at%, 63.45 at%, and 7.37 at%, respectively, which means that the TiO2 and gamma-ferrite phases were well separated and heterogeneous, as confirmed in the TEM image.  Figure 6a,b. Both Figure 6a,b show a ring-type diffraction pattern rather than a circle. This means that both samples were well crystallized. In the case of the amorphous phase, the pattern is diffused and shows a circular pattern. Figure 6a shows the diffraction pattern of the TiO2 nanofiber, and the (101) plane, which has a significant peak of the anatase phase, was identified. On Selected area electron diffraction results for each fiber are shown in Figure 6a,b. Both Figure 6a,b show a ring-type diffraction pattern rather than a circle. This means that both samples were well crystallized. In the case of the amorphous phase, the pattern is diffused and shows a circular pattern. Figure 6a shows the diffraction pattern of the TiO 2 nanofiber, and the (101) plane, which has a significant peak of the anatase phase, was identified. On the other hand, Figure 6b shows the diffraction pattern of the Fe-doped TiO 2 nanofiber, and it was indexed to the rutile phase. In general, when the heat treatment temperature increased, the anatase phase transitions to rutile and a high-temperature stable phase, but the calcination temperature of the two samples was the same at 350 • C. This result is the same as in a previous study where Fe doping reduced the high-temperature stability of TiO 2 and prompted the transition to the rutile phase [64].   Figure 7 shows the high-resolution transmission electron microscopy (HRTEM) images and the FFT pattern of the secondary phase particles on the Fe-doped TiO2 nanofibers. The analysis results reveal that the phase of the doped particles is γ-ferrite (austenite) and that it has a face-centered cubic structure. A photodegradation test was conducted to compare the photocatalytic activity of the TiO2 nanofibers and the Fe-doped TiO2 nanofibers. Figures 8-11 show the UV-Vis absorbance of various aqueous solutions sampled every 30 min. Figure 7a plots the absorption spectra of the UV exposed pure methylene blue solution without a photocatalyst, and b, c, and d plot the spectra of the mixture containing P25, TiO2 nanofibers, and Fe-doped TiO2 nanofibers, respectively. Figure 9 shows that AO7 is used as a dye instead of methylene blue, and Figure 10,11 show the samples where a Xenon lamp was used as a light source instead of UV. As shown in the graphs, dye degradation was not observed in the blank solution, but it was confirmed that the absorbances of the solutions that dispersed each nanofiber and P25 significantly decreased over time. However, when comparing the degradation tendency at the wavelengths of 663 nm and 483 nm, corresponding with the significant absorption edge of methylene blue and AO7, it can be seen that the photocatalytic performance of the TiO2 nanofibers in the anatase phase is slightly superior to that of Fe-doped TiO2 nanofibers with the rutile phase. The photocatalytic degradation performance was affected by the crystal phase and metal ion doping, and the effect of the ana-    Figure 7 shows the high-resolution transmission electron microscopy (HRTEM) images and the FFT pattern of the secondary phase particles on the Fe-doped TiO2 nanofibers. The analysis results reveal that the phase of the doped particles is γ-ferrite (austenite) and that it has a face-centered cubic structure. A photodegradation test was conducted to compare the photocatalytic activity of the TiO2 nanofibers and the Fe-doped TiO2 nanofibers. Figures 8-11 show the UV-Vis absorbance of various aqueous solutions sampled every 30 min. Figure 7a plots the absorption spectra of the UV exposed pure methylene blue solution without a photocatalyst, and b, c, and d plot the spectra of the mixture containing P25, TiO2 nanofibers, and Fe-doped TiO2 nanofibers, respectively. Figure 9 shows that AO7 is used as a dye instead of methylene blue, and Figure 10,11 show the samples where a Xenon lamp was used as a light source instead of UV. As shown in the graphs, dye degradation was not observed in the blank solution, but it was confirmed that the absorbances of the solutions that dispersed each nanofiber and P25 significantly decreased over time. However, when comparing the degradation tendency at the wavelengths of 663 nm and 483 nm, corresponding with the significant absorption edge of methylene blue and AO7, it can be seen that the photocatalytic performance of the TiO2 nanofibers in the anatase phase is slightly superior to that of Fe-doped TiO2 nanofibers with the rutile phase. The photocatalytic degradation performance was affected by the crystal phase and metal ion doping, and the effect of the ana- A photodegradation test was conducted to compare the photocatalytic activity of the TiO 2 nanofibers and the Fe-doped TiO 2 nanofibers. Figures 8-11 show the UV-Vis absorbance of various aqueous solutions sampled every 30 min. Figure 7a plots the absorption spectra of the UV exposed pure methylene blue solution without a photocatalyst, and b, c, and d plot the spectra of the mixture containing P25, TiO 2 nanofibers, and Fedoped TiO 2 nanofibers, respectively. Figure 9 shows that AO7 is used as a dye instead of methylene blue, and Figures 10 and 11 show the samples where a Xenon lamp was used as a light source instead of UV. As shown in the graphs, dye degradation was not observed in the blank solution, but it was confirmed that the absorbances of the solutions that dispersed each nanofiber and P25 significantly decreased over time. However, when comparing the degradation tendency at the wavelengths of 663 nm and 483 nm, corresponding with the significant absorption edge of methylene blue and AO7, it can be seen that the photocatalytic performance of the TiO 2 nanofibers in the anatase phase is slightly superior to that of Fe-doped TiO 2 nanofibers with the rutile phase. The photocatalytic degradation performance was affected by the crystal phase and metal ion doping, and the effect of the anatase phase was more dominant than the effect of metal ion doping. The photocatalytic degradation performance of P25 was also observed to be superior to that of the nanofibers since P25 has a larger surface area than the TiO 2 and the Fe-doped TiO 2 nanofibers.          Figure 12 shows the photodegradation rate obtained by the absorbance changes of various catalysts every 30 min at the absorption edge. As shown in Figures 8-11, there was no change in the blank sample. In the Fe-doped TiO 2 nanofibers, photodegradation rates of 38.3% and 27.9% were measured under UV irradiation and visible light, respectively. In both the acid AO7 and the methylene blue, the Fe-doped TiO 2 showed lower photodegradation performance than the pure TiO 2 photocatalysts, but the difference was decreased under visible light irradiation. Although other catalysts were not activated, the photodegradation rate in Fe-doped TiO 2 nanofibers was 9.6% when using AO7 and visible light. We considered two reasons for this tendency. The first is the difference in photocatalytic activity due to the TiO 2 crystal structure. As confirmed by the TEM analysis, the phase of the TiO 2 matrix in the Fe-doped TiO 2 is rutile, and previous studies proved that the rutile phase has lower photocatalytic performance compared to the anatase phase [65][66][67]. The second reason is an excessive dopant concentration. If the concentration of the dopant is too high, the photocatalytic efficiency can be suppressed because electron trapping competes with the surface transfer reaction and reduces the sites on the TiO 2 surface where the holes can react with pollution. As seen in Figure 12d, for the case using visible light and acid orange 7, although other catalysts were not activated, the photodegradation rate in the Fe-doped TiO 2 nanofibers was 9.6%. The band gap narrowed by Fe-doping increased the photocatalytic performance under visible light. Further research is needed to confirm the special benefits of the narrow band electron structure of Fe-doped TiO 2 nanofibers in low-energy incident light and azo dyes.
surface where the holes can react with pollution. As seen in Figure 12d, for the case using visible light and acid orange 7, although other catalysts were not activated, the photodegradation rate in the Fe-doped TiO2 nanofibers was 9.6 %. The band gap narrowed by Fedoping increased the photocatalytic performance under visible light. Further research is needed to confirm the special benefits of the narrow band electron structure of Fe-doped TiO2 nanofibers in low-energy incident light and azo dyes.

Conclusions
TiO2 is an important material that is being discussed for the removal of organic pollutants. To solve the low photoelectric conversion efficiency issue, we considered Fe-doping among various elements. TiO2 and Fe-doped TiO2 nanofibers were fabricated by an

Conclusions
TiO 2 is an important material that is being discussed for the removal of organic pollutants. To solve the low photoelectric conversion efficiency issue, we considered Fedoping among various elements. TiO 2 and Fe-doped TiO 2 nanofibers were fabricated by an electrospinning process. Fe-doped TiO 2 nanofibers were obtained in a single step by adding a dopant element to a solution in a precursor. The average diameter of the nanofibers electrospun with the proposed precursor solution composition was measured to be 181.5 nm for TiO 2 and 162.5 nm for Fe-doped TiO 2 . When heat treated at 350 • C, the crystal phase was anatase for the TiO 2 nanofibers and rutile for the Fe-doped TiO 2 nanofibers. The Fe-doped TiO 2 nanofibers showed lower photocatalytic performance compared to the TiO 2 nanofibers because of the rutile crystal phase and excessive Fe concentration. The photocatalytic degradation performance was mainly influenced by the crystal phase and the metal ion doping, and the effect of the anatase phase was more dominant than the effect of the metal ion doping. The photodegradation rate of 9.6% in Fe-doped TiO 2 nanofibers using visible light suggests a research direction for photocatalytic materials for environmental applications.