Hollow TiO2/Poly (Vinyl Pyrrolidone) Fibers Obtained via Coaxial Electrospinning as Easy-to-Handle Photocatalysts for Effective Nitrogen Oxide Removal

Herein, we present a method for fabricating hollow TiO2 microfibers from Ti (OBu)4/poly (vinyl pyrrolidone) sol-gel precursors and their effects on denitrification as a photocatalyst for air purification. Various sizes of hollow TiO2 fibers were developed using coaxial electrospinning by controlling the core flow rate from 0 to 3 mL h−1. At higher flow rates, the wall layer was thinner, and outer and core diameters were larger. These features are correlated with physical properties, including specific surface area, average pore diameter, and crystalline structure. The increase in the core flow rate from 0 to 3 mL h−1 leads to a corresponding increase in the specific surface area from 1.81 to 3.95 µm and a decrease in the average pore diameter from 28.9 to 11.1 nm. Furthermore, the increased core flow rate results in a high anatase and rutile phase content in the structure. Herein, hollow TiO2 was produced with an approximately equal content of anatase/rutile phases with few impurities. A flow rate of 3 mL h−1 resulted in the highest specific surface area of 51.28 m2 g−1 and smallest pore diameter size of ~11 nm, offering more active sites at the fiber surface for nitrogen oxide removal of up to 66.2% from the atmosphere.


Introduction
Nitrogen oxides such as NO and NO 2 (which are commonly known as NO x ), are major air pollutants primarily emitted during low-grade fossil fuel combustion [1]. NO x causes severe environmental and human health issues, such as respiratory diseases, lung cancer, global warming, ozone production, acid rain, and smog [2]. According to the World Health Organization guidelines, the upper limit of the allowed NO concentration is 40 µg m −3 [3].
Recently, many researchers have reported the need for highly effective photocatalysts targeting NO x . Owing to the increased attention toward denitrification, titanium dioxide has been extensively used as a photocatalyst for air purification because of its relatively low cost, nontoxicity, easy handling, and long-term stability [3]. TiO 2 can generate electron-hole (e − /h + ) pairs at its surface in the presence of ultraviolet (UV) light energy higher than its band gap energy of~3.2 eV. The exited e − /h + pairs then react with H 2 O and O 2 molecules in the atmosphere, leading to the formation of hydroxy radicals and superoxide oxygen via reactive oxygen species (ROS) [3,4]. Subsequently, these molecules can react with nitrogen oxides to form nitrates via photocatalytic oxidation.
However, brittleness and handling issues of TiO 2 nanoparticles are the major factors that restrict their applications as photocatalysts for air purification [5]. Therefore, it remains a challenge to develop a facile fabrication method to generate mesoporous TiO 2 as an effective air pollutant photocatalyst.
Electrospinning is an effective fabrication method to address the aforementioned limitations of TiO 2 nanoparticles because it provides a simple method to prepare ultrathin TiO 2 fibers with diameters ranging from several nano-to micrometers. Polymer, organic, inorganic, and hybrid core/shell or hollow materials have been fabricated via coaxial 2 of 11 electrospinning, in which a spinneret consisting of two coaxial capillaries form a core-sheath structure [6]. Controlling the variables using coaxial electrospinning can modify the architecture and multifunction of fibers.
Core-sheath or hollow materials have attracted extensive attention in recent years because of their unique heterogeneous or hollow structures that offer high surface-tovolume ratios and core loading spaces, rendering them useful for air purification [6]. For example, Chen et al. reported a multifluidic coaxial electrospinning method to fabricate photocatalytic TiO 2 nanowire-in-microtube structured nanofibers [7]. The mesoporous TiO 2 hollow nanofibers, which were fabricated via a facile single-capillary electrospinning technique, showed 99.5% degradation of Rhodamine B within 60 min, suggesting their promising application as efficient photocatalysts [8]. Additionally, polyaniline/TiO 2 bilayer microtubes with an average diameter of 200 nm were fabricated for photocatalysis [9]. Polyvinylidene fluoride (PVDF)-TiO 2 composite hollow fibers were prepared using the sol-gel method [10]. Furthermore, fluorinated titania-silica/PVDF hollow fibers were developed for CO 2 removal [11]. Electrospinning was used to prepare PVDF/poly dimethylsiloxane composites with TiO 2 nanoparticles (0.5 and 1 wt.%) for NOx adsorption [12].
Herein, coaxial electrospinning was used under four types of processing conditions to develop hollow TiO 2 fibers as photocatalysts for air purification. An ideal hollow architecture with a micro-nanoporous structure was achieved for excellent photocatalytic performance. Subsequently, physical properties, including morphology, specific surface area, pore diameter, and crystalline phase, of four types of TiO 2 fibers were evaluated. Moreover, their photocatalytic performance was evaluated for denitrification under UV irradiation. Additionally, the physical properties of hollow TiO 2 fibers were correlated to the photocatalytic performance. This study provides a comprehensive understanding and efficient design approach toward hollow TiO 2 fibers for various applications, such as surface coating, electronics, biomedicine, sensing, and water and air purification.

Preparation of Core-Sheath Sol-Gel
PVP was dissolved in a mixture of ethanol solution and acetic acid (1:8:1 w/w/w) in a flask, followed by the gradual addition of 40 wt.% Ti (OBu) 4 solution under stirring at room temperature for 4 h to form a homogeneous sol for outer sheath material, such as the Ti (OBu) 4 sol. For the core material, paraffin oil solution was prepared by dissolving 5 wt.% Tween 80 in 2 wt.% deionized water. Subsequently, the emulsion was stirred for 3 h at room temperature.

Fabrication of Hollow TiO 2 Fibers
The coaxial electrospinning experimental setup is illustrated in Figure 1. The spinneret was assembled using coaxial stainless steel with two capillaries. Typically, the outer and inner diameters were 3.0 and 2.5 mm and 0.6 and 0.37 mm, respectively. The Ti (OBu) 4 sol synthesized previously was injected into the outer capillaries at a flow rate of 8 mL h −1 , and the paraffin oil emulsion was simultaneously fed into the inner capillary at 0, 1, 2, or 3 mL h −1 . The work distance between the spinneret and collector was 15-20 cm, and the work voltage was 25-30 kV. After coaxial electrospinning, the collected fibers were calcinated at 550 • C for 1 h at a heating rate of 1 • C min −1 . The core was removed after calcination treatment (MF2-12GF, JeioTech, Daegeon, Republic of Korea) to obtain hollow TiO 2 fibers [13].

Chemical Structures and Physical Properties of TiO2 Fibers
The morphologies and average outer or core diameters of TiO2@C1, TiO2@C2, TiO2@C3, and TiO2@C0 were analyzed via scanning electron microscopy (SEM, Hitachi, Tokyo, Japan). For cross sections, samples were immersed in liquid nitrogen and then cut with a razor blade. The N2 adsorption and desorption analyses were performed (BEL-SORP-Mini II, Osaka, Japan) at 77 K. The specific surface areas were calculated from the obtained data using the Brunauer-Emmett-Teller (BET) method. The total pore volume was estimated from the amount adsorbed at a P/P0 of 0-0.99. The average pore diameter was derived using the Barret-Joyner-Halenda (BJH) model. The specific surface areas and pore diameters of TiO2@C1, TiO2@C2, TiO2@C3, and TiO2@C0 were evaluated using the BELSORP analysis software. On applying a suitable high voltage to the coaxial spinneret, the conductive outer fluids (Ti (OBu) 4 /PVP sol-gel) were elongated owing to an electrostatic force, while the inner fluids (paraffin oil/Tween 20/water) were subsequently stretched because of shear forces. The outer Ti (OBu) 4 sol-gel solidified easily during the coaxial electrospinning process because of the rapid evaporation of the ethanol solvent and the hydrolysis of metal alkoxide at appropriate humidity [7]. As shown in Figure 1, water is emulsified in paraffin oil to accelerate the hydrolysis and solidification of the core material during reactions [7]. If the core was selectively removed by calcination, then it would leave a vacant space to develop hollow TiO 2 fibers. Electrospinning was used to fabricate hollow and solid (non-hollow) TiO 2 fibers under different conditions and they were labelled with relevant codes, as listed in Table 1.

Chemical Structures and Physical Properties of TiO 2 Fibers
The morphologies and average outer or core diameters of TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C0 were analyzed via scanning electron microscopy (SEM, Hitachi, Tokyo, Japan). For cross sections, samples were immersed in liquid nitrogen and then cut with a razor blade. The N 2 adsorption and desorption analyses were performed (BELSORP-Mini II, Osaka, Japan) at 77 K. The specific surface areas were calculated from the obtained data using the Brunauer-Emmett-Teller (BET) method. The total pore volume was estimated from the amount adsorbed at a P/P 0 of 0-0.99. The average pore diameter was derived using the Barret-Joyner-Halenda (BJH) model. The specific surface areas and pore diameters of TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C0 were evaluated using the BELSORP analysis software.

Characterization of Synthesized TiO 2 Fibers
X-ray diffraction (XRD, D8 Advance, Bruker, Billerica, MA, USA) was used to analyze the crystalline phases of synthesized TiO 2 fibers. The diffractometer was operated in the reflection mode with Cu-K radiation (35 kV; 30 mA) and a diffracted beam monochromator, using a step scan mode with a step of 0.075 • at 4 s per step. Diffraction patterns of both anatase and rutile TiO 2 fibers at different flow rates were compared with a reference using the JCPDS database.

Photocatalytic Performance of TiO 2 Fibers for NO Removal
The photocatalytic activities of TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C0 for the removal of NO gas molecules were assessed based on ISO 22197-1:2016. Hollow TiO 2 microfiber (TiO 2 @C1, TiO 2 @C2, and TiO 2 @C3) or solid (TiO 2 @C0) TiO 2 fiber (5 cm × 10 cm) samples were placed in the middle of two plain glasses (5 cm × 10 cm) of non-photocatalytic blank samples in a photoreactor. Each sample was irradiated with UV-A light (10 W m −2 ) using a UV lamp system, consisting of two 6 W lamps, placed over the photoreactor, with an emission peak at 365 nm. An NO x analyzer (T-API, T200, San Diego, CA, USA) was used to measure nitrate concentrations at 1 min intervals. Under UV light irradiation, the NO gas flowed at a rate of 3 L min −1 containing 1 ppm v of NO in air with 50% relative humidity at 25 • C. The concentration of NO in the outlet stream was monitored for 20 min before the light was switched on and afterward during the UV irradiation for 60 min.

Results and Discussion
3.1. Morphology of TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C0 Figures 2 and 3 show the synthesized TiO 2 fibers before and after calcination, respectively. The synthesized TiO 2 fibers have smooth surfaces without a hole inside. After calcination of the synthesized TiO 2 fibers, they display a reduction in the average fiber diameter. The removal of the polymer (PVP and paraffin oil) by calcination shows a decrease in fiber diameter from about a few hundred nanometers, as shown in the SEM images in Figure 3. The average diameters of the core and sheath as well as the wall thickness of TiO 2 fibers were measured at randomly selected areas of SEM images. The hollow core corresponds to the vacancy of the core paraffin oil emulsion. In Table 2, the TiO 2 @C1 has a core and sheath with average diameters of 1.45 and 1.81 µm, respectively, with a wall thickness of 0.26 µm. For TiO 2 @C2, the average diameters of the core and sheath are 1.80 and 2.09 µm, respectively, with 0.38 µm wall thickness. TiO 2 @C3 displays core and sheath diameters of 2.83 and 3.94 µm, respectively, with a wall thickness of 0.53 µm, while TiO 2 @C0 has an average fiber diameter of 3.95 µm without the core. Increasing the flow rate of the core paraffin fluid from 1 to 3 mL h −1 increased the core diameter from 0.26 to 0.53 µm and the outer diameter from 1.81 and 3.94 µm. Di et al. reported that hollow fibers electrospun with increased inner liquid flow rates of 1.0, 3.0, and 5.0 mL h−1 showed increased inner diameter from 0.9 to 1.90 µm [14]. Another study reported that decreasing the flow rate resulted in thicker hollow fibers because there was deficient outer solution to enclose the outer shell continuously [15]. Furthermore, the core flow rate was very slow, resulting in the outer layer being too thick due to less shear stress [13]. In contrast, a high core flow rate would make the wall thinner and decrease the stability of the electrospinning process because there may not be sufficient outer fluid to effectively capture the inner contents, leading to their leakage. Therefore, an appropriate flow rate is a critical variable for fabricating hollow fibers.
The flow rate of the core is an important factor in modifying the morphology of hollow TiO 2 fibers, in terms of the core and outer diameter size. The higher the flow rate, the thinner the wall layer and the larger the core and outer diameters [16].   3.2. BET Analysis of TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C0 BET and BJH plots of TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C0 at 77 K are presented in Figure 4a,b. TiO 2 @C2 and TiO 2 @C3 showed a relatively narrow distribution of TiO 2 @C0, with values ranging from 5 to 30 nm. The order of the overall specific surface area is as follows: TiO 2 @C3 > TiO 2 @C2 > TiO 2 @C1 > TiO 2 @C0. The specific surface area of TiO 2 @C3 was approximately 51.2 m 2 g −1 . The increase in the core flow rate from 0 to 3 mL h −1 results in a corresponding increase in the specific surface area from 16.01 to 51.28 m 2 g −1 and a decrease in the average pore diameter from 28.9 to 11.1 nm, as listed in Table 2. TiO 2 @C0 has the lowest specific surface area of 16.01 m 2 g −1 and the highest average pore diameter size of 28.9 nm owing to the solid (non-hollow) structure of TiO 2 @C0. Previous studies have reported various advantages of hollow TiO 2 fibers owing to their mesoporous walls and unique hierarchical pore structure [8]. This hollow structure helps to improve the efficiency of air mass transport, which leads to a larger specific surface area and porosity [17,18].
Hou et al. developed hollow TiO 2 fibers with a BET surface area of~27.2 m 2 g −1 and an average pore diameter of 38 nm. However, our fabrication methods resulted in significantly higher specific surface area (~2 times higher than the aforementioned value), which provides more active surface sites for the adsorption of reactive molecules, resulting in more prominent photocatalytic effects [19]. The results suggest that hollow TiO 2 fibers (TiO 2 @C3) prepared via the proposed method might exhibit good photocatalytic activities. Polymers 2022, 14, x FOR PEER REVIEW 6 of 11

BET Analysis of TiO2@C1, TiO2@C2, TiO2@C3, and TiO2@C0
BET and BJH plots of TiO2@C1, TiO2@C2, TiO2@C3, and TiO2@C0 at 77 K are presented in Figure 4a,b. TiO2@C2 and TiO2@C3 showed a relatively narrow distribution of TiO2@C0, with values ranging from 5 to 30 nm. The order of the overall specific surface area is as follows: TiO2@C3 > TiO2@C2 > TiO2@C1 > TiO2@C0. The specific surface area of  TiO2@C3 was approximately 51.2 m 2 g −1 . The increase in the core flow rate from 0 to 3 h −1 results in a corresponding increase in the specific surface area from 16.01 to 51.2 g −1 and a decrease in the average pore diameter from 28.9 to 11.1 nm, as listed in Tab TiO2@C0 has the lowest specific surface area of 16.01 m 2 g −1 and the highest average diameter size of 28.9 nm owing to the solid (non-hollow) structure of TiO2@C0. Prev studies have reported various advantages of hollow TiO2 fibers owing to their mes rous walls and unique hierarchical pore structure [8]. This hollow structure helps to prove the efficiency of air mass transport, which leads to a larger specific surface area porosity [17,18]. Hou et al. developed hollow TiO2 fibers with a BET surface area of ~27.2 m 2 g −1 an average pore diameter of 38 nm. However, our fabrication methods resulted in si icantly higher specific surface area (~2 times higher than the aforementioned value), w provides more active surface sites for the adsorption of reactive molecules, resultin more prominent photocatalytic effects [19]. The results suggest that hollow TiO2 f (TiO2@C3) prepared via the proposed method might exhibit good photocatalytic activ

XRD Analysis of TiO2@C1, TiO2@C2, TiO2@C3, and TiO2@C0
After calcination, TiO2@C1, TiO2@C2, TiO2@C3, and TiO2@C0 contained ana brookite, and rutile phases. Generally, TiO2 has three representative crystalline ph anatase, brookite, and rutile phases. TiO2 is most likely to be a mixture of the aforem tioned phases rather than the pure anatase, brookite, or rutile structure; thus, quantit analysis is significantly important for analyzing the variations in photocatalytic activ Figure 5 and Table 3 [20]. TiO2@C3 and TiO2@C0 displayed an increase in the rutile phase content and sho diffraction peaks (yellow dot lines) at 27°, 36°, and 55°, corresponding to the crysta region of TiO2 [21]. The XRD results (Table 3) show that crystalline TiO2@C1 consis 53.5% of the anatase phase, 7.2% of the rutile phase, and 39.3% of the brookite ph TiO2@C2 and TiO2@C3 showed an increased rutile phase content of 11.8% and 24.3% spectively, while TiO2@C0 had 60.1% of the anatase phase and 9.5% of the rutile ph XRD data show that the increased core flow rate results in the higher composition o anatase form and lower content of the rutile phase in the structure. TiO2 has three m stable polymorphs, namely, anatase, rutile, and brookite. Among them, anatase struc generally shows the highest photocatalytic activity, with certain crystallographic pl Figure 4. (a) BET plots for N 2 adsorption/desorption isotherm, and (b) pore diameter distribution plots obtained using BJH plots of TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C0.

XRD Analysis of TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C0
After calcination, TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C0 contained anatase, brookite, and rutile phases. Generally, TiO 2 has three representative crystalline phases, anatase, brookite, and rutile phases. TiO 2 is most likely to be a mixture of the aforementioned phases rather than the pure anatase, brookite, or rutile structure; thus, quantitative analysis is significantly important for analyzing the variations in photocatalytic activities. Figure 5 and Table 3 present the XRD data of developed TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C0 in the 2θ range of 20 • to 80 • according to standard JCPDS card No. 21-1272. The anatase reflections dominated the reflection patterns, while rutile was present as well. All the diffraction peaks (blue dot lines) at 25 [20]. TiO 2 @C3 and TiO 2 @C0 displayed an increase in the rutile phase content and showed diffraction peaks (yellow dot lines) at 27 • , 36 • , and 55 • , corresponding to the crystalline region of TiO 2 [21]. The XRD results (Table 3) show that crystalline TiO 2 @C1 consists of 53.5% of the anatase phase, 7.2% of the rutile phase, and 39.3% of the brookite phase. TiO 2 @C2 and TiO 2 @C3 showed an increased rutile phase content of 11.8% and 24.3%, respectively, while TiO 2 @C0 had 60.1% of the anatase phase and 9.5% of the rutile phase. XRD data show that the increased core flow rate results in the higher composition of the anatase form and lower content of the rutile phase in the structure. TiO 2 has three major stable polymorphs, namely, anatase, rutile, and brookite. Among them, anatase structure generally shows the highest photocatalytic activity, with certain crystallographic planes of anatase being particularly reactive [22]. Anatase transforms to the brookite or rutile phase at temperatures below 600 • . A higher transformation rate of anatase to brookite, as compared with that of anatase to rutile, is observed and explained by the increased surface area of TiO 2 @C3, resulting from their aggregation, which act as sites for the rutile nucleation [23]. Upon heating, which is usually accompanied by a coarsening of the crystals, the crystal growth leads to alterations of phase stability [23]. The band gaps of anatase, rutile, and brookite are 2.13, 1.86, and 2.38 eV, respectively. Anatase is an indirect band gap semiconductor. In contrast, both rutile and brookite belong to the direct band gap semicon-ductor category [24]. It has been reported that a similar composition mixture of anatase and rutile crystalline phases of TiO 2 exhibited significantly higher photocatalytic activity than the pure anatase phase [8,25]. The primary reason is the enhanced charge transfer triggered by the energy gap between the band edges of crystalline phases of TiO 2 [8,26]. These photocatalysts show varying anatase/rutile ratios depending on the conditions of the core flow rate. Furthermore, the developed TiO 2 @C3 consists of a mixture of anatase/rutile (~1:1 ratio) crystalline phases, thereby exhibiting better photocatalytic performance than TiO 2 @C1, TiO 2 @C2, and TiO 2 @C0, as evident from our results. and rutile crystalline phases of TiO2 exhibited significantly higher photocatalytic activity than the pure anatase phase [8,25]. The primary reason is the enhanced charge transfer triggered by the energy gap between the band edges of crystalline phases of TiO2 [8,26]. These photocatalysts show varying anatase/rutile ratios depending on the conditions of the core flow rate. Furthermore, the developed TiO2@C3 consists of a mixture of anatase/rutile (~1:1 ratio) crystalline phases, thereby exhibiting better photocatalytic performance than TiO2@C1, TiO2@C2, and TiO2@C0, as evident from our results.   Figure 6 shows the NO removal performance of TiO2@C1, TiO2@C2, TiO2@C3, and TiO2@C0, which changes significantly with on-off UV lamp irradiation. To confirm the effectiveness of UV irradiation for triggering photocatalytic reaction, denitrification experiments were performed under UV irradiation (turning on UV lamp) and darkness (turning off UV lamp) conditions. The results show that the denitrification efficiency in the dark was ~0.01% during the initial 10 min. Moreover, under UV lamp irradiation, the Figure 5. XRD spectra of TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C0.  Figure 6 shows the NO removal performance of TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C0, which changes significantly with on-off UV lamp irradiation. To confirm the effectiveness of UV irradiation for triggering photocatalytic reaction, denitrification experiments were performed under UV irradiation (turning on UV lamp) and darkness (turning off UV lamp) conditions. The results show that the denitrification efficiency in the dark was 0.01% during the initial 10 min. Moreover, under UV lamp irradiation, the NO removal (%) for TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C0 showed a rapid increase up to 66.2%, 56.8%, 33.5%, and 31.2%, respectively. When the UV lamp was turned off, the removal efficiency of NO decreased rapidly in the dark, reaching 1 ppm of NO concentration at 70 min and maintaining the NO concentration constant without photocatalytic activity. This implies a significant photocatalytic effect of UV irradiation on NO removal. 58.65% of rutile and anatase phases with minimum impurities, resulting in enhanced charge transfer owing to the energy gap between the two crystalline phases [8,26]. The highest BET specific surface area of 51.28 m 2 g −1 leads to more active sites and higher capacity at the hollow TiO2 fiber surface for NO adsorption, resulting in more prominent photocatalytic effects. Amano et al. reported that the surface area and photocatalytic activity or capacity for reactant adsorption had a proportional relationship [27]. Therefore, TiO2@C3 is an efficient photocatalyst for air purification.

Conclusions
Herein, we demonstrated a facile method to fabricate easy-to-handle hollow TiO2 fibers as photocatalysts for air purification, particularly for denitrification. The hollow TiO2 fibers were fabricated using different core flow rates, ranging from 0 to 3 mL h −1 , with a fixed outer flow rate of 8 mL h −1 . The obtained TiO2 fibers showed mesoporous walls and a unique hierarchical pore structure. They were composed of mixed anatase and rutile phases with distinct hollow features. However, depending on the core flow rate, the fiber morphology changed in terms of the outer and core diameter sizes, wall thickness, BET specific surface area, and crystalline phase content as well as photocatalytic activity. Results show that TiO2@C3 had the largest core and outer diameters that improve the air Figure 6. Photocatalytic performance of TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C4 for NO removal.
The degradation of NO gas molecules (1 ppm) was achieved in 60 min under light irradiation in the presence of TiO 2 @C1, TiO 2 @C2, TiO 2 @C3, and TiO 2 @C0. The highest efficiency was observed for the photocatalyst TiO 2 @C3, containing a mixture of 41.12% and 58.65% of rutile and anatase phases with minimum impurities, resulting in enhanced charge transfer owing to the energy gap between the two crystalline phases [8,26]. The highest BET specific surface area of 51.28 m 2 g −1 leads to more active sites and higher capacity at the hollow TiO 2 fiber surface for NO adsorption, resulting in more prominent photocatalytic effects. Amano et al. reported that the surface area and photocatalytic activity or capacity for reactant adsorption had a proportional relationship [27]. Therefore, TiO 2 @C3 is an efficient photocatalyst for air purification.

Conclusions
Herein, we demonstrated a facile method to fabricate easy-to-handle hollow TiO 2 fibers as photocatalysts for air purification, particularly for denitrification. The hollow TiO 2 fibers were fabricated using different core flow rates, ranging from 0 to 3 mL h −1 , with a fixed outer flow rate of 8 mL h −1 . The obtained TiO 2 fibers showed mesoporous walls and a unique hierarchical pore structure. They were composed of mixed anatase and rutile phases with distinct hollow features. However, depending on the core flow rate, the fiber morphology changed in terms of the outer and core diameter sizes, wall thickness, BET specific surface area, and crystalline phase content as well as photocatalytic activity. Results show that TiO 2 @C3 had the largest core and outer diameters that improve the air mass transport. Furthermore, among the prepared photocatalysts, TiO 2 @C3 exhibited a balanced anatase/rutile phase ratio as well as the highest BET specific surface area of 51.28 m 2 g −1 and smallest pore size of 11.1 nm, offering more active sites for NO removal and resulting in the most effective denitrification up to 66.2%.
This study provides a comprehensive understanding and an ideal design for the fabrication of hollow TiO 2 fibers in terms of the crystalline phase composition as well as hollow and mesoporous structures for the desired photocatalytic performance. Our methodology presents a facile approach to fabricate hollow architecture for various applications including surface coating, electronics, biomedicine, sensing, and water and air purification.
Author Contributions: For this research article, J.K. confirms solo responsibility: conceptualization, methodology, validation, formal analysis, investigation, resources, data curation, writing-original draft preparation, writing-review and editing, visualization, project administration, and funding acquisition. The author has read and agreed to the published version of the manuscript.
Funding: This research was supported by the Korean Institute of Industrial Technology Convergence (PEO22020) and the National Research Foundation of Republic of Korea (NRF-2021R1F1A1061200).
Institutional Review Board Statement: Not applicable.