Potential of Nb2O5 nanobers in photocatalytic response for organic pollutants remediation

Due to the pollution caused by different organic pollutants, various photocatalytic nanomaterials for environmental remediation have been promoted. In this study, Nb 2 O 5 nanobers were obtained by electrospinning technique, presenting controlled crystallinity and high specic surface area to improve the photoactivity response. The structural characterization indicated Nb 2 O 5 nanobres with orthorhombic phase formation, and the photoluminescence measurements showed different energy levels contributing to the electronic transition events. The nanobers with a bandgap up to 3.6 eV were applied to photocatalysis of dyes [Rhodamine B (RhB) or Methylene Blue (MB)], and Prozac®, listed as an emergent pollutant. In the optimized condition (pH = 9), the RhB and MB photocatalysis was 59% and 93% more ecient than photolysis due to ζ = − 50 ± 5 mV for EtOH_550 sample that increased the interaction with MB (cationic) compared to RhB unprotonated (pKa = 3.7). Therefore, Prozac® (pKa = 10.7) was selected due to protonated form at pH = 9 and showed 68% ±1 adsorption in 30 min for EtOH_550. The Prozac® photocatalytic degradation under UV light irradiation was up to 17% higher than the photolytic degradation. The formation of hydroxyl radicals in the photocatalytic system (EtOH_550) was proven by the Coumarine probe assay, corroborating with the greater amount of α-[2-(Methylamino)ethyl]benzylalcohol (MAEB), a by-product obtained after Prozac® oxidation. Additionally, the material achieved specic catalytic activity for the different organic compounds (RhB, MB, or Prozac®), showing that only using dyes may not be ideal to conclude the great material applicability in environmental remediation studies. Therefore, Nb 2 O 5 nanobers were ecient for the degradation of three different pollutants under UV light, proving to be a viable alternative for environmental remediation.


Introduction
Pollution resulting from the inappropriate disposal of the varied industrial waste types has been the main route for contamination of environmental ecosystems (Kostich et al. 2014;Pinheiro et al. 2016;Lee et al. 2017). With the advancement of ne chemistry, new molecules have been created to meet society's demands in the elds of food (Kuenemann et al. 2017), health (Farooqui et al. 2018), agriculture (Sanaullah et al. 2020), and social well-being. Consequently, these new molecules have been increasingly introduced into ecosystems as contaminants, affecting environmental quality and increasing human health risks (Gouvêa et al. 2018;Grenni et al. 2018). Among these new molecules with high polluting potential, stand out the class of dyes, pesticides, and pharmaceutical products (Lapworth et al. 2012;Malafatti et al. 2020). Considering that environmental legislation does not keep up the speed with which these new molecules are created, this type of pollutant regulation is limited or non-existent in many countries (Petrie et al. 2015;Gorito et al. 2017). Watchful of this new reality, the scienti c community has found in nanotechnology the appropriate tools to minimize the impacts of this type of pollution (Khan and Tahir 2019;Paris et al. 2020). In this sense, advanced oxidation technologies, especially heterogeneous photocatalysis, have been the most e cient tools to remove these emerging contaminants from environmental ecosystems (Védrine 2019).

Page 3/30
The photocatalytic mechanism responsible for the degradation of organic pollutants is widely known. It can be summarized in the ability that materials with semiconductor properties exhibit to generate oxidizing (h + ) and reducing (e − ) sites on their surface (Rauf and Ashraf 2009;Gurgel et al. 2011).
Therefore, when absorbing a photon, these materials promote the electron from the valence band (VB) to the conduction band (CB), giving rise to the electron/hole pair responsible for triggering the redox reactions (Oliveira et al. 2013;Shen et al. 2019). According to the literature, photocatalytic success is associated with the electronic photoexcitation e ciency, recombination processes decrease, active surface increase, the interaction between molecule/catalyst, and homogeneity of the catalyst in the reaction system (Fang et al. 2017;Moreira et al. 2020a). This property is also broadly correlated with the ceramic material crystalline structures (Paris et al. 2010), which can be in uenced by the synthesis method (Leite et al. 2000) and heat treatment temperature (Alves et al. 2009). Therefore, TiO 2 application as a photocatalyst has been extensively investigated for environmental remediation, and nowadays, its physical-chemical modi cation has been proposed to optimize these processes (Shen et al. 2019).
However, the search for alternative materials to TiO 2 has promoted signi cant advances in the eld of heterogeneous photocatalysis, where semiconductors based on niobium oxide have been gaining prominence (Nico et al. 2016).
Nb 2 O 5 as a photocatalyst for environmental remediation is still little reported compared to other semiconductors such as TiO 2 and ZnO. However, its photoactivity has shown to be promising (Raba et al. 2015;Falk et al. 2017). The structural properties of Nb 2 O 5 favor chemical oxidation due to its ability to diffuse surface oxygen to form the reactive species • O 2 − and • OH (Nico et al. 2016). Moreover, the Nb 2 O 5 surface characteristics highlight its adsorptive properties, favoring chemical conversions (Zhuang et al. 2016). By comparing the TiO 2 and Nb 2 O 5 photoactivity, the niobium oxide structure electronic transitions are more varied, selecting its application in photocatalytic processes (Nb et al. 2018).
Thus, different synthesis methods have been investigated to obtain Nb 2 O 5 (Nogueira et al. 2017;Falk et al. 2017;Silva et al. 2019). The precipitation method by dissolution in an aluminum reduction system supplied Nb 2 O 5 in rod forms, and the photocatalytic properties for H 2(g) production was evidenced (Zhao et al. 2016). Additionally, the evaporation-induced self-assembly (EISA) (Hashemzadeh et al. 2014), oxidant peroxide (Lopes et al. 2014), microwave-assisted (Falk et al. 2017), and polymeric precursors (Raba et al. 2015) methods were also applied to obtain Nb 2 O 5 as a photocatalyst. When used for photocatalytic degradation of different organic pollutants, the e ciency is in the order of 45% for atrazine in 30 min (Lopes et al. 2014), > 95% for MB in 50 min (Shao et al. 2014), and > 45% for MB in 120 min (Falk et al. 2017). As an unconventional synthesis method, the electrospinning technique was applied to synthesize Nb 2 O 5 nano bers, exhibiting good performance to produce nanomaterials with structural defects that favor electromigration (Grishin et al. 2013). This method also had Nb 2 O 5 with small particle diameters (21 to 37 nm) and high speci c surface area (Leindecker et al. 2014). The Nb 2 O 5 nano ber photoactivity response was evaluated to the methyl orange degradation, reaching 62% in 3 h (Qi et al. 2013).
Under speci c heat treatment conditions, the textural properties of Nb 2 O 5 bers can be controlled to become optimal adsorbents for organic molecules (Ferreira et al. 2019). The adsorption process between nanomaterials and different compounds is responsible for optimizing the interaction between the nanomaterial and a speci c molecule (Pi et al. 2018). The literature has already shown the different compound chemical nature in uences in the photocatalytic process e ciency since their protonated/deprotonated form guides the material/molecule interaction (Bian et al. 2011). Thus, anionic/cationic dyes or molecules with different lipophilicity degrees can show distinct e ciency for the same catalyst during the photocatalytic process (Wang et al. , 2014. This aspect implies that the material photoactivity cannot be discussed just because of their physical-chemical properties but must be associated with the compound chemical nature to be degraded. Materials that exhibit surfaces with a high density of negative charge can optimally disperse the reaction system, increasing the active area and, consequently, the degradation e ciency (Pi et al. 2018). However, these same materials can repel negatively-charged molecules, decrease surface interaction, and impair photocatalytic e ciency (Bian et al. 2011). Therefore, in the photocatalysis eld, deepening investigations about the pollutant chemical nature and its interaction with nanomaterial can contribute to the optimization of photocatalytic processes.
In this study, Nb 2 O 5 nano bers were e ciently obtained after optimizing the electrospinning method proposed in the literature (Leindecker et al. 2014). The nano ber photocatalytic activities were monitored for the RhB and MB dyes and the Prozac® drug (emerging pollutant) degradation, chosen for the chemical nature (MB = cationic, RhB pKa = 3.7, and Prozac® pKa = 10.7). Also, the Prozac® degradation and some of its by-products were accompanied by high-performance liquid chromatography (HPLC), which provided important information regarding the degradation mechanism mediated by Nb 2 O 5 nano bers. Details apropos the dye degradation mechanism were presented in this study, showing the importance of seeking a new statement on the Nb 2 O 5 ceramic bers application for environmental remediation. Thus, expanding studies on the material capacity to form reactive oxygen species in uenced by parameters such as the reaction medium pH, the adsorption process contribution, and the degradation of different organic contaminants is a novelty.

Experimental
Nb 2 O 5 ceramic nano bers preparation Nb 2 O 5 ceramic nano bers were fabricated by the electrospinning technique (Leindecker et al. 2014).
Brie y, an alcoholic solution (10% w v − 1 ) of polyvinylpyrrolidone (PVP) was prepared using ethanol (EtOH) and methanol (MeOH). These organic solvents were selected to achieve the PVP solubility and the appropriate viscosity of the polymeric precursor. Solution viscosity was maintained between 45 to 65 cP.
Then, niobium ammonium oxalate (NAO) was added to a polymeric solution in the ratio of 2:1 PVP/NAO (w:w) in 1 mL of acetic acid the viscosity of 45-65 cP. After complete solubilization, 15 mL of polymeric blend solution (PVP/alcohol/NAO) was kept under stirring for 12 h. Then, 5 mL of blend solution was transferred into a 10 mL glass syringe with a 0.7 mm diameter metal rod needle. At a distance of 10 cm from the needle tip, a metallic cylindrical collector was covered with aluminum foil and kept under rotation for ceramic blanket formation. Here, a metal needle was connected to the electric system and submitted at a voltage of 26 kV. Simultaneously, the collector was grounded to close the electrical circuit, and the solution ow rate was maintained at 1.6 mL h − 1 . After 40 min, the blanket form composite was collected and dried at 80°C for 12 h in the oven for solvent remotion. The blankets were calcined at 550°C for 4 h, 600°C for 3 h in a traditional mu e furnace with 2°C min − 1 of heating/cooling rates. The obtained Nb 2 O 5 ceramic bers were named EtOH_550, EtOH_600, MeOH_550, and MeOH_600, according to solvent and the thermal-treatment temperature.

Characterization assays
The viscosity of the polymeric blend solution was measured by Brook Field viscometer. The thermal behavior of composite bers was evaluated using TA Instruments equipment (model SDT 650).
Approximately 20 mg of each sample was transferred to alumina crucible and heated from 30 to 1000°C with 10°C min − 1 heating rate, under 100 mL min − 1 synthetic air ow. Structural characterization was done by X-ray diffraction (XRD), using a Shimadzu XRD 6000 diffractometer (Cu-Kα / λ = 1,5406 Å), 30 kV operating voltage, 30 mA tube current, 10-85° angular scan (2θ interval), and 0.02° min − 1 dislocation. Crystallite size was calculated using the Scherrer equation (Hargreaves 2016), and the lattice parameters obtained using Bragg's law (Stern 1978). Raman spectroscopy was applied to evaluate the nano ber crystalline structures in the short-range order. The assays were realized a Renishaw spectrophotometer (model RM 2000) with backscattering in the region of 100-4000 cm − 1 , an argon laser of 1 µm 2 area, and a wavelength spot 632.8 nm as the excitation source. Nano ber morphologies were characterized by Field Emission Gun Scanning Electron Microscopy (FEG-SEM) using a Jeol microscope (model JSM-6701F). Moreover, the presence of functional groups on the material surface was analyzed by Fourier Transform Infrared Spectroscopy (FTIR) using a Bruker Equinox 55 spectrometer. The spectra were obtained between 4000 and 400 cm − 1 , adjusted in the transmittance module, 4 cm − 1 resolution, and 32 scans. Zeta potential was obtained using a Zetasizer nano-ZS analyzer ( Also, a reuse study of up to 4 cycles was applied to evaluate the extended photoactivity of the catalyst that showed better performance. Therefore, the material recovered after each cycle was washed with distilled water (1 time), pure ethanol (2 times), and dried at 60ºC for 12 h, before a new application of 300 min for RhB and 90 min MB.

Prozac® photodegradation
Prozac® (98 %, Santa Cecilia pharmacy), 4-Tri uoromethyl phenol (TFMP), MAEB, and 3-phenylpropyl methylamine (PPMA) solutions were separately prepared by dissolving the pure reagents (all > 97%, Sigma-Aldrich) in ultrapure water. Then, 10 mg L − 1 of Prozac® solution was applied in the photocatalytic assays, while TFMP, MAEB, and PPMA were used for calibration curve preparation by HPLC with UV-Vis detection (205 nm). More details on the analytical methodology and solution preparations can be obtained from the literature (Moreira et al. 2020a). For the photocatalytic assay, 5 mg of the Nb 2 O 5 nano bers were dispersed in 10 mL of Prozac® solution (10.0 mg L − 1 ) under constant stirring and kept in the dark for 30 min. Then, samples were irradiated in the time interval of 3 to 120 min, ltered in 0.22 µm membrane for catalyst removal, and subjected to chemical analysis by HPLC. Finally, dye or Prozac® solutions with pH-adjusted were obtained by dropwise addition of 0.1 mol L − 1 HCl or NaOH solution and monitored by a potentiometer coupled to a glass membrane electrode. Hydroxyl radical probe assay was performed following the literature method (Moreira et al. 2020b).

Results And Discussion
Structural characterization X-ray diffraction patterns of the Nb 2 O 5 ceramic bers were indexed according to JCPDS 30-0873 (Fig. 1).
The highest intensity peaks were observed at 2θ = 22.7°, 28.5°, and 36.7° attributed to the orthorhombic phase (T-Nb 2 O 5 ) (Zeng et al. 2017). The intense and well-de ned lines (001), (180), and (181) showed the crystalline nature of the materials that was in uenced neither by solvents (MeOH or EtOH) nor by heat treatment temperature (550 or 600°C). These results corroborated the higher thermal stability of the ceramic material above 550°C (Fig. S1). Due to a mass loss above 90% for the polymeric blanket by successive thermal events, a complete elimination of organic matter according to the infrared spectra ( Fig. S2) was observed. This study revealed that the con rmation of Nb 2 O 5 ceramic ber crystalline behavior agrees with the reported literature (Nowak and Jaroniec 2008;Leindecker et al. 2014).   (c) with those reported in JCPDS 30-0873, the approximation was 99% (a), 101% (b) and 100% (c). Thus, the results of structural characterization con rm that the orthorhombic phase with high crystallinity was obtained for all samples, and therefore, the application in heterogeneous photocatalysis shows potential since the literature has already con rmed the T-Nb 2 O 5 photoactivity (Lopes et al. 2014;Shao et al. 2014) Morphological characterization Figure 2 showed the FEG-SEM images of the EtOH_550, EtOH_600, MeOH_550, and MeOH_600 samples, con rming the Nb 2 O 5 nanometric bers (< 90 nm) obtaining. In the current study, the ber diameter was reduced by 40% compared to the similar synthesis method used to produce Nb 2 O 5 (Leindecker et al. 2014). This characteristic is preferred in photocatalysis studies since materials with a smaller diameter show a superior surface area. According to Table 1, the average sample diameters treated at 550°C were 35% smaller than the treated at 600°C. However, the samples processed in EtOH (550 and 600°C) showed a ber diameter of 15% smaller, concerning the samples processed in MeOH under the same temperature conditions. These results indicated that the methanol (MeOH = 64°C), the lowest boiling point solvent, is volatilized more e ciently than ethanol (EtOH = 78°C), favoring the Nb 2 O 5 crystal agglomeration. However, this same heat treatment temperature did not signi cantly in uence the crystallites size (Table 1). Thus, the ber diameter increased due to the particle agglomeration degree obtained after the polymeric structure decomposition. Samples processed with EtOH showed a surface with homogeneous roughness (inset) and without holes (Fig. 2b). Therefore, the Nb 2 O 5 growth and particle organization occurred more homogeneously to form the ceramic bers. As MeOH is the most volatile, the obtained bers with these solvents showed super cial deformations (Fig. 2d) and small holes distributed heterogeneously along with the structure (inset). Consequently, ceramic bers with large SSA values were obtained, increasing the material potential for heterogeneous photocatalysis.

Optical characterization
The Nb 2 O 5 bers optical property was investigated by diffuse re ectance analysis. The direct bandgap energy was calculated according to the literature (Lopes et al. 2014). For all samples, the diffuse re ectance spectrum exhibited a homogeneity that corroborates with the XRD data, which showed the presence of a single crystalline phase (T-Nb 2 O 5 ). Consequently, the average bandgap energy was E g = 3.87 ± 0.04 eV for MeOH_550, MeOH_600, and EtOH_600 samples. The EtOH_550 showed an E g = 3.6 eV slightly shifted to the lower energy. Therefore, the Nb 2 O 5 bers photoactivity investigation viewing heterogeneous photocatalysis applications requires the ultraviolet wavelength.  Figure 3 shows the PL spectra with top emission centers. The area percentage of each pro le was adjusted using the Voigt (area) function, as described in Sect. 2.2. All PL pro les are broadband type, showing an evident relationship between the contributions of various energy levels to the electronic transition process. In Fig. 3a, the Nb 2 O 5 bers bandgap values were higher than the maximum PL emission of each sample (~ 2.3 to 2.6 eV) at 477 to 545 nm, respectively. Therefore, broad PL spectra cover part of the visible electromagnetic spectrum, while band-to-band electronic transitions present a slight in uence. This optical phenomenon may be related to the prohibited zone additional energy levels due to the surface, interface, and bulk defects (Paris et al. 2007;Machado et al. 2017).
The differences observed in each PL spectra were treated individually with the insertion from three to six Voigt peaks, perfectly adjusted to the spectrum, as shown in Fig. 4. In all cases, a broadband emission pro le in which the maximum emission, intensity, and amplitude were signi cantly dependent on the heat treatment temperature and solvents used in the ber processing. Initially, there was a trend, with respect to the heat treatment variation in relation to the PL intensity (PL MeOH _ 550 > PL EtOH _ 550 > PL EtOH _ 600 > PL MeOH _ 600 ). The samples treated at 550°C / 4h showed higher PL intensity than those treated at 600°C / 3h, which was expected due to the higher particle ordination at a higher temperature and consequent decrease in defects. For the MeOH_550 sample (Fig. 4a), the PL spectrum showed broadband and the highest intensity, with a maximum centered at 507 nm (Eg = 2.44 eV) in the green color. The broadband emission of the MeOH_600 sample (Fig. 4b) showed a maximum at approximately 545 nm (Eg = 2.27 eV), resulting in green color. This pro le was the broadest of the samples analyzed, covering the most extensive spectrum from 450 to 750 nm. This effect on the PL spectrum may be due to a maximum amount of defects (surface, interface, and bulk) induced by structural changes (Phase TT-Nb 2 O 5 to T-Nb 2 O 5 ).
The Nb 2 O 5 orthorhombic phase begins to crystalize at 500°C (Nico et al. 2016). In this temperature vicinity, defects arising from atomic diffusion, plane displacements, and NbO 6 octahedral distortion are susceptible. These changes were possibly responsible for showing a greater PL for the MeOH_550 than the MeOH_600 sample, in whose temperature more signi cant structural ordering was expected and consequent decrease in defects (Zhou et al. 2008;Joya et al. 2017).
However, samples processed with ethanol showed PL spectra with smaller intensity, following the same trend as those processed with methanol. Although the EtOH_550 (Fig. 4c) showed a PL minor intensity pro le than MeOH_550, the rst one also showed a broadband spectrum with a maximum of 535 nm (Eg = 2.31 eV) resulting in green color. This same pro le was observed for EtOH_600 (Fig. 4c). However, the PL intensity was slightly lower than observed for EtOH_550. Therefore, the PL spectra changes were evidenced by processing changes in the precursors, solvents, or temperature. Finally, Nb 2 O 5 bers were evaluated to remove organic pollutants (dye and pharmaceuticals) through heterogeneous photocatalysis under UV radiation (254 nm).

Photocatalytic activity
Photocatalytic degradation of dyes The adsorption assays (on dark) performed for RhB (12 h) and MB (0.5 h) showed that while RhB adsorption was only 1%, MB reached 18% e cacy for the different nano bers. According to the literature, these results were shown more signi cant MB adsorption than RhB when other nanomaterials were applied (Choi et al. 2015;Radoń et al. 2019). Thus, after the adsorption equilibrium, the dyes were subjected to photocatalytic assays under UV light, and the degradation curves were shown in Fig. 5. Figure 5.
The average RhB photocatalytic degradation was 47 ± 2% for Nb 2 O 5 bers in 5h (Fig. 5a), highlighting the EtOH_550 showed the best performance (49%). Only 10% of RhB was degraded in the photocatalyst absence, con rming the tendency. The MB photocatalytic degradation was up to 78% ±1 in 90 min for EtOH_550 (Fig. 5b), while the photolysis removed only 2%. First-order kinetic constants were calculated for all processes and depicted in Table 2. The kinetic constants for all Nb 2 O 5 bers are shown in Table 2, con rming photocatalytic activity against RhB and MB degradation. EtOH_550 e ciency was highlighted due to the highest kinetic constants for RhB (1.77 x 10 − 3 ± 0.06 min − 1 ) and MB (12.9 x 10 − 3 ± 0.7 min − 1 ). The best performance for EtOH_550 corroborates with the smaller bers diameter, large surface area, and reduced bandgap (Tables 1 and 2). When comparing the highest kinetic constants obtained for RhB and MB, an average increase of 84% ±3 was con rmed. Con rmed the best performance for EtOH_550 against the RhB and MB degradation, this material was selected for the study of reuse of up to 4 cycles. Figure 6 shows that the RhB and MB removal after the 4 cycles were 45 ± 3% and 74 ± 2%, respectively. Simultaneously, the slight variation observed in the removal (%) after each cycle can be attributed to the minor material loss after the successive transfers. Therefore, these results indicate the maintenance of the EtOH_550 catalytic activity even after 1200 min for RhB and 360 min for MB.

Figure 6
Notably, the EtOH_550 performance was very different for the degradation of the different dyes in the investigated condition (pH = 6), suggesting that its performance was not associated with the material physicochemical properties. Thus, this enhanced catalytic performance of the bers associated with the dye chemical nature may have been the factor responsible for in uencing the process e ciency. As the RhB charge can be changed by the -COOH group protonation/deprotonation, the in uence of pH was investigated.
pH evaluation RhB and MB are cationic dyes that negative nanoparticle surfaces can attract to optimize the photocatalytic processes. However, only RhB presents a carboxylic acid group that can be deprotonated and increases the negative charge density in this molecule at pH > pKa = 3.7 (Obregón and Colón 2013; Aguilar et al. 2014). Then, the RhB (5 mg L − 1 ) and MB (3 mg L − 1 ) solutions were evaluated respectively for 5 h and 1.5 h to investigate the photocatalytic response at different pH conditions (pH = 3, 6, or 9), using EtOH_550 sample (Fig. 6).

Figure 7.
According to Fig. 7, the photocatalytic process was e cient for RhB and MB degradation in all pH conditions. When pH ~ pKa = 3.7, RhB molecules transit in their protonated or neutral form, facilitating the interaction with oxidizing species responsible for the dye degradation (Obregón and Colón 2013). In alkaline conditions, the -COOH group deprotonation increases the dye lipophilicity and signi cantly decreases the degradation rate (Aguilar et al. 2014). However, catalyst presence showed the best response for RhB degradation results under very different pH conditions (Fig. 7a). The RhB removal was 87% for pH = 3 and 72% for pH = 9, while the photocatalytic process performance was increased up to 28% and 59% compared to photolytic, respectively. This performance observed for the RhB photocatalytic degradation at alkaline pH was similar for MB degradation under the same conditions (Fig. 7b). MB photolytic degradation was insigni cant at 6 and 9 pH. Simultaneously, the addition of the EtOH_550 catalyst showed to improve performance up to 76% for pH = 6 and 93% for pH = 9 compared to the photolytic process. Therefore, Nb 2 O 5 bers were essential to MB e cient degradation, con rming the effectiveness as a catalyst for the dye degradation process.
In contrast to the RhB, the MB does not show the -COOH group affected by the acid-base equilibrium reaction, presenting a strictly cationic molecular structure. This aspect draws attention to the dissimilar performances, suggesting that the different dye charge distribution was the parameter responsible for optimizing the degradation process in an alkaline pH. Therefore, to better understand the pH in uence in the photocatalytic process, the zeta potential of the EtOH_550 bers was measured in the pH = 1 to 10 range (Fig. 7c). The zero charge point was identi ed at pH ~ 2, while the negative surface was con rmed in all conditions of pH > 2.
At pH = 3, RhB this protonated (pKa = 3.7), optimizing the interaction with the EtOH_550 ber which showed a slightly negative charge (ζ = − 15 ± 1 mV). At pH = 9, the EtOH_550 ber showed a signi cantly negative charge (ζ = − 50 ± 5 mV) capable of repelling the deprotonated RhB (Merka et al. 2011). Even with the less interaction between RhB and EtOH_550 at pH = 9, the photocatalysis e ciency was con rmed, showing that ζ = − 50 ± 5 mV could optimize the EtOH_550 ber dispersion and increase the active area. This justi cation is applied to the e cient MB degradation at pH = 9, optimized due to the better ber dispersions in the reaction medium. The -COOH group absence in the MB allowed the photocatalytic degradation to be 93% against the 59% observed for RhB. These RhB and MB photocatalytic degradation results follow the literature, showing the organic pollutant chemical nature is an important factor in understanding the photodegradation mechanism (Merka et al. 2011;Obregón and Colón 2013;Reeta Mary et al. 2018). Also, this increase of the negative material charge (Zeta) in less acidic conditions occurred due to the probable deprotonation of Nb-O groups that were identi ed by Raman analyses (Fig. 1), corroborating with the literature (Silva et al. 2019).
Due to the excellent results of dye photocatalytic degradations using EtOH_550 at pH = 9, this catalyst was applied in these conditions to degrade Prozac®. This pharmaceutical product choice was due to the high pKa = 10.7 (Do et al. 2017), which preferably shows its protonated form at pH = 9 and possibly a better interaction with EtOH_550, which led to a ζ = − 50 ± 5 mV. However, the -CF 3 group presence in the Prozac® molecular structure shows a negative charge density that can compete with the contribute repulsion from catalyst even for the protonated system. Thus, the photocatalytic degradation of Prozac® was investigated.

Prozac® Photocatalytic degradation
A 10 mg L − 1 Prozac® solution (pH = 9) was subjected to adsorption with EtOH_550 in 1 to 30 min range ( Fig. 8a). The results showed an adsorption > 85% ±2 in 1 min, indicating a rapid interaction between EtOH_550 and Prozac® induced by the ζ = − 50 ± 5 mV of the catalyst and the -NH 3 + group protonated in pH = 9 (Do et al. 2017). Then, was con rmed a 32% Prozac® desorption, previously adsorbed, until 5 min, suggesting that after the initial attraction, the -CF 3 group promoted the molecule repulsions close to the EtOH_550 surface. After this desorption step, the adsorptive process was retaken, removing Prozac® 67% ±1 in 20 min. Finally, in 30 min, the adsorption was 68 ± 1%, showing no variation and, therefore, reaching equilibrium. Thus, before being irradiated, Prozac solutions (10 mg L − 1 ) were subjected to adsorption for 30 min.  showed that up to 30 min, the C/Co ratio decreases considerably for both photochemical processes. Compared with photolytic, the Prozac®photocatalytic removal showed an average increase of 16% ±1 for 3, 5, and 10 min. These results corroborate those found during the degradation of the dyes, con rming that EtOH_550 is a photoactive catalyst capable of degrading the different organic pollutants.
The chromatographic analyses used to monitor Prozac® degradation also provided essential details regarding the quantitative by-product formations. According to the literature, TFMP, MAEB, and PPMA are the main by-products formed after the Prozac®degradation (Moreira et al. 2020a) and, therefore, were quantitatively monitored in the present study.
According to the literature, the conversion from Prozac® to MAEB + TFMP preferably follows a radical oxidation mechanism, while the conversion from Prozac® to PPMA + TFMP complies with the photolytic or hydrogenation mechanism (Moreira et al. 2020a). Figure 9a shows the maximum conversion rate obtained during the Prozac® photolytic and photocatalytic degradations calculated according to the literature (Moreira et al. 2020a). The results con rm that the conversion to TFMP was more signi cant in both processes precisely because this is present in the two degradation mechanisms. Furthermore, the EtOH_550 catalyst presence did not in uence the conversion rate to TFMP, which was similar to that found for the photolytic process. Therefore, this result suggests that the electrostatic repulsion between the -CF 3 group (present in the TFMP structure) and the EtOH_550 (ζ = − 50 ± 5 mV) prevents the interaction between both, allowing the TFMP to be available in the aqueous phase for a longer time. The 1: 1 stoichiometry is expected for conversion from Prozac® to MAEB or PPMA (Moreira et al. 2020a).
As different routes obtain PPMA or MAEB, the molar ratio between these by-products provides valuable information regarding the preferential Prozac® degradation mechanism. Therefore, the molar ratio [r molar = (n MAEB /n PPMA )] was calculated for the point of greatest conversion rate, where r molar = 0.96 for photolytic removal and r molar = 1.5 for EtOH_550. As the highest r molar value was obtained for EtOH_550, the Prozac® oxidation mechanism mediated by hydroxyl radicals to preferentially form the MAEB was suggested. A 1.5 mg L − 1 coumarin solution was irradiated in the absence and presence of the EtOH_550 catalyst to support this idea, and the reaction products being monitored by photoluminescence (Fig. 8b). The PL spectra of the samples submitted to photolysis and EtOH_550 showed a peak centered at 452 nm only for EtOH_550 (Fig. 9b).
According to the literature, umbelliferone is the hydroxylation by-product of coumarin responsible at 452 nm PL emitting, and the emission is not veri ed in the hydroxyl radical absence (Louit et al. 2005). When adding dimethylsulfoxide (DMSO) as a radical scavenger in the reaction medium, the PL spectra (EtOH_550 + DMSO) showed the peak-centered suppression at 452 nm, con rming that the radicals were not available to umbelliferone form. Therefore, the photoactivity of EtOH_550 when exposed to UV radiation was su cient to promote electronic excitation and generate the e − (CB) and h + (VB) pairs. These, in turn, were responsible for forming the oxidizing species capable of optimizing the Prozac® degradation for formed TFMP, PPMA, and MAEB. Finally, the Prozac® photocatalytic degradation mechanism in the presence of EtOH_550 can be represented by Fig. 10.

Conclusion
Nb 2 O 5 nano bers were e ciently obtained through the electrospinning technique, and their photocatalytic property was investigated from the degradation of three organic pollutants. XRD and Raman analysis showed that the structure was consistent with the high crystallinity orthorhombic phase (T-Nb 2 O 5 ). The FEG analysis con rmed the ceramic bers form with a diameter up to 47 nm ± 0.7, and BET results indicate an area up to 85.0 m 2 g − 1 ±1.7. The broadband PL spectra showed that different energy levels contribute to electronic transition events in the UV-vis region. When applied to RhB or MB photocatalytic degradation, the highest e ciency was obtained for EtOH_550, which degraded 47% ±2 of RhB in 5 h and 78% ±1 MB in 1.5 h. The kinetic constant in MB degradation was 86% higher than found for RhB, indicating that the strictly cationic dye was removed with greater e ciency by the catalyst. By investigating different pH conditions during photocatalytic degradation, a removal up to 87% at pH = 3 for RhB (5 h) and 93% for MB at pH = 9 (1.5 h) was achieved, suggesting the chemical nature of the dyes was the responsible factor for guiding photocatalytic e ciency. At pH = 9, the bers with surface charge could be dispersed in the reaction system with greater e ciency, reaching the optimized condition to degrade the protonated molecules. Besides, due to pKa = 10.7 and protonated at pH = 9, Prozac® was subjected to photocatalytic degradation with EtOH_550. For this compound, photocatalysis was up to 17% more e cient than photolysis. Additionally, Prozac® by-products (TFMP, MAEB, and PPMA) were monitored by HPLC, and the higher forms of MAEB in the photocatalytic process were consistent with the formation of hydroxyl radicals in this system. Therefore, Nb 2 O 5 nano bers can be e ciently obtained for application in heterogeneous photocatalysis and can be an alternative material to TiO 2 , which is still the most used semiconductor in these systems.

Declarations
Ethics approval and consent to participate Not applicable

Consent for publication
Not applicable Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests Funding Not applicable