Fabrication of ZnO and TiO2 Nanotubes via Flexible Electro-Spun Nanofibers for Photocatalytic Applications

Web-like architectures of ZnO and TiO2 nanotubes were fabricated based on a three-step process of templating polymer nanofibers produced by electrospinning (step 1). The electrospun polymer nanofibers were covered by radio-frequency magnetron sputtering with thin layers of semiconducting materials (step 2), with FESEM observations proving uniform deposits over their entire surface. ZnO or TiO2 nanotubes were obtained by subsequent calcination (step 3). XRD measurements proved that the nanotubes were of a single crystalline phase (wurtzite for ZnO and anatase for TiO2) and that no other crystalline phases appeared. No other elements were present in the composition of the nanotubes, confirmed by EDX measurements. Reflectance spectra and Tauc plots of Kubelka–Munk functions revealed that the band gaps of the nanotubes were lower than those of the bulk materials (3.05 eV for ZnO and 3.16 eV for TiO2). Photocatalytic performances for the degradation of Rhodamine B showed a large degradation efficiency, even for small quantities of nanotubes (0.5 mg/10 mL dye solution): ~55% for ZnO, and ~95% for TiO2.

Electrospinning is a high throughput, cost-effective, and versatile technology used at a large scale to fabricate nanofibers on large surfaces by applying an electrostatic force,

Step 1. Electrospinning of PMMA Fibers
All chemical reagents were acquired from Merck (Darmstadt, Germany) and used without further purification. In order to obtain polymer fibers, we prepared solutions of 10 wt% poly(methyl methacrylate) (PMMA) with Mw = 300,000, with a chemical formula of (C 5 O 2 H 8 ) n . Dimethylformamide (DMF) was used as the solvent.
To obtain the freestanding polymer fiber webs, large (10 × 10 cm 2 ) square copper frames were fabricated and used as collectors in an otherwise classical electrospinning setup, interposed between the syringe needle spinneret and a 20 × 20 cm 2 aluminum plate acting as a grounded electrode. The distance between the copper frame collectors and the aluminum plate and spinneret was approximately 6 cm and 15 cm, respectively. The 10% (m/v) PMMA/DMF solution was fed using a New Era Pump Systems (Farmingdale, NY, USA) syringe pump to a 0.8 mm diameter blunt needle at a 0.5 mL/h rate, with a 15 kV voltage applied to the needle using a Spellman (West Sussex, UK) SL300 HV source. Collection time was 60 min for each copper frame.

Step 2. RF Magnetron Sputtering of ZnO and TiO 2 Layers
After the electrospinning process, the copper frames containing the PMMA nanofibers were coated on both sides with a semiconducting thin layer (ZnO or TiO 2 ) by RF magnetron sputtering employing Tectra GmbH Physikalische Instrumente equipment (Frankfurt, Germany). In the deposition process, a zinc oxide target (ZnO) or a titanium dioxide (TiO 2 ) target (acquired from Kurt J. Lesker Company Ltd. Hastings, UK) with a diameter of 2 inches and a thickness of 0.125 inches was used. In the case of ZnO, the following parameters were applied: the working gas was an Ar atmosphere with a purity of 99.999%, the pressure in the deposition chamber was 5.4 × 10 −3 mbar, the RF power applied on the magnetron was 100 W, and the deposition time was 2 h for each side. In the case of TiO 2 , the following parameters were used: the working gas was an Ar atmosphere, the pressure in the deposition chamber was 4 × 10 −3 mbar, the RF power applied on the magnetron was 200 W, and the deposition time was 3 h for each side.

Step 3. Calcination of ZnO and TiO 2 Nanotubes
Subsequently, the PMMA nanofibers covered on both sides with either ZnO or TiO 2 films were transferred onto Si/SiO 2 substrates and calcinated using a convection oven from Nabertherm GmbH (Lilienthal, Germany). Hence, the calcination process was performed at 500 • C for ZnO and 600 • C for TiO 2 for 12 h in the air at ambient pressure. Afterward, 3D web-like networks of ZnO nanotubes (ZnO NT) and TiO 2 nanotubes (TiO 2 NT) were obtained, the PMMA being completely burnt out during the calcination process ( Figure 1). Subsequently, the PMMA nanofibers covered on both sides with either ZnO or TiO2 films were transferred onto Si/SiO2 substrates and calcinated using a convection oven from Nabertherm GmbH (Lilienthal, Germany). Hence, the calcination process was performed at 500 °C for ZnO and 600 °C for TiO2 for 12 h in the air at ambient pressure. Afterward, 3D web-like networks of ZnO nanotubes (ZnO NT) and TiO2 nanotubes (TiO2 NT) were obtained, the PMMA being completely burnt out during the calcination process ( Figure 1).

Characterization of ZnO and TiO2 Nanotubes
The morphological properties of the ZnO and TiO2 nanotube networks were investigated with a Zeiss (Oberkochen, Germany) Merlin Compact field emission scanning electron microscope (FESEM) working in high vacuum (HV), from 0.2 to 30kV, and equipped with a Zr/W emitter, InLens, and SE2 detectors.
The elemental composition of the samples was evaluated using a Zeiss (Oberkochen, Germany) Evo 50 XVP scanning electron microscope (SEM) equipped with a Bruker QUANTAX 200 energy dispersive X-ray spectrometer (EDS) with energy resolution < 129 eV at Mn-Ka and Peltier cooling.
The optical properties were investigated by reflectance and photoluminescence. The reflectance spectra of the nanotubes were recorded with a Perkin-Elmer (Rodgau, Germany) Lambda 45 spectrometer using an integrating sphere. The photoluminescence excited at λexc = 350 nm and was evaluated using an FL 920 Edinburgh Instruments (Livingston, United Kingdom) spectrometer with a 450 W Xe lamp excitation and double monochromators on both excitation and emission, respectively.
The photocatalytic properties of ZnO and TiO2 nanotubes were assessed by monitoring the photodegradation of the Rhodamine B (RhB) during the irradiation with solar light. Approximately 0.5 mg of either ZnO or TiO2 nanotubes deposited on Si substrates were immersed in a beaker containing 10 mL of RhB aqueous solution (10 −5 M). Deionized

Characterization of ZnO and TiO 2 Nanotubes
The morphological properties of the ZnO and TiO 2 nanotube networks were investigated with a Zeiss (Oberkochen, Germany) Merlin Compact field emission scanning electron microscope (FESEM) working in high vacuum (HV), from 0.2 to 30 kV, and equipped with a Zr/W emitter, InLens, and SE2 detectors.
The elemental composition of the samples was evaluated using a Zeiss (Oberkochen, Germany) Evo 50 XVP scanning electron microscope (SEM) equipped with a Bruker QUAN-TAX 200 energy dispersive X-ray spectrometer (EDS) with energy resolution < 129 eV at Mn-Ka and Peltier cooling.
The optical properties were investigated by reflectance and photoluminescence. The reflectance spectra of the nanotubes were recorded with a Perkin-Elmer (Rodgau, Germany) Lambda 45 spectrometer using an integrating sphere. The photoluminescence excited at λexc = 350 nm and was evaluated using an FL 920 Edinburgh Instruments (Livingston, UK) spectrometer with a 450 W Xe lamp excitation and double monochromators on both excitation and emission, respectively.
The photocatalytic properties of ZnO and TiO 2 nanotubes were assessed by monitoring the photodegradation of the Rhodamine B (RhB) during the irradiation with solar light. Approximately 0.5 mg of either ZnO or TiO 2 nanotubes deposited on Si substrates were immersed in a beaker containing 10 mL of RhB aqueous solution (10 −5 M). Deionized water was obtained using a Millipore system. During the optical measurements, the dye solutions were kept in the dark. Irradiation with white light was performed for 600 min using an SF300-A Small Collimated Beam Solar Simulator (Sciencetech, London, ON, Canada) equipped with an Air Mass AM1.5G Filter (spot size: 25 mm diameter at one Sun) and an integrated electrical shutter with a controller and a Xe lamp (300 W). The samples were positioned 10 cm from the source. During the irradiation, at different time intervals, the optical absorbance spectra of the samples at λ = 400-700 nm were measured using a UV-Vis-NIR CARY 5000 (Varian, Agilent Technologies Deutschland GmbH, Waldbronn, Germany) spectrophotometer provided with a quartz cell with a light path of 10 mm. The characteristic absorption band of RhB, peaking at~554 nm, was monitored, and its photodegradation was evaluated.

Results and Discussion
After the removal by calcination of the organic core template formed by the electrospun PMMA nanofibers, the ZnO and TiO 2 nanotubes were characterized from the morphological, structural, compositional, and optical points of view. water was obtained using a Millipore system. During the optical measurements, the dye solutions were kept in the dark. Irradiation with white light was performed for 600 min using an SF300-A Small Collimated Beam Solar Simulator (Sciencetech, London, ON, Canada) equipped with an Air Mass AM1.5G Filter (spot size: 25 mm diameter at one Sun) and an integrated electrical shutter with a controller and a Xe lamp (300 W). The samples were positioned 10 cm from the source. During the irradiation, at different time intervals, the optical absorbance spectra of the samples at λ = 400-700 nm were measured using a UV-Vis-NIR CARY 5000 (Varian, Agilent Technologies Deutschland GmbH, Waldbronn, Germany) spectrophotometer provided with a quartz cell with a light path of 10 mm. The characteristic absorption band of RhB, peaking at ~554 nm, was monitored, and its photodegradation was evaluated.

Results and Discussion
After the removal by calcination of the organic core template formed by the electrospun PMMA nanofibers, the ZnO and TiO2 nanotubes were characterized from the morphological, structural, compositional, and optical points of view. The semiconducting layers deposited on the nanofibers are continuous, as observed in the higher magnification FESEM images presented in Figure 2c,d, and have granular morphology for both ZnO and TiO2 nanotubes. The coverings with sputtered materials were of uniform thickness all around the surface of the polymeric fibers, as demonstrated by the FESEM images. Because the ZnO nanotubes' walls are thicker than the walls of the TiO2 nanotubes (ZnO nanotubes have a wall thickness of ~50 nm, whereas the TiO2 nano- The semiconducting layers deposited on the nanofibers are continuous, as observed in the higher magnification FESEM images presented in Figure 2c,d, and have granular morphology for both ZnO and TiO 2 nanotubes. The coverings with sputtered materials were of uniform thickness all around the surface of the polymeric fibers, as demonstrated by the FESEM images. Because the ZnO nanotubes' walls are thicker than the walls of the TiO 2 nanotubes (ZnO nanotubes have a wall thickness of~50 nm, whereas the TiO 2 nanotubes have a wall thickness of~35 nm (or less)), and due to the fact that the density of ZnO (~5.6 g/cm 3 ) is larger than the density of anatase TiO 2 (~3.8 g/cm 3 ), the web-like architecture of TiO 2 nanotubes has a higher active surface. Therefore, the same amount of nanotubes (~0.5 mg) leads to a larger number of nanotubes for TiO 2 than for ZnO. Although the diameters of the polymeric fibers are similar, the higher active surface provided by the TiO 2 nanotubes may drastically influence their properties, increasing the photocatalytic activity of those nanostructures.

XRD Measurements
The structural characterization of the ZnO and TiO 2 nanotubes fabricated via electrospinning was performed by XRD analysis, as presented in In order to better observe the results from the XRD measurements, Table 1 presents the XRD data for ZnO and TiO 2 nanotubes. It has to be remarked that all peaks appearing in the ICDD patterns are visible in the ZnO nanotubes and TiO 2 nanotubes samples' diffractograms. Moreover, no additional peaks were observed in the XRD patterns of the ZnO or TiO 2 nanotubes, confirming that the synthesized nanotubes were of a single crystalline phase and that no other additional crystalline structures were obtained after the calcination step.

204
It has to be remarked that all peaks appearing in the ICDD patterns are visible in the ZnO nanotubes and TiO2 nanotubes samples' diffractograms. Moreover, no additional peaks were observed in the XRD patterns of the ZnO or TiO2 nanotubes, confirming that the synthesized nanotubes were of a single crystalline phase and that no other additional crystalline structures were obtained after the calcination step.  Further analysis of the XRD results was performed. Thus, the mean sizes of the grains in the single-phase crystalline nanotubes were estimated from the full width at half maximum (FWHM) and Debye-Scherrer formula according to the equation D = 0.9λ/βcos θ, where 0.9 is the shape factor, λ = 0.154 nm is the wavelength of the incident CuKα radiation, β represents full-width at half maximum of the respective peak in radians, and θ is the Bragg diffraction angle. Further analysis of the XRD results was performed. Thus, the mean sizes of the grains in the single-phase crystalline nanotubes were estimated from the full width at half maximum (FWHM) and Debye-Scherrer formula according to the equation D = 0.9λ/βcos θ, where 0.9 is the shape factor, λ = 0.154 nm is the wavelength of the incident CuKα radiation, β represents full-width at half maximum of the respective peak in radians, and θ is the Bragg diffraction angle.
The mean size of the grains was calculated using the full width at half maximum (FWHM) of the most intense peaks in the diffractograms and was found to be~30 nm in the ZnO nanotubes and~35 nm in the TiO 2 nanotubes from this Debye-Scherrer equation.

EDX Analysis
The uniformity of the elemental composition of the deposited nanotubes, after the calcination step, was evaluated using the EDX mapping. Thus, the elemental maps of ZnO nanotubes (Figure 4a,c) and TiO 2 nanotubes (Figure 4b,d) confirm that the semiconducting material is uniformly distributed, and the network replication of the electrospun nanofibers has a uniform composition.

EDX Analysis
The uniformity of the elemental composition of the deposited nanotubes, after the calcination step, was evaluated using the EDX mapping. Thus, the elemental maps of ZnO nanotubes (Figure 4a,c) and TiO2 nanotubes (Figure 4b,d) confirm that the semiconducting material is uniformly distributed, and the network replication of the electrospun nanofibers has a uniform composition.

Optical Properties
UV-vis measurements were used in order to evaluate the optical properties of the investigated samples. The UV-vis band gap energies for the samples were determined from the room temperature reflectance (R) spectra. The reflectance spectra of the ZnO and TiO2 nanotubes deposited on Si substrates are presented in Figure 5a. It can be observed that, for both types of nanotubes, the reflectance values present slopes that have different angles at wavelengths between 300 nm and 400 nm.
In order to determine the band gap values of the samples, we calculated the Kubelka-Munk function F(R) = [(1 − R) 2 ]/2R, where R is the diffuse reflectance.

Optical Properties
UV-vis measurements were used in order to evaluate the optical properties of the investigated samples. The UV-vis band gap energies for the samples were determined from the room temperature reflectance (R) spectra. The reflectance spectra of the ZnO and TiO 2 nanotubes deposited on Si substrates are presented in Figure 5a. It can be observed that, for both types of nanotubes, the reflectance values present slopes that have different angles at wavelengths between 300 nm and 400 nm.
In order to determine the band gap values of the samples, we calculated the Kubelka-Munk function F(R) = [(1 − R) 2 ]/2R, where R is the diffuse reflectance.
Tauc's plots (F(R) × hν) n vs. hν), where hν is the photon energy and n = 1 2 for direct band gap semiconductors, were used in order to determine the band gap energies for the samples.
Thus, the band gap energies were estimated from the intersections of the tangents to the slopes in the curves with the photon energy axis, as illustrated in Figure 5b TiO 2 (3.2 eV for anatase) [17]. Both values calculated for the band gap energies suggest that the nanotubes will present photocatalytic properties at wavelengths higher than those corresponding to UV range.
The shift in the band gap energy's value is either an effect due to the modification of the structural properties of the semiconductor or an effect caused by doping the material. It is well known that the band gap values for semiconductors should increase with a decrease in size because of the modification of the spacing of the electronic levels. However, there can be situations when, even if the semiconductor is nanostructured, a decrease in band gap energy value may be observed. Usually, the narrowing of the band gap is due to the impurities present in the material. However, a decrease in the band gap can be explained based on the presence of a mixture of oxidation states similar to the ones induced by the doping of the material. The XRD confirmed that there is one single crystalline phase for the nanotubes; therefore, the decreasing of the bandgap is a result of the oxidation states induced by structural imperfections [44].

samples.
Thus, the band gap energies were estimated from the intersections of the tangents to the slopes in the curves with the photon energy axis, as illustrated in Figure 5b. The nanotubes show values of Eg of about 3.04 eV for ZnO and 3.16 eV for TiO2. The band gap of the ZnO nanotubes is consistent with studies presented in the literature that underline that the effective band gap of ZnO nanostructures is smaller than the value for the bulk material of 3.37 eV [16]. On the other hand, a very small shift to lower values of the energies is observed for the band gap of TiO2 nanotubes when compared with the band gap of bulk TiO2 (3.2 eV for anatase) [17]. Both values calculated for the band gap energies suggest that the nanotubes will present photocatalytic properties at wavelengths higher than those corresponding to UV range.
The shift in the band gap energy's value is either an effect due to the modification of the structural properties of the semiconductor or an effect caused by doping the material. It is well known that the band gap values for semiconductors should increase with a decrease in size because of the modification of the spacing of the electronic levels. However, there can be situations when, even if the semiconductor is nanostructured, a decrease in band gap energy value may be observed. Usually, the narrowing of the band gap is due to the impurities present in the material. However, a decrease in the band gap can be explained based on the presence of a mixture of oxidation states similar to the ones induced by the doping of the material. The XRD confirmed that there is one single crystalline phase for the nanotubes; therefore, the decreasing of the bandgap is a result of the oxidation states induced by structural imperfections [44]. The photoluminescence spectrum of the ZnO nanotubes presents two bands, typical for this semiconducting material (Figure 5c). Thus, a weak band may be observed peaking in the ultraviolet region, with a maximum at ~380nm, being related to the band-to-band transitions and appearing due to the recombination of the excitons [35]. The more intense and broad band that may be observed in the visible region, with the maximum at ~595 nm, is related to various defects that are formed in ZnO (vacancies, interstitials, etc.) [35]. The photoluminescence spectrum of the ZnO nanotubes presents two bands, typical for this semiconducting material (Figure 5c). Thus, a weak band may be observed peaking in the ultraviolet region, with a maximum at~380nm, being related to the band-to-band transitions and appearing due to the recombination of the excitons [35]. The more intense and broad band that may be observed in the visible region, with the maximum at~595 nm, is related to various defects that are formed in ZnO (vacancies, interstitials, etc.) [35].

Photocatalytic Activities
The photocatalytic activity of the nanotubes was evaluated by Rhodamine B degradation experiments under irradiation with a solar simulator's white light employing aqueous solutions in which the nanotubes were immersed. The self-degradation of Rhodamine B under the solar light in the absence of the samples was negligible. In order to evaluate the degradation that occurred when nanotubes were immersed in the dye's solutions, 3 mL of RhB solution was collected from the beaker at different intervals of time, and the absorption spectra were registered ( Figure 6). The irradiation with white light was started immediately after the immersion of the samples. The degradation process was monitored for 600 minutes in order to observe if, for any of the samples, a total degradation of the dye was obtained or until we achieve a saturation threshold.
During the absorption measurements, the irradiation with the solar simulator was interrupted. The solution was kept in the dark and, after the re-injection in the beaker of the measured amount, the whole solution was homogenized. Comparative with the total time of irradiation, the absorbance measurements were performed quickly, excluding a supplementary degradation due to the spectrometric evaluation.
In order to exclude the influence of secondary processes such as adsorption effects and prove that the degradation of Rhodamine B is not due to adsorption into the nanotubes, ZnO nanotubes were immersed in RhB solution in the dark for a total of 36 h ( Figure S1a). After the sample removal, there was no indication that adsorption processes took place ( Figure S1b), and the intensity of the absorption band of Rhodamine B after 36 h was similar to the initial intensity ( Figure S1c).
As observed in Figure 6, the photocatalytic effect registered for ZnO nanotubes (illustrated in Figure 6a) is less effective than the degradation observed for the TiO 2 nanotubes sample (illustrated in Figure 6b).
After increasing the reaction time, the photodegradation rate of the organic dye decreases. In order to observe the kinetics of the degradation, the degradation efficiency was calculated using the following equation: Degradation efficiency (%) = (C 0 − C)/C 0 × 100 where C 0 is the absorbance value before irradiation at t = 0 min, and C is the absorbance value at particular intervals of time (t = 15 min, 30 min, 60 min, 120 min, etc.).
The plot of degradation efficiency vs. time for both ZnO and TiO 2 nanotubes is presented in Figure 7a. It was observed that a total bleaching of the RhB solution was obtained for TiO 2 nanotubes after prolonged irradiation time, as illustrated in Figure 7b. The photocatalytic degradation efficiency is 95% for the TiO 2 nanotubes and 55% for the ZnO nanotubes, respectively. The degradation efficiency values of the nanotubes are consistent with data previously reported in the literature for ZnO and TiO 2 nanostructures [28,37,[45][46][47].
After increasing the reaction time, the photodegradation rate of the organic dye decreases. In order to observe the kinetics of the degradation, the degradation efficiency was calculated using the following equation: Degradation efficiency (%) = (C0 − C)/C0 × 100. where C0 is the absorbance value before irradiation at t = 0 min, and C is the absorbance value at particular intervals of time (t = 15 min, 30 min, 60 min, 120 min, etc.).  With the purpose of excluding the influence of secondary processes such as adsorption effects and to prove that the degradation of Rhodamine B is not due to adsorption into the nanotubes, experiments were performed in dark. Thus, ZnO nanotubes were immersed in RhB solution in the dark for a total of 36 h (Figure 8a). After sample removal, there was no indication that adsorption processes took place (Figure 8b), and the intensity of the absorption band of the Rhodamine B band after 36 h was similar to the initial intensity (Figure 8c). Similarly, TiO 2 nanotubes were immersed in Rhodamine B solution (Figure 8d), and no indication of the adsorption effect was observed after 36 h in the dark (Figure 8e), with no diminishing in the dye's absorption band (Figure 8f). mersed in RhB solution in the dark for a total of 36 h (Figure 8a). After sample removal, there was no indication that adsorption processes took place (Figure 8b), and the intensity of the absorption band of the Rhodamine B band after 36 h was similar to the initial intensity (Figure 8c). Similarly, TiO2 nanotubes were immersed in Rhodamine B solution (Figure 8d), and no indication of the adsorption effect was observed after 36 h in the dark (Figure 8e), with no diminishing in the dye's absorption band (Figure 8f).  In order to explain the mechanism of degradation of Rhodamine B in the presence of the nanotubes, the formation of the oxidant agents in the mechanism of degradation of Rhodamine B is presented schematically, using TiO 2 nanotubes to exemplify the processes. Figure 9 illustrates the types of reactions (oxidative and reduction) that take place in the presence of the metal oxide semiconductor and the photocatalysis process.
In photocatalytic degradation based on semiconducting nanotubes, a photon with an energy higher than the energy of the semiconductor's band gap activates the electron in the valence band to make the band-to-band transition, producing an electron-hole pair. The charge carriers from the surface of the nanotubes interact with the chemicals in the aqueous solution, degrading them and forming the reaction products. The electron in the conduction band is involved in the reduction reaction of O 2 , while the hole generated in the valence band participates in the oxidation reaction that produces the hydroxyl ion, as illustrated in Figure 9. Both types of oxidant agents may interact with Rhodamine B molecules leading to, after the formation of intermediate products, H 2 O and CO 2 molecules.
Although the pure ZnO and TiO 2 have lower degradation efficiencies when compared with doped ZnO and TiO 2 [48], nanostructuring these types of materials, i.e., reducing their dimensions, leads to larger specific surface areas. The nanotubes offer a doubled area for photocatalytic reactions. Moreover, the granulated morphology further increases that area. This is confirmed by the larger degradation efficiency of the TiO 2 nanotubes (~95%) compared with one of the ZnO nanotubes (~55%) due to the thinner walls observed by FESEM evaluations in the case of TiO 2 nanotubes.
The main advantage offered by the ZnO and TiO 2 nanotubes for their use as photocatalysts for pollutants' degradation is that they can be easily removed from the solution at the end of the degradation process. Further, the nanotubes can be used either immobilized on a substrate or by insertion in a porous composite. In order to explain the mechanism of degradation of Rhodamine B in the presence of the nanotubes, the formation of the oxidant agents in the mechanism of degradation of Rhodamine B is presented schematically, using TiO2 nanotubes to exemplify the processes. Figure 9 illustrates the types of reactions (oxidative and reduction) that take place in the presence of the metal oxide semiconductor and the photocatalysis process.
In photocatalytic degradation based on semiconducting nanotubes, a photon with an energy higher than the energy of the semiconductor's band gap activates the electron in the valence band to make the band-to-band transition, producing an electron-hole pair. The charge carriers from the surface of the nanotubes interact with the chemicals in the aqueous solution, degrading them and forming the reaction products. The electron in the conduction band is involved in the reduction reaction of O2, while the hole generated in ecules leading to, after the formation of intermediate products, H2O and CO2 molecules.
Although the pure ZnO and TiO2 have lower degradation efficiencies when compared with doped ZnO and TiO2 [48], nanostructuring these types of materials, i.e., reducing their dimensions, leads to larger specific surface areas. The nanotubes offer a doubled area for photocatalytic reactions. Moreover, the granulated morphology further increases that area. This is confirmed by the larger degradation efficiency of the TiO2 nanotubes (~95%) compared with one of the ZnO nanotubes (~55%) due to the thinner walls observed by FESEM evaluations in the case of TiO2 nanotubes. The main advantage offered by the ZnO and TiO2 nanotubes for their use as photocatalysts for pollutants' degradation is that they can be easily removed from the solution at the end of the degradation process. Further, the nanotubes can be used either immobilized on a substrate or by insertion in a porous composite.

Conclusions
A three-step fabrication process was used in order to obtain semiconducting nanotubes for optoelectronic and photocatalytic applications. Polymer fibers produced by electrospinning were covered over their entire surface via RF magnetron sputtering with a thin layer of semiconductor oxide. After being used as templates, the polymer fibers were removed by calcination leading ZnO and TiO2 nanotubes to assemble in web-like architectures. The single crystalline phase, wurtzite ZnO, with a band gap of 3.05 eV and intense luminescence, and anatase TiO2 with a band gap of 3.16 eV present interesting photocatalytic properties demonstrated by Rhodamine B degradation under irradiation with white light of a solar simulator, having a degradation efficiency of ~55% for ZnO nanotubes and ~95% for TiO2 nanotubes. The interesting capabilities of the ZnO and TiO2 nanotubes are given by the fact that the photocatalytic process takes place up to the total bleaching of contaminants in solutions, even for a small number of nanotubes, and by the possibility of removal of web-like constructed nanotubes after the degradation process.

Conclusions
A three-step fabrication process was used in order to obtain semiconducting nanotubes for optoelectronic and photocatalytic applications. Polymer fibers produced by electrospinning were covered over their entire surface via RF magnetron sputtering with a thin layer of semiconductor oxide. After being used as templates, the polymer fibers were removed by calcination leading ZnO and TiO 2 nanotubes to assemble in web-like architectures. The single crystalline phase, wurtzite ZnO, with a band gap of 3.05 eV and intense luminescence, and anatase TiO 2 with a band gap of 3.16 eV present interesting photocatalytic properties demonstrated by Rhodamine B degradation under irradiation with white light of a solar simulator, having a degradation efficiency of~55% for ZnO nanotubes and~95% for TiO 2 nanotubes. The interesting capabilities of the ZnO and TiO 2 nanotubes are given by the fact that the photocatalytic process takes place up to the total bleaching of contaminants in solutions, even for a small number of nanotubes, and by the possibility of removal of web-like constructed nanotubes after the degradation process.