Floating TiO 2 -Cork Nano-Photocatalysts for Water Puriﬁcation Using Sunlight

: In the present study, titanium dioxide (TiO 2 ) nano-photocatalysts were synthesized through microwave irradiation. In a typical microwave synthesis, TiO 2 nanomaterials were simultaneously produced in powder form and also directly covering cork substrates. The TiO 2 nanopowder was analyzed by X-ray diffraction (XRD), Raman spectroscopy and transmission electron microscopy (TEM), revealing that the solvothermal microwave synthesis resulted only in the TiO 2 anatase phase. From Fourier-transform infrared spectroscopy (FTIR), cork’s organic species, along with bands of TiO 2 , were detected. UV–VIS absorption spectrum revealed an absorption extension to the visible region, since a brown powdered TiO 2 product was obtained. Very ﬁne nanoparticles were observed displaying a nearly spherical shape that agglomerates in larger particles. These larger particles fully covered the surface of the honeycomb cork cells, originating TiO 2 functionalized cork platforms. The TiO 2 functionalized substrates were further tested as ﬂoating photocatalysts and their photocatalytic activity was assessed from rhodamine B degradation under solar simulating light and natural sunlight. Reusability tests were also performed under natural sunlight. The strategy applied in this research work allowed the production of green and low-cost cork platforms based on TiO 2 photoactive materials with the ability to purify polluted water under natural sunlight.


Introduction
The rapid development of industrialization and population growth have strongly contributed to the shortage of available clean water, thus increasing the demand for reliable and accessible solutions that enable water purification [1]. The uncontrolled discharge of toxic contaminants into the environment, which include organic pollutants, heavy metals, inorganic compounds, and other complex compounds, has significantly contributed to water contamination [2][3][4].
In this regard, significant research has been carried out to find economically viable and efficient solutions to face the issue of water pollution. Textile industries significantly contribute to this problem since around 60 thousand tons of textile dyes are discharged annually into the environment [5]. These textile dyes do not bind tightly to fabric, and, consequently, due to their high recalcitrance, they impair photosynthesis, inhibit plant growth, and enter the food chain, endangering animals and plants while promoting toxicity, mutagenicity, and carcinogenicity [6].
Among the most used methods for the treatment of pollutants and remediation of water, adsorption and membrane technologies, as well as biological treatments and advanced oxidation methods, appear as effective treatment methods for wastewater or water sources. complex experimental equipment or stirring during the photocatalytic reaction. Under these circumstances, an optimization of illumination and oxygenation should, hence, improve the radical formation rate and oxidation efficiency [50,51].
In terms of natural floating materials, cork presents several advantages, which makes it suitable in photocatalytic experiments, such as earth-abundance, compressibility, resilience, thermal stability and corrosion resistance, low density [52], non-toxicity and hydrophobicity [50]. Cork is extracted from the bark of the cork oak tree (Quercus suber L.) which flourishes in the specific regions of the Western Mediterranean (Portugal, Spain, Southern France, part of Italy, and North Africa) [53], and by being sustainable and renewable [52], it has become a valuable material. As proof of its sustainability, each time the bark is harvested from cork trees, these latter ones can absorb up to 5 times more CO 2 (with the aim of generating their bark) than a non-harvested tree [54]. Moreover, cork contributes to the circular economy, being extracted without damaging the tree and being regenerated over the years, besides being the base of several products, some of them produced by recycled cork [55]. In terms of cork's composition, it depends on several factors, such as geographic origin, climate and soil conditions, tree dimensions, age, among others. However, it is essentially composed by hydrophobic polymers such as suberin (which is mainly responsible for the elasticity and compressibility of cork representing~46% [56]) and lignin (~27% [56]), as well as by the hydrophilic polysaccharides (~12%), cellulose and hemicellulose [53,57], and extractives (~15% [58]) [2], s waxes and tannins [59].
Portugal is the world's largest producer of cork, with an estimated annual production of around 100 thousand tons. The Cork industry, despite being one of the most profitable markets in the country, generates a large amount of cork that has no commercial value, being usually burnt to generate energy [54]. The combination of green advanced materials and technology, which can tackle issues of sustainability, such as water pollution, has to be adopted for a sustainable future [60]. Hence, instead of wasting cork and to take advantage of its unique properties, it could serve as a platform in photocatalytic applications for the treatment of contaminated effluents, which would valorize both the cork industry and the effluent treatment industry.
Studies focused on the production of TiO 2 -based materials on cork and the investigation of their photocatalytic efficiency are scarce, and to the best of our knowledge, the synthesis of nano-TiO 2 photocatalysts on cork by a fast and low temperature microwave synthesis has never been reported. Regarding the previous studies, Sboui et al. reported the preparation of TiO 2 -Polyaniline (PANI) nanocomposites for the degradation of methyl orange (MO) dye by using a sol-gel method for the synthesis of TiO 2 nanostructures, followed by an additional step to immobilize the nanocomposite on cork surface [50]. In another study, TiO 2 /polyvinyl alcohol (PVA)/Cork nanocomposites were prepared through the immobilization of TiO 2 nanoparticles on cork and by employing PVA as a binder to anchor TiO 2 nanoparticles on the surface of the floating catalyst for the degradation of methylene blue (MB) [61]. For the photocatalyst preparation, several steps were involved, and higher temperatures were required to calcinate the TiO 2 nanostructures. Therefore, in this work, pure TiO 2 photocatalysts were synthesized on cork substrates by a fast and low temperature single step solvothermal method under microwave irradiation (120 • C for 1 h). The produced materials (nanopowder and/or functionalized substrates) were systematically characterized by XRD, Raman spectroscopy, FTIR, SEM coupled with EDS, and TEM. The optical properties were assessed through UV-VIS absorption measurements. Finally, the photocatalytic activity of the TiO 2 functionalized cork platforms was evaluated from the evolution of rhodamine B (RhB) degradation under solar simulated light and natural sunlight.  64-17-5) and stirred for 10 min, until a homogeneous solution was obtained. Afterwards, 4 mL of TTIP was added dropwise to the previous solution. Then, the obtained solution was left to stir overnight. This process guarantees the complete homogenization of the solution. Microwave synthesis was performed using a CEM microwave digestion system, Matthews, NC, USA (MARS one) and was carried out at 1000 W, 120 ± 10 • C for 1 h. Solution volumes of 20 mL were transferred into Teflon vessels of 75 mL, which were kept sealed. To produce the functionalized cork platforms, cork sheets with a thickness of 2 mm from Bi-Office (ref. RL044330 222468) were cut into pieces of 1.5 × 3.5 cm 2 , fixed with Kapton tape to ordinary glass squares and each cork piece was transferred to individual microwave vessels containing the solution. Each piece should be completely in contact with the solution. The functionalized cork platforms were afterwards cleaned in an ultrasonic bath, first with deionized water for 10 min and then with isopropyl alcohol (IPA) for another 10 min. To remove the excess solvents from the cork substrates, each piece was left to dry for a few minutes on top of lab paper and dried at 60 • C in a desiccator between 30 min and 2 h. TiO 2 nanopowder was simultaneously formed during microwave synthesis. The nanopowder was washed repeatedly several times with deionized water, followed by isopropyl alcohol (IPA), using a centrifuge at 4500 rpm for 5 min each time. Finally, the nanopowder was kept overnight in a desiccator for drying at 80 • C in vacuum. For comparison, a microwave synthesis was carried out without the cork substrates, and pure TiO 2 nanopowder was obtained by using the same experimental conditions as the TiO 2 nanopowder produced in the presence of cork substrates. A schematic of the experimental procedure is shown in Figure 1.

Synthesis of TiO2 Nanostructures on Cork via Microwave Irradiation
Titanium (IV) isopropoxide (TTIP) with 97% purity from Sigma-Aldrich, St. Louis, MO, USA (CAS: 546-68-9) was used as the titanium dioxide precursor. To prepare a solution of 120 mL, 10 mL of a 1 M acid solution (oxalic acid anhydrous from Sigma-Aldrich (CAS: 144-62-7) was added to 110 mL of absolute anhydrous ethanol from Carlo Erba reagents, Val-de-Reuil, France (CAS: 64-17-5) and stirred for 10 min, until a homogeneous solution was obtained. Afterwards, 4 mL of TTIP was added dropwise to the previous solution. Then, the obtained solution was left to stir overnight. This process guarantees the complete homogenization of the solution. Microwave synthesis was performed using a CEM microwave digestion system, Matthews, NC, USA (MARS one) and was carried out at 1000 W, 120 ± 10 °C for 1 h. Solution volumes of 20 mL were transferred into Teflon vessels of 75 mL, which were kept sealed. To produce the functionalized cork platforms, cork sheets with a thickness of 2 mm from Bi-Office (ref. RL044330 222468) were cut into pieces of 1.5 × 3.5 cm 2 , fixed with Kapton tape to ordinary glass squares and each cork piece was transferred to individual microwave vessels containing the solution. Each piece should be completely in contact with the solution. The functionalized cork platforms were afterwards cleaned in an ultrasonic bath, first with deionized water for 10 min and then with isopropyl alcohol (IPA) for another 10 min. To remove the excess solvents from the cork substrates, each piece was left to dry for a few minutes on top of lab paper and dried at 60 °C in a desiccator between 30 min and 2 h. TiO2 nanopowder was simultaneously formed during microwave synthesis. The nanopowder was washed repeatedly several times with deionized water, followed by isopropyl alcohol (IPA), using a centrifuge at 4500 rpm for 5 min each time. Finally, the nanopowder was kept overnight in a desiccator for drying at 80 °C in vacuum. For comparison, a microwave synthesis was carried out without the cork substrates, and pure TiO2 nanopowder was obtained by using the same experimental conditions as the TiO2 nanopowder produced in the presence of cork substrates. A schematic of the experimental procedure is shown in Figure 1. It should be noted that the final functionalized cork exhibited a slightly darker color after microwave synthesis. This has also been previously reported owing to an increase in temperature, assigned to the reactions of the extractives, without any degradation associated [62].

Characterization Techniques
XRD experiments were performed using a PANalytical's X'Pert PRO MPD diffractometer (Almelo, The Netherlands) equipped with a X'Celerator 1D detector and It should be noted that the final functionalized cork exhibited a slightly darker color after microwave synthesis. This has also been previously reported owing to an increase in temperature, assigned to the reactions of the extractives, without any degradation associated [62].
TEM observations were carried out with a Hitachi HF5000 field-emission transmission electron microscope operated at 200 kV (Mito, Japan). A drop of the sonicated dispersion was deposited onto 200-mesh copper grids covered with formvar and allowed to dry before observation. The average particle size and standard deviation were calculated from the dimensions of several nanoparticles (30 measurements) based on TEM images. Image J software was used to measure the nanoparticles.The absorption measurements were carried out using a Perkin Elmer lambda 950 UV/VIS/NIR spectrophotometer (Waltham, MA, USA) and the measurements were performed in the 250-800 nm range.
Raman spectroscopy measurements were obtained with an inVia Qontor confocal Raman microscope from Renishaw (Kingswood, UK). A 50 mW green diode pumped solid state laser operated at 532 nm was used as the excitation source, with a 10 s exposure time and settings of 3 accumulations. The Raman spectra were recorded as an extended scan in the range of 100-700 cm −1 . The laser beam was focused with a long working distance (8.2 mm) 50× Olympus objective. The results present are based on the average of several scans taken on the surface of the synthesized nanopowder and cork substrates. The 521 cm −1 peak of a silicon wafer was used to calibrate the spectrograph for possible fluctuations of the Raman system. All measurements were performed at RT.
FTIR transmittance spectrum of the produced powders was recorded on a Perkin-Elmer FT-IR Spectrometer Spectrum Two (Waltham, MA, USA), equipped with an attenuated total reflection (ATR) cell in the range of 400-4000 cm −1 . The measurements were performed at RT.

Photocatalytic RhB Degradation under Solar Simulating Light
The photocatalytic activity of TiO 2 on cork was evaluated at RT considering the degradation of RhB (C 28 H 31 ClN 2 O 3 ) from Sigma-Aldrich. The TiO 2 nanomaterials were synthesized directly on cork substrates to avoid the recovery of nanosized materials in powder form. All the experiments were performed according to the international standard ISO 10678 [45]. For each experiment, cork pieces were placed on the reaction recipient and before the reaction, the catalysts and the dye solution (50 mL of the RhB solution (5 mg/L)) were kept in the dark for 1 h to establish absorption-desorption equilibrium. No stirring was used during the reactions to avoid oxygenation of the solution and to simulate the use of a floating photocatalyst [50]. Solar light exposure was carried out by using a light-emitting diode (LED) solar simulator LSH 7320 (AM 1.5 spectrum) with an intensity of 100 mW/cm 2 (1000 W/m 2 ). Absorption spectra were recorded using a PerkinElmer lambda 950 UV/VIS/NIR spectrophotometer with different time intervals up to a total of 12 h. The measurements were performed in the 400-700 nm range. The blank RhB solution and pristine cork substrate were also exposed with identical conditions.

Photocatalytic RhB Degradation under Natural Sunlight
RhB photocatalytic degradation under natural sunlight was monitored around midday in the city of Caparica and in the month of May. The total exposure time was 4 h, since after that time (at the end of the day) the solar intensity significantly decreased. Moreover, if the exposure was continued on another day, the experimental conditions (solar intensity and UV level) would significantly differ, and for that reason, it has been considered 4 h of sunlight exposure for each photocatalytic experiment.
The sunlight intensity was measured three times by using a solar power meter from Sciencetech (Sciencetech-Inc., London, ON, Canada): before light exposure, in the middle and at the end of the experiment. Reusability tests were also performed with the functionalized cork substrates for the RhB degradation under three consecutive cycles and up to 4 h. These experiments consisted of recovering the cork substrates with further discard of the total liquid. The recovered cork pieces were dried in air prior to the next exposure. After that, the recovered materials were poured into a fresh solution and exposed to light considering the same exposure times. The pristine substrate was also exposed for comparison. Before light exposure, the sunlight intensity was 810, 800 and 880 W/m 2 , respectively, for the first, second and third experiments. In the middle of the experiments (after 2 h of light exposure), the intensity decreased to 730, 750 and 810 W/m 2 , respectively, for the first, second and third experiments. At the end of the experiments (after 4 h of light exposure), the sunlight intensity was 400, 350 and 300 W/m 2 , respectively, for the first, second and third experiments. During the photocatalysis experiments, the temperatures were between 17-26 • C, 19-28 • C and 20-26 • C in the first, second and third experiments, respectively. The UV levels were 7 out of 11 in the first experiment and 8 out of 11 in the second and third experiments, according to IPMA (Instituto Português do Mar e da Atmosfera) website [67].

Results and Discussion
Microwave synthesis was successfully used to produce TiO 2 nanopowders and also to synthesize TiO 2 nanostructures directly on cork substrates. The syntheses were carried out using oxalic acid at low temperatures (120 • C) for 1 h and without any seed layer or preliminary preparation step. The TiO 2 nanopowder, as well as the TiO 2 functionalized cork platforms, were systematically investigated, and the cork-based materials were tested as floating photocatalysts.  4 [45,68]. The XRD results also demonstrate that the materials are well crystallized and the broad diffraction peaks indicate the presence of very small sized crystallites [45,68]. Some contribution from the cork's chemical components released during synthesis, especially due to the brown color of the resulting powders, was expected. However, it was not confirmed by the XRD results. The average crystallite size of the synthesized TiO2 nanomaterials was obtained using the Scherrer equation. The calculated crystallite size value was found to be arou The average crystallite size of the synthesized TiO 2 nanomaterials was obtained by using the Scherrer equation. The calculated crystallite size value was found to be around 4.8 nm.

Raman Spectroscopy Measurements
Raman spectroscopy measurements were performed to confirm the purity of the synthesized materials. Since the cork substrates obscured the signal of the TiO 2 nanomaterials (see Figure S1), probably due to the high penetration depth of the laser's light (in this case with a wavelength of 532 nm), only the materials in powder form were investigated. It can be observed from the Raman spectrum in Figure 3 that the five typical TiO 2 anatase bands appear at 144 cm −1 (E g ), 198 cm −1 (E g ), 393 cm −1 (B 1g ), 515 cm −1 (B 1g + A 1g ) and 636 cm −1 (E g ) [68,69] with the six Raman active modes (A 1g + 2B 1g + 3 E g ) [70]. The purity of the produced TiO 2 nanopowder is thus confirmed and corroborates with the XRD data. The average crystallite size of the synthesized TiO2 nanomaterials was obta using the Scherrer equation. The calculated crystallite size value was found to be 4.8 nm.

Raman Spectroscopy Measurements
Raman spectroscopy measurements were performed to confirm the purity synthesized materials. Since the cork substrates obscured the signal of th nanomaterials (see Figure S1), probably due to the high penetration depth of the light (in this case with a wavelength of 532 nm), only the materials in powder for investigated. It can be observed from the Raman spectrum in Figure 3 that the five TiO2 anatase bands appear at 144 cm −1 (Eg), 198 cm −1 (Eg), 393 cm −1 (B1g), 515 cm −1 (B and 636 cm −1 (Eg) [68,69] with the six Raman active modes (A1g + 2B1g + 3 Eg) [70]. Th of the produced TiO2 nanopowder is thus confirmed and corroborates with the XR

FTIR
The TiO 2 nanopowder was also investigated using FTIR and the result is presented in Figure 4. The appearance of a broad band between 3000 and 3600 cm −1 , as well as the band at 1680 cm −1 , indicates the presence of hydroxyl groups, due to the adsorption of water from the surrounding environment [43,50,71,72]. For the lower wavenumber region, a peak located at around 1111 cm −1 is visible and attributed to the stretching vibration of the Ti-OH bond [50], whereas the peaks located at 438 cm −1 and 1396 cm −1 are attributed to Ti-O bending [73] and Ti-O-Ti stretching modes [74], respectively. Other peaks corresponding to vibrational modes of residual organic species were also observed. FTIR stretching band at 1680 cm −1 is likely related to C=O groups [75,76], while the broad and small peak that appears at 1260 cm −1 is probably ascribed to C-O stretching mode from groups such as ethers (common linkages in lignin [77]), alcohols, esters, lactones, phenols or carboxylic acids [78], all present in cork's chemical composition [79][80][81].
The brown nanopowder obtained from microwave synthesis may indicate the presence of organic species coming from the cork substrates (see Figure 1), as revealed by FTIR, and not detected by XRD and Raman spectroscopy techniques. Nevertheless, anatase TiO 2 phase was obtained, as seen previously from the XRD and Raman spectroscopy results, indicating the successful conversion of the TiO 2 precursors under microwave irradiation, which is based on the following Equations (1) and (2) [82]: in Figure 4. The appearance of a broad band between 3000 and 3600 cm −1 , as we band at 1680 cm −1 , indicates the presence of hydroxyl groups, due to the adsor water from the surrounding environment [43,50,71,72]. For the lower wave region, a peak located at around 1111 cm −1 is visible and attributed to the str vibration of the Ti-OH bond [50], whereas the peaks located at 438 cm −1 and 1396 attributed to Ti-O bending [73] and Ti-O-Ti stretching modes [74], respectively peaks corresponding to vibrational modes of residual organic species were also ob FTIR stretching band at 1680 cm −1 is likely related to C=O groups [75,76], while th and small peak that appears at 1260 cm −1 is probably ascribed to C-O stretchin from groups such as ethers (common linkages in lignin [77]), alcohols, esters, l phenols or carboxylic acids [78], all present in cork's chemical composition [79-81 The brown nanopowder obtained from microwave synthesis may indic presence of organic species coming from the cork substrates (see Figure 1), as reve FTIR, and not detected by XRD and Raman spectroscopy techniques. Never anatase TiO2 phase was obtained, as seen previously from the XRD and spectroscopy results, indicating the successful conversion of the TiO2 precursor microwave irradiation, which is based on the following Equations (1) and (2) [82] Ti(OCH(CH) ) ) Figure 5 shows the SEM and TEM images of the TiO2 nanopowder synthesize microwave irradiation. From the SEM image (Figure 5a), it canbe observed the fo of nano-sized particles largely agglomerated. Nevertheless, from TEM, it c perceived that these agglomerates are formed by very fine nanoparticles, displ nearly spherical shape (Figure 5c,d). The average particle size was found to be 4.8 nm. The inset in Figure 5b also shows the particle size distribution, revealing that  (Figure 5a), it canbe observed the formation of nano-sized particles largely agglomerated. Nevertheless, from TEM, it could be perceived that these agglomerates are formed by very fine nanoparticles, displaying a nearly spherical shape (Figure 5c,d). The average particle size was found to be 4.86 ± 1.31 nm. The inset in Figure 5b also shows the particle size distribution, revealing that smaller particles in the range of 4-6 nm are more likely to be found. The ring diffraction pattern in Figure 5b attested that these particles are solely from the anatase phase. Moreover, the fast Fourier transform (FFT) images and the lattice spacing (~0.35 nm) correspond to the (101) planes of anatase TiO 2 nanocrystals, Figure 5c,d. From the FFT image in Figure 5c, taken in the area indicated as A in the black square, it can be observed that two sets of lattices are present and that they are not perpendicular to each other with an equal lattice spacing of~0.35 nm, corresponding to the (101) and (011) planes of the anatase phase. Observed through the [111] zone axis, the angle inferred between both (101) and (011) is 82 • in accordance with the theoretical value reported for pure TiO 2 anatase (JCPDS 21-1272).

Electron Microscopy
The TiO 2 nanopowder synthesized without cork substrates was also observed by SEM ( Figure S2), revealing that, as observed for the TiO 2 nanopowder synthesized with cork substrates, the microwave synthesis resulted in very fine nanoparticles, in agreement with an analogous study [23].
fast Fourier transform (FFT) images and the lattice spacing (~0.35 nm) correspond to the (101) planes of anatase TiO2 nanocrystals, Figure 5c,d. From the FFT image in Figure 5c, taken in the area indicated as A in the black square, it can be observed that two sets of lattices are present and that they are not perpendicular to each other with an equal lattice spacing of ~0.35 nm, corresponding to the (101) and (011) planes of the anatase phase. Observed through the [111 ] zone axis, the angle inferred between both (101) and (011) is 82° in accordance with the theoretical value reported for pure TiO2 anatase (JCPDS 21-1272). The TiO2 nanopowder synthesized without cork substrates was also observed by SEM ( Figure S2), revealing that, as observed for the TiO2 nanopowder synthesized with cork substrates, the microwave synthesis resulted in very fine nanoparticles, in agreement with an analogous study [23].

Optical Characterization
The optical properties of the produced TiO2 nanopowder (in the presence of the cork substrates) were investigated by recording the UV-VIS absorption spectrum (see Figure 6). As observed in Figure 6, an absorption maximum is obtained in the UV region, around

Optical Characterization
The optical properties of the produced TiO 2 nanopowder (in the presence of the cork substrates) were investigated by recording the UV-VIS absorption spectrum (see Figure 6). As observed in Figure 6, an absorption maximum is obtained in the UV region, around 306 nm (~4.1 eV), and the absorption is seen to extend throughout the visible region. For comparison, the absorption spectrum of the TiO 2 nanopowder synthesized in the absence of the cork substrates was also recorded ( Figure S2). In this last case, even though the absorption maximum is placed at the same value as in Figure 6, no absorption occurs beyond 400 nm as expected, since pure TiO 2 presents a wide band gap value (around 3.2 eV for pure TiO 2 anatase nanoparticles [45,83,84]). It can thus be concluded that the visible light absorption in Figure 6 may be attributed to the presence of binders that were transferred from the cork during microwave synthesis and resulted in brown TiO 2 nanopowders. In fact, several studies have reported the impact of different binders on the TiO 2 nanostructures, with these extractives interacting with the nanostructures and modifying their morphology, size and optical properties [85][86][87].
absorption maximum is placed at the same value as in Figure 6, no absorption occurs beyond 400 nm as expected, since pure TiO2 presents a wide band gap value (around 3.2 eV for pure TiO2 anatase nanoparticles [45,83,84]). It can thus be concluded that the visible light absorption in Figure 6 may be attributed to the presence of binders that were transferred from the cork during microwave synthesis and resulted in brown TiO2 nanopowders. In fact, several studies have reported the impact of different binders on the TiO2 nanostructures, with these extractives interacting with the nanostructures and modifying their morphology, size and optical properties [85][86][87].

X-ray Diffraction
XRD analyses were also carried out for the pristine cork substrate and for the TiO2 functionalized cork substrate (Figure 7). It can be observed that the pristine cork substrate shows a broad peak centered at around 2θ = 21°, indicating its amorphous nature [88,89]. When it comes to the functionalized cork substrate, five weak and broad peaks were detected, which can be associated with the presence of anatase. The XRD peaks observed at 25.

X-ray Diffraction
XRD analyses were also carried out for the pristine cork substrate and for the TiO 2 functionalized cork substrate (Figure 7). It can be observed that the pristine cork substrate shows a broad peak centered at around 2θ = 21 • , indicating its amorphous nature [88,89]. When it comes to the functionalized cork substrate, five weak and broad peaks were detected, which can be associated with the presence of anatase. The XRD peaks observed at 25.

Scanning Electron Microscopy
SEM images of the pristine cork substrates together with the functionalized ones are shown in Figure 8. In Figure 8a-c, the pristine cork is shown, which displays several closed and air-hollow unit cells, forming polygons in a regular honeycomb arrangement and the number of sides in these polygons varies. As reported for Quercus suber L., hexagonal,

Scanning Electron Microscopy
SEM images of the pristine cork substrates together with the functionalized ones are shown in Figure 8. In Figure 8a-c, the pristine cork is shown, which displays several closed and air-hollow unit cells, forming polygons in a regular honeycomb arrangement and the number of sides in these polygons varies. As reported for Quercus suber L., hexagonal, pentagonal and heptagonal cells are the most frequent, representing around 95% of the total forms [90]. The diameter of the cork cells is around 15-20 µm (see Figure 8a) and their size depends on the season in which they are formed [90]. For instance, early cork cells (cells formed in the main growth period) are greater in height and have thinner cell walls, while late cork cells (cells formed at the end of the growth period) are smaller and have a thicker wall [91,92]. In the specific case of Quercus suber L., for early cork cells, prism height ranges from 30-40 µm, prism base edges from 13-15 µm, average base area from 4-6 × 10 −6 cm 2 and the cell wall thickness between 1 and 1.5 µm. Late cork cells have a reduced prism height, nearly 10 µm, and the cell wall thickness almost doubles. The number of cells per cm 3 fluctuates from 4 to 7 × 10 7 for early cork and late cork from 10 to 20 × 10 7 [90]. Another important characteristic is the non-uniform undulations of the cell walls, which vary from cell to cell, as well as the roughness of their surface [93]. Some totally corrugated cells are observed, together with a few collapsed cells (see Figure 8a). Concerning the irregularities on the surface of the cells, granulates and small deposits are visible with no defined shape (Figure 8c) [91]. It is also visible that the interior of these cork cells is closed, without internal porosity.

RhB Photocatalytic Degradation under Solar Simulating Light
The RhB decolorization in the presence of the cork substrates (pristine and functionalized with TiO2) was evaluated under solar simulating light up to a total of 12 h. The aliquots of the dye were taken after 1 h (up to 4 h) and, after that time, every 2 h by using the UV-VIS spectrophotometer. A gradual decrease in the absorption peak of RhB (at 554 nm [95,96]) was observed over time. The decolorization rate (%) of the RhB dye was calculated accordingly to Equation (3): After the microwave synthesis (Figure 8d-f), it is possible to observe that the honeycomb arrangement of the cork cells is preserved. Additionally, the TiO 2 agglomerates of nanoparticles are clearly discernible at the surface of the cork, completely filling the cork cells and covering the substrate entirely, Figure 8f.
The chemical composition of the substrates before (pristine cork) and after the microwave synthesis (TiO 2 functionalized cork substrates) was analyzed by EDS ( Figure S3). Figure S3a shows that the pristine cork substrates (C 123 H 182 O 56 N for Quercus suber L. [94]) are mainly composed of C ( Figure S3b) and O ( Figure S3c), consistent with its chemical composition [88]. For the TiO 2 functionalized cork substrates, a uniform distribution of Ti ( Figure S3g The RhB decolorization in the presence of the cork substrates (pristine and functionalized with TiO 2 ) was evaluated under solar simulating light up to a total of 12 h. The aliquots of the dye were taken after 1 h (up to 4 h) and, after that time, every 2 h by using the UV-VIS spectrophotometer. A gradual decrease in the absorption peak of RhB (at 554 nm [95,96]) was observed over time. The decolorization rate (%) of the RhB dye was calculated accordingly to Equation (3): where A 0 is the initial absorbance of the RhB solution before irradiation and A t is the absorbance of RhB solution after a certain exposure time (t) [23,97]. Blank experiments (only RhB solution) were also carried out under solar simulating light and natural sunlight (Figures 9c and 10c). Although a slight increase in the RhB concentration is observed for both cases, possibly due to the evaporation of the solvent over time [68], no degradation of the RhB dye was observed in both cases, as reported in other studies [18,98]. stabilization is visible (Figure 9c). After that time, fewer trap sites on the cork surface could be available, leading to a decrease in the absorption rate. According to the literature, the adsorption kinetics model of organic pollutants on cork presents a convex initial curvature, indicating that the adsorption rate decreases with time: it becomes more difficult for the adsorbate to find free sites to adsorb since the main sites are being occupied [99]. This overall significant adsorption of RhB molecules by the pristine cork, which is visible with and without irradiation in Figure 9a, has also been previously reported [100]. It is also well established that suberin, the main chemical component of cork, is hydrophobic due to the existence of aliphatic substances bound to the aromatic domain as triacylglycerol esters. Lignin, the second most abundant chemical component of cork, is characterized by the presence of aromatic alcohol monolignols and acids (in particular, cinnamic and ferulic acid) [59,101,102]. It is reported that the aromatic domain of both Meanwhile, with the TiO2 functionalized cork substrate (Figure 9b), the decolorization efficiency was substantially increased to 76% for the same exposure time. This indicates that the synthesized TiO2 nanomaterials effectively contributed to the degradation of RhB.
It is well known that different factors may influence the photocatalytic activity of TiO2, such as the phase structure, crystallite size, specific surface area and exposed crystal facets [106,107]. As confirmed by TEM and SEM, particles in the nanometer range were formed with an average size of ~5 nm. Smaller nanoparticles possess high specific surface area and surface-to-volume ratio [108]. As a consequence, more active sites on the catalyst are available to interact with the pollutant molecules [68]. Moreover, if the particle size is decreased, the distance that photogenerated electron-hole pairs need to travel is reduced, thus reducing the probability of recombination of photogenerated charge carriers [109]. It was also observed by TEM and SEM that these particles formed agglomerates. Such agglomerates influence the specific surface area of the nanoparticles and, consequently, their photocatalytic activity [23,110,111]. Several approaches and synthesis parameters can be used to minimize their formation including pH of the solution, addition of surfactants, among others. Nevertheless, the synthesized TiO2 nanoparticles enhanced the cork's performance on the removal of RhB molecules from water (Figure 9a,b).
The (101) anatase surface is found to dominate the structure of nanosized anatase crystallites due to its energetic preference and higher stability, being labeled as the majority surface [112]. As observed in Figure 5, (101) anatase near-spherical nanocrystals without a specific facet on the crystal surface are largely present in the nanopowder produced. So, it has not been considered any RhB degradation contribution from active anatase facets.
Moreover, in general, it is accepted that for pure TiO2 phases, anatase exhibits superior photocatalytic activity than the other two crystalline phases: rutile and brookite. Rutile and brookite are direct band gap semiconductors, while anatase presents an  Figure 9a shows a RhB decolorization of 44% without any catalyst on the surface of the cork after 12 h. The uptake of RhB molecules by the pristine cork stems from its honeycomb structure and cork cells' surface roughness, increasing the RhB adsorption capacity during the first few hours. However, after 8 h of light exposure, a tendency for stabilization is visible (Figure 9c). After that time, fewer trap sites on the cork surface could be available, leading to a decrease in the absorption rate. According to the literature, the adsorption kinetics model of organic pollutants on cork presents a convex initial curvature, indicating that the adsorption rate decreases with time: it becomes more difficult for the adsorbate to find free sites to adsorb since the main sites are being occupied [99]. This overall significant adsorption of RhB molecules by the pristine cork, which is visible with and without irradiation in Figure 9a, has also been previously reported [100].
It is also well established that suberin, the main chemical component of cork, is hydrophobic due to the existence of aliphatic substances bound to the aromatic domain as triacylglycerol esters. Lignin, the second most abundant chemical component of cork, is characterized by the presence of aromatic alcohol monolignols and acids (in particular, cinnamic and ferulic acid) [59,101,102]. It is reported that the aromatic domain of both suberin and lignin will determine the interaction between cork and the organic compounds, such as RhB [103], which are essentially hydrophobic [59,104]. As a result, these hydrophobic regions in cork will confer its special affinity to organic pollutants, and, hence, it is expected that an increase in the diffusion of hydrophobic dye molecules into cork will occur [104,105].
Meanwhile, with the TiO 2 functionalized cork substrate (Figure 9b), the decolorization efficiency was substantially increased to 76% for the same exposure time. This indicates that the synthesized TiO 2 nanomaterials effectively contributed to the degradation of RhB.
It is well known that different factors may influence the photocatalytic activity of TiO 2 , such as the phase structure, crystallite size, specific surface area and exposed crystal facets [106,107]. As confirmed by TEM and SEM, particles in the nanometer range were formed with an average size of~5 nm. Smaller nanoparticles possess high specific surface area and surface-to-volume ratio [108]. As a consequence, more active sites on the catalyst are available to interact with the pollutant molecules [68]. Moreover, if the particle size is decreased, the distance that photogenerated electron-hole pairs need to travel is reduced, thus reducing the probability of recombination of photogenerated charge carriers [109]. It was also observed by TEM and SEM that these particles formed agglomerates. Such agglomerates influence the specific surface area of the nanoparticles and, consequently, their photocatalytic activity [23,110,111]. Several approaches and synthesis parameters can be used to minimize their formation including pH of the solution, addition of surfactants, among others. Nevertheless, the synthesized TiO 2 nanoparticles enhanced the cork's performance on the removal of RhB molecules from water (Figure 9a,b).
The (101) anatase surface is found to dominate the structure of nanosized anatase crystallites due to its energetic preference and higher stability, being labeled as the majority surface [112]. As observed in Figure 5, (101) anatase near-spherical nanocrystals without a specific facet on the crystal surface are largely present in the nanopowder produced. So, it has not been considered any RhB degradation contribution from active anatase facets.
Moreover, in general, it is accepted that for pure TiO 2 phases, anatase exhibits superior photocatalytic activity than the other two crystalline phases: rutile and brookite. Rutile and brookite are direct band gap semiconductors, while anatase presents an indirect band gap. Therefore, anatase TiO 2 has a longer lifetime of photogenerated electrons and holes. Additionally, the average effective mass of photogenerated electrons is the lightest compared to rutile and brookite, and, as a result, it exhibits the fastest migration of photogenerated electrons and holes from the interior to the surface of anatase particles. Hence, a lower recombination rate is expected with pure TiO 2 anatase, thus showing a better photocatalytic performance [106,113].
The RhB decolorization process in the presence of cork substrates was investigated by using the Langmuir-Hinshelwood model (Equation (4)): in which r represents the degradation rate, C the concentration of the organic pollutant, and K ad the adsorption equilibrium constant. After integration in the interval [C, C 0 ], it is simplified to the pseudo-first-order-kinetics equation with an apparent rate constant (k ap = k r K ad ), and if we consider that the adsorption is weak, as well as the concentration of organic pollutants, the factor K ad C can be negligible and, thus, Equation (5) is obtained [110]: where k ap is the photodegradation apparent rate constant, t is the time, C 0 is the initial concentration and C is the concentration at a certain time [23,114]. Based on Equation (5), the rate constants k ap (min −1 ) can be determined by plotting − ln C C 0 versus time [115], where the slope of the linear regressions are the apparent rate constants [23].
The obtained photodegradation apparent rate constant was found to be 0.0019 min −1 with TiO 2 functionalized substrate (Figure 9e), and it can be concluded that the photocatalytic dye degradation follows the pseudo-first-order kinetics, since a good correlation for the fitted line was obtained (R 2 above 0.95 [116]). In contrast with the TiO 2 functionalized substrate, the pristine cork substrate (Figure 9d) did not show satisfactory fitting for the pseudo-first-order equation (which has also already been reported [100]), but fitted well (R 2 = 0.99) the pseudo-second-order equation, described with the following equation: Based on Equation (6), the rate constant can be obtained by a linear fit through the plot of 1 C − 1 C 0 versus time [117] and a value of 0.0012 L mg −1 min −1 was obtained. As mentioned before, this trend can be explained by a fast adsorption of RhB molecules in the first 8 h (480 min), followed by a slower uptake and a subsequent stabilization, since fewer adsorption sites are available.

RhB Photocatalytic Degradation under Natural Sunlight
RhB decolorization under natural sunlight was also monitored with the cork substrates (pristine and functionalized with TiO 2 ) within 4 h (240 min) ( Figure 10). For comparison, the degree of RhB decolorization under solar simulating light up to 4 h, in the presence of both cork substrates, was also calculated and is summarized in Table 1. As indicated before, no stirring nor oxygenation were needed during the reactions. Figure S4 shows the floating characteristic of cork during the photocatalytic experiments under natural sunlight. The substrates can thus be easily collected from the solution.  Table 1 shows that in the presence of the pristine substrate, a difference in the RhB decolorization is observed: 22% versus 39% of degradation under solar simulating light and natural sunlight, respectively. Cork's chemical components may degrade upon prolonged UV light irradiation [118]. In fact, the absorption of UV light by pristine cork under natural sunlight might lead to the photooxidation of cellulose, hemicellulose, and degradation of some lignin content present in cork, which could generate chromophore groups, such as carboxylic acids, quinones and hydroperoxide radicals. As a result, these chromophore groups could trigger photochemical reactions on the cork surface, thus promoting a faster decolorization of RhB, in comparison with solar simulating light exposure [119][120][121]. For the TiO 2 functionalized cork substrates, as observed, an increase in decolorization efficiency by 31% was obtained under natural sunlight when compared to solar simulating for 4 h. In contrast with the solar spectrum, which is composed of around 5% UV light, 43% visible light and 50% IR light [122], the LED solar simulator used in the photocatalytic experiments mainly provides visible light (the irradiance is cut for shorter wavelengths (high UV energy)), as observed in Figure S5. Since TiO 2 anatase has a wide band gap value (~3.2 eV), it is mostly photoactive in the UV region [123], and so the difference may be due to the UV light contribution of natural sunlight. This is also supported by the obtained apparent rate constants with the TiO 2 functionalized substrate under natural sunlight, as visible in Figure 10 and under solar simulating light (Figure 9e). The reaction was 2.5 times faster under natural sunlight (k ap = 0.0048 min −1 ) than under the solar simulating light (0.0019 min −1 ) in a third of the time (12 h under solar simulating light compared to 4 h under natural sunlight).

Reusability Tests under Natural Sunlight
Reusability tests are of great importance to evaluate the stability of the photocatalyst [124]. Since the best results were obtained under natural sunlight, reusability tests with the TiO 2 functionalized cork substrate were performed under three consecutive cycles within 4 h (240 min) under natural sunlight. Figure 11 shows the decolorization ratio (C/C o ) of the RhB dye, the pseudo-first-order model fitting for RhB decolorization under three cycles and a column chart of RhB decolorization efficiency with the number of uses. A summary of the obtained results is also presented in Table 2.

Conclusions
TiO2 nanomaterials were directly synthesized on cork substrates by a fast and low temperature solvothermal method under microwave irradiation (120 °C for 1 h), without any preliminary process by using oxalic acid. These nanomaterials were characterized in powder form, as well as functionalized on cork substrates. The characterization of TiO2 nanopowder by XRD, Raman spectroscopy and TEM revealed the presence of the TiO2 anatase structure. Although some FTIR bands indicated the presence of cork's organic species on TiO2 nanopowder, crystalline TiO2 anatase nanoparticles were formed on cork, as proved by XRD data. The obtained brown powdered TiO2 material showed a broad absorption band peaked at ~306 nm, which extended through the visible range up to 800 nm. SEM and EDS analyses of the TiO2 functionalized substrates showed the formation of nearly spherical TiO2 nanoparticles that uniformly covered the honeycomb cells of cork. The best RhB decolorization efficiency was obtained with the TiO2 functionalized cork substrate under natural sunlight, and a RhB degradation of 67% was obtained compared to 39% with the pristine cork substrate, in 4 h of natural light exposure. In the presence of the TiO2 functionalized substrates, the photocatalytic reaction was 2.5 times faster under natural sunlight than under solar simulating light, in a third of the time (12 h under solar simulating light compared to 4 h under natural sunlight). In summary, the results presented show that by using a simple microwave approach, cork floating catalysts based on TiO2-materials can be employed for removing organic pollutants from water under natural sunlight.
Supplementary Materials: The following supporting information can be downloaded at: www.mdpi.com/xxx/s1, Figure S1: Raman spectra of the cork substrates (pristine and TiO2 functionalized one). Figure S2: Absorption curve of the synthesized TiO2 nanopowder (in the absence of the cork substrates). The inset shows the SEM image of the TiO2 nanopowder. Figure    In Table 2 and Figure 11, a decrease in photocatalytic activity was observed with the number of exposures. The reaction rate decreased from 0.005 min −1 in the first cycle to 0.003 min −1 in the third cycle. After performing three consecutive cycles, a decolorization efficiency loss of around 22% was obtained. Since neither heat/UV treatments nor rinsing were applied to the substrate between cycles, a high percentage of RhB molecules could have remained from the previous cycle adsorbed on the photocatalyst surface, hindering the available pore sites on the TiO 2 functionalized cork substrate for reaction [13,125], thus decreasing its overall cycling efficiency. This effect has been previously reported in different studies [13,23,125,126].
It is worth mentioning that this work demonstrates the real contribution of the RhB molecules' adsorption on pristine cork substrates, with a direct effect on the overall photocatalytic activity of the floating photocatalysts.

Conclusions
TiO 2 nanomaterials were directly synthesized on cork substrates by a fast and low temperature solvothermal method under microwave irradiation (120 • C for 1 h), without any preliminary process by using oxalic acid. These nanomaterials were characterized in powder form, as well as functionalized on cork substrates. The characterization of TiO 2 nanopowder by XRD, Raman spectroscopy and TEM revealed the presence of the TiO 2 anatase structure. Although some FTIR bands indicated the presence of cork's organic species on TiO 2 nanopowder, crystalline TiO 2 anatase nanoparticles were formed on cork, as proved by XRD data. The obtained brown powdered TiO 2 material showed a broad absorption band peaked at~306 nm, which extended through the visible range up to 800 nm. SEM and EDS analyses of the TiO 2 functionalized substrates showed the formation of nearly spherical TiO 2 nanoparticles that uniformly covered the honeycomb cells of cork. The best RhB decolorization efficiency was obtained with the TiO 2 functionalized cork substrate under natural sunlight, and a RhB degradation of 67% was obtained compared to 39% with the pristine cork substrate, in 4 h of natural light exposure. In the presence of the TiO 2 functionalized substrates, the photocatalytic reaction was 2.5 times faster under natural sunlight than under solar simulating light, in a third of the time (12 h under solar simulating light compared to 4 h under natural sunlight). In summary, the results presented show that by using a simple microwave approach, cork floating catalysts based on TiO 2 -materials can be employed for removing organic pollutants from water under natural sunlight.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/su14159645/s1, Figure S1: Raman spectra of the cork substrates (pristine and TiO 2 functionalized one). Figure S2: Absorption curve of the synthesized TiO 2 nanopowder (in the absence of the cork substrates). The inset shows the SEM image of the TiO 2 nanopowder. Figure S3: SEM images of the cork substrates: (a) pristine substrate and (d) TiO 2 functionalized substrate. The corresponding EDS maps of C ((b) and (e)), O ((c) and (f)) and Ti (g) are also visible. Figure S4: Photocatalytic experiments with the cork substrates under natural sunlight: (a) pristine substrate and (b) TiO 2 functionalized substrate. Figure S5: Spectral irradiance versus wavelength (from 400 to 1100 nm) of the MiniSol model LSH-7320 and the reference AM 1.5.
Author Contributions: M.L.M. was responsible for producing and characterizing the materials, as well as writing the manuscript. M.M. and F.X.V. were responsible for the photocatalytic experiments. A.P. and D.N. were responsible for the overall scientific orientation and revising the manuscript. D.N. was also responsible for the TEM measurements. A.S.R.M. and J.R. participated in the review and editing. E.F. and R.M. were responsible for supervising all the process and getting funding. All authors have read and agreed to the published version of the manuscript.

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.

Conflicts of Interest:
The authors declare no conflict of interest.