Sustainable Solar Light Photodegradation of Diclofenac by Nano- and Micro-Sized SrTiO 3

: Currently, photocatalytic reactions under solar illumination have attracted worldwide attention due to the tremendous set of associated environmental problems. Taking sunlight into ac-count, it is indispensable to develop highly effective photocatalysts. Strontium titanate, SrTiO 3 (STO), is a cubic perovskite-type semiconductor, an inexpensive material with high thermal stability and corrosion resistance that exhibits a similar energy bandgap to TiO 2 and can represent an interesting alternative in photocatalytic applications. Particle size can signiﬁcantly affect both photocatalytic and photoelectrochemical properties of a photocatalyst, thus altering the photooxidation of organic pollutants in air or water. In this context, this research aims at investigating the photocatalytic features of nano- and micro-sized commercial STO powders towards the photodegradation of diclofenac (DFC), a non-steroidal, anti-inﬂammatory drug, widely used as analgesic, antiarthritic, and antirheumatic. Both nano- and micro-STO photocatalysts exhibited remarkable photocatalytic efﬁciency towards DCF, reaching photodegradation efﬁciency higher than 90% within one hour. Results obtained in simulated drinking water were also compared to those obtained in ultrapure water. Both STOs showed good stability during recycling tests, maintaining high performances after three cycles. Eventually, active species were identiﬁed using various scavengers by trapping holes and radicals generated during the photocatalytic degradation process.


Introduction
In the last decades, rapid industrialization and urbanization have raised the demand for new technologies for water decontamination to improve the health standards of world population [1][2][3][4]. In this context, it has been demonstrated that the main source of water pollution comes from sewage discharge and contaminated effluents in watercourses. Recent statistics report that lack of proper sanitation forces 2.4 billion people daily to use unsafe sources of water, experiencing diseases such as hepatitis, dysentery, and cholera, which in the worst cases can lead to cancer or death [5,6].
Traditionally, several treatment methods have been used to purify water, such as sedimentation, filtration, adsorption, biological digestion, and chemical disinfection, but, unfortunately, they could be insufficient in the removal of persistent pollutants [2,[7][8][9][10]. For these reasons, researchers are looking for economic and smart techniques able to uniformly eliminate all the dangerous molecules located in water matrices. Among all the possible approaches, advance oxidation processes (AOPs), such as heterogeneous photocatalysis, offer green and simple solutions to oxidize organic compounds to CO 2 [11][12][13][14]. In this field, the most studied photocatalyst is titanium dioxide (TiO 2 ), a semiconductor extremely active under UV light in producing oxidative hydroxyl radicals (OH) [15][16][17][18][19][20]. Despite As shown in Figure 1, their diffraction patterns fit to peak positions of the standard cubic perovskite structure (space group: Pm3m) of STO (JCPDS n. , whereas the other peaks can be attributed to traces of SrCO3 (JCPDS 05-0418). Moreover, as expected, if compared to those of STOm, the diffraction peaks of STOn show weakened intensity which, coupled to the peak broadening, suggest smaller particle size, as furtherly confirmed by Scherrer equation: 98.3 nm and 30.5 nm for STOm and STOn, respectively. N2 adsorption/desorption isotherms collected at −196 °C ( Figure 2) are of Type IV, according to the IUPAC classification [60,61]. They possess a hysteresis zone (H3 type, de Boer's classification), typical of mesoporous materials, and only minor microporosity [62,63]. In particular, STOm is characterized by a lower value of specific surface area than STOn (respectively, 2 vs. 21 m 2 ·g −1 ), in accordance with XRPD results. As shown in Figure 1, their diffraction patterns fit to peak positions of the standard cubic perovskite structure (space group: Pm3m) of STO (JCPDS n. , whereas the other peaks can be attributed to traces of SrCO 3 (JCPDS 05-0418). Moreover, as expected, if compared to those of STOm, the diffraction peaks of STOn show weakened intensity which, coupled to the peak broadening, suggest smaller particle size, as furtherly confirmed by Scherrer equation: 98.3 nm and 30.5 nm for STOm and STOn, respectively. N 2 adsorption/desorption isotherms collected at −196 • C ( Figure 2) are of Type IV, according to the IUPAC classification [60,61]. They possess a hysteresis zone (H3 type, de Boer's classification), typical of mesoporous materials, and only minor microporosity [62,63]. In particular, STOm is characterized by a lower value of specific surface area than STOn (respectively, 2 vs. 21 m 2 ·g −1 ), in accordance with XRPD results.
Surface properties (i.e., chemical composition and state of the main species present over sample surfaces) were investigated by X-ray photoelectron spectroscopy (XPS).
The obtained results are summarized in Table S1 and Figure 3, Figure 4 and Figure S1, in terms of surface concentrations (in atomic percent), survey spectra, and high resolution (HR) spectra of Sr, Ti, and O for STOn and STOm, respectively.
x FOR PEER REVIEW 4 of 17 Surface properties (i.e., chemical composition and state of the main species present over sample surfaces) were investigated by X-ray photoelectron spectroscopy (XPS).
The obtained results are summarized in Table S1 and Figures 3, 4 and S1, in terms of surface concentrations (in atomic percent), survey spectra, and high resolution (HR) spectra of Sr, Ti, and O for STOn and STOm, respectively.  In general, as expected, both survey spectra ( Figure S1) confirm the presence of Sr, Ti, O, and C at the photocatalyst surfaces. In any case, the presence of carbon besides the adventitious one could be ascribed to carbonate species on surfaces ( Figure S2), according to XRPD analyses ( Figure 1) [64].
Concerning the core energy levels (i.e., Sr 3d, Ti 2p, and O 1s, Figures 3 and 4 for STOn and STOm, respectively), both samples show very similar results. Sr 3d 5/2 core level centered at 133.1 eV in the STOn spectrum ( Figure 3a) corresponds to +2 valence of Sr [65,66].  In general, as expected, both survey spectra ( Figure S1) confirm the presence of Sr, Ti, O, and C at the photocatalyst surfaces. In any case, the presence of carbon besides the adventitious one could be ascribed to carbonate species on surfaces ( Figure S2), according to XRPD analyses ( Figure 1) [64].
Regarding the Ti 2p signals, the doublet at 458.7 and 464.5 eV can be assigned to Ti 2p 3/2 and ½ components of Ti 4+ . The second doublet at 456.7 and 462.3 eV is related to the presence of Ti 3+ species in the material (Figure 3b) [67].
Eventually, the O 1s spectrum (Figure 3c) shows a complex peak structure. The fitted two peaks located at binding energies of 527.8 and 528.9 eV are assigned to the lattice oxygen ion and oxygen vacancy, whereas the last two at 530.1 and 531.7 eV are related to the adsorption of OH species on the STO surface [68]. Similar results were obtained for STOn (Figure 4c).
Bandgap evaluation was carried out from UV-Vis diffuse reflectance spectra (UV-DRS) collected in the 200-800 nm interval ( Figure 5). Regarding the Ti 2p signals, the doublet at 458.7 and 464.5 eV can be assigned to Ti 2p 3/2 and 1 2 components of Ti 4+ . The second doublet at 456.7 and 462.3 eV is related to the presence of Ti 3+ species in the material (Figure 3b) [67].
Eventually, the O 1s spectrum ( Figure 3c) shows a complex peak structure. The fitted two peaks located at binding energies of 527.8 and 528.9 eV are assigned to the lattice oxygen ion and oxygen vacancy, whereas the last two at 530.1 and 531.7 eV are related to the adsorption of OH species on the STO surface [68]. Similar results were obtained for STOn ( Figure 4c).
Bandgap evaluation was carried out from UV-Vis diffuse reflectance spectra (UV-DRS) collected in the 200-800 nm interval ( Figure 5).
Both samples present higher absorbance at wavelengths shorter than 400 nm, resulting from electron transitions of valence band to conduction band, and the values of absorbance decrease as the photon energy is lower than the bandgap energy of each material.
As expected, the optical absorption of nano-sized oxide (STOn) is much stronger than that of the micrometric material (Figure 5a), in agreement with Liu and Reinosa et al. for TiO 2 -based materials [69,70]. In particular, this behavior has been detected in 230 nm < λ < 330 nm. The resulting values equal to about 3.23 and 3.24 eV for STOm and STOn, respectively, implying that the photocatalytic properties might exist under solar light irradiation, exploiting the UV component. Moreover, according to previous studies [69,70], the differences found between the two bandgap values could be ascribed to the particles' size, impacting on the crystalline structure and defects of the STO samples, also in agreement to XRPD results. Both samples present higher absorbance at wavelengths shorter than 400 nm, resulting from electron transitions of valence band to conduction band, and the values of absorbance decrease as the photon energy is lower than the bandgap energy of each material.
As expected, the optical absorption of nano-sized oxide (STOn) is much stronger than that of the micrometric material (Figure 5a), in agreement with Liu and Reinosa et al. for TiO2-based materials [69,70]. In particular, this behavior has been detected in 230 nm < λ < 330 nm. The resulting values equal to about 3.23 and 3.24 eV for STOm and STOn, respectively, implying that the photocatalytic properties might exist under solar light irradiation, exploiting the UV component. Moreover, according to previous studies [69,70], the differences found between the two bandgap values could be ascribed to the particles' size, impacting on the crystalline structure and defects of the STO samples, also in agreement to XRPD results.
To obtain some information about the functionalities eventually present at the surface of both STO materials, Fourier transform infrared (FT-IR) spectroscopy has been employed; results are depicted in Figure 6. To obtain some information about the functionalities eventually present at the surface of both STO materials, Fourier transform infrared (FT-IR) spectroscopy has been employed; results are depicted in Figure 6. Both samples present higher absorbance at wavelengths shorter than 400 nm, resulting from electron transitions of valence band to conduction band, and the values of absorbance decrease as the photon energy is lower than the bandgap energy of each material.
As expected, the optical absorption of nano-sized oxide (STOn) is much stronger than that of the micrometric material (Figure 5a), in agreement with Liu and Reinosa et al. for TiO2-based materials [69,70]. In particular, this behavior has been detected in 230 nm < λ < 330 nm. The resulting values equal to about 3.23 and 3.24 eV for STOm and STOn, respectively, implying that the photocatalytic properties might exist under solar light irradiation, exploiting the UV component. Moreover, according to previous studies [69,70], the differences found between the two bandgap values could be ascribed to the particles' size, impacting on the crystalline structure and defects of the STO samples, also in agreement to XRPD results.
To obtain some information about the functionalities eventually present at the surface of both STO materials, Fourier transform infrared (FT-IR) spectroscopy has been employed; results are depicted in Figure 6. The broad absorption located at ν > 3000 cm −1 is assigned to the stretching vibration of all the OH groups present on both STOs, including those deriving from undissociated water molecules present at the surface, whose spectroscopic counterpart (δHOH, bending mode) is included in the envelope right above 1600 cm −1 [65]. Eventually, the complex The broad absorption located at ν > 3000 cm −1 is assigned to the stretching vibration of all the OH groups present on both STOs, including those deriving from undissociated water molecules present at the surface, whose spectroscopic counterpart (δ HOH , bending mode) is included in the envelope right above 1600 cm −1 [65]. Eventually, the complex absorption bands (including the 1629 cm −1 component just discussed) located in the 1700-1300 cm −1 range can be related to some spectral modes due to traces of carbonate species still present as residues of the synthetic procedure [71]. Bands below 1000 cm −1 can be related to some SrTiO 3 crystal lattice vibrations [72]. 1300 cm −1 range can be related to some spectral modes due to traces of carbonate species still present as residues of the synthetic procedure [71]. Bands below 1000 cm −1 can be related to some SrTiO3 crystal lattice vibrations [72].
Sample morphology was investigated by conventional transmission and scanning electron microscopies (C-TEM and SEM). Micrographs of STOn (Figure 7a  In the Electronic Supplementary Material (ESI), the determination of isoelectric point (IEP) is provided (Paragraph S1 and Figure S3).

Diclofenac Abatement
Both samples were tested for DCF photodegradation under solar light irradiation. Before starting the photocatalytic investigation, a photolysis test was performed in ultrapure water (UW) and in the absence of any photocatalysts. Results indicate that photolysis has a negligible effect on the DCF degradation ( Figure S4).
The photocatalytic activity of samples was studied under different conditions to investigate the effect of some key parameters, such as photocatalyst dose and type of water matrix (ultrapure water, UW, and simulated drinking water, DW).
It is known that photocatalysis is a catalyst-mass-dependent reaction [73]. However, although large quantities of active species (hydroxyl radicals and/or electrons by absorbing more photons) correspond to large quantities of catalyst, a photocatalyst excess could increase the solution turbidity, reducing the energy absorption from light radiation, thereby decreasing the degradation efficiency. Moreover, proper doses of photocatalyst in the photodegradation system can reduce costs from an economic viewpoint. Therefore, at first, the effect of photocatalyst dose towards DCF photodegradation in ultrapure water In the Electronic Supplementary Material (ESI), the determination of isoelectric point (IEP) is provided (Paragraph S1 and Figure S3).

Diclofenac Abatement
Both samples were tested for DCF photodegradation under solar light irradiation. Before starting the photocatalytic investigation, a photolysis test was performed in ultrapure water (UW) and in the absence of any photocatalysts. Results indicate that photolysis has a negligible effect on the DCF degradation ( Figure S4).
The photocatalytic activity of samples was studied under different conditions to investigate the effect of some key parameters, such as photocatalyst dose and type of water matrix (ultrapure water, UW, and simulated drinking water, DW).
It is known that photocatalysis is a catalyst-mass-dependent reaction [73]. However, although large quantities of active species (hydroxyl radicals and/or electrons by absorbing more photons) correspond to large quantities of catalyst, a photocatalyst excess could increase the solution turbidity, reducing the energy absorption from light radiation, thereby decreasing the degradation efficiency. Moreover, proper doses of photocatalyst in the photodegradation system can reduce costs from an economic viewpoint. Therefore, at first, the effect of photocatalyst dose towards DCF photodegradation in ultrapure water (UW) under solar light irradiation was properly investigated. The main results are reported in Figure 8.  Dark adsorption tests led to a negligible DCF loss for both samples. In contrast, both oxides exhibit extraordinarily high activity. In general, the degradation efficiency increased as the dose decreased from 1.00 to 0.25 g·L −1 . In fact, as reported above, an excess of photocatalyst dose can increase the solution turbidity and simultaneously reduce the light penetration into the reaction system. Consequently, the lower light absorption of the photocatalyst decreases the degradation efficiency. This effect is much more pronounced for STOm than for the corresponding nanometric material. Nanoparticles' size is a crucial factor for their use in several fields. In particular, for photocatalytic applications, large particles or their aggregates can affect the optical properties of the solution by strong scattering and/or absorption of rays [74]. As shown in Figure 8, STOm is able to degrade 98% of the DCF within one hour when the catalyst concentration is 0.25 g·L −1 . This percentage decreases to 91% when the photocatalyst concentration is increased up to 0.50 g·L −1 and reaches 85% if the photocatalyst dose is doubled again. Conversely, the effect of photocatalyst dose is less marked for nanometric STO. In fact, both tests carried out using the concentration of 0.50 g·L −1 and 0.25 g·L −1 of STOn lead to 94% DCF degradation in one hour. However, even in this case, a further increase of photocatalyst amount causes a decrease in the efficiency of the catalytic system (up to 88%). Figure 9 suggests that DCF photodegradation carried out by both STOs follows second-order kinetics, according to the following equation: where qe and q are the amounts of photodegraded DCF by STO at the equilibrium and time t, respectively (mg·g −1 ), whereas k2 is the rate constant of pseudo second-order kinetics (min −1 ). Dark adsorption tests led to a negligible DCF loss for both samples. In contrast, both oxides exhibit extraordinarily high activity. In general, the degradation efficiency increased as the dose decreased from 1.00 to 0.25 g·L −1 . In fact, as reported above, an excess of photocatalyst dose can increase the solution turbidity and simultaneously reduce the light penetration into the reaction system. Consequently, the lower light absorption of the photocatalyst decreases the degradation efficiency. This effect is much more pronounced for STOm than for the corresponding nanometric material. Nanoparticles' size is a crucial factor for their use in several fields. In particular, for photocatalytic applications, large particles or their aggregates can affect the optical properties of the solution by strong scattering and/or absorption of rays [74]. As shown in Figure 8, STOm is able to degrade 98% of the DCF within one hour when the catalyst concentration is 0.25 g·L −1 . This percentage decreases to 91% when the photocatalyst concentration is increased up to 0.50 g·L −1 and reaches 85% if the photocatalyst dose is doubled again. Conversely, the effect of photocatalyst dose is less marked for nanometric STO. In fact, both tests carried out using the concentration of 0.50 g·L −1 and 0.25 g·L −1 of STOn lead to 94% DCF degradation in one hour. However, even in this case, a further increase of photocatalyst amount causes a decrease in the efficiency of the catalytic system (up to 88%). Figure 9 suggests that DCF photodegradation carried out by both STOs follows secondorder kinetics, according to the following equation: where q e and q are the amounts of photodegraded DCF by STO at the equilibrium and time t, respectively (mg·g −1 ), whereas k 2 is the rate constant of pseudo second-order kinetics (min −1 ). From the slope and the intercept of the graphics (t/qt vs. t, Figure 9) pseudo secondorder rate constants were calculated (k 2 ), and they the results were 0.016 and 0.006 min −1 for STOn and STOm, respectively.
This model predicts that intra-particle diffusion/transport process is the rate determining step that may involve valence forces through sharing or exchange of electrons between the pollutant and photoreactive species.
Total organic carbon (TOC) measurements showed that STO do not exhibit high mineralization ability ( Figure S5). From the slope and the intercept of the graphics (t/qt vs. t, Figure 9) pseudo secondorder rate constants were calculated (k2), and they the results were 0.016 and 0.006 min −1 for STOn and STOm, respectively.
This model predicts that intra-particle diffusion/transport process is the rate determining step that may involve valence forces through sharing or exchange of electrons between the pollutant and photoreactive species.
Total organic carbon (TOC) measurements showed that STO do not exhibit high mineralization ability ( Figure S5).
In more detail, only STOn used in large amount (1.00 g·L −1 ) permits one to reach 25% DCF mineralization, thanks to its nanometric dimension, which make the oxide particularly active. In all the other cases, the mineralization was poor. These findings are in line with those previously obtained by ZnO and TiO2 under UV light irradiation [12,18].
Based on all these results, the effect of water matrix was investigated, carrying out the DCF photodegradation by both STOm and STOn (0.25 g·L −1 ) in DW to evaluate the effect of the presence of inorganic salts on DCF photodegradation ( Figure 10).  In more detail, only STOn used in large amount (1.00 g·L −1 ) permits one to reach 25% DCF mineralization, thanks to its nanometric dimension, which make the oxide particularly active. In all the other cases, the mineralization was poor. These findings are in line with those previously obtained by ZnO and TiO 2 under UV light irradiation [12,18].
Based on all these results, the effect of water matrix was investigated, carrying out the DCF photodegradation by both STOm and STOn (0.25 g·L −1 ) in DW to evaluate the effect of the presence of inorganic salts on DCF photodegradation ( Figure 10). From the slope and the intercept of the graphics (t/qt vs. t, Figure 9) pseudo secondorder rate constants were calculated (k2), and they the results were 0.016 and 0.006 min −1 for STOn and STOm, respectively.
This model predicts that intra-particle diffusion/transport process is the rate determining step that may involve valence forces through sharing or exchange of electrons between the pollutant and photoreactive species.
Total organic carbon (TOC) measurements showed that STO do not exhibit high mineralization ability ( Figure S5).
In more detail, only STOn used in large amount (1.00 g·L −1 ) permits one to reach 25% DCF mineralization, thanks to its nanometric dimension, which make the oxide particularly active. In all the other cases, the mineralization was poor. These findings are in line with those previously obtained by ZnO and TiO2 under UV light irradiation [12,18].
Based on all these results, the effect of water matrix was investigated, carrying out the DCF photodegradation by both STOm and STOn (0.25 g·L −1 ) in DW to evaluate the effect of the presence of inorganic salts on DCF photodegradation ( Figure 10). All samples retained their degradation performance in both UW and DW, indicating that electrolytes do not have a detrimental effect on the DCF photodegradation. Previous studies, carried out with other oxides (ZnO and TiO2), reported different results, demonstrating that TiO2 worsens its photocatalytic performance in DW [18], whereas on the contrary, ZnO enhances its photocatalytic efficiency [12,18]. These findings are particularly important. In fact, a large number of investigations on DCF photodegradation confirmed the negative effects of electrolytes in different water matrices [75] due to the presence of inorganic species (chlorides, carbonates, bicarbonates, etc.) that can block active sites at the photocatalyst surface, as well as compete for free radicals. In addition, high values of All samples retained their degradation performance in both UW and DW, indicating that electrolytes do not have a detrimental effect on the DCF photodegradation. Previous studies, carried out with other oxides (ZnO and TiO 2 ), reported different results, demonstrating that TiO 2 worsens its photocatalytic performance in DW [18], whereas on the contrary, ZnO enhances its photocatalytic efficiency [12,18]. These findings are particularly important. In fact, a large number of investigations on DCF photodegradation confirmed the negative effects of electrolytes in different water matrices [75] due to the presence of inorganic species (chlorides, carbonates, bicarbonates, etc.) that can block active sites at the photocatalyst surface, as well as compete for free radicals. In addition, high values of ionic strength can also cause agglomeration phenomena with negative consequences on the available surface area of the photocatalyst. All these aspects do not seem to affect the photocatalytic properties of STO, which maintains the same efficiency regardless of size.
For practical applications, a useful photocatalyst requires stability to be reused many times without losing performance. In addition, it is important to remember that in the last years, great concerns have been risen regarding the effective toxicity of nanomaterials.
Particles with dimensions below 100 nm fall into the class of "ultrafine" materials [76], exhibiting several adverse health effects, such as lung tumors, inflammations, fibrosis, DNA damage, and so on [77,78], owing to their ability to deeply penetrate lungs and cell membranes [77].
Moreover, compared to nanometric materials, micro-sized particles are more easily recoverable by traditional filtration, greatly reducing process costs.
For all these reasons, reusability tests were carried out on STOm (1.00 g·L −1 ) in both UW and DW ( Figure 11).
ionic strength can also cause agglomeration phenomena with negative consequences on the available surface area of the photocatalyst. All these aspects do not seem to affect the photocatalytic properties of STO, which maintains the same efficiency regardless of size.
For practical applications, a useful photocatalyst requires stability to be reused many times without losing performance. In addition, it is important to remember that in the last years, great concerns have been risen regarding the effective toxicity of nanomaterials.
Particles with dimensions below 100 nm fall into the class of "ultrafine" materials [76], exhibiting several adverse health effects, such as lung tumors, inflammations, fibrosis, DNA damage, and so on [77,78], owing to their ability to deeply penetrate lungs and cell membranes [77].
Moreover, compared to nanometric materials, micro-sized particles are more easily recoverable by traditional filtration, greatly reducing process costs.
For all these reasons, reusability tests were carried out on STOm (1.00 g·L −1 ) in both UW and DW ( Figure 11). As can be observed, a good reproducibility of performances was observed after threerun recycling, guaranteeing a good stability of STOm under the used conditions, as confirmed by XRPD and FT-IR analyses carried out on the used sample ( Figure S6).
To understand the degradation pathway of various photocatalysts, active species trapping experiments are particularly useful. A number of photogenerated species, including h + , OH, and O2 − , play an important role in the photocatalytic degradation process. To trap h + , OH, and O2 − scavengers, such as disodium ethylenediaminetetraacetic acid (Na2-EDTA), p-benzoquinone (BQ), and isopropyl alcohol (IPA) can be used. The reactive species trapping experiments were carried out during DCF photodegradation in the presence of STOm (1.00 g·L −1 ) as photocatalyst in UW. Results are summarized in Figure 12. As can be observed, a good reproducibility of performances was observed after three-run recycling, guaranteeing a good stability of STOm under the used conditions, as confirmed by XRPD and FT-IR analyses carried out on the used sample ( Figure S6).
To understand the degradation pathway of various photocatalysts, active species trapping experiments are particularly useful. A number of photogenerated species, including h + , OH, and O 2 − , play an important role in the photocatalytic degradation process. To trap h + , OH, and O 2 − scavengers, such as disodium ethylenediaminetetraacetic acid (Na 2 -EDTA), p-benzoquinone (BQ), and isopropyl alcohol (IPA) can be used. The reactive species trapping experiments were carried out during DCF photodegradation in the presence of STOm (1.00 g·L −1 ) as photocatalyst in UW. Results are summarized in Figure 12.
It is possible to observe that the addition of BQ (an O 2 scavenger) greatly diminishes the percentage of DCF that degrades, suggesting that O 2 -photogenerated charge carriers are of much importance in this photocatalytic activity. In contrast, the addition of EDTA for trapping h + does slightly slow down photocatalytic DCF degradation, indicating that h + plays only a minor role in DCF degradation. Eventually, IPA (a OH scavenger) had no effect in photocatalytic DCF degradation. It is possible to observe that the addition of BQ (an O2 scavenger) greatly diminishes the percentage of DCF that degrades, suggesting that O2-photogenerated charge carriers are of much importance in this photocatalytic activity. In contrast, the addition of EDTA for trapping h + does slightly slow down photocatalytic DCF degradation, indicating that h + plays only a minor role in DCF degradation. Eventually, IPA (a OH scavenger) had no effect in photocatalytic DCF degradation.

Materials
All chemicals of analytical grade were purchased from Sigma-Aldrich products (Merck & Co, St. Louis, MO, USA). The two purchased catalysts were strontium titanate (99% purity) micro-and nano-powders; they were labelled as STOm and STOn, respectively. Pollutant mixtures were prepared in ultrapure water (UW) and in a simulated solution of drinking water (DW), whose composition is reported below. For HPLC/UV analyses, HPLC-grade acetonitrile and water were purchased from Carlo Erba reagents.

Characterization
Sample morphology was evaluated by scanning electron microscopy and conventional electron microscopy (SEM and C-TEM, respectively). Concerning the former, images were collected using a Zeiss LEO 1525 field emission microscope (Overcoached, Germany) without any sample pretreatment. Regarding the latter, images were collected by means of a Jeol JEM 3010-UHR instrument (LaB6 filament, acceleration voltage 300 kV, Milpitas, CA, USA), equipped with a Gatan Ultrascan 1000 camera. In this case, for sample preparation, powders were milled in an agate mortar and deposited on a Cu grid covered with a lacey carbon.
Crystal structure and phase composition of samples were determined by X-ray powder diffraction (XRPD) by means of a PANanalytical X'Pert PRO diffractometer (Cu Kα = 1.54060 Å and equipped with X-ray source operating at 40 kV × 40 mA). All the diffractograms were collected in 10-80° (2θ) range, with a step of 0.02° 2θ (or 0.05° 2θ), and the time for each step in the interval of 5-96 s. Before the analyses, powders were finely

Materials
All chemicals of analytical grade were purchased from Sigma-Aldrich products (Merck & Co, St. Louis, MO, USA). The two purchased catalysts were strontium titanate (99% purity) micro-and nano-powders; they were labelled as STOm and STOn, respectively. Pollutant mixtures were prepared in ultrapure water (UW) and in a simulated solution of drinking water (DW), whose composition is reported below. For HPLC/UV analyses, HPLC-grade acetonitrile and water were purchased from Carlo Erba reagents.

Characterization
Sample morphology was evaluated by scanning electron microscopy and conventional electron microscopy (SEM and C-TEM, respectively). Concerning the former, images were collected using a Zeiss LEO 1525 field emission microscope (Overcoached, Germany) without any sample pretreatment. Regarding the latter, images were collected by means of a Jeol JEM 3010-UHR instrument (LaB6 filament, acceleration voltage 300 kV, Milpitas, CA, USA), equipped with a Gatan Ultrascan 1000 camera. In this case, for sample preparation, powders were milled in an agate mortar and deposited on a Cu grid covered with a lacey carbon.
Crystal structure and phase composition of samples were determined by X-ray powder diffraction (XRPD) by means of a PANanalytical X'Pert PRO diffractometer (Cu Kα = 1.54060 Å and equipped with X-ray source operating at 40 kV × 40 mA). All the diffractograms were collected in 10-80 • (2θ) range, with a step of 0.02 • 2θ (or 0.05 • 2θ), and the time for each step in the interval of 5-96 s. Before the analyses, powders were finely ground and spread on an aluminum flat-plate horizontal sample holder. Patterns were identified by comparison with the JCPDS files from International Center for Diffraction Data Powder Diffraction. The average crystallite size (D) was evaluated by Scherrer equation (Equation (2)) using the peaks at 46.52 • (corresponding to 200 reflection), as follows: where K is the shape factor, λ is the wavelength of X-ray radiation, θ is the diffraction Bragg angle, and β is the corrected peak width at half the maximum. Specific surface area (SA) and porosity (pore volume, pore size, and pore size distribution) were determined by N 2 adsorption/desorption isotherms collected at −196 • C using an automatic analyzer of surface area (Coulter SA3100 instrument, Beckman Life Sciences, Los Angeles, CA, USA). Before each analysis, ca. 0.50 g dried sample was outgassed at 150 • C for 4 h under vacuum to remove water and other volatile organic compounds adsorbed on the surface. Specific surface area values were calculated by Brunauer-Emmet-Teller (BET) equation (two parameters, 0.05 < p/p 0 < 0.3), considering a cross-sectional area of ca. 16.2Ǻ LATIN CAPITAL/molecule N2 . Pore volume and pore size distribution (PSD) were determined by Barrett-Joyner-Halenda (BJH) model from the desorption branch of the collected isotherms (range: 0.3 < p/p 0 < 0.95).
Fourier transform infrared (FT-IR) spectra were collected in the 400−4000 cm −1 interval at room temperature through a Perkin-Elmer spectrometer (Perkin Elmer, Waltham, MA, USA). Prior to the analysis, samples were dried at 120 • C overnight, and then ca. 5 mg of each dried sample was padded in mixture with ca. 200 mg of dried KBr (sample to KBr weight ratio ca. 1:40).
UV-Vis diffuse reflectance (UV-DR) spectra were collected at room temperature in the 200−800 nm interval through a double-beam UV-Vis-NIR scanning spectrophotometer (Perkin Elmer Lambda 750s UV-Vis spectrophotometer, Perkin Elmer, Waltham, MA, USA) equipped with an integrating sphere assembly. Each sample in powder form was finely ground, uniformly pressed in a circular disk (E.D., ca. 4 cm), and included in the sampleholder. This latter was inserted in a special quartz cuvette and placed on the window of the integrating sphere for reflectance measurements. Spectra were measured using lab sphere reflectance standards consisting of BaSO 4 as the reference. The measured reflectance values (R%) were converted to absorbance (Abs, a.u.) by the following equation: The isoelectric point (IEP) of samples was determined according to the following procedure, previously reported in the literature [7]. In a typical experiment, ca. 50 mg of sample powder was weighed and introduced in NaNO 3 solutions (20 mL, 0.1 M) under stirring. Initial pH values (pH initial ) of NaNO 3 solutions were adjusted in 4.00−10.00 intervals, adding 0.1 M HNO 3 or NaOH. Suspensions were maintained under stirring (250 rpm) for 24 h, and successively, the final pH values (pH final ) were measured after suspension centrifuge (3000 rpm for 6 min). By plotting the difference between the pH final and pH initial (∆pH) along with the pH initial , pH pzc was determined as the intersection of the resulting line at which ∆pH = 0.
Surface composition of samples was studied by X-ray photoelectron spectroscopy (XPS) by means of an M-PROBE Surface Spectrometer with an Al (Kα) source and a spot size from 0.15 mm to 1 mm in diameter (USA). A 10 V applied voltage at a vacuum of 10 −7 -10 −8 Torr was used. Survey scans were investigated in 0−1100 eV binding energy range, using a spot size of 800 microns with 4 eV energy resolution (scan rate of 1 eV per step). ESCA Hawk Software was used for data curation. Charge neutralization was carried out by means of a low-energy electron beam. In any case, the resulting binding energy values (BE) were corrected by setting the C 1s peak (C-C) fixed at 284.6 eV as the reference. The C 1s photopeak comprised four components (C-C, C-O, and C=O bonds, as well as a specific feature corresponding to CO 3 2− ) with respective binding energies of 284.6, 285.9, 287.0, and 289.5 eV ( Figure S2). Normalized surface concentrations of all the species were computed by ruling out the contribution of adventitious carbon (C-C at 284.6 eV) [66,79].

Diclofenac Abatement
Photocatalytic activity was studied in the absorption and photodegradation of DCF in ultrapure and simulated drinking water (UW and DW, respectively). The latter was Experiments in UW and DW were performed by stirring 0.100, 0.050, and 0.025 g powder catalyst in a 250 mL batch glass reactor filled with 100 mL pollutant solution (10 ppm), maintained at spontaneous pH (about 5 for UW and 8 for DW, respectively).
The suspension was kept in the dark for 30 min, and it was successively irradiated for 180 min with a solar lamp (power density of irradiation of 35 W·m −2 , ULTRA VITALUX 300W-OSRAM, Osram, Munich, Germany) placed above the reactor. DCF abatement was monitored for 210 min by sampling aliquots every 15 min in the first half-hour and every 30 min in the subsequent 3 h. Suspensions were centrifuged for 10 min at 135 × 1000 rpm, and the liquid part was analyzed with high-performance liquid chromatography (HPLC). Chromatographic analysis was performed with an isocratic elution of a mobile phase, composed by water (50%), acetonitrile (50%), and formic acid (0.1%), at 1.00 mL·min −1 . The HPLC instrument (Agilent 1100 Series, Agilent, Santa Clara, CA, USA) was equipped with a C18 Supelco column (25 cm × 4 mm, 5 µm), a 20 µL autosampler, and a UV detector. DCF loss was monitored at 276 nm. In addition, the determination of total organic carbon (TOC) of the sampling aliquots at the beginning and at the end of experiments was carried out by means of a total organic carbon analyzer (TOC-L, Shimadzu, Kyoto, Japan).
After the first abatement test, the STO sample showing the best performances was recovered, washed with water, and then used for two further cycles.
Tests in the presence of disodium ethylenediaminetetraacetate (EDTA), benzoquinone (BQ), and isopropyl alcohol (IPA) as scavengers for holes (h + ), superoxide radical (O 2 − ), and hydroxyl radical (OH), respectively, were carried out. In these cases, the experimental procedures were conducted as previously reported for typical tests but with the addition of 3 mmol of each scavenger.

Conclusions
In this work, the potentialities of nano-and micro-sized strontium titanates in the photodegradation of diclofenac were comparatively investigated. Solar light as source, renewable in nature, was used for the photodegradation of diclofenac. Results indicate that both photocatalysts are promising for DCF abatement under solar light irradiation and maintain the same activity in a complex environment (simulated drinking water), indicating that the presence of electrolytes does not hinder SrTiO 3 performance. Moreover, it was demonstrated that the photocatalytic degradation of diclofenac is mostly due to O 2 − radical species, as indicated by radical trapping investigations. After recycling the catalyst three times, it maintains high photocatalytic performance. Based on these results, SrTiO 3 could be considered as a promising alternative to traditional materials to remove hazardous compounds from water.
Supplementary Materials: The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/catal12080804/s1, Table S1: Surface composition of the studied STO samples determined by XPS obtained by subtraction of adventitious carbon (at 284.6 eV); Figure S1: Survey spectra of STOm and STOn (left and right, respectively); Figure S2: HR C 1s spectrum of STOn; Figure S3: Isoelectric point (IEP) of STOm (red line) and STOn (blue line); Figure S4: Photolysis study for DCF decomposition without any catalyst; Figure S5: Percentage of DCF mineralization by STOm and STOn in UW; Figure S6: FT-IR spectra and XRPD pattern of used STOm. Funding: Velux Stiftung Foundation is gratefully acknowledged for their financial support through the project 1381 "SUNFLOAT-Water decontamination by sunlight-driven floating photocatalytic systems".

Data Availability Statement:
The data that support the plots within this paper are available from the corresponding author on reasonable request.

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