Photocatalytic Decomposition of Amoxicillin Trihydrate Antibiotic in Aqueous Solutions under UV Irradiation Using Sn / TiO 2 Nanoparticles

TiO2 and Sn/TiO2 nanoparticles were successfully synthesized by sol-gel method. The resulting nanoparticles were characterized by XRD, TEM, SEM, UV-Vis reflectance spectroscopy, and BET analysis methods. The effects of Sn-doping on the crystal structure, surface area, adsorption properties, pore size distribution, and optical absorption properties of the catalysts were investigated. The effect of different Sn content on the amount of hydroxyl radical was discussed by using salicylic acid as probe molecule. The photocatalytic activity of samples was tested by photocatalytic mineralization of amoxicillin trihydrate (AMOX) as a model pollutant. Sn/TiO2 nanoparticles exhibited high photocatalytic activity during the mineralization of AMOX under UV light due to increase in the generated hydroxyl radicals, band gap energy, specific surface area, and decrease in the crystallite size. The kinetic of the mineralization of AMOX can be explained in terms of the Langmuir-Hinshelwood model. The values of the adsorption equilibrium constant (KAMOX) and the kinetic rate constant of surface reaction (kc) were 0.56 (mg L−1)−1 and 1.86 mg L−1 min−1, respectively.


Introduction
Recently, there has been an increasing concern, particularly in highly developed countries, about penetration of pharmaceuticals into the environment and related risks [1].A large variety of pharmaceutical compounds have been frequently found in sewage treatment plant effluents and river streams at concentrations up to several μg L −1 [2].These compounds enter the aquatic environment after their ingestion and subsequent excretion either without modifications or in the form of nonmetabolized parent compounds [3].Among the various pharmaceutical compounds present in the environment, special emphasis has been given to antibiotics, which are the most often discussed pharmaceuticals because of their potential role in the development of antibioticresistant bacteria [4].The great concern is that waters contaminated with antibiotics can reach waters used for human consumption or municipal water treatment facilities [5].Due to their antibacterial activity, waters containing antibiotics commonly used in animal and human health (antibiotics included) are refractory to natural biological degradation or conventional wastewater treatments [6].Amoxicillin is a semisynthetic β-lactam antibiotic (7-[2amino-2-(4-hydroxyphenyl)-acetyl] amino-3,3-dimethyl-6oxo-2-thia-5-azabicyclo [3.2.0] heptane-4-carboxylic acid), used in humans and food-producing animals to treat several diseases.Some authors have found amoxicillin and cloxacillin in wastewater [7].
Several alternatives to destroy these kinds of compounds have been considered in recent studies in the literature.These include reverse osmosis, adsorption on activated carbons, or advanced oxidation technologies, such as Fenton reaction, ozonation, and peroxidation combined with UV light [8].Heterogeneous photocatalysis using a semiconductor is a new, effective, and rapid technique for the removal of pollutants from water [9].Among various semiconducting materials, much attention has been given to TiO 2 due to its outstanding stability, inexpensiveness, lack of toxicity, and strong photoactivity [10].Extensive researches have been focused on synthetic methods of semiconductor nanoparticles with better crystallinity and photocatalytic performance.Yu et al. [11] prepared CuO/Cu 2 O composite hollow microspheres with controlled diameter and composition and without the addition of templates and additives by hydrothermal synthesis using Cu(CH 3 COO) 2 •H 2 O as a precursor.However, there is still a problem that photocatalytic efficiency is not high because recombination of photogenerated electron-hole pairs influences low photo quantum efficiency [12].To enhance its photoactivity, TiO 2 is often doped with various metal ions and oxides such as Zn, Pt, Pd, Au, Ag, Cu, WO 3 , and V 2 O 5 [13].The dopants act as charge separators of photoinduced electron-hole pairs, and it is proposed that, after excitation, the electron migrates to the metal where it becomes trapped and electron-hole recombination is suppressed [14].Castro et al. [15] presented the performance of Ag-TiO 2 photocatalyst, obtained using the hydrothermal synthesis route, under different light irradiations setups such as solar-simulated and interior-lighting lamps with different irradiation powers, towards the photocatalytic inactivation of E. coli in water.A possible way to improve the photocatalytic performance of doped TiO 2 is to explore the cooperative effect by introducing more than one species of foreign elements to the host [16].Appropriate choice of the codoping pair is the key factor.Li et al. [17] investigated the structural and electronic properties of iron-fluorine (Fe-F) and iron-sulfur-(Fe-S-) codoped anatase TiO 2 .They reported that the photocatalytic performance under visible light of Fe-S-codoped TiO 2 is better than that of Fe-Fcodoped one, and Fe-S should be a better codoping pair.Photocatalytic oxidation of some antibiotics such as Lincomycin, tetracycline, oxolinic acid, and fluoroquinolone has been reported [2,[18][19][20].Xekoukoulotakis et al. [4] have reported that the use of UV/TiO 2 was able to achieve 90% TOC reduction after 90 min of reaction with 10 mg L −1 ERM and 250 mg L −1 TiO 2 .Abellán et al. [21] have reported 82% of sulfamethoxazole degradation and 23% TOC reduction by UV/TiO 2 in 6 hr.No study on photocatalytic oxidation of amoxicillin•3H 2 O, in aqueous solution using Sn/TiO 2 as a catalyst under UV-C irradiation has been reported.
In this work, preparation and characterization of pure and Sn/TiO 2 nanoparticles is reported.These nanosized catalysts were characterized by the techniques such as XRD, SEM, TEM, DRS, and BET analysis methods.The photocatalytic activity of the prepared nanoparticles has been studied on the mineralization of AMOX, as a prototype molecule.Chemical structure and characteristics of AMOX have been given in Table 1.We investigated in a systematic way the effect of various parameters such as doping content of Sn, the amount of photocatalyst, and AMOX initial concentration on mineralization in model aqueous solutions.A detailed kinetic description of the process was given based on well-known mechanistic/kinetic models, namely, the Langmuir-Hinshelwood (L-H) model whereby the organic regent was preadsorbed on photocatalyst surface prior to UV illumination.We intentionally decided to investigate sample mineralization rather than substrate decomposition since the rationale of micropollution abatement in aqueous matrices should involve its complete elimination rather than its transformation to other species.

Experimental
2.1.Materials.Titanium n-butoxide (TBOT, Ti(OC 4 H 9 ) 4 ), ethanol with absolute grade, and tin (IV) chloride were used without any further purification.All chemicals used in this study were analytical grade and purchased from Merck (Germany).

Preparation of
TiO 2 and Sn/TiO 2 Nanoparticles.According to [22,23], for the synthesis of TiO 2 nanoparticles, first titanium n-butoxide was slowly dissolved in ethanol.The prepared solution was sonicated in an ultrasonic bath.The hydrolysis process was then performed by adding of drop by drop H 2 O into a flask containing TBOT/EtOH mixture under reflux and magnetic stirring.The molar ratio of TBOT/EtOH/H 2 O was 1 : 1 : 65.The yellowish transparent sol was yielded after continuously stirring for 3 h.Finally, TiO 2 nanoparticles were gained by calcining the TiO 2 xerogel at 450 • C for 3 h and grinding.The preparation of Sn-doped TiO 2 nanoparticles was the same as that of TiO 2 , except that the water used for the synthesis contained the required amount of tin (IV) chloride.
2.3.Analytical Methods.Powder X-ray diffraction (XRD) was used for identification of crystalline phases and estimation of the crystallite size.The X-ray diffraction (XRD) patterns were recorded on a Siemens/D5000 X-ray diffractometer with Cu Kα radiation (λ = 0.15478 nm).The texture and morphology of the prepared samples were measured by scanning electron microscope (SEM) (Philips XL-30ESM).Transmission electron microscopy (TEM) observation was carried out on Zeiss EM 900-80 keV electron microscopy instrument.The sample for TEM was prepared by dispersing the final powder in ethanol, and the dispersion was then dropped on copper/carbon grid.The nitrogen adsorption and desorption isotherms at 77 K were measured using Belsorp mini II.The specific surface area and the pore size distribution (average pore diameter and mean pore volume) were measured from the adsorption isotherm using the Brunauer-Emmett-Teller (BET) method and from the desorption isotherm using the Barret-Joyner-Halender (BJH) method, respectively.Ultraviolet/visible diffuse reflectance spectra (DRS) were taken on Avaspec-2048 TEC spectrometer.Total organic carbon (TOC) was measured with Shimadzu 5000A TOC analyzer equipped with an autosampler (ASI-5000) and platinum-based catalyst.The carrier gas was synthetic air at the rate of 150 mL/min.The samples were acidified to pH < 4 before being sent to TOC analyzer to ensure that inorganic carbon would be released from the solution as CO 2 .

Photocatalytic Activity. Mineralization of AMOX under
UV light was used as a model reaction to evaluate the photocatalytic activity of prepared samples.Photocatalytic activity measurements were carried out at atmospheric pressure in a batch quartz reactor [24].Artificial irradiation was provided by 15 W (UV-C) mercury lamp (Philips, Holland) emitted around 254 nm positioned parallel to the reactor [25].In each run, desired concentration of AMOX (20 mg L −1 ) and TiO 2 was fed into the quartz tube reactor and it was allowed to equilibrate for 30 min in the darkness.The zero time reading was obtained from blank solution kept in the dark.Aliquots of the mixture were taken at periodic intervals during the irradiation, centrifuged for 10 min at 1000 rpm (Hettich EBA) to remove catalyst particles, and then analyzed for their total organic carbon (TOC) content.
All the concentration profiles can be correlated to irradiation time by the following exponential function with good agreement: Therefore, the photocatalytic mineralization of AMOX is pseudo-first-order reaction, and its kinetics may also be expressed as (2) In this equation, TOC 0 and TOC are the antibiotic concentrations (mg L −1 ) at times 0 and t, respectively, and k ap is the pseudo-first-order rate constant (min −1 ).

X-Ray Diffraction.
Based on the XRD spectra, the crystalline phases could be categorized into two primary components, an anatase (A) and a rutile (R) phase, and represented the intensity of the strongest anatase reflection of (101) plane at 2θ = 25.3 • ± 0.1 • , anatase reflection of (200) plane at 2θ = 48.0• ± 0.1 • , and the intensity of the strongest rutile reflection of (110) plane at 2θ = 27.4 • ± 0.1 • .Additionally, the anatase phase of TiO 2 crystal is a tetragonal system in lattice geometry [26].The average crystallite size of the particles was calculated according to following Scherrer's equation [27]; In this equation, λ the X-ray wavelength equal to 0.154 nm, k is a constant taken as 0.89, β is the line width at half maximum height, and θ is the diffracting angle.The phase content of a sample can be calculated by following equation [28]: where I A and I R are integrated intensities of the anatase and rutile peaks, respectively.Figure 1 shows the X-ray diffraction patterns of TiO 2 and 1.5 mol% Sn/TiO 2 samples.Intensity of the anatase main peak located in 2θ = 25.5 • indicates a high degree of crystallinity for these nanoparticles.Crystallite   size and amount of each phase nanoparticles have been reported in Table 2.The crystallite size of titania decreases with increasing doping concentration.This reduction in crystallite size is proposed due to segregation of the dopant cations at the grain boundary which inhibits the grain growth by restricting direct contact of grains.The phase composition varies with the increasing content of Sn.
According to the XRD patterns, the pure TiO 2 sample constituted pure anatase phase.On the other hand, the sample containing Sn shows a multiphase composition as the Sn content increases.The 3 mol% Sn sample is composed of crystallites of both anatase and rutile phase.The formation of crystallites of rutile phase at a low calcination temperature is induced by the presence of Sn, as previously reported [29].
Metal may have been finely dispersed, incorporated in the TiO 2 crystal structure, or crystals are too small for detection due to a relatively low metal doping [30].

SEM Analysis of Sn/TiO 2
Nanoparticles.The SEM image of 1.5 mol% Sn/TiO 2 is shown in Figure 2. Since less particle agglomeration occurred, the large surface area conveys high adsorption abilities of the catalysts.This image shows relatively uniform particles which are coherent together.
3.1.3.TEM Analysis of Sn/TiO 2 Nanoparticles. Figure 3 shows TEM images of TiO 2 and 1.5 mol% Sn/TiO 2 nanoparticles.It can be seen that the particles exhibit a relatively uniform particle size distribution.The average size of the primary particles estimated from the TEM image is about 7-10 nm, which is in good agreement with that calculated from the XRD pattern using Scherrer equation.3.1.4.BET Analysis.Figure 4 shows nitrogen adsorptiondesorption isotherms of TiO 2 and 1.5 mol% Sn/TiO 2 samples.Both samples show the isotherm of type IV.At high relative pressure range from 0.4 to 1, the isotherm exhibits a hysteresis loop of type H2 associated with the ink bottle pores, indicating that the powders contain mesopores due to the aggregation of crystallites [31].The hysteresis loop in the lower relative pressure range (0.4 < P/P0) was related to finer intra-aggregated pores formed between intraagglomerated primary particles, and that in the higher relative pressure range (0.8 < P/P0 < 1) was associated with larger interaggregated pores produced by interaggregated secondary particles [32].The porous structure is believed to facilitate the transporting of reactant molecules and products through the interior space due to the interconnected porous networks and favor the harvesting of exciting light due to enlarged surface area and multiple scattering within the porous framework [33].
Figure 5 shows the pore size distribution of 1.5 mol% Sn/TiO 2 as estimated according to the BJH method from the adsorption branch.It can be seen that the diameter range of pores located from 0.97 to 10 nm and the mean diameter of pores is 8.9 nm.With doping of Sn, the average pore size decreases slightly from 9.7 to 8.9 nm (Table 1).There are two possible factors causing the decrease in average pore size.One is that the aggregation of smaller crystallites forms smaller pores.The other is that some ions of doping probably insert into the pore of pure TiO 2 , which also causes pore size to become smaller [34].The specific surface area determined with the BET method is 80.03 m 2 g −1 which is higher than 47.03 m 2 g −1 of TiO 2 sample.The enhanced surface area of 1.5 mol% Sn/TiO 2 compared to nanosized TiO 2 is due to its increase in the mesopore size and mesopore volume [35].Loganathan et al. [36] have reported that metal doping at a specific concentration can have a favorable effect on the surface area.However, at an excess doping concentration, it could result in the damage of the porous framework, thus negatively affecting the surface area.
The specific surface area, pore volume, and pore size of TiO 2 and 1.5 mol% Sn/TiO 2 nanoparticles are presented in Table 3.The increase of surface area can be useful in the efficiency of the photocatalytic activity as it implies larger contact surfaces exposed to the reagents [37].It is accepted in heterogeneous photocatalysis process that higher surface area and pore volume can be useful in the formation of photogenerated electron and hole pairs.Hence, heterogeneous photocatalysis is influenced greatly by the surface area and pore structure [38].
International Journal of Photoenergy

DRS Analysis.
To investigate the optical absorption properties of synthesized samples, diffuse reflectance spectra (DRS) were analyzed.Sn-doping obviously affects light absorption characteristics of TiO 2 as shown in Figure 6.
The reflectance spectrum of TiO 2 nanoparticles show an absorption threshold at 400 nm, while for Sn/TiO 2 samples the absorption threshold is between 390 and 381 nm.
The absorbance of Sn/TiO 2 nanoparticles is shifted toward shorter wavelength than TiO 2 nanoparticles.In general, blue shift of the absorption onset of Sn-doped nanoparticles is associated with the increase of the carrier concentration blocking the lowest states in the conduction band, well known as the Burstein-Moss effect [39,40].The theory for band-gap widening for polar semiconductors, proposed by Sernelius et al. [41], in order to describe the experimentally found blue shift in Sn/TiO 2 nanoparticles, considers displacement polarization effects and structural disorder as additional factors that affect the shape of the band tails.The direct band-gap energy can be estimated from a plot of (αhν) 2 versus photon energy (hν).Sn/TiO 2 nanoparticles show higher band-gap energy than TiO 2 nanoparticles.The band gap energy estimated from DRS analysis for Sn/TiO 2 nanoparticles is between 3.17 and 3.25 eV, which is higher than 3.1 eV for TiO 2 .The band gap plays a critical role in deciding the photocatalytic activity of photocatalysts for the reason that it participates in determining the e − /h + recombination rate.According to Moss-Burstein theory, in heavily doped TiO 2 nanoparticles, the donor electrons occupy states at the bottom of the conduction band.Since the Pauli principal prevents states from being doubly occupied and optical transition is vertical, the valence electrons require extra energy of doped TiO 2 to be boarder than that of pure TiO 2 nanoparticles [42].

Measurement of Hydroxyl Radical.
It is well understood that hydroxyl radical is generated upon proper photon illumination to photocatalyst.The hydroxyl radical is a powerful oxidizing species, having potential oxidation of approximately 2.8 volt (versus NHE), which may lead to complete mineralization of pollutants.Generally, the greater the formation rate of * OH radicals is, the higher separation efficiency of electron-hole pairs is achieved.So, the photocatalytic activity is a positive correlation to the formation rate of * OH radicals, namely, a faster formation rate of * OH radicals leads to a higher photo-catalytic activity [43].The effect of different Sn content on the amount of hydroxyl radical was detected using salicylic acid as a probe molecule [44].The salicylic acid reacted with the hydroxyl radicals to form 2,3-dihydroxybenzoic acid.After reacting 20 min, the amounts of hydroxyl radicals absorbed by the solution at the wavelength of 510 nm were measured.A blank experiment was run before adding the photocatalyst.No change is observed in the absorbance, which indicated that there was no photolysis reaction in salicylic acid under UV irradiation.Figure 7 shows the hydroxyl radical with different Sn content.The amount of hydroxyl radicals increases as the Sn content increases.The hydroxyl radical of 1.5 mol% Sn/TiO 2 is very high, indicating that the sample has a high separation rate of photoinduced carriers.However, the amount of hydroxyl radicals can decrease if the Sn content is too high.The reason is that the amount of photoinduced electrons accumulating instantly at the SnO 2 conduction band is too much due to excess SnO 2 so that some photoelectrons can indirectly recombine with holes [45].Therefore, it can be concluded that the separation rate of photoinduced charge carriers can be improved by doping an appropriate amount of Sn.

Photocatalytic Mineralization of AMOX Using Sn/TiO 2
Nanoparticles.The results of mineralization of AMOX using Sn/TiO 2 nanoparticles are illustrated in Figure 8.As the irradiation time increased, AMOX degraded into small fragments and subsequently mineralized completely.It could be seen that the photocatalytic activity of Sn/TiO 2 nanoparticles was higher than that of the pure TiO 2 .Further observation showed that photocatalytic activity gradually increased with increasing content of Sn.The slight enhancement of photocatalytic activity of 0.5 mol% Sn/TiO 2 nanoparticles could be assigned to the fact that the excited electrons from the valence band to the conduction band could migrate to Sn nanoparticles and then migrated to O 2 molecules adsorbed on the surface of the Sn.Sn produced an Schottky barrier, which facilitated the electron capture [46].Especially, when the content of Sn increased to 1.5 mol%, the photocatalytic activity of Sn/TiO 2 nanoparticles significantly increased.The high photo-catalytic activity of 1.5 mol% Sn/TiO 2 sample is due to the following several factors.(i) TiO 2 is a photoactive semiconductor that when illuminated with photon energy equal or greater than its band gap energy, the following reaction took place on the surface of the photo-catalyst: SnO 2 and TiO 2 are both wide band gap semiconductors.Although the band gap of SnO 2 (3.8 eV) is wider than that of TiO 2 (3.1 V), the Fermi lever of SnO 2 is lower than that of TiO 2 .It means that the photo-generated electrons may easily transfer from TiO 2 to SnO 2 , but not to recombine with the photo-generated holes on the surface of TiO 2 immediately.Consequently, more and more holes are present on the surface and take part in the reactions of oxidizing OH − and H 2 O into hydroxyl radicals.Hydroxyl radicals would be finally responsible for the degradation of pollutants into H 2 O and CO 2 [47,48].Furthermore, photo-induced electrons easily transfer from TiO 2 to SnO 2 due to their potential difference of conduction band.According to [49], the conduction band (CB) edges of TiO 2 and SnO 2 are situated at −0.34 and +0.07 V versus normal hydrogen electrode (NHE) at pH 7. Thus, electron can easily flow into SnO 2 , which has a more positive conduction band.This results into the decrease in the chance of recombination of photoinduced carriers so that the separation rate of photoinduced electron-hole pairs can improve, that is, the amount of hydroxyl radicals can increase.The amounts of hydroxyl radicals and the efficient charge separation directly influenced the photocatalytic activity [45].As can be seen in Figure 8, the photocatalytic activity of 1.5 mol% Sn/TiO 2 is very high, indicating that the sample has a high separation rate of photoinduced carriers.
(ii) A small amount of metal ions can act as a photogenerated hole and a photo-generated electron trap and inhibit the hole-electron recombination: The trapped electron may thus be readily transferred to oxygen molecule to form a superoxide radical anion (O 2 •− ).In addition, the trapped hole can be easily transferred to hydroxyl anion adsorbed on the surface forming hydroxyl radical (OH • ), or it can also be transferred to adsorbed dye molecule to form a dye radical [50].
Sn/TiO 2 nanoparticles show higher band-gap energy than TiO 2 nanoparticles, which not only suppressed the electron-hole recombination but also generated more * OH radicals [35].
(iii) The superior degradation efficiency of Sn/TiO 2 nanoparticles could be attributed to a larger surface area compared to another photocatalyst.A large surface area may be an important factor in certain photocatalytic mineralization reactions, as a large amount of adsorbed organic molecules promotes the reaction rate [51].
(iv) Particle size is another important parameter influencing photocatalytic efficiency, since the electron-hole recombination rate may depend on the particle size.It is well known that in the nanometer-size range, physical and chemical properties of semiconductors are modified (compared with bulk).Small variations in particle diameters lead to great modifications in the surface/bulk ratio, thus influencing the recombination rates of volume and surface electrons and holes [52].
Moreover, a decrease in the activity is expected when the content of Sn becomes too large.The detrimental effect of tin on TiO 2 photoactivity has several reasons.
(v) The amount of photoinduced electrons accumulating instantly at the SnO 2 conduction band is too much due to excess SnO 2 so that some photoelectrons can indirectly recombine with holes [45].
(vi) An excess amount of Sn dopant can produce the recombination center of photoinduced electron and hole pairs.Recombination of e − -h + pairs reduces the rate of photocatalytic mineralization [53].
(vii) Excessive coverage of TiO 2 catalyst limits the amount of light reaching to the TiO 2 surface, reducing the number of photogenerated e − -h + pairs and lowering consequently the TiO 2 photoactivity [54].
(viii) Doped metal may occupy the active sites on the TiO 2 surface for the desired photocatalytic reactions causing the TiO 2 to lose its activity [55].
(ix) The probability of the hole capture is increased by the large number of tin particles at high tin dopings, which decrease the probability of holes reacting with adsorbed species at the TiO 2 surface [56].
Based on the results, the optimum content of Sndoped TiO 2 for treatment of AMOX in aqueous solution is 1.5 mol%.

Effect of the
Amount of Sn/TiO 2 Nanoparticles.Photocatalytic mineralization of AMOX may depend on the catalyst amount.In order to investigate the effect of catalyst amount, experiments were carried out by varying the amount of photocatalyst from 100 to 600 mg L −1 , and the mineralization profile is shown in Figure 9.The rate of reaction increases with the increase of photocatalyst up to 400 mg L −1 and a further increase leads to a decrease in the antibiotic mineralization.This observation can be explained in terms of availability of active sites on the catalyst surface and the penetration of UV light into the suspension.The total active surface area increases with increasing catalyst dosage [57].The reasons for decrease in mineralization rate were aggregation of Sn/TiO 2 nanoparticles at high concentration causing a decrease in the number of surface active sites and an increase in the opacity and light scattering of Sn/TiO 2 nanoparticles at high concentration.This tends to decrease the passage of irradiation through the sample [58].The catalyst concentration above which conversion levels off depends on several factors (e.g., reactor geometry, operating conditions, wavelength, and intensity of light source) and corresponds to the point where all catalyst particles, that is, all the surface exposed, are fully illuminated [4].In this study, the optimum concentration at which all subsequent experiments were conducted was about 400 mg L −1 .

Treatment of Real Water Containing AMX.
In order to investigate the efficiency of Sn/TiO 2 nanoparticles in mineralization of AMOX in the real water, 20 mg L −1 of antibiotic was added into a real water sample (carbonate hardness: 94 mg L −1 CaCO 3 , sulphate concentration: 175.1 mg L −1 SO 4 2− ) that was obtained from an irrigation well in Tabriz, Iran.Results in Figure 10 show the effect of presence of common anions such as sulphate, carbonate, and bicarbonate on the mineralization of AMX.It is clear from this figure that, in presence of SO 4 2− , HCO 3 − , and CO 3 2− , the percent of mineralization decreases.This inhibition is   undoubtedly due to their ability to act as hydroxyl radical's scavengers by the following reactions [59]: These ions may also block the active sites on the Sn/TiO 2 surface thus deactivating the catalysts towards AMOX and intermediate molecules.Although the generated radical anions have been shown to be an oxidant itself, but its oxidation potential is less than that of the hydroxyl radicals.

Effect of the Initial AMOX Concentration in the
Photocatalytic Reaction.The effect of varying AMOX initial concentration was studied in the range of 10-40 mg L −1 .Figure 11 shows a plot of ln (TOC 0 /TOC) versus time for all the experiments with different initial concentration of AMOX.
By applying a least square regression analysis, the values of k ap have been obtained.The results are shown in Figure 12.
Table 4 reports the values of k ap resulting from plot of ln ([TOC] 0 /[TOC]) versus "Time," which decreases as the initial reactant concentration increases.The presumed reason is that when the initial concentration of antibiotic is increased, more and more antibiotic molecules are adsorbed on the surface of Sn/TiO 2 nanoparticles.The large amount of adsorbed antibiotic is thought to have an inhibitive effect on the reaction of antibiotic molecules with photogenerated holes or hydroxyl radicals, because of the lack of any direct contact between them.This was attributed to the rise of internal optical density, which caused the solution to became impermeable to UV light [60].Once the concentration of antibiotic is increased, it also causes the antibiotic molecules to absorb light and the photons never reach the photocatalyst surface, and thus the photocatalytic mineralization efficiency decreases [61,62].
Several reports have established that the heterogeneous photo-oxidation rate fits well to the classic Langmuir-Hinshelwood (L-H) mechanism [63] which in terms of mineralization kinetics can be described as follows: where k AMOX and k c are the Langmuir-Hinshelwood adsorption equilibrium constant and rate constant of surface reaction, respectively.Using the data from photocatalytic experiments with different initial AMOX concentrations, the values of k AMOX and k c can be calculated using the linearized equation by plotting 1/k obs versus [TOC] 0 .As can be seen in Figure 13, a straight line fitted the experimental data reasonably well (the coefficient of linear regression, r, was 0.994), thus indicating that photocatalytic mineralization of ERM most probably follows Langmuir-Hinshelwood kinetics.From the slope of the straight line, k c was computed equal to 1.86 mg L −1 min −1 , while from the intercept, k AMOX was 0.56 (mg L −1 ) −1 .Yurdakal and his coworkers reported Langmuir-Hinshelwood equation constants for mineralization of Gemfibrozil (GEM) drug in the presence of two

Conclusions
TiO 2 and Sn/TiO 2 nanoparticles could be prepared by sol-gel method using titanium n-butoxide and tin (IV) chloride as precursors.The XRD results showed that the crystallite size greatly decreased due to Sn-doping but an increase in surface area, pore volume, and band gap energy was observed.Nitrogen adsorption-desorption isotherms showed that the adsorption ability was enhanced owing to Sn-doping.The absorbance of Sn/TiO 2 nanoparticles was shifted toward shorter wavelength than TiO 2 nanoparticles.The effect of Sn dopant on the photoinduced charge property was estimated by measuring hydroxyl radicals using salicylic acid as probe molecule.The photocatalytic efficiency for AMOX decomposition was remarkably enhanced owing to Sn-doping, and 1.5 mol% Sn/TiO 2 sample had the highest photocatalytic activity due to increase in the generated hydroxyl radicals, band gap energy, specific surface area, and decrease in the crystal size.Langmuir-Hinshelwood kinetic model provided a good fit to the photocatalytic mineralization of AMOX, used in this study.This study confirms the potentialities of heterogeneous photocatalysis to decontaminate wastewaters containing organic pollutants.

Figure 11 :
Figure 11: Determination of the pseudo-first-order kinetic rate constants, k ap .

Figure 12 :
Figure 12: Effect of initial AMOX concentration in the photocatalytic reaction.

Figure 13 :
Figure 13: Determination of the adsorption equilibrium constant, k AMOX , and the second order rate constant, k c , for the Langmuir-Hinshelwood kinetic model.

Table 1 :
Chemical structure and characteristics of AMOX.

Table 2 :
Phase content and crystallite size of prepared nanoparticles.