Synthesis of C-N-S-Tridoped TiO2 from Vietnam Ilmenite Ore and Its Visible Light-Driven-Photocatalytic Activity for Tetracyline Degradation

In this study, C-N-S-tridoped TiO2 composite was fabricated from TiO2 prepared from ilmenite ore and thiourea by means of hydrothermal method. The obtained material was characterized by X-ray diffraction, Raman scattering spectroscopy, UV-Vis diffuse reflectance spectroscopy, nitrogen adsorption-desorption isotherms, scanning electron microscopy (SEM), transmission electron microscopy (TEM), and X-ray photoelectron spectroscopy (XPS). It was found that C-N-S-tridoped TiO2 material has a large specific surface area, showing good photocatalytic activity on the degradation of antibiotic tetracycline in visible light region. The study on the mechanism of tetracycline photodegradation using the liquid chromatography with mass spectrometry was performed. It was found that tetracycline has been degraded over C-N-S-tridoped TiO2 catalyst into many different intermediates which can eventually be converted into CO2 and H2O. The kinetics of photocatalytic decomposition of tetracycline were investigated. In addition, the obtained material could catalyze well the degradation of other antibiotics (ciprofloxacin and chloramphenicol) and dyes (rhodamine-B, methylene blue, and organe red). The catalyst was stable after five recycles with slight loss of catalytic activity, which indicates great potential for practical application of C-N-S-tridoped TiO2 catalyst in treatment of wastewater containing tetracycline in particular or antibiotics in general.


Introduction
Tetracycline (denoted as TC) is one of the widely used antibiotics in veterinary medicine and aquaculture [1]. Due to poor absorption, most of them are not metabolized and excreted in the faces and urine as the original compounds. Tetracycline is usually detected at a concentration of 0.07-1.34 μg/L in surface water samples. The most dangerous effect of antibiotics in the environment is to create favorable conditions for the development of resistant strains of bacteria and cause the current phenomenon of antibiotic resistance [2]. Therefore, the elimination of tetracycline should be of considerable concern. Most of the amount of tetracycline detected in surface water, drinking water, and mud is due to the ineffective elimination of traditional and biological methods [1,3]. In order to reduce the amount of tetracycline released into the environment, new treatment methods need to be developed. Advanced oxidation processes (AOPs) such as photochemical processes [1,3], ozonation [4], ultrasonic waves [5], and Fenton processes [6] are considered as effective technologies to decompose antibiotic pollutants in aqueous solution. Among the AOPs, photocatalytic process based on titanium dioxide is one of the most promising technologies. TiO 2 , with the preeminent properties, such as high photocatalytic activity and super wetting, is very durable and nontoxic and has high reserves; it has been studied and applied widely [7]. However, with a band gap of about 3.2-3.5 eV, TiO 2 material can only give a catalytic effect in the ultraviolet (UV) light region. The portion of UV radiation in the solar spectrum to the earth's surface is only about 5%, so the use of this source of radiation for TiO 2 photocatalytic treatment is limited. In order to expand the use of solar radiation energy in the visible light area into the photocatalytic reaction, it is necessary to reduce the band gap energy of TiO 2 or shift the light absorption of TiO 2 from the UV to the visible region. Many studies have suggested that TiO 2 doped with nonmetal elements such as C, F, N, and S can extend the absorption spectrum to the visible light region [8][9][10][11].
Recently, the simultaneous doping of two or three nonmetals as a promising strategy to increase the photocatalytic activity of TiO 2 in the visible region exhibits higher photocatalytic activity than that of TiO 2 doping an element [12]. Zhou and Yu [13] reported that TiO 2 crystals simultaneously modified C, N, and S by solid phase reaction exhibited high photocatalytic activity for formaldehyde decomposition in visible light region. Ao et al. [14] prepared the C-N-S-tridoped TiO 2 through sol-gel method combined with surfactant cetyltrimethylammonium bromide (CTAB) as a soft template. Photocatalytic activity is assessed by the decomposition of Reactive Dyes Brilliant Red X-3B in aqueous solution. Wang et al. [15] prepared C-N-S-doped TiO 2 nanocrystals by hydrothermal method in the presence of TiO 2 and L-cysteine biological molecule. L-cysteine not only provided a source of carbon, sulfur, and nitrogen, but also controlled the final product of crystalline and morphological phases. The high activity of this catalyst in visible region is due to the synergic effect of the large surface area, the red shift at the absorption edge, the strong absorption in the visible light zone, and the phase structure mixture of the material.
In the present study, TiO 2 was prepared from ilmenite ore by using sulfuric acid method, then doping of C, N, and S elements into TiO 2 was performed through the hydrothermal process of TiO 2 and thiourea. Photocatalytic decomposition reaction in the visible region of tetracycline on C-N-Stridoped TiO 2 composite was investigated.  [16,17]. This solution was concentrated until scum formation. Then cool the solution at a temperature of about -2°C to -5°C for 8 hours. The precipitate was separated, and TiOSO 4 solution was obtained.

Experimental
2.1.2. Synthesis of C-N-S-Tridoped TiO 2 Materials. Take 2.27 grams of TiOSO 4 and a defined amount of thiourea (with molar ratio of thiourea: TiO 2 = 2 : 1) into a teflon flask, then dissolve with distilled water. Put the teflon bottle in the autoclave at 180°C for 12 hours. After the hydrothermal process, the autoclave was naturally cooled to room temperature. Filter the white precipitate and wash several times with distilled water until the filtrate has a constant pH. The product is dried and calcined at 400-700°C for 1 hour to obtain the C-N-S-tridoped TiO 2 (denoted as TH-TiO 2 -a (a = 400, 500, 600, and 700°C). TiO 2 samples were also prepared as the abovementioned procedure but without adding thiourea.

Catalytic
Activity. The catalytic activity of TiO 2 or TH-TiO 2 was estimated through the decomposition of tetracycline (30 mg/L) and the catalytic mass (0.6 g/L). The mixture was stirred in the dark for 30 minutes to reach the absorption/desorption equilibrium, then illuminated with a 60 W filament lamp (filter cutoff λ < 420 nm). The remaining tetracycline concentration was determined by HPLC-UV method on Thermo Scientific series 3300 HPLC (Thermo Scientific Technologies, CA, USA) (λ max = 355 nm).

Methods of Analysis.
The X-ray diffraction (XRD) was conducted in D8-Advance 5005 with Cu Kα λ = 0:154 nm. The element composition and oxidation state on the catalyst surface in the sample were determined by X-ray electron spectroscopy (XPS) (ESCALAB 250-Thermo VG, UK). The morphology and the element composition were measured by using scanning electron microscope (SEM) with energy dispersive X-ray spectroscopy (EDX) (Nova Nano SEM 450) and transmission electron microscopy (TEM) (JEOL JEM 1010). The photoluminescence (PL) was measured using a Hitachi F-7000 spectrophotometer over wavelength range of 375-600 nm with a 150 W xenon lamp. Raman spectrum and UV-Vis-DR were conducted by using T64000 Raman with 633 nm and GBC Instrument-2885, respectively. Textural properties of material samples were characterized by nitrogen adsorption/desorption isotherms method using Micromeritics ASAP 2000. The intermediates that appear

Results and Discussion
3.1. Characterization of Materials. The physical chemistry characterizations of TH-TiO 2 materials calcined at different temperature are presented in Figure S1-4 and Table S1.
The primary experiments show that the TH-TiO 2 calcined at 500°C provided the highest photocatalytic activity ( Figure S5). Therefore, the only discussion of TH-TiO 2 calcined at 500°C was focused in this text. The obtained materials were studied by XRD measurements (Figure 1)  [12,18], indicating that the titanium oxide in anatase form is the main phase in the obtained samples. However, TH-TiO 2 prepared from hydrothermal process of TiO (SO 4 ) and thiourea possesses the lower intensities of anatase peaks compared to TiO 2 from hydrothermal process of only TiO (SO 4 ). The crystallite sizes of the samples could be estimated from the broadening of the X-ray diffraction peak according to Scherrer formula [19]. It was calculated that TiO 2 has an average crystallite size of 14.39 nm, and it decreases to 9.54 nm for TH-TiO 2 indicated that the tridoping of C, N, and S suppressed grain growth. Raman spectrum ( Figure 2) shows only the existence of 4 spectral bands at 144 cm -1 , 399 cm -1 , 399 cm -1 , 519 cm -1 , and 639 cm -1 , indicating that both TH-TiO 2 and TiO 2 have only anatase phase [20]. The peaks at 144 cm -1 (E g ) and 639 cm -1 (E g ) are related to the symmetrical stretching of the O-Ti-O bond in TiO 2 while the peak of 399 cm -1 (B1g) corresponds to the bending oscillation of the O-Ti-O bond, and 513 cm -1 (A1g) could be attributed to the asymmetric bending vibration of the O-Ti-O bond [21]. This result is consistent with XRD analysis.
The products prepared from ilmenite ore are expected to contain a certain amount of iron element. The element composition was investigated by EDX spectrum ( Figure S6 and 7). It was found that the main elements in TiO 2 were Ti and O, and TH-TiO 2 were Ti, C, O, N, and S. It was notable that iron element was not detected in TiO 2 or TH-TiO 2 by XPS or EDX, this refuses the effect of iron oxide residue on catalytic activity of TiO 2 and TH-TiO 2 .
The oxidation states of C, N, and S in TH-TiO 2 were studied by XPS spectra (Figure 3). The survey XPS spectrum (Figure 3(a)) presents Ti2p peaked at 459.36 eV, C1s at 284.70 eV, O1s at 531.00 eV, N1s at 400.30 eV, and S2p at 168.01 eV. The XPS level core spectrum of C1s ( Figure 3(b)) shows a strong peak at 284.70 eV corresponding to the elemental carbon (graphite) that is formed by incomplete combustion of organic compounds under hydrothermal conditions. The spectral shoulders at 289.00 eV are associated with the presence of C=O bond of carbonate radicals on the catalytic surface [20]. Accord-ing to the XPS spectrum at 286.2 eV, there was a replacement of some Ti atoms by C in lattice and forming Ti-O-C structure [18,22]. The peak at 281 eV of the Ti-C bond was not found, indicating that C atom does not replace oxygen atoms in the TiO 2 lattice [23].
XPS core level spectrum of S2p ( Figure 3(c)) with peaks at 168.7 and 169.50 eV corresponding to S 6+ is replaced by Ti 4+ [24][25][26]. No peak presenting for Ti-S links was found around 160-163 eV. This result is consistent with some previous studies which suggested that S +6 replacing Ti 4+ in lattice is easier than replacing O 2with S 2- [24,27,28]. This substitution also contributes to reducing the band gap energy of TiO 2 by introducing the interband gaps and lowering conduction band. The XPS core level spectrum of N1s (Figure 3(d)) shows that a wide maximum range of 399-401 eV corresponding to the replacement of N into oxygen position in TiO 2 network [29,30] has contributed to increasing the valence band. Therefore, the band gap energy of TiO 2 decreases with nitrogen doping. The results of XPS analysis showed that carbon, nitrogen, and sulfur were added to the TiO 2 catalyst and their electronic structures change.
The UV-Vis diffuse reflectance spectroscopy (UV-Vis-DRS) is a useful technique to characterize the band gap energy of photocatalytic materials. In the present study, TC 20   Raman shift (cm -1 ) 3 Journal of Nanomaterials adsorption of TH-TiO 2 in the dark was conducted; therefore, TH-TiO 2 adsorbing tetracycline at equilibrium (TC adsorbed TH-TiO 2 ), along with TiO 2 and TH-TiO 2 , was characterized by means of UV-Vis-DRS. As can be seen from the figure, TH-TiO 2 sample had an absorption band shifting to higher wavelength compared with TiO 2 . The band gap energy of TiO 2 , TH-TiO 2 , and TC-adsorbed TH-TiO 2 samples calcu-lated by Kubelka-Munk function is 3.20, 2.88, and 2.74 eV, respectively (Figure 4(b)).
The red shift is thought to be C, N, and S doping to TiO 2 matrix, narrowing the band gap energy of the TiO 2 due to the occurrence of the hybridized states located in the band gap [31] or tetracyine as photosensitizer. The photoluminescence (PL) is widely used to study the recombination of photo-   Journal of Nanomaterials induced electron/hole pairs. The PL spectra of TiO 2 and TH-TiO 2 materials are shown in Figure 5. The materials were excited at 404 nm with a strong emission peak at about 468 nm. It was found that there is a significant decrease in luminescence intensity of TH-TiO 2 compared to TiO 2 . This demonstrates that the doping of C, N, and S into TiO 2 significantly improved the separation of photo-induced electrons, creating favorable conditions for reacting photo-generated holes and electrons with H 2 O or oxygen to active radicals [23,32]. From SEM images shown in Figures 6(a) and 6(b), it is observable that both of the pure and doped TiO 2 samples perform similar morphology which includes secondary particles with size varying from 500 to 600 nm constructed from primary subparticles. The only difference, however, could be observed is that, compared to pure sample, TH-TiO 2 exhibits significantly rougher surface with higher porosity (as clearly seen in the consistent insets) which could be related to the gas formation due to decomposition of thiourea during synthesis condition. The construction of secondary particles from subone could be clarified in HR-TEM images (as shown in Figure 6(c)) which indicates that the size range of the primary particles is around 12 to 18 nm. The high magnification TEM image in Figure 6(d) displays the observable lattice fringe corresponding to (101) plane with distance of 0.352 nm which is confirmed by Fast Fourier Transforms (FFT) (insets). The acceptable crystallinity of obtained sample was further proved via Selected Area Electron Diffraction (SAED) (Figure 6(e)) which includes separated rings formed form clear spot. The corresponding lattice planes were also indexed in SAED pattern.
The textural properties of the obtained materials were investigated by means of nitrogen adsorption/desorption isotherms (Figure 7). The curves of both materials belong to type V with type H1 loops according to the IUPAC classification which are all characterized for the mesoporous structure formed by intersecting particles [33,34]. The pore size curves exhibit the normal distribution with average sizes of 36.7 nm for TiO 2 and 13.8 nm for TH-TiO 2 (the inset of Figure 7). The doping of C, N, and S elements into TiO 2 matrix enhanced the specific surface area (the values of 36 m 2 ·g -1 for TiO 2 and 74 m 2 .g -1 for TH-TiO 2 as calculated by BET model) and reduced also the pore size of TH-TiO 2 .
3.2. Catalytic Activity. Some reports present that some natural minerals containing TiO shows photocatalytic activity in advanced oxidation process [35,36]; therefore, the present raw ilmenite was conducted to test photocatalytic activity as a reference. Figure 8(a) presents the kinetics of TC degradation over TiO 2 , raw ilmenite, and TH-TiO 2 . As can be seen in the figure, ilmenite does not exhibit any photocatalytic activity toward to oxidize TC due to the chemical inert property of ilmenite mineral. For TH-TiO 2 , the dark adsorption/desorption equilibrium is reached after 30 min, and it displays higher adsorption capacity than TiO 2 around 10% (F = 14:69% for TiO 2 and 24.63% for TH-TiO 2 in which F = 100 × C t /C 0 , where C 0 and C t are the initial concentration and at time t). It is worth noting that TH-TiO 2 yields a degradation efficiency of nearly 100% while TiO 2 yields only 50% after 120 min of visible light illumination. This may be due to larger surface area of TH-TiO 2 which increases the number of active sites for TC to adsorb and decompose. In the leaching experiment, the catalyst was filtered after 30 min, and the degradation of TC under illumination almost stops (Figure 8(a)). In addition, the degradation of TC without the catalyst is not observed after 120 min of illumination, revealing that TC is not photo-decomposed in the present condition. Therefore, it can be suggested that TH-TiO 2 acts as a heterogeneous catalyst in the TC photocatalytic degradation.  Journal of Nanomaterials transitions in benzene ring, and the latter is contributed to the conjugated double bonds. The decomposition of TC on TH-TiO 2 catalyst takes place quite quickly, after about 120 minutes of processing, the characteristic absorption peak of TC at 355 nm is almost disappeared. This result shows that TC decomposition has possible formed intermediates with smaller molecular masses. It is well-known that free radicals such as photo-induced electrons and holes, • OH and O 2 (•-) , play a critical role in the degradation reactions of organic compounds. In particular, tert-butylic alcohol (TB) is used as a quencher for • OH [37], 1,4-benzoquinon-(BQ-) quencher for O 2 (•-) [38,39], ammonium oxalate-(AO-) quencher for H + [37,39], and silver nitrate-(BN-) quencher for photosynthetic electrons e - [40]. The 2 mL of 10 mM quenching solutions was added just after 30 min of dark adsorption. The effect of extinguishing agents on TC degradation performance is shown in Figure 9. In general, the presence of free radicals reduces the efficiency of TC degradation. AO (quenching h + ), BQ (quenching O 2 (•-) ), and BN (quenching e -) reduce significantly the degradation rate of TC. However, TB seems not to affect TC degradation. This concludes that the free radicals (h + ; O 2 (• -) ; e -) take mainly part in degradation reactions of TC while • OH is negligible.
The tridoping of C, N, and S elements into TiO 2 lattice gives to rise new sublevels including S2p and C1s levels under Ti3d conduction band (CB); N2p level above the O2p valence band (VB) [41]. The sublevels of S2p and C1s served as trapping centers for photo-induced electrons that enhance the life times of the charge carriers due to the separation of   Journal of Nanomaterials

LC-MS Analysis.
The mechanism of TC photodegradation on TH-TiO 2 catalyst was studied by using the liquid chromatography-mass spectrometry (LC-MS) (see Scheme 2 and Figure S8-10 [45] (Scheme S1). For pathway 2 (Scheme S2), the ring opening products include compounds (5), (6), and (7), where (5) 4dimethylamino-2-formyl-3-hydroxybut-2-enamide (m/z = 171:8) is formed as a result of direct oxidation from TC (1) or from (4). The presence of 4-((dimethylamino) methyl)-2H-oxet-2-one (7) (m/z = 125:8) is due to the internal molecular reaction chain of (5). The reaction mechanism can be described as follows: (step 1) OH (= C-OH) acts as a nucleophile agent, interacting with C = O in the amide function group according to SN2 (CO) mechanism. At the same time, the dehydro group and  (6) is formed unstably because containing single electrons tends to turn into (7) through decarbonylation process. In solution, the -OH (H 3 C-C-OH) group of TC with high electron density will give way to H + , then dehydration process with hydrogen displacement also happens to form a more stable structural compound (9) (m/z = 427:0) [46,47] (Scheme S3). Compounds (11) (m/z = 320:9) and (12) (m/z = 238:9) are the result of NH 3 splitting and cutting off different bonds due to h + and e − , O 2•− . However, this process will not be clarified in this study because of the relatively complex mechanism and very short shelf life of intermediate compounds (pathway 3). Finally, compounds (13) and (14) are formed directly from TC or through intermediate compounds (4) (pathway 4). Besides, some compounds with small m/z (116.7, 91) were also found in the analysis process but were not identified because fragmentation data is relatively small. At the same time, the results of LC-ESI-MS analysis and the TOC measurement was subjected to reaction solution. Initial TOC was about 550 mg/L; however, it decreased 92% to 44 mg/L after 150 min treating, indicating the complete mineralization. It is concluded from LC-MS and TOC analysis that TC degradation over TH-TiO 2 catalyst proceeds via many different intermediates and they can eventually be converted into CO 2 and H 2 O.

Kinetic Study.
The pH of point of zero charge (pH PZC ) of TH-TiO 2 calculated by the pH drift method is 4.5  Journal of Nanomaterials ( Figure 10(a)). TC possesses three of pKa, i.e., pKa1 = 3:3, pKa2 = 7:68, and pKa3 = 9:7 [48].Then, TC is charged positively as pH is less than 3.3 and it is charged negatively as pH is more than 9.7. TC will be charged positively/negatively at pH within 3.3-7.68. The effect of pH on the TC degradation performance is illustrated in Figure 10(b). As can be seen from the figure, the TC degradation efficiency increases with an increase in pH and peaks around pH 4-6 and significantly decreases with further increase in pH. At pH is too low (<3. 3) or too high (>7.68), the degradation efficiency was decreased because pulse interaction of the same charge carries. At a range of pH of 3-6, isoelectronic interaction is predominated compared with pulse one.
3.5. Effect of Initial TC Concentrations. In this experiment, the initial TC concentration varied from 30 to 70 mg·L -1 , the other experimental conditions remained the same. It was found that when increasing the initial TC concentration from 30 to 70 mg·L -1 , decomposition efficiency decreases significantly from 96% to 55% after 120 minutes of visible light illumination (Figure 11(a)). The catalytic reduction performance by TC concentrations can be explained by the fact that the higher initial TC concentration, the more TC molecules adsorbed on the surface of the TH-TiO 2 . With a fixed amount of catalyst, excessive TC adsorption causing blockage of active sites on the substrate surface leads to negatively affecting photocatalytic activity. The Langmuir-Hinshelwood model 11 Journal of Nanomaterials was employed to analyze the kinetics data in which the linear plot of ln(C t /C o ) vs.t is constructed. Figure 11(b) presents the Langmuir-Hinshelwood plots at different concentrations. The high determination coefficients, R 2 (0.99-1) confirm that the kinetic degradation reaction of TC over TH-TiO 2 fixed well the Langmuir-Hinshelwood model. Table 2 presents the rate constants of TC degradation reaction over different catalysts. Although the comparison is lame because the reaction conditions are not the same. However, from Table 1, it can be seen that the rate constant of TH-TiO 2 is relative high compared to previous studies.
3.6. Reusability. Reusability is one of the very important factors when deciding to choose a catalyst for economic and environmental purposes. The used TH-TiO 2 material was washed many times with distilled water and dried at 80°C for 12 hours for regeneration. The TC degradation efficiency over reused catalyst is presented in Figure 12(a). This result shows a slight reduction in TC decomposition efficiency, but after four reuse times, effective TC decomposition still reached over 89.0%. The XRD patterns of TH-TiO 2 (Figure 12(b)) seem slightly changeable indicating TH-TiO 2 possessed excellent structural stability after the regeneration process.

Catalytic Decomposition of Other
Compounds. The photocatalytic activity of TH-TiO 2 was tested for some dyes and antibiotics ( Figure 13). It was found that the present TH-TiO 2 could catalyze excellently for dyes (rhodamine-B, methylene blue, and organe red) or antibiotics (ciprofloxacin and chloramphenicol). This offers a potential application of TH-TiO 2 material in the removal of antibiotics in particular and organic pollutants in general from shrimp pond wastewater.

Conclusions
The tridoping of S, N, and C into TiO 2 extracted from ilmenite ore by hydrothermal process with thiourea was demonstrated. The S-N-C-tridoped TiO 2 exhibited an excellent catatalytic activity toward the complete minerazation of tetracyline. The possible pathways of tetracyline phototocatalytic degradation over obtained catalyst include hydroxylation, dealkylation, and dehydration via intermediates with smaller molecular mass to final products of CO 2 and water. In addition, the S-N-C-tridoped TiO 2 are potential for photocatalytic degradation of other antibiotics (ciprofloxacin and chloramphenicol) and some dyes (rhodamine-B, methylene blue, and organe red).

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
The authors declare that they have no conflicts of interest. Figure S1: XRD patterns of TH-TiO2-a (a = 400, 500, 600, and 700°C). Figure S2: nitrogen adsorption/desorption isotherms of TH-TiO2-a (a = 400, 500, 600, and 700°C). Figure  S3: UV-Vis DRS spectra of TH-TiO2-a (a = 400, 500, 600, and 700°C) and TiO2. Figure S4: the Kubelka-Munk plots for TH-TH-TiO2-a (a = 400, 500, 600, and 700°C) and TiO2. Figure S5: the plot of C/C o vs. time for adsorption and photocatalytic reaction or TiO2 và TH-TiO2-a (a = 400, 500, 600, and 700°C), (C o = 30 mg/L and m catalyst = 0.6 gram·L-1); V = 100 mL) Figure S6: EDX spectrum and the elements content of TH-TiO2. Figure S7: EDX spectrum and the elements content of TiO2. Figure S8: LC diagram at the retention time of 11.9 min (upper) and its mass spectrum (lower). Figure S9: LC diagram at the retention time of 30 min (upper) and its mass spectrum (lower). Figure S10: LC diagram at the retention time of 16.4 min (upper) and its mass spectra (lower). Table S1: physical chemistry properties of TH-TiO2-a. Scheme S1: the proposed fragmentation mechanism of the compound at the retention time of 11.9 min. Scheme S2: the proposed fragmentation mechanism of the compound at the retention time of 30 min. Scheme S3: the proposed fragmentation mechanism of the compound at the retention time of 16.4 min. (Supplementary Materials)