Visible light photocatalytic mineralization of 17α-ethinyl estradiol (EE2) and hydrogen evolution over silver and strontium modified TiO2 nanoparticles: mechanisms and phytotoxicity assessment

The School of Civil Engineering and Architec – 241000, P. R. China. E-mail: wxyu1027@ Jiangsu Academy of Environmental Indus 210036, P. R. China Biofuels Institute, School of the Environme Zhenjiang – 212013, PR China School of Civil Engineering, Chongqing Thr P. R. China Ningbo Water Supply Co Ltd, Ningbo – 315 † Electronic supplementary informa 10.1039/c7ra12638g Cite this: RSC Adv., 2018, 8, 4329


Introduction
Hydrogen is a high energy density gas which has been described as a potential energy source for many green energy 'environmentally friendly' technologies. Statistical reports on energy necessitate the development and utilization of low price earth abundant materials as suitable materials for renewable energy technologies. However, these kinds of materials have some inadequate properties compared to that of currently existing efficient materials. These requirements may be fullled by titanium based compounds, since titanium is the ninth most abundant element in the earth's crust and the seventh most abundant metal. Titanium dioxide (TiO 2 ) is a well-known photocatalytic material with its high stability, non-toxicity, low cost and high performance in the mineralization of organic pollutants present in air and water. 1,2 However, TiO 2 catalysts are effective under UV irradiation, due to its high band gap ($3.2 eV), restricting the use of TiO 2 in practical applications since the UV region falls within 5% of total solar light spectrum. This urges researchers to develop facile modied materials to utilize the photocatalytic activity of the material in the visible region of solar spectrum. Modifying TiO 2 lattice by introducing other elements through physical and chemical processes, has been found to be a useful approach to obtain visible light sensitized catalysts which can initiate the water splitting mechanism. 3 Co-doping with two different atoms into TiO 2 /ZnO matrices has attracted signicant attention since they can exhibit higher photo-catalytic activity and unique characteristics compared to doping them with a single element. [4][5][6][7] However, these metal dopants have to be used in small quantities to prevent recombination of photo-generated electrons and holes. 8 Thus, low concentration co-doping of cations and anions could be an effective solution to enhance the visible light absorption efficiency and also to reduce the recombination of the photo-generated charges.
The greater production and usage of pharmaceuticals ($4000 types of 100-200 000 tons per year) are the major reason for their increased levels of contaminants in water all over the world. 9 The pharmaceuticals that are enforced and consumed by humans and animals, which are further excreted and nally land up in wastewater treatment plants (WWTPs). Hence, it is an urge to reduce these pharmaceutical contaminants, because they can be very toxic to the ecosystems although they present in very low concentrations (ppb). These active pharmaceutical ingredients (APIs) are deliberately made to be static in which to remain biologically active with the aim of being effective in health care. Therefore, the remnants of these pharmaceuticals are very hard to treat by conventional methods in WWTPs, since only a fraction of them are being destroyed in the process. 10 On account of this, large amounts of these pharmaceutical pollutants were detected in aquatic environments and even in drinking water during the past years. These emerging contaminants were declared by US EPA as chemical compounds which are not regulated properly and whose effect on the environment and human health is poorly understood yet. 11 17a-Ethinylestradiol (EE2) is a synthetic estrogen and endocrine-disrupting chemical (EDC) which is ranked in the top 100 of priority PPCPs. 12 17a-Ethinylestradiol is one of the major components in contraceptive pills and postmenopausal hormonal appurtenances. By means of its removal during the treatment, EE2 is more resistant than other normal estrogen due its complicated chemical nature. 13 Oen EE2 is also found in wastewater treatment plants (WWTPs) and is discharged into receiving waters due to its incomplete removal during the treatment process. 14 EE2 can induce the feminization in male sh and distort the reproductive potential and reduce the sh population even at trace levels of concentration. 15 Adverse effects on ecosystems, human health and drinking water safety, and its inefficient removal from WWTPs, it urges to explore effective techniques for EE2 removal within WWTPs.
Heterogeneous photocatalysis with visible light irradiation of metal modied TiO 2 nanoparticles has come forth as an effective technology for hydrogen generation and degradation of many pharmaceuticals, [16][17][18] for water reuse in agriculture and energy generation. The doping of two different atoms into TiO 2 lattice has attracted signicant interest since it can exhibit higher photo-catalytic activity and unique characteristics compared to doping with a single element in visible light. 19,20 Through electron trapping, Ag and Sr doping could possibly inhibit the recombination of electrons and positive holes. Furthermore, Sr is one of the widely available alkaline earth metals exhibiting good electrical properties on doping with other metal oxides. 21,22 The present study deals with Ag and Sr modied TiO 2 nanoparticles for effective degradation of EE2 and hydrogen generation under visible light. The possible degradation pathway was investigated by HPLC and LC-ESI/MS analysis. Water reuse necessitates the phytotoxicity assessment of EE2 and their corresponding degradation metabolites. Hence, the phytotoxicity of the degraded metabolites and EE2 was investigated upon two common plant seeds P. vulgaris and V. radiata to insure the environmental impact of the treated water.

Materials and methods
Chemicals Titanium(IV) isopropoxide, silver nitrate, strontium nitrate, 2propanol, tween 20, hydrazine hydrate, 17a-ethinylestradiol (EE2), from Alibaba chemicals used in the present work were analytical grade reagents and used without further purication, millipore water was used in all experiments.

Synthesis of Ag and Sr modied TiO 2 nanoparticles
Metal modied TiO 2 nanoparticles (Ag/Sr-TiO 2 ) were prepared by sol-gel synthesis. In brief, a mixture of 5 mL of titanium(IV) isopropoxide and 50 mL isopropanol was added drop-wise to 200 mL of distilled water maintained pH $ 1.5 while stirring. To the above solution required amounts of aqueous solutions of AgNO 3 and Sr (NO 3 ) 2 (0.2 mol% to 1.0 mol%) were added dropwise under stirring. A small aliquot of distilled water and 50 mmol hydrazine hydrate were added to it followed by 5 mL of tween 20 and stirred for 30 min. The resultant sol was sonicated at 80 MHz for 90 min and then dried at 100 C in a hot air oven for 24 h to get the dry gel which was calcined at 400 C for 5 h to get the desired nanoparticles powder.

Photocatalytic degradation experiments
A laboratory scale photoreactor with Xe lamp (300 W) (emission spectrum shown in Fig. S1 †) and a lter (l # 420 nm) was used for visible light photocatalytic degradation experiments as shown in Fig. 1. In the photocatalytic experiments, 50 mL of the EE2 aqueous solution (10 mg L À1 ) containing 50 mg of photocatalyst samples were rst ultrasonicated, and then stirred in the dark for 1 h to ensure absorption-desorption equilibrium. 23,24 A small aliquot of the sample was collected at regular intervals and analyzed in HPLC during the photocatalytic irradiation for 6 h.

Photocatalytic hydrogen generation experiments
The photocatalytic hydrogen generation tests were carried out in a quartz reactor (180 mL) equipped with a loading port, sealed with a gas-tight rubber septum. 50 mg of the photocatalyst was dispersed by magnetic stirrer in 80 mL of aqueous solution containing 0.5 M Na 2 S and 0.5 M Na 2 SO 3 solution. The suspension was sonicated for 10 min before irradiation and purged with nitrogen gas for 30 min in order to remove all air in solution and headspace to ensure the reactor was in an anaerobic condition. 300 W Xe lamp with a lter (l # 420 nm) was used to irradiate the sample for 5 h. The generated H 2 gas was collected at regular time intervals and analysed using an offline Shimadzu GC-2014 gas chromatograph (GC), equipped with a thermal conductivity detector (TCD) with a molecular sieve (5 A) column, at 70 C, using N 2 as the carrier gas. Control and the blank (without catalyst) experiments were carried out under identical conditions for comparison.

Analytical methods
The remnant concentration of EE2 was determined using High Performance Liquid Chromatography (HPLC) technique. Agilent 1260 series with Eclipse XDBC18 (5 mm) reverse phase column (4.6 Â 150 mm) was used for the separation. A uorescence detector with an excitation wavelength of 280 nm and an emission wavelength of 310 nm was used in this study. 25 The mobile phase with a mixture of water and methanol (30 : 70 v/v) was used with an injection volume of 1 mL min À1 for 6 minutes. The analysis of EE2 degradation metabolites was carried by liquid chromatography tandem mass spectrometry (Agilent 1290 Innity Binary LC system, Agilent 6460 Triple Quadrupole LCMS/MS system employing the Zorbax eclipse plus C18 column; rapid resolution, 2.1 Â 50 mm, 1.8 mm). The mobile phase of methanol and water at 30 : 70 (v/v) was used for EE2 about 60 min. To obtain mass spectra electro spray ionization (ESI) was used under the ow of helium gas at approximately 1 mL min À1 and 16 V of fragment voltage.

Photoelectrochemical measurements
The electrochemical measurements were carried out in a threeelectrode quartz cells using CHI760B as an electrochemical workstation. Pt wire as the counter electrode and saturated calomel electrodes (SCE) as the reference electrode was used during the measurements. For the working electrode typically  5 mg of the photocatalyst was dissolved in 1.0 mL of DMF solution and sonicated for 10 min, then 20 mL of the suspension was dropped on FTO with an xed area of 0.5 cm 2 , and then dried at ambient temperature for 24 h to form photocatalysts modied FTO. For light source 500 W Xe arc lamp was utilized in the photoelectrochemical measurements.
Characterization of nanoparticles X-ray diffraction patterns of the synthesized nanoparticles were recorded on BRUKER D8-Advance X-ray diffractometer with Cu Ka source (l ¼ 1.5406 A). TEM imaging and EDS analysis were recorded on JEOL JEM 2100 high resolution transmission electron microscope (HRTEM). Diffused reectance spectra were recorded with a JASCO V-670 UV-vis spectrophotometer.
The photoluminescence (PL) spectra were obtained on a HITA-CHI F-7000 uorescence spectrophotometer. XPS data was acquired with a Kratos Axis Ultra 165 Spectrometer with a monochromated Al Ka X-ray source (ha ¼ 1486.6 eV).

Phytotoxicity test
The phytotoxicity test was carried out on two common crop seeds P. vulgaris and V. radiata for EE2 and its corresponding degradation metabolites products (50 mg L À1 ) generated during photocatalytic degradation process. Ethyl acetate was used to extract the degradation metabolites of EE2, the extract was dried and dissolved in 10 mL of distilled water to make a nal concentration of 50 ppm. The phytotoxicity assessment test was carried out systematically based on the prescribed guidelines in the literature 26,27 with minor modications. From each type of plant variety 10 seeds were sterilized using 1-5% sodium hypochlorite solution for 15 min, and then rinsed thoroughly with distilled water to ensure complete removal of sodium hypochlorite from the surface of the seeds. Seeds were then immersed with EE2 (50 ppm) and the degradation metabolites (50 ppm) solutions in 100 mL Erlenmeyer asks separately. Seeds immersed in distilled water were included beside as control. Aer interaction was nished, the all 10 seeds were placed on wet cotton in Petri dish and incubated at 25 AE 1 C in dark for 24 h. Only germinated seeds were picked out for further phytotoxicity analysis. Further test was conducted in standard beakers contains 50 ppm of EE2 and its degradation metabolites dissolved in 30 mL of 1.5% of agar. The beakers were kept in freezer for few minutes to harden agar media immediately. The germinated seedlings were then placed on the surface of the agar in each beaker and incubated at 25 AE 1 C in darkness. Aer 7 days of incubation the plants were separated from the agar and lengths of root and shoot was measured. All the experiments were carried out in triplicate and average data was reproduced in the present study.

Results and discussion
Characterization of nanoparticles Fig. 2a shows X-ray diffraction pattern of Sr/Ag-TiO 2 nanoparticle which was supported by anatase phase with the 2q peaks at 25.3 , 37.9 , 47.9 , 55.0 and 62.8 respectively. All the values are completely matched with standard anatase phase TiO 2 , however mixed phase (anatase and rutile) was observed in Ag doped TiO 2 (Fig. 2b) samples. Due to lower concentrations of dopant no signals were observed corresponds to Ag and Sr in XRD patterns, this could be revealed that doping of metal atoms into TiO 2 did not alter its phase crystallinity. The UV-vis absorption spectra of pure TiO 2 , Ag/TiO 2 and Sr/ Ag-TiO 2 are shown in Fig. 3a. The red shi in absorption band from 350 nm to 400 nm could be due to the photo-excitation occurring at valence band to conduction band. TiO 2 nanoparticles modied with single atom (silver) has lower shi in comparison to two atoms (strontium and silver, 1.0 mol%). Kubelka Munk extrapolation plot of hn vs. (ahn) 2 was used to get optical energy band gap values (E g ) as depicted in Fig. 3b. These values for TiO 2 , Ag/TiO 2 and Sr/Ag-TiO 2 nanoparticles were found to be about 3.17 eV, 3.06 eV and 2.89 eV respectively. It revealed that the titanium dioxide modied with two atoms (silver and strontium) was followed with a decrease in the band energy and an increase in the wave length. Further to understand the stability and recombination of photogenerated electron photoluminescence (PL) study was carried out (Fig. 3c). The PL spectra of the doped and undoped TiO 2 are shown near band gap emission (NBE) and blue or deep level emission. These results suggest that the quelling of NBE on doping TiO 2 with silver and strontium, which may be due to the inhibition of recombination of the photogenerated electron-hole pairs on doped TiO 2 . Fig. 3d shows the photocurrent response of the synthesised TiO 2 , Ag/TiO 2 and Sr/Ag-TiO 2 nanoparticles. Upon visible light irradiation, the photocurrent rapidly increases and reaching a platform. The photocurrent of Sr/Ag-TiO 2 was higher than that of Ag/TiO 2 and TiO 2 nanoparticles, suggesting   a higher separation efficiency of photogenerated charge carriers. 28 The TEM images of Ag/TiO 2 (Fig. 4a) and Sr/Ag-TiO 2 (Fig. 5a) nanoparticles indicate that the particles are ranges from 7 to 22 nm and 13 to 31 nm respectively. The SAED pattern of this samples are shown in Fig. 4b and 5(b), in which the dark rings on the right correspond to the standard polycrystalline diffraction rings for the anatase phase (indexed). High resolution TEM images (Fig. 4c and 5c) were obtained to understand more evident microstructure information of the nanoparticles to enable accurately analyzes the single grains and grain boundaries. EDS spectrum (Fig. 4d and 5d) indicates elemental compositions that are present in the TiO 2 support.
In order to conrm the chemical state and surface composition XPS analysis was carried out for Sr/Ag-TiO 2 nanoparticles. The apparent XPS scans of Ag 3d, Sr 3d, Ti 2p and O 1s levels are executed to understand whether the dopant Ag and Sr atoms are interlinking in the lattice of TiO 2 or react with Ti and O to form other compounds. Fig. 6a shows high-resolution XPS spectrum of Ag at 3d core levels, indicates that the Ag 3d 5/2 and Ag 3d 3/2 binding energies are observed at 368.1 and 374.1 eV respectively, which are corresponded to metallic silver. 29,30 The Sr 3d core level peak was observed at 133.7 was in complete agreement with the literature (Fig. 6b). 31,32 The binding energies of Ti 2p 3/2 and Ti 2p 1/2 (Fig. 6c), photoelectron peaks was observed at 459.2 and 464.9 eV respectively. 33,34 The broad spectrum of Sr/Ag-TiO 2 nanocomposite was shown in Fig. 3d, indicated that silver, titanium and strontium metal ions were present in their highest oxidation state. The presence of Ag and Sr in broad XPS spectrum and detection of no new compounds between Ag, Sr, O, and Ti atoms, may conrm that Ag and Sr atoms were doped into TiO 2 crystal lattice.

Photocatalytic hydrogen generation
In order to determine the photocatalytic performance of the synthesized photocatalysts, hydrogen evolution experiments were carried out under visible light irradiation in the presence of Na 2 S (0.5 M) and Na 2 SO 3 (0.5 M) solution. From the Fig. 7a, from H 2 evolution kinetic curves shows that the amount of H 2 increases with the irradiation time by the visible light (l $ 420 nm). The results indicate that TiO 2 nanoparticles modied with Ag and Sr show efficient photocatalytic hydrogen generation activity as compared to their single components of Ag-TiO 2 , and TiO 2 . Further, the Ag and Sr modied TiO 2 nanoparticles shown the highest hydrogen production activity. Fig. 7b indicates the average hydrogen production rates of asobtained samples. Obviously no noticeable H 2 was generated for pure TiO 2 under visible light irradiation. The hydrogen production rate on Ag-TiO 2 is low, which is only 19.6 mmol per one hour, while the hydrogen production rate of Sr/Ag-TiO 2 is 49.4 mmol h À1 . These production rates are 2.6 times higher than that of Ag-TiO 2 nanocomposite. These results suggest that the hydrogen production rate of TiO 2 is remarkably enhanced by doping silver and strontium ions in the TiO 2 crystal lattice. The enhanced H 2 evolution activity of Sr/Ag-TiO 2 also attributed by the heterojunction electric eld among the TiO 2 , silver and strontium atoms and visible light  harvesters of Ag and Sr (narrow bandgap). 35,36 As described in UV-vis DRS section, upon visible light irradiation on the surface of the catalysts, the photo-generated electrons in the CB of Ag and Sr tends to transfer to the CB of TiO 2 , meanwhile the holes in the VB of Ag and Sr will move to the VB of TiO 2 . With this process, the photo-induced carriers within this ternary system can be successfully separated as supported by the PL studies and photocurrent results, thus excellent photocatalytic activity can be achieved. The pertinent series of reactions at the surface of the semiconductor causing the hydrogen generation and pollutant degradation can be expressed as follows: Sr/Ag-TiO 2 + hn / e À + h + (1) H + + OH À (VB) / cOH (4) EE2 + cOH and O 2 c À / degradation metabolites

Photocatalytic degradation of 17a-ethinyl estradiol
The photocatalytic degradation kinetics of EE2 can be depicted by a Langmuir-Hinshelwood model (eqn (1)), which exemplied that the reactions took place at a solid-liquid interface.
The plot of ln(C 0 /C) versus the irradiation time with various photocatalysts in 6 h under visible light irradiation is shown in Fig. 8. The results indicated that degradation of EE2 follows rst-order kinetics because the regression coefficients (R 2 ) are all above 0.8856. Apparently, in all the tested photocatalysts for degradation of EE2, Sr/Ag-TiO 2 nanocomposite was found to be the best performance with rate constant k ¼ 0.1699 min À1 .
To further investigate the plausible degradation pathway of Further to investigate the stability and recycle photocatalytic performance of the as-prepared Sr/Ag-TiO 2 nanoparticles, photocatalytic degradation of EE2 is cycled for four times and the results are shown in Fig. S4. † Aer each run, the photocatalyst was recycled by centrifuging, washing, and drying and was used for the next run. As can be seen from the results, no obvious decrease in photocatalytic degradation performance can be found aer four cycles (94%), which demonstrates that the Sr/Ag-TiO 2 nanoparticles have high photocatalytic stability and excellent reusability.

Phytotoxicity test
The disposal of the untreated/partially treated effluents of EDCs not only causes severe environmental and health hazards but also has a direct impact on the soil, affecting its fertility. This necessitates the assessment of phytotoxicity of EDCs and its corresponding degraded metabolites. 17a-ethinyl estradiol (50 ppm) highly inhibited the germination rate of P. vulgaris and V. radiata, reducing it to 50%. Interestingly, seeds exposed to 50 ppm of the photocatalytic degradation products and distilled water (control) exhibited 100% germination rate in both the types of seeds. Furthermore, EE2 also showed its toxicity on the length of root and shoots of the plants. In the case of P. vulgaris the shoot and root lengths are 2.98 AE 0.83 and 1.15 AE 0.17 respectively. Similarly, in the case of V. radiata the shoot and root lengths are found to be 2.90 AE 0.31 and 1.02 AE 0.17 respectively. In contrast to this higher shoot and root lengths were observed in the case of seeds grown in degraded products as well as in control (Table 1) and (Fig. 10). Thus, these results of phytotoxicity test indicated the ability of the prepared nanoparticles to not only degrade toxic pollutant EE2 but also to detoxify it. Values are mean of germination seeds treated with SMX, RR-194 and degraded metabolites, signicantly different from the seeds germinated with tap water at*P < 0.05, **P < 0.01, by one-way analysis of variance (ANOVA).

Fig. 10
Phytotoxicity assessment of EE2 and its visible light degradation products.

Conclusions
Novel metal modied TiO 2 nanoparticles were synthesized and used in the photocatalytic degradation of 17a-ethinyl estradiol under visible light and generation of hydrogen. The production rate of H 2 for double atom doped TiO 2 (Sr/Ag-TiO 2 ) is 2.6 times higher than that of single atom dopant (Ag-TiO 2 ). Further the plausible transformation pathway of EE2 and degradation products were determined by LC/ESI-MS analysis. According to phytotoxicity assessment it is suggested that the degraded products induces better growth in root and shoots compare to pure 17a-ethinyl estradiol in P. vulgaris, and V. radiata, which evidenced that less toxic intermediates were generated during visible light degradation process. Thus, we conceive that the present study provides the ability of multifunctional nanomaterials of Sr/Ag-TiO 2 to simultaneously degrade and detoxify the contaminated water and also utilized for energy generation during photocatalysis.

Conflicts of interest
There are no conicts to declare.