Effects of Ca2+ and fulvic acids on atrazine degradation by nano-TiO2: Performances and mechanisms

In this study, the adsorption and UV photocatalytic degradation of atrazine using nano-TiO2 particles were studied systematically, and the colloidal stability of nano-TiO2 particles in solution was also investigated to reveal the removal mechanism. Experiments which contained the first 6.0 hours darkness and 4.0 hours UV illumination later were conducted at different concentrations of Ca2+ and/or fulvic acids (FA) at pH = 7.0. Results showed that the adsorption rate of atrazine onto nano-TiO2 particles decreased with the increase of Ca2+ and/or FA concentrations, which could be explained well by the colloidal stability of nanoparticles. When the solution contained Ca2+ or Ca2+-FA, the nanoparticles were aggregated together leading to the decrease of the contact surface area. Besides, there existed competitive adsorption between FA and atrazine on the particle surface. During photocatalytic degradation, the increase of Ca2+ and/or FA concentration accelerated the aggregation of nano-TiO2 particles and that reduced the degradation efficiency of atrazine. The particle sizes by SEM were in accordance with the aggregation degree of nanoparticles in the solutions. Sedimentation experiments of nano-TiO2 particles displayed that the fastest sedimentation was happened in the CaCl2 and FA coexistent system and followed by CaCl2 alone, and the results well demonstrated the photodegradation efficiency trends of atrazine by nano-TiO2 particles under the different sedimentation conditions.

Generally speaking, pesticide contaminated wastewater usually contains many different ionic species, such as K + , Mg 2+ , NO 3 − , Cl − etc., and some natural organic matter 21 . These chemicals can influence the efficiency of photocatalytic materials during contaminants removal. A few studies have reported the effects of chemicals on photocatalytic activity of photocatalytic materials.  conducted experiments on the photocatalytic degradation of 2-chloro and 2-nitrophenol by TiO 2 in aqueous solution. They found that chloride ions seriously inhibited the photocatalytic reaction at pH 3.0, nitrate ions and sulfate ions had a slight inhibition effect 22 . Černigoj and co-workers (2010) studied the effects of dissolved ozone or ferric ions on the photodegradation of thiacloprid in the presence of TiO 2 catalysts 23 . They observed that dissolved iron(III) species did not promote the photocatalytic degradation of thiacloprid by TiO 2 . Cruz et al. 24 performed their experiments for the degradation of selected pesticides by bare TiO 2 and grapheme oxide TiO 2 under the conditions of ultrapure and natural water 24 . Their results showed that natural water decreased the degradation of four pesticides (diuron, alachlor, isoproturon and atrazine) using bare TiO 2 as the photocatalyst, however, the degradation of four pesticides was not affected in ultrapure water. They attributed the difference to inorganic and organic species in natural water that inhibited the photocatalytic process. Overview, most of researchers discussed the effects of chemicals on the generation of active free radicals by nanomaterials. Few examples regarding the effects of chemicals on the colloidal stability of nano-TiO 2 affecting photocatalytic activity are available.
In this work, atrazine was selected as the target pollutant and commercial nano-TiO 2 particles were employed as the photocatalyst. The effects of Ca 2+ and/or fulvic acids (FA) on the adsorption and photocatalytic degradation of atrazine were systematically studied. The degradation mechanism was the hypothesis that Ca 2+ and/or FA could influence the colloidal stability of nano-TiO 2 particles so as to influence degradation of atrazine. Therefore, the effects of Ca 2+ and/or FA on colloidal stability of nano-TiO 2 particles (zeta potential, hydrodynamic diameter (HDD) and sedimentation kinetics) were investigated. The possible relationship between colloidal stability and the photocatalytic properties of nano-TiO 2 particles in the presence of Ca 2+ and/or FA was discussed.

Results and Discussion
Characterization of commercial nano-TiO 2 . The characteristics of the commercial nano-TiO 2 with an average particle size of 5-10 nm used in the experiments were shown in Supporting Information (Figs S1-4). As observed in scanning electron microscopy (SEM) image (Fig. S1a), the shape of particles were irregular spheres, and they were aggregated strongly which probably due to thermodynamic stability 25 . Elemental analysis by energy dispersive X-ray (EDX) of the TiO 2 particles (Fig. S1b) showed that the particles were consisted of Ti (58.76 wt%), O (40.56 wt%) and a small amount of silicon impurity (0.58 wt%). The FT-IR spectrum (Fig. S2) showed that nano-TiO 2 appeared two absorption peaks at 516.92 and 3437.15 cm −1 , which corresponded to the vibration of Ti-O-Ti and the adsorption of -OH or H 2 O on nanoparticles respectively. Figure S3 showed that the diffraction peaks at 2θ values of 25.08°, 37.12°, 47.04°, 53.9° and 61.8° of the nanoparticles corresponded to the (101), (004), (200), (211) and (204) planes, respectively. The diffraction peaks were in consistent with the TiO 2 anatase which were in accordance with the parameters of nano-TiO 2 supplied by the company. Surface charges of the nano-TiO 2 particles over the pH range of 1.0-11 were investigated and the point of zero charge of pH (pH pzc ) was 6.2 ( Fig. S4) as with previous reports 26 . TiO 2 was an amphoteric oxide semiconductor. The nano-TiO 2 surface was positively charged when pH < 6.2, while at pH > 6.2, the nanoparticles were negatively charged 27 . www.nature.com/scientificreports www.nature.com/scientificreports/ Effect of Ca 2+ on the photocatalytic degradation of atrazine. The effect of Ca 2+ on the removal of atrazine by nano-TiO 2 was evaluated in suspensions with 10 mg/L of nano-TiO 2 , 1.0 mg/L of atrazine and different concentrations of CaCl 2 at pH 7.0. Results were presented in Fig. 1. In the first six hours of darkness, adsorption on nano-TiO 2 surface was the primary mechanism for the removal of atrazine, and then followed by photocatalytic degradation in the remaining four hours of UV irradiation.
During the dark period, adsorption of atrazine onto the nano-TiO 2 surface decreased significantly with the addition of Ca 2+ . Without Ca 2+ , the removal efficiency of atrazine by nano-TiO 2 was 49.2% after six hours. The addition of 1.0 mmol/L CaCl 2 resulted in a decrease in adsorption to 17.6%, and the removal efficiency decreased to 7.6% when the CaCl 2 concentration was increased to 100 mmol/L. During UV irradiation, the photocatalytic degradation efficiency of atrazine also decreased with the addition of Ca 2+ . With no addition of CaCl 2 , the photodegradation rate was approximately 0.46 C/C 0 /hr (from hour 6.5 to 7.5). However, the increase of CaCl 2 from 1.0 to 100 mmol/L resulted in a lower rate of 0.35 to 0.32 C/C 0 /hr during the same time period. Atrazine was completely degraded after 2.0 hours of UV irradiation with the absence of CaCl 2 . Nevertheless, when the concentration of CaCl 2 was 100 mmol/L, there was still 38.9% of atrazine left in the solution after 2.0 hours of photocatalytic process, and it wasn't completely degraded till the experiment finished. Similar results were reported by other investigators for other organic pollutants. Dionysiou and co-workers (2000) studied the influences of KNO 3 and H 2 O 2 on the removal of 4-chlorobenzoin (4-CBA) by TiO 2 powders 28 . They found that the removal efficiency of 4-CBA by TiO 2 decreased with the increase of KNO 3 concentration at both adsorption and photocatalytic degradation processes. Maybe due to the high TiO 2 loading, the complete degradation of 4-CBA was achieved at 3.0 h which was faster than that in our study.
In order to explained the mechanism of atrazine removal by nano-TiO 2 with the increase of Ca 2+ concentration, the colloidal stability (zeta potential and HDD) of nano-TiO 2 suspensions (pH 7.0) in different concentrations of CaCl 2 was studied systematically. The results were shown in Fig. 2. With the addition of CaCl 2 concentration, the HDD of nano-TiO 2 particles increased and the zeta potential changed from negative to positive values but still near zero, which meant the aggregation occurred between the nanoparticles. These changes could have reduced the actual contact surface area and active adsorption sites of nano-TiO 2 thus decreasing the adsorption capacity of nanoparticles for organic pollutants 29 . This suggested that the absorption and photodegradation mechanism of atrazine behaved much differently in distilled water and simulating natural waters.
According to the report by Chen and Liu 27 , the photocatalytic mechanism in the presence of TiO 2 could be described by Fig. 3 and the equations as follows: From the above equations and Fig. 3, the hydroxyl radical (OH) was the main reactant produced by UV irradiation on the surface of TiO 2 for atrazine photocatalytic degradation. When adding Ca 2+ , the aggregation of nanoparticles increased the contact thereby decreasing the exposed surface area and the generation of hydroxyl www.nature.com/scientificreports www.nature.com/scientificreports/ radicals. These were resulted in a reduction of photocatalytic degradation efficiency 30 . From Figs 1 and 2, the mechanism of atrazine removal by nano-TiO 2 at the CaCl 2 solution could be well explained by the colloidal stability of nano-TiO 2 . However, it should be noted that the decrease trend of photocatalytic degradation efficiency was not high as expected in Ca 2+ concentration.  Fig. 4. During the dark period of the experiments, the adsorption of atrazine onto the nano-TiO 2 decreased slightly with increasing FA concentration from 1.0 to 10 mg/L. During UV irradiation, complete photocatalytic degradation of atrazine by nano-TiO 2 was only obtained after 10 hours in the absence of FA, and over 90% of atrazine was degraded in the solutions containing FA. In addition, the higher amount of FA added, more residual atrazine was left in the solution, which suggested that FA inhibited the removal capacity of nano-TiO 2 . The results obtained in this study were agreed with the research by  which investigated the effects of pH, inorganic ions and humic acids on the photocatalytic degradation of 2-chlorobiphenyl (2-CB) by TiO 2 31 . However, in their study, the decrease of 2-CB degradation with the increase of humic acids concentration was ascribed to the competition between humic acid and 2-CB.

Effects of Ca
The results of atrazine degradation in the presence of Ca 2+ and FA could be attributed to the colloidal stability of nanoparticles suspensions and the competition for hydroxyl radicals by FA 32 . The presence of Ca 2+ and NOM, such as FA, could form ion bridge effects resulting in intensified aggregation of nanoparticles 33 . The aggregation could reduce the active adsorption sites of nano-TiO 2 , thereby decreasing the adsorption capacity of nanoparticles. The HDD and zeta potential of nano-TiO 2 suspensions in the presence of Ca 2+ and FA were investigated and the results were presented in Fig. 5. When the solution containing 10 mmol/L CaCl 2 , the HDD of nano-TiO 2 increased with the increase of FA concentration. Moreover, FA had a high adsorption capacity and could compete for active adsorptive sites of nano-TiO 2 34,35 . From Fig. 4, the adsorption efficiency of atrazine decreased with the addition of FA in the first six hours. The results could well validate the analysis acquired by the above-mentioned  www.nature.com/scientificreports www.nature.com/scientificreports/ research. During UV irradiation, on the one hand, the aggregation of nanoparticles decreased the generation of hydroxyl radical because of the recombination of generated holes with electrons from adjacent nanoparticles 30 . On the other hand, FA as one kind of organic matter could quench hydroxyl radical generated by nano-TiO 2 under the UV irradiation 31 . Moreover, the photocatalytic degradation efficiency of atrazine by nano-TiO 2 decreased with the increase of FA, which could also be related with the quenching effect of FA.
In order to well demonstrate the combined effect of Ca-FA, the effects of Mg 2+ and FA on the colloidal stability and photocatalytic activity of nano-TiO 2 were also studied. The experiments were performed as above and the results were shown in Figs 5 and 6. Comparing with the addition of 10 mmol/L CaCl 2 , the adsorption and photocatalytic degradation efficiencies of atrazine by nano-TiO 2 were almost the same in the presence of 10 mmol/L MgCl 2 (Fig. 6). And from Fig. 5, it could be seen that the nanoparticles were also similar in size. However, when in metal ion-FA coexistent system, the influence on the property of nano-TiO 2 was different. When the solution containing MgCl 2 -FA, the HDD of nano-TiO 2 was smaller and the degradation efficiencies of atrazine by nanoparticles were higher than in the CaCl 2 -FA solution. The reason could be that the ion bridge effects couldn't form in the Mg-FA system 36  From Fig. 7, it could be seen that the size of nano-TiO 2 was smallest in the solution cotaining 1.0 mmol/L CaCl 2 . Increasing CaCl 2 concentration, the degree of nano-TiO 2 aggregation was increased (Fig. 7a-c). When FA  www.nature.com/scientificreports www.nature.com/scientificreports/ was added to the 10 mmol/L CaCl 2 solution, nano-TiO 2 particles were gathered together and the size of nanoparticles were increased sharply (Fig. 7d). Even 1.0 mg/L FA added, the size of nanoparticles was bigger than that in the solution containing CaCl 2 . With the increase of FA concentration, the aggregation phenomena was getting serious (Fig. 7d-f). The results of SEM images were in accordance with the HDD of nanoparticles measured by the Nano ZS90 Malvern Zetasizer (Figs 2 and 5). The reduction in photocatalytic degradation efficiency of atrazine by nano-TiO 2 under sedimentation was explained by the decrease in the number of nano-TiO 2 particles (Fig. 9). Under sedimentation condition, the smaller number of nano-TiO 2 particles resulted in less surface area for UV exposure, and the output of hydroxyl radical was decreased for organic pollutant oxidation 37 . Fast sedimentation could decrease the contact time between nanoparticles and target pollutants, which could quench the hydroxyl radical produced by nano-TiO 2 38 . This decreased the photocatalytic performance of nano-TiO 2 . Sedimentation was fastest in the CaCl 2 and FA coexistent system and followed by CaCl 2 alone, supporting the contention that the addition of Ca 2+ and Ca 2+ -FA badly affected the colloid stability and catalytic activity of nano-TiO 2 particles.

Conclusion
In this study, attrazine degradation efficiency and colloidal stability of nano-TiO 2 were investigated systematically. The adsorption and photocatalytic degradation of atrazine by nano-TiO 2 were negatively affected by the addition of Ca 2+ and fulvic acids (FA) in aqueous solutions. The results suggested that the removal of atrazine by nano-TiO 2 was controlled by colloidal stability and adsorption interferences in the presence of Ca 2+ and/or FA. The addition of Ca 2+ could cause aggregation of nanoparticles by compressing the electric double layer, while FA could interfere by competitive adsorption. In photocatalytic degradation, the increase of particle size decreased the generation of hydroxyl radical. Besides, FA could quenche the hydroxyl radicals, thereby reducing the degradation efficiency of atrazine. Under sedimentation conditions, the number of nano-TiO 2 particles decreased in all solutions. Due to smaller available surface area, the photocatalytic degradation of atrazine decreased. During the sedimentation, the number of nano-TiO 2 particles remaining in solutions containing Ca 2+ and Ca 2+ -FA was less than the control, which demonstrated the negative effect of Ca 2+ and FA on the colloidal stability and catalytic activity of nano-TiO 2 particles.

Methods
Preparation of reagents. Commercial TiO 2 (anatase) nanoparticles (nano-TiO 2 ) were provided by Aladdin Chemistry Co. Ltd. The average particle size was 5-10 nm and the content of TiO 2 was over 99.8% as reported by the company. Atrazine of analytical grade was purchased from Shanghai Yuanye Bio-technology Co., Ltd, China, and stored at 4 °C before the experiment. FA with a molecular weight of 308.24 g/mol was obtained  www.nature.com/scientificreports www.nature.com/scientificreports/ from the Shanghai Luzong Chemical Reagent Co., Ltd without additional purification. Other chemical reagents employed in this study, including CaCl 2 , MgCl 2 , NaOH and HCl, were all of reagent grade and obtained from Damao Chemical Reagent Co, Tianjin, China.
A nano-TiO 2 stock solution (50 mg/L) was prepared immediately before use with ultrapure water (Barnstead D11911), and sonicated at 25 °C for 30 min with the ultrasonic power of 100 W and frequency of 40 kHz. Atrazine was dissolved in ultrapure water to obtain the 10 mg/L stock solution and stored at 4 °C without light. FA stock solution with a concentration of 1000 mg/L was prepared by dissolving FA in ultrapure water and stored at 4 °C before the experiment. A stock solution of CaCl 2 (1.0 mol/L) used as Ca 2+ was prepared in the same manner. The 1.0 mol/L stock solution of MgCl 2 was prepared as above. The chloride ion was chosen as anion in the experiments due to its little influence on the degradation and colloidal stability of nano-TiO 2 39 .
When performing a given experiment, the concentrations of all reagents used were prepared by diluting the stock solution. All containers used in the study were washed and dried carefully to prevent dust interference.
Atrazine degradation by nano-TiO 2 in the presence of Ca 2+ and/or FA. Photocatalytic reactor. In this study, a small self-made ultraviolet photocatalytic reactor which was also used in our other experiment was employed for atrazine removal 40 . Photocatalytic experiments were performed in a cylindrical Pyrex glass cylinder (diameter 9.0 cm, height 12 cm) containing a 100 mL aqueous sample under a 15 W tube-like ultraviolet lamp (GPH843T5VH, Longpro Co., Ltd, Guangzhou, China). The solutions were stirred by a 85-2 digital magnetic stirrer (Changzhou Guoyu instrument manufacturing co., LTD, Jiangsu, China) at 250 rpm. The distance between ultraviolet lamp and solution surface was 25 cm in order to maintain a fixed intensity of light. A lightproof casing was used for avoiding the contact of photocatalytic simulator with outside.
Photocatalytic experiments. The photocatalytic experiments containing nano-TiO 2 particles and atrazine were performed with CaCl 2 , MgCl 2 and/or FA. The experiments were carried out at pH 7.0 and a temperature of 30 °C. Each experiment lasted for 10 hours and was divided into two parts: 6.0 hours of darkness (0-6 h) and then followed by 4.0 hours of UV illumination (6-10 h) 28 . In the experiments, the concentrations of nano-TiO 2 and atrazine were 10 and 1.0 mg/L, respectively, and the total volume of the mixture was 100 mL. In order to obtain a well dispersed solution, the experimental timing began after stirring for 10 min at 250 rpm. 1.0 mL samples were taken out from the reactor at set times and filtered with a 0.45 μm nylon syringe filter. The atrazine concentration of the sample was measured. Sampling times were 0, 2.0, 4.0, 5.0, 6.0, 6.5, 7.0, 7.5, 8.0, 9.0, 10 h.
To study the effects of Ca 2+ concentrations on the atrazine degradation by nano-TiO 2 , proper volumes of stock CaCl 2 solution and ultrapure water were added to the mixture at the 4 th hour. To reduce experimental error, the additional total volume of stock solution and ultrapure water was 1.0 mL in all experiments. When CaCl 2 was added to the solutions, NaOH or HNO 3 was added quickly to readjust the pH to 7.0. To reduce the additional volume of acid or base, special care was taken to add as little HNO 3 or NaOH as possible. To evaluate the effects of Ca 2+ and FA on the atrazine degradation, 10 mmol/L CaCl 2 and FA (1.0, 5.0 10 mg/L) were added at the very beginning and the 4 th hour respectively. The other experiments were performed as above. The effects of Mg 2+ and FA on the atrazine degradation by nano-TiO 2 were conducted as above.
In order to study the effect of TiO 2 sedimentation on the atrazine photocatalytic degradation at the concentrations of 10 mmol/L CaCl 2 and/or 10 mg/L FA, the experiment of atrazine degradation by nano-TiO 2 was the same as above except that stirring was stopped during the 4.0 hours of UV illumination.
All photocatalytic experiments were conducted in duplicate and the average atrazine concentration was used to analyze the result.
Colloidal stability of nano-TiO 2 suspension. Colloidal stability of nano-TiO 2 particles were studied by examining the zeta potential and HDD of nano-TiO 2 particle suspensions in the presence of Ca 2+ and/or FA at www.nature.com/scientificreports www.nature.com/scientificreports/ pH 7.0. For all colloidal stability experiments, the concentrations of nano-TiO 2 particles were 10 mg/L. The suspensions containing CaCl 2 (1.0, 10, 100 mmol/L) and/or FA (1.0, 5.0, 10 mg/L) were prepared by adding appropriate volumes of the stock solutions and stirring at 250 rpm, 25 °C for 30 min. The zeta potential and HDD of the nano-TiO 2 suspensions were measured immediately after stirring. All aggregation experiments were carried out in duplicate and the average values were used for analysis.
To evaluate the sedimentation kinetics of nano-TiO 2 particles in the present of Ca 2+ and/or FA, the mixture containing 10 mg/L nano-TiO 2 , 10 mmol/L CaCl 2 and/or 10 mg/L FA was stirred at 250 rpm, 25 °C for 30 min. Then the absorbance (A) of nano-TiO 2 was measured in drive-time mode for 4.0 hours 41 . Control experiments containing 10 mg/L of nano-TiO 2 were carried out in parallel.
Photocatalyst characterization. The surface morphology and sample dimensions of the commercial nano-TiO 2 were determined by SEM (FEI QuANTA 200, USA) 42 . Quantitative detection and localization of elements in the photocatalyst were measured using an energy dispersive X-ray (EDX). The FT-IR spectrum was measured by a Fourier transform infrared spectrometer (Infinity-1, Shimadzu, Japan) in the range of 400-4000 cm −1 . A Bruker AXS D8 advance diffractometer with Cu radiation under 40 kV and 250 mA was employed for measuring the X-ray diffraction (XRD) patterns of nanoparticles. The pH pzc of the nano-TiO 2 particles was measured by a Nano ZS90 Malvern Zetasizer (Malvern Instrument, Worcestershire, UK) 43 .
In order to measure the morphology of nano-TiO 2 in solutions containing Ca 2+ and/or FA, the mixed solution was stirred for 30 min, and taken out a drop of sample to a clean silicon wafer (1.0 cm × 1.0 cm). Then the sample was dried for 24 h by a vacuum freeze dryer (WLFD-1-50, Beijing Bairui Weilai Analysis instrument co., LTD). After dried, the sample was sprayed gold for 45 s and was evaluated by the SEM (FEI QuANTA 200, USA).
Analytical methods. High performance liquid chromatography (HPLC) (Agilent 1100 Series with quaternary pump) with a C18 column UV detector (5 um, 4.6 × 150 mm) was employed to analyze the concentration of atrazine (C) 44 . 20 μL samples were injected into the instrument and monitored at 230 nm for 7.0 min. The mobile phase was kept constant at 30% HPLC grade water, 60% HPLC grade methyl alcohol and 10% HPLC grade acetonitrile. The flow rate was 1.0 mL/min and the measuring temperature was 40 °C.
The Nano ZS90 Malvern Zetasizer was employed to measure the zeta potential and HDD. The zeta potential was determined from the electrophoretic mobility by the Smoluchowski model, and HDD was obtained from the diffusion coefficient by the Stokes-Einstein equation 33,45 . For each sample, the zeta potential value was obtained from the average of 30 measurements and HDD was measured once. Before measuring, the Malvern Zetasizer was performed at 25 °C for 1.0 min to equilibrium 33 . A new disposable folded capillary cell and polystyrene cuvette were used to measure the zeta potential and HDD for each sample, respectively.
The absorbance of nano-TiO 2 was measured by an UV-Vis spectrophotometer (UV-2550, Shimadzu, Japan) at 343 nm for the sedimentation kinetics experiments. The absorbance values were obtained every minute for 4.0 hours for each sample. The temperature was kept at 25 °C during the experiment.