An Efficient Photocatalyst for Fast Reduction of Cr ( VI ) by Ultra-Trace Silver Enhanced Titania in Aqueous Solution

For the purpose of establishing a simple route to prepare a metal-semiconductor hybrid catalyst efficiently and reduce its cost through precise doping noble metals. In this study, ultra-trace silver doped TiO2 photocatalysts were fabricated via a “green” ultrasonic impregnation-assisted photoreduction strategy in an ethanol system, and its photocatalytic performance was systematically investigated by utilizing Cr(VI) as the model contaminant. A schottky energy barrier was constructed in Ag@TiO2, which served as a recombination center and possessed superior photocatalytic activity for Cr(VI) reduction. The obtained catalysts exhibited a significant e−/h+ separation efficiency which directly led to an obvious photocatalytic property enhancement. Then, the resultant Ag@TiO2 (0.06 wt %, 30 min irradiation) showed about 2.5 times the activity as that of commercial P25 NPs for Cr(VI) degradation. Moreover, after five cycles, it still maintained considerably high catalytic ability (62%). This work provides a deep insight into preparation techniques of metal-semiconductor photocatalyst and broadens their application prospect.


Introduction
Nowadays, heavy metals pollution in aquatic environments poses serious risks to public health and ecological systems since the rapid development of industrialization [1].As a non-degradable and toxic metal ion, Cr(VI) has been commonly identified in aquatic circumstances, due to its carcinogenic, mutagenic and teratogenic effects on living creatures [2].The Ministry of Ecology and Environment of the People's Republic of China and World Health Organization (WHO) have regulated the maximum concentration authorized in drinking water as being below 0.05 mg/L (Sanitary Standard for Drinking Water Quality (2001)) [3,4].However, the inevitable drawbacks of high cost or secondary contamination still exist among conventional methods to remove heavy metals from aqueous solutions, such as chemical reduction [5,6], membrane separation [7], electrolytic [8], ferrite method [9,10], precipitation [11], adsorption [12,13], ion exchange [14] and biological method [15].Hence, it is important to explore a cost-efficient method for Cr(VI) removal.
Heterogeneous photocatalysis is achieved by the excitation of electrons from semiconductors' valence band (VB) to the conduction band (CB) upon light irradiation, wherein the excited electrons and holes can be directly applied in redox processes.Owing to the highly reductive property of photo-generated electrons, photocatalysis [16] can be served as a desirable reductive.Compared to the conventional methods, it has relatively broader application prospects due to its recyclable and non-secondary pollution advantages [17].For example, reduced graphene oxide and ZnO hybrids (RGO/ZnO) was utilized in different particle sizes for photocatalytic reduction of Cr(VI) [18].In parallel, the photocatalytic Pb(II) reduction over WO 3 /TiO 2 nanomaterials, the reduction capacity was increased from 52.2% to 74.7% via WO 3 doping [19].Moreover, The photocatalytic removal of heavy metal ions by TiO 2 nanotubes could be further enhanced through Cu 2+ loadings [20].
Furthermore, to improve the semiconductors' catalytic performance [21], the noble metal deposition [22] recently has been proven to be a predominant way to heighten the activity of photocatalysis by creating the Schottky barrier at the semiconductor metal interface, which helps in separating the photo-excited electrons in the CB onto the metal nanoparticles (NPs) [23].During the past decade, various noble metals such as Pt, Au, and Ag have been frequently immobilized on TiO 2 [24,25] or other semiconductors to enhance their photocatalytic performance.However, among a wide variety of metal-semiconductor hetero-structures [26], Ag@TiO 2 [27] photocatalysts aroused tremendous attention as they showed the most efficient plasmon resonance which was favorable for catalytic enhancement [28].Therefore, various methods to obtain Ag@TiO 2 composites have been developed, such as hydrothermal [29], sol-gel [30], magnetron co-sputtering [31], photodeposition [32], sedimentation and impregnation [33].These methods are usually either complex or require rather extreme conditions to operate.To obtain an efficient Ag@TiO 2 photocatalyst, the silver dosage is still relatively high (>1 wt %), which greatly elevates its cost and limits its actual application.For instance, Chen and Zhang [30] synthesized an Ag@TiO 2 compound through a novel one-pot sol-gel strategy with 3.9 wt % silver addition and systematically studied its potent antibacterial activities.Liu et al. [34] demonstrated the deposition of Ag nanoparticles on TiO 2 nanofibers with 4.65 wt % silver dosage by a facile polyol synthesis, and the result showed superior natural antibacterial activity and high-performance photocatalytic disinfection under solar irradiation.Considering the conventional complex procedures and high silver consumption in Ag@TiO 2 preparation, how to simplify the fabrication steps and control the silver doping amount would be the crucial steps to control the cost and broaden the practical application of Ag@TiO 2 .Particularly, how to obtain an efficient catalyst under an ultra-trace noble metal deposition.
In this work, a facile photoreduction strategy was utilized to in situ synthesize Ag@TiO 2 photocatalysts with ultra and exhibited superior photocatalytic activity on reducing highly toxic Cr(VI) to low-toxic Cr(III).The photoreduction properties of Ag@TiO 2 were evaluated in different influencing parameters, such as the amount of Ag, the initial pH of solution, the amount of catalyst dosage, the light intensity and the coexisting ions.The results shown that the synthesized Ag@TiO 2 bears good Cr(VI) reduction performance, which provides a deep insight for the application of photocatalysis in wastewater treatment.

Preparation and Characterizations of Ultra-Trace Silver Enhanced Titania Photocatalyst
The ultra-trace silver doped TiO 2 photocatalysts were prepared via a "green" ultrasonic impregnation assisted photoreduction strategy, as shown in Figure 1.Firstly, the color of ethanol and TiO 2 mixture maintained no change under dark conditions after AgNO 3 addition.During the illuminated period, the mixture's color gradually turned from initial ecru to light blue (the color of products gradually deepened with the Ag dosage increasing, such as yellow-brown, purple-brown, etc., which mainly due to the photogenerated electrons (e − ) assisted Ag + reduction process (Figure 1a).
Then, the obtained catalysts were used for Cr(VI) reduction (Figure 1b).Cr(VI) was first adsorbed on the surface of Ag@TiO 2 , and then Cr(VI) was reduced to Cr(III) gradually by the CB formed e − .Figure 2a shows the typical morphology and surface composition of the resultant Ag@TiO2 compound.The transmission electron microscope (TEM) image clearly shows the sphere-like synthesized material with an average diameter of ca.24.5 ± 2 nm.High resolution transmission electron microscope (HRTEM, Figure 2b) observation revealed that TiO2 nanoparticles crystallized well with a lattice fringes spacing of 0.35 nm, which well matches the d-spacing of (111) plane of TiO2 (d = 0.35 nm).Furthermore, the d-spacing of 0.235 nm could be found either which fits the (001) plane d spacing of TiO2 (d = 0.235 nm) and no obvious Ag nanoparticles could be found since the rather trace silver addition of 0.06 wt %.However, as shown in Figure S1 and Table 1, silver nanoparticles exist on the surface of TiO2.According to the BET surface area analysis (Tables S1 and S2), the Ag doping ratio and h + scavenger (ethanol) concentration do not influence the specific surface area of Ag@TiO2 obviously.Considering the mild experimental conditions, the obtained nanocompounds exhibited similar X-ray diffraction (XRD) patterns to the pure TiO2, as shown in Figures 3 and S2.Specifically, the peaks of Ag@TiO2 at 2θ values of 25.1°, 37.6°, 47.9°, 53.7°, 55.1°, 62.6°, 68.8°, 70.1°, and 75.0° might belong Figure 2a shows the typical morphology and surface composition of the resultant Ag@TiO 2 compound.The transmission electron microscope (TEM) image clearly shows the sphere-like synthesized material with an average diameter of ca.24.5 ± 2 nm.High resolution transmission electron microscope (HRTEM, Figure 2b) observation revealed that TiO 2 nanoparticles crystallized well with a lattice fringes spacing of 0.35 nm, which well matches the d-spacing of (111) plane of TiO 2 (d = 0.35 nm).Furthermore, the d-spacing of 0.235 nm could be found either which fits the (001) plane d spacing of TiO 2 (d = 0.235 nm) and no obvious Ag nanoparticles could be found since the rather trace silver addition of 0.06 wt %.However, as shown in Figure S1 and Table 1, silver nanoparticles exist on the surface of TiO 2 .According to the BET surface area analysis (Tables S1 and S2), the Ag doping ratio and h + scavenger (ethanol) concentration do not influence the specific surface area of Ag@TiO 2 obviously.Figure 2a shows the typical morphology and surface composition of the resultant Ag@TiO2 compound.The transmission electron microscope (TEM) image clearly shows the sphere-like synthesized material with an average diameter of ca.24.5 ± 2 nm.High resolution transmission electron microscope (HRTEM, Figure 2b) observation revealed that TiO2 nanoparticles crystallized well with a lattice fringes spacing of 0.35 nm, which well matches the d-spacing of (111) plane of TiO2 (d = 0.35 nm).Furthermore, the d-spacing of 0.235 nm could be found either which fits the (001) plane d spacing of TiO2 (d = 0.235 nm) and no obvious Ag nanoparticles could be found since the rather trace silver addition of 0.06 wt %.However, as shown in Figure S1 and Table 1, silver nanoparticles exist on the surface of TiO2.According to the BET surface area analysis (Tables S1 and S2), the Ag doping ratio and h + scavenger (ethanol) concentration do not influence the specific surface area of Ag@TiO2 obviously.Considering the mild experimental conditions, the obtained nanocompounds exhibited similar X-ray diffraction (XRD) patterns to the pure TiO2, as shown in Figures 3 and S2.Specifically, the peaks  Considering the mild experimental conditions, the obtained nanocompounds exhibited similar X-ray diffraction (XRD) patterns to the pure TiO 2 , as shown in Figure 3 200), (220), and (311) planes of Ag (JCPDS card number: 04-0783), respectively.However, since the Ag addition was rather low and overlapping by the TiO 2 characteristic peak at 38.1 • , these peaks were relatively weak in the XRD pattern.Finally, according to the XRD and TEM analyses above, the results revealed that Ag nanoparticles were formed and successfully loaded on the TiO 2 surface.200), (220), and (311) planes of Ag (JCPDS card number: 04-0783), respectively.However, since the Ag addition was rather low and overlapping by the TiO2 characteristic peak at 38.1°, these peaks were relatively weak in the XRD pattern.Finally, according to the XRD and TEM analyses above, the results revealed that Ag nanoparticles were formed and successfully loaded on the TiO2 surface.It was noteworthy that the obvious improvement of visible light absorption in the range of 400-700 nm could be noticed under the addition of Ag, which fits the gradual color change from ecru to purple (Figure 3b, insert).A significant optical adsorption enhancement could be observed in different Ag contents of 0.3 wt %, 1.0 wt % and 3.0 wt %, whose absorption peaks appeared as a red shift moved from 480 nm to 580 nm.However, there were almost no differences between Ag@TiO2 doping ratios of 0.01 wt % and 0.06 wt %, which were similar to bare TiO2.The observed peaks should be attributed to the surface plasmon resonance (SPR) effect of spatially confined electrons in Ag nanoparticles and this red-shift might be related to the interaction between Ag and TiO2.Moreover, the adsorption peak range might be continuously broadened as the Ag dosage increases.

Catalytic Activity for Cr(VI) Photoreduction
The catalytic properties of Ag@TiO2 photocatalysts were evaluated by Cr(VI) imitated heavy metal wastewater.As shown in Figure 4a, Cr(VI) was firstly adsorbed onto the surface of Ag@TiO2 and adsorption equilibrium could be reached within 30 min under dark conditions.The aquatic Cr(VI) could be fully reduced to Cr(III) within 60 min under UV irradiation.The results illustrated that the Cr(VI) photocatalytic reduction efficiency got improved through Ag loading, which might be due to the following reasons.The photoresponse range could be enhanced, as presented in Figure 3b.Then, the Schottky junction would be formed at the interface with Ag deposition.Under light irradiation, photo-generated electrons (e − ) tend to migrate and accumulate on Ag nanoparticles which served as effective electron acceptors, and the extra acceptors subsequently promoted the e − /h + separation efficiency and ultimately enhanced the photocatalytic reduction of Cr(VI).It is noteworthy It was noteworthy that the obvious improvement of visible light absorption in the range of 400-700 nm could be noticed under the addition of Ag, which fits the gradual color change from ecru to purple (Figure 3b, insert).A significant optical adsorption enhancement could be observed in different Ag contents of 0.3 wt %, 1.0 wt % and 3.0 wt %, whose absorption peaks appeared as a red shift moved from 480 nm to 580 nm.However, there were almost no differences between Ag@TiO 2 doping ratios of 0.01 wt % and 0.06 wt %, which were similar to bare TiO 2 .The observed peaks should be attributed to the surface plasmon resonance (SPR) effect of spatially confined electrons in Ag nanoparticles and this red-shift might be related to the interaction between Ag and TiO 2 .Moreover, the adsorption peak range might be continuously broadened as the Ag dosage increases.

Catalytic Activity for Cr(VI) Photoreduction
The catalytic properties of Ag@TiO 2 photocatalysts were evaluated by Cr(VI) imitated heavy metal wastewater.As shown in Figure 4a, Cr(VI) was firstly adsorbed onto the surface of Ag@TiO 2 and adsorption equilibrium could be reached within 30 min under dark conditions.The aquatic Cr(VI) could be fully reduced to Cr(III) within 60 min under UV irradiation.The results illustrated that the Cr(VI) photocatalytic reduction efficiency got improved through Ag loading, which might be due to the following reasons.The photoresponse range could be enhanced, as presented in Figure 3b.Then, the Schottky junction would be formed at the interface with Ag deposition.Under light irradiation, photo-generated electrons (e − ) tend to migrate and accumulate on Ag nanoparticles which served as effective electron acceptors, and the extra acceptors subsequently promoted the e − /h + separation efficiency and ultimately enhanced the photocatalytic reduction of Cr(VI).It is noteworthy that the amount of metal deposition has a great influence on the photocatalytic ability of the obtained photocatalysts.However, the photocatalytic efficiency gradually decreases once the Ag loading exceeds 0.06 wt %, which is mainly due to the shielding effect by the excessive Ag coverage.In addition, when the Ag doping ratio gets higher (>0.06 wt %), the extra Ag sites tend to act as the e − /h + recombination centers which may lead to the recombination enhancement and result in photon quantum efficiency drop.The effect of different concentration of Ethanol on Cr(VI) reduction is shown in Figure S3. Figure 4b shows that the maximum UV-visible absorption of Cr(VI) solution was at 545 nm, which was close to the standard value of 540 nm.It can be obviously noticed that the Cr(VI) absorbance decreased rapidly in the first 20 min, but slowly later on.shown in Figure S3. Figure 4b shows that the maximum UV-visible absorption of Cr(VI) solution was at 545 nm, which was close to the standard value of 540 nm.It can be obviously noticed that the Cr(VI) absorbance decreased rapidly in the first 20 min, but slowly later on.The impact of the initial pH on the Cr(VI) reduction was investigated, and its degradation efficiency was gradually reduced with pH increasing from 3.0 to 11.0 (Figure 4c,d).Therefore, the initial pH plays a significant role in the whole Cr(VI) photocatalytic reduction processes, which not only affects the surface adsorption but also the chromium's redox potential [35].On the other hand, as the main catalytic reaction occurred on the interface, the adsorption ability to substrates would be a crucial role in the migration and captures of photogenerated carriers, which would directly affect the surface reactivity in the following photocatalytic reaction processes [36].First of all, the Cr(VI) adsorption procedure was the prerequisite step for photocatalytic reduction, and the influence of initial pH on adsorption as well as resultant removal rate are shown in Figure 4d.Cr(VI) interface adsorption rapidly increases with pH decreasing.The maximum adsorption rate of 25% (pH = 3) was The impact of the initial pH on the Cr(VI) reduction was investigated, and its degradation efficiency was gradually reduced with pH increasing from 3.0 to 11.0 (Figure 4c,d).Therefore, the initial pH plays a significant role in the whole Cr(VI) photocatalytic reduction processes, which not only affects the surface adsorption but also the chromium's redox potential [35].On the other hand, as the main catalytic reaction occurred on the interface, the adsorption ability to substrates would be a crucial role in the migration and captures of photogenerated carriers, which would directly affect the surface reactivity in the following photocatalytic reaction processes [36].First of all, the Cr(VI) adsorption procedure was the prerequisite step for photocatalytic reduction, and the influence of initial pH on adsorption as well as resultant removal rate are shown in Figure 4d.Cr(VI) interface adsorption rapidly increases with pH decreasing.The maximum adsorption rate of 25% (pH = 3) was five times that of the lowest one 1% (pH = 11).This phenomenon was mainly due to the electrostatic attraction that the negatively charged Cr 2 O 7 2− would spontaneously attach to positively charged subjects.According to the ζ potential testing in (Figure S4), the pH pzc (pH of zero-point charge) value of Ag@TiO 2 was 5-6.So, when initial pH <5, the surface would be positively charged which was in favor of the Cr 2 O 7 2− adsorption.On the contrary, its surface would become negatively charged when pH >5, which will result in a decrease in the adsorption capacity of Cr(VI).Therefore, it could be concluded that the adsorption capacity for Cr(VI) plays an important role in the photoreduction processes.
On the other hand, the metal pair's redox potential was either a significant factor that determines the photocatalytic ability of the resultant composites.As shown in Figure 4c, the Cr(VI) photoreduction rate gradually drops with pH increasing.Specifically, the equilibrium could be reached within 40 min under the condition of pH = 3, but no obvious reduction could be found at pH = 11, which indicated that the oxidation ability of Cr(VI) was strongly affected by the H + concentration.Moreover, as the initial pH got higher, the redox potential of Cr(VI) decreased which leads to the drop of its oxidation ability (See Supplementary Materials, Equations (S1)-( S3)).It could be concluded that the photoreduction of Cr(VI) should be easier under acidic conditions and relatively difficult under neutral or alkaline conditions could be drawn.And, it also indicated that the negative effect of the proton deficiency could be inhibited via providing enough protons to the reaction system.Moreover, it is noteworthy that Ag@TiO 2 exhibits a higher photocatalytic activity than pure TiO 2 at the same pH (Figure 4d, line), which may be due to the available free electrons on the Ag@TiO 2 surface or the active sites that can donate electrons to Cr(VI) being more than pure TiO 2 .Therefore, the initial pH impact on Cr(VI) reduction came from the integrative actions of various changes in the physicochemical properties of all the substances involved.The high efficiency of the obtained ultra-trace Ag doped TiO 2 photocatalyst under acidic, neutral and alkaline conditions suggests its promising potential for wastewaters purification and treatment.
The effect of the catalyst dosage on Cr(VI) reduction is shown in Figure 5a, the Cr(VI) reduction efficiency exhibited a positive dependence on the Ag@TiO 2 dosage.There was not any reduction that could be detected in the absence of catalyst addition and the Cr(VI) removal efficiency gradually increased with dosage increasing.When the dosage was increased to 2 g/L, Cr(VI) could be fully reduced to Cr(III) within 30 min, and the removal rate could still reach above 75% at a lower dosage (0.5 g/L).It was mainly due to that the increasing dosage providing more reaction sites for Cr(VI) adsorption and photoreduction, as well as to producing more photogenerated electrons.
efficiency exhibited a positive dependence on the Ag@TiO2 dosage.There was not any reduction that could be detected in the absence of catalyst addition and the Cr(VI) removal efficiency gradually increased with dosage increasing.When the dosage was increased to 2 g/L, Cr(VI) could be fully reduced to Cr(III) within 30 min, and the removal rate could still reach above 75% at a lower dosage (0.5 g/L).It was mainly due to that the increasing dosage providing more reaction sites for Cr(VI) adsorption and photoreduction, as well as to producing more photogenerated electrons.Figure 5b shows the Cr(VI) photocatalytic reduction under the UV-light intensity that varied from 0 W (blank) to 500 W. The removal rate for adsorption was maintained obviously at less than 10% until the equilibrium and the photoreduction was significantly improved under UV-light compared to the blank one.However, it should be noted that there are little differences between the 300 W and 500 W mercury lamps although the intense illumination might facilitate electron production.The Cr(VI) removal efficiency could reach about 95% within 50 min under the illumination intensities 300 W or 500 W. Therefore, it would be more advisable to utilize the 300 W lamp for considering the energy consuming.
Since industrial wastewater contains a variety of conventional materials which are usually comprised of routine anions and cations, these co-existing substances tend to reduce the catalytic performance of Ag@TiO 2 composites.To explore the potential synergistic effect between co-existing ions and Ag@TiO 2 composites, a series of co-existing ions, such as Cl − , SO 4 2− , NO 3 − and Na + , K + , Mg 2+ , Ca 2+ ions were utilized in the control group.As the amount of co-existing ions in practical industrial wastewater usually reaches thousands of ppm, in order to imitate the actual situation, its concentrations are all set at 0.1 mol/L, which in turn may affect the rate of photocatalytic reduction.As illustrated in Figure 5c, all co-existing ions have a significant impact on the Cr(VI) reduction process.Two sequences on the impact of co-existing ions were drawn through experiments, like Cl − > SO 4 2− > NO 3 − and Na + > K + > Ca 2+ > Mg 2+ .The main reason for the inhibiting effect by co-existing anions was the competition between anionic and dichromate adsorption and active sites on the surface.To be more specific, NO 3 − have very few effects whereas SO 4 2− or Cl − were more rapidly adsorbed by the catalyst than Cr(VI), which is similar to a previous report [37].Obviously, the Na + introduced together has a certain influence on the removal rate of the experiment (Figure 5d).As for cations, the impacts on photoreduction can be neglected since Na + , K + , Mg 2+ and Ca 2+ ions are already in their highest oxidation states, so they can neither capture radicals or photo-generated holes [38].Therefore, the inhibiting phenomenon by cations may be caused by the presence of Cl − .The Cr(VI) removal rates in Mg 2+ and Ca 2+ -containing systems are better than that in K + and Na + -containing systems, which may be due to the higher affinity of higher valence cations for Cr 2 O 7 2− , and enhances anion transfer rate in solution.For the same valence state, the larger the ion radius (K + > Na + or Ca 2+ > Mg 2+ ), the ion center contains more positive charge which promotes the anions' adsorption [39].
Additionally, the reusability of the Ag@TiO 2 photocatalyst was also evaluated.The performance of Ag@TiO 2 nanocomposite within five recycles is shown in Figure 6a.Obviously, the removal efficiency slowly decreases with the recycling continuing.It may be due to that the increasing amount of adsorbed Cr(III) on Ag@TiO 2 which will inhibit its surface adsorption and active site reduction properties.To further identify the adsorbed species on the catalyst surface, high resolution XPS spectra of the dry Ag@TiO 2 sample after the reaction were used to characterize the changes in the catalyst surface.The XPS spectra of Cr peaks were obviously presented in Figure 6b.The bands at binding energy of 577.2 eV and 585.9 eV were attributed to Cr(III) 2p 3/2 and Cr(III) 2p 1/2 , respectively.The peaks at 580.6 eV belonged to the Cr(VI) 2p 3/2 .The results show that the adsorption and reduction were simultaneously carried out during the Cr(VI) photoreduction and the generated Cr(III) was also adsorbed on the Ag@TiO 2 surface, which results in a low Cr(VI) reduction rate.Second, it was found that the XPS of Ag shifted to a certain extent after the reaction (Figure S5), indicating that more Ag 2 O was produced.It is possible that the ability to transfer electrons might decreased, which will lead to a decline in catalytic performance.However, after five consecutive cycles, Ag@TiO 2 removal rate could still reach 62%, indicating that the resultant photocatalyst has a relatively good reusability.
It is well known that the selective adsorption of Cr(VI) and the desorption of Cr(III) at the surface of photocatalysts are very important factors in determining the photocatalytic reduction.The modification of Ti 3+ and hydroxyl groups can enhance the positive charge on the surface of the catalyst and further increase the preferential adsorption of Cr(VI) and repulsion of Cr(III) by electrostatic action [40].Moreover, it might be a simpler way to make the catalyst aspects of active-site reusable by external force, like: thermal, chemical reactivation, calcine, and Ultrasonic cleaning et al.

Kinetic Modeling Study
The reaction between Ag@TiO2 photocatalysts and Cr(VI) was related to the multiphase surface reaction.Langumir-Hinshelwood first order kinetic model was applied to evaluate the reaction kinetics.
where r0 is the initial photocatalytic reduction efficiency (mg/L•min); t is the reaction time (min); k is the reaction efficiency constant (min −1 )); K is the reaction equilibrium constant; and Cs is the reactant concentration (mg/L); k and K are determined by many factors in the reaction system, including the amount of catalyst, light intensity, initial concentration of reactants, reaction temperature, physical properties of the reactants, gas phase oxygen concentration, etc.

Kinetic Modeling Study
The reaction between Ag@TiO 2 photocatalysts and Cr(VI) was related to the multiphase surface reaction.Langumir-Hinshelwood first order kinetic model was applied to evaluate the reaction kinetics.
where r 0 is the initial photocatalytic reduction efficiency (mg/L•min); t is the reaction time (min); k is the reaction efficiency constant (min −1 )); K is the reaction equilibrium constant; and C s is the reactant concentration (mg/L); k and K are determined by many factors in the reaction system, including the amount of catalyst, light intensity, initial concentration of reactants, reaction temperature, physical properties of the reactants, gas phase oxygen concentration, etc.When the reaction substrate concentration is rather low, KC 1, the equation can be simplified to the apparent first order equation: where k ap is the apparent rate constant of a pseudo first order reaction.In the beginning of the reaction, t = 0, C t = C 0 , the equation can be obtained after deformation: where C t is the concentration of Cr(VI) in the solution at t min; C 0 is the initial Cr(VI) concentration; b represents a constant; k ap is the apparent rate constant (min −1 ).
From Figure 7, it can be seen that ln(C t /C 0 ) − t is in a good linear relationship and its performance fits the first order reaction.The apparent reaction kinetic constant can be applied to estimate the total reaction rate, and then compare the photocatalytic efficiency under different conditions.In Figure 7, initial pH and Ag loading affected the obtained reaction kinetic constant.The first-order kinetic equation, reaction rate constant, reaction rate and the correlation coefficients (R 2 ) of the photocatalytic reaction under different conditions are shown in Table 2.It can be seen that the apparent rate constant k ap decreases as initial pH grows, and k ap under acidic condition (pH = 3) was about six times as much as the neutral one (pH = 7) and quickly dropped to about zero under alkaline conditions (pH = 11), indicating that acidic environment promotes the photocatalytic reduction processes.In addition, k ap presents a trend that gradually rises to 0.0534 min −1 until the silver doping ratio reaches 0.06 wt % and then continuously decreases to 0.0211 min −1 as the Ag loading grows to 3.0 wt %.Furthermore, the results show that Cr(VI) reduction efficiency by 0.06 wt % Ag loading was 2.5 times compared to pure TiO 2 , indicating that suitable Ag loading on TiO 2 surface promoted its photocatalytic ability significantly.Table 2.The relevant parameters obtained from kinetic study.ln(C t /C 0 ) = 0.0234t − 0.1560 0.0234 0.2340 0.9808 NO 3 − ln(C t /C 0 ) = 0.0320t − 0.1758 0.0320 0.3202 0.9843 Cl − ln(C t /C 0 ) = 0.0190t − 0.0600 0.0190 0.1900 0.9948

Mechanism of Photoreduction
To identify the radical species involved in the reduction of Cr(VI) by Ag@TiO 2 system, EPR experiments using DMPO as the spin-trapping agent were carried out.EPR analysis results of TiO 2 and Ag@TiO 2 solids are shown in Figure 8a; it can be clearly seen that one signal appeared after UV irradiation both for bare TiO 2 and Ag@TiO 2 ; the g-factors were 2.0024 and 2.0019, respectively, which was considered to be a free electron in the conduction band of TiO 2 or an oxyanion vacancy.The enhancement of the peak signal intensity proves that more free electrons and oxygen vacancies are generated in the Ag@TiO 2 system, which may be due to the photogenerated electrons being transferred to silver elements and stored therein.
Figure 8c,d shows the EPR spectra of DMPO-trapped O 2 •− in methanol medium and DMPO-trapped • OH in aqueous medium which have been investigated in the active radical species for the TiO 2 and Ag@TiO 2 .It be able to evidently noted that four characteristic peaks (1:2:2:1) of DMPO-• OH adducts and six characteristic peaks of the DMPO-O 2 •− adducts were observed on TiO 2 or Ag@TiO 2 nanocomposite systems under UV light irradiation, but no signal was detected in the dark.It could be noted that, the intensity signals of the hydroxyl radical and superoxide radical species formed in Ag@TiO 2 were stronger than those in TiO 2 under UV irradiation, which represent the sample of Ag@TiO 2 effectively promoted electron/hole pair separation and prolonged the lifetime.This, in turn, explains why there is a higher photocatalytic activity of Ag@TiO 2 than TiO 2 toward the contaminants removal.
Ag@TiO2 were stronger than those in TiO2 under UV irradiation, which represent the sample of Ag@TiO2 effectively promoted electron/hole pair separation and prolonged the lifetime.This, in turn, explains why there is a higher photocatalytic activity of Ag@TiO2 than TiO2 toward the contaminants removal.To further identify the elemental constitution and its valences, the resultant Ag@TiO2 photocatalyst was characterized via XPS spectra.The full spectrum of TiO2 loaded with Ag nanoparticles was illustrated in Figures 9a and S6, the main elements were Ti, O, C and Ag.The C emergence is due to the introduction of exogenous substances during the testing processes.As presented in Figure 9b, it can be seen that the loading of silver nanoparticles has obvious effects on the binding of Ti 2p orbitals.Compared with pure TiO2, the combination of Ti 2p1/2 and Ti 2p3/2 can change from to 465.13 eV and 459.45 eV to 458.77 eV and 464.57eV, respectively.The affection was due to that the load of Ag makes the heterojunction between Ag and TiO2, and the plasmonic effect produced by Ag makes the electron field which would influence TiO2. Figure 9c shows the XPS peak of O 1s, from which we can find the asymmetry peak of O. Therefore, two peaks were obtained by Gauss Lorentz method.The peaks at 531.87 eV and 530.79 eV were assigned to Ti-OH bond and the Ti-O-Ti bond in lattice oxygen, respectively.The binding energy at 533.57eV was due to the adsorbed oxygen on the catalyst surface.The X-ray spectrum of Ag 3d is shown in Figure 9d, indicating that the XPS peak of Ag 3d mainly contains two characteristic peaks.The binging energies of Ag 3d3/2 and Ag 3d5/2 were observed at 368.2 eV and 374.2 eV, respectively.The results showed that the difference of binding energies between Ag 3d5/2 and Ag 3d3/2 was 6.1 eV, revealing that the Ag element depositing on TiO2 surface exists in the form of zero valences, which was reported in the previous literature To further identify the elemental constitution and its valences, the resultant Ag@TiO 2 photocatalyst was characterized via XPS spectra.The full spectrum of TiO 2 loaded with Ag nanoparticles was illustrated in Figure 9a and Figure S6, the main elements were Ti, O, C and Ag.The C emergence is due to the introduction of exogenous substances during the testing processes.As presented in Figure 9b, it can be seen that the loading of silver nanoparticles has obvious effects on the binding of Ti 2p orbitals.Compared with pure TiO 2 , the combination of Ti 2p 1/2 and Ti 2p 3/2 can change from to 465.13 eV and 459.45 eV to 458.77 eV and 464.57eV, respectively.The affection was due to that the load of Ag makes the heterojunction between Ag and TiO 2 , and the plasmonic effect produced by Ag makes the electron field which would influence TiO 2 .Figure 9c shows the XPS peak of O 1s, from which we can find the asymmetry peak of O. Therefore, two peaks were obtained by Gauss Lorentz method.The peaks at 531.87 eV and 530.79 eV were assigned to Ti-OH bond and the Ti-O-Ti bond in lattice oxygen, respectively.The binding energy at 533.57eV was due to the adsorbed oxygen on the catalyst surface.The X-ray spectrum of Ag 3d is shown in Figure 9d, indicating that the XPS peak of Ag 3d mainly contains two characteristic peaks.The binging energies of Ag 3d 3/2 and Ag 3d 5/2 were observed at 368.2 eV and 374.2 eV, respectively.The results showed that the difference of binding energies between Ag 3d 5/2 and Ag 3d 3/2 was 6.1 eV, revealing that the Ag element depositing on TiO 2 surface exists in the form of zero valences, which was reported in the previous literature [27,41].To further elucidate the surface metal composition, Ag 3d peaks of the catalysts were deconvoluted, and the content of Ag was calculated to be 0.6655% (Table S3).Combined with the XRD and HRTEM results, the existence of Ag and TiO 2 nanoparticles in the resultant compound can be further proved.
Catalysts 2018, 8, x FOR PEER REVIEW 11 of 16 [27,41].To further elucidate the surface metal composition, Ag 3d peaks of the catalysts were deconvoluted, and the content of Ag was calculated to be 0.6655% (Table S3).Combined with the XRD and HRTEM results, the existence of Ag and TiO2 nanoparticles in the resultant compound can be further proved.The photocatalytic reaction was considered to be a heterogeneous interfacial reaction, and mainly contained two parts of the reaction mechanism (Figure 10).Firstly, the good adsorption of Cr(VI) on the surface of Ag@TiO2 was a prerequisite of photoreduction.Then, the Cr(VI) was reduced to Cr(III) by using photogenerated electrons from the light source.Adsorption of oxygen on the catalyst surface as an electron competitor was generally considered to have the potential to reduce the rate in the photoreduction reaction.Oxygen reduction state (O2 •− ) can promote the photoreduction of Cr(VI) in TiO2 system as electron transfer intermediates [42].From the results of photoreduction experiments and the EPR spectra of DMPO-trapped O2 •− , it can be seen that a substantial increase of O2 •− did not inhibit the reduction activity of photocatalysis, which may indicate the O2 •− as a reduced product formed by obtaining electrons and also has strong reducibility, resulting in the transformation of Cr(VI), a strongly oxidizing species, to Cr(III).Therefore, the photoreduction of Cr(VI) by Ag@TiO2 system under UV light is a complex synergistic processes with multi-actions.However, more photogenerated electrons were probably exposed, and resulted in a higher photocatalytic activity due to the loading of Ag on TiO2.The photocatalytic reaction was considered to be a heterogeneous interfacial reaction, and mainly contained two parts of the reaction mechanism (Figure 10).Firstly, the good adsorption of Cr(VI) on the surface of Ag@TiO 2 was a prerequisite of photoreduction.Then, the Cr(VI) was reduced to Cr(III) by using photogenerated electrons from the light source.Adsorption of oxygen on the catalyst surface as an electron competitor was generally considered to have the potential to reduce the rate in the photoreduction reaction.Oxygen reduction state (O 2 •− ) can promote the photoreduction of Cr(VI) in TiO 2 system as electron transfer intermediates [42].From the results of photoreduction experiments and the EPR spectra of DMPO-trapped O 2 •− , it can be seen that a substantial increase of O 2 •− did not inhibit the reduction activity of photocatalysis, which may indicate the O 2 •− as a reduced product formed by obtaining electrons and also has strong reducibility, resulting in the transformation of Cr(VI), a strongly oxidizing species, to Cr(III).Therefore, the photoreduction of Cr(VI) by Ag@TiO 2 system under UV light is a complex synergistic processes with multi-actions.However, more photogenerated electrons were probably exposed, and resulted in a higher photocatalytic activity due to the loading of Ag on TiO 2 .

Preparation of Ag@TiO2 Photocatalyst via Ultrasonic Impregnation Assisted Photoreduction Deposition
Silver was deposited onto the surface of TiO2 formed via a "green" ultrasonic impregnation assisted photoreduction strategy.Firstly, a suspension of titanium dioxide nanoparticles (P25, 10 g/L) with absolute ethanol solution (different concentrations of ethanol solution ca.0%, 25%, 50%, 75% and 100%, as an electron donor) and a precious metal precursor solution (AgNO3 0.074 mM) were stirred and ultrasound 30 min under dark conditions.Then, the suspension was transferred to a quartz reactor under UV irradiation (300 W mercury lamp) for another 30 min.Finally, the obtained Ag@TiO2 nanocomposites were filtered and washed with a large amount of ultrapure water to remove unreacted silver nitrate, and freeze-dried for further experiments.A typical silver loading was evaluated to be ca.0.0648 wt % (Ag/TiO2) by using ICP-AES (Table 1).

Batch Experiments of Photoreduction
The prepared photocatalyst was suspended in a quartz reactor containing 50 mL (10 mg/L) target contamination (Cr(VI)).The suspension was dispersed for about 2 min by ultra-sonication before irradiation, and then stirred in the dark for 30 min to allow equilibrium adsorption of the heavy metal.

Preparation of Ag@TiO 2 Photocatalyst via Ultrasonic Impregnation Assisted Photoreduction Deposition
Silver was deposited onto the surface of TiO 2 formed via a "green" ultrasonic impregnation assisted photoreduction strategy.Firstly, a suspension of titanium dioxide nanoparticles (P25, 10 g/L) with absolute ethanol solution (different concentrations of ethanol solution ca.0%, 25%, 50%, 75% and 100%, as an electron donor) and a precious metal precursor solution (AgNO 3 0.074 mM) were stirred and ultrasound 30 min under dark conditions.Then, the suspension was transferred to a quartz reactor under UV irradiation (300 W mercury lamp) for another 30 min.Finally, the obtained Ag@TiO 2 nanocomposites were filtered and washed with a large amount of ultrapure water to remove unreacted silver nitrate, and freeze-dried for further experiments.A typical silver loading was evaluated to be ca.0.0648 wt % (Ag/TiO 2 ) by using ICP-AES (Table 1). the broad-scale preparation of precious metal matrix composites, and might be used as a promising photocatalyst for wastewater treatment.

Supplementary Materials:
The following are available online at http://www.mdpi.com/2073-4344/8/6/251/s1. Table S1 BET data for different silver loadings.Table S2 BET data for different solvent concentration.Table S3 XPS date of atomic (Ag@TiO 2 , 0.6 wt %). Figure S1 The Energy Dispersive Spectrdmeter of Ag@TiO 2 catalyst.Figure S2 XRD patterns of the Ag@TiO 2 nanocomposites, in which A is anatase phase and R is rutile phase.

Figure 1 .
Figure 1.Schematic illustration of (a) synthesis of ultra-trace silver doped TiO 2 composites via ultrasonic impregnation assisted deposition through an in situ UV-assisted photoreduction strategy; (b) photoreduction of highly toxic Cr(VI) to nontoxic Cr(III) by resultant Ag@TiO 2 photocatalysts under UV irradiation.

Figure 3 .
Figure 3. (a) XRD patterns of the Ag@TiO2 photocatalysts, in which A is anatase phase and R is rutile phase (no obvious Ag characteristic peak was found due to the ultra-trace loaded amount); (b) UVvis diffuse spectra of Ag@TiO2 in different doping ratio (dried sample images with different Ag loadings, insert).

Figure 3 .
Figure 3. (a) XRD patterns of the Ag@TiO 2 photocatalysts, in which A is anatase phase and R is rutile phase (no obvious Ag characteristic peak was found due to the ultra-trace loaded amount); (b) UV-vis diffuse spectra of Ag@TiO 2 in different doping ratio (dried sample images with different Ag loadings, insert).

Figure 4 .
Figure 4. (a) Photocatalytic reduction efficiency by pure TiO2 and Ag@TiO2 composites with various Ag loading ratio (effect of different loading on adsorption balance before irradiation, insert); (b) UVvis full wave band scanning of Cr(VI) under different irradiation times (the color change of Cr(VI) simples with different reduction time, insert); (c) Effect of different initial pH on Cr(VI) reduction rate; (d) The contribution of adsorption and photoreduction in Cr(VI) removal with different initial pH value using bare TiO2 and Ag@TiO2.(Both (b,c,d) using the simple with 0.06 wt % Ag addition.The experimental conditions were [Cat] = 1 g/L, [Cr(VI)]0 = 10 mg/L, 300 W and 60 min irradiation.

Figure 4 .
Figure 4. (a) Photocatalytic reduction efficiency by pure TiO 2 and Ag@TiO 2 composites with various Ag loading ratio (effect of different loading on adsorption balance before irradiation, insert); (b) UV-vis full wave band scanning of Cr(VI) under different irradiation times (the color change of Cr(VI) simples with different reduction time, insert); (c) Effect of different initial pH on Cr(VI) reduction rate; (d) The contribution of adsorption and photoreduction in Cr(VI) removal with different initial pH value using bare TiO 2 and Ag@TiO 2 .(Both (b,c,d) using the simple with 0.06 wt % Ag addition.The experimental conditions were [Cat] = 1 g/L, [Cr(VI)] 0 = 10 mg/L, 300 W and 60 min irradiation.

Figure 6 .
Figure 6.(a) The reusability of synthetic catalysts by 5 recycle on Cr(VI) removal (using the Ag@TiO 2 with 0.06 wt %); (b) High resolution XPS of Cr on the surface of Ag@TiO 2 catalyst.The experimental conditions were [Cr(VI)] 0 = 10 mg/L).

Figure 7 .
Figure 7.The study of kinetics at different Ag loadings (a) and pH values (b).

Figure 7 .
Figure 7.The study of kinetics at different Ag loadings (a) and pH values (b).

Figure S3
Effect of different concentration of empty scavenger (Ethanol) on Cr reduction.Figure S4 Zeta potential of bare TiO 2 and Ag@TiO 2 catalyst.Figure S5 High resolution XPS of Ag 3d of Ag@TiO 2 catalyst before (a) and after (b) reaction.Figure S6 The survey spectrum XPS spectra of the Ag@TiO 2 composites.Author Contributions: S.P., R.Z., and W.C. initiated the concept; S.P. and Y.H. designed and conceived of the experiments; Y.H. and H.C. performed the experiments; S.P., Y.H. and Z.Y. analyzed the data; H.C., R.Z., and D.D. performed the characterizations, interpreted the data; S.P. contributed reagents/materials/analysis tools; S.P., Y.H. and S.X.drafted the manuscript, W.C. and Z.D. corrected the manuscript.All authors read and approved the final manuscript.Funding: This research received no external funding.

Table 2 .
The relevant parameters obtained from kinetic study.