WO3 in suit embed into MIL-101 for enhancement charge carrier separation of photocatalyst

Compositing nanoparticles photo-catalyst with enormous surface areas metal–organic framework (MOF) will greatly improve photocatalytic performances. Herein, WO3 nanoparticles are partly embedded into pores of MIL-101 or only supported on the outside of representative MIL-101, which were defined as embedded structure WO3@MIL-101@WO3 and coating structure WO3&MIL-101 respectively. Different pH, concentration and loading percentage were researched. XRD, TEM and BET were carried to analyze the composites. Compared with the pristine WO3, all WO3 loaded MOF nanocomposites exhibited remarkable enhancing for the efficiency of photocatalytic degradation methylene blue under visible light. Their activity of the same loading percentage WO3 in embedded structure and coating structure have increased for 9 and 3 times respectively compared with pure WO3. The WO3@MIL-101@WO3 has 3 times higher efficiency than WO3&MIL-101, because the shorter electron-transport distance can make a contribution to electron–hole separation. The further mechanism involved has been investigated by radical quantify experiment, XPS and photoluminescence spectroscopy.

synthesis of Wo 3 @MIL-101@WO 3 . Different sodium tungstate quality, pH and concentration were investigated, the different condition were listed as Table S1. Typically, 320 μL hydrochloric acid which was diluted to 5 mL with water, then 10 mg sodium tungstate was dissolved in 40 mL water was added. After adding 100 μL hydrogen peroxide, the mixture was stirred for 30 min. Then 100 mg MIL-101 was added to the solution, stirring for 24 h under room temperature. Then the mixture was stirring and heated with oil bath by stepwise warming method. After temperature raised to 30 °C, it was kept at 30 °C for 1 h. Then raise the temperature to 45 °C and kept it for 1 h. After that it was raised to 60 °C and kept for 3 h, the temperature was raised to 120 °C until the water was dried up. The substance was washed with water and ethanol several times. synthesis of Wo 3 &MIL-101. In order to deposit WO 3 completely outside of MIL-101, direct settlement method is used and with MIL-101 not being degassed. With the same steps to prepare WO 3 precursor, then 100 mg not degassed MIL-101 was added to the solution, stirring for 30 min, the temperature was raised to 120 °C for the water drying up. The substance was washed with water and ethanol several times.
Characterization. The crystal structure of the prepared samples were characterized by a Bruker D8 Advance X-ray diffractometer with Ni-filtered Cu KαIrradiation (λ = 0.15406 nm) under 40 kV and 40 mA. XPS diffraction patterns were carried out by an AXIS-His spectrometer (Kratos Corporation) with a Mg Kα X-ray source, and the spectra were adjusted to the C 1 s peak at 284.8 eV. The shape and size of the nanocomposites were characterized by a JEOL JEM-6700F field emission scanning electron microscope with an accelerating voltage of 20 kV, respectively. TEM and HRTEM images were obtained under a JEM-2100 transmission electron microscope with an accelerating voltage of 20 kV. The surface area and the pore size distribution were measured on Quantachrome Autosorb-IQ sorption system at 77 K. Optical absorption properties (DRS) were detected under a Shimadzu UV-3600 spectrometer with a reference of BaSO 4 . The photoluminescence (PL) emission spectra of samples were observed on a Hitachi F-4500 luminescence spectrometer.
photocatalytic Activity test. The photocatalytic degradation performance of Methylene Blue (MB) test was carried out under visible light irradiation 44 . A xenon lamp (300 W) with visual light filter was dispersed in an aqueous solution (50 mL) containing 30 mg/L MB dye by ultrasonic treatment for 5 min and maintained stirring for 30 min. Then, the solution was transferred to a quartz reaction vessel and agitated for some time. A liquid (5 mL) was sampled at scheduled irradiation time and the suspended catalyst were eliminated by centrifugation under 8000 rpm for 5 min. The UV-Visible absorption spectrum of the solution was carried out with a UV-Visible absorption spectrum of the solution was carried out with a UV-Vis spectrophotometer (UV-3600). The percentage of degradation was defined as −ln (C/C 0 ), herein, C 0 refers the absorption (λ max = 664 nm) of MB solution prior irradiation and C indicates the absorption of MB solution at each irradiated time interval.
Active Species Trapping and Superoxide Radical Quantification Experiments. For detecting the active species during photocatalytic reactivity 45 , hydroxyl radicals (·OH), the superoxide radical (O 2 − ·), and holes (h + ) were trapped by adding 2.0 mM (according to the reaction system) IPA 46 (a quencher of ·OH), AgNO 3 47 (a quencher of O 2 − ·), and TEOA 48 a quencher of h + respectively. The method was similar to the former photocatalytic activity test 45 . TA (5 × 10 −4 M in a 2 × 10 −3 M NaOH solution), which reacts readily with ·OH generating from WO 3 &MIL-101 and WO 3 @MIL-101@WO 3 . The production of ·OH was quantitatively analyzed by detecting the concentration of 2-hydroxyterephthalic acid (fluorescence peak at about 425 nm by excitation with the wavelength of 315 nm) with Shi-madzu spectro fluorophotometer (RF-5301 pc) after centrifugation 49 . The method was similar to the former photocatalytic activity test, with TA replacing the MB 48 .

Results and Discussion
Reaction process Illustration. As shown in Fig. 1, the WO 3 @MIL-101@WO 3 hetero-structure were synthesized by low temperature H 2 O 2 assistant sol-gel method. The process of adding H 2 O 2 was very important for the formation of peroxo-tungstate gel precursor, obvious Tyndall effect can be observed. Stirring for 24 h giving enough time for the slow kinetic reaction process of MIL-101 dipping into peroxo tungstate gel. The loading percentage, pH and concentration influencing the properties of precursor were also researched listed in Table S1. The resultant WO 3 &MIL-101 and WO 3 @MIL-101@WO 3 samples have been well characterized by various techniques. The actual loading percentages tungsten (5%-15%) at various precursor concentrations determined by atomic absorption spectrum (AAS) method matches well with the theoretical loading (5.27-15.5%), as shown in Table S2, indicating that the H 2 O 2 assistant sol-gel method is effective in loading WO 3 into MIL-101. structure, composition, and microstructure. As shown in Fig. 2a, the crystal structure of MIL-101 is in good agreement with the literature reported 41 , demonstrating the formation of MIL-101 with ultrapure and good crystallinity. After loading WO 3 , the characteristic XRD peaks of MIL-101 in all samples are maintained, demonstrating the treatment did not have the damage on the crystal structure of MIL-101. The weaker peaks of WO 3 @ MIL-101@WO 3 than WO 3 &MIL-101 should be due to a part of WO 3 have been embedded into pores of MIL-101 which resulted in small particle size. The patterns of different loading percentage were shown in Fig. 2b, as the loading percentage increased the intensity of the characteristic peaks increases. Different pH and concentration also show a significant influence on the intensity of MIL-101 shown in Fig. S1a,b.
The survey pattern of WO 3 , WO 3 @MIL-101@WO 3 , and MIL-101 was shown in Fig. 3a.The high resolution XPS of O element (Fig. 3b) show the position 530.6 eV for WO 3 @MIL-101@WO 3 which has right shift 0.21 eV compared with WO 3 , the absorbed oxygen at 532.67 eV was disappeared, indicating the O element has some changes in WO 3 @MIL-101@WO 3 compared with pure WO 3 . Figure 3c,d show the binding energy shift and half band width changes, W 4f2/7 and W 4f2/5 have right shift 0.12 eV and 0.13 eV, respectively. The half band width of WO 3 @MIL-101@WO 3 has been widen 0.44 eV compared with WO 3 , these results all show that W element has a good attachment with the linkages of MIL-101, there are interactions between WO 3 and MIL-101.  www.nature.com/scientificreports www.nature.com/scientificreports/ In order to further detect the relation of WO 3 and MIL-101, DLS method was adapted. In Table S3, The zeta potential of WO 3 , WO 3 &MIL-101,WO 3 @MIL-101@WO 3 and MIL-101 are −34.1 mV, −8.69 mV, −6.04 mV and 34.5 mV respectively. The result for WO 3 &MIL-101 (−8.69 mV) and WO 3 @MIL-101@WO 3 (−6.04 mV) shows electrostatic attraction between MIL-101 and oppositely charged WO 3 . A close interaction of the WO 3 &MIL-101 and WO 3 @MIL-101@WO 3 composite can be achieved with the electrostatic attraction. The more negative potential of WO 3 &MIL-101 than WO 3 @MIL-101@WO 3 indicates there exist more WO 3 nanoparticles on the surface of MIL-101.
As shown in Fig. 4a, octahedral structure with smooth surface of MIL-101 have the size of 400 nm-600 nm 27 . As shown in Fig. 4b, the size of pure WO 3 nanosheets are about 50 nm in thickness and 400 nm in width. After loading WO 3 , the surface of MIL-101 became rough and coating a slice WO 3 on the surface of MIL-101 for WO 3 @ MIL-101@WO 3 . For WO 3 &MIL-101, WO 3 particles growth and there are intensity aggregating together with each other. The average particle size is about 40 nm shown as Fig. S2d. For the exploration experiment, we investigated the WO 3 @MIL-101@WO 3 samples with different WO 3 loading proportion (5-15%), different pH, different concentration and not adding hydrogen peroxide and the SEM results are displayed in Figs S2 and S3, and S4. As can be seen from Fig. S2, WO 3 were thinly well coating on MIL-101 much like the morphology of WO 3 @MIL-101@WO 3 , as the loading proportion increasing, the amount of WO 3 slice increase. Figure S2c show no hydrogen peroxide added in the solution, WO 3 nanoparticles have grown much larger, indicating that the adding hydrogen peroxide can change the state of peroxo tungstate precursor gel, which is very important for WO 3 embed into the pores of MIL-101. As can be seen Figs S3 and S4 in all WO 3 @MIL-101@WO 3 samples, WO 3 are all well slice coating outside MIL-101, with pH and concentration changes, the state of WO 3 have some difference.
In order to study the morphology of MIL-101 (Fig. 5a,c insert), TEM and HRTEM measurements were carried 27 . The images of WO 3 @MIL-101@WO 3 (111) and (002) planes of WO 3 . These results suggest that an intimate contact which will be helpful for the charge separation and transferring between WO 3 and MIL-101. The energy dispersive X-ray spectroscopy (EDS) mapping (Fig. S5) was conducted to further confirmed the component and www.nature.com/scientificreports www.nature.com/scientificreports/ structure of WO 3 @MIL-101@WO 3 , the crystal structure of MIL-101 can be displayed by the uniform Cr elements in the background 27 .
To clarify the location of WO 3 relative to MIL-101, the N 2 adsorption measurement of MIL-101 was carried which shows type I property with secondary uptakes at p/p 0 of around 0.1 and 0.2 (Fig. 6a), which is a typical MIL-101 adsorption curve 50,51 . After loading WO 3 nanoparticles, the adsorption-desorption isotherm of WO 3 &MIL-101 sample has little changes but the adsorption-desorption isotherm of WO 3 @MIL-101@WO 3 has a significant change compared with MIL-101 and WO 3 &MIL-101 and N 2 adsorption decreased with the WO 3 content increased. The surface area of MIL-101 was measured to be 2480 m 2 /g and the total pore size value was estimated to be 1.193 cm 3 /g at a relative pressure of 0.99 (shown in Table 1), both are similar to the numbers reported in the literature. The surface area of WO 3 &MIL-101 was 2350 m 2 /g and pore size value was 1.153 cm 3 /g, both are close to MIL-101. The surface area of WO 3 @MIL-101@WO 3 samples gradually decreased from 1668 to 1255 m 2 /g, the pore size value change from 0.835 to 0.651 cm 3 /g, as WO 3 content increased from 5% to 15%. The pore size distribution is shown in Fig. S7, the pore size at 1.8, 2.6 and 3.2 nm are attributed to pure MIL-101 which has been reported in literature 50 , compared with pure MIL-101, the pores size distribution of the WO 3 &MIL-101 have little changes while the pore size of WO 3 @MIL-101@WO 3 have a significantly decrease than both pure MIL-101 and WO 3 &MIL-101. With the WO 3 content increased, the pore size shows a decreasing trend. All these phenomena could be caused by WO 3 embed into the pores of MIL-101. These results suggest that the WO 3 nanoparticles in WO 3 @MIL-101@WO 3 samples are successfully embedded in the cavities of MIL-101. WO 3 nanoparticle has little influence for surface area and pore size of MIL-101, so WO 3 nanoparticles are possibly on the surface of MIL-101 in WO 3 &MIL-101 sample.
Combined with the TEM and SEM results, it can be concluded the WO 3 particles just coat on the surface of MIL-101 in WO 3 &MIL-101 sample. The WO 3 particles partly embed into the pores and partly on the surface of MIL-101 in WO 3 @MIL-101@WO 3.

UV-Vis DRs analysis.
UV-Vis DRS spectra were used to analyze the optical properties of the MIL-101 and the different WO 3 loading proportion samples (Fig. 7). MIL-101 exhibits two characteristic absorption band centered at 450 nm and 600 nm, which coincides with that in literature 27 . The band of pure MIL-101 in the UV region belongs to π-π* transitions of ligands and the bands in the visible region can be assigned to the D-D spin-allowed transition of the Cr 3+ . WO 3 displays a sharp fundamental absorption edge rise at 475 nm as expected, corresponding to a band gap of 2.75 eV 7 . Compared with that of MIL-101, the WO 3 @MIL-101@WO 3 shows an enhanced board absorption in visible light region, this may correspond to the visible light enhanced of WO 3 . Compared with that of pure WO 3 , WO 3 @MIL-101@WO 3 composite shows an adsorption band centered at 600 nm, which is attributed to the absorption of MIL-101 matrix.
photocatalytic Degradation of organic pollutant. The MIL-101 loading WO 3 samples were evaluated for photocatalytic MB degradation. As shown in Fig. 8a, comparing with pure WO 3 , the embedded structure www.nature.com/scientificreports www.nature.com/scientificreports/ WO 3 @MIL-101@WO 3 and coating structure WO 3 &MIL-101 increased 9 times and 3 times, respectively, due to the closely contact between WO 3 and MIL-101 which can be concluded from XPS and DLS data. Also, the embedded structure has 3 times higher efficiency than coating structure due to the part of WO 3 have embeded into the pores of MIL-101 resulting in the shorter distance of the electrons transfer from WO 3 to MIL-101 comparing with coating structure 52 . Figure 8b show the degradation efficiency of different loading percentages and a series of control experiment, which shown 10% WO 3 loading sample has the best photocatalytic efficiency. From the control experiment, it can be seen that pure MIL-101 has no photocatalytic efficiency, MB cannot be degraded by self-sensitization and the light are the necessary condition during photocatalytic. Different pH and concentration  www.nature.com/scientificreports www.nature.com/scientificreports/ were also investigated in Fig. S8a,b which shown that pH and concentration influence the photocatalytic efficiency of MB degradation. The WO 3 @MIL-101@WO 3 has the best photocatalytic efficiency when the amount of HCl was 280 μL and the volume of water was 35 mL.
Mechanism Investigation on photocatalytic performance Improvement. To further unveil the higher photocatalytic efficiency of WO 3 @MIL-101@WO 3 than WO 3 &MIL-101, photocurrent measurement have been carried and the results show that the photocurrent for both MIL-101 supported WO 3 get enhanced as compared to the pristine WO 3 (Fig. 9a), revealing that the formation of WO 3 -MOF schottky junction helps to separate the photo-generated electron-hole pairs 52 . The WO 3 @MIL-101@WO 3 displays much stronger photocurrent response than WO 3 &MIL-101, suggesting the much higher efficiency of the charge transfer 52 . This result is also supported by the photoluminescence (PL) emission spectroscopy, which provides useful hints for the photo-excited charge transfer and recombination. The PL intensity is slightly weakened when the WO 3 only coating outside MOF, while get greatly suppressed when some WO 3    www.nature.com/scientificreports www.nature.com/scientificreports/ the MOF (Fig. 9b). These observations indicate that the irradiative electron-hole recombination is more effectively suppressed by extracting the electrons from internal WO 3 than coating WO 3 53 . Such distinctly different photoelectron-chemical properties in WO 3 @MIL-101@WO 3 and WO 3 &MIL-101 unambiguously demonstrate that the part of WO 3 in the pores of MIL-101 contribute mostly of the photocatalytic efficiency of MB degradation. For comparison, we also investigated the photoluminescence (PL) emission spectroscopy of different WO 3 loading percentage (Fig. S9), the PL intensity are corresponding with the photocatalytic efficiency. The WO 3 loaded MOF samples all have lower intensity than pure WO 3 , pure MIL-101 have a lower photoluminescence emission. It indicated that WO 3 -MOF schottky junction can well suppressed the electron-hole pairs recombination and pure MIL-101 cannot be excited by visible light. In order to further investigate the migration and interface transfer or recombination rates of charge carriers electrochemical impedance spectra (EIS) was detected in Fig. S10. It was found that the WO 3 @MIL-101@WO 3 and WO 3 &MIL-101 composite exhibits much smaller arc sizes than the pure WO 3 under visible light irradiation. It demonstrates that the heterojunction composite has faster electron transfer through an intimated interface between MIL-101 and WO 3 as compared to the pristine WO 3 , which is in good agreement with the photocatalytic performance.
The photocatalytic mechanism of WO 3 @MIL-101@WO 3 have been researched by active species trapping and ·OH quantify experiment during the photocatalytic process 48 . In order to study the active species of the photocatalytic reaction of WO 3 @MIL-101@WO 3 , the trapping experiment was investigated and showed in Fig. 10a. It can be concluded that the addition of AgNO 3 (a quencher of e-, which can hinder the formation of O 2− ·) have no influence on photocatalytic degradation of MB 48 . On the contrary, the addition of IPA (a quencher of ·OH) or TEOA (a quencher of h + ) have an obvious influence of decrease on the photocatalytic degradation of MB. Therefore, the conclusion can be drawn that photo-generated holes (h + ) and ·OH are the main effective species on MB degradation for WO 3 @MIL-101@WO 3 under visible light irradiation. The result consistent with the kind of effective species of pure WO 3 and WO 3 &MIL-101. It can be concluded that the kind of active species have not changed after the combined. ·OH production quantification experiments have been revealed by the fluorescent intensity of TAOH for WO 3 , WO 3 &MIL-101 and WO 3 @MIL-101@WO 3 in Fig. 10b. For WO 3 &MIL-101, the produce of ·OH just have little changed compared with pure WO 3 . But for WO 3 @MIL-101@WO 3 the fluorescent intensity of TAOH compared with pure WO 3 were totally different, there has a significant increase of ·OH. Above result can be explained: the electron produced from WO 3 conduction band can be easily recombined due to the  www.nature.com/scientificreports www.nature.com/scientificreports/ positive conduction level because of rapidly recombination of electron-hole pairs, so there are little ·OH produced for WO 3 . For WO 3 &MIL-101, because of the similar ·OH quantify result and the result of PL and photocurrent, the mechanism is the same as pure WO 3 . This can be explained that WO 3 coating outside of MIL-101 only, nanoparticles are tending to be aggregating and electrons are favoring to be stacking and recombination rate increase. For WO 3 @MIL-101@WO 3 the electrons produced from conduction band of WO 3 transferred to MIL-101 due to the shorter electrons transfer distance. Due to the transformation of electrons, the holes can be separated to a greater degree, as a result, ·OH can be easily produced from high valence position of h + .
This may also conclude that WO 3 in the pores of MIL-101 have a shorter distance to transfer electrons from WO 3 conduction band resulting in higher electrons-hole separation efficiency comparing with WO 3 &MIL-101, the same conclusion have been shown in the literature 52 . It may be the reason the embedded structure has higher efficiency than coating structure. The possible photocatalytic mechanism of WO 3 @MIL-101@WO 3