Au Modified F-TiO2 for Efficient Photocatalytic Synthesis of Hydrogen Peroxide

In this work, Au-modified F-TiO2 is developed as a simple and efficient photocatalyst for H2O2 production under ultraviolet light. The Au/F-TiO2 photocatalyst avoids the necessity of adding fluoride into the reaction medium for enhancing H2O2 synthesis, as in a pure TiO2 reaction system. The F− modification inhibits the H2O2 decomposition through the formation of the ≡Ti–F complex. Au is an active cocatalyst for photocatalytic H2O2 production. We compared the activity of TiO2 with F− modification and without F− modification in the presence of Au, and found that the H2O2 production rate over Au/F-TiO2 reaches four times that of Au/TiO2. In situ electron spin resonance studies have shown that H2O2 is produced by stepwise single-electron oxygen reduction on the Au/F-TiO2 photocatalyst.


Introduction
Hydrogen peroxide (H 2 O 2 ) is widely used as a clean oxidant in environmental purification and organic synthesis [1,2]. It is widely used in pulp bleaching, wastewater treatment, and disinfection of industrial and household wastes with only water as the by-product [3]. At present, most H 2 O 2 in industry is prepared by the anthraquinone method with H 2 and O 2 [4]. This method requires a lot of energy and organic solvents with complicated reaction steps and high risk of explosion. Therefore, finding a simple and direct method for H 2 O 2 synthesis has become the focus of research. H 2 O 2 can be effectively produced through photo-electrocatalysis [5] and photocatalysis. In recent years, the photocatalytic synthesis of H 2 O 2 with oxygen and sunlight as the input energy has attracted great attention. At present, many semiconductor materials with UV and visible light activities, such as ZnO [6,7], C 3 N 4 [8][9][10], BiVO 4 [11], and TiO 2 [12][13][14][15][16][17][18] have demonstrated the potential for direct synthesis of H 2 O 2 . Especially when these semiconductors are loaded with appropriate cocatalysts, the photocatalytic activity of the catalysts could be greatly improved. Au has been proved to be a very effective cocatalyst for promoting H 2 O 2 production.
As a classic photocatalyst, TiO 2 is one of the most frequent and promising semiconductors because of its low cost and high stability. Under UV irradiation, H 2 O 2 can be directly produced in aqueous solution in the presence of O 2 without hydrogen over TiO 2 . An important feature of photocatalytic H 2 O 2 synthesis is that the formation of H 2 O 2 from the oxygen reduction reaction (ORR) is accompanied by the decomposition process. Zhao et al. [19] reported that adsorption of H 2 O 2 on TiO 2 will readily form surface peroxide complexes in the form of ≡Ti-OOH, which can be easily photodegraded with a zero-order kinetic process, even with the irradiation of visible light, thus leading to the decrease in H 2 O 2 production. Maurino et al. [20] also reported that the production of H 2 O 2 increased remarkably after adding fluoride into the reaction suspension of TiO 2 . These studies showed the competition of the F − with superoxide/peroxide species for the surface sites of TiO 2 . The ≡Ti-F formation decreases the amount of ≡Ti-OOH and thus, inhibits H 2 O 2 degradation. This method is interesting but it will cause fluoride pollution to the reaction medium and the difficulty of H 2 O 2 purification. In order to solve these problems, we developed F − -modified TiO 2 by a hydrothermal method instead of adding NaF in the photocatalytic reaction medium and used Au as the cocatalyst of F-TiO 2 . The anchored F − on the TiO 2 surface will compete with the formation of peroxide species to suppress the decomposition of H 2 O 2 and increase the H 2 O 2 production rate. F-TiO 2 avoided adding fluoride into the reaction medium as used in a pure TiO 2 reaction system and thus, simplified the reaction procedure. In situ ESR reveals that the H 2 O 2 is efficiently formed through a stepwise single-electron ORR process on the Au/F-TiO 2 photocatalyst.

Experimental Materials Preparation
To produce the F-TiO 2 photocatalyst, 1 g commercial anatase TiO 2 and 0.42 g NaF (n F :n Ti = 0.5:1) were mixed with 25 mL absolute ethanol and 15 mL water for hydrothermal treatment. The powder mixtures were maintained at 180 • C for 4 h in a homogeneous reactor. Then, the mixtures were transferred into the deionized water for centrifugation, washing and drying. By changing the amount of NaF and TiO 2 with different molar ratios of F/Ti, we prepared a series of F-TiO 2 photocatalysts. The photocatalysts loaded with 0.1 wt% Au were obtained by the deposition-precipitation method reported previously [21].

Material Characterization
UV-Vis spectra were recorded with a Spectrum Lambda 750 S (Perkin-Elmer, Waltham, MA, USA). High-resolution transmission electron microscopy (TEM) characterization was performed with an 8000EX microscope (JEOL, Tokyo, Japan) operating at 200 kV. The S-4800 scanning electron microscope (SEM) from Hitachi Instruments was used to observe the morphology of the photocatalyst.

Photocatalytic Activity Test
A reaction kettle (200 mL) was used as a photocatalytic reactor; 0.2 g Au/F-TiO 2 was added into the reaction solution of alcohol (4 wt%) and deionized water. F − was directly modified on the surface of the TiO 2 by the hydrothermal method without adding F − into the reaction solution. The suspension was treated by ultrasonication for 2-3 min; then, oxygen was bubbled for 30 min before turning on the light. A 300 W Xe arc lamp (PLS-SXE300, Perfectlight Technology Co., Ltd., Beijing, China) was used as a light source. The reaction was carried out under magnetic stirring water cooling. The concentrations of H 2 O 2 generated were determined by using the DMP (2, 9-dimethyl-1, 10-phenanthroline) method [22].

Quantification of H 2 O 2 (DMP Method)
One milliliter of DMP (0.1 g/L), 1 mL of copper (II) sulfate (0.1 M), 1 mL of phosphate buffer (Ph 7.0) solution, and 1 mL of reaction solution were added to a 10 mL volumetric flask and mixed; then, deionized water was added to the volumetric flask to the tick mark. After mixing, the absorbance of the sample at 454 nm was measured. The blank solution was prepared in the same manner but without H 2 O 2 .
The concentrations of H 2 O 2 were calculated by the following formula: where A 454 is the difference of the absorbance between the sample and blank solutions at 454 nm, ζ is the slope of the calibration curve, and [H 2 O 2 ] is the H 2 O 2 concentration (µM).

In Situ ESR Test
Using 5,5-dimethyl-1-pyrroline N-oxide (DMPO) as spin trapping reagent, the reduction pathways of O 2 on different catalysts were determined by in situ electron spin resonance (ESR) analysis. An ESP 300E spectrometer (Bruker, Switzerland) was used to detect the ESR signals of radicals trapped by DMPO. Generally, the catalyst (1 mg) was put into a mixture containing 1 mL alcohol/water (4 wt%) and 0.125 mmol DMPO. After passing the O 2 for 3 min, the sample was irradiated under UV light for 5 min before testing.

Results and Discussion
Au/TiO 2 and TiO 2 have similar XRD test spectra ( Figure S1). The diffraction peak of Au was not observed. It is presumed that the content of Au is too low and it is highly dispersed in the catalyst, which makes it impossible to form obvious characteristic diffraction peaks. In order to explore the existence and state of F − in the catalyst, the XPS of Au/TiO 2 and Au/F-TiO 2 was tested ( Figure S2). Compared with the full spectrum of Au/TiO 2 , the full spectrum of Au/F-TiO 2 has a peak corresponding to F1s between 600 eV and 700 eV, which can preliminarily prove that F − has been successfully introduced into the catalyst. From the peak fitting results of the high-resolution XPS spectrum of F1s ( Figure 1), it is found that the F1s is mainly composed of two peaks. The low binding energy peak at 683.4 eV is the signal peak of the formation of complex ≡Ti-F due to the chemical adsorption of F − on the surface of TiO 2 . The small peak with high binding energy near 684.5 eV is attributed to the signal peak of the doped F atom in TiO 2 ; that is, the F atom substituting for the oxygen site in TiO 2 lattice.

In Situ ESR Test
Using 5,5-dimethyl-1-pyrroline N-oxide (DMPO) as spin trapping reagen tion pathways of O2 on different catalysts were determined by in situ electro nance (ESR) analysis. An ESP 300E spectrometer (Bruker, Switzerland) was u the ESR signals of radicals trapped by DMPO. Generally, the catalyst (1 mg) w mixture containing 1 mL alcohol/water (4 wt%) and 0.125 mmol DMPO. Afte O2 for 3 min, the sample was irradiated under UV light for 5 min before testing

Results and Discussion
Au/TiO2 and TiO2 have similar XRD test spectra ( Figure S1). The diffrac Au was not observed. It is presumed that the content of Au is too low and dispersed in the catalyst, which makes it impossible to form obvious chara fraction peaks. In order to explore the existence and state of F − in the catalys Au/TiO2 and Au/F-TiO2 was tested ( Figure S2). Compared with the full Au/TiO2, the full spectrum of Au/F-TiO2 has a peak corresponding to F1s bet and 700 eV, which can preliminarily prove that F − has been successfully intr the catalyst. From the peak fitting results of the high-resolution XPS spec (Figure 1), it is found that the F1s is mainly composed of two peaks. The energy peak at 683.4 eV is the signal peak of the formation of complex ≡Tichemical adsorption of F − on the surface of TiO2. The small peak with high ergy near 684.5 eV is attributed to the signal peak of the doped F atom in TiO F atom substituting for the oxygen site in TiO2 lattice. From the TEM image of TiO2 and F-TiO2 (Figure 2), it is found that the of F-TiO2 prepared by the hydrothermal method has a very obvious chang with that of TiO2. TiO2 has an irregular shape, while F-TiO2 is almost sphe because F − has an etching effect on TiO2 during hydrothermal treatment [23]. strong complexation ability with Ti on the surface of TiO2, which corrodes th corners of TiO2 particles and changes the irregular TiO2 into a spherical shap From the TEM image of TiO 2 and F-TiO 2 (Figure 2), it is found that the morphology of F-TiO 2 prepared by the hydrothermal method has a very obvious change compared with that of TiO 2 . TiO 2 has an irregular shape, while F-TiO 2 is almost spherical. This is because F − has an etching effect on TiO 2 during hydrothermal treatment [23]. The F − has a strong complexation ability with Ti on the surface of TiO 2 , which corrodes the edges and corners of TiO 2 particles and changes the irregular TiO 2 into a spherical shape [24]. The SEM and energy dispersive spectroscopy (EDS) of 0.1% Au/F-TiO2 ( Figure S3) show that there are F and Au elements on the surface of the catalyst. This also proved that the F − modification and Au loading on TiO2 were successfully realized in the sample preparation. The element mapping of Au/F-TiO2 (nF: nTi = 2.5) in Figure 3 shows that both F and Au are evenly distributed on the surface of TiO2, which is consistent with the element types shown in the EDS result ( Figure S3).  The SEM and energy dispersive spectroscopy (EDS) of 0.1% Au/F-TiO 2 ( Figure S3) show that there are F and Au elements on the surface of the catalyst. This also proved that the F − modification and Au loading on TiO 2 were successfully realized in the sample preparation. The element mapping of Au/F-TiO 2 (n F : n Ti = 2.5) in Figure 3 shows that both F and Au are evenly distributed on the surface of TiO 2 , which is consistent with the element types shown in the EDS result ( Figure S3). The SEM and energy dispersive spectroscopy (EDS) of 0.1% Au/F-TiO2 ( Figure S3) show that there are F and Au elements on the surface of the catalyst. This also proved that the F − modification and Au loading on TiO2 were successfully realized in the sample preparation. The element mapping of Au/F-TiO2 (nF: nTi = 2.5) in Figure 3 shows that both F and Au are evenly distributed on the surface of TiO2, which is consistent with the element types shown in the EDS result ( Figure S3).   Figure 4a shows the UV-Vis spectrum of F-TiO 2 , 0.1% Au/F-TiO 2 and pure TiO 2 . Besides the characteristic absorption bands of TiO 2 at lower than 370 nm, the absorption bands caused by the loading of Au nanoparticles are located between 500 nm and 650 nm, which is a typical Au surface plasma band [25]. According to the calculation, the band gap of TiO 2 is about 3.2 eV and F-TiO 2 is about 3.1 eV (Figure 4b).
Molecules 2021, 26, x FOR PEER REVIEW 5 of 8 Figure 4a shows the UV-Vis spectrum of F-TiO2, 0.1% Au/F-TiO2 and pure TiO2. Besides the characteristic absorption bands of TiO2 at lower than 370 nm, the absorption bands caused by the loading of Au nanoparticles are located between 500 nm and 650 nm, which is a typical Au surface plasma band [25]. According to the calculation, the band gap of TiO2 is about 3.2 eV and F-TiO2 is about 3.1 eV (Figure 4b). The photocatalytic activity of H2O2 synthesis on Au/F-TiO2 hybrids was tested under UV light and the concentration of H2O2 was quantified by spectrophotometry with copper ions and 2,9-dimethyl-1,10-phenanthroline (DMP). The standard curve showed that there was a good linear relationship between the absorbance and concentration of H2O2; the R squared value was 0.9996 ( Figure S3). Figure 5 shows the photocatalytic synthesis of H2O2 over Au-loaded F-TiO2 catalysts. Compared with the unmodified catalyst, the photocatalytic activity increased with the increase in F content, and the photocatalytic activity reached its highest when the F/Ti molar ratio increased to 2.5. The F − on the surface of TiO2 will compete with superoxide/peroxide species for the surface sites of TiO2 and inhibit the adsorption of peroxy radicals, thus suppressing the decomposition of H2O2. With the continuous increase in F content, when the F/Ti molar ratio is 3, the activity of the catalyst decreases. Excessive F − caused serious defects on the surface of TiO2 and destroyed the crystallinity of TiO2, thus decreasing the photocatalytic activity of the catalyst.  The photocatalytic activity of H 2 O 2 synthesis on Au/F-TiO 2 hybrids was tested under UV light and the concentration of H 2 O 2 was quantified by spectrophotometry with copper ions and 2,9-dimethyl-1,10-phenanthroline (DMP). The standard curve showed that there was a good linear relationship between the absorbance and concentration of H 2 O 2 ; the R squared value was 0.9996 ( Figure S3). Figure 5 shows the photocatalytic synthesis of H 2 O 2 over Au-loaded F-TiO 2 catalysts. Compared with the unmodified catalyst, the photocatalytic activity increased with the increase in F content, and the photocatalytic activity reached its highest when the F/Ti molar ratio increased to 2.5. The F − on the surface of TiO 2 will compete with superoxide/peroxide species for the surface sites of TiO 2 and inhibit the adsorption of peroxy radicals, thus suppressing the decomposition of H 2 O 2 . With the continuous increase in F content, when the F/Ti molar ratio is 3, the activity of the catalyst decreases. Excessive F − caused serious defects on the surface of TiO 2 and destroyed the crystallinity of TiO 2 , thus decreasing the photocatalytic activity of the catalyst.
Molecules 2021, 26, x FOR PEER REVIEW 5 of 8 Figure 4a shows the UV-Vis spectrum of F-TiO2, 0.1% Au/F-TiO2 and pure TiO2. Besides the characteristic absorption bands of TiO2 at lower than 370 nm, the absorption bands caused by the loading of Au nanoparticles are located between 500 nm and 650 nm, which is a typical Au surface plasma band [25]. According to the calculation, the band gap of TiO2 is about 3.2 eV and F-TiO2 is about 3.1 eV (Figure 4b). The photocatalytic activity of H2O2 synthesis on Au/F-TiO2 hybrids was tested under UV light and the concentration of H2O2 was quantified by spectrophotometry with copper ions and 2,9-dimethyl-1,10-phenanthroline (DMP). The standard curve showed that there was a good linear relationship between the absorbance and concentration of H2O2; the R squared value was 0.9996 ( Figure S3). Figure 5 shows the photocatalytic synthesis of H2O2 over Au-loaded F-TiO2 catalysts. Compared with the unmodified catalyst, the photocatalytic activity increased with the increase in F content, and the photocatalytic activity reached its highest when the F/Ti molar ratio increased to 2.5. The F − on the surface of TiO2 will compete with superoxide/peroxide species for the surface sites of TiO2 and inhibit the adsorption of peroxy radicals, thus suppressing the decomposition of H2O2. With the continuous increase in F content, when the F/Ti molar ratio is 3, the activity of the catalyst decreases. Excessive F − caused serious defects on the surface of TiO2 and destroyed the crystallinity of TiO2, thus decreasing the photocatalytic activity of the catalyst.  In general, H 2 O 2 from 2 e − ORR by CB electrons can be produced through stepwise coupled electrons and proton transfers (Equations (1)-(3)) [26]. In order to further study the mechanism of photocatalytic ORR for H 2 O 2 synthesis over Au/F-TiO 2 , DMPO was used as a trapping agent of free radical in situ ESR tests for different samples. In situ ESR spectra of Au/F-TiO 2 , Au/TiO 2 and pure TiO 2 under UV irradiation are shown in Figure 6. The results clearly show the signal of •OOH formed via equation (2) over various TiO 2 photocatalysts [27]. The DMPO-•OOH radical signal could be detected in both Au-loaded samples except pure TiO 2 . The superoxide radical is formed by the first combination of O 2 in the photocatalytic reaction medium with electrons and protons. The generated HO 2 • will continue to react with one electron and a proton and finally, generate H 2 O 2 . Therefore, the H 2 O 2 is formed by a stepwise single-electron ORR over Au/F-TiO 2 . In addition, compared with TiO 2 , Au/TiO 2 and Au/F-TiO 2 produced a more obvious HO 2 • signal, which implied that Au and F − promoted the formation of HO 2 •, and both of them promoted the photocatalytic synthesis of H 2 O 2 .
Molecules 2021, 26, x FOR PEER REVIEW In general, H2O2 from 2 e − ORR by CB electrons can be produced through s coupled electrons and proton transfers (Equations (1)-(3)) [26]. In order to furthe the mechanism of photocatalytic ORR for H2O2 synthesis over Au/F-TiO2, DM used as a trapping agent of free radical in situ ESR tests for different samples. In s spectra of Au/F-TiO2, Au/TiO2 and pure TiO2 under UV irradiation are shown in F The results clearly show the signal of •OOH formed via equation (2) over vario photocatalysts [27]. The DMPO-•OOH radical signal could be detected Au-loaded samples except pure TiO2. The superoxide radical is formed by the fi bination of O2 in the photocatalytic reaction medium with electrons and proto generated HO2• will continue to react with one electron and a proton and finally ate H2O2. Therefore, the H2O2 is formed by a stepwise single-electron OR Au/F-TiO2. In addition, compared with TiO2, Au/TiO2 and Au/F-TiO2 produced obvious HO2• signal, which implied that Au and F − promoted the formation o and both of them promoted the photocatalytic synthesis of H2O2.

Conclusions
In this work, we have designed Au/F-TiO2 as an efficient photocatalyst for duction of H2O2 in aqueous solution. The Au/F-TiO2 makes it possible to obtain H2O2 yield in fluoride-free reaction medium. The H2O2 production rate reach times that of Au/TiO2. The in situ ESR test showed that the synthesis mechanism was not changed by F − modification. The H2O2 was synthesized over Au/F-TiO2 th stepwise single-electron ORR.

Conclusions
In this work, we have designed Au/F-TiO 2 as an efficient photocatalyst for the production of H 2 O 2 in aqueous solution. The Au/F-TiO 2 makes it possible to obtain a high H 2 O 2 yield in fluoride-free reaction medium. The H 2 O 2 production rate reached four times that of Au/TiO 2 . The in situ ESR test showed that the synthesis mechanism of H 2 O 2 was not changed by F − modification. The H 2 O 2 was synthesized over Au/F-TiO 2 through a stepwise single-electron ORR.