Enhanced photocatalytic activity of a hollow TiO₂–Au–TiO₂ sandwich structured nanocomposite

Abstract

Combining TiO₂ photocatalysts with noble metal nanoparticles (NPs) has been deemed an effective method to enhance visible-light-induced photocatalytic activity. Herein, we report that the mesoporous hollow TiO₂–Au–TiO₂ (MHTAT) sandwich structured nanocomposite was synthesized using resorcinol–formaldehyde resin polymer nanospheres as templates, followed by TiO₂ coating, Au NP deposition, and external TiO₂ coating, and then hydrothermal treatment and calcination for post-treatment of template removal. The MHTAT nanocomposites have enhanced visible light harvesting efficiency through their unique hierarchical nanostructure and high surface-to-volume ratio, which are beneficial to enhancing visible photocatalytic performance. The Au NPs were sandwiched between two shells of TiO₂ in the MHTAT nanocomposites, which formed close Schottky contact for electron transfer through the interface between TiO₂ and Au in the photocatalytic reaction. The photocatalytic activity of MHTAT sandwiched nanocomposites was fully demonstrated in degradation reactions of a number of organic compounds under visible light irradiation, suggesting their intriguing potential as effective visible-light-induced photocatalysts.

---

1. Introduction

Titania is one of the most promising photocatalysts for water splitting and degradation of organic pollutants under UV light, because of its advantages such as low cost, low toxicity, high chemical stability and high photocatalytic efficiency.^1–4 However, the inferior utilization efficiency of visible light restricts the practical applications of TiO₂ nanomaterials due to its wide band gap (3.2 eV) and high recombination rate of electron–hole pairs.^5,6 Thus, a lot of efforts have been devoted to the development of visible-light-induced TiO₂ photocatalysts.^7–10 The mechanism of the photocatalytic degradation of organic pollutants under UV light (λ < 400 nm) and visible light (λ > 400 nm) is different.^11,12 Deposition of noble metal on the surface of TiO₂ enhances the photocatalytic efficiency under UV light by photogenerated electrons/holes which can lead to the interfacial charge transfer and the decrease in the recombination rate. However, the mechanism of the photocatalytic process under the visible light is determined that the promising strategy can be the loading of noble metal nanoparticles (NPs) on the surface of TiO₂ based on localized surface plasmon resonance (LSPR) photosensitization. When Au NPs are coupled with TiO₂ photocatalysts, excited electrons in Au NPs can be transferred through the interface to the conduction band (CB) of TiO₂ under visible light excitation. It is to be noted that unlike in the case of photocatalysis of TiO₂–Au under UV-light, electron–hole pairs are not generated in TiO₂ during visible light exposure. Under visible light irradiation, TiO₂ is not excited because of the wide band gap, whereas the SPR of Au NPs is excited and the energetic electrons are transferred to the empty CB of TiO₂. These electrons are then used for photocatalytic degradation of organic pollutants.^13–15 Thus, visible light can be utilized by TiO₂/Au photocatalysts.

Recently, TiO₂/Au photocatalysts have been used for the photodegradation of organic pollutants. Yin and co-workers have synthesized the TiO₂/Au catalyst with different morphology and showed the high catalytic performance.^16–18 Takeshi Sakamoto and co-workers prepared the submicrometer-sized, monodisperse titania particles incorporating Au NPs with the photocatalytic activity higher than the P25.¹⁹ However, the Au NPs were loosely attached to the surface of TiO₂ in reported methods, which were unstable during calcination and subsequent photocatalysis. Thus, the unstable Au NPs would migrate and coalesce, which significantly reduced or even decreased the catalytic activity of TiO₂/Au catalysts rapidly. To overcome this critical obstacle, the confinement of Au NPs within hollow mesoporous supports and encapsulation of Au NPs into core–shell sandwiched structure may be an effective strategy to improve the stability and recyclability of the photocatalysts. The encapsulation of Au NPs into TiO₂ shell nanostructures increases the contact area between Au NPs and TiO₂ matrix, and therefore allows for more efficient electron transfer. The MHTAT have the visible light harvesting efficiency and mass transfer through the hierarchical mesoporous structure.^20–22 The mesoporous hollow TiO₂ sandwich structure nanocomposite comprising plasmonic metal nanoparticles with outstanding photocatalytic performances have been seldom seen.

Herein, we utilized the novel sol–gel process, hydrothermal treatment and calcination process step-by-step to synthesize the mesoporous hollow TiO₂–Au–TiO₂ (MHTAT) sandwich structured nanocomposites. In a typical synthesis (Scheme 1), the resorcinol–formaldehyde resin polymer (RF) nanospheres were used as templates, loaded sequentially with TiO₂ coating, Au NPs deposition, and finally coated with a layer of TiO₂. Post-treatments including hydrothermal treatment and calcinations led to the formation of MHTAT sandwich nanocomposite, which exhibited the high crystallinity, stability, large specific surface area and superior adsorption capability. Moreover, the formation of close Schottky contact between Au and TiO₂ NPs after the calcination process is favor to electron transfer through this interface for photocatalytic reactions.²³

Scheme 1 Schematic illustration of the synthesis steps for MHTAT sandwich nanocomposite.

2. Experimental section

2.1 Materials

Acetonitrile, sodium citrate (SC), ammonia, triethanolamine (TEOA) and tert-butyl alcohol (TBA) were purchased from Shanghai Chemical Reagent Company (Shanghai, China). Resorcinol, formaldehyde, tetrachloroauric acid tetrahydrate (HAuCl₄·4H₂O), 3-mercaptopropionic acid (3-MPA), tetrabutyl titanate (TBOT), crystal violet (CV), Rhodamine 6G (RhB), hydroxypropyl cellulose (HPC), terephthalic acid and Methyl Orange (MO) were obtained from Sigma Company. Deionized water (>18.0 MΩ cm) was purified using a Millipore Milli-Q gradient system throughout the experiment. All chemicals were of analytical grade and used as received.

2.2 Synthesis of mesoporous hollow TiO₂–Au–TiO₂ (MHTAT) sandwich nanocomposites

The monodisperse RF NPs were synthesized according to a previously reported procedure,²⁴ typically by mixing ethanol (8 mL), deionized water (20 mL), and ammonia aqueous solution (0.1 mL). After stirring for 1 h, resorcinol (0.1 g) and formaldehyde solution (0.14 mL) were added to the reaction solution and stirred overnight. The RF NPs (50 mg) were dispersed in a mixture of ethanol (15 mL) and acetonitrile (7 mL) under vigorous stirring. After adding aqueous ammonia (0.2 mL, 28%), TBOT (0.2 mL) in a mixed solvent of ethanol (3 mL) and acetonitrile (1 mL) was injected into the mixture. After 3 hours, the composite was isolated by centrifugation and washed with ethanol to get RF–TiO₂ (RT) core–shell composites. The functionalizations were carried out by mixing 100 μL of 3-MPA which was sufficient to provide an approximately layer coating on the RT particles.

The Au NPs were synthesized by reducing HAuCl₄ with sodium citrate according to Grabar's method.²⁵ 5 mL 3-MPA functionalized RT particles (∼10 mg mL⁻¹) was added dropwise to the 50 mL freshly prepared Au NPs solution and stirred for 12 h in order to form RT particles covered with small gold NPs (RTA). Non-attached gold NPs were removed by centrifugation. Afterward, the RTA solution was dispersed in a mixture of ethanol (15 mL), acetonitrile (7 mL) and HPC (0.1 g) under vigorous stirring. After adding aqueous ammonia (0.2 mL, 28%), TBOT (0.2 mL) in a mixed solvent of ethanol (3 mL) and acetonitrile (1 mL) was injected into the mixture. The precipitate of RF–TiO₂–Au–TiO₂ (RTAT) was separated by centrifugation, and washed three times with ethanol.

The synthesized RTAT samples (55 mg) were dispersed into 70 mL deionized water under sonication for 10 minutes, and then the mixture was transferred into a 100 mL Teflon-lined stainless steel autoclave to get RTAT hydrothermal nanocomposites (RTAT-H). After that, the obtained products were collected by centrifugation, washed three times with ultrapure water, and then dried in vacuum at room temperature overnight. To obtain mesoporous hollow TiO₂–Au–TiO₂ (MHTAT) sandwich nanocomposite, the RTAT-H sample was calcined in air atmosphere at 400 °C for 2 h.

2.3 Photocatalytic activities measurements

Photocatalytic activity of each sample was evaluated by photodegradation of dyes. The photocatalyst (10 mg) was first added into a 50 mL quartz photoreactor containing 40 mL of 2 × 10⁻⁵ mM dyes solution. Before irradiation under simulated daylight (500 W xenon) irradiation, the mixture was ultrasonicated for 2 min and magnetically stirred for 30 min in the dark to ensure good dispersion of photocatalysts in solution and reached the adsorption–desorption equilibrium between dyes molecules and the surface of photocatalysts. The xenon lamp was used as the light source and the irradiation range was cut off below 420 nm using a Pyrex glass for the simulation of sunlight. At the given time intervals, 2 mL analytical suspensions were sampled from the mixture solution and immediately centrifuged at 8000 rpm for 5 min to purify the samples.

2.4 Characterization

The surface topography of nanocomposites was taken on environmental scanning electron microscope (FEI, Quanta 200 FEG, USA) at an accelerating voltage of 10.0 kV. X-ray scattering patterns were obtained by analyzing the powder samples on an X-ray diffractometer (Philips, X'Pert-PRO, Netherlands) with using the Ni-filtered monochromatic Cu Kα radiation (λKα1 = 1.5418 Å) at 40 keV and 40 mA. Transmission electron microscopy (TEM) images were recorded by a high resolution transmission electron microscope (JEOL, JEM 2010, Japan), equipped with X-ray energy dispersive spectroscopy (EDS) capabilities, operated at an acceleration voltage of 200 kV. The UV-Vis absorption spectra were recorded with a UV-vis spectrophotometer (Shimadzu, UV-2700, Japan). The specific surface area and the pore distribution of the samples were calculated by nitrogen adsorption (Micrometrics, ASAP 2020M, USA) at 77 K using the Brunauer–Emmett–Teller (BET) equation. The optical properties of photocatalysts were measured by UV-Vis diffuse reflectance spectra (Shimadzu, UV3600-MPC3100, Japan). The fluorescence spectra of terephthalic acid were measured on a fluorescence spectrophotometer (Horiba, FluoroMax-4, USA). The experimental procedure was similar to the photocatalytic experiment except using terephthalic acid (TA) solution (5 × 10⁻⁴ M) with NaOH concentration of 2 × 10⁻³ M instead of dyes molecules.

3. Results and discussion

The MHTAT sandwich structured nanocomposite was synthesized using RF NPs as templates, followed by TiO₂ coating, Au NPs deposition, external TiO₂ covering and finally hydrothermal treatment and calcination for post-treatments of template removal. At first, the RF NPs were synthesized by the modified Stöber method, using resorcinol and formaldehyde as the precursor, NH₃·H₂O as the catalyst. As revealed by the SEM image and the size histogram in Fig. 1A, the uniform RF NPs were obtained with a diameter of 450 ± 20 nm. Through a sol–gel process of hydrolysis and condensation of TBOT in acetonitrile/ammonia mixtures, the core–shell RT nanocomposites were obtained with a diameter of 547 ± 18 nm (Fig. 1B). By comparing the size histogram of Fig. 1A and B, the first TiO₂ shells appear to be a compact matrix with a thickness of 49 ± 14 nm. The Au NPs with a diameter of 17 ± 2 nm have a UV absorbance peak centered at 520 nm (Fig. S1A†).²⁶ The Au NPs were capable of sensitizing wide-band gap TiO₂ exhibiting much enhanced visible-light response.²⁷ The monodispersed Au NPs were coated on the TiO₂ shell by modification of 3-MPA as the coupling agent. The 3-MPA can dually provide driving force for the efficient loading of the Au NPs onto the TiO₂ nanocomposites. Fig. S1B† showed FTIR spectra of the RF–TiO₂–MPA and MHTAT nanocomposites. It was observed that the clear vibrational bands of 3-MPA (1711 cm⁻¹ from –COOH) emerged in the spectra of RF–TiO₂–MPA, while the 3-MPA were disappeared in the MHTAT nanocomposites. It demonstrated that 3-MPA was completely removed after calcination and led to the formation of the close TiO₂–Au Schottky contact.²⁸ In order to prevent the Au NPs from aggregation, another layer of TiO₂ was coated on the RTA nanocomposite. In the presence of HPC, the RTA nanocomposites were over coated with another layer of amorphous TiO₂ by hydrolyzing TBOT in ethanol solution. Fig. 1D showed the SEM image of the RTAT NPs with a diameter of 613 ± 22 nm. The average particle diameter increased from 560 to 613 nm after the coating process, which indicated the precipitation of a 27 ± 14 nm thick TiO₂ shell on the RTA core.

Fig. 1 SEM images of (A) the RF NPs, (B) RT nanocomposite, (C) RTA nanocomposite, (D) RTAT nanocomposite. Insert is the corresponding size histogram.

The TiO₂ shells can transform from the amorphous to crystalline structures after the hydrothermal treatment and calcination process. As shown in Fig. 2, it was clearly observed that the smooth surface changes to the rough surface after the hydrothermal treatment by comparison with Fig. 2A and B. The hydrothermal treatment can make the amorphous TiO₂ to the crystallite nanoparticles of the TiO₂.²² The calcination process can get the high surface area of mesoporous hollow nanospheres. Comparing Fig. 2B with 2C, it shown that the morphology and structure of these nanocomposites had no apparent change after calcinations process, which indicated the sandwiched structure was not destroyed after the calcinations process. Based on the TEM images (Fig. 2D), we can conclude that the products are the uniform hollow spheres with the sizes of 613 ± 22 nm, where the whole shell thickness is 82 ± 11 nm.

Fig. 2 (A) SEM image of RTAT nanocomposites after sol–gel process, (B) SEM image of RTAT-H, (C) SEM image and (D) TEM image of MHTAT.

In order to further investigate the elemental distribution and structure of the MHTAT nanocomposites, the elemental mapping was performed by HRTEM at HAADF (high angle annular dark field). As mentioned above, this nanocomposite has three layers which was TiO₂ layer, Au layer and the outmost TiO₂ layer, respectively. The EDS elemental mapping of Ti (Fig. 3B), O (Fig. 3C) and Au (Fig. 3D) in a single shell further confirmed the expected structure. Fig. 3A and B further proved this nanocomposite was the hollow mesoporous nanosphere. Fig. 3D showed that the Au element dispersed on the mesoporous TiO₂ shell. The EDAX spectrum of this nanocomposite showed the content of the nanocomposites was about 2 wt% Au and 39 wt% Ti.

Fig. 3 (A) The bright-field TEM image and the elemental mapping of the MHTAT nanocomposites from the HRTEM images, based on the (B) Ti Ka1, (C) O Ka1 and (D) Au La1, (E) the EDAX spectrum of the MHTAT nanocomposites.

The MHTAT nanocomposites were also characterized by HRTEM (Fig. S2†) which showed the presence of Au (fringe spacing = 0.24 nm) within the anatase TiO₂ (fringe spacing = 0.36 nm) crystallite domain. It was clearly shown that the lattice fringes with a lattice spacing of about 0.24 nm correspond to the (111) plane of Au and the lattice spacing of 0.36 nm corresponds to the (111) plane of the TiO₂.²⁹ In addition, the crystallographic orientations of the Au with TiO₂ indicated that the Au was embedded in the TiO₂ surface. More importantly, as shown in Fig. S2,† a close contact was formed at the interface of Au NPs and TiO₂, which would facilitate the electron or energy transfer in this composite photocatalyst.

The XRD patterns of MHTAT sandwich structured nanocomposites were shown in Fig. 4A. The identified peaks for TiO₂ were assigned to anatase TiO₂ (JCPDS, no. 73-1764), and the peaks located at 25.3°, 37.8°, 48.1°, 54° and 55° which were attributed to the (101), (004), (200), (105) and (211), respectively. In addition to the other four diffraction peaks at the 2θ values 38.1°, 44.3° and 64.4° correspond to the (111), (200) and (220) of the cubic phase of Au (JCPDS card no. 04-07834).³⁰ The UV/Vis diffuse reflectance spectra were conducted to uncover the optical properties of MHTAT sandwich nanocomposites (Fig. 4B). The addition of Au induced the increased light absorption intensity in visible light regions, as compared to the bare hollow TiO₂. The absorption band for MHTAT nanocomposites at around 552 nm appeared which was ascribed to the surface plasmon resonance of Au NPs. And then, we have compared the UV/Vis diffuse reflectance of TiO₂@Au and MHTAT and found the TiO₂@Au nanocomposites have the camel-like adsorption peak around the 520 nm to 680 nm. Many factors such as the particles sizes, the particles sizes distribution and particles morphology could influence the absorption band. When increasing the size of Au NPs, the surface plasmon absorption became wider and shifted to the longer wavelengths. This agglomerated Au NPs of TiO₂@Au could reduce the absorption efficiency of visible light.³¹ The corresponding N₂ adsorption–desorption isotherms and pore size distributions of MHTAT nanocomposites were characterized by N₂ physisorption experiments, as shown in Fig. 4C and D. It showed that the MHTAT nanocomposites have type III isotherms and type H₃ hysteresis loops, indicating the existence of slit-shape pores according to the IUPAC classification. The surface area as measured by the multi-point BET method from the adsorption branch was 421 m² g⁻¹. The pore size distribution of the sample as determined using the BJH method from the adsorption branch of the isotherm was shown in Fig. 4D, which suggested an average pore diameter of 2.2 nm. Therefore, the MHTAT sandwich structured nanocomposites were mesoporous structure.

Fig. 4 (A) XRD patterns of the MHTAT nanocomposites, (B) UV/Vis diffuse reflectance spectra of the TiO₂, TiO₂@Au and MHTAT photocatalysts, (C) nitrogen-sorption isotherm of the MHTAT nanocomposites, (D) the corresponding pore size distributions calculated by the Barrett–Joyner–Halenda method.

Before investigate the photocatalytic activity of MHTAT nano-composites for photocatalytic degradation of organic pollutants, we compared the performance of common TiO₂@Au composites and MHTAT to investigate the advantage of the sandwich structure. We found that the sandwich structure can protect the Au NPs from migration and coalescence compared to common TiO₂@Au composites, in which Au NPs were loosely attached on the surface of TiO₂ and unstable during calcination and subsequent photocatalysis. Fig. 5A showed that the small Au NPs on the surface of TiO₂ NPs fuse together and became lager Au particles after calcination. However, the Au NPs sandwiched between two TiO₂ layers still evenly dispersed in the MHTAT sandwich nanocomposites after the calcination, as shown in Fig. 5B. To explore the photocatalytic activity of TiO₂@Au and the MHTAT sandwich structured nanocomposites, the photodegradation of MB was investigated under visible light irradiation. As shown in Fig. 5C and D, simulated by under visible light irradiation completely decompose MB molecules within 120 min with the aid of MHTAT sandwich-structured photocatalysts, while the decomposition only reaches approximate 80% with TiO₂@Au during the same period. It attributed to the agglomerated Au NPs of TiO₂@Au reduced the absorption efficiency of visible light. Fig. 5D showed the rate constant of MHTAT (0.023 min⁻¹) was almost twice over the TiO₂@Au (0.012 min⁻¹). By comparing the photocatalytic activity of TiO₂@Au and MHTAT indicated that the MHTAT have higher light-harvesting efficiency and stability.

Fig. 5 SEM images of (A) the TiO₂@Au nanocomposites, (B) the MHTAT nanocomposites, (C) evolution of the MB concentration versus reaction time, (D) apparent reaction rate constant in the presence of MHTAT and TiO₂@Au as the photocatalysts.

To explore the photocatalytic activity of MHTAT nano-composites for real applications, the photodegradation of organic compounds was investigated under visible light irradiation. The nanocomposites were immersed in MO aqueous solution for 30 min to ensure that the products reached adsorption equilibrium. As shown in Fig. 6A, under visible light irradiation completely decomposed MO molecules within 110 min with the aid of MHTAT sandwich-structured photocatalysts, while the conversion only reaches approximate value of 37% with P25 during the same period because the P25 mainly absorb the ultraviolet light. The RTAT-H was reached approximate values of 81% during the same period, respectively. After adding the MHTAT nanocomposites, the conversion could reach approximate values of 90% because of its high crystallinity and larger specific surface area. The photocatalytic efficiency of MHTAT was increased after calcination at 400 °C, which might be attributed to increase the large specific surface area and improved contact between the Au NPs and TiO₂ NPs during annealing at high temperature. Fig. 6B also showed the photocatalytic activity of the catalysts for MO degradation follows the order: MHTAT > RTAT-H > P25.

Fig. 6 (A) Evolution of MO concentration versus reaction time and (B) apparent reaction rate constant under simulated daylight irradiation in the presence of P25, RTAT-H, and MHTAT, (C) evolution of MO concentration versus reaction time adding TBA and adding TEOA as the photocatalysts, (D) evaluation of the photocatalytic activities and cycling tests of simulated daylight irradiation photocatalytic activities of MHTAT.

The photodegradation mechanisms of MO by the MHTAT nanocomposites could be attributed to the direct hole oxidation or the photoreduction process. As we know, the triethanolamine (TEOA) is an effective hole scavenger and the tert-butyl alcohol (TBA) is the radical scavenger for photocatalytic reaction.³² In this experiment, the 10 vol% TEOA was introduced into the reaction system and the photocatalytic degradation was hardly suppressed in this experiment. However, the 10 vol% TBA was added to this reaction solution, the photocatalytic degradation efficiency significantly decreased (Fig. 6C). This experimental results demonstrated that the photocatalytic degradation of MO using MHTAT nanocomposites could be directly related to ˙OH radicals generated during photocatalysis.

The detail mechanism of electron injection from Au NPs to TiO₂ is as follows. In addition to enhanced visible light induced photocatalysis, MHTAT nanocomposites have a unique characteristic to interact with visible light through excitation of LSPR. The electron states in the Au NPs follow Fermi–Dirac distribution.³³ Upon plasmon excitation, many electrons of the Au NPs have higher energy than that of the conduction band (CB) of TiO₂. The Au (h⁺) was also the active site and could catalyze the H₂O to the ˙OH radicals. Moreover, the electrons generated by the excitation of LSPR were injected into the CB of TiO₂, which reduces the dissolved oxygen on the surface of TiO₂, and forms ˙OH radicals. The ˙OH radicals were highly reactive species and responsible for the degradation of pollutants (Scheme 2). The pathway (reactions (1)–(6)) was identified as involving formation of superoxide radical anions and hydroxyl radicals.³⁴

Au + hν → e⁻ + h⁺ (1)

TiO₂ + e⁻ → TiO₂ (e_CB⁻) (2)

e⁻ + O₂ → O₂⁻˙ (3)

O₂⁻˙ + H₂O → ˙OOH + OH⁻ (4)

˙OOH + H₂O + e⁻ → H₂O₂ + OH⁻ (5a)

2˙OOH → O₂ + H₂O₂ (5b)

H₂O₂ + e⁻ → OH⁻ + ˙OH (6)

Scheme 2 Charge transfer in monodisperse MHTAT sandwiched nanocomposite.

This mechanism of charge injection from Au NPs to TiO₂ was applicable when the Au NPs and TiO₂ are in close contact with each other, allowing for the facile electron transfer. In this experiment, the calcination process could improve the contact between the Au NPs and TiO₂ NPs to be conducive for the transport of electrons. Fig. 6C suggested that the photocatalytic degradation were attributed to the photoreduction process to generated the OH˙ with the strong oxidizing.

In general, hydroxyl radical is a key active species in the photocatalytic process. The oxidative ability of hydroxyl radical is high enough to attack many organic molecules. Terephthalic acid photoluminescence probing technique (TAPL), a highly sensitive and simple method, has been widely used in detection of hydroxyl radicals.³⁵ The reaction between active species ˙OH radicals and TA has been shown in Fig. S3.†

The fluorescence spectral of TA solution with MHTAT under visible light irradiation was shown in Fig. S4.† The PL intensity at around 426 nm increased with the irradiation time, indicating that the amount of ˙OH radicals generated on the surface of photocatalysts increases with the time. Fig. S4B† indicated that the photoluminescence intensity of the MHTAT solution linearly increased by the irradiation time and the ratio of hydroxyl radical generation rate (V) to time (t) followed linear dependence. Moreover, Fig. S4C† showed the evolution MO concentration versus reaction time and Fig. S4D† showed apparent reaction rate constant. The plots exhibited a linear characteristic which was in accordance with the hydroxyl radical generation rate, suggesting that the main reactive species may be the ˙OH in MHTAT photocatalytic process.

The as-prepared MHTAT sandwich nanocomposites exhibited both high photocatalytic activity and stability properties after the catalytic reduction. Fig. 6D showed the recyclable photocatalytic activity of MHTAT sandwich nanocomposites. The catalysts were successfully recycled and reused for at least five successive cycles of reaction with a stable conversion efficiency of approximately 90%. We also measured the morphology of MHTAT after five successive cycles of reaction by TEM, as shown in Fig. S5.† We observed the Au NPs were still uniform coated on the TiO₂ shell and the hollow nanospheres had no collapse. Therefore, we concluded that the second TiO₂ shell protected the Au NPs to avoid agglomeration in the photodegradation process. We also investigated the photocatalytic activity of MHTAT sandwich nanocomposites for degradation of other organic molecules under simulated daylight irradiation, e.g. RhB, methylene blue and crystal violet at the same initial concentration. As shown in Fig. S6,† all of those molecules can be decomposed almost completely (>80%) within 100 min.

4. Conclusions

In summary, we report a new kind of MHTAT sandwich nanocomposites, which were developed by the simple “sol–gel, hydrothermal treatment and calcination” method. In this study, the RF sphere were used as templates, loaded sequentially with TiO₂ coating, Au NPs deposition, and finally covered with a layer of TiO₂. Post-treatments including hydrothermal treatment and calcination led to the formation of MHTAT nanocomposite, which exhibited the high crystallinity, stability, large specific surface area and superior adsorption capability. This sandwich structure led to the formation of close Schottky contact between Au and TiO₂ NPs after the calcination process. This sandwich structure protected the Au NPs from migration and coalescence compared to common TiO₂@Au composites, in which Au NPs are loosely attached to the surface of TiO₂ such that they are unstable during calcination and subsequent photocatalysis. The MHTAT sandwich structured nanocomposites with the high crystallinity, stability, large specific surface area and superior adsorption capability are expected to have the great potential application in photocatalytic reaction and energy conversion in the future.

Acknowledgements

This work was supported by the Natural Science Foundation of China (Grant No. 51502296, 51472246, and 51272255), Anhui Provincial Natural Science Foundation (Grant No. 1608085QE89), and Hundred Talent Program of Chinese Academy of Sciences.

Notes and references

1. X. L. Liang, J. Li, J. B. Joo, A. Gutierrez, A. Tillekaratne, I. Lee, Y. D. Yin and F. Zaera, Angew. Chem., Int. Ed., 2012, 51, 8034–8036.
2. W. Li, Z. X. Wu, J. X. Wang, A. A. Elzatahry and D. Y. Zhao, Chem. Mater., 2014, 26, 287–298.
3. I. Lee, J. B. Joo, Y. D. Yin and F. Zaera, Angew. Chem., Int. Ed., 2011, 50, 10208–10211.
4. Y. Yu, C. Y. Cao, Z. Chen, H. Liu, P. Li, Z. F. Dou and W. G. Song, Chem. Commun., 2013, 49, 3116–3118.
5. C. H. Fang, H. L. Jia, S. Chang, Q. F. Ruan, P. Wang, T. Chen and J. F. Wang, Energy Environ. Sci., 2014, 7, 3431–3438.
6. X. B. Chen, L. Liu, P. Y. Yu and S. S. Mao, Science, 2011, 331, 746–750.
7. M. M. Gao, C. K. N. Peh, Y. L. Pan, Q. H. Xu and G. W. Ho, Nanoscale, 2014, 6, 12655–12664.
8. S. T. Kochuveedu, Y. H. Jang and D. H. Kim, Chem. Soc. Rev., 2013, 42, 8467–8493.
9. Y. X. Zhang, X. Y. Yu, Y. Jia, Z. Jin, J. H. Liu and X. J. Huang, Eur. J. Inorg. Chem., 2011, 5096–5104.
10. G. D. Li and Z. Y. Tang, Nanoscale, 2014, 6, 3995–4011.
11. Y. J. Jang and D. H. Kim, Chem.–Eur. J., 2011, 17, 540–545.
12. H. Li, Z. Bian, J. Zhu, Y. Huo, H. Li and Y. Lu, J. Am. Chem. Soc., 2007, 129, 4538–4539.
13. R. J. Dillon, J. B. Joo, F. Zaera, Y. D. Yin and C. J. Bardeen, Phys. Chem. Chem. Phys., 2013, 15, 1488–1496.
14. Z. K. Zheng, B. B. Huang, X. Y. Qin, X. Y. Zhang, Y. Dai and M. H. Whangbo, J. Mater. Chem., 2011, 21, 9079–9087.
15. R. Chalasani and S. Vasudevan, ACS Nano, 2013, 7, 4093–4104.
16. M. M. Ye, Q. Zhang, Y. X. Hu, J. P. Ge, Z. D. Lu, L. He, Z. L. Chen and Y. D. Yin, Chem.–Eur. J., 2010, 16, 6243–6250.
17. Q. Zhang, I. Lee, J. B. Joo, F. Zaera and Y. D. Yin, Acc. Chem. Res., 2013, 46, 1816–1824.
18. Q. Zhang, D. Q. Lima, I. Lee, F. Zaera, M. F. Chi and Y. D. Yin, Angew. Chem., Int. Ed., 2011, 50, 7088–7092.
19. T. Sakamoto, D. Nagao, M. Noba, H. Ishii and M. Konno, Langmuir, 2014, 30, 7244–7250.
20. H. Y. Liu, J. B. Joo, M. Dahl, L. S. Fu, Z. Z. Zeng and Y. D. Yin, Energy Environ. Sci., 2015, 8, 286–296.
21. J. B. Joo, Q. Zhang, M. Dahl, I. Lee, J. Goebl, F. Zaera and Y. D. Yin, Energy Environ. Sci., 2012, 5, 6321–6327.
22. J. Chen, D. W. Wang, J. Qi, G. D. Li, F. Y. Zheng, S. X. Li, H. J. Zhao and Z. Y. Tang, Small, 2015, 11, 420–425.
23. T. Sakamoto, D. Nagao, M. Noba, H. Ishii and M. Konno, Langmuir, 2014, 30, 7244–7250.
24. N. Li, Q. Zhang, J. Liu, J. Joo, A. Lee, Y. Gan and Y. D. Yin, Chem. Commun., 2013, 49, 5135–5137.
25. N. R. Jana, L. Gearheart and C. J. Murphy, Langmuir, 2001, 17, 6782–6786.
26. J. B. Joo, Q. Zhang, I. Lee, M. Dahl, F. Zaera and Y. D. Yin, Adv. Funct. Mater., 2012, 22, 166–174.
27. J. Liu, S. Z. Qiao, H. Liu, J. Chen, A. Orpe, D. Y. Zhao and G. Q. Lu, Angew. Chem., Int. Ed., 2011, 50, 5947–5951.
28. W. Li, J. P. Yang, Z. X. Wu, J. X. Wang, B. Li, S. S. Feng, Y. H. Deng, F. Zhang and D. Y. Zhao, J. Am. Chem. Soc., 2012, 134, 11864–11867.
29. C. Gao, J. Vuong, Q. Zhang, Y. Liu and Y. Yin, Nanoscale, 2012, 4, 2875–2878.
30. D. W. Ding, K. Liu, S. N. He, C. B. Gao and Y. D. Yin, Nano Lett., 2014, 14, 6731–6736.
31. Q. J. Xiang, J. G. Yu, B. Cheng and H. C. Ong, Chem.–Asian J., 2010, 5, 1466–1474.
32. Y. Yu, C. Y. Cao, Z. Chen, H. Liu, P. Li, Z. F. Dou and W. G. Song, Chem. Commun., 2013, 49, 3116–3118.
33. M. S. A. S. Shah, A. R. Park, K. Zhang, J. H. Park and P. J. Yoo, ACS Appl. Mater. Interfaces, 2012, 4, 3893–3901.
34. Y. H. He, D. Z. Li, G. C. Xiao, W. Chen, Y. B. Chen, M. Sun, H. J. Huang and X. Z. Fu, J. Phys. Chem. C, 2009, 113, 5254–5262.
35. K. Ishibashi, A. Fujishima, T. Watanabe and K. Hashimoto, Electrochem. Commun., 2000, 2, 207–210.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra26646g

---