Tiehu Han,
Huigang Wang* and
Xuming Zheng*
Department of Chemistry, Engineering Research Center for Eco-dyeing and Finishing of Textiles, MOE and Zhejiang Provincial Top Key Academic Discipline of Chemical Engineering and Technology, Zhejiang Sci-Tech University, Hangzhou 310018, China. E-mail: zdwhg@163.com; huigwang@uni-osnabrueck.de; Tel: +86-571-8684-3627
First published on 5th January 2016
Spindle-shaped nanoporous anatase TiO2 mesocrystals with exposed active {101} facets have been successfully prepared through a hydrothermal method with tetrabutyl titanate as the precursor. By a deposition–precipitation process, highly dispersed Au nanoparticles loaded on spindle-shaped mesoTiO2 exposed {101} facets, denoted as Aux/mesoTiO2, were firstly fabricated to establish close Schottky junctions to improve the visible light activity and the stability of Au on the catalyst surface. The photodegradation of methylene blue (MB) over Aux/mesoTiO2 was systematically investigated. The exposed active {101} facets together with the loaded Au nanoparticles dramatically enhanced the visible light photocatalytic activity of TiO2. The synergistic effect of the high intrinsic single-crystal-like nature of the anatase phase, the stability of gold and the strong interaction between Au and mesoTiO2 result in extraordinary photocatalytic stability of the catalyst. The detailed e− and h+ separation dynamics for the visible-light and UV-vis induced catalytic mechanisms were discussed.
As a novel class of TiO2 material, TiO2 mesocrystal has received rapidly increasing attention since anatase TiO2 mesocrystals (mesoTiO2) were first prepared by topotactic conversion from NH4TiOF3 mesocrystals in the presence of nonionic surfactants.16,17 Bian and co-workers reported that mesoTiO2 superstructures had significantly enhanced charge separation upon UV-light irradiation due to their remarkably long-lived charges.18 Hong and co-workers reported an experimental study of a new synthesis strategy for the formation of unique rutile TiO2 mesocrystals constructed from ultrathin nanowires in the absence of an additive. The rutile TiO2 mesocrystals were used for the first time as an electrode in LIBs and exhibited a large reversible lithium-ion charge–discharge capacity and excellent cyclic stability.19 However, the practical application of mesocrystals remains a great challenge, because their formation processes are poorly understood.
Recently, Au/TiO2, the representative “plasmonic photocatalyst”, has attracted much interest as a new type of visible light photocatalyst.20–23 Gold is a noble metal and does not undergo corrosion under photocatalytic conditions. In the photocatalytic process, the noble metal plays an important role: on one hand, due to surface plasmon resonance (SPR), gold nanoparticles (NPs) possess unique absorption in the whole visible region, which can be utilized to harvest visible light;20–23 on the other hand, the formation of a Schottky barrier between TiO2 and Au NPs inhibits the e–h pair recombination process.24 The photocatalytic reaction of Au/TiO2 mainly occurs on the surface of TiO2. The TiO2 surface can transfer the electrons from its conduction band (CB) to Au8 or accept the electrons from Au;25,26 this depends on whether the excitation occurs on TiO2 or on the surface plasmon band of Au.25,27 This obviously raises questions regarding the photocatalytic activity of mesoTiO2 loaded with Au NPs under visible light irradiation. However, the incorporation of plasmonic Au NPs onto mesoTiO2 has not been reported.
In this study, a simple deposition–precipitation (DP) method28,29 was used to deposit Au NPs on mesoTiO2 which was synthesized through a solvothermal method using tetrabutyl titanate (TBT) as the titanium source and acetic acid as the solvent.30,31 The photocatalytic activity of the Au/mesoTiO2 samples under simultaneous UV and visible light irradiation or visible light irradiation alone was evaluated by their capability to degrade methylene blue (MB). Furthermore, we also carefully investigated the influences of Au content on the microstructures and photocatalytic activity of the mesoTiO2 samples. It was found that superstructure-based Au/mesoTiO2 with suitable gold content has significantly enhanced photocatalytic activity.
![]() | ||
Fig. 1 FE-SEM images (a and b) and TEM image (c) of mesoTiO2. HRTEM image of mesoTiO2 (d) and selective area electron diffraction (SAED, inset d). FE-SEM and TEM images of Au5.0/mesoTiO2 (e and f). |
To investigate the surface composition and the chemical states of the elements in as-prepared Au/mesoTiO2, XPS studies were conducted for mesoTiO2 before and after Au loading. As shown in Fig. 2a, the wide-scan survey spectra of mesoTiO2 and Au/mesoTiO2 all contain O, Ti and C elements; the emergence of the C element can be attributed to residual carbon from the sample and adventitious hydrocarbon from the XPS instrument itself. The inset in (a) shows a comparison between the spectra of mesoTiO2 and Au/mesoTiO2 in the range of 0 to 100 eV; obviously, there are some weak peaks in the range of 80 to 100 eV for Au/mesoTiO2 which can be ascribed to Au4f. The Au4f core level spectrum, shown in Fig. 2b, is composed of two peaks at the binding energies of 83.3 and 87.2 eV, assigned to Au4f7/2 and Au4f5/2, respectively; these are in good agreement with the reported values of Au(0), suggesting that the Au species is present in the metallic state.33 However, the peak of Au (4f7/2) shifted slightly lower relative to that of free metallic Au(0) (∼83.8 eV). This difference indicates significant charge transfer from TiO2 to Au and thus confirms the strong Au/TiO2 interaction.24 A Schottky barrier is created between Au(0) and mesoTiO2. Furthermore, XPS was used to distinguish the surface change of mesoTiO2 before and after Au loading. Fig. 2c displays the XPS spectra of Ti2p of mesoTiO2 and Au/mesoTiO2; the binding energies of Ti2p3/2 and Ti2p1/2 are equal to 458.6 eV and 464.4 eV, respectively, suggesting the presence of Ti(IV) species.34 By comparison, there are no measurable changes in the peak positions for Ti2p in mesoTiO2 before and after Au loading. Meanwhile, the XPS spectra of O1s (Fig. 2d) shifts from 529.88 to 529.94 eV after Au NPs deposition, owing to the generation of surface oxygen vacancies.35
Several parameters influence the photocatalytic activity of plasmonic composite Au/mesoTiO2 photocatalysts, including gold particle size, morphology, and amount, as well as the interfacial contact between Au and titania. By adding an appropriate concentration of HAuCl4·3H2O (see Table 1), Au/mesoTiO2 photocatalysts with different weight percentages (wt%) of Au on mesoTiO2 were prepared using the DP method. The dominant exposed face of the spindle-shaped nanoporous anatase TiO2 mesocrystals is {101}, and the Au NPs do not change the anatase phase of mesoTiO2. Therefore, the dominant interfacial contact occurred between the Au nanoparticles and the {101} phase for all samples. Thus, the Au amount has a significant influence on the photocatalytic activity in this study.
Sample | C (HAuCl4·3H2O) (mM) | Phasea | SBETb (m2 g−1) | Dpc (nm) |
---|---|---|---|---|
a A: anatase, R: rutile.b BET surface.c Pore diameter. | ||||
MesoTiO2 | 0 | A | 73.7 | 6.9 |
Au1.0/mesoTiO2 | 0.51 | A | 51.8 | 6.8 |
Au3.0/mesoTiO2 | 1.57 | A | 50.3 | 7.0 |
Au5.0/mesoTiO2 | 2.67 | A | 68.3 | 7.2 |
Au7.6/mesoTiO2 | 4.19 | A | 59.2 | 7.3 |
P25 | 0 | A and R | 57 | — |
Au3.0/P25 | 1.57 | A and R | 54 | 7.5 |
XRD was used to identify the phase structures of the synthesized samples. Fig. 3 shows the X-ray diffraction patterns of pure mesoTiO2 and Au/mesoTiO2, in which all diffraction peaks of the calcined materials (with or without gold incorporation) can be indexed as anatase TiO2 with standard values in agreement with JCPDS card no. 21-1272. Oddly, no crystalline Au diffraction peaks were observed in any of the Au/mesoTiO2 composites. This may be ascribed to the fact that the gold is highly dispersed in the mesoTiO2 porous structures, and the low loading quantity of Au is beyond the XRD detection limit.36 It should be noted that single crystalline phase Au can be formed through this method.
To further identify the phases and the crystallinity of the samples, Raman studies were performed in the range of 100 to 1000 cm−1 (shown in Fig. 4). The Raman peaks at 143.3, 394, 513, and 636 cm−1 can be assigned to the Raman-active modes of anatase with Eg, B1g, A1g, and Eg symmetries, respectively.37 It is interesting to observe that the Eg peak at 143.3 cm−1 gradually shifted to a higher wavenumber as the Au content increased; however, the Eg peak at 636 cm−1 shifted to the opposite direction. This indicates that there was an interaction between the Au and mesoTiO2, and the created crystalline defects within the mesoTiO2 increase with increasing Au content.38 The crystalline defects affect the characteristic vibrational frequency of the anatase TiO2; it can act as a trap to capture photoelectrons, which makes a contribution to inhibiting the charge recombination. Thus, the XRD and Raman characterization of mesoTiO2 and Au/mesoTiO2 demonstrate the high crystallinity of the titania materials and the presence of anatase phase.
The light absorption properties of the composites were studied by UV-vis spectroscopy, and the effect of Au loading content on the UV-vis absorption properties are revealed in Fig. 5. This figure shows that the absorption of blank mesoTiO2 is only located in the ultraviolet (UV) region below 370 nm, whereas Au/mesoTiO2 shows two absorption bands, with the largest absorption edge located near 370 nm and the second near 558 nm. The presence of two absorption bands indicates two step transitions in the band gap. The pronounced low-energy absorption band at 558 nm is in good agreement with the reported values for Au nanoparticles and may be ascribed to the typical surface plasmon resonance (SPR) of Au NPs.25,39,40
It is clear that with the increase of Au content from 1.0 wt% to 7.6 wt%, the plasmon band intensity of Au/mesoTiO2 increased accordingly, while the light absorption of Au/mesoTiO2 in the UV region is the same as that of mesoTiO2 without Au. The UV-vis spectroscopy results indicate that Au/mesoTiO2 has significantly enhanced visible light absorption and can be photoexcited by visible light irradiation, by which electron–hole pairs can be generated; also, the two step-transition guarantees the prolonged separation lifetime of the electron–hole pairs, and thus, improved photoexcited performance can be expected.
It is known that the peak position and shape of the SPR absorption band are sensitive to Au particle size and morphology.41 In Fig. 5, no red-shifting phenomenon is observed, suggesting that all the Au/mesoTiO2 catalysts have similar Au NP sizes (8 nm).
The Brunauer–Emmett–Teller (BET) specific surface areas (SBET) and pore structures of mesoTiO2 and the Au/mesoTiO2 composites were investigated using N2 adsorption–desorption measurements at 77 K. All samples exhibit a characteristic type IV isotherm behavior with H2 hysteresis (Fig. 6), corresponding to mesoporous materials with ink-bottle structures. The pore distributions of all the samples are shown in the inset of Fig. 6. All the samples display similar narrow pore-size distributions, centered at about 4.0 to 6.0 nm. Detailed surface area and pore structure information are listed in Table 1. Usually, the larger the surface area of a photocatalyst, the more it promotes the adsorption of organic pollutants, which will result in a difference in the final photodegradation efficiency.
![]() | ||
Fig. 6 N2 adsorption–desorption isotherms and pore diameter distribution (inset) of undoped mesoTiO2 and the Au/mesoTiO2 nanocomposites at 77 K. |
The photocatalytic degradation of MB in the aqueous phase was selected as the probe reaction to evaluate the photocatalytic activity of mesoTiO2 before and after Au loading. The change of methylene blue concentration as a function of illumination time is shown in Fig. 7a and b. Under simultaneous UV and visible light irradiation, only 46% degradation was observed over mesoTiO2 within 10 minutes, as shown in Fig. 7a. However, the Au/mesoTiO2 samples with different Au contents demonstrated higher photodegradation efficiency and completely degraded MB within 10 minutes. It can be observed that the MB degradation efficiency continuously increased in the first 5 minutes as the Au loading content increased from 1 to 3 wt%. Further increasing the Au content, however, decreased the catalytic activity. The same rule was observed upon visible light irradiation (400 to 780 nm); however, an irradiation time of 3 hours is required for quantitative degradation. Among all these visible light photocatalysts, the highest activity was noted for Au3.0/mesoTiO2, which had a degradation efficiency of MB near to 100%. On the basis of a simplified Langmuir–Hinshelwood model, the linear relationship of ln(C/C0) versus time (see Fig. 7c) upon visible light irradiation indicates that MB degradation follows pseudo first order kinetics; the apparent rate constant (k) shown in Fig. 7d was calculated from the plot of ln(C/C0) vs. time. The highest apparent rate constant, obtained for Au3.0/mesoTiO2, is 1.739 × 10−2 min, which shows a 3.80, 1.31, 1.09 and 3.03-fold photocatalytic activity improvement over mesoTiO2, Au1.0/mesoTiO2, Au5.0/mesoTiO2 and Au7.6/mesoTiO2, respectively.
The stability of a photocatalyst is very important for practical applications; thus, a durability test was performed to confirm the stability of the Au3.0/mesoTiO2 photocatalyst, which achieved the highest performance. According to Fig. 8a, the MB photodegradation efficiency changed slightly after 5 cycles of the experiment under identical conditions, indicating that the photocatalyst has superior photocatalytic stability. Moreover, the crystallization was well maintained after the recycling test compared to that of Au3.0/mesoTiO2 before the reaction, which is confirmed by the results shown in Fig. 8b. This can be attributed to the synergistic effect of the high intrinsic single-crystal-like nature of the anatase phase, the stability of the gold, and the strong interaction between Au and mesoTiO2.
![]() | ||
Fig. 8 Durability study over Au3.0/mesoTiO2 for MB photodegradation under visible light irradiation (a); XRD patterns of Au3.0/mesoTiO2 before and after the recycling reaction. |
To investigate the main active species during the photodegradation process, additional examination was carried out via dissolving different trapping agents in the reaction solution before light irradiation. Under UV-visible light irradiation, as shown in Fig. 9a, the MB degradation was significantly suppressed when benzoquinone (BQ), a scavenger for ˙O2− radicals, was added to the reaction system.42 However, other radical scavengers presented no influence on photoactivity. This indicated that the ˙O2− radical was the main active species during the photodegradation process. It is interesting to note that, under visible light irradiation, the results were different, as shown in Fig. 9b. When the photo-generated holes and electrons were trapped with EDTA and AgNO3, respectively, the degradation was moderately suppressed. Similarly, when BQ was added to the reaction system, a weaker decrease of the degradation rate was also observed, indicating that ˙O2− radical was not the only active species under these conditions. Regardless of the kind of light excitation, no significant changes occurred in dye degradation when an ˙OH scavenger, TBA, was added to the degradation system.43 Accordingly, a possible reaction process can be proposed as follows. The mechanisms for the photocatalytic activity of Au/mesoTiO2 under UV light excitation and visible light excitation are different. As is demonstrated by XPS, Au and mesoTiO2 form a Schottky junction; the metal Au acts as the anode, and the n-type mesoTiO2 acts as the cathode. Under simultaneous UV and visible light irradiation (Scheme 1, left part), the electrons of TiO2 are rapidly promoted from the valence band (VB) to the conduction band (CB), while the holes remain in the VB (eqn (1)). Electrons can be readily transferred from CB to metallic Au through a Schottky junction (eqn (2)), which extends the lifetime of the electrons. Then the electrons reduce O2 adsorbed on the Au surface to form ˙O2− and further generate ˙HO2 (eqn (3) and (4)). Finally, MB molecules adsorbed on the Au surface are oxidized by ˙O2− and ˙HO2 together. Meanwhile, the holes can initiate the direct oxidation of MB molecules (eqn (5)).
MesoTiO2 + hν (≥Eg) → mesoTiO2(ecb− + hvb+) | (1) |
MesoTiO2(ecb−) + Au → mesoTiO2–Au(e−) | (2) |
MesoTiO2–Au(e−) + O2 → mesoTiO2–Au + ˙O2− | (3) |
˙O2− + H2O → ˙HO2 + OH− | (4) |
h+, ˙O2− or ˙HO2 + MB → degraded product | (5) |
![]() | ||
Scheme 1 Proposed mechanism for the photocatalytic activity of Au/mesoTiO2 under UV light excitation (left) and upon excitation of the gold surface plasmon band (right). |
Under visible light irradiation, mesoTiO2 exhibited low photocatalytic ability due to its slight absorption “tail” (Fig. 5). Once Au NPs are incorporated into mesoTiO2, upon visible light irradiation, as is shown in Scheme 1 (right part), intense SPR-enhanced EM fields and resonant photon scattering are generated on the Au NP surface, significantly increasing the yield of interfacial “hot electrons” with a higher potential energy than the Schottky barrier height at the interface (eqn (6));44 the position of the Fermi level will be shifted closer to the CB of mesoTiO2.45 Subsequently, the “hot electrons” are transferred to the CB of mesoTiO2 (eqn (7)). The Schottky barrier at the interface also helps the transferred “hot” electrons accumulate in the TiO2 CB, preventing them from traveling back to the Au NPs. Since no holes are generated in the valence band (VB) of TiO2, the transferred “hot electrons” in the TiO2 CB should have much longer lifetimes, offering more probability to reduce the O2 adsorbed on the mesoTiO2 surface and the consequent MB degradation.
It is known that the reduction potential of the electron scavenger Ag+ (+0.80 V vs. NHE at pH 7) is much higher than the TiO2 CB minimum (−0.1 V), while lower than that of Au+ (+1.70 V). Thus, Ag+ can easily be reduced by the transferred electrons within the TiO2 CB minimum but cannot be reduced by the electrons on Au. We observed this phenomenon in our experiments; MB degradation was not suppressed by the addition of electron scavengers under UV-vis light irradiation but was suppressed under visible light irradiation. Under visible light irradiation, the h+ remaining on Au greatly facilitates its chelation by EDTA, which is consistent with the experimental phenomenon, shown in Fig. 9b, that MB degradation was obviously suppressed by the addition of EDTA. The photoreaction process is very quick,46 and ˙O2− radicals, e− and h+ are simultaneously responsible for the photodegradation of MB under visible light irradiation. Obviously, when the abovementioned competition occurs, the h+ will mainly take the charge of the oxidation of MB molecules (400 to 780 nm).
Au + hν (≥Eg) → Au(e− + h+) | (6) |
Au(e−) + mesoTiO2 → Au–mesoTiO2(ecb−) | (7) |
This journal is © The Royal Society of Chemistry 2016 |