Introduction

The energy crisis has received much attention in the past few decades because of an increasing demand for energy and the gradual exhaustion of non-renewable energy resources1,2. Clean and renewable energy, such as solar energy, is thus very essential to the entire world. Photocatalysis, a photochemical approach to use solar energy, has been studied extensively since the discovery of water splitting using TiO2 in 19723. The photocatalytic H2 production from water splitting under solar irradiation has attracted huge attention because of its potential as one of the clean and environmentally friendly strategies to solve the energy crisis4,5,6,7. In addition to water splitting, photocatalysts have found many other applications, such as photocatalyzed degradation of organic pollutants8,9,10. Therefore, developing efficient photocatalysts is of high significance. Since the discovery of splitting of water by TiO2, TiO2 has been the subject of extensive investigations due to its high efficiency, low cost, nontoxicity, and high stability over the past few decades11,12,13,14,15,16.

TiO2, especially the anatase form of TiO2 that has more oxygen vacancies than the rutile phase, has a high photocatalytic efficiency under UV light17. For anatase TiO2, its (001) facet has a higher surface energy of 0.90 J/m2 than other facets, thus a highest chemical activity18,19,20. However, TiO2 has two major weaknesses: (1) TiO2 (anatase) has a band gap of 3.2 eV, and thus can only absorb UV light for photocatalysis21,22. UV light accounts for less than 5% of the total sunlight and thus a wide range of the solar spectrum is wasted during the process, which greatly restricts the practical applications of TiO2 under sunlight14,21,23. (2) The recombination of electrons and holes are very likely to happen after being excited by photons in TiO2 photocatalyzed reactions24,25. Therefore, the modifications of TiO2 that can address the above two issues, i.e., narrowing the band gap to broaden the working range of TiO2 to the visible-light range and separating the excitons more efficiently, have been studied extensively recently26,27,28.

Doping is a common approach to increase the photocatalytic efficiency of semiconductors under solar light29,30,31,32. For example, by doping metal ions into the semiconductor, impurity levels could be introduced into the forbidden band of the semiconductor. The recombination of electrons and holes can be restrained and relatively wider band gap photocatalysts can stay active in the visible-light region accordingly33,34. Doping non-metal ions (N, S, C, F, etc.) is another way to enhance the visible light efficiency by forming a new valance band, which is higher than the original one. In this way, the band gap would be narrowed and more visible light energy could be used35,36,37,38,39,40. Liu et al. reported TiO2 sheets doped with sulfur using TiS2 as a precursor, leading to significantly enhanced photocatalytic efficiency under visible light41. Fan et al. also introduced sulfur (using thiourea as a precursor) into TiO2 nanoparticle systems and obtained good results for the degradation of methyl orange (MO) under visible light42. These results show that doping sulfur is an efficient method to improve the photocatalytic activity of TiO2.

Graphene, a single layer graphite sheet composed of sp2-hybridized carbon atoms, has attracted significant interest due to its outstanding physiochemical properties, including high surface area, thermal and electrical conductivity43,44,45. Therefore, graphene can potentially promote electron-hole pairs to separate more effectively and thus improve photocatalytic efficiency because of its superior electron mobility46. Consequently, graphene-based composites have attracted much attention in the area of photocatalysis47,48. Reduced graphene oxide (rGO), consisting of graphene domains and interspersed with a few oxygen-containing functionalities, has been adopted to improve the electron mobility of photocatalysts49,50,51,52.

In this research, we designed the doping of anatase TiO2 with sulfur in order to narrow the band gap and achieve a red-shift on adsorption to potentially make use of the energy of visible light. Meanwhile, rGO was added into the doped semiconductor to expedite the transport of electrons and decrease the recombination of excitons. A facile one-pot hydrothermal synthesis approach was designed to prepare TiO2 nanosheets doped with sulfur on rGO via a single-step reaction.

Experimental

Materials

Graphene oxide (GO) was synthesized by the modified Hummers’ method53. Titanium n-butoxide (>99%), hydrofluoric acid (48–51%), thiourea (>99%), and methyl blue were purchased from Alfa Aesar and used as received without further purification.

Synthesis of TiO2 nanosheets doped with sulfur on rGO (TiO2-S/rGO hybrid)

A sample of 39.5 mg GO was mixed with 13.0 mL anhydrous ethanol in a beaker. After ultrasonication treatment for 3 hours, 0.076 g thiourea was added into the beaker and stirred for 5 minutes. Then 1.75 mL titanium n-butoxide and 0.2 mL hydrofluoric acid (48–51%) was added into the beaker while stirring. After stirring for 3 minutes, the dispersion was transferred into a Teflon® lined autoclave. The hydrothermal reaction was carried out for 24 hours at 180 °C. After reaction, the system was filtered and washed with deionized water for 3 times, then dried in vacuum at room temperature for 24 hours. A hybrid of TiO2 nanosheets doped with sulfur on rGO (TiO2-S/rGO hybrid) was obtained. The mass percentage of TiO2 in the hybrid was estimated to be ca. 95%.

Three control samples, including un-doped TiO2, sulfur doped TiO2 (TiO2-S) without rGO, un-doped TiO2 on rGO (TiO2/rGO), were synthesized using the same method under the same conditions and corresponding formulations.

Characterization

X-ray diffraction (XRD) patterns were recorded using a D5 Focus diffractometer (Bruker) with Cu Kα radiation. Raman spectra (Renishaw System 2000) were collected to characterize the rGO in the hybrids. The morphology and structure of the hybrids were studied by scanning electron microscopy (SEM, JEOL JSM-6335F field emission SEM with an accelerating voltage of 10 kV), transmission electron microscopy (TEM, FEI Tecnai T12 with an accelerating voltage of 120 kV), and high resolution TEM (JEOL 2010F field emission TEM with an accelerating voltage of 200 kV). The SEM samples were prepared by gently depositing the catalyst powders on a conductive tape. The TEM samples were prepared by dispersing catalyst powders in ethanol and depositing a drop of the suspension onto a carbon-coated copper grid. The elemental analysis was conducted using energy-dispersive spectroscopy (EDS) in the TEM. X-ray photoelectron spectroscopy (XPS, Thermo Scientific) was performed using a monochromated Al Kα X-ray source (1486.6 eV). The lateral dimension of the TiO2 nanosheets was estimated with a Nano Measurer (version 1.2). The band gap and reflectance were analyzed using UV-Vis diffuse reflectance spectroscopy (SHIMADZU UV-2450 Spectrophotometer).

Evaluation of photocatalytic activity

The photocatalytic activities of all the samples were evaluated by the degradation of organic dyes (10 mg/L methyl blue) under simulated sunlight, which was obtained from a 300 W Xenon lamp (Newport Corporation, Irvine, CA, USA). Control experiments under visible light and UV light were also carried out to better understand the mechanism of the photocatalyzed reactions. The simulated sunlight (200–20000 nm), visible light (420–630 nm), and UV light (280–400 nm) were obtained by using different mirrors on the same instrument. In each experiment, 0.010 g catalyst (TiO2-S/rGO hybrid, TiO2, TiO2-S, or TiO2/rGO) was dispersed in 50.0 mL (10 mg/L) methyl blue (MB) solution. After stirring in the dark for 30 minutes to reach adsorption equilibrium, the light was turned on for testing. A sample of 3.0 mL well-dispersed suspension containing photocatalyst particles was collected periodically. The collected suspension was centrifuged to remove the photocatalyst particles and the obtained supernatant was analyzed by UV-Vis spectroscopy (Varian Cary 5000 UV-Vis NIR). The absorbance of the characteristic peak at 664 nm is proportional to the concentration of MB54 according to the Beer-Lambert Law.

Results and Discussion

Structure and morphology of TiO2-S/rGO hybrid

A one-pot hydrothermal approach was developed with an aim to synthesize TiO2-S/rGO hybrids via a single-step reaction, during which anatase phase TiO2 doped with S was synthesized and deposited on rGO, which was reduced from graphene oxide during the hydrothermal reaction20,52,55,56.

Figure 1 shows the XRD patterns of TiO2-S/rGO hybrid and the control samples. The patterns confirmed that the TiO2 in TiO2-S/rGO hybrid and all the control samples is in the anatase phase57. The results confirmed that anatase phase TiO2 was successfully synthesized through the single-step hydrothermal method.

Figure 1
figure 1

XRD patterns of TiO2-S/rGO hybrid and the control samples TiO2, TiO2-S, TiO2/rGO.

Full size image

Among all the reported precursors for sulfur doping, TiS2, CS2, and thiourea are the most common and effective ones41,42,58,59,60,61,62,63. Considering the possible future practical application of these hybrids, thiourea was adopted to dope TiO2 in this project mainly because of its low cost. Energy-dispersive spectroscopy of the TiO2-S/rGO hybrid is shown in Fig. 2. Multiple elements, including carbon, oxygen, sulfur, and titanium were detected in the hybrid. Titanium is from TiO2 nanosheets, oxygen is from TiO2 nanosheets and rGO, while carbon is from rGO. Cu was also detected, which is from the Cu grid supporting the hybrid sample. The presence of sulfur confirmed the successful doping of TiO2 by S, and the doping ratio of sulfur was estimated to be ca. 1.5% according to elemental quantitative analyses.

Figure 2
figure 2

Energy-dispersive spectroscopy of TiO2-S/rGO hybrid.

Full size image

Figure 3 presents the XPS spectra of the synthesized samples. S element was clearly detected from the two S doped samples at 164.0 eV and 165.2 eV (Fig. 3a), representing 2p3/2 and 2p1/2 states of sulfur, respectively64,65. As for Ti (Fig. 3b), the Ti2p3/2 peak of the doped samples centered at ca. 459.6 eV, higher than that of the non-doped samples located at ca. 459.0 eV. Such a peak shift is expected and supports the successful doping of S into TiO2. For the XPS of carbon (C1s, Fig. 3c), the GO sample showed two peaks at 284.8 and 286.9 eV, attributing to sp2 C and the C bonded with O (C-O, etc.), respectively20. After hydrothermal reaction, the samples containing rGO exhibited only one peak at 284.8 eV, while the other peak at 286.9 eV disappeared. This suggested that during the hydrothermal reaction, GO was successfully reduced to rGO. It also indicates that no sulfur was doped into rGO during the hydrothermal reaction, since there is no peak reflecting the doping of sulfur.

Figure 3
figure 3

High resolution XPS spectra of S (a), Ti (b), and C (c) of the synthesized samples.

Full size image

Figure 4 shows the UV-Vis diffuse reflectance spectra of the TiO2-S/rGO hybrid and the control samples. The spectra show that the absorbance of the samples decreased sharply at ca. 400 nm. Among all the samples, the ones with rGO had a higher absorbance at higher wavelength (>400 nm). This result suggests that the samples containing rGO can absorb more visible light, which is expected. The TiO2 samples doped with sulfur had a higher absorption limit at ca. 411 nm, while the un-doped samples exhibited a lower absorption limit at ca. 372 nm. This indicates that the sulfur doped TiO2 indeed had a smaller band gap energy of ca. 3.0 eV while the un-doped samples have a band gap energy of ca. 3.3 eV (band gap energy is roughly estimated based on the absorption limit wavelength)66,67, which is slightly larger than the band gap of micron-sized TiO2 (3.2 eV) due to the quantum size effects of the nano-sized TiO2 (which will be discussed in Fig. 5d below) in this work68,69. Overall, the UV-Vis diffuse reflectance spectroscopy characterization suggests that a narrower band gap was achieved by doping TiO2 with sulfur as designed.

Figure 4
figure 4

UV-Vis diffuse reflectance spectra of TiO2-S/rGO hybrid and the control samples TiO2, TiO2-S, TiO2-rGO.

Full size image
Figure 5
figure 5

(a) SEM and (b) TEM images of TiO2-S/rGO hybrid, and (c) high resolution TEM image of TiO2 nanosheets; (d) size distribution of TiO2 nanosheets. The inset in (c) shows the FFT pattern of the same area.

Full size image

The morphology and structure of the TiO2-S/rGO hybrid was characterized by SEM and TEM. Figure 5a shows the SEM image of the TiO2-S/rGO hybrid, which is composed of an rGO film with folds and many TiO2 nanosheets attached. Because of their similar contrast under SEM, TiO2 nanosheets cannot be clearly differentiated from the rGO film under SEM. However, under TEM (Fig. 5b), in contrast to the very thin and thus very light rGO film, TiO2 nanosheets became clearly visible. They were uniformly deposited on an rGO film, which possesses folds and wrinkles, as indicated by the arrows in Fig. 5b. Such a uniform deposition is expected to be very beneficial for photocatalytic applications.

The high resolution TEM image of a randomly selected TiO2 nanosheet (Fig. 5c) shows that the lattice spacing is 0.19 nm, which corresponds to (200) or (020) planes. As such, the vertical facet facing up are (001) facets. The fast Fourier transform (FFT) pattern shown in the inset of Fig. 5c also suggests that the top facets are (001) facets55,70. This characterization result confirmed that with the assistance of HF70, TiO2 nanosheets with a large and well exposed (001) facet were successfully synthesized, which is very critical for photocatalytic reactions because the (001) facet is the most active facet for photocatalysis18,19,20. Overall, the TiO2 nanosheets have a square-like shape with an average lateral dimension of ca. 7.3 nm (Fig. 5d). It was reported that quantum size effect occurred in TiO2 when the particles were smaller than 5 nm71,72. While the synthesized TiO2 nanosheets have an average diameter of 7.3 nm, smaller nanosheets (including some < 5 nm) exists. In addition, the thickness of such nanosheets should be less than 5 nm. As a result, the band gap was shifted slightly from 3.2 to 3.3 eV.

Photocatalytic activity evaluation of TiO2-S/rGO hybrid

The photocatalytic activity of the TiO2-S/rGO hybrid was assessed by the widely accepted photo-degradation reaction of organic dye, MB. Figure 6 displays the degradation result of MB under simulated sunlight using different catalysts. Apparently, TiO2-S/rGO hybrid exhibited the highest photocatalytic activity compared to the control samples, which is believed to be due to the narrowed band gap of TiO2 and improved electron mobility, leading to the adsorption of visible light and minimized recombination of electrons and holes, respectively, both of which are beneficial for improving photocatalytic activity. Un-doped TiO2 nanosheets had a relatively low photocatalytic activity, mainly because of its relatively wide band gap and thus low efficiency under sunlight. Adding rGO into the system could help the transfer of electrons and thus enhance the separation of electron-hole pairs, leading to improved photocatalytic efficiency. As discussed above, doping sulfur into TiO2 nanosheets helped to narrow the band gap, therefore the sulfur doped TiO2 nanosheets could absorb a wider range of sunlight (the visible light part) thus resulting in a higher photocatalytic efficiency. As a result, the TiO2-S/rGO hybrid combining both sulfur doping and rGO exhibited the highest photocatalytic performance.

Figure 6
figure 6

Degradation of MB using TiO2-S/G hybrid under simulated sunlight.

Full size image

As shown in Fig. 6, the concentration-time curve of the degradation reaction of MB catalyzed by TiO2 is virtually linear. After the excitation of the electron-hole pairs on TiO2 nanosheets, the photon generated electrons will react with oxygen in the water to form O2•−, while the holes will react with hydroxyl groups in the water to form hydroxyl radicals. Then superoxide radicals will react with protons in water and more hydroxyl radicals can be formed. Those hydroxyl radicals can react with organic dye molecules to oxidize them. On the other hand, holes can oxidize organic dye molecules directly and generate degradation products73. Since the recombination of electrons and holes is rapid74, the degradation reaction rate mainly depends on the concentration of electrons and holes left without recombination. Therefore, the concentration of MB in the solution barely affects the degradation reaction rate catalyzed by TiO2, resulting in a virtually linear relationship of concentration versus time.

By adding rGO into the system, photon generated electrons can be quickly transferred to rGO thus reducing the recombination rate. Moreover, organic dyes like MB can be easily adsorbed onto the surface of rGO via π-π interactions, in comparison to the very limited adsorption of MB on TiO2 nanosheets. This adsorption process significantly increased the concentration of MB molecules close to the catalyst surface, which contributes to improve the degradation rate46. As such, a higher concentration of MB at the beginning of the reaction will lead to a higher reaction rate since more MB molecules will be adsorbed onto rGO. That is why the degradation rate of TiO2/rGO was faster than TiO2 in the beginning of the reaction. As the reaction progresses, the concentration of MB decreased, and therefore the accumulation effect by rGO turned out to be less and less significant. Since the same amount of catalyst was used for each reaction, the TiO2/rGO catalyst contained less TiO2 compared to the pure TiO2 catalyst, leading to a relatively low reaction rate at the late stages of reaction.

By doping sulfur into TiO2 nanosheets, besides the UV light, the photocatalyst could also absorb visible light. As a result, more electrons and holes can be generated under the sunlight. In addition, more oxygen vacancies may be introduced by doping, which may further improve the photocatalytic efficiency75,76,77,78.

To better understand the photocatalytic mechanism of the synthesized catalysts, a series of control experiments under visible light only (~420–600 nm) were conducted, and the results are shown in Fig. 7. TiO2-S/rGO hybrid still exhibited the highest photocatalytic activity under visible light, followed by TiO2-S, TiO2, and with TiO2/rGO as the lowest. Compared to Fig. 5, the photocatalysts without doped sulfur exhibited a much lower reactivity under visible light than under sunlight. This suggests that doping sulfur into TiO2 contributes to narrowing of the band gap of TiO2 to take advantage of visible light. As shown in Fig. 6, TiO2/rGO hybrid exhibited a slightly lower photocatalytic efficiency than TiO2 nanosheets. This is probably because TiO2 cannot effectively absorb visible light to catalyze the reaction, and therefore the reaction rate is mainly determined by the amount of TiO2 catalyst present. Since the TiO2/rGO contained less TiO2 compared to the pure TiO2 catalyst, TiO2/rGO has a lower efficiency under the visible light as compared to the pure TiO2 nanosheets.

Figure 7
figure 7

Degradation of MB using different photocatalysts under visible light.

Full size image

To further confirm the photocatalytic mechanism of the synthesized catalysts as proposed and discussed above, the degradation reactions of MB catalyzed by various TiO2 containing catalysts under UV light only (~280–400 nm) were also carried out, and the results are shown in Fig. 8. The overall performance trend of the four catalysts under UV light are very similar to the results under sunlight as shown in Fig. 6, with two major differences: (1) a slower reaction for all the four catalysts, which is due to the absence of absorption of visible light, and (2) a better relative performance of the un-doped TiO2, which is simply because TiO2 is most efficient under UV light. These results in turn suggest that the doping of sulfur helped the absorption of visible light as designed, leading to a higher reaction efficiency.

Figure 8
figure 8

Degradation of MB using different photocatalysts under UV light.

Full size image

Conclusions

In summary, TiO2-S/rGO hybrid was synthesized using a one-pot hydrothermal method via a single-step reaction. Doping sulfur proved to be effective in narrowing the band gap of TiO2, thus significantly enhancing the photocatalytic reaction rate under visible light. The introduction of rGO helped expedite the electron transfer and make the electron-hole pairs separate more efficiently. This effect works regardless of the wavelength range. As such, the designed TiO2-S/rGO hybrid catalyst exhibited a high photocatalytic activity under simulated sunlight. Considering both the facile and scalable reaction to synthesize TiO2-S/rGO hybrids, and their excellent photocatalytic performance, such TiO2-S/rGO hybrids are expected to find practical applications in environmental and energy sectors.

Additional Information

How to cite this article: Wang, W. et al. Single-step One-pot Synthesis of TiO2 Nanosheets Doped with Sulfur on Reduced Graphene Oxide with Enhanced Photocatalytic Activity. Sci. Rep. 7, 46610; doi: 10.1038/srep46610 (2017).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.