Introduction

As a member of layered two-dimensional material, molybdenum disulfide (MoS2) is promising for the applications in energy and environment1,2,3,4,5,6,7,8,9,10. The MoS2 nanosheets could be achieved by break the interlayer van der Waals forces. The band gap of MoS2 nanosheets is seriously depended on its layer number, which is varied from 1.3 (bulk) to 1.8 eV (monolayer)11,12,13,14. Therefore, the few-layered MoS2 could be used as an efficient visible light harvester. Meanwhile, the two-dimensional structure provides large contact interface and efficient charge transfer, as a result, the layered MoS2 nanosheets have been regarded as a low-cost co-catalyst candidate resently15,16,17,18,19,20,21,22. TiO2 is a wide band gap (3.6 eV) semiconductor and has exhibited potential in photoelectrochemical (PEC) water splitting and photocatalytic applications23,24,25,26,27,28,29. The narrow band gap of MoS2 can broaden the visible-light response. Additionally, the interface charge transfer between MoS2-TiO2 hetero-junction would accelerate the charge separation and enhance photocatalytic activity and increase the hydrogen generation28,29,30,31,32,33,34,35,36.

Gold nanoparticles (NPs) supporting tunable surface plasmon resonance in a wide region have been used for various light-matter interaction enhancement37,38,39,40,41, in which the main mechanism are broadening light-harvesting region and facilitating the charge separation42,43,44,45,46,47. Yung-Jung Hsu et al. have reported that the hot electrons in Au NPs can get over the Schottky barrier and be injected into the conduction band of the TiO2, which would supply additional charge carriers for photocatalytic reaction48. Xing-Hua Xia et al. also reported an efficient water splitting hydrogen evolution reaction of Au nanorods/MoS2 nanosheets hybrids through increase the carrier density in MoS2 by Au nanorods49.

In this paper, we report a three component hybrid (MoS2-TiO2)/Au including two-dimensional MoS2 nanosheets, self-ordered TiO2 nanorod arrays, and plasmonic Au nanorods. The microscopic structures and optical properties of (MoS2-TiO2)/Au are characterized. The photochemical activities of TiO2, MoS2-TiO2, and (MoS2-TiO2)/Au are comparatively investigated. The physical mechanisms of enhanced light-harvesting, hot electrons injection, and acceleration of separation of photo-excited charges are further discussed.

Results and Discussion

The preparation procedure of (MoS2-TiO2)/Au is shown in Fig. 1. The TiO2 nanorod arrays are firstly grown on the conductive FTO glass substrate. Then layered MoS2 nanosheets are deposited onto the TiO2. Finally, the as-prepared Au nanorods are introduced by a drop-casting method. As shown in Fig. 2a, the TiO2 nanorods are vertically grown from the FTO conductive glass. The average lateral dimension of TiO2 NRs is about 80 nm. Figure 2b shows the sheet-shaped MoS2 cover up the top of TiO2 nanorods and are also grown into the interspace of nanorod array. The estimated side-length of MoS2 nanosheets is in the range from hundreds of nanometers to micrometer-scale. The dimension and amount of MoS2 nanosheets can be controlled by the deposition reaction time. As the magnification TEM image shown in Fig. 2d, the locations of Au nanorods are randomly distributed on the MoS2-TiO2, including on the basal plane of MoS2 nanosheets, on the top-end and side-surface of TiO2 nanorods, and even on the junction of MoS2-TiO2.

Figure 1
figure 1

Schematic illustration of preparing (MoS2-TiO2)/Au nanocomposites.

Figure 2
figure 2

Microscopic structure evolution in prepration of (MoS2-TiO2)/Au. (ad) Top-view SEM images of TiO2 (a), MoS2-TiO2 (b), and (MoS2-TiO2)/Au (c,d). The arrows in (d) indictae the locations of Au nanorods.

For verifying the component in the hybrids, the HRTEM images and EDX analysis of the (MoS2-TiO2)/Au composites are shown in Fig. 3. The samples are extracted from the FTO glass and placed on the copper grids for TEM observation. The observed Au nanorods have the transverse size of 15 nm and the aspect ratio in the range of 3–4. The lattice fringes of an individual TiO2 nanorod with a spacing of 0.32 nm can be assigned to the (110) lattice planes of rutile TiO2. The MoS2 nanosheets show the lattice fringes with 0.23 nm spacing, corresponding to the (103) planes of MoS2. The EDX analysis of the prepared (MoS2-TiO2)/Au is presented in Fig. 3d. The composite mainly contains Ti and O, and the rest of the trace elements are S, Au, and Mo. The atomic ratios of Mo: S and Ti: O are both approximately 1: 2. In the XRD pattern (Fig. 3e), two sets of diffraction peaks are present, which are assigned to the TiO2 nanorod array phase (JCPDS No. 76-1939) and Au nanorods phase (JCPDS No. 04-0784).

Figure 3
figure 3

Component analysis of the (MoS2-TiO2)/Au composites. (ac) HRTEM images (d) EDX and (e) XRD pattern of (MoS2-TiO2)/Au composites.

Figure 4 displays the absorption spectra of pure TiO2, MoS2-TiO2, and (MoS2-TiO2)/Au. Pure TiO2 only absorbs UV light and exhibits an intense absorption edge before 400 nm, attributing to its band gap of 3.2 eV. The few-layered MoS2 nanosheets are reported to have two absorption bands near 400 nm and 600 nm in the visible region36, which are shown in the spectrum of MoS2-TiO2. In the experiment, the sample was tuned to yellowish-brown color when the MoS2 nanosheets were grown onto the TiO2. These results indicate the deposited MoS2 nanosheets have efficient light-harvesting in visible region. The absorption intensity around 700 nm of (MoS2-TiO2)/Au is obviously enhanced, which is attributed to the plasmon of Au nanorods.

Figure 4
figure 4

Absorption spectra of pure TiO2, MoS2-TiO2, and (MoS2-TiO2)/Au. The inset shows the absorption spectrum of as-prepared AuNRs.

The photon-electron conversion performance was performed by measuring the photocurrent response of three-electrode PEC cells with the hybrids as photoanode. Figure 5a shows the PEC It curves of the TiO2, MoS2-TiO2, and (MoS2-TiO2)/Au under the visible-light irradiation (wavelength >420 nm) with a bias of 0.6 V versus Ag/AgCl reference electrode. The electrolyte including Na2SO3 and Na2S solution can consume photo-excited holes on the photoanode. The photo-excited electrons are migrated to the Pt counter electrode through external bias circuit. As the arrows indicated in Fig. 5a, the light irradiation is switched ON/OFF for assessing the photocurrent responses. The average photocurrent densities of the three samples are plotted as bar charts in Fig. 5b. The current densities are 4.9, 18.9, 26.8 μA/cm2, for the samples of TiO2, MoS2-TiO2, (MoS2-TiO2)/Au, respectively. The current density of (MoS2-TiO2)/Au is 5.5 times that of TiO2 and 1.42 times that of MoS2-TiO2.

Figure 5
figure 5

Photocurrent responses of TiO2, MoS2-TiO2, (MoS2-TiO2)/Au electrodes recorded in 0.1 M Na2SO3 and Na2S aqueous solution under visible light by light-on and light-off cycles.

The pure TiO2 electrode shows a considerably low photocurrent density, because TiO2 has large band bap and only responds to UV light. The enhanced photocurrent response of MoS2-TiO2 electrode can be understood through two aspects of enhanced visible light absorption and accelerated photo-excited charge separation. As discussed in Fig. 4, MoS2 nanosheets exhibit efficient light absorption in visible region. The jungle-typed microstructure of TiO2 nanorod arrays could trap the incident light inside the arrays through multiple scatterings/reflections and guide the light pass through the MoS2 nanosheets multiply times, enhancing the visible light-harvesting. In addition, the band alignment between MoS2 and TiO2 is favorable for the electron transfer from the conduction band (CB) of MoS2 to the CB of TiO2 and suppresses the photogenerated carrier recombination of TiO2 effectively. Moreover, the inserted MoS2 nanosheets connect neighboring TiO2 nanorods and act as bridge routes which benefit the electron transfer along the TiO2 channel to the conductive substrate.

The highest photocurrent of (MoS2-TiO2)/Au electrode is benefits from the plasmon-enhanced light absorption and the plasmon-induced hot electron injections. In detail, the Au nanorods work as a reaction sensitizer and enhance the visible light absorption ability of MoS2. On the other hand, the Au nanorods have intense plasmon absorption and the plasmon-produced energetic electrons in the (MoS2-TiO2)/Au nanosystem could also contribute to the photon-to-electron conversion. The hot electrons can get over the Schottky barrier and be injected into the CB of MoS2 and TiO2.

Finally, the hot electron injection of Au nanorods, the enhanced visible light-harvesting and the accelerated charge separation in the (MoS2-TiO2)/Au hybrids is further demonstrated by testing the photocatalytic hydrogen generation. The H2 evolution rate of TiO2 and MoS2-TiO2 under visible light are barely observed, while that of (MoS2-TiO2)/Au is enhanced. Figure 6 shows the photocatalytic hydrogen generation under full spectrum, TiO2 alone shows a low photocatalytic activity with the H2 evolution rate of 48 μmol·h−1·g−1 because of the rapid recombination of electron-hole pairs. The introduction of MoS2 results in a significant improvement of photocatalytic H2 evolution rate to 125 μmol·h−1·g−1. The MoS2-TiO2 composite photocatalysts show enhanced photocatalytic activity because the layered MoS2 can help the charge separation also act as an efficient co-catalyst for H2 generation than TiO2. In the presence of a small amount of Au nanorods in the hybrid photocatalysts, the photocatalytic H2 evolution rate of (MoS2-TiO2)/Au hybrids is further enhanced to 190 μmol·h−1·g−1. The experimental result of photocatalytic hydrogen generation is consistent with that of the photocurrents under visible light. Figure S1 shows the photocatalytic hydrogen generation under visible light. The corresponding energy band structure and electrons transfer mechanism is schematically shown in the Fig. 7.

Figure 6
figure 6

Photocatalytic hydrogen production activities of TiO2, MoS2-TiO2, (MoS2-TiO2)/Au electrodes under full spectrum in the aqueous solution containing 20% methanol as sacrificial agents.

Figure 7
figure 7

Schematic illustration of energy band structure and electron-hole separation of (MoS2-TiO2)/Au nanocomposites in PEC cell (a) and in photocatalytic hydrogen generation (b).

Conclusion

In conclusion, we have prepared a composite of (MoS2-TiO2)/Au consisting of two-dimensional MoS2 nanosheets, self-ordered TiO2 nanorod arrays, and plasmonic Au nanorods. Acting as photoanode of PEC cells and photocatalysts for hydrogen generation, the current density of TiO2 is increased 2.8 times and the hydrogen generation rate is increased 2.6 time via the charge transfer from MoS2 nanosheets. Moreover, the PEC current density and hydrogen generation rate of MoS2-TiO2 is further enhanced 42% and 52% by plasmon resonance of Au nanorods. The intimate and large contact interface between AuNRs and MoS2-TiO2 leads to the efficient injection of hot electron, which plays a key factor in determining the high photocurrent response of (MoS2-TiO2)/Au. The efficient visible light absorption and the high carrier mobility of layered MoS2 nanosheets contribute the photocurrent response. In addition, the array-typed nanostructure can effectively trap incident light and the MoS2-TiO2 hetero-junction can lead to efficient photo-excited charge separation.

Methods

Materials Synthesis

Titanium butoxide (TBT, ≥99%), hydrochloric Acid (HCl, 37%), sodium molybdate (Na2MoO4∙2H2O, ≥99%), thioacetamide (TAA, ≥99%). All chemical materials were used without further purification.

Synthesis of TiO2 Nanorod Arrays on FTO

TiO2 nanorod arrays were fabricated on the FTO substrate through a hydrothermal method50. Before modification, the FTO substrates were washed with acetone, ethanol in an ultrasonic washer for 5 minutes. Then, 0.4 g of titanium butoxide was dissolved into 30 mL of 6 M HCl aqueous solution and then transferred into a Teflon-lined steel autoclave with a capacity of 50 mL. The FTO substrates were placed against the Teflon wall with the FTO side facing down. The autoclave was heated in an oven at 150 °C for 6 h and then cooled down to room temperature. The TiO2 nanorods were cleaned with deionized water and ethanol.

Synthesis of MoS2 on TiO2

0.3 g of sodium molybdate (Na2MoO4 · 2H2O) and 0.6 g of thioacetamide were added. After stirring for 5 minutes, the reaction solution was transferred into a 50 mL Teflon-lined stainless steel autoclave and kept in an electric oven at 200 °C for 10 h. The autoclave was then cooled down to room temperature in the oven.

Synthesis of (MoS2-TiO2)/Au

Au nanorods of various aspect ratios were synthesized using a seed-mediated growth method in aqueous solution51. Then, the as-prepared Au nanorods were dropped onto the FTO grown with the MoS2-TiO2 nanocomposites. The samples were thermally treated at 350 °C in N2 atmosphere for 0.5 h, and then dried at 70 °C for 10 h.

Characterizations

Transmission electron microscopy (TEM) and high-resolution TEM (HRTEM) images were taken on a JEOL 2100 F transmission electron microscope at an accelerating voltage of 200 kV. Energy-dispersive X-ray spectra (EDX) analysis was performed on an energy-dispersive X-ray spectrometer incorporated in the HRTEM. Scanning electron microscope (SEM) measurements were carried out with an FEI Sirion 200 scanning electron microscope operated at an accelerating voltage of 10.0 kV. Extinction spectra were recorded by the ultraviolet-visible-near infrared (UV-Vis-NIR) spectrophotometers (TU-1810 and Varian Cary 5000).

Photoelectrochemical Activity Measurement

A three-electrode configuration was adopted in a quartz cell on the VersaSTAT 3 electrochemical workstation (AMETEK, Inc., United States). A Pt plate and a commercially available Ag/AgCl electrode are used as the counter and reference electrodes respectively, and the sample modified FTO electrode was used as the work electrode. The 0.1 M Na2SO3 and Na2S aqueous solution was prepared to support electrolyte. The effective surface area of the work electrode was 1 × 2.5 cm2. Before measurement, the as-prepared samples of TiO2, MoS2-TiO2, (MoS2-TiO2)/Au were thermally treated at 350 °C in high-purity nitrogen atmosphere for 0.5 h. A 300 W Xenon lamp equipped with an ultraviolet cut-off filter (λ > 420 nm) was used as light source.

Photocatalytic H2 Evolution

Before measurement, the samples were dried at 70 °C for 10 h. The photocatalytic hydrogen evolution tests were conducted in a quartz reactor tube with a rubber septum. 20 mg photocatalyst powders were dispersed in 50 mL of aqueous solution containing 20% of methanol as sacrificial reagents. The system was evacuated by using a pump and the reaction solution was stirred for 30 min to remove any dissolved air. The light source was a 300 W Xenon lamp. The temperature of the suspension was maintained by an external water cooling system. The amount of hydrogen gas was automatically analyzed by an online gas chromatography (Tianmei GC-7806).