In Situ Growth of Metallic 1T-MoS2 on TiO2 Nanotubes with Improved Photocatalytic Performance.

1T-MoS2 is in situ grown on TiO2 nanotubes (TNTs) using a hydrothermal method, forming a 1T-MoS2@TNTs composite, which is confirmed by its physical characterization. The prepared composites show enhanced photocatalytic performance for the degradation of tetracycline hydrochloride under visible light, and the improved photocatalytic activity is closely related to the loaded amount of 1T-MoS2. Therein, 0.5 wt % 1T-MoS2@TNTs can degrade 57% in 1 h, which is the highest photocatalytic efficiency observed in experiments so far. It is speculated that the introduction of 1T-MoS2 may optimize light absorption and charge separation/transport. The active species are identified and the reaction mechanism is proposed here.

Introduction

Environmental pollution including water contamination is a serious problem threatening sustainable development. Antibiotic residue is one of the refractory pollutants in water, and diverse techniques have been evolved for efficient disposal.[1] Among these approaches, photocatalysis is one of the most attractive and booming methods because of its clean, nonpolluting, and efficient properties.[2] TiO2 is one of the most investigated photocatalysts; however, it experiences the intrinsic limits of a wide band gap and fast recombination of photogenerated carriers. One of the widely used approaches is to construct a heterojunction for facing these obstacles.[2]

Recently, transition-metal sulfide MoS2, a typical two-dimensional material with a graphene-like layered structure,[3] has attracted intense attention with wide applications.[4] Generally, MoS2 has several polymorphs including metallic 1T and semiconducting 2H and 3R phases. Previous studies have mainly focused on 2H-MoS2 due to its good stability, and the other two phases may convert into the 2H phase under specific conditions.[5] 2H-MoS2 has been widely used as a high-performance photocatalyst, and intensive attempts have been made to improve its photocatalytic activity by modification or exfoliation.[6] Additionally, MoS2 can work as an efficient cocatalyst by modifying other semiconductor photocatalysts such as TiO2, which is one powerful strategy to realize highly efficient photocatalytic performances[7] by tuning the light absorption and the separation/transport of carrier.[8]

As is known, metallic 1T-MoS2 has a single-layered S–Mo–S structure, in which each Mo atom is surrounded by six S atoms in an octahedral lattice.[9] Recently, 1T-MoS2 has become a promising candidate for a wide range of applications due to its plentiful active sites and high electronic conductivity (6 orders of magnitude higher than that of 2H-MoS2), exhibiting a better catalytic performance than the other semiconducting counterparts.[9] Lately, high-purity and stable 1T-MoS2 nanosheets were synthesized using a hydrothermal method showing an excellent activity. 1T-MoS2 may also work as a cocatalyst, which induces enhanced photocatalytic efficiency. Especially, its unique metallic feature may improve the separation/transport of carriers greatly,[10] which may lead to the potential replacement of Pt in practical applications.[11] However, the facile immobilization of 1T-MoS2 on other semiconductor photocatalysts with improved performance is still a challenge.

To date, considerable efforts have been made to optimize the MoS2 nanocomposites and expand their practical use.[12−14] For example, Zheng et al. applied a hierarchical MoS2 nanosheet on a TiO2 nanotube (TNT) array to enhance the photocatalytic and photocurrent performance.[12] Zhao et al. used the MoS2 quantum dots@TNTs composites to improve the visible-light-driven high-efficiency photocatalysis.[13] In this work, 1T-MoS2 is successfully grown on TNTs using an in situ hydrothermal method, which shows enhanced photocatalytic performance for the degradation of antibiotic residues in water. The prepared samples are characterized and the factors influencing the photocatalytic activity are discussed. The active species are studied and a reaction mechanism is suggested.

Results and Discussion

Powder X-ray diffraction (XRD) patterns of the synthesized 1T-MoS2, TNTs, and 1T-MoS2@TNTs samples are shown in Figure a. It is noticed that metallic 1T-MoS2 has two obvious characteristic peaks located at 13.9 and 33.2°, which could be assigned to its (002) and (100) facets.[9] The peak of the (002) facet is much greater than that of the (100) facet, indicating that the prepared 1T-MoS2 sample has a well-crystallized (002) facet.[15] As shown, the characteristic peaks of TNTs are completely consistent with that of anatase TiO2,[16] which illustrates that the obtained TNTs are anatase with good crystallinity.[17] As for 1T-MoS2@TNTs, the characteristic diffraction peaks of TNTs are all present without an obvious decrease of the relative peak intensity, which indicates that the growth of 1T-MoS2 does not destroy the crystal structure of TiO2. Additionally, quite weak diffraction peaks are also observed, which may come from the immobilized 1T-MoS2.

Figure 1

(a) XRD patterns of prepared 1T-MoS2, TNTs, and 1T-MoS2@TNTs; (b) Raman spectra of 1T-MoS2 and 1T-MoS2@TNTs.

Raman spectra are obtained to further identify the existence and the phase characteristics of the prepared 1T-MoS2 and 1T-MoS2@TNTs. As shown in Figure b, the characteristic peaks at 151, 201, 283, and 337 cm–1 are observed, which could be attributed to the J1, J2, E1g, and J3 of 1T-MoS2, respectively. Moreover, no obvious E2g peak of 2H-MoS2 is observed at 380 cm–1.[11] These results indicate that 1T-MoS2 is obtained and successfully grown and immobilized on TNTs.

Furthermore, X-ray photoelectron spectroscopy (XPS) is employed to analyze the prepared samples, as shown in Figures and S1. As shown in Figure S1a, all the elements of Mo, S, Ti, and O are present, suggesting the successful immobilization of 1T-MoS2 on TNTs. As shown in Figure a, two main deconvoluted peaks are observed at 227.4 and 230.6 eV on the Mo 3d spectra, which correspond to 3d3/2 and 3d5/2 of Mo4+ species.[17] The other two peaks at the binding energies of 231.8 and 234.7 eV belong to Mo6+, which may come from the unreacted MoO3 or the oxidized MoS2 in air.[15] Additionally, two weak peaks at 225.3 and 228.2 eV may belong to Mo3+, which should be derived from the S vacancy in MoS2 and/or the formed crystal defects.[16] As shown in Figure b, the peaks at 161.8 and 160.5 eV could be assigned to 2p3/2 and 2p1/2 of S2– from MoS2.[9] The other peaks at 168.7 and 167.6 eV can be assigned to 2p1/2 and 2p3/2 of S4+ in sulfate groups (SO32–).[18]Figure c shows two peaks at 463.6 and 457.8 eV, which could be attributed to Ti 2p of 1T-MoS2@TNTs. As shown in Figure d, two O 1s peaks at 530.5 and 529.1 eV are present. The first peak corresponds to the O atom bonded to Ti, and the latter one is ascribed to the surface hydroxyl species.[19] Compared with the curves in Figure S1b,c, it could be seen that both binding energies of Ti 2p and O 1s 1T-MoS2@TNTs are slightly shifted, which indicates the existence of electron interaction between 1T-MoS2 and TNTs.

Figure 2

XPS spectra of (a) Mo 3d, (b) S 2p, (c) Ti 2p, and (d) O 1s on 1T-MoS2@TNTs.

Scanning electron microscopy (SEM) images and energy-dispersive spectrometry (EDS) mapping images are used to investigate the surface morphology and elemental distribution of 1T-MoS2@TNTs, as shown in Figure . As shown in Figure a, TiO2 has an obvious one-dimensional morphology with diameters of about 10 nm. The existence of MoS2 is identified by the EDS mapping from an arbitrarily selected area, shown in Figure b–g. As shown, all the elements including Mo and S are evenly distributed in the 1T-MoS2@TNTs composites. In addition, no aggregated particles are observed on the surface of TNTs. It indicates that 1T-MoS2 has been successfully grown on the TiO2 surface with good dispersion.

Figure 3

(a) SEM image of 1T-MoS2@TNTs and (b) SEM images with the corresponding elemental mapping images of (c) all elements, (d) Ti, (e) O, (f) Mo, and (g) S.

Transmission electron microscopy (TEM) and high-resolution TEM (HRTEM) are applied to investigate the microstructure of the stable 1T-MoS2@TNTs composites, as shown in Figure . Figure a shows that 1T-MoS2@TNTs are completely composed of nanotubes with a diameter of about 7.5 nm. HRTEM images are shown in Figure b–d, and a more obvious tubular structure is observed in Figure b. The different contrast at the tube wall is also present, which might be derived from the loaded 1T-MoS2. For further understanding, high-resolution pictures of the circled parts in Figure b are given in Figure c,d. The lattice of the light part is 0.35 nm, which could be assigned to the (101) facet of TiO2. The tube wall is composed of several-layered TiO2, and the thickness of each layer is about 0.72 nm. As for Figure d, lattice fringes with a width of 0.26 nm can be attributed to the (100) facet of MoS2, which is at an angle of 120° from the TiO2 (101) facet.[20] To get further information about 1T-MoS2 in 1T-MoS2@TNTs, the selected zone in Figure S2a was filtered by fast Fourier transform filtration to remove the irrelevant noise based on the previous image-processing procedure.[21] The obtained atomic arrangement diagram and lattice dislocation are shown in Figure S2b,c, respectively. The characteristic trigonal lattice is evidently seen in Figure S2a and the hexagonal atomic arrangement of Mo of MoS2 is seen in Figure S2b, which confirms that the obtained MoS2 has the 1T phase.[21]Figure S2c shows the lattice fringes in the same direction with obvious dislocations indicating the presence of defects in 1T-MoS2, which is in correspondence with the above XPS results. Additionally, the a and b vectors give a length of 3.15 Å, which indicates that one Mo atom may be surrounded by six S atoms, as shown in Figure S2d.

Figure 4

(a,b) HRTEM images of the 1T-MoS2@TNTs composite; (c,d) enlarged images of the selected zone.

The light response of the obtained samples is analyzed by UV–vis diffuse reflectance spectroscopy, as shown in Figure a. Compared to TNTs, 1T-MoS2@TNTs has stronger absorption in the visible region. At the same time, slight red shifts of the spectrum could be observed for 1T-MoS2@TNTs as shown in Figure S3a, which indicates that the interaction between 1T-MoS2 and TNTs may influence the band energy.[20] The band gaps of prepared photocatalysts are calculated by using the Tauc equation,[9] shown in Figures b and S3b. The band gap of TNTs is 3.17 eV. However, the band gap of 1T-MoS2@TNTs is decreased to 3.09 eV. More decreases are observed by further increasing the contents of 1T-MoS2 and the values are 3.05, 3.03, and 2.84 eV for 1.0 wt % 1T-MoS2@TNTs, 1.5 wt % 1T-MoS2@TNTs, and 2.0 wt % 1T-MoS2@TNTs (see Figure S3b), respectively. It is speculated that the narrower band gap may improve the light harvesting, which facilitates enhanced photocatalytic activity.

Figure 5

(a) UV–vis diffuse reflectance spectra and (b) band gap of TNTs and 1T-MoS2@TNTs.

The degradation of tetracycline hydrochloride (TC-HCl) under visible light is carried out to investigate the photocatalytic performance of the obtained samples, as shown in Figure a. Pristine 1T-MoS2 and TNTs can only degrade about 10 and 23% TC-HCl in 1 h, respectively. In contrast, 1T-MoS2@TNTs (0.5 wt %) can degrade 57% TC-HCl under the same conditions, which is higher than previously reported efficiencies of 15 and 25%.[9,14] In addition, Figure S4 shows the degradation curves of 1T-MoS2@TNTs with different loaded amounts of 1T-MoS2, where it is observed that the prepared 0.5 wt % 1T-MoS2@TNTs exhibits the highest catalytic activity. The pseudo-first-order kinetics fitting of the photodegradation is given in Figure b. The reaction rate constants k of TC-HCl decomposition are 1.1 × 10–3 and 4.4 × 10–3 min–1 for 1T-MoS2 and TNTs, respectively. Enhanced reaction rates are observed for 1T-MoS2-loaded photocatalysts, and a maximum of 140 min–1 is achieved for 1T-MoS2@TNTs (0.5 wt %), which is 12.7 and 3.2 times more than those of 1T-MoS2 and TNTs. The full-wavelength absorption spectra of TC-HCl during degradation on 1T-MoS2@TNTs are shown in Figure S5. The absorption from TC-HCl is clearly observed and the absorption intensity decreases along with the reaction, indicating that no other intermediates are formed. Cyclic experiments for TC-HCl photodegradation have been conducted and the degradation rate is found to be reduced from 56.2 to 36.8% after four cycles, as shown in Figure S6.

Figure 6

(a) Photocatalytic degradation of TC-HCl by 1T-MoS2, TNTs, and 1T-MoS2@TNTs and (b) pseudo-first-order kinetics fitting of the photodegradation.

Trapping experiments are performed to figure out the active ingredients in the photocatalytic process by using 1T-MoS2@TNTs as the photocatalyst, as shown in Figure . Ethylenediaminetetraacetic acid (EDTA), benzoquinone (BQ), and tert-butyl alcohol (TBA) are used to scavenge the holes (h+), superoxide (•O2–) radicals, and hydroxyl (•OH) radicals, respectively. 10.0 mg of the photocatalyst is used and the operation is similar to the above photodegradation processes. After the addition of EDTA, BQ, and TBA, the degradation efficiency is obviously reduced, which indicates that h+, •O2–, and •OH may be the active species and responsible for the catalytic reaction.

Figure 7

Photocatalytic degradation of TC-HCl over 1T-MoS2@TNTs without and with addition of EDTA, TBA, and BQ.

Photocurrent response and electrochemical impedance spectroscopy (EIS) are employed to analyze the charge transfer in the photocatalytic process. As shown in Figure a, the 1T-MoS2@TNTs composite has a much stronger photocurrent than the pristine 1T-MoS2 and TNTs. This indicates that 1T-MoS2@TNTs photocatalyst has better charge separation efficiency, which accounts for its higher activity. EIS is used to further clarify the charge transport on the prepared samples, as shown in Figure b. 1T-MoS2@TNTs displays a much smaller EIS arc radius than TNTs, which indicates that the addition of metallic 1T-MoS2 to TNTs can reduce the resistance of TNTs.[20]

Figure 8

(a) Typical photocurrent response and (b) EIS spectra of samples 1T-MoS2, TNTs, and 1T-MoS2@TNTs.

The band structures of TNTs and 1T-MoS2 are obtained from the Mott–Schottky results in Figure S7, which indicate that they are all n-type semiconductors.[16] Together with the above discussions, a mechanism is proposed for the photodegradation of TC-HCl on 1T-MoS2@TNTs, as shown in Figure . The potentials of CB and VB of 1T-MoS2 are 0.18 and 1.98 eV, and those of TNTs are −0.4 and 2.8 eV, respectively. The electrons are excited to the CB of TiO2 under visible-light irradiation and migrated to the CB of 1T-MoS2 through the formed interface. Meanwhile, the holes in the VB of TiO2 may transfer to the VB of 1T-MoS2. Consequently, a type I heterojunction may be formed in the 1T-MoS2@TNTs composite, which optimizes the photocatalytic activity by improving the charge separation.[22]

Figure 9

Schematic illustrations for the photogenerated charge carrier transfer on 1T-MoS2@TNTs.

Conclusions

In summary, 1T-MoS2@TNTs composites have been successfully constructed by the in situ growth of 1T-MoS2 on the TNTs using a hydrothermal process, which show enhanced photocatalytic performance for the degradation of TC-HCl under visible light. The improved photocatalytic activity is closely related to the loaded amount of 1T-MoS2, and the highest photocatalytic efficiency is observed on 0.5 wt % 1T-MoS2@TNTs. The introduction of 1T-MoS2 may optimize light absorption and charge separation/transport. The active species are identified and a reaction mechanism is proposed. This work may provide an alternative for the construction of a new photocatalyst and also a new direction for application of 1T-MoS2 in the energy and environmental fields.

Materials and Methods

Synthesis of 1T-MoS2

The preparation of 1T-MoS2 was conducted according to the modified method from ref (8). Typically, 1.96 mg MoO3, 0.10 g thioacetic acid (TAA), and 0.96 g urea were dissolved in 80.0 mL of deionized water and stirred magnetically (with a cylindrical stirring bar, 8 × 35 mm) for 2 h at a speed of 400 rpm. Then, the solution was sealed in an autoclave and heated at 200 °C for 12 h. After that, the solution was cooled down to room temperature, and the solid product was collected and washed with deionized water several times. The prepared 1T-MoS2 was dispersed in deionized water before further use.

Synthesis of TNTs

TNTs were prepared using a hydrothermal method.[23] Typically, 0.96 g of urea and 1.92 g of titanium sulfate were dissolved in 80.0 mL of deionized water and stirred for 2 h. Then, the solution was transferred into a reaction vessel and reacted at 220 °C for 12 h to obtain TiO2 powder. The obtained TiO2 powder was added to 100.0 mL of 10.0 mol/L NaOH solution and stirred for 4 h. Then, the solution was transferred into a Teflon container and heated at 150 °C for 12 h. After washing with a 0.1 mol/L HNO3 aqueous solution and deionized water, white products were obtained and ground into powder for further use.

Synthesis of 1T-MoS2@TiO2

400 mg TiO2 powder was added into the mixed solution containing 2.1 mg TAA, 18 mg urea, and 1.8 mg MoO3 and stirred for 2 h. Then, the solution was sealed in an autoclave and heated at 200 °C for 12 h. After that, the solution was cooled down to room temperature, and the solid product was collected and washed with deionized water several times. The obtained samples were named as 1T-MoS2@TNTs (the content of 1T-MoS2 is 0.5 wt %). Other samples with different MoS2 contents were similarly synthesized by changing the concentrations of MoO3 precursors (Supporting Information).

Characterization

XRD was conducted at a scan rate of 10°/min on a Bruker D2 PHASER with Cu Kα-radiation. XPS was obtained on the Axis Supra XPS instrument using Al Kα as the source. SEM was conducted on a FE-SEM S-4800 (Hitachi). TEM was carried out on the FEI F200S equipment at an accelerating voltage of 200 kV.

Photoelectrochemical Measurements

Electrochemical testing was carried out on the CHI 660E electrochemical workstation. 15.0 mg prepared MoS2, TNTs, and 1T-MoS2@TNTs were separately dissolved in 3 mL of polyvinylidene fluoride solution and coated on the fluorine-doped tin oxide conducting glass by spin coating to form an electrode, respectively. A 300 W xenon lamp was used as the light source for photocurrent testing.

Photocatalytic Performance

The photocatalytic performance was studied by degrading the TC-HCl solution under visible light as the model reaction. A 300 W xenon lamp worked as the light source in the wavelength range of 420–780 nm, and the distance between the lamp and liquid level was set as 15 cm. 1.0 mg of the photocatalyst was dispersed in 100.0 mL of 20.0 mg/L TC-HCl solution. The reaction solution was stirred magnetically at a speed of 200 rpm in the dark for 1 h to achieve adsorption–desorption equilibrium. The concentration of TC-HCl was investigated at its maximum adsorption of 356 nm on the UV 2600 spectrophotometer. The pH values of the solution before and after the reaction are 4.23 and 4.28, respectively, and not much difference is observed.