Facile MoS₂ Growth on Reduced Graphene-Oxide via Liquid Phase Method

Introduction

Due to their unique electronic properties, atomically-thin two-dimensional (2D) materials such as graphene (Geim and Novoselov, 2007), hexagonal boron nitride (h-BN) (Dean et al., 2010) and transition-metal dichalcogenide (Radisavljevic et al., 2011; Lembke and Kis, 2012; Wang et al., 2012) have been attracting increasing attention. Molybdenum disulfide (MoS₂) is a layered 2D material that is traditionally used as a solid-state lubricant and is the major industrial catalyst for the hydrodesulfurization process (HDS) (Li et al., 2011; Kong et al., 2013). Recently, MoS₂ and hybrid MoS₂ materials have been extensively studied as catalysts in the hydrogen evolution reaction (HER) (Liao et al., 2013; Wang et al., 2014; Zheng et al., 2014). In recent years, ultrathin graphene and MoS₂ crystals have been obtained by either physical or chemical exfoliation methods (Coleman et al., 2011), while large-scale uniform layers have been grown by chemical-vapor-deposition (CVD) techniques (Chen et al., 2016; Samad et al., 2016; Zobel et al., 2016). Both graphene and MoS₂ in monolayer form exhibit unique optical and electrical properties compared with the respective multilayer analogs. MoS₂, which is referred to in the literature as the inorganic analog of graphene, is a semiconductor with a large direct band gap (Conley et al., 2013). The presence of a band gap in monolayer MoS₂ opens new possibilities in electronic, and photonic applications. Furthermore, the edges of MoS₂ monolayers are catalytically active so that increasing the number of such sites is critical for the design of advanced catalysts (Jaramillo et al., 2007).

During the last few years, there has been substantial interest in the literature concerning the synthesis of hybrid materials based on graphene and MoS₂. Kun Chang and Weixiang Chen reported the synthesis of single-layer MoS₂/graphene dispersed in amorphous carbon and tested its electrochemical performance for lithium batteries (Chang and Chen, 2011). The reaction takes place in water under autoclave conditions at 240°C for 24 h. Molybdenum and sulfur precursors Na₂MoO₄•2H₂O and sulfocarbamide (NH₂CSNH₂) were used and the final material was annealed at 800°C for 2 h in a H₂/N₂ atmosphere. Ma et al. reported the synthesis of few-layer MoS₂-graphene composites via a cationic surfactant-assisted hydrothermal method followed by a post-annealing step and showed that the electrochemical performance for lithium storage is greatly improved compared to bare MoS₂ (Ma et al., 2014). Following a similar experimental procedure in a two-step hydrothermal synthesis, Quanjun Xiang et al. reported the decoration of MoS₂/graphene heterostuctures with TiO₂ nanoparticles and tested them as non-precious metals photocatalysts for H₂ production (Xiang et al., 2012). In a one-step hydrothermal method, MoS₂/reduced GO was prepared by (NH₄)₂MoS₄ reduction in N,N'-dimethylformamide (DMF) at 200°C by hydrazine. A transmission-electron-microscopy (TEM) study showed that most of the MoS₂ lies flat on the reduced GO sheets, with some possessing-induced folded edges exhibiting parallel lines (Li et al., 2011).

Summarizing the literature concerning the synthesis of MoS₂/graphene composite nanomaterials via liquid phase reactions, most of the methodologies involve a hydrothermal process for at least 12 h in the 200–240°C temperature range in one- or two-step procedures. Here we report the synthesis of MoS₂/r-GO following a liquid-phase approach. The method is based on the thermolytic decomposition of (NH₄)₂MoS₄ by oleyl amine at 350°C in the presence of chemically-exfoliated reduced graphene oxide (r-GO). The presence of exfoliated r-GO nanosheets in the reaction mixture is crucial, providing a confinement environment for the nucleation and growth of the MoS₂. It is shown that r-GO is predominantly decorated by single-layer MoS₂ that lies flat on the r-GO sheets. The MoS₂/RGO composites are surface functionalized efficiently by oleyl amine molecules, making the material dispersible in non-polar solvents such as toluene and hexane.

Experimental Section

Materials

Ammonium tetrathiomolybdate ((NH₄)₂MoS₄, 99.97%), oleyl amine (90% C-18 content), carbon nanotubes, carbon nanofibers, absolute ethanol and n-hexane were purchased from Sigma-Aldrich and used without any further purification. Graphene oxide was synthesized using a modified Staudenmaier's method (Staudenmaier, 1898; Gengler et al., 2010; Stergiou et al., 2010). Briefly powdered graphite (purum, powder ≤0.2 mm, Fluka) were treated by a 2/1 mixture of concentrated H₂SO₄ (Riedelde Haën) and HNO₃ acid, while cooling in an ice–water bath. Subsequently powdered KClO₃ (Fluka) were added to the mixture in small portions under vigorous stirring while cooling in an ice-water bath. The reaction was quenched after 18 h by pouring the mixture into ultrapure water. The oxidation product was washed until the pH reached 6.0, and finally dried at room temperature.

Synthesis of MoS₂/Reduced Graphene Oxide

The reduction and functionalization of GO was carried out via a chemical approach. The pristine GO was exfoliated in oleyl amine via probe sonication at room temperature. Subsequently the mixture was heated up to 300°C under H₂-Ar reducing atmosphere for 1 h. The oleyl amine functionalized r-GO was separated by centrifuging at 12,000 rpm, washed twice in ethanol in order to remove the excess of oleyl amine, and finally dried at room temperature.

MoS₂/r-GO nanocomposites were prepared according to the following procedure. Oleyl-amine-functionalized r-GO (40 mg) was dispersed in 10 ml oleyl amine under probe sonication in an inert atmosphere for 15 min. The mixture was then transferred to a heating plate and 20 mg of (NH₄)₂MoS₄ was dissolved into the mixture under vigorous magnetic stirring. In the beginning, the temperature was raised to 120°C and kept there for 15 min under a high-purity N₂ blanket. The reaction finally took place at 350°C for 1 h. After cooling the mixture to room temperature, the composite material was precipitated by adding absolute ethanol and was separated by centrifugation (8,000 RPM, 25 min). The washing/separation procedure was repeated until the complete removal of excess oleyl amine molecules and unreacted components. Finally the composite material was dispersed and stored in toluene.

Characterization

The morphology of the composite MoS₂/RGO material was studied by a Nion Ultra STEM 100 aberration-corrected scanning transmission electron microscope (STEM), which is equipped with a Gatan Enfinium spectrometer for Electron Energy Loss Spectroscopy (EELS). The STEM was operated at 60 kV. Additionally, the material was characterized by atomic-force microscopy (AFM), Raman, and X-ray photoelectron spectroscopy (XPS).

XPS measurements were performed under ultrahigh vacuum conditions at a base pressure of 5 × 10⁻¹⁰ mbar in a SPECS GmbH instrument equipped with a monochromatic MgKα source (hν = 1253.6 eV) and a Phoibos-100 hemispherical analyzer. Pulverized samples were dispersed in toluene (1 wt%), and after short sonication and stirring, a minute quantity of the suspensions were drop-cast on evaporated gold films supported on mica substrates and left to dry in air before transferring them to ultrahigh vacuum. The energy resolution was set to 0.3 eV and the photoelectron take-off angle was 45° with respect to the surface normal. Recorded spectra were the average of three scans with energy step set to 0.05 eV and dwell time 1 s. All binding energies were referenced to the Si 2p level at 99.15 eV. Spectral analysis included a Shirley background subtraction and peak deconvolution employing mixed Gaussian–Lorentzian functions, in a least squares curve-fitting program (WinSpec) developed at the Laboratoire Interdisciplinaire de Spectroscopie Electronique, University of Namur, Belgium.

Raman spectra were recorded with a micro-Raman (μ-Raman) Renishaw RM1000 system using a laser diode excitation line at 532 nm in the frequency range of 200–2,000 cm⁻¹. Raman scatter was collected by means of an Olympus optical microscope, equipped with 50× and 100× lenses. Using the 50× lens, the probing spot was about 2 μm in diameter, while the laser was operated at 5 mW unless photodecomposition occurred when the power was decreased. The spectrometer was calibrated by recording the spectrum from a Si sample with characteristic Raman peak at 520.7 cm⁻¹. Raman spectra were obtained from samples in the form of drop casted films onto glass substrates. The reported spectra are an average of 3–5 scans.

AFM images were obtained in tapping mode with a 3D Multimode Nanoscope, using Tap-300G silicon cantilevers with a tip radius <10 nm and a force constant of ≈20–75 N m⁻¹. Samples were deposited onto silicon wafers (P/Bor, single side polished) from dilute dispersions of MoS₂/r-GO in toluene by drop casting.

Results and Discussion

The synthetic procedure for the decoration of r-GO with MoS₂ is illustrated in Scheme 1. Oleyl-amine-functionalized exfoliated r-GO sheets were dispersed in oleyl amine, providing nucleation centers for the growth of MoS₂. Then, the thermolytic decomposition of (NH₄)₂MoS₄ at 350°C resulted in the formation of a single layer of MoS₂ on the surface of the r-GO. The XRD patterns of MoS₂/r-GO nanocomposite are presented in Figure S1, in the Supplementary Material section. Because of the nanosize regime of the material, the diffraction peaks are quite broad, but it is possible to distinguish four diffraction peaks. The intense broad peak around 2θ = 24° corresponds to the (002) diffraction plane of r-GO, while the other peaks can be assigned to the (101), (103), and (110) lattice planes of MoS₂.

Scheme 1. Schematic diagram of the synthetic procedure.

The morphology of the MoS₂/r-GO nanocomposite was examined by atomic- resolution Scanning Transmission Electron Microscopy (STEM) measurements. In Figure 1, we present Z-contrast images of the MoS₂ nanosheets on the few-layer r-GO. It is obvious that the MoS₂ layers grow discontinuously on the surface of the r-GO sheets, covering most of their surface by a single- to few-layer MoS₂ films. Furthermore, the images confirm the few-layer nature of the r-GO as well as the presence of isolated single atoms, most likely Mo atoms, and Mo vacancies. The STEM observations indicate very clearly that the oleyl-amine-functionalized r-GO sheets could be an excellent substrate for the nucleation and growth of MoS₂. This effect can be attributed to the presence of oxygen-containing functional groups on the surface of the r-GO or on the weak van der Waals forces (Shi et al., 2012) between the two layered materials, which accommodate the relatively large lattice mismatch between MoS₂ and graphene (~28%) (Ma et al., 2011).

Figure 1. Z-contrast STEM images of the MoS₂ flakes grown on few-layer r-GO.

In order to clarify the role of the r-GO surface functionalization on the growth of the MoS₂ sheets, the synthesis was also done following a similar methodology using GO as staring material, without any previous reduction and surface functionalization. The TEM images, presented in the Supplementary Material section (Figure S2), clearly show that mainly isolated MoS₂ nanosheets grow on the GO surface, but they do not lie flat on the GO sheets. Furthermore, the formation of nanoparticles instead of nanosheets can be observed. This morphology is quite similar with a previously reported work in which the MoS₂/r-GO composites synthesized by mixing GO with (NH₄)₂MoS₄ in an alcohol/water medium followed by the (NH₄)₂MoS₄ decomposition by HCl acid addition (Koroteev et al., 2011). The above findings prove indirectly that the surface chemistry of the carbon material plays a very important role in the growth of MoS₂. The oxygen-containing functional groups, as already reported, affect the nucleation and growth of the MoS₂ nanosheets, but on the other hand probably suppress the van der Waals forces that are responsible for the flat MoS₂ growth. Finally, STEM-EELS mapping of sulfur and carbon (Figure 2b, which corresponds to the white square area of Figure 2a) shows the distribution of MoS₂ on the surface of the r-GO in the nanocomposite material.

Figure 2. Z-contrast STEM image (a) and EELS sulfur and carbon mapping (b) of MoS₂-decorated few layer r-GO. (b) corresponds to the white box area of (a).

The heterostructural nature of the MoS₂/r-GO nanocomposite was additionally identified by AFM. In Figure 3, we present the height and phase images (measured in tapping mode AFM) of MoS₂ nanosheets grown on few-layer r-GO. The AFM images of the MoS₂/r-GO nanocomposite deposited on a Si-wafer show the successful growth of MoS₂ layers on the graphene surface (Figure 3). The analysis of the height section profile (Figure 4) showed that the average thickness of the MoS₂ flakes on the r-GO sheets is around 4–8 nm, which corresponds to a few MoS₂ layers (~5–10). Furthermore, the height profile analysis obtained at the edges of the r-GO sheets, showed a 1.5 nm thickness, a value typically reported for very few r-GO layers as also demonstrated from phase images.

Figure 3. AFM height (left) and phase (right and bottom) images of MoS₂/r-GO nanocomposite.

Figure 4. AFM cross-section analysis of MoS₂/r-GO nanocomposite (a single MoS₂ layer can be distinguished from a package of few-layer MoS₂ sheets; picture below).

XPS measurements were used in order to confirm the successful formation of MoS₂ on the r-GO surface. C1s (a), Mo3d (b), and S2p (c) XPS spectra of MoS₂/r-GO sample are shown in Figure 5. After deconvolution with mixed Gaussian–Lorentzian functions, the C1s spectrum consists of five components. The main peak at a binding energy of 284.6 eV, representing 53.1% of the total C1s area, is attributed to the C-C and C-H bonds. The component at 285.4 eV, contributing a 32.7% to the total C1s signal, is assigned to the hydroxyl C-O bonds as well as the C-N bonds from the oleyl amine. The third peak at 286.2 eV (9.0%) originates from the C-O-C epoxide/ether groups, whereas carbonyl functional groups (C = O) are associated with the peak at 287.3 eV (3.5%). Finally, the higher binding energy contribution recorded at 288.8 eV (1.7%) corresponds to the carboxyl groups (O-C=O) (Dimos et al., 2017; Kouloumpis et al., 2017; Tzitzios et al., 2017).

Figure 5. C1s (a), Mo3d (b), and S2p (c), X-ray photoemission spectra of MoS₂/r-GO nanocomposite.

The domination of the C-C peak along with the high carbon content (80.8%) point out the presence of a r-GO substrate indicating the successful reduction of the starting GO. Due to the low concentration of the functionalizing oleyl amine and accordingly the even lower relative nitrogen atomic concentration, nitrogen and consequently oleyl amine were not detected by XPS. Nevertheless, its presence is verified by FTIR spectroscopy (see Figure S3), which is more effective for distinguishing organic molecules. The presence of oleyl amine in the r-GO and MoS₂/r-GO materials was indicated from the characteristic absorptions of the aliphatic and the amine groups in the IR spectra. The vibrations at 2,920 and 2,851 cm⁻¹ are due to the symmetric and asymmetric stretching modes of CH₂, the weak band at 2962 arises from the asymmetric C-H stretching mode of the terminal CH₃ groups, while the double broad bands around 3,240–3,340 cm⁻¹ arise from the stretching vibrations of the NH₂ group. The absence of these characteristic vibrations in the GO spectrum indirectly confirms the successful functionalization by the oleyl amine molecules. Furthermore, the absence of shifting of the N-H stretching band (around 3,300 cm⁻¹ in pure oleyl amine), to lower wavenumbers indicates that the interaction of the NH₂ groups on the surface of both r-GO and MoS₂/r-GO is weak. On the other hand, molybdenum and sulfur were detected with an atomic concentration of 0.8 and 2.1% respectively proving the development of MoS₂ layers on the composite materials as simultaneously the S/Mo ratio is close to 2. In fact, from the distinction of the Mo and S species further conclusions can be deduced.

More specifically, as observed in Figure 5b of the Mo3d energy region, the spectrum consists of two Mo components. From their binding energies (229.4 and 232.4 eV), these peaks are ascribed as Mo^IV 3d_5/2 and Mo^IV 3d_3/2, respectively, and are correlated with MoS₂ structures (Spevack and McIntyre, 1993; Merki et al., 2011; Kibsgaard et al., 2012; Zheng et al., 2014). No peaks or shoulders are detected at higher binding energies, indicating the lack of higher oxidation states (+5 or +6) which are possible for Mo and suggesting the growth of further Mo_xS_y species except MoS₂ (Spevack and McIntyre, 1993; Merki et al., 2011; Kibsgaard et al., 2012; Zheng et al., 2014). Additionally, the peak at 227.0 eV is defined as S2s whilst more information about sulfur is obtained by the S2p XPS spectrum at Figure 5c. As in the case of Mo, only two peaks are observed; a main peak at 162.3 eV and a shoulder at 163.7 eV, associated with the doublet from S²⁻ 2p_3/2 and S²⁻ 2p_1/2 respectively (Spevack and McIntyre, 1993; Merki et al., 2011; Kibsgaard et al., 2012; Zheng et al., 2014). Again, no extra peaks are detected affirming that no ${\text{S}}_2^{2-}$ or other species are present (Spevack and McIntyre, 1993; Merki et al., 2011; Kibsgaard et al., 2012; Zheng et al., 2014). In conclusion, the XPS study confirms the successful growth of MoS₂ structures on the composite material with a high content of 10.5 wt% (as deduced from the quantitative XPS analysis), excluding the creation of further Mo_xS_y species.

In addition, Raman spectroscopy is a powerful technique for both r-GO and MoS₂ materials, since it provides crucial information about the hybridization of atoms and the vibrational modes. As seen in the spectrum of the composite MoS₂/r-GO material (Figure 6), both r-GO-originating G- and D-bands are present as expected. The G-band at 1,594 cm⁻¹ is associated with sp² hybridized carbon atoms, while the D-band at 1,345 cm⁻¹ originates from the breathing modes of six-atom rings, requiring a defect for its activation, hence is correlated with lattice defects in the case of r-GO. Thus, the ratio of the D- to G-band intensities (I_D/I_G) states the defects ratio of the r-GO layers, (Tzitzios et al., 2017) and is equal to 0.86. A value slightly lower, as expected, than the one observed for starting GO (see Figure S4). Moreover, peaks at the lower region of Raman shift are attributed to the formed MoS₂ layers. This region of the spectrum is multiplied 4 times compared to the rest spectrum in Figure 6, in order to have a clear view of the area. The peak at 407 cm⁻¹ is assigned to the A_1g vibrational mode of MoS₂, while the one at 375 cm⁻¹ originates from the corresponding ${\text{E}}_{2\text{g}}^1$ mode (Lee et al., 2010; Kibsgaard et al., 2012; Sundaram et al., 2013; Zhang et al., 2013, 2015; David et al., 2014).

Figure 6. Raman spectrum of the MoS₂/r-GO nanocomposite.

The distance between these two peaks has been correlated in the literature with the number of MoS₂ layers (Lee et al., 2010; Sundaram et al., 2013; Zhang et al., 2015). Unfortunately in our case, we cannot support the formation of single layers as witnessed in TEM. This is because multilayers are also formed nearby and since the probing spot in Raman measurements was about 2 μm in diameter it was not possible to focus in a region with exclusively single layers of MoS₂. Nonetheless, the presence of both MoS₂ and r-GO peaks testify to the successful development of the composite material.

In order to study the generality of the methodology carbon nanotubes and carbon fibers were also used as support for the growth of MoS₂ layers. Both carbon nanotubes (CN), and carbon nanofibers (CNF), were first functionalized with oleyl amine followed by the growth of the MoS₂ using the same procedure. In Figure 7, we present STEM images of the MoS₂/CN composites. The images show similar behavior with the r-GO composites, i.e., the MoS₂ layers grow flat on the surface of the carbon nanotubes. It is worth mentioning here that, in the CN composites, most of the MoS₂ layers are single-layer, suggesting that the sp² nature of carbon is crucial for the uniform and single-layer growth of the MoS₂ on the surface of carbon-based materials. Atomic-resolution Z-contrast STEM images (Figure 7e) reveal the presence of Mo vacancies in the MoS₂ monolayers. The images of the corresponding composites with CNF are presented in the Supplementary Material section (Figure S5), and show clearly the uniform growth of mainly monolayer MoS₂ on the CNF surfaces.

Figure 7. Z-contrast STEM images (a,c), bright field image (b) and EELS sulfur and carbon elemental mapping (d), (with red and gray colors representing sulfur and carbon, respectively), of MoS₂/CN composites. Z-contrast image of monolayer MoS₂ on CN showing individual Mo vacancies (e).

In conclusion, in this work we report a facile liquid phase chemical approach for the synthesis of hybrid MoS₂/r-GO without autoclave conditions. The method is based on the thermolytic decomposition of (NH₄)₂MoS₄ inorganic salt in hot oleyl amine in the presence of chemically exfoliated r-GO. The r-GO sheets work as nucleation centers for the growth of MoS₂ and the composite material is composed mainly of single- to few-layer MoS₂ on few-layer r-GO. The use of oleyl amine functionalized r-GO as starting material for the growth of the MoS₂ seems to be crucial for the flat growth of MoS₂. In contrast, when GO is used as starting material, the MoS₂ layers do not grow flat. To our knowledge, it is the first time that MoS₂/r-GO is synthesized in non-polar media without autoclave conditions and the final hybrid nanocomposites are well dispersible in non-polar solvents such as n-hexane. The methodology is generally applicable and works in a similar way in the case of MoS₂ growth on the surface of other carbon-based materials such as multiwall carbon nanotubes (CN) and carbon nanofibers (CNF). This kind of colloidal dispersion can be applicable in the formation of films through inkjet printing techniques (Arapov et al., 2014).

Author Contributions

VT developed the synthesis concept, synthesized the materials and with SA, and GP prepared the manuscript; KD contributed to the manuscript preparation, and with AK, DG, and MAK performed the AFM, XPS and Raman spectroscopy studies; GB, YA, HJ, MK, and MF carried out the structural and the surface functionalization studies; RM, MR, NB, J-CI, SJP, AB, and STP carried out the STEM studies. All authors discussed the results and commented on the manuscript.

Conflict of Interest Statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Acknowledgments

This work is partially funded by the Petroleum Institute (a part of Khalifa University of Science and Technology). Electron microscopy work at Oak Ridge National Laboratory (ORNL) was supported by the U.S. Department of Energy (DOE) Office of Science, Basic Energy Sciences (BES), Materials Science and Engineering Directorate (MR) and through a user project conducted at the Center for Nanophase Materials Sciences, which is a DOE Office of Science User Facility (J-CI, RM).Work at Vanderbilt University was supported by the U. S. Department of Energy grant DE-FG02-09ER-46554 (RM, STP) and the McMinn Endowment.

Supplementary Material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmats.2018.00029/full#supplementary-material

References