Sara Cravanzola,
Federico Cesano*,
Giuliana Magnacca,
Adriano Zecchina and
Domenica Scarano
Department of Chemistry, NIS (Nanostructured Interfaces and Surfaces) Interdepartmental Centre and INSTM Centro di Riferimento, University of Torino, Via P. Giuria, 7, 10125 Torino, Italy. E-mail: federico.cesano@unito.it
First published on 10th June 2016
Graphene and its derivatives exhibit large surface area, being ideal templates to facilitate the nucleation, growth or interaction of a huge variety of structures. Among them, molybdenum disulphide, with its structural and morphological compatibility with graphene, can be a candidate for achieving an excellent integration, to make new hybrid nanocomposites with outstanding characteristics. Among the synthesis methods of graphene/MoS2 composites, the solution-phase exfoliation of a MoS2/graphite mixture, by means of ultrasounds, shows significant advantages in terms of large amount production without altering the main properties of 2D nanomaterials. Moreover MoS2, having a strong absorption in the visible, has been exploited as a novel visible light-sensitive semiconductor photocatalyst. But, due to the quick recombination of photo-generated charge carriers, the photocatalytic efficiency of MoS2 has to be further improved. Graphene and graphene related materials, as excellent electron-acceptor/transport materials, have been applied to photocatalysis, because they are able to decrease the photo-generated electron–hole recombination, thus improving the light absorption. Therefore, MoS2 and graphite oxide (GO) have been simultaneously sonicated in an ethanol/water mixture and characterized from the structure, morphology and electronic properties point of view. The composite was thermally treated in such a way to reduce GO and the photocatalytic activity of the reduced GO/MoS2 has been investigated by means of UV-vis spectroscopy, following the degradation of methylene-blue (MB) under solar-like irradiation.
In particular, as for graphene and its derivatives, due to their large surface area, they are ideal templating materials to facilitate the nucleation, growth or interaction of a huge variety of structures, from organic (polymers or biomolecules) to inorganic (oxides, oxide derivatives), for the production of functional hybrid nanocomposites.8
New opportunities and applications can come from the union of the peculiar properties of polymers, such as resistance to chemical corrosion, simplicity of manufacturing, low density and lightweight, together with the peculiar mechanical and electrical properties of the graphene-based conducting fillers,15,16 thus developing nanocomposites, whose well-known piezoresistive nature17 makes them suitable for sensing applications, including pressure,18 tactile,19 flow sensors20 and for monitoring the structural integrity of mechanical elements.21
Moreover, graphene could be beneficial in biomedical applications and electrochemical biosensing,22 being involved in different combinations with a wide range of biomolecules, from DNA, aminoacids and proteins, to cells and bacteria.23,24
Concerning carbon-inorganic composites, various metallic-based nanostructures like Pt, Co, Si, Al, Mg, Cu and Al/Au, Pd/Au, Mg/Sn have been successfully anchored to graphene nanosheets, to enhance performances ranging from catalysis to electronics and sensors.25
In addition, to further improve their properties, a number of metal oxides such as TiO2, SiO2, ZnO, SnO2, MnO2, Co3O4, Fe3O4 has been grown on graphene nanosheets via different synthetic approaches.26 Notably, synergy effects between nanocarbons and TiO2 have been shown for the photocatalytic degradation of organic pollutants compounds.3,27
It is noteworthy that molybdenum disulphide, a transition metal dichalcogenide (TMD), due to its physical, optical, electrical and structural properties, can be an excellent candidate for being combined with graphene and graphene derivatives, originating new hybrid nanocomposites with outstanding characteristics.28–31
The achievable applications of this kind of nanocomposite are depending on the adopted synthesis approach. In the bottom up approach, graphene-based structures are mixed with an aqueous solution of a MoS2 precursor, e.g. ammonium tetratiomolybdate,32 or with a mix of ammonium eptamolybdenate or sodium molybdate dihydrate33 and thiourea,34 thus obtaining a nanocomposite, which can be suitable as counter-electrode in dye-sensitized solar cells. The ion intercalation technique represents another possible way to synthesize layered MoS2/graphite composites, thus making them suitable for lithium-ions batteries.35 On the other hand, the top-down approach, consisting in breaking up the massive and bulk MoS2 material, allows to obtain particles of decreased dimensions. In particular, the solution-phase exfoliation of a MoS2/graphite mixture, by means of ultrasounds, shows significant advantages in terms of large amount production without altering the molecular structures and the intrinsic electronic properties of 2D nanomaterials.36
Beside all the promising applications of MoS2/graphite composites, it is important to underline that mono-layers and few-layers of MoS2, obtained from liquid exfoliation, suffer seriously from the restacking phenomena,37 which would directly suppress the charming advantages of 2D MoS2. This problem can be satisfactorily solved by introducing graphene as a sort of intercalating compound, which can avoid the complete restacking of MoS2 layer, as shown in literature for a system composed by MoS2 and WS2.38 Moreover, MoS2 has a strong absorption in the visible spectrum region, therefore it has been exploited as a novel visible light-sensitive semiconductor photocatalyst for photocatalytic applications.39,40 However, due to the quick recombination of photo-generated charge carriers,41 the photocatalytic efficiency of MoS2 has to be further improved. Along this line, graphene and graphene derivatives, as excellent electron-acceptor/transport materials, have been applied to photocatalysis, because they are able to decrease the photo-generated electron–hole recombination, thus improving the light absorption.42,43 The structural and morphological compatibility between graphene-like materials and single-layer MoS2 sheets should enable better intercalation to achieve the anticipated benefits of such integration.
Following this line, it would be of great interest to analyze the results of the simultaneous solvent-assisted ultrasonication of graphite and molybdenum disulphide.
As far as the choice of the sonication solvent is concerned, it is well known that the effectiveness of the exfoliation process of nanomaterials in liquids can be partially predicted by the theory of Hansen solubility parameters.44–46 In this domain, it has been shown47 that the use of alcohols, although generally considered “poor” solvents (low specific Hansen solubility) makes it possible to exfoliate and to fragment thick MoS2 flakes into thin sheets and small particles.
The features of the sonicated samples could give answers to unsolved questions, concerning: (i) the peculiar effects of sonication on both graphite and MoS2; (ii) the most appropriate solvent to be used; (iii) the enhanced properties of the obtained composite when compared to the single components; (iv) the morphology of the resulting composite material, and interactions established between graphite, molybdenum disulphide and solvent; (v) the effectiveness of the sonication in obtaining a composite with a structure characterized by intercalation of alternate MoS2 and graphene layers. Recent investigations have shown that stacked MoS2/graphene hybrids can be obtained48 and that the interlayer coupling, electric fields, and interface/contact regions may affect the electronic structure making these heterobilayers suitable for nanoelectronic devices.49,50 Among these property enhancements, recent computing and calculations have allowed the understanding about the best structure able to optimize interactions in layered materials.51
Taking into consideration these property enhancements, the coupling between graphite oxide (GO) and MoS2 slabs for photocatalysis has been investigated in this paper. In this domain, GO and MoS2 have been simultaneously sonicated in an ethanol/water mixture, for the first time. Notice that GO was used as a starting carbon-based material because of its high oxygen-based functionalization, that could help the establishment of interactions with MoS2 and, then, makes its surface hydrophilic. This allows therefore a more easy dispersion in a water-based solvent. The obtained composite materials have been characterized by means of scanning electron microscopy (SEM), energy dispersive X-rays (EDAX), X-ray diffraction (XRD). The GO/MoS2 composites have also been thermally reduced, in such a way to obtain reduced GO/MoS2 composite (rGO/MoS2). However, the thermal reduction of GO is a very complex process because of the multistep thermal removal of intercalated H2O molecules and oxide groups coming from carboxyl, hydroxyl and epoxy groups.52 The obtained rGO/MoS2 has been also characterized. The catalytic activity of the rGO/MoS2 has been investigated using UV-vis spectroscopy, following the degradation of methylene-blue (MB) under solar-like irradiation.
The dispersion was sonicated at 20 kHz for 6 hours by a VCX 500 Sonics Vibracell ultrasonic processor (power 500 W) equipped with a Ti alloy tapered microtip (d = 3 mm, 30% amplitude). In order to control the temperature and to avoid the evaporation of the solvent, the dispersion was put into an ice bath during the whole sonication step. The obtained dark solution was left overnight at RT to allow the evaporation of EtOH/H2O and the obtained dark powder was then collected to be analysed.
X-ray diffraction (XRD) patterns on MoS2, graphene oxide, GO/MoS2 composite and rGO/MoS2 composite have been collected with a diffractometer (PANalytical PW3050/60 X'Pert PRO MPD) by using a Ni-filtered Cu anode and working with reflectance Bragg–Brentano geometry.
UV-visible spectra have been acquired in the transmission mode at room temperature by using a double-beam UV-vis-NIR spectrophotometer (Varian Cary UV 5000) operating in the wavelength range of 190–1000 nm.
The choice of the reduction temperature for GO/MoS2 composite was made by examining the weight loss vs. temperature curve, as obtained by thermal gravimetric analysis (TGA, TA Instruments, Q600), in N2 atmosphere (Fig. S1, ESI†). It is observed that, starting from 200 °C, a rapid 20% mass loss, due to the removal of the oxygen-containing functional groups,54 in the form of CO, CO2 and steam, has occurred.55
Diffuse reflectance infrared Fourier transform spectroscopy (DRIFT) has been adopted to investigate the presence of functional groups (i.e. oxygen-containing species) on the different samples. Each sample was preventively dried in air at 100 °C. FTIR spectra were collected in air by using 128 scans and a resolution of 4 cm−1, with a Nicolet 6700 spectrophotometer equipped with a DRIFT Smart Accessory and a MCT detector. The reflectance spectra were converted in Kubelka–Munk units.
More in detail, bulk GO (grey pattern, a) shows a main peak at 2θ ≅ 10.5° (d-spacing of 0.84 nm), which indicates that GO retains a layered character without a strict crystalline lattice. As well-known, the d-spacing of GO planes along the c-axis is higher if compared to that of a typical graphite (3.35 Å), due to the presence of oxygen-based functional groups between planes. Besides, bulk MoS2 (black pattern, b) reveals its typical crystalline nature, with narrow peaks, the more intense of which at 2θ = 14.5° is due to the (002) diffraction planes, which corresponds to a d = 6.15 Å (PDF card #037-1492).
Moving to GO/MoS2 composite (red pattern, c), the main diffraction peaks of MoS2 are also observed, whereas a broad feature due to GO is appearing (marked with the asterisk). From this, we conclude that, during the sonication process, the MoS2 crystalline structure is maintained, while GO particles undergo a strong fragmentation, which gives rise to small and amorphous particles, as a result of the superior effectiveness of the sonication on GO rather than on MoS2. From the Scherrer's equation (L = Kλ/βcosθ, where: λ is the X-ray wavelength, β is the FWHM of the diffraction line corrected by the instrumental broadening, θ is the diffraction angle, and K is a shape constant, which has been assumed to be 0.76), the mean crystallite thickness of the MoS2 particles was calculated (for bulk MoS2 and for GO/MoS2). More in details, from the 2θ = 14.5° XRD peak, assigned to the (002) MoS2 crystalline planes, nanocrystals about 60 nm (99 layers) in thickness are obtained for bulk MoS2, while nanocrystals of about 18 nm (30 layers) are evaluated for GO/MoS2 after sonication. From this, a moderate effect of the sonication step on the size reduction of MoS2 is evidenced.
Finally, moving to rGO/MoS2 composite (green pattern, d), the GO feature is totally absent, while a small new feature is arising at 2θ ≅ 24.4°, corresponding to a d-spacing between the graphene planes of about 3.6 Å, which is closer to the position of (002) diffraction planes of hexagonal graphite (d = 3.35 Å). Besides the peak position, which is indicative of the crystal structure and symmetry of the contributing phase(s), a low crystallinity (i.e. high-defective character) of the rGO phase, can be highlighted from the FWHM (full-width-at-half-maximum) of the XRD band at 2θ ≅ 24.4°.56 From this, we can hypothesize only a partial restoration of the graphitic phase, as a consequence of the thermal treatment at 400 °C. It is worth noticing that the thermal treatment leads to an increase of the MoS2 particle thickness, up to about 34 nm (57 layers), as calculated by applying the Scherrer's equation.
The GO/MoS2 composite, as obtained after sonication in EtOH/H2O and evaporation of the solvent is SEM imaged in Fig. 2a and b. From these, a quite complex morphology of the composite is obtained, with aggregates of differently oriented platelets, having a heterogeneous distribution of sizes, ranging from some μm to a few nanometers. Nevertheless, the layered structure of the sample, made by the packing of bidimensional MoS2 and GO sheets can be highlighted.
More detailed information is obtained from the HRTEM images shown in Fig. 2c and d.
The in-focus image of a portion of GO/MoS2 composite reveals the presence of highly crystalline particles, whose interference fringes, highlighted by white lines, 6.1 Å and 2.7 Å spaced, are unequivocally associated with MoS2 (002) and (1−10) planes, respectively (Fig. 2c, left panel). Contextually, the same region of the sample is under-focus imaged (Fig. 2c, right panel), allowing to distinguish the amorphous portion of the composite, whose edge regions highlighted by white arrows, embed the MoS2 particle. Therefore, the non-crystalline area can be identified with GO decorating and covering the MoS2 surface, as confirmed by the before discussed XRD patterns, from which the amorphous nature of GO particles, due to strong sonication effects, has been shown.
Besides, rGO/MoS2 composite, as obtained after the thermal treatment at 400 °C, is HRTEM imaged in Fig. 2d, where white lines highlight lattice fringes 2.7 Å spaced, associated with (1−10) planes of few-layer thick crystalline MoS2. The MoS2 structure is confirmed by the corresponding fast-Fourier-transform (FFT) imaged in the inset at the top of Fig. 2d, which shows bright spots associated with {1−10} plane families. In the partially amorphous border region, interference fringes (white and light blue dots) 3.7 Å spaced, as shown in the inset at the bottom of the figure, are arising, due to the building up of graphitic phases, are observed in agreement with XRD investigations.
A model of the arrangement of GO and of rGO on MoS2 slabs is shown in Fig. 3, which summarizes the previously discussed HRTEM results.
MoS2 shows the A, B, C and D typical excitonic peaks (15000 cm−1, 16500 cm−1, 22300 cm−1 and 25100 cm−1 respectively), whose nature is well described in literature.47 Furthermore, GO presents a first band at about 43860 cm−1 attributed to π → π* transitions of CC and a second one at about 32680 cm−1 attributed to n → π* transitions of CO.57 Moving to the spectrum of GO/MoS2 composite (red line), all the typical features described for MoS2 and GO are observed, while rGO/MoS2 is highly absorbing in the whole 45000–10000 range as expected for a system characterized by extended CC sp2 domains.58 It is noteworthy that the GO bands are absent, meaning that the thermal treatment leads to the loss of functional groups containing oxygen and to the partial restoration of the sp2 conjugation. Some more the spectrum of rGO/MoS2 does not show clearly the typical MoS2 C and D excitonic bands (at 22300 cm−1 and 25100 cm−1 respectively) superimposed to graphene background. This fact is likely associated with photoluminescence effects that have been reported for MoS2 in monolayer and multilayer forms.41,59–63 The MoS2 A and B excitonic bands at 15000 cm−1 and 16500 cm−1 respectively appear to be “overturned” on rGO/MoS2 when compared to MoS2 and GO/MoS2. Specifically, a maximum of a peak of MoS2 (black line) corresponds exactly to a minimum of the same peak of rGO/MoS2 (green line). Although a detailed and deep study of this phenomenon is beyond the aim of this work, we hypothesize that the inversion of the shape of the bands can be due to specular reflectance effects, becoming important for a very highly absorbing material made by poorly dispersed particles with large dimensions.64 That unsufficient dispersion can be also inferred from the consideration that rGO/MoS2 derives from GO/MoS2 that was subjected to: removal of the solvent after sonication, thermal reduction treatment and finally redispersion in solvent for UV-vis measurements. The whole procedure plausibly undoes the effects of dispersion and exfoliation obtained during sonication.
Some more, to evaluate the stability of GO/MoS2 dispersion in EtOH/H2O, UV-vis spectra of the composite, just sonicated and then after 2, 4 and 9 days, were acquired. The obtained results are shown in Fig. 4b. In this figure and in the inset therein, no significant shift and intensity decrease of A and B excitonic peaks are observed. As reported by some authors, the position of the A and B bands is affected by the particles sizes due to a quantum size effect.47,65 Then, we can state that for our samples no restacking or strong precipitation and deposition phenomena are occurring through time, being the particles durably well dispersed in the solvent. This result is in agreement with the fact that the 45% vol EtOH/H2O mixture has resulted to be efficient in obtaining a good dispersion degree of particles, according to its Hansen solubility parameters.53
DRIFT spectroscopy has been adopted to investigate the presence of functional groups (i.e. oxygen species) on the different sample surfaces. DRIFT spectra of GO (gray line), GO/MoS2 (red line), and of rGO/MoS2 (green line) are compared with the spectrum of the bulk native MoS2 (black line) in the 3900–1300 cm−1 region (Fig. 5). Several common fingerprints for GO and GO/MoS2 can be highlighted. In particular, the wide absorption in the 3700–2500 cm−1 interval is associated with ν(OH) of alcoholic/phenolic and of carboxylic vibrational modes, while the absorption bands at 1800–1700 cm−1 and at 1640–1590 cm−1 are assigned to ν(CO) and at ν(CC) stretching modes. It is noteworthy that the intensity of the conjugated C-sp2 bonds belonging to graphitic islands, would be enhanced by the presence of oxygen atoms (i.e. increase of the dipole moment).66 As the intensity of ν(CO) and of ν(CC) vibrational modes is nearly vanishing for the rGO/MoS2 sample, it is concluded that the quantity of the polar oxygen groups is drastically reduced upon the thermal treatment as reported in literature.55,67 It is concluded that DRIFT spectroscopy provides a valuable and sensitive tool to gain information on the population of the polar groups on carbon and carbon hybrids and on its substantial decrement upon reductive treatment.
Fig. 5 DRIFT spectra of: (a) GO (grey line), (b) GO/MoS2 sonicated in EtOH/H2O (red line), (c) rGO/MoS2 obtained at 400 °C (green line), and of (d) bulk MoS2 (black line). |
By starting from an initial dye concentration of 12.5 mg L−1, the MB band evolution of the rGO/MoS2 composite, as a function of the exposure time under visible light, is shown in Fig. 6a.
In this figure it is shown that the intensity of the two main MB bands at 15100 cm−1 and at 16450 cm−1, which are assigned to monomeric and aggregated species,69,70 is decreasing with the exposure time. A series of experiments has been also conducted by using different initial concentration of MB (12.5 mg L−1, 6.25 mg L−1 and 3.2 mg L−1) on rGO/MoS2 composite (Fig. S2, ESI†), thus showing that depending on the concentration, MB in solution may show a distinct tendency to form agglomerates, made by monomeric and polymeric species, in thermodynamic equilibrium.71 In Fig. 6b the MB photodegradation performances of the rGO/MoS2 composite are compared to those of MoS2, of rGO and of the well-known P25 TiO2 photocatalyst, used as reference materials. It has been calculated that after 5 h under solar-light irradiation, the residual amount of the initial dye was about 4% for rGO/MOS2, which is close to 1.2% P25 performance, as compared to the bare rGO and MoS2 (10% and 20%, respectively). Although the activity of P25 TiO2 is definitely higher than that of composite, rGO/MoS2 composite shows a strong increment in the MB photodegradation, if compared to pure MoS2 and pure rGO.
As for the explanation of such an improvement, two points could be highlighted: the π–π conjugation between MB and aromatic regions of graphene domains and the step-wise structure of energy levels constructed in the MoS2/graphene composite (Fig. 7).72 The same considerations can be done in the case of the degradation of other organic dyes, due to their structure affinity with MB.
According to the data reported in literature on the conduction band, the valence band of MoS2 (ref. 39) and the work function of graphene,73 the energy levels are beneficial to transfer photo-induced electrons from the MoS2 conduction band to the graphene, which could efficiently separate the photo-induced electrons and hinder the charge recombination in the electron-transfer processes.74 In conclusion, rGO/MoS2 composites have higher photocatalytic performance due to: (i) facilitated electron transfer and separation; (ii) enhancement of the light absorption intensity as consequence of the introduction of rGO; (iii) increase of the adsorption of pollutants.74
Finally, it has been shown the enhanced photocatalytic activity toward the photodegradation of methylene blue of rGO/MoS2, as compared to pure MoS2, which exhibits a quick recombination of photo-generated charge carriers.
This has been explained with the remarkable role of rGO, an excellent electron acceptor/transport material, in separating the photo-induced electrons and in hindering the charge recombination during the electron transfer process, thus enhancing the photocatalytic performances.
Footnote |
† Electronic supplementary information (ESI) available: Additional figure and material characterization (TGA in N2). See DOI: 10.1039/c6ra08633k |
This journal is © The Royal Society of Chemistry 2016 |