Introduction

The ever-increasing global energy demands have spurred increased research into energy harvesting and storage systems1,2. The development of effective energy storage systems with high energy density as well as high power density is becoming increasingly important. Electrochemical capacitors, also termed as supercapacitors, have attracted significant interest as these devices bridge the energy density gap between conventional capacitors and batteries. Layered inorganic systems exhibit unusual properties that are technologically important. The unique mechanical, electronic, thermal and optical properties of graphene and other two-dimensional layered materials like the transition metal dichalcogenides (TMDCs) such as molybdenum disulfide (MoS2) have enabled them to be utilized for various novel applications3,4,5,6. One specific application for which these layered materials have been recently explored is in energy storage devices, such as lithium-ion batteries and supercapacitors7,8. The recent proliferation of research into 2D layered chalcogenides is the result of their intrinsic high ionic conductivity, high surface area, inherent chemical stability (under a variety of pH conditions) and propensity for charge storage9,10.

In this work, an electroactive metal oxide BiVO4 is combined with few-layered nanostructured (NS) MoS2 as a capacitive component to form a hybrid structure. It should, however, be noted that over the last few years several studies that quantify the capacitive storage of MoS2 have been reported11,12,13,14,15,16,17,18,19,20,21,22; these typically use MoS2 in conjunction with a variety of carbon allotropes, predominantly graphene (r-GO). Interestingly, as yet MoS2 has not been coupled with a viable electrochemical component to form a hybrid supercapacitor. Table 1 shows a summary of a few different approaches; the specific capacitance values vary between 100 and 550 Fg−1 with an average of ca. 250 Fg−1. Moreover this value is highly dependent on type of MoS2 (bulk vs. nanostructured), manner of synthesis of nanostructured MoS2 (exfoliated vs. synthesized from precursors), current density employed and the amount of loading of MoS2 to the carbon source.

Table 1 Comparison of supercapacitor performance of MoS2/carbon nanostructures and bare MoS2.

Another approach to augment the specific capacitance of electric double-layer capacitors is to incorporate a new electroactive material along with a potent capacitive component to enable high energy density supercapacitors while maintaining high power operations. It has been reported that MoS2 suffers from low conductivity, low theoretical specific capacity and easy restacking of the sheets23,24. Despite the aforementioned limitations of MoS2, it is being employed in electrochemical devices albeit in conjunction only with carbon nanostructures so as to overcome these limitations.

As mentioned above, here we report a BiVO4/MoS2 composite, with the first ever usage of bare MoS2 as the capacitive component. The MoS2 has been synthesized from thermolysis of (NH4)2MoS4 in the presence of H2. The resulting MoS2/BiVO4 nanocomposite has demonstrated much larger values of charge storage, longer discharge times and improved cycling stability in comparison to pure BiVO4 itself, or graphene/BiVO4 composites. NS MoS2/BiVO4 composites have been found to be superior to graphene/BiVO4 composites, and hence a promising candidate for supercapacitor applications.

Results and Discussion

In this study, three types of samples were investigated: pure BiVO4, a composite of 2.5 wt% graphene/BiVO4 and a composite of 2.5 wt% NS MoS2/BiVO4. These are labeled as BiVO4, BiVO4-G and BiVO4-M for succinctness in the following sections. Details of the synthesis of all these samples are discussed in the methods section.

BiVO4 was synthesized by a solvothermal route. The purity and crystallinity of solvothermally synthesized BiVO4 was analyzed by high-resolution X-ray diffraction (XRD), Fig. 1a. The peaks in the diffractogram could be indexed according to JCPDS card no. 75–2480, corresponding to a monoclinic scheelite structure. The inset in Fig. 1a shows an SEM image of BiVO4. It can be seen that 10–20 nm diameter spherical BiVO4 particles agglomerate to form larger porous oval shaped clusters with average size in the range 200–300 nm. Figure 1b shows a Raman spectrum of BiVO4. The most intense Raman band at around 823 cm−1 is assigned to the V-O symmetric stretching mode, while the weak shoulder at around 712 cm−1 is due to V-O antisymmetric stretch. The band at 211 cm−1 is related to external mode (translation/rotation) of BiVO4, and the bands at 367 and 327 cm−1 are ascribed to the symmetric and asymmetric deformation modes of the VO43− tetrahedron, respectively25.

Figure 1
figure 1

(a) XRD pattern of BiVO4, in the inset: SEM image of solvothermally synthesized BiVO4 and (b) Raman spectrum indicating various vibrational modes of BiVO4.

Figure 2a shows the XRD profile of the nanostructured MoS2 grown on sapphire via a two-step thermolysis process. The diffraction pattern from the sample ((002) peak) is indexed to MoS2, indicating periodicity along the c-axis. All other peaks in the XRD correspond to c-sapphire (substrate). The samples, after the first annealing step, were subjected to a second annealing at 1000 °C in the mixture of N2 and sulfur. The addition of sulfur before the second annealing process greatly improved the crystallinity26.

Figure 2
figure 2

(a) XRD showing (002) peak of MoS2, (b) AFM topographic image of nanostructured MoS2 grown on sapphire, and (c) Raman spectra of bilayer, trilayer and bulk MoS2.

An atomic force microscopy (AFM) image of the MoS2/sapphire sample obtained after second annealing in the presence of sulfur is shown in Fig. 2b. From the height of the flakes we can infer the presence of monolayer, bilayer and trilayer regions along with bulk MoS2, Fig. S1. The lateral dimension of the flakes are about 50–200 nm.

Figure 2c shows the Raman spectra for MoS2 flakes of different thickness on sapphire substrates, obtained using laser excitation at 532 nm. The energy difference (∆) between two Raman peaks (modes A1g and E2g) can be used to identify the number of MoS2 layers on the substrate26. The values of ∆ obtained for different parts of the sample are in the range 21–23 cm−1 and 25 cm−1, which suggests the presence of nanostructured MoS2 (mono/bi/tri-layer) and bulk respectively26. It should be noted that normally the spatial resolution of the Raman spectra, being limited by the laser spot size, is poorer than that of an AFM. So the Raman signal is actually averaged over different MoS2 flakes even when it is acquired from a single ‘spot’. We have therefore observed that it is difficult to differentiate a monolayer flake from a bilayer flake if the lateral sizes of the flakes are too small using only Raman spectroscopy. As such, further characterization was done by Raman imaging of a larger area (8μm x 8μm) of MoS2 on sapphire sample (Supporting Information Fig. S2). The image was acquired using a range of minimum value of 383.6 cm−1 (belonging to ideally bulk MoS2) to a maximum value of 384.8 cm−1 (ideally belonging to few layers of MoS2). As such, most of the image depicts that the sample shows a Raman signal closer to that of few layers of MoS2. From combined information provided by Raman mapping and AFM, it can therefore be concluded that the MoS2 film on the substrate consists of mostly nanostructured MoS2 and a few thicker flakes (≥5 monolayers).

To effectively recover nanostructured MoS2 from sapphire, the sample was simply peeled off using a surgical blade. Following this recovery, the MoS2/BiVO4 composite was prepared via ultrasonication. 2.5wt% NS MoS2 was loaded on BiVO4. Analogously, a graphene/BiVO4 composite was also prepared by ultrasonicating commercially-procured graphene with BiVO4. The two components in the composite (the 2D layered material and BiVO4) are expected to interact via van der Waals’ interactions. Figure 3a, shows XRD profiles of BiVO4-G and BiVO4-M, all the peaks can be indexed to BiVO4. XRD patterns show well resolved peaks of monoclinic BiVO4, and there were no deleterious effects observed due to ultrasonication. The peaks due to diffraction from the MoS2 lattice planes were however not observed in the diffractogram, perhaps due to the very small amount of MoS2 present and lack of a preferential orientation following delamination from the original sapphire substrate. Figure 3b shows Raman spectrum of BiVO4-G obtained using laser excitation at 532 nm. All vibrational modes of BiVO4 are observed along with vibrational signatures of graphene (D- and G- bands). Figure 3c and d show SEM images of BiVO4-G and BiVO4-M composites respectively. These images clearly show the two separate phases of capacitive components and BiVO4. The energy-dispersive X-ray pattern for BiVO4-M confirmed the presence of Bi, V and Mo (Fig. S3).

Figure 3
figure 3

(a) XRD profiles of BiVO4-G and BiVO4-M, all peaks can be indexed to BiVO4 which indicates chemical integrity of BiVO4 is intact, (b) Raman spectrum of BiVO4-G showing vibrational features of both BiVO4 and graphene, (c,d) show SEM images of BiVO4-G and BiVO4-M.

Electrochemical characterization

The electrochemical behavior of all three samples (BiVO4, BiVO4-M and BiVO4-G) was evaluated by performing cyclic voltammetry, charge discharge and cycling stability, using the same working electrode.

Cyclic voltammetry measurements were performed at various scan rates from 5 to 50 mVs−1, the data presented in Fig. 4a is for 20 mVs−1 scans (the other scans are shown in the supporting information). Redox peaks in the CV curve can be attributed to quasi-reversible faradaic process (Bi3+ ↔ Bi0). For bare BiVO4, a single peak with high current was assigned to reduction of Bi3+ to Bi0 (−0.85 V), on the other side, two anodic peaks were obtained at -0.52 V and -0.4 V for the oxidation27,28. The oxidation peak A1 is assigned to the oxidation of Bi0 to Bi+ and peak A2 is assigned to the oxidation of Bi+ to Bi3+. Further, the CV profiles for BiVO4-G and BiVO4-M in Fig. 4a matched well with CV curve of pristine BiVO4 indicating redox peaks were due to BiVO4 and the presence of MoS2 or graphene did not affect its electrochemical response. No extra peaks are visible for BiVO4-M suggesting that the small fraction of MoS2 did not show any charge transfer activity in the applied potential window that was operative. With increasing scan rates the oxidation peak potentials shift towards positive direction and reduction peak potentials shift towards negative direction, which is mainly attributed to IR drop component, which dominates at higher current values. The kinetics of interfacial faradaic redox reaction was rapid enough, as an increase in current response at higher scan rates was observed, (see Fig. S4a in the supporting information)28.

Figure 4
figure 4

(a) CV curves of BiVO4, BiVO4-G and BiVO4-M at a scan rate of 20 mVs−1, (b) CD profile of BiVO4, BiVO4-G and BiVO4-M at a current density of 7 Ag−1.

The electrochemical capacitance of BiVO4, BiVO4-M and BiVO4-G was evaluated by galvanostatic charge discharge (CD) measurements. For all the samples, 5 cycles for all current densities were carried out; Fig. 4b presents the data of the 2nd cycle in each case. CD profile of BiVO4 indicated pseudo-capacitive nature of material for charge storage applications. Similar to BiVO4, CD curves of BiVO4-M and BiVO4-G (Fig. 4b) between potential window −1.0 V to 0.0 V appeared to be non-symmetric, insinuating battery behavior and showing the IR drop. Despite the steep voltage drop, prolonged plateau of voltage output was observed, which is due to the involvement of faradaic process in BiVO4. Using equation, specific capacitance values were obtained. Specific capacitance of BiVO4-M was found to be (212 Fg−1 at 3 Ag−1) which is not only higher than bare BiVO4 (108 Fg−1 at 3 Ag−1) but at higher current densities it is surpassing specific capacitance obtained for BiVO4-G (Table S1). Figure 4b shows CD profiles of pristine BiVO4, BiVO4-M, and BiVO4-G at a current density 7 Ag−1, which clearly indicates nanostructured MoS2 composite with BiVO4 has maximum specific capacitance. Generally, the boost of current density results in fading of specific capacitance which is primarily due to inaccessibility of inner electroactive sites to the electrolyte ions due to diffusion limitations (Fig. S4b)29. However, even at higher current densities the specific capacitance values are found to be 166.6 Fg−1 and 156 Fg−1 at 10 Ag−1 and 15 Ag−1, respectively, which are impressive. High capacitance values at faster charging rates can be explained by the fact that transition metal dichalcogenides show higher ionic diffusivity as a consequence of large anionic polarizability30. For comparison, 2.5 wt% bulk MoS2 loaded BiVO4 was also prepared via ultrasonication and electrochemical behavior of the same was studied. Although the capacitance values of bulk MoS2/BiVO4 were higher than nanostructured MoS2/BiVO4, the composite with the bulk appears to show poor cycling stability (Fig. 5). It was observed that discharge capacity was retained only 10% up to 200 cycles for BiVO4-Bulk MoS2 composite. On the other hand, BiVO4-M exhibited better long term stability, as it could retain ~80% of the initial value after 200 cycles, for the same current density used for charging (3 Ag−1). The excellent structural and mechanical stability shown by BiVO4-M composite can be ascribed to the high elasticity of nanostructured MoS2. As depicted by the AFM topographic image, the synthesis method reported here yielded MoS2 with a lateral particle size of a few 100 nm. It has been purported that as the lateral size of MoS2 is decreased, there is a higher preponderance of step-edges, and low-coordination edge and corner atoms as compared to basal plane atoms31. The effect of these states dominate over those of the basal atoms and can contribute to higher charge storage sites. In addition, since the particles are made up of a few-layered sheets, it has been highlighted by Chhowalla et al.9 that such loosely stacked sheets are able to accommodate structural changes in a better manner upon cycling when compared to bulk MoS2. The latter has shown large structural instability especially as anodes in lithium ion batteries. As such, when evaluated in unison, the explicitly synthesized small lateral-dimensioned particles which consist of only a few layers of MoS2 appear to contribute substantially to the charge storage, discharge times and cycling stability of the MoS2/BiVO4 composite synthesized in this report.

Figure 5
figure 5

Cycling performance in terms of specific capacitance (%) of Bulk MoS2/BiVO4 and BiVO4-M at a current density 3 Ag−1.

Methods

Analytical reagent grade ammonium metavanadate (NH4VO3), sodium dodecyl sulfate (SDS) and anhydrous glycerol were purchased from Merck, India. Polyvinlydene fluoride (PVDF), N-methyl-2-pyrrolidinone (NMP) and activated carbon (AC) were purchased from Himedia, India whereas bismuth nitrate pentahydrate (Bi(NO3)3.5H2O) was purchased from Sigma Aldrich. Research grade graphene dispersion in water was procured from US Research Nanomaterials Inc. All chemicals were used as received without further purification.

Synthesis of monoclinic-BiVO4

BiVO4 was prepared by following earlier work reported by Khan et al.28 Typically, 3 mmol of SDS was dissolved in 40 mL of solvent (DI water and glycerol with volume ratio 1:1) in a flask at room temperature. 1 mmol Bi(NO3)3.5H2O and 1 mmol NH4VO3 was added to the above clear solution under constant stirring in sequence. After stirring for 7 minutes, the solution was transferred into a stainless steel autoclave with a Teflon liner and heated at 160 °C for 18 hours. After cooling to room temperature, the reaction mixture was centrifuged and the pellet was washed with water and ethanol. Finally, the product was dried under vacuum at 70 °C for ca. 5 hours.

Synthesis of nanostructured MoS2

Synthesis of few layered MoS2 was successfully achieved by following a published procedure with major modifications26. 25 mg of (NH4)2MoS4 was dissolved in 20 mL DMF and sonicated for 20 minutes. 10 μL of this precursor solution was spin coated on sapphire at 3000 rpm for 60 seconds. The substrate was then heated on a hot plate at 120 °C for 30 minutes. The annealing process was performed in a homemade rapid thermal annealing furnace. The freshly prepared thin (NH4)2MoS4 layer was placed on the graphite sample holder in the tube furnace flowing with a gas mixture N2/H2. It was then heated at 500 °C and was maintained for 60 minutes under constant flow of N2/H2 to efficiently remove residual solvent, ammonia molecules, and other byproducts from the precursor. After this step, the furnace was cooled down to room temperature, following which additional sulfur was introduced to the sample holder and the gas environment was changed to N2 and after 20 minutes purging to remove any air introduced, the temperature was raised to 1000 °C and was maintained for 30 minutes.

Synthesis of 2.5 wt% nanostructured MoS2/BiVO4 composite

0.2 mg of nanostructured MoS2 was recovered via scraping with a surgical blade and was dispersed in 15 mL water. To this 8 mg BiVO4 was added, and the dispersion was sonicated for 8 hours. The composite was recovered after centrifugation and dried under vacuum at 70 °C for 5 hours.

Synthesis of 2.5 wt% graphene/BiVO4 composite

1 mg of graphene and 40 mg of BiVO4 were dispersed in 30 mL ethanol, and the dispersion sonicated for 8 hours. The composite was recovered by evaporating the solvent on a hot plate at 90 °C.

Characterization

Powder X-ray diffraction (XRD) measurements from 10° to 70° 2θ were recorded using a PANalytical X’pertpro diffractometer with monochromatic Cu Kα source (λ = 1.54056 Å) operating at 40 kV and 30 mA. The elemental composition and surface morphologies of the samples were investigated by Field Emission Scanning Electron Microscopy (FE-SEM) on a Zeiss Ultra FEG 55 instrument at 5 kV operating voltage. Cyclic voltammetry and galvanostatic charge-discharge (CD) studies of the composites were carried out using Bio-Logic VMP3 Galvanostat/Potentiostat Instruments at room temperature.

Electrode material and electrochemical tests

Electrochemical measurements were carried out in a 2 M aqueous NaOH in a three electrode cell at room temperature. Hg/HgO/Ca(OH)2 and platinum plate were used as reference and counter electrodes, respectively. The active electrode was prepared by mixing electroactive material (80 wt%), activated carbon (15 wt%) and polyvinylidene fluoride (5 wt%) with 1 mL of NMP to form a slurry which was coated and dried on a small piece of graphite plate (area of coating, 1 cm2).

Additional Information

How to cite this article: Arora, Y. et al. Nanostructured MoS2/BiVO4 Composites for Energy Storage Applications. Sci. Rep. 6, 36294; doi: 10.1038/srep36294 (2016).

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