Nanostructured MoS2/BiVO4 Composites for Energy Storage Applications

We report the optimized synthesis and electrochemical characterization of a composite of few-layered nanostructured MoS2 along with an electroactive metal oxide BiVO4. In comparison to pristine BiVO4, and a composite of graphene/BiVO4, the MoS2/BiVO4 nanocomposite provides impressive values of charge storage with longer discharge times and improved cycling stability. Specific capacitance values of 610 Fg−1 (170 mAhg−1) at 1 Ag−1 and 166 Fg−1 (46 mAhg−1) at 10 Ag−1 were obtained for just 2.5 wt% MoS2 loaded BiVO4. The results suggest that the explicitly synthesized small lateral-dimensioned MoS2 particles provide a notable capacitive component that helps augment the specific capacitance. We discuss the optimized synthesis of monoclinic BiVO4, and few-layered nanostructured MoS2. We report the discharge capacities and cycling performance of the MoS2/BiVO4 nanocomposite using an aqueous electrolyte. The data obtained shows the MoS2/BiVO4 nanocomposite to be a promising candidate for supercapacitor energy storage applications.

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 BiVO 4 /MoS 2 composite, with the first ever usage of bare MoS 2 as the capacitive component. The MoS 2 has been synthesized from thermolysis of (NH 4 ) 2 MoS 4 in the presence of H 2 . The resulting MoS 2 /BiVO 4 nanocomposite has demonstrated much larger values of charge storage, longer discharge times and improved cycling stability in comparison to pure BiVO 4 itself, or graphene/BiVO 4 composites. NS MoS 2 /BiVO 4 composites have been found to be superior to graphene/BiVO 4 composites, and hence a promising candidate for supercapacitor applications.

Results and Discussion
In this study, three types of samples were investigated: pure BiVO 4 , a composite of 2.5 wt% graphene/BiVO 4 and a composite of 2.5 wt% NS MoS 2 /BiVO 4 . These are labeled as BiVO 4 , BiVO 4 -G and BiVO 4 -M for succinctness in the following sections. Details of the synthesis of all these samples are discussed in the methods section. BiVO 4 was synthesized by a solvothermal route. The purity and crystallinity of solvothermally synthesized BiVO 4 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 BiVO 4 . It can be seen that 10-20 nm diameter spherical BiVO 4 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 BiVO 4 . 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 BiVO 4 , and the bands at 367 and 327 cm −1 are ascribed to the symmetric and asymmetric deformation modes of the VO 4 3− tetrahedron, respectively 25 . Figure 2a shows the XRD profile of the nanostructured MoS 2 grown on sapphire via a two-step thermolysis process. The diffraction pattern from the sample ((002) peak) is indexed to MoS 2 , 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 N 2 and sulfur. The addition of sulfur before the second annealing process greatly improved the crystallinity 26 . An atomic force microscopy (AFM) image of the MoS 2 /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 MoS 2 , Fig. S1. The lateral dimension of the flakes are about 50-200 nm. Figure 2c shows the Raman spectra for MoS 2 flakes of different thickness on sapphire substrates, obtained using laser excitation at 532 nm. The energy difference (∆) between two Raman peaks (modes A 1g and E 2g ) can be used to identify the number of MoS 2 layers on the substrate 26 . 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 MoS 2 (mono/ bi/tri-layer) and bulk respectively 26 . 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 MoS 2 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 (8μ m x 8μ m) of MoS 2 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 MoS 2 ) to a maximum value of 384.8 cm −1 (ideally belonging to few layers of MoS 2 ). As such, most of the image depicts that the sample shows a Raman signal closer to that of few layers of MoS 2 . From combined information provided by Raman mapping and AFM, it can therefore be concluded that the MoS 2 film on the substrate consists of mostly nanostructured MoS 2 and a few thicker flakes (≥ 5 monolayers).
To effectively recover nanostructured MoS 2 from sapphire, the sample was simply peeled off using a surgical blade. Following this recovery, the MoS 2 /BiVO 4 composite was prepared via ultrasonication. 2.5wt% NS MoS 2 was loaded on BiVO 4 . Analogously, a graphene/BiVO 4 composite was also prepared by ultrasonicating commercially-procured graphene with BiVO 4 . The two components in the composite (the 2D layered material and BiVO 4 ) are expected to interact via van der Waals' interactions. Figure 3a, shows XRD profiles of BiVO 4 -G and BiVO 4 -M, all the peaks can be indexed to BiVO 4 . XRD patterns show well resolved peaks of monoclinic BiVO 4 , and there were no deleterious effects observed due to ultrasonication. The peaks due to diffraction from the MoS 2 lattice planes were however not observed in the diffractogram, perhaps due to the very small amount of MoS 2 present and lack of a preferential orientation following delamination from the original sapphire substrate.  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 (Bi 3+ ↔ Bi 0 ). For bare BiVO 4 , a single peak with high current was assigned to reduction of Bi 3+ to Bi 0 (− 0.85 V), on the other side, two anodic peaks were obtained at -0.52 V and -0.4 V for the oxidation 27,28 . The oxidation peak A1 is assigned to the oxidation of Bi 0 to Bi + and peak A2 is assigned to the oxidation of Bi + to Bi 3+ . Further, the CV profiles for BiVO 4 -G and BiVO 4 -M in Fig. 4a matched well with CV curve of pristine BiVO 4 indicating redox peaks were due to BiVO 4 and the presence of MoS 2 or graphene did not affect its electrochemical response. No extra peaks are visible for BiVO 4 -M suggesting that the small fraction of MoS 2 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 .
The electrochemical capacitance of BiVO 4 , BiVO 4 -M and BiVO 4 -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 2 nd cycle in each case. CD profile of BiVO 4 indicated pseudo-capacitive nature of material for charge storage applications. Similar to BiVO 4, CD curves of BiVO 4 -M and BiVO 4 -G (Fig. 4b) between potential  specific capacitance values were obtained. Specific capacitance of BiVO 4 -M was found to be (212 Fg −1 at 3 Ag −1 ) which is not only higher than bare BiVO 4 (108 Fg −1 at 3 Ag −1 ) but at higher current densities it is surpassing specific capacitance obtained for BiVO 4 -G (Table S1). Figure 4b shows CD profiles of pristine BiVO 4 , BiVO 4 -M, and BiVO 4 -G at a current density 7 Ag −1 , which clearly indicates nanostructured MoS 2 composite with BiVO 4 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 polarizability 30 . For comparison, 2.5 wt% bulk MoS 2 loaded BiVO 4 was also prepared via ultrasonication and electrochemical behavior of the same was studied. Although the capacitance values of bulk MoS 2 /BiVO 4 were higher than nanostructured MoS 2 /BiVO 4 , 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 BiVO 4 -Bulk MoS 2 composite. On the other hand, BiVO 4 -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 BiVO 4 -M composite can be ascribed to the high elasticity of nanostructured MoS 2 . As depicted by the AFM topographic image, the synthesis method reported here yielded MoS 2 with a lateral particle size of a few 100 nm. It has been purported that as the lateral size of MoS 2 is decreased, there is a higher preponderance of step-edges, and low-coordination edge and corner atoms as compared to basal plane atoms 31 . 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 MoS 2. 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 MoS 2 appear to contribute substantially to the charge storage, discharge times and cycling stability of the MoS 2 /BiVO 4 composite synthesized in this report.

Methods
Analytical reagent grade ammonium metavanadate (NH 4 VO 3 ), 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(NO 3 ) 3 .5H 2 O) 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-BiVO 4 . BiVO 4 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(NO 3 ) 3 .5H 2 O and 1 mmol NH 4 VO 3 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 MoS 2 . Synthesis of few layered MoS 2 was successfully achieved by following a published procedure with major modifications 26 . 25 mg of (NH 4 ) 2 MoS 4 was dissolved in 20 mL DMF and sonicated for 20 minutes. 10 μ L 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 (NH 4 ) 2 MoS 4 layer was placed on the graphite sample holder in the tube furnace flowing with a gas mixture N 2 /H 2 . It was then heated at 500 °C and was maintained for 60 minutes under constant flow of N 2 /H 2 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 N 2 and after 20 minutes purging to remove any air introduced, the temperature was raised to 1000 °C and was maintained for 30 minutes. 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