Green synthesis of in situ electrodeposited rGO/MnO2 nanocomposite for high energy density supercapacitors

This paper presents the preparation of in situ electrodeposited rGO/MnO2 nanocomposite as a binder-free electrode for supercapacitor application. The work describes and evaluates the performance of prepared electrode via green and facile electrodeposition technique of in situ rGO/MnO2-glucose carbon nanocomposites. The carbon content in the composite electrode increased after GO and D (+) glucose solution has been added in the deposition electrolyte. This study found that a suitable concentration of D (+) glucose in the deposition electrolyte can slow down the nucleation process of MnO2 particles and lead to uniform and ultrathin nanoflakes structure. The optimize electrode exhibited low transfer resistance and resulted on excellent electrochemical performance in three electrolyte systems viz. Na2SO4, KOH and KOH/K3Fe(CN)6 redox electrolytes. The optimum energy density and power density were 1851 Whkg−1 and 68 kWkg−1 at current density of 20 Ag−1 in mixed KOH/K3Fe(CN)6 electrolyte.

Scientific RepoRts | 5:16195 | DOi: 10.1038/srep16195 is being investigated. Graphene oxide holds great potential to be coupled with MnO 2, because it has high conductivity, good chemical stability, and a large surface area. In addition, the surfaces of graphene are capable of a reversible pseudo-reaction and electrochemical double layer formation, which is beneficial to the electrochemical performance of MnO 2 /graphene oxide composite materials 1,8,12,13 .
MnO 2 /graphene oxide composite electrode material synthesized by a hydrothermal method at various reaction times exhibited an optimized specific capacitance (SC) of 213 Fg −1 at current density of 0.1 Ag −1 13 . Under the same method, Deng et al. prepared MnO 2 nanorods/graphene composites and demonstrated that the composites ratio of MnO 2 on the graphene sheets is very important in order to obtain good electrochemical performance of the supercapacitor materials. The achieved specific capacitance is 218 Fg −1 and was determined by a cyclic voltammetry method at a scan rate of 5 mVs −1 in 1 M Na 2 SO 4 aqueous solution 14 . In the work of Sawangphruk et al., MnO 2 -rGO nanocomposites on graphitized carbon fiber paper was produced by a simple spray coating technique and exhibited a SC of 393 Fg −1 determined by a cyclic voltammetry method at a scan rate of 10 mVs −1 in 0.5 M Na 2 SO 4 . This high capacitance value is most likely attributed to the synergistic effect originating from the high surface area of MnO 2 nanoparticles, the high conductivity of rGO nanosheets, and the high porosity of MnO 2 -rGO nanosheets coated on CFP 12 . The composite electrode consisting of rGO and MnO 2 nanoneedles prepared through hydrazine hydrate-mediated reduction of graphene oxide (rGO)/MnO 2 displayed the highest SC among the composite electrodes studied with a SC as high as 371.74 Fg −1 at a scan rate of 10 mVs −1 . Kim et al. have validated the experimental results of the synergistic effects of the EDL capacity, the excellent electrical conductivity of graphene oxide, and the homogeneously dispersed MnO 2 nanoneedles 15 . A more green approach of reducing graphene nanosheet/urchin-like MnO 2 composite preparation for a supercapacitor electrode uses the glucose molecule as a reductant and the oxidized product is environmentally friendly. The obtained electrode exhibited a SC of 263 Fg −1 in a 1.0 M Na 2 SO 4 electrolyte solution calculated from constant current charge/discharge curves at 5 mAcm −2 . The excellent interfacial contact between MnO 2 and graphene is believed to have contributed to the fast transportation of electrons throughout the whole electrode matrix 1 .
Herein, we report a facile and green method to prepare reduced graphene oxide/manganese dioxide (rGO/MnO 2 ) with a glucose carbon composite (i.e., an in situ electrodeposition technique). The prepared electrode was heated to exceed the decomposition temperature of glucose with the hope that the presence of carbon from glucose in the electrode would increase the effectiveness of cation pathways from electrolyte to electrode. The schematic presentation of in situ electrodeposition of MnO 2 -rGO with and without glucose on stainless steel is shown in Fig. 1.
The dispersed GO in water is negatively charged due to some ionization of carboxyl and hydroxyl functional groups on the GO surface 16 . Those functional groups act as anchor sites, which allow the positive-charged ions to be absorbed on the surface and edge of the negatively-charged GO sheets 17 . When the Mn(CH 3 COO) 2 ·4H 2 O was added to the GO suspension solution, the Mn 2+ ion was bonded with the oxygen atom of the negatively-charged residual oxygen-containing functional groups on the graphene oxide via the electrostatic force. Then, upon electrodeposition, the GO sheets with absorbed Mn 2+ ions were deposited together to form manganese hydroxide and were reduced to graphene oxide. It is believed that adding glucose causes some molecules of Mn 2+ will bind together with the hydroxyl group of glucose via electrostatic interaction and will deposit together in the SS. Manganese hydroxide will be converted to manganese oxide after being heated at 300 °C for 6 h.

Results
In this study, we performed an in situ electrodeposition of manganese hydroxide, graphene oxide, and glucose to yield rGO/MnO 2 nanocomposite electrode materials after heat treatment at 300 °C. The XRD patterns of as-prepared rGO-MnO 2 deposits with and without glucose on top of SS and pure SS as references are shown in Fig. 2. Compared to the XRD pattern of stainless steel, there is an additional peak at 2θ = 28.3° in the diffractograms of deposited M30, M60, and M90 electrodes, Fig. 2a. This peak is attributable to the (310) plane of MnO 2 and the intensity of the peak increased as the content of manganese ion in the deposition electrolyte increased. For clearer evidence of deposited rGO-MnO 2 , the XRD diffractogram for scraped-off deposits powder of M30, M60, and M90 was carried out as shown in Fig. 2b. The characteristic peaks of MnO 2 observed at 2θ = 28.7°, 36.9°, 42.9°, and 50.3° are attributed to the (310), (211), (301) and (411) planes, which can be indexed to a tetragonal phase of α -MnO 2 with lattice constant a = 9.7847 Å, c = 2.8630 Å (JCPDS 44-0141) 15,18 . In all XRD patterns of rGO-MnO 2 scraped-off powders, an appreciable peak is observed at 2θ = 42.9°, which is associated with the (301) plane of MnO 2 . The peaks of GO in the deposit powders of the M30, M60, and M90 electrodes are hardly noticeable, suggesting that reduction of GO has taken place 19,20 . Figure 2c shows the XRD patterns of deposited rGO-MnO 2 on top of SS with the presence of different concentrations of glucose molecules in the deposition electrolyte. The XRD results show that the addition of glucose molecules did not change the structure of the sample, since only one peak is observable that centred at 28.3°, which is attributed to the (310) plane of MnO 2 .
To study the effect of glucose on the deposited rGO-MnO 2 in more detail, Raman microscopy was performed. Figure 3a,b displays the Raman spectrum of as-heated rGO/MnO 2 with and without the addition of glucose in deposition electrolytes. The characteristics peaks of GO centered at 1358 cm −1 , 1579 cm −1 , and 2675 cm −1 were attributed to D-band, G-band, and 2D-band, respectively. The D-band is related to the vacancies, edge defects, grain boundaries, and disordered carbon species in graphite layers, and G-band is due to the vibration of sp 2 hybridized C-C bonds in two dimensional hexagonal lattice [20][21][22] . The intensity ratio of the D and G-bands (I D /I G ) can be used to evaluate the sp 2 domain size of a carbon structure and partially ordered crystal structure of graphenes 23 . The I D /I G ratio of GO powder is 0.78 and the M30 and M60 electrodes resulted in increments of the I D /I G ratio to 2.37 and 1.78 (Fig. 3a), which can be attributed to an increase in defects on the surface of the reduced GO that were induced during the synthesis process. Furthermore, the G band and D band in the prepared electrode got shifted to lower wavenumbers of around 1242 cm −1 and 1568 cm −1 for the M30 and M60 electrode, revealing that rGO are deposited 24 . The addition of glucose content in the deposition electrolyte has resulted in an increase of D band intensity (Fig. 3b), indicating an increase of disorder carbon in the graphite layers. The increment of D peak intensity might be attributed to the bands combination of D1, D2, D3 and D4 in the region from 1000 to 1800 cm −1 , thus the deconvoluted of Raman spectra in this region for selected samples M30 and G03 is displayed in Fig. 3c,d. The deconvolution results for both samples clearly showed that the peak at around 1560-1598 cm −1 is related to the G peak. Peaks of D1 and D2 can be observed at around 1301-1317 cm −1 and 1599-1624 cm −1 , respectively. Depiction of another two peaks at around 1489-1545 cm −1 and 1127-1200 cm −1 are corresponded to D3 and D4 peaks. The I D /I G ratio (area) and vibration mode 25 of M30 and G03 are listed in Table 1. The increment of the D band intensity mainly arises from overlapping of D1 and D4 peaks in the band region of 1100-1400 cm −1 , suggesting that the disordered carbon in graphitic lattice has increased. In all deposited samples, the presence of a sharp peak at 650 cm −1 corresponds to MnO 2 that is attributed to stretching vibrations of MnO 6 octahedral 20 .
The elements of deposits on top of SS were also studied using EDX spectrums, the pristine MnO 2 , M30, and G03 electrodes are shown in Fig. 4. The presence of Cr, Fe, Ni, Si, and Mo belongs to SS (Fig. 4a) and the presence of new elements such as Mn and O is assigned to formation MnO 2 in electrode. The increment of carbon weight percentage in the M30 and G03 electrodes can be seen in Fig. 4c,d. The increment of carbon content in the M30 electrode is contributed from deposited of rGO carbon, while the G03 electrode is from rGO carbon and glucose decompose carbon. Glucose is considered to be the one of the most organic compounds that will decompose to a carbon solid element (known as carbon sources) 26 . Figure 4e displays the TGA analysis of D(+ ) glucose, in which the decomposition temperature of glucose was found to be around 250 °C. The heating temperature of our deposited electrodes was 300 °C, which exceeds the glucose decomposition temperature. Therefore, the increase of carbon content of the G03 electrode in the Raman and EDX analyses is believed from carbon of glucose decompose.
FESEM was carried out to characterize the morphology structures of as-heated rGO/MnO 2 deposited from different deposition electrolyte contents as shown in The morphological alteration of M30 electrode was obvious when a 0.01 M glucose molecule was added to the electrolyte deposition Fig. 6. The MnO 2 bunches no longer existed and clearly showed that the porous structure was formed and regularly arranged. The pore diameter was ~60-100 nm when the glucose concentration was 0.01 M, Fig. 6a 1 . The pore diameter decreased dramatically (~30 nm) when the glucose concentration was increased to 0.03 M, Fig. 6b 1 , indicating that the pore structure of rGO/   MnO 2 samples can be tuned by adjusting the glucose concentration in the electrodeposition electrolyte. The average thickness of the G03 electrode was also reduced to 265 nm Fig. 6b 2 . A high number of pores that built from interconnected nanoflakes structures in the G03 electrode is believed to improve the porosity and provide a unique conductive network. This observation might be due to a slower rate of MnO 2 electrocrystallization, which allows the atoms to arrange themselves at the lowest energy site. According to Babakhani et al., the influential factors, such as the concentration of the deposition electrolyte, would affect the electrocrystallization rate of MnO 2 28 . Further increases of glucose concentration 0.06 M (G06), deposition rate became slower and lead to flakes not readily growth uniform during the electrodeposition period, Fig. 6c 1 . The TEM of G03 electrode is shown in Fig. 6d 1 . The deposits have a uniform and well-spread rGO sheet, which is covered by MnO 2 nanoflakes. The interconnected structure, which creates the porous structure, can also be clearly observed at a high magnification of TEM (Fig. 6d 2 . This unique structure has several advantages: (i) the porous structure greatly facilitates the ion diffusion from the electrolyte into the electrode matrix, which promotes the specific capacitance due to high utilization of MnO 2 ; (ii) thin deposited materials that are able to shorten the diffusion path Scientific RepoRts | 5:16195 | DOi: 10.1038/srep16195 of electrons and ions; (iii) the interconnected flakes structure without agglomeration could exhibit the excellent electrochemical performance as an electrode for a supercapacitor 13 .
The electrochemical properties of the materials were characterized by cyclic voltammetry (CV) and galvanostatic charge-discharge measurements. CV has been considered to be a suitable technique to investigate the occurrence of faradic or non-faradic reactions in the electrode 29 . The CV curve of pristine MnO 2 , M30, M60, and M90 electrodes in the potential range from − 1 V to + 1 V at scan rate of 5 mVs −1 in the 0.5 M Na 2 SO 4 electrolyte are shown in Fig. 7a. A pair of distinct anodic and cathodic peaks can be clearly observed around 0.2 V and − 0.1 V, while other less-intense anodic and cathodic peaks are around 0.9 V/− 0.7 V. These peaks are believed to be derived mainly from the redox pairs of Mn 2+ /Mn 3+ . The current response for M30, M60, and M90 is much higher than that of pure MnO 2 , inferring that rGO/ MnO 2 samples have better charge transfer kinetics, due to higher utilization of active Mn species. The calculated specific capacitances from the CV curve for pristine MnO 2 , M30, M60, and M90 electrodes are 167 Fg −1 , 264 Fg −1 , 220 Fg −1 , and 175 Fg −1 , respectively. The highest specific capacitance is found in the M30 electrode, which can be attributed to the combined contribution of redox pseudocapacitance of MnO 2 and the electrical double layer capacitance of the rGO. The improved performance of the M30 electrode may also be due to a high content of graphene carbon, which is beneficial to shortening the cation path into the electrode matrix and reducing the transfer resistance. CV curve of the deposited G03 electrode at a scan rate of 5 mVs −1 in a Na 2 SO 4 electrolyte solution is shown in Fig. 7b. Compared to the CV curve of M30, CV curve of G03 electrode has a similar shape and potential position of anodic and cathodic peaks. However, the current response of G03 is higher, indicating that the effective utilization of the MnO 2 increased 27 . The calculated specific capacitances are 377 Fg −1 , 430 Fg −1 , and 361 Fg −1 for G01, G03, and G06 at a scan rate of 5 mVs −1 . Electrode G03 exhibits 63% specific capacitance improvement compared to M30, which is attributed to the uniform morphology structure, less thickness, and low transfer resistance. A further increase of glucose concentration decreased the capacitance to 37% (G06), which might be due to less MnO 2 available for the reaction.
The characteristic of ion transport resistance for all samples has been investigated by electrochemical impedance spectra (EIS). The Nyquist plot of all deposits electrode in a frequency range of 0.1 Hz until 100 kHz in 0.5 M Na 2 SO 4 electrolyte is shown in Fig. 7c,d. The equivalent circuit in accordance with Nyquist plots is fitted using Nova software and parameters are shown in Table 2. The intercept of the arc on the x-axis at high frequency region is called an equivalent series resistance (R s ) which represents combination resistance of ionic resistance of electrolyte, contact resistance and internal resistance of the material. This value is almost same for all electrodes within the logical magnitude error. The semicircle region in high frequency region corresponds to charge transfer resistance, R ct [30][31][32] . Overall, G03 electrode has lowest transfer resistance among all samples, indicating G03 has a better electrochemical performance. The incorporation of 0.03 M glucose in M30 electrode leads to improvise the access for intercalation/deintercalation of cation to electrode matrix. The two constant phase elements of CPE 1 and  Fig. 7c were used to replace the double layer capacity and Warburg impedance resistance for semi-infinite linear diffusion, respectively 33 . In glucose system, CPE 2 represents the faradic impedance which is due to redox transition within the electrode, Fig. 7d 34 . As discuss in previous reports 33,35 , two constant phase elements can be describe as Z CPE1 = [Q(jω ) n1 ] −1 and Z CPE2 = [Q(jω ) n2 ] −1 with − 1 ≤ n ≤ 1. The component n is correction factor represents the roughness of electrode and it has value ranging from 0 to 1. Pure capacitance yields n = 1, pure resistance yields n = 0, while n = 0.5 represents Warburg impedance. The value of n 1 ~ 0.8 in G03 indicates that G03 has a nature porous of electrode, in agreement with TEM result.
In order to get more information on the capacitive performance of the best prepared G03 electrode, G03 electrode was selected to study the performance in three different electrolytes (i.e., 0.5 M Na 2 SO 4 , 0.5 M KOH, and a 0.5 M KOH/0.04 M K 3 Fe(CN) 6 electrolyte solution. The capacitive performance is believed to be influenced by the size of the cation, cation mobility, and rate of adsorption/desorption at the electrode-electrolyte surface 30 . All CV and charge-discharge performances were studied in a potential range of − 0.5 V to 0.5 V. The CV curve of the G03 electrode in three different electrolytes at a scan rate of 5 mVs −1 is shown in Fig. 8. There is no pitting corrosion of SS substrate found in the potential range of − 0.5 to 0.5 V in this three different electrolyte, displayed in supplementary Fig. 1 (supplementary information).
The CV curve of the G03 electrode in 0.5 M Na 2 SO 4 and 0.5 M KOH electrolyte solutions shown in Fig. 8a,b, the electrode reaction occurred according to Eq. (1, 2) 30,36 . The current response in 0.5 M KOH is found to be higher than in the 0.5 M Na 2 SO 4 electrolyte. This behaviour might be due to the smaller K + size, which can enhance the chemisorption reaction rate, thus optimizing the pseudocapacitance 37 where C is Na 2+ or K + The best performance of the G03 electrode in KOH was obtained. The redox mediator electrolyte, 0.04 M of K 3 Fe(CN) 6 , was then added into a 0.5 M KOH electrolyte solution with a volume ratio of 1:1. The CV curve in Fig. 8d reveal that the additional of redox mediator has increase the current response drastically. The highest anodic/cathodic peaks at 0.27 V/0.15 V is assign to the charging and discharging process of K 4 Fe(CN) 6 to K 3 Fe(CN) 6 (Fig. 8c), which undergoes the reaction shown in Eq. 3 38 , while other less-intensely observed redox peaks originated from electrode reactions with KOH electrolytes. Both instances of redox reactions of rGO/MnO 2 in KOH and the reaction of K 3 Fe(CN) 6 in the CV curve indicates that the reactions occur simultaneously and independently 38 . The calculated specific capacitance is 5135 Fg −1 from the curve, the specific capacitance was found to increase 538% after an addition of 0.04 M K 3 Fe(CN) 6    where M is Mn 2+ cations, and 1 ≤ n ≤ z. Figure 9a displays the charge-discharge (CDC) profile of the G03 electrode in three different electrolytes at current density of 20 Ag −1 . The CDC profile showed a slightly non-linear curve, which indicates the occurrence of a redox reaction within this voltage range 41 . The charge-discharge curve of G03 in the Na 2 SO 4 and KOH electrolyte solutions has a small plateau potential around − 0.1 V in the discharging curve, which corresponds to the reduction of MnO 2 . In the KOH/K 3 FeCN 6 electrolyte solution, two potential plateaus were found. The first potential at − 0.1 V is due to the reduction of MnO 2 , and the second potential at 0.2 V corresponds to the redox reaction of K 4 Fe(CN) 6 . Because both reactions occured, the G03 electrode in the KOH/K 3 Fe(CN) 6 electrolyte solution at a current density of 20 Ag −1 has a higher specific capacitance of 13,333 Fg −1 , with a power density of 68.35 kW.kg −1 and an energy density of 1851 Wh.kg −1 .
The stability test of the 0.03 M electrode in Na 2 SO 4 , KOH, and KOH/ K 3 Fe(CN) 6 electrolyte solutions was continuously performed in a voltage range from − 0.5 V to 0.5 V at a scan rate of 10 mVs −1 until 2000 cycles, as shown in Fig. 9b. The specific capacitance of G03 in the KOH/ K 3 Fe(CN) 6 electrolyte solution has retained up to 46% of its initial specific capacitance after 2000 th cycle, it has good stability compared to other electrolytes. In comparison with pass studies, rGO/MnO 2 electrode prepared using hydrothermal and sol gel methods [42][43][44] showed good stability but low specific capacitance which is due to high carbon content in the electrode and no obvious redox peaks were observed 42 . In our work, the high capacitance, energy and power density can be assigned to high redox activity of the electrode. The high degradation of the capacitance retention may be due to a high rcurrent pass through the electrode during cyclability test 20,30 .

Discussion
In summary, rGO/MnO 2 nanocomposites has successfully synthesized by using electrodeposition method. The rGO was found to be clearly covered by MnO 2 in M30 electrode sample. An ultrathin-deposits electrode with uniform nanoflake structure was obtained after 0.03 M of glucose solution was added into GO/Mn(CH 3 COO) 2 ·4H 2 O electrolyte solution. It is believed that the additional 0.03 M glucose solution is a suitable concentration for slowing down the nucleation process of MnO 2 , which results in an excellent flake morphology structure. The depositing of glucose onto SS electrodes was also found to increase the carbon content, which helps to reduce the transfer resistance of cation diffusion paths to the electrode matrix. As a result, the effective use of MnO 2 resulted in a high specific capacitance of 13,333 Fg −1 with a power density of 68.35 kW.kg −1 , and an energy density of 1851 Wh.kg −1 at a current density of 20 Ag-1 in a 0.5 M KOH/ 0.04 M K 3 Fe(CN) 6 electrolyte solution. Nevertheless, after 2000 cycles at a scan rate of 10 mVs −1 , a retention percentage of 46% was observed. This result may be ascribed to the loss of the active materials in the electrode. The preparation of G03 electrode is a simple, low-cost, and environmentally-friendly method that holds great potential for producing cost-effective and high-energy-density supercapacitors.

Methods
The MnO 2 -rGO composite was synthesized by a potentiodynamic method on 4 cm 2 of stainless steel (SS) in the voltage range of 0 to 2 V. The scan rate of 50 mVs −1 was applied. The electrochemical cell consisted of the steel substrate as the working electrode, a carbon rod as the counter electrode, and Ag/AgCl as the reference electrode. The stainless steel subtracts were sonicated and rinsed with acetone three times and dried at room temperature. Prior to deposition, 50 mg of GO was dispersed in 50 mL of distilled water and ultrasonication treatment for 0.5 h. The in situ electrodeposition process was performed in an aqueous solution containing a fixed volume of GO (10 mL) and different volumes (30 mL, 60 mL, and 90 mL) of 0.01 M manganese acetate aqueous solution. The obtained electrodes were dried at a temperature of 300 °C for 6 h and allowed to cool to room temperature before the characterisation step. The prepared samples were denoted as M30, M60, and M90, respectively. The MnO 2 -rGO-glucose carbon composite was prepared by the same aforementioned procedure. In a typical experiment, the deposition electrolytes were prepared by dissolving 0.01 M, 0.03 M, 0.06 M of glucose in 30 mL of distilled water under magnetic stirring. These solutions were then separately mixed with 30 mL manganese acetate (0.01 M) aqueous solution. Consequently, 10 mL GO solution was added drop-wise in the solution and the mixture was sonicated for 22 mins at room temperature. The prepared samples after heating treatment are labeled as G01, G03, and G06.
The X-ray diffraction (XRD) patterns of deposited samples on SS and the powders of deposits were obtained using a PANalytical Empyrean and D8 Advance X-Ray diffractometer-Bruker AXS with CuK α monochromatized radiation at 40 kV and 40 mA. The Raman spectra were obtained using Renishaw inVia Raman microscope with a green beam. The FESEM and TEM images of the electrodes were captured using a Jeol JSM 7600 and Jeol JEM 2100F, respectively. The Energy Dispersive X-ray (EDX) spectrum was collected using an Oxford Instruments apparatus for the elemental analysis of the electrode sample. Thermal gravimetric analysis (TGA) of glucose was performed using a TA Instrument, Q500 with heating rate of 0.01 °C min −1 . The electrochemical test of charge/discharge (CD), cyclic voltammetry (CV), and electrical impedance spectroscopy (EIS) were conducted using potentiostat (Autolab, PGSTAT30). The platinum wire was used as a counter and Ag/AgCl was used for reference electrodes.