Effect of Potassium Ions on the Formation of Mixed-Valence Manganese Oxide/Graphene Nanocomposites

One-pot synthesis of mixed-valence manganese oxide (MnOx)/potassium ion-doped reduced graphene oxide (rGO) composites for efficient electrochemical supercapacitors is introduced. Using manganese nitrate and potassium permanganate as co-precursors for the MnOx and by directly annealing the rGO without tedious purification steps, as described herein, MnOx/rGO composites with a high specific capacitance of 1955.6 F g−1 at a current density of 1 A g−1 are achieved. It is found that the presence of potassium ions helps in the development of mixed-valence MnOx on the surface of the rGO.


Introduction
Recently, composite electrode materials comprising electrochemical double-layer capacitors (EDLCs) and pseudocapacitors have come to represent a promising avenue for providing greater energy density for supercapacitors. The use of a pseudocapacitive material enables fast electron transfer at the surface of the electrode, leading to a higher energy density than can be achieved by using EDLCs alone. As a promising candidate for a composite electrode material, nanocomposites composed of manganese oxide (MnO x ) and graphene have been widely investigated [1][2][3][4][5][6][7][8]. The properties of graphene, such as high electrical conductivity, large specific surface area, transparency, and flexibility, are beneficial to use as a supercapacitor electrode material [9][10][11][12][13][14][15][16][17]. Manganese oxide is attractive as a pseudocapacitor because of its low cost, natural abundance, environmental compatibility, and high theoretical specific capacitance [18]. The shortcoming of manganese oxide as a supercapacitor electrode is its lack of electrical conductivity, which can be compensated for by the use of graphene composites. In addition, the high surface area and high electrical conductivity of graphene can improve the electrochemical performance by double layer formation, and this is also helped by a uniform distribution of the manganese oxides on the basal plane of the graphene.
The electrochemical performance of the manganese oxides are directly influenced by their crystal structures and oxidation states. Several studies have reported on how the electrochemical performance depends on the oxidation states of manganese oxides [19,20]. In particular, it is reported that the presence of mixed-valence manganese oxides can result in superior performance of capacitive behavior, due to the coexistence of aliovalent manganese cations that drive the defect-accelerated kinetics of the surface reactions [21].
In this study, we report on a one-pot synthesis of a mixed-valence manganese oxide (MnO x )/potassium (K + ) ion-doped, reduced graphene oxide (rGO) composite materials for an efficient supercapacitor electrode. By using manganese nitrates and potassium permanganate as co-precursors for MnO x and then using direct annealing without any intervening washing steps, mixed-valence MnO x /rGO composites with a K + -doped rGO structure were successfully achieved. Due to the synergetic effects created by the presence of mixed-valence MnO x and K + ions, the resulting composite structure showed excellent capacitive properties, reaching a maximum specific capacitance of 1955.6 F g −1 at a current density of 1 A g −1 .

Preparation of GO
A modified Hummers' method was used to prepare the Graphite oxide (GO), as described in previous reports [22]. Briefly, 2 g of graphite flakes were mixed with 46 mL of 98% sulfuric acid in a 250 mL round bottom flask and placed in an ice bath with constant stirring. Then 6 g of potassium permanganate was slowly added to the mixture. After 2 h, the supernatant mixture was transferred to an oil bath and kept at a constant temperature of 35 • C for 6 h. Then 92 mL of de-ionized (DI) water was gradually added to the reaction mixture. The mixture was then stirred for 1 h. The whole reaction mixture was then poured into a 1 L beaker containing 240 mL of water. Then 35% hydrogen peroxide solution was added until the color of the mixture changed to bright yellow. Hydrochloric acid diluted in 5% water was added in order to remove the metal cations. Finally, the resulting solution was washed with DI water, and dialysis was performed until a neutral graphite oxide solution was obtained.

Preparation of MnO x /rGO
Nanocomposites 100 mg of GO was uniformly dispersed in 200 mL of DI water using water bath sonication for 1 h. Then 20 mL each of 0.01 M Mn(NO 3 ) 2 ·4H 2 O and KMnO 4 solutions were simultaneously added, dropwise, to the GO solution with stirring at a constant speed for 1 h. The supernatant mixture was then dried at room temperature, and finally the remaining moisture was removed by vacuum drying at room temperature. No washing with water was involved in this process. The resulting solid powder-type mixtures were annealed at 500 • C in an Ar atmosphere. For a comparative study, samples annealed at 400 • C and 600 • C were also prepared and compared for their electrochemical performances.

Characterization
A Tecnai G2 F20 (FEI, Hillsboro, OR, USA) was used to take transmission electron microscopy (TEM) images and perform energy dispersive X-ray spectroscopy (EDS) analysis. X-ray diffraction (XRD) patterns were recorded on a Rigaku SmartLab diffractometer (Rigaku, Tokyo, Japan) using Cu Kα radiation. A Jobin Yvon LabRAM HR UV-Visible-NIR spectrometer (Horiba, Kyoto, Japan) was used to obtain Raman spectra. X-ray photoelectron spectroscopy (XPS) information was obtained on a Thermo Fisher Scientific (Waltham, MA, USA) K-alpha using an Al source. Fourier transform infrared (FT-IR) spectra were recorded on a FR/IR-6600 (JASCO, Tokyo, Japan). A Verios 460L (FEI, Hillsboro, OR, USA) was used to take field emission scanning electron microscope (FE-SEM) images and also used to perform energy dispersive spectroscopy (EDS) mapping analysis.

Electrochemical Measurements
Electrochemical responses of the MnO x /rGO composites were measured using a three-electrode system. A 6 M KOH aqueous solution was used as the electrolyte, a platinum plate was used as the counter electrode, and Ag/AgCl was used as the reference electrode. The working electrode was prepared by mixing 80 wt% active material, 15 wt% activated carbon, and 5 wt% polyvinylidene fluoride (PVDF) in N-methyl-2-pyrrolidone (NMP). The slurry was then spread onto a glassy carbon electrode, which was used as the current collector. The electrode was then heated at 60 • C for 24 h to evaporate the solvent. The rate of total deposited mass throughout our study was 1.5 mg cm −2 . To investigate the electrochemical performance of the electrodes, cyclic voltammetry (CV) and galvanostatic charge-discharge techniques were employed. Electrochemical performance in the three-electrode configuration was determined in a CH660D electrochemical station (CH Instruments, Inc., Austin, TX, USA).

Materials Characterization
The MnO x /rGO nanocomposites were synthesized using manganese nitrate and potassium permanganate as co-precursors for the MnO x . First, electrostatic binding of manganese (II) cations on the surface of GO was achieved. With the addition of potassium permanganate, permanganate , respectively. It was found that using the annealing temperature of 500 • C was beneficial for the development of the mixed-valence manganese oxides in this work (Figure 1b). At lower annealing temperatures, such as 300 • C or 400 • C, negligible peaks of manganese oxides were detected, indicating that such annealing temperatures were insufficient for the development of MnO x crystals. In addition, when the annealing temperature reached 600 • C, most MnO x were thermally converted to a Mn 3 O 4 single phase [23]. The nominal peaks of K + -intercalated MnO x structure (K x Mn 8 O 16 ) was also detected alongside in case of the composite annealed at 500 • C.

Electrochemical Measurements
Electrochemical responses of the MnOx/rGO composites were measured using a three-electrode system. A 6 M KOH aqueous solution was used as the electrolyte, a platinum plate was used as the counter electrode, and Ag/AgCl was used as the reference electrode. The working electrode was prepared by mixing 80 wt% active material, 15 wt% activated carbon, and 5 wt% polyvinylidene fluoride (PVDF) in N-methyl-2-pyrrolidone (NMP). The slurry was then spread onto a glassy carbon electrode, which was used as the current collector. The electrode was then heated at 60 °C for 24 h to evaporate the solvent. The rate of total deposited mass throughout our study was 1.5 mg cm −2 . To investigate the electrochemical performance of the electrodes, cyclic voltammetry (CV) and galvanostatic charge-discharge techniques were employed. Electrochemical performance in the three-electrode configuration was determined in a CH660D electrochemical station (CH Instruments, Inc., Austin, TX, USA).

Materials Characterization
The MnOx/rGO nanocomposites were synthesized using manganese nitrate and potassium permanganate as co-precursors for the MnOx. First, electrostatic binding of manganese (II) cations on the surface of GO was achieved. With the addition of potassium permanganate, permanganate (MnO4 − ) ions were bonded with Mn 2+ ions. Subsequent annealing at 500 °C in an Ar atmosphere led to the resulting MnOx/K + -doped rGO composites. No washing was involved in this process; therefore, abundant K + ions were introduced in the composite structure. The X-ray diffraction (XRD) patterns of the compositions and crystal structures of the MnOx/rGO nanocomposites and the GO are shown in Figure 1a respectively. It was found that using the annealing temperature of 500 °C was beneficial for the development of the mixed-valence manganese oxides in this work (Figure 1b). At lower annealing temperatures, such as 300 °C or 400 °C, negligible peaks of manganese oxides were detected, indicating that such annealing temperatures were insufficient for the development of MnOx crystals. In addition, when the annealing temperature reached 600 °C, most MnOx were thermally converted to a Mn3O4 single phase [23]. The nominal peaks of K + -intercalated MnOx structure (KxMn8O16) was also detected alongside in case of the composite annealed at 500 °C.  Typical microstructures of the MnO x /rGO nanocomposites, investigated by TEM, are shown in Figure 2. As seen in Figure 2a, MnO x nanoparticles in the range of 6 to 12 nm grew on the surface of the graphene sheets. Our proposed approach allowed us to achieve uniformly distributed MnO x particles without nanoparticle aggregation. The loaded amount of the MnO x on the surface of the graphene can be controlled by varying the amounts of precursors used for the growth of the MnO x . Figure 2b shows a lattice spacing of 2.6 nm, which corresponds to the (111) plane of MnO. Figure 2c shows the 0.28 nm lattice spacing of the (200) plane of Mn 3 O 4 . Figure 2d shows the 0.26 nm lattice spacing of the (222) plane of the Mn 2 O 3 . The MnO x phase is found to be one of the following: MnO, Mn 2 O 3 , and Mn 3 O 4 . In addition, the nominal presence of the K + -intercalated MnO x structures were observed through TEM analyses ( Figure S1), which are in agreement with the XRD results. Typical microstructures of the MnOx/rGO nanocomposites, investigated by TEM, are shown in Figure 2. As seen in Figure 2a, MnOx nanoparticles in the range of 6 to 12 nm grew on the surface of the graphene sheets. Our proposed approach allowed us to achieve uniformly distributed MnOx particles without nanoparticle aggregation. The loaded amount of the MnOx on the surface of the graphene can be controlled by varying the amounts of precursors used for the growth of the MnOx. Figure 2b shows a lattice spacing of 2.6 nm, which corresponds to the (111) plane of MnO. Figure 2c shows the 0.28 nm lattice spacing of the (200) plane of Mn3O4. Figure 2d shows the 0.26 nm lattice spacing of the (222) plane of the Mn2O3. The MnOx phase is found to be one of the following: MnO, Mn2O3, and Mn3O4. In addition, the nominal presence of the K + -intercalated MnOx structures were observed through TEM analyses ( Figure S1), which are in agreement with the XRD results.  It also shows that the K + ions were uniformly distributed on the rGO surface and near the grain boundary of the polymorphs of the MnOx. Also, a small amount of the K + ions were detected within the MnOx structure (Figure 3e,f). This is due to the small number of K + -intercalated MnOx structures present in the composites, which is in accordance with the XRD study. During calcination at 500 °C, it is known that NO3 2− ions can easily evaporate, but K + ions remain stably along the grain boundary of the polymorphs of MnOx, as well as on the surface of the rGO [24].
The Raman spectra of the GO and the MnOx/rGO nanocomposite are shown in Figure 4a. A D band at 1346 cm −1 and a G band at 1570 cm −1 were detected in the spectrum of GO. The characteristic D and G bands are observed for both the MnOx/rGO nanocomposite and the GO products. The intensity ratios of the D to G bands for GO and the MnOx/rGO nanocomposite show an obvious change: 0.75 and 0.98, respectively. The increasing value of the D/G intensity ratio of the MnOx/rGO nanocomposite in comparison to that of the GO is due to the reduced size of the sp 2 domains and the creation of new graphitic domains with smaller sizes after thermal reduction at  It also shows that the K + ions were uniformly distributed on the rGO surface and near the grain boundary of the polymorphs of the MnO x . Also, a small amount of the K + ions were detected within the MnO x structure (Figure 3e,f). This is due to the small number of K + -intercalated MnO x structures present in the composites, which is in accordance with the XRD study. During calcination at 500 • C, it is known that NO 3 2− ions can easily evaporate, but K + ions remain stably along the grain boundary of the polymorphs of MnO x , as well as on the surface of the rGO [24]. The Raman spectra of the GO and the MnO x /rGO nanocomposite are shown in Figure 4a. A D band at 1346 cm −1 and a G band at 1570 cm −1 were detected in the spectrum of GO. The characteristic D and G bands are observed for both the MnO x /rGO nanocomposite and the GO products. The intensity ratios of the D to G bands for GO and the MnO x /rGO nanocomposite show an obvious change: 0.75 and 0.98, respectively. The increasing value of the D/G intensity ratio of the MnO x /rGO nanocomposite in comparison to that of the GO is due to the reduced size of the sp 2 domains and the creation of new graphitic domains with smaller sizes after thermal reduction at 500 • C [25]. In addition, the peaks at 639.7 cm −1 for the MnO x /rGO are assigned to the Mn-O stretching vibration in the basal plane of MnO 6 and/or the symmetric stretching vibration of the MnO 6 group [26].    500 °C [25]. In addition, the peaks at 639.7 cm −1 for the MnOx/rGO are assigned to the Mn-O stretching vibration in the basal plane of MnO6 and/or the symmetric stretching vibration of the MnO6 group [26].   The surface electronic state and the chemical bonding state of MnO x /rGO were analyzed in detail by XPS. As shown in Figure 4b, the survey XPS spectrum for the MnO x /rGO nanocomposite revealed the presence of carbon, manganese, potassium, and oxygen. In particular, the omission of the rinsing step after the reaction of the KMnO 4 in the preparation of the MnO x /rGO nanocomposite may account for the presence of the K + inside the MnO x /rGO after annealing. The C 1s peak originates from the graphene nanosheets. The Mn 2p peak was further inspected by high-resolution XPS analysis, as shown in Figure 4c. The two peaks at 641.5 eV and 653.2 eV can be assigned to Mn 2p 3/2 and 2p 1/2 , respectively, which are characteristic of mixed-valence MnO x [27,28]. The oxidation state of the manganese was confirmed by the multiplet splitting of the Mn 3s state. As shown in Figure 4d, the splitting width was 5.5 eV, which is in accordance with a previous report on the XPS spectrum of Mn 3 O 4 [19]. Fourier transform infrared spectroscopy (FT-IR) was used to characterize the GO and the MnO x /rGO nanocomposites, and the results are shown in Figure S2. The GO showed a broad band at 3300-3700 cm −1 and a distinct band at 1620~1730 cm −1 , corresponding to O-H and C=O, C-O respectively. After the composite was made, all bands decreased significantly, and a new peak, for the MnO x band, at 600 cm −1 was confirmed [29][30][31].

Electrochemical Properties
To evaluate the temperature-dependent electrochemical behavior and quantify the capacitance performance of the MnO x /rGO nanocomposites, samples annealed at 400 • C, 500 • C, and 600 • C were prepared, and cyclic voltammetry (CV) measurements were taken at different scan rates under the operating potential of −0.6-0.7 V in a three-electrode system. The results are shown in Figure 5a. It can be observed that GM 500 shows a larger capacitive area than either GM 400 or GM 600 . This is because the pseudocapacitance is dependent on the structure of the metal oxide. Recently, it was established that a multivalent oxide system is capable of adsorbing anions as well as transporting the oxygens into vacant sites of manganese ions, thus promoting high redox reactions and faster electron transitions [32,33]. From the above XRD study, we observed that the crystal structure and the valence state of the resulting MnO x and the following CV performances are dependent on the annealing temperature. GM 400 showed very low capacitive current compared to GM 500 , probably due to the insufficient growth of the mixed-valence MnO x . GM 600 also demonstrated an inferior performance to that of GM 500 . This is attributed to the reduced redox active formation of the Mn 3 O 4 , which could, in turn, result in a reduction of the capacitive current for the increased annealing temperature of 600 • C. However, upon annealing at 500 • C, various mixed-valence MnO x phases, such as MnO, Mn 2 O 3, and Mn 3 O 4 , were generated, which is found to be beneficial for a larger capacitive current area.
A CV analysis of GM 500 at different scan rates was carried out to investigate the current response of GM 500 . All of the CV curves had a quasi-rectangular shape for both low and high scan rates (Figure 5b). Typically, at the high scan rates-above 50 mV s −1 -the CV curves are distorted from their rectangular shape. This indicates unbalanced ion diffusion with respect to the charging and discharging currents, caused by the polarization between the metal oxide and electrolyte with limited incubation of electrolyte inside the electrode materials. However, the graphs are almost the same with respect to the zero-current axis. To quantify the capacitance value of GM 500 , we conducted galvanostatic charge/discharge measurements. Figure 5c shows a faint hump in the 0.09 V signal during charging, and a slight bend appears in the discharging graph. This is due to the significance of the redox reaction during electrochemical performance, which varies due to the presence of the mixed-valence MnO x and the abundant K + ions. Figure 5d presents specific capacitance values calculated from the following equation: where C s (F g −1 ) is the specific capacitance, i m = I/m (A g −1 ) is the current density, where I is the current and m is the mass of the active material, Vdt is the integral current area, where V is the potential with initial and final values of V i and V f , respectively [34]. The highest and lowest specific capacitance values were 1955.6 F g −1 and 840.7 F g −1 at current densities of 1 A g −1 and 4.5 A g −1 , respectively. These high capacitive performances might be due to the presence of the K + ions on the grain boundary of the MnOx and the surface of the rGO, which were easily transported from the electrode to the electrolyte. In previous studies, the charge storage mechanisms of MnOx-based composites were controlled mainly by the intercalation/de-intercalation of alkali cations and the structural changes that occurred during the electrochemical performances [35]. In our study, most of the K + ions were doped on the surface of the rGO, rather than intercalated within the MnOx. Since the K + ions doped on the rGO induce an n-doping effect of graphene, the electrical conductivity of the rGO is expected to be enhanced [18]. In this regard, transportation of electrons from the redox center to the current collector can be promoted, resulting in an increase of capacitance performance. Moreover, the formation of mixed-valence MnOx can result in superior performance of capacitive behavior, which benefits from the coexistence of aliovalent manganese cations that drive the defect-accelerated kinetics of the surface reactions. Both GM400 and GM600 show capacitive properties inferior to that of GM500, which indicates that during the annealing process, there is insufficient development of the mixed-valence MnOx (Figures S3 and S4). In addition, GM500 exhibited the largest surface area, 210 m 2 g −1 , from Brunauer-Emmet-Teller (BET) analysis (Belsorp-max, Bel Japan Inc., Toyonaka, Japan) compared to that of GM400 and GM600, supporting the superior capacitive property of GM500 ( Figure S5). The cyclic stability test results shown in Figure 6 reveal that the specific capacitance is retained as 96%, even after 4000 charge/discharge cycles. Based on these results, we conclude that GM500 is an effective material for use as a supercapacitor electrode. The highest and lowest specific capacitance values were 1955.6 F g −1 and 840.7 F g −1 at current densities of 1 A g −1 and 4.5 A g −1 , respectively. These high capacitive performances might be due to the presence of the K + ions on the grain boundary of the MnO x and the surface of the rGO, which were easily transported from the electrode to the electrolyte. In previous studies, the charge storage mechanisms of MnO x -based composites were controlled mainly by the intercalation/de-intercalation of alkali cations and the structural changes that occurred during the electrochemical performances [35]. In our study, most of the K + ions were doped on the surface of the rGO, rather than intercalated within the MnO x . Since the K + ions doped on the rGO induce an n-doping effect of graphene, the electrical conductivity of the rGO is expected to be enhanced [18]. In this regard, transportation of electrons from the redox center to the current collector can be promoted, resulting in an increase of capacitance performance. Moreover, the formation of mixed-valence MnO x can result in superior performance of capacitive behavior, which benefits from the coexistence of aliovalent manganese cations that drive the defect-accelerated kinetics of the surface reactions. Both GM 400 and GM 600 show capacitive properties inferior to that of GM 500 , which indicates that during the annealing process, there is insufficient development of the mixed-valence MnO x (Figures S3 and S4). In addition, GM 500 exhibited the largest surface area, 210 m 2 g −1 , from Brunauer-Emmet-Teller (BET) analysis (Belsorp-max, Bel Japan Inc., Toyonaka, Japan) compared to that of GM 400 and GM 600 , supporting the superior capacitive property of GM 500 ( Figure S5). The cyclic stability test results shown in Figure 6 reveal that the specific capacitance is retained as 96%, even after 4000 charge/discharge cycles. Based on these results, we conclude that GM 500 is an effective material for use as a supercapacitor electrode. To clarify the effect of the K + ions, GM500 without K + ions was also prepared. K + -free GM500 was prepared by repetitive washing after the reaction of the manganese precursors and subsequent annealing at 500 °C. From TEM and XRD studies, it is found that the resulting K + -free GM500 is composed of Mn3O4, without showing any multivalency ( Figure S6). We employed scanning electron microscopy (SEM) and EDS mapping to compare the composition of elements before and after washing the GM500 (Figure 7). The EDS elemental mapping image of GM500 showed the K-edge signals of Mn, O, and K with atomic percentages of 22%, 39%, and 39%, respectively (Figure 7a). In contrast, the EDS mapping of K + -free GM500 indicates an increased amount of O and a dramatically decreased amount of K after washing. The atomic percentages of Mn, O, and K were 9%, 84%, and 7%, respectively (Figure 7c). Figure 7b shows the XPS survey scan of the K + -free GM500, presenting an absence of bonding associated with K ions. The above results indicate that the K + ions are removed by washing. The electrochemical performances also show that K + -free GM500 is very inferior to GM500 with K + doping, exhibiting a specific capacitance of 272 F/g at a current density of 1 A/g (Figure 7d). From these observations, we can conclude that the presence of K + ions is essential for developing multi-valence MnOx on the surface of the rGO upon annealing at a temperature of 500 °C, leading to the superior capacitive performance of the composite electrode.  To clarify the effect of the K + ions, GM 500 without K + ions was also prepared. K + -free GM 500 was prepared by repetitive washing after the reaction of the manganese precursors and subsequent annealing at 500 • C. From TEM and XRD studies, it is found that the resulting K + -free GM 500 is composed of Mn 3 O 4 , without showing any multivalency ( Figure S6). We employed scanning electron microscopy (SEM) and EDS mapping to compare the composition of elements before and after washing the GM 500 (Figure 7). The EDS elemental mapping image of GM 500 showed the K-edge signals of Mn, O, and K with atomic percentages of 22%, 39%, and 39%, respectively (Figure 7a). In contrast, the EDS mapping of K + -free GM 500 indicates an increased amount of O and a dramatically decreased amount of K after washing. The atomic percentages of Mn, O, and K were 9%, 84%, and 7%, respectively (Figure 7c). Figure 7b shows the XPS survey scan of the K + -free GM 500 , presenting an absence of bonding associated with K ions. The above results indicate that the K + ions are removed by washing. The electrochemical performances also show that K + -free GM 500 is very inferior to GM 500 with K + doping, exhibiting a specific capacitance of 272 F/g at a current density of 1 A/g (Figure 7d). From these observations, we can conclude that the presence of K + ions is essential for developing multi-valence MnO x on the surface of the rGO upon annealing at a temperature of 500 • C, leading to the superior capacitive performance of the composite electrode. To clarify the effect of the K + ions, GM500 without K + ions was also prepared. K + -free GM500 was prepared by repetitive washing after the reaction of the manganese precursors and subsequent annealing at 500 °C. From TEM and XRD studies, it is found that the resulting K + -free GM500 is composed of Mn3O4, without showing any multivalency ( Figure S6). We employed scanning electron microscopy (SEM) and EDS mapping to compare the composition of elements before and after washing the GM500 (Figure 7). The EDS elemental mapping image of GM500 showed the K-edge signals of Mn, O, and K with atomic percentages of 22%, 39%, and 39%, respectively (Figure 7a). In contrast, the EDS mapping of K + -free GM500 indicates an increased amount of O and a dramatically decreased amount of K after washing. The atomic percentages of Mn, O, and K were 9%, 84%, and 7%, respectively (Figure 7c). Figure 7b shows the XPS survey scan of the K + -free GM500, presenting an absence of bonding associated with K ions. The above results indicate that the K + ions are removed by washing. The electrochemical performances also show that K + -free GM500 is very inferior to GM500 with K + doping, exhibiting a specific capacitance of 272 F/g at a current density of 1 A/g (Figure 7d). From these observations, we can conclude that the presence of K + ions is essential for developing multi-valence MnOx on the surface of the rGO upon annealing at a temperature of 500 °C, leading to the superior capacitive performance of the composite electrode.

Conclusions
In conclusion, we presented a facile synthesis of mixed-valence MnO x /K + -doped rGO nanocomposites for efficient supercapacitor electrodes. By using the electrostatic binding of the manganese precursor and KMnO 4 , and subsequent annealing, we successfully fabricated the MnO x /rGO composites with uniformly distributed MnO x on the surface of K + -doped rGO. The abundant K + ions remaining on the surface of the rGO helps to develop mixed-valence MnO x and leads to a superior capacitive property by promoting a faster reversible reaction. Furthermore, it was found that an annealing temperature of 500 • C was suitable for sufficient growth of mixed-valence MnO x /rGO nanocomposites. The resulting composite materials yield a specific capacitance of 1955.6 F g −1 at 1 A g −1 , which demonstrates that these K + -doped MnO x /rGO nanocomposites would be attractive for various energy applications.  Figure S2: FT-IR spectra of the GO and the MnO x /rGO composites. Figure S3: (a) GCD curves of GM 400 ; (b) specific capacitance variation of GM 400 according to the change of current density; (c) GCD curves of GM 600 ; (d) specific capacitance variation of GM 600 according to the change of current density. Figure S4. Electrochemical impedance spectroscopy (EIS) plots of GM 400 , GM 500, and GM 600 . Figure S5. BET surface area analysis of the GM 500 . Figure S6. K + -free GM 500 prepared from repetitive washing.

Conflicts of Interest:
The authors declare no conflict of interest.