Manganous-Manganic Oxide@Carbon Core-Shell Nanorods for Supercapacitors with High Cycle Retention

Very few research has focused on Mn 3 O 4 @carbon nanorod-structured materials for supercapacitor electrode. A facile process has been developed to prepare Mn 3 O 4 @carbon core-shell hybrid nanorods for supercapacitor electrode materials. The core Mn 3 O 4 polycrystals, which are 50 nm in diameter and 500 nm in length, offer faradaic pseudo-capacitance, while at the same time they serve as supporting template for 4.5 nm shell carbon surface, which functions as electrically conductive material and is also beneﬁcial for improving the capacitive performance. It is highly desirable that the hybrid nanorods exhibit an excellent cycle ability with 95% retention after 5000 cycles at 4 A g − 1 . The hybrid nanorods present a capacitance of 168 F g − 1 and good rate capability (125 F g − 1 at 5 A g − 1 ). The results indicate that the Mn 3 O 4 @carbon core-shell nanorods may have a promising future in applications which need durable, stable, long-lasting power supply. under the terms of the

Sustainable and renewable resources have become more and more important due to climate change and the decreasing availability of fossil fuels. As a result, a continuous and dramatic increase in renewable energy production from wind and sun is observed, as well as the rapid development of vehicles fully powered by electricity with zero carbon dioxide emission. Wind energy and solar power, however, both suffer from instable and inconsistent supply, while a practical car model should be able to run at least a few hours on its own. So largescale, efficient energy storage industry has been flourishing to solve this problem. Batteries and supercapacitors stand at the very front of energy storage industry. [1][2][3] Supercapacitors, also called ultracapacitor, are efficient energy storage units. Those using fast surface redox reaction are called pseudo-capacitors, and those using ion adsorption-desorption are called electrochemical double layer capacitors. They have attracted wide attention around the world over the past decades because of their higher power density, longer cycle life, safer working conditions, higher retention, better environment-friendliness and wider range of working temperatures compared with secondary batteries. And their energy density is much higher than those available in conventional electrical double-layer capacitors. It is undeniable that in order to develop an advanced supercapacitor device, high performance electrode material is indispensable. Active carbon materials, conducting polymers and transition-metal oxides are three fundamental candidates for supercapacitor electrode materials. 1,4,5 Unfortunately, none of them are entirely satisfactory. Active carbon materials have long cycle life but low specific capacitance. 8 Conducting polymer is well-known for its high flexibility but poor cycle ability. 9,10 Transition-metal oxides, such as RuO 2 and MnO x have their unique advantages in their variable oxidation states, good chemical and electrochemical stability, convenience in preparation and high theoretical specific capacitance. However, low porosity, low natural abundance, toxicity and the high cost of RuO 2 have made them unlikely candidates for commercialization of supercapacitors. 11,12 In contrast, manganese oxides are somehow attractive due to their low cost, abundance, high theoretical capacitance (about 1370 F g −1 over a potential window of 1.0 V in theory) and environment friendliness. [14][15][16][17][18] Unfortunately, pure and bulk MnO 2 has a much lower specific capacitance than its theoretical value. According to the charge-discharge mechanism of MnO 2 involving: 7 where M + = Li + , N a + , K + etc. [1] In order to improve the electrochemical performance of MnO 2 , a hybrid of conductive materials and MnO 2 has been widely rez E-mail: wangca@mail.tsinghua.edu.cn searched, especially those bind-free electrode materials. Bind-free electrode materials come from a nanofabrication which applies MnO 2 as a thin film coating on the surface of electrically conductive nanomaterials such as Au, 19 Cu, 20 25 polyaniline, 26 three dimensional graphene. 27 Of those, carbon/MnO 2 hybrid materials have been most intensely investigated, such as carbon@MnO 2 core shell nano-spheres, 4 carbon nano-tube MnO 2 composite 6,13 etc. This structure, carbon material with manganese oxide thin film coating, can obtain high specific capacitance for MnO 2 electrodes, but their relatively complicated fabrication process (As for the carbon@MnO 2 core shell nano-spheres, 4 carbon nanospheres are produced from hydrothermal reaction and in order to synthesize carbon@MnO 2 nanospheres hydrothermal reaction has to be conducted for a second time. The whole process is slightly convoluted.), relatively common specific capacitance, not high cycle retention may have restricted their industry applications in electrical mobiles. Besides MnO 2 , Mn 3 O 4 is also a potentially and promising electrode material for high performance supercapacitors. Nam synthesized Mn 3 O 4 film using electrostatic spray deposition. 28 Cui prepared multiwall carbon nanotubes/Mn 3 O 4 composite by dip-casting method. 30 Lee obtained graphene/Mn 3 O 4 composite in hydrothermal. 31 Although these electrode materials synthesized with Mn 3 O 4 showed a certain degree of electrical performance, there is still large room to improve the electrical performance of Mn 3 O 4 composite electrode materials.
In order to simplify fabrication process, to decrease the impedance of electrode materials and to improve electrochemical performance, also considering the fact that very few researches focus on Mn 3 O 4 @carbon nanorod-structured materials for supercapacitors, we came up with a method in preparing powdery Mn 3 O 4 @carbon core-shell hybrid nanorods electrode materials for supercapacitor with improved electrochemical performances. The new structure Mn 3 O 4 @carbon core-shell hybrid nanorods electrode materials exhibit an excellent cycle ability with 95% retention after 5000 cycles at 4 A g −1 , a good specific capacitance of 168 F g −1 at 0.1 A g −1 , good rate capability (125 F g −1 at 5 A g −1 ), relatively low combination of electrolyte resistance R s (1.5 ohm) and charge-transfer resistance R ct (2 ohm).

Experimental
Synthesis of Mn 3 O 4 @carbon core-shell composites.-All of chemical reagents were analytically pure and used without any further purification. Firstly, 3.16 g KMnO 4 (Sinopharm Chemical Reagent Co., Ltd) were dispersed in 150 mL deionized water by magnetic stirring for 5 min at room temperature until uniform transparent purple aqueous solution was obtained. 1 mL aniline (C 6 H 5 NH 2 ) (Sinopharm Chemical Reagent Co., Ltd) was added to the above solution by magnetic stirring for 20 min. The dark brown precipitates and puce suspension were isolated by centrifugation at a rate of 10000 r min −1 , and then rinsed with deionized water and alcohol respectively. Three deionized water rinsing/centrifugation/alcohol rinsing/centrifugation cycles were carried out till the supernatant became colorless and transparent. The obtained amorphous MnO 2 was dried at 70 • C for 3 h in air. The synthesis of amorphous MnO 2 is based on the redox reaction as below: [2] The amorphous MnO 2 was calcinated at 500 • C for 5 hours to obtain MnO 2 nanorods, seen in Fig. 1b. 0.4 g PEO (Polyethylene Oxide) (A Johnson Matthey Company) was dispersed into 80 mL deionized water. To prevent the PEO from aggregating and precipitating, magnetic stirring for 1 h. 0.8 g MnO 2 nanorods were added into PEO solution with stirring for 3 h. The black precipitates and suspension were isolated by centrifugation at a rate of 10000 r min −1 for 60 min. The MnO 2 @PEO nanorods were dried at 70 • C for 3 h in air, followed by calcinating at 500 • C for 5 h in argon atmosphere and then Mn 3 O 4 @carbon nanorods were obtained, seen in Fig. 1c and Fig. 1d.
Structural characterization.-Transmission electron microscopy (TEM) was carried out on a Tecnai G 2 20 instrument in bright field at 200 kV. Scanning electron microscopy (SEM) was conducted on JSM-7001F (JEOL, Japan). The X-ray diffraction (XRD) data were collected using a Bruker X-ray diffractometer (D8 ADVANCE A25) with Cu K α (λ = 0.154178 nm) radiation. The diffraction patterns were recorded from 10 • to 90 • at a scanning rate of 6 • min −1 . Xray photoelectron spectroscopy (XPS) data were obtained with an ESCALAB 250 Xi electron spectrometer from VG Scientific using 300 W Al K a radiation.
Electrochemical measurements.-The working electrode materials were prepared by mixing the Mn 3 O 4 @carbon, acetylene black and polytetrafluoroethylene (PTFE) in a weight ratio of 80:10:10 with ethanol. Then the electrode slurry was coated on nickel foam round sheet (r = 7 mm) and dried at 70 • C for 3 h. The thickness of the tested electrodes are 0.1 mm and typical mass of the loaded Mn 3 O 4 @carbon electrode materials in each nickel foam sheet is about 5 mg cm −2 .
Electrochemical measurements were carried out in a three-electrode system: A Ni foam coated with Mn 3 O 4 @carbon composites as the working electrode, a platinum foil as the counter electrode and a saturated calomel electrode (SCE) as the reference electrode. The measurements were carried out in a 0.5 mol L −1 Na 2 SO 4 (Sinopharm Chemical Reagent Co., Ltd) aqueous electrolyte at room temperature. Cyclic Voltammetry (CV), galvanostatic charge/discharge and electrochemical impedance spectroscopy (EIS) were measured by a CHI 760E electrochemical workstation. CV tests were done between 0 and 1.0 V (vs. SCE) at different scan rate of 2, 5, 10, 20, 50, 100, 150 mV s −1 . Galvanostatic charge/discharge curves were measured at different current densities of 0.1, 0.2, 0.5, 1, 2, 3 and 5 A g −1 , and the electrochemical impedance spectroscopy measurements were performed by applying an AC voltage with 5 mV amplitude in a frequency range from 0.1 Hz to 100 kHz. The specific capacitances were calculated from galvanostatic charge/discharge curves respectively according to the following equation: 5 Where I is the constant discharge current, tis the discharging time, V is the potential window, and m is the mass of the Mn 3 O 4 @carbon electrode materials. Two-electrode system is also used to evaluate the capacitive behavior of Mn 3 O 4 @C//AC in 0.5 mol L −1 Na 2 SO 4 aqueous electrolyte at room temperature. The method of preparing activated carbon (AC) is from former reports. 36

Results and Discussion
The Mn 3 O 4 @carbon nanorods were synthesized via a series of reactions between KMnO 4 , aniline and PEO. Firstly, amorphous manganese oxide was prepared. Fig. 2a shows a typical TEM image of amorphous manganese oxide. It is obvious in Fig. 2a that tiny crystals distribute extensively in the view of sight and the SAED figure on the left-above shows the material is amorphous.
The MnO 2 nanorods showed in Fig. 2b are about 50 nm in diameter and 500 nm in length. The diameter and length of the nanorod are determined by the calcination time of armorphous manganese oxide (Shown in Fig. S2 (a), (b) and Fig. S3). Because when those armorphous manganese oxides were calcinated at 500 • C, the thermodynamic free energy of Mn 2+ and O 2− ions were raised above their diffusion barrier energy, so Mn 2+ and O 2− atoms transferred to form MnO 2 monocrystal in order to decrease the whole system energy as much as possible. The longer the calcination time, the more unanchored atoms would shift onto the outside surface of crystal nucleus and consequently longer, wider nanorod would form. As we can see, the broadside of the nanorod is polyhedral and each plane of the polyhedral is a lattice plane of MnO 2 . The top/bottom edge of the MnO 2 nanorod is rough, because the edge consists of several crystal planes of MnO 2 . These monocrystal nanorods can serve as ideal templates for further coating. The high resolution TEM (HRTEM) and selected area electron diffraction (SEAD) methods were also used to characterize structure of the MnO 2 grains. Fig. 2c shows a HRTEM image of one MnO 2 nanorod. Visible lattice spacing was measured to be 0.49 nm in one MnO 2 grain, a perfect match of the distance of (200) plane of MnO 2 (JCPDS 44-0141). For the sake of clarity, Fig. 2e demonstrates a simulation image of Inverse Fast Fourier Transform (IFFT) taken from the red circle region in Fig. 2c. Fig. 2d is a SAED of Fig. 2c, which displays a dot diffraction pattern, a typical character of single crystalline MnO 2 with d-spacing values of various planes. Typical SEM images of MnO 2 nanorods are shown in Fig. S2 (c), (d).
The MnO 2 nanorods were dispersed in PEO aqueous solution to form coating on the surface of nanorods. TEM images of MnO 2 /PEO are presented in Figs. 2g, 2h. The as-resulted Mn 3 O 4 @carbon nanorods were fabricated when MnO 2 @PEO nanorods were calcinated at 500 • C for 5 h in argon atmosphere. It is apparent that there is a thin carbon coating on the surface of Mn 3 O 4 nanorod which has been pointed out in Fig. 2f, and this denotes that most precursor may has been changed into Mn 3 O 4 @C. Typical SEM pictures of MnO 2 , MnO 2 nanorods and Mn 3 O 4 /C are presented in Fig. S2 and Fig. S3, which could provide more morphological information.
where a, b and x are three parameters in reaction equation (x = 0). The PEO was converted into amorphous carbon which is attached firmly   Fig. 3e presents the energy-dispersive X-ray spectroscopy (EDS) of the selected area taken from the red rectangular frame in Fig. 3c, where the atomic ratio of C: O: Mn: Cu is calculated to be about 18: 33.6: 29.9: 18.5. Elements Cu was also detected, which comes from Cu grid probably. Fig. 3d is a HRTEM image of the selected area taken from the red circle in Fig. 3c. The carbon atom layers show various kinds of orientation and the thickness of this carbon coating is about 4.5 nm. A cluster of Mn 3 O 4 @carbon nanorods is given in Fig. 3d and it is easy to distinguish a thin, uniform carbon coating surrounding the Mn 3 O 4 nanorods. In Fig. 3c, visible lattice spacing was measured as 0.49 nm on one Mn 3 O 4 grain, well matching the distance of (101) plane of Mn 3 O 4 (JCPDS 24-0734). Meanwhile, the manganese oxidation state is confirmed by XPS. As is shown in Figs. 3f and 3g, in the Mn 2p region, the binding energy of 642.01 eV and 653.51 eV corresponds to Mn2p 1/2 and Mn2p 3/2 , respectively. The splitting width of 11.50 eV is in good agreement with earlier reports on Mn 3 O 4 . 28 In the Mn 3s region in Fig. 3g, the spin-energy separation of 5.04 eV is well in accordance with reported data. 28 XPS spectra of survey scan is presented in Fig. 3h. In Fig. 3i, the weight loss from 25 • C to 100 • C is probably caused by evaporation of moisture. It could be calculated that the carbon content in the Mn 3 O 4 @carbon composite is approximately 11.8% (100 • C to 366 • C) and the weight of the Mn 3 O 4 @carbon composite remains steady which is 85% when temperature is higher than 366 • C.
Cyclic voltammetry (CV), galvanostatic charging-discharging (GCD) and electrochemical impedance spectroscopy (EIS) measurements were tested in a three-electrode system. Fig. 4a shows typical quasi-rectangular and symmetric CV curves at low scan rate and CV curves are still symmetric at high scan rate (150 mV s −1 ), indicating that the Mn 3 O 4 @carbon nanorods electrode material has a good electrical double-layer capacitive behaviour. Cations could penetrate carbon coating easily during charging and discharging process and redox reaction can take place beneath the porous and amorphous carbon coating in Mn 3 O 4 active materials. Curves in Fig. 4b demonstrate fairly linear slopes, electrochemical reversibility and capacitive behavior. In Figure 4b (especially at 0.1 A g −1 ) and Figure 4d (inset graph), the IR drops are clearly observable. The voltage (IR) drop becomes larger with the current density going up indicating that the conductivity of Mn 3 O 4 @carbon nanorod material needs to be further improved to subdue the drop of IR. In Fig. 4c, it is clear that the specific capacitance of Mn 3 O 4 @carbon is at least twice as high as that of MnO 2 nanorod at the same current density and the rate capability of Mn 3 O 4 @carbon is also better than that of pure MnO 2 nanorod (The capacitive performance of pure MnO 2 nanorod is presented in Fig. S1). It reveals that a simple adjustment of MnO 2 nanorod into Mn 3 O 4 @carbon nanorods will increase its original specific capacitance by more than 100%, indicating a fairly effective and dramatic improvement. The Mn 3 O 4 @carbon nanorods display a specific capacitance of 168 F g −1 at 0.1 A g −1 , (The specific capacitance of Mn 3 O 4 @carbon at 0.1 A g −1 , 0.2 A g −1 , 0.5 A g −1 , 1 A g −1 , 2 A g −1 , 3 A g −1 and 5 A g −1 are 168 F g −1 , 160 F g −1 , 151 F g −1 , 147.5 F g −1 , 140 F g −1 , 135.5 F g −1 and 125 F g −1 respectively. The coulombic efficiency of Mn 3 O 4 @carbon at 0.1 A g −1 , 0.2 A g −1 , 0.5 A g −1 , 1 A g −1 , 2 A g −1 , 3 A g −1 and 5 A g −1 are 71%, 81%, 91%, 93%, 95.3%, 96.2% and 97% respectively.) which is slightly higher than the results on the basis of Mn 3 O 4 or MnO 2 material reported by other groups, such as Mn 3 O 4 film, 28 36 and whisker-like MnO 2 arrays on carbon fibers paper. 37 The Mn 3 O 4 @carbon nanorods also show a good rate ca-pability. At a current density 5 A g −1 , it remains 125 F g −1 . A good rate capability may derive from the carbon surface, which could improve the conductivity and guarantee high transfer ability of ions in bulk material. So even at relatively high current density of 5 A g −1 , it could still remain 75% of the specific capacitance at 0.1 A g −1 . Two-electrode test system is also conducted to evaluate the capacitive behaviour of Mn 3 O 4 @C. Activated carbon is employed as negative electrode material. The test is performed in a two-electrode cell in 0.5 mol/L Na 2 SO 4 aqueous electrolyte solution. CV curves and GCD plots are presented in Fig. S4. From Fig. S4 (b), the specific capacitance of the two electrode system is 50 F g −1 , 47.5 F g −1 , 46 F g −1 , 44 F g −1 , 40.5 F g −1 , 36 F g −1 and 33 F g −1 at 0.
Generally, energy storage mechanism of Mn 3 O 4 @carbon supercapacitors is derived from the ionic charge accumulation in the electric double layer which exists at the electrolyte/electrode interface. Manganese in Mn 3 O 4 is present in two oxidation states: +2 and +3, and the formula Mn 3 O 4 is sometimes written as MnO · Mn 2 O 3 . In the neutral aqueous Na 2 SO 4 electrolytes, the charge/discharge mechanism may be described as the following reaction: where M represents hydrated protons (H 3 O + ) and/or alkali Na + cation. The outmost electron distribution of Mn in [Mn 2 O 3 ] 2− is 3d 3 4s 2 . The reaction equation implies two kinds of possible processes, i.e., an adsorption/desorption process of cations at the material surface and/or an insertion/extraction process of cations into the bulk material. Recently, a research reports that the insertion/extraction process happens mostly on well crystallized bulk materials while the adsorption/desorption process occurs on weakly crystallized materials. 38 Based on this assumption, there being no obvious redox peaks in CV curves implies that the charge storage process in Mn 3 O 4 @carbon nanorods electrode system may be dominated by adsorption/desorption of cations at the Mn 3 O 4 surface with abundant crystalline defects and amorphous interspaces.
The cycling stability of the Mn 3 O 4 @carbon nanorods was also probed by galvanostatic charge/discharge at a high density of 4 A g −1 , as is shown in Fig. 4d. It is obvious that the capacitance retention keeps going up (slightly above 100%) in the first 250 cycles, which may be caused by the gradual activation process of the Mn 3 O 4 aggregates. From 250 cycles to 2700 cycles, the capacitance retention drops from 101.5% to 97%. After that, between 2700 cycles to 3000 cycles, retention begins to go up again to 97.5% and this may be a second time activation process. During the first time activation process, not all of the aggregated Mn 3 O 4 are contacted with Na 2 SO 4 aqueous electrolyte thoroughly; however, with the increasing number of cycling, those Mn 3 O 4 which are firmly attached may be totally activated. When the activation process is completed, the retention performs a steady state from then on. With the whole system getting into a stable state after 3500 cycles, 95% of initial specific capacitance can be finally retained and the shape of charge/discharge curves for the last thirteen cycles still remain nearly symmetric, which indicates the excellent electrochemical stability of the Mn 3 O 4 @carbon nanorods electrode materials.
In addition, electrochemical impedance spectroscopy (EIS) was employed to detect the properties of charging and ion transfer in the Mn 3 O 4 @carbon and MnO 2 nanorods electrode. Fig. 4e shows the Nyquist plots for MnO 2 nanorods electrode. Fig. 4f is the Nyquist plots for Mn 3 O 4 @carbon nanorods electrode, the inset is well-fitted equivalent circuit, in which C dl is double layer capacitance, R s is bulk solution resistance, R ct is the Faradic charge-transfer resistance and Z w is the Warburg impedance. The sectional semicircle reveals the combination of electrolyte resistance R s and charge-transfer resistance R ct . The slope of the line indicates the Warburg impedance which represents the diffusive behaviors of the electrolyte in electrode pores and ions in active materials. It is clear that the Mn 3 O 4 @carbon has a slightly smaller semicircle diameter than the MnO 2 electrodes in high-frequency region, implying smaller resistance from the electrochemical system and charge transfer. From Fig. 4f, it could be estimated that the combination of electrolyte resistance R s and charge-transfer resistance R ct of Mn 3 O 4 @carbon are 1.5 ohm and 2 ohm respectively. Meanwhile, the difference among the slopes of the lines in low-frequency region indicates that the diffusive resistance of the Mn 3 O 4 @carbon is much smaller than that of MnO 2 . The reason for smaller resistance of Mn 3 O 4 @carbon than MnO 2 may be caused by the outside amorphous carbon coating. When the electrochemical tests are performed in three-electrode testing system, 0.5 mol L −1 Na 2 SO 4 aqueous electrolyte at room temperature is used for all tests. The mobility of ions in electrolyte would maintain at a same level for all electrochemical tests. The difference lies in the transfer ability of ions in bulk material. Ions could transfer easier in Mn 3 O 4 @carbon than MnO 2 nanorod, because the thin carbon film is conductive and the poly-crystal Mn 3 O 4 structure would also facilitate the transportation of ions in bulk material. The Mn 3 O 4 @carbon nanorods in contact with each other in amorphous carbon coating makes charges easier to transfer, and consequently, the charge transfer resistance can become smaller than that of MnO 2 . A smaller diffusive resistance of Mn 3 O 4 @carbon than MnO 2 could be induced by Mn 3 O 4 polycrystals, because these small grains make the whole system more porous and provide more defects, Fig. 3c can illustrate the polycrystalline structure. As for MnO 2 , seen in Fig. 2c, the degree of crystallinity is high and this leads to a condense bulk with less defects than Mn 3 O 4 @carbon which may increase the diffusive resistance of MnO 2 .

Conclusions
In summary, a facile and low-cost preparation of Mn 3 O 4 @carbon core-shell hybrid nanorods for supercapacitor electrode materials has been developed. The average thickness of amorphous carbon coating is 4.5 nm. The diameter and length of inner Mn 3 O 4 nanorod are about 50 nm and 500 nm respectively. Electrochemical measurements show that the Mn 3 O 4 @carbon nanorods exhibit a good electrochemical performance: a high specific capacitance of 168 F g −1 at the current density of 0.1 A g −1 , a good rate capability with 125 F g −1 at 5 A g −1 and an excellent cycle ability with 95% retention after 5000 cycles at the large current density of 4 A g −1 . Considering the efficiency, low cost, good controlability of the synthesis process and superior electrochemical performance, the Mn 3 O 4 @carbon core-shell nanorods have a promising future in applications where durable, stable and long-lasting power supply is needed. Due to the large size of Mn 3 O 4 nanorod, the specific capcitance of Mn 3 O 4 @carbon nanorod is not outstanding. We still believe that the performance of electrode materials based on Mn 3 O 4 still can be further improved, so future researches shall focus on fabricating a small-sized Mn 3 O 4 nanorod electrode materials.